TBE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING A STUDY OF OXIDATION-REDUCTION POTENTIALS APPLIED TO -SEWAGE SLUDGE DIGESTION Alan Hao Molof December, 1960 mP485

Doctoral Committee: Associate Professor Jack A. Borchardt, Chairman Professor Earnest Boyce Professor Lloyd L. Kempe Professor Clarence J. yelz Professor Brymer Williams ii

ACKNOWLEDGEMENTS The author wishes to express his appreciation to Professor J. A. Borchardt for his help in the preparation of this thesis and for the invaluable assistance and encouragement he gave during the course of the investigation. To Professor G. B. Williams whose helpful advice was a great aid during the author's graduate studies. To Mr, G, Geissendorfer of the Civil Engineering Laboratory for his suggestions and help during the construction of the experimental apparatus. iii

TABLE OF CONTENTS Page ACKNOWIEDGMENNT o n. *.,o....................... ii LIST OF TABLES.. vi LIST OF FIGURES...............00e. * * 9. *. vii LIST OF APPENDICES e * e f * * a a i o S 6 6 a e * * 4 e n * iX INTRODUCTION............. *... * o. o *. 1 SLUDGE DIGESTION. o. ~. 0 a. s..a.~ a 0 a 0 8 ~ a w~*. o o,...oe. 3 OXIDATION,.-REDUCTION POTENTIAL9S a * e e * e o 8 10 METHODS OF MEASURING THE POTENTIAL.., o.........o....o..o.. 16 FACTORS TO BE CONSIDERED IN THE USE OF METAL ELECTRODES 19 Types of Electrodes. *.. o... o. o.... o......... o 19 Cleaning the -Electrode;,...........*............ *.. o...,.* 20 Effect of Gases on the Electrode,.,,.,.......,........00 24 OXIDATION-REDUCTION POTENTIALS IN BACTERIOLOGY AND BIOCHEMISTRY.... 25 ELECTRODES USED FOR THE QUINHYDRONE SYSTEM 4.............o.o. 26 OXIDATION-REIDUCTION POTENTIALS IN SEWAGE TREATMENT........,...... 49 EXPERIMENTAL APPARATUS...................................... 54 Sludge Digestion Apparatus. a........ o..O... a.......... 54 Oxidation-Reduction Potential Apparatus........,................ 59 Lactose Broth Feed Apparatus *'..09....................... 67 EXPERIMENTAL PROCEDURE a............... 69 Lactose Broth Feed.....................o..............o.... 69 Sludge Digestion 9.................9........*.....99...* 70 Electrode Potential Measurements................... o.o.... 70 BEHAVIOR OF ELECTRODES IN QUINHYDRONE......o........ **......' 73 ADJUSTMENT OF THE METAL EIE3CTRODE TO THE SLUDGE POTENTIAL. 78

TABlE OF CONTENTS (cont'd) Page LACTOSE BROTH FEEDING OF A DIGESTED SLUDGE................... 0... 99 INITIAL POTENTIAL STUDIES ON SLUDGE DIGESTION o o oo..............o 113 Raw Sludge Feedings.....................1................116 Electrode Check...............................o.............. 121 BATCH- TYPE FEEDING.......... o o o.....o o...............o.o 126 ACTIVATION OF A DORMANT SLUDGE.. o...................o o.. o...o 130 Parallel Feeding...........................o.....O.O 131 Non-Parallel Feeding........ o... o.o 134 Overall Results of the Feeding..................e.......v o. 137 LOADING TO A TEMPORARY DIGESTION FAILURE.................... 144 TIE EFFECT OF AGITATION ON THE ELECTRODE POTENTIAL o........o 161 DISCUSSION..................O.O..OOOOOOOOOO... O..O. 168 SUMMARY AND CONCLUSIONS......... 193 BIBLIOGRAPHY o............... o a.. s o o. O o O O 214

LIST OF TABLES Table Page I OXIDATION-REDUCTION POTENTIALS OF SOME BACTERIAL CULTURES AND BIOCHEMICAL REACTIONS................ a... 28 II COMPARISON OF GOLD AND PLATINUM ELECTRODES IN QUINHYDRONE SOLUTIONS WITH VARYING SOLUTE CONCENTRATIONS..... 4**.. 75 III SUMMARY OF ELECTRODE POTENTIALS AND GAS PRODUCTION DURING THE PERIOD OF CONTINUOUS AGITATION............... 16 IV SUMMARY OF ELECTRODE POTENTIALS AND GAS PRODUCTION DURING THE FEEDING PERIOD STARTING WITH A DORMANT SLUDGE. 132 V SUMMARY OF THE POTENTIAL CHANGES FOLLOWING THE MAXIMUM POSITIVE POTENTIAL...*........................... 136 VI SUMMARY OF THE ELECTRODE POTENTIALS DURING THE FEEDING PERIOD BEFORE DIGESTION FAILURE.............. 146 VII GAS RATE CHANGES AND THE CORRESPONDING ELECTRODE POTENTIALS.................o. o o o 150 'III POTENTIAL-TIME RELATIONSHIP DURING THE LINEAR POTENTIAL DROP a a o m. a.0. aa. e o.a o ob a a o * 0 0 0 0 Q a 155 IX AVERAGE CARBON DIOXIDE CONCENTRATIONS FOR VARIOUS TIME INTERVALS..........o... o o. o.o * o o o * 156 vi

LIST OF FIGURES Figure Page 1 Relation Between the Oxidation-Reduction Potential and the Percentage Oxidation of the System............. 13 2 Circuit for the Measurement of the Electrode Potential... 18 3 Schematic Drawing of a Sludge Digestion Unit Containing the Oxidation-Reduction Electrodes o....o......... o 56 4 View of the Two Level Incubator from the Front.........57 5 View of Apparatus from the Rear......................... 57 6 View of the Apparatus from the Top.....o........ 60 7 View of Gold and Platinum Electrodes o.ooo..o......... 64 8 Comparison of Platinum and Gold Electrodes in Varying Concentrations of Quinhydrone Solution................... 77 9 Adjustment of a Platinum Electrode in a Digested Sludge o. 94 10 Comparison of the Potential Changes of Platinum and Gold Electrodes Due to Stirring a Digested Sludge Exposed to Air 1o o E f o f eo ~ ~ ~ Ot t o o oo 'o o th e A dj n 0 a P l a t i n u 0 a a * 0 95 11 Effect of Agitation on the Adjustment of a Platinum Electrode in Digested Sludge o........e o..... o 96 12 Comparison of the Adjustment of Gold and Platinum Electrodes in Digested Sludge After Being Conditioned in the Sludge and Exposed to Air e o.. ~............. 97 13 Adjustment of a Gold Electrode in a Digested Sludge with a High Volatile Solids Content ao.o..... 98 14 Electrode Potentials and Gas Production After the Addition of Ten Grams of Lactose Broth Concentrate to a Digested Sludge Media...........o...... 109 15 Electrode Potentials and Gas Production After the Addition of Four Grams of Lactose Broth Concentrate to a Digested Sludge Media............................. 1. l10 vii

LIST OF FIGURES (Continued) Figure Page 16 Electrode Potentials and Gas Production Resulting From a Short Lag Period Digestion After the Addition of Four Grams of Lactose Broth Concentrate to a Digested Sludge Media....oll............... 111 17 Electrode Potentials Over a Long Time Period and the Gas Production Resulting from a Short Lag Period Digestion.After the Addition of Four Grams of Lactose Broth Concentrate to a Digested Sludge Media.......... 112 18 Electrode Potentials and Gas Production Resulting from an Increase in the Temperature of the Sludge from 25 to 355C....o..... o o o. o.. o 0..... e 123 19 Electrode Potentials and Gas Production Resulting from Digester Feeding with Continuous Agitation... 124 20 Electrode Potentials and Gas Production Resulting from Digester Feeding After the Continuous Agitation was Stopped..0...9..............O.O*.OOOaO. 125 21 Electrode Potentials and Gas Production Resulting from a Batch-Type Feed................................ 129 22 Electrode Potentials and Gas Production Resulting from Feeding a Dormant Digested Sludge............. 141 23 Electrode Potentials and Gas Production Resulting from Different Feeding Levels.................o... 142 24 Electrode Potentials and Rates of Gas Production Resulting from Activating a Dormant Sludge by Feeding o 143 25 Electrode Potentials and Rates of Gas Production Resulting from Overloading a Digester 0.... 159 26 Total Gas Production Resulting from Overloading,a Digester o. a0a4a aaaaOo a.O eO 160 27 Effect of Agitation on the Potential of Gold Electrodes 166 28 The Initial Effect of Agitation on the Potential of Gold Electrodes...............O..O.. o.....o 167 viii

LIST OF APPENDICES Appendix Page A. DERIVATION OF OXIDATION-REDUCTION EQUATION................ 196 B. CLEANING METHOD -EFFECT ON THE METAL ELECTRODE............ 199 C. GAS-METAL ELECTRODE EFFECT....................... 209 ix

INTRODUCTION The cost of sludge digesters under present day construction prices comprises a substantial portion of the total first cost of sewage treatment plants. To reduce this cost, efforts have been made to diminish the required tank volume or digestion time by improving present sludge digestion practices. More efficient use of digester capacity has been accomplished by such techniques as higher feeding levels and lower detention times in conjunction with adequate agitation of the digester contents and prethickening of the raw sludge. The increased trend toward maximum utilization of digester capacity has resulted in a need for more adequate knowledge of the process and its control. This is due to the increased sensitivity of the digesting sludge to adverse conditions. Like many process developments, the understanding of its mechanism and control has lagged behind its use. Knowledge of the factors affecting the process will not only aid in the understanding of its mechanism and control, but will also stimulate further process improvements. Sludge digestion has been extensively studied by the use of such parameters as pH, quantity and quality of the gas production, volatile acids, volatile solids and methane-forming bacteria. These parameters help define the necessary conditions for successful sludge digestion as well as aid in the evaluation of the bacteriological, chemical, and physical processes taking place during the digestion. However, because of the complex biochemical and biological systems present, the entire mechanism of the digestion has not been completely established. 1 -

-2 -The digestion of sewage sludge is a process involving the anaerobic degradations of complex organic materials. A succession of bacterial systems produce enzymes which enable the degradation reactions to proceed. The many steps involved in the breakdown of these complex compounds to simpler materials include various oxidation-reduction reactions. Since an oxidation-reduction reaction involves an electron transfer, it is possible to make quantitative studies of these reactions by measuring the resulting potential change. This technique has been used so successfully in pure bacteriological and biochemical systems that it would appear possible to apply the same techniques to the mixed cultures of sludge digestion with some degree of success. If such would be the case, a better understanding permitting better control of the process might result. The measurement of oxidation-reduction potentials in digesting sewage sludge has never been investigated on a continuous basis. In a few instances data has been published giving a few potential values, but the information was so meager that its value was limited. The objectives of this thesis are to present and evaluate methods of measuring oxidation-reduction potentials in anaerobic sludge systems, to study the potential changes in a digesting sludge, and to contribute to the further understanding of oxidation-reduction potentials and their application to sludge digestion. The research included in this dissertation was begun in September 1953 and concluded in June 1957.

SLUDGE DIGESTION Sludge digestion is the biochemical process by which complex organic materials in sewage sludge are broken down into simpler substances. This organic matter is gasified, liquified, mineralized or converted into more stable organic products. The result is a residue which can be economically disposed of by drying or burning without creating a nuisance. A valuable by-product of the digestion is the production of methane gas. This is used mainly as a fuel, but also has been employed as a raw material for chemical synthesis. The biological, chemical and physical changes occurring in sludge digestion have been the subject of many investigations. A complete review of all this work would be beyond the scope of this thesis. There are, however, many descriptions of sludge digestion available in books and review articles concerning sewage treatment (1)(2)(3)(). The anniversary book of the Federation of Sewage Works Associations(5) is an outstanding reference with contributions by several authors. Because of these comprehensive references, this presentation of the sludge digestion process will be reduced to a brief summary. The sludge digestion process has been studied on both a batch and a continuous basis. In the former process the digestion of sewage sludge is divided into three well-defined stages: acid production, acid regression, and alkaline fermentation. At a temperature of 60~F, the acid production phase lasts about two weeks. The bacteria attack easily

available substances such as carbohydrates and soluble nitrogenous compounds releasing small quantities of gas, organic acids and acid carbonates. The gas is composed chiefly of carbon dioxide, with some methane and hydrogen sulfide present. The pH is around six or lower. Acid regression. lasts about three monthso In this phase, the organic acids and nitrogenous compounds are attacked, thereby producing ammonia compounds, acid carbonates and gaso The small quantity of gas produced is composed chiefly of carbon dioxide and methane, along with some hydrogen. The pH gradually rises to around 6.8 and the sludge odor is very offensive. With the rise of the pH, alkaline fermentation sets ino This phase lasts about one month, during which the organisms attack the more resistant materials such as proteins and organic acids. The pH continues to rise, reaching pH 7~0 or above, and the decrease in organic acids which results makes a more favorable enviroment for proteolysiso The resulting mixture of reactants is a welbuffered system in which acids are neutralized by the previously formed products of alkaline decomposition. Large volumes of gas are released. consisting of methane, along with some carbon dioxide. When the bulk of the digestible material is exhausted,, the sludge becomes stable and a gradual cessatlion of bacterial activity sets in. During the phases outlined above, there exists a succession of bacterial agents which are influenced both by intermediate and final chemical products and also are greatly affected by the changing environrrmental conditions.

-5 -Additions of small.quantities of fresh solids to the sludge in the alkaline fermentation period result in a continuous type of digestion where the acid production and regression periods are not distinct and the gas production is high. This is due to the well buffered, highly developed bacterial culture developed within the sludge. At 600F the time for comparable digestion is reduced to onetenth of that of the batch process. On a continuous basis sludge digestion can be divided into two phases. The first is the breakdown of complex materials by hydrolytic enzymes. This includes protein breakdown to amino acids, carbohydrate breakdown to simple sugars and lipid breakdown to glycerol and fatty acids. The amino acids are in turn converted to organic acids, aldohols and amines, while the sugars are converted to organic acids. The formation of these organic acids is considered a primary end product. The second phase involves the breakdown of lower fatty acids, alcohols and other molecular fragments with the production of methane and carbon dioxide. This is completed by a group designated as methane-producing bacteria. The gas produced from a continuously digesiting sludge can include methane, carbon dioxide, hydrogen, hydrogen sulfide, and nitrogen. The analysis of this gas under normal digestion would be about 70 to 75 percent methane, 23 to 28 percent carbon dioxide and 1.5 per". cent nitrogen. Hydrogen and hydrogen sulfide comprise the remaining small portion of the total.

46. The amount of gas produced by digesting sludge was reported throughout the literature and also summarized in many texts on the sub(3) ject, i.e., Imhoff and Fair. More detailed studies have been concerned with the composition of the gas (6)(7)(8), the gas yield from fat, protein and carbohydrate(6), the time required to produce a certain percentage of the total yield(9)(10)(11)(12), and gas production rates(l3). The production of hydrogen was of particular concern to this research since the effect of this gas on the platinum electrode could be appreciable(l4)(15)(16)(17). Reports of sludge gas analysis did not always include hydrogen, although most authors reported that hydrogen was usually present. Buswell and Neave (18) found an average of approvxi (19) mately two percent hydrogen, Sawyer and Roy reported two to three (20) percent in high rate laboratory tests, and Steel reported from one to three percent. Imhoff and Fair(3) have stated that hydrogen was produced in the acid regression period of a batch digestion of raw sludge. There were other authors however, who did not feel that hydrogen was a normal component of sludge gas or, if present, would be so only in very small amounts, Hawkes and Jenkins(21) found that hydrogen was sometimes (22) produced in immature digesting sludge and Rudolfs stated that the gas production included hydrogen only under certain conditions. Fulweiler(23) reported that there was littleg if any, hydrogen in the sludge gas and Phelps(24) stated that hydrogen did not appear in any quantity.

67 -The question of whether or not hydrogen was present in sludge gas was discussed by Buswell and Neave (18) and Hawkes and Jenkins(21) Possible sources of this gas were presented by Buswell et al.(25) From the above information it appeared that hydrogen was usually, but not always, present in sludge gas. Large ratios of raw to digested sludge appeared to favor hydrogen productiot. Conditions more favorable to the production of hydrogen have been created by the recent trend toward higher feeding levels. This condition has been created by the attempt of engineers to utilize more of the available capacity of the sludge digesters. On a plant scale, gas recirculation was used by Morgan(26) to agitate the sludge; pre-thickening (27) in conjunction with continuous Seeding was tried by Torpey In a pilot digester Torpey used continuous mechanical agitation, pre-thickening and feeding every two hours. Laboratory scale experiments have been performed by Sawyer and Roy utilizing continuous gas agitation and two feedings per day~ This experimental work helped to define certain of the- factors involved in successful high rate feeding. These were: 1) concentrating the sludge feed, 2) continuous feeding of the sludge, and 3) continuous agitation to get effective utilization of the tank volume. Higher feeding levels created a need for better process control than was necessary with lower loadings since the digestion was more

susceptible to overloading, and hence failure. Because of their sensitive environmental. pattern, the metb.ane-producing bacteria might determine the point in th.e digestion where fa~ilure was m.ost li.kely to occur. These (28)(29) bacteria have been. extensivrely investigated 'by Stadtman and Barker (30)(3 )e Heukelekian and H1einemnan2)(332 )(5)34(35) and. Buswell et a.16)(3 7) (38)(39)(40)~ The work or:Heukelekian and iHeneman. was concerned with sewage sludge, while the ot.her in,,restigators concentrated. on the properties of' the bacteria per seo The effect of volatile acids on these bacteria was evidenced from. the fact that an ac..d accumul.ation could retard or completely stop the action of these methan.eproduac.:..ng bacteria. The temperature effect to Bu.swe 11 ( 37) The effect o:f olatile aci.ds on. the digestion9 particularly the gas producing phase9 has resulted in mary studies of these aci.ds(37)(25) (153)(413) L2)(4) 443(45) (46)0.It was ',r tlat vol.atile acids were produced by th.e fi rst o.r- hyd.roly'tic pLhase of t-he digestion9 an.d destroyed by the second or met+hane -prod'c.,.i.ng bacte;niai. p.h.,- ase In. regard to th.is, Kaplovsky '43) bhas sta-ted. that a looering of T.e acid c:on.ten-t was due -to a greater rate of des'truct. ion. not a.erate o produ.ction. This in.dicatsed thaty the acid balance was cont rol-ed by r'he?7iwthane~-'bacterial. phase and not by- the hydrolytic.e phase0

-9 -Control of the sludge digestion process has usually been based on the factors affecting the methane bacteria since this bacterial system was the most sensitive link in the chain of reactions. These factors included volatile acids, pH and temperature. The reduction of volatile matter, alkalinity, and composition and quantity of the gas have also been used as parameters since they reflect the condition of the bacterial environment. Although it was found in one plant that a rise in the carbon dioxide content preceded by at least one week a reduction in the gas (47) rate,7) the volatile acid content has been considered by most authorities to bh the most sensitive indicator of proper digestion(25)(26)(47), Apparently it was not the complete answer, since Kaplovsky(43) stated that the volatile acid accumulation was only a secondary factor because without the proper bacterial flora, trouble would be encountered even at low acid levels. The length of time necessary to determine the volatile acid content of a sludge has been a disadvantage for process control. The approximate analytical determination of volatile acids required over an hour, while the more precise tests lasted several hours. If oxidationreduction potentials would indicate the condition of the digesting sludge, it would make available a rapid test for control purposes and thus provide a new tool in sludge digestion operation.

OXIDATION- REDUCTION POTENTIALS The digestion of sewage sludge inv.o'lves oxidation-reduction reactions which play an important role in the degradation of complex biochemical materials. There is a chain of steps from the initial reactant to the final product, many of which are imperfectly understood. Included in this chain are oxidation-reduction reactions, some reversible and some irreversible. The reversible reactions enable the living cell to store energy for future utilization. Hewitt(48) states that oxidation-reduction systems are so essential that life might be defined as a continuous oxidation-reduction reaction. These reactions can be studied by the use of electrode potential measurements. However, their quantitative interpretation depends upon the reaction being reversible. Michaelis(16) states that the use of potential measurements -has provided a more clear and thermodynamically correct approach to cell energetics and the metabolism of the living cello The potential measurement is termed electrode potential, oxidation-reduction potential, ORPE, or redox potential by different authors and these terms are...generally used synonymously. A complete review of the basic principles of oxidation-reduction potentials would be beyond the scope of this thesis. Several sources of this information are available, of which Michaelis(6) Clark et al., (49) and Hewitt(48) are outstanding. Because of these references, the presentation of the basic principles of oxidation-reduction potentials is reduced to a brief summary. 10 -

Oxidation signifies more than simply the gain of oxygen by an element or compound such as the formation of carbon dioxide from carbon or acetic acid from acetaldehyde. C + o2 = C02 HCHO + 0 = HCOOH It also denotes the loss of hydrogen or dehydrogenation and the loss of an electron. These comprise the three methods of oxidation. Dehydrogenoration of ethane to ethylene and a gain of an electron by ferric from ferrous iron are examples of the latter two methods. CH-CH =H — C.-=H2 + E2 Fe++ = Fe+++ + e The modern definititon.6f dehydrogenation by Michaelis (16) involves the loss of an electron with the simultaneous loss of a proton or hydrogen ion. The proton is not involved in the oxidation process and the proton attachment is termed the level of acidic ionization. Referring to the chain of respiratory processes, he states that the process is not entirely a transfer of hydrogen to another molecule, but is a transfer of electrons with or without protons. It is important to note that for every oxidation, there must be a corresponding reduction process~ The oxidation of ferrous chloride to ferric chloride illustrates this: FeC12 + C1 - FeC13 or in ionic form: Fe++ 2C1- + Cl~ = Fe+++ 3C1'

From this can be seen the oxidation and corresponding reduction. Oxidation: Fe++ = Fe+++ + e Reduction: C1~ + e - C1 -Various methods are presented by different authors which relate the transfer of electrons in an oxidation-reduction reaction to the potential (49) by the use of a quantitative equation. These include Clark et al. Michaelis (16) Glasst (50) and Hewitt(48) All of these methods lead to the same equation for which the derivation is illustrated in Appendix A. Eh = Eo + R in (O n11 IF- (Red) This is the basic equation udder reversible conditions which quantitatively relates the oxidative state of the system with the oxidationreduction potential. Eh is the potential of the system referred to the normal hydrogen electrode and Eo is a specific constant for the system being measured. They are measured in volts. R is the universal gas constant,. T is the absolute temperature, F is the Faraday constant and n is the number of electrons transferred in the oxidation-reduction reaction. If the ratio of oxidant to reductant gets larger, the potential becomes more positive and if the ratio of oxidant to reductant gets smaller, a more negative potential results. A high positive is shown by chlorine, an excellent oxidizing agent. Conversely, hydrogen as a good reducing agent will exhibit a high negative potential. We can then speak of an aerobic solution in the positive potentialrarige and an anaer-obic solution:in -t.be -negativs potential range.

It can be seen that the potential is dependent only on relative amounts and not absolute quantities. Thus, oxidation-reduction potentials determine the level of intensity, not capacity of the oxidizing system. Intensity resembles the pH and temperature of a solution in that nothing is indicated about the buffering ability or heat content, respectively. The ability to resist change in oxidation-reduction systems is called the "poise" of the system.. As temperature alone does not allow prediction of the total cooling or heating ability of the solution, neither does the electrode potential yield any information of the total oxidizing or reducing ability of the solutionn This factor is especially important in biological systems where many have well-defined potentials but are poorly poised. A typical oxidation-reduction curve as shown in Figure 1 illustrates the poising effect. The maximum stability or poise of a solution 0 o ( ) C100 Percent Oxidized Figure 1. Relation Between the Oxidation-Reduction Potential and the Percentage Oxidation of the System will be found at the point where the curve is flattest and most resistent to change. By mathematical means Michaelis(l6) shows this point to be located where there are equal concentrations of oxidant and reductant. Thus a system with small quantities of either the oxidized or reduced state is poorly poised and its potential is very susceptible to other redox

-14 -systems which may have lower concentrations but are in a more poised state. The effect of concentration on the.capacity of the solution is based on the law of mass action. If the solution is poorly poised, based on the above definitions, it still could resist changes of potential if the concentration is sufficiently high. The effect of pH on the potential is due to the altering of ionic equilibria. Because of the fact that acid or basic groups are created or destroyed in redox reactions, it follows that oxidation-reduction is an interdependent factor in the acid-base equilibria. The dissociation constants can be affected and this plus the pH may have a considerable influence according to Clark et al. The formulation of a general rule for the pH effect was very complicated in many cases. Hewitt(48) felt that for biological systems it was impossible to formulate a general rule since the pH effect varied in different ranges of pH. With systems that were not exhaustively studied, the Eh as well as the pH should be noted. Hewitt stressed that even though the effect of pH on Eh was complex, it did not imply that the effect was necessarily great. A usual E4h to pH ratio of 0.03 for biological systems at a pH near seven meant a difference of three my for 0.1 pH unit change. The derivation of the electrode equation presented previously was based on a thermodynamically reversible reaction. The equation, therefore, could not theoretically be applied to an irreversible reation and was independent of any time considerations. However, there have been attempts to relate the potential to irreversible reactions and kinetics.

15 -A convenient, but empirical, method of characterizing irreversible systems by potential measurements was developed by Conant and Lutz(51). The measured potential was referred to as 'the apparent reduction potential," which was defined as "the oxidation-reduction potential of a particular reducing agent which will just cause an appreciable amount of reduction of the substance in question." In this way it was possible to treat irreversible organic reductions by the use of potential measurements and place them on a scale based on the standard hydrogen electrode. Gershinowitz(52) has pointed out that Conant and co-workers have shown that there was a definite relationship be tween rates of an irreversible reaction and OR potential, and as a result they have derived an equation for this condition which gave satisfactory experimental agreement. He concluded from this and other work that there was some connection between kinetic rate constants and thermodynamic quantities; and he derived theoretical equations relating these. This work was based on the assumption of the formation of an activated complex and a reversible first step. In regard to this assumption it was of interest to note that LuValleand.Goddard(53) in discussing the mechanisms and kinetics of enzymatic oxidation-reduction, assumed that a complex was formed between the enzyme, electron donor and the electron acceptor; and that the reactions proceeded through a series of univalent electron exchanges. Longsworth and MacInnes(54) found a connection between 'H and the rate of acid production in a growing culture. They used the term apparent ORP" and said the j- measured was not an equilibrium potential but the result of several kinetic processes~ Kinetic properties of potential measurements have also been observed by others(5)(6)

METHODS OF MEASURING THE POTENTIAL There are two methods of obtaining the oxidation-reduction potential inert metal electrodes and oxidation-reduction dyes. Much of the early work was done with dyes due to the siccess and simplicity of the use of pH colorimetric indicators. Clark et al.(9) made an extensive study of dye systems but no simple reliable method was found which compared to pH measurements. The methods of obtaining pH andoxidation-reduction potentials are identical except that one uses a dye which reacts to pH in the first case, while in the other a dye is used which reacts to the oxidative state of the system. The majority of dyes used are indophenols or derivatives of indigo and their color is due to the quinoid structure of the benzene ring. Methylene blue used in the "relative stability" of sewpge is an example of this technique. But the disadvantages for biological systems are many. Dyes can poise the solution by oxidizing a system of low capacity without being reduced appreciably. The time lag may be very great and the system may be changing. There can be catalytic as well as toxic effects on the system. A great disadvantage lies in the small range the dye can span and when potential changes are great, more than one dye system would be needed. Metal electrodes provide the most reliable method of obtaining oxidation-reduction pbtentials... The measurement reflects the difference in potential between an inert metal electrode, i eo., platinum or gold, and a reference electrodes such as a calomel cello Since the metal electrode is only one-half a cell, the reference electrode is needed to complete the 16~

17. measuring circuit. Potential measurements are identical to pH electrometric determinations except for the measuring electrodeo The glass electrode is replaced by an inert metal for the redox potential. The metal electrode must be inert to the system in which it is usedo An inert electrode can be defined as one that does not participate in the reaction but merely acts as a conductor of electrons other than its own. It is important that the inherent characteristics of the metal do not affect the potential measurementso There are various types of reference electrodes but the most commonly used is the calomel electrode. The hydrogen electrode is not very convenient to construct or use, and it is used only as a standard for the other reference electrodes. The values obtained with the calomel electrode can be referred to the normal hydrogen scale since these electrodes are standardized against the hydrogen electrode at different ion concentrations and temperatureso The reversible oxidation-reduction equation for the calomel or mercurous chloridesmercury-potassium chloride cell is as follows: Hg2Cl12 + 2e = 2Hg(l) + 2C1 -(s) In order to join the reference electrode with the solution to be measured, maintain electrical neutrality between the solutions involved, and to reduce the liquid junction potential (caused by the contact of two different solutions), a salt bridge is used. Potassium chloride is generally used for this connecting bridge because the liquid junction is minimized due to the almost identical mobilities of the potassium and chloride ions. This does not eliminate completely the junction potential, but its value is Iegligible where measurements do not have to be of the highest accuracy.

18 -The principle generally used in the measurement of potentials is based on the Poggendorff compensation method. The voltage of the cell is measured when no current is being produced due to balancing of the resistances. Since a cell behaves reversibly only when an i.nfinitesimally small current is produced, this open circuit where no current flows is ideal for deters mining the reversible potential.. This is accomplished as shown in Figure 2. E-=Cell of known EMF R_______ _ R-Variable resistance iG- Galvanometer V-Voltmeter C S X-Solution Half Cell C-Reference Half Cell B-Salt Bridge S,-Switch X C Figure 2. Circuit for the Measurement of the Electrode Potential In order to obtain the potential of the solution, X, the switch, S, is closed and the resistance, R, is balanced so that the Galvanometer, G, shows no current flowingO The potential is then read. from the potentiometer. At this point the potential from the known cell, E, is exactly compensated by the measuring system, X and CO Since the curren.t drawn during the balancing of the circuits must be kept very small in. order to avoid polarization of the metal electrode, a very high resistence galvanometer must be used. Vacuum tube potentiomreters with very high resistances are presently used to minimize or completely eliminate this problem. This is especially important in biological systems where slight polarization of the electrode can cause errors in the readings.

FACTORS T:O BE CONSIDERED IN THE USE OF METAL ELECTRODES Of the previously described components required for potential measurement, the metal electrode presents the most complications. This is a result of the requirement that the metal be inert and the fact that no metal actually is inert. From the initial selection of the metal to be used as the electrode to the final evaluation of the potential measurements, the investigator will require a knowledge of the factors involved. These factors are: the metal to be used in the electrode, the size and shape of electrode required, the method of cleaning the metal surface, the effect of cleaning on the metal, both from the physical and the electrical viewpoint, the effect of the experimental conditions on the electrode, i.e., gas adsorption, and finally the characteristics of potential measurements in various systems. The literature survey conducted by the author failed to reveal an adequate coverage of the various factors outlined above and in no instance was there a compilation of what is presently known about the subject. One of the, purposes of this dissertation, therefore, was to summarize what could be found in the literature regarding the factors influencing a potential measurement. Types of Electrodes Many different metals have been used in oxidation-reduction studies. Platinum has been the most popular, with gold in a secohdary position. Other choices have included iridium(57), tungsten(58) graphite (59, mercury, rhotanium alloy(49) and palladium(60). Clark(49) -19 -

-20 -used a wide variety of electrodes which L. cluded platinum wire and foil, gold foil, gold plated wire and foil) rhotanium alloy) burnt-on platinum, gold plated and amalgamated burnt-on platinum and mercury. Electrodes have been used in the form of straight wire9 spiral wire9 gauze, foil, and plated glass. The first type of oxidation-reduction electrode was the hydrogen electrode used in pH measurements. In this electrode9 platinum was used as the base metal for the plating of platinum black. The metal deposit was necessary to attain rapid e4uilibrium with the hydrogen gas atmosphere. Gold was in some cases used as the base metal, With the development of the quinhydrone oxidation-reduction system as used in measuring pH, more attention was paid to the electrode base metal. An inert metal surface was essential in this technique9 and platinum and gold were the common metals used for the electrode. The electrode most commonly used in other oxidationreduction systems9 both inorganic and organic, was platinum. The other metals mentioned previously9 except for gold. were used only in isolated instances o The use of electrodes in bacteriological and biochemical systems presented a more difficult problerem to many workers since these systems often yielded erratic potential results, Many types of metal were tried, but platinum was the most popular with gold in a secondary position. For sewage treatment systems, platinum was used by all workers. Since sludge digestion presented problems siw1lar to the bacteriological systems, it was decided by the author that either platinum or gold should be used9 no better metal. having yet been discovered. The following discussion was therefore limited to these two metals.

.21 -Cleaning the Electrode In order to achieve reproducibility of an electrode and agreement between different electrodes, various techniques have been employed to cleanthe metal surface. A lack of agreement existed between various workers regarding which technique was best, and the reasons for a particular preference were frequently not mentioned. Methods employed in cleaning have ranged from a water rinse(61) to replating the surface(62) However, most of the methods for cleaning involved the use of an acid, an oxidizing solution, heat, or an abrasive. Often two or more of these methods were used in conjunction with each other. One of the most frequently used techniques has been nitric acid cleaning. This acid has been used concentrated or diluted, hot or cold. Another popular cleaning method required a potassium dichromatesulfuric acid mixture, or chromic acid solution0 Hydrochloric acid has been used alone or in combination with nitric acid to form aqua regia0 In addition to the frequent use of water as a rinse, it has also been used as the sole cleaning agent. Heating of the metal to redness by an alcohol flame has been~acommon technique, while polarization was mentioned in many research paperso Abrasive cleaning was done by rubbing the metal surface with filter paper, sand paper, or materials such as levigated alumina or Bon Ami. A few authors have compared some of these methods in an attempt to find a superior process for cleaning the electrode. Borchardt(63) tried the alcohol flame, a hot chromic-sulfuric acid mixture, hot nitric

acid, and levigated alumina for platinum and gold electrodes~ The alumina was preferred for measurements in quinhydrone si~nce the other mrethods resulted in poor electrode agreement. In addition, he found that when using the gold electrodes in amino acid-chlorination studies, it was necessary to clean with alumina before reuse. HEowever, if the electrode was used in quinhydrone, a water rinse was sufficient. i (6 ) Remington and Trimble found that the potential in hypochlorite solutions showed no indication of equilibrium, rarely or never was twice the same value, fell lower and lower in successive experiments and that different electrodes behaved in the same manner. Treatment with chromic acid, alkaline potassium permangenate and nitric acid did not, improve the performance. They found cleaning with hot nitric acid, washing, drying, heating to cherry red and cooling before lmme=sion and reading, yielded fairly consistent results in successive experiments. These investigators felt that the heating degasif ied the electrode0 To achieve the cherry red heat, an electric current was preferred to a gas burner due to possible gas con.tamination and cracking of the glass-metal seal. Si.mi arly, Clark(65) used an alcohol flame to prevent contamination by the sulfur Jin ordinary gas. Ward(66) tri ed hot chromic acid, nitrtic acid, and aqua regia for platinum electrodes in sterile and cultured synthetice media working with -0 colt Although none of these cleaning methods proved to be superior, nitric acid was adopted since its action was less vigorous than aqua regia. Tobin(67) found that a hot potassium dichromate-sulfuric acid mixture followed by a water wash was theonly method which woauld give reproducible results with platinumn electrodes in yeast fermentrlatia ons

As may be concluded from above, a divergence of opinion was the usual situation found in the literature with regard to the preference for a cleaning method. Although different test solutions were used in the above citations, the same divergence of opinion existed for vari-ous authors using the same metal in the same solution~ It was therefore felt that a knowledge of the effect of the cleaning methods on the metal electrode would be an invaluable aid in the selection of a cleaning technique to be used in this research as well as contributing to the understanding and evaluating of the potential measurements obtained. This literature study concerning the effect of cleaning methods on the metal included the effect of acids and other chemicals, heat, abrasives, and electrolysis with regard to their influence on the surface structure from the physical and electrial viewpoints. The presentation of this study as Appendix B was considered desirable since a compilation of information on this important subject did not exist. The results of the above study indicated that cleaning methods can change the metal surface, can create catalytic activity, can react with the metal, can only with great difficulty degasif~y the surface, and can affect the function of the metal as an electrode. Although these were all possibilities, the actual effect on the electrode potential by the usual cleaning techniques could not be definitely ascertained. A very important objective of this discussion would be achieved if it made the investigator more aware of the possible effects of different cleaning procedures on the electrode.

Effect of Gases on the Electrode Because of the production of various gases from. the sludge digestion process, it was necessary to determine if any or all of these gases would affect the potenti.al. measurement by producing a gas-metal effect. In addition, the effect of oxygen was also of interest due to the procedure used in measuring the potential. As was the case in the previous discussion, this summary presented as Appendix C was concerned only with the metals platinum and goldo The information, presented established the possibility of the sludge gas affecting the potentaial measurement. This was especially t.rue for the platinum electrode although the gold was not inert to the gas-metal effect. It would seem that the differences between the two 'types of elect-rodes was substantiagl and it migh.t be indicated that gold was more inert to gas effects in digesting sludge.

OXIDATION-REDUCTION POTENTIALS IN BACTERIOLOGY AND 'BIOCHEMISTRY The fields of bacteriology and biochemistry provide an excellent source of background information for the application of electrode potential measurements to digesting sludge. The usefulness of this information arises from the fact that sludge digestion is a biological process containing various bacteriological and biochemical systems. Although sludge digestion with its multitude of organisms, continuous operation, and naturally buffered media, differs somewhat from the conditions of normal bacterial cultures, potential observations derived from bacterial growth remain valuable. These observations are even more applicable to batch-type digestion because of the lack of a well-buffered sludge medium which more closely simulates the conditions of the average bacterial culture. Because of the excellent treatise on the relation of potentials and bacterial cu.ltures by Hewitt8, the discussion of 'the basic cornsiderations involved in this relation shall be limited to a brief summary. Anaerobic organisn*, cannot grow under aerobic conditions, but they can grow if aerobic organisms first reduce thepotential to a given level. A distinguishing feature of aerobes and anaerobes is the potential range in which they can initiate growth. The different potential levels at which organisms multiply and the effects they produce are a reflection of their metabolic effects and the enzyme systemwhic h they possess, According to Hewitt the potential (Ed) range of an aerobe is generally from 400 to -200 my and of an anaerobe is 50 to -400 my or lower. Facultativatorgamiarne such as E, coli span the enttle potential range.

Reduction of the media is a necessary result ofe metabolism. In order to obtain the energy necessary for growth, any oxidizing agent such as electron acceptors, hydrogen acceptors, or oxygen is utilized by the cell and one after another of these oxidizing systems will be reduced during growth. As a- result, there is a drop in the potential level. The magnitude of this fall is dependent on which enzymes or other hydrogen acceptors are reduced and this depends on the enzymes which are present and able to function. Therefore, a more carefully constituted medium is needed for anaerobic than for aerobic growth since hydrogen acceptors must be present due to the lack of oxygen. Reducing conditions alone are not a sufficient condition for bacterial growth. This can be compared to the effect of pH where a correct value is necessary, but not sufficient for growth. It is stressed by Hewitt(48) that bacterial cultures develop reducing conditions more because of the exertion of reducing activities than due to the liberation of reducing substances. The metabolic activities of the bacterial culture can be divided into four periods. These are the lag phase, logarithmic growth phase, stationary period and declining period. The first two are of particular interest as they add insight into the biological and potential changes that occur after adding raw sludge to the digesting solution. During the lag phase, intense bacterial action alters the environment for the best growth and the oxidation-reduction potential begins to fall. The logarithmic phase takes place after the bacteria have

"27. altered the environment and active multiplication takes place. It is characterized by intense enzymatic and synthetic activities. The lowest potential is reached in this phase where the reducing conditions are symptomatic of the dynamic equilibrium between the enzymes and the constituents of the medium which are always in a state of change. Items which are constantly changing in this system abe the potential, pH, chemical composition, number and condition of organisms, etc. The next two steps are characterized by a rising potential, a decrease in bacterial growth and activity, and the gradual death of the bacteria. The potential characteristics of many bacterial cultures are given by Hewitt(48). Many of the biochemical reations taking place during bacterial growth or participating in other metabolic processes are also covered. A few of these potentials are given in Table I. Included in this thbie are potentials values from other sources(68)(69) The values listed are the potentials of the half-reduced system, E'o at pH 7, except for bacteriological cultures. In these cases the minimum potentials reached during growth are given. Although definite potentials are established in bacterial cultures, it is not fully known what substances are responsible for these potentials or what oxidation-reductions systems are actually being measured. Hewitt(48) states that some systems may be present in too small a quantity to affect the potentials while others may have a large poising power and therefore would act to inhibit any oxidation-reduction processes which are different since they would have to be altered themselves. Therefore, the quantity of the redox system as well as its potential level will determine

W28 -Table I OXIDATION-REDUCTION POTENTIALS OF SOME BACTERIAL CULTURES AND BIOCHEMICAL REACTIONS System E (volts) Succinate - fumarate............................. -0.01 Glutamate - ketoglutanate....................... -0.03 Amoeba in air......o.......................... -0.07 Alanine - pyruvate.............................. -0.07 Malate - oxaloacetate..................... -OO10 Oral lactobacillii cultures..........o.......... -0.13 Pneumococcus cultures.o.......,..oo.......... -0.15 Ethanol - acetaldehyde....................... -0o.16 Lactate - pyruvate........o.....o............ -0.18 Bacillus polymyxa cultures..................... -0.2 Flavines...................................... -0 o23 Glutathione (pH 7o15)...................... -0o233 Lactobacillus acidophilus cultures.............. -0.24 Amoeba in nitrogen...................... -0o25 Bacillus macerans....................-0.27 3-phosphoglyceraldehyde: 1, 3-di.phosphoglycerate. -0.28 Beta-hyhroxy-butyrate~ acetoacetate............ -029 Isocitrate - oxalosuccinate...............O... -0.30 Coenzyme I (pH 7.4, 20~C)........... o.... -0.325 E. coli cultures.............e............. -0.40 Hypoxanthine-uric acid (pH 7.2)................. -0.41 Glucose-gluconic adid.......................... -0.45 the final potential. In addition there are the effects of components which do not directly affect the potential, but do so indirectly by poising, catalysis, and poisoning of the electrode surface. Finally, the oxidationreduction systems which are irreversible or semi-reversible affect the equilibrium and complicate the results. The irreversible systems can affect the electrode in two ways according to Michaelis(l6). There can be no effect on the electrode and no direct effect on the reversible system potential. The only effect is

by changing the ratio of the components of the reversible system which may oxidize or reduce the irreversible system. The other type of behavior consists of the determination of the potential at the electrode. Time is needed for the potential establishment and the potential is apparently independent of the quantitative ratio of the reduced reactant to the oxidized product. In spite of the lack of knowledge concerning the substances responsible for the potential in bad Lerial cultures, the use of the electrode potential has been successful in providing.a better understanding of the behavior of these cultures. In a similar fashion it can be anticipated that the electrode potentials of digesting sludge will also provide a further understanding of the sludge digestion process even though the problem of what the potential is actually measuring becomes more complex. Although the potential of a biological solution can be measured without any knowledge of what is causing the potential, the characteristic of the metal electrode in these solutions are an integral part of the potential measurement. Many authors have had difficulty in measuring the potential of biological systems. Hewitt(48) felt that this was due to the fact that no two organisms were absolutely identical and no organism was quite the same on two different occasions. However, the large majority of investigators felt the difficulty was due to the electrode itself. Erratic electrode behavior in biological systems resulted in an extensive study of organic oxidation-reduction dyes in relation to electrode potentials by Clark et al. (9) This work was carried on to obtain more

-30 -basic data on electrode.. potentials in order that explanations would be more feasible for biological systems. With biological material they found disagreement between electrodes of different metals, of different surface structure or even between electrodes of the same metal. It was found that cultures, especially in the early growth phase, were not well-poised and different electrodes did not yield the same potential. Such electrodes did tend to agree after a time duration. There was no case of static equilibrium. The authors stated, "The criterion of stability with time, which is one of the most important criterions of significance in potentials observed with definite reversible systems, is lacking." They believed that mainly ii-r reversible systems were present. Another very extensive study of electrodes was done by Michaelis(l6)o He stressed that the important criterion for a potential measurement was that it must be independent of the individual electrode and show' agreement in the same solution. In biological systems two apparently identical gold or platinum electrodes in the same solution differed by more than 100 mv and in such cases he felt these values could not be said to be that of the system. This divergence could be due to (1) a poisoned electrode surface or (23 a system so weakly poised that it failed to overcome the slight inherent individuality of the electrode. The influence of the electrode metal on the potential was considered by Barron, Flexner and Michaelis(70). They stated that an answer concerning the inertness of the "unattackable" electrodes was necessary for the correct interpretation of all potentials measured in tissues and

531l bacterial culture. They concluded that this was "of such great importance that it justifies a separate study of platinum and gold which is soon to be reported." Unfortunately, however9 no record was found by the author covering this proposed study. The question concerning which metal, platinum or gold, was best for oxidation-reduction electrodes has resulted in differences of opinion. In order to properly determine the type of electrode to be used in this research, it was considered desirable to review the work of other investigators who have studied this problem. Because of this restricted viewpoint, the following literature survey on the use of electrode potentials in bacteriology and biochemistry did not represent the large majority of cases where platinum was used as the electrode, The emphasis was on experimental work where gold and platinum were compared or where the characteristics of either metal as the electrode were discussed. Other investigations which contributed directly to the evaluation of potential measurements of the sludge digestion process were also included in the review. It was first noted by Potter(71) in 1911 that the electrode potential of an inoculated culture was more negative than that of the sterile medium~ Gillespie(72) in 1920 was the first to observe the development of reducing conditions in bacterial cultures by following the fall in potential, Although he discussed the use of gold, only mercury, platinum and platinum black were used as electrodes. Michaelis(l6) stated that the germs of all later investigations were contained in this work and the ideas credited to Gillespie were. (1) interpretation of the potential

as an oxidation-reduction potential, (2) comparison with the hydrogen electrode at the same pH, (3) exclusion of oxygen, and (4) the fact that the addition of glucose to the culture increased the velocity of the establishment of the maximum reduction potential. The potential drop resulting from bacterial growth as well as (49) the other findings of Gillespie were confirmed by Clark et al. and many others. It would be beyond the scope of this thesis to even attempt to list the many authors in this field. However, an excellent summary of the quantity of work done with bacterial cultures and electrode potentials can be found in the excellent reference book by Hewitt(48)o A relationship between the oxidation-reduction potential and the hydrogen electrode potential was noticed first by Gillespie(72) with a growth of E, coli. This was also observed by Clark et al (49) with E. coli using platinum electrodes for the redox potential and platinized platinum for the hydrogen electrode. The curves of Efh and the hydrogen potential crossed and remained close together after one hour at approximately -380 my and then drifted upward. In spite of a alight but repeated depolarizing current, there developed an over-voltage which was maintained. This was observed many times. The gas produced from this culture consisted of slightly more carbon dioxide than hydrogen. Thus, the partial pressure of hydrogen was less than 0.5 atmosphere. A case was cited where the addition of a hydrogen suppressor, asparagine, resulted in a potential slightly positive to that of the hydrogen electrode. Organisms of the dysentery group in media

from which no hydrogen was produced brought the potential almost to the hydrogen electrode. They concluded that the evaluation of these observations was not possible uhtil more experimental work was done, Among the others who investigated the growth of Eo coli, the work of Boyd and; Reed(73) was of special interest since they published their original experimental data for each of two platinum electrodes. This provided an example of the kind of eleetrode agreement that might be expectedo The usual practice was simply plotting the average value if more than one electrode was used. The electrodes were allowed to remain overnight in the sterile 'broth befor.e inoculating the broth0 For the first three hogrs after inocu -lation of the broth the eleetrodes differed by 70 myo Ho wever, during the active growth period the electrodes differed by only 10 to 20 myn After 14E hours the values began to diverge until at 50 hours *they were 130 mv aparto Although this large divergence between electrodes was probably mo~re thanr was the usual situationr it did illu.strate how the usi of one electrode could give misleading data0 Addition of glucose to the broth before inoculation induced an. earlier and greater fall. in the potential. The beginning of the fall was coincident with the first evolution of gas and they concluded that hydrogen was a factor in the potential change0 In previous work(14) it was stated that the potential change was not due to the carbon dioxide production. The electrodes behaved similarly to those in non-glucose broth exhibiting a large disagreement at first which betame. less during the active growth period0 At 10 hours when the readings were terminated, the electrodes differed by 35 mvo

Another instance where individual electrode values were reported was by Ward(66) in the investigation of E. coli growth in synthetic media. He found large disagreements up to 95 my between electrodes in the same vessel and up to 450 my between the vessels themselves. Various types of platinum electrodes such as foil, coil, gauze and straight wire were used due to "somewhat conflicting statements in the literature", With nutrient broth and agitation there was much better agreement between the electrodes and between different vessels. The electrode agreement was usually within 5 my with an occasional 20 my difference. Between cultures the difference ranged from 5 to 10 my with a greater variation during the active growth period. It was concluded that potential determinations in poorly poised media in the presence of oxygen were dependent on the type of bright platinum electrode used. This was believed due to the possibility of an oxygen electrode effect, Platinum was used almost exclusively although a few gold electrodes were ti-ied. The use of gold was the result of the work of Longsworth and MacInnes(54) and Elema et al.(7) who reported that gold was more sensitive than platinum. However> Ward did not find any advantage in using gold, and there was no apparent sluggishness of platinum as compared to gold. Ward found that stratification caused different readings from temporary poisoning of the electrode and agitated to get better agreement by using nitrogen. When a synthetic broth was intermittently agitated, better agreement between duplicate electrodes was exhibited although a considerablp difference still existed between different broths, When stirring was started or stopped there was no definite trend, some potentials rose and some fell slightly.

Nitrogen gas was used in many experiments to deeaerate the culture for anaerobic growth and as a result, agitated the solution. Differences in agitation d.id not change the potential, but did change the time neces(14)(75) (75) sary to reach the ultimate potential. Coulter used gold electrodes for potential-time experiments on sterile beef bouillon. Initial potentials were variable and. poorly poised and were between +250 and +150 rv, Individual electrodes might disagree by 50 mv but often by only 5 my when purified by polarization. A common limiting value and time were found for various bouillon samples. These were between -50 and -60 amy and the time was five to six hours for a rapid rate of oxygen removal by nitrogen. This gas rate was sufficient for agitation also. With less deaeration time, the potential change was slowear, taking up to five days to reach -50 to -60 mv. Coulter found that close agreement was not possible with electrodes of the purest metals obtainable and therefore purified the gold himself. This was fused on. platinum wire and enclosed in glass. Four electrodes were used simultaneously. It was frequently necessary to make the electrodes sensiti~ve by slight polarization. Deaeration of a sterile meat broth by nitrogen was used by Knight(62) to study the effect of pH on the potential. The final limiting value was reached in six to eight hours, Soon after the d.eaeration was started, large differences in the potential occurred.. Fowever, as the potential became more negative, the poterntial values tended to come together until they differed by 10 mv in the final value. Occasionally one of the three electrodes used in each run woluld behave in a different manner.

This was theorized as being due to & crack in the glass seal with the metal. The electrodes were made by freshly gilding a platinum foil electrolytically for each experiment. The foil was welded to a platinum wire and the contact made by mercury. Of more importance to this research than the addition of nitrogen to create anaerobic conditions by dieaertbing the broth was the reverse case of adding oxygen to the anaerobic broth. This was because the addition of an oxidizing agent to a reduced bacterial culture or broth closely corresponded to the condition of feeding a more oxidized raw sludge to a more reduced digesting sludge. It was found by Clark, et al. (49) that an addition of an oxidant to a cell suspension caused a positive displacement of the potential followed by a recovery toward the originally more negative potential. It was concluded that cell suspensions appeared to be of low active capacity acting as if small quantities of active poising material were fed from a large reservoir. This reservoir was slowly but decidedly mobilized. The effect of oxygen on the potential of anaerobic bacterial growth was studied by Aubel, Rosenberg and GCnberg( They stated that a zone of defense was set up by hydrogen and reduced substances to counteract the oxygen. Using Clstridium saccharobutyricum it was found that 1.4 percent oxygen merely delayed the negative potential drop, but at 2 percent the culture was unable to cause any reduction. Knight(77) used oxygen in nitrogen to poise the potential of a sterile and hnoculated broth at a constant level.

Instead of varying the oxygen input, Reed and Orr(78) added a definite volume of air to anaerobic broth containing different quantities of a reducing agent and two platinum electrodes, One ml of air passed into a peptone broth caused a sharp potential rise of approximately 40 mvy, The potential returned to its original value in a few hours with 0,1 percent sodium thioglycollate but continued to rise with only 0.01 percent sodium thioglycollate in the broth. The addition of oxygen or any other gas to a bacterial culture or sterile broth could affect the potential directly by a gas-metal effect, or indirectly by changing the oxidant-reductant ratio. The direct method would depend not only on the particular gas used, but also on the type of electrode metal. Studies of the sulphydryl system, cysteine-cystine, have produced a considerable amount of information on different electrode metals due to the difficulties in obtaining potential readings by different investigators. The electrode potentials in this system did not fit the behavior of a perfectly reversible system. This system is of the type~ 2RSH = RSSR + H2. Specifically the reaction is: 2 CH2SHCHNH2COOH = COOHCHIH2CH2S SCH2NH2CHCOOH + H2 Cysteine Cystine Such compounds are present in biological systems and are involved in many vital processes. These amino acids are present in digesting sludge and it was therefore felt that the results obtained on this system could aid materially in the understanding of electrode potential behavior in digesting sludge.

-38 -Dixon and Quastel(61) tried platinum and gold in the cysteine-cystine system. The drift exhibited with platinum was explained by the hydrogen pressure. Methods that were tried to control the drift included passing a polarizing current through bright platinum and the use of platinum black. The platinum black merely slowed the drift, The bright platinum was not affected by stirring or bubbling with nitrogen and was independent of the initial state, i.e,, ignited or not. The electrodes were satisfactory in alkaline solutions but not in acid solutions where they slowly reached steady values and then gave considerably different values. Since biological reactions were near pH 7 or less, they concluded platinum was unsatisfactory for this type of system. They also were not able to get extreme accuracy in a solution such as quinhydrone. Other electrodes, gold-plated platinum and massive gold, were also investigated but the only satisfactory electrode was pure gold. It gave very steady readings independent of the rate of stirring, was good in acid or alkaline solution and returned to the intial values after slight anodic or cathodic polarization. But these results could not always be reproduced with great accuracy with definite concentrations of cysteine. Such small deviations were "in no case sufficiently large to affect appreciably the magnitude of the potential". Differences of up to sixty millivolts caused much comment by Michaelis(l6) concerning the reproducibility of these electrodes, The only cleaning felt necessary was washing the electrode with water. The gold plate electrode used hid a surface area of 5.35 square centimeters.

539 -Kendall. and Loewen(79) compared gold, platinum, mercury, and dyes in cysteineo With cysteine and platinum there was a six to eight hour drift to a value of -140 mv. But the reduction potential was higher and reached equilibrium quicker if the electrodes were exposed to the air for a time longer than usual. They concurred with the statement by Dixon and Quastel(6l) that the reduction potential "was determined by the previous history of the platinum electrode used". It appeared that the electrodes were influenced by the nitric acid used in cleaning as well as the air. Anodic polarization of platinlaru in cysteine delayed the development of the reduction potential. Cathodic polarization caused a greater reduction potential which driLfted back to the original value of 140 mva With gold and cathodic polarization the potential was obtained with a:little greater speed than with platinum, and with anodic polarization gold promptly returned to the original potential. They felt the difference in behavior was the speed with which the sulphydryl group could react with oxygen on the surface of the metal, The smaller fluctuations with gold than platinum also observed by Dixon and Quastel were explained by the traces of oxygen or oxidizing agents adsorbed on the platinum whereas gold failed or didn't adsorb oxygen or oxidizing agents as well. Tests with reducing dyes showed that gold responded more rapidly than platinrum and was less affected upon the subsequent addition of hydrogen peroxide. If cysteine was then added, the gold responded more quickly to a steady potential. The final values attained less than an hour after the addition of cysteine were the same for the gold and platinum ~ Subsequent additions of the peroxide yie ded a greater response by platinum.

In evaluating the platinum as a hydrogen acceptor for the cysteine reduction potential, they found this could not be shown to occur and they added that therefore "the probability of its occurring with the gold may be entirely dismissed". These authors therefore expressed the belief that gold was more inert to hydrogen than platinum. Michaelis aand Flexner(80) have compared platinum, platinized platinum, gold-plated platinum, and solid gold in cysteine solutions. For blank platinum a long time was necessary to establish a final, definite potential. It was four to six hours after the start of deaeration by nitrogen before they felt justified in calling it the ultimate potential. They found platinum very sensitive to oxygen. After the potential was reduced by the system in several hours, oxygen was passed through which caused the oxidation potential to rise. However, nitrogen rapidly returned the electrode to its equilibrium valueo This indicated to them a change in the electrode itself with the possible formation of an oxide or suboxide on the surface.which had to be reduced, and with each metal differing in this respect. With small amounts of oxygen. the potential depended on the speed of bubbling gas, the concentration of cysteine, and stirring. Without oxygen the stirring effect was not noticed. They attempted to get equilibrium with hydrogen but the velocity was slow and the process was incomplete. By polarizing the return velocity was fast with cathodic and slow with anodic polarization. Different time intervals were necessary to obtain the same final potential even though all the electrodes checked with quinhydrone. Platinized platinum drifted as long as nine hours which was longer than blank

platinum. Gold plate was similar to platinum except there was less response to hydrogen and the final potential was reached more Slowly~ There were individual differences in the time needed to reach equilibrium. If all traces of oxygen were excluded, they found the final potential was the same for platinum, gold-plated platinum,, or mercury. Pure gold was different in that it was the quickest to reach a constant final potential and there was less variation with a change from nitrogen to oxygen than for gold plate. The difficutly was that the final potential was different for each electrode and never was as negative as with the other electrodes. They behaved as though the surfaces contained oxygen, with each electrode having a different degree of oxidation. These experimenters felt that in the work of Dixon and Quastel the Eo constant of the potential equation varied too much for different gold electrodes and there was also variation for the same electrode in different experiments. Gold was discarded since acceptance of the potential as an intrinsic property of the solution was felt to be impossible. The long period of time necessary for the establishing of the potential with platinum. and gold indicated that the mechanism of the cysteine potential was different from truly reversible systems which established their potentials almost immediatelyu The authors felt that sulphydry1 compounds were highly responsible for the pote4ntials of noble metal electrodes in a tissue liquid. The cysteine potential in the presence of oxygen was discussed (16) by Michaelis6) Cysteine consumed oxygen slowly and differed from wellpoised positive potential systems like quinhydrone where oxygen caused

little change in the potential. It also differed from reduced components such as dyes in the negative potential zone which had a great affinity for oxygen and hence determined the final potential by affecting the oxidantreductant ratio. The potential of cysteine was sensitive to the slightest trace of oxygen and therefore, oxygen was of significance in the potential measurement. The largest variation in the platinum potential was at very low oxygen concentrations in nitrogen, For one part oxygen in 40,000 parts nitrogen the potential varied by 17 mv from pure nitrogen. The stirring effect was most noticed at very slight oxygen pressures. Without oxygen stirring was not a definitely noticeable factor:. A constant oxygen consumption at all pH's indicated that oxygen was deposited on the electrode surface and not consumed by the cysteine. For the oxygen effect on the electrode, two possible explanations given were: 1) tendency to behave as an oxygen electrode, and 2) catalytic effect on hydrogen and oxygen. Comparisons of platinum and gold were made by other investigators which resulted in differ4ing viewpoints, Boyd and Reed(14) studied differences in platinum and gold in anaerobically aged beef extract broth at OC. The electrodes were polished foil in glass tubes8 A gas flow speeded equilibrium slightly and slightly decreased the time to reach the limiting potential, while it did not affect the final value, Slight polarization had very little effect, For the sterile broth without a gas flow, they found slight variations in the potential in any one experiment with gold or platinum, With two or more samples under the same conditions, the potential varied considerably. A statistical study was presented to

-43 -analyse these differences. An excerpt from this is shown below. All values are in millivolts. No. of Initial Std. Probable Error P. E. in Metal Samples Mean Dev. in Mean Std. Dev. Platinumh 14 86 77 15 10 Gold 8 122 41 11 8 Thus platinum gave a mean potential of 71 to 101 myv. with a possible variation from -100 to +163. Gold gave a mean potential of 111 to 133 my. with a possible variation of from 73 to 171 my. The variation within any one sample was much less than variation from sample to sample. Nod of No. of Std. Dev. of P. Eo in Std. Mean Readings Samples Initial Mean Dev. Platinum 75 3 7.2 1.7 Gold 101 4 22. 4.4 Longsworth and MacInnes(54) compared a gold and a platinum foil electrode in a broth containing Lactobaccilus acidophilus under anaerobic conditions. The platinum was found more sluggish and had more lag than the gold. The gold showed symmetry to the acid production rate curve and they suggested this meant the values were nearer to the instantaneous ratio of the oxidant and reductant0 They felt the difference was due to the fact that the metals attained equilibrium at different rates with the electro(81) motively active materials in the solution. They had found previously that oxygen in the nitrogen affected the potential of the gold electrodes~ The gold electrodes used were foils one cm square welded to platinum.

The exposed platinum wire was covered by a thin film of gold by melting. The gold electrodes agreed closely within three mvy. The maximum difference was nine my when a rapid change occurred in the solution potential. When the electrodes were placed in the media from the sterilizer, gold usually gave +250 my. The platinum value Vas from +400 to +500 mvY indicating a larger response to the atmospheric oxygen dissolved in the media. Saturating the media with a nitrogen-carbondioxide mixture resutted in a gold potential equal to zero which corresponded to previously reported values. Platinum approached this value but in a much slower manner requiring a period of time of two days duration. Lepper and Martin(57) found gilded platinum and mirrored iridium disagreed up to 30 mv in a cooked meat media. The meat media was autoclaved with the electrodes and allowed to stand overnight. When very fresh the potential of the media was -200 mvy, but the potential usually was -174 mv. Stirring dropped the potential 50 mv at once and it took several hours to return to the original value. These electrodes were compared in the meat media following the inoculation of anaerobic bacteria (82) With no obvious evolution of gas from a culture of B. tetani, the gold and iridium agreed and the potential dropped to ~400 mnv. However, with gas evolution using B, welchii, the iridium rapidly dropped to a potential of -450 while the gold remained approximately constant at less than -200 mvo They explained this by stating that iridium functioned as a hydrogen electrode showing the potential oei the buffer solution surrounding the meat, while the) gold was inert to hydrogen and remained

poised by the feeble reversible system shown previously to be at approximately -200 mv, By adding a dye of lower potential with ES = -0~289 they found the gold dropped to a potential of -330 mvo. ten they reduced the gas-forming period by adding a suppressor, the iridium started to rise in four hours and agreed with the gold in seven hours. Clark(49) in his extensive work with many types of electrodes did not draw any conclusions regarding the best electrode metal, although he appeared to use platinum and gold-plated platinum in biological and related systems, Borsook and Schott(83) used gold-plated platinum foil in studying the succinate-enzyme-fumarate equilibrium. Platinum was also used and the results were the same for both types of electrodes. Hewitt(48) preferred platinum, and stated that with gold wire and foil he did not obtain the reliable reproducible results obtained with platinum. No quantitative information was given. After reviewing the different electrodes used in the field of biochemistry and bacteriology, he concluded "that no more suitable electrode than bright platinum has been described." Although this discussion does not provide a definite answer to the question, "What is the best type of electrode for use in bacteriological and biochemical systems?", it does present the advantages and disadvantages of the metals, platinum and goldo From this and the general characteristics of the metal electrode in question, a much more valid basis is possible for the selection of the electrode. In addition, this information can be of assistance in the understanding of oxidation-reduction potentials as applied to sludge digestion

ELECTRODES USED FOR THE QUINHYDRONE SYSTEM Another source of information for the investigation of the applicability of gold or platinum as an oxidation-reduction electrode, is the quinhydrone system. This organic compound is often used in conjuction with investigations of the potentials of bacteriological, biochemical and other systems to provide a means of di.terminng: whether or not the electrode is functioning properly. Quinhydrone is an excellent reference standard for electrodes because of its potential characteristics which include stability, reproducibility, being well-poised, rapid equilibrium between the electrode and the solution, and a large capacity to resist change from environmental factors such as the dissolved oxygen content of the solution. The impetus for the study of the quinhydrone electrode was due to its use as a pH indicator before the glass electrode was developed. The usefulness of quinhydrone resulted from the fact that the ionization of the compound in a saturated solution produced an equal concentration of oxidant and reductant which provided a linear relationship between the potentialand pHo The widespread use of quinhydrone resulted in many investigations of the relative merits of platinum and gold as the inert metal electrode for this oxidation-reduction system. A survey of the electrode preferred by various workers and a comparison of platinum, gold-plated platinum and gold of different sizes, ages and sources with air and nitrogen were given by Morgan and Lammert(84)-46 -

They used an ool N hydrochloric acid solution with quinhydrone. They found more people preferred gold and those who found platinum and gold similar, also preferred gold or gold plate due to sharper readings. These authors preferred platinum because2 (1) platinum could be sealed directly into glass. (2) gold was easily alloyed with mercury and contamianated, (3) a larger current capacity was available for platinum, and (4) gold was at times variable. They found nitrogen improved the reproducibility of platinum and gold plate but did not affect gold or graphite. This effect meant to them that possibly platinum catalyzed the reaction with oxygen better than goldo They found in addition that the methodof gold plating added a variable factor. They felt that the foil or wire should be larger than one square centimeter and concluded that gold and platinum were interchangeable. Corran and Lewis(85) tried platinum foil and abandoned this metal since they obtained unreasonable values. They felt that the platinum foil acts other than a' an inert metal in this system and gold foil was used. After each measurement the gold was cleaned with chromic acid, boiled in distilled water and tested in a buffer solution. This cleaning was felt to be essential for the tests. However, Grossman(86) abandoned gold.in favor of platinum for the pH of organism fluids including blood. For similar solutions Mislowitzer(87) also used platinum in preference to goldo The use of platinum was not recommended by some investigators who also did not consider gold and platinum interchangeable. Borchardt(63) found platinum wire gave erratic results in quinhydrone after trying

different cleaning methods. These were: (1) nitric acid, (2) alcohol flame, (3) hot chromic and sulfuric acids, and (4) a mild abrasive such as Bon Ami. Plating the platinum with gold gave good results but the plate did not stand up with the cleaning method adopted and therefore solid gold electrodes were adopted. Dixon and Quastel(61) also found that platinum did not yield the expected accuracy in quinhydrone. The results of the survey on the metal preferred for the quinhydrone electrode indicated either that gold was usually preferred or the metals could be used interchangeably. The preference of platinum was not a common occurrence. That there should be any difference in the behavior of gold and platinum was interesting due to the fact that quinhydrone was regarded as a well defined reversible oxidation-reduction system which could overcome the inherent characteristics of the metal(16) Therefore, the electrode should independently indicate the potential of the solution without regard to the metal used.

OXIDATION-REDUCTION POTENTIALS. IN SEWAGE TREATMENT There are many opportunities for the successful application of oxidation-reduction potentials to sewage treatment. Hood(88) has summarized the possible benefits of potential measurements in plant control as follows~ 1) Detection of deposits and growths in sewers 2) Detection of industrial wastes and toxic conditions 3) Proper sludge pumping scheduling 4) Detection of primary tank stagnation 5) Determination of conditions in filters, activated sludge tanks, sludge digestion tanks, influents and effluents. In addition to their use simply as a control device, potential measurements can also contribute to the further understanding and evaluation of sewage treatment processeso The application of oxidation-reduction potentials in the sewage treatment field has been concentrated on the activated sludge process. The large potential changes, which occurred after aeration was stopped, provided a very fertile field for laboratory study. This was also the situation for the anaerobic aging of raw sewage. In contrast to this the anaerobic digestion of sludge received little attention. The use of an oxidation-reduction indicator was first proposed by Spitta and Weldert(89) in 1906. They recommended the use of methylene blue as an indicator of the oxidative state of a sewage effluent. This

-50 -was further developed by Phelps and Winslow(90), and called the relative (49) stability test. Clark et alo gave an excellent discussion of methylene blue and electrodes in sewage. Data was presented on the change of potential with time in buffered sewage at different pH levels using potentiometric measurements. They recommended the use of electrodes rather than dyes. This was the first case of the use of electrode measurements in sewage. The first electrode measurement of activated sludge was by Dickinson(91), in 1940 who made a bioelectrode by growing activated sludge on a platinum foil of one square centimeter. No direct EMF measurements were takpn; only deflections of a galvanometer were used. There was no quantitative agreement. Rohlich, Sarles, and Kessler(92) in 1941 were the first to obtain potential readings on activated sludge mixtures and the first to do a plant survey with potentials. This work was expanded by Rohlich(93). One platinum electrode was used in the plant survey and three electrodes in laboratory aeration studies. These uasually checked withimtntemm<v. Trottble encountered with the electrodes was believed due to a film on the metal surface and for successful cleaning, the electrode was polished with a levigated alumina paste, dipped into aqua regia or hot nitric acid, washed thoroughly, and then placed in distilled water for several hours. This vecommended cleaning method was later changed by Rohliah(94) to polishing with a clean filter paper or a mild abrasive.

-51l Moore and Ruchhoft used three platinum electrodes in activated sludge studies(95). The metal was shaped as a coil after comparison with foil and straight wire electrodes in potential studies of inoculated (96) filtered fresh sewage (9 The maximum difference between the three electrodes in the activated sludge studies was 80 my, but agreement was usually within five mvu Cleaning was done by alternately immersing the electrodes in hot dilute nitric acid and sodium bisulfite followed by washing and storing in distilled water. Other potential investigations of an activated sludge media were done with a single electrode(88)(97)(98)(99) except in the case of Keefer and Meisel who usually used three electrodes (l0)(ll)(l0 2)o The only workers who specified a cleaning technique were Nussburger(99) who cleaned his electrodes by rubbing with Bon Ami and Hood(88) who used nitric acid followed by distilled water. The first investigation of the potential of digester sludge and supernatant was done by Hood(88)o Potentials of other-plant units were also measured. The platinum electrode used was 15 to 18 cm long and although it was not specifically mentioned, it appeared as if only one electrode was used. This was cleaned by immersing in nitric acid and rinsing with distilled water for the sludge studies. The time interval used for readings varied from 10 to 30 minutes at different plants. Readings were taken every minute until there was equilibrium or a reverse trend. No agitation was used since it was felt that this would disqualify the readings, especially in poorly poised

solutions. It was stated by Hood that recent verbal advice indicated platinum might be replaced, one reason being that values were obtained which were erroneously high. Hood presented the following potential values from different plant surveys. Digester supernatant:............ -132, -121, -67, and -139 my. Primary sludge:.................... +44 and -75 my. Digested sludge:................... -54 my. It was concluded that the hydrogen potential for anaerobic units should be from -100 to -200 my for maximum efficiency and that when the potential approached zero, it was no longer optimum. The other reference pertaining to potentials of digesting sludge was by Eckenfelder and Hood(98). To obtain the potential one platinum electrode was inserted into the sludge to be tested which was agitated by a magnetic mixer, Readings were taken every few minutes for 30 minutes. They stated that the resulting time-potential curve was a measure of the dynamics of the system, Sulfate was added to a digesting sludge with a potential of 0 to -100 my. The potential changed to a range of -125 to -175 my and the digestion was improved with better gas production. Other sludge potential measurements were made on a digested sludge with a potential of -75 my and a dormant digester with a low pH, high volatile acid content, and a potential of +200 mvy Additional potential measurements were presented on other types of wastes.

From the literature survey presented above, it can be seen that the use of oxidation-reduction potentials in the field of sludge digestion has been very limited. Since only a few scattered points in the digestion period were sampled, the value of the data was restricted. The lack of information on the potential behavior in a digesting sludge seemed to underline the need for a study on the application of oxidation-reduction potentials to the process of sludge digestion. In this thesis, the many factors affecting the potential measurement have been studied. In addition the use of potential measurements has been explored to determine whether or not changes in the condition of the digesting sludge could be detected. Methods of measuring the potential of sludge have also been investigated. This thesis, as, a whole then, was an attempt to study the use of oxidation-reduction potentials as an indicator of proper digestion, and to aid in the further understanding of sludge digestion.

EXPERIMENTAL APPARATUS The primary objective in the design of thei.,experimental apparatus was to permit the measurement of the oxidation-reduction potential within the digestion compartment. This, in turn, governed the design of the sludge digestion equipment. This was especially true in determining the method of agitation and the size of the digester; in other respects conventional procedures in the design of sludge digesters were followed. Sludge Digestion Apparatus The sludge digestion apparatus consisted of three single stage digesters housed in an enclosed chamber with provisions for temperature control, sludge feeding and withdrawal, gas collection, agitation and the electrode assembly. A schematic drawing of a sludge digestion unit containing the electrodes is shown in Figure 3. The digesters were made from one gallon glass jugs. The digesters were placed on the second level of a two level incubator containing light bulbs as a source of heat. The construction is illustrated in Figure 4. The air temperature inside the incubator was controlled at 35+1~C by means of a thermoregulator. The selection of this temperature was based on the fact that the optimum for mesophilic digestion lies between 95 and 100~F. In addition, the large amount of work done with sludge digestion at this temperature, both on a laboratory and plant scale, provided a reference for the experimental work. -94 -

-55 -The gas collection system was constructed so that digestion progressed under constant pressure. A typical gas collector consisted of two cans. The smaller can was inverted and placed inside the larger one which was filled with an acid salt solution recommended for gas retention by the AOAC(103). Each digester had two gas receivers connected in parallel. This is shown in Figure 5. Gas sampling for carbon dioxide was done in a Burrel contact adsorption pipette containing thirty percent potassium hydroxide solution. The confining liquid used was mercury. The desirability of continuously agitating the sludge digesters was based primarily on the work of various authors who found that agitation was helpful in the measurement of the electrode potential(65)(66)(96). In addition some form of agitation for the sludge digestion process has also been recommended(l9)(26)(104)(105)(106) Preliminary potential measurements in this research with hand mixing alone also indicated that agitation would be helpful. In order to select a method of providing continuous agitation, the effect of various mixing methods on both the potential measurement and sludge digestion were considered. Agitation of the digester contents has been done with mechanical mixers, gas recirculation, and sludge recirculation. The preferable method for this research was considered to be gas recirculation because of the resultant gentle rolling action of the 'digester contents with the sludge flowing upward at the center and downward at the periphery. This rolling action would allow good sludge agitation without violent mixing which would

TO POTENTIOMETER GAS TO TO MOTOR, t4 TO CALOMEL COLLECTIONELECTRODE FEED TUBE ~ -SALT BRIDGE MIXEMR-IXE 0 B 0 COPPERlllCOPPER ~~~~SECTION BB ~RUBBER ESLUDGEE TUBERUBBER VANS the Oxidation-Reduction Electrodes Figre3,Scemti Dawngo Ssd Dgti fn Ctaig ~h xdto-eucinEetoe

-C,-:) ( -II-,-.-,-.-,-.-,-.-,-, —. ---.-,-,-.-,-,-,-.-.-"-,-.,,. I I .-.-..-..-................... -;:;- — !.::::::..]:]::::::::::]:::::::;.,:. 1:.1.7:::::::: :::::... 1.. -....:...:...:.`.,.! —.,-. —,-.-.-..., -:.,...,...::::::::::::::::.,::,,1 ....... ----........... -...-......... I I... I I...%.:::: I:....".'..........'....,.,....I...,.-11.!:,:::::::::::::::::::::::::::::::::::::,:-::::; -.,,-.,-I.I..-. -..-.-....I.I...I'I..:::::::,:::::::::::::7!!.,.",..,.,.......,.,..,......,..,............:....,...",.,..".,.,...,."",.,",,.,.""I."', I I I I. I.. 1. I... - ............-.'I,,,.,.. —...,..-... .... ,......, qI.I.I...",.,..".",.,.,.,.",.,".",.', I. I. I I I I ".,.,.,.",.",".,,,."..".....'.."..'. ..............:,:.:.:.:.:::::::.,.:::-::.-.,.. :i: 'I.I"1.1"."",.,."""""I.I."",.",.I."""",.. ". "."""."",.,""."",".",.,.,.......",........ t".: 1..,............".........., i...I...I... I. I,'%'.%,.%%%,.,,%%I',.-......... -,:ii]I —.:::: -... - - - -,,, I I ".",."",""".,,,.,.",".,.",".,,..",.,.,.",..'..'..............I.:... -............,,-..,, —...., I I I I I. I I I "I I."",.,."" 1. I I I I. I I I. 1. 1. I.. I. I.. I I I I...I.........-...... — ,,-%.....::::::::7::,..:: ----- --....-,''''I'll''....... ----------.................I...-..........-.I --............................................. 1. 1. I.. I.......... - ... -- --... -.,, —,, -,''''I'll, 11 ''I'll'I'll''''I.111,Ill''''I'll'',.'','',.I.............. --.....-... ''..,..,..''.''.'',.,.,'',.,,.''..,.'',..''I... 1. I.. I.. I.......... -... -... -,-.-..................................................................... I.....,.............,. - ii:iv::i:;i;i;i]:::].....,..''.''.''..''.'',..,.'',.,''''.''.''..''.I.... I —..................... I.. I................................................................... I —,.........::....... - I.... I —,.... 1.1.1.11,11,11111.1.1,...-...-................................. I-.................................................................... I..''.,.................I.I-.. - 1. I11.111.1..... ---.. —., I.... -..-.-.-...... I I..... - - .............. -....................................................................... 11.1.1...........- I.... I I''..''... I.. I... I.. I... 11-1.1-1.111.1-111.11..... I.. I....................... -.................................................................... 11.11.11...............I-111-.............-.... 11.1... I I... I I......... I.. I I.. .....................11.11, I I I I I I I - 1. 1. I I I I I I I I I I I I...................................... I'll.,....... 1.11,..................I.........I.... I.-I...-....-....... ---..... ---------................. -................. 1.11,11,111,111,1111,'ll''''I'll''''Il. i.,..-,. I!,............ 1.11111111111111.1111....I.............I.......I...I..............................................................,..,................................................A,, ''I'll'I'll''''I'll.,'ll''''I'll1l.11.1111,'ll'I.................................................................... -,''''I'll, ''I'll'-, 11.111.111,111,..................I.... I.. I I I.. I..........................1.1111,-I.,.,...,.,.,......... "'."I",......-,..,,.-A:iii'...,.,.,..".I............ 11.1..1.1111.1.1.11, 1.1111,.............................. I..... 111.1.................................1.111,....-.................................................. 11.''..'',...............................................- 11.11,-.............,.................... I1.111, I".""'I'll ]-.-.......I ''I'll-:....I, -. —... I..'',,,I — I I I..'',..................,I.1.111,.,I II'll, - I.... I...I...... I"'I'll,:11,! - "I ''I'll -111.1 ''''''I'll -111. I,,-1111 II'I'll" I-I'l -.1"I", I'llI]:i:;;,, -,, I'''',....... I —',',,, I'll "I'll'' -1- -111 I''I'll...I'''' 11-I'll — -''I'll...... 11.......I'',I''I'll. -111.... -I'llI-.....-111.......... -.1I','- 11", -111..:::::::::: I...,:. I.:.::::"."', I 'I',.,.,.,..II — 11 1.11 11., 11... -.: I"Ill, 11 'I.". I —..-,.-.,-4, I.........-,........--.... --.-..1 -...%.. —. -Mi -II -"::::::. — 11 -''I'',-%.., —I'll, 11 -:,-..,.,... — 1:111,...,I., -.-.1 11 -11:::::::::::::::::::.., -I.xxxxx-I -.-.1 — 1..111 -—.I11 —.'.-1-1,11,...... -1..-. — ---. .... --,-.....- II.................... 11 --.. ---.........-... I, —.......... I...,.....-.... —.. 1-1.111 ---... - II-.....-............. I,....... '', ---... I -... 11................. I11 1.1.111. 1.11,-.....-... ---........ 11....... 1, 11.-""""",...'', "::::::::::::::::::::::::::::::::::::::::::: ....... 41,11:::x]]xx:::.-,..I,',",""., I'-1...11 -1111. 1.11.-,........1.]:xxx::.%%1.1'I...I',...,...,.,.,.,.-.I 11111-1 11.1 I - — 1111''........:.:,:.:.:.:: -.......-11.1.1.111.1111111.1'' -.1111.... ---............... I —, -1- I'l-I I....,... ''....-I.I.I.I.,..... —...-.....:.:.:.::::::::: 1:I-1.11.111..%..11111 —. I.%....... I..., 11... 1.11-1111 11.1111,.....11,11'.. -111..........-............. 11...... I....1.1.1''.. 1.11,11,.................. 11I.111,-........................I I —.,.1.1-111 1-11111.... II.....:I...''.........-......... .III....''... 1.11 III''Ill","...-..........-.... '"'I.,".,]:: I.1 ''I'll.-, 11.111, I'll, 11-.111:-....... —.......... III-. 11: -%- I1.1,.....I...I.I.I.,.....1.1".1".1"'I".1 III...11,... - I1-1.1.1.1...1.1.1..''..'',''''.'''',.,.,,.,:,..................,:"::::::: 1-1....%'...I...I.I., I..".",."",...... I.1.... 1.11, ''I'll, -—. 11 11., I.........I....... ''I'll'',..........., —,''''I'll'', 1 -11 ''I'll,......-....... I.. -................. 11.11....,ilI1111111 —. 1.1-11.-,.,...,.,.,.'I,.'I 11.-............ 1:.'. I11:::::::::::::: - 11.. ".1-1..::::::::::::::::::::::::::::::::::::.."... I11.. 1''.. "....'',........ 1.11.1 ",...".""","'"'".,.""""".",.,",...1,I —,.,..,.......",.,.,.","""",""",...",.,",'',"","', - I —.... 1.11111-111111111111.-......................... 1: 1111".,....'",........,.,.,.,.,.,".,.","."",.,.,"""'',.,,'I'l::::::,:::::::::::. ::.1.... I..''.'..,.,.,.,.,".,"".."""".,.",""'""'I'll.11: ",., —,...,.'I,.,I'I-'.,.'I,.—.1 I.......",.,"".,.",".,,""..:::::::::::::::::::::::::::::::::::::::::::::..'.I:1"I...'..''',,.'...,.,.",.,.,.,.,.,."",.,.,'''''''''',:::,,,::::::::iI 1111- 1.11, — l-1.1.111.1.1111.111111. ,,............................. I1-1- 1.111,-....I..... ---—,.........................., II11111- I — -.1... 11-111111111111111............................... 11-11 ''I'll''...1 I.....I.... --- —----—, 1.I.................................... -I ---..1111,... 11.11.11,111,111,111,'ll''''I'll.11......................- ''''..... I. 111-.111... I..'','''','''''''''''''ll''''I'll''''...............,, -, - ''. '''' I- -.111 "..,.",".,'',."","."'"""',"'"'"'"l", "I'll, "',',,"'-j I.111''...1 I... I..'','''','''''''''''''ll''''..'',I II:111''.,.,.,. I,.,.",.,.,.,".,'"',11"1.'"""'''"."""I'l,"''''..''.'',f............................ III."''"".,..1.11.."",.",.",."",,""I'l,"""'"'"""I'l,"'"',,,.i I1-111"'.'. ".-II,'..""."",""'',."'I'll'I'l,'''"',"",l"."''ll,'''':, I1, -11""'.. "...'',...",.."""""'""',",""I'l,''""''''"""I", -— I I%11 -.-. --- —- ''I'll 111111-.. I.-.................... ''I''',.. I..'',.... 11.11,111,111,111,11,Ill''''I'll'','',.. I'll.-.....................''I'll. I-... - I..,.''.''''.''''''''''ll,'',.'','''''''''I -I% 11.1.%.................,,,,, -, - -'::. - I ----. --- —-- I................ '', '''' — 11- I...... 1.11,11,111,111,1111,...............,. 1''...................................... ''I'llI-1,111- I.. -.-. —. --- —'''' ''..-..................... '',.%1.1.1.1.1.11 -1,11-,",.,.,""".,"'"',""I'll"'',.","",",.,""'"','"',"I'lli...1111. "....,...,..".,.",.,"'"',",'I'l,'"""''"'"","I'll''''II.". ''I''.....I.... -. --- —--—,, -................ '', '',,,,, '',.. I -,-..... ----------- I'll,.....................- -" II ... "...,...,.,.",.","""","I'll'I'll'I",":::::::::::::::::::::::.". II-", ''.. I......... "',',,-"""-"'I', 11 A 11.1... -.-, --- — -........................ ''I'll, I.1,.1.1 -.1.1-1111.11111111111111111... ''I'll, I-........................... . %,::::: I I I I I I I I I —.-k II 1'1......III.I" I..:,,.... 1:1III11.1" I111...-I''..'' I —..........11.1.11I-.1 -...,-11111 1I- "Ill.,.... ''I'll''..... - -,..... I, 11 -,''''I'll I II..1111111111,111 I II 11I.II IIII-11,,,I III II. III.IIII- I-I,..11,1111, 11 11...-I1,II-I...1111"'I"I'll "I'll -, I 1.1I,III'I'," I....-I''','', ''Ill-,''..I ---II I1.11 11........................ '',,''I'll....I-1.1.1.11,11,111,1111,".""",I I I 1.1,-I I - 11II II III -........... "I'll,:.-.................-,-I 11I-...I... ''I'll'',I ---II -......... - II1., :,,::::::,.., 11,I1.... ''I'llI11 - I1.III.I..11,,,I, I 1I 1I-I-....,;:;::.1I11 '-.: i,,I.g.....-,,::::.11111II I.. ,,Ppww-.-. —.. --.11 11,111, 1''..., _-..,,.,-,.,.,...... ---.....-....... -.......... %_....... 11.11:Iw —. —.-I... --- I ''.."'.'.1"'I"'I'l......... ..........".."""".,.....,.,,....,...""......... ''I'll,.....-.........1-1..... - --—, I.....-....-..............................111.11,-................................... —.-. —. —....-... — I-I'..'','', ''I'll'', ''Ill., 1 I I 1 11.11.1... ''I'll'',........................-........I.-.1% ''I'll" -, I I I I I 1,111, 11-111 I 'I, "I"I'll"'I''.111111 "I', "I.I-.'I.-'I. —..... ''I'll.......,,, - ........ ''I'll '' I I I I. 1 1. 1.1 11 -,''..-,, 1, - - - -.- -.... 1.11, kw-,.,., --- —-- I IIII-.1". 1.111. —,"'-,"" —:: ''I'l.111",""""""'',11.111. 111,111,111,11............. 11.11,111,111,1111,I.1 "''I I --—:::",:::: ::::',, - IIIIII'I.", -, I — ',,, .-,.: I -k ,.. -,''''I'll, I I. 1..''I'll,,,.... ---.... ''I'llI11II 11.11,111,I,......,.,.....,...",.,.",",,,, I II"II::::::::::::::::""I', I I I 'I',"..,."..,. I I I I I II..........,..::::;.,......., -,... 1.1111.1...... 111.11I-I,''I'll 11-111111... -—.. ''I'll'' 11. —..................I...11111......... -—,I —I II''I''II — -..'',''.... ''Ill-.-. ... I III-. --- — 1''.., -.11.1111 - - 1-11,I....IIIIII'll.,I.............. 1-1111...........I.11,-.I I-1.11111-....I.11111111, -......''I''..I-IIII ---... 11-1111............ -111.1111111 111.1,..I11.11 11 11111-11........... ''I'll....I-.1I''I'll'' I.1111......11.1.... -.1.1.1.1111 1.I-........... 1. I I.,.,...""",""," I I. I I I I I I I I I I I1.11 I II11I-II.1.1111.11 1, :::::,.I........... 111.1.11,111,11,........IIII I-I11-1111.11,. 1.11,11.1.11,11,111,Ill""Ill.,"""""'ll""I'll''I:X- 111 ---- I 11 11, 11 I,,"-, I::::::::::I:-...... ----— I-,-,,I III --- —, ''I'll'',,11, 1-11111, 11 11IIWI...... -' ----, —II —''I'll ''Ill.1-111 I I I,,I.... — -. --- —--- I, I I I I II-.... I-.II I 11 I 11 - - -.-,- I...... 1. I..'',''''III,,,...................... 1.11,11,111,11'', 11,111.1111,,I I-,.,. 11111.1 —... 1.11.111.1.111,..... -1-1-1 —.1.......... -.11111I.....,.............. ---.................. --.... --..........-...... -,''''I'llI''I'll, I.-II,''I'll.,...-.... 1.1111.11........................... - -............. -:,.",., -.I.1.1I.I.I.I.I.I.,"",,.,I - II I-"I',.,...4 I -II..:..,. "."",.,.,...,......,.....,.,.,.,.,.,.,.,.,.,.,...,.............,..'...,.,.,....,...........'......,.,...,.,...,.,.....-........-''I'll''..... I'll, I —............I....................... --....-.........I.-I.I.I.I.I...1.1.1"... ',' ---- I ', —,',' ---IIIIGUII''"I'll,,-.111-1,.................................. --......-.............................................. ",.,".,.,..'............,...",.,.,.,.""..,.,.,.,...,...'...,.,....'........................,....... -::: --- —......I..,..."""'"'I''I'll'.,,-I''I'll'',-,.........1-11 -II....II- I ---------.",.,.,"......,......,.....,.,.,.,...,.,.",.,.,.,.,.,....',.,........'.................'.......,.,...,.,.,..,.,.,.,.I.I'll, %I..-I'I I II-',., I, - -— "-:,.. "".""..........,...,.,.,...,...,.,.,...,......'.,,.,.,...''...,.....'...,............,.,...,.",.....,....".. I --- —,..... 111.1 I..''... ---.........I...I.................-............... -...I.I., 11 11111111111-.....''I'll ".""",.,,.,.......,...",.",.,...,.,..,.,.,.,.,...,...,.,....................:.............. -.11 I.11.II I — l'',...................................... -.1.... —..''..'',..- I'll,... 11 I'll.II ---- -.....'','', 11.11.... ---.....-...........-............-............ 1-11.111III, --........................................................................ I.....,.,... I'll 1.111,111, "I'll",-11 IIIIIIIIIIII-......"."I"', "".",.,............'.".,.",.,.,.,.,".,.,.",.,.,.,.,.,...,...,.,.,.,....,.....,...,.,...",.....",.,...,.,.,...'.,...,........ 1. ''I'll ""I'll, 'I',.,', - II - --- '','.AZT,',,'"' ""I","...'..........",.,"".,.,.",...",...,...",.,.,"...,.,.....,.,.,.,.,.,.,.,.,...,...,.,.,...,.,.,.,.,.,.,...,.,I,,.....I.I.I.1"'.. ''I'll.11111I"".",AI I 1111,11111''........411,11,11 ""I",".,...,.....,.,.",.,.",.,.,".,...,.,.,.",."..,.,.,...,...,.....,.,...,.,...,.,...,.,.,.",.,.,.,.,".,.,.,.,..."..,.,."" ''I'll.,.- 1, I'-., I'll. ''I'l''''I'll"........."""I ""I'l,""...,......,.",.,"".,.,.,.,".,.,.,...,.,.,.,.,.,.,,.,.,.,.,.,.,.,.,.,.,..,.".,.,...",.,.,.,.,.,.,.,.,.,.,.,....... I'll, 1.11,I —,''''I'll......-I'll,.... -.1....... 11............I.....I..........I..... I.1-11.........,-.... 'I'll""...,......,.,.,.,.,".,.,."",.,...,.,.,.,.,..",...'.,.,...,.,.,.,...,...,.,.",.",.,.",.,.,.",.,.,.",.,.,,. 1.11,....-.111,.................................................................. I...,.,., ".""', ""I'll."'I"', II""." ''I'll ''I'll, I.'"',............,.,.,.",.,".,".,,,.",.,.",.,.,.,.,...,.,.,.,.,...",.,.,.,.,.,.,.,.",.,.,.,.",..."",.,.,.,".,...-.1.-..-'I ""I'll.."'."",I"".""I''I'll, 11.111"".,.........",.,.,".,.,.,.,.,,,.,.,.,.,.,.,".,.,...'.,.,.,.,.,.,.,.,.,.,".,".,".,.,.,.",.,.,.,.,.,".,".,.,.. —.",."", ''I'll.,.'I,.,II""."'. I II I I - ::!$j$ ''..........l.,.,.",.,.,.,."",.I.I.I...''.''.,...,.,.,.,.,"".,.,.,.,.,"".,.",.",".,.,.,".,.,.,.,.,.,.. 1.11,...............I... -11....... 11.111.1................................... .....-''I'll.1.11I....I''I'll, —........................... 1.11... -—,...I........ 11 '',,......... I I "I'll -......''I''',....,...,.,.,.....,.,.,",,.,.,.,.,.,.I.I.I..."".,.,.,.,...",.,.".,.,.,.,.,.,.,.",".,."",.,.,.",.,., 1.11,.......................I..... ''I''....11,111.11,111,111,11,ll,'''I ........ I'llIII.....I''I'll''''I'll,-...............................I.... 1.1.11,111,........''.-..... —1111.....I-.11,111,111,111.11,1111.11,11,.................................. I'll, 11II..... I ''I'll'',.................................... 11-..... I'll-............... ''I'll -11111 I II.....II- I, - 111111111:]],.-.xx:]...,.,.,.,.,.,.,.,...,.,.,.,.,...,.,.,.",.,.,.I...,.",.,.,.,.",.,"...,.,.,.1.1.1.1,"",.",.,.,.I.I...... ---................-I....11.1111.11,111.11,11'll'.11,111,111,111,11'll''''I'll'.........I''I'll -1111 II.....II —I'll -.......-........................................................... I...-.................-''I'll... I,.... --.... 1''..''.'''','''''',.,'','ll''''I'll''''I'll'',ll.l.......... ''I'll,... I'llI...-... ---......... 11..................................................,'ll'',..... -,..... '- -:............... ''II.Ill''''I'll''''IIll''''I'll''''.''''. —..*,,, - - - - - I ''I'll........,...,.",.,.,...,.....",.,,.,.,.,.",.,.,.",.,.,""".,"".,.,.,.,.,...,.1.1.1.1.1,1.1.1.1.1".1.1".,.,.......,.,.,.,.,I'- I',",- -."'."'I'lI:-x: ------—.:I:I-.-..... '.,.,.,I'.,-,."'I"I'll,....'."'I"'I",I... 11I-.....- "I 1 I I'll", 11 "I'l.'.........'.,......,..."...,...,...,.,.,.",.,.,.,.,,.,.,.,.,.,".,.,.",.,.,...,.",.,.'...,...""",.,.,.,".,.,..".,.......I..... 11 I'll,... ''I'llIII.... —:,;........I I 1 1 I ''I'll'', -.......''.......,.....,.,..".,.......,.,.,.,.,.,...,.,.,.,.,.,.,.,.,.,,,.,.,.,.,.,.,.,...,.,.",.,..".,.,.,.,.",...,.,.,.,............. 11.........-.............'', '', 1-11-...-......-11.1111,.111.1111 I I ----I.............. - ----—.-.. --..-.................I............I...I..............................., —.............. 11 I 11 I,, ''.-.,.,, 111.. - -1.1.11...............................................................I1.1111, -.111111I11 11 11... - I'',I.............."....................,..-...-....... -... — 1111 ''I'll, -.1.1111I''I'll -- I I"-I - - - 1.11,11.11,11, 1.11''... - -.................I.... I —......I... -.1............................ - I......"-.1......................... --- .................................................................-...............................................................,.,.,...'I'I'I'I."I'll".-,.,.' —I""I'll''....-.11" 1 II'll I.....,.,.,.,...,.,.,......,..,......., I "I I. 11,111,11'1,1'1'::]X:iii................................................................................. 111. 1.1111, -.111111 I I''I'll,... - -.1..............-...............................-, -..................:..........,...,.'I,.,I'.,I.'.1,11"',..-,.,-'I - I-',-''I'll,..............'....,...,.,.,.,...,.....,.,.",......,.,.,.,.,.,.",.,.,.,.,.",.,".,.,.,.,.,"".",.,.,.,"..".,.,.,.,.,.,.,. -........... —, -1.1111 I::.,,'.....""",.,.,.,.,.,.,...,.,.,.,.,.,. ''...''..1.1.11.1....... ''I'll-......-..........I....... 1.11.... ----.... 11..........................:]x x::: -....................,, - - -.......................................................... -..,.,.",.,.,."" I'll',",..-'I,.,-I.I I -.'ll''''I'll,.......:::::-" "I'll, 11-I.-...........,.,.,.,.....,...,.,.,.,.,.,.,.,.,.,.,..,.",.,.,.",.,.,.",.,.,.,".",.,.,",..,.",.,.,.,...,...,...,.,.. 1-11.1.......... 1-,::::::::::::::::::::: - ,.,.......,.,...'..,...'...,.,.,.,.,.,.,.,.",.,.,.,.,.,.,.,."",,,.,.",.,.,.,.,.,".,".,.,.,.,...,...,.,..,. —.,.,.,.,.,.,", 'I',",'','', -1111111 -, I ],.....".,.",.,",..",.""",.,...,..-I....-''I'll,... I'll,-I I.............-... -—,... I'll.,... 11...-......-.................,............................................................ -..,.,.,.,.,I'.,.'I "I',"'', —,.,I'-I I-,",.::I,.....'',.,"""".,.,."",.,.,.. -I'll, — """I'll......,...,........'....'..,.,.,.,...,.,.,.,.,.,.,.",.,.,.,."",.,.",..".,..".",.,.,"".,.,.,.,.,.,...,.,.,..,..I...I.I...I.I.I.IIII.11111,,I 11 I I......"",."",...",.",.,.,..., 1 I ""I'll"''::",:,,:,,::::":""",''I II............,..,...,.,.,'...,.,.,.",.,.,.,.,.",.,.."".,.,.,.,..".,.,.,"".,.",.,...,..''.,.,,"::::::::::::..........,.............,.,.....,.,.,.,...,...",.,".,,"",.,.,.,".,.,.",.,".,.",.,.,...,.,........ -.......... ''I'll'',...... I'll, I'll, I I I..-I........... -111I -... ''I'll., 11......,,, -...............1.111,111.1 -----....... I —......-.... —I.I.,...,.,.,...',.............. ''I'll'',.... ---- I - --.... 1.1- ---—,'' -''I'll........I...................... 11 -—.-.............................................. -.....I...I...... 11.1.1111111111111 I --...-.... I, --- —---—,I. ---.......-....I...-...I.....-........... ........, '''',....................................................... -........I11.11,11, 11 I I -." I I.,... 111. 11 -,''''I'll'', I.-,...-........ 111.1 ----........ I —.... I..'',..................., —..... 1.111,11, -.11111111111 II'll. "I.11,111,111,111,111,11'll""I'll""Il'I'l,"I...........,.,...'...,.",...".,.,...,.,.,.,.,.,.,.,.,.,.",.,.,".,.,.,.,.,.,.,...,.,.,.....,.'...: I.",."'-...,.",- I ---".%, I'll'Il""I'l""I'l1,11,11",""."'"','I'l'I ............,."...'.,.,...,.,.,.,.,.,.,.,.,.,,,.,.,.,.,.,...,.,.,.,.,,.",....,.,...,....,. 1-11I...... —.... --.......... 1.111111111111111111.11111.11...... I.........I''I'll, -.1111 I I -- I -----------... --.........................................-.-....... ''I'll......,. —...,.,.,.- -.1111111 111-1I1. - - I I I'll, 'I',"". -- ---------—.-, -.....-............. -—,................ —.... -.......................... - I -------------....... --..........-............-.... -- —.111....1.1.... ----. ----.-.........1..... -.11 I 1 I. ---................................. -.-....1.111-.....I.......... 1.11........ -— I.-.....I... - .... ----...................................1O....I..."'. -.."'." I I I1-1 I... -.1-1.1111111.1111111111.111....I..... ... - -I.. ---............................... ---.......,........ 1.1111.1-11-11.11-1... --.0,...... I.I'...,.,.....'..,.....,.,...,..............4.I.I.I......"'.", 11 111,11- ---............... -.11111 11 1 I - I I 1......-............ --...... -.11111........,, -1-1.1... 1.1.11.11.... --.......... -.11111 - I'll,''.. I'll,................. ---..II ".......,, - -.11.1.......-.........-... 1.11,11,1111'..'', --...-........ ----...... ''I'll'',....... -, -,-...I... 1-1-1.... —. I............ ---—,......... -----..... P.......... 11 ''I'll'' -.1111,. —... 11...-... --.... I..... --........... -.11111 - 11 1 I I. 11 11.....-...............''I'll''........I - ''. -. :*: 'I.""",.""".-."'.", 'I'll 11111-1., -......,..,....'....,.,.,...'......., ',.I.,.,".,."""... -,''''I'll —,,:.... -.1.11 11 - 11 11 I. 1 I I: -.1 ''I'll'',... I'll,. —... -.1........................ -.11 11 11 11 -- 11..11.11,11, E.......,.,.,.'I,.,.,I'I'. -.11111 ''I'...'', -.... 1.1.11-...... - 111-1111..............1111I. ---..... I'll, 1.1111- I I......... I'll, 11 -.....I... - -11 11 liiiiiiii! :::: ''..''I'll:-:-:-:-,:-:-Xiii! -..... -1 ''I'll ----.,-::::::::::-::::::::1.1.."... —'...,.:: -- - -----------................... I.I...''...,.,..,...,.,.,",.,,..'........... iiii;i]: I.%I.I...I.I.I.I.I.I.I.::I ---....- ---.... ............,.,..........,...,.....,.,.,.,...,...,.....'...,...,.,.,...,.,.,.,.....,.,.,....,....,."",...,.,.,...,...,..,.. -—. ---... ----... --............I...I............................. -—. —......................... -..''-,... 1.1111.111,11''..'','',.''..,..,.''.,.I....... -- I-.1 —. —. —... ---..............-...............................................-........ --- -- -.11......I.............-.................................... 1.1.11,111,111,111,11l''..'',,.''.I... 11-........ --.......... ---... -----...........-........ --....................................................-.... ----.......................................... I'll..''..,..,..'',.,.,''''''''''.''.1.11.1111.1.11.1..................-.......1-............... ---......-.............. --........................................................ ----........................................ ''I'll.......... I.I..'','''','''''''',.,,1.1.1.1.11........................ --..1.............................. 11......................-............... 1.1.111,...................-...... ---- -,........,''..''.'''','''''''''''I'll'', ------—,-. ---........................................-..... -:-:.-...............................................I............................. 111.11......................... I...,...11,11'..'',''''.,..'','''','''''''''''I................................................................................................................................................. --................................. ----- I............................................................................................................. I — ......... I.'''':::,:::::::::::::::::::::::::::::::::::::::::::::::::::::::.'...,.,.,...,.,.,.,.,.,.,.,...,.,.,...,.,.,.,...,.,.,.,...,.,.,.,.,.,.,...,.,.",...."""",.,.,.,.,.,.,.,.,.,...".......................................I..............I.... 11.1'... ---- ''I'll'I'll''''I.111,Ill''''.,.'',''''''I.............................................. ............................................................................................... ..................................................................................................... — .....:''....",.,".""""""""I'll",".",.,."""",.",".,"".,.,.,.,.,...,..,.,.,.,.",.",......,.,.,.....'......,.,.,.,.,.,.,.,.,.,.",.,.",.,.,.",".,'',.,.,".,.,."",".. -....-''I'l''''I'll''''I.I.1.1.111.1.11......I.........I................. 1.1-1.1.11 1, -11111111111111........................................................................ 11.... --......... 11.11,111,111,111,11, ''I'l''''I'll''''I.I.1.111111.11.1.1..I...................... I..'',.....-.... ''I'll''...........................................................I.... --............... 1.11.1111,......... ''I'll 1.111,11I.I.''..'',,..'','''',''.''''I...........I.......... 1-11.1.111.1. — l-11.111111.1111111 ''I'll'',........................................................................ --......................-......... ''I'll'.11,.11,1111.111.1.11111.1Ill''''.'''''I...I...I...I..........I....I........... 1.111.1111.1.11.11,11,..............................................................-........................ ''I'l''''I'll''''I'll'll1l.11.11,11.1111.1111.11,11,111.11,11,11111..111.111.1111'', -,''''I'll..''IIIIIIIIIIIIIIII ---------- -.............................. --------—,.................... ''I'll'-,........................ --- —----- --- I -,''''I'll'', ''I'll......................... -,''''I'll,........... IIIIII.-.11.................. 111.11.......11,111,111,111,1111,'ll''''I'll'I ---- 11-1111.1111111-............................................... 1.11,11,111,................. -,''''I'll'','' 11.1111, -,''''I'll - ''I'll''-....................... ''I'll'I'll''''I.11.1'll''''I'll''''I.... 1-11.11........ -,''''I'll... ''I'll'' -,''''I'll'''' -------— l ''I'll........................... ''I'll.....................-....... -.11IIIIIIIIIIIIIIIIIIIIIIII

-58 -tend to decrease the sludge filterability. This method has been used successfully on both a laboratory and plant scale(26)(100) However, the possibility of a gas-metal effect on the electrode potential as discussed previously and the possible effect of the gas on the digestion itself led to the elimination of this method as a possible source of agitation. The effect of gases on bacteriological cultures served as a guide for the possible gas (29)(40)(54)(108)(109)(110)(111) effect on the digestion process, Some method of mechanical mixing was therefore considered. The centrifugal type mixer was given primary consideration since it would be designed to develop a lifting force in the center of the tank which seemed to be the most desirable type of agitating action. The use of vertical vanes arranged in a circle proved to have desirable characteristics. The spacing of the vanes had to be such that no plugging would occur and yet the desired action would be obtained at a reasonable speed. After trials with single and double units on one shaft, vanes on the top and bottom, and large holes in conjunction with vanes, it was decided to use one single bottom vane unit placed near the top of the sludge. The best design is illustrated in Figure 3. The mixer consisted of a 1-3/16 inch OD, 7/8 inch ID rubber tube with four vanes cut in the tubing. The vanes were 1/2 inch long and 5/16 inch wide based on the outside circumference. The mixer shaft was inserted into a hollow spindle stirring unit with a three step pulley on top.

-59 -Dental chord was used to turn the mixers in the three digesters at the same speed. This type of belting was found satisfactory after the pulleys haid black tape put on the riding surface for additional grip. The final agitation setup included two dummy pulleys, three spindle pulleys, the motor pulley and three short chords as shown in Figure 6. The motor used to drive the mixers was rated at one-tenth horsepower and continuous duty. Although it was single speed, the use of a pulJley system as well as a variable transformer allowed variable speed control. The speed of the mixers under full load was near 600 RPM as measured by a stroboscope. Oxidation-Reduction Potential Apparatus The basic components needed for a potential measurement of a solution were the inert metal electrode, reference electrode, salt bridge and potentiometer. These components were similar to those used in the pH measuring system except for the replacement of the glass pH electrode by the metal electrode. Although the glass electrode has been universally adopted as the best method of measuring the pH, no such unanimity existed in the acceptance of the metal to be used for the oxidation-reduction electrode. Platinum has been used by the majority of experimenters with gold in a seasndaryposition. Investigators who have considered the use of both metals did not particularly agree on which gave the best results.

e l> 1X,..S;: iS ~~~~~~~~~~~~~~~....... e ~~~~~~~~~~~~~~~~~~~~~~.......................S:::::::............ _:.. s ~ t.Ei44: E:ii E i: i................... - S''"'.'lS'.'.''."'"''''S'i00"'''i'"""4..................................:i':::.~I ~ ~ li~lr" ~~': ~ 0 ', ~~~~~~~~~~~~~~~~~~~~~..s...'s.......,..... ~2 O S ggg~~~~~~~~~~~~~~~~~~~.g.:g............',ss.S ++'S."0.....'..:,:j. i. E iEREERE............,:..... g..:ss.s:.s s, EiE.E.i. #TE t ~ ~ ~ ~ ~ ~ - i-:~ _~~ ~i s~t00: E+++...~~~~~~~~~.......... sss++ i:::::::400:::::::::: i 4l g sd:X1440.................................::::~~~ilii:~~~lill~: ~3:::::i~I1~~~~& 8~-:~ ~: III:::::-kiil':~~:~1. ~~~~~~~~J1 ~ ~ ~...... "t~~~~~~~~~~~~~~:: ~ ~ ~ ~ ~.......................i:,::::::::::::::: ~~~~~~~~~~~~~~~ii;::' ~~3::::::::::j~~~~~~~~~~ ~~lii~~~~~~~i~~i~~i~~......................iiiiiiiiiiiiiiiiiiiiii i~iiii~iiriiiiii iiiiiii~~~:~i:~~~~~% ~ O ii~~~~~~iiiiiiiiiliiiil: "~~~~~~~~~~~~~~~..................... ii —iii lii ii r-5::::~~~~~~~~~~~~~~~~~~~~~~~~..................... WA~~~~~~~~~~~~~~~~~~~~~~~~~~'I~rrii::::.....................iiii~i::lii~:,,.........................illiiiilil- lilll:lllll~lllllli lll~.'Blilillli~~d~l::?~~ 8 ~lliiBB'iiliiiil: Fs ~iiiiiiiiiiiiiiiiiiilii~~~~~~~~iiiiiiiiri........................................ i;: i;:::::::7 1~~~~~................. * ~ ~ liiiiiiiiiiiiiiiiiiiiiiiiiiiii.....................:'Biliiiiiiiiir'iliiiii~~~~~~~~~~~i~i:5i llll~~~~~l:-:: ---':i~~~~~~i~~i~~iliiiiil::................ ~~J -:::::::::::::::::::::::::::::::::::I:I: I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....... l-....... Y~~~~~~~~~~~~~~~~~~~......... E3 r~~~~~~~~~~~~~~~~~~~~~~~~~..................................................................................-' l'~'::::: Y~~~~~~~~~~~~~~~~~~~~~~~~~~~.....................:liill llllll illlll llllll l::..............rrri!?llilr ii ii..............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iiilj...........iiillli~li~li ~iiiiis~ i~ cC ~ ~ ii~iiiiii:~iiiiiiiiiii..................'iii ~ ~~~~~~~3" ~ ~ ~ ~ ~......................................................iiiiiiiii.................i, i:::s:::::::::::::::: i::::::'':'':i7,: j iijiiiiiiiii~(i~ii~)ijl. ll..l.l:::.::::::::::::::::::::::::: ~ ~ ~ ~ ~.......... ~i::::::::::::::::::::::............ ~ ff~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.... iiiiiiiii~~~~~~~~~~~~iiiii i iiii~~......... ''' '''''''' ''' ':'.............................. ~~~:~::-:~:~ ~ ~ Cicg........ rxliliisisr ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~.............::::::::::::::::::: ~ ~ ~ ~ ~..................:::::::::::::::::::: ~ ~..........................................:::::::::::::::::::: ~ ~....................................::::::::::::::::::: ~~~~~~~~~~~~~~~~~~~~~:......................,alrlali lal i~~~~~::::::::............................................................................................................................................iiiiiii iiiii liii::::::::::::::::::: ~ ~ ~ ~ ~ ~ ~.......................::i::::::::::::::: ~ ~................... i::::::::::::::::::~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..........................................................................................................................................................i:: i::::::::::::::............................ ~~inlccli l Is:~:~:~~:~:~:::~~'t'''::::':1:1:1:1: -i:::......................................iiiiiiii~lii~iill ~ ~~~:~~~.:: ~ ~ ~.:.:::::::: ''''''''''''''''..........................................................

The most thorough investigations covering the.comparison of gold and platinum as an electrode were performed oP the same system by two different authors(l6)(61). The result produced two different opinions. Gold has been said to be more sensitive in bacteriological studies(54)(74), but another investigator did not find that the two metals differed in this respect (66) Although reference to the greater sensitivity of gold was found, the author did not fihd any instance where platinum was alleged to be more sensitive. The choice of a particular metal by different investigators often seemed based on the ease of the electrode construction. In this respect platinum had a decided advantage over gold due to the fact that its coo efficient of expansion was much closer to that of glass. Thus, sealing of the metal into glass was possible and the electrode metal could be fully immersed in the solution to be tested. It was of interest to note that the usual commercial gold electrodes consisted of a gold.plated platinum wire since only the platinum could be sealed into the glass. Another possible reason for the preference of platinum might have been that so-called gold electrodes were actually gold.plated platinum. This interchanging of nomenclature was observed during the literature study. Therefore, the preference of platinum over gold might actually have been over gold.plated platinum. This distinction was important because gold and plated.gold electrodes did not always behave in the same fashion, and because of the variability introduced by the plating itself.

The investigators who used gold, constructed the electrodes by fusing the gold to platinum which was then sealed into the glass. The exposed platinum was then gold-plated. However, the development of a special glass in recent years with a similar coefficient of expansion to that of gold has made it possible to construct gold or platinum electrodes in a similar manner. Thus, the advantage of platinum described above would no longer be a factor in the choice of an electrode metal. From the literature survey presented previously, the use of gold as the electrode appeared to have several advantages over platinum. The gas-metal effect was the most important. The production of hydrogen from digesting sludge would cause platinum, unlike gold, to assume a gasmetal potential and mask the effect of weakly poised oxidation-reduction systems. The presence of catalytic activity due to the metal was always a factor with platinum, not gold, The possibility of the greater sensi^ tivity of gold to bacteriological doautions, and the experimental work in this research showing gold was more sensitive to quinhydrone were other factors in favor of this selection of gold. Although the advantages of gold made it appear to be better for digesting sludge, the literature study did not provide a clear-cut answer to the definitely superior electrode. As a result, gold was selected as the standard electrode for this research with platinum being used in a supplemental manner so that the author could compare the two metals experimentally. The use of gold was the first attempt to measure the potential of the sludge digestion process with an electrode other than platinum.

The construction of either type of electrode was similar. The metal was fused to copper wire and then sealed into glass. The possibility of sealing gold into glass was made possible by a special lead glass from the Corning Glass Company with an exansion coefficient similar to that of gold. The Corning code number for this glass was 1990, and the glass ranged in OD from 6.5 to 7.0 mm. Platinum was sealed into ordinary soft glass with an OD of 6 mm. The first step in the construction of the electrode was the joining of the copper conductor to the inert metal. One inch of the inert metal was fused over a Bunsen burner to the copper. The inert metal was then sealed ovez4 a Bunsen burner into the glass. The metal was pushed slightly forward while the glass was still hot in order to insure a watertight seal, The electrode is shown in Figure 7. The gold and platinum wire were number 20 gage and the copper wire was number 16 gage. The exposed length of the metal wire after sealing was about 0.25 inches giving a surface area of about 0.0259 inches or 0.162 square centimeters, The length of the exposed metal comprising the electrode was kept small to minimize the area of measurement. This length was felt to be sufficient based on the work of Vennesland and Hanke(ll2) who found that large platinum foil with a surface area of four square centimeters gave no better results than small wire ones in anaerobic bacteriological studies.

-.6. Figure 7. View of gold and platinuxa electrodes.

-65 -In order to determine the effect of the fusion of the metals on the potential, some platinum electrodes were constructed with a mercury seal. This provided the electrical contact instead of the fused joint~ In quinhydrone solutions the difference between the two types of joints was found to be insignificant and it was concluded that the fusion method was wholly satisfactory for this work. The reference electrode or half cell used in this research was a Beckman calomel fiber type with a salt reservoir. Since this electrode could not be used directly in the digesters, a salt bridge extension was made using glass tubing. The salt 'bridge solution contact point in the digesters consisted of a short rubber sleeve placed over a hole made in the glass tubing about one inch, from the sealed end. This procedure proved very successful after unsuccessful trials with contact points pinched at the end as a dropper or an upturned end. Diffusion of the sludge solution into the bridge was the major problem in these open-end types. Rohlich(94) evidently experienced this with activated sludge since it was stated that, the liquid junctions were a source of trouble~ Although agar-saturated potassium chloride bridges have been -used widely in bacteriological work9 their use was not possible here as the electrode was in the sludge over very long periods. Possible growths or other contamination would cause muach difficulty~ The sleeve type bridge using saturated potassium chloride worked very sluccessfully and never needed to be removed from the sludge solution. In addition this salt bridge worked very well for potential measurements in a lake bottom mad.

sample. Renewal of the solution contact was very simply accomplished by squeezing the connecting rubber tube. Saturated potassium chloride solution was used as the salt bridge fluid. Although the temperature coefficient was higher than the 0.1 or lN solutions, there was very little chance of error due to concentration changes. The reference electrode was placedcin a saturated solution of potassium chloride which was connected to the salt bridges in the digesters and the electrical connection went to the potentiometer. An air space was put on top of the salt bridge of each digester to eliminate difficulties due to air bubbles in the line. This is illustrated in Figure 6. The vacuum tube high resistence potentiometer used as the measurhng, instrument was the Beckman model H-2 pH meter. A specially-made rubber stopper was used to hold the electrodes and maintain the digester pressure. This can be seen in Figure 3. The stopper was made of two sections. The base which was inserted in the digester was a rubber stopper with an opening to allow the electrodes to pass through unobstructed. The top was another rubber stopper with four holes for the three electrodes and one salt bridge. A piece of rubber tubing was placed around the electrode stem so that a gas tight seal would result 'when the electrodes and salm bridge were inserted into the electrode holder, The electrodes were connected to the potentiometer through a ten point switch. This switch was placed in a boE in which the inside was lined with aluminum. The wires from the electrodes to the potentiometer

were shielded and the shielding part of the wire was connected to the aluminum-lined box. The shielding was used to diminish the influence of any disturbt'.g electromotive forces outside the circuit, Stray currents from external sources such asthe.agitation apparatus could have caused difficulties. The aluminum lining was not grounded to any external source. This was based on two factors. White(llO) in discussing shielding requirements stated that the shielding preferably should not be connected to the earth. In the initial experimental phase of this thesis it was found that a vibrating needle could be eliminated if the potentiometer was grounded to a water pipe. But it was also discovered that renewing the salt bridge produced the same result. From this it was felt that there was a possible danger in external grounding and it was not used in any further work. Lactose Broth Feed Apparatus The design of the apparatus used for the feeding of a lactosepeptone-beef extract mixture to a digested sludge embodied the basic rem quirements of electrode measurements as discussed above. The digester was a one quart wide mouth mason jar and contained about 700 ml of sludge. The salt bridge, three electrodes, a submerged feed tube, and a gas outlet were held in place by the jar cap. The electrode leads were shielded and a four point switch was used between the electrode and the potentiometer. The potentiometer Used was the Beckman Model H-2 pH meter. The salt bridge system was the same as described previously.

The gas collection system was initially a water displacement type. But variable pressure caused difficulties in electrode measurements and henceforth, a constant pressure inverted can type was used. The difficulty was due to the pressure release when the gas collector was emptied. Between the digester and the gas collector a flowmeter was placed to indicate instantaneous rates of the gas flow. It consisted of a glass tube packed with glass wool yielding.a linear relationship between the rate and pressure drop. This was calibrated using a wet test gas meter. The entire apparatus was placed in a vented hood in which the temperature remained at 27+ 1.5~C.

EXPERIMENTAL PROCEZDURE Lactose Broth Feed The feed used was Difco Bacto dehydrated lactose broth which contained a lactose-peptone-beef extract mixtureo The ratios of the components were 5-5-3 for the lactose, peptone, and beef extract respectively, and the dry solid contained about 95 percent volatile matter. This was suspended in a minimum amount of water. The quantities fed varied from one to ten grams of solid. The quantities added were so small that usually no digested sludge was withdrawn. The digested sludge at the beginning of the test was an equal mixture of two and ten month old sludge aged in the laboratory. This was obtained from the Ann Arbor Sewage Treatment Plant which used activated sludge treatment. The ten month old sludge was considered as a completely digested sludge and the two month old sludge as a digested sludge still retaining good seeding properties. It was felt that the potential changes would result only from the feeding and the digested sludge would produce no metabolic action on its own. After the feeding, the mixture was shaken by hand to insure a complete blending of the materials. During the run. measurements were taken of the oxidation-reduction potential, gas rate and gas quantity. The gas production rate was obtained by a manometric reading and the quantity by the rise of the inverted can gas collector. The pressure of the system and temperature were checked at intervals and the action of the digestion on the volume of sludge and supernatant was followed.

-.70 -Sludge Digestion The digested and raw sludge were obtained from the Ann Arbor Sewage Treatment Plant. The raw sludge from the primary tanks was picked up at intervals and stored until used in a refrigerator at a temperature of 4~C. The temperature of the feeds was either at 10~C or 30~C. The feed sequence usually consisted of the following steps: (1) mixer on, (2) digested sludge withdrawn, (3) mixer off, (4) feed, (5) mixer on, and (6) mixer off. The mixer was on about ten minutes before a sludge withdrawal in order to insure that a uniform sludge mixture would be withdrawn. The feeding was followed by ten more minutes of mixing to blend the raw and digested sludge. The major measurements obtained were oxidation-reduction potential values and gas production. From the gas production, gas rates were calculated on the basis of a ten hour interval. The samples for analysis of carbon dioxide were taken from the gas collectors. Volatile acids, pH, total and volatile solids were determined according to Standard Methods(ll4)o Electrode Potential Measurements For the digestion studies the electrodes were allowed to remain in the sludge for long time intervals. This allowed the continuous measurement of the potential in situ so that the metabolic activity of the digesting sludge could be followed at all times. After placing an electrode in the digester, approximately two days were allowed for the electrode to come to equilibrium with the solution potential. By this procedure the potential changes resulting from the adjustment of the electrode and the metabolic changes of the sludge would not be conflicting.

71 -In addition to the digestion studies, electrodes were used for the potential measurement of sludge samples from plant digesters. Before reusing an electrode, an experimental procedure for cleaning the metal had to be developed. From the literature survey presented previously, it was found that there existed a large divergence on which cleaning procedure was most effective. In addition, most of these cleaning techniques could affect the metal surface and, in turn, the potential measurement. The method of using a weak abrasive appeared to be the best possibility, especially since this technique has been successfully used in sewage-activated sludge studies. To better evaluate the effect of cleaning on the electrode to be used in digested sludge, two methods were tested. In order to provide a contrast, one method involved wiping the metal with a wet cheesecloth and the other involved soaking the electrodes in hot nitric acid. Cleaning with nitric acid was tried not only because of its ability as a powerful oxidizing agent, but also because it has been used by many workers ineluding those who have investigated digested sludge. Both platinum and gold -eleo~trodes were immersed in the hot acid for a period of 15 minutes. This was followed by a thorough rinsing and placing of the electrodes in distilled water for ten minuteso Upon inn sertion into the sludge the electrodes showed wide disagreement. The -trend to the negative or reducing potential was initially much more retarded for the platinum indicating more of a residual effect from the oxidizing agent. (79) This was also found by'Kendall and Loewen with the qasteine system. Since both the platinum and gold electrodes appeared to be sluggish, it was concluded that this technique would not be satisfactory0

-72 -Electrodes cleaned by wiping with a wet cheesecloth and soaking in distilled water produced good results. The agreement and reproducibility that was obtained in the sludge indicated that this method would be satisfactory. Although this technique was very mild in comparison to most cleaning methods used, a similar method was used by Dixon and Quastel(61) who found that washing the electrode in water was sufficient for tests in the cysteine system. In the discussion of the experimental work, the changes in potential are usually referred to as rises or falls. The former signifies a change in the more positive or oxidizing direction, and the.latter signifies a change in the more negative or reducing direction. 'The potential values are expressed in millivolts referred to the calomel electrode, Ec. If the hydrogen electrode is used as a reference, Eh, it is specifically indicated. The potential measurement is identified by various terms, all used synonomously. These are: oxidation-reduction potential, electrode potential, redox potential, and potential. Identification of a particular electrode is used when needed, This consists of a number followed by the letter byL the. let.ter G'.or P. The:former r.e.pr.ese'nt's gold and the lattef, platinui,

BEHAVIOR OF ELECTRODES IN QUINHYDRONE In order to check the condition of the electrode after construction and during the life of the electrode, an oxidation-reduction system was sought which has been thoroughly investigated and would provide a reference potential solution. The system had to be stable, reproducible, well-poised, result in rapid electrode equilibrium, and have a large capacity to resist change. Quinhydrone was such a solution. Because the relationship of potential, temperature and pH of this medium had been thoroughly investigated and published widely, it was thought that the use of this solution would add no unknown variable. The solution was made by adding a small amount of quinhrdrone to a buffer solution of known pH. The actual quantity added was not determined due to the supposedly small error resulting if the solution was not saturated. This would amount to less and 1 my at 0.1 of saturation, or 0.377 grams quinhydrone per liter(ll5) Between tests the electrodes were wiped, washed and stored in distilled water. Different types of gold electrodes wtre first tried to see whether the history or handling had any effect One was freshly made and sealed into a special glass, another was approximately five years old and had been soaking in distilled water for at least two years, and the last was a Coleman gold-plated platinum electrode. The first and last type were fully immersible, but the second type did not allow full immersion of the metal.. After wetting with water and drying with a cloth, the electrodes were inserted into the q-inhydrone solution and exhibited excellent agreeiento -735 -

The new type of electrode, so-called as the metal was bonded directly to the glass, was slightly cold-worked by bending. Insertion into the, solution resulted in a slow adjustment. Not even in 15 minutes had it come to the correct value. This effect of physical handling has been shown by others. This factor was considered in constructing and handling electrodes, Comparison of the gold and platinum electrodes in quinhy4rone resulted in a surprising lack of accuracy and precision by the platinum electrode. By repeating the procedure several times it was discovered that the potentials of some platinum electrodes were dependent on the concentration of the quinhydrone solution as shown in Figure 8. However,.this was not the case for the gold electrodes. The first test solution, I, was made by adding a small amount of quinhydrone to the buffer solution and the potential was then determined. Another small amount of quinhydrone was then added resulting in solution II. A final addition of quinhydrone resulted in a definitely saturated solution III. The platinum potential in solution I showed a wide scatter of values which were brought closer together by increasing the solute concentration. A summary of these results is given in Table II in terms of the arithmetic mean, X, standard deviation, S.D.o and high and low potential values. The sample consisted of twelve platinum and nine gold electrodes. The only solution in which these electrodes were previously used was quinhydrone.

-75 -Table II COMPARISON OF, GOLD AND PIATINTUM EIECTRODES IN QUINHYDRONE SOLUTIONS WITH VARYING SOLUTE CONCENTRATIONS Quinhydrone High L6w Solution X $ ' PO Reading Reding Electrode Number (m) (m) (mv) ( Platinum I 86 33 12.6 107 65 II 81,33 5,6 91 72 III 77.67 l.6 81 75 Gold I 75.67 1.14 77 74 III 75~33 0~ 84 76 74 The greater sensitivity of gold to quinhydronre, a well-poised system yielding rapid potential equilibriums could be significant for the measurement of biological systems which were of low capacity and often behaved in a sluggish manner. If platinum was less sensitive in quinhydrone, perhaps, but not necessarily, it would be the same in biological solutions. No reference was found in the litedrature concerning the relative sensitivity of gold and platinum in quinhydroneo However, platinum was found to be more sensitive than gold-plated platinum(l5) Investigations of the potassium ferri-ferro-cyanide system by various authors also indicated a difference in electrode sensitivity. Lewis and Sargent(ll6) first tried plain platinum foil. Later the platinum was platinized, iridized or gold plated. They stated, "The gold gave by far the mast consistent and accurate results," Comparison of the work of two other authors clearly demonstrated the difference in electrode sensitivity. Michaelis stated that a gold electrode yielded a potential at a concentration of 0,0001 M while Shaffer and Williams(ll7) found with platinum that the potential was indefinite at 0.001 Mo

476 -The general effect of the solute content on the potential was (49) discussed by Clark et al. 9) They found that electrodes of different or the same metal agreed perfectly when appreciable amounts of oxidant and reductant of an active reversible O-R system were present. But if extremely dilute solutions were used or the ratio of oxidant to reductant approached zero or infinity, the electrodes would disagree. A very practical result of"the solute concentration on the potential was seen in a statement by Hewitt(48), "The concentration of quinhydrone has little effect on the potential... "o The results of this research indicated that by following this procedure, supposedly erratic platinum electrodes would be discarded even though they might be perfectly satisfactory. However, the correct potential reading in quinhydrone by an electrode did not insure that the electrode was not poisoned. It has been found that.an electrode which appeared erratic in a sludge solution, acted satisfactorily in quinhydroneo The main value of the quinhydrone was in detecting distinct malfunctioning of the electrode metal before use in biological solutions.

-77 -110 100 Solute concentration: *e~ Solution III>II>X Each symbol represents >E So a given electrode E O 90 A 70 A PLATINUM GOLD 60 I II III I I i m SOLUTION NO. SOLUTION NO. Figure 8. Comparison of Platinum and Gold Electrodes in Varying Concentrations of Quinhydrone Solution

ADJUSTMENT OF THE METAL EIECTRODE TO THE SLUDGE POTENTIAL The determination of the oxidation-reduction potential of a digested sludge did not appear initially to be a complex technique. Although previous work on this subject was very limited, the recommended procedures consisted of short time electrode immersido. Hood(88) found that 10 to 30 minutes was sufficient with readings taken every minute until equilibrium or a reverse trend occurred. It was recommended that no agitation be used as he considered the readings might not be valid, especially for a poorly poised solution, However, agitation with a magnetic stirrer was recommended by Eckenfelder and Hood(98). Theystated that readings should be taken with one platinum electrode every few minutes for a total of 30 minutes. Initial experiments using a 30 minute time period appeared to substantiate the recommended short time immersion. However, further work disclosed that this procedure was greatly in error since the time needed to fully stabilize the potential reading was at least ten hours or mores*. Because the determination of the correct potential of a sludge was considered to be of utmost importance, a considerable amount of time was spent on this subject. The fact that the only published technique for obtaining sludge potentials was erroneous added to the importance of this study. A study of the literature of the bacteriological and biochemical fields disclosed that the electrode adjustment period, or time necessary for the electrode to come to equilibrium with the solution, used by various

X79 -workers was much longer than 30 minutes. Six to eight hours were used by Knight (62) with freshly gold-plated electrodes; Lepper and Martin (57) with special go4ds.plated and iridium electrodes autoclaved a meat media after inserting the electrodes and let the mixture equilibrate overnight; Reed and Orr(78) found a few hours were needed with platinum for a reducing agar with sodium,thioglycollate; and Michaelis(16) used four to six hours in a cysteine solution for platinum and a slightly longer time for gOld-plated plat inum A very important characteristic of practically all the adjustment times reported was that nitrogen or autoclaving was used to deaerate the media to create anaerobic conditions after the electrode was in the media. Although the work of Reed and Orr(78) was an exception, unfortunately the anaerobic conditions were created by chemical means. Thus, no quantitative information on the potential measurement of' an already existing anaerobic bacteriological or biochemical system was available. However, the adjustment periods listed above did indicate that a longer time than 30 minutes would be necessary. It was of interest to note that the adjustment period was practic cally always reported as a time interval and not a fixed time. From this it would appear that a variable factor existed either in the solution or the electrode or both simultaneously. The work of investigators who have used more than one electrode in a solution indicated that the major variable in the 'testing procedure was the electrode although the solution effect could also be appreciable.

-80. On examination of one of the many adjustment curves obtained in this research, some very sig4ificant observations were made, This potential-time relationship shown in Figure 9 was representative of the results obtained in a digested sludge under quiescent conditions, The platinum electrode, one of six used, was inserted into a sludge which had been removed two hours previously from a plant digester. By plotting the abscissa in minutes, the curve appeared to approach a constant potential value near 60 minutes, However, by following the potential over a longer period and expressing the abscissa in hours, it became apparent that the potential adjustment actually took a considerable length of time, The short time abscissa in minutes yielded a pseudo-equilibrium which evidently was what caused Eckenfelder and Hood(88)(98) to recommend that 30 minutes was Eufficient for a digested sludge potential measurement, Although the curve was still dropping at 30 minutes, the slope had decreased substantially to about 0,5 mv/minute and this could possibly have been aFsumed by them to be indicative qf the approach of equilibrium, The reading at 30 minutes was -388 my and at equilibrium it was -488o This indicated an error of 100 my would be made if the short time adjustment period were used in this example. If these potential readings were expressed as Eh, the 30 minute and equilibrium readings would be -142 and -242 my respectively. Therefore, these values could be compared to those published by these authors. For the digester supernatant the following values were given: -132, -121, -139, and -67 my. Three values were given for digester sludges -75, -54, and -125 to -175 mvy Since practically all 30 minute readings obtained in this thesis-twere within the Eh range of -60 to -160 my, it could be seen that most of the values cited above fell within this potential rangeo

An adjustment trial was made with a magnetic stirrer.-t. follow the recommended procedure of these authors(98), The results in the short time interval were not significantly different from the quiescent condition. This was also shown by other trials with agitation. From these results it must be concluded that the potential values given by these authors were not valid for the potential of digested sludge. After it had been established that no valid sludge potentials had been published, this section of the experimental work was directed towards the factors in the single potential determination of a digested sludge. This type of measurement basically consisted of inserting a metal electrode in a sludge solution and following the potential change with time until equilibrium was reached. However, -ttr. existed two variables in this measurement, the electrode and the sludge. This investigation was concerned primarily with the electrode and methods to obtain the sludge potential. Adjustment trials were done with various sludges ranging from a freshly drawn plant digester sludge to a sludge remaining dormant for a three year period in a laboratory digester. A major portion of the testing program was done in the laboratory digesters where the metabolic activity level could be approximated from the history of the digestion. The work outside the digesters was mainly concerned with plant digester sludge. Both platinum and gold electrodes were used and the number of electrodes at one time in a sludge ranged from one to twelve. Tests were run with both quiescent and agitated sludges.

-82 -The agitation tests were done in the laboratory digesters under a slight gas pressure. In this manner it was assured that there would be no oxygen diffusion into the sludge which could affect the potential( (112)o An example of the effect of oxygen adsorption on the electrode potential in sludge is shown in Figure 10o The sludge had been taken from a plant. digester and placed in a two liter beaker which was open to the air. Twe.lve electrodes, six platinum and six gold, were then placed in the sludge. Five days later the sludge was gently stirred with a glass rod. This is indicated in Figure 10 as "hand mix". The second mixing was done 23 hours later for 22 minutes with a vane type mixer similar to that used in the laboratory digesterso The degree of mixing was much more thorough as can be seen by the much larger potential rise. The potential values were averaged from six electrodes. With this number of electrodes it was possible to discard erratic values and still be reasonably assured that the average value was representative0 Electrode agreement was better for gold after the first mix and better for platinum after the second mix. During the second mixing the platinum started to drop after the initial rise of about 100 mva The potential reduced its change rate near -455 mv which approximately corresponded to a similar situation for the gold electrodes0 However, the gold rose 58 mv, much less than the platinum0 The potential rise after the first mix was also much less for gold, mvar than for platinum, 22 mvy This clearly indicated the greater susceptibility of platinum to oxygen,

The cause of the potential change to a more positive value was oxygen since it had been established during agitation trials in the laboratory digesters that mixing caused a trend toward a more negative potential. However, the mechanism of the effect was not as clearly known. The electrodes were over three inches from the surface and therefore could not have had any direct contact with the atmosphere. Thus, it must have been adsorbed oxygen which created a gas-metal electrode effect. The apparatus used by Eckenfelder and Hood(98) was of interest in this connection. The electrode was placed in a rubber stopper which in turn was placed in an Erlenmeyer flask containing a magnetic bar for agitation. Although it was mentioned that air must be kept from entering the flask, no provision was made for the air in the gap between the sample and the stopper. This would definitely cause erroneous readings as seen from the previous dicsaussion. A solution to this problem would be to fill the flask to the top and'replace the stopper so that no oxygen would be entrained. The quiescent tests were done both in the laboratory digesters and in beakers open to the air. The two liter beaker was filled with about 1500 ml of sludge and the electrode metal was about three inches below the surface, The exposure of the sludge surface to the air did not affect the potential results when quiescent. It appeared as if the sludge formed a protective zone at the surface against oxygen diffusion, Tests with quiescent sludge were emphasized in this research in order to contribute Atwoards the goal of the simplest possible method for potential determinations in the field. In addition, insertion of an electrode into a digester would be under conditions of a relatively quiscent sludge. Without agitation, the only apparatus necessary would be a portable pH meter and the electrodes,

-84 -However, the length of time necessary to reach the equilibrium potential under quiescent conditions would be a disadvantage for routine process control. Any reduction in this time would be a valuable goal. One method of achieving a more rapid electrode response was by the use of agitation (14)(75) Although the experimental results indicated that this technique greatly reduced the adjustment time, erratic behavior in both the gold and platinum electrodes was a disadvantage. It was not usually possible to determine the potential with any accuracy due to sudden changes in the electrode reading. One of the more successful tests is. shown in Figure 11. Agitation was started two minutes after the platinum electrode was inserted into the sludge which had a potential of -505 mvy. When the potential reached -528 my, the agitation was stopped. This occurred near 4.5 hours. The potential then rose and in a short time leveled off near -502 my. This value was maintained for the next 20 hours. Unsuccessful trials differed in that the potential rose much more after agitation was stopped before dropping again to-the correct potential. It appeared in these cases as if the electrodes were indicating a temporary potential due to the agitation. The potential that the electrode rose to was near a value corresponding to a quiescent sludge adjustment trial. Since only one speed was used for the agitation tests, it was possible that an optimum stirring rate existed which would eliminate the erratic electrode behavior.

In all agitation trials the first 20 to 30 minutes of adjustment did not differ from the results of quiescent adjustment. The erratic behavior took place after this period. Although agitation did not affect the early adjustment phase, it did improve the adjustment after the potentialtime slope began to decrease. It was not the initial potential drop to between -300 to -400 mv that created the long adjustment phase, but rather the subsequent drop over the last potential interval. It appeared as if the driving force became less as the potential indicated by the electrode approached the equilibrium potential. Although. agitation had decreased the adjustment time to about three to five hours, it was felt that the time could be shortened by another technique, pre-conditioning the electrode in an anaerobic media. In this manner it might be possible to overcome the existing lag in the latter phase of the adjustment period, By removing the electrodes from the sludge and reinserting them after different exposure periods to air, it would be possible to determine the effect of preconditioning the electrodes in the digested sludge to be tested, The exposure times uped for both gold and platinum were 0,25, 1p 3, and 5 minutes, The results are shown in Figure 12. The electrodes had been in the sludge for seven days before being removed and no agitation was used, The gold electrodes differentiated the effect of these different pretreatments very clearly. The electrodes with the 0.25 and one minute exposure resulted in a more rapid potential drop than the electrodes with the three and five minute exposure, There appeared to be a critical exposure

-86 -time between the one and three minute values since the potentials of the two electrodes with the low exposure were similar as was the case for the more exposed electrodes. The time required to reach within ten my of the equilibrium potential was three to 305 hours for the 0,25 and one minute exposed electrodes, The other two electrodes required a longer time of eight to nine hours. Although the platinum electrodes did not reflect the various pretreatments as clearly as 'gold, the overall effects were similar, The longer exposure times did not appear to change the adjustment for about the first hour. However, after this time the electrodes began to shown differences. Since the platinum equilibrium potential was about ten mv more negative than gold, a longer time of about 12 hours was required to reach within ten my of this potential. If the equilibrium potential indicated by gold was used, then the platinum approached within ten my in about three to five hours for the 0',25 and one minute exposures. The various exposures to air resulted in a greater differentiation in the gold than in the platinum, Although platinum was more susceptible to oxygen effects, it usually had a more rapid adjustment than gold., Sowever, the adjustment times were very close for the gold and platinum electrodes exposed for 0.25 and one minute, The three and five minute exposures provided the difference. From this it. would appear that the gold lost the residual effect of the sludge more easily than platinum when exposed to air for more than approximately two minutes, This result was similar to the data obtained in gas-metal studies where it was found that by changing

-87. the gas from hydrogen to oxygen, platinum changed slowly while gold changed rapidly(141)o Conversely, the changing from oxygen to hydrogen produced a rapid change in platinum and a slow change in gold. The use of a short exposure to air seemed to affect the platinum electrodes by creating temporary potential rises. This occurred during the first ten minute period for the electrodes exposed for 0.25 and one minute. This was again repeated near six hours for the former time exposure and near 3.5 hours for the latter time exposure. The gold electrodes exposed for one minute also exhibited a temporary rise starting near four hours. Although the temporary changes did not appear to affect the overall adjustment curve of any electrode, it did indicate a certain degree of instability was possible after very short exposures to air. This was especially true for the platinum. This technique of sludge conditioning reduced the adjustment time under quiescent conditions. But it did nbt offer a method which could be done within a short time interval such as 30 minutes. To do this the electrode would probably have to be conditioned with a more negative solution so as to make the initial potential more negative than the sludge and thus adjust from the more negative tothe less negative sludge potential. A trial was made by conditioning a platinum electrode in a strongly basic sodium thiosulfate solution with a potential around -610 mv. The electrode was transferred as rapidly as possible to the sludge and the potential changes were followed. The potential did not retain its negative value and at 2,5 minutes was at -412 mv, At 21 minutes the potential had

dropped linearly to -478 my, and then continued to drop at a lower rate but still in a linear manner. The potential finally leveled off at six hours at a potential of -578 mvy Four days later the potential has risen to -510 mv. The change in slope at -478 mv and 21 minutes was very distinct and the indicated potential approximated that of the sludge. Thus, this method might enable rapid sludge potential determinations to be made, There were many possible highly negative solutions which could be used as an electrode conditioner and this approach would warrant further study of the technique. The conditioning of the electrode in an anaerobic solution before use in a digested sludge was opposed to the accepted practice of cleaning the metal by techniques such as acids, oxidizing agents, abrasives and heat. As was demonstrated previously, the history of the metal which includes cleaning was an important factor in the adjustment time of the electrode. In addition, the metal history could also play an important role in determining the shape of the adjustment curve, especially in the early phases of the adjustment. However, the final equilibrium potential was usually dependent on the sludge and not on the electrode. It was this initial period after the electrode was immersed that could possibly be developed into a means of determining the metabolic activity level of the sludge. No method was available for determining the sludge act iily a1Md otherefore., this would provide a valuable supplement to the existing tests for sludge. However, certain important factors must first be considered.

,89 It has been found in this research that the electrode could be more important than the sludge in determining the potential-time relationship. With apparently identical electrodes with identical backgrounds, a more dormant sludge could appear more dynamic than an active sludge. A survey of the literature also disclosed many instances of the lack of electrode agreement after the initial insertion(l6)(49)(66)(73)(75) In addition it has been found that metal surfaces could not be considered the same even though apparently treated identically(l38)(41)o On the basis of the above information, it should be stressed that the electrode must be in a reproducible conditicon before being placed in the sludge if the adjustment period is to be used as an indicator of sludge activity. The use of a controlled oxygen electrode(l )(132) would appear to be an excellent method to determine the dynamics of the sludge. The metal would contain a predetermined amount of an oxidizing agent and in this way provide the same initial electrode characteristics. Thus, the removal of these oxidizing materials on the electrode would provide a potential measurement reflecting the activity level of the sludge. The treatment of the electrode metal before use has been shown to affect the adjustment curve,: In addition the inherent characteristics of the metal Could also be a factor as shown previously in-Figures 10 and 12 Another example of a distinct difference in the behavior of gold and platinum was found when the electrodes were immersed into a sludge freshly dolbwn from a plant digester. The adjustment of the gold Was responsible for the difference since it appeared to form a temporary plateau between

a90 --350 and -400 mv before finally reaching the equilibrium potential. One of the more distinct examples of this is shown in Figure 13. Platinum exhibited a slight tendency towards this behavior, but it was not nearly as distinct a phenomenon. Rather, the adjustment followed more closely the usual situation where the potential-time relationship tended to be a continuously decreasing slope without any interruption in the continuity. The cause of this unusual behavior became more clear when the electrodes were reinserted into the sludge 12 days after the initial trial. The usual type of continuous curve resulted with practically no trend towards a plateau. Tests in other fresh sludges showed some tendency towards the formation of a plateau, but never as distinctly as shown in Figure 13. The reason for this behavior could have been due to both the sludge and the electrode. The use of six electrodes ruled out any possibility of an erratic electrode being the cause. It appeared as if a weak oxidation-reduction system in the s budge was affecting the gold more than the platinum. This possibility existed since platinum has been shown to be less sensitive than gold in this research and by other workers (16 54, 74) Another possible reason for the differene in the metals is that platinum may have been more reactive to the equilibrium potential than gold. Thus, weakly poised systems would not interfere with its adjustment. The work of Lepper and Martin(15) with cooked meat media in a buffer solution produced similar results, It was found that the special gold-plated electrodes were not affected by the hydrogen gas present and remained poised by the feeble redox system. However, iridium rapidly converted to a hydrogen electrode reaching a more negative potential.

-9l The fact that this platlau effect with the gold electrodes was not as noticeable 12 days after the initial trial and that other sludges did not produce as distinct an effect could be explained by the sludge itself. The volatile solids content was at a high level of 65 percent which was not present in any other sl-tdge tested. Therefore, it was possible that the sludge possessed a redox system which disappeared when the sludge was fur't.her along in the digestion period. In contrast to the experiments with fresh digested sludge described above, adjustment trials were conducted with an extremely well-digested sludge which could be classified as dormant because of a three year storage period. Before insertion of three gold electrodes, the sludge was mixed to blend the supernatant with the settled sludge. Since the sludge was three years old and had been exposed to the atmosphere during this period, it was anticipated that the solution would be more aerobic than the other sludges used in adjustment studies. In addition, it was considered possible that the potent tial might reflect the oxidative state of a fully digested sludge. Although no characteristic potential could be determined due to the very large disagreement of up to 250 mv between the electrodes, the adjustment time and potential reached was very surprising. One of the electrodes reached -715 my and another reached -650 mv. These values were much more negative than any potential obtained in digestion studies. The value of -715 mv closely approximated the potential of a hydrogen electrode. However, it should be mentioned that these values could not be considered reproducible.

-92 -Another large difference with normal sludges was indicated by the shortness of the adjustment periods. One electrode fell from -320 my to a constant potential at -680 my in one hour. Another fell from -340 my to -690 mv in about the same period. The electrodes had been exposed to the air for 15 minutes between their removal and reinsertion into the sludge. The rapid adjustment could indicate a well-poised solution, but-'the poor reproducibility of the potential and lack of agreement between electrodes did not make this assumption valido Therefore, it appeared that this three year old sludge differed from a normal digested sludge in its ability to produce a residual effect on the electrode. This sludge seemed to possess the dynamic property of establishing a rapid potential even though it was completely dormant. Therefore, it must be placed in a special cate ory for the use of the adjustment period as an indicator of the metabolic activity level of a sludge. The highly negative potential indicated by the electrodes also placed this sludge in a special category. Although the cause of this low potential was not known, the electrodes were not considered to be faulty since they exhibited normal behavior in all other sludges in which they were usedo In addition to the adjustment trials with various digested sludges, the author had an opportunity to follow an adjustment curve in a lake bottom mud sample. The results were included here because of the surprising similarity to the sludge adjustment curves. The sample was obtained from the bottom of Ford Lake, an artificial lake near Ypsilaht.i,

- 935. Michigan. The first similarity to sludge was in the adjustment period. The platinum electrode took about ten hours to come to a constant potential which remained at this level until 25 hours when the test was completed. The shape of the adjustment curve was also similar to that resulting from sludge. Furthermore, the potential reached was -465 my which also fit into the results of sludge trials. A brief survey of the literature disclosed that Mortimer, who worked extensively with lake muds, used a two hour adjustment period(ll9) It was not known whether this shorter time was correct or not for lake mud potentials since this experiment was the only trial made. But, the results did indicate that both anaerobic solutions, lake mud and digested sludge, could exhibit very similar potential characteristics,

-300 -3 50 -400 -450 -500 0 10 20 30 40 50 60 0 10 20 30 40 50 60 TIME (MINUTES) TIME (HOURS) Figure 9o Adjustment of a Platinum Electrode in a Digested Sludge

-95 --360 -380 3 -400 E o -420 PLATINUM -440 -460 - -480 -400 -420 I. HAND MIX E I I I I I ill 1 2. MECHANICAL w -440 3. MECHANICAL MIXER OFF — 460 -4 80 0 5 10 15 20 25 30 35 40 45 TIME (HOURS) Figure 10. Comparison of the Potential Changes of Platinum and Gold Electrodes due to Stirring a Digested Sludge Exposed to Air

-300 -.. -350 -400 o 'MIXER W ON -450 MIXER OFF -500 -550 _ - _ 0 20 40 60 80 100 120 140 160 180 0 5 10 15 20 25 TIME (MINUTES) TIME (HOURS) Figure 11. Effect of Agitation on the Adjustment of A Platinum Electrode in Digested Sludge

-350 -375 *~Xx~ ~GOLD A -400 E w -425 -4 50 -475 -350 -375 75.. -- EXPOSURE E -400 TIME ~U~~~~~~~~~~~~~~~~~~ -~ -._.- ~(MIN.) W ~0 0.25 -425 ______ OI' X3 -450. -AS PLATINUM -475 0 20 40 60 80 100 120 140 160 180 0 5 10 15 20 25 30 TIME (MINUTES) TIME (HOURS) Figure 12. Comparison of the Adjustment of Gold and PXatinum Eleotrodes in Digested Sludge after Being Conditioned in the Sludge and Exposed to Air

-280 -300 -320 -340 -360 o-380 -400 -420 -440 -460 0 5 10 15 20 25 30 35 40 45 50 55 60 TIME (HOURS) Figure 13. Adjustment of a Gold Electrode in a Digested Sludge with a High Volatile Solids Content

SACTOSE BROTH FEEDING OF A DIGESTED SLUDGE The composition of the raw sludge fed to a digester is very difficult or impossible to reproduce exactlyo Therefore, to aid in the evaluation of the,; potential changes produced by the natural feed, it was decided to use a synthetic constant composition solution as a feed to a digested sewage sludge. This would provide controlled loadings as well as reproducibility The use of a sugar in an n.oculated bacterial media would result in rapid and large potential changes. Since much work on this subject has been done, a comparison would be possible by using a synthetic feed containing a sugar. In addition, this type of feed would provide an indication of the magnitude of the potential changes. possible in a digested sl-Udge media0 The feed selected was Bacto-Lactose Broth (hereafter called LPB) in which one gram contained 03.85 grams of lactose and peptone, and 0.230 grams of beef extract. This particular feed was selected primarily bee cause of its sugar and peptone content. Furthermore, a similar feed was used in sludge digestion studies by Morgan(13)o Tgree goald eleatrodes were placed in 700 ml of a well digested sludge which consisted of equal quantities of two and ten month old sludgeo Additions of the LPB mixture were started after the electrodes had reached equilibrium. The feedings were primarily on a batch basis although the results could be placed together so that it could be regarded as a aethie continuous basis0 The initial digested sludge was used throughout as the

-100 -media. An addition of digested sludge and removal of supernatant was made when the mixture had become too dilute due to the water used to suspend the solid feed. Feedings of two and three grams of LPB were spaced twenty hours apart and corresponded to a loading of 0.17 and 0.28 pounds of volatile solids per cubic foot per day. The potential exhibited an instantaneous rise after the feeding fOll'owed by a fall, a rise and another fall. This same general pattern was followed with a ten gram feed (0.96 pounds of volatile solids per cubic foot per day) as seen in Figure 14. After the feed, the potential became more negative, leveled off, and then continued to its maximum negative value. The leveling off period appeared to be related to the lag phase as indicated by the gas production. This level period ended as the gas rate approached its maximum value and the beginning of the.potential rise corresponded to the drop in the gas production. The gas collection system was not a constant pressure type and the high gas production with the ten gram feed necessitated a release of the pressure at 7.7 and 9.1ol hours after the feeding. Although the latter time corresponded to the beginning of a rapid potential rise, it was not felt that this was responsible for the potential change, since the gas production had become very low at this point. To avoid the-release of pressure, a constant pressure gas collection system was adopted.

A feeding of four grams of LPB gave similar results in a constant gas pressure system. This is shown in Figure 15o Since a relationship between the potential and gas production was shown by various feeds, it could be concluded that the potential was a sensitive indicator of the metabolic changes occurring in this batch-type system. Gas rates were measured by a glass tube packed with glass wool in which the pressure drop was a linear function of the gas rate. The gas rate meter was used to establish the rate of gas production since calculation of the rate from total production was not as accurate because of the rapid gas changes. However, the gas rates yielded the same conclusion as total gas production; namely, the relationship of gas flow and potential. The rates also confirmed the work of Morgan(l3) as to the rapid change in gas production with a synthetic type feed0 In, addition 'to this,, he has shown that- a dextrose-peptone-beef extract mixture fed to a digested sludge resulted in an increase followed by a decrease in the volatile acid content, and that the acid content was proportional to the rate of gas production0 From this finding and the results of this research, it could be said that the potential changes also reflected the volatile acid changes0 The previous results have shown that the lag phase was reflected by the potential0 Therefore, it would be anticipated that if the lag in gas production were small, the potential should reflect this difference by reaching the maximum negative potential without first exhibiting a plateau. Examples of this result are shown in Figures 16 and 17 for which the feed was four grams of LPB for each case0 This was equivalent to

loading of 0.38 pounds of volatile solids per cubic foot per day. The decrease in the lag time was probably due to the effect of previous feeds in stimulating the bacterial activity. The potential in Figure 16 fell to a constant maximum negative level and then rose slowly. The rise started soon after the maximum gas rate was reached. Figure 17 exhibited a more distinct rise in potential after falling to a short constant maximum negative potential period. This rise started near the beginning of the maximum gas rate period which extended over several hours and did not exhibit a distinct maximum rate. In all cases studied the potential reflected the end of the rapid gas production by rising to a more positive potential. The immediate effect of the feed upon the potential Was in a positive direction in all cases but one. The one exception was due to erratic electrode behavior and therefore could be discounted. The solid feed was suspended in small amounts of distilled water and because of its more positive potential, should have caused a positive change in the potential as was the result. A factor which could not be overlooked when discussing the immediate feed effect on the potential was the effect of mixing on the electrode. After the feeding the sludge solution was mixed by hand. The results of mixing when no feed was added indicated that the potential in most cases became more negative. Thus, the overall positive effect of feeding included the opposing negative tendency of the electrode due to mixing.

The potential resulting from any feeding exhibited one general pattern. After the feeding, the potential fell and then reversed itself within a tEelve hour period. This potential fall was expected as the metabolic activities of the ana;erobe initially created a more negative potential. The work of many authors using pure cultures of bacteria indicated that after the maximum negative potential was reached, a more positive potential resulted. Using synthetic sewage, Moore, Rucchoft and Wattie(96) found that after the maximum negative potential was reached, a positive trend followed. With Eo coli a stabilized condition soon resulted due to their inability to utilize the end products. In contrast, A. aerogenes being able to utilize the end products continued to show a positive rise in the potential. Certain bacteriological methods of anaerobiosis often allowed air to enter the system which could partially explain positive potential changes. However, in this research anaerobic conditions were maintained by the sludge itself in a sealed system under pressure. Therefore, no access of air was possible which could cause a positive potential trend. It would appear that this potential change was due to metabolic action alone. The first rise and fall in potential could be referred to as a primary rise and fall for comparison with the other potential changes. The behavior of the potential after the gas production had practically ceased exhibited a definite pattern. After the primary potential rise, it was noticed that a secondary fall in potential developed without an appreciable increase in gas production as was the case for the primary fall. This can be seen in Figure 14o The mixing

of the solution near 24 hours was not responsible for this fall. It has already been demonstrated in several instances that the fall would occur without mixing. This is shown in Figure 17. The effect of the mixing was in creating a more negative potential instantaneously which did not begin to fall in a continuous manner until at least two hours after the mix. Further potential measurements over longer time intervals revsealed that there existed both a primary and secondary fall-and rise followed by a third fall which appeared to level off. This is shown most clearly in Figure 16 in which the time period covered about 75 hours. The primary fall and rise occurred in the O - 24 hour period, the secondary fall and rise in the 24 - 48 hour period, and the tertiary fall and plateau after 48 hours. The similarity in the potential curves after various feeds led to the conclusion that an LPB type of synthetic feed resulted in three zones of falling and two zones of rising potential in the first 72 hours. The fact that this feed brought into play different bacterial and enzymatic systems than a raw sludge feed was shown clearly by the resulting potentialtime curves. Raw sludge feeds produced a very level potential curve without periodic ups and downs. These reversals of potential could not be attributed to any one bacterial system due to the complex flora that existed, Possibly these reversals were a reflection of the changes n acid production as well as different bacterial systems,

- l05 -The potential was followed beyond 75 hours for one of the feedings. This is shown in Figure 17. The potential started to fall at 95 hours reaching a plateau of about r519 mv near 145 hours0 This fall to a new low level probably indicated a more highly reducing system was created as the bacterial systems became dormant. The potential did not vary widely until at 1.95 hours a rise began which reached a maximunam near 245 hours of -405 mv. The potential started to fall at 260 hours until, it leveled off near 295 hours at -500 mv0 It then. started to fall slowly at 325 hours to -520 mv. This level was maintained from about 430 hours to 480 hours at which ti.me readings were terminated. It should be noted that the potential. returned to the same level, as before the last rise and fallo The potential changes were probably due to the metabolic activities of certai.n. bacterial systemso The rise at 195 hours could hare been influenced by a temporary release of: the press-ure to 6 mm at 1,75 hours and a mix of the soluaton at 193 hourso The mixing would have made possilble the renewal of the activity by exposing new food sour@es and removing localized waste products0 The~ potential before the feed was -467 my and the primary fall. reached:481 mv. The fall to -519 mv at 145 hours represented a change of 38 myr based on the maximum primary fall potential. This could be comr pared to the ten gram feed shown in Fi.gure 14 The maximum of th.e priimrary fall, at eight hours wastp 496 mv and at 145 hours the potential was 539 mynar Thi.s resultant differience of 43 my exhibited a remarkable agree.mdnt which. seemed to indicate that there existed a, certain overall reproducibility for this type of synthetic feed0

The first potential fall was probably due to the increase an metabolic action shown by an increased gas production which resulted in more negative potentials. The following rise after the peak gas production was reached was probabtly due to more oxidized materials accumulating in the solution. The second fall was not as easy to explain. Possibly a new bacterial system was brought into action by the waste products created initially, or the same initial system after adjusting to the waste products created this change. The two systems could operate in series with the new system changing the waste into new tfood or destroying the. toxic effects so the original system could continue to metabolize. Gas production was not large enough to show any relationship to the second flall and rise although gas production did not cease. However, rapid gasification was not apparent after 12 hours. In Figgre 17 there was more gasification than usually was the case up to about 40 hours but this was an exception. This would indicate the sugar must be nearly completely used in the first 12 hours and the peptones must be the source of the additional gas. This would indicate a new bacterial system was in operation after the initial phase. The third fall at about 50 hours was not followed by a rise in potential which could indicate the food supply was consumed. The maintaining of a constant potential level could be attributed to the fact that the oxidized substances which had altered the potential were decomposed into substances which poised the sludge at a certain potential level.

1t07h There have been other workers who have observed potential changes after the primary fall and rise. A double reversal was found by Vennesland and Hanke( ) for a synthetic media containing one percent glucose inoculate ed with Alcaligenes fecalis, a non-fermenting obligate aerobe, and Bacterordes (48) vulgatus, a non-spore forming anaerobeo Hewitt also observed this phenomenon with E. coli in a one percent glucose broth. With a media of two percent neopeptone and eight percent galactose, Longworth and MacInnes found. that Lactobacillus acidophilus also yielded a secondary fall and rise in the potential..Although the digested sludge provided a mixed dulture and a different media base, it was of interest that a potential, reversal also occurred under these conditions All of the above authors concluded that the potential reversal was not due to the pH, but to the production of acid~ Checks on the pH during the experimental work revealed that the sludge was a good buffer since the pH was not found below seven. Thus, the potential reversals were probably due to acid production based on the work of the above authors, In order to establish a characteristic potential for this system, the variation in potential after feeding should be considered. For the ten gram feed in Figure 14 the variation was 84 mvo Four gram feeds in Figures 15, 16, and 17 yielded differences of 66, 42, and 32 mv respectively. These variations were for the time period from the feeding to 55 hours. By consider 'ng the midpoint of the variations in the potential for the above and other feedings, the characteristic potential zone of this system ranged from [440 to -470 mv or an-Eh of -194 to -22~4 mvo

Upon extension of the time to 150 hours after the fdeding, the variations for the ten and four gram feeds in Figures 14 and 16 were 128 and 71 mv respectively. The resulting potential due to the ten gram feed was at -540 my after 150 hours. Subsequent feedings resulted in a normal primary fall and rise but the secondary fall and rise caused the potential to become more positive. The time between feedings did not exceed 55 hours. There appeared to be a strong tendency for the potential to return by steps to the characteristic potential zone. This return substantiated the location of the characteristic potential zone.

Ec (mv) TOTAL GAS (cc) (' - P v P P P P P N ( J 0 a1 s o 0 a:) O - n 0 0 0 0 0 0o o 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0 o 0 -- - ---— ~ — -- - -n (11 aO C,' KnH 0 1 Figure 14. Electrode Potentials and Gas Production after the Addition of Ten Grams of Lactose Broth Concentrate to a Digested Sludge Media Grams of ]Lactose Broth Concentrate to a Digested Sludge Media

300 Ii~~ ~~~~~~~~~~~~~~~~~~~~~~ _ _ii 'j~~~~~~~~ Y~~~ O0 ~" 200 cn \-n0 ~k- 100 (P H*.() 0 cic~ CD CD c td H -50t FEED ID-5 10 lx FJ ~~~~~~~~~~~M IX,Y t-J~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ iH' H' ~ ' -5 0t/' (D CO ct ~~~~~~~~~~~~~~~~~~~~~~-5 - o o t 53 P0 o ~~cn E cio+ ~-5 -0 0 0! ct ct 0H- -560 c-!ct o P CD -560 CD 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 TIME (HOURS)

- 111 -0 (D 0 tIn _ _ _ - In _o _ _0 — 0 (.() Stt ivLOl (AW)33 Figure 16. Electrode Potential and Gas Production Resulting from a Short la;g Period Digestion after the '&ddition of Four Grams of Lactose Broth Concentrate to a Digested Sludge Media CV~~~I

-112 --425 > -450 E E -475 -500 -525 0 100 200 300 400 500 TIME (HOURS) 500 400 -44 --450 0E 460 -480 -490 0 5 10 15 20 25 30 35 40 TIME (HOURS) Figure 17. Electrode Potentials over a Long Time Period and the Gas Production Resulting from a Short Lag Digestion after the Addition of Four Grams of I&etose Broth %oncentrate to a Digested Sludge

INITIAL POTENTIAL STUDIES ON SIALTDGE DIGESTION The lactose broth experiments have shown that potential measurements in a digested sludge media could provizde information on the mechanism and control of this type of biological process0 This was primarily dxue to the magnitude and rate of the potential changes3 which were large enough. too (1) overeome the media poi.se or resistancee to change and (2) mi.nimize any variability introduced by the metal electrode per se. Although, the potential. changes produced by a digesting sludge with raw sludge feedings would not be as distinct as those with lactose broth feedings3 it nevertheless appeared that an evaluation of the former system would be possible~ The lactose broth data also indicated the reproducibility of electrode potential measuremernts was possible in an active di.gested sludge0 In addition, these experiments helped to define the best means of measuring the potential. This resulted i.n the use of three digesters, each with three electrodes3 and provision for continuous agitation. of the sludgeo Three electrodes in each, digester eliminated any reliance on a malfunctioning electrode which would not be possible if' only one electrode 'was available. Three digesters allowed comparison of the behavior of electrodes in different digesters under identical conditions as well as the electrodes within one digester0 By establishing this variability under identical conditions3 it would then be possible to e'valuate the results when the feeds were not the same in each digester0 ool.

The impdrtance of understanding the behavior of the metal electrode in digesting sludge was considered to be an essential component in the evaluation of any potential data. Much valuable data had been obtained from previously presented studies on electrodes in quinhydrone, digested sludge, and digested sludge with lactose broth additions~ However, since the digestion of sewage sludge had never been followed by potential measurements, the initial studies on sludge digestion were primarily to observe and establish the general characteristics of the metal electrode in thia system. The feedings were begun on a daily basis to provide a 25 day detention period by feeding 100 ml of raw sludge corresponding to a loading of 0.091 pounds of volatile solids per cubic foot per day. The digesters were continuously agitated. The digested sludge had been removed from a plant digester about two weeks before and was placed in the laboratory digesters at 250Co There was no noticeable gas production at this temperature. The first feeding was set at 50 ml after which the 100 ml feeds were started. However, unusual behavior of the electrodes led to the cessation of daily feeding after the third addition in order to follow the subsequent potential changes. It was felt that in this way more valuable data on electrode behavior would be gained. The first instance of unusual potential behavior occurred prior to feeding when the temperature of the sludge was raised from 250C to 35~C in continuously agitated digesterq. The electrodes had been inserted into the sludge five days previously and the electrode adjustment followed.

-115 -The temperature was raised in order to bring the digesters to the proposed operating level of 355OC This procedure allowed the opportunity of following the potential changes resulting from an increase in sludge activity by a temperature change, not an addition of raw sludge. The three digesters did not yield the same potential changes as can be seen in Figure 18. The gas production yielded an "S" type gas curve simulating the results of a batch-type digestion. The maximum gas rate occurred on the second day after the temperature was raised. Before the temperature rise, the electrodes averages -496, -498 and -488 my in Digesters 1, 2 and 3 respectively. Digester three first showed a rise in the potential followed by a falling potential until at about two days, a temporary leveling occurred. This was followed by a further drop until seven days when the average potential was about 519 mvJ The initial potential rise was probably caused by the entrance of air into the digester diet to a leak which was fixed after the first day. Digester one showed a dropping potential which became constant in six days at a value near -515 mvo In contrast to the above behavior was the action of Digester two since the potential in this case did not drop to the lower potential of the other two digesters. In six days the potential averaged about -472 mv after becoming essentially constant after only one day. The cause of the initial rapid rise from -498 to -479 my was unknown, There was also no apparent reason for the potential rising instead of dropping and being about 40 mv more positive than the other two digesters. The difficulty in finding an explanation arose from the identical experimental conditions of the digesters and the good agreement of the electrodes. This is discussed further at the end of this chapter~

l116 -Raw Sludge Feedings Raw sludge feedings were begun after the potentials had become approximately constant and the gas production was at, a very low level. This occurred about seven days after the temperature was raised. At zero hours 50 ml of raw sludge at 8~C were fed to each digester~ This was followed by 100 ml feeds at 24o5 and 5175 hours0 The resulting potential change and gas production are shown in Figure 19 and Table IIIo The potentials rose after each feeding and there was always a Subsequent trend to the more negative prefeed value. The potential values used were the average of either two or three electrodes. At times one electrode acted erratically which, if it were included, would not give an accurate picture of the changes. Table III SUMMARY OF ELECTRODE POTENTIALS AND GAS PRODUCTION DURING THE PERIOD OF CONTINUOUS AGITATION Change In Potential (Ec) Potential Total Gash Time Before Time from Before Production. (hours) Feed Indicated (my) Feed One (my) (cc) Digester Digester Digester 1 2 3 1 2 3 1 2 3 0 1 512 -472 518 - - - - 24.5 2 -509 -469 -522 +3 - 4 380 520 380 48o5 3sa 5o~5 =469 -50o +- 7 0 +18 1190 1340 1020 51o8 3 -500 -471 -470 +12 +1 +48 - - 72o5 - -497 -468 -489 +15 +4 +29 1910 2030 1490 960. -511 -481 -516 + 1 -9 + 2 4830 55300 5300 a~ - Sludge withdrawal before feed 3

-117 -The second feed of 100 ml yielded similar potential changes and gas yields for Digesters one and two. The behavior of Digester three differed by an immediate rise to a new potential level followed by changes similar to the other two digesters. There was no tendency toward recovery of the prefeed potential, only toward the displaced potential level. The gas production appeared to reflect this rapid initial displacement as it was about 18 percent lower than the other digesters for the 23 hour period after the feedingo After the third feed digesters one and two exhibited similar electrode behavior with the potential rise due to feeding higher for Digester one. The gas production was also slightly higher for this digester. A leak and pressure loss in Digester three, which was caused by the sludge withdrawal before the feed, resulted in a large potential rise. The pressure was restored at the time df the feeding. The potential then began to drop toward the potential existing prior to the pressure losso The gas production corrected for the loss was less by about 33 percent than the other digesters for the 20 hour period after the feeding. The reason for the potential rise in Digester three was probably due to the gas-metal effect since air had entered the digester. But the large difference in the gas production and gas rate indicated that the air affected more than just the behavior of the metal electrode. Since the gas necessary, for the pressure restoration was -tWiken into account, there could have been a residual effect on the digester prdcess itself. That this was a lag in the gas production was shown by the total gas production at 150 hours where the yields for all three digesters were about equaLO

-118 -The.evaluation of the potential measurements could be done on the basis of potential changes, as had been done above, or on the basis of the potential level. Digester two in the 72.5 hour period with continuous agitation remained largely in the -460 to -470 my range. Digester one varied from around -475 to -510 my with about -500 my being the most representative level. The largest variation was -470 to -520 mv for.Digester three. The most representative level was difficult to determine due to.the large variations but it could be approximated as the same as Digester one, -500 mv. Similar gas yields for Digesters one and two rqspectively at 72.5 hours indicated that the potential level did not necessarily indicate the production efficiency. Digester three with a lower yield did not bear comparison due to adverse experimental occurrences. The gas production at 145 hours indicated the digesters were still producing gas comparable to the rate obtained several hours after the third feeding. The potentials of Digesters one and two had fallen slightly since 72.5 hours. Digester three had become more negative exceeding the potential after the second feed by falling from the more positive potential caused by the pressure loss. The potentials appeared to be approximately constant until a trend to a more negative level occurred near 170 hours. This potential drop for Digesters one and two corresponded to the beginning of the rapid drop in the gas production rate. The correlation for Digester three was not as apparent since the gas rate did not begin the rapid drop until ten hours after the new potential level had been reached.

After the fall of the gas rate was nearly completed the continuous agitation was stopped at 200 hours. Divergence of at least one electrode in a digester and certain erratic electrode behavior prompted this action to determine if the agitation was in any way responsible. Since the results after stopping the mixer were very favorable, agitation-was henceforth used only to mix the sludge before a sludge withdrawal and after a feeding~ Separate experimental work on the effect of agitation on the electrode is discussed later in this thesis.. Small potential rises followed by a constant potential resulted from the cessation of the agitationo The beginning of the constant potent tial could be related to the beginning of a secondary rise in the gas rate, However, the constant potential gave no indication of the fall of the gas rate, Before the 8gitation was stopped at 200 hours, the potentials were very close to the' values before the first foe0.~ Digester two had not varied widely from this prefeed potential of J472 mv and at 200 hours was near -480 my, Digester one with a variation of nearly 40 mv was at -510 my very close to the prefeed potential of -512 mvo The widest variation by Digester three of almost.55 my found the potential at -515 ma9v slightly higher than the prefeed value of -519 myv. These results indicated the prefeed potentials could be called base levels which the potentials sought after an addition of a solution with a different potential.

Ninety hours after the cessation of agitation and 291 hours after the first feeding when the potentials appeared to be constant, anothier feed of 100 ml was given to the digesters. The previous feeding had taken place ten days before and gas was still being produced. The only difference from previous feedings was that mixing was done only for a total of 45 minutes during which time the digesters were fed. Included in this time period was 30: minutes of agitation after the feed to insure good mixing. The potential results are shown in Figure 20. The potentials before this fourth feeding were at -506 for Digesters one and three and at k464 mv for Digester two. After 170 hours the potential of each digester returned to within a few millivolts of the original potential before the feeding. The behavior of Digester two differed from the others by dropping to a potential more negative than the prefeed value and rising, instead of leveling off, after this potential was reached. The cause of this behavior could have been the prefeed potential which was 40 mv more positive than the other digesterso This would indicate a tendency to the more negative potential level. Although the potential changes were not large, the relationship between the production of gas and potential could be comparedo The constant potential zone for Digester one and three reached near 50 hours corresponded closely to the peak gas rate and start of the rapid gas rate drop. The rapid change in the rate was reflected by a slight drop in the potential to a new constant level in Digester one. The potential of

.121 -Digester two leveled off near 35 hours, and did not reflect the maximum gas rate. The peak gas rate of Digester two near 45 hours fell between the con. stant potential zone starting at 30 hours and the rising potential zone starting at 60 hours~ The sharp change in the gas rate starting near 70 hours did show a relationship to the potential which exhibited a drop start-:ihg at 60 hours. The total gas production could not be considered wholly on a pure batch basis as there was a residual condition from the previous feed. This was particularly true for Digester one which exhibited a spurt in gas production at 35 hours unmatched by the other digesters. In the 460 hour period after the first of the four feeds the total gas production was about 9550, 9050 and 9800 cc for Digesters one, two and three respectively~ Electrode Check The potential at 460 hours after the initial feed was at -501, -462, and.502 mv for Digesters one, two and three respectively0 The more positive potential of Digester two had persisted for over a month beginning with the raising of the incubator temperature to 355C and continued during subsequent feedingso Since the three digesters had been treated identically, the cause of this was not apparent. In order to investigate this potential difference of 40 my between Digester two and the other two digesters) the electrodes in. Digester three were strtched with those in Digester two to discover if the electrodes were at fault. The potentials of the digesters were known from previous experiments and this provided a reference for the electrode readings, The digesters had not been fed for seven days;

-122 -the potentials were constant and agreed well; and the gas production was practically nil. Before the transfer electrodes 10, 11 and 12 G in Digester two averaged -462 my and 13, 14 and 15 G in Digester three averaged r502 mvo Electrodes 10, 11 and 12G were put in Digester three after being exposed to the atmosphere for about eight minutes. The normal adjustment curve leveled off at about 30 hours. The average was at -499 my which agreed well with the previously determined digester potential of -502 myv Thus, it was shown that inherent characteristics of the electrodes were not causing the lower negative potential in Digester two. The behavior of electrodes 13, 14 and 15 G when placed in Digester two did not help explain the potential difference. The potential first leveled off near -492 my at 60 hours. This was followed by a slight drop near 90 hours to a potential of -497 which closely paralleled the potential of the other two digesterso The previously determined sludge potential of -462 my was not reproduced or even approached closelyo The potential changes exhibited by Digester two had been similar to the changes of the other two digesters even though at a different potential levelo The gas production did not indicate any distinct reflection of this difference, It thus appeared that the more positive potential level was a temporary, although persistent, condition that had been set up by some unknown action during the temperature increase period.

1500 -ioo..I,,1000 -j 500 0 I- — 0 0 * DIG. I HEATING UNIT ON 0 DIG. 2 -460 - X DIG.3 3.Q ----'")' — 0-: —. — '0 -0 I 0~ > -480 ', E 'X w -500 -520 0 20 40 60 80 100 120 140 160 180 200 TIME (HOURS) Figure 18. Electrode Potential and Gas Production Resulting from an Increase in the Temperature of the Sludge from 25 to 350C.

6000 XO-.x _Xx — x 5000 ~ 4000 C), 3000 0 - 000 I- ~~~~~~~~~~~ir i 000 H 0 AMP _ _G FEED -- -DIG. I FEED FliED MIXER OFFDIG. -460 X DIG. 3 -4 -0~ E -480 'I~~~~~~~~~~~_ -500 ~~~~~~~~~~~~~- oo.~ -520 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 TIME (HOURS) Figure 19. Electrode Potentials and Gas Production Resulting from Digester Feeding with Continuous Agitation

10000 -X.x-* -. -- 9000 Ie 0 ~DIG. I ole~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r 7000 I,-~~~~~~'e 6000 FEED -460 > -480 E ow -500 I ML ' —X — -520~~~~J - -K5. 0 20 40 60 80 100 120 140 160 180 200 TIME (HOURS) Figure 20. Electrode Potentials and Gas Production Resulting from Digester Feeding after the Continuous Agitation was Stopped.

BATCH TYPE FEEDING The results of the lactose broth feedings have shown that the poten tial would reflect a lag in the gas production after feeding. Since the last raw sludge feeding in Figure 20 yielded a very short lag period, it was decided to attempt to lengthen the lag period and observe whether the potential would reflect this with raw sludge feedings. By using a sludge not fed for 30 days, a ripe but non-gassifying sludge, a lag period should be observed. By employing a single feed and creating a batch type digestion, it was also felt that the potential changes would be more distinct than in a continuous system. This was important since the previous experimental work had shown that the variability in the metal electrodes could mask the actual potential change if the magnitude of the change was small. The 100 ml feed simulated a batch condition wherein a ripe sludge received a feed equivalent to 0.091 pounds volatile solids per cubic foot per day or to a 25 day displacement without any direct influence of previous feedings. On a batch basis the ratio of fresh to digested sludge was one to ten based on the volatile solids content, The prefeed potentials were at -499, -493 and -501 for Digesters one, two and three respectively. The expected rise in potential after feeding took place followed by a rapid recovery to the prefeed potential as seen in Figure 21. In this case the feed made the potential more positive while the five minute mixing about 20 minutes later started the trend toward more negative values. At 165 hours the potentials did not differ from the prefeed values by more than two millivolts. -126 -

127b - The increase of the gas rate occurred in two stages. The first plateau reflected the lag phase and the second reflected the maximum rate. This is shown in Figure 21. The average rate based on the first ten hour period was at eight cc/hour which increased from a negligible gas produce tion before the feeding. There was some relationship between the potential and the gas rate. After the feeding the potential dropped and the gas rate rose but the potential did not reflect the leveling off of the gas rate near 20 hours. The potential drop ceased at 30 hours and the gas rate started to increase to its maximum at 40 hourso It could be observed that the midway point between the beginning of the initial gas rate plateau and the rapid rise correlated with the start of the potential plateau. The cessation of the rising gas rate at 70 hburs was only reflected by a rising potential in, Digester two which started at 50 hours and ended at 68 hours, The start of the rapid drop at 120 hours -as not reflected by the potential. The next significant change in the potential was the beginning of the potential rise toward a new constant level, the potential before before zthe feeding. The end of the rapid gas rate drop preceded this potential rise by about ten hours. This was opposite to the relationship of potential, and gas rate before the start of the rapid gas rise. In this case the gas rise followed the beginning of the potential platdau by about 15 hours.

l128 -From these results it appeared that it was necessary for the potential to reach a certain level before the gas production could approach its maximum rate, Before this potential level was reached, the sludge exhibited a lag period with a constant gas rate equal to about one-half the maximum rate. After the gas rate drored to a low level, the potential.eturri. to its original value representing a ripe sludge with a very low gas production A significant result of tiis experiment was the small potential changes. This was rather surprising as the conditions were similar to a batch feed condition with a definite lag phase where it would be expected that large changes would resulo, Excluding the immediate potential changes due to the feed, the change from a to 200 hours was 10, 14 and 11 mv for Digesters one, two and three respectively, It was during this period that all the gas rate changes occurred. It would appear that che digested sludge was very well poised to resist the potential changes resulting from the feed. Even though the potential change was not large, the excellent agreement of the electrodes within a digester as well as between digesters indicated that the results shown in Figure 21 were an accurate representation of the potential changes. The electrode agreement indicated thpt potential measurements couldlbi;a reproducible tool for sludge digestion studies. However, it should be mentioned that the initial studies previously presented as well as other studies did not exhibit the fine electrode agreement shown here. Thus, it was considered desirable to produce larger potential changes in order to overcome any electrode variability as discussed in the next experiment o

Ec(mv) 10 HOUR GAS PRODUCTION RATE TOTAL GAS (cc) I I I (cc/HOUR) C, (A 0' o o N 0 OD N 0 0 0 0 0 0 0.0 0 0 0 0 0 CD R) 0al-M ~~~~t~~~mim N __ F-J 0 ~ Ct 0~ 0 CD 1 fld i c+ OM _ i Co3, 0 _ No o: _ _ __ __ _ Co ~~~~ CD ~3 F-~~C M O X! - C~~o ':@0 N 0i 0 0

ACTIVATION OF A DORMANT SLUDGE The ability of a ripe sludge to poise the potential after additions of raw sludge resulted in small changes in the potential. Because of the variability in the metal electrodes, these small potential changes were difficult to evaluate accurately. The lack of a wide potential range in the batch type feeding previously discussed with a sludge aged for one month led to the use of a sludge media aged for four months. It was thought that this aging period would produce a dormant sludge '(lack good seeding characteristics) which would yield larger potential changes after additions of raw sludge. The aging time to classify a digested sludge as dormant was based on the work of Heukelekian(120) who found that after three months at 200C the seeding value of the sludge was lost rapidly based on gas production. He felt the optimum time was after the decomposition of the digesting mixture was completed and up to one month after. In addition, Heukelekian and Heinemann(33) found that the maximum number of methane bacteria were present at the peak of gasification and that the organisms maintained their values for at least 90 days with a decrease after gasification stopped. The use of a dormant sludge also offered an opportunity to study a long lag period digestion and the resulting relationship between the potential and gas production. In this manner the results could be compared to the shorter lag periods where an active digested sludge was used and to the lactose broth experiments.

-131 -After several widely spaced feedings. the identical feed quantity to each digester was varied, This was done to study the potential differences arising from different feed quantities. The varied or non-parallel feedings were initially begun on a daily basis, However, the potential exa hibited unusual trends resulting in the discontinuance of the feedings. The digestion was then followed without any further sludge additions since it was felt the results would be more meaningful than if further additions were made. There has a total of -slx feedings given to each digester with the first four being equal and spaced about 90 hours apart. The two non-parallel feedings were given 24 hours apart with one digester receiving the same quantity as in the previous parallel feeding sequence. The feed was preheated to 30~C before being added to the digesters. The digesters were agitated before and after each feed for five minutes. The potentials during the feeding period are summarized in Table IV. Parallel Feeding The first feed of 50 ml of raw sludge produced very little gas as seen in Figure 22, The first few hours after feeding showed a negligible gas production which was contrary to the effect of feeding an active sludge. The second feed of 100 ml at 91.5 hours also yielded negligible gas produce tion for the first eight hours. However, 71 hours after this feed there was production of gas which., compared to an equal time period after the first feed, was over five times as great. This indicated the methane bacteria were becoming active. This was also shown by a rise in the gas rate after

TABLE IV SUMMiARY OF ELECTRODE POTENTIALS DURING THE FEEDING PERIOD STARTING WITH A DORMANT SLUDGE Time Time Feed Feed Quantity Potential (E.) Resulting Potential Change (hours) Interval No. (m!) Before Feed Potential Based on Value (hours) (my) Change (my) Before Feed One (my~) Digester Digester Digester Digester 1 2 3 1 2 3 1 2 3. 12 3 0 1 50 50 50 -496 =481 -478... 91.42 91:42 2 100 100 100 =519 [495 -526 23 14 - 48 23 14 = 48 163.25 71.9 100 100 100 -506 =491 =519 + 13 + 4 + 7 - 10 - 10 - 41 258.9 95.6 4 100 100 100 =492 -475 =500 + 14 + 16 + 19 + 4 + 6 = 22 355.25 96.6 5 300 100 200 =484 -468 -500 + 8 + 7 0 + 12 + 13 22 379~92 24.6 6 300 100 200 =477 =470 =500 + 7 = 2 0 + 19 + 11 22 404.2 24,3 -- --- --- -469 -470 -491 + 8 0 + 9 + 27 + 11 13

133-5 the third feed of 100 ml at 163 hours. This was the first time the rate had increased soon after the feeding. The fourth feeding of 100 ml at 259 hours also exhibited this rise. The first feed produced a rapid negative potential change which was most severe for Digester three. This is shown in Figure 22. The pattern of this fall to a more negative potential after feeding was not exhibited in other experiments where the usual result was a rapid positive trend followed by a recovery to the more negative prefeed level. This difference was probe ably due to the condition of the dormant sludge. The large change to a more negative potential could have been due principally to the hydrolytic enzyme systems since the gas production was so small. However, the changes were not distinct enough from other potential measurements to warrant any breakdown concerning the system causing this potential change. Mixing of the sludge at 79 hours for seven minutes produced a sudden positive change in Digester one, a gradual positive change in Digester two and practically no change in Digester three. Since the usual effect of agitation was in creating a more negative potential, the results here appeared to be due to the sludge and not due to the electrodes. The potential. changes after the second feed at 91 hours were not as rapid as were those after the first feed. There was a slow trend toward a more positive potential. This was the period in which the gas production began to increase. There was no direct indication by the potential. of this increase,

The third feed at 163 hours produced more rapid potential changes than the second feed as well as an immed.ate increase in the gas rate. The potentials became more positive~ A distinct change in the gas rate at 210 hours was not reflected by the potential~ The potential changes after the fourth feed were not large and showed fairly constant behavior during the 96 hour period following the feed. The potential, 96 hours after feed four, was at -4849 4468 and -500 mxr for the Digesters one9 two and three respectively0 The activity of the sludge had reached a level where the gas was produced at an approximately constant rate0 Although this rate was not as high as obtained on the other runs, it appeared that the sludge was functioning normally. On the basis of the gas production.as well as the similarity of the potential curves after feed four, it was felt that the parallel operas tion could.be abandoned and non-parallel. feedings started. Non-Parallel Feeding The loadings for Digesters one, two and three were set at 300, 1.00 and 200 ml, of sludge. These were equi.valent to a displacement of 80339 25 and 12o5 days and a loading of 0o25, 0~083 and 0o166 pounds volatile solids per cubic foot per day. The feed to Digester two was kept the same as in the parallel feeding section in. order to use this digester as a reference0 Although the potential. after feeds five and six reflected the different loading levels, the results when compared to the previous feeds were following a different pattern0 It was therefore decided to postpone the next feeding 24 hours to see if a longer time after feeding would throw

7l35 -more light on this situation. By the end of this period the potentials had exhibited a potential drop for all the digesters. Since this had not occurred with any of the previous feedings, it was decided to follow the resulting trend without any additional feedings. The beginning of this dropping potential occurred 35 hours after the sixth feed. By reference to the time of the first non-parallel feed, number five, as zero, then the drop for all three digesters began near 60 hours. These potential changes are summarized in Table Vo The potential of Digester two remained essentially constant from the time the negative trend ended near 102 hours until the completion of the run at 310 hours. All the major changes in the gas rate of this digester occurred before 102 hours0 The potential of Digester three varied little after 145 hours and near 280 hours rose slightly to a potential of 511 mar The potential of Digester one started to drop near 220 hours until reaching a new low near 290 hours of -509 mv. This drop was probably a reflection of higher metabolic activity due to the larger-sfeed quantities. The agreement of Digesters one and three within two my was of special interest since they had not been that close since after feed one which took place 675 hours previouslyo After four parallel and two nonparallel feedings which provided differences up to 30 mva, the agreement provided a basis for the determination of the potential level, of ripe suldge. According to Figure 24 it was possible to relate the gas rate and potential in many instances. However, the data also indicated that neither the potential nor the gas rate changes could be used to predict the behavior

-136 -of each other. But it was found that the relative production of gas and the relative feed quantities could be predicted from the change in the potential0 By considering the negative trend or recovery period starting at 60 hours, the differences in potentials at the first point where all the digesters exhibited constant potentials would yield this result. TABLE V SUMMARY OF THE POTENTIAL CHANGES FOLLOWING THE MAXIMUM POSITIVE POTENTIAL Time Potential (Ec) Potential Potential Time Interval (mv) Change (mv) Change from (hours) (hours) 61 hours (my) Digester Digester Digester 1 2 31 2 1 22 61 a- -468 -469 -492.... -- -- - 78 17 -478 -478 -502 -10 - 9 -10 -10 - 9 -10 102 24 -488 -481 -511 -10 -3 -9 -20 -12 -19 126 24 -492 -480 -511 - 4 + 1 0 -24 -11 19 145 19 -500 -480 -514 - 8 0 - 3 -32 -11 -22 310 165 -509 -480 -11 - 9 0 + 3 -41 -11 -19 The first time this occurred was at 145 hours. Previously at 110 hours two of the three digesters were constant but the comparison appeared to be more valid when all digesters exhibited constant potential zones. This could be compared to the gas production in Figure 23 where the relative production could not be accurately measured until all three digesters exhibited relative flat curves. Before this time the results would be misleadi.ng. At 350 hours after feed five the production was approximately

10,000, 4,000 and 7,000 cc for the three digesters. The ratios were not even multiples of the feeding ratios, but by correcting these figures for the gas contributed by the previous parallel feedings, the ratios became 3.05/1, 1/1 and 2.08/1 for Digesters one, two and three respectively. The corresponding feed ratios were 3/1, 1/L and 2/1 and this indicated equal efficiency for each digester at a different feeding level. The negative potential changes in the 60-145 hour period were 32, 11 and 22 mv for Digesters one, two and three respectively. If the lowest change were used as a basis, the ratios would be 2.91/1, 1/1 and 2.0/1. This correspon.ded very closely to the nonparallel feeding ratios as did the relative prod-uction of gas. These ratios were only approximate as the potential changes were not sufficiently large enough to decrease the sensitivity of this ratio to small errors in the potential. But the approximate trend gave results that proved very enlightening as these ratios were indicative of the three different feed quantities. Overall Results of the Fee~dfngs The two methods of feeding the digesters have been considered separately in order to better present the experimental data. However, the overall results which included both the parallel and non-parallel feedings also provided some significant results as seen in Figure 24. The feed of Digester two was Kept at the same constant feed level so that the overall result could be considered without the effect of a varying loading comnplicating the results, In addition, this digester could be used as a reference for the other digesters which were at different feeding levels for

the last two feedings. The potential in the dormant sludge before the first feed was at -481 my for Digester two. After six feedings and 680 hours later the potential had dropped to -504 mv., risen to -468 my and dropped again to -479 mv. Although Digesters two and three showed similar patterns, the resuiting potential of Digester three was not nearly as close to the initial value as was Digester two. This possibly- was due to the last two feedings being higher. But it was more probably due to temporary readings in the dormant sludge as evidenced by the larger changes upon feeding than that exhibited by Digester two. Digester three began at -479 mv, dropped to -530 mv, rose to -491 mv and dropped again to -511 mv, Digester one started at -496 mvy, dropped to -527 mv, rose to -467 mv and dropped again to -509 mv. In the case of Digester two where all six feedings were at the same levelS an unexpected drop in the potential occurred after feed five at 355 hours. The drop was followed by a rising potential which after 25 hours was two ma lower than the prefeed potential. The potential had also dropped after feed one, and continued to drop whereas after feed five the potential rose, Feed six followed feed five by 25 hours and produced practically no change. At 420 hours or 35 hours after feed six the potential began to drop and continued dropping until near 460 hours. This potential varied very slightly until the end of the run at 670 hours. The dropping of the potential 35 hours after feed six was not observed after the first four feeds which had time intervals following them of more than 60 hours. The potentials after feeds three and four were taken in

an actively gasifying sludge while feeds one and two were given to the sludge when in the beginning stages of activation. The fairly constant gas rates after feed four and the rising rates after feed five indicated full activity was being developed during this interval. It was possible to explain this potential drop by means of the gas rate since the rate was always rising or remaining essentially constant after the first five feeds, However, about 20 hours after feed six the gas rate began to drop near 400 hours. This could be related to the potential drop at 420 hours >by stating that the start of the dropping gas rate resulted in the starting of the potential drop. However, the heavier feedings in Digesters one and three did not follow this pattern. The drop in potential of all three digesters began near 420 hours but the drop in gas rate *began near 425 hours for Digester three and near 455 hours for Digester one. Although the magnitude of the potential change varied proportionately to the feed as previously discussed, the time of the start of the potential drop did not reflect the feeding loads. This indicated that the beginning of the rapid drop in the gas rate was not rest ponsible for the potential drop in Digester two. This could possibly be part of the reasons but the consistency of the time of the potential drop for the three differently fed digesters must lead to a different conclusion. By considering the surprising immediate potential drop after feedd five in Digester two, the condition of the digested sludge 'before feed one9 and the potential drop at 420 hours, it was concluded that the overall potential pattern was a reflection of the process of activating a dormant sludge. The natural tendency of the potential to go towards the characteristic

-14ovalue of a, ripe sludge after becoming actively digesting occurred at 60 hours after feed five or 420 hours after feed one. Therefore, the buildup of the sludge activity was reflected by a cyclic potential pattern consisting of a drop, a rise and a final drop to the characteristic potential of ripe sludge.

4 3 co _J low- DI I — D G.I -i I - I ~L_~F I x0 DIG. 2 0 F ED I MIX FEED 2 FEED 3 FEED 4 -460 -470 -480 X H -510~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-490 - 0 50 23 0 =-a~~~~~~~~xL X )CX`X -500 X -As x_ X-X Xt'-,X- (* w X-( X -510 IXX -X -520 -5 30 ~Sx~ 0 50 100 150 200 250 300 5 TIME (HOURS) Figure 22. Electrode Potentials and Gas Production Resulting from Feeding a Dormant Digested Sludge

10 8 X-~~~~~~~ I-~ ~ ~~~~~~~~~~~~x i- _ _ _ _ _ _ _ _ _ _ _ _ _ X_. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _-.... _-.- _- _- - ~ _x~.x_-x- -x- $~- - --- 2 0 -460 0~~~~ ~~~~~~~~~~~~~. ---- ---....-_ o -— __ * I 300 0 2 100 H -4 70 0 0 1 0 x 3 2C -480 C 0DC ~~~~~~~~~~~~~~~~o -490 ". w -500 -510 -=_ _ _ _ _ _ _ _ xX x- -X ---x — 7 — x - -0 -520 0 50 100 150 200 250 300 350 TIME (HOURS) Figure 23. Electrode Potentials and Gas Production Resulting from Different Feeding Levels

.j 60 - I50 -- - z 0 oo~ x \'x'\ I.I 40 - '- '_ _ 0'~ o0 3 X~~~~~~~~~~~\ -40 20 At 1~t- o -~ W -5 I 0 -0I SEx E 5(m1)t>__h_0 __XX _- -5zo IT - t x-e | 11 1|-41 5 1 6 r 8 l l | 0 --53~~~0 FEED W ~ ~ ~ ~ ~e.. Fx —,x-x- "- 0-~ ~~~~~~FEED4 w~~~~~~~~~~~~~~~~~ _460 -FEED ' ' -540 50-100 50 200 2 50 300 350 400 450 50 550 600 65 0 E~~~~~~~ ~~~~~~~~~~~TI ME (HOURS ) ~~~~~~~~~~~~~~~~~~~~V-4,X~.-\-.,r '- -- ~ - xx ~ ~ ~ ~ ~ ~ ~ ~rFigure 24~. Electrode Potentials and. Rates of Gas Production Resulting from Activating a Dormant Sludge by Feeding -520 I-,~~~~~ ~~~x1 3xx-,IX.. -54I0 ~~~~~~~~~~~~~~~~~Figur 24. ElcraXoetasaa ae fGsPouto X... -. foXAtvtigaDrmn lug y ean

LOADING TO A TEMPORARY DIGESTION FAILURE The main objective of this experiment was to evaluate the behavior of the oxidation-reduction potential in a sludge undergoing some type of digestion failure, The failure of a sludge digester could be detected by various parameters such as a rise in the volatile acid content, cessation of gas production or a rise in the carbon dioxide content of the gas. Volatile acids have been generally considered the best indication but the time necessary to perform the test would detract from its value in control operations. The use of oxidation-reduction potentials offered the possiblity of a more rapid method for the detection of failure. Another objective of this experiment was to observe the relative effectiveness of platinum and gold electrodes in de' tecting failure conditionso Since these metals could exhibit differences in their potential behavior, this was considered to be an important phase of the experiment. The experimental conditions were determined according to these objectives0 Feed levels of 300 ml of raw sludge were given to Digesters one and two to create a failure condition0 The former contained one gold and two platinum electrodes and the latter contained one platinum and two gold electrodes0 Digester three with three gold electrodes was used as the reference digester with a feeding level of 100 mlo This low feeding was used in all other experiments and thus provided a reference point0 The digested sludge had stopped producing gas eight weeks before the run0 Although it could not be strictly considered a dormant sludge, it was desired to use a sludge that would not be as able to resist failure

conditions as well as a riper sludge. It was felt that a more rapid approach to failure conditions would yield greater potential changes in a shorter time and set an extreme condition to test the use of potential measurements. The feeding sequence consisted of eight additions of raw sludge. The feedings were stopped when it appeared that the digestion was in a failure condition. This was done in order to follow the potential of the sludge recovery from the temporary failure condition. The purpose of the first feed at 1000 was to check potential changes immediately after adding a cold raw slu.dge The subsequent feeds were preheated to about 355C in order to achieve failure not by temperature causes, but by loading per seo The feeding schedule is shown in Table VIo Mixing, instead of feeding, was done at two points for ten minutes after the first and second feedings in addition to allowing large time intervals after deAds one and two in order to assist in the buildup of the digester activity0 The mixing was done at 23,5 and 59.4 hourso The resulting potential changes during the feeding period are given in Table VI and the Overall potential trends are given in Figure 25, This first feed at 100C produced little immediate change in Digester three. However, Digesters one and two at the higher loadings showed a considerable effect from the feed by an immediate rise in potential followed by a recovery and then a positive trend. This immediate effect was influenced largely by the low temperature of the feed, The gas production rate which is shown in Figure 25 lagged for the first few hours and then began to fall off 30 hours after the feedingo The gas production curves shown in Figure 26 were similar except that Digester three produced less gas0

The gas rate after feed three was higher initially and began to drop off about seven hours after the feed fpr all three digesters. The total gas produced before feed four at 107 hours was 1236, 1130 and 758 cc for Digesters one, two and three respectively and the gas rates were approximately 19, 17 and 8 cc/hour. These figures indicated that the methane bacteria were not fully developed. TABLE VI SUMMARY OF THE ELECTRODE POTENTIALS DtRING THE FEEDING PERIOD BEFORE DIGESTION FAILURE Time Time Feed Potential (Ec) Potential Potential (hours) Interval No. Before Feed Changes Change (hours) (my) (my).Based on Potential Before Feed One (my) Digester Digester Digester 1 2 3 1 2 3 1 2 3 0 -— 1 -501 -488 -528.. x. 47.5 47.5 2 -495 -478 -529 + 6 +10 - 1 + 6 +10 - 1 96.1 48.6 3 -494 -476 -529 + 1 + 0 + 7 +12 - 1 107,1 11 4 -493 -467 -523 + 1 + 9 + 6 + 8 +21 + 5 119.7 12.6 5 -475 -466 -521 +18 + 1 + 2 +26 +22 + 7 125-2 5.5 6 -471 -464 -516 + 4 + 2 + 5 +30 +24 +12 145.5 20.2 7 -455 -455 -506 +16 + 8 +10 +46 +32 +22 148.3 2.9 8 -453 -455 -513 + 2 + 1 - 7 +48 +33 +15 195 46.7 -.442 -433 -504 +11 +22 -+9 +59 +5-5- +A4Jv After the fourth feed, Digesters two and three yielded little change in the potential level. However, Digester one showed a large immediate rise of 14 mv. The immediate potential rise could possibly be

better described as a displacement of the potential level by the feedo This.was followed by a.slight drop and then a rise to a potential 18 myr positive than before feed fouro This was the largest overall potential change and initial displacement caused by any of the feedings~ After the seventh feed, Digester two for the first time exhibited a drop after the feed instead of. a more positive potentialo.The 'gas rates after feed seven for Digesters one and two approximated those shortly before the feedo Digester three exhibited a decline in the rate0 Three hours later at 148 hours the eighth rfeed was given to the digesters. The pH of the withdrawn sludge indicated that the digesters were not operating properly, The pH for the more heavily fed Digesters one and two was 6.5 and 6.58 while the lighter fed Digester three yielded 7.1. It was fe.lt that the lower pH.of the sludge was indicative of an impending cessation of the rising gas rate and the feedings were suspended temporarily0 The gas rate for the three digesters exhibited an initial short rise followed by a slight decrease. The rate then increased to the highest values obtained up to that time before falling sharply near 170 hours indicating a temporary failure of the digestion~ Because of the drop in the gas rate, no more sludge was added to the digesters.. The.second phase of the experiment involaved the potential changes after the discontinuing Off the feed and up to the.ceSsation of the gas production which occurred 625 hours later~.However, since the potential changes practically ceased 300 h6drs after the feed stoppage, the following discussion will be largely concerned with this period.

After the eighth feed at 148 hours, there was 36 percent by volume of raw sludge in Digesters one and two not over 28 hours digested. Frty, nine percent was not over 51 hours in the digesters. Twenty percent of the digester volume was fed within the last three hours. For the two days preceding the last feed there were six feeds of 300 ml each. This corresponded to a three day displacement time. There was only 40.4 percent of the original seed sludge left. Digester three was not nearly as heavily loaded. At 148 hours there was 15.5 percent by volume of raw sludge not over 28 hours digested. The amount of sludge not over 51 hours in the digester was 19,6 percent. Seven percent of the digester volume was fed within the last three hours. Nearly 75 percent of the original seed sludge was left. Based on the ultimate gas production and the gas produced up to the time of the final feed at 148 hours, the percentage of digestion that occurred was about nine percent for the more heavily fed digesters. From the 'volatile solid content of the raw and digested sludge, the volatile solids of the raw sludge comprised 75 percent of the total volatile solids in Digesters one and two, This was equivalent to a batch digestion with a ratio of three parts of volatile solids from the raw sludge to one part volatile solids from the ripe sludge. The lighter feeding to Digester three resulted in a ratio of one part from the raw sludge to two parts from the ripe sludge at 148 hours.

l l49 After the final feed, the potential of Digester two rose until reaching its maximum positive potential of -432 mv near 200 hours, 52 hours after the feed. At 250 hours the potential started to drop slowlyo At 270 hours the slope became greater and the potential fell linearly to near -495 my at 335 hours. After a ten hour steady period the potential continued to drop until becoming constant again at 360 hours and at -508 mvy This was 20 my more negative than the potential before the initial feeding which was -488 mv. The potential inDilgester one exhibited a slight rising trend. after an initial sharp drop and rise until reaching a maximum of 441s mv at 190 hours. The potential began to drop slowly after this -and at 250 hours and -452 imv the slope became greater and'followed a linear path. The potential began. to level off at 515 hours. and.495 mJ until reaching a constant zone of -500 my at 330 hourso This potential corresponded very closely to the original potential before feedling of -501 mvo The potential of Digester three rose very slowly to its maximum positive potential of -504 my near 195 hours. This was followed by a slow drop to 224 hours and -507 mvr where the potential began a linear drop at a more rapid rate of change~ *At 300 hoaurs the potential leveled off near -528 mya and this level d.id not change appreciably up to 365 hours. This potential was practically the same as the 'value before the first feed. The digesters were mixed for 20 minutes after the potentials had leveled. off at 365 hours, At 455 hours electrode adjustment experi, ments were run. inDilgesters one and two. This lasted through to 490 hours. Digester three adjustment runs were from 480 to 525 hours0 These experiments

-150 -interrupted the normal potential readings due to changes in electrodes and agitation. Even though normal readings were not possible during the adjustment period, the results indicated no other changes occurred in addition to those due to the adjustment procedure. The gas rate changes after the last feed at 148 hours and corresponding potential levels are summarized in Table VII. From this table and Figure 25 the relationship of changes in the gas rate and potential could be evaluated'o TABLE VII GAS RATE CHANGES AND CORRESPONDING ELECTRODE POTENTIALS Time (hours) Potential (Ec) at Time t (my) Digester Digester 1 2 3 1 2 End of rising gas rate and point of gas failure............... 172 175 170 -447 -440 -508 End of falling rate one..l....... 194 198 180 -441 -432 -506 Beginning of rising rate one....~ 215 229 190* -446 -433 505* End of rising rate one.. ....o 263 276 260 -459 -446 -515 Beginning of falling rate two... 286 315 2826 -464 -475 523 End of falling rate two O........ 5 v90 5.-.528 Beginning of rising rate two..... 313* 348 495*.500 Beginning of dropping rate three o 438 375 - -504 -507 Beginning of rising rate three.... --- 420* --- -510* -- Maximum rate 0...00O..........0... 440 498 275 -504 -509 -520 Beginning of dropping rate four oo -- 548 -- --- -511 End of dropping rate four.. 507 586. --- -506 508 * Beginning of rise to maximum gas rate

The gas rate in Digester two began to drop before the maximum positive potential was reached near 200 hourso This latter time coincided with the end of the dropping gas rate. The renewal of the rate Qoccurred about 30 hours after the maximum potential where the potential was only one my more negative. The above gas rate changes occurred within an eight my potential range which was larger than that for Digester one but could still be considered small in relation to the overall potential change. The first gas rate rise ended shortly after the start of the linear potential drop and the subsequent drop occurred in this zon.e. After the potern tial leveled off and then dropped. the gas rate began to rise again.~ This rate dropped and finally at 420 hours began to rise to a ma>xmumo This rise occurred shortly after a small -drop in the potential to a new constant level. The beginning of the gas rate drop for Digester one occurred about 20 hours before the maximum positive potential was reached. At this -point the rate drop ended. The renewal of the rising rate occurred about 20 hours after the time of maaximum positive potential and 40 hours before the start of the linear drop~ The potential range in which these three rate changes occurred was six m.o This small range was slightly larger than that of Digester three.o The end of the rising gas rate and subsequent falling rate as well as the renewal of the rising rate for Digester three were spaced in the period where the potential changes were very small. The beginning and end of the drop in the gas rate occurred before the maximum positive potential was reached as did the beginning of the gas rate rise to a nmaximum. The maximum.difference in potential during this period was four mvo

There were further changes after 410 hours in the gas rates of Digesters one and two although the potential showed little variation. At 790 hours where the gas rates were approximately constant and less than five cc/hour, the potentials were -502, -512 and -513 mv for Digesters one, two and three respectively. These values would seem to indicate the potential range of a ripe sludge. The corresponding Elh values would be -256, -266 and -267 my. BefOre 410 hours the changes in the potential pattern-usually reflected a gas rate change although the reverse did not yield as good a correlation. After 410 hours the potential varied little even though the gas rate was still undergoing changes. The initial period resembled more a batch type of.digestion where the potential and gas rate were changing. The latter period partly resembled a continuous type of system in which the potential changes were small. In a true continuous system both the gas rate and potential would be relatively constant. Therefore, a relationship between the gas rate and potential could only be expected in a batch type of feeding in which the metabolic bacterial systems went through all phases of their growth pattern. This was shown earlier in this thesis where a synthetic feed containing sugar was fed on a batch basis.:This also was exhibited by the sludge feeds on a bathh basis. An interesting relationship was shown by the potential and the beginning of the gas rate rise to its maximum value. The beginning of the rate rise to the maximum was at a potential of -495 and -510 my for Digesters one and two..o For Digester three this occurred at.505 mvy. From this it could be concluded that there existed a characteristic potential level necessary for the gas rate to reach its maximum,

-l535 However, the inverse did not seem to apply as seen from the results of Digester two, The.508 my mark was reached about 60 hours before the rate began its rise to the maximum, This would indicate that the potential can define a necessary, but not a sufficient experimental condition, This could be referred to the results found by Kaplovsky() that high volatile acids hindered digestion but a low content did not insure good digestiono Another significant relationship was exhibited between the potential and gas rate. The linear potential drop for Digester one preceded that of Digester two even though they were fed identically0 This would appear to indicate that the activity of Digester two lagged behind that of Digester one. By observing the gas rate curves for the initial rise after failure, it could be seen that the same conditions were describedo The similarity in the potent tial and gas rate curves which both indicated an activity lag for Digester two provided an excellent example of a relationshlp between the potential and sludge activity Another instance where the potential could be used to indicate.certain. characteristics of the d.gestion was by the use of potential changes'..Since the digester loading was different, this difference, as well as the relative gas production, could possibly be reflected by the potential as demonstrated in previous experimental worko After the potential recovery from the minimum negative value had started, the first sample of all three digesters yielding a- constant potent tial levels at the same time occurred near 335 hours. The potentials were at -500.-494 and 528 my.for Digesters one, two and three respectivelyo. The difference between these potentials and the maximum positive values were 59, 62 and 24 1-my0 The highest positive potential was shoiwy by Digester two

which reached -432 mv. Digester one rose to -441 my and Digester three rose to -504 mv. The ratio of the changes would then be 2.46 and 2.58 to one for Digesters one and two based on. Digester three. These ratios did not reflect the feeding ratios of three to one for.Digesters one and two to three. The corresponding ratios for the total gas production over the complete digestion period yielded values of 2.47 and 2.5 to one for Digesters one and two compared to Digester three. This corresponded very well to the potential change ratios. The gas production ratio indicated that-'tee-;m mbte hearvily.-ix,4-;I.(. ed digesters were not as efficient in the breakdown of the organic substances in the feed as the normally loaded digester. Since the potential directly reflected the gas production ratios, the potential also indicated the less c efficient digestion. This same technique yielded similar results for another potential recovery curve presented previously. In that instance both the feeding and gas production ratios were similar and were directly reflected by the potential ratio calculated on the same basis as above. The ratios indicated the differently loaded digesters were operating at efficiency equal to the lowest loading. The gas ratios were also reflected by the potential change between the potential before feeding and the maximum positive level. The changes were 6o0 56 and 24 mv for the three digesters respectively. These yielded ratios of 2.29 and 2.5 to one for-Digesters one and two compared to Digester three. These results were similar to the ratios based on the recovery potentials which reflected the lower efficiency of the-,more.. heavily loaded digesters.

*~155 -The linear portion of the potential recovery curve was also evaluated for indications of the digester operation. The duration of this straight line zone and the corresponding potential levels are given in Table VIIIe The slope ratio indicated a possible relationship to the different feeding levels of three times as much for Digester one and two than for three0 The potential change ratio indicated only that the loading was differento TABLE VIII POTENTIAELTIME RELATIONSHIP DURING THE LINEAR POTENTIAL DROP Di 0 Slope ~ --- ~ -e ~c~ —~-~ ea~~-~-~- _ ---- - ----— ~ --- -Slope Pot. Dig~ Time (Hours) Potential (Ec) Slope Pot0 (my) (mv/hro) -.Ratio Change Ratio Begin -End Diff. Begin:End Diff, 1 253 313 6o -452 -495 -43 0 725 2.75 2. 15 2 271 334 63 A441 -494.53 -o.831 3014 2 65 3 224 300 76 -507.527 -20 -0.264 1o0 1 0 The carbon dioxide of the gas was determined at several points in the experiment. The samples tested were taken from the gas holders after a certain amount had accumulatedo By calculation an average carbon dioxide content of the gas was determined for the accumulation period. Although certain.assumptions were used for the calculations. it -was felt that the resultl.ing values wsould yield a fairly representative picture of the situatl.on These results are givern.,in Table IX as the average for a definite time interval 4 -

The cause of the failure condition indicated by the pH, carbon dioxide production and sharp drop in the gas production was due to overloading the digesters by feeding. In order to determine if the potential gave an indication of this digestion failure, a time of failure needed to be defined, The low pH at 148 hours possibly indicated an impending failure condition. By use of the gas production as the basis, the time of the digestion failure occurred near 170 hours. The actual time of the failure must have been before the drop in the gas production. This was based on the work of Morgan(26) who found that a sudden drop in gas production followed a sudden rise of the volatile acids to 1000 ppm. He defined failure as the time the volatile acids reached 1000 ppm for two days or more since it was found that the drop in gas production and a further rise in volatile acids followed. TABLE IX. AVERAGE CARBON DIOXIDE CONCENTRATIONS FOR VARIOUS TIME INTERVALS Time Interval Calculated Average C02 in Gas (percent) (hours).. Digester 1 Digester 2 Digester 5 144=178 57~9 56.7 35 8 178-219 66.2 62.2 47.3 21L9246 57.9 7205 39.0 2462-99 43.3 51.0 27.5 299-340 29.7 38a8 27.0 340=399 26.2 29.0 22.5 3999420 26.2

157 A volatile acid determination on Digester two at 148 hours yielded 1.900 ppm as acetic aci.d~ If the defrinition of failure as deefined by Morgan were used, there would be a lag of about two days to reach this acid content, or to reach the drop in gas prod-uction~ This.wou:d place the time of failure between 100 and 120 hourso The potential ehanges after feed three at 96 hours exhibited possible indications of the failure~ In normal digest ion the potential exhibited a, rise due -to the feed f.ollowTed by a drop toward the prefeed potential and then remained essentially constant, It had 'been thought t:hat an a'bnormal -feed cond.ition owould. result in some deviat.i.on -from this normal~ proced~ure. Digester one after this feed did not change but Digester two rose, fell. and then rose again. The second rise by Digester two possibly indicated an abnormal condition~ Hgever~, the potential fob Digester two did not exhibit -this second rise after feed four at 107 hourso Tle pohtential otf Digester one was displaced by the feed to a more positive le.;vel which then dropped slightly and then rose by six my.o Feed five at 120 hours was followred 'by feed.six at ~25 hours. DuLring the fivfe hour period the potential level.rose but no definite trends were apparent. Feed six produced a very large secondary rise in the poten.tial for Digester one0, Digester two also rose b-ut not as muLch.. The difference bqtween the maximarm and minirmum potential. l during the 21. hour period after the feed was 16 and n.ne mv for Digesters one and two respectiLve.ly a6nd the potentia1s were at 455 and 456 mv.O

-158 -The type of potential curve after feed six at 125 hours in conjunction with the potential level reached appeaied to be the best indication of the digestion failure, This time was near the previously assumed failure time range of 100 to 120 hours. Although Digester two after feed three exhibited an abnormal curve and Digester one after feed four did likewise, a time was sought when both digesters would react abnormally since their feeding loads were identical and the failure times were the same, The continued rising potential after the eighth feed at 148 hours could also have indicated the impending failure. It was of interest to note the maximum positive potentials were -441 my at 190 hours and -432 at 198 hours for Digesters one and two respectively. These relatively positive potential levels were not reached in any other digestion experiment. Since other analytical data beside the potential and gas measurements were not taken regularly due to the experimental objectives, the time of failure was not definitely known. The potential measurements did not definitely indicate the failure time. ThBt the results were sufficiently promising to indicate that the ioxidation-reduction potential could be a valuable tool for pxocess control in sludge digestion.

w Ir z o 80.... ' 0.r_ I 0 — xo 21: o 0, D 20............ 1~~~~~~~~~~~ \X- - o-o FEE D FEED FEED MIX 2 mix 3 4 56 78 mix -430 -44C -45C -.. -46 - 'dl -470 - ____ ____ DIGESTER SLUDGE FEED LVELS MI.) >` 300 E -480 300 b, 3 100 W -490" -5-00 -520 - X. ~~~~~~~X/,x' —X X I MI~~~~~X - MX-3456 7 IX -530 X- _ _ _xx -540A 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 420 440 460480 TIME (HOURS) Figure 25. Electrode Potentials and Rates of Gas Production Resulting from Overloading a Digester

30 I~~~~~~~~~~~~~~~~~~-~~ ~ ~ ~ ~ 25 15 uI) 0 u,~~~~~ 0/ I — ~~o — C X ---.X --- - o X. —X ~~~~~~~~~~~~~~ —X- -X-..X —~ —X --- ~- I1~~~~~~~~0 x-.X~~~.-.-XSLUDGE FEE DINGS DGSE SLUDGE FEED DIGESTER LEVEL (ml) * I 300 0 2 300 x 3 100 ~~~~~~~~~~~~~~~~~~~~~~0O 0 100 200 300 400 500 600 700 800 900 TIME (HOURS) Figure 26. Total Gas Production Resulting From Overloading a Digester

THE EFFECT OF AGITATION ON THI ELECTRODE POTENTIAL Agitation of the solution in which oxidation-re.duction electrodes were immersed introduced a new variable due to the effect of the agitation on the poten.tial measurement~ French and Kahlenberg(17) found that stirring resulted in. potential changes in a potassium chloridde-gas solution. They concluded that this was due to a condensation film on the electrode surface. They also stated that the film was a variable quantity and was affected by stirring, dirt on the surface, etco This reference was the only on.e found to present quantitative data on the agitation effect, Continuous agitation of the digester during the first set of feedings resulted in some erratic electrode behavior0 Potential trends could be detected but the electrode agreement was not always good. In an attempt to improve the condition, the contienous agitation was stopped0 This proved very.successful o Although it did not completely eliminate all erratic behavior, the electrodes inan individuaal digester showed. excellent agreement. This effect was a little surp.rising sl.nce it was thought that the stirring wounld give more accurate and rapidly responding electrodes due -to mass transfer effects as had been found by various authors. Moore, Rucchoft and Wattie(96) noted more divergence in. electrode readings in E.o aoli sewage cualtures when quiescent than. when. agitated0 They felt this was due to stratification and possible poisoning of the electrodeo Ward(66) used intermittent agitation in order to get better agreement of duplicate electrodes in. a synthetic cul-' ture media since stratification caused different readings. However, the ooL6I

-162 -agreement between different synthetic broths was not as good and when the stirring was put on or off, a potential fall resulted in some and a potential rise in others. It should be added that agitation in bacteriological and biochemical experiments was almost always done by the passage of.nitrogen gas through the solution and not by mechanical means. The reason for the adverse effect of agitation found in this research in contrast to the above authors was probably due to the type and degree of agitation used. The methlod used in this research involved movement of a heavy sludge and therefore, a more violent type of agitation was needed. ThA movement of sludge solids in the digester could have provided more: than a simple stirring effect due to impingement on the electrode surface. Because of the improved condition resulting.from the cessation of agitation, it was decided to use agitation only as a mixing device during the feeding sequence. Therefore, the effect of this agitation on the potential was investigated in order to attempt to differentiate the potential changes produced by the feed and the mixing. The sludge solution was periodically agitated in order to investigate this effect on the gold and platinum. electrodes. The agitation was usually done at a time when the effects of the previous feeding would not be a complicating factor. The usual potential trend was in the negative direction but the magnitude of the change was not consistent. Certain of these agitation periods will be presented to illustrate some of the general characteristics since no previous quantitative data has been published for any bacteriological or biochemical systems.

t.63 -The test shown in Figure 27 was done in two stages. The first part consisted of a 23 hour agitation period in order to investigate the long term effect, Three days later the second part was done with an agitation. period of five hours in order to check the reproducibility of the agitation effect on a shorter time basis. The results indicated that a rapid negative trend developed when the sludge was agitated~ The potential then became more positive reaching an approximately constant level which was more negative than the premix potential0 The time of the beginning of the constant period ranged from two to twelve hours depending on the electrode. In Figure 28 this occurred near ten, hours0 The cessation of the agitation yielded a rising potential towards the premix potential levelo The time duration of this change was a variable depending on the electrode, the length of agitation, and the potential difference between the premix potential and the potential, when. the agitation is stopped. The first stage required about 20 hours while the second never did quite reach the premix potential0 -A difference of five myv was important in determining the time necessary to reach the premix potenti al and this made the time estimation. difficult in many cases,~, IHowever, the curves did illustrate that a certain degree of reproducibility was possible and the effect of agitation was mainly transitory and not permanent0 The initial period of the agitation test described above is show~', more clearly in Fiugre 28~ In addition another test which was done with these electrodes at a later time.is also included in this graph0 This latter test produced good electrode agreement and a lack of erratic

.164. electrode behavior. The beginning of the five hour mixing was more representative of the many agitation tests with the exception of erratic electrode behavior. This was shown in the initial stage of the 23 hour mixing. However, all did indicate the usual negative potential trend when agitation was begun. The tests described above were done with the same three gold electrodes. There were many other attempts to establish the potential pattern resulting from agitation with both gold and platinum electrodes The results were not always as favorable as the above tests and it became apparent that there was no way to definitely predict the potential behavior. However, certain generalizations could be made concerning the electrode behavior~ The direction of the potential change was mainly toward the more negative potentialo There were instances of positive changes as well as no change at allo The reasons for these differences were obscured as the experimental conditions were often similar. The activity of the solution did not appear to bear any relationship to the direction and magnitude of the change. The negative potential trend due to agitation was found by others. Boyd and Reed using sterile beef broth as the media stated that this was the result using gas stirring. Lepper and Martin(57) found that when a cooked meat media was stirred, a sudden drop of 50 mv was produced, The media took several hours to recover its original potential.

'165 -The quantitative evaluation of the potential change was not possible since the direction-and magnitude of the change varied, If the negative changes within the first ten minutes of agitation were considered, the range covered zero to twenty my excluding any sudden or erratic potential changes. The positive changes were in a smaller range 6f zero to ten my. From these results it could possibly be concluded that the best prediction of the potential change upon agitation would be five to ten my in the negative direction. It must be stressed that this range represented a semi-quantitative estimation with a good possibility of being violated. Gold seemed to have more tendency to sudden large changes than did platinum. However, both metals did exhibit this phenomenon. The magnitude of the change tended to be larger for gold although the differences were often not appreciable. Some gold electrodes showed little change and some large changes. This applied also to platinum0 All in all the agreement between the gold electrodes was usually as good as the platinum electrode agreement and the agreement of the gold and platinum electrodes. The variability that existed in each electrode pointed out the necessity of using more than one electrode in experiments in which potential measurements are. taken0 The main object of these tests was to differentiate the potential changes produced by the feeding and byr-agitation~ On a qualitative basis, the direction oft-teapotential change was positive for the feeding and usually negative for the agitation. Thus, the potential changes would tend to counteract each other. The quantitative evaluation of the net effect was not found to be possible as both the feeding and agitation gave variable results.

— 470 23 HOUR TEST 5 HOUR TEST -480 MIXER MIXER MIXER ON OFF ON OFF -490 -500 50-o > X/ _ I -550 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 IN016 U~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ TIME (HOURS) Fig~ure 27-. Ef fect of Agitation on the Potential of Gola Electroaes -540! -550 -I - - -- - 0 I0 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 TIME (HOURS) Figure 27. Effect of Agitation on the Potential of Gold Electrod~es

-470 23 HOUR TEST 5 HOUR TEST -480 MIXER MIXER MIXER -ON ON ON -490 -5.00..:~ -5ISO 0 10 20 0 10 20 30 0 10 20 '.TIME (':,MINUTES) Figure 28. The Initial Effect of Agitation on the Potential of Gold Electrodes.

DISCUSSION Although the indicating electrode provides the only difference -in the components used to measure the redox potential and the pH, it is, however, this element which is responsible for the need of different considerations in then application and evaluation of either determination in a biological mixture such as digesting sludge. This sludge is a complex) heterogeneous mixture involving many bacterial and enz;ymatic systems, some of which contain oxidation-reduction reactions. These include reversible, semi-reversible oritreversible reactionso The glass pH electrode is inert in this mixture and only indicates the hydrogen ion concentration. In contrast to this, the metal redox electrode cannot be considered as an inert indicator and the system or systems responsible for the potential are not known. These characteristics of the potential determination result in a situation whereby the measuring and measured systems are imperfectly understood due to their complex nature, Thus, interpretation of the data can become somewhat speculative since evaluation can be difficult even when the characteristics of either the measuring or measured system are known. Although an equation exists to aid in the evaluation of the potential, the development of this basic oxidationsreduction equation relating the concentration of the oxidized and reduced Components in a quantitative manner is based on reversible systems for which the equation for the reaction is known. Unfortunately, a digesting sludge cannot meet these requirements and the equation, therefore, can only be applied inma qualitative or semi-quantitative manner. P~C~

Although a simple9 direct explanation canrot always, be given for the electrode potential changes that occurred in the sludge9 these potential measurements can still be of considerable value. They can nowa, as in the pasts contribute towards an rimproved knowledge of the biological. processes and provide an analytical device for better plant coontrolo Although this research involves two imperfectly understood systems9 digesting sl:dge and electrode potentials, it is felt that the results will aid 'in the further understanding of eacho There are two methods by which the oxidationsreduction potential of digesting sludge can be determined. The first consists of an indirvidual measurement on a sludge withdrawn from the digester. This is similar to most analytical tests such as the volat..lle acid test in which a single spot value is obtained from a sludge -eapleo:If the electrode is allowed to remain. in the digesting sludge9 a series of potential valui.es can be obtained which would povide a cont.nuous indicator o:f the digestion process0 This second technique would add to the existing m.ethods of ontin.uously foTllowing the di.gestlion su.ch as the measurement of -the gas productiorn and ceposi'tion0 It shoauld be noted that these tests plus the pH are the only methods of continuously following the digestion ~trocess in situ. Therefore9 the addition of the electrode potentl.al to this list is a valuable goal. The single potential measaurement can be done on the sludge sample withdrawn by the sewage tasent plaentl operato:. for tests such as volatil.e solids. Since the required experimental apparatus ip similar to that used

170 -for pH measurements, facilities would be available for most sewage plants to utilize the potential measurement in control operations and to add this technique to the analytical tests commonly done for digester operation~ There-are, however, certain precautions to follow when using this technique. The previous investigators,.who have published data on sludge potential values(88)(98) apparently were not conscious of these precautions since their recommended procedure and resulting potentials were found to be erroneous. This was due to the two factors that are of prime importance in this measurement —the time for the electrode to reach equilibrium and the condition of the metal electrodeo It war found in this work that the adjustment period, or time necessary for the electrode to reach equilibrium with the sludge potential, was a variable factor. However, it was much longer than the previously recommended period of 30 minutes. Under quiescent conditions the adjustment period ranged from ten to fifty hours depending on the condition of the sludge and the electrode. Since this time would hot _khave been conducive to field measurements and rapid process control, several methods were cons sidered which would reduce this time interval. Agitation of the sample- was successful in reducing the required time to near four hours. However9 erratic electrode behavior proved to be a disadvantage. To combat this, mild agitation was tried but, as would be expected, the necessary time increased~ Although the overall adjustment time was decreased by stirring, it was found that the initial phase of the adjustment was not appreciably changed.

The general form -that an electrode followed when inserted in a sludge consisted of a rapid initial drop in potentlal followed by a decreasing rate of change~ As the electrode approached equilibriums the slope became smaller and thereby created long adjustment periods as seen in Figure 9. It was this latter phase that agitation aided by increasing the potential-'time slope. The initial slope during the first 30 minutes was not appreciably affected., Since the sluldge was agitated quite vigorousl.y' it was felt that this extreme condito.n reliably defined the extent of the agitation effect0.When utilizing agitation as an aid in the adjustment,9 there must be no accessibility to air or the results would 'be affected as shown. in Figure 10,, The investigator should 'be aware that merely placing,a rubber stopper in a sampling container woul.d stl.ll alLow contact wilth the air gap between th.e letvel of' the sample and the stopper. Therefores the sludge lev-el sho.uSld be in contact wlth -the stopper suc.h as was done in the dissolved oxygen test. If the test was done under qiaLescent c.onditins9 no special ppre cautions were necessary for possible oxygen interference,, The results of thlis research indicated thatJ the -test eould 'be done jin. an open beaker with the electrodes about three inches below thb surface, It appeared as if the sludge surface prorvided a buffer against the rentrance of' air. The next atatermpt in reducing th.e adjustma ent time c on.sisted of conditioning the electrodes in the slu6dge to be tested. The results were encouraging, although the adjustment was sernsitive to the time that the

-l72 -electrode was exposed to the air before reinsertion into the sludge. It appeared that part of the residdal effect of the conditioning was lost rapidly. This led to the use of a sdlution with -a more negative potential than the sludge for conditioning the electrode. Although the work done was very limited, it appeared that this technique could provide a method of measuring the potential within a 30 minute period0 It was felt that further work on this technique would be a Very valuable contribution to the use of potential measurements in the sludge digestion processo The use of a conditioned electrode was opposed to the standard practice of cleaning the electrode surface by methods ranging from a 'water rinse to soaking in aqsua regia. Since the condition of the electrode affected the adjustment period and cleaning affected the electrode condition, the method of cleaning the metal became of great concern. An extensive literature survey revealed thatlthere were many diverse opinions on which cleaning method was besto However, the effect on the -adjustment period was not a factor in the selection. As a result, the literature survey was broadened in an attempt to evaluate the effect on the surface of the different cleaning agents as acids, oxidizing agents, heat, electrolysis9 and abrasives. It was found that all the cleaning agents affected the metal surface and reproducibility of a surface was practically impossible by these techniques~,Experimentally, it was found that hot nitric acid yielded very variable results while rubbing with a wet cheesecloth proved to be a better method. It was interesting to note that nitric acid was used in the work

1735 that had been published on sludge potentials. Thus, in additon to the erroneous time interval used, their results were probably also influenced by the cleaning technique. (98) Whether or not this was responsible for a wide potential range given for a digesting sludge potential was not known~ But the figures were possibly significant. In describing the.improvement.of digestion by the additLon of sulfate, the potential (Eh) range changed from zero to -100 mv to a range of -125 to -170 mv. Thus, the potential change ranged from 25 to 170 mv, a sevenfold difference. The experience in this research indicated that these values much be regarded as more qualitative than quantitative. Although nitric cleaning has been used successfully by various investigators, the solutions tested were not digested sludge. The reason that this technique yielded variable results in sludge was believed to be due to the nature4 of.the solution0.Digested sludge -was already an anaerobic biological solution and the conditions of testing were quite different from.inserting electrodes into an aerobic broth, Thus, when inserted into sludge after a nitric acid cleaning, the electrode surface containing an oxidizing agent had to have the oxidizing condition of the surface reduced before the electrode could record the sludge potential. Since digested sludge did not exhibit the tremendous ability of.E. coli to produce reducing conditions rapidly in a sugar broth, the residual effect on the surface due to cleanr ing had to be considered important, The previous investigators of sludge potentials(98) in referring to the results obtained by taking measurements every few minutes and plotting them as a time-potential curve stated: "This slope will then be a measure of the dynamic character of the subject. 0 " By following their

procedure, it has been found that this conclusion was completely erroneous. Apparently identical electrodes would ycield different slopes..and an active sludge could appear to be more sluggish than a less active sludge. The concept of indicating the dynamics of a system by a potential-time curve was not new, but any approach to make it a realization must be concentrated on the electrode itself since the condition of the electrode was often more important than the actual dynamics of the sludge. The use of the potential to indicate other characteristics.of the sludge beside the absolute potential would be a valuable goalo The absolute potential is a measure.of the environment, but the adjustment period could indicate the capacity as well-as the metabolic activity level of. the sludge No test exists for measuring the activity level of a sludge~ All analytical determinations simply measure the state of the sludge at one pointo Methylene blue, an oxidation-reduction dye, is used to indicate the dynamics of more aerobic solutionso This dye has been applied to the sewage field as the relative stability test and is described in Standard Methods(ll4)) The extension of this dye technique to sludge would not be feasible due to color interference from the sludge. -Separation and clarificatiop to produce the supernatant would not give a true picture of the situation. Thus, the metal electrode is the only method of obtaining these results by potential measurements

In order to achieve this goal, esom. m.ethod must be used to provide reproducibil ity of the electrode surface so that the reference or sta:rting potential will always be the sam.e In. addition~ it would be. desirable to have the surface in an oxidative state~ Thi s has been done successfully for preparing electrodes for anodi.c oxidations (l31) (l2) After finding that the surface could not be reproduced by ordinary cleaning techinules9 the electrodes werel allowed to s.oak in a buffered solution. of potassl.aum permangenate to create a "fc. ontrol.ed oxygen electrode,-," If this or a sirmilar c ondiLoning were applied to the eleetrode9 an extensive testing program in establishing thi.s technique could well. contribute a reliable mnethod of determini.ng th.e metabo i.c actvl.srl.-ty level and capacity of a.digesting sludge~O I.n order to follow the potential charges in a digesting sludge.< the el.ectrodies wee allows d to reairl in She sluldge.for long time intervals. It wouild heave been i.n-possible to detect rapid poterntial changes by the withdrawal, of slu.dge samples an.d single potential, determinations This eontiniued nmmersionr o:f the elec'tloie$s 'in the siu.de T 'woulld also allow the mfeasur;emaenets to be taken. in. situ. whic,.h 'was advantageous.for biological. s.ystems. Howe.vers no previ.ous experience exil.ted. on. the long,-time immtersi.on..of.electrodes since no ebontinuos'y d,y. iological. systems hare been in vestigated u.ti.li.,zi.ng this ameasurilngs -eLhn q'mue The iongest that an eleetrode had been imumersed was for aboue.t 20 days in an inoculated bacterial, culture with. apparently suceessful relts(l 2)

-176. The experimental results indicated that very long time periods did not affect the electrode behavior. Although this was contrary to the standard practice of cleaning the electrode to insure that the metal surface was not poisoned, this recommended procedure was not basedmon work in a digested sludge media. The insertion of new eelectrodes into the sludge at various times was done to determine if their behavior would differ from the electrodes immersed for several months. In addition the electrodes were removed from the sludge and reinserted after exposing the metal to air or rinsing the sirface with distilled water. All tests indicated the electrodes were not adversely affected by the lengthy immersion period which extended up to six months. The ability of the electrode to remain unaffected by the sludge was a very important result. This made it possible to follow the potential changes resulting from feeding the digesters without mEp'ic~atsIz arising from the electrode adjustment. ThusVl. the potential can be added to those tests that can be used to provide a continuous record of the digestion. The first use of continuously immersed electrodes was in following the potential changes resulting from feeding a lactose broth concentrate to a digested sludge. Since there were successive additions of the broth to the same media, this marked the first attempt of tracing the potential changes in a continuously fed, anaerobic, biological system. The work that has been published has been concerned only.with batch bacter4_ial growths. It is interesting that the feedings in this experiment could be. classified as both batch and continuous because of the usual few days

lapse between feedings and the rapid potential changes0 Thus9 the discussion of the data could be based on the results of a single feed or the overall results of the successive feeds, The feedings of the lactose broth concentrate, a lactosespeptonebeef extract mixture, yielded a wide potential range of -405 to -560 myv during the series of feedings0 This was a much wider range than shown by raw sludge feedso The potential changes were also much more rapid as would be expected from a sugar-rich feed. In addition the potential. changes cor.re lated well with the gas production after a feedingo If a distinct lag occurred in the gas. the potential also exhibited a l.ag. before reaching its most negative potential. With no lag in the gas production9 the potential also reached its maximum negative potential without a lad. The potential became more positive after the gas production started to decreaseo There appeared subsequent potential changes after the gas produce tion had nearly ceased which indicated a non-gaseous oxidation-reduction system was activeo The pattern of these changes could be classified as a double reversal. These potential changes differed from. the effects due to feeding raw sludge where there were no systematic rises and falls in the potential after gasification had ceased0 This was not surprising since the feed composition was markedly different. This indicated that the measuring of gas rates due to a synthetic feed could possibly be misleading if the result was used to predict behavior due to the raw sludge feedingso This was because the potential indicated that the bacterial systems operating in the sludge were not the same. The use of potentalas can be applied to check the assumption that synthetic sewage feeds act similarly to natural feeds0

The characteristic potential level of this system was difficult to determine due to the wide variation. However, it could be approximated at -440 to -470 my or an Eh of -194 to -224 my. The potential level did not seem to affect the behavior of an electrode. If one of the electrodes was at a different level than the others, the changes were still identical, This would seem to indicate that for this system the potential changes of the electrodes were relatively independent of the absolute potential within a certain potential range. The raw sludge feedings did not yield rapid and relatively large potential changes as did the lactose-peptone-beef extract mixture discussed above. This was primarily due to the characteristics of the digesting sludge which acted like a solution with a high capacity to resist change. In addition, the metabolic 3reakdown of the complex materials in the raw sludge appeared to be slower than the utilization of the sugar in the synthetic feed, Another very important factor concerns the highly anaerobic condition and potential of the sludge before the feeding. Batch type bacterial systems which have been widely investigated are relatively aerobic before inocuiLti6ndi and as a result, very large potential changes occur because of the natural tendency of the bacteria to create reducing conditions. However, the highly negative potential of sludge eliminated this initial large change and subsequently, the magnitude of the change was not great. Perhaps this is one of the reasons why potential measurements have not been applied to following the changes in a digesting sludge.

1(.79 -One of the more extreme examples of a small potential change resultinrg from feeding was found in a batch type feed where the maximum potential change during most of the gasification period was.about ten. my, The electrqes yielded very consistent behavior and. the average of the three electrodes in. the digester was not difficult to determine,. Electrode behavior often was not as consistent as in this case and the determination of an average representative potential became a very difficult task, Critical situations occurred where potential differences of less than five my 'would alter the picture of the potential trend., n these situations the experience and judgment of the experimenter were a very 'vital factor, Agreement of three electrodes within two 'mr was considered excellent, five m:r goodg and ten my more the usual case. However, for averaging electrode values the agreement of the electrodes was not as important a factor as the consistency of the potential change of eac'h electrode since a slight error in the potential level was not significant. But, the path of the potential change was very importdn6o0 Part of the diff.ic,.lty in. determining the average potential was du.e to the impossibility of d.eterminl.ng the potential over 24 hour periods during run.s of' which.some extended over 800 hours, The solution tlo this wou.lid be a recordaing potentiometer which would record a continuous set of EreadLings and detect changes that could. occur while the experimenter was not p.resent0 Another aid for increased accuracy in the potential measaurement would be more than three el.ectrodes in a d:igester if a recordimng potentdio meter was not avrallable, Prequent readings would redu.ce the recornmended

number to three since potential patterns could be evaluated accurately even with some erratic electrode behavior. It was found that three electrodes were not always adequate and it was felt that any decrease in this number would greatly affect the accuracy of the potential evaluation. From the results of this research it cannot be stressed too much that the use of a single electrode can very easily yield erroneous potential values and should never be relied on. Agitation has been used as a means of increasing the electrode agreement. Attempts were made at first to use continuous agitation but erratic electrode behavior made it necessary to abandon this procedure, Improvement of the electrode behavior resulted from this although other workers had found stratification caused diverse readings. Settling in the digesters did not cause differences evien when the electrodes were in the supernatant layer. There appeared to be no potential difference between the sludge and supernatant layer. This was possibly-, due to the "in situ" laboratory arrangement. In addition to the electrode agreement within one digester, the potential agreement among different digesters under identical conditions was also a very important consideration. The simultaneous use of three digesters treated identically made it possible to find the degree of reproducibility that could be expected as well as the agreement between different digesters so that different feeding levels could be reliably evaluated, The results indicated that the potential behavior was not identical, but often similar for the same feeding levels. The batch feeding shown in Figure 21

exhibited excellent potential similarity and indicated that close agreemtent between samples was possilble~ However9 in a biological system as complex as sludge there always existed a certalin amount of diverfgence between identical samples In.addition the possible variability of the electrodes was also a factor especially when the potential changes were not large..xlring the parallel feeding experiments it became apparent that low feeding levels to a normally digesting sludge would resu.lt i.n small potential changes and fairly constant, potential levels, This would be expected for equilibrium condltions ~ On this basis further experi.mental work with nonrparallel feedings was done when the sludge was not as capable of resisting potential changes. In this mann.er it would be possible to accentuate potential changes to aid in both. the evaluation of the data and the potential comparison with the gas production. The ability of electrode potentials to indicate the uniformity of the digestion was similar! to 'the conclusions of Kaplovsky(43) who found that.volatile acids were the least affected:from day to day of all the analyses. He stated that wit:h constant loadings volatile acids should be fairly constan.t and that a constant level of 'aci.d accumu.lation was a good indicator of digestion uniformity0 Relationships were 'sought in. the non.parallel, feeding studies between the potential change.9 feeding level.9 and gas production. The short period after feeding yi.elded some instances where the dfl.ference in feeding level was reflected by the potential~ However, m:l.ch better results were obtained by using the potential changes after feeding -was discontinued in

batch-type experiments involving several successive feedings. This behavioral pattern was similar to that of the gas production which did not differentiate the different quantities of feed untiltwiultimate production was reached as shown in Figure 23, In this experiment the feeding rations of 3/1, 1/1 and 2/1 for Digesters one, two and three irespectively corresponded to the same ratios based on total gas production. In addition the change in the potential from the mintinum negative value to the first point where all three digesters exhibited constant potentials also corresponded to these ratios, Another experiment with feeding ratios of 3/1, 3/1, and 1/1 for Digesters one, two and three only yielded gas production ratios of 2.5/1,, 2,5/1 and 1/1 for the same digesters (Figure 26). This indicated that-t'ihe efficiency of the digestion was somewhat less at the higher loadiiag level. The results of the potential change were very significant as they reflected the gas production, not the feeding level, and thus indicated a reduced digestion efficiency, Although the calculation of the potential change ratios could be somewhat ar'bitrary due to the small change of theJdlightest loaded digester, the results were still of considerable value in evaluating digester operation. From the gas production curves, gas rates were calculated to investigate the relationship of the changes in the gas rate and the potential. Since the gas rate was an indicator of the sludge activity, the comparison would indicate to what extent potentials reflect this activity. Under conditions simulating a batch experiment where rapid changes are characteristic, variations in the potential and gas rate could be related.

However, in continnous digestion it was possible to have the gas rate constant and the potential changing and when the feeding was stopped,. the gas rate changed while tie potential remained constant. This latter condition would indicate that the environment was not changing but the metabolic activity was. Thus, the potential measurxed changes in the enlvironment, and the gas rates were an indication of the activity..From this, it appeared that the use of continuous potential measurements alone w.ould not yield. a clear picture of the 'digestiono From the results of this research it has been found that a single potential determination oSf a s araple of sludge woul.d not be of much..value unless the potentlial was outside the range of 'about -460 to 540 mv or a corresponding E: of 220 to -300 mvy The range was determined from the potential values of normsally digestingl sludge in the laboratory digesters. Below this mini:mum negative value the sludge did not appear to be digesting ~normally. This was not an attempt to say that a potential more positive than -460 my indicated the sludge was approaching failure, However, when sludges fresh.ly drawn. from plant digesters exhibl.ted potentials near i460 mv, the sludges also had high volatile solid contents that were not in.dicative of a well digested sludge. Ak;.er 'any series of feedings was terminated, the potential re. turned to a constant value. During this period the last stages of gasification occurred. This potential was felt; to be the most indicative of a ripe digested sludge., The most representative of these values was about -500 mv or an Ei of '260 my. But it must be cautioned that a 'value of '500 my was exhibited by sludges dri4ng:active gasification and this wvalue could not be used to state that the sludge was in a rl.pe stage.

The Eh potential range of -220 to -300 myv did not agree with that stated by Hood(88) of -100 to -200 my for optimum digestion. The reason for this discrepancy was felt to be due to the method of determining the potential which was discussed previously and shown to be erroneous. It was of interest to note that the -100 to -200 my range has been included as a standard in the sewage treatment text by Babbitt(t) It was encouraging, however, that the oxidation-reduction potential was included as an analytical test for the:sewage field~ Another very important phase of the study was concerned with the potential changes starting immediately after the feeding. Because of the ability of the electrodes to remain immersed in the sludge, it was possible in this work to follow continuously the conditions resulting from the feed. This type of potential response could provide a valuable contribution to sludge digestion control if it was possible to determine the ability of a digesting sludge to absorb a feed without deleterious results. This possibility was based on the concept that when a lore aerobic solution was added to a less aerobic solution, the resulting mixture would yield a potential dependent on the original solutions. The quantities of each solution, their capacity and poising power would determine this final potential which would be between the limits of eaqh individual potential, The effect of oxygen on the sludge potential was an excellent example of adding an oxidant to a more reducing solution. This was shdwn in Figure 10. The degree of mixing andijoxyen:.diffusion.was 'ref lcteld.-in).-::, Ptohe.;p-otential riaeia4tich illust'rated the possible effect of different feed levels to the digester, The sludge was 4abe to overcome the immediate rapid potential rise and..return the potential to the premix value.

Additions of o~l7dants to anaerobic solutions has been done by others Using oxrgen in nitrogen gas it was found that a reversal of the negative potential trend was caused by a two percent oxygen content, temporary checking (76) at 1o4 percent and a 'very sli~ght effect at 0o1 percent A fixed amount of air added to a peptone broth containing 0.Ol percent sodium thioglycollate caused the potential, to rise rapidly and continue to drift in a positive direct tion whereas at a 0.1 percent concentration of the reducing.agent the potential rose about 45 mv followed by a return to the original potential in, about three hours(78)o Clark et al.,(49) stated that 'the action of small doses of oxidants would produce temporary checks C reversals on a time-potential curve followed by a very decided recovery of the negative drift Thus, it would be anticipated t'ht the raw sludge feed would cause a temporary positive trend.followed by a slower negative recovery to the original potential. of the solu'tion. Also, the larger the quantity fed, the longer the recovery would be prolonged. It might be possible that for a feed too large for the capacity of the sludge, the negative recovery would be retarded or eliminated as was the case with high concentrations of oxygen as observed by others. This could ndicate 'a failure condition of the digestion. The initial effect of feeding was *toward a more positive potential0 However, the addition of raw sludge did not consistently produce the rapid rise and recovery of the potential described above. Some examples of the the potential becoming more positive after feeding,and then returning towards the more negative prefeed potential were seen in Figure 19o

-186 -In addition to the sludge feeding the potentials were also influenced by agitation of the digester contents. Thus the effect of agitation onthe-potential had to be determined since it.was necessary to mix the digesters after feeding in order to blend the raw sludge with the digesting sludge. Intermittent mixing was necessitated by the erratic behavior of the electrodes resulting from continuous agitation. Separate agitation tests wOjre.,attempted to evaluate this effect quantitatively but this.was not successfulo The potential change usually went in a more negative direction. This was a transient effect after which the potentiali returned to its premix level. The magnitude of this change was not consistent. Thus, mixing of the feed put two opposite forces in action. This mifked the effect of feeding upon the potential and no quantitative measure was possible. A different method of agitation might improve this situation and the relative potential could then be accurately compared for different feeding levels. The largest immediate changes in the potential resulted when the feed was not preheated. This was anticipated by evaluation of the temperature effect on the standard oxidation-reduction equation. Part of this effect was probably due to the metal electrode since during the initial stages of the mixing there still existed a colder mass of sludge due to the feed, The temperature of the digester sludge on an overall basis was lowered very slightlyo Large feed quantities usually caused greater immediate changes than the lower amounts, However, the results were not always consistent possibly due to the effect of agitation. In addition equal feeds could also give different responses which falsely indicated unequal loadings,

Constant loadings to an actively digesting sludge should result in a fairly constant potential levelo The trend to a more positive potential due to the feed was opposed by the capacity of the sludge to resist this change. Since this capacity was not rapidly mobilized9 the potential became more positive at first and then.returned to the prefeed potential as the reaction of the sludge increased. If the loading were more than. the sludge capacity could handle9 it would be anticipated that the potential would rise rapidly after the feed. This would be followed by a drop and a rise in the potential. The second rise would indicate an abnqrnml condition~ On the other hand, after the initial rise the potential could continure to become more positive witholut any tendency to return to its prefeed potential. Both of these types of poten:tial responses occurred where overloading conditions were present as shown in Figure 250 Although these results appeared to be indicative of abnormal conditions9 the potential changes often were not dlistinct enough to yield conclusive results after a single feeding0 Thus the overall potential trend would be needed to su;pple-. ment the results of one feed. However9 the potential did not appear to be able to detect a fa'l.2.Lure of the sludge digestion. process clearly enough to supplant other testso From the results of this research9 it was felt that further experil mental work along this line could piossibe establ.sh a def.inite pattern.. which would indi.cate impending fallure without the volatile acid test0 It was noticed that the earbon dioxide content of the gas increased appreeiably

before the rapid drop in the gas rate occurred. Since this test could be done on a q6ntinuous basis, it would be an excellent supplement to the potential measurements which could also be taken continuously. This would aid in the investigation of the metabolic changes occurring in the critical phases of digestion and give more insight into the entire process. A rising Dptential did not necessarily indicate an approaching failure condition. The process of activ&ting a dormant sludge produced a rising potential level after an initial drop to a more negative level. This rising potential reached a maximum and then became more negative again. The trend t-oward a more positive potential did not indicate an impending failure but was simply a reflection of the activation of the sludge. The most positive potential reached was near -465 mv which fell within the limits; previously defined where digestion was still satisfactoy'y. The first reaction to the feed by the dormant sludge was toward a more negative potential. Since there was practically no gas production during this period, it was felt that the change was due more to the hydrolytic systems than the methane bacterial systems. However, because both systems were probably affecting the potential, the relative effect was impossible to determine. There were various oxidation-reduction reactions in both systems and the study of each phase by potentials would be of great value in evaluating the mechanism. Since the most sensitive part of the process was the production of gas by the methane bacteria, the determination of the potential characteristics of this phase would set the optimum level for sludge digestion and gas production. This would not only aid digestion operation per se, but would also help ih the evaluation of the basic factors affecting the process.

l189~ A very important part of this study was the comparison. of gold. and platinum electrodes. Based on experimental work in Olqlinhydrone which. in.dicated gold was more sensitive than platinaum and a literature survey which also indicated gold might be a better electrodes gold was selected as the electrode metal~ However since no data was available on the relative value of each metal in digesting sludge, platinum., was used in a supplemental manner to provide a comparison~ The immediate response to the raw sludge feeding, as well, as the overall potential results during the digestion did not show any significant difference between the two metals. Similarly9 the response to digester overloadring did not appear to be differentiated. The differences- that were exhibited were not beyond the normal, 'variabi'li.ity of the electrodes. The only contrast was provided by agitation of the sludge whilch. caused a greater amount of erratic behavitor in the gold electrode although both metals were affected Comparison of the gol.d and platinuam electrodes during the ad. justment period iEdic.ated more dissimi.lal.ity than dl;uring the digestion experiments. This was especially apparent in fresh p1rant digesther sludge where the average equil.ibrium potential of the platl.inum electrodeswas more negative. However. oan an ind.ividual basis a few gold electrodes were more negative than some of the platinulm group~ The d.ifference between. the average potential of each metal was usually within 20 mv*

The platinum electrodes usually required less time to reach the equilibrium potential. The potential of the gold electrodes in a high volatile solids plant sludge exhibited a plateau before reaching the final value. The phenomenon of poising the electrodes temporarily at an intermediate potential was not apparent with platinum. The cause was probably due to a weak oxidation-reduction system which could not be detected by the platinum as was the case in quinhydrone. Electrode adjustment tests done in the laboratory digesters also indicated that platinum required less time to achieve an equilibrium potential. However, the difference between platinum and gold was less than in the fresh plant digester sludge and agreement between the electrodes was usually within ten mvo The form of the adjustment curve was also similar for the two metals. This improvement in theeclectrode agreement was apparently due to the characteristics of the sludge since the equilibrium potential was usually independent of the electrode metal. The most apparent difference between the plant and laboratory sludges was in the degree of digestion. This was based on the volatile solids which indicated that the laboratory sludge was usually more digested. It therefore appeared that the less digested sludges contained substances which caused greater divergence between gold and platinum electrodes. These substances evidently were removed upon further digestion and decreased the effect of the specific metal used as the electrode. An important factor in the explanation of electrode behavior arose from the gas-metal effect. This was especially significant in digested sludge studies since carbon dioxide, methane and possibly hydrogen

-19l1 were present in the sludge~ The electrode metal and9 in turn9 the potential could be affected by these gaseso Another gas. oxygen, was also of interest when the electrode adjustment was studied~ The studies of French and Kahlenberg(l7) have shown that when the gas flow was changed from oxygen to hydrogen, the platinum electrode adjusted rapidly while the gold responded slowlyo From hydrogen -to oxygen9 the platinum responded slowly while the gold changed rapidly. This possibly indicated why platinum would adjust to the anaerobic equilibrium potential more rapidly than gold since the oxidative condition of the metal had to be changed to a reduced conditiono Another possibility arose from the fact that platinum was a better catalyst than gol.d Thus9 the removal of oxygen from the surface would be facilitated yielding a more rapid adjustment of the platinumi The effect of air on the electrode when immersed in sludge was greater on the platinum than the gold as seen in Figure l0o This did not correspond to the fact that platinum. responded slower than gold when the gas was changed from hydrogen 'to oxygen0 However9 when. the electrodes were removed from the sludge temporarily9 platinum 'was less- affected by' the air exposure than golds It was of interest 'that a 0,25 minute exposure of the platinum made a readjustment necessary0 This indicated the effect of oxygen was a definite factor with both tmetals

U192 -The gap-metal effect could become quite complex and, in addition, could be influenced by the inertness of the metal. Although the complexity of this subject extended beyond the scope of this thesis, it was felt that an adequate understanding of the oxidation-reduction potential would not be possible without some consideration of the gas-metal effect, The choice of either platinum or gold as the electrode metal has not beem completely resolved by this research. Although their behavior could be very similar, there were instances where large differences occurred..o This was especially true for the single potential measurement for which the adjustment period played such an important role. It appeared that the more well digested sludge produced better electrode similarity, However, for digested sludges with high volatile solids, there existed distinct differences. Unfortunately, it was not possible to determine which metal yielded the most representative potential results, It is possible that both metals by their differences could contribute more together than if used separately,

SUMMARY AND C0NCLUTSIONS 1. Oxidation-reduction potentials are of cons derable value in detecting abnormal, digestion. condit ions and. in providing a continuous measure of the slu'dge dl.gestionr proeess which can lead to better process control, and to a better understan.ding of -the process itself. 20 The additi.ox. of a Lactose broth coneentrate to digested sludge exhibits different potential behavior than do raw sludge additions. This indicates different bacterial. systems are operating and. therefore comparison of gas yields from the two types of feeds can be misleading~ 3. Additions of raw sludge to a normally digesting sludge will usually resuit in a potential rise followed by a.re.turn to the more negative prefeed potenti.al The potential rise is increased by low temperature feeds and decreased by agitation. 4. Under normal digestion conditions the potential will remai.n fairly constant. A continued positive trend of the potential. can be indicati've of overloading. 5- The optimum range for sludge dlgestion was from -460 to 5530 my or an Eah of '220 to =290 mar The most representati,v:ie potertial of a ripe sludge was =500 my or an E. of 260O myrrr 6. A single electrode epotential reading otf a digesting sludge can r.ot'define the degree of digestion.;..f the rjalue is wi.thin this potential. range. The potential chanLges are usually more inad.izcativ e thar the absolute potential valu.e for digester operation.

7. The potential change in certain cases reflects the ratios of the ultimate gas production and relative efficiency of differently loaded digesters after feeding is discontinued. 8, The potential changes and gas rate changes are related for batch experimentso For continuous operation where conditions are relatively steady, there is no relationship exhibited, Gas rate changes can take place and not be reflected by a constant potential level indicating that the metabolic activity, but not the environment, is changingo 9. The time for an electrode to reach eguilibrium with the sludge potential ranges from 10 to 50 hours under quiescent conditions depending on the condition of the electrode and the sludge. Agitation can reduce this time to four hours with the electrodes often exhibiting a certain amount of erratic behavior. The time can be reduced further by preconditioning the electrode in digested sludge or in a solution more anaerobic than the sludge to be tested, 10. Agitation usually results in a negative trend of the potential which returns towards the premix value when the agitation is stopped. 11o Electrodes immersed for long periods in digesting sludge do not loose their sensitivity. 12. A minimum of three electrodes should be used for any sludge measurement, This is not just important, it is necessary for reliable data, 13. Goldand platinum did not differ significantly during the laboratory digestion studies, Adjustment trials on sludge withdrawn from plant digesters revealed distinct differences which became less as the sludge

195. became more digested. Platinum usually adjusted rnpre rapidly and was less affected by agitation than gold. Gold appeared more sensitive to weak oxidation-reduction systems.Platinum was more sensitive to oxygen when immersed in the sludge~

APPENDIX A. DERIVATION OF OXIDATION-REDUCTION EQUATION The basic oxidation-reduction equ tion relates the oxidative state of the system with the potential. The derivation of Michaelis based on chetmical affinity will be followed. If a ferrousr-ferric ion solution is at equilibrium, there will be no tendency for change. This can be represented as system I: Fe++' = Fe+++ + e Fe+++ Fe++ k If another ferrous-ferric ion solution is not at equilibrium, there will be a tendency to change, ioe., system II~ Fe+++ \ Fe++ / However, there will be no change since the electronic equilibrium would be distufrbedo If an inert electrode which does not participate in the system, but merely acts as a conductor of electrons, is placed in each solution, and if the solutions are joined by a semi-permeable membrane and the electrodes connected to complete the circuit, then there will be a tendency of Fe+++ in system II to be reduced to Fe++ and this will be accompanied by an dxidation in system I. This results in electrons being transferred through the electrodes and an electromotive force, E, is produced, The volumes are assumed to be so large so that no measurable concentration ratio change is produchdo -196.

7 197 -Sufficient current is passed so that one mol of Fe+++ is reduced to Fe++ in system II and the converse occurs in system I o Thus one mol of Fe+++ will disappear in system II and appear in system I. The work done in this reversible isothermal reaction for the transportation of Fe++ from concentration (Fel+++) to concentration (Feo+++) through the semi-permeable membrane is equal to~ Fel+++ RT 1 d(Fe+++) ~1 =~RTln ~(Feo+++) Feo+++ The work for the Fe++ change i.so RTln (Fe~++) (Fel++) The total work will be the sum of these two9 and is equivalent to the electrical work for the passage of one faraday across the cell with electromotive force E0 Rt (Feo +.+) Rt (Fel+++) -n ( I.. F (Feo+++) R F (Fe l+++) Since.(Fe.) = k; E k + -- in (Feo++) F (Fe-++)

-198 -As this electromotive force constitutes only one-half a cell, another half cell will have to be added in order to complete the circuit. This reference cell is taken as the reversible hydrogen electrode at one atmosphere pressure in a solution containing hydrogen ions at unit activity. The potential of this electrode is assumed to be zero at all temperatures. The general reaction equation fori the comtleted cell is: Reduced state + nH+f (aH+ = 1) = Oxidized State + 1/2nH2 (1 atm)o Since the activities of the hydrogen gas and ions are unity in the standard hydrogen electrode, the following equation results: Eh = Eo + F n (nRd) where Eh = potential of the system referred to the normal hydrogen electrode (volts) Eo = specific constant for system being measured (volts) R = universal gas constant T = absolute temperature F = Faraday constant n number of electrons transferred in the oxidation-reduction reaction

APPENDIX B. CLEANING METHOD EFFECT ON THE MTAL ELECTRODE To evaluate the different cleaing methods, the effects on the elects rode metal of chemicalsg heat, gases, and abrasives were studied. Acids will be considered first since they are most commonly used in cleaning electrodes. Based on the electromotive scale, metals above hydrogen react with acids replacing hydrogen. Since noble metals fall below hydrogen, they are therefore inert to this type of action. Ho-wever9 acids can have functions other than behaving as an acid in the normally defi.ped mannero Hydrochloric acid can act as a reducing agent releasing chlorine in. the presence of strong oxidizing agents such as manganese dioxide, potash sium permanganate, potassium dichromat3 and nitric acid0 Cold or dilute sulfuric acid acts only as an acid, with the concentrated form acting less rapidly to displace hydrogen than the dilute form, However, concentrated sulfuric acid is an active oxidizing agent when heated, being reduced to sulfur dioxide. oIn this capacity it reacts with metals above platinum in the electromotive serieso It is not as strong an acid as hydrochloric or nitric. In combination with potassium dichromate it is a strong oxidizing agent releasing atoms of available oxygen. This material can also release free halogens from the halic acid. Nitric acid can act as an acid, oxidizing agent or nitrating agent. It will react with all metals except gold, platinum, iridium, titanium, tantalum and rhodium. The concentration and temperature of the acid determine its oxidizing ability. At 68~C or higher, it is readily decomposed by heat to water, oxygen and nitrogen djioxide. If mixed with three parts of hydrocloric acid, aqua regia is formed. This will ~-99o

-200 -dissolve gold, platinum and other metals which are not affected by the acids individually. The chlorides of gold and platinum are formed, probably due to the action of atomic chlorine. From the above general description, it would appear that platinum and gold are relatively inert. However, a study of Mellor's( 1l122) extensive work on inorganic compounds yields a different picture. Platinum is readily dissolved by aqua regia and solutions containing chlorine. Under reducing conditions it is attacked by alkalis, nitrates, cyanides and phosphates. Carbon alloys with it, and therefore the metal should not be heated in a smoky flame. There is some solution in boiling sulfuric acid, and boiling a long time in water yields a small (123) quantity of colloidal platinum. Pierce and Haenisch in discussing the care of platinum crucibles, state that the metal should be kept away from oxidizing agents such as chlorine, bromine and ferric chloride. No compounds of sulfur, phosphorus, arsenic, selenium, mercury, lead, bismuth, silver, tin or ait'iMon should be ignited in platinum since they will alloy with the platinum if in elemental form. A reducing flame causes formation of crystalline deposits which will penetrate the metal. Strong bases, or salts which give strong bases, i.e., barium carbonate, should not be heated in platinum. Scouring with talc and cleansing with hydrochloric acid is the best method of cleaning. Gold is not attacked by sulfuric acid if pure and no oxides such as manganese and lead are present. It is not completely inert in hydrochloric acid unless air is kept out. Alkalis do not appreciably attack gold. It is dissolved by aqua regia, mixtures of hydrochloric acid with an oxidizing agent which liberate chlorine, water containing the halogens (chlorine, bromine, iodine in solution), and solutions which can generate halogens, i.e.,

chlorides, iodi:des and sulfuric acid. Gold is attacked by boiling ferric chloride, alkaline sulfides and thiosulfates, and substances which can give large quantities of oxygen, i.e.., manganese dioxide, potassium permangenate, nitric and sulfuric acids, and potassium cyanide solution in the presence of air. Chlorine water slowly attacks gold. Since both of these metals are affected by electrolysis, and polarization has been used to treat the metal electrode, its effect on the metal is of interest. In electrolysis, platinum becomes rough and dull in hydrochloric acid, forms a finely divided film in potassium hydroxide, and forms platinum black in hypochlorite solution. As an anode, its behavior depends on the previous history of.thi electrode. There is a possible oxide film when used as an anode, according to Bowden(l24). Glasstone(50) states that platinum and gold are regarded as unattackable, due to the fact that they are nearly always in the passive state. Although less noble than gold, platinum is rendered passive more easily, especially in solutions of oxyacids. If ammonia or hydrochloric acid is in the solution, the anode will suffer appreciable attack, particularly if in the finely divided form. Rideal and Wansboroaigh-Jones(125) point out that a platinum anode is soluble in hydrochloric acid and the amount of oxygen in the platinum bears a constant and simple relationship to the amount dissolved. With gold as an anode the previous history is also important. It has been stated that it is more readily polarized than platinum. It will dissolve in a hydrochloric acid-cyanide solution, and. inca neutral solution of sodium or potassium nitrate.

The effect of heating on the metal warrants considerable attention since this operation is usually performed at some time during the life of the electrode. This will most probably occur during the construction of the electrode Although heating of the electrode metal is sometimes used as a cleaning method in an attempt to degasify the metal b~ to burn off imbedded material. In or4er to evaluate the actual extent of the metal s the following factors are considered: (1) to what extent are different gases sorbed on a metal surface; (2) does heating degasify the surface; and (3) how can the gas be removed. Sorption of gases occurs on all metal surfaces. The metals of interest, platinum and gold, sorb all gases but each to a different degree. From reviews of the literature, gold appears to sorb less quantities of oxygen and hydrogen than platinum(121)(122)(126) The sorption of gases on metals can be important when the metal is heated in a gas flame because of thecpossible contamination of the surface. It is of interest to note that in various laboratories different gas compositions are used in the burner used in heating the electrode. This can cause different degrees of contamination due to the varying percentages of conponenht gases and the specific adsorption characteristic of tAe..metal for each gas. Also, thel.:1ocation of the metal in the flame can be important due to incomplete combustion of the gas in certain zones. The next point of interest is how easily these sorbed gases are removedi. -Aiording to Mellor(l22) ordinary platinum contains much occluded gas which is-very difficult to remove. Once freed, the surface is recharged

with more gas than before. The gas is held very tenaciously and can be recovered only by heating to redness in vacuo(122)(l26). When describing the temperature of platinum, authors often use the color of the metal. The color is dull red, cherry red and orange in the temperature range of 600-850, 850-1000, and 1000-1100~C, respectively. At 300~C9 Langmuir(l27) finds that oxygen is wholly irreversibly Wdsorbed on platinum and can only be removed by reacting with a reducing agent such as hydrogen. At least 360~C and the best vacuum is needed to remove oxygen from platinum. The oxygen film is not removed by heating alone: even at 12000C. At higher temperatures oxygen begins to attack the platinum. Although degasification of the gold surface has not been as extensively studied as the platinum surface9 it appears that the same general conclusions apply to both metals(121)(128)(12l9) For oxygen removal at 130C, Benton and Elgin(l30) used hydrogen gas since the use of a vacuum was unsuccessfulo The above data seem to indicate that heating alone will not degasify the surface. This points out the fallacy heJld by some investigators that heating alone can remove gases from the electrode surface. A technique has been reported in the literature which overcomes the difficulty of removing the sorbed gas. It is done by controlling the initial oxygen coontent of the metal and in this manners it is possible to standardize the initial oxygen-electrode equilibrium. This was done by Goard and Rideal(131) who first tried boiling with hydrochloric acid, awashing and heating in a methyl alc'phol flame. However, the best accuracy obtained was up to 30 mv. Then, in addition to

the above procedure, the electrodes were placed in a solution of 90cc borate buffer pH 9 and 10cc 0.lN potassium permangenate for 30 to 45 minutes with vigorous stirring. They were then washed and put into the measuring system. The authors found good.reproducibility and accuracy by this technique for irreversible systems. Glasstone and Hickling(132) had difficulty in reproducibility for the anodic oxidation of thiosulfate with platinum. This was overcome by controlling the initial oxygen content of the anode. This was accomplished by washing the electrode with water and warm nitric acid, heating to redness, stirring in O.lN potassium permangenate and buffer solution of borate pH 11 for 30 minutes, washing with water and using immediately. Although heating alone will not degasify the metal surface, it can affect the metal itself. Platinum is not acted upon by air or oxygen at ordinary temperatures and scarcely affected chemically at any temperature. It is doubtful if oxidation occurs below 1300~C. Gold is not acted upon by air or oxygen at any temperature. It only combines indirectly with oxygen as in the case for platinum. Platinum.does not volatilize appreciably at 9000C but can be volatilized to some extent at higher temperatures. Gold begins to volatilize just below its melting point of 10635C and the loss Ls- probably nil at or below this pointo Besides volatilization of the metal, there can also be changes in the metal surface due to heating. McBain(l26) stated that Berlinerts work(133) showing that platinum heated to redness in vacuo sputtered when given up occluded gases, made it evident that a smooth surface could

-205 -(134) no longer be considered perfect after such treatment, Stromberg presented photomicrographs of platinum foil evacuated for two hours at bright red heat. He stated that the furrowed surface was due to the twinning of countless visible crystallites. On the assumption of a wholly superficial effect, he allowed 40 percent of extra surface. Mendenhall found from photoelectric studies that there were complications due to gases on the surface~ Heating in vacuum not only affected the removal of gases, but the surface characteristics were also changed by the heat treatment Frazer, Patrick, and Smith stated that platinum was roughened by alternate heating and cooling. This.was so called "activation of surface." They believed that the activation of platinum by heating in.air was due to the formation of finely divided platinum on its surfaceo.Finch.and Stimson(l28)(l29) found that heating resulted in an electrical charge on the metal surface which became dissipated as the metal was cooledo However, their data indicated that platinum would have a residual charge at 25~C if heated to 970~C or cherry red. This was not apparent for goldo To achieve more insight into the heating effect on the metal surface, it will be necessary to briefly look into sintering, the preliminary step in annealing. The sintering temperature of gold is 2000C and platinum is 700C0 This is the temperature where the microcrystalline and amorphous arrangements of the atoms on the surface begin to change rapidly into the macrocrystalline form, This indicates that a structural change can result from heating. Rideal(l37) stated that after sintering, the surface activity of the deposited metal progressively became less in

.206o6 adsorption and catalytic activity. Finch and Stimson[129) found that freshly rolled platinum was in an unstable state and needed alternate heating in oxygen and hydrogen to achieve a constant and reproducible charge-temperature relationship. This was also the case for nickel where heating alone was not sufficient. However, it was found that heating alone was sufficient for gold. They felt that the sintering temperature was possibly the reason for this and perhaps the normalizing involved a structural change. Other cleaning methods also affected the metal surface. After burnishing, boiling in 96 percent alcohol, boiling in concentrated HN03, washing with distilled water and drying at 150~C, Stromberg(l3) allowed 100 percent for unevenness for gold foil. Photomicrographs showed innumerable crystallites and furrows.Frazer, Patrick, and Smith3) found that surfaces were not plane from a molecular point of view after cleaning with acid or Subbingo Highly polished platinum in x-ray diffraction no longer showed characteristic lines, but the structure of an amorphous substance. McBain(126) added that it was very likely the surface was also contaminated with impacted polishing material. By changing the metal surface and increasing the accessible area, acids could create a catalytically active surface(l38). Langmuir(l39) described Faradays' experiment of boiling platinum foil in concentrated sulJ furic acid aird washing. The platinum was brought to a condition where it caused combination of oxygen and hydrogen at rbom temperature' Platinum boiled in aqua regia, hydrochloric acid, or nitric acids was made more active in its ability to displace gold from its salt(122). This did not occur with sulfuric acid~

-207,Attempts have been made to quantitatively measure the change on the available surface resulting from various treatments which affect the metal. From Bowden and O'Connor(l38) some accessible area-apparent area ratios based on nickel were obtained. They were- rolled metal, 2:. sandpapered, 6-10; alternate oxidation and reduction, 46, and metal solidified from liquid state, 1o5-1o7, With the latter methodthe surface was fairly smooth but when the surface layer was removed by polishing or etching, cracks and irregularities developed which relieved the strains. They stated there was evidence of narrow cracks penetrating deeply below the surface of polished metals~ McBain(l26)summarized the work of Bowden and others on platinum, nickel, gold and other metals. For bright foil platinum the ratio between the true surface and apparent surface was 2.2. If cleaned in acid and heated in a flame the ratio was 3.39 an increase of 50 percent. Silver after being finely sandpapered had a ratio of 16. Newly polished nickel had a ratio of 75 which decreased to a ratio of 9.7 when oldo These figures indicated that the metal surface was an unknown varying factor and could help to account for different potential readings when the electrode.s were given identical treatment. The effect of mechanical treatment has been investigated in terms of potential measurements in addition to being exprbssed as the apparent area change as presented above. Differendes of potential in distilled water between platinrum and platinum rubbed with sandpaper, filter paper or linen were found by Gaugain(l40) In addition to possible mechanical strain resulting from cleaning, the construction of the electrode offers another instance where the metal was mechanically handled, Any resulting strain could set up small voltaic cells since the strained portion was less noble

-208. than the unstrained part. Also small crystals are less noble than large ones. Glasstone(50) stated that the effect on the single electrode potential of bending or straining the electrode could be measured on a student potentiometer except in the case of the softest of metals. Gold and platinum have a Moh's hardness of 2o5-3 and 4-4~5 respectively. -Since the hardness scale ranges from one to ten, it would appear that gold would be less affected than platinum.

APPENDIX C. GAS-METAL.E;ECTRODE EFFECT The observations of Fredenhagen(l41) concerning the effect of gases on a potential reading indicated this effect could be appreciable. He found that the platinum potential in an oxidizing solution was sometimes reached quickly and sometimes slowly and concluded that.the velocity of reaction of ions in the oxidizing solution with gases absorbed on platinum was quite variable. It was further stated, that in oxidation elements platinum always became charged with gases; in.some cases the potential was entirely due to the gas charge and was in no way conditioned by the giving up of electrons from ions of the solution to the electrode. Chang(i) also concluded that the evaluation of the platinum potential (66) was complicated by the presence of dissolved oxygen..According to Ward a poorly poised media exposed to air during bacterial growth created an oxygen-platinum electrode effect. An extensive investigation by French and Kahlenberg(17) resulted.in the conclusion that metals formed gas-metal electrodes with gases. The resulting electrode potentials were specific for each metal and gas, and were related to the absorbing power of the metal for the gas and the affinity of the metal for each gas in a mixture. The total potential was due too (1) absorbO4 gas and (2) a condensed film on the surface. The potential.depended ndt only on the proportion of gas, but the?elative affinity for the metal, -209

-210. They tried. many metal electrodes including platinum and goldo The electrodes were freshly sanded and put in an air-saturated normal solution of potassium chloride. The zero time was recorded when the potential leveled off and gas was then passed through the solutiono The electrode was a flat plate half-immersed in the solution. The results were not reproducible but checked in a semi-quantitative mannero Carbon monoxide and methane altered the potential of platinum and paIladium materiallly while on other metals the effect was slight. With hydrogen, gold remained essentially constant but fell off after the gas was shut off. Platinum quickly dropped approximately 500 my before leveling off and contiizu6edl to drop and then rose slightly after the gas was shut off. All metals except those of the platinum group were rapidly changed from hydrogen to oxygen potentials while it took considerable time to get the hydrogen potential. Platinum was opposite with the hydrogen potential being rapidly attained while the oxygen changed the potential: slowly. Platinum was found to be unsatisfactory unless stablizied with a gas such as hydrogen in which the metal assumed a constant potential.q It exhibited a specific behavior with gases, quite different from most other metalso Gold was the only metal which could not be-polarized by hydrogen after being exposed to oxygen for some time. To prevent condensation of the gas film, either the electrodes were coated with a material like collodion or else the solution was agitated, With the gas flow on, the electrodes read the same whather coated or not. But if not coated, the tptentials fell after the gas was shut off. This condensation

-211l fiJm on the surface was given as the source of variable electrode iehavior upon agitating:or accidentally moving the electrode or electrode vessel, The effect of a rapid change was particularly noticeable with hydrogen and platinum or nickel and slightly less noticeable with gold. It was thought tiat the potential produced was du- to the dissolved gas, not to the'gas i':- iri'the:gas phase. However, greater changes of potential which occurred when the electrode was completely immersed opposed the authors theoretical conclusion that a partly immersed electrode gave a greater effect, The effect of hydrogen could be further considered in connection with the hydrogen -electrode. This type of electrode has been extensively studied and yielded a linear relationship between the potential and pH by catalyzing the reaction: IF = 2HE+ + 2eo Beans and Hammett(l43) concluded that bare platinum was a poor catalyst and that platinized platinum was an excellent one. Michaelis(l6) found that blank platinum acted somewhat differently than platinized platEnumb The latter gave a potential dependent upon the pressure of the hydrogen atmosphere due to its catalytic effect on the hydrogen equilibrium~ The blank platinum acted qualitatively like the hydrogen electrode but it was a slow reaction and equilibrium was never reached. The potential was almost always uncertain, and varied as to the individuality of the electrode? Gold-plated platinum usually gave a potential further from the true value than pure platinum. Solid gold reacted but slightly.

The gas effect has also been investigated for electrodes in buffer solutions. Gold yielded a well-defined potential which returned to its original value a short time after polarization accordi'g to Michaelis(l6) The potential of the gold electrode was scarcely altered whether air, hydrogen, or nitrogen was introduced. There was always a -small difference between air in the positive direction and hydrogen in a negative direction. Some sensitive electrodes gave differences up to one volt but gold and sometimes blank platinum gave differences much lesstiafi 0.1 volto The differences with platinum were larger and depended on the surface. In spite of the differences with respect to hydrogen and oxygen, various "indifferent electrodes" behaved alike in well-poised solutions such as quinhydrone where the oxygen content did not affect the readings. Lepper and Martin(l5) investigated special gold-plated platinum and iridium electrodes in weakly poised and unpoised bufferosolutions under the effect of gas. Iridium is a member of the platinum family. The gold plate electrode was made by heavily gilding a platinum plate electrolytically, drying, heating several times in a blowpipe until a deep gold color resulted and then finally electrolytically depositing gold. These were found much less sensitive to hydrogen and oxygen than merely gilded platinum. In a weakly poised solution the gold plate changed little with decreasing concentrations of poising agent but iridium at the lower concentrations approached the hydrogen electrode potential under the influence of hydrogen. In an u4poised buffer the gold plate showed little

change with hydrogen. However, the iridium showed a marked drop in potential. Oxygen, bubbled immediately after the hydrogen, caused a slight rise in potential with gold, but the iridium rose much more in relation to its low potential with hydrogen. The iridium rose to a higher positive potential than the gold did. The gold appeared to be more sensitive to oxygen than hydrogen in this case. In order to create conditions more closely associated with biological growth than salt or bdffer solutions, Boyd and Reed(l4) studied the effect of hydrogen, nitrogen, carbop dioxide and air on platinum, gold and mercury in an anaerobically aged beefg extract broth at 0~Co The platinum and gold were used in the form of polished foilo Air caused a gradual increase in the potential of 100 to 200 myo The potential rose -when the gas was shut off, Due to the slow response, the authors felt this indicated the inert nature of the electrode to air. Nitrogen had no effect on platinum and a very slight effect on gold and mercury. Hydrogen had no.effect on gold and mercury beyond that predicted by the standard deviation and caused by stirring. With platinum the change was much greater producing a. drop tb approximately -250 my. The potential dropped further after the gas was shut off. Since this fall was not exhibited with the other gases it was concluded that the hydrogen effect was not due to stirring. These investigators also investigated the effect of bubbling hydrogen-through broth for five hours. They compared their results with the work of French.and Kahlenbe*t who treated a potassium chloride solution in the same fashion. In both media the drop in the platinum potential of 500 my was practidally the same, They felt this general agreement of the platinum increased the significance of the gas-metal electrode.

BIBLIOGRAPHY 1. Babbitt, H. ]. Sewerage and Sewage Treatment, 7th Edition, New York: John Wiley arid Sonr7, Inc., 1953. 2, Fair, G. M. and Geyer, Jo C. Water Supply and Waste-Water Disposal. New York: John Wiley and Sons, Inc., 1954. 30 Imbdff, K. and Fair, G. M. Sewage Treatment. New York: John Wiley and Sons, Inc., 1940. 4. Keefer, C. E. Sewage Treatment Works. New York: McGraw-Hill Book Co., 1940. 5. Pearse, L. (ed.)o Modern Sewage Disposal. New York: Federation of Sewage Works Assn., 1938. 6. Buswell, A. M. and Boruff, C. S. "The Relationship between the Chemical Composttion of Organic Matter and Quality and Quantity of Gas Produced during Sludge Digestion," Sewage Works Jour., 4, (1932), 454, 7. Haseltine, T. R. "The Cause and EffOct of Variations in Composition of Gas," Sewage Works Juurnal, 5, (1933), 482. 8. Symons, Go E. and Buswell, A. M. "The Methane Fermentation of Carbohydrates," Jour. Am. Chem. Soc. 55, (1933), 2028, 9. Garber, W. F, "Plant-Scale Studies of Thermophilic Digestion at Los Angeles," Sewage and Ind. Wastes, 26, (1954), 1202. 10. Hatfield, W, D,, Symons, G. E, and Mills, Ro Ro "Gases from Sewage Sludge Digestion," Ind. Eng. Chem., 20, (1928), 174, 11. Heukelekian, H. "Thermophilic Digestion of Daily Charges of Fresh Solids-aand Activated Sludge," Sewage Works Jour., 3, (1931), 3. 12, Rawn, A. M., Banta, A. P. and Pomeroy, R. "Multiple-Stage Sewage Sludge Digestion," Proc. Am. Soc. Civil Engo, 6, (1937), 1673, 13. Morgan, P. F, "Efficient Utilization of Digester Capacity," Sewage and Ind. Wastes, 26, (1954), 1340. 14, Boyd, E. M, and Reed, G. Bo "Gas-Metal Electrode Potentials in Sterile Culture Media for Bacteria," Can. Jour. Res,, 4, (1931), 54, 15. Lepper, E. H and Martin, C,., "On the Behavior of Indifferent Electrodes when Used for the Determination of Oxidation-Reduction in the Presence of Hydrogen," Biochem. Jourg 25, (1931), 45, B 2 L4~I

2.19 16o Michael.s, Lo Oxidation-Reduction Potentials. Translated by L, B, Flexnero Phil. ade1ph ia- J. B. L-ppincott C pany, 1930o 17. French.- SO J. and Kahlenberg, L, "Gas-Metal. Electrodes,," Trans. Am. Electrocherm. Soc 54o 9 (1928), 163o 18. Buswell, A. Mo and Neave, S. L. 'Laboratcry Studies of Sludge DigestiOno Illinois State WaterD Supply Bull No. 30, 19300 19o Sawyer, Co N. and Roy, Ho K. "A Laboratory Evaluation of Higa-Rate Sludge D.igeston," Sewage and Ind, Wastes, b (1955), 1356. 20. Steel,. E, W, Water Suppy ad Sewerage~ 3rd Edition, New YorkMcGraw-Hill Book Co., Inc., 1953. 21E Hawkes, H. A. and Jenkinss S, H, "BiOological PriniEples in Sewage Purification," Jour, Proc. Inst. Sewage Purif,, Part 3, (1951), 300, 22. Rudolfs, W o Principles of Sewage Treatmente,._1l 212, National Lime Association, Washington D. C.,, 1955. 23, Fil.we[iler, W H, '!'se of Sewage Gas as City Gas," Sewage Works Jour,, 2, (1930)9 424. 24. PhelpsY E. Bo Stream Sanitation, New York, John Wiley and Sons, Inc., 1944, 25, Buswell, A, M, et-, al. aerobic Fermentations, Illinois State Water Supply B ulo No. 32, (1936, 43 — 26, Morgan., P, F, "Studies of Ac.elerated Digestion of Sewage Sludge, " Sewage and Ind, Wastes, 26, (1954), 462~ 270 Torpey, W. No "High Rate Digestion of Concentrated Primary and Activated Sludge," Sewage and Ind, Wastes, 26 (1954), 479. 28, Barker, Ho A, "On the Bioochemistry of the Methane Fermentation, " Arch, f. o Mik obi.ol, (1936), 404_ 29, Barker, Ho A. "Studies upon the Methane=Producing Bacteria," Arch~ f. Mikrobiol,, (1936), 420. 300 Stadtman, T. C. and Barker, H. Ao "Tracer Experiments of Fat; y,yAcid O*idation by Methane Bacteria," Jour, Bact., 61, (1951), 67~ 31. Stadtman, T,. C and Barker, H. Ao "The Origin of Methane in the Acetate and Methanol Ferment attlons by Methanosarcina," Jour, Bact., 61, (1951), 81. 320 He kelekiJan, H. and Hein.Lonrun, B '"Studies on MethanesProducing Bateriq..T Development' of a Method for Enuameration," Sewage Works JoO >,, ELJ, <z..39), 1 426,

33. Heukelekian, H. and Heinemann, B. "Studies on Methane-Producing Baat:eriU. II. Enumeration in Digesting Sewage Solids," Sewage Works Jour. 11, (1936), 436. 34. Heukelekian, H. and Heinemann, B, "Studies on Methane-Producing Bacterisa, III. Digestion of Sewage Solids by the Addition of Enriched Cultures of Methane-Producing Organisms," Sewage Works Jour., 11, (1939), 571. 35. Heukelekian, H. and Heinemann, B. "Studies on Methane-Producing Baoteria, IV. Growth of Methane-Producing Organisms in Supernatant Sludge Liquors," Sewage Works Jourw, 11, (1939), 965. 36. Buswell, A. M. "Microbiology and Theory of Anaerobic Digestion," Sewage ~Works Jour,, 19, (1947), 280 37. Buswell, A. M. "Fermentations in Waste Treatment," Industrial Fermentations, Edited by Underkofler, L. Ai(and Hickey, R. J, Vol. II, 518,- Nliew:-iYork': Cherhical Publishing Co., 1954, 38. Buswell, A. M. and Mueller, H, F. "Mechanism of Methane Formation," Ind._Eng. Chemo, 44, (1952), 550. 39. Buswell, A. M., Pagano, J. F, and Sollo, F. W. "Effect of Sodium Sulfate and Chloride on Methane Fermentation," Ind..Eng, Chem., 41, (1949), 596. 40. Buswell, A, M. and Sollo, F. WO "Mechanism of Methane Fermentation," Jour. Am, Chem. Soc., 70, (1948), 1778. 41, Haukelekian, Ho "Volatile Acids in Digesting Sewage Sludge, " Ind. Eng., Chem,, 20, (1928), 752. 42. Heukelekian, H. and Kaplovsky, A, J. "Improved Method of V0latile Acid Recovery from Sewage Sludges," Sewage Works Jour., 21, (1949), 974. 43. Kaplovsky, A. J. "Volatile Acid Production during Digestion of Seeded, Unseeded, and Limed Fresh Solids," Sewage and Ind, Wastes, 23, (1951), 713. 44, Kaplovsky, Ao J. "Volatile Acid Production during Digestion of Seye'al Industrial Plant Wastes," Sewage and Ind Wastes, 24, (1952), 1940 45. Pearson, E, L, and Buswell, A. M, "Sludge Ripeness Studies, "nd. Eng. Chem,, Anal. Ed,, 3, (1931), 359. 46, Rudolfs, W. and Heinemann, B. "Growth Promoting Substances in Sewage Sludge," Sewage Works Jour., 11, (1939), 587. 47. Schlenz, H. Eo "Important Considerations in Sludge Digestion. I. Practical Aspects," Sewage Works Jour., 1_, (1947), 19o

48, Hewitt~ L. F., Oxidation:Reduction Potentials in Bacteriology and Biochemistry, 6th Edition. EdiEnburgh:.E. S. Livingstone, Ltd,, 2950. 49, Clark, W. M. et al. Studies on Oxidation-Reduction I-X, U.S. Public Health Service, Hygienic Laboratory Bull. No. 151, U.S. Government Printing Office, Washington, 1928o 50. Glasstone, SO An Introduction to Electrochemistry. New York: Do Van Nostrand Co., Inc., 1942, 51. Conant, }J B, and Lutz, R. E. "Irreverqible Reduction of Organic Compounds," Jouro Amo Chem. Soc,, 46, (1924), 1254. 52. Gershinowitz, H, "The Relation between the Rate and Oxidation-Reduction Potential," Jbur, Chem. Physics, 4, (1936), 3630 530 LuValle, jo. E and Goddard, P. R, "The Mechanism of Enzymatic Oxidation and Reduction," Quart0 Rev, Biol, 23, (1948), 1970 54, Longsworth, L. G, and Macinnes, D. A. "Appexent Oxidation-Reduction Potential, Acid, and Population Studies of.,L. Acido hilus under Anaerobic Conditions, " Jour. Bact,, 32, (193-6), 57 55, Barron, E. S. G. "The Rate of Auto-Oxidation of Oxidation-Reduction Systems and Its Relation to Their Frde Energy," Jour. '-Biolo Chem,, 97, (1932), 287. 56, Clark, W, M.' "The Potential Energies of Oxidation-Reduction Systems' and Their Biochemical Significance," Medicine, 13, (1934), 207 -57. Lepper, E, H. and Martin, C. J. "The Oxidation-Reduction Potential of Cooked Mest Media," Brit0 JOur. Exp0 Path0, 11, (1930), 137. 58., Flexner, L. B, and Barron, E..S. G. "Oxidation-Reduction Potentials at Carbon and Tungsten Electrodes," JOuro Am, Chem Soc., 52, (1930) 27730 590 Tuttle, C,. Do and Huddleson, I. F. "Determination of Oxidation-Reduction Potentials of Sterile Culture Media with Graphite Elec#tiodes," Jour. Inf, Dis.,, (1934), 273. 60o Hakampnri, S. "Potential of Hydrogen Peroxide, " Tech, Rep, Tohoku, 9 (1931),, 1o06a 61, Dixon, Mo and Quastel, J. H. "A New Type of Oxidation-Ren.uction System0 I. Cysteine and Glutathione," Jour, Chem. Soco, 123, (1923), 2943, 62. Knight, B, C, "Oxidation-Reduction Potentials of Sterile Meat Broth," Biochemn Jour,;, 24, (1930), 1066. f~~~~~~~~~~ ~,., *ol.

63. Borchardt, J. A. 'A Study of the Chlorine Breakpoint by Means of Electrode Potentials. Ph..D. Thesis, University of Wisconsin, 1948. 64. Remington, V. H. and Trimble, H. M. "The Oxidation Potentials of Some Hypochlorite Solutions," Jour. Phys. Chem., 33, (1929), 424. 65. Clark, W. M. The Determination of Hydrogen Ions, 3rd Edition. Baltimore: The Williams and Wilkins Co., 1928] 66, War%, W.E. "Apparent OxidatiOn-Reduction Potentials of Bright Platinum Ele drodes in Synthetic Medium Cultures of Bacteria," Jour. Bact., j (, (1938), 337. 67. Toben, L, C. "Convenient Measurement of Oxidation-Reduction Potentials in Fermentation Studies," Chemist Analyst, 35, (1946), 63. 68. Barron, E. G. So "M1phanisms of Carbohydrate Metabolism," Advances in Enzymology, 3, (1943), 182. 6+o Porter, J. R.0 Bacterial Chemistry and Physiology, New York: John Wiley and sdns, Inc., 1948. 70. Barron, E. S. G., F-lexner, L. B. and Michaelis, L. "Mechanism of the Cysteine Potential at the Mercury Electrode," Jour. Biol. Chem., 81, (1929),- 743. 71. Potter, Mo C. "Electrical Effects Accompanying the Decomposition of Organic Compounds," Proco Roy. Soc. 13, 1i911), 84, 260. 72. Gillespie, L. JO "Reduction Potentials of Bacterial Cultures and WatWLo6gge;jd, Soils, Soil Sciencs 9, (1920), 199, 73. Boyd, E. M. and Reed, G. B. "Oxidation-Reduction Potentials in Cultures of E. Ctoli, " Can, Jour. Res., 4, (1931), 6050 74. Elena,/ B. et al. "Uber die Berziehurgen zwischen den stoffwechselvorgangen der microoxr'anismen und dem oxydoreduktionspotential im Medium," Biochem. Ztschr., 270, (1934), 317. 750 Coulter, C. B. "Reduction Potentials of Sterile Culture Bouillon,: Jour. Gen, Physbol., 12, (1928), 139. 76. Aubel, Eo, Rosenberg, A. G. and Grunberg, M. "Limiting OxidationReduction Potentials for Anaerobes," Helv. Chem. Actv,, 29, (1946), 1267. 77. Knight, B. Co "A Method of Poising the Oxidatlion-Reduction Potential of Bacteriological Culture Media," Biochem, Jour., 24, (1930), 107V). 78. Reed, Go B. and Orr, J. B. "Cultivation of Anaerobes and OxidationReduction Potentials," Jour. Bact,, 45, (14~), 309.

r219 -79, Kendall, E, C, and Loewen, D. F "The Oxidation-Reduction Potential of Cysteine and Cystine, " Biochemo Jouro, 22, (1928), 6690 80, Michaelis, L, and Flexner, L. B, "The Reduction of Cysteine~ Its Measurement and Significance, " Jour Biol Chem,, 79, (1928&), 689~ 81, Longsworth, L. G, and MacIrnes, D. A, "Bacterial Growth at Constant pH," Jour. Bact., 31, (1936), 287~ Lepper, Eo H, and Martin, Co J. "The Oxidation-Reduction Potential of Cooked'Meat Media Following:IiocuIlation1. of Bacteria, 1 Brit, Jouro Exp, Path,,' 11, (1930), 140, 830 Borsook, H, and Schott, H. FP "The Role of the Enzyme in the SuccinateEnzyme-Fumarate Equilibrium," Jour, Biolo Chem., 92? (1931), 535o 84, Morgan, J. Lo R. and Lammert, 0o Mo "The Quinhydrone Electrode," Jour, Am. Chem.. Soc, 53, i1931)- 21,54. 85, Corran, J. W. and Lewis, W. Co "The Hydrogen Ion Concentration of the Whole Blood of Normal Males and Cancer Patients Measured by Meags of the Quinhydrone Electrode, " Biochemo Jour,, 21, (1924), 1358o 86. Grossman, F, "The Use of the Hydroquinhydrone Electrode for pH Determination in the Fluids of the Organism," Biochem, Jour,, 21, (1927), 2670 87, Mislowitzer, E, "Determination of the pH of Plasma, Serum, and Blood by the Quinhydrone Method," Klino Wochschr,, 5, (1926), 1863. 88, Hood, J, Wo "Measurement and Control of Sewage Treatment Process.Efficiency by Oxidation-Reduction Pote4]tial," Sewage Works Jour,, 20, (1948), 6400 89, Spitta, Do and Weldert, Fo "Indikatoren fur die Beurteilung biolo isch gereinigter Abwasser," Mitto Prufungsanst, Wasserversorg. Berlin,., (1906), 161, 90, Phelps, E, B, and Winilow, Co. E. A, "On the Use of Methylene Blue in Testing Sewage Effluents," Jour. Inf, Ds,, Supplement No, 3, (1907), 2 91o Dickinson, DI "Changes in Oxidation-Reduction Potential with the Sludge Floc 'J5ruo Soc. Chem, IndL, 5, (1940), 257. 92, Rohlich, G, A,, Sarles, W. B, and Kessler, L. H, "Oxidation-Red4~tion Potentials in Activated Sludge, " A Symposium of Hydrobiology, Un;versity of Wisconsin, (1941), 288. 930 Rohlich, Go A, "Oxidation-Reduction Potential Measurements in Activlated Sludge and Sewage Mixtures," Sewage Works Jour,, 16, (1944), 5400

-220 -94. Rohlich, G. A. "Discussion — 'Measurement and Control of Sewage Treatment Process Efficiency by Oxidation-Reduction Potential,' by Hood, J. W,, Sewage Works Jour., 20, (1948), 640," Sewage Works Jour., (1948), 650. 95. Moore, W. A, and Ruchhoft, C. C. "Oxidation-Reduction Potentiald Developed in Sewage and Sewage-Activated Sludge Mixtures," Sewage Works Jour., 15, (1943), 880. 96. Moore, Wo Ao, Ruchhoft, C. CO and Wattre,. E. "Oxidation-Reduction Potentials Developed by Pure Cultures in Sewage," Sewage Works Jour., 14, (1942), 980. 97. Dawson, P. So So and Jenkins, S. H. "Oxygen Requirement of Activated Sludge Determined by Manometric Methods," Sewage and Ind. Wastes, 21, (1949), 643. 98. Eckenfelder, W. W. and Hood, J.Wo. "The Application of OxidationReduction Potential to Biological Waste-Treatment Process Controls" Proc. of the 6th Ind. Waste Conf., Purdue Univ., (1951), 2210 99. Nussberger, F. G. "Application of Oxidation-Reduction Potentials to the Control of Sewage Treatment Processes," Sewage and Indo Wastes, 25, (1953), 1003. 100..Keefer, Co Eo and Meisel, J "'Activated Sludge Studies. I, Effect of Sludge Age on Oxidation Capacity," Sewage and Indo-Wastes, 22, (1950), 1117. 101. Keefer, C. E. and Moisel, J. "Activated Sludge Studies. III. Effect of pH on the Process," Sewage and Indo Wastes, (1951), 982. 102. Keefer, Co.Eo and Meisel, Jo "Activated Sludge Studies. IV. Sludge Age and Itis Effects on Activated Sludge Process," Sewage and Ind..Wastes, 25, (1953), 898. 103. Official Methods of Analysis of the Association of Official Agricultural Chemists, 7th Edition.o Assno of Off. Agro Chemists, Washingtion, D. C., 1950. 104. Rudolfs, W. and Trubnick, E. H. "Biological Treatment i~i Laboratory Pilot Units, t Sewage Works Jour., 21, (1949), 100. 105. Torpey, W No."Loacing to Failure of a Pilot High-Rate Digester," Sewage and Indo Wastes, 27, (1955), 121o 106. Heukelekian, H. "Discussion —'Loading to Failure of a Pilot High-Rate Digester,' by Torpey, Wo N., Sewage and Indo.Wastes, 27 (1955), 121," Sewage and Ind. Wastes, 27, (1955), 142o

~221 -107. Blodg.ett, J. H. "Discussion — 'Studies of Accelerated Digestion of Sewage Sluge, I by Morgan, P. F., Sewage and Ind. Wastes, 26, (1954), 462," Sewage and Ind. Wastes, 26, (1954), 476. 108, Ajl, S. J. and Werkman, C. H. "Replacement of Carbon DiOxide in Heterotrophic Metabolism," Arch, Biochem., _, (1948), 483, 109. Jordan, E, 0. and Burrows, W. Textbook of Bacteriology, 14th Edition. Philadelphia: W. B. Sanders Co., 1946. 110. Salle, A. J. Fundamental Principles of Bacter yy, 4th Edition. New Yoak: McGraw-Hill Book Co.., Inc., 1954. 111. Walker, H, H. "Carbon Dioxide as a Factor Affecting Lag in Bacterial Growth," Science, 76, (1932), 602. 112, Vennesland, B. and Hanke, M. E "~he Oxidation-Reduction Potential Requirements of a Non-Spore- orming, Qbligate Anaerobe, " Jour, Bact., 39, (1940), 139. 113. White, W. PO "Leakage Prevention by Shielding, Especially in Potentiometer Systems, " Jour. Am. Chem, Soc., 36, (1914), 2011. 114. 'Standard Methods for the Examination of Water, Sewage, and Ind-ugtr'ial Wastes, 10th Edition. Amer, Pub. HegaltIhjAdsn., New York, 1955o 115. MacIcnies, D:, A., Prin ip]Bes of Electrochemistry. New York~ Reinhold $ 1_, M. r..9:. r.......... ~., -,.. Pub. Co., 1939. 116, Lewis, G. N. and Sargent, L. W. "The Potential of the Ferro-Ferricyanide Electrode," Jour. Am. Chem, Soc., 31, (1909), 355. 117. Shaefer, P. A, and Williams, R. D. "Sugar Detection by the Ferricyanide Electrode, " Jour. Biolo Chem,, 111, (1935), 707. 118. Toole, F. J. and Johnson, F. M. G. 'The Solubility of Oxygen in Gold and in Certain Silver-Gold Alloys," Jour. Phys. Chem., 37, (1933), 331. 119, Mortimer, C, H. "Exchange of Dissolved Substances between Mud and Water in Lakes," Jour. Ecaol,, 30, (1942), 147. 120. Heukelekian, H. "Further Studies on Thermophylic Digestion of Sludge, " Sewage Works Joure, 2, (1930), 2. 121. Mellor, J. W. A Comprehensive Treatise on Inorganic and Theoretical Chemistry, Volume III. Cop; r, Silver, Gold, Calcium, Strontium and Barium. London' Longmans ahd.op,, 1923. 122, Mellor, Jo W. A Comprehensive Treatise on Inorganic' and Theoretical Chemistry, Jbli'de XVI, Platinum and General lndex. London: Longmans and Co., 1937.

-222 -123. Pierce, W. C, and Haenisch,. E. L. Quantitative Analysis, 2nd Edition. New York: John Wiley and Sons, Inc., 19400. 124. Bowden, F. P. "The Amount of Hydrogen and Oxygen Present on the Surface of a Metallic Electrode," Proco Roy. Soc. (London) A, 125, (1929), 446. 125. Rideal, E. K. and Wansborough-Jones, 0. H. "An Investigation of the Combustion of Platinum," Proc. Roy, Soc. (London) A, 123, (1929), 202. 126, McBain, J. W. Sorption of Gases and Vapours by Solids, London: George Routledge and Sons, Ltdo, 1932. 127, Langmuir, I. "The Adsorption of Gases on Plane Surfaces of Glass, Mica, and Platinum," Jour. Amo Chem, Soc., 40, (1918), 13610 128o Finch, G. I. and Stimson, J. C. "The Electrical Condition of Hot Surfaces during the Adsorption of Gases. I. Gold and Silver Surfaces at Temperatures up to 850 C," Proc. Roy. Soco (London) A, 116, (1927), 3790 129. Finch, G. I. and Stimson, J. C. "The Electrical Condition of Hot Surfaces during the Adsorption of Gases, IIIo A Platinum Surface at Temperatures- up to 850~C," Proc. Roy Soc. (London) A, 124, (1929), 356, 130. Benton, A. F. and Elgin, J, C, "The Catalytic Synthesis of Water Vapor in Contact with Metallic Gold," Jour, Am, Chem. Soc, 49, (1927), 2426. 131, Goard, A. K. and Rideal, E. K. "On The Measurement of Certain Inaccessible Potentials with a Controlled Oxygen Electrode," Trans. Farad.. Soc., 19, (1924), 740. 1320 Glasstone, S. and Hickling, A. "The Electrolytic Oxidation of Sodium Thiosulfate at a PiBtinum Anode," Jouro Chemo Soc. (London), (1934), 23450 133. Berliner, A, "About the Pulverization of Glowing Me als, Ann, der Physik, 33, (1888), 289, 134, Stromberg, R. "Adsorption Measurements with an Improved Microbalance," Kgl. Svenska Vetenskapsakad Handl. (3), 6, No. 2, 1.(1928). 135. Mendenhall, C. E, "Recent Developments in Photoelectricity," Science, 73, (1931), 107. 136. Frazer, J. C. W., Patrick, W. A.. and Smith, H, E. "The Thickness of Adsorbed Vapor Films," Jour, Physo Chem., 31, (1927), 897. 137o Rideal, E. Ko An Introduction to Surface Chemistry_ London: Cambridge University Press, 1930.

,2235 138. Bowden, F, P. and O Connor, Eo A. "The Change:in the Area and Catalytic Activity of Metallic Surfaces on Passing from the Solid to the Liquid State," Proc. R oy. Soc. (LondOn) A, 128, (1930), 317* 139, Langmuir, I. "Chemical Reactions on Surfaces, " Trans. Farad. Soc., 17, (1922), 607. 140o Gaugain, J. M, "Electromotive Force, " Compt. Rend., 7k, (1872), 1332. 141. Fredenhagen, C. "On the Theory of Oxidation and Reduction," ZeIt. Anorg. Chem,, 29, (1902), 407~ 142. Chang, S L. "Application of Oxidation-Reduction Potential Measurements in Determining the Concentration of Germicidally Active Chlorine in Water," Jour. New Eng. Water Works Assn,, 59, (1945), 79 -143. Beans, H. T. and Hammett, L. P. "Experimental Studies on the Hydrogen Electrode," Jour. Am. Chem. Soc., 47 (1925), 12150 Electrode --

UNIVERSITY OF MICHIGAN 3 9015 03483 6497