H H I III I i“ l ~ I Av____d, 7”“ ,,# -77 g , ’_K #. I if” '-7 7 VA): _— I) 1 W ‘ I II W I II I I», I I I —1_._. IbN (nub DEVELOPMENT OF A METHOD FOR DETERMINING THE CHEMICAL OXYGEN DEMAND OF PICKLE MANUFACTURING WASTES Thesis Ior the Degree OI M. S. MICHIGAN STATE UNIVERSITY Albert Leo Saari 1960 DEVELOPMENT OF A .METHOD FOR DETERMINING THE CHEMICAL OXYGEN DEMAND OF PICKLE MANUFACTURING WASTES B y Al bcrt Loo SEtLtl‘l AN ABSTRACT Submittcd to thc Collcgc of Agriculture Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirenlcnts for the degree of MASTER OF SCIENCE Dcpartmcnt of Food Science 1960 A p p I: O V. (3 Cl I"\) '\.. A . A‘S\3CF/J‘\ "_‘i)/ / / ‘~~ U / \ -.L ’l I \. Albert Saari ABSTRACT At the present time, a great deal of attention is being focused on the problem of lake and stream pollution by oxidizable organic. matter. The standard procedure, which requires five days, for measuring the amount of pollution is the dilution biological oxygen demand (BOD) method. An attempt was made to modify a dichromate chemical oxygen demand (COD) procedure so that it could be used to measure in 50 nninutes the concentration of oxidizable organic matter in pickle manufacturing effluents. A reflux method employing 10 ml of aqueous dichrornate, 5 ml sample and 15 ml sulfuric acid was found most effective. When the COD procedure was applied to pure. samples of acetic acid, between 50 and 75 percent of oxidation was attained whereas 90 to 95 percent of theoretical oxidation was attained with pure samples of lactic acid or carbohydrate. When all three of these components were present in mixture, 90 to 95 percent of the non-acetate components were oxidized and apparently none. or little. of the acetate was oxidized. Oxidation of chloride was prevented by adding an excess of silver sulfate to precipitate the chloride in the sample. This cor— rection step proved adequate for a control type of determination but the silver chloride should be removed before digestion for most reliable results. The BOD procedure was applied to synthetically prepared mix— tures which were similar in composition to pickle plant effluents. The amounts of the reSpective components in the mixture were ii ’--‘—“"“I Albert Saari determined by analysis before and after incubation. The following COInponents were present in the mixtures: sucrose, starch, lactic acid, and acetic acid. The. levels of oxidation attained in 5 days were 80, 47, 98 and 97 percent respectively. The BOD and COD of both synthetic "effluents" and actual plant effluents were compared and the. agreement between COD and BOD values was poor. In all cases the COD values were higher. For the synthetic sample, the BOD values were approximately 80 to 90 percent of the COD value, and for actual plant effluents, the BOD values were only approximately 50 to 65 percent of the COD values. iii DEVELOPMENT OF A METHOD FOR DETERMINING THE CHEMICAL OXYGEN DEMAND OF PICKLE MANUFACTURING WASTES By Albert Loo Saari A THESIS Submitted to the College of Agriculture Michigan State University of Agriculture and Applicd Science in partial fulfillmvnt of the requircmcnts for the degree of MASTER OF SCIENCE Department of Food Science 1960 ACKNOWLEDGMENT The author wishes to express his sincere appreciation to Dr. Irving J. Pflug and Dr. Andrew Tininick for their suggestions, criticisms, guidance, indulgence and assistance during the investigation and preparation of this manuscript. Grateful acknowledgment is extended to Dr. Richard C. Nicholas for his suggestions, help and constructive criticism of the thesis. The author wishes to thank Dr. Ralph N. Costilow. Dr. Karl L. Schulze and all of the others who helped in any way to make this investigation successful. A special acknowledgment is extended to Mr. Roy Gielow of Aunt Jane's Foods Inc. for furnishing effluent for the determinations and for providing an opportunity to carry out work at the Croswell pickle plant. ....................... aaaaaaaaaaaaaaaaaaaaaaaaaa TA BLE OF CONTENTS INTRODUCTION . REVIEW OF LITERATURE . EXPERIMENTAL, RESULTS AND DISCUSSION . Evaluation of the Oxidizers. . Corrections for Chloride Oxidations. . . . . . BOD's of lVIixtures of Lactic Acid, Acetic Acid and Carbohydrates. . . . . . . . . . BOD and COD of Pickle Plant Effluents CONCLUSIONS . SUGGESTIONS FOR FURTHER STUDY SUMIVIARY . APPENDIX . LITERATURE CITED vi Page 20 29 33 36 37 4O 43 INTRODUCTION In the. orderly balance of nature, there are mechanisms to utilize and degrade organic materials. The bulk of these degrad- ation processes are carried on by microorganisms that oxidize these materials to carbon dioxide. and water. When organic materials are added to bodies of water. there is danger that the microbiological oxidations will deplete the oxygen in the water causing the death of higher animals that require'oxygen. The maximum solution of oxygen from the air in water is only about 9 mg/l or 9 ppm, (parts per million), which is such that one pound of sucrose will completely use the oxygen from 2000 cubic feet of water. Water conservation and stream polution control agencies have established minimum oxygen levels for public bodies of water. The Michigan State Health Department's Ininimuni oxygen level for fish survival is 4 ppm, therefore. the. maximum amount of sucrose that could be added to a lake or stream is 1 pound of sucrose per 4000 cubic feet or 30, 000 gallons of water. Food products contain 10 to 30% soluble carbohydrates, proteins and assorted simple organic materials. Operations such as grinding, slicing, peeling and washing remove up to 5‘36 of these soluble con- stituents. Since a pound of soluble solids requires about a pound of oxygen for complete oxidation, it is readily apparent that a food plant handling several thousand tons of food products a day has a major problem with the disposal of waste water. The majority of this sewage is in a highly diluted state and because of the quantity, it must be disposed of into an available body of water or a municipal sewerage system. The introduction of the waste from a food plant into a large sewerage system can place a heavy load on the system; however. if to the waste from a large food plant is introduced into a small system, the high oxidative load can cause the sewage plant to revert to anaerobic processes. Food plants with quantities of soluble wastes can either develop their own sewage disposal facilities or separate the more concentrated waste from the more dilute waste. The dilute waste can be disposed of into a body of water or sent. to the municipal sewage. treatment plant. The more concentrated wastes can be disposed of by other means; for example. ponding. To separate high concentration effluent from low effectively, it is necessary to have a rapid method of analysis. The standard BOD method is useless for quick measurement because of the 5—day incubation period. Determining soluble solids with a refractonneter has been proposed. but this method is inaccurate at the low concentrations usually encountered in pickle cannery wastes. An acid titration has been suggested for use in pickle canneries where the effluent may contain appreciable quantities of acetic or lactic acids. This method too is inaccurate because the acid level is usually quite low and the relationship between total soluble solids and total acid is indeterminate. This study was undertaken to evaluate biological oxygen demand (BOD) and chemical oxygen demand (COD) procedures for evaluating food, plant wastes. and to find a COD method which might be used for monitoring and control so that the concentration of pickle cannery wastes could be determined rapidly with ease. and accuracy. REVIEW OF LITERATURE The measurement of BOD has been studied extensively. The Standard Methods for the Examination of Water, Sewage and Industrial Wastes (1955) usually referred to as "Standard Methods, " gives the following definition: "The biochemical oxygen demand, BOD, of sewage, sewage effluents. polluted waters, or industrial wastes is the quantity of dissolved oxygen in mg/liter required during the stabilization of decomposable organic matter by aerobic biochemical action. " The generally accepted method for measuring BOD described in ”Standard Methods" is a dilution method designed to simulate the. oxidative conditions in a typical body of water. The following procedure is used in the standard BOD determination: a dilution water is prepared by minerali‘zing and aerating distilled water. A suitable amount of sample is placed in a specially designed BOD bottle which is then filled with the dilution water. After 5 days of incubation at 200C, the residual oxygen content is determined by the Winkler (1888) method. A blank on the dilution water is run simul- taneously and the difference between the blank and the sample is taken as the oxygen consumed or the oxygen demand. Babbitt (1957) indicates that BOD is a unimolecular reaction; that is dL/dt : —kL, where I: is the remaining concentration of oxidizable organic matter, and _t_ is the time. Busch (1958) and Gotaas (1948) differ from this view. They hold that the utilization of soluble substances is essentially complete in 24 hours and that a second stage of degradation begins in which protozoans and other ' microbes consume the bacteria of the first stage as well as some of the remaining organic material. This view obviously cannot be interpreted as a simple unimolecular relationship. Any factor that influences the growth of microorganisms will influence the BOD measurement. Lea and Nichols (19.36) illustrated the need for a source of nitrogen, phosphorous, magnesium and calcium in the dilution water. Gellinan and Heukelekian (1951) dis- cussed the influence of pH. nutrients. substrate concentration, and seeding as factors affecting BOD. They found that the optimum pH is in the range of 6. 0 to 8.0 depending on the nature of the substrate. "Standard Methods“ requires that the substrate concentration should be such that 25 to 75 percent of the dissolved oxygen is consumed. Gotaas (1948) studied the temperature versus BOD relationship and reported that 200C was the most satisfactory temperature for accurate BOD's. The temperature required in "Standard Methods" is 200C 3: 10. In practice. seeding of the dilution water with organisms is seldom done. Seeding is necessary, however, if the sample has been treated with a bactericidal agent. According to Oberton and Stack (1957), the best results will be obtained with acclimated seed organ- isms. Gellman and Heukelekian (1951) showed the oxidation of up to 2000 ppm phenol by acclimated seed whereas settled sewage seed exhibited a long lag phase and correspondingly lower oxidation as well as a larger variation among samples. Analytically, the standard BOD leaves much to be desired: it requires five days to make the determination; the determination is sensitive to the microbe content and the microbial activity; and it has an accepted accuracy of i5 percent. Manonietric methods for measuring BOD have been suggested as possible substitutes for the dilution method. For example, Sierp (1928) used a gas buret to measure the volume of gaseous oxygen consumed from the atniOSphere above the waste water under standard (J1 BOD conditions. Gellman and Heukelekian (1951) eliminated the CO; evolution error of the Sierp apparatus by inserting a vial of alkali into a flask. Following the development of the sophisticated Warburg nianometric device (see Umbreit, 1957), Caldwell and Langelier (1948) applied it to BOD determination. They showed that it had some promise as a commercial waste analysis method but that further work would be necessary to evaluate its usefulness. Lee and Oswald (1954) compared the dilution and Warburg methods for the measurement of BOD. They concluded that the Warburg method is better suited to research than to industrial applications. It is obvious that the BOD of a waste is the sum of the BOD's of the compounds making up the waste. A number of investigators have been interested in and have measured the BOD of naturally occurring organic compounds often found in waste waters. Heukelekian and Rand (1955) summarized the results of. many workers. The reported figures varied a great deal. Examples of the range of BOD's reported as grams of oxygen consumed per gram of compound are given below: Compound R_a_ng_e Theoretical Acetic acid 0. 34 to 0.88 1.07 Lactic acid 0.65 to 0.64 1.07 Sucrose 0.49 to 0.76 1.12 Casein 0.25 to 1.17 1.2 Gaffney and Heukelekian (1958) found that the lag phase in the growth of the oxidative organisms in water could. extend as long as 5 days. They found that the BOD of acetic acid was as low as 6 percent of the theoretical value when a lag phase of 5 days was observed. A number of chemical oxidation procedures have been proposed for estimating the BOD. They are often called chemical oxygen demand, COD, chemical oxygen consumed, COD, or simply, oxygen consumed, OC. The chemical methods employ vigorous oxidizing conditions for short periods of time with strong oxidizing agents such as dichromate, periodate. permanganate. or ceric salts. Chemical oxidizers oxidize carbohydrates very readily but they are slow in attacking most other commonly occurring natural organic cmnpounds such as fatty acids and proteins. Burtle and Bushwell (1937) compared the oxidation of carbon in organic compounds by persulfate. permanganate, and dichromate. They found by Ineasuring the CO; evolved that dichromate in a strongly acid Inedium was most effective for oxidizing organic compounds. Moore, it 8:}. ,(1951) COInpared the oxidizing power of ceric sulfate, potassium permanganate, periodic acid, and dichromate reflux on a large number of assorted organic materials such as carbohydrates, proteins. fatty acids, hydrocarbons and alcohols. They too concluded that dichromate gave the most complete oxidation of the various organic compounds. Moore, at a_1.. (1949) prOposed a dichromate reflux scheme which appeared as a tentative COD method in the 1955 edition of ”Standard Methods. " They used 25 ml of a standard solution of potassium dichromate plus 25 ml of sample and 50 ml of concentrated. sulfuric acid in a reflux apparatus to effect the complete oxidation of most organic compounds. A silver sulfate catalyst was found to speed the oxidation of compounds barely affected by two hours of reflux. Winneberger. Kerrigan and Whittier (1960) streamlined this method of eliminating the daily water blank and using smaller portions of the reactants. The reflux period was also shortened to 15 minutes. Rharne (1947, 1.949) heated a reaction mixture of sample and Schollenberger's oxidizer (1916). He found good reproducibility and a comparison of the COD with the BOD of settled sewage gave fair correlation. Mercer and Rose (1956) used the same oxidizer but digested the Inixture for 10 minutes at 92 d: 20C. /They found excellent reproducibility on fruit cannery and tomato processing wastes. Their results also showed good correlation of the COD’S with BOD's run on the same wastes. Johnson (1947) described a micro method utilizing chroniic acid as the oxidizer for the determination of total organic matter in the various stages of enzyme isolations. He used 0.4 ml of sample in 1 ml of dichromate in concentrated sulfuric acid solution and digested in a test tube for 20 Ininutes at 1000C. Kieselbach (19541) devised a continuous carbon analyzer using 0.1 M CrO3 in 95 percent sulfuric acid as an oxidizer at 2500C. Kieselbach claimed complete oxidation of all organic materials. Snethlage (19.38) and Aynsley (1950) found negligible decompo- sition of dichromate in sulfuric acid solution at temperatures below 1000c. EXPERIMENTAL, RESULTS AND DISCUSSION This study was divided into four parts: 1) evaluation of various dichromate solutions as oxidizers of acetic acid, 2) correction for the oxidation of chloride when the effluents contain high concen- trations of salt. 5) comparison of BOD and COD on an artificial substrate. and 4) comparison of BOD and COD on actual plant effluent. The experimental procedure and results and discussion are combined and presented in the above order. Quantitative chemical procedures were followed throughout the experimental work. All glassware such as burets. pipets and volu- metric flasks met the American Chemical Society standards for quantitative glassware. Volumes in the macro burets were read to 0. 02 ml. When "micro buret” is specified, the volumes were read to 0.01 ml. The amounts of oxidizer and sample for the micro Schollenberger oxidizer method were determined to the nearest mg using an analytical balance. The accuracy of the other weights given in the results was that necessary to give the reported number of significant figures. Evaluation of the Oxidizers Oxidation of acetic acid by modified Moore method. The modifi— cation by Winneberger. et a1., (1960) of the dichromate reflux method of Moore. et a1. , (1949) was used to measure the percentage of acetic acid oxidized. Approximately 0.2 to 0. 3 g of silver sulfate was introduced into a 250 ml round-bottomed flask with standard taper neck. Five ml of sample was then pipetted into the flask and the contents were thoroughly Inixed. Next. 10 ml of 0. 1000 N aqueous potassium di- chromate solution was added followed by 15 ml of concentrated sulfuric acid. The flask was immediately fixed. onto a water cooled condenser with the standard taper connection and refluxed for 15 minutes. The flask was then removed from the condenser and the contents were cooled and diluted to about 75 ml with tap water. The residual di- chromate was titrated with 0. 092 N ferrous ammonium sulfate solution to a potentiornetrically indicated endpoint of 0. 30 v. This emf level was selected because in preliminary potentiometric titrations of standard dichromate solutions under identical conditions, the emf drOpped rapidly through the range 0. 45 to 0. 10 V on addition of less than 0.1 ml of reagent. Glass and platinum electrodes were used with a Beckman model G pH meter. The literature reported that acid dichromate does not oxidize acetic acid at an appreciable rate without a catalyst. Silver sulfate was the only catalyst mentioned in the literature so various other compounds were tried as substitutes for the silver salt. In all of the experiments with the assorted catalytic agents. 1 g of the catalyst and 1 m1 of 1.0 N acetic acid was added to the oxidizer. This quantity of acetic acid was greatly in excess of the amount required to reduce all of the dichromate present. Table I gives a resume of the results of the experiments which indicate that silver is the only effective catalyst among the compounds tried. The reason for the reduction of dichromate by chromic acetate is not known. The oxidation of acetic acid in the presence of chromic sulfate does not occur. Micro oxidations using the Schollenberger oxidizer. Initially. the micro scale oxidations with Schollenberger’s oxidizer (dichromate dissolved in a mixture of concentrated sulfuric and phosphoric acids) were carried out by the volumetric method proposed by Johnson (1947). 10 TABLE I: THE CATALYTIC ACTIVITY OF VARIOUS COMPOUNDS TOWARDS THE OXIDATION OF ACETATE BY ACID DICHROMATE AT 1300C. Catalyst. 1 g Time Observation Catalysis none 2 hours remained orange none AgNO3 1 minute turned green positive NH4VO3 1 hour remained orange none CuSO4 1 hour remained orange none (NH4)ZFe(SO4)Z 1 hour remained orange none CrZ(SO4)3 1 hour remained orange none Cr(CZH3O;)3 5 minutes turned green positive ? 11 Because this method proved inconsistent due to the high viscosity of the Schollenberger oxidizer, a weight titration method with weighing burets was used. The procedure followed in making a micro oxidation deterxnin~ ation was as follows: Test tubes were Cleaned with hot nitric acid 0 of the and rinsed three times with distilled water. Approximately 2 g micro Schollenberger oxidizer was weighed to the nearest mg into each tube. Approximately 0. 3 to 0. 4 g Of sample, also weighed to the nearest trig. was then added and the mixture in the test tube was digested in a boiling water bath for 20 minutes. After the digestion period, the test tubes and contents were cooled to room temperature and diluted to about 10 ml with tap water. Approximately 1 g of potassium iodide crystals was added and the iodine released by the residual dichromate was titrated with O. 142 N thiosulfate from a micro buret. A starch solution was used to indicate the endpoint. A water blank was run simultaneously and the COD was calculated from the amount of dichromate consumed which is the difference between the thiosulfate required for the blank and the samples. The micro Schollenberger procedure was used to determine whether or not a lower concentration of silver sulfate than the amount mentioned in the literature would be adequate for catalysis. Varying amounts of silver nitrate ranging from O. 3 mg to 95 rng were weighed into test tubes. Then Schollenberger oxidizer, prepared without silver nitrate, and a solution of acetic acid having a known theoretical COD of 800 ppm was weighed into the tubes as in the procedure described above. Similarly, a number of water blanks were run with the silver nitrate varying from 0. 0 to 143 mg AgNO3/g oxidizer. It was found that good oxidation Of the acetic acid took place at 14.2 mg AgNO3/g oxidizer and above. As shown in Table II there was a great deal of variation in the water blanks above a level Of TABLE II: DATA OF THE EFFECT OF SILVER CONCENTRATION ON THE OXIDATION OF ACETIC ACID AND THE WATER BLANK USING MICRO SCHOLLENBERGER TECHNIQUE Oxidation of Acetic Acid Water Blank mg AgNO3/g oxidizer ppm COD mg AgNOL/g oxidizer meq/g“~ 0.3 0 0.0 0.126 0.6 155 0.0 0.126 1.1 0 1.3 0.124 4.2 13 1.9 0.131 14.2 833 3.0 0.126 13.2 395 23.3 0.127 18.2 blank 31.8 0.129 21.8 355 48.7 0.124 52.5 710 60.2 0.152 85.4 662 143 0.239 .‘ . ’/ . . =.~ — meq dichroniate/ g ox1dizer 13 48. 7 Big AgNO3/‘g oxidizer. The decision to use a silver salt level of o oxidizer was based on theSe data. The reason for the about 15 nig/g variation in the amount of oxidation at various silver salt levels is not known. An investigation to determine this variation could prove interesting. Some Of the Schollenberger oxidizer with 15 nig AgNO3/g was prepared and set aside to determine deterioration Of the oxidizer. The concentration of dichromate in the oxidizer was 0. 113 meq/lg on January 27, 1960 and 0.113 meq/g on July 31, 1960. A value of 0.115 meq/g was Obtained on April 25. 1960. From these values, it can be concluded that there is no decomposition of the dichrome in silver- Schollengerger oxidizer in 6 months. An experiment was conducted to determine whether the silver- Schollenberger oxidizer solution was stable at 980C. Trials with water blanks were run for 20 minutes at 980C and 250C (see Table III). The mean, r3, for 980 of 4 replicates. 2, was 0. 257 meq/g and the standard deviation, 3’ was 0.0026. At 250, four replicates were run to give 133: 0.259 meq/g and s : 0.0017. These results indicate that heating for 20 minutes at 980C produces no significant decomposition of dichromate. Oxidation of acetic acid using macro Schollenberger method. The macro scale oxidations of various substrates with Schollenberger oxidizer were carried out. by adding 0.2 to 0.4 g of silver sulfate to 5 ml of sample in a 125 ml Erlenmeyer flask. Twenty—five m1 of Schollenberger oxidizer was added to the sample in the flask and swirled to insure thorough mixing. The flask was placed in a boiling water bath for 20 minutes to digest the oxidation mixture. After the digestion, the flask was removed from the bath and the contents cooled and diluted to about 75 ml with tap water. The residual dichromate was TABLE III: 14 EFFECT OF TEIVIPERATURE ON THE STABILITY OF THE SCHOLLENBERGER OXIDIZER CONTAINING DISSOLVED SILVER SALTS U) 11 standard deviation Time Temperature Time meq CrzO7Z/g ox ma, sb 1 980C 20 min 0. 254 3 98 20 0. 257 4 98 20 0.260 398 = 0.0026 5 250C 20 min 0. 259 6 25 20 0. 260 1:12.; 2 0. 259 7 25 20 0. 258 8 25 20 0.262 335: 0.0017 a : mean titrated with 0.092 N ferrous ammonium sulfate solution to a potentio- metrically indicated endpoint of 0. 30 v. A water blank was run daily to check the oxidizer concentration. The COD was calculated from the difference between the oxidizer concentration in the sample and the water blank. In the first trials the digested, diluted, cooled sample mixture was treated with 1. 5 g Of potassium iodide crystals and the iodine released by the residual dichromate was titrated to the starch endpoint with 0. 142 N thiosulfate. Thiosulfate forms a complex with silver ion whereas ferrous ammonium sulfate solution does not; therefore. in later trials ferrous ammonium sulfate solution was used for determining the residual dichromate. Comparison Of COD oxidative procedures. A statistical com- parison Of reflux COD and the Schollenberger oxidizer methods is given in Table IV. It should be noted that the difference in the standard deviation of the two types of oxidizers is not significant on water blanks. The reflux method appears to be somewhat more effective and consistent for oxidizing acetic acid than are the Schollenberger oxidizer methods. The reflux method appears to be very satisfactory for the carbohydrate oxidations as indicated by the oxidation of sucrose. The mean is within 1‘72 of the theoretical COD value and the coefficient Of variation is about 1%. The micro Schollenberger oxidizer method gave an accurate mean, within 1’}; Of theory, but the coefficient of variation was 23%. The macro Schollenberger oxidation appeared to be more consistent than the micro procedure for carbohydrate oxidation as indicated by results from glucose oxidation. The reasons for the large carbohydrate oxidation variation of the micro method should be investigated. l6 111111111111 Ema .emogjw om. 0mm: w n . n3 Roam E .emo . .. - .. emu o. 2 m A: ow: e , . Nbudhfl>£0£km~w0 an Us ofleom f o a. 3: m: s m: ow: e E e. . . . s N - - - Moo .0 E; .o w moo .o s; o a. H O to 38 imam w. H m. m. H M M ed em. BED was Chums OHUMHC UCSOQEOU sonar/a0 powgoflceZOSOm xsflvm l mQZDOnflZOU Hman ZO mQOEHmfix/H DOD rmO ZOmHmH Mimm <8 , .... A“; .. 17 Chemical oxidations are strongly temperature dependent. The oxidizing power of dichromate increases with temperature. The greater oxidation of acetic acid by the reflux method is probably due to the fact that the reflux temperature was Ineasured at more than 1300C whereas the Schollenberger oxidizers were digested at 1000C. This temperature difference, although small, could account for the slightly better oxidation Of the acetic acid by the micro Schollenberger compared with the macro Schollenberger oxidizer method. The micro method insures better heat. transfer because of the much smaller volumes. Moreover. it is easier to maintain boiling temperatures throughout a small water bath than throughout a large one. Corrections for Chloride Oxidations Sodium chloride is found in pickle cannery effluents at concen- trations up to 1 percent. Although 1.0 N dichromate in 1 N acid at room temperature is not strong enough to oxidize chloride to chlorine. more strongly acid dichromate solutions at higher temperatures can be strong enough. This fact can be verified if the electrochemical potentials and the Nernst equation are considered: __ ' K (Ox) E — E0 T n Log (Red) The constant K is equal to 0.060 at 30°C, at 1000C the value is 0.075. The reduction of dichromate in acid medium occurs by the following equation: + : ' __ 4-” CrlO7 +13HT+6e ——-—>2Cr T+7HZO Assuming that this is a reversible reduction. and that E0 is 1. 30 v, the Nernst equation takes the following form: l8 _ 14 2 0.075 Log (11) (Cr307) O at 100 C. 6 ‘* (Cr+++)2 E:l.30+ If the available acidity is 10 N and the dichromate concentration is 0.1000 N or 0. 01667 M. the potential of the system at 50 percent reduction is: 0. 1 14 , t E :1.30+ ——9—7—§- 102(0) (0 (138): 1.44 v. 6 5 (0.0167) This oxidation potential is probably high enough to oxidize chloride to chlorine since the standard oxidation potential for this reaction is -l.36 v. Correction by titration. "Standard Methods" suggests that an aliquot of the sample be titrated with standard silver nitrate and that the oxidation of chloride be corrected for by subtracting the oxidizer consumed in the oxidation of chloride from the total used in the sample. This Inethod is unsatisfactory for use in the special case of pickle effluents because the correction would amount to 90 percent of the COD, obviously leading to large errors because the short period oxidations do not oxidizer the chloride quantitatively. Therefore, in general. the correction would be larger than the total COD measured. The result is a negative COD for the sample of waste which is impossible. Three methods were tried to correct for the presence of salt: 1) the oxidizing power of the dichromate was limited by addition of a large quantity of chromic ion. 2) the chloride was precipitated by silver nitrate and the precipitate removed by centrifugation, and 3) the chloride was inactivated by the addition of excess silver ion; the silver chloride was left in the oxidation mixture. lVIodification of the oxidizing power of the dichromate. In an attempt to modify the oxidizing power of the dichromate. a large amount of chromic ion was added to the oxidizer. This addition was 19 not effective in eliminating the oxidation of chloride. A possible difficulty which would arise from the use of this method is that the ability of the oxidizer to oxidize acetic acid would be decreased. Precipitation and separation. The separation of the chloride by silver nitrate precipitation followed by centrifugation was very effective in eliminating chloride interference. This method consists of determining the volume of a concentrated silver nitrate solution which is necessary to precipitate the chloride in an aliquot of the sample. An equal amount of the concentrated silver solution is added to another aliquot of the sample in a centrifuge tube. After centrifu- gation to settle all of the AgCl. the centrifuge was added to the oxidizer and the oxidation procedure continued as before. This method has several disadvantages. First, sample dilution is inconsistent. The dilution is in direct proportion to the amount of salt in the sample. larger volumes of silver nitrate solution being required for samples of higher chloride content. Since the rate of chemical oxidation is dependent on the concentrations of the reactants. such an irregular dilution will cause sonie variation in the COD de- termination. Secondly, the finely divided AgCl precipitate often requires 15 minutes or longer to settle out of suspension. Thirdly, the titration procedure to determine the amount of silver required and then the addition of this amount to another aliquot makes this procedure cumber some . Inactivation of the chloride. Because the oxidation potential of a system is dependent on the concentrations of the reactants, it would be possible to inhibit the oxidation of chloride by reducing its concen— tration in solution. The addition of excess silver ion will accomplish this by tending to form the insoluble silver chloride. Experimentally this was accomplished by adding an excess of silver sulfate to the sample. The precipitate of silver chloride was allowed to remain in the flask, There was no noticeable evolution of chlorine from the oxi- dations of chloride when about 0.1 g excess silver sulfate was used. A strip of starch-iodide paper was used to test for evolution of chlorine. When pure washed AgCl was added to the oxidizer and this mixture was refluxed, a trace of chlorine was observed. On refluxing l g of silver chloride with 0.1 g of silver sulfate for 30 minutes in the reflux COD procedure, four trials gave r_r_1 : 74 ppm withs = 10. Experiments showed that silver chloride exposed to light gave much higher values, m: 35.3, _s_: 33. Since the purpose of this study was to evaluate and develop a simple. fast method for estimating BOD of pickle plant wastes. the inactivation of the chloride by adding excess silver salt was considered sufficiently accurate and was used for all of the evaluations of samples containing salt. BOD‘s of Mixtures of Lactic Acid. Acetic Acid and Carbohydrates A series of BOD experiments was performed on mixtures of standard solutions of lactic acid. acetic acid. starch, and sucrose. The method of preparation of the standard solutions is described in the Appendix. The relative proportions of lactic. acid, acetic acid, and carbohydrate material were in a range normally found in a pickle manufacturing plant effluents. No proteinaceous material was included in these "artificial sewages” because refluxing a sample of effluent in alkaline permanganate gave off negligible quantities of ammonia. The. BOD of these mixtures was determined according to the procedure in ”Standard Methods. " A portion of the sample was analyzed at the end of the BOD incubation period to determine the amount of each component remaining. The theoretical BOD was 2.1 compared to the actual BOD and the effect of the various components in the mixture on the result was observed. The ”artificial sewages” were formulated so that there would be a theoretical BOD of about 1000 ppm. To run BOD determinations, 10 to 1 dilutions \\ ere made. Ten, 15 or 20 ml of these dilutions was placed into a BOD bottle; three bottles or replicates were run at each of the three volumes of substrate. This was in effect placing 1.00, 1.50 or Z. 00 ml of the original 1000 ppm sewage into each BOD bottle. Dilution water, which had been seeded with 5 nil/l of water from East Lansing's Red Cedar River, was added and the bottles placed into an incubator for 5 days. After the incubation period, sonne of the water from the BOD bottle was carefully siphoned into a 125 ml Erlenmeyer flask. The dissolved oxygen in this water was determined by the Winkler method as described in "Standard Methods. " One hundred ml of the water‘ remaining in the BOD bottle was treated with 1 drop of 0.1 N NaOH solution and placed in a 250 ml beaker. This beaker was set on a hot plate to evaporate at about 950C. When the volume in the beaker decreased to below 10 ml. the concentrate was poured into a 10 ml graduated cylinder and made up to 10.0 ml. One ml of this concentrate was analyzed colorimetrically for lactic acid by the Markus (1950) modification of the Barker-Summerson method (1941). One drop of 4 percent CuSO4 solution was added to the sample of concentrate in a test tube. Six ml of concentrated sulfuric acid was added with a syringe. After standing for 5 minutes in air, the reactants were cooled to 150C. One drop of 1 percent o~hydroxy~ diphenyl solution was carefully added and mixed thoroughly into the solution. The tubes were held for 12 hours at room temperature and the absorbancy was 'measured at 570 mu by a Bausch and Lomb "Spectronic ZO" spectrophotometer. A calibration curve was made by treating similarly various dilutions of a known solution of lactic acid. The optical density of the known solutions was plotted against the concentration and the unknown solution concentrations were determined frozn this curve. Morris‘ (1948) anthrone method was tried for carbohydrate determinations. Apparently the concentrations encountered. even in the concentrated sample. were too low for this determination and all of the tubes gave negative results. A more sensitive method would have to be used for carbohydrate determinations in future work. The colorimetric method of Pesez (1956) for the determination of acetic acid. gave negative results in all cases. This failure should have been investigated but the investigation was beyond the scope of this study. It was found necessary to make a detailed evaluation of two of the steps in the preceding procedure, namely, the transfer of oxygen depleted water from the incubated BOD bottles to the 125 ml Erlen- meyer flasks and the reproducibility of the concentrating of the incubated dilution to a volume at which the components could be determined quantitatively: Table V shows the values obtained for the reproducibility of lactic acid recovery by the evaporation technique. This evaluation was made by taking 10 m1 of a known dilution of a standard lactic acid solution, adding 1 drop of alkali and diluting to 100 ml with distilled water. The evaporation was carried out in a beaker as described above for the samples from the BOD dilution. The beaker containing the residue in trial 2, marked ”just dried, ” was removed from the hot plate before the residue was completely dry, the final drying occurring while the beaker containing the residue cooled off. The final drying took place at lower temperatures so that the residue was not over- heated. Residues in trials 5, 6, and 9 dried while on the hot plate TABLE V: RECOVERY OF LACTIC ACID BY EVAPORATION OF DILUTION WATER TREATED WITH ALKALI . , Percent Trial ppm Recovered T reatment Recovered 1 10. 5 normal to 10 ml 105 2. 10. 0 just dried 100 3 1i. 5 normal to 10 ml 125 4 11.5 normal to 10 ml 115 5 8. 5 dried and overcooked 85 6 Z. 0 dried and overcooked 20 7 11.7 normal to 10 ml 117 8 11.0 normal to 10 ml 110 9 7. 5 dried and overcooked 75 Z4 and as a result these were heated for several minutes without moisture at a temperature over 1000C. It is readily apparent from the data that evaporation under carefully controlled conditions can be used for concentrating highly diluted lactic acid samples for colorimetric analysis. The siphoning transfer technique was evaluated by the following procedure. Sonne oxygen depleted water was prepared by boiling distilled water in a flask. After cooling, 8 BOD bottles were filled by siphoning from the flask. A transfer was made frOIn 5 of the BOD bottles to five 125 ml Erlenmeyer flasks and the dissolved oxygen was determined by the Winkler method. The dissolved oxygen was also determined in the other .3 BOD bottles. The results are tabulated in Table VI. It should be noted that the standard deviation of both is essentially the same and that there is only about a 2. percent difference in the means. Apparently there is slight absorption of oxygen during the transfer but it can be considered negligible. BOD‘s of artificial sewage components. The concentration of lactic acid was determined by the colorimetric method. The percent of lactic acid oxidation was calculated by simply dividing the difference between the original concentration and the final concentration by the initial concentration and multiplying by 10.0. The percent oxidation of the other components was calculated from the BOD data by simultaneous equations. For example, if x : percent oxidation of acetic acid, y : the percent oxidation of sucrose and z z the percent oxidation of starch. Ifa is the. initial concentration of acetic acid in sewage l. _b_ is the initial concentration of sucrose in sewage 1, _c_ is the initial concentration of starch in sewage 1, and a' , _l_)_'_, and c' are the. respective initial concentrations a———— in sewage 2., and a", b", and c" are the reSpective initial concentrations Z5 TABLEYVL EVALUATKDQOF7UHZTRANSFERCH‘OXYGEN DEPLETED WATER FROM A BOD BOTTLE TO AN ERLENIVIEYER FLASK .. . ppniCDZ T rial Action ppm Oz mean std. dev. I transferred 3. 68 Z transferred 3.68 3 transferred 3.56 3.61 0.06 4 transferred 3.60 5 transferred 3.56 6 nottrans. 3.60 7 nottrans. 3.46 3.54 0.07 8 nottrans. 3.56 26 in sewage 3, the following series of simultaneous equations can be set up: ax + by + cz : BOD of sewage l a’z + b'y + c’z : BOD of sewage 2. a"x + b"y + c112. = BOD of sewage 3. Table VII lists the values for percent oxidation of the organic compounds as obtained from the BOD of artificial sewage. The values for lactic acid are from colorimetric data. When simultaneous equations were used to calculate the percent oxidation of lactic acid, a value of about 90 percent of theory was obtained compared to the 98 percent oxidation arrived at by colorimetric data. This discrep— ancy might be due to the fact that the microbial System present is multiplying. The lactate is probably used as a source of carbon in the formation of the protoplasni of the new cells which would have a BOD if subjected to oxidation. Although the lactic acid itself is entirely used up, some of its BOD is converted to protoplasm which is not oxidized completely in 5 days. TABLE VIII lists the results of the determinations made on several artificial sewage combinations. The standard deviations of the BOD values indicate that there is at least 5 percent variation from a mean of three or more detern1inations. It was found that the COD standard deviation was about the same expressed as ppm but the coefficient of variation was somewhat smaller with COD because the mean was larger. The artificial sewage number 2 differs from sewage number 1 in that some of the sucrose of sewage l was replaced by acetate. The difference in the COD's can be explained if the acetate was not oxidized at all by the COD method in either case. The non-acetate COD of sewage l is 990 ppm and the theoretical non-acetate portion of sewage 2 totals‘878 ppm. About 94 percent of the non—acetate TABLE VII: Z7 DEGREE OF OXIDATION OF PURE ORGANIC COM- POUNDS IN AN ARTIFICIAL SEWAGE BY STANDARD BOD C 1 Percent Oxidized onnpounc n high low 1T1 _s_ literature sucrose 14 87 73 80 5 49 to 76 starch .10 65 0 ~17 26 — lactic acid ll 100 95 98 3 63 acetic acid 6 100 95 97 3 34 to 88 Z8 flood ONN Lehman mm. mac m 00 IN hm uflvum N m f 03 a. om; 3: 282 00 :5 m 004 i I 1 00m wmouodm m Om 2 Egg . m. I 55% M m we owe m 0m .N +0 .4333 me. 2;. m 004 I n u cow vapudm #4 $3 8% ex wives QB 00 com M 0m 4 mm: 0333 +m 3.0 m 00.. u u u I: meMUSm m E as me ofloom mo USUNH : .. u u u we com o I: mmOpoSm N trod 9:5 3: on» m 00 .N em ofloom a; N; mmn m cm; +0 3733 cm 000 m 004 Ii :0 m cow smokesm A QdeLU ocsusonw .ml ml: w: amok/um it um: mm H. QOm GOO coCflmOQCMOU REP MO<3Mm J 3 CO‘2 + Z HlO A 0. 0156 N acetic solution has a theoretical COD of 1000 ppm. Standard lactic acid. The preparation and standardization procedure was the same for lactic acid as it was for acetic acid described above. The COD for lactic acid was based on the following equation: CH3CHOHCOOH + 6[O] ——> 3 coZ + 3 H30 A 0. 0156 N lactic acid solution has a theoretical COD of 1000 ppm. Standard glucose. Sonae C. P. glucose was recrystalized from water. This was ground, dried for an hour at 1000C. One gram was weighed out on an analytical balance to the nearest 0. 1 mg and dissolved in 1 liter of distilled water. Such a solution is 1000 ppm in glucose having a theoretical COD or BOD of 1066 ppm as calculated from the following relationship: Standard sucrose. Exactly l. 0000 g of commercial beet sugar, previously dried at 1000C was dissolved in 1 liter of water. This solution was 1000 ppm in sucrose having a theoretical BOD or COD of llZO ppm. The following oxidation was the theoretical basis for the value for the COD or BOD: C13H22011+ Z4[O] —-—> 12 CO; ~'~ ll HZO 42 Standard starch solution. Exactly 1.000 g of “soluble starch" was dissolved in about 25 ml of hot distilled water. This solution was then diluted to 1 liter with distilled water. The theoretical COD-BOD was calculated from the following: The 1000 ppm starch solution therefore represented a solution having about 1170 ppm COD. ~13 LITERATURE CITED APHA. AWWA and FSIWA, 1955, Standard Methods for the Exarni- nation of Water, Sewage and Industrial Wastes, 10111 Ed.. APHA, New'York. Aynsley, E. E.. 1950. Decomposition of chromic acid in sulfuric acid. J. Chem. Soc.. 19502568. Babbitt. H. E.. 1957, Sewerage and Sewage Treatment, 7th Ed.. J. Wiley & Sons. New York. Barker, S. B. and Suninierson. W. H.. 19—11, The colorimetric determination of lactic acid in biological medium, J. Biol. Chem. . 1581535. Burtle, J. and Buswell.. A. M.. 1937, Oxygen demand studies. Sew. W'ks. J.. 9222—1. Busch, A. W., 1958, BOD progression in soluble substrates, Sew. & Ind. Wastes,_3_0213’>6. Caldwell. D. H. and Langelier. W. F.. 1948. lvfanonietric measure- ment of BOD of sewage. Sew. Wks. J.. 22:202. Costilow. R. N.. 1960. Private Communication. Gaffnev. P. E. and Heukelekian, H. . 1958, Oxygen demand Ineasure- ' . . 7 ment errors 1n pure organic compounds. Sew. & Ind. Wastes, 30:505. and , 1958. Oxygen demand of the lower fatty acids. Sew. & Ind. Wastes, 10:673. Crotaas. H. B. . 19-18. Effect of temperature on biochemical oxidation of sewage. Sew. Wks. J.. 30:441. Gellinan. I. and Heukelekian, H.,1951. Studies of biochemical oxidation by direct methods, Sew. & Ind. Wastes, 2:126? Heukelekian, H. and Gellrnan, I.. 1951, Studies of biochemical oxidation by direct methods, Sew. Ind. Wastes, 3311546. 4—1 Heukelekian, H. and Rand. l\1. C. , 1955, Biochemical oxidation of pure organic compounds, Report of the Research Committee-- FSIWA, Sew. & Ind. Wastes, 27:10-10. Johnson, M. J. . 1949, Determining organic matter by chromic acid oxidation. J. Biol. Cheni., 1812707. Kieselbach, R., 195—1. Continuous recording of concentration of organic I'natter in waste water, Anal. Chem. . 26: 1312. Lea. W. L. and Nichols, M. S.. 1936, Influence of substrate on BOD, Sew. WkS. J.. £32435. Lee, E. W. and Oswald. W. J.. 1954, Comparative studies of the dilution and Warburg methods for determining BOD, Sew. & Ind. Wastes, 26: 1097. Markus, R. L. . 1950, Colorimetric determination of lactic acid in body fluids utilizing cation exchange for deproteinization, Arch. Biocheni., 29:159. Mercer, W. A. and Rose, W. W.. 1956, Research Laboratory Report 816-56-B, National Canners Association. Moore, W. A., Kroner. R. C. and Ruchhoft. C. C., 1949. Dichromate reflux for determining oxygen consumed, Anal. Cheni., _2_._1_:953. Ludzack, F. J. and . 1951. Determination of f M oxygen consumed values of organic wastes. Anal. Chem., 23: 1297. and Ruchhoft, C. C., 1951, Chemical oxygen consumed and its relationship to BOD, Sew. & Ind. Wastes, _2-3-:705 lVIorris, D. L., 1948. Quantitative determination of carbohydrates with Dreywood's anthrone reagent. Science 1071254. Muers. M. M., 1936, J. Soc. Chem. Ind., 55:71T. Oberton, C. E. and Stack, Jr., B. T., 1957, Biochemical oxygen demand of organic chemicals, Sew. & Ind. Wastes, 3.221267. Pesez, M. and Ferrero, J.. (abst. A. E. Meyer. CA 2127224), 1956, Ann. Pharm. Franc., 1:558. Rhame. C}. A.. 1947, Determination of BOD values by chemical oxidation. Water & Sewage Works. 94: 192. 1949. Determination of first—stage oxygen demand, Water 3‘ Sevvage Works, 96: 317. Schollenberger, C. J. . 1916. Total carbon in soil by wet combustion. Ind. Eng. Cheni., 8:11.36. Sierp. E.. (trans. S. L. Ncave), 1928. A new method for determining biochemical oxygen demand. Ind. Eng. Chem“ 20:247. Snethlage. H. C. S.. 1938. Influence of solvent and temperature on decomposition of chromic acid, Rec. Trav. Chiin.. 57: 1341. Unibreit. W. W.. Burris. R. H. and Stauffer, J. F.. 1957. hianometric Techniques and Tissue Metabolism, 3rd Ed. . Burgess Publishing Co.. lVlinneapolis. Minnesota. Winkler, L. W.. 1888. Die bestininiung des iin wasser gelosten sauerstoffes, Ber. 21:2843. Winneberger. J.. Kerrigan. J. and Whittier. R., C&EN May 2. 1960:46. 13.911331 1.331;. ill-1L IGAN STATE UNIVERSITY LIBRARIES mum HI Ill ILI lllll III! | 3 1293 O3 69 1771 MICH