r ^r^"" 
 
 L I E> HAHY 
 
 OF THE 
 
 U N IVER_SITY 
 
 Of ILLINOIS 
 
 628 
 lL6Sc 
 
 no- I 9-20 
 
 L "- i ""- , jii\-t£u'JW£j 
 
 ** COHF. ROO'r 
 
"?JWI 
 
 The person charging this material is re- 
 sponsible for its return on or before the 
 Latest Date stamped below. 
 
 Theft, mutilation, and underlining of books 
 are reasons for disciplinary action and may 
 result in dismissal from the University. 
 
 University of Illinois Library 
 
 SEP 9 «■ 
 
 
 .- 
 
 L161— O-1096 
 
Digitized by the Internet Archive 
 in 2013 
 
 http://archive.org/details/presenceoforgani19robi 
 
CIVIL ENGINEERING STUDIES 
 
 SANITARY ENGINEERING SERIES NO. 19 
 
 /sy 
 
 THE PRESENCE OF ORGANIC MATTER 
 
 AND ITS EFFECT ON IRON REMOVAL 
 
 IN GROUND WATER 
 
 By 
 L. R. ROBINSON, JR. 
 
 Supported By 
 
 DIVISION OF WATER SUPPLY AND POLLUTION CONTROL 
 
 U. S. PUBLIC HEALTH SERVICE 
 
 RESEARCH PROJECT WP-17 
 
 DEPARTMENT OF CIVIL ENGINEERING 
 
 UNIVERSITY OF ILLINOIS 
 
 URBANA, ILLINOIS 
 
 NOVEMBER, 1963 
 
THE PRESENCE OF ORGANIC MATTER 
 
 AND ITS EFFECT ON IRON REMOVAL 
 
 IN GROUND WATER 
 
 by 
 LLOYD R. ROBINSON, JR. 
 
 Supported by 
 
 Division of Water Supply and Pollution Control 
 U. S. Public Health Service 
 Research Project WP-17 
 
 Department of Civil Engineering 
 
 University of Illinois 
 
 U rbana , 1 1 1 i noi s 
 
 November, 1 963 
 

 £ ENGINEERING LIBRARY 
 
 ID 
 
 THE PRESENCE OF ORGANIC MATTER AND ITS EFFECT ON IRON 
 
 REMOVAL IN GROUND WATER 
 Lloyd R. Robinson, Jr., Ph.D. 
 Department of Civil. Engineering 
 University of Illinois, 1964 
 
 Studies were made of several ground waters in Illinois in which the 
 presence of organic matter was demonstrated. The concentrations of this 
 organic matter and iron in these waters were compared with reported iron 
 removal efficiencies by the treatment plants at these towns. A town which 
 had always reported satisfactory iron removal was found to have a raw 
 water with a low and fluctuating iron content of less than 1. mg/1 and 
 total organic extracts in the amount of about 1. 5 mg/l. The towns which 
 had reported at least occasional difficulties with iron removal were found 
 to have raw water which contained in excess of 1. mg/1 of iron and total 
 organic extracts in amounts averaging about 5 mg/1. 
 
 Carbon filters were used to concentrate the organic matter from 
 the waters studied. It was found that two filters in series, with the water 
 acidified before it was applied to the second filter, would produce at least 
 twice the amount of extract as the unacidified filter alone. Chloroform, 
 ethanol, ethanol plus ammonia, and ethanol plus hydrochloric acid were 
 used serially as solvents to extract the organic matter from the activated 
 carbon. Each solvent in turn produced considerable extract; however, the 
 extracts obtained with ethanol plus hydrochloric acid contained considerable 
 
ammonium chloride and were of little value in subsequent analyses. 
 
 The extracts thus obtained were analyzed for possible interference 
 with iron removal by filtration, Some of the extracts obtained with polar 
 solvents were found to maintain as much as one milligram of iron in a fil- 
 terable condition per mg of extract. The filters used for these experiments 
 were 0, 22 |i and 0.45(i membrane filters. It was found that, if sufficient 
 additional iron was added to a solution prior to filtration and after the first 
 appearance of iron on the filter, practically all the iron could be filtered 
 from the solution. It thus appeared that the interference with iron filtra- 
 tion was caused by colloidal dispersion and not by chelate formation as 
 had been proposed by some investigators. 
 
 The extracts were characterized by color, odor, characteristic 
 titration curves, infra-red spectroscopy, chemical oxygen demand, and 
 carbon analysis by wet combustion, and found to exhibit the characteris- 
 tics of humic acids or acids formed during the decomposition of vegetable 
 matter. 
 
 The extracts were applied to raw waters taken directly from the 
 wells under investigation. Effects on the rate of oxidation from ferrous 
 to ferric iron and the rate of change from a filterable to a nonfilterable 
 condition were measured. No appreciable interference was noted. Tar- 
 taric acid was used as a reference in these experiments and in the fil- 
 tration study. 
 
 Three items were noted in regard to possible organic interference 
 with iron removal by rapid sand filtration. The presence of organic matter 
 
is apparently associated with difficulties with iron removal; but the extracts 
 failed to exhibit interference with iron removal in the field studies. How- 
 ever, previous studies at the University of Illinois appear to link the pas- 
 sage of iron through .the filters with reduction of iron in the filters. These 
 filters in which iron was reduced were all found to have a surface slime 
 layer. It is thus hypothesized that the actual interference with iron removal 
 is biological activity, which is supported by the organic matter in the ground 
 water. Thus the slime layer creates reducing conditions at the surface of 
 the filters, which reconvert the iron to the soluble state. 
 
Ill 
 ACKNOWLEDGEMENTS 
 
 The experimental work presented in this thesis has been accom- 
 plished under Research Grant WP-17 (formerly RG6436) from the United 
 States Public Health Service, National Institutes of Health, entitled 
 "Fundamental Factors in the Treatment of Iron Bearing Waters, " Per- 
 mission to use the data is gratefully acknowledged. The author was also 
 supported during the last year of his research by a Traineeship from the 
 same organization. This too is greatly appreciated. 
 
 The author also wishes to thank Dr. R, S. Engelbrecht, advisor, 
 Dr. J. T. O'Connor, Assistant Professor, and Mr. G. E. Margrave of 
 the Bureau of Public Water Supplies, Illinois Department of Public Health 
 for their guidance and assistance. 
 
 The help of the laboratory staff, especially Mr. Miles Norton s 
 laboratory mechanic, and Mr. Steve Young, undergraduate assistants is 
 gratefully acknowledged. 
 
 This report was submitted as a thesis in partial fulfillment of the 
 requirements for the degree of Doctor of Philosophy in Sanitary Engineer- 
 ing under the direction of Dr. R. S. Engelbrecht, Professor of Sanitary 
 Engineering. 
 
TABLE OF CONTENTS 
 
 IV 
 
 Chapter 
 
 Page 
 
 ACKNOWLEDGEMENTS 
 LIST OF TABLES 
 LIST OF FIGURES 
 I. INTRODUCTION 
 
 A. Nature of the Problem 
 
 B. Purpose and Scope of the Study 
 
 II. PRESENT KNOWLEDGE AND THEORETICAL CON- 
 
 SIDERATIONS 
 
 Extracts 
 
 D. Field Testing of the Organic Extracts 
 
 E. Characterization of the Organic Extracts 
 
 F. Chemical Analyses 
 
 111 
 
 VI 
 
 VI 1 
 
 9 
 
 A. The Inorganic States of Iron in Water 9 
 
 B. The Reactions of Iron with Colloids in Water 13 
 
 C. The Reactions of Iron with Organic Chelating 
 Agents in Water 23 
 
 III. EXPERIMENTAL EQUIPMENT AND PROCEDURES 30 
 
 A. Isolation of Organic Matter from Ground Water 30 
 
 B. Experimental Technique for Isolation of Organic 
 Matter from Ground Water 33 
 
 C. Interference with Iron Removal by Organic 
 
 49 
 56 
 58 
 62 
 
Chapter Page 
 IV. EXPERIMENTAL RESULTS AND DISCUSSION 63 
 
 A. Characterization of -the Waters Studied 63 
 
 B. Measurement and Recovery of Organic 
 
 Matter in Well Water 66 
 
 C. Effects of Organic Extracts on the Filter ability 
 
 of Iron 7 9 
 
 D. Characterization of the Organic Extracts 90 
 
 E. Field Study of the Effects of Organic Extracts 
 
 on Iron Removal 97 
 
 F. Proposed Source of the Interference with 
 
 Iron Removal 101 
 
 V. CONCLUSIONS 104 
 
 VI. AREAS OF FUTURE STUDY 106 
 
 VII. BIBLIOGRAPHY 107 
 
 APPENDIX A 116 
 
 CHEMICAL ANALYSES 116 
 
 APPENDIX B 120 
 
 LOG OF FILTERABILITY STUDY 120 
 
 APPENDIX C 134 
 
 INFRA-RED SPECTROGRAPHS 134 
 
 APPENDIX D 138 
 
 LOG OF OXIDATION STUDY 138 
 
 VITA 140 
 
LIST OF TABLES 
 Table Number Page 
 
 i Mineral characteristics of waters studied 65 
 
 f-i 
 
 2 ANALYSES OF WATERS AS APPLIED TO CARBON 
 FILTERS 67 
 
 3 ORGANIC EXTRACTS RECOVERED FROM WELL 
 WATERS 69 
 
 4 CONCENTRATION OF ORGANIC MATTER EXTRACTED 
 FROM WELL WATERS 70 
 
 5 REMOVAL OF CHEMICAL OXYGEN DEMAND BY 
 ACTIVATED CARBON FILTERS 74 
 
 6 CHEMICAL OXYGEN DEMAND OF ORGANIC EXTRACTS 
 AND PURE COMPOUNDS 7 5 
 
 7 TOTAL CARBON ANALYSES 77 
 
 8 EFFECTS OF ORGANIC EXTRACTS ON THE 
 FILTERABILITY OF IRON 81 
 
 9 PAPER CHROMATOGRAPH OF ORGANIC ACIDS 92 
 10 SOLVENT SEPARATION OF ORGANIC EXTRACTS 96 
 
Vll 
 
 LIST OF FIGURES 
 
 Figure Number Page 
 
 1 COLLOID PARTICLE 16 
 
 2 ZONES OF FLOCCULATION 20 
 
 3 FILTER ARRANGEMENT 35 
 
 4 COMMERCIAL GLASS FILTER 37 
 
 5 LOCALLY FABRICATED PLASTIC FILTER 38 
 
 6 CARBON FILTERS IN OPERATION 39 
 
 7 SOXHLET EXTRACTORS IN OPERATION 43 
 
 8 SUMMARY OF EXTRACTION SCHEME 47 
 
 9 ASCENDING PAPER CHROMATOGRAPH 91 
 
 10 TITRATION CURVES FOR ORGANIC EXTRACTS 95 
 
 11 IRON OXIDATION AND FILTRATION STUDY AT 
 
 PHILO s ILLINOIS . 98 
 
 12 IRON OXIDATION AND FILTRATION STUDY AT 
 CLINTON, ILLINOIS 100 
 
I. INTRODUCTION 
 
 A. Nature of the Problem 
 
 Iron is a common constituent of ground water in many parts of the 
 country. Two of every three well water supplies in Illinois contain in 
 excess of 0.4 mg/l of iron. Forty percent of the public water supplies 
 contain iron in excess of 1.0 mg/l (1). In Arkansas, Louisiana, Oklahoma, 
 and Texas, seventy-seven percent of the water supplies are well supplies 
 and this well water generally has a high iron content (2). Much of the 
 literature which will be cited later will bear out the fact that problems 
 with iron in water supplies are fairly general throughout the world. 
 
 The recommended U. S. Public Health Service limit for iron in 
 public water supplies is 0.3 mg/l (3). Mohler (4) reports troublesome 
 taste and turbidity problems with water containing as little as 0. 25 mg/l. 
 Hauer (2) has cited additional reasons for limiting the amount of iron in a 
 water supply. Iron in excess of 0. 1 mg/l causes discoloration of pulp and 
 paper in the paper industry; the rayon industry prefers the iron content of 
 water to be below 0. 05 mg/l; iron causes spots and discolorations in the 
 tanning of leather; it discolors, clouds, and affects the taste of liquor; 
 stains plumbing fixtures and laundered clothes; makes tea black; darkens 
 boiled vegetables; and iron provides an energy source for iron bacteria in 
 water mains. 
 
 Salbach (5) in 1868 first used aeration and filtration for iron removal 
 and the first plant was constructed in Charlottenburg, Germany xn 1874. An 
 
iron-removal plant was serving Champaign and Urbana, Illinois in 1916 (6) 
 and by 1957 there were about 170 municipal, plants in Illinois providing iron 
 removal (1). This rapid growth in the number of iron removal plants has 
 not been limited to Illinois alone, but in other areas of high iron content 
 as well. As the number of plants increased, so did the problems associ- 
 ated with the successful treatment of certain waters. In order to solve 
 these problems, modifications of the conventional treatment of aeration., 
 sedimentation, and filtration were tried, as were entirely different meth- 
 ods of treatment. Applebaum and Bretschger (7), Babcock (8), Fosnot (9) 3 
 Longley (10), and others have discussed the more common methods of iron 
 removal which, include: 
 
 1. Aeration followed by filtration. 
 
 2. Aeration, sedimentation, and filtration. 
 
 3. Aeration, addition of lime to raise the pH, sedimentation, 
 and filtration. 
 
 4. Same as three but with the addition of special coagulants 
 in addition to lime. 
 
 5. Aeration and filtration through a manganese zeolite. 
 
 6. Filtration through a sodium zeolite without prior aeration. 
 
 7. Substitution of chemical oxidants for aeration in any of the 
 above processes. 
 
 Van der Wal (11) has reported that iron and manganese are diffi- 
 cult to remove from water by sand filtration without prior floe formation. 
 There are several references in the literature pointing to the necessitv of 
 
3 
 aeration not only to provide oxygen for oxidation of iron but to strip out 
 dissolved carbon dioxide followed by the addition of lime to raise the pH 
 to the alkaline range where ferric iron is the least soluble (12, 13 3 1.4). 
 With the iron formed as an insoluble precipitate, alum or one of the fer- 
 ric salts may be added to aid in floe formation resulting in more efficient 
 sedimentation and longer filter runs. Babcock (15) reported that sedimen- 
 tation could be further improved by using an upflow clarifier to aid clari- 
 fication by contact with a previously formed sludge blanket. Careful pH 
 control, in the range of 8. 5 to 9. 1 was essential to the successful forma- 
 tion of a good floe with this type of treatment on the water studied. 
 
 Several methods of oxidation have been employed for waters 
 which appeared resistant to simple aeration. Mathews (16) has demon- 
 strated successful oxidation and iron removal with the maintenance of a 
 free chlorine residual in the filter effluent. Contact aerators using various 
 media with and without injected aeration have proved satisfactory (17, 18, 19). 
 As little as 0. 5 mg/l of copper added to the water prior to filtration acts as 
 an oxidation catalyst and aids in adsorption of iron on sand filters (11). Potas 
 sium permanganate has been found by some to be a satisfactory oxidizing 
 agent (20, 21, 22). If a cation exchange resin is treated with manganese, 
 the sodium is replaced with manganese to form a resin which has proved 
 satisfactory for iron and manganese removal. The manganese in the resin 
 is oxidized to a valence of +4 (manganic form) with potassium permanganate; 
 and then, when iron and manganous manganese are passed through, this gene- 
 rated resin, they are oxidized to higher insoluble forms and then removed 
 
4 
 by filtration. When the capacity of the medium is exhausted, the filter is 
 backwashed and then treated again with potassium permanganate. It has been 
 found, however, that if the potassium permanganate is added to the water to 
 be treated prior to filtration, the filter can be operated continuously until 
 increased head loss dictates the necessity of backwashing. 
 
 Ion exchange or the use of the zeolite process for iron removal as 
 well as for softening has been practiced in some, instances but the zero hard- 
 ness water and the increased cost of softening 100 percent of the water is 
 somewhat of a drawback. Davy (23) reported that several towns in Wisconsin 
 have had no trouble with complete hardness reduction; however, the pH must 
 be controlled between 8. 2 and 8.4 to control corrosion. This writer has 
 observed that the same procedure is used satisfactorily at Atwood, Illinois. 
 Will (24) stated that some iron will remain on the zeolite even after regene- 
 ration and he recommended aeration and sedimentation ahead of softening 
 to reduce this difficulty. Accumulated iron can be removed by treating 
 the filter with 10 percent hydrochloric acid. 
 
 No matter what method of iron removal is employed, difficulties 
 and incomplete removal often are apparent. As early as 1909, Weston (25) 
 noted that colloids so small that they would not settle out of a water inter- 
 fered with iron sedimentation. In 1914, he said that "certain kinds of 
 organic, matter . . . humic acids . . , interfere with coagulation of iron 
 when the water is excessively aerated (5). " Since that time there have been 
 many references to colloidal interference with the removal of iron. Chemi- 
 cal coagulation, with and without chemical oxidation by chlorination or 
 
treatment with potassium permanganate, has been found to aid in overcoming 
 this interference (16, 22, 26, 27, 28, 29, 30, 31, 32, 33). However Adams 
 (34), Beaujean (35), Beneden (36), Boorsma (37), McCrea (38), Nordell (39), 
 and Scharp (40) have reported the interference to be due to the formation of 
 complexes with organic matter or chelation of humic acids with the iron 
 and have found that conventional treatment methods will not provide ade- 
 quate treatment. 
 
 In a study by Longley (10), further discussed by Engelbrecht, et al. 
 (41), waters, both natural and synthetic, with varying concentrations of 
 sulfate, nitrogen, alkalinity, hardness, chloride, and iron were subjected 
 to iron removal treatment consisting of aeration, sedimentation, and rapid 
 sand filtration in a pilot plant. Iron removal efficiency reached 99 percent 
 in all cases and it was concluded that these inorganic mineral constituents 
 or their concentrations have little or no effect on iron removal . 
 
 Ghosh (42) studied the rate of oxidation of iron in aerated ground 
 wa,ter and found that small temperature changes, dissolved oxygen if 
 present in excess of the stoichiometric requirement, sulfate,, and chloride 
 had no effect on the oxidation rate. He found that the equilibrium pH after 
 aeration and the alkalinity were the controlling factors in the rate of iron 
 oxidation. Using a trivariate regression analysis, he proposed the fol- 
 lowing equation for the half life of ferrous iron after oxidation (the time 
 required for one half of the ferrous iron present to be oxidized to the fer- 
 ric state); 
 
14 . 2 
 
 T 1/2 = 521.854 - 0. 3278 x 10 x (OH") -182.931 log x(Alk. )+ 8. 10. 
 
 T 1/2 - half life in minutes. 
 
 (OH~) = hydroxyl ion concentration in mol/ 1 (from equilibrium pH) . 
 
 Alk. = total alkalinity as mg/1 of CaCO . 
 
 This equation was considered valid for equilibrium pH values between 
 7.48 and 7.78 and alkalinity values between 354 and 610 mg/l as CaCO_. 
 Since two of the towns studied, Clinton and Deland, Illinois, were in the 
 present study found to contain considerable amounts of organic matter, the 
 presence of organic matter in ground water does not appear to influence the 
 rate of oxidation of ferrous iron. . 
 
 The work on ground waters in Illinois by Longley (10) seems to 
 eliminate such inorganic mineral content as sulfate, nitrogen, alkalinity, 
 hardness, chloride, and iron as major factors causing interference with 
 iron removal, at least for the Illinois ground waters studied. Ghosh (42) 
 found the rate of oxidation to be affected principally by alkalinity and pH 
 but only the rate was affected and not the overall, amount of oxidation. The 
 question then presents itself: Is the interference then due to the presence 
 of organic compounds in the water? If the problem lies in the presence of 
 organic matter, is the problem one of complex formation or chelation so 
 as to convert ferric iron to a soluble state, or is it a problem of colloidal 
 formation and dispersion? Marsh (43) made a field study of possible inter- 
 ference with oxidation of iron by organic extracts. The extracts used in 
 the study by Marsh were some of the extracts obtained by this author in 
 the present study. Marsh reported that the extracts showed no interference 
 
with oxidation of the iron in ground waters even in concentrations much 
 higher than those found naturally. 
 
 Bo Purpose and Scope of the Study 
 
 This study was designed to answer the question: Are there organic 
 compounds present in well waters which interfere with the removal, of 
 iron by aeration, sedimentation, and filtration? 
 
 The first step in answering this question was to determine if there 
 was actually any organic matter in the well water in Illinois. For the study, 
 cities in Illinois were selected which had reported some difficulties with 
 iron removal. As a control or reference, a study was also made at Philo, 
 Illinois where satisfactory iron removal had always been reported. 
 
 Organic matter was extracted from the well water using a carbon 
 filter and then extracted from the activated carbon with suitable solvents, 
 dried, and weighed to determine the amount of organic matter obtained. 
 The organic matter thus obtained was then dissolved in water to which 
 iron was added. The water was then membrane filtered to determine how 
 much iron could be held in a filterable condition by a given amount of 
 organic extract. Plant operational records were also compared with the 
 above information to determine the correlation between the presence of 
 organic matter and effectiveness of iron removal treatment. 
 
 Field tests were then made to determine if the organic matter natu- 
 rally present in the ground water or if addition of extracted organic matter 
 could hold iron in natural waters in a filterable condition even after 
 
8 
 aeration, sedimentation^ and filtration through a membrane filter. 
 
 A correlation seemed to exist between the presence of organic 
 matter and iron removal efficiency. This possible correlation was not 
 apparent for the water at Atwood probably because of the practice of 
 treating all the water by cation exchange. Therefore 3 the extracts were 
 characterized by measuring their chemical oxygen demand, by studying 
 their infra-red spectra, and their titration curves. Further character- 
 ization was made using paper chromatography and polarography. These 
 extracts exhibited the characteristics of the hydroxyl acids commonly 
 called humic acid. Thus, the study of these acids as they affect the 
 removal of iron from well water is the subject of this thesis. 
 
9 
 II. PRESENT KNOWLEDGE AND THEORETICAL CONSIDERATIONS 
 
 A. The Inorganic States of Iron in Water 
 
 Iron is abundant in nature, composing about five percent of the 
 
 earth's crust. It exists mainly as anhydrous ferric oxide or hematite, 
 
 Fe„0 ; but also in a number of hydrated forms and in combination with 
 
 other anions. Other forms are turgite, Fe " 0. 5H 0; gothite and lepido- 
 
 crocite, Fe *H 0; limonite, Fe • 1. 5H 0; xanthosiderite, Fe • 2H 0; 
 2 3 2 2 3 2 2 3 2 
 
 limnite, Fe * 3H 0; magnetite, Fe ; siderite, FeCO ; pyrite, FeS ; 
 2 3 2 6 3 4 3 2 
 
 and, in addition, iron is usually contained in the silicate minerals of 
 dark-colored igneous rocks, in the cementing material in sandstones, 
 and in shales and most carbonate rocks (10, 44, 45). 
 
 Bass Becking, et al. (46) and Schoeller (47) have reported that all 
 natural, environments fit into a limited area of pH - Eh values and that the 
 pH, Eh, and carbon dioxide and sulfate concentrations dictate the types of 
 iron present in nature. Hem (48) has also reported that in the absence 
 of excessive carbonates or chemical complexing agents, pH and iron con- 
 tent of a ground water serve as indicators of the Eh of the water in its 
 natural state. The Eh, or the measured electrode potential in volts with 
 reference to the standard hydrogen electrode, is an indication of the oxi- 
 dation-reduction potential of the water at the instant of measurement; 
 however, as pointed out by Stumm in a discussion of Hem's article and in 
 a study by Komolrit (49) » ambiguous interpretations of Eh measurements 
 in natural waters are all too possible and a selective electrode system is 
 
10 
 needed. Low Eh values are encountered where reducing conditions exist. 
 Reducing conditions are usually created by microorganisms even to the 
 extent of reducing iron from the ferric to the ferrous state at. the ground 
 surface by some facultative heterotrophs. Under these conditions of low 
 potential, iron exists in the ferrous state and, as the Eh is raised by the 
 addition of oxygen or some other oxidizing agent, ferric iron becomes more 
 and more predominant. Ferrous iron is generally more soluble than ferric 
 iron with the solubility being greatly affected by pH. Hem and Cropper (50) 
 have developed a stability -field diagram of the state of iron in water as a 
 function of Eh and pH. The amount of iron that theoretically could be pres- 
 ent in solution is generally below 0. 01 mg/1 if the pH is between five and 
 eight and the Eh is between 0, 30 and 0. 50 volts. However, at a pH of five 
 and an Eh of 0. 30 volts, ferrous iron in solution could reach a concentra- 
 tion in excess of 100 mg/1. From this it can be seen that, unless iron is 
 held in solution by chelation or in suspension by colloidal dispersion, the 
 normal pH range and dissolved oxygen content of the water treated in con- 
 ventional iron removal plants should be sufficient to reduce the iron con- 
 tent of a water to a satisfactory level. 
 
 The reactions involved in the removal of iron from well water are 
 not. simply oxidation of ferrous ions to ferric ions followed by the formation 
 of insoluble ferric hydroxide at neutral or basic pH values. Stumm and 
 Lee (51) and Stumm and Morgan (52) have presented a thorough study of the 
 various ionic states in which iron is found in water. Soluble ferrous iron 
 occurs mainly as [Fe(H 0)1++, [Fe(H 0) (0H)J + , and [Fe(H 0) (OH) ] ". 
 
11 
 
 The solubility of these constituents is primarily controlled by the solubility 
 of ferrous hydroxide, ferrous carbonate^ and ferrous sulfide. The solu- 
 bility of ferric iron is controlled in natural waters by the solubility of ferric 
 hydroxide. The reactions of ferric iron in water are of importance not only 
 in iron removal processes but also in processes where ferric iron is used 
 as a chemical coagulant. 
 
 Iron usually has a coordination number of six (53) which means 
 that it can share electron pairs •with six donor atoms through coordinate 
 bonds. The donor atoms are quite often oxygen from water 'molecules. 
 Stumm and Lee (51) and Stumm and Morgan (52) reported that the hydrated 
 ferric i^on thus formed is an acid in the Bronsted sense and the complex 
 ion reacts again with water to form the acid -base equilibrium: 
 
 £Fe(H 2 0) 6 ]+++ + H 2 = [ Fe(H 2 0) 5 (0H)]++ + H3O+. 
 The conjugate base of this reaction can again transfer a proton: 
 
 [Fe(H 2 0) 5 (0H)] f+ + H 2 =[Fe(H 2 0) 4 (0H) 2 ] + + H 3 0+. 
 These reactions tend to decrease the pH of the solution, but in very alka- 
 line solutions these hydrolytic reactions can continue through two more 
 steps until the anion [Fe(H 2 0) 2 (0H) 4 ]~ is formed. 
 
 These ferric hydroxo complexes have a pronounced tendency to 
 polymerize. Two molecules of the trivalent ferric ion can combine., 
 probably through two hydroxo bridges, to form a dimeric ion. 
 
 2[Fe(H 2 0) 5 (0H)] ++ = [Fe 2 (H 2 0) 8 (0H) 2 J 4+ + 2H 2 .0. 
 
 or 
 
 H 
 
 (H 2 0) 4 Fe' ^Fe(H 2 0) 4 
 H 
 
 4+ 
 
12 
 
 •2 
 
 It has been estimated that in a 10 ~ J M solution of ferric iron only 
 
 about twenty percent of the ferric iron is present in the form of tripositive 
 ferric ion. The balance is made up of about 40 percent [FeCH^O^OH) ] ++ , 
 five percent [Fe(H 0) (OH) J+, and 35 percent [Fe 2 (H 0) (OH) ] 4+ . As a 
 solution is allowed to stand, more and more hydrolytic (olation) reactions 
 take place and the net charge on the iron becomes less positive. At higher 
 pH values, the hydrolysis reactions become more involved. Riddick (54) 
 reported that the anionic species [Fe^oOWOH^]- becomes predominant 
 to the point where the iron in suspension exhibits a net negative charge. 
 Polymerization (oxolation) reactions or dehydration continue to take place 
 as a solution ages leading to a progressive coordination of ferric ions 
 through hydroxy! links until, eventually, colloidal hydroxo polymers 
 and, ultimately, insoluble hydrous ferric oxide precipitates are formed 
 (51, 52). 
 
 Black, in a discussion of the article by Stumm and Morgan (52), 
 pointed out that ligands other than (OH)" may be present in the solvate 
 shell of the free metal cation. Stumm and Lee (51) listed chlorine, sul- 
 fate, and phosphate as ligands which could replace hydroxide in the iron 
 hydroxo complexes. The solubility of ferric iron at pH 7 is about 18 
 micrograms per liter but when the hydroxide in the complex ions is 
 replaced by other ligands the solubility is increased. Many organic 
 bases form strong soluble complexes with iron. In natural waters, high 
 concentrations of organic material such as humic acids and lignin deriv- 
 atives are frequently associated with high concentrations of soluble iron. 
 
13 
 Complex formation may be at least partially responsible for the high solu- 
 bility; however, it is also possible that, colloidal ferric, hydroxide is 
 stabilized and protected as a sol by organic compounds. 
 
 B. The Reactions of Iron with Colloids in Water 
 
 Colloids or colloidal particles are generally defined as particles 
 dispersed in a solvent medium which are too small to be seen with an 
 ordinary laboratory microscope but which are larger than individual 
 molecules. Colloids are thus particles with diameters between one 
 micron and one millimicron (10 angstrom units). Where ions and mole- 
 cules carry one or a few electronic charges, the much larger colloidal 
 particles may carry thousands. With these larger charges complete dis- 
 sociation becomes the exception rather than the rule. In fact these inter- 
 actions are one of the determining factors of their behavior. As the 
 particles increase in size from that of simple molecules, they may grow 
 into large spheres, cylinders, rods, plates, or threads. The shape thus 
 attained gives colloids their properties of high viscosities in dilute so- 
 lutions or the formation of iridescent layers. When atoms or molecules 
 are involved in interactions, the whole or a significant part of the whole 
 becomes involved., but with colloids many interactions are limited to the 
 surface. As the proportion of atoms located in the surface is significant 
 and their number large, the total amount of interaction and its affect on 
 the colloid often becomes very important. This unique field of chemistry 
 is discussed in detail in volumes by Ware (55), Weiser (56), Mysels (57) , 
 
14 
 and was the subject of a recent Rudolfs Research Conference at Rutgers 
 University (58). 
 
 Colloids can be classed in two general catagories; lyophilic or 
 solvent-loving colloids where the two phases show a marked combining 
 power and both phases are continuous such as starch in water, and lyo- 
 phobic or solvent -hating colloids where the solvent phase is continuous 
 and the colloids remain dispersed by forces acting on the colloidal parti- 
 cles such as clay or silica in water. Iron, finely dispersed in water, is 
 of this latter type. Lyophobic or hydrophobic (water -hating) colloidal 
 solutions or sols are the systems generally encountered in water treat- 
 ment and will be the type to be considered here more fully. 
 
 Colloids are subject to three principle driving forces. Van der Waals 
 forces are the mutual attractions between two particles brought about by the 
 mutual repulsion of like charges of the inherent molecular polarity. With 
 the like charges repulsed, one set of opposites is brought closer together 
 providing an average net attraction. This charge is usually too small to 
 overcome the surface or electrochemical forces which may either attract 
 or repel the particles as dictated by the surface charges on the individual 
 particles. When the electrochemical forces are of the same charge, they 
 are quite often strong enough to prevent agglomeration of the particles. 
 
 This mutual repulsion is what insures the stability of a colloidal suspen- 
 
 i 
 
 sion. Although colloids are much larger than simple molecules, their mass 
 is so small that sedimentation due to the third driving force, gravity, is 
 negligible. 
 
15 
 The electrochemical force or charge on the surface of a colloid may 
 
 be formed by the preferential adsorption of one ion of the supporting electro 
 lyte and/or by direct ionization of some of the surface molecules. Silver 
 iodide provides one example of how the electrolyte can affect the charge. 
 If silver iodide, which has a definite crystal lattice, is suspended in a solu- 
 tion of hydriodic acid, the iodide ions in solution would be preferentially 
 adsorbed on the surface of the silver iodide crystals and the colloids would 
 acquire a negative charge. If silver nitrate were the supporting electrolyte, 
 silver ions in solution would be preferentially adsorbed and the colloids 
 would acquire a positive charge. Attracted to these charged particles are 
 oppositely charged ions from the solution. Some of these attracted ions are 
 held rigidly to the particles or micelles while the rest of the attracted ions 
 form a diffuse or second layer surrounding the inner attached layer. The 
 adsorbed ions which constitute the inner portion of the double layer are 
 called the stabilizing or potential-determining ions and the ions which 
 constitute the outer diffuse portion of the double layer are called the coun- 
 ! ter ions. 
 
 The total, or Nernst potential on a particle is reduced with, distance 
 from the colloid surface as more and more counter ions are encountered. 
 The reduction in potential caused by counter ions in the fixed portion of the 
 double layer is called the Stern potential and the potential in the diffuse 
 layer is called the zeta potential. Figure 1 is a diagram of a colloid parti- 
 cle showing how the potentials are established. 
 
 In order to destabilize colloids so that they can agglomerate or 
 
16 
 
 Rigid 
 
 Attached 
 
 Particle 
 
 Solu- 
 tion- 
 
 j^ 
 
 -Stern 
 Potential 
 
 S 
 
 Nernst 
 Potential 
 
 Zeta 
 Potential 
 
 DISTANCE 
 
 FIGURE 
 
 COLLOID PARTICLE 
 
17 
 flocculate into particles large enough to be affected by gravity and settle from 
 solution, the Nernst potential must be overcome sufficiently to allow Van 
 der Waals forces to hold the particles when brought into contact with one 
 another. As the Stern potential reduces the Nernst potential by the amount 
 of the charges on the counter ions in the fixed portion of the double layer 
 and as the fixed layer is so thin that particles so close that their fixed layers 
 touch are usually agglomerated by mutual attraction, only the potential in the 
 diffuse layer, the zeta potential, must be overcome in order to cause floc- 
 culation. If the zeta potential is sufficiently low, this can be accomplished 
 by mechanical agitation and thus forcing the particles to collide. If this is 
 not sufficient, the zeta potential must be reduced. 
 
 There are two general ways in which the charge associated with a 
 sol can be reduced to the point where flocculation can take place. First, 
 ( counter ions with a higher valence than those originally present can be 
 added in sufficient concentration to substitute for the low valence ions 
 originally present and thus cause a much more rapid drop in the Stern and 
 zeta potentials. As hydrogen and hydroxl ions quite often make up a con- 
 siderable portion of the stabilizing and counter ions, a change in pH can 
 sometimes effectively lower the zeta potential. Either separately or in 
 conjunction with the first method, the potential can be reduced by adding 
 a similar concentration of colloids with the opposite charge and then the 
 particles can flocculate by mutual neutralization of charges. 
 
 The valence of the counter ions is very important in determining 
 their effectiveness in sol destabilization. It has been found that divalent 
 
18 
 ions are about one hundred times as effective as monovalent ions and tri- 
 valent ions are about one hundred times again as effective as divalent ions. 
 This effect of valence on the counter ions is called the Schulze -Hardy rule 
 after the men who first enphasized it (57) . Even all ions of the same sign 
 and valence do not all exhibit the same flocculation values. The structure 
 and especially the size of the ions dictate how easily they can fit into the 
 colloidal structure with the smaller ions being able to fit closer and more 
 effectively. Although all c611oids do not fit exactly into the same pattern, 
 a general scheme of the order of effectiveness, called the Hofmeister 
 series, has been developed. One series for the monovalent anions in 
 decreasing order of effectiveness is: F~VI0,~, ^PO^", BrO, = , CI", CIO "'. 
 Br", NO., , CIO , I", CNS". One series for cations prepared for the pre- 
 cipitation of a silver sol is: A1+++, Ba++, Sr++, Ca++, H+, Cs+, Rb+ S K+, 
 Na+, Li"*" (55, 57). It is to be noted that the same ion will not always be the 
 most effective in flocculation; the total ionic makeup of the solvent system 
 as well as the colloidal particles will dictate which ion will be the most ef- 
 fective under a given set of conditions. 
 
 In natural waters the charge on turbidity particles is predominantly 
 electronegative and is often strong enough to result in sufficient mutual, 
 repulsion to cause the formation of colloidal sols. These sols consist of 
 finely divided silt and clay, and organic matter undergoing microbial decom- 
 position (54). Riddick (59) found these electronegative colloids to have a. 
 zeta potential in the range of 15 to 25 millivolts. Shapiro (60) in studying 
 the yellow coloring matter of pond water, found it to be mainly hydroxy 
 
19 
 carboxylic acids either in solution or in colloidal suspension. These are 
 the acids which are generally called humic acids which are intermediate 
 decay products of decomposing vegetation. One importance of these col- 
 ored organic compounds is their general association with iron in water. 
 Black, in his discussion of the article by Stumm and Morgan (52) , stated 
 that "colored waters have been collected from many parts of the country 
 and without a single exception, the color has been found to be complexed 
 with iron. " 
 
 If the iron in a water is held in an organic colloidal complex, floc- 
 culation to destabilize the sol would be the way in which to remove the iron. 
 Ferric iron, either as the chloride or sulfate, and aluminum sulfate or 
 filter alum are the most common coagulants used in water treatment, so 
 excess ferric iron should be useful in flocculating organic colloids, 
 
 Longelier, et al. (61) have studied the relationship between coagu- 
 lant dosage and exchange capacity (the capacity of the counter ions to 
 exchange with ions of a higher valence) and the affect of this relationship 
 on floe formation. Normal floe formation occurs when sufficient counter 
 ions and hydrolysis products are present to destabilize the sol. A more 
 satisfactory, larger, and more rapidly settling floe is obtained when suf- 
 ficient agglomerating agent of opposite charge is added. Figure 2 is a 
 graph of the flocculation zones. 
 
 When sufficient iron is present in a solution in the ferric state, the 
 ferric hydroxo complexes can form positive, neutral, or negative sols 
 depending on the pH of the solution. This charge on the coagulant particles 
 
20 
 
 < 
 
 0_ 
 
 < 
 
 UJ 
 
 o 
 
 z 
 < 
 
 X 
 
 o 
 
 X 
 
 UJ 
 
 
 
 
 
 Zone 5 
 
 
 
 
 
 Exchange capacity of natural Col- 
 
 
 
 
 
 loids so high that stable zone for 
 
 
 
 
 / ^ 
 
 hydrous oxide sols formed by pre- 
 
 
 
 
 / £ 
 
 cipitation of metal oxide coagulant 
 
 
 
 
 / w 
 
 no longer appears. If water is low 
 
 
 
 
 / f 
 
 / o 
 
 in natural colloids, Zone 5 con- 
 
 
 
 
 
 
 dition may be attained by adding 
 
 
 4-> 
 
 <u 
 
 
 / O 6c 
 / <+* 2! 
 
 clay or other negative colloids. 
 
 
 a 
 
 
 / o o 
 
 
 
 d 
 cr 
 
 
 / o 3 
 
 
 
 <u 
 
 
 / •" 5 
 
 
 
 XI 
 
 
 / *j 3 
 
 
 
 •3 
 
 
 1 rtf *J 
 / N 3 
 
 
 
 
 / ^ £ 
 
 
 
 co 
 
 •r-< 
 
 
 / <tf <u c 
 
 
 
 0) 
 
 
 1 *J +o .£ 
 
 
 
 bJO 
 
 
 / « ftf "O 
 
 
 _ 
 
 
 
 CO 
 •rH 
 
 
 / \ Zone 4 
 
 
 T> 
 
 
 
 1 — 1 
 
 / * • o 2 > 
 
 ' \ Second zone of coag- 
 
 0) 
 
 j c 
 o 
 
 CO 
 
 a 
 
 o 
 
 CD 
 
 1 — 1 
 
 
 u 
 
 CD 
 > 
 4-> 
 
 / o o £ <tf / 
 / N £ "* c / 
 
 / * i? o / 
 
 1 So *< tf / 
 
 \ ulation. Read- 
 \ sorption of anions 
 \ by the hydrous ox- 
 
 
 
 
 
 / ° 09 / 
 
 \ ide sol probably 
 
 
 <d 
 
 / ° 3 / 
 
 \ reduces its zeta 
 
 
 • H 
 
 c 
 
 ' '-* / 
 
 
 
 
 CD i 
 
 ftf ^ / 
 
 \ potential and 
 
 
 r— 1 
 
 3 
 
 N / 
 
 • H 1 
 
 £ £ / 
 
 o ^ / 
 
 ** / 
 
 \ coagulation 
 \ occurs. 
 
 
 
 ■— 1 / 
 • H / 
 
 -° / 
 
 
 
 
 o 
 
 
 Zone 3 \ 
 
 
 
 
 CO / 
 CD / 
 X! / 
 
 / Hydrous oxide sol becomes \ 
 
 
 
 
 / stable 
 
 No mutual coagu- \ 
 
 
 
 / lation 
 
 and binding. \ 
 
 
 A 
 
 
 
 
 DOSAGE OF COAGULANT 
 
 FIGURE 2 
 
 ZONES OF FL0CCULATI0N 
 
21 
 becomes quite important in clarification by sedimentation. Riddick (54, 59, 
 
 62), Stanley (63), and Black, et al. (64, 65, 66) discuss methods in which 
 electrophoretic measurements are made of zeta potentials by measuring 
 the current of opposite direction which will just stop migration of the col- 
 loids under consideration in an electrophoresis cell. In noting the zeta poten- 
 tial of a solution and the efficiency of removal of unwanted colloids, the zeta 
 potential at optimum coagulant doses and optimum pH can be determined. 
 
 Stumm and Morgan (52), Riddick (54), Stanley (63) , and Riddick in 
 a discussion of an article by Black (66) all point out that alum and ferric 
 salts produce negative colloids in alkaline solutions. Although in natural 
 waters the coagulants can produce floe under these conditions: their zeta 
 potentials, which can be as high as -25 millivolts, hinder their effective- 
 ness in removing the negative colloids present in the water. Fair and 
 Geyer (67) summarize the practical aspects of the behavior of alum and 
 ferric iron as coagulants as follows: 
 
 1. The pH range of relative insolubility is five to seven for 
 alum and above four for ferric iron. 
 
 2. The presence or addition of negative ions extends the use- 
 ful range of pH in the acid range. For example, sulfate 
 ions become counter ions to neutralize any coagulant 
 remaining after the negative colloids have been agglomerated. 
 
 3. In soft waters, negatively charged color colloids are coagu- < 
 lated most effectively at pH values of four or less. The 
 trivalent ions of aluminum and iron are then the precipitating 
 
22 
 agents. (In practice a pH as low as four is seldom attained.) 
 
 4. Stirring increases the opportunity for contact and decreases 
 the time for floe formation. It also promotes floe growth. 
 
 5. Iron and manganese naturally present in water can be called 
 into use as coagulants and to speed their own removal. 
 
 Riddick (54, 59, 62), and Black et al. (64, 65, 66, 68) have made 
 studies to determine the effectiveness of polyelectrolyte coagulant aids in 
 further reducing zeta potentials when used in conjunction with the usual 
 coagulants. Riddick found that a zeta potential of JL5 millivolts gives opti- 
 mum turbidity removal. In order to attain this low a potential with alum 
 or ferric salts, all the natural alkalinity would be consumed and excess 
 coagulant would then remain in colloidal suspension. In order to prevent 
 this, sufficient coagulant is added to lower the zeta potential to about -10 
 millivolts, providing that at least 6 to 8 mg/1 of alkalinity still remains, 
 and then one of the new commercial coagulant aids or cationic poly electro- 
 lytes is added to lower the potential to zero. Using an electrophoresis 
 cell as an operational tool should enable operators to accomplish practi- 
 cally complete turbidity and color removal. 
 
 As this study is related to iron removal, it is more concerned with 
 the complete removal of the coagulant (ferric iron compounds) than with 
 the removal of the organic colloids except as they are associated with 
 iron. However, as can be seen from the above discussion, if the problem 
 with iron removal is one of sol stabilization of the iron in organic colloids, 
 then iron removal can only be accomplished by sol destabilization. Zaides (69) 
 
23 
 studied the interaction of iron with certain organic acids. He found that the 
 sodium salts of acetic acid and formic acid form colloidal suspensions with 
 ferric iron while sodium oxalate forms a definite anionic complex: Fe^CCoO^o, 
 This reaction of iron with sodium oxalate points to another possible source of 
 organic interference with iron removal. Iron has the ability to form stable 
 soluble complexes or chelates with certain organic compounds. 
 
 C The Reactions of Iron with Organic Chelating Agents in Water 
 
 Metal ions in aqueous solution are highly solvated with a number of 
 water molecules having the negative (oxygen) end of their dipoles directed 
 toward the positive metal. Some common configurations for this complex 
 formation are tetrahedrans, square planes, and octahedrans with the metal 
 in the center. This is an idealized condition in which the number of water 
 molecules is only a statistical average and the number, the angles involved, 
 and the bond distances are all a function of the metal involved, the tempera- 
 ture, concentration and the ionic strength. Because of the positive charge 
 on the metal ions, one or more of the water molecules can lose a hydrogen 
 ion with the result that a hydroxyl ion then becomes the ligand or complex- 
 ing agent. 
 
 Ligands, other than hydroxyl ions, can complex with metal ions 
 through several types of bonding. Sometimes the bonds are electrostatic 
 such as the attraction of a positive metal ion for a negative ion or the 
 negative part of the dipole of a neutral molecule, or the bonding can be 
 covalent with an unshared pair of electrons from the ligand helping to fill 
 
24 
 out the electron ring of the metal to that of an inert gas. In either case or 
 any intermediate system, the ligand is always the electron donor while the 
 metal is the electron acceptor. Thus it can be seen that metal ions in solu- 
 tion can affect pH, and pH can affect complex formation. 
 
 These complex or coordination compounds are called metal com- 
 plexes if each ligand enters into one bonding with the metal ion. If, how- 
 ever, a ligand has two or more donor groups and forms a ring with the 
 metal ion as a part of the ring, the resulting structure is a metal chelate 
 compound and the donor is said to be a chelating agent. Some of the chemi- 
 cal and physical properties of metal chelates resemble those of the simple 
 complexes while other properties are fundamentally different. These dif- 
 ferences set chelates apart as a separate and distinct group of chemical 
 compounds which have become important as aqueous sequestering agents 
 in chemistry, biology, and agriculture. Books by Martell and Calvin (70), 
 Chaberek and Martell (71), and one edited by Wallace (72) present a thorough 
 description of the current knowledge concerning the chelate compounds. 
 
 Many metals can react with chelating agents but only the strongly 
 non-metallic elements of Groups V and VI in the Periodic Table cf the Ele- 
 ments can serve as the donor atoms. Nitrogen, sulfur, and oxygen are the 
 only common examples. Covalent compounds with a metal bonded directly 
 to a carbon atom are called organometallic compounds and are an entirely 
 different class of compounds. There is a definite order of stability of the 
 i complexes as influenced by the metal involved, irrespective of the nature 
 of the ligands considered; for the trivalent metals in order of decreasing 
 
25 
 stability: Tl, Fe, Ba, Al, Cr , and for bivalent metal ions: Pd, Cu, Ni, 
 Pb, Co, Zn, Cd, Fe, Mn, Mg. These are not all the metals which can 
 form chelates. A complete list with their common coordination require- 
 ments can be found in the literature (70, 71). Iron in either the ferrous or 
 the ferric state commonly has a coordination number of 6 which accounts 
 for iron hydrolyzing with six molecules of water in aqueous solutions. 
 
 The properties of metal ions in aqueous solutions become altered 
 when they become chelated. Chelation decreases the tendency toward 
 olation or the building of high molecular weight metal hydroxyl complex 
 ions. Olation is not entirely eliminated as a number of olation reactions 
 have been observed with some metal chelates. 
 
 When metal ions are masked from the solvent water by the chelat- 
 ing ligands, the solubility becomes that of the chelating agent. If ionizable 
 groups a,re present making the chelating agent water soluble, the metal 
 chelate formed is also water soluble. The only apparent change in the 
 
 solution might then be a typical color formed by the chelate. Shapiro (60) 
 
 _ _ 
 
 . in his studies of lake water found that, although at pH 7 the solubility pro- 
 
 Q 
 
 duct for iron is 4 x 10, as much as 0. 03 mg/ 1 was actually in solution. 
 
 | This ability to increase the solubility of iron in solution is called seques- 
 tering and the chelating agents involved are sequestering agents. Wallace 
 
 j (72) reports that organic chelates in increasing the solubility of iron pro- 
 vide a method for vegetative plants to obtain iron from the soil as iron 
 chelates. Chelating agents such as ethylenediamine tetraacetic acid (EDTA) 
 provide metals with solubility and definite color changes to the extent that. 
 
26 
 they can be used as indicators for the quantative determinations of such 
 metals as calcium and magnesium. Chelating agents like orthophenanth.ro- 
 line produce color in proportion to the amount of metal entering into chelate 
 formation and can thus be used for the colorimetric determination of such 
 metals as iron. 
 
 The increase in electron density surrounding a chelate increases 
 the tendency of a metal to oxidize to a higher valence and $hus increases 
 the oxidation potential. Also, as these complexes are usually negative, an 
 increase in oxidation state of the metal results in a stabilizing reduction in 
 the charge of the complex. The tendency to form stable electron configu- 
 rations in the orbitals may result in either an increase or a decrease in 
 the oxidation potential. In order to complete the available orbitals, the 
 lower valence state for iron is favored, but the high negative charge on 
 most natural chelating agents, aids in maintaining iron in the ferric state. 
 The phenanthroline compounds have sufficient positive charge to reverse 
 the potential, and thus insure that iron will remain in the ferrous state 
 which is essential for proper color development in the quantative determi- 
 nation of iron. 
 
 Chelation of metal ions may also affect the rate of chemical reac- 
 tions through specific catalytic action of the chelate compound or by the 
 reduction in concentration of free metal ions to give an "unreactive" 
 chelate compound. 
 
 In the introduction it was pointed out that many instances have been 
 reported where iron is present in natural, waters complexed with organic 
 
27 
 Compounds generally classified as humic acids. Febeck (7 3) defines a 
 humic acid as "soil organic matter sufficiently degraded so that its source 
 and original structure are no longer identifiable . " He was able to account 
 for 20 percent of the organic component of a soil by using gas chromatog- 
 raphy and found oxygen containing compounds containing carboxyl groups. 
 He found that these organic compounds combine with minerals to create 
 soils rather than merely mineral aggregations. These organic compounds 
 were also found to permit biological activity, provide a source of nitrogen, 
 phosphorous, sulfur, and trace elements, and increase the tilth, aeration, 
 and water holding capacity of soil. Chelate compounds and their functions 
 can explain, at least in part, all of these phenomena. Walther (74) extrac- 
 ted humic acids from Minnesota peat and found them to be a heterogenous 
 mixture of nonspherical molecules which are probably hydrated polyelec- 
 trolytes. Their molecular weights range from 2 x 10 to 3 x 10 and they 
 
 contain a wide and continuous range of ionization constants. Scharp 
 found that the strongest complexing agents are ions of the smaller dicar- 
 boxylic and the alpha -hydroxy carboxylic acids. He further stated that 
 when two acidic groups are close enough to be mutually interacting, strong 
 chelates are formed with iron. DeKock (75) demonstrated that iron com- 
 plexed with EDTA or a water extract of peat promoted a translocation of 
 otherwise insoluble iron in plants. Muller (76) observed carboxyl and 
 hydroxy! grr»ups to be the sights of base exchange in humus and Waksman 
 and Iyer (77) found that one gram of ferric oxide could be complexed by 
 from 0.9 to 2. 79 grams of humus. 
 
28 
 Shapiro (60, 78, 7 9) in his study of the yellow coloring matter of 
 pond water found it to be hydroxy -car boxylic acids with an average molecu- 
 lar weight of 456 and an equivalent, weight of 228 as compared with soil 
 humic acids which have an average molecular weight: of 1400. The lakes he 
 studied contained between 4. 5 and 9. mg/1 of humic acids, which he con- 
 cluded were not chelates because they were not. dialyzable and would not 
 pass through a 0. 10|j, filter. This does not necessarily eliminate the pos- 
 sibility that chelates are formed and then are hydrolyzed into larger parti- 
 cles with colloidal properties. Adams (34) found hydroxy -polybasic acids 
 such as citric, tartaric, and malic acid to interfere with iron sedimentation. 
 
 Chaberek and Martell (71) indicate that the most important classes 
 of sequestering agents are those which can form five- and six-membered 
 chelate rings. Four-membered rings are rather rare. Rings containing 
 more than six links are generally unstable. It can thus be seen that the 
 alpha -hydroxy carboxylic acids can form stable five -member ed ring che- 
 late compounds such as: 
 
 H 
 
 ! 
 
 R -C - C = 
 
 i i 
 
 °\ -° 
 
 Te 
 
 In this type of rea,ction the iron replaces the proton from the carboxyl 
 group and thus lowers the pH of the solution. Conversely, in very acid 
 solutions, chelate formation would be greatly hindered. 
 
 The foregoing discussion points to three possible sources of 
 
29 
 
 interference with iron removal bv aeration, sedimentation and filtration. 
 They are: 
 
 1. The formation of soluble hydroxo polymers of iron. 
 
 2. The formation of colloidal dispersions with organic 
 colloids. 
 
 3. The formation of soluble chelates with organic seques- 
 tering agents. 
 

 30 
 III. EXPERIMENTAL EQUIPMENT AND PROCEDURES 
 
 A. Isolation of Organic Matter from Ground Water 
 
 Several methods have been successfully used to remove and isolate 
 organic matter from natural waters in order to make chemical character- 
 izations and quantitative measurements of the concentrations of organic 
 matter present. The method in most general use is to pass the water to 
 be analyzed through an activated carbon filter. The organic matter, which 
 is adsorbed on the carbon, is then extracted with suitable solvents. After 
 the solvents are driven off by evaporation, semi -quantitative measurements 
 can be made of the organic residue and the extracts can be used for further 
 characterization. Middleton, et al. (80, 81, 82) at the Robert A. Taft 
 Sanitary Engineering Center have done considerable work with carbon 
 filters. They found that a standard carbon filter (which will be described 
 later) removed 63 percent of the organic matter from filtered city water 
 at Cincinnati, Ohio and 82 percent of the organic matter from raw water 
 from Nitro, West Virginia. Various solvents have been used to extract 
 the organic matter from the carbon. Middleton, et al. (81) found that 
 extraction with chloroform followed by extraction with ethanol obtained 
 the greatest recovery of organic matter adsorbed from river water. 
 
 As the carbon filter technique does not. recover 100 percent of the 
 organic matter present in a water, it does not give absolute quantitative 
 results. However, it can be used to establish a base concentration of 
 organic matter present to use as a reference and then for monitoring 
 
31 
 organic pollution and changes in concentration. In order to obtain higher 
 yields of organic extract, Myrick and Ryckman (83), SprouL and Ryckman 
 (84), Atkins and Tomlinson (85), and Spicher and Skrinde (86) have used 
 filters containing from 1. 3 to 2. cu. ft. of carbon instead of the 0.074 
 cu. ft. in the standard filter; however, they were designed for the same 
 contact time. As adsorption efficiency is higher for waters of low turbid- 
 ity and as the water used in their studies came from the Missouri River 
 at St. Louis, Missouri, the water was presettled and filtered through a 
 diatomite filter before it was applied to the carbon filters. Sproul and 
 Ryckman (84) reported that the adsorption efficiency drops off above pH 8.5. 
 Therefore, they maintained the pH between 7. and 7. 5 for maximum adsorp- 
 tion. Hyndshaw, et al. (87) used activated carbon to remove odors from a 
 water supply and found that the efficiency was only 10 percent at pH 10. 6 
 and increased to 70 percent at pH 9° 6.. Spicher and Skrinde (86) found 
 that, if the effluent from a carbon filter was acidified with sulfuric acid 
 to pH 3. 2, at least: twice as much organic matter was obtained as with the 
 first filter alone. Middleton (80) also. made this same observation. The 
 Water Research Association, at Buckinghamshire, England (88) obtained 
 I a high recovery of colored organics from some lakes by using a combina- 
 tion of activated carbon mixed with an anion exchange resin. The extracts 
 
 [were eluted from the filter with sodium chloride, but the sodium chloride 
 
 i 
 
 | was difficult to remove from the extracts. 
 
 Bikerman (89), in discussing the adsorption of organic liquids on 
 
 activated carbon, has stated that the relative size of pores on one hand 
 
32 
 and of the molecules of solvent and solute on the other influences the acces- 
 sible surface area of the carbon. Carbon is activated by heating at high 
 temperatures in an atmosphere o£ carbon dioxide. All types of activated 
 carbon adsorb uncharged molecules, but, when significant amounts of ash 
 remain, ion exchange reactions take place. With very little ash present, 
 adsorption of charged ions is by hydrolytic or electrochemical action. 
 Adsorption is a process at the carbon-liquid interface and usually is so 
 rapid that the efficiency is not greatly dependent on the rate of application. 
 
 Shapiro (60, 78, 79, 90), has made extensive studies of the yellow 
 coloring matter in several lakes in New England. His usual process of 
 concentration is a freezing-out technique. A sample to be concentrated 
 is continuously stirred while being frozen. As the water freezes, the 
 central, unfrozen core becomes more and more concentrated containing 
 about 99 percent of the impurities in the water. A twenty to one concen- 
 tration has been made in this manner. The concentrate is applied to a 
 hydrogen ion exchange resin after having been filtered through a 0. 45u 
 membrane filter. The organics, freed of their metal cations, are eluted 
 with n-butanol and dried under a vacuum. 
 
 Shapiro (60) and Mueller, et al. (91, 92) have used vacuum distil- 
 lation to concentrate the organic matter in surface waters. Shapiro extrac- 
 ted the organic fraction with, a mixture of ethanol and hydrochloric acid, 
 redissolved in hydrochloric, acid and then re-extracted with ethyl acetate. 
 Mueller and his co-workers extracted the organic fraction with ethyl ether 
 and then used silicic, acid column chromatography to separate the organic. 
 
33 
 matter into distinct fractions. 
 
 In choosing a method of extracting the organic matter from well 
 water for this study, it was desired to find one which would be relatively 
 quantitative for the organic matter present and still be free of iron in 
 order to permit a study of the interaction of the organic matter obtained 
 with iron. It was decided to first try variations of the carbon filter method 
 since it had been so widely used with apparent success. The vacuum dis- 
 tillation method and the freezing=out method tend to limit recovery to 
 organic compounds which are soluble in solvents which can be used in 
 liquid-liquid separations. Also s as the concentrates would necessarily 
 contain a high concentration of inorganic material, the possibility of 
 considerable inorganic carryover during solvent extraction seemed 
 highly possible. It was hoped that, with the lower concentrations during 
 adsorption on activated carbon, there would be a greater selectivity for 
 organic matter which would eliminate the necessity of using a cation 
 exchange resin to remove metal ions from the extracts. 
 
 B. Experimental Technique for Isolation of Organic Matter from Ground 
 Water 
 
 In order to obtain more complete adsorption of the organic matter 
 i in the water, two carbon filters were connected in series with sulfuric 
 i acid injected after the first, filter. The pH of the water, as it was applied 
 i to the second filter, was maintained between 2. 5 and 3. 0. A check on the 
 amount of organic matter passing the acidified filter was made by adding 
 
34 
 a third filter into the series during two of the experiments. Figure 3 is a 
 diagram of the filter arrangement as it was used in the field. The system 
 consisted of: 
 
 1. A flow regulator valve. 
 
 2. The raw water sampling tap. 
 
 3. Rubber hose connecting the flow regulator valve with the 
 first filter. 
 
 4. The unacidified filter hereafter called filter A. 
 
 5. Filter A effluent sampling tap. 
 
 6. Hose connecting filter A with the water meter. 
 
 7. A 3/4 inch water meter used to meter the flow. 
 
 8. Hose connecting the water meter with the acid injection tee. 
 
 9. The acid injection tee. 
 
 a. Five gallon plastic bottle containing approximately 
 0. 8 N sulfuric acid. 
 
 b. Siphon tube from acid bottle to Sigma* pump. 
 
 c. Sigma pump connected through a low ratio pulley 
 drive to electric motor. 
 
 d. Rubber hose connecting Sigma, pump to the acid 
 injection tee. 
 
 10. Hose connecting tee with the first acidified filter, 
 
 11. First acidified filter hereafter called filter B. 
 
 # 
 
 The Sigma pump is manufactured by Sigmamotor, Middleport, New York. 
 
35 
 
 4) 
 
 :- m 
 
 as 2 <? 
 
 W *- -C Z (/) 
 
 S o» « (M 
 
 a. oo * oi 
 
 7^ 
 
 UJ 
 UJ 
 
 e> 
 
 < 
 
 =2 < 
 o 
 
 u. 
 
 (T 
 LU 
 
 u 2 
 
 a> o 
 
 UJ 5 
 
 n 
 
 o o> 
 £ 2 
 
 J 
 
 =d 
 
36 
 
 12. Filter B effluent sampling tap. 
 
 13. Hose connecting filter B with the second acidified filter. 
 
 14. Second acidified filter hereafter called filter C. 
 
 15. Effluent hose. When filter C was not used, the hose 
 normally connecting filter B with filter C was used as 
 the effluent hose. 
 
 All hoses and valves were 3/4 inch except those leading from the acid 
 bottle to the acid injection tee, 
 
 Filter A was a standard glass filter, as described by Middleton, 
 et al. (81), made up mainly of commercially available parts. It is detailed 
 in Figure 4. Filters B and C were fabricated locally from lucite plastic. 
 They are detailed in Figure 5. Figure 6 shows the sampling equipment 
 in operation in the field. 
 
 The filters used were all charged with a 400cc layer of dry Cliffchar* 
 . of 4 to 10 mesh size, and then filled with approximately 1700cc of 30 mesh 
 dry Nuchar C-190* , The Cliff char served as a coarse filter for removing 
 any turbidity which might otherwise clog the filters. The carbon was 
 extracted with chloroform and ethanol prior to use to insure that there 
 were no organic contaminates present in the activated carbon which could 
 give a false indication of the presence of organic matter in the well waters 
 tested. The extraction procedure was the same as for extraction of the 
 filters after use. This procedure will be discussed later. 
 
 * Both Cliff char and Nuchar are products of West Virginia Pulp and Paper 
 Company. 
 
37 
 
 3/4"x2" 
 
 Galvanized 
 
 Nipple 
 
 3/8"* 2" 
 Bolts and 
 Nuts 
 (Not Shown) 
 
 Pyrex Glass 
 
 Pipe 
 
 3"l.D.xl8" 
 
 End Assem- 
 bly Same 
 as Above 
 
 V 
 
 6 " D . x ,/ 2 "_ 
 
 Aluminum Plat 
 
 Locally 
 
 Fabricated 
 
 Neoprene Gasket 
 l/4 M ThicK 
 with 3 M Hole 
 Slotted 3/8 
 
 for Screen 
 
 Flange to Fit 
 3 Pipe 
 
 Topped for 
 3/4 M Nipple 
 
 Stainless Steel 
 Screen 40 Mesh 
 3/4" D. 
 
 Asbestos Insert 
 
 FIGURE 4 
 COMMERCIAL GLASS FILTER 
 
38 
 
 3/4"0.D.x3 
 Plastic Pipe 
 
 |/4"D.x2l 
 Threaded 
 Brass Rods 
 and Nuts 
 
 Plastic Pipe 
 3" I.D. x 18' 
 
 End Assem< 
 bly Same 
 as Above- 
 
 Brass Screen 
 
 6 D.x 1/2" Alignment 
 Plate 
 
 Note: AH Plastic is Lucite 
 
 FIGURE 5 
 
 LOCALLY FABRICATED PLASTIC FILTER 
 
39 
 
 ■ 
 
 ..^■f. 
 
 %^H>^2?' 
 
 ■""' 
 
 ,<£& 
 
 
 FIGURE 6 
 CARBON FILTERS IN OPERATION 
 
40 
 In early experiments, a sand filter was connected directly to the 
 raw water supply, and before the water meter. This arrangement was used 
 so as to protect the water meter from becoming jammed by either sand com- 
 ing from the aquifer with the raw water or by carbon dust from the filter. 
 Even though the effluent from the sand filter was refiltered through a glass 
 wool plug, fine sand passed through on every attempt to use this system. 
 A more satisfactory arrangement proved to be the one shown in Figure 3. 
 It was successfully used by passing about five gallons of water through 
 filter A and discharging it through the filter A effluent sample tap before 
 the remainder of the system was connected. The water meter never 
 failed when this procedure was followed. 
 
 The water meter was a 3/4 inch service meter provided by the 
 Northern Illinois Water Corporation, Urbana, Illinois.- The meter was 
 tested in the Corporation's meter shop and found to be 99 percent accurate 
 it flows as low as 0. 25 gpm. 
 
 The rate of flow was regulated by observing the time required to 
 :ollect one gallon of effluent from the system. As there were no auto- 
 matic flow regulators employed in the equipment and as there was no emer- 
 gency shutoff provided in case of a failure of the acid injection system, 
 requent attention was required. For this reason, it was decided to pass 
 he water through the filters at the rate of one gpm if at all possible instead 
 i>f the 0.25 to 0. 50 gpm rate recommended by Middleton, et al. (81). With 
 (his procedure, 4000 to 5000 gallons of water could be extracted in a 24 
 tour period. Repeated experiments were made at Clinton and Champaign 
 
41 
 which provided some comparison of filter performance at different rates. 
 Initial experiments were in both cases made at low rates of about 0. 5 gpm 
 and later experiments were made at higher rates of approximately 1. gpm, 
 Although the quantities of extracts obtained were not identical at the two 
 rates, they were of the same order of magnitude. As the rate chosen for 
 applying water to the carbon filters was higher than the rate normally used, 
 it was expected that the percentage of the organic matter in the water obtained 
 by this method would be slightly lower than the percentage which would be 
 obtained at a lower rate. However, from the results of the above men- 
 tioned repeated experiments and the quantities of organic extracts obtained 
 in all the experiments, it is felt that the adsorption efficiency at the high 
 rate is sufficiently high to provide data which could prove useful for compar- 
 ison of concentrations of organic matter in various waters. 
 
 After the carbon had adsorbed organic matter from a water, it was 
 removed from the filters, placed in trays, spread in a layer about two inches 
 thick, and dried two or three days in a constant temperature room at 95°F. 
 The dry carbon was then charged into large capacity Soxhlet type all glass 
 extractor s-''. The bottom plates were removed and the bottoms of the extrac- 
 tors were packed with glass wool, which had previously been extracted with 
 chloroform and ethanol, to prevent small particles of carbon from carry- 
 ing over into the extracts. One extractor was just sufficient to handle 
 the carbon from one filter. 
 
 *E. H. Sargent and Company Catalog No. S-31400 
 
42 
 
 Solvent was added through the carbon and allowed to siphon into the 
 heating flasks. Two siphoning cycles were usually sufficient to add the 
 2500 mis of solvent needed. In addition to rapidly wetting the carbon to 
 prevent excess heat buildup through hydrolysis, two extraction cycles were 
 gained with this procedure. Three or four Hengar granules were added to 
 each heating flask to insure smooth boiling. The flasks ordinarily used 
 with the large Soxhlet extractors are three liter round bottom flasks with 
 29/42 ground glass neck connections. In this work, however, three liter 
 flasks with three 24/40 necks were used. The University of Illinois glass 
 blower removed the center necks from the flasks and replaced them with 
 the required 29/42 necks. The side openings were used for addition of 
 solvent when necessary, the monitoring of pH, and measurement of the 
 boiling temperature of the solvent. The flasks were heated with three 
 liter electric heating mantles. Power was regulated with a 115 volt vari- 
 able power transformer. Because the extractors were left unattended for 
 considerable periods of time, they were used in a laboratory hood with the 
 exhaust fan running. Figure 7 shows two extractors in operation. 
 
 Some difficulties were observed in the use of the Soxhlet extractors. 
 When the voltage was set too low and the boiling rate was too slow, solvent 
 was able to drip past the top of the siphon without forcing the siphoning pro- 
 :edure to start. This, of course, interfered with the complete extraction 
 pycle and reduced the extraction efficiency. When the glass wool was packed 
 r-oo loosely, fine carbon passed into the solvent flask; and when it was packed 
 |oo tightly, the increased head loss reduced the siphoning rate to such an 
 
43 
 
 FIGURE 7 
 SOXHLET EXTRACTORS IN OPERATION 
 
44 
 extent that volatilization was able to match the siphoning rate and the siphon 
 operated continuously without allowing the extraction chamber to ever become 
 filled with solvent. Ethanol was usually more difficult to use than chloro- 
 form, probably because of its greater viscosity. For most trouble-free 
 operation, it was found that a setting of 90 volts provided the best rate when 
 chloroform was the solvent being used and 100 to 115 volts was the optimum 
 for ethanol. When carbon carried over into the solvent, it was necessary 
 to filter the solvent before further concentration of the organic extract. 
 It was found that vacuum filtration through Whatman No. 40 filter paper 
 followed with several washings with the solvent being used provided satis- 
 factory removal of the carbon. 
 
 Middleton, et al. (81) studied the extraction efficiency on Cincinnati 
 finished water. With chloroform as a solvent and a cycle period of approx- 
 imately one hour, the following results were obtained: 
 
 Number of Cycles Percent of total amount 
 
 of extract recovered 
 
 1st 5 48. 3 
 
 2nd 5 26. 5 
 
 3rd 7 17.6 
 
 4th 16 2. 
 
 5th 35 3. 6 
 
 6th 60 1. 9 
 i 
 jit was further reported that chloroform followed by ethanol gave about the 
 
 largest recovery of extract. 
 
 In this study with the voltages noted above, the length of the extrac- 
 tion cycles were from 45 minutes to one hour so long as the siphon func- 
 tioned properly. A solvent was cycled for about 30 hours providing about 
 
45 
 
 30 to 35 cycles. This should have accomplished about 95 percent removal. 
 When the completeness of the extractions were determined by recycling a 
 carbon sample for 30 hours with fresh solvent, no appreciable additional 
 extract was obtained. 
 
 For most waters, an aqueous layer formed on top of the chloroform 
 solvent even though the carbon had apparently been completely dried. This 
 water layer rapidly changed in color from yellow to orange to dark brown 
 as the cycles progressed. The chloroform layer gradually changed from 
 yellow to orange. After the extraction period had been completed, the 
 aqueous and chloroform layers were separated with a separatory funnel 
 and retained and studied as separate fractions. The aqueous extracts for 
 each experiment will be called: 
 
 "Water A" (from the unacidified filter) 
 
 "Water B" (from the first acidified filter) 
 
 "Water C" (from the second acidified filter). 
 Correspondingly, the chloroform extracts will be called "Chloroform A, " 
 "Chloroform B, " and "Chloroform C. " 
 
 Upon completion of the chloroform extractions, the extractors were 
 purged of residual chloroform by attaching an air line to the siphon tube and 
 massing air through the carbon for about four hours. The carbon was then 
 extracted for about 30 hours with 95 percent ethanol. In most instances 
 he color of the solvent, darkened rapidly to a very dark brown. The ex- 
 racts will be designated as above with the extract from the unacidified 
 lilter being called "ethanol A," etc. 
 
46 
 Since pH has been reported to affect adsorption on activated carbon, 
 it was decided to try extractions with ethanol at various pH values. After 
 the ethanol extractions were completed, fresh ethanol was added followed 
 by concentrated ammonium hydroxide. Sufficient ammonium hydroxide was 
 added to raise the pH to at least nine. It soon became apparent that addi- 
 tional extract was being obtained as the same dark brown color appeared. 
 The extract from the unacidified filter will be called "ammonia A," etc. 
 
 After this extraction procedure was completed, the Soxhlet was 
 purged with air for 24 hours in an attempt to remove the ammonia. The 
 extractor was reassembled, fresh ethanol was added and then sufficient 
 hydrochloric acid was added to lower the pH to less than two. The color 
 change in the solvent indicated that considerable extract was being obtained, 
 but in lesser amounts than with the other ethanol solvents. The extract 
 from the unacidified filter will be called "HC1 A, " etc. As not all the 
 ammonia had been removed from the carbon before the hydrochloric acid 
 was added, large amounts of ammonium chloride precipitated from solu- 
 tion as concentration was attempted. The extracts were concentrated, fil- 
 tered, diluted, reconcentrated, and refiltered but so much ammonium 
 chloride remained that it was impossible to tell how much extract was 
 present so the hydrochloric acid extracts were not included in the final 
 calculations of organic concentrations in the well waters studied. A sum- 
 mary of the extraction scheme is shown in Figure 8. 
 
 After the solvent-extract mixtures were obtained, the solvents were 
 idriven off by distillation. The chloroform extracts were distilled at 58 to 
 
47 
 
 
 u 
 
 
 
 
 
 u 
 
 a 
 
 u 
 
 U 
 
 u 
 
 
 Pi 
 w 
 
 
 
 <u 
 
 i — i 
 
 
 
 d 
 
 •iH 
 
 u 
 
 1 — 1 
 
 H 
 i— ( 
 
 
 
 o 
 
 . — 1 
 
 A 
 
 u 
 
 cq 
 
 
 nJ 
 
 w 
 
 o 
 
 u 
 
 X 
 
 CQ 
 
 £ 
 
 cq 
 
 pq 
 
 fq 
 
 
 W 
 
 In 
 
 
 
 
 t—i 
 
 
 
 d 
 
 TO 
 
 d 
 
 CQ 
 •—I 
 
 H 
 
 
 
 4-> 
 
 ro 
 
 
 
 u 
 
 i— i 
 
 u 
 
 
 
 I— 1 
 
 u 
 
 
 ■M 
 
 £ 
 £ 
 
 X 
 
 
 < 
 
 
 
 
 
 <J 
 
 g 
 
 «j 
 
 < 
 
 <J 
 
 
 
 u 
 
 
 
 
 i— 1 
 
 
 
 d 
 
 TO 
 
 •r-l 
 
 < 
 
 1 — 1 
 
 H 
 
 
 
 ■M 
 
 to 
 
 
 
 £ 
 
 (J 
 
 
 u 
 
 
 
 1— 1 
 A 
 
 TO 
 
 A 
 +-> 
 
 W 
 
 HI 
 
 
 o 
 
 
 
 TO 
 • r-i 
 ti 
 O 
 
 £ 
 
 1— 1 
 
 
 £ 
 
 
 
 <j 
 
 H! | 
 
 
 
 
 
 
 + 
 
 + 
 
 
 *4-l 
 
 /^ 
 
 ^H 
 
 <—* 
 
 ^H 
 
 H 
 
 
 
 M 
 
 
 
 
 
 
 
 > 
 
 M 
 
 0) 
 
 a 
 
 d 
 
 d 
 
 
 
 
 
 +-> 
 
 TO 
 
 TO 
 
 ro 
 
 A 
 
 to 
 
 43 
 
 X! 
 ■(-> 
 
 A 
 
 •M 
 
 ■w 
 
 U 
 
 w 
 
 w 
 
 w 
 
 UJ 
 
 
 LU 
 
 
 X 
 
 
 cj 
 
 
 CO 
 
 
 2 
 
 
 o 
 
 /"O 
 
 h- 
 
 UJ 
 
 o 
 
 
 < 
 
 LU 
 
 a: 
 
 a: 
 
 h- 
 
 3 
 
 X 
 
 (D 
 
 LU 
 
 a: 
 < 
 
 3 
 CO 
 
48 
 62°C. and when the temperature started to rise, voltage on the heating man- 
 tle was reduced and the temperature was allowed to rise to 100°C. in order 
 to distill off any residual water. As soon as the temperature started to 
 rise above 100°C. , the mixtures were transferred to tared flasks. The 
 flasks were placed on a steam table which was swept by a stream of air 
 from an exhaust fan. When the extracts became solidified,, they were dried 
 in a dessicator and weighed. This weight was taken as the total weight of 
 the subject extract. With the ethanol and pH adjusted ethanol extracts, 
 the only difference was that the original distillation temperature was 78. 5°C. 
 It was realized that any organic materials with low boiling points would be 
 lost with this procedure. However, considering the large quantities recov- 
 ered by the process used, the proportion of the extracts lost through vola- 
 tilization is probably small and thus would not be expected to add greatly 
 to the total amount of iron which could apparently be held in a filterable 
 condition. 
 
 The process of freeze-drying with liquid nitrogen under a partial 
 vacuum was tried, but owing to the large volumes to be handled and the 
 apparent success with the distillation method, this method was not pursued. 
 
 The solvents used were all reagent grade, except for the 95 percent 
 ethanol, which was technical grade received in bulk. The ethanol was redis- 
 tilled before use and only that portion which came over between 78 and 80°C. 
 was used. The solvents given off during concentration of the chloroform 
 and ethanol extracts were collected and redistilled again for reuse. The 
 iethanol which was collected for reuse distilled at between 78 and 80°C. and 
 
49 
 the chloroform between 58 and 61 C. Ethanol which contained either ammo- 
 nium hydroxide or hydrochloric acid was discarded as contaminated. 
 
 The carbon from one filter was divided in half and one half was 
 extracted in the usual manner. The other half was extracted with dimethyl 
 formamide. At the time of this test, it was not known that the ethanol plus 
 hydrochloric acid extract was practically all ammonium chloride; there- 
 fore, the volume of extract obtained. was compared with all the extracts 
 obtained in the original procedure. Using this comparison, the dimethyl 
 formamide extraction was not as effective as the solvent series, but, when 
 the ethanol plus hydrochloric acid extract was excluded,, it was found to be 
 more effective. As more samples had been extracted using the solvent 
 series before this discrepancy was noticed, the solvent series was used 
 in all subsequent tests. The solvent series did offer the added advantage 
 of providing some division of the extracts according to solubility. 
 
 C. Interference with Iron Removal by Organic Extracts. 
 
 As chelate formation and/or formation of colloidal dispersions 
 
 were thought to be the most likely sources of interference with iron removal, 
 
 it was these properties of the organic extracts which were to be determined. 
 
 [f the fractions causing interference had not been recovered, or if the extracts 
 aad been altered in the extraction process, measurements of the sought-after 
 
 Properties would be in error. However, the most probable reaction would 
 >e esterification which would have had the effect of blocking chelation sites 
 ind final results would indicate a lesser interference with iron removal. 
 
50 
 Probably the only way in which high results could have been produced 
 would have been through the formation of compounds which had an increased 
 tendency to form colloidal dispersions. Therefore, any positive results 
 obtained could probably be considered significant. 
 
 Pari and Sarup (93), Beckwith (94), and Khanna and Stevenson (95) 
 used titration methods to measure the chelation capacity of humic acids in 
 soils. The organic acids were titrated with a strong base with and without 
 the addition of alkali metals. In some instances the metals displaced pro- 
 tons from the carboxyl groups, which caused the end point of the titration 
 curves to be displaced. Beckwith pointed out, however, that not all acids 
 reacted the same way. The end point was not displaced at all with some 
 hydroxy carboxylic acids, in some instances hydroxyl groups gave up pro- 
 tons, and it was even possible for some metals to be present in the acids 
 tested which blocked the reaction sites. 
 
 Khanna and Stevenson (95) measured chelation capacity by running 
 titration curves with ever increasing amounts of the metal under consider- 
 ation. The end point of the curves were displaced more and more until the 
 chelation capacity was exceeded. At this point the curves were observed 
 to form a horizontal shoulder, in proportion to the excess metal added, 
 because the uncomplexed metal was consuming the hydroxide in forming 
 metal hydroxide. This prevented the pH from rising until, all the uncom- 
 Dlexed metal had reacted. This method of measuring chelation capacity 
 was not always found to be satisfactory 3 because not all organic, acids 
 reacted the same way, and hydrolysis of some metals is too involved to 
 
51 
 
 produce the simple reactions necessary for the curves to be meaningful. 
 This procedure was attempted with two of the extracts obtained in the cur- 
 rent study, but the addition of iron to the extracts merely produced progres- 
 sively displaced titration curves with carbon dioxide free sodium hydroxide. 
 The displacements were such that it was impossible to tell where chelation 
 was exceeded and iron hydrolysis began. 
 
 Meites (96) and Kolthoff and Lingane (97) have prepared texts dis- 
 cussing the uses of the electrochemical method of polarography. Electro- 
 lysis or the occurrence of chemical reactions at electrodes immersed in 
 solutions is characterized by the transfer of electrons between the electrode 
 and substances in solution. This unbalance of electrons is counteracted by 
 the migration of positive and negative ions in the solution. This flow of 
 current is measured by the flow of electrons through the external portion 
 of the circuit connecting the two electrodes. If the current caused by the 
 electrode reactions is spontaneous, the system is a galvanic cell. If the 
 reactions are forced to occur by an externally applied voltage, the system 
 is an electrolysis cell. The cathode is the electrode at which electrons 
 are transferred from the electrode to substances in solution causing reduc- 
 tion of the substance in solution. The anode is the electrode at which some 
 of the electrolyte is oxidized. As the voltage is increased, the current 
 flow increases. Most elements can be oxidized or reduced if sufficient 
 voltage is applied. The minimum voltage at which an element will react 
 is typical for given elements and compounds. The greater the concentra- 
 tion of the element present, the greater will be the change in the current flow. 
 
52 
 
 Thus a graph of current flow at a whole range of voltages can be used to 
 qualitatively and quantitatively identify a number of substances in a solution. 
 
 For negative potentials, a dropping mercury electrode is used with 
 a saturated calomel reference electrode. However, at even slightly nega- 
 tive potentials, oxygen is reduced, producing a current flow which can mask 
 a current flow caused by the ions under consideration. Elaborate measures 
 must be taken to purge all oxygen from the system and keep it out during 
 the analysis. At very positive voltages, mercury tends to dissolve from 
 the anode; therefore, a rotating platinum microelectrode can be used for 
 positive voltages. The platinum electrode cannot be used for negative 
 voltages as hydrogen deposition begins at very low potentials. 
 
 As metals formed in chelate compounds have oxidation potentials 
 different than the metal ions, polarography offers a method for measuring 
 chelation capacity. Ferrous ion is oxidized to ferric ion at a voltage less 
 positive than is ferrous iron complexed with organic chelate compounds. 
 If a potential intermediate between the two values is set, no current will 
 flow as long as ferrous iron added is chelated. As soon as excess ferrous 
 ion is added, a current would be noted indicating the chelation capacity of 
 the quantity of organic material originally added. This procedure using 
 the rotating platinum electrode is called an amperometric titration. Unfor- 
 tunately it does not work readily for the iron system. If the pH is high 
 enough for the organic material to enter into chelation reactions, excess 
 iron will, upon oxidation to the ferric state, be precipitated as ferric hydrox- 
 ide instead of providing ions in solution which would produce an increase in 
 
53 
 current. The reduction of ferric iron to ferrous iron can be measured 
 using a dropping mercury electrode and more negative voltages. Although 
 free ferrous iron is more soluble than ferric iron, the reaction is still 
 quite pH dependent and changes in current flowing are too readily masked 
 by any oxygen which might enter the reaction flask. 
 
 Finally it was decided to measure the ability of the organic extracts 
 to hold iron in a filterable condition. This technique has been used by 
 Shapiro (60, 78, 79, 90) in his study of the yellow organic coloring matter 
 in pond waters. Increasing amounts of ferric iron as ferric chloride were 
 added to a water of a controlled pH and then filtered through membrane 
 filters. Iron was added until ferric hydroxide was observed to precipitate 
 on the filter papers. 
 
 In the present study the extracts were dissolved in a solvent con- 
 sisting of 50 percent 0. 2 N potassium chloride and 50 percent dioxane. 
 The potassium chloride was added to the extracts followed by the dioxane 
 except, when chloroform extracts were being dissolved, the order of sol- 
 vent addition was reversed. Dioxane, or diethylene dioxide, was used as 
 a solvent aid as it is infinitely soluble in water, alcohol, ether, and most 
 other organic liquids, and many substances which are practically insoluble 
 in water are soluble in dioxane. This extreme range of solubility greatly 
 assisted in putting the extracts in solution. 
 
 All solutions were prepared in the concentration of 1. mg organic 
 )er ml. A, few of the extracts, especially the chloroform extracts, did 
 liot dissolve completely at this time. However, all of the brown color left 
 
54 
 the particles and went into solution. All solutions were brown in color. Ten 
 mis of each solution was diluted with distilled water to two liters, providing 
 a working solution with an organic concentration of 5 mg/1. As will be seen 
 later, this is in the range of concentrations of the organics in ground water. 
 Of each of these solutions, 1500 mis was used as the test solution and the 
 other 500 mis was retained to use as make-up solution after a portion of 
 the original test solution had been removed to test for filterability. The 
 solutions were stirred constantly with a magnetic stirrer and the pH was 
 monitored continuously. The initial pH of the solutions were measured 
 and then the pH was raised to 8. with dilute sodium hydroxide. At this 
 pH, the extracts were all in solution. Pari and Sarup (93) noted that humic 
 acids precipitate at low pH values, approximately 3. 5, and alkali metal 
 humates are formed at about pH 7.5. 
 
 In this concentration the solutions were colorless. Iron was usually 
 added as ferrous sulfate from a stock solution which contained 1. mg/ml 
 oi *on neid in acid soiution with suUuric acid. After the iron was added, 
 the pH was quickly raised again to 8. Q. Usually the first addition of iron 
 was 0.4 to 0. 5 mg/l and almost immediately the solutions began to turn 
 yellow or brown. The solutions were allowed to react approximately five 
 ! Tiinutes before samples were taken; however, at no time, after the ferrous 
 .r on was added and mixed, was any ferrous iron ever found in the reaction 
 
 )eaker. After the reaction times were completed, 35 ml samples were 
 
 ... 
 
 lltered through 0. 2 Z M- cellulose nitrate membrane filter papers 5 !'. The 
 
 ; !c Millipore Filter Corporation, Bedford, Massachusetts 
 
55 
 filter papers used were 7/8 inches in diameter hand cut with a steel die to 
 fit a stainless steel filter. An aspirator was used to provide vacuum for 
 the filtration. The filter papers were examined for a yellowish precipitate 
 of ferric hydroxide and the filtrates were analyzed for total iron concentra- 
 tion. After this step was completed, more extract solution with a concentra^ 
 tion of 5. mg/l was added to the reaction beaker to bring the volume back 
 up to 1500 mis. More iron was added and the pH was again adjusted to 8. 0. 
 Again 35 ml samples were filtered and analyzed as before. Increasing 
 amounts of iron were added even after iron began to appear on the filter 
 papers. The tests were stopped when iron precipitate became so heavy 
 that filtration became difficult. As a cross check and always when iron 
 appeared on the filter papers, total iron determinations were made on 
 anfiltered samples. 
 
 In order to obtain larger samples for iron analysis when most of 
 the iron was being retained on the 7/8 inch membrane filter, 7 5 ml sam- 
 ples were filtered through two inch diameter membrane filters of the same 
 pore diameter. To test for the apparent diameter of the floe particles, 
 0. 45p. membrane filters were used to compare filterability of the two sizes. 
 The pH of the solutions were changed, when iron just began to precipitate, 
 !to determine the pH dependance of filterability. Samples were allowed to 
 stand for periods up to 24 hours to see if flocculation was time dependent. 
 Ferric nitrate was used as the iron source in some instances to see if the 
 
 | : 
 
 oxidation state, of the iron at the time it first contacted the organic matter, 
 jwould cause any change. Another variable determined was the use of water 
 
56 
 instead of the usual initial solvent of dioxane and potassium chloride. 
 
 Tartaric acid, which is a dihydroxy-dicarboxylic acid, was used as 
 a reference of known composition. Adams (34) had reported that tartaric 
 acid formed complexes with iron and its structure provided ideal sites for 
 the formation of five-membered chelate rings with iron. If one end formed 
 a chelate with iron, the other was available to ionize and insure the solu- 
 bility of the chelate in water. The capacity of tartaric acid to react with 
 iron was compared with the extracts throughout the research. 
 
 D. Field Testing of the Organic Extracts 
 
 Organic extracts, which demonstrated ability to hold iron in a fil- 
 terable condition even in the ferric state, were added to well waters to 
 determine if the same interference could be produced with natural waters. 
 Tests were conducted at two cities where organic extracts from the water 
 supplies had shown a definite ability to maintain iron in a filterable condi- 
 tion. Experiments were also conducted at a city where little organic matter 
 was found in the well water. This water also did not present any problems 
 with iron removal. These tests were conducted as follows: 
 
 1. A covered nine liter battery jar was completely filled 
 with water, through a submerged tube, directly from 
 the well under consideration. 
 
 2. The following determinations were then made: 
 
 a. Dissolved oxygen (which was always practically nil). 
 
 b. Ferrous iron. 
 
57 
 
 c. Total iron. 
 
 d. pH (taken with electrodes inserted in rubber stoppers 
 in the cover of the jar). 
 
 e. Temperature. 
 
 3. The jar was drained to a point where it contained eight 
 liters after a measured quantity of organic extract was 
 added. 
 
 4. The sample was aerated for two minutes with 2000 cubic 
 centimeters of air per minute using a portable air compres- 
 sor and a carborundum diffuser. 
 
 5. The following tests were then made at measured intervals 
 until no ferrous iron remained in the reaction jar: 
 
 a. Dissolved oxygen. 
 
 b. Ferrous iron. 
 
 c. pH. 
 
 d. Temperature. 
 
 e. A 35 ml sample was filtered through a 0. 22|~i membrane 
 filter and total iron was measured in the filtrate. 
 
 6. An additional sample was also aerated to the same pH and 
 the alkalinity of the water after aeration was measured on 
 this sample. This information was used to compare the iron 
 oxidation rate with the formula proposed by Ghosh (42). 
 
 7. In all. cases, samples were taken with a siphon to protect 
 against additional aeration during sampling procedures. 
 
58 
 Marsh (43) used the extracts obtained in this study to further search 
 for interferences with iron removal. In addition to the above procedures, 
 he also ran a parallel study using iron-59 as a tracer and followed the rate 
 of conversion of soluble to insoluble iron instead of measuring the conver- 
 sion of ferrous to ferric iron. 
 
 E. Characterization of the Organic Extracts 
 
 Mueller, et al. (91, 92) separated organic acids, extracted from 
 river water, by eluting them through a silicic acid column. Identification 
 of some acids, separated in this manner was accomplished by using paper 
 chromatography. 
 
 The method used for identification of dicarboxylic acids (92) was 
 used in this study with several of the extracts and compared to several 
 known dicarboxylic acids as a reference. The solvent system was pre- 
 pared by mixing n-amyl alcohol and 5 M formic acid volume for volume. 
 The solvent was separated with a separatory funnel into an alcoholic and 
 an aqueous phase. The aqueous phase was retained in a beaker in the 
 chromatography jar. After the acids were applied to the paper, the jar 
 was sealed with the paper suspended above the solvent for one hour to 
 allow for the atmosphere to come to equilibrium. The paper was then 
 Lowered into the solvent which was allowed to rise through the paper for 
 about six hours. The paper was then dried for about two hours in a flow- 
 ng current of air. The chromato grams were developed by spraying with 
 i 0. 04 percent solution of bromophenol blue in 95 percent ethanol adjusted 
 
59 
 to pH 6. 7. The acids appeared as yellow spots on a field of blue. The equip 
 ment used was as follows: 
 
 1. Chromatography jars, 8 3/4 in. 0. D. x 1.8 in. , with 
 ground glass top edge. 
 
 2. Flexiglass tops drilled for stoppers through which hangers 
 were inserted and lowered. 
 
 3. Whatman No. 40 chromatography paper cut to 14 x 16 in. 
 with acids applied one inch up along the 14 inch side. Paper 
 was stapled into a 16 inch long cylinder. 
 
 4. Acids were applied with melting point capillary tubes. 
 Takem, et al. (98) and Buch, et al. (99) have also used paper chroma- 
 tography for identification of organic acids. 
 
 Gas chromatography is another technique which has been used 
 successfully by some to identify organic material. Febeck (7 3) sepa- 
 rated a number of compounds from soil extracts using gas chromatogra- 
 phy. Lamar and Goerlitz (100) used this technique to characterize the 
 ''■ 
 carboxylic acids in unpolluted streams. 
 
 Even with the use of separation techniques such as those mentioned 
 I above, more elaborate methods are needed to actually identify functional 
 groups and/or actual compounds. A satisfactory approach, currently 
 being used, is infra-red spectroscopy. In a book by Bellamy (101) the 
 technique is discussed and the characteristics of all, the major functional 
 groups are discussed as they appear on spectrographs. Infra-red spectrom- 
 ieters measure the frequencies of the vibrations of the various linkages in 
 
60 
 molecules. As the actual frequencies of vibrations of connecting bonds can 
 only be predicted for very simple molecules, the analyst must rely on 
 empirical data which has been accumulated relating infra-red absorption 
 bands with structural units in making an interpretation of a spectra. Bellamy 
 has presented a summary of the data available and included many charts of 
 the characteristic absorption bands for many structural units. This tech- 
 nique has been used by Shapiro (60, 78, 79, 90), Middleton (81), and Rice, 
 et al. (102) in identifying some of the organic matter in water as carboxylic 
 acids. 
 
 Infra-red spectrographs of the organic extracts from the Illinois 
 ground waters used in this study were made in the laboratory of the Illinois 
 State Health Department at Springfield, Illinois and in the Chemistry Depart- 
 ment at the University of Illinois. Two methods of determining the spectra 
 were used. The extracts were fused with potassium bromide into pellets 
 under high pressure or they were smeared in solution on sodium chloride 
 crystals. 
 
 In order to determine the percent organic matter in samples, vari- 
 ous oxidation techniques have been used. Standard Methods (103) suggests 
 that a residue from a water sample be ignited at 600 C. and the loss on 
 ignition corresponds to orga.nic matter plus inorganic matter lost due to 
 decomposition and volatilization. Pickhardt, et al. (104) suggest that, for 
 the determination of the total carbon, in organic materials, the sample 
 should be ignited at 750°C. in the presence of a copper oxide catalyst. In 
 jthis study the organic composition of the extracts was made by determining 
 
61 
 the total carbon content of the extracts. This analysis is similar to the one 
 discussed by Prickhardt, et ah (104). This determination consisted of burn' 
 ing the extracts in a carbon dioxide -free environment and measuring the 
 total amount of carbon dioxide given off. This analysis was performed by 
 the Organic Chemistry Microanalysis Laboratory at the University of 
 Illinois. 
 
 Chemical oxidation can also be used as a measure of the amount 
 of organic matter in a sample. Puri and Sarup (105) used boiling potas- 
 sium permanganate to measure the amount of oxidizable material in a sam- 
 ple extracted from soil with sodium hydroxide. They found the results to 
 be comparable with those obtained when potassium dichromate was used as 
 the oxidizing agent. 
 
 In this study potassium dichromate was used as directed in Standard 
 Methods (103) except that, as the extracts were added in known weights, the 
 results were obtained in mg oxygen demand per mg extract. Analyses of 
 pure compounds were also made so as to determine the completeness of 
 this oxidation procedure. 
 
 An attempt was made to classify the extracts on the basis of solu- 
 bility in various solvents using the scheme described by Middleton (81). 
 However, as this scheme is basically for chloroform extracts, and as the 
 chloroform extracts were the least significant in the study, this method of 
 separation and classification met with little success as almost all fractions 
 
 1 ell in the water-soluble fraction. 
 
 > 
 
 Several samples were dissolved in water and used to produce 
 
62 
 titration curves. They were titrated with 0.02 N sodium hydroxide and 0.02 N 
 hydrochloric acid and the pH titration curves were plotted in an attempt to 
 determine the number of ionizable groups present. Puri and Sarup (93) used 
 this method to characterize humic acids and their results compare closely 
 with the curves obtained in this study. 
 
 F. Chemical Analyses 
 
 In general the other chemical analyses performed in this study were 
 carried out in accordance with Standard Methods (103). The actual methods 
 used and any deviations made will be listed in Appendix A. 
 
63 
 IV. EXPERIMENTAL RESULTS AND DISCUSSION 
 
 A. Characterization of the Waters Studied 
 
 Earlier studies of factors affecting the removal of iron from well 
 water conducted at the University of Illinois by Longley (10)., Ghosh (42), 
 Komolrit (49) , and at the Illinois State Health Department by Weart and 
 Margrave (1), were used as a guide in selecting waters for this study. The 
 towns of Oakwood., Clinton s and Deland.. Illinois, were chosen because of 
 reported difficulties with iron removal and persistent slime growth on fil- 
 ter media. This growth seemed a good indication that there was organic 
 matter present in the water being treated. After two preliminary experi- 
 ments at Deland, this location had to be abandoned because of major con- 
 struction at the water plant. Iron removal at Oakwood was so unsatisfactory 
 that the well supply was soon to be abandoned in favor of a surface supply 
 for which a treatment plant was nearing completion at the time of this study. 
 
 Philo and Atwood, Illinois, were chosen as plants where iron removal 
 was being satisfactorily accomplished. Philo has a low and fluctuating iron 
 content in its raw water supply. One of the wells in operation was selected 
 for determination of its organic content. However, when a later attempt 
 was made to measure organic interference with iron removal, the iron con- 
 tent of this particular well had dropped to nearly zero. The study was then 
 made at another well several hundred yards away. Atwood has a higher 
 iron content in the raw water but was passing 100 percent of its water through 
 a cation exchanger to insure a high and satisfactory degree of iron removal 
 
64 
 
 as well as an unusually soft finished water. Two sets of preliminary tests 
 were also made on the local Champaign -Urbana, Illinois, tap water, mainly 
 for the purpose of testing the equipment before taking it into the field. This 
 water was also known to present difficulties with iron removal from time to 
 time. As the supply consisted of a number of wells in several different 
 well fields, the water varied constantly in quality as the various wells were 
 placed in and out of service. Soon after this study was completed, the iron 
 removal treatment plant was replaced by two modern lime softening plants. 
 
 All the plants considered in this study consisted of coke tray aerators, 
 reaction tanks, and pressure filters, except for Champaign -Urbana which 
 had a spray aerator, reaction tank, and gravity filters. Table 1 lists the 
 usual mineral characteristics of raw water from the wells under consider- 
 ation and finished water from the treatment plants. Since, in all cases, 
 the plants treat water from several wells, the finished water characteris- 
 tics reported are for a mixture of waters. These figures have been obtained 
 from plant operational reports on file at the Illinois State Health Department. 
 From this table it can be seen that Philo and Atwood consistently produce 
 satisfactory iron removal while Clinton and Oakwood exceed the maximum 
 recommended concentration of 0. 3 mg/1 during a considerable proportion 
 of the time. The mineral characteristics of the waters do not indicate any 
 pattern which would point to an interference with iron removal. The finished 
 water data represents data collected for the first six months of 1962. 
 
 At Clinton, Oakwood, and Philo the carbon filters were attached to 
 raw water sample taps at the wells. The wells tested were: Clinton, Well 
 

 Q 
 
 
 
 W 
 
 
 
 i—i 
 
 
 
 Q 
 
 
 
 O 
 
 
 
 H 
 
 
 
 co 
 
 
 
 CO 
 
 
 
 Pi 
 
 +J 
 
 ti 
 
 
 H 
 
 <u 
 
 
 «J 
 
 a 
 
 
 £ 
 
 
 
 fc 
 
 R} 
 
 ft 
 
 
 
 
 <D 
 
 1— 1 
 
 
 Q CM 
 
 
 CO 
 
 O 
 
 w 
 
 PQ 
 
 O 
 
 i— i 
 
 H 
 co 
 
 Xi o 
 
 +-> 1— 1 
 t— 1 
 
 co - 
 
 < 
 
 2 
 w 
 
 4_) 
 
 
 H 
 
 RJ 
 
 
 U 
 < 
 
 CO 
 
 
 ti 
 
 CO 
 
 •H 
 
 
 < 
 
 
 
 
 X 
 
 •H 
 
 
 u 
 
 r— 4 
 1—4 
 
 w 
 
 h £ a 
 
 X 
 ft 
 
 CD 
 
 -I 
 
 fl to 
 
 DO 
 
 "S 6 
 
 H 
 
 cu r5 
 
 ^ 
 -M GO 
 
 < 2 
 
 1—4 
 
 CO 
 
 r— 4 
 
 co 
 
 a a 
 
 CO 
 
 ro 
 
 GO 
 
 rj 
 
 a u 
 
 bO 
 
 ^ 
 
 co 
 
 r3 co 
 
 I** ° 
 
 rJ r ^ cj 
 
 h w a u 
 
 0) 
 
 0) 
 
 1—1 
 
 ft 
 
 
 
 a 
 
 2 
 
 Rj 
 
 
 
 CO 
 
 CO 
 
 
 ^ 
 
 • 
 
 +J 
 
 
 
 ft 
 
 K 
 
 <u 
 
 1—1 
 
 p 
 
 i—4 
 
 0) 
 
 Xi 
 
 £ 
 
 Pi 
 
 co 
 
 CM 
 
 © 
 
 I 
 CO ^ o 
 
 CO O 
 lO ^ 
 
 CM 
 
 00 00 
 
 00 
 
 o 
 
 LO O ^ 
 
 n r- h 
 
 CM 00 
 
 CO 
 
 >> 
 
 CO 
 
 i i ™ 
 
 CO CO -H 
 
 tf tf Q 
 
 
 --H CM 
 
 X* 
 O 
 
 o 
 
 <1 
 
 o 
 c\i 
 
 
 d 
 
 i 
 
 o o 
 
 CM O 
 
 ■^ O 
 
 
 o 
 
 o 
 
 
 oo 
 
 00 
 
 00 
 
 o 
 
 LO vD 
 
 a 
 
 ■ 
 
 • • 
 
 r*- 
 
 r- 
 
 I s - r- 
 
 co 
 
 IT) 
 
 00 
 
 CM 
 
 CO 
 
 o 
 oo 
 
 IT) 
 
 O 
 CM 
 
 CM 
 
 O 
 
 4-> 
 CO 
 
 CO 
 
 CO 
 
 Rj 
 
 LO 
 
 oo 
 
 
 s 
 
 o 
 
 +-> 
 ti 
 
 •H 
 1—4 
 
 u 
 
 CO 
 
 oo 
 
 o 
 
 0O 
 
 +J 
 
 CO 
 
 CO 
 
 £ CO 
 
 CO -i-4 
 
 ti Q 
 
 CM 
 
 ro 
 
 XI 
 
 o 
 
 o 
 
 I 
 
 O 
 
 CM 
 
 © 
 I 
 
 O O O 
 
 65 
 
 o 
 
 
 o 
 
 r~ 
 
 
 oo 
 
 sO 
 
 uD 
 
 LD 
 
 ■ 
 
 • 
 
 • 
 
 r~ 
 
 r- 
 
 r~ 
 
 LO 0O 
 oO 0O 
 
 —4 00 
 
 CM r-4 
 
 CM 
 
 ■—4 s 
 
 CO CM 
 
 CO 
 
 CO 
 
 RJ R5 .H 
 
 tf Cd Q 
 
 O LO 
 
 CM vO 
 
 oo ^ 
 
 A' 
 
 Ph 
 
66 
 
 No. 6; Oakwood, Well No. 3; and Philo, Well No. 3. As the pumps at 
 Atwood did not run continuously, the carbon filters were attached to a fau- 
 cet in the distribution system. Analyses of the waters at the time of the 
 extractions are given in Table 2. Even though considerable iron was pres- 
 ent in the waters at the time the extractions were made, the filters were 
 not plugged during operation and, as will be shown later, organic extracts 
 were obtained which contained a minimum concentration of iron. 
 
 B. Measurement and Recovery of Organic Matter in Well Water 
 
 Only one carbon filter was used to adsorb organic matter from well 
 water in the early experiments. During these experiments, a rate of flow 
 between 0. 25 and 0. 50 gpm was attempted, as recommended by Middleton, 
 et al, (81). It was hoped that these filters would stay in operation long 
 enough to adsorb the organic matter from about: 5000 gallons of water as 
 recommended by Middleton. In the first three experiments, the water meter 
 was upstream from the carbon filter and was protected from sediment from 
 the wells by a sand and glass wool pre -filter. In each instance the filter 
 did not prevent the passage of very fine sand into the water meter. Since 
 the meter became jammed during these experiments, the total flow through 
 the carbon filters was estimated from the measured rates of flow and the 
 time of operation. As the wells were out of service during these periods, 
 lestimates of total time of operation were obtained from the plant operators. 
 
 These estimates were then used to translate the amounts of extracts obtained 
 
 I. 
 
 into concentrations in terms of mg/1. Using these estimates, the concentrations; 
 
67 
 
 M 
 
 w 
 
 co 
 U 
 
 w 
 
 H 
 
 -1 
 
 h 
 
 O 
 
 « 
 
 u 
 o 
 
 H 
 
 Q 
 W 
 
 Oh 
 < 
 
 co 
 
 < 
 
 CO 
 
 W 
 
 H 
 
 fa 
 
 
 to 
 
 W 
 
 CO 
 
 < 
 < 
 
 h 
 
 
 CO S 
 
 Q 
 O 
 
 u 
 
 d 
 W 
 
 OJO 
 
 Ed DJO 
 
 CO 8 
 
 9 
 
 <u oj oo 
 
 ti Tl H O 
 
 a s ^ ^ 
 
 a 
 
 9 
 
 H 
 U 
 
 13 jS 
 
 en 
 
 ,H O 
 
 0D 
 
 U 
 rd 
 
 o k a u 
 
 r-j Tj "H O 
 
 h ffi a u 
 
 CO 
 
 
 
 tn 
 
 o 
 
 00 
 lO 
 
 o 
 
 r 
 
 I— I 
 
 PQ 
 < 
 
 o — . 
 
 rO OX) 
 
 00 
 LT) 
 
 N °" 
 
 r-. X) 
 
 • d 
 
 o nj 
 
 +-> 
 o d 
 <u 
 +-> 
 d 
 o 
 y 
 
 • § 
 
 N 
 
 •H 
 
 cd 
 
 fd 
 
 d 
 o 
 
 00 o 
 
 i-H 0) 
 rd 
 
 y 
 
 CO j3 
 
 CO O 
 
 PQ 
 
 (M 
 
 00 
 
 o 
 
 <! 
 
 oo 
 
 00 
 00 
 
 00 
 
 
 o 
 ro 
 
 
 oo 
 
 ro 
 
 rd 
 
 0O 
 
 U 
 
 oo 
 OJ 
 
 
 o 
 
 ro 
 
 I s - 
 
 
 o 
 
 ro 
 
 OO 
 
 r^ 
 
 o 
 
 i— i 
 
 o 
 
 ro 
 
 oo 
 
 ro 
 
 ^ 
 
 o 
 o 
 
 IT) 
 
 U 
 
68 
 calculated for the two experiments at Deland and the comparison of the early 
 estimated experiment at Clinton with the two metered experiments made 
 later, agree fairly closely. This tends to somewhat, validate the estimates. 
 A comparison of the extracts, obtained from the early Clinton experiment 
 with the two later experiments, and the two Champaign experiments, indi- 
 cates that the carbon filters did not function as efficiently at 1. gpm as 
 they did at 0. 5 gpm. However, it is felt that the rather small decrease 
 in efficiency was more than offset by the greater facility with which samples 
 were collected. In Table 3 are listed the amounts of extracts obtained from 
 the carbon filters during the early experiments when only one filter was in 
 use. These preliminary experiments, in which only one filter was used, 
 are numbered with lower-case Roman numerals to differentiate them from 
 the more complete experiments performed later. 
 
 In Table 4 are listed the amounts of extracts obtained from the car- 
 bon filters when both unacidified and acidified water was passed through, 
 carbon filters. When filter C was used, it was found that 50 to 60 percent 
 as much extract was obtained from filter C as from filter B. From this 
 it is apparent that not all the organic matter applied to filter B was retained. 
 From this, it may be assumed that the effluent from filter C may have also 
 contained some organic matter. Thus the total amount of organic extract 
 obtained from a water has been reported as the amount obtained from fil- 
 ter A plus the amount obtained from filter B and the amount obtained from 
 filter C when it was in use. Although the total, concentrations thus deter- 
 mined are undoubtedly less than the actual concentrations present in the 
 
69 
 
 co 
 W 
 
 ffl 
 < 
 
 CO 
 
 w 
 
 H 
 W 
 
 o 
 
 tn 
 
 Q 
 
 W 
 
 > 
 
 u 
 w 
 
 co 
 H 
 U 
 <! 
 
 a! 
 H 
 X 
 
 w 
 u 
 
 g 
 
 
 
 
 u 
 
 w 
 
 o 
 
 m 
 O 
 
 tuO 
 
 GuD 
 
 b£ 
 
 s ^ 
 
 co D 
 
 0) CO 
 
 d d 
 
 ,-4 5 O 
 
 iS •- , ■-' 
 
 O O aJ 
 
 H > M 
 
 a 
 
 Oh [ih M 
 
 So 
 
 CD 
 
 H 
 W 
 
 d 
 
 o 
 
 H 
 
 o 
 
 o o 
 
 O -H 
 
 o 
 
 o 
 u 
 o 
 
 ,d +3 
 
 o 
 o 
 CM 
 
 CO 
 
 d 
 
 d 
 i— < 
 
 Q 
 
 CM 
 
 t-< oo 
 
 O CM 
 
 o o 
 
 u 
 
 
 
 
 "4-1 
 
 .— < 
 
 
 
 u 
 
 
 
 d 
 
 o 
 
 CtJ 
 
 I— 1 
 
 rd 
 
 ,d 
 
 +-> 
 
 U 
 
 W 
 
 o 
 
 LD 
 O 
 IT) 
 
 LT) 
 O 
 
 d 
 o 
 +-> 
 d 
 
 <H 
 
 1—4 
 
 u 
 
 o 
 
 CM 
 
 i>- 
 
 co 
 
 i— I LO 
 
 O CM 
 lO — i 
 
 ■^ O vO O 
 O i— i i— i vD 
 
 O O 
 
 O O 
 
 O O O O 
 
 o <$ 
 
 8 w 
 
 o 
 
 CM 
 
 i—4 
 
 00 
 
 CM 
 
 -* 
 
 CO 
 
 00 
 
 ^ 
 
 IT) 
 
 L<0 
 
 d 
 
 d 
 
 d 
 
 -d 
 d 
 (3 
 
 i—i 
 co 
 
 Q 
 
 d 
 a, 
 
 cc3 Oh 
 
 d 
 
 ojO 
 
 • H 
 
 ccj 
 
 ccj a, 
 
 rd d 
 
 r- 
 
 Tf 
 
 r- 
 ^ 
 
 o o 
 
 LD -^ 
 
 IT) O 
 
 CM O 
 
 a- in so ^ 
 
 LO 
 
 O CM 
 
 !— 1 IT) 
 
 O CM 
 
 00 CM 
 
 O CO CO ^ 
 
 o 
 
 a 
 
 
 'cb 
 id 
 
 u 
 
 s + + 
 
 -< 'd 
 
 
 
 m r— 1 
 
 o 
 
 m r—i i—i i—i 
 
 5 Si 
 
 
 
 O O 
 
 U d 
 
 n S d d 
 
 2 a 
 
 rt 
 
 O ccj ccj nj 
 
 3£ 
 
 tj ,d a a 
 
 rQ -(-> -M +-> 
 
 •H 
 
 y w 
 
 o w w w 
 
 Q h 
 
 U H U H 
 
 PH 
 
 
 ccj 
 
 
 co 
 
 ro 
 
 cu 
 
 
 •(-> 
 
 ID 
 
 ccj 
 
 nj 
 
 (J 
 
 a 
 
 1—1 
 
 +■> 
 
 
 
 co 
 
 p 
 
 w 
 

 
 
 
 
 ri 
 
 
 
 (H 
 
 
 
 +-> 
 
 
 
 X 
 
 
 
 w 
 
 r— t 
 
 
 .—1 
 
 00 
 
 
 
 
 s 
 
 
 H 
 
 
 CO 
 
 
 1—1 
 
 u 
 
 U +> 
 
 ojo 
 
 W 
 H 
 
 % d 
 
 a 
 
 < 
 
 5 n 
 
 
 £ 
 
 3 X 
 
 
 J 
 
 fa w 
 
 a 
 
 on 
 
 J 
 
 
 
 W 
 
 
 
 £ 
 
 
 r-i 
 
 5 
 
 
 
 
 oo 
 
 * 
 
 h 
 
 j3 *H 
 
 3 X 
 
 
 Q 
 fa 
 
 fa fa 
 
 OO 
 
 H 
 
 
 
 U 
 
 
 i— i 
 
 « 
 
 < 4, 
 
 OX) 
 
 H 
 X 
 
 S £ 
 
 a 
 
 W 
 
 fa w 
 
 a 
 
 H 
 
 
 
 H 
 
 
 
 4 
 
 
 
 
 
 
 
 
 fa 
 
 
 
 
 
 H 
 <! 
 
 H 
 fa 
 
 u 
 
 
 
 u 
 
 d 
 
 > 0) 
 
 w 
 
 2^5 
 
 d 
 <D 
 
 a . 
 
 •H O 
 
 a? Z 
 ft 
 
 fa 
 
 d 
 
 o 
 
 H 
 
 70 
 
 r— o -^ co 
 
 N O H M 
 
 i— i oo cm r- m 
 Tf cm oo ^ -^i 
 
 o o o o ^ 
 
 r^ o oo o oo 
 
 r— < O " H I— I 0— 
 
 r- in cm tri oo 
 
 oo -^ IT) -^ ■—* 
 
 o o o o — • 
 
 lD 00 i-H I s - vO 
 
 CI iH (Vj H ^t 
 
 s ,— i O IT) — l 
 
 oo oo oo cm r>- 
 
 O O O O *-* 
 
 no oo cm o r- 
 
 o 
 
 o 
 u 
 o 
 
 oO-h 
 4- 4- 
 
 
 
 
 
 
 
 >H S 
 
 d 
 
 a 
 
 <l) (d 
 
 cd 
 
 rd 
 
 3 -fj 
 
 X 
 
 ^ 
 
 u ^ w w w 
 
 o 
 
 T3 
 
 
 o 
 
 I 
 
 cd 
 
 
 oo oo oo ^ un 
 
 ^ i— i o s oo 
 
 m 
 
 o 
 
 o o o o 
 
 (>• h f. ^ N 
 
 o o r~- cm >-< 
 
 © o o o o 
 
 fl rl 00 CO I 1 
 © © CM © © 
 
 o r^ I s - oo oo 
 
 O O r— I 1—1 sO 
 
 o o o" o o 
 
 M d vO ifl 1< 
 
 o o o o cm 
 
 u 
 o 
 
 o 
 
 
 
 oo _. 
 
 x u 
 
 4- 4- 
 
 , 
 
 a ti ti d 
 
 p rt ni rt 
 
 ,3 ™ 4-> +-> +j 
 
 u £ w w w 
 
 o 
 
 IT) 
 
 o 
 
 CM 
 
 o 
 
 IT) 
 
 o 
 
 ^ O O •-< 
 
 oo r- o h« 
 
 — < O O CM O 
 
 O O 00 CM o 
 
 co r^ co Tt 1 o 
 O O ~i o o 
 
 *• ^ CM OO 00 
 W W o CM oo' 
 
 ■^ O C58 O O 
 O O S CM ^ 
 
 CM 
 
 00 
 
 
 o 
 +-> 
 d 
 
 •W 
 
 r— < 
 
 u 
 
 
 f 
 
 oo 
 
 o 
 
 o 
 
 o 
 
 O CM 
 
 
 CM 
 
 o 
 
 r- 
 
 vD uO 
 
 
 o 
 
 o 
 
 o 
 
 i-* 00 
 
 00 r-. 
 
 X o 
 
 
 a 
 
 
 
 Z X 
 
 
 u 
 
 
 
 
 
 + +- 
 
 
 <-H 
 
 
 I—) 
 
 ^-1 r— 1 
 
 
 
 
 
 o 
 
 O 
 
 
 !h 
 
 u 
 
 a 
 
 a ti 
 
 i— i 
 
 
 
 <X) 
 
 ri 
 
 ccj ccj 
 
 cd 
 
 1—1 
 
 cd 
 
 Jl 
 
 A jd 
 
 +j 
 
 aJ 
 
 
 +-> 
 
 +-> +J 
 
 
 
 U 
 
 ^ 
 
 W 
 
 W W 
 
 H 
 
 d 
 
 CJ 
 
 CD 
 ft 
 
 X! 
 W 
 
 d 
 o 
 
 4-> 
 
 d 
 
 • H 
 r-l 
 
 u 
 
 
 o 
 
 <u 
 
 d 
 
 h 
 
 ■H 
 
 rd 
 
 o 
 
 Xi 
 
 d 
 
 CD 
 
 
 UJ 
 
 CO 
 
 d 
 
 •M 
 
 
 o 
 
 CQ 
 
 ctj 
 
 CD 
 
 u 
 
 d 
 
 +J 
 
 i—i 
 
 X 
 
 cd 
 
 CL) 
 
 > 
 
 i— I 
 
 „ 
 
 u 
 
 ■4-> 
 
 rn 
 
 K 
 
 
 
 
 r-l 
 
 + 
 
 
 
 UJ 
 
 i— i 
 
 CD 
 
 o 
 
 ^-1 
 
 d 
 
 ft 
 
 
 a 
 
 4-> 
 
 cd 
 
 H 
 
 CO 
 
CO 
 ■t-> 
 U 
 ctj 
 
 U 
 +j 
 
 X 
 
 w 
 
 
 
 H 
 
 U 
 
 bO 
 
 GO 
 
 CQ 
 
 M 
 
 * fi 
 
 5 x 
 
 fa W 
 
 B 
 
 cj 
 
 M 
 
 $2 
 
 .-J + - > 
 
 H 
 
 tn W 
 
 H 
 
 
 GO 
 
 c 
 
 > 
 
 i—i 
 
 
 
 £ 
 
 •^ CM 
 
 sO IT) 
 
 LD i— I oO r— ■— ) 
 
 N h (js oo Tf 
 
 O i-h o o o 
 
 CM LT) LO CO O 
 •-* IT) •<# -^ (\j 
 
 o 
 
 i-H 
 
 00 
 
 00 
 
 ■sO 
 
 o 
 
 vO 
 
 ^ 
 
 CM 
 
 i-i 
 
 o 
 
 o 
 
 o 
 
 O 
 
 i— ♦ 
 
 ^ 
 
 o 
 
 i-H 
 
 "* 
 
 h- 
 
 o 
 
 CO 
 
 CM 
 
 i-H 
 
 cn 
 X 
 
 
 s 
 
 
 
 z w 
 
 
 
 -f- 
 
 +' 
 
 
 
 
 r— 1 
 
 I— 1 
 
 ,-1 
 
 MH 
 
 
 
 
 
 
 
 
 
 
 u 
 
 a 
 
 d 
 
 a 
 
 h 
 
 <v 
 
 ctj 
 
 crt 
 
 ctj 
 
 
 
 ctj 
 
 rCi 
 
 rd 
 
 .d 
 
 
 -M 
 
 ■w 
 
 -M 
 
 ^ww 
 
 CO 
 
 
 
 
 
 o 
 
 
 
 
 
 CM 
 
 
 
 
 
 I s - in lo i—i oo 
 m un oo oo o 
 
 -tf o ■— i oo cm 
 
 I s - I s - O I s - ^ 
 
 ■— « ^o o un o 
 
 o o <— i o o 
 
 CTJ 
 
 4-> 
 
 
 
 H 
 
 h- O 0O LT) 00 
 
 O CM -^ cm' cm' 
 
 o oo cm m r# 
 oo no r^ o vD 
 
 o o ■— i <— ( o 
 
 CO O LO sO 00 
 "i OO t^ -"^ CM 
 
 00 O O -sO O 
 O O IT) tJh o 
 
 O O O O CM 
 
 "^ O sO O ■— i 
 O O CM CM O 
 
 00 _) 
 
 K u 
 Z X 
 
 + + 
 
 o.ooo 
 % ft rt d CI 
 £ ctj ctj ctj 
 
 7l (Tj ,£! X! ,C| 
 
 An ^ +J +J 4-> 
 
 u £ w w w 
 
 00 
 
 LT) 
 
 71 
 
 CM 
 
 o 
 H 
 
 •a o 
 u Z 
 cu 
 
 ft 
 
 w 
 
 IT) 
 
 
 H 
 
 o 
 o 
 
 CI 
 
 o 
 d 
 
 •H 
 
 1—1 
 
 u 
 
72 
 
 waters under consideration, they should be of sufficient accuracy to facili- 
 tate comparisons of the organic, concentrations of these waters with waters 
 in other regions. 
 
 The quantities of extracts obtained with ethanol plus hydrochloric 
 acid as a solvent were not used in calculating total concentrations of organic 
 extracts obtained from the waters because the major part of the extraction 
 residues was found to be ammonium chloride formed by a reaction between 
 immonia, remaining in the carbon after the ethanol plus ammonia extrac- 
 :icns. and hydrochloric acid from the final extractions. This observation 
 )f the presence of ammonium chloride was confirmed by analyzing all eth- 
 mol plus hydrochloric acid extracts for chloride. The extracts were found 
 o contain from 37 to 57 percent chloride. With such a high percentage of 
 hloride in these extracts, any adjustment in the weights obtained would 
 Lecessarily be so large that the final values for organic extracts present 
 fould be entirely meaningless. 
 
 Table 3 reveals a large difference in the amount of chloroform 
 xtract obtained in the two experiments at Champaign. In the first experi- 
 ment 0. 50 mg/1 "Chloroform A" extract was obtained while in the second 
 jXperiment only 0.04 mg/l was obtained. This difference is probably 
 lerely a reflection of different waters being used in the experiments. The 
 ater for Champaign is from a number of wells in several different well 
 elds. As the same wells are not always in operation, differences in organic: 
 mcentration of the finished water can be expected to occur. 
 
 The chemical oxygen demand of the waters was measured before and 
 
73 
 after the water was passed through the carbon filters. Table 5 indicates 
 that organic matter was being removed from waters as they passed from one 
 filter to the next. In comparing this data with, the total amounts of extracts 
 recovered (Table 4), however, there is no direct numerical correlation 
 between COD and amounts of organic extracts obtained. Some samples were 
 rendered useless because fine carbon from the filters was eluted into the 
 sample flasks at the time of collection. This carbon gave high, and erratic 
 COD values. 
 
 As the methods used to obtain the extracts did not insure the elimi- 
 lation of inorganic matter, two methods were used in an attempt to deter - 
 nine the percentage of organic matter in the extracts. The organic extracts 
 yere oxidized chemically using boiling potassium dichr ornate in concentrated 
 ulfuric acid as the oxidizing agent. Silver sulfate was used as a catalyst 
 o insure more complete oxidation. As the samples were added by weight 
 f volume, the chemical oxygen demand of the extracts were obtained as mg 
 /OD per mg extract instead of in mg/l. In order to see if this method would 
 ive a true picture of the COD of organic acids, the COD values of several 
 are organic acids were determined and compared with theoretical values 
 »r total oxidation. In Table 6 are given the results of this study. It is to 
 <i noted that the experimental COD values for the pure compounds agree 
 T ith the theoretical values within 2. 5 percent except for malfcmic acid which 
 
 ■ ows a difference of 1.0 percent. In most instances, the COD values of the 
 
 ! 
 
 «! tracts fell, somewhere between those of the short chain dicarboxylic acids 
 jammed and pure carbon which would have the maximum COD expected of 
 
74 
 
 TABLE 5 
 
 REMOVAL OF CHEMICAL OXYGEN DEMAND 
 BY ACTIVATED CARBON FILTERS 
 
 Town 
 
 Experiment 
 No. 
 
 Cjh emical. Oxygen Demand 
 Before After After After 
 
 Filter A Filter A Filter B Filter C 
 mg/1 mg/1 mg/1 mg/l 
 
 Oakwood 
 
 1 Begin 
 
 15. 2 
 
 
 
 
 1 End 
 
 10. 5 
 
 
 9. r 
 
 Philo 
 
 2 Begin 
 
 0. 
 
 0. 
 
 0.0 
 
 
 2 End 
 
 0. 
 
 0. 
 
 0.0 
 
 Clinton 
 
 3 Begin 
 3 End 
 
 38.0 
 37. 6 
 
 27. 5 
 
 25. 5 
 
 Atwood 
 
 4 Begin 
 4 End 
 
 23. 1 . 
 
 
 18.0 
 
 Clinton 
 
 5 Begin 
 
 37. 1 
 
 21. 7 
 
 13. 9 
 
 
 5 End 
 
 40. 5 
 
 36.6 
 
 27.3 
 
 21.8 
 
75 
 
 TABLE 6 
 CHEMICAL OXYGEN DEMAND OF ORGANIC EXTRACTS 
 AND PURE COMPOUNDS 
 
 Town 
 
 Experiment 
 No. 
 
 Extract 
 
 Chemical Oxygen Demand 
 Measured Theoretical 
 
 mg/mg Ext. mg/mg Ext, 
 
 Dakwood 
 
 Philo 
 
 1-lmton 
 
 D ure Compounds 
 
 Et.han.ol A. 
 
 0. 
 
 71 
 
 
 
 Ethanol B 
 
 2. 
 
 08 
 
 
 
 Ethanol A 
 
 0„ 
 
 50 
 
 
 
 Ethanol B 
 
 0. 
 
 12 
 
 
 
 Ammonia B^ 
 
 0. 
 
 22 
 
 
 
 Ethanol A 
 
 1. 
 
 27 
 
 
 
 Ammonia A 
 
 1. 
 
 39 
 
 
 
 Chloroform B 
 
 1. 
 
 67 
 
 
 
 Water B 
 
 1. 
 
 04 
 
 
 
 Ethanol B 
 
 1. 
 
 23 
 
 
 
 D Tartaric Acid 
 
 0. 
 
 517 
 
 0. 
 
 533 
 
 D Tartaric Acid 
 
 0, 
 
 539 
 
 0. 
 
 533 
 
 Malonic Acid 
 
 0. 
 
 680 
 
 0. 
 
 615 
 
 DL Malic Acid 
 
 0. 
 
 714 
 
 0. 
 
 716 
 
 Succinic Acid 
 
 0. 
 
 928 
 
 0. 
 
 948 
 
 Carbon 
 
 
 
 2. 
 
 66 
 
 ample Calculations for theoretical COD: 
 
 ) Tartaric Acid: C 4 H 6 6 + 5 2 = 3 C0 2 + 2H 2 
 
 2 
 
 lolecular weight: 
 
 150 
 
 80 
 
 80 
 150 
 
 = 0. 533 mg COD per mg tartaric acid 
 
76 
 organic compounds. From these results, it may be concluded that the 
 extracts must be nearly 100 percent organic or it would have been practi- 
 cally impossible for the COD values to have been so high. These values 
 ire in the same range as those determined by Myrick and Ryckman (106) 
 or extracts recovered from the Missouri River. Using a microchemical 
 inalytical technique, they determined empirical formulas for the chemical 
 omposition of their extracts. They determined the theoretical oxygen 
 lemand of an ethanol extract, based on its empirical formula to be 1. 39 
 ng/1 as compared to a measured COD of the same extract of 1. 37 mg/1. 
 
 Total carbon analyses of several of the extracts were performed 
 y the Organic Chemistry Microanalysis Laboratory at the University of 
 .linois. The results are given in Table 7. Using this data and the rough 
 stimate that the extracts consisted entirely of carbon, hydrogen, oxygen, 
 nd residue, the theoretical COD values of the extracts were calculated, 
 he values are in close agreement with the measured values (Table 6), 
 specially when it is remembered that the difference between the total 
 eight of the extracts and the carbon, hydrogen, and residue was cer- 
 
 inly not entirely oxygen. It would thus seem likely that the extracts 
 
 I 
 
 ;re composed almost entirely of organic compounds or metal salts thereof, 
 
 i it can be assumed that the other extracts would give similar results. 
 
 r ie presence of ammonium chloride is again demonstrated in the ethanol 
 as hydrochloric acid extracts by the high volatility and low carbon con- 
 it in the sample analysis. 
 
 It is realized that the extracts obtained do not represent the actual 
 
 J 
 
77 
 
 TABLE 7 
 TOTAL CARBON ANALYSES 
 
 Extract Percent Percent Percent Percent Calculated 
 
 Carbon Hydrogen Residue Oxygen* COD 
 
 mg/mg 
 
 Ul extracts are from Clinton Experiment 5. 
 
 Cthanol A 44.05 5.41 18.45 32.09 1.58 
 
 :hloroform B 57.75 6.98 5.15 30.12 1.99 
 
 Vater B 37.93 4.72 9-88 48.47 1.02 
 
 Ithanol B 45.37 5.15 3.11 46.37 1.20 
 
 .mmonia B 47.71 5.52 0.87 45.90 1.27 
 ;C1 B 3.25 6. 80 6. 59 
 
 imple calculations for COD: 
 
 thanol A: 100 percent - percent carbon - percent hydrogen - 
 
 percent residue = percent oxygen. 
 
 100 - 44. 05 -5.41 - 18. 45 = 32. 09 percent oxygen 
 
 100 mg extract = 18.45 mg residue plus 
 
 44. 05 + 5. 41 + 32. 09 = 81. 55 mg organic material 
 
 Weight . 44. 05 5.41 32. 09 
 
 lolecular Weight 12 1 16 
 I 
 
 Inpirical formula: C 3 6? H 5 41 2 Q1 + 4.02 2 = 3.67 CO z + 2.7 1 H20 
 
 * Secular weight: 81.55 128.6 
 
 128. 6 
 81. 55 
 
 1. 58 mg COD per mg extract 
 
 * 
 
 xygen assumed to be remaining percentage for the purpose of making 
 ie COD calculations. 
 
78 
 
 quality a.nd quantity of organic matter present in ground waters to a defined 
 degree of accuracy. Not all of the organic matter in the water was completely 
 adsorbed on the activated carbon, and not all. the organic: matter adsorbed on 
 the carbon was extracted with the solvents used. Also, any compounds, which 
 were volative at or below 100°C. , may have been lost. Despite all these los- 
 ses, the fact that the purity of the extracts was not precisely determined, 
 and the possibility that some structural changes might have occurred during 
 the extraction procedures: considerable, consistent, and meaningful quan- 
 tities of organic matter were obtained from the waters tested. The water 
 at Philo was found to contain about one -third as much organic matter as 
 the other waters studied. This water ..also contained much less iron and 
 presented no difficulty with iron removal. 
 
 Surprisingly, these well waters contain considerably more organic 
 matter than many surface waters presently being studied for organic pol- 
 lution. These well waters were found to contain 1. 5 to 7. 3 mg/l total 
 organic, extract while the Missouri River at Saint Louis, Missouri was 
 I found by Myrick and Ryckman (1.06) to contain less than 1. mg/l. Mueller, 
 et al. (92) reported that carbon filter samples from the Mississippi and 
 Ohio Rivers revealed from 0.19 to 0. 38 mg/l while low temperatures dis- 
 tillation and extraction with ethyl ether yielded from 1. to 2.4 m/ 1 of 
 organic material. Of this higher figure 33 to 62 percent was nonvolatile. 
 jNot all surface waters reported, however, are lower in organic content. 
 Shapiro (60) found the highly colored water of Lower Linsley Pond in 
 'Connecticut to contain between 4. 5 and 9. mg/l of organic acids. It is 
 
79 
 
 evident that the concentration of organic matter in some well waters is as 
 high as in many surface waters. 
 
 Effects of Organic Extracts on the Filterability of Iron 
 
 After the extracts had been obtained, demonstrating the presence 
 of organic matter in the ground waters under consideration, studies were 
 designed to determine what effects if any these extracts would have on the 
 filterability of iron. In previous work at the University of Illinois by 
 Longley (10), Ghosh (42), and Komolrit (49), studies of the mineral char- 
 acteristics of waters and oxidation rates of iron in the various waters had 
 revealed no sources of interference with iron removal. In fact, the iron 
 was always found to be in the ferric state when it reached the filters. Be- 
 cause of the low solubility of ferric iron, satisfactory treatment should 
 have been assured in all these waters. It thus seemed that the interfer- 
 ence with iron removal was an inability of the iron to form flocculant 
 particles of sufficient size to be retained by rapid sand filters. 
 
 As has been discussed previously, it was decided not to measure 
 either chelation capacity or zeta potential, but to measure directly the 
 quantities of iron which could be held in a filterable condition by known 
 amounts of the various extracts. In order to guard against the possibil- 
 ity of side reactions which might give erroneous results, only a minimum 
 of chemicals were added to the test solutions. Sodium hydroxide and hydro- 
 chloric acid were used to adjust the pH to 8.0; small quantities of dioxane 
 and potassium chloride were used to dissolve the extracts; distilled water 
 
80 
 
 vas used as the dispersion medium; air was used as the oxidizing agent; 
 
 errous sulfate was used as the ferrous iron source in most experiments, 
 
 md ferric nitrate was used when it was desired to add iron already in the 
 
 erric state. A pH of 8. was used as this was close to the pH of the natu- 
 
 •al waters studied after they had been aerated. When iron first started to 
 
 ippear on the membrane filters, the pH was raised or lowered in several 
 
 if the experiments so as to determine the pH dependency of the extracts. 
 
 ieveral of the extracts were studied after having been put into a strictly 
 
 ;ater solution and were found to produce effects slightly greater than when 
 
 loxane and potassium chloride were used as the original solvents for the 
 
 rganic extraxts. As it was very difficult to put some of the extracts into 
 
 olution with water only and as the solvent system was found to give results 
 
 n the conservative side, the solvent system was used in most of the experi' 
 
 T.ents. Air was introduced into the solutions by diffusion and by the vortex 
 
 roduced by stirring with a magnetic stirrer. There was undoubtedly a 
 
 tight amount of carbonate alkalinity present from the carbon dioxide 
 
 bsorbed with the oxygen from the air but it did not appear to affect the 
 
 ssults. Although the iron was added as ferrous sulfate,, it oxidized so 
 
 ! ipidly, when it contacted the aerated water at pH 8.0, that it was always 
 
 I 
 
 npossible to detect ferrous iron in the test solutions even though deter - 
 
 inations were made immediately after the addition of the iron. Actual 
 
 jgs of these experiments are given in Appendix B. A summary of these 
 
 jjsults is given in Table 8. The table lists the maximum concentrations 
 
 ; 
 
 \ iron found in the filtrates. 
 
i-< ^ o 
 
 o 
 o 
 
 o 
 o 
 
 o 
 
 o 
 
 o 
 
 IT) 
 
 m c *, ---. 
 
 » o 2^b 
 
 a £ 
 
 y O to 
 
 ro h U 
 
 ^ ^ is 
 
 —I -H 
 
 rt rt £ 
 
 £ $ ti 
 
 ao 
 
 o o 
 o o 
 
 o o 
 
 o ^ 
 
 o o 
 o o 
 
 o o 
 
 r-l O 
 
 o o 
 
 o o 
 
 O O ^f 
 
 o o o 
 
 o o o 
 
 r>_ o 
 
 ^H 00 
 
 oo r>- 
 
 ^ ^ vd o 
 
 o o 
 
 O C4 
 
 «-H LO 
 
 O ,— 1 ■-! O 
 
 o o o o 
 o o o o 
 
 o o o o 
 
 O 00 O 00 
 
 O >— i >— i o 
 
 o o o o 
 
 
 o o 
 
 o o o o 
 
 OOOO'-tOOOO-H 
 
 o o 
 o o 
 
 o o 
 
 O r-H 
 
 p o 
 
 00 o 
 
 o o 
 
 o o r- o 
 
 O O <NI O 
 
 O0 s OOOO'-<t^OO 
 
 oraoooo^oo—<o 
 
 o o 
 
 o o o o 
 
 oooooooooo 
 
 o o 
 
 o ^ 
 
 00 o 
 
 O O "St* o 
 
 o o 
 
 O vO 
 
 CO O 
 
 O O CO o 
 
 s^k 
 
 1 
 
 
 ) 
 
 ^^ 
 
 
 4-> —* 
 
 { 
 
 o ^ 
 
 ) 
 
 a m 
 
 j-l 
 
 g a 
 
 
 k 
 
 1 
 
 fll£. 
 
 ; 
 
 i 
 
 •H ^ . 
 M C 
 
 ! 
 
 «j s £ 
 
 
 
 
 w 
 
 
 
 h 
 
 0JD 
 
 o o 
 
 00 CJ 
 
 o o 
 
 <: 
 
 o o 
 
 co o 
 
 O >-! 
 
 o o 
 
 ■^ o 
 o on 
 
 o o 
 
 < 
 
 o ^ 
 
 o ^ 
 
 8« 
 
 h <1 
 o 
 
 <-H r-H 
 
 <-H i— 1 
 
 <+■! ■— 1 
 
 *H i— 1 
 
 
 
 
 
 
 
 
 
 u a 
 
 u a 
 
 u a 
 
 n rt 
 
 ti 
 
 o a 
 
 ctf 
 
 rt 
 
 Th ^ 
 
 Ti & 
 
 ^ a' 
 
 T! ^ 
 
 rd +-> 
 
 rJ3| +* 
 
 rd -M 
 
 si ■*-> 
 
 u w 
 
 u w 
 
 o w 
 
 u w 
 
 o o 
 
 ■^ ^ o o 
 
 oo ro O O 
 
 O O r-« O 
 
 a <^ 
 
 rt 
 
 I, 
 
 a u 
 
 
 
 
 OB 
 
 
 
 
 • H 
 
 Td 
 
 ti 
 
 T3 
 
 Oh 
 
 rt 
 
 r— 1 
 
 
 
 +-> 
 
 rt 
 
 i— ! 
 
 a 
 
 <D 
 
 i—i 
 
 0) 
 
 ,d 
 
 Q 
 
 u 
 
 p 
 
 U 
 
 O-^OOOO^sOsOO 
 O^OOOOOOOOO 
 
 o--<oooorJ"-ioo 
 
 iMOO<M00<NIOOPJC<l 
 
 HO^fHrOvO^inc') 
 
 orvioooorJoJ-HO 
 
 a 
 
 
 
 
 M 
 
 
 a, 
 
 2*3 a O 3 * £ 
 
 X< 
 
 o 
 o 
 
 I 
 
 O 
 
 i— i 
 
 ■3 
 
 H 
 
 rt 
 a 
 
 i— i 
 H 
 
 o 
 
S o w 
 
 Ph i — i • O 
 
 *""* -4-> .— I 
 
 h 4) o rj 
 
 d ^ CO <U k A 
 
 o ^ j: x d 
 
 M 0) m P A 
 
 O 
 O 
 
 
 82 
 
 to 5 v ao 
 
 o< £ Q « 
 
 ooooooooo 
 ooooooooo 
 
 ooooooooo 
 
 o 
 o 
 
 o en 
 o o 
 
 o o 
 
 o r- o -^ 
 
 O LD O CO 
 
 o o o o 
 
 . TO 
 
 H tf 
 
 £ «5 ti 
 
 00' - *' -, ^Oor^(Mr-< 
 
 oooooooooo 
 
 t^ (MOOOOOOfvJO 
 
 o •— i n Tt* m r» oo ^ \D 
 
 O OOCMOO'-'OO 
 
 w 
 
 ofl 
 
 u 
 
 TO 1 
 
 ■M 
 
 w 
 
 M 
 
 tuO 
 
 JJjj H £ OX) 
 
 oooooooooo o 
 oooooooooo o 
 
 ooooo'ooooo o 
 
 oooooooooo o 
 oooooooooo o 
 
 oooooooooo 
 
 NO^1ivO1iv000v0v0 
 
 oooooooooo 
 
 " flj M 
 
 oooo--<ocnio 
 oooooooooo 
 
 o'ooooooo 
 
 o cm o 
 
 O LD O 
 
 vO 
 
 O O O -X- 
 
 O 00 o 
 
 o o p 
 o ^ o 
 
 CD 
 CO 
 
 CD 
 
 Ph 
 CD 
 
 o 
 
 
 
 
 ft 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 CO 
 
 
 
 O 
 
 oo 
 
 o 
 
 +-> 
 
 +J o 
 
 xtf 
 
 fNj 
 
 ^ 
 
 o 
 
 o 
 
 CO 
 
 CO oo 
 
 xO 
 
 rO 
 
 o 
 
 r— ) 
 
 o 
 
 
 
 
 
 J o 
 
 ^° 
 
 o 
 
 CQ 
 
 ffl 
 
 s ^ 
 
 nJ £ FQ cti 
 
 £ <J ^ 
 
 •rH «-l -H 
 
 j-i m p-| ,_i j-j 
 
 L, 
 * « ti 
 
 g jg cj 
 
 2 <! 2 ^ § 8 « 
 
 A £ & 
 U ^ w 
 
 Su3^ a u 
 
 <! ffi £ w <J UJ 
 
 CM 
 
 CO 
 
 m 
 
 a 
 o 
 
 +j 
 
 pi 
 
 • H 
 
 .—I 
 
 u 
 
 a 
 
 o 
 ^i 
 m 
 
 CD 
 
 U 
 cti 
 
 co 
 CD 
 
 i— ( 
 > 
 
 Oh 
 
 
 +-> 
 ti 
 
 • H 
 
 i—l 
 
 u 
 
 i 
 
 CO 
 
CO g 
 
 ? 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X £ 
 
 . 
 
 
 
 t-H 
 
 
 
 
 
 
 
 00 
 
 
 
 
 un 
 
 — i Q) 
 
 o £ 
 
 
 
 4-> 
 
 CP 
 
 i 
 
 n 
 
 00 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 o 
 in 
 
 Lh CO 
 
 H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r-l 
 
 4J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 D WD 
 
 — i 
 
 W 
 
 i-H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u a 
 
 
 
 00 
 
 
 o 
 
 on 
 
 ^ 
 
 i-H 
 
 o 
 
 t^- 
 
 so r- 
 
 o 
 
 o o 
 
 O LD uT) sO 
 
 o <-* -^ -^ 
 
 en ? 
 
 — 1 
 
 a 
 
 
 o 
 
 m 
 
 o 
 
 O 
 
 o 
 
 s 
 
 00 lD 
 
 o 
 
 o o 
 
 o in vfl o 
 
 O I s - I s - P0 
 
 00 
 
 
 K 
 
 
 W 
 
 
 00 
 
 +J 
 
 a 
 
 u 
 
 •~». 
 
 Rj 
 
 00 
 
 +- 
 
 a 
 
 X 
 
 
 ±1 w 
 
 OJO 
 
 00 
 
 a 
 
 o o o o 
 
 o o o o 
 
 o o 
 
 o o •— ' •— • 
 
 o o o o 
 
 OvO^N«-iNHOcO^ 
 
 00 
 
 o 
 
 oooooor<iLo^t~~r-or-'^ 1 
 Lfi^orovOr-o^O'- tsootno 
 
 OOOO'-'O'-iOOO 
 
 OOtMOO'-H'-iOOO'-'OO 
 
 O^O^OOr-oOOO 
 
 oin— hoooooosoo 
 
 o 
 o 
 
 OOOO 1 — isD"— "OO^O^O^O 
 
 oooocoooooor-m^ 
 
 oooooooooo 
 
 OOOCONOO^^OO^ 
 OvD^MOOoOvOCN]^ 
 
 o 
 
 o 
 
 OOOOOO— lOO^HQOO 
 
 oooo^fo-spoooocOvOcM 
 ooooocoooor^i>-OLn 
 
 ocMoooo-^-^ooo 
 
 oooo^-^iriooLnoorvio 
 
 N^^OOOOOMO^O 
 "-icvJOOOOCMOOJOOO 
 
 
 n 
 
 QJ 
 CO 
 CD 
 
 Oh 
 
 0) 
 
 a 
 o 
 
 oooosoovOvo^oocaoo 
 r-o-^^c^oocMOooc^mo 
 
 oco— iooo^loco— • o^oooo^mLoooi-n'^coi-i 
 
 ^ s- ~ 
 
 U ^ W 
 
 ^ 
 
 T3 
 
 
 o 
 
 
 3 
 
 
 s < 
 
 
 & < - 
 
 cq 
 
 L, 
 
 h * d 
 
 i— t 
 O 
 
 hlo 
 ate 
 
 tha 
 
 w 
 
 U ^ w 
 
 IT) 
 
 o 
 
 +-> 
 
 a 
 
 •rH 
 
 r— 1 
 
 u 
 
 83 
 
 o 
 
 O vD 
 LD LT) 
 
 00 o 
 
 •—I LO 
 
 ^ o 
 
 (V3 
 
 
 TJ 
 
 'u 
 
 (i) 
 
 • f-1 
 C ) 
 
 TJ 
 
 <i 
 
 
 u 
 
 ■P 
 
 ■ H 
 
 o 
 
 IH 
 
 rri 
 
 
 ■M 
 
 
 0) 
 
 H 
 
 
 
 o 
 
 Q 
 
 S3 
 
84 
 The conditions of the tests as reported in Table 8 are as follows: 
 
 1. Concentration of organic extract in the test solution is 5.0 mg/1. 
 
 2. Dioxane and dilute potassium chloride were used initially to dis- 
 solve the extracts. 
 
 3. Iron was added as ferrous sulfate in increments as indi- 
 cated in Appendix B, 
 
 4. The pH was 8. at the time samples were filtered. 
 
 5. Five minutes was allowed for reaction time after iron was 
 added before the samples were filtered. 
 
 6. The volume filtered was 35 ml. 
 
 7. Membrane filters with a 0. 22u pore diameter were used. 
 
 8. The filter diameter was 7/8 inch. 
 
 9. Mixing and aeration was with a magnetic stirrer. 
 
 It was realized that if the extracts retained very much residual 
 iron through the extraction procedures, any effects, which might have 
 otherwise been attributed to the iron added during the filtration studies, 
 would be masked or at least reduced. Fortunately, however, when the 
 extracts were analyzed for iron, they were found to contain less than one 
 
 percent iron. 
 
 I 
 
 To illustrate the total effect of the extracts on holding iron in solu- 
 tion or suspension, the concentrations of the extracts, as found in the ground 
 
 j waters, are listed in Table 8. These concentrations are used to show the 
 
 concentrations of iron which could be held in a filterable condition with these 
 I 
 amounts of organic matter. Filtration studies, made using ferrous sulfate 
 
without adding any organic extracts, showed that approximately 0. 5 mg/1 
 of iron could pass through the filters without assistance from any organic 
 sequestering agents or colloids. It appears that when the iron was present 
 in low concentrations, the hydrated ferric oxide present exhibited little if 
 any tendency to flocculate and thus the particles were small enough to pass 
 through a 0. 22u membrane filter. It is again noted that ferrous iron deter- 
 minations made on the solution in the reaction vessel, always gave negative 
 results even though the determinations were made immediately after adding 
 iron as ferrous sulfate and allowing no more than a minute for mixing. 
 Therefore, the maximum amount of iron in the filtrates which could act- 
 ually be attributed to the organic extracts was approximately 0. 5 mg/1 
 less than the observed amounts. With, an adjustment of 0. 56 mg/1, many 
 of the extracts exhibited no apparent effect on the filterability of iron. 
 
 The effects of the various extracts from a particular water were 
 considered to be additive. This assumption was made as it appeared that 
 the various fractions obtained were not duplications but actually indicated 
 that additional portions of the organic matter present in the water were 
 being extracted with appropriate solvent adjustments. It is further noted 
 that the only extracts which exhibited any significant effects on the fil- 
 terability of iron were obtained with polar solvents and, as will be shown 
 later, were found to exhibit similar chemical characteristics. 
 
 In Appendix B, the actual logs of these experiments, the column 
 headed 'Iron on Filter" refers to the membrane filters used. A "-" signi- 
 fies that no iron precipitate could be detected on the filter. A "+" means 
 
86 
 
 that a barely detectable yellow precipitate could be seen on the filter. A 
 "+" indicates that a definite yellow to orange precipitate was present on 
 the filter. Finally a "++" designates a very heavy precipitate and that the 
 filter was actually starting to clog. In addition to listing the amounts of 
 iron in the test solutions, the increasing amounts of iron in the filtrates, 
 and the points where iron first appears on the filters; results of increased 
 times of reaction, variations in pH, changes in pore size of the filter paper, 
 and changes in the diameter of the filter paper are listed. 
 
 Ferric nitrate was added to several of the extracts in order to observe 
 the amounts of iron which could be held in solution if the iron was already in 
 the ferric state when it was added to the organic extracts. In most cases 
 the maximum amounts of iron found in the filtrates were less than the 
 amounts observed when ferrous sulfate was used as the iron source. This 
 is most probably because the ferric nitrate had had time to undergo suf- 
 ficient polymerization reactions so that the large ferric hydroxo complex ions 
 thus formed were less subject to being stabilized as colloids or entering into 
 chelation reactions with the organic acids present. Actually this condition 
 probably does not exist in ground waters as the iron is already in contact 
 with the organic matter before and when it is oxidized to the ferric state. 
 
 Perhaps the most revealing fact to be gleaned from a study of the 
 logs in Appendix B is that, no matter how much iron an extract was capable 
 of holding in a filterable condition, the addition of sufficient additional iron 
 
 ! 
 
 'would cause practically all the iron to be retained by the filters. If the re- 
 daction involved was chelation, it would be reasonable to expect the chelating 
 
87 
 agents to hold their maximum capacities of iron in solution no matter how 
 much additional iron was added. However, it appears that the organic 
 extracts are either negative colloids or are large molecules in solution 
 which possess sufficiently high negative charges through ionization and 
 Van der Waals forces to physically act as though they were colloids. Thus 
 when iron was first added and oxidized to the ferric state., it was drawn to 
 the negatively charged organic particles and held in colloidal, suspension 
 until sufficient iron was added to reduce. the zeta potential to a point where 
 the particles could flocculate. It is possible that iron floe first started to 
 appear on the filters only when the zeta potential had been reduced suffi- 
 ciently to allow some particles to grow sufficiently to be stopped by the 
 membrane filters being used. As more iron was added, coagulation and 
 flocculation became more efficient until almost 100 percent of the iron had 
 formed into particles large enough to be filtered from solution. This size 
 dependency is further demonstrated by the use of the 0. 45\i membrane fil- 
 ters instead of the 0. 22u membranes generally used. When membrane 
 filters with the larger pore diameter were used, more iron was required 
 to cause a precipitate to become visible and the maximum amount of iron 
 
 'which could be held in a filterable condition was always greater. 
 ! 
 
 It was always possible to filter a greater concentration of iron 
 through the two inch diameter filter than through the 7/8 inch diameter 
 filter. It appears reasonable that when the zeta potential had been reduced 
 sufficiently so that some particles were beginning to build to sufficient size 
 to be retained by the filter medium or pore diameter, the particles which 
 
88 
 lodged on the filter surface not only tended to partially block the filter pas- 
 sages but smaller particles in the solution being filtered were forced into 
 close proximity with the larger particles and the flocculation tendency was 
 increased. Thus under such conditions when flocculation was just begin- 
 ning, the smaller the surface area of the filter, the greater was the ten- 
 dency for particles to build up in a greater density and increase the 
 flocculation forces on the remainder of the solution. 
 
 When additional reaction time was allowed prior to filtration, very 
 little difference was noted in the filterability of the iron. This would seem 
 to indicate that the dispersions were formed fairly rapidly and their sta- 
 bility was not affected by time. 
 
 i 
 
 Changes in pH provided other evidence that the holding of iron in a 
 filterable condition was caused by colloidal dispersion. When iron began 
 to appear on the filter membranes, a higher concentration of iron was held 
 [in a filterable condition if the pH of the solution was raised from 8. to 9. 0. 
 1 the pH was lowered from 8. to 6. 0, a greater portion of the iron was 
 •etained by the filter. Thus it appears that as more hydroxyl ions became 
 ivailable to enter the double layer surrounding the colloids, the greater 
 ;>ecame the negative zeta potential. Although this is another demonstra- 
 
 ion of the possibility that the interference is colloidal, pH dependence does 
 |iot rule out the possibility of chelation as organic chelating agents can 
 
 ncrease and decrease in chelation capacity as ionization of carboxyl groups 
 
 t increasing pH values makes them more readily available as chelation 
 
 i. 
 ites. 
 
89 
 This work substantiates and further explains some of the observa- 
 tions made by Shapiro (60, 7 8, 79, 90) if it can be concluded that the extracts 
 obtained from ground waters were similar in character to those from the 
 surface waters studied by Shapiro. Although much more thorough character- 
 izations have been made of the surface water extracts, all results obtained 
 in characterizing the ground waters have been in agreement: with the surface 
 water analyses. These analyses will be discussed more fully in the next 
 section. Also the iron filteration studies seem to be in substantial agree- 
 ment at all points of comparison. 
 
 Shapiro found that most of the extracts were dialyzable through a 
 :ellophane membrane and thus not colloidal. He, however, found that, when 
 ron was added to his extracts, it was held in a filterable condition when fil- 
 er ed through 0. 45 and 0. 22 u membrane filters, but this iron was not fil- 
 erable through 0. 10u filters. Filter ability of iron with the ground water 
 extracts was not tested through 0. lOu filters but the other observations hold 
 rue. Shapiro suggested that the iron had been peptized in a colloidal state. 
 ;Ie further demonstrated this colloidal condition by adding electrolytes such 
 
 s potassium chloride and aluminum chloride and found that they were cap- 
 
 i 
 
 ble of destabilizing the iron colloids. Thus it would appear that the main- 
 
 I 
 
 -nance of large quantities of iron in suspension is caused by organic 
 
 impounds with the capacity to stabilize colloidal dispersions. When the 
 
 ltration studies were being made, it was noticed that 5 mg/l of the extracts 
 
 1 d not produce a color detectible by the human eye but the addition of as 
 
 i 
 
 ttle as 0. 2 mg/l of iron quickly produced a definite yellow color which 
 
90 
 
 increased in intensity as more iron was added. This color was more 
 intense than the color which would be caused by iron alone and appeared 
 similar to the color imparted to water from decaying vegetable matter. It 
 is thus quite possible that the humic acids in surface and ground waters have 
 the same source and are similar to the humic acids in soils. 
 
 D. Characterization of the Organic Extracts 
 
 One characterization of. the organic extracts was, of course, the 
 measurement of their ability to hold iron in a filterable condition. The 
 solvent system described by Mueller, et al. (92) to identify dicarboxylic 
 acids with ascending paper chromatographs was attempted with several of 
 the extracts. Three of the five extracts spotted were acidic enough to re- 
 act with methylene blue, which was used to develop the chromato graph, to 
 form yellow spots which indicated their final locations on the chromato- 
 graph. The "Rf" values of these extracts, or the ratio of movement of 
 the extracts to the movement of the solvent front, were compared with 
 several known acids. Only one of the extracts was found to move while 
 all the acids tested ascended the paper and produced distinct spots with 
 i the exception of oxalic acid which produced a streak. The chromatograph 
 is reproduced to scale in Figure 9. In Table 9 are tabulated the com- 
 pounds and their Rf values. 
 
91 
 
 «W"0»»— "»«»«^l«»"P"»W»^™"T*"W»P*»»»P»»pn»™"»PP^ 
 
 W*W!!!W"WF 
 
 ■ l^il^J^ifSfe^ l^ftf 
 
 
 pxoV ^IT^O 
 
 
 pioy dtj;tq 
 
 
 pioy OTT^l^V 
 
 <; . : > V - '• .' ':■-'■•■..■'• . 
 
 ptoy OT 110 !" 8 !^ 
 
 ■•*o'" 
 
 piDV Dtp'eq 
 
 
 
 pioy oiuToong 
 
 o 
 
 
 
 piDV OTXITDDng 
 
 V - 
 
 piDV 3XJ-e;j-ex a 
 
 i*"j.i*-V.'.i ^'\.i '.•'■/'• -:*'Z ■!•'■ * •■'.'■• 
 
 pioy oixv^a-vx a 
 
 • -.•-■■ ,'% ....... •-.',<•-- ■■':• i'-i.t'. 
 
 (g) uo^uno 
 
 .- 
 
 g leTuouiuiv 
 
 •'.«• ''•'*;v"---'-'-^ -X T:'; -'-'•'•- '^•" 
 
 \ (S) uo: » u HO 
 
 * 
 
 I 9 J91T2M 
 
 c 
 
 w (g) ucnuno 
 
 g uij:ojojoxxj3 
 
 
 1 (5) UOIUTTQ 
 
 ,*.*.*••-""•• ■"**" : !'/'»"'. **** "- : '""• 
 
 — v "BTuouiury 
 (Z) °TT^d 
 
 '•.'■■'•"■ : / *'■•.'* '.•>■*•. ■."■*'•'. "J "■.-. , *- '*■'*. 
 
 g Xou^t[i3 
 
 ",*-/.'* '■'•■.■' '•'., '--"-' **o--?--. v *; : -v.; •/.'-"■ 
 
 
 n 
 
 ■ 
 
 1 
 
 c 
 
 
 •o : 
 o> 
 
 a. 
 
 a. 
 < 
 
 '5 
 
 < 
 
 a 
 
 X 
 Q_ 
 < 
 
 a: 
 o 
 o 
 
 !* 
 
 o 
 
 at 
 
 o> x 
 
 o 
 
 LU 
 
 5 a: 
 
 o 
 
 LU 
 O 
 (J) 
 < 
 
 in ■! ■■ i ii i i r ■ : 
 
Philo 
 
 2 
 
 0. 03 
 
 Clinton 
 
 5 
 
 % 
 
 Clinton 
 
 5 
 
 •A* 
 
 '1* 
 
 Clinton 
 
 5 
 
 0. 02 
 
 Clinton 
 
 5 
 
 0. 13 
 0. 14 
 0. 14 
 0. 59 
 0. 61 
 0. 61 
 0. 54 
 0. 32 
 0. 23 
 
 (To top of 
 
 streak) 
 
 0. 31 
 
 92 
 TABLE 9 
 
 PAPER CHROMATOGRAPH OF ORGANIC ACIDS 
 
 Acid or Town Experiment Rf Values 
 
 Extract Number 
 
 Ethanol B 
 Ammonia A 
 Chloroform B 
 Water B 
 Ammonia B 
 D Tartaric Acid 
 D Tartaric Acid 
 Succinic Acid 
 Succinic Acid 
 Lactic Acid 
 Malonic Acid 
 Malic Acid 
 Citric Acid 
 Oxalic Acid 
 
 *Final location not visible on developed chromatography 
 
 This test indicates that three of the five extracts were of sufficient 
 
 acidity to be detectable in the determination. Also there is a possible indi- 
 
 :ation that tartaric acid might be present in the ethanol plus ammonia 
 
 j 
 
 extract from the acidified filter at Clinton. These extracts had possibly 
 
 iot been purified sufficiently to give satisfactory results with paper chroma- 
 tography. Even the possible presence of the acids as their metal salts 
 
 ;ould interfere with the results. Also there was no assurance that the 
 
 i 
 
 Reference acids used covered even a small portion of the acids which 
 
 ould possibly be present in the extracts. However, this brief attempt 
 
 oes demonstrate that paper chromatography can be a useful tool in the 
 
 lore complete analysis of organic acids present in ground water. 
 
 A more definite characterization of several of the extracts was 
 
93 
 made using infra-red spectroscopy. The spectra are reproduced in Appen- 
 dix C. The spectra are essentially the same whether the extracts were 
 scanned as smears on sodium chloride crystals or molded into potassium 
 bromide pellets. As the extracts had not undergone extensive purification 
 prior to making these analyses, several undefined peaks appear on several 
 of the spectra. Bellamy (101) has listed the wavenumbers in cm"* of the 
 carboxylic acids to have one or two peaks between 3500 and 2500 cm and 
 two other peaks in the region between 1725 and 1300 cm" . Spectra from 
 the Sadtler Research Laboratories (107) were used to determine the spec- 
 trum of distilled water which can mask spectra in samples which are not 
 
 - 1 
 completely dry. Major peaks are found at 3400 and 1650 cm . Thus these 
 
 peaks can mask the appearance of peaks from carboxylic acids. 
 
 Of major interest is the fact that most of the extracts show definite 
 
 peaks in the characteristic regions for carboxylic acids. This is in close 
 
 agreement with Shapiro (60, 78, 79, 90) and Rice, et al. (102) who found 
 
 organic acids in surface waters to have similar infra-red spectra. Frisch 
 
 ind Kunin (108) found the organic foulant, from anion-exchange resin used 
 
 o treat Delaware River water at Philadelphia, produced a similar spec- 
 
 rum indicating organic acids. They also found that tea and an infusion 
 
 riade from leaves produced a similar spectrum, indicating the probable 
 
 rigin of all these organic acids and that, these acids are "humic acids" or 
 
 ;ecomposition products from decaying vegetable matter. 
 
 Puri and Sarup (93) reported that humic acids usually show an equiva 
 
 mce, when titrated with a strong base, at about pH 7. 5. Also humic acids 
 
94 
 were reported to demonstrate titration curve equivalence points about 4 pH 
 units above the original pH values of the acids in distilled water. In Figure 
 10 are reproduced the titration curves for several of the extracts. All the 
 extracts obtained with polar solvents show just such characteristic curves 
 which further demonstrate the likelihood that these extracts are "humic 
 acids." Tartaric acid and the chloroform extract titrated did not show the 
 same leveling off of pH above the equivalence point as did the other extracts. 
 As the ionizable hydrogen ions, from both carboxyl groups in tartaric acid, 
 are both given off below pH 5, it and the chloroform extract, which is likely 
 to be highly non-polar, show no further effect on the titration curves and 
 the pH continues to rise just as it would if the strong base was added to dis- 
 tilled water. 
 
 In Appendix B are listed the initial pH values of the test solutions 
 used in the filtration study. These pH measurements were made on the 
 test solutions prior to pH adjustments or the addition of any iron. Although 
 they offer no quantitative information, the fairly close agreement of these 
 values seems to indicate that the functional groups of all the extracts have 
 :>een subjected to similar pH influences during the extraction procedures. 
 
 Using the method described by Middleton (81), attempts were made 
 o identify several of the extracts from Clinton Experiment 5 by solvent 
 separations. As had been expected, the polar water-soluble extracts 
 exhibited very little solubility in the non -polar solvent, ethyl ether. In 
 i-able 10 are given the results of these separations. 
 
95 
 
 
 fl 
 
 o 
 
 
 H 
 
 
 
 
 G 
 
 CO 
 
 ^•^ 
 
 
 
 
 an 
 
 O 
 
 
 a 
 
 o 
 
 
 
 a 
 
 oo 
 
 PQ 
 
 CO 
 
 
 
 o 
 
 CO 
 
 pq 
 
 •r-t 
 
 d 
 
 
 
 a 
 
 u 
 
 
 
 
 m 
 
 o 
 
 CO 
 
 cq 
 
 cq 
 d 
 
 
 
 a 
 a 
 < 
 
 So 
 
 a 
 
 m 
 
 r— 1 
 
 a 
 
 o 
 
 
 
 d 
 rd 
 X! 
 
 a 
 a 
 < 
 i 
 
 
 
 U 
 
 ■M 
 
 
 CO 
 
 u 
 
 0) 
 
 W 
 
 1 
 
 i 
 
 i 
 
 < 
 
 
 1 
 
 LT) 
 
 in 
 
 in 
 
 LfT 
 
 
 
 £ 
 
 , , 
 
 
 -— ' 
 
 — 
 
 — ' 
 
 ^H 
 
 TJ 
 
 ro 
 
 d 
 
 d 
 
 d 
 
 d 
 
 rd 
 
 -t-> 
 
 0) 
 
 
 o 
 
 
 
 
 
 
 
 U 
 
 i—i 
 
 n 
 
 
 •i-> 
 
 4-> 
 
 4-> 
 
 flj 
 
 • H 
 
 •H 
 
 d 
 
 d 
 
 d 
 
 ■H 
 
 d 
 
 •H 
 
 H 
 
 en 
 
 ,r| 
 
 
 i— 1 
 
 . — 1 
 
 r— 1 
 
 
 •H 
 
 Ph 
 
 U 
 
 u 
 
 u 
 
 u 
 
 Q 
 
 Q 
 
 sO 
 
 ^ m 
 o* Z 
 
 o 
 
 N £ 
 
 a o 
 
 o 
 
 d SU 
 
 o 
 
 © a 
 
 CO 
 O 
 
 < 
 cr 
 
 X 
 
 LU 
 
 O 
 
 z 
 < 
 o 
 cr 
 o 
 
 w o 
 
 CO 
 
 111 
 > 
 
 cr 
 
 3 
 O 
 
 2 
 
 g 
 cr 
 
 00 
 
 Hd 
 
 sO 
 
 rs] 
 
Ether 
 
 Soluble 
 
 Per cent 
 
 6. 
 
 25 
 
 0. 
 
 93 
 
 7. 
 
 45 
 
 67 
 
 .05 
 
 96 
 TABLE 10 
 SOLVENT SEPARATION OF ORGANIC EXTRACTS 
 Clinton - Experiment Number 5 
 Extract 
 
 Ammonia 
 Water B 
 Ammonia B 
 Chloroform B 
 
 The results of the solvent separation study , the results of the titra- 
 tion study, the tentative identification of carboxyl groups by infra-red 
 spectroscopy, and the fact that several of the extracts possessed sufficient 
 acidity to produce acid reactions on the paper chromatograph, provide a 
 reasonable indication that organic acids are present in the ground waters 
 studied and were obtained as extracts when polar solvents were used. The 
 extracts had a physical appearance similar to the extracts described by 
 Shapiro (60, 78, 79, 90) and Rice, et al. (102). The extracts were a deep 
 orange to brown color, very hard and brittle, but they chipped off easily 
 from the walls of the drying flasks and were ground to a dark brown powder 
 which gave off a characteristic burnt sugar odor. Shapiro (60) suggested 
 a molecular structure for the extracts from surface waters of (CcqH/ ?^27^^n 
 or C iH^zOi i. The average molecular weight was found to be 456 and the 
 equivalent weight was found to be exactly one -half the molecular weight or 
 228, indicating two ionizable acid groups per molecule. Although no attempt 
 has been made to determine the molecular weight of the extracts from ground 
 water, it is quite possible that the findings would be similar. 
 
97 
 
 E. Field Study of the Effects of Organic Extracts on Iron Removal 
 
 After the extracts had been obtained from the well, waters and had 
 been shown to interfere with the satisfactory filtration of iron in the labora- 
 tory, it was still necessary to see if these same extracts could be made to 
 cause an interference with filtration of iron from natural water. Although 
 the extracts had shown no retarding effect on the rate of oxidation of fer- 
 rous iron in the laboratory, the fate of ferrous iron was observed in the 
 field tests and compared to similar samples which did not contain organic 
 extracts. In addition to observing oxidation, a comparison was made of 
 the filterability of iron with and without the addition of organic extracts. 
 Field studies were made at Phil o, and Clinton, but no study was 
 made at Oakwood because a preliminary oxidation test showed the iron in 
 Oakwood well water to be completely oxidized within two to three minutes 
 after air was added. This time interval was too short to allow for adequate 
 observations. Studies were also made at Atwood with the results having 
 been previously reported by Marsh (43). For the waters studied, a sam- 
 ple was drawn without aeration. Dissolved oxygen (which was always 
 practically nil), ferrous iron, total iron, pH, and temperature were meas- 
 ured. At this point, if a particular organic extract was to be studied, it 
 was added, the sample was aerated, and then the rate of change of filter - 
 ability and the rate of ferrous iron oxidation were observed. 
 
 Figure 11 is a graphical, representation of the rate of oxidation of 
 i iron to the ferric state and the rate of change of the iron from a filterable 
 I to a nonfilterable condition. This study was made at Well No. 4 at Philo 
 
98 
 
 0. 6 
 
 0.0 
 
 10 20 30 40 
 
 Time in Minutes After Aeration 
 
 50 
 
 FIGURE II 
 IRON OXIDATION AND FILTRATION STUDY 
 AT PHILO, ILLINOIS 
 
99 
 because at the time of the study there was insufficient iron present in the 
 water from Well Number 3 to make the necessary observations possible. 
 Experiments were made using the ethanol plus ammonia extract from the 
 acidified filter used at Philo Well Number 3, using tartaric acid, and, as 
 a control, using a sample to which no extract had been added. A similar 
 graph for a study made at Clinton is given in Figure 12. In most tests 5 mg/1 
 of extract was added but in one of these tests 50 mg/l of extract was added 
 to see if an interference could be detected with this increased concentration. 
 The logs of these tests are given in Appendix D. All the studies produced 
 similar results. Neither the extracts nor tartaric acid produced any notice - 
 ible change in the normal oxidation rate of iron for these waters. This would 
 jeem to indicate that the presence of organic matter in water has little bear- 
 ng on the rate of oxidation of iron to the ferric state. Tartaric acid did 
 neasurably, but not greatly, retard the conversion of iron to a nonfilterable 
 :ondition, whereas the extracts exhibited no measurable influence on the fil- 
 erability of iron. Tartaric acid in the laboratory tests also showed a much 
 igher capacity for holding iron in a filterable condition. Thus, it appears 
 hat organic, matter in water could possibly have some influence on the rate 
 ! f change of iron from a filterable to a nonfilterable condition; however, 
 lis influence seems to be so slight as to be of no consequence for the con- 
 entrations of organic matter normally present in ground waters. 
 
 These tests were extended by March (43) to a study of all extracts 
 yailable for the water from Atwood and from Clinton. He noted that the 
 
 ! 
 
 aange from ferrous to ferric iron was more rapid than the change from 
 
100 
 
 2.4 
 
 2.0 
 
 1.6 
 
 s 
 
 1.2 
 
 0. 8 
 
 0.4 
 
 0. 
 
 
 
 
 No Extract Added 
 
 
 
 
 1. Filterable Iron 
 
 
 
 
 2. Ferrous Iron 
 
 
 
 
 Tartaric Acid (5 mg/1) 
 
 
 
 
 3. Filterable Iron 
 
 
 
 
 4. Ferrous Iron 
 
 
 
 
 Ammonia B (5 mg/ 1) 
 
 
 
 
 5. Filterable Iron 
 
 
 
 
 6. Ferrous Iron 
 
 
 
 
 Ammonia B (50 mg/ 1 
 
 
 
 
 7. Filterable Iron 
 
 
 
 
 8. Ferrous Iron 
 
 1 
 
 \ \ 
 
 V 
 
 \. 
 
 
 
 n 
 
 
 
 N. h~^~ 
 
 
 
 ^. 
 
 sȣ<r 4^ __ 
 
 10 20 30 40 
 
 Time in Minutes After Aeration 
 
 FIGURE 12 
 IRON OXIDATION AND FILTRATION STUDY 
 
 50 
 
 AT CLINTON, ILLINOIS 
 
101 
 a filterable to a. nonfilterable condition. This is the same phenomenon 
 observed in the laboratory study when a. gradual increase in the iron remain- 
 ing on the filter paper catalyzed the precipitation of the remaining iron. 
 The rates of oxidation of the iron were considerably higher than 
 those predicted by the equation proposed by Ghosh (42). The calculated time 
 for the oxidation of one -half of the iron was 78 minutes for Philo and 27 min- 
 utes for Clinton while in both cases more than one -half the iron was oxidized 
 in less than 10 minutes. It is to be noted, however, that the water at Philo 
 
 lad a much lower alkalinity than the range covered in the determinations 
 
 ■ 
 )y Ghosh, 
 
 T. Proposed Source of the Interference with Iron Removal, 
 
 None of the field studies revealed any interference with iron removal 
 >y oxidation and filtration. This occurred even though the extracts added 
 o the waters had shown the ability to retain iron in a filterable condition in 
 he laboratory, and these same waters had been previously reported to ex- 
 ibit inadequate iron removal with just the organic concentration present 
 a the natural waters. The laboratory determinations demonstrated that the 
 ormal concentrations of organic matter present in all the waters studied, 
 xcept the water at Philo, were sufficient to hold significant quantities of 
 "on in a filterable condition. Yet in none of these studies, the studies by 
 ;[arsh (43), or the studies by Longley (10) was it ever possible to filter a 
 ater through any type of pilot plant filter without obtaining at least 99 per- 
 snt iron removal. 
 
102 
 There seems to be only one explanation remaining for the interfer- 
 
 erence with iron removal for the waters studied. The waters, which were 
 
 ound to contain considerable concentrations of organic matter, were the 
 
 vaters which had exhibited the most unsatisfactory iron removal in actual 
 
 tlant operation. Thus it appears that the interference with iron removal 
 
 s due to the presence of organic, matter but, as previous discussion in 
 
 his study has demonstrated, the interference is apparently not interfer- 
 
 :nce with filtration by colloid stabilization or chelate formation. 
 
 Holluta and Eberhardt (109) in a study of iron filtration found that 
 
 .11 of the iron passing through a sand filter was in the ferrous state in 
 
 he filter effluent. Komolrit (49) , in a study of the efficiency of operation 
 
 i iron removal plants for several Illinois ground waters, found that, when 
 
 here was iron in a finished water, the major portion of the iron passing 
 
 hrough the filter was in the ferrous state as it left the filter. It seems 
 
 hen that the problem is not one of chelation or colloidal dispersion but 
 
 >ne of reconversion of the iron back into the soluble ferrous form. 
 
 Komolrit also observed that the plants experiencing trouble with 
 
 ron removal were also showing much less dissolved oxygen in the finished 
 
 r ater than in the water as it was applied to the filters. The waters in 
 
 Tinois which were difficult to treat also were observed to present problems 
 
 ith persistent slime growths on top of the filter media. This then sug- 
 
 ests that the interference is actually biological reduction of the iron as 
 
 part of the metabolism of bacteria or other simple organisms which are 
 
 sing the humic acids present as a food source. This prevalence of slime 
 
103 
 
 md the reducing conditions present, as indicated by the reduction of iron 
 ind dissolved iron concentration, seem to tie together. In the Netherlands 
 26) it was found that by using dry filtration, i. e. holding the water sur- 
 ace always just below the surface of the filter sand, iron filtration was 
 ilways satisfactory. This and Mathews ' findings (1.6) that break-point 
 :hlorination insures satisfactory iron removal both indicate the possibil- 
 ty that the prevention of the formation of a slime growth can insure satis - 
 actory iron removal. 
 
104 
 V. CONCLUSIONS 
 
 1. Carbon filters can satisfactorily be used to extract organic mat- 
 ter from ground water. However, in order to insure adequate extraction 
 of particular fractions of interest, proper solvents and pH must be used. 
 The three inch diameter by eighteen inch standard carbon filter will pro- 
 duce satisfactory reproducible results at an application rate of one gallon 
 per minute or at about four times the application rate normally recom- 
 mended. Two filters in series, with the water acidified before it is applied 
 to the second filter, will produce at least twice the extract as the unacidified 
 filter alone. Although the carbon filter cannot be expected to provide an abso 
 lute measure of the quantity of organic matter present in a well water, the 
 filter does provide for the extraction of meaningful quantities. 
 
 2. Organic extracts can be obtained from ground waters in a condi- 
 tion reasonably separated from the inorganic constituents present in the 
 water by the carbon filter technique. 
 
 3. At least some ground waters contain significant quantities of 
 organic matter. Most of this organic matter is soluble in polar solvents. 
 The extracts, obtained from the Illinois ground waters studied, exhibited 
 :olor, odor, titration, and infra-red spectrum characteristics of the humic 
 icids or acids formed during the decomposition of, vegetable matter. 
 
 4. Most of these extracts can hold iron in a filterable condition in 
 
 aboratory tests in alkaline conditions but upon the addition of sufficient 
 
 i 
 
 .excess iron, all the iron is retained by filtration. This indicates that the 
 
105 
 
 i 
 
 Organic matter creates colloidal dispersions and not soluble chelates. 
 
 5. Although the extracts obtained from the waters under consider- 
 ation were able to maintain the iron in a filterable condition in laboratory 
 
 ests, field tests were never able to reproduce this phenomenon. It is 
 hypothesized that the actual interference with iron removal, is biological 
 ictivity, supported by the humic acids in the well water, creating reducing 
 :onditions in a slime layer at the surface of or within the filters. ^These 
 •educing conditions may reconvert the iron to the soluble ferrous form. 
 
 6. The organic extracts obtained did not interfere with oxidation 
 )f ferrous iron to the ferric state. 
 
 7. The organic acids found in the ground water are similar to the 
 mmic acids found in soils and in surface waters. 
 
 8. The ground waters considered in this study, which demon- 
 strated difficulties with iron removal, were all found to contain significant 
 imounts of organic matter. This, and the fact that organic extracts from 
 hese waters can hold iron in a filterable condition at least in laboratory 
 studies, indicates the possibility that organic acids are at least partially 
 Responsible for the solution and transport of iron in ground water. 
 
106 
 VI. AREAS OF FUTURE STUDY 
 
 1. A study should be made of the bacteria present in rapid sand 
 ilters in an attempt to demonstrate their ability to cause reduction of iron 
 n filters. 
 
 2. If this is found to be the case, then a search should be made for 
 satisfactory method to control slime growth. 
 
 3. A study should be undertaken to further classify and actually 
 dentify the various organic compounds present in ground water. This 
 hould be undertaken, not only from the standpoint of interference with 
 ron removal, but also to establish methodology for monitoring ground 
 rater for possible pollution by intrusion of such things as insecticides and 
 ther industrial pollutants. 
 
 4. This study should be extended to other areas of the United States 
 ) see if similar correlations exist generally in all areas with iron removal 
 roblems. 
 
107 
 VII. BIBLIOGRAPHY 
 
 1. Weart, J. G. and Margrave, G. E. , "Oxidation -Reduction Potential 
 Measurements Applied to Iron Removal, " Jour. AWWA, 49, 9, 1223, 
 (1957). 
 
 Hauer, G. E. , "Iron and Carbon Dioxide Removal," Jour. AWWA, 
 42, 6, 555, (1950). 
 
 3. , "Public Health Service Drinking Water Standards," U. S. P. 
 
 H. S. , (1962). 
 
 4. Mohler, H. , "Iron and Manganese as Disturbing Factors in Ground 
 Water," Sweiz. Ver. Gas-u. Wasserfach Mona.ts - Bui. (Swiss), 31, 
 293, (1951). Abs. Jour. AWWA, 44, 10, P and R 42, (1952). 
 
 5. Weston, R. S. , "Some Recent Experiences in the Deferrization and 
 Demanganization of Water, " Jour. NEWWA, 28, 1, 27, (1914). 
 
 6. Amsbary, F. C. , "Further Development of Iron Removal Plant and 
 Storage," Jour. AWWA, 3, 6, 400, (1916). 
 
 7. Applebaum, S. B. and Bretschgen, M. E. , "Removal of Iron and Man- 
 ganese from Water," Ind. and Engr. Chem. , 26, 9, 25, (1934). 
 
 8. Babcock, R. H. , "Iron and Manganese in Water Supplies and Methods 
 of Removal, " W. and Sew. Wks. , 98, 10, 442, (1951). 
 
 9. Fosnot, H. R. , "Seven Methods of Iron Removal, " Pub. Wks., 86, 11, 
 81, (1955). 
 
 0. Longley, J. M. , "The Removal of Iron from Water by Aeration and 
 Filtration," Thesis, U. of 111. , (1961). 
 
 1. Van der Wal, S. S. , "Removal of Iron from Ground Water by Adsorp- 
 tion, " Water (Neth. ), 36, 1, (.1952). Abs. Jour. AWWA, 44, 10, 
 42 P. and R. , (1952). 
 
 I. Willcomb, G. E. , "Iron and Manganese in Water , " Jour. AWWA, 28, 
 12, 1896, (1936). 
 
 3. Fulkman, J. A. , and Taylor, E. J. , "Iron Removal Plant at Selma, 
 Alabama, " Jour. AWWA, 17. 1, 69, (1927). 
 
108 
 
 4. Craig, E. C. , Bean, E. L. , and Sawyer, R. W. , "Iron and Lime in 
 Removal of Manganese, " Jour. AWWA, 24, 11, 1762, (1932). 
 
 .5. Babcock, R. H. , "Iron and Manganese Removal, in Spaulding Precipi- 
 tator," Jour. NEWWA, 64, 2, 138, (1950). 
 
 [6. Mathews, E. R. , "Iron and Manganese Removal by Free Residual 
 Chlorine," Jour. AWWA, 39, 7, 680, (1947). 
 
 .7. Ley, G. , "A Unique Iron Removal Plant, " Jour. AWWA, 30, 9, 1493, 
 (1938). 
 
 8. Erickson, D. L. , "The New Iron and Manganese Removal Plant for 
 Lincoln, Nebraska," Jour. AWWA, 27 , 3, 337, (1935). 
 
 9. Charles, R. S. "Treatment of Subsurface Supplies for Removal of 
 Iron," Canadian Engr. 74, x, 34, (1938). 
 
 ,0. Kirchoffer, W. E. , "The Use of Beds of Manganese Ore in Iron and 
 Manganese Removal, " Jour. AWWA, 28, 10, 1488, (1936). 
 
 ,1. Moore, E. W. and Snow, E. A. , "Studies on Removal of Iron and 
 Manganese from Water, " Jour. NEWWA, 56, 3, 320, (1942). 
 
 2. Humphrey, S. B. , and Eikleberry, M. A. , "Iron and Manganese 
 Removal Using Potassium Permanganate, " 53rd An. Conf. , 111. 
 Sect. AWWA, Chicago, (1962). 
 
 3. Davy, P. S. , "Red Water and Its Prevention, " Jour . AWWA, 45, 1, 
 10, (1953). 
 
 4. Will, E. G. , "Report on Effect of Iron on High-Capacity Zeolite, " 
 Jour. AWWA, 48, 6, 739, (1956). 
 
 5. Weston, R. S. , "The Purification of Ground Water Containing 
 Iron and Manganese, " Trans. ASCE, 64, 112, (1909). 
 
 6. j "Aeration and Deferrisation, " International W. Sup. Cong. 
 
 and Exhib. , London, (1955). 
 
 7. Frisk, P. W. , "Control, of Iron and Manganese Accumulated in Sand 
 Filters," W. Wks. Engr., 86, 26, 1307, (1933). 
 
 3. McCracken, R. A. , "Study of Color and Iron Removal, by Means of 
 Pilot Plants at Amesbury," Jour. NEWWA, 75, 2, 102, (1961). 
 
109 
 
 29. McNamee, R. L. , "The Removal of Iron from Hard Ground Waters, " 
 Jour. AWWA, 21, 6, 758, (1929). 
 
 30. Cox, C. R. , "Laboratory Control of Water Purification, " Case - 
 Shepherd Mann Pub. Co. , New York, (1946). 
 
 31. Gilcreas, F. W. , "Treating Water to Eliminate Iron and Manganese 
 Troubles, Part XI, " W. Wks. Engr. , 109, 9, 832, (1956). 
 
 32. Willcomb, G. E. , "Coagulation of Soft Water with Iron Salts," Jour. 
 AWWA, 31, 10, 1703, (1939). 
 
 33. Saunders, D. J., "Iron Removal and Water Softening," W. and Sew. 
 Wks., 98, 3, 112, (1951). 
 
 34. Adams, B. A. , "Certain Complex Combinations of Iron and Their 
 Removal from Water, " W. and W. Engr., 33, 249, (1931). 
 
 35. Leclerc, E. , and Beaujean, P. , "Research Relating to Iron Com- 
 plexes," Bull. Mens. Centre Beige Et. Document Eaux, 56, 170, 
 (1955). Abs. WPA, 1069, (1958). 
 
 36. Beneden, G. V., Bull. Centre Beige Etude et Document (Liege), 
 43, 49, (1959). Abs. Jour. AWWA, 52, 7, 64 P and R, (I960). 
 
 37. Boorsma, H. J. , "Les Principes de la Deferrisation, de la Deman- 
 ganisation et de la Nitrification dans la Pratique de L'epuration des 
 Eaux Souterraines, " XXVII Congfes de Chimie Industrielle a 
 Bruxelles, I, 340, (1954). 
 
 38. McCrea, T. R. , "The Removal of Organic Bound Iron from Highly 
 Colored Water, " Jour. AWWA, 25, 7, 931, (1933). 
 
 39. Nordell, E. , "Municipal Iron and Manganese Removal, " Pub. Wks. , 
 87, 3, 88, (1956). 
 
 40. Schoap, W. B. , "Electrochemical Studies of Some Coordination 
 Compounds of Iron, " PhD Thesis, U. of 111. , (1950). 
 
 41. Engelbrecht, R. S., Margrave, G. E. , Longley, J. M. , Robinson, L. R. , 
 and Weart, J. G. , "Fundamental Factors in Treatment of Iron Bearing 
 Waters," San. Engr. Ser. No. 9, U. of 111. , (1961). 
 
 42. Ghosh, M. M. , "A Study of the Rate of Oxidation of Iron in Aerated 
 Groundwaters," San. Engr. Ser. No. 12, U. of 111. , (1962). 
 
110 
 
 43. Marsh, B. A. , "The Effect of Organics Extracted from Gro\ind Water 
 on the Rate of Oxidation of Iron, " San. Engr. Ser. No. 13, U. of 111. , 
 (1962). 
 
 44. Hem. J. D. , "Study and Interpretation of the Chemical Character- 
 istics of Natural Water, " Geol. Surv. W. -Sup. Paper 1473, 
 
 U. S. Gov. Printing Off . , (1959). 
 
 45. Weiser, H. B. , "The Hydrous Oxides, " McGraw-Hill, New York, 
 (1926). 
 
 46. Baas Becking, L. G. M. , Kaplan, I. R. , and Moore, D. , "Limits 
 of the Natural. Environment in Terms of pH and Oxidation -Reduction 
 Potentials," Jour, of Geol. , 68, 3, 243, (I960). 
 
 47. Schoeller, M. , "La Solubilite Du Fer Dans Les Eaux Souterraines, " 
 Annales de L'Institut d'Hydrologie et de Climatologie, XXVI, 1, 
 
 78, (1955). 
 
 48. Hem, J. D. , "Stability Field Diagrams as Aids in Iron Chemistry 
 Studies," Jour. AWWA, 53, 2, 211, (1961). 
 
 49. Komolrit, K. , "Measurement, of Redox Potential, and Determination 
 of Ferrous Iron in Ground Waters,". San. Engr. Ser. No. 11, 
 
 U. of 111. , (1962). 
 
 50. Hem, J. D. , and Cropper, W. H. , "Survey of Ferrous-Ferric 
 Chemical Equilibria and Redox Potentials," Geol. Surv. W, -Sup. 
 Paper 1459-A, U. S. Gov. Printing Off. , (1959). 
 
 51. Stumm, W. , and Lee, G. F. "The Chemistry of Aqueous Iron," 
 Schweiz. Zeitschr. , Hydrology, 22, 295, (I960). 
 
 52. Stumm, W. , and Morgan, J. J., "Chemical Aspects of Coagulation, " 
 Jour. AWWA, 54, 8, 971, (1962). 
 
 ,53. Cotton, F. A. , and Wilkinson, G. „, "Advanced Inorganic Chemistry, " 
 Interscience Publishers, New York, (1962). 
 
 |54. Riddick, T. M. , "Zeta Potential and Its Application to Difficult 
 Waters," Jour. AWWA, 53, 8, 1007, (1961). 
 
 55. Ware, J. M. , "The Chemistry of the Colloidal State," John Wiley 
 and Sons, New York, (1930). 
 
Ill 
 
 56. Weiser, H. B. , "Colloid Chemistry," John Wiley and Sons, New York, 
 (1939). 
 
 57. Mysels, K. J. , "Introduction to Colloid Chemistry, " Interscience 
 Publishers, New York, (1959). 
 
 58. Proceedings Rudolfs Research Conference, "Principles of Colloidal 
 Behavior and their Application to Water Sanitation, " Rutgers U., (I960) 
 
 59. Riddick, T. M. , "Zeta Potential: New Tool for Water Treatment, 
 Parti," Chem. Engr. 68, 13, 121, (1961). 
 
 bO. Shapiro, J. , "Chemical and Biological Studies on the Yellow Organic 
 Acids of Lake Water," Limnology and Oceanography, II, 3, 161, 
 (1957). 
 
 ol. Langelier, W. F. , Ludwig, H. F. , and Ludwig, R. G. , "Flocculation 
 Phenomena in Turbid Water Clarification, " Proc. ASCE, 78, Sep. 
 No. 118. (1952). 
 
 )2. Riddick, T. M. , "Zeta Potential: New Tool for Water Treatment, 
 Part II," Chem. Engr., 68, 14, 141, (1961). 
 
 >3. Stanley, P. M. , "The Zeta Potential of Clay Colloids and Their 
 Coagulation in Water Treatment J' MS Thesis, NMSU, (1962). 
 
 )4. Black, A. P., Ball, A. I., Black, A. L. , Boudet, R. A., and 
 
 Campbell, T. N. , "Effectiveness of Polyelectrolyte Coagulant Aids 
 in Turbidity Removal," Jour. AWWA, 51, 2, 247, (1959). 
 
 i5. Black, A. P. , "Current Research on Coagulation, " Jour. AWWA, 
 
 51, 12, 1545, (1959). 
 
 >6. Black, A. P., "Basic Mechanisms of Coagulation," Jour. AWWA, 
 
 52, 4, 492, (I960). 
 
 »7., Fair, G. M. , and Geyer, J. C. , "Water Supply and Waste-Water 
 Disposal," John Wiley and Sons, New York, (1954). 
 
 ;>8. Black, A. P. , "Some applications of the Principles of Colloidal 
 Behavior to Water Treatment, " Proc. Rudolfs Research Conf, , 
 106, Rutgers U. , (I960). 
 
 I 
 9. Zaides, A. , "Interaction of Trivalent Iron Salts with Organic Acids, " 
 
 Jour. Gen. Chem. (USSR), 5, 1530, (1935). Abs. Chem. Abs., 30, 
 
 21?3 6 , (1936). 
 
112 
 
 70. Martell, A. E. , and Calvin, M. , "Chemistry of the Metal Chelate 
 Compounds," Prentice-Hall, New York, (1952). 
 
 71. Chaberek, S. and Martell, A. E. , "Or ganic Sequestering Agents, " 
 John Wiley and Sons, New York, (1959). 
 
 72. Wallace, A. , "A Decade of Synthetic Chelating Agents in Inorganic 
 Plant Nutrition, " Arthur Wallace, Los Angeles, (1962). 
 
 73. . "New Method Gives Components in Soil's Organic Matter," 
 Chem. and Engr. News, 39, 50, 49, (1961). 
 
 74. Walther, H. C. , Jr. , "Some Physico- Chemical. Properties of the 
 Humic Acids, " PhD. Thesis, U. of Minn. , (1959). Dissertation Abs. , 
 Mar. -June, (I960). 
 
 75. Dekock, P. C. , "Influence of Humic Acids in Plant Growth, " Science, 
 121, 473, (1955). 
 
 76. Muller, J. F. , "Some Observations on Base Exchange in Organic 
 Materials," Soil Sc. , 35, 229, (1933). 
 
 77. Waksman, S. A. and Iyer, K. R. N. , "Contributions to Our Knowledge 
 of the Chemical Nature and Origin of Humus, III, The Base Exchange 
 Capacity of Synthesized Humus (Ligno -Protein) and of Natural Humus 
 Complexes," Soil Sc. , 36, 57, (1933). 
 
 78. Shapiro, J., "Yellow Acid - Cation Complexes in Lake Water," 
 Science, 127, 3300, 702, (1958). 
 
 79. Shapiro, J. , "Natural Coloring Substances of Water and their Relation 
 to Inorganic Compounds," Mtg. Div. of W. and Waste Chem. , ACS, 
 Los Angeles, (1963). 
 
 30. Middleton, F. M. , Braus, H. , and Ruckhoft, C. C. , "Fundamental 
 Studies of Taste and Odor in Water Supplies," Jour. AWWA, 44, 6, 
 538, (1952). 
 
 81. Middleton, F. M. , Rosen, A. A., and Burttschell, R. IL , "Manual 
 for Recovery and Identification of Organic Chemicals in Water, " 
 USPHS, Robt. A. Taft San. Engr. Ctr . , Cincinnati, (1957). 
 
 '32. Middleton, F. M. , Rosen, A. A., and Burttschell, R. H. , "Taste 
 and Odor Research Tools for Water Utilities," Jour. AWWA, 50, 
 1, 21, (1958). 
 
113 
 
 83. Myrick, N. and Ryckman, D. W» , "Bio -Oxidation of Extracts of 
 Organic Chemical Pollutants," Proc. 16th Ind. Waste Conf. , Purdue 
 U., (1961). 
 
 84. Sproul, O. J. and Ryckman, D. W. , "The Significance of Trace 
 Organics in Water Pollution, " Jour. WPCF, 33, 11, 1189, (1961). 
 
 85. Atkins, P. F. , Jr. and Tomlinson, H. D. , "Evaluation of Daily 
 Variations of Carbon Chloroform Extracts with the Presently Used 
 Two-Week Carbon Adsorption Method, " Proc. 17th Ind. Waste Conf. , 
 Purdue U. , (1962). 
 
 86. Spicher, R. G. and Skrinde, R. T. , "Chemical Oxidation of 
 Organic Contaminants in Water Supplies," Presented before 83rd 
 Ann. Conf. AWWA, Kansas City, (.1963). 
 
 87. Hyndshaw, A. Y, , Laughlin, H. F. , Coleman, D. C. , and Filicky, 
 J. G. , "Factors Influencing the Efficiency of Activated Carbon, " 
 Jour. NEWWA, 66, 1, 36, (1952). .. i 
 
 88. ___, "The Isolation of Organic Matter, " Water Res. Assoc. 
 
 Ann. Rept. , Research Sta. , Medmenham, Buckinghamshire, 
 England, (1960-61). ■ ' 
 
 89. Bikerman, J. J., "Surface Chemistry," Academic Press, New 
 York, (1958). 
 
 90. Shapiro, J. , "Freezing-Out - A Safe Technique for the Concentra- 
 tion of Dilute Solutions, " Science, 133, 2063, (1961). 
 
 91. Mueller, H. F. , Larson, T. E. , and Lennarz, W„ J. , "Chromato- 
 graphic Identification and Determination of Organic Acids in Water, " 
 Circ. 69, 111. State W. Surv. , (1958). Anal. Chem. , 30, 41, (1958). 
 
 92. Mueller, H. F. , Larson, T. E. , and Ferretti, M. , "Chromato- 
 graphic Separation and Identification of Organic Acids, " Circular 
 80, 111. State W. Surv., (I960). Anal. Chem., 32, 687, (I960). 
 
 93. Puri, A. N. and Sarup, A. , "Studies in Soil Humus: II. Potentiometric 
 Study of the Function of Humic Acid and Humates, " Soil Sci. , 45, 165, 
 (1938). 
 
 194. Beckwith, R. S. , "Titration Curves of Soil Organic Matter, " Nature, 
 184, 4687, 745, (1959). 
 
 95. Khanna, S. S. and Stevenson, F. J. , "Metallo-Organic Complexes in 
 Soil: I. Potentiometric Titration of Some Soil Organic Matter Isolates 
 in the Presence of Transition Metals, " U. of 111. , (1961). 
 
114 
 
 96. Meites, L. , "Polarographic Techniques," Interscience Publishers , 
 New York, (1955). 
 
 ,97. Kolthoff, I. M. and Lingane, J. J., "Polar ogr aphy , " (2 Volumes), 
 Interscience Publishers, New York, (1952). 
 
 98. Higuchi, Takeru, Hill, N. C. , and Corcoran, G. B. , "Chromato- 
 graphic Separation and Determination of Dicarboxylic Acids, C . to Ciq, " 
 Anal. Chem. , 24, 3, 491, (1952). 
 
 99. Buch,. M. L. , Montgomery, R. , and Porter, W. L. , "Identification of 
 Organic Acids on Paper Chromato grams, " Anal. Chem. , 24, 3, 489, 
 (1952). 
 
 )0. Lamar, W. L. and Goerlitz, D. F. , "Characterization of Carboxylic 
 Acids in Unpolluted Streams by Gas Chromatography, " Jour. AWWA, 
 55, 6, 797, (1963). 
 
 31. Bellamy, L. J. , "The Infra-red Spectra of Complex Molecules," 
 Methuen and Co. , London, (1958). 
 
 ;32. Rice, J. K. , Simon, D. E. , and Rice, R. C. , "Determination of 
 
 Naturally Occurring Organic Acids in Surface Waters," Ind. W. and 
 Waste, 7, 3, 75, (1962). 
 
 33. , "Standard Methods for the Examination of Water and Waste- 
 water," Eleventh Edit. , APHA, New York, (I960). 
 
 ■ - 
 
 34. Prickhardt, W. P., Oemler, A. N. , and Mitchell, J., Jr., "Deter- 
 mination of Total Carbon in Organic Materials by Wet-Dry Combustion 
 Method," Anal. Chem., 27, 11, 1784, (1955). 
 
 05. Puri, A. N. and Sarup, A. , "Studies in Soil Humus: I. Estimation 
 of Soil. Humus by Oxidation with Alkaline Permanganate, " Soil Sc. , 
 44, 323, (1937). 
 
 j06. Myrick, H. N. and Ryckman, D. W. , "Considerations in the Isolation 
 and Measurement of Organic Refractories in Water," Jour. AWWA, 
 55, 6, 783, (1963). 
 
 07. , "The 25 Most Useful Spectra for the IR Lab," Sadtler 
 Research Laboratories, Philadelphia. 
 
 08. Frisch, N. W. and Kunin, R. , "Organic Fouling of Anion -Exchange 
 Resins," Jour. AWWA, 52, 7, 875, (I960). 
 
115 
 
 0, Holluta, J. and Eberhardt, M. , "Uber geschlossene Enteisenund 
 durch Schnellfiltration, " VomWasser, (German), 24, 79, (1957). 
 
 1 , Lee, G. F. and Stumm, W. , "Determination of Ferrous Iron in the 
 Presence of Ferric Iron with Bathophenanthroline, " Jour. AWWA, 
 52, 12, 1567, (I960). 
 
116 
 APPENDIX A 
 
 CHEMICAL ANALYSES 
 
 . Total Iron 
 
 Total iron was determined by the Phenanthroline* Method described 
 n Standard Methods (103). A Beckman Model DU Spectrometer was used 
 o measure transmittance at 512 xn[X with a 1. cm light path. To insure 
 hat a sufficient amount of 1, 10 -phenanthroline was present in samples 
 nth. high iron concentrations, 10 ml of the phananthroline solution was 
 dded to the samples instead of the recommended 2. ml. When measur- 
 ng percent iron in the dry extracts, the samples were ashed as per 
 itandard Methods prior to color development. 
 
 . Ferrous Iron 
 
 The method described by Lee and Stumm (110) of measuring fer- 
 ous iron using bathophenanthroline was used for this determination, 
 'he method is as follows: 
 a. Reagents 
 
 (1). Bathophenanthroline*, 0.001 M solution. The solution 
 
 was prepared by dissolving 0. 0332 gm of 4, 7-diphenyL-l s 
 10 -phenanthroline in 50 ml of reagent grade absolute 
 ethanol and diluting to 100 ml with distilled water. 
 
 Both Phenanthroline and Bathophenanthroline are products of the 
 G. Frederich Smith Chemical Company, Columbus, Ohio. 
 
117 
 
 (2). Extracting alcohol. Reagent grade isoamyl alcohol 
 
 was used. 
 (3). Diluting alcohol. Reagent grade absolute ethanol was 
 
 used. 
 (4). Sodium acetate, 10 percent solution. The solution was 
 
 prepared by dissolving 10 gm of sodium acetate in 
 
 100 ml of distilled water. 
 (5). Standard iron solution. A solution made of 20 ml of 
 
 concentrated sulfuric acid added to 50 ml of distilled 
 
 water was used to dissolve 0.7022 gm of ferrous 
 
 ammonium sulfate (Fe(NH 4 ) 2 (S0 4 ) 2 « 6H 2 0). This 
 
 was then diluted to 1000 ml. Appropriate dilutions 
 
 were made to establish a standard curve, 
 b. Procedure for the determination of ferrous iron 
 
 (1). A 5 to 15 ml sample was pipetted into a 125 ml separ- 
 
 atory funnel. 
 (2). Then 4 ml of sodium acetate solution was added to 
 
 the sample. 
 (3). This was followed by 15 ml of bathophenanthroline 
 
 after which the sample was swirled gently. 
 (4). Then 10 ml of the isoamyl alcohol was added. The 
 
 funnel was shaken vigorously and let stand. 
 (5). After the phases had separated, the lower aqueous 
 
 layer was discarded and the colored upper fraction 
 
118 
 
 was transferred to a 50 ml volumetric flask. 
 (6). The separatory funnel was washed thoroughly with 
 
 absolute ethanol and the washings were transferred 
 
 to the volumetric flask. The sample was then 
 
 diluted to 50 ml with absolute ethanol. 
 (7). A reagent blank was prepared using 10 ml of distilled 
 
 water instead of a sample. 
 (8). The intensity of color developed was measured with 
 
 a Beckman Model DU Spectrometer at a wavelength 
 
 of 533 mjj, and using a 1. cm light path. 
 
 3. Dissolved Oxygen 
 
 Dissolved oxygen was measured by the Alsterberg (Azide) Modifi- 
 cation of the Winkler Method as described in Standard Methods (103). 
 
 k. Hardness 
 
 Hardness was determined by titration with ethylenediamine 
 tetraacetic acid (EDTA) as per Standard Methods (103). However, instead 
 
 L ■ . 
 
 ,of using the reagents and inhibitors described, the prepared reagents 
 CalVer II and UniVer II* were used. UniVer II was used for determining 
 total hardness and CalVer II and 1 ml of 8 N potassium hydroxide were 
 !ased for determining calcium hardness. 
 
 ;: Both CalVer II and UniVer II are products of the Hach Chemical 
 Company, Ames, Iowa. 
 
119 
 
 . Chloride 
 
 The mercuric nitrate method described in Standard Methods (103) 
 Vas used for determining chloride concentrations. 
 
 . Alkalinity- 
 Alkalinity was determined by titration with, dilute sulfuric acid as 
 escribed in Standard Methods (103). The end points of the alkalinity 
 itrations were detected with pH meters. The pH meters were standard- 
 zed with pH 4. 01 buffer made from Beckman buffer powders*. 
 
 . Chemical Oxygen Demand 
 
 Chemical oxygen demand (COD) was measured in accordance with 
 tandard Methods (103). For all natural waters the 0. 1 dilution method 
 r as followed and for the extracts the undiluted method was used. Silver 
 ulfate was used as a catalyst in all cases. 
 
 Beckman Instruments, Inc., Fullerton, California. 
 

 
 
 
 
 
 120 
 
 
 
 APPENDIX B 
 
 
 
 
 
 LOG OF FILTER ABILITY 
 
 STUDY 
 
 
 
 Extract 
 
 Initial 
 pH 
 
 pH 
 
 When 
 Filtered 
 
 Total 
 
 Iron 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Iron 
 on 
 
 Filter 
 
 Iron 
 in 
 
 Filtrate 
 mg/1 
 
 Special 
 Conditions 
 
 Deland - 
 
 Experiment i 
 
 
 
 
 
 
 Chloroform A 6. 80 
 
 Cthanol A 
 
 6.80 
 
 8.0 
 
 0. 29 
 
 8.0 
 
 1. 10 
 
 8. 
 
 0. 30 
 
 8.0 
 
 1, 10 
 
 Cl inton - Experiment ii 
 Chloroform A 7.35 8.0 
 
 Cthanol A 
 
 7.00 
 
 Deland - Experiment iii 
 
 Chloroform A 6. 90 
 iCthanol A 8. 30 
 
 iimmonia A 7. 80 
 
 0. 20 
 
 8.0 
 
 0. 28 
 
 8.0 
 
 1. 20 
 
 8. 
 
 2. 12 
 
 Chloroform A 7. 05 
 
 8.0 
 
 0.21 
 
 
 8.0 
 
 1. 12 
 
 
 8. 
 
 1. 96 
 
 Cthanol A 6. 80 
 
 8.0 
 
 0. 30 
 
 
 8. 
 
 I. 04 
 
 Champaign - Experiment 
 
 V 
 
 
 + 
 
 + 
 
 + 
 
 0. 
 
 29 
 
 0. 
 
 48 
 
 0. 
 
 30 
 
 0. 
 
 42 
 
 0. 12 
 
 0. 16 
 1.20 
 0.04 
 
 0. 
 
 21 
 
 0. 
 
 94 
 
 0. 
 
 14 
 
 0. 
 
 30 
 
 0. 
 
 
 
 1C1 A 
 
 8.10 
 
 8. 
 
 0. 34 
 
 + 
 
 0. 34 
 
 8. 
 
 0. 34 
 
 + 
 
 0. 34 
 
 8.0 
 
 0. 20 
 
 - 
 
 
 8. 
 
 1.00 
 
 - 
 
 • • • 
 
 8. 
 
 1. 70 
 
 - 
 
 1.70 
 
 8.0 
 
 2. 00 
 
 + 
 
 1. 90 
 
 8. 
 
 0. 20 
 
 + 
 
 0. 
 
;xtract 
 
 Initial pH 
 pH When 
 
 Filtered 
 
 Total 
 Iron 
 
 in 
 
 Sol- 
 
 _m£/l_ 
 
 121 
 
 Iron Iron Special 
 
 on in Conditions 
 
 Filter Filtrate 
 
 mg/1 
 
 )akwood - Experiment 1 
 
 Chloroform A 7. 55 
 
 8. 
 
 0.20 
 
 + 
 
 0. 12 
 
 Vater A 
 
 7.30 
 
 8. 
 
 
 
 0. 50 
 
 - 
 
 • • • 
 
 
 
 
 8. 
 
 
 
 1.00 
 
 - 
 
 • • ♦ 
 
 
 
 
 8. 
 
 
 
 2. 10 
 
 + 
 
 2. 00 
 
 
 
 
 8. 
 
 
 
 2.40 
 
 + 
 
 1.60 
 
 
 
 
 6. 
 
 9 
 
 2.40 
 
 + 
 
 1. 60 
 
 pH lowered 
 
 Vater A 
 
 7.30 
 
 8. 
 
 
 
 1. 60 
 
 + 
 
 1. 52 
 
 Ferric iron 
 used. 
 
 Cthanol A 
 
 7.50 
 
 8. 
 
 
 
 0._40 
 
 - 
 
 0.40 
 
 
 
 
 8. 
 
 
 
 0. 50 
 
 + 
 
 0.22 
 
 
 
 
 7. 
 
 
 
 0. 50 
 
 + 
 
 0. 10 
 
 pH lowered 
 
 Cthanol A 
 
 7. 50 
 
 8. 
 
 
 
 0. 20 
 
 - 
 
 " 
 
 Ferric iron 
 
 
 
 8. 
 
 
 
 0. 96 
 
 + 
 
 0. 10 
 
 used 
 
 Ammonia A 
 
 7.40 
 
 8. 
 
 
 
 0. 20 
 
 _ 
 
 • • • 
 
 
 
 
 8. 
 
 
 
 0. 50 
 
 - 
 
 • 
 
 
 
 
 8. 
 
 
 
 0. 88 
 
 + 
 
 0.42 
 
 
 
 
 7. 
 
 
 
 0. 88 
 
 + 
 
 0. 32 
 
 pH lowered 
 
 tmmonia A 
 
 7.40 
 
 8. 
 
 
 
 0. 60 
 
 + 
 
 0. 56 
 
 Ferric iron 
 used 
 
 IC1 A 
 
 7.50 
 
 8. 
 
 
 
 0. 36 
 
 + 
 
 0. 04 
 
 
 
 
 7. 
 
 
 
 0. 36 
 
 + 
 
 0. 18 
 
 pH lowered 
 
 IC1 A 
 
 7.50 
 
 8. 
 
 
 
 0. 20 
 
 + 
 
 0. 14 
 
 Ferric iron 
 used 
 
 Chloroform B 
 
 7.35 
 
 8. 
 
 
 
 0. 32 
 
 + 
 
 0. 32 
 
 
 Chloroform B 
 
 7.35 
 
 8. 
 
 
 
 0. 28 
 
 + 
 
 0. 16 
 
 Ferric iron 
 used 
 
 Vater B 
 
 6.55 
 
 8. 
 
 
 
 1. 00 
 
 - 
 
 • • • 
 
 
 
 
 8. 
 
 
 
 1.40 
 
 - 
 
 • • f 
 
 
 
 
 8. 
 
 
 
 2.00 
 
 - 
 
 1.88 
 
 
 
 
 8. 
 
 
 
 2. 50 
 
 + 
 
 2.44 
 
 
 
 
 8. 
 
 
 
 2.72 
 
 + 
 
 2.60 
 
 
 
 
 7. 
 
 
 
 2. 72 
 
 ++ 
 
 2.48 
 
 pH lowered 
 
122 
 
 Extract 
 
 Initial 
 
 PH 
 
 
 Total 
 
 Iron 
 
 Iron 
 
 Special 
 
 
 PH 
 
 When 
 
 
 Iron 
 
 on 
 
 in 
 
 Conditions 
 
 
 
 Filtered 
 
 in 
 
 Filter 
 
 Filtrat 
 
 e 
 
 
 
 
 
 SoL 
 
 
 mg/1 
 
 
 
 
 
 
 mg/I 
 
 
 
 
 )akwood - Exp 
 
 eriment 
 
 1 (Continued) 
 
 
 
 
 Cthanol B 
 
 6.55 
 
 8. 
 
 
 
 L 00 
 
 - 
 
 • . . 
 
 
 
 
 8. 
 
 
 
 2.20 
 
 - 
 
 2. 12 
 
 
 
 
 8. 
 
 
 
 2.48 
 
 + 
 
 2.40 
 
 
 
 
 7. 
 
 
 
 2.48 
 
 + 
 
 2.40 
 
 pH lowered 
 
 
 
 9. 
 
 
 
 2.48 
 
 + 
 
 2.40 
 
 pH raised 
 
 ilthanol B 
 
 6. 55 
 
 8. 
 
 
 
 2.00 
 
 + 
 
 1.40 
 
 Ferric iron 
 used 
 
 Ammonia B 
 
 7. 30 
 
 8. 
 
 
 
 1.00 
 
 r 
 
 • • « • 
 
 
 
 
 8. 
 
 
 
 2.00 
 
 .+ 
 
 1.52 
 
 
 
 
 7. 
 
 
 
 2.00 
 
 + + 
 
 0.98 
 
 pH lowered 
 
 Ammonia B 
 
 7. 30 
 
 8. 
 
 
 
 1.74 
 
 ± 
 
 1.64 
 
 Ferric iron 
 
 
 
 8. 
 
 
 
 2. 24 
 
 + 
 
 0.80 
 
 used 
 
 IC1 B 
 
 7.00 
 
 8. 
 
 3 
 
 0.40 
 
 + 
 
 0. 32 
 
 
 
 
 8. 
 
 1 
 
 0. 50 
 
 + 
 
 0. 10 
 
 
 
 
 9. 
 
 
 
 0. 50 
 
 + 
 
 0.26 
 
 pH raised 
 
 
 
 7. 
 
 
 
 0. 50 
 
 + 
 
 0.08 
 
 pH lowered 
 
 iCl B 
 
 7.00 
 
 8. 
 
 
 
 0.40 
 
 + 
 
 0,. 28 
 
 Ferric iron 
 used 
 
 3 hilo - Experiment 2 
 
 ~~~ ~~~ — —— — ^— — 
 
 Chloroform A 7. 50 
 
 Vater A 
 
 Cthanol A 
 
 7.60 
 
 7.60 
 
 Ammonia A 7. 60 
 
 ICl A 7.60 
 
 Chloroform B 7. 50 
 
 ! Vater B 5. 50 
 
 8.0 
 
 0. 
 
 20 
 
 + 
 
 0. 
 
 12 
 
 
 8. 1 
 
 0. 
 
 28 
 
 + 
 
 0. 
 
 20 
 
 
 8.0 
 
 0. 
 
 34 
 
 + 
 
 0. 
 
 54 
 
 
 8. 1 
 
 0. 
 
 28 
 
 + 
 
 0. 
 
 14 
 
 
 8. 1 
 
 0. 
 
 22 
 
 + 
 
 0. 
 
 16 
 
 
 8.0 
 
 0. 
 
 28 
 
 + 
 
 0. 
 
 14 
 
 
 8. 1 
 
 0. 
 
 40 
 
 + 
 
 0. 
 
 16 
 
 
 6.0 
 
 0. 
 
 40 
 
 + 
 
 0. 
 
 04 
 
 pH lowered 
 
 9.0 
 
 0. 
 
 40 
 
 + 
 
 0. 
 
 24 
 
 pH raised 
 
123 
 
 Extract 
 
 Initial 
 
 PH 
 
 Total 
 
 Iron 
 
 Iron Special 
 
 PH 
 
 When 
 
 Iron 
 
 on 
 
 in Conditions 
 
 
 Filtered 
 
 in 
 
 Sol. 
 
 mg/ 1 
 
 Filter 
 
 Filtrate 
 mg/ 1 
 
 Phi lo - Experiment 2 (continued) 
 Water B 5. 50 8. 1 
 
 Ethanol B 6. 90 
 
 Ethanol B 
 
 6.90 
 
 Ammonia B 7. 50 
 
 Ammonia B 7. 50 
 
 HC1 B 
 
 7.60 
 
 Clinton - Experiment 3 
 Chloroform A 6.45 
 
 Water A - None present 
 Ethanol A 6.70 
 
 Ammonia A 
 
 6.7 5 
 
 HC1 A 6.55 
 
 Chloroform B - Lost 
 Water B - Lost 
 
 8.0 
 
 0.40 
 
 8. 1 
 
 0. 30 
 
 8. 1 
 
 0. 50 
 
 8. 1 
 
 0. 60 
 
 6.8 
 
 0. 60 
 
 8.0 
 
 0.40 
 
 8.0 
 
 0. 64 
 
 8.0 
 
 0. 32 
 
 6.0 
 
 0. 32 
 
 8. 
 
 0. 32 
 
 8. 
 
 0. 30 
 
 0.28 
 
 + 
 
 + 
 + 
 
 + 
 + 
 
 ± 
 
 + 
 
 + 
 + 
 
 + 
 
 + 
 
 0. 
 
 10 
 
 Ferric iron 
 used 
 
 0. 
 
 26 
 
 
 0. 
 
 48 
 
 
 0. 
 
 48 
 
 
 0. 
 
 12 
 
 pH lowered 
 
 0. 
 
 34 
 
 
 0. 
 
 04 
 
 Ferric iron 
 used 
 
 0. 
 
 16 
 
 
 0. 
 
 14 
 
 pH lowered 
 
 0. 
 
 32 
 
 Ferric iron 
 used 
 
 0. 
 
 06 
 
 
 8.0 
 
 0.23 
 
 + 
 
 0. 16 
 
 8. 
 
 1. 00 
 
 + 
 
 0.0 
 
 8. 
 
 0. 27 
 
 - 
 
 0.24 
 
 8. 
 
 1. 00 
 
 + 
 
 0.46 
 
 8.0 
 
 2. 16 
 
 ++ 
 
 0. 12 
 
 8.0 
 
 0. 16 
 
 _ 
 
 0. 08 
 
 8. 
 
 1. 08 
 
 - 
 
 1. 08 
 
 8. 
 
 2. 10 
 
 + 
 
 0. 80 
 
 8. 
 
 3. 16 
 
 ++ 
 
 0. 12 
 
 0. 
 
124 
 
 Extract Initial 
 
 PH 
 
 
 Total 
 
 Iron 
 
 Iron 
 
 Special 
 
 PH 
 
 When 
 
 Iron 
 
 on 
 
 in 
 
 Conditions 
 
 
 Filtered 
 
 in 
 
 Filter 
 
 Filtral 
 
 :e 
 
 
 
 
 Sol. 
 
 
 mg/1 
 
 
 
 
 
 mg/1 
 
 
 
 
 Clinton - Experiment 
 
 3 (continued) 
 
 
 
 
 
 Ethanol B 6. 7 5 
 
 8. 
 
 
 
 0. 26 
 
 + 
 
 0. 30 
 
 
 
 8. 
 
 
 
 1.06 
 
 + 
 
 0. 
 
 
 Ethanol B 6. 7 5 
 
 8. 
 
 
 
 0. 31 
 
 + 
 
 0. 16 
 
 Repetition 
 
 
 8. 
 
 
 
 1. L4 
 
 + 
 
 0. 24 
 
 
 Ammonia B 6. 80 
 
 8. 
 
 
 
 0. 30 
 
 - 
 
 0. 34 
 
 
 
 8. 
 
 
 
 1.06 
 
 - 
 
 0. 88 
 
 
 
 8. 
 
 
 
 2.02 
 
 - 
 
 2. 00 
 
 
 
 8. 
 
 
 
 3. 16 
 
 - 
 
 2.88 
 
 
 
 8. 
 
 
 
 4. 76 
 
 + 
 
 4. 64 
 
 
 
 8. 
 
 
 
 6. 14 
 
 ++ 
 
 4. 32 
 
 
 Ammonia B 7. 10 
 
 8. 
 
 
 
 0.40 
 
 _ 
 
 0.42 
 
 Repetition 
 
 
 8. 
 
 
 
 1. 24 
 
 - 
 
 1. 16 
 
 
 
 8. 
 
 
 
 2. 08 
 
 - 
 
 2. 36 
 
 
 
 8. 
 
 
 
 3. 00 
 
 + 
 
 3.08 
 
 
 
 8. 
 
 
 
 3. 00 
 
 + 
 
 2. 92 
 
 10 min. mix 
 before filtering 
 
 
 8. 
 
 
 
 3. 92 
 
 + 
 
 2.28 
 
 
 
 8. 
 
 
 
 4. 82 
 
 ++ 
 
 0. 28 
 
 
 Ammonia B 7. 10 
 
 8. 
 
 
 
 3. 32 
 
 _ 
 
 3. 20 
 
 
 
 8. 
 
 
 
 3. 32 
 
 ~ 
 
 3. 32 
 
 24 hrs. reaction 
 before filtering 
 
 Ammonia B 6. 70 
 
 8. 
 
 
 
 0.40 
 
 _ 
 
 0. 38 
 
 Equivalent 
 
 
 8. 
 
 
 
 L 10 
 
 : -. 
 
 1. 00 
 
 amount of 
 
 
 8. 
 
 
 
 2. 10 
 
 - 
 
 2. 12 
 
 extract but 
 
 
 8. 
 
 
 
 2. 98 
 
 - 
 
 2.88 
 
 not heated at 
 
 . 
 
 8. 
 
 
 
 4. 26 
 
 + 
 
 4. 04 
 
 100°C - still 
 
 
 8. 
 
 
 
 5. 52 
 
 + 
 
 4. 32 
 
 liquid 
 
 HC1.B 7.00 
 
 8. 
 
 
 
 0. 22 
 
 _ 
 
 0.24 
 
 
 
 8. 
 
 
 
 1. 02 
 
 + 
 
 0. 12 
 
 
 HC1 B 7.00 
 
 8. 
 
 
 
 0. 36 
 
 + 
 
 0. 04 
 
 Repetition 
 
 
 8. 
 
 
 
 1. 14 
 
 + 
 
 0. 32 
 
 
125 
 
 Extract 
 
 Initial 
 
 PH 
 
 Total 
 
 Iron 
 
 Iron 
 
 Special 
 
 pH 
 
 When 
 
 Iron 
 
 on 
 
 in 
 
 Conditions 
 
 
 Filtered 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Filter 
 
 Filtrate 
 mg/1 
 
 
 A twood - Experiment 4 
 Chloroform A 7.50 
 Water A 6. 85 
 
 Ethanol A 
 
 6. 80 
 
 Ethanol A 
 
 Ammonia A 
 
 6.80 
 
 6. 30 
 
 Ammonia A 6. 30 
 
 HC1 A 
 
 6. 50 
 
 Chloroform B 7. 30 
 
 Water B 
 
 6.40 
 
 8. 
 
 0. 20 
 
 8. 
 
 - 0.20 
 
 8.0 
 
 0. 50 
 
 8.6 
 
 1. 00 
 
 8.0 
 
 2. 00 
 
 8. 
 
 3. 00 
 
 8. 
 
 3. 50 
 
 8. 
 
 4. 00 
 
 8.0 
 
 0. 30 
 
 8. 
 
 1. 04 
 
 9.0 
 
 1. 04 
 
 6.0 
 
 1. 04 
 
 8. 
 
 0. 60 
 
 8. 
 
 1. 00 
 
 8.0 
 
 0. 20 
 
 8. 
 
 0. 88 
 
 9.0 
 
 0. 88 
 
 6.0 
 
 0. 88 
 
 8. 
 
 0. 80 
 
 8. 
 
 0. 80 
 
 8. 
 
 0.20 
 
 8. 
 
 0. 30 
 
 8. 
 
 0. 20 
 
 8. 
 
 1. 00 
 
 8. 
 
 2. 00 
 
 8. 
 
 3,00 
 
 8. 
 
 4. 00 
 
 8.0 
 
 5. 00 
 
 8,0 
 
 6. 00 
 
 8.0 
 
 6. 00 
 
 + 
 
 + 
 + 
 
 + 
 + 
 + 
 + 
 
 ++ 
 ++ 
 
 0. 12 
 0.28 
 
 3. 12 
 3.24 
 
 0. 24 
 0.76 
 1.04 
 0. 60 
 
 0. 60 
 0. 04 
 
 0. 22 
 
 0. 80 
 0. 88 
 0. 02 
 
 0. 04 
 
 0. 08 
 
 0. 
 
 0. 28 
 
 0.20 
 
 0. 56 
 0.46 
 
 pH raised 
 pH lowered 
 
 Ferric iron 
 used 
 
 pH raised 
 pH lowered 
 
 Ferric iron 
 
 used 
 
 0. 45 u filter 
 
 0. 45 ^ filter 
 0. 45.U filter 
 0. 45 ^ filter 
 0. 45^ filter 
 0. 45^ filter 
 0. 45|i filter 
 0. 45|j. filter 
 0. 22 u filter 
 
126 
 
 Extract 
 
 Initial pH 
 pH When 
 
 Filtered 
 
 Total 
 
 Iron 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Iron Iron Special 
 
 on in Conditions 
 
 Filter Filtrate 
 
 mg/1 
 
 Atw ood - Experiment 4 (continued) 
 Water B 6.40 
 
 Water B 
 
 6.40 
 
 Water B 
 
 5. 90 
 
 Water B 
 
 5. 90 
 
 Water B 
 
 5. 90 
 
 Water B 
 
 5. 90 
 
 8. 1 
 
 5. 00 
 
 8. 
 
 5. 52 
 
 7.0 
 
 5. 52 
 
 8. 2 
 
 0. 50 
 
 8. 1 
 
 1. 00 
 
 8. 1 
 
 4. 56 
 
 8. 1 
 
 4. 56 
 
 8. 2 
 
 5. 20 
 
 8. 2 
 
 5. 20 
 
 8. 2 
 
 5. 92 
 
 8. 2 
 
 5. 92 
 
 8.0 
 
 0. 31 
 
 8. 
 
 1. 12 
 
 8. 
 
 4. 04 
 
 8. 
 
 4. 90 
 
 8. 
 
 6. 24 
 
 8. 
 
 7. 60 
 
 8. 
 
 0. 31 
 
 8. 
 
 1. 12 
 
 8. 
 
 4. 04 
 
 8.0 
 
 4. 90 
 
 8.0 
 
 6.24 
 
 8.0 
 
 7. 60 
 
 7.0 
 
 0. 18 
 
 7.0 
 
 1.46 
 
 7. 
 
 2. 00 
 
 7. 
 
 4. 08 
 
 7. 
 
 5. 70 
 
 7. 
 
 0. 18 
 
 7. 
 
 1.46 
 
 7. 
 
 2. 00 
 
 7. 
 
 4. 08 
 
 7. 
 
 5. 70 
 
 + 
 
 + 
 
 + 
 
 ++ 
 ++ 
 
 + 
 ++ 
 
 ++ 
 
 . 
 
 . . 
 
 Repetition 
 
 4. 
 
 92 
 
 
 2. 
 
 84 
 
 pH lowered 
 
 0. 
 
 60 
 
 Repetition 
 
 0. 
 
 92 
 
 
 4, 
 
 48 
 
 
 4. 
 
 48 
 
 0.45u filter 
 
 5. 
 
 20 
 
 0. 45u filter 
 
 4. 
 
 88 
 
 0. 22u filter 
 
 5. 
 
 88 
 
 0.45^1 filter 
 
 5. 
 
 92 
 
 0.45^1 filter 
 72 hrs. reac- 
 tion before 
 filtering 
 
 0. 
 
 34 
 
 Repetition 
 
 1. 
 
 04 
 
 
 4. 
 
 00 
 
 
 4. 
 
 60 
 
 
 1. 
 
 92 
 
 
 0. 
 
 32 
 
 
 0. 
 
 30 
 
 Same series 
 
 1. 
 
 10 
 
 as above but 
 
 4. 
 
 00 
 
 with 2 in. dia. 
 
 4. 
 
 78 
 
 filters 
 
 5. 
 
 52 
 
 
 1. 
 
 26 
 
 
 0. 
 
 12 
 
 pH lowered 
 
 1. 
 
 48 
 
 
 1. 
 
 92 
 
 
 4. 
 
 04 
 
 
 0. 
 
 04 
 
 
 0. 
 
 17 
 
 Same series 
 
 1. 
 
 52 
 
 as above but 
 
 1. 
 
 92 
 
 with 2 in. dia 
 
 
 m # 
 
 filters. 
 
 ++ 
 
 0. 30 
 
127 
 
 Extract 
 
 Initi al 
 
 PH 
 
 Total 
 
 Iron 
 
 Iron 
 
 Special 
 
 
 PH 
 
 When 
 
 Iron 
 
 on 
 
 in 
 
 Conditions 
 
 
 
 Filtered 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Filter 
 
 Filtrat 
 mg/ L 
 
 e 
 
 Atwood - 
 
 Experiment 4 (continued) 
 
 5.60 8.0 0.28 
 
 
 
 
 Water B 
 
 Ferric iron 
 
 
 
 8. 
 
 1. 08 
 
 - 
 
 0. 92 
 
 used 
 
 
 
 8. 
 
 2. 16 
 
 + 
 
 2. 12 
 
 
 
 
 8.0 
 
 3. 12 
 
 + 
 
 1. 52 
 
 
 
 
 8. 
 
 4.84 
 
 ++ 
 
 0.40 
 
 
 Water B 
 
 5.60 
 
 8.0 
 
 0. 28 
 
 - 
 
 0.26 
 
 Same series 
 
 
 
 8. 
 
 1. 08 
 
 - 
 
 1.00 
 
 as above but 
 
 
 
 8. 
 
 2. 16 
 
 - 
 
 2. 10 
 
 with 2 in. 
 
 
 
 8.0 
 
 3. 12 
 
 + 
 
 2. 84 
 
 dia. filters 
 
 
 
 8.0 
 
 4. 84 
 
 ++ 
 
 0.76 
 
 
 Water B 
 
 6. 90 
 
 8. 
 
 0.40 
 
 _ 
 
 0. 34 
 
 Water used as 
 
 
 
 8.0 
 
 1.02 
 
 - 
 
 1. 08 
 
 original sol- 
 
 
 
 8.0 
 
 2.48 
 
 - 
 
 2. 60 
 
 vent for 
 
 
 
 8. 
 
 3. 80 
 
 - 
 
 3.76 
 
 extract 
 
 
 
 8.0 
 
 4. 30 
 
 + 
 
 4. 24 
 
 
 
 
 8. 
 
 5.70 
 
 + 
 
 5.76 
 
 
 
 
 8. 
 
 9. 60 
 
 ++ 
 
 8. 32 
 
 
 
 
 8. 
 
 10. 2 
 
 ++ 
 
 7. 60 
 
 
 
 
 8. 
 
 L6. 2 
 
 + + 
 
 0. 20 
 
 
 Water B 
 
 6.90 
 
 8.0 
 
 0.40 
 
 _ 
 
 0. 29 
 
 Same series 
 
 
 
 8.0 
 
 1.02 
 
 - 
 
 1.00 
 
 as above but 
 
 
 
 8. 
 
 2.48 
 
 - 
 
 2. 52 
 
 with 2 inch 
 
 
 
 8.0 
 
 3. 80 
 
 - 
 
 3. 68 
 
 dia. filters 
 
 
 
 8.0 
 
 4. 30 
 
 - 
 
 4. 20 
 
 
 
 
 8.0 
 
 5.70 
 
 + 
 
 5.84 
 
 
 
 
 8.0 
 
 9. 60 
 
 + 
 
 8.66 
 
 
 
 
 8.0 
 
 10. 2 
 
 + 
 
 9. 60 
 
 
 
 
 8. 
 
 16. 2 
 
 ++ 
 
 0. 19 
 
 
 Water B 
 
 7.05 
 
 8. 
 
 1. 12 
 
 _ 
 
 1. 08 
 
 Water used as 
 
 
 
 8.0 
 
 2. 06 
 
 - 
 
 2.00 
 
 original sol- 
 
 
 
 8. 
 
 2. 98 
 
 - 
 
 3. 00 
 
 vent for extract 
 
 
 
 8. 
 
 3.74 
 
 - 
 
 3.76 
 
 
 
 
 8.0 
 
 4. 68 
 
 - 
 
 4.72 
 
 Sample was 
 
 
 
 8.0 
 
 5. 34 
 
 + 
 
 5. 20 
 
 aerated during 
 
 
 
 8.0 
 
 6.26 
 
 + 
 
 6.24 
 
 testing 
 
 
 
 8. 
 
 8. 20 
 
 + 
 
 8. 32 
 
 
 
 
 8.0 
 
 10.00 
 
 +* 
 
 8. 56 
 
 
 
 
 8.0 
 
 12. .8 
 
 + + 
 
 2.44 
 
 
128 
 
 Extract 
 
 Initial 
 
 PH 
 
 Total 
 
 Iron 
 
 Iron 
 
 Special 
 
 PH 
 
 When 
 
 Iron 
 
 on 
 
 in 
 
 Conditions 
 
 
 Filtered 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Filter 
 
 Filtrate 
 mg/1 
 
 
 Atwood - Experiment 4 (continued) 
 
 Water B 
 
 7.05 
 
 Ethanol B 
 
 6.40 
 
 Ethanol B 
 
 6.40 
 
 Ammonia B 
 
 6.60 
 
 Ammonia B 
 
 6. 60 
 
 8. 
 
 1. 12 
 
 8. 
 
 2. 06 
 
 8. 
 
 2. 98 
 
 8. 
 
 3.74 
 
 8. 
 
 4. 68 
 
 8. 
 
 5. 34 
 
 8.0 
 
 6.26 
 
 8.0 
 
 8.20 
 
 8.0 
 
 10. 00 
 
 8.0 
 
 12.8 
 
 8.0 
 
 0. 20 
 
 8.0 
 
 1.00 
 
 8.0 
 
 3.00 
 
 8. 
 
 5,00 
 
 8.0 
 
 5.25 
 
 8.0 
 
 5. 35 
 
 8.0 
 
 5.35 
 
 8. 
 
 5. 35 
 
 6.0 
 
 5. 35 
 
 8. 
 
 1.00 
 
 8. 
 
 2.00 
 
 8. 
 
 4. 28 
 
 8.0 
 
 4. 28 
 
 8.0 
 
 1. 00 
 
 8.0 
 
 2.00 
 
 8.0 
 
 4. 00 
 
 9.0 
 
 4. 00 
 
 9.0 
 
 4.20 
 
 8.0 
 
 4. 20 
 
 8.0 
 
 4. 20 
 
 8.0 
 
 2. 04 
 
 + 
 
 + 
 
 ++ 
 
 + 
 + 
 
 + 
 + 
 
 + 
 
 ++ 
 + 
 
 1. 16 
 
 Same series 
 
 1. 92 
 
 as above but 
 
 2. 92 
 
 with 2 in. 
 
 3. 68 
 
 dia. filters 
 
 4.66 
 
 
 4.28 
 
 
 6. 38 
 
 
 8.04 
 
 
 9.68 
 
 
 6. 20 
 
 
 0. 22 
 
 5. 
 
 20 
 
 
 5. 
 
 12 
 
 
 5. 
 
 16 
 
 72 hrs. reac 
 tion before 
 filtering 
 
 9 
 
 . 
 
 0. 4S|i filter 
 
 1. 
 
 28 
 
 pH lowered 
 
 , 
 
 . . 
 
 Ferric iron 
 
 . 
 
 . . 
 
 used 
 
 0. 
 
 24 
 
 
 0. 
 
 88 
 
 0.45 u filter 
 
 3. 84 
 
 . 
 
 e • 
 
 pH raised 
 
 3. 
 
 84 
 
 pH raised 
 
 2. 
 
 52 
 
 pH lowered 
 
 4. 
 
 12 
 
 0. 45(i filter 
 
 2, 
 
 00 
 
 Ferric iron 
 used 
 
129 
 
 Extract 
 
 Initial 
 PH 
 
 pH 
 When 
 
 
 Total 
 Iron 
 
 Iron 
 on 
 
 Iron 
 in 
 
 Special 
 Conditions 
 
 
 
 Filtered 
 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Filter 
 
 Filtrat 
 mg/1 
 
 e 
 
 Atwood - 
 
 Experiment 4 
 
 (continue 
 
 8. 
 8. 
 
 d) 
 
 0.40 
 1. 26 
 
 + 
 
 0.42 
 0.88 
 
 
 HC1 B 
 
 6.70 
 
 
 
 
 8.0 
 
 
 1. 26 
 
 + 
 
 1. 00 
 
 0.45|~i filter 
 
 
 
 9.0 
 
 
 1. 26 
 
 + 
 
 1. 12 
 
 pH raised 
 
 
 
 6.0 
 
 
 1. 26 
 
 ++ 
 
 0. 20 
 
 pH lowered 
 
 HC1 B 
 
 6.70 
 
 8.0 
 
 
 0. 92 
 
 + 
 
 0. 14 
 
 Ferric iron 
 used 
 
 Clinton - Experiment 5 
 Chloroform A 7. 20 8. 
 
 Water A - None present 
 Ethanol A 7. 25 
 
 Ammonia A 7.30 
 
 Ammonia A 7. 30 
 
 HC1 A 
 
 7.30 
 
 Chloroform B 7. 20 
 
 Water B 
 
 7.20 
 
 8.0 
 
 0. 30 
 
 8.0 
 
 0.40 
 
 8.0 
 
 1. 32 
 
 9.0 
 
 1. 32 
 
 8. 1 
 
 0. 20 
 
 8. 
 
 0. 50 
 
 8.0 
 
 1. 00 
 
 8.0 
 
 2. 16 
 
 9.0 
 
 2. 16 
 
 8.0 
 
 1. 50 
 
 8.0 
 
 2.00 
 
 0.20 
 
 8.0 
 
 0.40 
 
 8.0 
 
 0. 68 
 
 8.0 
 
 0. 50 
 
 8. 
 
 1. 00 
 
 8.0 
 
 2. 00 
 
 8.0 
 
 3. 00 
 
 8.0 
 
 4. 00 
 
 8. 
 
 4. 72 
 
 9.0 
 
 4.72 
 
 6.0 
 
 4. 72 
 
 + 
 + 
 
 + 
 
 + 
 
 + 
 + 
 
 + 
 
 0.24 
 
 0. 36 
 0.42 
 0. 64 
 
 0.40 
 0. 16 
 
 pH raised 
 
 0. 
 
 70 
 
 
 2. 
 
 12 
 
 pH raised 
 
 1. 
 
 56 
 
 Ferric iron 
 
 0. 
 
 72 
 
 used 
 
 + 
 
 4. 
 
 60 
 
 
 4. 
 
 76 
 
 pH raised 
 
 3. 
 
 32 
 
 pH lowered 
 
130 
 
 Extract 
 
 Initial 
 
 pH 
 
 Total 
 
 Iron 
 
 Iron Special 
 
 pfT 
 
 When 
 
 Iron 
 
 on 
 
 in Conditions 
 
 
 Filtered 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Filter 
 
 Filtrate 
 mg/1 
 
 C linton - Experiment 5 (continued) 
 7.20 
 
 Water B 
 
 Ethanol B 
 
 7.05 
 
 Ammonia B 6. 95 
 
 HC1 B 
 
 7. 10 
 
 Chloroform C 6. 90 
 
 8. 1 
 
 2.00 
 
 - 
 
 . . . 
 
 Ferric iron 
 
 8. 1 
 
 3.00 
 
 + 
 
 2. 92 
 
 used 
 
 8.0 
 
 0. 17 
 
 _ 
 
 0. 34 
 
 
 8.0 
 
 1. 08 
 
 - 
 
 0. 92 
 
 
 8.0 
 
 1. 92 
 
 _ 
 
 2. 12 
 
 
 8. 
 
 3.04 
 
 - 
 
 3. 00 
 
 
 8.0 
 
 4.00 
 
 - 
 
 4. 04 
 
 
 8. 
 
 5. 16 
 
 + 
 
 4. 64 
 
 
 8.0 
 
 5. 16 
 
 + 
 
 5.08 
 
 10 min. mix 
 before filter 
 ing 
 
 8. 
 
 6. 24 
 
 + 
 
 5. 36 
 
 
 8. 
 
 7. 60 
 
 ++ 
 
 2.08 
 
 
 8. 
 
 7. 60 
 
 ++ 
 
 1.04 
 
 10 min. mix 
 before filter 
 ing 
 
 8. 
 
 1.02 
 
 _ 
 
 0. 96 
 
 
 8. 
 
 1. 88 
 
 - 
 
 1. 88 
 
 
 8.0 
 
 3.04 
 
 _ 
 
 3. 12 
 
 
 8.0 
 
 4. 00 
 
 - 
 
 3. 96 
 
 
 8.0 
 
 5. 28 
 
 + 
 
 5.28 
 
 
 8.0 
 
 5. 28 
 
 + 
 
 5. 16 
 
 10 min. mix 
 before filter 
 • ing 
 
 8. 
 
 6. 88 
 
 + 
 
 5. 60 
 
 
 8. 
 
 8.78 
 
 ++ 
 
 2. 96 
 
 
 8.0 
 
 10. 80 
 
 ++ 
 
 0. 64 
 
 
 8.0 
 
 0.40 
 
 _ 
 
 0. 26 
 
 
 8.0 
 
 1.36 
 
 + 
 
 0. 
 
 
 8.0 
 
 0. 29 
 
 + 
 
 0. 06 
 
 
 8. 
 
 1. 20 
 
 + 
 
 0. 
 
 
131 
 
 Extract Initial 
 
 PH 
 
 Total 
 
 Iron 
 
 Iron 
 
 Special 
 
 PH 
 
 When 
 
 Iron 
 
 on 
 
 in 
 
 Conditions 
 
 
 Filtered 
 
 in 
 
 Sol. 
 
 mg/1 
 
 Filter 
 
 Filtrat 
 mg/1 
 
 e 
 
 Clinton - Experiment 
 
 5 (continued) 
 8. 
 
 0. 30_ 
 
 
 0. 32 
 
 
 Water C 6.70 
 
 
 
 8.0 
 
 1. 12 
 
 - 
 
 1. 04 
 
 
 
 8. 
 
 2. 14 
 
 - 
 
 2. 20 
 
 
 
 8. 
 
 3. 16 
 
 - 
 
 3. 12 
 
 
 
 8. 
 
 4.42 
 
 - 
 
 4. 08 
 
 
 
 8. 
 
 5. 88 
 
 + 
 
 5. 84 
 
 
 
 8.0 
 
 5. 56 
 
 + 
 
 4. 08 
 
 
 Ethanol C 6.70 
 
 8.0 
 
 0. 22 
 
 - 
 
 0.24 
 
 
 
 8.0 
 
 2.42 
 
 - 
 
 2. 56 
 
 
 
 8. 
 
 3.84 
 
 - 
 
 3. 80 
 
 
 
 8. 
 
 5. 56 
 
 + 
 
 4. 28 
 
 
 
 8.0 
 
 7. 68 
 
 ++ 
 
 2. 28 
 
 
 Ammonia C 7. 10 
 
 8. 
 
 1. 08 
 
 _ 
 
 1. 12 
 
 
 
 8.0 
 
 3.70 
 
 + 
 
 3. 52 
 
 
 
 8.0 
 
 5.06 
 
 ++ 
 
 0.24 
 
 
 HC1 C 7.20 
 
 8.0 
 
 0. 24 
 
 _ 
 
 0. 20 
 
 
 
 8.0 
 
 1.02 
 
 - 
 
 1.08 
 
 
 
 8.0 
 
 2.08 
 
 + 
 
 0.0 
 
 
 Control Experiment 
 
 
 
 
 
 
 No extract added 
 
 8. 
 
 0. 30 
 
 _ 
 
 0. 30 
 
 
 
 8.0 
 
 0.44 
 
 + 
 
 0.44 
 
 
 
 8. 
 
 0. 92 
 
 + 
 
 0. 56 
 
 
 
 8. 
 
 1. 28 
 
 + 
 
 0. 20 
 
 
 
 8.0 
 
 1. 58 
 
 + 
 
 0. 12 
 
 
 
 8. 
 
 2. 00 
 
 + 
 
 0. 08 
 
 
 
 8.0 
 
 2. 30 
 
 + 
 
 0. 08 
 
 
 No extract added 
 
 8. 
 
 0. 30 
 
 _ 
 
 0. 30 
 
 Same series 
 
 
 8. 
 
 0.44 
 
 - 
 
 0.44 
 
 as above 
 
 
 8.0 
 
 0. 92 
 
 + 
 
 0. 92 
 
 with 2 in. 
 
 
 8.0 
 
 1. 28 
 
 + 
 
 0. 68 
 
 dia. filters 
 
 
 8.0 
 
 1. 58 
 
 + 
 
 0. 29 
 
 
 
 8.0 
 
 2. 00 
 
 + 
 
 0.08 
 
 
 
 8. 
 
 2. 30 
 
 + 
 
 0. 08 
 
 

 
 
 
 
 
 132 
 
 Extract Initial 
 
 PH 
 
 
 Total 
 
 Iron 
 
 Iron 
 
 Special 
 
 PH 
 
 When 
 
 Iron 
 
 on 
 
 in 
 
 Conditions 
 
 
 Filtered 
 
 in 
 
 Filter 
 
 Filtrat 
 
 e 
 
 
 
 
 Sol. 
 
 
 mg/1 
 
 
 
 
 
 mg/.l 
 
 
 
 
 Control Experiment (continued) 
 
 
 
 
 
 No extract added 
 
 8. 
 
 
 
 0.48 
 
 + 
 
 0.46 
 
 
 
 8. 
 
 
 
 0.48 
 
 + 
 
 0.40 
 
 10 min. mix 
 before filter- 
 ing 
 
 
 8. 
 
 
 
 0. 54 
 
 + 
 
 0.08 
 
 15 min. mix 
 before filter- 
 ing 
 
 No extract added 
 
 8. 
 
 
 
 0.48 
 
 _ 
 
 0.46 
 
 Same series 
 
 
 8. 
 
 
 
 0.48 
 
 - 
 
 0. 54 
 
 as above but 
 
 
 8. 
 
 
 
 0. 54 
 
 + 
 
 0.44 
 
 with 2 in. 
 dia. filters 
 
 No extract added 
 
 8. 
 
 
 
 0. 16 
 
 + 
 
 0. 16 
 
 Diffused air 
 
 
 8. 
 
 
 
 0. 52 
 
 + 
 
 0.46 
 
 used 
 
 
 8. 
 
 
 
 0.52 
 
 + 
 
 0. 36 
 
 10 min. mix 
 before filter- 
 ing 
 
 
 8. 
 
 
 
 1. 36 
 
 ++ 
 
 0. 12 
 
 
 
 8. 
 
 
 
 1. 36 
 
 ++ 
 
 0.42 
 
 10 min. mix 
 before filter- 
 ing 
 
 
 8. 
 
 
 
 2. 10 
 
 ++ 
 
 0. 22 
 
 
 
 8. 
 
 
 
 2. 10 
 
 ++ 
 
 0.0 
 
 10 min. mix 
 before filter- 
 ing 
 
 
 8. 
 
 
 
 2. 52 
 
 ++ 
 
 0. 
 
 
 No extract added 
 
 8. 
 
 
 
 0. 16 
 
 _ 
 
 0. 17 
 
 Same series 
 
 
 8. 
 
 
 
 0. 52 
 
 + 
 
 0.44 
 
 as above but 
 
 
 8. 
 
 
 
 0. 52 
 
 + 
 
 0. 50 
 
 with 2 in. 
 
 
 8. 
 
 
 
 1. 36 
 
 + 
 
 0. 60 
 
 dia. filters 
 
 
 8. 
 
 
 
 1. 36 
 
 + 
 
 0. 92 
 
 
 
 8. 
 
 
 
 2. 10 
 
 + 
 
 0.44 
 
 
 
 8. 
 
 
 
 2. 10 
 
 + 
 
 0. 33 
 
 
 
 8. 
 
 
 
 2. 52 
 
 + 
 
 0. 07 
 
 
133 
 
 Extract 
 
 c t Initial pH 
 
 pH When 
 
 Filtered 
 
 Total 
 Iron 
 in 
 Sol. 
 
 mg/1 
 
 Iron 
 
 on 
 
 Filter 
 
 Iron 
 in 
 
 Filtrate 
 rag/ 1 
 
 Special 
 Conditions 
 
 ol Experiment (continued) 
 
 
 
 
 
 Tartaric acid 6. 30 
 
 Tartaric acid 6. 30 
 
 Tartaric acid 6. 30 
 
 8. 
 
 1.02 
 
 8.0 
 
 2. 88 
 
 8. 
 
 5. 20 
 
 8.0 
 
 7. 28 
 
 8.0 
 
 . 9. 36 
 
 8. 
 
 11. 28 
 
 8.0 
 
 12.44 
 
 8. 
 
 12.44 
 
 1. 12 
 
 2. 84 
 
 5. 00 
 
 6. 92 
 9. 52 
 
 11.24 
 12.24 
 
 12. 04 
 
 10 min mix be- 
 fore filtering 
 
 8.0 
 
 15.04 
 
 + 
 
 14. 84 
 
 
 8. 
 
 15. 04 
 
 + 
 
 15*08 
 
 30 min. mix be 
 fore filtering 
 
 8.0 
 
 16.48 
 
 + 
 
 15. 92 
 
 
 8.0 
 
 19. 8 
 
 + 
 
 19. 6 
 
 
 8. 
 
 24.4 
 
 + 
 
 23. 5 
 
 
 8.0 
 
 29. 3 
 
 ++ 
 
 24. 1 
 
 
 8.0 
 
 33. 5 
 
 ++ 
 
 5. 80 
 
 
 8.0 
 
 62.6 
 
 +++ 
 
 0. 08 
 
 - 
 
 8.0 
 
 1.02 
 
 _ 
 
 0. 92 
 
 Same series 
 
 8.0 
 
 2. 88 
 
 - 
 
 2. 86 
 
 as above but 
 
 8.0 
 
 5.20 
 
 - 
 
 5.06 
 
 with 2 in. 
 
 8.0 
 
 7.28 
 
 - 
 
 7. 28 
 
 dia. filters 
 
 8.0 
 
 9. 36 
 
 - 
 
 9. 04 
 
 
 8. 
 
 11. 28 
 
 - 
 
 11. 56 
 
 
 8. 
 
 12.44 
 
 - 
 
 12.08 
 
 
 8. 
 
 12.44 
 
 - 
 
 12. 08 
 
 1.0 min. mix be 
 fore filtering 
 
 8.0 
 
 15.04 
 
 + 
 
 15. 00 
 
 
 8.0 
 
 15. 04 
 
 + 
 
 15. 00 
 
 30 min. mix be 
 fore filtering 
 
 8. 
 
 16.48 
 
 ± 
 
 16.48 
 
 
 8.0 
 
 19. 8 
 
 + 
 
 24. 5 
 
 
 8. 
 
 24.4 
 
 + 
 
 24. 2 
 
 
 8.0 
 
 29. 3 
 
 + 
 
 20.7 
 
 
 8.0 
 
 33. 5 
 
 ++ 
 
 22. 9 
 
 
 8. 
 
 62.6 
 
 ++ 
 
 0.07 
 
 
134 
 
 APPENDIX C 
 
 INFRA-RED SPECTROGRAPHS 
 
135 
 
136 
 
137 
 
Extract 
 
 APPENDIX D 
 
 LOG OF OXIDATION STUDY 
 
 Phil o - Raw water : Total Iron: 0. 66 mg/1 
 
 Alkalinity: 227 mg/1 as CaC03. 
 
 138 
 
 ct Lapsed 
 
 Ferrous 
 
 Total 
 
 pH 
 
 Dissolved 
 
 Time 
 
 Iron 
 
 Iron 
 
 of 
 
 Oxygen 
 
 After 
 
 Before 
 
 After 
 
 Test 
 
 in Test 
 
 Aeration 
 
 Filtration 
 
 Filtration 
 
 Sol. 
 
 Sol. 
 
 Min. 
 
 mg/ 1 
 
 mg/ 1 
 
 
 mg/ 1 
 
 Mo extract 
 
 Before air. 
 
 0.66 
 
 0. 66 
 
 7.48 
 
 0. 
 
 25 
 
 idded 
 
 
 
 0. 54 
 
 0. 32 
 
 7. 75 
 
 
 
 
 3 
 
 0. 30 
 
 0. 26 
 
 7. 78 
 
 7. 
 
 20 
 
 
 10 
 
 0. 20 
 
 0. 18 
 
 7. 77 
 
 
 
 
 20 
 
 0.06 
 
 0. 10 
 
 7. 77 
 
 
 
 
 35 
 
 0. 04 
 
 0. 06 
 
 7. 78 
 
 
 
 
 50 
 
 0.0 
 
 0. 04 
 
 7. 78 
 
 
 
 fYmmonia B 
 
 Before air. 
 
 0. 50 
 
 0. 52 
 
 7. 54 
 
 0. 
 
 30 
 
 3 mg/1 
 
 
 
 0.45 
 
 0.48 
 
 7. 90 
 
 
 
 
 3 
 
 0. 32 
 
 0. 24 
 
 7. 92 
 
 7. 
 
 30 
 
 
 10 
 
 0.04 
 
 0. 12 
 
 7. 94 
 
 
 
 
 20 
 
 0.04 
 
 0. 20 
 
 7. 94 
 
 
 
 
 32 
 
 0.04 
 
 0.06 
 
 7. 94 
 
 
 
 Tartaric Acid 
 
 Before air. 
 
 0.40 
 
 0. 60 
 
 7.70 
 
 0. 
 
 30 
 
 5 mg/1 
 
 
 
 0.26 
 
 0. 56 
 
 7. 98 
 
 
 
 
 3 
 
 0. 26 
 
 0. 60 
 
 8.00 
 
 7. 
 
 40 
 
 
 10 
 
 0. 12 
 
 0. 58 
 
 8. 00 
 
 
 
 
 20 
 
 0. 10 
 
 0. 60 
 
 8. 00 
 
 
 
 
 35 
 
 0.08 
 
 0.46 
 
 8. 00 
 
 
 
 
 55 
 
 0.06 
 
 0. 30 
 
 8. 00 
 
 
 
 
 75 
 
 0.04 
 
 0. 18 
 
 7. 96 
 
 
 
 
 105 
 
 0.04 
 
 0. 18 
 
 7. 93 
 
 
 
 31inton - Raw water : Total Iron: 1.88mg/.l 
 
 Alkalinity: 4.30 mg/1 as CaC0 3 
 
 ^o extract 
 idded 
 
 Before air. 
 
 1.85 
 
 2. 04 
 
 7. 56 
 
 0. 05 
 
 
 
 1, 18 
 
 0. 96 
 
 7. 83 
 
 
 3 
 
 0.71 
 
 0. 80 
 
 7. 88 
 
 6.70 
 
 10 
 
 0. 22 
 
 0. 10 
 
 7. 89 
 
 
 18 
 
 0. 18 
 
 0. 14 
 
 7. 89 
 
 
 26 
 
 0.0 
 
 0. 
 
 7. 89 
 
 
139 
 
 Istract 
 
 Lapsed 
 
 Ferrous 
 
 Total 
 
 PH 
 
 Dissolved 
 
 Time 
 
 Iron 
 
 Iron 
 
 of 
 
 Oxygen 
 
 After 
 
 before 
 
 After 
 
 Test 
 
 in Test 
 
 Aeration 
 
 Filtration 
 
 Filtration 
 
 Sol. 
 
 Sol. 
 
 Min. 
 
 mg/1 
 
 mg/1 
 
 
 mg/1 
 
 (Linton (continued) 
 
 'irtaric 
 
 Acid 
 
 B 
 
 efore 
 
 air. 
 
 1.95 
 
 1. 
 
 88 
 
 7. 
 
 61 
 
 
 img/1 
 
 
 
 
 
 
 1.25 
 
 1. 
 
 68 
 
 7. 
 
 82 
 
 
 
 
 
 3 
 
 
 0.78 
 
 1. 
 
 84 
 
 7. 
 
 83 
 
 6.70 
 
 
 
 
 10 
 
 
 0.22 
 
 1. 
 
 32 
 
 7. 
 
 82 
 
 
 
 
 
 18 
 
 
 0. 17 
 
 1. 
 
 00 
 
 7. 
 
 82 
 
 
 
 
 
 28 
 
 
 0.06 
 
 0. 
 
 60 
 
 7. 
 
 82 
 
 
 
 
 
 38 
 
 
 0. 10 
 
 0. 
 
 68 
 
 7. 
 
 82 
 
 
 .mmonia 
 
 B 
 
 B 
 
 efore 
 
 air. 
 
 2.02 
 
 1. 
 
 88 
 
 7. 
 
 58 
 
 
 kp. 5 
 
 
 
 
 
 
 1.25 
 
 1. 
 
 00 
 
 7. 
 
 85 
 
 
 ?mg/l 
 
 
 
 3 
 
 
 0.75 
 
 0. 
 
 36 
 
 7. 
 
 85 
 
 6.75 
 
 
 
 
 10 
 
 
 0.21 
 
 0. 
 
 30 
 
 7. 
 
 85 
 
 
 
 
 
 18 
 
 
 0. 17 
 
 0. 
 
 
 
 7. 
 
 85 
 
 
 
 
 
 26 
 
 
 0.03 
 
 0. 
 
 16 
 
 7. 
 
 85 
 
 
 jnmonia 
 
 B 
 
 B 
 
 efore 
 
 air. 
 
 2,04 
 
 1. 
 
 72 
 
 7. 
 
 65 
 
 0. 05 
 
 kp. 5 
 
 
 
 
 
 
 1, 15 
 
 1. 
 
 68 
 
 7. 
 
 80 
 
 
 I mg/1 
 
 
 
 3 
 10 
 
 
 0.63 
 0. 19 
 
 1. 
 
 16 
 
 7. 
 7. 
 
 80 
 80 
 
 6.60 
 
 
 
 
 18 
 
 
 0. 10 
 
 0. 
 
 16 
 
 7. 
 
 80 
 
 
 
 
 
 26 
 
 
 0.03 
 
 0. 
 
 
 
 7. 
 
 80 
 
 
140 
 
 VITA 
 
 Lloyd R. Robinson, Jr. was born April 13, 1931 in Kansas City, 
 Missouri, and received his primary and secondary education in that com- 
 munity. 
 
 He attended Kansas City, Missouri, Junior College and received 
 the A. S. Certificate in 1950. He then attended the University of Kansas 
 where he received the B.S. Degree in Civil Engineering in 1953. During 
 the summer of 1952, he was employed as an Engineering Assistant at 
 Black and Veatch, Consulting Engineers. 
 
 He was commissioned as an Ensign in the Civil Engineer Corps, 
 U. S. Naval Reserve, in July, 1953 and served first as a personnel offi- 
 cer and then as a transportation officer before his release from active 
 duty in 1956. 
 
 In 1956 he accepted a position as Instructor in Civil Engineering 
 and part-time graduate student at the University of Kansas. During the 
 summer of 1957, he served as a Junior Engineer for the Kansas State 
 Board of Health. In 1958 he received the M. S. Degree in Civil Engi- 
 neering. 
 
 He then accepted a position in the Plans and Grants Section of 
 the Water Pollution Control Division of the Texas State Health Depart- 
 ment. He resigned this position in January, I960 to accept a position 
 as Research Assistant at the University of Illinois. During this period, 
 ,he was also a part-time graduate student. During the school year 1961-62, 
 
141 
 
 ie pursued his studies full-time with the aid of a Traineeship from the 
 J. S. Public Health Service. 
 
 He is presently an Assistant Professor of Civil Engineering at 
 Mew Mexico State University. 
 
 He is a Member of the Society of the Sigma Xi, the American Water 
 Works Association, the Water Pollution Control Federation, the American 
 Society for Engineering Education, an Associate Member of the American 
 Society of Civil Engineers, and a Lieutenant in the Civil Engineer Corps, 
 J. S. Naval Reserve.