Polarography of Humic Acid-Like Oxidation Products of Bituminous Coal

Phthalide produced one well-defined wave with a half-wave potential nearly the same as that for the second wave of diethyl phthalate (Table I). Thus, ...
0 downloads 0 Views 634KB Size
ANALYTICAL CHEMISTRY

362 increment in voltage (Table I ) occurred between the first and second waves of diethyl phthalate. This increment was enough to allow a large scale controlled electrolysis of a solution of diethyl phthalate to be made. I n addition, samples of phthalide and phthalaldehyde were examined polarographically. Phthalaldehyde produced two polarographic waves at more positive potentials than diethyl phthalate. The half-wave potential of the first wave us. a mercury pool was -1.15 volts, while the second wave us. a mercury pool was - 1.57 volts. The first wave disappeared on Ptanding. The second wave appeared to overlap slightly n i t h the first wave for diethyl phthalate (Table I ) . Phthalaldehyde was therefore not considered to be the intermediate product in the reduction process. Phthalide produced one n ell-defined wave with a half-wave potential nearly the same as that for the second wave of diethyl phthalate (Table I ) . Thus, if phthalide were the intermediate product in the electrode reaction, the addition of phthalide to a solution sf diethyl phthalate should increase only the second wave. ildditions of phthalide to a 1mJl solution of diethyl phthalate did increase only the second a-ave. Three large scale electrolysis experiments failed t o produce phthalaldehyde as a reduction product. I n each case phthalide was isolated in good yields. 0-Toluic acid was a t first thought t o be the end product in the reaction. Hoaever, no o-toluic’acid lyas isolated from the reaction mixture upon reducing phthalide a t a mercury cathode. I n this case the cathode potential was raised to -2.20 volts us. S.C.E. This potential was well on the limiting value of the diffusion current for phthalide as determined polarographically. The product isolated in the reduction of phthalide appeared to have the properties of an aldehyde as shown by infrared spectruni and chemical tests. It TTas noncrystalline as determined from an x-ray diffraction pattern of a powdered sample. On the basis of these evperimental data the intermediate product in the polarographic reduction of diethyl phthalate is phthalide (Equation 2). The end product in the electrode process is so far unknown. From these experiments i t seems likely that the second step in the polarographic reduction is the follon ing:

//

0

C-H

However, under the conditions of the large scale electrolysis experiments, the product is not stable and only a resinous material was recovered from the catholyte. ACKNOWLEDGMENT

The authors wish to express their appreciation to Julian JI. Xielson for assistance in measuring diffusion coefficients. This paper is published by permission of W. B. JIcLean, technical director of the U. S. Xaval Ordnance Test Station, Inyokern, China Lake, Calif. LITERATURE CITED (1) Britton, H. T. S., and Robinson, R. rl., J . Chem. Soc., 1931,1456.

(2) Gardner, J. H., and Naylor, C. A., Jr., in “Organic Syntheses,” Coll. Vol. 11, p. 526, Wiley, Ken, York, 1943. (3) Kolthoff, I. &I., and Lingane, J. J., “Polarography,” 2nd ed., p. 70, Interscience Publishers, New York, 1952. (4) Ibid., pp. 373-5. (5) Ibid.,p. 374. (6) Lingane, J. J., and Jones, S. L., AKAL.CHEM.,22, 1169 (1950). (7) Radin, Nathan, and DeVries, Thomas, Ibid., 24, 971 (1952). (8) Shirley, D. A,, “Preparation of Organic Intermediates,” p.*261, Wiley, New York, 1951. (9) Stokes, R. H., J . A m . Chem. Soc., 72, 763 (1950). (10) Whitnack, G. C., and Gantz, E. S.C., ANAL.CHEW,24, 1060-1 (1952). (11) Ibid., 25, 553-6 (1953). (12) Whitnack. G. C., Kielson, J. N., and Gantz, E. S. C.. J . Am. Chem. Soc., 76, 4711-14 (1954). RECEIVED for review July 12, 1954. Accepted November 26, 1954. Presented before the Division of Analytical Chemistry a t the 1 2 6 t h Meeting of the . ~ M E P . I C A Y CHEMICAL B o c I E r Y , New York, September 1954.

Polarography of Humic Acid-Like Oxidation Products of Bituminous Coal A. F. CODY, S. R. MILLIKEN, and C. R. KINNEY The Pennsylvania State University, State College, f a . ; i study of the polarographic behavior of humic acids was undertaken because of the sensitivity of polarography to changes in conditions imposed during electrolysis. It was assumed that if the humic acids have abncrmal properties, indications of such abnormalities might be observed at the dropping mercury electrode. Under suitable conditions three well-defined reduction waves were observed. The waves were long and drawn out but otherwise appeared to be normal and were affected by drop time, temperature, buffer, carrier ions, pH, and concentration in a typical manner. The first wave was most persistent of the three and was believed to be due to the reduction of nitro groups. The second wave was best developed in alcoholic solutions of 30 to 40’% alcohol. The third wave was best developed in aqueous solutions containing a salt such as potassium chloride. The structures responsible for the second and third waves have not been identified. The polaro-

graphic behavior of the humic acids indicates that these substances can be studied successfully by this means. Diffusion currents observed for the first wave suggest that the molecular weight of these acids is less than 1000.

C

E R T A I S properties of the humic acid-like oxidation products of bituminous coal appear to be anomalous ( 1 ) and for this reason the behavior of these acids at the dropping mercury electrode was studied. Polarography was not selected to prove structural features of the humic acids but rather because of its sensitivity to changes in conditions imposed during electrolysis If the humic acids have abnormal properties, it was assumed that indications of such abnormalities might be observed a t the dropping mercury electrode. The humic acids selected for study were of a sample prepared by the treatment of a high-volatile 2 . bituminous coal TTith hot concentrated nitric arid ( 2 ) . The humic acids constituted about

V O L U M E 27, NO. 3, M A R C H 1 9 5 5 65% of t h e carbon of the original coal and therefore are believed to contain an important part of the structure of the coal unit molecules. The nitric acid treatment introduces 4.6% nitrogen into the humic acids, and it has been estimated that 68.4% of the nitrogen (total nitrogen including t h a t part of the original nitrogen in the coal remaining after treatment) was in an oxidized state, probably nitro groups. These groups were expected to undergo polarographic reduction, as well as possible ketonic or quinoidal structures t h a t might have been produced by t h e oxidizing action of the hot nitric acid or olefinic structures preqent in the original coal molecules, although natural humic acids have been reported as not being reducible (10).

363 Table I.

Source of Coal and .inalyses of Humic Acids on Jloisture-Free Basis Amount of Elements Present, % Sample 1 , Upper Freeport a

Sample 3, Pittsburgh b

61.4 3.2 4.6 0.4 28 :4 2.0

62.9 2.2 1.4 1.0

Carbon 61.4 Hydrogen 3.2 Kitrogen 4 6 Sulfur 0 5 Oxygen ( b y difference) 28.9 Ash 1.5 a Oxidized with concentrated nitric acid. b Oxidized with air a t ZOOo C .

Table 11. EQUIPM EiYT

A Leeds and Xorthrup Electrochemograph Type E polarograph was used. T o minimize large oscillations during the growth of individual mercury drops and to facilitate thereby the measurement of the current-voltage curves, a high degree of damping was imposed on the recording galvanometer. As a result, the half-v, ave potentials recorded were shifted to more negative values. T h e amount of the shift was -0.04 volt. Consequently, the half-wave potentials reported on the humic acids presumably should be increased by this amount. The capillary used was a 10-cm. length of marine barometer tubing. A constant mercury level apparatus described by Lingane and Laitinen (11) was employed to maintain a constant drop rate. T h e capillary characteristics were: drop time, 4.0 seconds per drop; m for the capillary, 1.59 mg. per second; and for m2i3t1'6, 1.72 mg.2/3per second (open circuit). T h e cells used were standard H-type, permanent external anode cells described by Lingane and Laitinen (11 ), using the saturated calomel electrode (S.C.E.). T h e cell in use was immersed in a constant temperature water bath a t 25' & 0.5" C. Cells were used until the porous plugs were harmed by the strongly alkaline solutions and then they Fere replaced. T h e resistance of the cells was about 600 ohms. A Beckman 1Iodel H-2 pH meter TI as used to obtain the p H of all solutions studied.

Sample 2 , Pittsburgha

~

b

First Wave Ei/1

3.5 4.0 5.0 6.0

pH:

4.2

Effect of 3Iercury Drop Time"

Drop Time, Seconds

a

28.3

Second Wave _ . Ei/a

Zd

-0.64 -0.63 -0.64 -0.65

0 85 0.75 0 58 0.65

id

-1.68 -1.66

0.45 0.18 b

b

Humic acids. 0.100 gram per liter; Clark and Lubs borate buffer: 0 . 1 M , 9.1.

These tiares were too small to be measured.

XIATERIALS

In addition to the nitric acid-oxidized coal humic acids mentioned above, tlvo other samples were examined. The source of the coal, the method used in preparing t h e humic acids, and the analyses appear in Table I. Sample 1, used in most of the present work, was a portion of a large sample prepared previously from Upper Freeport' seam coal and stored under nitrogen ( 2 ) . Sample 2 m-as prepared from Pittsburgh-seam, high-volatile A bituminous coal (9), using the same method as for sample 1. Sample 3 was prepared from the same coal ( 1 2 ) using air oxidation (4)a t 200" C. for 234 hours. Humic acids from sample 1 were also reduced with an equal quantity of Devarda's alloy a t room temperature in dilute sodium hydroxide solution. Following reduction, the acids were precipitated with hydrochloric acid and washed with water until free from chloride ion. This resulted in some loss by peptization of the humic acids and consequent recovery of 88%. All ot,her reagents used met ARIERICAXCHEMICALSOCIETY standards of purity and xere tested v-ith the polarograph t,o ensure that no interference occurred in the potential range under study. Clark and Lubs borate buffer was used from p H 7.5 to (3.0 and Kolthoff and Vleeschouwer sodium carbonate-Boras and disodium phosphate buffers from 9.4 to 10.6 and 10.9 to 11.7, respectively. At p H 12.5 no buffer was added to the 0 . 1 s sodium hydroxide solution. PROCEDURE

Humic acid solutions were prepared by dissolving weighed amounts of the acids in predetermined volumes of 0.1N sodium hj-droxide, the volume depending upon the final pH desired. After solution of the acids was complete and the buffer added, any additional solvent, salts, or other material was introduced, and the solution was diluted to volume in a volumetric flask. When the sodium carbonate-Borax buffer was used with no sodium hydroxide, it was found difficult to get the acids dispersed satisfactorily in sodium carbonate, but if the acids were first dissolved in 0.1N sodium hydroxide followed by the addition of an equivalent amount of sodium bicarbonat,e, the acids were Tell dispersed and the net result so far as the buffer and the pH Tvere concerned was the same as though the acids had been dis-

00

/

l 02

I

l 04

!

l 06

1

I CB

,

,I

I

I

IO I2 PO?ENTIAL,-VOI?B

l

1 14

1

1 16

1

1 IS

1 20

Figure 1. Polarograms of humic acids I. S i t r i c acid-oxidized coal humic acids 11. Humic acids reduced b y Devarda's alloy

solved in sodium carbonate solution. The type of buffer seemed to have little effect on either the half-wave potentials or the diffusion currents. T o prepare solutions of different concentrations of humic acids, aliquots were removed from the stock solutions as prepared above and diluted with an identically prepared solution containing no humic acids. The pH of each solution was determined before polarographing. After the humic acid solutions had been placed in the cell, nitrogen freed from oxygen by passing through alkaline pyrogallol was bubbled through the solution for 15 minutes before beginning the run. Ordinarily the nitrogen was bubbled through water before passing through the test solution, but when alcoholic solutions were studied, an alcohol-water solution of the same concentration x a s used. When check runs were made on the same solution, nitrogen was bubbled through the solution for 5 minutes before beginning the check run. The effect of temperature on the first wave of sample 1, over the range of 3' to 38" C., was to increase the diffusion current an average of 1.65% per degree. Since this temperature coefficient is nithin the normal range, it appears that the humic acids behave normally with respect to temperature. ,411 subsequent data were obtained a t 25' f 0.5" C. T h e effect of mercury drop time on diffusion currents is shown in Table I1 for the first and third waves. il normal decrease in diffusion cuirent with increasing drop time is observed. Further studies on the effect of drop time indicated that more reproducible rvaves were obtained a t 4 seconds and, although greater diffusion currents were observed a t smaller drop times, 4 seconds wae chosen as standard.

ANALYTICAL CHEMISTRY

364 Table 111. Effect of Humic Acids on Reduction of Thallium(1)"Ion TlCl Molality

Humic Acids Gram per Lit&

i d , pa. Calcd. i d , N U . 0.000 4.55 ... 0.025 4.77 4.82 0.100 5.40 5.25 0.200 5.69 5.79 ... 0.025 0.27 .. .. .. ... 0,100 0.70 ... 0.200 1.24 ... Clark and Lubs borate buffer 0.1M;P H 9.0;potassium chloride: 0.25M.

0.001 0.001 0.001 0.001

'

a

and the thallium wave were superimposed and could not be distinguished. When the concentration of the thallium ion was such that its wave was more than three times the humic acid wave, the slope of the thallium wave was practically unchanged and the wave height due to the presence of both reducing substances was additive within the limitations imposed by the inaccuracies of reading the humic acid waves (see Table 111). 4 s no decided variation in the thallium wave resulted from the addition of humic acids, it was concluded that the reduction waves for the humic acids are normal for these substances and that their dope and height are not affected by the possible presence of an adsorbed layer on the mercury droplets.

DISCUSSION OF RESULTS

The nitric acid-oxidized coal humic acids (sample 1) dissolved in buffered alkaline solutions gave polarograms of the type shown by Curve I in Figure 1. Previous reduction of the acids with Devarda's alloy resulted in polarogram 11. Crider more favorable conditions, the unreduced acids exhibited a t least three reduction waves. Polarograms of this type are shoxn in Figure 2, .which also shows the effect of adding varying amounts of potassium chloride as a supporting salt. All of the waves obtained under varying conditions were of the sloping character shown in Figures 1 and 2, but considering the heterogeneous character of the organic matter from which coals xvere made, it is significant that the waves are as well defined as they are. Possibly the sloping character of the waves reflects the heterogeneity of the general structure of the humic acid molecules, although the appearance of discernible waves must be due to the reduction of characteristic structures present in the majority of the molecules. The mercury droplets falling to the bottom of the electrode chamber showed a remarkable resistance toward coalescence, suggesting that humic acids or ions were adsorbed on the surface of the mercury droplets (8). As this might have an adverse effect on the polarography of the humic acids, an investigation of the behavior of the acids in the presence of a "pilot" ion, which gives a well-defined wave in the same region, was made. For this purpose the thallous ion was selected, not only because it meets the requirement above but also because it is one of the few ions that does not precipitate the humic acids. In buffered alkaline solutions the first iwve of the humic acids

i -

l

'

l

~

l

'

l

'

l

'

l

'

l

'

l

~

l

cow w a i c t m s : o o % q / i pH-95

-

i-

EE-

+-

2

OM

KU

! W

02

04

Figure 2.

06

08

IO 12 WTE~lAL,-m?s

1.4

IS

(8

ZC

Effect of adding potassium chloride

The addition of ethyl alcohol to the solvent medium also had a pronounced effect upon the reduction waves of the humic acids. The polarographic constants obtained from solutions containing 0.500 gram of humic acids and up to 50% alcohol are shonn in Table IV. Since the addition of alcohol to aliquots of the original solution, made with varying amounts of 0.1N sodium hydroxide and 0 . l X boric acid, resulted in progressively more alkaline solutions, the p H of each solution is recorded also. The effect on the first wave of adding alcohol was t o increase the negative half-wave Table IV. Polarography of Alcoholic Solutions of Humic Acids" Dotential. but as the addition Alcohol of alcohol increased the pH of p H of p H of H Concn., p H of the solutions, which also has vel. % soln. - E ~ / , id Soln. - E ~ / ~ i d Soin. - ~ l / t id k'oln. - ~ : i r id First Wave the effect of increasing the 0 8.2 0.62 1.15 8.4 0.63 1.22 9.0 0.67 1.44 9.6 0.75 2.00 potential, it may be presumed 10 8.4 0.63 1.07 8.8 0.63 1.04 9.3 0.68 1.46 9.9 0.77 1.91 that the primary cause of the 20 8.6 0.62 0.81 9.0 0.66 1.08 9.5 0.70 1.41 10.1 0.77 1.79 30 9.0 b ... 9.6 0.67 1.01 9.8 0.73 1.45 10.4 0.76 1.90 shift was the change in pH. 40 9.3 C ... 9.8 d 10.1 0.74 1.66 10.7 0.78 1.75 50 9.7 a ... 9.9 0 . m i.'i4 10.3 0.76 1.59 11.0 0 ... On the other hand, 40 to 50% Second Wave concent,ration of alcohol in0 8.2 h' ... 8.4 1.45 0 9.0 1.42 Q 9.6 1.57 1.10 duced this wave to split into 10 8.4 1,45 0.17 8.8 1.54 0.51 9.3 1.47 0.46 9.9 1.56 1.24 20 8.6 I 0.18 9.0 1.59 0.69 9.5 1.53 1.19 10.1 1.62 2.75 two waves a t the lower p H 30 9 0 1.62 0.41 9.6 1.65 1.88 9.8 1.65 3.00 10.4 1.62 3.42 40 9.3 b ... 9.8 1.66 2.98 10.1 1.64 3.32 10.7 1.60 2.85 values. The second of these 50 9.7 1.61 1.79 9.9 1.68 2.77 10.3 1.52 1.48 11.0 1.58 2 03 waves appeared in the range of Third Wave -0.9 to -1.1 volts. KO at0 8.2 0 ... 8.4 1.75 0.18 9.0 1.81 0.35 9.6 Q ... 10 8.4 g ... 8.8 ... 9.3 1.72 0.30 9.9 0 .,. tempt was made to establish 20 8.6 1.67i 0.18 9.0 1.75 0.40 9.5 1.74 0.53 10.1 0 ... 30 9.0 0 .. . 9.8 9.6 ... 9.8 0 . . . 10.4 Q ... the chemistry underlying this 40 9.3 ... g . . . 10.1 Q 10.7 9 ... behavior. The effect of alcohol 50 9.7 0 ... 9.9 . . . 10.3 1.76 0.90 11.0 0 Concentration, 0.500 gram per liter; no buffer. on the wave height on the first b Wave distorted. wave seemed to show no defiTWO waves (1)E l / % -0.69, i d 1.61; (2)EI/Z - 1.11, i d 0.82. d TWO waves (1)& I / % = -0.66, i d = 0.79;(2)E I / ~ = -0.91, i d = 0.65. nite trend. Very likely the a Two wavea, difficult to read. f Extended, b u t could be read. variations shown in Table I\Wave too small t o read. h Two small waves a t - 1.39 and - 1.57. are due essentially to the i Wave drawn out from - 1.40t o - 1.55. difficulties of measuring the j Also a fourth wave E112 = -1.81, i = 0.22. sloping waves. E

Q

-

'

V O L U M E 27, NO. 3, M A R C H 1 9 5 5

365

The second wave was best Table Y. Polarographic Constants of Humic Acids Dissolved in Aqueous Buffered developed in alcoholic solutions Solutions of Varying pH Containing 0.5M Potassium Chloride of 30 to 40% and a t p H values Concn. Humic Acids, pH 0.025 0.050 0.100 0.200 above 8.8 to 9.0. The halfGrams/Liter Obsvd. -El/: id -Ei/a id -Ei/i i d -E112 i, wave potentials tended t o inFirst Wave crease markedly with alcohol 7. . 5 0.59 1.36 0.80 Borate buffer 0.49 0.19 0.51 0.44 0.55 content at the lower p H values, 1.30 0.61 0.80 0.22 0.52 8.0 0.53 0.44 0.58 1.42 0.64 0.83 0.23 0.57 0.55 8.4 0.44 0.60 but a t higher values the effect 1.45 0.67 0.56 0.80 0.24 0.59 8.8 0.61 0.48 0.67 1.39 u as less important. Diffusion 9.0 0.78 0.60 0.46 0.58 0.28 0.63 0.67 1.35 0.77 Carbonate-borax buffer 9.4 0.60 0.26 0.59 0.48 0.63 currents rose rapidly with in1.23 0.68 10.0 0.77 0.63 0.42 0.25 0.62 0.66 1.53 0.71 10.4 0.83 0.S6 0.28 0.65 0.66 0.67 c r e a s i n g c o n c e n t r a t i o n of 1.37 0.72 10.7 0.79 0.27 0.66 0.67 0.61 0.68 1.53 alcohol to 30 to 40% and then 0.73 1.02 Disodium phosphate buffer 10.9 0.32 0.68 0.67 0.71 0.60 1.51 0.75 11.3 0.91 0.36 0.70 0.68 0.72 0.60 declined. S o direct proof is 1.38 0.74 0.86 11.7 0.72 0.72 0.33 0.74 0.61 1.50 0.80 Yo buffer 0.95 12.5 0.29 0.76 0.78 0.55 0.80 available, hut it is suspected T h u d Wave that in the higher alcohol con1.68 1.93 1.76 8.40 1.65 1.07 centrations the humate ions Borate buffer 7.5 1.63 0.53 1.67 1.11 1.70 3.06 8.0 1.62 0.26 1.63 0.47 tend to associate ( I S ) in a way 8.4 1.66 0.71 1.69 2.18 1.61 0.18 1.68 0.30 1.18O 1.67 0.33 1.68 8.8 1.62 0.15 1.65 0.24 that interferes with the second 1.60* 1.08' 1.72 0.33 9.0 1.61 0.14 1.66 0.10 0.37 1.69 0.25 1.73 1.65 0.09 wave reduction to a greater exCarbonate-borax buffer 9.4 1.63 0.11 0.44 1.69 0.25 1.72 10.0 1.62 0.10 1.66 0.08 t>ent than the first wave, be10.4 1.63 0.07 0.40 1.67 0.27 1.72 1.65 0.18 0.50 1.69 0.23 1.71 10.7 1.64 0.10 1.66 0.12 if ''lore 0.51 1.68 0.30 1.69 Disodiumphosphatebuffer 10.9 1.62 0.15 1.69 0.18 0.55 1.67 0.26 1.69 1.63 0.14 1.63 0.16 added, sodium humates pre11.3 0.48 1.67 0.20 1.69 11.7 1.64 0.20 1.61 0.11 1.67 0.45 cipitate. S o buffer 12.5 1.59 0.18 1.64 0.32 1.62 0.23 The third wave was erratic Total id for coalesced second and third waves. b Rerun a t a lower current range (2)a n d two waves observed. (1) E l / * - 1.46,i d 0.47; (2)Ei/z = - 1.74, and poorly defined in alcoholic id = o,74, solutions. U s u a l l y , o n l y a slight inflection indicated its presence and often the appearance of a niaximurn or dip in the trace, as shown in Figure 2, rate of t h e electrode reaction in which the limiting current is not a linear function of concentration (7). Considering the interfered with measuring the wave. Attempts were made to improve the character of the reduction strong tendency of humic acids to associate, it seems probable waves in alcoholic solutions by the addition of salts, but were that the effect of concentration and perhaps p H on both halfahandoned because of the salting out effect. The addition of wave potentials and diffusion currents can be explained on this salts to aqueous solutions, however, was marked and was studied basis. The tendency for the diffusion currents to rise viith inin greater detail. Potassium chloride gave the best results creasing p H is shown in Figure 3. of those tried, which include potassium bromide, potassium nitrate, and sodium sulfate. The effect of increasing concentration on the first xvnve was to increase the wave height, as shown - P k FRCM P I T T S W SEAM COAL in Figure 2. Small concentrations tended to develop the second wave, while larger concentrations, 0 . 5 M , produced a striking improvement in the third wave a t the expense of the second. On the basis of these results a detailed study of aqueous solutions $containing 0.531 potassium chloride was made. Polarographic 8 constants for the first and third waves are given in Table 1 ' for varying concentrations of humic acids in buffered solutions of varying p H values. The data in Table V show that the half-wave potentials of the 6 first wave were not constant when either the concentration of the ? humic acids or the p H was varied and tended to shift t o more negative values with increasing concentration and pH. The shift of half-wave potentials t o more negative values with increasing p H has been discussed by Kolthoff and Lingane ( 9 ) , who note that this effect appears to be connected n-ith the 64 -L I de Ib ,? I 1I14 1I S ' i'B 20 mechanism of reduction. The diffusion currents, also, are not FOTENTW, -Mlh proportional to the concentration of the humic acids. Instances Figure 4. Comparison ofpolarograms of air-oxidized and of thip kind are known and appear t o be governed by the actual nitric acid-oxidized bituminous coal humic acids

-

-

kr

1r

75

85

I05

1 L5

PH

Figure 3.

Diffusion current of first w-ave cs. pH

LeJ

The structural grouping responsible for the first wave appears to be the nitro group introduced into the humic acid molecules by the concentrated nitric acid used in oxidizing the coal. The observed half-wave potential not only agrees with reported values of known nitro compounds and reduction with Devarda's alloy eliminates this wave, but when nitric acid is replaced by another method of oxidation, such as air oxidation, this wave does not appear. This is shown in Figure 4, in which polarograms obtained from both nitric acid-oxidized and air-oxidized, Pittsburgh-seam coal humic acids, samples 2 and 3 of Table I, are compared under identical conditions. The polarogram of the

ANALYTICAL CHEMISTRY

366 air-osidized coal humic acids shows no wave in the -0.6 to -0.i-volt region characteristic of the nitric acid-oxidized coal humic acids. Two other waves do appear at, about -1.0 and - 1.65 volts. The first wave seems t o be characteristic of airoxidized c o a l h u m i c acids and no doubt is caused by the reduction of a group introduced by air oxidation. The second wave appears in the same

where AT = molecular weight n = 4 electrons R = 8.315 X 107 ergs per O I