Tải bản đầy đủ - 0trang
Chapter 7. STRUCTURAL CHEMISTRY OF SOIL HUMlC SUBSTANCES
T. FELBECK, JR.
The progress in almost any area of scientific investigation depends
to a large extent on the adequacy of the techniques used in its investigation. This is particularly true in the study of the structural chemistry
of soil humic substances. The confusing and often contradictory results
that have been obtained are the direct result of the inadequacies of the
techniques of extraction, fractionation and purification, degradation, and
final isolation and characterization of the products obtained.
Therefore, it is the purpose of this review to discuss improvements in
old techniques and developments of new techniques that have occurred
in slightly more than a decade since the last review of the chemistry of
soil organic matter in this series (Broadbent, 1953).
Although the principal emphasis is on techniques, the only way these
techniques can be evaluated is by the results that are obtained. At the
end of each section an attempt will be made to evaluate the results and,
insofar as possible, point the way in which, in the reviewer's opinion,
further progress might be made. Finally, a summary will be made of
those characteristics of soil humic substances that seem most firmly established and how these characteristics may best be fitted into some current
hypotheses of the molecular structure of humic substances.
Many reports have appeared on all phases of research with soil organic matter, but this review will be limited to those that appear to
provide definite insight into the humic molecular structure. Likewise,
there have been many reports on the characteristics and formation of
model compounds that are supposed to be related in one way or another
to actual soil humic substances. As far as the reviewer is aware, none
of these reports has shown a model substance to be identical with a soil
humic substance, and therefore only those that appear to have a direct
relation to soil humic substances will be included.
Soil organic matter can be conveniently divided into two groups:
nonhumic substances and humic substances (Dubach and Mehta, 1963;
Kononova, 1961; Scheffer and Ulrich, 1960).
Nonhumic substances include all those classes of compounds occurring in plants and microorganisms that appear to have relatively
definite characteristics. In this group would be included carbohydrates,
proteins, fats, waxes, resins, pigments, and low molecular weight compounds. Quantitatively speaking, most of these could be relatively easily
attacked by soil microbes when in the pure state, and have a rather rapid
STRUCTURAL CHEMISTRY OF SOIL HUMIC SUBSTANCES
turnover in the soil. This appears to be true even though a very small
percentage of such compounds as amino acids can survive through
geologic time periods when protected in an anaerobic environment by
adsorption on clay minerals or on other kinds of organic compounds
( Abelson, 1963).
Humic substances are those materials produced in soils that are either
yellow or brown-to-black colored, acidic, polydisperse substances of
relatively high molecular weight. Based on solubilities, the group is
generally divided into three classes: fulvic acid, which is thought to be
of the lowest molecular weight and which is alkali and acid soluble;
humic acid, of median molecular weight, alkali soluble and acid insoluble; and humin, apparently of the highest molecular weight and
insoluble in both alkali and acid except under the most drastic conditions.
It has generally been believed that these materials are chemically homogeneous, and are heterogeneous only as to molecular weight. However,
recent evidence, which will be discussed later, indicates that fulvic
acid might also be chemically distinct from the other two classes.
Similar kinds of compounds appear to be present in geologic deposits
other than soils. As is the case with model substances, it has not been
shown to date that geologic humic substances are identical to soil
humic substances, although a much stronger case can be made for such
similarity than can be made for model compounds.
A definite word of caution must be added at this point. It might
appear from the above definitions that there are clear-cut differences
between nonhumic and humic fractions of soil organic matter and among
the three classes of humic substances. Unfortunately, such does not appear
to be the case. As far as is known continuous gradations exist among the
chemical characteristics of degraded lignins, tannins, and humic substances as well as among and within the fulvic acid, humic acid, and
humin fractions of many soils. At the present time these fractions are
largely defined on an operational basis, although there appears to be
some progress away from this kind of definition. As Dubach and Mehta
(1963) have stated, it is possible that perhaps no two molecules of
humic substance are exactly alike.
A number of reviews representing a variety of views on soil organic
matter studies have appeared in the past decade (Bremner, 1954, 1956;
Broadbent, 1953, 1955; Dawson, 1956; Deuel et al., 1960; Dubach and
Mehta, 1963; Dubach et al., 1964; Duchaufour and Jacquin, 1964;
Erdtman, 1962; Flaig, 1960b; Fraser, 1955; Kononova, 1961; Mortensen,
G. T. FELBECK, JR.
1963; Mortensen and Himes, 1964; Scheffer and Ulrich, 1960; Steelink,
1963; Swain, 1963).
The problems of exbacting the various fractions of soil humic substances from the mineral portion of the soil and separating these organic fractions from each other have been reviewed by Dubach and
Mehta (1963). Little, if any, progress has been made in these areas of
research since the appearance of that review. The inherent difficulties
and improbability of success in these areas of investigation are well
summarized by the suggestion repeated above that it may be that no two
molecules of humic substance are alike. Suffice it to say that at the
present time there has been no evidence presented to indicate that a
definite fraction has ever been obtained from a humic substance. Each
fraction can be refractionated into subfractions by some other technique,
and this can be repeated, apparently, ad infiniturn. The nearest approach
to a satisfactory fractionation seems to be the method of curtain electrophoresis ( Burges 198Oa,b; Johnston, 1959). Two fractions appear to
be partly resolved by this procedure. One fraction, representing a relatively small portion of the applied sample, is light colored and highly
fluorescent, whereas the other, representing by far the larger portion of
the sample, is dark colored and noduorescent. Whether the light-colored
fraction differs in chemical structure from the dark-colored fraction or
whether it merely represents a different molecular weighst range is not
known at the present time.
The preponderance of data indicates that the molecular weights of
soil humic substances range from 2,000 or 3,000 for the alkali-soluble,
acid-soluble fractions to well over 300,OOO for the fractions insoluble in
both alkali and acid. The upper ranges are largely a matter of speculation,
since few, if any, useful data have been obtained in this region
(Dubach and Mehta, 1963).
For the purposes of this review it will be assumed that soil humic
substances in the higher molecular weight ranges are chemically homogeneous, but heterogeneous in molecular weight. The lower molecular
weight fractions, as will be suggested in the final section of the review,
appear to be somewhat different from the larger fractions, having almost
all of the oxygen atoms in functional groups and none in the central
units of the molecule. This appears to impart greater acidity to the fulvic
acid fraction than is possessed by the larger molecular weight fractions.
The central structure of fulvic acid probably consists entirely of carbon
STRUCTURAL CHEMISTRY OF SOIL HUMIC SUBSTANCES
and hydrogen whereas the central structure of humic acid also contains
oxygen and possibly nitrogen.
The elucidation by techniques presently available of the structure
of polymeric materials such as soil humic substances requires, at some
step in the process, that the polymer be degraded into a form that can
be completely characterized. From the degraded forms, along with data
on bond linkages, a clearer picture of the polymeric structure can be
Many kinds of degradative procedures have been applied to humus
fractions. The results of these procedures generally fall into one of two
classes. First, the procedure generally does not appear to degrade humus
to any appreciable extent, the yields of any products identified being of
the order of a few per cent or less. Second, the more drastic procedures
produce the opposite result, degrading the humus into such small molecular fragments, e.g., CO2, HzO, NH3, acetic acid, and oxalic acid, that
almost all information on the original structure is lost.
It is the reviewer’s opinion that these observations can be explained
if it is assumed that the bonds between the monomers within the polymer are essentially equivalent in strength to the bonds within the
monomer itself. If this is the case, only a small increase in the degradative ability of a procedure would be necessary to go from the state of
no reaction to the state of essentially complete destruction of the polymer. This assumption would eliminate from consideration those bonds
between monomers that are substantially weaker than the intramonomeric bonds. Included in this category would be peptide, glycosidic,
and ionic bonds, since these bonds generally do not occur within the
aromatic, alicyclic, aliphatic, or heterocyclic structures that are assumed
to be among the monomeric units in humic substances. Bonds that might
occur between monomers are C-C, ether, C-N-C,
and possibly esters
whose reactivity is reduced by H-bonding or steric hindrance. Additional
stability could be produced by multiple bonds between monomers.
Hydrolysis of polymeric materials usually involves suspending the
substance in an aqueous solution of an acid and heating the mixture for
a period of time. The degraded products are either soluble in the acid
solution or separate as a second phase. In either case they can be separated from the aqueous phase and examined free from the original
G . T.
Approximately one-third to one-half of the total organic matter in
most mineral and organic soils can be dissolved by acid hydrolysis.
Included in the soluble products are a variety of sugars, amino acids,
uronic acids, and pigmented molecules. In podzolic soils the organic
matter in the BH horizon is almost entirely soluble in acid.
1. Acid Hydrolysis of Podzol BH Organic Matter
Coffin and DeLong (1960) extracted the BH horizon of a Quebec
Podzol with 0.1 N 8quinolinol and fractionated the product with laurylpyridinium chloride. They attempted acid hydrolyses of the various
fractions using HCl and H2S04, under reflux, for 6 to 48 hours with
acid concentrations varying from 1.0 to 6.0N. In several fractions, dark
brown precipitates not appreciably soluble in water were found. Paper
chromatograms of the hydrolyzates revealed traces of ninhydrin-reactive
substances and, in one fraction, very small amounts of several sugars.
Likewise, paper chromatograms revealed negligible amounts of phenolic
Jakab et al. (1962) hydrolyzed humic and fulvic fractions of a Podzol
BH horizon with water, HCl, H2S04,and HC104 for 16 hours at 120°C.
Ether extracts of the hydrolyzates represented 0.5 to 2.5% of the organic
fractions. In these extracts protocatechuic acid, p-hydroxybenzoic acid,
vanillic acid, and vanillin were detected. A lignin preparation treated
in the same manner yielded similar products. The authors concluded that
these aromatic products could have been derived from the lignin impurities in the humic substances.
2. Acid Hydrolysis of Mineral and Organic Soils
Waldron and Mortensen (1961) rduxed a Brookston silty clay loam
with 6 N HC1 and fractionated the hydrolyzate on a curtain electrophoresis apparatus. They examined the fractions obtained for a-amino N,
amino sugars, total N, and organic matter (by dichromate oxidation).
However, no positive identification of products was made.
In a later study with the same soil (Waldron and Mortensen, 1962)
50 ninhydrin-positive compounds were detected, 25 of them being identified as specific amino acids.
Kosaka et al. (1961) subjected humic acids from volcanic ash soils
to prolonged reflux with 6 N HCl. This procedure reduced the CH30and N content of the humic acid. About 16% of the total C in the humic
acid was released after treatment at 100°C. for 5 minutes; further treatments with stronger acid up to 16 hours at 110°C. released only a slight
additional amount of C.
STRUCTURAL CHEMISTRY OF SOIL HUMIC SUBSTANCES
Swaby (1960) attempted to degrade humic acids from Australian
soils with prolonged acid treatment at high temperatures. The hydrolyzates were examined for phenols and organic acid without notable
Coulson et al. (1959) hydrolyzed peat humic acids with 6 N HC1 for
24 hours at 100°C. and reduced the hydrolyzates electrolytically. Paper
chromatograms were dipped in FeC13/K3Fe( CN ) to reveal phenols, but
the quantity of phenolic materials was not determined. In addition sugars
and amino acids were detected in acid hydrolyzates. These workers concluded that sugars were not an integral part of the humic central structure, but that phenols and amino acids were, lending support to Swaby’s
(1958) theory that humic acids consist of phenols linked together by
In a preliminary acid hydrolysis of the total organic matter in a
muck soil with 3 N H2S04 at 90°C. and 72% H2S04at room temperature
(Felbeck, 1!365), it was found that about 32% of the C, 50% of the N,
36% of the 0 S, and 39% of the CH30- were dissolved by the more
dilute acid whereas about 17% of the C, 27% of the N, 36% of the 0 S ,
and no CH30- were dissolved by the 72% acid. No products were
isolated from the hydrolyzates, although in each case it was observed
that a substantial quantity of nondialyzable pigment was dissolved.
W. H. Conrad (unpublished data, 1963) examined the 3 N H2S04
hydrolyzate of this soil by paper chromatography and detected six sugars.
Farmer and Morrison ( 1960) examined the effect of acid hydrolysis on
the infrared spectra of peat humic acid. The hydrolysis removed the
sharp bands of lignin, and the absorption at 6.2 p was increased. No
attempt was made in this study to characterize the material chemically.
From the results discussed above, it appears that acid hydrolysis is
quite effective in hydrolyzing various carbohydrates and proteins in soil
organic matter, but has little effect in hydrolysis of the humic acid
fraction. Therefore, it can be assumed that glycosidic and peptide bonds
of the type occurring in carbohydrates and proteins are not those
existing in soil humic substances (except for those protein fragments that
may be present). Prolonged acid hydrolysis releases additional small
amounts of amino acids from humic acids, and it is not known whether
these amino acids are derived from proteinaceous material occluded in
humic substances or whether the amino acids are an integral part of
the humic molecule. The general consensus seems to be that carbohydrates are not a fundamental part of the humic molecular structure.
C. T. FELBECK, JR.
B. ALKALINE HYDROLYSIS
Alkaline hydrolysis is considered in this review to include both hydrolysis with relatively dilute aqueous alkaline solutions and alkali fusion.
Hydrolysis with dilute aqueous alkaline solutions is carried out in
somewhat the same manner as acid hydrolyses except that foaming is
often a difficult problem under reflux conditions. In addition, the presence
of air often causes concurrent oxidations to occur (Bremner, 1950).
Weedon (1963) has reviewed alkali fusion and some related processes
as applied to the elucidation of organic structures.
1. Alkaline Hydrolysis of Podzol BH Organic Matter
In the study discussed previously Coffin and DeLong ( 1960) degraded
the Ba humates of Podzol BH extracts by fusion with KOH in presence
of air. The ether extracts were analyzed by paper chromatography, and
several phenols and phenolic acids were detected. Substances actually
identified were p-hydroxy-, 2,4-dihydroxy-, m-hydroxy-, and 3,5-dihydroxybenzoic acids; these accounted for approximately 12% of the total
organic matter in the soil. Of particular interest was the fact that the
last two acids are not generally found in lignin degradation products
but were thought to be derived from microbial products. In addition to
the identified substances, seven other phenols were detected but not
identified. These unidentified phenols were found in about the same
amounts as the identified phenols. Thus about one-quarter of the total
organic matter could be accounted for as being derived from phenolic
Jakab et al. (1963) degraded the humic acid from a Swiss Podzol BR
horizon with 5 N NaOH at 170 to 250°C., with and without CuSO4.
Over 30 phenolic compounds were detected, probably of both lignin and
microbial origin. Yields of phenols accounted for 6% of the total organic
matter for reactions at 250°C. but only 2% for the 170°C. study.
Podzols from England and California were subjected to fusion by
KOH and NaOH
CuS04 by Steelink et al. (1960). Ether extracts of
the reaction mixture were examined qualitatively for lignin degradation
products. Protocatechuic acid, catechol, and resorcinol were definitely
identified, and vanillic acid and a phloroglucinol derivative were provisionally identified. No estimate was given as to what fraction of the
total organic matter was accounted for by these products.
Later Greene and Steelink (1962) obtained 20% ether-soluble products from the same two Podzols using the CuO-NaOH oxidation process.
However, they were able to identify compounds accounting for only
about 2% of the organic matter, among which were both resorcinol-
STRUCTURAL CHEMISTRY OF SOIL HUMIC SUBSTANCES
derived and guaiacyl-derived compounds in about equal amounts. They
concluded that other plant polyphenols were of equal significance to
lignin as possible sources of humus precursors. On the basis of their evidence microbial processes could not be excluded. They believed that
humic acid from a Podzol BH horizon included a variety of plant and
microbial phenols that copolymerized under oxidative conditions.
2. Alkaline Hydrolysis of Mineral and Organic Soils
Swaby (1960) subjected humic acids from several Australian soils to
KOH fusion and detected four to six phenols and two to three organic
acids, but no degradation products were completely identified.
Waldron and Mortensen (1961, 1%2) treated Brookston silty clay
loam with 0.5N NaOH at 105°C. In the extracts they detected 53
ninhydrin-positive compounds, of which 24 amino acids were identified.
Diphenylcarbazone was used to detect nucleic acid material, but the
separation of material was not complete and no definite identifications
Alkaline hydrolysis seems to be sufficiently vigorous to produce a
variety of phenolic substances from humic substances without completely
degrading the central unit of the molecule. From this observation it
appears to this reviewer that these phenolic units may not be a part of
the central unit but instead may be attached to the central unit by a
bond type hydrolyzed by vigorous treatment with alkali. In addition to
ether, C-C, and C-N bonds usually thought to be attacked by alkali
,is possible that ester bonds the reactivity of which
has been reduced by H-bonding and steric hindrance might also be included in this category. The presence of an ester bond was indicated by
Farmer and Morrison (19eO), who found evidence for ester linkages in
the infrared spectra of peat humic acids. I. V. Tyurin (as cited by
Kononova, 1961, p. 78) suggested that fulvic acids were combined with
humic acids by ester linkages, since this would explain ,the nonextractability of fulvic acid from the soil prior to treatment with alkali.
According to Fieser and Fieser (1956, p. 178) both the formation
and hydrolysis of the ester linkage are subject to acid catalysis. A possible
explanation of the observation that acid hydrolysis produces substantially
lesser amounts of phenolic materials than does alkaline hydrolysis might
be based on the fact that alkaline reactions favor complete hydrolysis
of ester linkages by removing one of the products, whereas under acid
conditions an equilibrium is set up between the ester and the cleavage
G . T. FELBECK, JR.
Humic substances appear to be quite sensitive to oxidation. In contrast to acid and alkaline hydrolyses in which low yields of products
are common, oxidative processes generally are too drastic, yielding
large amounts of COz and acetic and oxalic acids. Most efforts in the
application of oxidative procedures have been directed toward moderating the effect to the point where a suitable yield of large molecular
fragments is obtained. The application of oxidative methods to C-C
single bond fission has been reviewed by Potts (1963).
1. O ~ i d a t i ~with
Mehta et al. (1963) extracted humic acid from a Podzol BH with
ethylenediaminetetraacetic acid and oxidized it with 6, 15, and 30%
H202 for 5 days and 15 days at room temperature. They also allowed
humic acid to react with 30% Hz02 for 2 hours at 100°C. The principal
products were COz and HzO with a maximum of 5% of the original
material soluble in ether (probably various acids and phenols), Malonic
acid in a yield of 1.5%, in addition to phthalic acid, benzoic acid, and
oxalic acid, was identified in the peroxide oxidation products.
Savage and Stevenson (1961) examined humic acid from a Brunizem
soil by HzO2 oxidation and concluded that this method was not too satisfactory for producing characterizable intermediate products. No benzenoid structures were observed, and they suggested that humic acid did
not contain significant amounts of stable aromatic structures.
2. Alkaline Potassium Pemnanganate Oxidations
The A, and BH horizons of a Canadian Podzol (Prince Edward
Island) were examined by Wright and Schnitzer (1959a) using alkaline
permanganate oxidations. From the A, horizon, yields of 65% COP,
23% oxalic acid, and 2% acetic acid were obtained, 10% of the material
being resistant. From the BH horizon the yields were 92% COz, 7%
oxalic, and 1%acetic acid. Less drastic procedures apparently produced
benzene di- and tricarboxylic acid in one fraction and aliphatic acids
in another. No pure products were isolated. The fact that no mellitic
acid was detected suggested that neither horizon contained complex
benzenoid ring structures.
In a later study (Schnitzer and Wright, 1960a) a series of Cs to Ce
dicarboxylic aliphatic acids were isolated from oxidation products of
the A,, but not from the BH horizon. Benzene carboxylic acids and
aliphatic monocarboxylic acids were detected in both horizons. They
concluded that the A, horizon contained appreciable amounts of aliphatic
STRUCTURAL CHEMISTRY OF SOIL HUMIC SUBSTANCES
and/or alicyclic, in addition to aromatic structures, whereas the BH
organic matter consisted predominantly of aromatic structures.
By means of gas chromatography the methyl esters of several benzene
carboxylic acids were quantitatively measured in the oxidation products
of the Canadian Podzol BH horizon by Schnitzer and Desjardins (1964).
The total yield of benzene carboxylic acids accounted for 0.4% of the
original organic matter. From this observation, the relatively high carbon
aromaticity, and the high level of phenolic hydroxyl groups, Schnitzer
and Desjardins concluded that relatively few of the aromatic rings in
this soil preparation were not substituted by hydroxyl or other electron
Similar results were obtained by Hayashi and Nagai (1961) in a
study of volcanic ash soils and low and high moor peats. They obtained
large amounts of COz, oxalic acid, and acetic acid. The difference
between these products and the total carbon content was assumed to be
“aromatic compounds.” No specific aromatic compounds were identised
among the alkaline permanganate oxidation products.
In another study on volcanic ash soils Kumada et al. (1961) qualitatively identified anthraquinone in the ether extract of an alkaline permanganate oxidation product.
3. Nitric Acid Oxidations
In the studies on a Canadian Podzol previously mentioned (Schnitzer
and Wright, 1960a, b) HNO, oxidation gave 5.5% picric acid plus
benzene carboxylic acids and aliphatic monocarboxylic acids in both
the A,, and B H horizons. Total yields of material identified were over
twice the yields of alkaline KMn04 oxidation. These workers suggested
that salicyclic acid could have been the source of picric acid, cyclohexanol
the source of adipic acid, glutaric acid, and succinic acid, and cyclopentanone another possible source of glutaric acid. The yield of acidsoluble products from the total organic matter of the A,, horizon was 31%
and the yield of similar products from the alkali-soluble fraction of the
BR horizon was likewise 31%.
Jakab et al. (1962) in a study of HNO3 oxidation of a Swiss Podzol
BH horizon, obtained 20% ether-soluble products. Only picric acid was
identified among the reaction products.
Hayashi and Nagai ( 1961) were able to obtain high yields (30 to
60% ) of ether-soluble products from HN03 oxidation of organic fractions
isolated from a volcanic ash soil and two peats. Among the products
qualitatively identified were nitrophenols, nitrobenzoic acid, and hydroxybenzoic acids. An interesting aspect of this study was the observation that whereas lignin was completely decomposed by HN03, up to
G. T. FELBECK, JR.
50% of the organic fractions from the volcanic ash soil were resistant to
HN03 oxidation (i.e., not made acid soluble). Organic fractions extracted from two peat soils were, with one exception, completely decomposed by HNOs.
4. Alkaline Nitrobenzene Oxidations
The technique of alkaline nitrobenzene oxidation, which has been
of great value in the elucidation of lignin structure, has been applied
to soil organic fractions by Morrison (1958,1963). He was able to identify
only about 1% of the organic fraction of mineral soils as lignin-derived
material, and about 5% of such products from peats. Morrison concluded
from these observations that it seemed unlikely that alkaline nitrobenzene oxidations would be of further value in the study of soil humic
5. Other Oxidative Techniques
A variety of other oxidative techniques has been applied recently to
the study of soil humic substances. Among these have been oxidations
with C102 ( Mehta et al., 1963; Murphy and Moore, 1960), NaI04 (Mehta
et al., 1963), HGlz
NaOH (Morita, 1962), and HI04 (Murphy and
Moore, lW), but none of these procedures gave identifiable products
in higher yields than the techniques previously discussed. An interesting
observation by Murphy and Moore was that calculations of the amount
of HI04 reduced by natural humic acid indicated that there were 15.6%
by weight of glycol linkages in the material.
Of the variety of oxidative techniques applied to soil humic substances, the HN03 technique gave the highest yields of ether-soluble
materials. Both aromatic and aliphatic products have been detected in
the ether-soluble fraction. The reagent does not appear to be specific
in its action, and careful control of conditions is required if extensive
decomposition to low molecular weight acids is to be avoided (Potts,
1963). Therefore, it does not appear possible at the present time to
ascertain the structural type from which the aliphatic acids were obtained. On the other hand, it seems clear that the aromatic structures
could only have been obtained from other aromatic structures of equal
or greater complexity existing in the humic materials. Further work with
HN03 oxidations on humic substances from several soils may be of
value, but any extension of work with the other oxidants reviewed would
seem to be far less promising.