the supramolecular structure of humic … science 2002... · the supramolecular structure of humic...

810

0038-075C/01/16611-810–832 November 2001Soil Science Vol. 166, No. 11Copyright © 2001 by Lippincott Williams & Wilkins, Inc. Printed in U.S.A.

HUMIC substances (HS), natural organic sub-stances that are ubiquitous in water, soil, and

sediments, are of paramount importance in sus-taining plant growth and controlling both thefate of environmental pollutants and the biogeo-chemistry of organic carbon (OC) in the globalecosystem (Piccolo, 1996). Despite their role inthe sustainability of life, the basic chemical natureand the reactivities of HS are still poorly under-stood. The scientific community of humic scien-tists has so far failed to provide a unified under-standing of this field of science, and there is still,therefore, a poor awareness of fundamental as-pects of humic structures and reactivities. Never-theless, the implications of the relevance ofawareness of HA structure should extend far be-yond the interests of a few chemists; HA struc-tures affect the ways that the soil ecosystem work,as well as the bioavailabilty of organic substances

(including potential pollutants) in the soil envi-ronment (Tate, 1999).

Most of the difficulties encountered in chem-ically defining the structures and reactivities of HSderive from their large chemical heterogeneityand geographical variability. The substances areundoubtedly mixtures that develop randomlyfrom the decay of plant tissues, from microbialmetabolism-catabolism, or from both. Thus, thechemistry is not only highly complex,but it is alsoa function of the different general properties ofthe ecosystem in which it is formed, such as veg-etation, climate, topography, etc. It is not surpris-ing that, despite the efforts of many excellent sci-entists in the distant and recent past (Kononova,1961; Stevenson, 1994), there is still much to bedone to achieve an appropriate awareness of hu-mic chemistry.

The objective of this contribution is to reporton recent experimental findings leading to a newunderstanding of the conformational nature ofHS and to mention the profound implicationsthat this may have on our understanding of soilorganic matter (SOM) functions and reactivities.

THE SUPRAMOLECULAR STRUCTURE OF HUMIC SUBSTANCES

Alessandro Piccolo

Dipartimento di Scienze Chimico-Agrarie, Università di Napoli “Federico II” 80055,Portici, Italy. E-mail: [email protected].

Received June 4, 2001; accepted Aug. 21, 2001.

The scientific understanding of the molecular size and shape of humicsubstances (HS) is critically reviewed. The traditional view that HS arepolymers in soil is not substantiated by any direct evidence but is as-sumed only on the basis of laboratory experiments with model moleculesand unwarranted results produced by incorrectly applying either analyt-ical procedures or mathematical treatments developed for purified andundisputed biopolymers. A large body of evidence shows an alternativeunderstanding of the conformational nature of HS, which should be re-garded as supramolecular associations of self-assembling heterogeneousand relatively small molecules deriving from the degradation and de-composition of dead biological material. A major aspect of the humicsupramolecular conformation is that it is stabilized predominantly byweak dispersive forces instead of covalent linkages. Hydrophobic (vander Waals, �-�, CH-�) and hydrogen bonds are responsible for the ap-parent large molecular size of HS, the former becoming more importantwith the increase of pH. This innovative understanding of the nature ofHS implies a further development of the science and technology for thecontrol of the chemistry and reactivity of natural organic matter in thesoil and the environment. (Soil Science 2001; 166:810–832)

Key words: Humic substance structure, size exclusion chromatogra-phy, supramolecular associations.

Traditional Paradigms in Humus ChemistryThe amount of HS in soils is several times

greater than the amount in waters.Up to 70–80%of the OC in mineral soils may be composed ofhumic material, with recognizable plant remainsconstituting a small percentage of the organicmatter (OM) of mineral soils. Even though theabundance of OC in HS is from 2 to 3 timesgreater than that in the terrestrial biomass, the lat-ter greatly influences the OM dynamics in soils(Insam, 1996). Biological activity rapidly decom-poses labile plant materials on entering aerobicsoil environments with adequate water supplies,but more resistant components transform slowlyin the same environment. The compositional di-versities and the differences in the modes oftransformation of the components make it ex-tremely difficult to define accurately the grossmixtures that compose SOM, or the dissolved(DOM),or the particulate organic matter (POM)of waters.

Because of the heterogeneity of humic mix-tures, simplification and reductionism must beadopted. Stevenson (1994), in summarizing pre-vious reports and definitions, stated that humusincludes a broad spectrum of organic con-stituents, many of which have their counterpartsin biological tissues. He distinguished betweennon-HS and HS, the former of which consists ofcompounds belonging to the well known classesof organic chemistry such as amino acids, carbo-hydrates, lipids, lignin, and nucleic acids, whereasthe latter are unspecified, transformed, dark col-ored, heterogeneous, amorphous and high mole-cular weight (MW) materials.

The classical definitions of HS are opera-tional only and are based on solubility propertiesin the aqueous solutions used as soil extractants.The generalized terms humic acids (HAs), fulvicacids (FAs), and humins cover the major fractionsstill used to describe HS components, but theboundaries between these fractions have notbeen yet clarified in chemical terms. Becausemodern analytical techniques for organic com-pounds were not available during the first half ofthe 20th century, the data that were availablefrom elemental analyses (C, H, O) and determi-nations of acidic functional groups suggested thatHAs and FAs from many different soils were rel-atively similar (Kononova, 1961). Such simplecorrelations, although not based on any molecu-lar understanding, encouraged scientists to con-sider humic fractions to be chemical entities hav-ing specific properties rather than complexmixtures of nonspecific compounds.

Many of the modern concepts of HS derivefrom theories illustrated in the book of Ko-nonova (1961). In a review of hypotheses ad-vanced predominantly by Russian scientists,Kononova introduced the concept of HS as a sys-tem of polymers, based on the observation thatelemental composition, optical properties, ex-change acidities, electrophoretic properties, andMW characteristics varied consistently with soilclasses. Using this concept, the various fractionsof HS obtained on the basis of solubility charac-teristics are imagined to be part of a heteroge-neous mixture of molecules, which, in any givensoil, range in MW from as low as several hundredsto perhaps over 300,000 Da, and exhibit a con-tinuum of any given chemical property (Steven-son, 1994).

It must be noted that despite the existence ofdata showing that the molecular dimensions (asmeasured by osmometry, viscometry, and diffu-sion) of some HS were scarcely beyond one ortwo thousand Da (Scheffer and Ulrich, 1960;Schnitzer and Khan, 1972), more reliance wasplaced on the early work of Flaig and Beu-telspacher (1958) who showed, using the ultra-centrifuge, that MW values were in the range of30,000 to 50,000 Da for HAs and about 10,000Da for FAs. One reason for such bias towardshigh molecular weight (HMW) structures maybe found in the historical hypotheses that consid-ered HS to be products of biologically-assistedsyntheses from compounds derived from degra-dations of lignin, polyphenols, cellulose, andamino acids. Evidence for this polymeric as-sumption was researched in many classical labo-ratory experiments that indicated possibilities foreither abiotic or biotic condensations of simplemolecules into humic-like materials (Kononova,1961). Some of these early laboratory studieshave been researched in more carefully definedconditions in recent times (Haider and Martin,1967; Martin and Haider, 1969; Flaig et al., 1975;Flaig, 1988; Hedges, 1988). However, no directevidence has ever been provided for the occur-rence of such polymer build-up processes in nat-ural soil systems.

The polymeric view of HS has included thegeneral concept of polydispersity (Dubach andMetha, 1963), in which the HS are made of poly-mers with different MW values, similar to thatwhich applies for other natural biological macro-molecules such as proteins, polysaccharides, andlignin. The experimental difficulties encounteredin isolating chemically homogeneous fractions ofHS was compared with those observed for bio-

VOL. 166 ~ NO. 11 SUPRAMOLECULAR STRUCTURE OF HS 811

polymers with varying molecular dimensionssynthesized in the living cell. This unwarrantedsimilarity supported the polydisperse polymericconcept of HS and justified the observation thatthese are mixtures of compounds.The concept ofHMW and polydisperse polymers became a par-adigmatic part of the descriptive definitions ofHS proposed thereafter (Schnitzer and Khan,1972; Aiken et al., 1985; Malcolm, 1990; Steven-son, 1994).

The assumption that HS are polymers, whichhas no sound molecular basis, has promulgated theuse of simple physical-chemical measurements tocharacterize HS. An example is the E4/E6 index,the ratio of the absorbance of HS at 465 nm to thatat 665 nm, introduced by Welte (1955) and repro-posed by Kononova (1961). This ratio supposedlyindicates a reverse relationship with progressivehumification as well as increased condensation as-sumed to produce large amounts of polyconden-sated aromatic-ring structures. Despite its wide-spread and continued application, the E4/E6 ratiohas repeatedly been shown not to hold the pre-dicted relationship with MW.At variance with thehypotheses related to this ratio, Campbell et al.(1967) had already indicated that humic materialwith the lowest mean residence time in the soilhad the highest E4/E6 ratio, and Anderson andHepburn (1978), using gel permeation chro-matography, showed that humic fractions withlarge molecular sizes and low E4/E6 ratios weremainly aliphatic, whereas those with molecularsize had the highest aromatic contents. Piccolo(1988) compared the E4/E6 ratios of various HSwith their gel permeation chromatograms andfound that the results were comparable only whenthe HS had been subjected to extensive purifica-tion. Summers et al. (1987) reported on the limi-tations of E4/E6 ratios and found that the ratiovalues varied considerably with the concentrationsof ultrafiltered fractions of HS. By comparing HSextracted from soils and soil particles before andafter long term amendments with organic wastes,Piccolo and Mbagwu (1990) found that theE4/E6 ratio values did not account for the increasein molecular sizes in amended samples as observedby gel filtration

Despite a lack of evidence, various reasons ledthe scientific community to accept the polymericconcept for HS. One of these reasons lies in theacceptance of the description by Staudinger(1935) of macromolecular polymers occurring inliving cells. Based on the Staudinger concept, itseemed convenient to assume that HS were alsopolymers biologically synthesized from plant tis-

sues components, notwithstanding that HS arisefrom cell death rather than cell biosynthesis, as oc-curs with other biomolecules.The rapid degrada-tion and decomposition in soil of biopolymersliberated from cell lysis after death is now a wellaccepted process from both a biological and athermodynamic perspective ( Jenkinson, 1981;Haider, 1987; Clapp and Hayes, 1999; Spaccini etal., 2001). Surprisingly, Staudinger’s view with re-gard to macromolecularity is still advocated by de-fenders of the polymeric nature of HS (Swift,1999).

A second reason for acceptance of the poly-meric concept is that a stable polymeric structurewould account for the refractory characteristics ofHS in soil, rather than considerations of physicaland chemical protection,as provided, for example,by interactions with inorganic soil particles. TheOC in stable humic materials is known (Steven-son,1994) to have extended residence times (from250 up to 3000 years) in soil. In addition, the clas-sical hypothesis of HS formation through con-densation between amino acids and componentsof degraded lignin promoted the assumption thatHS had a polymeric structure similar to that oflignin (Waksman, 1936). Lignin is known to bepolydisperse in MW, with values ranging from�1000 to several million Da (Goring, 1971), andits resistance to microbial degradation in soil hasbeen attributed repeatedly to its macromolecularstructures (Waksman, 1936; Amalfitano et al.,1992). Similarly, polymerization processes carriedout in laboratory or in confined conditions led toother classical hypotheses of HS formation, suchas the polyphenol theory (Flaig et al., 1975) andthe melanoidin theory based on the Maillard re-action (Maillard, 1912). Another aspect of HScontributing to the polymeric view is their col-loidal properties, which were related to those ofpolyelectrolytes in aqueous media (van Dijk,1972; Flaig et al., 1975). Most of the properties of polyelectrolytes, such as processes of floccula-tion and dispersion, responses to electrolytes, anddouble-layer behavior were also observed for HS,and thus concepts of HMW properties were read-ily attributed to HS.

The macromolecularity of HS, despite thewide acceptance of the concept, has never beendemonstrated unambiguously in chemical andphysical-chemical terms for soil-extracted humicfractions. The instrumentation for organic andphysical chemistry analyses was not generallyavailable in the past, nor was it completely reli-able. This situation has changed dramatically inthe present world of sophisticated biophysical

812 PICCOLO SOIL SCIENCE

technology. Thus, the polymeric paradigm of HSshould either be proven beyond any doubt orabandoned in favor of another description.

CONCEPTS OF CONFORMATIONALMODELS FOR HUMIC SUBSTANCES

Determination of the true MW of a humicfraction is central to HS research. Conformationalaspects of structures of HS (i.e., shapes and sizes),and ultimately their reactivities in soil and the en-vironment, are determined by MW. It is quite ev-ident from reviews that different methods formeasuring MW values of HS do not give the sameanswers (Wershaw and Aiken, 1985; Swift, 1989;Clapp et al., 1989;Stevenson,1994).Generally, thedifferences are not slight but can be several ordersof magnitude. The great differences in values havebeen attributed either to the variability of the HSor the intrinsic limitations of the methods whenapplied to polydisperse systems. However, thepolymeric paradigm has not been questioned inanalyses of the differences, and very few attemptshave been made to decrease significantly polydis-persity by classical chemical methods.

Much of the confusion with regard to MWdeterminations of HS has arisen as a result ofsedimentation-velocity and diffusion methods thatbecame fashionable with the upsurge of biochem-istry research in the 1950–1970 era. Despite theclear indication that such semiempirical methodsare not suitable for polydisperse systems because ofthe multiple diffusion coefficients and sedimenta-tion constants for different size particles (Wershawand Aiken, 1985), studies were carried out withwhole HA extracts, and MW values ranging from25,000 to more than 200,000 Da were reported(Stevenson et al., 1953; Flaig, 1958; Piret et al.,1960). Interestingly, Flaig and Beutelspacher(1968), when working with polydisperse humicsolutions, found that MW values varied from77,000 Da, when 0.2 M NaCl was added to de-press repulsive negative charges, to only 2050 Da in the absence of the background electrolyte.Though the latter is commonly added to polyelec-trolytes to reduce interferences from charge repul-sion in monodisperse systems,a jump of about twoorders of magnitude in MW for the same humicpolydisperse system could suggest a molecular as-sociation rather than an extreme interference insedimentation-velocity of polyelectrolytes.

In an attempt to decrease the polydispersity ofHS, Cameron et al. (1972a) fractionated soil-extracted HAs by ultrafiltration and by gel per-meation chromatography (GPC) extensively, andthey subjected the fractions to sedimentation-

velocity ultracentrifugation. The fractions ob-tained were more homogeneous than the unfrac-tionated solution, but these could still not be con-sidered to be monodisperse. They reported MWvalues for the fractions ranging from 2600 to asurprising 1,360,000 Da. It is most likely that theymeasured associations of smaller molecules ran-domly aggregated by different forces during thefractionations procedures. Moreover, Cameron etal. (1972a) overlooked the difficulties in assessingsedimentation coefficients for associations of mol-ecules such as HS.Measurement of sedimentationcoefficients of polydisperse materials that includesubunits invariably leads to erroneous values ofMW (Laue and Rodhes, 1990). Other studies us-ing sedimentation-velocity ultracentrifuge studies(Ritchie and Posner, 1982) and the even moremathematically sound equilibrium centrifugation(Posner and Creeth, 1972; Reid et al., 1990) alsoshowed polydispersity in HS, and, for this reason,confirmed the ambiguity of the MW values ob-tained by ultracentrifuge methods.

There is more reliability in MW obtained forFAs, or for OM dissolved in waters. There is ageneral agreement that the MW of these humicmolecules is in the 400–1500 range, as deter-mined by a variety of methods (Aiken andGillam, 1989; Wershaw, 1989; Stevenson, 1994).As will be seen below, the smaller discrepanciesamong MW values obtained for FAs can be as-cribed to their larger hydrophilicity (small veryacidic molecules), which prevents strong molec-ular associations by hydrophobic forces. How-ever, a process of molecular associations in solu-tion can produce a polydisperse system ofapparently HMW also for FAs. In fact, FA MWswhen analyzed by ultrafiltration and gel filtrationmethods, were consistently higher than whenmeasured with osmometry or cryoscopy (Thur-man et al., 1982).

Similar contradictions (as described for theMW values of HS) apply for the concepts ofshapes attributed to humic polymeric macromol-ecules. Globular shapes (Visser, 1964), flexible lin-ear configurations (Mukherjee and Lahiri, 1959),ellipsoidal shapes (Orlov et al., 1975), spheroidpolyelectrolytes (Ghosh and Mukherjee, 1971),and randomly folding long chains (Cameron et al.,1972a) have all been proposed to describe the everelusive polydisperse humic system. Ghosh andSchnitzer (1980) reconciled the different views bymeasuring surface pressures and viscosities of HSat different pH values and neutral salt concentra-tions and adapting the results to relationships (theFlory and Fox and the Staudinger equations) that


had been developed for real polymers. They ex-plained the observed behavior of HS (unchargedmatter at low pH and polyelectrolytes at high pH)on the basis of the polymeric theory and they pro-posed that HS are rigid spherocolloids at high sam-ple concentration and ionic strength and at lowpH, whereas at high pH values, low sample con-centrations,and ionic strength, they behave as flex-ible linear polymers. This understanding had twomajor flaws: it was based on studies with wholehumic extracts with full polydispersity and,despitethe lack of a direct knowledge of the real molecu-lar structure, the data were arbitrarily used in equa-tions specifically derived for polymers. Neverthe-less, this reversible coiling model for humicconfigurations soon became the most widely usedto describe HS, although it does not explain all ofthe behavior of HS.

In applications of small-angle X-ray scatter-ing to determine the particle sizes of whole HAsin solution, or of fractions separated by adsorp-tion chromatography using cross-linked dextrangels (see Wershaw, 1989), it was found that HSformed molecular aggregates in solution andtheir sizes were a function of pH. It was con-cluded that the various fractions were differentchemically and that the differences in aggregationbehavior were a reflection of the interactions ofdifferent bonding mechanisms. These findings,coupled with other results that showed that hu-mic fractions from different sources have surfaceactive properties (Hayase and Tsubota, 1983),have led to a description of HS that is an alterna-tive to the random coil polymeric structure.Wer-shaw (1986, 1993) proposed that HS consist ofordered aggregates of amphiphiles, composedmainly of relatively unaltered plant polymer seg-ments possessing acidic functionality. In thismodel, HS are aggregates held together by hy-drophobic (�-� and charge-transfer bonds) andH-bonding interactions, and the hydrophobicparts of the molecules are in the interiors withthe hydrophilic parts make up the exterior sur-faces. Ordered aggregates of humus in soils weredepicted to exist as bilayer membranes coatingmineral grains and as micelles in solutions.

Wershaw’s model represented a major break-through because it introduced the concept of ag-gregation of different particle sizes of humic con-stituents in contrast to the traditional view ofpolydisperse linear humic polymers. Neverthe-less, the micelle-like model did not yet solve theissue of MWs of HS. The spontaneous aggrega-tion of humic molecules into micellar aggregateswas advocated by other authors (Engebretson and

von Wandruszka, 1994, 1997) to explain fluores-cence quenching of pyrene, but the explanationof the results was still based on the polymeric na-ture of the aggregating humic molecules. How-ever, the classical concept of an ordered micelle ishardly applicable because of the heterogeneity ofHS: the CMC (Critical Micelle Concentration)reported in the literature for HS, in the range of1 to 10 g L�1 (Hayase and Tsubota, 1983), ismuch higher than those found for surface activecompounds giving regular micellar structures(Tanford, 1980).

Despite its limitations, the concept of aggre-gation of hydrophobic parts of HS could explain:

1) results from light-scattering which showedthat addition of Cu ions to dilute soil FA in-creased the amount of light scattered by thesolution (Ryan and Weber, 1982);

2) the increased solubility of nonpolar com-pounds in humic solution because of parti-tion/adsorption in the hydrophobic interiorof HS (Carter and Suffet, 1982); and

3) the further release of humic matter throughdialysis bags from an already extensivelydialyzed HS when this is treated with anamphiphilic compound such as acetic acid(AcOH) or an electrolyte (De Haan et al.,1987; Nardi et al., 1988).

The advances that the molecular aggregationmodel made for an understanding of the envi-ronmental behavior of HS (compared with thepolymeric paradigm) did not prevent recent hy-potheses, based on pyrolytic analyses of HS, indi-cating macromolecular structures for HAs withMW values up to about 100,000 Da (Schulten etal., 1998). Despite the many limitations inherentin the analytical pyrolysis of HS (Saiz-Jimenez,1995, 1996), compounds identified by pyrolysis-mass spectrometry techniques were arbitrarilylinked together by covalent bonds in computermolecular models to give graphical representa-tions of large branched polymers. Such largemacromolecules are even being proposed asmodels of humic structures and used to explainthe behavior of HS in soil (Schulten and Leinwe-ber, 2000).

SIZE EXCLUSION CHROMATOGRAPHYOF HUMIC SUBSTANCES:

HISTORICAL PERSPECTIVE

Gel permeation chromatography or low-pres-sure size exclusion chromatography on Sephadexcross-linked dextran gel columns, a method de-


vised to desalt and purify proteins, has been ap-plied extensively to humic fractions to evaluatemolecular sizes and to obtain more size-homoge-neous materials (see reviews of Wershaw andAiken, 1985; De Nobili et al., 1989). It soon be-came clear that separation by GPC is not a puresize fractionation (Determann and Walter, 1968)and that interferences may occur with HS,namely ionic exclusion and adsorption chro-matography. In the first process, electrostatic re-pulsion between the negative charges present onboth dissolved HS and the dextran gel enhancesthe chromatographic velocity,whereas in the sec-ond process, hydrophobic interactions betweenHS and the stationary phase retard the chromato-graphic elution of HS (Lindqvist, 1967). Theionic strength of mobile phases should be suffi-ciently high to prevent electrostatic interactionsbut not too high (�0.5 M) to drive hydrophobicinteractions (Chicz and Regnier, 1990).

Much of the inconsistency found in HS be-havior in gel permeation has been ascribed re-peatedly to either ionic (electrostatic) exclusionor to hydrophobic gel-solute interactions. Swiftand Posner (1971), by eluting a Sephadex G-100with distilled water, showed that increasing con-centrations of HS produced shifts of peaks fromhigh to low molecular size ranges. They disre-garded the possible hydrophobic retardation andqualitatively attributed the behavior they ob-served to the repulsion between HS and the neg-atively charged Sephadex.They also assumed thatrepulsion would become stronger if a decreasingHS concentration lowered the ionic strength.Us-ing a classical polyelectrolytic (polymeric) view,they claimed that “charged double layers on thesolute and the gel extend further into the solu-tion, resulting in effectively larger solute mole-cules and smaller pore sizes. Charge repulsion ef-fects therefore occur at great distances leading toincreased exclusion with decreasing sample con-centration.” However, in order to corroboratethese theoretical assumptions, the authors did notcarry out measurements of actual charge densityat different sample concentrations or in the gelbed. Notwithstanding the theoretical and exper-imental inconclusiveness of HS behavior in gelpermeation, buffers of high ionic strength wererepeatedly recommended as mobile phases tosuppress interferences caused by ionic exclusion(Swift and Posner,1971;Swift,1989;De Nobili etal., 1989).

High ionic-strength buffers reduce the molarvolume of HS in solution and, by thermodynam-ically favoring the hydrophobic associations of

humic molecules, invariably produce chromato-grams with a bimodal distribution.However, theyalso enhance hydrophobic adsorption on the gelsolid phase (Specht and Frimmel, 2000).Yonebayashi and Hattori (1987) showed that theaddition of 2 M urea to either phosphate or bo-rate buffers eliminates the hydrophobic adsorp-tion observed during the GPC of HS. An expla-nation for the urea effect will be given below.Adsorption of HS on gel columns has tradition-ally been accounted for either by salinity or bychanges in ionic strength (Lindqvist, 1967; DeNobili et al., 1989). Anderson and Hepburn(1978) attributed the appearance of new HS elu-tion bands to adsorption when sodium acetatewas added to the mobile phase and regarded thesebands as unspecified artifacts created by adsorp-tion effects. However, De Haan et al. (1987)demonstrated for the first time that HAs in salinesolution were able to diffuse more freely across adialysis membrane than when in solutions oflower ionic strength.After comparing dialysis andgel filtration experiments, they suggested thatchanges in elution profiles caused by ionic-strength variation had to be attributed to real al-terations of humic conformations rather than toany interactions between the gel and HS. Thiswas at variance with the classical explanations.

Further understanding came from the workof Yonebayashi and Hattori (1987) who studiedseveral HS by gel filtration through Sephadex G-75. The eluent was 0.1 M phosphate or boratebuffers in 2 M urea, and they thereby avoided in-terferences by both ionic exclusion and adsorp-tion on the column. They first found that by in-creasing the pH of the buffer eluent from 4.7 to11.2, the molecular sizes of the HS increased pro-gressively, although the fraction excluded at highmolecular-size (Vo) was larger at pH 4.7. Inter-estingly, the elution pattern at pH 4.7 showed ad-sorption of humic material beyond the total vol-ume (Vt) of the column. They attributed thebehavior at low pH to an aggregation process ofhumic molecules but still advocated ionic repul-sion as the reason for the size increase at high pH.However, their experimental conditions (largeionic strength and urea) should have excluded anelectrostatic effect.

Yonebayashi and Hattori (1987) also showedthat for both phosphate and borate buffers andfor all pH values from 4.7 to 11.2, the molecularsizes of HS increased significantly with time ofstanding (from 5 min to 24 h) in the buffer solu-tion before injection into the gel column. To ex-plain these findings, they had to invoke again a


process of molecular association into micelle-likeaggregates. In order not to contradict the poly-meric paradigm of HS, they assumed that associ-ation occurred between humic molecules oflarge and small molecular sizes. However, this ex-planation did not account for what they furtherobserved when HS were treated with ethanol-benzene, 6 M HCl, 0.5 M H2SO4, or 5 M NaOHbefore dilution in the buffer and separation byGPC. After the ethanol-benzene or acid treat-ments, the gel chromatograms were almost thesame as those for the original HA, but the frac-tion excluded increased with an increase ofstanding time in the buffer solution. Conversely,the excluded fraction disappeared after the alka-line treatment, and the gel chromatogram re-mained unchanged regardless of the standingtime in the buffer solution before injection. Al-though they invoked only micellar aggregation toexplain their findings, these results were clearly asign of conformational changes caused by the es-tablishment of hydrophobic interactions (see dis-cussion below). The conformational rearrange-ment could take place in the buffer solution forall pretreated samples but not in 5 M NaOH,where a full humic dispersion (disruption of allhydrogen bonds) and hydrophobic rearrange-ment had already taken place.

In an attempt to clarify the formation of mi-cellar aggregates,Yonebayashi and Hattori (1987)also measured the surface tension of HS. Theyfound that the surface tension of the whole HSmixture in neutral phosphate buffer containingurea decreased with an increase in sample con-centration. This suggested that molecular associa-tions increase the surface activities of humic ma-terials. In fact, four fractions isolated by GPCrevealed that only the first HMW size fraction be-haved as a surface active agent, whereas the threesmaller fractions showed low surface activity. Acluster analysis grouping the gel behavior of sev-eral HS with their chemical and physical-chemi-cal properties showed that HS with high surfaceactivities (and a large fraction excluded at Vo ingel filtration) were rich in long aliphatic chains,whereas HS with low surface activities (and smallexcluded fractions) were composed of aromaticrings, a high content of -COOH groups, and alow content of -OCH3 groups. Based on the re-sults of Yonebayashi and Hattori (1987), whichwere similar to those of Anderson and Hepburn(1978), it may be concluded that the associativenature of HS was proven by gel filtration studies.Fractions had different chemical compositions,and hydrophobic attractive forces between ali-

phatic components favor their association into ap-parently large and surface-active fractions.

The work of Ceccanti et al. (1989) has pro-vided new evidence that HS behavior in GPCwhere changes in ionic strength occurred cannotbe ascribed to simple electrostratic interactions(ionic exclusion), as for a polyelectrolyte (a poly-mer holding multiple charges). They fractionatedHS by ultrafiltration and gel filtration using bothwater and salt solutions as mobile phases andfound that 100% of the humic C was in the low-est molecular size fraction (less than nominal10,000 Da) when HS were ultrafiltered in 100mM of pyrophosphate/HCl buffer at pH 7.1.However, no humic C was in the same fractionwhen ultrafiltration was in water only,when 43%of the C was in the highest molecular size frac-tion (greater than nominal 100,000 Da). Con-trary to previous qualitative research, Ceccanti etal. (1989) did measure by isoelectric focusing thenegative net charge of HS fractions isolated bygel filtration. Despite the charge being very sim-ilar, refractionation of fractions in the SephadexG-50 column in water gave distinctly differentelution patterns. Moreover, fractionation by wa-ter elution from the polydextran Sephadex,which has some residual negative charges, gave amolecular size distribution in accordance withthat obtained on the uncharged PM-10 (polysul-fone) ultrafiltration membrane. They concludedthat gel filtration occurred by a true size frac-tionation rather than by a charge-dependent elu-tion. The work of Ceccanti et al. (1989) alsoshowed that the strength and the shape of humicassociations controlled the elution volumes of HSfractions.

A major problem in using size exclusionchromatography to determine the sizes of humicfractions is the lack of adequate standards to cali-brate the gel column. A calibration of Sephadexcolumns with presumably size-controlled humicfractions differed substantially from that obtainedwith globular proteins (Cameron et al., 1972b).This indicated that the hydrodynamic behaviorof HS is different from that of globular macro-molecules, and their molecular size may be moreapparent than real and could be ascribed to mol-ecular aggregation phenomena. Hayes (1997) re-ported that the ability to obtain humic fractionsof homogeneous size by repetitive gel fraction-ation is not effective because reprocessing offractions results in the separation of smaller sizedcomponents.

The difficulty of obtaining MW values of HSby gel-filtration was noted by Reuter and Perdue


(1981). They measured by osmometry a value of1231 Da for the number-average MW of the HAfraction excluded from G-50. The publishedglobular-protein exclusion limit for the gel is30,000 Da.Again, the reason for the exclusion ofthe humic material at such an apparent high MWcut off may be attributable to the molecular ag-gregation of small molecules into apparent largesizes. This interpretation is substantiated by thelack of differences between the IR spectrum ofthe original HA and the spectra of fractions iso-lated by means of gels with different MW cutoffvalues (Reuter and Perdue, 1981). The associa-tion of small humic molecules into different ag-gregate sizes may have also been the reason whyCornel et al. (1986) could not find sufficient sim-ilarity between the diffusion behavior of HS andthat of synthetic polymers of known composi-tions such as polyethylene oxides and polystyrenesulphonates. Further questions arose when Sum-mers et al. (1987) did not find differences in theIR spectra of HS size fractions isolated by ultra-filtration, a separation technique that is slowerthan the velocity by which humic molecules ag-gregate or disaggregate in solution.

EVIDENCE FOR SUPRAMOLECULARASSOCIATIONS OF SMALL

HUMIC MOLECULES

Using low-pressure size exclusion chromatogra-phy (SEC), Piccolo et al. (1996a and b) reportedthat the 280 nm absorption of HAs was reversiblyshifted from high to low molecular size ranges(column total elution volume, Vt) when organicacids were added to lower the pH of a humic so-lution from 9.2 to 2 before the elution in an 0.02M alkaline borate buffer. To explain their resultsthey suggested that, instead of being stable poly-mers, HS at neutral or alkaline pH values aresupramolecular associations of relatively smallheterogeneous molecules held together by weakdispersive forces, such as van der Waals, �-�,CH-�, interactions. The addition of organicacids altered such unstable humic conformationsthrough the formation of energy rich hydrogenbonds, and the subsequent chromatographic elu-tion separated the resulting smaller subunits andprevented the reassociation that would have oc-curred in static conditions.

Supporters (Swift, 1999) of the traditionalpolymeric model of HS have criticized the aboveresults, not on the basis of an experimental repli-cation but on theoretical and qualitative interpre-tations of gel-solute interactions and interfer-ences caused by a supposed polymeric form of

borate. However, the results of Piccolo et al.,(1996a and b) could not be attributed to a buffer-ing action of the organic acid toward the alkalineeluent because the amount of the different or-ganic acids varied by two orders of magnitude,but the shift to larger elution volumes remainedthe same for all acids. Nor could an elution delaycaused by a solid deposition on the gel and sub-sequent resolubilization by the eluent have beenthe cause of the shift since the treated samples re-mained soluble at low pH and entered the chro-matographic elution immediately after deposi-tion on the column.Nevertheless, if this had beenthe case, a progressive neutralization of the acidicbuffering capacity by the alkaline eluent wouldhave caused a smearing out in the column of thepolydisperse humic mixture, and a diffuse chro-matographic band from lower to larger elutionvolumes, instead of the sharp peak at the totalvolume (Vt) of the column. Furthermore, theionic strength effect (De Nobili et al., 1989)could not be invoked for the reversible peak shiftbecause elution in an ionic strength quencher,such as a borate buffer 10 times more concen-trated, gave the same chromatographic changeupon AcOH and KOH additions.

Piccolo et al. (1996a and b) considered theirfindings to be an expression of the associative na-ture of relatively small humic molecules that self-assemble only into apparent high-molecular sizematerials. This interpretation was in accordancewith previous research, which showed that HSbehaved as molecular associations when studiedby SEC (De Haan et al., 1987; Yonebayashi andHattori, 1987;Ceccanti et al., 1989;Piccolo et al.,1990a). Moreover, laboratory observations haveindicated that when AcOH is added to HS thathave been already extensively dialyzed, furthersmall size components are released during subse-quent dialysis (Nardi et al., 1988). These lowmolecular size fractions are a product of a con-formational rearrangement and a chemical com-position different from the bulk HS. The sepa-rated fractions were found to stimulate specificbiological properties in plants and were more bi-ologically active than the whole humic materialsfrom which they were separated (Piccolo et al.,1992).

An experimental replication of the work ofPiccolo et al. (1996a and b) that applied high per-formance SEC was reported by Varga et al.(2000). They obtained the same results as Piccoloet al., (1996a and b), but they observed that by us-ing organic acids to lower the pH of humic solu-tions before injection from 9.2 to 1–3, the shift of


absorption to lower molecular size ranges mayhave been caused by hydrophobic interactions onthe column upon protonation of the humic mat-ter and the consequent change in ionic strength.A contribution to SEC of hydrophobic interac-tions on the column has been long recognized inthe gel-filtration studies of HS (Lindqvist, 1967)and of other biomolecules (Chicz and Regnier,1990). Nevertheless, hydrophobic interactions onsize-exclusion gels have often been usefully ex-ploited to fractionate humic aggregates into sim-pler molecular associations that were character-ized chemically more easily than the bulk humicmaterial (Anderson and Hepburn, 1978;De Haanet al., 1987;Yonebayashi and Hattori, 1987).

It was necessary to verify the innovative expla-nation of the conformational changes of HS usingadditional experimentation, and High-PressureSize Exclusion Chromatography (HPSEC) wasadopted for further investigation. HPSEC is, for anumber of reasons, a more valid chromatographicsystem than low-pressure GPC (Piccolo andConte, 2000;Piccolo et al., 2001a).The major ad-vantages of HPSEC are:

1) high reproducibility of chromatograms (5%);2) relatively higher rapidity of analysis (about 60

minutes per chromatogram);3) longer column life (more than 1000 injections

per HPSEC column); and4) higher sensitivity to chemical changes in in-

jected samples because of a much lower load-ing mass.

Conte and Piccolo (1999a) have comparedtwo commercial HPSEC columns with respect totheir capacity to measure the molecular sizes ofHS accurately and precisely.

In the HPSEC mode,Piccolo et al. (1999) re-produced the same shift of chromatographicpeaks observed earlier by low-pressure gel filtra-tion (Piccolo et al., 1996a and b) when the pH ofhumic solutions was adjusted from 7 to 3.5 by us-ing only small amounts (�0.5� 10�3 M) of anumber of monocarboxylic acids and, thus,with-out significantly changing the ionic strength ofthe sample. The larger resolution and repro-ducibility of HPSEC columns, in comparisonwith the low-pressure gel phases, allowed the au-thors to assess whether the alteration of theHPSEC molecular size distribution was depen-dent on the number of C atoms in the organicacids and on the hydrophilic/hydrophobic (HI/HB) C ratio of HS, as measured by CPMAS 13C-NMR spectroscopy. It was found that the

higher the C content of organic acids and thelower the HI/HB ratio of humic materials, thelarger the decrease in the average molecular sizeof HS.

Conte and Piccolo (1999b) conducted a sec-ond HPSEC experiment in which, instead oflowering the pH of humic solutions before injec-tion, they allowed the HPSEC eluent to beslightly modified.A control eluent at pH 7 (0.05M NaNO3, not absorbing light at 280 nm) wasmodified by addition of 2.0 � 10�6 M of eithermethanol, HCl, or AcOH to pH 6.97, 5.54, and5.69, respectively, without changing the ionicstrength (I � 0.0504 M). In this way the hy-drophobic adsorptions, which may occur at in-creased ionic strengths (Lindqvist, 1967; Chiczand Regnier, 1990; Specht and Frimmel, 2000),were avoided.Moreover, sodium humate and ful-vate solutions were previously titrated to pH 7 sothat their dissolution in the mobile phase at pH 7would have prevented any random occurrence ofnegative charges. This was to avoid any uncon-trolled formation of negative charges on thesolute, ascribed to changes in the ionic strength ofhumic solutions, thereby affecting the volume ofsample elution (Swift and Posner, 1971). UV-Visand Refractive Index (RI) detectors were used to record the chromatograms indicating the molecular-size distributions of the humic mate-rials with the objective of comparing the chro-matographic behavior of the chromophores (UVat 280 nm) with that of the real humic mass (RI).Both UV and RI detectors showed major alter-ations of the humic molecular-size distributionsand dramatic decreases in the weight-averageMW values in the modified eluents.The decreasein molecular size was revealed by the shift towardincreasingly larger elution volumes for both theUV and RI detectors and by the concomitant re-ductions in peak absorbance at the UV-Vis de-tector. In the constancy of ionic strength, the ob-served changes could be ascribed to theinteractions of the added chemicals with theweakly associated molecules giving rise to an ap-parent macromolecular structure of the humicmatter. The slight modifications of the mobilephases may have caused the collapse of the het-erogeneous humic aggregate into molecular asso-ciations of smaller dimensions but of greaterthermodynamic stabilities than for the controlsolution.This was attributed to the gain in energyobtained by the formation of intermolecular hy-drogen bonds (from 10 to 20 kJ mol�1 for eachhydrogen bond) among humic molecules in themodified mobile phases (Schwarzenbach et al.,


1993). The large decrease in molecular size sug-gests that the weak association of apparently highmolecular size, as observed for the HA in thecontrol solution, must, therefore, have been duepredominantly to weak intermolecular hydro-phobic forces such as van der Waals, �-�, andCH-� (Nishio et al., 1998) bonds, which holdsmall molecules together.

The work of Conte and Piccolo (1999b) pro-vided further direct evidence for the conforma-tional model based on the reversible self-associationof small humic molecules rather than on themacropolymeric random-coil concept.Molecularsize decreases caused by the breaking of ester andother covalent linkages by such small amounts ofmodifiers could not be proposed considering thecustomary strongly alkaline and acidic conditionsto which the HS had been subjected during iso-lation and purification. Furthermore, usingHPSEC, it was shown that the molecular size dis-tribution of HS must be interpreted by a combi-nation of two factors: the elution volume, and themolar absorptivity of the chromatographic peaks.Earlier research had failed to address the combi-nation of these factors either because closely sim-ilar HAs were analyzed by UV detection only andin less sensitive low-pressure size-exclusion sys-tems (Swift and Posner, 1971) or because weaklyUV absorbing FAs and/or (aqueous) dissolved or-ganic matter (DOM) samples were used in studieswith HPSEC systems without the support of RIdetector systems (Becher et al., 1985; Berden andBerggren, 1990; Chin and Gschwend, 1991; Chinet al., 1994). More recent HPSEC investigationswhich coupled UV, RI, and MALS (Multi-AngleLight Scattering) detectors, have also indicateddifferent molecular size distributions for HS ac-cording to the detector employed (von Wan-druszka et al., 1999).

When ionic-exclusion interferences werecarefully eliminated in the work of Conte andPiccolo (1999b), permanent adsorption of HSon the column was not found to occur (Conteand Piccolo, 1999a; Piccolo et al., 2001c), in ac-cordance with other HPSEC studies in similarconditions (Becher et al., 1985; Mueller et al.,2000). However, a modification of the mobilephase by addition of methanol, HCl, and AcOH,although in very small amounts, may have pro-duced pore-size changes that influenced non-SEC. In order to verify that mobile phase modi-fication did not alter the exclusion properties ofthe HPSEC column, Piccolo et al. (2001b) sub-jected polymeric standards of known MW, suchas the negatively charged polystyrenesulphonates

(PSS) and neutral polysaccharides (PYR), toSEC in the same mobile phases used by Conteand Piccolo (1999b) for the HS studies. Theundisputed polymeric nature of the covalentlylinked units of both nonionic PYR and poly-electrolytic PSS made a comparison of theirchromatographic behavior (using the same ex-perimental conditions) with that of HS relevant.Regardless of the charge density of the polymer,the similar chromatographic behavior of each indifferent mobile phases indicated that the slightvariations in the compositions of the mobilephases were not sufficient to alter either the sta-bility provided by strong covalent bonding ortheir interactions with the stationary phase.Conversely, the size-exclusion chromatograms,based on UV or RI detector systems, of threedifferent HAs (HA1 from a volcanic soil, HA2from an oxidized coal, and HA3 from a lignite)varied dramatically (in both peak absorbance andelution volumes) with the compositions of mo-bile phases.

Piccolo et al. (2001b) concluded that,whereasslight modifications in the mobile phase did notaffect the column capacity for size exclusion, thesubstantial difference between the response toHPSEC of the polymeric standards and that ofHS in the very same chromatographic conditionsproved that humic materials have different struc-tural assemblies.As proposed earlier, this may wellbe a self-assembling association of relatively smalland heterogeneous molecules rather than a coil ofpolymeric macromolecules. To verify that suchvariations were not specific for the wavelengthused to record the chromatograms (280 nm), andthereby simply accountable to shifts of peak max-ima,Piccolo et al. (2001b) also recorded UV spec-tra of HS solutions over a range of wave lengths.Their results showed that the three humic materi-als also produced different absorbance values uponmodification of their solutions over a wide rangeof wavelengths. This was regarded as confirmingthat the molar absorptivity of the bulk HS variedwith the composition of the solution. The ab-sorptivity was decreased by adding chemicalswhich disrupt their weakly-stabilized molecularassociations.

Piccolo et al. (2000a) also compared the chro-matographic behavior of HS with that of neutralPYR and polyelectrolytic PSS when solutions (in0.05 M NaNO3) before injection into theHPSEC system were titrated from pH 7 to 3.5with either HCl or AcOH, as was done for HS inthe work outlined above (Piccolo et al., 1999).Contrary to the work of Piccolo et al. (2001b),


the mobile phase was kept constant while the so-lutions under analysis were modified. It wasfound that plots of retention volumes (obtainedusing either the RI detector for PYR and PSSpolymers of known MW or the UV detector foronly PSS standards) versus Ln (MW) did notshow significant differences between the controland samples added with either HCl or AcOH.This result indicated that the capacity of theHPSEC column in aqueous media to exclude bysize was not altered by injecting the polymer so-lutions brought to pH 3.5 with a very smallamount of acid. However, when HS solutionswere similarly modified by lowering pH andchromatographed in the same mobile phase, theelution of the acid-added humic samples differedsignificantly from the control humic solution.Asin previous experiments, these results were con-sidered to be evidence that a humic association,stabilized only by weak dispersive forces at pH 7(mainly hydrophobic forces), varies considerablywith respect to the arrangements of its compo-nent molecules and its size distribution whentreated with an acid. In contrast, covalently linkedpolymers have more stable molecular arrange-ments and their HPSEC behavior is not affectedby an interaction with either a mineral or an am-phiphilic acid such as AcOH.

Cozzolino et al. (2001) studied the effect oforganic acids of plant, microbial, or anthropicorigins on the molecular size distribution of dis-solved HS. They used a Phenomenex BiosepS2000 column and eluted with a 0.05 M NaN03solution to evaluate size changes in four differentHS upon addition of hydroxy- (glycolic andmalic), keto- (glyoxylic), and sulfonic (benzene-sulfonic and methanesulfonic) acids. All HSshowed a decrease in peak absorbance when hu-mic matter was dissolved in the HPSEC mobilephase at pH 7, and the pH of the solution waslowered to 3.5 by acid addition before analysis.This effect was generally accompanied by an in-crease in peak elution volumes. The overall de-creases of the total areas of the chromatogramscompared with the control HS solutions wasagain explained with a disruption of supramole-cular humic aggregates into smaller sized butmore energy-rich associations brought about bythe formation of mixed intermolecular hydrogenbonds after treatment with acid. The hydroxy-bicarboxylic malic acid was the most effective indisrupting the original humic associations. Thiswas attributed to its greater capacity to form newhydrogen bonds with complementary functionsof HS. Malic acid is a bicarboxylic acid (HOOC-

CH2-CHOH-COOH) with two carboxylgroups (pKa1 � 3.4; pKa2 � 5.11) that can eitherbe protonated or partially dissociated at pH 3.5,thereby allowing a larger number of mixed hy-drogen bonds to form on its oxygen-containingfunctions than could be formed either by thestrong hydrochloric acid (which is a protondonor only) or by the more weakly acidic mono-carboxylic glycolic and glyoxylic acids. The ex-tent of molecular association variation was re-lated not only to the pKa values of the acids butalso to the chemical and stereochemical affinitiesof the humic components that would allow pen-etration of acids into the inner humic domains,depending on the acid structures. For example,the strongly acidic methanesulfonic and ben-zenesulfonic acids showed effects that varied withthe humic properties. Methanesulfonic acid wasmore effective in the disruption of the molecularassociations of HS that contained predominantlyaliphatic and alkyl moieties, and, equally, ben-zenesulfonic acid disrupted the distribution ofthe aromatic-rich humic material because of theprobable larger �-� interactions with aromatichumic components.

The indications of the supramolecular natureof HS provided by the analytical HPSEC studieshas caused Piccolo et al. (2001c) to employ apreparative HPSEC column to separate humicsize fractions that might be more homogeneouschemically, and they have subjected these frac-tions to chemical and spectroscopic characteriza-tion.A HPSEC fractionation was carried out onthe same HS before and after adjusting the pH ofthe humic solution (0.05 M ionic strength) from7 to 3.5 with 0.5 � 10�3 M AcOH. Six fractionswere collected from the HPSEC separation ofthe HA solution at pH � 7.0,whereas eight frac-tions were obtained from the separation of thehumic solution treated with AcOH to pH � 3.5before HPSEC injections. Fractionation was runcontinuously using an auto-sampler and a frac-tion collector, and the chromatographic patternmonitored by means of a UV detector. Repro-ducibility was excellent (CV � 5%) for morethan 100 runs for each fractionation series.Moreover, elemental analyses on the sum of frac-tions from both untreated and AcOH-treated HSexcluded the possibility of any significant ad-sorption of humic matter on the column station-ary phase. The molecular size distribution of theuntreated HS was significantly different fromthat of the AcOH-treated HS. The latter showedmuch less peak absorbance (280 nm), and severaladditional peaks appeared at higher elution times


(120–180 min). Size-fractions were analyzed by aCurie point (610 �C) Pyrolysis-Gas Chromatog-raphy-Mass Spectroscopy (Py-GC-MS) tech-nique and by 1H-NMR spectroscopy. Total ionchromatograms by Py-GC-MS showed that frac-tions had a significantly different chemical com-position after AcOH treatment, thereby confirm-ing that AcOH had caused rearrangement of thehumic associations in solution and that prepara-tive HPSEC is adequate for HS fractionation.Py-GC-MS spectra showed that the AcOH addi-tion altered the distribution of humic molecularcomponents in the size-fractions. The unsatu-rated alkyl chains were moved from size-frac-tions of larger molecular size into those of lowermolecular sizes. Most of the aromatic moieties,which were found in larger molecular size frac-tions for the untreated HA, were spread intofractions of lower molecular size after AcOH ad-dition to HA. Carbohydrates, which were unde-tectable in any fraction of the untreated HA, ap-peared instead in the pyrogram of the lowestmolecular size (and the most hydrophilic frac-tion) after treatment with AcOH. Our resultssuggested that AcOH disrupted the weaklybound association of humic constituents andHPSEC elution separated size-fractions of differ-ent compositions. The fractions with the largestapparent molecular sizes were the richest in alkylchains, suggesting that humic molecules werestabilized into supramolecular associations bymultiple weak interactions among apolar groupssuch as alkyl chains and aromatic moieties. 1H-NMR spectra of HPSEC fractions not only con-firmed the findings by Py-GC-MS analyses butalso showed that the chemical compositions offractions were less complex after AcOH treat-ment. This was attributed to a less complex mol-ecular association in the size separates, and thatallowed a larger solubility in the NMR solventand more favorable spin-lattice relaxation times.Piccolo et al. (2001c) showed,by combining frac-tion-separation by HPSEC with chemical andspectroscopic characterization, that the fractionswith larger apparent molecular sizes were com-posed predominantly of a mixture of alkyl com-pounds of relatively low molecular sizes, whereasfractions of lower apparent molecular sizes con-tained small aromatic systems and hydrophiliccompounds. These conclusions were in har-mony with previous research which character-ized HS fractions separated by either low-pres-sure GPC or ultrafiltration (Anderson andHepburn, 1978; Yonebayashi and Hattori, 1987;Piccolo et al., 1990a).

SOLUTE-GEL-ELUENT AS ANINTERACTIVE SYSTEM IN

EXCLUSION CHROMATOGRAPHY

Although the above results by HPSEC shouldbe regarded as free of interferences from ionicexclusion and adsorption (coefficient of variationfor more than five runs was consistently less than5%, and there was a lack of adsorption of humicmatter on the columns), a further check for pos-sible hydrophobic adsorption phenomena wasconducted. Seven molecules that might be con-sidered monomeric constituents of HS (Steven-son, 1994) were subjected to HPSEC separatelyand in a full mixture (1 g L�1). Elution was witha phosphate buffer at pH 3. The individual mol-ecules that were size-excluded gave the elutionvolumes reported in Table 1, whereas the mix-ture of the seven molecules produced an elutionprofile with only two peaks (Fig. 1).

It must be noted that the retention volumes ofthe two peaks for the mixture did not correspondto any of the elution volumes found for the singlemolecular species (Table 1). The first peak for themixture eluted (23.07 mL) before any of the sin-gle species, with the exception of the highly hy-drophilic glucosamine, and the more intense sec-ond peak was eluted before all other monomersexcept dihydroxyphenylacetic and gallic acids.These results suggest that when the single molec-ular species are mixed together in solution, theyform weak mutual associations giving rise to hy-drodynamic radius values larger than when theyare eluted alone. Based on the larger single elutionvolumes of the more hydrophobic monomers (hy-drocaffeic acid, resorcinol, and catechol), it can beinferred that their elution is retarded by adsorptionon the stationary phase of the column.Conversely,


TABLE 1

HPSEC retention volumes (mL) of different monomers and mixtures of the same monomers on a column

Polysep GFC-P3000 eluted with a phosphate buffer at pH 3

Monomers/Mixtures mL

Dihydroxyphenylacetic acid 23.39Gallic acid 25.49Protocatechuic acid 25.94Hydrocaffeic acid 26.60Glucosammine (RI) 20.23Resorcinol 26.90Catechol 26.95

Mixture of the above (0 h) First peak: 23.07; second peak: 25.91

Mixture of the above First peak: 23.01; second (after 144 h standing) peak: 25.70

when the same compounds are in mixtures withother monomers the thermodynamic drive to de-crease exposure of hydrophobic components tothe aqueous medium increases the mutual attrac-tion of monomers. This interpretation is substan-tiated by the observation that the retention vol-umes of mixture peaks decreased in value withtime of standing before injection (Table 1), therebyindicating a further hydrophobic strengthening ofthe monomer associations. A consequence is thatthe hydrophobic interactions of single monomerswith the column matrix are reduced while exclu-sion of the monomer association is expedited.

Evidence that hydrophobic adsorption on thecolumn is not irreversible and affects only elutionvolumes was obtained by eluting progressivelyless concentrated solutions of the mixture. Theabsorbance decrease of both peaks given by themixture gave a highly significant linear correla-tion with the decrease in the concentration of themixture, as expected by the Lambert and Beerlaw (Fig. 2). This suggests that there were nolosses attributable to adsorption on the column of

chemical species that might be regarded as repre-sentative of monomeric structures in HS.

The behavior of the monomer mixture inHPSEC may have similarities with that of associ-ations of humic molecules. When the apparentlylarge-size, weakly bound humic associations arealtered by the action of organic acids or by otherspecies (uncharged compounds, electrolytes, ca-tions), smaller humic clusters are separated andtheir elution is retarded by size exclusion; in ad-dition, the hydrophobic interactions with thecolumn matrix will have changed.

Other evidence that humic molecules aresuprastructural associations in aqueous solutionsis given by the elution patterns of HS (in HPSECexperiments) in 8 M urea. Concentrated urea isemployed to disrupt protein-protein interactionsand to solubilize aggregating hydrophobic pro-teins for their further separation (Hjelmeland,1990). Concentrated urea has also been proposedfor low-pressure size exclusion and polyacryla-mide gel electrophoresis (PAGE) separations ofHS because it is considered able to produce elu-tion patterns similar to 0.1 M Tris-HCl buffer atpH 9 (Trubetskoj et al., 1997).

The elution profiles of three different puri-fied HS (0.2 g L�1) dissolved in 8 M urea andeluted in the same solution are shown in Fig. 3.The chromatograms were detected by both UVand RI detectors in a series. This experiment isillustrative of the capacity of urea to separate thehydrophobic from the hydrophilic componentsof HS.The chromatograms show a net separationof humic matter: a high molecular size fraction atthe column Vo visible only by the UV detector


Fig. 1. HPSEC chromatogram of a mixture (1 g.L�1) ofseven monomeric precursors of HS eluted by a phos-phate buffer at pH 7 through a Phenomenex column(Polysep GFC-P3000). 1. Dihydroxyphenylacetic acid; 2.Gallic acid;. 3:Protocatechuic acid;4. Hydrocaffeic acid;5. Glucosammine (RI detected); 6. Resorcinol; 7. Cate-chol.

Fig. 2. Relationship of peak absorbances (first and secondpeak of Fig. 1) versus concentration of a mixture of sevenmonomeric precursors of HS eluted by a phosphatebuffer at pH 7 through a Phenomenex column (PolysepGFC-P3000). 1. Dihydroxyphenylacetic acid; 2. Gallicacid;. 3. Protocatechuic acid; 4. Hydrocaffeic acid; 5. Glu-cosammine (RI detected); 6. Resorcinol; 7. Catechol.

and a low molecular-size fraction eluting at thecolumn Vt detected only by the RI detector. Thehydrophobic humic molecules are strongly asso-ciated through hydrophobic effects induced bythe highly concentrated urea. The urea interactsmore favorably than hydrophobic molecules withthe network of the water structure. Moreover,since urea is known to form complexes withnonionic detergents (Hjelmeland, 1990), it mayalso be possible that complexes between urea andnonionic humic hydrophobic compounds in-creased the hydrodynamic radius of the hy-drophobic associations, causing these to elute atthe earliest exclusion volume of the column.

Differences among HS could also be noted.The HA from an oxidized coal was the only oneto show two well separated peaks in the RI de-tector system (Fig. 3, I-RI), whereas only onevery intense peak (�1000 mV) at the Vo columnwas shown by the UV detector (Fig. 3, I-UV).This indicates that while light-absorbing chro-mophores (280 nm) with high molar absorptivi-ties were excluded rapidly from the column, theirconcentrations in the humic samples were rele-vant and of a similar order of magnitude as thenonabsorbing hydrophilic, probably ionized,compounds detected by the RI detector at thetotal exclusion volume of the column. The HAfrom lignite gave a similar complete separationbetween the large-size hydrophobic chro-mophore fraction (Fig. 3, II-UV) and the small-size ionized and hydrophilic components shownby the RI detector (Fig. 3, II-RI). However, al-though the molar absorptivity of the excludingchromophores was still high (as suggested by anabsorption intensity � 800 mV), the lack of cor-responding peaks at the RI detector indicatedthat their concentrations in the sample weremuch less than those of the hydrophilic compo-nents excluded at large elution volumes. Identicalbehavior was shown by the HA extracted from anagricultural soil. Also for this sample, the chro-mophores association, excluded at the columnvoid volume (Fig. 3, III-UV), was hardly repre-sentative of the mass of humic sample since thecorresponding peak at the RI detector was al-most irrelevant compared with the signal of the


Fig. 3. UV- and RI-(Refractive Index) detected HPSECchromatograms of humic acids dissolved and eluted in8M urea solution. I. humic acid from an oxidized coal; II.humic acids from a lignite; III. humic acid from a Danishagricultural soil.

small size, nonabsorbing, hydrophilic constituents(Fig. 3, III-RI).

The two experiments reported above suggestthat in the HPSEC analysis of associations ofmolecules such as HS, the stationary and mobilephases lose the properties associated with specificsingle molecules alone. Rather, they become partof an interactive system involving the mixture ofsolutes. It is the whole solute-gel-eluent systemthat produces the size-exclusion of the associa-tion under analysis. Moreover, it is the humicconformational structure, and the degree of po-tential hydrophobic interaction between the as-sociated molecules and either the stationary orthe mobile phases, that determines the elutionvolume of a humic association in aqueous solu-tion. This applies more than ionic exclusion, sooften considered in the past to explain the SECbehavior of HS. These results also point out thatthe molecular sizes of HAs are not related to themolar absorptivity of chromophores in the hu-mic samples, and they should be assessed by de-tection means that are related to the total mass ofthe HAs.

POLYMERIZATION OF HUMICSUBSTANCES BY OXIDATIVE

CATALYTIC REACTIONS

To understand humus as a supramolecular as-sociation of small molecules it is necessary toovercome the limitations imposed by the para-digmatic polymeric model. If HS are seen asweakly bound supramolecular associations theirunstable conformations could then be stabilizedin real polymeric structures. This could beachieved by increasing the number of intermolec-ular covalent bonds via an oxidative coupling re-action catalyzed by oxidative enzymes such as thephenoloxidases. This class of enzymes has beenshown to promote, through a free-radical mech-anism,oligo- and poly-merization of phenols andanilines, and, hence, is believed to contribute tosoil detoxification from related organic contami-nants (Kim et al., 1997).

Piccolo et al. (2000b) turned a loosely boundhumic superstructure into a covalently linkedpolymer by treating a humic material dissolved in0.1 M phosphate buffer at pH 7 with horseradishperoxidase (HRP) and hydrogen peroxide (oxi-dant). They used HPSEC to evaluate the changesin molecular size distribution brought about bythe oxidative reaction with HRP catalysis.More-over, AcOH was added to the reacted humicmixture to bring the pH to 4, and HPSEC injec-tion was then used to assess the stability of the

humic conformation following the polymeriza-tion reaction. The polymerized HS had HPSECabsorptions of larger intensities, and the elutionvolumes were shifted to lower values than thecontrol.Moreover, treatment with AcOH did notalter significantly the chromatographic appear-ance of the polymerized HS,whereas it produceddisruption of the loosely bound association of theuntreated HS resulting from a significant reduc-tion of the intensities of the peaks and their shiftsto larger elution volumes. These changes indi-cated a significant increase in the molecular sizesof the humic materials resulting from oxidationcatalyzed by HRP, which was attributable to atrue polymerization of humic molecules via theformation of C-O or C-C bonds. Furthermore,DRIFT (Diffuse Reflectance Infrared FourierTransform) spectroscopy was used by Piccolo etal. (2000b) to verify the effect of the oxidativepolymerization reaction on the molecular struc-ture of the HS. In comparison with the control,the DRIFT spectrum of the humic material sub-jected to oxidative coupling showed a substantialchange in the 1500–900 cm�1 frequency intervalwith the appearance of three main bands at 1247,1097, and 947 cm�1 and a decrease in the 1400and 1227 cm�1 bands. The absorption shown at1247 and at 1097 cm�1 was reasonably assignedto bond deformation of aryl and alkyl ethers,respectively, which were formed during free-radical coupling reactions catalyzed by HRP and,hence, confirm the interpretation of HPSECmeasurements.

The HPSEC and DRIFT results of Piccolo etal. (2000b) suggest that the small heterogeneousmolecules present in HS, as in weakly associatedsuperstructures, can be covalently bound intotrue oligo- or polymers by an oxidative couplingreaction catalyzed by a peroxidase enzyme. Theextent of covalent polymerization should be afunction of the amount of humic molecules,mainly phenolic or benzencarboxylic acids de-rived from lignin and microbial biosynthesis,which may undergo oxidative coupling reac-tions. However, it should be imagined that otherclasses of compounds may become assimilatedinto the macromolecular structures of polymer-ized humus.

Cozzolino and Piccolo (2001) extended thepolymerization catalysis by HRP to other HS andstudied the effects of solution pH (4.7 and 7) andcompositions of humic associations. By HPSECexperiments they confirmed that an increase inweight-average MW occurred invariably for allHS subjected to oxidative polymerization. More-


over, a comparison of chromatograms and of MWvalues obtained by treating humic solutions withAcOH to pH 3.5 before HPSEC injection con-firmed that the increase in molecular size by HRPcatalysis was stable and caused by the formation ofcovalent bonds in the reacting humic molecules.However, covalent polymerization of humic mol-ecules was found to proceed to a greater extent atpH 7 than at pH 4.7, despite the fact that HRP ismost active at the latter pH. The difference in re-activity was attributed to the large mobility of re-acting molecules in the hydrated and relativelysmaller humic associations stabilized only by weakdispersive (hydrophobic) forces at pH 7. Reactivehumic molecules are more mobile at neutral pH,and the polymerization via a free-radical mecha-nism is more efficient. Conversely, intermolecularhydrogen bonds formed at pH 4.7 confer a largersize and rigidity to humic associations, whereasthe mobility as well as the reactivity of small mol-ecules are thereby decreased.

Synthetic complexes formed with Al-(hydr)oxide preparations of montmorillonite were madeusing HAs from an oxidized coal and a lignite(Violante et al., 1999) to model organo-mineralcomplexes of soils using humic matter similar tothat of soil. In a separate experiment, these syn-thetic clay-humic complexes were subjected tothe oxidative reaction with HRP as a catalyst andH2O2 as an oxidant. No changes in the OC con-tents were observed for the oxidative conditionsapplied. Extraction of the complexes with alka-line-pyrophosphate solutions allowed the deter-mination of the amounts of HS solubilizable be-fore and after the treatment with HRP. Figure 4shows that the yields extracted in the cases ofboth humic-clay complexes decreased signifi-cantly after the oxidative coupling reaction, rang-ing from about 42 to 32% and from 40 to 29%for the complexes made of HA from oxidizedcoal and lignite, respectively. These results indi-cate that polymerization of humic molecules oc-curred also in the solid phase of the clay-humiccomplexes and the increase in molecular sizes ofthe humic materials was the most probable causefor the reduction in extraction yields. Generalsimilarities to soil HS of the humic material wereused to form the model clay-humic complexes,and, thus, it would also seem to be possible to in-duce the polymerization of HS in natural soilsamples in order to control or change the proper-ties of native SOM.

The evidence shown here that humicsupramolecular associations can be turned intomore stable covalently linked structures of larger

molecular size can be interpreted as additionalevidence that HS may be present in soils as asso-ciations of relatively small molecules.

CHEMICAL AND SPECTROSCOPICEVIDENCE OF SUPRAMOLECULAR

ASSOCIATIONS

The model of self-assembling supramolecularassociation of HS is related to mutual affinities ofcertain molecules in aqueous solutions. Mole-cules tend to associate by intermolecular forces(Israelachvili, 1994), and the strength of the asso-ciation depends on their molecular structures.Particularly strong associations are formed byapolar compounds via the hydrophobic effect(Tanford, 1980).A humic supramolecular associ-ation in solution is thus formed by the self-organization of hydrophobic and amphiphiliccompounds. The associations are isolated pro-gressively from the network of water structure.Such separation results in an increase in the en-tropy of the system and in the overall energy sta-bilization as the different humic molecules forminto a superstructure.

The importance of hydrophobic humic com-ponents in phenomena of aggregation in solutionand on surfaces, as well as in controlling the reac-tivity of HS, is well documented (Wershaw,1999). Investigations by fluorescence quenching


Fig. 4. Yields (%) of H. S. extracted with an alkaline-py-rophosphate solution from synthetic OH-AL-humate-montmorillonite complexes formed by using humicacids from oxidized coal and lignite before (COX-M;LIG-M) and after polymerization reaction catalyzed by (HRP) horseradish peroxidase (COX-M�HRP; LIG-M�HRP).

techniques provided evidence of hydrophobicmicrodomains in loose humic associations (Morraet al., 1990; Engebretson and von Wandruszka,1994). By following the diffusion of 1,2-dichloroethane into humic matter, Aochi andFarmer (1997) showed that there are discrete mi-croregions of different polarities in humic struc-tures. Chien et al., (1997) studied by 19F-NMRspectroscopy the interactions of a trifluoromethyl-ated atrazine with a soil HA by measuring theNMR relaxation of atrazine in the presence ofboth hydrophilic and hydrophobic paramagneticprobes. They confirmed the existence of hy-drophobic domains by showing that atrazine oc-cupied a domain of HS accessible only to neutralhydrophobic molecules. Similar results were ob-tained by Piccolo et al. (1998) who studied theisotherms for the adsorption of atrazine by HSextracted from two soils with four different ex-tractants. The most hydrophobic humic extract,isolated from soil by an acetone-HCl solution, ad-sorbed more and desorbed less atrazine than themore hydrophilic humic materials extracted byeither an alkaline or an alkaline-pyrophosphatesolution. Since the acetone-HCl extract repre-sents only a small fraction of soil HS (Piccolo etal., 1990b), the study of Piccolo et al. (1998) in-ferred that humic microdomains of different po-larities are also present in soil and their distribu-tions determine their reactivities.

Direct evidence for the presence of hy-drophobic domains in HS were further given byKohl et al. (2000),who studied the sorptive uptakeof hexafluorobenzene by two peat samples usingsolid-state 19F-NMR spectroscopy. They foundthat the sorption process was rapid and related tothe soil lipid content, whereas removal of thelipids decreased significantly the sorptive capacityof SOM. Hu et al. (2000), using solid-state NMRand wide-angle X-ray scattering (WAXS) de-tected crystalline domains composed of poly-(methylene) chains in several samples of SOM,humins, and HAs from soil and coal. Their resultsindicated a crystallite thickness equivalent to ap-proximately 25 CH2 units to give hydropho-bic crystalline domains. A comparable amount ofnoncrystalline and more isotropically mobile poly(methylene) chains were also found and, togetherwith the noncrystalline materials, larger aggre-gates were formed. In accordance with thesupramolecular association model described here,Hu et al. (2000) concluded that the crystallites areexpected to be resistant to environmental attackand, thus, inert in the soil and likely to have longresidence times, whereas amorphous regions may

play a role in the sorption of nonpolar moleculesin soil.

Disruption of humic supramolecular associa-tions as a result of the formation of hydrogenbonds stronger than the hydrophobic forces stabi-lizing the original conformation was shown byMiano et al. (1992). They used fluorescence andinfrared spectroscopy to investigate the effects ofthe addition of glyphosate [N-(phosphonomethyl)glycine] herbicide to a purified humic solution tobring the pH from 9 to progressively lower val-ues. The excitation spectra revealed increasingfluorescence quenching effects with increasingglyphosate contents at high wavelengths. Syn-chronous spectra showed a decrease in the mainpeak intensity with increasing additions of herbi-cide. These results were interpreted in terms of adisaggregation of the humic supramolecular asso-ciation as the result of the formation of multiplehydrogen bonds between glyphosate and thesmall humic molecules. This implied that a de-crease in electron delocalization, attributable tothe apparent large molecular size of the HS, wasresponsible for the bands at high wavelengths.The molecular and conformational structures ofHS were confirmed as determinants of the ad-sorption of glyphosate on different humic mate-rials (Piccolo et al., 1996c). The high content ofaliphatic components and the large and flexiblemolecular size increased adsorption of the herbi-cide on HS, possibly because of the involvementof dispersive hydrophobic interactions, togetherwith hydrogen bonds.

Conte et al. (1997) showed that the formationof tightly bound hydrophobic domains in solu-tions of HS could alter the quantitative evaluationof humic C distribution in 13C-NMR spectrarecorded in the liquid state. A comparison be-tween liquid-state NMR spectra and solid-stateCPMAS-NMR spectra of several HS showed thatthe amounts of alkyl groups measured by the for-mer technique were invariably lower than for thelatter method.The authors attributed this result tothe separated phase that the humic hydrophobiccomponents create to decrease the total solvationenergy in solution and to the consequent limitedspin-lattice relaxation time of hydrophobic car-bons.Although liquid-state spectra may, therefore,show lower signals in the alkyl-C region, this doesnot occur in the solid-state mode, where signalrecording occurs by polarization transfer fromprotons near the carbons measured.

In another approach, Kenworthy and Hayes(1997) used the fluorescence quenching ofpyrene by bromide to investigate the nature of


humic associations. They found that pyrene in aHS solution was protected from bromidequenching. This protection was lost, however,when AcOH, followed by a base, was added tothe medium. The authors considered that hy-drophobic associations of the humic moleculesprotected the pyrene from the bromide. Treat-ment of the humic solution with AcOH, as pro-posed by Piccolo et al. (1996a and b), removedthat protection because of disaggregation of theloose humic superstructure. This suggested thatthe HS in solution were associations of low MWmasses held together by hydrophobic bondingand in which pyrene was enveloped.

Supramolecular associations of HS werefound by Haider et al. (2000) and Ricca et al.(2000) to be disrupted into smaller componentsby the action of derivatizing reagents (trimethylsilyl or halogenated alkyl compounds) to silylateor methylate acidic oxygenated functions. Suchsimple derivatizations, which could not breakether and ester linkages, disaggregated the weaklyassembled humic supramolecular structures intosmaller entities that readily dissolve in organicsolvents, elute in the low molecular size ranges byHPSEC (Haider et al., 2000), and produce betterresolved NMR and FTIR spectra (Ricca et al.,2000). Wanner et al. (2000) used the concept ofdisaggregation of humic supramolecular associa-tions to explain the shift in gel permeation tolarger elution volumes of the radioactivity of a14C-labeled dithianon fungicide bound to HSextracted from soil after 64 days of incubation to-gether with straw. The same material extractedfrom soil without incubation with straw indi-cated that the radioactivity was mainly in thelarge molecular size fraction of the HS. The au-thors reasoned that organic acids or other am-phiphilic compounds produced during microbialdegradation of maize would have disrupted theassociation of humic molecules as proposed byPiccolo et al. (1996a and b).

Data from Simpson et al. (2000), who usedDiffusion Ordered Spectroscopy (DOSY) NMR,which has enhanced sensitivity because the probecircuitry is kept at 22 oK, support the supramolec-ular model proposed by Piccolo and coworkers.In their preliminary study, the authors observedthat the Nuclear Overhauser Effect (NOE) of ahumic material was explained more by an associ-ation of small molecules rather than by a macro-molecular structure.

Buurman et al. (2001) studied the thermal sta-bilities of soil humic extracts exchanged with H,Na,Ca,or Al before and after wetting with organic

solvents such as methanol, formic acid,and AcOH.The thermal stabilities of Na-humates, upon addi-tion of the organic solvents, shifted to high tem-peratures, whereas hardly any effect was observedfor H-, Ca-, and Al-humates. This was explainedby the mobilities of relatively small humic mole-cules that were forced by the organic solvent toenhance their intermolecular hydrophobic inter-actions and strengthen their Na-humate supra-molecular structures. More energetic bonds, suchas hydrogen bonds in H-humates, and electrostaticbridges with divalent and trivalent cations in Ca-and Al-humates, provided an increased humicconformational rigidity and prevented a molecularrearrangement in the organic solvents.

CONCEPTS OF SUPRAMOLECULARASSOCIATIONS OF HUMIC

SUBSTANCES

The results of the experiments described thatused either analytical or preparative SEC cannotbe explained by analytical interferences or by thetraditional polymeric model of HS. Rather, theycan be interpreted by the concept of looselybound humic supramolecular associations. In thisconcept, one can imagine HS to be relativelysmall and heterogeneous molecules of variousorigins that self-organize in supramolecular con-formations. Humic superstructures of relativelysmall molecules are not associated by covalentbonds but are stabilized only by weak forces suchas dispersive hydrophobic interactions (van derWaals, �-�, and CH-� bondings) and hydrogenbonds, the latter being progressively more impor-tant at low pH values. Hydrophilic and hy-drophobic domains of humic molecules can becontiguous to or contained in each other and, inhydration water, form apparently large molecularsize associations. In humic supramolecular orga-nizations, the intermolecular forces determinethe conformational structure of HS, and thecomplexities of the multiple noncovalent inter-actions control their environmental reactivity.

The definition given by Lehn (1995) maywell be applied to HS:“Supramolecular assemblies(are) molecular entities that result from the spon-taneous association of a large undefined numberof components into a specific phase having moreor less well-defined microscopic organization andmacroscopic characteristics depending on its na-ture (such as films, layers,membranes,vesicles,mi-celles, mesomorphic phases, solid state structures,etc.).”

On consideration of the concept of supra-molecular associations, the classical definitions of


HAs and of FAs should be reconsidered. The FAsmay be regarded as associations of small hy-drophilic molecules in which there are enoughacidic functional groups to keep the fulvic clus-ters dispersed in solution at any pH. The HAs arecomposed by associations of predominantly hy-drophobic compounds (polymethylenic chains,fatty acids, steroid compounds) that are stabilizedat neutral pH by hydrophobic dispersive forces(van der Waals, �-�, and CH-� bondings). Theirconformations grow progressively in size whenintermolecular hydrogen bonds are increasinglyformed at lower pH values until they flocculate.

FUTURE PERSPECTIVES IN RESEARCH AND TECHNOLOGY

A clarification of the aggregate structures ofHS represents a major innovation in humuschemistry. The concepts presented here empha-size that HS are not macromolecular polymers, asthey have been described for so long, but arerather superstructures of apparent large size onlyand are self-assembled from relatively small het-erogeneous molecules held together mainly byhydrophobic dispersive forces. These conceptsprovide a new opportunity to enlarge knowledgeof both the detailed chemistry and the manage-ment of HAs in soil and in the environment.

Chromatographic methods of separation suchas HPSEC were found to produce reproducibleand more homogenous fractions of humic super-structures.Awareness of the weak forces that causethe self-assembling of humic molecules allowsmethods to be devised based on interactions withchemical species, such as amphiphilic organicacids,urea,mono- and polyvalent cations, that candisrupt the apparently large humic associationsand obtain fractions that are chemically simplerand more homogeneous.

The combination of chromatographic meth-ods with spectroscopic techniques such as NMR,IR, and ESR spectroscopy, together with the dif-ferent modern variations of mass-spectrometry,has increased enormously the potential to derivea complete picture of the secondary molecularstructure of HS. The time is near when there willbe good grasp of the molecular structures and as-sociations of HS, based on chemical methodsrather than on computer models. Our ability toobtain a full molecular structure of any humicmaterial or natural organic matter (NOM) mol-ecule from any environment and ecosystem willthen be limited by only the advances in analyticalautomation.When we consider the rapidity withwhich certain biological research fields can ad-

vance, this limitation could easily be overcome,provided there is the necessary public and privateinterest to pursue knowledge in a field that is vi-tal for the well-being of our planet and the bio-logical life on it.

The novel understanding of HS as supramol-ecular associations as described here has great im-plications in soil and environmental manage-ment. One example is the possibility of turningthe loose humic superstructures into real cova-lently linked polymers by catalytic technologythat can ensure polymerization of humic mole-cules in both water and soil environments. Suchtechnology can improve our capacity to controlSOM management, reduce the risk of erosion,and limit soil desertification with the obvious im-provement of soil productivity.Another exampleof the potential of the polymerization of soilhumic molecules in situ is the possibility of con-trolling CO2 emission from agricultural soils bysequestering the organic C in more stable poly-merized humus. Finally, a large number of obser-vations indicate that humic molecules releasedfrom large supramolecular associations can influ-ence nutrient uptake by plants and increase cropyields significantly. Combinations of refinedchemical analyses of humic molecules with phys-iological studies on their effect on plants mayclarify the mechanism(s) by which soil HS in-crease crop yields and, possibly, revive the formerinterests in humus based fertility. The use of hu-mic molecules, either native or exogenous, incombination with inorganic fertilizers, to maxi-mize plant nutrient uptake and final yields, mayalso have a tremendous impact on increasing theeconomic efficiency of fertilizers and protectingthe environment from the pollution caused byexcess uses of fertilizers.

ACKNOWLEDGMENTS

The author thanks his former doctoral stu-dents, Drs. P. Conte and R. Spaccini, for most ofthe original research presented here and Ms. A.Cozzolino,who is now completing her doctorate.

REFERENCES

Aiken, G. R., D. M. McKnight, R. L. Wershaw, and P.MacCarthy. 1985. Introductory remarks. In HumicSubstances in Soil, Sediment, and Water. G.R.Aiken, D.M. McKnight, R.L. Wershaw, and P.MacCarthy (eds.).Wiley, New York, pp. 1–12.

Aiken,G.R., and A.H.Gillam.1989.Determination ofmolecular weights of humic substances by colliga-tive property measurements. In Humic SubstancesII. In Search of Structure. M.H.B. Hayes, P. Mc-


Carthy,R.L.Malcolm, and Swift R.S. (eds.).Wiley,New York, pp. 516–543.

Amalfitano, C.,V. Pignalosa,, L.Auriemma. and A. Ra-munni. 1992. The contribution of lignin to thecomposition of humic acids from a wheat-strawamended soil during 3 years of incubation in pots.J. Soil Sci. 43:495–504.

Anderson, H.A., and A. Hepburn. 1978. Fractionationof humic acid by gel permeation chromatography.J. Soil Sci. 28:634–643.

Aochi, Y. O., and W. J. Farmer. 1997. Role of mi-crostructural properties in time-dependent sorp-tion/desoprtion behaviour of 1,2-dichloroethaneon humic substances. Environ. Sci. Technol.31:2520–2526.

Becher,G.,G.E.Carlberg,E.T.Giessing, J.K.Hongslo,and S. Monarca. 1985. High-performance size ex-clusion chromatography of chlorinated natural hu-mic water and mutagenicity studies using the mi-croscale fluctuation essay. Environ. Sci. Technol.19:422–426.

Berden,M., and D.Berggren.1990.Gel filtration chro-matography of humic substances in soil solutionsusing HPLC-determination of the molecularweight distribution. J. Soil Sci. 41:161–172.

Buurman, P., B. van Lagen, and A. Piccolo. 2001. In-crease in thermal stability of soil humic substancesas a result of self-association. Org. Geochem. (inpress).

Cameron, R. S., B. K. Thornton, R. S. Swift, and A. M.Posner. 1972a. Molecular weight and shape of hu-mic acid from sedimentation and diffusion measure-ments on fraction extracts. J. Soil Sci. 23:394–408.

Cameron, R. S., R. S. Swift, B. K. Thornton, and A. M.Posner. 1972b. Calibration of gel permeation chro-matography materials for use with humic acid. J.Soil Sci. 23:343–349.

Campbell, C.A., E.A. Paul, D.A. Rennie, and K. J. Mc-Callum. 1967. Applicability of the carbon datingmethod of analysis to soil humus studies. Soil Sci.104:217–224.

Carter, C. W., and I. H. Suffet. 1982. Binding of DDTto dissolved humic material. Environ. Sci. Technol.16:735–740.

Ceccanti, B., M. Calcinai, M. Bonmati-Pont, C. Ciardi,and R. Tarsitano. 1989. Molecular size distributionof soil humic substances with ionic strength. Sci.Total Environ. 81/82:471–479.

Chicz, R. M., and F. E. Regnier. 1990. High-performance liquid chromatography: Effective pro-tein purification by various chromatographicmodes. In Guide to Protein Purification. M.P.Deutscher (ed.).Methods of Enzymology,Vol. 182.Academic Press, San Diego, pp. 392–421.

Chien, Y.-Y., E.-G. Kim, and W. F. Bleam. 208. 1997.Paramagnetic relaxation of atrazine solubilized byhumic micellar solutions. Environ. Sci. Technol.31:3204–3203.

Chin, Yu-P., and P. M. Gschwend. 1991. The abun-dance, distribution, and configuration of porewater

organic colloids in recent sediments. Geochim.Cosmochim.Acta 55:1309–1317.

Chin, Yu-P., G. R. Aiken, and E. O’Loughlin. 1994.Molecular weight, polydispersity, and spectroscopicproperties of aquatic humic substances. Environ.Sci. Technol. 28:1853–1858.

Clapp, C. E., W. W. Emerson, and A. E. Olness. 1989.Sizes and shapes of humic substances by viscositymeasurements. In Humic Substances II. In Search ofStructure. M.H.B. Hayes, P. McCarthy, R. L. Mal-colm and R.S. Swift (eds.). Wiley, New York, pp.497–514.

Clapp,C.E., and M.H.B.Hayes.1999.Sizes and shapesof humic substances. Soil Sci. 164:777–788.

Conte, P.,A. Piccolo, B. van Lagen, P. Buurman, and P.A. de Jager. 1997. Quantitative differences in evalu-ating soil humic substances by liquid- and solid-state13C-NMR spectroscopy. Geoderma 80:339–352.

Conte,P., and A.Piccolo. 1999a.High pressure size ex-clusion chromatography (HPSEC) of humic sub-stances. Molecular sizes, analytical parameters, andcolumns performance. Chemosphere 38:517–528.

Conte, P., and A. Piccolo. 1999b. Conformationalarrangement of dissolved humic substances.Influenceof solution composition on association of humicmolecules. Environ. Sci. Technol. 33:1682–1690.

Cornel, P. K., R. S. Summers, and P.V. Roberts. 1986.Diffusion of humic acid in dilute aqueous solution.J. Colloid Interface Sci. 110:149–164.

Cozzolino,A., and A. Piccolo. 2001. Polymerization ofdissolved humic substances catalyzed by peroxidase.Effects of pH and humic composition. Org. Geo-chem. (in press).

Cozzolino,A., P. Conte, and A. Piccolo. 2001. Confor-mational changes of soil humic substances inducedby some hydroxy-, cheto-, and sulphonic acids. SoilBiol. Biochem. 33:563–571.

De Haan, H., R. I. Jones, and K. Salonen. 1987. Doesionic strength affect the configuration of aquatichumic substances, as indicated by gel filtration?Freshwater Biol. 17:453–459.

De Nobili,M.,E.Gjessing, and P.Sequi.1989.Sizes andshapes of humic substances by gel chromatography.In Humic Substances II. In Search of Structure.M.H.B. Hayes, P. McCarthy, R. L. Malcolm andR.S. Swift (eds.).Wiley, New York, pp. 562–587.

Determann,H., and I.Walter. 1968.Source of aromaticaffinity of Sephadex dextran gels. Nature (London)219:604–605.

Dubach, P., and N. C. Metha. 1963. The chemistry ofsoil humic substances. Soils Fertil. 26:293–300.

Engebretson, R., and R. von Wandruszka. 1994.Micro-organization of dissolved humic acids. Env-iron. Sci. Technol. 28:1934–1941.

Engebretson, R., and R. von Wandruszka. 1997. Theeffect of molecular size on humic acid associations.Org. Geochem. 26:759–767.

Flaig, W. 1958. Die Chemie der organischer Stoffe inBoden und deren physiologische Wirkung. Ver-handl., II u. IV Komm. int. Bodenk. Ges., 2.


Flaig,W. 1988. Generation of model chemical precur-sors. In Humic Substances and Their Role in theEnvironment. F.H. Frimmel and R.F. Christman(eds.).Wiley, Chichester, pp. 75–92.

Flaig,W., and H. Beutelspacher. 1968. Investigations ofhumic acids with the analytical ultracentrifuge. InIsotopes and Radiation in Soil Organic MatterStudies. IAEA,Vienna, pp. 23–30.

Flaig,W., H. Beutelspacher, and E. Riets. 1975. Chem-ical composition and physical properties of humicsubstances. In Soil Components,Vol. 1. J.E. Giesek-ing (ed.). Springer-Verlag, Berlin, pp. 1–219.

Ghosh, K., and S. K. Mukherjee. 1971. Hymatomelanicacids as polyelectrolytes: 1.Viscosimetric and osmo-metric studies. J.Appl. Polym. Sci. 15:2073–2077.

Ghosh, K., and M. Schnitzer. 1980. Macromolecularstructures of humic substances. Soil Sci. 129:266–276.

Goring, D.A. I. 1971. Polymer properties of lignin andlignin derivatives. In Lignins: Occurrence, Forma-tion, Structure and Reactions. K.V. Sarkanen andC.H.Ludwig (eds.).Wiley,New York, pp. 695–768.

Haider, K. 1987. The synthesis and degradation of hu-mic substances in soil. In Trans. 13th Congr. Intern.Soc. Soil Sci. Hamburg, Germany, pp. 644–656.

Haider, K., and J. P. Martin. 1967. Synthesis and trans-formation of phenolic compounds by Epicoccus ni-grum in relation to humic acid formation. Soil Sci.Soc.Am. Proc. 31:766–772.

Haider, K., M., Spiteller, J. Dec, and A. Schäffer. 2000.Silylation of soil organic matter. Extraction of hu-mic compounds and soil-bound residues. In SoilBiochemistry, vol X. J.-M. Bollag and G. Stotzky(eds.). Marcel Dekker, New York, pp. 139–170.

Hayase, K., and H. Tsubota. 1983. Sedimentary humicand fulvic acid as surface active substances. Geo-chim. Cosmochim.Acta 47:947–952.

Hayes, M. H. B. 1997. Emerging concepts of the com-positions and structures of humic substances. InHumic Substances, Peats and Sludges. M.H.B.Hayes and W.S.Wilson (eds). The Royal Society ofChemistry, Cambridge, pp. 3–30.

Hedges, J. I. 1988. Polymerization of humic substancesin natural environments. In Humic Substances andTheir Role in the Environment. F.H. Frimmel andR.F.Christman (eds.).Wiley,Chichester, pp. 45–58.

Hjelmeland, L. M. 1990. Solubilization of native mem-brane proteins. In Guide to Protein Purification.M.P. Deutscher (ed.). Methods of Enzymology,Vol. 182.Academic Press, San Diego, pp. 253–264.

Hu, W.-G., J. Mao, B. Xing, and K. S. Rohr. 2000.Poly(methylene) crystallites in humic substancesdetected by Nuclear Magnetic Resonance. Envi-ron. Sci. Technol.34:530–534.

Insam, H. 1996. Microorganisms and humus in soils. InHumic Substances in Terrestrial Ecosystems. A.Piccolo (ed.). Elsevier,Amsterdam, pp. 265–292.

Israelachvili, J. N. 1994. Intermolecular and SurfaceForces, 2nd Ed.Academic Press, London, pp. 282,

Jenkinson, D. S. 1981. The fate of plant and animal

residues in soil. In The Chemistry of Soil Processes.D.J. Greenland and M.H.B Hayes (eds.). Wiley,Chichester, pp. 505–561.

Kenworthy, I. P., and M. H. B. Hayes. 1997. Investiga-tions of some structural properties of humic sub-stances by fluorescence quenching. In Humic Sub-stances, Peats and Sludges. M.H.B. Hayes and W.S.Wilson (eds.). The Royal Society of Chemistry,Cambridge, pp. 39–45.

Kim, J-E., E. Fernandes and J-M. Bollag. 1997. Enzy-matic coupling of the herbicide betazon with hu-mus monomers and characterization of reactionproducts. Environ. Sci. Technol. 31:2392–2398.

Kohl, S. D., P. J. Toscano,W. Hou, and J.A. Rice. 2000.Solid-state 19F NMR investigation of hexafluo-robenzene sorption to soil organic matter. Environ.Sci. Technol. 34:204–210.

Kononova,M.M.1961.Soil organic matter. Its Nature,Its Role in Soil Formation and in Soil Fertility.Pergamon Press, New York.

Laue,T.M., and D.G.Rodhes. 1990.Determination ofsize, molecular weight and presence of subunits. InGuide to Protein Purification.M.P.Deutscher (ed).Methods of Enzymology,Vol. 182.Academic Press,San Diego, pp. 566–587.

Lehn, J.-M. 1995. Supramolecular Chemistry. VCH,Weinheim, p. 7.

Lindqvist, I. 1967.Adsorption effects in gel filtration ofhumic acid.Acta Chem. Scand. 21:2564–2566.

Maillard, L. C. 1912. Formation d’humus et de com-bustibles mineraux sans intervention de l’oxygeneatmospherique, des microorganismes des hautestemperatures ou des fortes pressions. C. R. Acad.Sci. Paris 154:66.

Malcolm, R. L. 1990. The uniqueness of humic sub-stances in each soil, stream, and marine environ-ment.Anal. Chim.Acta 232:19–30.

Martin, J. P., and K. Haider. 1969. Phenolic polymersof Stachibotrys atra, Stachibotrys chartarum and Epicoc-cus nigrum in relation to humic acid formation. SoilSci. 107:260–270.

Miano, T. M., A. Piccolo, G. Celano, and N. Senesi.1992. Infrared and fluorescence spectroscopy ofglyphosate-humic acid complexes. Sci. Total Envi-ron. 123/124:83–92.

Morra, M. J., M. O. Corapcioglu, R. von Wandruszka,D. B. Marshall, and K. Topper. 1990. Fluorescencequenching studies of Naphtalene and 1-Naphtholinteraction with humic acid. Soil Sci. Soc. Am. J.54:1283–1289.

Mueller, M., D. Schmitt, and F. Frimmel. 2000. Frac-tionation of natural organic matter by size exclu-sion chromatography-properties and stability offractions. Environ. Sci. Technol. 34:4867–4872.

Mukherjee, P. N., and A. Lahiri. 1959. Polyelectrolyticbehaviors of humic acids. Fuel 37:220–226.

Nardi, S., G. Arnoldi, and G. Dell’Agnola. 1988. Re-lease of the hormone-like activities from Allophob-ora rosea (sav.) and Allophobora caliginosa (sav.) feces.Can. J. Soil Sci. 68:563–567.


Nishio, M., M. Hirota, and Y. Umezawa. 1998. TheCH/p Interaction, Evidence, Nature and Conse-quences.Wiley, New York.

Orlov, D. S., I. A. Ammosova, and G. I. Glebova. 1975.Molecular parameters of humic acids. Geoderma13:211–229.

Piccolo, A. 1988. Characteristics of soil humic sub-stances extracted with some organic and inorganicsolvents and purified by the HCl-HF treatment.Soil Sci. 146:418–426.

Piccolo, A., and J. S. C. Mbagwu. 1990. Effects of dif-ferent organic waste amendments on soil microag-gregates stability and molecular sizes of humic sub-stances. Plant Soil 123:27–37.

Piccolo, A., R. Rausa, and G. Celano. 1990a. Charac-teristics of molecular size fractions of humic sub-stances derived from oxidized coal. Chemosphere24:1381–1387.

Piccolo,A., L. Campanella, and B. M. Petronio. 1990b.13C-NMR spectra of humic substances extractedwith different mechanisms. Soil Sci. Soc. Am. J.54:750–755.

Piccolo,A., S. Nardi, and G. Cancheri. 1992. Structuralcharacteristics of humic substances as related to ni-trate uptake and growth regulation in plant systems.Soil Biol. Biochem. 24:373–380.

Piccolo,A.1996.,Humus and soil conservation, In Hu-mic Substances in Terrestrial Ecosystems. A. Pic-colo (ed.). Elsevier,Amsterdam, pp. 225–264.

Piccolo, A. S. Nardi, and G. Cancheri. 1996a. Macro-molecular changes of soil humic substances in-duced by interactions with organic acids.Eur. J. SoilSci. 47:319–328.

Piccolo, A. S. Nardi, and G. Cancheri. 1996b. Micelle-like conformation of humic substances as revealedby size-exclusion chromatography. Chemosphere33:595–600.

Piccolo, A., G. Celano, and P. Conte. 1996c. Interac-tions of Glyphosate with humic substances. J.Agric.Food Chem. 44:2442–2446.

Piccolo, A., P. Conte, I. Scheunert, and M. Paci. 1998.Atrazine interactions with soil humic substances ofdifferent molecular structure. J. Environ. Qual.27:1324–1333.

Piccolo, A., P. Conte, and A. Cozzolino. 1999. Effectsof mineral and monocarboxylic acids on the mole-cular association of dissolved humic substances.Eur. J. Soil Sci. 50:687–694.

Piccolo, A., and P. Conte. 2000. Molecular size of hu-mic substances. Supramolecular associations versusmacromolecular polymers. Adv. Environ. Res.3:508–521.

Piccolo,A., P. Conte, and A. Cozzolino. 2000a. Differ-ences in high performance size exclusion chro-matography between humic substances and macro-molecular polymers. In Humic Substances.VersatileComponents of Plant, Soils and Water. E.A. Gabb-hour and G. Davies (eds.). Royal Society of Chem-istry, Cambridge, UK., Special Publication no. 259,pp. 111–124.

Piccolo, A., A. Cozzolino, P. Conte, and R. Spaccini.2000b. Polymerization of humic substances by anenzyme catalyzed oxidative coupling. Naturwis-senschaften 87:391–394.

Piccolo, A., P. Conte, A. Cozzolino, and R. Spaccini.2001a.Molecular sizes and association forces of hu-mic substances in solution. In Humic Substancesand Chemical Contaminants. C.E. Clapp, M.H.B.Hayes, N. Senesi, P. Bloom, and P.M. Jardine (eds.).Soil Science Society of America, Madison, WI, pp.89–118.

Piccolo, A., A. Cozzolino, and P. Conte. 2001b. Chro-matographic and spectrophotometric properties ofdissolved humic substances compared with macro-molecular polymers. Soil Sci. 166:174–185.

Piccolo,A.,P.Conte,E.Trivellone,B. van Lagen and P.Buurman. 2001c. Size-fractionation of a humicacid by preparative HPSEC and characterization ofsize-separates by Pyr-GC-ms and 1h-nmr spec-troscopy. Environ. Sci. Technol. (In Press).

Piret, E. L., R. G. White, H. C. Walther, and A. J.Madded.1960.Some physicochemical properties ofpeat humic acids. Sci. Proc. R. Dublin Soc., Ser. A.1:69–79.

Posner, A. M., and J. M. Creeth. 1972. A study of hu-mic acids by equilibrium centrifugation. J. Soil Sci.23:333–341.

Ricca, G., F. Severini, G. Di Silvestro, C. M. Yuan, andF.Adani. 2000. Derivatization and structural studiesby spectroscopic methods of humic acids fromLeonardite. Geoderma 98:115–125.

Reid,P.M.,A.E.Wilkinson,E.Tipping, and N. J.Mal-colm.1990.Determination of molecular weights ofhumic substances by analytical (UV scanning) ul-tracentrifugation. Geochim. Cosmochim. Acta 54:131–138.

Reuter, J. H., and E. M. Perdue. 1981. Calculation ofmolecular weights of humic substances from col-ligative data: Application to aquatic humus and itsmolecular size fractions. Geochim. Cosmochim.Acta 45:2017–2022.

Ritchie, G. S. P., and A. M. Posner. 1982. The effect ofpH and metal binding on the transport propertiesof humic acids. J. Soil Sci. 33:233–247.

Ryan, D. K., and J. H. Weber. 1982. Fluorescencequenching titration for determination of complex-ing capacities and stability constants of fulvic acid.Anal. Chem. 54:986–990.

Saiz-Jimenez, C. 1995. Analytical pyrolysis of humicsubstances: Pitfalls, limitations and possible solu-tions. Environ. Sci. Technol. 28:1773–1780.

Saiz-Jimenez, C. 1996. The chemical structure of hu-mic substances: Recent advances. In Humic Sub-stances in Terrestrial Ecosystems. A. Piccolo (ed.).Elsevier,Amsterdam, pp. 1–44.

Schwarzenbach, R. P., P. M. Gschwend, and D. M. Im-boden. 1993. Environmental Organic Chemistry.Wiley, New York.

Scheffer, F., and B. Ulrich. 1960. Lehrbuch derAgrikulturchemie und Bodenkund. Teil III. Hu-



mus und Humusdüngung,2nd Ed.Ferdinand Enke,Stuttgart.

Schnitzer,M., and S.U.Khan.1972.Humic Substancesin the Environment. Marcel Dekker, New York.

Schulten, H.-R., P. Leinweber and M. Schnitzer. 1998.Analytical pyrolysis and computer modeling of hu-mic and soil particles. In Structure and Surface Re-actions of Soil Particles. P.M. Huang, N. Senesi andJ. Buffle (eds.).Wiley, New York, pp. 281–324.

Schulten, H.-R. and P. Leinweber. 2000. New insightsinto organic-mineral particles: composition, prop-erties, and models of molecular structure. Biol. Fer-til. Soils 30:399–432.

Simpson, A. J., W. L. Kingery, M. H. B. Hayes, M.Spraul, E. Humpfer, P. Dvorstsak, R. M. Kesser-baum, M. Godejohann, and M. Hofmann. 2000.The structure and association of organic moleculesin the terrestrial environment. In Proceedings ofthe 10th International meeting of the Humic Sub-stances Society, 24–28 July 2000, Vol. 2. Toulouse,France, pp. 1109–1112.

Spaccini, R., A. Piccolo, G. Haberhauer, M. Stemmer,and M. H. Gerzabek. 2001. Decomposition ofmaize straw in different European soils as revealedby DRIFT spectra of soil particle fractions. Geo-derma 99:245–260.

Specht, C. H., and Frimmel, F. H. 2000. Specific inter-actions of organic substances in size-exclusion chro-matography. Environ. Sci. Technol. 34:2361–2366.

Staudinger, H. 1935. Die hochmolekularen organis-chen Verbindungen, Kautschuk und Cellulose.Springer, Berlin.

Stevenson, F. J., Q. van Winkle and W. P. Martin. 1953:Physico-chemical investigations of clay-adsorbed or-ganic colloids: II. Soil Sci. Soc.Am. Proc. 17:31–34.

Stevenson,F. J.1994.Humus Chemistry.Genesis,Com-position, Reactions, 2nd Ed.Wiley, New York.

Summers, R. S., P. K. Cornel, and P.V. Roberts. 1987.Molecular size distribution and spectroscopic char-acterization of humic substances. Sci. Total Envi-ron. 62:27–37.

Swift, R. S., and A. M. Posner. 1971. Gel chromatogra-phy of humic acid. J. Soil Sci. 22:237–249.

Swift, R. S. 1989. Molecular weight, shape, and size ofhumic substances by ultracentrifugation. In HumicSubstances II. In Search of Structure. M.H.B.Hayes, P. McCarthy, R. L. Malcolm and R.S. Swift(eds.).Wiley, New York, pp. 449–466.

Swift, R. S. 1999. Macromolecular properties of soilhumic substances: Fact, fiction, and opinion. SoilSci. 164:790–802.

Tanford,C. 1980.The Hydrophobic Effect: Formationof Micelles and Biological Membranes.Wiley,NewYork.

Tate,R.L. III. 1999:Micelles, subunits, and the Mediter-ranean sun. Soil Sci. 164:775–776.

Thurman,E.M.,R.L.Wershaw,R.L.Malcolm,and D.J. Pinkney. 1982. Molecular size and weight mea-

surements of humic substances. Org. Geochem.4:27–35.

Trubetskoj, O. A., O. E., Trubetskaja, G. V., O. I.Afanas’eva, Reznikova, and C. Saiz-Jimenez. 1997.Polyacrilamide gel electrophoresis of soil humic acidfractionated by size-exclusion chromatography andultrafiltration. J. Chromatogr. 767:285–292.

van Dijk, H. 1972. Colloidal chemical properties ofhumic matter. In Soil Biochemistry, Vol. 2. A.D.McLaren and G.H. Peterson (eds.). Marcel Dekker,New York, pp. 16–34,.

Varga, B., G. Kiso, I. Geleucser, J. Hlovcy, and Z.Krivécsy.2000.Secondary structure of humic acids:Can micelle-like conformation be proved by aque-ous size-exclusion chromotography? Environ. Sci.Technol. 16:3303–3306.

Violante, A., M. Arienzo, F. Sannino, C. Colombo,A. Piccolo, and L. Gianfreda. 1999. Formation and characterization of OH-Al-humate-montmo-rillonite complexes. Org. Geochem. 30:461–468.

Visser, S. A. 1964. A physico-chemical study of theproperties of humic acids and their changes duringhumification. J. Soil Sci. 15:202–219.

von Wandruszka, R., M. Schimpf, M. Hill, and R. En-gebretson. 1999. Characterization of humic acidsize fractions by SEC and MALS. Org. Geochem.30:229–235.

Waksman,S.A.1936.Humus.Williams & Wilkins,Bal-timore.

Wanner, U., P. Burauel, and F. Führ. 2000. Characteri-zation of soil-bound residue fractions of the fungi-cide dithianon by gel permeation chromatographyand polyacrilamide gel elctrophoresis.Environ.Pol-lut. 108:53–59.

Welte,E. 1955.New results of humus research.Angew.Chem. 67:153–155.

Wershaw, R. L. 1986.A new model for humic materi-als and their interactions with hydrophobic chemi-cals in soil-water and sediment-water systems. J.Contam. Hydrol. 1:29–45.

Wershaw, R. L. 1989. Size and shape of humic sub-stances by scattering techniques. In Humic Sub-stances II. In Search of Structure. M.H.B. Hayes, P.McCarthy, R. L. Malcolm and R.S. Swift (eds.).Wiley, New York, pp. 545–559.

Wershaw, R. L. 1993. Model for humus in soils andsediments. Environ. Sci. Technol. 27:814–816.

Wershaw, R. L. 1999. Molecular aggregation of humicsubstances. Soil Sci. 164:803–813.

Wershaw, R. L., and G. R.Aiken. 1985. Molecular sizeand weight measurements of humic substances. InHumic Substances in Soil, Sediment, and Water.G.R. Aiken, D.M. McKnight, R.L. Wershaw, andP. MacCarthy (eds.). Wiley, New York, pp.477–492.

Yonebayashi, K., and T. Hattori. 1987. Surface activeproperties of soil humic acids. Sci. Total Environ.62:55–64.

the supramolecular structure of humic … science 2002... · the supramolecular structure of humic...

Documents