Spectroscopic characterization of the structural and functional properties of natural organic matter fractionsJie Chen a ,Baohua Gua,*,Eugene J.LeBoeuf c ,Hongjun Pan d ,Sheng Daiba Environmental Sciences Division,Oak Ridge National Laboratory (ORNL),P.O.Box 2008,Oak Ridge,TN 37831,USAbChemical and Analytical Sciences Divisions,ORNL,P.O.Box 2008,Oak Ridge,TN 37831,USAcDepartment of Civil and Environmental Engineering,Vanderbilt University,P.O.Box 1831,Station B,Nashville,TN 37235,USAdDepartment of Chemistry,University of Tennessee,Knoxville,TN 37996,USAReceived 13July 2001;received in revised form 17January 2002;accepted 21January 2002AbstractNatural organic matter (NOM)is known to be complex in nature with varying structural and functional charac-teristics.In this study,an aquatic NOM was fractionated into the polyphenolic-rich (NOM-PP)and the carbohydrate-rich (NOM-CH)fractions in an attempt to better characterize their chemical and structural properties along with a reference soil humic acid (SHA).Various spectroscopic techniques were employed for the study,including ultraviolet–visible (UV/Vis),13C-nuclear magnetic resonance,Fourier-transform infrared,fluorescence,and electron paramagnetic resonance spectroscopies.Results indicate that the relative abundance of aromatic C @C and methoxyl (–OCH 3)functional groups are in the order of SHA >NOM-PP >NOM-CH.However,the aquatic NOM-PP and NOM-CH fractions are characterized by high contents of carboxylic and alcoholic functional groups relative to the SHA.In particular,the NOM-PP fraction appears to contain more phenolic and ketonic functional groups than the NOM-CH and SHA fractions,and it gives a strong fluorescence and high paramagnetic spin count.On the other hand,the NOM-CH fraction possesses a relatively low amount of carbon but a high amount of oxygen or oxygen-containing structural features,such as carbohydrate-OH and carboxylic groups,and shows the least fluorescence intensity and paramagnetic spin counts.Results of these spectroscopic studies confirm the heterogeneous nature of NOM,and point out the im-portance of isolation and improved characterization of various NOM subcomponents in order to better understand the behavior and roles of NOM in the natural environment.Ó2002Elsevier Science Ltd.All rights reserved.Keywords:Natural organic matter;Humic substances;Fluorescence;FTIR;UV/Vis;NMR;EPR1.IntroductionNatural organic matter (NOM)is widely distributed in soil,natural water,and sediments and consists of a mixture of the decomposition products of plant and animal residues and of substances synthesized biologi-cally and/or chemically from decomposed products or intermediate products (Aiken et al.,1985;Suffet andMacCarthy,1989).The size,chemical composition,structure,and functional groups,and polyelectrolytic characteristics of NOM may vary greatly,depending on the origin and age of the material (Aiken et al.,1985;Chin et al.,1994;Gu et al.,1995;Chin et al.,1998).It is this heterogeneity that has made the structural and functional characterization of NOM extremely chal-lenging.Since NOM is also known to play important roles in the fate and transport of many toxic organic or inorganic chemicals and in nutrient cycling throughout the environment,characterization of NOM is particu-larly useful for studying these reactions and processes (Chen et al.,1977;Choppin,1992;Chen et al.,1996;Chemosphere 48(2002)59–68/locate/chemosphere*Corresponding author.Tel.:+1-865-574-7286;fax:+1-865-576-8543.E-mail address:b26@ (B.Gu).0045-6535/02/$-see front matter Ó2002Elsevier Science Ltd.All rights reserved.PII:S 0045-6535(02)00041-3Vairavamurthy et al.,1997;Chorover et al.,1999;Bur-gos et al.,2000;Fredrickson et al.,2000).More spe-cifically,a better understanding of the structural and functional properties of NOM may greatly improve our understanding of the underlying mechanisms responsi-ble for the complexation,reduction,and mobilization or immobilization of heavy metals,radionuclides,pesti-cides,and other toxic chemicals with NOM.This,in turn,may improve our predictive capabilities of the behavior of NOM and environmental pollutants in natural ecosystems.No single analytical tool,however,can provide de-finitive structural or functional information about NOM because of its heterogeneous,ill-defined nature.Rather, as Schnitzer and Khan(1972)suggested:‘‘significant progress in our knowledge of the chemistry and reac-tions of humic substances may be possible only by the combined application of a variety of spectroscopic and wet-chemical techniques.’’Among analytical character-ization methods,nondestructive spectroscopic tech-niques appear to be most useful,and they are directly applicable to both solid and liquid sample analyses. These methods have a number of attractive advantages including:(a)they are nondestructive;(b)only small amounts of samples are needed;(c)they are experi-mentally simple and do not require special manipulative sample preparation;and,perhaps most importantly,(d) they provide valuable information on molecular struc-ture and chemical or functional properties of NOM (Stevenson,1994).Spectroscopic techniques such as nu-clear magnetic resonance(NMR),Fourier-transform in-frared(FTIR),electron paramagnetic resonance(EPR), ultraviolet–visible(UV/Vis),andfluorescence have been previously applied for both quantitative and qualitative characterization of NOM(Stevenson and Goh,1971; Schnitzer and LeVesque,1979;Schnitzer and Preston, 1986;Leenheer et al.,1987;Senesi et al.,1989,1991;Gu et al.,1994,1995).Significant spectral overlapping and peak broadening often occur,however,because of the multicomponent nature of NOM,thus making the iden-tification and interpretation of spectral signatures par-ticularly difficult.This study aimed to fractionate the bulk NOM into well-defined subcomponents,and then systematically characterize their structural and functional properties by a combination of spectroscopic techniques including NMR,FTIR,EPR,UV/Vis,andfluorescence.Frac-tionation of total NOM offers advantages by selectively separating one group of organic compounds(or a sub-component)from the others on the basis of their phys-ical and chemical properties.Relatively well-defined NOM subcomponents or fractions are thus obtained, and their chemical and structural features can be better characterized.It is known that humic substances can be classified into two major components,namely humic acid(HA)and fulvic acid(FA),on the basis of their solubility in acid and base.HAs are those materials that are insoluble under acidic conditions(at pH<$2)but soluble at higher pH values.FAs are soluble at all pH values and are generally characterized by lower molec-ular weight but higher oxygen-containing organic com-ponents.FAs can be further fractionated on the basis of their adsorptive behavior on synthetic resins(Leenheer et al.,1995a,b)or on a cross-linked polyvinyl pyrroli-done(PVP)polymer(Swincer et al.,1968;Lowe,1975; Gu and Doner,1992).Fractionation of NOM on syn-thetic resins such as XAD-8can provide two distinct hydrophobic and hydrophilic subcomponents and has been applied in the separation and characterization of many aquatic NOMs(Gu et al.,1995;Leenheer et al., 1995a,b).Similarly,the PVP polymer has been used for separating colored organic components rich in aromatic C@C moieties because of their preferential adsorption on PVP under acidic conditions(Swincer et al.,1968; Lowe,1975).Those organic components not adsorbed by PVP are primarily comprised of carbohydrates, proteins,amino acids,and uronic acids in solution.This latter fractionation scheme was utilized in the present study as part of our major effort to understand the dif-ferent roles of polyphenolic-or carbohydrate-rich NOM components and HAs in their complexation and redox reactions with environmental contaminants such as heavy metals and radionuclides(Choi et al.,2000;Gu et al.,2000).2.Experimental2.1.Selection and isolation of NOM fractionsAn aquatic NOM(GT-NOM)was obtained from a wetland pond in Georgetown,South Carolina,as de-scribed previously(Gu et al.,1994,1995).Because GT-NOM represents a bulk aquatic NOM,it was further fractionated on the basis of its adsorptive behavior on a cross-linked PVP synthetic polymer(Lowe,1975),to which those organic compounds rich in aromatic C@C double bonds(or colored fractions)adsorb by forming covalent bonds(Watanabe and Kuwatsuka,1992).In brief,about1g of GT-NOM was dissolved in500ml of deionized water by adjusting to pH$7:0with dilute NaOH.The NOM solution was then acidified to pH1.5 with6M HCl to separate the HA(precipitants)from the FA fraction(supernatant).The precipitated HA was collected by centrifuging it at11000rpm for30min,and by repeated washing with small volumes($20ml)of purified water(Milli-Q system,Barnstead,Corp.),fol-lowed by freeze drying.The FA supernatant solution was then decanted and passed through a glass column packed with the pre-washed PVP.Most of the colored materials(designated as the polyphenolic-rich NOM or60J.Chen et al./Chemosphere48(2002)59–68NOM-PP fraction)were retained by the PVP,whereas the eluent contained the carbohydrate-rich NOM frac-tion(designated as the NOM-CH fraction)and soluble salts.The column was then backwashed with approxi-mately300ml of0.1M NaOH at aflow rate of0.5ml/ min to desorb the NOM-PP from the PVP column.The NOM-PP and NOM-CH fractions were further purified by passing through a column containing AG50W-X4 Hþcation exchange resin,followed by a rinse with100 ml of deionized water.Both the NOM-PP and NOM-CH fractions were concentrated to approximately100ml on a vacuum evaporator(Speed-Vac,Savant Instru-ments,Farmingdale,NY)and freeze dried.All freeze-dried samples were placed into airtight glass bottles and stored in a desiccator.Because only a small fraction of HA could be isolated from the aquatic GT-NOM,a soil HA(SHA)sample was obtained from the International Humic Substances Society(IHSS);it represents a high-molecular weight NOM component that is rich in aromatic functional groups.This SHA was isolated from a mollic epipedon near Joliet,Illinois,and the sample was used as received without further purification.2.2.Spectroscopic analysis of NOM fractionsUV/Vis spectra of NOM samples were recorded on a Hewlett–Packard8453spectrophotometer(Hewlett–Packard Company,Wilmington,DE)in a1-cm quartz cuvette and scanned from190to800nm.Total organic carbon(TOC)concentrations of the NOM samples were 10mg C/l,and pH was adjusted to$7.0prior to the analysis of their absorbance spectra.For comparison, each absorbance spectrum was plotted as the molar absorptivity by normalizing the UV/Vis absorbance to the molar mass of dissolved organic carbon concentra-tion(Traina et al.,1990;Gu et al.,1995,1996).FTIR spectra of NOM samples were collected in the transmission mode using a Nicolet Magna760FTIR spectrophotometer equipped with an MCT(Hg–Cd–Te) detector(Nicolet Instrument Corporation,Madison WI). About1mg of the freeze-dried NOM sample and200mg of KBr powder were ground together and hydraulically pressed into a small pellet($13mm in diam.and$1mm thick)(Stevenson and Goh,1971;Gu et al.,1994).The palletized KBr samples were dried in an oven at100°C for2h prior to analysis to minimize interferences from absorbed water.Spectra resolution was8cmÀ1.Solid-state CPMAS13C-NMR spectra of NOM fractions were obtained with a Nicolet NT-200NMR spectrometer operated at the13C frequency of50.3 MHz.The instrument was run under the following conditions:contact time,3ms;spinning speed,4kHz; 90°1H pulse,5l s;acquisition delay,4s;and number of scans,from5000to70000.The relative abundance of specific functional groups of each NOM fraction was estimated by integrating the peak area or intensity of the given spectral region of interest relative to the alkyl re-gion between0and50ppm.Total aromaticity of each NOM fraction was then calculated as the ratio of total aromatic carbon in the spectral region112–163ppm to total carbon in the spectral region0–190ppm(Schnitzer and Preston,1986;Malcolm,1990).Fluorescence spectra were collected using a Spex Fluorolog-3fluorescence spectrophotometer(Johin-Yvon,Edison,NJ)equipped with a450-W Xe-arc lamp source.All NOM samples were dissolved in0.01M KCl solutions with afinal TOC concentration of20mg C/l and a pH of$7.Approximately1.4ml of the sample was placed in a Spectrosul Far UV quartz triangular fluorescence cell,and thefluorescence spectra were ac-quired as follows(Miano et al.,1988;Senesi et al.,1991). Emission spectra were measured from400to600nm at the excitation wavelength k ex¼340nm.Synchronous spectra were recorded with the excitation monochro-mator scanned from300to600nm and D k¼20nm (Miano and Senesi,1992).All spectra were collected with a5-nm band pass on both emission and excitation monochromators.The wavelength step size was1nm, and each spectrum intensity was measured as the aver-age of three signal samplings.EPR spectra at the X-band of9.8GHz were recorded by a Bruker EMX Spectrometer equipped with a TM110 cavity(Bruker Instruments,MA).The NOM samples werefirst dissolved in a0.1M NaOH solution to obtain a1%weight per unit volume solution concentration. Samples were placed in50-l l glass capillaries(VWR Scientific,West Chester,PA).Sample temperature was controlled at37Æ0:1°C during data acquisition by means of standard Bruker variable temperature units. Effective microwave and modulationfields of the sample were calibrated with peroxylamine disulfonate(Aldrich, Milwaukee,WI)(Beth et al.,1983).The EPR was op-erated at a peak-to-peak modulation amplitude of2.0G, a microwave attenuation of25db,a modulation fre-quency of100kHz,a conversion time of10.24ms,and a time constant of1.28ms.Linear signal intensity under the operating conditions of the instrument indicated avoidance of power saturation.Records of each spec-trum were obtained by digital signal averaging and were electronically integrated by standard Bruker data ac-quisition software.The standard compound3-carba-moyl-2,2,5,5,-tetramethyl-3-pyrrolin-1-yloxy was used as a reference for spin count calibration.3.Results and discussion3.1.UV/Vis spectroscopic analysisHumic substances generally show strong absor-bance in the UV–Vis range(from$190to800nm),J.Chen et al./Chemosphere48(2002)59–6861particularly in the UV region,because of the presence of aromatic chromophores and/or other organic com-pounds(Schnitzer and Khan,1972;Traina et al.,1990; Gu et al.,1995,1996).In fact,the UV absorptivity at260 nm(ABS260)is commonly used to determine the relative abundance of aromatic C@C content of NOM because p–pÃtransitions in substituted benzenes or polyphenols occur in this wavelength region(Ghosh and Schnitzer, 1979;Traina et al.,1990).Accordingly,the UV/Vis spectra of the three NOM fractions showed a generally decreased absorptivity(or optical density)as the wave-length increased(Fig.1).Although the spectra appeared to be broad and featureless,showing no maxima or minima,the absorption intensity varied greatly among the three NOM fractions;their relative order of inten-sity was SHA>NOM-PP>NOM-CH.The spectra of the NOM-CH fraction exhibited about a threefold lower UV absorbance compared to the NOM-PP and SHA fractions,with a relative order of absorptivities at 260nm of SHAð0:82Þ>NOM-PPð0:57Þ>NOM-CH ð0:12Þ.These observations suggest that both SHA and NOM-PP contain a relatively high amount of aromatic or polyphenolic organic compounds in contrast with the NOM-CH fraction.In addition,the NOM-CH fraction showed a more rapid decrease in absorptivity as the wavelength in-creased in comparison with the NOM-PP and SHA fractions,and its absorptivity approached zero at>350 nm wavelength.The relative order of the ratios of ab-sorptivity at465to665nm,commonly referred to as the E465=E665ratio,was NOM-CHð>100Þ>NOM-PP ð8:6Þ>SHAð4:3Þ.This ratio is independent of the concentration of the humic substances but is character-istic of different NOM fractions or humic substances obtained from different sources.The E465=E665ratio for HAs is usually<5.0;the ratio for FAs ranges from6.0 to8.5(Chen et al.,1977).Similarly,the ratios of E265=E465for the NOM-CH,NOM-PP,and SHA were calculated to be228,15.3,and4.29,respectively,and appeared to be consistent with the E465=E665ratio.These observations could be explained by the fact that UV/Vis absorption of NOM is affected by the relative abun-dance of aromatic C@C and ketonic C@O functional groups(or chromophores)and auxochromes such as C–OH and C–NH2.The ketonic C@O functional groups give a weaker absorption in the visible range,whereas auxochromes do not confer color by themselves but in-crease the color of chromophores(Schnitzer and Khan, 1972;Leenheer et al.,1987).Elemental analyses of these NOM fractions indicated that the NOM-CH and NOM-PP fractions contained a higher amount of oxy-gen but a lower amount of carbon than the SHA(Table 1).Strong UV absorption(at<250nm),particularly by the SHA,may result from both the aromatic C@C and ketonic C@O functional groups.However,UV ab-sorption by the NOM-CH fraction may result primarily from the absorption by ketonic C@O functional groups since absorption at wavelengths>350nm was approx-imately zero;this explains a high E465=E665or E265=E465 ratio(or a rapid decrease in absorptivity with an in-crease in wavelength).On the other hand,a low E465=E665or E265=E465ratio for SHA may be largely at-tributed to the absorption by aromatic C@C functional groups.Additionally,a high degree of condensation of the aromatic rings and the large molecular weight of SHA are believed to contribute to its relatively high absorption in the visible range(Schnitzer and Khan,1972).Table1Elemental analysis of NOM fractions a;bSample Carbon(mass%)Hydrogen(mass%)Oxygen(mass%)Nitrogen(mass%)Sulfur(mass%) NOM-CH40.7 4.948.4 1.4 5.66NOM-PP46.0 4.948.4 1.0 1.06Soil HA c53.5 4.539.3 4.00.44a Analysis was performed by Huffman Laboratories,Golden,CO.b All data were corrected for ash content.c Obtained from the International Humic Substances Society.62J.Chen et al./Chemosphere48(2002)59–683.2.13C-NMR spectroscopic analysisSolid-state13C-NMR spectroscopy provides a com-plementary technique to the UV/Vis spectroscopy for investigating the chemical and structural properties of NOM fractions(Schnitzer and Preston,1986;Chen et al.,1996;Cook and Langford,1998;Xing and Chen, 1999;Mao et al.,2000).As shown in Fig.2,differences in structural signatures were observed among the three NOM fractions.The following major spectral bands were identified and integrated(Table2):(a)0–50ppm, representing mainly aliphatic or paraffinic carbon chains;(b)50–75ppm,representing primarily methoxyl(OCH3) groups,and75–112ppm for carbohydrate RC–OH or RC–OR functional groups;(c)112–145ppm,as a result of resonance from aromatic carbons;(d)145–163ppm, representing the phenolic groups;and(e)163–180ppm, corresponding to carboxylic,carbonyl,amine,and ester carbons in NOM.These major spectral bands are con-sistent with those observed for soil and aquatic humic substances by a number of investigators although the relative intensities of the spectra varied greatly among three NOM fractions(Newman et al.,1987;Malcolm, 1989;Newman and Tate,1991;Chen et al.,1996;Cook and Langford,1998;Xing and Chen,1999;Mao et al., 2000).The SHA displayed a relatively larger amount of aromatic carbon at$130ppm and methoxyl(OCH3) carbon at$55ppm than the NOM-PP and NOM-CH fractions.Integration of the aromatic region between 112and163ppm revealed that the SHA contained $44.7%of aromatic carbon,an amount which is sub-stantially higher than those of NOM-PP and NOM-CH fractions.This amount is also higher than the aver-age aromatic carbon content of most humic substances ($35%)(Malcolm,1989).Relatively low quantities of carboxylic carbon were observed in the SHA at163–180 ppm and only trace quantities of phenolic carbons at145–163ppm.On the other hand,NOM-PP and NOM-CH showed relatively higher amounts of carboxyl and carbohydrate or alcoholic carbons at163–180and 50–112ppm.In particular,the NOM-CH fraction ex-hibited the highest amounts of carbohydrate or alco-holic carbons at$75ppm,but the lowest amounts of aromatic carbons at$100–145ppm.In addition,the spectral region180–220ppm is commonly assigned to ketone and aldehyde carbons(Bortiatynski et al.,1996), and both NOM-CH and NOM-PP fractions appeared to contain more of these functional groups than SHA (Fig.2).Results of the13C-NMR spectroscopic analysis of NOM fractions are therefore consistent with the ele-mental and UV/Vis spectroscopic analyses discussed previously.The relative abundance of aromatic C@C and methoxyl functional groups appears to be in the order of SHA>NOM-PP>NOM-CH.Contents of carboxylic and alcoholic functional groups,however,are in the relative order of NOM-CH>NOM-PP>SHA, and the ketonic and phenolic functional groups are in the relative order of NOM-PP>NOM-CH>SHA. 3.3.FTIR spectroscopic analysisInfrared spectroscopy has also been widely used for gross characterization of humic substances and can pro-vide valuable information on the structural and func-tional properties of NOM molecules.Like13C-NMR Table2Peak areas of the interested13C-NMR spectral bands of NOM fractionsSample Alkyl0–50ppm O-Alkyl50–112ppmAromatic112–145ppmPhenolic145–163ppmCarboxyl163–190ppmPercentaromaticity a(%)NOM-CH 1.00 1.160.230.170.2614.2NOM-PP 1.00 2.48 1.680.74 1.1834.2SHA 1.00 1.48 1.970.420.4844.7The integrated areas are relative to the alkyl groups between0and50ppm.a Percent aromaticity¼(peak area of13C-NMR spectrum112–163ppm)/(total peak area of13C-NMR spectrum0–190ppm) (Malcolm,1990).J.Chen et al./Chemosphere48(2002)59–6863spectra,FTIR spectra of the three NOM fractions(Fig.3)show rather similar absorption bands since many similar structural and functional groups co-exist in these NOM macromolecules.As expected,these spectra also show similarities to those of humic substances isolated from both soil and aquatic environments(Stevenson and Goh,1971;Geyer et al.,1988;Senesi et al.,1989; Gu et al.,1994;Leenheer et al.,1995a;Barancikova et al.,1997).Differences in the relative intensity of some specific absorption bands were observed among these NOM fractions,although detailed interpretation of infrared spectra with respect to NOM structural features is chal-lenging due to significant overlapping of individual ab-sorptions at different vibrational modes by different functional groups.Major spectra bands were assigned as follows:2920cmÀ1(aliphatic C–H stretching),1725 cmÀ1(carboxyl and ketonic carbonyl stretching),1620 cmÀ1(aromatic C@C and conjugated carbonyl C@O), 1400cmÀ1(symmetrical stretching of COOÀ,OH de-formation,and C–O stretching of phenolic groups), 1200cmÀ1(C–O stretching and O–H deformation of COOH groups),and1050cmÀ1(stretching of carbohy-drate or alcoholic C–O).The SHA expectedly displayed a relatively higher IR absorption intensity and a broader peak width at$1620cmÀ1and at$2920cmÀ1than the NOM-PP and NOM-CH fractions.These results suggest that the relative order of abundance of aromatic com-pounds and aliphatic CH3/CH2groups in these NOM fractions is SHA>NOM-PP>NOM-CH,and is con-sistent with the13C-NMR and UV/Vis spectroscopic analyses(Figs.1and2).On the other hand,the relative absorption intensity at$1725cmÀ1for the NOM-CH fraction(in comparison with other absorption bands within the same NOM-CH spectrum)was stronger than that for NOM-PP and SHA.These results again confirm our previous observations that the NOM-CH fraction is among the most abundant in oxygen-containing func-tional groups as revealed by both elemental and13C-NMR analyses.In a study of the characterization of aquatic humic and fulvic materials by cylindrical inter-nal reflectance infrared spectroscopy,Marley et al. (1992)also reported an increased absorption intensity of carboxylic functional groups in the smaller size fractions of an aquatic NOM.Because the NOM-CH fraction is among the most soluble components of NOM,the NOM-CH is believed to consist of mostly small molec-ular weight organic compounds such as uronic acid and low molecular weight carbohydrates.3.4.Fluorescence spectroscopic analysisFluorescence spectroscopy has been used as a tech-nique for classifying and distinguishing between humic substances of various origins and natures(Miano et al., 1988;Senesi et al.,1991;Miano and Senesi,1992).The fluorescence emission spectra of NOM fractions and SHA are shown in Fig.4.The SHA exhibits a broad, featureless emission band with a maximum at$510nm; this band is consistent with data published by IHSS and other researchers(Senesi et al.,1989).Spectra of NOM-CH and NOM-PP fractions possess unique broad bands,with a maximum at$435–450nm and a shoulder at$475nm.The long wavelength and low intensity measured for thefluorescence peaks of SHA indicate the presence of condensed aromatic ring and other unsatu-rated bond systems,a high degree of conjugation,and electron-withdrawing groups such as carbonyl and carboxyl groups.The short wavelength and high inten-sities measured for NOM-PP and NOM-CH are asso-ciated with low aromatic content,low molecularweight64J.Chen et al./Chemosphere48(2002)59–68components,and electron-donating groups such as hy-droxyls,methoxyls,as has been commonly observed for HAs and FAs(Senesi et al.,1989,1991;Senesi,1990b).The emissionfluorescence of humic substances in the region440–480nm,however,is also influenced by polycondensation between carbonyls and amines,or phenolic structures.Becausefluorescence spectra result from the reactions of many differentfluorophores pre-sent in the NOM molecules,it would be difficult to identify the relevant structural and functional compo-nents responsible forfluorescence in these NOM com-ponents without other complementary spectroscopic data as discussed above(Senesi et al.,1989,1991;Gal-apate et al.,1990;Miano and Senesi,1992;Pullin and Cabaniss,1995).Synchronousfluorescence has been reported to have several advantages over conventionalfluorescence exci-tation and emission spectroscopy in studying NOM (Miano and Senesi,1992).Synchronous scan excitation spectra are obtained by measuring thefluorescence in-tensity while simultaneously scanning over both the ex-citation and emission wavelengths and maintaining a constant value for D k¼k emÀk ex.Synchronousfluo-rescence can provide better sensitivity and improved peak resolution compared to the conventional emission fluorescence technique,and possibly allow differentia-tion of thefluorescence spectra of samples of various origins(Galapate et al.,1990;Miano and Senesi,1992).As shown in Fig.5,the synchronous spectra of NOM-PP,NOM-CH,and SHA fractions exhibited spectral line shapes that were distinct from each other. The NOM-CH fraction showed a major spectrum with a single peak at380nm;the NOM-PP and SHA, however,exhibited spectra with a broad major peak at 350–450nm and a number of less defined minor peaks at$450nm.The SHA spectrum exhibited sharp peaks at a longer wavelength around460nm and two minor shoulders at480and490nm.In addition,the spectrum of SHA showed lower intensity and longer wavelength maximum than the NOM-PP spectrum,suggesting that SHA contains higher amounts of conjugated aromatic p-electron systems with electron-withdrawing functional groups,which are responsible for thefluorescence shift to lower energy levels or longer wavelengths(Guibault, 1973;Pullin and Cabaniss,1995).In comparison,the shorter wavelength and higherfluorescence intensity for the NOM-PP and the low intensity of NOM-CH indi-cate more simple structural features with fewer aromatic functional groups,conjugated chromophores,and elec-tron-donating substituents such as hydroxyl,methoxyl, and amino groups.3.5.EPR analysisEPR spectroscopy is a technique used to measure the free radical content of humic substances—a property that can greatly influence NOMs’redox property and reactivity with many organic and inorganic substrates in surface or subsurface environments(Wilson and Weber, 1977;Senesi et al.,1989;Senesi,1990a;Scott et al., 1998).EPR studies are based upon the measurement of the interaction between an external magneticfield and the magnetic moment of an unpaired(free radical) electron.The energy of the unpaired electron,E,can be described by the equation:E¼g b HM z;where g is the spectroscopic splitting factor(or so-called ‘‘g-value’’),b is the Bohr magneton,H represents the external magneticfield,and M z represents the two ori-entations of the electron spin(þ1/2orÀ1/2).Measure-ment of the samples’response to the external magnetic field can thus yield information about the number of unpaired electrons,in spins per gram of sample.A nu-clear hyperfine structure may arise from dipole–dipole interactions between unpaired electrons and their nu-cleus(Senesi,1990b).These hyperfine structures,when present,can thus provide information about the number and type of nuclei interacting with the free electron, which can be further reduced in terms of the chemical structure of the free radical.The g-values may also provide insights to the structure of free radicals because g-values will deviate from free electron values(g¼2:00232)by the molecular character of the paramagnetic species(Senesi,1990b).Fig.6illustrates thefirst-derivative EPR spectra of the three NOM fractions at pH10,and Table3pro-vides a summary of relevant EPR physical parame-ters.The spectra are characterized by an intensesingle J.Chen et al./Chemosphere48(2002)59–6865。
