Anal. Chem. 2002, 74, 2985-2993
Fast Quantification of Humic Substances and Organic Matter by Direct Analysis of Sediments Using DRIFT Spectroscopy Luc Tremblay and Jean-Pierre Gagne´*
Institut des sciences de la mer de Rimouski (ISMER), Universite´ du Que´ bec a` Rimouski, 310 Alle´ e des Ursulines, Rimouski, Que´ bec, G5L-3A1, Canada
A simple method based on diffuse reflectance coupled with infrared Fourier transform spectroscopy (DRIFTS) has been developed for the quantification and the characterization of sedimentary (or soil, peat, etc.) humic substances. Under optimized conditions, the quantification of humic substances or total organic matter is possible with DRIFTS at a frequency of 2930 cm-1 using whole dry sediment samples. A study of the operational parameters that affect the DRIFTS signal shows the importance of normalizing analysis conditions, especially the diffuse reflectance accessory alignment, the particle size and compaction, and the homogeneity of the powdered samples, to obtain reproducible quantitative analyses. The quantification of total humic substances by DRIFTS correlates well with the concentrations determined using classical extraction methods. DRIFTS analysis requires only a few minutes instead of tedious extractions of humic substances. Moreover, the distribution of total organic matter and of fulvic acids, humic acids, and humin can also be obtained. Analysis of natural samples indicates that a calibration using humic material representative of the studied area provides the most accurate quantification. The fast screening of organic matter fractions by DRIFTS on intact natural samples provides useful quantitative and qualitative information that can be used in environmental or monitoring studies. The accumulation and global cycling of organic matter (OM) in aquatic or terrestrial environments are of great ecological and economic interest. To understand the biogeochemistry of natural OM and its interactions with other compounds, it is essential to know the nature of the studied material. However, OM represents a complex mixture of molecules of mostly unidentified structures.1,2 Despite this lack of characterization, it is known that humic substances (HS) represent the main components of sedimentary organic material.3-6 These heterogeneous biogenic * Corresponding author. Tel: 418-723-1986, ext 1870. Fax: 418-724-1842. E-mail:
[email protected]. (1) Wakeham, S. G.; Lee, C.; Hedges, J. I.; Hernes, P. J.; Peterson, M. L. Geochim. Cosmochim. Acta 1997, 61 (24), 5363-5369. (2) Colombo, J. C.; Silverberg, N.; Gearing, J. N. Mar. Chem. 1996, 51, 295314. (3) Huc, A. Y. Ph.D. Thesis, Universite´ de Nancy, 1973. (4) Naik, S.; Poutanen, E. L. Oceanol. Acta 1984, 7 (4), 431-439. 10.1021/ac011043g CCC: $22.00 Published on Web 05/16/2002
© 2002 American Chemical Society
compounds are generally described as being refractory to degradation and having high molecular weight.7,8 The “classical” quantification method of HS, and the majority of the analytical techniques used for their study, require the extraction of these compounds. The extraction arbitrarily classifies organic material by its relative solubility in solutions of different pH and by its retention/elution on specific adsorbents. This fractionation scheme has the major disadvantage of requiring numerous time-consuming steps. Moreover, the extraction alters the OM.9-12 To overcome these limitations, there is a need for new approaches that perform direct analysis of OM and HS in natural matrixes.12,13 A fast and direct quantification technique for HS would provide a tool that could be used to improve understanding of their geochemistry, to identify HS deposits of commercial potential, and to satisfy the requirements of product registration and labeling. Solid-state 13C NMR using cross polarization with magic angle spinning is a spectroscopic technique frequently used to analyze OM and HS14 in their natural matrixes without extraction. However, this technique is difficult to use for quantitative analysis. In addition to the possible presence of paramagnetic impurities and spinning sidebands, the anisotropy is not easily corrected in the solid state and the variable relaxation rates observed for the numerous carbon bonds of HS widen the bands and distort signal (5) Yen, T. F. Chemistry of marine sediment; Ann Arbor Science: New York, 1977; p 265. (6) Rashid, M. A. Geochemistry of Marine Humic Compounds; Springer-Verlag: New York, 1985; p 300. (7) Aiken, G. R.; McKnight, D. M.; Wershaw, R. L.; MacCarthy, P. In Humic Substances in Soil, Sediment and Water; Aiken, G. R., McKnight, D. M., Wershaw, R. L., MacCarthy, P., Eds.; John Wiley and Sons: New York, 1985; Chapter 1. (8) Schnitzer, M.; Khan, S. U. Humic Substances in the Environment; Marcel Dekker: New York, 1972; p 327. (9) Francois, R. Rev. Aquat. Sci. 1990, 3 (1), 41-80. (10) Bremner, J. M. J. Agric. Sci. 1949, 39, 280-282. (11) Swift, R. S.; Posner, A. M. J. Soil Sci. 1972, 23, 381-393. (12) Swift, R. S. In Methods of Soil Analysis; Sparks, D. L., et al., Eds.; SSSA Book Series 5; Soil Science Society of America and American Society of Agronomy: Madison, WI, 1996; Chapter 35. (13) Thurman, E. M.; Aiken, G. R.; Ewald, M.; Fischer, W. R.; Fo¨rstner, U.; Hack, A. H.; Mantoura, R. F. C.; Parsons, J. W.; Pocklington, R.; Stevenson, F. J.; Swift, R. S.; Szpakowska, B. In Humic Substances and Their Role in the Environment; Frimmel, F. H., Christman, R. F., Eds.; John Wiley and Sons: Chichester, U.K., 1988; 31-43. (14) Wershaw, R. L. In Humic Substances in Soil, Sediment and Water; Aiken, G. R., McKnight, D. M., Wershaw, R. L., MacCarthy, P., Eds.; John Wiley and Sons: New York, 1985; Chapter 22.
Analytical Chemistry, Vol. 74, No. 13, July 1, 2002 2985
intensities.15,16 Moreover, solid-state NMR typically requires long acquisition times. Infrared spectroscopy (IR) is another common technique used for characterizing the nature and arrangement of functional groups in HS.17-23 However, the majority of these studies have been performed on extracted HS using KBr pellets. An alternative to classical transmission IR spectroscopy is diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). DRIFTS has several advantages: it is fast (5-20 min/sample), simple to use, sensitive, inexpensive, and nondestructive, requires only a few milligrams of solid material, and allows the analysis of opaque samples such as whole sediments or soils. In the last twenty years, DRIFTS has been used for the analysis of different compounds including polymers,24 glass fibers,25 quartz,26 coals,27 and cellulose products.28 In addition, DRIFTS does not have the drawbacks encountered with the traditional IR KBr pellet method, principally the incorporation of water.12,20,29 DRIFTS was previously applied to extracted HS,20-23 but like the DRIFTS analysis carried out on whole soils and sediments,30-32 these studies focused on organic functional group analysis or mineral characterization. No quantitative analysis was performed in these studies. The first objective of this research was to identify the most important factors that affect the analysis of HS by DRIFTS. The second objective was to develop an optimized protocol to directly quantify the different HS fractions and the total OM in whole dry sediments.
incident radiation is reflected at the solid’s surface to produce Fresnel reflectance (i.e., mirror reflection) while another part travels through the particles, where it interacts, and reemerges in many directions. Absorption and scattering of light into many directions by a powdered sample produce what is known as diffuse reflectance.33,34 The description of the diffuse reflectance generated by particles is given by the radiation transfer equation.33,34 Kubelka and Munk (K-M) proposed a simplified solution for this equation.29,33-36 According to K-M theory, the absolute diffuse reflectance of an “infinitely thick” layer is represented by R∞ in the K-M equation:
THEORY The interaction of light with a mixture of particles of different sizes, shapes, and composition is a complex process. Part of the
By substituting eq 2 into eq 1 for the relative diffuse reflectance, R′∞, we have
(15) Wershaw, R. L., Mikita, M. A., Eds. NMR of humic substances and coal; Lewis Publishers: Chelsea, MI, 1987. (16) Wilson, M. A. In Humic Substances in soil and crop sciences: selected readings; MacCarthy, P., Clapp, C. E., Malcolm, R. L., Bloom, P. R., Eds.; American Society of Agronomy and Soil Science Society of America: Madison, WI, 1990; Chapter 10. (17) Stevenson, F. J.; Goh, K. M. Geochim. Cosmochim. Acta 1971, 35, 471483. (18) MacCarthy, P.; Rice, J. A. In Humic Substances in Soil, Sediment and Water; Aiken, G. R., McKnight, D. M., Wershaw, R. L., MacCarthy, P., Eds.; John Wiley and Sons: New York, 1985; Chapter 21. (19) Tao, S.; Deng, B.; Hermosin, B.; Saiz-Jimenez, C. In Humic Substances in the Global Environment and Implication on Human Health; Senesi, N., Miano, T. M., Eds.; Elsevier Science: Amsterdam, 1994; pp 367-372. (20) Baes, A. U.; Bloom, P. R. Soil Sci. Soc. Am. J. 1989, 53, 695-700. (21) Niemeyer, J.; Chen, Y.; Bollag, J.-M. Soil Sci. Soc. Am. J. 1992, 56, 135140. (22) Francioso, O.; Sanchez-Cortes, S.; Tugnoli, V.; Marzadori, C.; Ciavatta, C. J. Mol. Struct. 2001, 566, 481-485. (23) Ding, G.; Amarasiriwardena, D.; Herbert, S.; Novak, J.; Xing, B. In Humic Substances: Versatile Components of Plants, Soil and Water; Ghabbour, E. A., Davies, G., Eds.; Royal Society of Chemistry: London, 2000; pp 53-62. (24) Garbassi, F.; Morra, M.; Occhiello, E. Polymer Surfaces, From Physics to Technology; John Wiley & Sons: New York, 1996. (25) Graf, R. T.; Koenig, P. R.; Ishida, H. Anal. Chem. 1984, 56, 773-777. (26) Koretsky, C. M.; Sverjensky, D. A.; Salisbury, J. W.; D’Aria, D. M. Geochim. Cosmochim. Acta 1997, 61 (11), 2193-2199. (27) Marinov, S. P.; Butuzova, L. F.; Krzton, A. Oxid. Commun. 2000, 23, 266. (28) Sereti, V.; Stamatis, H.; Pappas, C.; Polissiou, M.; Kolisis, F. N. Biotechnol. Bioeng. 2001, 72, 495-500. (29) Griffiths, P. R.; Fuller, M. P. In Advances in infrared and Raman spectroscopy; Clark R. J. H., Hester R. E., Eds.; Heyden and Son: London, 1982; pp 63129. (30) Nguyen, T. T.; Janik, L. J.; Raupach, M. Aust. J. Soil Res. 1991, 29, 49-67. (31) Ristori, G. G.; Sparvoli, E.; de Nobili, M.; D’Acqui, L. P. Geoderma 1992, 54, 295-305. (32) Bishop, J. L.; Koeberl, C.; Kralik, C.; Fro ¨schl, H.; Englert, P. A.; Andersen, D. W.; Pieters, C. M.; Wharton, R. A. Geochim. Cosmochim. Acta 1996, 60 (5), 765-785.
2986
Analytical Chemistry, Vol. 74, No. 13, July 1, 2002
f(R∞) ) (1 - R∞)2/2R∞ ) k/s
(1)
where f(R∞) is the K-M function (in K-M units), k is the absorption coefficient, and s is the scattering coefficient. In practice, the sample’s diffuse reflectance is measured with respect to a nonabsorbing reference solid, such as KCl or KBr. Thus, R∞ is divided by the reference reflectance and is replaced by R′∞. The absorption coefficient is associated with the sample absorptivity, a (at a given frequency) and concentration, c, by
k ) 2.303ac
f(R′∞) ) 2.303ac/s
(2)
(3)
Equation 3 suggests that f(R′∞) varies linearly with sample concentration. In practice, this equation is valid only if the Fresnel reflectance is negligibly small and if the scattering coefficient remains constant as the concentration varies.35 Since the scattering coefficient is controlled primarily by the particle refractive index, particle size, shape, and compaction,29,33 which also affect the magnitude of the Fresnel reflectance, control and optimization of the sample preparation parameters are essential to achieve the linearity of eq 3 and to obtain reproducible, quantitative analysis. Quantitative analysis by DRIFTS requires a very low level of Fresnel reflectance. When the Fresnel reflectance component is small, the DRIFTS spectrum of a sample has the same appearance as its transmittance spectrum.20,29 On the other hand, high Fresnel reflectance produces anomalous band dispersion for high-absorptivity bands34,35,37 and reduces the linear range of a DRIFTS calibration curve by producing a K-M signal approaching a constant value at high concentrations.29,37 As predicted by the Fresnel equation, the Fresnel reflectance component of the detected signal increases with the sample absorptivity and refrac(33) Kortu ¨ m, G. Reflectance Spectroscopy; Springer-Verlag: New York, 1969; p 366. (34) Hapke, B. Theory of Reflectance and Emittance Spectroscopy; Cambridge University Press: Cambridge, 1993; p 455. (35) Brimmer, P. J.; Griffiths, P. R.; Harrick, N. J. Appl. Spectrosc. 1986, 40, 258-265. (36) Kubelka, P.; Munk, Z. Z. Tech. Phys. 1931, 12, 593-601. (37) Brimmer, P. J.; Griffiths, P. R. Appl. Spectrosc. 1988, 42, 242-250.
tive index.29,34,35,38 Therefore, some experimental procedures can be used to substantially reduce the Fresnel reflectance. The dilution of the analyte in a matrix that does not absorb at the selected frequencies reduces the absorption, gives a contact medium with a refractive index closer to that of the analyte, compared to air, and thus reduces the Fresnel reflectance.29,33,35 However, the particle must be of the same size or smaller than the wavelength in order to consider effective contact of the analyte with the matrix.35 Moreover, large particles increase the mirror reflections.33,35 Fresnel reflectance is also reduced by a rough sample surface (i.e., not glossy), with particles randomly oriented.35,39 To keep a rough surface, the sample should not be tightly packed or mechanically leveled. A reduction of the particle size, or an increase in the packing density, diminishes the penetration depth of the incident radiation, produces more scattering, and attenuates the absorption and the K-M function value.33,38 Particle size and compaction of the sample can be controlled by using regulated and constant sample preparation, grinding, and packing procedures. EXPERIMENTAL SECTION Products and Sample Preparation. Potassium chloride (random cuttings, International Crystal Laboratories Co., Garfield, NJ) was used as a nonabsorbing reference material for the spectroscopic background measurement. Powdered KCl and fused quartz (99.995+% SiO2) from Quartz Scientific Inc. (Q.S.I., Fairport Harbor, OH) were used to dilute the humic acids (HA) in the preparation of standard solid solutions. Humic acids, containing less than 3% ash, were obtained from the International Humic Substances Society (IHSS) (Soil HA standards, Lot No. 1S1O2H) and from the extraction of a homogeneous batch of “reference” sediment. The reference sediment and all sediment samples for this study were collected at different stations in the St. Lawrence estuary ecosystem (Que´bec, Canada) including the Saguenay Fjord, a major tributary. Site descriptions can be found elsewhere.40,41 These sediments contain between 0.7 and 3.2% organic carbon as determined by elemental analysis. Sediment samples were freeze-dried prior to the extraction or DRIFTS analysis. Extraction of sedimentary fulvic and humic acids was performed according to the method developed by the IHSS for soil samples,12 adapted here for dry marine sedimentary material. The quantification of humin was done by combustion of the extraction residue. These results were used for the validation of the developed DRIFTS method. Before spectroscopic measurements, all solids were stored in a desiccator with CaSO4 and then ground with a Wig-L-Bug grinder (Crescent Mfg. Co., Lyons, IL) using stainless steel vials and balls. One minute of grinding with a single ball was used to process KCl cubes or humic acids. Fused quartz, purchased as 2-mm beads, was ground for 2 min using two balls in the vial (38) Pieters, C. M., Englert, P. A. J., Eds. Remote geochemical analysis: Elemental and mineralogical composition; Cambridge University Press: Cambridge, 1993; p 594. (39) Salisbury, J. W.; Walter, L. S.; Vergo, N.; D’Aria, D. M. Infrared (2.1-25 µm) Spectra of Minerals; Johns Hopkins University Press: London, 1991; p 267. (40) Deflandre, B.; Mucci, A.; Gagne´, J.-P.; Grignard, C.; Sundby, B. Geochim. Cosmochim. Acta 2002, in press. (41) El-Sabh, M. I., Silverberg, N., Eds. Oceanography of a large-scale estuarine system, the St. Lawrence; Springer: New York, 1990; p 434.
while freeze-dried sediments were ground for 5 min with one ball. For the preparation of standard solid solutions, the ground humic acids and diluents (KCl or fused quartz) were weighed into the vial cap and then shaken together for 1 min with no ball present in the vial, according to Hamadeh et al.42 Instrumentation and DRIFTS Analysis. A Perkin-Elmer (PE) model 1605 FT-IR spectrometer (Perkin-Elmer Corp., Norwalk, CT) was used in this study. All measurements were made with the PE model 0186-0791 diffuse reflectance accessory (DRA) (full π steradians, 38° angle of incidence) using an off-line geometry. The data processing was done by the PE Spectrum for Windows software. Figure 1 summarizes the procedure we developed. The selection of these analytical conditions is validated in the next section. Before background and sample analysis, the powder was poured into a sample cup, held between the thumb and the forefinger, and tapped several times on a hard surface to level the sample surface. Then the sample cup and the mirrors were aligned by maximizing the collected throughput energy. To control the energy conditions, a KCl background measurement was performed every 5 h because the IR energy detected by the spectrometer can fluctuate slightly depending on the ambient humidity and temperature conditions. To control these fluctuations, the ratio of the energy collected with the KCl matrix (EDRA) to the energy collected without the diffuse reflectance accessory (E-DRA) was monitored and kept constant, close to the maximal value, here 0.10. This value indicates that the background is measured with an optimized mirror alignment that determines the reference reflectance (i.e., KCl reflectance) used to calculate the R′∞ of the samples. The measurement chamber was flushed with nitrogen for 2 min before the scan acquisitions and during all spectrum acquisitions. For all spectra, 200 scans were collected between 4000 and 600 cm-1 using a resolution of 4 cm-1. The signal was transformed to K-M units by the spectroscopic software. Calibration with the humic acids standards and the quantification of HS and OM in sediments was done by measuring the signal maximum at 2930 cm-1 ((1 cm-1) and a baseline passing at 3010 and 2800 cm-1 on the spectra. For quantification in sediments, the calibration curve obtained with the humic acids extracted from Saguenay reference sediment was used. The particle size distributions used to determine the optimized grinding conditions were obtained with a Coulter Counter model LS100 (Coulter Electronics Inc., Hialeah, FL). Sedimentary total OM content was determined with a PE CHN model 140 elemental analyzer. The factor 2.5 was used to convert the organic carbon concentration (µg of C mg-1) into OM concentration (mg of OM g-1).43 RESULTS AND DISCUSSION Method Development. Spectroscopic analysis of a compound requires the choice of frequencies representative of that compound. Past qualitative analyses of HS by transmission or reflectance IR spectroscopy have produced spectra with relatively few bands that are often very broad.17-20 This apparent simplicity can be attributed to the signals’ summation and superimposition of functional groups that exist in diverse chemical environments (42) Hamadeh, I. M.; Yeboah, S. A.; Trumbull, K. A.; Griffiths, P. R. Appl. Spectrosc. 1984, 38, 486-491. (43) Saliot, A. Bioge´ ochimie organique marine; Oceanis 20; Institut Oce´anographique: Paris, 1994; p 93.
Analytical Chemistry, Vol. 74, No. 13, July 1, 2002
2987
Figure 1. Details of the developed methodology.
Figure 2. DRIFTS spectra of reference sediment separated by sieving the sample to give particle diameters (from top to bottom) of (A) 125-250, (B) 45-63, and (C)
