Water Interfaces. IV

Reddy , Timothy A. Johnson , Bruce Arey , Mark Bowden , and R. James Kirkpatrick ... L. Schmidt , Stephen J. Wall , Andrew Buhrlage and Heather A...
0 downloads 0 Views 280KB Size
Download PDF
5002

Langmuir 2005, 21, 5002-5012

Adsorption of Organic Matter at Mineral/Water Interfaces. IV. Adsorption of Humic Substances at Boehmite/Water Interfaces and Impact on Boehmite Dissolution Tae Hyun Yoon,† Stephen B. Johnson,† and Gordon E. Brown, Jr.*,†,‡ Surface and Aqueous Geochemistry Group, Department of Geological and Environmental Sciences, Stanford University, Stanford, California 94305-2115, and Stanford Synchrotron Radiation Laboratory, SLAC, 2575 Sand Hill Road, MS 69, Menlo Park, California 94025 Received September 23, 2004. In Final Form: March 12, 2005 The adsorption of Suwannee River fulvic acid (SRFA) and Pahokee peat humic acid (PPHA) at the boehmite (γ-AlOOH)/water interface and the impact of SRFA on boehmite dissolution have been examined over a wide range of solution pH conditions (pH 2-12), SRFA surface coverages (ΓSRFA, total SRFA binding site concentration normalized by the boehmite surface area) of 0.0-5.33 µmol m-2, and PPHA surface coverages (ΓPPHA, PPHA binding site concentration normalized by boehmite surface area) of 0.0-4.0 µmol m-2, using macroscopic adsorption and in situ attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy. At relatively high SRFA surface coverages (ΓSRFA ) 5.33 µmol m-2), in situ ATR-FTIR spectral features of adsorbed SRFA are very similar to those measured for SRFA in solution at approximately 1-3 pH units higher. At sub-monolayer surface coverages (ΓSRFA ) 1.20 and 2.20 µmol m-2), several new peaks and enhancements of the intensities of a number of existing peaks are observed. The latter spectral changes arise from several nonorganic extrinsic species (i.e., adsorbed carbonate and water, for alkaline solution conditions), partially protonated SRFA carboxyl functional groups (nearneutral pH conditions), and small quantities of inner-spherically adsorbed SRFA carboxyl groups and/or Al(III)-SRFA complexes (for acidic conditions). The spectra of PPHA adsorbed at boehmite/water interfaces also showed changes generally consistent with our observations for SRFA sorbed on boehmite. These observations confirm that SRFA and PPHA are predominantly adsorbed at the boehmite/water interface in an outer-sphere fashion, with minor inner-sphere adsorption complexes being formed only under quite acidic conditions. They also suggest that the positively charged boehmite/water interface stabilizes SRFA and PPHA carboxyl functional groups against protonation at lower pH. Measurements of the concentration of dissolved Al(III) ions in the absence and presence of SRFA showed that the boehmite dissolution process is clearly inhibited by the adsorption of SRFA, which is consistent with previous observations that outerspherically adsorbed organic anions inhibit Al-(oxyhydr)oxide dissolution.

Introduction Naturally occurring organic macromolecules (collectively referred to as humic substances) are ubiquitous in many natural settings and are formed primarily through microbial and chemical transformations of plant and animal wastes.1-4 As a result, humic substances consist of a highly heterogeneous mixture of organic macromolecules possessing widely varying molecular weights, functional group types and concentrations, and solution properties.5 Due to their roles as strong sorbents of many pollutant species, it is well-known that they have significant impacts on the chemical speciation and transport of pollutants.6 Moreover, when present in environments * Corresponding author. Phone: +1 650 723-9168. Fax: +1 650 725-2199. E-mail: [email protected]. † Stanford University. ‡ Stanford Synchrotron Radiation Laboratory. (1) Davies, G.; Ghabbour, E. A. Humic substances: Versatile components of plants, soil and water; Royal Society of Chemistry: Cambridge, U.K., 2000. (2) Hayes, M. H. B. Humic substances II: In search of structure; John Wiley & Sons: New York, 1989. (3) MacCarthy, P. Humic substances in soil and crop sciences: Selected readings. Proceedings of a Symposium Cosponsored by the International Humic Substances Society, Chicago, IL, Dec 2, 1985; American Society of Agronomy, Soil Science Society of America: Madison, WI, 1990. (4) Schnitzer, M. Adv. Agron. 2000, 68, 1-58. (5) Huang, W.; Peng, P.; Yu, Z.; Fu, J. Appl. Geochem. 2003, 18, 955-972. (6) Aiken, G. R. Humic substances in soil, sediment, and water: Geochemistry, isolation, and characterization; John Wiley & Sons: New York, 1985.

containing metal (oxy)hydroxide minerals (e.g., aluminum and iron (oxy)hydroxides), they are typically strongly associated with these mineral phases, particularly under acidic to near-neutral pH conditions, as demonstrated by numerous laboratory studies of the interaction of humic substances with these solids.7-11 Similarly, several investigations of natural sediment and soil samples have shown that the mineral components possess at least a partial surface coating of macromolecular organic materials.12-14 The presence of such organic coatings is of significant environmental importance because they can dramatically change the characteristics of colloidal mineral particles and thus can substantially alter the sorption capacity of metal (oxy)hydroxide minerals for a variety of solution-based metal and metalloid species,12,15-17 making (7) Evanko, C. R.; Dzombak, D. A. Environ. Sci. Technol. 1998, 32, 2846-2855. (8) Vermeer, A. W. P.; van Riemsdijk, W. H.; Koopal, L. K. Langmuir 1998, 14, 2810-2819. (9) Avena, M.; Koopal, L. Environ. Sci. Technol. 1999, 33, 27392744. (10) Filius, J. D.; Lumsdon, D. G.; Meeussen, J. C. L.; Hiemstra, T.; Van Riemsdijk, W. H. Geochim. Cosmochim. Acta 2000, 64, 51-60. (11) Filius, J. D.; Meeussen, J. C. L.; Lumsdon, D. G.; Hiemstra, T.; Van Riemsdijk, W. H. Geochim. Cosmochim. Acta 2003, 67, 14631474. (12) Neihof, R. A.; Loeb, G. I. J. Mar. Res. 1974, 32, 5-12. (13) Ransom, B.; Bennett, R.; Baerwald, R.; Shea, K. Mar. Geol. 1997, 138, 1-9. (14) Mayer, L. M.; Xing, B. S. Soil Sci. Soc. Am. J. 2001, 65, 250258. (15) Tipping, E. Geochim. Cosmochim. Acta 1981, 45, 191-199. (16) Davis, J. A. Geochim. Cosmochim. Acta 1984, 48, 679-691.

10.1021/la0476276 CCC: $30.25 © 2005 American Chemical Society Published on Web 04/28/2005

Organic Matter Adsorption at Mineral/Water Interfaces

Langmuir, Vol. 21, No. 11, 2005 5003

Figure 1. ATR-FTIR spectra (left) and schematic drawings (right) of (a) aqueous oxalate, (b) inner-spherically adsorbed oxalate, (c) aqueous maleate, and (d) outer-spherically adsorbed maleate at boehmite/water interfaces.

them potentially important environmental regulators of heavy metal contaminants. Additionally, sorption processes can (1) significantly impact the mechanism and extent of transport of organic matter, (2) protect organic substances against biodegradation through so-called carbon sequestration processes,18 and (3) influence mineral dissolution processes through ligand-promoted dissolution.19 Partitioning of natural organic matter (NOM) from aqueous solutions onto mineral surfaces can be described in terms of two fundamental processes: inner-sphere adsorption, which involves a ligand-exchange process between surface hydroxyl groups and organic functional groups; and outer-sphere adsorption, in which no ligandexchange reaction takes place, and the organic species is instead held at the surface through hydrogen bonding and/or electrostatic interactions. In general, inner-sphere modes of adsorption have been employed to interpret data obtained from macroscopic adsorption and/or titration studies of the interactions of natural organic matters10-11,20-21 with a variety of metal (oxy)hydroxide minerals. A number of studies have further indicated that carboxyl functional groups, particularly because of their high relative abundance, are strongly involved in innersphere complexation reaction(s),22,23 with the extent of inner-sphere binding increasing with decreasing pH.10,11 In addition to macroscopic adsorption and/or titration studies, several authors have utilized spectroscopic techniques to directly probe the complexation mode(s) of humic substances at metal (oxy)hydroxide surfaces. For example, Parfitt et al.20 and Gu and co-workers22,24 utilized ex situ infrared spectroscopic measurements on dried samples to examine the binding modes of humic substances on aluminum and/or iron (oxy)hydroxides. Marked changes in the spectra measured for the humic substances before and after adsorption to the mineral phase were interpreted as evidence for an inner-sphere mode of complexation at the mineral surface. These findings are in contrast to a recent communication by Yoon et al.,25 in which in situ attenuated total reflectance (ATR) infrared spectroscopy was used to examine the complexation modes of a model Suwannee River fulvic acid (SRFA) on boehmite (γAlOOH) particles in aqueous solution. The results of that study indicated that fulvic acid was bound predominantly in an outer-sphere complexation mode over a broad range of solution pH (pH ) 2.4-11.2). The surface concentration of fulvic acid used, however, was quite high (5.33 µmol/ m2), and led to substantial concentrations of nonadsorbed fulvic acid being present at all but the lowest pH conditions

examined. Spectral contributions from this solution-based fulvic acid could therefore potentially have obscured contributions due to inner-spherically bound fulvic acid. This possibility is examined in more detail in the present study using a larger range of surface coverages (0.0-5.33 µmol m-2), and the adsorption behavior of a reference humic acid has also been examined. An additional complication comes from changes in the macromolecular structure of SRFA as a function of pH and ionic strength, as shown by Myneni et al.26 using synchrotron-based transmission X-ray microscopy. In previous ATR-FTIR studies of the interaction of low molecular weight (LMW) dicarboxylic acid molecules with boehmite surfaces,27-29 we have shown that these molecules adsorb to boehmite in a manner largely determined by their molecular structures and the relative stabilities of their aqueous and adsorbed species, which can be determined by the combined use of in situ ATR-FTIR spectroscopy and quantum chemical simulation methods.27 For example, as illustrated in Figure 1, ATR-FTIR spectroscopy can distinguish between inner-sphere and outer-sphere adsorption modes of oxalate and maleate on the basis of the following reasoning. During inner-sphere adsorption of oxalate, the molecular structure around carboxyl functional groups experiences dramatic changes in bond angles, bond lengths, and molecular symmetry, resulting in significant changes in IR vibrational features (e.g., number, positions, and relative intensities of peaks) (17) Vermeer, A. W. P.; McCulloch, J. K.; Van Riemsdijk, W. H.; Koopal, L. K. Environ. Sci. Technol. 1999, 33, 3892-3897. (18) Kaiser, K.; Guggenberger, G. Org. Geochem. 2000, 31, 711-725. (19) Furrer, G.; Stumm, W. Geochim. Cosmochim. Acta 1986, 50, 1847-1860. (20) Parfitt, R. L.; Fraser, A. R.; Farmer, V. C. J. Soil Sci. 1977, 28, 289-296. (21) Hur, J.; Schlautman, M. A. J. Colloid Interface Sci. 2003, 264, 313-321. (22) Gu, B.; Schmitt, J.; Chen, Z.; Liang, L.; McCarthy, J. F. Environ. Sci. Technol. 1994, 28, 38-46. (23) McKnight, D. M.; Bencala, K. E.; Zellweger, G. W.; Aiken, G. R.; Feder, G. L.; Thorn, K. A. Environ. Sci. Technol. 1992, 26, 1388-1396. (24) Gu, B.; Schmitt, J.; Chen, Z.; Liang, L.; McCarthy, J. F. Geochim. Cosmochim. Acta 1995, 59, 219-229. (25) Yoon, T. H.; Johnson, S. B.; Brown, G. E., Jr. Langmuir 2004, 20, 5655-5658. (26) Myneni, S. C. B.; Brown, J. T.; Martinez, G. A.; Meyer-Llse, W. Science 1999, 286, 1335-1337. (27) Yoon, T. H.; Johnson, S. B.; Musgrave, C.; Brown, G. E., Jr. Geochim. Cosmochim. Acta 2004, 68, 4505-4518. (28) Johnson, S. B.; Yoon, T. H.; Kocar, B. D.; Brown, G. E., Jr. Langmuir 2004, 20, 4996-5006. (29) Johnson, S. B.; Yoon, T. H.; Slowey, A. J.; Brown, G. E., Jr. Langmuir 2004, 20, 11480-11492.

5004

Langmuir, Vol. 21, No. 11, 2005

of adsorbed (Figure 1b) vs aqueous oxalate27 (Figure 1a). In contrast, maleate forms dominantly outer-sphere complexes with the boehmite surface,28,30 and the IR spectrum of the resulting sorption complex (Figure 1d) shows only minor changes in peak positions and intensities relative to the aqueous maleate complex (Figure 1c). Similar reasoning has been applied in a number of previous ATR-FTIR studies of the adsorption modes of organic anions on mineral surfaces27-41 and has also been used in the present study to determine the dominant adsorption mode(s) of SRFA at the boehmite/water interface. In the present study, we have extended the work of Yoon et al.25 and have confirmed that the major adsorption mode of humic substances to boehmite surfaces is outer sphere over a wide range of SRFA and Pahokee peat humic acid (PPHA) concentrations (surface coverage of 0.0-5.33 µmol m-2 for SRFA and 0.0-4.0 µmol m-2 for PPHA, assuming complete sorption) and pH values (2.44-11.18 for SRFA and 3.85-11.13 for PPHA). Due to their high concentration relative to other functional groups in SRFA and PPHA,42 our primary focus has been on the role(s) of the carboxyl groups in the adsorption process. Additionally, we performed experiments on the impact of SRFA adsorption on boehmite dissolution, which showed that adsorbed SRFA inhibits boehmite dissolution. Boehmite was chosen as a representative mineral particle because it is a common fine-grained constituent of soils and sediments due to its thermodynamic stability under hydrous conditions,43 and it has a very high surface area, which facilitates detection of small amounts of SRFA adsorbed at the boehmite/water interface. Suwannee River fulvic acid and Pahokee peat humic acid were selected as representative natural organic matter (NOM) compounds, since both have been relatively well characterized using various spectroscopic and macroscopic tools and both have been used as reference NOM materials in many previous studies.44-47 Where appropriate, results have been compared with spectroscopic and adsorption findings from previous studies of the interactions of humic substances (30) Rosenqvist, J.; Axe, K.; Sjoberg, S.; Persson, P. Colloids Surf., A 2003, 220, 91-104. (31) Persson, P.; Nordin, J.; Rosenqvist, J.; Lovgren, L.; Ohman, L. O.; Sjoberg, S. J. Colloid Interface Sci. 1998, 206, 252-266. (32) Degenhardt, J.; McQuillan, A. J. Chem. Phys. Lett. 1999, 311, 179-184. (33) Dobson, K.; McQuillan, A. Spectrochim. Acta, Part A 1999, 55, 1395-1405. (34) Dobson, K.; McQuillan, A. Spectrochim. Acta, Part A 2000, 56, 557-565. (35) Boily, J. F.; Nilsson, N.; Persson, P.; Sjoberg, S. Langmuir 2000, 16, 5719-5729. (36) Boily, J. F.; Persson, P.; Sjoberg, S. Geochim. Cosmochim. Acta 2000, 64, 3453-3470. (37) Nordin, J.; Persson, P.; Laiti, E.; Sjoberg, S. Langmuir 1997, 13, 4085-4093. (38) Nordin, J.; Persson, P.; Nordin, A.; Sjoberg, S. Langmuir 1998, 14, 3655-3662. (39) Axe, K.; Persson, P. Geochim. Cosmochim. Acta 2001, 65, 44814492. (40) Kubicki, J. D.; Schroeter, L. M.; Itoh, M. J.; Nguyen, B. N.; Apitz, S. E. Geochim. Cosmochim. Acta 1999, 63, 2709-2725. (41) Kubicki, J. D.; Itoh, M. J.; Schroeter, L. M.; Apitz, S. E. Environ. Sci. Technol. 1997, 31, 1151-1156. (42) IHSS Standard and Reference Collection Data Sheet; International Humic Substances Society: St Paul, MN, 2002. (43) Navrotsky, A. Rev. Mineral. Geochem. 2001, 44, 73-104. (44) Leenheer, J.; Rostad, C.; Gates, P.; Furlong, E.; Ferrer, I. Anal. Chem. 2001, 73, 1461-1471. (45) Leenheer, J.; Wershaw, R.; Reddy, M. Environ. Sci. Technol. 1995, 29, 393-398. (46) Leenheer, J.; Wershaw, R.; Reddy, M. Environ. Sci. Technol. 1995, 29, 399-405. (47) Leenheer, J.; McKinght, D.; Thurman, E.; MacCarthy, P. In Humic Substances in the Suwannee River, Georgia: Interactions, Properties and Proposed Structures; Averett, R. C., Leenheer, J. A., McKnight, D. M., Thorn, K. A., Eds.; U. S. Geological Survey: Denver, CO, 1994.

Yoon et al.

with mineral surfaces. The implications of SRFA adsorption mode(s) for the dissolution behavior of boehmite have also been examined and are discussed. Experimental Methods A Suwannee River reference fulvic acid (1R101F-1) and Pahokee peat humic acid (1R103H-2) were obtained from the International Humic Substances Society (IHSS) and were used as received. They contained carboxyl and phenolic concentrations of 11.21 and 2.89 mol kg-1 for SRFA and 8.87 and 2.05 mol kg-1 for PPHA.47 Boehmite (γ-AlOOH) used in this study was purchased from Condea Chemie GmBH and characterized using BET surface area measurements (BET area, 270 m2/g) and X-ray diffraction (XRD) analysis. Milli-Q water and a background electrolyte concentration of 0.01 M NaCl were used in all experiments. Solution and suspension pH values were adjusted using small volumes of concentrated (1-10 mol dm-3) HCl and NaOH. The surface area normalized concentration of SRFA (ΓSRFA) and PPHA (ΓPPHA) were obtained by dividing the total concentration of functional groups (both carboxylic and phenolic) concentrations of SRFA or PPHA by the total surface area of boehmite. Boehmite-fulvic acid or boehmite-humic acid suspensions were prepared using a standard batch equilibration technique. Briefly, quantities of a pre-dispersed boehmite suspension and SRFA (or PPHA) were added to polypropylene centrifuge tubes to give a final boehmite concentration of 25 g dm-3 and a total surface concentration of major functional groups (both carboxyl and phenolic) of 1.20, 2.20, and 5.33 µmol m-2 for ΓSRFA and 1.0, 2.0, 3.0, and 4.0 for ΓPPHA when fully adsorbed. The pH of the samples was adjusted with HCl or NaOH, and the sample tubes were tumbled end-over-end in the dark for 48 h. No control of carbonate in this system was made during our experiments in order to examine the adsorption modes of humic substances under conditions close to those in “natural” settings (i.e. in the presence of CO2 and bulk water at ambient temperature). The final pH was then measured before the samples were centrifuged at 10 000 rpm for 30 min, and the supernatants were passed through 20 nm syringe filters. Residual fulvic acid concentrations in the supernatant were determined by UV spectrophotometry (HP 8452A diode array spectrophotometer) at pH ) 3 and λ ) 220 nm, a wavelength that was experimentally determined to be insensitive to both slight changes in pH and the presence of Al(III) in solution. Dissolved aluminum concentrations in these filtered supernatants were also determined by inductively coupled plasma (ICP) spectrometry (TJA IRIS Advantage/1000 radial ICAP spectrometer) and were used to check the progress of boehmite dissolution during these batch experiments for various SRFA concentrations and pH conditions. We developed this macroscopic adsorption methodology carefully in order to ensure that experimental uncertainties were minimized and the effects of dissolved Al(III) were negligible. Therefore, macroscopic adsorption experiments are consistent with previously published studies. ATR-FTIR measurements were performed on aqueous fulvic and humic acids and centrifuged boehmite-SRFA suspensions using a Nicolet Nexus 470 FTIR spectrometer equipped with a horizontal attenuated total reflectance accessory and a Ge crystal (PIKE Technologies). Data collection and spectral calculations were accomplished using OMNIC software (version 6.0a, Thermo Nicolet, Wisconsin). In each case, the FTIR spectrum was extracted from the overall ATR-FTIR response by subtracting the spectrum of an aqueous 0.01 M NaCl solution measured at an equivalent pH condition. In this manner, the FTIR contributions of water, which would otherwise dominate the FTIR spectra, were eliminated. Some ATR-FTIR experiments were replicated to check the reproducibility of our observations, and any spectra that were in doubt were also confirmed by duplicate measurements. In general, our ATR-FTIR measurements are reproducible in terms of both peak positions and intensities.

Results and Discussion ATR-FTIR Study of Aqueous SRFA and PPHA. ATR-FTIR spectra of aqueous fulvic and humic acids in the 2000 to 1100 cm-1 region are shown as a function of

Organic Matter Adsorption at Mineral/Water Interfaces

Langmuir, Vol. 21, No. 11, 2005 5005

Figure 2. ATR-FTIR spectra of (a) Suwannee River fulvic acid in powdered form and in aqueous solution as a function of pH, and (b) Pahokee peat humic acid in aqueous solution as a function of pH.

pH in Figure 2a,b. The data of Figure 2a have been presented in a previous communication25 and are shown again here for the purposes of comparison with the new data for PPHA. The total concentration of major functional groups (carboxyl and phenolic) was approximately 0.2 M for all measurements. Data are presented for pH values down to 2.20 for aqueous SRFA, but only down to pH ) 3.50 for PPHA, as PPHA was found to become significantly less soluble under more acidic conditions. At the highest pH values examined (12.44 and 12.05 for FA and HA, respectively), two prominent peaks are observed (Figure 2a,b), centered at 1568 ( 3 and 1391 ( 4 cm-1. These peaks are characteristic of COO- stretching and gradually diminish in intensity as pH decreases. They can be assigned to the asymmetric (νas, 1568 ( 3 cm-1) and symmetric (νs, 1391 ( 4 cm-1) stretching modes of deprotonated carboxyl moieties,33,34,48-50 which are abundant in SRFA and PPHA.47,56 Interestingly, the spectra of both SRFA and PPHA at high pH contain quite broad features compared with those observed for more simple organic anions under similar pH conditions. For example, the full width at half-maximum (fwhm) of the νas peak observed at high pH in Figure 2a,b is 75 ( 6 cm-1, compared with only 52 cm-1 fwhm observed for the deprotonated oxalate anion.29 The relatively broadened nature of the SRFA and PPHA peaks can be attributed to the heterogeneous nature of the SRFA and PPHA molecules, which leads to a wide distribution of chemical environments for carboxyl moieties (e.g., aliphatic versus aromatic environments, with a range of different neighboring functional groups).44-47 Parts a and b of Figure 2 show that both νas and νs gradually shift to higher wavenumbers as pH is reduced. For example, for SRFA at pH ) 3.74, νas is shifted by +28 (48) Cabaniss, S. E. Anal. Chim. Acta 1991, 255, 23-30. (49) Cabaniss, S. E.; McVey, I. Spectrochim. Acta, Part A 1995, 51, 2385-2395. (50) Cabaniss, S. E.; Leenheer, J.; McVey, I. Spectrochim. Acta, Part A 1998, 54, 449-458.

cm-1 compared with the position observed at the highest pH values examined. Similarly, for PPHA at pH ) 3.50, νas is shifted to 1588 cm-1, a 17 cm-1 increase from the value observed at high pH. Such findings are likely to be due to the effects of intramolecular hydrogen bonding between neighboring carboxyl groups, the extent of which will presumably increase as the degree of carboxyl protonation rises at low pH. Interestingly, similar lowpH behavior has recently been reported in the ATR-FTIR spectra of aqueous aromatic LMW molecules containing multiple carboxyl groups (e.g., mellitate).51 At the lowest pH conditions examined, Figure 2a,b show two peaks in addition to those observed at high pH. These are positioned at 1250 and 1715 cm-1 for SRFA and at 1263 and 1717 cm-1 for PPHA. These peaks are a function of the increased protonation of the SRFA and PPHA molecules and can be assigned to the C-O (1257 ( 7 cm-1) and CdO (1716 ( 1 cm-1) stretches of the protonated carboxyl groups. In addition to the carboxyl vibrations discussed above, Figure 2a,b show a broad absorption peak in the spectral region from 1650 to 1800 cm-1. This peak is present at all but the highest pH values investigated and can be attributed to CdO stretching vibrations characteristic of esters and ketones, both of which are present in substantial quantities in SRFA.47 Additionally, a small peak at 1265 cm-1 was observed for the SRFA over the pH range examined and lies in close proximity to the C-O vibration previously reported for phenolic type groups in LMW organic acids (e.g., νC-O ) 1240 cm-1 for phenol,52 1255 cm-1 for salicylate,53,54 and 1256 cm-1 for catecholate55). (51) Johnson, B. B.; Sjoberg, S.; Persson, P. Langmuir 2004, 20, 823828. (52) Tejedor-Tejedor, M.; Yost, E.; Anderson, M. Langmuir 1990, 6, 979-987. (53) Humbert, B.; Alnot, M.; Quiles, F. Spectroc. Acta, Part A 1998, 54, 465-476. (54) Yost, E. C.; Tejedor-Tejedor, I.; Anderson, M. A. Environ. Sci. Technol. 1990, 24, 822-828. (55) Connor, P.; Dobson, K.; Mcquillan, A. Langmuir 1995, 11, 41934195.

5006

Langmuir, Vol. 21, No. 11, 2005

Yoon et al.

Table 1. Assignments of Various IR Peaks from Aqueous Phase Suwannee River Fulvic Acid and Pahokee Peat Humic Acid peak position (cm-1) SRFA

PPHA

assignment

1250 1265 1395 1565 1715 1745

1263 1264 1387 1571 1717 1747

C-O stretching of carboxylic acid Ar-O stretching of phenolic group symmetric COO- stretching of carboxylate asymmetric COO- stretching of carboxylate CdO stretching of carboxylic acid CdO stretching of ester or ketone

It is therefore assigned to the C-O stretch of phenolbased groups that are prevalent in SRFA molecules. Interestingly, a similar but almost indistinct peak in the vicinity of 1264 cm-1 was observed for PPHA under neutral pH conditions (see Figure 2b), a finding that is presumably a function of the substantially larger heterogeneity and lower phenolic content of PPHA compared with that of SRFA.47,56 It should be noted that many of the above observations and peak assignments for SRFA and PPHA, which are summarized in Table 1, are consistent with previous ATR-FTIR results reported for aqueous fulvic acids from Lake Drummond.57 In addition to these pH-dependent changes of aqueous SRFA and PPHA ATR-FTIR spectra, it is also of interest to examine the effect of hydration (and therefore the impact of in situ versus ex situ spectroscopic measurements) on the measured ATR-FTIR spectra. An example of this effect is demonstrated in Figure 2a, in which the ATRFTIR spectra of SRFA in its aqueous and dry (dehydrated) forms are shown. The FTIR spectrum of dry, powdered fulvic acid is mostly similar to the aqueous spectrum obtained under very acidic conditions (pH ) 2.20), but it also displays a number of distinct features compared with the aqueous SRFA spectrum obtained at pH ) 2.20. These include (1) broadening of peak shapes and weakening of the intensities of the symmetric stretching vibration (νs) of the deprotonated carboxyl moiety and the C-O stretching vibration of the protonated carboxyl group, (2) a further shift of the asymmetric stretching vibration (νas) of the deprotonated carboxyl moieties, and (3) disappearance of peaks from minor functional groups such as phenolic (1265 cm-1) and ketone/ester (1745 cm-1) type moieties. These differences between the aqueous and dry, powdered SRFA ATR-FTIR spectra indicate that hydration has a substantial impact on the local chemical environments of the SRFA carboxyl functional groups, and therefore on their characteristic vibrational frequencies. These marked differences between dried and aqueous SRFA ATR-FTIR spectra highlight the importance of using an in situ spectroscopic method for studying fulvic acid- and humic acid-containing systems under hydrous conditions. Based on the above peak assignments, FTIR investigations in the present study have been focused on the role of the most prevalent SRFA functional groups, the carboxyls, in the adsorption of SRFA on boehmite. Other functional groups such as amino acids, sulfate esters, and phosphate esters are present in only minor concentrations (e.g., 0.7% N, 0.6% S, and 8 (Figure 3). The corresponding ATRFTIR spectra obtained at ΓSRFA ) 2.20 and 1.20 µmol m-2 are shown in Figure 4b,c, respectively. Many of the major spectral features displayed by aqueous-based SRFA (Figure 2a) are again present in the spectra of SRFA adsorbed on boehmite (Figure 4b,c), consistent with outer-sphere adsorption of a significant proportion of the SRFA carboxyl groups at ΓSRFA ) 2.20 and 1.20 µmol/m2. When compared with the spectra in Figure 4a (ΓSRFA ) 5.33 µmol/m2), a number of additional features are evident in Figure 4b,c, most notably the emergence of several new peaks and enhancements of the intensities of a number of existing peaks. For example, in addition to major spectral features observed for the ΓSRFA ) 5.33 sample, significant increases in the intensities of shoulder peaks around 1445 and 1600 cm-1 were observed as pH decreased below 3 (Figure 4b,c). Similarly, a substantial increase in the intensity of the peak positioned at 1398 cm-1, and the emergence of two additional peaks at 1506 and 1664 cm-1, can be seen at solution pH values above 7.12 and 5.02 for ΓSRFA ) 2.20 and 1.20 µmol m-2, respectively, while these peaks were only observable at pH values above 11.18 for ΓSRFA ) 5.33 µmol m-2. In addition, under very acidic solution conditions (pH ) 2.03 and 2.24 for ΓSRFA ) 2.20 and 1.20 µmol m-2, respectively), we observed a substantially higher relative intensity of the 1568 cm-1 versus the 1712 cm-1 peak (i.e., I1568/1712 ) 3.82 and 3.63 for ΓSRFA ) 2.20 and 1.20 µmol m-2, respectively), compared with that found under the acidic solution condition (pH ) 2.44) for ΓSRFA ) 5.33 µmol m-2 (i.e., I1568/1712 ) 0.87). This latter observation indicates that SRFA carboxyl groups are less protonated at these undersaturated (i.e., sub-monolayer) coverages. This suggestion is consistent with a greater extent of interaction between the SRFA carboxyl groups and boehmite surface groups (which stabilize SRFA against protonation25) at ΓSRFA ) 1.20 and 2.20 µmol m-2. The presence of new spectral features at 1445, 1600, 1398, 1506, and 1664 cm-1 observed in the ΓSRFA ) 1.20 and 2.20 µmol m-2 samples could be indicative of an innersphere adsorption mode of the fulvic acid carboxyl groups, but they may also have arisen due to background species including boehmite and/or other nonorganic extrinsic

Organic Matter Adsorption at Mineral/Water Interfaces

Langmuir, Vol. 21, No. 11, 2005 5009

Figure 5. ATR-FTIR spectra of (a) SRFA adsorbed to boehmite at pH ) 6.95, ΓSRFA ) 1.20 µmol m-2; (b) boehmite at pH ) 7.02, ΓSRFA ) 0.00 µmol m-2; and (c) spectrum a - spectrum b × 0.72.

species at the boehmite/water interface. Such species may not have been clearly observed in ATR-FTIR spectra obtained at higher SRFA concentrations (see Figure 4a) if they were present in minor quantities and/or have small absorption coefficients and so were potentially obscured by the presence of the dominant outer-spherically adsorbed carboxyls and/or SRFA carboxyls not directly interacting with the boehmite surface. To further investigate the possibility of significant spectral contributions from extrinsic species or the boehmite surface, pH-dependent spectra of boehmite in 0.01 M NaCl solution alone (i.e., ΓSRFA ) 0.0 µmol m-2) were measured. These spectra are presented in Figure 4d and contain a number of features in common (i.e., peaks at 1664, 1506, and 1395 cm-1) with previous spectra obtained for SRFA adsorbed to boehmite at ΓSRFA ) 2.20 and 1.20 µmol m-2 (Figure 4b,c). In particular, two prominent peaks located at 1506 and 1395 cm-1 in Figure 4d increase markedly in intensity under alkaline pH conditions. They are attributable to adsorbed carbonate species27,61 and arise due to solubilization of (and subsequent adsorption of the anionic form of) CO2 from the atmosphere. In contrast, the peak at 1664 cm-1 can be explained as being due to incomplete subtraction of the spectral contributions from a small amount of physisorbed water at the boehmite/ water interface. The above peak assignments are further examined in Figure 5, where the ATR-FTIR spectrum of boehmite in the absence of SRFA at pH ) 7.02 (from Figure 4d) is subtracted from that of SRFA-boehmite at ΓSRFA ) 1.2 µmol m-2 and pH ) 6.95 (from Figure 4c). As is shown in Figure 5c, only the peaks due to νas and νs remain following this subtraction. This result provides additional support for the assignment of the new spectral features observed at alkaline pH conditions and low surface coverages of SRFA to nonorganic extrinsic species (i.e., to adsorbed carbonate and water) present at the boehmite/water interface rather than to inner-spherically bound SRFA species. There are several potential explanations for the new spectral features observed under acidic pH conditions (i.e., the increased intensities of the shoulders near 1445 and (61) Bargar, J.; Reitmeyer, R.; Davis, J. Environ. Sci. Technol. 1999, 33, 2481-2484.

Figure 6. ATR-FTIR spectra of (a) SRFA adsorbed to boehmite at pH ) 3.06, ΓSRFA ) 1.20 µmol m-2; (b) boehmite at pH ) 3.10, ΓSRFA ) 0.00 µmol m-2; and (c) aqueous SRFA at pH ) 2.20. (A) Contribution from adsorbed water; (B) contribution from νas and νs stretching of carboxylate group shifted due to the protonation of neighboring carboxylic groups; and (C) contribution from adsorbed carbonate.

1600 cm-1 in Figure 4b,c). One plausible explanation is that these new features are due to νas and νs stretching frequencies of the carboxyl group, which are shifted to higher wavenumber as protonation of neighboring carboxyl groups proceeds and as the contribution from intramolecular hydrogen bonding increases. Similar behavior was observed in aqueous SRFA spectra (Figure 2a) as well as in the spectra of aqueous aromatic LMW organic molecules containing multiple carboxyl groups (e.g., mellitate)51 as pH decreased. As shown in Figure 6, the features at 1445 and 1600 cm-1 in both the spectrum of adsorbed SRFA (Figure 6a) and the spectrum of aqueous SRFA under acidic conditions (Figure 6c) match reasonably well in terms of position. These observations indicate that these spectral features are mainly caused by partial deprotonation of SRFA at the boehmite/water interface. However, pH-dependent changes of aqueous SRFA at 1445 and 1600 cm-1 (Figure 2a) are more dramatic than those observed for the adsorbed SRFA species (Figure 4a-c). Thus, we believe that this explanation could not account for all these observed changes, although they may have contributed to the spectral features in these regions to a significant extent. A second possible explanation for the spectral changes observed in the vicinity of 1445 and 1600 cm-1 at low pH (Figure 4b,c) is the formation of inner-sphere complexes between SRFA carboxyl groups and boehmite functional groups. For example, Leenheer et al.45,46 previously reported that strongly binding moieties such as malonatelike groups may be present in SRFA. Recent ATR-FTIR studies30 have shown that malonate anions adsorb in a dominantly inner-sphere manner at the boehmite/water interface under acidic pH conditions, a process that

5010

Langmuir, Vol. 21, No. 11, 2005

Yoon et al.

Table 2. Assignments of Various IR Peaks from Suwannee River Fulvic Acid and Pahokee Peat Humic Acid Adsorbed at Boehmite/Water Interfaces peak position (cm-1)

assignment

sample conditions

1157 1395 1398 1445

Al-O stretching of boehmite adsorbed carbonate symmetric COO- stretching of carboxylate symmetric COO- stretching of carboxylate involving partial protonation of neighboring carboxyl groups adsorbed carbonate asymmetric COO- stretching of carboxylate asymmetric COO- stretching of carboxylate involving partial protonation of neighboring carboxyl groups adsorbed water CdO stretching of carboxylic acid CdO stretching of ester or ketone

all pH neutral to alkaline neutral to alkaline acidic to neutral

1506 1568 1600 1664 1712 1745

generates new peaks in the vicinity of 1603 and 1437 cm-1. However, while the latter peak positions are similar to those observed in Figure 4b,c, malonate-type groups are likely to be present in SRFA in only very minor concentrations45,46 and so cannot alone account for the marked spectral changes observed in the vicinity of 1445 and 1600 cm-1. In addition, however, phthalate-type groups are likely to be present to a significantly greater extent in SRFA molecules, where they have been implicated as potential major metal binding sites.47 Using in situ ATRFTIR spectroscopy, Persson et al.31 demonstrated that phthalate is capable of forming inner-sphere complexes with boehmite at pH e 5, as evidenced by a new ATRFTIR peak emerging at 1422 cm-1. The formation of minor inner-sphere complexes by phthalate-like moieties in SRFA may, therefore, also have contributed to the spectral changes observed in the vicinity of 1445 cm-1 at low pH and ΓSRFA ) 1.20 and 2.20 µmol/m2. A third plausible explanation of the spectral features at 1445 and 1600 cm-1 is that solution-facing carboxyl functional groups complex Al(III) cations as they dissolve from the boehmite surface. As can be seen in the following section, the dissolution rate of boehmite increases rapidly as pH decreases below 4.5, and as a consequence, a significant concentration of Al(III) cations is available to participate in the formation of Al(III)-SRFA complexes. A recent ex situ FTIR study of Al(III)-SRFA complexes62 found two new peaks at 1638 and 1460 cm-1 when SRFA binds to Al(III) in solution. The peaks are within 38 and 15 cm-1, respectively, of the new spectral features observed at 1600 and 1445 cm-1 in the SRFA-boehmite system under acidic conditions and (given the broadness of all features in the SRFA ATR-FTIR spectra) may therefore have contributed to these absorption peaks. The fact that the latter study was undertaken under ex situ conditions (compared with the in situ conditions used in this study) may well mean that a better spectral match between these two systems would have been obtained if both spectra had been obtained under in situ conditions. We conclude that the new ATR-FTIR peaks under alkaline conditions are mostly generated by nonorganic extrinsic species (i.e., adsorbed carbonate and water, for alkaline condition), whereas those new peaks observed under acidic pH conditions are mainly due to contributions from (1) partially protonated SRFA carboxyl functional groups, (2) small quantities of inner-spherically adsorbed SRFA carboxyl groups, and (3) Al(III)-SRFA complexes caused by dissolved Al(III) cations (see Table 2 and Figure 6 for detailed assignments). Although we found evidence for the possible existence of inner-spherically adsorbed SRFA species under acidic conditions, their contribution (62) Elkins, K.; Nelson, D. J. Inorg. Biochem. 2001, 87, 81-96.

neutral to alkaline neutral to alkaline acidic to neutral neutral to alkaline acidic all pH

is minor compared to the dominant contributions from outer-spherically adsorbed SRFA carboxyl groups and/or intermolecular interactions at higher SRFA surface coverage. Thus, SRFA is predominantly outer-spherically adsorbed to the boehmite surface over the entire range of pH conditions investigated, with minor contributions from inner-spherically adsorbed carboxyl groups emerging only under quite acidic pH conditions. Our results are generally consistent with the previous findings of Filius et al.,11 based on surface complexation modeling, which predict the dominance of outer-sphere adsorption modes for SRFA carboxyl groups on goethite over a broad pH range (pH ) 3-11), and the systematic emergence of inner-sphere modes as pH is decreased. In the study of Filius et al.,11 however, the relative importance of inner-spherically versus outer-spherically adsorbed carboxyl species was predicted to be somewhat higher than observed heresa finding that may be attributable to the different metal (oxyhydr)oxide substrates utilized in the two studies. In addition to carboxyl functional groups, inner-sphere mode adsorption involving phenolic groups (e.g., salicylic or catecholic groups) may also be possible. For salicylate and catecholate, the formation of inner-sphere mode surface complexes has been previously reported,55,63,64 and the lack of salicylate-specific peaks at 1502, 1454, and 1256 cm-1 in samples of aqueous/adsorbed SRFA at higher pH64 and adsorbed catecholate-specific peaks at 1480 and 1250 cm-1 in samples of adsorbed SRFA55,63 imply that these types of functional groups are relatively insignificant compared to the predominant carboxyl functional group in the adsorption of SRFA on boehmite. However, phenolic group-specific peaks (i.e., ν(Ar-O-) ∼ 1250 cm-1)53 have relatively low absorption coefficients and/or low site densities in SRFA and also are located near the large boehmite peaks. Therefore, it is difficult to determine if phenolic groups are involved in SRFA-boehmite interactions. More systematic in situ spectroscopic studies are necessary to unequivocally determine SRFA adsorption modes involving phenolic functional groups. ATR-FTIR spectra taken from PPHA adsorbed at the boehmite/water interface under various pH conditions (3.85-11.13) and surface coverages (0-4 µmol/m2) are presented in Figure 7a,b, respectively. Despite the different characteristics of PPHA relative to SRFA (e.g., different molecular weights, binding site concentrations, and sources of materials56), the spectra of the PPHA/ boehmite samples are generally consistent with our observations for SRFA sorbed on boehmite. For example, when no PPHA is present (ΓPPHA ) 0.0 µmol/m2), only (63) McWhirter, M.; Bremer, P.; Lamont, I.; McQuillan, A. Langmuir 2003, 19, 3575-3577. (64) Tunesi, S.; Anderson, M. Langmuir 1992, 8, 487-495.

Organic Matter Adsorption at Mineral/Water Interfaces

Langmuir, Vol. 21, No. 11, 2005 5011

Figure 7. ATR-FTIR spectra of Pahokee peat humic acid adsorbed to boehmite (a) as a function of pH (ΓPPHA ) 2.0 µmol m-2) and (b) surface coverage (pH ) 5.0). Spectra have been normalized to the boehmite peak at 1157 cm-1.

spectral features from adsorbed carbonate (peaks at 1506 and 1395 cm-1) and water (peak at 1664 cm-1) are observed. As the concentration of PPHA is increased, however, spectral features from adsorbed PPHA gradually emerge (major peaks in the vicinity of 1745, 1712, 1568, and 1398 cm-1), and their similarity with the spectra of aqueous PPHA indicate a dominant outer-sphere mode of interaction with the boehmite surface (Figure 7b). Again, the adsorbed PPHA spectra tend to be most similar with those obtained 1-2 pH units higher in aqueous solution (particularly at low pH: see Figure 7b), indicating that the boehmite surface is promoting deprotonation of the adsorbed PPHA molecules. Interestingly, Figure 7b shows that, at low pH, no additional peaks are observed at ΓPPHA ) 2 µmol/m2, as were evident at a similar concentration and low pH for SRFA. The implications of the latter findings are, however, significantly less certain than those for SRFA, as humic acid molecules are substantially bulkier than those of fulvic acids. As a result, even at low ΓPPHA, only a small proportion of PPHA functional groups may be involved in interactions with the boehmite surface. Under such circumstances, minor inner-sphere interactions are unlikely to be detected using bulk ATR-FTIR spectroscopy. Impact of SRFA Adsorption on the Boehmite Dissolution Process. The impact of SRFA adsorption on mineral dissolution rate can provide indirect information on the adsorption modes of SRFA at the boehmite/ water interface. For example, inorganic ligands adsorbed in an inner-sphere, bidentate, binuclear manner tend to inhibit mineral dissolution,65-67 whereas ligands adsorbed as inner-sphere, bidentate, mononuclear complexes can (65) Stumm, W. Colloids Surf., A 1997, 120, 143-166. (66) Biber, M. V.; Afonso, M. D.; Stumm, W. Geochim. Cosmochim. Acta 1994, 58, 1999-2010. (67) Bondietti, G.; Sinniger, J.; Stumm, W. Colloids Surf., A 1993, 79, 157-167.

enhance dissolution.19,25,65,68,69 In addition, several recent studies undertaken in our laboratory have demonstrated that outer-spherically adsorbed ligands such as maleate28 and pyromellitate70 can significantly inhibit the dissolution kinetics of corundum under acidic conditions through steric protection of dissolution-active sites against attack by dissolution-enhancing species such as protons. A reduction in the protolytic dissolution rate results. The dissolution behavior of boehmite as a function of pH and various SFRA concentrations is shown in Figure 8. In the absence of SRFA, very low concentrations of dissolved Al(III) are observed at pH 5-7, consistent with a slow rate of dissolution typical of aluminum (oxyhydr)oxides71,72 and/or the low solubility of Al(III)73,74 previously reported in this intermediate pH range. The concentration of dissolved Al(III) increases markedly under acidic (pH < 5) conditions, which is consistent with an increased attack on the mineral surface by dissolution-enhancing H+ species. Figure 8 also shows that as the concentration of SRFA is increased, the concentration of dissolved Al(III) in solution decreases for pH < 4. These findings indicate that SRFA inhibits dissolution of boehmite under acidic conditions when it is strongly adsorbed to the boehmite surface (see Figure 3). The results shown in Figure 8 are consistent with the recent observation by Johnson et al.28,70 that outer-spherically adsorbed maleate anions inhibit corundum dissolution under acidic condi(68) Stumm, W.; Furrer, G. In Aquatic Surface Chemistry; Stumm, W., Ed.; Wiley-Interscience: New York, 1987; pp 197-217. (69) Stumm, W. Adv. Chem. Ser. 1995, 244, 1-32. (70) Johnson, S. B.; Yoon, T. H.; Brown, G. E., Jr. Langmuir 2005, 21, 2811-2821. (71) Carroll-Webb, S. A.; Walther, J. V. Geochim. Cosmochim. Acta 1988, 52, 2609-2623. (72) Wesolowsky, D.; Palmer, D. Geochim. Cosmochim. Acta 1994, 58, 2947-2969. (73) Bethke, C. M. The Geochemist’s Workbench; University of Illinois: Champaign, IL, 2002. (74) Baes, C. F.; Mesmer, R. E. The Hydrolysis of Cations; WileyInterscience: New York, 1976.

5012

Langmuir, Vol. 21, No. 11, 2005

Yoon et al.

indicated by the ATR-FTIR findings in the previous section. Again, the dominant mechanism underlying the inhibition of dissolution is likely to be one based on steric protection of surface sites, with the strong outer-sphere association of SRFA with boehmite surface sites under acidic conditions blocking access to those surface groups for dissolution-enhancing H+ ions. As a result, the protolytic dissolution rate of boehmite is substantially reduced. In addition, an aluminum trapping mechanism may also have contributed to the boehmite dissolution behavior observed in the presence of SRFA in Figure 8, whereby solution-facing SRFA functional groups complexed Al(III) cations as they were dissolved from the boehmite surface, thus holding them in the vicinity of the interface. A reduction in the apparent dissolution rate would result, as has recently been reported by Johnson et al.70 for the interaction of SRFA with corundum. Given the prevalence of high molecular weight NOM compounds in the environment, it is likely that they, along with other LMW organic anions with dominant outer-sphere mode adsorption, such as maleate, o-phthalate, and pyromellitate,28,70 will play a significant role in passivating metal (oxyhydr)oxides against dissolution, particularly under acidic pH conditions.

Figure 8. Concentration of Al(III) dissolved from the boehmite surface as a function of pH at various SRFA/boehmite ratios. Γ ) 0.00 (filled circles), 1.20 (hollow circles), 2.20 (hollow triangles), and 5.33 (filled triangles) (all in µmol m-2).

tions. We take the latter finding as further indirect evidence that SRFA is adsorbed in a predominantly outersphere manner at the boehmite/water interface, as

Acknowledgment. We gratefully acknowledge the financial support of NSF Grant CHE-0089215 (Stanford University CRAEMS on Chemical and Microbial Interactions at Environmental Interfaces), NSF Grant CHE0431425 (Stanford Environmental Molecular Science Institute), and EPA STAR Grant R827634. We also thank Prof. Scott Fendorf (Stanford University) for the use of a Nicolet Nexus 470 FTIR spectrometer. LA0476276