Effect of Randomly Methylated β-Cyclodextrin on Physical Properties

May 24, 2003 - The application of cyclodextrins in several soil remediation technologies has been increasingly studied, but little is known about thei...
7 downloads 11 Views 134KB Size
Download PDF
Environ. Sci. Technol. 2003, 37, 3012-3017

Effect of Randomly Methylated β-Cyclodextrin on Physical Properties of Soils G R Z E G O R Z J O Z E F A C I U K , * ,† ATTILA MURANYI,‡ AND EVA FENYVESI§ Institute of Agrophysics of the Polish Academy of Sciences, Doswiadczalna 4 Street, Lublin, Poland, Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences, H-1022 Budapest, Herman Otto ut15, Hungary, and Cyclodextrin Research and Development Laboratory Ltd., H-1097 Budapest, Illatos ut 7, Hungary

The application of cyclodextrins in several soil remediation technologies has been increasingly studied, but little is known about their effects on soil physical properties. One of the popular soil remediation additives, randomly methylated β-cyclodextrin (RAMEB), was found to significantly alter surface and pore properties of soil clay minerals. Therefore, in this paper we studied the effect of various RAMEB doses on physical properties of selected soils, representing a wide range of clay content (3-49%). The results showed that soil physical properties were greatly modified by RAMEB treatment. Analysis of water vapor adsorption isotherms revealed that RAMEB increased water adsorption and surface area in sandy soils and decreased them in clayey soils. An increase in adsorption energy of water in RAMEB-treated soils indicated that desorption of nonpolar pollutants can be enhanced. Water vapor desorption isotherms showed that the volumes and radii of micropores (nanometers range) increased above 1% RAMEB concentration. The micropores became more rough and complex after RAMEB treatment as deduced from an increase in fractal dimensions. The volume of soil mesopores measured by mercury intrusion porosimetry (micrometers range) gradually decreased in most soils with an increase in RAMEB concentration whereas the average mesopore radius increased, indicating that finer mesopores were blocked by RAMEB. Measurements of the granulometric composition of soils by sedimentation analysis showed that the amount of coarse-size soil fractions increased on the expense of finer fractions due to aggregation of smaller particles. Behavior of the studied soils after RAMEB treatment depended on their clay content and the dose of cyclodextrin. In clay-rich soils, strong interactions of cyclodextrins with the soil solid phase governed the resulting soil properties. In clay-poor soils, the cyclodextrin excess (not interacted with clays) played a dominant role. Modification of surface, pore, and aggregation properties of soils by RAMEB can have a significant effect in soil remediation technologies. * Corresponding author telephone: +48-81-7445061; fax: +4881-7445067; e-mail: [email protected]. † Institute of Agrophysics of the Polish Academy of Sciences. ‡ Research Institute for Soil Science and Agricultural Chemistry of the Hungarian Academy of Sciences. § Cyclodextrin Research and Development Laboratory Ltd. 3012

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 13, 2003

Introduction Since the possibility of using cyclodextrins (CDs) for soil remediation was first mentioned in 1992 (1), two main soil treatment technologies have been developed: washing of contaminated soils with a relatively concentrated CD solution (sugar flushing) (2) and using small amounts of CDs as a bioavailability-enhancing additive to accelerate the biodegradation of organic pollutants (3). These technologies are based on the solubilizing effect of CDs forming inclusion (host/guest) complexes with typical soil pollutants, like polycyclic aromatic hydrocarbons (PAHs) (4), chlorinated hydrocarbons (5), chlorinated biphenyls (PCBs) (6), and explosives (7). CDs, ring-shaped carbohydrates prepared by enzymatic conversion of amylose (8), are able to include into their cavity these xenobiotics or at least a part of their molecule. The hydrophobic character of the cavity provides a favorable microenvironment for the guest while the hydrophilic character of the rim and the outer surface enhances usually an apparent water solubility of the included compounds. This amphyphilic character of CDs provides surface activity that also plays an important role in solubilization (5), so these special carbohydrates unify the advantages of the hosts forming inclusion complexes and of the surfactants reducing the interfacial tension. In consequence, CDs reduce the adsorption of the organic contaminants on the soil surface (9) and improve their desorption (10) even in case of aged contamination (11, 12). The biodegradation-enhancing effect of CDs is explained partly by their carrier function, resulting in enhanced bioavailability due to the increased substrate concentration in the biofilm where the microflora lives and partly by the toxicity-decreasing effect (3). Two CD derivatives are most commonly used for soil remediation: hydroxypropyl and random methylated β-cyclodextrins (HPBCD and RAMEB, respectively), both being mixtures of isomers with different degree and pattern of substitution, extremely soluble in water (more than 50%) (13), and nonvolatile. Although RAMEB has a higher solubilizing effect, HPBCD is more feasible for the “sugar flushing” technology because of lower surface activity (5). Aqueous solution of methylated β-CD is not the best choice for soil washing because the mobility of the nonaqueous phase liquids (NAPLs) might be increased (2). The other unfavorable property of methylated β-CD may be its slight adsorption on clay minerals (illite) while HPBCD is not adsorbed (14). The high solubilizing effect of RAMEB is utilized in bioremediation techniques requiring only low amounts (0.1-1% w/w soil) of the additive (16, 17). In this case RAMEB acts as a catalyst improving the transfer of pollutants from the solid phase to the aqueous phase of the soil. β-CD was also reported to improve the biodegradation of a single hydrocarbon (dodecane) (15). γ-CD, HPBCD, and RAMEB were effective in the intensification of PCB biodegradation in soils (10, 16). Especially remarkable bioavailability-enhancing properties were exhibited by RAMEB in hydrocarbon-polluted soils (17). Because soil bioremediation needs months to years depending on type and concentration of the contaminants, soil properties, and microflora, an additive slowly biodegraded in the soil is required. RAMEB meets this requirement. Its half-life time is about 1 yr in a soil contaminated with motor oil (18) while HPBCD is decomposed rapidly (19). RAMEB treatment markedly modifies surface areas, adsorption energies, and porosities of typical soil clay minerals (20). In RAMEB-treated soils, changes in soil appearance were observed: the soil color darkened and the consistency altered, especially in soil surface layer (21), suggesting that surface, pore, and aggregate properties of soils may be altered as 10.1021/es026236f CCC: $25.00

 2003 American Chemical Society Published on Web 05/24/2003

TABLE 1. Basic Physicochemical Characteristics and Surface and Pore Parameters of Nontreated Soil Samples and RAMEB ( 95% Confidence Intervalsa soil pH % OM % sand % silt % clay S (m2 g-1) -Eav/RT v (mm3 g-1) rav (nm) D V (mm3 g-1) Rav (µm)

S1

S2

S3

S4

S5

S6

S7

5.1 0.5 87 10 3 8 ( 0.7 3.84 ( 0.12 6.9 ( 0.5 30.4 ( 2.7 2.41 ( 0.09 62 ( 5.4 2.63 ( 0.2

7.7 11.9* 79 13 8 13 ( 0.6 4.13 ( 0.09 10.2 ( 0.7 41.5 ( 1.8 2.47 ( 0.09 15 ( 2.2 1.67 ( 0.08

5.8 1.9 20 69 11 46 ( 2.4 4.12 ( 0.13 25.6 ( 1.2 13.8 ( 1.4 2.51 ( 0.06 208 ( 7.1 1.49 ( 0.09

7.1 3.2 11 73 16 74 ( 4.3 3.99 ( 0.14 37.1 ( 2.4 15.9 ( 1.0 2.46 ( 0.07 231 ( 9.5 1.30 ( 0.13

7.3 3.6 19 56 25 86 ( 4.6 3.17 ( 0.07 61.1 ( 3.1 27.0 ( 2.1 2.42 ( 0.06 192 ( 4.6 3.71 ( 0.08

5.6 1.3 23 41 36 156 ( 5.7 3.34 ( 0.09 90.8 ( 3.7 13.9 ( 0.7 2.38 ( 0.06 151 ( 5.1 0.59 ( 0.04

7.4 3.9 4 47 49 135 ( 5.1 2.92 ( 0.09 80.9 ( 4.2 19.2 ( 0.9 2.38 ( 0.05 152 ( 6.0 0.37 ( 0.04

RAMEB 7.0

428 ( 15.8 1.5 ( 0.05 630 ( 23 19.3 ( 0.71 1.93 ( 0.01 64 ( 1.8 0.74 ( 0.02

a Abbreviations: OM, organic matter (* oil pollution included); S, surface area; E /RT, average water vapor adsorption energy (in RT units); av V, volume of mesopores; v, volume of micropores; Rav, average mesopore radius; rav, average micropore radius; D, fractal dimension of micropores. Data for RAMEB are from ref 20.

well. Checking this hypothesis on soils representing a wide range of clay content was the aim of this paper. A practical importance of such studies is that the above soil properties control (among others) soil water retention, hydraulic conductivity, ventilation, solute transport, and sorption of pollutants, which can modify the efficiency of remediation processes.

Materials and Methods Materials and Pretreatments. Technical-grade randomly methylated β-cyclodextrin (RAMEB, CAWASOL M1.8, Wacker Chemie, Germany) having an average of 12.6 methyl groups per molecule and containing less than 0.1% β-cyclodextrins and 2% chlorides as impurities was used. Samples of upper horizons of four soils from Hungary (H) and three from Poland (Pl), varying in typology and mineral composition (Table 1), were studied. For the sake of convenience, these soils are numbered in order of increasing clay content: S1, Typic Haplohumod (Nyirlugos, H); S2, Typic Haplaquoll (Dunaujvaros, H); S3, Typic Eutrochrept (Tarnawatka, Pl); S4, Typic Rendoll (Machnow, Pl); S5, Typic Hapludoll (Buda, H); S6, Dystric Eutrochrept (Rudnik, Pl); and S7, Umbric Dystrochrept (Kisujszalas, H). The S2 sample was taken from a site severely polluted (2% w/w) with motor oil where our RAMEBbioremediated experimental plot is located. Aqueous RAMEB solutions of different concentrations were added to the soils (1:1 liquid:soil v/v ratio) and slowly air-dried for 1 week in a laboratory. At the end of the drying process, the samples were carefully mixed with a glass rod to minimize nonuniform RAMEB distribution. Dry samples were gently ground in a mortar and used for further experiments. The final contents of RAMEB in the samples were 0.0 (control), 0.03, 0.10, 0.30, 1.0, 3.0, and 9.0% w/w (dry mass), which were adjusted by varying the concentrations of the RAMEB solutions added. Surface Parameters. Water vapor adsorption isotherms were measured using a vacuum chamber method at a temperature (T) of 294 ( 0.1 K. The relative water vapor pressure (p/p0) in the chamber was controlled by sulfuric acid of stepwise decreasing concentrations (increase of p/p0). An amount of adsorbed water (kg/kg) at a given p/p0 was measured after 48 h of equilibration by weighing. Having completed the isotherm measurements, the dry mass of the samples was estimated by 24-h oven-drying at 378 K. In the beginning of the adsorption measurements, a decrease in the sample mass of oil-polluted soil (S2) occurred because of evaporation of volatile organic compounds under vacuum. Therefore, this sample prior to the adsorption measurements was evacuated for 2 weeks to reach a constant mass. This mass was taken as the dry mass of the S2 sample (without

oven-drying). The Aranovich adsorption model (22) was applied to find the surface area and energetic heterogeneity (23, 24) of the studied samples. The idea of the estimation of the surface area is to find a number of adsorbate molecules that cover the adsorbing surface as a monolayer and to multiply this number by the area occupied by a single molecule. Surface heterogeneity is estimated assuming that different surface sites bind adsorbate molecules with different forces (and energies), thus influencing adsorption pathways. Because the free energy is constant throughout the system in equilibrium, the energy of water vapor at a given relative pressure is associated with adsorption energy on a given site. The number of these sites is estimated from the amount adsorbed. An adsorption energy distribution function is constructed showing fractions of sites of distinct adsorption energies. From this function, average adsorption energy can be easily derived. Micropore Parameters. Water vapor desorption isotherms were measured using the vacuum chamber method as described above, but sulfuric acid of stepwise increasing concentrations (decrease of p/p0) was used. From desorption values, characteristics of pores ranging from ca. 1 to 175 nm (micropores) were evaluated (25). Assuming that during desorption the condensed water evaporates from micropores, then the micropore radius (r) is associated with p/p0 and the micropore volume (v) is associated with the amount of evaporated water. The first derivative of v versus r function gives pore size distribution, from which the average micropore radius is calculated. The degree of complexity of the irregular microporous surface may be described in terms of fractal geometry by a single number, called fractal dimension or D (26). The D value should fall between 2 (flat, twodimensional surface) and 3 (geometrically most complex, rough surface, approaching volumetric properties). The micropore fractal dimension (D) can be calculated from the slope of the linear part of the ln-ln plots of adsorption a versus adsorption potential A ) RT ln(p0/p) in the multilayer adsorption regime (i.e., for ln(A) equal or higher than 0) (27). Mesopore Parameters. Mercury intrusion porosimetry (MIP) (24) was used to characterize pores ranging from ca. 0.005 to 10 µm (mesopores). Samples of S3-S7 soils were studied as hand-made aggregates, pretreated with four wetting-drying cycles to stabilize the structure, and crushed onto a few millimeter pieces. Samples of S1 and S2 soils were studied as obtained after wetting-drying. These soils developed strong aggregates only at the maximum RAMEB dose. When the intrusion pressure is increased, the mercury is forced into the narrower pores of the sample. The intrusion pressure is thus associated with the mesopore radius (R), VOL. 37, NO. 13, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3013

FIGURE 1. Water vapor adsorption isotherms for RAMEB-treated soils. Soil symbols are as in Table 1. The number in parentheses following the soil symbol is the dose of RAMEB (%).

FIGURE 2. Mercury intrusion porosimetry curves for RAMEB-treated soils. Soil symbols are as in Table 1. The number in parentheses following the soil symbol is the dose of RAMEB (%).

and the intruded volume of mercury (V) gives the pore volume. The first derivative of V versus R function gives pore size distribution, from which the average micropore radius is calculated. Particle/Microaggregate Size. Samples of selected soils (S1, S5, and S7) were suspended either in water (control) or in a 0.1% RAMEB solution and dispersed in an end-over-end shaker for 4 h. The size distribution of the suspended particles/microaggregates was determined using the routine sedimentation procedure (corrected for presence of RAMEB if necessary). All measurements mentioned above were performed in three replicates. All details on experimental techniques, theoretical approach, and calculation schemes are given in ref 20.

Results and Discussion Effect of RAMEB on Water Vapor Adsorption on Soils. Experimental adsorption isotherms for the RAMEB-treated soils are presented exemplary in Figure 1. As pure RAMEB sorbs a very high amount of water (ca. 1 g g-1 at p/p0 ) 0.99), an increase in water sorption was expected after RAMEB addition to all soils. However, the isotherms for RAMEBtreated clay-rich S6 and S7 soils show lower adsorption than the original soils, which is illustrated for S7 soil having 49% of clay. This potentially indicates that RAMEB decreases the amount of water-available surfaces in clay-rich soils, similar to what was observed for pure clay minerals (20). In sandy soils (S1-S4), the water sorption markedly increases, particularly at higher RAMEB doses, as is illustrated for S2 soil. This may be attributed to the water sorption by the excess of RAMEB that does not interact with soil clay components. For soil S5 of medium clay content (25%), the effect of RAMEB on water sorption is small, which can reflect a balance between the two tendencies described above. Effect of RAMEB on MIP Patterns. Figure 2 shows examples of mesopore volume versus radius curves measured by mercury intrusion. In general, the MIP curves for RAMEBtreated soils are situated below those for the untreated soils, indicating that the addition of RAMEB decreases soil mesoporosity, due probably to pore filling with RAMEB and/or shrinking of aggregate structure. The opposite trend of the MIP curves for the S5 soil (illustrated) indicates aggregate structure loosening. This effect may be due either to the illitic composition of this soil (similar behavior of aggregates composed from pure illite was noticed in our previous studies (20)) or to specific composition of soil organic matter. This soil was taken from a forest and thus may contain not decomposed raw organic residues, forming more rigid spatial 3014

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 13, 2003

FIGURE 3. Changes in surface areas (S) of the studied soils under RAMEB treatment. On the y-axis, the ratio of the surface area of the soil + RAMEB to that of the control soil is shown. Soil symbols are as in Table 1. structures with RAMEB than more humified material present in other field soils. Effect of RAMEB on Soil Physical Parameters. Surface and pore parameters of the control soils, calculated from the isotherms and MIP curves, are presented in Table 1, while changes of these parameters for soils after various RAMEB treatments are summarized in Figures 3-9. Figure 10 shows RAMEB-induced changes in granulometric soil composition. In these figures, the ratio of a given parameter for the RAMEBtreated soil to its value for the control soil is presented on the y-axis, which allows for easier comparison of the behavior of different soils. Surface Area. Changes in surface area of the studied soils due to RAMEB treatment are illustrated in Figure 3. Because RAMEB itself has a very high surface area, one could expect an increase in surface area for all RAMEB-enriched soils. However, the surface area of the clay-rich (S6 and S7) soils decreases after RAMEB addition. This suggests the strong interaction of RAMEB and soil clay phase blocking water vapor adsorbing surfaces. A similar decrease in surface area was also observed for pure minerals (20). An increase in surface area for clay-poor soils with increasing RAMEB concentration is observed. With a lower clay content, the effect is more pronounced. This reflects a high contribution of the excess cyclodextrin (not interacting with soil clay components) on water sorption. Surface area of the S5 soil having intermediate (25%) clay content is the least sensitive to RAMEB.

FIGURE 4. Changes in average water vapor adsorption energy (Eav) of the studied soils under RAMEB treatment. On the y-axis, the ratio of the adsorption energy of the soil + RAMEB to that of the control soil is shown. Soil symbols are as in Table 1. Adsorption Energy. Water vapor sorption on pure RAMEB is a low-energy process because water molecules interact with cyclodextrins practically only via hydrogen bonds with energies close to water-water interactions. A high amount of low-energy sites and small amounts of more energetic sites is seen in sorption energy distribution function of RAMEB (20). Despite the fact that water vapor sorption energy on RAMEB is low, for all of the soils at low RAMEB doses we observed an increase in the relative amount of higher energy centers and a decrease of the lower energy centers, reflecting strong interactions of RAMEB and soil constituents. This tendency extended toward high RAMEB doses for clay-rich soils, while for clay-poor soils low-energy centers started to dominate above ca. 1% RAMEB dose. Changes in energy distribution patterns affect average adsorption energy values, which is illustrated in Figure 4. Instead of an adsorption energy decrease, as expected because of very low water vapor sorption energy of the RAMEB, the adsorption energy of all soils increases at low RAMEB concentrations. For clay-rich soils, this tendency holds in all RAMEB concentration ranges. For clay-medium S5 soil at 9% RAMEB, the adsorption energy starts to decrease. Similarly, for clay-poor soils, the initial increase in average adsorption energy is followed by its drop at higher RAMEB doses, which can be related to the overload of RAMEB. The adsorption energy increase indicates that the overall water binding forces become higher after cyclodextrin treatments, which may reduce the amount of water available for soil biota. However, the increase in interaction energy of polar water molecules with RAMEB-treated soils shows that the soil surface tends to be less hydrophobic, which may result in enhanced desorption of nonpolar contaminants and their inclusion into the cyclodextrin. Microporosity. Changes in micropore volumes of the studied soils versus RAMEB addition are illustrated in Figure 5. Although RAMEB micropore volume (630 mm3 g-1) is much higher than that of the soils, for clay-rich soils this hardly changes after RAMEB treatment. For clay-poor soils, the micropore volume increases at higher RAMEB doses. In general, small changes in micropore size distributions (data not presented) were noted for clay-rich soils at lower than maximum RAMEB doses, for which the fraction of large micropores (few tens of nanometers) increased on the expense of smaller micropores. In clay-poor soils, the fraction of large micropores started to increase markedly at 1% RAMEB load. However, small changes in micropore size distribution of oil-contaminated clay-poor soil (S2) were observed under all RAMEB concentrations applied, suggesting that until a significant portion of the added RAMEB reacts with the oil,

FIGURE 5. Changes of micropore volumes of the studied soils under RAMEB treatment. The y-axis shows the ratio of the micropore volume of the soil + RAMEB to this of the control soil. Soil symbols are as in Table 1.

FIGURE 6. Changes of average micropore radius of the studied soils under RAMEB treatment. On the y-axis, the ratio of the micropore radius of the soil + RAMEB to this of the control soil is shown. Soil symbols are as in Table 1. the micropore sizes may not alter markedly. Moreover, the S2 soil has very large micropores (see Table 1), which can be least sensitive on size changes. Small micropores in this soil are probably filled with the oil. Changes in average micropore radius of the studied soils due to RAMEB addition are illustrated in Figure 6. In general, an increase in the average micropore radius with increasing RAMEB load is seen. For soils having initial micropore radius lower than that of the RAMEB, one may explain the micropore radius increase as a simple effect of RAMEB addition. However, if the micropore radius of the initial soil is larger, the increase of micropore radius may be connected with the filling of smaller pores by RAMEB and/or by gluing their walls together. As a direct consequence of the smallest changes in micropore size distributions, changes in average micropore radius for oil-polluted S2 soil are the lowest. The microporous structure exhibits fractal scaling for all of the samples as evidenced by the linearity of ln-ln plots of a versus A (data not presented). The upper limit of the linearity interval is different for different soils (ln(A) ) 0.7 for S5 and S7, 1.3 for S6, 1.4 for S3, 1.6 for S2 and S4, and 1.8 for S1); however, the lower limit is the same (ln(A) ) -1). Thus, the fractal scaling intervals extend below the traditional multilayer regime (ln(A) ) 0). Value of ln(A) ) -1 corresponds to pore radius of ca. 1 nm. Such small pores can be referred as surface roughness. The fractal scaling intervals do not change significantly after RAMEB addition. The micropore VOL. 37, NO. 13, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3015

FIGURE 7. Changes of micropore fractal dimension of the studied soils under RAMEB treatment. On the y-axis, the ratio of fractal dimension of the soil + RAMEB to this of the control soil is shown. Soil symbols are as in Table 1.

FIGURE 9. Changes of average mesopore radius of the studied soils under RAMEB treatment. On the y-axis, the ratio of the mesopore radius of the soil + RAMEB to this of the control soil is shown. Soil symbols are as in Table 1.

FIGURE 8. Changes of the mesopore volume of the studied soils vs RAMEB content. On the y-axis, the ratio of the mesopore volume of the soil + RAMEB to this of the control soil is shown. Soil symbols are as in Table 1.

FIGURE 10. Changes in soil particle/microaggregate sizes after RAMEB addition. Average sizes of the particles are on the x-axis. The y-axis shows ratio of the w/w fraction of particles of a given size for soils dispersed in 0.1% RAMEB solution to that of the soils dispersed in water. Soil symbols are as in Table 1.

fractal dimension of the studied soils (Figure 7) drops slightly at minimum RAMEB dose and increases next. For sandy soils this decreases again at high RAMEB contents. This trend may be explained as follows: initially RAMEB covers the most complex parts of original soil surfaces, next a rearrangement of the micropore structure or grouping of the RAMEB as “surface islands” increases surface complexity, and finally (as observed for sandy soils) the excess of RAMEB tend to exhibit its own properties. Mesoporosity. The mesopore volume of the RAMEB is lower than this in all but S1 and S2 soils. The oil-polluted S2 soil has very low mesopore volume (15 mm3 g-1), even lower than the most sandy soil S1, suggesting that oil can fill the mesopores. The mesopore volume of all but S5 soils decreases after RAMEB addition, indicating that the aggregate structure of these soils becomes generally more compacted and/or that the mesopores are filled with RAMEB (Figure 8). The structure compaction and closing mesopores after RAMEB treatments may alter soil water capacity and ventilation. For S5 soil, an increase in mesopore volumes is observed, which may be connected with a rearrangement of soil particles within the aggregate network. A similar increase in mesopore volumes was observed for aggregates of pure illite (20). Mesopore size distribution functions (data not presented) have a single-peak shape for all but S2 and S6 soils. The maximum of this peak (i.e., largest fraction of pores) occurs at the pore radius R ) 10 µm for S1, 0.2 µm for S3 and S4, 3016

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 13, 2003

8 µm for S5, and 1.6 µm for S7. Two broad peaks are found for S2 (at 5 and 0.01 µm), and two narrow peaks are found for S6 (at 1 and 0.008 µm). In general, RAMEB addition did not markedly change mesopore size distributions for all but the two most sandy soils, S1 and S2. For S1 soil, the narrow peak at 10 µm markedly broadened at 9% RAMEB addition, thus resembling the pore size distribution function of the RAMEB itself; probably all mesopores in this soil were filled with 9% RAMEB. In the oil-contaminated S2 soil, fine (0.1 µm) mesopores decreased sharply with increasing RAMEB concentration and vanished completely at 1% RAMEB addition. Similar decrease in fine (0.008 µm) mesopores occurred also for S6; however, part of these pores remained at all RAMEB loads. Average mesopore radii for most of the soils tend to increase with RAMEB dose (Figure 9), which can reflect the gluing of finer mesopore walls by the cyclodextrin. For the two most sandy soils (S1 and S2), the mesopore radius sharply decreases at 9% RAMEB: most mesopores are filled up, and properties of the RAMEB start to dominate. The increase in mesopore radius due to the RAMEB addition is the largest in oil-contaminated S2 soil, which may be connected with rearrangement of organic matter architecture (this soil is very rich in organic material) and partially with entering RAMEB cavities by oil, thus emptying the pores and cleaning their walls from oil layers.

Particle/Aggregate Size. Addition of RAMEB to water suspensions of the selected soils affected sizes of the suspended particulates, which is presented in Figure 10. RAMEB increases the amount of coarser soil particulates on the expense of finer fractions because of aggregation of smaller particles. It is necessary to mention that no aggregation of the clay minerals illite, bentonite, and kaolin was observed in up to 2% RAMEB solutions in a parallel experiment. This indicates that, at the excess of water, RAMEB either tends to remain in the solution or is partially adsorbed on single mineral particles. Binding together of soil organic matter and/or of organomineral particulates by the cyclodextrin most probably causes the aggregation in soils: entering of hydrophobic radicals of individual organic particles into the RAMEB cavity increases the mass and size of the resulting formation and so its sedimentation ability. Our results demonstrated that RAMEB strongly interacts with soils, modifying their surface, pore, and aggregate properties. These effects can affect soil remediation technologies.

Acknowledgments Part of this work was granted by Polish Research Committee KBN under Project 6P04G05613 and by NATO Science for Peace Program SFP-973720.

Literature Cited (1) Szejtli, J. In Minutes, 6th International Symposium on Cyclodextrins; Hedges, A. R., Ed.; Sante: Paris, 1992; pp 380-389. (2) Boving, T. B.; Wang, X.; Brusseau, M. L. Environ. Sci. Technol. 1999, 33, 764-770. (3) Gruiz, K.; Fenyvesi, E.; Kriston, E.; Molnar, M.; Horvath, B. J. Inclusion Phenom. Mol. Recognit. Chem. 1996, 25, 233-236. (4) Wang, X.; Brusseau, M. L. Environ. Sci. Technol. 1993, 27, 28212825. (5) McCray, J. E.; Boving, T. B.; Brusseau, M. L. Ground Water Monit. Rem. 2000, 20, 94-103. (6) Fava, F.; Bertin, L.; Fedi, S.; Zannoni, D. Biotechnol. Bioeng. 2003, 81, 381-390. (7) Hawari, J.; Paquet, L.; Zhou, E.; Halasz, A.; Zilber, B. Chemosphere 1996, 32, 1929-1936.

(8) Schmid, G. Compr. Supramol. Chem. 1996, 3, 41-56. (9) Brusseau, M. L.; Wang, X.; Hu, Q. Environ. Sci. Technol. 1994, 28, 952-956. (10) Fenyvesi, E.; Szeman, J.; Szejtli, J. J. Inclusion Phenom. Mol. Recognit. Chem. 1996, 25, 229-232. (11) Fava, F.; Di Gioia, D.; Marchetti, L. Biotechnol. Bioeng. 1998, 58, 345-355. (12) McCray, J. E.; Brusseau, M. L. Environ. Sci. Technol. 1998, 32, 1285-1293. (13) Szente, L.; Szejtli, J. Adv. Drug Delivery Rev. 1999, 36 (1), 17-28. (14) Sheremata, T. W.; Hawari, J. Environ. Sci. Technol. 2000, 34, 3462-3468. (15) Steffan, S.; Bardi, L.; Marzona, M. Special Issue: Cyclodextrins. Biol. J. Armenia 2001, 53, 218-225. (16) Fava, F.; Ciccotosto, F. V. Appl. Microbiol. Biotechnol. 2002, 58, 393-399. (17) Molnar, M.; Fenyvesi, EÄ .; Gruiz, K.; Leitgib, L.; Balogh, G.; Muranyi, A.; Szejtli, J. J. Inclusion Phenom. Mol. Recognit. Chem. 2002, 44, 447-452. (18) Fenyvesi, E.; Csabai, K.; Molnar, M.; Gruiz, K.; Muranyi, A.; Szejtli, J. J. Inclusion Phenom. Mol. Recognit. Chem. 2002, 44, 413-416. (19) Verstichel, S.; De Wilde, B.; Fenyvesi, E.; Szejtli, J. J. Polym. Environ. (in press). (20) Jozefaciuk, G.; Muranyi, A.; Fenyvesi, E. J Environ. Sci. Technol. 2001, 35, 4947-4952. (21) Leitgib, L.; Gruiz, K.; Molnar, M.; Fenyvesi, E. Proceedings of 8th International FZK/TNO Conference on Contaminated Soil; 2003. (22) Aranovich, G. L. Langmuir 1992, 8, 736-739. (23) Jaroniec, M.; Rudzinski, W.; Sokolowski, S.; Smarzewski, R. J. Colloid Polymer Sci. 1975, 253, 164-166. (24) Jaroniec, M.; Brauer, P. Surf. Sci. Rep. 1986, 6, 65-117. (25) Roquerol, R.; Avnir, D.; Fairbridge, C. W.; Everett, D. H.; Haynes, J. H.; Pernicone, N.; Ramsay, J. D. F.; Sing, K. S. W.; Unger, K. K. Pure Appl. Chem. 1994, 66, 1739-1758. (26) Mandelbrot, B. B. The Fractal Geometry of Nature; Freeman: San Francisco, 1983. (27) Neimark, A. V. Russ. J. Phys. Chem. 1990, 64, 2593-2605.

Received for review October 13, 2002. Revised manuscript received April 30, 2003. Accepted May 3, 2003. ES026236F

VOL. 37, NO. 13, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3017