Characterization of Charcoal Adsorption Sites for Aromatic

Charcoal and carbon storage in forest soils of the Rocky Mountain West. Thomas H DeLuca , Gregory H Aplet. Frontiers in Ecology and the Environment 20...
1 downloads 0 Views 293KB Size
Environ. Sci. Technol. 2005, 39, 1606-1615

Characterization of Charcoal Adsorption Sites for Aromatic Compounds: Insights Drawn from Single-Solute and Bi-Solute Competitive Experiments MICHAEL SANDER† AND J O S E P H J . P I G N A T E L L O * ,†,‡ Department of Chemical Engineering, Environmental Engineering Program, Yale University, New Haven, Connecticut 06511, and Department of Soil and Water, The Connecticut Agricultural Experiment Station, New Haven, Connecticut 06511

Charcoal, the residue of incomplete biomass burning that is found in many soils and sediments, is considered a high affinity sorbent for organic pollutants. However, little is known about the microscopic processes controlling sorption. The purpose of this study was to gain molecularscale insight into the sorption on a charcoal of three weakly soluble aromatic compounds [benzene (BEN), toluene (TOL), and nitrobenzene (NBZ)] by conducting both singlesolute and bi-solute experiments. The charcoal (420 m2 g-1) was produced from maple wood shavings by oxygenlimited pyrolysis at 673 K. Solute affinity for charcoal followed the order NBZ > TOL > BEN. Commonly employed sorption models did not adequately describe the singlesolute isotherms. Competition in both TOL-BEN and the TOLNBZ bi-solute systems was strong. Normalization of the isotherms for the hydrophobic driving force by using an existing free energy correlation between sorption and partitioning to an inert solvent (benzene or n-hexadecane) with a nonpolar aromatic compound calibration set resulted in a finding of enhanced sorption of NBZ relative to the coalesced BEN and TOL isotherms, indicating some specificity in the interaction of NBZ. The competitive data indicated 1:1 molar competition between BEN and TOL and between NBZ and TOL, showing conclusively that this specificity was not due to a subpopulation of sorption sites unique to NBZ. H-bonding was ruled out, as the relative affinity for the sorbent among the solutes did not change at all when increasing the solution pH from 6.5 to 11. 1H NMR experiments showed molecular complexation in chloroform between NBZ and model graphene polycyclic aromatic units (naphthalene, phenanthrene, and pyrene) which was absent for BEN and TOL. This result, in combination with the results of a companion study (Zhu and Pignatello, Environ. Sci. Technol. (in press)), is used to support the existence of π-π electron donor-acceptor interactions between NBZ (electron acceptor) and the polycyclic aromatic charcoal surface (electron donor) as the cause of enhanced NBZ sorption. * Corresponding author telephone: (203)974-8518; fax: (203)9748502; e-mail: [email protected]. † Yale University. ‡ The Connecticut Agricultural Experiment Station. 1606

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005

Introduction Black carbon (BC) refers to the residues of incomplete combustion of biomass and fuel and includes soot and charcoal or char (1-3). Soots are formed in the gas phase, whereas charcoals result from incomplete solid-phase burning. BCs are very stable in the environment (4) and are found in atmospheric aerosols (5), ice, off-shore and deep-sea sediments (6-10), and soils (11). The interest in BCs is interdisciplinary, as these carbonaceous materials are believed to play a crucial role in atmospheric chemistry, the global carbon cycle (12), and the environmental fate of organic contaminants. This study focuses on wood charcoal as an adsorbent. Charcoals consist of stacked, condensed, and highly disordered polycyclic aromatic sheets (13, 14). These sheets vary in size (i.e., between a few to several tens of fused rings), the degree of functionality, and the extent to which pentagonal (i.e., fullerene-like) and heptagonal rings are dispersed throughout the hexagonal carbon framework (14). Despite that little is known about the pore structure of charcoal, it has often been compared to that of activated carbon (AC). Although structural similarities between these two carbonaceous materials may exist, a generalization seems unsubstantiated. AC is formed under special atmospheres and at higher temperature than charcoal. Also, AC typically has greater specific surface area than charcoal [≈900-1100 m2 g-1 (15) vs 200-500 m2 g-1]. The sorption of organic contaminants to chars and soots has been demonstrated to be strong and nonlinear (7, 1621). Although consensus has been established that adsorption as opposed to absorption is the dominant process, little is yet know about the exact molecular-scale mechanisms of the adsorption process. Previous studies primarily focused on single-solute sorption. Mechanistic insights are often drawn based on indirect evidence such as the fit of a particular sorption model. Information on competitive effects during multi-solute sorption, however, is scant. Bi-solute experiments may provide additional information on the nature of the sorption process, such as the fraction of sorption sites being shared between the two solutes tested, and their relative affinities toward these sites. Furthermore, bi-solute experiments are of practical relevance from a fate and risk assessment standpoint, as single-solute sorption likely is the exception rather than the rule for environmental systems. In the present study, we investigate the sorption of three weakly soluble aromatic compounds (BEN, TOL, and NBZ) on a maple wood charcoal. In contrast to BEN and TOL, NBZ is fairly polar and can undergo H-bonding (as H-acceptor only). This charcoal likely represents a subset of charcoals found in the environment. The researchable questions are two-fold: (i) Are all three solutes sorbing to identical sites in the charcoal? If not, what is the fraction of sites they share? (ii) What are the relative affinities of these compounds for the charcoal? Can the sorption of all three compounds be explained by hydrophobic driving forces alone, or are specific interactions involved? To answer these questions, we conducted single-solute and bi-solute aqueous sorption experiments.

Experimental Section Materials. The charcoal was produced by atmospheric pyrolysis of maple wood shavings at 400 °C for 2 h as previously described (16). The charcoal of eight batches (24.9% ( 0.8% yield by mass) was combined, crushed gently in a mortar, passed through a sieve no. 100 (0.150 mm) and 10.1021/es049135l CCC: $30.25

 2005 American Chemical Society Published on Web 01/25/2005

TABLE 1. Selected Properties of Benzene (BEN), Toluene (TOL), and Nitrobenzene (NBZ)a water mol wt solubility, SW log KOW log KHW log KBW sorbate [g mol-1] [mmol L-1]b (T ) 25 °C)b (T ) 25 °C)c (T ) 20 °C) BEN TOL NBZ

78.1 92.1 123.1

22.9 5.6 17.0

2.13 2.69 1.83

2.15 2.72 1.50

could be compared without having to set up an additional BEN-NBZ bi-solute system. Sorption Models. Among the sorption models most frequently used to describe the sorption of organics to carbonaceous sorbents are the Freundlich model (FM, eq 1) and the Langmuir model (LM, eq 2):

2.70 3.25 2.65

q ) KFCnW

a KOW ) 1-octanol-water partition coefficient, KHW ) n-hexadecanewater partition coefficient, and KBW ) benzene-water partition coefficient, all in [L L-1]. b Ref 22. c Ref 23.

stored sealed. The specific surface area of the charcoal was determined to be 420 m2 g-1 by nitrogen adsorption at 77 K. This value is consistent with the value of 400 m2 g-1 of a charcoal produced under identical experimental conditions from identical maple wood shavings (14). The latter charcoal was reported to be ultra-microporous: pores with diameters smaller than 10 Å accounted for 50%, and pores with diameters smaller than 20 Å accounted for 80% of total porosity (16). Given the identical formation procedure, we anticipate similar pore size distribution for the charcoal in the present study. BEN, TOL, and NBZ (all HPLC grade, 99.9+ %) were purchased from Sigma Aldrich (St. Louis, MO). Selected properties are given in Table 1. Naphthalene (NAPH), phenanthrene (PHEN), pyrene (PYR), chloroform-d containing tetramethylsilane (TMS), and NaN3 (SigmaUltra grade) were purchased from Sigma Aldrich (St. Louis, MO). Benzene-water partition coefficients of TOL and NBZ were obtained by a procedure described elsewhere (24). Sorption Experiments. A known mass of charcoal (25 mg) was added to each of several 24.5 mL glass screw-cap vials. Each vial was then filled to capacity with water containing 0.01 M CaCl2 and 200 mg (L)-1 NaN3 to inhibit biodegradation by incidental bacteria. The vials were stoppered with a PTFE-lined cap and shaken for 4 days to allow the charcoal to hydrate. Solute was then added in a methanol carrier, while keeping the fraction of methanol below 0.1 vol %. Following solute addition, the glass tubes were shaken for 35 days in the dark at 20 ( 1 °C to reach apparent sorption equilibrium (16). [The qualifier “apparent” is used to take into account the possibility that the system resides in a metastable state.] After equilibration, BEN, TOL, and NBZ in the liquid phase were extracted by hexanes containing 1,3dibromopropane as internal standard and subsequently analyzed by GC-FID. Calibration curves were constructed by running duplicate control samples in the absence of charcoal. “Bottle losses” over the sorption period were determined in sorbent-free controls using external standards and were found to be small ( 10-4 M, the differences in affinities are less pronounced and follow the order BEN < NBZ < TOL. However, based on the common assumption that charcoal is a “highly hydrophobic material” and judging from molecular properties in Table 1, such as solubility and solventwater partition coefficients (octanol-water, n-hexadecanewater, and benzene-water), which are common indices of hydrophobic character, “intuition” might predict that NBZ would have the least affinity for the sorbent. Figure 2 demonstrates strong competition between the solutes in both the BEN-TOL and the NBZ-TOL bi-solute systems. It is clear from this figure that the order of competitive ability at a given solution-phase concentration is BEN < TOL < NBZ; this is not surprising, as it is the same order as the single-solute affinity shown in Figure 1. As the initial competitor concentration CW_0(Competitor) increases, the target solute isotherm (i) shifts downward toward lower q, (ii) becomes less curved on the log scale, and (iii) tends toward linearity (slope of 1). It will later be shown that the hydrophobic-normalized isotherm describing the sum of BEN and TOL sorption in bi-solute systems is characterized by the same isotherm as the respective single solutes and remains highly nonlinear. It follows that effects (ii) and (iii) are merely the apparent result of looking at one solute at a time and that the trend toward linearity as competitor concentration increases does not signify a change in the mechanism of sorption of the principal solutesfor example, from adsorption to absorption.

FIGURE 2. Competitive sorption isotherms of the benzene-toluene bi-solute system (top pair) and the nitrobenzene-toluene bi-solute system (bottom pair). CW_0(X) corresponds to the time zero molar concentration of competitor X, where BEN ) benzene, TOL ) toluene, and NBZ ) nitrobenzene. 1608

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005

FIGURE 3. Model fits to single-solute sorption data. (a) Freundlich model FM, (b) Langmuir model LM, (c) dual-mode model DMM, (d) dual-Langmuir model DLM, and (e and f) Polanyi-Manes model PMM. Benzene data are represented by solid circles and solid regression curves, toluene data are represented by empty circles and dashed regression curves, and nitrobenzene data are represented by solid triangles and dash-dotted regression curves. Isotherm Modeling. The fits of the FM, LM, DMM, DLM, and PMM to the single-solute data of BEN, TOL, and NBZ are shown in Figure 3. Respective fitting parameters are given in Table 2. Figure 3 clearly shows that the FM (Figure 3a), LM (Figure 3b), and DMM (Figure 3c) fail to describe well the log-scale curvature of the isotherms. Inadequate fits are reflected in the high standard error of estimates SEE (Table 2). Furthermore, for all three test compounds, the DMM and especially the LM greatly underestimate q at CW higher than 10-4 mol L-1. The model results indicate that the sorption process of BEN, TOL, and NBZ on charcoal (i) is not one of monolayer formation on a homogeneous surface (i.e., LM); (ii) cannot be described by assuming a continuous energy distribution of sorption sites (i.e., the FM); and (iii) is not adequately described by allowing for a partition domain in addition to a Langmuir-type adsorption domain (i.e., DMM). Better fits are obtained by the DLM (Figure 3d, Table 2). However, good DLM model fits might partially result from over-parameterization, as the DLM is the only tested model with four fitting parameters. Furthermore, Figure 3d shows that the slope of the model fits for BEN, and particularly NBZ, starts to flatten out at concentrations above 10-3 mol L-1, indicating that the capacity of the low-affinity sorption site is approached. This contradicts previous data on sorption

of BEN to a different batch of charcoal prepared under identical conditions (16), which shows an isotherm that is convex to the CW axis at elevated CW. Given that the BEN sorption data presented here are in very good agreement with the previous data (16), it is reasonable to assume that a similar trend would have been observed for our isotherms had they been extended to higher concentrations. Further doubt on the validity of the DLM arises from the fitted molar site capacities of BEN and TOL. Fitted capacities for TOL are larger than fitted capacities for BEN by a factor of 2 (S10) and 2.6 (S20), although the molecular-free surfaces of approximately 108 and 128 Å2 for BEN and TOL, respectively, are very similar (39). Under the assumption of similar sorption domains for BEN and TOL (we will later see that this assumption is correct), the results are contrary to the expectation that S0 values are the same for BEN and TOL or perhaps slightly lower for TOL. Despite the PMM having lowest SEE values among all tested models, it is questionable as to whether the PMM mechanistically captures the single-solute sorption. According to PMM theory, the sorption data of chemically similar compounds (e.g., here BEN and TOL) are expected to fall onto a single curve (the so-called “correlation curve”) when plotted as sorbed volume qV [cm3 kg-1] versus the ratio of the VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1609

TABLE 2. Results of Model Fits to Single-Solute Sorption Isotherms of Benzene, Toluene, and Nitrobenzenea Freundlich Model (FM) KF [mol (L)N (molN kg)-1]

compound

22.64 ( 1.751 59.02 ( 5.023 16.95 ( 1.033

benzene toluene nitrobenzene

S0LM [mol kg-1]

benzene toluene nitrobenzene

0.3471 ( 1.845e-2 0.3990 ( 2.165e-2 0.3376 ( 1.598e-2

KD [L mol-1]

benzene toluene nitrobenzene

2.205e+2 ( 1.540e+1 9.780e+2 ( 6.001e+1 2.424e+2 ( 1.499e+1

benzene toluene nitrobenzene

7.205e-2 ( 1.064e-2 1.439e-1 ( 1.554e-2 1.410e-1 ( 1.175e-2

a

0.5433 ( 0.00723 0.5457 ( 0.00731 0.4302 ( 0.00495

0.2205 0.1808 0.2882

bLM [L mol-1]

SEE

2.268e+4 ( 1.948e+3 9.832e+4 ( 8.359e+3 3.877e+5 ( 3.037e+4

0.4070 0.4493 0.5041

Dual Model Model (DMM) S0DMM [mol kg-1]

compound

S01_DLM [mol kg-1]

SEE

Langmuir Model (LM)

compound

compound

N

1.664e-1 ( 1.158e-2 2.088e-1( 1.550e-2 2.396e-1 ( 1.350e-2

bDMM [L mol-1]

SEE

6.683e+4 ( 6.848e+3 2.384e+5 ( 2.534e+4 6.585e+5 ( 5.704e+4

0.2357 0.1791 0.3224

Dual-Langmuir Model (DLM) b1_DLM [L mol-1] S02_DLM [mol kg-1] 1.777e+5 ( 3.304e+4 3.724e+5 ( 5.303e+4 1.354e+6 ( 1.581e+5

compound

log QV0 [cm3 kg-1]

benzene toluene nitrobenzene

2.027 ( 4.099e-3 2.399 ( 6.960e-3 2.199 ( 4.415e-3

9.562e-1 ( 8.944e-2 2.639e+0 ( 4.192e-1 1.310e+0 ( 1.032e-1

Polanyi-Manes Model (PMM) aPMM [(cm3)b+1 (kg Jb)-1] -3.553e-4 ( 1.475e-5 -2.978e-3 ( 1.599e-4 -2.764e-4 ( 1.598e-5

b2_DLM [L mol-1]

SEE

1.220e+3 ( 2.558e+2 8.666e+2 ( 2.335e+2 1.911e+3 ( 3.987e+2

0.0409 0.0993 0.1708

bPMM

SEE

1.539 ( 7.090e-3 1.216 ( 9.475e-3 1.567 ( 9.948e-3

0.0417 0.0678 0.0692

Weighted nonlinear regression. SEE: standard error of estimates.

adsorption potential SW to molar volume VS (16). However, the curves of all three tested sorbates substantially deviate from one another (Figure 3e). This deviation likely results from the ultra-microporosity of the charcoal. Given its pore dimensions, it is neither likely that solute liquification occurs to a significant degree nor that potentially formed condensates have bulk liquid properties. Consequently, molar volume does not seem to be an appropriate abscissa scaling factor in this case. Therefore, the traditional approach to explain competitive sorption data (i.e., by bi-solute modeling based on singlesolute model parameters) is not viable for the present system. In the following, we will present a novel technique to analyze the competitive data which allows answering the questions posed in the Introduction. Normalization for Hydrophobic Effects. The approach used here, which is outlined in detail elsewhere (24), is to hyd partition the free energy of adsorption into a term (∆Gch-W,i ) for “hydrophobic effects”, which is understood to be the sum of processes that drive organic solutes out of solution onto the “inert” (dispersion forces only) surface, and a term e,elec (Gch,i ) for excess “electrostatic” interactions of adsorbate with the surface relative to the reference state, the pure liquid i: hyd e,elec ∆Gch-W,i(q) ) ∆Gch-W,i (q) + Gch,i (q)

(6)

hyd The term ∆Gch-W,i (q) represents cavity effects, the difference in dispersion (London and Debye) interactions between solution and sorbed states, dipole-dipole (Keesom) interactions with water excess to the reference state, and H-bonding interactions with water excess to the reference state. Electrostatic effects include dipole-dipole interactions, Hbonding, and π-π electron donor-acceptor (EDA) interactions with the surface, all excess to the reference state. ∆Gch-W,i

1610

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005

is concentration-dependent because the isotherm is nonlinear. The hydrophobic driving force may be related to the process of partitioning into an inert organic solvent from water. An equation similar to eq 6 may be written for partitioning into this solvent. With the choice of an aphyd is related to the molar free energy propriate solvent, ∆Gch-W,i of solvent-water partitioning, ∆GSW,i [J mol-1] by: hyd (q) ) aS(q)∆GSW,i + b′S(q) ∆Gch-W,i

(7)

where aS is factor of proportionality and b′S [J mol-1] is the offset due to comparing free energies of different processes. The value of aS is not necessarily equal to 1 because cavity effects and dispersion interactions each do not necessarily vary systematically with i identically for adsorption and partitioning; adsorption, in contrast to partitioning, may involve displacement of adsorbed water molecules whose interactions contribute to the change in free energy. Combining eqs 6 and 7 and substituting equilibrium constant expressions yields:

ln Kch-W,i(q) ) aS(q) ln KSW,i -

e,elec e,elec (q) - aS(q)GS,i Gch,i + RT bS(q) (8)

where Kch-W,i is the charcoal-water distribution ratio [L kg-1]; KSW,i is the solvent-water partition coefficient [L L-1]; R is the universal gas constant [J (mol K)-1]; T is the absolute e,elec temperature [K]; and GS,i is the excess electrostatic component of free energy of dissolution of i in the reference solvent. The constant bS includes b′S and other terms that depend on the adsorption reference state. In the following, we therefore present the sorption data as q [mol kg-1] versus the concentration C′S [mol L-1] that is

FIGURE 4. Single-solute sorption isotherms of benzene, toluene, and nitrobenzene normalized for hydrophobic effects. (a) n-Hexadecane as reference solvent and (b) benzene as reference solvent. For notation, see eq 9. normalized for the hydrophobic driving force according to:

C′S ) CWKSW,iaS(q)ebS(q)

(9)

Two reference partition solvents S, n-hexadecane (SdH) and benzene (SdB), were considered. Hexadecane allows only nonspecific, dispersion interactions between i and solvent. Benzene was chosen because it presents a face of π electrons that may more closely represent the graphene surface in charcoal. Values of KHW,i and KBW,i are given in Table 1. Due to the difficulty of applying a universal thermodynamic sorption model to the single-solute isotherms, for simplicity, aS and bS are assumed to be concentration independent for sorption to charcoal (see ref 24). The data were fit to the plane ln q ) ln CW + aS ln KSW,i + bS. Note that the value of bS does not affect the positions of the isotherms relative to one another. Regression of BEN, TOL, 1,2-dichlorobenzene, and 1,4-xylene sorption data on the same charcoal (1) yielded for hexadecane: aH ) 1.00 ( 0.08 and bH ) 1.1 ( 0.2; for benzene: aB ) 1.00 ( 0.08 and bB ) 1.2 ( 0.2. It was gratifying to find that substitution of BEN and TOL isotherms constructed independently by one of us (M.S.) for the corresponding ones of the companion study resulted in fitted aS and bS values that were identical within experimental uncertainty. The finding that for both inert solvents, aS values are indistinguishable from unity implies that the contribution of hydrophobic driving forces to the molar free energy of sorption of BEN and TOL is fully captured by the compound’s partition coefficient into the inert solvent. This was not the case for adsorption on graphite, a model for the charcoal surface, where aS substantially exceeded 1 (24). Note that the above approach is similar to a recent one in which the dilute solution in n-hexadecane was proposed as a universal reference state for all solutes (40-42). The approaches would be identical if the proportionality factor aS is assumed equal to 1. The single-solute sorption isotherms of BEN, TOL, and NBZ are replotted using normalized solute concentrations according to eq 9 in Figure 4, panels a and b, for reference solvents n-hexadecane and benzene, respectively. The graphs show the following. First, the isotherms of BEN and TOL overlap. This is consistent with the sorption of these two compounds occurring to identical sites by the same mechanism. Second, the NBZ isotherm is shifted toward higher q as compared to the “coalesced” BEN-TOL isotherm, indicating additional forces governing sorption of NBZ besides the hydrophobic effect. Third, the deviation between the NBZ and the BEN-TOL isotherm is more pronounced in Figure 4a as compared to Figure 4b. This suggests that part of the additional driving force for NBZ sorption to the charcoal as compared to BEN and TOL is canceled by

FIGURE 5. Sorption data plotted as the sum of sorbed concentrations q vs the sum of normalized concentrations for the reference solvent n-hexadecane CH′(X) ) CW(X)KHW(X)aHebH, where X ) BEN and X ) TOL for solutes benzene and toluene. The benzene and toluene single-solute isotherms are represented by solid and open circles, respectively. All other symbols represent bi-solute data, where CW_0(TOL) corresponds to the time zero toluene concentrations. Isotherms also overlapped when using benzene as the reference solvent (not shown). intermolecular interactions between NBZ and benzene solvent. Similar results have been found for a series of nitrotoluenes (24). In the past it has been popular to normalize hydrophobic effects by plotting isotherms as sorbed concentration versus the “reduced” solution concentration, which is the experimental value divided by the compound’s solubility. Applying this approach also results in coalescence of BEN and TOL isotherms and leaves NBZ isotherm displaced above the coalescence line (data not shown). However, this method has been criticized as misleading because the reference state for each solute (i.e., the pure liquid or subcooled liquid state) is unique (42, 43). While the reference state for BEN and TOL may be chemically similar, the reference state for NBZ includes dipole-dipole interactions among NBZ molecules. Evaluation of Competitive Effects. Having answered the secondquestionraisedintheIntroductionsectionaffirmativelys namely, that NBZ sorption does exceed the hydrophobic effectsthe first question can therefore be reformulated: Does the enhanced NBZ sorption result from NBZ sorbing to identical sites as BEN and TOL but with a higher affinity? Or is NBZ sorbing to a subpopulation of sorption sites uniquely suited to NBZ? This question will be addressed in this section by comparing BEN-TOL and NBZ-TOL bi-solute systems. Whereas traditional bi-solute models (such as Ideal Adsorbed VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1611

FIGURE 6. Effect of the competitors BEN and NBZ on the sorption of TOL plotted as R vs the normalized concentrations using reference solvent n-hexadecane CH′(X) ) CW(X)KHW(X)aHebH. R ) qbs(CW)(qss(CW))-1, where qbs(CW) [mol kg-1] is the sorbed concentration of TOL in the presence of the competitor at a given CW(TOL), and qss(CW) [mol kg-1] is the hypothetical, single-solute sorbed concentration of TOL for the same measured CW(TOL). Solution Theory or the competitive Langmuir model) require single-solute data fits (and therefore automatically make certain assumptions on the sorption process), the approach presented here is novel in that it does not assume any particular sorption model. To confirm the suggestion above that BEN and TOL sorb by the same mechanism to identical sites on the charcoal, we replotted the BEN-TOL bi-solute data as the sum of the sorbed concentrations q(BEN) + q(TOL) [mol kg-1] versus the sum of the corrected solute concentrations CH′(BEN) + CH′(TOL) (Figure 5). Figure 5 clearly shows that all BENTOL bi-solute sorption data presented in Figure 2 fall onto one “isotherm”, which itself overlaps with the single-solute BEN and TOL sorption isotherms. The same results are obtained if benzene is used as the reference solvent (not shown). Hence, TOL and BEN are “ideal competitors” and replace each other on the charcoal surface in a 1:1 molar ratio. We now compare the two bi-solute systems by studying the effects of NBZ and BEN on the adsorption of TOL along 1612

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005

its single-solute isotherm. More specifically, we look at the effect of the competitor on the ratio

R)

qbs[CW(TOL)] qss[CW(TOL)]

(10)

where qbs[CW(TOL)] [mol kg-1] is the sorbed concentration of TOL in the bisolute system at a given solution concentration of toluene CW(TOL), and qss[CW(TOL)] [mol kg-1] is the hypothetical, single-solute sorbed concentration of TOL at the same measured solution concentration of TOL (i.e., same chemical potential). Values of qss[CW(TOL)] were determined by fitting a localized DLM to the single-solute sorption isotherm of TOL around the experimental CW(TOL) in the presence of the competitor (r2 ) 0.998 for all fits). For bisolute data points close to the single-solute isotherm of TOL, R ≈ 1. R decreases with increasing degree of competition. In Figure 6, the ratio R is plotted over the hexadecane normalized concentration of the competitor CH′ for six

FIGURE 7. Effect of the competitors BEN and NBZ on the sorption of TOL plotted as R vs the sorbed concentration of competitors. R ) qbs(CW)(qss(CW))-1, where qbs(CW) [mol kg-1] is the sorbed concentration of TOL in the presence of the competitor at a given CW(TOL), and qss(CW) [mol kg-1] is the hypothetical, single-solute sorbed concentration of TOL for the same measured CW(TOL). different TOL sorption points along the isotherm (i.e., from CW_0(TOL) ) 1.11 × 10-5 to 6.18 × 10-4 mol L-1). Figure 6 clearly shows that for any given CH′, ratios R are lower for the NBZ-TOL bi-solute system than for the BEN-TOL bi-solute system. Therefore, the following conclusions can be drawn: (i) sorbed concentrations of TOL are suppressed more by NBZ than by BEN, (ii) the population of NBZ sorption sites at least partly overlaps with the populations of TOL and BEN, and (iii) for the fraction of sorption sites all compounds compete for, NBZ has a higher affinity for sorbent than BEN and TOL. To determine what fraction of sorption sites is shared by all compounds, we replotted R versus the sorbed concentration of the competitor, q(competitor) [mol kg-1] (Figure 7). This plot allows comparison of the competitive effects between the three compounds on the basis of their absolute molar sorbed concentrations. Figure 7 shows that for any given sorbed concentration of competitor, R for TOL is equally affected by BEN and NBZ. It is therefore convincingly established that (i) all three compounds replace each other on the sorbent in molar ratios of 1:1:1; that (ii) NBZ has a

higher affinity for sorption sites than BEN and TOL; and finally that (iii) NBZ sorption exhibits some driving force in addition to nonspecific van der Waals interactions that primarily govern the sorption of BEN and TOL. Addressing Interactions of Nitrobenzene. The result that NBZ exhibits an additional driving force raises the obvious question: To which specific interactions can the enhanced sorption affinity of NBZ be ascribed? A number of different solute-sorbent interactions are possible, namely, (a) dipole-induced dipole interactions between NBZ and the surface; (b) hydrogen bonding where the surface of the charcoal is the H-donor and NBZ is the H-acceptor; or (c) π-π electron donor-acceptor interactions, where the electron-rich, conjugated ring system of the graphene unit on the charcoal surface is the electron donor and the relatively electron-poor aromatic ring of NBZ is the electron acceptor. Hunter at al. (44) remark that there is little evidence to suggest that dipole-induced dipole interactions are of major importance in interactions between aromatic systems. Consistent with this, we found that when isotherms are normalized for solvophobic effects, those of chlorinated VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1613

FIGURE 8. Sorption isotherms of benzene, toluene, and nitrobenzene on maple wood charcoal under neutral (pH 6.5; circles and triangles) and alkaline conditions (pH 11, squares and inverted triangles). Sorption data are presented as sorbed concentration q [mol kg-1] over the normalized concentrations in solvent n-hexadecane CH′. benzenes which possess a dipole moment coalesce with those of methylated benzenes which possess no significant dipole moment (24). H-donor groups on the charcoal surface comprise carboxylic acid (-COOH), alcoholic and phenolic (-OH), with traces of thiolic (-SH) and ammonium (-NH3+). Therefore, increasing the pH to 11 should eliminate or greatly reduce the H-donor capabilities of the charcoal (45). If the enhancement of NBZ sorption results primarily from Hbonding, then the normalized isotherm of NBZ under alkaline conditions is expected to approach those of BEN and TOL. Figure 8 shows the result of single-solute sorption experiments for BEN, TOL, and NBZ on the same charcoal under alkaline (pH 11) conditions in addition to the sorption data under neutral conditions (pH 6.5, replotted from Figure 1). The plot shows that (i) all the isotherms shift downward slightly at pH 11 relative to pH 6.5 by about the same degree and that (ii) the sorption isotherms of BEN and TOL overlap under alkaline as well as neutral conditions. As the relative position of the NBZ isotherm to the BEN and TOL isotherm is unaffected by elevating the pH, the specificity in the sorption of NBZ to charcoal does not result from H-bonding between NBZ and surface functional groups of the charcoal. In the companion paper (1), we demonstrated π-π EDA molecular complexation in solution between π-acceptor compounds (benzonitrile, 4-nitrotoluene, 2,4-dinitrotoluene, and 2,4,6-trinitrotoluene) and a series of π-donor compounds [naphthalene (NAPH), phenanthrene (PHEN), and pyrene (PYR)] as simple models representing the graphene surface. In other studies we have observed molecular complexation in aqueous and nonaqueous solution between donor PAHs and model π-acceptor humic acid subunits including quinones, N-heteroaromatic cations (46), and carboxyl-substituted aromatic compounds (47). In this study, we use 1H NMR to demonstrate the existence of molecular complexes between the π-acceptor NBZ and model π-donor charcoal graphene subunits including NAPH, PHEN, and PYR in chloroform solution. Due to the parallel-planar (stacked or offset-stacked) geometry of π-π complexes, protons on or near each ring are expected to show upfield chemical shifts (toward lower ppm values) due to shielding by “ring current” effects originating in the opposing ring (47). Figure 9 clearly demonstrates upfield chemical shifts of ring hydrogens of NBZ with increasing concentrations of π-donor (PAH). Figure 9 further shows that the upfield shift increases with increasing fused ring size of the PAH (donor polarizability). These results are consistent with π-π EDA interactions of NBZ. Only slight changes in the observed 1614

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005

FIGURE 9. Chemical shifts δ of 2,6-H of nitrobenzene (NBZ), 2,6-H of toluene (TOL), and benzene (BEN) vs concentration of naphthalene, phenanthrene, or pyrene in deuterated chloroform. chemical shift are detectable for BEN or TOL when paired with the strongest donor PYR. The notion that π-π EDA interactions play a crucial role in the sorption of nitroaromatics on condensed, polycyclic aromatic surfaces is further supported by sorption studies of a serious of nitrotoluenes on charcoal (1). We also note that Haderlein and co-workers (48, 49) have proposed n-π EDA interactions between electron pairs on the siloxane surface of clays and polynitroaromatic compounds.

Acknowledgments We thank the National Science Foundation, Bioengineering Program (BES-0122761) for funding. Furthermore, M.S. thanks the Environmental Education and Research Foundation for a scholarship (Francois Fiessinger Scholarship 2003). We thank Dongqiang Zhu for NMR measurements.

Literature Cited (1) Novakov, T. The role of soot and primary oxidants in atmospheric chemistry. Sci. Total Environ. 1984, 36, 1-10. (2) Goldberg, E. D. Black Carbon in the Environment; John Wiley & Sons: New York, 1985. (3) Schmidt, M. W. I.; Noack, A. G. Black carbon in soils and sediments: Analysis, distribution, implications, and current challenges. Global Biogeochem. Cycles 2000, 14, 777-793. (4) Kuhlbusch, T. A. J. Black carbon and the carbon cycle. Science 1998, 280, 1903-1904. (5) Lighty, S. J.; Veranth, J. M.; Sarofim, A. F. Combustion aerosols: Factors governing their size and composition and implications to human health. J. Air Waste Manage. Assoc. 2000, 50, 15651618. (6) Masiello, C. A.; Druffel, E. R. M. Black carbon in deep-sea sediments. Science 1998, 280, 1911-1913. (7) Accardi-Dey, A.; Gschwend, P. M. Assessing the combined roles of natural organic matter and black carbon as sorbents in sediments. Environ. Sci. Technol. 2002, 36, 21-29. (8) Gustafsson, O ¨ .; Haghseta, F.; Chan, C.; MacFarlane, J.; Gschwend, P. M. Quantification of the dilute sedimentary soot phase: implications for PAH speciation and bioavailability. Environ. Sci. Technol. 1997, 31, 203-209. (9) Gustafsson, O ¨ .; Gschwend, P. M. In Molecular Markers in Environmental Geochemistry; Eganhouse, R. P., Ed.; American Chemical Society: Washington, DC, 1997. (10) Gustafsson, O ¨ .; Bucheli, T. D.; Kukulska, Z.; Andersson, M.; Largeau, C.; Rouzaud, J.-N.; Reddy, C. M.; Eglinton, T. I. Evaluation of a protocol for the quantification of black carbon in sediments. Global Biogeochem. Cycles 2001, 15, 881-890. (11) Skjemstad, J. O.; Taylor, J. A.; Smernik, R. J. Estimation of charcoal (char) in soils. Commun. Soil Sci. Plant Anal. 1999, 30, 22832298. (12) Seiler, W.; Crutzen, P. J. Estimates of gross and net fluxes of carbon between the biosphere and the atmosphere from biomass burning. Clim. Change 1980, 2, 207-247. (13) Palota´s, AÄ . B.; Rainey, L. C.; Feldermann, C. J.; Sarofim, A. F.; Vander Sande, J. B. Soot morphology: An application of image

(14) (15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25) (26)

(27)

(28) (29)

(30) (31)

analysis in high-resolution transmission eletron microscopy. Microsc. Res. Tech. 1996, 33, 266-278. Harris, P. J. F. On charcoal. Interdiscip. Sci. Rev. 1999, 24, 301306. Manes, M. Activated carbon adsorption fundamentals. In Encyclopedia of Environmental Analysis and Remediation; Meyers, R. A., Ed.; John Wiley: New York, 1998; pp 26-68. Braida, W.; Pignatello, J. J.; Lu, Y.; Ravikovitch, P. I.; Neimark, A. V.; Xing, B. Sorption hysteresis of benzene in charcoal particles. Environ. Sci. Technol. 2003, 37, 409-417. Kleineidam, S.; Rugner, H.; Ligouis, B.; Grathwohl, P. Organic matter facies and equilibrium sorption of phenanthrene. Environ. Sci. Technol. 1999, 33, 1637-1644. Karapanagioti, H. K.; Kleineidam, S.; Sabatini, D. A.; Grathwohl, P.; Ligouis, B. Impacts of heterogeneous organic matter on phenanthrene sorption: equilibrium and kinetic studies with aquifer material. Environ. Sci. Technol. 2000, 34, 406-414. Bucheli, T. D.; Gustafsson, O. Quantification of the soot-water distribution coefficient of PAHs provides mechanistic basis for enhanced sorption observations. Environ. Sci. Technol. 2000, 34, 5144-5151. Chiou, C. T.; Kile, D. E.; Rutherford, D. W.; Sheng, G.; Boyd, S. A. Sorption of selected organic compounds from water to a peat soil and its humic-acid and humin fractions: Potential sources of the sorption nonlinearity. Environ. Sci. Technol. 2000, 34, 1254-1258. Jonker, M. T.; Koelmans, A. A. Extraction of polycyclic aromatic hydrocarbons from soot and sediment: Solvent evaluation and implications for sorption mechanism. Environ. Sci. Technol. 2002, 36, 4107-4113. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental Organic Chemistry, 2nd ed.; John Wiley & Sons: New York, 2003. Abraham, M. H.; Chadha, H. S.; Whiting, G. S.; Mitchelll, R. C. Hydrogen bonding. 32. An analysis of water-octanol and wateralkane partitioning and the delta log P parameter of Seiler. J. Pharm. Sci. 1994, 83, 1085-1100. Zhu, D.; Pignatello, J. J. Characterization of aromatic compound sorptive interactions with black carbon (charcoal) assisted by graphite as a model sorbent. Environ. Sci. Technol. (in press). Sposito, G. The Surface Chemistry of Soils; Oxford University Press: New York, 1984. Allen-King, R. M.; Grathwohl, P.; Ball, W. P. New modeling paradigms for the sorption of hydrophobic organic chemicals to heterogeneous carbonaceous matter in soils, sediments, and rocks. Adv. Water Resour. 2002, 25, 985-1016. Kamiya, Y.; Hirose, T.; Mizoguchi, K.; Naito, Y. Gravimetric study of high-pressure sorption of gases in polymers. J. Polym. Sci. B 1986, 24, 1525-1539. Barrer, R. M. Diffusivities in glassy polymers for the dual mode sorption model. J. Membr. Sci. 1984, 18, 25-35. Berens, A. Transport of organic vapors and liquids in poly(vinyl chloride). Makromol. Chem., Macromol. Symp. 1989, 29, 95108. Koros, W. J.; Paul, D. R.; Huvard, G. S. Energetics of gas sorption in glassy-polymers. Polymer 1979, 20, 956-960. Xing, B.; Gigliotti, B.; Pignatello, J. J. Competitive sorption between atrazine and other organic compounds in soils and model sorbents. Environ. Sci. Technol. 1996, 30, 2432-2440.

(32) Xing, B.; Pignatello, J. J. Dual-mode sorption of low-polarity compounds in glassy poly(vinyl chloride) and soil organic matter. Environ. Sci. Technol. 1997, 31, 792-799. (33) Lu, Y.; Pignatello, J. J. Demonstration of the “conditioning effect” in soil organic matter in support of a pore deformation mechanism for sorption hysteresis. Environ. Sci. Technol. 2002, 36, 4553-4561. (34) Stumm, W.; Morgan, J. J. Aquatic Chemistry, 3rd ed.; John Wiley & Sons: New York, 1996. (35) Xia, G.; Ball, W. P. Adsorption-partitioning uptake of nine lowpolarity organic chemicals on a natural sorbent. Environ. Sci. Technol. 1999, 33, 262-269. (36) Xia, G.; Ball, W. Polanyi-based models for the competitive sorption of low-polarity organic contaminants on a natural sorbent. Environ. Sci. Technol. 2000, 34, 1246-1253. (37) Grathwohl, P.; Rahman, M. M. Partitioning and pore-filling: solubility-normalized sorption isotherms of nonionic organic contaminants in soils and sediments. Isr. J. Chem. 2002, 42, 67-75. (38) Kleineidam, S.; Schu ¨ th, C.; Grathwohl, P. Solubility-normalized combined adsorption-partitioning sorption isotherms for organic pollutants. Environ. Sci. Technol. 2002, 36, 4689-4697. (39) Gavezzotti, A. Molecular free surface: A novel method of calculation and its uses in conformational studies and in organic crystal chemistry. J. Am. Chem. Soc. 1985, 107, 962-967. (40) Borisover, M. D.; Graber, E. R. Specific interactions of organic compounds with soil organic carbon. Chemosphere 1997, 34, 1761-1776. (41) Borisover, M.; Reddy, M.; Graber, E. R. Solvation effect on organic interactions in soil organic matter. Environ. Sci. Technol. 2001, 35, 2518-2524. (42) Borisover, M.; Graber, E. R. Hydration of natural organic matter: Effect on sorption of organic compounds by humin and humic acid fractions versus original peat material. Environ. Sci. Technol. 2004, 38, 4120-4129. (43) Goss, K.-U.; Schwarzenbach, R. P. Rules of thumb for assessing equilibrium partitioning of organic compounds: Successes and pitfalls. J. Chem. Educ. 2003, 80, 450-455. (44) Hunter, C. A.; Lawson, K. R.; Perkins, J.; Urch, C. J. Aromatic interactions. J. Chem. Soc., Perkin Trans. 2 2001, 651-669. (45) Radovic, L. R., Morena-Castilla, C., Rivera-Utrilla, J., Eds. Chemistry and Physics of Carbon: A Series of Advances; Marcel Dekker: New York, 2001; Vol. 27. (46) Wijnja, H.; Pignatello, J. J.; Malekani, K. Formation of pi-pi complexes between phenanthrene and model pi-acceptor humic subunits. J. Environ. Qual. 2004, 33, 265-275. (47) Jackman, L. M.; Sternhell, S. Applications of Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry, 2nd ed.; Pergamon Press: Oxford, UK, 1969. (48) Haderlein, S. B.; Weissmahr, K. W.; Schwarzenbach, R. P. Specific adsorption of nitroaromatic explosives and pesticides to clay minerals. Environ. Sci. Technol. 1996, 30, 612-622. (49) Haderlein, S. B.; Schwarzenbach, R. P. Adsorption of substituted nitrobenzenes and nitrophenols to mineral surfaces. Environ. Sci. Technol. 1993, 27, 316-326.

Received for review June 8, 2004. Revised manuscript received November 29, 2004. Accepted December 1, 2004. ES049135L

VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1615