Environ. Sci. Technol. 2006, 40, 221-227
Effects of Dissolved Carboxylates and Carbonates on the Adsorption Properties of Thiuram Disulfate Pesticides PANAGIOTA STATHI,† KONSTANTINOS C. CHRISTOFORIDIS,† ATHANASIOS TSIPIS,‡ DIMITRA G. HELA,§ AND Y I A N N I S D E L I G I A N N A K I S * ,† Laboratory of Physical Chemistry, Department of Environmental and Natural Resources Management, University of Ioannina, Seferi 2, 30100 Agrinio, Greece, Section of Inorganic and Analytical Chemistry, Department of Chemistry, University of Ioannina, Panepistimioupoli Douroutis, 45110 Ioannina, Greece, and Department of Farm Organization and Management, University of Ioannina, Seferi 2, 30100 Agrinio, Greece
The adsorption of thiram and disulfiram onto R-Al2O3 and montmorillonite clay has been studied in the presence of small carboxylate anions, bicarbonate, formate, and oxalate. At natural concentrations, HCO3- enhances dramatically the adsorption of both pesticides on R-Al2O3 and clay. An analogous significant enhancement of pesticide adsorption is also observed in the presence of formate and oxalate. Density functional theory calculations demonstrate that in solution a stable molecular complex between one molecule of thiram and one molecule of HCO3is formed with interaction energy -35.6 kcal/mol. In addition, two H2O molecules further stabilize it by an interaction energy of -3.6 kcal/mol. This clustering [thiramHCO3--2H2O] leads to a change of the electronic structure and the ultraviolet-visible spectrum of thiram that is observed experimentally. Surface complexation modeling shows that the molecular cluster [thiram-HCO3-2H2O], which bears a total net charge of -1, is responsible for the observed enhanced adsorption on the charged surface of alumina and clay at pH below their points of zero surface charge. The results reveal a novel pervasive role of carboxylate anions and particularly HCO3- on the adsorption of dithiocarbamate pesticides in natural waters.
Introduction Thiram (C6H12N2S4, tetramethyl thiuram disulfide) and disulfiram (C10H20N2S4, tetraethyl thiuram disulfide) are among the thiuram disulfides as animal repellents to protect fruit trees and ornamentals from damage by rabbits, rodents, and deer or as rubber vulcanization accelerators (1). The half-life for thiram in soil has been reported to be between 2 and 3 weeks (2). Both thiram and disulfiram are * Corresponding author phone: +003 26410 39516; fax: +003 26410 39576; e-mail:
[email protected]. † Department of Environmental and Natural Resources Management. ‡ Department of Chemistry. § Department of Farm Organization and Management. 10.1021/es051451s CCC: $33.50 Published on Web 11/22/2005
2006 American Chemical Society
hydrophobic neutral molecules that are strongly adsorbed on soils (1, 3, 4), making their elimination from the environment difficult. However, the physicochemical parameters for the adsorption of thiram and disulfiram in soils have not been evaluated. Strong adsorption of thiram on lignin has been reported (5). The carboxyl groups of lignin were postulated (5) to play an important role in the observed strong adsorption of thiram; however, no further evidence was provided. The adsorption mechanisms of organic molecules and pesticides on soils are quite diverse (2). Appropriate studies on soil constituents allow the identification of the controlling factors (6-12) and mechanisms (8, 11-12). Among the soil constituents, clay minerals (2, 4, 7) and mineral oxides (2, 5, 6) strongly influence the fate of pesticides in soils due to their large surface area and abundance in agricultural soils. In this context, in the present work we have studied the adsorption properties of thiram and disulfiram on aluminum oxide and montmorillonite clay. Carbon dioxide is ubiquitous in natural environments and is present either as gaseous CO2 or dissolved in water, i.e., as CO2(aq) and carbonate species HCO3- and CO32-. A typical range of carbonate species in groundwater is expected to be in the range of 0.5-10 mM (13, 14). Atmospheric CO2 contributes more than 60% of the carbonate present in river water, while 31% comes from carbonate minerals and 7% from oxidation of organic carbon in sediments (15). The ubiquity of CO2 and carbonates therefore makes it necessary to include these compounds in studies of adsorption and mobility of chemical compounds in the natural environment. However, dissolved species, i.e., CO2, HCO3-, and CO32-, have mainly been neglected or specifically avoided in adsorption studies, and their possible impact on parameter estimation of adsorption phenomena anions has not been fully appreciated (14, 16). In the present work the role of dissolved carbonates on the adsorption of thiuram disulfides has been examined in detail. The objectives of the present work were (a) to investigate the adsorption mechanism of thiram and disulfiram on alumina and clay and (b) to elucidate the role of small carboxylates, particularly HCO3-, on the adsorption of these pesticides. The observed phenomena were interpreted by the combined use of quantum chemical calculations and macroscopic surface complexation modeling (17).
Experimental Procedures Materials. Aluminum oxide R-Al2O3 was obtained from Sigma-Aldrich (Lot No. 0.5716CU-414). The clay used was a synthetic sodium-saturated montmorillonite, Kunipia F (KUN) [Na0.87[Al3.12 Fe(III)0.20Mg0.61Ti0.01](Si7.90Al0.10)O20(OH)4], purchased from Kunimine Industries Co. This is a very high purity montmorillonite and was used as the model clay in order to eliminate potential interferences from spurious contaminants often occurring in natural montmorillonite (18). The surface charge and protonation constants were determined by mass titration and potentiometric titration. Analytical quantitative determination of thiram and disulfiram in aqueous solution was done with HPLC (19, 20). Light absorption spectra were recorded in the ultraviolet-visible (UV-vis) wavelengths (190-800 nm). Full experimental details are provided in the Supporting Information. Adsorption Experiments. For the adsorption studies, batch samples were prepared by spiking pesticides from a stock solution in methanol with appropriate volumes, typically 20 mL, of Milli-Q water. This resulted in methanol concentrations below 1%. Adsorbent suspensions, e.g., either VOL. 40, NO. 1, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
221
TABLE 1. Physicochemical Parameters of r-Al2O3 and Clay Used in This Study R-Al2O3 BET 7 m2/ga PZNC 8.65,a 7.7-9.00b
montmorillonite clay CEC 119 mequiv/100 ga PZNC 10.55a particle size 200 nmb
a This work. b Kosmulski, M.; pH-dependent surface charging and point of zero charge II. Update: J. Colloid Interface Sci. 2004, 24, 214-224.
R-Al2O3 or montmorillonite, at different pH were prepared and suitable volumes of pesticide solutions were added. The pH was then adjusted to the desired value, using small volumes of HNO3 and NaOH. After stirring at room temperature for a predetermined time never exceeding 120 min to avoid degradation of the pesticides, the suspension was centrifuged and the concentration of thiram or disulfiram in the solution phase was determined by either HPLC or UVvis spectophotometry. To avoid CO2 contamination (samples without CO2) the suspension was purged with nitrogen gas for 20 min prior to the addition of the pesticide. The CO2free aliquots were stored in airtight Teflon tubes. In all cases there was 3 mg of adsorbents and the concentration of thiram or disulfiram was 3 or 4 mg L-1. Surface Complexation Modeling. The surface reactions used for surface complexation modeling (17) are described in Table 2. FITEQL 4.0 (22) was used to determine the best fit of various surface complexation reactions or combinations of reactions to the experimental thiram adsorption data. The Davies equation was used for activity (18) corrections of aqueous species only. The relative experimental errors for the FITEQL input values of total thiram concentration and log [H+] were 5%. Density Functional Theory (DFT) Calculations. The structural and electronic properties of all compounds were computed at the Becke’s three-parameter hybrid functional (26) combined with the Lee-Yang-Parr (27) correlation functional, termed as the B3LYP level of DFT. All calculations were performed using the Gaussian03 (G03) (28) series of programs and spin-restricted wave functions, as no unpaired electron exists in the systems. The energy minimum for each system was verified by calculating the vibrational frequencies that result in absence of the imaginary eigenvalue. The DFT calculations show that in aqueous solution HCO3- is stabilized in a hydrated form, i.e., by two water molecules, in agreement with previous studies (29, 30). Thus in the geometry
optimization of thiram in the presence of HCO3-, the influence of H2O molecules has been examined. Calculated Electronic Spectrum. DFT and time-dependent DFT (TD-DFT) calculations were used to obtain the ultraviolet-visible (UV-vis) absorbance spectrum of thiram and of the system [thiram + HCO3 + 2H2O]. The calculation was done using the RB3LYP functional and the 6-31G** basis set. TD-DFT calculations were performed by including 50 excited states. The calculated energy states and the oscillator strengths are described in detail in Table S1 of the Supporting Information.
Results and Discussion Carboxylates Modify the UV-Vis Spectra of Thiram and Disufliram. Thiram-Bicarbonate. The UV-vis spectrum of thiram under open-air conditions (partial pressure PCO2 ) 10-3.5 atm) has a characteristic light absorbance band at 240265 nm [see Figure 1A(a) solid line]. Removal of CO2, i.e., by purging the solution with pure N2 gas, while the pH was kept at 6.5 [see Figure 1A(b,c)], progressively converted the spectrum to that shown in Figure 1A(d). Repetitive cycles of exposure at ambient CO2 at pH >6 resulted in reversible conversion of the spectrum between the forms shown in Figure 1A(a-d) (not shown). This spectral change appeared after relatively short incubation periods, ca. 20 min. The observed spectral enhancement at 240-265 nm required CO2 to be present in the aqueous solution and pH to be over pH 6. These requirements can be explained if we assume that the carbonate species involved is HCO3according to the elementary pH dependence of carbonates in solution.
CO2(aq) + OHHCO3-
pK1)6.34
pK1)10.33
T
T
HCO3-
(1)
H+ + CO32-
(2)
To determine whether HCO3- was indeed the species involved, we recorded the UV-vis spectrum of thiram after addition of 100 µM NaHCO3 in a CO2-free solution of thiram. Indeed, the light absorbance at 240-265 nm was enhanced after addition of NaHCO3 (Figure 1B). A subsequent decrease of the pH of this sample to 5.6 without purging the CO2 converted the spectrum to that seen prior to the addition of NaHCO3. Thus, on the basis of the present data, we conclude that the change in the UV-vis spectrum of thiram was correlated exclusively with the presence of HCO3- in the
TABLE 2. Equilibrium Equations and Optimized Constants of Reactions log K (IS ) 0 M) reaction water dissociations [OH-] ) [H+]-1 surface sites acid-base reactions [XOH2+] ) [XOH][H+] [XO-] ) [XOH][H+]-1 carbonate acid-base reactions [HCO3-] ) [CO3][H+]a [HCO3-] ) [CO3][H+]2 a bicarbonate sorption reaction [XOH2-HCO 3] ) [XOH][CO3][H+]2 thiram sorption reaction [XOH2-thiram] ) [XOH][thiram][H+] thiram-bicarbonate complexation reaction [thiram-HCO3] ) [thiram][CO3][H+]b complex sorption reaction [XOH2-HCO3-thiram] ) [XOH][CO3][thiram][H+]2 a
r-alumina
montmorillonite
-14
-14
8.40 ( 0.05 8.90 (0.05
10.9 ( 0.2 10.2 ( 0.2
10.35 16.50
10.35 16.50
3.4 ( 0.8
2.8 ( 0.8
2.5 ( 0.5
2.8 ( 0.5
-0.046
-0.046
21 ( 2
22 ( 2
Data taken from Matrell, E. A.; Motekaitis, R. J. Determination and Use of Stability Constants, 2nd ed.; Wiley-VCH: New York, 1992. on the basis of the interaction energy by DFT.
222
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 1, 2006
b
Calculated
FIGURE 1. (A) Effect of CO2 on the UV-vis spectrum of 12.5 µM thiram (a, solid line) in open-air solution at pH 6.5. Traces b-d are UV-vis spectra of sample a after bubbling with N2 for 10, 20, and 40 min, respectively. In all cases the pH was adjusted to 6.5. (B) Influence of dissolved HCO3- and other small carboxylates on the UV-vis spectra of thiram: (dotted line) 10 µM thiram in CO2-free solution, pH 6.35; (solid line) 10 µM thiram plus of 100 µM NaHCO3, pH 6.45; (dashed line) 10 µM thiram plus 10 µM oxalate, pH 6.34; (dashed-dotted line) 10µM thiram plus 10 µM formate, pH 6.36. solution. The reversibility of the observed UV-vis spectral changes, i.e., reversibility as a function of the solution pH and/or the removal of HCO3-, indicates that the observed UV-vis spectral change is due to a specific interaction between thiram and HCO3-, which takes place in aqueous solution. Note that the HPLC data excluded dissociation or other molecular transformation of thiram. Thiram-Oxalate, Thiram-Formate. The spectral change observed for thiram due to bicarbonate was also observed in the presence of other small carboxylate anions such as formate and oxalate. The spectra in Figure 1B show that the presence of 10 µM formate or oxalate induced the characteristic enhancement in the UV-vis spectrum of thiram at 240-265 nm. As in the case of bicarbonate, this spectral change was observed rapidly, i.e., within 20 min at room temperature. Disulfiram-Bicarbonate. Under open-air conditions (partial pressure PCO2 ) 10-3.5 atm), a significant light absorbance at 250-270 nm was observed in the UV-vis spectrum of disulfiram [see spectrum S1(a) of the Supporting Information]. As in the case of thiram, the effect of CO2 on the UVVis spectrum of disulfiram was reversible. Purge of the solution with N2 progressively diminished the light absorbance at 250-270 nm [spectra S1(a-e)]. Overall the data show that thiram and disulfiram interact in a specific, reversible manner with the carboxylate anions bicarbonate, formate, and oxalate. In the case of HCO3-, the kinetics of this interaction at room temperature is in the range of 20 min. The HPLC data show that during this treatment thiram and disulfiram remain unmodified. Overall, from this information it seems inevitable to conclude that (a) for natural concentrations of CO2(aq) the UV-vis spectral changes are due to the reversible interaction between thiram
FIGURE 2. TD-DFT (B3LYP, 6-31G**)-calculated UV-vis spectrum for thiram (A) and [thiram + HCO3- + 2H2O] (B). The bars correspond to the calculated oscillator strengths described in detail in Table S1 of the Supporting Information. or disulfiram with HCO3-. (b) Similar effects are observed for formate and oxalate. The molecular details of this effect can be explored with the aid of DFT calculations. DFT Calculations. Thiram. The calculated electronic transitions for thiram in the UV-vis region are displayed in Figure 2A. According to the TD-DFT calculations (see Table S1 of the Supporting Information), most of the transitions are determined from contributions between the five higher occupied molecular orbitals and the five lower unoccupied molecular orbitals. The strong spectral transitions predicted by DFT at 200220 nm in Figure 2 are responsible for the strong absorbance seen in the experimental spectrum in Figure 1. Noteworthy, the TD-DFT results show that at each wavelength more than one electronic transition contributes. For example, the dominant transition at 213 nm is a superposition from transitions between the five orbitals with energies just below the HOMO, i.e., H-1, H-2, H-3, H-4, H-6, to the LUMO or the next two orbitals above it, i.e., L, L+1, L+2 (see Table S1). [Thiram-HCO3-]: Energies. The interaction of thiram with HCO3- results in its stabilization by about -250 atomic units (see Figure 3B). Translated into kcal/mol, this is an interaction energy of -35.6 kcal/mol. The H2O molecules have a further stabilizing effect, i.e., an interaction energy of -3.6 kcal/ mol. Despite this energetic stabilization effect, we found that the H2O molecules have no resolvable effect on the calculated UV-vis transitions. Thus, to keep the discussion to a tractable level, we discuss in more detail the calculations for the system [thiram + HCO3- + 2H2O]. Overall the interaction energy of thiram with HCO3- + 2H2O is -39.2 kcal/mol. This interaction is weaker than a covalent bond and this explains the experimentally observed reversibility. VOL. 40, NO. 1, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
223
FIGURE 3. (A) DFT (B3LYP, 6-31G**)-optimized structure of [thiram + HCO3- + 2H2O] that was used for the electronic spectrum calculations. (B) DFT-B3LYP (6-31G*(+))-calculated energies for optimized structures of thiram interacting with HCO3- and one or two H2O molecules. Structure. The DFT-optimized structure of [thiram + HCO3-+ 2H2O] (Figure 3A) reveals that thiram adopts an almost symmetric U-shaped conformation where the four CH3 groups of N1 and N2 form a network of H-bonds with the two H2O and the O-atoms of HCO3-. An H-bond, R(O‚‚‚H) ) 1.77 Å, is formed between the two water molecules (see Figure 3A). In addition, one water molecule is H-bonded to one CH3 group of thiram at 2.37 Å. Bicarbonate interacts with thiram via its two carboxyl oxygens. O15 is H-bonded to the H24 of thiram, R(O15‚‚‚H24) ) 2.24 Å. The carboxyl oxygen O16 of HCO3- plays a central role in this structure. This oxygen interacts with the hydrogen atoms H20 and H28 of the methyl groups of thiram, plus the H34 atom from one of the waters. An inspection of the HOMO shows that a nonzero covalency, i.e., 5/100 orbital overlap, is developed between the electrons of O16 of bicarbonate and the electron cloud of thiram via H20 and H28. This weak “pseudocovalent” overlap of O16 with H20 and H28, together with the H-bonds, is responsible for the thermodynamic stabilization of the structure at room temperature. These weak interactions are in agreement with the observed reversibility of the interaction between thiram and HCO3-. Finally, it is worth mentioning that the two interacting waters are in close proximity to N1 of thiram. Their eventual conversion to OH-, i.e., at alkaline pH, is likely to lead to a nucleophilic attack at the N1, which is known to be responsible for the decomposition of thiram at pH >9 (5). UV-Vis Spectra. The TD-DFT calculations predicted that the electronic transitions in the UV-vis region are strongly influenced by the interaction between thiram and HCO3(see Figure 2). The main electronic transitions are shifted in the region above 245 nm up to 265 nm, in agreement with the experimental observation (see Figure 1A,B). According to the TD-DFT calculations, the main effect is due to the 224
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 1, 2006
FIGURE 4. (A) Adsorption of thiram on r-Al2O3 in CO2-free solution (b) or under open-air (PCO2 ) 10-3.5atm) conditions (9). (B) Adsorption of 3 ppm (1) or 4 ppm (2) thiram on clay in CO2-free solution or under open-air (PCO2 ) 10-3.5atm) conditions with 3 ppm (9) or 4 ppm ([) thiram. Open symbols are theoretical predictions based on FITEQL by using the parameters listed in Table 2. At the top of the figures the Schindler plots help the visualization of the charges of the interacting species as a function of the pH. enhancement of the intensity at 244 and 250 nm, which originate from transitions involving the L + 1 and L + 2 molecular orbitals (see Table S1 of the Supporting Information). In summary, the DFT calculations provide a theoretical basis for the observed phenomena. The carboxyl group of HCO3- is mainly responsible for the association of thiram with bicarbonate. On the bais of the similarity of the experimental observations for thiram and disulfiram, the present theoretical findings can be extrapolated to the case of disulfiram interacting with bicarbonate as well as to the interaction of the disulfide pesticides with formate and oxalate. The importance of the carboxyl group together with H-bond from water might be of relevance to the interaction of these thiuram disulfides with carboxyl-rich soil organic matter. In this context, we may suggest the association with carboxyl groups to be responsible for the observed irreversible adsorption of thiram in lignin (5). Adsorption of Thiram-Bicarbonate on r-Al2O3 and Clay. Figure 4 shows the adsorption edge of thiram onto R-Al2O3 (part A) or clay (part B) as a function of the pH. In CO2-free aqueous solution, thiram was only slightly adsorbed onto R-Al2O3 with a maximum of ca. 18% (Figure 4A). A slightly higher adsorption on the clay of about 24% was observed (Figure 4B). The clay is positively charged at pH
