Article pubs.acs.org/JPCA
Theoretical Investigation on Monomer and Solvent Selection for Molecular Imprinting of Nitrocompounds Julia Saloni,†,‡,* Kiara Walker,‡ and Glake Hill, Jr‡ †
Copiah-Lincoln Community College, 11 Co-Lin Circle, Natchez, Mississippi 39120, United States Department of Chemistry and Biochemistry, Jackson State University, 1400 J. R. Lynch Street, Jackson, Mississippi 39217, United States
‡
ABSTRACT: The aim of this work is to serve as a guideline for the initial selection of monomer and solvent for the synthesis of the nitrocompound-based molecularly imprinted polymers, MIPs. Reported data include evaluation of six systems with the ability to form noncovalently bonded monomer−template complexes. These systems are represented by the following aliphatic and aromatic molecules: acrolein, acrylonitrile, 2,6-bisacrylamide, 4-ethylenebenzoic acid, methyl methacrylate, and 2vinylpyridine. Cave models for selected monomers are also presented and supported by binding energy analysis under various conditions. Solvent effects on monomer− template binding energy have been studied for four solvents: acetone, acetonitrile, chloroform, and methanol. Additionally, systems such as 2,4-dinitrotoluene (2,4DNT), 2,6-dinitrotoluene (2,6-DNT), pentachlorophenol (PCP), and 3,6-dichloro-2methoxybenzoic acid (Dicamba) have been used to study selectivity of acrolein-based MIP toward TNT detection. The density functional theory, DFT, method has been used for all structural, vibrational frequency, and solvent calculations.
1. INTRODUCTION Application of theoretical calculations can significantly limit the number of necessary search syntheses performed for the selection of the best compound for a particular experiment. Computationally-based evaluation extracts the most useful substances out of the pool of possible systems quickly and cost efficiently, thereby accelerating the experiment itself. Molecular imprinting, MI, is an experimental technique for creating receptor structures on a polymer surface that can selectively bind to molecules of interest. MI has a broad spectrum of applications, including, but not limited to, mixture separation,1,2 extraction,3,4 detection,5,6 and drug delivery.7,8 As presented in Figure 1, the molecular imprinting process9 can be divided into
of the functional groups. The formed MIP is ready to rebind the template for its detection, separation, or extraction. Nitrocompounds, especially 2,4,6-trinitrotoluene,10,11 are the group of highly energetic materials for which detection is in demand due to security5,12 as well as environmental6 reasons. Besides TNT causing an immediate danger, as an explosive material, this substance can also contaminate soil and groundwater.6 Hence, it has been proposed to apply the MIP technique for the detection and possible removal of the TNT and its derivatives from the environment. In our previous work, the theoretical model for the evaluation of the binding energies between the monomer and template for both DNT13 and TNT14 species has been designed. The presented model allows us to predict qualitatively and quantitatively possible interactions in the MIP system. Theoretical data reported in both manuscripts have been verified by experimental FT-IR spectra. The above-mentioned model has been utilized in the course of this work to perform theoretical calculations of six selected aromatic and aliphatic monomers, for selection of the best one for the imprinting of TNT. As presented in Figure 2, studied monomers include acrolein, acrylonitrile, 2,6-bisacrylamide, 4ethylenebenzoic acid, methyl methacrylate, and 2-vinylpyridine. The effects of the four most common solvents (acetone, acetonitrile, chloroform, and methanol) on the 1:1 monomer− template and selected cave−template systems have also been reported. Additionally, the selectivity of acrolein-based MIP
Figure 1. MIP process: (A) self-assembly; (B) polymerization; (C) solvent extraction.
three main stages. During the first stage (monomer selfassembly) the molecule of interest, the template, is surrounded by monomers positioned to interact with functional groups of the template. The second stage (polymerization) occurs when monomers polymerize with cross-linking agents to form a cave around the template. During the final, third, stage (extraction), the template is removed, by a solvent extraction process, from the formed cavity. The empty cavity possesses the characteristics of the template: complementary shape, size, and positions © 2013 American Chemical Society
Received: December 26, 2011 Revised: January 16, 2013 Published: January 23, 2013 1531
dx.doi.org/10.1021/jp2124839 | J. Phys. Chem. A 2013, 117, 1531−1534
The Journal of Physical Chemistry A
Article
study solvent effects on the binding energy. The cave in first model, Figure 4a, consists of twelve acrolein molecules forming six dimers around the TNT template. Figure 4b shows an empty acrolein cave. The second cave is built from eight aromatic 2-vinylpyridine molecules, forming four dimers around TNT, as presented in Figure 4c. Figure 4d shows an empty 2-vinylpyridine cave. The study on the selectivity of MIP toward the TNT molecule has been performed using a previously optimized acrolein cave, reported in Figure 4b. The geometry of the cave has been preserved to mimic the real solid state polymeric structure of formed MIP. Binding energies of four species, selected due to their presence in groundwater and soil and structural similarities to TNT, have been calculated. The structures of 2,4-DNT, 2,6-DNT, PCP, and Dicamba have been reported in Figure 5. The density functional theory,15 DFT, method with the B3LYP16−18 functional has been applied to optimize all selected species and their complexes with 2,4,6-trinitrotoluene and to perform solvent studies. No symmetry constraints were imposed during the geometry optimization processes. Geometry searches for a variety of possible configurations were performed to obtain the global minimum, which has been verified by DFT vibrational frequency calculations. The standard Pople basis set, 6-31G(d,p), was used in this study.19 Basis set superposition error,20,21 BSSE, has been included in calculations of the total binding energy of 1:1 complexes and their components. The effect of the solvent on the studied systems was modeled by the CPCM model.22 Calculations presented in this work have been carried out utilizing the Gaussian0323 suite of programs.
Figure 2. Structures of organic monomers.
toward detection of TNT has been studied using a cave model interacting with 2,4-DNT, 2,6-DNT, PCP, and Dicamba molecules, as possible TNT competitors. Overall, this work aims to provide a guideline for the monomer and solvent selection process for the molecular imprinting of nitroaromatic compounds, as well as an insight into MIP selectivity.
2. COMPUTATIONAL DETAILS This study utilizes previously developed 1:1 model for TNT imprinting,14 where TNT interacts with a single monomer molecule through hydrogen bonding. Although the monomer can form three possible isomers (ortho, meta, and para) with respect to the methyl group of TNT, the model is based on the lowest energy ortho complex, As presented in Figure 3, monomer (acrolein) interacts with methyl and the ortho nitro group of the TNT molecule through hydrogen bonds. Two cave models made of the aliphatic and aromatic monomers (underlined in Table 1) have been designed to
3. RESULTS AND DISCUSSION Six, the most popular, monomers used for molecular imprinting, MI, have been selected for computational study of monomer−template complex formation occurring during imprinting of 2,4,6-TNT. Study reports binding energy, BE, hydrogen-bonding distances and monomer−TNT BE affected by the presence of solvent. This work refers to previously reported models13,14 designed for modeling of monomer selfassembly process occurring during molecular imprinting of selected nitrocoumpounds, where the most effective binding site for the functional monomer (acrylic or methacrylic acid) involves the methyl group and ortho positioned nitro group of the nitroaromatic’s molecule. As a continuation, this work presents summary of monomer−template binding properties for aliphatic (acrolein, acrylonitrile, methyl methacrylate), and aromatic (2,6-bisacrylamide, 4-ethylenebenzoic acid, and 2vinylpyridine) functional monomers interacting with 2,4,6TNT. Binding energies, BE, and binding distances for all six studied complexes, calculated in a gas phase, and selected solvents have been reported in Table 1. Presented values of BE, calculated in the gas phase, decrease from 7.03 kcal/mol for 2ethylenebenzoic acid-TNT to 3.10 kcal/mol for the 2vinylpyridine−TNT interacting systems. Gas phase calculated binding energies have been corrected by counterpoise correction calculations. BSSE lowers the value of BE for about 2.5 kcal/mol, which, according to Grabowski,24 classifies them among weak hydrogen interactions. Monomer−template binding distances are placed within the range of ∼1.90 to ∼3.10 Å. Taking into account a fact that the MI polymerization process occurs in a solution, it is crucial to mimic solvent
Figure 3. 1:1 model of TNT interacting with monomer (acrolein). 1532
dx.doi.org/10.1021/jp2124839 | J. Phys. Chem. A 2013, 117, 1531−1534
The Journal of Physical Chemistry A
Article
Table 1. Bonding Properties of Studied Monomer−Template Complexes Including: Binding Energies, BE, Calculated in a Gas Phase as Well as Selected Solvents, and Hydrogen Binding Distancesa complex monomer− template
BE in a gas phase
H-bonding type
H-bonding distance 2.417 2.585 2.516 2.549 2.886 2.327
acrolein
5.83
acrylonitrile
5.69
2,6-bisacrylamide pyridine
6.04
CO···H−C C−H···O−N C≡N···H−C C−H···O−N C−H···O−N N−H···O−N
4-ethylenebenzoic acid
7.05
N−H···O−N ring N···H−C O−H···O−N
2.364 2.868 1.931
methyl methacrylate
6.22
2-vinylpyridine
3.1
CO···C ring CO···H−C C−H···O−N C−H···O−N ring N···H−C C−H···O−N C−H···O−N
3.095 2.182 2.391 2.467 2.438 2.508 2.547
a
complex BE in methanol
complex BE in acetone
complex BE in chloroform
complex BE in acetonitrile
2.42
2.48
3.12
2.36
1.76
1.74
2.48
1.67
−0.66
−0.07
1.15
−0.42
−0.61
−0.77
0.57
−0.95
2.42
2.29
3.03
2.26
0.71
0.57
2.36
0.56
Calculations performed at the B3LYP/6-31G(d,p) level of theory. Energies in kcal/mol, angles in angstroms.
Figure 5. Structures of selected organic molecules present in groundwater.
acetone, acetonitrile, or methanol results in disintegration of the complex (no monomer−template interaction is observed). Hence, those three solvents may not be the best medium for TNT imprinting with 2,6-bisacrylamide or 4-ethylenebenzoic acid. Out of three initially selected aromatic functional monomers only 2-vinylpyridine has a potential for molecular imprinting of nitrocoumpounds in the presence of all four studied solvents. The 2-vinylpyridine−TNT calculated BE for a 1:1 ratio (monomer:template) in solvent solution amounts to between 0.56 kcal/mol for acetonitrile and 2.36 kcal/mol for chloroform. Summarizing, results of the solvent effect on the aliphatic monomer−TNT interaction, reported in Table 1 show the following trends: • All four studied solvents are suitable for molecular imprinting of nitrocompound template. • In all calculated cases the value of BE decreases in the order acrolein > methyl methacrylate > acrylonitrile for monomer−template complexes. • The single value of BE is larger for aliphatic systems (for about 1.5 kcal/mol) when compared to BEs of studied aromatic systems.
Figure 4. Cave model of TNT encapsulated in polymer and empty cave (models for acrolein and 2-vinylpyridine).
environment in computational studies. One of the goals of this work is to provide solvent analysis for the polymerization process, and overall help with selection of the best solvent for imprinting of nitrocompounds. Using the CPCM22 solvation model at the DFT level of theory, we attempt to instigate solvent effects on the strength of H-bonding between monomer and template during self-assembly. Acetone, acetonitrile, chloroform, and methanol have been selected for individual study of their influence on BE. As presented in Table 1, introduction of solvent dramatically changes the value of BE when compared to one calculated in the gas phase. The most spectacular effect on binding energy can be observed for the two TNT−aromatic complexes, 2,3-bisacrylamide−TNT and 4ethylenebenzoic acid−TNT systems, where addition of 1533
dx.doi.org/10.1021/jp2124839 | J. Phys. Chem. A 2013, 117, 1531−1534
The Journal of Physical Chemistry A
■
Cave models for TNT imprinting, using acrolein and 2vinylpyridine monomers, have been designed to create more realistic view of monomer−template binding in MIPs, as shown in Figure 4. In the first model the TNT molecule has been surrounded by six acrolein dimers and in the second by four 2vinylpyridine dimers. Cave−template binding energies in a gas phase and solvents have been reported in Table 2. The analysis
BE in a gas phase
BE in methanol
BE in acetone
BE in chloroform
BE in acetonitrile
acrolein 2vinylpyridine
21.59 14.8
7.99 3.37
8.28 2.2
10.86 5.33
7.80 3.61
a
Calculations were performed at the B3LYP/6-31G(d,p) level of theory. Energies are in kcal/mol.
of collected data shows that introduction of the chloroform the system lowers cave−template biding energy for 10.73 kcal/mol for the acroleine cave and 9.47 kcal/mol for the 2-vinylpyridine cave. Introduction of other solvents lowers cave−template binding energies even more significantly. However, all computationally studied solvents seems to be suitable as a medium for the TNT imprinting with acrolein and 2vinylpyridine. Finally, the acrolein cave model has been used to study the selectivity of acrolin-based MIP toward the TNT molecule. As acrolein-based MIP is a solid state polymer, in our study the geometry of the acrolein cave model has been preserved. No symmetry constraints have been imposed on the four species chosen for the selectivity study. These pollutants have been selected due to their structural resemblance to TNT and abundance in ground waters and soil. Binding energies between all four species and the acrolein cave have been calculated in a gas phase. The study indicates that both DNT molecules bind to the acrolein cave but the binding energy is significantly smaller than for TNT, reported as 21.59 kcal/mol. Binding energies are 10.98 and 3.73 kcal/mol for 2,4-DNT and 2,6DNT, respectively. PCP and Dicamba do not bind to the acrolein cave; performed calculations show that they are strongly repelled by the cave. On the basis of the performed calculations, out of the set of four studied solvents the most universal solvent for MIP of TNT, regardless of the type of used functional monomer, seems to be chloroform. Acetone, acetonitrile, and methanol are proposed to be used for aliphatic monomer imprinting or for template extraction of aromatic monomers, with the exception of 2-vinylpyridine.
■
REFERENCES
(1) Jo, S.-H.; Lee, S.-Y.; Park, K.-M.; Yi, S. C.; Kim, D.; Mun, S. J. Chromatogr. A 2010, 45, 7100−7108. (2) Rezaei, B.; Jafari, M. T.; Khademi, R. Talanta 2009, 79, 669−675. (3) Pap, T.; Horvath, V.; Tolokan, A.; Horvai, G.; Sellergen, B. J. Chromatogr. A 2002, 973, 1−12. (4) Ersoz, A.; Denizli, A.; Sener, I.; Atilir, A.; Diltemiz, S.; Say, R. Sep. Purif. Technol. 2004, 38, 173−179. (5) Toal, S. T.; Trogler, W. C. J. Mater. Chem. 2006, 16, 2871−2883. (6) Alizadeh, T.; Zare, M.; Ganjali, M. R.; Norouzi, P.; Tavana, B. Biosens. Bioelect. 2010, 25, 1166−1172. (7) Piacham, T.; Nantasenamat, C.; et al. Molecules 2009, 14, 2985− 3002. (8) Alvarez-Lorenzo, C.; Concheiro, A. J. Chromatogr. B 2004, 804, 231−245. (9) Cormack, P. A. G.; Elorza, A. Z. J. Chromatogr. B 2004, 804, 173−182. (10) Politzer, P.; Murray, J. S.; Koppes, W. M.; Concha, M. C.; Lane, P. Cent. Eur. J. Energet. Mater. 2009, 6, 165. (11) Clarkson, J.; Smith, W. E.; Batcheldar, D. N.; Smith, D. A.; Coats, A. M. J. Mol. Struct. 2003, 648, 203−214. (12) Bunte, G.; Hurttlen, J.; Pontius, H.; Hartlieb, K.; Krause, H. Anal. Chim. Acta 2007, 591, 49−56. (13) Saloni, J.; Dasary, S. S. R.; Yerramilli, A.; Yu, H.; Hill, G., Jr. Struct. Chem. 2010, 21, 1171−1184. (14) Saloni, J.; Lipkowski, P.; Dasary, S. S. R.; Anjaneyulu, Y.; Yu, H.; Hill, G., Jr. Polymer 2011, 52, 1206−1216. (15) Parr, R. G.; Yang, W. Density-Functional Theory of Atoms and Molecules; Oxford University Press: New York, 1994. (16) Becke, D. J. Chem. Phys. 1993, 98, 5648−5652. (17) Vosko, S. S.; Wilk, L.; Nusiar, M. Can. J. Phys. 1980, 58, 1200− 1211. (18) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789. (19) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54, 724−728. (20) Simon, S.; Duran, M.; Dannenberg, J. J. J. Chem. Phys. 1996, 105, 11024−11031. (21) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553−566. (22) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735−746. (23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; et al. Gaussian 03, revision C 02; Gaussian Inc.: Pittsburgh, PA, 2004. (24) Grabowski, S. Annu. Rep. Prog. Chem. Sect. C 2006, 102, 1−37.
Table 2. Bonding Energies, BE, for Studied Cave−Template Complexes Calculated in a Gas Phase and Selected Solventsa cave−template
Article
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Phone: 601-446-1234. Fax:601446-1296. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This project is supported by the U.S. Department of Defense through the Engineer, Research and Development Center (Vicksburg, MS), Contract #W912HZ-10-C-0107, and The Mississippi Center for Supercomputing Research (MCSR). 1534
dx.doi.org/10.1021/jp2124839 | J. Phys. Chem. A 2013, 117, 1531−1534
