Study of Prepolymerization Complex Formation in the Synthesis of

Article pubs.acs.org/ac

Study of Prepolymerization Complex Formation in the Synthesis of Steroid-Based Molecularly Imprinted Polymers Corinne Sanglar,*,† Tim Jansen,‡ Marius C. Silaghi,‡ Julien Mernier,† Pierre Mignon,‡ and Henry Chermette‡ †

Département Service Central d’Analyse, Université de Lyon, Institut des Sciences Analytiques, UMR 5280 CNRS, Université Lyon1, 5 rue de la Doua, 69100 Villeurbanne, France ‡ Département Laboratoire des Sciences Analytiques, Université de Lyon, Institut des Sciences Analytiques, UMR 5280 CNRS, Université Lyon1, 5 rue de la Doua, 69100 Villeurbanne, France ABSTRACT: The preparation of steroid-based molecularly imprinted polymers (MIPs) based upon noncovalent interaction is particularly suited for selective capture of steroid hormones in biological and environmental samples. The success of this method lies in the optimization of the interaction between steroids (template) and methacrylic acid (functional monomer) in the prepolymerization mixture. NMR techniques coupled with DFT calculations were used to evaluate the capacity of the methacrylic acid to bind a steroid for future applications. The androstane derivative steroids considered in the present study have two functional groups at C3 and C17, which may interact with the methacrylic acid. These can be hydroxyl or ketone groups. Experimental results show that the steroids can be divided in three groups corresponding to the ketone type at C3, the H-bond strength increasing with the number of double bonds. DFT calculations are in very good agreement with experimental results, showing increasing binding energy from no bonds, a single bond, and two double bond steroids. For steroids holding a hydroxyl group the binding energy obtained in the solvent model is comparable to the binding energy of single bond ketone steroids.

T

matrices and preliminary to the instrumental analysis, needs to be developed. To simplify the cleanup step, interest in green microextraction and molecular imprinting is growing during this past decade. These techniques are particularly suited to the selective capture of small molecules in complex matrices9 and steroid hormones in biological and environmental samples.10 Molecular imprinting, through specific binding sites and chemical functions positioned in the cavities of cross-linked material, allows trapping a chemical compound and its structural homologues. The synthesis of molecularly imprinted polymers (MIPs) is based on radical polymerization reactions between a monomer and a dimer usually of vinylic nature. First, the monomer develops a noncovalent interaction with a functional group of the target molecule to form a complex. In the next step the dimer called “crosslinker” is fixed to the complexed monomer via polymerization. Thus, a spatial organization around the template is formed. For the implementation of the MIP the template has to be removed. The obtained cavities correspond to the template regarding the size, shape and arrangement of functional groups.11 The present work aims to study monomer-template interactions and affinities by using experimental techniques

he last introduced European legislation dealing with water quality aims to reach a good chemical and ecological status in water by 2015. Therefore one needs reliable analytical methods to ensure the monitoring of waters. In addition to conventionally monitored substances (e.g., metals, organochlorines, pesticides, hydrocarbons), other chemicals such as antibiotics, hormones or steroids, used for pharmaceutical products are under suspicion of affecting the aquatic ecosystem. These substances can be classified as emerging contaminants as there is little data reported onto their effect in the aquatic environment.1 Among these emerging pollutants, steroids include pharmaceutical hormones used for contraception (estrogenic and progestin) and nonpharmaceutical hormones derived from testosterone used as anabolic steroids.2 These steroids are known to affect the endocrine function and thus identified as endocrine disruptors. They are already found in fresh water, surface water, and groundwater.3−5 In another area, the illegal use of corticosteroidal hormones as veterinary drugs leads to the presence of residue in food.6 This is the result of misuse both in livestock and humans (doping). Some substances like clenbuterol7 are prohibited in the European Union.8 The presence of emerging contaminants like endocrine disruptors in water and in meat products is also a public health problem. Thus, monitoring of food and water quality is one of the most important issues of analytical chemistry. In that sense a selective and rapid extraction method, well suited to complex © 2012 American Chemical Society

Received: February 1, 2012 Accepted: April 12, 2012 Published: April 12, 2012 4481

dx.doi.org/10.1021/ac3003159 | Anal. Chem. 2012, 84, 4481−4488

Analytical Chemistry

Article

Figure 1. (a) Structural formula of androstane and (b) Structural formulas of androstane derived steroids divided in three classes depending on the substituent at C3 and its adjacent double bonds.

double bonds and the degree of conjugation (Figure 1). Methacrylic acid (MAA) was also purchased from SigmaAldrich France and used as received. Deuterated chloroform, acetonitrile and toluene were purchased from Euriso-Top France. NMR Measurements: All NMR spectra were recorded on a Bruker Avance400 spectrometer equipped with a QNP probe 5 mm at 30 °C. The proton spectra were acquired with a pulse angle of 30°, a spectral width of 6082 Hz and 32768 data points. The stoichiometry of the complexes was determined by Job’s plot analysis14 in deuterated chloroform, acetonitrile and toluene. The ratios of MAA to steroid were systematically varied using equimolar solutions (0.05 M) with a constant volume sample of 700 μL. 1HNMR was used to study the Hbonding between steroids (Figure 1) and the monomer MAA in deuterated chloroform, acetonitrile and toluene using an analysis volume of 700 μL. The samples were prepared with a fixed concentration of steroid (7 mmol L−1) and various concentrations of MAA (0−57 mmol L−1). At least, 8 different MAA/steroid ratios were analyzed for each interaction from 1:0 to 1:8. Computational Details. All calculations were performed using density functional theory methods provided by the Amsterdam Density Functional program (ADF2010).15 As a general protocol geometry optimizations were performed using

combined with molecular modeling methods. This study will help to design a new structural candidate able to build steroidbased molecular imprinted polymer. We investigate the binding characteristics of the complex formed between different steroids and methacrylic acid (MAA), a monomer widely used in the literature for molecularly imprinted polymers.12 Experimental results from NMR spectroscopy and binding energies obtained from quantum chemical calculations will be compared to understand the binding properties in prepolymerization complexes. In that way, although rather small chemical shifts are observed, the complementary combination of experimental and theoretical techniques constitutes a state of the art approach that will help to assess the binding properties of the considered steroids.

■

MATERIALS AND METHODS Experimental Section. Within the field of molecular imprinting, NMR spectroscopy is widely used to study non covalent interactions such as hydrogen bonds. 1H NMR spectroscopy allows the prediction of the molecular ratio template to methacrylic acid using the Job’s method,13 allowing to identify out the number of interaction sites. Materials: All steroids used in this work were purchased from Sigma-Aldrich France. They are classified according to their chemical structure (androstane based), the number of 4482



Article

Figure 2. 1H-NMR spectrum displaying the downfield shift as a function of the steroid/MAA ratio black 1:0, red 1:4; (a) proton at C3 of ADT, (b) proton at C17 of TEST, and (c) proton at C4 of TEST.

the B3LYP16 functional in combination with a triple-ζ polarized basis set, built from Slater type orbitals. Optimizations were carried out both in the gas phase and in the polarized continuum model to treat solvent effects, the COSMO approach17−19 was used for this task. Binding energies were computed as the difference between the energy of the complex and the energy of each molecule. Each complex has been optimized separately, that is, complexes with two MAA as well as those with one. Net atomic charges on the oxygen atoms, q(O), of the steroids have been computed from a Mulliken population analysis20 on the isolated steroid. As well, the local softness,21,22 s−(O), relative to the basic character of those oxygen atoms, has been computed from the product between the softness of the molecule and the Fukui function, f−(O), obtained from a finite difference between the charge of the oxygen atom in the neutral molecule and the charge on the oxygen atom in the positively charged molecule. This approach is used to gauge the Lewis basicity of atoms in molecules as it has already been

successfully used in numerous previous studies.23,24 It is wellknown that the “chemical accuracy” is still a goal of quantum chemistry in its DFT formulation. Many benchmarks of several modern exchange-correlation functionals do indeed provide an accuracy amounting a few kJ.mol−1, i.e. still larger than the aim of 1 kJ.mol−1.25 However, the DFT calculations are very robust and do provide significant energy differences provided that a given basis set is selected, coupled to a fixed exchangecorrelation functional, and that the calculation wears on a given type (class) of molecules.26 This is typically the case of the present work, where all the molecules are androstane steroid derivatives, of similar size, interacting with the same MAA monomer. In these conditions, relative energy differences can be trusted if they are larger than 2 kJ·mol−1.

■

RESULTS AND DISCUSSION Spectral Analysis of Steroid−MAA Interaction in Prepolymerization Solutions. The chemical structure (see

4483



Article

Figure 1) of the steroids studied in this work suggests two possible interaction sites because the presence of either two hydroxyl groups or a hydroxyl and a carbonyl group, on rings A and D, available to form H-bonds with the MAA. Most MIPs are synthesized in aprotic solvents such as toluene, chloroform and acetonitrile, favoring hydrogen bonding interactions. Hence, the stoichiometry of the complexes using Job’s plot analysis was determined in all solvents to confirm the number of interaction sites. The Job’s plot analysis in toluene and chloroform resulted in a maximum at 0.65 suggesting a stoichiometry steroid:MAA of 1:2, whereas in acetonitrile a maximum at 0.50 was obtained, suggesting a stoichiometry of 1:1. Chloroform and toluene seem to favor two interaction sites relative to acetonitrile. Therefore, complexes obtained in these solvents may offer better stability of the prepolymerization mixture leading to enhanced recognition properties. All steroids studied in this work are derived from the same chemical structure. Depending on the chemical groups at C3 and C17, the change of the proton’s chemical shift at these sites can be monitored as a function of the steroid:MAA ratio in the complex. Figure 2 shows the superposition of 1H NMR spectra of steroids and MAA in chloroform with ratios of 1:0 and 1:4, respectively. One can see a peak shift corresponding to the protons located at C3 and C17 as well as those located at C4, adjacent to a ketone function on C3. One can see a downfield shift (3.53 Hz) corresponding to the proton at C3 of androsterone (ADT) (Figure 2a) which demonstrates a Hbond formation between the ADT hydroxyl group and the MAA. In this interaction both the hydroxyl group of ADT and the carboxylic acid function of MAA can play the role of Hbond acceptor and donor, although the (ADT)−OH···OC− (MAA) is probably the strongest H-bond. The same interpretation can be drawn for the proton at C17 of testosterone (TEST) (Figure 2b). The downfield shift is about 2.63 Hz. To gauge the interaction force of the H-bond between the MAA and the carbonyl function of TEST at C3, the change in the chemical shift (3.36 Hz) of a proton at C4 of TEST is considered (Figure 2c), since no proton is present at C3. As a whole these results confirm that TEST and more generally all considered steroids display two possible interaction sites with the MAA. Figure 3 shows the change of chemical shift corresponding to the 1H NMR signal at C4 for testosterone measured in acetonitrile, chloroform and toluene. The largest chemical shift change corresponding to the strongest MAA···steroid interaction, occurs in toluene then follows chloroform and acetonitrile. One can notice that the less polar is the solvent the strongest is the interaction. Indeed since a polar solvent may engage H-bond with the steroid, it may compete with the MAA to bind one of the two sites of the steroids, that is, at C3 or C17, and prevent the MAA···steroid interaction. Subsequently, the titration of different steroids with MAA was performed in chloroform followed by 1H NMR in order to estimate the influence of the nature of a functional group at C3 and C17 regarding the interaction forces between a monomer and a template. Figure 4a shows, as expected, a nearly identical change in the chemical shift at the C3 carbons of ADT and dihydroandrosterone (DHA) during the titration. The curves in Figure 4b show that the four steroids: dihydrotestosterone (DHT), TEST, boldenone (BOLD), and DHA, display similar interaction forces with the MAA at C17. All these steroids all hold a hydroxyl group at C17, and as could be expected the Hbond between MAA and the hydroxyl group is of similar

Figure 3. Increasing variation of the chemical shift of a proton at C4 for TEST as a function of the MAA:TEST ratio in three different porogenic solvents: In the most polar solvent, that is, acetonitrile, the variation of the chemical shift is the less affected compared to the less polar solvent, that is, toluene, explaining the increased H-bond formation in the last mentioned.

strength and is neither related to the nature of the functional group nor the presence of double bonds at C3. Figure 4c focuses on the force of the interaction between the MAA and the ketone function at C3 by measuring the change in the chemical shift of a proton at C4. Three groups of steroids can be identified related to the presence of double bonds on ring A. Steroids without double bond on ring A (DHT) show a lower chemical shift than steroids with one double bond on ring A (TEST, 17α-methyltestosterone (MTEST), androstenedione (ANDRO)), which is also lower than those with two double bonds on ring A (BOLD, methandienone (METH)). One can conclude that the presence and the number of double bond on ring A increase the molecular recognition properties of steroids in molecularly imprinted polymer applications. Baggiani27 et al. obtain insights on the steroidal structural motifs able to increase or decrease the molecular recognition of corticosteroids. Their study indicates that the presence of a double bond on the steroidal ring A increases the molecular recognition toward corticosteroids through the presence of double bond which forces a planar configuration of ring A of steroids. Whatever the interaction site studied the effect on chemical shifts is weak (2− 7 Hz). If one may observe some trends about steroid groups with chemical shift in NMR study, these results have to be confirmed by calculations of high level. Computational Results. Binding energies have been computed at the DFT level in gas phase and in a continuum model (Table 1) to treat solvent effects for the steroids interacting with one MAA at C3 or C17 or with two MAA at each side. MAA···Steroid Interaction at C3. Gas phase calculations for ADT and DHA holding a hydroxyl group at C3 show similar binding energies of 14.8 and 14.7 kcal/mol, respectively. However for DHT, which holds a carbonyl group on the saturated A ring the binding energy decreases to 13.0 kcal/mol. If one considers the steroids with only one double bond on ring A, such as TEST, MTEST, and ANDRO, the binding energies at C3 are larger, and lie between 13.8 and 14.1 kcal/mol. For BOLD and METH presenting two double bonds the binding energies increases up to 14.8 and 14.6 kcal/mol, respectively. These latters show comparable binding energy to those obtained for the steroids holding a hydroxyl group (ADT, DHA). Therefore the steroids can be categorized in three 4484



Article

groups according to their interaction force with MAA as a function of the functional group present at C3 as follows: ketone < single-conjugated ketone < double-conjugated ketone ≈ hydroxyl groups. The energy difference between each group is about 1 kcal/mol. The steroids holding a hydroxyl group exhibit the largest interaction strength together with those with a double-conjugated ketone on ring A, which is expected since the hydroxyl group accepts but also gives a H-bond to the MAA. The binding energies in solvent show more or less the same trends: the ketone binding energy for DHT is 8.1 kcal/mol, followed by the single-conjugated ketones TEST, MTEST, and ANDRO at 9.3, 9.3, and 9.2 kcal/mol together with the hydroxyl groups of ADT and DHA at 9.1 and 9.2 kcal/mol respectively. With 9.6 kcal/mol, the largest binding energy is found for the double conjugated ketone groups of BOLD and METH. The inclusion of the solvent effects to model chloroform decreases the interaction strength although the polar groups become more polarized. In fact, the steroid···MAA interaction is now in competition with the interaction between the MAA or the steroid and the solvent continuum. One notices also that the binding energy between the MAA and the steroids holding an hydroxyl group (ADT and DHA) is now, with the inclusion of a solvent continuum, lower than the double-conjugated ketone systems (BOLD and METH). In comparison with NMR results, the three groups observed in Figure 4c corresponding to the ketone, single-conjugated ketone and double-conjugated ketone systems are well represented by the values of the binding energies computed here, with and without solvent effects. The acidic function of MAA at C3 is almost coplanar to the conjugated carbonyl system of BOLD and METH (Figure 5). This allows a well stabilizing electrostatic interaction between the hydrogen of the steroid and oxygen atom of the acid to be formed between a CH of the steroid. The distance MAAO···HC−steroid is 2.35 and 2.30 Å for BOLD and METH, respectively, in the C3 complexes including solvent effects. The MAA−OH···O− steroid distances given in Table 2 are almost identical for all steroids except for DHT with the nonconjugated ketone function for which the distance is greater. In Table 2, the net atomic charges, q(O), as well as the local softness relative to the Lewis basicity, s−(O), of the oxygen atoms of the steroid are shown. For gas phase calculations, when looking at the oxygen located at C3, the three groups of ketone are well divided according to the sequence found experimentally: DHT (q(O) = −0.523 au)< TEST, MTEST, ANDRO (q(O) ≈ −0.537 au) < BOLD, METH (q(O) ≈ −0.545 au). The more negatively charged on oxygen atom of the isolated steroid is, the stronger it may interact with the MAA hydrogen atom. The charge on the oxygen atom of the hydroxyl group at C3 is the largest for ADT and DHA, displaying binding energies as large as that of BOLD and METH. Including solvent effects in the calculations, lets observe the same sequence for the ketones. As expected, the atomic charge is greater due to the effects of the continuum, while the net atomic charge of the hydroxyl is now slightly lower than that of the single-conjugated ketone. This is also in good agreement with the binding energy differences found for the C3 complexes ΔEC3 (Table 1) for which the interaction strength for the steroid carrying a hydroxyl group is comparable to that of the single-conjugated ketone steroids. A similar conclusion can be drawn from the local softness values for the three different ketone types, except for ANDRO which is the only one carrying a ketone group at

Figure 4. Increasing variation of the 1H chemical shift with increasing MAA/steroid ratio (M) in chloroform: (a) proton at C3, (b) proton at C17, and (c) proton at C4.

Table 1. Binding Energies in (kcal/mol) Computed in Gas Phase and in a Solvent (Chloroform) Obtained from Optimized Structuresa gas phase ADT DHA DHT TEST MTEST ANDRO BOLD METH

solvent

ΔEC3

ΔEC17

ΔE2MAA

ΔEC3

ΔEC17

ΔE2MAA

14.8 14.7 13.0 14.0 14.1 13.8 14.8 14.6

11.2 14.9 13.7 14.8 13.9 12.4 14.8 14.1

26.1 29.6 26.9 28.7 28.1 26.2 29.4 28.7

9.1 9.2 8.1 9.3 9.3 9.2 9.6 9.6

7.3 8.8 8.7 8.7 7.9 7.4 8.7 8.3

16.4 17.9 16.6 17.9 17.3 17.4 18.1 17.8

ΔEC3 is the binding energy between the steroid and one MAA interacting via an H-bond at C3, equivalently for ΔEC17 at C17, ΔE2MAA corresponds to the binding energy between a steroid and two MAA forming H-bonds on both sides (C3 and C17). a

4485



Article

Figure 5. Optimized structures of DFT calculations using the COSMO solvent model for androstane derived steroids forming an interaction with two MAA at C3 and C17.

Table 2. Geometric Parametersa gas phase d(O···H) 2MAA ADT DHA DHT TEST MTEST ANDRO BOLD METH

d(O···H) 1MAA

C3

C17

C3

1.73/1.93 1.71/1.96 1.75 1.70 1.70 1.72 1.69 1.70

1.75 1.72/1.91 1.69/2.07 1.74/1.89 1.74/1.92 1.71 1.74/1.88 1.74/1.94

1.72/1.93 1.69/2.00 1.74 1.70 1.70 1.71 1.70 1.70

d(O···H) 2MAA ADT DHA DHT TEST MTEST ANDRO BOLD METH

q(O) isolated steroid C17

1.74 1.72/1.94 1.71/2.07 1.74/1.89 1.74/1.92 1.78 1.74/1.87 1.73/1.94 solvent

d(O···H) 1MAA

s−(O) isolated steroid

C3

C17

C3

C17

−0.556 −0.556 −0.523 −0.537 −0.538 −0.535 −0.545 −0.545

−0.518 −0.566 −0.563 −0.562 −0.557 −0.513 −0.561 −0.555

0.51 0.46 1.72 2.33 2.24 1.95 2.58 2.48

1.74 0.78 0.90 0.96 1.01 1.63 0.93 1.00

q(O) isolated steroid

s−(O) isolated steroid

C3

C17

C3

C17

C3

C17

C3

C17

1.69/2.00 1.66/2.10 1.71 1.66 1.67 1.68 1.68 1.67

1.71 1.70/1.98 1.67/2.28 1.73/1.96 1.72/1.96 1.69 1.73/1.94 1.73/1.99

1.69/2.02 1.66/2.19 1.72 1.67 1.68 1.67 1.66 1.66

1.70 1.71/1.99 1.66/2.39 1.72/1.96 1.73/1.98 1.75 1.73/1.92 1.73/2.02

−0.616 −0.615 −0.603 −0.626 −0.627 −0.624 −0.638 −0.638

−0.594 −0.621 −0.620 −0.618 −0.611 −0.589 −0.618 −0.608

0.12 0.36 2.40 3.51 3.24 1.79 3.96 3.89

2.58 0.76 0.55 0.39 0.52 2.09 0.36 0.53

a

Distances (in Å) between a steroid oxygen atom and a hydrogen atom of MAA, that is, steroid−O···H−MAA (first value) and between a steroid hydrogen atom and an oxygen atom, that is, steroid−H···O−MAA (second value, only present if the steroid contains an hydroxyl group) for complexes where the steroid interacts with one (1MAA) or two MAA (2MAA) at C3 or C17. The net atomic Mulliken charge q(O)on the oxygen atom of the steroid at C3 or C17 is given in au from a Mulliken population analysis. The local softness s−(O), relative to the basic character of the oxygen atom at C3 or C17 is also given in au. Net atomic charges have been computed from the isolated steroids. 4486



Article

for future applications. Androstane derived steroids considered in the present study present two important binding sites, that is, the C3 and C17 carbon atoms, possessing an hydroxyl or ketone group implicated in the interaction with methacrylic acid. Experimental results lead to the conclusion that steroids can be divided in three groups depending on the functional group at C3 and its adjacent double bonds. This functional group being a ketone at C3 is part of a six-membered ring having none, one or two neighboring double bonds. An important finding is the increasing H-bond strength with increasing number of double bonds around the carbonyl group at C3. Calculated interaction energies, using density functional theory, are in very good agreement with 1H NMR shifts obtained from experimental results. Both approaches display an increasing binding energy following the sequence no double bond, one double bond and two double bonds on the ring carrying the ketone. Net atomic charges on the steroidal oxygen atoms as well as Fukui functions relative to the oxygen Lewis basicity confirmed this trend. For steroids holding a hydroxyl group at C3 the binding energy obtained via the COSMO solvent model is comparable to that of ketone groups at C3 with a vicinal single bond. In addition 1H NMR measurements show similar H-bond strengths for steroid bearing a hydroxyl group at C17. This is fully confirmed by the binding energies computed with DFT including solvent effects. Moreover, DFT calculations showed that the presence of a methyl at C17 may contribute to a slight lowering of the H-bond interaction with methacrylic acid. This works underlines the profit to couple NMR techniques with DFT calculations when small effects are analyzed.

C17, probably leading to a different charge distribution for the cationic form of the molecule. One may observe that the values of the local softness computed in Table 2 reproduce even better the trend observed from NMR shifts and binding energies. Nevertheless, the strength of H-bonding predicted from the atomic charges and local softness values are in good agreement with NMR experimental results for gas phase as well as solvent calculations. However it is more difficult to compare the results of hydroxyl group with those of ketones which present different electronic distribution behavior. MAA···Steroid Interaction at C17. Concerning the Hbonding capacity at C17 site, NMR experiments could only be performed for DHA, DHT, TEST, and BOLD. The results shown in Figure 4b reveal that these steroids are involved in an H-bond with the MAA of a similar strength, which could be expected since all these steroids hold an hydroxyl group. Looking at the results obtained from gas phase calculations (ΔEC17 in Table1), the binding energies of these steroids are all around 14.8 kcal/mol (except for DHT), while the interaction strength for the other steroids is lower by almost 1 kcal/mol. The results of the calculations including solvent effects show a similar trend. DHT displays now identical binding energy as DHA, TEST, and BOLD of about 8.7−8.8 kcal/mol, while the binding energies of all remaining steroids are lesser. When looking to atomic distances, no clear trend can be observed. All steroids possessing a hydroxyl group show steroid−(OH)···OMAA distances around 1.72−1.74 Å except for DHT which has a shorter H-bond. For the latter, the H-bond donated by the hydroxyl group in return is longer than that of the other steroids, which might thus compensate. Looking to the charge on the oxygen atom at C17, one can see that they are similar for DHA, DHT, TEST, and BOLD. The net atomic charge of the oxygen is lower for all remaining steroids in gas phase (q(O) ≈ −0.563 au) as well as in the solvent continuum (q(O) ≈ −0.619 au). In that case however, the values of the local softness do not confirm the H-bonding ability of the hydroxyl oxygen atoms as clearly as the net atomic charge. One may however remark that the values of the local softness of the oxygen of MTEST and METH are close to those of DHA, DHT, TEST, and BOLD as compared to ADT and ANDRO. This is rather reasonable since the formers possess a hydroxyl group at C17 and the two latters a ketone group. That seems to decrease the negative charge on the oxygen, as it can be seen from the corresponding computed binding energy at C17. Global MAA···Steroid interaction. From the energetical point of view, the interaction through C3 or C17 does not seem to change in the presence of two MAA. In fact, the energy difference between the sum of the isolated interactions between one MAA and the steroid and the global interaction between the two MAA and the steroid is rather small. This can also be seen from the geometric parameters which only marginally change with respect to the inclusion of a second MAA. As a result, the largest interaction are found for the steroid holding a hydroxyl group at C17 and another hydroxyl group at C3 or a double-conjugated ketone on ring A, that is, BOLD, METH, DHA, and exceptionally TEST that only has a singleconjugated bond on ring A.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS P.M and H.C gratefully acknowledge the GENCI/CINES for HPC resources/computer time (Project cpt2130). REFERENCES

(1) Amalric, L. Composés perturbateurs endocriniens et autres molécules organiques émergentes dans les eaux souterraines: état des lieux; BRGM/ RP-54484-FR, 69 p, 14 ill., 2 ann., 2006. (2) Richardson, S. D. Anal. Chem. 2002, 74, 2719. (3) Benito-Pena, E.; Urraca, J. L.; Sellergren, B.; Cruz Moreno-Bondi, M. J. Chromatogr. A 2008, 1208, 62. (4) Le Noir, M.; Lepeuple, A.-S.; Guieysse, B.; Mattiasson, B. Water Res. 2007, 41, 2825. (5) Yang, M.; Gu, W.; Sun, L.; Zhang, F.; Ling, Y.; Chu, X.; Wang, D. Talanta 2010, 81, 156. (6) Courtheyn, D.; Le Bizec, B.; Brambilla, G.; De Brabander, H. F.; Cobbaert, E.; de Wiele, A. V.; Vercammen, J.; De Wasch, K. Anal. Chim. Acta 2002, 473, 71. (7) Blomgren, A.; Berggren, C.; Holmberg, A.; Larsson, F.; Sellergren, B.; Ensing, K. J. Chromatogr. A 2002, 975, 157. (8) Council directive 88/146/EEC, 1988. (9) Kloskowski, A.; Pilarczyk, M.; Przyjazny, A.; Namiesnik, J. Crit. Rev. Anal. Chem. 2009, 39, 43. (10) Aufartova, J.; Mahugo-Santana, C.; Sosa-Ferrera, Z.; Juan Santana-Rodriguez, J.; Novakova, L.; Solich, P. Anal. Chim. Acta 2011, 704, 33. (11) Arshady, R.; Mosbach, K. Macromol. Chem. Phys. 1981, 182, 687.

■

CONCLUSION NMR techniques combined with DFT calculations were used to evaluate the capacity of the methacrylic acid, a crucial component for molecularly imprinted polymers, to bind steroid 4487



Article

(12) Turiel, E.; Martin-Esteban, A. Anal. Chim. Acta 2010, 668, 87. (13) Cruz, J. R.; Becker, B. A.; Morris, K. F.; Larive, C. K. Magn. Reson. Chem. 2008, 46, 838. (14) O’Mahony, J.; Molinelli, A.; Nolan, K.; Smyth, M. R.; Mizaikoff, B. Biosens. Bioelectron. 2005, 20, 1884. (15) Baerends, E.J.; J. A., Bashford, D.; A. Bérces, Bickelhaupt, F.M.; Bo, C.; Boerrigter, P.M.; Cavallo, L.; Chong, D.P.; Deng, L.; Dickson, R.M.; Ellis, D.E.; M. van Faassen, Fan, L.; Fischer, T.H.; C. Fonseca Guerra, Ghysels, A.; Giammona, A.; S. J. A. van Gisbergen, A. W. Götz, Groeneveld, J.A.; Gritsenko, O.V.; M. Grüning, Harris, F.E.; P. van den Hoek, Jacob, C.R.; Jacobsen, H.; Jensen, L.; G. van Kessel, Kootstra, F.; Krykunov, M.V.; E. van Lenthe, McCormack, D.A.; Michalak, A.; Mitoraj, M.; Neugebauer, J.; Nicu, V.P.; Noodleman, L.; Osinga, V.P.; Patchkovskii, S.; Philipsen, P.H.T.; Post, D.; Pye, C.C.; Ravenek, W.; J. I. Rodríguez, Ros, P.; Schipper, P.R.T.; Schreckenbach, G.; Seth, M.; Snijders, J.G.; M. Solà, Swart, M.; Swerhone, D.; G. te Velde, Vernooijs, P.; Versluis, L.; Visscher, L.; Visser, O.; Wang, F.; Wesolowski, T.A.; E. M. van Wezenbeek, Wiesenekker, G.; Wolff, S.K.; Woo, T.K.; Yakovlev, A.L.; , Ziegler, T.. ADF2010.02, SCM. Theoretical Chemistry; Vrije Universiteit: Amsterdam, The Netherlands, 2010; http://www.scm.com. (16) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623. (17) Klamt, A. J. Phys. Chem. 1995, 99, 2224. (18) Klamt, A.; Jonas, V. J. Chem. Phys. 1996, 105, 9972. (19) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin Trans. 2 1993, 799. (20) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1833. (21) Fukui, K. Science 1982, 218, 747. (22) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and Molecules; Oxford University Press: New York, 1989. (23) Chermette, H. J. Comput. Chem. 1999, 20, 129. (24) Geerlings, P.; De Proft, F.; Langenaeker, W. Chem. Rev. 2003, 103, 1793. (25) Lepetit, C.; Chermette, H.; Gicquel, M.; Heully, J.-L.; Chauvin, R. J. Phys. Chem. A 2007, 111, 136. (26) Bremond, E.; Pilard, D.; Ciofini, I.; Chermette, H.; Adamo, C.; Cortona, P. Theor. Chem. Acc. 2012, 131, 1184. (27) Baggiani, C.; Baravalle, P.; Giovannoli, C.; Anfossi, L.; Giraudi, G. Biosens. Bioelectron. 2010, 26, 590.

4488


Study of Prepolymerization Complex Formation in the Synthesis of

Recommend Documents