Evaluation of Molecularly Imprinted Thin Films for Ephedrine Recognition

ELENA B. STOICA, ANA-M. GAVRILA, CATHERINE BRANGER, HUGUES BRISSET, ANTON V. DYSHLYUK, OLEG B. VITRIK, HORIA IOVU, ANDREEA MIRON, ANDREI SARBU*, TANTA-V. IORDACHE* National Institute for Research&Development in Chemistry and Petrochemistry ICECHIM, Advanced Polymer Materials and Polymer Recycling Group, 202 Splaiul Independentei, 060021,Bucharest, Romania Laboratoire MAPIEM of Toulon University, SeaTech, Bât. X, Avenue de l'Université, 83130 LA GARDE FRANCE Institute of Automation and Control Processes (Far Eastern Branch of Russian Academy of Sciences) and Far Eastern Federal University, 5 Radio Str., Vladivostok, Russia University Politehnica of Bucharest, Faculty of Applied Chemistry and Materials Science, Advanced Polymer Materials Group, 1-7 Polizu, 011061, Bucharest, Romania

Microscopy and topography of films was assessed using digital microscopy -Hirox RH-2000 Digital, atomic force microscopy in non-contact mode -NT-MDT NTEGRA Probe NanoLaboratory system (NT-MDT NSG01 cantilever with tip radius of 10 nm) and scanning electron microscopy -Zeiss Supra 40VP.
The structure analysis was carried out using Nicolet iS50 FT-IR device with ATR -SMART iTR (Termo SCIENTIFIC) and thermal analysis was performed using TA Instrument SDT Q600 Thermo-Gravimetric Device.
For batch rebinding experiments Thermo Nicolette UV-VIS Spectrometer was used in order to determine the adsorption of ephedrine at the specific wavelength, λ=257 nm.

Preparation of film precursor solutions and film deposition
The films were synthesized by a sol-gel derived technique in basic conditions, according to a similar recipe developed by Florea et al. [51], which involved the hydrolysis of methoxy groups of the monomer, followed by their condensation for obtaining the rigid polymer network. For each sample, two solutions were prepared as follows: solutions A, which consists in dissolving in ethanol the monomer and the template, i.e. ephedrine hydrochloride, and solution B obtained from ammonium hydroxide and water. After the two solutions were homogenized separately, solution A was added over B to yield the MIP film precursor solution. After 2 h of mechanical stirring the precursor was casted on glass substrates by air-spraying ( Figure 1). The same procedure was used for obtaining the control films (NIPs) but without adding the template. Using this technique two film series with various thicknesses, namely MIP C/NIP C and MIP D/NIP D pairs (Table 1), were obtained by dilution of solution A. After deposition, the films were maturated at room temperature for 48 hours, then for another 48 hours at 80 º C in an oven. All MIP films were washed with ethanol to extract the ephedrine and to remove the unreacted monomer. In this step, the films were immersed in 100 mL of ethanol and placed in the ultrasonic bath for 4 hours at room temperature. To reduce errors, the same procedure was applied to the NIPs, as well.

Rebinding capacity and imprinting factor of films
The adsorption capacity is one of the most important characteristics of MIP sensitive films. Depending of this property the sensor is capable or not to recognize the template, in this case ephedrine. Therefore, batch rebinding measurements are highly useful to validate the specificity of MIP films by measuring the adsorbed amount of ephedrine relative to the NIP films. In this respect, dried MIP and NIP films were immersed in 40 mL of aqueous solution containing 2.74 mmoles of ephedrine/L. The quantification method referred, in this case, to measuring the initial and final concentration of the feed aqueous solutions (i.e. before and after contact with the films) using a UV-Vis Spectrophotometer. The specific parameters used to determine the adsorption performance were calculated using the following equations: Where: -Q is the binding capacity of films, (g of ephedrine/ g of film); - Where: -F is the imprinting factor, which quantifies the specificity of adsorption; - is the adsorption capacity of MIP; is the adsorption capacity of NIP.
Calibration was performed in the 5.48 mmoles/L-0.137 mmol/L ephedrine concentration range. The regression equation, Eq. (3), used to calculate the concentration of ephedrine was determined at the specific wavelength, λ=257 nm with a regression coefficient, r=0.982. Where: -A is the sample absorbance at λ=257 nm; -N c (g/L) is the concentrations of ephedrine in the analyzed sample;

Results and discussions Thickness assessment and optical proprieties
According to the synthesis recipes of films that use two different concentrations for the precursor solution, the final films after maturation should present either nano-or micron-sized thickness. Ellipsometry is one of the fewest methods to allow proper evaluation of thickness for transparent films. In this respect, ellipsometric data were fitted according to the Cauchy Model for transparent materials (given in Table 2), which allowed thickness evaluation of films. Thickness differences were observed for both the MIP D/ NIP D and MIP C/NIP C films as summarized in Table 2.  Figure 2 reveals the refractive index variation of the analyzed films obtained from the ellipsometric data modelling. As shown in Figure 2a and b, the refractive indexes were higher for the MIP films due to ephedrine presence in the films structure. Hence, both MIP film types presented optical response for the target molecule regardless of thickness. The refractive index difference, Δn, in the 600-1000 nm range (visible range) between MIP C and NIP C varied from 0.004 to 0.005, while for MIP D and NIP D was constant, Δn =0.003. This latter result indicated again that thinner films are more homogenous.
Further on, in Figure 2c and d, the transmittance profiles of MIP and NIP film pairs was evaluated. Herein, the glass substrate presented the highest transmission, over 92%, across the 350-900 nm wavelength range. Transmission of films was lower than that of glass alone, maximum being attained by NIP D with over 87% T, which is explainable due to its low thickness. Yet, the MIP D transparency was even lower, compared to that of MIP C and NIP C, indicating the formation of different networks due to monomer-template auto-assembly.

FT-IR analysis
All the films were evaluated by FT-IR and their spectra are shown in Figure 3. The characteristic band for -OH stretching (terminal groups of the organosilica network) was present in the range of 3260-3280 cm -1 . Methylene groups from the DAMO-T monomer were spotted around 2930 cm -1 . The stretching of Si-O-Si backbone was noticed in the 850-1108 cm -1 range. Deformation vibrations of the primary (-NH2-) and secondary (-NH-) amines were observed around 1650 cm -1 and 1470 cm -1 , respectively, and can be attributed to both the template (ephedrine) and the monomer. The presence of ephedrine in the MIP films (Figure3a and c) was attested by the shift of the NH2 band with around 10 nm; this being the functional group in the monomer/polymer responsible for template interactions via hydrogen bonds.

TGA analyses
For determining the thermal behavior, the MIP and NIP films were scratched off the glass substrate and analyzed by thermo-gravimetry (thermograms and derivatives depict in Figure 4.a-d). In the first interval, the mass loss of all films was associated with the loss of water and ethanol. In the 200-300 °C range, thermal degradation of ephedrine occurred (attaining a maximum at 220 °C), as observed from the thermogram of MIP C and MIP D (Figure 4.a and c). The next range valid for all films, 300-500 °C, can be attributed to degradation/fragmentation of the pendant groups (i.e. amino, amino-ethyl, amino-methyl, methyl and ethyl fragment types), and the last degradation step, between 500-800 °C, represented the degradation of the polymer backbone [51]. At 800 °C all films registered high amounts of residue, which also confirmed the good thermal stability of the films. Here it can be mentioned that the residue was higher for MIP D film (Figure 4.a), which indicated a more discrete auto-assembly between the monomer and the template at lower concentrations of precursors. Concluding, the monomer-template auto-assembly in the imprinting stage has led to more homogenous and compact structure of films.

Microscopy analysis
Besides the recognition properties for ephedrine, the sensor elements should also be continuous with no cracks (to have homogenous properties on bulk) and also to present a high surface/contact area to enhance the sensitivity of the future sensor. As a result, the films morphology was analyzed using digital, atomic force and electron microscopy. For discrete analysis of morphology on thin films SEM was very much suited. AFM was used to calculate changes in roughness and surface area variations due to differences in films deposition. The optical microscopy was also used to observe if the films were homogenous on larger surfaces after maturation; it is important for sensors applications to have a homogenous film with no cracks as the electrical properties and reproducibility are very much influenced by this factor. As it can be observed from the digital snapshots, MIP D and NIP D films ( Figure 5.c and d) present less asperities and bubbles than MIP C and NIP C ( Figure 5.a and b). Hence, the thin films were indeed more homogeneous as revealed by the thermal analyses, as well.   (Figure 6c and  d), random peaks can be observed on the surface, indicating either a non-homogenous deposition or the fact that ephedrine addition improved the films homogeneity for the following MIP C film (in Figure 6a and b), as it auto-assembles with the functional monomer. Nevertheless, at lower concentrations of precursor monomer, the imprinted and the control films, (Figure 7c and d, MIP D and NIP D, respectively) presented better thickness distributions and surface uniformity. Further on, SEM micrographs (Figure 8 a-d) were consistent with the topography assessment, revealing different surface morphologies of the MIP and NIP films compared with each other, but also compared from the thickness point of view. The surface of MIP D (Figure 8c) was more homogenous comparing to the MIP C and presented a tortuous aspect. The roughness of films (presented in Figure 9) pointed toward higher surface homogeneity of MIP films relative to their control pairs and also validates the unusual surface morphology of NIP C.

Batch Rebinding of Ephedrine
Finally, to evaluate the binding affinity for ephedrine of the MIP relative to the NIP films, rebinding experiments were performed. After drying, the films were weighed and immersed in a feed aqueous solution of ephedrine with known concentration. For MIP C films the concentration of ephedrine increased in the supernatant in the studied time range, which can be due to a phenomenon called leaching; this means that ephedrine was not entirely extracted from the MIP C film and it was desorbed during batch rebinding tests. Therefore, determining Q and F for this MIP C/ NIP C pair was not possible. However, MIP D film seemed to rebind ephedrine quantitatively with high specificity relative to the reference NIP D film. The concentration of supernatants was assessed by UV-vis measurements at 1 minute, 5, 10, 15, 20 minutes after contact (in duplicate) and the specific parameters, in terms of binding capacity, Q (µg ephedrine/ g of film), and imprinting factor, F, were determined and outlined in Table 3 and Figure 9b, respectively. http://www.revmaterialeplastice.ro The difference of binding capacity from MIP D to NIP D was approximately one order of magnitude and definitely due to the molecularly imprinted cavities formed during synthesis. At the same time, the imprinting factors, F, attained high values between 4.6 and 6.2 (maximum attained at 15 minutes) on the studied time range (Figure 9b). Thus, confirming the efficiency of the imprinting method, in terms of strong interactions between the MIP D film and ephedrine, fast and specific recognition and binding of ephedrine [40,51]. These results are by far superior relative to other literature reports [52][53][54]. For instance, Tian et al. [52] reported ephedrine binding capacities of approximately 95 µg/ g for MIP and 86 µg/ g for NIP, which resulted in a 1.1 imprinting factor.

Conclusions
In this study, ephedrine-MIP films were obtained by air-sparing on glass supports sol-gel precursors in order to develop sensitive elements for ephedrine recognition. Structural, thermal, optical and morphological assessments were complementary, helping toward revealing and explaining the discrete modifications of the films structure that occurred during molecular imprinting. According to the batch re-binding experiments, only the imprinted thin film, MIP D was performant, attaining a 10 times higher rebinding capacity relative to the control film. The recorded imprinting factor in the first minute of contact was above 5 and increased at 6.2 after 15 minutes, reveling fast and specific adsorption of ephedrine.