Scholarly article on topic 'Novel electro-polymerized protein-imprinted materials using Eriochrome black T: Application to BSA sensing'

Novel electro-polymerized protein-imprinted materials using Eriochrome black T: Application to BSA sensing Academic research paper on "Chemical sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Electrochimica Acta
OECD Field of science
Keywords
{"Eriochrome black T" / "Protein imprinting" / "Electrochemical impedance spectroscopy" / "Conducting polymer" / "Molecularly imprinted polymer"}

Abstract of research paper on Chemical sciences, author of scientific article — Ana P.M. Tavares, M. Goreti F. Sales

Abstract A novel material produced in-situ by electropolymerization of Eriochrome black T (EBT) is presented for the first time to produce a molecularly-imprinted polymer (MIP) tailored for protein recognition. This monomer is particularly useful because it contains in the same structure different functions that may interact with different sites within the same protein (by ionic interaction of hydrogen bonding). The polymer was poly(EBT) (PEBT) and was obtained by applying on a carbon support a suitable range of potential values, established by cyclic voltammetry (CV) along consecutive cycles. In a parallel approach, the carbon support was modified by electropolymerizing 3,4-ethylenedioxythiophene (EDOT) prior to the MIP synthesis, thereby yielding a substrate of better electrical properties and checking this effect upon the resulting biosensor. The two above approaches used BSA as model target protein. The polymeric material acted as a plastic antibody for BSA and was obtained through a bulk imprinting strategy, by electropolymerizing EBT in a solution that also contained the target protein. The chemical features were followed by Raman spectroscopy while the electrical properties were followed by electrochemical impedance spectroscopy (EIS). The electrical properties tested were the stability of polymeric film within time, the main analytical features of the calibration curves under different media and the selectivity properties. The thermal stability was also tested by thermogravimetric assays. Overall, the novel polymeric film displayed good thermal and storage stabilities, which are fundamental features in biosensor development. Both MIP and MIP-PEDOT displayed linear responses over a wide range of concentrations and similar detection limits. The MIP-PEDOT material was 9 × more sensitive to the presence BSA concentration. The analytical responses of the biosensor to spiked serum confirms the promising features of the described approach.

Academic research paper on topic "Novel electro-polymerized protein-imprinted materials using Eriochrome black T: Application to BSA sensing"

Accepted Manuscript

Novel electro-polymerized protein-imprinted materials using Eriochrome black T: Application to BSA sensing

Ana P.M. Tavares, M. Goreti F. Sales

PII: S0013-4686(17)32765-2

DOI: 10.1016/j.electacta.2017.12.191

Reference: EA 30979

To appear in: Electrochimica Acta

Received Date: 4 October 2017 Revised Date: 29 December 2017 Accepted Date: 31 December 2017

Please cite this article as: A.P.M. Tavares, M.G.F. Sales, Novel electro-polymerized protein-imprinted materials using Eriochrome black T: Application to BSA sensing, Electrochimica Acta (2018), doi: 10.1016/j.electacta.2017.12.191.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Novel electro-polymerized protein-imprinted materials using Eriochrome

Black T: application to BSA sensing

Ana P. M. Tavares, M. Goreti F. Sales BioMark-CINTEISS/ISEP, School of Engineering, Polytechnic Institute of Porto, Portugal

* To whom correspondence should be addressed: Goreti Sales, School of Engineering of the Polytechnique School of Porto, R. Dr. Antonio Bernardino de Almeida, 431, 4200-072 Porto, Portugal. Tel: +351228340544; Fax: +351228321159. goreti.sales@gmail.com; mgf@isep.ipp.pt.

Abstract

A novel material produced in-situ by electropolymerization of Eriochrome black T (EBT) is presented for the first time to produce a molecularly-imprinted polymer (MIP) tailored for protein recognition. This monomer is particularly useful because it contains in the same structure different functions that may interact with different sites within the same protein (by ionic interaction of hydrogen bonding). The polymer was poly(EBT) (PEBT) and was obtained by applying on a carbon support a suitable range of potential values, established by cyclic voltammetry (CV) along consecutive cycles. In a parallel approach, the carbon support was modified by electropolymerizing 3,4-ethylenedioxythiophene (EDOT) prior to the MIP synthesis, thereby yielding a substrate of better electrical properties and checking this effect upon the resulting biosensor.

The two above approaches used BSA as model target protein. The polymeric material acted as a plastic antibody for BSA and was obtained through a bulk imprinting strategy, by electropolymerizing EBT in a solution that also contained the target protein. The chemical features were followed by Raman spectroscopy while the electrical properties were followed by electrochemical impedance spectroscopy (EIS). The electrical properties tested were the stability of polymeric film within time, the main analytical features of the calibration curves under different media and the selectivity properties. The thermal stability was also tested by thermogravimetric assays.

Overall, the novel polymeric film displayed good thermal and storage stabilities, which are fundamental features in biosensor development. Both MIP and MIP-PEDOT displayed linear responses over a wide range of concentrations and similar detection limits. The MIP-PEDOT material was 9x more sensitive to the presence BSA concentration. The analytical responses of the biosensor to spiked serum confirms the promising features of the described approach.

Keyword: Eriochrome Black T; Protein imprinting; Electrochemical impedance spectroscopy; conducting polymer; molecularly imprinted polymer.

1. Introduction

Biosensor technology has emerged in recent decades, addressing many challenges in different areas, such as the environment, agriculture, biotechnology and health [1, 2]. Despite the great diversity of approaches when conceiving a biosensor, it always combines a biorecognition element capable of interacting with a given target analyte and a suitable transduction mode that transforms this interaction into a measurable event. The main objective of the biorecognition element is to provide the sensor the ability to interact with the target analyte and distinguish it from other species that may be present in the sample. Proteins, antigens, antibodies, DNA, RNA, membrane receptors or nucleic acids are amongst the several compounds used as biorecognition elements.

The most common biosensors published are immunosensors, where the biorecognition element is an antibody. Their mode of operation is inspired in the immune system. The antibody captures the target analyte to which it was developed, and thereby generates a change in the transduced signal. Some immunosensors are also linked to enzymes, where their catalytic response contributes to additional chemical alterations on the sensing layer [3]-[4]. Despite their great selectivity, these use of antibodies in biosensors, combined or not with enzymes, also justifies the lack of commercial use of these immunosensors. They offer little chemical stability, displaying high sensitivity to pH and temperature that may lead to protein denaturation and biological inactivity [5]. Their cost is high and the response mostly irreversible, which leads to expensive biosensors of single use [6, 7].

Aiming to solve the previous disadvantages, the use of synthetic biomimetic materials that mimic the biological activity of antibodies in biosensors has emerged along the last decade [8]. In this context, biomimetic materials of natural antibodies have been designed to act as their natural analogues, and may display similar or superior proprieties. These materials simulate the recognition process made by natural receptors, by holding selective binding sites within a polymeric matrix that have been moulded with the target molecule acting as template [9, 10]. This process is well known as molecular imprinting and the generated materials known as molecularly imprinted polymers (MIP).

In this, monomers and cross-linkers are polymerized around the target molecule; the subsequent extraction of the target molecule leads to the formation of vacant sites that hold the same shape as the target and provide a favourable electrostatic environment for rebinding. Overall, many biomimetic materials produced so far have displayed similar affinity and specificity features to their biological analogues, also having the advantages of low-cost, ability to be tailored on-demand, little storage requirements and long shelf life [11-13].

MIP materials may be produced by many approaches. The most simple and straightforward process includes polymerization in situ by free-radical polymerization. Polymerization in situ involves a chemical reaction between functional and cross-linking monomers that organized around a template (to be imprinted) and copolymerize after to form a highly cross-linked polymeric structure. This process may require several hours along with high temperature (60-80°) to initiate polymerization, and it has been proven effective in the preparation of selective MIPs for compounds of small molecular weight (~200-1200 Da) such as drugs or amino acids.

When the target template is sensitive to any of the above conditions, other approaches are required. This is the case of proteins. Besides their little stability under different pH and ionic strength conditions, they are mostly incompatible with high temperatures and their imprint should be carried out under experimental conditions that resemble their native environment. This is mostly difficult, also because most of the monomers/initiators employed in radical polymerizations have little water solubility. Besides, the large size of proteins hinders their extraction from the polymeric network and its subsequent rebinding [14].

As an alternative to the classical chemical imprinting process, electropolymerization has been successfully established for imprinting proteins [14, 15]. It is a suitable approach to generate plastic antibodies and to open new horizons into scaling-up processes of current "immunosensors". In electropolymerization, the application of a potential/current on a conductive substrate triggers and sustains the polymerization [16]. Such electro-synthesis creates thin-films that are few nanometers

thick and offers the advantage of controlling the rate of polymeric growth around the protein. Overall, the rate of initiation is controlled by the selected electrochemical parameters, because the species responsible for initiating the polymerization are formed according to the electrochemical parameters applied, mostly current density or potential. In addition, this approach does not require any external initiator, such as thermal or UV light. [14, 17], which also helps controlling the chemical composition of the final polymer. In turn, the conductive features of the final polymeric network are also relevant (especially when biosensors of electrical transduction are involved), which in turn depends of the monomers involved, the electrical conditions applied and the chemical composition of the medium where the polymerization takes place.

Overall, the use of electropolymerization in protein imprinting requires the selection of suitable monomers that are able to undergo electro-oxidation or electro-reduction under specific electrochemical parameters, as long as the target protein remains stable under the electrochemical conditions selected. Usually, most processes described in the literature involve electro-oxidation polymerizations and the oxidation of aromatic compounds. Monomers used for the purpose of small molecule imprinting include ethylenedioxythiophene (EDOT) [18], pyrrole (Py) [19], aniline [20], phenol [17], aminophenol [21] and phenylenodiamine [22]. In turn, monomers as o-aminophenol [21] and aniline boronic acid [23] have been employed for protein imprinting in electrically generated polymers [24, 25]. Thus, it is easy to conclude that the selection of monomers for electropolymerization is rather limited and that each monomer used so far holds a limited number of functionalities, which in turn may limit the electrostatic recognition of the protein upon rebinding.

Thus, this paper aims to introduce a novel monomer that displays proper features for electropolymerization and its application in protein imprinting. The monomer selected for this purpose is Eriochrome black T (EBT). EBT has been used to generate polymers by an electrical stimulus, [26] [27], [28] but never with the purpose of molecular imprinting. Herein, EBT may lead to the production of excellent complementary rebinding sites because it has different chemical

functions within the same structure (nitro, sulfonic and hydroxyl), which may interact easily with proteins, also containing a wide diversity of functionalities. As model target protein, Bovine serum albumin (BSA) was selected. It is a low cost protein, with similarities with the Human serum albumin, [29-31] involved in many physiological processes, including the regulation of blood osmotic pressure and pH [32]. The performance of the imprinted sensor was investigated by electrochemical impedance spectroscopy (EIS). The proposed imprinted sensor exhibited good sensitivity, improving its sensitivity when increasing the conductivity of the solid support. It also showed good thermal and long-term storage stability.

2. Experimental section

2.1 Apparatus

A Metrohm Autolab potentiostat/galvanostat Autolab PGSTAT302N computerized electrochemical instrument, controlled by dedicated NOVA software was employed in all electrochemical measurements, including construction and analytical evaluation stages. Screen Printed electrodes (SPEs) of carbon ink were purchased from DROPSENS, DRP-C110, with a working electrode of carbon with 4 mm diameter, an auxiliary electrode of the same material and a silver pseudo-reference. All electrochemical measurements of the SPEs were made connecting these to a switch box interface, also from DROPSENS. The chemical modification of electrode surface was confirmed by Raman from Thermo Scientific DXR Raman coupled to a confocal microscope.

2.2 Reagents

Ultrapure Milli-Q water was employed in all experiments, and chemicals were pro-analysis, used without further purification. Potassium hexacyanoferrate III (K3[Fe(CN)6]) and Potassium hexacyanoferrate II trihydrate (K4[Fe(CN)6].3H2O) was produced by Riedel-de-Haen. Phosphate buffered saline (PBS) tablets, Myoglobin (Myo), Glucose monohydrate (Glu), Albumin from Bovine

serum (BSA), Hemoglobin (Hb) and Immunoglobulin G, (IgG) EDOT were purchased from Sigma-Aldrich. Creatinine (Crea) was obtained from Fluka and Sulphuric Acid (H2SO4) from Scharlau. EBT was purchased from Biochem.

2.3 Solutions

Stock standard solutions for calibration curve of 1.00x10-5 mol/L BSA were prepared in 1.00x10" mol/L PBS buffer with pH 7.4 (obtained by dissolution of the commercial tablets, as indicated in the label). Less concentrated BSA standard solutions were needed to calibrate the biosensor and were prepared by accurate dilution of the stock solution in PBS buffer. Electrochemical measurements used a solution containing 5.00x10-3 mol/L K3[Fe(CN)6] and 5.0x10-3 mol/L K4[Fe(CN)6], also prepared in PBS buffer. The modification of the working electrode area required a 0.50 mol/L H2SO4 solution, prepared in water, and 1.00x10- mol/L EBT, prepared in the same PBS buffer.

Simulated serum was prepared with the same concentrations as in real serum and used without dilution. Its composition was: 0.85 g/L of Glu, 9.00* 10-3 g/L of Crea and 4.30* 10-5 g/L of Myo. Interfering globulin solutions were prepared in PBS buffer. Their composition was 1.20 *

10-2 g/L Hb

or 0.7 * 10-3 g/L IgG.

2.4 Electro-synthesis of biomimetic material

The working electrode of the SPE was first "cleaned" by oxidation with H2SO4 by cyclic voltammetry (CV), under a potential range of -0.20 V and +1.50 V, for 5 cycles, with a scan rate of 0.050 V/s. The protein-imprinted film was synthesized on the previously cleaned surface, by eletropolymerizing a 1.00x10- mol/L solution of EBT prepared in PBS buffer, and containing 1.00x10-4 mol/L BSA. Polymerization was achieved by CV, between -0.45 V and +0.90 V, under a scan rate of 0.10 V/s, for 4 cycles. Finally, the BSA was extracted from the surface with PBS, pH 7.0, and under CV cycling, from -0.30 V to +0.80 V, with a scan rate of 0.10 V/s for 25 cycles.

Control SPEs were produced by synthesising non-imprinted polymer (NIP) films, using a similar approach but where the protein was missing.

2.5 Electrochemical assays

CV measurements were performed in the standard redox probe solution, containing 5.00x10"5 mol/L in K3[Fe(CN)6] and in K4[Fe(CN)6], prepared in PBS buffer. In CV assays, the potential was scanned from -0.30 V to +0.70 V, at a scan rate of 0.05 V.s-1, along 2 successive CV cycles. EIS measurements were conducted in the same redox couple [Fe(CN)6]3"4", at a standard potential of 0.12 V (± 0.01 V) with 50 scans of frequencies, and a sinusoidal potential peak-to-peak with amplitude 0.01 V in the 0.01 Hz-100000 Hz frequency range. Square Wave Voltammetry (SWV) analysis was performed roughly in the same condition as CV; the potential was scanned from -0.10 V to +0.80 V.

2.6 Electrochemical assays

Selectivity studies required the preparation of synthetic serum sample and its spiking with BSA. The concentrations of BSA selected for this purpose ranged between the concentrations applied to

calibration curves having PBS buffer has background medium (1.00x10" - 1.00x10" mol/L). The simulated serum was prepared with 2.50x10-9 mol/L Myo concentration, Glu (4.72x10-6 mol/L) and Crea (7.96x10-5 mol/L), also in PBS buffer pH 7.4. These concentrations values have been selected according to their relative concentration to BSA in biological serum samples.

3. Results and Discussion

3.1 Approach to produce the protein-imprintedfilm

A carbon-working electrode acts as support of the MIP-film. This electrode was first cleaned/oxidized and after subject to electropolymerization in a solution containing EBT and a known amount of BSA. The overall process is presented in Figure 1.

The cleaning stage of the electrode consisted on promoting an electrochemical oxidation of all materials standing at the carbon area. This was achieved by consecutive CV cycling in a solution of H2SO4. This procedure oxidizes the impurities standing at the carbon electrode area, thereby improving their solubility in aqueous solution and allowing their lixiviation from the electrode surface. Thus, this oxidation stage enhanced the electron transfer capability of each blank carbon electrode and contributed for a higher reproducibility among the different commercial units of SPEs [33].

The imprinting stage was performed simply by covering the three-electrode system of the SPEs with a solution of EBT and BSA, and imposing CV conditions capable of yielding the formation of a thin-film of poly(EBT). These CV conditions selected are in agreement with the oxidation potential of EBT in a PBS solution. As shown in Figure S1-A, there are two oxidation potentials of EBT and one reduction potential. Under consecutive CVs, an EBT solution lead to the formation of a polymeric film. The collected data indicated that this polymer had non-conductive features, because its current decreased as the polymer was growing. On the other hand, the polymer was not highly isolating, as its current drop was moderate. No electrochemical alteration of BSA was evidenced under the same CV conditions, suggesting that BSA was chemically stable by the moment the polymer was growing.

Please Insert Figure 1

The subsequent protein removal was achieved also by consecutive CV cycles, using only a PBS buffer. This treatment is expected to remove adsorbed compounds at the surface, also facilitating the exit of the protein exposed towards the outer surface of the polymer. A biosensor carrying a NIP thin-film was also produced in parallel (Figure S1-B) following the procedure employed in the

production of the MIP film but without involving the protein. In this case, the imprinting stage involved a solution containing only EBT.

3.2 Optimization of relevant variables

The variables tested in the assembly of the protein-imprinted film were the concentration of the protein at the imprinting stage and the time given for the protein incubation. The concentration of

BSA as varied between 1.00*10- mol/L and 1.00*10- mol/L, for a fixed concentration of EBT,

always equal to 1.00*10- mol/L. The time given for the protein incubation was varied between 15 and 30 minutes.

The change in BSA concentration was evaluated by EIS. EIS monitors the resistance to chargetransfer (Rct) of a redox standard system on a conductive surface. Herein, EIS data was plotted as Nyquist plots (Figure S2) and followed the Rct changes occurring at that surface after each stage of chemical modification. Overall, the electropolymerization increased the Rct of the previously oxidized carbon surface. The magnitude of such increase was higher for higher concentrations of protein. As BSA is not a conductive material, when it becomes entrapped within the polymer matrix it contributes for a further increase in the Rct. The subsequent protein removal also indicated that BSA would be leaving the surface by decreasing the value of Rct. This decrease was more intense up to concentrations of 1.00^10-5 mol/L BSA, suggesting the formation of a higher number of binding sites under this concentration, which was consistent with the higher number of BSA molecules per polymer formation. For higher concentrations of BSA the exit of the protein became limited, as the Rct increased (instead of decreasing, as expected); it seemed that BSA adsorbed onto the MIP surface firmly, thereby limiting the access to other BSA molecules for rebinding. Considering these results, the concentration of BSA selected in this study was 1.00^10-5 mol/L of BSA at the imprinting stage. The time of incubation of BSA at the rebinding period was 15, 20 or 30 min at room temperature, for

complete calibrations with 7 standard BSA solutions, ranging from 5.00x10 and 1.00x10" mol/L.

These results were also followed by EIS and were consistent with the previous studies, revealing an Rct increase after BSA incubation (Figure S3). It was also observed that the longer is the incubation of BSA, the greater is the Rct. This observation revealed time as variable that strongly affected the sensitivity of the biosensor. The calibration with 15 minutes incubation periods presented a linear

response ranging from 5.00x10- mol/L to 1.00x10- mol/L and a slope 360.3 Q/decade; the 20 minutes incubation showed a higher slope, equal to 641.0 Q/decade, for the same linear range; the 30 min incubation presented a slope 1137 Q/decade for a linear response between 5.00x10- mol/L and 5.00x10-6 mol/L. Although, the incubation of 30 min had the best sensitivity when compared with other time conditions, the Rct change saturated earlier, narrowing the concentration range of linear response. Thus, an intermedium time of incubation of 20 min was selected, offering intermedium conditions of slope and linear range of concentration.

3.3 Electrochemical follow-up of the assembly

All stages of the SPE modification leading to the production of a protein-imprinted sensor were followed by EIS and SWV. The data obtained is presented in Figure 2. The Bode plots corresponding to the Nyquist plots presented are also shown in Figure S4; the Bode phase expressed the capacitance of the electrochemical cell [34, 35]. The double layer capacitance of each step (Figure S4B) was calculated from the frequency at the maximum of the semicircle present in the Nyquist plot (Figure 2).

In the Bode plots, the oxidized carbon showed ^ values of 35.89° and 30.71° for MIP and NIP electrodes, at 2 Hz of frequency, respectively. This accounted the usual variability of the commercial devices, but readings in relative values leads to consistent and repeatable evaluations. Then, the MIP material shifted from 35.89° to 51.07° in coupled with a frequency shift (1.43 Hz) after electropolymerization, and the angle decreased back to 48.30° and its frequency was 1.57 Hz after template removal. In turn, the NIP material increased around 10° (^ = 40.44°) with a frequency of

1.57 Hz after eletropolymerization, and showed no frequency shift after template removal, although coupled to a decrease in Overall, the fact that after protein removal the frequency shifted only in the MIP film (and not in the NIP films) and the ^ of MIP and NIP are in the same frequency, suggested that the protein molecules used as template successfully exited the polymeric film. Moreover, the Bode modulus reflected the solution resistance Rs, which consisted in the resistance between reference and working electrodes, mediated by the redox probe, and the Rct. All modification stages showed approximately the same Rs, of ~2.54 Q, but the Rct increased after the electropolymerization, both in MIP (from 3.27Q to 3.70 Q) and NIP (from 30.71 Q to 40.44 Q) films. After template removal, the Rct of the MIP decreased ~0.12 Q, while the NIP decreased ~0.07Q.

In the Nyquist plots (Figure 2B), the values for Rs and Rct were determined by the semicircle. The Randles equivalent circuit was used herein because it was the model that best fitted the behaviour of biosensor along the different stages of its assembly, which all elements present error below to 7% (Table S4). This circuit combined the resistance of the solution phase (Rs), with the constant phase element (CPE, which can behave as a resistance if n=0, capacitance n=1 or Warburg impedance if n=0.5, and where n is the value of the exponent of the constant phase element ) [36], the Rct and the Warburg diffusion element (W) (Figure 2). Overall, the MIP films showed that the Rct of electropolymerization of EBT in MIP increased 2890 Q (up to 4180 Q) and then decreased by 1090 Q (down to 3090 Q) after template removal (Figure 2B1). This observation accounted the fact that the protein and the polymer act as non-conductive materials. In turn, the NIP had an Rct increase of only 704 Q (Figure 4 A) after polymerization and an additional 190 Q (but little) increase after template removal (Figure 2B2). The small Rct increase in the NIP polymerization stage reflected the absence of the protein among the polymeric matrix, while the little increase in the template removal stage reflected the reorganization/stabilization of the polymeric surface just formed.

In the SWV assays (Figure 2A), the corresponding data was evaluated in terms of peak current, which changed inversely to the Rct values recorded previously. After MIP polymerization, the peak current decreased from 42.2 ^A to 31.1 ^A and the peak position shifted from 0.32 V to 0.43 V; the template removal yielded intermedium features, with a current increase up to 34.1 |iA and a peak shift to 0.40 V (Figure 2A1). The SWV analysis of NIP showed a lower current after polymerization, followed by a slight current increase in the template removal stage (Figure 2A2). Overall, the SWV data was consistent with the EIS data.

Please Insert Figure 2

In general, all electrical data used to follow-up the assembly of the biosensor was consistent with a successful polymer formation and the efficient template removal. The stability of the polymer formed is an additional feature that deserves attention considering the possible commercial use of such devices and was studied next.

3.4 Stability of the imprinted Poly(eriochrome black T)

A biosensor shall present good stability features to provide precise and accurate measures along its storage, and allow its subsequent commercialization. For this purpose, the thermal stability of the materials was evaluated and the signals presented by the biosensor were followed daily for a given period, by EIS (Figure 3B).

The MIP-EBT polymer was evaluated daily with an iron redox probe within 8 days of its construction always kept under cold storage. Results of the electrochemical analysis showed that the signal variations ranged 0.24 % - 7.07 %, with an average relative standard deviation of 3.07 % (corresponding to triplicate measurements). Longer periods were not evaluated as the biosensors are expected to have a single use, while these assays checked the stability under consecutive measures within ~1 week (not occurring under commercial use). Overall, these results suggest that the

biosensor will offer enough stability for a possible commercialization under proper packaging, also allowing stable measurements at least for 8 days after opening the (commercial) package.

Please Insert Figure 3

The thermal stability was also evaluated, as it may infer about its long term stability under normal ambient conditions. In this test, the blank signal of the biosensor was recorded; then the biosensor was subject to a constant temperature of 45 °C for 2h30min (Figure 3A) and let stand until room-temperature was reached; finally, its blank signal was measured again. The MIP film had an Rct of ~7120 Q before heat treatment, and its Rct varied ~10 Q after heat treatment. Overall, the results obtained support the long-term stability of the device.

3.5 General analytical features 3.5.1 Calibration curves in buffer

EIS calibration curves were obtained first by incubating increasing concentrations of BSA standard

solutions, ranging from 1.00x10- mol/L to 1.00x10- mol/L and prepared in PBS buffer. After each incubation, the Nyquist plot was obtained for a solution of 5.0x10-3 mol /L [Fe(CN)6]3-4-, prepared in PBS buffer, pH 7, at a standard potential of 0.12 V ± 0.02 V. This procedure applied to MIP and NIP films.

The data so obtained indicated an Rct increase for increasing concentration of BSA (Figure 4A), accounting the presence of a non-conductive biomolecule bound to the sensing film. The resulting calibration curves so obtained plotted Rct against logarithm BSA concentration (Figure 4C). In the MIP biosensor, a linear regression of Rct = 644.28 Log [BSA] + 7504.8 was obtained, with a squared

correlation coefficient of 0.9966, from 5.00x10 mol/L to 1.00x10 mol/L. The data obtained with the corresponding NIP films and in the same concentration range corresponded to smaller slope (555.6 Q/decade) and a low squared correlation coefficient of 0.8097. Thus, the NIP film displayed

an uncontrolled BSA rebinding profile, justified by the existence of nonspecific binding at the sensing surface (Figure S7 A).

Overall, the slope of the MIP was 16% higher than that of the NIP. The linear response of the NIP was unacceptable for quantitative application purposes, in contrast to that of the MIP, where the linear response offered small differences between the experimental data points and the linear regression established. These results suggest that the rebinding positions of BSA generated among the polymeric matrix are dominating the analytical response. In turn, non-specific binding positions exist in smaller extent (as expressed by the NIP response), but generate an uncontrolled rebinding behaviour (as expressed by the very small correlation coefficient).

Please Insert Figure 4 3.5.2 Calibration curves in simulated serum

In a second stage, the background medium used for calibration was similar to serum. Considering that HSA would be a major component in serum, the use of blank serum this purpose was not possible. Instead, other ions that could coexist with BSA in real serum samples and are likely to interfere were used, considering their relative levels in biological fluids. The calibrating solutions were then prepared in a solution containing Crea, Glu and Myo in the same concentrations as expected in human serum (to allow determinations in samples without dilution). The BSA

concentrations changed from 1.00x10- mol/L to 1.00x10- mol/L (as in calibrations with standards in PBS buffer). Accordingly, EIS measurements were performed in

5.0x10 mol/L of

K3[Fe(CN)6]/K4[Fe(CN)6], prepared in the same background serum (Figure 4B). As in buffer, the Rct increased for increasing BSA concentrations (Figure 4B) and the resulting calibration curves plotted Rct against logarithm BSA concentration (Figure 4D). The MIP biosensor displayed a slope of 935.3 Q/decade, with a squared correlation coefficient of 0.9904, from 5.00x10-mol/L to 1.00x10-5 mol/L. In the same concentration range, the NIP biosensor showed lower slope

(834.1 Q/decade) and a poor squared correlation coefficient comparing with MIP (0.9392) (Figure S7 B). The detection limit (LOD) of the MIP biosensor was 4.51x10-7 mol/L.

Overall, the slope of the MIP was higher (12 %) than that of the NIP and the linear response was of a much superior quality, suggesting higher precision and reproducibility of the analytical data. Moreover, the use of simulated serum instead of buffer contributed to a 45% slope increase. The higher ionic strength of the simulated serum, when compared to the PBS buffer is the probable cause for this result. 3.5.3 Selectivity

The selectivity test evaluated the effect of chemical compounds present in biological fluids. This was achieved by a competing test between BSA (in 4.00x10-6 mol/L) and other biomolecule present in human serum. The interfering species selected for this purpose, were the same selected to simulate the serum sample: Crea, Glu, and Myo. The incubation of the binary solutions (BSA + interfering species) was set 20 min, the same period used in the calibration procedure, and compared to that of a single BSA solution. The results obtained (in Figure S5) indicated a low binding of ability of Crea (4.45 %), Glu (3.82 %), Myo (7.64 %), Hb (11.4%) and IgG (-2.62%), revealing the good selectivity features of the MIP film.

Overall, the BSA dominated the analytical response, even in the presence of very high concentrations of other competing molecules. These results are consistent with the fact that the rebinding positions are controlling the analytical response. In fact, Myo is a small protein that would fit the huge cavities linked to BSA rebinding and it was unable to change significantly the response promoted by BSA.

3.6 Application of the biosensor

Simulated serum samples were analysed next. First, a MIP electrode was prepared and calibrated. Then, spiked simulated serum samples were tested in another sensor, just as a calibration is done. The concentrations of these spiked samples were calculated by means of the calibration curve of the

first electrode. As different electrodes are involved in this process, only relative values are considered. In general, the experimental errors linked to the known amount of BSA present ranged from -6.60 % to +15.4 % with an average relative standard derivation of ± 1.94 %. The response of the corresponding NIP was also evaluated by same way. The experimental errors obtained for the NIP were much higher than for the MIP, ranging between -14.4% and -77.9% (Figure 5). Considering that this study was made in serum, it also suggested that the MIP was more selective than the NIP.

Please Insert Figure 5

Overall, it seemed clear that the MIP sensor is more selective and reproducible than the NIP, indicating a good accuracy and precision for the obtained MIP data.

3.7 Modifying the conductive support

The ability of given MIP biosensor to distinguish BSA from other co-existing compounds is directly linked to the polymer surface interacting directly with the sample, but the sensitivity by which this discrimination is evidenced depends on the nature of the signal and the materials involved. Herein, as the signal is of electrical nature, we launched the hypothesis that the use of a highly conductive polymer could improve the sensitivity of the analytical response, i.e., the relative Rct change between consecutive concentrations. For this purpose, PEDOT was electrodeposited on the carbon-working electrode. PEDOT exhibits high conductivity and is among the most stable conducting polymers currently available [37]. 3.7.1 Electrochemical follow-up

The EIS data involved in the construction of PEDOT modified sensor is presented in Figure 6. The PEDOT layer was obtained by electropolymerization of EDOT, and this resulted in a significant decrease of the Rct recorded in the Nyquist plots (Figure 6A1 and 6B1). The Rct increased again after the formation of the PEBT/BSA (305.7 Q, Figure 6A1) and PEBT (130.0 Q, Figure 6B1) films,

corresponding to MIP and NIP biosensors. The subsequent template removal step decreased again the Rct; this decrease was more significant in the MIP than in the NIP control, thereby confirming the exit of the protein from the polymeric matrix.

Please Insert Figure 6

Accordingly, the SWV data was consistent with the previous EIS data. The PEDOT electropolymerization step increased the current of the oxidized carbon (6.68 |iA in the MIP and 4.83 |iA in the NIP), respectively, decreasing again after the electropolymerization step of PEBT. Such decrease was more evident in the MIP, confirming the presence of BSA within the polymeric matrix. The template removal caused a current increase, which was smaller in the NIP. In addition, a peak potential shift was observed, but only in the MIP film (of -5.04 mV), which confirmed again the exit of the protein.

3.7.2 Raman spectroscopy follow-up

The Raman spectra of the working electrode was recorded for the several stages of electrode modification. The obtained spectra are shown in Figure 7, and evidenced three peaks occurring typically in carbon materials, located at 1370, 1580 and 2700 cm-1 Raman shift. These are known as G, D and 2D peaks, respectively. The G peak represented the bond-stretching vibrations of sp2 hybridization carbon atoms, expressing the C=C stretching; the D peak expressed the vibrations of the carbon atoms of dangling bonds or sp3 hybridized of carbon atoms, indicating the presence of disordered and/or defected in the carbon[38, 39]. The 2D peak represented the second order of the D band, involving a two-phonons lattice vibrational process, without the presence of any kind of disorder or defects [40].

—Please insert figure 7 here — The intensity ratio of ID/IG peaks is often used to quantify the defect density of the carbon material and was used herein the check the occurrence of a given chemical modification (table 1). The oxidized carbon was the starting material and yielded an ID/IG ratio of 0.94. When EDOT was

electropolymerized on top, this ratio increased to 1.04. This signalled the Cp-Cp stretching in the PEDOT matrix and was coupled by a D peak shift to higher frequencies (from 1352.9 to 1366.8 cm-1) and a G peak shift to lower frequencies (from 1583.4 to 1580.4 cm-1) [41]. Two additional peaks appeared at this stage: (1) one at 1434.3 cm-1, the strongest one, and probably assigned to Ca=Cp (—O) symmetric stretching vibration; (2) and at 1504.9 cm-1 assigned to the C=C stretching. Overall, the changes occurring at the Raman spectra upon EDOT electropolymerization confirmed the presence of PEDOT on top of the carbon electrode.

—Please insert table 1 here — The addition of a polymeric imprinted layer on the PEDOT is expected to contribute to disorder the sp2 carbon system, thereby leading to an ID/IG ratio increase. This was confirmed by a moderate ratio increase, from 1.04 to 1.07. Considering that the PEDOT film yielded two additional peaks, the ratio between these peaks (I1/I2) was also evaluated; the intensity ratio decreased from 1.64 to 1.53, thereby also confirming that the PEDOT layer had been modified. Moreover, the 2 peaks had a Raman shift (from 1504.91 to 1509.28 cm-1) as well as the G peak (from 1580.41 to 1597.95 cm-1), thereby also confirming the successful modification of the PEDOT layer.

The template removal stage was in general consistent with the previous observations. The ID/IG ratio decreased (from 1.07 to 0.95), was well as the I1/I2 ratio (from 1.53 to 1.28). The 2 peaks had a Raman shift (from 1509.3 to 1504.8 cm-1) as well as the G peak (from 1597.9 to 1587.8 cm-1), thereby also confirming the successful protein exit.

Overall, Raman spectra confirmed the successful chemical modification of the sensing surface. The usefulness of such modification was evaluated next by checking the resulting analytical features.

3.8 Re-evaluation of the Analytical Features

To assess the effect of the PEDOT film under the PEBT imprinted layer on the performance of the BSA biosensor, the analytical features of this novel biosensor were also evaluated. This was done by

calibration curves and selectivity studies, employing a similar approach to that reported for the biosensors without PEDOT. The obtained EIS data and the corresponding calibration curves results are present in Figure 8.

— Please insert figure 8 here —

3.8.1 Calibrations in PBS buffer

In PBS buffer solutions (Figures 8A and 8C), the MIP displayed a linear trend from 5.00x10- mol/L to 1.00x10-5 mol/L, fitting the equation Rct = 257.6 x log [BSA] + 1951.6, with a squared correlation coefficient of 0.9896 (Figure 8C). In the same concentration range, the NIP material showed a much lower slope (91.27 Q/decade, corresponding to only 35 % of the MIP slope) and an unacceptable squared correlation coefficient (0.7281). These results indicated clearly that the non-specific response became negligible and that the sensitivity improved, only by placing a PEDOT film under the sensing polymeric layer. Although the absolute Rct values seemed smaller than in the biosensor without PEDOT, its relative changes between consecutive concentrations were higher. This revealed and revealed an increased sensitivity, which was expected to lead to an increased accuracy.

3.8.2 Calibrations in simulated serum

In simulated serum conditions, a linear response was observed from 5.00x10- to 1.00x10- mol /L, with a slope of 285.4 Q/decade (Figure 8B), and a squared correlation coefficient of 0.9859 (Figure

8D). The limit of detection (LOD) and limit of quantification of the MIP were 4.95x10- mol/L and

5.00x10- mol /L, respectively, similar to the ones presented. As reported before, the NIP presented lower slope and poorer regression correlation coefficient than the MIP; the slope was 69.6 Q/decade (39% slope than that of the MIP) and the squared correlation coefficient was 0.866.

Overall, the relative Rct changes increased from 0.31 to 2.69, when PEDOT was added to the carbon support before producing the MIP film of PEDOT on TOP (Figure S7). Thus, this tremendous sensitivity increase was clearly assigned to the presence of a conductive material under the antibody-like film. This fact seems to be related with positive charges that PEDOT has along

polymeric structure [42], which it allows to electrostatic interaction between BSA (negative charged) and polymer.

3.9 Comparison to previousy reported BSA sensors

Molecular imprinted polymers for BSA detection published throughout the years are listed in table 2. Comparing all these sensors, the surface to which the imprinted MIP is attached are very different, as well as the kind of monomers employed and the type of polymerization. However, these sensors present nanostructure materials with huge conductivity features and it seems to be related with reach very low limits of detection. In this work, it was possible to observe that the sensibility of MIP had a significant increase after introduction the high conductive material, PEDOT, confirming the influence of nanostructured materials at the sensing element. Despite of the detection limit of the present sensor to be high, it shows LOD similar or better that some biosensors present in literature. This work also offers enough detection capability to real samples analysis, along with lower cost material and higher construction simplicity than all other concurring works. In addition, the proposed work fits well the disposable requirements in biological analysis in point-of-care in terms of cost.

—Please insert table 2 here —

4. Conclusions

EBT was applied successfully as monomer in the preparation of protein-MIP films by electroplolymerization, having a protein as molecular target. The obtained polymeric films assembled on a carbon support displayed good rebinding features for BSA, offering also suitable thermal and long-term stability, and exceptional reproducibility within time.

The test of the biosensor for increasing BSA concentrations revealed good sensitivity/selectivity features, over a wide concentration range and ability to discriminate BSA among other compounds. The calibration in simulated serum was more sensitive than that in PBS buffer. Moreover, the inclusion of a highly conductive film before assembling the MIP layer (represented herein by

PEDOT) lead to a significant increase upon the relative signal variations. The absolute resistance signals decreased because PEDOT improved the conductivity of the system, but the difference between the relative signals along a calibration were much more significant.

The control materials (NIP) showed random or uncontrolled signal variations. This confirmed that the MIP response was mainly controlled by the specific rebinding sites tailored at the imprinting stage. In turn, the non-specific rebinding sites had little or none contribution to the analytical signals generated by MIP devices. Since these non-specific rebinding sites account the direct affinity between the protein and the polymeric film, this suggests that the PEBT holds a beneficial composition for the biorecnognition of proteins, establishing little interactions with these.

Overall, this biosensor presented here displays strong simplicity in designing, short measuring time, good analytical performance within the biological levels of BSA, and good selectivity against other possible interfering proteins. Thereby, the proposed biosensor offered suitable analytical features for point-of-care applications and seems to be a suitable approach to create other protein-imprinted materials.

Acknowledgements

European Research Council is acknowledge for funding this work through the Starting Grant 3P's (GA 311086) to MGFS.

References

[1] A. Hasan, M. Nurunnabi, M. Morshed, A. Paul, A. Polini, T. Kuila, M. Al Hariri, Y.-k. Lee, A.A. Jaffa, Recent advances in application of biosensors in tissue engineering, BioMed research international 2014 (2014) 307519-307519.

[2] A.A. Koyun, Esma. ipek, Yeliz Koca, Biosensors and Their Principles, in: P.S. Kara (Ed.), A Roadmap of Biomedical Engineers and Milestones, InTech2012.

[3] D.R. Thevenot, K. Toth, R.A. Durst, G.S. Wilson, Electrochemical biosensors: Recommended definitions and classification, Analytical Letters 34(5) (2001) 635-659.

[4] S. Mohammadali, P.M. D., C. Pu, Electrochemical Biosensor for Glycated Hemoglobin (HbA1c), 2011.

[5] C. Huang, H. Bai, C. Li, G. Shi, A graphene oxide/hemoglobin composite hydrogel for enzymatic catalysis in organic solvents, Chemical Communications 47(17) (2011) 4962-4964.

[6] P.B. Luppa, L.J. Sokoll, D.W. Chan, Immunosensors - principles and applications to clinical chemistry, Clinica Chimica Acta 314(1-2) (2001) 1-26.

[7] K.G. Yang, L.H. Zhang, Z. Liang, Y.K. Zhang, Protein-imprinted materials: rational design, application and challenges, Analytical and Bioanalytical Chemistry 403(8) (2012) 2173-2183.

[8] V. Scognamiglio, A. Antonacci, M.D. Lambreva, S.C. Litescu, G. Rea, Synthetic biology and biomimetic chemistry as converging technologies fostering a new generation of smart biosensors, Biosensors and Bioelectronics 74 (2015) 1076-1086.

[9] T.J. Li, P.Y. Chen, P C. Nien, C.Y. Lin, R. Vittal, T.R. Ling, K.C. Ho, Preparation of a novel molecularly imprinted polymer by the sol-gel process for sensing creatinine, Analytica Chimica Acta 711 (2012) 83-90.

[10] G. Vasapollo, R. Del Sole, L. Mergola, M.R. Lazzoi, A. Scardino, S. Scorrano, G. Mele, Molecularly Imprinted Polymers: Present and Future Prospective, International Journal of Molecular Sciences 12(9) (2011) 5908-5945.

[11] M.L. Yola, T. Eren, N. Atar, Molecularly imprinted electrochemical biosensor based on Fe@Au nanoparticles involved in 2-aminoethanethiol functionalized multi-walled carbon nanotubes for sensitive determination of cefexime in human plasma, Biosensors & Bioelectronics 60 (2014) 277-285.

[12] J. Ji, Z. Zhou, X. Zhao, J. Sun, X. Sun, Electrochemical sensor based on molecularly imprinted film at Au nanoparticles-carbon nanotubes modified electrode for determination of cholesterol, Biosensors & bioelectronics 66 (2015) 590-5.

[13] Z.Y. Li, H.J. Quan, C.B. Gong, Y.Z. Yang, Q. Tang, Y.B. Wei, X.B. Ma, H.W. Lam, Photocontrolled solid-phase extraction of guanine from complex samples using a novel photoresponsive molecularly imprinted polymer, Food Chemistry 172 (2015) 56-62.

[14] J. Erdossy, V. Horvath, A. Yarman, F.W. Scheller, R.E. Gyurcsanyi, Electrosynthesized molecularly imprinted polymers for protein recognition, Trac-Trends in Analytical Chemistry 79 (2016) 179-190.

[15] D.R. Kryscio, N.A. Peppas, Critical review and perspective of macromolecularly imprinted polymers, Acta Biomaterialia 8(2) (2012) 461-473.

[16] P. Monti, G. Calia, S. Marceddu, M.A. Dettori, D. Fabbri, S. Jaoua, R.D. O'Neill, Q. Migheli, G. Delogu, P.A. Serra, Low electro-synthesis potentials improve permselectivity of polymerized natural phenols in biosensor applications, Talanta 162 (2017) 151-158.

[17] P.S. Sharma, A. Pietrzyk-Le, F. D'Souza, W. Kutner, Electrochemically synthesized polymers in molecular imprinting for chemical sensing, Analytical and Bioanalytical Chemistry 402(10) (2012) 31773204.

[18] A. Menaker, V. Syritski, J. Reut, A. Opik, V. Horvath, R.E. Gyurcsanyi, Electrosynthesized Surface-Imprinted Conducting Polymer Microrods for Selective Protein Recognition, Advanced Materials 21(22) (2009)2271-+.

[19] A. Ramanavicius, A. Ramanaviciene, A. Malinauskas, Electrochemical sensors based on conducting polymer- polypyrrole, Electrochimica Acta 51(27) (2006) 6025-6037.

[20] H.J. Chen, Z.H. Zhang, D. Xie, R. Cai, X. Chen, Y.N. Liu, S.Z. Yao, Surface-Imprinting Sensor Based on Carbon Nanotubes/Graphene Composite for Determination of Bovine Serum Albumin, Electroanalysis 24(11)(2012)2109-2116.

[21] F.T.C. Moreira, S. Sharma, R.A.F. Dutra, J.P.C. Noronha, A.E.G. Cass, M.G.F. Sales, Protein-responsive polymers for point-of-care detection of cardiac biomarker, Sensors and Actuators B-Chemical 196 (2014) 123-132.

[22] X.D. Wang, J. Dong, H.M. Ming, S.Y. Ai, Sensing of glycoprotein via a biomimetic sensor based on molecularly imprinted polymers and graphene-Au nanoparticles, Analyst 138(4) (2013) 1219-1225.

[23] S.Q. Liu, L. Bakovic, A.C. Chen, Specific binding of glycoproteins with poly(aniline boronic acid) thin film, Journal of Electroanalytical Chemistry 591(2) (2006) 210-216.

[24] N.S. Ferreira, A.P.T. Moreira, M.H.M. de Sa, M.G.F. Sales, New electrochemically-derived plastic antibody on a simple conductive paper support for protein detection: Application to BSA, Sensors and Actuators B: Chemical 243 (2017) 1127-1136.

[25] T.S.C.R. Rebelo, C.M. Pereira, M.G.F. Sales, J.P. Noronha, F. Silva, Protein Imprinted Material electrochemical sensor for determination of Annexin A3 in biological samples, Electrochimica Acta 190 (2016) 887-893.

[26] H. Yao, Y.Y. Sun, X.H. Lin, Y.H. Tang, L.Y. Huang, Electrochemical characterization of poly(eriochrome black T) modified glassy carbon electrode and its application to simultaneous determination of dopamine, ascorbic acid and uric acid, Electrochimica Acta 52(20) (2007) 6165-6171.

[27] H. Yao, Y.Y. Sun, X.H. Lin, Y.H. Tang, A.L. Liu, L. Guangwen, L. Wei, S B. Zhang, Selective determination of epinephrine in the presence of ascorbic acid and uric acid by electrocatalytic oxidation at poly(eriochrome black T) film-modified glassy carbon electrode, Analytical Sciences 23(6) (2007) 677-682.

[28] L.H. Wang, X.L. Liao, Y.T. Ding, F. Gao, Q.X. Wang, DNA biosensor based on a glassy carbon electrode modified with electropolymerized Eriochrome Black T, Microchimica Acta 181(1-2) (2014) 155162.

[29] X. Li, G. Wang, D. Chen, Y. Lu, P-Carotene and astaxanthin with human and bovine serum albumins, Food Chemistry 179(0) (2015) 213-221.

[30] E.L. Gelamo, C.H.T.P. Silva, H. Imasato, M. Tabak, Interaction of bovine (BSA) and human (HSA) serum albumins with ionic surfactants: spectroscopy and modelling, Biochimica et Biophysica Acta (BBA) -Protein Structure and Molecular Enzymology 1594(1) (2002) 84-99.

[31] Y. Inoue, A. Kuwahara, K. Ohmori, H. Sunayama, T. Ooya, T. Takeuchi, Fluorescent molecularly imprinted polymer thin films for specific protein detection prepared with dansyl ethylenediamine-conjugated O-acryloyl L-hydroxyproline, Biosensors & Bioelectronics 48 (2013) 113-119.

[32] J. Anguizola, R. Matsuda, OS. Barnaby, K.S. Hoy, C. Wa, E. DeBolt, M. Koke, D.S. Hage, Review: Glycation of human serum albumin, Clinica Chimica Acta 425(0) (2013) 64-76.

[33] W. Wang, G. Xu, X.T. Cui, G. Sheng, X. Luo, Enhanced catalytic and dopamine sensing properties of electrochemically reduced conducting polymer nanocomposite doped with pure graphene oxide, Biosensors & bioelectronics 58 (2014) 153-6.

[34] F. Lisdat, D. Schafer, The use of electrochemical impedance spectroscopy for biosensing, Analytical and Bioanalytical Chemistry 391(5) (2008) 1555-1567.

[35] C.M.A.B. Ana Maria Oliveira Brett, Electrochemistry: Principles, Methods, and Applications, Oxford science Publications, Oxford University, 1993.

[36] U. Piratoba Morales, E. Vera Lopez, C. Ortiz Otalora, BASIC ASPECTS IN THE INTERPRETATION OF THE DIAGRAMS OF ELECTOCHEMICAL IMPEDANCE, Dyna-Colombia 77(162) (2010) 13-19.

[37] Q. Zhao, R. Jamal, L. Zhang, M.C. Wang, T. Abdiryim, The structure and properties of PEDOT synthesized by template-free solution method, Nanoscale Research Letters 9 (2014).

[38] A.C. Ferrari, Raman spectroscopy of graphene and graphite: Disorder, electron-phonon coupling, doping and nonadiabatic effects, Solid State Communications 143(1-2) (2007) 47-57.

[39] Z. Luo, T. Yu, Z. Ni, S. Lim, H. Hu, J. Shang, L. Liu, Z. Shen, J. Lin, Electronic Structures and Structural Evolution of Hydrogenated Graphene Probed by Raman Spectroscopy, Journal of Physical Chemistry C 115(5) (2011) 1422-1427.

[40] F. Withers, S. Russo, M. Dubois, M.F. Craciun, Tuning the electronic transport properties of graphene through functionalisation with fluorine, Nanoscale Research Letters 6 (2011).

[41] S. Garreau, G. Louarn, J.P. Buisson, G. Froyer, S. Lefrant, In situ spectroelectrochemical Raman studies of poly(3,4-ethylenedioxythiophene) (PEDT), Macromolecules 32(20) (1999) 6807-6812.

[42] H.-S. Park, S.-J. Ko, J.-S. Park, J.Y. Kim, H.-K. Song, Redox-active charge carriers of conducting polymers as a tuner of conductivity and its potential window, Scientific Reports 3 (2013) 2454.

Table 1 - Analytical data extracted from the collected Raman spectra of the several materials.

Material

Raman intensity

Raman Shift

Peak ratio

G Peak D Peak

G Peak D Peak ID/IG I1/I2

Oxidized Carbon

135.29 127.55

1583.41 1352.90 0.94 —

232.09

141.03 151.34 157.8

1434.32 1504.91 1580.41 1366.80 1.04 1.64

Electrical polymerization

164.00

106.87 127.30 136.62 1434.93 1509.28 1597.95 1364.00 1.07 1.53

Template Removal

155.83

121.50 168.06 160.34 1434.36 1504.80 1587.78 1361.17 0.95

Table 2 - Previous works reporting molecularly-imprinted polymers for BSA detection.

MIP LOD Linear

Transducer Support Material Technique (mg/mL) response (mg/mL) Ref.

Optical MPTS-modified ZnO particles Acrylamide, methacrylic acid and N,N'-methylene- bisacrylamide surface grafting copolymerization 0.2 0.8-2.0 [43]

Optical Hydrogel Acrylamide, N,N-methylenebis-acrylamide Hydrogel-based protein imprinting 3.2 3.2-5.7 [44]

Optical metal layer (Cr and Au) 3-aminophenylboronic acid Electropolymerization 0.02 0.02-0.8 [45]

Piezoelectrical Gold Methacrylic acid, acrylic acid and vinylpyrrolidone Free Radical Polymerization 10 10-100 [46]

Electrical Carbon nanotubes decorated with magnetic nanoparticles Pyrrole Electropolymerization 2.8 x 10-8 1x10-7-1x10-1 [47]

Figure 1 - Design of synthesis of BSA sensor. (A) oxidation of the carbon electrode área; (B) Electropolymerization of monomers in presence of protein; (C) template removal and (D) protein rebinding.

Electropolymerization

--Template Removal

-0.3 -0.025 0.25 0.525 0.8 Potential Applied (V)

-0.3 -0.025 0.25 0.525 0.8 Potential Applied (V)

2000 T

0 1350 2700 4050 5400 Z'(O)

0 1350 2700 4050 5400 Z'(Q)

Figure 2 - SWV (A) and EIS data in the carbon SPE in [Fe(CN)6]3-/4- solution prepared in PBS buffer, pH 7, when assembling MIP (1) or NIP (2) materials, starting with the oxidation of the carbon electrodes, the electrical polymerization for assembling the imprinted or non-imprinted material and the template removal.

Figure 3 - Stability of the EIS readings after thermal treatment (A) and along time (B).

0.10 (imol/L 0.50 (mol/L 1.00 (mol/L -»-2.50 (mol/L A 5.00 (mol/L 07.50 (mol/L x 10.00 (mol/L

(A) \ X -a. *

/ \v \ ^

- * ............

0 1500 3000 4500 6000 0 1500 3000 4500 6000

-6.5 -5.5 -4.5 Log [BSA, mol/L]

I I I I I I I I I-r-

-6.5 -5.5 Log [BSA, mol/L]

Figure 4 - Nyquist plots of MIP biosensors (A, B) and the corresponding calibrations (C, D), obtained after incubation of BSA standard solutions prepared in PBS buffer (A, C) or simulated serum (B, D), followed by EIS reading in standard in [Fe(CN)6]3-/4- solution prepares in the same medium. The response of the corresponding NIP

films is also included.

4500 T

4000 -

g, 3500 -u

3000 -

2500-6.5

MIP NIP 0

—I— -6

Log[BSA, mol/L]

Figure 5 - Calibration curves of different MIP and NIP BSA electrodes evaluated in simulated serum, and the standard deviation between the different electrodes.

-Oxidation ~ ■ Eletrodeposition of PEDOT .....Eletropolymerization--Template Removal

200 300 400 500 600

150 T Z'(Q) 500 r

; (Bi) ^

- >4 250

/ 0 390 780 1170 1560

£ 70.0 =

35.0 0.0

-0.4 0 0.4 0.8

140 0 Potential Applied (V)

' T(B2)

0 390 780 1170 1560

200 300 400 500 600 Z'(O)

0.0 111

-0.4 0 0.4 0.8

Potential Applied (V)

Figure 6 - EIS (1) and SWV (2) data for the different stages of the assembly of the MIP-PEDOT (A) and NIP-PEDOT (B) biosensor using the redox probe [Fe(CN)6]3-/4- prepared in PBS buffer.

Figure 7 - Raman spectra of several films concerning the several stages of modification of the assembly the MIP-PEDOT on the carbon electrode. 1) and 2) Characteristic peak of

PEDOT.

-•-0.10 (imol/L0O.5O (mol/L01.00 (mol/L 0 2.50 (mol/L * 5.00 (mol/L ^7.50 (mol/L * 10.00 (m0l/L

Z'(Q) Z'(Q)

(C) (D)

-7.5 -6.5 -5.5 -4.5 -7.5 -6.5 -5.5 -4.5 Log [BSA, mol/L] Log [BSA, mol/L]

Figure 8 - Nyquist plots of MIP-PEDOT biosensors (A, B) and the corresponding calibrations (C, D), obtained after incubation of BSA standard solutions prepared in PBS buffer (A, C) or simulated serum (B, D), followed by EIS reading in standard in [Fe(CN)6]3-/4- solution prepares in the same medium. The response of the corresponding NIP

films is also included.

Highlights for

Novel monomer to produce stable and sensitive protein-imprinted materials by electropolymerization: application to BSA sensing

Ana P. M. Tavares andM. Goreti F. Sales

> Eriochrome black T as monomer to produce protein-imprinted materials

> On-site electropolymerization of Eriochrome black T

> Poly(eriochrome black T) formed with great stability

> Conductive polymer film produced to improve the sensitivity

> Application to BSA as model protein