Organosilane-functionalization of nanostructured indium tin oxide films

Abstract

Fabrication and organosilane-functionalization and characterization of nanostructured ITO electrodes are reported. Nanostructured ITO electrodes were obtained by electron beam evaporation, and a subsequent annealing treatment was selectively performed to modify their crystalline state. An increase in geometrical surface area in comparison with thin-film electrodes area was observed by atomic force microscopy, implying higher electroactive surface area for nanostructured ITO electrodes and thus higher detection levels. To investigate the increase in detectability, chemical organosilane-functionalization of nanostructured ITO electrodes was performed. The formation of 3-glycidoxypropyltrimethoxysilane (GOPTS) layers was detected by X-ray photoelectron spectroscopy. As an indirect method to confirm the presence of organosilane molecules on the ITO substrates, cyclic voltammetry and electrochemical impedance spectroscopy (EIS) were also carried out. Cyclic voltammograms of functionalized ITO electrodes presented lower reduction-oxidation peak currents compared with non-functionalized ITO electrodes. These results demonstrate the presence of the epoxysilane coating on the ITO surface. EIS showed that organosilane-functionalized electrodes present higher polarization resistance, acting as an electronic barrier for the electron transfer between the conductive solution and the ITO electrode. The results of these electrochemical measurements, together with the significant difference in the X-ray spectra between bare ITO and organosilane-functionalized ITO substrates, may point to a new exploitable oxide-based nanostructured material for biosensing applications. As a first step towards sensing, rapid functionalization of such substrates and their application to electrochemical analysis is tested in this work. Interestingly, oxide-based materials are highly integrable with the silicon chip technology, which would permit the easy adaptation of such sensors into lab-on-a-chip configurations, providing benefits such as reduced size and weight to facilitate on-chip integration, and leading to low-cost mass production of microanalysis systems.

1. Introduction

Indium tin oxide (ITO) is a degenerate n-type semiconductor with wide band gap energy around 3.2 eV [1] that has been highly exploited in the field of optoelectronics due to its simultaneously good conducting and transparency properties [2,3]. ITO thin films have been widely used to develop organic light emitting diodes (OLEDs), solar cells, flat panel displays and transparent conducting electrodes, among others [4–7]. Moreover, some nanostructured materials have been extensively used for developing ultrasensitive, low-cost and miniaturized sensors, as they present large surface-to-volume ratio, favourable electrocatalytic activity and excellent electronic properties [8,9]. Actually, some applications of sensors based on thin-film structured ITO can be found in the literature [10–13], yet no electrochemical sensor application based exclusively on nanostructured ITO has been developed to our knowledge.

Covalent attachment of specific biomolecules to nanostructured conductive surfaces is of great importance for the development of sensitive biosensors based on molecular recognition. The attachment of organic biomolecules to inorganic substrates requires an intermediate layer of molecules. A group of molecules that have been widely used are the so-called organosilanes, which participate in the first steps of immobilization procedures for the fabrication of on-chip biodevices [14]. Although being quite unstable in aqueous solutions at the time of forming the dense and homogeneous layer on the electrode's surface, organosilanes have proved to be very useful for biosensing [15,16]. Many authors have reported the fabrication of oxide-based biosensors functionalized with organosilanes for the detection of biomolecules in aqueous environments. For instance, Wang & Wang described hydrogen peroxide detection with a horseradish-functionalized gold-modified ITO electrode in phosphate-buffered solution (PBS, pH 7.0) with a limit of detection of 8 µM estimated at a signal-to-noise ratio of 3 [17]. Yang & Li measured Escherichia coli O157:H7 by electrochemistry in aqueous solution of [Fe(CN)₆]^3−/4− in PBS [18]. Other systematic investigations on functionalization of ITO-based devices have been reported [19–21]. Hanson et al. [19] described a method for surface modification of ITO with α-quarterthiophene-2-phosphonate to enhance charge transport across anodic and cathodic interfaces in OLEDs. High current densities in simple single-layer devices and double-layer light-emitting devices were obtained compared with those with untreated ITO anodes. Similarly, Hatton et al. [20] reported functionalization of ITO-coated glass thin films with small molecule chlorosilanes, dramatically improving the performance of ITO anodes in OLEDs. Cossement et al. [21] reported n-hexyltrichlorosilane and 6-(1'-pyrrolyl)-n-hexyltrichlorosilane modification of ITO substrates, for further polypyrrole polymerization of the substrates, which is halfway optoelectronics and sensing applications. Such conducting organic polymers have been found to have extensive sensing applications [22–24]. The intermediate molecule used to functionalize ITO substrates in the present work is (3-glycidoxypropyl)trimethoxysilane (GOPTS), and its use has been widely reported in the literature [18,25,26]. This molecule helps controlling ITO physical and chemical properties for biosensing purposes. Hillebrandt & Tanaka [26] reported ITO thin films functionalization with self-assembled monolayers of octyltrimethoxysilane and octadecyltrimethoxysilane, demonstrating the behaviour of the alkylsiloxane monolayer as a barrier for ions in the electrolyte, as well as the passivation effect of the monolayers against electrochemistry. This feature may be used as the basic principle for the development of a biosensor. For instance, Ruan et al. reported the use of GOPTS for immobilizing anti-E. coli antibodies onto ITO thin films in order to fabricate an immunosensor for the detection of E. coli O157:H7, achieving a detection limit of 6 × 10³ cells ml⁻¹ [27]. The same authors presented a study on the AFM and impedance spectroscopy characterization of the specific recognition of E. coli O157:H7 cells by the immobilized anti-E. coli antibodies [18]. They observed a systematic increase in the impedance spectra on GOPTS-functionalized substrates, anti-E. coli antibodies immobilization and biorecognized E. coli O157:H7 with respect to bare ITO substrates. The same principle is investigated in the present work.

Our long-term aim is to test the feasibility of developing robust electrochemical sensors taking advantage of the structural and technological characteristics of nanostructured ITO, i.e. enhanced surface-to-volume ratio, together with other useful properties such as good conductivity and high stability under physiological conditions because of its polarizability [28,29]. In this paper, we report on the interaction of GOPTS with –OH terminated nanostructured ITO surfaces, to test the viability of functionalizing such nanostructured ITO surfaces. The experiments were carried out under two conditions: with as-deposited nanostructured ITO electrodes and also with annealed electrodes, thus observing the effect of the annealing treatment on organosilane adsorption. As-deposited nanostructured ITO substrates were observed by scanning electron microscopy (SEM) and characterized by atomic force microscopy (AFM) for the determination of geometrical surface area. The electroactive surface area was measured by cyclic voltammetry (CV) and results are shown elsewhere [30]. Surface functionalization was monitored by X-ray photoelectron spectroscopy (XPS) as a microscopic characterization. XPS results may point to a better organosilane-functionalization of as-deposited ITO slides, so CV and electrochemical impedance spectroscopy (EIS) were performed on as-deposited substrates for better characterizing them at the macroscopic level. Electrochemical measurements also confirmed the presence of organosilanes on the ITO electrode.

2. Material and methods

2.1. Substrate fabrication: electron beam evaporation

ITO was deposited on one face of double-sided polished n⁺-type Si substrates by electron beam evaporation using a Pfeiffer vacuum classic 500 growth chamber. Substrate temperature was set at 500°C, and deposition rate was set at 1 Å s⁻¹ for depositing an equivalent thin film thickness of 200 nm of ITO. Then, half the samples were left as-deposited, and the rest were submitted to an annealing process, at 600°C under nitrogen atmosphere for 1 h. There is an enhancement of the surface conductivity when films are annealed [31,32], due to the decrease in sheet resistance produced by the crystalline organization of In₂O₃ and SnO₂ molecules in the evaporated ITO. That is the reason why we extended our study to explore the possibilities that annealed substrates may open in the field of surface functionalization.

2.2. Substrate characterization: atomic force microscopy, X-ray photoelectron spectroscopy and electrochemical measurements

AFM observations of as-deposited nanostructured ITO films were carried out in a Multimode 8 with a Nanoscope V electronics from Bruker. Height images were obtained under peak force tapping mode using a 25 µm scanner and SNL tips with force constant of 0.35 N m⁻¹ purchased from Bruker.

XPS measurements of as-deposited and annealed nanostructured ITO substrates were performed in a PHI 5500 Multitechnique System (Physical Electronics) with a monochromatic X-ray source (Al Kα line of 1486.6 eV energy and 350 W), placed perpendicular to the analyser axis and calibrated using the 3d_5/2 line of Ag with a full width at half maximum of 0.8 eV. The analysed area was a circle of 0.8 mm in diameter, and the selected resolution for the spectra was 23.5 eV of pass energy and 0.1 eV step⁻¹. All measurements were made in ultra-high vacuum chamber pressure between 5 × 10⁻⁹ and 2 × 10⁻⁸ torr.

All electrochemical measurements (CV, EIS) were carried out in a conventional three-electrode Teflon electrochemical cell, with a solution of 5 mM ferricyanide/ferrocyanide () in 0.1 M KCl at 25°C. The ITO geometrical projected working electrode area, defined by the inner diameter of an O-ring, was 0.7 cm². A platinum wire was used as a counter electrode, and an Ag/AgCl reference electrode was placed inside the cell. The electrochemical cell was connected to a Biologic-EC-Lab SP150 potentiostat.

2.3. Sample cleaning and functionalization

The samples were cleaned and prepared for the chemical deposition of epoxysilane compounds. ITO slides were treated with acetone and dichloromethane for 10 min each, then rinsed with ultrapure water and finally immersed in 5 : 1 H₂O + H₂O₂ 30% for another 10 min. Then, all slides were rinsed with ultrapure water and dried with a gentle stream of nitrogen. For nanostructured ITO films silanization, a solution of GOPTS at 4% (v/v) in toluene was prepared and slides were immersed in it for overnight reaction at room temperature and physiological pH. Figure 1 shows the schematic of the assembly of the epoxysilane layer onto the ITO electrodes, which consists in a hydrolysis (2.1), described as follows:

2.1

followed by a condensation

2.2

Figure 1.

Schematic of the formation of (3-glycidoxypropyl)trimethoxysilane layer on the ITO substrates. Adapted from [14].

3. Results and discussion

3.1. Atomic force microscopy and scanning electron microscopy

Surface morphology of the clean as-deposited nanostructured ITO electrodes is represented as an AFM topographical scan in figure 2a: nanostructuration is clearly appreciated in the image. The effective geometrical surface area can be calculated by triangularization algorithms from the AFM data. We found that nanostructured geometrical surface area increases 2.5 times with respect to projected area. There is direct correlation between the latter and the increase in electroactive surface area with respect to thin ITO films, which yields enhanced sensitivity in terms of electrochemical performance [30]. This AFM image shows that we are able to fabricate nanostructured ITO films by controlling the electron beam evaporation conditions, according to recent publications [33,34]. Figure 2b shows the nanostructured surface morphology observed by SEM as a complement to AFM measurements. Nanowire formation can be clearly appreciated.

Figure 2.

(a) AFM 25 × 25 µm² topographical scan and (b) SEM micrograph of as-deposited nanostructured ITO electrode.

3.2. X-ray photoelectron spectroscopy

Our XPS analysis is focused on the intensity and position of XPS binding energy peaks for O 1s and Si 2p bonds. All data provided hereby can be found in the electronic supplementary material. Figure 3a–d shows general binding energy spectra for as-deposited ITO electrodes, GOPTS-coated as-deposited ITO electrodes, annealed ITO electrodes and GOPTS-coated annealed ITO electrodes. GOPTS-coated ITO electrodes spectra show an overall intensity baseline reduction with respect to non-functionalized ITO electrodes. The same effect can be observed between as-deposited ITO electrodes and annealed ITO electrodes. Besides, Si 2p bonds (red circles in figure 3b,d) can be appreciated in GOPTS-functionalized ITO electrodes spectra. Insets in each graphic correspond to spectral regions where In 3d and Sn 3d peaks can be found. In all cases, Sn 3d bonds are split into two peaks centred at 495 and 485 eV for Sn 3d₅ and Sn 3d₃, respectively, while In 3d bonds are split into peaks at 451 eV for In 3d₃ and 444 eV for In 3d₅. The separation between two peaks of the same element stands for spin orbital splitting. Spin orbital splittings of a core level of a particular element in different compounds are nearly the same, and so it happens with peak area ratios. This is characteristic of ITO, as some references in the literature show [34,35].

Figure 3.

XPS spectra of (a) as-deposited ITO, (b) GOPTS-functionalized as-deposited ITO, (c) annealed ITO and (d) GOPTS-functionalized annealed ITO. Each image contains zoom plots corresponding to energy regions of Sn 3d and In 3d peaks, respectively. Si peaks are found and circled in graphics (b,d).

The first sign of ITO functionalization appears when observing the O 1s bonds. Indeed, this is shown in figure 4, where a peak splitting due to Si–O bond formation can be appreciated in figure 4b,d, which corresponds to GOPTS-coated as-deposited and annealed ITO slides, respectively. GOPTS is attached to the ITO surface by covalent binding with –OH terminals in the ITO surface. Indeed, the Si core atom binds to surface oxygen and this produces a shift in the corresponding binding energy peak. Figure 4a,c corresponds to non-functionalized ITO slides, thus showing no oxygen peak separation. Non-functionalized as-deposited and annealed ITO O 1s peak can be fitted by two Gaussians, centred at 530.0 and 531.1 eV (as-deposited films) and 530.0 and 531.5 eV (annealed films), and resulting in a single oxygen peak in the generalized spectra. This is not the case for GOPTS-coated ITO surfaces, which do show an O 1s peak separation. The corresponding fitting Gaussians are centred at 530.1 and 532.4 eV for both as-deposited and annealed films, respectively. This can be correlated with some studies that show the variations in XPS spectra for functionalized ITO. Brewer et al. [25] investigated the formation of thiolate and phosphonate adlayers on ITO. Similar to our results, they found that the fitting Gaussians for the O 1 s peak were centred at 530.0 and 531.7 eV for bare ITO, and 530.8 and 532.2 eV for adlayer-coated ITO.

Figure 4.

Gaussian fit of XPS O 1s peak for (a) as-deposited ITO, (b) GOPTS-functionalized as-deposited ITO, (c) annealed ITO and (d) GOPTS-functionalized annealed ITO. GOPTS-functionalized substrates spectra clearly show a peak splitting due to Si–O bond formation.

Figure 5 shows the amplified spectral region for Si 2p binding energies. Figure 5a,c corresponds to bare as-deposited and annealed ITO, respectively, and they show weak features in the low energy range, which is probably a result of impurities due to ITO evaporation onto Si substrates. Figure 5b,d corresponds to GOPTS-functionalized as-deposited and annealed ITO slides, respectively, and a five-time magnified peak at ≈102 eV with higher signal-to-noise ratio can be observed. This may be indicative of the presence of Si on the substrate surface, as a result of GOPTS functionalization. The Si 2p peak position is in accordance with the information provided by Materne et al. [36] in their review on organosilane technology in coating applications. Table 1 contains a numerical description of Gaussian curves used for peak fitting and depicted in figures 4 and 5, where μ is the mean binding energy in which Gaussian curves are centred, σ is the standard deviation of the mean and α is a scaling factor, attending to the following equation: . In the particular case of oxygen, the lowest peak corresponds to O 1s and the highest binding energy corresponds to SiO₂, as reported in the literature [37]. The Si-related peaks shown in figure 5b,d correspond to SiO binding energy [37]. A small shift (≈1 eV) can be observed for the Si peak. More notorious is the case of the O 1s peak splitting, which is increased by more than 1 eV in the case of GOPTS-functionalized slides.

Figure 5.

XPS Si 2p peak for (a) as-deposited ITO, (b) GOPTS-functionalized as-deposited ITO, (c) annealed ITO and (d) GOPTS-functionalized annealed ITO.

Table 1.

Summary of Gaussian fit characteristics for peaks shown in figures 4 and 5. All data are expressed in electronvolt.

As-dep. ITO	101.3	1.09	530.0	0.84	531.1	1.70
Func. As-dep. ITO	102.4	1.19	530.1	0.97	532.4	1.35
Ann. ITO	101.5	1.33	530.0	0.92	531.5	1.61
Func. Ann. ITO	102.4	1.16	530.1	0.96	532.4	1.33

Table 2 summarizes the atomic concentrations in percentage of different elements on the differently treated surfaces. This analysis was done taking into account the background noise level and atomic characteristics of the elements under analysis. Quantitative analysis of XPS spectra leads to the determination of the atomic concentration of the ith element by knowing the intensity of its peak, , where N_i is the average atomic concentration of element i on the surface under analysis, σ_i is the photoelectron cross-section for element i as expressed by peak p, λ_i stands for the inelastic mean free path of a photoelectron from element i as expressed by peak p and K combines all other factors related to quantitative detection of a signal. Clearly, there is a correlation between the Si atomic concentrations and the intensity of the corresponding binding energy peaks, which is coherent with whether the substrates were GOPTS-coated or not.

Table 2.

Atomic concentrations expressed in %.

	O 1s	In 3d	Sn 3d	Si 2p
As-dep. ITO	50.0	38.7	11.3	0.0
Func. As-dep. ITO	57.9	27.4	7.6	7.1
Ann. ITO	51.5	35.0	12.5	1.0
Func. Ann. ITO	60.9	25.7	7.6	5.8

3.3. Cyclic voltammetry and electrochemical impedance spectroscopy

Electrochemical measurements were performed on bare and functionalized as-deposited nanostructured ITO films in order to test the presence of epoxysilane compounds on the sensor surface at a macroscopic level. Figure 6a shows CV measurements on bare (blue line) and GOPTS-functionalized (red line) ITO electrodes. The experiment was carried out in a 5 mM solution in 0.1 M KCl buffer. Potentials were scanned at 50 mV s⁻¹ from −0.4 to 0.9 V. Oxidation peaks appear around 0.3 V, which is in accordance with published information [38,39]. GOPTS-functionalized ITO presents lower intensity current than non-functionalized ITO electrodes, which indicates that the organic layer of epoxysilane compounds acts as an insulator between the medium and the conducting ITO working electrode. Figure 6b shows the impedance response of the redox couple probe on the bare (blue lines) and the GOPTS-functionalized (red lines) ITO electrodes in a Bode impedance plot representation. Frequency was varied from 200 kHz to 100 mHz, and the excitation signal consisted of a 10 mV amplitude sine waveform centred at 200 mV DC. The Bode impedance plot shows the normalized absolute impedance and phase response as a function of frequency in a linear-logarithmic scale representation. No significant difference can be observed in the absolute impedance values for frequencies above 10² Hz. This could indicate that the resistance of the solution was not significantly affected by the immobilization of the epoxysilane layer on the ITO electrodes. On the contrary, there is a difference in the absolute impedance values for bare and functionalized electrodes at frequencies below 10² Hz, indicating a change in the double-layer capacitance and also in the polarization resistance values as a result of the chemical surface modification, supporting the hypothesis that the epoxysilane layer behaves as a diffusion barrier for ions in the electrolyte [18,26]. An increase in electron transfer resistance can be observed in the low-frequency range for organosilane-functionalized substrates with respect to bare substrates.

Figure 6.

Electrochemical measurements to prove the presence of GOPTS on the ITO surface. (a) CV before (blue line) and after (red line) GOPTS immobilization on the ITO surface. Three repetitive potential cycles were applied from −0.4 V to 0.9 V at a scan rate of 50 mV s⁻¹. (b) Bode impedance plot from 200 kHz to 100 mHz for as-deposited ITO (blue line) and after GOPTS immobilization (red line), demonstrating the presence of functionalizing molecule on the working electrode surface.

Future work will involve the optimization of annealing conditions (time of annealing and temperature) for a better adhesion of organosilanes on the electrode surface. Besides, an in-depth study of ITO substrates as electrochemical transductors will be done, analysing their behaviour under different electrochemical conditions.

4. Conclusion

In summary, the current study presents an analysis on the viability of developing nanostructured ITO electrodes for biosensing applications. The AFM image shows very densely distributed ITO nanostructures grown on Si substrates during electron beam evaporation at high temperatures. Nanostructures have higher surface-to-volume ratio and thus higher geometrical and electroactive surface area than thin films, permitting the flow of elevated currents and increasing the sensitivity of such working electrodes as electrochemical sensors. As-deposited and annealed nanostructured ITO electrodes were analysed by XPS at the microscopic level, before and after organosilane-functionalization. XPS allowed qualitative and semi-quantitative information to be extracted from the films, indicating that the level of attachment of silanes may be slightly higher on as-deposited ITO electrodes than on annealed ITO electrodes. This could be ascribed to a better adhesion of GOPTS molecules on disorganized atomic distributions (as-deposited ITO films) rather than on crystalline organized annealed ITO substrates. Finally, the macroscopic electrochemical behaviour of both bare and functionalized as-deposited nanostructured ITO electrodes was analysed by CV and EIS. These techniques also permitted semi-quantitative analysis of the effect of the organic epoxysilane layer on the electron transfer between redox species in solution and conducting nanostructured ITO substrates. A slight increase in polarization resistance can be observed in organosilane-functionalized ITO electrodes, which accounts for the behaviour of the organic layer as an electronic barrier.

The obtained results may indicate that a basic and systematic functionalization of nanostructured ITO is possible, which would open the door to a wide scope of possible applications in fields such as biomedicine and environmental sciences.

Authors' contributions

R.P. conducted the experiments and wrote the manuscript, with inputs from all other authors. R.P. and M.L. conceived the initial goal of the project. O.B., S.H. and B.G. were involved in the fabrication of the nanostructured surfaces. F.P., M.M. and M.L. participated in the chemical modification of the substrates and the interpretation of the results. All authors gave final approval for publication.

Competing interests

We declare we have no competing interests.

Funding

This project was supported by the 7th Frame Program European project ‘Real-time monitoring of SEA contaminants by an autonomous lab-on-a-CHIP biosensor (SEA-on-a-CHIP)’, grant no. 614168.