Hollow microtubes made of carbon, boron and gold: novel semiconducting nanocomposite material for applications in electrochemistry and temperature sensing

Abstract

Carbon based nanocomposites have recently been intensively investigated as a new class of functional hybrid materials. Here, we present a procedure to obtain a new nanocomposite material made of carbon, boron and gold for applications in electrochemistry and electronics. The presented fabrication protocol uses cellulose fibers as a template that is first modified with an inorganic nanocomposite material consisting of gold nanoparticles (AuNPs) embedded in a polyoxoborate matrix, and then is subjected to the process of thermal decomposition. The as obtained material has a form of tubes with a diameter of a couple of micrometers that are composed of carbonized cellulose coated with the polyoxoborate–AuNP nanocomposite. This inorganic shell, which covers the outer surface of the carbon microtubes, serves as a scaffold that makes the structure stable. The obtained material exhibits electrical properties of a semiconductor with the width of the band gap of about 0.6 eV, and forms Schottky contact with a metal electrode. We show that the new material is suitable for preparation of the NCT-type thermistor. We also demonstrate application of the new nanocomposite in electrochemistry for modification of the surface of a working electrode. Experiments carried out with three exemplary redox probes show that the electrochemical performance of the modified electrode depends greatly on the amount of AuNPs in the nanocomposite.

---

Introduction

The most common route to fabricate functional nanocomposite materials is to combine nanoparticles (NPs) with a natural or synthetic polymer matrix. The resulting composite possesses advantageous qualities, exceeding those achievable with the separate components. Application of the NPs often gives rise to substantial improvement in the mechanical, thermal, and electrical properties of the host material.¹ It is often beneficial to use templates for the formation of desired nanostructures. For instance, various templates have been used to organize NPs on the nanoscale. The examples include block copolymers,² DNA,³ peptides,⁴ nanotubes,⁵ and others.⁶

In an effort to promote sustainable and eco-friendly materials, the U.S. Department of Energy (DOE) elaborated The Technology Road Map with the goal to rich 10% of plant derived materials as chemical building blocks by 2020.⁷ Cellulose is one of the most abundant renewable polymeric materials with exceptional properties that is almost inexhaustible. In this respect, cellulose is an ideal candidate to be used as both a matrix or template material in the fabrication of sustainable functional nanocomposites. Yet, the great potential of cellulose has not been fully exploited. Cellulose can be easily modified into derivatives – ethers and esters – and regenerated materials such as fibers and films.⁸ It can be also transformed into a carbonized form either by combustion or calcination process.⁹ Modification of cellulose surface with NPs offers a facile way to the synthesis of novel composite materials of multifunctional properties. Synthesis and applications of nanocomposites obtained from cellulose functionalized with a variety of organic and inorganic NPs has recently been reported.¹⁰ Enhanced performance of cellulose-based nanocomposites modified with variety of NPs, such as Ag,^11–14 Au,^15–17 Cu,¹⁸ CdS,¹⁹ and ZnO,^20,21 has also been demonstrated. Such materials have proved to be of great importance for catalysis, medicine and food industry.

Natural and synthetic polymer fibers with modified surface can be further processed into new forms of nanostructured materials. For example, cotton, which has a uniform morphology, can serve as both template and precursor for the preparation of carbon based nanocomposites by simple carbonization process.²² Synthesis of fiber-like Fe₂O₃,^23,24 TiO₂,^25–27 NH₄V₄O₁₀,²⁸ and ZrTiO₄ (ref. 29) materials of this type has already been reported.

We present synthesis of a new type of the cellulose-based nanocomposite material possessing semiconducting properties. The fabrication protocol employs cotton originated cellulose fibers as a template. The fibers are first modified with an inorganic nanocomposite material (BOA) that has a form of building blocks consisting of gold nanoparticles (AuNPs) embedded in a polyoxoborate matrix. The shape of the individual BOA block can be approximated by a cuboid of the length of dozens of nanometers and the thickness ∼10 nm. These blocks chemisorb onto the surface of the cellulose fibers. The cellulose modification process consists of two steps described in our recent work.³⁰ First, cellulose sample is immersed in aqueous solution of the AuNPs containing oxoborates. Next, to trigger the condensation reaction of the oxoborate species and formation of the BOA blocks, the solution is acidified by adding hydrochloric acid. One of the most promising applications of BOA-modified cellulose is in medicine, as it exhibits strong antibacterial activity while being completely harmless for human cells.³⁰

Here we show that properties of the BOA-functionalized cellulose change dramatically upon removal of the cellulose template by thermal decomposition. The obtained semiconducting material has a form of hollow tubes of the diameter of a couple of micrometers that are composed of carbonized cellulose, and are coated with a thin layer of the polyoxoborate–AuNP nanocomposite. This layer plays a role of a scaffold providing stability of the structure. Also, it has an effect on the properties of the obtained material. The method of fabrication of the new material is fast, inexpensive and satisfy the requirements of green chemistry approach. To demonstrate the potential applicability of the new nanocomposite, we apply it successfully in electronics and electrochemistry.

Experimental section

Materials

HAuCl₄·3H₂O (99.8%, Aldrich, USA), NaBH₄ granules (99%, Fluka, UK), HCl (analytical grade, POCH, Poland), NaOH (99.8%, POCH, Poland), H₂SO₄ (min 95%, POCH, Poland), H₂O₂ (30%, Chempur, Poland). Ultra-pure water characterized by surface tension of 72.75 mN m⁻¹ at 20 °C and resistivity 18.2 MΩ cm was obtained from the Milli-Q water purification system. Silicon wafers were provided by the Institute of Electronic Materials Technology (Warsaw, Poland). Before use, the wafers were cleaned in a freshly prepared piranha solution (H₂SO₄ : H₂O₂ = 7 : 3 (v/v)), rinsed several times with water, acetone and dried in air.

Synthesis of AuNPs

AuNPs were prepared according to the procedure described previously.³¹ Briefly, an aqueous stock solution of 50 mM gold chloride anions (AuCl₄⁻) in glass vial was prepared by dissolving HAuCl₄·3H₂O with equimolar amount of HCl. An aqueous solution of 50 mM borohydride anions (BH₄⁻) in a glass vial was made by dissolving NaBH₄ with the same molar amount of NaOH. 100 μL of AuCl₄⁻/H⁺ solution was added to a glass vial with 10 mL of deionized water. Afterwards 300 μL of the BH₄⁻/OH⁻ was added rapidly while stirring. Colour of the solution turned from light yellow to brown-orange immediately and then to wine-red during further stirring. The diameter of the synthesized AuNPs, determined using UV-Vis spectroscopy and DLS measurements (see Fig. S1†), was around 4.2 nm. The resulting AuNPs are coated with the layer of oxoboron ions that provide stabilization against aggregation due to the electrostatic repulsion. The as prepared AuNPs were further utilized to modify the surface of the cellulose fibers.

Fabrication of the carbon-polyoxoborates–AuNPs (carBOA) nanocomposite material

To prepare the carBOA material we employed two-step protocol that is illustrated schematically in Fig. 1A. First, the surface of the cellulose/nitrocellulose fibers was modified with the polyoxoborates–AuNPs (BOA) nanocoating according to the method described in great details in our recent work.³⁰ In brief, the material to be modified was first immersed in aqueous solution of the AuNPs. To induce the condensation of the oxoborate species and formation of the BOA nanocomposite, the AuNP solution was acidified by adding a 0.5 M solution of hydrochloric acid. In acidic conditions free borate moieties present at the surface of the BOA nanocomposite bond chemically with OH groups of the cellulose fibers to form a layer on its surface. The modification procedure took from several minutes to hours, depending on the AuNP concentration and pH. During the process, the vial was shaken on mechanical shaker (400 rpm). After the surface modification, the samples were dried under vacuum and then the cellulose/nitrocellulose template was removed. We explored two different routes of the template removal: (i) rapid combustion of the sample, and (ii) calcination of the sample in controlled temperature in a resistance furnace (in the temperature range from 250 to 750 °C). Three-zone resistance furnace (Carbolite, UK) was used for the controlled calcination of the modified cotton.

Fig. 1 (A) Schematic representation of the procedure of the fabrication of the carBOA nanocomposite. (B) Comparison of the SEM images of the product of rapid combustion of pure cotton sample (left) and the carBOA composite (right). Inset: EDS image representing distribution of gold in carBOA. (C) SEM images revealing the morphology of the carBOA material obtained using cellulose and nitrocellulose fibers removed upon rapid combustion.

Estimation of the gold saturation amount adsorbed on unit mass of cellulose

To estimate the maximal amount of AuNPs that can be embedded on the surface of cellulose, we performed the following experiments employing scoured and bleached cotton samples consisting of almost pure cellulose.³² In the first experiment, a sample of raw cotton (0.1 g) was placed in a glass beaker. Then, small amounts (1 mL) of the aqueous solution of the AuNPs (2.2 × 10⁻⁷ moles per liter (M)) were added along with 2.54 μL of a 0.5 M solution of hydrogen chloride. This procedure was continued until free (unbound) BOA material remained in the solution. The presence of the free BOA nanocomposite was confirmed with UV-Vis spectroscopy (preformed with Thermo Scientific Evolution 201 apparatus). The end point corresponded to 17 mL of the AuNP solution added. In the second experiment, to enhance the estimation accuracy, we used a collection of samples of raw cotton wool of equal weight (0.1 g) immersed in the same amounts of distilled water. To each sample decreasing volumes of the AuNPs solution, along with appropriate amount of the 0.5 M HCl solution, were added, starting from the volume of 17 mL. The volume of the AuNP solution added to the consecutive samples was decreased by 0.1 mL. The procedure continued until no unbound BOA residues were present in the solution. As we found, 15.75 mg of gold in the form of the AuNPs was needed to saturate 1 g of raw cotton. Throughout the article the cellulose sample covered with the BOA nanocomposite in an amount corresponding to the above proportion is referred to as the “BOA-saturated” material.

Synthesis of nitrocellulose

Samples of nitrocellulose of different esterification level were obtained by nitrating cellulose (cosmetic cotton) through exposure to nitric acid in the presence of sulfuric acid as a catalyst for different time periods (5, 20, and 60 min). After exposure to acid, the cotton was removed and washed in excess of cold distilled water, then in 10% solution of NaHCO₃, and again in distilled water. Finally, the as-prepared nitrated cotton samples were dried under vacuum.

SEM/EDS analysis

Scanning electron microscope (SEM) and energy dispersive X-ray (EDS) analysis was performed using Zeiss Ultra (Germany) and FEI Nova NanoSEM 450 (USA) apparatus. Zeiss microscope was equipped with micro-analytical EDS setup (Quantax 400 by Bruker, USA) with ultrafast (up to 300 kcounts per s) detector with resolution of 127 eV and active surface of 30 mm².

Current–voltage characteristics and resistance measurements

Two copper wires (electrodes) of diameter of around 590 μm were placed in the glass capillary tube with internal diameter of 600 μm, and separated by around 1 mm from each other. Next, the void between the wires was filled with the carBOA material (about 0.2 g). Then, the glass capillary was thermally sealed over the copper wires to isolate the inside of the tube from the environment, and to fix the position of the wires (see Fig. 3A). The current–voltage characteristics of the such constructed device were performed with BioLogic SP-300 (Bio-Logic Science Instruments) electrochemical system with EC-Lab software v.10.19. The measurements were carried out at 22 ± 2 °C. The dependence between temperature and resistance of the device was measured with the Keithley 196 and CHY24CS Digital Multimeters. Measurements in the range from −50 to 100 °C were performed in a silicone oil bath. Negative temperatures were achieved by addition of dry ice to the system.

Cyclic voltammetry

Cyclic voltammograms were performed with Autolab (Methrom) electrochemical system with dedicated software in a conventional three electrode cell. Working electrode (d = 2 mm diameter), platinum wire (d = 0.5 mm) and Ag|AgCl|3 M KCl were used as a the working, counter, and reference electrode, respectively. To prepare the working electrodes based on the carBOA material we applied the following procedure: carbon tape was used to allow for adhesion between studied material and ITO as conductive substrate. On the as-prepared substrate, mask was applied with hole of diameter of 2 mm. The carBOA material was then applied on the uncovered carbon tape (see Fig. 4A). Copper tape allowed the electrical conductivity between ITO substrate/carbon tape/material electrode and power supply. The electrodes were conditioned before each experiment by submerging in isopropanol in order to evacuate oxygen from between the fibers. The conditioning process took up to 24 h. The solutions used in experiments were prepared by dissolving the redox probes (K₃FeCN₆, Fc(CH₂OH)₂, and Ru(NH₃)₆Cl₃) in 0.1 M KCl aqueous electrolyte. All measurements were carried out at the scan rate 0.02 V s⁻¹ at 22 ± 2 °C.

Thermogravimetry combined with IR spectroscopy

Thermogravimetric analysis of the carBOA material was performed with the use of DSC-TG STA 449 F1 Jupiter (Netzsch, Germany) analyzer connected with FTIR spectrometer Tensor 27 (Bruker Optics, USA). The FTIR spectrum of the released gas product was recorded at 350 °C, corresponding to the temperature of the thermal decomposition of the analyzed sample.

X-ray diffraction studies

Powder XRD data were collected on Empyrean diffractometer (PANalytical). Measurements employed Ni-filtered Cu Kα radiation of a copper sealed tube charged with 40 kV voltage and 40 mA current and Bragg–Brentano geometry with beam divergence of 1 deg. in the scattering plane. Diffraction patterns were measured in the range of 4–50 deg. of scattering angle by step scanning with step of 0.02 degree. The sizes of the gold crystallinities within BOA and carBOA were determined from the Scherrer equation^33,34 based on the (111) peak, with the shape factor K = 0.94.

BET isotherm studies

The BET studies of the carBOA material were performed with the Micromeritics Instrument Corporation (Norcross, Georgia, USA) ASAP 2020 system. Approximately 100 mg of the corresponding solid product was transferred to a preweighed sample tube and evacuated under vacuum at 50 °C on the gas adsorption apparatus until the outgas rate was <5 μm Hg. All gases used were of 99.999% purity. Helium was used for the free space determination after sorption analysis. Adsorption isotherms were measured at 77 K in a liquid nitrogen. The surface area (m² g⁻¹) was determined by fitting the N₂ gas isotherm at 77 K to the BET equation.

Results and discussion

Material preparation and characterization

The new nanocomposite material was fabricated by employing two-step protocol. First, a cellulose sample was modified with an inorganic nanocomposite, BOA, containing AuNPs embedded in a polyoxoborate matrix, according to the recently developed technique.³⁰ The individual BOA building block has a form of a cuboid of the length of dozens of nanometers and the thickness of about 10 nm. These blocks chemisorb onto the surface of the cellulose. Then, the BOA-modified sample was subjected to a process in which the cellulose fibers – playing a role of a template in the fabrication process – were removed by thermal decomposition. The fabrication method is explained schematically in Fig. 1A. In our experiments, thermal decomposition of the cellulose template was achieved either by incomplete combustion (burning) or heat treatment in a furnace in a limited supply of oxygen (calcination). These processes involve both oxidation and pyrolysis that lead to partial carbonization of the cellulose. The products of thermal decomposition of cellulose comprise aromatic (C=C), aliphatic (C–C), carbonyl (C=O), and epoxy (C–O–C) carbon.⁹ Because of the chemical composition (carbon and BOA), in the following, we call this material by its acronym carBOA.

SEM studies

SEM image of the carBOA nanocomposite obtained using rapid combustion is shown in Fig. 1C and compared with combusted unmodified cotton fibers. As seen, the material is composed of tube-like structures having the diameter in the order of 5 μm. The walls of these tubes consist of a layer of partially carbonized cellulose covered with a layer of the BOA composite. The thickness of these walls varies in the range 0.5–1.3 μm. The carBOA material shown in Fig. 1C was obtained from cotton samples covered with the BOA nanocomposite in an amount corresponding to the saturation value (that is referred to as the BOA-saturated one). To demonstrate the role that the BOA nanocomposite embedded on the surface of the fibers plays in the formation of the tubes, we compared morphology of the carBOA composite with the product of combustion of pure cellulose. SEM images of these two materials are presented in Fig. 1B. As seen, combustion of pure cellulose resulted in a completely amorphous structure. This result indicates clearly that the presence of the inorganic shell on the surface of the cellulose fibers provides a scaffold that enables formation of the tubes.

We also employed the same fabrication procedure for nitrocellulose used as the template. We found that the only morphological difference between the carBOA materials obtained using cellulose and nitrocellulose template was the thickness of the walls of the resulting tubes. In case of nitrocellulose the walls of the tubes were roughly six times thinner than those obtained from the cellulose template. This difference can be attributed to the fact that nitrocellulose is much more prone to thermal decomposition, which results in a more effective calcination process. The representative SEM pictures of the carBOA nanocomposite material obtained using nitrocellulose templates are shown in Fig. 1C.

The effect of the temperature and time of the calcination process on the morphology of the resulting carBOA material was also investigated. SEM pictures of the nanocomposites obtained for three different calcination temperatures (350, 400, and 500 °C) and three calcination times (30, 120, and 180 min.) are shown in Fig. S2.† As can be seen, all materials are composed of similar tube-like structures coated with BOA and gold clusters. The inner diameter of the tubes decreases with increasing heat treatment temperature and calcination time. This effect can be explained by the increased degree of the carbonization of the cellulose fibers. Also, at the highest temperatures and prolonged calcination times a tendency of gold to separate from the BOA composite was observed. This is a result of the process of sintering (coalescence) of gold nanoclusters that is greatly accelerated at elevated temperatures.

Note that the SEM imaging of both raw cotton and cotton modified with BOA was difficult to perform due to very small electrical conductivity of these materials, resulting in charge buildup. After the template removal we did not observe charging of the specimen. This observation indicates that the investigated materials became conductive upon the thermal decomposition of the cellulose template.

EDS analysis

To characterize chemical composition of the nanocomposite, EDS analysis was carried out. The studies revealed the presence of carbon, oxygen, gold, and boron elements in the investigated samples (see Fig. S3†). Atomic content of gold, boron, carbon, and oxygen in the sample was, respectively, 0.62, 4.03, 82.3 and 7%, yielding the oxygen-to-carbon ratio O/C = 0.084. The presence of oxygen in the EDS spectrum confirmed partial oxidation of the product of the thermal decomposition of cellulose. Similar values of the O/C ratios were reported⁹ for cellulose samples subjected to the heat treatment at temperatures above 400 °C, and indicates a high degree of carbonization. EDS analysis revealed also that gold nanoparticles are distributed uniformly over the surface of the carbonized fibers (see inset in Fig. 1B).

Thermogravimetry-FTIR

Thermogravimetric analysis combined with IR spectroscopy (Fig. 2A and B) gave more insight into the process of template removal. We found that the cellulose fibers started to decompose at around 350 °C, what was in a good agreement with the literature data.^23,26 The FTIR measurements revealed also that carbon dioxide was the main product released during the cellulose decomposition process.

Fig. 2 (A) Results of thermogravimetric analysis combined with IR analysis of carBOA. The template decomposition occurred in the temperature range from 300 to 350 °C. (B) The IR spectroscopy revealed that CO₂ was the main product released upon decomposition of the template (spectra correspond to marked area in panel A). (C) BET isotherm of nitrogen adsorption recorded at 77 K for the carBOA composite. Inset: schematic representation of the structure the tubes. (D) XRD patterns of raw cotton, BOA modified cotton, and the carBOA material obtained via rapid combustion.

BET isotherm

SEM studies showed that the nanocomposite possess a porous structure. To investigate further the morphology of the carBOA material, we characterized its specific surface area by means of the BET adsorption isotherm. The N₂ adsorption isotherm recorded at 77 K, shown in Fig. 2C, corresponds to the unrestricted monolayer gas adsorption (type II isotherm). The isotherm displayed no significant gas uptake up to 1 bar of the pressure. We found that the material had rather low specific BET surface area; S_BET = 5.97 m² g⁻¹. The observed gas-sorption behavior is characteristic for a macroporous material. The low value of S_BET obtained can be to extent attributed to the presence defects inside the tubes that were generated during the thermal decomposition of the cellulose fibers. These defects can block the channels impairing diffusion and adsorption of the gas (see inset, Fig. 2C).

X-ray studies

The material preparation procedure was monitored by means of X-rays scattering (Fig. 2D). The X-ray diffraction pattern recorded for the unmodified cellulose exhibited characteristic peaks located at around 15.4, 16.2, and 23 deg. Upon modification of the surface with BOA composite two new peaks appeared at around 38.2 and 44.5 deg, corresponding, respectively, to the (111) and (200) peaks of gold crystal. Note that the XRD pattern of the BOA composite does not show signals related to the polyoxoborates that are much less intense than those of gold and cellulose. After the calcination only the peaks of gold are present in the spectrum. This result confirms that the cellulose was completely decomposed or removed during this process. The analysis of the broadening of the diffraction peaks of Au revealed that the average size of the gold clusters was about 6 and 20 nm in the BOA and carBOA material, respectively. This result indicates that during the calcination the AuNPs present in the BOA nanocomposite undergo coalescence.

Application of carBOA in electronics: thermistor

To characterize electrical properties of the nanocomposite, we measured resistance of a sample of carBOA over a range of temperatures, from −50 to 100 °C. To carry out the measurements, we placed a sample of the material between two copper wires and sealed it within a glass capillary (see Fig. 3A). We found that the measured electrical resistance, R, was strongly dependent on the temperature (Fig. 3C). As can be seen in the plot shown in Fig. 3D, the quantity ln(R) is a linear function of the reverse temperature in the whole temperature range investigated. It means that the resistance of the carBOA material follows the exponential dependency, R(T) ∼ exp(E_g/2k_BT), that is characteristic of pure (i.e. intrinsic) semiconductors. Here, k_B is the Boltzmann constant and E_g is the width of the energy gap. Fitting the measured dependency of ln(R) on T⁻¹ to the experimental data yielded the value of the energy gap E_g = 0.573 ± 0.004 eV, which is typical for semiconducting materials. Thus, the carBOA-based device designed represents a prototype of a NTC (negative temperature coefficient) thermistor operating in the employed temperature range.

Fig. 3 (A) A scheme explaining construction of thermistor based on the carBOA composite material. (B) The current–voltage (I–V) plot recorded for the thermistor at room temperature. The I–V curve displays two conducting regimes with the critical voltage ∼3 V. (C) Dependence of the resistance of the thermistor on the temperature. (D) Logarithm of the resistance plotted as a function of the reverse temperature.

The electrical conductance of the product of the calcination of cellulose results from the presence of highly conducting carbon nanoclusters, which are embedded in a matrix of low conducting amorphous carbon. Formation of these clusters starts at the heat treatment temperature ∼400 °C.⁹ Both the amount and size of the carbon clusters increases with the heat treatment temperature. The appearance of the electrical conductivity is accompanied by the drop of the O/C ratio in the carbonized cellulose below ∼0.1. Interestingly, the studied carBOA material was characterized with the ratio O/C = 0.084, which may suggest a similar conduction mechanism.

To further investigate electrical properties of carBOA, we determined the current–voltage (I–V) characteristics of this material in the voltage range from −10 to 10 V at room temperature. The obtained I–V curve (see Fig. 3B) is typical for a symmetric metal–semiconductor–metal device having double Schottky barrier. The dependency of the current, I, on the bias voltage, V, of such device is described by the following equation:

where β = (k_BT)⁻¹, e is the elementary charge, and I_s is reverse saturation current of the Schottky barrier. The quantity V_R in eqn (1) represents the voltage drop due to the resistance of the semiconductor, V_R = I(V)R. For small bias voltages eqn (1) takes the following form:

For higher values of the bias voltages applied, above the breakdown (critical) voltage, all of the voltage changes occurs across the semiconductor that is placed between the metal electrodes, that is V ≈ V_R = IR. In this regime the device is described by the ohmic relation,

Both the regimes, given by eqn (2) and (3), are clearly displayed by the I–V curve obtained for the carBOA material, which is plotted in Fig. 3D. For the sample analyzed we found R ∼ 1110 MΩ, and I_s ∼3.3 × 10⁻¹¹ A. Based on the I–V plot we also determined the value of the breakdown voltage to be V_B ∼ 3 V. The observed rectifying (non-ohmic) character of the metal–carBOA junction is expectable because it is typical for lightly doped or intrinsic semiconductors. (Note that the I–V curve and the R(T) dependence shown in Fig. 3 were recorded for two different thermistors, displaying different resistance at the room temperature.)

Application of carBOA in electrochemistry: electrode modification

We employed the carBOA nanocomposite for modification of the surface of working electrode for applications in electrochemistry. In our experiments we investigated the carBOA material obtained from the BOA-saturated cotton samples as well as obtained from cotton samples covered with the BOA nanocomposite in an amount corresponding to about half of the saturation value. These two types of the material are referred, respectively, as to the “saturated carBOA” and “unsaturated carBOA”. For both saturated and unsaturated carBOA, the cellulose template was removed by rapid combustion. Construction of the electrode used in experiments is shown in Fig. 4A. The electrodes were applied for three redox analytes: potassium hexacyanoferrate(III) K₃FeCN₆ (negatively charged analyte), hexaammineruthenium(III) chloride Ru(NH₃)₆Cl₃ (positively charged analyte), and 1,1′-dimethylferrocene Fc(CH₂OH)₂ (neutral analyte). Cyclic voltammograms recorded for these three redox analytes are shown in Fig. 4B–D, respectively. For each analyte the measurements were performed for the electrodes covered with BOA-saturated cotton, unsaturated carBOA, and saturated carBOA. As can be seen, for each type of the analyte and electrode coating, we observed the presence of the characteristic pair of peaks at the expected values of the redox potentials. The obtained peak-to-peak separation (∼240, ∼190, and ∼200 mV for K₃FeCN₆, Ru(NH₃)₆Cl₃, and Fc(CH₂OH)₂, respectively) is higher than the theoretical value of 59.16 mV for a reversible one electron process³⁵ for all of the samples studied. This result indicates that the electrode surface was partially blocked, electron transfer from the electrode to the analyte was slow, and the resulting electrode resistance was high. This effect can be to some extent attributed to the negative surface charge of the modified electrode due to the presence of polyoxoborates in the BOA nanocomposite. The currents recorded for the BOA-saturated cotton and for the unsaturated carBOA were quite similar. A significant increase of the current was observed for the saturated carBOA. We also observed that capacitance and faradaic currents observed for the saturated carBOA was larger than that measured for BOA and unsaturated carBOA material. The huge difference in the current at the working electrode between the saturated and unsaturated carBOA proves that the electrical properties of this material are determined by the amount of the BOA nanocomposite. This material is fundamentally different from the product of incomplete combustion of pure cellulose.

Fig. 4 (A) Schematic representation of the working electrode modified with carBOA nanocomposite used in electrochemical experiments. Cyclic voltammograms recorded for three redox probes: K₃FeCN₆ (B), Fc(CH₂OH)₂ (C), and Ru(NH₃)₆Cl₃ (D). For each probe the measurements were carried out for the working electrode modified with BOA-saturated cotton, unsaturated carBOA, and saturated carBOA composite. Cyclic voltammograms obtained for the electrode modified with saturated carBOA for three different (E) calcination times (30, 120, and 180 min) and (F) calcination temperatures (350, 400, and 500 °C).

Since the carBOA composite material is negatively charged, one should expect differences in the cyclic voltammograms obtained using neutral, positive and negative redox analytes. Indeed, as can be seen in Fig. 4, the currents recorded for 1,1′-dimethylferrocene (neutral probe) were higher compared to potassium hexacyanoferrate(III) (negative probe). In case of hexaammineruthenium(III) chloride (positive probe), accumulation of the analyte at the surface of the electrode was observed. These effects are attributed to the electrostatic attraction between the surface of the modified electrode and the molecules of the probe.

To check whether the method of the template removal (calcination vs. combustion) affects the electrochemical performance of the electrode, we carried out experiments using carBOA obtained by the calcination in a furnace in controlled temperatures ranging from 350 to 500 °C. Also, we tested different calcination times varying from 30 to 180 min (Fig. 4E). For all electrodes Fc(CH₂OH)₂ was employed as the redox probe. We found (see Fig. 4F) that the cyclic voltammograms obtained for all calcination temperatures and times were quite similar to that recorded for the combusted sample. We observed, however, that for the highest calcination temperatures (450 and 500 °C) the peak-to-peak separation was smaller, as it decreased from ∼200 to ∼100 mV. This effect is most probably due to the enhanced electric conductivity of the samples.

Conclusions

We synthesized novel nanocomposite material composed of carbon, gold, and polyoxoborates (called by its acronym carBOA). The fabrication route involves modification of cellulose fibers – that are employed as the template – with an inorganic nanocomposite comprising AuNPs and polyoxoborates (BOA), followed by their thermal decomposition. The template can be removed either by rapid combustion or by heat treatment in temperature range between 350 to 500 °C. The resulting carBOA nanocomposite has a form of hollow tubes of the diameter of a couple of micrometers and a length of hundreds of microns. The walls of these tubes are composed of carbonized cellulose and are coated with the BOA nanocomposite. This nanostructured shell plays a double role: it provides scaffold making the tubes stable as well as affects their electrical properties. The carBOA nanocomposite exhibits properties of intrinsic semiconductor of the energy gap E_g = 0.573 eV. We demonstrated that this material can be employed as a NCT thermistor operating in the temperature range from −50 to 100 °C. The new nanocomposite material was also used as an electroactive agent for electrode functionalization. We successfully tested the electrodes modified with carBOA for three exemplary electrochemical probes: potassium hexacyanoferrate(III) K₃FeCN₆, hexaammineruthenium(III) chloride Ru(NH₃)₆Cl₃, and 1,1′-dimethylferrocene.

In comparison to conventional crystalline semiconductors, carBOA exhibits a significantly lower mechanical strength. Brittleness is the main drawback of this material, which may impede its processing during the device fabrication. Also, because the synthesis of carBOA is based on cellulose – which is a natural product – the properties of the resulting material may change slightly depending on the source of cellulose used. This factor is the second disadvantage of carBOA to mention. However, the utilization of plant derived materials as chemical building block is not only in line with the green chemistry approach, but also allows to lower the fabrication costs. The proposed cellulose-based approach fits well into the broader context of engineering of bioelectronic devices,³⁶ offering facile tools for fabrication of new type of electronic materials of unique properties.

The carBOA material can be employed for functionalization of surfaces of electrodes with a porous semiconducting layer of tunable electrical properties. Electrodes modified in this manner are widely applied in a variety of electronic devices, including solar cells, capacitors, and sensors. Also, the unique morphology of carBOA makes possible utilization of single microtubes as microwires connecting components of electronic systems.

Author contribution

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflict of interest

The authors declare no competing financial interest.

Acknowledgements

Presented scientific work was financed by Ministry of Science and Higher Education of Republic of Poland from the budgetary resources for years 2013–2015 within project Iuventus Plus 3 IP2012 046572 and the Foundation for Polish Science within Team Programme co financed by the EU European Regional Development Fund, TEAM/2010-6/4. The work of MW and RH was supported by the National Science Centre within the grant Maestro 2011/02/A/ST3/00143. JP was a scholar within START programme (81.2014) operated by Foundation for Polish Science. We are grateful to Andrzej Żywociński for inspiring discussions.

Notes and references

1. D. R. Paul and L. M. Robeson, Polymer, 2008, 49, 3187.
2. (a) W. A. Lopes and H. M. Jaeger, Nature, 2001, 414, 735; (b) C. Minelli, C. Hinderling, H. Heinzelmann, R. Pugin and M. Liley, Langmuir, 2005, 21, 7080; (c) R. Shenhar, E. Jeoung, S. Srivastava, T. B. Norsten and V. M. Rotello, Adv. Mater., 2005, 17, 2206; (d) J. van Herrikhuyzen, R. A. J. Janssen, E. W. Meijer, S. C. J. Meskers and A. P. H. J. Schenning, J. Am. Chem. Soc., 2006, 128, 686; (e) J. van Herrikhuyzen, S. J. George, M. R. J. Vos, N. A. J. M. Sommerdijk, A. Ajayaghosh, S. C. J. Meskers and A. P. H. J. Schenning, Angew. Chem., Int. Ed., 2007, 46, 1825; (f) J. Chai and J. M. Buriak, ACS Nano, 2008, 2, 489.
3. (a) E. Braun, Y. Eichen, U. Sivan and G. Ben-Yoseph, Nature, 1998, 391, 775; (b) L. Jiang, H. Zhang, J. Zhuang, B. Yang, W. Yang, T. Li and C. Sun, Adv. Mater., 2005, 17, 2066; (c) Z. Deng and C. Mao, Nano Lett., 2003, 3(11), 1545.
4. F. Patolsky, Y. Weizmann, O. Lioubashevski and I. Willner, Angew. Chem., Int. Ed., 2002, 41, 2323.
5. (a) D. Ugarte, A. Chatelain and W. A. de Heer, Science, 1996, 274, 1897; (b) G. Wei, F. Xu, Z. Li and D. Jandt, J. Phys. Chem. C, 2011, 115, 11453.
6. (a) S. Chen, Langmuir, 2001, 17, 2878; (b) Z. Guo, Y. Zhang, L. Huang, M. Wang, J. Wang, J. Sun, L. Xu and N. Gu, J. Colloid Interface Sci., 2007, 309, 518; (c) O. Krichevski and G. Markovich, Langmuir, 2007, 23, 1496; (d) Y. X. Zhang and H. C. Zeng, J. Phys. Chem. B, 2006, 110, 16812; (e) J. Wang, X. Zhao, J. Li, X. Kuang, Y. Fan, G. Wei and Z. Su, ACS Macro Lett., 2014, 3, 529.
7. A. K. Mohanty, M. Misra and L. T. Drzal, J. Polym. Environ., 2002, 10, 19.
8. H.-P. Fink, P. Weigel, H. J. Purz and J. Ganster, Prog. Polym. Sci., 2001, 26, 1473.
9. Y.-R. Rhim, D. Zhang, D. H. Fairbrother, K. A. Wepasnick, K. J. Livi, R. J. Bodnar and D. C. Nagle, Carbon, 2010, 48, 1012.
10. I. S. Bayer, D. Fragouli, A. Attanasio, B. Sorce, G. Bertoni, R. Brescia, R. Di Corato, T. Pellegrino, M. Kalyva, S. Sabella, P. P. Pompa, R. Cingolani and A. Athanassiou, ACS Appl. Mater. Interfaces, 2011, 3, 4024.
11. W. K. Son, J. H. Youk and W. H. Park, Carbohydr. Polym., 2006, 65, 430.
12. W.-L. Chou, D.-G. Yu and M.-C. Yang, Polym. Adv. Technol., 2005, 16, 600.
13. A. Fernández, E. Sorianoa, G. López-Carballoa, P. Picouetb, E. Lloretb, R. Gavara and P. Hernández-Muñoza, Food Res. Int., 2009, 42, 1105.
14. M. S. Khalil-Abad, M. E. Yazdanshenas and M. R. Nateghi, Cellulose, 2009, 16, 1147.
15. T. Ishida, H. Watanabe, T. Bebeko, T. Akita and M. Haruta, Appl. Catal., A, 2010, 377, 42.
16. N. Sosibo, P. Mdluli, P. Mashazi, R. Tshikhudo, A. Skepu, S. Vilakazi and T. Nyokong, Mater. Lett., 2012, 88, 132.
17. H. Dong and J. P. Hinestroza, ACS Appl. Mater. Interfaces, 2009, 1, 797.
18. N. C. Cady, J. L. Behnke and A. D. Strickland, Adv. Funct. Mater., 2011, 21, 2506.
19. D. Ke, S. Liu, K. Dai, J. Zhou, L. Zhang and T. Peng, J. Phys. Chem. C, 2009, 113, 16021.
20. R. Liu, Y. Huang, A. Xiao and H. Liu, J. Alloys Compd., 2010, 503, 103.
21. S. Ye, D. Zhang, H. Liu and J. Zhou, J. Appl. Polym. Sci., 2011, 121, 1757.
22. B. Li, J. Zai, Y. Xiao, Q. Han and X. Qian, CrystEngComm, 2014, 16, 3318.
23. S. Liu, L. Zhang, J. Zhou, J. Xiang, J. Sun and J. Guan, Chem. Mater., 2008, 20, 3623.
24. S. Chaudharia and M. Srinivasan, J. Mater. Chem., 2012, 22, 23049.
25. J. Zeng, R. Li, S. Liu and L. Zhang, ACS Appl. Mater. Interfaces, 2011, 3, 2074.
26. X. Liu, Y. Gu and J. Huang, Chem.–Eur. J., 2010, 16, 7730.
27. Y. Lu, Q. Sun, T. Liu, D. Yang, Y. Liu and J. Li, J. Alloys Compd., 2013, 577, 569.
28. H. A. Abbood, H. Peng, X. Gao, B. Tan and K. Huang, Chem. Eng. J., 2012, 209, 245.
29. P. Rouhani, E. Salahinejad, R. Kaul, D. Vashaee and L. Tayebi, J. Alloys Compd., 2013, 568, 102.
30. K. Wybrańska, J. Paczesny, K. Serejko, K. Sura, K. Włodyga, I. Dzięcielewski, S. T. Jones, A. Śliwa, I. Wybrańska, R. Hołyst, O. A. Scherman and M. Fiałkowski, ACS Appl. Mater. Interfaces, 2015, 7, 3931.
31. M. N. Martin, J. I. Basham, P. Chando and S.-K. Eah, Langmuir, 2010, 26, 7410.
32. A. Timar-Balazsy and D. Eastop, in Chemical Principles of Textile Conservation, Routledge, New York, 2012, p. 33.
33. P. Scherrer, Bestimmung der Grosse und der Inneren Struktur von Kolloidteilchen Mittels Röntgenstrahlen, Nachr. Ges. Wiss. Goettingen, Math.-Phys. Kl., 1918, 2, 98–100.
34. A. Patterson, Phys. Rev., 1939, 56, 978.
35. A. J. Bard and L. R. Faulkner, in Electrochemical methods: fundamentals and applications, Wiley, New York, 2001.
36. M. Zhou and S. Dong, Acc. Chem. Res., 2011, 44, 1232.

---

Footnotes

† Electronic supplementary information (ESI) available: UV-Vis and DLS characterization of synthesized AuNPs suspension, additional SEM pictures and EDS spectrum of carBOA material. See DOI: 10.1039/c5ra12146a

‡ J. Paczesny and K. Wybranska contributed equally.

---