Fabrication of a peptide–AuNP–TiO₂ nanocomposite and its application as a VOC sensor

Abstract

We fabricated a volatile organic compound (VOC) sensor with a peptide–Au nanoparticle (AuNP)–TiO₂ nanocomposite in which AuNPs were linked with TiO₂-coated conductive peptide nanowires. The conductive peptide nanowires were formed between the AuNPs via self-assembly through the complexation of amphiphilic peptides, LESEHEKLKSKHKSKLKEHESEL, and Co(II). Furthermore, TiO₂ mineralization on the surface of the peptide nanowires yielded mixed crystals of rutile and anatase, which exhibited highly effective photolytic activity. In particular, the obtained TiO₂ exhibited three times greater photodecomposition activity in the unsintered state toward organic matter than did commercially available TiO₂. Next, we constructed a VOC sensor by immobilizing peptide–AuNP–TiO₂ nanocomposites on a comb electrode. The electrochemical properties of the nanocomposite changed drastically under light irradiation in the presence of VOCs, indicating transport of the VOC-decomposition-generated photoexcited electrons of TiO₂ to AuNPs through conductive peptide nanowires, which prevented electron–hole recombination. The obtained sensor exhibited a sensing range of 2–100 ppm for dichloromethane, which was used as a representative VOC. Therefore, nanocomposites made of AuNPs linked with conductive TiO₂ nanotubes may be highly effective for TiO₂-driven VOC decomposition. Moreover, we believe that this nanocomposite has high sensitivity for sensing VOCs.

Introduction

Environmental, water, and air pollution have emerged as significant challenges owing to industrial development in recent years. Although air pollution occurs due to many factors, volatile organic compounds (VOCs) may be a major factor [1,2,3,4,5,6]. Photocatalytic degradation is among the most economical and effective strategies for reducing the concentrations of various environmental pollutants owing to the strong redox capabilities of photocatalysts [7,8,9,10,11]. In particular, titanium dioxide (TiO₂) is a widely recognized and promising photocatalytic material because of its lack of toxicity, low production cost, and high photosensitivity due to a suitable band gap [12,13,14,15,16]. However, this photocatalyst is limited by its low separation rate for photoexcited carriers as well as low gas adsorption capacity [17,18,19]. Li et al. demonstrated the high photocatalytic activity of nanocomposites comprising TiO₂, carbon nanotubes (CNTs), and Au nanoparticles (AuNPs) [20]. Zhang et al. reported a high-activity photocatalyst comprising self-assembled TiO₂ nanowires on carbon nanofibers coated in metal–organic frameworks [21]. Furthermore, Li et al. reported high photocatalytic activity for TiO₂ nanodots anchored on N-doped CNTs encapsulated with Co NPs [22]. In these systems, the recombination of electrons and holes was generally prevented by transporting photoinduced electrons to Au or Co NPs through conductive nanowires, thus improving the photocatalytic efficiency of TiO₂.

To freely generate nanolevel conductive pathways, various scholars have constructed conductive nanowires, including nanofibers consisting of α-helix peptides [23] and polymer nanowires self-assembled through π-conjugation [24]. Although organic nanowires can be readily self-assembled between functional parts, they have lower mechanical durability than inorganic materials. We previously reported the construction of an organic/inorganic hybrid nanowire to resolve this drawback [25]. Nanowire-linked electrodes were self-assembled through the complexation of amphiphilic peptides and Co(II)s, and the surface of the electrodes was coated with silica via mineralization. Thereafter, the Co(II) complexes inside the nanowire were conjugated to form a conductive path, improving the conductivity of the nanowire via silica coating–induced structural stabilization. Organic/inorganic hybrid materials consisting of self-assembled functional nanostructures produced from organic and inorganic materials exhibit great promise as novel functional materials because of their high mechanical stability and functionality [26, 27].

In this study, we fabricated organic/inorganic hybrid materials for application as compact and lightweight VOC sensors. First, we constructed conductive self-assembled nanowires by complexing amphiphilic peptides with Co(II)s (Scheme 1a). Thereafter, the surface was coated with photocatalytic titanium oxide to form TiO₂ nanotubes with high photocatalytic activity.

Scheme 1

a Construction of the conductive nanowire. The amphiphilic peptides formed peptide bundles via complexation with Co(II), and the bundle was elongated in the axial direction via dipole–dipole interactions among the peptides. The TiO₂ nanotubes were formed via TiO₂ coating by mineralization on the surface of the peptide nanowires, which was catalyzed by functional groups. b Illustration of the construction of the peptide–AuNP–TiO₂ nanocomposite. The AuNPs were connected by conductive peptide nanowires, after which the nanowire surfaces were coated with TiO₂ via mineralization. c Electron-transfer model of the peptide–AuNP–TiO₂ nanocomposite during VOC decomposition. The photoinduced electrons generated by the TiO₂ coating during the photolysis of VOCs could be efficiently transported to AuNPs without recombination with holes

We constructed a peptide–AuNP–TiO₂ nanocomposite as a VOC sensor (Scheme 1b). AuNPs were connected with conductive peptide nanowires in this nanocomposite. Then, the nanowire surfaces were coated with TiO₂ by mineralization to obtain the peptide–AuNP–TiO₂ nanocomposites. We hypothesized that the photoinduced electrons generated by the TiO₂ coating during VOC photolysis could be efficiently transported to AuNPs via the self-assembled nanowires, preventing electron–hole recombination. Scheme 1c shows a model diagram of electron transfer in the peptide–AuNP–TiO₂ nanocomposites during VOC decomposition.

Furthermore, Wang et al. revealed that electron transfer in a system comprising AuNPs connected with a conductive polymer that acts as a Coulomb island proceeds via the Coulomb blockade effect [28]. Therefore, in our system, the photoinduced electrons generated by VOC photolysis were transported to AuNPs via Co(II) complexes, suppressing the Coulomb blockade effect between AuNPs. This effect induced changes in the electrochemical properties of the composite.

In this study, we constructed peptide–AuNP–TiO₂ nanocomposites, which facilitated easy and high-sensitivity detection of VOCs on the basis of the electrochemical properties of the composite.

Experimental

Fabrication of titanium dioxide nanotubes

The amphiphilic peptide LESEHEKLKSKHKSKLKEHESEL (Pep-GLS) [25] was designed and synthesized via an Fmoc solid-phase method. The hydrophilic amino acids, lysine (K), glutamic acid (E), and serine (S), were concentrated on one side, and the hydrophobic amino acids, leucine (L) and histidine (H), bearing an imidazole group, were concentrated on the opposite side when the peptide formed an α-helical conformation. The three imidazole groups of the H residues were arranged in the same direction as the hydrophobic face. Co(II) was selected as the coordinating metal, as it forms a complex with four ligands consisting of imidazole groups, with a square-planar geometry and two apical counteranions capped on the square planar face. The complexation between Co(II) and the imidazole groups of the α-helical Pep-GLSs resulted in the formation of a stable α-helical bundle [25], which grew axially via dipole–dipole interactions. The carboxyl and amino groups in the hydrophilic face of the peptide promoted TiO₂ mineralization via the charge relay effect between the functional groups. The hydroxyl group of the S residue acted as a binding site for the mineralized TiO₂ on the hydrophilic surface opposite the H residues (Scheme 2). We designed Pep-GLS-L with lipoic acid at the N-terminus of Pep-GLS to fix the peptide on the surface of the AuNPs. Pep-GLS and Pep-GLS-L were synthesized via solid-phase peptide synthesis. The peptides were characterized via matrix laser desorption/ionization time‒of‒flight mass spectrometry (MALDI‒TOF‒MS) on a JMS-S3000 instrument (JEOL Ltd.). Pep-GLS and Pep-GLS-L had molecular weights of 2775.1 and 2962.8, respectively, which are in agreement with the calculated values of 2775.1 and 2962.4, respectively (Fig. S1). Using MALDI-TOF-MS, we confirmed the successful solid-phase synthesis of the designed peptides.

Scheme 2

Schematic images of the amino acid arrangement in the α-helical Pep-GLS: (a) top view and (b) side view

To form a conductive peptide nanowire via complexation between Co(II) and the imidazole groups of Pep-GLS, 10 µL of a Co(II) acetate tetrahydrate (Wako Pure Chemical Industries, Ltd.)–methanol (Nacalai Tesque Co., Ltd.) solution (0.6 M) was added to 10 mL of a Pep-GLS–ethanol (Nacalai Tesque Co., Ltd.) solution (1.0 mM). Furthermore, 2 μL of the Co(II)–methanol solution was added sequentially every 7 days until the molar ratio of Co(II) to the imidazole group, [Co(II)]/[His], reached 0.15. Next, 10 µL of the Co(II)–methanol solution was added every three days until the molar ratio of Co(II) to the imidazole group was 1:4, which was the stoichiometric ratio of the desired complex.

TiO₂ mineralization was performed using a conductive peptide nanowire comprising an amphiphilic peptide–Co(II) complex as a template. Tetraisopropyl orthotitanate (TTIP; Tokyo Kasei Kogyo Co., Ltd.) was used as the TiO₂ precursor. Equal volumes of TTIP were added to the obtained amphiphilic peptide–Co(II) complex solution in a glove box filled with nitrogen and shaken at 25 °C for 20 days. After the reaction, the mixture was centrifuged at 10,000 × g for 30 min (AS185, ASONE) under a nitrogen atmosphere, after which ethanol was added to the precipitate to redisperse it. This process was repeated three times to remove TTIP.

Preparation of peptide-coated gold nanoparticles

To optimize the surface density and particle size of the peptide and AuNPs, respectively, the AuNPs were prepared at a peptide:spacer molar ratio of 1:34. Furthermore, the molar ratio of thiol to Au ions was fixed at 10:23 [29]. Furthermore, Pep-GLS-L, 2-mercaptoethanol, and hydrogen tetrachlorocuprate(III) tetrahydrate (HAuCl₄; Wako Pure Chemical Industries, Ltd.) were added to 100 mL of a 1:4 mixture of deionized (DI) water and dimethylformamide (DMF) and stirred. Additionally, 20 mL of an aqueous solution of sodium borohydride (Wako Pure Chemical Industries, Ltd.) was added as a reducing agent and stirred well. The final concentration of Au ions was 0.2 mM. The amount of sodium borohydride that was added was 10 times greater than the amount of Au ions. After stirring at room temperature for 1 day, the peptide-coated AuNPs produced by the reaction were centrifuged three times at 5 °C and 30,000 × g for 1 h via a high-speed cooling centrifuge (Suprema21, Tommy Seiko Co., Ltd.) to remove unreacted peptides, 2-mercaptoethanol, and inorganic ions; the dispersant was replaced with DI water or ethanol. The resulting peptide-coated AuNP dispersion was used for various measurements and for preparing the nanocomposites.

Fabrication of peptide–AuNP–TiO₂ nanocomposites

An ethanol dispersion of peptide-coated AuNPs and a Pep-GLS–ethanol solution were mixed to obtain 10 mL of ethanol. The total concentration of surface peptide (Pep-GLS-L) and free peptide (Pep-GLS) was fixed at 0.01 mM. The Co(II) acetate tetrahydrate–methanol solution was added in the same manner as that used for the fabrication of the TiO₂ nanotubes. Peptide–AuNP networks in which AuNPs are connected by conductive peptide nanowires were formed via this process.

TiO₂ mineralization was performed using the prepared peptide–AuNP networks as templates with the same procedure employed for the constructing the TiO₂ nanotubes. Equal volumes of TTIP were added to the peptide–AuNP network dispersion in a glove box filled with nitrogen, and the mixture was shaken at 25 °C for 14 days under a nitrogen atmosphere. After the reaction, the precipitate was centrifuged at 10,000 × g for 30 min and redispersed by adding ethanol to the precipitate under a nitrogen atmosphere. This operation was repeated three times to remove unreacted TTIP.

Transmission electron microscopy observations

We observed the morphologies of the TiO₂ nanotubes, peptide-coated AuNPs, peptide–AuNP networks, and peptide–AuNP–TiO₂ nanocomposites under a scanning transmission electron microscope (JEM-z2500, JEOL Ltd.) at an acceleration voltage of 200 kV. The observations were performed in transmission mode without staining, and the images were recorded with an ultrascan CCD camera (US1000, Gatan, Inc.). Sample dispersions of the 10 µL solution were dropped onto elastic carbon-coated scanning transmission electron microscope grids (100 mesh; Ohkenshoji Co., Ltd.). After adsorption for 5 min, excess solution was removed via filter paper.

Elemental analysis and mapping were performed on an energy-dispersive X-ray (EDX) spectrometer (EX-37001, JEOL Ltd.). The same samples used for transmission electron microscopy (TEM) observations were used to map N, C, O, Co, Ti, Au, and S atoms in the bright-field region in scanning transmission (STEM) mode.

The crystal phase of the mineralized titanium oxide of the TiO₂ nanotube and peptide–AuNP–TiO₂ nanocomposites was determined via selected-area electron diffraction (SAED).

Spectroscopic measurements

Fourier transform infrared reflection absorption spectroscopy

We investigated secondary structural changes occurring in the peptides during the construction of the peptide–AuNP networks via Fourier transform infrared (FT–IR) reflection absorption spectroscopy (FTIR-RAS) via an FT–IR spectrometer (6700S JASCO Co., Ltd.) equipped with a high-sensitivity direct reflection attachment (RAS PR0410-H type, JSCO Co., Ltd.). We performed FTIR-RAS on the AuNP dispersion containing the free peptide Pep-GLS before and after Co(II) coordination. The sample dispersions were cast onto Au-deposited glass substrates. The spectra were collected with a Hg/Cd/Te detector (resolution: 2 cm⁻¹; wavenumber range: 1500–1800 cm⁻¹; number of scans: 1024). The fraction of the second-order structure of the peptide was calculated via peak deconvolution of the amide I and II bands [30]. The 1800–1500 cm⁻¹ regions of the spectra were analyzed as the sum of the individual bands. When the Gaussian/Lorentzian ratio was 9/1, the sum of the calculated individual bands exhibited the best fit to the experimental spectra.

Fluorescence spectroscopy

The photocatalytic activity of the obtained TiO₂ nanotubes was evaluated on the basis of fluorescence spectral changes occurring due to dye degradation following ultraviolet (UV) light irradiation via a spectrofluorophotometer (RF-5300pc, SHIMAZU Corp.). The TiO₂ nanotubes (1 mg) were dispersed in 3 mL of a 1.0 × 10⁻⁵ M rhodamine 6GX (Nacalai Tesque Co., Ltd.) aqueous solution and sealed in a quartz cell in which the solution was stirred. An ultrahigh-pressure mercury lamp (UVD-500, Ushio, Inc.) was placed 3 cm from the quartz cell with the stirred solution, and the solution was irradiated with UV light at an output of 500 W for 120 min. A sharp-cut filter (L-39; Asahi Techno Glass) and a bandpass filter (B-390, Asahi Techno Glass) were used to selectively irradiate the sample with UV light in the 390–470 nm range. The excitation wavelength of Rhodamine 6GX was 528 nm, and the fluorescence spectra were recorded every 20 min from 540 to 600 nm. For comparison, the photocatalytic activity of commercially available TiO₂ (ST-31; ISHIHARA SANGYO KAISHA, LTD.) was measured in the same manner. The estimated photocatalysis rate constants of the commercially available TiO₂ and our TiO₂- nanotubes were as follows: semilogarithmic plots of I_t/I₀ against the UV-light-irradiation time were obtained, where I_t and I₀ are the fluorescence intensities of Rhodamine 6GX at 553 nm before and after UV irradiation, respectively. The photocatalysis rate constants were calculated from the slopes of the respective curves.

Electrochemical analysis

We evaluated the formation of Co(II) complexes inside the peptide nanowires as well as the photoexcitation of TiO₂ in the peptide–AuNP–TiO₂ nanocomposite via cyclic voltammetry (CV). The peptide–AuNP–TiO₂ nanocomposite dispersion was dropped onto a comb electrode (012257 IDA electrode (Au) 2 μm, BAS Corp.) with an electrode spacing of 2 µm. The dropping and drying processes were repeated 20 times at room temperature to deposit the peptide–AuNP–TiO₂ nanocomposites on the comb electrode. By employing an optical microscope (BX51, OLYMPUS Corporation, 100× magnification), we confirmed that the electrodes were bridged by the nanocomposites. We applied voltage between the working and reference electrodes via a function generator (NPS-2, NIKKO KEISOKU).

The resulting response currents were measured by a potentiostat (NPGS-2501-10nA, NIKKO KEISOKU). The applied voltage was a triangular wave in the scanning range of −1.0 to 3.5 V at a scanning rate of 0.3 s/V. The redox activity of the TiO₂ that formed on the peptide nanowire surface inside the nanocomposite was evaluated. Furthermore, the photoinduced redox-reaction changes in the surface TiO₂ under UV-light irradiation were evaluated from the shape changes in the CV curves. The comb electrode deposited on the composite was fixed in a plastic cell with a quartz window on one side; the cell was installed in a Faraday cage. For UV light irradiation, commercially available white LED lamps (LDA11LG100V1, TOSHIBA Corp.) were used, and samples were irradiated through a quartz window (Scheme S2). The VOC-sensing ability of the peptide–AuNP–TiO₂ nanocomposites was evaluated from the change in the anodic current of the CVs under light irradiation in the presence of various concentrations of VOCs. Dichloromethane (Wako Pure Chemical Industries, Ltd.) was used as a representative VOC in this study because of its known susceptibility to degradation by TiO₂, as well as its widespread industrial applications; it is also known to pose significant health threats. VOCs containing saturated water vapor were circulated in the cell via a peristaltic pump, and their concentrations were held constant.

Furthermore, the redox reactions of the Co(II) complexes inside the peptide nanowires were evaluated by applying a triangular wave in the operating range of −200 to +500 mV at a scanning rate of 7.0 s/V. The measurements were performed at room temperature.

Rest potential measurements

The VOC-sensing ability of the peptide–AuNP–TiO₂ nanocomposites was also evaluated on the basis of the change in light irradiation–induced rest potential. We used the same apparatus as that used for the CV measurements (Scheme S2). However, the changes in the photoinduced rest potential were recorded in the rest potential mode of the potentiostat. Furthermore, different concentrations of VOCs were introduced, as in the case of the CV measurements.

Results and Discussion

Characterization of titanium dioxide nanotubes and their capacity to photodecompose organic matter

In a previous study [25], we reported that Pep-GLS–formed nanowires made of α-helical bundles that connected axially because of complexation between Co(II) and H residues and due to the macrodipole interactions of the α-helical peptides. The complexes arranged along the wire axis within the nanowire were conjugated, acting as a conductive path. Furthermore, we stabilized the structure of the conductive nanowires via surface coating with silica mineralization and increased the persistence length of the conjugated complexes. In this study, the surface of the conductive peptide nanowires were coated with TiO₂ instead of silica.

First, the conductive nanowires were constructed via the same method used in the previous study. Next, TiO₂ mineralization was performed under mild conditions at room temperature using the conductive peptide nanowire as the template and TTIP as a precursor. We believe that TiO₂ mineralization was induced by the peptide nanowires via a two-step reaction mechanism similar to that of silica mineralization. The first step involves TTIP hydrolysis using the TiO₂ precursor. TTIP is hydrolyzed into orthotitanic acid in ethanol containing trace water. The second step involves a dehydration condensation reaction between the –OH groups of the orthotitanic acid catalyzed by the basic (K) and acidic (E) side chain pairs on the peptide nanowire. TiO₂ is formed during the second step of the reaction, during which a condensation reaction also occurs between the –OH group of the S side chain and the Ti(OH)₄ molecule. Therefore, the TiO₂ layer is fixed on the surface of the peptide nanowire.

The structure of the fabricated TiO₂ nanotubes was evaluated via TEM observation and EDX analysis. The TEM image displayed wire-like structures (Fig. 1a), whereas the EDX mapping images revealed N, Co, and Ti and O, which were assigned to the peptide, to the complex in the nanowire, and to TiO₂, respectively, in the same locations (Fig. 1b). Figure 1c shows the SAED results for the same sample. The diffractions attributed to the (101) plane of the anatase phase (Fig. 1c: A) and the (202) and (212) planes of the rutile phase (Fig. 1c: B, C, respectively) were observed. The

Fig. 1

a TEM image of the TiO₂ nanotubes, (b) EDX element mappings displaying the Co, N, Ti, and O in the same area, and (c) SAED patterns displaying the (101) plane of the anatase phase and (202) and (212) planes of the rutile phase

mineralized TiO₂ layer on the conductive peptide nanowire was confirmed to consist of a mixture of rutile and anatase crystals, indicating the possibility of constructing TiO₂ nanotubes comprising mixed crystals with high pollutant degradation activity [16].

The decomposition of organic matter by the constructed TiO₂ nanotubes was evaluated from the photodegradation rate of a dye (rhodamine 6GX was employed as the dye). Fluorescence was measured at regular intervals during light irradiation in the presence of the TiO₂ nanotubes, and the amount of decomposition was determined from the rate of decrease in the fluorescence intensity. Figure 2a shows the spectral changes that occurred in Rhodamine 6GX in the presence of the TiO₂ nanowire.

Fig. 2

Changes in the fluorescence spectra of rhodamine 6GX induced by photodecomposition using (a) TiO₂ nanotubes and (b) commercially available sintered TiO₂. Each spectrum was normalized to the before-light-irradiation spectrum as a reference. c Semilogarithmic plots of the relative fluorescence intensity, I_t/I₀, as a function of the light irradiation time for each system. I_t and I₀ are the maximum fluorescence intensities at each light irradiation time and before irradiation, respectively

Each spectrum was normalized to the before-light-irradiation spectrum as a reference. For comparison, photodecomposition experiments were conducted using commercially available sintered TiO₂, a photocatalyst (Fig. 2b). For each system, the ratio I_t/I₀ was plotted against the irradiation time on a semilogarithmic basis (Fig. 2c). Here, I_t is the maximum fluorescence intensity at each light-irradiation time, and I₀ is the maximum fluorescence intensity before irradiation. The photocatalysis rate constants of rhodamine 6GX catalyzed by TiO₂ nanotubes and commercially available sintered TiO₂, as calculated from the respective slopes, were 1.13 × 10⁻³ and 4.08 × 10⁻⁴ min⁻¹, respectively. The results confirmed that the obtained TiO₂ nanotubes exhibited approximately 2.8 times greater photoresolution than the sintered photocatalyst. This might have occurred because the crystalline phase of the TiO₂ nanotubes consisted of mixed crystals of rutile and anatase, which enhanced decomposition activity even without sintering. Furthermore, the photoexcited electrons produced by light irradiation of the TiO₂ layer on the conductive peptide nanowire were transported to the Co(II) complex inside the nanowires, stabilizing the holes in TiO₂ and further improving TiO₂ decomposition activity.

We believe that we can construct VOC sensors that detect VOC decomposition as an electrical signal on the basis of the organic matter-decomposing activity of TiO₂ nanotubes.

Structures of the peptide–AuNP–TiO₂ nanocomposites

First, we prepared peptide-coated AuNPs via a reduction reaction of HAuCl₄ in the presence of 2-mercaptoethanol and Pep-GLS-L. We evaluated the size of the AuNPs formed via TEM observations. Figure 3a, b show the TEM and HR-TEM images of the peptide-coated AuNPs in ethanol, respectively. HR-TEM revealed the presence of lattice fringes of 2.36, 2.04, and 1.44 Å, corresponding to the planar spacings of the (111), (200), and (220) planes of Au, respectively. The diameters of the AuNP cores were measured from the TEM images, after which a histogram of the particle size was generated (Fig. 3c). The particle size distribution of the AuNP core was 3.95 ± 0.72 nm. These results confirmed the possibility of fabricating homogeneously dispersed peptide-coated AuNPs in ethanol.

Fig. 3

a TEM image of peptide-coated AuNPs in an ethanol dispersion and (b) HR-TEM image showing the lattice fringes of the peptide-coated AuNPs. c Particle size distribution of the AuNP cores obtained from the TEM images of the peptide-coated AuNPs in ethanol. We measured the diameters of the AuNP cores in the TEM images and created a histogram

We calculated the amount of peptide coated on the AuNPs as follows [31]:

$$\frac{1}{6}\pi {d}^{3}\rho n{N}_{A}\times \frac{1}{M}\times {10}^{-21}\,\left(\frac{{{\rm{molecules}}}}{{{\rm{particle}}}}\right)$$

where d, ρ, n, and M represent the diameter of the AuNP core obtained via TEM observation (d = 3.95 nm), the density of Au (ρ = 19.3 g/cm³), the peptide/Au weight ratio (n = 7.9 × 10⁻²), and the molecular weight of the peptide (M = 2962.8 g/mol), respectively. N_A represents the Avogadro number. The peptide/Au weight ratio was calculated from the gravimetric concentration of the AuNP dispersion, 0.136 mg/mL (Table S4), and the concentration of the peptide coating on AuNPs, 0.01 mg/mL (Fig. S3), in the aqueous dispersion. From these values, the number of peptides per AuNP was calculated, confirming that each AuNP was coated with approximately 10 peptides. This indicated that the fabricated peptide-coated AuNPs could form three-dimensional (3D) assemblies connected by Co(II)-coordinated peptide nanowires. Next, Pep-GLSs and Co(II) were added to the obtained peptide-coated AuNP dispersion to construct the 3D network structure via the formation of Co(II)-coordinated peptide nanowires between AuNPs. The molar ratio of the coated peptide to the added free peptide was 1:19. At this ratio, the NPs were connected with six 4α-helical bundles, resulting in an interparticle spacing of ~20 nm. We evaluated the conformational transition of the peptide following the addition of Co(II) during the construction of the peptide–AuNP network on the basis of the FTIR-RAS results. In a previous study, circular dichroism spectroscopy was employed to evaluate the conformational transition [25]. The peptide-coated AuNPs formed large assemblies following the addition of Co(II). Therefore, we cast the obtained peptide–AuNP networks on Au substrates and characterized their structure via FTIR-RAS. Figure 4 shows the FTIR-RAS spectra of the mixtures containing peptide-coated AuNPs and free peptides before and after the addition of Co(II). The peak of the acetate ion, which is the counterion of Co, was detected in the FTIR-RAS spectra only after Co(II) addition. Notably, the secondary structure of the peptides increased and the α-helix and β -sheet fractions decreased, respectively, following the addition of Co(II) to the system.

Fig. 4

FTIR-RAS of peptide-coated AuNPs containing Pep-GLS (a) before and (b) after the addition of Co(II). The red, blue, and green lines show the peak deconvolutions of the amide I band into α-helical, β-sheet, and random coil, respectively. The yellow line shows the peak of CH₃COOH⁻, which is the counterion of Co(II). The peak indicated by the broken line might be derived from the antiparallel β -sheet conformation or from DMF that did not volatilize

In a previous study [25], we observed a similar conformational transition when peptide nanowires were formed. This result might have been due to the formation of the four α-helical bundle structures generated by the complexation between the four His residues and Co(II), which stabilized the individual α-helical conformations. Moreover, the individual particles were more aggregated in the peptide–AuNP networks formed by Co(II) addition (Fig. S5a) than in the peptide-coated AuNPs formed in ethanol (Fig. 3a). Furthermore, we performed EDX mapping of the same area for which TEM observations were performed (Fig. S5b). The “S” derived from the coated Pep-GLS-L and 2-mercaptoethanol and the “Au” derived from AuNPs were confirmed to occur in the same location. Furthermore, peptide-derived “N” and added “Co” were observed in the same area. This further indicated that the peptide-coated AuNPs were connected by conductive peptide nanowires containing Co(II) complexes.

We constructed a peptide–AuNP–TiO₂ nanocomposite by TiO₂ mineralization using peptide–AuNP networks as the template. The structure of the nanocomposite was observed via TEM (Fig. 5a). The crystalline phases of TiO₂ and Au were characterized via SAED (Fig. 5b), and the elemental distribution was evaluated via EDX analysis (Fig. 5c). The TEM image revealed the presence of substances with different contrasts around the AuNPs, which appeared dark and black (Fig. S5a). The SAED pattern of the area circled red in the TEM image revealed diffractions that were assigned to the rutile (111) and (301) and anatase (200) phases of the TiO₂ and gold (200) crystals, respectively. It has been confirmed that mixed crystals of rutile and anatase exhibit high pollutant-degrading activities. This EDX mapping further confirmed that “Au” and peptide-derived “N,” nanowire-derived “Co,” and “Ti” were observed in the same locations. These results confirmed that TiO₂ consisting of the rutile and anatase phases was prepared in the peptide–AuNP network connecting the conductive nanowires by TiO₂ mineralization.

Fig. 5

a TEM image, (b) electron diffraction pattern, and (c) EDX elemental mappings showing Au, Co, N, and Ti in the peptide–AuNP–TiO₂ nanocomposite

Electrochemical properties of the peptide–AuNP–TiO₂ nanocomposite

To evaluate the electrochemical properties of the obtained peptide–AuNP–TiO₂ nanocomposites, the nanocomposites were deposited on comb electrodes. Our optical microscopy observations confirmed that the electrodes were connected by peptide–AuNP–TiO₂ nanocomposites (Fig. S6). CV measurements of the peptide–AuNP–TiO₂ nanocomposites were performed under two conditions. First, to determine the redox potential of TiO₂, a triangular wave with a scanning range of −1.0 to 3.5 V was applied at a scanning rate of 0.3 s/V under Condition a (Fig. 6a).

Fig. 6

Cyclic voltammograms of the peptide–AuNP–TiO₂ nanocomposites obtained under different measurement conditions. a Triangular wave at a scanning range of −1.0 to 3.5 V was applied at a scanning rate of 0.3 s/V. b) Low sweep rate (7 s/V) and narrow sweep range (−200 to 500 mV compared with Condition a)). The redox potential of TiO₂ was measured in the dark (black line in a)) and under light irradiation (red line in a))

For Condition b, a triangular wave with a scanning range of −200 to 500 mV was applied at a scanning rate of 7 s/V to evaluate the redox potential of the Co(II) complex in the peptide nanowire (Fig. 6b). In the measurement under Condition a, the anode and cathode peaks appeared at +2.85 and −0.22 V, respectively. These values agreed well with the redox potentials of TiO₂, which are +2.9 and −0.3 V [32]. Conversely, no TiO₂-derived redox peaks could be confirmed in the CV measurement at a low sweep rate and under a narrow sweep range (Condition b)), although the redox peaks derived from the Co(II) complex inside the peptide nanowire were observed at +0.22 V and +0.19 V [33].

Furthermore, the anodic and cathodic currents of TiO₂ increased following photoirradiation by the LED lamp. This was considered to occur due to the photoexcited electrons emitted by TiO₂ upon photoirradiation. These results indicated the presence of TiO₂ and Co(II) complexes as conductive pathways in the peptide–AuNP–TiO₂ nanocomposite. This confirmed that the obtained nanocomposite could act as a sensor for detecting electron transfer associated with the decomposition of organic matter by TiO₂ as an electrical signal.

Volatile organic compound-sensing capacity of the peptide–AuNP–TiO₂ nanocomposites

CV measurements were performed by exposing the comb electrode containing the peptide–AuNP–TiO₂ nanocomposites to DCM at arbitrary VOC concentrations. As moisture in the air acts as a medium through which TiO₂ decomposes organic matter, we introduced VOCs saturated with water vapor. The current increased at the anodic and cathodic peaks with increasing VOC concentration (Fig. S7a, S7b). The anodic peak currents were plotted for each VOC concentration (Fig. S7c). We observed a rapid increase in the anodic current at a VOC concentration of ~5 ppm, and saturation was reached at ~50 ppm. However, the changes in the anodic current were negligible.

VOC sensing in real environments requires simplification of the measurement equipment. Therefore, we attempted a sensing method that does not require sweeping of the applied potential. Specifically, we measured the VOC-sensing ability via changes in the resting potential of the peptide–AuNP–TiO₂ nanocomposites. Via this method, VOC sensing could be performed using only a voltmeter. The resting potential of the nanocomposite increased by ~25 mV during light irradiation when 2 ppm VOC gas was introduced (Fig. 7a). The potential change upon light irradiation became constant at ~20 s, indicating a rapid response. This change was caused by the consumption of light-irradiation-generated holes inside TiO₂, and electrons were transferred to AuNPs via Co(II) complexes inside the peptide nanowires. This might have occurred due to an overvoltage induced by the accelerated oxidation reaction [34]. We measured the changes in the resting potential of the nanocomposite due to light irradiation at various VOC concentrations (Fig. 7b). The resting potential increased rapidly with increasing VOC concentration up to 100 ppm and reached saturation at 250 ppm. This result confirmed the fabrication of a VOC sensor that can be used at concentrations of < 100 ppm. The control and permissible concentration of DCM specified by the Japan Society for Occupational Health is 50 ppm [35, 36]. Therefore, this sensor can effectively measure DCM in the environment where it is used. In this experiment, the introduced gas was saturated with water vapor; therefore, humidity adjustments may be necessary.

Fig. 7

a Changes in the resting potential of the peptide–AuNP–TiO₂ nanocomposite caused by light irradiation. The nanocomposites were exposed to saturated water vapor containing 2 ppm DCM. The light-irradiation period is shown in yellow. b Changes in the resting potential of the peptide–AuNP–TiO₂ nanocomposite exposed to saturated water vapor containing VOC (DCM) at various concentrations under light irradiation

Conclusions

In this work, we report the construction of organic/inorganic hybrid nanocomposite materials that exhibit high photocatalytic activity and novel VOC-sensing capabilities with low environmental impact.

First, TiO₂ nanotubes were fabricated via TiO₂ mineralization under mild conditions at room temperature using conductive peptide nanowires containing Co(II) complexes as the template. The obtained TiO₂ nanotubes were highly functional photocatalysts that did not require environmentally burdensome methods, such as sintering, and could pave the way for further development in the field of photocatalysis.

Next, we fabricated a VOC sensor made of peptide–AuNP– TiO₂ nanocomposites in which AuNPs were connected by the constructed TiO₂ nanotubes. The connected conductive nanowires could transport photoexcited electrons from surface TiO₂ to AuNPs through the Co(II) complexes inside the nanowires during pollutant decomposition. This electron transfer inhibited the recombination of the generated holes and photoelectrons, enabling the detection of pollutants, such as VOCs, with high sensitivity via changes in the electrochemical properties of the nanocomposite. With respect to the electrochemical changes in the nanocomposite, the anodic and cathodic currents, as well as the resting potential, changed with the concentration of VOC under light irradiation. In particular, the potential change under light irradiation occurred at a low VOC concentration of ~2 ppm, indicating the high sensitivity and responsiveness of the sensor. This VOC-sensing ability was within the control and permissible concentration ranges established by the Japan Society for Industrial Hygiene. We believe that our VOC-sensing method is superior to other methods, as it only requires a voltmeter. However, in this study, VOC sensing was performed under saturated water vapor. As moisture is involved in the VOC decomposition process by TiO₂, the sensing ability of our sensor may change depending on humidity levels.

This sensing system has a small sensing component and can be measured by a simple device, making it valuable for the development of portable VOC sensors that can be incorporated into other devices.