Novel TiO₂/graphene oxide functionalized with a cobalt complex for significant degradation of NO_x and CO

Abstract

We are reporting a well-designed nanostructured composite, consisting of titanium dioxide (TiO₂) and cobalt (Co) imidazole (Im) complex functionalized graphene oxide (FGO). The Co–Im complex was attached on the GO by covalent bonding of Im with the graphene functional groups. The films were characterized by XRD, XPS, TEM, UV-vis, FTIR and Raman spectroscopy in order to subsequently prove the evidence of hybridization of GO with the Co–Im complex. The band gap calculated for the TiO₂ thin film was 3.10 eV, 2.96 eV for TiO₂/GO and 2.776 eV for TiO₂/FGO. Experimental results confirmed high-percentage deterioration of NO_x (51%) and CO (46%). The photodegradation experiments revealed a significant photocatalytic character of TiO₂, physical adsorption of GO, and the affinity of cobalt to undergo complex formation with pollutant gases. The TiO₂/FGO composite increases air pollutant degradation by nearly three times that of the bare TiO₂ thin film.

---

1. Introduction

Currently, human beings have a severe problem called air pollution, particularly contributing to asthmatic diseases of children and old people who live in populated cities. Gases such as NO_x (NO, NO₂) and CO are toxic and have many dreadful impacts on public health such as cancers.^1–5

National Ambient Air Quality Standards (NAAQS) have limited the maximum indoor level of carbon monoxide to 35 ppm and for nitrogen dioxide to 100 ppb (in 1 hour).^6,7 Scientists offered and used different techniques for the removal of NO_x and CO. Selective catalytic reduction (SCR) has been proposed with a high amount of NO elimination; however, the procedure is cost-effective and requires elevated temperatures.^8–10

NO_x and CO are difficult to oxidize and remove; consequently, it is essential to implement new and economical methods for their destruction. Photocatalysis could be one alternative, because it has a low price and efficiency, the energy source (both solar and artificial) is light, the catalysts are not toxic (semiconductors such as TiO₂), and the decomposition products are harmless (CO₂ and H₂O).¹¹ Fujishima and Honda explored the photocatalytic nature of TiO₂,¹² which is preferable to other semiconductor oxides because of its chemical stability, low cost and lack of toxicity.^13,14 Detailed mechanisms of TiO₂ photochemical reactions are well documented.^15–17

The successful synthesis of a TiO₂ photocatalyst is essential for achieving the best photocatalytic activity. We have selected the film instead of the powder form, because of its superior properties like optical transparency, further mechanical stability, durability and adhesion on the surface. Additionally, a thin film can be easily separated and assembled for regeneration or replacement purposes.^18–23

TiO₂ thin films could be arranged by numerous methods, such as sol–gel,^24,25 hydrothermal,²⁶ anodizing Ti foil,²⁷ chemical vapor deposition,²⁸ metal–organic chemical vapor deposition,²⁹ pulsed laser deposition³⁰ and RF magnetron sputtering,³¹ with diverse crystalline structures that greatly rely on the preparation method. Among them, RF magnetron sputtering is one of the easiest to industrialize.³² This technique is suitable for optical coatings, owing to their high density, high adhesion, hardness and uniformity.³³

TiO₂ in the anatase form demonstrates photocatalytic activity only after UV light irradiation (λ < 384 nm).^34,35 Practical composites may be created to overcome the large band gap of TiO₂ (3.2 eV). Li et al. proved that combining carbon nanotubes (CNTs) with TiO₂ could assist in the separation of electron–hole charges.³⁶ Choy et al. reported a TiO₂/clay composite with excellent photocatalytic activity.³⁷ TiO₂ composited with zeolite templated carbon (TiO₂–ZTC) successfully adsorbed organic molecules due to its high surface area.³⁸ TiO₂ composited with mesoporous silicate MCM-41 exhibited good photocatalytic activity for oxidizing phenol.³⁹ Hybrid SiO₂–TiO₂ material has been synthesized and characterized as a composite with a tunable surface area (500–1170 m² g⁻¹);⁴⁰ however, in our case, we needed a composite with the ability of surface functionalization to handle both photocatalytic activity and complex formation with pollutants. Graphene oxide (GO) is our choice to make a composite due to its high surface area, non-toxicity and functionability. GO is analogous to benzene and polycyclic aromatic hydrocarbons, consisting of sp² and sp³ carbon atoms.^41–43 GO as an absorbent, owing to its large theoretical surface area (2630 m² g⁻¹), is very fascinating. As a result, GO is potentially applicable in environmental remediation as an effective sorbent for pollutants. Recently, GO has been used to remove MB,⁴⁴ fluoride,^45,46 aromatics,^47–49 dyes,^50,51 arsenic⁵² and other heavy metals from aqueous solutions, with high adsorption amounts and fast adsorption rates.^53,54

GO is capable of functionalization through its alcohol, epoxide, carboxyl and ketone groups.^55–57 Chemically modified graphenes (CMGs) have been achieved by adding functional groups to GO.^58–63 Prior to our research, only a few cases of research employing semimetals or metals were released on CMGs.^64,65 Functionalization of GO with cobalt has been used for various applications like an unusual catalytic effect,⁶⁶ enhanced photocatalytic H₂ production⁶⁷ and ameliorated electrochemical performances.⁶⁸ Furthermore, in particular, cobalt has attracted extensive attention because of its outstanding properties like being an earth-abundant metal,⁶⁹ having interfacial hole transfer promotion, and consequently a band gap reduction,^70,71 and above all, Co(II) ions are excellent metal centers in complex formation with NO_x and CO.⁷² Interactions between cobalt and gases such as CO and NO lead to complex formation, and thereby gaseous pollutants are eliminated.^73–75 An effective approach is to fabricate transition metal compounds for complex formation with NO and CO ligands on the GO surface.

As a matter of fact, this paper deals with the synthesis of a composite photocatalyst in the form of a thin film, consisting of TiO₂ synthesized by RF magnetron sputtering on a quartz surface with uniform thickness; subsequently, functionalized GO with a cobalt–imidazole complex is deposited on the TiO₂ surface with the purpose of air purification.

2. Experimental

2.1. Synthesis

2.1.1. Preparation of the TiO₂ thin film. The TiO₂ thin film was prepared using the RF magnetron sputtering method with a Ti target (grade: 99.99%, diameter: 100 mm, thickness: 5 mm) as the source material, Ar (99.995%) was the sputtering gas and O₂ (99.995%) was the reactive gas. Unheated substrates were used for deposition. The RF generator operated at a frequency of 13.56 MHz and at 200 W. A turbo pump was utilized for vacuuming the stainless-steel chamber with a pressure less than 2.7 × 10⁻³ Pa. Mixed gases of Ar and O₂ were purged to the vacuum chamber with a pressure ratio of 9 : 1 while the total pressure was 1.1 Pa. Before the deposition, the chamber was kept in the pre-sputtering period for 20 min for the purpose of removing contaminants and stabilizing the deposition parameters.

Quartz substrates were ultrasonically washed with acetone, ethanol and deionized water, each for 15 min. After the substrates were cleaned, they were placed on a sample holder facing the target at a fixed distance of 50 mm in a vacuum chamber. A TiO₂ film with a thickness of 900 nm was deposited onto each quartz substrate. In our experiment, the deposition rate was 2.5–5.8 nm min⁻¹ and the deposition time was approximately 160 min. The deposited films were annealed at 550 °C for 1 h in air.

2.1.2. Preparation of GO. The adjusted Hummers method was employed to oxidize graphite.⁷⁶ H₂SO₄ (50 mL) was added into a 500 mL flask containing graphite (2 g) and NaNO₃ (1 g) at 0 °C in an ice bath. KMnO₄ (6 g) was added to the above mixture slowly (over 1 h) to prevent the temperature increasing. The suspension was stirred continuously for 2 h, diluted with 350 mL of deionized water, and left for 2 days before H₂O₂ (30%) was added in order to reduce any residual manganese dioxide and permanganate and to solubilize manganese sulfate. This mixture was filtered with a glass filter, followed by washing with dilute HCl (10% vol). GO was dried under vacuum at ambient temperature.

2.1.3. Acylation of GO. GO was acylated with thionyl chloride (SOCl₂) to introduce acyl chloride groups on the surface. Acyl chloride-modified GO was prepared by reacting GO (0.2 g) with SOCl₂ (85 mL) and dimethylformamide (DMF; 4 mL) at 70 °C for 24 h inside a dry Ar glove box to activate the carboxylic units by forming the corresponding acyl chlorides. The reaction took place between the C=O group in the carboxyl and the chloride ions (eqn (1)). Anhydrous tetrahydrofuran (THF) was used to remove excess oxalyl chloride from the remaining acyl chloride-functionalized GO (GO-COCl), and then the product was washed with anhydrous toluene three times. The resulting solid material was centrifuged and dried at room temperature under vacuum.

SOCl₂ + RCO₂H → RC(O)Cl + SO₂ + HCl (1)

2.1.4. Modification of GO with imidazole. Acyl chloride-functionalized GO (0.2 g) was re-dispersed in DMF (40 mL) in a dry Ar-filled glove box under dry conditions with excess imidazole (1 g) and refluxed at 70 °C for 72 h. Imidazole (C₃H₄N₂) is a weak base with two N atoms, which was chemically deposited onto the resulting acylated GO through an acid–base reaction. The resulting brown-gray solid was separated by filtration. The product was washed with methylene chloride (4 × 20 mL) and dried under vacuum (see ESI Fig. S1†).

2.1.5. Functionalized GO with the cobalt complex. GO functionalized with cobalt was obtained from the reaction of imidazole-functionalized GO with CoCl₂·6H₂O (0.1 mmol, 2.379 g) in methanol (50 mL). The mixture was refluxed for 1 h and then reserved for a few days at ambient temperature; pink cobalt complex crystals emerged. A cobalt tetrakis imidazole chloride complex [Co(Im)₄]Cl₂ was obtained (see ESI Fig. S2†).

2.1.6. Deposition of functionalized GO on TiO₂. Functionalized GO (FGO) was suspended in ethylene glycol (1 mg mL⁻¹). Then, the solution was subjected to ultrasonication with a probe-type sonicator for 2 h in order to obtain a uniform aqueous solution. The FGO suspension was spin-coated at 2000 rpm for 30 s with a HOLMARC Spin coater model: HO-TH-05 on TiO₂-sputtered quartz to achieve a 30 ± 2 nm FGO film thickness⁷⁷ (see ESI Fig. S3†).

2.2. Reagents

Graphite (powder, <45 μm, ≥99.99%, Sigma-Aldrich), potassium permanganate (KMnO₄, 158.03 g mol⁻¹, extra pure, Merck), sulfuric acid (H₂SO₄, 95%, extra pure, Merck), hydrogen peroxide (H₂O₂, 30%, Merck), sodium nitrate (NaNO₃, 84.99 g mol⁻¹, ≥99.0%, Merck), imidazole (C₃H₄N₂, 68.08 g mol⁻¹, ≥99%, Sigma-Aldrich), cobalt chloride (CoCl₂·6H₂O, 237.90 g mol⁻¹, ≥98%), methanol (CH₄O, 32.04 g mol⁻¹, Merck), THF (C₄H₈O, 72.11 g mol⁻¹, Merck), thionyl chloride (SOCl₂, 99.5%, 118.97 g mol⁻¹, Sigma-Aldrich), and dilute hydrochloric acid (HCl, 37%, Merck, diluted in H₂O) were used in this work.

2.3. Instrumentation

The crystal structure of the TiO₂ thin film was characterized by X-ray diffraction with Cu radiation (X’Pert Pro MPD, PANalytical). The accelerating voltage and applied current were 40 kV and 40 mA, respectively and an incident angle of 1° was used in the region of 2θ = 20–80°.⁷⁸ Field-emission scanning electron microscopy (FESEM) (Tescan Mira2) was applied to understand the morphology of the products. Before the FESEM measurements, the sample surface was covered with a thin gold film to prevent charge build up. X-ray photoelectron spectroscopy (XPS) was performed on an X-ray photoelectron spectrometer (ESCALAB 250; VG Scientifics) with an Al Kα irradiation source (1486.7 eV) performed at 15 kV and 10 mA. A hemispherical energy analyzer (Specs EA 10 Plus) performing in a vacuum better than 10⁻⁷ Pa was used to determine the core level binding energies of the photoelectrons emitted from the surface. All of the peaks were deconvoluted using the SDP software program (version 4.1) with an 80% Gaussian-20% Lorentzian peak fitting. All binding energy (BE) values were calibrated by fixing the C(1s) core level with a BE of 285.0 eV. Fourier-transform infrared (FTIR) spectra were documented by Bruker tensor 27 and Raman spectra of the samples were scripted by Horiba Jobin Yvon at λ = 532 nm with a laser power of 1.7 mW, a 100× objective lens, and a 0.9 NA. The Brunauer–Emmett–Teller (BET) specific surface areas of the samples were measured with BEL Belsorp-Mini II 475.⁷⁹ Adsorption–desorption isotherms of N₂ were measured at 77 K. The BET surface area was calculated in the relative pressure range of P/P₀ = 0.05–0.2 and the pore size dimensions were calculated using the Barrett–Joyner–Halenda (BJH) model.⁸⁰ Samples were degassed for 12 h at 110 °C and 10⁻⁶ Torr. Transmission electron microscopy (TEM) images were collected with a Philips EM 208 microscope at 100 kV. The diffuse reflection spectroscopy (DRS) and absorbance of the films were recorded using an Avantes Avaspec-2048-TEC spectrometer in the wavelength range of 200–900 nm. The optical band gap, E_g, can be related to the absorption coefficient, α [α = 2.303 log(T/d)], where d is the thickness and T is the transmission, using the Tauc expression, and it can be obtained by extrapolating the linear portion of (αhν)² as a function of the photon energy (hν).⁸¹

2.4. Photocatalytic measurements

Photocatalytic activity data were collected by the degradation of CO and NO_x at the ppm level in a batch-flow reactor at room temperature. TiO₂ thin film photocatalyst plates were placed in the air-tight photoreactor vessel and their positions were fixed in order to have a constant UV intensity of 3.0 mW cm⁻². The thin film photocatalysts contribute to the apparent surface area of 50 cm². The vessel was placed under a UV reactor equipped with an 8 W UV lamp (OSRAM, Italy). One small ventilator was attached near the lamp to avoid the temperature changing. The assembled system was regularly checked for leaks by means of immersing the system in water or by applying a leak-test solution at every connection (see ESI Fig. S4†).

Analytical grade gases were acquired from a compressed gas stainless-steel cylinder at a concentration of 1 ppm NO_x and 50 ppm CO (N₂ balance, STG gas) with traceable National Institute of Standards and Technology and National Association of Testing Authorities standards. Gas regulators were supplied from Messer. The initial concentration of NO_x was diluted to about 1 ppm and CO to 50 ppm by the air stream. The relative humidity of the flow was measured and controlled at 40%, using a water-filled Drechsel bottle. The flow rate was fixed at 200 mL min⁻¹ by a mass-flow controller. After fulfilling the adsorption–desorption equilibrium the lamp was turned on. The concentrations of NO_x and CO were measured by a Modular gas analyzer (GMS800 gas analyzer, SICK), which monitors NO_x and CO with a sampling rate of 0.7 L min⁻¹.^82–84

3. Results and discussion

3.1. Characterization

3.1.1. XRD measurements. Fig. 1a illustrates the XRD patterns of the TiO₂ film deposited on quartz; the XRD peaks are shown at 25.3°, 37.8°, and 48.6°, which confirm the anatase phase planes at (101), (004), and (200) for the nanostructured TiO₂ thin film, respectively. The maximum background signal at about 20° is due to the amorphous phase of the quartz substrate.

Fig. 1 (a) XRD pattern of the TiO₂ thin film. (b) XRD plots of GO and FGO.

In previous reports by other scientists, it was shown that the electronic and optical properties of the anatase phase enable it to be chosen as a photocatalyst rather than the rutile or brookite phases, which is favorable for our research strategy.^85,86

Fig. 1b demonstrates the XRD patterns of GO and FGO; in the XRD pattern of GO, a big peak can be seen (2θ = 10.6°; d-spacing = 7.62 Å), according to the (001) diffraction peak of GO.^87–89 The large interlayer spacing (7.62 Å) of GO corresponds to the oxygen functional groups of GO such as carboxyl, hydroxyl and epoxy. The broad peak that appears at approximately 42.5° corresponds to the (101) plane of the graphitic framework that remains in the GO structure.^90,91

In the XRD pattern of FGO (Fig. 1b), the main peak can be recognized at 2θ = 23.79°. This gives an interlayer spacing of approximately 3.7 Å which is smaller than the 7.62 Å for GO. Additionally, it has a broad shoulder at 2θ = 18.5°, which was presumably induced by the bimodal or multimodal character of the interlayer spacing of the FGO powder. The XRD pattern of FGO shows a broad peak centered at 2θ = 16.81° (d-spacing = 5.27 Å), which is lower than the d-spacing of GO (7.62 Å) due to functionalization.

The diffraction peaks attributed to cobalt and imidazole had a high impact on the GO surface, which is detectable by XRD and proves the functionalization of GO. In comparison, Yu et al. reported XRD results assuming the incorporation of AuNPs in GO instead of chemical bonding.⁹² As well as the study on Ag-doped GO, Kim et al. observed that XRD shows no diffraction peak for the silver content, indicating the dispersion of Ag in GO.⁹³ In our research on the covalent bonding of cobalt, the diffraction peak at 2θ = 10.6° is diminished in intensity, most probably because the carboxylic functional groups on the surface of GO were removed after functionalization.

3.1.2. FESEM and TEM images. The FESEM image of the TiO₂ thin film (Fig. 2a) shows particle size uniformity with a small grain size (D = 19.72 nm). The morphology of the particles is dense, resulting in an increasing mechanical strength. The surface coating is uniform without cracks.

Fig. 2 (a) FESEM photograph of the top surface of the TiO₂ thin film. (b) FESEM image of GO. (c) FESEM image of FGO. (d) TEM image of FGO.

In the FESEM image of GO in Fig. 2b, a layered structure can be observed, which consists of ultrathin and homogeneous graphene sheets.⁹⁴ The corresponding films are gently folded with ordered accumulation.

The morphology of FGO was characterized using FESEM (Fig. 2c). The FGO is porous, fluffy and irregular, and layered with folded wrinkles. The FGO is rather more exfoliated than GO and the outer layers are considerably delaminated.⁹⁵

The TEM image of nano-sized FGO is similar to GO with a high surface area which is crucial for the catalytic purpose (Fig. 2d).⁹⁶ FGO is a transparent sheet with folding edges. These wrinkles on the graphene sheets are due to the attachment of functional groups on both sides of the carbon grid.

3.1.3. Nitrogen adsorption–desorption isotherm. Fig. 3 illustrates the nitrogen adsorption–desorption isotherm of FGO at 77 K and at a saturated vapor pressure (P₀) of 89.785 kPa. Keeping the temperature constant and varying the external gas pressure, by recording the amount adsorbed at each pressure, the adsorption isotherm can be obtained. Pore condensation happens at a gas pressure P lower than the bulk saturated vapor pressure P₀. The isotherm of N₂ adsorption–desorption at 77 K is typical of type-IV with a hysteresis loop, which indicates the existence of uniform pore distributions. The isotherm is sharp, and the hysteresis loop at P/P₀ > 0.87 can be observed; therefore the catalyst is principally mesoporous, according to the IUPAC classification on pores (2–50 nm). The specific surface area calculated from the BET method is 205.8 m² g⁻¹, and the average pore radius and average pore volume calculated by the BJH method are 3.09 nm and 0.501 cm³ g⁻¹. The surface area of FGO (ca. 206 m² g⁻¹) is lower than that of GO (ca. 487 m² g⁻¹). Despite the fact that FGO has mostly sustained its layered structure, functionalization likewise takes place at the edges of the GO particles. Therefore, parts of the basal planes near the edges snap together, due to π–π interactions, which narrows the interlayer distance and the specific surface area decreases. The specific surface area is one of the necessary tools to individualize graphene-based materials as absorbents and catalysts.⁹⁷

Fig. 3 N₂ adsorption–desorption isotherm of FGO at 77 K and at a saturated vapor pressure P₀ of 89.785 kPa, with the hysteresis loop at P/P₀ > 0.87.

3.1.4. XPS analysis. X-ray photoelectron spectroscopy (XPS) is a practical instrumentation to detect functional groups and elemental compositions. Fig. 4a demonstrates the full range XPS survey of FGO. The N1s peak of FGO at 399 eV (15.5%) reveals the existence of imidazole and the Co2p peak at 779 eV (1.1%) represents the presence of cobalt. Successful functionalization of GO with the cobalt–imidazole complex was confirmed by XPS. The peak related to the oxygen functional groups, O1s (532 eV), in the spectrum of FGO is minor (9.5%); however, the spectrum still displays a small peak in that region, indicating a higher removal degree of oxygen-containing functional groups. The peak area calculation of C and O from the XPS survey scan in Fig. 4a reveals that the ratio of C to O atoms increased from 2.7 in GO to 7.12 in FGO, according to the functionalization.

Fig. 4 (a) XPS survey spectrum of FGO, (b) XPS C1s region spectrum for FGO with the results of curve fitting, and (c) N1s region of FGO with the results of curve fitting.

Fig. 4b shows the high resolution XPS spectrum of C1s after functionalization. The noticeable but small peak at around 288.5 eV is attributed to a small portion of the carboxyl groups (1.1% of total C) in the FGO due to functionalization. The appearance of a new peak at 285.8 eV (C–N) further confirms the presence of imidazole nitrogen in the graphene sheets.

The N1s high resolution spectrum of the FGO sample is shown in Fig. 4c. The deconvoluted N1s spectrum shows three types of N atom. The one at 401.0 eV corresponds to N–Co, the second peak at 399.8 eV stands for N–C, and the peak at 398.7 eV belongs to N=C which further confirms the chemical bonding between the GO surface imidazole with Co.

3.1.5. Raman spectroscopy. Raman spectroscopy is a helpful instrumentation to characterize GO. GO frequently engages in two main aspects: a D peak at 1349 cm⁻¹ and a G peak at around 1597 cm⁻¹, originating from the first-order scattering of the E2g phonon of sp²-hybridized C atoms and a breathing mode of k-point photons with A1g symmetry, respectively.⁹⁸ In the Raman spectrum of FGO (Fig. 5), the D and G bands are correspondingly shifted to 1353 and 1604 cm⁻¹ after GO functionalization. Thus, it could be deduced that the functionalization induced a certain decrease in the size of the in-plane sp² domains, and increased the edge planes, as well as expanding disorder in the prepared FGO.

Fig. 5 Raman spectrum of FGO.

3.1.6. FTIR spectroscopy. The FTIR spectrum of GO in Fig. 6 indicates a broad peak at 3409 cm⁻¹ in the high frequency area together with a sharp peak at 1630 cm⁻¹, corresponding to the stretching and bending vibrations of water OH molecules adsorbed on GO, respectively. The peak at 846 cm⁻¹ is attributed to the stretching vibration of C=O in the carboxylic acid and carbonyl groups present at the edges of GO. Finally, the absorption peaks at 1223 and 1051 cm⁻¹ correspond to the stretching vibration of C–O in carboxylic acids and the C–OH bending of alcohol groups, respectively. The presence of these oxygen-containing groups reveals that the starting material (graphite) has been oxidized successfully.

Fig. 6 FTIR spectra of GO and FGO.

The FTIR spectrum of FGO represents a stretching vibration of the N–H group at 2945 and 3474 cm⁻¹, depending on whether the N–H stretching is strong or broad, confirming the presence of hydrogen bonding (Fig. 6). The stretching vibration of the C=N group of the imidazole ring is present at 1535 cm⁻¹. The peaks at 1433 and 1488 cm⁻¹ represent the stretching vibrations of C=C and N=N groups, respectively, whereas the peaks at 1252 and 1325 cm⁻¹ belong to the stretching vibration of C–N=N–C and C=N–N=C. The strong band at 1065 cm⁻¹ and medium band at 1094 cm⁻¹ are assigned to the C–N stretching vibrations. The peaks at 853 and 937 cm⁻¹ correspond to the vibration of C–N=N–C and C–N of imidazole, whereas the peaks at 612, 664, and 752 cm⁻¹ prove the complex formation between the cobalt and the nitrogen atoms of imidazole.

3.2. Optical properties

3.2.1. UV-vis spectra and band gap determination. UV-vis spectroscopy is a helpful technique for understanding photocatalyst structure. Fig. 7 displays the UV-vis spectra of bare TiO₂, and the GO/TiO₂ and FGO/TiO₂ composites.

Fig. 7 UV-vis spectra of the TiO₂ thin film, and the TiO₂/GO and TiO₂/FGO composites.

All of the samples expose a sharp transition in the UV region, which could be ascribed as the peculiar band gap absorption of TiO₂, concluding that electrons transit from the valence band to the conduction band (O_2p → Ti_3d). In comparison to TiO₂ (400 nm), a redshift to higher wavelengths for the GO/TiO₂ (421 nm) and FGO/TiO₂ (447 nm) composites can be seen, proving the narrowness of the band gap with the addition of GO and FGO. The approximated band gaps for TiO₂, GO/TiO₂ and FGO/TiO₂ were 3.10, 2.96 and 2.77 eV, respectively. This supports the detection of a redshift in the absorption of the composites, as compared to the uncombined TiO₂. Bare TiO₂ exhibits no absorption above its absorption edge (400 nm), however introducing GO and FGO results in a continuous absorption band at 421–447 nm, which is in agreement with the gray-black color of the sample. The optimized absorption intensity of light for the composites can lead to an enhancement in photocatalytic activity.^99,100 In the case of FGO, the reduction of the band gap energy is referred to as the GO carboxylic groups substitution with imidazole bonded to the cobalt.

3.3. Photocatalytic activity for the removal of NO_x and CO from air

3.3.1. Photocatalyst pre-treatment. Prior to the degradation experiments, the photocatalyst was illuminated by UV irradiation in the presence of a purified air stream (5 mL min⁻¹) in order to convert any adsorbed carbonaceous impurities, known as the carbon burn-off period.

3.3.2. Photocatalytic activity measurements. Setup in batch-mode adsorption experiments were performed to check the photocatalytic degradation of NO_x and CO. The experiments were tested on three different photocatalysts, such as the bare TiO₂ thin film, and the TiO₂/GO and TiO₂/FGO composites.

Fig. 8a and b display the composite photocatalyst performance for NO_x and CO degradation (C/C₀%) under UV light irradiation. The initial concentration of CO used in this study was 50 ppm. The conversion percentage of CO as a function of residence time using the bare TiO₂ thin film photocatalyst was calculated and is shown in Fig. 8a. Only 10% degradation of CO was found for a residence time of 120 min using the TiO₂ thin film photocatalyst. The CO degradation was 27% for the TiO₂/GO composite and 46% for TiO₂/FGO. It has probably occurred because the adsorbed amount of gas on the TiO₂ thin film was rather low. The NO_x degradation ability of the TiO₂ thin film was 16%, whereas it was 31% for TiO₂/GO and 51% for TiO₂/FGO. Adding GO to the composite improves the NO_x removal, though adding FGO is more effective for the same residence time. It could be seen that the existence of FGO noticeably increased the NO_x removal, because FGO could accelerate the NO adsorption, trapping the VB holes and CB electrons along with complex formation with the pollutant gases.

Fig. 8 (a) Degradation of CO with irradiation time over the samples under UV light irradiation and (b) degradation of NO_x with irradiation time over the samples under UV light irradiation.

In the case of NO_x mixed with CO, the NO_x degradation on the TiO₂ thin film has not been changed, whereas for TiO₂/GO and TiO₂/FGO, it has increased by 5% and 7% consequently in the same residence time. This is probably related to the GO presence. GO on one hand increases CO adsorption and photodegradation, followed by elevated production of hydroxyl radicals. On the other hand, GO helps the electron–hole pair separation. In the case of TiO₂/FGO the impact of CO is more obvious than in the case of TiO₂/GO, because there are more sites in the cobalt structure for complex formation with pollutant gases.^101,102

CO degradation in the presence of NO_x has not increased. It may be due to the small initial concentration of NO_x rather than CO, hence surface saturation does not take place and there is no competition or deactivation effect between the pollutant gases.¹⁰³

Chuensab and Watcharenwong reported the photocatalytic activity of WO₃ films. The removal efficiency of CO was 45% by UV light and 31% by visible light; both cases are lower than our results.¹⁰⁴ Toma et al. reported 35% NO and around 20% NO_x decomposition under the optimum experiment conditions,¹⁰⁵ which are less than our achievements. 23% NO_x and 34% CO removal by a ZnO photocatalyst has been proposed by Kowsari and Bazri.¹⁰⁶ In another study, the maximum conversion efficiency of NO₂ to harmless products was 33%.¹⁰⁷ According to the results presented by Gomez et al., TiO₂ supported on a zeolitic composite material had acceptable degradation results; however, the formation of organophosphorous intermediates lowered the photocatalytic reaction.¹⁰⁸

In the case of bare TiO₂, the band gap energy is 3.10 eV and when photocatalyst is irradiated with appropriate photon energy, the electrons (e_CB⁻) and holes (h_VB⁺) are generated in the conduction and valence bands, and these two high energy parts start oxidizing pollutants. Electrons react with O₂ and accomplish superoxide radicals (O₂˙⁻). Holes combine with water vapor and produce hydroxyl radicals (˙OH). Hydroxyl radicals attack CO and NO, resulting in the oxidation of pollutants (eqn (2)–(8)).

CO is a poisonous gas, and its oxidation to CO₂ is the only feasible way for its destruction. The mechanism of CO photocatalytic oxidation is simple. Previous theoretical calculations and experiments suggested that O₂ adsorbs at the oxygen vacancies on the surface, which could be responsible for the oxidation of CO.¹⁰⁹

When using a TiO₂ thin film, we should pay attention to the crystal structure by preventing electron–hole recombination and adsorbing pollutants at the photocatalyst surface.

The high performance of the selected catalyst can be explained by combining TiO₂ with GO, as excited electrons of TiO₂ could move from the conduction band (CB) to graphene.¹¹⁰ The barrier formed at the network, termed the Schottky barrier, separates the electron–hole pairs, thus overcoming charge recombination. The close contact between TiO₂ and GO favors the transfer of photogenerated electrons from TiO₂ to GO, leading to an adequate charge anti-recombination and raising the photocatalytic performance (eqn (9)–(11)).

TiO₂ + hν → TiO₂ (e_CB⁻ + h_VB⁺) (2)

h_VB⁺ + OH_ads⁻ → OH˙ (3)

O₂ + e_CB⁻ → O₂˙ (4)

2OH˙ + 2CO + 1/2O₂ → 2CO₂ + H₂O (5)

2OH˙ + NO → H₂O + NO₂ (6)

3NO₂ + 2OH˙ → H₂O + NO + 2NO₃⁻ (7)

O₂˙ + NO_x → NO₃⁻ (8)

Graphene/TiO₂ + hν → graphene (e_CB⁻)/TiO₂ (h_VB⁺) (9)

Co–Im + hν → Co–Im* (excited state) (10)

Co–Im* → Co–Im⁺ + graphene (e_CB⁻) (11)

In the FGO photocatalyst composite, the existence of Co–Im attachment on the GO surface is crucial for preventing electron–hole recombination, considering that Co–Im transfers electrons to the GO conducting band, and thus charge separation is achieved.

Zhang et al. first demonstrated the appreciable photocatalytic activity of graphene-based semiconductor composites,¹¹¹ followed by Ng et al.,¹¹² Shen et al.,¹¹³ and Zhou et al.,¹¹⁴ who performed a one-step hydrothermal method to supply graphene–TiO₂ hybrid materials. They presented that the composite had enhanced photocatalytic activity towards organic degradation compared to bare TiO₂. Fan et al. synthesized P25-graphene composites, which illustrated the significant photocatalytic performance compared to pure P25.¹¹⁵

To the best of our knowledge, this is the first study on the use of graphene functionalization with an imidazole–cobalt complex-GO/TiO₂ composite on the photocatalytic degradation of CO and NO_x.

The TiO₂/FGO nanocomposite exhibits a new photocatalytic mechanism, in which cobalt acts as a complexing metal and pollutant gases act as coordinating agents, according to the reaction equations (eqn (12)–(16)).

This photocatalytic mechanism is specifically unusual compared to previous research on GO–semiconductor photocatalysts, where GO is declared to behave as an electron scavenger to catch and then transport the photogenerated electrons from the semiconductor (see ESI Fig. S5†).

The Hard Soft Acid Base (HSAB) theory confirms that soft Lewis bases bind to soft Lewis acids preferably and vice versa. Considering Co²⁺ as a borderline Lewis acid and carbon monoxide (CO) as a soft Lewis base, the reaction takes place. CO (Lewis base) donates a pair of electrons to Co (Lewis acid) in order to form a coordination complex.¹¹⁶

Nitric oxide is different from CO in its mechanism, because of the existence of one odd electron in an anti-bonding orbital. NO can miss one electron and become isoelectronic (NO⁺) like CO. Hence, cobalt complex formation with nitrosyl takes place by transmitting electrons to cobalt and forming a covalent bond, respectively. Besides, NO is a strong π-acceptor and can overlap with filled d-orbitals in Co with its empty anti-bonding orbitals.¹¹⁷

[Co(Im)₄]Cl₂ + 2NO → [Co(Im)₄(NO)₂]Cl₂ (12)

[Co(Im)₄]Cl₂ + 2CO → [Co(Im)₄(CO)₂]Cl₂ (13)

[Co(Im)₄(NO)₂]Cl₂ + O₂ → [Co(Im)₄(NO₂)₂]Cl₂ (NO → NO₂) (14)

[Co(Im)₄(NO₂)₂]Cl₂ + O₂ → [Co(Im)₄(NO₃)₂] (NO₂ → NO₃) (15)

[Co(Im)₄(CO)₂]Cl₂ + O₂ → [Co(Im)₄(CO₂)₂]Cl₂ (CO → CO₂) (16)

XPS analysis of Co after the photocatalytic reaction confirms that Co addition can greatly enhance the chemisorption of NO_x and CO species; at the same time, graphene can cooperate with holes, by inhibiting the recombination of photogenerated electrons and holes which leads to a photocatalytic reaction improvement.

In the case of FGO, the XPS data of the quantitative surface atomic percentage analyses of Co2p after photocatalytic reactions of CO show concentrations of 25% Co–C–O and 8% Co–C–(O)2 which reveals the oxidation of CO at the cobalt surface. After the photocatalytic oxidation of NO_x, the XPS results represent Co–N–O as 27%, Co–N–(O)2 as 7% and Co–N–(O)3 as 3% owing to the Co–NO_x species forming (Table 1).

Table 1 XPS elementary atomic surface ratios of the FGO catalyst before and after the photocatalytic reactions

Sample	Atomic ratios (%)
	Co–N=C	Co–N–C	Co–Cl	Co–C–O	Co–C–(O)2	Co–N–O	Co–N–(O)2	Co–N–(O)3
FGO before photocatalytic reactions	45	39	26	—	—	—	—	—
FGO after photocatalytic reactions of CO	30	35	10	25	8	—	—	—
FGO after photocatalytic reactions of NOx	28	33	2	—	—	27	7	3

---

Two photocatalytic reactions are considered for air pollutant removal. The first one concerns the photocatalytic activity of TiO₂ for the removal of NO_x and CO, whereas the other is the cobalt forming a complex with CO and NO followed by the oxidation of NO to NO₂ and NO₃ as well as CO to CO₂. Both are applicable to treat dilute concentrations of CO and NO_x.

The roles of GO in this case are, first, to adsorb NO_x and CO, second, to store the species in its surface and third, to separate the electrons and holes. Adding FGO to bare TiO₂ significantly increases the NO_x and CO elimination. The NO_x and CO removal per unit of FGO/TiO₂ composite is approximately three times greater than that of the pure TiO₂ thin film.

4. Conclusions

In summary, a GO-based TiO₂ composite photocatalyst was developed. The anatase TiO₂ thin film was deposited on quartz substrates using the RF magnetron sputtering method. FGO was composed of cobalt–imidazole binding on the surface carboxylic groups of GO. The activity of the FGO/TiO₂ composite was tested by the photocatalytic degradation of NO_x and CO under UV irradiation. The TiO₂/FGO composite displayed a high performing photocatalytic activity of 51% for NO_x and 46% for CO, which is higher than that achieved by GO/TiO₂ and the TiO₂ thin film. Incorporating FGO into the composite led to a decrease of the band gap and increased the sensitivity to visible light irradiation (λ > 400 nm), with superior functional properties. In addition, the photoinduced electrons can easily travel around FGO, continuing with efficient charge separation and prolonging the charge recombination time. These improvements, along with the increase of the reactant adsorbing at the surface, are the key factors for augmenting the photocatalytic operation. We hope our findings could be directed to an expandable and cost-effective approach for obtaining powerful photocatalytic materials.

Acknowledgements

The authors wish to gratefully thank the Research Affairs Division at Amir Kabir University of Technology (AUT), Tehran, for financial support (Grant no. 4/360).

References

1. S. A. Edgerton, M. W. Holdren and D. L. Smith, J. Air Pollut. Control Assoc., 1989, 39, 729.
2. C. W. Sweet and S. J. Vermette, Environ. Sci. Technol., 1992, 26, 165.
3. R. Kostiainen, Atmos. Environ., 1995, 29, 693.
4. R. Mukund, T. J. Kelly and W. Chester, Atmos. Environ., 1996, 30, 3457.
5. R. M. Alberici and W. E. Jardim, Appl. Catal., B, 1997, 14, 55.
6. National Ambient Air Quality Standards (NAAQS), http://epa.gov/air/criteria.html, (accessed May 2015).
7. B. J. Finlayson-Pitts and J. N. Pitts Jr, Science., 1997, 267, 1045.
8. W. Cha, S. Yun and J. Jurng, Phys. Chem. Chem. Phys., 2014, 16, 17900.
9. F. Guo, J. Yu, M. Chu and G. Xu, Catal. Sci. Technol., 2014, 4, 2147.
10. F. Liu, Y. Yu and H. He, Chem. Commun., 2014, 50, 8445.
11. S. Das and W. M. A. W. Daud, RSC Adv., 2014, 4, 20856.
12. A. Fujishima and K. Honda, Nature, 1972, 238, 37.
13. A. Fujishima, K. Hashimoto and T. Watanabe, TiO₂ photocatalysis: fundamentals and applications, BKC Inc., Tokyo, 1999, vol. 58, pp. 14–176.
14. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253.
15. A. Mills and S. Le Hunte, J. Photochem. Photobiol., A, 1997, 108, 1.
16. M. R. Hoffmann, S. T. Martin, W. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69.
17. A. Ajmal, I. Majeed, R. N. Malik, H. Idrissc and M. A. Nadeem, RSC Adv., 2014, 4, 37003.
18. U. L. Štangar, M. Kete, A. Šuligoj and M. Tasbihi, Mater. Sci. Eng., 2012, 30, 012001.
19. I. Arabatzis, S. Antonaraki, T. Stergiopoulos, A. Hiskia, E. Papaconstantinou and M. Bernard, J. Photochem. Photobiol., A, 2002, 149, 237.
20. K. Weng and Y. Huang, Surf. Coat. Technol., 2013, 231, 201.
21. D. Hwang, J. Moon, Y. Shul, K. Jung, D. Kim and D. Lee, J. Sol-Gel Sci. Technol., 2003, 26, 783.
22. R. Fateh, A. Ismail, R. Dillert and D. Bahnemann, J. Phys. Chem., 2011, 115, 10405.
23. R. Fateh, R. Dillert and D. Bahnemann, ACS Appl. Mater. Interfaces, 2014, 6, 2270.
24. S. Okunaka, H. Tokudome, Y. Hitomib and R. Abe, J. Mater. Chem. A., 2015, 3, 1688.
25. E. V. Skorb, D. G. Shchukin, H. Möhwald and D. V. Sviridov, J. Mater. Chem., 2009, 19, 4931.
26. Z. Zhao, H. Tan, H. Zhao, D. Li, M. Zheng, P. Du, G. Zhang, D. Qu, Z. Sun and H. Fan, Chem. Commun., 2013, 49, 8958.
27. D. Lee, H. Kim, S. Yu, H. J. Kim, W. I. Lee and D. Jang, J. Mater. Sci., 2014, 49, 3414.
28. J. Altmayer, S. Barth and S. Mathur, RSC Adv., 2013, 3, 11234.
29. R. Bhakta, R. Thomas, F. Hipler, H. F. Bettinger, J. Müller, P. Ehrhart and A. Devi, J. Mater. Chem., 2004, 14, 3231.
30. T. Nakajima, T. Nakamura, K. Shinoda and T. Tsuchiya, J. Mater. Chem. A., 2014, 2, 6762.
31. R. Sellappan, J. Sun, A. Galeckas, N. Lindvall, A. Yurgens, A. Y. Kuznetsov and D. Chakarov, Phys. Chem. Chem. Phys., 2013, 15, 15528.
32. J. B. Bellam, M. A. Ruiz-Preciado, M. Edely, J. Szade, A. Jouanneaux and A. H. Kassiba, RSC Adv., 2015, 5, 10551.
33. S. Ameen, M. S. Akhtar, H. Seo and H. Shin, CrystEngComm, 2014, 16, 3020.
34. T. Sun, Y. Wang, M. Al-Mamun, H. Zhang, P. Liu and H. Zhao, RSC Adv., 2015, 5, 12860.
35. H. Kim, H. Kim, W. I. Lee and D. Jang, J. Mater. Chem. A, 2015, 3, 9714.
36. Q. Zeng, H. Li, H. Duan, Y. Guo, X. Liu, Y. Zhang and H. Liu, RSC Adv., 2015, 5, 13430.
37. J. Yang, H. Piao, A. Vinu, A. A. Elzatahry, S. Paek and J. Choy, RSC Adv., 2015, 5, 8210.
38. W. Donphai, T. Kamegawa, M. Chareonpanich, K. Nueangnoraj, H. Nishihara, T. Kyotani and H. Yamashita, Phys. Chem. Chem. Phys., 2014, 16, 25004.
39. S. Zheng, L. Gao, Q. H. Zhang and J. K. Guo, J. Mater. Chem., 2000, 10, 723.
40. X. D. Gao, X. M. Li, X. Y. Gan, Y. Q. Wu, R. K. Zheng, C. L. Wang, Z. Y. Gu and P. He, J. Mater. Chem., 2012, 22, 18930.
41. H. He, T. Riedl, A. Lerf and J. Klinowski, J. Phys. Chem., 1996, 100, 19954.
42. G. Eda, Y. Y. Lin, C. Mattevi, H. Yamaguchi, H. A. Chen, I.-S. Chen, C. W. Chen and M. Chhowalla, Adv. Mater., 2010, 22, 505.
43. M. Li, S. K. Cushing, X. Zhou, S. Guo and N. Q. Wu, J. Mater. Chem., 2012, 22, 23374.
44. C. Wang, J. Zhou and L. Chu, RSC Adv., 2015, 5, 52466.
45. M. Barathi, A. S. K. Kumar, C. U. Kumar and N. Rajesh, RSC Adv., 2014, 4, 53711.
46. Y. Li, P. Zhang, Q. Du, X. Peng, T. Liu, Z. Wang, Y. Xia, W. Zhang, K. Wang, H. Zhu and D. Wu, J. Colloid Interface Sci., 2011, 363, 348.
47. G. Zhao, L. Jiang, Y. He, J. Li, H. Dong, X. Wang and W. Hu, Adv. Mater., 2011, 23, 3959.
48. G. Zhao, J. Li and X. Wang, Chem. Eng. J., 2011, 173, 185.
49. J. Wang, X. Gao, Y. Wang and C. Gao, RSC Adv., 2014, 4, 57476.
50. B. Yang, Y. Guo, S. Zhang, T. Wen and C. Zhao, RSC Adv., 2014, 4, 64771.
51. Q. Fang and B. Chen, J. Mater. Chem. A., 2014, 2, 8941–8951.
52. L. Yu, Y. Ma, C. N. Ong, J. Xie and Y. Liu, RSC Adv., 2015, 5, 64983.
53. V. Chandra and K. S. Kim, Chem. Commun., 2011, 47, 3942.
54. R. Sitko, E. Turek, B. Zawisza, E. Malicka, E. Talik, J. Heimann, A. Gagor, B. Feist and R. Wrzalik, Dalton Trans., 2013, 42, 5682.
55. H. Y. He, J. Klinowski, M. Forster and A. Lerf, Chem. Phys. Lett., 1998, 287, 53.
56. A. Lerf, H. Y. He, M. Forster and J. Klinowski, J. Phys. Chem. B, 1998, 102, 4477.
57. H. C. Schniepp, J. L. Li, M. J. McAllister, H. Sai, M. Herrera Alonso, D. H. Adamson, R. K. Prud’homme, R. Car, D. A. Saville and I. A. Aksay, J. Phys. Chem. B, 2006, 110, 8535.
58. S. Park, K. S. Lee, G. Bozoklu, W. Cai, S. T. Nguyen and R. S. Ruoff, ACS Nano, 2008, 2, 572.
59. Y. Xu, Z. Liu, X. Zhang, Y. Wang, J. Tian, Y. Huang, Y. Ma, X. Zhang and Y. A. Chen, J. Adv. Mater., 2009, 2, 1275.
60. Z. B. Liu, Y. F. Xu, X. Y. Zhang, X. L. Zhang, Y. S. Chen and J. G. Tian, J. Phys. Chem. B, 2009, 113, 9681.
61. X. Zhang, Y. Huang, Y. Wang, Y. Ma, Z. Liu and Y. Chen, Carbon, 2009, 47, 334.
62. S. Wang, P. J. Chia, L. L. Chua, L. H. Zhao, R. Q. Png, S. Sivaramakrishnan, M. Zhou, R. G. S. Goh, R. H. Friend, A. T. S. Wee and P. K. H. Ho, J. Adv. Mater., 2008, 20, 3440.
63. H. Yang, C. Shan, F. Li, D. Han, Q. Zhang and L. Niu, Chem. Commun., 2009, 26, 3880.
64. K. H. Cheon, J. Cho, Y. H. Kim and D. S. Chung, ACS Appl. Mater. Interfaces, 2015, 7, 14004.
65. Y. Yamadaa, Y. Suzukia, H. Yasudaa, S. Uchizawaa, K. Hirose-Takai, Y. Sato, K. Suenagab and S. Satoa, Carbon, 2014, 75, 81.
66. C. Xu, X. Wang, J. Zhu, X. Yang and L. Lu, J. Mater. Chem., 2008, 18, 5625.
67. J. Wang, K. Feng, H. H. Zhang, B. Chen, Z. J. Li, Q. Y. Meng, L. P. Zhang, C. H. Tung and L. Z. Wu, Beilstein J. Nanotechnol., 2014, 5, 1167.
68. L. Wang, D. Wang, X. Hu, J. Zhu and X. Liang, Electrochim. Acta, 2012, 76, 282.
69. S. Losse, J. G. Vos and S. Rau, Coord. Chem. Rev., 2010, 254, 2492.
70. J. Tian, H. Li, A. Asiri, A. Al-Youbi and X. Sun, Small, 2013, 9, 2709.
71. X. Dong, H. Xu, X. Wang, Y. Huang, M. Chan-Park, H. Zhang, L. Wang, W. Huang and P. Chen, ACS Nano, 2012, 6, 3206.
72. H. Yu and Z. Tan, Environ. Sci. Technol., 2014, 48, 2453.
73. W. Cruz, P. Leung and K. Seff, Inorg. Chem., 1979, 18, 1692.
74. Y. Lee, Y. Kim and K. Seff, J. Phys. Chem. B, 2005, 109, 4900.
75. O. Alnaji, M. Dartiguenave, Y. Dartiguenave, M. Simard and A. Beauchamp, Inorg. Chim. Acta, 1991, 187, 31.
76. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
77. J. T. Robinson, M. Zalalutdinov, J. W. Baldwin, E. S. Snow, Z. Wei, P. Sheehan and B. H. Houston, Nano Lett., 2008, 8, 3441.
78. C. Suryanarayana and M. G. Norton, X-Ray Diffraction: A Practical Approach, Plenum Press, New York, 1998.
79. K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603.
80. F. Rouquerol, J. Rouquerol and K. Sing, Adsorption by Powders & Porous Solids, Academic Press, London, 1999.
81. J. Tauc, Optical Properties of Solids, North-Holland, Amsterdam, 1972, pp. 278–313.
82. C. H. Huang, I. K. Wang, Y. M. Lin, Y. H. Tseng and C. M. Lu, J. Mol. Catal. A: Chem., 2010, 316, 163.
83. ISO 22197-1: 2007, ‘Fine ceramics, advanced technical ceramics) – Test method for air-purification performance of semiconducting photocatalytic materials – part 1: Removal of nitric oxide, ISO, Geneva, 2007.
84. ISO 10677: 2011, Fine ceramics, advanced technical ceramics –Ultraviolet light source for testing semiconducting photocatalytic materials, ISO, Geneva, 2011.
85. J. N. Hart, D. Menzies, Y. B. Cheng, G. P. Simon and L. Spiccia, Sol. Energy Mater. Sol. Cells, 2007, 91, 6.
86. H. Kim and D. Jang, CrystEngComm, 2015, 17, 3325.
87. X. Li, Y. Hu, J. Liu, A. Lushington, R. Li and X. Sun, Nanoscale., 2013, 5, 12607.
88. X. Tong, H. Wang, G. Wang, L. Wan, Z. Ren and J. Bai, J. Solid State Chem., 2011, 184, 982.
89. A. Buchsteiner, A. Lerf and J. Pieper, J. Phys. Chem. B, 2006, 110, 22328.
90. Q. Du, M. Zheng, L. Zhang, Y. Wang, J. Chen, L. Xue, W. Dai, G. Ji and J. Cao, Electrochim. Acta, 2010, 55, 3897.
91. Z. Xu, Y. Huang, C. Min and L. Chen, Radiat. Phys. Chem., 2010, 79, 839.
92. H. Yu, P. Xu, D. W. Lee and X. Li, J. Mater. Chem. A., 2013, 1, 4444.
93. K. S. Kim, I. J. Kim and S. J. Park, Synth. Met., 2010, 160, 2355.
94. M. Chhowalla and G. Eda, Adv. Mater., 2010, 22, 2392.
95. Y. Zhu, S. Murali, W. Cai, X. Li, J. W. Suk, J. R. Potts and R. S. Ruoff, Adv Mater., 2010, 22, 3906.
96. C. Zhu, S. Guo, Y. Fang and S. Dong, ACS Nano, 2010, 4, 2429.
97. Q. Yu, L. A. Jauregui, W. Wu, R. Colby, J. Tian, Z. Su, H. Cao, Z. Liu, D. Pandey, D. Wei, T. F. Chung, P. Peng, N. P. Guisinger, E. A. Stach, J. Bao, S. Pei and Y. P. Chen, Nat. Mater., 2011, 10, 443.
98. S. Pisana, M. Lazzeri, C. Casiraghi, K. S. Novoselov, A. K. Geim, A. C. Ferrari and F. Mauri, Nat. Mater., 2007, 6, 198.
99. H. Yamashita, M. Harada, J. Misaka, M. Takeuchi, K. Ikeue and M. Anpo, J. Photochem. Photobiol., A, 2002, 148, 257.
100. S. G. Babu, R. Vinoth, D. P. Kumar, M. V. Shankar, H. Chou, K. Vinodgopal and B. Neppolian, Nanoscale., 2015, 7, 7849.
101. O. Carp, C. L. Huisman and A. Reller, Solid State Chem., 2004, 32, 33.
102. M. L. Sauer and D. F. Ollis, J. Catal., 1996, 163, 215.
103. C. H. Ao, S. C. Lee, C. L. Mak and L. Y. Chan, Appl. Catal., B, 2003, 42, 119.
104. A. Chuensab and A. Watcharenwong, International Conference on Chemical, Environmental and Biological Sciences (ICCEBS), Malaysia, 2012.
105. F. L. Toma, G. Bertrand, D. Klein and C. Coddet, Environ. Chem. Lett., 2004, 2, 117.
106. E. Kowsari and B. Bazri, Appl. Catal., A, 2014, 475, 325.
107. A. Gandolfo, V. Bartolomei, E. Gomez Alvarez, S. Tlili, S. Gligorovskia, J. Kleffmann and H. Wortham, Appl. Catal., B, 2015, 166, 84.
108. S. Gomez, C. L. Marchena, M. S. Renzini, L. Pizzio and L. Pierella, Appl. Catal., B, 2015, 162, 167.
109. A. V. Vorontsov, E. N. Savino, E. N. Kurkin, O. D. Torbova and V. N. Parmon, Catal. Lett., 1997, 62, 83–88.
110. G. Williams, B. Seger and P. V. Kamat, ACS Nano, 2008, 2, 1487–1491.
111. H. Zhang, X. Lv, Y. Li, Y. Wang and J. Li, ACS Nano, 2009, 4, 380.
112. Y. H. Ng, I. V. Lightcap, K. Goodwin, M. Matsumura and P. V. Kamat, J. Phys. Chem. Lett., 2010, 15, 2222.
113. J. Shen, B. Yan, M. Shi, H. Ma, N. Li and M. Ye, J. Mater. Chem., 2011, 21, 3415.
114. K. Zhou, Y. Zhu, X. Yang, X. Jiang and C. Li, New J. Chem., 2011, 35, 353.
115. W. Fan, Q. Lai, Q. Zhang and Y. Wang, J. Phys. Chem. C, 2011, 115, 10694.
116. Organomettalic Hyper Text Book, http://www.ilpi.com/organomet/carbonyl.html, (accessed June 2015).
117. P. T. Manoharan and H. B. Gray, Inorg. Chem., 1966,(5), 823.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra13620b

---