Cu@CuO promoted g-C₃N₄/MCM-41: an efficient photocatalyst with tunable valence transition for visible light induced hydrogen generation

Abstract

A series of ternary Cu@CuO–g-C₃N₄/MCM-41 photocatalysts have been synthesized by varying the percentage of Cu using simple impregnation and co-condensation methods. The physico-chemical characterization of all the samples was determined using XRD, FTIR, UV-Vis DRS, PL, N₂ ads–des studies, SEM and XPS HRTEM, EDAX, EIS and MS. The structural advantages of MCM-41, allow the uniform distribution of g-C₃N₄ and coexistence of Cu²⁺ along with Cu⁰ without using a reducing agent. The presence of g-C₃N₄ helps to shift the Fermi level of CuO towards more negative values due to accumulation of photogenerated electrons on the surface. It favours charge separation by creating a Schottky barrier at the junction. The 4 wt% Cu loaded over g-C₃N₄/MCM-41 exhibits a maximum 750 μmol 2 h⁻¹ of H₂ evolution under visible light irradiation with an apparent energy conversion efficiency of 24.8%. The enhancement in catalytic activity has been explained on the basis of synergism between g-C₃N₄ and Cu²⁺ and the SPR effect of Cu which also acts as a co-catalyst present on the surface of photocatalysts.

---

1. Introduction

Energy and environmental issues are among the biggest technological challenges being faced by chemists and technologists in the 21^st century. To satisfy the rising energy demands with minimum environmental impact, research work by different groups of scientists to develop a sustainable and cost-effective method for harvesting solar energy is on-going. Hydrogen produced using solar energy will play an important role in the system because it is an ultimate clean energy and it can be used in fuel cells. So the exploration of high-efficiency visible light induced photocatalysts has attracted more and more attention.

Recently it has been found that as a pi-conjugated material graphitic carbon nitride exhibits improved performance towards energy conversion owing to its ring structure, high chemical and thermal stability due to high condensation, faster charge transport and its suitable band gap (2.7 eV).¹ Still the photocatalytic activity is unsatisfactory because of (1) limited utilisation of broad spectrum of solar radiation; (2) restricted mobility of charge carriers due to the absence of inter layer hybridisation of the electronic states; (3) grain boundary effect and most importantly (4) its low surface area. To overcome these drawbacks various strategies like morphological variation, doping, construction of hetero junction, designing the material with an appropriate textural porosity to enhance its surface area have been emphasised.²

In recent years various group of scientists extensively studied the photocatalytic activity of g-C₃N₄ based composites especially incorporating Cu nanoparticles. For example, Zhou et al. showed improved H₂ evolution by loading Cu(OH)₂ nanoparticles as cocatalyst on the surface of g-C₃N₄.³ Tian et al. explained enhanced activity by constructing a p–n heterojunction of Cu₂O and g-C₃N₄.⁴ Peng et al. also reported degradation of acid orange-II by Cu₂O/g-C₃N₄ composite.⁵ Fan et al. showed efficient hydrogen evolution by dispersing Cu nanoparticles over g-C₃N₄.⁶ Liu et al. explained enhanced photostability and H₂ evolution by core@shell structured Cu₂O@g-C₃N₄ octahedra photocatalyst.⁷ Wang et al. reported the synthesis of porous nanorods of g-C₃N₄/CuO composite by using Cu(Ac)₂ solution in presence of a reductant to study its photocatalytic activity towards degradation of RhB.⁸ Li et al. also constructed a p–n heterojunction between CuO and g-C₃N₄ in presence of reductant to facilitate the photocatalytic H₂ production process.⁹ Shi et al. fabricated a g-C₃N₄/CuO_X heterostructure in presence of ethanolamine as reductant to adjust the valence state of CuO_X.¹⁰

The discovery of modified mesoporous materials are mainly known suitable for their surface active properties. In addition to high surface area, they possess semiconducting properties (modified by transition metals/metal oxides) which are active in UV and visible light.^11–14 The transition metal and metal oxide modified MCM-41 possess enough surface reactive sites and nano scale channels facilitating transfer of photo generated charges to the surface. Xing et al. studied the application of Cu/BiVO₄–MCM-41 catalyst for the photo-catalytic degradation of methylene blue.¹⁵ Pradhan et al. in our group reported the visible light induced degradation of phenolic compounds by using Cu/Al₂O₃–MCM-41 and Fe@MnO₂–MCM-41 and efficient hydrogen evolution by Fe/Al₂O₃–MCM-41.^16–18 Wang et al. developed (Fe–Fe) hydrogenase mimics over MCM-41 for photochemical hydrogen production having exceptional activity.¹⁹ Shen et al. reported visible light induced photocatalytic H₂ evolution of Ti–MCM-41/g-C₃N₄ composite.²⁰

Considering the structural advantage of MCM-41, here in the present work we have tried to prepare a g-C₃N₄/MCM-41 composite to improve the dispersion of g-C₃N₄ on the large surface of MCM-41 and to extend the absorption of the composite in the visible region. Further Cu nanoparticles are deposited on the semiconductor photocatalyst in two different oxidation states without using a reductant to increase H₂ production activity. Because Cu is well known as earth-abundant, low cost transition metal which promotes charge separation by creating a Schottky barrier at the junction and able to catalyze reduction of proton to H₂ molecules. Owing to the presence of silanol groups in MCM-41 and g-C₃N₄ having lone pairs of electrons, in the process of Cu loading Cu²⁺ ions are partially converted into Cu⁺ and Cu⁰. Presence of CuO (E_g = 1.5 eV) along with g-C₃N₄ (E_g = 2.7 eV) enhances visible light absorption. The improvement in intrinsic photo catalytic activity of Cu cocatalyst at nanolevel also increases the number of active sites due to stronger interfacial interaction and enlarged surface area. The great advantage of loading Cu is to use it for large scale practical applications because the price is much lower than that of other novel metals as cocatalyst. The improved photocatalytic H₂ production can be attributed to more number of surface active sites of Cu cocatalyst at the nanoscale and superior interfacial interactions, which results efficient electron transfer due to surface plasmonic resonance from the semiconductor to the cocatalyst to catalyze water reduction to produce H₂.

2. Experimental

2.1. Synthesis of g-C₃N₄ (CN)

Pure CN was synthesized by the well reported pyrolysis method.²¹ Typically, a certain amount of melamine was put into an alumina crucible with a half cover, and calcined at 550 °C in air for 4 h with a ramping rate of 5 °C min⁻¹.

2.2. Synthesis of g-C₃N₄/MCM-41 (CM)

Graphitic carbon nitride modified MCM-41 composites were synthesized by in situ incorporation method. During the standard synthesis method²² of MCM-41, CN was added just before addition of NH₄OH. By varying the amount of CN (1, 2, 4, 6 and 8 wt%) five catalysts of CN/MCM-41 (CM) were synthesized. The composite materials were annealed at 550 °C for 5 h. The formulated mesoporous materials are designated as CM-1, CM-2, CM-4, CM-6 and CM-8 depending the amount of CN from 1 to 8 wt%.

2.3. Synthesis of Cu@g-C₃N₄ (CC)

Different wt% of Cu modified gC₃N₄ can be prepared by impregnation method. 1 g of neat CN was dispersed in 50 mL of distilled water. Required amount of CuSO₄·5H₂O (0.3–0.1.2 g) solution was added to the above solution and then stirred. The liquid phase was removed by a 4 h treatment at 50 °C in a rotary evaporator. The prepared samples are termed as CC-x (x varies from 2 to 8 wt%).

2.4. Synthesis of Cu@CuO–g-C₃N₄/MCM-41 (CCM)

Modification of the surface textural properties for utilizing the materials as photocatalysts, different wt% of copper such as 2, 4, 6 and 8 was introduced onto the surface of CM-2 by wetness impregnation method. CuSO₄·5H₂O is taken as the source of copper. The copper-containing catalysts (2–8 wt% copper loading) were prepared by impregnation of 1 g of CM-2 with an aqueous solution of CuSO₄·5H₂O. The required amount of CuSO₄ (0.07–0.31 g) solution was added to the parent CM-2 and then stirred. The liquid phase was removed by a 4 h treatment at 50 °C in a rotary evaporator. After drying, the samples were calcined in airflow at 500 °C for 6 h. The prepared samples are termed as CCM-x (x varies from 2 to 8 wt%).

2.5. Formation mechanism

The ternary nano composite CCM was synthesised by a facile in situ incorporation method. By incorporating CN on the surface of MCM-41 was helpful to deposit Cu nanoparticles in two different oxidation states (Cu²⁺, Cu⁰) without using a reductant. During synthesis the presence of silanol group of MCM-41 and the available lone-pairs of electrons on the N-atom of CN were capable of reducing Cu²⁺ → Cu⁺ → Cu⁰. When Cu loading is less Cu nanoparticles are well dispersed and get reduced over CN forming a Cu–O–Si bond indicating strong interaction between Cu species and SiO₂. Owing to which complete reduction of Cu²⁺ to Cu⁰ was restricted which is also in accordance with FTIR data. In case of higher wt% of Cu, it may encapsulate in the pores of MCM-41 and form large CuO particles (Scheme 1).

2.6. Characterization techniques

Powder X-ray diffraction (PXRD) patterns of the samples were taken in the 2θ range of 1–10° at a scanning rate of 2° min⁻¹ in steps of 0.01° (Rigaku Miniflex set at 30 kV and 15 mA) using Cu Kα radiation. The wide angle X-ray diffraction (XRD) patterns of powdered samples were taken in the 2θ range of 10–70° at a rate of 1.2° min⁻¹ (Philips analytical 3710) using CuKα radiation. The FTIR spectra of the samples were recorded using Varian 800-FTIR in KBr matrix in the range of 4000–400 cm⁻¹. The Brunauer–Emmett–Teller (BET) surface area, average pore diameter, mesopore distribution, total pore volume, and micropore volume were determined by the multipoint N₂ adsorption–desorption method at liquid N₂ temperature (−196 °C) by an ASAP 2020 (Micromeritics). Prior to analyses, all the samples were degassed at 300 °C and 10⁻⁶ Torr pressure for 5 h to evacuate the physically adsorbed moisture. The mesopore structure was characterized by the distribution function of mesopore volume calculated by applying the Barrett–Joyner–Halenda (BJH) method. The co-ordination environment of the samples were examined by UV-Vis diffuse reflectance spectroscopy. The spectra were recorded in JASCO V-750 UV-visible spectrophotometer in the wavelength range of 200–800 nm. The PL spectra are recorded by a JASCO FP-8300 spectro fluorometer. The scanning electron microscopic figures were recorded using Hitachi S3400N. The X-ray photoelectron spectra of Cu, C, N, Si and O were recorded using KRATOS with Mg, Al, and CuKα as X-ray sources. The photoelectrochemical measurement was performed using a potentiostat/galvanostat under illumination conditions (λ ≥ 420 nm). The current voltage was measured using a conventional Pyrex electrochemical cell consisting of the prepared electrode as the working electrode, and a platinum wire and a Ag/AgCl electrode were used as the counter and reference electrodes, respectively. The potential of the working electrode was controlled by a potentiostat. The cell was filled with an aqueous solution of 0.1 M Na₂SO₄ and the pH of the solution was adjusted to 6.

2.7. H₂ production experiment

The photocatalytic activities of all the pure and composite materials were examined for H₂ gas evolution under visible light irradiation (>400 nm). For H₂ evolution study, alternatively hole suppression can be achieved by using a sacrificial electron donor CH₃OH. The photochemical reaction is carried out by taking 20 mL of 10% methanol solution and 0.02 g of powder photocatalyst in the quartz batch reactor. For the photocatalytic reaction, 125 W medium pressure mercury visible lamp is used as the light source and 1 M NaNO₂ as UV filter. The solution (pH = 5) was stirred continuously with the help of a magnetic stirrer preventing the setting of nanoparticles at the bottom of the photochemical reactor. For elimination of dissolved gas the mixture solution was purged by nitrogen gas at a flow rate of 10 mL min⁻¹ for 30 min. The reaction was carried out at an interval of 2 h, throughout the experiment. The evolved H₂ gas was collected by downward displacement technique and it was analysed by GC-17A (Shimadzu) with 5 Å molecular sieves column and a thermal conductivity detector (TCD). The peak which appeared on the chromatogram along the standard affirms that the evolved gas was hydrogen. The photocatalytic tests for each sample were repeated at least three times. The approximate average incident radiation power on the photocatalyst and stored chemical energy for CCM-4 photocatalyst under the irradiation of visible light has been calculated and it has been found to be 120 mW and 59.53 mW, respectively. Therefore, using eqn (1) the apparent energy conversion efficiency of the photocatalyst has been calculated.

Conversion efficiency (%) = stored chemical energy/incident light energy × 100 (1)

3. Result discussion

3.1. XRD analysis

The crystal structure, structural identity and phase composition of the synthesised nano-composites were studied by X-ray diffraction as shown in Fig. 1. The small angle XRD pattern shows characteristic diffraction peaks for MCM-41. A strong diffraction peak for 100 plane indicates the mesoporosity and other two less intense reflections were indexed as 110 and 200, which suggests the existence of a periodic hexagonal long range order of the channel. The mesoporosity remains intact after modification of the silica network with Cu and CN. There is a little bit reduction and broadening of the (100) peak of MCM-41 after modification indicating a slight reduction in hexagonal symmetry.¹⁴ The XRD patterns of neat g-C₃N₄ shows diffraction peaks at 2θ = 13.1° and 27.4° representing in planar structural packing motif indexed as 100 plane and inter layer stacking of conjugated aromatic system indexed as 002 plane respectively (Fig. S1†). However the wide angle XRD patterns suggest that there is no significant peak for CN due to uniform distribution of very low wt% of CN on the large surface of MCM-41. The peak at 2θ = 22.6° represents the silica which was slightly shifted to lower value due to the hybridization of CN with MCM-41. Upon Cu loading a series of narrow and sharp peaks are obtained indicating well crystallisation of Cu-nanoparticles. Meanwhile when the Cu loading increases (more than 4 wt%) on CM-2 composite, the high angle peaks appear at 2θ = 44.1° and at 64.6°. These are attributed to the crystalline phase of Cu(0) present as bulk amount in the material. The narrow and sharp diffraction peaks due to CuO at 38.0° and 34.9° indicates well crystallization of CuO particles representing (111) and (002) planes respectively. Presence of Cu–O–Si bond indicates strong interaction between Cu species and SiO₂. Owing to which complete reduction of Cu²⁺ to Cu⁰ was restricted, in accordance to FTIR data.²³ Less intense peaks at 33.2° and 42.1° indicate the presence of Cu₂O phase in case of CCM-4 only. The existence of metallic Cu along with CuO favours charge separation and thus enhance the photocatalytic activity of the heterostructure nanocomposite.

Fig. 1 Wide angle XRD patterns of CCM-2 (c), CCM-4 (d), CCM-6 (e) and CCM-8 (f) (low angle XRD patterns of (a) MCM-41 (b) CM-2 are inserted).

3.2. N₂ adsorption–desorption studies

The N₂ ads–des isotherms of CM-2, CCM-2, CCM-4 and CCM-6 materials are studied and given in Fig. 2a–d. All the figures show a type IV hysteresis confirming the mesoporous nature of the materials. Initially there is slow increase in the adsorption volume and then a sharp rise in the curve. In mesoporous materials there is fluid–fluid interaction in addition to fluid–wall attraction. Hence there is multilayer adsorption of fluid and capillary condensation occurs at the pore walls. Again the interconnectivity of pores facilitate the condensation process. The desorption process of N₂ doesn't follow the same path as that of adsorption hence a loop is observed in the graphs. The tensile strength effect and capillary condensation inside the pores are mainly responsible for the loop formation.

Fig. 2 N₂ sorption isotherm study of (a) CM-2, (b) CCM-2, (c) CCM-4 and (d) CCM-6.

The specific surface area of all the materials are calculated by BET method (Table 1). The parent MCM-41 and CN shows a surface area of 878 m² g⁻¹ and 10 m² g⁻¹ respectively. There is decrease in the surface area value of MCM-41 with loading in Cu and CN over the surface. CM-2 composite shows a surface area of 863 m² g⁻¹ which gradually decreased by increasing Cu loading. Though the surface area of modified materials is less compared to the parent MCM-41, the active sites are uniformly distributed over the support to provide excellent activity towards H₂ production. The pore size and pore volume also show the same trend as that of surface area. The pore size of parent MCM-41 is 3.4 nm and that of modified sample is 2–3 nm that is in mesoporous range.

Table 1 Surface characterisation properties of MCM-41, CM-2, CCM-2, CCM-4, CCM-6 and CCM-8

Sl no.	Sample	Specific surface area (m^2 g^−1)	Pore size (nm)	Pore volume (cm^3 g^−1)
1	MCM-41	878	3.4	0.65
2	CM-2	853	2.8	0.52
3	CCM-2	827	2.5	0.35
4	CCM-4	796	2.3	0.29
5	CCM-6	772	2.2	0.21
6	CCM-8	753	2.0	0.17

---

3.3. FTIR spectral analysis

The different functional groups present in the prepared samples can be identified by FTIR spectra. Fig. 3 represents the FTIR-spectra of CM-2, CCM-2, CCM-4, CCM-6 and CCM-8. A strong absorption band in the region 3100–3700 cm⁻¹ ascribed to the O–H and N–H stretching vibration frequencies. A medium band near 1633 cm⁻¹ is due to the water of hydration assigned to H–O–H bending motion. A small absorption band in the range of 1243 cm⁻¹ is assigned to the stretching vibration mode of triazine and in addition to that, another peak at 810 cm⁻¹ is due to breathing mode of triazine units. The characteristic peak intensities are smaller for CM-2 composite as compared to neat CN due to the lower loading of CN into MCM-41 (2 wt%).²⁰

Fig. 3 FTIR spectra of (a) CM-2 (b) CCM-2 (c) CCM-4 (d) CCM-6 (e) CCM-8.

The characteristic peak at 1056 cm⁻¹ is assigned to Si–O–Si for MCM-41. But the peak intensity gradually reduced due to the incorporation of CN and followed by the Cu metal in the framework of MCM-41. It confirms from the shoulder peak at 968 cm⁻¹ which is assigned to the stretching vibrations of surface Si–O–Cu bond.²³ This is generally considered to be a proof of the incorporation of metal into the MCM-41 framework. Another small Cu–O stretching peak at 630 cm⁻¹ is shown only in case of CCM-8 as Cu concentration is maximum in the sample but it is absent in all other Cu modified samples.

3.4. UV-Vis DRS

The optical properties, absorption edges, effect of modification on the electronic structure and corresponding band gap energy of the as-prepared CN, CM and CCM photocatalysts were studied using UV-Vis diffuse reflectance spectroscopy and the results are shown in Fig. 4. All the composites showed significantly improved absorption in visible region. The absorption of CM-2 composite at 445 nm is attributed to π–π* transition.²⁰ As CuO shows a broad absorption in visible region from 400 to 600 nm.^24,25 After copper loading, the absorptions of CCM composites remarkably increase and there is a red shift in comparison with CM composite due to high dispersion of CuO and CN. After Cu modification the absorption bands are due to the charge transfer transition of O²⁻ → Cu²⁺ ion which indicates the presence of isolated Cu²⁺ ions surrounded by surface oxygen ions. The absorption band at around 430 nm can be assigned to the formation of well dispersed small crystalline CuO species on the surface of mesoporous materials.²⁶ The intensity of this band is found to decrease with increase in copper content and the charge transfer bands of copper ions become broader. When the Cu content is increased (i.e. in CCM-4, CCM-6 and CCM-8) another absorption band appeared in between 600 and 720 nm due to atomic copper cluster and plasmon resonance band in MCM-41 matrix.²⁷ In the case of copper nanoparticles, as the plasmon resonance band overlaps with the inter-band transitions of Cu, owing to which a broad peaked plasmon resonance was observed. It is found that in case of materials with higher Cu content two absorption bands are more prominent, one due to octahedral Cu²⁺ ions and another due to atomic copper cluster. The enhanced absorbance intensities along with the enlarged surface areas are favourable to improve the photocatalytic performance significantly.

Fig. 4 (a) UV-Vis DRS absorption spectra of (a) neat CN (figure inserted) (b) CM-2, (c) CCM-2, (d) CCM-4, (e) CCM-6, (f) CCM-8. (b) Calculated band gap of CN, CM-2 and CCM-4 catalyst.

The band gap of the synthesised samples were calculated by using the expression:

(ahν)^1/n = A(hν − E_g)

where h, ν, E_g and A are the absorption coefficient, planck's constant, frequency of the light radiation, band gap energy and a constant respectively (Fig. 4b). The neat CN shows a band width of 2.7 eV active in visible light. As very low wt% of CN was distributed on MCM-41 surface the band gap of CN was greatly influenced by MCM-41 and shifted to higher side. Here in for the highest active CCM-4 composite two band gap energies are found. One at 3.1 eV due to CN modified MCM-41 (direct band gap where n = 1/2) for direct transition and another at 1.5 eV due to Cu²⁺ (indirect band gap where n = 2) for indirect transition.^28–30 Also CuO was prepared by simple thermal treatment method and its bandgap energy (S3†) was calculated to compare with that of CN, CM-2 and CCM-4.

3.5. Photoluminescence spectra

PL spectra is widely used to describe the migration, charge transfer and electron–hole pair recombination processes in a photo catalytic system. The spectra are able to record the separation capacity of the photo-induced charge carriers which improves the quantum yield of the photo catalyst. To know the electron–hole recombination of charge carriers, Fig. 5 depicts the PL spectroscopy of CM-2 and CCM-4 with inset the neat CN at room temperature with an excitation of 380 nm. A strong PL band of CN is centered at 450 nm which is very similar to UV-visible spectra. After incorporation of CN into MCM-41 the PL band was slightly shifted to 440 nm. A significant PL emission quenching is observed in CCM composites with an emission peak in the region 420–480 nm. Again PL emission intensity is directly proportional to the recombination of excited electron–hole pair. The figure depicts a significant lower PL intensity of CCM-4 as compared to parent CM and neat CN. This is because of the incorporation of Cu onto the surface of CM effectively inhibit the recombination of photo generated charge carriers and charge transfer established between Cu and CN. For that reason the prepared catalysts are very promising for the catalytic process with satisfying efficiency.

Fig. 5 PL spectra of (a) CN (inset), (b) CM-2 and (c) CCM-4.

3.6. XPS analysis

To study the chemical composition and formal oxidation states of various elements present in the composite, XPS spectra were recorded. The typical XPS scan of CCM-4 composite is depicted in Fig. 6 which indicates the existence of C, N, Cu, Si and O in the composite.

Fig. 6 XPS study of CCM-4 composite material (a) C 1s (b) N 1s (c) O 1s (d) Si 2p (e) Cu 2p.

The high-resolution XPS spectra of C 1s after Gaussian curve fitting was de convoluted into two peaks, at 284.9 eV and at 287.6 eV representing two different oxidation states of carbon.¹² The peak observed at binding energy of 284.9 eV corresponds to surface adventitious sp²-hybridized carbon atom attached to nitrogen atom present in the aromatic ring of CN lattice. Again the peak at 287.6 eV corresponds the carbon atom bonded to three nitrogen atoms in the same way as that of C atoms in melamine molecules. The XPS spectra of N 1s in CN was deconvoluted into three peaks at 397.3 eV, 398 eV, and 400 eV indicating different chemical environments of N atom.¹² The major peak observed at 397.3 eV was assigned to sp² hybridised N-atom where the N-atom was bonded to two carbon atoms in two different ways (C=N–C). Another two peaks at 398 eV and 400 eV can be ascribed to tertiary N-atoms bonded to three carbon atoms (N–(C)₃) in the CN matrix and one H-atom as N–H bonding.

According to the reported data the binding energy of Si 2p in SiO₂ is 101.5 eV. However in the composite when Cu-nanoparticles were incorporated the peak of Si 2p was shifted to 103.4 eV which indicates strong interaction between CuNPs onto the surface of CM photocatalyst. The XPS study shows a peak for O 1s at 532.6 eV which might be due to a Si–O bond and represents a close interaction between oxygen and silicon. Further two intensive XPS peaks at 933.7 eV and 953.3 eV were found, which were ascribed to binding energy of doublet Cu 2p^3/2 and Cu 2p^1/2 levels for Cu¹⁺.¹⁶ Slight shift in binding energy value was due to the interaction between copper with the support. This higher shifting of BE of Si 2p value indicates strong interaction between Cu and Si through oxygen atoms and confirms the presence of Cu–O–Si bond in the network. Slight shift in the values of Cu suggests that during Cu loading partially Cu²⁺ ions are converted into Cu⁺ and Cu⁰ (support from XRD data). Accordingly, we conclude that the photocatalytic active sites on the surface of CM composite are mainly due to coexistence of Cu²⁺ and Cu⁰ species along with very small amount of Cu⁺. However due to extremely close binding energy, it is difficult to differentiate between the Cu⁺ and Cu⁰ valence states from the Cu 2p peaks.

3.7. Scanning electron micrographs

The surface topography and surface dispersion of active components are studied by the scanning electron micrographs. The SEM of CCM-2 and CCM-4 are shown in the Fig. 7. From the figures it is clear that the particles formed are spherical and almost uniformly distributed over the support surface with some agglomerations formed. The average particle size is 200–250 nm. It provides the maximum active sites for photo catalytic water splitting and H₂ production.

Fig. 7 SEM images of CCM-2 and CCM-4.

3.8. HRTEM study

The detailed morphology of the materials can be studied by high resolution transmission electron microscopic images. The HRTEM images of CCM-4 are given in Fig. 8a–f. The material shows a spherical structure which is also observed in SEM images. It is clearly seen that the Cu nano particles are uniformly distributed on the ordered hexagonal mesoporous surface of MCM-41. The g-C₃N₄ nano sheets are stuffed onto the mesoporous structure of metal modified MCM-41 spheres. In addition to the HRTEM images the composite shows the lattice spacing of the Cu and g-C₃N₄ interface. The lattice spacing of 0.265 nm corresponds to the 002 plane of Cu phase which is in accordance to our XRD data.

Fig. 8 HRTEM images of CCM-4.

Again Fig. 8f depicted the EDAX image of the CCM-4 photocatalyst. The presence of Si, O and Cu can be confirmed, which can demonstrate the loading of Cu nanoparticles on the surface of MCM-41. Also the real atomic% of Si, O and Cu were calculated as 37.5, 7.4 55% respectively by EDAX.

4. Electro-chemical study

4.1. Electrochemical impedance spectra

The photocatalytic mechanism, charge transfer resistance and interfacial charge separation efficiency of the prepared photocatalysts were investigated by electrochemical impedance spectra as represented by the Nyquist plots. The diameter of Nyquist semicircle is a function of the resistance at the interface between the working electrode and electrolytic solution and reflects the rate of photo catalytic reaction on the surface of the semiconductor photocatalyst. The smaller diameter supports the faster charge transfer and greater efficiency for the separation of electron–hole pairs and maximum activity. Fig. 9 shows the Nyquist plot of CN, CM-2 and CCM-4. There is a steady decrease of the diameter of the semi-circle which represents lowest resistance in CCM-4 than that of CM-2 and CN. Again it suggests more effective charge separation and increased activity of CCM-4, which is also in agreement with the results of photo catalytic activity.

Fig. 9 The Nyquist plot of CN (a), CM-2 (b) and CCM-4 (c).

4.2. Mott–Schottky plot

To investigate the band alignment, to analyze the conductivity type and charge transfer the Mott–Schottky plots of neat g-C₃N₄ and CuO were used (Fig. 10). The M–S plot of neat g-C₃N₄ and CuO studied under constant frequency revealed positive slope of the line segment suggesting n-type characteristic for g-C₃N₄ and negative slope indicates the p-type characteristic for CuO which is in consistent with previously reported data.^31,32 According to the M–S equation,³³ the flat band potential of neat g-C₃N₄ and CuO are calculated. From the X-axis intercept of the linear region of M–S plot, flat band potential for neat g-C₃N₄ and CuO are found to be −0.86 eV and +1.97 eV respectively. As g-C₃N₄ is a n-type semiconductor the flat band potential represents its CB and in case of CuO as it is a p-type semiconductor the flat band potential represents its VB potential. From the band gap energies of neat g-C₃N₄ and CuO the VB potential of neat g-C₃N₄ and CB potential of CuO are calculated to be +1.85 eV and +0.42 eV respectively. From the calculated band edge potentials in the electrochemical study formation of a heterojunction between g-C₃N₄ and CuO is established. The SPR effect of copper nanoparticles established an equilibrium with the electrolytic solution along with an upward shift of the band level of CuO owing to the built in electric field resulted by the space charge layer at the g-C₃N₄/CuO/Cu interface. As the Fermi level of Cu⁰ is at lower level than that of p-type CuO, the electrons flow from g-C₃N₄ to Cu⁰ through CuO until the system establishes equilibrium.^34,35

Fig. 10 The Mott–Schottky plot of g-C₃N₄ and CuO.

4.3. H₂ evolution measurement using photocatalyst

The photo induced holes irreversibly oxidizes methanol instead of water, when the reaction was carried out in aqueous solution of methanol and it improves the hydrogen evolution process. Fig. 11 shows the H₂ evolution rate of different wt% of CM composites. From the figure it is clear that no H₂ was detected in the absence of either photocatalyst or light irradiation. H₂ production significantly increases for the CM composite than that of pure CN and MCM-41. H₂ evolution rate improved 1.7 times (i.e. 289 μmol 2 h⁻¹) after loading of 1 wt% of CN on MCM-41 surface. The activity of the catalyst is further improved with increasing concentration of CN upto 2 wt% (575 μmol 2 h⁻¹) and then after decreased with higher loading because low wt% of CN is well dispersed on MCM-41 surface but high wt% are encapsulated the inner/outer wall of MCM-41 surface. The CM composites display photocatalytic activity of H₂ evolution in a sequence as follows: CM-2 (575 μmol 2 h⁻¹) > CM-4 (481 μmol 2 h⁻¹) > CM-6 (419 μmol 2 h⁻¹) > CM-8 (401 μmol 2 h⁻¹) > CM-1 (289 μmol 2 h⁻¹) > CN (165 μmol 2 h⁻¹). Again after impregnation of Cu²⁺ and Cu on the surface of CM-2 catalyst, the H₂ evolution rate considerably increased than CM composites (Fig. 12a). This is due to the reduced recombination of electron–hole in the Cu modified samples, which is also supported by the PL study. H₂ evolution rate for CCM nanoparticles are in a sequence as follows: CCM-4 (750 μmol 2 h⁻¹) > CCM-2 (651 μmol 2 h⁻¹) > CCM-1 (588 μmol 2 h⁻¹) > CCM-6 (521 μmol 2 h⁻¹) > CCM-8 (428 μmol 2 h⁻¹). CCM-4 shown highest H₂ gas evolution among the all composites due to presence of dual oxidation state of Cu and maximum amount of photon flux absorbance by the material. The apparent energy conversion efficiency was calculated to be 24.8%. In comparison with other materials, it has shown better result than Fe–Al₂O₃–MCM-41 composite (146 μmol h⁻¹) which is previously reported by our group.¹⁸ Mao et al.²⁰ reported Ti–MCM-41/g-C₃N₄ composite giving 80.76 μmol of H₂ evolved in 1 h time. Again Xiao et al.⁶ synthesized a Cu/g-C₃N₄ catalyst showing 20.5 μmol of H₂ for gram of sample in 1 h. Cu(OH)₂ modified over g-C₃N₄ produced 48.7 μmol of H₂ which is also less than our result for CCM-4 composite.³ Different wt% of Cu (2–8 wt%) modified gC₃N₄ materials are also studied for H₂ evolution reaction and the data are as follows: CC-4 (618 μmol 2 h⁻¹) > CC-2 (590 μmol 2 h⁻¹) > CC-1 (506 μmol 2 h⁻¹) > CC-6 (432 μmol 2 h⁻¹) > CC-8 (376 μmol 2 h⁻¹). Compared to the CCM composites all the samples show less activity indicating the importance of MCM-41 support. It provides the Cu@gC₃N₄ materials a huge surface to act effectively.

Fig. 11 Evolution of H₂ gas (μmol 2 h⁻¹) by different catalysts CM-1, CM-2, CM-4, CM-6 and CM-8.

Fig. 12 (a) Evolution of hydrogen gas (μmol 2 h⁻¹) (b) amount of hydrogen gas evolved in five consecutive cycle, run every 2 h.

Besides photocatalytic activity, the stability of the photocatalyst is important for its commercial application. The stability of the CCM-4 composite was evaluated by performing recycling experiments on the photocatalyst under similar conditions. After each run the photocatalyst was collected, washed several times with distilled water and ethanol and used for the next run. The activity was found to be almost same in three repeated runs (Fig. 12b) and then there is slight decrease in the activity.

After five runs the XRD of the CCM-4 composite was again studied to confirm the stability of the photocatalyst and the figure is shown in S2 as ESI data.† There is a slight reduction in the intensities of the peaks at 38.0° due to (111) plane of CuO and 44.1° for Cu(0) present as bulk. With subsequent runs the CuO may be reduced to metallic Cu with the evolved H₂ during the process and that caused the reduction of catalytic activity. The characteristic peak of SiO₂ at 22.6° remained intact after five subsequent runs. Again the peaks at 33.2° and 42.1° for Cu₂O phase are completely evident in both the fresh and used catalysts. Hence it is confirmed that the photocatalyst remains stable and active after five runs in the present reaction conditions.

4.4. Mechanism of photocatalytic water splitting study

From the results of the photocatalytic H₂ evolution study, it could be said that the CCM composites are novel photocatalytic materials for H₂ generation. The excellent photocatalytic activity can be explained by efficient light absorption, particle size, high surface area, mesoporosity, high pore volume and more importantly the coexistence of Cu²⁺ and Cu⁰ in the ternary nanocomposite. The activity is mainly attributed to the supporting MCM-41 surface, well dispersed visible light active CN and SPR effect of Cu metal on the inner and outer wall surface of CM composite. The strong association with CN further facilitates visible light absorption, charge transport and improves the photocatalytic performance of the catalyst.^36–38

Moreover, it is widely accepted from PL spectra that the PL signal of the CCM-4 composites significantly reduced than CM-2 after the incorporation of Cu²⁺ and Cu⁰ in the CM composites, indicating remarkably suppression of the electron–hole recombination. The combined effect of both Cu²⁺ and Cu⁰ in the CCM nanocomposite enhances the photocatalytic activity because of the formation of n–p hetero junction as well as the presence of strong plasmonic effect of Cu. Under visible-light irradiation, both CuO (p-type semiconductor with band gap 1.5–2.2 eV) and CN (n-type semiconductor with band gap 2.7 eV) can absorb visible-light photons to produce photo-generated electrons and holes which are effectively transferred at the interface to suppress their recombination.^39,40 As the conduction band (CB) and valence band (VB) potentials of CN are −0.86 and +1.84 eV and those of CuO are +0.42 eV and +1.97 e V respectively photo-generated electrons can transfer easily from the CB of CN to that of CuO via the interface.^41,42 But the position of CB of CuO is more positive than the potential of H⁺/H₂ (NHE) and not suitable for H₂ production. Owing to that there is an accumulation of photo-generated electrons in the CB of CuO. Consequently, the Fermi level of CuO shifts upward at the interface as a result of electron accumulation which is thought to have a significant effect on interfacial charge transfer.²⁷ Although the energy band barrier does not allow water reduction reaction on the surface of CuO, but it plays a dual role to increase the efficiency of photocatalytic water splitting. Firstly, due to the accumulation of charge on the CB of CuO the Fermi level shifts towards more negative side and creates an over potential for proton reduction. Secondly, it improves the transfer of photo-generated electrons through the interface to enhance the SPR effect of Cu nanoparticles as co-catalyst.

During synthesis the available lone pairs of electrons on the N-atom of CN is capable of reducing Cu²⁺ → Cu¹⁺ → Cu⁰. Hence the surface plasmonic resonance of metallic Cu also favours channelization of photoinduced electrons as its Fermi level lies just below the CB of CN.^42,43 An effective charge separation can be achieved in this case, which results enhancement of photocatalytic activity and inhibition of photo corrosion.

Surface area is also one of the most important parameters which play a great role in photocatalytic reaction. With the increase in the surface area of catalyst, the interaction of reacting species (water, methanol) and catalyst will be more which leads to high hydrogen production. N₂ adsorption–desorption study showed narrow pore size (2–3 nm) and large pore volume that lead to high surface area.

Scheme 1 Formation mechanism of Cu@CuO–g-C₃N₄/MCM-41.

Hence it is concluded that the prepared CCM-4 composite photocatalyst, due to superior electric conductivity of Cu and the strong interaction between Cu and CN creates a Schottky barrier and is suitable for water reduction leading to a spatial separation of electrons and holes.^44,45 The whole mechanism is shown in Scheme 2.

Scheme 2 Mechanism of H₂ evolution over CCM-4 in visible light irradiation.

5. Conclusion

In summary a simple impregnation and co-condensation method is followed to disperse Cu@CuO–gC₃N₄ nano particles on mesoporous MCM-41 support to develop a ternary composite photocatalyst. The high surface area with large number of silanol groups on the surface and suitable pore size of MCM-41 are the key factors for even dispersion of the active components. Without using any extra reducing agent during the synthesis procedure the CuO is partly reduced to metallic Cu by utilising the free electrons from graphitic C₃N₄. The synergetic effect of both Cu and CuO with g-C₃N₄ endorses highest photocatalytic activity of the material. This activity attributed to extended light absorption, effective transfer of photogenerated carriers and well dispersion of CuO and Cu as co-catalyst with strong SPR effect. This new material can enlighten the prospects of effective H₂ evolution.

References

1. Z. Wang, W. Guan, Y. Sun, F. Dong, Y. Zhou and W. K. Ho, Nanoscale, 2015, 7, 2471–2479.
2. S. Patnaik, S. Martha and K. M. Parida, RSC Adv., 2016, 6, 46929–46951.
3. X. Zhou, Z. Luo, P. Tao, B. Jin, Z. Wu and Y. Huang, Mater. Chem. Phys., 2014, 143, 1462–1468.
4. Y. Tian, B. Chang, J. Fu, B. Zhou, J. Liu, F. Xi and X. Dong, J. Solid State Chem., 2014, 212, 1–6.
5. B. Peng, S. Zhang, S. Yang, H. Wang, H. Yu, S. Zhang and F. Peng, Mater. Res. Bull., 2014, 56, 19–24.
6. M. Fan, C. Song, T. Chen, X. Yan, D. Xu, W. Gu, W. Shi and L. Xiao, RSC Adv., 2016, 6, 34633–34640.
7. L. Liu, Y. Qi, J. Hu and W. Cui, Appl. Surf. Sci., 2015, 351, 1146–1154.
8. J. Wang, H. Xu, X. Qian, Y. Dong, J. Gao, G. Qian and J. Yao, Chem.–Asian J., 2015, 10, 1276–1280.
9. B. Li, Y. Wang, Y. Zeng and R. Wang, Mater. Lett., 2016, 178, 308–311.
10. Y. Shi, Z. Yang, Y. Liu, J. Yu, F. Wang, J. Tong, B. Su and Q. Wang, RSC Adv., 2016, 6, 39774–39783.
11. D. P. Shoo, D. Rath, B. Nanda and K. M. Parida, RSC Adv., 2015, 5, 83707–83724.
12. M. Sui and L. She, Front. Environ. Sci. Eng., 2013, 7, 795–802.
13. Z. A. ALOthman, Materials, 2012, 5, 2874–2902.
14. K. M. Parida and D. Rath, J. Mol. Catal. A: Chem., 2006, 258, 381–387.
15. Y. Xing, J. Wang, L. Chen, A. Wang, F. Li, C. Wang and E. Zhong, J. Mater. Sci.: Mater. Electron., 2016, 27, 8633–8640.
16. A. C. Pradhan, K. M. Parida and B. Nanda, Dalton Trans., 2011, 40, 7348–7356.
17. A. C. Pradhan, B. Nanda, K. M. Parida and G. R. Rao, J. Phys. Chem. C, 2015, 119, 14145–14159.
18. A. C. Pradhan, S. Martha, S. K. Mahanta and K. M. Parida, Int. J. Hydrogen Energy, 2011, 36, 12753–12760.
19. W. Wang, T. Yu, Y. Zeng, J. Chen, G. Yang and Y. Li, Photochem. Photobiol. Sci., 2014, 13, 1590–1597.
20. S. Shen, D. Zhao, J. Chen, L. Guo and S. S. Mao, Appl. Catal., A, 2016, 521, 111–117.
21. L. Yin, Y. P. Yuan, S. W. Cao, Z. Zhang and C. Xue, RSC Adv., 2014, 4, 6127–6132.
22. A. C. Pradhan, B. Nanda, K. M. Parida and M. Das, Dalton Trans., 2013, 42, 558–586.
23. A. Thomas, A. Fischer, F. Goettmann, M. Antonietti, J. O. Muller, R. Schlogl and J. M. Carlsson, J. Mater. Chem., 2008, 18, 4893–4908.
24. L. Gou and C. J. Murphy, J. Mater. Chem., 2004, 14, 735–738.
25. A. Paracchino, V. Laporte, K. Sivula, M. Grätzel and E. Thimsen, Nat. Mater., 2011, 10, 456–461.
26. G. Ghadimkhani, N. R. deTacconi, W. Chanmanee, C. Janaky and K. Rajeshwar, Chem. Commun., 2013, 49, 1297–1309.
27. H. Hou, M. Shang, F. Gao, L. Wang, Q. Liu, J. Zheng, Z. Yang and W. Yang, ACS Appl. Mater. Interfaces, 2016, 31, 20128–20137.
28. S. Nayak, L. Mohapatra and K. M. Parida, J. Mater. Chem. A, 2015, 3, 18622–18635.
29. N. Zhang, C. Han, Y.-J. Xu, J. J. Foley, D. Zhang, J. Codrington, S. K. Gray and Y. Sun, Nat. Photonics, 2016, 10, 473–482.
30. S. Liu and Y. J. Xu, Sci. Rep., 2016, 6, 22742,  DOI:10.1038/srep22742.
31. Q. Huang, F. Kang, H. Liu, Q. Lia and X. Xiao, J. Mater. Chem. A, 2013, 1, 2418–2425.
32. J. Fang, H. Fan, M. Li and C. Long, J. Mater. Chem. A, 2015, 3, 13819–13826.
33. Y. Liu, Y. X. Yu and W. D. Zhang, J. Phys. Chem. C, 2013, 117, 12949–12957.
34. D. Dvoranova, V. Brezova, M. Mazur and M. A. Malati, Appl. Catal., B, 2002, 37, 91–105.
35. M. Al-Mamun, Y. Kusumoto, B. Ahmmad and M. A. Islam, Top. Catal., 2010, 7–10, 571–577.
36. Q. Yan, M. Sun, T. Yan, M. Li, L. Yan, D. Wei and B. Du, RSC Adv., 2015, 5, 17245–17252.
37. C. Chen, L. Kuai, Y. Chen, Q. Wang, E. Kan and B. Geng, RSC Adv., 2015, 5, 98254–98259.
38. Z. Zhang and P. Wang, J. Mater. Chem., 2012, 22, 2456–2464.
39. J. S. Zhang, M. W. Zhang, C. Yang and X. C. Wang, Adv. Mater., 2014, 26, 4121–4126.
40. P. D. Tran, S. K. Batabyal, S. S. Pramana, J. Barber, L. H. Wong and S. C. J. Loo, Nanoscale, 2012, 4, 3875–3878.
41. S. Martha, A. Nashim and K. M. Parida, J. Mater. Chem. A, 2013, 1, 7816–7824.
42. X. H. Li and M. Antonietti, Chem. Soc. Rev., 2013, 42, 6593–6604.
43. J. Yu and J. Ran, Energy Environ. Sci., 2011, 4, 1364–1371.
44. J. Yang, D. Wang, H. Han and C. Li, Acc. Chem. Res., 2013, 46, 1900–1909.
45. S. Xiao, P. Liu, W. Zhu, G. Li, D. Zhang and H. Li, Nano Lett., 2015, 15, 4853–4858.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra24358d

---