Systematic synthesis and analysis of change in morphology, electronic structure and photoluminescence properties of 2,2′-dipyridyl intercalated MoO₃ hybrid nanostructures and investigation of their photocatalytic activity

Abstract

An organic–inorganic hybrid structure was synthesized by using 2,2′-dipyridyl and MoO₃ nanorods through a simple hydrothermal method. The as-prepared dipyridyl–MoO₃ hybrid samples looked like rod shaped micro crystals. The starting material used for this work was MoO₃ nanorods which had a width of 150 nm and a length of several microns, whereas the resulting hybrid structure had a width of one micron and a length of 10 to 30 microns. Here, dipyridyl has acted as a stretching molecule and bonded the MoO₃ nanorods together along their length to form hybrid micro crystals. By calcinating the hybrid sample at 400 °C, intercalated dipyridyl was removed, while maintaining the microscale morphology. These deintercalated MoO₃ samples looked like micro slabs having a width of 5 micrometers. The presence and intercalation of dipyridyl was confirmed by the change in the XRD [0 k 0] peak positions. As the cross sectional size of the dipyridyl is close to the van der Waals gap of the orthorhombic MoO₃ crystal, this space was effectively used for this intercalation process. The deintercalation process, i.e. the removal of dipyridyl was confirmed by TGA, and XRD measurements. The influence of dipyridyl in the valence band electronic density of states (DOS) of MoO₃ was also analyzed by XPS and XES methods. A photoluminescence study was also conducted, reflecting the intercalation effect on the emission characteristics of the MoO₃ nanostructures. A photodegradation study on Procion Red MX 5B was also carried out, showing that the dipyridyl deintercalated MoO₃ micro slab like samples had the highest photodegradation efficiency.

---

1. Introduction

Molybdenum has several common oxidation states, namely, +2, +3, +4, +5, and +6, but it readily oxides into MoO₂ and MoO₃. In MoO₃, the α-phase is the most stable phase. Molybdenum oxides are n-type semiconductors with a band gap in the range of 2.8–3.6 eV,^1–4 and a work function in the range of 5.3–6.5 eV.^5–8 Because of these reasons this material is attractive for many advanced applications as catalysts, sensors, photochromic and electrochromic materials, recording materials, and field emitters.^9–12 Molybdenum oxides can also be widely used as electrode materials for battery applications.¹³ An orthorhombic α-MoO₃ crystal structure consists of MoO₆ octahedral shared edges and corners, resulting in a zigzag network.¹⁴ The edge sharing zigzag row forms in the [0 0 1] direction, whereas the corner sharing row forms in the [1 0 0] direction. Along the [0 1 0] direction,¹⁵ in addition to the network of atoms, there also exist a gap which is usually known as van der Waals gap. Recently, researchers have tried to incorporate organic molecules into the unit cell of MoO₃. Through literature, it has been found that intercalation work was done with larger sized organic molecules like 4,4′-bipyridyl,¹⁶ poly-(p-phenylene),¹⁷ 2,2′-bipyridine,¹⁸ nicotinamide,¹⁹ hexamethylenetetramine,¹⁹ pyrrolidinedithiocarbamate,²⁰ and polyaniline.²¹ All these molecules were intercalated along the [0 0 1] direction, and in the process, it connected the zigzag network resulting in new morphologies. For example, Xu Ming Wei et al. have found that when an organic bipyridyl molecule was intercalated into MoO₃, the morphology was changed from bulk to nanoplatelets.¹⁶ A similar work on bipyridine intercalated into WO₃, which is also having a similar crystal structure as that of MoO₃, has resulted in a morphology transformation from bulk to rectangular nanostructure.²² Similar kind of work is also done on organic–cadmium telluride hybrid composites.^23,24 Thus the work on MoO₃ nanocrystal becomes interesting as well as progressive. In the present work, for our intercalation study, we have chosen the organic molecule dipyridyl, which is entirely different from the molecules, such as bipyridyl¹⁶ and pyrazine^25,26 intercalation into MoO₃ lamellar structure. In fact, differences in 4,4′-bipyridyl and 2,2′-dipyridyl molecules are very important concerning their possible intercalation with molybdenum atoms in the van der Waals gaps of MoO₃. The intercalation effect on changes in morphology, structure, electronic density of state, photoluminescence and photodegradation properties have been systematically investigated in detail. Though a report is available on molybdenum trioxide (MoO₃)–bipyridyl hybrid structure,¹⁶ the present work is completely different and constitutes a novel one in terms of materials and methods (of preparation) and of the systematic investigations performed. It is worth mentioning that organic–inorganic hybrids are established to be very prospective materials for gas sensors, devices adopting second harmonic generation effects, light-emitting diodes and photodiodes, etc.²⁶ MoO₃ based organic–inorganic hybrid structures hold great promise towards engineering efficient electrode blocking layers in a piezoelectric nanogenerator,²⁷ interconnection layer in perovskite–organic solar cells,²⁸ hole transporting layer for inverted polymer solar cells,²⁹ electrochemical performance of cathode materials, ethanol gas sensors,³⁰ volatile organic compounds sensors,³¹ catalysts,³² and photocatalyst.^26,33 In particular, intercalation of pyrazine molecules into the MoO₃ lattice causes a blue shift of the photoluminescence spectrum.²⁵ Furthermore, a formation of an additional Raman mode at 714 cm⁻¹ attributed to the pyrazine ring deformation vibration is a characteristic of pyrazine–MoO₃ hybrids as compared with pure molybdenum trioxide.²⁵

2. Experimental section

2.1 Preparation method

The synthesis process of MoO₃ nanorods is mainly based on the acidification of sodium molybdate (Na₂MoO₄), as reported elsewhere.^25,26 We have used the same MoO₃ nanorods whose field emission scanning electron microscopy (FESEM) and high resolution transmission electron microscopy pictures (HRTEM) have been published recently²⁶ when elaborating studying controlled synthesis of MoO₃ microcrystals by subsequent calcination of hydrothermally grown pyrazine–MoO₃ nanorod hybrids and their photodecomposition properties. In accordance with FESEM/HRTEM results,²⁶ the average width and length of the MoO₃ nanorods used in the present work as a starting material for synthesis of the 2,2′-dipyridyl intercalated MoO₃ hybrid nanostructures were found to be about 150 nm and 10–30 microns, respectively, with the nanorod aspect ratio of about 23–26. We obtained a small amount of the final product which was characterized further to know the structure, morphology, chemical composition and optical property. The techniques used were X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), X-ray emission spectroscopy (XES), and photoluminescence spectroscopy (PS). From SEM observations it was found that the as obtained product is a nanorod in shape, and through chemical composition, as well as XRD analysis, it was confirmed that the nanorod sample is orthorhombic α-MoO₃. To study the intercalation effect of dipyridyl into orthorhombic α-MoO₃ nanorods, a separate experiment was conducted with the as prepared MoO₃ nanorods and dipyridyl as starting materials. That is, the as prepared MoO₃ nanorods and dipyridyl were put in a 50 mL Teflon lined stainless steel autoclave with a weight ratio of 3 : 1 for MoO₃ nanorods and dipyridyl, respectively, (2.88 g of as-prepared MoO₃ powder was dispersed in 0.3 M of dipyridyl solution, prepared in 40 mL double distilled water) and the mixture was kept at 180 °C for 24 h in hydrothermal condition. After completion of the reaction, the final product was obtained in the form of a powder which was again characterized by using techniques such as XRD, Raman spectroscopy, XPS, XES, SEM, TEM, and PS.

2.2 Characterization

The surface morphology of the samples such as dipyridyl free pure MoO₃, dipyridyl intercalated MoO₃, and their calcined counterpart at 400 °C, were examined using field emission scanning electron microcopy (FESEM-JEOL JSM 6500). High-resolution transmission electron microscopy (TEM-2011 JEOL STEM) was also used to examine the crystalline nature of MoO₃ and its dipyridyl intercalation effect. Crystallographic structures of the samples including the intercalated and calcined samples were investigated with X-ray diffraction method (PANalytical X-ray diffractometer – XPERT PRO). Molecular vibrational structure analysis was carried out with Horiba Jobin Yvon Raman spectrometer, in the reflection mode, at a wavelength of 532 nm. Thermal decomposition analysis was made on the hybrid samples under atmospheric pressure with a Thermo Gravimetric and Differential Thermal Analysis (TG-DTA) instrument (Perkin Elmer STA-6000). Electronic structure and stoichiometry details were also analyzed by X-ray photoelectron spectroscopy (XPS – ion-pumped chamber of an ES-2401 spectrometer equipped with a source of Mg Kα radiation; photon energy of 1253.6 eV) and X-ray emission spectroscopy (XES – RSM-500 and SARF-1 spectrometers) techniques. Optical measurements were carried out with a UV-Vis spectrophotometer (SHIMADZU 3600 UV-Vis-NIR spectrophotometer) and photoluminescence properties were analyzed using a Horiba Jobin Yvon spectrofluromax spectrometer at room temperature.

3. Results and discussion

3.1 Structural characterization

Fig. 1a shows the XRD pattern of MoO₃ nanorods. The XRD pattern belongs to orthorhombic phase and the cell parameters are 3.963, 13.85, and 3.696 Å for a, b, and c lattice constants, respectively. These values are in good agreement with the JCPDS card no. 35-0609 of MoO₃.^25,26 A similar structural characterization was conducted on 24 h dipyridyl intercalated MoO₃ sample. For comparison, dipyridyl deintercalated sample was also obtained by calcinating the dipyridyl containing sample up to 400 °C. This calcinating temperature was optimized from TGA thermal analysis study. Fig. 2a shows the TGA curve of dipyridyl intercalated MoO₃ sample. We can clearly see that there exists a weight loss of about 20% in the temperature interval ranging from 300 to 400 °C. As there is no such a weight loss from dipyridyl free pure MoO₃ samples in Fig. 2b, we believe that this weight loss is due to the removal of dipyridyl and hence we fixed the calcinating temperature at about 400 °C. Fig. 1b and c shows the XRD patterns of intercalated and deintercalated samples. Though the XRD pattern of deintercalated sample (Fig. 1c) is similar to that of the as prepared pure MoO₃ nanorod samples (Fig. 1a), the pattern of intercalated sample (Fig. 1b) is different. This observation indicates that the intercalated dipyridyl molecule has influenced the crystal structure of MoO₃.

Fig. 1 XRD spectra of three different samples: (a) as-prepared MoO₃, (b) dipyridyl–MoO₃ and (c) calcined dipyridyl–MoO₃ at 400 °C.

Fig. 2 (a) and (b) Thermo gravimetric curve of dipyridyl intercalated MoO₃ and as prepared MoO₃ nanorod samples, respectively.

From this observation, it was confirmed that, the intercalation of dipyridyl was along the [0 k 0] direction in the MoO₃ crystal structure, where it can be accomodated freely, i.e., in the van der Waals gaps. On the other hand, there is no possibility to accommodate dipyridyl along the ‘c’ axis of the orthorhombic crystal structure of the parent MoO₃. If this was possible, it would not be an intercalation compound. In this situation, diffraction pattern of Fig. 1b should be a combination of dipyridyl and MoO₃ crystal systems, which is not exist in our present case.

Fig. 1b, shows entirely different diffraction pattern of the dipyridyl intercalated MoO₃ crystal system. MoO₃ has a van der Waals gap along the ‘b’ axis, and intercalation can take place by expanding this lattice dimension without breaking any strong Mo–O bonds. If the structure was to expand along the ‘c’ axis, many strong covalent Mo–O bonds would be broken. Hence, if we consider ‘c’ axis, it cannot be an intercalation reaction. So the only chance for intercalation is along the van dar Waals gap, i.e., ‘b’ axis. It is because of the vertical cross sectional size of the dipyridyl, i.e., about 0.58 nm is close to the van der Waals gap of the α-MoO₃, which is about 0.574 nm.²⁵ As the size of dipyridyl molecule is double the size of pyrazine,²⁵ addition of this one more carbonaceous ring present in the dipyridyl in to the entire MoO₃ structure, will form new crystal lattices in throughout the crystal systems, so different diffraction patterns were observed in the intercalated samples. During the intercalation process, due to this smaller size difference and shape of the organic molecule will produce the peaks and peak shifts. It means that, the dipyridyl molecular arrangement within the MoO₃ crystal structure could be arranged vertically to the ‘b’ axis, i.e., binding to the left and right side of the crystal network of MoO₃ through its ‘nitrogen’ atoms as shown in Fig. 3. This kind of binding thus helps to bridge the MoO₃ network along the [0 k 0] direction converting the nanorods to micron sized morphology. Fig. 4 shows the schematic illustration of the formation of slab like MoO₃ micro-structures during the calcinations of dipyridyl intercalated MoO₃ sample.

Fig. 3 Schematic illustration of (a) MoO₃ layered structure and (b) binding nature of 2,2′-dipyriyl between the layered structures of MoO₃.

Fig. 4 A schematic illustration of the formation of plate-like MoO₃ single crystals.

Fig. 5a and c shows the SEM and TEM images of dipyridyl intercalated sample, respectively. Here, the nanorod shaped morphology is not seen (Fig. 5e, left), instead, a micron sized rod shaped bulk morphology was noticed. A similar morphology analysis on calcinated sample is shown in Fig. 5b and d. For size comparison, an enlarged view of a MoO₃ nanorod and a slab like morphology of calcined dipyridyl–MoO₃, are shown in Fig. 5e. Here the morphology is similar to that of the intercalated one with micron sized crystals. This observed feature indicates that, during calcination, the morphology does not change much from micron sized rectangular slab-like structure into other forms. HRTEM image from calcinated sample has also shown (Fig. 5f) a clear lattice fringe pattern with a lattice spacing value of 3.636 Å for the [0 0 1] orientation. Further the SAED pattern of deintercalated sample (inserted image) is similar to that of nanorods sample and thereby suggested that the deintercalation process has resulted in a single crystalline nature.

Fig. 5 SEM and TEM images of two different samples; (a) and (c) dipyridyl intercalated MoO₃ and (b) and (d) calcined dipyridyl–MoO₃ at 400 °C, respectively, (e) enlarged view of a MoO₃ nanorod and a slab like morphology of calcined dipyridyl–MoO₃, a size comparison (f) HRTEM image of the calcined sample, and (inserted image) SAED pattern of the calcined sample.

During the hydrothermal experiment, in the presence of nanorods and dipyridyl, dipyridyl intercalates into the MoO₃ unit cell by sharing its nitrogen atoms in place of oxygen at Mo⁶⁺=O sites. Replacement of oxygen atomic position into nitrogen is possible because of similar atomic size and moreover, nitrogen atom's lone pair electrons can involve in the bond formation process. As this intercalation process continues into the entire crystal system, at the end, the dipyridyl molecule sees the edge of the nanorods along the entire length. Fig. 3 schematically represents the binding nature of dipyridyl at the edges of MoO₃ nanorods. The free end of dipyridyl can bind to the MoO₃ unit cell of other nanorods along its length and thereby helps to grow micron sized structures (Fig. 3). During the prolonged thermal treatment, there is a possibility for the as grown microstructure to break into pieces along its length and it might be the reason why micron sized crystals of reduced length have resulted.

3.2 Raman analysis

Fig. 6a shows the recorded Raman spectrum with three predominant higher frequency peaks at 667, 817, and 996 cm⁻¹, which are the signature of the orthorhombic alpha MoO₃ crystal system. A peak at 667 cm⁻¹ corresponds to the triply coordinated oxygen (Mo₃–O) stretching mode which results from the edge shared oxygen atoms in common to three octahedra. The peak at 817 cm⁻¹ represents the doubly coordinated oxygen (Mo₂–O) stretching which is due to the corner shared oxygen atoms in common to two octahedral.¹⁵ The peak at 996 cm⁻¹ is assigned to the terminal oxygen (Mo⁶⁺=O) stretching mode along the a and b-axis which results from an unshared oxygen and is responsible for the layered crystal structure of alpha MoO₃. Fig. 6c and d, shows the recorded Raman spectra of dipyridyl intercalated and deintercalated MoO₃ samples. For comparison, a Raman spectrum of pure dipyridyl molecule was also given (Fig. 6b). In the case of nonintercalated pure as prepared MoO₃ samples, we noticed the bands as explained earlier. The three major Raman modes were Mo₃–O, Mo₂–O and Mo⁶⁺=O with frequency values 667, 817, and 996 cm⁻¹, respectively. When it is intercalated with dipyridyl molecule, Raman modes, especially the modes at high frequency have been changed, either shifting towards lower or higher wavenumbers, as shown in Fig. 6c. Once calcinated, then the samples structure is returned to its original crystal structure of MoO₃ and therefore the original Raman modes of MoO₃ were observed as shown in Fig. 6d.

Fig. 6 Raman spectra of four different samples: (a) as-prepared MoO₃, (b) as-purchased dipyridyl, (c) dipyridyl–MoO₃ and (d) calcined dipyridyl–MoO₃ at 400 °C.

These results indicate that, intercalating the dipyridyl molecule has greatly influenced the crystal structure of MoO₃ and hence its vibrational modes as well. In order to explain the observed shifts in the Raman modes of intercalated samples, we need to consider the XRD results. According to XRD results, we come to the conclusion that, dipyridyl has intercalated in to the crystal network of MoO₃ along its ‘b’ axis. Fig. 3 schematically represents the nature of intercalation of dipyridyl into the crystal network of MoO₃. One can see that the dipyridyl molecule is holding together two MoO₃ crystal network through its nitrogen atoms, present on top and bottom corner of the molecules. In this arrangement, there are four possibilities for the change in the vibration modes during the intercalation into MoO₃ crystal structures. In the first, since ‘N’ atoms of dipyridyl are replacing ‘O’ atoms at Mo₃–O positions as indicated in Fig. 3, one can expect a change in the vibrational mode of the Mo₃–O vibration. Whether the frequency increases or decreases is an interesting one. Our recent pyrazine intercalation work has shown a red shift in the frequency, because of the involvement of ‘N’ atoms.²⁵ We also expected the same result in the dipyridyl intercalated MoO₃ Raman spectra. Interestingly we obtained an unusual mode at 608 cm⁻¹ which is red shifted with respect to that of Mo₃–O band vibration of pure MoO₃ and we believe that this unusual mode must be due to the ‘N’ replacement effect.

The second possibility is a change in the Mo₂–O vibration mode. When the crystal structure is arranging into the three dimensional network, Mo₂–O vibrational mode (Fig. 3) also seems to be getting effected. The third possibility is the change in the Mo⁶⁺=O vibrational state. Practically speaking, the ‘O’ at Mo⁶⁺=O is not involved in the bond formation with the dipyridyl molecule and therefore one should not expect a change in this frequency mode. But if the crystal arrangement is as shown in Fig. 3, naturally less room is available for the ‘O’ atom at the Mo⁶⁺=O site, because of the possible Coulomb interaction between the dipyridyl ring. As a result we assume that a reduced vibration mode is possible for ‘O’ at Mo⁶⁺=O. In the Raman spectra of dipyridyl intercalated sample we obtained a red shifted mode at 910 cm⁻¹, and we attributed this to dipyridyl interaction effect at the Mo⁶⁺=O site. The fourth possibility is related to dipyridyl molecule, when it is in the MoO₃ crystal network. The most intense mode of the dipyridyl molecule is the breathing mode at 999 cm⁻¹. This breathing mode seems to be shifted towards high frequency wavenumbers. But, the reason why the breathing mode is blue shifted is not clear. However, the blue shift could be attributed to the compressed state of dipyridyl molecule. That is because of the compressed state the molecule started to vibrated at high frequencies and therefore it has shown a blue shifted mode. Raman modes of as purchased 2,2′-dipyridyl, interacted dipyridyl–MoO₃, and as prepared MoO₃ nanorods samples were tabulated (Table 1).

Table 1 Raman modes of 2,2′-dipyridyl, dipyridyl–MoO₃, and as prepared MoO₃ nanorods samples

Raman assignments in cm^−1
	Pure dipyridyl		Shift during intercalation in MoO3	Pure MoO3
1	βCH I	1050	1022, 1044, 1061	—
2	Ring breathing	999	—	996 for Mo^6+=O
3	πNH (νII)	820	871, 891, 912	820 for Mo2–O
4	αCCC I, II (νII)	770	—	—
5	αCCC I (ν7)	620	770	667 for Mo3–O

---

3.3 Electronic structure analysis

X-ray photoelectron spectroscopy (XPS) is a quantitative spectroscopic technique that gives information regarding the elemental composition, empirical formula, chemical state and electronic state of the elements, components of a sample under study. XPS spectra are originated due to irradiation of a sample with a beam of X-rays while simultaneously measuring the kinetic energy (KE) and number of electrons that escape from the top 1 to 10 nm of the sample being analyzed. For our XPS analysis we have used Mg Kα (1253.6 eV) photon energy to check the stoichiometry of the dipyridyl intercalated MoO₃. Our previous XPS Mo 3d core level spectrum of the MoO₃ nanorods, precursors for synthesis of the dipyridyl intercalated MoO₃ sample, have revealed²⁵ that the spectrum is a simple spin-doublet with the Mo 3d_5/2 and Mo 3d_3/2 binding energies at 232.70 ± 0.05 and 235.85 ± 0.05 eV, respectively, corresponding to those of molybdenum in the formal valence +6. The XPS O 1s core-level binding energy of the MoO₃ nanorods was found to be 531.05 ± 0.05 eV, as reported elsewhere.²⁵ From the relative intensities of the above XPS spectra, we have calculated the compositional stoichiometry between ‘Mo’ and ‘O’ and it has been found that the Mo/O ratio in the studied MoO₃ nanorods corresponds to 1 : 3 within the accuracy of the XPS measurements.²⁵ To analyze the intercalation effect of dipyridyl into the MoO₃ nanorods, we have made XPS measurements on this dipyridyl–MoO₃ sample and compared with the as prepared MoO₃ samples.

The XPS spectra for the dipyridyl intercalated MoO₃ sample are shown in Fig. 7. The survey spectrum of the sample (Fig. 7a) shows that all the spectral features, except the XPS C 1s core-level, are attributed to the constituent element core-levels or Auger O KLL line. The presence of the C 1s line on the survey spectrum is explained by hydrocarbons adsorbed on the dipyridyl–MoO₃ sample and probably carbon belonging to dipyridyl in the intercalated sample. The presence of the N 1s line that originates from dipyridyl intercalated into MoO₃ is not obvious in the survey spectrum presented in Fig. 7a because the line superimposed on the Mo 3p_3/2 spectrum in the dipyridyl–MoO₃ sample. The result of such a superposition of the XPS N 1s and Mo 3p_3/2 spectra in the above sample is obvious from Fig. 7b. To make clear the above superposition, we presented in Fig. 7b the Mo 3p spectrum of pure MoO₃ nanorods. As it has been established in ref. 25, the XPS Mo 3p core-level spectrum of pure MoO₃ nanorods is a spin-doublet with the Mo 3p_3/2 and Mo 3p_1/2 lines binding energies positioned at 398.4 ± 0.1 and 415.7 ± 0.1 eV, respectively. As can be seen from Fig. 7b, the intercalation of dipyridyl into MoO₃ causes the decreasing Mo 3p core-level binding energies by about 0.3 eV. This shift of the Mo 3p core-levels towards the lower binding energies when going from the pure MoO₃ nanorods to the dipyridyl intercalated MoO₃ sample is confirmed also by the measurement of the Mo 3d core-level spectrum. As can be seen from Fig. 7c, the XPS Mo 3d spectrum of the dipyridyl–MoO₃ is a spin-doublet with the Mo 3d_5/2 and Mo 3d_3/2 binding energies positioned at ∼398.1 ± 0.1 and 415.5 ± 0.1 eV, respectively. These values are smaller by about 0.3 eV compared to those of the Mo 3d_5/2 and Mo 3d_3/2 core-level binding energies of the pure MoO₃ nanorods.²⁵

Fig. 7 XPS (a) survey, (b) core-level Mo 3p, and (c) core-level Mo 3d spectra for the dipyridyl intercalated MoO₃ sample; for comparison, the XPS Mo 3p core-level spectrum obtained in ref. 23 for pure MoO₃ nanorods is presented on the panel (b).

To verify the above XPS core-level data for the dipyridyl intercalated MoO₃ sample, the XPS valence-band spectrum as well as the X-ray emission Mo Lβ_2,15 (L_III → N_IV,V transition) and O Kα (K → L_II,III transition) bands, representing the energy distribution of mainly the Mo 4d- and O 2p-like states, respectively, have been recorded. The technique of measuring the above XES bands was analogous to that applied successfully when studying the electronic structure of molybdenum oxides.³⁴ Fig. 8 shows the XPS valence-band spectrum and XES Mo Lβ_2,15 and O Kα bands of the dipyridyl intercalated MoO₃ provided that a common energy scale is used. The method of matching the above spectra on a common energy scale is completely the same as described elsewhere.³⁴ Results of studies of the X-ray emission bands presented in Fig. 8 reveal that in the dipyridyl intercalated MoO₃ sample the main contributions of the Mo 4d- and O 2p-like states occur at the bottom and near the top of the valence band, respectively, with contributions of the mentioned states throughout other portions of the valence band. These experimental results resemble those derived for pure MoO₃ nanorods,²⁵ however the center of gravity of the XPS valence-band spectrum of dipyridyl intercalated MoO₃ sample is shifted by about 0.4–0.5 eV towards the Fermi level, E_F, compared to its position on the spectrum of pure MoO₃ nanorods. This effect can be explained by the intercalation of pyrazine into MoO₃. It is well-known that, replacement of ‘O’ sites by nitrogen, or ‘N’ sites by carbon in solids is expected to cause a shift of the valence-band spectrum towards E_F.^35–39 Such a shift is detected experimentally due to the replacement of the ‘O’ sites by nitrogen when intercalating dipyridyl in the MoO₃ nanorods.

Fig. 8 Comparison on a common energy scale of the X-ray emission Mo Lβ_2,15 and O Kα bands as well as the XPS valence-band spectrum of the dipyridyl intercalated MoO₃ sample.

Comparison on a common energy scale of the XES and XPS spectra measured for the dipyridyl intercalated MoO₃ sample (Fig. 8) displays that a weak shoulder A of the XPS valence-band spectrum is originated mainly from contributions of the Mo 4d-like states, the maximum B is formed by contributions of the Mo 4d- ad O 2p-like states, whilst shoulders C and D are originated mainly from contributions of the O 2p-like states. We have made an attempt to measure the X-ray emission N Kα band (K → L_II,III transition), representing the energy distribution of the N 2p-like states, in the dipyridyl intercalated MoO₃ sample. However, the intensity of this band in the sample was found to be too low to discuss its fine structure peculiarities. We used an RSM-500 spectrometer equipped with a diffraction grating (600 lines per mm; radius of curvature of R ≈ 6026 mm) and operating at accelerating voltage, U_a = 5.0 kV, and anode current, I_a = 4.5 mA. When increasing the anode current to gain better intensity of the N Kα band we detected partial loss of oxygen (and, in all probability, nitrogen) in the dipyridyl intercalated MoO₃ sample.

3.4 Photoluminescence properties

To study the optical property of the as prepared nanorod samples, photoluminescence (PL) spectrum was obtained. Initially the samples excitation spectrum was recorded, which showed a maximum (not shown) at 248 nm. Using this excitation maximum, the emission property of MoO₃ nanorod samples were recorded in the wavelength range of 250 to 450 nm. Fig. 9A(a) shows the as recorded spectrum with an emission maximum at 332 nm. In addition to this, a sub band emission at 360 nm was also noticed.²⁵ Though an excitation energy at 248 nm was used, the emission band was stoke shifted much with an emission band at around 332 nm. This emission with equivalent energy value of 3.74 eV is not the band gap energy of MoO₃, instead, it is an excitonic emission band and it originates from crystal field splitted energy levels of the Mo⁵⁺ ion. Similarly, the sub band emission at around 360 nm is also due to the crystal field splitted energy levels. The orthorhombic MoO₃ is an interesting layered crystal structure and it is described to be formed by distorted octahedron as explained elsewhere.¹⁵ At the centre of each octahedron, there exist Mo ions and at the surrounding positions, there are oxygen atoms. Each Mo ion has four neighboring oxygen at a distance ranging from 167 to 195 pm and the other two oxygen atoms completing the octahedron are at 225 and 233 pm, thus forming a distorted octahedron network.⁴⁰ The central metal ion's ‘d’ levels, get splitted into discrete energy levels, because of crystal field effect. Upon excitation, the transition between these levels gives rise to excitonic emission bands, which are usually blue shifted with respect to the energy band gap E_g of MoO₃. In our present case, the emission at 332 and 360 nm (3.74 and 3.45 eV) are similar to the one from MoO₃ as reported elsewhere,⁴¹ and their origin is due to two different crystal environments. One is due to the pentacoordinated Mo⁵⁺ ion having no oxygen vacancy, and the other one is due to the hexacoordinated Mo⁵⁺ ion having an oxygen vacancy. We believe that these two different environments are existing in our as prepared samples as bulk defect states. Earlier X-ray photoelectron spectroscopy technique didn't find any deviation in the stoichiometry, but the PL results show the presence of oxygen vacancy. This may be due to the more sensitive nature of the PL to both bulk and surface related emissions, whereas XPS is surface sensitive technique only.

Fig. 9 (A) Comparative photoluminescence spectra (PL) of as prepared MoO₃ nanorod, dipyridyl intercalated MoO₃, and calcined samples, (B), a separate PL spectrum of dipyridyl intercalated MoO₃ and (C) a separate PL spectrum of calcined sample.

Fig. 9A(b) and (c), show the comparative PL spectra recorded for dipyridyl intercalated and deintercalated MoO₃ samples. Spectrum 9B shows the enlarged view of PL emission band of the dipyridyl intercalated MoO₃ sample. From this spectrum, we observed that, a blue shift is present at around 315 nm (3.95 eV), similar to the pyrazine intercalated MoO₃ sample.²⁵ This blue shift is unusual, because it will appear only when quantum confinement takes place. Since it is present in the dipyridyl intercalated MoO₃ micro-rods, it means that it is only due to the intercalant dipyridyl molecule. As already discussed, ionic and covalent mixture of MoO₃ bondings, produces change in energy level splitting. So, similar to pyrazine interaction with MoO₃ crystal field, dipyridyl intercalated MoO₃ sample also exhibits blue shift emission bands. Spectrum 9C, shows the enlarged view of PL emission spectrum from dipyridyl deintercalated, i.e., calcined sample. Here the emissions are at 315 and 352 nm, which were close to the emissions of the intercalated MoO₃ sample, but the broad spectrum indicates that, the sample is not a pure one, and that it has some traces of dipyridyl remnants. At the same time, the improved intensity in comparison to the intercalated sample, shows the improved crystallinity of the calcined sample.

3.5 Photodecomposition measurements

Photodecomposition activities of the as-prepared MoO₃ samples were analyzed with Procion Red MX-5B. We have chosen this dye because it is one of the harmful environmental pollutants. Attempts were made to decompose this pollutant by using WO₃ nanoparticles in a typical photocatalytic setup.⁴² Since MoO₃ nanostructures could also exhibit good photocatalytic performance, it was used for our photodecomposition study.^43,44 The dye solution was prepared by mixing 0.25 mmol of Procion Red MX-5B reactive dye in 20 mL of double distilled water (DDW). Three photodecomposition experiments were carried out with the three different MoO₃ morphologies by mixing 0.05 g of the MoO₃ samples with the dye solution. The experiments were done under a 2.0 mW UV light source at a wavelength of 365 nm. Three experiments were carried out with the three different MoO₃ morphologies. The absorption spectra changed during the photodecomposition of the reactive dye resulting in a decrease of the absorption maximum. Absorption values a, b, and c in Fig. 10, respectively, correspond to pure as-prepared dye, dye in the dark for 24 h, and dye kept in UV light for 24 h, all without the presence of MoO₃. From the observed results, we come to the conclusion that there is no noticeable change when the dye solution is maintained for 24 h in the dark, but a slight decrease in absorption was observed after 24 h exposure to UV.

Fig. 10 UV-Vis absorption spectral changes of Procion MX-5B dye. The absorption spectra correspond to pure Procion MX-5B dye, Procion MX-5B dye in the dark for 24 h, and Procion MX-5B dye kept under UV illumination for 24 h, in the absence of photocatalyst nanoparticles.

For this initial study, MoO₃ nanorods, dipyridyl intercalated and deintercalated MoO₃ nanostructure samples were dispersed in the dye solution individually and illuminated with UV light for 3, 6, 9, 12 and 24 hours and their absorptions were recorded. The results are shown in Fig. 11A–C for all the three morphologies. From these results, it was found that the MoO₃ nanorods and deintercalated MoO₃ microslab like samples dispersed medium has responded quickly in decomposing the dye.

Fig. 11 UV-Vis absorption spectral changes of Procion MX-5B dye mixed with MoO₃ samples at different timings. (A) MoO₃ nanorods + procion dye, (B) dipyridyl intercalated MoO₃ + procion dye, and (C) dipyridyl deintercalated MoO₃ + procion dye. Absorption spectra corresponds to UV illumination time; 3, 6, 9, 12, and 24 h in the presence of MoO₃ nanostructures.

In general, catalysis is the property of any substance that makes possible chemical reactions without being consumed in them. A broad classification of catalysis is homogeneous and heterogeneous. Homogeneous catalysts are present in the same phase as reactants and products, whereas heterogeneous are present in a different phase. An application of quantum size effects in chemical reactions, in particular, is called photocatalysis, which use photons as accelerators in presence of nanomaterials, that is mostly in solid phase and acts against harmful dyes, which are in liquid phase. In this present work, a solid catalyst (MoO₃ and its derivatives) is in contact with fluid (Procion Red dye) reactants. So, obviously this mechanism is heterogeneous catalysis. In general, all heterogeneous reactions are complex phenomena because they evolve in between different phases.

While being involved in chemical reactions, they can have different orders, like first order or unimolecular reactions, second order or bimolecular reactions, third order or termolecular reactions and so on. Here in unimolecular reaction, a reactant becomes products using catalysts, simply A → products. In bimolecular reaction, A + B → products, and A + B + C → products is obviously termolecular reaction.

The reaction rate for the first order reaction is

where k is the reaction coefficient and A is the concentration of the reactant. By integrating, the expression for concentration [A] can be obtained

[A]_t = [A]_o e^−kt (3)

where A_o is the initial concentration, A_t is the concentration after time ‘t’.

Similarly for second and third order reactions, the reaction rates will be as follows:

In the second order reaction, if [A] ≫ [B] and [A] can be assumed to remain constant during the reaction, so the second order reaction rate can be written as

with

k⁽¹⁾ = k⁽²⁾[A] (8)

where k⁽¹⁾ is the pseudo-first order rate coefficient. With this approximation the second-order reaction becomes pseudo-first-order and follows the kinetics of a first order reaction.⁴⁵

The present case is that Procion Red dye decomposes using MoO₃ catalysts, in presence of photon. During the photocatalytic process, electrons and holes were generated by absorbing photons on the surface of the semiconductors. These electrons will split water molecules as hydroxyl radicals (OH^.) and hydrogen ions (H⁺), whereas holes divide oxygen as super oxide ions (O₂⁻˙). These short living species will act on Procion Red dye and let it get degraded. So this reaction shall not come under first order chemical reaction and at the same time concentration of this highly energetic hydroxyl radical and super oxide ions are constant for the particular semiconductor material, as in eqn (7). So this chemical reaction becomes pseudo first order.

To estimate the rate of the reaction in the present photodegradation experiments, the eqn (4) was used. The rate of the reaction k was calculated from all three photodecomposition experiments by drawing a graph between ln(A_o/A_t) and time.⁴²

Fig. S1† shows the relation between ln(A_o/A_t) and time. The rate of the reaction k was obtained from the slope of these graphs. The obtained values are 0.339 h⁻¹, 0.279 h⁻¹ and 0.382 h⁻¹, which respectively corresponds to the photodecomposition reaction with MoO₃ nanorods, dipyridyl intercalated, and deintercalated MoO₃ microslabs like structure. Usually in photodecomposition, two processes are important. The first one is an effective generation of electron–hole pairs on the semiconductor surface and the second one is the separation of these pairs effectively. Generation and separation of these electron–hole pairs are different in different materials. In this present work, optical properties of the materials and calculations of the reaction rate of the corresponding materials, shows that MoO₃ nanorods/calcined MoO₃ could produce more number of electrons and holes, which are essential for the photodecomposition, than the dipyridyl intercalated MoO₃. In the case of intercalated MoO₃, the presence of organic molecule, which perturbs the production and separation process, reveals a reduced reaction rate. This case has been presented in a schematic diagram, Fig. 12. A high photodecomposition was noted from the deintercalated MoO₃ microslabs like samples which have a high cross section for absorption of photons due to their high aspect ratio and surface area, which could be the reason for the quick response in decomposition. It is worth mentioning that, from the BET analysis, we observed an increased surface area from 0.0472 m² g⁻¹ for MoO₃ nanorods to 1.1783 m² g⁻¹ for dipyridyl deintercalated micron size MoO₃ sample.

Fig. 12 A schematic illustration of the photosynthesis of electron–hole pairs.

In the case of intercalated sample the rate of decomposition is reduced and this should be because of the presence of dipyridyl molecule in the crystal structure. In photogenerated catalysis, the photocatalytic activity depends on the ability of the catalyst to create electron–hole pairs, which generate free radicals (hydroxyl radicals: ˙OH) able to undergo secondary reactions. When the molecule is present in the crystal structure naturally the electron hole pair separation is perturbed and therefore it is reflected in the photocatalytic activity as well.

4. Conclusions

Self assembled MoO₃ nanorods with high aspect ratio were obtained by hydrothermal method. An organic molecule, dipyridyl, was added into the as obtained MoO₃ nanorods at ambient condition, and it resulted in dipyridyl intercalated MoO₃ single crystals of dimensions in the micrometer range. The intercalation of dipyridyl into the MoO₃ crystal structure along the [0 k 0] direction has bridged the nanorods by bringing them very close. Detailed XRD and Raman analysis confirmed the intercalation and deintercalation of dipyridyl. The electronic density of states details of the as prepared intercalated MoO₃ sample was analyzed by XPS and XES methods. Dipyridyl intercalation has not changed the chemical composition, because of its effective replacement of the oxygen atoms by nitrogen atoms. However the valance band was found to be shifted towards the Fermi level due to intercalation of dipyridyl into the MoO₃ nanorods. Once dipyridyl was deintercalated by calcinating the samples at 400 °C under atmospheric pressure, the electronic density of states retained the original feature of pure MoO₃ state. The atmospheric calcination process has helped to retain the stoichiometry and hence there was no oxygen deficiency effect in the calcined sample. PL measurements have shown excitonic emission features due to crystal field splitted energy levels, from both the dipyridyl free and dipyridyl intercalated samples. However, there was a blue shift in the emission bands of the dipyridyl intercalated sample due to a local change in the crystal field seen by the metal ion through an improved lattice coupling, because of the intercalated dipyridyl. From the photodecomposition experiments, a relatively high photodecomposition activity was observed from deintercalated MoO₃ microslab like samples due to their relatively high absorption. We believe that the concepts presented in this study provide a general route for designing single crystalline metal oxides by intercalation and deintercalation of organics through hybrid technique and can also be used to modify other catalytic materials. In summary, the present successful synthesis of 2,2′-dipyridyl–MoO₃ hybrid structure and its physical and chemical properties studied in the present work allow for suggesting that it can be considered as a very promising material for light-emitting diodes and photodiodes. Since 2,2′-dipyridyl molecules can be rather easily removed from the 2,2′-dipyridyl–MoO₃ hybrid structure by its calcination, as the present study indicates, the above property gives a unique opportunity for obtaining well-defined MoO₃ micro slabs. Further, in accordance to our findings, dipyridyl deintercalated MoO₃ micro slabs possessing very high photodegradation efficiency are also expected to be of significant importance for practical applications.

Acknowledgements

One of the authors, S. Rajagopal, would like to thank Bharathiar University for awarding University Research Fellowship to carry out this work. The author would like to thank Dr K Swaminathan, Professor and Head, Department of Microbial Biotechnology, Bharathiar University for his support in utilizing the UV-vis absorption spectrophotometer and Dr S Velmathi, Associate Professor, Department of Chemistry, National Institute of Technology, Tiruchirappalli for her support in providing the BET analysis.

References

1. B. Hu, L. Mai, W. Chen and F. Yang, ACS Nano, 2009, 3, 478.
2. R. Sivakumar, R. Gopalakrishnan, M. Jayachandran and C. Sanjeeviraja, Curr. Appl. Phys., 2007, 7, 51.
3. R. S. Patil, M. D. Uplane and P. S. Patil, Appl. Surf. Sci., 2006, 252, 8050.
4. X. Yin, H. Han and A. Miyamoto, J. Mol. Model., 2001, 7, 207.
5. K. S. Yook and J. Y. Lee, Synth. Met., 2009, 159, 69.
6. Y. Liang, C. Tracy, E. Weisbrod, P. Fejes and N. D. Theodore, Appl. Phys. Lett., 2006, 88, 081901.
7. A. A. Knizhnik, I. M. Iskandarova, A. A. Bagatur'yants, B. V. Potapkin and L. R. C. Fonseca, J. Appl. Phys., 2005, 97, 064911.
8. A. R. Blythe and D. Bloor, Electrical Properties of Polymers, Cambridge Solid State Science Series, Cambridge University Press, Cambridge, 2nd edn, 2005.
9. J. Zhou, N. S. Xu, S. Z. Deng, J. Chen, J. C. She and Z. L. Wang, Adv. Mater., 2003, 15, 1835.
10. J. Zhou, S. Z. Deng, N. S. Xu, J. Chen and J. C. She, Appl. Phys. Lett., 2003, 83, 2653.
11. J. Liu, Z. Zhang, C. Pan, Y. Zhao, X. Su, Y. Zhou and D. Yu, Mater. Lett., 2004, 58, 3812.
12. A. Khademi, R. Azimirad, A. A. Zavarian and A. Z. Moshfegh, J. Phys. Chem. C, 2009, 113, 19298.
13. S. H. Lee, Y. H. Kim, R. Deshpande, P. A. Parilla, E. Whitney, D. T. Gillaspie, K. M. Jones, A. H. Mahan, S. Zhang and A. C. Dillon, Adv. Mater., 2008, 20, 3627.
14. L. Mai, B. Hu, W. Chen, Y. Qi, C. Lao, R. Yang, Y. Dai and Z. L. Wang, Adv. Mater., 2007, 19, 3712.
15. K. Kalantar-zadeh, J. Tang, M. Wang, K. L. Wang, A. Shailos, K. Galatsis, R. Kojima, V. Strong, A. Lech, W. Wlodarski and R. B. Kaner, Nanoscale, 2010, 2, 429.
16. X. M. Wei and H. C. Zeng, Chem. Mater., 2003, 15, 433.
17. T. A. Kerr, F. Leroux and L. F. Nazar, Chem. Mater., 1998, 10, 2588.
18. P. J. Zapf, R. C. Haushalter and J. Zubieta, Chem. Mater., 1997, 9, 2019.
19. R. F. de Farias, Mater. Chem. Phys., 2005, 90, 302.
20. R. F. de Farias, Int. J. Inorg. Mater., 2001, 3, 931.
21. R. Bissessur, D. C. DeGroot, J. L. Schindler, C. R. Kannewurf and M. G. Kanatzidis, J. Chem. Soc., Chem. Commun., 1993, 687.
22. S. Pang, F. Jian and L. Wang, Inorg. Chem., 2008, 47, 344.
23. H. B. Yao, X. B. Li, S. Y. Ai and S. H. Yu, Nano Res., 2010, 3, 81.
24. H. Wei, H. Sun, H. Zhang, C. Gao and B. Yang, Nano Res., 2010, 3, 496.
25. S. Rajagopal, D. Nataraj, O. Y. Khyzhun, Y. Djaoued, J. Robichaud, D. Mangalaraj and K. Senthil, CrystEngComm, 2011, 13, 2358.
26. S. Rajagopal, D. Nataraj, O. Y. Khyzhun, Y. Djaoued, J. Robichaud and C.-K. Kim, Mater. Chem. Phys., 2013, 141, 383.
27. G. C. Yoon, K. Shin, M. K. Gupta, K. Y. Lee, J. Lee, Z. L. Wang and S. Kim, Nano Energy, 2015, 12, 547–555.
28. J. Liu, S. Lu, L. Zhu, X. Li and W. C. H. Choy, Nanoscale, 2016, 8, 3638–3646.
29. Y. Wang, Q. Luo, N. Wu, Q. Wang, H. Zhu, L. Chen, Y. Li, L. Luo and C. Ma, ACS Appl. Mater. Interfaces, 2015, 7, 7170–7179.
30. X. Wei, L. Jiao, J. Sun, S. Liu and H. Yuan, J. Solid State Electrochem., 2010, 14, 197–202.
31. T. Itoh, J. Wang, I. Matsubara, W. Shin, N. Izu, M. Nishibori and N. Murayama, Mater. Lett., 2008, 62, 3021–3023.
32. D. Chen, M. Liu, L. Yin, T. Li, Z. Yang and X. Li, et al., J. Mater. Chem., 2011, 21, 9332.
33. C. Ma, J. Zhou, H. Zhu, W. Yang, J. Liu, Y. Wang and Z. Zou, ACS Appl. Mater. Interfaces, 2015, 7, 14628–14637.
34. O. Y. Khyzhun, V. L. Bekenev and Y. M. Solonin, J. Alloys Compd., 2008, 459, 22.
35. P. Steiner, H. Höchst, J. Schneider, S. Hüfner and C. Politis, Z. Phys. B: Condens. Matter Quanta, 1979, 33, 241.
36. V. P. Zhukov, V. A. Gubanov, O. Jepsen, N. E. Christensen and O. K. Andersen, J. Phys. Chem. Solids, 1988, 49, 841.
37. V. E. Henrich and P. A. Cox, The Surface Science of Metal Oxides, Cambridge University Press, Cambridge, 1994.
38. O. Y. Khyzhun and V. A. Kolyagin, J. Alloys Compd., 2004, 363, 32.
39. A. A. Lavrentyev, B. V. Gabrelian, V. B. Vorzhev, I. Y. Nikiforov and O. Y. Khyzhun, J. Alloys Compd., 2009, 472, 104.
40. G. Blasse and M. Wiegel, J. Alloys Compd., 1995, 224, 342.
41. J. Song, X. Ni, D. Zhang and H. Zheng, Solid State Sci., 2006, 8, 1164.
42. S. Rajagopal, S. D. Nataraj, D. Mangalaraj, Y. Djaoued, J. Robichaud and O. Y. Khyzhun, Nanoscale Res. Lett., 2009, 4, 1335.
43. I. Shakir, M. Shahid and D. J. Kang, Chem. Commun., 2010, 46, 4324.
44. P. Agarwal, I. Paramasivam, N. Shrestha and P. Schmuki, Chem.–Asian J., 2010, 5, 66.
45. L. D. Socaciu-Siebert, PhD thesis, Freie Univeritat, Berlin, 2004.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra13558g

---