Low temperature synthesis of NbC/C nano-composites as visible light photoactive catalyst

A facile carbothermal route was adopted to obtain niobium carbide nanoparticles (NPs) embedded in carbon network from Nb2O5 to study photocatalytic behavior. Optimization of synthesis parameters to obtain single phase NbC NPs has been successfully done. The phase identification, morphology and nature of carbon were determined with the help of X-ray diffraction, transmission electron microscopy (TEM) and Raman spectroscopy. X-ray photoelectron spectroscopy (XPS) suggested the presence of multiple oxidation states of Nb associated to NbC and NbCxOy centers on the surface of NPs. Due to the presence of NbCxOy on the surface of NPs, absorption under visible region of EM spectrum has been observed by UV-visible spectroscopy. Different organic dyes (RhB, MB and MO) were used to study the effect of holding time on the photocatalytic performance of as-synthesized samples. RhB dye was found to be the most sensitive organic molecule among all the considered dyes and degraded 78% in 120 min.


Results and Discussion
XRD analysis. For the synthesis of single phase carbon coated NbC NPs, synthesis temperatures and holding times were optimized on the basis of XRD results which are shown in Fig. 1a,b respectively. In both figures, peaks marked with 'α' depict the rock salt structure (space group Fm-3m) NbC which were matched with (ICDD 01-089-3690). The relative carbon content of synthesized NbC samples was also calculated by a = 4.09847 + 0.7 128x − 0.3457x 2 where a = lattice parameter in Å and x = atomic ratio C/Nb and listed in Table 1 55 . For the optimization of synthesis parameters (temperature and holding time) single phase NbC NPs were synthesized using experiments discussed below.
Optimization of synthesis temperature. XRD patterns of samples synthesized at low temperatures (600, 700 and 800 °C) with 10 h holding is shown in Fig. 1a. The synthesis at 600 °C results in the reduction of Nb 2 O 5 to NbO (sample 10C600). While, at 700 and 800 °C, mixtures of (NbC + NbO 2 ) and NbC were obtained respectively. The absence of the NbC phase at 600 °C (Fig. 1a) may be due to the unavailability of enough thermal energy to further reduce and carburize NbO to transform into NbC which can be seen in the case of 10C600. Sample   Optimization of holding time. After obtaining single phase NbC at 800 °C, the holding time has been varied to study the behavior of reduction and carburization of Nb 2 O 5 with optimized temperature which is shown in Fig. 1b. Sample 5C800 shows a reduction of Nb 2 O 5 after carburization and the formation of a mixture of NbO 2 and NbC. Sample 7C800 showed contrary results in which carburization exceeded the reduction process as shown by NbO 2 as a minor phase and NbC as a major phase. In the case of 10 h, complete reduction followed by carburization of Nb 2 O 5 was observed in sample 10C800 (Fig. 1b) resulting in single phase NbC. Furthermore, the increment of holding time to 11 h led to decarburization of NbC with the formation of NbO 2 . The decarburization of NbC may be caused by a lower CO/CO 2 ratio (CO 2 rich reaction) beyond 10 h at 800 °C 56,57 . While the formation of NbO 2 as a result of decarburization at longer holding (11 h) instead of NbO (as observed in the case of lower temperature) may be attributed to the higher stability of NbO 2 than NbO 58 . Optimization of synthesis parameters to obtain single phase NbC also affects the constituted crystallographic distortion. In order to understand all the distortional transitions that occurred in NbC due to different synthesis temperatures and holding times, stress-strain analysis (Williamson-Hall) of the obtained samples has been done. Detailed theory and assumptions of Williamson-Hall analysis has been provided in Appendix-II.
Williamson-Hall analysis. Table 2 depicts a comparative study of all the postulates of W-H analysis and Scherrer criteria for all the synthesized NbC samples. The least magnitude of strain in NbC lattice at higher synthesis temperature (sample 10C800) is very obvious due to complete conversion of Nb 2 O 5 to NbC, while presence of NbO 2 at 700 °C (sample 10C700) distorts the crystallites which shows higher strain in crystallites. As a function of holding time at 800 °C, the induced strain tends to decrease as carburization occurs. Figure 1b shows that up to 10 h of holding, minor phases are being eliminated (5 h (NbC + NbO 2 ) to 7 h (NbC + NbO 2 ) then 10 h (pure NbC)) leading to a large reduction in strain. The diffusion of carbon in the Nb lattice is increased by increasing the holding time which supports the enhancement of lattice parameters by reducing the strain and supporting the attainment of equilibrium positions of Nb and C atoms in the NbC unit cell. Beyond 10 h (11C800), decarburization of NbC led to increase the distortion which might be associated to incorporation of oxygen resulting NbO 2 . XPS analysis. XPS analysis of all the samples synthesized at 800 °C was carried out to understand the surface chemical variations (valence states) that reactants undergo during reduction and carburization processes. The survey spectrum confirms the presence of Nb, C and O on the surface of as-synthesized samples which is shown in Figs 2a and S1. High resolution XPS (HR-XPS) spectra of Nb3d, C1s and O1s transitions are shown in Figs 2(b-d) and S1 respectively in which different chemical states of Nb, C, and O have been illustrated. Peak positions and relative contents of all the elements in various valence states have been listed in Table 3. The convoluted HR-XPS spectrum of Nb 3d is composed of doublets corresponding to spin orbital splitting of 3d 5/2 and 3d 3/2 as shown in Figs 2b and S1. Figure 2b revealed the presence of multiple oxidation states of Nb associated to different compounds (NbC, NbC x O y , NbO 2 and Nb 2 O 5 ) even after obtaining single phase XRD pattern of 10C800 as shown in Fig. 1a. Such HR-XPS spectra of single phase NbC (10C800) suggests the generation of different ionic states of Nb during the reduction-carburization process. Variation in the holding time at 800 °C, led to change the chemical composition of samples (suggested by XRD results) which can also be observed in XPS results as shown in Fig. S1. Volume fraction (VF) of different phase present on the surface has also been calculated by deconvoluting the XPS spectrum and listed in Table 3. Further, the increment in the holding time led to enhance carburization up to 10 h as VF of NbC has increased (4.66% to 13.72%) and decreased in 11C800 (7.17%) supporting the XRD results. VF of Nb 4+ (NbO 2 ) and Nb 5+ (Nb 2 O 5 ) has increased (7C800; 94.45% and 11C800; 90.74%) as the holding time was shifted from 10 h (81.72%). Moreover, binding energy associated to NbC x O y (~204.8 eV) was not observed in 5C800. HR-XPS data for Nb3d suggest that direct conversion of Nb 2 O 5 to NbC involved reduction and carburization simultaneously which was responsible for the multiple transitions of oxidation states of Nb in the lattice as earlier discussed for XRD analyses. The HR-XPS spectrum of C 1 s as shown in Figs 2c and S2 revealed the presence of peaks associated with NbC, C-C sp 2 and C-C sp 3 Fig. 3b,c (marked with dotted lines at the periphery of the particle). Figure 3d shows the selected area electron diffraction pattern of the area shown in Fig. 3a, indicating nano crystallinity by concentric diffraction spots corresponding to various atomic planes. Furthermore, a narrow particle size distribution was observed having 12 nm as average particle size of as-synthesized sample as shown in Fig. S3. Further, to observe the distribution of Nb, C and O in the synthesized NbC NPs (10C800), elemental profile was also carried out as shown in Fig. 4 in which distribution of carbon is nearly same as carbon is also present as amorphous network. While, concentration of oxygen is relatively low throughout the particle confirming the presence of oxygen centers in the NbC nanoparticle resulting NbO x or NbC x O y . Further, elemental line profile (STEM) was also taken to observe the linear distribution of each elements from periphery to core which is also shown in Fig. 4. Extreme left region of line profile shows the higher C content on the left side of nanoparticle which may be associated to the amorphous carbon network. Moreover, TEM micrograph of considered region for EDAX analysis is shown in Fig. S2 and the presence of Nb, C and O were confirmed along with Cu (unmarked peak at 8 keV) which is associated to grid. On the nanoparticle region marked by arrow (green colored), increase in the concentration of Nb (red) and C (green) is more as compared to O (blue) confirming the presence of oxygen centers in NbC nanoparticles. Such concentration profile of Nb and O throughout the nanoparticle suggest the presence of multiple oxidation states of Nb corresponding to NbC and NbC x O y (NbO x centers) as observed from XPS analysis.  Proposed synthesis mechanism. Normally NbC is synthesized at >1000 °C 23,25,66 which is higher as compared to the present work. The transition of Nb 2 O 5 to NbC at relatively low temperature (800 °C) in a high pressure closed chamber is a temperature driven transition. The mixture of reactants contains Mg and charcoal, which act as reducing agents providing MgO, CO and CO 2 as byproducts. The gaseous byproducts (CO and CO 2 ) of the reactions lead to increase the pressure inside the autoclave. Various chemical reactions occurring inside the autoclave have been listed in the supplementary information.
All the possible chemical reactions are categorized into two categories; (i) reduction and (ii) carburization. Reactions (S1-S24) represents the reduction of Nb 2 O 5 forming NbO 2 , NbO and (NbO 2 + NbO) mixture in the presence of reducing agents (Mg and C) and their mixture. All the reduction reactions suggested that Mg (alone) reduced Nb 2 O 5 more efficiently to obtain NbO 2 , NbO and (NbO 2 + NbO) mixture as compared to C and (C + Mg) mixture as represented by reaction (S1, S9, S17) in Fig. 6. Further, C and CO (in the presence of Mg) reduced Nb 2 O 5 equivalently to obtain similar products as reaction (S5, S13, S21) and (S7, S15, S23), respectively, which are shown in Fig. 6. Rest of the other reduction reactions are non-feasible due to their positive ΔG suggesting that individually C and CO (in the absence of Mg) are not capable enough to reduce Nb 2 O 5 .
As a result of reduction reaction, feasibility of the formation of NbO is higher than that of NbO 2 (more negative ΔG of reaction S9, S17 than S1) which is also supported by the XRD pattern of sample 10C600. Further, the formation of NbC can be obtained via in-situ reduction carburization of oxide products (NbO 2 and NbO) as Nb metal was not obtained in any of the synthesized sample (as shown in Fig. 1). In the similar pattern, reaction (S25-S42) represent the possible chemical reactions for the formation of NbC from NbO 2 , NbO and (NbO 2 + NbO) mixture, all are feasible with negative ΔG values. Unlike reduction reactions, reduction-carburization of NbO 2 and NbO follow different paths as shown in Fig. 6 by reaction (S25-S30) and (S31-S36), respectively. While, reduction-carburization of (NbO 2 + NbO) mixture is represented by reaction (S37-S42) which are more spontaneous than previous chemical reactions (S25-S36) with high difference of ΔG values. As single phase NbC was obtained at 800 °C, reduction-carburization reactions follow the following order according to ΔG at 800 °C; S37 Thermodynamically higher feasibility of the formation of NbO 2 + NbO mixture is also evidenced in the XRD pattern (Fig. 1a) of 10C600 where NbO (major) + NbO 2 (minor) were obtained at 600 °C. This is further converted to NbC (major) + NbO 2 (minor) at 700 °C and NbC at 800 °C. In this sequence, charcoal encapsulated NbO particles which is followed by reduction-carburization reactions. As a result, CO and CO 2 gases evolved through coated carbon resulting to porous particles. Moreover, the transition leads to more evolution of gases, which enhances the porosity on the particle surfaces, which is demonstrated in Fig. 7.

Spectroscopic analysis.
Raman spectroscopy was used to analyze the nature of carbon in samples 5C800 and 10C800. Figure S5 shows the Raman spectra with two characteristic bands corresponding to disordered (Dband) and graphitic carbon (G-band) near 1282.1 nm −1 and 1589.8 nm −1 respectively. Ferrari and Robertson 67 suggested that the disordered scattering and bond stretching of sp 2 carbon atoms contribute to the formation of D-and G-band respectively. The position of the G-band near to 1600 nm −1 suggests nano-crystalline graphite.
The observed higher intensity of the D-band than the G-band in these samples depicts the amorphous nature of carbon as has also been observed from TEM analysis.  Figure 8a shows the UV-visible absorption spectra of all the samples synthesized at 800 °C suggesting broad multiple absorption humps in the visible region, which may be associated with the presence of either oxide centers (NbO 2 and Nb 2 O 5 ) or continuous carbon network (highly disordered graphic carbon) in the nanocomposite powder samples [68][69][70] . The band gap for all the samples was calculated with the help of Tauc's relationship between photon energy (hν) and absorption coefficient (α) as where A is a constant and 'n' depends on the type of transition containing values 0.5, 2.0, 1.5 and 3.0 corresponding to allowed direct, allowed indirect, forbidden indirect, and forbidden direct transitions, respectively 71 . The values of band gap were calculated by extrapolating the linear portion of the (αhν) 2 vs. hν curve as shown in Fig. 8b. The absorption observed at 600-650 nm and 550 nm might be associated to the presence of disordered carbon 70 and Nb-O centers 69 , respectively. Table 4 shows the band gaps (~1.8 eV & ~2.1 eV) of as-synthesized  Moreover, to observe the photoemission characteristics of the synthesized samples photoluminescence (PL) spectroscopy was conducted, which is shown in Fig. 9. A broad emission spectrum in the visible region was observed for all the samples, which is associated with disordered carbon (observed in RAMAN and XPS analysis) and other functional groups present in the powder samples. Photoemission decreases as the amount of lattice carbon increases and lattice distortion decreases from a holding time of 5 h (5C800) to 10 h (10C800).
For 11 h of holding time (11C800) more intense photoemission is observed as compared to 10C800. This is because of the presence of some functional groups, which reduce the efficiency of non-radiative decay. The XPS spectra showed the energy corresponding to π → π* transition which also promotes the photoluminescence quenching as observed in Fig. 9 which is in good intact with its volume fraction on the surface of different photo catalysts (5C800 and 7C800) 72 . Moreover, the presence of NbC x O y on the surface also promotes the photoluminescence quenching.   Table 4. Band gap of the samples synthesized at 800 °C. Photocatalytic study. Photocatalytic activity of prepared nanocomposites was monitored by observing the degradation of RhB, MB and MO dyes. All of them are cationic dyes, which acquire electrons from donor excited sites of a photocatalyst. Figure S6 shows the diminishing absorption spectra of dyes (RhB, MB and MO) under the exposure of visible radiation for 120 min which might be associated to the decolorization or degradation of dyes with respect to irradiation time. The fractional variation in the dye concentration after 120 min are shown in Fig. 10(a-c). The controlled experiments showed that ~2.8% decolorization of dyes (without catalyst) was observed in a dark chamber. Moreover, 3.2% photo bleaching was observed under visible irradiation without photocatalyst which is negligible. The experiments under visible irradiation with the as synthesized photo catalysts showed decolorization/degradation as a function of exposure time for the derivatives of RhB, MB, and MO, respectively as shown in Fig. 10(a-c).
Moreover, Fig. 11 shows a comparison chart of percent degradation after 120 min of exposure which elucidates the better performance of 10C800 followed by 11C800, 7C800 and 5C800 with 78.6%, 67.8%, and 57.1% change in concentration of RhB, MB, and MO, respectively. The presence of oxides in the samples may affect the generation of excitons and their transfer towards the surface. Sample 5C800 shows the least photocatalytic activity while minor content of oxides in 7C800 and 11C800 resulted in better activity than 5C800. Among all samples, 10C800 shows better photocatalytic activity while other samples contribute differently with all the dyes (Fig. 11). Such behavior may be associated with different molecular structures of the dyes due to generated excitons interacting differently at targeted sites e.g. aromatic rings, N, S and unsaturated bonds etc.
The decolorization/degradation kinetics of photocatalytic reactions were studied with the help of pseudo first and second order law, which can be expressed as follows: where C 0 , C, K 1 and K 2 are the solution concentrations at irradiation time t = 0, at different times and pseudo first and second order rate constant (per minute), respectively 71 . A plot of −ln(C/C 0 ) vs. t (min) and t/C vs. t (min) as shown in Fig. 12 represents pseudo first order kinetics (a,c,e) represents the pseudo second order kinetics  (b,d,f), respectively. The respective reaction rate constants (K 1 and K 2 ; min −1 ) and regression coefficients (R 2 P) have been listed in Table 5, which illustrated that 10C800 has the highest photochemical reaction rate constants (K 1 ) of 0.0143, 0.0102, and 0.0075 min −1 for RhB, MB, and MO dyes, respectively, following pseudo first order kinetics. Moreover, based on quality of fitting (R 2 ), it can be suggested that 10C800 and 11C800 followed pseudo first order kinetics while, 5C800 and 7C800 followed second order kinetic (Table 5). Similar trend has also been observed by Yang et al. 73 and Younis et al. 74 for photoelectron reduction of Cr 4+ on MoS 2 @TiO 2 nanotubes and photodegradation activity of undoped and doped CeO 2 nanocrystals. Moreover, the observed diminishing absorption patterns may be associated with the interaction of excitons with different reactive oxygen species (ROS) such as hydroxyl radicals (OH · ) and superoxide anion radicals ( ⋅− O 2 ). With the help of mass spectrometry, the decreasing absorption spectra were attributed to the degradation of dye molecules to lighter organic molecules which are shown in Fig. 13 as lower m/z molecules along with the parent molecule (Figs S7-S9). To ascertain the results obtained from absorption spectra and MS analysis, TOC was carried out to check the decrease of concentration of organic carbon. In association with the MS results, 41.8%,  Fig. 14a. These results demonstrate the partial degradation of parent dye molecules due to the presence of some organic carbon in the form of aromatic rings which can be asserted that the decolorization (absorption study) rate was faster than the degradation (MS and TOC study) rate of dye molecules.
The photocatalytic activity of the synthesized samples was increased with increasing holding time at 800 °C up to 10 h, which again was decreased at 11 h. As observed from XRD analysis, carbon content in the lattice of NbC has increased from 0.79 (5C800) to 0.959 (10C800) resulting in less distorted crystallites (Table 2) which is further decreased at longer holding time (11 h) with higher distortion as discussed in the W-H analysis. Moreover, a similar trend was also observed from BET analysis where SSA was increased from 292 m 2 /g (5C800) to 506 m 2 /g (10C800) and then decreased to 475 m 2 /g (11C800). Further, PL spectroscopy data show that recombination decreases with the decreased impurities in the synthesized samples resulting in better mineralization of dye with 10C800 than other samples. Further, the reusability of the nanocomposite photocatalyst is a significant parameter  which maneuvers the practical usefulness of photocatalyst. To observe the reusability and stability of as-prepared C coated NbC NPs, recycle reactions were carried out for the degradation of all the dyes over 10C800 as shown in Fig. 14b. To observe the recyclability of as prepared photocatalyst, previously used photo catalysts (10C800) were removed from the solution by centrifugation and then reused for the photodegradation of all the dyes with the same concentration of the solution and irradiation intensity. The excellent stability of photocatalyst (10C800) is well-illustrated from the XRD pattern of 10C800 after the 3 rd cycle of photodegradation reaction as shown in Fig. S10. These results suggest sufficient stability of the photocatalyst during photodegradation of the dyes and can be reused without significant decrease in performance.
To establish the mechanism responsible for such a good photocatalytic behavior of synthesized photocatalyst, the generation of reactive oxygen species (ROS) has to be confirmed by using e − , h + , OH · and ⋅− O 2 scavengers. As an effect of these scavengers, degradation of dyes got retarded differently, which is shown in Fig. 15. In case of RhB dye, the addition of AO (hole scavenger) in dye solution retarded the degradation rate by 38% as compared to other scavengers and degradation of MB dye was reduced by 15% with the addition of SS, IPA, and AO (Fig. 15a,b,  respectively). Furthermore, addition of superoxide free radical scavenger i.e. AA does not affect the degradation of MB dye. On the other hand, degradation of MO dye was affected significantly in the presence of scavengers among which IPA reduced the efficiency by 33% which is higher than other scavengers (Fig. 15c). The observed retardation in photocatalytic degradation of different dyes can be correlated with the generation of the above-mentioned ROS. Figure 16 illustrates the proposed mechanism responsible for the photocatalytic degradation of dyes under visible light irradiation. As per the XRD results, it cannot be stated that nanocomposite sample consists of NbC alone, but, presence of O-centers (observed form TEM and XPS analysis) inside the NPs induced the optical active sites generating the charge carriers upon visible exposure. Earlier, Ohgi et al. 75 and Ishihara et al. 76 also suggested the enhanced generation and transportation of charge carriers by inducing oxygen centers (through partial oxidation) in the carbide nanoparticles. Figure S11 shows the valence band XPS spectra of 10C800 and 5C800 depicting different VB edges of ~1.97 and ~1.71 eV and with the help of optical absorption analysis, the position of the CB can be estimated. The CB edge potential of both the samples (−0.15 eV and −0.38 vs NHE, respectively) are more negative than the standard redox potential of OH / OH (+1.99 eV vs NHE) is lower than VB of photocatalyst 77 . Being metallic in nature, NbC supported the charge transfer towards the surface of NPs where the reaction with the hydroxyl anion and dissolved oxygen in water produces hydroxyl free radicals ( ⋅ OH ) and superoxide free radical anions ( .− O 2 ), respectively. These reagents are considered as strong oxidizing agents and thus oxidize the dyes adsorbed on the surface of the NPs. Various reactions, which are responsible for the degradation of dyes and are expressed as follows:  (6) For all the samples, degradation efficiency is a function of phase content (NbC, NbC x O y , and Nb-O), π→π* transitions and oxygen vacancies present on the surface. Among all of these parameters, the presence of NbC x O y and π→π* governs whole catalytic reaction (as discussed in PL analysis). Moreover, oxygen vacancies promotes the delayed recombination (PL emission peak at 468 nm) which in turn provide more excitons 78 . But, excess vacancies can also quench the photocatalytic reaction which might be responsible for the observed reduced photodegradation in 7C800 as compared to 10C800. While, 11C800 possess insufficient vacancies providing lesser degradation efficiency than 10C800.

Degradation mechanism.
Except for 10C800, both 7C800 and 11C800 contain higher carbonyl (C=O) group, considered as hole scavengers 79,80 , which in turn reduce the generation of OH · . Since, OH · is a strong oxidizing agent as compared to ⋅− O 2 , reduction in the generation of OH · retards the photocatalytic degradation efficiency of 7C800 and 11C800. Furthermore, in 7C800 and 11C800, Nb exhibits 4+ (NbC x O y , NbC, NbO 2 ) and 5+ (Nb 2 O 5 ) oxidation states in which Nb 5+ tries to attain a more stable state, i.e. Nb 4+ by scavenging an electron to retard the photodegradation. Sample 10C800 consists of Nb in the 3+ (NbC x O y , NbC) and 4+ (NbC x O y , NbC, NbO 2 ) states in which Nb 3+ will try to achieve Nb 4+ by ejecting an electron to enhance the degradation efficiency which can be expressed by the following equations: 3 4 As the degradation can be achieved by the transformation of chromophore groups and the generation of energetic electrons in visible light, photocatalysis is thermodynamically and kinetically limited which might cause the partial degradation of RhB, MB and MO dyes even with the visible range band gap of photo catalysts based   81 . With the help of results of detection of ROS species, the generation of ROS species is a function of both photocatalyst and dye solution. In the case of MO dye, all the species are being generated among which hydroxyl radical is the most active oxidizing agent. But, degradation of MB dye occurred by the hole, electron, and hydroxyl radical equivalently where, holes dominated the degradation of RhB dye as compared to other ROS species. Thus, smaller crystallite size, high pore volume and specific surface area, smaller recombination, and fine tuning of bandgaps are considered to be prominent factors for the enhanced photocatalytic properties of single phase carbon coated NbC NPs. Further, to check the toxicity of the best degraded dye solutions where 10C800 photocatalyst was used, a toxicity test using IS: 6582-2001 method was done. In the adopted method, 100% survival of fish species was observed after 96 h indicating the non-toxicity of the degraded solution as shown in Fig. S12.

Summary
From a series of experiments, carbon coated NbC x NPs has been successfully synthesized by using Nb 2 O 5 , Mg, and charcoal at relatively low temperature (800 °C). However, the holding time at 800 °C for the synthesis has been optimized as 10 h to obtain carbon coated NbC x NPs in which x = 0.95 is obtained. The distorted lattice due to simultaneous transportation of carbon and oxygen has been supported by W-H analyses having the least distortion in 10C800. The as synthesized NPs are thermally stable up to 400 °C in air atmosphere (Fig. S4). The existence of multiple phases in XRD patterns at different temperatures and holding times conveys a multistep in situ reduction-carburization process for the formation of NbC. This results in agglomerated mesoporous NPs.
The XRD of 10C800 sample shows the formation of single phase NbC, however, XPS results show the multiple oxidation states of Nb (associated to NbC, NbC x O y , NbO 2 and Nb 2 O 5 ) along with the both sp 2 and sp 3 hybridized carbon, oxygen vacancies confirming the removal of oxygen from Nb 2 O 5 (reactant). Raman spectroscopy also confirmed the presence of disordered graphitic carbon as coatings of NPs. UV-visible spectroscopy affirmed the absorption of visible radiation corresponding to small bandgaps. The higher O-vacancies, less π → π* transition and lower NbC/NbC x O y content are responsible for the increased PL emission intensity in 5C800, 7C800, and 11C800 as compared to 10C800. Small crystallites, high SSA, and less PL emission intensity gave enhanced photocatalytic properties of optimized sample (10C800) that follow pseudo first order kinetics. Further, visible light photocatalysis is very complex phenomenon comprising of decolorization and degradation of dye solution with the help of photocatalyst. For visible light photocatalysis with synthesized single phase NbC (10C800), rate of decolorization is faster than that of degradation of dye solution in 120 min under visible irradiation, which is much better than that reported by Chen et al. 52 .

Experimental Methods
Synthesis. To synthesize carbon coated NbC nano-composite, Nb 2 O 5 (1.329 g), metallic Mg powder (2.0 g) and activated charcoal (1.5 g) were used as the niobium source, reducing agent, and carbon source respectively. The mixture of reactants was heated to different temperatures (600, 700 and 800 °C) at a heating rate of 5 °C/min for different holding times (5, 7, 10 and 11 h) in a specially designed stainless steel autoclave 50 . The autoclave was allowed to cool within the furnace to room temperature. Thereafter, the obtained black powder was leached in 50% diluted HCl (v/v) and then washed with distilled water. The washed powder sample was further dried at 120 °C. The obtained powder samples were assigned as 10C600, 10C700, 5C800, 7C800, 10C800, and 11C800 corresponding to their synthesis temperature and holding time as listed in Table 1.
Characterization. For the identification of phases, X ray diffraction (XRD) analysis of synthesized samples was performed on a PANalytical Xpert-Pro diffractometer with Cu-Kα radiation (λ = 1.5406 Å, Ni filter, step size = 0.1301°) over 20 ≤ θ ≤ 80° and X'pert High Score was used to match the diffraction peaks with standard ICDD cards. Surface chemical compositions of synthesized samples (synthesized at 800 °C) were analyzed by X-ray photoelectron spectroscopy (XPS) (PHI 5000 Versa Prob II, FEI Inc.) using Al-Kα radiation (1486.7 eV) and C1s (284.5 eV) for the calibration of binding energies of all the elements. All the high resolution XPS spectra were analyzed with the help of XPSPEAK41 software by considering the equal FWHM of doublet peaks of Nb3d keeping area ratio of 3:2 (d 5/2 :d 3/2 ) with doublet separation of 2.74 eV. Morphologies of synthesized powder samples were observed with transmission electron microscopy (TEM) (JEOL 2100 F, operating at 200 kV). Line profile of nanoparticle was also taken using Cu grid in STEM mode to study the linear distribution of elemental concentration. Thermal analysis (TG) of the samples were carried out with a NETZSCH STA 449F3 at a heating rate of 5 °C/min up to 850 °C in air atmosphere to predict the thermal stability as a function of temperature. The surface area and pore size distribution calculations were done with the help of N 2 adsorption-desorption analyses (Tristar 3000). Optical absorption spectra of the samples were recorded on a double beam UV-Visible spectrophotometer (Hitachi U-3900H) in the range of 350-700 nm. Raman spectroscopic signals were recorded with a Renishaw Raman Microscope at 50X magnification with 785 nm (wavelength) laser source over a range of 100-1800 cm −1 .

Photo-catalysis.
For photocatalytic studies, a concentration of 1 mg/L of dye (RhB, MO and MB) was used.
The photocatalyst (0.002 g) was suspended in the 100 mL aliquot after homogenizing the dye solution. For the establishment of adsorption-desorption equilibria, the suspended solution was stirred for 60 min in a dark chamber. Thereafter, the solution was exposed to visible radiation of around 8900 lux intensity for 120 min in which the sample aliquots were extracted at an interval of 30 min. The photo reactor was open to air atmosphere to avoid the depletion of dissolved oxygen during the irradiation. Each sample was centrifuged at 4500 rpm to extract photocatalyst from the suspension and supernatant was characterized with a double beam UV-Visible spectrophotometer (Hitachi U-3900H) to observe the absorption spectra of dye solution at regular time interval. To confirm the variation of absorbance of dye solution due to decolorization or degradation, the filtered aliquots (WHATMAN 0.45 µm, PTFE) after 120 min irradiation were characterized by mass spectroscopy (MS) and total organic carbon (TOC). MS was carried out on UPLC-XEVO-G2-XS/QTOF mass spectrometer equipped with electron ionization source operated in positive ion mode. The mass spectrometer ion source and desolvation temperatures were set to 120 °C and 350 °C, respectively. Further, TOC was measured by MULTI N/C 3100 (N3-800/O) with the gas flow rate 160 mL/min and furnace temperature 800 °C. Further, the generation of reactive oxidative species (ROS) in the photodegradation of dyes were observed with in situ trapping experiments. The detection process was similar to the photodegradation experiments in which four different scavengers, such as, isopropanol (IPA, hydroxyl free radical scavenger, 1 mM), sodium sulphate (SS, electron scavenger, 1 mM), ammonium oxalate (AO, hole scavenger, 1 mM) and ascorbic acid (AA, superoxide radical scavenger, 1 mM) were added to dye solution in a 1:10 volume ratio prior to irradiation in separate systems.