Article pubs.acs.org/JPCC
Cite This: J. Phys. Chem. C 2018, 122, 1477−1485
Photocatalytic Hydrogen Evolution by Flexible Graphene Composites Decorated with Ni(OH)2 Nanoparticles J. Oliva,†,⊥ C. Gomez-solis,‡,⊥ L. A. Diaz-Torres,*,‡ A. Martinez-Luevanos,† A. I. Martinez,§ and E. Coutino-Gonzalez∥ †
CONACYT, Facultad de Ciencias Quimicas, Universidad Autónoma de Coahuila, 25280 Saltillo, Coahuila, Mexico Laboratorio de Fotocatalisis y Fotosintesis Artificial (F&FA), GEMANA, Centro de Investigaciones en Optica, A.P. 1-948, León, Guanajuato 37160, Mexico § Cinvestav Unidad Saltillo, Parque Industrial, Ramos Arizpe, Coahuila 25900, Mexico ∥ CONACYT, Centro de Investigación y Desarrollo Tecnológico en Electroquímica, Parque Industrial Querétaro, Sanfandila s/n, Pedro Escobedo, Querétaro 76703, Mexico ‡
ABSTRACT: This work presents the hydrogen evolution produced by flexible graphene composites (FGCs) fabricated by a casting method. Ni(OH)2 nanoparticles were also grown on the FGCs by using a wet chemical method. Those nanoparticles present spherical shapes and are uniformly distributed on the surface of the FGCs. The hydrogen generation activity in water for the FGCs with and without Ni(OH)2 nanoparticles was produced by UV light excitation. The FGCs decorated with Ni(OH)2 nanoparticles had a hydrogen generation rate 2.66 times higher than the FGCs without nanoparticles. It was also observed that the surface of the FGCs is oxidized during the photocatalytic process (formation of graphene oxide); this in turn helped to create actives sites for the generation of the electron−hole pairs during the irradiation under UV light and to transfer of charge from the FGCs’ surface to the electron trapping centers (Ni(OH)2 nanoparticles). Further, reuse experiments of the FGCs demonstrated that their stability for the hydrogen generation improved due to the presence of Ni(OH)2 nanoparticles, since the hydrogen generation rate after three cycles of use decreased by 46% and by 84% in the FGCs with and without Ni(OH)2 nanoparticles, respectively. The flexibility of the graphene composites facilitated their introduction and removal from the water container where the photocatalytic generation of H2 occurred. Hence, our results suggest that the FGCs could be a feasible option for water splitting in industrial reactors.
1. INTRODUCTION
composites have been used to generate hydrogen in the range of 2−900 μmol/h by photocatalysis under solar or UV irradiation.4−9 Particularly, the nanocomposites formed by decorating CdS nanorods or TiO2 nanograins with Ni(OH)2 nanoparticles are able to produce high hydrogen generation rates by photocatalysis in the range of 5085−7280 μmol/h· g.10,11 Although these systems are “efficient” for hydrogen generation, they employ expensive and rare materials (Pd, La, Rh, W, and Ta), which increases the H2 production costs. In addition, the materials employed for the fabrication of these nanocomposites such as CdS and TiO2 nanoparticles could be toxic for the environment and living organisms if they have a nanosize.12,13 As an alternative to produce less toxic materials with enhanced hydrogen generation, it has been reported nanocomposite powders made with carbon materials such as
Scientists in the field of environmental science are currently focusing their efforts to find a clean and ecofriendly alternative for energy generation, since the fossil fuels are seriously contaminating the air in our environment and promote the green house effect on the earth, which affects drastically the world weather conditions.1 Hydrogen (H2) is an alternative source of energy that can be produced from renewable sources of sunlight and water; then, research about how to produce it efficiently became important during the past 10 years.1,2 In fact, hydrogen as a source of energy would be able to satisfy all the energetic needs which are increasingly growing worldwide.3 However, the current H2 generation methods are not economical due to the low process efficiencies and therefore require further advances to be commercially viable.1 Several photocatalysts powders such as NaTaO3−NiO, pervoskite H1.9K0.3La0.5Bi0.1Ta2O7 loaded with Pd or Rh, Ni(OH)2/TiO2 nanocomposite photocatalyst, nickel/CdS photocatalyst, WO3/ g-C3N4/Ni(OH)x, and Ni(OH)/tantalum oxinitride (TaON) © 2018 American Chemical Society
Received: October 19, 2017 Revised: December 30, 2017 Published: January 2, 2018 1477
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C
The use of films for hydrogen generation is interesting because they can be adapted to the shape of the container or reactor where the hydrogen production is achieved. There are no studies about flexible graphene composites for H 2 generation by photocatalysis to the best of our knowledge. Therefore, this work reports the use of flexible graphene composites (FGCs) with and without Ni(OH)2 nanoparticles (NPs) on their surface for water splitting. Those last NPs were used because they have enhanced the hydrogen generation according to literature.5,6,27 We found that the hydrogen generation rate was enhanced 2.66 times in the FGCs with Ni(OH)2 nanoparticles.
graphene (G), graphene oxide (GO), reduced graphene oxide (rGO), or carbon nanotubes (CNT): (1) Agegnehu et al. reported graphene oxide nanosheets decorated with Ni and NiO nanoparticles and demonstrated hydrogen production as high as 70 μmol/h in methanol solution. Nevertheless, those authors could not observe the hydrogen generation in water under UV light during the first 8 h;14 (2) Dai et al. published the synthesis of Pt@MWCNT-TiO2 nanocomposite by a hydrothermal process and obtained a H2 production of 300− 350 μmol/h (under UV−vis−NIR irradiation), but the problem of this system was the use of the expensive Pt element;15 (3) Cheng et al. studied 3D graphene flowers decorated with TiO2 and Au nanoparticles for hydrogen generation by photocatalysis and showed that Au helped to improve the photoelectrocatalytic efficiency by plasmonic effects, but the main inconvenience for this nanocomposite is the cost of Au;16 (4) Peng. et al. demonstrated that GO-CdS nanocomposites produced 30% more H2 with respect to bare CdS under visible light irradiation, reaching a maximum H2 rate of 314 μmol/h.17 Further, other groups reported composites such as CdS/MoS2/ G and CdS QDs−ZnIn2S4/RGO with hydrogen generation rates in the range of 1800−2700 μmol/h.18 However, the main problem of these systems with CdS and graphene is the use of toxic CdS, since there is a possibility for the release of Cd ions in the water, causing contamination. Moreover, other authors have studied G-TiO2, G-CdS, or BiVO4-G nanocomposite powders but found low H2 generation rates in the range of 0.2− 16 μmol/h.19−21 The enhanced photocatalytic activity observed in the previous systems for hydrogen generation was due to the fast transfer of electrons generated from graphene, GO, or CNT toward the catalysts, NiO, Ni nanoparticles, and TiO2 or from CdS to GO or G. Those last materials also worked as electron traps for proton reduction or served as hole traps for water oxidation; this in turn suppressed the electron−hole recombination and enhanced the H2 production.14,15 Although these nanocomposite materials with carbon compounds have improved the hydrogen generation with respect to those without carbon, all of them are powders, and their removal from the solvent (alcohol or water) is difficult because they remain dispersed as colloids in the solvent after the H2 production. Therefore, some catalytic materials have been fabricated as flexible films or sponges to facilitate their removal after the water splitting: (1) Miao et al. published an easy method to deposit Ni−Mo−S sheets on carbon fiber textiles and produced hydrogen by electrocatalysis in neutral electrolyte,22 but they employed a biomolecule L-cysteine which could be degraded during the hydrogen evolution; (2) Rao et al. fabricated a Ni(OH)2/nickel foam with good electrocatalytic activity and durability for hydrogen evolution reaction (HER), but those authors needed relatively high overpotentials (0.49 V) for the catalytic reaction;23 (3) Sankir et al. reported the use of a flexible film composed of platinum and Teflon and produced hydrogen with a high rate of 0.695 mol/h in the presence of HCl,24 but the main problem of this film is the cost of Pt metal and Teflon. Furthermore, other groups have reported the use of solid electrodes formed by the deposition of WO3 and NiSe2 films on FTO and Au-glass solid substrates for H2 production25,26 by photoelectrochemical or electrocatalytical methods. Thus, the hydrogen production has been studied by photocatalysis with nanocomposites powders or by photoelectrochemical or by electrocatalytic methods employing flexible or solid films according to the previous reports.
2. EXPERIMENTAL SECTION 2.1. Fabrication of Flexible Graphene Composites (FGCs). The FGCs were fabricated using a casting method. Basically, a blend of graphene nanoplatelets, acetone, water, and ethanol with a weight ratio of 0.04:20:20:20 was prepared and subjected to vigorous stirring during 30 min. This graphene solution was dropped on a mold. After this, a second solution made of an acrylic monomer (AM), ethanol (ETHO), acetone (AC), water (WT), and a photoinitiator (PI) with a weight ratio of AM:ETHO:AC:WT:PI = 1:1:1:1:0.4 was dropped on the mold. Subsequently, the whole mixture was cured with UV light at 254 nm during 3 min to promote the polymerization process (formation of the acrylic polymer). Finally, the flexible graphene composite was removed from the mold. The dimensions of the FGCs characterized in this work were 1.4 cm × 1.4 cm × 0.1 cm (length × width × thickness). 2.2. Growth of Ni(OH)2 Nanoparticles on FGCs. All the chemicals were purchased from Sigma-Aldrich and used without further treatment. The Ni(OH)2 nanoparticles were grown on the FGCs using a following process: an amount of 5 mmol of Ni(NO3)2 was dissolved in a mixture of ethylene glycol and N,N-dimethylformamide with a weight ratio of 0.5:1. After this, the FGCs were immersed into the solution and stirred under UV-light irradiation of 365 nm during 20 min. Subsequently, 1 mmol of NaBH4 was added to the mixture and stirred for 5 h. Finally, the FGCs were washed with deionized water and ethanol several times and dried at 150 °C during 12 h.28,29 For discussion purposes, the FGCs with Ni(OH)2 nanoparticles on their surface will be named FGCs-Ni(OH)2. 2.3. Morphological, Structural, and Optical Characterization. Field emission scanning microscopy (FESEM) images were also obtained from a JEOL JSM-7800F microscope, and electron dispersion spectroscopy (EDS) analysis was done with an Oxford Instruments detector. The structural characterization of our samples was done by using X-ray diffraction (XRD) in a Bruker D8 equipment with radiation Kα1 of copper (λ= 1.540 56 Å) at an incidence angle of 2° in the range of 8−80° in 2θ. Raman measurements in the range of 100−1800 cm−1 were carried out using a Witec CRC200 at a wavelength of 514.5 nm. The power used for all measurements was 2.4 mW, and the spot area is around 189 μm2. A 50× objective lens and a total time integration of 30 s were used. The Fourier transform infrared (FTIR) spectra of samples were recorded in the range of 2700−3100 cm−1 on an ABA (MB300) spectrophotometer using the KBr method. 2.4. Mechanical and Electrical Conductivity Characterization. The samples were prepared for the tensile tests according to the ASTM D-FGC638 standard method. The tensile properties of the FGCs were measured by using a homemade universal testing machine controlled with a Labview 1478
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C software with a cross-head speed of 7 × 10−4 mm/s. A GW Instek (GOM-802) four probe test unit is used to measure the sheet resistance of the graphene composites with dimensions of 2 cm × 2 cm × 0.1 cm. The probes were first calibrated using an indium tin oxide (ITO) substrate before measurements, and five readings were taken in 10 different FGCs to obtain an average value of sheet resistance. 2.5. XPS and BET Measurements. X-ray photoelectron spectroscopy (XPS) spectra were obtained by using a Thermo Scientific K-Alpha XPS instrument with monochromatic Al Kα radiation of 1486.68 eV. The specific surface area of the FGCs was measured by physisorption of N2 through the BET method using a Quantachrome NOVA 2000e equipment. 2.6. Photocatalytic Hydrogen Generation. The photocatalytic hydrogen generation was carried out by using a quartz reactor (total volume 250 mL) coupled to gas chromatography equipment. The procedure for the hydrogen generation was as follows: 0.68 g of the FGCs-Ni(OH)2 catalyst was dispersed into the quartz reactor with 200 mL of water. Next, the solution was deoxygenated by bubbling N2 gas, and the photocatalytic reaction started by turning on a UV lamp (λexc= 254 nm). The progress of the photocatalytic reaction was followed by a gas chromatograph, which has a column with a TCD detector (Thermo Scientific 3GC Ultra). The amount of gases generated (hydrogen and oxygen) was registered every 30 min during the 3 h reaction.
3. RESULTS AND DISCUSSION 3.1. Structural and Morphological Properties of Flexible Composites. The fabrication process of the FGCs by a casting method produced overlapped graphene nanoplatelets (GNPs) in the in-plane direction; see the surface (top view) of the flexible graphene composite in Figure 1a. Those FGCs also have pores on their surface with sizes in the range of 15−40 μm; see red circles. It is also observed GNPs with edges partially folded and without polymer on their surface, which means that all the polymer is located in the bulk of the FGCs and provides high mechanical properties; see Figure 1b. Further, the surface of FGCs after the deposition of Ni(OH)2 (sample of FGCs-Ni(OH)2) shows coalescence of the graphene nanoplates; this was a consequence of the annealing treatment at 150 °C to induce the growth on Ni(OH)2 nanoparticles; see Figure 1c. Those nanoparticles had spherical shape with an average size of 55 nm as corroborated by a higher magnification image in Figure 1d. The Ni(OH) 2 nanoparticles are homogeneously distributed on the surface of FGCs, and their separation distance was beneficial for the photocatalytic hydrogen generation because it favors the charge migration from the surface of graphene toward the Ni(OH)2, which in turn promotes the proton reduction reactions.30−32 However, there were some small areas in the FGCs-Ni(OH)2 composites where conglomerations of Ni(OH)2 nanoparticles were observed; see Figure 1e. If the nanoparticles decorating the surface of the FGCs are agglomerated, the photocatalytic activity diminishes because the physical contact among nanoparticles favors the electron−hole recombination, decreasing their availability for the water splitting reaction.30−32 The presence of C, O, and Ni elements in the FGCs was confirmed by the EDS spectra in Figure 1f. The weight percentage composition of the FGCs-Ni(OH)2 composites was 84%, 14.4%, and 1.6% for C, O, and Ni, respectively. The low content of Ni (only 1.6%) was high enough to enhance the hydrogen evolution by approximately 2.66 times as explained
Figure 1. (a) and (b) are SEM images of FGCs with low and high magnification, and (c) and (d) are SEM images of FGCs-Ni(OH)2 composites with low and high magnification. (e) shows FGCs with conglomerations of Ni(OH)2 nanoparticles on their surface, while (f) shows the EDS spectra obtained from the surface of FGCs-Ni(OH)2 composites.
later. Inset in Figure 1c illustrates a picture of a FGC with Ni(OH)2 nanoparticles with an elastic modulus of 34.5 ± 0.02 MPa (this value allows us to obtain FGCs with flexibility similar to that of a piece of paper). Moreover, the EDS mapping of the FGCs-Ni(OH)2 composites (see Figure 2a and Figure 2b) reveals the presence of carbon and oxygen at the same time, which indicates a possible formation of graphene oxide. The presence of Ni element was also detected by the EDS mapping; see Figure 2c. This was due to the growth of Ni(OH)2 nanoparticles on the FGCs as corroborated later by the XPS analysis. In addition, Figure 2c shows that Ni was homogeneously distributed on the surface of the FGCs even though there are some small areas with high concentration of it; see the bottom part of Figure 2c. XRD and Raman measurements were useful to demonstrate the structural quality of the graphene composites. The XRD pattern of the pure GNPs powder in Figure 3a shows diffraction peaks at 26.63°, 44.9°, and 54.82° that correspond to the planes (002), (100), and (004) and are in good agreement with the well-known diffraction peaks for graphene according to previous reports.33,34 Figure 3a also shows the diffraction peaks that correspond to the FGCs. An extra broad band is observed, which indicates the presence of the amorphous polymer in the composite. Moreover, the diffraction peak corresponding to the planes (100) disappeared while the ones for the planes (002) and (004) were displaced ∼0.3° toward lower angles. Such displacement confirms the formation 1479
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C
Figure 2. EDS elemental mapping of FGCs-Ni(OH)2 composites.
of the graphene composite. Figure 3b shows the XRD pattern of the FGC-Ni(OH)2 composite, where a broad band centered at 40° is observed. This band corresponds to the crystalline plane (101), which is associated with the hexagonal phase of βNi(OH)2 according to the JCPDS no. 14-0117.35,36 In addition, the diffraction peak at 26.63°, corresponding to graphene, is hindered due to the presence of the β-Ni(OH)2 broad band, probably due to the diffusion of the acrylic polymer from the bulk toward the upper surface of the FGCs. This was caused by the expansion of the graphene−polymer network during the annealing treatment at 150 °C; in consequence, the detection of the diffraction peaks corresponding to graphene was not possible. The Raman spectra in Figure 3c show the typical D, G, and 2D graphene bands for the graphene nanoplatelets (GNPs) powder and the flexible graphene composites (FGCs). In the case of the Raman spectra for FGCs, a broad band with overlapped peaks was observed due to the presence of the acrylic polymer while only a very intense peak is observed for the GNPs powder. The D, G, and 2D bands are located at 1358 cm−1, 1583 cm−1, and 2866 cm−1, respectively. The D-to-G peak intensity ratios (ID/IG) were calculated for GNPs powder and for FGCs in order to know the degree of defects’ formation at each stage of fabrication. The values of ID/IG were 0.045 and 0.24 for the GNPs powder and FGCs, respectively. Since these intensity ratios are below 0.3, the GNPs and the FGCs should have a very low density of defects.37 However, the value of ID/ IG was higher for FGCs; this means that the presence of the acrylic polymer increased slightly the amount of defects in the structure of GNPs incorporated in the FGCs. The peak intensity ratios IG/I2D were also calculated in order to elucidate whether we have a bilayer or multilayer GNPs. The IG/I2D values were 3.75 and 2.62 for GNPs powder and FGCs,
Figure 3. XRD patterns of (a) GNPs and FGCs and (b) FGCsNi(OH)2. (c) is the Raman spectra for GNPs and FGCs.
respectively. The IG/I2D ratios of >2 indicate the presence of multilayer graphene.37 The reduction from 3.75 to 2.62 indicates a detaching of graphene layers after the mixture of GNPs with the acrylic polymer. The surface of the GNPs is interacting with the acrylic polymer, and an excess of this last compound could be detrimental for the electrical conductivities of the FGCs. Therefore, an appropriate balance between these components is required to maximize the conductive properties on the surface of the FGCs. FTIR measurements could help to determine at least qualitatively the presence of the acrylic polymer. Figure 4a shows the FTIR spectra of the pristine graphene powder before its dispersion in the water/alcohol/acetone solution (see red curve). There are some weak peaks at 616, 1083, and 1424 cm−1 that correspond to graphene oxide (GO). Those peaks are related to the bonds of C−O−C, C−O−H, and C−O.38,39 The band centered at 3440 cm−1 indicates the presence of water adsorbed on the GNPs surface. The peak centered at 1623 cm−1 corresponds to the skeletal vibrations of graphene (CC).40,41 The presence of the acrylic polymer into the FGCs was detected due the appearance of new vibrational peaks at 795, 1018, 1108, 1261, 1529, 1694, 2849, 2922, and 3309 cm−1; see black curve in Figure 4a. Those peaks are attributed to the C−O−C, C−H, C−CH3, and CH2 stretching of the acrylic polymer.42−45 In fact, the vibrational bands related to OHs (3440 cm−1), graphene (1623 cm−1), and GO (616 cm−1) were hindered due to the presence of the acrylic 1480
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C
Figure 5. (a) and (b) show the XPS spectra for C 1s and O 1s orbitals in FGCs, respectively. (c) and (d) show the XPS spectra for Ni 2p3/2 and O 1s orbitals in FGCs-Ni(OH)2 composites.
O), and C−C bonds, respectively.48 Moreover, sp2 and sp3 hybridization of C−C and C−H bonds were observed at 284.6 and 284.1 eV, respectively.49 The sp2 bonds are dominant in the carbon layer, and the sp3 bonds correspond mainly to the acrylic polymer. The O 1s spectrum of the FGCs in Figure 5b shows a dominant band centered at 532.5 eV, which corresponds to an aliphatic C−O bond.50 An aromatic CO bond is observed as well. Figure 5c shows the Ni 2p spectra of the FGCs-Ni(OH)2 composites, and the deconvolution band centered at 861.1 eV corresponds to the satellite peak while the band centered at 586 eV is attributed to the Ni(OH)2 compound.51 This confirms the nature of the nickel hydroxide grown on our samples. In fact, the Figure 5d shows that the O 1s spectrum of the FGCs changed due to the presence of Ni(OH)2; that is, a shift toward lower energies occurred due to the appearance of an additional band (see pink curve), which is associated with Ni(OH)2.52 Figure 6a depicts the results of hydrogen generation of the FGCs with and without Ni(OH)2 nanoparticles. Both the FGCs and the FGCs-Ni(OH)2 composites generated hydrogen, with the hydrogen production rate for the FGCs-Ni(OH)2 composites being around 2.66 times higher than that for the bare FGCs. The presence of the Ni(OH)2 nanoparticles enhanced the hydrogen generation rate from 90 to 330 μmol· g−1 (see the values of hydrogen generation after 3 h of photocatalytic activity), that is, from 30 to 110 μmol h−1 g−1. This maximum hydrogen generation rate produced by FGCsNi(OH)2 composites is ∼56% lower than that of 250 μmol h−1 g−1 reported in other based graphene composites such as rGON749-Pt (which employs expensive materials such as ruthenium dye and platinum).53 Compared with other hydrogen generation rates produced by nanocomposite powders such as TiO2 or CdS decorated with Ni(OH)2 nanoparticles, the hydrogen generation rate of the FGCsNi(OH)2 composites is 46−65 times lower.10,11 However, one of the advantages of our work is the fact that our material is flexible and thus can be adapted to the shape of the photoreactors. This could increase the performance of the H2 production by increasing the available contact area between the water and the photocatalytic FGCs. Another advantage is the fact that our catalytic material can be easily removed from water
Figure 4. FTIR spectra of: (a) GNPs and FGCs, (b) FGCs-Ni(OH)2 composites, and (c) FGCs after their use for H2 generation.
polymer. The presence of the Ni(OH)2 on FGCs (sample FGC-Ni(OH)2) was also suggested by the FTIR spectra; see the peaks at 489 and 669 cm−1 associated with Ni−O and Ni− O−H, respectively, in Figure 4b.46 It is worthy to mention that the presence of the O2 during the UV-light irradiation could provoke the oxidation of the composite for the formation of graphene oxide (GO), which in turn would facilitate the transfer of charge toward the Ni(OH)2 nanoparticles (electron trapping centers), and this will increase the hydrogen generation. The formation of graphene oxide (GO) on the surface of FGCs was confirmed by the FTIR spectra of the FGCs without Ni(OH)2 after their use for hydrogen generation, since some peaks at 3410, 1721, 1459, and 1081 cm−1 attributed to OH, CO, C−OH, and C−O vibrations, respectively, were observed;47 see Figure 4c. In fact, the vibrational peaks of GO in Figure 4c are very notorious in comparison with the peaks in the red curve of Figure 4a, suggesting that the GO was completely formed after the irradiation with UV light during the photocatalytic generation of H2. 3.2. Hydrogen Generation by FGCs and FGCs-Ni(OH)2. The XPS characterization technique was used to determine the chemical states of elements such as carbon (C 1s) and oxygen (O 1s) as well as to confirm the nickel bonding nature. The C 1s spectrum of the FGCs in Figure 5a shows the deconvolution bands centered at 287, 286.1, and 285.4 eV, which correspond to the carboxyl (O−CO), carbonyl (CO), epoxide (C− 1481
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C
leading to the formation of active sites from which the electron−hole pairs are generated after excitation with UV light. Those charges could move on the surface of the FGCs to react with H2O molecules. The sheet resistance increased from 8.5 ± 0.1 Ω/□ to 17.5 ± 0.1 Ω/□ for the FGCs and from 11.3 ± 0.1 Ω/□ to 24.7 ± 0.1 Ω/□ for the FGCs-Ni(OH)2 composites after their use for H2 production. This was due to the oxidation of the FGCs (formation of GO on their surface) during the photocatalytic generation of hydrogen as corroborated by FTIR spectra in Figure 4c. Although the electrical conductivity was lower in the oxidized FGCs-Ni(OH)2 composites due the increase of their sheet resistance, the presence of Ni(OH)2 nanoparticles enhanced the H2 production in comparison with the FGCs without nanoparticles because the Ni(OH) 2 nanoparticles acted as effective electron trapping centers.6 Furthermore, the high surface area of the oxidized FGCs (18.6 m2/g) and the high porosity of the graphene network in the bulk of the FGCs both enhance the photocatalytic generation of hydrogen due to the increase of contact area between the water molecules and the FGCs. Figure 6b shows a cross-section image of the FGCs-Ni(OH)2 composites where rounded pores and grooves are observed. The inset of Figure 6b indicates that the Ni(OH)2 nanoparticles were not found in the porosity of the FGCs-Ni(OH)2 composites; this means that all the Ni(OH)2 nanoparticles that improved the hydrogen evolution remain on the surface of the FGCs. The size distribution of the pores in the bulk of FGCs is depicted in Figure 6c. It is clear that most of the pores (84%) have sizes in the range of 15−30 μm. Also, the absence of organic binding agents to attach the Ni(OH)2 nanoparticles to the FGCs favors a fast transfer of electrons from GNPs to Ni(OH)2, which in turn promotes the hydrogen generation. The presence of the Ni(OH)2 catalyst helped to increase the absorption of UV light and acted as trapping centers for electrons, which enhanced the hydrogen generation by avoiding the recombination of the photogenerated electron−hole pairs.7 The energy level diagram in Figure 7a indicates that the oxidized FGCs (OFGCs) have a conduction band (CB) of around 0.52 eV (vs NHE), the valence band (VB) should be close to −1.48 eV (NHE), and the level for Ni(OH)2 is around 0.23 eV according to literature.14,35,58 Since the CB of OFGCs is located in a higher positive potential than that for Ni(OH)2 nanoparticles, the electrons photogenerated under UV excitation from the VB in OFCGs can jump toward the CB of FGCs and be easily transferred to the CB of Ni(OH)2 nanoparticles. The empty spaces left by electrons in the VB of the OFGCs are generated holes; this in turn produces a water oxidation reaction for the formation of H+ (see eq 1), while the electrons captured quickly by the Ni(OH)2 nanoparticles promote the H+ reduction reaction on the surface of the cocatalyst Ni(OH)2 (see eq 2):
Figure 6. (a) Hydrogen generation rates for FGCs and FGCsNi(OH) 2. (b) shows a cross-section image of FGCs-Ni(OH)2 composites obtained by SEM, and (c) shows the pore size distribution of the pores in the bulk of FGCs-Ni(OH)2 composites.
manually, while the catalyst powders need a process of centrifugation to produce their precipitation before taking them out from water. Additionally, some catalysts for hydrogen generation such as TiO2/Ni(OH)2 and NiO/GO have a nanosize and can remain dispersed in the water, which makes their removal difficult.8−14 If we consider that the weight of the FGCs-Ni(OH)2 composite was 0.68 g, a hydrogen generation rate of ∼75 μmol/h is obtained, which is higher than these reported in other graphene based systems such as G-TiO2, GCdS, or BiVO4-G (0.2−16 μmol/h).19−21 According to literature, the graphene nanoplates and graphene oxide photocatalysts are not able to produce oxygen and/or hydrogen under photocatalytic reaction using water because their oxidation overpotential is not high enough to oxidize water for O2 generation. Further, their reaction kinetics cannot favor the generation of hydrogen (graphene or GO can only transport electron or holes, but they cannot generate the electron−hole pairs necessary for the oxidizing−reducing reaction).20,54−56 However, the presence of impurities, the use of a cocatalyst, or the generation of defects in the structure of graphene or graphene oxide can improve their photocatalytic activity for hydrogen generation.6,54,57 In our case, we used an acrylic polymer to make the FGCs, which were oxidized during the photocatalytic activity, and this composite was able to generate hydrogen by the following process: There was an overlapping of the π orbitals from the polymer, GO, and the carboxyl functional groups formed on the surface of FGCs,
4h+ + 2H 2O → O2 + 4H+
(1)
2e− + 4H+ → 4H 2
(2)
Figure 7b shows how the electrons are captured by the Ni(OH)2 nanoparticles, which are homogeneously distributed on the surface of OFGCs, leaving the holes available to perform the oxidation of water molecules. In order to analyze the stability of the FGCs and FGCsNi(OH)2 for the hydrogen generation, reuse experiments were carried out using the FGCs with and without Ni(OH)2 nanoparticles. Figure 8 shows the results of hydrogen 1482
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C
protecting passivation layer which avoided the surface degradation of the composite during the generation of O2.
4. CONCLUSIONS The hydrogen generation activity produced by UV excitation was enhanced 2.66 times in FGCs with Ni(OH)2 nanoparticles decorating their surface with respect to the bare FGCs. Those nanoparticles acted as electron trapping centers, which in turn avoided the electron−hole recombination and promoted the hydrogen production. In fact, the separation and the homogeneous distribution of these nanoparticles were a key factor to improve the photocatalytic activity. Moreover, reuse experiments demonstrated that the reduction of the hydrogen generation rate was lower in the FGCs decorated with Ni(OH)2 nanoparticles (after three cycles). One of the main advantages of the FGCs is the fact that they maintain a low electrical resistance in spite of the deposition of Ni(OH)2 nanoparticles on their surface, and this helps to transfer the electrons rapidly to the trap centers. The property of flexibility in the graphene composites also provided two more advantages: (1) the active photocatalytic material can be removed easily from water and (2) the FGCs can be adapted to the shape of the container, which would allow increase of the contact area between the water and the photocatalytic FGCs. Due to the physical characteristics and performance for hydrogen production, the FGCs could be useful for water splitting in chemical reactors.
■
Figure 7. Schematic illustration of H2 production in FGCs and FGCsNi(OH)2.
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
J. Oliva: 0000-0001-7361-9304 L. A. Diaz-Torres: 0000-0002-1281-9916 E. Coutino-Gonzalez: 0000-0001-8296-0168 Author Contributions ⊥
J.O. and C.G.-s. contributed equally to this work.
Notes
The authors declare no competing financial interest.
■
Figure 8. Hydrogen generation rates obtained after 1−3 reuse cycles of the FGCs and FGCs-Ni(OH)2.
ACKNOWLEDGMENTS J.O. is thankful for the financial support of UC-MEXUSCONACYT postdoctoral research program. The authors also appreciate the technical work performed by C. Albor from Centro de Investigaciones en Ó ptica in León, México.
generation rates after three cycles of use (each cycle lasted 3 h). In the case of the FGCs, the hydrogen generation rate decreased by 74% and by 84%, after the second and third cycles of use, respectively. The decrease of the hydrogen generation in FGCs was probably due to the adsorption of oxygen molecules, generated during the reaction on the FGCs, which in turn produced C−O groups that physically degrade the surface of the flexible composite.7 In other words, the formation of covalent bonds by O2 damaged the bonds between π and π* orbitals in the acrylic polymer, and the bonds between the π orbitals from the polymer and the sp2 orbitals of GNPs in OFGCs. This modified the band gap of the OFGCs in such a way that the electron transfer from the CB to the VB is not allowed anymore. The FGCs-Ni(OH)2 composites presented better stability because the decrease of the hydrogen generation rate was only 27% and 46% after the second and third cycles of use, respectively. This improvement of stability was probably caused by the fact that Ni(OH)2 nanoparticles acted as a
■
REFERENCES
(1) Muhich, C. L.; Ehrhart, B. D.; Al-shankiti, I.; Ward, B. J.; Musgrave, C. B.; Weimer, A. W. A Review and Perspective of Efficient Hydrogen Generation Via Solar Thermal Water Splitting: A Review and Perspective of Efficient Hydrogen Generation. Adv. Rev. 2016, 5, 261−287. (2) Nguyen, S. C.; Vu, N.; Chabot, S.; Kaliaguine, S.; Do, T. Role of CxNy-Triazine in Photocatalysis for Efficient Hydrogen Generation and Organic Pollutant Degradation Under Solar Light Irradiation. Sol. RRL 2017, 1, 1700012. (3) N.R.C.N.A. of Engineering, The Hydrogen Economy: Opportunities, Costs, Barriers, and R&D Needs, 2004. (4) Zhang, Q.; Li, Q.; Wang, S.; Li, R.; Zhang, X.; Liang, Z.; Han, H.; Liao, S.; Li, C. Effect of Redox Cocatalysts Location on Photocatalytic Overall Water Splitting Over Cubic NaTaO3 Semiconductor Crystals Exposed with Equivalent Facets. ACS Catal. 2016, 6, 2182−2191. 1483
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C
(24) Sankir, M.; Semiz, L.; Serin, R. B.; Sankir, N. D. Hydrogen Generation from Nanoflower Platinum Films. Int. J. Hydrogen Energy 2015, 40, 8522−8529. (25) Swesi, A. T.; Masud, J.; Liyanage, W. P. R.; Umapathi, S.; Bohannan, E.; Medvedeva, J.; Nath, M. Textured NiSe2 Film: Bifunctional Electrocatalyst for Full Water Splitting at Remarkably Low Overpotential with High Energy Efficiency. Sci. Rep. 2017, 7, 2401−2412. (26) Mai, X. M.; Ma, X.; Zhou, H.; Ye, M.; Li, T.; Ke, S.; Lin, P.; Zeng, X. Effect of Oxygen Pressure on Pulsed Laser Deposited WO3 Thin Films for Photoelectrochemical Water Splitting. J. Alloys Compd. 2017, 722, 913−919. (27) Yu, J.; Wang, S.; Cheng, B.; Lin, Z.; Huang, F. Noble Metal-Free Ni(OH)2-gC3N4 Composite Photocatalysts with Enhanced VisibleLigth Photocatalytic H2-Production Activity. Catal. Sci. Technol. 2013, 3, 1782−1789. (28) Li, S.; Dai, D.; Ge, L.; Gao, Y.; Han, C.; Xiao, N. Synthesis of Layer-Like Ni(OH)2 Decorated ZnIn2S4 Sub-Microspheres with Enhanced Visible-Light Photocatalytic Hydrogen Production Activity. Dalt. Trans. 2017, 46, 10620−10629. (29) Hu, J.; Cao, Y.; Xie, J.; Jia, D. Simple Solid-State Synthesis and Improved Performance of Ni(OH)2-TiO2 Nanocomposites for Photocatalytic H2 Production. Ceram. Int. 2017, 43, 11109−11115. (30) Gómez-Solís, C.; Ballesteros, J. C.; Torres-Martínez, L. M.; Juárez-Ramirez, I. RuO2−NaTaO3 Heterostructure for its Application in Photoelectrochemical Water Splitting Under Simulated Sunlight Illumination. Fuel 2016, 166, 36−41. (31) Gómez-solís, C.; Ruiz-Gómez, M. A.; Torres-martínez, L. M.; Juárez-Ramírez, I.; Sánchez-Martínez, D. Facile Solvo-Combustion Synthesis of Crystalline NaTaO3 and its Photocatalytic Performance for Hydrogen Production. Fuel 2014, 130, 221−227. (32) Yang, J.; Wang, D.; Han, H.; Li, C. Roles of Cocatalysts in Photocatalysis and Photoelectrocatalysis. Acc. Chem. Res. 2013, 46, 1900−19009. (33) Rashad, M.; Pan, F.; Tang, A.; Lu, Y.; et al. Effect of Graphene Nanoplatelets (GNPs) Addition on Strength and Ductility of Magnesium-Titanium Alloys. J. Magnesium Alloys 2013, 1, 242−248. (34) Ara, K.; Wadumesthrige, K.; Meng, T.; Salley, S. O.; Ng, K. Y. S. Effect of Microstructure and Sn/C Ratio in SnO2−Graphene Nanocomposites for Lithium-Ion Battery Performance. RSC Adv. 2014, 4, 20540−20547. (35) Hao, J.; Wang, X.; Liu, F.; Han, S.; Lian, J.; Jiang, Q. Facile Synthesis ZnS/ZnO/Ni(OH)2 Composites Grown on Ni Foam: A Bifunctional Materials for Photocatalysts and Supercapacitors. Sci. Rep. 2017, 7, 1−12. (36) Malara, F.; Carallo, S.; Rotunno, E.; Lazzarini, L.; Piperopoulos, E.; Milone, C.; Naldoni, A. A Flexible Electrode Based on Al-Doped Nickel Hydroxide Wrapped around a Carbon Nanotube Forest for Efficient Oxygen Evolution. ACS Catal. 2017, 7, 4786−4795. (37) Shtein, M.; Nadiv, R.; Buzaglo; Kahil, K.; Regev, O. Thermally Conductive Graphene-Polymer Composites:Size, Percolation, and Synergy Effects. Chem. Mater. 2015, 27, 2100−2106. (38) Xian, T.; Yang, H.; Di, L.; Ma, J.; Zhang, H.; Dai, J. Photocatalytic Reduction Synthesis of SrTiO3-Graphene Nanocomposites and Their Enhanced Photocatalytic Activity. Nanoscale Res. Lett. 2014, 9, 327−332. (39) Hayyan, M.; Abo-Hamad, A.; Alsaadi, M. A.; Hashim, M. A. Functionalization of Graphene Using Deep Eutectic Solvents. Nanoscale Res. Lett. 2015, 10, 324−329. (40) Zhang, J.; Xie, X.; Li, C.; Wang, H.; Wang, L. The Role of Soft Colloidal Templates in the Shape Evolution in Flower-Like MgAlLDH Hierarchcal Microstructures. RSC Adv. 2015, 5, 29757−29765. (41) Jiang, N.; Xiu, Z.; Xie, Z.; Li, H.; Zhao, G.; Wang, W.; Wu, Y.; Hao, X. P. Reduced Graphene Oxide − CdS Nanocomposites with Enhanced Visible-Light Photoactivity, New. New J. Chem. 2014, 38, 4312−4320. (42) Szilasi, S. Z.; Huszank, R.; Szikra, D.; Váczi, T.; Rajta, I.; Nagy, I. Chemical Changes in PMMA as a Function of Depth Due to Proton Beam Irradiation. Mater. Chem. Phys. 2011, 130, 702−707.
(5) Chen, X.; Chen, S.; Lin, C.; Jiang, Z.; Shangguan, W. Nickels/ CdS Photocatalyst Prepared by Flowerlike Ni/Ni(OH)2 Precursor for Efficiently Photocatalytic H2 Evolution. Int. J. Hydrogen Energy 2015, 40, 998−1004. (6) Chen, W.; Chu, M.; Gao, L.; Mao, L.; Yuan, J.; Shangguan, W. Ni(OH)2 Loaded on TaON for Enhancing Photocatalytic Water splitting Activity Under Visible Light Irradiation. Appl. Surf. Sci. 2015, 324, 432−437. (7) Chen, W.; Wang, H.; Mao, L.; Chen, X.; Shangguan, W. Influence of Loading Pt, RhO2 Co-catalysts on Photocatalytic Overall Water. Catal. Commun. 2014, 57, 115−118. (8) Yu, J.; Hai, Y.; Cheng, B. Enhanced Photocatalytic H2-Production Activity of TiO2 by Ni(OH)2 Cluster Modification. J. Phys. Chem. C 2011, 115, 4953−4958. (9) He, K.; Xie, J.; Luo, X.; Wen, J.; Ma, S.; Li, X.; Fang, Y.; Zhang, X. Enhanced Visible Ligth Photocatalytic H2 Production Over Z-scheme g-C3N4 Nanosheets/WO3 Nanorods Nanocomposites Loaded with Ni(OH)x Cocatalysts. Chinese Journal of Catalysis 2017, 38, 240−252. (10) Ran, J.; Yu, J.; Jaroniec, M. Ni(OH)2 Modified CdS Nanorods for Highly Effficient Visible-Ligth-Driven Photocatalytic H2 generation. Green Chem. 2011, 13, 2708−2713. (11) Wang, P.; Lu, Y.; Wang, X.; Yu, H. Co-modification of Amorphous-Ti(IV) Hole Cocatalyst and Ni(OH)2 Electron Cocatalyst for Enhanced Photocatalytic H2-Production Performance of TiO2. Appl. Surf. Sci. 2017, 391, 259−256. (12) Liu, L.; Sun, M.; Li, Q.; Zhang, H.; Alvarez, J. J. P.; Liu, H.; Chen, W. Genotoxicity and Cytotoxicity of Cadmium Sulfide Nanomaterials to Mice: Comparison Between Nanorods and Nanodots. Environ. Eng. Sci. 2014, 31 (7), 373−380. (13) Zhu, X.; Zhou, J.; Cai, Z. TiO2 Nanoparticles in the Marine Environment: Impact on the Toxicity of Tributyltin to Abalone (Haliotis diversicolor supertexta) Embryos. Environ. Sci. Technol. 2011, 45, 3753−3758. (14) Agegnehu, A. K.; Pan, C.-J.; Rick, J.; Lee, J.-F.; Su, W.-N.; Hwang, B.-J. Enhanced Hydrogen Generation by Cocatalytic Ni and NiO Nanoparticles Loaded on Graphene Oxide Sheets. J. Mater. Chem. 2012, 22, 13849−13854. (15) Dai, K.; Zhang, X.; Fan, K.; Zeng, P.; Peng, T. Multiwalled Carbon Nanotube-TiO2 Nanocomposite for Visible-Light-Induced Photocatalytic Hydrogen Evolution. J. Nanomater. 2014, 2014, 1−8. (16) Cheng, Q.; Xu, J.; Wang, T.; Fan, L.; Ma, R.; Yu, X.; Zhu, J.; Xu, Z.; Lu, B. Double Quantum Dots decorated 3D Graphene Flowers for Higly Efficient Photoelectrocatalytic hydrogen production. Appl. Surf. Sci. 2017, 422, 528−535. (17) Peng, T.; Li, K.; Zeng, P.; Zhang, O.; Zhang, X. Enhanced Photocatalytic Hydrogen Production over Graphene Oxide−Cadmium Sulfide Nanocomposite Under Visible Light Irradiation. J. Phys. Chem. C 2012, 116, 22720−22726. (18) Li, Q.; Li, X.; Wageh, S.; Al-Ghamdi, A. A.; Yu, J. CdS/ Graphene Nanocomposite Photocatalyts. Adv. Energy Mater. 2015, 5, 1500010. (19) Kim, H.; Moon, G.; Monllor-Satoca, D.; Park, Y.; Choi, W. Solar Photoconversion Using Graphene/TiO2 Composites: Nanographene Shell on TiO2 Core versus TiO2 Nanoparticles on Graphene Sheet. J. Phys. Chem. C 2012, 116, 1535−1543. (20) Xiang, Q.; Yu, J.; Jaroniec, M. Graphene-Based Semiconductor Photocatalysts. Chem. Soc. Rev. 2012, 41, 782−796. (21) Pham, T.; Nguyen-Huy, C.; Shin, E. W. Facile One-Pot Synthesis of Nickel-Incorporated Titanium Dioxide/Graphene Oxide Composites: Enhancement of Photodegradation Under VisibleIrradiation. Appl. Surf. Sci. 2016, 377, 301−310. (22) Miao, J.; Xiao, F.; Yang, H. B.; Khoo, S. Y.; Chen, J.; Fan, Z.; Hsu, Y.; Chen, H. M.; Zhang, H.; Liu, B. Hierarchical Ni-Mo-S Nanosheets on Carbon Fiber Cloth: A Flexible Electrode for Efficient Hydrogen Generation in Neutral Electrolyte. Nanomaterials 2015, 14, 1−14. (23) Rao, Y.; Wang, Y.; Ning, H.; Li, P.; Wu, M. Hydrotalcite-like Ni(OH)2 Nanosheets in Situ Grown on Nickel Foam for Overall Water Splitting. ACS Appl. Mater. Interfaces 2016, 8, 33601−33607. 1484
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485
Article
The Journal of Physical Chemistry C (43) Cui, Y.; Jiang, X.; Feng, C.; Gu, G.; Xu, J.; Huang, X. First Double Hydrophilic Graft Copolymer Bearing a Poly(2-hydroxylethyl acrylate) Backbone Synthesized by Sequential RAFT Polymerization and SET-LRP. Polym. Chem. 2016, 7, 3156−3164. (44) Chen, X.; Zhou, J.; Ma, J. Synthesis of a cationic fluorinated polyacrylate emulsifier-free emulsion via ab initio RAFT emulsion polymerization and its hydrophobic properties of coating films. RSC Adv. 2015, 5, 97231−97238. (45) Ye, M.; Zhang, D.; Han, L.; Tejada, J.; Ortiz, C. Synthesis, Preparation, and Conformation of Stimulus-Responsive End-Grafted Poly (methacrylic acid-g-ethylene glycol) Layers. Soft Matter 2006, 2, 243−256. (46) Zhao, Y. L.; Wang, J. M.; Chen, H.; Pan, T.; Zhang, J. Q.; Cao, C. N. Al-Substituted α-Nickel Hydroxide Prepared by Homogeneous Precipitation Method with Urea. Int. J. Hydrogen Energy 2004, 29, 889−896. (47) Ciplak, Z.; Yildiz, N.; Calimli, A. Investigation of Graphene/Ag Nanocomposites Synthesis Parameters for Two Different Synthesis. Fullerenes, Nanotubes, Carbon Nanostruct. 2015, 23, 361−370. (48) Yoon, Y.; Zheng, M.; Ahn, Y.; Park, W. K.; Yang, W. S.; Kang, J. Synthesis of Magnetite/Non-Oxidative Graphene Composites and Their Application for Arsenic Removal. Sep. Purif. Technol. 2017, 178, 40−48. (49) Jerng, S.; Yu, D. S.; Lee, J. H.; Kim, C.; Yoon, S.; Chun, S. Graphitic Carbon Grown on Crystalline and Amorphous Oxide Substrates Using Molecular Beam Epitaxy. Nanoscale Res. Lett. 2011, 6, 565. (50) Smith, M.; Scudiero, L.; Espinal, J.; McEwen, J.; Garcia-Perez, M. Improving the Deconvolution and Interpretation of XPS Spectra from Chars by Initio Calculations. Carbon 2016, 110, 155−171. (51) Ran, J.; Yu, J.; Jaroniec, M. Ni(OH)2 Modified CdS Nanorods for Higly Efficient Visible-Ligth Driven Photocatalytic H2 Generation. Green Chem. 2011, 13, 2708−2713. (52) Dupin, J.; Gonbeau, D.; Vinatier, P.; Levasseur, A. Systematic XPS Studies of Metal Oxides, Hydroxides and Peroxides. Phys. Chem. Chem. Phys. 2000, 2, 1319−1324. (53) Huang, J.; Wang, D.; Yue, Z.; Li, X.; Chu, D.; Yang, P. Ruthenium Dye N749 Covalently Functionalized Reduced Graphene Oxide: A Novel Photocatalyts for Visible Ligth H2 Evolution. J. Phys. Chem. C 2015, 119, 27892−27899. (54) Yeh, T.-F.; Cihlar, J.; Chang, C.-Y.; Cheng, C.; Teng, H. Roles of Graphene Oxide in Photocatalytic Water Splitting. Mater. Today 2013, 16, 78−84. (55) Zhang, N.; Zhang, Y.; Xu, Y. Nanoscale Recent Progress on Graphene-Based Photocatalysts : Current Status and Future Perspectives. Nanoscale 2012, 4, 5792−5813. (56) Sampaio, M. J.; Silva, C. G.; Marques, R. R. N.; Silva, A. M. T.; Faria, J. L. Carbon Nanotube−TiO2 Thin Films for Photocatalytic Applications. Catal. Today 2011, 161, 91−96. (57) Hu, C.; Teng, H. Structural Features of P-type Semiconducting NiO as a Co-catalyst for Photocatalytic Water Splitting. J. Catal. 2010, 272, 1−8. (58) Jiang, X.; Nisar, J.; Pathak, B.; Zhao, J.; Ahuja, R. Graphene Oxide as a Chemically Tunable 2-D Material for Visible-Light Photocatalyst Applications. J. Catal. 2013, 299, 204−209.
1485
DOI: 10.1021/acs.jpcc.7b10375 J. Phys. Chem. C 2018, 122, 1477−1485