Co 2+ , Fe 2+ , and Ni 2+ : Modifiers for Photocatalytic Deposition of Highly Active Pt on Graphene-Based Supports

: This study focuses on photocatalytic syntheses, in which transition metal ions (Co 2+ , Fe 2+ , or Ni 2+ ), as the hole scavengers and surface modifiers of partially reduced graphene oxide, PRGO, were utilized to photoreduce Pt 4+ . A pulsed UV reactor was used to illuminate the precursors. The electrostatic interaction between the metal ions (Co 2+ , Fe 2+ , or Ni 2+ ) and the oxygen functional groups on PRGO was the main parameter, proposed to be the reason controlling Pt 4+ reduction and Pt structure and activity. The alternative assumption in managing the oxidation states of Pt was the variation in the oxidation rates of hole scavengers. Pt electrocatalysts’ structural and electrochemical characteristics revealed that utilizing the cobalt-based hole scavenger caused a dominant growth phase of Pt particles at preferred positions on PRGO, with metallic states and improved electrocatalytic activities (ECSA value of 191 m 2 · g − 1 for Co 2+ vs 141 m 2 · g − 1 and 127 m 2 · g − 1 for Fe 2+ and Ni 2+ , respectively). Density functional theory (DFT) calculation, on the other hand, suggested that the greater affinity of cobalt and iron ions to oxygen groups could detach more “O” from the graphene plane. Based on the DFT results, less “O” groups in the vicinity of Pt particles gave an amorphous morphology to Pt and facilitated the hydrogen oxidation reaction (HOR).


INTRODUCTION
Environmental issues, the fossil fuel crisis, and population growth have elevated the demand for advanced and green energy storage or conversion facilities.One of the basic obstacles to achieve the commercialization stage in developing polymer electrolyte fuel cells (PEMFCs) is the few expensive options, platinum or platinum alloys, which could be used as the catalyst layer.In addition, the sluggish nature of oxygen reduction reaction (ORR) and its high overpotential necessitate a high loading of expensive catalysts. 1−10 Carbon black has been considered to be an acceptable option for this purpose. 11,12However, its low corrosion resistance could not provide long-term performance for PEMFCs. 13−16 Ag/TiO 2 , 17,18 Pt/TiO 2 , 19 and Cu/ZnO 20 could be a few from the pool of metal(s)/semiconductor(s) composites being used in areas such as water splitting and hydrogen production, 21−25 dye degradation and organic materials oxidation, 19,26,27 removing water contamination, 28 ozonation of pesticides, 29 conversion of CO 2 to organic fuels like methanol, 23,30 and bacterial disinfection. 31hen semiconductor materials are illuminated by a light source with the right wavelength, electrons in the valence band (VB) are excited to move into the conduction band (CB).The produced hole could be used to oxidize the adjacent chemicals, while the electron could be the source of reduction. 15,32,33The key point here is the redox potential of the in-contact ions or chemicals.If the redox potential is more positive than the electron in the CB, the reaction would be a reduction.However, if the reduction potential is more negative than the potential of the hole, oxidation would happen (Figure 1a). 23ome chemicals could be used to scavenge the hole in the VB by their oxidation preference, 33 so the electron in the CB would be responsible for the reduction of metal ions. 16,20,34A study conducted by Wenderich et al. 15 demonstrated that the existence of hole scavengers could change the phase and composition of deposited metal particles.They showed a very high dependence of [PtCl 6 ] 2− adsorption and Pt photodeposition on the duration of illumination and methanol as the hole scavenger.Without using the hole scavenger, the Pt particles had a smaller size with higher oxidation states, whereas using methanol resulted in the kinetically faster deposition of clustered metallic Pt particles. 15−37 There are four types of oxygen functional groups on GO: epoxy and hydroxyl on the basal plane and carboxyl and carbonyl on the edges (Figure 1b).These functional groups create sp 3 zones while disrupting the sp 2 structure. 38The heterogeneous electrical properties of graphene oxide are because of sp 2 and sp 3 hybridizations. 39O's band gap is also affected by the oxidation level. 40Band gap modulation has opened up a new area for the application of GO in photovoltaic devices 34,41 and solar cells. 39GO has been used, as in an aqueous solution using methanol as the hole scavenger, for water splitting and hydrogen generation. 41here are also other studies in which GO was used as a support for the deposition of gold 42 and a photocatalyst for reduction of Cr(VI) 43 for preparing dye-degrading photocatalysts.Partially reduced graphene oxide, PRGO, could be referred to as modulated band gap GO that shows semiconducting properties. 34,41,44,45he photocatalytic deposition of metals, M, on PRGO could be explained with the following steps: 46 adsorption of M X+ containing a complex on the surface; electron−hole separation by light illumination; stepwise reduction of M X+ to the M 0 state, and at the same time the oxidation of the second chemical, the hole scavenger; formation of the Mcryst (crystallites of M) by two possible pathways: agglomeration of the reduced atoms or reduction of the complexes on the surface of previously deposited nuclei.The light source and the types of hole scavengers are crucial factors in regulating these steps. 14raphene could also be among novel support materials to be used in catalyst development. 2,4,44,45,47The beneficial properties of GO as a promising support in PEMFCs' catalyst layer, CL, have been studied and mentioned elsewhere. 2,4,44,47It was demonstrated that graphene has H 2 up-taking ability, 48 which led to less mass transport resistance and better kinetics of hydrogen oxidation.It has been reported that a photocatalytic deposition could result in carbon/Pt composites with strong interaction between graphitic domains of carbon and Pt. 49owever, there is a gap in the literature about the role of hole scavengers, and their interactions with graphene-based semiconductors, on the photocatalytic deposition of metal/metal oxides.One of the reasons for this gap could be the greater emphasis that has been put on the band energy level and less focus on the role of functional groups.Graphene oxide functional groups contain electronegative oxygens, which could lead to different interactions with ionic or organic hole scavengers, affect the charge transfer between species, and hence reduce metal ions at different spots.Studies have shown different types of interactions between graphene oxide and metal ions or alcohol molecules in aqueous solutions. 50,51n this study, a novel technique for synthesizing Pt electrocatalysts by photocatalytic deposition was proposed.In addition to the pulsed photocatalytic approach, to modify the Pt particles' formation, metal ions (Co 2+ , Fe 2+ , or Ni 2+ ) were utilized as the hole scavengers and Pt deposition modifiers.PRGO was the support and the photocatalyst. 13o specific chemicals for the reaction were needed, except the transition metal ions that, depending on their interaction with PRGO sheets, determined the Pt deposition kinetics and final activity.Following the proposed approach and the synthesis procedure could enable one to gain a deeper insight into the deposition of metals on graphene-based photocatalysts and could clarify that why, in some cases, graphene/metal catalysts show enhanced electrocatalytic activities, but, elsewhere, the performances could be poor.This study also proposed a new idea in the bimetallic synthesis of electrocatalysts, through which the second metal just assisted the formation of a highly active form of the main one, by an initial interaction with the graphene functional groups, and played almost no role in the final performance.

GO Synthesis. A detailed explanation of GO (by modified
Hummers' method 39 ) and PRGO synthesis, the reactor design, and the method used for the metallic particle deposition on semiconductor materials by the proposed method are explained in previous studies and video articles. 44,45For PRGO preparation, the same volumes of the GO suspension and 4 M sodium hydroxide (NaOH, >98%, Sigma-Aldrich) solution were mixed and stirred/ refluxed for 8 h at 90 °C.The product was then centrifuged and washed repeatedly with DI water.

Photocatalytic Deposition of Pt.
A more diluted PRGO suspension (50 mg/L) was prepared to make sure that UV light penetrated the suspension passing through the setup's central quartz tube (Figure 2a).
Cobalt chloride, nickel chloride, or iron chloride salt (>98%, Sigma-Aldrich) was added to 1500 mL of the PRGO suspension (60 mg of CoCl 2 •6H 2 O, and similar molarity for other salts).For the comparison, lower concentration of cobalt chloride (30 mg, half of the first sample) was chosen to prepare the control sample.N 2 gas was purged into the suspension while stirring for 30 min before the UV illumination.The N 2 purging was continued throughout the synthesis to maintain a positive pressure of N 2 gas inside the circulation path and to eliminate the generated gas caused by reactions.Pt 4+ ions were added to the suspension (20 wt % with respect to initial GO) by the addition of hexachloroplatinic acid (H 2 PtCl 6 , 8 wt % solution in water, Sigma-Aldrich) and mixed well before the illumination.
The reactor's design, Figure 2a, provided illumination in a pulsed form on the circulating suspension.Considering the quartz tube dimensions, 50 cm (L) × 0.5 cm (Ø), the volume of the reservoir, the length of the opaque pipe connections, and the type of the pump, we were able to define the time intervals by which the suspension faced the UV light (on-time or t on ) and by which the suspension is not   under UV illumination (off-time or t off ). Figure 2b demonstrates the pulse wave and the time intervals.t on and t off define the duty cycle parameter based on eq 1.
In this experiment, the parameters were t on = 1.5 s, t off = 66 s, and θ = 2.2%.

Post-treatment of Electrocatalysts.
It was expected that, in addition to Pt, other metals would be deposited in an oxide form during the photocatalytic oxidation.To remove those species to gain a higher electrocatalytic activity, a post-treatment process was conducted.The prepared samples which were centrifuged and washed with DI water were dispersed in 140 mL of highly concentrated ascorbic acid (60 g•L −1 , C 6 H 8 O 6 , 99%, Sigma-Aldrich) solution as a reducing agent and stirred/refluxed at 90 °C for about 2.5 h.They were subsequently centrifuged, repeatedly rinsed by DI water, and kept for further characterizations.
2.4.Structural Characterizations.The size, morphology, and distribution of Pt particles were studied by high-resolution transmission electron microscopy (HRTEM, JEOL JEM-ARM200CF).Also, for the elemental distribution analysis, a high-angle annular darkfield (HAADF) imaging was performed.The sample preparation for TEM was done by collecting a diluted aqueous suspension of the final product on holey carbon-coated TEM grids.For examining the oxidation states of Pt and Co, as well as the structural variations of graphene, a high-resolution Thermo Fisher K-alpha X-ray photo-electron spectrometer (XPS) system using a 400 elliptic radius spot size of Al Kα monochromatic source was adopted.

Electrocatalytic Characterizations.
A three-compartment electrochemical cell and a rotating disc electrode (RDE) setup (Gamry Instruments Reference 3000 potentiostat/galvanostat/ZRA and Pine rotating control) were employed for electrochemical evaluations.The ink preparation was simply done by adjusting the amount of DI water to the solid amount (3.375 g/L).Then the suspensions were ultrasonically agitated for 1 h to prepare a homogeneous ink.16.66 μL of the resulted ink was drop-cast onto glassy carbon electrodes, with the surface area of 0.19625 cm 2 and then dried in an oven at 60 °C for 10 min.These ink-covered glassy carbon electrodes were used as the working electrodes (WE) and a Pt wire as the counter electrode (CE).The identical parameters from our previous research were used for the cyclic voltammetry (CV) tests and ORR experiments. 2,52.6.Computational Method.To understand the interaction of Co, Ni, and Fe ions with the GO plane, and the structural properties of Pt particles and adsorption of H 2 on platinum deposited on graphene without oxygen functional groups, density functional theory (DFT) calculations were performed via the quantum-espresso simulation package. 53,54The parameters of the two systems are presented in the Table 1.
The PBE 55 exchange and correlation potential was used for the systems containing Pt; for the other samples containing Co, Ni, or Fe, the PBESOL 56 exchange and correlation energy functional was used.Projector-augmented wave (PAW) pseudopotentials were used for elements of systems, unless the system contained Ni or Fe in which    supports.To assess the interaction between the graphenebased supports and the transition metal ions, during the first step, DFT calculations were conducted.After optimizing the necessary parameters such as the kinetic energy cutoff and the K-sampling points (Table 1), the "relax" calculations were performed.The geometry and distance of the atoms were  similar in all three systems (Figure 3).The relaxed structures demonstrated a lower affinity of Ni ions toward oxygen groups compared to samples mixed with Co or Fe ions, which caused a relatively higher amount of remaining oxygen groups on the graphene plane in the former case.In this regard, Fe ions seemed to detach more planar oxygen functional groups.Modifying the concentration of oxygen groups led to deposition of Pt with various structural and morphological properties which we will see in the following sections.
The X-ray diffraction (XRD) spectra of the Pt/rGO electrocatalyst prepared by premodification of PRGO samples with Co 2+ , Ni 2+ , or Fe 2+ are shown in Figure 4a,b.The (111), ( 200), (220), and (311) planes (with respectable peaks) of the platinum face-centered cubic structure are visible in all samples at 2θ = 39.8°,46.5°, 68.7°, and 82.1°, respectively.The most remarkable aspect of the effect of different metal ions on the final structure of Pt crystallites is the difference in peak intensities and sharpness, which is broad for Co 2+ -assisted synthesis and sharp for Ni 2+ -assisted synthesis.The more crystalline structure of the samples might be one factor contributing to the greater intensities in diffraction patterns.Broadened peaks are present in smaller crystallites or amorphous formations.However, because to GO's amorphous nature and a unique graph adjustment to eliminate the background peak, it is extremely important to evaluate peak intensities and relative crystallinity with great care.To evaluate the particle distribution and their morphology, HRTEM with HAADF imaging was performed.The results are demonstrated in Figure 4c−k.
Pt distribution was different in the three samples (Figure 4e,h,k).In the case of using Co 2+ , more agglomeration of the Pt was seen.The use of Ni 2+ caused the best uniform distribution of Pt, and Fe 2+ resulted in less coverage of Pt on graphene.To assess the effect of transition metals on D and G vibration modes of the rGO/Pt electrocatalysts, Raman spectroscopy was conducted (Figure 4l,m).Defects in the carbon lattice, vacancies, oxygen functional groups, or deposited particles change the sp 2 structure to sp 3 , and the intensity of the D band increases. 58In completely oxidized graphene, all carbon atoms are sp 3 -hybridized and the D band disappears.The intensity of the D band peaks as the amount of deficits rises.In Figure 4l,m, the highest value of I D /I G can be seen (letter "I" represents the intensity) for rGO-Pt prepared by Ni 2+ as the hole scavenger and the lowest one for the Fe 2+ pretreated one.By comparing Raman and TEM images (before ascorbic acid treatment), the highest I D /I G belonged to the sample with more even distribution of Pt particles with less agglomeration.This result highlights the hypothesis that small and good distribution of Pt particles might further distort the sp 2 .
In Figure 5a, schematic steps from Pt 4+ adsorption to the Pt agglomeration or crystallite formation are demonstrated.It has been claimed that Pt 4+ is reduced to Pt 0 in two steps: (1) Pt 4+ to Pt 2+ and (2) Pt 2+ to Pt 0 . 59The oxidation rate of hole scavengers may depend on their adsorption mechanism, availability (concentration), and redox potential or kinetics. 60,61One of the factors in studying GO behavior in aqueous solutions containing metal ions or alcohol molecules is GO's interaction with its surrounding species.Based on a study conducted by Wang et al., 50 metal ions, M + , could be attracted to negatively charged GO sheets and cause partial neutralization of GO (eq 2).
x GO M GOM However, methanol (as an organic hole scavenger 62 ) and water molecules tend to form hydrogen bonds with GO functional groups.The probability of forming a hydrogen bond between water molecules and GO is much more than forming the bond between the methanol molecule and GO, especially when the GO has already been stored in aqueous solutions. 51s a result, metal ions are more accessible for oxidation on the surface of PRGO and reducing the holes.Figure 5d shows the change of zeta potential value on PRGO particles in solutions with different types of hole scavengers and metal ions.All of these factors could have a determining effect on the oxidation kinetic of hole scavengers and hence the mechanism, position of deposition, and oxidation state of reduced Pt X+ particles.
The XPS method was employed to gain more detailed information about the chemical structure of the electrocatalysts.Figure 6 illustrates the binding energy of platinum, carbon hybridization states, and oxygen functional groups of two samples prepared with low and high concentrations of Co 2+ .The XPS spectra of Pt 4f demonstrated a doublet peak at 71.8 and 75.2 eV.Comparison of Figure 6a to Figure 6c revealed that the binding energy of Pt has shown more contribution from Pt 2+ and Pt 4+ states in samples with lower amount of Co 2+ (shifting the binding energy higher values and less reduction of Pt).
Not accomplishing the Pt 4+ reduction to Pt 0 would lead to higher Pt oxidation states rather than the metallic state. 63In Figure 6d and comparing it to Figure 6b, which were the deconvolution of the C 1s from the graphene-based supports, showed a relative increase in the binding energy of carbon electrons by shifting to the right, showing a greater contribution of oxygen functional groups including hydroxides (C−OH), epoxy (C−O−C), carbonyl (−COOH), and carboxyl (C�O) in samples with a higher amount of Co 2+ .In other words, by evaluating Figure 7, cobalt has been deposited on the samples mostly in Co 3 O 4 form, and the disappearance of cobalt element after the ascorbic acid reduction shifted the binding energy of electrons of carbon atoms to higher values.

Electrocatalytic Activities.
To assess the number of active sites for the hydrogen oxidation reaction, HOR, CV experiments were conducted (Figure 8).Results from previous investigations have shown that the current approach produces Pt ions with a very high deposition efficiency and stability. 44,45,62Based on those results, the ECSA and ORR performance were calculated with respect to 20 wt % Pt loading.The most promising electrochemical activity was for the sample prepared with a high concentration of Co 2+ , with the ECSA value of 191 m 2 •g −1 .The ECSA values for the other samples were 141, 127, and 46 m 2 •g −1 for the ones prepared with Fe 2+ , Ni 2+ , and a low concentration of Co 2+ , respectively.
The possible reason for the lower electrocatalytic activity of some of the samples might be the Pt deposition spots, morphology, or oxidation states (TEM and XPS data in Figures 4 and 6).Pt precursors, reducing agents, additives, reaction temperature, and time are only a few examples of the variables that might modify the rate of Pt reduction from the ionic states. 64o evaluate the ORR performance of the synthesized samples, linear sweep voltammetry (LSV) tests by employing rotating disk electrodes at the rotation speed of 100, 400, 900, and 1600 rpm were conducted.From Figure 9, a mixed kineticdiffusion-controlled region appeared from 0.8 to 0.95 V. Therefore, to calculate the mass activity value, the potential of 0.9 V was chosen.One of the ways to interpret the efficiency of ORR electrocatalysts is the electron transfer number (n), which could also reveal the mechanism of the ORR.Lower n is desired for inexpensive and safe production of H 2 O 2 , but a high concentration of H 2 O 2 has a deteriorating effect on the fuel cells' membrane.To calculate n based on the RDE experiment, the following equation could be used.
In eq 3, j, j K , and j L are the measured, kinetic-limited, and masstransfer-limited currents, respectively.ω is the angular velocity of the RDE, and B is the proportionality coefficient which depends on the diffusion coefficient of the reactant, the viscosity of the electrolyte, and the concentration of the reactant in the bulk electrolyte.In this regard, n could be deduced from the slope of the j −1 versus ω −1/2 .The number of transferred electrons, n, for the three groups of electrocatalysts were calculated to be 1.8, 2.1, and 2.2 for the electrocatalysts prepared by Co 2+ , Fe 2+ , and Ni 2+ , respectively.The mechanism for ORR could be followed by the 2 or 4 electron transfer path, or a mixture of these two.The reason for the low number of electron transfer in the mentioned electrocatalysts could be related to the graphene support oxidation.Based on DFT calculations (Figure 3), the pretreatment by the three metal ions caused the detachment of oxygen groups from the graphene surface.During the ORR, these bare surfaces of graphene are prone to oxidation and could cause the dissolved oxygen in the acidic electrolyte involved in the oxidation process, instead of H 2 O or H 2 O 2 production.Based on the sensitive nature of ORR, the samples' performance could highly depend on the structure of the catalyst and support (see XPS and TEM results).The LSV test demonstrated a higher performance for ORR behavior of the graphene/Pt electrocatalyst prepared with the high concentration of Co 2+ as the hole scavenger (0.49 A•mg −1 ).The mass activities of the other two samples were 0.39 A•mg −1 and 0.25 A•mg −1 for the ones prepared by Fe 2+ and Ni 2+ , respectively.In different studies in which the synthesis and/or support materials were different types of carbon materials, mass activities were reported to be in the range of 0.063−0.272A•mg −1 . 2,52,62The summary of the electrochemical performances is presented in Table 2.
To further assess the effect of oxygen groups coverage on the final morphology and activity of Pt/graphene oxide samples, two systems, one with graphene and the other with oxygen containing graphene, have been modeled in DFT calculations.As it is demonstrated in Figure 10, the existence of oxygen groups close to Pt particles could affect the crystal shape of Pt particles in a way that, without the functional groups, the Pt particles gain more amorphous and unstable structures.TEM and XRD results in Figure 4 also confirmed that, in the case using cobalt or iron ions as the modifier, Pt crystallinity and morphology of Pt particles were more amorphous and heterogeneous.Not only the "O" groups affected the morphology of Pt particles on graphene plane, but they also hindered the H 2 adsorption (the first step of HOR) on Pt electrocatalysts.This effect could also be confirmed with the cyclic voltammetry results in which Co-modified samples showed superior performance compared to that of Ni-and Femodified ones.

CONCLUSION
This paper has investigated the photocatalytic deposition of Pt nanoparticles on PRGO supports, using metal ions (Co 2+ , Fe 2+ , and Ni 2+ ) as the hole scavengers and surface modifiers, and the morphology and electrocatalytic performance of the synthesized electrocatalysts.Here, the catalyst support, PRGO, the synthesis procedure using a pulsed UV illumination, DFT modeling of the interaction between the metal ions/particles and graphene's "O" groups, and finally the whole procedure's effect on the morphology of Pt particles and their electrocatalytic activities may highlight the role of the chosen topic(s) in the synthesis of electrocatalysts.Based on the electrochemical tests and computational results, we concluded that an even distribution of uniform Pt nanoparticles on graphenebased materials was not always the sign of higher electrocatalytic activities, as the position and distribution of "O" groups would not be uniform.Yet, it was the Pt complex adsorption, Pt 4+ → Pt 2+ → Pt 0 reduction progress, and Pt's final oxidation state and its deposition spot with respect to the functional groups that determined the crystallinity and performance of Pt/graphene.The different affinities of transition metal ions (Co 2+ , Fe 2+ , and Ni 2+ ) to oxygen groups caused a random detachment of oxygen.The DFT modeling also showed that "O" groups in the vicinity of Pt particles hindered the adsorption of H 2 molecules on Pt, although it led to a more crystalline and stable structure of deposited Pt.Here, the transition metal precursor played a modifying role in the synthesis procedure of highly active Pt particles and almost did not have a contribution to the final products' catalytic activity, as it was removed with the final ascorbic acid treatment.Future studies could focus on the interaction between transition metal ions and other types of functional groups at defect sites and edges of graphene oxide.

Figure 1 .
Figure 1.(a) Mechanism of the photocatalytic deposition of (a 1 ) Pt along with oxidation of methanol and (a 2 ) Pt or reduction of O 2 along with oxidation of Co 2+ and (b) oxygen functional groups on graphene oxide plane.

Figure 2 .
Figure 2. (a) Schematic of the photocatalytic deposition setup and (b) resulting UV pulse wave.

Figure 4 .
Figure 4. (a,b) X-ray diffractogram of the Pt/graphene-based samples prepared with three different types of transition metal ions, before and after ascorbic acid treatment, respectively.TEM micrographs of Pt particles and their morphology on PRGO supports: (c,d) Co 2+ -assisted Pt/PRGO and (e) its Pt elemental distribution; (f,g) Fe 2+ -assisted Pt/PRGO and (h) its Pt elemental distribution; (i,j) Ni 2+ -assisted Pt/PRGO and (k) its Pt elemental distribution; (l,m) Raman spectra of samples prepared with different types of transition metal ions, before and after ascorbic acid treatment, respectively.

Figure 5 .
Figure 5. (a) Alternative mechanisms for Pt deposition with oxidation of different types of hole scavengers, (b,c) hypothetical relative position of hole scavengers, (b) CH 3 OH and (c) Co 2+ , in an aqueous suspension containing PRGO), and (d) zeta potential value of PRGO in aqueous and nonaqueous (methanol) suspensions with the probability of PRGO precipitation after 24 h.

3 .
RESULTS AND DISCUSSION 3.1.Structural and Compositional Properties.The two major steps of the experimental procedures consisted of initially mixing the partially reduced graphene-oxide-containing aqueous solution with one of the transition metal ions and then deposition of platinum on the modified graphene-based

Figure 6 .
Figure 6.X-ray photoelectron spectroscopy spectrum of Pt 4f and C 1s for (a,b) Pt/rGO deposited with a low concentration of Co 2+ and (c,d) Pt/ rGO deposited with a high concentration of Co 2+ .

Figure 7 .
Figure 7. (a) Survey XPS spectrum of two post-treated (P.T.: chemically reduced with ascorbic acid solution) samples prepared with Co 2+ ions and one without post-treatment and (b) XPS spectrum of Co 2p of the sample without post-treatment.

Figure 8 .
Figure 8.(a) Cyclic voltammetry of Pt/rGO prepared with Ni 2+ , Fe 2+ , or two concentration of Co 2+ after ascorbic acid treatment and (b) CV of three samples before the treatment.

Figure 10 .
Figure 10.(a,b) Proposed initial structure of deposited Pt nanoparticle on graphene plane, with and without planar oxygen functional groups, respectively, in an environment with 6 H 2 molecules.(c,d) "Relaxed" structure of the Pt/graphene systems (without oxygen groups): the less stable structure of Pt and more H 2 adsorption in the samples lacking "O" groups.

Table 1 .
DFT Calculation Parameters in the Two Systems