Effective and Durable Co Single Atomic Cocatalysts for Photocatalytic

Nov 21, 2017 - This research reports for the first time that single cobalt atoms anchored in nitrogen-doped graphene (Co-NG) can serve as a highly eff...
0 downloads 10 Views 1MB Size
Subscriber access provided by Columbia University Libraries

Article

Effective and Durable Co Single Atomic Cocatalysts for Photocatalytic Hydrogen Production Qi Zhao, Weifeng Yao, Cunping Huang, Qiang Wu, and Qunjie Xu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b13566 • Publication Date (Web): 21 Nov 2017 Downloaded from http://pubs.acs.org on November 29, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Effective and Durable Co Single Atomic Co-catalysts for Photocatalytic Hydrogen Production Qi Zhao, 1 Weifeng YAO, 1* Cunping Huang, 2 Qiang Wu1, Qunjie Xu1* 1

Shanghai Key Laboratory of Materials Protection and Advanced Materials in Electric Power, College of Environmental & Chemical Engineering, Shanghai University of Electric Power, Shanghai, P. R. China. 2Aviation Fuels Research Laboratory, Federal Aviation Administration William J. Hughes Technical Center, Atlantic City International Airport, NJ 08405, U.S.A. * Corresponding author: [email protected]

Abstract This research reports for the first time that single cobalt atoms anchored nitrogen-doped graphene (Co-NG) can serve as a highly effective and durable cocatalyst for visible light photocatalytic hydrogen production from water. Results show that under identical conditions the hydrogen production rate (1382 µmol/h) for 0.25 wt.% Co-NG loaded CdS photocatalyst (0.25 wt.% Co-NG/CdS) is 3.42 times greater than that of nitrogen-doped graphene (NG) loaded CdS photocatalyst (NG/CdS) and about 1.3 times greater than the greatest hydrogen production rate (1077 µmol/h) for 1.5 wt.% Pt nanoparticle loaded CdS photocatalyst (1.5 wt.% Pt-NPs/CdS). At 420 nm irradiation the quantum efficiency of the 0.25 wt.% Co-NG/CdS photocatalyst is 50.5%, the highest efficiency among those literature-reported non-noble metal cocatalysts. The Co-NG/CdS nanocomposite based photocatalyst also has an extended durability. No activity decline was detected during three cyclic photocatalytic lifespan tests. The very low cocatalyst loading, along with the facile preparation technology for this non-noble metal cocatalyst, will significantly reduce hydrogen production costs and will finally lead to the commercialization of the solar catalytic hydrogen production process. Based on experimental results we conclude that Co-NG can successfully replace noble metal cocatalysts as a highly effective and durable cocatalyst for renewable solar hydrogen production. This finding will point to a new way for the development of highly effective, long lifespan, non-noble metal based cocatalysts for renewable and cost effective hydrogen production. Keywords: Single Atom, Cocatalyst, Photocatalysis, Water-splitting, Nitrogen-doped graphene 1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1. Introduction Hydrogen production via visible light photocatalytic water splitting requires a photocatalyst consisting of two materials: a semiconductor-based main catalyst and a transition metal or alloy cocatalyst. The main catalyst absorbs and converts solar photonic energy into electron-hole pairs. Cocatalyst particles are loaded onto the surface of the main catalyst to form metal/semiconductor junctions (Schottky junctions) for the separation and storage of photo-induced electrons. Precious metals (Pt and Pd) and their alloys are excellent cocatalysts due to their optimized hydrogen adsorption energies, catalyzing proton reduction reaction and high corrosion resistance.1-3 For decades the discovery of non-noble metal based cocatalysts has been a hydrogen economy research goal. Although some non-noble metal catalysts have been reported as cathode or anode electrocatalysts for hydrogen proton exchange membrane fuel cells, due to the complexity of photocatalytic processes3-5 few successful results have been reported in literature for photocatalytic hydrogen production. The earth abundant elements Cobalt (Co), Nickel (Ni) and Copper (Cu) are the most promising candidates for cocatalysts in a photocatalytic process. Differing from a nickel metal catalyst that tends to corrode in acidic media, Co has currently emerged as a promising hydrogen evolution reaction (HER) electrocatalyst for water electrolysis in a relatively wider pH range.5 However, due to the greater hydrogen adsorption energy, the corrosion tendencies of cobalt and low efficiencies, only a few Co compounds, such as Co(OH)2, Co3O4, CoO and CoS have been investigated as potential cocatalysts.6 In recent years Co single atom dispersed nitrogen doped graphene oxide (Co-NG) has been attracting attention in electrochemical fields as an effective electrocatalyst for hydrogen production.7 Due to its high electrochemical activity, this material can potentially be used as a valuable cocatalyst for renewable visible light photocatalytic hydrogen production.8-9 Unlike electrocatalysts, however, a photocatalyst requires cocatalyst particles to be loaded onto the surface of a semiconductor photocatalyst, forming a composite material. The complex interaction of cocatalyst and semiconductor makes it more difficult to apply metal atomic cocatalysts in photocatalytic fields. This research reports for the first time the application of Co-NG as novel non-noble-metal cocatalysts for photocatalytic H2 evolution. The activity and durability of Co-NG loaded CdS (Co-NG/CdS) photocatalysts are evaluated against traditional Pt loaded CdS (Pt/CdS) 2

ACS Paragon Plus Environment

Page 2 of 17

Page 3 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

photocatalysts. Both electrochemical and photochemical characterizations have proven the activity and durability of this new and less expensive cocatalyst.

2. Experimental 2.1 Synthesis of graphene oxide (GO) GO was synthesized using an improved Hummer method.10 Briefly, a stoichiometric amount of graphite and H2SO4 mixture was stirred in an ice bath. Then 60 g KMnO4 powder was slowly added to the mixture. The mixture was first heated at 35 °C for 2 hours and then to 98 °C for 10 minutes. Finally, 2.8 L water and 50 mL H2O2 was added to stop the reaction. The precipitate was collected by vacuum filtration, washed with 5% HCl and water, and finally dried in air at 60 °C. 2.2 Synthesis of Single Co atoms anchored nitrogen doped graphene (Co-NG) Single Co atoms anchored nitrogen doped graphene (Co-NG) was prepared according to a method reported by Fei et al.7 In detail, 100 mg prepared GO was initially sonicated in 50 ml deionized water for 2 h for the formation of a homogenous suspension. Then 1.0 ml CoCl2·6H2O (3 mg/ml) aqueous solution was added into the prepared GO suspension and sonicated for another 10 min. This precursor solution was freeze-dried for one week to produce a brownish powder. The dried powder was then heat-treated in a gas mixture containing 25 vol.% of NH3 in Ar at 750 °C for 1 h at a heating rate of 20 °C/min. After NH3 treatment the sample was cooled to room temperature under an Ar atmosphere. The final product was identified as single Co atoms dispersed in nitrogen doped graphene (Co-NG). 2.3 Synthesis of Co-NG/CdS photocatalysts The prepared Co-NG cocatalysts were loaded onto a CdS photocatalyst surface using a method introduced by this group previously.8-9, 11 An appropriate amount of prepared Co-NG was added to 100 ml of deionized water containing 0.1 g commercial CdS photocatalyst. The resulting suspension was stirred for 2 h at room temperature. After centrifugation the obtained composite powders were washed and dried at 60 °C. UV−vis absorption spectral analyses indicated that the Co-NG was completely deposited onto cadmium sulfide particles. The Co-NG loading concentration (mass%) was calculated as Co-NG mass/(CdS mass + Co-NG mass) * 100%. 2.4 Photocatalytic activity measurements Visible light photocatalytic hydrogen production was carried out as follows: 0.10 g prepared Co-NG/CdS powder was suspended in a 100 ml 1.0 M aqueous (NH4)2SO3 solution. The solution 3

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

was then transferred to a Pyrex glass reactor cell (Perfectlight Co., Labsolar-III). The system was vacuum-degassed and irradiated using a 300 W Xe lamp (PLS-SXE300/300UV, Perfectlight Co.). A water filter and an optical UV cutoff filter (> 420 nm) were used to remove the infrared and UV radiation portions of the spectra. Hydrogen evolution was measured using an online gas chromatograph (Techcomp Limited Co., GC7890II) via a thermal conductivity detector. The apparent quantum efficiency (Q.E.)

12-15

of the photocatalyst toward H2 production is defined as:

Q.E.(%) = 2*NH2/NP*100%. Where, NH2 and NP refer to the moles of hydrogen evolution and the total moles of incident photons absorbed by the photocatalyst, respectively. 3. Results and discussion 3.1 Characterizations of Single Co atoms anchored Co-NG Single cobalt (Co) atoms anchored in nitrogen-doped graphene (Co-NG) was prepared by heating a mixture of graphene oxide (GO) and cobalt salt in NH3 atmosphere, based on a method reported by Fei et. al.7. Fig. 1 shows the TEM and High Angle Annular Dark Field Scanning Transmission Electron Microscopy (HAADF-STEM) images of prepared Co-NG powders. No Co nanoparticles but N-doped graphene (NG) sheet can be observed under low resolution TEM analysis (Fig. 1(A)), suggesting that cobalt species are highly dispersed either as sub-nano sized clusters or single atoms whose small sizes are beyond the detection of low level TEM resolution. The XRD patterns of GO show an intense and sharp diffraction peak at 2θ = 10.6°, which is attributed to the (001) lattice plane of GO (Supporting Information, Fig. S1). This is consistent with the lamellar structure of GO sheets. The number of sheets can be reduced during the NH3 treatment process. After treating GO with ammonia, a broad (002) peak at 2θ = 26.6° appears with the disappearance of the (001) (2θ = 10.6°) peak for NG and Co-NG XRD spectra. This result suggests that GO has been reduced to r-GO or nitrogen doped graphene sheets (NG) due to the removal of functional groups from the GO during the NH3 treatment process. No XRD peaks of cobalt metal crystal particles or cobalt oxide particles were obtained, indicating that Co species in nitrogen doped graphene are highly dispersed or amorphous. The HAADF-high resolution STEM images show a large number of bright single Co atoms uniformly dispersed on the NG matrix (Figs. 1(B) and 1(C)). The diameters of 99% of Co atoms are less than 0.3 nm, indicating that Co on NG consists of isolated single atoms.7 The single Co atoms were further verified using a new HAADF image produced by a probe-aberration-corrected cold field emission JEM-ARM200CF 4

ACS Paragon Plus Environment

Page 4 of 17

Page 5 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

operating at 200 kV (Figs. 1(D)). Elemental mapping results (Fig. S2 of the Supporting Information) confirm the existence of cobalt (Co), nitrogen (N), carbon (C) and oxygen (O) atoms in the prepared Co-NG cocatalyst. The elemental mapping also confirms that no impurity atoms were involved during the synthesis of Co-NG. Fig. 2(a) shows the Raman spectra of the prepared GO, NG and Co-NG samples. Two prominent peaks at about 1362 and 1597 cm-1 are assigned to the D and G bands of pure graphite oxide. The D band is attributed to the vibration of sp3 carbon atoms in the disordered GO nanosheet, while the G band is related to the vibration of sp2 carbon atom domains of graphite.16 It is noteworthy that the intensity ratios of D- and G-bands (ID/IG) of GO, NG and Co-NG increase from 0.83 to 1.02 and 1.03, respectively. This result indicates that most oxygenated groups in GO were removed during the simultaneous nitrogen doping and Co dispersion processes. The comparable value of ID/IG for NG and Co-NG also suggests that the atomic Co deposition has no obvious effect on the disorder/defects of NG nanosheets.17-18 The FT-IR spectrum (Fig. 2(b)) characterizes four absorption bands at 1076 cm-1 (νC-O), 1394 cm-1 (νC-OH), 1639 cm-1 (νC=O) and 3462 cm-1 (νO-H).19-20 The absorption bands of oxide groups in GO, NG and Co-NG (νC-OH and

νC=O) decrease greatly from GO to NG and Co-NG, suggesting that GO has been partially reduced as indicated in Raman spectra. XPS is an effective method for studying the oxidation states of atoms on the surface of a material. The high-resolution Co 2p and N 1s XPS spectra of Co-NG are shown in Fig 2(c) and 2(d), respectively. The high-resolution Co 2p spectra are split into 2p1/2 and 2p3/2 spin-orbits. The main intensive peak at 779.2 eV and 794.2 eV can be ascribed to Co 2p3/2 and Co 2p1/2, respectively

21-22

(Fig. 2(c)). The 15.0 eV difference between these two peaks indicates the

presence of the Co(III) ions7. The N 1s binding energy peaks of Co-NG (Fig. 2(d)) can be considered as four separate peaks: 398.4 (pyridinic N and N-Co), 399.8 (pyrrolic N), 401.2 (graphitic N) and 402.8 (N oxide) eV.23-24 The binding energies for pyridinic N and N-Co are so close that the two peaks cannot be separated further.25 Binding energy intensity comparison shows that pyridinic N and N-Co are the major species of nitrogen in Co-NG. Co and N concentrations in Co-NG are 2.00 and 13.87 wt.%, respectively, according to ICP analyses. 3.2 Photocatalytic activities of Co-NG/CdS Photocatalysts The visible light photocatalytic activities of Co-NG/CdS photocatalysts for hydrogen 5

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

production were evaluated via the photocatalytic oxidation of an aqueous (NH3)2SO3 solution (Fig. 3(a)). An adsorption technology8-9, 11 was applied to load prepared Co-NG onto CdS. TEM images show the deposition of CdS particles on Co-NG (Supporting information, Fig. S3). No photocatalytic activity was observed under dark conditions. The H2 production rate of pure CdS is only 57 µmol/h (Fig. 3(a)) as compared with 1382 µmol/h for 0.25 wt.% Co-NG/CdS photocatalyst. By comparison, the activity of 0.25 wt.% NG/CdS photocatalyst is 404 µmol/h, 3.42 times lower than that of 0.25 wt.% Co-NG/CdS. This result indicates that single Co atoms dispersed in NG matrix are the major proton adsorption and reduction centers. BET specific surface areas of NG/CdS and Co-NG/CdS are not significantly different due to the very low cocatalyst loading concentration (0.25 wt%). As shown in Table S1 (Supporting information), the surface area of 0.25 wt.% Co-NG/CdS is nearly the same as those of the bare CdS and 0.25 wt.% NG/CdS. However, the rate of hydrogen production over 0.25 wt.% Co-NG/CdS photocatalyst is much greater than that of 0.25 wt.% NG/CdS or bare CdS. These results indicate that the BET specific surface area for a single atom-based catalyst is not the critical factor determining the photocatalytic activity of Co-NG/CdS. Results also showed that the photocatalytic activity of 0.25 wt.% Co-NG/CdS photocatalyst can be even much greater than that of Pt noble metal loaded CdS (Pt/CdS). As shown in Fig. 3(a), under identical conditions the activity of 0.25 wt.% Pt/CdS is only 455.6 µmol/h, 3.03 times lower than that of 0.25 wt.% Co-NG/CdS. The quantum efficiency of 0.25 wt% Co-NG/CdS for hydrogen production is 50.5% at 420 nm of irradiation under the current reaction conditions. This is the highest quantum efficiency among all reported non-noble metal based cocatalysts (Table S2 of the Supporting Information). All of these results suggest that Co-NG can be an excellent non-noble metal cocatalyst and could significantly reduce the cost of solar photocatalytic hydrogen production. The experimental results have shown that hydrogen production rates of Co-NG/CdS photocatalysts are heavily dependent on Co-NG loading quality. A very narrow concentration range, centered at 0.25 wt.%, was found to have maximum activity. When the Co-NG loading concentration was outside this range the activity of Co-NG/CdS dropped rapidly to only ¼ the highest rate for 0.25 wt.% optimal loading. When Co-NG loading was increased from 0.00 to 0.25 wt.% the hydrogen rate of Co-NG/CdS dramatically increased from 57 to 1382 µmol/h. However, further increasing Co-NG loading beyond 0.25 wt.% resulted in a significant decline in the activity. 6

ACS Paragon Plus Environment

Page 6 of 17

Page 7 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

This is attributed to the optimization of the light blocking effect of Co-NG on the CdS surface. In contrast, the activity of Pt/CdS photocatalysts as a function of Pt loading does not have a very significant critical activity range. As shown in Figure 3(c), 1.5 wt% Pt/CdS achieves a maximum 1077 µmol/h hydrogen evolution rate. Differing from Co-NG/CdS, when Pt loading increases from 0.5 to 2.0 wt.% the activity of Pt/CdS does not change significantly. It can be concluded that Co-NG/CdS requires a lower level of Co-NG loading than does Pt/CdS photocatalyst (Fig. 3(b) and (c)). However, under identical conditions, the maximum efficiency of 0.25 wt.% Co-NG/CdS photocatalyst can be 1.3 times greater than 1.5 wt.% Pt/CdS photocatalyst (1382 vs. 1077 µmol/h). This very significant result indicates that Co-NG can be an effective non-noble metal cocatalyst for the replacement of Pt or Pd, thereby dramatically reducing renewable solar hydrogen production cost. Fig. 3(d) illustrates the stability of 0.25 wt.% Co-NG/CdS. In three photocatalytic reaction cycles the hydrogen rates of 0.25 wt.% Co-NG/CdS are almost identical. In total, 20.13 mmol of pure H2 was produced during a 15-hour reaction. The turnover numbers (TON), defined as the molar ratio of the total hydrogen atoms evolved (20.13 x 2 = 40.26 mmol) per mole of CdS photocatalyst (0.692 mmol) or per mole of Co cocatalysts (8.48 * 10 -5 mmol), are 58.2 and 474,764 for CdS and Co, respectively. The turnover frequency (TOF), calculated based on hydrogen atoms produced per Co metal atom per second over the 0.25 wt.% Co-NG/CdS

photocatalyst, is about 8.8 S-1. These very high TON and TOF numbers indicate that the produced hydrogen has resulted from the photocatalytic reduction of water rather than from the photoetching of Co-NG/CdS photocatalysts. It must be pointed out that under the same conditions the TOF numbers of Pt-Pd nanocubes/CdS and Pt-Pd nanooctahedra/CdS, two previously reported efficient noble metal cocatalysts, are only 3.28 S-1 and 0.97 S-1, respectively.8 All the results of activity, stability and higher TON and TOF of Co-NG/CdS indicate conclusively that Co-NG is a highly effective and durable cocatalyst for cost-effective photocatalytic hydrogen production and for replacement of Pt or Pd noble cocatalysts. 3.3 Influence of Co-NG loading on the photocatalytic performance of CdS There are many factors affecting the activity and durability of a cocatalyst. The three main factors can be summarized as: (1) An active cocatalyst must also be a highly efficient proton reduction catalyst with 7

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

optimized hydrogen adsorption energy. In comparison to noble metal Pt, the hydrogen adsorption energy of Co is not the most favorable to catalyze proton reduction. However, the integration of single Co atoms into NG reduces the hydrogen adsorption energy level of Co atoms and maximizes their activity as a highly effective proton reduction catalyst. Additionally, reported hydrogen adsorption energy “volcano curves” are based on bulk or nanoparticles of transition metals that may not be completely applicable to single Co atoms in NG. (2) Impedance between cocatalyst particles and CdS can significantly affect the electron transfer from CdS to cocatalyst. Higher impedance means more electrical energy loss, leading to a lower hydrogen production rate. NG is an efficient electrical conductor which promotes the electron transfer from CdS to Co atoms. NG’s two dimensional structure with large surface area enhances the contact of cocatalyst with CdS and further reduces the impedance of the Schotty junction. (3) Resistance of electron migration in a crystalline cocatalyst can be much great than that of a single atomic cocatalyst. To better understand the effect of Co-NG loading concentration on the rate of electron transfer between the interfaces of CdS and Co-NG we carried out some tests for Co-NG/CdS using photoluminescence (PL) spectroscopy, electrochemical impedance spectra (EIS), linear-sweep voltammogram (LSV) and photocurrent measurements. PL spectroscopy is used to measure the separation and transfer efficiency of the photogenerated charge carriers. When excited by 325 nm light irradiation at room temperature CdS shows broad luminescence emission centered at about 550 nm (Fig. 4(a)). The width of the emission band ranges broadly from 500 to 600 nm. This wide PL emission spectrum for CdS is attributed to: (1) an overlapping band edge emission; (2) shallow trapping states and (3) sub-bandgap energy states.26 A notable decrease in the CdS emission band was observed after Co-NG loading, suggesting that the photogenerated electron-hole recombination in Co-NG/CdS can be effectively suppressed due to the existence of Co-NG cocatalyst particles on the surface of the semiconductor. This effect could originate from electron trapping and accumulation in the Co-NG cocatalyst. The increased transportation efficiency of photoinduced electrons from CdS to Co-NG effectively suppresses the recombination rates of photoinduced electron-hole pairs and finally leads to the improvement of photocatalytic activity of Co-NG/CdS. 8

ACS Paragon Plus Environment

Page 8 of 17

Page 9 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

To study the interfacial electron transfer resistance between the interface of Co-NG and CdS, electrochemical impedance spectral (EIS) measurements, Fig. 4(b), for CdS, NG/CdS and Co-NG/CdS samples were conducted using a typical three-electrode electrochemical cell. In comparing with CdS and NG/CdS photocatalyst samples, the semicircular diameter of the Nyquist plot for Co-NG/CdS, Fig. 4(b), has been significantly reduced. This observation suggests that the electron transfer efficiency of Co-NG/CdS is greater than that of NG/CdS and pure CdS.27-28 In other words, Co-NG cocatalyst accelerates electron transport, leading to higher effective charge separation for the CdS main catalyst. The facilitated electron transfer from CdS to Co-NG was further verified by the photocurrent measurements, Fig. 4(c). The photocurrents generated by CdS, NG/CdS and Co-NG/CdS photocatalysts were determined using Fe3+ (aq) as an electron shuttle and methanol as a hole scavenger. The photocurrent for CdS loaded with 0.5 wt.% Co-NG is significantly greater than that of CdS and NG/CdS, Fig. 4(c). The much greater photocurrents of Co-NG/CdS indicate that Co-NG cocatalyst can decrease the electron-hole recombination rate due to the electron separation and trapping effects that subsequently accelerate the interfacial electron transfer from Co-NG to Fe3+(aq) ions in the electrolyte solution. As has been pointed out, proton reduction during photocatalytic hydrogen production is an important step occurring on the cocatalyst surface.9 The efficiency of a Co-NG/CdS composite depends strongly on the proton adsorption, reduction and hydrogen molecule desorption occurring on the surface of Co-NG cocatalysts.8-9, 11 Fig. S4 shows that the prepared Co-NG cocatalyst exhibits an excellent electrocatalytic activity for hydrogen evolution reaction (HER) at a very small overpotential onset (ηonset, ~ 23 mV), beyond which the current density increases sharply. The overpotential, η, of Co-NG is about 210 mV at a current density of 10 mA cm-2. The corresponding Tafel slope of the Co-NG (126 mV decade-1) derived from a linear-sweep voltammogram (LSV) is significantly greater than that of the commercial Pt/C catalyst (32 mV decade-1). This result suggests that the HER process over Co-NG is based on the Volmer-Heyrovsky mechanism, which is controlled by hydrogen absorption on the active sites.4 The LSV characteristics of CdS, 0.5 wt.% NG/CdS and 0.5 wt.% Co-NG/CdS catalysts were detected in a 10 vol.% lactic acid aqueous solution. As shown in Fig. 4(d), the overpotentials corresponding to the 10 mA/cm2 cell current density are -0.64 V, -0.93 V and -1.1 V (vs. RHE) for 0.5 wt.% Co-NG/CdS, 0.5 wt.% NG/CdS and CdS, respectively. A lower overpotential at the same 9

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

current density for 0.5 wt.% Co-NG/CdS indicates a lower potential energy loss, and results in a higher photocatalytic activity for Co-NG/CdS catalyst rate during proton reduction reaction29. In summary, electron impedance reduction, higher photocurrent and lower overpotential for Co-NG cocatalyst have all substantiated that Co-NG is a highly effective cocatalyst for solar hydrogen production as it not only reduces the recombination rate of electron-hole pairs, but it is also a highly active proton reduction catalyst. Conclusion Single Co atom anchored nitrogen-doped graphene (Co-NG) is a highly effective and durable cocatalyst for visible light photocatalytic hydrogen production. Under identical conditions, the photocatalytic hydrogen production rate (1382 µmol/h) for 0.25 wt.% Co-NG loaded CdS photocatalyst (0.25 wt.% Co-NG/CdS) is 24.2 and 3.42 times greater than the rates for bare CdS and 0.25 wt.% NG/CdS photocatalysts, (404 µmol/h), respectively. This rate (1382 µmol/h) is 1.3 times greater than the greatest hydrogen production rate (1077 µmol/h) for 1.5 wt.% Pt nanoparticle loaded Pt/CdS photocatalyst. The quantum efficiency of 0.25 wt.% Co-NG/CdS can be as high as 50.5% at 420 nm light irradiation, the greatest reported quantum efficiency for non-noble metal cocatalysts. The results derived from this research have verified that single Co atoms dispersed in NG can serve as a high efficiency and long lifespan non-noble metal cocatalyst for photocatalytic hydrogen production via water splitting. This finding will attract new interest in more fundamental research in the fields of hydrogen adsorption energy, charge separation and proton reduction kinetics for non-noble metal single atoms in nitrogen doped graphene oxide. The success of all these research efforts will ultimately lead to the replacement of high cost precious metals in a variety of hydrogen related fields, including photocatalytic hydrogen production, water electrolysis and hydrocarbon hydrogenation.

Supporting Information The electrochemical and photoelectrochemical measurements and catalyst characterizations are summarized in detail in the Supporting Information.

Acknowledgments This work was financially supported by the National Natural Science Foundation of China (21103106, 21107069), Shanghai Key Project for Fundamental Research (13JC1402800), the “Dawn” Program of Shanghai Education Commission (11SG52) and Science and Technology 10

ACS Paragon Plus Environment

Page 10 of 17

Page 11 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Commission of Shanghai Municipality (14DZ2261000). The invaluable support of all the above funding agents is greatly appreciated.

References 1.

Chen, X.; Shen, S.; Guo, L.; Mao, S. S., Semiconductor-based Photocatalytic Hydrogen

Generation. Chem. Rev. 2010, 110 (11), 6503-6570. 2.

Zou, Z.; Ye, J.; Sayama, K.; Arakawa, H., Direct splitting of water under visible light irradiation

with an oxide semiconductor photocatalyst. Nature 2001, 414 (6864), 625-627. 3.

Zou, X.; Zhang, Y., Noble metal-free hydrogen evolution catalysts for water splitting. Chem. Soc.

Rev. 2015, 44 (15), 5148-5180. 4.

Qu, Y.; Medina, H.; Wang, S. W.; Wang, Y. C.; Chen, C. W.; Su, T. Y.; Manikandan, A.; Wang,

K.; Shih, Y. C.; Chang, J. W.; Kuo, H. C.; Lee, C. Y.; Lu, S. Y.; Shen, G.; Wang, Z. M.; Chueh, Y. L., Wafer Scale Phase-Engineered 1T- and 2H-MoSe2/Mo Core-Shell 3D-Hierarchical Nanostructures toward Efficient Electrocatalytic Hydrogen Evolution Reaction. Adv. Mater. 2016, 28 (44), 9831-9838. 5.

Wang, J.; Cui, W.; Liu, Q.; Xing, Z.; Asiri, A. M.; Sun, X., Recent Progress in Cobalt-Based

Heterogeneous Catalysts for Electrochemical Water Splitting. Adv. Mater. 2016, 28 (2), 215-230. 6.

Zhang, L. J.; Zheng, R.; Li, S.; Liu, B. K.; Wang, D. J.; Wang, L. L.; Xie, T. F., Enhanced

Photocatalytic H2 Generation on Cadmium Sulfide Nanorods with Cobalt Hydroxide as Cocatalyst and Insights into Their Photogenerated Charge Transfer Properties. ACS Appl. Mater. Inter. 2014, 6 (16), 13406-13412. 7.

Fei, H.; Dong, J.; Arellano-Jiménez, M. J.; Ye, G.; Dong Kim, N.; Samuel, E. L. G.; Peng, Z.;

Zhu, Z.; Qin, F.; Bao, J.; Yacaman, M. J.; Ajayan, P. M.; Chen, D.; Tour, J. M., Atomic cobalt on nitrogen-doped graphene for hydrogen generation. Nat. Commun. 2015, 6, 8668. 8.

Luo, M.; Lu, P.; Yao, W.; Huang, C.; Xu, Q.; Wu, Q.; Kuwahara, Y.; Yamashita, H., Shape and

Composition Effects on Photocatalytic Hydrogen Production for Pt-Pd Alloy Cocatalysts. ACS Appl. Mater. Inter. 2016, 8 (32), 20667-20674. 9.

Luo, M.; Yao, W.; Huang, C.; Wu, Q.; Xu, Q., Shape effects of Pt nanoparticles on hydrogen

production via Pt/CdS photocatalysts under visible light. J. Mater. Chem. A 2015, 3 (26), 13884-13891. 10. Hong, Y.; Shi, P.; Wang, P.; Yao, W., Improved photocatalytic activity of CdS/reduced graphene oxide (RGO) for H2 evolution by strengthening the connection between CdS and RGO sheets. Int. J. Hydrogen Energy 2015, 40 (22), 7045-7051. 11. Luo, M.; Hong, Y.; Yao, W.; Huang, C.; Xu, Q.; Wu, Q., Facile removal of polyvinylpyrrolidone (PVP) adsorbates from Pt alloy nanoparticles. J. Mater. Chem. A 2015, 3 (6), 2770-2775. 12. Fontelles-Carceller, O.; Muñoz-Batista, M. J.; Conesa, J. C.; Fernández-García, M.; Kubacka, A., UV and visible hydrogen photo-production using Pt promoted Nb-doped TiO2 photo-catalysts: Interpreting quantum efficiency. Appl. Catal. B: Environ. 2017, 216, 133-145. 13. Fontelles-Carceller, O.; Muñoz-Batista, M. J.; Rodríguez-Castellón, E.; Conesa, J. C.; Fernández-García, M.; Kubacka, A., Measuring and interpreting quantum efficiency for hydrogen photo-production using Pt-titania catalysts. J. Catal. 2017, 347, 157-169. 14. Escobedo Salas, S.; Serrano Rosales, B.; de Lasa, H., Quantum yield with platinum modified TiO2 photocatalyst for hydrogen production. Appl. Catal. B: Environ. 2013, 140-141, 523-536. 15. Guayaquil-Sosa, J. F.; Serrano-Rosales, B.; Valadés-Pelayo, P. J.; de Lasa, H., Photocatalytic 11

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

hydrogen production using mesoporous TiO2 doped with Pt. Appl. Catal. B: Environ. 2017, 211, 337-348. 16. Pan, Z.; Hisatomi, T.; Wang, Q.; Chen, S.; Iwase, A.; Nakabayashi, M.; Shibata, N.; Takata, T.; Katayama, M.; Minegishi, T., Photoreduced Graphene Oxide as a Conductive Binder to Improve the Water Splitting Activity of Photocatalyst Sheets. Adv. Funct. Mater. 2016, 26 (38), 7011-7019. 17. Wang, H.; Robinson, J. T.; Li, X.; Dai, H., Solvothermal reduction of chemically exfoliated graphene sheets. J. Am. Chem. Soc. 2009, 131 (29), 9910-9911. 18. Xu, Y.; Mo, Y.; Tian, J.; Wang, P.; Yu, H.; Yu, J., The synergistic effect of graphitic N and pyrrolic N for the enhanced photocatalytic performance of nitrogen-doped graphene/TiO2 nanocomposites. Appl. Catal. B: Environ. 2016, 181, 810-817. 19. Li, Y.; Zhao, X.; Zhang, P.; Ning, J.; Li, J.; Su, Z.; Wei, G., A facile fabrication of large-scale reduced graphene oxide–silver nanoparticle hybrid film as a highly active surface-enhanced Raman scattering substrate. J. Mater. Chem. C 2015, 3 (16), 4126-4133. 20. Nie, R.; Miao, M.; Du, W.; Shi, J.; Liu, Y.; Hou, Z., Selective hydrogenation of C C bond over N-doped reduced graphene oxides supported Pd catalyst. Appl. Catal. B: Environ. 2016, 180, 607-613. 21. Miao, X.; Pan, K.; Wang, G.; Liao, Y.; Wang, L.; Zhou, W.; Jiang, B.; Pan, Q.; Tian, G., Well˗Dispersed CoS Nanoparticles on a Functionalized Graphene Nanosheet Surface: A Counter Electrode of Dye˗Sensitized Solar Cells. Chem. ˗Eur. J 2014, 20 (2), 474-482. 22. Xu, J.; Wu, J.; Luo, L.; Chen, X.; Qin, H.; Dravid, V.; Mi, S.; Jia, C., Co3O4 nanocubes homogeneously assembled on few-layer graphene for high energy density lithium-ion batteries. J. Power Sources 2015, 274, 816-822. 23. Xue, Y.; Wu, B.; Jiang, L.; Guo, Y.; Huang, L.; Chen, J.; Tan, J.; Geng, D.; Luo, B.; Hu, W., Low temperature growth of highly nitrogen-doped single crystal graphene arrays by chemical vapor deposition. J. Am. Chem. Soc. 2012, 134 (27), 11060-11063. 24. Ferrandon, M.; Kropf, A. J.; Myers, D. J.; Artyushkova, K.; Kramm, U.; Bogdanoff, P.; Wu, G.; Johnston, C. M.; Zelenay, P., Multitechnique characterization of a polyaniline–iron–carbon oxygen reduction catalyst. J. Phys. Chem. C 2012, 116 (30), 16001-16013. 25. Wang, Y.; Nie, Y.; Ding, W.; Chen, S.; Xiong, K.; Qi, X.; Zhang, Y.; Wang, J.; Wei, Z., Unification of catalytic oxygen reduction and hydrogen evolution reactions: highly dispersive Co nanoparticles encapsulated inside Co and nitrogen co-doped carbon. Chem. Commun. 2015, 51 (43), 8942-8945. 26. Acharya, K. P.; Nguyen, H. M.; Paulite, M.; Piryatinski, A.; Zhang, J.; Casson, J. L.; Xu, H.; Htoon, H.; Hollingsworth, J. A., Elucidation of two giants: challenges to thick-shell synthesis in CdSe/ZnSe and ZnSe/CdS core/shell quantum dots. J. Am. Chem. Soc. 2015, 137 (11), 3755-3758. 27. Yu, J.; Fan, J.; Cheng, B., Dye-sensitized solar cells based on anatase TiO2 hollow spheres/carbon nanotube composite films. J. Power Sources 2011, 196 (18), 7891-7898. 28. Klahr, B.; Gimenez, S.; Fabregat-Santiago, F.; Hamann, T.; Bisquert, J., Water oxidation at hematite photoelectrodes: the role of surface states. J. Am. Chem. Soc. 2012, 134 (9), 4294-4302. 29. Li, L.; Deng, Z.; Yu, L.; Lin, Z.; Wang, W.; Yang, G., Amorphous transitional metal borides as substitutes for Pt cocatalysts for photocatalytic water splitting. Nano Energy 2016, 27, 103-113.

12

ACS Paragon Plus Environment

Page 12 of 17

Page 13 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

(A) (A)

(B)

Fig. 1. TEM image of Co-NG nanosheet on a lacey carbon TEM grid (A) with a 100 nm scale bar, HAADF-STEM image (B) of well-dispersed Co-single atoms in nitrogen doped graphene (NG) with a 2 nm scale bar. The red arrow in image (B) points to a single Co atom.

(C)

(B)

13

ACS Paragon Plus Environment

(D)

ACS Applied Materials & Interfaces

(b)

GO Co-NG NG

3462

1639

Intensity(a.u.)

Intensity (a. u.)

GO NG Co-NG

1076

D G

1394

(a)

500

1000

1500

2000

2500

3000

4000

3500

3000

2500

2000

1500

1000

500

-1

Wavenumber (cm )

Raman Shift (cm ) -1

(c)

(d)

Co 2p

N1s Pyridinic/N-Co

Intensity (a.u.)

Intensity (a.u.)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 17

805

800

795

790

785

780

Pyrrolic Quaternary

N-oxide

408

775

406

404

402

400

398

396

394

B.E.(eV)

B.E. (eV)

Fig. 2. Raman (a) and FTIR (b) spectra of synthesized graphene oxide (GO), nitrogen doped graphene oxide (NG) and Co single atoms dispersed NG (Co-NG). (c) Co 2p and (d) N 1s XPS spectra of Co-NG.

14

ACS Paragon Plus Environment

Page 15 of 17

0 0

(c)

1

2

3

4

400 200 0

5

Co-NG/CdS Photocatalysts

Irradiation Time (h)

(d)

1600

7000

Evacuated

1400 1200

1.5 wt.% 2.0 wt.% 0.5 wt. % 1.0 wt. %

1000 800 600

0.25 wt. %

400

1 wt. %

1000

600

0.5 wt. %

2000

800

0.375 wt.%

3000

1000

0.185 wt.%

4000

1200

0.125 wt.%

5000

1400

Bare CdS

6000

H2 Production (µ mol)

H2 Production (µmol)

7000

H2 Production Rate (µ mol/h)

Bare CdS 0.25 wt.% NG/CdS 0.25 wt.% Co-NG/CdS 0.25 wt.% Pt/CdS

0.25 wt. %

(b) 1600

(a) 8000

H2 Production Rate (µ mol/h)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Evacuated

6000

First run

Second run

Third run

5000 4000 3000 2000 1000

200 Bare CdS

0

0 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Irradiation Time / h

Pt/CdS photocatalysts

Fig. 3. (a) Photocatalytic H2 evolution over 0.25 wt.% Co-NG/CdS, 0.25 wt.% NG/CdS, 0.25 wt.% Pt/CdS and bare CdS photocatalysts. (b) Rates of hydrogen production over various Co-NG loaded CdS (Co-NG/CdS) photocatalysts. (c) Activities of Pt/CdS with various Pt loadings. (d) Catalyst lifespan tests over 0.25 wt.% Co-NG/CdS photocatalyst. (Catalyst: 0.1 g; Photolyte: 100 ml 1.0 M (NH3)2SO3 aqueous solution; Light source: Xe light with 420 cutoff filter).

15

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

Bare CdS 0.5 wt.% NG/CdS 0.5 wt.% Co-NG/CdS

140

(b)

120

-Z'' (kΩ )

Intensity (a.u.)

(a)

100 80 60 40

Bare CdS 0.5 wt.% NG/CdS 0.5 wt.% Co-NG/CdS

20 0

480

510

540

570

600

630

0

20

40

60

80

100

120

Z' (kΩ)

Wavelength (nm) 6

Co-NG 0.5 wt.% /CdS -2

(c)

Current density (mA cm )

Photocurrent density(µAcm-2)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 17

4

0.5 wt. % Co-NG/CdS

2

0.5 wt. % NG/CdS

0

Bare CdS

0

(d)

-5 -0.93

-1.1

-10

-0.64

-15 -20 -25

0.5 wt.% NG/CdS 0.5 wt.% Co-NG/CdS Bare CdS

-30 -35

300

600

900

1200

1500

-1.5

-1.0

-0.5

0.0

0.5

Overpotential (V vs RHE )

Time (sec)

Fig. 4. Photoelectrochemical measurements for CdS, 0.5 wt.% NG/CdS and 0.5 wt.% Co-NG/CdS photocatalysts: (a) photoluminescence spectra, (b) electrochemical impedance spectra (Nyquist plots) and (d) linear-sweep voltammogram (Electrolyte: 10 vol.% lactic acid aqueous solution; Scan rate: 2.0 mV s-1)

16

ACS Paragon Plus Environment

Page 17 of 17

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Table of Contents (TOC) graphic

17

ACS Paragon Plus Environment