Xiaotian Wang,a Mingyun Zhu,a Wanlin Fu,a Chengqian Huang,a Qing Gu,a Tingying Helen Zeng,b Yunqian Dai*a and Yueming Sun*a

a School of Chemistry and Chemical Engineering, Southeast University, Nanjing,  Jiangsu 211189, P. R. China  Email: daiy@seu.edu.cn and sun@seu.edu.cn  bAcademy for Advanced Research and Development, One Kendall Square,  Cambridge, MA 02135, US.

Abstract  

This work presents an ultrafast and low-cost method to prepare reduced graphene  oxide (RGO) by photocatalytic reduction of graphene oxide (GO) sheets in the  assistance of dumbbell-like Au nanorods. Au nanoparticles are effectively loaded on  the surface of GO sheets by electrostatic interaction and serve as the initiating points  for the splitting of GO sheets. The existence of Au nanoparticles tremendously  accelerates the process of reducing reaction as well, in which the reduction completes  in 15 min, ascribing to hot electrons generated by the localized surface plasmonic  resonance (LSPR) property of Au nanorods. Interestedly, in this process, the function  of Au nanorods is not limited in prompting the reduction. The more important merit is  that they serve as pore generators which help to cut the whole GO sheets into small  pieces. This photocatalytic method of RGO preparation provides a novel strategy in  this fields and creates various possibilities for future applications.

Introduction

Graphene, a two-dimensional (2D) monolayer of atomic-scaled carbon atoms  consisted lattice, has attracted enormous attention and research interest,1 due to its  superior electronic property,2 excellent thermal stability3 and unique mechanical  flexibility.4-5 Efficient methods to prepare graphene has been extensively explored,  which are classified into two main categories, known as physical methods and  chemical methods. Physical methods include chemical vapor deposition,6 micromechanical exfoliation7 and epitaxial growth,8-9 while chemical methods are  mainly focused on the chemical reduction of graphene oxide.10 Solvothermal  reduction11 is one of the most widely used chemical method to prepare Graphene, but  endures the drawbacks of high temperature process. Other approaches using  hydrazine as the reducing agent to synthesize reduced graphene oxide,12 a method  which is highly effective but also suffers from the disadvantages of high toxicity.  Thus, more eco-friendly reducing agents were developed including vitamin C and  chitosan.13-14 However, efficiency of these process is relatively low and it always  takes a long time for them to make a complete reduction of GO sheets. Besides, all of  the methods above are unable to allow a controlled process to regulate the lateral size  of RGO sheets in this reaction, which is highly important for multiple applications.  Therefore, efforts should be made to establish a fast, controlled and low-toxic process  for the preparation of graphene from chemical reduction.

Metallic nanoparticles with plasmonic property are always among the popular  research interests for the last decades, due to their wide range absorption in the  visible-light and near-infrared regions.15-16 The localized surface plasmonic resonance  (LSPR) of these nanoparticles, normally Au and Ag, occurs from photo-induced  collective oscillation of conduction band electrons in these plasmonic structures.  These electrons are typically referred to as “hot electrons” which is generated from  the non-radiative decay of plasmons,17 for the reason that they are not in thermal  equilibrium with the atoms in the metallic nanoparticles. The generation of these hot  electrons have proved to be a great promotion for catalytic reactions because they can either participate into the reaction or produce heat by self-collision to accelerate the  reaction process.18

Herein, in this present study, we utilized the surface modified dumbbell-like Au  nanorods to serve as photocatalyst for GO sheets reduction. On one hand, hot  electrons generated in Au nanorods under the irradiation of visible light tremendously  accelerate the process of GO reduction. More importantly, these Au nanorods, which  tightly anchored onto the surface of GO sheets, keep etching the GO surface and  creates lots of pores in the presence of reducing agents and visible light irradiation.  These generated pores, whose size are in agreement with the size of Au nanorods,  work as the initiating points for the continuous cracking and splitting of the GO sheets,  which finally cuts the whole sheets into small pieces. We choose to use rod-like Au  particles instead of conventional Au nanospheres not only because they exhibits better  catalytic ability but more importantly, for the reason that the holes they create are  much more discernable in TEM images, which makes it possible to determine the  cracking process and mechanism. In the assistance of Au nanorods, the process of GO  reduction is ultrafast, complete and efficient. Moreover, the pore generating and  lateral-size-controlling ability of Au nanorods towards 2-dimensional GO sheets is  significantly promising because this strategy of graphene structure designing can be  applied in many useful applications such as catalytic reaction19 and DNA detect.20-21

Experimental section

The original Au nanorods was synthesized by the seed-mediated method  according to Murphy’s group. Typically, a brown-yellow seed solution was first  obtained by adding 0.6 mL of ice-cold 0.01 M NaBH4 into a 10 mL of mixed solution  containing 0.1 M CTAB and 0.25 mM HAuCl4. The seed solution was kept steady for  at least 2 h for further use. To synthesize desired Au nanorods, 80 µL of 0.01 M  AgNO3 solution, 65 µL of 0.1 M ascorbic acid (AA) and 60 µL of the as-prepared  seed solution were sequentially added into a 10 mL of 0.1 M CTAB and 0.5 mM HAuCl4 solution. The obtained product was centrifuged twice using 18.2 MΩ cm-1 ultrapure water to remove excessive CTAB and unreacted reagent. The obtained Au  nanorods were re-dispersed in ultrapure water in a fixed ratio.  

Synthesis of dumbbell-like Au nanorods was based on the obtained straight Au  nanorods without centrifuging purification. For example, 46.4 µL of 10 g/ml HAuCl4 solution was first added into the uncentrifuged Au nanorods solution. The mixed  solution was then kept undisturbed for at least 30 min after adding 80 µL of 0.1 M AA  dropwise with gentle stirring. The obtained solution was then centrifuged using the  same method mentioned above.

2. Photo-assisted synthesis of RGO catalyzed by gold nanorods  

GO was prepared by a modified Hummers method which has been discussed in  our previous works. The obtained GO was diluted into 1.5 mg/mL in water solution  for further use. For a typical run of reaction, 0.15 mL~1.5 mL of GO solution and a  certain amount of Au nanorods or surface-modified Au nanorods solution were  initially mixed together to achieve the efficient adhesion of Au nanorods onto the  surface of GO. The mixed solution was then added into 9 mL of water followed by  injecting 0.1 mL of 90 mg/mL NaBH4 solution to initiate the reaction. the reaction  was conducted under the irradiation of a Xe lamp (CEL-HXF300/CEL-HXUV300,  CEAULIGHT Co., Beijing, China) equipped with a UV-cut filter (420 nm-780 nm).  The intensity of visible light (UV-cut filter) was measured to be 250 mW cm-2 .

Samples were extracted at different time intervals to investigate the evolution of  the reaction. Small amount of 1 M HCl solution were injected into the samples to  terminate further reduction reaction. The samples were subsequently washed by water  until the pH recovered to 7. For the final product, it was washed by centrifugation for  three times to eliminate the remaining reagents in the solution.

3. Characterization  

XRD pattern was carried out by a Bruker D8 Advanced diffractometer using  Cu-Kα (1.5406 Å) radiation. TEM images were collected using a transmission  electron microscope (Tecnai G2 T20, FEI) operated at 200 k. The Au mass in the  solution was determined by Inductively Coupled Plasma Optical Emission  Spectrometer (ICP-OES) (Optima 7300DV, Perkin Elmer Corporation). Scanning  electron microscope (SEM) images were captured with a field emission microscope  (FEI Sirion 200) at an accelerating voltage of 5.0 kV. The height profiles of GO and  RGO were collected by atomic force microscope (AFM, Veeco) in tapping mode.  Samples at different reaction times were taken and characterized by UV-vis  spectrophotometer (Cary-60).

Result and discussion

Dumbbell-like Au nanorods was synthesized by conducting a second step of Au  ions reduction on the straight rod-like nanorods, which is obtained based on a typical  seed-mediated method.22 It can be found in Fig. S1 that the absorption peak of  dumbbell-like Au nanorods underwent a blue shift due to the shrink of aspect ratio  (Au nanorods: 3.89, dumbbell-like Au nanorods: 2.79). The model reactions of  4-nitrophenol reduction were launched in the presence of NaBH4 to determine the  catalytic ability of straight rods and dumbbells. The dumbbell-like Au nanorods  undoubtedly exhibited much better catalytic property, the reaction rate constant of  which is 5.5 times higher than that of straight rods (Fig. S2). Therefore,  Dumbbell-like Au nanorods was utilized instead of the original straight rods for their  considerably enhanced catalytic ability and LSPR property.23 Fig. 1A is the typical  TEM image of the dumbbell-like Au nanorods in which the length and width of  nanorods exhibit to be 41.43 nm and 14.84 nm, respectively. The high-resolution  TEM image of the enlarged end of dumbbell-like Au nanorods clearly reveals two  distinct kinds of lattice fringes with lattice distance of 2.0 Å and 2.3 Å, which are  assigned to Au (200) and Au (111), respectively (Fig. 1B). GO sheets are  demonstrated to be an excellent carrior for cetyltrimethylammoniumbromide  (CTAB)-capped Gold nanocraystals,24 the fact of which are verified in our research by the phenomenon that Au nanorods can be well dispersed in GO aqueous solution. The  yellowish GO solution will turn to brown after the addition of Au nanorods and the  extent of color evolution is depended on the amount of Au nanorods. Fig. 1C is the  transmission electron microcopy images of Au nanorods loaded GO sheets. Au  nanoparticles are tightly and uniformly distributed on the surface of GO. The  mechanism of effective binding between Au nanoparticle and GO is believed to be  caused by electrostatics. GO sheets in water is negatively charged (ζ potential: -30.2  mV) due to the ionization of their carboxylic group and phenolic hydroxyl groups,  which we have fully discussed in our previous study. 25 CTAB, as a cationic surfactant,  always make the surface of particles positively charged when serving as the capping  agents. The ζ potential of CTAB-capped Au nanoparticles was measured to be 34.6  mV in aqueous solution. Therefore, it is possible for GO sheets to attract  CTAB-capped gold nanoparticles as soon as they are injected into the GO solution.  The electrostatic attraction establishes a stable binding between GO sheets and Au  nanoparticles for further applications.  

Previous work by Sun’s group26 has reported that RGO can be obtained via  LSPR-induced photocatalytic reduction of GO in the assistance of considerable  amount of Ag nanoparticles. Interestedly, it is observed that in our present work, only  tiny amount of noble metallic Au nanoparticles can stimulate the GO sheets to  undergo a rapid reaction process in aqueous solution under the irradiation of visible  light, initiated by a minor amount of reducing agents (NaBH4). The general  mechanism is that the existence of metal in the reaction accelerates the hydrolysis of  NaBH4, creating hydrogen atoms to reduce the GO sheets.27 In a typical reaction, a  mixed solution containing 0.2 mg of GO and 6 µg of dumbbell-like Au nanorods were  firstly added to a 9 mL of water solution. Certain amount of NaBH4 solution was  added which made the concentration of NaBH4 to be 50 mM. The intensity of the  incident visible light was set to be 250 mW cm-2. The color of the solution changed  from nearly transparent to slight yellow in 1 min, and then to brown in 4 min.  Suspended black particles were observed after around 5 min, the amount of which enlarges over time, due to the hydrophobic nature of reduced graphene oxide. This is  caused by the reduction of oxygen-containing groups and the splitting of large GO  sheets, which will be demonstrated and discussed in details below. The optical images  of the product at each time interval are shown in Fig. 1E, which illustrates the  evolution of the reaction. The reaction product was monitored by UV-vis spectroscopy  at designated time intervals (Fig. 1D). The pristine sample at 0 min shows a typical  absorption peak of GO at 233.5 nm, which rapidly red shifts with the reaction going  on. The red-shift of absorption peak to > 260 nm and the initial increase in absorption  intensity indicates that GO has been reduced to RGO and the aromatic structure and  electronic conjugations of GO have been restored.28 The distance of red-shift is  enlarged with the development of the reaction, which indicates the increasing extent  of GO reduction.29 The decrease in absorption intensity after 6.5 min is due to the  assembly of RGO in aqueous solution, the extent of which increases as reaction goes  on. The assembly of RGO occurs because more RGO is produced over time with the  drastic decrease of oxygen-containing group and shrink of the size of RGO. It can be  concluded from UV-vis spectra and optical images that the reaction completely  finishes before 15 min, which is an ultrafast and easy process comparing to all kinds  of previous methods.  

To confirm the efficient reduction of GO, XPS was used to investigate the  evolution of oxidation states for C. Fig. 2A shows C1s peaks of GO at different  reaction times. Typically, 4 identical peaks at 284.6, 286.8, 288.4 and 289.0 eV were  exhibited, which were assigned to C-C/C=C bond in graphitic carbon, C-O in  hydroxyl and epoxy groups, C=O bond in carbonyl groups and O-C=O bond in  carboxyl groups, respectively.30 The peak area ratios of oxygen-containing groups to  C-C/C=C are 1.49, 0.49 and 0.45 for pristine GO (a), reducing product at 6.5 min (b)  and reducing product at 15 min (c), respectively. It is evident that comparing with  pristine GO, the intensity of oxygen-related peaks decrease drastically for the final  product at 15 min, indicating the effective removal (67 %) of oxygen-containing  groups in GO. This further confirms the reduction process of GO which yields RGO after the reaction. The C1s peak of the intermediate product at 6.5 min also shows that  the reaction process is fast and efficient, by the fact that most part (96 %) of the  reduced oxygen-containing groups have been successfully removed before 6.5 min.  Raman spectroscopy was also utilized to investigate the intrinsic structure of GO and  its final reduction product (Fig. 2B). Two distinct bands are detected at ~1350 cm-1  and ~1588 cm-1, which are known as D band and G band for carbon-based materials.  It has been reported that D band is related to the disordered carbon structure induced  by lattice defects (the E2g mode of sp2 carbon atoms) and G band is assigned to  ordered structure (the symmetry A1g mode).31 The intensity ratios expressed as ID/IG are measured to be 0.78 for pristine GO and 1.18 for the final reducing product. The  remarkably enhancement of ID/IG also indicates the formation of new graphitic  product. Since the Raman ID/IG is inversely proportional to the average size of the sp2 domains, the increase of ID/IG indicates that smaller in-plane sp2 domains are formed  during the reduction process of GO.12-13 This result demonstrates the effective  reduction of GO in our catalytic system. Furthermore, the value of ID/IG is particularly  large, comparing to our previous work which used UV light irradiation and  hydrothermal method.25 ID/IG of RGO produced using these ways were among  ~0.91-0.96, which indicates that the current method utilizing Au nanorods as the  photocatalyst creates much more defects and disordered structure on the obtained  RGO. To investigate the photothermal effect of Au nanorods in the reaction,  temperature evolution of aqueous solution on different conditions under visible light  irradiation were monitored (Fig. 2C). Comparing to pure water solution, the addition  of 2 mg of RGO brings little effect on the elevation of temperature, while the addition  of 6 µg of Au nanorods make a considerable 3.5 °C of enhancement in the final  temperature. According the method based on heat balance equation which has been  described in our previous work,18 the photothermal conversion efficiency of the  system with Au is 20.1 %, a remarkable value which indicates outstanding  photothermal effect of Au. The temperature evolution of the reaction process was also  investigate to confirm this photothermal effect of Au nanorods. Excluding the heat  released by the reaction and RGO which is only able to increase the temperature  around 3.2 °C according to experimental result and calculation, the overall  enhancement of 7 °C demonstrates that the existence of Au nanorods exerts superb  photothermal effect on the reaction system which is vital to the whole process.  

To further investigate the influence of various factors in the reaction, some  controlled experiments were conducted for comparison. The reducing reaction  without Au catalyst under visible light irradiation (Fig. S3A) was much slower, which  was incomplete at 15 min with removal of only 50 % of oxygen-containing groups  (Fig. S4), comparing to that of 67 % when Au is added (Fig. 2A). This conclusion is  consistent with optical images, which show that the color of the solution only turns  dark yellow in the reduction process (Fig. S3C). It demonstrates that NaBH4 itself has  limited reducing effect towards GO without the assistance of Au catalyst. And such a  reaction condition cannot effectively split the GO sheets and reduce the size of  obtained RGO sheets, due to the fact that no black particles generate in this process.  Also, we investigated the effect of visible light irradiation on the reaction. When Au  assisted reaction was conducted without visible light irradiation, the reaction rate  seems to be only slightly slower than that with visible light irradiation, according to  the optical images (Fig. S3D). However, the absorption peaks at 260-270 nm on  UV-vis spectra, which are assigned to RGO, are almost indiscernible (Fig. S3B), in  contrary to the reducing product obtained using visible light, the absorption peak of  which is quite notable at the region of 260-270 nm (Fig. 1D). The fact that no obvious  absorption peaks can be found in that region reveals that the extent of reducing  without visible light irradiation is far more incomplete, comparing to the condition in  photocatalysis which exhibits a solid evidence of RGO’s existence. To further verify  this assumption, some experimental parameters were tuned to have a more  comprehensive understanding and strong support for the influence of visible light  irradiation. Firstly, concentration of Au catalyst was adjusted to 1/10 of its original  concentration. It can be easily concluded that the reaction with light irradiation is  much faster than that in dark, owning to the different increasing rate of absorption  intensity (Fig. 3A, B). The similar phenomenon was observed when the reducing  agent was reduced to 1/5 of its original amount (Fig. 3C, D). In both of these  situations, the absorption peaks of RGO at 260-270 nm produced by the methods with  visible light irradiation are evident, while the peaks without visible light irradiation  are almost negligible, as well. Therefore, it can be confirmed that the LSPR property  of Au nanorods induced by visible light irradiation plays a major part in the reducing  process. Besides, the influence of transverse-mode and longitudinal-mode of LSPR  was studied separately by using a 520 nm filter or an above-600 nm filter on the Xe  lamp. The reactions take place normally but are both slower than the original reaction  in Fig 1D which can be found by the fact that the absorption intensity decreases after  10.5 min, comparing to 6.5 min for original reaction (Fig. S5 and Fig. 1). Thus, both  of the transverse and longitudinal modes contribute to the reaction and black particles  are produced which indicates the splitting of GO sheets in both conditions. We also  used conventional spherical Au nanoparticles to catalyze the reduction for comparison.  The size of the spherical particles are almost equal to dumbbell-like Au nanorods (Fig.  S6). The result in Fig. S7 shows that in the assistance of spherical Au nanoparticles,  the reduction process is fairly slow. The reaction solution just turns yellow after 2 h  (inset in Fig. S7). And it becomes dark yellow even after 24 h, which still has no  black particles in the solution. It means that the GO sheets can be reduced but they  can hardly be split in this process. Therefore, the fact that the catalytic property of  traditional Au nanoparticles is relatively poor is one of the reasons why we choose to  use dumbbell-like Au nanorods.  

To investigate the intrinsic mechanism of the reducing process, intermediate  product of the reducing reaction was obtained by adding tiny amount of dilute HCl to  terminate the reaction. The HCl can react with the excessive NaBH4, preventing  further reaction on graphene sheets. This intermediate product is believed to be an  indicator of the developing process of reaction. The morphology of the intermediate  product was investigated by transmission electron microscopy. Lots of Au nanorods  related pores on the sheets are observed in these images. One typical situation is that a lot of tiny pores with the diameter between 5 nm to 25 nm are found on the product at  1 min, in which Au has been removed after being treated by chloroazotic acid (Fig.  4A). It is certain that these pores are previously occupied by the end side of  dumbbell-like Au nanorods, because the distances between two nearby pores are  measured to be between 28.6-49.2 nm, which is corresponded to the length of Au  nanorods, and the size of the pores are consistent with the size of the end part of  dumbbell-like Au nanorods as well. Interestedly, in some images of product at 2.5 min,  we not only observed that the nanorods were embedded into the GO sheets, but also  found lots of pores exposed outside without any nanorods on it (Fig. 4B). The  phenomenon that the shape of pores is approximately the same with the shape of  nanorods reveals the high possibility that these exposed pores were previously  occupied by nanorods, which moved afterwards. In addition, the lateral size of  graphene sheets was also investigate by atomic force microscope (AFM). Height  profile in Fig. 4E shows that the pristine GO sheets has a thickness of around 0.4 nm,  confirming that they are one-layer sheets with large size (Fig. 4C) but the reduction  process in the presence of Au nanorods cut the whole pieces of GO sheets into small  pieces (Fig. 4D,F and Fig. S8). Moreover, some parts of the sheets are stacked  together in this process, making some high thickness area in the sheets. We also  observed the phenomenon from the TEM image of the final reduction product that  most part of the Au nanorods disappear on the surface of GO and very little amount of  Au nanorods still keeps attaching on the surface (Fig. S9). On the basis of these  observations above, the mechanism of photocatalytic reduction of GO is proposed and  illustrated in Fig. 5. The Au nanoparticles have been efficiently attached on the  surface of a GO sheet by electrostatic attraction. (i) The cracking/reduction of the GO  sheet initially takes place in the area where the end parts of dumbbell-like Au  nanorods locate. Photo-induced thermal effect by the LSPR property of surface  modified Au nanorods is supposed to be the main reason for the selective reduction on  the GO sheets. Firstly, hot electrons are generated by non-irradiative decay of  plasmons produced by LSPR of Au nanoparticles. The work functions of GO and Au  nanorods are both 4.70 eV, according to literature reports.32-33 Thus, it is possible for hot electrons to transfer from Au nanoparticles onto the surface area of GO sheets  which closely sticks to Au nanoparticles, and to participate into the reduction process  of these area in GO sheets by assisting in generating hydrogen atoms in the hydrolysis  of NaBH4. Some literature works have found the influence of hot electrons in catalytic  reactions such as water splitting33 and the reduction of nitrobenzene34. Moreover,  Wu’s group26 have researched in a similar situation in which photo-generated hot  electrons produced by Ag nanoparticles are injected into GO to participate in the  reduction. The possible involvement of these hot electrons in the reaction may also  lead to the preference of reduction initiated in the area of GO sheets which are  adhered to Au nanoparticles. Also, most of these hot electrons transfer their energy to  heat by electron-electron or electron-phonon collisions, making the Au nanoparticles  become heating points which tremendously accelerate the reduction rate of GO under  them. The reason why the cracking takes place right at the end part of nanorods (Fig.  4A) is because these areas with edges and corners have much stronger intensity of  LSPR property.35 The temperature of the dumbbell-like Au nanorods, especially in the  end parts, would become extremely high after light irradiation.36-38 The surface of GO  attached by Au nanorods begins to crack due to this partial high temperature, the  mechanism of which is similar with the traditional hydrothermal method that creates  the cracking in GO sheets by high temperature treatment. Thus, it can be concluded  that Au nanorods work as pore generators which initiate and accelerate the reaction.  (ii) The “destructed” area in the GO sheet is further expanded for the same reason in  step (i). The attraction between Au nanoparticles and GO in that area is consequently  weakened, because Au nanoparticles are losing contact with GO while the cracking  develops. Eventually, Au nanoparticles exfoliate from the surface of GO sheet. (iii)  Some of the exfoliated Au nanoparticles are re-located on the surface of GO sheet  again by electrostatic attraction. And the cracking takes place again as is described in  step (i). (iv) The cracking area expands and dislocates the Au nanoparticles once more.  The re-adhering and re-exfoliating of Au nanoparticles on GO sheet may occur  repeatedly until most of oxygen containing group in GO sheets is reduced, which  make it more and more difficult for Au nanoparticles to be adhered onto the surface of  GO, due to the severe decline in electrostatic interaction. Some enlarged pores with  the diameter between 100 nm to 200 nm in Fig. S8 verifies the existence of pore  expanding process. (v) The edge of the destructed areas induced by dislocated Au  nanoparticles gradually extends and develops in this process, and finally cut the whole  piece of GO sheet into small pieces of RGO. The color evolution of GO sheets from  yellow to black indicates the gradual reducing of oxygen-containing groups in the  whole procedure. Moreover, the fact that most of the oxygen-containing groups are  reduced tremendously weakens the electrostatic attraction between Au nanorods and  graphene sheet, leading to the result that little Au nanorods are still able to locate on  the surface of graphene sheets. Dumbbell-like Au nanorods, in this reduction process,  serve as scissors which cut the large pieces of GO sheets into small pieces, while  making promotion on the reducing of GO sheets as well. The fact that the obtained  RGO possess much more defects and disordered structures by this method, according  to the high ID/IG value, is also a substantial support for this mechanism. In a word, this  photocatalytic method demonstrates to be an ultrafast and efficient approach to obtain  RGO. More impressively, the pore-generating nature of Au nanorods in this process  provides very promising prospect for creating size-controlled pores in GO sheets and  precise control of lateral size in graphene materials, the property of which is important  in various applications.  

Conclusions

In summary, an ultrafast and efficient method to obtain RGO from GO sheets was  established in the assistance of dumbbell-like Au nanorods under the irradiation of  visible light. Oxygen-related group in the carbon structure is extensively eliminated in  the obtained RGO, indicating an extraordinary reducing effect. The existence of Au  nanoparticles demonstrates to be a key point in the reducing reaction. The LSPR  property of Au nanoparticles in this present work makes themselves serve as pore  generators on the surface of GO sheets, which not only accelerates the progress of  reduction, but help to cut the large pieces of GO sheets into small pieces as well, due  to their excellent photothermal and photoelectrical property. This novel strategy of  graphene synthesis provides an ultrafast and low-cost method to obtain RGO, which  is promising for further applications due to its potential possibility of lateral size  control and size-controlled pore generation on GO sheets.

This work was financially supported by the National Basic Research Program  (973 program, 2013CB932902), the National Natural Science Foundation of China  (21201034, 21173042 and 21310102005) and the Science and Technology Support  Program (Industry) Project of Jiangsu Province (BE 2013118).