Hong-Suk Shin‡
ac,
Ki Woong Kim‡b,
Yong-goo Kangc,
Sung Myungb,
Jong Seung Kimc,
Ki-Seok Anb,
Ill Young Lee*a and
Sun Sook Lee*b
aEco-Friendly New Materials Research Center, Korea Research Institute of Chemical Technology, 141, Gajeong-ro, Yuseong-gu, Daejeon, Republic of Korea 34114. E-mail: iylee@krict.re.kr
bThin Film Materials Research Center Korea Research Institute of Chemical Technology, 141, Gajeong-ro, Yuseong-gu, Daejeon, Republic of Korea 34114. E-mail: sunsukl@krict.re.kr
cDepartment of Chemistry, Korea University, 145, Anam-ro, Seongbuk-gu, Seoul, Republic of Korea 02841
First published on 9th February 2016
We developed a one-flask method for the thermal reduction of graphene oxide (GO) with triphenylphosphine dihalide (Ph3PX2) at 180 °C. Our approach offers a potential to cost-effective mass-production of graphene nanosheets under mild and environmentally friendly conditions and to avoid the use of strong acids or reducing agents. Significantly, this reduced graphene oxide (rGO) by utilizing a Ph3PX2 reductant has a C/O ratio higher than 15 and an electrical conductivity of 400 S cm−1, which indicate that this synthetic method allows us to achieve graphene nanosheets with high quality when comparing with previous reduction methods.
In the present study, triphenylphosphine dihalide (Ph3PX2) was used to reduce GO nanosheets by heating GO with Ph3PBr2 at 180 °C for 1 h. First, GO was prepared from natural graphite according to the modified Hummers method.52,53 Natural graphite flakes were oxidized into graphite oxide using NaNO3, H2SO4, KMnO4, and H2O2. The as-obtained graphite oxide was washed several times by using centrifugation. The product was then dispersed in distilled water and the exfoliated GO sheets were subjected to ultra-sonication. The crude graphite oxide was washed several times with 10% HCl solution during the centrifugation. The product was then washed with deionized water until the pH level was 4 and the exfoliated GO sheets were dried in vacuum oven at 40 °C for 12 h. Although commercially available Ph3PBr2 (6 mmol) could be used for the reduction of GO, PPh3 was first dissolved in nitrogen-degassed acetonitrile (8 ml), and bromine (6 mmol) was added dropwised into the reaction mixture for 10 min at 0 °C for getting fresh Ph3PBr2. GO (100 mg) suspended in nitrogen-degassed acetonitrile was added to the Ph3PBr2 in acetonitrile solution, and acetonitrile and bromine species in reaction mixture slowly distilled out at 90 °C under atmospheric pressure. The resulting concentrated residue was heated at 180 °C for 1 h under a nitrogen atmosphere. After cooling down to room temperature, the resulting solid was dissolved in chloroform and stirred for 30 min. The rGO dispersion was filtered through PTFE membranes and filter paper and then washed successively with methanol, ethanol, and dichloromethane. Finally, the reduced product was dried overnight in a vacuum oven at 40 °C, and rGO nanosheets were collected as a black filter cake for further use.
Our proposed mechanism to reduce GO to rGO is illustrated in Fig. 1. Generally, the conversion of acid,49 cyclic ester,54 epoxide50 and aryl alcohol51,55 functional group into each bromine substituent has been extensively studied and mechanism carefully proposed. The reaction of acid, cyclic ester, aryl alcohol and epoxide group on GO with Ph3PBr2 may give ionic phosphonium intermediate 1 followed by substitution by bromide ion to proceed by forming anticipated brominated rGO intermediate 2 with the generation of triphenylphosphine oxide. After that, rGO can be formed by reductive debromination by generating triphenylphosphine from Ph3PBr2 by thermal heating and further aromatization of 2 which mechanism would be similar to debromination at activated positions such as fullerene or α-bromoketone by triphenylphosphine as mechanism proposed in the literatures.56–59 In this reduction procedures, we could isolate and purify triphenylphosphine oxide (mp = 150–157 °C) as a side product of Appel conditions48 from the washing organic solvents of rGO. Significantly, since this synthetic method is a cost effective and scalable procedure, to apply this nanosheet to various photonic devices such as solar cells, light-emitting diodes photodetectors is currently under investigation.
Fourier transform infrared spectroscopy (FTIR) was used to determine the functional groups on GO and rGO (Fig. 2a).26,28,33,36 In the FTIR spectrum of GO, strong bands due to CO and C–O stretching vibrations in COOH groups were present at 1727 and 1045 cm−1, respectively. The strong band at 1620 cm−1 was assigned to the vibration of adsorbed water molecules and skeletal vibrations of the graphene sheets. The broad band at 3358 cm−1, which is particularly visible in unfunctionalized GO, corresponded to O–H stretching vibrations of carboxylic groups and adsorbed water molecules.35,39,60–63 Bands at 1223 cm−1 and 1040 cm−1 corresponded to the epoxy CO and alkoxy CO stretching vibrations, respectively. In the rGO spectrum, the bands at 3358 cm−1 (OH), 1223 cm−1 (epoxy), and 1040 cm−1 (alkoxy) became dramatically small. Only small peaks at 1727 cm−1 and 1045 cm−1 due to C
O and C–O were observed after the reduction reaction.
We also used thermogravimetric analysis (TGA) to confirm the thermal stability of GO and rGO reduced with Ph3PBr2 (Fig. 2b). Here, GO and rGO were heated in a nitrogen atmosphere from room temperature to 900 °C at a rate of 10 °C min−1. The weight of the GO decreased at temperatures higher than 100 °C due to loss of water molecules on the hydrophilic GO surface. A large mass loss occurred at about 200 °C, possibly due to decomposition of labile, oxygen-containing functional groups. rGO had a much greater thermal stability than did GO, indicating that GO functional groups containing oxygen were removed successfully during reduction with Ph3PBr2.
In the X-ray diffraction (XRD) analysis of graphite, GO, and rGO, shown in Fig. 2c, graphite demonstrated a major diffraction peak at 2θ ≈ 26.3°. GO had a diffraction peak at 2θ ≈ 10.3°, which corresponds to an interlayer d-spacing of 0.84 nm. After the reduction process, the peak at 2θ ≈ 10.3° was no longer present in the GO XRD pattern, and rGO had a broad peak at 2θ ≈ 24°, which was attributed to the agglomeration of graphene sheets.27 Hydrazine (N2H4) is a common and effective reducing agent for converting from GO to rGO via hydrothermal treatment. However, hydrazine is highly toxic, explosive chemical and can potentially functionalize the GO with nitrogen heteroatoms. We used Raman spectroscopy to compare the structural properties of GO, rGO reduced by hydrazine, and rGO reduced by Ph3PBr2. Raman spectra were measured at an excitation of 514 nm, and all spectra were normalized to the G-band. Fig. 2d shows the Raman spectrum of GO, in which the Raman fingerprints of GO, including the D-band (1358 cm−1) and G-band (1596 cm−1), are apparent. The Raman spectrum of rGO reduced by hydrazine had a D-band at 1349 cm−1 and a G-band at 1589 cm−1, and rGO reduced by Ph3PBr2 had a D-band at 1357 cm−1 and a G-band at 1590 cm−1. The G-band of rGO reduced by Ph3PBr2 exhibited a significant redshift, similar to that of rGO reduced by hydrazine. In addition, the ID/IG ratio increased from 0.8 (GO) to 1.0 (rGO by Ph3PBr2), indicating that the graphene domain decreased during the reduction process. This result indicates that our reducing method using Ph3PBr2 is a good candidate for the reduction method that does not require strong acidic conditions or highly toxic agents such as hydrazine. Additionally, we took AFM topography images of a GO to measure the stability during reducing process (Fig. S3 in ESI†). The results indicate that we achieved GO nanosheets with ∼1.1 nm. It is clear that we can obtain uniform rGO flakes without the structure destruction of GO nanosheets after reduction processes.
X-ray photoelectron spectroscopy (XPS) was used to quantify the atomic compositions and stoichiometric ratios of GO, rGO reduced by hydrazine, and rGO by Ph3PBr2. Fig. 3a–c show the core level of C1s XPS spectra of GO, rGO reduced by Ph3PBr2, and rGO reduced by hydrazine. As shown in Fig. 3a, XPS spectra of GO exhibits six peaks at 284.5, 285.3, 286.7, 288.4 and 289.2, which were assigned to C–C, C–O, C–O–C, CO and O–C
O functional groups, respectively. In cases of rGO reduced by Ph3PBr2 and rGO by hydrazine, the intensity of the oxygenated groups in the C1s peak was reduced dramatically, compared to GO. This result shows that the oxygen-containing group was eliminated during the reduction process, and the sp2 carbon peaks were recovered. Also, a new carbon peak at 290.8 eV corresponding to π–π* satellite was appeared at a high binding energy, which corresponds to a carbon bound to another atom. Significantly, the C/O ratio of rGO reduced by hydrazine is only 7.9, while that of rGO by Ph3PBr2 is 16. Also, rGO by Ph3PBr2 has a larger C/O ratio than rGO by dehydrobromination in a previous work.41
Fig. 3d shows the optical transmittances at 550 nm of thin films based on rGO reduced by hydrazine and rGO reduced by Ph3PBr2. Thin GO film on a transparent substrate was prepared by spin-coating, and the sheet resistance was measured by using the four-point probe technique. The sheet resistance of rGO reduced by hydrazine was 350.6 kΩ per sq at 69.4%, and the film based on rGO reduced by Ph3PBr2 had a sheet resistance of 26.8 kΩ per sq with an optical transmittance of 69.7%.
The relative atomic ratio of functional groups of GO, rGO reduced by Ph3PBr2, and rGO reduced by hydrazine were determined by XPS analysis and are summarized in Table 1 rGO reduced by Ph3PBr2 had a higher percentage of C–C (sp2) after the reduction process and a hetero carbon percentage of 18.8%. The hetero carbon percentage of rGO reduced by Ph3PBr2 was lower than that of rGO reduced by hydrazine. This result indicates that our approach may allow us to mass-produce rGO nanosheets with high quality. The atomic percentages of rGO reduced by Ph3PX2 in various synthetic conditions according to XPS analysis are also given in Table 2. GO was reacted with Ph3PBr2 (6 equiv.) at 180 °C without solvent to yield rGO containing 5.8% oxygen. In a similar reaction using PPh3I2 instead of Ph3PBr2, rGO contained 6.1% oxygen. When we increased the GO weight up to gram scale in this condition, the rGO was prepared up to gram scale with similarly C/O ratio. In addition, the dependence of the reduction reaction on temperature was investigated. When the reaction was conducted with Ph3PBr2 (6 mmol) under neat conditions or in chlorobenzene at 120 °C, the rGO contained 11.2% or 10.5% oxygen, respectively. rGO and Ph3PBr2 (6 mmol) in DMF by microwave assist reaction at 120 °C (400 W) had a lower oxygen percent (7.1%) than with an ultrasonic-assisted reaction in acetonitrile at 40 °C (O: 8.6%). Refluxing GO and Ph3PBr2 (6 mmol) in acetonitrile provided bromine attached partially reduced GO (C: 84.8, O: 13.0 and Br: 1.4% content) as expected in proposed mechanism (Fig. 1). When only PPh3 without dihalide was used, the rGO contained 18.3% oxygen, which was lower than the 23.8% for rGO heated without reagents or solvent. The various phosphine regents, solvents, and temperature were optimized to reduce GO as described in detail (37 reaction conditions) in the ESI.†
Sample | Relative atomic percentage (%) | Hetero-carbon (%) | |||||
---|---|---|---|---|---|---|---|
C–C | C–O, C–N | C–O–C | C![]() |
C![]() |
π–π* | ||
GO | 38.5 | 5.3 | 46.2 | 8.1 | 1.2 | 0.7 | 60.9 |
rGO (Ph3PBr2) | 77.7 | 10.2 | 5.8 | 1.8 | 1.0 | 3.5 | 18.8 |
rGO (H4N2) | 73.3 | 8.3 | 10.6 | 5.2 | 0.8 | 1.8 | 25.0 |
Reagent (equiv.) | Temp. (°C) | C (%) | O (%) |
---|---|---|---|
a GO (100 mg) and reagents (6 mmol) were dispersed in CH3CN and then reaction mixture was concentrated for neat reaction.b Br content: <0.5%.c Br content: 1.4%. | |||
PPh3Br2 (6)/neata | 180 | 93.8 | 5.8 |
PPh3I2 (6)/neata | 180 | 93.1 | 6.1 |
PPh3Br2 (6)/neata | 120 | 88.4 | 11.2b |
PPh3Br2 (6)/ClPh | 120 | 88.1 | 10.5b |
PPh3Br2 (6)/DMF/microwave | 120 | 92.4 | 7.1 |
PPh3Br2 (6)/CH3CN/sonic wave | 40 | 90.9 | 8.6 |
PPh3Br2 (6)/CH3CN | 90 | 84.8 | 13.0c |
PPh3 (6)/neata | 180 | 81.5 | 18.3 |
Pristine GO | — | 64.2 | 35.4 |
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra26046a |
‡ These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2016 |