- Email: [email protected]
Chemical instability of graphene oxide following exposure to highly reactive radicals in advanced oxidation processes
Accepted Manuscript Regular Article Chemical instability of Graphene Oxide following exposure to highly reactive radicals in advanced oxidation processes Zhaohui Wang, Linyan Sun, Xiaoyi Lou, Fei Yang, Min Feng, Jianshe Liu PII: DOI: Reference:
S0021-9797(17)30872-X http://dx.doi.org/10.1016/j.jcis.2017.07.105 YJCIS 22634
To appear in:
Journal of Colloid and Interface Science
Received Date: Revised Date: Accepted Date:
6 May 2017 28 July 2017 28 July 2017
Please cite this article as: Z. Wang, L. Sun, X. Lou, F. Yang, M. Feng, J. Liu, Chemical instability of Graphene Oxide following exposure to highly reactive radicals in advanced oxidation processes, Journal of Colloid and Interface Science (2017), doi: http://dx.doi.org/10.1016/j.jcis.2017.07.105
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Chemical instability of Graphene Oxide following exposure to highly reactive radicals in advanced oxidation processes Zhaohui Wangab*, Linyan Suna, Xiaoyi Lou c, Fei Yanga, Min Fenga, Jianshe Liu a
a
State Environmental Protection Engineering Center for Pollution Treatment and
Control in Textile Industry, College of Environmental Science and Engineering, Donghua University, Shanghai, 201620, China b
International Center for Balanced Land Use (ICBLU), The University of Newcastle,
Callaghan, NSW 2308, Australia c
East China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences,
Shanghai, 200090, China *E-mail: [email protected] (Z. Wang)
1
Abstract The rapidly increasing and widespread use of graphene oxide (GO) as catalyst supports, requires further understanding of its chemical stability in advanced oxidation processes (AOPs). In this study, UV/H2O2 and UV/persulfate (UV/PS) processes were selected to test the chemical instability of GO in terms of their performance in producing highly reactive hydroxyl radicals (•OH) and sulfate radicals (SO4•-), respectively. The degradation intermediates were characterized using UV–visible absorption spectra (UV-vis), x-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), Raman spectroscopy, and matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF-MS). Experimental data indicate that UV/PS process was more effective in enhancing GO degradation than the UV/H2O2 system. The overall oxygen-containing functionalities (e.g. C-O, C=O and O-C=O groups) dramatically declined. After radical attack, sheet-like GO was destructed into lots of flakes and some low-molecular-weight molecules were detected. The results suggest GO is most vulnerable against SO4•radical attack, which deserves special attention while GO acts as a catalyst support or even as a catalyst itself. Therefore, stability of GO and its derivatives should be carefully assessed before they are applied to SO4•--based AOPs. Keywords: Graphene oxide; Sulfate radical; Hydroxyl radical; Decomposition; Advanced Oxidation Processes
2
1. Introduction Graphene oxide (GO), a structural analog to a graphene-based material, has attracted great interest in the science community all over the world because of its fundamental properties and potentially broad applications [1,2]. Different from the pristine graphene, GO has hydroxyl and epoxide functional moieties on the basal plane, while carbonyls and carboxyl groups dominate on the edges [3]. These oxygen functionalities of GO lead to an enhanced dispersibility in aqueous media, which makes it feasible for their applications in biomedicines, sensors, and environmental and energy applications (such as ultracapacitors, photocatalysts and adsorbent) [4-9]. The strong hydrophilicity of GO also enables strong H-bonding interaction to water, particularly benefiting photocatalytic water splitting applications [10]. As a consequence its properties (e.g. large surface area and good electrical conductivity), GO and related materials, such as reduced GO (rGO), have been extensively used as supports in environmental photocatalysis through improving the adsorption of organic pollutants, increasing charge separation and facilitating charge transport [11, 12]. A number of GO-iron catalysts hybrid materials (such as, GO-Fe3O4, GO-Fe-aminoclay, GO-Fe2O3) have been fabricated as efficient heterogeneous Fenton-like catalysts for degradation of organic contaminants [13-15]. Graphene oxide was found to outperform the macroscale carbon systems (e.g. activated carbon) in improving catalytic activity for H2O2 decomposition [16], which is inextricably linked to surface functionalities, morphology and unique chemical feature of GO. Under the support of GO, decomposition of H2O2 that is catalytically 3
driven by iron catalysts leads to formation of highly reactive •OH radical (Eqs. 1-3) (where ≡ denotes the iron species bound to the catalyst surface).
≡ Fe III + H 2O 2 ← → ≡ Fe III H 2O 2
(1)
≡ FeIII H 2O 2 → ≡ FeII + HO•2 + H +
(2)
≡ Fe II + H 2O 2 → ≡ Fe III + • OH + OH −
(3)
Recently, sulfate radical (SO4•-) based advanced oxidation processes (SR-AOPs) have been extensively investigated in degrading toxic organic pollutants and treating contaminated soil or groundwater, owing to its relatively long life-time (30-40 ps) and the strong oxidation capacity (E=2.5-3.1 V versus NHE) of SO4•- [17, 18]. Persulfate (PS) is frequently used to produce SO4•- via activation by transition metals [19], heat [20], base [21] and ultraviolet (UV) radiation [22]. To further improve the catalytic activity of heterogeneous catalysts towards PS or peroxymonosulfate (PMS) activation, GO is selected as a support. Shi et al. reported a significantly enhanced catalytic activity of Co3O4/GO hybrid in degradation of Orange II due to a synergistic effect of Co3O4 and GO [23]. Wang and his colleagues [24, 25] even discovered that GO’s derivatives themselves, as metal-free catalysts, are capable of activating PMS efficiently. Despite of its extensive use in AOPs, GO’s stability in aqueous solution where highly reactive radicals are abundant has often been ignored, while its roles in enhancing heterogeneous catalytic activities has been highlighted. Pristine graphene is generally considered stable [26], but recent investigations indicate precursor-GO can be oxidized in TiO2-photocatalysis [27, 28], photo-Fenton [29-31], direct photolysis 4
[32-34] or by enzymatic reaction with horseradish peroxidase and H2O2 [35]. For example, under the attack of hydroxyl radicals (•OH), rGO can be cut into polycyclic aromatic hydrocarbon (PAH) like compounds then be further mineralized after longer irradiation [27]. Bai et al identified the oxidized PAHs as byproducts of GO degradation in a photo-Fenton reaction [31]. Graphene oxide could be completely mineralized by a photo-Fenton reaction after 28 days [29]. There are few studies concerning the stability of GO as exposed to SO4•-as compared to the typical •OH-based AOPs, such as UV/H2O2. The applicability of GO-based catalytic oxidation systems may be incomplete or even overestimated without the stability data of GO following exposure to highly reactive radicals (e.g. •OH, SO4•-). The objective of this study is to compare the chemical instability of GO in UV/PS and UV/H2O2 which represents one of the most efficient and simple way to produce SO4•- and •OH [36, 37], respectively. The decomposition intermediates were characterized by UV–visible absorption spectra (UV-vis), x-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), Raman spectroscopy, and matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF-MS). According to the MALDI-TOF-MS data of the degradation intermediates, GO degradation mechanism in these two systems is discussed.
2. Experimental section 2.1. Materials
Graphene oxide powder (diameter: 100~200 nm; thickness: 0.8~1.2 nm; single 5
layer ratio: > 99%; purity: > 99%) was purchased from Nanjing Jicang Nano Tech Company. A dialysis bag was obtained from Nanjing Mengyi Commercial Center. Hydrogen
peroxide
(30%),
ethylenediaminetetraacetic
acid
disodium
salt
(EDTA•2Na), sodium hydrogen carbonate (NaHCO3), potassium iodide (KI), soluble starch, sodium nitrite (NaNO2), ethyl alcohol were of reagent grade and all from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China. Potassium persulfate (K2S2O8, 99%) was obtained from Alfa Aesar. All materials were used without further purification. Barnstead UltraPure water (18.2 MΩ·cm) was used for all experiments. 2.2. Reaction procedure
Graphene oxide powder was stirred in water for 30 min, then treated with ultrasound for 2 h with ice bags to form graphene oxide suspension and stored in refrigerator without light. A dialysis bag was pretreated by heating membranes (8-10 cm) at 105 ℃ in 10 g NaHCO3, 186.6 mg EDTA•2Na and 500 mL water for 10 min, and stored in 50% ethanol. The degree of dialysis can be measured either by iodometry or evaluated by its electrical conductivity. The chemical stability of GO was examined in a photochemical reactor equipped with the irradiation lamps having different emission wavelengths and powers (Xujiang Electromechanical Plant, Nanjing, China). The light source in our experiment was a 300 W medium-pressure Hg lamp (maximal emission: 365.0-366.3 nm). Its intensity at reactor vessel positions was 16.8 mW/cm2 measured by a UV-A irradiation meter (Photoelectric Instrument Factory of Beijing Normal University, China). The reaction was carried out in a 50 mL quartz tube under continuous 6
magnetic stirring and cooled by circulating water with the temperature maintained at 31±2 ℃. During this study, several series of tubes were prepared for irradiation. At specific times during irradiation, one tube was removed from the reactor and sacrificed for chemical analysis. Dark control tubes were covered with aluminum foil and irradiated concurrently. In some experiments, duplicate samples were irradiated for the same time and studied to verify experimental reproducibility. In a typical experiment, 6.25 mL of 4 mg/10 mL GO aqueous suspension and 5 mL of 100 mM H2O2 were mixed in a quartz tube by vigorous stirring. At the same time, for comparison, 6.25 mL of 4 mg/10 mL GO aqueous suspension and 10 mL of 50 mM K2S2O8 (PS) were also mixed concurrently. After the reaction, the samples were quenched with NaNO2 (aq) immediately and then dialyzed (100 Da) against ultrapure water for 7 days to remove residual H2O2, PS, and finally freeze-dried. 2.3. Instrumentations
2.3.1 UV-vis absorption spectrum A UV–visible spectrophotometer (Hitachi Model U-2910) was employed for absorbance measurements using quartz cells of 1 cm path length. Samples obtained at 0h, 1h, 2h and 4 h were analyzed immediately without any purification. 2.3.2 XPS X-ray photoelectron spectroscopy (XPS) measurement was carried out on a PHI-5000C ESCA system (Perkin-Elmer, USA). The samples were lyophilized for tests. The XPS spectra was analyzed and fitted using XPS peak 4.1 software. 7
2.3.3 TEM The nanostructures of graphene oxide ware characterized by using a transmission electron microscopy (TEM) system (JEOL, Model JEM-2100) operating at 200 kV. Approximately 10 µL of samples (aq) were drop-coated onto copper grids and dried under ambient conditions. 2.3.4 AFM Atomic force microscopic (AFM) images were acquired using Agilent 5500 AFM system (Agilent Technologies, USA) in tapping mode for thickness, phase, and sectional analysis. Approximately 10 µL of samples (aq) were drop-coated onto mica plates and dried under ambient conditions. 2.3.5 Raman Raman spectra were collected on a Renishaw InVia Reflex Raman spectrometer in a backscattering configuration. One drop of sample (aq) was drop-coated onto glass sheets and dried in a vacuum oven several times. 2.3.6 MALDI-TOF-MS The MS analysis was performed with 4800 Plus MALDI TOF/TOF Analyzer (AB Sciex). Mass spectra was acquired in the reflector mode in the m/z 50-1000 and 1000-5000 mass range. The samples were tested without any matrix. Approximately 10 µL of the sample (aq) was dropped onto a MALDI plate and dried under ambient conditions. 3. Results and discussion 3.1. UV-vis absorption spectrum 8
The preliminary data indicate that neither PS or H2O2 can effectively decompose GO within the experimental duration (see Fig. S1). All suspensions containing GO, H2O2-GO and PS-GO were exposed to UV irradiation for 4 h. The irradiation of the solutions containing H2O2 or PS generated highly reactive •OH or SO4•- radicals. UV exposure darkened the color of GO sample (Fig. S2), leading to a monotonic increase in the UV-vis absorbance (Fig. S3), which has been observed in previous reports (Hou et al., 2015; Matsumoto et al., 2010). Fig. 1 illustrates the UV-vis absorption spectra of GO suspension under UV/H2O2 and UV/PS after 0 h, 1 h, 2 h and 4 h of oxidation. GO shows two characteristic peaks in UV-vis spectra: the peak of 230 nm represents the Π→Π* transition of the C=C bonding and the shoulder at 300 nm is related to the n→Π* transition of the C=O groups [38]. In UV/H2O2 system, no obvious change occurred during the initial 3 h and a decrease in absorbance happened at 4 h. The absorbance increase resulting from the direct photolysis of GO may be offset over time by GO oxidation by •OH which can cause colour bleaching of GO. In contrast, a monotonic decrease in UV-Vis absorbance of GO sample in UV/PS system was observed. After 4 h, GO sample in this suspension turned colorless, implying a predominant role of SO4•- oxidation against direct photolysis. A high content of GO is likely to decrease the UV absorption of the oxidants, thereby affecting the degradation efficiency of GO. Three more levels of GO dosages (0.03, 0.08, 0.13 g/L) were examined. Figs. S4-S6 show that GO degradation was significantly affected by GO loading. However, the changes in UV-vis absorption spectra with time were very similar to those at a GO 9
loading of 0.05 g/L (Fig. 1), that is, GO in the UV/PS system is much more rapidly degraded than that in the UV/H2O2 system. Collectively, these comparisons demonstrate that to some extend the UV/PS process is more effective towards GO degradation than the UV/H2O2 process. Some studies reported that the distance of O-O bond in PS (1.497Å) is longer than in H2O2 (1.453Å), suggesting that PS could be more easily cleaved than H2O2 [39]. In addition, PS absorbed more strongly at 254 nm than H2O2 [40]. It is therefore supposed that PS can be activated more easily by UV.
Fig. 1. UV-visible absorption spectra of GO suspension collected after 0 h, 1 h, 2 h 10
and 4 h irradiation. (a) UV/H2O2; (b) UV/PS. 3.2 ．XPS
To probe into the chemical composition variation of the GO samples during the radical oxidation, the XPS spectra were acquired. Figures. 2a and 2b depict the XPS spectra of C 1s of GO before and after the UV/H2O2 and UV/PS reactions for 4 h. The C 1s spectra were deconvoluted into four sub-peaks corresponding to C atoms with a variety of chemical states. The peaks at 285.0, 286.8-287.1, 287.8-288.0 and 289.0-289.3 eV are assigned to C=C, C-C, C-H bonds, C-O (C-O-C and C-OH), C=O, and O=C-OH bands, respectively [33]. The XPS spectra show the substantial changes in GO surface O-functional groups upon oxidation treatment (Fig. 2c). The graphitic content of photolyzed GO samples increased from 23% to 61% in the UV/H2O2 system, from 24% to 72% in UV/PS system. In contrast, the overall oxygen containing functionalities (including C-O, C=O and O-C=O groups) decreased dramatically from 77% to 39% in the UV/H2O2 system, from 76% to 28% in UV/PS system. Disappearance of C/O containing groups caused GO mineralization and relative percentage increase in graphitic C=C, C-C and C-H groups.
11
Fig. 2. XPS spectra of C 1s of GO suspension obtained after 0 h and 4 h. (a) UV/H2O2.
(b) UV/PS. (c) the percentage of carbon groups. 12
3.3. Raman spectroscopy
Raman spectroscopy was employed to further analyze the degree of oxidation of GO by free radicals (Fig 3). The observed D and G bands are distinctive of graphitic materials: the D band generally appears at 1300-1400 cm-1 representing the disorder present in sp2-hybridized carbon systems and the G band generally appears at 1560-1620 cm-1 representing the stretching of C-C bonds [41, 42]. The intensity ratio (ID/IG) is usually used to evaluate the degree of disorder of graphitic materials, such as defects, ripples, and edges [38, 42]. The higher the ratio is, the more defects the material has. In our experiments, the calculated ratios were decreased from 0.983 to 0.971 and 0.938, after the two radical oxidation. This means that their degrees of order were both improved. The lower ID/IG value in UV/PS system compared with UV/H2O2 system indicates that the SO4•- radical is more efficient in removing oxygen functional groups of GO, is consistent with the XPS analysis.
Fig. 3. Raman spectra of pristine GO, UV/H2O2-4 h, UV/PS-4 h. 3.4. TEM 13
Using TEM, we observed morphologic changes of GO before and after the radical attack (Fig. 4). Fig. 4a illustrates the transparent and sheet-like structure of GO with large number of wrinkles and scrolls. The size of pristine GO was more than 500 nm and it was not a single layer. Fig. 4b shows many GO flakes were found in the UV/H2O2 process after 4 h. While under the same reaction time in UV/PS, these GO flakes became smaller which indicates the higher efficiency of UV/PS system for GO decomposition.
Fig. 4. TEM images of GO. (a) pristine GO; (b) UV/H2O2-4 h; (c) UV/PS-4 h. 3.5. AFM
AFM is employed to analyze the morphology and thickness of GO after the attack of the two free radicals. Oval and punctiform morphologies of GO samples 14
were observed in the UV/H2O2 and UV/PS system, respectively (Figs. 5a, 5b). The AFM thickness of the GO flakes in the UV/H2O2 treatment was about 1.3 nm (Fig. 5c), half of that after oxidation in the UV/PS system, demonstrating that GO samples after PS treatment consisted of an aggregation of flakes. According to the XPS data, the oxygen functional groups on the basal plane of GO were reduced. As a result, the electrostatic repulsion between the flakes of GO decreased while van der Waals forces dominated, leading to the formation of aggregates.
15
Fig. 5. Tapping mode AFM images of GO (a, c) UV/H2O2-4h; (b, d) UV/PS-4h.
3.6. MALDI-TOF-MS
The specific structure of GO can lead us to envision it as many fused polyaromatic hydrocarbons (PAHs), a significant environmental concern. Based on the understanding of its morphology and size, chemical functional groups on its surface, we trace its MS changes which could not be detected by TEM during the oxidation processes. MALDI-TOF-MS has an important advantage in analytical chemistry: absolute molecular weights of fragments can be determined, as opposed to obtaining relative molecular weights by chromatographic techniques. According to these obtained accurate molecular weights, one can get their molecular formulas (CxHyOz). The corresponding molecular structures are tentatively assigned during the molecular formulas and structural similarity with the parent compound. Figure 6 depicts the MALDI-TOF-MS spectra corresponding to products obtained at 0, 1, 2 and 4 h after the beginning of UV/H2O2 or UV/PS reaction. The distinct differences between the two processes were: low molecular weight (LMW) products and the purity of suspension which is indicative of the reaction efficiency. Because of the complexity of the data, only selected compounds were chosen for chemical composition and possible structural analysis. In the process of UV/H2O2, peaks between m/z 49 to 200 mainly disappear after 1 h, implying LMW substances were degraded quickly. Known from peaks at m/z 701 ([C40H28O12 + H]+) and 537 ([C36H40O4 + H]+), they were degraded to form other species, such as peak at m/z 489 16
([C25H28O10+H]+) which represents the most abundant compound in the system. New peaks lower than 200 m/z appeared and maintained during the last two hours. However, no peaks lower than m/z 202 (C14H18O) were found in samples treated by UV/PS. Moreover, in the latter two hours of experiment, no peaks smaller than m/z 263 ([C12H6O7 + H]+) were detected. The MS results prove that the UV/PS system is more efficient in the degradation of GO. In other words, this means applications of GO should be taken more concerns where SO4•- radical exists. Changes of peaks in the range of m/z 1000-5000 were also considered. Different to the results of Bai et al., the changes of carbon clusters were not distinct [31]. The possible reasons are that there were no pretreatments of our samples before the MS test and it records the whole compounds the instrument can detect. But from the signal intensity, it is clear that differences do occur during the reaction. In the UV/PS system at 4 h, there was no signal, meaning no compounds in m/z 1000-5000 can be found (Figs. S7 and S8). Some tentative chemical structures of typical compounds in those systems are suggested. However, considering the existence of allotropes, further efforts are needed to identify the decomposed byproducts of GO.
17
Fig. 6. MALDI-TOF-MS spectra for GO samples after 0, 1, 2 and 4 h. (a) UV/H2O2;
(b) UV/PS; (c) possible chemical structures of products observed in spectra. 18
3.6. Mechanism of GO oxidative degradation pathway in •OH and SO4•- radicals
Recently, reactions of graphene-related materials with
•
OH have been
investigated under different scenarios, such as engineering nano-porous GO, degrading GO in the wastewater and the transformation of GO in the natural environments [30, 32, 43]. Graphene oxide would be degraded in the presence of Fenton reagents (Fe2+/3+, H2O2) under UV irradiation. Radich et al. reported that rGO was bleached with a significant structural degradation upon reaction with •OH [28]. Yu et al. attempted to etch GO sheets by •OH generated from water radiolysis by g-irradiation and nano-porous GO sheets with well controlled nano-pores and ‘lacelike’ edges [43]. However, there are few studies on reactions of GO with SO4•radicals. It is generally accepted that •OH and SO4•- radicals react with organic compounds by three ways: electron transfer, hydrogen abstraction and addition mechanisms [44]. Among these, •OH is more likely to participate in hydrogen abstraction or addition reactions, while SO4•- is apt to electron transfer reactions [45]. Yu et al. reported that hydroxylation of •OH is a major and crucial step for initial GO oxidation [43]. Hydroxyl (•OH) undergoes electrophilic addition on unsaturated bonds to produce hydroxylated GO, which would be further oxidized to quinones, then carboxylic acids. Hydroxyl (•OH) imparts carboxylic acids at the defect sites, leading to mineralization of those oxygen moieties [46], as evidenced from TOC and XPS analyses. UV/PS system seems more efficient towards GO degradation, because SO4•- (E=2.6-3.1 V vs NHE) could be more reactive than •OH (E=1.9-2.7 V vs NHE) [47]. To compare their 19
oxidation capacity of UV/H2O2 and UV/PS systems, degradation performance of a typical dye, AO7 was evaluated with the same level of the oxidant. It is observed that AO7 was much rapidly degraded in the UV/PS system (Fig. S9), implying that UV/PS process is more efficient to generate radicals and oxidize AO7 and GO. Therefore, it is believed that SO4•--induced degradation can result in a better mineralization rate than •OH [48]. On the other hand, SO4•- is an electrophilic oxidant, and prefers to react with electron-donating groups, such as hydroxyl (-OH) and organic compounds containing unsaturated bonds. Therefore, SO4•- is supposed to selectively attack C=C bonds on the GO structure, facilitating the GO destruction and cleavage.
4. Conclusion
This study examined the structural changes of GO in UV/H2O2 and UV/PS processes and found that a monotonic decrease in UV-vis absorbance of aqueous GO sample was observed in a UV/PS system, while no obvious changes of UV-vis spectra occurred in a UV/H2O2 solution. GO was decomposed to small flakes after oxidation and some low-molecular-weight molecules. Collectively, UV/PS process seems more effective towards GO decomposition than the UV/H2O2 system, because of stronger oxidation potential and electrophilicity of SO4•- radicals. This study provides, to the best of our knowledge, the first evidence for GO’s instability in aqueous solution where highly reactive radicals are abundant, although its roles in enhancing heterogeneous catalytic activities are widely recognized. GO would be destructed upon the radical attack when it is used as a support or catalyst in producing SO4•20
radicals from persulfate or peroxymonosulfate activation. Therefore, special attention on GO stability should be given when GO-based nanocomposites are used in advanced oxidation processes (AOPs) where highly reactive radicals are abundant. On the other hand, UV/PS technology can be envisaged as a novel method for “etching” GO sheets and modifying porous structures of GO due to the stronger oxidative capacity of SO4•- radicals than hydroxyl radicals generated in water radiolysis [43]. This engineering process can be conducted under a mild condition only by a simple control of the treatment time. In addition, comparative toxicity assessment of GO suspensions and destructed GO would be a very interesting topic and deserves in-depth investigations [49].
Acknowledgments
This work was supported by the National Natural Science Foundation of China (NSFC) (Nos. 21377023 and 21677031), National Key Research and Development Program of China (2016YFC0400501/2016YFC0400509), Shanghai Pujiang Program and DHU Distinguished Young Professor Program. The authors thank Dr. Nick Ward at Southern Cross University for his careful proofreading. Appendix A. Supplementary material
Supplementary data associated with this article can be found, in the online version. Reference
[1] S. Stankovich, D.A. Dikin, G.H.B. Dommett, K.M. Kohlhaas, E.J. Zimney, 21
E.A. Stach, R.D. Piner,
S.T. Nguyen,
R.S. Ruoff, Nature 442 (2006)
282-286. [2] Compton, S.T. Nguyen, Small 6 (2010) 711-723. [3] A. Dimiev, D.V Kosynkin, L.B. Alemany, P. Chaguine, J.M. Tour, J. Am. Chem. Soc. 134 (2012) 2815-2822. [4] D. Chen, H. Feng, J. Li, Chem. Rev. 112 (2012) 6027-6053. [5] C. Chung, Y.K. Kim, D. Shin, Acc. Chem. Res. 46(2013), 2211-2224. [6] F. Peng, T. Luo, L. Qiu, Y. Yuan, Mater. Res. Bulletin. 48 (2013) 2180-2185. [7] H. Yan, X. Tao, Z. Yang, K. Li, H. Yang,
A. Li, R. Cheng, J. Hazard. Mater.
268 (2014) 191-198. [8] M.L. Yola, T. Eren, N. Atar, Sens. Actuators. B: Chem. 210 (2015) 149-157. [9] M.L. Yola, V.K. Gupta, T. Eren, A.E. Şen, N. Atar, Electrochim. Acta. 120 (2014) 204-211. [10] T.F. Yeh, J.M. Syu, C. Cheng, T.H. Chang, H. Teng, Adv. Funct. Mater. 20 (2010) 2255-2262. [11] G. Chen, M. Sun, Q. Wei, Y. Zhang, B. Zhu, B. Du, J. Hazard. Mater. 244 (2013) 86-93. [12] P. Gao, J. Liu, D.D. Sun, W. Ng, J. Hazard. Mater. 250 (2013) 412-420. [13] S. Guo, G. Zhang, Y. Guo, J.C Yu, Carbon 60 (2013) 437-444. [14] Y.C. Lee, S.J. Chang, M.H. Choi, T.J Jeon, T. Ryu, Y.S. Huh, Appl. Catal. B: Environ. 142 (2013) 494-503. [15] N.A. Zubir, C. Yacou, X. Zhang, J.C. Diniz da Costa, J. Environ. Chem. Eng. 22
2 (2014) 1881-1888. [16] K.V. Voitko, R.L. Whitby, V.M. Gun’ko, O.M. Bakalinska, M.T. Kartel, K. Laszlo, A.B. Cundy, S.V. Mikhalovsky, J. Colloid. Interf. Sci. 361 (2011) 129-136. [17] F. Ghanbari, M. Moradi, Chem. Eng. J. 310 (2017) 41-62. [18] Z. Wang, R. Yuan, Y. Guo, L. Xu, J. Liu, J. Hazard. Mater. 190 (2011) 1083-1087. [19] G.P. Anipsitakis, D.D. Dionysiou, Appl. Catal. B: Environ. 54 (2004) 155-163. [20] R.L. Johnson, P.G. Tratnyek, R.O.B Johnson, Environ. Sci. Technol. 42 (2008) 9350-9356. [21] Furman, A.L. Teel, R.J. Watts, Environ. Sci. Technol. 44 (2010) 6423-6428. [22] M.G. Antoniou, A.A. de la Cruz, D.D. Dionysiou, Appl. Catal. B: Environ. 96 (2010) 290-298. [23] P. Shi, R. Su, F. Wan, M. Zhu, D. Li, S. Xu, Appl. Catal. B: Environ. 123-124 (2012) 265-272. [24] W. Peng, S. Liu, H. Sun, Y. Yao, L. Zhi, S. Wang, J. Mater. Chem. A 1 (2013) 5854-5859. [25] H. Sun, Y. Wang, S. Liu, L. Ge, L. Wang, Z. Zhu, S. Wang, Chem. Comm. 49 (2013) 9914-9916. [26] P.M. Ajayan, B.I. Yakobson, Mater. Sci: Nat. 441 (2016) 818-819. [27] O. Akhavan, M. Abdolahad, A. Esfandiar, M. Mohatashamifar, J. Phys. Chem. C 114 (2010) 12955-12959. 23
[28] J.G. Radich, A.L. Krenselewski, J. Zhu, P.V. Kamat, Chem. Mater. 26 (2014) 4662-4668. [29] C.Z. Zhang, T. Li, Y. Yuan, J. Xu, Chemosphere 153 (2016) 531-540. [30] X. Zhou, Y. Zhang, C. Wang, Acs Nano 6 (2012) 6592-6599. [31] H. Bai, W. Jiang, G.P. Kotchey, W.A. Saidi, B.J. Bythell, J.M. Jarvis, A.G. Marshall, R.A. Robinson, A. Star, J. Phys. Chem. C: Nano. Interfac. 118 (2014) 10519-10529. [32] W.C. Hou, I. Chowdhury, D.G Goodwin, Jr, W.M. Henderson, D.H. Fairbrother, D. Bouchard, R.G. Zepp, Environ. Sci. Technol. 49 (2015) 3435-3443. [33] Y. Matsumoto, M. Koinuma, S. Ida, S. Hayami, T. Taniguchi, K. Hatakeyama, H. Tateishi, Y. Watanabe, S. Amano, J. Phys. Chem. C 115 (2011) 19280-19286. [34] M. Minella, M. Demontis, M. Sarro, F. Sordello, P. Calza, C. Minero, J. Mater. Sci. 50 (2015) 2399-2409. [35] G.P. Kotchey, B.L. Allen, H. Vedala, ACS. Nano. 5(2011), 2098-2108. [36] Y. Guo, J. Zhou, X. Lou, R. Liu, D. Xiao, C. Fang, Z. Wang, J. Liu, Chem. Eng. J. 254 (2014) 538-544. [37] X. Lou, D. Xiao, C. Fang, Z. Wang, J. Liu, Y. Guo, S. Lu, Environ. Sci. Pollut. Res. 23 (2015) 4778-4785. [38] Y. Zhao, C.T. Jafvert, Environ. Sci.: Nano 2 (2015) 136-142. [39] S. Yang, P. Wang, X. Yang, L. Shan, W. Zhang, X. Shao, R. Niu, J. Hazard. 24
Mater. 179 (2010) 552-558. [40] X. Gu, S. Lu, Z. Qiu, Q. Sui, Z. Miao, K. Lin, Y. Liu, Q. Luo, Ind. Eng. Chem. Res. 51 (2012) 7196-7204. [41] L.M. Malard, M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, Phys. Rep. 473 (2009) 51-87. [42] G.-H. Moon, Y. Park, W. Kim, W. Choi, J. Carbon 49 (2011) 3454-3462. [43] C. Yu, B. Zhang, F. Yan, J. Zhao, J. Li, L. Li, J. Li, Carbon 105 (2016) 291-296. [44] C. Liang, Z.S. Wang, C.J. Bruell, Chemosphere 66 (2007) 106-113. [45] Y.T. Lin, C.
Liang, J.H. Chen, Chemosphere 82 (2011) 1168-1172.
[46] W. Li, Y. Bai, Y. Zhang, M. Sun, R. Cheng, X. Xu, Y. Chen, Y. Mo, Synth. Metal. 155 (2005) 509-515. [47] W.-D. Oh, Z. Dong, T.-T. Lim, Appl. Catal. B: Environ. 194 (2016) 169-201. [48] M. Sanchez-Polo, M.M. Abdel daiem, R. Ocampo-Perez, J. Rivera-Utrilla, A. J. Mota, Sci. Total Environ, 463-464 (2013) 423-431. [49] K. Yang, Y. Li, X. Tan, R. Peng, Z. Liu, Small 9 (2013) 1492-1503.
25
Graphical abstract
UV/H2O2
GO
UV/PS
26
S0021-9797(17)30872-X http://dx.doi.org/10.1016/j.jcis.2017.07.105 YJCIS 22634
To appear in:
Journal of Colloid and Interface Science
Received Date: Revised Date: Accepted Date:
6 May 2017 28 July 2017 28 July 2017
Please cite this article as: Z. Wang, L. Sun, X. Lou, F. Yang, M. Feng, J. Liu, Chemical instability of Graphene Oxide following exposure to highly reactive radicals in advanced oxidation processes, Journal of Colloid and Interface Science (2017), doi: http://dx.doi.org/10.1016/j.jcis.2017.07.105
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Chemical instability of Graphene Oxide following exposure to highly reactive radicals in advanced oxidation processes Zhaohui Wangab*, Linyan Suna, Xiaoyi Lou c, Fei Yanga, Min Fenga, Jianshe Liu a
a
State Environmental Protection Engineering Center for Pollution Treatment and
Control in Textile Industry, College of Environmental Science and Engineering, Donghua University, Shanghai, 201620, China b
International Center for Balanced Land Use (ICBLU), The University of Newcastle,
Callaghan, NSW 2308, Australia c
East China Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences,
Shanghai, 200090, China *E-mail: [email protected] (Z. Wang)
1
Abstract The rapidly increasing and widespread use of graphene oxide (GO) as catalyst supports, requires further understanding of its chemical stability in advanced oxidation processes (AOPs). In this study, UV/H2O2 and UV/persulfate (UV/PS) processes were selected to test the chemical instability of GO in terms of their performance in producing highly reactive hydroxyl radicals (•OH) and sulfate radicals (SO4•-), respectively. The degradation intermediates were characterized using UV–visible absorption spectra (UV-vis), x-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), Raman spectroscopy, and matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF-MS). Experimental data indicate that UV/PS process was more effective in enhancing GO degradation than the UV/H2O2 system. The overall oxygen-containing functionalities (e.g. C-O, C=O and O-C=O groups) dramatically declined. After radical attack, sheet-like GO was destructed into lots of flakes and some low-molecular-weight molecules were detected. The results suggest GO is most vulnerable against SO4•radical attack, which deserves special attention while GO acts as a catalyst support or even as a catalyst itself. Therefore, stability of GO and its derivatives should be carefully assessed before they are applied to SO4•--based AOPs. Keywords: Graphene oxide; Sulfate radical; Hydroxyl radical; Decomposition; Advanced Oxidation Processes
2
1. Introduction Graphene oxide (GO), a structural analog to a graphene-based material, has attracted great interest in the science community all over the world because of its fundamental properties and potentially broad applications [1,2]. Different from the pristine graphene, GO has hydroxyl and epoxide functional moieties on the basal plane, while carbonyls and carboxyl groups dominate on the edges [3]. These oxygen functionalities of GO lead to an enhanced dispersibility in aqueous media, which makes it feasible for their applications in biomedicines, sensors, and environmental and energy applications (such as ultracapacitors, photocatalysts and adsorbent) [4-9]. The strong hydrophilicity of GO also enables strong H-bonding interaction to water, particularly benefiting photocatalytic water splitting applications [10]. As a consequence its properties (e.g. large surface area and good electrical conductivity), GO and related materials, such as reduced GO (rGO), have been extensively used as supports in environmental photocatalysis through improving the adsorption of organic pollutants, increasing charge separation and facilitating charge transport [11, 12]. A number of GO-iron catalysts hybrid materials (such as, GO-Fe3O4, GO-Fe-aminoclay, GO-Fe2O3) have been fabricated as efficient heterogeneous Fenton-like catalysts for degradation of organic contaminants [13-15]. Graphene oxide was found to outperform the macroscale carbon systems (e.g. activated carbon) in improving catalytic activity for H2O2 decomposition [16], which is inextricably linked to surface functionalities, morphology and unique chemical feature of GO. Under the support of GO, decomposition of H2O2 that is catalytically 3
driven by iron catalysts leads to formation of highly reactive •OH radical (Eqs. 1-3) (where ≡ denotes the iron species bound to the catalyst surface).
≡ Fe III + H 2O 2 ← → ≡ Fe III H 2O 2
(1)
≡ FeIII H 2O 2 → ≡ FeII + HO•2 + H +
(2)
≡ Fe II + H 2O 2 → ≡ Fe III + • OH + OH −
(3)
Recently, sulfate radical (SO4•-) based advanced oxidation processes (SR-AOPs) have been extensively investigated in degrading toxic organic pollutants and treating contaminated soil or groundwater, owing to its relatively long life-time (30-40 ps) and the strong oxidation capacity (E=2.5-3.1 V versus NHE) of SO4•- [17, 18]. Persulfate (PS) is frequently used to produce SO4•- via activation by transition metals [19], heat [20], base [21] and ultraviolet (UV) radiation [22]. To further improve the catalytic activity of heterogeneous catalysts towards PS or peroxymonosulfate (PMS) activation, GO is selected as a support. Shi et al. reported a significantly enhanced catalytic activity of Co3O4/GO hybrid in degradation of Orange II due to a synergistic effect of Co3O4 and GO [23]. Wang and his colleagues [24, 25] even discovered that GO’s derivatives themselves, as metal-free catalysts, are capable of activating PMS efficiently. Despite of its extensive use in AOPs, GO’s stability in aqueous solution where highly reactive radicals are abundant has often been ignored, while its roles in enhancing heterogeneous catalytic activities has been highlighted. Pristine graphene is generally considered stable [26], but recent investigations indicate precursor-GO can be oxidized in TiO2-photocatalysis [27, 28], photo-Fenton [29-31], direct photolysis 4
[32-34] or by enzymatic reaction with horseradish peroxidase and H2O2 [35]. For example, under the attack of hydroxyl radicals (•OH), rGO can be cut into polycyclic aromatic hydrocarbon (PAH) like compounds then be further mineralized after longer irradiation [27]. Bai et al identified the oxidized PAHs as byproducts of GO degradation in a photo-Fenton reaction [31]. Graphene oxide could be completely mineralized by a photo-Fenton reaction after 28 days [29]. There are few studies concerning the stability of GO as exposed to SO4•-as compared to the typical •OH-based AOPs, such as UV/H2O2. The applicability of GO-based catalytic oxidation systems may be incomplete or even overestimated without the stability data of GO following exposure to highly reactive radicals (e.g. •OH, SO4•-). The objective of this study is to compare the chemical instability of GO in UV/PS and UV/H2O2 which represents one of the most efficient and simple way to produce SO4•- and •OH [36, 37], respectively. The decomposition intermediates were characterized by UV–visible absorption spectra (UV-vis), x-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), Raman spectroscopy, and matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF-MS). According to the MALDI-TOF-MS data of the degradation intermediates, GO degradation mechanism in these two systems is discussed.
2. Experimental section 2.1. Materials
Graphene oxide powder (diameter: 100~200 nm; thickness: 0.8~1.2 nm; single 5
layer ratio: > 99%; purity: > 99%) was purchased from Nanjing Jicang Nano Tech Company. A dialysis bag was obtained from Nanjing Mengyi Commercial Center. Hydrogen
peroxide
(30%),
ethylenediaminetetraacetic
acid
disodium
salt
(EDTA•2Na), sodium hydrogen carbonate (NaHCO3), potassium iodide (KI), soluble starch, sodium nitrite (NaNO2), ethyl alcohol were of reagent grade and all from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China. Potassium persulfate (K2S2O8, 99%) was obtained from Alfa Aesar. All materials were used without further purification. Barnstead UltraPure water (18.2 MΩ·cm) was used for all experiments. 2.2. Reaction procedure
Graphene oxide powder was stirred in water for 30 min, then treated with ultrasound for 2 h with ice bags to form graphene oxide suspension and stored in refrigerator without light. A dialysis bag was pretreated by heating membranes (8-10 cm) at 105 ℃ in 10 g NaHCO3, 186.6 mg EDTA•2Na and 500 mL water for 10 min, and stored in 50% ethanol. The degree of dialysis can be measured either by iodometry or evaluated by its electrical conductivity. The chemical stability of GO was examined in a photochemical reactor equipped with the irradiation lamps having different emission wavelengths and powers (Xujiang Electromechanical Plant, Nanjing, China). The light source in our experiment was a 300 W medium-pressure Hg lamp (maximal emission: 365.0-366.3 nm). Its intensity at reactor vessel positions was 16.8 mW/cm2 measured by a UV-A irradiation meter (Photoelectric Instrument Factory of Beijing Normal University, China). The reaction was carried out in a 50 mL quartz tube under continuous 6
magnetic stirring and cooled by circulating water with the temperature maintained at 31±2 ℃. During this study, several series of tubes were prepared for irradiation. At specific times during irradiation, one tube was removed from the reactor and sacrificed for chemical analysis. Dark control tubes were covered with aluminum foil and irradiated concurrently. In some experiments, duplicate samples were irradiated for the same time and studied to verify experimental reproducibility. In a typical experiment, 6.25 mL of 4 mg/10 mL GO aqueous suspension and 5 mL of 100 mM H2O2 were mixed in a quartz tube by vigorous stirring. At the same time, for comparison, 6.25 mL of 4 mg/10 mL GO aqueous suspension and 10 mL of 50 mM K2S2O8 (PS) were also mixed concurrently. After the reaction, the samples were quenched with NaNO2 (aq) immediately and then dialyzed (100 Da) against ultrapure water for 7 days to remove residual H2O2, PS, and finally freeze-dried. 2.3. Instrumentations
2.3.1 UV-vis absorption spectrum A UV–visible spectrophotometer (Hitachi Model U-2910) was employed for absorbance measurements using quartz cells of 1 cm path length. Samples obtained at 0h, 1h, 2h and 4 h were analyzed immediately without any purification. 2.3.2 XPS X-ray photoelectron spectroscopy (XPS) measurement was carried out on a PHI-5000C ESCA system (Perkin-Elmer, USA). The samples were lyophilized for tests. The XPS spectra was analyzed and fitted using XPS peak 4.1 software. 7
2.3.3 TEM The nanostructures of graphene oxide ware characterized by using a transmission electron microscopy (TEM) system (JEOL, Model JEM-2100) operating at 200 kV. Approximately 10 µL of samples (aq) were drop-coated onto copper grids and dried under ambient conditions. 2.3.4 AFM Atomic force microscopic (AFM) images were acquired using Agilent 5500 AFM system (Agilent Technologies, USA) in tapping mode for thickness, phase, and sectional analysis. Approximately 10 µL of samples (aq) were drop-coated onto mica plates and dried under ambient conditions. 2.3.5 Raman Raman spectra were collected on a Renishaw InVia Reflex Raman spectrometer in a backscattering configuration. One drop of sample (aq) was drop-coated onto glass sheets and dried in a vacuum oven several times. 2.3.6 MALDI-TOF-MS The MS analysis was performed with 4800 Plus MALDI TOF/TOF Analyzer (AB Sciex). Mass spectra was acquired in the reflector mode in the m/z 50-1000 and 1000-5000 mass range. The samples were tested without any matrix. Approximately 10 µL of the sample (aq) was dropped onto a MALDI plate and dried under ambient conditions. 3. Results and discussion 3.1. UV-vis absorption spectrum 8
The preliminary data indicate that neither PS or H2O2 can effectively decompose GO within the experimental duration (see Fig. S1). All suspensions containing GO, H2O2-GO and PS-GO were exposed to UV irradiation for 4 h. The irradiation of the solutions containing H2O2 or PS generated highly reactive •OH or SO4•- radicals. UV exposure darkened the color of GO sample (Fig. S2), leading to a monotonic increase in the UV-vis absorbance (Fig. S3), which has been observed in previous reports (Hou et al., 2015; Matsumoto et al., 2010). Fig. 1 illustrates the UV-vis absorption spectra of GO suspension under UV/H2O2 and UV/PS after 0 h, 1 h, 2 h and 4 h of oxidation. GO shows two characteristic peaks in UV-vis spectra: the peak of 230 nm represents the Π→Π* transition of the C=C bonding and the shoulder at 300 nm is related to the n→Π* transition of the C=O groups [38]. In UV/H2O2 system, no obvious change occurred during the initial 3 h and a decrease in absorbance happened at 4 h. The absorbance increase resulting from the direct photolysis of GO may be offset over time by GO oxidation by •OH which can cause colour bleaching of GO. In contrast, a monotonic decrease in UV-Vis absorbance of GO sample in UV/PS system was observed. After 4 h, GO sample in this suspension turned colorless, implying a predominant role of SO4•- oxidation against direct photolysis. A high content of GO is likely to decrease the UV absorption of the oxidants, thereby affecting the degradation efficiency of GO. Three more levels of GO dosages (0.03, 0.08, 0.13 g/L) were examined. Figs. S4-S6 show that GO degradation was significantly affected by GO loading. However, the changes in UV-vis absorption spectra with time were very similar to those at a GO 9
loading of 0.05 g/L (Fig. 1), that is, GO in the UV/PS system is much more rapidly degraded than that in the UV/H2O2 system. Collectively, these comparisons demonstrate that to some extend the UV/PS process is more effective towards GO degradation than the UV/H2O2 process. Some studies reported that the distance of O-O bond in PS (1.497Å) is longer than in H2O2 (1.453Å), suggesting that PS could be more easily cleaved than H2O2 [39]. In addition, PS absorbed more strongly at 254 nm than H2O2 [40]. It is therefore supposed that PS can be activated more easily by UV.
Fig. 1. UV-visible absorption spectra of GO suspension collected after 0 h, 1 h, 2 h 10
and 4 h irradiation. (a) UV/H2O2; (b) UV/PS. 3.2 ．XPS
To probe into the chemical composition variation of the GO samples during the radical oxidation, the XPS spectra were acquired. Figures. 2a and 2b depict the XPS spectra of C 1s of GO before and after the UV/H2O2 and UV/PS reactions for 4 h. The C 1s spectra were deconvoluted into four sub-peaks corresponding to C atoms with a variety of chemical states. The peaks at 285.0, 286.8-287.1, 287.8-288.0 and 289.0-289.3 eV are assigned to C=C, C-C, C-H bonds, C-O (C-O-C and C-OH), C=O, and O=C-OH bands, respectively [33]. The XPS spectra show the substantial changes in GO surface O-functional groups upon oxidation treatment (Fig. 2c). The graphitic content of photolyzed GO samples increased from 23% to 61% in the UV/H2O2 system, from 24% to 72% in UV/PS system. In contrast, the overall oxygen containing functionalities (including C-O, C=O and O-C=O groups) decreased dramatically from 77% to 39% in the UV/H2O2 system, from 76% to 28% in UV/PS system. Disappearance of C/O containing groups caused GO mineralization and relative percentage increase in graphitic C=C, C-C and C-H groups.
11
Fig. 2. XPS spectra of C 1s of GO suspension obtained after 0 h and 4 h. (a) UV/H2O2.
(b) UV/PS. (c) the percentage of carbon groups. 12
3.3. Raman spectroscopy
Raman spectroscopy was employed to further analyze the degree of oxidation of GO by free radicals (Fig 3). The observed D and G bands are distinctive of graphitic materials: the D band generally appears at 1300-1400 cm-1 representing the disorder present in sp2-hybridized carbon systems and the G band generally appears at 1560-1620 cm-1 representing the stretching of C-C bonds [41, 42]. The intensity ratio (ID/IG) is usually used to evaluate the degree of disorder of graphitic materials, such as defects, ripples, and edges [38, 42]. The higher the ratio is, the more defects the material has. In our experiments, the calculated ratios were decreased from 0.983 to 0.971 and 0.938, after the two radical oxidation. This means that their degrees of order were both improved. The lower ID/IG value in UV/PS system compared with UV/H2O2 system indicates that the SO4•- radical is more efficient in removing oxygen functional groups of GO, is consistent with the XPS analysis.
Fig. 3. Raman spectra of pristine GO, UV/H2O2-4 h, UV/PS-4 h. 3.4. TEM 13
Using TEM, we observed morphologic changes of GO before and after the radical attack (Fig. 4). Fig. 4a illustrates the transparent and sheet-like structure of GO with large number of wrinkles and scrolls. The size of pristine GO was more than 500 nm and it was not a single layer. Fig. 4b shows many GO flakes were found in the UV/H2O2 process after 4 h. While under the same reaction time in UV/PS, these GO flakes became smaller which indicates the higher efficiency of UV/PS system for GO decomposition.
Fig. 4. TEM images of GO. (a) pristine GO; (b) UV/H2O2-4 h; (c) UV/PS-4 h. 3.5. AFM
AFM is employed to analyze the morphology and thickness of GO after the attack of the two free radicals. Oval and punctiform morphologies of GO samples 14
were observed in the UV/H2O2 and UV/PS system, respectively (Figs. 5a, 5b). The AFM thickness of the GO flakes in the UV/H2O2 treatment was about 1.3 nm (Fig. 5c), half of that after oxidation in the UV/PS system, demonstrating that GO samples after PS treatment consisted of an aggregation of flakes. According to the XPS data, the oxygen functional groups on the basal plane of GO were reduced. As a result, the electrostatic repulsion between the flakes of GO decreased while van der Waals forces dominated, leading to the formation of aggregates.
15
Fig. 5. Tapping mode AFM images of GO (a, c) UV/H2O2-4h; (b, d) UV/PS-4h.
3.6. MALDI-TOF-MS
The specific structure of GO can lead us to envision it as many fused polyaromatic hydrocarbons (PAHs), a significant environmental concern. Based on the understanding of its morphology and size, chemical functional groups on its surface, we trace its MS changes which could not be detected by TEM during the oxidation processes. MALDI-TOF-MS has an important advantage in analytical chemistry: absolute molecular weights of fragments can be determined, as opposed to obtaining relative molecular weights by chromatographic techniques. According to these obtained accurate molecular weights, one can get their molecular formulas (CxHyOz). The corresponding molecular structures are tentatively assigned during the molecular formulas and structural similarity with the parent compound. Figure 6 depicts the MALDI-TOF-MS spectra corresponding to products obtained at 0, 1, 2 and 4 h after the beginning of UV/H2O2 or UV/PS reaction. The distinct differences between the two processes were: low molecular weight (LMW) products and the purity of suspension which is indicative of the reaction efficiency. Because of the complexity of the data, only selected compounds were chosen for chemical composition and possible structural analysis. In the process of UV/H2O2, peaks between m/z 49 to 200 mainly disappear after 1 h, implying LMW substances were degraded quickly. Known from peaks at m/z 701 ([C40H28O12 + H]+) and 537 ([C36H40O4 + H]+), they were degraded to form other species, such as peak at m/z 489 16
([C25H28O10+H]+) which represents the most abundant compound in the system. New peaks lower than 200 m/z appeared and maintained during the last two hours. However, no peaks lower than m/z 202 (C14H18O) were found in samples treated by UV/PS. Moreover, in the latter two hours of experiment, no peaks smaller than m/z 263 ([C12H6O7 + H]+) were detected. The MS results prove that the UV/PS system is more efficient in the degradation of GO. In other words, this means applications of GO should be taken more concerns where SO4•- radical exists. Changes of peaks in the range of m/z 1000-5000 were also considered. Different to the results of Bai et al., the changes of carbon clusters were not distinct [31]. The possible reasons are that there were no pretreatments of our samples before the MS test and it records the whole compounds the instrument can detect. But from the signal intensity, it is clear that differences do occur during the reaction. In the UV/PS system at 4 h, there was no signal, meaning no compounds in m/z 1000-5000 can be found (Figs. S7 and S8). Some tentative chemical structures of typical compounds in those systems are suggested. However, considering the existence of allotropes, further efforts are needed to identify the decomposed byproducts of GO.
17
Fig. 6. MALDI-TOF-MS spectra for GO samples after 0, 1, 2 and 4 h. (a) UV/H2O2;
(b) UV/PS; (c) possible chemical structures of products observed in spectra. 18
3.6. Mechanism of GO oxidative degradation pathway in •OH and SO4•- radicals
Recently, reactions of graphene-related materials with
•
OH have been
investigated under different scenarios, such as engineering nano-porous GO, degrading GO in the wastewater and the transformation of GO in the natural environments [30, 32, 43]. Graphene oxide would be degraded in the presence of Fenton reagents (Fe2+/3+, H2O2) under UV irradiation. Radich et al. reported that rGO was bleached with a significant structural degradation upon reaction with •OH [28]. Yu et al. attempted to etch GO sheets by •OH generated from water radiolysis by g-irradiation and nano-porous GO sheets with well controlled nano-pores and ‘lacelike’ edges [43]. However, there are few studies on reactions of GO with SO4•radicals. It is generally accepted that •OH and SO4•- radicals react with organic compounds by three ways: electron transfer, hydrogen abstraction and addition mechanisms [44]. Among these, •OH is more likely to participate in hydrogen abstraction or addition reactions, while SO4•- is apt to electron transfer reactions [45]. Yu et al. reported that hydroxylation of •OH is a major and crucial step for initial GO oxidation [43]. Hydroxyl (•OH) undergoes electrophilic addition on unsaturated bonds to produce hydroxylated GO, which would be further oxidized to quinones, then carboxylic acids. Hydroxyl (•OH) imparts carboxylic acids at the defect sites, leading to mineralization of those oxygen moieties [46], as evidenced from TOC and XPS analyses. UV/PS system seems more efficient towards GO degradation, because SO4•- (E=2.6-3.1 V vs NHE) could be more reactive than •OH (E=1.9-2.7 V vs NHE) [47]. To compare their 19
oxidation capacity of UV/H2O2 and UV/PS systems, degradation performance of a typical dye, AO7 was evaluated with the same level of the oxidant. It is observed that AO7 was much rapidly degraded in the UV/PS system (Fig. S9), implying that UV/PS process is more efficient to generate radicals and oxidize AO7 and GO. Therefore, it is believed that SO4•--induced degradation can result in a better mineralization rate than •OH [48]. On the other hand, SO4•- is an electrophilic oxidant, and prefers to react with electron-donating groups, such as hydroxyl (-OH) and organic compounds containing unsaturated bonds. Therefore, SO4•- is supposed to selectively attack C=C bonds on the GO structure, facilitating the GO destruction and cleavage.
4. Conclusion
This study examined the structural changes of GO in UV/H2O2 and UV/PS processes and found that a monotonic decrease in UV-vis absorbance of aqueous GO sample was observed in a UV/PS system, while no obvious changes of UV-vis spectra occurred in a UV/H2O2 solution. GO was decomposed to small flakes after oxidation and some low-molecular-weight molecules. Collectively, UV/PS process seems more effective towards GO decomposition than the UV/H2O2 system, because of stronger oxidation potential and electrophilicity of SO4•- radicals. This study provides, to the best of our knowledge, the first evidence for GO’s instability in aqueous solution where highly reactive radicals are abundant, although its roles in enhancing heterogeneous catalytic activities are widely recognized. GO would be destructed upon the radical attack when it is used as a support or catalyst in producing SO4•20
radicals from persulfate or peroxymonosulfate activation. Therefore, special attention on GO stability should be given when GO-based nanocomposites are used in advanced oxidation processes (AOPs) where highly reactive radicals are abundant. On the other hand, UV/PS technology can be envisaged as a novel method for “etching” GO sheets and modifying porous structures of GO due to the stronger oxidative capacity of SO4•- radicals than hydroxyl radicals generated in water radiolysis [43]. This engineering process can be conducted under a mild condition only by a simple control of the treatment time. In addition, comparative toxicity assessment of GO suspensions and destructed GO would be a very interesting topic and deserves in-depth investigations [49].
Acknowledgments
This work was supported by the National Natural Science Foundation of China (NSFC) (Nos. 21377023 and 21677031), National Key Research and Development Program of China (2016YFC0400501/2016YFC0400509), Shanghai Pujiang Program and DHU Distinguished Young Professor Program. The authors thank Dr. Nick Ward at Southern Cross University for his careful proofreading. Appendix A. Supplementary material
Supplementary data associated with this article can be found, in the online version. Reference
[1] S. Stankovich, D.A. Dikin, G.H.B. Dommett, K.M. Kohlhaas, E.J. Zimney, 21
E.A. Stach, R.D. Piner,
S.T. Nguyen,
R.S. Ruoff, Nature 442 (2006)
282-286. [2] Compton, S.T. Nguyen, Small 6 (2010) 711-723. [3] A. Dimiev, D.V Kosynkin, L.B. Alemany, P. Chaguine, J.M. Tour, J. Am. Chem. Soc. 134 (2012) 2815-2822. [4] D. Chen, H. Feng, J. Li, Chem. Rev. 112 (2012) 6027-6053. [5] C. Chung, Y.K. Kim, D. Shin, Acc. Chem. Res. 46(2013), 2211-2224. [6] F. Peng, T. Luo, L. Qiu, Y. Yuan, Mater. Res. Bulletin. 48 (2013) 2180-2185. [7] H. Yan, X. Tao, Z. Yang, K. Li, H. Yang,
A. Li, R. Cheng, J. Hazard. Mater.
268 (2014) 191-198. [8] M.L. Yola, T. Eren, N. Atar, Sens. Actuators. B: Chem. 210 (2015) 149-157. [9] M.L. Yola, V.K. Gupta, T. Eren, A.E. Şen, N. Atar, Electrochim. Acta. 120 (2014) 204-211. [10] T.F. Yeh, J.M. Syu, C. Cheng, T.H. Chang, H. Teng, Adv. Funct. Mater. 20 (2010) 2255-2262. [11] G. Chen, M. Sun, Q. Wei, Y. Zhang, B. Zhu, B. Du, J. Hazard. Mater. 244 (2013) 86-93. [12] P. Gao, J. Liu, D.D. Sun, W. Ng, J. Hazard. Mater. 250 (2013) 412-420. [13] S. Guo, G. Zhang, Y. Guo, J.C Yu, Carbon 60 (2013) 437-444. [14] Y.C. Lee, S.J. Chang, M.H. Choi, T.J Jeon, T. Ryu, Y.S. Huh, Appl. Catal. B: Environ. 142 (2013) 494-503. [15] N.A. Zubir, C. Yacou, X. Zhang, J.C. Diniz da Costa, J. Environ. Chem. Eng. 22
2 (2014) 1881-1888. [16] K.V. Voitko, R.L. Whitby, V.M. Gun’ko, O.M. Bakalinska, M.T. Kartel, K. Laszlo, A.B. Cundy, S.V. Mikhalovsky, J. Colloid. Interf. Sci. 361 (2011) 129-136. [17] F. Ghanbari, M. Moradi, Chem. Eng. J. 310 (2017) 41-62. [18] Z. Wang, R. Yuan, Y. Guo, L. Xu, J. Liu, J. Hazard. Mater. 190 (2011) 1083-1087. [19] G.P. Anipsitakis, D.D. Dionysiou, Appl. Catal. B: Environ. 54 (2004) 155-163. [20] R.L. Johnson, P.G. Tratnyek, R.O.B Johnson, Environ. Sci. Technol. 42 (2008) 9350-9356. [21] Furman, A.L. Teel, R.J. Watts, Environ. Sci. Technol. 44 (2010) 6423-6428. [22] M.G. Antoniou, A.A. de la Cruz, D.D. Dionysiou, Appl. Catal. B: Environ. 96 (2010) 290-298. [23] P. Shi, R. Su, F. Wan, M. Zhu, D. Li, S. Xu, Appl. Catal. B: Environ. 123-124 (2012) 265-272. [24] W. Peng, S. Liu, H. Sun, Y. Yao, L. Zhi, S. Wang, J. Mater. Chem. A 1 (2013) 5854-5859. [25] H. Sun, Y. Wang, S. Liu, L. Ge, L. Wang, Z. Zhu, S. Wang, Chem. Comm. 49 (2013) 9914-9916. [26] P.M. Ajayan, B.I. Yakobson, Mater. Sci: Nat. 441 (2016) 818-819. [27] O. Akhavan, M. Abdolahad, A. Esfandiar, M. Mohatashamifar, J. Phys. Chem. C 114 (2010) 12955-12959. 23
[28] J.G. Radich, A.L. Krenselewski, J. Zhu, P.V. Kamat, Chem. Mater. 26 (2014) 4662-4668. [29] C.Z. Zhang, T. Li, Y. Yuan, J. Xu, Chemosphere 153 (2016) 531-540. [30] X. Zhou, Y. Zhang, C. Wang, Acs Nano 6 (2012) 6592-6599. [31] H. Bai, W. Jiang, G.P. Kotchey, W.A. Saidi, B.J. Bythell, J.M. Jarvis, A.G. Marshall, R.A. Robinson, A. Star, J. Phys. Chem. C: Nano. Interfac. 118 (2014) 10519-10529. [32] W.C. Hou, I. Chowdhury, D.G Goodwin, Jr, W.M. Henderson, D.H. Fairbrother, D. Bouchard, R.G. Zepp, Environ. Sci. Technol. 49 (2015) 3435-3443. [33] Y. Matsumoto, M. Koinuma, S. Ida, S. Hayami, T. Taniguchi, K. Hatakeyama, H. Tateishi, Y. Watanabe, S. Amano, J. Phys. Chem. C 115 (2011) 19280-19286. [34] M. Minella, M. Demontis, M. Sarro, F. Sordello, P. Calza, C. Minero, J. Mater. Sci. 50 (2015) 2399-2409. [35] G.P. Kotchey, B.L. Allen, H. Vedala, ACS. Nano. 5(2011), 2098-2108. [36] Y. Guo, J. Zhou, X. Lou, R. Liu, D. Xiao, C. Fang, Z. Wang, J. Liu, Chem. Eng. J. 254 (2014) 538-544. [37] X. Lou, D. Xiao, C. Fang, Z. Wang, J. Liu, Y. Guo, S. Lu, Environ. Sci. Pollut. Res. 23 (2015) 4778-4785. [38] Y. Zhao, C.T. Jafvert, Environ. Sci.: Nano 2 (2015) 136-142. [39] S. Yang, P. Wang, X. Yang, L. Shan, W. Zhang, X. Shao, R. Niu, J. Hazard. 24
Mater. 179 (2010) 552-558. [40] X. Gu, S. Lu, Z. Qiu, Q. Sui, Z. Miao, K. Lin, Y. Liu, Q. Luo, Ind. Eng. Chem. Res. 51 (2012) 7196-7204. [41] L.M. Malard, M.A. Pimenta, G. Dresselhaus, M.S. Dresselhaus, Phys. Rep. 473 (2009) 51-87. [42] G.-H. Moon, Y. Park, W. Kim, W. Choi, J. Carbon 49 (2011) 3454-3462. [43] C. Yu, B. Zhang, F. Yan, J. Zhao, J. Li, L. Li, J. Li, Carbon 105 (2016) 291-296. [44] C. Liang, Z.S. Wang, C.J. Bruell, Chemosphere 66 (2007) 106-113. [45] Y.T. Lin, C.
Liang, J.H. Chen, Chemosphere 82 (2011) 1168-1172.
[46] W. Li, Y. Bai, Y. Zhang, M. Sun, R. Cheng, X. Xu, Y. Chen, Y. Mo, Synth. Metal. 155 (2005) 509-515. [47] W.-D. Oh, Z. Dong, T.-T. Lim, Appl. Catal. B: Environ. 194 (2016) 169-201. [48] M. Sanchez-Polo, M.M. Abdel daiem, R. Ocampo-Perez, J. Rivera-Utrilla, A. J. Mota, Sci. Total Environ, 463-464 (2013) 423-431. [49] K. Yang, Y. Li, X. Tan, R. Peng, Z. Liu, Small 9 (2013) 1492-1503.
25
Graphical abstract
UV/H2O2
GO
UV/PS
26