- Email: [email protected]
Electricity generation across graphene oxide membranes
Accepted Manuscript Title: Electricity generation across graphene oxide membranes Authors: Lu Huang, Junxian Pei, Haifeng Jiang, Changzheng Li, Xuejiao Hu PII: DOI: Reference:
S0025-5408(17)30721-3 http://dx.doi.org/10.1016/j.materresbull.2017.08.049 MRB 9528
To appear in:
MRB
Received date: Revised date: Accepted date:
21-2-2017 23-8-2017 26-8-2017
Please cite this article as: Lu Huang, Junxian Pei, Haifeng Jiang, Changzheng Li, Xuejiao Hu, Electricity generation across graphene oxide membranes, Materials Research Bulletinhttp://dx.doi.org/10.1016/j.materresbull.2017.08.049 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.
Electricity generation across graphene oxide membranes
Lu Huang, Junxian Pei, Haifeng Jiang*, Changzheng Li, Xuejiao Hu* Key Laboratory of Hydraulic Machinery Transients of Ministry of Education, School of Power and Mechanical Engineering, Wuhan University, Wuhan, Hubei 430072, China *Corresponding authors. E-mail addresses: [email protected] (H. Jiang), [email protected] (X. Hu).
Graphical abstract
1
Highlights Three factors were studied to control the electricity generation using GO films. A novel energy harvesting device with GO films was demonstrated. This provided new insights into the utilization of GO films for energy harvesting.
ABSTRACT Currently, energy conversion methods in nanochannels have been developed for both fundamental sciences and technological applications. Graphene oxide membranes with excellent properties possess potential applications in numerous fields. Due to the confinement of nanocapillaries within graphene oxide membranes and the mobility difference of ions, an electricity generation process has been achieved using graphene oxide membranes. In this study, for the first time, the influence of source concentration, applied pressure and membrane thickness was investigated on electricity generation in aqueous environment. The results revealed that these three factors could successfully control the electricity generation with graphene oxide membranes. Electrical potential difference would become larger with increasing the source concentration and membrane thickness. In contrast, the applied pressure was not well-suited to capture energy from sources. This provided new insights into the design and utilization of graphene oxide films for highly efficient energy harvesting.
Keywords: A. nanostructures, A. thin films, B. microstructure, D. energy storage, D. ionic conductivity.
1. Introduction Graphene oxide (GO) attracted recently significant attention in the field of water purification [1,2],
2
ionic and molecular sieving [3,4], gas leakage testing [5] and desalination [6-14], because of its remarkable properties such as adequate mechanical strength [5,15] and excellent ion selectivity [2-4,16]. Typically, GO sheets with polar oxygenated functional groups on the basal plane and the edges, resulting in the formation of sp2-sp3 hybridized structure, could hold them to establish lamellar structure with stable nanochannels [17, 18]. These nanocapillaries allow unimpeded permeation of water but are completely impermeable to liquids, vapors, and gases [5, 19-21]. On the other hand, due to the deprotonation of oxygenated functional groups which are attached to GO surface, the films are negatively charged, resulting in the strong attracting force to cations by the contrast of the significant repulsion force to anions. Therefore, some researchers hold an optimistic attitude that GO membranes could act as cation-exchange membranes for separation and filtration [22-24], and even electricity generation [20, 25]. Recent studies demonstrated that graphene-based materials provided considerable opportunities for development of future-generation energy-transform devices [20, 25, 26]. Guo et al. reported a twodimensional (2D) nanofluidic generator system based on the electrokinetic ion transportation through a layered graphene hydrogel membrane. In aqueous conditions, 2D nanocapillaries were formed between adjacent graphene sheets and the negatively charged 2D nanocapillaries showed surface-chargegoverned ion transportation that preferentially permeated counter-ions and rejected co-ions [26]. This work provided a new idea to design GO-based membranes for energy-transform devices. Subsequently, Zhao et al. presented a power-generating phenomenon of single GO film with a preformed oxygencontaining group gradient (g-GOF) and the as-prepared g-GOF was able to provide moisture-enabled voltage output with efficient energy conversion efficiency [25]. Sun et al. reported that it was practical to realize synchronous energy harvesting and ion separation through nanochannels within GO films during the process of ion trans-membrane permeation [20]. The presence of GO membrane between sources and drains causes significant confinement to the ion free-migration from source to drain, which results in the establishment of electrochemical dialysis equilibrium and the formation of a space charge zone [27].
3
The different interactions of cations and anions with GO membranes give rise to the unequal diffusivities of them, respectively. These results that an excessive amount of cations and anions present in drains and sources as well as on both surfaces of GO membranes, respectively, lead to the generation of transmembranes electrical potential and Donnan potential [28]. Electrical potential differences across source and drain solutions can be sustained for a long time due to the strong confining effect of GO sheets [27]. Few previous studies have investigated the influence of some factors on electricity generation with GO membranes such as the mobility difference of ions, the intensity of magnetic fields and temperature difference etc. Sun et al. indicated that GO membranes could find potential applications in the purification of wastewater, while producing electricity simultaneously [20]. The results demonstrated that applying magnetic fields and improving temperature could increase electrical potential differences. Moreover, the mobility difference of different ions would cause electrical potential differences, and hybridization of various source solutions could also improve electrical potential difference. In this work, for the first time, we found that the source concentration, the applied pressure and membrane thickness could successfully give birth to great change of electrical potential differences in energy harvesting using GO membranes. This provided a new avenue for designing GO films for efficient energy harvesting.
2. Experimental procedure 2.1. Preparation of well-stacked GO membranes The preparation of GO was carried out by a modified Hummer Method using natural flake graphite (325 mesh), which was exposed to a mixture of concentrated sulfuric acid, phosphoric acid and potassium permanganate for oxidation according to our previous method [29]. As-prepared GO flakes were dispersed in ultrapure water by vigorous sonication (125 W for 5 min) to form a stable suspension with a concentration of 2 mg·mL-1. At first, we produced a series of different thicknesses of GO membranes assembled on the nylon micro-filters (diameter: 50 mm, pore size: 0.2 μm) by vacuum 4
filtration [30-32]. In order to improve GO membranes stability in solutions, GO films were not exfoliated from the support micro-filter. The thickness of GO films was turned by just controlling the volume of the filtrated 2 mg·mL-1 GO aqueous suspension. The interlayer spacing of the obtained GO membrane was characterized by X-ray diffraction (XRD, X' Pert Pro, Panalytica) in the dry state. The morphologies of membranes were examined by the scanning electron microscopy (SEM, SIGMA, Carl Zeiss).
2.2. Experimental Setups The electricity generation experiments were carried out with two sets of self-made apparatus. We used the same apparatus to investigate the effect of the membrane thickness and source concentration, and it consisted of a feed reservoir and a permeate reservoir separated by a PMMA plate with a leak hole (1.2 mm in diameter) in the center，as shown in Figure 1 (a, b). The studied GO membranes were sandwiched between two pieces of identical copper foils with an opening of 5 mm in diameter to ensure the GO films directly contact the solutions in the sources and drains. Two pieces of copper foils were clamped between two seal rings, which formed a tight seal between the two compartments. In order to eliminate the influence of micro-filters support, we performed control experiments using bare nylon micro-filters with other conditions remaining the same. In a typical experiment, we filled one of the compartment (referred to as feed reservoir abovementioned) with 100 ml of certain concentration MgCl2 source solution. Equivalent volume deionized water was injected into the other (permeate) compartment with the same speed. Meanwhile, gentle magnetic stirring was applied in both feed and permeate solutions to avoid concentration polarization effect. During the penetration process, the conductivities of the drains were measured by a conductivity meter (LEICI, DDSJ-308A). The voltage ΔVDS (ΔVDS = VD - VS. In this equation, VD and VS are the electrical potential of drains and sources respectively, ΔVDS is the electric potential difference between the drains and sources) was measured by a pair of identical silver electrodes (99.99% in purity) which 5
were fastened to the PMMA plate and separated by the GO film in the middle with a total distance of 12 mm. The voltage was measured by the multimeter (UNI-T, UT805A). Before the permeation experiments, a thin tape was glued to one of copper foils to substitute the GO membrane with other steps remaining the same. No obvious conductivity variations in drains occurred with time using a high concentration solution in the source compartment, which illustrated that the experimental setup possessed an excellent seal. However, due to study the effect of pressure on electricity generation with graphene oxide membranes, the pressure-driven electricity generation experiments were performed using the second sets of apparatus. The set of apparatus was pressurized with N2 flow. The schematic diagram and photograph of pressuredriven device are showed in Figure 1 (c, d). In a typical experiment, the whole device was placed on a flat plane, and then 50 ml of a certain 0.1M HCl feed solution and deionized water were injected into the source and drain compartments from respective circular tube in the vertical direction with the same speed. The feed solution was allowed to withstand different N2 flow pressure in the horizontal direction. Meanwhile, these nanochannels within GO membranes would collapse and shrink at high pressures [33]. Therefore, a low pressure of less than 0.2 MPa was applied in this study in order to avoid destroying the structural stability of GO interlayer nanochannels.
3. Results and discussion A typical preparation of GO membranes was indicated by vacuum filtration with 0.75 ml suspension from 2 mg·mL-1 original prepared solution, as shown in Figure 2a. The well-stacked GO films are highly stable in water for a long time, as illustrated in Figure 2b. The result of XRD revealed that the interlayer spacing of the GO membrane was about 0.857 nm (Fig. 2c). The SEM image of the cross-sections indicated that the thinnest thickness of studied GO flakes from 0.25 mL original 2 mg·mL-1 GO suspensions was about 1 μm (Fig. 2d). Before formal experiments, the control experiment was performed to investigate the effect of nylon micro-filter, which indicated that it could also produce
6
voltage with bare micro-filter because of the effect of electrical double layer [34-36]. The voltage generated consistently dropped over time but the voltage generated of GO membranes can maintain a relatively stable value after 1.5 h as shown in Fig. 3a. The mechanism of the voltage generated from GO membrane and microfilter attributes to the effect of electric double layer in nanochannels and microchannels [26, 37, 38]. For GO membranes, due to the deprotonation of oxygenated functional groups which are attached to GO surface, the GO surface is negatively charged. The negatively charged 2D nanocapillaries show surface-charge-governed ion transportation that preferentially permeate counter-ions and exclude co-ions [20, 26]. Consequently, such mobility difference of cations and anions would lead to the presence of an excess amount of cations or anions in both drains and sources, which in turn would attain the establishment of an electrochemical dialysis equilibrium [20, 27]. Ultimately, a Donnan potential is generated on the GO surfaces, thus the voltage generated can maintain a stable value after 1.5 h. But for common nylon micro-filters, the pore size is 0.22 μm (much larger than ion diameters). Most cations and anions hardly interact with the microfilters surface because of the broad microchannel and weaker surface charge relative to GO films which possess narrow nanochannel and higher surface charge. Therefore the voltage generated of nylon microfilter consistently dropped over time but the voltage generated of GO membranes can maintain a relatively stable value after 1.5 h. In short, compared with GO films, the common nylon-filter are unsuitable for the electricity generation during the process of transmembrane transport in practice. Firstly, the effect of GO membrane thickness on the generated voltage between the sources and drains was investigated (Fig. 3b), and it exhibited that the permeation of source ions effectively decreased with the increasing of the GO membrane thickness by observing the conductivity of drains. Apart from that, we found the voltage across the source and drain solution could adequately increase with increasing the GO membranes thickness. The results revealed that the mobility difference of cations and anions led to variations of the voltage generated, and the nanochannels length within GO membranes dominated the mass transport process and the ion selectivity [39]. The nanochannels length would increase with the
7
increasing of the GO membrane thickness, which enhanced the selectivity of cation and anion ions. Therefore, the thicker GO membranes with longer nanochannels could cause a greater mobility difference, resulting in a larger ΔVDS. However, the growth of total permeation ions by reducing the GO membranes thickness could not improve the voltage generated.
Next, the influence of the source concentration was studied on ΔVDS, as illustrated in Figure 3c. The result appeared that increasing the source concentration could effectively improve the voltage generated, but the increase of voltage became extremely not obvious when the concentration increased to a certain degree. That’s because during the ion trans-membrane transport process, the negatively charged GO membranes repulsed anions and attracted cations, thus cations would be the first ions transported into membranes through GO nanocapillaries [20]. Moreover, the total permeation of cations would also increase with the growth of source concentration due to a higher diffusion driving force. Consequently, compared with low source concentration, more cations would be enriched on GO surfaces that directly contact with drain solutions. In addition, the anions trans-membrane transport was driven by the electrostatic attraction force from the corresponding cations embedded in GO sheets and penetrate through the negatively charged GO membranes passively, resulting in excessive anions enriched on GO surfaces that directly contact with source solutions. Due to the presence of Donnan effect, an electrochemical dialysis equilibrium would be established and the generated voltage could tend to be a stable value after about one hour [20, 27]. Thus more cations and anions would enrich in drains and sources respectively with increasing the source concentration, which ultimately producing a lager ΔVDS. Furthermore, the effect of the applied pressure was examined using another set of experimental device (Fig. 1 (c, d)). According to recent studies on the pressure-driven separation performance of GO films,
8
Huang et al. has reported that nanochannels within GO membranes were relatively stable at the low pressure range due to their rigidity and would begin to collapse and shrink at high pressures [33]. Therefore, we applied a low pressure of less than 0.2 MPa to keep the structural stability of interlayer nanochannels. Sun et al. has reported that the applied pressure weakened water-ion interactions confined in GO nanochannels and further decreased the selectivity of cations and anions, resulting in poor source solution rejection [39]. In our pressure-driven experiments, the result illustrated that the voltage would decrease when the pressure was applied on feed solution. At the same time, the generated voltage had a drop tendency with the increase of the applied pressure, as illustrated in Fig. 3d. From the experimental phenomenon, it proved to be convincing that the salt rejection of GO films was poor in pressure-driven separation performance and the selectivity of ions would decrease with increasing the applied pressure. The results confirmed that increasing the thickness of GO films could sharply weaken the permeation of source ions. However, electrical potential differences would effectively increase with the increase of GO membrane thickness. In the meanwhile，increasing the source solution concentration could increase concentration gradient between the source and drain solutions, which would enhance electrical potential differences. Remarkably, electrical potential differences unexpectedly reduced while pressure was applied on GO membranes.
4. Conclusions In summary, the effect of the membrane thickness, solution concentration and applied pressure on electricity generation with GO membranes has been studied, revealing that tuning each of three factors could affect the generated voltage. Applying thicker GO films and highly-concentrated source solutions could efficiently increase voltage generated, which has a significant engineering application. In the meanwhile, our experimental results prove that it is beneficial to avoid applying pressure on electricity generation with GO membranes because the applied pressure decreases the rejection of solution and weakens the selectivity of cations and anions. The results presented in this study indicate that there is a further breakthrough for GO membranes to realize efficient energy conversion. In addition, we can roughly identify its concentration based on the difference of generated voltage relative to the same source solution.
Acknowledgements 9
The authors acknowledge the support of the National Natural Science Foundation of China (No. 51706157), the China Postdoctoral Science Foundation (No. 2017M612498), and the Fundamental Research Funds for the Central Universities (No. 2042016kf0023).
References [1] T. Ramanathan, A. A. Abdala, S. Stankovich, D. A. Dikin, M. Herrera-Alonso, R. D. Piner, D. H. Adamson, H. C. Schniepp, X. Chen, R. S. Ruoff, Nat. Nanotechnol. 3 (2008) 327-331. [2] F. Peng, T. Luo, L. Qiu , Y. Yuan, Mater. Res. Bull. 48 (2013) 2180-2185. [3] H. Li, Z. Song, X. Zhang, Y. Huang, S. Li, Y. Mao, H. J. Ploehn, Y. Bao, M. Yu, Science 342 (2013) 95-98. [4] B. Mi, Science 343 (2014) 740-742. [5] R. R. Nair, H. A. Wu, P. N. Jayaram, I. V. Grigorieva, A. K. Geim, Science 335 (2012) 442-444. [6] D. Cohen-Tanugi, J. C. Grossman, Nano Lett. 12 (2012) 3602-3608. [7] S. P. Surwade, S. N. Smirnov, I. V. Vlassiouk, R. R. Unocic, G. M. Veith, S. Dai, S. M. Mahurin, Nat. Nanotechnol. 10 (2015) 459-464. [8] K. P. Lee, T. C. Arnot, D. Mattia, J. Membr. Sci. 370 (2011) 1-22. [9] C. Charcosset, Desalination 245 (2009) 214-231. [10] A. Nicolai, B. G. Sumpter, V. Meunier, Phys. Chem. Chem. Phys. 16 (2014) 8646-8654. [11] H. Deng, P. Sun, Y. Zhang, H. Zhu, Nanotechnology 27 (2016) 274002. [12] M. Bhadra, S. Roy, S. Mitra, Desalination 378 (2016) 37-43.
10
[13] D. An, L. Yang, T. J. Wang, B. Liu, Ind. Eng. Chem. Res. 55 (2016) 4803-4810. [14] D. Cohen-Tanugi, L. C. Lin, J. C. Grossman, Nano Lett. 16 (2016) 1027-1033. [15] Y. Jiang, W. N. Wang, D. Liu, Y. Nie, W. Li, J. Wu, F. Zhang, P. Biswas, J. D. Fortner, Environ. Sci. Technol. 49 (2015) 6846-6854. [16] P. Sun, H. Liu, K. Wang, M. Zhong, D. Wu, H. Zhu, Chem. Commun. 51 (2015) 3251-3254. [17] K. P. Loh, Q. Bao, G. Eda, M. Chhowalla, Nat. Chem. 2 (2010) 1015-1024. [18] G. Eda, M. Chhowalla, Adv. Mater. 22 (2010) 2392-2415. [19] B. Chen, H. Jiang, X. Liu, X. Hu, J. Phys. Chem. C, 121 (2017) 1321-1328. [20] P. Sun, F. Zheng, M. Zhu, K. Wang, M. Zhong, D. Wu, H. Zhu, Sci. Rep. 4 (2014) 5528-5532. [21] S. C. O'Hern, M. S. Boutilier, J. C. Idrobo, Y. Song, J. Kong, T. Laoui, M. Atieh, R. Karnik, Nano Lett. 14 (2014) 1234-1241. [22] T. Sata, T. Sata, W. Yang, J. Membr. Sci. 206 (2002) 31-60. [23] P. M. V. D. Velden, C. A. Smolders, J. Appl. Polym. Sci. 21 (1977) 1445-1457. [24] H. Strathman, A. Grabowski, G. Eigenberger, Ind. Eng. Chem. Res. 52 (2013) 10364-10379. [25] F. Zhao, H. Cheng, Z. Zhang, L. Jiang, L. Qu, Adv. Mater. 27 (2015) 4351-4357. [26] W. Guo, C. Cheng, Y. Wu, Y. Jiang, J. Gao, D. Li, L. Jiang, Adv. Mater. 25 (2013) 6064-6068. [27] P. Sun, H. Deng, F. Zheng, K. Wang, M. Zhong, Y. Zhang, F. Kang, H. Zhu, 2D Mater. 1 (2014) 034004. [28] H. Ohshima, S. Ohki, Biophys J. 47 (1985) 673-678. [29] J. Pei, X. Zhang, L. Huang, H. Jiang, X. Hu, RSC Adv. 6 (2016) 101948-101952. [30] B. S. Kong, J. Geng, H. T. Jung, Chem. Commun. 16 (2009) 2174-2176. [31] S. Zhang, Y. Li, N. Pan, J. Power Sources 206 (2012) 476-482. [32] Q. Chen, P. Liu, C. Sheng, L. Zhou, Y. Duan, J. Zhang, RSC Adv. 4 (2014) 39301-39304. [33] H. Huang, Y. Mao, Y. Ying, Y. Liu, L. Sun, X. Peng, Chem. Commun. 49 (2013) 5963-5965. [34] S. T. Senthilkumar, B. Senthilkumar, S. Balaji, C. Sanjeeviraja, R. K. Selvan, Mater. Res. Bull. 46
11
(2011) 413-419. [35] K. Liu, P. Yang, S. Li, J. Li, T. Ding, G. Xue, Q. Chen, G. Feng, J. Zhou, Angew. Chem. Int. Ed. 55 (2016) 8003-8007. [36] G. Xue, Y. Xu, T. Ding, J. Li, J. Yin, W. Fei, Y. Cao, J. Yu, L. Yuan, L. Gong, J. Chen, S. Deng, J. Zhou, W. Guo, Nat. Nanotechnol. 12 (2017) 317-322. [37] D. Grahame, Chem. Rev. 41 (1947) 441-501. [38] T. Ding, K. Liu, J. Li, G. Xue, Q. Chen, L. Huang, B. Hu, J. Zhou, Adv. Funct. Mater. 27 (2017), 1700551. [39] P. Sun, R. Ma, H. Deng, Z. Song, Z. Zhen, K. Wang, T. Sasaki, Z. Xu, H. Zhu, Chem. Sci. 7 (2016) 6988-6994.
12
Fig. 1 (a, b) The photograph of electricity generation experimental apparatus for studying the effect of the membrane thickness and solution concentration. (c) The photograph of the pressure-driven electricity generation experimental setup. (d) Schematic diagram of pressure-driven filtration experimental setup for electricity generation.
13
Fig. 2 (a) The photograph of GO membranes and the inset shows corresponding suspension. (b) The photograph shows the stability of GO papers by vacuum-filtrated through nylon microfilters in water after 1 day. (c) XRD patterns of GO films in a dry state. (d) The corresponding cross-section SEM characterization of as-prepared GO membranes.
14
Fig. 3 (a) Voltage generations across sources and drain (ΔVDS) of bare microfilters and GO on microfilters with 1M MgCl2 source solution in control experiments. (b) Voltage generations of different thickness GO films across the source and drain solution, and corresponding conductivities of the drains for 3 h. (c) Stable voltage generations across GO membranes based on the penetrations of different concentration MgCl2 feed solutions for 3 h. (d) Voltage generations across GO membranes based on 0.1M HCl feed solution with different pressure for 3 h. The error bars come from the corresponding error estimations and are formed by three times experimental results. The error bar of conductivities variations is neglected because of a very slight variation in experiments.
15
S0025-5408(17)30721-3 http://dx.doi.org/10.1016/j.materresbull.2017.08.049 MRB 9528
To appear in:
MRB
Received date: Revised date: Accepted date:
21-2-2017 23-8-2017 26-8-2017
Please cite this article as: Lu Huang, Junxian Pei, Haifeng Jiang, Changzheng Li, Xuejiao Hu, Electricity generation across graphene oxide membranes, Materials Research Bulletinhttp://dx.doi.org/10.1016/j.materresbull.2017.08.049 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.
Electricity generation across graphene oxide membranes
Lu Huang, Junxian Pei, Haifeng Jiang*, Changzheng Li, Xuejiao Hu* Key Laboratory of Hydraulic Machinery Transients of Ministry of Education, School of Power and Mechanical Engineering, Wuhan University, Wuhan, Hubei 430072, China *Corresponding authors. E-mail addresses: [email protected] (H. Jiang), [email protected] (X. Hu).
Graphical abstract
1
Highlights Three factors were studied to control the electricity generation using GO films. A novel energy harvesting device with GO films was demonstrated. This provided new insights into the utilization of GO films for energy harvesting.
ABSTRACT Currently, energy conversion methods in nanochannels have been developed for both fundamental sciences and technological applications. Graphene oxide membranes with excellent properties possess potential applications in numerous fields. Due to the confinement of nanocapillaries within graphene oxide membranes and the mobility difference of ions, an electricity generation process has been achieved using graphene oxide membranes. In this study, for the first time, the influence of source concentration, applied pressure and membrane thickness was investigated on electricity generation in aqueous environment. The results revealed that these three factors could successfully control the electricity generation with graphene oxide membranes. Electrical potential difference would become larger with increasing the source concentration and membrane thickness. In contrast, the applied pressure was not well-suited to capture energy from sources. This provided new insights into the design and utilization of graphene oxide films for highly efficient energy harvesting.
Keywords: A. nanostructures, A. thin films, B. microstructure, D. energy storage, D. ionic conductivity.
1. Introduction Graphene oxide (GO) attracted recently significant attention in the field of water purification [1,2],
2
ionic and molecular sieving [3,4], gas leakage testing [5] and desalination [6-14], because of its remarkable properties such as adequate mechanical strength [5,15] and excellent ion selectivity [2-4,16]. Typically, GO sheets with polar oxygenated functional groups on the basal plane and the edges, resulting in the formation of sp2-sp3 hybridized structure, could hold them to establish lamellar structure with stable nanochannels [17, 18]. These nanocapillaries allow unimpeded permeation of water but are completely impermeable to liquids, vapors, and gases [5, 19-21]. On the other hand, due to the deprotonation of oxygenated functional groups which are attached to GO surface, the films are negatively charged, resulting in the strong attracting force to cations by the contrast of the significant repulsion force to anions. Therefore, some researchers hold an optimistic attitude that GO membranes could act as cation-exchange membranes for separation and filtration [22-24], and even electricity generation [20, 25]. Recent studies demonstrated that graphene-based materials provided considerable opportunities for development of future-generation energy-transform devices [20, 25, 26]. Guo et al. reported a twodimensional (2D) nanofluidic generator system based on the electrokinetic ion transportation through a layered graphene hydrogel membrane. In aqueous conditions, 2D nanocapillaries were formed between adjacent graphene sheets and the negatively charged 2D nanocapillaries showed surface-chargegoverned ion transportation that preferentially permeated counter-ions and rejected co-ions [26]. This work provided a new idea to design GO-based membranes for energy-transform devices. Subsequently, Zhao et al. presented a power-generating phenomenon of single GO film with a preformed oxygencontaining group gradient (g-GOF) and the as-prepared g-GOF was able to provide moisture-enabled voltage output with efficient energy conversion efficiency [25]. Sun et al. reported that it was practical to realize synchronous energy harvesting and ion separation through nanochannels within GO films during the process of ion trans-membrane permeation [20]. The presence of GO membrane between sources and drains causes significant confinement to the ion free-migration from source to drain, which results in the establishment of electrochemical dialysis equilibrium and the formation of a space charge zone [27].
3
The different interactions of cations and anions with GO membranes give rise to the unequal diffusivities of them, respectively. These results that an excessive amount of cations and anions present in drains and sources as well as on both surfaces of GO membranes, respectively, lead to the generation of transmembranes electrical potential and Donnan potential [28]. Electrical potential differences across source and drain solutions can be sustained for a long time due to the strong confining effect of GO sheets [27]. Few previous studies have investigated the influence of some factors on electricity generation with GO membranes such as the mobility difference of ions, the intensity of magnetic fields and temperature difference etc. Sun et al. indicated that GO membranes could find potential applications in the purification of wastewater, while producing electricity simultaneously [20]. The results demonstrated that applying magnetic fields and improving temperature could increase electrical potential differences. Moreover, the mobility difference of different ions would cause electrical potential differences, and hybridization of various source solutions could also improve electrical potential difference. In this work, for the first time, we found that the source concentration, the applied pressure and membrane thickness could successfully give birth to great change of electrical potential differences in energy harvesting using GO membranes. This provided a new avenue for designing GO films for efficient energy harvesting.
2. Experimental procedure 2.1. Preparation of well-stacked GO membranes The preparation of GO was carried out by a modified Hummer Method using natural flake graphite (325 mesh), which was exposed to a mixture of concentrated sulfuric acid, phosphoric acid and potassium permanganate for oxidation according to our previous method [29]. As-prepared GO flakes were dispersed in ultrapure water by vigorous sonication (125 W for 5 min) to form a stable suspension with a concentration of 2 mg·mL-1. At first, we produced a series of different thicknesses of GO membranes assembled on the nylon micro-filters (diameter: 50 mm, pore size: 0.2 μm) by vacuum 4
filtration [30-32]. In order to improve GO membranes stability in solutions, GO films were not exfoliated from the support micro-filter. The thickness of GO films was turned by just controlling the volume of the filtrated 2 mg·mL-1 GO aqueous suspension. The interlayer spacing of the obtained GO membrane was characterized by X-ray diffraction (XRD, X' Pert Pro, Panalytica) in the dry state. The morphologies of membranes were examined by the scanning electron microscopy (SEM, SIGMA, Carl Zeiss).
2.2. Experimental Setups The electricity generation experiments were carried out with two sets of self-made apparatus. We used the same apparatus to investigate the effect of the membrane thickness and source concentration, and it consisted of a feed reservoir and a permeate reservoir separated by a PMMA plate with a leak hole (1.2 mm in diameter) in the center，as shown in Figure 1 (a, b). The studied GO membranes were sandwiched between two pieces of identical copper foils with an opening of 5 mm in diameter to ensure the GO films directly contact the solutions in the sources and drains. Two pieces of copper foils were clamped between two seal rings, which formed a tight seal between the two compartments. In order to eliminate the influence of micro-filters support, we performed control experiments using bare nylon micro-filters with other conditions remaining the same. In a typical experiment, we filled one of the compartment (referred to as feed reservoir abovementioned) with 100 ml of certain concentration MgCl2 source solution. Equivalent volume deionized water was injected into the other (permeate) compartment with the same speed. Meanwhile, gentle magnetic stirring was applied in both feed and permeate solutions to avoid concentration polarization effect. During the penetration process, the conductivities of the drains were measured by a conductivity meter (LEICI, DDSJ-308A). The voltage ΔVDS (ΔVDS = VD - VS. In this equation, VD and VS are the electrical potential of drains and sources respectively, ΔVDS is the electric potential difference between the drains and sources) was measured by a pair of identical silver electrodes (99.99% in purity) which 5
were fastened to the PMMA plate and separated by the GO film in the middle with a total distance of 12 mm. The voltage was measured by the multimeter (UNI-T, UT805A). Before the permeation experiments, a thin tape was glued to one of copper foils to substitute the GO membrane with other steps remaining the same. No obvious conductivity variations in drains occurred with time using a high concentration solution in the source compartment, which illustrated that the experimental setup possessed an excellent seal. However, due to study the effect of pressure on electricity generation with graphene oxide membranes, the pressure-driven electricity generation experiments were performed using the second sets of apparatus. The set of apparatus was pressurized with N2 flow. The schematic diagram and photograph of pressuredriven device are showed in Figure 1 (c, d). In a typical experiment, the whole device was placed on a flat plane, and then 50 ml of a certain 0.1M HCl feed solution and deionized water were injected into the source and drain compartments from respective circular tube in the vertical direction with the same speed. The feed solution was allowed to withstand different N2 flow pressure in the horizontal direction. Meanwhile, these nanochannels within GO membranes would collapse and shrink at high pressures [33]. Therefore, a low pressure of less than 0.2 MPa was applied in this study in order to avoid destroying the structural stability of GO interlayer nanochannels.
3. Results and discussion A typical preparation of GO membranes was indicated by vacuum filtration with 0.75 ml suspension from 2 mg·mL-1 original prepared solution, as shown in Figure 2a. The well-stacked GO films are highly stable in water for a long time, as illustrated in Figure 2b. The result of XRD revealed that the interlayer spacing of the GO membrane was about 0.857 nm (Fig. 2c). The SEM image of the cross-sections indicated that the thinnest thickness of studied GO flakes from 0.25 mL original 2 mg·mL-1 GO suspensions was about 1 μm (Fig. 2d). Before formal experiments, the control experiment was performed to investigate the effect of nylon micro-filter, which indicated that it could also produce
6
voltage with bare micro-filter because of the effect of electrical double layer [34-36]. The voltage generated consistently dropped over time but the voltage generated of GO membranes can maintain a relatively stable value after 1.5 h as shown in Fig. 3a. The mechanism of the voltage generated from GO membrane and microfilter attributes to the effect of electric double layer in nanochannels and microchannels [26, 37, 38]. For GO membranes, due to the deprotonation of oxygenated functional groups which are attached to GO surface, the GO surface is negatively charged. The negatively charged 2D nanocapillaries show surface-charge-governed ion transportation that preferentially permeate counter-ions and exclude co-ions [20, 26]. Consequently, such mobility difference of cations and anions would lead to the presence of an excess amount of cations or anions in both drains and sources, which in turn would attain the establishment of an electrochemical dialysis equilibrium [20, 27]. Ultimately, a Donnan potential is generated on the GO surfaces, thus the voltage generated can maintain a stable value after 1.5 h. But for common nylon micro-filters, the pore size is 0.22 μm (much larger than ion diameters). Most cations and anions hardly interact with the microfilters surface because of the broad microchannel and weaker surface charge relative to GO films which possess narrow nanochannel and higher surface charge. Therefore the voltage generated of nylon microfilter consistently dropped over time but the voltage generated of GO membranes can maintain a relatively stable value after 1.5 h. In short, compared with GO films, the common nylon-filter are unsuitable for the electricity generation during the process of transmembrane transport in practice. Firstly, the effect of GO membrane thickness on the generated voltage between the sources and drains was investigated (Fig. 3b), and it exhibited that the permeation of source ions effectively decreased with the increasing of the GO membrane thickness by observing the conductivity of drains. Apart from that, we found the voltage across the source and drain solution could adequately increase with increasing the GO membranes thickness. The results revealed that the mobility difference of cations and anions led to variations of the voltage generated, and the nanochannels length within GO membranes dominated the mass transport process and the ion selectivity [39]. The nanochannels length would increase with the
7
increasing of the GO membrane thickness, which enhanced the selectivity of cation and anion ions. Therefore, the thicker GO membranes with longer nanochannels could cause a greater mobility difference, resulting in a larger ΔVDS. However, the growth of total permeation ions by reducing the GO membranes thickness could not improve the voltage generated.
Next, the influence of the source concentration was studied on ΔVDS, as illustrated in Figure 3c. The result appeared that increasing the source concentration could effectively improve the voltage generated, but the increase of voltage became extremely not obvious when the concentration increased to a certain degree. That’s because during the ion trans-membrane transport process, the negatively charged GO membranes repulsed anions and attracted cations, thus cations would be the first ions transported into membranes through GO nanocapillaries [20]. Moreover, the total permeation of cations would also increase with the growth of source concentration due to a higher diffusion driving force. Consequently, compared with low source concentration, more cations would be enriched on GO surfaces that directly contact with drain solutions. In addition, the anions trans-membrane transport was driven by the electrostatic attraction force from the corresponding cations embedded in GO sheets and penetrate through the negatively charged GO membranes passively, resulting in excessive anions enriched on GO surfaces that directly contact with source solutions. Due to the presence of Donnan effect, an electrochemical dialysis equilibrium would be established and the generated voltage could tend to be a stable value after about one hour [20, 27]. Thus more cations and anions would enrich in drains and sources respectively with increasing the source concentration, which ultimately producing a lager ΔVDS. Furthermore, the effect of the applied pressure was examined using another set of experimental device (Fig. 1 (c, d)). According to recent studies on the pressure-driven separation performance of GO films,
8
Huang et al. has reported that nanochannels within GO membranes were relatively stable at the low pressure range due to their rigidity and would begin to collapse and shrink at high pressures [33]. Therefore, we applied a low pressure of less than 0.2 MPa to keep the structural stability of interlayer nanochannels. Sun et al. has reported that the applied pressure weakened water-ion interactions confined in GO nanochannels and further decreased the selectivity of cations and anions, resulting in poor source solution rejection [39]. In our pressure-driven experiments, the result illustrated that the voltage would decrease when the pressure was applied on feed solution. At the same time, the generated voltage had a drop tendency with the increase of the applied pressure, as illustrated in Fig. 3d. From the experimental phenomenon, it proved to be convincing that the salt rejection of GO films was poor in pressure-driven separation performance and the selectivity of ions would decrease with increasing the applied pressure. The results confirmed that increasing the thickness of GO films could sharply weaken the permeation of source ions. However, electrical potential differences would effectively increase with the increase of GO membrane thickness. In the meanwhile，increasing the source solution concentration could increase concentration gradient between the source and drain solutions, which would enhance electrical potential differences. Remarkably, electrical potential differences unexpectedly reduced while pressure was applied on GO membranes.
4. Conclusions In summary, the effect of the membrane thickness, solution concentration and applied pressure on electricity generation with GO membranes has been studied, revealing that tuning each of three factors could affect the generated voltage. Applying thicker GO films and highly-concentrated source solutions could efficiently increase voltage generated, which has a significant engineering application. In the meanwhile, our experimental results prove that it is beneficial to avoid applying pressure on electricity generation with GO membranes because the applied pressure decreases the rejection of solution and weakens the selectivity of cations and anions. The results presented in this study indicate that there is a further breakthrough for GO membranes to realize efficient energy conversion. In addition, we can roughly identify its concentration based on the difference of generated voltage relative to the same source solution.
Acknowledgements 9
The authors acknowledge the support of the National Natural Science Foundation of China (No. 51706157), the China Postdoctoral Science Foundation (No. 2017M612498), and the Fundamental Research Funds for the Central Universities (No. 2042016kf0023).
References [1] T. Ramanathan, A. A. Abdala, S. Stankovich, D. A. Dikin, M. Herrera-Alonso, R. D. Piner, D. H. Adamson, H. C. Schniepp, X. Chen, R. S. Ruoff, Nat. Nanotechnol. 3 (2008) 327-331. [2] F. Peng, T. Luo, L. Qiu , Y. Yuan, Mater. Res. Bull. 48 (2013) 2180-2185. [3] H. Li, Z. Song, X. Zhang, Y. Huang, S. Li, Y. Mao, H. J. Ploehn, Y. Bao, M. Yu, Science 342 (2013) 95-98. [4] B. Mi, Science 343 (2014) 740-742. [5] R. R. Nair, H. A. Wu, P. N. Jayaram, I. V. Grigorieva, A. K. Geim, Science 335 (2012) 442-444. [6] D. Cohen-Tanugi, J. C. Grossman, Nano Lett. 12 (2012) 3602-3608. [7] S. P. Surwade, S. N. Smirnov, I. V. Vlassiouk, R. R. Unocic, G. M. Veith, S. Dai, S. M. Mahurin, Nat. Nanotechnol. 10 (2015) 459-464. [8] K. P. Lee, T. C. Arnot, D. Mattia, J. Membr. Sci. 370 (2011) 1-22. [9] C. Charcosset, Desalination 245 (2009) 214-231. [10] A. Nicolai, B. G. Sumpter, V. Meunier, Phys. Chem. Chem. Phys. 16 (2014) 8646-8654. [11] H. Deng, P. Sun, Y. Zhang, H. Zhu, Nanotechnology 27 (2016) 274002. [12] M. Bhadra, S. Roy, S. Mitra, Desalination 378 (2016) 37-43.
10
[13] D. An, L. Yang, T. J. Wang, B. Liu, Ind. Eng. Chem. Res. 55 (2016) 4803-4810. [14] D. Cohen-Tanugi, L. C. Lin, J. C. Grossman, Nano Lett. 16 (2016) 1027-1033. [15] Y. Jiang, W. N. Wang, D. Liu, Y. Nie, W. Li, J. Wu, F. Zhang, P. Biswas, J. D. Fortner, Environ. Sci. Technol. 49 (2015) 6846-6854. [16] P. Sun, H. Liu, K. Wang, M. Zhong, D. Wu, H. Zhu, Chem. Commun. 51 (2015) 3251-3254. [17] K. P. Loh, Q. Bao, G. Eda, M. Chhowalla, Nat. Chem. 2 (2010) 1015-1024. [18] G. Eda, M. Chhowalla, Adv. Mater. 22 (2010) 2392-2415. [19] B. Chen, H. Jiang, X. Liu, X. Hu, J. Phys. Chem. C, 121 (2017) 1321-1328. [20] P. Sun, F. Zheng, M. Zhu, K. Wang, M. Zhong, D. Wu, H. Zhu, Sci. Rep. 4 (2014) 5528-5532. [21] S. C. O'Hern, M. S. Boutilier, J. C. Idrobo, Y. Song, J. Kong, T. Laoui, M. Atieh, R. Karnik, Nano Lett. 14 (2014) 1234-1241. [22] T. Sata, T. Sata, W. Yang, J. Membr. Sci. 206 (2002) 31-60. [23] P. M. V. D. Velden, C. A. Smolders, J. Appl. Polym. Sci. 21 (1977) 1445-1457. [24] H. Strathman, A. Grabowski, G. Eigenberger, Ind. Eng. Chem. Res. 52 (2013) 10364-10379. [25] F. Zhao, H. Cheng, Z. Zhang, L. Jiang, L. Qu, Adv. Mater. 27 (2015) 4351-4357. [26] W. Guo, C. Cheng, Y. Wu, Y. Jiang, J. Gao, D. Li, L. Jiang, Adv. Mater. 25 (2013) 6064-6068. [27] P. Sun, H. Deng, F. Zheng, K. Wang, M. Zhong, Y. Zhang, F. Kang, H. Zhu, 2D Mater. 1 (2014) 034004. [28] H. Ohshima, S. Ohki, Biophys J. 47 (1985) 673-678. [29] J. Pei, X. Zhang, L. Huang, H. Jiang, X. Hu, RSC Adv. 6 (2016) 101948-101952. [30] B. S. Kong, J. Geng, H. T. Jung, Chem. Commun. 16 (2009) 2174-2176. [31] S. Zhang, Y. Li, N. Pan, J. Power Sources 206 (2012) 476-482. [32] Q. Chen, P. Liu, C. Sheng, L. Zhou, Y. Duan, J. Zhang, RSC Adv. 4 (2014) 39301-39304. [33] H. Huang, Y. Mao, Y. Ying, Y. Liu, L. Sun, X. Peng, Chem. Commun. 49 (2013) 5963-5965. [34] S. T. Senthilkumar, B. Senthilkumar, S. Balaji, C. Sanjeeviraja, R. K. Selvan, Mater. Res. Bull. 46
11
(2011) 413-419. [35] K. Liu, P. Yang, S. Li, J. Li, T. Ding, G. Xue, Q. Chen, G. Feng, J. Zhou, Angew. Chem. Int. Ed. 55 (2016) 8003-8007. [36] G. Xue, Y. Xu, T. Ding, J. Li, J. Yin, W. Fei, Y. Cao, J. Yu, L. Yuan, L. Gong, J. Chen, S. Deng, J. Zhou, W. Guo, Nat. Nanotechnol. 12 (2017) 317-322. [37] D. Grahame, Chem. Rev. 41 (1947) 441-501. [38] T. Ding, K. Liu, J. Li, G. Xue, Q. Chen, L. Huang, B. Hu, J. Zhou, Adv. Funct. Mater. 27 (2017), 1700551. [39] P. Sun, R. Ma, H. Deng, Z. Song, Z. Zhen, K. Wang, T. Sasaki, Z. Xu, H. Zhu, Chem. Sci. 7 (2016) 6988-6994.
12
Fig. 1 (a, b) The photograph of electricity generation experimental apparatus for studying the effect of the membrane thickness and solution concentration. (c) The photograph of the pressure-driven electricity generation experimental setup. (d) Schematic diagram of pressure-driven filtration experimental setup for electricity generation.
13
Fig. 2 (a) The photograph of GO membranes and the inset shows corresponding suspension. (b) The photograph shows the stability of GO papers by vacuum-filtrated through nylon microfilters in water after 1 day. (c) XRD patterns of GO films in a dry state. (d) The corresponding cross-section SEM characterization of as-prepared GO membranes.
14
Fig. 3 (a) Voltage generations across sources and drain (ΔVDS) of bare microfilters and GO on microfilters with 1M MgCl2 source solution in control experiments. (b) Voltage generations of different thickness GO films across the source and drain solution, and corresponding conductivities of the drains for 3 h. (c) Stable voltage generations across GO membranes based on the penetrations of different concentration MgCl2 feed solutions for 3 h. (d) Voltage generations across GO membranes based on 0.1M HCl feed solution with different pressure for 3 h. The error bars come from the corresponding error estimations and are formed by three times experimental results. The error bar of conductivities variations is neglected because of a very slight variation in experiments.
15