Ultrahigh water permeation with high multivalent metal ions rejection rate through graphene oxide membranes

Graphene oxide (GO) membranes show exceptional molecular permeation properties and have gained tremendous attention in the area of wastewater treatment. However, they still suffer from some limitations, such as low water permeance when the ion rejection rate is at a satisfactory level and unstable performance. Here, we develop a sort of GO membrane that exhibits ultrahigh water permeance up to 75.2 L m −2 h −1 bar −1 while still maintaining a high rejection rate of 99.9% for multivalent metal ions. Importantly, of all state-of-the-art nanoltration membranes, this is the most permeable membrane with a satisfactory level of rejection rate for multivalent ions. Furthermore, the GO membrane has outstanding stability over long-time operation. Our work provides a simple way to fabricate GO membranes with outstanding water purication performance.


Introduction
Water puri cation is an attractive pathway to fresh water 1,2 , and membrane technologies, one of the most important energy-e cient and environmentally friendly strategies for water treatment, have attracted intensive interest in recent years 3,4 . An ideal water treatment membrane should exhibit key performance benchmarks, including high water permeance with an e cient rejection rate and good stability [5][6][7] .
Here, we develop a sort of GO membrane that achieves an unprecedented water permeance while maintaining an ultrahigh ion rejection rate for multivalent metal ions. The water permeance reaches up to 75.2 L m −2 h −1 bar −1 , which is about one or two orders of magnitude higher than the permeances through most GO membranes reported previously 18,37,38,41 . Remarkably, the GO membrane has superior ltration performance compared to other nano ltration membranes, as the highest water permeance is about 62 L m −2 h −1 bar −1 with a rejection rate of 98 % for multivalent metal ions very recently 42 . Also, the GO Page 3/13 membranes have outstanding stability. Thus, this work provides a simple avenue to fabricate GO membranes with exceptional ltration performance.

Results
As illustrated in Fig. 1 are still superior to most other membranes, as shown in Fig. 1(b), (c), and the Table S1 in Supporting Information section PS 3. Thus, the GO membranes in the present work exhibit ultrahigh water permeances with high rejection rates for multivalent ions.
The GO membranes also showed outstanding stability with superior performance. Fig. 1(e) shows that the water permeance decreases from 95.5 L m −2 h −1 bar −1 and reaches a stable value of 75.2 L m −2 h −1 bar −1 within 100 min, while the corresponding rejection rate is maintained at more than 97% for 5mg/L FeCl 3 . We noted that the feed salt concentration continuously increased in the dead-end ltration set-up.
In addition, the increased salt concentration (concentration polarization) could affect the membrane performance. Therefore, a semi-continuous process was employed for the long-term operation experiment of the high-concentration solution ltered by the GO membranes 43 : the surface and ltration cleaning of the GO membranes by DI water were performed before the next ltration (see details in Supporting Information section PS 7). The water permeances slightly decreased slightly from 49.0 L m −2 h −1 bar −1 to 46.5 L m −2 h −1 bar −1 , while the corresponding rejection rates increased from 85.0% to 96.4%, showing an outstanding stability with superior ltration performance of the GO membranes (see Fig. S5 in Supporting Information section PS7). We believe that the increase in the rejection rate and the slight decrease in the water permeance were mainly due to the improved self-assembly and more compacted structures between the GO sheets under the long-term vacuum ltration conditions.
In addition, we analysed the the ions adsorption by the GO membranes in our ltration experiments 44 . As shown in Fig. 2, the adsorption e ciency of the GO membranes for ZnSO 4 , CuSO 4 , FeCl 3 , and Pb(NO 3 ) 2 were 13%, 19%, 26%, and 42%, respectively. As mentioned above, the rejection rates of the GO membranes were 83% ~ 99.9% (see Fig. 1(b)), which are much higher than 13% ~ 42% removed by adsorption. These results indicate that the signi cant effect on the removal multivalent ions at a low concentration of 5 mg/L is mainly due to rejection by the GO membranes, even though the adsorption effect cannot be negligible. Further, we performed adsorption experiments using FeCl 3 solution with a high concentration of 500 mg/L. As shown in Fig. S6 in Supporting Information section PS8, the FeCl 3 solution with 500 mg/L showed stable concentration for over 120 min. The removal rate by adsorption was < 4%. Considering the rejection rate for 500 mg/L FeCl 3 solution was 89.8 %. Thus, for the high salts concentrations, the salts removal are almost entirely attributed to the rejection rather than adsorption by the GO membranes, which was further con rmed by the mass balance experiment (see details in Supporting Information section PS 9).
Our previous experimental studies showed that the order in which the GO membrane is exposed to ions is important for e cient ion sieving 41 Fig. 3(c)). The rejection rates for the FeCl 3 -controlled GO membranes are much lower than the rejection rates for the ZnSO 4 -controlled GO membranes.
To illustrate the underlying physical mechanism, the interlayer spacings of the GO membranes fabricated in the present work were analyzed by X-ray diffraction. As shown in Fig. 1(d) The large interlayer spacings of the GO membranes prepared by the present method can be further demonstrated by the interlayer spacing of the GO membrane immersed in pure water, which was 16.2 Å ( Fig. 1(d)), this is also ~3 Å larger than the interlayer spacing of 13.0 Å for the GO membrane fabricated by the drop-casting method immersed in pure water, as reported earlier 41

Discussion
We think that the larger interlayer spacings of the GO membranes in the present work are attributed to the fabrication process without drying treatment. We note that a drying treatment process was previously used during the fabrication of GO membranes prepared by the drop-casting method 41,45,46 . During the drying treatment, the water molecules between the GO sheets are lost, which leads to a reduction in spacing between the GO sheets. In some parts of the spacings between two GO sheets, there is no water, which causes direct contact between the two GO sheets, and thus induces the π-π stacking 47 formed between the aromatic rings in the upper and lower GO sheets. Due to those π-π stackings, it is very di cult for water molecules to enter the spacings between the GO sheets when the GO membranes prepared by the conventional drop-casting method are again immersed in an aqueous solution. Thus, due to the drying treatment, some spacings in the GO membranes fabricated by the drop-casting method are very small. As a result, the distribution of the spacings between the GO sheets in membranes fabricated by the conventional drop-casting method shifts toward small values due to the drying treatment. In contrast, the GO membranes in the present work have larger spacings due to the lack of the drying treatment. We note that the interlayer spacing corresponds to the largest distribution of the spacings. Therefore, the GO membranes in the present work have larger interlayer spacings.
Considering that the sizes of the hydrated multivalent metal ions are much larger than the sizes of the hydrated anions, the hydrated anions should not pass through the membrane due to the electrostatic interactions between the cations and the anions when the cations are rejected. To demonstrate that the effect of anions in these multivalent ions rejection is negligible, we further performed ions permeation tests with different anions. Taking Fe 3+ as an example, the three Fe 3+ salts solutions with different anions and 98.3%, respectively. Therefore, the differences in the water permeances and rejection rates were negligible. Thus, the anions have less contribution to the high water permeances for multivalent metal ions rejection using the GO membrane.
One may argue that the GO membranes in the present work with larger interlayer spacings still have high rejection rates for the multivalent ions. Clearly, the hydrated ions x the spacings between the neighboring GO sheets due to the strong hydrated cation-π interactions 41,48 . The key for ion rejection is the gate size determined by these hydrated ions 10 . The interlayer spacing, corresponding to the largest distribution of the spacings in the GO membranes, is not the gate for the ion rejection but usually correlates with the gate size 10 . We note that, when the hydrated multivalent ions enter the space between GO sheets, the hydration structures of the multivalent ions will be distorted due to the strong hydrated cation-π interactions from the DFT calculations (see details in Methods), we observed that, when the hydrated multivalent ions enter the space between GO sheets, the hydration structures of the multivalent ions are distorted due to the strong hydrated cation-π interactions. The stable optimized geometries of hydrated Fe 3+ -(H 2 O) 6 and hydrated Fe 3+ -(H 2 O) 6 @Graphene between graphene sheets from DFT computations were shown in Fig. 4. The radius of hydration structures of Fe 3+ ions distorted from 3.92 Å to 2.55 Å inbetween two graphene sheets, indicates that size of the hydrated Fe 3+ ion inbetween two graphene sheets is smaller than that in solution. This result is similar to the behavior of hydrated K + in GO membranes reported in our previous study 41 . Furthermore, we noted that this distorted behavior of the hydrated cations is attributed to the strong hydrated cation-π interactions 48,49 . Thus, this makes the sizes of multivalent ions with undistorted hydrated water in bulk solutions larger than the gate size in the GO sheets, resulting in the higher rejection rates of undistorted hydrated ions Why ZnSO 4 -controlled GO membranes can realize effective sieving for the other three multivalent ions, while FeCl 3 -controlled GO membranes cannot? It is very di cult to directly measure the structures of the hydration shells of these ions. Fortunately, we can reasonably assume that the order of the sizes of the hydration shells correlate well with the order of the interlayer spacings. As shown in Fig. 1(d), the order of the interlayer spacing from widest to narrowest is FeCl 3 > Pb(NO 3 ) 2 > CuSO 4 ≈ ZnSO 4 . In mixed ion solutions, the interlayer spacings of GO membranes controlled by ZnSO 4 or FeCl 3 keep the order and only differ slightly from the interlayer spacings of GO membranes with pure ZnSO 4 or FeCl 3 solutions, as shown in Fig. 3 (b), (d). These results clearly demonstrate that the rst added multivalent ions can effectively control the spacing in mixed solutions. Since the ZnSO 4-controlled GO membranes have the smallest interlayer spacing, the ZnSO 4 -controlled GO membrane can reject most of the other multivalent ions used in the present work. In contrast, the FeCl 3 -controlled GO membranes have the largest interlayer spacing, hence, other ions are able to easily pass through the GO membranes controlled by FeCl 3 .
Overall, we developed a sort of GO membrane with an ultrahigh water permeance while still maintaining a high rejection rate of multivalent metal ions. In particular, the permeance reaches an unprecedented 75.2 L m −2 h −1 bar −1 , while maintaining the high rejection rate of 99.9%. We note that this is the most permeable membrane with a satisfactory rejection rate for multivalent ions of all state-of-the-art nano ltration membranes. The ultrahigh water permeances are attributed to the larger interlayer spacings of the GO membranes resulting from the lack of the drying treatment during the fabrication process. The drying treatment is usually used in conventional methods such as the drop-casting method. Moreover, the GO membranes in the present work showed outstanding stability with superior performance. Furthermore, we found that ZnSO 4 -controlled GO membranes and GO membranes fabricated by conventional drop-casting method were prepared by drop-casting the GO suspension (4 mg/mL, 1 mL) droplets onto a smooth paper substrate. The GO membranes fabricated by conventional drop-casting method were drying thoroughly at 60 ºC for 12 hours. After that, they were peeled off, rinsed and soaked with DI water for more than half an hour to remove the absorbed metal ions, then dried in a dry dish at room temperature for three days. These drop-casting GO membranes were used for ion controlling experiments.
Filtration experiment. 5~500 mg/L different multivalent metal cations (FeCl 3 , CuSO 4 , Pb(NO 3 ) 2 or ZnSO 4 ) solutions were added into the feed side after the GO membrane were prepared using vacuum ltration, respectively. Under a pressure of 1 bar, the salt solutions were ltered through GO membranes. Filtrates were collected when the ltration process became steady. Water permeances were calculated by measuring the time required to collect 30 mL ltrate. Rejection rates for the multivalent ions were calculated from the concentration of feed and permeate solution. The water permeance (J w) and Rejection rate (R) was measured by using the following equation (1) and (2): where J W is the water permeance (L m −2 h −1 bar −1 ), V is the volume of the ltrate (L), A is the effective membrane area (m 2 ). ∆t is the permeation time (h) and the P is the ltration pressure (bar). C p and C f are the concentration of permeation and feed ions solution which were measured by inductive coupled plasma-optical emission spectrometry (ICP-OES), respectively.
Adsorption experiment. GO membranes were prepared from 40 mL, 7.5 mg/L GO suspensions on substrates using vacuum ltration. Then, 100 mL, 5 mg/L FeCl 3 , Pb(NO 3 ) 2 , ZnSO 4 , or CuSO 4 solutions were added to the feed side, respectively. These salt solutions under ambient conditions (without vacuum ltration) were stirred with a blender at ~180 RPM. Next, samples were collected at different predetermined time intervals in 120 min to evaluate the residual salt concentrations of the solutions. We noted that the salt solutions in the feed side permeated very slowly though the GO membranes due to their own gravity, the ltrates were only ~ 5 mL within 120 min. Therefore, the effect of the small volume of ltrates on the salt concentration during the adsorption experiments could be negligible.
The adsorption e ciency (AE) was de ned by using the following equation (3): where, C o and C a is the initial feed concentration and the concentration of salts after adsorption equilibrium, respectively.
Calculation Methods. The B3LYP 50,51 method is used to study the intermolecular interactions. For geometry optimizations, the double-ζ basis was employed, and a d-polarization function was added (marked with 6-31 G(d)). The pseudopotential function with Lanl2dz is introduced into the basis set for Fe 3+ ion. Two graphene surfaces of 15.69 × 12.26 Å 2 were used for the DFT study, which were large enough to obtain results with a tolerable error 48 . All calculations were carried out using the Gaussian-09 program 52 .

Data availability
The authors declare that all the data supporting the ndings of this study are available within the article (and its Supplementary Information le), or available from the corresponding author on reasonable request.