Design of hyperporous graphene networks and their application in solid-amine based carbon capture systems

1. Graphite powder, 10 g was stirred with cold concentrated H2SO4 (230 ml at 0 °C). 2. Then KMnO4 (30 g) was added to the suspension slowly to prevent a rapid rise in the temperature (less than 20 °C). The solution underwent a colour change at this point from black to a very dark green. 3. After removal of the ice-bath the mixture was stirred at room temperature for another 2 h. 4. DI water (230 ml) was slowly added to the reaction vessel to keep the temperature under 98 °C. 5. The diluted suspension was stirred for an additional 15 min and further diluted with DI water (1.4 l) before adding H2O2 (100 ml). Upon addition vigorous foaming occurred and the solution turned to brown. 6. The mixture stirred for 2 h at room temperature and left overnight. 7. Product settled at bottom was separated from the excess liquid by decantation followed by centrifugation. 8. The product was washed by centrifugation until the pH reached neutral. 9. Freeze dried to obtain a final product, called GO-A.

We demonstrate a simple and fully scalable method for obtaining hierarchical hyperporous graphene networks of ultrahigh total pore volume by thermal-shock exfoliation of graphene-oxide (exfGO) at a relatively mild temperature of 300 C. Such pore volume per unit mass has not previously been achieved in any type of porous solid.We find that the amount of oxidation of starting graphene-oxide is the key factor that determines the pore volume and surface area of the final material after thermal shock.Specifically, we emphasize that the development of the hyperporosity is directly proportional to the enhanced oxidation of sp 2 C]C to form C]O/COO. Using our method, we reproducibly synthesized remarkable meso-/macroporous graphene networks with exceptionally high total pore volumes, exceeding 6 cm 3 g À1 .This is a step change compared to #3 cm 3 g À1 in conventional GO under similar synthetic conditions.Moreover, a record high amine impregnation of >6 g g À1 is readily attained in exfGO samples (solid-amine@exfGO), where amine loading is directly controlled by the pore-structure and volume of the host materials.Such solid-amine@exfGO samples exhibit an ultrahigh selective flue-gas CO 2 capture of 30-40 wt% at 75 C with a working capacity of z25 wt% and a very long cycling stability under simulated flue-gas stream conditions.To the best of our knowledge, this is the first report where a graphene-oxide based hyperporous carbon network is used to host amines for carbon capture application with exceptionally high storage capacity and stability.
3][4][5][6][7] Considerable performance improvements have been achieved by synthesizing and manipulating their functional features and porous structures. 6In the case of CO 2 capture, obtaining high enough CO 2 uptake capacities by using sorbent based materials under ue-gas conditions is still a challenging task.Many sorbent materials show good CO 2 uptake capacities but only at 0 C or 25 C and for 100% dry CO 2 .Porous sorbents including zeolites, activated carbons and MOFs do not show desirable selective CO 2 uptake under humid conditions and/or at >50 C. 8 This is a major setback for the implementation of CCS.The recent efforts on the introduction of CO 2 -philic amine groups into nanoporous structures have led to a large enhancement in selective carbon (CO 2 ) capture and effluent ue-gas tolerance. 4,7,8However, the practical feasibility of these materials for CO 2 capture is compromised by the lack of a simple and large-scale synthesis method.Substantial improvements in both the structural stability and scalable synthesis of these materials are therefore urgently needed.In this context, carbon based systems are particularly attractive, due to their relatively high thermal stability, chemical resistance and potential for large-scale production at low-cost. 2,3,91][12][13][14] Current synthesis techniques aimed at such structures involve salt assisted chemical activation and templates of pre-designed silica/salt or other molecular architectures.
6][17] In most cases, such structures are synthesized by a top-down approach of graphite oxidation followed by chemical or thermal reduction.Thus, the achievable conductivity and porosity in the nal graphene based networks is directly related to the strengths of the oxidation of the starting graphite (to yield graphene-oxide, GO), and the reduction (of GO) processes.In this work, we report a simple and wellcontrolled method for obtaining highly hierarchical mesoand macro-porous graphene networks, prepared by thermalshock exfoliation of GO at moderate temperatures for a short time of about 5 minutes.Specically, we demonstrate that the degree of oxidation of the GO controls the strength of exfoliation to yield extraordinary porosity: both the specic surface area (SSA, z800 m 2 g À1 ) and ultrahigh total pore volume ($6 cm 3 g À1 ).Such hyperporous exfGO samples in turn exhibit remarkable cyclic working ue-gas CO 2 capacities in solidamine-impregnated systems (solid-amine@exfGO).
Firstly, precursor GO samples with increasing degree of oxidation (named GO-A, -B, -C & -D, in the ascending order of oxidation state) were synthesized by oxidation of graphite powder using a modied Hummers' method (see the Experimental and ESI, Fig. 1 and S1 †).Note that there are environmentally green approaches for the synthesis of graphene-oxide at industrial quantities. 18The increased oxidation is conrmed by an enhanced layer spacing (by powder X-ray diffraction, PXRD), increased concentration of epoxy (-C-O-C-), hydroxyl (-C-OH) and carboxylic (-C-O-OH) groups on the graphene basal plane and edge (by X-ray photoemission spectroscopy, XPS and Raman spectroscopy), and excess mass-loss and exothermicity during the decomposition in thermal reduction (by combined thermogravimetry-differential scanning calorimetry-mass spectrometry, TG-DSC-MS) (Fig. 1a, b, S2-S7 and Table S1 †).Specically the C]O/COO content is increased from z7.5% to $15.0% at the expense of sp 2 C]C carbon, which is reduced from z54.0% to #44.0% in the samples GO-A to GO-D.In addition to this, the TG reveals about 9 wt% of excess decomposition in GO-D with respect to the GO-A sample.PXRD also shows a larger interlayer distance of >0.92 nm in GO-D compared to z0.79 nm in GO-A.All these characteristics are consistent with a greater degree of oxidation in the GO-D sample.
Next, the exfoliation of these GO samples was carried out by direct thermal-shock at a relatively mild temperature of 300 C, and for less than 5 minutes to achieve complete exfoliation (Fig. 1c-g and S6-S10 †).The exfoliation is associated with an exothermic reaction, where an increase in the heating rate leads to a highly enhanced exothermicity at the decomposition point of GO (Fig. S6 and S7 †).A relatively high exothermicity is observed for the highly oxidized GO-D sample compared to the less oxidized GO-A sample.Since the GO structure incorporates surface oxy-functional groups and hydrogen bonded interlamellar water molecules, the sudden volatility of these labile species leads to a pressure build-up between the graphene sheets.Therefore, by this process, a highly oxidized GO sample always leads to a high degree of exfoliation (Fig. S9 and S10 †).This is exactly what is observed through the continuous increase in the volume of the exfGO samples, packed by tapping in a 4 ml vial at the constant mass of 32.5 mg (Fig. 1d and S9 †).The exfoliation is identied with the loss of the layered structure and the reduction of oxygen content, to 12.5 at% from 32.0 at% in precursor GO (Fig. 1, S1-S12, Tables S1 and S2 †).
N 2 adsorption-desorption isotherms (type II or S-type) indicate a considerable enhancement in the porosity of the samples from exfGO-A to exfGO-D (Fig. 2, S13-S15 and Table S3 †).For instance, the additional oxidization of the starting GO increases the SSA from 360 m 2 g À1 to 830 m 2 g À1 , the total N 2 uptake (at P/P 0 of #0.994) from 1748 cm 3 g À1 to 3953 cm 3 g À1 , and the corresponding total pore volume from 2.71 cm 3 g À1 to 6.10 cm 3 g À1 .A total pore volume of over 6.00 cm 3 g À1 in the exfGO-D samples is consistently achieved when the synthesis and exfoliation were repeated for four different batches of the samples.Such a total pore volume of $6 cm 3 g À1 is extremely high considering that our synthesis route is very simple and scalable.[5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] Furthermore, the highly hierarchical pore sizes and their distributions across the micro-, meso-and macro-porous regimes of the samples can be understood from the qualitative behaviour of N 2 adsorption-desorption isotherms (Fig. 2 and S13 †).The type II or S-type isotherms with a sharp hysteresis at high relative pressures indicate highly hierarchical pore sizes and unsaturated condensation of N 2 in macroporous networks of a plate-like geometry in the exfGO samples.We note that the majority of the meso-porosity is distributed in the range 3 to 50 nm that accounts for more than 75% of the total pore volume and yields ultrahigh meso-pore volume of up to 4.65 cm 3 g À1 .
In a controlled experiment, slow heating of GO samples to 300 C without exfoliation (GO-300C) leads to negligible porosity, similar to the as-synthesized precursor GO (asGO) samples (Fig. S9, S10, S14 and Table S3 †).Note that the exfoliation of GO-D at a very high temperature, 600 C to 1000 C, induces no further enhanced porosity compared to the samples exfoliated at 300 C (Fig. S17 †).
Moreover, our critical analysis shows, for the rst time, that the development of porosity is directly proportional to the strength of the oxidation of the precursor (Fig. 2e and f).Both the SSA and the total pore volume are linearly increased with respect to the oxidation of sp 2 (C]C) carbon into C-O/C]O/COO.Here a high degree of reduction in exfGO-D samples with an oxygen content of z12.5 at% compared to the z14.0 at% in exfGO-A is noticed, which further supports the high exothermic heat generation during the thermal-shock of GO-D (Fig. S7, S11 and Table S2 †).Thus, a high degree of oxidation of the GO with an increased concentration of carboxylic groups, and their induced large exothermic heat generation during the thermal decomposition are the desirable parameters to achieve high enough porosities under simple thermal-shock conditions.
Such porosity characteristics of the samples are further well supported by SEM and TEM micrographs (Fig. 3

and S18-S22 †).
A high degree of exfoliation and large pore networked graphenes can be observed for the exfGO-D compared to exfGO-A.The exfGO-D samples show highly interconnected graphene networks resembling a kirigami structure.A closer inspection of the SEM micrographs reveals a characteristic pore-size development, directly relevant to the exfoliation with respect to the increased degree of oxidation of the precursor GO sample.The highly oxidized GO samples exhibit a larger pore structure aer exfoliation.The structures consist of interconnected pores, with the macro-pores leading to meso-pores, and the meso-pores leading to micro-pores.
21][22][23][24][25][26][27][28] Here, triethylenetetramineimpregnated exfGO (TETA@exfGO) samples were obtained by methanol solution inltration (Fig. 4 and S23 †).Note that TETA is a low-cost short-amine and exhibits low thermal stability compared to long-chain poly-amines. 8,22In a controlled synthesis, the asGO sample showed a wet surface amine coating when the TETA loading was increased over 1 g g À1 of sample, due to a limited layer/pore space (Fig. S24 and S25 †).Due to their ultrahigh pore volume, exfGO-D samples can accept a record high TETA loading of z10 g g À1 of sample, without any sign of surface  wetting (Table S6 †). 4,7,8,13-15,21-28Indeed, SEM micrographs still show void spaces for a TETA loading as high as 7.0 g g À1 in the exfGO-D sample (Fig. S26 †).A systematic study of amineimpregnation conrms that the pore volume of the sample is the controlling factor for determining the amine loading (see SEM, TG, XPS and 77 K N 2 uptake porosity, Fig. 4a, b, S26-S30, Tables S7 and S8 †).The amine impregnation results in denite shis in the XPS binding energy of the C 1s and N 1s peaks due to the pore connement and/or interaction of surface/ functionality. 8Relative changes in the O 1s peaks show amine interaction with the surface oxygen functional groups.Amine impregnation is also conrmed from the N 2 adsorption isotherms of TETA-loaded samples, which show a negligible uptake.
For the rst time, we also show that the TETA@exfGO samples exhibit high room temperature ue-gas capture capacities of z30 wt% at 30 C aer humidication (Fig. S34 †).It is interesting to note that the same samples show reduced CO 2 uptake capacities for pure CO 2 without humidication (Fig. S36 and S37 †). 8 Dry and humidied state CO 2 interactions in the TETA@exfGO samples aer uptake runs are shown by XPS analysis (Fig. S38 †).It is worth pointing out here that the exfGO samples alone without amine impregnation show very low CO 2 uptake capacities of #1 mmol g À1 (Fig. S39 †).Direct amine loading on asGO without exfoliation (TETA@asGO) and bulk TETA samples shows a maximum CO 2 capture capacity of only 12 wt% and 10 wt% respectively with very slow uptake kinetics at 75 C under similar ue-gas conditions (Fig. 4d and S40 †).
Although the TETA@exfGO samples show exceptional CO 2 working capture capacities for both ue-gas CO 2 and singlecomponent CO 2 , the amine-group stability/volatility of the TETA is a concern during repetitive temperature swing cycles (Fig. S31-S33 †).32,33 Such amines can be more easily stabilized with a strong connement effect in the large mesopores of the exfGO consisting of considerable surface oxygen.comparison is made between TEPA@exfGO and the best literature samples of TEPA@carbon, 8 TEPA@silica, 14 TEPA@clay, 26 and also PEI@silica, 13,28 and diamine-appended-MOFs. 4,24,25 This is exactly the case with the TEPA (tetraethylenepentamine) @exfGO (Fig. 5 and S41 †).For instance, a 6.0 g g À1 TEPA impregnated exfGO-D sample shows a record-high CO 2 working capacity of over 28 wt% for ue-gas when investigated using cyclic temperature swing measurements, at 75 C and 100 C for uptake and desorption respectively.Aer 50 cycles of continuous operation for 120 h, the sample still exhibits an impressive working CO 2 capture of 22 wt%.2][23][24][25][26][27][28]32,33 Here it is worth noting that the capacity loss of z21% in our TEPA@exfGO sample is arguably smaller than the z40% capacity loss from TEPA@mesoporous silica capsules, under similar synthetic and experimental conditions. 14A notable capacity loss of up to 23% within the rst 5-10 uptake cycles is also observed for TEPA@silica samples at an extremely low loading, 1 g g À1 , of TEPA (see the references in Table S6 †). 8 Similarly, TEPA@activated carbon shows a capacity loss of up to 36% within 20 uptake cycles.A capacity loss of 14% is estimated in TEPA@clay, within the rst 10 uptake cycles, for a given loading, 1.5 g g À1 , of TEPA and at a much reduced desorption temperature of 90 C. 26 Such capacity losses are primarily ascribed to the leaching of amines in successive temperature swing desorption runs at high temperature, around 100 C.
We also note that the reversible CO 2 capacity of our TEPA@exfGO is comparable or superior to high molecular weight poly-and/or branched amines, e.g., PEI (polyethylenimine)@silica samples at similar working temperatures (Fig. 5b).In amine@silica systems, most of the amine is graed on the surface of the particles.Due to their relatively small pore sizes and pore volumes of #1.0 cm 3 g À1 , they invariably show unfavourable CO 2 capture performances (Table S6 †).A similar problem exists for solid-amine@mesoporous carbons.The other main drawback with these carbon based systems is the negligible surface oxygen, required to obtain a strong aminesurface interaction or amine graing.Therefore, the enhanced working capacity and stability of our solid-amine@exfGO are directly attributed to the well-impregnated amine within the oxygen-rich pores of ultrahigh pore volume and less oxidation due to the rapid dissipation of exothermic heat of reaction (Fig. 5 and S41 †). 28Furthermore, a volumetric CO 2 uptake capacity of 6.4 mmol g À1 (z28 wt%) in the exfGO-D Â 6.0TEPA sample, measured at 75 C and 0.15 bar, is in good agreement with the gravimetric ue-gas uptake (Fig. 5b and c).Similar to the TETA@exfGO samples, the pore volume directed TEPA loading and CO 2 uptake capacity can be understood from the comparative CO 2 uptake isotherms of the exfGO-D Â 6.0TEPA and exfGO-A Â 3.0TEPA samples.Moreover, as demonstrated earlier, these stable amines impregnated in the highly exfoliated GO based structures in their pre-humidied state can be well-considered for room temperature CO 2 scrubbers in onboard and conned living places (Fig. S34 †).
Finally, our results show that the graphene based materials can be tuned for high performance energy storage and guest molecular capture/storage via a simplied synthesis route by simply enhancing the pore population in the meso-/macro-porous region (see the summary of the synthesis methods of literature samples in Tables S4-S6 †).Most importantly, our solid-amine@exfGO samples can be favourably considered for an efficient column separation in a breakthrough method.These samples can exhibit tuneable volumetric densities, between 0.3 and 0.6 g cm À3 , depending on the strength of exfoliation and amine loading to achieve desirable permeation/ diffusion selectivities.The long-term cyclic uptake and further high temperature stabilities of solid-amine@exfGO can be easily achieved by choosing more stable polyamines.Graphene networks of ultrahigh meso-/macro-pore volumes can be effectively applied to other specic energy storage and conversion technologies.For example, these highly open-pore networked graphene samples can be ideal substrates for further functionalization.

Conclusions
We have achieved exceptional functional performances from meso-and macro-pore rich graphene-oxide based networks with ultrahigh total pore volumes, by developing a simple but facile route of thermal-shock exfoliation at a relatively mild temperature, 300 C. We have discovered that the degree of oxidation of the precursor graphene-oxide determines the nature and extent of exfoliation and porosity development.Through the control of the oxidation of the GO source material, a step-change in the total pore volume reproducibly exceeding 6 cm 3 g À1 is achieved.The meso-to macro-pore volume governed CO 2 capture and the capacitive energy storage performance of these structures is illustrated via amine loading and stability, and cyclic working CO 2 capture in solid-amine@exfGO.Specically, the solidamine@exfGO system, which is the rst material of its kind to use a graphene-oxide based porous structure for solid-amineimpregnation, exhibits such extraordinarily stable working CO 2 capture capacities of z25 wt% for a ue-gas stream.The simple exfoliation method described in this work is clearly a viable technique not only for scaling-up, but also for reducing the cost of synthesis and therefore materials production.

Experimental
Disclaimer Certain commercial equipment, instruments, or materials are identied in this paper to foster understanding.Such identication does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identied are necessarily the best available for the purpose.

Materials
All chemicals, such as synthetic graphite powder of <20 micron (Sigma-Aldrich), potassium permanganate, $99.0%(Sigma-Aldrich), phosphorus pentoxide, 99.0%, reagent plus (Sigma-Aldrich), potassium persulfate, 99+%, A.C.S. reagent (Sigma-Aldrich), sulphuric acid, 95-97%, puriss (Sigma-Aldrich), phosphoric acid, $85 wt%, A.C.S. reagent (Aldrich), hydrogen peroxide solution, 34.5-36.5% (Sigma-Aldrich), hydrochloric acid, 35%, technical 38% (VWR), triethylenetetramine, $97.0%(T) (Aldrich) and methanol, $99.5%, Emplura (Merck), were of commercial grade and used as-received.All gases or gas mixtures used were of research grade purity and purchased from BOC, UK.A, -B, -C & -D) samples of different degrees of oxidation were prepared by Hummers' and improved/modied Hummers' methods.A full detailed step-by-step synthesis is described in ESI.† Briey, to synthesize the GO-D sample, graphite powder (2.0 g) was added to a 9 : 1 mixture of concentrated H 2 SO 4 (45.0 ml) and H 3 PO 4 (5.0 ml) under stirring under cold conditions (z0 C).KMnO 4 (12.0 g) was then added slowly to prevent a sudden temperature rise, not more than 10 C. The solution became very dark green.The reaction mixture was then heated to 50 C and le for 3 hours under stirring.At this point the mixture has turned into a brown paste.De-ionized (DI) water (240.0ml) was added very slowly, again to prevent a sudden temperature rise of the highly exothermic reaction (water to a concentrated acid), to the mixture, and the mixture turned to a dark brown colour.The reaction mixture was le to stir for several minutes.Finally, H 2 O 2 (16.0 ml) was slowly added to the solution causing vigorous foaming and a colour change to bright yellow.The solution was stirred for another 30 minutes in a warm state (50 C) and le to settle overnight at room temperature.The solid product was separated from the excess liquid by centrifugation followed by decantation.The product was then washed in dilute HCl (3.4%, 750 ml) to remove any remaining salts, followed by further washing in DI water until the washings were pH neutral.The GO-D sample was obtained aer freeze-drying.

Exfoliation of the GO samples
Thermal exfoliation has been carried out at 300 C (unless specied) in a pre-heated vertical tube furnace.A step-by-step guide for the synthesis is described in ESI, Fig. S8-S10.† About 200-500 mg of GO sample was charged into a glass tube of 1.5 inch diameter and 12 inch long under ambient air.The tube was then sealed with a paper towel and placed in a furnace.From this point, the exfoliation can be observed in about 2 minutes; some of the samples have exfoliated in multiple steps.Thus the sample tube was still le in the furnace for additional two minutes for completing the process.Aer a total of 4 minutes the sample tube was removed from the furnace and le on a bench to cool down.Exfoliated samples named exfGO were collected for further characterization.No further chemical treatment has been performed.The as-exfoliated samples were examined straight away for all the characterizations including gas adsorption and amine loading.
A controlled sample, asGO-300C (without exfoliation), was also prepared as follows: the sample was placed in a tube furnace at room temperature and heated slowly at a heating rate of 3 C per minute to prevent exfoliation, and the furnace was cooled immediately aer reaching the temperature of 300 C.

Amine loading
TETA (or TEPA) impregnation in exfGO and the as-synthesized GO samples was carried out by methanol solution blending.A full detailed step-by-step method is described in ESI, Fig. S23.† 8 The required amount of amine (between 1 and 12 times the weight of the host sample) was added to the already weighed exfGO samples from the amine-methanol stock solution and later these solidsolution mixtures were aged for several hours on a shaker.Handling of all these samples was done at room temperature and under ambient air.The sample vials were then transferred into an air tight vacuum oven and le at 60 C up to a day for drying.The amine loading was estimated from the initial and the nal dry sample mass.The samples were labelled according to the amine loading with respect to the exfGO.For example, the loading of 4.0 g TETA for 1 g of exfGO sample is represented by exfGO Â 4.0, or exfGO Â 4.0TETA, i.e., equal to 80% loading.

Structure & porosity
Powder X-ray diffraction (PXRD, on a Stoe Stadi-P, Cu-K-alpha) was carried out by lling the sample in a 0.5 mm diameter borosilicate glass capillary under ambient conditions.Raman spectroscopy (514.5 nm laser, Renishaw) was carried out on hand pressed powder samples on a glass slide.X-ray photoemission spectroscopy (XPS, on an Al-K-alpha, Thermo Scientic), scanning electron microscopy (SEM, Jeol) and transmission electron microscopy (TEM, Jeol) measurements were carried out on the samples supported on a carbon tape or a carbon coated copper TEM grid.The porosity and gas adsorption-desorption isotherms up to 1 bar were determined at 77 K and 298 K using liquid nitrogen and water bath, respectively, on a Quantachrome Autosorb-iQC.The specic surface area was measured from the 77 K N 2 isotherm in a relative pressure range between 0.03 and 0.15, according to the Brunauer-Emmett-Teller (BET) method.The QSDFT (quenched solid density functional theory) method with slit/cylindrical pores (#50 nm) was applied to obtain pore size distribution, micro-pore volume and cumulative pore volume.The DH (Dollimore and Heal) method was applied to estimate meso-pores (>3 nm) and macro-pores (>50 nm).The total pore volume was estimated from the amount of N 2 adsorbed at a relative pressure, P/P 0 , of #0.994.Note that the total pore volume at P/P 0 #0.994 is fairly in good agreement with the total pore volume obtained through the DH method (meso-pore + macro-pore) + QSDFT method (micro-pore).The sample was degassed at 180 C overnight under dynamic vacuum prior to the actual gas adsorption measurement.Combined thermogravimetry-differential scanning calorimetry (TG-DSC, on a Setsys from Setaram) and mass spectrometry (MS, on an OmniStar from Pfeiffer Vacuum) up to 1000 C measurements were carried out on a dry sample under an Ar ow with a heating rate of 3-6 C per min.The TG mass-loss was recorded aer background correction to the empty alumina crucible.

CO 2 uptake measurements
For CO 2 uptake, all the samples were initially screened with volumetric adsorption isotherms using 100% dry CO 2 and measured up to 1 bar and at different temperatures between 30 and 75 C (maintained by using a water bath) on a Quantachrome Autosorb-iQC.The sample was degassed at 70 C prior to the actual CO 2 adsorption isotherm measurement.
The gravimetric and temperature swing cyclic CO 2 uptake tests were performed by TG under a constant gas ow at 100 ml min À1 around 1 bar.The tests were carried out under 3 different conditions of the test gas: (1) humidied (bubbled through a water bubbler) 15% CO 2 balanced with 85% N 2 , (2) humidied 100% N 2 , and (3) 15% dry CO 2 , as a reference.In each case the CO 2 desorption cycle was obtained with a dry N 2 ow (100 ml min À1 ) at z100 C. The sorption tests were conducted at different temperatures of 30, 50, 65, and 75 C.As the experiment involved switching between the gases for each sorption and desorption run, the measurements were carried out in the daytime and le with CO 2 overnight at room temperature to start the subsequent cycling the next day.
Note that the volumetric CO 2 uptakes are shown in "mmol g À1 ", whereas the gravimetric capacities are in "wt%".

Fig. 1
Fig. 1 Top: schematic representation of GO-samples from GO-A to GO-D structures with increasing order of oxidation state.Bottom: (a and b) PXRD patterns and XPS C 1s spectra of GO samples.(c and d) Digital images of GO and exfGO samples of equal mass, 32.5 mg in a 4 ml vial.(e-g) Schematic mechanism of reduced GO (rGO) and exfGO formation at different heating rates; GO (e) thermogravimetric curves at a slow heating rate of 3 C min À1 to yield rGO (f), and fast heating or thermal-shock to produce exfGO (g).

Fig. 2
Fig. 2 Development of porosity against increasing degree of oxidation of the GO samples: (a and b) N 2 adsorption-desorption isotherms at 77 K (a) and their derived pore size (line data on the left Y-axis) and pore volume (dotted data on the right Y-axis) distributions (b).The inset in (a) shows the pressure region near P/P 0 z 1 and the clear maximum uptake differences between samples.(c and d) The comparison of the total pore volume (c) and SSA (d) of our exfGO and a large number of samples (over 250) from the literature.(e and f) Porosity (pore volume and SSA) generation as a function of oxidation state of the precursor GO samples.The development of the porosity is directly proportional to the oxidation strength of GO, where a drastic enhancement in the porosity is observed with an increased degree of oxidation, i.e., conversion of sp 2 C]C to C-O/C]O/COO.Insets (e and f) showing the linear increase of porosity with respect to the carboxylic group C]O/COO formation (see the ESI †).

Fig. 3
Fig. 3 Surface morphology and porosity development determined by electron microscopy: (a-d) comparative SEM micrographs of exfGO-A (a and c) and exfGO-D (b and d) samples.(a and b) and (c and d) respectively represented at a similar magnification.Middle-bottom rows show the SEM and TEM micrographs of exfGO-D samples with highly networked structures.

Fig. 4
Fig. 4 (a) SEM micrograph of exfGO-D Â 7.0TETA.(b) XPS C 1s core level spectra.(c) Volumetric CO 2 (100% dry) uptake isotherms at 75 C.(d) Gravimetric flue-gas stream uptake isotherms at 75 C. Solid and dotted line data represent the flue-gas uptake and humidified N 2 uptake, respectively.TETA loading in all plots is given in g g À1 and indicated by multiplication numbers.