General and Predictable Synthesis of Ultrahigh-Surface-Area Porous Carbons with Superior Yield via Preferential Removal of sp2-Hybridized Atoms


 A grand challenge in the state-of-the-art porous carbons is the lack of reliable synthesis strategy for achieving ultrahigh surface areas while maintaining a high carbonization yield. Ultrahigh surface area generally depends on trial and error activation with poor understanding of structural information in the starting carbonaceous matter to predict the ultrahigh porosity. Meanwhile, excessive development of porosity (> 3500 m2 g− 1) will undoubtedly give rise to low carbonization yield (< 10%), thus far restricting cost-effective applications. Here, we report a general and predictable protocol via constructing nitrogen-doped sp2-hybridized carbon atoms in the carbonaceous matter, which guides the pore-creating agents (e.g., KOH) to preferentially etch over sp2- rather than sp3-hybridized atoms, thus greatly increasing the activation reaction efficiency to simultaneously accomplish ultrahigh porosity without sacrificing carbonization yield, a critical paradox in producing carbons. A highest surface area (4482 m2 g− 1) with 10 wt.% carbonization yield and 3500 m2 g− 1 with an unparalleled yield of 35% are achieved so far, which enables great potential in adsorptive-related applications as exemplified by their record-high gas adsorption and supercapacitve performances. Our findings reveal important insights on directed synthesis of ultrahigh-surface-area carbons and provide an impetus for their on-demand applications.


Introduction
Carbon materials, such as graphenes 1 , carbon nanotubes 2 , fullerenes 3 , and porous carbons 4 , constitute a class of versatile materials that nd indispensable utility in a wide range of potential and practical applications. Porous carbons, integrating the advantages of carbon materials with porous solids, have specially gathered intense interests due to their outstanding physicochemical properties and their extensive application elds 5,6 . Bene ting from the fast-growing synthesis strategy and nanotechnology during the two past decades, porous carbons have reached a new level of precision control over pore structure with rapid advancement of surface areas. Despite tremendous previous efforts, porous carbons produced by activation still stand out to be competitive taking into account of its feasibility to attain much higher surface area and the low cost, thus dominating the market of porous carbons.
Essentially, pursuing ultrahigh speci c surface areas has been a long-pursued goal for the eld of porous materials due to the high porosity for providing large surface area for dispersion of active sites could lead to unconventional properties, and thus opens unprecedented opportunities for breakthroughs in the emerging applications or further expands in their application scopes. Continuous progress in speci c surface area has been achieved in the state-of-the-art porous materials, such as, zeolites(i.e., 4100 m 2 g -1 ) 7 , metal-organic framework (i.e., MOF, 9140 m 2 g -1 ) 8 and porous organic polymers (i.e., POP, 6461 m 2 g -1 ) 9 . For porous carbons, chemical activation of biomass and polymer has been well established to prepare porous carbons with large surface area generally up to 3000 m 2 g -1 , and only very few examples have been reported to prepare higher values over 3500 m 2 g -110, 11 by selecting speci c precursors.
Nevertheless, the realization of these high surface areas is on account of trial and error activation, and there are no material parameters or characteristics of the starting carbonaceous matter that can be used to predict the ultrahigh porosity. Moreover, the harsh chemical activation condition usually gives rise to a low yield below 10 wt.% when the surface area up to 3500 m 2 g -1 , which is mainly caused by the burn-off effect during the forming of pores. This yields a high preparation cost of super activated carbons, such as the price of classical activated carbon YP-50 up to about 257 $ per kilogram. Therefore, obtaining the ultrahigh BET surface area along with the satis ced yield can vitalize the high performance activated carbon market for industrialized producing at a low cost. However, the research in this eld lacks attention.
In this contribution, we show a concept for combination of ultrahigh porosity and large yield in carbon materials through preferential removal of sp 2 -hybridized carbon atoms. Key to this strategy is construction of nitrogen-doped (N-doped) sp 2 -hybridized carbon atoms in the precursors, which act as "magnet" with a strong attraction to pore-creating agents, e.g., KOH, to guide and preferentially etch over the sp 2 -hybridized carbon atoms. Such a priority of removal of sp 2 -hybridized carbon atoms but not sp 3hybridized carbon atoms greatly increases the activation reaction e ciency when compared with routine approaches, in which the sp 3 -hybridized carbon atoms are more reactive and preferred to be etched. As a consequence, the speci c surface area and the yield are enhanced simultaneously. In this way, we prepared a carbon material with the highest Brunauer-Emmett-Teller (BET) speci c surface area (i.e., 4482 m 2 g − 1 ) reported so far, accompanying with an unconventionally large yield of 10.0 wt.%. These values are not only 1.3 times (for speci c surface area) or 3.2 times (for yield) enhancement compared to those of prepared by conventional means, but also hold a signi cant superiority over all those reported in the literatures (Supplementary Table 1). The present synthetic strategy is rather general, which can be utilized to synthesize a series of porous carbons with simultaneously ultrahigh speci c surface area over 3500 m 2 g − 1 and large yield up to 35.1 wt.% from various carbon precursors including biomass (e.g., mushroom), natural macromolecule (e.g., cellulose) and synthetic polymer (e.g., phenolic resin).
Bene ting from the well-developed porous structure, the resulting porous carbons offer a promising opportunity to boost the performances to a new stage. For instance, high H 2 adsorption capacity of 8.4 wt.% at 40 bar, large CO 2 capture capacity of 39.6 mmol g − 1 at 40 bar, huge supercapacitive capacitance of 634 F g − 1 can be achieved, which exceed those of nearly all the benchmark porous carbons. Atoms Peifeng Yu, Weicai Zhang, Yingliang Liu, Fei Xu, Yeru Liang Abstract A grand challenge in the state-of-the-art porous carbons is the lack of reliable synthesis strategy for achieving ultrahigh surface areas while maintaining a high carbonization yield. Ultrahigh surface area generally depends on trial and error activation with poor understanding of structural information in the starting carbonaceous matter to predict the ultrahigh porosity. Meanwhile, excessive development of porosity (> 3500 m2 g− 1) will undoubtedly give rise to low carbonization yield (< 10%), thus far restricting cost-effective applications.

Results And Discussion
Here, we report a general and predictable protocol via constructing nitrogen-doped sp2-hybridized carbon atoms in the carbonaceous matter, which guides the pore-creating agents (e.g., KOH) to preferentially etch over sp2-rather than sp3-hybridized atoms, thus greatly increasing the activation reaction e ciency to simultaneously accomplish ultrahigh porosity without sacri cing carbonization yield, a critical paradox in producing carbons. A highest surface area (4482 m2 g− 1) with 10 wt.% carbonization yield and 3500 m2 g− 1 with an unparalleled yield of 35% are achieved so far, which enables great potential in adsorptiverelated applications as exempli ed by their record-high gas adsorption and supercapacitve performances.
Our ndings reveal important insights on directed synthesis of ultrahigh-surface-area carbons and provide an impetus for their on-demand applications. KEYWORDS ultrahigh-surface-area porous carbons, synthesis strategy, superior yield, sp2-hybridized atoms Edit Fulltext Introduction Carbon materials, such as graphenes 1 , carbon nanotubes 2 , fullerenes 3 , and porous carbons 4 , constitute a class of versatile materials that nd indispensable utility in a wide range of potential and practical applications. Porous carbons, integrating the advantages of carbon materials with porous solids, have specially gathered intense interests due to their outstanding physicochemical properties and their extensive application elds 5,6 . Bene ting from the fast-growing synthesis strategy and nanotechnology during the two past decades, porous carbons have reached a new level of precision control over pore structure with rapid advancement of surface areas. Despite tremendous previous efforts, porous carbons produced by activation still stand out to be competitive taking into account of its feasibility to attain much higher surface area and the low cost, thus dominating the market of porous carbons.
Essentially, pursuing ultrahigh speci c surface areas has been a long-pursued goal for the eld of porous materials due to the high porosity for providing large surface area for dispersion of active sites could lead to unconventional properties, and thus opens unprecedented opportunities for breakthroughs in the emerging applications or further expands in their application scopes. Continuous progress in speci c surface area has been achieved in the state-of-the-art porous materials, such as, zeolites(i.e., 4100 m 2 g -1 ) 7 , metal-organic framework (i.e., MOF, 9140 m 2 g -1 ) 8 and porous organic polymers (i.e., POP, 6461 m 2 g -1 ) 9 . For porous carbons, chemical activation of biomass and polymer has been well established to prepare porous carbons with large surface area generally up to 3000 m 2 g -1 , and only very few examples have been reported to prepare higher values over 3500 m 2 g -110, 11 by selecting speci c precursors.
Nevertheless, the realization of these high surface areas is on account of trial and error activation, and there are no material parameters or characteristics of the starting carbonaceous matter that can be used to predict the ultrahigh porosity. Moreover, the harsh chemical activation condition usually gives rise to a low yield below 10 wt.% when the surface area up to 3500 m 2 g -1 , which is mainly caused by the burn-off effect during the forming of pores. This yields a high preparation cost of super activated carbons, such as the price of classical activated carbon YP-50 up to about 257 $ per kilogram. Therefore, obtaining the ultrahigh BET surface area along with the satis ced yield can vitalize the high performance activated carbon market for industrialized producing at a low cost. However, the research in this eld lacks attention.
In this contribution, we show a concept for combination of ultrahigh porosity and large yield in carbon materials through preferential removal of sp 2 -hybridized carbon atoms. Key to this strategy is construction of nitrogen-doped (N-doped) sp 2 -hybridized carbon atoms in the precursors, which act as "magnet" with a strong attraction to pore-creating agents, e.g., KOH, to guide and preferentially etch over the sp 2 -hybridized carbon atoms. Such a priority of removal of sp 2 -hybridized carbon atoms but not sp 3hybridized carbon atoms greatly increases the activation reaction e ciency when compared with routine approaches, in which the sp 3 -hybridized carbon atoms are more reactive and preferred to be etched. As a consequence, the speci c surface area and the yield are enhanced simultaneously. In this way, we prepared a carbon material with the highest Brunauer-Emmett-Teller (BET) speci c surface area (i.e., 4482 m 2 g − 1 ) reported so far, accompanying with an unconventionally large yield of 10.0 wt.%. These values are not only 1.3 times (for speci c surface area) or 3.2 times (for yield) enhancement compared to those of prepared by conventional means, but also hold a signi cant superiority over all those reported in the literatures (Supplementary Table 1). The present synthetic strategy is rather general, which can be utilized to synthesize a series of porous carbons with simultaneously ultrahigh speci c surface area over 3500 m 2 g − 1 and large yield up to 35.1 wt.% from various carbon precursors including biomass (e.g., mushroom), natural macromolecule (e.g., cellulose) and synthetic polymer (e.g., phenolic resin). Bene ting from the well-developed porous structure, the resulting porous carbons offer a promising opportunity to boost the performances to a new stage. For instance, high H 2 adsorption capacity of 8.4 wt.% at 40 bar, large CO 2 capture capacity of 39.6 mmol g − 1 at 40 bar, huge supercapacitive capacitance of 634 F g − 1 can be achieved, which exceed those of nearly all the benchmark porous carbons.
Preview Carbon materials, such as graphenes1, carbon nanotubes2, fullerenes3, and porous carbons4, constitute a class of versatile materials that nd indispensable utility in a wide range of potential and practical applications. Porous carbons, integrating the advantages of carbon materials with porous solids, have specially gathered intense interests due to their outstanding physicochemical properties and their extensive application elds5, 6. Bene ting from the fast-growing synthesis strategy and nanotechnology during the two past decades, porous carbons have reached a new level of precision control over pore structure with rapid advancement of surface areas. Despite tremendous previous efforts, porous carbons produced by activation still stand out to be competitive taking into account of its feasibility to attain much higher surface area and the low cost, thus dominating the market of porous carbons. Essentially, pursuing ultrahigh speci c surface areas has been a long-pursued goal for the eld of porous materials due to the high porosity for providing large surface area for dispersion of active sites could lead to unconventional properties, and thus opens unprecedented opportunities for breakthroughs in the emerging applications or further expands in their application scopes. Continuous progress in speci c surface area has been achieved in the state-of-the-art porous materials, such as, zeolites(i.e., 4100 m2 g-1)7, metal-organic framework (i.e., MOF, 9140 m2 g-1)8 and porous organic polymers (i.e., POP, 6461 m2 g-1)9. For porous carbons, chemical activation of biomass and polymer has been well established to prepare porous carbons with large surface area generally up to 3000 m2 g-1, and only very few examples have been reported to prepare higher values over 3500 m2 g-110, 11 by selecting speci c precursors. Nevertheless, the realization of these high surface areas is on account of trial and error activation, and there are no material parameters or characteristics of the starting carbonaceous matter that can be used to predict the ultrahigh porosity. Moreover, the harsh chemical activation condition usually gives rise to a low yield below 10 wt.% when the surface area up to 3500 m2 g-1, which is mainly caused by the burn-off effect during the forming of pores. This yields a high preparation cost of super activated carbons, such as the price of classical activated carbon YP-50 up to about 257 $ per kilogram. Therefore, obtaining the ultrahigh BET surface area along with the satis ced yield can vitalize the high performance activated carbon market for industrialized producing at a low cost. However, the research in this eld lacks attention. In this contribution, we show a concept for combination of ultrahigh porosity and large yield in carbon materials through preferential removal of sp2-hybridized carbon atoms. Key to this strategy is construction of nitrogen-doped (N-doped) sp2-hybridized carbon atoms in the precursors, which act as "magnet" with a strong attraction to pore-creating agents, e.g., KOH, to guide and preferentially etch over the sp2-hybridized carbon atoms. Such a priority of removal of sp2-hybridized carbon atoms but not sp3-hybridized carbon atoms greatly increases the activation reaction e ciency when compared with routine approaches, in which the sp3-hybridized carbon atoms are more reactive and preferred to be etched. As a consequence, the speci c surface area and the yield are enhanced simultaneously. In this way, we prepared a carbon material with the highest Brunauer-Emmett-Teller (BET) speci c surface area (i.e., 4482 m2 g− 1) reported so far, accompanying with an unconventionally large yield of 10.0 wt.%. These values are not only 1.3 times (for speci c surface area) or 3.2 times (for yield) enhancement compared to those of prepared by conventional means, but also hold a signi cant superiority over all those reported in the literatures (Supplementary Table 1). The present synthetic strategy is rather general, which can be utilized to synthesize a series of porous carbons with simultaneously ultrahigh speci c surface area over 3500 m2 g− 1 and large yield up to 35.1 wt.% from various carbon precursors including biomass (e.g., mushroom), natural macromolecule (e.g., cellulose) and synthetic polymer (e.g., phenolic resin). Bene ting from the well-developed porous structure, the resulting porous carbons offer a promising opportunity to boost the performances to a new stage. For instance, high H2 adsorption capacity of 8.4 wt.% at 40 bar, large CO2 capture capacity of 39.6 mmol g− 1 at 40 bar, huge supercapacitive capacitance of 634 F g− 1 can be achieved, which exceed those of nearly all the benchmark porous carbons. Results And Discussion The chemical etching processes for generating porous structure are schematically illustrated in Fig. 1, in which KOH is used as the pore-creating agents for a model demonstration. As the starting materials, semi-carbonized precursors obtained at low temperatures usually consist of sp 2 -hybridized and sp 3hybridized carbon atoms 12 . With increasing the activation temperature, the KOH will start to react with carbon atoms. In a normal KOH activation procedure, the sp 3 -hybridized carbon atoms are preferred to be etched because of the more reactive of sp 3 -hybridized carbon atoms than sp 2 -hybridized carbon atoms ( Fig. 1a) 13 . It is known that the disordered packing of the carbon sheets that is crosslinked by sp 3hybridized carbon atoms synergistically signi cantly contribute to the generation of the pore structure 12,14 . Therefore, the preferential removal of sp 3 -hybridized carbon atoms not only easily collapse the porous architectures, but also leads to a poor activation e ciency, resulting in relatively low porosity and low yield (Fig. 1a).
To circumvent this intrinsic limitation, we proposed to construct N-doped sp 2 -hybridized carbon atoms in the semi-carbonized precursors to guide KOH to preferentially etch over the sp 2 -hybridized carbon atoms.
We rst performed density functional theory (DFT) to evaluate the effect of the N species on KOH etching process. In the DFT calculations, KOH is simulated as an individual hydroxyl group for convenience. As shown in Fig. 2a, site 1, site 2 and site 3 indicate three different distances between the hydroxyl groups adsorbed on carbon atoms and N atoms. The corresponding adsorption energy (∆E) between KOH and Ndoped carbon was calculated and listed in Fig. 2b. It can be seen that the hydroxyl group at site 1 nearest to N atom has the most negative ∆E of -2.64 eV. Moreover, the ∆E gradually increased from − 2.07 to -2.00 when increasing the distance between the hydroxyl group and N atom. However, all the ∆E of Ndoped carbon is smaller than that of pristine carbon without N atom (i.e., -1.35 eV), which indicates that the hydroxyl group has a stronger interaction with N-doped carbon than pristine carbon. Meanwhile, the reactant and product were predicted by intrinsic reaction coordinate by calculating the transition state ( Fig. 2c). As shown in Fig. 2d, the energy barrier between the hydroxyl group and N-doped carbon (i.e., 0.12 eV) is obviously smaller than that of pristine carbon (i.e., 0.66 eV). All above results indicate that KOH is preferred to react with N-doped carbon; especially, the region of N atom is preferentially etched by KOH.
Guided by the theoretical calculations, we then set out to introduce N heteroatoms by pyrolysis of carbon precursor and N-enrich resource, which utilize mushroom and melamine as typical demonstrations, respectively. As shown in Supplementary Fig. 1, the resultant semi-carbonized mushroom possesses a high N content of 10.9 at.% as determined by X-ray photoelectron spectroscopy (XPS). High-resolution N1s spectrum shows that the main peak is centered at 399.18 eV assigned to pyridinic N, illustrating that the hybridization between N and carbon atoms is mainly sp 2 in the semi-carbonized mushroom scaffolds ( Supplementary Fig. 2) 15 .
We further heated the mixture composed of semi-carbonized mushroom and KOH at high temperature in N 2 ow. After washing and drying, we obtained the mushroom-based porous carbon (M-PC). Its pore structure characteristic was probed by N 2 adsorption-desorption measurements at 77 K ( Supplementary   Fig. 3). As shown in Supplementary Fig. 3a, the M-PC exhibits type I isotherm with signi cant adsorption at low relative pressure (P/P 0 < 0.01), which indicates the presence of tremendous micropores. After that, the adsorption amount increases gradually until the P/P 0 near 0.4, demonstrating the formation of abundant small-sized mesopores. According to the pore size distribution curves in Supplementary Fig. 3b, the size of these micropores and mesopores are mainly centered at about 1.2 and 2.5 nm, respectively. High-resolution transmission electron microscopy (TEM) image con rmed the existence of these nanopores distributed in the carbon framework (Fig. 3a). A BET calculation reveals that the BET speci c surface area of M-PC is as high as 4482 m 2 g -1 (Fig. 3b). To the best of our knowledge, this is the highest speci c surface area for all carbon materials reported so far (Fig. 3c and Supplementary Table 2). In sharp contrast, a mushroom-based control carbon sample (M-control sample) prepared by a normal strategy without N atoms has an obviously smaller BET speci c surface area of 3428 m 2 g -1 . More importantly, although possessing an ultrahigh speci c surface area, M-PC displays an unusual enhanced yield as high as 10.0 wt.%, which is almost 3.2 times larger than that of the control sample, i.e., 3.1 wt.% (Fig. 3b).
In addition to natural mushroom, this e cient strategy can be well extended to other kinds of carbon precursors such as natural macromolecule (e.g., cellulose) and synthetic polymer (e.g., phenolic resin). Similarly, both BET speci c surface area and yield of the as-prepared carbon materials are enhanced at the same time. For example, as shown in Fig. 3d, the BET speci c surface areas of the cellulose-based porous carbon (C-PC) and phenolic resin-based porous carbon (PR-PC) are calculated to be 3504 and 3954 m 2 g -1 , respectively, which are signi cantly higher than those of the corresponding control samples (i.e., 2929 and 3167 m 2 g -1 ). Meanwhile, the yields of C-PC and PR-PC are remarkably enhanced from 6.0 wt.% and 20.8 wt.% to 10.0 wt.% and 35.1 wt.%, respectively (Fig. 3e). As shown in Fig. 3f and Supplementary Table 1, these resulting porous carbons are super to these reported previous considering the comprehensive properties of speci c surface area and yield. These results indicate that ultrahigh porosity and large yield, which is an intrinsic con ict under normal circumstances, can be simultaneously combined into carbon material by constructing N heteroatoms in the semi-carbonized precursors for preferential removal of sp 2 -hybridized carbon atoms. Considering its thermally decomposable characteristics ( Supplementary Fig. 4), the N-enrich resource, i.e., melamine makes very little contribution to carbon yield and is believed to just provide N heteroatoms to affect the pore-marking process.
To understand the detailed pore-making mechanism with the introduction of N heteroatoms, we evaluated the phase and structural change of the intermediate products. As well known, the pore formation in the KOH activation underling the process complexity usually originates from the combustion of carbon species treated with KOH 36 . It has been demonstrated the traditional KOH activation generally proceeds as 6KOH + C = 2K + 3H 2 + 2K 2 CO 3 followed by reaction of K/K 2 CO 3 /CO 2 with carbon species and/or decomposition of K 2 CO 3 (Supplementary Eqs. 1 ~ 5) 37 . Typically, KOH starts to react with carbon atoms to yield K 2 CO 3 at around 400 o C according to Supplementary Eq. 1 38 , which was con rmed by the XRD measurement (Fig. 4a). At this stage, the sp 3 -hybridized carbon atoms but not the sp 2 -hybridized carbon atoms are preferred to be etched due to its stronger reactivity (Fig. 1a) 13 . This reaction continues far till 700 o C (Fig. 4a). When the reaction temperature achieves to 700 o C, the K 2 CO 3 reacts further as the Supplementary Eqs. 2 ~ 5 with the release of CO and CO 2 gas to create the pores 38 . Since those sp 3hybridized carbon atoms plays important roles in crosslinking the graphite-like microcrystals within the semi-carbonized framework ( Supplementary Fig. 5), the graphite-like microcrystals are no longer restricted upon the removal of the sp 3 -hybridized carbon atoms 39 . Correspondingly, the carbonaceous porous framework is destroyed with constantly increasing the temperatures, collapsing into nanosheetlike structure with a low porosity (Fig. 4b and Supplementary Fig. 6) rather than the original irregular bulk ( Supplementary Fig. 5). Furthermore, in order to construct a high porous structure, more carbon species have to be consumed, leading to a low yield (Supplementary Fig. 7).
On the contrary, with the presence of N heteroatoms, KOH is inclined to react with carbon atoms close to N atoms (i.e., sp 2 -hybridized carbon atoms) due to the strong interaction between KOH and N atoms.
Then the KOH reacts as the Eq. 1 with the N atoms and their neighboring sp 2 -hybridized carbon atoms to form KCN and KOCN with gradually increasing the activation temperatures (Fig. 4a). Such a priority of removal of the sp 2 -hybridized carbon atoms not only retains the original carbonaceous porous bulk framework ( Fig. 4b and Supplementary Fig. 8), but also makes fewer carbon atoms etched to realize a high porosity when the activation process is performed (Eqs. 2 ~ 4), thus increasing the activation reaction e ciency accompanied by ultrahigh porosity and large yield. As a consequence, the emerged carbon materials possessed a higher proportion of sp 3 -hybridized carbon atoms (i.e., a lower proportion of sp 2 -hybridized carbon atoms) when compared with those materials obtained in a regular activation procedure. For example, the M-PC exhibited a decreased (002) diffraction intensity assigned to the sp 2hybridized carbon in the XRD patterns ( Supplementary Fig. 9), a lower characteristic peak intensity around 1580 cm − 1 (G-band) ascribed to the sp 2 -hybridized carbon in the Raman spectra ( Supplementary   Fig. 10) and a calculated higher value of I D /I G as compared to the control carbon sample ( Supplementary   Fig. 11). Moreover, the calculated values of stack height (L c ) and stack width (L a ) based on the results of XRD and Raman measurements for M-PC is much lower than the control sample, intuitively highlighting that the lower ratio sp 2 -hybridized carbon atoms in the sample M-PC ( Fig. 4c and 4d). Quantitative calculation from X-ray photoelectron spectroscopy (XPS) analysis showed that the ratio between the sp 3hybridized carbon and sp 2 -hybridized carbon atoms in the M-PC is as high as 0.9, which is 3 times higher than that in the control sample (i.e., 0.3) as illustrated in Supplementary Fig. 12 and Fig. 4e. This result is well consistent with those obtained from Raman and XRD measurements, further con rming that the introduction of N heteroatoms indeed affects the removal process of carbon atoms and improve the activation e ciency and yield concurrently.
The exceptionally well-developed porosity together with the remarkable yield makes the as-fabricated porous carbons very attractive in their practical application elds. Particularly, the ultrahigh speci c surface area of 4482 m 2 g -1 in M-PC provides an unprecedented platform for boosting their performance to a new stage. Herein, we focused on the adsorption (i.e., H 2 uptake and CO 2 capture) and electrochemical (i.e., supercapacitive) performances. Given that H 2 storage in porous carbons is favored by high speci c surface area, we assessed the H 2 uptake properties of M-PC at 77 K in the pressure range from 20 to 40 bar (Supplementary Fig. 13). As shown in Fig. 5a, the excess hydrogen uptake of M-PC is 5.8 wt.% at 20 bar and increases in line with the rise in adsorption pressure. An impressive excess hydrogen uptake achieves to 8.4 wt.% at 40 bar, which is at the top end of those values previously reported for carbon materials, and indeed is comparable to or even superior to those ones obtained in metal organic framework with BET speci c surface area exceeding 6000 m 2 g -1 (Fig. 5c and Supplementary Table 3). Such superiority in adsorption also enables great potential for their application in CO 2 capture (Supplementary Fig. 14). As summarized in Fig. 5b, the M-PC delivers CO 2 adsorption capacities of 24.4, 30.2 and 39.6 mmol g -1 at 20, 30 and 40 bar, respectively. So far as known, these values are the highest adsorption capacities ever reported for porous carbons (Fig. 5c and Supplementary Table 4).
The preponderance of the presented carbons in application is further demonstrated in energy storage. For example, when used as supercapacitor electrode, the M-PC can supply su cient active sites for ion storage and then possesses outstanding electrochemical performance. Galvanostatic charge-discharge test was utilized to evaluate the capacity of the as-constructed supercapacitors ( Supplementary Fig. 15). A calculation based on these galvanostatic discharge curves demonstrates that M-PC showed an intriguing speci c capacitance value as high as 634 F g -1 at 0.1 A g -1 , which is higher than that of commercial activated carbon benchmark and is one of the highest values obtained from all porous carbons ( Fig. 5d and Supplementary Table 5). Even increasing the current density 500 times to an ultrahigh value, i.e., 50 A g -1 , a remarkable value of 266 F g -1 can be achieved (Fig. 5d), indicative of its fast ion storage ability and rate performance. These very attractive electrochemical performances highlight the great potential of suitability of M-PC for advanced supercapacitors, which presents a positive potential for bridging the performance gap between supercapacitors and lithium-ion battery.
In summary, we have developed an e cient strategy to prepare porous carbons with ultrahigh porosity and large yield by preferential removal of sp 2 -hybridized carbon atoms. The introduction of N-doped sp 2hybridized carbon atoms was demonstrated to guide the KOH to preferentially etch over the sp 2hybridized carbon atoms, which greatly improved the activation reaction e ciency to simultaneously intensify the porosity and yield. The resulting carbon material from natural mushroom showed an ultrahigh speci c surface area of 4482 m 2 g − 1 and an impressively large yield of about 10 wt.%, which gave rise to excellent adsorption and electrochemical performance in practical application. Also, this strategy had proved its generalization to other kinds of carbon precursors such as natural macromolecule (e.g., cellulose) and synthetic polymer (e.g., phenolic resin), which could provide a powerful conceptual approach to generate advanced porous carbon materials. Overall, this study provides deep understating and opens up the possibilities for overcoming the major issues that how to cost-effectively prepare porous carbon with high speci c surface area and desirable yield through an ingenious design for the precursors.

Materials
Mushroom (Pleurotus geesteranus) was purchased from Dongyuanxing, Jingdong Mall. The mushroom stalk was used as the starting material obtained after dried at 70 o C for 12 h. Cellulose with an average particle size of 90 μm was purchased from Macklin. Potassium hydroxide (KOH, Guangzhou Chemical Reagent Factory), ethanol (Guangzhou Chemical Reagent Factory), hydrochloric acid (HCl, Guangzhou Chemical Reagent Factory) and other chemicals were used as received. The phenolic resin was synthesized in our lab according to the literature previously reported 64 . Typically, 100 g resorcinol, 126 mL formaldehyde (40 wt.%), and 0.17 g hexadecyl trimethyl ammonium bromide were dissolved in 148 mL deionized water with stirring for 0.5 h, and then placed in an oven for 5 days at 85 o C. After that, the product was taken out in a ventilated and dry environment and air-dried for 3 days to readily obtain phenolic resin.

Materials synthesis
Preparation of highly porous carbons. In a typical process, the dried mushroom stalk was carbonized at 400 o C for 1 h. Hereafter, the carbonized mushroom stalk was mixed with melamine (the mass ratio of melamine/carbonized mushroom is 1) and KOH (the mass ratio of KOH/carbonized mushroom weight is 4), and the mixture was rstly maintained at 100 o C for 1 h then increased to 800 o C for 1 h with a ramp rate of 5 o C min -1 in the N 2 atmosphere (about 60 ml min -1 ). The resulting products were washed with 2 mol L -1 HCl and distilled water until the pH was ~ 7. After drying in 105 o C for 4 h, the targeted sample mushroom-based porous carbon (M-PC) was obtained. The synthetic procedures of other highly porous carbons were similar to M-PC except for using different carbon precursors.
Preparation of control samples. For comparison, the control samples were synthesized by using traditional KOH activation methods. In detail, the dried mushroom stalk was carbonized at 400 o C for 1 h. Hereafter, the carbonized mushroom stalk was mixed with KOH at a weight ratio of 1:4, and the mixture was rstly maintained at 100 o C for 1 h then increased to 800 o C for 1 h with a ramp rate of 5 o C min -1 in the N 2 atmosphere (about 60 ml min -1 ). The resulting products were washed with 2 mol L -1 HCl and distilled water until the pH was ~ 7. After drying in 105 o C for 4 h, the control sample (M-control sample) was obtained. The synthetic procedures of other corresponding control samples were similar except for using different carbon precursors.

Density functional theory calculations
All calculations were performed by using the rst-principles plane-wave pseudopotential formulation [65][66][67] as implemented in the Vienna ab-initio Simulation Package (VASP). The exchange-correlation functional was in the form of Perdew-Burke-Ernzerhof (PBE) 68 with the generalized gradient approximation (GGA).
The cut-off energy of 400 eV for the plane-wave basis, and Γ-cantered k-mesh 69 of 3×4×1 for the calculations were applied to insure the energy convergence was 10 −5 eV and the residual force acting on each atom was less than 0.01 eV Å -1 . It should be noted that we adopted graphene with one carbon atom of supercell substituted by nitrogen atom (C 47 N) to model the N-doped graphitic, and the pristine graphene sheet was used as non N-doped graphitic. To eliminate interactions between the 2D atomic layer and its periodic images, we used a vacuum distance larger than 14 Å for the supercell geometry.
Especially, the adsorption energy was calculated by the following formula: where E G , E G+OH and E OH was the total energy of pristine or N-dopded graphene (G), the hydroxyl (OH) adsorbed system and isolated OH molecule, respectively.
Preparation operations for ex-situ XRD and ex-situ SEM analysis Ex-situ XRD. The preparation of the intermediate products was similar to that of x-PC except for different heating temperature and without any post-treatment. Typically, the mixture containing semi-raw material, KOH and melamine was heated to a desired temperature (e.g., 400, 500, 600, 700 and 800 o C) with a heating rate of 5 o C min -1 in the N 2 atmosphere and cooled down naturally. The resulting intermediate products without any were collected for ex-situ XRD analysis.
Ex-situ SEM. The resulting intermediate products of ex-situ XRD were washed with 2 mol L -1 HCl and distilled water until the pH was about 7.0, and then dried at 105 o C for 4 h. The nal products were collected for ex-situ SEM analysis.

Material characterization
The morphology of the samples was acquired by eld emission electron microscopy (FESEM, HITACHI SU8220) and high-resolution transmission electron microscopy (HRTEM, JEM2100 HR). Raman spectra were obtained by a Jobin-Yvon HR800 micro Raman spectrophotometer (λ=457.9 nm). XRD data were obtained using a powder X-ray diffractometer (Rigaku-Ultima IV, Cu Kα, λ=0.15405 nm). The surface chemical composition was analyzed by X-ray photoelectron spectroscopy (XPS, Nexsa Thermo Fisher Scienti c, USA). Thermogravimetry (TG) was performed using an METZSCH STA 449 F5 system. The pore structure of all the samples was determined from N 2 adsorption-desorption isotherms in 77 K in an ASAP 3Flex automatic volumetric sorption analyzer (Micromeritics sorption analyzer). All samples were outgassed at 350 o C for 8 h before analysis. The speci c surface area was calculated by Brunauer-Emmett-Teller (BET) theory. The pore size distribution was analyzed by the nonlocal density functional theory.

Gas adsorption measurements
The purity of H 2 and CO 2 in the gas adsorption measurements was 99.9999%. The H 2 and CO 2 uptake tests were performed by high-pressure gas adsorption analyzer (SETARAM, PCT Pro) at the pressure range of 20-40 bar, where the analysis of H 2 and CO 2 adsorption were measures at -196 o C and 25 o C, respectively. All samples were outgassed in a vacuum state at 200 o C for 2 h before analysis. Before analysis, all the samples were degassed at 200 o C for 2 h. In all cases, the hydrogen uptakes of all samples were excess hydrogen uptakes.

Supercapacitive measurements
The supercapacitor electrodes were prepared by mixing the as-prepared porous carbon, carbon black and poly (tetra uorethylene) binder in ethanol according to the weight ratio of 8: 1: 1. The obtained slurry was dispersed on nickel foam (current collector) as a 1 cm×1 cm sheet followed by being pressed together and dried at 105 o C for 4 h. The mass of active materials loaded on each electrode was about 3.5 mg. Three-electrode system was executed in aqueous solutions of 6 mol L -1 KOH. In the test process, a Hg/HgO electrode was used as the reference electrode and the Pt foil (1 cm 2 ) was employed as counter electrode. Galvanostatic charge-discharge was carried out on Chenhua electrochemical workstation (CHI660E, Shanghai, China). The speci c capacitance (C g ) was calculated from the discharge curves by the following equation: where the I is constant discharge current, △t is the discharge time, m is the active materials, △V is the voltage widow that reject the ohmic drop to the end of the discharge process.

Data and materials availability
All data needed to evaluate the conclusions in the paper are present in the paper or the supplementary materials.
Declarations electrochemical measurements, gas adsorption measurements. P.F.Y., F. X., and Y.R.L. co-wrote and revised the manuscript. All the authors contributed to the overall scienti c interpretation and edited the manuscript.

Competing interests
Authors declare no competing interests.