& Synthesis Design

Molecular-Based Design of Microporous Carbon Nanosheets

Lei He, Wen-Cui Li, Shuang Xu, and An-Hui Lu*[a]

Chem. Eur. J. 2019, 25, 3209 – 3218 T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3209

MinireviewDOI: 10.1002/chem.201804747

Abstract: Microporous carbons afford high surface areas,

large pore volumes, and good conductivity, and are fascinat-ing over a wide range of applications. Traditionally synthe-

sized microporous carbon materials usually suffer from somelimitations, such as poor accessibility and slow mass trans-port of molecules due to the micrometer-scale diffusionpathways and space confinement imposed by small pore

sizes. Two-dimensional microporous carbon materials, denot-ed as microporous carbon nanosheets (MCNs), possessnanoscale thickness, which allows fast mass and heat trans-

port along the z axis; thus overcoming the drawbacks of

their bulk counterparts. Herein, recent breakthroughs in the

synthetic strategies for MCNs are summarized. Three typicalmethods are discussed in detail with several examples: py-

rolysis of organic precursors with 2D units, a templatingmethod that uses wet chemistry, and the molten saltmethod. Among them, molecular-based assembly of MCNsin the liquid phase shows more controllable morphology,

thickness, and pore size distribution. Finally, challenges inthis research area are discussed to inspire future explora-tions.

1. Introduction

A microporous material is considered to have a major part of

its porosity in pores of less than 2 nm in width and to exhibitapparent surface areas usually higher than 200 m2 g@1; these

materials have been widely used as adsorbents, catalyst sup-ports, and electrode materials with a long history.[1] Traditional

microporous materials, such as activated carbon, are usually

synthesized by pyrolysis or activation of precursors, which arenormally composed of micrometer units with relatively broad

pore size distributions.[2] Accordingly, the key drawbacks forthese microporous materials include: 1) relatively slow mass

transport of molecules due to the long diffusion distanceamong micrometer units and space confinement imposed by

small pore sizes; 2) high pressure drop arising from the in-

creased resistance at high flow rate; and 3) poor separation se-lectivity because of the random pores without controlled ar-

rangement. Recent advances in novel 2D microporous materi-als, with nanoscale thickness, abundant porosity, and large

aspect ratios in the other two dimensions, are one of the inno-vative developments to overcome the abovementioned draw-

backs.[3] Their ultrahigh surface-to-volume ratio and thin geo-

metric structure greatly enlarge the accessible surfaces andshorten the diffusion pathways; thus allowing them to exhibitsuperior performances compared with those of their bulkcounterparts.

Typical 2D materials include graphene and boron nitride.Zeolites, metal–organic frameworks (MOFs), covalent organic

frameworks (COFs), microporous carbon nanosheets (MCNs)and various novel materials with single-atomic or single-poly-hedral thicknesses can also be classified as 2D materials, which

are often denoted as nanosheets, nanoplates, nanofilms, andso forth.[4] Among them, MCNs with nanoscale thickness have

shown great potential in a wide range of applications due totheir large surface areas; abundant porosity; and good chemi-

cal-, thermal-, and moisture-resistant properties.[5] Although

porous carbons have been widely used in many applicationswith a long history, MCNs are artificial 2D materials that have

been developed during the past decade. Several methodshave been developed for the synthesis of MCNs, including

chemical vapor deposition (CVD) of gaseous precursors,[6] tem-plating and self-assembly in the liquid phase,[7] and pyrolysis of

biomass[8] or other sheet-like carbon precursors.[9] Attempts

have been made to adjust the thickness, degree of graphitiza-tion, and porous structure by varying precursors, synthetic

conditions, and post-treatment processes. However, the prepa-ration of MCNs with uniformed micropores and a narrow pore

size distribution remains challenging because of the randomorientation and cross-linking of turbostratic carbon nanodo-

mains produced during pyrolysis.

Novel 2D materials, such as MOF nanosheets, have recentlygained particular attention as precursors to construct MCNs

through simple pyrolysis. As-prepared MCNs inherit periodicand well-defined framework structures with 2D morphology

and enriched microporous structure. However, only a fewtypes of MOFs have been reported to form MCNs due to the

aggregation of primary particles into crystallized large particles

rather than nanosheets. Comparably, the oriented fabricationof carbon-containing molecules into nanosheets in the liquid

phase is considered to be more controllable and a universalsynthesis of MCNs. Templates with sheet-like morphology areoften necessary for this method because it is easier to obtain aspherical morphology, instead of a sheet-like structure,through the self-assembly of carbon precursors to minimize

surface energy.[10] For example, a series of MCNs with a sand-wich structure have been prepared by means of graphene-di-

rected polymerization following carbonization.[11] The thicknessand porous structure can be easily tuned by changing the

carbon precursor and carbonization conditions. Another exam-ple is a soft-templating method assisted by a thermoregulated

phase-transition process, which has been reported to obtain

MCNs with uniform micropore distribution due to the orientedgrowth of carbon crystallites.[12] Furthermore, the successful

fabrication of MCNs in molten salt medium has also attractedmuch interest.[13] At high temperatures (>200 8C), the molten

salt creates an inorganic liquid phase, in which carbon precur-sors are assembled into nanosheets alongside carbonization.

[a] Dr. L. He, Prof. W.-C. Li, S. Xu, Prof. A.-H. LuState Key Laboratory of Fine ChemicalsSchool of Chemical Engineering, Dalian University of TechnologyDalian 116024, Liaoning (P.R. China)E-mail : anhuilu@dlut.edu.cn

The ORCID identification number(s) for the author(s) of this article can befound under : https ://doi.org/10.1002/chem.201804747.

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3210

Minireview

The molten salts not only play the role of reaction medium,but also act as templates. Interestingly, the surface areas of as-

prepared MCNs are also significantly improved compared withtheir counterparts without using salts ; thus indicating the po-

rogen effect of the salts.Herein, we introduce synthetic strategies for MCNs with rep-

resentative examples in the last five years, focusing on molecu-lar fabrication in the liquid phase through a templating

method under mild conditions or the molten salt method at

relatively high temperature. The mechanism, or proposedmechanism, of each method is discussed to shed light on the

principle of how to control the thickness, morphology, and mi-croporous structure of these materials. The challenges in this

research field are discussed in the last section to inspire futuredevelopments.

2. Pyrolysis of Organic Precursors with 2DUnits

The pyrolysis of carbon precursors with nanosheet-like units is

an easy and efficient way to obtain MCNs. A series of biomass

materials,[8] for example, cotton fibers,[14] eucalyptus leaves,[15]

and Perilla frutescens (PF),[16] are considered to be recyclable

and available as MCN precursors. The obtained MCNs generallyshow a high surface area of 600–800 m2 g@1, and can be doped

with various heteroatoms, including N, S, and O. Upon furtheractivation (e.g. , KOH), the surface area of the MCNs can reach

over 2000 m2 g@1 (Table 1).[17] However, the pore size distribu-

tion is usually random with a relatively broad range throughthis method. Recently, novel organic materials with nanosheet

units, for example, MOFs, have attracted increasing interest asprecursors to prepare MCNs by simple pyrolysis.[4a, 18] The struc-

tures of as-obtained MCNs are mainly determined by the pre-cursors, usually with periodic and well-defined frameworks.

Therefore, the molecular assembly of the precursors into a

sheet-like morphology is a key step for this method. For exam-ple, ZIF-8 is a typical MOF precursor to prepare porous car-

bons, which are generally synthesized with particle-like isotrop-ic morphologies both for monoclinic structures of low symme-

try and cubic sodalite topology (SOD) structures of high sym-metry.[18b] Jiang et al. developed an organic-solvent-free ap-proach to synthesize ZIF-8-derived MCNs.[18b] The surface areacould reach over 3000 m2 g@1 with a large amount of micro-

pores. They proposed that the nature of the solvent changedthe morphologies from 0D to 2D. Compared with methanol,water possesses stronger polarity and a more asymmetric

structure, which influences the capabilities of solvents to coor-dinate with metals and the configurations of functional

groups. To date, only a few types of MOFs have been reportedfor the synthesis of MCNs due to their complex preparation

processes and relatively low carbon yields. Additionally, it is dif-

ficult to obtain MCNs with small thicknesses (<30 nm) due tothe crystalline structure of MOFs.

Graphene, a typical carbon nanosheet with an ideal one-atom thickness, has been extensively explored during the last

decade. Graphene-based 2D solids, including porous graphenefilms and membranes, can be prepared by top-down or

bottom-up methods. Both methods are widely investigatedand have been introduced in several reviews.[19] However,

defect-free graphene-based 2D solids are limited in adsorptionapplications due to restricted gas permeability along the z axis

to gas molecules.[3, 20] Although the open-pore structures origi-nating from the stacking of graphene sheets can theoretically

Lei He received her Ph.D. degree from DalianInstitute of Chemical Physics, Chinese Acade-my of Sciences, in 2014. Since then, she hasbeen working at the State Key Laboratory ofFine Chemicals, School of Chemical Engineer-ing, Dalian University of Technology, as a lec-turer and a postdoctoral researcher. Her re-search interests now mainly focus on thedesign of 2D nanomaterials and their applica-tion in catalytic CO2 conversion processes.

Wen-Cui Li received her Ph.D. degree from theSchool of Chemical Engineering, Dalian Uni-versity of Technology (DUT), before her post-doctoral training at Universit-t Werzburg andMax-Planck-Institut fer Kohlenforschung. Shehas been working at DUT as a professor since2006. Her research interests include the designof nanostructured materials and their applica-tions in catalysis, energy storage, and energyconversion.

Shuang Xu received her Master’s degree fromHeilongjiang University in 2015. She is cur-rently a Ph.D. student in Prof. An-Hui Lu’sgroup. Her research focuses on the synthesisand application of 2D nanomaterials withcontrolled morphology.

An-Hui Lu received his Ph.D. degree from theInstitute of Coal Chemistry, Chinese Academyof Sciences, in 2001. He is currently AssociateDirector of the State Key Laboratory of FineChemicals, and Dean of the School of Chemi-cal Engineering, Dalian University of Technol-ogy. His research interests include the synthe-sis of porous materials for heterogeneous cat-alysis, adsorption, energy storage, and energyconversion.

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3211

Minireview

Table 1. Examples of MCNs discussed herein.[a]

Carbon source Synthesis (brief description) Morphology SBET

[m2 g@1]pore volume[cm3 g@1]

Pore size[b]

[nm]Doping Ref.

cotton fiber cotton microfibers are presoaked with fumingsulfuric acid, mechanically ground, and an-nealed at 800–850 8C

N-doped carbon nanosheetswith abundant micro- andmesopores

324–912

Vtotal = 237.8–556.0

0.8–10 6–8.9 wt %N, 7–10.1 at % O

[14]

Eucalyptus leaves leaves are dried at 120 8C; pulverized; mixedwith KHCO3 in a mass ratio of 1:4; and carbon-ized at 750, 850, or 950 8C for 5 h under Ar

porous carbon nanosheets 2133 Vtotal = 0.71 <3 1.71 at %N, 7.7 at %O

[15]

PF PF leaves are crushed and sieved into powder,with particle sizes of &200 mm, and pyrolyzedat 600–800 8C under N2 for 2 h

wrinkled nanosheets with arough surface, thickness:&120 nm; flat nanosheets,thickness: 30–40 nm

480–712

Vmicro = 0.20–0.25Vtotal = 0.44

1.2 18.76 at %N, 1.7 at %O

[16]

biosources (gelatinand dopamine)

gelatin solution added to dispersion of mont-morillonite template, intercalation reaction at70 8C for 24 h, followed by further treatmentat 600 8C for 1 h under N2, activated by KOH at800 8C

wrinkled carbon nanosheetassemblies

2774–3106

Vmicro = 0.46–0.90Vtotal = 1.48–2.35

<2 0.3–0.6 at% N,4–13.2 at%O

[17b]

glucose spheres direct pyrolysis of a mixture of KOH and pGSsat 800 8C

oriented and interlinkedcarbon nanosheets

2633 Vtotal = 1.86 0.6–0.8, 1.1–1.4, and 10–100

– [17a]

ZIF-8 ZIF-8 nanosheets pyrolyzed at 920 8C for 2 hunder Ar, followed by KOH activation at 700 8C

aggregated nanosheets,thickness &100 nm

3052 Vtotal = 2.35 <2, 2.4, 2.8 1.251–3.417 wt %N

[18b]

Zn-TDPAT Zn-TDPAT powder is heated at different tem-peratures (800, 900, and 1000 8C) at 5 8C min@1

for 5 h under Ar, and then washed with 1 mHCl

flowerlike MCNs 746–1432.6

Vmicro = 0.26–0.56

<0.6 16.4, 6.71,and 2.92 %N

[18c]

asparagine, resorci-nol, formaldehyde,GO

GO as a shape-directing agent, polymerizationat 90 8C, pyrolysis at 800 8C for 2 h under N2

thickness: 9.9, 17, 71, and82 nm

350–1086

Vtotal = 0.31–0.64Vmicro = 0.12–0.26

0.6 &4 % O,&1.2 % N

[11a]

melamine, aromaticdialdehydes, AGO

modified AGO directed condensation at 180 8Cfor 72 h under an inert atmosphere, pyrolysisat 700–900 8C

thickness: 60, 32, and 4 nm 323–762

Vtotal = 0.653–1.011

0.63 up to12.3 % N

[11b]

GO, PDA carbonization of GO-PDA-S hybrids under N2

at 400 8C for 2 h, followed by further treat-ment at 800 8C for 3 h

mesoporous MCNs, thick-ness : &3.0 nm, 1.2 nm

273 Vtotal = 0.33 <2, 3.6 4.1 % Nand 6.1 %S after py-rolysis

[11c]

resorcinol, mela-mine, formalde-hyde

hydrothermally treated at 120 8C for 24 h,mixed with KOH, annealed at a rate of5 8C min@1 to 800 8C, and kept for 2 h under N2

flow

spherical structure com-posed of microporous nano-sheets

580.4–2118

Vmicro = 0.112–0.23

0.8, 1.0, and2.3

1.585 wt %N

[10c]

resorcinol, formal-dehyde, propyl-amine, SA

thermoregulated phase transition methodwith SA as a template, polymerization and car-bonization at certain temperatures

nanosheets with orientatedgrowth of sp2 carbon, thick-ness: 65, 54, 43 to 30 nm

463 Vmicro = 0.23Vtotal = 0.26

0.58 (singlesize distribu-tion of mi-cropores)

&6 wt % O [12]

glucose monohy-drate

glucose monohydrate mixed with activatingoxysalts and LiCl/KCl (45/55 mass ratio) in amass ratio of 1/1/10, flushed with N2 for100 min, and heated at 5 K min@1 to the reac-tion temperature and left for 5 h

thickness: &0.5–10 nm 64–1912

Vmicro = 0.02–0.705,Vtotal = 0.314–0.930

<2 and 10–100

– [13b]

d-glucose monohy-drate

d-glucose monohydrate is mixed with LiCl/KCl,ground, and heated to the carbonization tem-perature at 5 8C min@1 for 5 h under N2 ; thencrushed into particles and washed with waterto obtain porous carbon

thickness: 0.65 nm, 1.35 nm,and thicker sheets from&1.5 up to 5 nm

595 – – – [30]

pyromellitic di-anhydride, benz-idine

pyromellitic dianhydride, benzidine and metalchloride salts (solvent: KCl/ZnCl2 51/49 bymolar ratio) in a 1:9 weight ratio of reactant/solvent are mixed and ball-milled for 5 min;the mixture was heated to 325–1000 8C at3.33 8C min@1 for 1 h under N2

graphene-like structure witha crumpled sheet

2607 Vtotal = 3.12 <2, 3.5 5.9 at % N [31]

melamine, tere-phthalaldehyde

melamine and terephthalaldehyde are groundwith 9.0 g LiCl/KCl (45/55 by weight); the mix-ture is heated to 150 8C for 4 h, then to 400 8Cfor 2 h, and 600 8C for another 2 h under N2

sheet-like nanostructure 332–696

Vmicro = 0.14–0.31Vtotal = 0.21–0.39

– 21.34–30.5 wt %N, 9.83–12.78 wt %O

[32]

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3212

Minireview

deliver fast sorption kinetics, the adsorption potential is too

weak for effective trapping of gas molecules, such as CO2,

which consequently leads to a low adsorption capacity. Post-treatments are necessary to create pores and further increase

the porosity in carbon nanosheets through activation[21] andcarbothermal reactions.[22] Nevertheless, it is still difficult to pre-

cisely control the pore sizes and distribution during the post-treatment process.

3. Templating Methods

In the liquid phase, the self-assembly of carbon-containingmolecules into a 2D morphology is often difficult due to the

high surface energy compared with that of a spherical mor-phology. Templates or shape-directing agents are often neces-sary for obtaining a sheet-like morphology through wet

chemistry under mild conditions. Typically, templating methodsinclude the following steps: the preparation of templates, thefabrication of templates and carbon precursors, pyrolysis orcarbonization, and the removal of templates if necessary.

Therefore, the physical and chemical structures of MCNs are in-fluenced by the original structure of the templates, molecular

recognition between the carbon precursor and template, andpost-treatment methods. Templating is considered to be themost promising method to precisely control the thickness,

pore size distribution, and uniformity of the final MCNs. In thissection, two carbon-containing templates are introduced as

representatives for this method.

3.1. Graphene-based templates

Atomic-thickness GO obtained by the Hummer method usually

possesses a charged surface and is suitable for use as a shape-directing agent to prepare MCNs. During the synthesis, poly-

merization of carbon precursors occurs on the surfaces of GOto form polymer nanosheets. The composition, thickness, and

porous structure of the final MCNs are tailorable by changing

the precursors or controlling the polymerization conditions.

For example, the poly(benzoxazine-co-resol) system is em-ployed as a carbon precursor, and the amino acid asparagine is

used as the bridging molecule that interacts with GO to attractpolymer molecules onto the GO surfaces to form a sandwich

structure (Figure 1).[11a, 23] Notably, the graphene content usedin all syntheses was in the range of 0.5–2.6 wt %, which was

significantly lower than that of the content of common surfac-

tants. This indicates an efficient shape-directing function ofGO. The synthesized MCNs possess precisely tunable thickness-

es from 20 to 200 nm, according to a fitted linear correlationbetween the carbon precursor/GO mass ratio and the coating

thickness. Abundant micropores contribute to a large capacityof gas adsorption.

Table 1. (Continued)

Carbon source Synthesis (brief description) Morphology SBET

[m2 g@1]pore volume[cm3 g@1]

Pore size[b]

[nm]Doping Ref.

melamine, tere-phthalaldehyde

0.2 g carbon precursors is mixed with 20 gLiCl/KCl (1/1 molar ratio) and 0.2 g KOH byball milling for 6 h; heated to 150 8C for 4 h,then to 400 8C for 2 h, and to activation tem-perature (500, 600, or 700 8C) for 2 h

MCNs 170–1415

Vtotal = 0.2–1.4 <0.8, 1.8,2–4

3.47–5.89 wt %N

[33]

aromatic nitrilemonomers (e.g. ,1,4-dicyanoben-zene)

1,4-dicyanobenzene is mixed with ZnCl2 (1/100 molar ratio), transferred into an ampoulein a glove box, sealed under vacuum, andheated at 400 8C for 40 h

thickness: 3–20 nm 537 Vtotal = 0.45 0.55 <1.5 wt %N

[34]

benzidine, 3,3’,4,4’-benzophenone tet-racarboxylic di-anhydride

polyimide is fabricated in a typical two-stepmethod, including pre-polymerization and sol-vothermal polymerization; the polymer is fur-ther pyrolyzed to obtain microporous carbonconstructed by 2D nanosheets

3D architectures constructedby 2D nanosheets with di-ameter of 1.5–2.5 mm andthickness of 20 @25 nm

983–1375

Vtotal = 0.533–1.015

1.6, 1.9, and3.9

2.95–3.13 at % N

[35]

[a] pGS = polymerized glucose sphere, TDPAT = 2,4,6-tris(3,5-dicarboxylphenylamino)1,3,5-triazine, GO = graphene oxide, AGO = aminated graphene oxide,SA = stearic acid, PDA = polydopamine. [b] Average pore sizes or the range of pore size distributions mentioned in the literature are listed here for compar-ison.

Figure 1. Schematic illustration of the formation of MCNs. a) The negativelycharged GO sheet. b) Positively charged amino acids are equally dispersedon both surfaces of GO. c) In situ copolymerization of preadsorbed aspara-gine, resorcinol, and formaldehyde. d) The polymer is transformed into a mi-croporous carbon layer during pyrolysis in Ar.[11a]

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3213

Minireview

As an extension, MCNs with hybrid N atoms were successful-ly prepared by using nitrile-containing ionic-liquid-functional-

ized graphene oxide (IL-functionalized GO) sheets as a shape-directing agent.[24] Similar to the abovementioned example,

the resulting MCNs possess a nanosheet structure associatedwith graphene as an inner layer and microporous carbon coat-

ings on both sides. The function of the ionic liquid used herewas to stabilize the GO sheets and functionalize them in an

aqueous solution. By simply changing the ratio of reactants to

GO content, the thickness of the microporous carbon layer canbe precisely controlled over a wide range below 100 nm. As-

obtained MCNs exhibited excellent performance if used as theelectrode material for supercapacitors due to their abundant

micropores with a narrow size distribution, short diffusionpaths, highly electrically conductive networks, and good wetta-bility.

The doping of hybrid atoms is an attractive research topic tomanipulate the electronic properties of MCNs, and can be real-ized through this template method by changing the carbonprecursors. Schiff base type 2D porous polymer (TPP) nano-

sheets were successfully prepared by using graphene as ashape-directing agent. The sheets were further carbonized to

obtain nitrogen-enriched 2D porous carbon (TPC) nanosheets

(Figure 2).[11b] The precursors include melamine and aromaticdialdehydes, such as 1,3-phthalaldehyde and 1,4-phthalalde-

hyde, which result in a high nitrogen doping content in thefinal MCNs. Qiao et al. also reported the synthesis of MCNs by

using GO as a template with N and S codopants.[11c] Modifica-tion of GO surfaces is a key step for this graphene-directed

method to enhance the interactions between GO and carbon

precursor, which determines the nanosheet morphology forthe final MCNs.

Benefitting from the short diffusion paths and high micro-porosity, these MCNs exhibited high sorption capacity, excel-

lent selectivity for guest molecules, and fast sorption kinet-ics.[25] However, perforation along the z axis is still restricted

due to the impermeability of a defect-free graphene intermedi-ate layer for this type of MCN, which should be further im-proved for application in the field of gas separation. Mean-

while, the existence of a graphene network shows good con-ductivity, which is beneficial for electrochemistry applications.

3.2. Soft templates

Soft templating is a common and well established method forthe synthesis of porous nanomaterials. However, the sheet-like

morphology is difficult to prepare in aqueous solution, be-

cause the soft template has good mobility and is tends to bespherical to minimize its surface energy. Recently, we devel-

oped a novel method to synthesize free-standing carbon nano-plates (FCPs) with uniformly sized ultra-micropores (5–6 a) and

controlled thickness in the range of 30–65 nm (Figure 3).[12] Inprinciple, these FCPs are prepared through a molecular tem-plating polymerization with the assistance of thermoregulated

phase transition process. Hydrophobic SA, with a meltingpoint of 66 8C, was first liquefied at 80 8C and then dispersedinto an aqueous solution of F127 surfactant to form a uniformmicroemulsion. Subsequently, the as-obtained microemulsionwas cooled to 28 8C, which caused SA to transform from aliquid to a solid following Ostwald ripening. During this pro-

Figure 2. a) Schematic illustration of the formation of 2D porous polymersand heteroatom-doped MCNs based on Schiff base chemistry. b) and c) TEMimages of TPC at different magnifications. d) Nitrogen adsorption–desorp-tion isotherms of TPC at 77.3 K (inset shows the pore size distribution). e) Ni-trogen content derived from X-ray photoelectron spectroscopy (XPS) analy-sis for TPP and TPC prepared at different heating temperatures. f) N 1 s core-level spectra for TPP and TPC prepared at different heating temperatures.[11b]

Figure 3. a) Scheme of the synthesis for the FCPs. b) SEM image of FCPs.c) Relationship between mass ratio and thickness of FCPs. d) TEM ande) HRTEM images of FCPs. f) N2 adsorption isotherms, and g) Pore size distri-bution of FCPs.[12]

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3214

Minireview

cess, a suspension of SA with an ultrathin nanosheet structurewas successfully obtained. Compared with the sample without

the addition of SA, a spherical morphology was obtained in-stead of nanosheets. This proves that the linear asymmetric ge-

ometry of SA determines the formation of anisotropic sheetcolloids, as opposed to conventional spherical structures. After-

wards, the abundant carboxylic acid groups on these SA nano-sheets induced interface assembly and subsequent polymeri-

zation of resorcinol/formaldehyde/propylamine through hydro-

gen bonding. After pyrolysis of the polymer, MCNs with en-riched, uniformly distributed micropores were obtained.

The unique porous structure of these FCPs originated fromtheir 2D morphology and played an important role in acceler-

ating selective gas separation. The diffusion path is extremelyalong on the z axis, which leads to very fast release of volatilespecies formed during pyrolysis and guarantees the uniformity

of the porous structure. More interestingly, the thin, sheet-likemorphology facilitates the orientated growth of carbon crystal-

lites instead of cross-linking between neighboring carbon crys-tallites. The FCPs consist of about 80 % sp2-carbon with a low

content of oxygen-containing functional groups and sp3-carbon, which consequently weakens the cross-linking of

carbon crystallites. The structure thus tends to be more com-

pact and neighboring crystallites have a strong tendency tostack into an almost parallel orientation.[26] Meanwhile, molecu-

lar-level micropores are uniformly distributed over the entireFCPs, which are provided by vertically aligned carbon crystal-

lites. On one hand, the ultrathin structure effectively shortensthe diffusion paths and facilitates the transfer of gas molecules

into the inner micropores. On the other hand, concentrated

and permeable ultra-micropores enhance host–guest van derWaals interactions and enhance the adsorption capacity and

selectivity. As a result, the FCPs exhibited excellent selectivityand adsorption kinetics for gas separation in both static and

dynamic adsorption circumstances. This method could be ex-tended by changing the surfactants or carbon precursors to

further manipulate the porous structure and meet the de-

mands for different applications.In addition to the two abovementioned examples, other

templates have been reported for the preparation of carbonnanosheets. For example, by using MgO sheets as templates,

Fan et al. prepared mesoporous carbon nanosheets.[27] Zhaoet al. developed a solution deposition method to synthesize

2D ordered mesoporous graphene/carbon nanosheets over dif-ferent substrates.[28] Although these studies are irrelevant tomicroporous structures, the mechanism may provide inspira-

tion for the future design of MCNs.

4. Molten Salt Method

The molten salt method is an important complementary route

to conventional liquid-phase synthesis, which can be carriedout at higher temperatures (>200 8C) than that with traditional

solvents and yield a wide range of nanostructured materials.[29]

Recently, inorganic salts with melting points of 300–800 8C,

such as metal halides, nitrates, and sulfates, have attractednew interest for the synthesis of materials with novel struc-

tures. Typically, the carbon precursors were initially mixed withsalt in the solid phase. Afterwards, the mixture was heated toa certain temperature under an inert atmosphere. During thisprocess, the salt became liquid to create a homogeneousmedium for carbonization. After cooling and removing the saltby simply washing, MCNs can be obtained.[13] Depending onthe nature of the salt, the carbonization temperature rangesfrom 100 to 1000 8C, which consequently diversifies the struc-tures of the produced MCNs. It is a one-step method of creat-

ing 2D structures with a large amount of micropores.Several types of MCNs have been synthesized, from simple

porous frameworks to complex heteroatom (nitrogen or sulfur)doped materials. For example, the group of Antonietti devel-oped glucose-derived carbon nanosheets by using a mixtureof LiCl/KCl as a molten salt medium (Figure 4).[13b, 30] The car-

bonization of glucose proceeds gradually through the elimina-

tion of water and condensation, which results in the conver-sion of sp3-C@X bonds (X = C, H, or O) of glucose into sp2-carbon. Interestingly, at a low concentration of glucose, few-layered graphene was synthesized through this method, which

suggested that a solution growth mechanism was involved. Inother words, the nanosheet morphology is formed only if thecarbon-containing intermediates dissolve in liquefied themolten salt medium. Meanwhile, they also investigated the in-fluence of adding different dissolved oxysalts, and found that

carbonate, nitrate, sulfate, and chlorate were efficient in en-hancing the formation of carbon nanosheets with improved

porosity. Upon investigating the mechanism of the formationof the 2D morphology, it was speculated that sp2-C@C bondingenergetically favored the 2D assembly of carbon intermediates

mediated by the ionic environment.[13b] The growth of 2Dcarbon nanosheets then started by organizing these primitive

structures into large nanosheets. In addition, the authors alsosuggested that the salts assisted the creation of a large

Figure 4. Graphene prepared from glucose by means of salt melt synthesis(SMS). a) Schematic illustration of the conversion of glucose into porouscarbon (or graphene) at a high (or low) glucose concentration (Cg) in LiCl/KCl melt. b) TEM and c) AFM images for the graphene synthesized from glu-cose at low Cg (glucose/salt = 1/100).[30]

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3215

Minireview

amount of micropores. It is similar to a chemical activation pro-cess in which weakly bonded carbon atoms are oxidized by

the activation agents.The group of Dai prepared very thin MCNs by using KCl/

ZnCl2 as the mixed salt.[31] The salt mixture possesses a relative-ly low melting temperature (&200 8C) and a high decomposi-

tion temperature (>1000 8C), which provides a wide range oftemperatures for fabricating porous carbon materials. By usingdifferent monomers, the porous structure of the carbon sheets

can be adjusted with well-dispersed and through-plane nano-holes (diameter : 5–10 nm). Furthermore, the authors suggest-

ed that inorganic salts played the role of “molten salt scaffolds”for condensing and rearranging 1D aromatic polymers to 2D

nanosheets. The as-prepared MCNs were used as electrode ma-terials in supercapacitors, and showed a specific capacity of

131.4–275.5 F g@1 at a scan rate of 2000 mV s@1. Additionally, in

combination with an active transition-metal oxide (Co3O4), theMCNs also exhibited good performance in the aerobic oxida-

tion of alcohols and amines.Our group also synthesized a series of MCNs by using mela-

mine and terephthalaldehyde as carbon precursors throughthe Schiff base reaction in a molten salt medium (Figure 5).[32]

The as-obtained MCN possesses a large amount of nitrogen

(30.51 wt %) and high surface area with enriched porosity. No-tably, the specific surface area for all MCNs derived from the

molten salt system was at least a tenfold increase comparedwith that of the counterpart prepared without salt, which fur-

ther indicated that the molten salt acted as a porogen as wellas a solvent. As-obtained MCNs were used as anode materials

for lithium-ion batteries, which exhibited an initial coulombic

efficiency of about 63.1 %, and a constant reversible capacityof 605 mA h g@1 at a current density of 100 mA g@1 after

100 cycles. Moreover, they showed a high-rate capability,which was tenfold that of the capacity of commercial graphite.

The excellent electrochemical performance is due to theunique 2D structure, with high porosity and a high nitrogen

content of about 30 wt %. By adding KOH to the carbon pre-cursors, Hao et al. prepared MCNs with improved microporosi-

ty and enlarged surface area through the molten salt methodcombined with chemical activation.[33] As-obtained MCNs as

electrode materials in supercapacitors showed a maximum ca-pacitance of 243 F g@1 at a current density of 0.5 Ag@1 in aque-

ous electrolyte, 179 F g@1 rate capability at 20 Ag@1, and goodcycling stability. The group of Chen synthesized a type of 2Dmicroporous carbonaceous polymer nanosheet through poly-

merization of aromatic nitrile monomers in molten ZnCl2 at arelatively low temperature (400 8C).[34] The resulting micropo-rous nanosheets possessed a thickness of 3–20 nm and well-defined microporosity. These carbonaceous polymer nano-

sheets exhibited excellent CO2 adsorption capacity (8.14 wt %at 298 K and 1 bar) and 25.6 CO2/N2 selectivity, which might

find potential applications as adsorbents for gas separation

and other applications.The molten salt method is an efficient and facile one-step

strategy for preparing MCNs with high microporosity at rela-tively low temperature. However, the mechanism is still under

debate and requires in situ observation. The formation of a 2Dmorphology is related to the ionic environment of the molten

salt. Such a salt mixture may play the role of a scaffold during

the condensation reactions and concomitant molecular assem-bly of 2D nanosheets.

The surface area, pore volume, and range of pore sizes ofthe abovementioned MCNs are summarized in Table 1 for easy

comparison.

5. Summary and Outlook

We intended to provide an overview of the most recent devel-

opments of novel synthetic methods for MCNs, by highlightingseveral examples based on molecular assembly. The advantag-

es and limitations of each method are discussed, along with an

introduction of the principles of tuning the thickness and mi-croporous structure of the final MCNs. The applications are

also briefly introduced after each method, including adsorp-tion, catalysis, and electrochemical storage. Among the pro-

posed strategies, the templating method is considered to beone of the most promising methods for creating controllablemicroporous structures with a very narrow pore size distribu-tion. In addition, other methods, such as exfoliation and micro-etching, are also reported for the synthesis of porous graphene

nanosheets or mesoporous carbon nanosheets, which havebeen introduced in many other excellent reviews and not dis-cussed herein in detail.[36] We hope that this article will provideuseful information for the design of future MCNs with unique

structures to meet the demands of practical applications.Although these emerging methods are promising for obtain-

ing MCNs, several challenges still need to be overcome before

their usage in practical applications. One challenge is to obtainMCNs without conglutination during high-temperature anneal-

ing. Due to the high surface energy of 2D-structured carbonprecursors, the tendency of incidental condensation and sinter-

ing often result in nondispersible and conglutinated bulky ma-terials. Consequently, this creates a complex network of porous

Figure 5. a) Schematic illustration of the preparation strategy for carbonnanosheets. b) SEM image of the prepared MCNs. c) Raman spectra of theprepared MCNs.

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3216

Minireview

structures with disordered micro-, meso-, or macropores. More-over, the restacking of nanosheets often leads to a decrease in

the exposed surfaces and micropores. One example of solvingthis problem is the fabrication of monolithic porous carbon

constructed from nanosheet units by means of graphene-di-rected preparation.[11a] The connection between the edges ofthe nanosheets provides excellent strength and creates macro-pores for diffusion without restacking. In this way, the shapingprocess can be simplified without the necessity to add binders.

Additionally, this facile method can be used to make porouscarbons with a variety of shapes on the centimeter scale,

which can be directly equipped as adsorbents, catalyst sup-ports, and in many other applications. By combining the novelsynthetic strategy with post-treatment, such as activation orfunctionalization, more types of MCNs can be designed with

tunable porosity and surface properties.Another challenge lies in the investigation of the fabrication

mechanism of MCNs. By using in situ characterization methods,it is possible to monitor changes to the skeleton structure;thus recognizing the driving force of the molecular assembly

or revealing the growth regularity of the primary units into a2D morphology. Nevertheless, these methods can rarely work

at high temperatures, for instance, pyrolysis or molten salt con-

ditions. In this case, theoretical calculations and computer-as-sisted simulations are powerful tools to investigate what exact-

ly happens at the molecular level. These studies may guide thedesign of novel MCNs to allow better control of a uniform mi-

croporous structure.

Acknowledgements

This work was supported by the National Natural ScienceFoundation of China (21473021, 21506022), Cheung Kong

Scholars Program of China (T2015036), National Science Foun-dation for Post-doctoral Scientists of China (2015M581336).

Conflict of interest

The authors declare no conflict of interest.

Keywords: carbon · microporous materials · nanostructures ·self-assembly · template synthesis

[1] D. H. Everett, Manual of Symbols and Terminology for PhysicochemicalQuantities and Units, Appendix II : Definitions, Terminology and Symbols inColloid and Surface Chemistry, Vol. 31 (Ed. : D. H. Everett), IUPAC, Wash-ington DC, 1972, p. 577.

[2] a) C. Liang, Z. Li, S. Dai, Angew. Chem. Int. Ed. 2008, 47, 3696 – 3717;Angew. Chem. 2008, 120, 3754 – 3776; b) A.-H. Lu, G.-P. Hao, X.-Q. Zhangin Porous Carbons for Carbon Dioxide Capture (Eds. : A.-H. Lu, S. Dai),Springer, Heidelberg, 2014, pp. 15 – 77.

[3] M. Lozada-Hidalgo, S. Hu, O. Marshall, A. Mishchenko, A. N. Grigorenko,R. A. W. Dryfe, B. Radha, I. V. Grigorieva, A. K. Geim, Science 2016, 351,68 – 70.

[4] a) X. Kuang, Z. Wang, X. Sun, Y. Zhang, Q. Wei, Chem. Commun. 2018,54, 264 – 267; b) G. Li, K. Zhang, T. Tsuru, ACS Appl. Mater. Interfaces2017, 9, 8433 – 8436; c) Q. Song, S. Jiang, T. Hasell, M. Liu, S. Sun, A. K.Cheetham, E. Sivaniah, A. I. Cooper, Adv. Mater. 2016, 28, 2629 – 2637.

[5] a) H. L. Fan, W. Z. Shen, ChemSusChem 2015, 8, 2004 – 2027; b) H. E. Lei,X.-Q. Zhang, A.-H. Lu, Acta Phys. Chim. Sin. 2017, 33, 709 – 728.

[6] a) J. Wang, M. Zhu, R. A. Outlaw, X. Zhao, D. M. Manos, B. C. Holloway,Carbon 2004, 42, 2867 – 2872; b) B. Xu, D. Zheng, M. Jia, H. Liu, G. Cao,N. Qiao, Y. Wei, Y. Yang, Mater. Lett. 2015, 143, 159 – 162.

[7] a) M. Inagaki, M. Toyoda, Y. Soneda, S. Tsujimura, T. Morishita, Carbon2016, 107, 448 – 473; b) V. Malgras, Q. M. Ji, Y. Kamachi, T. Mori, F. K.Shieh, K. C. W. Wu, K. Ariga, Y. Yamauchi, Bull. Chem. Soc. Japan 2015,88, 1171 – 1200; c) H. Nishihara, T. Kyotani, Adv. Mater. 2012, 24, 4473 –4498.

[8] J. Deng, M. M. Li, Y. Wang, Green Chem. 2016, 18, 4824 – 4854.[9] a) X. Yan, Y. Yang, X. Hu, M. Zhou, S. Komarneni, Microporous Mesopo-

rous Mater. 2016, 234, 162 – 165; b) J. K. Sun, Q. Xu, Energy Environ. Sci.2014, 7, 2071 – 2100.

[10] a) A. H. Lu, T. Sun, W. C. Li, Q. Sun, F. Han, D. H. Liu, Y. Guo, Angew.Chem. Int. Ed. 2011, 50, 11765 – 11768; Angew. Chem. 2011, 123, 11969 –11972; b) S. Wang, W. C. Li, G. P. Hao, Y. Hao, Q. Sun, X. Q. Zhang, A. H.Lu, J. Am. Chem. Soc. 2011, 133, 15304 – 15307; c) S. Wang, J. Zhang, P.Shang, Y. Li, Z. Chen, Q. Xu, Chem. Commun. 2014, 50, 12091 – 12094.

[11] a) G.-P. Hao, Z.-Y. Jin, Q. Sun, X.-Q. Zhang, J.-T. Zhang, A.-H. Lu, EnergyEnviron. Sci. 2013, 6, 3740 – 3747; b) X. Zhuang, F. Zhang, D. Wu, X.Feng, Adv. Mater. 2014, 26, 3081 – 3086; c) K. Qu, Y. Zheng, S. Dai, S. Z.Qiao, Nano Energy 2016, 19, 373 – 381.

[12] L. H. Zhang, W. C. Li, H. Liu, Q. G. Wang, L. Tang, Q. T. Hu, W. J. Xu, W. H.Qiao, Z. Y. Lu, A. H. Lu, Angew. Chem. Int. Ed. 2018, 57, 1632 – 1635;Angew. Chem. 2018, 130, 1648 – 1651.

[13] a) X. Liu, N. Fechler, M. Antonietti, Chem. Soc. Rev. 2013, 42, 8237 – 8265;b) X. Liu, M. Antonietti, Carbon 2014, 69, 460 – 466; c) X. Deng, B. Zhao,L. Zhu, Z. Shao, Carbon 2015, 93, 48 – 58.

[14] X. Lin, X. Wang, L. Li, M. Yan, Y. Tian, ACS Sustainable Chem. Eng. 2017,5, 9709 – 9717.

[15] A. K. Mondal, K. Kretschmer, Y. Zhao, H. Liu, C. Wang, B. Sun, G. Wang,Chem. Eur. J. 2017, 23, 3683 – 3690.

[16] B. Liu, Y. Liu, H. Chen, M. Yang, H. Li, J. Power Sources 2017, 341, 309 –317.

[17] a) X. Zheng, W. Lv, Y. Tao, J. Shao, C. Zhang, D. Liu, J. Luo, D.-W. Wang,Q.-H. Yang, Chem. Mater. 2014, 26, 6896 – 6903; b) X. Fan, C. Yu, J. Yang,Z. Ling, C. Hu, M. Zhang, J. Qiu, Adv. Energy Mater. 2015, 5, 1401761.

[18] a) J. Cong, H. Xu, M. Lu, Y. Wu, Y. Li, P. He, J. Gao, J. Yao, S. Xu, Chem.Asian J. 2018, 13, 1485 – 1491; b) Y. Jiang, H. Liu, X. Tan, L. Guo, J.Zhang, S. Liu, Y. Guo, J. Zhang, H. Wang, W. Chu, ACS Appl. Mater. Inter-faces 2017, 9, 25239 – 25249; c) X.-J. Hong, X.-Y. Tang, Q. Wei, C.-L. Song,S.-Y. Wang, R.-F. Dong, Y.-P. Cai, L.-P. Si, ACS Appl. Mater. Interfaces 2018,10, 9435 – 9443.

[19] a) F. Bonaccorso, L. Colombo, G. Yu, M. Stoller, V. Tozzini, A. C. Ferrari,R. S. Ruoff, V. Pellegrini, Science 2015, 347, 1246501; b) N. Song, X. L.Gao, Z. Ma, X. J. Wang, Y. Wei, C. J. Gao, Desalination 2018, 437, 59 – 72.

[20] H. W. Kim, H. W. Yoon, S.-M. Yoon, B. M. Yoo, B. K. Ahn, Y. H. Cho, H. J.Shin, H. Yang, U. Paik, S. Kwon, J.-Y. Choi, H. B. Park, Science 2013, 342,91 – 95.

[21] Y. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. Cai, P. J. Ferreira, A. Pirkle,R. M. Wallace, K. A. Cychosz, M. Thommes, D. Su, E. A. Stach, R. S. Ruoff,Science 2011, 332, 1537 – 1541.

[22] D. Zhou, Y. Cui, P.-W. Xiao, M.-Y. Jiang, B.-H. Han, Nat. Commun. 2014, 5,4716.

[23] G.-P. Hao, A.-H. Lu, W. Dong, Z.-Y. Jin, X.-Q. Zhang, J.-T. Zhang, W.-C. Li,Adv. Energy Mater. 2013, 3, 1421 – 1427.

[24] a) D.-C. Guo, J. Mi, G.-P. Hao, W. Dong, G. Xiong, W.-C. Li, A.-H. Lu,Energy Environ. Sci. 2013, 6, 652 – 659; b) Z.-Y. Jin, A.-H. Lu, Y.-Y. Xu, J.-T.Zhang, W.-C. Li, Adv. Mater. 2014, 26, 3700 – 3705.

[25] Z. Y. Jin, Y. Y. Xu, Q. Sun, A. H. Lu, Small 2015, 11, 5151 – 5156.[26] R. E. Franklin, Proc. Royal Soc. London A. 1951, 209, 196 – 218.[27] Z. Fan, Y. Liu, J. Yan, G. Ning, Q. Wang, T. Wei, L. Zhi, F. Wei, Adv. Energy

Mater. 2012, 2, 419 – 424.[28] a) Y. Fang, Y. Lv, J. Tang, H. Wu, D. Jia, D. Feng, B. Kong, Y. Wang, A. A. El-

zatahry, D. Al-Dahyan, Q. Zhang, G. Zheng, D. Zhao, Angew. Chem. Int.Ed. 2015, 54, 8425 – 8429; Angew. Chem. 2015, 127, 8545 – 8549; b) Y.Fang, Y. Lv, R. Che, H. Wu, X. Zhang, D. Gu, G. Zheng, D. Zhao, J. Am.Chem. Soc. 2013, 135, 1524 – 1530.

[29] S. V. Volkov, Chem. Soc. Rev. 1990, 19, 21 – 28.[30] X. Liu, C. Giordano, M. Antonietti, Small 2014, 10, 193 – 200.

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3217

Minireview

[31] P. Zhang, Z.-A. Qiao, Z. Zhang, S. Wan, S. Dai, J. Mater. Chem. A 2014, 2,12262 – 12269.

[32] B. He, W.-C. Li, A.-H. Lu, J. Mater. Chem. A 2015, 3, 579 – 585.[33] X. Hao, J. Wang, B. Ding, L. Shen, Y. Xu, Y. Wang, Z. Chang, H. Dou, X.

Lu, X. Zhang, Chem. Eur. J. 2016, 22, 16668 – 16674.[34] M. Zhang, L. Liu, T. He, G. T. Wu, P. Chen, Chem. Asian J. 2016, 11, 1849 –

1855.[35] Z. Xu, X. Zhuang, C. Yang, J. Cao, Z. Yao, Y. Tang, J. Jiang, D. Wu, X.

Feng, Adv. Mater. 2016, 28, 1981 – 1987.

[36] W. Li, J. Liu, D. Zhao, Nat. Rev. Mater. 2016, 1, 16023; M. J. Allen, V. C.Tung, R. B. Kaner, Chem. Rev. 2010, 110, 132 – 145; B. Mendoza-S#nchez,Y. Gogotsi, Adv. Mater. 2016, 28, 6104 – 6135.

Manuscript received: September 17, 2018Revised manuscript received: November 6, 2018

Accepted manuscript online: November 17, 2018

Version of record online: January 2, 2019

Chem. Eur. J. 2019, 25, 3209 – 3218 www.chemeurj.org T 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim3218

Minireview