nchem.1628

NATURE CHEMISTRY | VOL 5 | JUNE 2013 | www.nature.com/naturechemistry 453

Most commercial products contain synthetic polymers, which are synthesized industrially on a hundred-million-ton annual scale. Recently developed controlled polymerization methods14 provide increasingly well defined chain lengths, excellent functional-group tolerance, and access to diverse block copolymer and novel architectures. In contrast, comparatively few polymeriza-tion methods provide periodic two-dimensional structures, a chal-lenge that lies at the interface of polymer science and molecular self-assembly. Diverse strategies have emerged to synthesize and characterize 2D polymers from a disparate range of research tradi-tions, including surface science, crystal engineering, scanning probe microscopy and molecular frameworks. These advances will inspire further synthetic innovation, enable the first rigorous investigations of the properties unique to this topology, and finally, provide a means for organic chemists to design within extended 2D and 3D space.

A 2D polymer is a covalently linked network of monomers with periodic bonding in two orthogonal directions. These polymers have been isolated in three forms: layered crystals, multilayer and monolayer58 films on surfaces, and free-standing sheets. Individual or disordered 2D polymer sheets are a subset of 2D crystals, and periodic layered 2D polymers are 3D crystals in which covalent bonding is confined to two dimensions and non-covalent interac-tions direct their organization in the third. It should also be noted that multiple definitions of 2D polymers have been used histori-cally. The term 2D polymer has been used more broadly to include linear polymerizations performed at interfaces or in layered non-covalent assemblies, or to irregularly crosslinked polymers confined to surfaces or layered films. Each of these methods provides 2D objects that lack periodic bonding9,10. However, spatially confined polymerizations represent pioneering examples of controlled reac-tions within organized molecular assemblies, an important step towards 2D polymerization. The term 2D polymer has also been defined more restrictively to include only materials isolated as single sheets11, a milestone that was achieved only recently12.

For this Review, we describe polymers with 2D topologies in all of the above isolated forms, just as we consider a linear polymer to be inherently one-dimensional whether it is found in the bulk, dissolved in solution, or isolated as an individual macromolecule. However, we consider 2D networks linked by coordination bonds or other non-covalent interactions to be outside the scope of this Review.

Rationally synthesized two-dimensional polymersJohn W. Colson and William R. Dichtel*

Synthetic polymers exhibit diverse and useful properties and influence most aspects of modern life. Many polymerization meth-ods provide linear or branched macromolecules, frequently with outstanding functional-group tolerance and molecular weight control. In contrast, extending polymerization strategies to two-dimensional periodic structures is in its infancy, and successful examples have emerged only recently through molecular framework, surface science and crystal engineering approaches. In this Review, we describe successful 2D polymerization strategies, as well as seminal research that inspired their development. These methods include the synthesis of 2D covalent organic frameworks as layered crystals and thin films, surface-mediated polymerization of polyfunctional monomers, and solid-state topochemical polymerizations. Early application targets of 2D poly mers include gas separation and storage, optoelectronic devices and membranes, each of which might benefit from pre-dictable long-range molecular organization inherent to this macromolecular architecture.

Graphite, graphene13 and hexagonal boron nitride14,15 are proto-typical 2D polymers whose optical, electronic and mechanical prop-erties have attracted intense contemporary interest. Such materials feature distinct properties from other 2D materials, such as ionic 2D crystals16, as a result of their covalent bonding. These 2D poly-mers are prepared by exfoliation17, epitaxial growth18 or pyrolytic decomposition19,20 techniques, which preclude rational modifica-tion of their structure. A wide variety of nanometre-size graphene fragments21 have also been synthesized using both top-down and bottom-up approaches. Structurally precise nanographenes22, the largest of which contains 222 carbon atoms23, are derived from multi step synthetic pathways rather than a single polymerization reaction. Such approaches are likely to prove inefficient for prepar-ing significantly larger systems. Nevertheless, these hydrocarbons are useful models of structurally precise graphene subunits, and exhibit desirable optoelectronic properties as a consequence of quantum confinement and their specific edge structure.

Although the properties of synthetic 2D polymers are only now being explored, they offer several features that distinguish them from linear macromolecules. For example, 2D polymers derive their topology from the directionality of their covalent link-ages and offer a means of organizing chemical functionality with atomic precision over long distances. In contrast, subtle changes in chemical structure can dramatically and unpredictably affect the solid-state structure of molecular compounds24,25 and linear polymers2628. Many non-covalent assemblies show similar struc-tural order2932, but 2D polymers promise to be far more robust. The mechanical properties and stability of both free-standing 2D polymer sheets and their multilayer assemblies are as yet untested, but should reflect their covalently bonded structure, in contrast to the interchain interactions of entangled 1D polymers (Fig.1a,b). Two-dimensional polymers also frequently feature permanent pores of defined size and shape, providing high specific surface areas and free volume suitable for postsynthetic functionaliza-tion. As such, they might prove useful as membranes and for filtration, molecular recognition and sensing, catalysis, optoelec-tronic devices and many other applications that leverage specific molecular organization. Realizing these opportunities will require general and robust methods to synthesize, isolate and characterize 2D polymers in their various forms.

Department of Chemistry and Chemical Biology, Cornell University, Baker Laboratory, Ithaca, New York 14853, USA. *e-mail: [email protected]

REVIEW ARTICLEPUBLISHED ONLINE: 12 MAY 2013|DOI: 10.1038/NCHEM.1628

2013 Macmillan Publishers Limited. All rights reserved
mailto:[email protected]://www.nature.com/doifinder/10.1038/nchem.1628

454 NATURE CHEMISTRY | VOL 5 | JUNE 2013 | www.nature.com/naturechemistry

Two-dimensional polymerizations are distinct from their linear counterparts in that they must precisely control monomer assem-bly before or during covalent bond formation, otherwise, aperiodic, crosslinked networks are obtained. Two strategies provide such con-trol: Thermodynamic approaches perform monomer assembly and polymerization simultaneously under dynamic-bond-forming condi-tions (Fig.1c), providing 2D polymers commonly known as covalent organic frameworks (COFs). Kinetic approaches (Fig.1d) polymerize building blocks preorganized into 2D assemblies, and are thus com-patible with irreversible reactions. Both approaches rely on shape-persistent monomers and directional interactions to enforce specific bonding geometries, and all existing examples employ an interface or layered crystal to template the formation of the desired 2D structure. This Review describes emerging 2D polymerization methods and highlight important developments that influenced these approaches.

Synthesis of 2D polymers under thermodynamic controlThe most general existing 2D polymerization strategy provides layered periodic networks, linked by covalent bonds, known as COFs. These networks are synthesized by condensing appropriately designed monomers under reversible bond-forming conditions. The dimensionality (2D or 3D) and topology of the COF result from the shape of the monomers and the relative orientations of their reac-tive groups (Fig.2). The synthesis of COFs is modular, as chemically distinct monomers produce identical network topologies provided that these design criteria are preserved. No other 2D polymerization strategy currently gives comparable scope and a straightforward means of predicting and tuning the network structure.

COF synthesis was informed by prior advances in other classes of framework materials and designed non-covalent assem-blies. Seminal work by Lehn33,34, Robson35, Wuest36,37 and Fujita38

Monomer Polymer Solid-state packing

1D

2D

a

b

c

d

Dynamic covalentbond formation

Non-covalentassembly Polymerization

Error correctionand crystallization

Figure 1 | Schematic illustration of the structures of one- and two-dimensional polymers, and strategies for 2D polymerization. a, Some 1D polymers pack into crystalline domains (shown as blue-shaded regions) within a disordered matrix, whereas others are amorphous in the solid state. b, 2D polymers can be isolated as multilayer crystals or as individual sheets. c, 2D polymerization under thermodynamic control, in which bond formation and crystallization occur simultaneously. d, 2D polymerization under kinetic control relies on monomer assembly prior to bond formation. Red squares represent monomers where blue and green triangles are functional groups. Solid black lines represent covalent bonds formed during polymerization. Dashed black lines are non-covalent interactions holding the monomers in a pre-organized position before polymerization.

REVIEW ARTICLE NATURE CHEMISTRY DOI: 10.1038/NCHEM.1628

http://www.nature.com/doifinder/10.1038/nchem.1628


B

BOHHO

HO OH

HO OH

OH

OHHO

HO

OH

OH

O

O

BOH

OH

CN

B

BOH

OH

OH

OH

H

O

H

O

3

3

OB O B

OB

OB

O

OB

O

N

NH

NO

NH2

NH

ONH2

N

NN

2 H2O

O

H2O

H2O

H2O

3 H2O

cat. H+

cat. H+

ZnCl2

400

Mesitylene / dioxane90 C

BF3 OEt2

OB

O

O BOO

B O

HO OH

OH

OHHO

HO

OH

OH

HO

HO

N

NN

N

N

N

N

N

OHOH

HO

HO

HOHO

OHOH

M

OO

OO

OO

N

NN

N

N

N

N

NM

OO

OO O

O

OO

NN

O

O

O

O

BBB B BB

BB

BB

BBOH

OHHO

HO

HO

HO OH

OHHO

HO OH

OH

HO

HO OH

OH

HO

HO OH

OH

OH

OHHO

HO

R

R

OHHO

OH

OH

HO

HO

B

BB

HO OH

OH

OH

HO

OHB

B

BHO

OH

OH

OH

HO OH

NS

N

BBHO

HO OH

OH

R = H, Me, Et, Pr

M = H2, Zn, Ni

N

N

N

N

B

B

B

B

OHC

M

M = H2, Zn, Cu

CHOOHC

CHO

CHOOHC

CHO

M = H2, Ni

O

HNO

NH

NH2

H2N

NH2H2N

N

NH

NHN

NH2

NH2H2N

H2N

CHO

HHTP

BDBA

RR

BBOH

OHHO

HO

R = C6H13

NH2BH3

BH

NBHNHB

N2 H23

NH2BH3

NH2BH3H3BH2N

HO

HO OH

OH

O

O

R

R R = H, Me

R

R

R

R = H, OH

RR

R RR = H, F

HO OH

HO

HO

OH

OH

NS

N

BOH

OHB

HO

HO

HOOH

OHOH

HOHO

S

SB B

OH

OHHO

HO

R

RR = H, Me

R = H, CH2N3

R

R

CHO

CHOOHC

OHC

Equilibria used to prepare 2D COFs

a

b

(ref. 80)

(ref. 81)

(ref. 82)

Other reactions used to prepare 2D COFs

Amines, hydrazines and ammonia boranes

Aldehydes

HO

OH HOOH

OHHO OH

HO

Boronic acids

Catechols and acetonides

+

+

+

+

+ +

+

+

+

+

+

+

C

Figure 2 | Thermodynamic synthesis of COFs. a, Condensation of HHTP and BDBA provides a prototypical 2D COF (COF-5). Its topology and chemical structure (inset) are displayed. b, Equilibria that have provided bulk 2D COFs are given, as well as all monomers used to synthesize layered 2D COFs by the date of submission. References are indicated for monomers that have been used to construct COFs that are not discussed in the text.

REVIEW ARTICLENATURE CHEMISTRY DOI: 10.1038/NCHEM.1628



established the connection between network structure and coordi-nation geometry within 2D coordination networks, 3D coordination polymers, 3D hydrogen-bonded networks and coordination cages, respectively. These concepts were elaborated to synthesize perma-nently porous 3D coordination polymers known as metalorganic frameworks (MOFs)39 or porous coordination polymers (PCPs)40. These high-surface-area materials have attracted intense recent interest for gas storage, separation and catalysis. The MOF/PCP approach has proved to be remarkably versatile, with thousands of new materials reported in the past decade4145.

One of the challenges of extending MOF design concepts to cova-lently linked polymers is that many organic functional groups do not undergo dynamic exchange processes as readily as coordination complexes, which typically possess lower bond strengths. Synthesis of COFs relies on reversible covalent-bond-forming reactions, such as condensations that furnish boroxines or imines. These and other

equilibria have been used in various dynamic molecular and poly-meric systems. Relevant examples include the concept of dynamic covalent chemistry46, in which a single receptor is amplified from a complex, equilibrating mixture in the presence of a template, as well as in the syntheses of organic polyhedra4749, ladder polymers50,51 and reorganizing, self-repairing polymers5254. The synthesis of COFs expands on these examples of dynamic covalent bond formation to achieve predictable long-range order in 2D or 3D.

Yaghi and co-workers synthesized the first two 2D COFs, each containing 1,4-benzenediboronic acid (BDBA)55. A boronate ester-linked network designated COF-5 (Fig.2a), obtained from condensing BDBA and 2,3,6,7,10,11-hexahydroxytriphenylene (HHTP), represents a prototypical 2D COF structure. The COF-5 network adopts a hexagonal unit cell (a = b = 2.97 nm) whose macromolecular sheets stack in an eclipsed fashion, similar to that found in hexagonal boron nitride, with a 0.34nm interlayer spacing, similar to that of graphite. This structure, as well as all subsequent COFs, was assigned by comparing its refined pow-der X-ray diffraction (PXRD) pattern with simulated patterns of eclipsed and staggered stacking arrangements; no single-crys-tal X-ray structure of a 2D COF has been reported so far. The COF-5 structure exhibits permanent porosity with high surface area (1,590m2g1), a narrow pore-size distribution, and excellent thermal stability (>350C). A 2D COF, COF-1, derived from the dehydration of BDBA was reported simultaneously with COF-5, and remains the single example of staggered interlayer stacking. Its benzene subunits are arranged above boroxine rings in adjacent layers, probably due to the size and electronic complementarity of its boroxine and benzene subunits not found in networks with larger aromatic surfaces56.

In several recent computational studies, the structure of 2D COFs has been investigated in greater detail than has been pos-sible to measure experimentally. Aromatic systems do not typi-cally crystallize into fully eclipsed cofacial structures, as suggested by refinement of 2D COF PXRD data. It is likely that they adopt slightly offset structures that balance attractive forces with repulsive electrostatic interactions. Indeed, density functional theory calcu-lations (Fig.3) by Clancy57, Heine58,59 and Zhou60, identified offset structures as energy minima. Clancy and co-workers also correlated the preferred offset to the density of boronoxygen linkages in each COF61, suggesting that interlayer boronoxygen or boronaromatic interactions may also influence COF structure. These offsets have no single preferred direction with a minimum energy observed in a symmetric ring in the potential energy surface (Fig.3b), indicat-ing that the layers probably adopt random offsets from a vector normal to the stacking direction. However, these offsets have yet to be confirmed experimentally because peak broadening in powder samples is too large to differentiate between eclipsed and slightly offset structures. High-resolution microscopy or complemen-tary spectroscopy will be necessary to confirm these calculations in lieu of higher-resolution X-ray diffraction. Recently, Jiang and co-workers further engineered COF offsets by incorporating arene-perfluoroarene interactions62, which are expected to induce closer interlayer packing.

Thirty-seven unique 2D COFs have now been synthesized as layered crystals, most linked by boronate esters, including several 3D networks6365 outside the scope of this Review. Monomers uti-lized in 2D COFs are shown in Fig. 2b. Most are planar, though Zheng and co-workers crystallized a 2D COF using a bowl-shaped monomer, hexahydroxycyclotricatechylene (CTC), linked by BDBA (CTC-COF)66. Adjacent columns of stacked CTC units face in opposite directions, forming rippled, rather than planar, 2D pol-ymer sheets. A 2D COF incapable of cofacial stacking that incor-porated hexahydroxytriptycene linked by BDBA (TDCOF-5)67 was reported by El-Kaderi and colleagues, which exhibited an out-of-plane lattice parameter commensurate with this difference

a

c

1.8

2.02.2

2.4

2.6

2.8

Energy (eV

)

Laye

r offs

et (

)

3210123Layer offset ()

b

Layer offset ()

3

2

1

0

3

1

2

BBO

O O

O

OB O

OBO O

BO

O BO

Figure 3 | Computational studies indicate that COFs adopt offset layered structures. a, Partial structure of a 2D COF whose preferred interlayer offsets were calculated using DFT. b, Potential energy surface as a function of interlayer offset of the COF in a. Energy minima corresponding to several degenerate offset structures suggest that no particular direction is preferred. c, The lowest energy structure of the COF represented in a. Parts a and b reproduced with permission from ref.57, 2011 ACS.




in packing. Many polyfunctional boronic acids have provided 2D COFs, but the structural diversity of the catechol compo-nent has remained limited. Only HHTP55,6871, octahydroxyph-thalocyanine7274 and 1,2,4,5-tetrahydroxybenzene7578 have been incorporated into multiple materials. The low solubility and oxi-dative sensitivity of planar polyfunctional catechols complicates large-scale synthesis and storage, and oxidized catechol building blocks are known contaminants of isolated frameworks. Spitler and Dichtel developed an alternative method73 in which boronate esters are formed from soluble and bench-stable catechol acetonides in the presence of the Lewis acid catalyst BF3OEt2. This method pro-vided square free-base and Ni phthalocyanine-containing frame-works, as well as COF-5 and the HHTP-biphenyl network known as COF-10 (ref.68). A mechanistic study of boronate ester forma-tion79 revealed several complex equilibria, including non-produc-tive BF3ArB(OH)2 complexes, suggesting that alternative Lewis acids might further improve this approach.

Although boronic acid condensations now dominate the 2D COF literature8082, other equilibria also provide 2D crystalline networks. Trimerization of the nitrile moieties of 1,4-dicyanobenzene under ionothermal conditions (molten ZnCl2, 400C) yields a hexagonal triazine-linked framework83. However, this COF showed limited crystallinity, and these reaction conditions are unlikely to tolerate many functional groups. This contrasts with triazine formation under mild acid-catalysed conditions, which is irreversible, provid-ing high-surface-area, but amorphous, porous polymers. Carbonyl condensation reactions also show promise, as demonstrated by 2D imine-84 and hydrazone-linked85 COFs synthesized under solvo-thermal acid-catalysed conditions. Notably, two enamine-linked COFs derived from the condensation of linear diamines and trifor-mylglucinol exhibited enhanced stability to concentrated aqueous acid and boiling water86. Borazine-linked frameworks have also been reported, which were synthesized from the thermal decom-position of an ammonia-borane precursor87. Other reversible reac-tions might also provide crystalline networks, and broadening the range of COF linkage chemistries represents an important challenge for increasing the scope of 2D polymerization.

Inspired by intense interest in MOFs and other high-surface-area materials for storing, sequestering or separating technologically relevant gases, 2D COFs were initially evaluated for these applica-tions. The COF networks comprise lightweight elements that could provide improved gravimetric storage capacity, and many adsorb significant amounts of CH4, CO2 and H288,89. The COF-10 network demonstrated notable NH3 uptake90, outperforming zeolite 13X, sulfonic acid-containing polymer Amberlyst 15, and mesoporous silica MCM-41. The strong interaction between the COF and NH3 derives from the Lewis acidity of its boronate ester linkages. Decreased NH3 uptake observed over three adsorptiondesorption cycles was attributed to turbostratic disorder induced into the 2D layered crystals, suggesting that using Lewis acidic 3D COFs might improve performance. Generally, 2D COFs display smaller surface areas than their 3D counterparts and many MOFs88, in part because their 2D architectures are inherently more dense.

Post-synthetic modification of 2D COFs uses the periodicity of the framework to organize secondary functionality, either through covalent or coordinative bonding. Wang and co-workers84 pre-pared a 2D imine-linked framework, COF-LZU1, by condensing 1,3,5-benzene tricarboxaldehyde and 1,4-diaminobenzene. This framework adopts a nearly eclipsed layered crystal with 1.8-nm-wide pores, and is hypothesized to bind metal ions through bidentate coordination to proximal N atoms in adjacent layers. For example, 0.5 equivalents of Pd2+ per unit cell were incorpo-rated by soaking COF-LZU1 in a solution of Pd(OAc)2, inducing only minor changes to its PXRD pattern. The resulting Pd-loaded COF is a highly active catalyst for the SuzukiMiyaura cross-cou-pling reaction, achieving high yields (often >97%) at low catalyst

loadings (1mol% Pd). Control experiments suggested that bound Pd ions were the active catalysts, and the insoluble COFs were eas-ily recovered and recycled at least four times without loss of activ-ity. This study demonstrates a promising future for 2D polymer catalysts to combine specific metal-coordination environments, the operational simplicity of heterogeneous catalysis, and sub-strate selectivity imparted by the pore structure, though further improvements in stability are needed. Another example of COF functionalization was reported by Jiang and co-workers70, who performed Cu-catalysed azide-alkyne cycloaddition (CuAAC) reactions within the pores of 2D COFs bearing azide moieties. Several cycloaddition partners and functionalization densities demonstrate the versatility of this strategy to distribute external functionality throughout the porous crystal.

A distinctive aspect of 2D layered COFs is their long-range -orbital overlap in the stacking direction, which gives rise to high exciton- and charge-mobilities of interest for optoelectronic devices. First, their pores run parallel to the stacking direction, providing an opportunity to introduce complementary organic or inorganic semiconductors. The resulting vertically aligned p-n junctions embody a morphology long thought to be ideal for pho-tovoltaic performance91. Second, the stacked -systems found in COFs adopt zero-degree dihedral angles between layers, maximiz-ing their orbital interactions, in contrast to discotic liquid crystals, which often stack into less ideal twisted interlayer structures due to steric crowding of their pendant mesogens. Finally, COFs generally show excellent thermal stability and do not undergo thermal phase transitions, unlike polymeric semiconductors, which frequently require thermal or solvent annealing to attain morphologies suit-able for reasonable photovoltaic performance.

Jiang and co-workers reported the first photoconductive COFs, each based on 2,7-pyrene diboronic acid, either self-con-densed into a boroxine-linked structure92 (Fig. 4a) or co-crys-tallized with HHTP to yield a boronate-ester-linked material69. Photoconductivity observed in devices consisting of the COF powder sandwiched between Au and Al electrodes (Fig.4a, right panel, red trace) suggested long-range exciton delocalization, pre-sumably through the stacked pyrene moieties. Charge mobilities of phthalocyanine- and porphyrin-based COFs were later studied using laser flash-photolysis time-resolved microwave conductiv-ity, which measures the intrinsic charge-carrier mobility of the material independent of defects or impurities. Phthalocyanine-based COFs display hole mobilities72 as high as 1.3 cm2 V1 s1, whereas mobilities as high as 8.1cm2V1 s1 were reported for a porphyrin-containing imine-linked COF93. Charge separation in COFs has been studied for COFs in which the donor and accep-tor are both incorporated into the framework94, and also for non-covalent post-synthetic functionalization using a soluble fullerene derivative acceptor95. In both cases, ultra-fast charge transfer on the timescale of picoseconds occurred from the electron donor to the acceptor. Charge-separated states with long-lived lifetimes of 1.5s at 280K and 1,500s at 80K, respectively, were observed in the donoracceptor framework94. Further, the first COF-based photovoltaic device was reported95 using a thienothiophene-con-taining 2D COF/fullerene hybrid, albeit with an unoptimized pho-toconversion efficiency of 0.053%.

An outstanding issue in studies of COF charge mobility is how interlayer offsets impact their optoelectronic properties. For exam-ple, calculations suggest that HHTP-containing COFs with per-fectly eclipsed stacking could potentially delocalize charge carriers over the entirety of the HHTP pillar96, resulting in band-like charge transport with mobility as high as 10m2V1s1. However, the degree of lateral sliding (for example, offset) between adjacent HHTP mol-ecules in a pillar strongly affects the transfer integral between lay-ers97. Charge is delocalized to the greatest extent when the lateral offset approaches 3, whereas offsets of 1 between layers have




extremely small transfer and overlap integrals, resulting in charge localization to a single HHTP molecule in a stack. Therefore, con-trolling the nature of interlayer stacking in 2D COFs is likely to strongly influence their charge-transport properties.

One of the most attractive aspects of COFs for optoelectronic applications is that their 2D layered structures are preserved even as the molecular structure of the monomers is varied. In contrast, even subtle changes in structure often induce dramatic changes in the packing behaviour of molecular24 and polymeric26 semiconduc-tors. For example, Dichtel and co-workers systematically increased the pore size of a square Zn phthalocyanine framework to diagonal widths of 2.7, 3.4, 4.0 and 4.4nm by employing increasingly long and structurally diverse diboronic acid linkers98. Similarly, hexag-onal COF lattices have also been expanded57,99 to 4.0and 4.7nm. These larger pore sizes might accommodate complementary semi-conductors or other functional guests. Jiang and co-workers uti-lized the modularity of COF synthesis to characterize the charge mobility of various cofacially stacked -systems100. They found the semiconducting properties of porphyrin-containing 2D COFs (Fig. 4b) to depend strongly on the identity of the central metal ion: the free-base porphyrin COF conducts holes, Zn porphyrin COFs are ambipolar, and Cu porphyrin COFs conduct electrons. Separately, Jiang and co-workers co-condensed complementary triphenylene donors and benzothiadiazole acceptors into a frame-work that retained its 2D layered morphology, a rare example of self-segregated donoracceptor systems capable of simultaneous hole and electron conduction101.

Their high intrinsic mobilities and structural versatility confirm the promise of 2D COFs as novel organic semiconductors and set the stage for incorporating these materials into working devices. The performance of molecular or polymeric semiconductors in working devices is strongly affected by impurities and local disorder, and the identity and prevalence of relevant defects and impurities within 2D COFs remain largely uncharacterized. Such measurements will be facilitated by preparing these materials as thin films on both con-ducting and insulating surfaces.

Thermodynamic and kinetic surface-supported synthesesTwo-dimensional COFs are typically obtained as insoluble powders, an isolated form that limits their utility. Microcrystalline powders are suitable for many applications traditionally associated with porous materials, including gas storage, separations and catalysis. However, their crystallites are randomly oriented and are not easily interfaced to surfaces, including electrodes. Two-dimensional poly-mer membranes and scaffolds will demand few-layer, free-standing or surface-bound materials with controlled orientation. Likewise, interfacing COFs reliably to electrodes and insulators will be neces-sary to realize their optoelectronic potential. Therefore, methods of synthesizing multilayer or monolayer 2D polymer films are of great contemporary interest. Surfaces also offer a natural interface to tem-plate molecular assembly and polymerization in two dimensions.

Using the same solvothermal conditions that produce COF pow-ders, Dichtel and co-workers grew COF thin films on single-layer-graphene-functionalized substrates102. Figure 5 compares grazing incidence X-ray diffraction (GIXD) data of powders and films, a diffraction technique that measures the periodicity and orientation of thin films relative to a substrate surface. These data reveal that the powder and film exhibit identical network topologies. However, diffraction from powders is observed as an isotropic ring of inten-sity from low to high out-of-plane angle (Q, where Q|| represents the plane of the sample), including the out-of-plane COF stacking, observed in the powder at 1.84 1. In contrast, diffraction from thin films is localized to small Q, and the stacking peak is absent, indicating that the -systems of the COF film are oriented normal to the substrate. Scanning electron microscopy and focused ion beam milling of the thin films revealed a continuous COF film that con-forms to the substrate. This method was generalized to pyrene- and several phthalocyanine-based COFs, which each crystallized as an oriented thin film. One of the attractive features of this approach is that single-layer graphene can be transferred to arbitrary substrates, and has shown promise as a transparent electrode material103. For example, single-layer graphene was deposited onto fused SiO2, after which COF films were grown and characterized spectroscopically

h

Al

Au

I

=

h off

h on

10

5

0

012 1 2

5

I (nA

)

Bias voltage (V)

AmbipolarHole Electron

h = 3.5 cm2 V1 s1

e = 0.016 cm2 V1 s1

Linker =

OB

OOB

O

N

HN

N

NH

N

N

N

N

N

N

N

N Cu

h = 0.032 cm2 V1 s1 e = 0.19 cm2 V1 s1

Zn

a

b

Figure 4 | 2D COFs exhibit photoconductivity and variations in charge transport as a function of their structure. a, 2,7-pyrene diboronic acid self-condenses into a boroxine-linked 2D COF. Blue, carbon; white, boron; red, oxygen. The individual crystallites are represented as purple spheres with a grey shell for clarity (left). Crystalline powder sandwiched between two electrodes (middle) exhibits photoconductivity (right). b, The nature of charge transport through 2D porphyrin COFs depends on the presence or nature of the coordinated metal ion. h and e stand for hole and electron mobilities, respectively. Figure reproduced with permission from: a, ref.92, 2009 Wiley; b, ref.100, 2012 Wiley.




in transmission mode. The generality of this procedure will open new applications and characterization methods for 2D COFs that are unavailable in their powder forms. This method also represents a means of incorporating COFs into optoelectronic devices.

The above solvothermal conditions produce multilayer COF films that range in thickness from 20400 nm but have not accessed crystalline few-layer or monolayer films. Islands of COFs consisting of 1025 layers were isolated by Zamora and co-work-ers104 by sonochemically exfoliating bulk COF powders. These multilayers were dispersed in solvent and deposited onto highly oriented pyrolytic graphite (HOPG), where their thickness was measured by atomic force microscopy. These results suggest that other 2D COFs might also be exfoliated to few or single layers, and the properties and stability of materials isolated in this way remain undetermined.

Individual 2D COF layers have been grown on metal surfaces through ultra-high vacuum (UHV) deposition of the mono-mers8,105,106, an approach pioneered by Porte and co-workers8. These single-layer 2D polymers are typically characterized using scanning tunnelling microscopy (STM), which provides atomic-scale resolu-tion of the polymer structures and periodicity over approximately 100100nm field of view (Fig.6). For example, macrocyclic pen-tagon and heptagon defects were observed107 in COF-1 monolayers alongside the expected fused hexagons. Zamora and co-workers108 optimized the deposition parameters of acid chloride and alcohol monomers to synthesize a single-layer polyester COF that was defect-free over areas of 100nm2. This example demonstrates that fine control of monomer deposition can provide long-range order even in the absence of reversible bond formation, as this reaction is unlikely to be reversible and no polyester COFs have been prepared

OHOH

HO

HO

OHOH

BHO OH

BOHHO

BOO

OBO

BO O

O BOO

B O

OBO

BO

O

OB O

OBO

B

B O

O

O

O

OBO

O BO

O BO

OB

O

B

BOO

O

O

OB

O

HHTP BDBA

COF-52.7 nm

a

b

c

Substrate-supportedgraphene

mesitylene /dioxane90 C

Substrate-supported graphene

25-400 nm COF-5 filmCOF-5 powder

0.2

0.6 1,000

100

10

0.4

0.2

0.00.4 0.6

Q (

1)

Q (

1)

QII (1)

0.8 1.0 1.2 1.4 1.6 1.8 2.0

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

0.6

0.4

0.2

0.0

+

Figure 5 | COF thin films supported on single-layer graphene. a, 2D COF thin films were prepared by including a single-layer graphene-functionalized substrate in the polymerization solution. Structure (left), schematic representation of the samples (middle), and schematic of the resulting COF. b, GIXD pattern of a 2D COF powder whose crystallites are randomly oriented (inset). c, GIXD pattern of the corresponding film, whose crystallites are oriented with the stacking direction normal to the graphene surface (inset). Q and Q|| are the out-of-plane and in-plane angles, respectively. Figure reproduced with permission from ref.102, 2011 AAAS.




in bulk form. Monolayers of COFs were also synthesized by evapo-rating boroxines with pendant aryl bromides onto a gold surface to provide non-covalent dimers, which undergo arylaryl coupling to yield a hexagonal lattice on heating109. Preparation of COFs by UHV is a 2D polymerization technique under kinetic control, in contrast to solvothermal polymerizations under thermodynamic control.

Alternative methods for forming surface-bound COF monolay-ers utilize reversible bond formation to access or improve the long-range order of the polymer. Such approaches tolerate a variety of monomer deposition or initial polymerization strategies, followed by a final annealing procedure. Dienstmaier and co-workers pio-neered this approach by depositing COF-1 nanocrystals of 7.8nm average diameter onto HOPG5 (Fig. 6b). These crystallites were annealed in a humid environment, enabling dynamic hydroly-sis and reformation of the boroxine linkages. Under these condi-tions, the few-layer, small crystallites ripen into a COF monolayer with ~40 nm crystalline domains. Similar COF monolayers were obtained when the BDBA monomer itself was used in place of the nanocrystal seeds. The generality of this method was subsequently demonstrated for monolayer COFs with pore sizes as large as 3.8 nm by lengthening the diboronic acid monomer110. Wan and co-workers also improved the order of UHV-deposited monolayers on annealing in the presence of trace amounts of water released by heating a hydrated metal salt at pressure111. These results demon-strate that reactivating dynamic bond exchange processes provides a means of annealing COF monolayers and improving their long-range order, while also mitigating the need to deposit COF mono-mers under careful conditions.

Monolayer and multilayer COF films offer a wide range of oppor-tunities for further study. First, the full range of linkage chemistries used in COF powders has yet to be generalized to oriented thin films. Likewise, the nucleation processes of COFs are unknown; control of domain size, film thickness and selective growth on patterned substrates might follow from improved understanding. The porosity of multilayer films is also undemonstrated. Ultimately, COF multi-layer and monolayer films will enable many advanced functions for these materials, and properties such as their ionic and electronic conductivity, dielectric constants, etch resistance, patternability and hydrolytic stability are each areas of immediate interest.

Surface-mediated molecular networks and polymerizations Another strategy for synthesizing single-layer 2D polymers involves depositing shape-persistent monomers onto a crystalline metal sur-face. The surface is then heated to speed molecular diffusion and activate covalent bond-forming reactions, often arylaryl couplings. The resulting 2D polymers remain adsorbed to the substrate and are characterized by STM, revealing their structure, defects and domain sizes with atomic precision. This 2D polymerization strat-egy emerged from pioneering STM studies of 2D molecular crys-tallization and directed non-covalent assembly, which have been reviewed previously32,112. Covalent networks have appeared com-paratively recently and remain much more rare.

Hecht and Grill reported the first covalent two-dimensional polymers of this type by developing the above approach to various halogenated porphyrins6. The substitution pattern of the porphyrin monomer determines the dimensionality of the resulting coupled products. For example, a porphyrin bearing a single aryl-bromide dimerizes on the surface, and linear disubstituted monomers pro-vide 1D polymers, a method recently applied to surface-bound molecular wires113115 and graphene nanoribbons116. Tetrafunctional porphyrins produce 2D polymers with dimensions of approxi-mately 5 5 monomers (Fig. 7). The authors later modified this approach to perform sequential polymerizations based on the lower activation temperature of aryl iodides relative to aryl bromides7. Heating 5,15-bis(4-bromophenyl)-10,20-bis(4-iodophenyl)por-phyrin on Au(111) to 120C initiates polymerization at the trans aryl-iodide positions, providing linear porphyrin polymers that retain their aryl-bromide moieties. These aryl bromides are acti-vated on increasing the temperature to 250 C. The resulting 2D polymers were more extended than the single-step approach (88 monomer units and larger) due to the preorganization of the intermediate 1D polymer chains in the subsequent 2D crosslinking step. This two-step protocol also afforded the opportunity to incor-porate additional dibrominated monomers selectively during the second polymerization, although the resulting materials were not periodic in two dimensions.

Abel and co-workers demonstrated117 2D polymerization of 1,2,4,5-tetracyanobenzene (TCB), which tetramerizes to form phthalocyanine networks on Au(111), Ag(111) or NaCl/Ag(100)

10 nm

UHV deposition

UHV

BOB O B

O

BO B O

BO

B O BOBO

B OBOB

O

BOBO

B O

BOBO B O

B OBOB

O

B O BOBO

BO B O

BO

BOB O B

O

BO B O

BO

B O BOBO

B OBOB

O

B O BOBO

BO B O

BO

BOB O B

O

BOB O B

O

BO B O

BO

B O BOBO

B OBOB

O

BOBO

B O

BOBO B O

B OBOB

O

B O BOBO

BO B O

BO

BOB O B

O

BO B O

BO

B O BOBO

B OBOB

O

B O BOBO

BO B O

BO

BOB O B

O

Dropcast

Au(111) or Ag(111)

Highly orientedpyrolytic graphite

a

b

Dropcast COF-1

H2O annealing

Solvent / H2Oannealing

24 nm

B

BOHHO

HO OH

10 nm 10 nm

15 nm

Figure 6 | Synthesis of surface-supported single-layer 2D polymers. a, BDBA self-condenses onto crystalline metal surfaces following UHV deposition. The STM image (middle) shows formation of COF-1 monolayers containing various 5- and 7-membered ring defects. Annealing this film in a humid environment improves the order of the film (right). b, COF-1 nanocrystals dropcast on to HOPG (middle) were similarly annealed to provide a monolayer with improved long-range order (right). Figures reproduced with permission from: a, middle image, ref. 8, 2008 ACS; right image, ref. 111, 2012 RSC; b, ref.5, 2011 ACS.




surfaces when co-evaporated with Fe. The polymerization depended strongly on the TCB:Fe stoichiometry. At a 1:2 ratio, complete cyclization to a 2D polymer was observed, with domain sizes on the order of 400nm2, or about 812 phthalocyanines per side. A 1:4 TCB:Fe ratio instead provided isolated phthalocya-nines with pendant nitrile groups, which assemble into periodic non-covalent arrays. The nature of the surface also played a key role in network formation, as polymerizations that took place on Au or Ag gave small domain sizes, whereas, remarkably, those on NaCl formed a single polymeric domain (103104 nm2). This large domain size is notable for a 2D polymer synthesized under kinetic control and illustrates the important role of the substrate on the polymerization outcome. Future studies of the optoelectronic properties of these conjugated phthalocyanine polymers will also be of great interest.

Much like COFs, a key advantage of surface-mediated 2D polym-erizations is that their topology is predictable based on monomer structure. Although the scalability of these methods is inherently limited, their characterization by STM provides direct evidence of their structure, domain size and defects. These studies also moti-vate fundamental studies of molecular diffusion on surfaces. For example, Mllen and Fasel correlated surface mobility and degree of polymerization as a function of monomer size and structure118, though it is not fully understood how this mobility changes dur-ing the polymerization. Surface-mediated reactions have since expanded in scope to include imide119, imine120 and urea121 con-densations, which might also be amenable to forming 2D periodic networks. Additional opportunities for these polymers include organizing guest molecules into periodic arrays, as was shown for C60 by Champness and co-workers122. Transferring these polymers to insulating substrates might facilitate their electronic characteri-zation and use as templates or coatings, particularly as larger 2D polymers are accessed through on-surface polymerization.

Reactions in organized 2D assembliesNon-covalent 2D assemblies can organize monomers to place reactive groups in close proximity for subsequent polymerization. Such assem-blies can derive their 2D form from layered crystals, surface confine-ment or amphiphile assembly in solution. This strategy offers a route to synthesizing high-molecular-weight networks because molecular organization can occur over long length-scales. The polymerization

reaction must be efficient, selective and sufficiently mild to preserve the assembly; however, it need not be reversible, greatly expanding the number of compatible bond-forming reactions compared with solvothermal COF synthesis or surface-mediated polymerization.

Elaborating on prior work in polymerization within organized assemblies, Stupp and co-workers confined two distinct polymeri-zations within spatially segregated domains of an organized molec-ular assembly9. Amphiphiles containing acrylate (A) and nitrile (B) polymerizable functionalities were incorporated into a mesogen that assembled into layered 2D sheets (Fig.8). Thermally initiated

Figure 8 | Two distinct polymerizable groups A and B were incorporated into a mesogen. Thermal polymerization results in a bilayer superstructure with independent linear polymers confined to two distinct planes. Black lines represent covalent bonds resulting from the polymerization. Figure reproduced from ref.9, 1993 AAAS.

Bond formation

4 nm

Activation

Br

NH N

HNN

Br

BrBr

Br

a

b

Figure 7 | Two-dimensional polymers prepared by surface-catalysed arylaryl coupling. a, Surface-mediated 2D polymerization scheme of the tetrafunctional porphyrin monomer in b. b, Structure of the monomer and STM image of resulting 55 2D networks; reproduced from ref.6, 2007 NPG.

O

OO

O

CN

O

NN

N

O O O O O O

(i) Polymerize group A

(ii) Polymerize group B

O

C5H11

O

Group A

Group B

Mesogen




acrylate and nitrile polymerizations crosslink the mesogens into a bilayer superstructure. Although the resulting polyacrylate and polynitrile chains lack periodicity in two dimensions, this exam-ple represents pioneering work towards 2D polymerizations within organized assemblies.

Ozaki and co-workers demonstrated a 2D polymerization within a crystalline monolayer of alkyne-containing monomer 1 on HOPG123,124 (Fig. 9). This example is also the first true covalently linked 2D polymer of which we are aware. Analysis by STM of the monolayers prior to polymerization indicated that both the internal diyne and terminal alkyne moieties were aligned appropriately for topochemical polymerization. Indeed, UV irradiation yielded a 2D polymer linked by two distinct types of polyacetylenes. The polymer was first characterized using Penning ionization electron spectros-copy123 to show the change in electronic energy levels before and after polymerization. As determined by STM124, the lattice parame-ters of the 2D polymer were nearly unchanged from the monomeric form, which is likely to be a key design criterion for polymeriza-tions of this type. Mllen and co-workers125 subsequently polymer-ized an isophthalic acid derivative using either light or an STM tip to prepare a similar 2D polymer linked by parallel polyacetylenes. Like other surface-mediated polymerization techniques, STM-tip-induced polymerizations provide small quantities and are rela-tively rare. Furthermore, care must be taken to characterize such

polymerizations using complementary analyses, as the STM itself does not provide definitive proof that polymerization has occurred. Nevertheless, this method does offer the possibility of patterning with nanometre-scale resolution126.

Sakamoto and co-workers12 elegantly expanded the scope of topo-chemical 2D polymerizations to take place within layered crystals. Their design utilizes a trigonal cup-shaped monomer bearing reac-tive groups on each of its faces. The monomer crystallizes into a layered structure in which adjacent monomers face in opposite direc-tions, placing an alkyne of one monomer near an anthracene of its neighbour. Irradiation of these crystals induced intermolecular [4+2] cycloaddition reactions, providing a layered two-dimensional poly-mer (Fig.10). Remarkably, the 2D polymer multilayers were exfoliated to dispersed single-layers in suitable solvents, representing the first example of a rationally synthesized 2D polymer isolated as an individ-ual sheet. Transmission electron microscopy revealed few-layer bun-dles of polymer that sometimes retained the same macroscopic shape as the parent multilayer crystal, and selected-area electron-diffraction revealed that the covalently linked polymer retains the topology of its supramolecular precursor. Because molecular crystals can be large and synthesized in bulk quantities, this approach provides access to high-molecular-weight 2D polymers on a preparative scale.

Two-dimensional polymerizations in non-covalent assem-blies, though rare, have been reinvigorated by the recent work of

a b

c

5 nm

bg, ag

PD PA

PA

PD

PA

PD

PA

PD

n

n

n

n

n

nn

n

n

n

n

n

n

n

n

nn

n

n

n

n

nn

n

n

n

n

n

n

n

n

n

n

n

n

n

n

n

n

n

1, n = 5 h

2b

3a

Figure 9 | Irradiation of alkyne-containing monomers induces simultaneous topochemical polymerizations that yield a 2D polymer comprising linear polyacetylene and polydiacetylene chains linked by alkanes. a, Monomer 1 was crystallized onto graphite to align its alkyne functions for the topochemical photochemical reaction. b, STM image of the 2D polymer on graphite. c, The proposed structure of the 2D polymer on graphite generated from the Fourier transform of the image in b. The dark lines represent the polydiacetylene (PD); the dotted lines represent the polyacetylene (PA). Graphite is represented by dark periodic circles. a is the unit vector along the polydiacetylene chain, b is the unit vector between polydiacetylene chains, and represents the angle between vectors a and b. ag and bg represent the unit vectors of the underlying graphite. Parts b and c reproduced with permission from ref.124, 1997 Wiley.




Sakamoto and co-workers. Demonstrating the generality of this approach is an important next step, as it remains difficult to predict molecular crystallization with the precision needed for successful topochemical reactions. The possibility of a monomer crystallizing into multiple polymorphs might also complicate design, although computational crystal structure prediction is improving rapidly.

Conclusion and outlookSynthesizing 2D polymers represents a longstanding challenge in polymer science. Several complementary strategies have emerged recently, fuelled in part by advances in characterization that have enabled unambiguous structural information of these insoluble materials. Further advances will provide insight into the proper-ties of these polymers, which will determine their applications and inspire new methods to produce them. Understanding the fundamental processes of chemical assembly that occur during 2D polymerization, including nucleation, error correction and defect formation, are important chemical questions that remain poorly understood. Ultimately, 2D polymers represent an excit-ing class of materials with major opportunities for fundamental study, the promise of novel properties that emerge from their unique topology and structural precision, and an as-yet undevel-oped application space.

Received 14 November 2012; accepted 7 March 2013; published online 12 May 2013.

References1. Bielawski, C.W. & Grubbs, R.H. Living ring-opening metathesis

polymerization. Prog. Polym. Sci. 32, 129 (2007).2. Matyjaszewski, K. Atom transfer radical polymerization (ATRP): Current

status and future perspectives. Macromolecules 45, 40154039 (2012).3. Matyjaszewski, K. & Tsarevsky, N.V. Nanostructured functional materials

prepared by atom transfer radical polymerization. Nature Chem. 1, 276288 (2009).

4. Semsarilar, M. & Perrier, S. Green reversible addition-fragmentation chain-transfer (RAFT) polymerization. Nature Chem. 2, 811820 (2010).

5. Dienstmaier, J.F. etal. Synthesis of well-ordered COF monolayers: Surface growth of nanocrystalline precursors versus direct on-surface polycondensation. ACS Nano 5, 97379745 (2011).

6. Grill, L. etal. Nano-architectures by covalent assembly of molecular building blocks. Nature Nanotech. 2, 687691 (2007).

7. Lafferentz, L. etal. Controlling on-surface polymerization by hierarchical and substrate-directed growth. Nature Chem. 4, 215220 (2012).

8. Zwaneveld, N.A. etal. Organized formation of 2D extended covalent organic frameworks at surfaces. J.Am. Chem. Soc. 130, 66786679 (2008).

9. Stupp, S.I., Son, S., Lin, H.C. & Li, L.S. Synthesis of two-dimensional polymers. Science 259, 5963 (1993).

10. Schultz, M.J. etal. Synthesis of linked carbon monolayers: Films, balloons, tubes, and pleated sheets. Proc. Natl Acad. Sci. USA 105, 73537358 (2008).

11. Sakamoto, J., van Heijst, J., Lukin, O. & Schluter, A.D. Two-dimensional polymers: Just a dream of synthetic chemists? Angew. Chem. Int. Ed. 48, 10301069 (2009).

12. Kissel, P. etal. A two-dimensional polymer prepared by organic synthesis. Nature Chem. 4, 287291 (2012).

13. Geim, A.K. & Novoselov, K.S. The rise of graphene. Nature Mater. 6, 183191 (2007).

a

b

20 nm

Light

Exfoliate

Polymerize

10 nm

Figure 10 | Topochemical polymerization in a layered crystal provides a multilayer 2D polymer. a, Cup-shaped monomers crystallize with their reactive groups in close proximity with adjacent monomers. UV-light-induced polymerization, followed by exfoliation, yields isolated single-layer 2D polymers. b, Large-view transmission electron microscope image of exfoliated few-layer 2D polymer sheets (left). Higher magnification (right) reveals that the polymers maintain their trigonal shape and topology after irradiation. Figure reproduced from ref.12, 2012 NPG.




14. Whittell, G.R. & Manners, I. Advances with ammonia-borane: Improved recycling and use as a precursor to atomically thin BN films. Angew. Chem. Int. Ed. 50, 1028810289 (2011).

15. Golberg, D. etal. Boron nitride nanotubes and nanosheets. ACS Nano 4, 29792993 (2010).

16. Novoselov, K.S. etal. Two-dimensional atomic crystals. Proc. Natl Acad. Sci. U.S.A. 102, 1045110453 (2005).

17. Novoselov, K.S. etal. Electric field effect in atomically thin carbon films. Science 306, 666669 (2004).

18. Berger, C. etal. Electronic confinement and coherence in patterned epitaxial graphene. Science 312, 11911196 (2006).

19. Li, X. etal. Large-area synthesis of high-quality and uniform graphene films on copper foils. Science 324, 13121314 (2009).

20. Huang, P.Y. etal. Grains and grain boundaries in single-layer graphene atomic patchwork quilts. Nature 469, 389392 (2011).

21. Chen, L., Hernandez, Y., Feng, X. & Mllen, K. From nanographene and graphene nanoribbons to graphene sheets: Chemical synthesis. Angew. Chem. Int. Ed. 51, 76407654 (2012).

22. Pisula, W., Feng, X. & Mllen, K. Charge-carrier transporting graphene-type molecules. Chem. Mater. 23, 554567 (2011).

23. Simpson, C.D. etal. Synthesis of a giant 222 carbon graphite sheet. Chem. Eur. J. 8, 14241429 (2002).

24. Mas-Torrent, M. & Rovira, C. Role of molecular order and solid-state structure in organic field-effect transistors. Chem. Rev. 111, 48334856 (2011).

25. Rivnay, J., Mannsfeld, S.C.B., Miller, C.E., Salleo, A. & Toney, M.F. Quantitative determination of organic semiconductor microstructure from the molecular to device scale. Chem. Rev. 112, 54885519 (2012).

26. Peet, J., Senatore, M.L., Heeger, A.J. & Bazan, G.C. The role of processing in the fabrication and optimization of plastic solar cells. Adv. Mater. 21, 15211527 (2009).

27. Beaujuge, P.M. & Frchet, J.M.J. Molecular design and ordering effects in functional materials for transistor and solar cell applications. J.Am. Chem. Soc. 133, 2000920029 (2011).

28. Tsao, H.N. & Mllen, K. Improving polymer transistor performance via morphology control. Chem. Soc. Rev. 39, 23722386 (2010).

29. Carlucci, L., Ciani, G., Proserpio, D.M. & Porta, F. Open network architectures from the self-assembly of AgNO3 and 5,10,15,20-tetra(4-pyridyl)porphyrin (H2tpyp) building blocks: The exceptional self-penetrating topology of the 3D network of [Ag8(ZnIItpyp)7(H2O)2](NO3)8. Angew. Chem. Int. Ed. 42, 317322 (2003).

30. Dai, F., Dou, J., He, H., Zhao, X. & Sun, D. Self-assembly of metal-organic supramolecules: From a metallamacrocycle and a metal-organic coordination cage to 1D or 2D coordination polymers based on flexible dicarboxylate ligands. Inorg. Chem. 49, 41174124 (2010).

31. Finke, A.D., Gross, D.E., Han, A. & Moore, J.S. Engineering solid-state morphologies in carbazole-ethynylene macrocycles. J.Am. Chem. Soc. 133, 1406314070 (2011).

32. Bartels, L. Tailoring molecular layers at metal surfaces. Nature Chem. 2, 8795 (2010).

33. Lehn, J.M. etal. Spontaneous assembly of double-stranded helicates from oligobipyridine ligands and copper(I) cations: Structure of an inorganic double helix. Proc. Natl. Acad. Sci. USA 84, 25652569 (1987).

34. Baxter, P.N. W., Lehn, J.-M., Fischer, J. & Youinou, M.-T. Self-assembly and structure of a 3 3 inorganic grid from nine silver ions and six ligand components. Angew. Chem. Int. Ed. Engl. 33, 22842287 (1994).

35. Hoskins, B.F. & Robson, R. Infinite polymeric frameworks consisting of three dimensionally linked rod-like segments. J.Am. Chem. Soc. 111, 59625964 (1989).

36. Simard, M., Su, D. & Wuest, J.D. Use of hydrogen bonds to control molecular aggregation. Self-assembly of three-dimensional networks with large chambers. J.Am. Chem. Soc. 113, 46964698 (1991).

37. Wang, X., Simard, M. & Wuest, J.D. Molecular tectonics. Three-dimensional organic networks with zeolitic properties. J.Am. Chem. Soc. 116, 1211912120 (1994).

38. Fujita, M., Yazaki, J. & Ogura, K. Preparation of a macrocyclic polynuclear complex, [(en)Pd(4,4-bpy)]4(NO3)8 (en=ethylenediamine, bpy=bipyridine), which recognizes an organic molecule in aqueous media. J.Am. Chem. Soc. 112, 56455647 (1990).

39. Stock, N. & Biswas, S. Synthesis of metal-organic frameworks (MOFs): Routes to various MOF topologies, morphologies, and composites. Chem. Rev. 112, 933969 (2012).

40. Bae, Y.-S. & Snurr, R.Q. Development and evaluation of porous materials for carbon dioxide separation and capture. Angew. Chem. Int. Ed. 50, 1158611596 (2011).

41. Zhou, H.-C., Long, J.R. & Yaghi, O.M. Introduction to metal-organic frameworks. Chem. Rev. 112, 673674 (2012).

42. Btard, A. & Fischer, R.A. Metal-organic framework thin films: From fundamentals to applications. Chem. Rev. 112, 10551083 (2012).

43. Li, J.-R., Sculley, J. & Zhou, H.-C. Metal-organic frameworks for separations. Chem. Rev. 112, 869932 (2012).

44. Sumida, K. etal. Carbon dioxide capture in metal-organic frameworks. Chem. Rev. 112, 724781 (2012).

45. Kreno, L.E. etal. Metal-organic framework materials as chemical sensors. Chem. Rev. 112, 11051125 (2012).

46. Rowan, S.J., Cantrill, S.J., Cousins, G.R. L., Sanders, J.K. M. & Stoddart, J.F. Dynamic covalent chemistry. Angew. Chem. Int. Ed. 41, 898952 (2002).

47. Mastalerz, M. Shape-persistent organic cage compounds by dynamic covalent bond formation. Angew. Chem. Int. Ed. 49, 50425053 (2010).

48. Jin, Y., Jin, A., McCaffrey, R., Long, H. & Zhang, W. Design strategies for shape-persistent covalent organic polyhedrons (COPs) through imine condensation/metathesis. J.Org. Chem. 77, 73927400 (2012).

49. Nishiyabu, R., Kubo, Y., James, T.D. & Fossey, J.S. Boronic acid building blocks: Tools for self assembly. Chem. Commun. 47, 11241150 (2011).

50. Hartley, C.S., Elliott, E.L. & Moore, J.S. Covalent assembly of molecular ladders. J.Am. Chem. Soc. 129, 45124513 (2007).

51. Hartley, C.S. & Moore, J.S. Programmed dynamic covalent assembly of unsymmetrical macrocycles. J.Am. Chem. Soc. 129, 1168211683 (2007).

52. Lehn, J.-M. Dynamers: Dynamic molecular and supramolecular polymers. Prog. Polym. Sci. 30, 814831 (2005).

53. Maeda, T., Otsuka, H. & Takahara, A. Dynamic covalent polymers: Reorganizable polymers with dynamic covalent bonds. Prog. Polym. Sci. 34, 581604 (2009).

54. Montarnal, D., Capelot, M., Tournilhac, F. & Leibler, L. Silica-like malleable materials from permanent organic networks. Science 334, 965968 (2011).

55. Ct, A.P. etal. Porous, crystalline, covalent organic frameworks. Science 310, 11661170 (2005).

56. Zheng, C. & Zhong, C. Estimation of framework charges in covalent organic frameworks using connectivity-based atom contribution method. J.Phys. Chem. C 114, 99459951 (2010).

57. Spitler, E.L. etal. A 2D covalent organic framework with 4.7-nm pores and insight into its interlayer stacking. J.Am. Chem. Soc. 133, 1941619421 (2011).

58. Lukose, B., Kuc, A., Frenzel, J. & Heine, T. On the reticular construction concept of covalent organic frameworks. Beilstein J.Nanotechnol. 1, 6070 (2010).

59. Lukose, B., Kuc, A. & Heine, T. The structure of layered covalent-organic frameworks. Chem. Eur. J. 17, 23882392 (2010).

60. Zhou, W., Wu, H. & Yildirim, T. Structural stability and elastic properties of prototypical covalent organic frameworks. Chem. Phys. Lett. 499, 103107 (2010).

61. Koo, B.T., Dichtel, W.R. & Clancy, P. A classification scheme for the stacking of two-dimensional boronate ester-linked covalent organic frameworks. J.Mater. Chem. 22, 1746017469 (2012).

62. Chen, X., Addicoat, M., Irle, S., Nagai, A. & Jiang, D. Control of crystallinity and porosity of covalent organic frameworks by managing interlayer interactions based on self-complementary -electronic force. J.Am. Chem. Soc. 135, 546549 (2013).

63. El-Kaderi, H.M. etal. Designed synthesis of 3D covalent organic frameworks. Science 316, 268272 (2007).

64. Hunt, J.R., Doonan, C.J., LeVangie, J.D., Ct, A.P. & Yaghi, O.M. Reticular synthesis of covalent organic borosilicate frameworks. J.Am. Chem. Soc. 130, 1187211873 (2008).

65. Bunck, D.N. & Dichtel, W.R. Internal functionalization of three-dimensional covalent organic frameworks. Angew. Chem. Int. Ed. 51, 18851889 (2012).

66. Yu, J.-T., Chen, Z., Sun, J., Huang, Z.-T. & Zheng, Q.-Y. Cyclotricatechylene based porous crystalline material: Synthesis and applications in gas storage. J.Mater. Chem. 22, 53695373 (2012).

67. Kahveci, Z., Islamoglu, T., Shar, G.A., Ding, R. & El-Kaderi, H.M. Targeted synthesis of a mesoporous triptycene-derived covalent organic framework. CrystEngComm 15, 15241527 (2013).

68. Ct, A.P., El-Kaderi, H.M., Furukawa, H., Hunt, J.R. & Yaghi, O.M. Reticular synthesis of microporous and mesoporous 2D covalent organic frameworks. J.Am. Chem. Soc. 129, 1291412915 (2007).

69. Wan, S., Guo, J., Kim, J., Ihee, H. & Jiang, D. A belt-shaped, blue luminescent, and semiconducting covalent organic framework. Angew. Chem. Int. Ed. 47, 88268830 (2008).

70. Nagai, A. etal. Pore surface engineering in covalent organic frameworks. Nature Commun. 2, 18 (2011).

71. Du, Y. etal. Experimental and computational studies of pyridine-assisted post-synthesis modified air stable covalent-organic frameworks. Chem. Commun. 48, 46064608 (2012).

72. Ding, X. etal. Synthesis of metallophthalocyanine covalent organic frameworks that exhibit high carrier mobility and photoconductivity. Angew. Chem. Int. Ed. 50, 12891293 (2011).

73. Spitler, E.L. & Dichtel, W.R. Lewis acid-catalysed formation of two-dimensional phthalocyanine covalent organic frameworks. Nature Chem. 2, 672677 (2010).




74. Ding, X. etal. An n-channel two-dimensional covalent organic framework. J.Am. Chem. Soc. 133, 1451014513 (2011).

75. Feng, X., Chen, L., Dong, Y. & Jiang, D. Porphyrin-based two-dimensional covalent organic frameworks: Synchronized synthetic control of macroscopic structures and pore parameters. Chem. Commun. 47, 19791981 (2011).

76. Tilford, R.W., Gemmill, W.R., zur Loye, H.-C. & Lavigne, J.J. Facile synthesis of a highly crystalline, covalently linked porous boronate network. Chem. Mater. 18, 52965301 (2006).

77. Tilford, R.W., Mugavero, S.J., Pellechia, P.J. & Lavigne, J.J. Tailoring microporosity in covalent organic frameworks. Adv. Mater. 20, 27412746 (2008).

78. Lanni, L.M., Tilford, R.W., Bharathy, M. & Lavigne, J.J. Enhanced hydrolytic stability of self-assembling alkylated two-dimensional covalent organic frameworks. J.Am. Chem. Soc. 133, 1397513983 (2011).

79. Spitler, E.L., Giovino, M.R., White, S.L. & Dichtel, W.R. A mechanistic study of Lewis acid-catalyzed covalent organic framework formation. Chem. Sci. 2, 15881593 (2011).

80. Feng, X., Dong, Y. & Jiang, D. Star-shaped two-dimensional covalent organic frameworks. CrystEngComm 15, 15081511 (2013).

81. Dogru, M. etal. Facile synthesis of a mesoporous benzothiadiazole-COF based on a transesterification process. CrystEngComm 15, 15001502 (2013).

82. Rabbani, M.G. etal. A 2D mesoporous imine-linked covalent organic framework for high pressure gas storage applications. Chem. Eur. J. 19, 33243328 (2013).

83. Kuhn, P., Antonietti, M. & Thomas, A. Porous, covalent triazine-based frameworks prepared by ionothermal synthesis. Angew. Chem. Int. Ed. 47, 34503453 (2008).

84. Ding, S.Y. etal. Construction of covalent organic framework for catalysis: Pd/COF-LZU1 in Suzuki-Miyaura coupling reaction. J.Am. Chem. Soc. 133, 1981619822 (2011).

85. Uribe-Romo, F.J., Doonan, C.J., Furukawa, H., Oisaki, K. & Yaghi, O.M. Crystalline covalent organic frameworks with hydrazone linkages. J.Am. Chem. Soc. 133, 1147811481 (2011).

86. Kandambeth, S. etal. Construction of crystalline 2D covalent organic frameworks with remarkable chemical (acid/base) stability via a combined reversible and irreversible route. J.Am. Chem. Soc. 134, 1952419527 (2012).

87. Jackson, K.T., Reich, T.E. & El-Kaderi, H.M. Targeted synthesis of a porous borazine-linked covalent organic framework. Chem. Commun. 48, 88238825 (2012).

88. Han, S.S., Furukawa, H., Yaghi, O.M. & Goddard, W.A. III Covalent organic frameworks as exceptional hydrogen storage materials. J.Am. Chem. Soc. 130, 1158011581 (2008).

89. Furukawa, H. & Yaghi, O.M. Storage of hydrogen, methane, and carbon dioxide in highly porous covalent organic frameworks for clean energy applications. J.Am. Chem. Soc. 131, 88758883 (2009).

90. Doonan, C.J., Tranchemontagne, D.J., Glover, T.G., Hunt, J.R. & Yaghi, O.M. Exceptional ammonia uptake by a covalent organic framework. Nature Chem. 2, 235238 (2010).

91. Benanti, T. & Venkataraman, D. Organic solar cells: An overview focusing on active layer morphology. Photosynth. Res. 87, 7381 (2006).

92. Wan, S., Guo, J., Kim, J., Ihee, H. & Jiang, D. A photoconductive covalent organic framework: Self-condensed arene cubes composed of eclipsed 2D polypyrene sheets for photocurrent generation. Angew. Chem. Int. Ed. 48, 54395442 (2009).

93. Wan, S. etal. Covalent organic frameworks with high charge carrier mobility. Chem. Mater. 23, 40944097 (2011).

94. Jin, S. etal. Charge dynamics in a donoracceptor covalent organic framework with periodically ordered bicontinuous heterojunctions. Angew. Chem. Int. Ed. 52, 20172021 (2013).

95. Dogru, M. etal. A photoconductive thienothiophene-based covalent organic framework showing charge transfer towards included fullerene. Angew. Chem. Int. Ed. 52, 29202924 (2013).

96. Patwardhan, S., Kocherzhenko, A.A., Grozema, F.C. & Siebbeles, L.D. A. Delocalization and mobility of charge carriers in covalent organic frameworks. J.Phys. Chem.C 115, 1176811772 (2011).

97. Senthilkumar, K., Grozema, F.C., Bickelhaupt, F.M. & Siebbeles, L.D.A. Charge transport in columnar stacked triphenylenes: Effects of conformational fluctuations on charge transfer integrals and site energies. J.Chem. Phys. 119, 98099817 (2003).

98. Spitler, E.L. etal. Lattice expansion of highly oriented 2D phthalocyanine covalent organic framework films. Angew. Chem. Int. Ed. 51, 26232627 (2012).

99. Dogru, M., Sonnauer, A., Gavryushin, A., Knochel, P. & Bein, T. A covalent organic framework with 4nm open pores. Chem. Commun. 47, 17071709 (2011).

100. Feng, X. etal. High-rate charge-carrier transport in porphyrin covalent organic frameworks: Switching from hole to electron to ambipolar conduction. Angew. Chem. Int. Ed. 51, 26182622 (2012).

101. Feng, X. etal. An ambipolar conducting covalent organic framework with self-sorted and periodic electron donor-acceptor ordering. Adv. Mater. 24, 30263031 (2012).

102. Colson, J.W. etal. Oriented 2D covalent organic framework thin films on single-layer graphene. Science 332, 228231 (2011).

103. Pang, S., Hernandez, Y., Feng, X. & Mllen, K. Graphene as transparent electrode material for organic electronics. Adv. Mater. 23, 27792795 (2011).

104. Berlanga, I. etal. Delamination of layered covalent organic frameworks. Small 7, 12071211 (2011).

105. Clair, S., Ourdjini, O., Abel, M. & Porte, L. Two-dimensional polymer as a mask for surface nanopatterning. Adv. Mater. 24, 12521254 (2012).

106. Gutzler, R. etal. Surface mediated synthesis of 2D covalent organic frameworks: 1,3,5-tris(4-bromophenyl)benzene on graphite(001), Cu(111), and Ag(110). Chem. Commun. 45, 44564458 (2009).

107. Ourdjini, O. etal. Substrate-mediated ordering and defect analysis of a surface covalent organic framework. Phys. Rev. B 84, 125421125429 (2011).

108. Marele, A.C. etal. Formation of a surface covalent organic framework based on polyester condensation. Chem. Commun. 48, 67796781 (2012).

109. Faury, T. etal. Sequential linking to control growth of a surface covalent organic framework. J.Phys. Chem. C 116, 48194823 (2012).

110. Dienstmaier, J.F. etal. Isoreticular two-dimensional covalent organic frameworks synthesized by on-surface condensation of diboronic acids. ACS Nano 6, 72347242 (2012).

111. Guan, C.Z., Wang, D. & Wan, L.J. Construction and repair of highly ordered 2D covalent networks by chemical equilibrium regulation. Chem. Commun. 48, 29432945 (2012).

112. Slater, A.G., Beton, P.H. & Champness, N.R. Two-dimensional supramolecular chemistry on surfaces. Chem. Sci. 2, 14401448 (2011).

113. Lipton-Duffin, J.A., Ivasenko, O., Perepichka, D.F. & Rosei, F. Synthesis of polyphenylene molecular wires by surface-confined polymerization. Small 5, 592597 (2009).

114. Lipton-Duffin, J.A. etal. Step-by-step growth of epitaxially aligned polythiophene by surface-confined reaction. Proc. Natl Acad. Sci. USA 107, 1120011204 (2010).

115. Lafferentz, L. etal. Conductance of a single conjugated polymer as a continuous function of its length. Science 323, 11931197 (2009).

116. Cai, J. etal. Atomically precise bottom-up fabrication of graphene nanoribbons. Nature 466, 470473 (2010).

117. Abel, M., Clair, S., Ourdjini, O., Mossoyan, M. & Porte, L. Single layer of polymeric Fe-phthalocyanine: An organometallic sheet on metal and thin insulating film. J.Am. Chem. Soc. 133, 12031205 (2011).

118. Bieri, M. etal. Two-dimensional polymer formation on surfaces: Insight into the roles of precursor mobility and reactivity. J.Am. Chem. Soc. 132, 1666916676 (2010).

119. Treier, M., Richardson, N.V. & Fasel, R. Fabrication of surface-supported low-dimensional polyimide networks. J.Am. Chem. Soc. 130, 1405414055 (2008).

120. Weigelt, S. etal. Surface synthesis of 2D branched polymer nanostructures. Angew. Chem. Int. Ed. 47, 44064410 (2008).

121. Jensen, S., Frchtl, H. & Baddeley, C.J. Coupling of triamines with diisocyanates on Au(111) leads to the formation of polyurea networks. J.Am. Chem. Soc. 131, 1670616713 (2009).

122. Blunt, M.O., Russell, J.C., Champness, N.R. & Beton, P.H. Templating molecular adsorption using a covalent organic framework. Chem. Commun. 46, 71577159 (2010).

123. Ozaki, H. etal. Formation of atomic cloth observed by penning ionization electron spectroscopy. J.Chem. Phys. 103, 12261228 (1995).

124. Takami, T. etal. Periodic structure of a single sheet of a clothlike macromolecule (atomic cloth) studied by scanning tunneling microscopy. Angew. Chem. Int. Ed. Engl. 36, 27552757 (1997).

125. Miura, A. etal. Light- and STM-tip-induced formation of one-dimensional and two-dimensional organic nanostructures. Langmuir 19, 64746482 (2003).

126. Okawa, Y. & Aono, M. Nanoscale control of chain polymerization. Nature 409, 683684 (2001).

AcknowledgementsThe authors acknowledge financial support from the National Science Foundation (CHE-1056657and CHE-1124754), a Sloan Research Fellowship, the Arnold and Mabel Beckman Foundation, and the Research Corporation for Science Advancement. J.W.C. acknowledges the National Science Foundation for a Graduate Research Fellowship.

Additional InformationReprints and permissions information is available online at www.nature.com/reprints. Correspondence should be addressed to W.R.D.

Competing financial interestsThe authors declare no competing financial interests.


http://www.nature.com/reprintshttp://www.nature.com/doifinder/10.1038/nchem.1628Rationally synthesized two-dimensional polymersSynthesis of 2D polymers under thermodynamic controlThermodynamic and kinetic surface-supported synthesesSurface-mediated molecular networks and polymerizations Reactions in organized 2D assembliesConclusion and outlookFigure 1 | Schematic illustration of the structures of one- and two-dimensional polymers, and strategies for 2D polymerizationFigure 2 | Thermodynamic synthesis of COFsFigure 3 | Computational studies indicate that COFs adopt offset layered structuresFigure 4 | 2D COFs exhibit photoconductivity and variations in charge transport as a function of their structureFigure 5 | COF thin films supported on single-layer grapheneFigure 6 | Synthesis of surface-supported single-layer 2D polymersFigure 7 | Two-dimensional polymers prepared by surface-catalysed arylaryl couplingFigure 8 | Two distinct polymerizable groups A and B were incorporated into a mesogenFigure 9 | Irradiation of alkyne-containing monomers induces simultaneous topochemical polymerizations that yield a 2D polymer comprising linear polyacetylene and polydiacetylene chains linked by alkanesFigure 10 | Topochemical polymerization in a layered crystal provides a multilayer 2D polymerReferencesAcknowledgementsAdditional InformationCompeting financial interests

nchem.1628

Documents