doi.org/10.26434/chemrxiv.11586510.v1

Single Crystal of a One-dimensional Covalent Organic FrameworkHaisen Xu, Yi Luo, Xing Li, Pei Zhen See, Zhongxin Chen, Tianqiong Ma, Lin Liang, Kai Leng, IbrahimAbdelwahab, Lin Wang, Runlai Li, Xiangyan Shi, Yi Zhou, Xiu Fang Lu, Xiaoxu Zhao, Cuibo Liu, Junliang Sun,Kian Ping Loh

Submitted date: 12/01/2020 • Posted date: 13/01/2020Licence: CC BY-NC-ND 4.0Citation information: Xu, Haisen; Luo, Yi; Li, Xing; See, Pei Zhen; Chen, Zhongxin; Ma, Tianqiong; et al.(2020): Single Crystal of a One-dimensional Covalent Organic Framework. ChemRxiv. Preprint.https://doi.org/10.26434/chemrxiv.11586510.v1

Although polymers have been studied for well over a century, there are few examples of covalently linkedpolymer crystals synthesized directly from solution. One-dimensional (1D) covalent polymers that are packedinto a framework structure can be viewed as a 1D covalent organic framework (COF), but making singlecrystal of this has been elusive. Herein, by combining labile metal coordination and dynamic covalentchemistry, we discovered a strategy to synthesize single-crystal metallo-COF under solvothermal conditions.The single-crystal structure was rigorously solved using single-crystal electron diffraction (SCED) technique.The non-centrosymmetric metallo-COF allows second harmonic generation (SHG). Due to the presence ofsyntactic pendant amine groups along the polymer chains, the metallopolymer crystal can be furthercross-linked into a crystalline woven network.

File list (3)

download fileview on ChemRxivSingle-crystal 1D COF.pdf (1.14 MiB)

download fileview on ChemRxivSupplementary_Materials.pdf (2.46 MiB)

download fileview on ChemRxivmCOF-Ag.cif (6.16 KiB)

1

Single Crystal of a One-dimensional Covalent Organic Framework

Hai-Sen Xu,1† Yi Luo,2,3† Xing Li,1† Pei Zhen See,1 Zhongxin Chen,1 Tianqiong Ma,2 Lin Liang,4

Kai Leng,1 Ibrahim Abdelwahab,1 Lin Wang,1 Runlai Li,1 Xiangyan Shi,5 Yi Zhou,6 Xiu Fang Lu,1

Xiaoxu Zhao,1 Cuibo Liu,1 Junliang Sun,2* Kian Ping Loh1*

1Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 5

117543, Singapore.

2College of Chemistry and Molecular Engineering, Beijing National Laboratory for Molecular

Sciences, Peking University, Beijing 100871, China.

3Department of Materials and Environmental Chemistry Stockholm University, SE-10691

Stockholm (Sweden). 10

4State Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical

Engineering, Lanzhou University, Lanzhou, Gansu 730000, China.

5School of Physical and Mathematical Sciences, Nanyang Technological University, 21 Nanyang

Link, Singapore 637371.

6School of Physical Science and Technology, Shanghai Tech University, Shanghai 201210, China. 15

† These authors contributed equally to this work.

*Correspondence to: junliang.sun@pku.edu.cn (J.S.); chmlohkp@nus.edu.sg (K.P.L.)

Abstract: Although polymers have been studied for well over a century, there are few examples

of covalently linked polymer crystals synthesized directly from solution. One-dimensional (1D)

covalent polymers that are packed into a framework structure can be viewed as a 1D covalent 20

organic framework (COF), but making single crystal of this has been elusive. Herein, by

combining labile metal coordination and dynamic covalent chemistry, we discovered a strategy to

synthesize single-crystal metallo-COF under solvothermal conditions. The single-crystal structure

was rigorously solved using single-crystal electron diffraction (SCED) technique. The non-

centrosymmetric metallo-COF allows second harmonic generation (SHG). Due to the presence of 25

syntactic pendant amine groups along the polymer chains, the metallopolymer crystal can be

further cross-linked into a crystalline woven network.

Main Text: Framework solids have been one of the hottest materials over the past 30 years

because they represent mankind’s attempt to control chemical bonding in space versus random

polymerisation1,2. Depending on the nature of the strongest bond (coordination bond, covalent 30

bond, hydrogen bond, etc.) used in constructing the solids, framework materials are categorized

into metal-organic frameworks (MOFs)3-5, covalent organic frameworks (COFs)6-10, and

hydrogen-bonded organic frameworks (HOFs)11,12. The synthesis of framework materials needs to

2

be carried out under conditions where bond formation is highly reversible to facilitate the self-

correction process necessary for crystal growth. In this regard, COFs are among the most difficult

to crystallize, owing to the lesser reversibility of their covalent linkages compared to coordination

bonds and hydrogen bonds in MOFs and HOFs. The ease of encoding functionalities in COFs and

their structural robustness render them potentially useful in wide-ranging applications13-19. 5

However, an in depth understanding of the structure-property correlation in COFs is lacking,

owing to fact that most synthesized COFs are polycrystalline, which hampers structural

determination. Recently, the addition of crystal seeds or modulators have been used to grow single-

crystalline two- or three-dimensional (2D or 3D) COFs20,21. Nevertheless, only a few examples of

3D COFs have their crystal structure rigorously solved21-23. It is generally recognized that single 10

crystalline frameworks of lower dimensionality are more difficult to grow compared to higher

dimensionality ones24.

Although the vast choices of COF building units and covalent linkages give rise to diverse

structural motifs in the synthesized product, the produced 2D and 3D COFs tend to crystallize in

high symmetric space groups due to the high symmetries of the building blocks and low freedom 15

of intermolecular packing8,9. As such, conventionally synthesized COFs are mostly

centrosymmetric and do not exhibit second harmonic generation or ferroelectricity25. In this regard,

one dimensional (1D) COFs, which possess a high degree of freedom in molecular packing, can

be a candidate to construct non-centrosymmetric crystals. Conceptually, the framework structure

in 1D COF is constructed from 1D-confined covalent linkages, and non-covalent interactions (such 20

as π-π interactions, hydrogen bonding, etc.) in the other two dimensions help to pack the 1D chains.

However, the extremely high anisotropy and the entropy-driven random packing of organic chains

impose a huge challenge on the synthesis of 1D COFs. Using its covalent analogue, linear

polymers, as examples: since the invention of the first synthetic polymer (Bakelite) in 1907, nearly

all the polymers synthesized are amorphous or poorly crystalline. To obtain a well-defined 1D 25

COF, two basic aspects have to be considered: (i) How to get a framework, not just a densely

packed organic polymer; (ii) how to control the periodic packing of organic chains to get a single

crystalline structure. Although topochemical polymerisation approach has been used to produce

single crystalline polymers from pre-packed molecular crystals, the approach is rather limited in

scope26-28. It is highly desirable to search for a direct crystallization method from solution, which 30

has been so far elusive.

Herein, we demonstrate a strategy to synthesize single crystalline 1D metallo-COFs by

combining metal-ligand coordination29 and dynamic covalent chemistry (DCC)30 (Fig. 1). A one-

pot reaction combining self-assembly31,32 and imine condensation was conducted under

solvothermal conditions. Along with the poly-condensation reaction, the constituents are 35

consumed and regenerated from the reactant “pool”33. An advantage of this method is that building

blocks can evolve into multi-intermediates through reversible reactions or interactions without the

need to synthesize each building block individually, which is distinct from the conventional

methods of making crystalline organic networks using predetermined building blocks34,35. We first

attempt to construct 1D conjugated polymer solely based on DCC without a metal template. Due 40

3

to the absence of secondary interactions, the 1D polymer chains pack randomly to afford a poorly

crystalline polymer (Supplementary Fig. 1). We then introduce AgBF436 to initiate ligand exchange

and provide an additional reversible process for the poly-condensation, which should be beneficial

for the crystal nucleation process. When one of the building units, 4,4'-(1,10-phenanthroline-2,9-

diyl)dianiline (I), is in excessive amount, it can be anchored onto the backbone chains via AgI 5

coordination in a syntactic fashion, thus providing periodic spacers to induce π-π stacking and

hydrogen bonding between the 1D chains (Fig.1). Using this approach, we synthesized single

crystalline 1D metallo-COF (mCOF-Ag) with micrometer-sized particles, allowing the rigorous

characterisation of the crystal structure via single-crystal electron diffraction (SCED)37.

In line with the strategy, a one-pot reaction by combining self-assembly and imine 10

condensation was conducted under solvothermal conditions. In 2,9-bis(4-

(dimethoxymethyl)phenyl)-1,10-phenanthroline (II), an acetal group38 is preferred for its good

solubility which is beneficial for the self-assembly process; besides, the relatively low reactivity

of acetal with amine groups can reduce the crystal nucleation rate to enable good crystals. Once

AgBF4 was added into the suspension of I and II, an immediate colour change was observed 15

indicating the occurrence of self-assembly via coordination. The amount of AgBF4 added has a

strong influence on crystal qualities. According to the strategy, the stoichiometry of I, II, and

AgBF4 is 3:1:2. However, adopting this ratio produces a large number of Ag particles on the

material. The presence of Ag particles is attributed to the reversible nature of imine bond; in situ

formed aldehyde groups can reduce Ag ions into elemental Ag, which then aggregated into 20

particles. When the ratio of AgBF4 is reduced to 1, no Ag particles were detected under the same

reaction conditions. The use of excess I and II can protect Ag ions from being reduced. After

carefully screening the reaction conditions, mCOF-Ag was obtained in a mixture of 1-butanol, 1,2-

dichlorobenzene, and 6 M aqueous acetic acid (1/9/1, v/v/v) at 120 °C for three days with a feeding

ratio of I, II, and AgBF4 equals to 3:1:1. After copiously washing with DMSO and THF, and 25

drying at 100 °C for 8 hours, mCOF-Ag was obtained as a light yellow solid (74% yield, based on

AgBF4).

The chemical bonding in mCOF-Ag was assessed by Fourier-transform infrared

spectroscopy (FT-IR) (Supplementary Figs. 3 and 4) and 13C cross-polarization magic-angle

spinning (CP/MAS) NMR (Supplementary Fig. 5). mCOF-Ag exhibits sharp powder X-ray 30

diffraction (PXRD) peaks, indicating its high crystallinity (Supplementary Fig. 2). Scanning

electron microscopy (SEM) image reveals that mCOF-Ag exhibits a uniform rod-like morphology

with the crystal size larger than 2 μm (Fig. 2b). Ordered lattice fringes in the high-resolution

transmission electron microscopy (HRTEM) image (Fig. 2c) and selected-area electron diffraction

(SAED) pattern (Fig. 2c, inset) confirm its single-crystal nature. 35

The single-crystal structure of mCOF-Ag is determined directly from the SCED data39,40

(Fig. 2a). From the observed electron density maps (Fig. 2d and Supplementary Fig. 8), all

positions of the non-hydrogen atoms (C, N, and Ag) on the framework structure and the locations

of the guests (BF4‾ anions) are identified. The SCED data was collected on a typical rod-like

4

crystal, with a resolution of ~0.95 Å, which reflects the single-crystal nature of mCOF-Ag. A C-

centered monoclinic cell with a = 15.66 Å, b = 31.00 Å, c = 10.87 Å, and β = 123.31° is identified

by indexing the SCED data using the program XDS (Supplementary Table S1). As indicated by

the reflection conditions (hkl: h+k; 0kl: k = 2n; hk0: h+k; h00: h = 2n; and 0k0: k = 2n), the

possible space groups are deduced to be C2/c (No.15), Cc (No. 9), C2/m (No. 12), Cm (No. 8), or 5

C2 (No. 5). Based on this SCED data, the structure model of mCOF-Ag was solved directly in the

space group C2/c using the program SHELXT (dual-space method). In the determined structure

model, it is found that the C and N atoms on the connection bonds (–C=N–) are symmetry related.

In order to make them symmetry independent, the symmetry of the structure model was reduced

into its subgroup C2. The determined structure model of mCOF-Ag was further confirmed and 10

refined against its synchrotron powder X-ray diffraction (SPXD) data (Supplementary Fig. 11).

Using the SVD-index method implemented in the software Topas, the SPXD data of mCOF-Ag is

indexed by a C-centered monoclinic cell of a = 15.83 Å, b = 29.97 Å, c = 10.69 Å, and β = 123.96°,

which is consistent with the unit cell obtained from the SCED data. The final agreement residuals

for the Rietveld refinement are RI = 0.075, Rwp = 0.108, with Rexp = 0.049. 15

Further analysis of the crystal structure reveals more insights about the packing of the 1D

chains. The adjacent phenanthroline chains are arranged in a parallel fashion to generate a 2D

corrugated layer (Supplementary Fig. 10), and the 2D layers are further stabilized by the

interactions of amines with BF4‾ anions (Supplementary Fig. 9), as well as the π-π stacking of

interlayer phenanthroline rings (Supplementary Fig. 10), to form a permanent structure (Fig. 3a). 20

The Ag and pendant groups I are arranged in a syntactic version, favouring ordered packing of the

polymer chains. mCOF-Ag belongs to the C2 space group, which is non-centrosymmetric. This is

validated by second harmonic generation (SHG)41 at 425 nm when the crystal is excited by

fundamental laser wavelength of 850 nm (Fig. 3b and Supplementary Fig. 16). Polarization-

dependent SHG was recorded on an isolated crystal. Using polarized laser excitation, a SHG map 25

of the COF-Ag crystal could be obtained (Fig. 3b, inset). The strongest SHG response was

observed under parallel polarized excitation, this means that the maximum SHG response

originates from the longitudinal direction of the mCOF-Ag crystal.

Due to the packing of zigzag chains, the orientation and rotation of amine groups in

building block I are spatially well defined and the two diagonally positioned –NH2 groups are 30

separated by ~3.9 Å (Fig. 4a). mCOF-Ag is an ideal scaffold for crystalline state polymerisation

because the adjacent amine groups can be linked by bi-functional compounds. Glyoxal, a molecule

with two aldehyde groups, was chosen for its suitable molecular size and good reactivity with

amine groups. mCOF-Ag was reacted with glyoxal solution (40 wt. % in H2O) at 70 °C in 1,4-

dioxane. After incubation for five days and washing, the product was collected and dried to yield 35

a cross-linked woven network, termed as wCOF-Ag (Fig. 4a). The crystalline-state polymerisation,

which occurs via “stitching” of the inter-chain amine groups by the aldehyde molecules, was

verified using a range of techniques. In the FT-IR spectra, the characteristic bands at 3361 cm-1,

3301 cm-1, and 3197 cm-1 for N-H stretching, and at 1618 cm-1 for N-H bending, have almost

vanished after cross-linking, indicating the near-complete consumption of amine groups in mCOF-40

5

Ag (Supplementary Fig. 17). The polymerisation process was further verified by the CP/MAS

NMR spectrum of wCOF-Ag, where the peak corresponding to the aldehyde carbon in glyoxal at

190 ppm is absent (Supplementary Fig. 18), indicating the full connection of amine and aldehyde

groups. The PXRD pattern appears very similar with mCOF-Ag except for slight differences at

high angles (Fig. 4b). These results affirm that cross-linking process did not affect the major part 5

of the crystal structure. One evidence that the cross-linked solid is different from the pristine

structure is provided by the photoluminescence (PL) spectroscopy, where the PL of wCOF-Ag

showed a dramatic six-fold enhancement in intensity compared to that of mCOF-Ag

(Supplementary Fig. 23). We have also examined the mechanical properties of the crystals before

and after cross-linking using nano-indentation. By indenting on an isolated micron-sized crystal 10

of each of these two samples, the Young’s moduli of mCOF-Ag and wCOF-Ag were determined

to be 9.0 (Supplementary Fig. 24) and 19.1 GPa (Supplementary Fig. 25), respectively (Fig. 4c).

The value of wCOF-Ag is comparable with that of COF-505 (~12.5 GPa)34. The distinct increase

in Young’s modulus of wCOF-Ag is in line with its covalently connected structure in a 3D space,

which imbues greater rigidity on the resulting material. 15

In this study, by combining metal-ligand coordination and dynamic imine bond formation,

a single crystalline material, mCOF-Ag, was successfully constructed via solvothermal conditions;

its crystal structure was rigorously solved with SCED. Due to its non-centrosymmetric structure,

mCOF-Ag shows an obvious SHG signal, demonstrating its potential as nonlinear optical

materials. Moreover, due to the presence of interlaced pendant amine groups along the polymer 20

chains, a high degree of control on the polymer backbone is obtained by crystalline-state

polymerisation to form a woven network. From a synthetic perspective, the strategy we developed

combines the powerful templating effect of metal ions and cooperative organization of building

blocks to maximize the number of secondary bonds in solution, thus it can be a promising method

for the solution-phase synthesis of single-crystal covalent metallopolymers with unique topologies 25

and functionalities.

6

Fig. 1 | Strategy of synthesizing single crystalline 1D metallo-COF. Colour scheme: C on

organic chains, blue; C on pendants, grey; N, purple; Ag, yellow; B, pink; F, green. Hydrogen

atoms are omitted for clarity.

7

Fig. 2 | Structure characterisation and analysis of mCOF-Ag. a, 3D reciprocal lattice of mCOF-

Ag reconstructed from the SCED data (left) and 2D slices cut from the reconstructed reciprocal

lattice (right). b, SEM image of mCOF-Ag with uniform rod-like morphology. Scale bar: 1 µm. c,

HRTEM image of mCOF-Ag and SAED pattern (inset) confirmed the single crystalline nature of 5

the material. Scale bar: 20 nm. d, Observed electron density map of the initial structure model

determined from SCED data along the c-axis.

8

Fig. 3 | Single-crystal structure and nonlinear optical measurements of mCOF-Ag. a, Crystal

view of mCOF-Ag along the c-axis. Colour scheme: C on organic chains, blue; C on pendants,

grey; N, purple; Ag, yellow. Hydrogen atoms and BF4‾ anions are omitted for clarity. b, The SHG

spectrum of an isolated crystal of mCOF-Ag. The excitation wavelength is 850 nm, the peak at 5

425 nm is the SHG signal. Inset: polarization dependent SHG response. Red curve is the fitting

result of experimental data (blue dots). The units are in degree.

9

Fig. 4 | Crystalline-state polymerisation and characterisation. a, Confined environments of

mCOF-Ag (left) and the ideal structure of the cross-linked framework, wCOF-Ag (right). One of

the cross-linked parts is highlighted with a red circle. Colour scheme: C on organic chains, blue;

C on pendants, grey; N, purple; Ag, yellow. Hydrogen atoms and BF4‾ anions are omitted for 5

clarity. b, PXRD patterns of mCOF-Ag and wCOF-Ag. c, The Young’s moduli of mCOF-Ag and

wCOF-Ag.

10

References:

1. Yaghi, O. M., O'Keeffe, M. & Kanatzidis, M. Design of solids from molecular building blocks:

Golden opportunities for solid state chemistry. J. Solid State Chem. 152, 1-2 (2000).

2. Yaghi, O. M. et al. Reticular synthesis and the design of new materials. Nature 423, 705-714

(2003). 5

3. Zhou, H.-C., Long, J. R. & Yaghi, O. M. Introduction to metal–organic frameworks. Chem. Rev.

112, 673-674 (2012).

4. Kitagawa, S., Kitaura, R. & Noro, S.-i. Functional porous coordination polymers. Angew. Chem.

Int. Ed. 43, 2334-2375 (2004).

5. Furukawa, H., Cordova, K. E., O’Keeffe, M. & Yaghi, O. M. The chemistry and applications of 10

metal-organic frameworks. Science 341, 1230444 (2013).

6. Cote, A. P. et al. Porous, crystalline, covalent organic frameworks. Science 310, 1166-1170

(2005).

7. Diercks, C. S. & Yaghi, O. M. The atom, the molecule, and the covalent organic framework.

Science 355, eaal1585 (2017). 15

8. Huang, N., Wang, P. & Jiang, D. Covalent organic frameworks: A materials platform for

structural and functional designs. Nat. Rev. Mater. 1, 16068 (2016).

9. Ding, S.-Y. & Wang, W. Covalent organic frameworks (COFs): From design to applications.

Chem. Soc. Rev. 42, 548-568 (2013).

10. Segura, J. L., Mancheño, M. J. & Zamora, F. Covalent organic frameworks based on schiff-20

base chemistry: Synthesis, properties and potential applications. Chem. Soc. Rev. 45, 5635-5671

(2016).

11. Brunet, P., Simard, M. & Wuest, J. D. Molecular tectonics. Porous hydrogen-bonded networks

with unprecedented structural integrity. J. Am. Chem. Soc. 119, 2737-2738 (1997).

12. He, Y., Xiang, S. & Chen, B. A microporous hydrogen-bonded organic framework for highly 25

selective C2H2/C2H4 separation at ambient temperature. J. Am. Chem. Soc. 133, 14570-14573

(2011).

13. Lin, S. et al. Covalent organic frameworks comprising cobalt porphyrins for catalytic CO2

reduction in water. Science 349, 1208-1213 (2015).

14. Xu, H., Gao, J. & Jiang, D. Stable, crystalline, porous, covalent organic frameworks as a 30

platform for chiral organocatalysts. Nat. Chem. 7, 905-912 (2015).

15. DeBlase, C. R. et al. β-ketoenamine-linked covalent organic frameworks capable of

pseudocapacitive energy storage. J. Am. Chem. Soc. 135, 16821-16824 (2013).

16. Li, X. et al. Covalent-organic-framework-based Li–CO2 batteries. Adv. Mater. 0, 1905879

(2019). 35

17. Guan, X. et al. Chemically stable polyarylether-based covalent organic frameworks. Nat.

Chem. 11, 587-594 (2019).

18. Li, X. et al. Tuneable near white-emissive two-dimensional covalent organic frameworks. Nat.

Commun. 9, 2335 (2018).

11

19. Dogru, M. et al. A photoconductive thienothiophene-based covalent organic framework

showing charge transfer towards included fullerene. Angew. Chem. Int. Ed. 52, 2920-2924 (2013).

20. Evans, A. M. et al. Seeded growth of single-crystal two-dimensional covalent organic

frameworks. Science 361, 52-57 (2018).

21. Ma, T. et al. Single-crystal x-ray diffraction structures of covalent organic frameworks. Science 5

361, 48-52 (2018).

22. Beaudoin, D., Maris, T. & Wuest, J. D. Constructing monocrystalline covalent organic

networks by polymerization. Nat. Chem. 5, 830 (2013).

23. Sun, T. et al. Atomic-level characterization of dynamics of a 3D covalent organic framework

by cryo-electron diffraction tomography. J. Am. Chem. Soc. 141, 10962-10966 (2019). 10

24. Navarro, J. A. R. The dynamic art of growing COF crystals. Science 361, 35-35 (2018).

25. Ok, K. M., Chi, E. O. & Halasyamani, P. S. Bulk characterization methods for non-

centrosymmetric materials: Second-harmonic generation, piezoelectricity, pyroelectricity, and

ferroelectricity. Chem. Soc. Rev. 35, 710-717 (2006).

26. Dou, L. et al. Single-crystal linear polymers through visible light–triggered topochemical 15

quantitative polymerization. Science 343, 272-277 (2014).

27. Lauher, J. W., Fowler, F. W. & Goroff, N. S. Single-crystal-to-single-crystal topochemical

polymerizations by design. Acc. Chem. Res. 41, 1215-1229 (2008).

28. Kissel, P. et al. A two-dimensional polymer prepared by organic synthesis. Nat. Chem. 4, 287-

291 (2012). 20

29. Lewis, J. E. M., Beer, P. D., Loeb, S. J. & Goldup, S. M. Metal ions in the synthesis of

interlocked molecules and materials. Chem. Soc. Rev. 46, 2577-2591 (2017).

30. Rowan, S. J. et al. Dynamic covalent chemistry. Angew. Chem. Int. Ed. 41, 898-952 (2002).

31. Gil-Ramírez, G., Leigh, D. A. & Stephens, A. J. Catenanes: Fifty years of molecular links.

Angew. Chem. Int. Ed. 54, 6110-6150 (2015). 25

32. Nitschke, J. R. Construction, substitution, and sorting of metallo-organic structures via

subcomponent self-assembly. Acc. Chem. Res. 40, 103-112 (2007).

33. Corbett, P. T. et al. Dynamic combinatorial chemistry. Chem. Rev. 106, 3652-3711 (2006).

34. Liu, Y. et al. Weaving of organic threads into a crystalline covalent organic framework. Science

351, 365-369 (2016). 30

35. Liu, Y. et al. 3D covalent organic frameworks of interlocking 1D square ribbons. J. Am. Chem.

Soc. 141, 677-683 (2019).

36. Khlobystov, A. N. et al. Supramolecular design of one-dimensional coordination polymers

based on silver(I) complexes of aromatic nitrogen-donor ligands. Coord. Chem. Rev. 222, 155-192

(2001). 35

37. Wang, Y. et al. Elucidation of the elusive structure and formula of the active pharmaceutical

ingredient bismuth subgallate by continuous rotation electron diffraction. Chem. Commun. 53,

7018-7021 (2017).

38. Li, Z.-J. et al. Synthesis of -C=N- linked covalent organic frameworks via the direct

condensation of acetals and amines. Chem. Commun. 52, 7217-7220 (2016). 40

12

39. Ma, T. et al. Observation of interpenetration isomerism in covalent organic frameworks. J. Am.

Chem. Soc. 140, 6763-6766 (2018).

40. Gao, C. et al. Isostructural three-dimensional covalent organic frameworks. Angew. Chem. Int.

Ed. 58, 9770-9775 (2019).

41. EATON, D. F. Nonlinear optical materials. Science 253, 281-287 (1991). 5

Acknowledgments

K.P.L. acknowledges NRF-CRP grant “Two Dimensional Covalent Organic Framework:

Synthesis and Applications”. Grant number NRF-CRP16-2015-02, funded by National Research

Foundation, Prime Minister’s Office, Singapore. J.S. and Y.L. acknowledge National Natural

Science Foundation of China and Swedish Research Council. H.-S.X. and K.P.L. thank Q. Zhang 10

(National University of Singapore) for help with PL tests, Z. Zhu, X. Wu, Q. Guo (National

University of Singapore) for help with SHG tests, and L. Zhang, P. Chen (Peking University) for

help with single-crystal structure analysis.

Author contributions

K.P.L. supervised the project. J.S. supervised the crystal structures analysis. H.-S.X. designed and 15

performed the experiments. Y.L. conducted SCED characterisation and Rietveld Refinement.

X.L., Z.C., T.M., L.L., K.L., X.F.L. and C.L. discussed the synthesis and characterisation. I.A.

conducted the SHG test. L.W. conducted the AFM test. R.L. helped to analyze the nano-

indentation data. X.S. conducted CP/MAS NMR characterisation. Y.Z. conducted HRTEM

characterisation. X.Z. helped to analyse the HRTEM data. K.P.L., H.-S.X., X.L., P.Z.S. and Y.L. 20

wrote the manuscript.

Competing interests

The authors declare no competing interests.

download fileview on ChemRxivSingle-crystal 1D COF.pdf (1.14 MiB)

1

Supplementary Materials for

Single Crystal of a One-dimensional Covalent Organic Framework

Hai-Sen Xu,† Yi Luo,† Xing Li,† Pei Zhen See, Zhongxin Chen, Tianqiong Ma, Lin Liang, Kai

Leng, Ibrahim Abdelwahab, Lin Wang, Runlai Li, Xiangyan Shi, Yi Zhou, Xiu Fang Lu, Xiaoxu

Zhao, Cuibo Liu, Junliang Sun,* Kian Ping Loh*

†These authors contributed equally to this work.

*Correspondence to: junliang.sun@pku.edu.cn (J.S.); chmlohkp@nus.edu.sg (K.P.L.)

This PDF file includes:

Materials and Methods

Supplementary Text

Supplementary Figures 1-25

Supplementary Table 1

Supplementary References

2

Materials and Methods

Reagent

All commercially available starting compounds and solvents were purchased from commercial

sources and used without further purification. 2-(4-(dimethoxymethyl)phenyl)-4,4,5,5-

tetramethyl-1,3,2-dioxaborolane was synthesized through a two-step reaction as reported in

literature1.

Instrumentation and characterization

Liquid 1H and 13C NMR spectra were recorded on a Bruker AVANCE I 300 MHz or a Bruker

AVANCE I 500 MHz NMR spectrometer and chemical shifts (δ-scale) were reported in parts per

million (ppm). Solid-state NMR experiments were performed on a Bruker Avance III HD 600

MHz wide-bore NMR spectrometer. The 13C cross-polarization magic-angle spinning (CP/MAS)

NMR spectra were recorded with a 4 mm double-resonance MAS probe. High-resolution mass

spectrometry (HRMS) data were obtained on a Finnigan/MAT 95XL-T spectrometer. Fourier-

transform infrared spectroscopy (FTIR) was carried out on a Bruker VERTEX 80V spectrometer

in transmission mode under vacuum. Photoluminescence (PL) spectra were obtained with a laser

confocal microscope (NT-MDT, NTEGR Spectra) with 532 nm (continuous wave (cw)) excitation

lasers under ambient conditions. H, C and N contents were determined via elemental analysis

performed on an Elementar vario MICRO cube. Thermogravimetric analysis (TGA) was carried

out on a Discovery TGA within the temperature range of 30 °C to 800 °C with a heating rate of 10

°C/min.

Powder X-ray diffraction (PXRD) data were collected on a Bruker D8 Focus Powder X-ray

Diffractometer using Cu Kα radiation (40 kV, 40 mA) at room temperature. Synchrotron powder

X-ray diffraction (SPXD) data were collected on the sample in a 0.5 mm capillary on the BL14B1

beamline (λ = 0.69005 Å) at the Shanghai Synchrotron Radiation Facility in Shanghai, China. The

collected data were ranging from 2.0 to 20.0° with 0.004° data binning.

Scanning electron microscopy (SEM) images were obtained with a JEOL JSM-6701F microscope.

A JEM-2100Plus (LaB6 filament) operated at 200 kV was used for high resolution transmission

electron microscopy (HRTEM) imaging. Single-crystal electron diffraction (SCED) data were

collected using a JEOL JEM2100 TEM (LaB6 filament) operating at 200 kV. The reciprocal space

reconstruction was carried out using the RED software, and the reflection intensity extraction was

conducted by the program XDS2,3.

For the second harmonic generation (SHG) characterization, a Yb:KGW femtosecond PHAROS

laser system was used as the pump of a collinear optical parametric amplifier ORPHEUS with a

LYRA wavelength extension option (Light Conversion Ltd, pulse duration of 150 fs, tunable

excitation wavelength of 310 nm – 2700 nm, and repetition rate of 100 Hz –100 kHz). The laser

beam was focused onto the COF samples with a 100x (NA = 0.9) air objective from Nikon (1 μm2

spot size). The nonlinear emission was collected in a back-scattering configuration via the same

3

objective and detected by a spectrograph (PI Acton SP2300 by Princeton Instruments) for spectral

measurements. The power of the collected SHG emission was measured using a calibrated silicon

photodetector (Newport), while the excitation power was measured using a germanium

photodetector (Thorlabs). For polarization dependent SHG measurements, a half-wave plate was

used to vary the orientation of the linearly polarized laser pulses.

X-ray photoelectron spectroscopy (XPS) was performed using PHI Quantera II with a

monochromatic Al-Kα source (energy = 1486.6 eV).

Atomic force microscopy (AFM) nano-indentation experiments were performed using the

Dimension Icon instrument operating under the indentation mode, equipped with a Bruker

Tap525A probe. Its spring constant and contact sensitivity have been calibrated as 156.760 N/m

and 83.01 nm/volt, respectively.

4

Supplementary Text

Synthesis of 4,4'-(1,10-phenanthroline-2,9-diyl)dianiline (I)

Compound S2. To a two-necked round bottom flask were added S1 (1.014 g, 3.0 mmol), tert-

butyl (4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)carbamate (2.490 g, 7.8 mmol),

K2CO3 (1.242 g, 9.0 mmol), Pd(PPh3)4 (347 mg, 0.3 mmol), 1,4-dioxane (48 mL), and H2O (12

mL). The reaction mixture was deaerated by slow bubbling of N2 for 30 min, and then heated to

reflux at 100 °C for 8 h. After cooling to room temperature, the solvent was removed under reduced

pressure, the residue was treated with H2O (60 mL), extracted with CH2Cl2 (4 × 40 mL), the

combined organic phase was washed with brine, dried over Na2SO4, and concentrated. The crude

product was purified by silica gel column chromatography using CH2Cl2/CH3OH = 100/0.5 as the

eluent to give the product S2 (1.204 g, 71% yield) as a white solid. 1H NMR (300 MHz, CDCl3):

δ = 8.48-8.39 (m, 4H), 8.26 (d, J = 8.5 Hz, 2H), 8.10 (d, J = 8.5 Hz, 2H), 7.75 (s, 2H), 7.64-7.53

(m, 4H), 6.65 (s, 2H), 1.57 (s, 18H). 13C NMR (125 MHz, CDCl3): δ = 156.1, 152.5, 139.7, 136.8,

134.1, 128.4, 127.7, 125.7, 119.4, 118.4, 80.8, 28.4. HRMS: m/z calcd for C34H35N4O4 [M + H]+:

563.2653, found: 563.2664.

4,4'-(1,10-phenanthroline-2,9-diyl)dianiline (I). To a round bottom flask were added S2 (562.3

mg, 1.0 mmol) and 4 M HCl/1,4-dioxane (10 mL). The mixture was stirred at room temperature

for 2 h, and then filtered to get a dark red solid. The solid was dissolved in deionized water (50

mL), and then neutralized with KOH solution. The resulting suspension was extracted with CH2Cl2

(3 × 40 mL), the combined organic phase was washed with brine, dried over Na2SO4, and

concentrated to afford the product I (335 mg, 92% yield) as a light yellow solid. 1H NMR (300

MHz, DMSO-d6): δ = 8.36 (d, J = 8.5 Hz, 2H), 8.24 (d, J = 8.6 Hz, 4H), 8.16 (d, J = 8.5 Hz, 2H),

7.81 (s, 2H), 6.77 (d, J = 8.6 Hz, 4H), 5.57 (s, 4H). 13C NMR (125 MHz, DMSO-d6): δ = 155.6,

150.4, 145.2, 136.5, 128.3, 126.7, 126.3, 124.9, 118.3, 113.8. HRMS: m/z calcd for C24H19N4 [M

+ H]+: 363.1604, found: 363.1611.

5

Synthesis of 2,9-bis(4-(dimethoxymethyl)phenyl)-1,10-phenanthroline (II)

2,9-bis(4-(dimethoxymethyl)phenyl)-1,10-phenanthroline (II). To a two-necked round bottom

flask were added S1 (676 mg, 2.0 mmol), 2-(4-(dimethoxymethyl)phenyl)-4,4,5,5-tetramethyl-

1,3,2-dioxaborolane (1.45 g, 5.2 mmol), K2CO3 (828 mg, 6.0 mmol), Pd(PPh3)4 (231 mg, 0.2

mmol), 1,4-dioxane (32 mL), and H2O (8 mL). The reaction mixture was deaerated by slowly

bubbling N2 for 30 min, and then heated to reflux at 100 °C for 8 h. After cooling to room

temperature, the solvent was removed under reduced pressure. The residue was treated with H2O

(60 mL), extracted with CH2Cl2 (4 × 40 mL), the combined organic phase was washed with brine,

dried over Na2SO4, and concentrated. The crude product was purified by silica gel column

chromatography using hexane/ethyl acetate = 3/1 as the eluent to give the product II (758 mg, 79%

yield) as a white solid. 1H NMR (300 MHz, CDCl3): δ = 8.51-8.40 (m, 4H), 8.33 (d, J = 8.4 Hz,

2H), 8.16 (d, J = 8.4 Hz, 2H), 7.81 (s, 2H), 7.68 (d, J = 8.3 Hz, 5H), 5.54 (s, 3H), 3.39 (s, 11H). 13C NMR (125 MHz, CDCl3): δ = 156.5, 146.1, 139.6, 139.2, 136.9, 127.9, 127.6, 127.3, 126.0,

120.1, 102.8, 52.5. HRMS: m/z calcd for C30H29N2O4 [M + H]+: 481.2122, found: 481.2131.

Synthesis of the conjugated polymer

I (11.6 mg, 0.032 mmol) and II (15.4 mg, 0.032 mmol) were weighed into a 10 mL Schlenk tube.

To the mixture were added 1-butanol (0.1 mL), 1,2-dichlorobenzene (0.9 mL) and 6 M aqueous

acetic (0.1 mL). The Schlenk tube was frozen in a liquid nitrogen bath, evacuated to an internal

pressure of 0 mbar and sealed. After warming to room temperature, the Schlenk tube was placed

into an oven and heated at 120 °C for 3 days yielding a yellow orange solid at the bottom of the

tube. The Schlenk tube was opened when the mixture was still warm and the solid was transferred

into a vial, separated by centrifugation, washed copiously with DMSO, THF, and EtOH. After

drying at room temperature, the resulting solid was dried under vacuum at 100 °C for 12 h to obtain

the conjugated polymer.

Crystallization of single-crystalline mCOF-Ag

I (17.4 mg, 0.048 mmol) and II (7.7 mg, 0.016 mmol) were weighed into a 10 mL Schlenk tube.

To the mixture, a solution of AgBF4 (3.1 mg, 0.016 mmol) in 1-butanol (0.1 mL) and 1,2-

6

dichlorobenzene (0.9 mL) was added. After sonicating at room temperature for 10 min, 0.1 mL 6

M aqueous acetic was added into the suspension. The Schlenk tube was frozen in a liquid nitrogen

bath, evacuated to an internal pressure of 0 mbar and sealed. After warming to room temperature,

the Schlenk tube was placed into an oven and heated at 120 °C for 3 days yielding a light yellow

solid at the bottom of the tube. The Schlenk tube was opened when the mixture was still warm and

the solid was transferred into a vial, separated by centrifugation, washed copiously with DMSO,

THF, and EtOH. After drying at room temperature, the resulting solid was dried under vacuum at

120 °C for 8 h to obtain mCOF-Ag as a light yellow solid (10.8 mg, 74% yield, based on AgBF4).

Anal. Calcd for (C49H33N7AgBF4)n: C 64.35; H 3.64; N 10.72. Found: C 64.29; H 3.61; N 10.67.

Crystalline-state polymerization

wCOF-Ag. To a 10 mL pressure tube were added mCOF-Ag (5.0 mg), 1,4-dioxane (1.0 mL), and

glyoxal solution (40 wt. % in H2O) (5.0 μL). The tube was sealed and placed in an oven at 70 °C

for 5 days. After cooling to room temperature, the solid was transferred into a vial, and washed

with THF, H2O, EtOH, and Et2O in sequence. The resulting solid was dried at room temperature

to afford the polymerized material, wCOF-Ag. Anal. Calcd for (C51H31N7AgBF4)n: C 65.41; H

3.34; N 10.47. Found: C 65.92; H 3.26; N 10.35.

Temperature is an important factor to obtain a high quality material. Increasing the temperature

from 70 °C to 80 °C will generate Ag particles on the surface of the crystal. The Ag particles are

attributed to the labile nature of Ag ions, which can be reduced to elemental Ag by glyoxal at a

higher temperature.

7

Supplementary Figure 1 | Experimental PXRD patterns of the conjugated polymer (black) and

mCOF-Ag (red). The synthesized conjugated polymer has a semi-crystalline feature, which is

attributed to the π-π stacking in the rigid and conjugated structure of phenanthroline backbones;

besides, the reversibility in imine bond formation can impart self-correction during the

hydrothermal conditions.

8

Supplementary Figure 2 | Experimental PXRD patterns of mCOF-Ag (black), building block a

(red), and building block b (blue).

9

Supplementary Figure 3 | FT-IR spectra of mCOF-Ag (black), building block a (red), and

building block b (blue).

10

Supplementary Figure 4 | FT-IR spectra of mCOF-Ag before (black) and after (red) drying the

KBr pellet at 100 °C to eliminate the influence of water. In the red curve, the characteristic peaks

of N–H stretches and N–H bending vibration are apparent.

11

Supplementary Figure 5 | 13C CP/MAS NMR spectra of mCOF-Ag. The asterisks denote the

spinning sidebands.

12

Supplementary Figure 6 | TGA curve of mCOF-Ag. The decomposition temperature is about 320

ºC. The ~1.4% weight loss below 320 °C is attributed to the small amount of solvents trapped in

the crystal.

13

Supplementary Figure 7 | Nitrogen adsorption (filled symbols) and desorption (empty symbols)

isotherms of mCOF-Ag. The low uptake in the isotherm line is in line with its crystal structure,

whereby the pores are occupied by BF4– anions.

14

Supplementary Table 1 | SCED: Experimental and Crystallographic parameters of mCOF-Ag.

Tilt range (°) -41.40 ~ 39.96

Tilt step (°) 0.23

Wave length (Å) 0.0251

No. of frames 336

Exposure time per image (s) 0.5

Crystal system Monoclinic

Possible space group C2/c, Cc, C2/m, Cm, C2

Unit cell parameters a = 15.66 Å, b = 31.00 Å, c = 10.87 Å,

α = 90°, β = 123.3°, γ = 90°

Resolution (Å) 0.95

I/σ(I) 4.90

Completeness (%) 56

CC1/2 98.8

Rmeas (%) 11.8

No. of total reflections 4024

No. of unique reflections 1522

15

Supplementary Figure 8 | Observed electron density map of the structure of mCOF-Ag

determined from SCED data along the a-axis. All positions of the non-hydrogen atoms (C, N, and

Ag) on the framework and the locations of the guests (BF4‾ anions) are resolved.

16

Supplementary Figure 9 | Crystal view along the c-axis, in which the H∙∙∙F hydrogen bond

interactions between amine and BF4‾ anions are highlighted with a red square. Colour scheme: C

on organic chains, blue; C on pendants, grey; N, purple; Ag, yellow; B, pink; F, green; H of amine

groups, white.

17

Supplementary Figure 10 | Tilted side view of the crystal structure, in which a single

phenanthroline chain is highlighted in bright yellow. The corrugated layer and the π-π stacking of

interlayer phenanthroline rings are also identified. C on organic chains, blue; C on pendants, grey;

N, purple; Ag, yellow. Hydrogen atoms and BF4‾ anions are omitted for clarity.

18

Supplementary Figure 11 | Structure refinement of mCOF-Ag against the SPXD data (Rietveld

refinement). Experimental (red), Rietveld refined (blue), as well as difference profiles (black) are

presented. The positions of Bragg reflections under the patterns are shown in green.

19

Supplementary Figure 12 | Experimental SPXD pattern and the simulated profile from the single

crystal structure of mCOF-Ag. The simulated data is highly consistent with the experimental data,

indicating the very high quality of the synthesized crystals.

20

Supplementary Figure 13 | Uv-vis spectra of mCOF-Ag, building block I, and building block

II.

21

Supplementary Figure 14 | PL spectra of mCOF-Ag, building block I, and building block II

excited by 532 nm laser.

22

Supplementary Figure 15 | SHG spectra of an isolated mCOF-Ag crystal at different pump

wavelengths.

23

Supplementary Figure 16 | Pump power dependent SHG spectra of an isolated mCOF-Ag crystal

at 850 nm. Inset shows the linear relationship between SHG intensity and pump power, the fitting

curve (red line) has a slope of 1.94, confirming the second-order nature of the nonlinear process.

24

Supplementary Figure 17 | FT-IR spectra showing time evolution of vibrational bands when

mCOF-Ag was cross-linked with glyoxal. With longer reaction time, the characteristic peaks of

N−H stretches (blue band) and N−H bending vibration (pink band) decrease, indicating the

consumption of amine groups, while the peak at 1600 cm-1 (green band) due to newly formed

imine bond increases gradually. The red curve corresponds to the control experiment without

glyoxal, which was heated in 1,4-dioxane at 70 °C for five days. All KBr pellets were dried at 100

°C before testing to eliminate background signals due to water.

25

Supplementary Figure 18 | 13C CP/MAS NMR spectra of mCOF-Ag and wCOF-Ag. The

asterisks denote the spinning sidebands.

26

Supplementary Figure 19 | SEM image of wCOF-Ag.

27

Supplementary Figure 20 | 3D reciprocal lattice of wCOF-Ag reconstructed from the SCED data.

As wCOF-Ag was synthesized via post-synthesis, the local bonding is not very order. It is difficult

to directly observe the new formed bonds via SCED data.

28

Supplementary Table 2 | SCED: Experimental and Crystallographic parameters of wCOF-Ag.

Tilt range (°) -53.22 ~ 50.17

Tilt step (°) 0.23

Wave length (Å) 0.0251

No. of frames 401

Exposure time per image (s) 0.5

Crystal system Monoclinic

Possible space group C2/c, Cc, C2/m, Cm, C2

Unit cell parameters a = 15.74 Å, b = 31.11 Å, c = 11.00 Å,

α = 90°, β = 122.6°, γ = 90°

Resolution (Å) 1.15

I/σ(I) 2.54

Completeness (%) 72.4

CC1/2 95.6

Rmeas (%) 27.2

No. of total reflections 2775

No. of unique reflections 1145

29

Supplementary Figure 21 | Pawley fitting of wCOF-Ag. Experimental (red), calculated (blue),

as well as difference profiles (black) are presented. The position of Bragg reflections under the

patterns are shown in green. The refined unit cell parameters are a = 15.6958 Å, b = 29.9200 Å, c

= 10.6661 Å, and β = 123.31° (space group: C2), which are in good agreement with those of

mCOF-Ag, indicating the parent structure of mCOF-Ag has been well inherited by wCOF-Ag. The

final residuals for the Pawley fitting are Rwp = 0.135, Rexp = 0.123.

30

Supplementary Figure 22 | X-ray photoelectron spectroscopy (XPS) Ag3d data of mCOF-Ag

and wCOF-Ag. The XPS Ag3d spectrum of wCOF-Ag is nearly identical with that of mCOF-Ag,

indicating that Ag ions in both materials have similar chemical environment.

31

Supplementary Figure 23 | PL spectra of mCOF-Ag (before crosslinking) and wCOF-Ag (after

crosslinking) excited by 532 nm laser.

32

Supplementary Figure 24 | AFM nano-indentation force−displacement curve of an isolated

crystal of mCOF-Ag. Following the standard analysis of nano-indentation, the Young’s modulus

is related to the indentation curve by Equation 1: Hertzian model 𝐹 = 4

3

𝐸

(1−𝑣2)√𝑅𝛿3/2, where F =

force (from force curve), E = Young's modulus (fit parameter), ν = Poisson's ratio (sample

dependent, 0.3), R = radius of the indenter (30.0 nm), δ = indentation.

33

Supplementary Figure 25 | AFM nano-indentation force−displacement curve of an isolated

crystal of wCOF-Ag. Following the standard analysis of nano-indentation, the Young’s modulus

is related to the indentation curve by Equation 1: Hertzian model 𝐹 = 4

3

𝐸

(1−𝑣2)√𝑅𝛿3/2, where F =

force (from force curve), E = Young's modulus (fit parameter), ν = Poisson's ratio (sample

dependent, 0.3), R = radius of the indenter (30.0 nm), δ = indentation.

34

Supplementary Figure 26 | 1H-NMR spectrum of compound S2 (25 °C, 300 MHz, CDCl3).

Supplementary Figure 27 | 13C-NMR spectrum of compound S2 (25 °C, 125 MHz, CDCl3).

35

Supplementary Figure 28 | 1H-NMR spectrum of I (25 °C, 300 MHz, DMSO-d6).

Supplementary Figure 29 | 13C-NMR spectrum of I (25 °C, 125 MHz, DMSO-d6).

36

Supplementary Figure 30 | 1H-NMR spectrum of II (25 °C, 300 MHz, CDCl3).

Supplementary Figure 31 | 13C-NMR spectrum of II (25 °C, 125 MHz, CDCl3).

37

Supplementary References:

1. Clary, J. W. et al. Hydride as a leaving group in the reaction of pinacolborane with halides under

ambient Grignard and Barbier conditions. One-pot synthesis of alkyl, aryl, heteroaryl, vinyl, and

allyl pinacolboronic esters. J. Org. Chem. 76, 9602-9610 (2011).

2. Kabsch, W. XDS. Acta Crystallogr. Sect. D. 66. 125-132 (2010).

3. Wan, W., Sun, J., Su, J., Hovmöller, S. & Zou, X. Tree-dimensional rotation electron difraction:

sofware RED for automated data collection and data processing. J. Appl. Crystallogr. 46, 1863-

1873 (2013).

download fileview on ChemRxivSupplementary_Materials.pdf (2.46 MiB)

Other files

download fileview on ChemRxivmCOF-Ag.cif (6.16 KiB)