1

All reagents and organic dyes, unless otherwise noted, were obtained from the Energy Chemical and used

without further purification. The solvents (ethanol, o-dichlorobenzene, 1,4-dioxane, mesitylene and n-

butanol) for COF synthesis condition optimization were purchased from Energy Chemical. Deionized

water was used for the preparation of 6 M acetic acid. Solvents for removal of guests from the pores of

COF were obtained from Sinopharm Chemical Reagent Beijing Co., Ltd.

Methods: Powder X-ray diffraction measurements were carried out on a Bruker D4 ENDEAVOR X-Ray

diffractometer (Bruker, Germany) at 40 kV and 40 mA with Cu Ka radiation (λ= 1.5406 Å). Solid-phase

synchrotron X-ray-scattering experiments were performed on the BL16B beamline of Shanghai

Synchrotron Radiation Facility, using a fixed wavelength of 0.124 nm, a sample-to-detector distance of

1.84 m and an exposure time of 200 s. The 2D scattering pattern was collected on a charge-coupled device

camera, and the curve intensities versus q were obtained by integrating the data from the pattern. Small-

angle X-ray scattering experiments were conducted both on the NanoStar U SAXS System (Bruker,

German) using Cu Kα radiation (40 kV, 35 mA) and exposure time of 2 minutes, and on the in SAXSess

mc2 system (Anton Paar, Austria) with Ni filtered Cu Kα radiation source. The power of X-ray source

was operated at 50 mA and 40 kV. The d-spacing values were calculated by the formula d = 2π/q. Scanning

electron micrographs of the samples were obtained on a Nova nano SEM 450 Field Emission Scanning

Electron Microscope at 3.00 kV with the material adhered to the SEM sample holder directly or on a

Phenom China Scanning Electron Microscope at 15.00 kV after the material that adhered to the sample

holder was been gilded to 10-1 -10-2 vacuum degree. Transmission electron micrographs were recorded on

a JEM 2011 FETEM microscope at 200 kV aligned for low dose (10 e Å-2 s-1) diffractive imaging. TGA

experiments were performed on a Model TGA/SDTA 851 instrument. Samples were placed in alumina

pans and heated at a rate of 5 °C per minute from 100 to 900 °C under a nitrogen atmosphere. Fourier

transform infrared spectroscopy was carried out with an Avatar-360 FT-IR spectrometer (Nicolet, USA).

The samples for IR spectra were prepared as KBr pellets. Solid-state NMR spectra were recorded at

ambient pressure on a Bruker 400WB AVANCE III spectrometer using a standard Bruker magic angle-

spinning (MAS) probe with 4 nm zirconia rotors. UV-Vis spectra were performed on a Perkin-Elmer 750s

instrument from 200-800 nm at the scan rate of 3 nm/ internal. Gas adsorptions of N2 at 196 C and CO2

at 0 C were measured on a Micromeretics Model ASAP 2020 gas adsorption analyzer. About 40 mg of

activated sample was degassed at 180 C for 10 h by using the “outgas” function of the surface area

analyzer. Helium gas was used to estimate the dead volume. The saturation pressure (P0) was measured

throughout the N2 analyses via a dedicated saturation pressure transducer, which helped to monitor the

vapor pressure for each data point. For CO2 isotherm measurements, activated sample was transferred into

a pre-weighed glass sample tube. The tube was then sealed and quickly transferred to a system providing

10-4 torr dynamic vacuum. The sample was kept under this vacuum at 180 C for 10 h and then used for

CO2 adsorption measurements.

Adsorption experiments. In a typical experiment, activated PC-COF was added to a solution of MO,

AR-27, AG-25, DFBM, or IC of fixed concentration in water. The amount of PC-COF was calculated by

assuming no defects. The UV-vis spectra of the solutions were recorded after shaking for different time

Electronic Supplementary Material (ESI) for Polymer Chemistry.This journal is © The Royal Society of Chemistry 2016

2

and standing for 15 minutes at room temperature. The absorption of the dye was then recorded and used

to calculate the amount of dye that was absorbed by PC-COF.

Desorption experiments. In a typical experiment, MO-adsorbed PC-COF (5 mg) containing 8.9 10-6

mole of MO was immersed into a solution of ethanol (85 mL) and acetic acid (15 mL). The UV-vis spectra

of the solution were recorded at 60 C after shaking for different time and then standing for 15 minutes.

The absorption of the acidized MO was recorded and used to calculate the amount of the acidized MO

that was released from the MO@PC-COF sample.

TAPB and BFBP2+•2PF6- was synthesized according to the reported precedure.1,2 BFBP2+•2Cl- was

obtained from BFBP2+•2PF6- by ion-exchange with tetrabutylammonium chloride in acetonitrile. The

precipitation was filtrated and washed with acetonitrile, and dried in vacuo to give BFBP2+•2Cl- as a pale

grey solid. FBP2+•2PF6- and BFBP2+•2Cl- gave rise to the indentical 1H and 13C NMR in DMSO-d6.

Synthesis of model compound 1,1'-bis(4-((E)-(phenylimino)methyl)phenyl)-[4,4'-bipyri-dine]-1,1'-

diium: The model compound was synthesized by the reaction of BFBP2+•2PF6- (98 mg, 0.15 mmol) and

aniline (28 mg, 0.30 mmol) in 10 mL acetonitrile under nitrogen at 90 ºC for 12 hours or standing in

atmosphere at r.t. for 3 days. Under the standing condition, the grey precipitate was collected and filtration

was dried under vacuum to get the yellow powder. 1H NMR (400MHz, DMSO-d6): 9.77 (s, 4 H), 9.12 (s,

4 H), 8.87 (s, 2 H), 8.34 (s, 4 H), 8.15 (s, 4 H), 7.51-7.47 (t, 4 H), 7.39-7.37 (d, 4 H), 7.33 (s, 2H) ppm. 13C NMR (400 MHz, DMSO-d6): 159.5, 151.2, 149.6, 146.34, 144.2, 139.1, 130.5, 129.8, 127.3, 127.1,

126.0, 121.7 ppm. HRMS (ESI): Calcd for C36H28N4: 516.2314 [M-2Cl]2+. Found 516.2282.

Synthesis of PC-COF. 1,3,5-Tris(4-aminophenyl)benzene (TAPB) (42 mg, 0.12 mmol) and 1,1-bis(4-

formylphenyl)-4,4’-bipyridinium dichloride (BFBP2+•2Cl-) (79 mg, 0.18 mmol) were weighed into a

schlenk tube. To the mixture were added 3.6 mL o-dichlorobenzene and 0.4 mL ethanol as solvent

combination and 0.8 mL aqueous acetic acid (6 M) as catalyst at the ratio 9:1:2. The schlenk tube was

then flash frozen at 77 K (liquid N2 bath) and degassed by three freeze-pump-thaw cycles. The tube was

sealed off with a little N2 in and then heated at 120 °C for 7 days. The powders formed were filtered out,

and washed with ethanol and acetone, and then dried under vacuum at 120 °C for 12 hours to give PC-

COF as a pale brown powder (0.10 g, 83% based on TAPB).

Synthesis of PC-COF-2. TAPB (14 mg, 0.04 mmol) and BFBP2+•2Cl- (39 mg, 0.06 mmol) were weighed

into a schlenk tube. To the mixture was added 1.5 ml o-dichlorobenzene and 0.5 ml ethanol as solvent

combination and 0.35 ml 6 mol/L aqueous acetic acid as catalyst at the ratio 15:5:3.5. Then the schlenk

tube was then flash frozen at 77 K (liquid N2 bath) and degassed by three freeze-pump-thaw cycles. The

3

tube was sealed off with a little N2 in and then heated at 120 °C for 7 days. After the reaction the COF

powders are filtered out, washed with dichloromethane and acetone and then dried under vacuum at 120 °C

for 12 hours to give PC-COF-2 as a pale grey powder (40 mg, 73% based on TAPB).

Adsorption of dyes by PC-COF-2. Activated PC-COF-2 (1.2 mg) were added to a solution of MO, IC,

or AG-25 in 10 ml water to prepare solutions of required concentrations. The molar quantity of the anionic

units of dyes was equal to that of the cationic unit of PC-COF-2 (1.2 mg) with the assumption that the

PC-COF-2 was of non-defects. The UV-vis spectra of the solution were recorded after different shaking

time at room temperature and at least 10-minute standing time. The decreased absorption of dyes was then

used to calculate the amount of dyes that was absorbed by the PC-COF-2 sample.

4

2000 1500 1000 5000

25

50

75

100

NN

N

N

2 Cl-

cis/ trans

Wavenumber (cm-1)

1635 cm-1

C=N

a)

3500 3000 2500 2000 1500 1000 500

OH

3419 cm-1

Wavenumber (cm-1

)

1635 cm-1

C=N

model compound

PC-COF

b)

Figure S1. The FT-IR spectrum of a) model compound and b) PC-COF.

5

2000 1800 1600 1400 1200 1000

wavnumber cm -1

BFPBP 2Cl -

-CHO

-C=N

-C=C

TAPB

PC-COF

Figure S2. The FT-IR spectrum of PC-COF (upper), TBPA (middle) and BFBP2+2Cl- (down).

Figure S3. 13C NMR of BFBP2+2Cl-, TBPA, model compound, and PC-COF.

6

Figure S4. Field-emission SEM images of PC-COF.

Figure S5. TEM images of PC-COF.

7

200 400 600 800

30

40

50

60

70

80

90

100

44.87% loss

1.90% loss

32.89% loss

24.98% loss

17.75% loss

6.11% loss

norm

aliz

ation

Temperature

12.57% loss

Figure S6. Thermogravimetric analysis of PC-COF.

0.0 0.2 0.4 0.6 0.8 1.0

0

20

40

60

80

100

100 200 300 400 500 600

0.0000

0.0002

0.0004

0.0006

0.0008

0.0010

dV

p/(

cm

-3g

-1)

angstrom

pore-size-distribution

N2 u

pta

ke

(cm

3/g

)

P/P0

adsorption

desorption

a)

0 200 400 600 800

0

3

6

9

12

15

CO

2 u

pta

ke

(cm

3/g

)

Pressure (mmHg)

adsorption

desorption

b)

Figure S7. a) N2 adsorption/desorption isotherms at -196 C (insert pore-size-distribution) and b) CO2

adsorption/desorption at 0 C of PC-COF as synthesized.

8

0 5 10 15 20 25 30 35 40 45 50

10 15 20 25 30 35 40 45 50

2

001

b)

2 4 6 8 102 (

o)

100

110200

210300 220 320 330

2

a)

100

200

Figure S8. The PXRD profile of PC-COF.

2.5 5.0 7.5 10.0

2 ()

200

100a)

1 2 3 4 5 6

200

q (nm-1)

100

110

b)

0.05 0.10 0.15 0.20

q (angstron-1)

100c)

Figure S9. a) PXRD, b) and c) SAXS profiles of PC-COF-2.

9

0.00 0.05 0.10 0.15 0.20

q (angstrom-1)

a)

0.00 0.05 0.10 0.15 0.20

q (angstrom-1)

b)

0.00 0.05 0.10 0.15 0.20

q (angstrom-1)

c)

0.00 0.05 0.10 0.15 0.20

q (angstrom-1)

d)

Figure S10. The SAXS profiles of PC-COF after soaking in water for a) 0, b) 1, c) 2, and d) 10 days.

The latter three samples were obtained after being filtrated off and dried in vacuo.

0.0 0.1 0.2 0.30

1

2

3

IC at 611 nm

AG-25 at 611 nm

MO at 464 nm

AR-27 at 521 nm

DFBM at 395 nm

Ab

s

[dye] (mM)

a)

0.000 0.025 0.050 0.075 0.100

0.0

0.5

1.0

1.5

2.0

2.5

MO with equivilent H+ from H

2SO

4

MO with excess H+ from CH

3CO

2H

Ab

s

[dye] (mM)

b)

Figure S11. a) The plots of the absorbance of dyes by PC-COF versus the concentrations of the dyes in

water. b) The absorbance of MO versus its concentrations in ethanol containing acids (2.5 mM).

10

0 20 40 60 80 100 120

0.0

0.2

0.4

0.6

0.8

1.0

IC+1eq. PC-COF

up

take

%

Time (h)

40.4%

a)

0 20 40 60 80 100 120

0.0

0.2

0.4

0.6

0.8

1.0

AG-25+1eq. PC-COF

AG-25+3eq. PC-COF

up

take

(%)

Time (h)

11.4%

37.8%

b)

0 20 40 60 80 100 120

0.0

0.2

0.4

0.6

0.8

1.0

up

take

(%)

Time (h)

MO+1eq. PC-COF

MO+3eq. PC-COF

32.7%

85.1%

c)

0 20 40 60 80 100 120

0.0

0.2

0.4

0.6

0.8

1.0

AR-27+1eq. PC-COF

AR-27+3eq. PC-COF

up

take

(%)

Time (h)

d)

33.9%

89.5%

0 20 40 60 80 100 120

0.0

0.2

0.4

0.6

0.8

1.0

64.5%

DFBM+1eq. PC-COF

DFBM+3eq. PC-COF

up

take

(%)

Time (h)

e)

38.2%

Figure S12. The uptake of organic dyes at the same charge concentration of 0.21 mM by 1 and 3

loadings of PC-COF at different times.

11

350 400 450 500 550 600

0.0

0.5

1.0

1.5

2.0

2.5

MO+1 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=25h

t=34h

t=48h

t=59h

t=70h

t=82h

t=106h

a)

350 400 450 500 550 600

0.0

0.5

1.0

1.5

2.0

2.5

t=0

t=1h

t=34h

t=48h

t=59h

t=70h

t=82h

t=106h

MO+3 eq. PC-COF

Ab

s

wavelength (nm)

b)

350 400 450 500 550 600

0.0

0.2

0.4

0.6

0.8

MO+3 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=2.5h

t=6h

t=16h

t=27h

t=40h

t=69h

t=103h

c)

350 400 450 500 550 600

0.0

0.2

0.4

0.6

0.8

t=0

t=2h

t=5h

t=15h

t=27h

t=39h

t=65h

t=87h

t=103h

Ab

s

wavelength (nm)

MO+3 eq. PC-COF

d)

350 400 450 500 550 600

0.0

0.2

0.4

0.6

0.8

t=0

t=1h

t=3h

t=25h

t=34h

t=48h

t=59h

t=70h

t=48h

t=59h

t=70h

Ab

s

wavelength (nm)

MO+5 eq. PC-COF

e)

Figure S13. UV-vis spectra of MO samples at the initial concentrations of a) and b) 0.21 mM, and of c),

d) and e) 0.032 mM after uptaking by different loadings of PC-COF.

12

500 550 600 650 700 750 800

0.0

0.4

0.8

1.2

A

wavelength (nm)

t=0

t=1.5h

t=3h

t=8h

t=20.5h

t=33h

t=45h

t=57h

t=71h

t=94h

t=107h

AG-25+1 eq. PC-COFa)

500 550 600 650 700 750 800

0.0

0.4

0.8

1.2AG-25+3 eq. PC-COF

A

wavelength (nm)

t=0

t=1.5h

t=3h

t=8h

t=20.5h

t=33h

t=45h

t=57h

t=71h

t=94h

t=107h

b)

500 550 600 650 700 750 800

0.00

0.05

0.10

0.15

0.20

AG-25+1 eq. PC-COF

A

wavelength (nm)

t=0

t=2h

t=5h

t=15h

t=27h

t=39h

t=65h

t=87h

t=103h

c)

500 550 600 650 700 750 800

0.00

0.05

0.10

0.15

0.20AG-25+3 eq. PC-COF

A

wavelength (nm)

t=0

t=2h

t=5h

t=15h

t=27h

t=39h

t=65h

t=87h

t=103h

d)

500 550 600 650 700 750 800

0.00

0.05

0.10

0.15

0.20

AG-25+5 eq. PC-COF

A

wavelength (nm)

t=0

t=1.5h

t=3h

t=8h

t=20.5h

t=33h

t=45h

t=57h

t=71h

t=94h

t=107h

e)

500 550 600 650 700 750 800

0.00

0.05

0.10

0.15

0.20

AG-27+10 eq. PC-COF

A

wavelength (nm)

t=0

t=1h

t=3h

t=25h

t=34h

t=48h

t=59h

t=70h

t=82h

f)

Figure S14. UV-vis spectra of AG-25 samples at the initial concentrations of a) and b) 0.11 mM, and of

c), d) and e) 0.032 mM after uptaking by different loadings of PC-COF.

13

350 400 450 500 550 600 650 700

0.0

0.4

0.8

1.2

DFBM + 1 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=7h

t=18h

t=29h

t=41h

t=54h

t=78h

t=102h

a)

350 400 450 500 550 600 650 700

0.0

0.4

0.8

1.2 DFBM + 3 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=7h

t=18h

t=29h

t=41h

t=54h

t=78h

t=102h

b)

350 400 450 500 550 600 650 700

0.00

0.04

0.08

0.12

0.16 DFBM + 3 eq. PC-COF

Ab

s

wavelength (nm)

t=2

t=2h

t=5h

t=15h

t=27h

t=39h

t=65h

t=87h

t=103h

c)

350 400 450 500 550 600 650 700

0.00

0.04

0.08

0.12

0.16 DFBM + 3 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=2.5h

t=6h

t=16h

t=27h

t=40h

t=69h

t=103h

d)

350 400 450 500 550 600 650 700

0.00

0.04

0.08

0.12

0.16DFBM + 5 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=7h

t=18h

t=29h

t=41h

t=54h

t=78h

t=102h

e)

Figure S15. UV-vis spectra of DFBM at the initial concentrations of a) and b) 0.11 mM, and of c), d)

and e) 0.032 mM after uptaking by different loadings of PC-COF.

14

450 500 550 600 650 700

0.0

0.5

1.0

1.5

2.0IC+1 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=5h

t=10h

t=22h

t=35h

t=59h

t=83h

t=105h

a)

450 500 550 600 650 700

0.00

0.06

0.12

0.18

0.24

0.30

0.36IC+1 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=2h

t=5h

t=15h

t=27h

t=39h

t=65h

t=87h

t=103h

b)

450 500 550 600 650 700

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35IC+3 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=5h

t=10h

t=22h

t=35h

t=59h

t=83h

t=105h

c)

450 500 550 600 650 700

0.00

0.06

0.12

0.18

0.24

0.30

0.36 IC+5 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=25h

t=34h

t=48h

t=59h

t=70h

t=82h

t=106h

d)

Figure S16. UV-vis spectra of IC samples at the initial concentrations of a) 0.11 mM, and of b), c) and

d) 0.032 mM after uptaking by different loading of PC-COF.

15

400 450 500 550 600 650

0.0

0.5

1.0

1.5

2.0AR-27+1 eq. PC-COF

Abs

wavelength (nm)

t=0

t=1h

t=3h

t=7h

t=18h

t=29h

t=41h

t=54h

t=78h

t=102h

a)

400 450 500 550 600 650

0.0

0.5

1.0

1.5

2.0

AR-27+3 eq. PC-COF

Abs

wavelength (nm)

t=0

t=1h

t=3h

t=7h

t=18h

t=29h

t=41h

t=54h

t=78h

t=102h

b)

400 450 500 550 600 650

0.00

0.05

0.10

0.15

0.20

0.25

0.30 AR-27+3 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=2h

t=5h

t=15h

t=27h

t=39h

t=65h

t=87h

t=103h

d)

Figure S17. UV-vis spectra of AR-27 samples at the initial concentrations of a) and b) 0.07 mM, and of

b), c) and d) 0.011 mM after uptaking by different loadings of PC-COF.

400 450 500 550 600 650

0.00

0.05

0.10

0.15

0.20

0.25

0.30AR-27+5 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=2h

t=5h

t=15h

t=27h

t=39h

t=65h

t=87h

t=103h

c)

400 450 500 550 600 650

0.00

0.05

0.10

0.15

0.20

0.25

0.30 AR-27+5 eq. PC-COF

Ab

s

wavelength (nm)

t=0

t=1h

t=3h

t=7h

t=18h

t=29h

t=41h

t=54h

t=78h

t=102h

e)

16

0 30 60 90 120

0.00

0.25

0.50

0.75

1.00

99.7%

77.2%

IC+1eq. PC-COF

IC+3eq. PC-COF

IC+5eq. PC-COFup

take

%

adsorption time (h)

24.3%

a)

0 30 60 90 120

0.00

0.25

0.50

0.75

1.00

99.6%

54.5%

22.2%

AG-25+1eq. PC-COF

AG-25+3eq. PC-COF

AG-25+5eq. PC-COF

AG-25+10eq. PC-COF

up

take

(%

)

adorption time (h)

b)

7.8%

0 30 60 90 120

0.00

0.25

0.50

0.75

1.00

MO+1eq. PC-COF

MO+3eq. PC-COF

MO+5eq. PC-COFup

take

(%)

adorption time (h)

c)

25.4%

75.6%

96.9%

0 30 60 90 120

0.00

0.25

0.50

0.75

1.00

86.9%

42.5%

AR-27+1eq. PC-COF

AR-27+3eq. PC-COF

AR-27+5eq. PC-COF

up

take

(%)

adorption time (h)

d)

16.1%

0 30 60 90 120

0.00

0.25

0.50

0.75

1.00

60.3%

35.1%

DFBM+1eq. PC-COF

DFBM+3eq. PC-COF

DFBM+5eq. PC-COF

up

take

(%)

adorption time (h)

e)

22.1%

0 100 200 300

0.00

0.25

0.50

0.75

1.00

69.4%

AR-27 +5 eq. PC-COF

DFBM +5 eq. PC-COF

up

take

%

adsorption time (h)

f)

97.8%

Figure S18. The normalized uptake of dyes at the same ion concentration of 0.32 mM by 1, 3, 5, and 10

loadings of PC-COF with the adsorption time being a-e) 110 h and f) 300 h.

17

0.05 0.10 0.15 0.20

q (ansgtrom-1)

PC-COF

PC-COF + Ar-27 after immersed in water for 29 d

PC-COF + DFBM after immersed in water for 18 d

(100)

Figure S19. SAXS profiles of PC-COF before and after the uptake of AR-27 and DFBM.

Figure S20. SEM images of PC-COF after uptaking IC from water.

18

350 400 450 500 550

0.00

0.04

0.08

0.12

0.16

0.20

0.24

Ab

s

wavelength (nm)

t=2h

t=8 h

t=25h

t=48h

t=72h

t=96h

a)

0 20 40 60 80 100

2

3

4

5

6

7

8

9

rele

ase

(1

0-6 m

ol)

shaking time (h)

b)

0 20 40 60 80 100

0

20

40

60

80

100

rele

ase

yie

ld (

%)

shaking time (h)

c)

97.9%

Figure S21. UV-vis spectra of MO released to ethanol containing AcOH (2.5 mM) from MO-adsorbed

PC-COF.

400 500 600

0.0

0.2

0.4

0.6

0.8

Ab

s

Wavelength (nm)

0

2 h

4 h

8 h

12 h

28 h

34 h

47 h

60 h

77 h

100 h

a)

0 20 40 60 80 100

0.0

0.2

0.4

0.6

0.8

1.0

MO + 5 eq. PC-COF with AcO- as counterion

MO + 5 eq. PC-COFup

take

(%

)

Adsorption time (h)

b)

97%

34%

Figure S22. a) UV-vis spectra of MO (0.032 mM) of PC-COF (5 equivalent) with AcO- as counterion,

b) the adsorption efficiency for MO versus time.

450 500 550 600 650 700

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Ab

s

wavelenght (nm)

t=0

t=0.33h

t=1h

t=2

t=3h

a)

500 550 600 650 700 750 800

0.0

0.5

1.0

1.5

2.0

2.5

Ab

s

wavelength (nm)

t=0

t=1h

t=2h

t=3h

b)

350 400 450 500 550 600

0.0

0.5

1.0

1.5

2.0

Ab

s

wavelenght (nm)

t=0

t=0.25h

t=1.25h

t=2.25h

c)

Figure 23. Time-dependent UV-vis spectra of a) IC (0.16 mM), b) AG-25 (0.16 mM), and c) MO (0.32

mM) after mixing with 1 equivalent of PC-COF-2.

19

0 20 40 60 80 100 120

0

2

4

6

We

igh

t o

f a

dso

rbe

d w

ate

r

Time (h)

Figure S24. Water adsorption experiments of PC-COF. The sample (80 mg) was exposed to air of 100%

relative humidity at 25 C for 5 days. The Y-axis values represent the weight of water adsorbed by the

sample.

Figure S25. 1H NMR spectrum of model compound in DMSO-d6.

20

Figure S26. 1H NMR spectrum of model compound (upper), aniline (middle), and BFBP•2PF6– (down).

Figure S27. 13C NMR spectrum of model compound in DMSO-d6.

21

General methods for structural simulation of PC-COF. Structure modelling of PC-COF was

performed using the Accelrys Materials Studio 7.0 program suite.3 Geometric optimization of COF

structure in classic forcefield was carried out with Forcite module, and PXRD pattern prediction was done

in Reflex module. In order to simulate the electrostatic interaction between layers of the polycationic

framework and anions, simple model of eclipsed (AA stacking) PC-COF was built and geometrically

optimized in Universal forcefield with no specific charge treatment with reference to previously reported

methods4,5 (Figure S25). The lattice was in P6/M space group, and the parameters are a = 59.493Å, b =

59.493 Å, c = 3.439 Å, α = 90°, β = 90°, γ = 120°. PXRD prediction was in good agreement with

experimental data even though some difference in peak strength. However, structural check revealed that

inside the model, the framework was neutral and Cl- anions were omitted. Thus, we considered that the

above model could not represent the actual structure inside the material.

Figure S28. Simulated structure for non-charged eclipsed (AA stacking) PC-COF viewed from a) 001

direction, b) 110 direction, and c) the predicted PXRD pattern (down) and experimental PXRD pattern

(upper).

With further investigation, different treatments were selected to incorporate charges of bipyridinium N

atoms into the eclipsed PC-COF model. The optimization was carried out with COMPASSII forcefield,

with charges specified by Gasteiger method (Gast_polygraf 1.0 parameter set) to support calculation of 4-

valent nitrogen, which was used in following models without specification (Figure S26). The lattice

parameters were a = 58.281 Å, b = 58.929 Å, c = 19.999 Å, α = 89.958°, β = 89.840°, γ = 119.792°, and

Cl- anions were put in random coordination in the pores, thus the symmetry group changed to P1. The

resulting structure gave a lattice with a c parameter of ~20 Å but still did not reach a convergence, resulting

in a 001 signal at ~4.4º, which did not match the experimental data. This indicated a strong repulsion

between positive charges of the stacked layers of PC-COF.

22

Figure S29. Simulated structure of charged eclipsed (AA stacking) PC-COF viewed from a) 001 direction,

b) 110 direction, and c) the predicted PXRD pattern and the experimental PXRD pattern.

Staggered (AB stacking) structure of honeycomb COFs, with the same stacking pattern as graphite, was

considered to be able to release the direct repulsion between cations of paralleled layers. Optimization

with the same treatment as shown above gave the structure of c = 7.043Å, which was consistent to the

experimental data, but the relative intensity of peaks between 1~10 ºdid not match ideally (Figure S27),

indicating this might not be the best structure. However, it was observed that the randomly placed Cl-

anions were drawn to the framework, “harboring” in between positively charged bipyridinium -N atoms

of adjacent A (or B) layers, causing an interlayer distance of ~7 Å. The N-Cl distances were measured as

3.31 Å and 5.14 Å (average). This indicated that counter ions harboring close to or inserted in between

layers of the framework might alleviate the electrostatic repulsion and stabilize the structure to some extent.

Figure S30. Simulated structure of charged staggered (AB stacking) PC-COF viewed from a) 001

direction, b) 110 direction, and c) 1,-1,0 direction (gray: C, white: H, blue: N, green: Cl), and d) the

predicted PXRD pattern and the experimental PXRD pattern.

The above result was introduced back to the eclipsed model of PC-COF, and the interlayer distance was

decreased to 6.5 Å, which was a large improvement but still inconsistent with the observed distance. Small

23

arbitrary shift of 4.44 Å was tried according to similar treatment of a formerly reported COF model,4

leading to a model shown in Figure S28, with the lattice parameters of a = 58.100 Å, b = 60.229 Å, c =

6.944 Å, α = 91.388°, β = 92.500°, γ = 121.849°. The shift was drawn back by the strong stacking of

TAPB residues. However, probably because of its perturbation to the symmetry, the BP linkers were

slightly bent (measured 156.6°) sideward and the interlayer distance was 3.47 Å. This model showed a

favored conformation of π-π stacking, where the bipyridinium rings packed sideward with the C-H bonds

of one stretching towards center of another, and Cl- anions were close aside to neutralize the charge

repulsion. From (110) view of the simulated crystal, the components actually formed two bipyridinium-

Cl-- bipyridinium arrays bound by π-π sideward stacking. The N-Cl distance were measured to be 3.36-

3.67 Å, similar to those of the staggered model. To determine whether this pattern was a coincidence,

larger shift and arbitrary shift directions of simulation input were tried with the same algorithm, but they

converged to almost identical ones like the above. We therefore proposed that this was a favored

conformation of PC-COF.

Figure S31. Simulated structure of charged eclipsed (slipped AA stacking) PC-COF form a) 001 direction,

b) 110 direction, and c) 1,-1,0 direction (gray: C, white: H, blue: N, green: Cl) and d) its predicted PXRD

patterns with real ones.

References

1) Zhao, Y.; Li, J.; Chun, C.; Yin, K.; Ye, D.; Jia, X. Green Chem. 2010, 12, 1370.

2) Chen, L.; Wang, H.; Zhang, D.-W.; Zhou, Y.; Li, Z.-T. Angew. Chem. Int. Ed. 2015, 54, 4028.

3) Accelrys, Material Studio Release Notes, Release 7.0, Accelrys Software, San Diego 2013.

4) Guo, J.; Xu, Y.; Jin, S.; Chen, L.; Kaji, T.; Honsho, Y.; Addicoat, M. A.; Kim, J.; Saeki, A.; Ihee,

H.; Seki, S.; Irle, S.; Hiramoto, M.; Gao, J.; Jiang, D. Nat. Commun. 2013, 4, 3736.

5) Dogru, M.; Sonnauer, A.; Gavryushin, A.; Knochel, P.; Bein, T.; Chem. Commun. 2011, 47, 1707.