Synthetic control of the pore dimension and surface area in conjugated microporous polymer and copolymer networks

Article abstract of DOI:10.1021/ja8010176

A series of rigid microporous poly(aryleneethynylene) (PAE) networks was synthesized by Sonogashira-Hagihara coupling chemistry. PAEs with apparent Brunauer-Emmet-Teller surface areas of more than 1000 m²/g were produced. The materials were found to have very good chemical and thermal stability and retention of microporosity under a variety of conditions. It was shown that physical properties such as micropore size, surface area, and hydrogen uptake could be controlled in a "quantized" fashion by varying the monomer strut length, as for metal-organic and covalent organic frameworks, even though the networks were amorphous in nature. For the first time, it was demonstrated that these properties can also be fine-tuned in a continuous manner via statistical copolymerization of monomer struts with differing lengths. This provides an unprecedented degree of direct synthetic control over micropore properties in an organic network.

Full text of DOI:10.1021/ja8010176

Published on Web 05/24/2008
Synthetic Control of the Pore Dimension and Surface Area in
Conjugated Microporous Polymer and Copolymer Networks
Jia-Xing Jiang, Fabing Su, Abbie Trewin, Colin D. Wood, Hongjun Niu,
James T. A. Jones, Yaroslav Z. Khimyak, and Andrew I. Cooper*
Department of Chemistry and Centre for Materials DiscoVery, UniVersity of LiVerpool,
LiVerpool L69 3BX, United Kingdom
Received February 15, 2008; E-mail: aicooper@liv.ac.uk
Abstract: A series of rigid microporous poly(aryleneethynylene) (PAE) networks was synthesized by
Sonogashira-Hagihara coupling chemistry. PAEs with apparent Brunauer-Emmet-Teller surface areas
2
of more than 1000 m /g were produced. The materials were found to have very good chemical and thermal
stability and retention of microporosity under a variety of conditions. It was shown that physical properties
such as micropore size, surface area, and hydrogen uptake could be controlled in a “quantized” fashion by
varying the monomer strut length, as for metal-organic and covalent organic frameworks, even though
the networks were amorphous in nature. For the first time, it was demonstrated that these properties can
also be fine-tuned in a continuous manner via statistical copolymerization of monomer struts with differing
lengths. This provides an unprecedented degree of direct synthetic control over micropore properties in an
organic network.
Introduction
air, atmospheric moisture, and, often, much more rigorous
conditions. These advantages notwithstanding, there are rela-
tively few approachs for the direct synthesis of microporous
Microporous materials (pore diameters <2 nm) are interesting
because of their applications in areas such as molecular
separations, heterogeneous catalysis, and gas storage. Significant
advances have been made recently, particularly in the prepara-
tion of microporous crystalline metal-organic frameworks
organic polymers with high specific surface areas (surface areas
2
>
1000 m /g). Polymers of intrinsic microporosity (PIMs) can
exhibit apparent Brunauer-Emmet-Teller (BET) specific sur-
face areas of 500-1065 m /g. Hyper-cross-linked polymers
HCPs) are a class of molecularly porous organic polymer
networks with BET surface areas that can exceed 2000 m /g.
We have shown in a recent paper that amorphous conjugated
microporous polymers (CMPs) can be synthesized with surface
2
7
1,2
3–6
(
MOFs) and covalent organic frameworks (COFs). In some
(
cases it has been possible to tune material properties such as
density, pore volume, and surface area for series of MOFs and
COFs by varying the length and geometry of the rigid organic
2
8–11
12
1
–6
linkers.
2
areas in the range 500-800 m /g by using Pd(0)/Cu(I)-catalyzed
Organic microporous polymers possess a number of advan-
tages. First, they can be composed solely of light elements (C,
H, O, N, etc.). Second, it is possible to achieve control over
functionality in organic polymers by a wide variety of synthetic
strategies. Third, most organic polymers are highly stable to
13
Sonogashira-Hagihara cross-coupling chemistry. These poly-
(
aryleneethynylene) (PAE) materials represent the first examples
(
7) (a) Budd, P. M.; Ghanem, B. S.; Makhseed, S.; McKeown, N. B.;
Msayib, K. J.; Tattershall, C. E. Chem. Commun. 2004, 230–231. (b)
McKeown, N. B.; Gahnem, B.; Msayib, K. J.; Budd, P. M.; Tattershall,
C. E.; Mahmood, K.; Tan, S.; Book, D.; Langmi, H. W.; Walton, A.
Angew. Chem., Int. Ed. 2006, 45, 1804–1807. (c) Ghanem, B. S.;
Msayib, K. J.; McKeown, N. B.; Harris, K. D. M.; Pan, Z.; Budd,
P. M.; Butler, A.; Selbie, J.; Book, D.; Walton, A. Chem. Commun.
2007, 67–69. (d) Budd, P. M.; Butler, A.; Selbie, J.; Mahmood, K.;
McKeown, N. B.; Ghanem, B.; Msayib, K.; Book, D.; Walton, A.
Phys. Chem. Chem. Phys. 2007, 9, 1802–1808.
(
1) (a) Yaghi, O. M.; Li, H. L.; Davis, C.; Richardson, D.; Groy, T. L.
Acc. Chem. Res. 1998, 31, 474–484. (b) Cheetham, A. K.; F e´ rey, G.;
Loiseau, T. Angew. Chem., Int. Ed. 1999, 38, 3268–3292. (c) Kitagawa,
S.; Kitaura, R.; Noro, S. Angew. Chem., Int. Ed. 2004, 43, 2334–
2
375. (d) Li, H.; Eddaoudi, M.; O’Keeffe, M.; Yaghi, O. M. Nature
999, 402, 276–279. (e) Mellot-Draznieks, C.; Ferey, G. Prog. Solid
1
State Chem. 2005, 33, 187–197. (f) Zhao, X. B.; Xiao, B.; Fletcher,
A. J.; Thomas, K. M.; Bradshaw, D.; Rosseinsky, M. J. Science 2004,
(8) (a) Tsyurupa, M. P.; Davankov, V. A. React. Funct. Polym. 2006, 66,
768–779. (b) Germain, J.; Hradil, J.; Fr e´ chet, J. M. J.; Svec, F. Chem.
Mater. 2006, 18, 4430–4435.
3
06, 1012–1015. (g) Lin, X.; Jia, J. H.; Zhao, X. B.; Thomas, K. M.;
Blake, A. J.; Walker, G. S.; Champness, N. R.; Hubberstey, P.;
Schr o¨ der, M. Angew. Chem., Int. Ed. 2006, 45, 7358–7364.
2) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keeffe,
M.; Yaghi, O. M. Science 2002, 295, 469–472.
(9) Lee, J. Y.; Wood, C. D.; Bradshaw, D.; Rosseinsky, M. J.; Cooper,
A. I. Chem. Commun. 2006, 2670–2672.
(
(
(
(10) Wood, C. D.; Tan, B.; Trewin, A.; Niu, H. J.; Bradshaw, D.;
Rosseinsky, M. J.; Khimyak, Y. Z.; Campbell, N. L.; Kirk, R.; St o¨ ckel,
E.; Cooper, A. I. Chem. Mater. 2007, 19, 2034–2048.
(11) Germain, J.; Fr e´ chet, J. M. J.; Svec, F. J. Mater. Chem. 2007, 17,
4989–4997.
3) C oˆ t e´ , A. P.; Benin, A. I.; Ockwig, N. W.; O’Keeffe, M.; Matzger,
A. J.; Yaghi, O. M. Science 2005, 310, 1166–1170.
4) El-Kaderi, H. M.; Hunt, J. R.; Mendoza-Cortes, J. L.; C oˆ t e´ , A. P.;
Taylor, R. E.; O’Keeffe, M.; Yaghi, O. M. Science 2007, 316, 268–
(12) Jiang, J. X.; Su, F.; Trewin, A.; Wood, C. D.; Campbell, N. L.; Niu,
H.; Dickinson, C.; Ganin, A. Y.; Rosseinsky, M. J.; Khimyak, Y. Z.;
Cooper, A. I. Angew. Chem., Int. Ed. 2007, 46, 8574–8578.
(13) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975,
4467–4470.
2
72.
(
5) C oˆ t e´ , A. P.; El-Kaderi, H. M.; Furukawa, H.; Hunt, J. R.; Yaghi, O. M.
J. Am. Chem. Soc. 2007, 129, 12914–12915.
(
6) Mastalerz, M. Angew. Chem., Int. Ed. 2008, 47, 445–447.
7
710 9 J. AM. CHEM. SOC. 2008, 130, 7710–7720
10.1021/ja8010176 CCC: $40.75  2008 American Chemical Society

Pore Dimension and Surface Area Control in Polymers
A R T I C L E S
of conjugated microporous polymer networks with high surface
area. We also showed that the micropore size distribution was
controlled in these PAE networks by the rigid node-strut
under a nitrogen atmosphere. Copper(I) iodide (50 mg) and
dichlorobis(triphenylphosphine)palladium(II) (150 mg) were added
to the mixture. (Trimethylsilyl)acetylene (1.96 g, 20.0 mmol) was
added, and the mixture was heated to 50 °C for 12 h. After the
mixture was cooled to room temperature, the precipitate that was
formed was filtered off and washed with ether. The crude
intermediate was purified by column chromatography (silica gel,
light petroleum) to give 4,4′-bis[(trimethylsilyl)ethynyl]biphenyl as
an intermediate. Deprotection was carried out by treatment with a
mixture of MeOH (50 mL)/NaOH (50 mL, 1 M) under stirring at
room temperature for 12 h. The product was isolated by evaporation
of the organic solvent, extraction of the residue with ether, drying
1
2
topology, in particular by the average strut length. This
suggests that extended long-range order as found in MOFs and
COFs is not necessarily a prerequisite for fine control over
micropore properties. This opens the door for a range of
14
amorphous polymer materials with controlled pore dimensions.
Very recently, we also prepared homocoupled conjugated
microporous poly(phenylenebutadiynylene)s with surface areas
2
15
of 840 m /g and mesoporous poly(phenylenevinylene)s with
2
16
surface areas of 761 m /g, although micropore size control in
these cases was complicated by phase behavior and competing
side reactions.
2 4
with Mg SO overnight, and finally removal of the solvent under
reduced pressure. The crude product was purified by column
chromatography (silica gel, light petroleum) to give 4,4′-diethy-
1
nylbiphenyl as white needlelike crystals (1.5 g, 74.2%). H NMR
A number of synthetic strategies have been employed
1
3
1
7,18
19,20
21,22
(CDCl
NMR (CDCl
8.08. Anal. Calcd for C16
3
): δ (ppm) 7.57 (d, 4H), 7.52 (d, 4H), 3.13 (s, 2H).
): δ (ppm) 140.58, 132.94, 127.18, 121.55, 83.41.
10: C, 95.05; H, 4.95. Found: C, 94.26;
C
previously to produce linear,
hyperbranched,
dendritic,
23
3
and cross-linked PAEs. Most soluble PAEs have incorporated
flexible solubilizing alkyl or alkoxyl substituents. Such materials
have been used as emissive layers in light-emitting diodes
and in nonlinear optics. There are very few reports concerning
the direct synthesis of porous PAE-type materials,
7
H
+
4
H, 4.97. MS: m/z 220.11 [M + NH ] .
1
8,20,22
Synthesis of Poly(aryleneethynylene) Networks. All of the
17
poly(aryleneethynylene) networks were synthesized by palladium-
1
2,15
although
catalyzed Sonogashira-Hagihara cross-coupling polycondensation
19
12,13
Kobayashi et al. have recently described the indirect prepara-
tion of porous pyrolytic polymers via treatment of alkyl-
substituted PAE precursors at high temperatures (>350 °C).
We describe here a series of PAE networks synthesized by
direct (A3 + B2) Pd-catalyzed cross-coupling polycondensation.
of arylethynylenes and aryl halides.
All reactions were carried
out using a 1.5 M excess of the ethynyl functionality since this
was found to maximize surface areas in the polymers (see the
Supporting Information, Table S2). A typical experimental proce-
2
4,25
dure for CMP-0 is given as follows:
1,3,5-Triethynylbenzene
(
450.5 mg, 3 mmol), 1,3,5-tris(4-iodophenyl)benzene (1368 mg,
Conjugated microporous PAEs with surface areas greater than
2
2.0 mmol), tetrakis(triphenylphosphine)palladium(0) (100 mg), and
copper(I) iodide (30 mg) were dissolved in a mixture of toluene
1
000 m /g were obtained, and the average micropore size was
controlled by the molecular dimensions of the monomers. The
results were rationalized using atomistic simulations of frag-
ments of these networks. For the first time, we show that
micropore size can be fine-tuned by statistical copolymerization
of monomers with different strut lengths; this can be done in a
continuous fashion with amorphous polymers but would be
difficult to achieve in microporous crystalline solids.
(
2.5 mL) and Et
3
N (2.5 mL). The reaction mixture was heated to
80 °C and stirred for 72 h under a nitrogen atmosphere to rigorously
exclude oxygen and to prevent homocoupling of the alkyne
monomers. The mixture was cooled to room temperature, and the
precipitated network polymer was filtered and washed four times
(once each) with chloroform, water, methanol, and acetone to
remove any unreacted monomer or catalyst residues. Further
purification of the polymers was carried out by Soxhlet extraction
with methanol for 48 h. The product was dried in vacuum for 24 h
at 70 °C. Yield: 67.3%. IR (KBr, cm- ): 3297.6 (-C′CsH), 2201.7
(-C′C-). Anal. Calcd for C H : C, 95.25; H, 4.75. Found: C,
Experimental Section
1
Chemicals. 1,3,5-Triethynylbenzene, 1,4-diethynylbenzene, 1,4-
diiodobenzene, 4,4′-diiodobiphenyl, 1,3,5-tris-(4-iodophenyl)ben-
zene, tetrakis(triphenylphosphine)palladium(0), (trimethylsilyl)acetyl-
ene, dichlorobis(triphenylphosphine)palladium(II), copper(I) iodide,
other chemicals, and solvents were all purchased from ABCR, TCI,
or Sigma-Aldrich and either recrystallized or used as received.
Synthesis of 4,4′-Diethynylbiphenyl Monomer. 4,4′-Diiodobi-
phenyl (4.06 g, 10.0 mmol) was dissolved in diethylamine (40 mL)
3
6
18
2
86.15; H, 4.41. Apparent BET surface area: 1018 m /g.
Gas Sorption Analysis. Surface areas and pore size distributions
were measured by nitrogen adsorption and desorption at 77.3 K
using either a Micromeritics ASAP 2420 or a Micromeritics ASAP
2020 volumetric adsorption analyzer. Samples were degassed at
110 °C for 15 h under vacuum (10- bar) before analysis. Hydrogen
isotherms were measured at 77.3 K up to 1.13 bar using a
Micromeritics ASAP 2420 volumetric adsorption analyzer with the
same degassing procedure.
5
(
(
14) Weder, C. Angew. Chem., Int. Ed. 2008, 47, 448–450.
15) Jiang, J. X.; Su, F.; Niu, H. J.; Wood, C. D.; Campbell, N. L.; Khimyak,
Y. Z.; Cooper, A. I. Chem. Commun. 2008, 486–488.
Isosteric Heats of Sorption. Heats of adsorption were deter-
mined from hydrogen adsorption isotherms up to a pressure of 1.13
bar at both liquid nitrogen (77.3 K) and argon (87.2 K) temperatures
(
16) Dawson, R.; Su, F.; Niu, H.; Wood, C. D.; Jones, J. T. A.; Khimyak,
Y.; Cooper, A. I. Macromolecules 2008, 41, 1591–1593.
10
using a Micromeritics ASAP 2420 instrument and the standard
(
17) (a) Moroni, M.; Lemoigne, J.; Luzzati, S. Macromolecules 1994, 27,
calculation routines in the Datamaster offline data reduction software
(Micromeritics).
5
62–571. (b) Weder, C.; Wrighton, M. S.; Spreiter, R.; Bosshard, C.;
Gunter, P. J. Phys. Chem. 1996, 100, 18931–18936.
(
18) (a) Weder, C.; Wrighton, M. S. Macromolecules 1996, 29, 5157–
5
165. (b) Babudri, F.; Colangiuli, D.; Di Bari, L.; Farinola, G. M.;
Omar, O. H.; Naso, F.; Pescitelli, G. Macromolecules 2006, 39, 5206–
(24) Since completing this series of experiments, we have found that
significantly lower Pd catalyst concentrations can be used while
obtaining very similar surface areas and pore properties in the resulting
polymers. For example, CMP-1 was prepared using [Pd] (1.0 mol %)
5
212.
(
(
19) Kobayashi, N.; Kijima, M. J. Mater. Chem. 2007, 17, 4289–4296.
20) Mendez, J. D.; Schroeter, M.; Weder, C. Macromol. Chem. Phys. 2007,
2
2
08, 1625–1636.
to give an apparent BET surface area of 864 m /g. This also decreases
(
21) (a) Peng, Z. H.; Pan, Y. C.; Xu, B. B.; Zhang, J. H. J. Am. Chem.
Soc. 2000, 122, 6619–6623. (b) Pan, Y. C.; Lu, M.; Peng, Z. H.;
Melinger, J. S. J. Org. Chem. 2003, 68, 6952–6958.
substantially the amount of Pd entrained in the polymer (see the
discussion regarding H2 isosteric heats for CMP-0-CMP-5).
(25) See the Supporting Information for full synthetic details for all
networks, as well as details on the influence of the monomer
concentration and monomer ratio upon the surface area (CMP-1-
(
(
22) Atas, E.; Peng, Z. H.; Kleiman, V. D. J. Phys. Chem. B 2005, 109,
1
3553–13560.
13
1
23) Trumbo, D. L.; Marvel, C. S. J. Polym. Sci., Part A: Polym. Chem.
986, 24, 2311–2326.
CMP-4), plus solid-state C{ H} MAS NMR for CMP polymers and
1
H2 sorption isotherms for networks CPN-1-CPN-6.
J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008 7711

A R T I C L E S
Jiang et al.
Table 1. Series of Five PAE Networks with Increasing Average Strut Lengths
a
b
Data from ref 12. Surface area calculated from the N2 adsorption isotherm using the BET method. The number in parentheses is the Langmuir
c
surface area calculated from the H2 adsorption isotherm by application of the Langmuir equation. Micropore surface area calculated from the N2
adsorption isotherm using the t-plot method. Total pore volume at P/P0 ) 0.99. Micropore volume derived using the t-plot method based on the
Halsey thickness equation. Average node-to-node strut length (measured between connected quaternary carbons) derived from polymer fragment
d
e
f
g
models. Average overall fragment diameter derived from polymer fragment models.
1
0,27
Solid-State NMR. Solid-state NMR spectra were measured on
COMPASS force field.
A seed molecule was constructed for
a Bruker Avance 400 DSX spectrometer operating at 100.61 MHz
each system by fully substituting the 1,3,5-benzene node monomer
with three of the corresponding linear connecting monomers. Each
model was built to have an average number molar mass in the region
of 11120-11210 g/mol. Any terminal halogen atoms on the
periphery of the fragments were replaced with hydrogen atoms for
clarity.
1
3
1
1
13
for C and 400.13 MHz for H. H- C cross-polarization magic
angle spinning (CP/MAS) NMR experiments were carried out at
an MAS rate of 10.0 kHz using zirconia rotors 4 mm in diameter.
1
The H π/2 pulse was 3.3 µs, and two-pulse phase modulation
2
6
(
TPPM) decoupling was used during the acquisition. The
Hartmann-Hahn condition was set using hexamethylbenzene. The
spectra were measured using a contact time of 2.0 ms and a
relaxation delay of 10.0 s. Typically 2048 scans were accumulated.
Results and Discussion
Effect of the Monomer Strut Length on the Pore Struc-
ture. We showed in a preliminary paper that the pore structure
1
13
The H- C dipolar dephasing spectra were recorded at an MAS
12
rate of 10 kHz and using dephasing delays of 20, 50, 100, and 200
1
13
for a series of three PAE networks could be varied by changing
the length of the rigid connecting strut, much as for crystalline
µs. The variable contact time H- C CP/MAS NMR spectra were
1
3
1
measured using t
m
) 0.02-12.0 ms. The C{ H} MAS NMR
1
–6
MOFs and COFs,
even though these PAE networks were
spectra were measured at an MAS rate of 10.0 kHz using TPPM
1
3
decoupling. The C π/2 pulse was 3.5 µs. The spectra were
measured using a recycle delay of 30.0 s. Typically, 1024 scans
were accumulated. The values of the chemical shifts are referred
to that of TMS. The analysis of the spectra (deconvolution and
integration) was carried out using Bruker TOPSPIN software.
Atomistic Simulations. Molecular models for the network
fragments were generated using the Materials Studio Modeling 4.0
package (Accelrys Inc., San Diego, CA, 2005). The fragments were
constructed by adding the repeating unit for each system in a
stepwise manner. All models fully relaxed using the Discover
molecular mechanics and dynamics simulation module with the
amorphous. In this study, we extend this concept by synthesizing
networks with shorter (CMP-0) and longer (CMP-5) struts
(Table 1). All five polymer networks were designed to have a
three-dimensional net structure arising from the three-pronged
aryleneethynylene linkages (Scheme 1). As shown in Scheme
1, CMP-1 and CMP-4 had the same notional repeat unit structure
but CMP-1 was produced from 1,3,5-triethynylbenzene and 1,4-
diiodobenzene (A3 + B2), whereas CMP-4 was produced from
1
,3,5-tribromobenzene and 1,4-ethynylbenzene (A2 + B3); as
such, these materials are topologically identical but produced
by two different routes. All polymerization reactions were
(
26) Bennett, A. E.; Rienstra, C. M.; Auger, M.; Lakshmi, K. V.; Griffin,
R. G. J. Phys. Chem. 1995, 103, 6951–6958.
(27) Sun, H. J. Phys. Chem. B 1998, 102, 7338–7364.
7
712 J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008

Pore Dimension and Surface Area Control in Polymers
A R T I C L E S
a
Scheme 1. Representative Molecular Structures for Networks CMP-0-5
a
The benzene “nodes” in the networks (derived from the trifunctional A3 or B3 monomer in each case) are shown in red.
carried out at a fixed molar halogen monomer concentration
400 mmol/L), a fixed reaction temperature (80 °C), and a
croporous network where the pore dimensions are controlled
primarily by the molecular structure of the network, rather than
(
2
8
reaction time of 72 h. It was found that the surface area and
pore size distribution for these networks were relatively unaf-
fected even by large changes (×10) in the monomer concentra-
tion (Supporting Information, Table S1 and Figure S1). This
contrasts with macroporous vinyl polymers, for example, where
surface areas can be very sensitive to even small changes in
by phase behavior and phase separation. The molar ratio of
ethynyl to halogen functionalities in this “strut length series”
was fixed at 1.5:1 since maximum surface areas and H2 uptakes
were observed consistently at this molar ratio (see also the
Supporting Information, Table S2).
All polymers in the series were characterized at the molecular
2
8,29
1
13
monomer concentration.
This is consistent with a mi-
level by H- C CP/MAS NMR, and the assignment of the
J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008 7713

A R T I C L E S
Jiang et al.
1
13
Figure 1. Solid-state H- C CP/MAS NMR spectra for conjugated microporous PAE networks recorded at an MAS rate of 10 kHz. Asterisks denote
spinning sidebands.
1
13
resonances (Figure 1) was confirmed using H- C CP/MAS
kinetics and dipolar dephasing experiments. The low-intensity
lines at ca. 76 and 82 ppm can be ascribed to -Ct CH end
groups (the line at 82 ppm corresponds to the quaternary
acetylene carbons). The end group content was found to be lower
in the polymers produced from 1,4-ethynylbenzene (CMP-3)
and 4,4′-diiodobiphenyl (CMP-5), possibly because these longer
struts give rise to a greater degree of overall conformational
freedom, which allows more efficient linking (i.e., fewer
unreacted end groups) in the resulting network. The ratio of
the intensities of the acetylenic and aromatic peaks was
bearing residual iodine atoms. The relative intensities of the
peaks for all samples are consistent with their proposed
formulae, although it should be noted that the accurate deter-
mination of populations of different sites is hampered by both
the similarity of chemical shifts for sites in different monomers
and the significant overlap of lines in the spectra of the polymer
networks. A more detailed analysis of the dynamic and structural
heterogeneities of these materials using advanced solid-state
NMR methodology is in progress. These insoluble networks
have some similarities to soluble hyperbranched polymers, and
more quantitative NMR data should allow the degree of
branching to be calculated.
1
13
calculated using variable contact time H- C CP/MAS NMR
spectra as follows: CMP-0, 0.12 (expected value 0.18); CMP-
The porous structure of this series of PAE networks was
investigated by gas sorption analysis using N and H . Figure
1
0
(
, 0.27 (expected value 0.40); CMP-2, 0.18 (expected value
.25); CMP-3, 0.10 (expected value 0.18); CMP-5, 0.08
expected value 0.14). These values were verified using C{ H}
MAS NMR spectra. This trend follows the expected values
in accordance with the nominal structural formulae presented
in Scheme 1, calculated assuming a complete condensation of
2
2
2a shows the N adsorption and desorption isotherms for the
2
13
1
resulting polymers, vertically offset for clarity. All networks
2
5
gave rise to type I N gas sorption isotherms according to
2
3
0
IUPAC classifications. The apparent BET surface area for
-
Ct CH groups and ideal stoichiometry of the resulting
(
28) Sherrington, D. C. Chem. Commun. 1998, 2275–2286.
polymeric frameworks. The higher than predicted content of
aromatic groups indicates that these linkages are likely to be
the dominant chain ends in the polymer network. The spectrum
of CMP-1 shows a broad shoulder at ca. 137 ppm which may
originate from the protonated carbons of aromatic end groups
(29) (a) Cooper, A. I.; Holmes, A. B. AdV. Mater. 1999, 11, 1270–1274.
(b) Hebb, A. K.; Senoo, K.; Bhat, R.; Cooper, A. I. Chem. Mater.
2
003, 15, 2061–2069.
(
30) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti,
R. A.; Rouquerol, J.; Siemieniewska, T. Pure Appl. Chem. 1985, 57,
603–619.
7
714 J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008

Pore Dimension and Surface Area Control in Polymers
A R T I C L E S
CMP-1, CMP-4 had a slightly lower surface area; this may be
a result of the lower reactivity of aryl bromides with respect to
aryl iodides in Pd/Cu-catalyzed cross-coupling reactions. In
general, we have found that, all other things being equal, iodine-
substituted monomers give rise to slightly higher surface area
PAE materials.
Atomistic simulations were used to build representative
fragments for each of the five networks. Each fragment was
built to have a similar molar mass (between 11 120 and 11 210
g/mol). We believe that the amorphous three-dimensional
structure of these PAE networks arises from the bending of the
struts out of plane of the benzene nodes, from the dihedral angle
between struts on adjacent connected nodes, from bending of
the struts themselves, and from in-plane deviation of the angle
between struts from the hypothetical 120° imposed by the 1,3,5-
substituted node geometry. CMP-0 has the shortest strut length
of 0.82 nm. However, the presence of the 1,3,5-triphenyl-
substituted benzene node leads to a broadened distribution range
of in-plane and out-of-plane bending angles (103-131° and
3
-57°, respectively). CMP-1, while having a longer average
strut length of 1.11 nm, has the most geometrically simple
structure: the out-of-plane bending of struts ranges from 0° to
10°, while in-plane angles range from 115° to 130°. The greatest
flexibility is seen for the simulation of CMP-5, which contains
both biphenyl moieties and the 1,3,5-triphenyl-substituted
benzene node. This model also has the longest strut length,
Lav(strut), of 2.550 nm. The in-plane and out-of-plane bending
angles in this model range between 70° and 143° and between
2
6° and 95°, respectively.
Figure 3 shows cluster fragments for CMP-0-CMP-5 derived
Figure 2. (a) N2 adsorption-desorption isotherms measured at 77.3 K (the
from these molecular models. The overall average fragment
diameter, Lav(cluster) in Table 1, increases in the series CMP-
0-CMP-5 in keeping with the increase in the average strut
length, Lav(strut). The increased degree of conformational freedom
in the extended struts in this series of models allows for greater
intramolecular intercalation and space-filling within the frag-
ments, and hence, the increase in cluster size does not scale in
direct proportion to the average strut length. We believe that
this conformational freedom may also contribute to the reduction
in micropore volume for the networks with longer struts (Table
1). These models represent fragments of the networks rather
than the pore structure itself. It is likely, for example, that there
is significantly more intermolecular catenation and entanglement
of fragments in the actual samples, consistent with the fact that
these fragment models significantly overestimate the micropore
volume in comparison with measured values. The hydrogen
sorption properties of the polymers were also investigated by
adsorption branch is labeled with filled symbols). For clarity, the isotherms
of CMP-0, CMP-1, CMP-2, and CMP-3 were shifted vertically by 400,
3
3
00, 200, and 100 cm /g, respectively. (b) NL-DFT pore size distribution
curves. The dashed arrow indicates the shift to larger average pore size in
this series. (c) Cumulative pore volume curve calculated by application of
NL-DFT.
2
these networks varied between 512 m /g (CMP-5) and 1018
2
m /g (CMP-0). The small rise in the N2 uptake at high relative
pressures (P/P0 > 0.8) in the adsorption isotherms for CMP-2
and CMP-5 may stem from interparticulate porosity associated
with the meso- and macrostructures of the samples. Figure 2b
shows pore size distribution curves for the five CMP networks
using the nonlocal density functional theory (NL-DFT). In
general, the micropore size distribution is shifted systematically
to larger pore diameters for the series CMP-0-CMP-5 as the
monomer strut length is increased, and the overall micropore
3
volume also falls (from 0.38 to 0.16 cm /g) over this series.
9
,10
This is illustrated in the NL-DFT cumulative pore volume plots
for the five networks (Figure 2c). In particular, CMP-3 and
CMP-5, which were formed using longer struts (Scheme 1,
Table 1) exhibit almost no ultramicropore volume below a pore
diameter of about 0.9 nm. The observations for CMP-0 and
volumetric methods.
Figure 4 shows the H2 adsorption
isotherms for the PAE networks at 77.3 K, up to a maximum
H2 pressure of 1.13 bar. CMP-0 has the smallest strut length,
the highest micropore volume, and the highest apparent BET
surface area, and this sample exhibited the largest H2 uptake
1
2
3
3
CMP-5 are consistent with our previous results and further
illustrate that micropore dimensions and surface areas can be
tuned synthetically by varying the monomer strut length. For
example, CMP-0 (which has the shortest struts in this series)
(155 cm /g, ∼1.4 wt %; approximately 215 cm /g at 8 bar of
3
H2). CMP-5 showed the lowest H2 uptake (67 cm /g, ∼0.6 wt
%). In general, there was a systematic change in H2 uptake as
a function of molecular strut length in the polymers (Figure 4).
These results can be explained by the fact that ultramicropores,
rather than “large” micropores or mesopores, contribute mostly
to H2 adsorption at this pressure and temperature; as such, H2
is a sensitive probe for the variation in micropore structure in
this series. The H2 uptake per unit BET surface area (as
calculated by N2 sorption) for these PAE networks is signifi-
cantly greater than observed for our highest surface area hyper-
2
has an apparent BET surface area (1018 m /g) which is higher
than that of CMP-1 or any of the other networks reported in
1
2
our preliminary paper. The polymer CMP-4, which was
topologically equivalent to CMP-1, exhibited porous properties
2
2
3
(
)
SBET ) 744 m /g, Smicro ) 596 m /g, Vtotal ) 0.39 cm /g, Vmicro
3
0.29 cm /g) which were very similar to those of CMP-1,
although it was produced by a different route. Compared to
J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008 7715

A R T I C L E S
Jiang et al.
Figure 3. Atomistic simulations for PAE networks with different strut lengths. (a) Node-strut topology for simulated network fragments for CMP-0 (left),
CMP-1 (center), and CMP-5 (right). A single 1,3,5-connected benzene node connecting three other nodes via rigid struts is highlighted (in gray/white) in
each case. (b) Atomistic simulations of network fragments for CMP-0, CMP-1, CMP-2, CMP-3, and CMP-5 (left to right). The solvent-accessible surface
is shown (in green) in each case, as calculated using a solvent diameter of 0.182 nm. Each model was built to have an average number molar mass in the
range 11120-11210 g/mol.
Figure 5. Isosteric heats of sorption for H2 on networks CMP-0-5.
33
Figure 4. Volumetric H2 adsorption isotherms for CMP-0-5 at 77.3 K.
centers. The isosteric heats at higher H2 coverages for networks
such as CMP-5 are comparable to values reported recently by
1
0
11
cross-linked polymers under equivalent conditions (e.g., 1.36
Svec and colleagues for hyper-cross-linked polyanilines.
A
-5
2
-6
2
×
10 g of H2/m for CMP-0 versus 8.46 × 10 g of H2/m
preponderance of small ultramicropores (<0.7 nm) might in
principle give rise to multiwall interactions and enhanced
isosteric heats for H2, but no simple correlation was observed
between the average pore size and isosteric heat for these
materials. In fact, CMP-5 s which had the largest average
micropore size s exhibited the highest overall isosteric heat. It
should be noted, however, that these materials vary in both
chemical structure (Scheme 1) and average micropore size and
do not in this respect form a fully homologous series. A number
of factors other than pore size may influence isosteric heats.
First, these materials contain small, variable quantities of residual
palladium, copper, and iodine (typically <1 atom % for each
10
for hyper-cross-linked polymer 43). Similarly, the H2 sorption
per unit micropore volume for CMP-0 is about 20% higher than
-2
-2
3
for 43 (3.63 × 10 and 2.98 × 10 g/cm , respectively).
The isosteric heat of sorption for H2 on these PAE materials
was calculated as a function of H2 coverage from comparison
1
0,31
of the adsorption isotherms at 77.3 and 87.2 K (Figure 5).
All five networks exhibited low gas coverage enthalpies of
adsorption of ∼10 kJ/mol, falling to between 4 and 8 kJ/mol
with increasing H2 coverage. In general, these isosteric heats
are higher than we observed for hyper-cross-linked polymers.
The initial isosteric heats at low H2 coverage for the PAE
10
2
5
element). Species such as Pd could certainly influence H2
sorption, but we did not find any clear correlation between the
quantity of residual Pd, Cu, or I and the average isosteric heat
3
2
networks are comparable to the highest reported for MOFs,
despite the fact that these polymers do not contain charged metal
(Supporting Information, Figure S3). Second, ethynyl end groups
are present in these networks, and it was noted, for example,
(
(
31) (a) Dinc a˜ , M.; Yu, A. F.; Long, J. R. J. Am. Chem. Soc. 2006, 128,
8
904–8913. (b) Rowsell, J. L. C.; Yaghi, O. M. J. Am. Chem. Soc.
006, 128, 1304–1315.
2
(33) Panella, B.; H o¨ nes, K.; M u¨ ller, U.; Trukhan, N.; Schubert, M.; P u¨ tter,
32) Dinc a˜ , M.; Long, J. R. J. Am. Chem. Soc. 2007, 129, 11172–11176.
H.; Hirscher, M. Angew. Chem., Int. Ed. 2008, 47, 2138–2142.
7
716 J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008

Pore Dimension and Surface Area Control in Polymers
A R T I C L E S
Controlling the Micropore Size and Surface Area by
Statistical Copolymerization. A well-known advantage of
organic polymers is the ability to produce copolymers over a
continuous range of relative monomer compositions and hence
achieve systematic control over material properties such as glass
transition temperature, melting temperature, solubility, and pH
response. Statistical copolymerization has not until now been
used to fine-tune the micropore size in materials. Here, the
amorphous nature of these PAE systems is an advantage. While
pore properties in MOFs and COFs have been varied systemati-
1
–6
cally by changing the connecting ligands, the structures are
quantized” in terms of the unit of variation, for example, the
“
difference in length between ligands in a series of materials.
As such, the gradations between structures in a series may be
small but are nevertheless finite, and this strategy requires that
each ligand in the series reacts in a similar way to form a stable,
crystalline porous structure. Similarly, while MOF “copolymers”
Figure 6. Thermogravimetric analysis for PAE networks CMP-0-CMP-
5
. The heating rate is 5 °C/min under nitrogen.
3
5
have been prepared using more than one connecting ligand,
that the two polymers with the highest low-coverage isosteric
heats (CMP-5 and CMP-3) appeared to have lower -Ct CH
end group concentrations (see the NMR discussion above).
One cannot infer from these data alone, however, that this is
a dominant factor. Third, the various networks in this series
have different electronic conjugation patterns and varying
alternation of benzene and alkyne functionalities (Scheme
the physical properties of the materials are a result of the
accessible thermodynamically stable crystal structures and are
not continuously variable as a function of the relative ligand
ratio. By contrast, it should be possible to “tune” properties such
as micropore size and surface area in amorphous PAE networks
in a completely continuous fashion, such that the degree of fine
control over these properties is limited only by the inherent
variability in the polymerization process itself. To test this idea,
we synthesized a series of copolymers of 1,3,5-triethynylbenzene
with 1,4-diiodobenzene and/or 4,4′-diiodobiphenyl (CPN-1-6;
Table 2, Scheme 2). In principle, the average micropore size
and surface area would be expected to increase systematically
as the percentage of the shorter monomer strut, 1,4-diiodoben-
zene, is increased.
1
). Second-order Møller-Plesset calculations for small
34
aromatics and fused-ring systems have suggested that the
interaction energy between H2 and benzene π-systems is
affected by the degree of conjugation and also by the presence
of substituents. For example, extended conjugation (coronene
versus benzene) led to enhanced H2-arene interactions. For
monosubstituted benzenes such as cyanobenzene or fluo-
robenzene, H2 was found to interact preferentially with the
3
4
To check that the two comonomers had similar reactivities,
the relative monomer incorporation in the networks was
3
4
center of the aromatic ring in an end-on fashion. As such,
the presence of ethynyl or halogen end groups in our
polymers might influence interactions with H2, either directly
as substituents or because these groups are indicative of a
lower degree of cross-linking and, thus, a reduction in
extended electronic conjugation. It is therefore very hard at
this time to fully deconvolute the potential effects on the
isosteric heat of the pore size distribution, chemical func-
tionality, and electronic conjugation for these materials.
These porous PAE networks were found to have good
chemical and thermal stability and did not require any special
handling to maintain microporosity. The workup procedure
involved multiple washing steps in air followed by extended
Soxhlet extraction. The polymers were also stable toward
water; indeed, no change in surface area was observed when
the materials were treated with aqueous solutions of acid or
base. By contrast, MOFs and COFs can vary in stability from
1
13
investigated by solid-state NMR spectra (Figure 7). The H- C
CP/MAS solid-state NMR spectra for the CPN networks
confirmed the presence of two different types of building units
in the copolymers. The intensity of the peaks at 140.4 and 126.1
ppm, which are characteristic of the 4,4′-diiodobiphenyl mono-
mer, increased systematically with its content in the initial
monomer feed.
The BET surface areas for these networks were found to
2
2
decrease from 856 m /g for CPN-1 to 643 m /g for CPN-6, as
measured by nitrogen sorption. Figure 8a shows the full nitrogen
adsorption and desorption isotherms for the six copolymer
networks. All of the copolymers gave rise to type I gas sorption
isotherms, indicating that all of the materials were microporous.
Figure 8b shows the NL-DFT pore size distribution curves for
the six networks. The micropore size distribution is shifted to
larger pore diameters for the series of networks from CPN-1 to
CPN-6 as the ratio of the longer strut monomer, 4,4′-diiodobi-
phenyl, increases. This is shown by the NL-DFT cumulative
pore volume plots for these networks (Figure 8c). Even though
1
–6
being highly air sensitive to being quite robust. In general,
however, the coordinate bonding which allows the broad
structural diversity in crystalline MOFs and COFs may not
in general be compatible with harsher physical conditions,
such as acidic and basic environments, and hence, these PAE
materials may offer some advantages in such applications.
The microporous PAE networks were also found to have good
thermal stability as a result of their cross-linked structure
and strong covalent bonds. The polymers showed little
thermal decomposition under nitrogen gas below temperatures
of around 400 °C (Figure 6).
(
35) (a) Wang, C. C.; Kuo, C. T.; Yang, J. C.; Lee, G. H.; Shih, W. J.;
Sheu, H. S. Cryst. Growth Des. 2007, 7, 1476–1482. (b) Wang, C. C.;
Tseng, S. M.; Lin, S. Y.; Liu, F. C.; Dai, S. C.; Lee, G. H.; Shih,
W. J.; Sheu, H. S. Cryst. Growth Des. 2007, 7, 1783–1790. (c) Yang,
E. C.; Zhao, H. K.; Ding, B.; Wang, X. G.; Zhao, X. J. Cryst. Growth
Des. 2007, 7, 2009–2015.
(36) (a) Deng, W. Q.; Xu, X.; Goddard, W. A. Phys. ReV. Lett. 2004, 92,
No. 166103. (b) Lee, H.; Choi, W. I.; Ihm, J. Phys. ReV. Lett. 2006,
9
7, No. 056104.
(
37) Cooper, A. I.; Poliakoff, M. Chem. Commun. 2007, 2965–2967.
(
34) H u¨ bner, O.; Gl o¨ ss, A.; Fichtner, M.; Klopper, W. J. Phys. Chem. A
004, 108, 3019–3023.
(38) Trewin, A.; Darling, G. R.; Cooper, A. I. New J. Chem. 2008, 32,
2
17–20.
J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008 7717

A R T I C L E S
Jiang et al.
a
Table 2. Copolymerization with Different Molar Ratios of Two Halogen Monomers
a
b
All reactions were carried out under a monomer concentration of 400 mmol/L, 80 °C/72 h. Surface area calculated from the N2 adsorption
isotherm using the BET method. The number in parentheses is the Langmuir surface area calculated from the H2 adsorption isotherm by application of
c
d
the Langmuir equation. Micropore surface area calculated from the N2 adsorption isotherm using the t-plot method. Total pore volume at P/P0 )
e
f
0
.99. Micropore volume derived using the t-plot method based on the Halsey thickness equation. Data were obtained at 1.13 bar and 77.3 K.
Scheme 2. Synthetic Route of Copolymerization for CPN-1-6
the changes in the cumulative curves are small (particularly for
CPN-2-CPN-4), the shift is systematic and all six cumulative
curves fall in the predicted order. Likewise, the hydrogen
adsorption isotherms for the six copolymers (Supporting Infor-
mation, Figure S4) followed the same trend and order, again
suggesting that H2 uptake is a sensitive probe for the micropore
surface area in materials of this type. The systematic variation
in surface area for these PAE materials as a function of
comonomer ratio is shown in Figure 9. These data also illustrate
the high reproducibility of these polymerization reactions:
networks CPN-1 and CPN-6 are repeats of CMP-1 and CMP-
2, respectively, and both the BET surface areas and H2 isotherm
Langmuir surface areas differ by less than (3% for the two
pairs of samples (cf. Tables 1 and 2). In effect, this series of
CPN copolymers corresponds to the materials that lie “between”
networks CMP-1 and CMP-2 in the aforementioned strut-length
series, as illustrated graphically in Figure 9 (also cf. Figures 2c
and 8c). This demonstrates the essentially continuous synthetic
variation that can be achieved in terms of micropore size and
surface area, at least within the apparent (3% reproducibility
of the system, and we believe that this degree of direct,
continuous synthetic control is unprecedented for any mi-
croporous material.
While control over surface area and pore size was very good,
the isosteric heat of sorption for H2 was found, as for the CMP
series, not to fall into a simple, rational order (Supporting
Information, Figure S5). For example, we did not observe the
isosteric heat to simply increase with decreasing pore size. Also,
the isosteric heat for CNP-1 was found to be slightly lower than
observed for CPN-6, which is the reverse trend observed for
the corresponding samples CMP-1 and CMP-2. Again, this
suggests that variability at the molecular level may be dominat-
ing. It should be stressed however that these differences are
relatively small (low-coverage heats for the whole CPN series
vary between 6.0 and 7.5 kJ/mol) and that all of these energies
are much too low, for example, to allow close-to-ambient-
temperature H2 sorption.
1
13
Figure 7. Solid-state H- C CP/MAS NMR spectra for conjugated
microporous CPN networks recorded at an MAS rate of 10 kHz. Asterisks
denote spinning sidebands.
These PAE materials, for example, CMP-0, exhibit ap-
parent BET surface areas that are higher than or equal to
7
718 J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008

Pore Dimension and Surface Area Control in Polymers
A R T I C L E S
1
,2
3–6
some similarity to MOFs and COFs.
Indeed, CMP-1
2
2
(
SABET ) 834 m /g, Smicro ) 675 m /g, pore volume 0.33
cm /g) exhibited a N2 isotherm and pore size distribution
Figure 2) which were both very similar to those reported
3
(
2
2
for COF-1 (SABET ) 711 m /g, Smicro ) 587 m /g, pore
3
3
volume 0.32 cm /g). The properties of CMP-2 were also
similar to those of COF-1; in general, CMP-1 was somewhat
more microporous than COF-1, whereas CMP-2 was some-
3
what less so. COF-1 is crystalline, whereas CMP-1 is
1
2
amorphous; nevertheless, the close similarity in sorption
properties may be more than superficial since both materials
are composed of a rigid 1,3,5-substituted node-strut topol-
ogy. Moreover, in both cases, struts and nodes may form
“
eclipsed” conformationssCOF-1 by eclipsing of extended
3
ordered hexagonal sheets and CMP-1 by interpenetration
of the amorphous polymer network fragments (see the
simulation discussion above). While it is possible to continu-
ously fine-tune the surface area and micropore size by
changing the strut length for PAEs, the correspondence
between “design” and structure is perhaps more indirect than
observed for COFs. For example, COF-10 (with a relatively
long biphenyldiboronic acid strut) was reported to be
5
mesoporous. By contrast, PAEs with extended strut lengths
(
CMP-2-CMP-5) showed a shift to larger average micropore
sizes (Figure 2), but all networks were microporous, not
mesoporous, and exhibited type I isotherms. Similarly, the
pore volume for the COF series (COF-6, COF-8, COF-12)
5
was found to increase with increasing spacer length, whereas
the pore volume for our PAE materials decreases in the series
CMP-0-CMP-5. This is a fundamental difference between
these two classes of material and reflects the fact that the PAE
pore structure is not crystallographically defined by the monomer
structure. Rather, we propose that the average pore size distribution
envelope is related statistically to the average strut length and
polymer topology and that the decrease in overall pore volume
with increasing strut length is a consequence of greater confor-
mational freedom and chain interpenetration, as suggested by
atomistic simulations. We have exploited this, uniquely, by
continuous fine-tuning of pore size distributions and surface areas
via statistical copolymerization of monomers with different strut
lengths (Figure 9). Our PAE materials exhibit apparent BET surface
areas which are much lower than the highest reported for COFs
Figure 8. (a) N2 adsorption-desorption isotherms measured at 77.3 K (the
adsorption branch is labeled with filled symbols, and the desorption branch
is labeled with empty symbols). For clarity, the isotherms of CPN-1, CPN-
2
2
, CPN-3, CPN-4, and CPN-5 were shifted vertically by 500, 400, 300,
00, and 100 cm /g, respectively. (b) Pore size distribution curves. (c) NL-
3
DFT cumulative pore volume curve.
2
4
(e.g., 4210 m /g for COF-103). It may be very difficult to approach
surface areas of this magnitude in amorphous polymers such as
these without introducing new strategies: for example, molecular
templating approaches to significantly enhance the creation of
micropore volume. Nevertheless, PAE materials may have
unique advantages as a result of combining fine control over
micropore properties with very good chemical and thermal
stability. These PAEs are also highly conjugated, unlike MOFs,
COFs, PIMs, or hyper-cross-linked polymers. There is therefore
a wealth of opportunity for producing robust microporous
materials with specific functionality; for example, by attaching
metals to these networks to facilitate catalysis or new electrical
properties or to introduce high-binding energy sites such as
Figure 9. Surface area as a function of polymer structure for polymers
CMP-1-CMP-5 (closed symbols) and copolymers CPN-1-CPN-6 (open
symbols). Both the apparent N2 BET surface area (circles) and the Langmuir
surface area (as calculated from the H2 adsorption isotherm, squares) are
shown. The series of CPN copolymers falls between the respective
homopolymers, CMP-1 and CMP-2, for both the N2 and H2 sorption
measurements.
36
37
38
metals, transition-metal complexes, or “naked” halides for
gas storage applications.
7
,28,29
those of most microporous organic polymers,
although
Conclusion
substantially lower than those observed for some hyper-cross-
8
–11
linked materials.
Unlike hyper-cross-linked polymers,
We have synthesized microporous conjugated PAE networks
with BET surface areas of more than 1000 m /g and shown
that the surface area and pore size can be controlled by the
monomer structure. For the first time, we have shown that the
2
however, it is possible to fine-tune the micropore size and
surface area in PAE networks by systematic variation of the
monomer strut length. In this regard, PAE networks have
J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008 7719

A R T I C L E S
Jiang et al.
average micropore size and surface area can be fine-tuned in a
continuous fashion by statistical copolymerization, something
which is possible with amorphous copolymers but very difficult
to achieve with crystalline materials. This may be useful in
applications such as molecular separations where particular
optimal pore sizes are required. These PAE networks combine
some of the advantages of microporous and mesoporous carbon,
organic polymers, while offering an unprecedented degree of
pore size control for an amorphous microporous polymer system.
Acknowledgment. We are grateful to EPSRC (Grant EP/
C511794/1) and the University of Liverpool for funding. We thank
Dr. N. L. Campbell for carrying out the EDX measurements.
Supporting Information Available: Full synthetic details,
3
9
13
1
for example, chemical robustness, tunable pore size, and
extended conjugation, with the broad synthetic versatility of
solid-state C{ H} MAS NMR for CMP polymers, and H
2
sorption isotherms and isosteric heat data for networks CPN-
-CPN-6. This material is available free of charge via the
1
Internet at http://pubs.acs.org.
(39) Jun, S.; Joo, S. H.; Ryoo, R.; Kruk, M.; Jaroniec, M.; Liu, Z.; Ohsuna,
T.; Terasaki, O. J. Am. Chem. Soc. 2000, 122, 10712–10713.
JA8010176
7
720 J. AM. CHEM. SOC. 9 VOL. 130, NO. 24, 2008