Micro- and mesoporous polycyanurate networks based on triangular units

Article abstract of DOI:10.1002/cplu.201300090

Hyper-cross-linked polycyanurate networks (CE-P1 and CE-P2) were synthesized by means of thermal self-cyclotrimerization from two triangular cyanate resin monomers 1,3,5-tri(4-cyanatophenyl)benzene and 1,3,5-tricyanatobenzene, respectively. Interestingly, it was found that CE-P1 exhibited microporous characteristics and a moderately large BET surface area. The two narrow peaks in the nonlocal density functional theory (NLDFT) curve appeared at 0.57 and 1.01 nm. In contrast, the CE-P2 sample had a small surface area and broad pore-size distribution with major pores of around 3.39 nm, which indicated a mesoporous material. The reason for this was interpreted in terms of the geometric configuration, steric hindrance, and reactivity of the cyanate monomers. The adsorptions of CO₂, H₂, benzene, n-hexane, and water vapors were investigated by correlating the data with the porosity parameters, chemical structure, and composition of the two networks. The results showed that the vastly distinct pore properties significantly influenced the adsorptions of gases and vapors. In particular, organic vapors such as benzene and n-hexane tended to be adsorbed on the pore surface owing to their affinity and thereby the adsorption amounts were tightly attached to the surface area of the samples. On the contrary, the hydrophobic nature of polymers made the water molecules preferentially condense within the pores so that the pore size rather than the surface area became the dominant factor influencing the adsorption of water vapor. Copyright

Full text of DOI:10.1002/cplu.201300090

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201300090
Micro- and Mesoporous Polycyanurate Networks Based on
Triangular Units
Hao Yu, Changjiang Shen, and Zhonggang Wang*[a]
Hyper-cross-linked polycyanurate networks (CE-P1 and CE-P2)
were synthesized by means of thermal self-cyclotrimerization
from two triangular cyanate resin monomers 1,3,5-tri(4-cyana-
tophenyl)benzene and 1,3,5-tricyanatobenzene, respectively.
Interestingly, it was found that CE-P1 exhibited microporous
characteristics and a moderately large BET surface area. The
two narrow peaks in the nonlocal density functional theory
(NLDFT) curve appeared at 0.57 and 1.01 nm. In contrast, the
CE-P2 sample had a small surface area and broad pore-size dis-
tribution with major pores of around 3.39 nm, which indicated
a mesoporous material. The reason for this was interpreted in
terms of the geometric configuration, steric hindrance, and re-
activity of the cyanate monomers. The adsorptions of CO2, H2,
benzene, n-hexane, and water vapors were investigated by
correlating the data with the porosity parameters, chemical
structure, and composition of the two networks. The results
showed that the vastly distinct pore properties significantly in-
fluenced the adsorptions of gases and vapors. In particular, or-
ganic vapors such as benzene and n-hexane tended to be ad-
sorbed on the pore surface owing to their affinity and thereby
the adsorption amounts were tightly attached to the surface
area of the samples. On the contrary, the hydrophobic nature
of polymers made the water molecules preferentially condense
within the pores so that the pore size rather than the surface
area became the dominant factor influencing the adsorption
of water vapor.
Introduction
The development of microporous organic polymers has
emerged as a hot topic in polymer science and materials di-
rected toward applications in gas storage, gas sensors, mem-
brane separation, heterogeneous catalysis, microreactors, and
so forth.[1] Polycyanurates, that is, cured cyanate resins, are one
of the most important classes of thermosetting polymers, and
are widely used in structural adhesives, electronics, and resin
matrixes for high-performance composite materials in aeronau-
tics and astronautics industries owing to their low dielectric
loss, high heat resistance, excellent dimensional stability, low
moisture absorption, and good mechanical properties.[2] Build-
ing micro- and mesopores within polycyanurates[3,4] is of great
significance in extending their applications further to other ad-
vanced technology fields, for example, gas and vapor adsorp-
tion, gas separation, optical-fiber coating, and ultra-low dielec-
tric materials.
(CMPs),[9] polybenzimidazole,[10–12] polyimides,[13–18] and polyary-
lates.[19] However, their studies on adsorption have been fo-
cused mainly on gases like H2, CO2, and CH4. In fact, they can
also be utilized potentially as adsorbents to selectively remove
volatile organic vapors such as aromatic hydrocarbons from air
since, in addition to their large surface area, microporous or-
ganic polymers are usually composed of benzene rings, and
consequently, have an intrinsically strong affinity toward pollu-
tant vapors like benzene or toluene. Furthermore, through
chemical modification, polar groups can be introduced to en-
hance their capacity to adsorbe water vapor by virtue of the
hydrogen-bonding interaction, which provides the possibility
for their use as a moisture adsorbent in gas dehumidification
applications.[4,12,18]
The present study was undertaken to design and synthesize
triangular cyanate monomers 1,3,5-tricyanatobenzene and
1,3,5-tri(4-cyanatophenyl)benzene, which were thermally cyclo-
trimerized to yield hyper-cross-linked polycyanurate networks
with 1,3,5-triazine rings connected through aryl ether linkages.
The large number of electron-rich oxygen and nitrogen atoms
in the networks will be advantageous for CO2 affinity owing to
the dipole–quadrupole interactions. In addition, the presence
of rigid cross-linking nodes such as the triazine ring, benzene,
and triphenylbenzene is expected to effectively prop up the
segments to generate micropores or mesopores, and the po-
rosity parameters such as specific surface area, pore size, and
distribution may be affected by the chemical structures of the
net nodes and struts. The influence of the pore structure,
chemical composition, and hydrophilic/oleophilic property of
the inner pore surface of the porous networks on the adsorp-
Over the past several years, various microporous polymers
have been published in the literature, such as hyper-cross-
linked polymers (HCPs),[5] polymers of intrinsic microporosity
(PIMs),[6] covalent organic frameworks (COFs),[7] triazine-based
frameworks (CTFs),[8] conjugated microporous polymers
[a] H. Yu, C. Shen, Prof. Dr. Z. Wang
State Key Laboratory of Fine Chemicals
Department of Polymer Science and Materials
School of Chemical Engineering, Dalian University of Technology
Dalian 116024 (P.R. China)
Fax: (+86)411-84986096
E-mail: zgwang@dlut.edu.cn
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201300090.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 498 – 505 498

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
tions of H2 and CO2, organic, and water vapors will be investi-
gated and discussed in detail.
As illustrated in Scheme 2, thermal polycyclotrimerization re-
actions were performed in diphenyl sulfone solution to synthe-
size two polycyanurate networks CE-P1 and CE-P2 from 1,3,5-
tri(4-cyanatophenyl)benzene and 1,3,5-tricyanatobenzene, re-
spectively. The polymerization conditions for the two polymers,
such as the system concentration and temperature program,
were the same. In addition, nonyl phenol (1.0 wt%) was added
as a catalyst to facilitate the reaction.
Results and Discussion
Synthesis of cyanate monomers and cross-linked networks
The synthesis routes to the cyanate monomers are depicted in
The chemical structures of CE-P1 and CE-P2 were confirmed
by FTIR spectra, solid-state 13C CP/MAS NMR spectra, and ele-
mental analysis. As shown in Figure 1, after the polymerization,
Scheme 1. 1,3,5-Tricyanatobenzene (1) could be synthesized by
Figure 1. FTIR spectra of porous polycyanurate networks CE-P1 and CE-P2.
the absorptions of the OCN groups at 2280 and 2237 cmꢀ1
completely disappear, while the strong bands corresponding
to the triazine rings appear at 1564 and 1365 cmꢀ1, which is in-
dicative of the formation of polycyanurate networks. In addi-
tion, it is worth noting that the cyclotrimerization of cyanate
groups should undergo three stages. At first, two cyanate
groups are dimerized to yield the intermediate diazine ring,
Scheme 1. Synthesis routes to cyanate monomers 1,3,5-tri(4-cyanatophenyl)-
benzene and 1,3,5-tricyanatobenzene.
the reaction between 1,3,5-trihydroxybenzene and cyanogen
bromide, whereas 1,3,5-tri(4-cyanatophenyl)benzene (5) was
obtained mainly in three steps: the Suzuki cross-coupling be-
tween 1,3,5-tribromobenzene and 2 led to 1,3,5-tri(4-
methoxyphenyl)benzene (3), which was converted
further to 1,3,5-tri(4-hydroxyphenyl)benzene (4) by
demethylation. Then 4 was reacted with cyanogen
bromide to yield the target monomer 5.
The above intermediates and final monomer prod-
ucts were characterized by FTIR and 1H NMR spec-
troscopy. The characteristic absorptions in the FTIR
spectra (Figure 1S in the Supporting Information) at
2280 and 2237 cmꢀ1 are attributed to the NCO
groups of the two cyanate monomers. As shown in
the 1H NMR spectra (Figure 2S and 3S) for 1,3,5-tricya-
natobenzene, there is only a single signal at d=
7.35 ppm owing to the protons in the 1,3,5-substitut-
ed benzene ring. By contrast, for 1,3,5-tri(4-cyanato-
phenyl)benzene, the former single peak shifts from
d=7.35 to 7.71 ppm, and two new double peaks
appear at around d=7.76 and 7.45 ppm, which are
assigned to the protons in the 1,4-substituted ben-
zene ring.
Scheme 2. Synthesis of polycyanurate networks by means of a cyclotrimerization reac-
tion.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 498 – 505 499

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
which is then ring-opened to produce the intermediate linkage
ꢀOCONC(NH2)Oꢀ, and finally, the intermediate is further treat-
ed with another cyanate group to form the triazine ring. Al-
though the above reaction mechanism still remains a matter of
debate, the existence of a stable ꢀOCONC(NH2)Oꢀ group over
the course of the polymerization has been confirmed.[20] In our
case, indeed, the FTIR spectra show the adsorption bands at
approximately 3400 and 1720 cmꢀ1 corresponding to the
stretching vibrations of NꢀH and C=O, respectively. The inten-
sities of the NꢀH and C=O bands for CE-P1 are much lower
than those of CE-P2, which suggests that CE-P1 has a higher
conversion of cyanate groups to triazine rings. Furthermore,
the chemical structures of the networks were analyzed by
solid-state 13C CP/MAS NMR spectroscopy (Figure 2). Both of
Figure 3. TGA curves of porous polycyanurate networks CE-P1 and CE-P2.
the reduction of the thermal stability, which agrees well with
the FTIR results.
The field-emission scanning electron microscopy (FESEM)
images show that the two products have completely different
surface morphologies. As illustrated in Figure 4, CE-P1 is com-
Figure 2. Solid-state 13C CP/MAS NMR spectra of porous polycyanurate net-
works CE-P1 and CE-P2. Asterisks (*) indicate peaks arising from spinning
side bands.
Figure 4. FESEM images of porous polycyanurates for (a) CE-P1 and (b) CE-
P2.
them display carbon signals at about d=173 and 152 ppm,
which belong to the carbon atom of the triazine ring and the
O-substituted phenyl carbon, respectively. For CE-P2, the
broad peak centered at d=106 ppm corresponds to the un-
substituted phenyl carbon. As for CE-P1, the peak at d=
141 ppm is assigned to the phenyl-substituted phenyl carbon,
whereas the peak at d=127 ppm and its shoulder peak at d=
120 ppm are attributed to the unsubstituted phenyl carbon.
Nevertheless, the carbonyl signals of residual intermediate seg-
ments in both polymers are not clearly detected in the 13C CP/
MAS NMR spectra. They are probably embedded within the
broad peak of the carbon signals of phenyl groups or triazine
rings.
posed of tiny, loose particles with relatively uniform size and
rough surface. The diameters of the particles are between 30
and 50 nm, which are similar to other microporous polymers
with large surface areas.[8] In contrast, at the same magnifica-
tion, CE-P2 displays much larger particle sizes and the particle
surfaces are relatively smooth. The surface morphology of
a product is tightly related to its pore-formation mechanism
and pore property, which will be discussed in the following
sections.
Porous structures of polycyanurate networks
The cured products are insoluble in all common solvents,
which reflect their hyper-cross-linked structure. Thermogravi-
metric analysis (TGA) traces (Figure 3) show that CE-P1 exhibits
a significantly greater thermal stability than CE-P2. For exam-
ple, the 5% weight losses for CE-P1 and CE-P2 take place at
462 and 4078C, and the char yields at 8008C are 72.6 and
50.1 wt%, respectively. The reason can be attributed to two
factors: first, the higher phenyl content in CE-1 is advanta-
geous for its resistance to heat; second, the presence of more
intermediate segments ꢀOCONC(NH2)Oꢀ in CE-P2 results in
The nitrogen physisorption isotherms of the two samples mea-
sured at 77 K are presented in Figure 5. For CE-P1, the nitro-
gen adsorption displays a steep rise at low relative pressure (P/
P0 <0.01), and then the curve becomes level, thereby exhibit-
ing the typical characteristic of a microporous material. The
significant increase of N2 uptake at P/P0 values over 0.8 is
a result of the presence of interparticulate voids caused by the
loose aggregation of tiny particles, which is consistent with
the observation in the FESEM micrographs. In the case of CE-
P2, the nitrogen uptake displays a continual rise with the rela-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 498 – 505 500

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
condensed firstly to yield the intermediate linkage ꢀOCONC-
(NH2)Oꢀ, which would then randomly react with the cyanate
group from any other adjacent segment to yield the triazine
ring so that, compared to CE-P1, the topological and porous
structures in CE-P2 networks are inhomogeneous. Moreover,
the FTIR results have indicated that the triazine reaction for
CE-P2 is less complete. The residual segments may occupy par-
tial voids, which leads to the loss of the specific surface area to
some extent.
Adsorption of CO2 and H2 and its correlation with pore
structure
CO2-uptake capacities for the polycyanurate networks were
measured at 273 and 298 K in the pressure range from 0 to
1 bar (Figure 6). The data in Table 2 show that CE-P1 has a max-
Figure 5. a) Nitrogen adsorption–desorption isotherms at 77 K. b) pore-size
distributions calculated by NLDFT for CE-P1 and CE-P2.
tive pressure, and the hysteresis loop formed by the adsorp-
tion and desorption curves closes at P/P0 =0.47. According to
the classification of IUPAC, this sorption is attributed to type IV,
which is indicative of its mesoporous structure.
The BET surface areas of the two networks were calculated
from the N2-adsorption isotherms. The data in Table 1 reveal
that, although the two cyanate monomers have a similar ster-
Table 1. Pore-structure parameters of the porous polycyanurate net-
works.
[a]
[b]
Polymer SBET [m2 gꢀ1
]
Vmicro [cm3 gꢀ1
]
Vtotal [cm3 gꢀ1
]
Pore size [nm][c]
CE-P1
CE-P2
630
195
0.17
0
0.594
0.198
0.57, 1.01
3.39
Figure 6. CO2-adsorption isotherms at 273 and 298 K, and N2-adsorption iso-
therms at 298 K.
[a] Micropore volume derived from the N2-adsorption isotherm using the
t-plot method based on the Halsey thickness equation. [b] Total pore
volume determined from the N2 isotherm at P/P0 =0.97. [c] Pore size de-
rived from the N2 isotherm using the NLDFT method.
Table 2. Gas and vapor uptake of the porous polycyanurate networks.
Polymer CO2
CO2/ H2
Benzene
n-Hexane
[wt%][d]
Water
[wt%][e]
[b]
[wt%][a] N2
[wt%][c] [wt%][d]
eoconfiguration and the same functionality, the porous proper-
ties of the cured products are remarkably different from each
other. For example, CE-P1 has a large BET surface area of
630 m2 gꢀ1 and a microporous volume of 0.17 cm3 gꢀ1, whereas
the BET surface area of CE-P2 is only 195 m2 gꢀ1 and the micro-
porous volume cannot be detected.
CE-P1
CE-P2
9.21
5.79
37.8
14.2
1.05
0.45
47.8
15.1
35.6
9.0
8.5
16.3
[a] CO2 uptake determined volumetrically from the adsorption isotherm
at 273 K and 1 bar. [b] Uptake selectivity of CO2 to N2 at 1 bar and 298 K.
[c] H2 uptake determined volumetrically from the adsorption isotherm at
77 K and 1 bar. [d] Determined volumetrically from the adsorption iso-
therm at P/P0 =0.9 measured at 298 K. [e] Water-vapor uptake deter-
mined volumetrically from the adsorption isotherm at P/P0 =0.9 mea-
sured at 293 K.
The pore-size distributions for the two polycyanurate net-
works were analyzed from the N2-adsorption isotherms by
using nonlocal density functional theory (Figure 5). Interesting-
ly, CE-P1 exhibits two quite narrow peaks at 0.57 and 1.01 nm.
CE-P2 displays a very broad pore-size distribution in the range
from 1.2 to 4.1 nm, and the major pores are located at around
3.39 nm. The reason for this can be attributed to the chemical
structure of the cyanate monomer 1,3,5-tricyanatobenzene, in
which the three cyanate groups are attached on one benzene
ring. The steric hindrance and great molecular tension made it
very difficult for the cyclotrimerization to generate a very small
pore size. Instead, the mutual reaction of cyanate groups be-
comes more random. In other words, two cyanate groups are
imum uptake value of 9.21 wt%. The CO2 isosteric enthalpy of
gas adsorption (Qst) was calculated from its sorption isotherms
at different temperatures according to the Clausius–Clapeyron
equation [Eq. (1)]:[21]
Qst
RT
ð1Þ
ln P ¼
þ C
in which R is the gas constant; C is a constant; and P and T are
ChemPlusChem 2013, 78, 498 – 505 501
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
were obtained from the slope of the plot, whereas Henry’s law
constant (KH) was derived from the equation KH =exp(A0).
These data are listed in Table 3. At either 273 or 298 K, CE-P2
exhibits less negative A0 values than CE-P1, which is indicative
of the stronger interactions between the CO2 molecule and
CE-P2 network, which agrees with the comparative results of
Qst. Based on KH at different temperatures, the limiting enthal-
pies of adsorption (Q0), that is, Qst at zero surface coverage,
were derived from the slope of the plot of lnKH versus 1/T. CE-
P1 and CE-P2 exhibit the Q0 values of 36.9 and 37.1 kJmolꢀ1,
respectively, which are comparable to MOFs with organic am-
monium ions or Lewis basic amine in the pores (30–
55 kJmolꢀ1).[23]
To obtain CO2/N2-selectivity data, the N2-adsorption iso-
therms for the two samples at 298 K were also measured
(Figure 6). The ratio of uptake of CO2 to N2 at 1 bar and 298 K
was taken as a measure of the selectivity for the two gases,
and the data are presented in Table 2. CE-P1 has a high CO2/N2
selectivity of 37.8. Moreover, CE-P1 exhibits a higher CO2/N2 se-
lectivity than CE-P2 even though the latter shows the stronger
affinity for the CO2 molecule. This is because the small pore
and narrow pore-size distribution in CE-P1 are more sensitive
to the difference in molecular diameter of various gases.
H2-adsorption isotherms for both samples are presented in
Figure 8. The data in Table 2 shows that, at 77 K and 1.0 bar,
CE-P1 has the maximum H2 uptake of 117.4 cm3 gꢀ1
(1.05 wt%). The Qst values of the two polycyanurate networks
were calculated from the H2-adsorption data at 77 and 87 K ac-
Figure 7. Variation of CO2 isosteric enthalpies with the adsorbed amount for
CE-P1 and CE-P2.
the pressure and temperature at the equilibrium state, respec-
tively.
The variation of the CO2 isosteric enthalpies with its adsorp-
tion amounts are illustrated in Figure 7. As can be seen for the
two samples, the Qst values are high at low CO2 surface cover-
age, and then apparently decrease with the adsorbed amount,
which indicates that the CO2 molecule preferentially adsorbs
on the pore walls of the polymer network rather than aggre-
gates with itself. The strong affinity of CE-P1 and CE-P2 for
the CO2 molecule is probably as a result of the high content of
triazine rings and ether bonds in the networks. It has been re-
ported that the introduction of heteroatoms like oxygen and
nitrogen in a porous polymer is advantageous for isosteric
heat of CO2 adsorption by virtue of the dipole–quadrupole in-
teractions.[11] The data of the elemental composition in
Table 1S show that the O/C and N/C atomic ratios of CE-P2 are
three times those in CE-P1. As a result, CE-P2 exhibits the sig-
nificantly higher Qst than CE-P1.
In addition, the CO2-adsorption isotherms in CE-P1 and CE-
P2 were analyzed by using the following virial equation
[Eq. (2)]:[22]
lnðn=PÞ ¼ A0 þ A1n þ A2n2 þ :::
ð2Þ
in which n is the amount adsorbed at pressure P, and A0, A1,
and so forth, are virial coeffi-
Figure 8. H2-adsorption isotherms of CE-P1 and CE-P2 at 77 and 87 K.
cients. At low surface coverage,
A2 and higher terms can be ne-
Table 3. A0, A1, KH and Q0 values of gas adsorptions in the polycyanurate networks.
glected. A0 is related to the gas-
material interaction, whereas A1
reflects the gas–gas interactions.
At the lower adsorption
amount, the virial plots of CO2
adsorption for the two networks
have very good linear relation-
ships at different temperatures
(see Figure 4S). The A0 values
Polymer
Gas
T [K]
KH [molgꢀ1 Paꢀ1
]
A0 [ln(molgꢀ1 Paꢀ1)]
A1 [gmolꢀ1
]
Q0 [kJmolꢀ1
]
CE-P1
H2
H2
H2
77
87
77
1.6ꢁ10ꢀ6
3.8ꢁ10ꢀ7
2.9ꢁ10ꢀ7
1.1ꢁ10ꢀ7
1.2ꢁ10ꢀ7
3.0ꢁ10ꢀ8
1.5ꢁ10ꢀ7
3.9ꢁ10ꢀ8
ꢀ13.3
ꢀ14.8
ꢀ15.0
ꢀ16.0
ꢀ16.0
ꢀ17.3
ꢀ15.7
ꢀ17.1
ꢀ827.2
ꢀ776.7
8.1
CE-P2
CE-P1
CE-P2
ꢀ2528.1
ꢀ2259.1
ꢀ1039.4
ꢀ1012.6
ꢀ2385.4
ꢀ2358.5
5.2
36.9
37.1
H2
87
CO2
CO2
CO2
CO2
273
298
273
298
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 498 – 505 502

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Figure 9. Variation of H2 isosteric enthalpies with the adsorbed amount for
CE-P1 and CE-P2.
Figure 10. Organic and water-vapor adsorption isotherms as a function of
the relative pressure at 298 and 293 K for CE-P1 (+20%) and CE-P2.
cording to Equation (1), which were plotted against the H2-ad-
sorbed amounts. The results in Figure 9 show that the Qst
values of CE-P1 are significantly higher than those of CE-P2.
The virial plots are illustrated in Figure 5S, and the KH, A0, A1,
and Q0 values are listed in Table 3. Q0 values were obtained
from KH. Compared to other porous materials,[24] CE-P1 shows
less negative A0 values of ꢀ13.3 ln(molgꢀ1 Paꢀ1) at 77 K and
ꢀ14.8 ln(molgꢀ1 Paꢀ1) at 87 K, which indicates the stronger H2–
material interaction. Moreover, the Q0 of CE-P1 is up to
8.1 kJmolꢀ1, which is higher or comparable to many other mi-
croporous materials. In contrast, CE-P2 has a rather low Q0
(5.2 kJmolꢀ1). The reason for the great difference of Q0 values
for CE-P1 and CE-P2 can be attributed to their porous struc-
ture. The pores in CE-P1 are located mainly in the ultrami-
croporous region. Studies have revealed that the micropore,
especially ultramicropore, can enhance the affinity for hydro-
gen in the porous materials.[25] However, for CE-P2, its major
pores are mesopores, which result in its low Q0 value.
Compared to organic vapors, the adsorption of water vapor
shows type III sorption, that is, the water vapor displays no
clear uptake initially, but its uptake amount continually increas-
es with the relative pressure. The above type III sorption char-
acteristics are especially apparent for CE-P2, which suggests
that the networks are hydrophobic in nature and the water
uptake mainly depends on the mutual agglomeration of water
molecules within the pores of networks.
In addition, the above experimental results reveal that, for
CE-P1 and CE-P2, the oleophilic pore surfaces cause the
uptake of benzene and n-hexane to be closely related to the
specific surface area of the sample. For example, the larger BET
surface area of CE-P1 results in its remarkably higher uptake
amount than that of CE-P2. Nevertheless, for water vapor, the
weak water–polymer interaction leads to the pore size becom-
ing the dominant influencing factor in its adsorption ability.
Relative to the micropores or ultramicropores in CE-P1, the
mesopores of CE-P2 can provide more sufficient space for the
condensation of water molecules within the voids. As a conse-
quence, the water uptake in CE-P2 is almost twice that in CE-
P1.
Adsorptions of organic and water vapors in polycyanurate
networks
The adsorption isotherms of benzene and n-hexane measured
at 298 K and water vapor measured at 293 K are shown in
Figure 10. For clarity, the uptake curves of CE-P1 were moved
upward 20 wt%. It is observed that the adsorptions of ben-
zene and n-hexane vapors, especially for benzene, exhibit
a steep increase at low relative pressure (P/P0 <0.1), and then
the growing trend becomes flat gradually, which is assigned to
type I sorption, and indicates that the organic vapors have
a strong affinity for the pore surface of networks. The data in
Table 2 show that the benzene uptake for CE-P1 is up to
47.8 wt%, higher than those of carbon material F42C
(39.5 wt%).[26] The excellent benzene-adsorption performance
can be due to the presence of a large number of phenyl
groups, which have strong p–p interactions with benzene mol-
ecules. In addition, it is rational to explain that the aliphatic
structure of n-hexane gives rise to an apparently lower vapor
uptake than benzene in the two networks.
Conclusion
Two highly symmetrical cyanate monomers 1,3,5-tri(4-cyanato-
phenyl)benzene and 1,3,5-tricyanatobenzene were synthesized
and polymerized by means of a thermal cyclotrimerization re-
action to obtain two new polycyanurate networks CE-P1 and
CE-P2, respectively. The products are insoluble in common or-
ganic solvents and thermally stable with 5% weight-loss tem-
peratures over 4008C. It was surprising to find that a subtle dif-
ference in the chemical structure of the cyanate monomer led
to vastly distinct porous properties. For example, the nitrogen-
adsorption isotherm of CE-P1 belongs to type I, which exhibits
typical characteristics of microporous and ultramicroporous
material. Moreover, the two narrow peaks in the NLDFT curve
appear at 0.57 and 1.01 nm. Nevertheless, CE-P2 displays
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 498 – 505 503

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
105 mmol) over 1 h. The resulting mixture was stirred at ꢀ788C for
another hour, and then triisopropyl borate (24.3 mL, 105 mmol)
was added dropwise. The mixture was stirred at ꢀ788C for a further
hour and was then allowed to warm to room temperature over-
night. White precipitation occurred on the addition of water
(20 mL). The product mixture was stirred vigorously for 30 min and
then vacuum filtered, washed with water, and then with hexane.
The resulting white solid powder was dried under vacuum at ambi-
ent temperature and used without further purification. Yield: 90%;
type IV nitrogen sorption. The pores are mainly located in the
mesoporous region and have a very broad distribution. The
reason for this was interpreted in terms of the geometric con-
figuration, steric hindrance, and reactivity of the cyanate mon-
omers.
The adsorption enthalpies of CO2 on the two networks are
up to 37.1 kJmolꢀ1, which suggests that the presence of tria-
zine rings and ether bonds can significantly enhance the inter-
actions between CO2 molecules and the pore surface. It is
noteworthy that the narrow pore-size distribution results in
the high CO2/N2 selectivity of 37.8 for CE-P1. In addition, the
ultramicropore structure of CE-P1 gives rise to its limiting en-
1
m.p. 202–2048C; H NMR (400 MHz, [D2]H2O, 258C, TMS): d=7.28–
7.26 (d, 2H), 6.69–6.67 (d, 2H), 4.61 (s, 2H), 3.58 ppm (s, 3H); IR
(KBr): n˜ =3740–3055 (OH), 3022 (ArꢀH), 2956, 2839 (CH3), 1606 (Ar),
1248 cmꢀ1 (BꢀOH).
thalpy of H2 adsorption of 8.1 kJmolꢀ1
.
The vapor-adsorption results show that the pore size, sur-
face area, and hydrophilic/oleophilic surface of the polycyanu-
rates as well as the physicochemical nature of the vapor mole-
cules play a synergistic role in the vapor-uptake ability. For or-
ganic vapors such as benzene and n-hexane, the intrinsic affini-
ty for the oleophilic surface makes their uptake amounts more
related to the surface area of the networks. On the contrary,
the water molecules tend to condense in the pores so that the
larger pore size is more advantageous for the adsorption of
water vapor.
Synthesis of 1,3,5-tri(4-methoxyphenyl)benzene (3)
1,3,5-Tribromobenzene (2.00 g, 6.35 mmol), 2 (4.34 g, 28.59 mmol),
and [Pd(PPh3)4] (1.1 g, 0.95 mmol) were dissolved in deoxygenated
dioxane (40 mL) under N2 purging. A solution of deoxygenated 2m
K2CO3 (6.91 g, 50.0 mmol) was added, and then the mixture was
heated under reflux conditions for 48 h. The white needle crystals
were obtained by vacuum-filtration, washed with water, and puri-
fied through silica gel column chromatography (CH2Cl2/hexane 1:
1
3). Yield: 95%; m.p. 146–1478C; H NMR (400 MHz, [D1]HCCl3, 258C,
TMS): d=7.64 (s, 3H), 7.62–7.60 (d, 6H), 7.00–6.98 (d, 6H),
3.84 ppm (s, 9H); IR (KBr): n˜ =3013 (ArꢀH), 2957, 2836 (CH3), 1608,
1512 (Ar), 1253, 1031 cmꢀ1 (ArꢀOꢀC).
Experimental Section
Materials
Synthesis of 1,3,5-tri(4-hydroxyphenyl)benzene (4)
1,3,5-Tribromobenzene, 1,3,5-trihydroxybenzene, cyanogen bro-
mide, 1-bromo-4-methoxybenzene, boron tribromide (BBr3), triiso-
propyl borate, and n-butyllithium were purchased from J&K Chemi-
cal Co. and used without further purification. Nonyl phenol, ace-
tone, diethyl ether, THF, dichloromethane, and triethylamine were
supplied by Shanghai Chemical Reagent. Diethyl ether and THF
were purified under reflux conditions over sodium with the indica-
tor benzophenone complex. Dichloromethane and triethylamine
were purified by distillation over calcium hydride. Acetone was pu-
rified by reflux over phosphorus pentoxide and was distilled prior
to use. Other chemical reagents were of reagent grade and used
as received.
BBr3 (8 mL of 1m CH2Cl2 solution) was added dropwise at ꢀ788C
to a solution of 3 (4.00 g, 10.10 mmol) in anhydrous CH2Cl2
(64 mL). The resulting mixture was stirred overnight. The reaction
mixture was poured into water (250 mL) and stirred vigorously for
30 min. The white needle crystals were obtained by vacuum-filtra-
tion, washed with water, and purified by recrystallization from
methanol. Yield: 92%; m.p. 238–2408C; 1H NMR (400 MHz,
[D6]acetone, 258C, TMS): d=8.44 (s, 3H), 7.69 (s, 3H), 7.67–7.65 (d,
6H), 6.98–6.96 ppm (d, 6H); IR (KBr): n˜ =3318 (OꢀH), 3030 (ArꢀH),
1607, 1513 (Ar), 1220 cmꢀ1 (ArꢀO).
Synthesis of 1,3,5-tri(4-cyanatophenyl)benzene (5)
A solution of cyanogen bromide (3.36 g, 31.69 mmol) in anhydrous
Synthesis of 1,3,5-tricyanatobenzene (1)
acetone (60 mL) was stirred at ꢀ308C and added dropwise to a so-
A solution of cyanogen bromide (6.87 g, 64.86 mmol) in anhydrous
acetone (70 mL) was stirred at ꢀ308C and added dropwise to a so-
lution of 1,3,5-trihydroxybenzene (2.33 g, 18.48 mmol) and triethyl-
amine (7.7 mL, 55.44 mmol) in acetone (80 mL) over 1 h. After the
addition had been completed, the mixture was stirred rapidly at
ꢀ108C for 1 h, and at 108C for another hour. The resulting mixture
was evaporated to dryness and washed with water. The white
needle crystals were obtained by purification through silica gel
column chromatography (CH2Cl2). Yield: 75%; m.p. 107–1098C;
1H NMR (400 MHz, [D1]HCCl3, 258C, TMS): d=7.35 ppm (s, 3H); IR
(KBr): n˜ =2281, 2236 cmꢀ1 (CꢁN).
lution of 4 (3.20 g, 9.03 mmol) and triethylamine (3.76 mL,
27.05 mmol) in acetone (60 mL) over 1 h. After the addition was
completed, the mixture was stirred rapidly at ꢀ108C for 1 h, and at
108C for another hour. The resulting mixture was evaporated to
dryness and washed with water. The white needle crystals were
obtained by purification through silica gel column chromatography
1
(CH2Cl2). Yield: 72%; m.p. 281–2838C; H NMR (400 MHz, [D1]HCCl3,
258C, TMS): d=7.76–7.74 (d, 6H), 7.71 (s, 3H), 7.45–7.43 ppm (d,
6H); IR (KBr): n˜ =2272, 2237 cmꢀ1 (CꢁN).
Preparation of porous polycyanurate networks
The two cyanurate resin networks CE-P1 and CE-P2 were obtained
by thermal cyclotrimerization from 1,3,5-tricyanatobenzene and
1,3,5-tri(4-cyanatophenyl)benzene, respectively. They were pre-
pared under the same conditions, so only a typical procedure for
CE-P1 is described: nonyl phenol (0.013 g) was added to a mixture
Synthesis of 4-methoxyphenylboronic acid (2)
A solution of 4-bromoanisole (13.5 mL, 105 mmol) in anhydrous
THF (300 mL) was stirred at ꢀ788C under dry N2 and treated drop-
wise with a solution of n-butyllithium (42.0 mL, 2.5m in hexane,
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 498 – 505 504

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
[1] a) N. B. McKeown, P. M. Budd, Chem. Soc. Rev. 2006, 35, 675–683; b) A.
Thomas, Angew. Chem. 2010, 122, 8506–8523; Angew. Chem. Int. Ed.
2010, 49, 8328–8344; c) M. Mastalerz, Chem. Eur. J. 2012, 18, 10082–
10091.
[2] I. Hamerton, Chemistry and Technology of Cyanate Ester Resins, Blackie,
Glasgow, 1994.
[3] B. F. Zhang, Z. G. Wang, Chem. Commun. 2009, 5027–5029.
[4] H. Yu, C. J. Shen, M. Z. Tian, J. Qu, Z. G. Wang, Macromolecules 2012, 45,
5140–5150.
[5] a) M. P. Tsyurupa, V. A. Davankov, React. Funct. Polym. 2006, 66, 768–
779; b) C. D. Wood, B. Tan, A. Trewin, H. J. Niu, D. Bradshaw, M. J. Ros-
seinsky, Y. Z. Khimyak, N. L. Campbell, R. Kirk, E. Stçckel, A. I. Cooper,
Chem. Mater. 2007, 19, 2034–2048.
[6] N. B. McKeown, P. M. Budd, K. J. Msayib, B. S. Ghanem, H. J. Kingston,
C. E. Tattershall, S. Makhseed, K. J. Reynolds, D. Fritsch, Chem. Eur. J.
2005, 11, 2610–2620.
[7] a) A. P. Cote, A. I. Benin, N. W. Ockwig, M. O’Keeffe, A. J. Matzger, O. M.
Yaghi, Science 2005, 310, 1166–1170; b) H. M. El-Kaderi, J. R. Hunt, J. L.
Mendoza-Cortes, A. P. Cote, R. E. Taylor, M. O’Keeffe, O. M. Yaghi, Science
2007, 316, 268–272.
[8] a) P. Kuhn, M. Antonietti, A. Thomas, Angew. Chem. 2008, 120, 3499–
3502; Angew. Chem. Int. Ed. 2008, 47, 3450–3453; b) P. Kuhn, A. D.
Forget, S. Su, A. Thomas, M. Antonietti, J. Am. Chem. Soc. 2008, 130,
13333–13337.
[9] R. Dawson, A. Laybourn, R. Clowes, Y. Z. Khimyak, D. J. Adams, A. I.
Cooper, Macromolecules 2009, 42, 8809–8816.
[10] M. G. Rabbani, H. M. El-Kaderi, Chem. Mater. 2011, 23, 1650–1653.
[11] M. G. Rabbani, T. E. Reich, R. M. Kassab, K. T. Jackson, H. M. El-Kaderi,
Chem. Commun. 2012, 48, 1141–1143.
[12] H. Yu, M. Z. Tian, C. J. Shen, Z. G. Wang, Polym. Chem. 2013, 4, 961–968.
[13] J. Weber, M. Antonietti, A. Thomas, Macromolecules 2008, 41, 2880–
2885.
[14] B. S. Ghanem, N. B. McKeown, P. M. Budd, N. M. Al-Harbi, D. Fritsch, K.
Heinrich, L. Starannikova, A. Tokarev, Y. Yampolskii, Macromolecules
2009, 42, 7881–7888.
[15] Z. G. Wang, B. F. Zhang, H. Yu, L. X. Sun, C. L. Jiao, W. S. Liu, Chem.
Commun. 2010, 46, 7730–7732.
[16] Z. G. Wang, B. F. Zhang, H. Yu, G. Y. Li, Y. J. Bao, Soft Matter 2011, 7,
5723–5730.
[17] K. V. Rao, R. Haldar, C. Kulkarni, T. K. Maji, S. J. George, Chem. Mater.
2012, 24, 969–971.
[18] C. J. Shen, Y. J. Bao, Z. G. Wang, Chem. Commun. 2013, 49, 3321–3323.
[19] B. F. Zhang, Z. G. Wang, Chem. Mater. 2010, 22, 2780–2789.
[20] a) M. F. Grenier-Loustalot, C. Lartigau, F. Metras, P. Grenier, J. Polym. Sci.
Part A 1996, 34, 2955–2966; b) K. F. Lin, J. Y. Shyu, J. Polym. Sci. Part A
2001, 39, 3085–3092.
of 3 (1.3 g) and diphenylsulfone (15 g). The resulting mixture was
heated under a nitrogen atmosphere at 1708C for 4 h, 1908C for
4 h, 2308C for 8 h, 2708C for 8 h, 2908C for 4 h, and 3308C for 2 h.
The product obtained was ground into powder, which was extract-
ed with THF in a Soxhlet apparatus for 48 h, and then dried under
vacuum at 1508C for 2 days. Yield 93%.
Instrumentation
FTIR spectra were recorded using a Nicolet 20XB FTIR spectropho-
tometer in the 400–4000 cmꢀ1 region. Samples were prepared by
dispersing the complexes in KBr and compressing the mixtures to
form disks. 1H NMR spectra were measured on a 400 MHz Varian
INOVA NMR spectrometer, using tetramethylsilane as an internal
reference. Solid-state 13C CP/MAS (cross-polarization with magic-
angle spinning) spectra were recorded on a Varian Infinity-Plus
400 spectrometer at 100.61 MHz at an MAS rate of 10.0 kHz using
zirconia rotors 4 mm in diameter using a contact time of 4.0 ms
1
and a relaxation delay of 2.0 s. The H p/2 pulse was 3.75 ms, and
two-pulse phase modulation (TPPM) decoupling was used during
the acquisition. FESEM experiments were performed on a Nova
NanoSEM 450 scanning electron microscope. Before measurement,
the samples were sputter-coated with gold. TGA measurements
were performed under a nitrogen atmosphere on a NETZSCH TG
209 thermal analyzer at a heating rate of 108Cminꢀ1. The tempera-
ture range was from room temperature to 8008C. To eliminate the
effect of possible adsorption of moisture from the atmosphere, all
the samples were dried at 1508C under vacuum for 10 h just prior
to measurements. All the gas and vapor adsorption and desorption
measurements were performed on an Autosorb iQ (Quantachrome)
surface-area and pore-size analyzer. Before measurements, the
samples were degassed at 1208C under high vacuum for 24 h. Ad-
sorption and desorption isotherms of N2 were measured at 77 K.
The surface areas (SBET) were calculated using the BET model in the
relative pressure (P/P0) range from 0.007 to 0.1 for CE-P1 and from
0.125 to 0.275 for CE-P2. Pore-size distributions were calculated
from the N2 and CO2 adsorption isotherms by applying the nonlo-
cal density functional theory (NLDFT). Micropore volumes were de-
rived using the t-plot method based on the Halsey thickness equa-
tion. H2-adsorption isotherms were measured at 77 and 87 K up to
1 bar. CO2-adsorption isotherms were measured at 273 and 298 K
up to 1 bar. Water-vapor adsorption isotherms were measured at
293 K. Benzene and n-hexane vapor adsorption isotherms were
measured at 298 K. All the vapor uptake values were determined
at P/P0 =0.9. P0 is the saturation pressure (water 17.54 torr, ben-
zene 94.52 torr, n-hexane 121.41 torr).
[21] V. Krungleviciute, L. Heroux, A. D. Migone, J. Phys. Chem. B 2005, 109,
9317–9320.
[22] J. H. Cole, D. H. Everett, C. T. Marshall, A. R. Paniego, J. C. Powl, F. Rodri-
guez-Reinoso, J. Chem. Soc. Faraday Trans. 1 1974, 70, 2154–2169.
[23] J. An, N. L. Rosi, J. Am. Chem. Soc. 2010, 132, 5578–5579.
[24] X. Lin, I. Telepeni, A. J. Blake, A. Dailly, C. M. Brown, J. M. Simmons, M.
Zoppi, G. S. Walker, K. M. Thomas, T. J. Mays, P. Hubberstey, N. R. Champ-
ness, M. Schrçder, J. Am. Chem. Soc. 2009, 131, 2159–2171.
[25] M. Rzepka, P. Lamp, M. A. de La Casa-Lillo, J. Phys. Chem. B 1998, 102,
10894–10898.
Acknowledgements
[26] M. A. Lillo-Rꢂdenas, D. Cazorla-Amorꢂs, A. Linares-Solano, Adsorption
2011, 17, 473–481.
We thank the National Science Foundation of China (nos.
51073030 and 51273031) for financial support of this research.
Keywords: adsorption
·
carbon storage
·
mesoporous
Received: March 12, 2013
materials · microporous materials · polymers
Published online on April 15, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 498 – 505 505