Polymers of Intrinsic Microporosity derived from a carbocyclic analogue of Tr?ger's base

Article abstract of DOI:10.1016/j.polymer.2017.03.037

Tr?ger's base (TB) is often used as a building block for the synthesis of Polymers of Intrinsic Microporosity (PIMs) due to its rigid bicyclic V-shaped structure. In this study the TB component in the structure of a PIM is replaced by 2,3:6,7-dibenzobicyclo[3.3.1]nonane, a purely carbocyclic analogue of TB. This modification results in only a slightly reduced amount of microporosity as determined using nitrogen adsorption. Further comparisons with previously reported PIMs indicate that this building unit (and therefore TB) is significantly less effective for the generation of intrinsic microporosity than spirobisindane, a commonly used structural unit for PIM synthesis. It appears that the V-shape of the 2,3:6,7-dibenzobicyclo[3.3.1]nonane and TB units allows closer contact between polymer chains thereby enhancing packing efficiency.

Full text of DOI:10.1016/j.polymer.2017.03.037

Polymer xxx (2017)
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Polymers of Intrinsic Microporosity derived from a carbocyclic
€
analogue of Troger's base
,
Mariolino Carta a **, C. Grazia Bezzu a, James Vile b, Benson M. Kariuki b,
,
*
Neil B. McKeown a
a School of Chemistry, Cardiff University, Main Building, Park Place, Cardiff, CF10 3AT, UK
b School of Chemistry, University of Edinburgh, David Brewster Road, Edinburgh, EH9 3FJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
€
Troger's base (TB) is often used as a building block for the synthesis of Polymers of Intrinsic Micropo-
Received 30 January 2017
Received in revised form
7 March 2017
Accepted 17 March 2017
Available online xxx
rosity (PIMs) due to its rigid bicyclic V-shaped structure. In this study the TB component in the structure
of a PIM is replaced by 2,3:6,7-dibenzobicyclo[3.3.1]nonane, a purely carbocyclic analogue of TB. This
modification results in only a slightly reduced amount of microporosity as determined using nitrogen
adsorption. Further comparisons with previously reported PIMs indicate that this building unit (and
therefore TB) is significantly less effective for the generation of intrinsic microporosity than spi-
robisindane, a commonly used structural unit for PIM synthesis. It appears that the V-shape of the
2,3:6,7-dibenzobicyclo[3.3.1]nonane and TB units allows closer contact between polymer chains thereby
enhancing packing efficiency.
Keywords:
Polymers of intrinsic microporosity
€
Troger’s base
© 2017 Published by Elsevier Ltd.
1. Introduction
one gas over another [12]. This prompted the recent development
of PIMs derived from highly rigid bridged bicyclic structural units
Over recent years there has been increasing interest in the
preparation of new microporous materials using organic compo-
nents [1]. For example, there are a number of different types of
porous organic polymers including structurally ordered Covalent-
Organic-Frameworks (COFs) [2] and amorphous network poly-
mers such as Hypercrosslinked Polymers (HCPs) [3], Microporous
Conjugated Polymers (MCPs) [4] and Porous Aromatic Frameworks
(PAFs) [5]. Polymers of Intrinsic Microporosity (PIMs) differ in that
they do not possess a network structure and, hence, are often so-
lution processable materials [6]. PIMs generate porosity from their
rigid and contorted macromolecular chains that do not pack effi-
ciently in the solid state [7,8]. The solubility of PIMs allows them to
be processed into self-standing films and coatings and, therefore,
they are suitable for making devices such as sensors [9,10] or for the
fabrication of polymer membranes, particularly for gas separations
[11]. It is well established that increasing the rigidity of polymers
used for gas separation membranes enhances their selectivity for
such as triptycene [13e16], ethanoanthracene [17] and the amine-
based bicyclic system 6H, 12H-5,11-methanodibenzo[b,f ][1,5]
€
diazocine [17], which is more commonly known as Troger's base
(TB). The V-shaped TB unit is used in numerous applications
including making components for supramolecular chemistry [18].
TB has been introduced into PIMs both by using polymerisation
reactions that incorporate monomers that are TB derivatives, such
as diamines suitable for PIM-polyimide synthesis [19e26], or by
using
TB
formation
as
the
polymerisation
reaction
[12,14,16,27e29]. The resulting TB-based PIMs have demonstrated
excellent potential as gas separation membranes with high
permeability and good selectivity for one gas over another, for
which the latter can be partially attributed to the rigidity of TB.
Therefore, it is of interest to determine the structural contri-
bution to the generation of intrinsic microporosity by the TB unit.
With this objective, we report the synthesis of monomers and PIMs
based on a purely hydrocarbon analogue of the TB unit e i.e.
2,3:6,7-dibenzobicyclo[3.3.1]nonane [30e32]. Comparisons are
made between these polymers, structurally related TB PIMs [21],
and the more typical spirobisindane PIMs that have been described
previously [33].
* Corresponding author.
** Corresponding author.
E-mail addresses: mariolino.carta@ed.ac.uk (M. Carta), neil.mckeown@ed.ac.uk
(N.B. McKeown).
http://dx.doi.org/10.1016/j.polymer.2017.03.037
0032-3861/© 2017 Published by Elsevier Ltd.
€
Please cite this article in press as: M. Carta, et al., Polymers of Intrinsic Microporosity derived from a carbocyclic analogue of Troger's base,
Polymer (2017), http://dx.doi.org/10.1016/j.polymer.2017.03.037

2
M. Carta et al. / Polymer xxx (2017)
2. Experimental
HRMS Calc for C17H12O6 312.0634, found 312.0631. IR (DCM): 3200,
1653, 1596, 1517, 1348, 1299 cmꢀ1
.
2.1. Materials and methods
2.2.3. 5,6,11,12-Tetrahydro-2,3,8,9-tetrahydroxy-5,11-
methanodibenzo[a,e]cyclooctene [3]
Commercially available reagents were used without further
purification. Anhydrous dichloromethane was obtained by distil-
lation over calcium hydride under a nitrogen atmosphere. Anhy-
drous N,N-dimethylformamide was bought from Aldrich. All
reactions using air/moisture sensitive reagents were performed in
oven-dried or flame-dried apparatus, under a nitrogen atmosphere.
Using the general procedure diketone 5 gave 3 (0.4 g, 86%) as a
white solid, m.p. > 300 ꢁC. 1H NMR (400 MHz, CDCl3)
d
6.56 (s, 2H),
6.31 (s, 2H), 3.05 (m, 4H), 2.54 (m, 2H), 1.98 (m, 2H). 13C NMR
(125 MHz; MeOD) 144.1, 143.9,133.4, 126.6,116.0,115.9, 39.8, 33.3,
26.0. HRMS Calc for C17H16O4 284.1049, found 284.1047. IR (DCM):
3386, 1630, 1520, 1266 cmꢀ1
d
Flash chromatography was performed on silica gel 60A (35e70
mm)
.
chromatography grade (Fisher Scientific). Melting points were
recorded using a Gallenkamp Melting Point Apparatus and are
uncorrected. Infrared spectra were recorded in the range
4000e600 cmꢀ1 using a Perkin-Elmer 1600 series FTIR instrument
either as a thin film or as a nujol mull between sodium chloride
2.2.4. Tetracyclo[7.7.1.02,7.010,15]heptadecane-160,900-fluorene]-
20(70),30,50,100,120,14'-hexaene- 40,50,120,13'-tetrol [4]
Using the general procedure diketone 6 gave 4 (0.43 g, 81%) as a
light brown solid, M.p. > 300 ꢁC. 1H NMR (400 MHz, (CD3)2SO)
plates. The positions of absorption bands are quoted in cmꢀ1
.
1H
d
8.34 (s, 2H), 8.18 (s, 2H), 7.00 (m, 16H), 5.44 (s, 2H), 5.39 (s, 2H),
NMR spectra were recorded in the solvent stated using an Avance
Bruker DPX 400 instrument (400 MHz), with 13C NMR spectra
recorded at 100 MHz. Chemical shifts (dH and dC) were recorded in
parts per million (ppm) from tetramethylsilane (or chloroform) and
are corrected to 0.00 (TMS) and 7.26 (CHCl3) for 1H NMR and 77.00
(CHCl3), centre line, for 13C NMR. The abbreviations s, d, t, q, m and
br. denote singlet, doublet, triplet, quartet, multiplet and broad-
ened resonances; all coupling constants were recorded in Hertz
(Hz). Low-resolution mass spectrometric data were determined
using a Fisons VG Platform II quadrupole instrument using electron
impact ionization (EI) unless otherwise stated. High-resolution
mass spectrometric data were obtained in electron impact ioniza-
tion (EI) mode unless otherwise reported, on a Waters Q-TOF
micromass spectrometer.
2.99 (m, 2H), 2.79 (m, 2H). 13C NMR (100 MHz; (CD3)2SO)
d
156.7,
152.2, 144.2, 142.6, 140.3, 138.6, 130.8, 128.0, 127.9, 127.2, 127.1,
127.0, 125.9, 124.0, 120.0, 119.3, 119.1, 114.3, 60.0, 42.6, 25.8. HRMS
Calc for C41H28O4 584.1988, found 584.1989. IR (DCM): 3523, 3046,
2914, 1602, 1516, 1443, 1267 cmꢀ1
.
2.2.5. 2,3,8,9-Tetrahydroxy-6H,12H-5,11-methanodibenzo[1,5]-
diazocine [7]
Using the general procedure diketone 8 gave 7 (2.57 mg, 88%).
Mp > 300 ꢁC; 1H NMR (400 MHz; MeOD)
d 6.53 (s, 2H), 6.32 (s, 2H),
4.45 (d, 2H, J ¼ 16.5 Hz), 4.21 (s, 2H), 3.89 (d, 2H, 16.5 Hz); 13C NMR
(100 MHz; MeOD)
d 145.9, 145.2, 144.0, 139.9, 119.4, 113.6, 112.2,
111.4, 67.8, 58.9, 49.0; HRMS Calc. for C15H15N2O4 287.1032, found
287.1026.
Low-temperature (77 K) nitrogen adsorption/desorption mea-
surements of PIM powders were made using a Coulter SA3100.
Samples were degassed for 800 min at 120 ꢁC under high vacuum
prior to analysis. Thermo Gravimetric Analysis (TGA) was per-
formed using the device Thermal Analysis SDT Q600 at a heating
rate of 10 ꢁC/min from room temperature to 1000 ꢁC. Single crystal
XRD data were collected at Cardiff University using a Bruker-Nonius
Kappa CCD area-detector diffractometer equipped with an Oxford
Cryostream low temperature cooling device operating at 150(2) K,
2.2.6. 5,6,11,12-Tetrahydro-2,3,8,9-tetramethoxy-5,11-
methanodibenzo[a,e]cyclooctene [5] [32]
Potassium hydroxide (3.04 g, 54.4 mmol) and ethylene glycol
(75 ml) were heated to 80 ꢁC and to this mixture hydrazine mon-
ohydrate (2.72 g, 54.4 mmol) and diketone 1 (2.50 g, 6.8 mmol)
were added and the mixture was refluxed at 200 ꢁC for 16 h. The
mixture was cooled to room temperature, quenched with water
(200 ml) and acidified slowly with aqueous hydrochloric acid until
neutral. The solid was collected by suction filtration and purified by
column chromatography (hexane/EtOAc, 7:3) and recrystallized
from MeOH to give the desired product (0.92 g, 40%) as a white
solid, m.p. 206e208 ꢁC (209e210 ꢁC) [32]. 1H NMR (400 MHz,
Mo K(
a
) radiation (
l
¼ 0.71073 Å).
2.2. Synthetic procedures
The key diketone intermediate 2,3,8,9-tetramethoxy-6,12-
methanodibenzo[a,e]cyclooctene-5,11(6H, 12H)-dione 1 was pre-
pared using a literature procedure [32] (see Supplemental Data).
Crystallographic data for 1 is deposited in the Cambridge Structural
Database (CCDC 1529466).
CDCl3)
d
6.68 (s, 2H), 6.44 (s, 2H), 3.86 (s, 6H), 3.77 (s, 6H), 3.17 (m,
147.7,
4H), 2.72 (m, 2H), 2.07 (m, 2H). 13C NMR (125 MHz; CDCl3)
d
147.4, 133.0, 126.5, 112.0, 111.9, 56.0, 55.6, 38.9, 32.7, 29.2. HRMS
Calc For C21H24O4 340.1675, found 340.1664. IR (DCM): 2907, 1609,
1514, 1464, 1355, 1258, 1216, 1125, 1033 cmꢀ1. Crystals were pre-
pared by a slow diffusion of hexane into THF solution of monomer.
Crystal size: 0.4 ꢂ 0.15 ꢂ 0.02 mm, orthorombic, space group P 21
2.2.1. General procedure for demethylation using BBr3
The tetramethoxy compound (i.e. one of 1, 5, 6, 8) was dissolved
in DCM (30 ml) under a nitrogen atmosphere at 0 ꢁC. Boron tri-
bromide (5 molar equivalents) was added slowly and the reaction
was stirred at room temperature for 12 h. The reaction was
quenched with water (5 ml) and the solid was collected by suction
filtration and dried under vacuum for 12 h at 30 ꢁC to give the
desired biscatechol product.
21 21,
a
g
¼
7.6363(4),
b
¼
8.2848(6),
c
¼
27.736(2) Å,
a
¼
b
¼
¼ 90 V ¼ 1754.7(2) Å3, Z ¼ 4:
m
¼ 0.088 mmꢀ1, 1828
reflections measured, 1828 unique reflections (Rint ¼ 0.0000), 1596
reflections with I > 2
data), R ¼ 0.0706 and
s
(I), R ¼ 0.057 and
u
R2 ¼ 0.1179 (observed
u
R2 ¼ 0.1127 (all data). Crystallographic data
was deposited in the Cambridge Structural Database (CCDC
1529467).
2.2.2. 2,3,8,9-Tetrahydroxy-6,12-methanodibenzo[a,e]cyclooctene-
5,11(6H,12H)-dione [2]
2.2.7. 40,50,120,13'-tetramethoxydispiro[fluorene-9,8'-tetracyclo
[7.7.1.02,7.010,15]heptadecane- 160,900-fluorene]-20(70),30,50,100,120,14'-
hexaene [6]
In a two-necked round bottom flask was added diketone 1
(2.00 g, 5.4 mmol) and dry THF (60 ml). 2-Biphenyl magnesium
bromide (5.54 g, 21.6 mmol) in dry THF (60 ml) was added slowly at
Using the general procedure diketone 1 [32] gave 2 (0.44 g, 80%)
as a grey solid, m.p. > 300 ꢁC. 1H NMR (400 MHz, MeOD)
d
7.27 (s,
2Hc), 6.76 (s, 2H), 3.65 (m, 2H), 2.87 (m, 2H). 13C NMR (100 MHz;
MeOD) 197.1, 153.8, 147.6, 136.8, 122.5, 115.8, 115.2, 57.1, 35.3.
d
€
Please cite this article in press as: M. Carta, et al., Polymers of Intrinsic Microporosity derived from a carbocyclic analogue of Troger's base,
Polymer (2017), http://dx.doi.org/10.1016/j.polymer.2017.03.037

M. Carta et al. / Polymer xxx (2017)
3
0 ꢁC under vigorous stirring and the mixture was refluxed for 48 h.
The reaction was quenched with water (150 ml) and the THF was
evaporated under reduced pressure. The precipitate was filtered
under suction and dried. The crude product was stirred in Eaton's
reagent (50 ml) for 16 h, quenched carefully with water (150 ml)
and extracted with DCM (2 ꢂ 150 ml). The organic layers were
evaporated under reduced pressure to give the crude product. Pu-
rification by column chromatography (hexane/EtOAc) yielded 6
(1.04 g, 30%) as a white solid. M.p. 175e180 ꢁC. 1H NMR (400 MHz,
potassium carbonate (1.78 g, 12.90 mmol) were reacted according
to the general procedure to give a yellow solid (521 mg, 75% based
on the molecular weight of the repeated unit). Apparent BET sur-
face area ¼ 46 m2 gꢀ1; total pore volume ¼ 0.15 cm3 gꢀ1 estimated
from nitrogen adsorption at P/P0 ¼ 0.98; TGA analysis (nitrogen):
4.55% loss of weight occurred at ~ 100 ꢁC. Initial weight loss due to
thermal degradation commences at ~416 ꢁC. IR (DCM): 3200, 2250
(CN), 1653, 1596, 1517, 1348, 1299.
CDCl3)
d
6.96 (m, 16H), 5.68 (s, 2H), 5.45 (s, 2H), 3.35 (s, 6H), 3.33 (s,
2.3.2. Polymer P2
6H), 3.05 (t, J ¼ 3.0 Hz, 2H), 2.93 (t, J ¼ 3.0 Hz, 2H). 13C NMR
Monomer
3
(0.304
g,
1.07
mmol),
2,3,5,6-
(100 MHz; CDCl3)
d
156.1, 152.1, 148.1, 145.8, 140.8, 139.0, 132.4,
tetrafluoroterephthalonitrile (0.214 g, 1.07 mmol) and dry potas-
sium carbonate (1.78 g, 12.90 mmol) were reacted according to the
general procedure to give a yellow solid (367 mg, 85% based on the
molecular weight of the repeated unit). Apparent BET surface
129.4,128.3, 127.2, 127.1, 127.0,126.2,124.3,120.0,119.4, 114.9,110.5,
61.1, 55.6, 55.0, 43.9, 25.5. HRMS Calc for C45H36O4 640.2614, found
640.2636. IR (DCM): 3059, 2936, 1507, 1443, 1361, 1256, 1212 cmꢀ1
.
Crystals were prepared by a slow diffusion of hexane into Toluene
solution of monomer. Crystal size: 0.4 ꢂ 0.15 ꢂ 0.02 mm, mono-
clinic, space group P 21, a ¼ 14.221 [5], b ¼ 10.214 [5], c ¼ 15.384 [5]
area ¼ 437 m2
g
ꢀ1; total pore volume ¼ 0.36 cm3 gꢀ1 estimated
from nitrogen adsorption at P/P0 ¼ 0.98; TGA analysis (nitrogen):
5.51% loss of weight occurred at ~ 100 ꢁC. Initial weight loss due to
thermal degradation commences at ~505 ꢁC. IR (Nujol): 2924, 2238
Å,
b
¼ 115.536 [5] V ¼ 2016.3 [14] Å3, Z ¼ 2:
m
¼ 0.08 mmꢀ1, 4640
reflections measured, 4640 unique reflections (Rint ¼ 0.0000),
(CN), 1600 cmꢀ1. SS 13C NMR (100.5 MHz):
67.8, 39.1, 32.3, 25.7.
d
¼ 138.3, 133.9, 94.0,
3809 reflections with I > 2
s
(I), R ¼ 0.070 and
u
R2 ¼ 0.1834
(observed data), R ¼ 0.0895 and
u
R2 ¼ 0.1666 (all data). Crystal-
lographic data was deposited in the Cambridge Structural Database
(CCDC 1529468).
2.3.3. Polymer P3
Monomer
4
(0.338
g,
0.58
mmol),
2,3,5,6-
tetrafluoroterephthalonitrile (0.116 g, 0.58 mmol) and dry potas-
sium carbonate (0.64 g, 4.62 mmol) were reacted according to the
general procedure to give a yellow solid (0.224 g, 55% based on the
molecular weight of the repeat unit). Apparent BET surface
2.2.8. 2,3,8,9-Tetramethoxy-6H,12H-5,11-methanodibenzo[1,5]-
diazocine [8] [34]
3,4-Dimethoxyaniline (3.00 g, 19.58 mmol) and para-
formaldehyde (1.17 g, 39.16 mmol) were added in portions and
under vigorous stirring to trifluoroacetic acid (20 ml) at 0 ꢁC. The
mixture was allowed to cool down to room temperature and left
under stirring for 16 h. The mixture was quenched with water and a
30% aqueous NH3 solution was added until pH~9. The resulting
precipitate was collected under suction and repeatedly washed
with water and MeOH. The solid was dried under vacuum to obtain
8 as light purple powder (2.50 g, 75%). Mp 198e200 ꢁC; 1H NMR
area ¼ 684 m2
g
ꢀ1; total pore volume ¼ 0.63 cm3 gꢀ1 estimated
from nitrogen adsorption at P/P0 ¼ 0.98; TGA analysis (nitrogen):
0.80% loss of weight occurred at ~ 100 ꢁC. Initial weight loss due to
thermal degradation commences at ~477 ꢁC. 1H NMR (400 MHz,
CDCl3):
d
¼ 6.6e7.2 (br m, 20H), 2.5e3.3 (br m, 6H). Analysis by GPC
(THF): Mn ¼ 3700 Mw ¼ 10,300 g molꢀ1 relative to polystyrene. IR
(CHCl3): 3062, 2923, 2852, 2240 (CN), 1501, 1450, 1324, 1265 cmꢀ1
.
(400 MHz; CDCl3)
J ¼ 16.3 Hz), 4.27 (s, 2H), 4.06 (d, 2H, 16.3 Hz), 3.85 (s, 6H), 3.77 (s,
6H); 13C NMR (100 MHz; CDCl3)
148.3, 146.1, 140.4, 118.8, 108.8,
107.9, 67.1, 57.9, 55.9; HRMS Calc. for C19H23O4N2 343.1658, found
343.1669.
d
6.65 (s, 2H), 6.37 (s, 2H), 4.60 (d, 2H,
2.3.4. Polymer P4
Monomer
7
(1.000
g,
3.49
mmol),
2,3,5,6-
d
tetrafluoroterephthalonitrile (698 mg, 3.49 mmol), and dry potas-
sium carbonate (3.85 g, 27.92 mmol) were reacted according to the
general procedure to give a yellow solid (1.13 g, 80% based on the
molecular weight of the repeated unit). Apparent BET surface
2.3. General procedure for polymer synthesis
area ¼ 570 m2
g
ꢀ1; total pore volume ¼ 0.35 cm3 gꢀ1 estimated
from nitrogen adsorption at P/P0 ¼ 0.98; TGA analysis (nitrogen):
5% loss of weight occurred at ~ 380 ꢁC. Initial weight loss due to
thermal degradation commences at ~440 ꢁC.
In a two-necked round bottom flask was added, under inert
atmosphere, equimolar amount of the bis-catechol monomer (2, 3,
4 or 5) and 2,3,5,6-tetrafluoroterephthalonitrile, were added to
anhydrous dimethylformamide (25 ml per g of catechol). The
mixture was heated to 65 ꢁC, until the two starting materials were
completely dissolved, then dry potassium carbonate (8 equivalents)
was added and the mixture kept to stirring for 96 h. The solution
was quenched with water (80 ml per g of catechol), filtrated and
washed repeatedly with water and acetone. For the insoluble
polymers P1, P2 and P4, the crude polymer was washed under
reflux using solvent in the following sequence: THF, CHCl3, acetone
and methanol. The product was then refluxed overnight in meth-
anol, filtered off and dried under vacuum. For soluble polymer P3,
the solid was dissolved in CHCl3 (15 ml per g of solid), the solution
filtered through cotton wool and poured into a mixture of acetone/
methanol (2/1, 40 ml/g). The product was dried and refluxed with
methanol overnight, collected and dried under vacuum.
3. Results and discussion
3.1. Monomer synthesis
Following the work of Ogura et al. [32], the key intermediate
diketone 1 was conveniently prepared from commercially available
(3,4-dimethoxyphenyl)acetonitrile in three steps in a good overall
yield of 40% (Scheme 1; Supplementary Data).
Biscatechol monomers 2e4, which are suitable for PIM syn-
thesis via polymerisation using dibenzodioxine formation, were
prepared from diketone 1 as outlined in Scheme 2. Simple removal
of the methyl groups from 1 using boron tribromide gave the
diketone monomer 2. Wolff-Kishner reduction was successfully
performed on 1 to give 5 followed by methyl group removal (BBr3)
to give biscatechol monomer 3, which contains the simplest
carbocyclic analogue to the TB unit. Fused fluorene substituents
were introduced by the reaction of 1 with the Grignard reagent
derived from 2-bromobiphenyl to give 6, which was demethylated
2.3.1. Polymer P1
Monomer
2
(0.503
g,
1.61
mmol),
2,3,5,6-
tetrafluoroterephthalonitrile (0.322 g, 1.61 mmol) and dry
€
Please cite this article in press as: M. Carta, et al., Polymers of Intrinsic Microporosity derived from a carbocyclic analogue of Troger's base,
Polymer (2017), http://dx.doi.org/10.1016/j.polymer.2017.03.037

4
M. Carta et al. / Polymer xxx (2017)
Scheme 1. Reagents and conditions. i. CH2I2, NaOMe, THF, reflux 2 h; ii. KOH, EtOH, 88%. iii. PPA, 80 ꢁC.
Scheme 2. Reagents and conditions. i. BBr3, DCM, 4 h; ii. Hydrazine, KOH, ethylene glycol, 200 ꢁC, 24 h; iii. Grignard reagent from 2-bromobiphenyl, THF, reflux, 24 h; iv. Eaton's
reagent 12 h 30%.
to give monomer 4.
3.2. Polymer synthesis
To allow the direct comparison between a TB PIM and a PIM
derived from a carbocyclic TB analogue, the biscatechol-containing
TB monomer 7 was prepared using a previously reported procedure
(Scheme 3) via tetramethoxy-substituted TB derivative 8 [19,21,34].
The structure of the 2,3:6,7-dibenzobicyclo[3.3.1]nonane units
within the novel monomers was confirmed by growing crystals
suitable for x-ray diffraction analysis from intermediates 5 and 6
and this was compared with the structure reported by Bu et al. [34]
Monomers 2e4 and 7 where polymerised using the typical
conditions optimised previously for PIM preparation using diben-
zodioxin formation to give polymers P1eP4, respectively (Scheme
4) [6,35]. With the exception of P3, all of the polymers proved to be
insoluble in common organic solvents preventing their structural
characterisation via solution-based techniques such as Gel
Permeation Chromatography (GPC). In contrast, P3 is fully soluble
in THF and chloroform, presumably due to the bulky fused fluorene
groups enhancing solubility. Hence P3 was readily characterised by
1H NMR and GPC (THF), which revealed a value for Mw of
10,300 g molꢀ1 relative to polystyrene standards (Table 1). This
modest molecular mass was insufficient for the formation of robust
self-standing films.
€
of the related tetramethoxy derivative of Troger's base 8 (Fig. 1).
There is a clear similarity between the TB unit (Fig. 1a) and the
TB analogous structures (Fig. 1b and c). For example, for 5 and 8 the
angle between the two planes formed by the dimethoxy benzene
moieties was found to be 88ꢁ for both structures. For monomer 6,
instead, the same angle was found of 114ꢁ presumably due to the
bulky fluorene substituents.
Scheme 3. Reagents and conditions. i. Dimethoxymethane, TFA, RT; ii. BBr3, DCM.
€
Fig. 1. Solid state crystal structures of (a) tetramethoxy Troger's base 8, (b), 5 (b) and (c) 6.
€
Please cite this article in press as: M. Carta, et al., Polymers of Intrinsic Microporosity derived from a carbocyclic analogue of Troger's base,
Polymer (2017), http://dx.doi.org/10.1016/j.polymer.2017.03.037

M. Carta et al. / Polymer xxx (2017)
5
for Polymer P3 relative to P2 or P4 suggests swelling during anal-
ysis that indicates weaker cohesion between chains, which is
consistent with the solubility of Polymer P3 in organic solvents. A
direct comparison of the nitrogen adsorption isotherms of the
structurally related polymers P2 and P4, shows that 2,3:6,7-
dibenzobicyclo[3.3.1]nonane of P2 appears to be only slightly less
efficient at generating intrinsic microporosity than the TB unit of
P4. In addition, previous work demonstrated that the equivalent
spirobisindane-based PIMs to polymers P1 and P3, containing ke-
tone and fused fluorene substituents, had apparent BET surface
areas of 501 and 895 m2 gꢀ1, respectively [33]. Therefore, it can be
deduced that 2,3:6,7-dibenzobicyclo[3.3.1]nonane unit is signifi-
cantly less efficient at inducing intrinsic microporosity than the
spirobisindane unit, which is an often used component in the
synthesis of PIMs.
Fig. 2. N2 adsorption isotherms for polymers P1eP4 obtained at 77 K.
4. Conclusions
3.3. Analysis of microporosity
Replacing the TB component in the structure of a PIM with
2,3:6,7-dibenzobicyclo[3.3.1]nonane, a purely carbocyclic analogue
of TB, results in only a slightly reduced microporosity as deter-
mined using nitrogen adsorption. Indeed direct comparison with
previously reported PIMs indicates that thus building unit (and
therefore TB) is significantly less effective for the generation of
intrinsic microporosity than spirobisindane, a common structural
unit used in PIM synthesis. It appears likely that the shape of the
2,3:6,7-dibenzobicyclo[3.3.1]nonane unit allows closer contact
between polymer chains thereby enhancing packing efficiency.
Greater cohesive interaction between chains is apparent from the
lack of solubility of the resulting polymers, although placing bulky
fused fluorene substituents onto the polymer improves solubility
and microporosity, presumably by increasing the distance between
chains in the solid state.
Nitrogen adsorption isotherms obtained at 77 K from powdered
samples of the polymers (Fig. 2; Supplementary Data) allowed
apparent BET surface area and pore volumes to be estimated
(Table 1). Isotherms from polymers P2, P3 and P4 all show signif-
icant nitrogen adsorption at low relative pressure (P/Po < 0.01),
which indicates the presence of intrinsic microporosity. In contrast,
polymer P1 proved non-porous, which may be attributed to the
large cohesive interactions between polymer chains due to the
polar ketone groups. Polymer P3 displays relatively high surface
area, which suggests that the rigid, spiro-fused fluorine units in-
crease free volume by maintaining a larger distance between
polymer chains. The larger hysteresis between the nitrogen
adsorption and desorption isotherms (Fig. 2: Supplemental data)
Scheme 4. Reagents and conditions. i. 2,3,5,6-tetrafluoroterephthalonitrile, K2CO3, DMF, 65 ꢁC, 3 days.
Table 1
Physical properties of polymers P1eP4.
Polymer
Monomer
Solubility
BET surface
Pore Volumea
Mw
PDI
area (m2 gꢀ1
)
(ml gꢀ1
)
P1
P2
P3
P4
2
3
4
7
No
No
45
0.15
0.36
0.63
0.35
n/a
n/a
10,300
n/a
n/a
n/a
2.78
n/a
440
685
570b
Yes (CHCl3)
No
a
Calculated from the amount of nitrogen adsorbed at 77 K and relative pressure P/P0 ¼ 0.98.
b
This apparent BET surface area in good agreement with the previously reported value of 565 m2 gꢀ1 [21].
€
Please cite this article in press as: M. Carta, et al., Polymers of Intrinsic Microporosity derived from a carbocyclic analogue of Troger's base,
Polymer (2017), http://dx.doi.org/10.1016/j.polymer.2017.03.037

6
M. Carta et al. / Polymer xxx (2017)
[15] R. Swaidan, M. Al-Saeedi, B. Ghanem, E. Litwiller, I. Pinnau, Macromolecules
Acknowledgements
47 (2014) 5104e5114.
[16] I. Rose, M. Carta, R. Malpass-Evans, M.-C. Ferrari, P. Bernardo, G. Clarizia,
J.C. Jansen, N.B. McKeown, Acs Macro Lett. 4 (2015) 912e915.
[17] M. Carta, R. Malpass-Evans, M. Croad, Y. Rogan, J.C. Jansen, P. Bernardo,
F. Bazzarelli, N.B. McKeown, Science 339 (2013) 303e307.
[18] O.V. Runarsson, J. Artacho, K. Warnmark, Eur. J. Org. Chem (2012) 7015e7041.
[19] M. Carta and N. B. McKeown, (2010), Patent WO 2012/035328.
[20] Z. Wang, D. Wang, F. Zhang, J. Jin, ACS Macro Lett. 3 (2014) 597e601.
[21] Z.G. Wang, X. Liu, D. Wang, J. Jin, Polym. Chem. 5 (2014) 2793e2800.
[22] Y. Zhuang, J.G. Seong, Y.S. Do, H.J. Jo, Z. Cui, J. Lee, Y.M. Lee, M.D. Guiver,
Macromolecules 47 (2014) 3254e3262.
[23] B. Ghanem, N. Alaslai, X.H. Miao, I. Pinnau, Polymer 96 (2016) 13e19.
[24] M. Lee, C.G. Bezzu, M. Carta, P. Bernardo, G. Clarizia, J.C. Jansen, N.B. McKeown,
Macromolecules 49 (2016) 4147e4154.
[25] Y. Zhuang, J.G. Seong, Y.S. Do, W.H. Lee, M.J. Lee, M.D. Guiver, Y.M. Lee,
J. Membr. Sci. 504 (2016) 55e65.
[26] Y.B. Zhuang, J.G. Seong, Y.S. Do, W.H. Lee, M.J. Lee, Z. Cui, A.E. Lozano,
M.D. Guiver, Y.M. Lee, Chem. Commun. 52 (2016) 3817e3820.
[27] M. Carta, M. Croad, J.C. Jansen, P. Bernardo, G. Clarizia, N.B. McKeown, Polym.
Chem. 5 (2014) 5255e5261.
[28] M. Carta, R. Malpass-Evans, M. Croad, Y. Rogan, M. Lee, I. Rose, N.B. McKeown,
Polym. Chem. 5 (2014) 5267e5272.
[29] Y. Xiao, L. Zhang, L. Xu, T.-S. Chung, J. Membr. Sci. 521 (2017) 65e72.
[30] M.C. Kimber, A.C. Try, L. Painter, M.M. Harding, P. Turner, J. Org. Chem. 65
(2000) 3042e3046.
[31] J. Ju, L. Zhang, R.M. Hua, Tetrahedron Lett. 55 (2014) 3374e3376.
[32] F. Ogura, A. Nakao, M. Nakagawa, Bull. Chem. Soc. Jpn. 52 (1979) 1165e1168.
[33] M. Carta, K.J. Msayib, P.M. Budd, N.B. McKeown, Org. Lett. 10 (2008)
2641e2643.
We thank the Universities of Cardiff and Edinburgh for support.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.polymer.2017.03.037.
References
[1] A.G. Slater, A.I. Cooper, Science 348 (2015) 988.
[2] S.Y. Ding, W. Wang, Chem. Soc. Rev. 42 (2013) 548e568.
[3] M.P. Tsyurupa, V.A. Davankov, React. Funct. Polym. 53 (2002) 193e203.
[4] A.I. Cooper, Adv. Mater. 21 (2009) 1291e1295.
[5] C. Pei, T. Ben, S. Qiu, Mater. Horiz. 2 (2015) 11e21.
[6] P.M. Budd, B.S. Ghanem, S. Makhseed, N.B. McKeown, K.J. Msayib,
C.E. Tattershall, Chem. Commun. (2004) 230e231.
[7] N.B. McKeown, P.M. Budd, Macromolecules 43 (2010) 5163e5176.
[8] N.B. McKeown, ISRN Mater. Sci. (2012), 513986.
[9] Y. Wang, N.B. McKeown, K.J. Msayib, G.A. Turnbull, I.D.W. Samuel, Sensors 11
(2011) 2478e2487.
[10] N.A. Rakow, M.S. Wendland, J.E. Trend, R.J. Poirier, D.M. Paolucci, S.P. Maki,
C.S. Lyons, M.J. Swierczek, Langmuir 26 (2010) 3767e3770.
[11] P.M. Budd, N.B. McKeown, B.S. Ghanem, K.J. Msayib, D. Fritsch,
L. Starannikova, N. Belov, O. Sanfirova, Y. Yampolskii, V. Shantarovich,
J. Membr. Sci. 325 (2008) 851e860.
[12] M.D. Guiver, Y.M. Lee, Science 339 (2013) 284e285.
[13] B.S. Ghanem, R. Swaidan, E. Litwiller, I. Pinnau, Adv. Mater. 26 (2014)
3688e3692.
[34] X.-H. Bu, M. Du, L.-J. Zhao, K. Tanaka, M. Shionoya, M. Shiro, J. Chem. Res.
Synop. (2001) 243e245.
[35] P.M. Budd, E.S. Elabas, B.S. Ghanem, S. Makhseed, N.B. McKeown, K.J. Msayib,
C.E. Tattershall, D. Wang, Adv. Mater. 16 (2004) 456e459.
[14] M. Carta, M. Croad, R. Malpass-Evans, J.C. Jansen, P. Bernardo, G. Clarizia,
K. Friess, M. Lanc, N.B. McKeown, Adv. Mater. 26 (2014) 3526e3531.
€
Please cite this article in press as: M. Carta, et al., Polymers of Intrinsic Microporosity derived from a carbocyclic analogue of Troger's base,
Polymer (2017), http://dx.doi.org/10.1016/j.polymer.2017.03.037