Poly(phenylene) and m-Terphenyl as Powerful Protecting Groups for the Preparation of Stable Organic Hydroxides

Article abstract of DOI:10.1002/anie.201511184

Four benzimidazolium hydroxide compounds, in which the C2-position is attached to a phenyl group possessing hydrogen, bromine, methyl groups, or phenyl groups at the ortho positions, are prepared and investigated for stability in a quantitative alkaline stability test. The differences between the stability of the various protecting groups in caustic solutions are rationalized on the basis of their crystal structures and DFT calculations. The highest stability was observed for the m-terphenyl-protected benzimidazolium, showing a half-life in 3 m NaOD/CD₃OD/D₂O at 80 °C of 3240 h. A high-molecular-weight polymer analogue of this model compound is prepared that exhibits excellent mechanical properties, high ionic conductivity and ion-exchange capacity, as well as remarkable hydroxide stability in alkaline solutions: only 5 % degradation after 168 h in 2 m KOH at 80 °C. This is the most stable hydroxide-conducting benzimidazolium polymer to date.

Full text of DOI:10.1002/anie.201511184

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201511184
German Edition: DOI: 10.1002/ange.201511184
Alkaline-Stable Benzimidazolium Very Important Paper
Poly(phenylene) and m-Terphenyl as Powerful Protecting Groups for
the Preparation of Stable Organic Hydroxides
Andrew G. Wright, Thomas Weissbach, and Steven Holdcroft*
Abstract: Four benzimidazolium hydroxide compounds, in
which the C2-position is attached to a phenyl group possessing
hydrogen, bromine, methyl groups, or phenyl groups at the
ortho positions, are prepared and investigated for stability in
a quantitative alkaline stability test. The differences between
the stability of the various protecting groups in caustic
solutions are rationalized on the basis of their crystal structures
and DFT calculations. The highest stability was observed for
the m-terphenyl-protected benzimidazolium, showing a half-
life in 3m NaOD/CD3OD/D2O at 808C of 3240 h. A high-
molecular-weight polymer analogue of this model compound
is prepared that exhibits excellent mechanical properties, high
Figure 1. Chemical structures of four C2-substituted benzimidazoliums
prepared herein, where XÀ is the counteranion.
ionic conductivity and ion-exchange capacity, as well as
remarkable hydroxide stability in alkaline solutions: only
5% degradation after 168 h in 2m KOH at 808C. This is the
most stable hydroxide-conducting benzimidazolium polymer
to date.
over that of quaternary ammoniums.[17] In 2015, Coates et al.
demonstrated that the same o-dimethylphenyl C2 protecting
groups also protect imidazolium molecules.[18] To date, only
three polymers have been reported that utilize this type of C2-
protection strategy for (benz)imidazoliums,[3,16,19] yet it pro-
vides the most likely strategy towards alkaline-stable, hy-
droxide-conducting polymers. The discovery that o-dimethyl-
phenylenes can protect the C2 position opens the door to
other o-substituted phenylene variants, such as halogen or
aryl protecting groups.[20,21]
Herein, we present differences in hydroxide stability of
four o-substituted phenylene C2 groups, each bearing either
ortho-positioned hydrogen atoms (HB), bromine atoms
(BrB), methyl groups (MeB), or phenyl groups (PhB). As
BrB and PhB had never been reported, a novel and versatile
synthetic route (Scheme 1) was designed to prepare function-
alized aryl-protected benzimidazoliums on a multi-gram
scale. After directed ortho-metalation and electrophilic
aromatic substitution of 1,3-dibromobenzene,[22] an acid
condensation yields compound 2 in near quantitative yield.
The controlled methylation of 2 to produce 3 allows access,
via Suzuki coupling, to various aryl-protected benzimidazoles,
such as 4 and 5. A second methylation of 3 and 4 yields BrB
and PhB, respectively. MeB and HB were prepared according
to the Supporting Information, Schemes S1 and S2.
I
mmobilized quaternary ammoniums are a class of cationic
head groups that support the conduction of anions.[1–7] They
have been used in a range of technologies, such as anion-
exchange resins,[8] hydrogen fuel cells,[1,9,10] water electro-
lyzers,[11] redox-flow batteries,[12] and reverse dialysis.[13]
However, of the numerous reported cationic groups,[14,15]
only a few show promise of long term stability under strong
alkaline conditions at elevated temperatures (for example,
808C). A sub-class of cationic head groups that are attracting
increasing attention is sterically protected imidazoliums and
benzimidazoliums.
The first example of a benzimidazolium that showed
promise of stability in strongly alkaline conditions was
reported in 2012. The compound, MeB (Figure 1), bears two
methyl groups, each attached to the ortho-position of an
adjoining C2-substituted phenyl. Both the small molecule and
the analogously structured polymer were found to be stable
for extended periods in 2m KOHaq at 608C.[16] The methyl
groups serve to increase the dihedral angle compared to that
of HB and sterically-protect the C2 position from hydroxide
attack and its subsequent ring-opening degradation. The
stability of these molecules was supported by density func-
tional theory (DFT) calculations of Long and Pivovar,
showing that methyl protecting groups greatly enhance the
stability of imidazolium and benzimidazolium hydroxides
Each of the four model compounds (Figure 1) was
subjected to the same accelerated hydroxide stability test,
which involved dissolution of the model compound (0.02m) in
3m NaOD/CD3OD/D2O (7:3 CD3OD:D2O by mass). The
solutions were heated to 808C for up to 240 h. Aliquots were
1
intermittently extracted and analyzed by H NMR spectros-
[*] A. G. Wright, T. Weissbach, Prof. S. Holdcroft
Department of Chemistry, Simon Fraser University
8888 University Dr., Burnaby BC, V5A 1S6 (Canada)
E-mail: holdcrof@sfu.ca
copy (Supporting Information, Figures S26–S29). The extent
of degradation was quantified using Equation S5 and plotted
in Figure S30.
Compound HB began degrading immediately after its
dissolution in the basic solution at room temperature and was
Supporting information for this article can be found under http://dx.
doi.org/10.1002/anie.201511184.
4818
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 4818 –4821

Angewandte
Communications
Chemie
Table 1: Properties of the model compounds based on experimental data
and DFT calculations.
À
À
Compound
Solid-state
Solution-state
C2 I
t1/2
dihedral angle[a]
dihedral angle[b]
[Š][c]
[h][d]
HB
54.40/55.02
70.45/73.26
73.08/81.92
79.21/83.77
65.03/68.58
62
88
3.704
5.497
5.587
4.743
6.218
<0.1
<10
BrB[e]
MeB
PhB
86
71
436
3240
[a] Measured between the benzimidazolium and C2 phenyl planes in the
iodide form from XRD below 908. [b] DFT calculated solution structures.
[c] The shortest C2 carbon–iodide distance(s) for the iodide-form X-ray
structures. [d] The half-life of the compound dissolved in 3m NaOD/
CD3OD/D2O at 808C. [e] BrB (XRD) possessed two unique structures
within one unit cell.
exchange capacity (IECOHÀ ) of 2.56 meqgÀ1. In its fully
hydrated state and in air, it exhibited a mixed hydroxide/
carbonate conductivity of 13.2 Æ 1.4 mScmÀ1 (228C), which is
twice the conductivity of methyl-protected poly(benzimida-
zolium) of similar IEC and similar water uptake of 81 Æ
10%.[19] Additionally, the conductivity is in the same order
of magnitude as that of pendant alkyl ammoniums and
imidazoliums.[24–28] After immersion of the membrane in 1m or
2m KOH at 808C for 168 h, only 1.7% and 5.3% degradation
was observed, respectively (Figure S35), which is unprece-
dented for a benzimidazolium-containing polymer. A plot of
the stability of PPMB in 2m KOH at 808C over time is shown
in the Supporting Information, Figure S36, which did not
appear to follow first-order kinetics. As this is unlike the small
molecule stability tests, we presume that this is due to the
distinct differences between homogeneous and heterogene-
ous degradation experiments.
To investigate the origin of the stability differences
between the C2-protected benzimidazolium small molecules,
single crystals were grown and characterized by XRD.
Furthermore, the structures were compared to those found
using DFT calculations. Each compound was crystallized in its
iodide form. The methods for their crystallization as well as
their relevant crystal data can be found in the Supporting
Information, Table S1. Refined crystal structures are shown in
Figure 2. Using the crystal structures, the dihedral angles
between the benzimidazolium plane and that of the C2-
substituted phenyl plane (Figure S38), as well as the shortest
distance between the C2 carbon and iodide, were measured
(Table 1).
The solid-state dihedral angles within each molecule were
unique for each quadrant owing to the non-planarity of the
benzimidazolium ring. BrB possessed the largest variation of
dihedral angles, and also possessed two molecular structures
in its unit cell, leading to eight different dihedral angles. The
average dihedral angles increased in the order HB < PhB <
BrB < MeB. As this trend does not follow the trend in half-life
in strong base, the dihedral angle alone cannot be used as
a measure of hydroxide stability. However, the C2 carbon–
iodide distance does match the trend in half-life, with the
longer distance translating to a longer half-life. The exception
to this trend is BrB, as its protecting bromine groups are
strongly susceptible to nucleophilic displacement.
Scheme 1. Synthetic route to the model compounds (BrB and PhB)
and polymers PPB and PPMB.
fully degraded to its amide product by time of the first
measurement (Figure S29), demonstrating extreme lability of
unprotected benzimidazoliums in strongly alkaline media.
BrB appeared to be stable at room temperature, but fully
degraded after 17 h when the temperature was raised to 808C
(Figure S28). The transition of the resonance from 4.00 ppm
to 3.90 ppm in the 1H NMR spectrum suggests that new
dimethylated benzimidazoliums are produced, which form
from the nucleophilic displacement of bromide for hydroxy
groups, as well as amide products, which appear at 3.0–
2.7 ppm.
Degradation of MeB and PhB followed exponential
decay, indicative of a pseudo-first order reaction. By fitting
the data to exponential functions, the rate constants and half-
life (t1/2) at 808C in those solutions were calculated, as shown
in Table 1. The rate of degradation of PhB (t1/2 3240 h) was ꢀ 7
times slower than that of MeB (t1/2 436 h), which represents
the highest alkaline stability for a benzimidazolium hydroxide
reported to date.
Motivated by the stability of PhB, we designed a synthetic
route to prepare the polymeric analogue, PPMB (Scheme 1).
The neutral polymer, PPB, was prepared by Yamamoto
coupling of 5 to produce a high-molecular-weight (intrinsic
viscosity of 2.10 dLgÀ1; Figure S25) poly(phenylene) back-
bone bearing 1-methylbenzimidazole pendant groups. If the
intrinsic viscosity of PPB dissolved in N-methyl-2-pyrroli-
done behaved similarly to that of poly(benzimidazole) in N,N-
dimethylformamide, which has Mark–Houwink constants a =
0.75 and Kw = 3.2 ” 10À4 dLgÀ1,[23] PPB would have a molec-
ular weight of 129000 gmolÀ1. Complete methylation of PPB
with iodomethane produced PPMB in its iodide form. As
a membrane, PPMB was strong and flexible, possessing a high
tensile strength of 72 MPa, elongation at break of 49%, and
Youngꢀs modulus of 1.29 GPa (Figure S37). In its hydroxide
form, the colorless and transparent film possessed an ion-
Angew. Chem. Int. Ed. 2016, 55, 4818 –4821
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4819

Angewandte
Communications
Chemie
Scheme 2. The two degradation pathways for benzimidazolium hydrox-
ides (ring-opening and de-methylation).
As observed in Figure 3, the nucleophilic addition–
elimination reaction on the C2 carbon of the benzimidazo-
lium leads to the formation of an intermediate state (IS1) after
overcoming the first transition state (TS1). HB has a reaction
free energy barrier (DG°) of 10.6 kcalmolÀ1 for TS1, which is
considerably lower in energy compared to MeB (22.9 kcal
molÀ1), and is similar to findings of Long and Pivovar. As DG°
is greatest for TS1, the higher the energy for this rate-limiting
step, the slower the ring-opening degradation. As such, MeB
should have improved stability over that of HB, which is in
good agreement with experimental observation. PhB is even
more resistant to ring-opening degradation, consistent with
the larger DG° (24.2 kcalmolÀ1).
The second transition state (TS2) may proceed by one of
two ways depending on the orientation of the two N-methyl
groups (TS2,trans or TS2,cis) and results in two different
configurational isomers of the amide product (Figures S40,
1
S41). The H NMR spectra of degraded MeB reveal numer-
ous amide products being formed, as only two alkyl peaks are
expected for a single isomer configuration in the 3.0–2.0 ppm
resonance region. Degraded products were also isolated (see
the Supporting Information for methods) and analyzed by
mass spectrometry (Figure S31). Only the amide product was
observed, with various amounts of deuterium exchange on the
methyl groups. However, when the same process was
performed on the isolated PhB degradation products, two
products were observed (Figure S32) that were not present in
the mass spectrum prior to the degradation test (Figure S34).
The ring-opened amide was present alongside the de-methy-
lated product, which is the first observation of its kind for an
alkali-degraded benzimidazolium hydroxide.
Figure 2. X-ray crystal structures[29] of model compounds in their
iodide form (ellipsoids set at 50% probability) alongside the dihedral
angles measured (A represents the 2-phenyl plane and B represents
the benzimidazolium plane). Only one of the two unique BrB struc-
tures is shown for clarity and PhB co-crystallized with H2O (where the
hydrogen atoms of H2O are not shown).
DFT calculations indicated that the activation energies of
de-methylation differ only slightly between HB, MeB, and
PhB (DG° of TSSN2 of 27.4, 26.9, and 27.3 kcalmolÀ1,
respectively). As TS1 is generally significantly lower than
TSSN2, the de-methylation product is usually not observed.
However, the substantial increase in the DG° of TS1 for PhB
has decreased the energetic advantage of ring-opening
degradation over that of de-methylation, with a difference
of only 3.1 kcalmolÀ1. While the effects of methanol were not
considered in the DFT calculations, the estimated differences
between the degradation rate and mechanism of individual
model compounds are in good agreement between DFT and
experiment.
To compare the hydroxide stability differences between
HB, MeB, and PhB, DFT was used to calculate the energy
barriers and states along two possible degradation pathways,
which are graphically displayed in Figure 3 (see Table S2 for
values). The overall reaction for each pathway is shown in
Scheme 2. The first pathway represents the nucleophilic
addition–elimination reaction of hydroxide on the C2
carbon of the benzimidazolium, resulting in the amide
“ring-opened” product. The second pathway represents the
nucleophilic substitution of hydroxide with the N-methyl
carbon, resulting in a 2-substituted-1-methylbenzimidazole,
which we term “de-methylation” degradation.
In summary, through examination of benzimidazolium
hydroxide model compounds, XRD, and DFT calculations,
4820 www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 4818 –4821

Angewandte
Communications
Chemie
Figure 3. Reaction profiles for the two hydroxide-mediated degradation pathways (de-methylation and ring-opening) for HB, MeB, and PhB. The
dotted lines represent the higher energy, TS2,cis, ring-opening degradation pathway. No barrier was found between IS1 and IS2.
[10] X. Lin, X. Liang, S. D. Poynton, J. R. Varcoe, A. L. Ong, J. Ran,
Y. Li, Q. Li, T. Xu, J. Membr. Sci. 2013, 443, 193 – 200.
[11] D. Pletcher, X. Li, Int. J. Hydrogen Energy 2011, 36, 15089 –
15104.
[12] M. Skyllas-Kazacos, M. H. Chakrabarti, S. A. Hajimolana, F. S.
Mjalli, M. Saleem, J. Electrochem. Soc. 2011, 158, R55 – R79.
[13] G. Z. Ramon, B. J. Feinberg, E. M. V. Hoek, Energy Environ.
Sci. 2011, 4, 4423 – 4434.
[14] J. R. Varcoe, P. Atanassov, D. R. Dekel, A. M. Herring, M. A.
Hickner, P. A. Kohl, A. R. Kucernak, W. E. Mustain, K.
Nijmeijer, K. Scott, et al., Energy Environ. Sci. 2014, 7, 3135 –
3191.
[15] M. G. Marino, K. D. Kreuer, ChemSusChem 2015, 8, 513 – 523.
[16] O. D. Thomas, K. J. W. Y. Soo, T. J. Peckham, M. P. Kulkarni, S.
Holdcroft, J. Am. Chem. Soc. 2012, 134, 10753 – 10756.
[17] H. Long, B. Pivovar, J. Phys. Chem. C 2014, 118, 9880 – 9888.
[18] K. M. Hugar, H. A. Kostalik, G. W. Coates, J. Am. Chem. Soc.
2015, 137, 8730 – 8737.
the effect on stability of four C2-protecting groups have been
rationalized. A benzimidazolium is designed to be pendant on
a poly(phenylene)-backbone in a manner that makes use of
the sterically-protecting function of ortho-disubstituted phe-
nylenes. Such polymers provide exceptional stability in
alkaline solutions at 808C. Moreover, a versatile synthetic
route is presented that facilitates further investigations of
numerous other C2-protecting groups.
Acknowledgements
This work was financially supported by the Natural Sciences
and Engineering Research Council of Canada (NSERC). We
thank Dr. Jeffrey S. Ovens and Prof. Daniel B. Leznoff for
XRD training. Computing resources were provided by West-
Grid (www.westgrid.ca) and Compute Canada Calcul Canada
(www.computecanada.ca).
[19] A. G. Wright, S. Holdcroft, ACS Macro Lett. 2014, 3, 444 – 447.
[20] T. Zimmermann, K. Schmidt, J. Heterocycl. Chem. 1996, 33,
1717 – 1721.
[21] Zimmermann and Schmidt synthesized 2,4,6-substituted-aryl-
protected benzimidazoliums but these were never investigated
for ion-exchange properties nor were hydroxide versions
formed.
Keywords: alkaline stability · anion exchange membranes ·
benzimidazolium · fuel cells · nitrogen heterocycles
[22] C. Heiss, E. Marzi, M. Schlosser, Eur. J. Org. Chem. 2003, 2003,
4625 – 4629.
[23] Y. Yuan, F. Johnson, I. Cabasso, J. Appl. Polym. Sci. 2009, 112,
3436 – 3441.
How to cite: Angew. Chem. Int. Ed. 2016, 55, 4818–4821
Angew. Chem. 2016, 128, 4898–4902
[24] C. G. Arges, L. Wang, M. Jung, V. Ramani, J. Electrochem. Soc.
2015, 162, F686 – F693.
[25] Y. Yang, J. Wang, J. Zheng, S. Li, S. Zhang, J. Membr. Sci. 2014,
467, 48 – 55.
[26] B. Lin, H. Dong, Y. Li, Z. Si, F. Gu, F. Yan, Chem. Mater. 2013,
25, 1858 – 1867.
[27] A. L. Ong, S. Saad, R. Lan, R. J. Goodfellow, S. Tao, J. Power
Sources 2011, 196, 8272 – 8279.
[28] J. Zhou, M. Ünlü, I. Anestis-Richard, P. A. Kohl, J. Membr. Sci.
2010, 350, 286 – 292.
[29] CCDC 1439721 (HB), 1439723 (BrB), 1439722 (MeB), and
1439724 (PhB·H2O) contain the supplementary crystallographic
data for this paper. These data are provided free of charge by
The Cambridge Crystallographic Data Centre.
[1] J. R. Varcoe, R. C. T. Slade, E. Lam How Yee, S. D. Poynton,
D. J. Driscoll, D. C. Apperley, Chem. Mater. 2007, 19, 2686 –
2693.
[2] N. Li, Y. Leng, M. A. Hickner, C.-Y. Wang, J. Am. Chem. Soc.
2013, 135, 10124 – 10133.
[3] J. Wang, S. Gu, R. B. Kaspar, B. Zhang, Y. Yan, ChemSusChem
2013, 6, 2079 – 2082.
[4] F. Gu, H. Dong, Y. Li, Z. Si, F. Yan, Macromolecules 2014, 47,
208 – 216.
[5] M. Zhang, J. Liu, Y. Wang, L. An, M. D. Guiver, N. Li, J. Mater.
Chem. A 2015, 3, 12284 – 12296.
[6] A. D. Mohanty, C. Y. Ryu, Y. S. Kim, C. Bae, Macromolecules
2015, 48, 7085 – 7095.
[7] P. Jannasch, ECS Trans. 2015, 69, 369 – 375.
[8] H. P. Gregor, J. Belle, R. A. Marcus, J. Am. Chem. Soc. 1955, 77,
2713 – 2719.
[9] O. I. Deavin, S. Murphy, A. L. Ong, S. D. Poynton, R. Zeng, H.
Herman, J. R. Varcoe, Energy Environ. Sci. 2012, 5, 8584 – 8597.
Received: December 2, 2015
Revised: January 18, 2016
Published online: March 7, 2016
Angew. Chem. Int. Ed. 2016, 55, 4818 –4821
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4821