Characterizing Cation Chemistry for Anion Exchange Membranes - A Product Study of Benzylimidazolium Salt Decompositions in the Base

Article abstract of DOI:10.1021/acs.joc.9b02493

Imidazolium functionality has played a prominent role in research on anion exchange membranes for use in alkaline electrochemical devices. Base stability and degradation of these materials has been much studied, but in many instances, product pathways have not been thoroughly delineated. We report an NMR study of base-induced decomposition products from three benzylimidazolium salts bearing varying extents of methyl substitution on the imidazolium ring. The major products are consistent with a hydrolytic ring fragmentation pathway as the principal mode of decomposition. We observe several new products not previously reported in the literature on imidazolium salt degradation, including benzilic acid rearrangement products formally derived from intermediate 1,2-dicarbonyl compounds or their equivalents. However, the overall reactions are complex, the yields of observed products do not account for all consumed starting materials, and mechanistic ambiguities remain.

Full text of DOI:10.1021/acs.joc.9b02493

Subscriber access provided by Stockholm University Library
Article
Characterizing Cation Chemistry for Anion Exchange Membranes --
A Product Study of Benzylimidazolium Salt Decompositions in Base
Mark James Pellerite, Marina M. Kaplun, and Robert J. Webb
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.9b02493 • Publication Date (Web): 25 Oct 2019
Downloaded from pubs.acs.org on October 29, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Characterizing Cation Chemistry for Anion Exchange
Membranes -- A Product Study of Benzylimidazolium
Salt Decompositions in Base
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Mark J. Pellerite,* Marina M. Kaplun, and Robert J. Webb
Corporate Research Laboratories
3M Company
3M Center
Saint Paul, Minnesota 55144, United States
Abstract
Imidazolium functionality has played a prominent role in research on anion exchange membranes for
use in alkaline electrochemical devices. Base stability and degradation of these materials has been
much studied, but in many instances product pathways have not been thoroughly delineated. We
report an NMR study of base-induced decomposition products from three benzylimidazolium salts
bearing varying extents of methyl substitution on the imidazolium ring. The major products are
consistent with a hydrolytic ring fragmentation pathway as the principal mode of decomposition. We
observe several new products not previously reported in the literature on imidazolium salt degradation,
including benzilic acid rearrangement products formally derived from intermediate 1,2-dicarbonyl
compounds or their equivalents. However, the overall reactions are complex, the yields of observed
products do not account for all consumed starting material, and mechanistic ambiguities remain.
*Corresponding author: mjpellerite1@mmm.com
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 30
1
2
3
4
Introduction
5
6
7
8
Anion exchange membranes based on onium functionalities figure prominently in current research on
electrolytic methods for renewable fuel and chemical raw material production.1-3 In particular, carbon
dioxide (CO2) electrolysis using renewable energy as an input provides an appealing pathway to turn
waste CO2 into valuable fuels and chemicals. It offers a means of closing the carbon cycle as well as
storing renewable energy in the form of carbon-based fuels. There have been many excellent reviews
on CO2 electrolysis,4-13 but in most previous work, selectivities for CO formation in preference to
hydrogen were modest, and currents were low at industrially relevant potentials. Recently, Masel and
co-workers14 showed that a CO2 electrolyzer utilizing an anion exchange membrane based on an
imidazolium-substituted styrenic copolymer could maintain high current densities with >95% selectivity
for CO formation and stable performance.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Consequently, benzyltrialkylammonium- and benzylalkylimidazolium-substituted polymers such as those
utilized in the above work figure prominently in the literature on anion exchange membranes.1-3 Since
these membranes will generally be exposed to highly alkaline environments for long durations during
use in an electrolytic device, base stability of the cationic functionality is of paramount importance to a
commercially viable device. A wide range of cationic group structures have been studied in monomeric
model compounds as part of an effort to develop materials chemistry appropriate for use in robust,
chemically stable anion exchange membranes.1-3 Benzyltrimethylammonium (BTMA) salts 1 (X = Cl, Br)
have been the flagship in this work, and a large literature15-21 has grown around the characterization of
this cation’s behavior in strongly basic solutions. Benzylimidazolium salts bearing various degrees of
alkyl substitution on the imidazolium ring have also been extensively studied, and despite much effort,
questions remain as to decomposition mechanisms and ultimate products.1-3,17,18,20-28
+
N
_
X
1
Coates and coworkers17 published an in-depth study of base stability in KOH/methanol-d3 of monomeric
model imidazolium compounds bearing a wide range of substituents on the imidazolium ring. They
reviewed the various mechanisms for base-induced decomposition that have been proposed in the
literature, and compared relative base stabilities of many compounds, both new as well as known, with
the much-studied compound 1. While the work suggests a path to new and highly base-stable
2
ACS Paragon Plus Environment

Page 3 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
imidazolium cations based on use of phenyl ring substituents, products from decomposition of less
stable compounds also examined during the study were not in general well-characterized. Also,
commercial-scale electrolytic devices containing anion exchange membranes are expected to utilize
aqueous electrolytes, and the potential effect of aqueous vs. methanol-based decomposition media on
base stability and decomposition products is in general unknown.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In this work we present an NMR study of products from decomposition of benzylimidazolium salts 2-4 in
aqueous KOH solutions. Compounds featuring cations of 2 and 3 have been studied by others;17,18,20-28
while to our knowledge previous data on compound 4 decomposition in basic media is found only in the
study17 by Coates and coworkers in which the decomposition products (from the Br- salt) were not
characterized. Our data show formation of products not previously identified in other studies.
However, the overall reactions are complex, the observed products do not account quantitatively for all
consumed starting materials, and mechanistic ambiguities remain.
+
+
+
N
N
N
_
N
N
_
N
_
Cl
Cl
Cl
2 -- BMImCl
3 -- BDMImCl
4 -- BTMImCl
Results
General Considerations
We chose undeuterated KOH solution as the salt degradation medium for two primary reasons.
Electrolytes for use in most electrolytic device applications are anticipated to be aqueous solutions, with
no organic solvent content,4-14 so we wanted to mimic this condition in our studies. Also, methylated
imidazolium salts have been shown to undergo extensive H/D exchange under basic conditions in
deuterated media,18,25,30 and use of undeuterated water would avoid the attendant signal loss that could
cloud or complicate interpretation of NMR spectra. This choice of degradation medium is somewhat of
a compromise, as the large water peak obscured the region of the proton NMR spectra between 4.5 and
5.0 ppm. Although we could not use peaks that would appear in this region (such as benzylic
methylenes) in spectral analysis, there were sufficient other signals for these salts and their
decomposition products well removed from the solvent peak that yielded a useful picture of these
reactions as they unfolded in the undeuterated reaction media.
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 30
1
2
3
4
5
6
7
8
9
Our choice of FEP (fluorinated ethylene-propylene) NMR tube liners for use in the salt degradation
studies was motivated by the desire to find a readily available reaction vessel that was reasonably inert
to the highly basic reaction media. Other workers have reported observation of voluminous
precipitates, presumably siliceous, arising from basic attack when running similar reactions in
glass.18,21,25 We wanted to avoid this complication if possible and its unknown effects on reaction
chemistry and product analysis. FEP is reported to be resistant to a wide variety of chemicals, including
organics as well as strong aqueous base, up to its maximum recommended use temperature of 205
⁰C.31a Control experiments involving heating 4.25 M KOH in FEP tube liners at 80 ⁰C showed only trace
levels of contaminants including fluoride ion, and none of the products derived from the salt
decompositions. However, as discussed below, although these tube liners are convenient and
inexpensive, and allow collection of NMR data on degraded samples in situ, we found that FEP comes
with its own disadvantages. Neither it nor glass are ideal reaction vessel materials of construction for
study of these reactions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reaction conditions – salt and KOH concentrations, degradation times and temperatures – were chosen
to give reliably integratable proton NMR spectra (conversions of starting material on the order of at
least 10 mole %) in a time frame of no more than a few weeks. Thus, compounds 2 and 3 were studied
in 1 M KOH at 60-80 ⁰C, while the much more stable 4 (and 1) required the higher 4.25 M KOH
concentration and 80 ⁰C. Higher temperatures, while undoubtedly speeding decomposition reactions,
would also lead to faster evaporative loss of water from the FEP sample tubes, and we wanted to
minimize the need to open tubes and replenish fluid levels. We arrived empirically at the conditions
utilized here, and found them to provide a workable balance of these factors. Reaction solutions
started off clear and remained so for the duration of the experiment, except for some long runs on
compound 4 that were allowed to proceed to high degrees (>50 %) of starting material degradation.
These tended to develop a small amount of an insoluble yellow oil phase. Attempts at analysis of this oil
phase were largely unsuccessful in that they showed extremely complex, intractable mixtures containing
the starting salt as the only identifiable component. Comparative runs in glass vials also produced solid
precipitates and oil phases, as noted in later discussion.
Pivalate (5) proved to be an ideal choice for an internal integration standard. It is stable to KOH under
the conditions employed in this study, freely soluble in the reaction media, and gives a single sharp,
4
ACS Paragon Plus Environment

Page 5 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
intense peak in a region of the proton NMR spectrum free of absorptions from starting salts and
decomposition products. We found DSS (sodium 4,4-dimethyl-4-silapentane-1-sulfonate), used in
previous work,17 difficult to dissolve in 4.25 M KOH solution. Thus, we chose to use it as an external
reference in the D2O lock solvent in the annular space between the FEP NMR tube liner and the glass
NMR tube. These conditions gave very reproducible chemical shifts, with variability in proton spectra on
the order of ±0.02 ppm and in carbon-13 spectra on the order of ±0.5 ppm. Proton chemical shifts vs.
the external reference tended to vary between 1 M and 4.25 M KOH by about 0.15-0.2 ppm.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
_
O
5
Benzylmethylimidazolium Chloride (BMImCl, 2)
In accord with previous literature results,17,22,23,26 we found compound 2 to be the least base-stable of
the series we investigated. In 1 M KOH, 5 days at 80 ⁰C or 19 days at 60 ⁰C was sufficient to decompose
≥95% of the starting salt (see Figure S5 in the Supporting Information for an illustrative example of a
reaction mixture proton NMR spectrum). As shown in Scheme 1a, the major decomposition products
-
were formate, glycolate (HOCH2CO2 ), benzylamine, and methylamine. Table 1 shows formation of
products per mole of starting material consumed, at both reaction temperatures and as a function of
heating time. While levels of the carboxylate products were stable, we observed a steady decrease in
the signal intensity for the neutral products methylamine and benzylamine as the reaction progressed.
-
Among the minor products we have assigned are benzoate and lactate (CH3CH(OH)CO2 ).
Benzyldimethylimidazolium Chloride (BDMImCl, 3)
While more stable than 2, again in accord with literature observations,17,22,23,26 compound 3 underwent
extensive degradation under conditions used in our experiments. In 1 M KOH, we observed
approximately 8 % loss over 19 days at 60 ⁰C, and 38 % loss over 20 days at 80 ⁰C (see Figure S6 in
Supporting Information for an example of a reaction mixture proton NMR spectrum). The product
distribution shown in Scheme 1b is like that for compound 2 except acetate appeared as a major
product in place of formate, the latter now a minor product. Table 2 shows formation of products per
mole of starting material consumed, at both reaction temperatures and as a function of heating time.
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 30
1
2
3
4
5
6
We again observed stable signals for the carboxylated products, but a decrease in signal intensity for the
neutral products methylamine and benzylamine as the reaction progressed.
7
8
9
+
N
KOH
_
_
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
HCO2
CH3NH2
a)
b)
HOCH2CO2
PhCH2NH2
+
+
+
Δ
glycolate
2
+
_
_
N
KOH
PhCH2NH2
CH3NH2
HOCH2CO2
CH3CO2
+
+
+
Δ
_
HCO2
+
3
O
+
N
KOH
_
_
_
c)
N
CH3CO2
CH3NH2
+
HCO2
+
+
O
Δ
_
OH
PhCO2
+
4
2-hydroxyisobutyrate
6
Scheme 1a-c – Principal products identified in decomposition of salts 2-4 in KOH solution, in FEP tubes.
Counterions omitted for clarity. Formate was a minor product in reaction 1b; formate and benzoate
were minor products in reaction 1c. Other products formed at trace levels not listed, see text.
6
ACS Paragon Plus Environment

Page 7 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
Table 1 – Product yields in FEP as function of salt degradation for 2. Reproducibility for molar product
analyses was approximately ±0.05 based on repeatability of NMR spectra integrals.
8
9
Salt 2/1 M KOH
Product Analysis
Mole Products per Mole 2 Reacted
Glycolate Methylamine Benzylamine
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
T
Rxn Time Mole Frac 2 Mole Frac 2
(days)
(°C)
Reacted
Remaining
Formate
0.00
60
60
60
60
0
4
0.00
1.00
0.00
0.71
0.77
0.77
0.00
0.47
0.36
0.28
0.00
0.63
0.55
0.47
0.59
0.41
0.89
11
19
0.84
0.16
0.98
0.95
0.05
1.04
80
80
80
80
80
0
2
0.00
0.90
1.00
1.00
1.00
1.00
0.10
0.00
0.00
0.00
0.00
0.75
0.78
0.79
0.79
0.00
0.45
0.32
0.15
0.05
0.00
0.64
0.55
0.33
0.16
0.00
0.92
0.97
0.96
0.97
5
12
20
Table 2 – Product yields in FEP as function of salt degradation for 3. Reproducibility for molar product
analyses was approximately ±0.05 based on repeatability of NMR spectra integrals.
Salt 3/1 M KOH
Product Analysis
Mole Products per Mole 3 Reacted
Acetate Glycolate Methylamine Benzylamine Formate
T
Rxn Time Mole Frac 3 Mole Frac 3
(°C)
(days)
Reacted
Remaining
60
60
60
60
0
4
0.00
1.00
0.00
0.34
0.61
0.65
0.00
0.32
0.50
0.48
0.00
0.25
0.45
0.40
0.00
ND*
ND*
0.73
0.00
0.00
0.07
0.02
0.04
0.96
11
19
0.05
0.95
0.08
0.92
80
80
80
80
80
0
2
0.00
0.08
0.14
0.27
0.38
1.00
0.92
0.86
0.73
0.62
0.00
0.66
0.77
0.81
0.87
0.00
0.24
0.73
0.68
0.79
0.00
0.47
0.51
0.38
0.29
0.00
0.00
0.65
0.41
0.37
0.00
0.02
0.12
0.12
0.15
5
12
20
*Benzylamine methylene NMR signal appeared as shoulder on N-CH3 signal at 3.67 ppm, and could not
be accurately integrated at lowest conversions.
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 30
1
2
3
4
Benzyltetramethylimidazolium Chloride (BTMImCl, 4)
5
6
7
8
Compound 4 showed substantially more base stability than 2 and 3, as was also reported by Coates and
coworkers.17 In our studies, it required 4.25 M KOH and 80 ⁰C to produce 15 % loss of starting material
over 27 days, 23 % after 44 days, and 40% loss after 64 days (see Figure S7 in Supporting Information).
The major products observed (Scheme 1c, Table 3) were acetate and 2-hydroxyisobutyrate (2-HIB, 6);
formate, methylamine, and benzoate also appeared, the latter two only as minor products. Benzoate
appeared in greater than trace amounts only in reactions carried to higher conversions at reaction times
>30 days, and the yield of formate per mole of starting salt consumed also increased as the reaction
progressed. Benzylamine and benzyl alcohol were not found above trace levels if at all. Much of the
expected aromatic product absorption was missing, having disappeared from the spectrum along with
the consumed starting material.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 3 – Product yields in FEP as function of salt degradation for 4. Reproducibility for molar product
analyses was approximately ±0.05 based on repeatability of NMR spectra integrals.
Salt 4/4.25M KOH
Product Analysis
Mole Products per Mole 4 Reacted
Acetate 2-HIB Methylamine Formate Benzoate
T
Rxn Time Mole Frac 4 Mole Frac 4
(days)
(°C)
Reacted
Remaining
80
80
80
80
80
80
0
0.00
0.07
0.10
0.15
0.23
0.40
1.00
0.93
0.90
0.85
0.77
0.60
0.00
0.95
0.93
0.81
0.78
0.72
0.00
0.57
0.54
0.68
0.64
0.63
0.00
0.07
0.07
0.05
0.04
0.03
0.00
0.23
0.22
0.44
0.53
0.53
0.00
0.00
0.00
0.04
0.11
0.20
8
14
27
44
64
Discussion
Mechanistic Aspects – Compounds 2 and 3
Although decomposition products were not thoroughly delineated, previous work17,18,26,27 with salts 2
and 3 concluded that under strongly basic conditions these cations degrade by base attack at the C2
position of the imidazolium ring followed by ring opening. Whereas BTMA 1 appears to undergo mainly
SN2 reactions to yield products such as benzyl alcohol,19,20 benzyl methyl ether,17,19 and N,N-
dimethylbenzylamine,17 imidazolium salts are believed to follow other pathways besides SN2.17,26,27 Our
results on 2 and 3 are entirely consistent with this proposal. The suite of products in Scheme 1a-b,
particularly the observations of benzylamine as a major product, suggests hydrolytic fragmentation of
8
ACS Paragon Plus Environment

Page 9 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
the starting intact imidazolium, as opposed to SN2 displacement at the benzylic methylene. The major
products observed for salts 2 and 3 can all be traced as arising from different parts of the imidazolium
ring as shown in Scheme 2. Acetate has also been proposed by Sun and coworkers18 as a product from
base-induced degradation of the cation in salt 3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In this regard, the observation of glycolate as a major product from degradation of 2 and 3 merits
further comment. To our knowledge, this is the first report of this species being observed as a product
in base-induced hydrolysis of imidazolium salts. A reasonable pathway for its formation, with ample
literature precedent,32 involves benzilic acid rearrangement of intermediate glyoxal or an equivalent
nitrogen-functionalized species18 formed in the hydrolysis of the imidazolium ring as shown in Scheme 2.
O
R
NH
N
R
H O
N
R
_
H
+
_
OH
OH
H
N
N
N
+
O
H
H
H
R
H
2 (R = H)
3 (R = CH3)
N
H
N
H
H
_
OH
A
_
NH2
RCO2
CH3NH2
+
+
+
O
_
HOCH2CO2
glycolate
Benzilic acid
rearr
H
H
_
O
OH
_
glyoxal
?
HCO2
Scheme 2 – Proposed schematic mechanism for formation of products arising from hydrolytic
fragmentation of benzylimidazoliums 2 and 3. Counterions omitted for clarity. Species indicated in
brackets are proposed intermediates and are not isolated.
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 30
1
2
3
4
5
6
7
8
9
This could be glyoxal itself or perhaps the ring-opened products analogous to those proposed by Ye and
Elabd27 and Sun and coworkers18 and indicated as intermediates A in Scheme 2. In this regard,
degradation of salt 2 showed an intermediate minor product with a proton NMR signal at 8.05 ppm (see
Figure S5 in Supporting Information) consistent with the formamide-functional ring-opened
intermediates (A in Scheme 2, R = H) of Ye and Elabd.27 These species would be expected to be unstable
in KOH solution and yield formate and possibly glyoxal-derived products upon further hydrolysis. A
base-induced secondary fragmentation pathway of glyoxal to formate (also shown in Scheme 2) may
account for the observation of this minor product in decomposition of 3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Tables 1 and 2 show that the main carboxylated ring fragmentation products – formate and glycolate
from 2, acetate and glycolate from 3 -- account for most, but not all, of the disappearance of starting
materials. For instance, heating for five days at 80 ⁰C in 1 M KOH produced complete disappearance of
2 and resulted in formation of 0.97 mole of formate and 0.79 mole of glycolate per mole of salt 2
decomposed. For 3, similar treatment after twenty days gave decomposition of 38% of the starting salt
and appearance of 0.87 mole of acetate and 0.79 mole of glycolate per mole of 3 decomposed.33
However, while these carboxylates were stable under the reaction conditions, retention and detection
of neutral products was considerably lower than for the ionic products. A glance at Tables 1 and 2
shows that levels of methylamine and benzylamine products peaked during the reactions only to drop
off as consumption of starting material continued. Loss of methylamine is perhaps not unexpected due
to its low boiling point;34 we believe volatility is the main factor behind disappearance of the
methylamine signal, as the seal afforded by the PTFE plug in the FEP tube is likely imperfect even under
modest pressure. However, this seems unlikely to explain the loss of the very high-boiling
benzylamine,34 and additional experiments were undertaken to explore this phenomenon further.
A solution of 0.4 wt % benzylamine in 4.25 M KOH containing 0.05 wt % pivalic acid as internal standard
was prepared in a screw top polypropylene vial. Aliquots were transferred into a FEP NMR tube liner
and a small glass screw top vial for aging at 80 ⁰C. Proton and carbon-13 NMR data on the fresh solution
and after 10 days’ heat aging in FEP and in glass appear in Figure 1a-b. Clearly, whereas the
benzylamine signals were stable when the solution was aged in glass, in FEP the signals decayed relative
to the internal standard and no new species appeared in either the proton or carbon-13 spectra. This
suggests that the disappearance is not due to chemical decomposition in the strongly basic medium, and
10
ACS Paragon Plus Environment

Page 11 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
indeed we have found no indications in the literature regarding base instability of benzylamine. As
mentioned above, the high boiling point of benzylamine makes loss by volatility unlikely. Thus, data on
chemical resistance of FEP notwithstanding,31a we believe these results indicate that benzylamine is
absorbed out of solution into the FEP tube wall where it is no longer detected by the NMR
experiment.31b The time scale for this process is on the order of days at 80 ⁰C, competitive with the
decomposition of salts 2 and 3; this results in what appears as a reduced yield of benzylamine relative
to the stable carboxylated products, and benzylamine yield that peaks then falls off as the
decompositions proceed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1a – Proton NMR spectra for 0.4 wt % benzylamine in 4.25 M KOH heated in FEP and in glass.
Spectra are staggered for clarity in illustrating changes in peak amplitude.
us246889.02-122-1.011.001.1r
1.0
PhCH2NH2
PhCH2NH2
Pivalate
internal std
0.5
Pivalate
internal std
=
CO3
Initial
0
10 days/80C in FEP
-0.5
180
160
140
120
100
80
60
40
Chemical Shift (ppm)
Figure 1b – Carbon-13 NMR spectra for benzylamine/KOH solution, fresh and heated in FEP.
12
ACS Paragon Plus Environment

Page 13 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
In hopes of shedding additional light on the mechanisms of BMImCl and BDMImCl decompositions, and
in view of the above results in FEP tubes, reactions were also run in glass vials. Samples of 0.83 wt %
BMImCl in 1 M KOH and 0.89 wt % BDMImCl in 4.25 M KOH with pivalate internal standard were heated
in glass screw top vials at 80 ⁰C for five days, and proton NMR spectra on fresh and aged solutions were
compared. Although benzylamine yields were slightly higher than when the reactions were run in FEP,
they were still lower than the yields of the principal carboxylated products formate and acetate,
respectively. Surprisingly, glycolate was only observed as a trace product in both cases, in stark contrast
to mixtures run in FEP where this product is prominent and easily observed. The reaction mixtures after
heating in glass were hazy, and showed development of solid precipitate and droplets of a brown oil
phase. Neither of these features were seen in FEP, where solutions remained clear, single-phase, and
colorless. This dependence of decomposition product mix on reaction vessel material of construction
indicates an additional layer of mechanistic complexity, perhaps arising from attack of KOH on the wall
of the glass vial, resulting in formation of insoluble silicates25 and a possible effect on the reaction
pathways. These aspects of the reactions were beyond the scope of our study, and were not analyzed
further. At this point our data support the view that most (>60%) of the products of base-induced
decomposition of salts 2 and 3 in FEP tubes are adequately described by the reactions shown in Scheme
1a-b, although there are other minor pathways not encompassed by this mechanism and that are not
understood in detail at present.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Mechanistic Aspects – Compound 4 and the Case of the Missing Aromatic Absorption
The major products detected in degradation of BTMImCl (4) in 4.25 M KOH – acetate and 2-
hydroxyisobutyrate (2-HIB, 6) – are consistent with a ring fragmentation mechanism as discussed above
6
for 2 and 3. Acetate is the expected product from ring opening and hydrolysis at the C-2 carbon, while
2-hydroxyisobutyrate can be formally rationalized by a benzilic acid rearrangement of a 2,3-butanedione
intermediate or an equivalent species (Scheme 3). This is the analogue of the glycolate product
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 30
1
2
3
4
5
6
O
7
8
NH
N
9
H O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
_
+
_
OH
N
OH
N
N
N
+
O
4
N
H
N
_
OH
_
NH2
CH3CO2
CH3NH2
+
+
+
O
?
O
_
Benzilic acid
rearr
_
O
O H
_
O
OH
2-hydroxyisobutyrate
Loss into FEP liner?
6
_
2,3-butanedione
?
CH3CO2
Scheme 3 – Proposed schematic mechanism for formation of acetate and 2-hydroxyisobutyrate from 4.
Counterions omitted for clarity. Benzylamine is not observed as a major product; see text. Species in
brackets are proposed intermediates and are not isolated.
observed from compounds 2 and 3. While benzilic acid rearrangements32 involving migration of primary
alkyl groups are far less common than those involving aromatic groups, there is precedent for the
process in this case. For example, 2-hydroxyisobutyrate has been observed in the product mixture
obtained from elevated-temperature treatment of cellulose in strongly basic media,35 and its formation
was attributed to benzilic acid rearrangement of intermediate 2,3-butanedione. Also, in control
experiments we observed this product (along with acetate36) when tetramethylimidazole and 2,3-
butanedione were separately treated37 in 4.25 M KOH at 80 ⁰C. Yields of these two products (along with
the minor ones assigned to date) per mole of starting material 4 consumed at various stages of the
decomposition are shown in Table 3. While the yields were not quantitative, formation of these two
carboxylated products accounted for most of the starting material loss. Methylamine was also observed
14
ACS Paragon Plus Environment

Page 15 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
in low yields as a neutral product; we believe that the observed levels are not indicative of its actual
yields due to evaporative loss, as discussed above for 2 and 3. Thus, it appears that hydrolytic ring
fragmentation is a major pathway for decomposition of 4 in aqueous KOH, as is the case for the less
substituted analogues 2 and 3. However, we also note from Tables 1-3 that decomposition of 4 appears
to be more complicated mechanistically than those of 2 and 3. Whereas the mixtures of carboxylate
products for the latter two reactions eventually leveled off to be relatively constant as a function of
starting material decomposition, this was not the case for 4 which showed increases in production of
formate and benzoate as the reaction progressed. At this point we cannot offer an interpretation of
these data, except to point out that yields of acetate and 2-hydroxyisobutyrate from decomposition of 4
in FEP tubes were in the range of 65-75 mole % based on consumption of starting material, implying that
there is a significant portion of the reaction chemistry not accounted for by the mechanistic schemes
discussed here.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As mentioned above, we saw no accompanying aromatic products (such as benzylamine) at the levels of
the major products – acetate and 2-hydroxyisobutyrate – seen in base-induced degradation of
compound 4. Data in Table 3 illustrate the issue. While benzoate was observed, it was a minor product
at best. Formate was also observed,37 as a minor product initially but becoming more significant as the
decomposition progressed. At this point we have no mechanistic explanation for the origin of formate
from 4, and we do not know whether it relates to the loss of aromatic proton absorption. The simplest
mechanistic explanation is that all three compounds 2-4 degrade primarily by the same pathway,
producing ring hydrolysis products including benzylamine. Since the degradation rate for 4 is much
slower than those for 2 and 3, the loss rate of benzylamine by absorption into the FEP tube liner
(demonstrated and discussed above) could be competitive with, or greater than, the rate of production
of benzylamine from 4. This concurrent production and consumption could produce an extremely low
steady-state concentration during the reaction such that benzylamine is not detected. In contrast, the
decomposition rates of 2 and 3, and hence the production rates of benzylamine, are much faster,
allowing it to build up to easily detectable (although transient) levels in those cases.
We hoped to test this explanation by examining the degradation reaction of 4 in glass where the above
control experiments showed stability of benzylamine NMR signals. Reaction conditions were kept
identical, changing only the reaction vessel – a glass vial in place of a FEP tube liner. Analysis of the
resulting reaction mixture was complicated by the formation of voluminous white precipitate, as
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 30
1
2
3
4
5
6
7
8
9
mentioned above and seen in other studies21,25 of base-induced imidazolium salts. Also, the rate of
disappearance of starting salt 4 was much slower than that in FEP, consistent with consumption of KOH
via reaction with the glass container and reduction of the KOH concentration. We observed only 6 mole
% consumption of 4 even out to 36 days at 80 ⁰C. While acetate and 2-hydroxyisobutyrate were again
seen as the major products, and the yield of methylamine was higher than in FEP, no benzylamine above
trace levels was found, as confirmed by spiking experiments. Unfortunately, this exercise did little to
shed light on the missing aromatic absorption. The presence or absence of benzylamine in the product
mixture from 4 is an important factor in whether base decomposition of the three salts of this study
proceeds via a common mechanism or divergent ones.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In view of these results, another scenario we considered involves base attack at the benzylic methylene
to produce benzyl alcohol and tetramethylimidazole (Scheme 4). This mechanism has the appealing
feature that it accounts for the formation of acetate and 2-hydroxyisobutyrate (by hydrolysis of a
tetramethylimidazole intermediate product) as well as the absence of benzylamine. However, one
might hope to also observe benzyl alcohol formation, and we did not detect it under any conditions for
experiments run in FEP and in glass. Spiking experiments showed that the NMR analysis used here was
easily capable of detecting benzyl alcohol in these reaction mixtures were it present. These results are
consistent with those of Mohanty and Bae20 who reported observation of benzyl alcohol formed in base
decomposition of several quaternary ammonium salts, but not BDMIm (3), leading them to conclude
that products from that cation involved other rearrangement processes that were not specified further.
In the case of 3 we believe this to be ring hydrolytic fragmentation.
+
OH-
OH
N
N
N
N
+
Δ
4
Scheme 4 – Proposed minor SN2 pathway in degradation of 4. Counterions omitted for clarity.
Our experience with instability of benzylamine NMR signals in FEP tubes led us to analogous
experiments with benzyl alcohol. Results were unexpected, and potentially relevant to the mechanistic
discussion at hand. Figure 2a-b shows spectra obtained from a solution of 0.4 wt % benzyl alcohol in
4.25 M KOH containing 0.05 wt % pivalic acid as internal standard prepared in a screw top
16
ACS Paragon Plus Environment

Page 17 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
polypropylene vial. Aliquots were transferred into FEP NMR tube liners and small glass screw top vials
for aging at 80 ⁰C. Benzyl alcohol showed behavior like that of benzylamine in that the NMR signal was
stable for the sample aged in glass but not that in FEP, where most of the signal decayed over the 13-
day duration of the experiment. However, we observed two modes of loss for benzyl alcohol in contrast
to only one for benzylamine. Whereas both species showed signal loss in FEP, about 46% of the loss of
benzyl alcohol signal was due to conversion to benzoate. The remainder was, we believe, due to
absorption into the FEP tube walls (like what we observed for benzylamine). In glass, only a trace of
benzoate formation was detected (Figure 2a). We believe the appearance of benzoate in these
experiments is due to alkaline oxidation of benzyl alcohol38 by air, since we did not use inerted
atmosphere in these studies. The high oxygen permeability of FEP31 is likely responsible for the large
disparity in the yield of benzoate in FEP and in glass.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2a – Proton NMR spectra for 0.4 wt % benzyl alcohol in 4.25 M KOH heated in FEP and in glass.
Spectra are staggered for clarity in illustrating changes in peak amplitude.
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2b – Carbon-13 NMR spectra for benzyl alcohol/KOH solution, heated in glass or FEP.
These results with benzyl alcohol suggest that observation of benzoate as a minor product from reaction
of BTMImCl 4 in KOH solution may be a proxy or indicator for formation of low levels of benzyl alcohol,
arising from an SN2 reaction occurring as at most a minor decomposition pathway. However, the failure
to observe benzyl alcohol in significant amounts when the reaction was run in glass means that this is
also not an explanation for the loss of aromatic absorption from 4. At this point we have no good
rationale for what is happening, and more data are needed to provide further mechanistic
understanding into the behavior of this compound in strong base. One possibility is that benzylamine
and/or benzyl alcohol are the aromatic products, but they were absorbed by the FEP tube; at the same
time, formation of solid precipitates when the decomposition was run in glass interfered with their
detection in that experiment. However, at present we have no direct evidence for this somewhat
contrived interpretation. Another possibility is that degradation of aromatic products all the way to
=
species such as oxocarbons like rhodizonic acid (7) salts or carbonate (CO3 ) accounts for disappearance
O
O
O
OH
OH
O
7
18
ACS Paragon Plus Environment

Page 19 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
of the aromatic proton absorption. Arguing against this, all observable peaks in the carbonyl region of
the carbon-13 spectra of our reaction mixtures are accounted for by the species listed in Table S1, with
no evidence for formation of species such as 7. In some instances, we did see increases in carbonate
levels by carbon-13 NMR as the reactions proceeded. However, carbonate can also form by scavenging
of atmospheric CO2 by KOH,39 and quantitation of this species by NMR is problematical due to extremely
long acquisition times,40 so this avenue was not pursued. Also, we have seen no indication in our control
experiments or in the literature that compounds such as benzylamine and benzyl alcohol degrade to
oxocarbons or carbonate under the conditions used here. A reaction vessel material less permeable to
air, CO2, and organics than FEP, and more stable to KOH than glass, would help immensely.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Thus motivated, we ran trial experiments on 4 in a sealed PTFE reactor19b as a preliminary investigation
as to whether this approach might offer a more robust and versatile (although much less convenient)
reaction vessel for these base-induced decompositions. Detailed results are included in the Supporting
Information. More data are certainly needed, but results were encouraging in that the resulting
reaction mixture was clear and single-phase after treatment. Rates of decomposition in FEP and PTFE
were similar under the conditions utilized, and benzylamine and benzyl alcohol were confirmed as
products although in yields lower than that of the principal products acetate and 2-hydroxyisobutyrate.
While the results do not explain the loss of aromatic absorption discussed above (unless benzylamine
and benzyl alcohol can also undergo losses into PTFE as they appear to do in FEP), they do demonstrate
the full set of hydrolytic ring fragmentation products analogous to those seen in Scheme 2 for BMIm and
BDMIm cations. (See Scheme 3, with the addition of benzylamine and methylamine (also observed at a
much higher level than in FEP) to the product mix.) This suggests the possibility of a common
degradation mechanism for all three compounds. Observation of benzyl alcohol also provides possible
evidence for the SN2 process as a minor reaction pathway as shown in Scheme 4. However,
consumption of starting salt 4 was still larger than yield of any of the observed products, so even in PTFE
there appears to be a significant portion of the reaction chemistry eluding characterization and
description by the mechanisms discussed herein. Also complicating mechanistic interpretations is the
possibility of low yields of benzyl alcohol via hydrolysis of benzyl chloride impurity in the starting salt.
Finally, we undertook some experiments to compare behavior of BTMACl 1 with BTMImCl 4 in 4.25 M
KOH at 80 ⁰C. Existing data from Coates and coworkers17 indicated that in the d3-methanol/KOH system
utilized in that work, BTMIm was the more stable cation of the two for the bromide salts. Our results in
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 30
1
2
3
4
5
6
7
8
9
aqueous 4.25 M KOH (with chloride salts) indicated the opposite: whereas we have consistently
observed about 10-15% loss of BTMImCl over about 30 days at 80 ⁰C, BTMACl under the same salt/KOH
ratio and temperature conditions produced no detectable loss (<2-5 %) of starting material relative to
the internal standard, and no observable new products. This is consistent with work of Pivovar and
coworkers19b in which the BTMA cation showed less than 5 % loss in 2 M KOH at 80 ⁰C over 1000 hr.
Other work30b has shown that addition of organic solvents such as methanol can have a dramatic effect
on the base-induced decomposition reactions of organic cations in aqueous media. Thus, while BTMIm
is already a relatively stable cation41 in 4.25 M KOH at 80 ⁰C, and much more so than the less substituted
2 and 3, it appeared to be less stable than BTMA as the chloride salts under the conditions studied here.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Summary and Conclusions
An NMR study of base-induced decomposition of three benzylimidazolium chloride salts 2-4 in aqueous
potassium hydroxide solutions revealed a suite of products consistent with imidazolium ring hydrolytic
fragmentation. Among these products -- along with formate and acetate, and the neutral products
methylamine and (in some cases) benzylamine -- there are species previously unidentified in earlier
-
literature on imidazolium salt decompositions in base. These include glycolate (HOCH2CO2 ) from 2 and
3, and 2-hydroxyisobutyrate 6 from 4. Both products can be rationalized as arising formally from
benzilic acid rearrangements of 1,2-dicarbonyl compounds (glyoxal and 2,3-butanedione or equivalents)
produced by base hydrolysis of the imidazolium ring. While the ring fragmentation mechanism shown in
Scheme 2 accounts for most of the disappearance of the least stable starting salts 2 and 3, the
decomposition mechanism for the more stable 4 is more involved, in that the fate of the aromatic
proton absorption cannot yet be fully accounted for. Also, yields of ionic products were modest to good
based on starting material disappearance. However, for all three salts the complete material balance
was complicated by loss mechanisms involving the neutral products. Some but not all of this can be
attributed to absorption of benzylamine and evaporative and/or absorptive loss of methylamine in the
fluoropolymer NMR tube liners used as reaction vessels. These liners have advantages and
disadvantages when compared with the standard glass used in much of the literature. Details of these
decomposition reactions are complex, featuring gaps in mechanistic understanding which can provide
impetus for future studies. To our knowledge, this work is the most detailed examination to date of
products from base-induced decomposition of these three model compounds. Our results provide
additional support for the mechanistic conclusions in the literature17 that imidazolium salts degrade via
20
ACS Paragon Plus Environment

Page 21 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
ring fragmentation in highly basic media. This also appears to be the case for the much less-studied 4,
which is a model compound for anion exchange membranes finding increased utilization in carbon
dioxide electrolysis,14 and for which no degradation products had previously been identified. Our work
provides additional chemical understanding into the nature of the reactions involved, as well as some
new tools for use in study of base-induced decompositions of imidazoliums and other cationic systems
to aid development of improved anion exchange membranes.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental
Materials
Benzyl chloride (Alfa Aesar), 1-methylimidazole (Sigma Aldrich), 1,2-dimethylimidazole (Alfa Aesar),
1,2,4,5-tetramethylimidazole (TCI), potassium hydroxide 1 M aqueous solution (JT Baker), potassium
hydroxide pellets (BDH VWR Analytical), pivalic acid (TCI), benzylamine (Alfa Aesar), glycolic acid (Alfa
Aesar), 2-hydroxyisobutyric acid (Alfa Aesar), 2,3-butanedione (Alfa Aesar), benzoic acid (Alfa Aesar),
benzyl alcohol (Macron), formic acid (Macron), and sodium 4,4-dimethyl-4-silapentane-1-sulfonate
(Sigma Aldrich) were all obtained from commercial sources and used without further purification. All
process and NMR solvents were standard available commercial grades and used without further
purification.
Synthesis of Compounds 2-4
Equimolar amounts of benzyl chloride and the appropriate imidazole were charged to a small glass vial.
The vial was capped, shaken to mix the starting materials, and placed in a forced-air oven at 80 ⁰C for 3
hr. Upon cooling, the crude products 2 and 4 were taken up in chloroform and precipitated by addition
of diethyl ether.17 Solvent was decanted off, then the product layer was washed with additional ether
and dried first under a nitrogen stream then in a vacuum desiccator at room temperature to complete
solvent removal. Product purity was checked by 1H NMR spectroscopy (see Supporting Information).
Compound 3 was recrystallized from acetonitrile. Compound 2 was a viscous oil at room temperature,
while 3 and 4 were white crystalline solids. Compound 4 was also found to recrystallize beautifully from
acetonitrile, but NMR data on the resulting solid showed residual solvent that was not easily removed.
It also interfered with product analysis due to hydrolysis upon base treatment, so this purification
method was not utilized. Proton NMR spectra on the salts as prepared agreed well with those reported
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 30
1
2
3
4
by Coates and coworkers.17 Spectra for 2-4 in their respective KOH solutions prior to decomposition
5
6
reactions appear in Figures S2-S4 of the Supporting Information.
7
8
9
Degradation Studies
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Degradation media used in this work were 1 M (5.3 wt %) and 4.25 M (20 wt %) potassium hydroxide
aqueous solutions29 to which had been added 0.05 wt % pivalic acid as a base-stable internal integration
standard. Compounds 2 and 3 were studied at 0.83 and 0.89 wt % respectively in 1 M KOH (molar ratio
KOH:imidazolium salt = 23.7) at 60 and 80 ⁰C, while compound 4 was studied at 1 wt % and compound 1
at 0.74 wt % in 4.25 M KOH (molar ratio KOH:imidazolium salt = 89.6) at 80 ⁰C. Solutions were mixed in
small polypropylene bottles and dispensed using plastic pipettes. No effort was made to exclude air
from the samples. Unless otherwise specified, reactions and control samples were run in FEP
(fluorinated ethylene-propylene) NMR tube liners fitted with a PTFE (polytetrafluoroethylene) plug (Bel-
Art SP Scienceware Wilmad 3 mm FEP Sample Tube Liner, obtained from Fisher Scientific). The FEP tube
liner containing the sample solution was plugged, inserted into a glass 5 mm NMR tube, and placed in a
forced-air oven for the desired time. At 80 ⁰C we observed that the liquid level in the tube tended to
drop slightly during long-duration experiments (a week or more), presumably due to evaporative loss of
water. In these cases, the samples were uncapped and topped off with deionized water to the original
fluid levels to restore the target KOH concentration before continuing the experiment. For NMR
analysis, deuterium oxide (D2O) containing sodium 4,4-dimethyl-4-silapentane-1-sulfonate (DSS) was
added to the annular space between the FEP liner and the glass as an instrument lock and external
reference (see Figure S1 in Supporting Information). Spectra were recorded immediately after sample
preparation and at selected intervals during elevated temperature treatment. Initial spectra showed no
peaks above trace levels other than water and those attributed to starting material and pivalate internal
standard.
Acquisition of NMR Spectra
NMR data were obtained at 298 ⁰K on a Bruker Avance-III NMR spectrometer operating at 500.13 MHz
for proton and equipped with a cryogenically cooled BBFO probe head. In situ NMR data were collected
for the 1 M and 4.25 M KOH samples using 3 mm Wilmad fluorinated ethylene propylene (FEP) tube
liners prepared as described above. The D2O in the annular space allowed for both locking and
shimming operations. The proton and carbon-13 NMR 90 ° pulse widths were determined for these
lossy samples using a 4.25 M hydroxide solution with a mixture of potassium and tetramethyl-
22
ACS Paragon Plus Environment

Page 23 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
ammonium cations. The 1H 90 ° pulse width was 99 µS at 12W, while the 13C 90 ° pulse width was 33 µS
at 87 W. Proton NMR data were collected with the standard Bruker zg15 pulse sequence utilizing 15 °
excitation pulses, ~8 kHz spectral width, 4 sec acquisition time, and 32 scans, with a 0.1 sec interpulse
delay. 13C data were collected with the Bruker power gated decoupled pulse sequence (zgpg30), 30 °
excitation pulses, ~33 kHz spectral width, 1 sec acquisition time, 2 sec interpulse delay, 2048 scans, and
composite pulse decoupling. The two-dimensional (2D) proton-carbon HMQC NMR data were collected
using the standard Bruker gradient-selected, quantum-filtered pulse sequence (hmqcgpqf), with a ~8
kHz (512 complex points) 1H and ~21 kHz (128 slices) 13C spectral window, adiabatic decoupling, and a
1.5 sec interpulse delay. Typical data processing included forward linear prediction and zero filling in
the f1 dimension, followed by squared sine bell weighting prior to Fourier transformation in both
dimensions. The HMBC NMR data were collected with a somewhat larger spectral window in f1 (i.e. ~28
kHz, 1024 complex points) using the standard Bruker gradient selected (hmbcgplpndqf) pulse sequence
optimized for long range couplings and utilizing a low pass J-filter to suppress the one-bond correlations.
Again, typical data processing included forward linear prediction and zero filling in f1, followed by sine
bell weighting and Fourier transformation in both dimensions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Product Assignments
Structural assignments for the products from salt decompositions were made using spiking experiments
and comparison of proton NMR spectra and two-dimensional 1H-13C correlations from HMQC and HMBC
spectra with those of authentic samples. Chemical shift data for starting materials and products in 1 M
and/or 4.25 M KOH solutions used in the assignments and analyses are shown in Table S1 in the
Supporting Information.
Base-Induced Decomposition of Benzyltetramethylimidazolium Chloride (4) in a Sealed PTFE Reactor.
A solution of 1 wt % 4 in 4.25 M KOH solution containing 0.05 wt % pivalic acid as internal integration
standard was prepared in a polypropylene screw-top vial. The solution was analyzed by proton NMR
spectroscopy using methods described in the text, producing a spectrum essentially identical to that in
Figure S4. A 16.3 g aliquot of this solution was transferred to a 125 ml Parr 4748 acid digestion vessel
equipped with a PTFE cup and cover. The headspace was purged briefly with a nitrogen stream, and the
vessel was sealed and placed in a forced-air oven at 80 ⁰C for 27 days. The vessel was removed from
the oven and allowed to cool to room temperature before opening. The product solution was clear,
single-phase, and colorless. Proton NMR analysis of the product solution gave the spectrum and peak
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 30
1
2
3
4
5
6
7
8
9
assignments shown in Figure S8 of the Supporting Information. Analysis of peak integrals gave the
product composition shown in Table S2 of the Supporting information. Consumption of starting
material compares favorably with that observed in FEP under the same conditions (data from Table 3
also included in Table S2). Yields of acetate, 2-hydroxyisobutyrate, and formate were lower, while those
of methylamine, benzoate, benzylamine, and benzyl alcohol were higher than in FEP (the latter two not
observed when the reaction was conducted in FEP, possibly due to loss into the tube wall as discussed in
the text.)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Acknowledgments
The authors are grateful to Drs. C. Hartmann-Thompson, J. C. Thomas, M. Yandrasits, and N. Jing (3M
Corporate Research Laboratory) for engaging and helpful discussions. Portions of this work were
supported in part by ARPA-E under contract DE-AR0000950. The opinions here are those of the
authors and may not reflect the opinions of ARPA-E.
Supporting Information: Schematic diagram of NMR tube assembly; Table of NMR chemical shifts used
in product assignments; NMR spectra of starting salts 2-4 in KOH solution before and during
decomposition; Summary of data from degradation of salt 4 in a PTFE-lined sealed reactor.
24
ACS Paragon Plus Environment

Page 25 of 30
The Journal of Organic Chemistry
1
2
3
4
References
5
6
1) Arges, C.G.; Zhang, L. Anion Exchange Membranes’ Evolution toward High Hydroxide Ion
7
8
Conductivity and Alkaline Resiliency. ACS Applied Energy Materials 2018, 1, 2991-3012.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2) Hagesteijn, K. F. L.; Jiang, S.; Ladewig, B. P. A Review of the Synthesis and Characterization of
Anion Exchange Membranes. J. Mater. Sci. 2018, 53, 11131-11150.
3) Sun, Z.; Lin, B.; Yan, F. Anion-Exchange Membranes for Alkaline Fuel-Cell Applications: The
Effects of Cations. ChemSusChem 2018, 11, 58–70.
4) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois, D. L.; Dupuis, M.; Ferry, J. G.;
Fujita, E.; Hille, R.; Kenis, P. J. A.; Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F,; Portis, A. R.;
Ragsdale, S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.; Waldrop, G. L.
Frontiers, Opportunities, and Challenges in Biochemical and Chemical Catalysis of CO2 Fixation.
Chem. Rev. 2013, 113, 6621-6658.
5) Costentin, C.; Robert, M.; Saveant, J.-M. Catalysis of the Electrochemical Reduction of Carbon
Dioxide. Chem. Soc. Rev. 2013, 42, 2423-2436.
6) DuBois, D. L. Electrochemical Reactions of Carbon Dioxide, in Encyclopedia of Electrochemistry,
Bard, A. J.; Stratmann, M.; Scholtz, F.; Pickett, C. J. eds., 2006, 7A, 202-224, Wiley-VCH,
Weinheim, Germany.
7) Gattrell, M.; Gupta, N.; Co, A. A Review of the Aqueous Electrochemical Reduction of CO2 to
Hydrocarbons at Copper. J. Electroanal. Chem. 2006, 594, 1-19.
8) Hori, Y. Electrochemical CO2 Reduction on Metal Electrodes, in Modern Aspects of
Electrochemistry, Vayenas, C. G.; White, R. E.; Gamboa-Aldeco, M. E. eds., 2008, 42, 89-189,
Springer.
9) Inglis, J. L.; MacLean, B. J.; Pryce, M. T.; Vos, J. G. Electrocatalytic Pathways Towards
Sustainable Fuel Production from Water and CO2. Coord. Chem. Rev. 2012, 256, 2571-2600.
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 30
1
2
3
4
5
6
7
8
9
10) Kondratenko, E. V.; Mul, G.; Baltrusaitis, J.; Larrazabal. G. O.; Perez-Ramirez, J. Status and
Perspectives of CO2 Conversion Into Fuels and Chemicals by Catalytic, Photocatalytic and
Electrocatalytic Processes. Energy Environ. Sci. 2013, 6, 3112-3135.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11) Lim, R. J.; Xie, M.; Sk, M. A.; Lee, J. M.; Fisher, A.; Wang, X.; Lim, K. H. A Review on the
Electrochemical Reduction of CO2 in Fuel Cells, Metal Electrodes and Molecular Catalysts. Catal.
Today 2014, 233, 169-180.
12) Ma, J.; Sun, N.; Zhang, X.; Zhao, N.; Xiao, F.; Wei, W.; Sun, Y. A Short Review of Catalysis for
CO2 Conversion. Catal. Today 2009, 148, 221-231.
13) Qiao, J.; Liu, Y.; Hong, F.; Zhang, J. A Review of Catalysts for the Electroreduction of Carbon
Dioxide to Produce Low-Carbon Fuels. Chem. Soc. Rev. 2014, 631-675.
14) a) Kutz, R. B.; Chen, Q.; Yang, H.; Sajjad, S. D.; Liu, Z.; Masel, R. I. Sustainion™ Imidazolium-
Functional Polymers for Carbon Dioxide Electrolysis. Energy Technol. 2017, 5, 929. b) Masel, R.
I.; Sajjad, S. D.; Gao, Y.; Liu, Z.; Chen, Q. Ion-Conducting Membranes. US Patent 9,849,450
(2017).
15) There are many literature reports describing work on 1; see for instance refs 1-3. Only a
representative sampling is given herein.
16) Ponce-Gonzalez, J.; Varcoe, J. R.; Whelligan, D. Commercial Monomer Availability Leading to
Missed Opportunities? Anion-Exchange Membranes Made from meta-Vinylbenzyl Chloride
Exhibit an Alkali Stability Enhancement. ACS Appl. Energy Mater. 2018, 1, 1883−1887.
17) Hugar, K. M.; Kostalik, H. A.; G. W. Coates, G. W. Imidazolium Cations with Exceptional Alkaline
Stability: A Systematic Study of Structure−Stability Relationships. J. Am. Chem. Soc. 2015, 137,
8730-8737 and Supporting Information.
26
ACS Paragon Plus Environment

Page 27 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
18) Yang, C.; Wang, S.; Jiang, L.; Ma, W.; Sun, G. Comparison of Alkaline Stability of Quaternary
Ammonium- and 1,2-Methylimidazolium-Based Alkaline Anion Exchange Membranes. J. Membr.
Sci. 2015, 487, 12–18 and Supporting Information.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
19) a) Long, H.; Kim, K.; Pivovar, B. S. Hydroxide Degradation Pathways for Substituted
Trimethylammonium Cations: A DFT Study. J. Phys. Chem. C 2012, 116, 9419−9426. b)
Sturgeon, M. R.; Macomber, C. S; Engtrakul, C.; Long,. H.; Pivovar, B. S. Hydroxide based
Benzyltrimethylammonium Degradation: Quantification of Rates and Degradation Technique
Development. J. Electrochem. Soc. 2015, 162 (4), F366-F372.
20) Mohanty, A.; Bae, C. Mechanistic Analysis of Ammonium Cation Stability for Alkaline Exchange
Membrane Fuel Cells. J. Mat. Chem. A 2014, 2, 17314-17320.
21) Deavin, O. I.; Murphy, S.; Ong, A. L.; Poynton, S. D.; Zeng, R.; Herman, H.; Varcoe, J. R. Anion-
Exchange Membranes for Alkaline Polymer Electrolyte Fuel Cells: Comparison of Pendent
Benzyltrimethylammonium- and Benzylmethylimidazolium-Head-Groups. Energy Environ. Sci.
2012, 5, 8584−8597.
22) Liu, F. H.; Lin, C. X.; Hu, E. N.; Yang, Q.; Zhang, Q. G.; Zhu, A. M.; Liu, Q. L. Anion Exchange
Membranes with Well-Developed Conductive Channels: Effect of the Functional Groups. J.
Membr. Sci. 2018, 564, 298–307 and Supporting Information.
23) Yang, C.; Wang, S.; Jiang, L.; Hu, J.; Ma, W.; Sun, G. 1,2-Dimethylimidazolium-Functionalized
Cross-Linked Alkaline Anion Exchange Membranes for Alkaline Direct Methanol Fuel Cells. Int. J.
Hydrogen Energy 2015, 40, 2363-2370.
24) Si, Z.; Sun, Z.; Gu, F.; Qiu, L.; Yan, F. Alkaline Stable Imidazolium-Based Ionomers Containing
Poly(arylene ether-sulfone) Side Chains for Alkaline Anion Exchange Membranes. J. Mater.
Chem. A 2014, 2, 4413-4421.
25) Lin, B.; Dong, H.; Li, Y.; Si, Z.; Gu, F.; Yan, F. Alkaline Stable C2-Substituted Imidazolium-Based
Anion-Exchange Membranes. Chem. Mater. 2013, 25, 1858-67.
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 30
1
2
3
4
5
6
7
26) Long, H.; Pivovar, B. Hydroxide Degradation Pathways for Imidazolium Cations: A DFT Study.
J. Phys. Chem. C 2014, 118, 9880−9888.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
27) Ye, Y.; Elabd, Y. A. Relative Chemical Stability of Imidazolium-Based Alkaline Anion Exchange
Polymerized Ionic Liquids. Macromolecules 2011, 44, 8494-8503.
28) Yang, C.; Wang, S.; Ma, W.; Jiang, L.; Sun, G. Highly Alkaline Stable N1-Alkyl Substituted 2-
Methylimidazolium Functionalized Alkaline Anion Exchange Membranes. J. Mater. Chem. A
2015, 3, 8559-8565.
29) Caustic Potash Handbook, Occidental Chem. Corp., March 2018
(https://www.oxy.com/OurBusinesses/Chemicals/Products/Documents/CausticPotash/KOH-
Handbook-2018-ed.pdf).
30) a) Sowmiah, S; Srinivasadesikan, V.; Tseng, M.-C.; Chu, Y.-H. On the Chemical Stabilities of
Ionic Liquids. Molecules 2009, 14, 3780-3813. b) Sun, Z.; Pan, J.; Guo, J.; Yan, F. The Alkaline
Stability of Anion Exchange Membrane for Fuel Cell Applications: The Effects of Alkaline Media.
Adv. Sci. 2018, 5, 1800065 (1-11).
31) a) Teflon™ FEP Fluoropolymer Film Information Bulletin, The Chemours Company,
https://www.chemours.com/Teflon_Industrial/en_US/assets/downloads/teflon-fep-film-
technical-information.pdf. b) We have been unable to find any literature addressing migration
of low-molecular weight amines into fluoropolymers such as FEP. However, adsorption of
polyethyleneimine on fluoropolymer surfaces has been reported: Coupe, B.; Evangelista, M. E.;
Yeung, R. M.; Chen, W. Surface Modification of Poly(tetrafluoroethylene-co-
hexafluoropropylene) by Adsorption of Functional Polymers. Langmuir 2001, 17, 1956-1960.
32) Gill, G. B. Benzil-Benzilic Acid Rearrangements. Comprehensive Org. Synth. 1991, 3, 821-838.
33) If hydrolytic ring fragmentation were the only decomposition route, yield of acetate would be
1.00 mole per mole of salt 3 decomposed, and the yield of formate would be 1.21 mole per
28
ACS Paragon Plus Environment

Page 29 of 30
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
mole of salt 2 decomposed (assuming that formate has two sources – hydrolysis at the
imidazolium C2 carbon and base-induced scission of glyoxal side decomposition product -- and
there are no other losses for the glyoxal-derived portion of the molecule besides benzilic acid
rearrangement.)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
34) Boiling points at 1 atm: methylamine, -6.3 ⁰C; benzylamine, 185 ⁰C. The Merck Index, Eighth
Edition, Merck & Co., Rahway, NJ, 1968, pp 138 and 680.
35) Niemela, K. The Formation of Hydroxy Monocarboxylic Acids and Dicarboxylic Acids by Alkaline
Thermochemical Degradation of Cellulose. J. Chem. Tech. Biotechnol. 1990, 48, 17-28.
36) Acetate has been reported as a product from treatment of 2,3-butanedione in base: Machell, G.
The Action of Alkali on Diacetyl. J. Chem. Soc. 1960, 683-687.
37) a) A solution of 0.25 wt % 2,3-butanedione in 4.25M KOH was prepared in a polypropylene
screw top vial, and an aliquot was transferred to a FEP NMR tube liner which was heated at 80
⁰C for five days. Analysis by proton NMR showed a mixture of 57 mole % acetate, 12 mole % 2-
hydroxyisobutyrate, and 31 mole % formate as the major products. b) A mixture of 0.5 wt %
tetramethylimidazole in 4.25M KOH was prepared in a polypropylene screw top vial and allowed
to stand at room temperature for 12 days, during which period the initially two-phase mixture
of a small amount of solid in liquid had become a clear solution. An aliquot was transferred to a
FEP NMR tube liner and the solution was heated at 80 ⁰C for four days. Analysis by proton NMR
showed a mixture of 16 mole % unreacted tetramethylimidazole, 42 mole % acetate, 28 mole %
2-hydroxyisobutyrate, 12 mole % methylamine, and 2 mole % formate. c) While formate was
definitely observed as a product in both of these reactions, its relative yield appears to be
sensitive to the details of the reaction. The mechanism for production of formate from base-
induced decomposition of 2,3-butanedione and tetramethylimidazole, and the role it might play
in decomposition of 4, is currently a mystery.
38) Ferraz, C. P.; Garcia, M. A. S.; Teixeira-Neto, E.; Rossi, L. M. Oxidation of Benzyl Alcohol Catalyzed by
Gold Nanoparticles Under Alkaline Conditions: Weak vs. Strong Bases. RSC Adv. 2016, 6, 25279–
25285.
29
ACS Paragon Plus Environment