Rapid Analysis of Tetrakis(dialkylamino)phosphonium Stability in Alkaline Media

Article abstract of DOI:10.1021/acs.organomet.7b00663

Hydroxide-stable organic cations are crucial components for ion-transport processes in electrochemical energy systems, and the tetrakis(dialkylamino)phosphonium cation is a promising candidate for this application. These phosphoniums are known to be highly resistant to alkaline media; however, very few investigations have systematically evaluated how these cations decompose in the presence of hydroxide or alkoxide anions. The excellent stability of several tetraaminophosphoniums in 2 M KOH/CH₃OH at 80 °C led us to design experiments for the rapid assessment of phosphonium degradation in homogeneous solution and under phase-transfer conditions. The analysis illustrated how substituents around the cation core affect both degradation pathways and rates. β-H elimination and direct attack at the phosphorus atom are the most common degradation pathways observed in an alcoholic solvent, while α-H abstraction and direct attack are observed under phase-transfer conditions (PhCl and 50 wt % NaOH/H₂O). The collected data provided a relative stability comparison for this family of cations to enable future design improvements and illustrated the utility of using multiple tests for degradation studies.

Full text of DOI:10.1021/acs.organomet.7b00663

Article
pubs.acs.org/Organometallics
Rapid Analysis of Tetrakis(dialkylamino)phosphonium Stability in
Alkaline Media
C. Tyler Womble,† Jamie Kang,† Kristina M. Hugar,‡ Geoffrey W. Coates,‡ Stefan Bernhard,†
and Kevin J. T. Noonan*,†
†Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213-2617, United States
‡Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell University, Ithaca, New York 14853-1301, United States
S
* Supporting Information
ABSTRACT: Hydroxide-stable organic cations are crucial compo-
nents for ion-transport processes in electrochemical energy systems,
and the tetrakis(dialkylamino)phosphonium cation is a promising
candidate for this application. These phosphoniums are known to
be highly resistant to alkaline media; however, very few
investigations have systematically evaluated how these cations
decompose in the presence of hydroxide or alkoxide anions. The
excellent stability of several tetraaminophosphoniums in 2 M KOH/
CH3OH at 80 °C led us to design experiments for the rapid
assessment of phosphonium degradation in homogeneous solution
and under phase-transfer conditions. The analysis illustrated how
substituents around the cation core affect both degradation
pathways and rates. β-H elimination and direct attack at the phosphorus atom are the most common degradation pathways
observed in an alcoholic solvent, while α-H abstraction and direct attack are observed under phase-transfer conditions (PhCl and
50 wt % NaOH/H2O). The collected data provided a relative stability comparison for this family of cations to enable future
design improvements and illustrated the utility of using multiple tests for degradation studies.
and OH−.11 Additionally, these cations are derived from a class
of organic superbases known as phosphazenes.12 Though the
INTRODUCTION
■
Materials that enable selective ion transport under a wide range
utility of these phosphoniums has been well documented, few
reports have detailed the limits of stability for these cations in
the presence of hydroxide anions (Figure 1). Marchenko and
of conditions (temperature, pH, oxidative stress) are necessary
for chemical separation processes and electrochemical energy
systems (e.g, redox-flow batteries and fuel cells).1 Solid polymer
electrolytes are a natural choice for these applications, since
polymers can be easily decorated with various ionic groups to
provide pathways for charge movement. While polymers with
pendant anions have been employed extensively as proton
conductors for ion transport,1a anion-exchange membranes
(AEMs) with pendant cations to conduct hydroxide often suffer
from limited chemical stability.2
Preparation of hydroxide-stable AEMs has attracted signifi-
cant attention for fuel cell devices,3−5 since non-noble metals
can be used for O2 reduction under alkaline conditions.2 The
limited chemical durability of most AEMs is often directly
related to the cation appended to the polymer chain.2
A
number of model compound studies have appeared on organic
cations in an effort to better understand decomposition
pathways and improve alkaline stability (e.g., ammonium,6
imidazolium,7 guanidinium,8 and arylphosphonium9). These
studies offer a direct means to determine how cations
decompose before they are appended to a polymer chain.
Tetrakis(dialkylamino)phosphonium cations have attracted
our attention as candidates for AEMs10 due to their impressive
resistance to hydroxide. These cations have already been
explored as counterions for anion delivery, including F−, Cl−,
Figure 1. Decomposition pathways of [P(NR2)4]+ cations.
Received: August 31, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
co-workers first identified the potential decomposition path-
ways of [P(NR2)4]+ cations with linear alkyl substituents (R =
13
n
n
Me, Et, Pr, Bu). Schwesinger significantly expanded on this
work and investigated the nature of the R group and the size of
the phosphorus−nitrogen framework for phase-transfer catal-
ysis.14 Both studies established a clear link between molecular
structure and alkaline stability; however, decomposition
pathways were only determined for a few select systems.13,14
Herein, we systematically evaluated how a series of
tetrakis(dialkylamino)phosphoniums degrade in homogeneous
solution and under phase-transfer conditions. The experiments
were designed to enable rapid screening of decomposition
pathways for these systems and inform future molecular design.
The utilization of harsh conditions and short experiment times
are key to accelerated discovery, ensuring that the best cations
are targeted for membrane applications.
Figure 2. Alkaline stability studies of 1 in 2 M KOH/CD3OH at 80 °C
(experimental setup according to ref 7b).
RESULTS AND DISCUSSION
■
A number of factors must be considered when evaluating cation
degradation under alkaline conditions. The reaction medium,
the concentration of the reactants, and the temperature all
significantly affect degradation rates for these ions. Subtle
changes in any of these parameters can dramatically influence
the rate of decomposition and the pathway by which the cation
degrades. Consequently, direct comparisons between inde-
pendent studies are difficult, as reaction setups are rarely
identical.6b,c Most stability investigations of model compounds
are conducted in H2O as the solvent: with a concentration of
cation between 0.01 and 1 M, a concentration of hydroxide
between 1 and 6 M, and temperatures between 60 and 160
°C.6c While H2O is the most relevant solvent for H2-based fuel
cells, the solvating power of water limits the nucleophilicity and
basicity of the hydroxide anion and leads to longer experiment
times.6b,h Rather than explore stability over periods of days or
months, rapid screening (hours to minutes) could enable
molecular design improvements much more quickly. This
approach would be complementary to longer-term stability
studies in H2O but help delineate how different cations degrade
much more quickly.
Changing the reaction medium to solvents with lower
dielectric constants will result in more aggressive reaction
conditions. Methanol has already been used in a number of
stability studies, since it accelerates degradation and the data are
relevant for direct methanol fuel cells (DMFCs).6a,7b,9,10b When
investigating tetrakis(dialkylamino)phosphoniums, we initially
selected CH3OH as the solvent for decomposition studies;
however, it did not facilitate rapid screening. We have
previously reported that [P(N(Me)Cy)4]+ exhibits no signs of
degradation in NaOD/CD3OD over a period of 20 days.10b For
comparison, we explored [P(N-pyrrolidinyl)4]+ under similar
conditions and found that this much smaller cation also shows
no signs of change in 30 days (Figure 2).
dissolving KOH (2 M) and the phosphonium (0.05 M) in
degassed 2-(2-methoxyethoxy)ethanol followed by heating at
160 °C (Scheme 1). Aliquots were removed from the mixture
after 4 h and analyzed to determine which phosphorus species
had formed (P-imide, P-oxide, and Phosphine, Scheme 1).
Three types of dialkylamino groups were evaluated: cyclics,
straight-chain alkyls, and branched alkyls.
Cyclic Substituents. The first class of phosphonium
compounds investigated included cyclic amino groups on
phosphorus (compounds 1 and 2 in Figure 3). These two
systems were the least stable of all tetraaminophosphoniums
investigated in this study, and temperatures below 160 °C could
bring about decomposition. Compound 1 degraded completely
within several hours at 120 °C with formation of a single
phosphorus species in the 31P{1H} NMR spectrum (Figure S24
in the Supporting Information). Mass spectrometry confirmed
that this signal corresponds to the tris(N-pyrrolidinyl)-
phosphine oxide (P-oxide). Two decomposition pathways are
possible to form the P-oxide: direct attack of the hydroxide at
the P atom (pathway A, Figure 1) and subsequent formation of
the phosphine oxide or α-H abstraction and oxidation of the
three-coordinate phosphine (pathway C, Figure 1). Since no
three-coordinate phosphine was observed in the NMR
spectrum, direct nucleophilic attack at the phosphorus atom
by the anion is most likely.
Direct attack is commonly postulated for decomposition of
phosphonium hydroxides ([PR4]OH) with alkyl or aryl
groups.15 Initial attack of OH− occurs at phosphorus
(R4POH), and a second equivalent of OH− deprotonates this
species to form an oxyanion (R4PO−). The rate-determining
step is loss of a carbanion (R−) to produce the neutral
phosphine oxide (with rapid protonation of R−). This
mechanistic interpretation is consistent with a reaction that is
second order in OH− and first order in phosphonium.15 In our
case, the large excess of base (2 M) in comparison to the
tetraaminophosphonium (0.05 M) did not permit determi-
nation of the reaction order for the hydroxide. However, we
were able to ensure that phosphine oxide formation from the
phosphonium species is a pseudo-first-order process, as
evidenced by a series of spectra collected over 2 h at 120 °C
(Figure 4).
Much harsher conditions were necessary to bring about
[P(NR2)4]+ degradation on a fast time scale; therefore, methyl
carbitol (2-(2-methoxyethoxy)ethanol) was selected as a
solvent. It has a low dielectric constant (14.8), is miscible
with water, and has a boiling point of 194 °C. This alcohol,
combined with KOH, was selected to access higher reaction
temperatures for screening.
Eight tetrakis(dialkylamino)phosphoniums were synthesized
to systematically probe decomposition with different amino
groups. Decomposition pathways for these structures are
summarized in Figure 3. Degradation was achieved by
We synthesized compound 2 with N-piperidinyl substitu-
ents16 and subjected it to the similar harsh conditions applied
to compound 1. Slightly higher temperatures were needed to
bring about decomposition of 2 in comparison to 1, suggesting
B
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 3. Summary of the decomposition pathways observed for the [P(NR2)4]+ cations in 2 M KOH/2-(2-methoxyethoxy)ethanol.
Scheme 1. Possible Decomposition Products of
Tetrakis(dialkylamino)phosphoniums in Alkaline Media
Table 1. Ratios of the Different Phosphorus Species after 4 h
in 2 M KOH/2-(2-Methoxyethoxy)ethanol
a
Percentage of unidentified signals in the 31P{1H} NMR spectrum.
Half-life for compound 2 was estimated after 0.5 h.
b
consistent with Marchenko’s observations.13 Elimination is
likely the preferred pathway for compound 3 with the β-Me
groups. The longer alkyl chains in compound 4 provide
shielding to slow β-H elimination but cannot prevent direct
attack at phosphorus.
Figure 4. Plot of phosphonium decay (compound 1) versus time in 2
M KOH/2-(2-methoxyethoxy)ethanol at 120 °C. Inset: pseudo-first-
order kinetic analysis.
Branched Substituents. Branched groups offer a greater
degree of steric shielding around the central atom in these
cationic structures. The combination of compound 5 with
KOH in methyl carbitol at 160 °C produced the P-imide as the
major decomposition product by loss of an isopropyl group,
supporting the β-H elimination hypothesis (confirmed by ESI).
Only a minor amount of P-oxide is formed in the
decomposition of compound 5 (Table 1). A plot of the
reaction progress for 5 revealed that pseudo-first-order behavior
is still observed when the elimination pathway is operational
(Figure 5).
Extension of the branched alkyl group to a 3-pentyl (6)
rather than an isopropyl group impeded the rate of elimination
but did not completely prevent it (Table 1). β-H elimination is
minimal for compound 7, as expected from the work of
Schwesinger and co-workers.14 Compound 8, with the ethyl
substituents, exclusively forms the P-imide, as these groups
promote the elimination pathway.
an increase in steric shielding around phosphorus. At 140 °C,
some loss of 2 was detected, but 160 °C was used to ensure
complete decomposition, also by direct attack (Figure S25 in
the Supporting Information).
Straight-Chain Substituents. The diethylamino and
dibutylamino derivatives 3 and 4 had both been synthesized
previously.13,17,18 Decomposition of these compounds was
nearly complete after 4 h at 160 °C. Combination of 3 with
KOH in methyl carbitol (Figure S26 in the Supporting
Information) produced the P-imide as the major decom-
position product (loss of an ethyl group). Some concurrent P-
oxide formation was also observed with an ∼3:1 ratio of the
phosphazene to the phosphine oxide (Table 1). Loss of the
ethyl group from the structure was confirmed by analysis of the
reaction mixture using mass spectrometry (ESI). Interestingly,
the formation of these two products, as well as their ratios, are
consistent with the observations of Marchenko and co-workers.
They explored decomposition of 3 by reaction of the bromide
salt with anhydrous metal hydroxide under vacuum at high
temperature.13 Extension of the alkyl chains to butyl groups
afforded nearly exclusive formation of P-oxide (Table 1), also
A summary of all the collected data is included in Table 1.
The relative amounts of phosphorus products were determined
by integration of the final NMR spectrum for each reaction
mixture. In some of the spectra, minor unassigned signals were
C
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
low-polarity solvents while ensuring only OH− is present.20 We
explored phase-transfer reactions as a possible means of
decomposition for 5−7 using a method similar to that of
Landini and co-workers.21 We focused on the branched systems
due to their performance in the alcoholic KOH studies.
Phase-transfer catalysis is typically used to accelerate
reactions between an organic substrate and a reactive anion.
The anion (insoluble in organic solvents) and the substrate
(insoluble in aqueous) are combined in a two-phase system. A
catalytic amount of a lipophilic transport reagent facilitates
extraction of the reactive anion into the organic phase, where
the reaction takes place. Interestingly, researchers have noted
previously that transport agents such as quaternary ammoniums
and phosphoniums display limited stability under strongly
alkaline conditions and degrade by Hofman elimination and
direct attack, respectively.22 Landini and co-workers extensively
explored the decomposition of these quaternary onium salts
using two-phase reaction systems (chlorobenzene and NaOH/
H2O).21 Consequently, phase transfer should serve as an
effective tool to evaluate the stability of lipophilic cations.
Four parameters are extremely important for carrying out
these phase-transfer reactions, and each variable affects the
decomposition rate. First, the counterion selected for the
organic cation has a dramatic effect on the amount of OH−
extracted into the organic phase (KOH/Xsel).21 To maximize the
concentration of OH− in the organic phase for decomposition,
if a halogen counterion is used, then Cl− is best.21 This will
result in accelerated tests in comparison to other halides, since
extraction of anions in low-polarity solvents follows the
sequence OH− ≪ Cl− < Br− < I−.20 Second, higher
temperatures will result in faster decomposition with more
extraction of OH− into the chlorobenzene.21 Third, as the
concentration of NaOH is increased in the aqueous phase,
hydration of the OH− in the organic phase is diminished,
increasing its reactivity.22c Finally, high stirring rates are
necessary to ensure rapid extraction of the reactive anion into
the organic phase.21 To compare directly with prior work from
Schwesinger, we exposed compounds 5−7 with a Cl−
counterion to 50 wt % NaOH/H2O in chlorobenzene at 100
°C (Scheme 2). Decomposition reactions proceeded rapidly
Figure 5. Plot of phosphonium decay (compound 5) versus time in 2
M KOH/2-(2-methoxyethoxy)ethanol at 160 °C. Inset: pseudo-first-
order kinetic analysis.
observed, which were also given a percentage (less than 5% in
all cases). The two most common decomposition pathways are
direct attack at phosphorus and β-H elimination. The trends are
clear: direct attack of the anion at the P atom occurs with cyclic
substituents such as pyrrolidine and piperidine. With linear
alkyl chains on the amino group, the predominant decom-
position pathways are β-H elimination and direct attack at
phosphorus. Branched alkyl groups can enhance steric
shielding, but β-H elimination tends to dominate as the
mode of decomposition with these derivatives unless
substituents (e.g, cyclohexyl groups) are employed to block
this pathway. Considering pseudo-first-order decay was
observed for both 1 and 5, the half-lives of all cations were
estimated from the 4 h time point on the basis of the relative
ratios of different phosphorus species in the mixture. This
information was used to provide a relative stability comparison
for the different derivatives (Table 1).
Linear alkyls on the amino group provide some measure of
protection in alkaline media, as compounds 3 and 4 do not
completely decompose after 4 h under the harsh conditions
employed (12% and 5% remaining, respectively). However,
branched substituents substantially improve the stability of
tetrakis(dialkylamino)phosphonium salts (higher t1/2 observed
for compounds 5−8). The branching offers increased steric
protection, impeding direct attack at phosphorus. Compound
7, first synthesized by Schwesinger, exhibited the highest degree
of alkaline resistance in methyl carbitol due to the cyclohexyl
group. The role of this substituent is 2-fold: first, it imparts
steric bulk around the central atom and limits attack at P.
Second, the preferred conformation of the cyclohexyl group will
have the substituent in an equatorial position, impeding β-H
elimination.14 The decomposition of compound 7 was much
slower than that for all the other derivatives with a half-life
greater than 10 h in 2 M KOH/2-(2-methoxyethoxy)ethanol at
160 °C, a testament to its excellent alkaline resistance.
Scheme 2. Phase-Transfer Conditions Employed for
Decomposition
under these conditions.23 Interestingly, significant changes in
the major decomposition pathways were observed in
comparison to our data obtained in methyl carbitol (Table
2). For compounds 5 and 7, only phosphine oxide was formed,
indicating direct attack, while, for compound 6, α-H abstraction
is predominant.
In stark contrast to the alcohol study where β-H elimination
is dominant, this pathway was not observed in the two-phase
system. In methyl carbitol, the bulkier alkoxide is present and
thus elimination is favored.24 Under phase-transfer conditions,
the smaller hydroxide is the only anion present, and direct
attack dominates. For alkaline fuel cells (AFCs), this
information is relevant, since fuel choices (ROH versus H2)
will likely influence which anions are present in the electro-
Phase-Transfer Reactions. In alcoholic solvents, the
mixture of OR− and OH− is likely affecting decomposition
pathways and degradation rates. Researchers have already
observed differences in products from [PR4]+ decomposition
depending on reaction conditions, solvents, and anions.19 To
ensure only hydroxide promotes decomposition in solution,
water must be used as the reaction medium. While
decomposition studies in H2O can be slow, phase-transfer
reactions can be used to obtain strongly alkaline solutions in
D
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
conditions. This is in contrast to the study in methyl carbitol,
where 7 is more stable. We suspect in the alcohol that the
bulkier alkoxide anions favor elimination, which the cyclohexyl
groups prevent. In the phase-transfer reaction with only
hydroxide anions present, 6 has improved steric shielding
(limiting direct attack) due to the increased conformational
flexibility provided by the 3-pentyl group.25 This enhances the
stability of 6 in comparison to 7. Space-filling structures of the
computed cations (Figure 7) illustrate the steric constraints
Table 2. Ratios of the Different Phosphorus Species in PhCl
(0.04 M) with 50% NaOH/H2O at 100 °C
a
Percentage of unidentified signals in the 31P{1H} NMR spectrum.
chemical cell and could promote different decomposition
pathways.
Interestingly, we observed rapid degradation of phospho-
nium salts 5 and 7 in comparison to the prior report.14 Nearly
complete degradation of 5 occurred within 1 h (94%) and 90%
degradation of 7 within 2 h (the previously reported t1/2 value
for 7 was 67 h14). A slightly lower concentration of cation (0.04
M versus 0.07 M) was used in our study, along with larger
amounts of NaOH in the aqueous layer, but the change in half-
life is dramatic. We suspect the large difference can be traced
back to different stir rates and the counterion exchange
procedures. We used an anion-exchange resin to obtain
−
phosphonium chlorides, while Schwesinger conducted a BF4
to Cl− anion exchange using methanol as a solvent. Landini’s
report indicated how critical the Cl− counterion is to maximize
extraction of the hydroxide into the organic phase.21 If anion
exchange to the Cl− were incomplete, this would affect
decomposition rates.
Figure 7. Space-filling structures of the computed (B3PW91/6-
31G(d,p)) phosphonium cations with 3-pentyl and cyclohexyl
substituents.
We completed phase-transfer studies on [7]PF6 and
compared it to a 50/50 mixture of [7]PF6 and [7]Cl to
illustrate how counterions can alter decomposition rates. The
percent cation remaining was determined after 2 h, and the
differences are remarkable (Figure 6). No degradation is
imposed by the two branched substituents on P and highlight
the limited accessibility of the N−H groups for the 3-pentyl
substituent in comparison to the cyclohexyl precursor.
Another indicator of the steric crowding imposed by this
substituent is clear in the attempted ethylation of [P(N(H)3-
Pent)4]+. To date, we have been unable to achieve complete
ethylation, whereas [P(N(Et)Cy)4]+ could be synthesized at
higher temperatures. This suggests that substitution reactions
with cyclohexyl substituents on the aminophosphonium are
more facile. This analysis is also indicative that further
optimization around the amino phosphonium core can lead
to improvements in stability.
Redox Stability. Finally, phosphonium cations for AFCs
must exhibit good electrochemical stability in addition to
alkaline stability. To ensure this, we conducted a cyclic
voltammetry experiment with [7]PF6 as the supporting
electrolyte (Figure 8). Compound 7 was selected, since it has
been used in polyolefin AEMs previously.10b Compound 7
functioned as an excellent supporting electrolyte for the
ferrocene/ferrocenium redox couple, rivaling the behavior of
[NBu4]PF6. The redox stability of [7]PF6 provides further
support for use of these cations in AEMs.
Figure 6. Change in decomposition rate as a function of counterion
choice for compound 7 in PhCl (0.04 M) and 50% NaOH/H2O.
observed at all with only PF6− as the counterion, while a 50/50
mixture of the chloride and hexafluorophosphate forms
produced 50% loss of 7 in only 2 h. If only [7]Cl is present,
90% loss of the cationic structure is observed. Consequently,
counterion exchange is an important consideration when
stability is investigated.
CONCLUSION
■
The promise of tetrakis(dialkylamino)phosphoniums for AEMs
led us to explore the possible decomposition pathways of these
cations in the presence of oxyanions. This family of cations
required extremely harsh conditions to bring about decom-
position, which is a strong indicator of their stability.
Another interesting feature was the lower stability of
compound 7 in comparison to 6 under phase-transfer
E
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
phosphonium hexafluorophosphate(V) ([P(N(H)Cy)4]PF6) was
synthesized according to a previous report.14
NMR Analysis. All NMR spectra were recorded on either a 300 or
1
500 MHz Bruker Avance spectrometer. The H NMR spectra were
referenced to residual protio solvents (CHCl3, 7.26 ppm; CHDCl2,
5.32 ppm; CHD2OH, 3.31 ppm; DMSO-d5, 2.50 ppm; acetone-d5,
2.05 ppm), and 13C{1H} NMR were referenced to the solvent signal
(CDCl3, 77.23 ppm; CD2Cl2, 54.00 ppm; DMSO-d6, 39.51 ppm;
acetone-d6, 29.92 ppm). The 31P{1H} NMR spectra were electronically
referenced using internal Bruker software according to a universal scale
determined from the precise ratio, Ξ, of the resonance frequency of
1
the 31P nuclide to the H resonance of TMS in a dilute solution (φ <
1%).26 Quantitative 31P{1H} NMR spectra were recorded using
inverse-gated decoupling for the decomposition studies (compounds 5
and 7 required 256 and 2000 scans, respectively, due to limited
solubility, and spectra were recorded at 65 °C). A relaxation delay of
20 s was employed to ensure accurate integrations for the final time
points. The percent cation remaining was determined by integration of
all phosphorus signals present in the mixture and calculating a ratio
Figure 8. Cyclic voltammogram of 5 mM ferrocene (black trace)
recorded at 100 mV/s in degassed MeCN with [7]PF6 as the
supporting electrolyte (0.05 M). The blue trace is a voltammogram
with only the supporting electrolyte ([7]PF6).
−
(the PF6 signal was excluded from the analysis).
Mass Spectrometry (MS). High-resolution MS was performed in
the School of Chemical Sciences Mass Spectrometry Laboratory at the
University of Illinois, Urbana−Champaign, IL. Low-resolution ESI-MS
to determine decomposition products was performed on an LCQ
Classic (ThermoFisher Scientific Corp., San Jose, CA) APCI
quadrupole ion trap mass spectrometer operating with the XCalibur
software (Version 1.3). Data were recorded continuously with a
scanned mass range of 150−1000 m/z using a spray voltage of 3.0 kV,
a capillary temperature of 200 °C, and nitrogen as the carrier gas.
GC-MS Analysis. GC-MS analysis was performed on a Hewlett-
Packard Agilent 6890-5973 GC-MS workstation. The GC column was
a Hewlett-Packard fused silica capillary column cross-linked with 5%
phenylmethylsiloxane. Helium was used as the carrier gas. The
following conditions were used for all GC-MS analyses: injector
temperature, 250 °C; initial temperature, 70 °C; temperature ramp, 10
°C/min; final temperature, 280 °C.
Degradation studies were conducted in methyl carbitol with
KOH at high temperatures to better understand the modes of
decay for these cationic structures. In this alcoholic solvent, the
predominant decomposition pathways were direct attack at
phosphorus and β-H elimination. The observed decomposition
pathways changed when degradation studies were carried out
under phase-transfer conditions. This is attributed to the
oxyanion driving decomposition (OR− versus OH−) in the
different reaction media. These studies indicate that both steric
bulk and conformational flexibility play a role in ion stability.
The different fuels employed in AFCs will affect cation stability,
a feature to consider when preparing AEMs.
Synthesis of Tetrakis(pentan-3-ylamino)phosphonium
Hexafluorophosphate(V) ([P(N(H)3-Pent)4]PF6). A 250 mL
Schlenk flask was evacuated and back-filled with N2 three times. The
flask was charged with 150 mL of dry dichloromethane, 3-
aminopentane (5.0 mL, 43 mmol), and triethylamine (6.0 mL, 43
mmol). The reaction vessel was then cooled to 0 °C using an ice bath.
The rubber septum was removed from the top of the flask, and under a
stream of N2, solid PCl5 (2.13 g, 10.2 mmol) was added slowly to the
reaction mixture over a period of 5 min. After complete addition of the
PCl5, the ice bath was removed, the reaction mixture was warmed to
room temperature, and the mixture was stirred for 17 h. An aliquot
was removed and analyzed using 31P{1H} NMR spectroscopy to
ensure complete conversion to the desired product (δ ∼23 ppm). The
reaction mixture was transferred to a separatory funnel, and the
organic solution was washed twice with a saturated KPF6 solution (2 ×
100 mL) and once with water (50 mL). The organic layer was
subsequently dried over anhydrous sodium sulfate and concentrated
using rotary evaporation. The crude solid was dissolved in a minimal
amount of dichloromethane and precipitated into 200 mL of diethyl
ether. The white solid was collected using vacuum filtration and
There is tremendous difficulty comparing degradation rates
between individual studies. This study was meant to provide a
relative comparison for the tetrakis(dialkylamino)phosphonium
salts and determine their decomposition pathways to inform
cation design. Model studies using water have been employed
in the past for ammoniums, but more lipophilic cations (e.g.,
phosphoniums and imidazoliums) are often insoluble in water,
making comparisons difficult. Phase-transfer studies offer a
means to directly compare lipophilic cations while using the
base concentration and temperature as parameters to tune the
basicity of the reaction mixture. Phase transfer does require the
cation to be soluble in chlorobenzene, which may not be
suitable for all AEM cation candidates.
Future work will evaluate degradation in more depth for
these aminophosphoniums: particularly, the oxidation process
and how OR− or OH− interacts with the phosphonium to bring
about P-oxide formation. In the future, other systematic
changes to the substituents on the tetraaminophosphoniums
will be explored to further optimize these structures. Other
phosphonium cations will also be evaluated using the above
methods, as the different alkaline conditions and electro-
chemical measurements are crucial to assess potential AEM
compounds.
1
recrystallized from hot water/ethanol (2.42 g, 45% yield). H NMR
(500 MHz, CD2Cl2): δ 3.13−3.01 (m, 4H), 2.83 (t, J = 11.0 Hz, 4H),
1.60 (dtd, J = 14.5, 7.4, 6.0 Hz, 16H), 0.94 (t, J = 7.4 Hz, 24H).
1
31P{1H} NMR (202 MHz, CD2Cl2): δ 22.2 (s), − 143.4 (sep, JPF
=
2
711.8 Hz). 13C{1H} NMR (126 MHz, CD2Cl2): δ 55.8 (d, JPC = 1.4
Hz), 28.4 (d, 3JPC = 4.8 Hz), 9.9. HRMS (EI) (m/z): calculated [M]+
for C20H48N4P, 375.3617; found, 375.3609.
EXPERIMENTAL SECTION
Synthesis of Tetrakis(methyl(pentan-3-yl)amino)-
phosphonium Hexafluorophosphate(V) (Compound 6). [P(N-
(H)3-Pent)4]PF6 (0.197 g, 0.38 mmol), chlorobenzene (0.64 mL),
0.64 g of 50% (w/w) KOHaq, and methyl iodide (0.25 mL, 4 mmol)
were placed in a 25 mL flask. The flask was placed in an oil bath set at
60 °C, and the reaction mixture was stirred for 17 h. An aliquot from
the organic layer was removed and analyzed by 31P{1H} NMR
■
General Methods. Salts of compounds 1−5 and 7 were prepared
according to previous literature reports and washed with saturated
KPF6 solutions to furnish the expected products (1H and 31P{1H}
NMR spectra are included in the Supporting Information).14,16,18
Benzyltrimethylammonium bromide ([BTMA]Br) was prepared
according to a previous report.7b Tetrakis(cyclohexylamino)-
F
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
spectroscopy to ensure complete conversion to the desired product (δ
∼47 ppm). The mixture was transferred to a separatory funnel and
diluted with 15 mL of water and 20 mL of dichloromethane. The
layers were separated, and the aqueous layer was extracted twice more
with dichloromethane (2 × 20 mL). The organic extracts were
combined and washed once with saturated KPF6 solution (15 mL) and
once with water (15 mL). The combined organic extracts were dried
over anhydrous sodium sulfate and then concentrated using rotary
evaporation. The residue was precipitated into 150 mL of diethyl
ether, and the solids were collected using vacuum filtration. The
product was isolated as a white powder (0.147 g, 67% yield). 1H NMR
(500 MHz, CD2Cl2): δ 3.03 (qt, J = 8.3, 5.2 Hz, 4H), 2.65 (d, J = 10.0
Hz, 12H), 1.83 (dp, J = 15.1, 7.5 Hz, 8H), 1.64−1.50 (m, 8H), 1.01 (t,
J = 7.5 Hz, 24H). 31P{1H} NMR (202 MHz, CD2Cl2): δ 46.6, − 143.6
(sep, 1JPF = 709.5 Hz). 13C{1H} NMR (126 MHz, CD2Cl2): δ 59.5 (d,
removed for an initial time point after the solution became
homogeneous (within a few minutes). The reaction mixture was
stirred at the above temperature for 4 h, and aliquots were removed at
2 and 4 h for analysis using 31P{1H} NMR spectroscopy. We have
noted some variation in the PF6− counterion signal, which we suspect
is due to solubility in the solvent when the mixture is cooled for NMR
analysis. We ensured that the mixtures are homogeneous at the desired
decomposition temperature. For photographs of the experimental
setup, see the Supporting Information.
After 4 h, characterization was completed as follows: a 0.5 mL
aliquot was removed for NMR analysis, and percent cation remaining
was determined by integration of all phosphorus signals present in the
−
mixture and calculation of a ratio (excluding the PF6 signal). A 0.5
mL aliquot was also removed, neutralized with 10% HCl (v/v),
extracted with ethyl acetate (15 mL), and analyzed using GC-MS.
Finally, mixtures which contained P-imide were analyzed using ESI-
MS. Samples were prepared by removing a 0.5 mL aliquot and
neutralizing with 10% HCl (v/v). The mixture was extracted with
dichloromethane (3 × 5 mL), and the organic layer was concentrated
via rotary evaporation. The residue was diluted with 200 mL of
acetonitrile and analyzed.
Example Procedure for Degradation of Compound 1. A 20
mL scintillation vial with a septum-sealed cap was charged with 1 (100
mg, 0.22 mmol) and KOH (490 mg, 8.7 mmol). The vial was
evacuated and back-filled with Ar or N2 seven times, and then degassed
2-(2-methoxyethoxy)ethanol (4.4 mL) was added via syringe. The vial
was placed in an aluminum block set to 120 °C. Aliquots were
removed after 2 and 4 h for analysis by 31P{1H} NMR spectroscopy.
After 4 h, an aliquot was removed and analyzed using GC-MS.
Pseudo-first-order kinetic analysis for compound 1 and 5 were
conducted in an identical manner, except a series of 31P{1H} NMR
spectra were collected at periodic intervals and analyzed.
3
2JPC = 4.2 Hz), 30.6 (d, JPC = 3.8 Hz), 26.4 (br s), 11.9. HRMS (EI)
(m/z): calculated [M]+ for C24H56N4P, 431.4243; found, 431.4240.
Synthesis of Tetrakis(cyclohexyl(ethyl)amino)phosphonium
Hexafluorophosphate(V) (Compound 8). [P(N(H)Cy)4]PF6
(0.198 g, 0.35 mmol), chlorobenzene (1.1 mL), 1.12 g of 50% (w/
w) KOHaq, and ethyl iodide (1.60 mL, 20 mmol) were placed in a 25
mL round-bottomed flask equipped with a reflux condenser. The flask
was placed in an oil bath at 115 °C, and the reaction mixture was
stirred for 48 h. An aliquot from the organic layer was removed and
analyzed by 31P{1H} NMR spectroscopy to ensure complete
conversion to the desired product (δ ∼56 ppm). The reaction mixture
was transferred to a separatory funnel and diluted with 15 mL of water
and 20 mL of dichloromethane. The layers were separated, and the
aqueous layer was extracted twice more with dichloromethane (2 × 20
mL). The organic extracts were combined and washed once with
saturated KPF6 solution (15 mL) and once with water (15 mL). The
combined organic extracts were dried over anhydrous sodium sulfate
and then concentrated using rotary evaporation. The residue was
precipitated into 150 mL of diethyl ether, and the solids were collected
using vacuum filtration. The product was isolated as a white powder
General Procedure for Degradation Using a Two-Phase
Reaction Mixture. The [P(NR2)4]PF6 salts were converted to the
[P(NR2)4]Cl salts by dissolution in MeOH and stirring in the
presence of Amberlite IRA-400 Cl− Exchange Resin (8 g of resin per 1
g of phosphonium salt). The mixture was gently stirred for 17 h, and
the resin beads were removed by filtration. 31P{1H} NMR spectros-
1
and recrystallized from hot water/acetone (0.211 g, 89% yield). H
NMR (500 MHz, CD2Cl2): δ 3.30 (qq, J = 11.7, 4.0 Hz, 4H), 3.13 (p,
J = 7.3 Hz, 8H), 1.91 (dt, J = 13.4, 3.3 Hz, 8H), 1.85−1.67 (m, 20H),
1.39 (t, J = 7.1 Hz, 12H), 1.30 (qt, J = 12.8, 3.7 Hz, 8H), 1.17 (qt, J =
13.2, 3.5 Hz, 4H). 31P{1H} NMR (202 MHz, CD2Cl2): δ 56.5, −
−
copy was used to ensure loss of the PF6 counterion. MeOH was
removed by rotary evaporation, and the slightly brown residue was
dissolved in 20 mL of dichloromethane and washed with saturated
NaCl solution (2 × 10 mL) and water (1 × 5 mL). The organic layer
was concentrated by rotary evaporation, and the salt was further dried
in vacuo at 70 °C. The compounds were obtained as white solids.
The two-phase method employed was similar to prior reports (an
example procedure is included below).14,21 A phosphonium chloride
salt (0.1 g) was placed in a 20 mL scintillation vial and dissolved in
chlorobenzene (0.04 M). A 50% w/w solution of NaOH in deionized
water was prepared and placed in the vial. The weights of NaOH and
H2O (in grams) were matched to the number of milliliters of
chlorobenzene. This ensured a minimum of 300-fold excess of OH− in
the water layer to the phosphonium cation in the organic layer. The
vial was capped (septum-seal), and 50 μL was removed from the
chlorobenzene layer via syringe. This aliquot was diluted with 0.6 mL
of DMSO-d6 for an initial NMR analysis. The vial was placed in an
aluminum block set to 100 °C, with a stirring rate of 800 rpm. After
the specified times, the reaction mixture was cooled and a 50 μL
aliquot was removed, diluted with 0.6 mL DMSO-d6, and analyzed
using NMR spectroscopy. 1H NMR spectroscopy was conducted using
a 30° tip angle, 64 scans, and 1 s delay. 31P{1H} NMR spectroscopy
was conducted using a 30° tip angle, 1000 scans, and 2 s delay. The
percent cation remaining was determined by integration of all
phosphorus signals present in the mixture and calculation of a ratio.
For photographs of the experimental setup, see the Supporting
Information.
1
143.6 (sep, JPF = 709.8 Hz). 13C{1H} NMR (126 MHz, CD2Cl2): δ
2
2
56.5 (d, JPC = 4.3 Hz), 41.1 (d, JPC = 4.6 Hz), 33.2, 27.1, 25.8, 17.4
(d, 3JPC = 6.0 Hz). HRMS (EI) (m/z): calculated [M]+ for C32H64N4P,
535.4869; found, 535.4870.
Stability Investigation in CD3OH. This study was conducted
according to a published report.7b Stock solutions of basic methanol
were prepared by dissolving KOH (1 or 2 M) and 3-(trimethylsilyl)-1-
propanesulfonic acid sodium salt standard (0.025 M) in CD3OH.
Compound 1 (0.05 M in 1 M KOH or 0.03 M in 2 M KOH) was
dissolved in 1 mL of the methanol stock solution and transferred to an
1
NMR tube. The NMR tube was flame-sealed and analyzed by H
NMR spectroscopy for the t = 0 h point. Integration of a selected
signal relative to a signal for the standard provided the initial
concentration of phosphonium salt. The tube was heated in an 80 °C
oil bath, and every 5 days, the tube was removed, cooled to 22 °C, and
1
analyzed by H NMR spectroscopy to determine the percent cation
remaining. A very small signal was observed at 8.6 ppm in some of the
1H NMR spectra, which we believe corresponds to a formate
(confirmed using a spiking experiment). This forms due to some
carbon monoxide from flame-sealing the NMR tubes. For a
representative example calculation and photographs of the exper-
imental setup, see the Supporting Information.
General Procedure for Degradation Experiments in 2 M
KOH/2-(2-Methoxyethoxy)ethanol. A 20 mL scintillation vial with
a septum-sealed cap was charged with a phosphonium salt (0.1 g) and
KOH. The contents were evacuated and back-filled with Ar or N2
seven times. Degassed 2-(2-methoxyethoxy)ethanol was syringed into
the vial (phosphonium concentration 0.05 M, KOH concentration 2
M), and the reaction mixture was placed into an aluminum block at a
predetermined temperature (either 120 or 160 °C). An aliquot was
Example Procedure for Degradation of Compound 5. A 20
mL scintillation vial was charged with [5]Cl (100 mg, 0.28 mmol) and
chlorobenzene (7.0 mL). A 50% w/w NaOH/H2O solution was
prepared by dissolving 3.5 g of NaOH in 3.5 g of deionized water, and
this alkaline solution was placed in the vial (equal grams of 50%
G
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
NaOH solution to milliliters of chlorobenzene). The vial was capped,
and a 50 μL aliquot was removed from the organic layer and then
diluted with 0.60 mL of DMSO-d6 for an initial NMR analysis. The vial
was placed in an aluminum block set to 100 °C, with a stirring rate of
800 rpm. The reaction mixture was stirred for 1 h, and the vial was
removed from the aluminum block and cooled (5 min) before
removing a 50 μL aliquot from the organic layer, diluting with DMSO-
d6, and analyzing using NMR spectroscopy.
Coughlin, E. B. Macromolecules 2016, 49, 4714−4722. (b) Arges, C.
G.; Wang, L.; Ramani, V. RSC Adv. 2014, 4, 61869−61876. (c) Ye, Y.;
Stokes, K. K.; Beyer, F. L.; Elabd, Y. A. J. Membr. Sci. 2013, 443, 93−
99. (d) Arges, C. G.; Parrondo, J.; Johnson, G.; Nadhan, A.; Ramani,
V. J. Mater. Chem. 2012, 22, 3733−3744. (e) Gu, S.; Cai, R.; Yan, Y.
Chem. Commun. 2011, 47, 2856−2858. (f) Arges, C. G.; Kulkarni, S.;
Baranek, A.; Pan, K.-J.; Jung, M.-S.; Patton, D.; Mauritz, K. A.; Ramani,
V. ECS Trans. 2010, 33, 1903−1913. (g) Gu, S.; Cai, R.; Luo, T.;
Jensen, K.; Contreras, C.; Yan, Y. ChemSusChem 2010, 3, 555−558.
(h) Gu, S.; Cai, R.; Luo, T.; Chen, Z.; Sun, M.; Liu, Y.; He, G.; Yan, Y.
Angew. Chem., Int. Ed. 2009, 48, 6499−6502.
(5) For some examples of organometallic-based membranes see:
(a) Gu, S.; Wang, J.; Kaspar, R. B.; Fang, Q.; Zhang, B.; Coughlin, E.
B.; Yan, Y. Sci. Rep. 2015, 5, 11668. (b) Disabb-Miller, M. L.; Zha, Y.;
DeCarlo, A. J.; Pawar, M.; Tew, G. N.; Hickner, M. A. Macromolecules
2013, 46, 9279−9287. (c) Zha, Y.; Disabb-Miller, M. L.; Johnson, Z.
D.; Hickner, M. A.; Tew, G. N. J. Am. Chem. Soc. 2012, 134, 4493−
4496.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.7b00663.
NMR and mass spectra, kinetic data, and sample
calculations (PDF)
All computed molecule Cartesian coordinates (XYZ)
(6) (a) Nunez, S. A.; Capparelli, C.; Hickner, M. A. Chem. Mater.
̃
2016, 28, 2589−2598. (b) Marino, M. G.; Kreuer, K. D.
ChemSusChem 2015, 8, 513−523. (c) Sturgeon, M. R.; Macomber,
C. S.; Engtrakul, C.; Long, H.; Pivovar, B. S. J. Electrochem. Soc. 2015,
162, F366−F372. (d) Long, H.; Pivovar, B. S. ECS Electrochem. Lett.
2015, 4, F13−F16. (e) Long, H.; Kim, K.; Pivovar, B. S. J. Phys. Chem.
C 2012, 116, 9419−9426. (f) Edson, J. B.; Macomber, C. S.; Pivovar,
B. S.; Boncella, J. M. J. Membr. Sci. 2012, 399−400, 49−59.
(g) Chempath, S.; Boncella, J. M.; Pratt, L. R.; Henson, N.; Pivovar,
B. S. J. Phys. Chem. C 2010, 114, 11977−11983. (h) Chempath, S.;
Einsla, B. R.; Pratt, L. R.; Macomber, C. S.; Boncella, J. M.; Rau, J. A.;
Pivovar, B. S. J. Phys. Chem. C 2008, 112, 3179−3182. (i) Einsla, B. R.;
Chempath, S.; Pratt, L. R.; Boncella, J. M.; Rau, J.; Macomber, C.;
Pivovar, B. S. ECS Trans. 2007, 11, 1173−1180.
AUTHOR INFORMATION
Corresponding Author
*E-mail for K.J.T.N.: noonan@andrew.cmu.edu.
ORCID
Geoffrey W. Coates: 0000-0002-3400-2552
Stefan Bernhard: 0000-0002-8033-1453
Kevin J. T. Noonan: 0000-0003-4061-7593
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(7) (a) Wright, A. G.; Weissbach, T.; Holdcroft, S. Angew. Chem., Int.
Ed. 2016, 55, 4818−4821. (b) Hugar, K. M.; Kostalik, H. A., IV;
Coates, G. W. J. Am. Chem. Soc. 2015, 137, 8730−8737. (c) Dong, H.;
Gu, F.; Li, M.; Lin, B.; Si, Z.; Hou, T.; Yan, F.; Lee, S.-T.; Li, Y.
ChemPhysChem 2014, 15, 3006−3014. (d) Long, H.; Pivovar, B. J.
Phys. Chem. C 2014, 118, 9880−9888. (e) Price, S. C.; Williams, K. S.;
Beyer, F. L. ACS Macro Lett. 2014, 3, 160−165. (f) Wang, W.; Wang,
S.; Xie, X.; Lv, Y.; Ramani, V. Int. J. Hydrogen Energy 2014, 39, 14355−
14361. (g) Wang, W.; Wang, S.; Xie, X.; Lv, Y.; Ramani, V. K. J.
Membr. Sci. 2014, 462, 112−118.
K.J.T.N. is grateful to the ACS for a PRF Award (56836-ND7)
and the NSF for a Career Award (CHE-1455136). This work is
part of the Energy Materials Center at Cornell (emc2)
supported by the New York State Energy Research and
Development Authority (NYSERDA) under contract number
38304 and the U.S. Department of Energy, Office of Science,
and Office of Basic Energy Sciences, under Award DE-
SC0001086. The authors are grateful to Dr. Wei You and Dr.
Isaac Mills for help with experiments.
(8) Homer, R. B.; Alwis, K. W. J. Chem. Soc., Perkin Trans. 2 1976,
781−784.
REFERENCES
■
(9) Zhang, B.; Kaspar, R. B.; Gu, S.; Wang, J.; Zhuang, Z.; Yan, Y.
ChemSusChem 2016, 9, 2374−2379.
(1) (a) Tanaka, Y. Ion exchange membranes: fundamentals and
applications, 1st ed.; Elsevier: Amsterdam, Boston, 2007. (b) Kreuer, K.
D. Chem. Mater. 2014, 26, 361−380. (c) Zhang, H.; Shen, P. K. Chem.
Rev. 2012, 112, 2780−2832. (d) Hickner, M. A.; Ghassemi, H.; Kim,
Y. S.; Einsla, B. R.; McGrath, J. E. Chem. Rev. 2004, 104, 4587−4612.
(2) (a) Varcoe, J. R.; Atanassov, P.; Dekel, D. R.; Herring, A. M.;
Hickner, M. A.; Kohl, P. A.; Kucernak, A. R.; Mustain, W. E.;
Nijmeijer, K.; Scott, K.; Xu, T.; Zhuang, L. Energy Environ. Sci. 2014, 7,
3135−3191. (b) Hickner, M. A.; Herring, A. M.; Coughlin, E. B. J.
Polym. Sci., Part B: Polym. Phys. 2013, 51, 1727−1735.
(10) (a) Womble, C. T.; Coates, G. W.; Matyjaszewski, K.; Noonan,
K. J. T. ACS Macro Lett. 2016, 5, 253−257. (b) Noonan, K. J. T.;
Hugar, K. M.; Kostalik, H. A., IV; Lobkovsky, E. B.; Abruna, H. D.;
̃
Coates, G. W. J. Am. Chem. Soc. 2012, 134, 18161−18164.
(11) (a) Yonemoto-Kobayashi, M.; Inamoto, K.; Kondo, Y. Chem.
Lett. 2014, 43, 477−479. (b) Boileau, S.; Illy, N. Prog. Polym. Sci. 2011,
36, 1132−1151. (c) Mathiessen, B.; Jensen, A. T. I.; Zhuravlev, F.
Chem. - Eur. J. 2011, 17, 7796−7805. (d) Lemaire, C. F.; Aerts, J. J.;
Voccia, S.; Libert, L. C.; Mercier, F.; Goblet, D.; Plenevaux, A. R.;
Luxen, A. J. Angew. Chem., Int. Ed. 2010, 49, 3161−3164. (e) Kim, M.-
Y.; Seo, G.; Kwon, O. Z.; Chang, D. R. Chem. Commun. 2009, 3110−
3112. (f) Sakakura, A.; Katsukawa, M.; Hayashi, T.; Ishihara, K. Green
Chem. 2007, 9, 1166−1169. (g) Schwesinger, R.; Link, R.; Wenzl, P.;
Kossek, S. Chem. - Eur. J. 2006, 12, 438−445. (h) Grzelka, A.;
Chojnowski, J.; Fortuniak, W.; Taylor, R. G.; Hupfield, P. C. J. Inorg.
Organomet. Polym. 2004, 14, 85−99. (i) Kim, K.-S.; Kim, J.-H.; Seo, G.
(3) For some examples of imidazolium-based membranes see:
(a) Wright, A. G.; Holdcroft, S. ACS Macro Lett. 2014, 3, 444−447.
(b) Lin, X.; Varcoe, J. R.; Poynton, S. D.; Liang, X.; Ong, A. L.; Ran, J.;
Li, Y.; Xu, T. J. Mater. Chem. A 2013, 1, 7262−7269. (c) Lin, X.; Liang,
X.; Poynton, S. D.; Varcoe, J. R.; Ong, A. L.; Ran, J.; Li, Y.; Li, Q.; Xu,
T. J. Membr. Sci. 2013, 443, 193−200. (d) Thomas, O. D.; Soo, K. J.
W. Y.; Peckham, T. J.; Kulkarni, M. P.; Holdcroft, S. J. Am. Chem. Soc.
2012, 134, 10753−10756. (e) Ran, J.; Wu, L.; Varcoe, J. R.; Ong, A.
L.; Poynton, S. D.; Xu, T. J. Membr. Sci. 2012, 415−416, 242−249.
(f) Thomas, O. D.; Soo, K. J. W. Y.; Peckham, T. J.; Kulkarni, M. P.;
Holdcroft, S. Polym. Chem. 2011, 2, 1641−1643.
Chem. Commun. 2003, 372−373. (j) Rexin, O.; Mulhaupt, R.
̈
Macromol. Chem. Phys. 2003, 204, 1102−1109. (k) Rexin, O.;
Mulhaupt, R. J. Polym. Sci., Part A: Polym. Chem. 2002, 40, 864−
̈
873. (l) Grzelka, A.; Chojnowski, J.; Cypryk, M.; Fortuniak, W.;
(4) For some examples of phosphonium-based membranes see:
(a) Zhang, W.; Liu, Y.; Jackson, A. C.; Savage, A. M.; Ertem, S. P.;
Tsai, T.-H.; Seifert, S.; Beyer, F. L.; Liberatore, M. W.; Herring, A. M.;
Hupfield, P. C.; Taylor, R. G. J. Organomet. Chem. 2002, 660, 14−26.
(m) Baskaran, D.; Muller, A. H. E. Macromol. Rapid Commun. 2000,
̈
H
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
21, 390−395. (n) Hupfield, P. C.; Taylor, R. G. J. Inorg. Organomet.
Polym. 1999, 9, 17−34.
(12) Superbases for Organic Synthesis: Guanidines, Amidines,
Phosphazenes and Related Organocatalysts; Ishikawa, T., Ed.; Wiley:
Chichester, West Sussex, United Kingdom, 2009.
(13) (a) Marchenko, A. P.; Koidan, G. N.; Pinchuk, A. M. J. Gen.
Chem. (U.S.S.R.) 1984, 54, 2405−2409. (b) Marchenko, A. P.; Koidan,
G. N.; Pinchuk, A. M. Zh. Obshch. Khim. 1984, 54, 2691−2696.
(14) Schwesinger, R.; Link, R.; Wenzl, P.; Kossek, S.; Keller, M.
Chem. - Eur. J. 2006, 12, 429−437.
(15) (a) Dawber, J. G.; Skerratt, R. G.; Tebby, J. C.; Waite, A. A. C.
Phosphorus, Sulfur Silicon Relat. Elem. 2012, 187, 1261−1268.
(b) Dawber, J. G.; Tebby, J. C.; Waite, A. A. C. J. Chem. Soc., Perkin
Trans. 2 1983, 1923−1925. (c) Hays, H. R.; Laughlin, R. G. J. Org.
Chem. 1967, 32, 1060−1063. (d) McEwen, W. E.; Axelrad, G.; Zanger,
M.; VanderWerf, C. A. J. Am. Chem. Soc. 1965, 87, 3948−3952.
(e) Aksnes, G.; Brudvik, L. J. Acta Chem. Scand. 1963, 17, 1616−1622.
(f) Aksnes, G.; Songstad, J. Acta Chem. Scand. 1962, 16, 1426−1432.
(g) Hoffmann, H. Justus Liebigs Ann. Chem. 1960, 634, 1−8.
(h) Zanger, M.; VanderWerf, C. A.; McEwen, W. E. J. Am. Chem.
Soc. 1959, 81, 3806−3807. (i) Fenton, G. W.; Ingold, C. K. J. Chem.
Soc. 1929, 2342−2357.
(16) Schiemenz, B.; Wessel, T.; Pfirmann, R.; Beck, A.; Hahn, W. US
6,407,029B1, Jun 18, 2002.
(17) (a) Koidan, G. N.; Marchenko, A. P.; Kudryavtsev, A. A.;
Pinchuk, A. M. J. Gen. Chem. (U.S.S.R.) 1982, 52, 1779−1790.
(b) Koidan, G. N.; Marchenko, A. P.; Kudryavtsev, A. A.; Pinchuk, A.
M. Zh. Obshch. Khim. 1982, 52, 2001−2011.
(18) Cameron, T. B.; Hanson, H. N.; de la Fuente, G. F.; Huheey, J.
E. Phosphorus, Sulfur Silicon Relat. Elem. 1993, 78, 37−46.
(19) (a) Eyles, C. T.; Trippett, S. J. Chem. Soc. C 1966, 67−71.
(b) Trippett, S. Proc. Chem. Soc. 1963, 19−20.
(20) Montanari, F.; Landini, D.; Rolla, F. Top. Curr. Chem. 1982, 101,
147−200.
(21) Landini, D.; Maia, A.; Rampoldi, A. J. Org. Chem. 1986, 51,
3187−3191.
(22) (a) Dehmlow, E. V.; Slopianka, M.; Heider, J. Tetrahedron Lett.
1977, 18, 2361−2364. (b) Cristau, H. J.; Long, A.; Christol, H.
Tetrahedron Lett. 1979, 20, 349−352. (c) Landini, D.; Maia, A. J.
Chem. Soc., Chem. Commun. 1984, 1041−1042.
(23) Landini, D.; Maia, A.; Rampoldi, A. J. Org. Chem. 1986, 51,
5475−5476.
(24) (a) Caldin, E. F.; Long, G. J. Chem. Soc. 1954, 3737−3742.
(b) Caldin, E. F.; Long, G. Nature 1953, 172, 583−584.
́
(25) Karlstrom, A.; Unden, A. Int. J. Pept. Protein Res. 1996, 48, 305−
̈
311.
(26) Harris, R. K.; Becker, E. D.; De Menezes, S. M. C.; Goodfellow,
R.; Granger, P. Pure Appl. Chem. 2001, 73, 1795−1818.
I
DOI: 10.1021/acs.organomet.7b00663
Organometallics XXXX, XXX, XXX−XXX