Pairing heterocyclic cations with closo-icosahedral borane and carborane anions, II: Benchtop alternative synthetic methodologies for binary triazolium and tetrazolium salts with significant water solubility

Article abstract of DOI:10.1080/00397911.2010.520543

Two efficient processes for the synthesis of 12 relatively water-soluble binary triazolium and the first tetrazolium borane [B12H12] and carborane [CB₁₁H₁₂] salts by a one-step, open-air metathesis reaction have been developed. First, a combination of exhaustive trituration of the two solid reactant salts with refluxing anhydrous acetonitrile followed by flash filtration through a plug of silica gel afforded excellent recovery for a broad series of otherwise water-soluble heterocyclium salts. Second, an alternative aqueous metathesis, driven to completion by precipitation of silver halides, followed by removal of water, redissolution in acetonitrile, and filtration through a silica-gel plug, also yielded such heterocyclium borane and carborane salts. Mixed 1:1 dication heterocyclium borane salts were first synthesized using this second procedure, and one example showed melting-point depression behavior. Taylor & Francis Group, LLC.

Full text of DOI:10.1080/00397911.2010.520543

This article was downloaded by: [Columbia University]
On: 26 September 2014, At: 09:33
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Synthetic Communications: An
International Journal for Rapid
Communication of Synthetic Organic
Chemistry
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/lsyc20
Pairing Heterocyclic Cations with closo-
Icosahedral Borane and Carborane
Anions, II: Benchtop Alternative
Synthetic Methodologies for Binary
Triazolium and Tetrazolium Salts with
Significant Water Solubility
John L. Belletire a , Stefan Schneider b , Brett A. Wight a , Steven L.
Strauss c & Scott A. Shackelford b
a ERC Incorporated, Air Force Research Laboratory , Edwards AFB ,
California , USA
b Space and Missile Propulsion Directorate, AFRL/RZSP , Edwards
AFB , California , USA
c Department of Chemistry , Colorado State University , Fort
Collins , Colorado , USA
Accepted author version posted online: 27 Jun 2011.Published
online: 14 Sep 2011.
To cite this article: John L. Belletire , Stefan Schneider , Brett A. Wight , Steven L. Strauss & Scott
A. Shackelford (2012) Pairing Heterocyclic Cations with closo-Icosahedral Borane and Carborane
Anions, II: Benchtop Alternative Synthetic Methodologies for Binary Triazolium and Tetrazolium Salts
with Significant Water Solubility, Synthetic Communications: An International Journal for Rapid
Communication of Synthetic Organic Chemistry, 42:2, 155-169, DOI: 10.1080/00397911.2010.520543
To link to this article: http://dx.doi.org/10.1080/00397911.2010.520543
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,

and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions

Synthetic Communications1, 42: 155–169, 2012
Copyright # Taylor & Francis Group, LLC
ISSN: 0039-7911 print=1532-2432 online
DOI: 10.1080/00397911.2010.520543
PAIRING HETEROCYCLIC CATIONS WITH CLOSO-
ICOSAHEDRAL BORANE AND CARBORANE ANIONS,
II: BENCHTOP ALTERNATIVE SYNTHETIC
METHODOLOGIES FOR BINARY TRIAZOLIUM
AND TETRAZOLIUM SALTS WITH SIGNIFICANT
WATER SOLUBILITY
John L. Belletire,1 Stefan Schneider,2 Brett A. Wight,1
Steven L. Strauss,3 and Scott A. Shackelford2
1ERC Incorporated, Air Force Research Laboratory, Edwards AFB,
California, USA
2Space and Missile Propulsion Directorate, AFRL=RZSP, Edwards AFB,
California, USA
3Department of Chemistry, Colorado State University, Fort Collins,
Colorado, USA
GRAPHICAL ABSTRACT
Abstract Two efficient processes for the synthesis of 12 relatively water-soluble binary
triazolium and the first tetrazolium borane [B12H12] and carborane [CB11H12] salts by
a one-step, open-air metathesis reaction have been developed. First, a combination of
exhaustive trituration of the two solid reactant salts with refluxing anhydrous acetonitrile
followed by flash filtration through a plug of silica gel afforded excellent recovery for a
broad series of otherwise water-soluble heterocyclium salts. Second, an alternative aqueous
metathesis, driven to completion by precipitation of silver halides, followed by removal of
water, redissolution in acetonitrile, and filtration through a silica-gel plug, also yielded such
heterocyclium borane and carborane salts. Mixed 1:1 dication heterocyclium borane salts
were first synthesized using this second procedure, and one example showed melting-point
depression behavior.
Keywords closo-Icosahedral borane dianions; closo-icosahedral carborane anions;
exhaustive trituration; heterocyclium halides; mixed heterocyclium boranes; silver-halide-
mediated metathesis
Received June 3, 2010.
Address correspondence to John L. Belletire or Scott A. Shackelford, AFRL=RZSP, 10 East Saturn
Blvd., Edwards AFB, CA 93524-7680. E-mail: jbelletire@roadrunner.com; scottshack@roadrunner.com
155

156
J. L. BELLETIRE ET AL.
INTRODUCTION
In a previous publication,[1] a group of new water-insoluble binary heterocyclium
(imidazolium and triazolium) closo-icosahedral borane [B12H12] and carborane
[CB11H12] salts were synthesized with a one-step aqueous metathesis reaction in excel-
lent yields. These unique salts were generated using an open-air benchtop procedure
conducted by combining stoichiometric quantities of potassium borane or potassium
carborane with a variety of heterocyclium chloride, bromide, or iodide salts. Mixing
the two metathesis component salts in hot deionized (DI) water, followed by slow cool-
ing in a refrigerator, produced these water-insoluble salt products and permitted trivial
product isolation and efficient purification. These resulting binary heterocyclium borane
and carborane salts were verified by NMR, infrared (IR), high-resolution m=z, and=or
single-crystal x-ray analyses and were chemically free of any significant contamination
by their associated by-product potassium halide precursor counterions. This latter issue
was verified by direct ion chromatography analysis for the [Cl]– titer.
Other new binary heterocyclium (triazolium and tetrazolium) closo-icosahedral
borane and carborane metathesis salts, however, possessed high solubility in water,
and innovative alternative synthesis metathesis methodologies were needed for their
preparation. These new synthetic procedures were developed as described in this
article. Several other publications[2–4] also have been reported that used closed-
system nonaqueous preparations of various binary solid salts containing only the
closo-icosahedral carborane anion [CB11H12]– paired with a moderate selection of
imidazolium cations and one pyridinium cation example.
RESULTS AND DISCUSSION
We sought a further generalization of this open-air aqueous benchtop metath-
esis methodology[1] for synthesizing water-soluble salts where additional triazolium
cations and a new tetrazolium cation could be paired with the closo-borane and
closo-carborane anions (Scheme 1). In pursuing this endeavor, at least two serious
Scheme 1. Typical reactions of trituration-based metatheses.

WATER-SOLUBLE HETEROCYCLIUM BORANE-BASED SALTS
157
complications had to be solved. First, when attempting to perform a simple
metathesis by combining 2.0 equivalents of [4-amino-1-H-1,2,4-triazolium][Cl] with
1.0 equivalent of K2[B12H12] in warm DI water, instead of the desired adduct, a
beautiful highly faceted crystalline solid deposited (Scheme 2), which consisted of
the pure [4-amino-1-H-1,2,4-triazolium][K][B12H12] (1) salt product. The identity
of this adduct was unambiguously ascertained by proton NMR integration and
single-crystal x-ray crystallography (Fig. 1). This result likely occurred from a far
lower solubility of the mixed mono-K=triazolium borane salt (1) compared to either
of the starting materials or of the desired metathesis salt product (2) seen in Fig. 2.
The mono-K salt (1) insolubility precluded a direct water-driven metathesis reaction
and necessitated developing an alternative methodology. Second, expansion of aque-
ous metathesis combinations that pair these borane and carborane anions beyond
previous imidazolium and triazolium examples[1] gave a variety of highly water-
soluble salts. For example, the metathesis reaction of [5-amino-x-H-tetrazolium][Cl]
with either the borane, K2[B12H12], or the carborane, K[CB11H12], reactants yielded
hydrated crystals of the desired product salts 4 and 5 as illustrated in Figs. 2 and 3,
respectively. However, because of their high water solubilities, only very poor yields
of the metathesis targets were recovered. Even highly concentrated aqueous solu-
tions afforded poor product yields. Thus, direct water-driven metathesis reactions
proved to be a highly inefficient process, especially given the expense of the com-
mercial potassium borane or potassium carborane precursors. Development of an
alternative nonaqueous methodology with a greater generality for water-soluble
metathesis salt products became the important challenge addressed in this study.
Acetonitrile Trituration (Procedure A)
We investigated the combined use of exhaustive triturations with acetonitrile,
followed by a flash filtration through silica gel (Scheme 1). This method exploited
the significant solubility of many heterocyclium borane and carborane salts in boiling
anhydrous acetonitrile versus the relative insolubility of the corresponding KCl
by-product. It also required determining the amount of KX by-product that a given
dry volume of silica gel might have retained using a reasonable volume of eluting
solvent. For KCl, 0.4473 g (6.0 mmol) was partially solubilized in 100 mL of warm
acetonitrile. After cooling to room temperature, the resulting turbid liquid was passed
through 40 mL (dry volume) of silica gel, overlaid with 20 mL (dry volume) of sea
sand, and eluted with another 100 mL of acetonitrile. Less than 1 mg of KCl residue
was collected after solvent removal.
Scheme 2. Formation of the mixed mono-K=triazolium borane salt.

158
J. L. BELLETIRE ET AL.
Figure 1. Graphical and molecular drawing of the mixed [4-amino-1,2,4-triazolium][potassium][B12H12
salt 1 with thermal ellipsoids shown at 50% probability level. (Figure is provided in color online.)
]
In this metathesis reaction process, once both solid salt reactants were mixed,
our acetonitrile trituration=flash filtration approach was followed by careful removal
of volatiles to give a series of finely divided amorphous powders. While many prelimi-
nary experiments were performed to seek a general recrystallization solvent mixture
for these powders, and while some individual successes resulted, a general efficient
solvent system for all salt product recrystallizations could not be found. So, rather
than recrystallizing each individual adduct, spectroscopic examination of the
amorphous solids was performed. This allowed a direct comparison for the effective-
ness of the trituration=filtration procedures across a group of reactants. These
isolated crude powders were indeed spectroscopically pure and had a moderately
low content of [Cl]– (Table 1), in fair agreement with the blank experiments done
on a bolus of KCl suspended in acetonitrile. While a couple of values were greater
than expected (Table 1, salts 4 and 5), varying the number of triturations, trituration
volume, or a silica-gel-plug rechromatography might have reduced these halide
concentrations. The [Cl]– measurement procedure used can be viewed in the
Experimental section.
The use of exhaustive trituration=plug filtration worked for a considerable
variety of heterocyclium halide starting materials. Replicate experiments with reac-
tions involving 2–6 mmol of the starting heterocyclium halide were examined with
no discernable variation in yield. For each preparative run, we performed a total
of 10 separate triturations and then pooled the triturant samples in a flask, followed
Figure 2. Synthesized binary heterocyclium borane salts.

WATER-SOLUBLE HETEROCYCLIUM BORANE-BASED SALTS
159
Figure 3. Synthesized binary heterocyclium carborane salts.
by cooling to room temperature prior to the flash filtration. For the trituration
process to work at maximum efficiency and to minimize elution of the by-product
KCl, using anhydrous acetonitrile was essential for successful flash filtration. High-
performance liquid chromatography (HPLC)–grade acetonitrile from Aldrich con-
veniently was employed without any further purification by carefully wrapping the
screw-cap-sealed bottle top with Parafilm after initially opening a new vessel. For
this investigation, commercial silica gel of moderate sieve size (70–230 mesh; Merck)
was optimal. Semiproportional scaling of the required silica gel from 22 mL (dry
volume) to 40 mL (dry volume) and the additional eluting solvent volume from
80 mL to 120 mL were found to be satisfactory as the scale of reaction was increased
from 2 to 6 mmol, respectively. After evaporation of the trace acetonitrile remaining
in the stirbar=recovery flask apparatus, a proton NMR (dimethylsulfoxide,
DMSO-d6, solvent) performed on the residue indicated essentially complete removal
of the starting metathesis reactant salts. When generalized to additional heterocy-
clium chloride examples beyond this article, such a trial NMR would indicate
whether any further triturations were needed. This result could then be easily incor-
porated as feedback into a specific modified procedure. Heterocyclium iodides
showed significant solubility in acetonitrile, so there could be instances where a
greater amount of silica gel would be needed to remove all the KI. Our experimental
efforts with heterocyclium iodides, however, focused on procedure B (discussed
later) where silver nitrate was employed to drive an aqueous metathesis reaction
to completion by silver halide by-product removal.
Eliminating the final traces of acetonitrile from the metathesis salt products
initially proved to be difficult. The sluggishness in removal of the acetonitrile for
Table 1. Percentage of residual chloride ion content in synthesized salts 2–11
Compound
Parameter
2
3
4
5
6
7
8
9
10
11
Chloride by weight (%)
Detection limit (%)
0.47
0.02
0.19
0.02
3.06
0.01
2.41
0.01
0.45
0.02
0.71
0.01
<udla
0.01
0.30
0.02
<udla
0.02
0.06
0.02
aBelow detection limit.

160
J. L. BELLETIRE ET AL.
many examples of borane and carborane salts suggested some sort of significant
borane-based solvent interaction. A simple azeotrope-type process was developed
where a few milliliters of deionized water were added to the sample followed by
removal of all volatiles at high vacuum (final P ꢀ 40–50 mTorr). Occasional modest
heating of the flask with a heat gun, while under vacuum azeotrope conditions and
after water addition, was repeated three times. This process effectively removed all
the residual acetonitrile. Incorporating this final solvent ‘‘distillation’’ into the over-
all process gave isolated yields of spectroscopically pure anhydrous metathesis
product salts (2–11) as finely divided amorphous solids in overall yields that ranged
from 85% to 98% (Figs. 2 and 3). For completeness, moderately water-insoluble
examples from our earlier publication,[1] [4-amino-1-H-1,2,4-triazolium][closo-CB11
H12] (3) and [1-amino-3-H-1,2,3-triazolium]2[closo-B12H12] (11), were examined by
procedure A. The isolated flash-filtered products were obtained in good yields and
with spectroscopic data identical to those previously reported.[1]
There was no significant difference in the results whether one employed either
the potassium borane or potassium carborane as starting materials for this acetoni-
trile trituration metathesis procedure. The starting boron-containing materials
were purchased from Katchem in the Czech Republic and used without any further
purification.[1]
Silver-Ion-Driven Metathesis (Procedure B)
A second method for driving the metatheses reactions to completion used a
stoichiometric amount of aqueous silver nitrate to precipitate the halide gegenions
as AgX. This process produced several useful advantages and one possible disadvan-
tage. The procedure was particularly valuable in driving this metathesis reaction to
completion by the extreme insolubility of the AgX by-product. Furthermore, as was
previously suggested for borane salts,[1] this process conveniently permitted the first
intentional example for synthesizing potential ‘‘ternary’’ꢁ mixed diheterocyclium
borane salts, where two different singly charged heterocyclium cations were paired
with the one borane dianion (Scheme 3). This was possible by reacting 1 mol each
of two different heterocyclium halides with 1 mol of the potassium borane and using
2 equivalents of silver nitrate in water solvent to yield mixed dication heterocyclium
borane salts like 12–14 seen in Fig. 4. The immediate silver halide precipitate was
filtered through a bed of moist Celite, affording a colorless aqueous filtrate that
contained the desired product salt as well as 2 equivalents of KNO3. Removal of
the water via a high vacuum line gave a fluffy white powder. Dissolution of a small
sample in DMSO-d6 afforded a crude proton NMR spectrum that conformed
ꢁThe term binary fits the carborane salts that stoichiometrically contain one heterocyclium cation
paired with the carborane anion. For discussions in this report and others,[1,6] the term binary also is
extended to the borane salts that contain the same two heterocyclium cations paired with the borane
dianion. The term ternary is reserved for the mixed borane salts where two different heterocyclium
cations are stoichiometrically paired in a 1:1 ratio with the borane dianion. Similarly, salt 12 with a
heterocyclium [Het]þ cation and potassium [K]þ cation stoichiometrically paired with the same borane
dianion is considered to be a ternary salt.

WATER-SOLUBLE HETEROCYCLIUM BORANE-BASED SALTS
161
Scheme 3. Silver-ion-driven metathesis and mixed borane salt capability.
exactly to the expected relative correct integrations of all key peaks.y The initial mass
balance was near theory when the presence of 2 equivalents of KNO3 were included.
Upon filtration of a resuspension of the crude solid in acetonitrile through a plug of
silica gel, overlaid by sand, followed by removal of all volatiles (again using the thrice
‘‘water azeotrope treatment’’), an amorphous white solid was isolated that generally
conformed to the expected proton and carbon NMR. To ensure the KNO3
by-product was removed on the silica-gel plug, a blank run using 0.6066 g (6.0 mmol)
of KNO3 was partially solubilized in 100 mL of warm acetonitrile. After cooling to
room temperature, the resulting suspension was passed through 40 mL (dry volume)
of silica gel, overlaid with 20 mL (dry volume) of sea sand, and eluted further with
100 mL of acetonitrile. Solvent removal afforded less than 1 mg of recovered residue.
In experiments to synthesize a mixed heterocyclium borane salt (12) containing
two different mono-cations (Fig. 4), we initially were highly gratified that, after the
plug filtration, this approach indeed gave the two heterocyclium cations in a precise
1:1 ratio based on proton NMR integrations of their distinctive peaks. However,
when comparing the integration of the very sharp N-methyl singlets (3H) to the
borane B-H multiplet (12H), the borane integration was always slightly (ca. 5%)
higher than expected. Similar results appeared for the other mixed heterocyclium
boranes (15 and 16) displayed in Fig. 4. We were not certain what caused this slight
yThe synthetic method used, the significantly greater than 50% yields, and the proton NMR
integrations are consistent with the formation of mixed salts 12–14. However, it is possible the same
result would be observed if the isolated solids contained either an exact one-third mixture of the two
pure salts and the mixed salt or of some mixture containing an equal percentage of the two pure salts
with the mixed salt constituting the remaining 100% sample. In any case, two different heterocyclic
mono-cations are in some arrangement of intimate contact with one borane dianion, which provided at
least one interesting property, eutectic melting-point depression. Recrystallization of ternary salts 12–14
was not attempted because of a concern that a cation redistribution might result.

162
J. L. BELLETIRE ET AL.
Figure 4. Synthesized mixed diheterocyclium borane salts.
anomaly but hypothesized that it probably resulted from overloading the adsorbent
active sites on the silica gel with both the metathesis components as well as with the
KNO3 by-product. Attempts to modify the elution by taking smaller numbers of
column volumes of elutant basically lowered the overall yield, but this anomalous
shortfall in the ratio of heterocyclium peak integrations versus the borane dianion
B-H integration was not improved. Nevertheless, within the caveat of the 5% inte-
gration anomaly, moderately pure crude solids of various mixed dication heterocy-
clium salts with the desired borane dianion were obtained. These results simply are
noted and accepted as an inherent limitation of procedure B.
The possible synthesis of ternary mixed diheterocyclium borane salts, which
pair two different singly charged heterocyclium cations with the borane dianion
(Fig. 4), raised an interesting question regarding eutectic melting-point behavior.
While it is well known that a physical admixture of two neutral compounds often
affords eutectic behavior, resulting in a lower melting point, a recent literature article
revealed similar behavior with heterocyclium salts.[5] When three solid salts were
physically mixed in equimolar ratios, the admixture gave a lower intersalt eutectic
melting point that resulted in a room-temperature melt. This led to the supposition
that a ternary mixed diheterocyclium borane salt, such as 12 with its two dissimilar
heterocyclium cations, might provide a lower melting point via an intrasalt eutectic
behavior than the melting points exhibited by the two related binary salts 15[1] and
16.[1] Binary salts 15 and 16 each paired two identical heterocyclium cations with a
borane dianion, where the two cations of one salt (e.g., 15 or 16) were the same as
one each of the two dissimilar dications in the mixed diheterocyclium borane salt
(12). A melting-point depression would also be expected, relative to each pure salt,
when the two individual binary salts, 15 and 16, were physically mixed via an intersalt
eutectic behavior.[5] Thus, similar intrasalt (12) and intersalt (15 and 16) melting-point
behaviors might be expected, and indeed, this was the case.
Preparation of the ternary mixed diheterocyclium borane salt (12) by the silver
nitrate process (procedure B) afforded a white solid with a crude, unrecrystallized mp
of 136–142 ꢂC. Any attempt at recrystallizing the 1:1 mixture of cations in 12 could
have afford an enrichment in one or the other of the heterocyclium cations. The flash
filtration, by virtue of eluting the two cations and the dianion in electrostatic

WATER-SOLUBLE HETEROCYCLIUM BORANE-BASED SALTS
163
Figure 5. Comparison of intersalt and intrasalt melting-point depressions.
neutrality, preserves the 1:1 ratio to within experimental error of the proton NMR
integration. When the two recrystallized individual binary salts having mp of
197–199 ꢂC for 15 and 178–179 ꢂC for 16 were physically mixed in a 1:1 ratio by
weight, a mp of 142–148 ꢂC was obtained for this admixture (Fig. 5). Subsequent
improved melting points of pure salts 15 and 16 were obtained from that originally
determined for Ref. 1. For completeness, a 2:1 ratio physical admixture of salts 15
and 16 gave a mp of 128–132 ꢂC, and a 1:2 ratio displayed a mp of 132–135 ꢂC. Thus,
either a physical mixture of individual binary heterocyclium salts 15 and 16, or a
synthesized ternary mixed diheterocyclium borane salt (12), respectively, provided
a similar intersalt or intrasalt eutectic melting-point lowering (Fig. 5).
EXPERIMENTAL
Caution! While no special precautions were taken in synthesizing and handling
these borane-based salts, they were shown to be high-energy materials and could be
thermally,[6] impact, friction, and=or spark (ESD) initiated to rapid energy-releasing
phenomena. Hazards testing was conducted on salts 2, 4, 5, and 11. Impact initiation
testing determined the 100% value where no initiation occurs, and salt 11 was the
most sensitive at 84 Kg-cm, while salt 5 was the least sensitive at 250 Kg-cm. For
comparison, the following more sensitive impact Kg-cm values where 50% sample
initiation results for known explosives were CL-20 at 33, PETN at 67, HMX at
115, and RDX at 117. Salt 2 was most sensitive to friction initiation at 14.4 Kg-cm,
while salts 4 and 5 were the least friction sensitive at 21.6 Kg-cm. Salt 2 also was the
most sensitive to spark initiation (ESD test) at 0.0025 J, and salt 5 was the least ESD
sensitive at 1.00 J.
All neutral heterocycles that were converted to the hydrochloride (aqueous con-
centrated HCl followed by recrystallization from EtOH=Et2O) or to the methiodide
(CH3I followed by recrystallization from EtOH=Et2O) were purchased commercially
with one exception. The 1-amino-1,2,3-triazole was synthesized as described in the
literature.[7] DI water was obtained from an in-house Millipore MILL-Q reagent-
grade water system at an 18-megaohm-cm purity level. All organic solvents were

164
J. L. BELLETIRE ET AL.
commercially purchased and were either reagent grade or HPLC purity. Most impor-
tant, the acetonitrile used in the triturations and flash filtrations was HPLC grade,
and the solvent container was carefully re-Parafilmed after each use to minimize
any accumulated moisture.
A Bruker Avance 400 digital NMR instrument was used to obtain both proton
¨
(1H) and carbon (13C) spectra. Because this project is primarily one involving process
chemistry, we used commercial samples of Katchem K2[B12H12] that contained a
small amount of [NHEt3]2[B12H12] salt as an impurity. By direct proton integration
against the main component, this impurity was approximately 1.5%. One component
of this impurity peak overlapped with the B-H multiplet, but this integration contri-
bution was negated (by simple subtraction) from the (B-H) integrations for each new
metathesized salt. However, this [NHEt3]2[B12H12] ‘‘impurity’’ provided a constant
marker for our reaction mixtures because the product salts were not recrystallized.
The sole exception to this for the salts (1–14) was (1) which, when recrystallized in
water solvent, was completely devoid of the [NHEt3]2[B12H12] impurity.
Fourier transform–infrared spectra (FTIR) were taken on powdered samples
using a Nicolet 6700 spectrometer in air with a horizontal attenuated total reflec-
tance (HATR) optical system.
Commercial Opti-Melt apparatus was used with video playback software.
For typical acetonitrile-trituration runs, the isolated amorphous solids were not
subjected to melting-point determination.
The [Cl]– ion concentration was determined by ion chromatography using a
Waters HPLC equipped with a Waters 432 conductivity detector and a Phenomenex
Star-Ion A300 100 ꢃ 4.3 mm ID (PEEK) column. A borate=gluconate eluent was
used, and system conditions were set according to Waters method 980895. Each salt
sample was weighed between 0.01 mg to 0.05 mg in a plastic class B centrifuge tube
and diluted to 25 mL using type I ultrapure water. Samples that did not readily
dissolve were heated to 80 ꢂC with a plastic centrifuge cap. Each sample was passed
through an IC Millex LG 0.2-mm syringe filter prior to injection. A three-point
calibration curve was generated using blank, 1-ppm, and 10-ppm NIST traceable
chloride standard. Concentrations were determined by comparing peak area
response of the samples to the standard calibration curve.
Procedure A Synthesis: General Procedure
After weighing out the corresponding stoichiometric quantities of the heterocy-
clium halide salt and the appropriate potassium carborane or potassium borane salt,
the solids were placed in a recovery flask equipped with a stirbar (flask and stirbar are
tarred together) and a Vigreaux reflux tube. Addition of pure acetonitrile (HPLC
grade only) was followed by placing the apparatus in a hot oil bath (internal
T ¼ 161–165 ꢂC). After the contents of the flask reached a vigorous reflux, the suspen-
sion was then cooled to near room temperature, and the supernatant was carefully
transferred to an Erlenmeyer flask. Exhaustive trituration was repeated 10 times,
using an empirically derived volume of acetonitrile (10 mL for a 6-mmol scale and
6 mL for a 2 to 3-mmol scale of carborane or borane), and resulted in complete
removal of the reactive species while most of the by-product potassium halide
remained in the original recovery flask. After cooling to room temperature, the

WATER-SOLUBLE HETEROCYCLIUM BORANE-BASED SALTS
165
combined trituration solutions in the Erlenmeyer flask were poured over an
acetonitrile-moistened bed of 35 mL (dry volume) of silica gel and 8 mL (dry volume)
of washed sea sand packed in a 125-mL coarse frit glass filter. An additional 100 mL
of acetonitrile was passed through the adsorbent bed. The combined solvent was then
rotovapped to dryness. To remove the last traces of acetonitrile, addition of 2 mL of
DI water was followed by pumping at high vacuum, with periodic heat gun treatment,
thereby affording a solid residue. This ‘‘semi-azeotrope’’ treatment was repeated a
total of three times. This afforded a clean white amorphous powder whose proton
NMR integration was excellent, whose [halide] weight percentage was generally small
(Table 1), and whose yields were acceptable. Individual samples were not recrystal-
lized because each product required considerable experimentation to find a suitable
single or mixed optimal solvent combination for recrystallization.
Procedure B Synthesis
This procedure has been adapted for a mixed cation system and was based on
the reaction used to prepare the mixed diheterocyclium borane, [4-amino-1-Me-
1,2,4-triazolium][1-amino-3-Me-1,2,3-triazolium][closo-B12H12] (12). The crystalline
iodide [4-amino-1-Me-1,2,4-triazolium][I] (0.6781 g; 3.00 mmol), the crystalline iodide
[1-amino-3-Me-1,2,3-triazolium][I] (0.6781 g; 3.00 mmol), the borane K2[B12H12]
(0.6601 g; 3.00 mmol), and the silver nitrate (1.0192 g; 6.00 mmol) were carefully
weighed out separately, and each was transferred separately to four different
50-mL beakers. To each of the beakers were added 4 mL of DI water, and the contents
were warmed on a hot plate. Slowly, in the previous order, the contents of the first
three beakers were transferred to the fourth beaker, and each of the first three beakers
was rinsed with 2.0 mL of DI water. The resulting combined grayish suspension was
stirred with a glass rod and then filtered through a pre-prepared moist bed of Celite.
The beakers were rinsed with a final 5 mL of DI water. The filtrate was placed in a
tared recovery flask and pumped for 40 h on a high vacuum line (final P ꢀ 35 mTorr).
The contents of the recovery flask were swirled with 70 mL of hot anhydrous
acetonitrile, cooled to RT, and poured through a 125 mL (coarse frit) glass funnel
with a pre-equilibrated (acetonitrile) bed of silica gel (38 mL dry volume) and overlain
with sea sand (10 mL dry volume). Elution with a total of 200 mL of anhydrous
acetonitrile was followed by removal of solvent (rotovapped and high vacuum line).
To the crude residue was added 3 mL of DI water, and the contents were heated
repeatedly with a heat gun, while the volatiles were slowly removed under high
vacuum. This process was repeated a total of three times to afford an off-white
amorphous solid (0.9253 g of spectroscopically pure residue. See this ternary salt
(12) below for spectroscopic details.
Selected Data
[4-Amino-1-H-1,2,4-triazolium][K][closo-B12H12]
(1). From
2.410 g
(20.0 mmol) of [4-amino-1-H-1,2,4-triazolium][Cl] and 2.2003 g (10.0 mmol) of
K2[B12H12] in 11.0 mL DI water held at reflux for 10 min and cooled to 4 ꢂC overnight,
there was isolated 1.3754 g (51.7%) of massive distorted rhombs [mp 207.2 ꢂC (soften-
ing), 252.3–254.9 ꢂC liquification with rapid gas evolution]. This product was

166
J. L. BELLETIRE ET AL.
subjected to single-crystal x-ray crystallography. 1H NMR [400.132 MHz, DMSO-d6
(center peak ¼ 2.50)] d 9.47 (s, 2H), ꢀ8.6 (br exchangeable peak 3H), 1.7–0.0 (complex
multiplet, 12H). 13C NMR [100.624 MHz, DMSO-d6 (center peak ¼ 39.51)] d 144.03.
FTIR (HATR method; significant peaks) 3318, 3226, 3145, 3109, 3061, 3036, 2479,
2461, 2435, 1747, 1733, 1623, 1546, 1524, 1425, 1360, 1334, 1202, 1070, 1034, 979,
946, 880 cmꢄ1
.
[4-Amino-1-H-1,2,4-triazolium]2[closo-B12H12] (2). Yields: proc. A ¼ 85.8%;
proc. B ¼ 91.7%. 1H NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d 9.43 (s,
4H), ꢀ7.95 (br exchangeable peak, 6H), 1.9–0.0 (complex multiplet, 12H). 13C NMR
[100.624 MHz, DMSO-d6 (center peak ¼ 39.51)] d 144.05. FTIR (HATR method;
significant peaks) 3322, 3305, 3224, 3113, 2464, 2439, 1622, 1552, 1520, 1422,
1356, 1328, 1302, 1205, 1152, 1062, 1029, 976, 939, 870, 711, 682, 612, 576 cmꢄ1
[4-Amino-1-H-1,2,4-triazolium][closo-CB11H12] proc.
(3)[1]. Yield:
.
1
A ¼ 93.7%. H NMR (400.132 MHz, DMSO-d6 (center peak ¼ 2.50) d 9.49 (s, 2H),
ꢀ7.95 (br exchangeable peak, 3H), 2.38 (br s, 1H), 2.3–0.7 (complex multiplet,
11H). 13C NMR [100.624 MHz, DMSO-d6 (center peak ¼ 39.51)] d 144.03, 50.75.
FTIR (HATR method; significant peaks) 3612, 3346, 3285, 3126, 3052, 3023,
3003, 2929, 2514, 1634, 1610, 1512, 1422, 1352, 1319, 1201, 1144, 1094, 1062,
1025, 939, 919, 874, 711, 612 cmꢄ1
.
[5-Amino-x-H-Tetrazolium]2[closo-B12H12] (4). Yields: proc. A ¼ 98.4%;
proc. B ¼ 96.5%). 1H NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)]
d ꢀ 8.16 (br exchangeable peak, 8H), 1.7–0.0 (complex multiplet, 12H). 13C NMR
[100.624 MHz, DMSO-d6 (center peak ¼ 39.51)] d 156.22. FTIR (HATR method;
significant peaks) 3559, 3497, 3403, 3330, 3199, 2480, 1699, 1634, 1581, 1442,
1409, 1311, 1258, 1140, 1066, 1046, 1001, 911, 774, 735, 715, 682, 625 cmꢄ1
.
[5-Amino-x-H-Tetrazolium][closo-CB11H12] (5). Yields: proc. A ¼ 94.3%;
1
proc. B ¼ 96.8%. H NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d ꢀ 7.13
(br exchangeable peak, 4H), 2.38 (br s, 1H), 2.2–0.6 complex multiplet, 11H). 13C
NMR [100.624 MHz, DMSO-d6 (center peak ¼ 39.51)] d 155.89, 50.73. FTIR
(HATR method; significant peaks) 3608, 3534, 3444, 3346, 3318, 3056, 2533, 1708,
1642, 1605, 1450, 1438, 1381, 1278, 1127, 1091, 1058, 1029, 760, 715, 666, 592,
559 cmꢄ1
.
1
[1-H-4-H-1,2,4-Triazolium]2[closo-B12H12] (6). Yield: proc. A ¼ 94.7%. H
NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d ꢀ 13.19 (br exchangeable
peak, 4H), 9.22 (s, 4H), 1.7–0.0 (complex multiplet, 12H). 13C NMR [100.624 MHz,
DMSO-d6 (center peak ¼ 39.51)] d 143.54. FTIR (HATR method; significant peaks)
3438, 3297, 3278, 3252, 3225, 3126, 3058, 3020, 2461, 1644, 1557, 1523, 1398, 1264,
1143, 1063, 1033, 949, 877, 714, 653 cmꢄ1
.
1
[1-H-4-H-1,2,4-Triazolium][closo-CB11H12] (7). Yield: proc. A ¼ 94.7%. H
NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d ꢀ 12.52 (br exchangeable
peak, 2H), 9.23 (s, 2H), 2.37 (br s, 1H), 2.3–0.7 (complex multiplet, 11H). 13C NMR
[100.624 MHz, DMSO-d6 (center peak ¼ 39.51)] d 144.06, 50.72. FTIR (HATR
method; significant peaks) 3469, 3367, 3281, 3130, 3028, 2517, 1740, 1556, 1520,
1418, 1352, 1172, 1152, 1086, 1025, 886, 866, 711, 620 cmꢄ1
.

WATER-SOLUBLE HETEROCYCLIUM BORANE-BASED SALTS
167
[3-Amino-1-H-x-H-1,2,4-triazolium]2[closo-B12H12]
(8). Yield:
proc.
1
A ¼ 90.1%. H NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d ꢀ 13.37 (br
exchangeable peak, 4H), 8.30 (s, 2H), ꢀ7.99 (br exchangeable peak, 4H), 1.8–0.0
(complex multiplet, 12H). 13C NMR [100.624 MHz, DMSO-d6 (center peak ¼ 39.51)]
d 150.90, 139.48. FTIR (HATR method; significant peaks) 3379, 3309, 3260, 3183,
3154, 3113, 2464, 1695, 1610, 1593, 1573, 1389, 1340, 1258, 1066, 1033, 952, 866,
715, 645, 604, 564 cmꢄ1
.
1
[1-H-3-H-1,2,3-Triazolium]2[closo-B12H12] (9). Yield: proc. A ¼ 90.7%. H
NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d 13.26 (br s, 4H), 7.96
(s, 4H), 1.7–0.0 (complex multiplet, 12H). 13C NMR [100.624 MHz, DMSO-d6
(center peak ¼ 39.51)] d 130.35. FTIR (HATR method; significant peaks) 3220,
3154, 3134, 3060, 3044, 2999, 2460, 1560, 1454, 1409, 1368, 1287, 1242, 1127,
1074, 968, 833, 776, 764, 715, 670 cm1.
[1-H-3-H-1,2,3-Triazolium][closo-CB11H12] (10). Yield: proc. A ¼ 96.2%. 1H
NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d 12.44 (br s, 2H), 7.96 (s, 2H),
2.38 (br s, 1H), 2.3–0.7 (complex multiplet, 11H). 13C NMR [100.624 MHz,
DMSO-d6 (center peak ¼ 39.51)] d 130.30, 50.79. FTIR (HATR method; significant
peaks) 3604, 3362, 3240, 3158, 3136, 3051, 3001, 2951, 2905, 2516, 1946, 1854,
1657, 1611, 1553, 1503, 1453, 1407, 1357, 1230, 1118, 1091, 1068, 1022, 960, 883,
795, 760, 710, 671, 621 cmꢄ1
.
[1-Amino-3-H-1,2,3-triazolium]2[closo-B12H12]
(11)[1]. Yield:
proc.
A ¼ 96.7%. 1H NMR [400.132 MHz, DMSO-d6 (center peak ¼ 2.50)] d 8.41
(extremely br s, 6H), 7.942 (d, J ¼ 0.8 Hz, 2H), 7.715 (d, J ¼ 0.8 Hz, 2H), 1.7–0.0
(complex multiplet, 12H). 13C NMR [100.624 MHz, DMSO-d6 (center peak ¼ 39.51)]
d 132.10, 124.19. FTIR (HATR method; significant peaks) 3301, 3224, 3187, 3154,
3126, 2484, 2451, 1593, 1544, 1466, 1401, 1295, 1168, 1148, 1107, 1074, 1058, 1021,
992, 944, 886, 792, 727, 656 cmꢄ1
.
[4-Amino-1-Me-1,2,4-triazolium][1-amino-3-Me-1,2,3-triazolium][closo-
1
B12H12] (12). Yield: proc. B ¼ 90.8%. H NMR (400.132 MHz, DMSO-d6 (center
peak ¼ 2.50) d 10.04 (s, 1H), 9.15 (s, 1H), 8.730 (d, J ¼ 1.2 Hz, 1H), 8.597 (d,
J ¼ 1.2 Hz, 1H), 8.25 (s, 2H), 6.92 (s, 2H), 4.22 (s, 3H), 4.03 (s, 3H), 1.7–0.0 (complex
multiplet, 12H). 13C NMR (100.624 MHz, DMSO-d6 center peak ¼ 39.51) d 145.02,
142.91, 131.51, 126.85, partially resolved N-CH3 peaks on the DMSO-d6 carbon
heptet 39.75 vs. 39.71 and 38.92 vs. 38.88. FTIR (HATR method; significant peaks)
3628, 3555, 3326, 3301, 3248, 3220, 3117, 3097, 3076, 2464, 2018, 1765, 1712, 1638,
1605, 1573, 1532, 1479, 1434, 1401, 1381, 1315, 1209, 1172, 1062, 960, 866, 796, 715,
657, 608 cmꢄ1
.
[4-Amino-1-Me-1,2,4-triazolium][4-amino-1-H-1,2,4-triazolium][closo-
1
B12H12] (13). Yield: proc. B ¼ 81.2%. H NMR [400.132 MHz, DMSO-d6 (center
peak ¼ 2.50)] d 10.04 (s, 1H), 9.44 (s, 2H), two overlapping singlets with essentially
the same chemical shift), 9.14 (s, 1H), 7.00 (extremely br s, 5H), 4.03 (s, 3H),
1.8–0.0 (complex multiplet, 12H). 13C NMR [100.624 MHz, DMSO-d6 (center
peak ¼ 39.51)] d 145.06, 144.05, 142.95, and 38.97. FTIR (HATR method; significant
peaks) 3739, 3583, 3563, 3542, 3506, 3318, 3228, 3109, 3064, 3036, 2942, 2901, 2476,

168
J. L. BELLETIRE ET AL.
1748, 1732, 1618, 1556, 1520, 1425, 1324, 1262, 1205, 1168, 1058, 1033, 980, 939, 886,
866, 754, 711, 678, 653, 612, 555 cmꢄ1
.
[4-Amino-1-H-1,2,4-triazolium][1-Amino-3-Me-1,2,3-triazolium][closo-
1
B12H12] (14). Yield: proc. B ¼ 71.7%. H NMR [400.132 MHz, DMSO-d6 (center
peak ¼ 2.50)] d 9.44 (s, 2H), 8.728 (d, J ¼ 1.2 Hz, 1H), 8.595 (s, J ¼ 1.2 Hz, 1 H),
8.20 (extremely broad s, 5H), 4.22 (s, 3H), 1.7–0.1 (complex multiplet, 12H).
13C NMR [100.624 MHz, DMSO-d6 (center peak ¼ 39.51)] d 144.04, 131.53,
126.87 (N-Me peak not resolvable within the DMSO-d6 heptet). FTIR (HATR
method; significant peaks) 3546, 3318, 3224, 3109, 3032, 2460, 1748, 1732, 1622,
1564, 1524, 1475, 1434, 1401, 1364, 1324, 1238, 1209, 1062, 1038, 980, 948, 907,
874, 809, 715, 682, 618 cmꢄ1
.
Crystal Data for C2H17B12KN4 (1)
Mr ¼ 266.02, orthorhombic, space group Pna2(1), a ¼ 12.166(2), b ¼ 11.443(2),
3
c ¼ 9.868(2) A, a, b, c ¼ 90 , V ¼ 1373.6(4) A , F-(000) ¼ 544, qcalcd ¼ 1.286 gcmꢄ3
,
ꢂ
˚
˚
Z ¼ 4, l ¼ 0.361 mmꢄ1, crystal size 0.3 ꢃ 0.2 ꢃ 0.2 mm3, h range 2.44 to 28.34, index
˚
ranges ꢄ16 ꢅ h ꢅ 16, ꢄ15 ꢅ k ꢅ 15, ꢄ13 ꢅ l ꢅ 13, MoKa (k ¼ 0.71073 A), T ¼ 296(2)
K, 296. The single-crystal x-ray diffraction data were collected on a Bruker 3-circle-
platform diffractometer equipped with a Smart Apex 2 detector with the v axis fixed
at 54.74ꢂ and using MoKa radiation from a fine-focus tube. The goniometer head,
equipped with a nylon Cryoloop and magnetic base, was used to mount the crystals
using perfluoropolyether oil. The data collection as well as structure solution and
refinement were carried out using standard procedures with the APEX2 V.2.1–4,
SMART V.5.622, SAINT 7.24A, SADABS, and SHELXTL software packages
and programs. Measured data 29305 of which 3418 (Rint ¼ 0.0231) were unique.
Transmission factors min=max 0.641=0.746, final R indices [I > 2r(I)]: R1 ¼
0.0231, wR2 ¼ 0.0624, R indices (all data): R1 ¼ 0.0246, wR2 ¼ 0.0613, GOF on
F2 ¼ 1.052. CCDC-787684 contains the supplementary crystallographic data for this
article. These data can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre via http://www.ccdc.cam.ac.uk/data+request/cif.
ACKNOWLEDGMENT
The authors acknowledge the Air Force Research Laboratory (AFRL=RZSP,
Edwards AFB, Calif.) for financial and administrative support.
REFERENCES
1. Shackelford, S. A.; Belletire, J. L.; Boatz, J. A.; Schneider, S.; Wheaton, A. K.; Wight, B.
A.; Hudgens, L. M.; Ammon, H. L.; Strauss, S. H. Pairing heterocyclic cations with closo-
icosahedral borane and carborane anions, I: Benchtop aqueous synthesis of binary triazo-
lium and imidazolium salts with limited water solubility. Org. Lett. 2009, 11, 2623–2626.
2. Larsen, A. S.; Holbrey, J. D.; Tham, F. S.; Reed, C. A. Designing ionic liquids: Imidazolium
melts with carborane anions. J. Am. Chem. Soc. 2000, 122, 7264–7272.

WATER-SOLUBLE HETEROCYCLIUM BORANE-BASED SALTS
169
3. Zhu, Y.; Ching, C.; Carpenter, K.; Xu, R.; Selvaratnam, S.; Hosmane, N. S.; Maguire, J. A.
Synthesis of the novel ionic liquid [N-pentylpyridinium][closo-CB11H11]. Appl. Organome-
tal. Chem. 2003, 17, 346–350.
4. Nieuwenhuyzen, M.; Seddon, K. R.; Teixidor, F.; Puga, A. V. Ionic liquids containing
cluster anions. Inorg. Chem. 2009, 48, 889–901.
5. Bai, Y.; Cao, Y.; Zhang, J.; Wang, M.; Li, R.; Wang, P.; Zakeeruddin, S. M.; Grtzel, M.
High-performance dye-sensitized solar cells based on solvent-free electrolytes produced
from eutectic melts. Nat. Mater. 2008, 7, 626–630.
6. Shackelford, S. A.; Belletire, J. L. Heterocyclic salt synthesis and rational properties
tailoring. Centr. Europ. J. Energ. Mater. 2009, 6, 219–237.
7. Kaplan, G.; Drake, G.; Tollison, K.; Hall, L.; Hawkins, T. Synthesis, characterization, and
structural investigations of 1-amino-3-substituted-1,2,3-triazoliium salts, and a new route
to 1-substituted-1,2,3-triazoles. J. Heterocycl. Chem. 2005, 42, 19–27.