Lewis Acid Catalyzed Synthesis of Cyanidoborates

Article abstract of DOI:10.1002/ejic.201501485

The reactions of [BF₄]^- salts with Me₃SiCN were studied at different temperatures and in the presence of several different Lewis acids, which led to the generation of the well-known [BF_4-n(CN)_n]^- (n = 1-4) salts. Depending on the catalyst and the reaction conditions, the whole series of [BF_4-n(CN)_n]^- salts is now accessible under ambient conditions, and long preparation times and the excessive production of waste materials can be avoided. The best results were obtained when the F^-/CN^- substitution reactions were performed in ionic liquids.

Full text of DOI:10.1002/ejic.201501485

DOI: 10.1002/ejic.201501485
Full Paper
Cyanidoborates
Lewis Acid Catalyzed Synthesis of Cyanidoborates
[
a]
[c]
[a]
[a,b]
[a]
Kevin Bläsing, Stefan Ellinger, Jörg Harloff, Axel Schulz*
Katharina Sievert,
[c]
[a]
[c]
Christoph Täschler, Alexander Villinger, and Cornelia Zur Täschler
–
Abstract: The reactions of [BF ] salts with Me SiCN were stud- salts is now accessible under ambient conditions, and long
4
3
ied at different temperatures and in the presence of several preparation times and the excessive production of waste mate-
different Lewis acids, which led to the generation of the well- rials can be avoided. The best results were obtained when the
–
–
–
known [BF4–n(CN) ] (n = 1–4) salts. Depending on the catalyst F /CN substitution reactions were performed in ionic liquids.
n
and the reaction conditions, the whole series of [BF4–n(CN)n]–
Introduction
tures and large amounts of chemicals. Recently, Finze et al.[17]
The first attempts to generate tetracyanidoborates, M[B(CN) ], reported more efficient syntheses of M[BF(CN) ] (M = Na, K)
SiCl as a catalyst independently to our work on cata-
4
3
[
1–6]
date back to the middle of the 20th century.
claimed the isolation of M[B(CN) ] (M = Cu, Ag) through the lyzed F /CN substitution reactions.
In 1977, Bessler with Me
3
–
–
[18–24]
4
Cyanidoborate salts are used widely,[
25–29]
for example, as
reactions of BCl3 and MCN, but this was later proved to be
[
30]
wrong. The first synthesis, isolation, and full characterization of starting materials for light-element ceramics, the generation
–
salts bearing the [B(CN) ] ion were achieved by Willner et al., of 3D coordination polymers, or as anions in ionic liquids
4
[
8,11,13,31–33]
who studied cyanidoborates intensively between 2000 and (ILs).
Cyanidoborate-based ILs can be obtained read-
[
7–16]
[34–40]
2010.
As shown in Scheme 1, different routes to M[B(CN)4] ily through salt metathesis reactions.
Generally, ILs con-
–
salts have been reported, including a sinter process [Scheme 1, taining the [B(CN) ] anion are chemically robust and have re-
4
Equation (5)], which has been the most efficient and popular markably low viscosities {e.g., EMIm[B(CN) ]: 21 mPa s, EMIm =
4
[
10]
[41]
route until now. The latter reaction is performed at high tem- 1-ethyl-3-methylimidazolium},
peratures (280–290 °C) to form a melt of all three salts (K[BF4], series of applications, for example, as lubricants, electrolytes in
which lead to their use in a
[
42–44]
KCN, and LiCl), from which Li[B(CN) ] can be isolated through a batteries, in electric condensers, and solar cells.
Owing to
4
rather laborious workup procedure. In summary, all of these the wealth of (possible) applications and the problems in the
–
routes [Scheme 1, Equations (1)–(5)] are either expensive, time hitherto known syntheses of [B(CN) ] salts, there is a need for
4
consuming, low yielding, nonselective, or require high tempera- a straightforward, less-expensive synthesis. Herein, we report
Scheme 1. Different synthesis routes to [BF4–n(CN)
n
]– salts (n = 1–4, M = Li, K, X = Cl, Br).
–
–
[
a] Institut für Chemie, Universität Rostock,
Albert-Einstein-Str. 3a, 18059 Rostock, Germany
E-mail: axel.schulz@uni-rostock.de
catalytic F /CN substitution reactions that lead to the desired
–
[BF4–n(CN) ] ions (n = 1–4) depending on the catalyst and reac-
n
tion conditions at ambient temperatures.
http://www.schulz.chemie.uni-rostock.de/
[
[
b] Abteilung Materialdesign, Leibniz-Institut für Katalyse e.V. an der
Universität Rostock,
Albert-Einstein-Str. 29a, 18059 Rostock, Germany
c] Lonza Ltd., Valais Works,
Results and Discussion
Synthesis without Catalyst
Lonzastrasse, 3930 Visp, Switzerland
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/ejic.201501485.
The selective syntheses of [nBu N][BF(CN) ] and [nBu N][B(CN) ]
4
3
4
4
are readily achieved when [nBu N][BF ] is treated with an excess
4
4
Eur. J. Inorg. Chem. 2016, 1175–1183
1175
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
of Me SiCN in an autoclave under autogenous pressure [Ph P][BF ] with Me SiCN did not generate the pure cyanidobo-
3
4
4
3
[
18,19]
(
Scheme 2).
Depending on the reaction conditions, espe- rates, and either decomposition of the cation or product mix-
[
18,19]
cially the temperature profile during the reaction, either tures were observed.
By salt metathesis reactions with the
[
nBu N][BF(CN) ] or [nBu N][B(CN) ] can be isolated. These reac- silver salt, the generation of the alkali metal salt is readily
4
3
4
4
[
7,10]
tions were optimized with respect to the temperature and achieved.
From [nBu N][B(CN) ], Ag[B(CN) ] could be iso-
4 4 4
stoichiometry of Me SiCN. For the isolation of [nBu N][B(CN) ], lated in 96 % yield, and K[B(CN) ] was obtained in 90 % yield
3
4
4
4
the optimized parameters are 180 to 200 °C, 15–30 h reaction from the Ag[B(CN)4].
time, and 6 equiv. of Me SiCN [Scheme 2, Equation (7)]. After
3
3
Table 2. Reaction conditions and isolated yields for the synthesis of [BF(CN) ]
salts from [BF4] salts (nOct = n-octyl).
workup and careful purification, [nBu N][B(CN) ] was isolated in
–
4
4
good yields of 70–75 %.
Cation
Temperature [°C]
Time [h]
Isolated yield [%]
[
[
[
[
[
Me
Et MeN]
nBu
nOct
4
N]+
150
150
140–210
150
13
13
20–0.25
13
82
53
83–58
76
+
3
+
4
N]
+
4
N]
+
nBu
4
P]
160/200
20
79/64
Synthesis with Lewis Acids as Catalysts
Scheme 2. Syntheses of [nBu
under autogenous pressure without a catalyst.
4
N][BF(CN)
3
] and [nBu
4
4
N][B(CN) ] in an autoclave
–
–
Recently, we discovered that the F /CN substitution in fluorido-
cyanidophosphates can be accelerated by the addition of Lewis
As depicted in Scheme 2 [Equation (6)], [nBu N][BF(CN) ] was
obtained under considerably milder conditions and shorter re-
action times. The optimized parameters are listed in Table 1.
[22]
–
–
4
3
acids.
Therefore, we studied the F /CN substitution at the
boron center in the presence of a Lewis acid.
Upon the addition of Me SiCN to [Ph C][BF ] at ambient tem-
3
3
4
Compared to the synthesis procedure for [nBu N][B(CN) ], a
–
4
4
perature, the exclusive formation of the [B(CN) ] ion was de-
4
smaller excess of Me SiCN (3.5–4.0 equiv.) and lower tempera-
11
3
tected after 20 h through B NMR spectroscopy experiments.
tures were used. As can be seen from Table 1, the lower the
temperature is, the longer it takes to obtain [nBu N][BF(CN) ]
exclusively as the main product. At a temperature of 130 °C,
only a reaction mixture consisting of [nBu N][BF (CN) ] with
n = 2 and 3 was isolated after 30 h. This procedure also works
–
After 2 h reaction time, the [BF(CN) ] ion was formed exclu-
3
4
3
sively. Unfortunately, a rather fast oligomerization of CN species
was observed, and the reaction mixture turned into a black
viscous product. Oligomerization in this reaction has been dis-
4
4–n
n
[7,9]
[45]
cussed previously by Willner et al.
and Bessler et al.
For
nicely for other symmetrically substituted tetraalkylammonium
this reason, we changed the Lewis acidic catalyst (see next para-
graph). In the first reaction step, it can be assumed that Ph C–
+
[18,19]
salts ([R N] with R = alkyl).
4
3
+
–
CN is formed in addition to [Me Si–CN–SiMe ] (and [BF ] ), the
3
3
4
Table 1. Reaction conditions and isolated yields for the synthesis of
nBu N][BF(CN) ] from [nBu N][BF ] and 3.5–4 equiv. of Me SiCN.
actual Lewis acidic catalyst. [Me Si–CN–SiMe ][BF ] can readily
3
3
4
[
4
3
4
4
3
–
–
exchange F ions with CN ions to yield Me SiF and the adduct
3
Time [h]
Temperature [°C]
Isolated yield [%]
Me SiCN·BF . The formation of the latter adduct has been de-
3
3
2
1
1
5
1
0
0
3
0
140
150
170
190
200
210
83[a]/76
74
74
scribed previously by Willner et al., who observed adduct for-
mation in the reaction of Et O·BF with Me SiCN. This group
2
3
3
[
9]
also reported the X-ray data of Me SiNC·BF(CN) . Ph C–CN
3
2
3
61
was isolated in large amounts and fully characterized, including
by X-ray structure elucidation.
60
[
b]
.25
58
[
a] 83 % yield when the reaction was scaled up (20-fold). [b] The product was
+
impure owing to the start of the decomposition of the [nBu
4
N] cation at
Variation of the Lewis Acid
this temperature.
–
–
To study the influence of Lewis acids on the F /CN exchange
reaction in borates, we set up a standard experiment:
In the temperature range 140–210 °C, only the exclusive for-
mation of [nBu N][BF(CN) ] (within rather short reaction times)
[
nBu N][BF ] was treated with 10 equiv. of Me SiCN in the pres-
4 4 3
4
3
ence of 5 mol-% of a Lewis acid (LA) at ambient temperature
was observed (Table 2). However, with increasing reaction tem-
perature, a decrease of the isolated yield (76 → 58 %) was ob-
served. This decrease was attributed to the beginning of the
decomposition of the cation, as indicated by variable-tempera-
1
1
19
under an argon atmosphere. B and F NMR spectroscopy was
used to evaluate the progress of the substitution reaction. For
comparison, the same reaction was performed in the absence
of the LA catalyst, as listed in Table 3.
1
ture H NMR spectroscopy studies. According to these experi-
+
ments, for the [nBu N] ion, the temperature limit seems to be
4
Table 3. Product distribution [%] for the treatment of [nBu
4
N][BF
4
] with
200 °C, above which significantly enhanced decomposition was
11
1
0 equiv. of Me SiCN at 25 °C, as determined by B NMR spectroscopy.
3
–
observed. Contrary to the synthesis of [B(CN) ] salts, [R R′N]-
4
3
Time [h]
1
6
23
47
100
[
BF ] (R,R′ = alkyl) and [nBu P][BF ] can also be used (in addition
4
4
4
–
–
[BF (CN)]
99
1
93
7
80
20
79
21
82
18
to [R N][BF ]) to generate the corresponding [BF(CN) ] salts.
3
4
4
3
[
BF
2
(CN)
2
]–
However, the treatment of [NH ][BF ], [Et NH][BF ], and
4
4
3
4
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org
1176
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
When no catalyst was utilized, the almost exclusive forma- –15.7 ppm, δ F = –183.5 ppm, and J = 40 Hz),[17] which we
1
9
1
B,F
tion of [nBu N][BF (CN)] was observed after 1 h at 25 °C. With only found when P(CN) was used as the LA catalyst. In this
4
3
3
–
11
increasing reaction time, the amount of [nBu N][BF (CN) ] also case, in addition to the resonances of [BF (CN) ] (δ B =
increased. After 23 h, a [BF (CN)] /[BF (CN) ] ratio of 80:20 was –7.3 ppm, t, J = 42 Hz) and [BF(CN) ] (δ B = –17.7 ppm, d,
detected.
All of the product distributions derived from the 11B NMR –15.7 ppm with a 1
spectroscopy studies of the same reaction in the presence of –182.4 ppm, cf. [BF(CN)
4
2
2
2
2
–
–
1
–
11
3
2
2
B,F
3
1
11
JB,F = 45 Hz), an additional doublet was detected at δ B =
19
J
B,F coupling constant of 40 Hz {δ F =
–
19
–
] : δ F = –211.2 ppm and [BF
3
(CN)
] :
2
2
1
9
19
different Lewis acids are shown in Table 4 (10 equiv. of Me SiCN, δ F = –153.2 ppm}. All of the signals in the F NMR spectra
% catalyst, 25 °C). Upon the addition of NbCl , FeCl , or SiCl , appear as quartets owing to coupling with the B nucleus (I =
5 3 4
3
1
1
5
the formation of [nBu N][BF (CN) ] was detected exclusively in 3/2). According to our NMR spectroscopy studies, the maximal
4
2
2
the 1 B and F NMR spectra after 3 h at ambient temperature. amount of the [BF(NC)(CN)
1
19
] ion reached was ca. 7 %. In all
–
2
–
The utilization of PCl , TiCl , or GaCl led to the formation of other cases, we did not observe the [BF(NC)(CN)
] ion as inter-
5
4
3
2
[
nBu N][BF(CN) ] in almost quantitative yields under the same mediate, and this can either be attributed to a low concentra-
4 3
thermodynamic conditions (10 equiv. of Me SiCN, 5 % catalyst, tion or a different reaction pathway (see below).
3
2
5 °C) as those discussed for NbCl , FeCl , and SiCl . For the
As the formation of [nBu N][B(CN) ] was detected in the reac-
5
3
4
4 4
GaCl -catalyzed reaction only, the formation of [nBu N][B(CN) ] tion with GaCl even at ambient temperature, the same reaction
3
4
4
3
was also detected with increasing reaction time, even at ambi- was performed under reflux conditions (Me SiCN b.p. 114–
3
ent temperature. After 44 h, 7 % of [nBu N][B(CN) ] was ob- 117 °C). When [nBu N][BF ] was treated with 5 mol-% of GaCl
4
4
4
4
3
served in addition to 93 % of [nBu N][BF(CN) ]. The activity of and 10 equiv. of Me SiCN at 120 °C, the quantitative formation
4
3
–
3
–
the used Lewis acids with respect to the F /CN exchange reac- of [nBu N][B(CN) ] was observed after 3 h. After workup,
4
4
tion increases along the series: CrCl ≈ MnCl < NbCl ≈ FeCl
[nBu N][B(CN) ] was obtained in rather good yields of ca. 80 %.
3
2
5
3
4 4
≈
SiCl < PCl ≈ TiCl < GaCl . Notably, all of these reactions Interestingly, the utilization of PCl or TiCl instead of GaCl as
4 5 4 3 5 4 3
can be used to prepare either [nBu N][BF (CN) ] or the catalyst did not lead to the exclusive formation of
4
2
2
[
nBu N][BF(CN) ] in rather good isolated yields (69–80 %).
[nBu N][B(CN) ].
4
3
4 4
Finally, as shown in Table 4, all cyanidoborates of the type
Table 4. Product distribution [%] for the treatment of [nBu
0 equiv. of Me SiCN in the presence of 5 % LA catalyst at 25 °C, as deter-
mined by B NMR spectroscopy.
4 4
N][BF ] with
[
nBu N][BF (CN) ] (n = 1–4) can be generated in isolated
4
4–n
n
1
3
11
yields of 69–90 % by utilizing the appropriate catalyst (see Sup-
porting Information).
LA
Time [h]
[BF
3
(CN)]– [BF
2
(CN)
2
]
[BF(CN)
3
]–
[B(CN)
4
]–
–
CrCl
100
25
20
3
3
3
3
3
3
82
–
–
–
–
–
–
–
–
–
–
–
18
100
100
100
100
100
95
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
7
3
Spectroscopic Characterization
MnCl
NbCl
FeCl
SiCl
P(CN)
PCl
TiCl
2
The 11B and 19F NMR spectroscopy data of all cyanidofluorido-
5
–
3
borates of the type [BF4–n(CN) ] (n = 0–4) described in this
n
4
–
1
1
19
work are listed in Table 5. B and F NMR spectroscopy are
particularly well-suited to distinguish between cyanido-
5[
a]
3
5
100
100
100
93
–
[
7,9,10,13,14,16,17,31]
fluoridoborates.
With the increasing number of
4
C]+
19
11
[
Ph
3
19
44
cyanido groups attached to the boron atom, the ( F, B, and
C) chemical shifts are increasingly high-field-shifted as shown
in Table 5. The large difference in the F chemical shifts (20–
0 ppm) allows the use of this method to check the purity
of salts containing fluoridoborates. The observed J coupling
GaCl
GaCl
3
3
13
[
b]
[b]
[b]
3
100
]– and 4 %
] . [b] The temperature was increased to 120 °C (reflux condi-
1
9
[
a] Sum of cyanido- and isocyanidoborate with 1 % [BF(CN)
3
6
–
[
BF(NC)(CN)
2
1
B,F
tions).
1
constants lie in the expected range of 28–45 Hz, the J values
B,C
2
As mentioned in the introduction, independently to our are 70–90 Hz, and JC,F coupling constants of 30–80 Hz are ob-
[
7,9,10,13,14,16,17,31]
work, Finze et al. described a new synthesis route to M[BF(CN)3] served.
(
M = Na, K) from M[BF ] and 9 equiv. of Me SiCN in the pres-
The melting points as well as selected IR and Raman data of
4
3
17]
[
ence of 10 mol-% Me SiCl (100 °C, 6 h).
The authors dis- all considered [nBu N][BF (CN) ] salts (n = 0–4) are listed in
3
4
4–n
n
cussed the action of Me SiCl as that of a Lewis acid. Moreover, Table 6. The compounds show sharp bands that can be as-
3
–
11
the [BF(NC)(CN) ] ion was observed as an intermediate (δ B = signed to the νCN stretching frequencies in the expected region
2
Table 5. 11B, 13C, and 19F NMR spectroscopic data of all [BF4–n(CN)
]– ions.
δ13C [ppm]
n
Ion
δ11B [ppm]
δ19F[a] [ppm]
1JB,F [Hz]
1JB,C [Hz]
2JC,F [Hz]
–
[b]
[
[
[
[
[
BF
BF
BF
4
3
2
]
–1.2 (s)
–3.6 (q)
–7.2 (t)
–17.9 (d)
–38.2 (s)
–151.6
–131.5
–153.1
–211.7
–
–
–
86
81
74
–
(CN)]–
(CN)
]–
]–
]–
131.5 (qq)
130.0 (tq)
127.0 (dq)
28
42
45
–
78
50
37
–
2
BF(CN)
B(CN)
3
[c]
71[c], 23[c]
4
122.5
[
a] Signals appear as quartets. [b] Not observed. [c] Quartet due to coupling with 11B (I = 3/2) and septet due to coupling with 10B (I = 3).
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org © 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1177

Full Paper
–
1
ν˜ = 2200–2250 cm . If more than one cyanido group is at- Direct Synthesis in Ionic Liquids – Synthesis of Ionic
tached to the boron atom, an in-phase and out-of phase cou- Liquids
pling is observed. A shift to higher wavenumbers was observed
with increasing number of CN groups at the boron atom.
The reaction of [BMIm][BF ] (BMIm = 1-butyl-3-methylimida-
4
zolium) with Me SiCN in the presence of GaCl is rather fast.
3
3
After 2 h, quantitative conversion to [BMIm][B(CN) ] was found,
and the isolated yield was 85 % after the workup and drying
process (see Experimental Section). For the synthesis of pure
Table 6. Melting points [°C] and νCN [cm–1] of [nBu
4
4
N][BF4–n(CN)
n
] (n = 0–4).
n =
0
1
2
3
4
[
a]
M.p.
153
–
–
48
2210
2211
62
2214
2216
83
2222
2222
EMIm[BF(CN) ], EMIm[BF ] was treated with Me SiCN in the
3 4 3
ν
ν
CN,IR
2206
2207
presence of [Ph C][BF ] as a Lewis acidic catalyst at ambient
3
4
CN,Raman
temperature. The best isolated yield (85 %) was obtained after
20 h of stirring at ambient temperature with a catalyst concen-
[
a] Broad glass transition.
tration of 3.5 mol-% and 10 equiv. of Me SiCN.
3
The melting point of [nBu N][BF ] is 153 °C. The melting
As shown previously, in addition to the temperature,
4
4
points decrease significantly as cyanido groups are attached to stoichiometry, catalyst, and reaction time, the cation has a con-
the B atom. All [nBu N][BF (CN) ] salts (n = 1–4) can be re- siderable influence on the product distribution. Furthermore,
4
4–n
n
ferred to as ionic liquids as their melting points are lower than with respect to facile cation-exchange reactions, we were inter-
00 °C (Table 6).
ested in the preparation of [nPr NH][BF4–n(CN) salts
Scheme 3). Such salts allow facile cation-exchange reactions
1
3
]
n
(
when treated with bases, as illustrated in Scheme 3 [Equa-
tion (10)]. We tested three Lewis acids (FeCl , PCl , and GaCl )
3
5
3
Direct Synthesis of Alkali Cyanidoborates
that we knew would have different reactivities (see the above
+
discussion on [nBu N] salts, Table 4). In the presence of FeCl ,
In 2003, Willner et al. described the synthesis of
4
3
only [nPr NH][BF (CN) ] was isolated after 23 h at ambient tem-
Li[BF (CN) ]·2Me SiCN, which was isolated in 60 % yield, when
3
2
2
2
2
3
perature. With catalytic amounts of PCl , [nPr NH][BF(CN) ] was
Li[BF ] was treated with 7 equiv. of Me SiCN under reflux condi-
5
3
3
4
3
[
9]
isolated. The desired [nPr NH][B(CN) ] was isolated within 6 h
tions for 4 h [Scheme 1, Equation (2) with M = Li and n = 2].
3 4
–
under reflux conditions when GaCl was used as the Lewis acid.
3
Within 7 d under the same reaction conditions, the third F ion
–
It should be noted that [nPr NH][BF (CN)] could not be ob-
was exchanged by a CN ion. A similar situation was observed
3 3
tained by this approach. The isolated yields were in the range
6 to 78 %; however, scaling up these reactions should lead to
considerably improved yields, e.g., the best yield of
nPr NH][B(CN) ] was more than 78 % when the reaction was
when K[BF ] was used. We studied the same reactions
4
5
[
Scheme 1, Equation (2) with M = Li, K] but in the presence of
a Lewis acid catalyst. In the presence of 5 mol-% GaCl , an
3
[
immediate, rather violent reaction was observed after the addi-
tion of Me SiCN to M[BF ] (M = Li, K). After 4 h under reflux, a
3
4
scaled up. All [nPr NH][BF (CN) ] salts are highly viscous oily
3
4–n
n
3
4
ionic liquids.
mixture of 35 % Li[BF(CN) ] and 65 % Li[B(CN) ] was detected
3
4
1
1
19
in the B and F NMR spectra; therefore, these reactions are
also strongly enhanced by the addition of Lewis acids. After the
addition of another 3 equiv. of Me SiCN and 12 h under reflux,
3
Synthesis of M[BF4–n(CN) ] in [nPr NH][BF (CN) ] Ionic
Liquids
n
3
4–n
n
almost pure Li[B(CN) ] could be isolated (isolated yield 63 %).
4
Similar results were obtained for the potassium salt, that is, a
fast conversion to K[BF(CN) ] and a very slow reaction for the [nPr NH][BF (CN) ] salts represent ionic liquids with a labile
3
3
4–n
n
last substitution step (Table 7).
cation as it readily decomposes under basic conditions
[
[
Scheme 3, Equation (8)]. Therefore, the utilization of
nPr NH][BF (CN) ] ILs allows neutralization reactions with
3
4–n
n
metal hydroxides or amides to form the M[BF4–n(CN) ] salts
n
Table 7. Product distribution [%] for the reaction of K[BF
4
] with 10 equiv. of
30
without any problematic workup procedures such as the use of
expensive silver or lithium salts. In this case, the only side prod-
ucts are H O and nPr N, which can be removed readily in vacuo.
Me
3
SiCN and 5 mol-% GaCl
3
under reflux conditions.
Time [h]
K[BF(CN)
6
15
22
2
3
3
]
96
4
93
7
93
7
93
7
By this route, costs and chemicals are saved. Furthermore, each
4
K[B(CN) ]
of the M[BF4–n(CN) ] (n = 1–4) salts are readily and quickly ac-
n
3 n
Scheme 3. Synthesis of [nPr NH][BF4–n(CN) ] and application in salt metathesis (M = Li, K).
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org
1178
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cessible by this new route. To test this procedure, we treated process are the salt metathesis step with silver salts (Scheme 5)
[
nPr NH][B(CN) ] with KOH to afford K[B(CN) ] in 80–90 % yield. and the necessary use of autoclaves. To our minds, the LA catal-
3 4 4
ysis process represents a route that is fast, high-yielding, and
easy to perform not just for K[B(CN) ] but also for almost all
4
Comparison of Different Routes to K[B(CN)4]
metal tetracyanidoborate salts. Moreover, the number and
–
+
–
+
amount of impurities (e.g., Cl , Li , CN , or Ag ions) in the final
products is much smaller than for the other methods.
Finally, we have listed the advantages and disadvantages of the
different routes for the synthesis of K[B(CN) ] in Table 8. The
4
major advantage of the sinter process of Willner (Scheme 4) is
the use of the less-expensive KCN as the cyanide source instead
Thermodynamic Considerations – Catalysis
of Me SiCN for the autoclave and LA-catalyzed reactions
3
At the B3LYP/aug-cc-pvTZ level of theory, the thermodynamic
(
Schemes 5 and 6). However, for the sinter process, large
–
–
data of the successive F /CN substitution reactions were com-
puted; in agreement with the experimental observations, the
gas-phase Gibbs energies for the consecutive substitution reac-
amounts of LiCl, higher temperatures, and more reaction steps
are needed, and a large amount of waste product is generated.
Furthermore, the sinter process is not easy to perform, as hot
spots need to be avoided, and the purity of the starting materi-
als plays an essential role in the reaction yields; however, the
reported yields of up to 60 % are hard to reproduce. The disad-
vantages of the autoclave reaction compared to the LA catalysis
2
98
tions are all exergonic (Table 9: ΔG = –13.2, –8.6, –11.5, and
–1
–
16.3 kcal mol ).
Table 9. Calculated gas-phase Gibbs energies [kcal mol–1] for the consecutive
–
n
substitution reactions of the [BF4–n(CN) ] ion.
Reaction step[a]
ΔG289
]– + TCN → [BF
(CN)] + TF
–
–13.3
–8.6
–11.5
–16.3
–49.6
[
[
[
[
[
BF
BF
BF
4
3
2
3
–
–
(CN)] + TCN → [BF
2
(CN)
2
]
]
+ TF
+ TF
–
–
(CN)
2
]
+ TCN → [BF(CN)
3
–
BF(CN)
3
]– + TCN → [B(CN)
4
]
+ TF
–
–
BF
4
]
+ 4TCN → [B(CN)
4
]
+ 4 TF
Scheme 4. Synthesis of K[B(CN)
4
] by the sinter process.[10]
[
a] T = Me
3
Si.
The natural bond orbital (NBO) analysis data revealed a suc-
cessive decrease of the partial charge at the boron atom
–
–
{
[BF ] : +1.3 to –0.22 e in [B(CN) ] , Table 10}, whereas the
4
4
charge at the F atoms only decreases slightly (–0.57 to –0.50 e).
The charges in the neutral Lewis acids BF3–n(CN) also decrease
n
(1.41–0.47 e, Table 10). Interestingly, the polarities of the B–
F (ca. 20/80 %) and B–C bonds (ca. 35/65 %) do not change
significantly upon substitution (Table 10).
Scheme 5. Synthesis of [nBu
sis reaction to form K[B(CN)
4
N][B(CN)
].
4
] in an autoclave process and metathe-
4
Table 8. Reaction parameters of different routes to K[B(CN)
Sinter process[10]
4
].
Autoclave
[nBu N][BF
Me SiCN
200
15
LA catalysis
[nPr NH][BF
Me SiCN
120
6
Starting material
Cyanide source
T [°C]
Reaction time [h]
Product
Intermediate yield [%]
Metathesis product
Steps
K[BF
KCN
300
1
4
]
4
4
]
3
4
]
3
3
Li[B(CN)
4
][a]
[nBu
75
4
N][B(CN)
4
]
3 4
[nPr NH][B(CN) ]
[
a]
–
70–80 %
none
2
[nPr
3
3
NH][B(CN)
4
]
4
Ag[B(CN) ]
3
Overall time [h]
Waste products
Solvents
20
25
[nBu
MeOH, CH
60–70 %
10
KCN, LiCl, LiF, KCl, nPr
CH Cl , THF
<60
3
N
4
N]NO
3
, AgBr
CN
nPr
CH
60–80 %
3
N
Cl
2
2
b]
3
2
2
[
Yield K[B(CN)
4
] [%]
[
a] The lithium salt is not isolated from the reaction mixture. [b] It is very hard to achieve 60 %; we only achieved isolated yields of 20–35 % with the sinter
method. However, we do want to stress that the overall yield depends strongly on the purity of the starting materials and the avoidance of hot spots during
the reaction.
3 4 4
Scheme 6. Synthesis of [nPr NH][B(CN) ] in an LA-catalyzed process and metathesis reaction to form K[B(CN) ].
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org
1179
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 10. NBO partial charges [e] and bond polarities [% localization] in
acids). Both Lewis acidic centers of [LA1–CN–LA2] can attack
–
BF3–n(CN)
n
and [BF4–n(CN)
n
] .
–
the [BF4–n(CN) ] anions to trigger the fluoride abstraction (as
n
–
well as the CN abstraction). So far, we do not have any experi-
Species
q(B)
q(F)
B–C
B–F
mental proof (in situ NMR spectroscopy experiments) for the
generation of the proposed intermediates; this suggests either
a fast reaction on the NMR time scale or a very low concentra-
BF
BF
BF(CN)
B(CN)
3
1.41
1.18
0.88
0.47
1.30
1.06
0.75
0.32
–0.22
–0.47
–0.44
–0.41
0.00
–0.57
–0.54
–0.52
–0.50
0.00
–
18/8
18/82
19/80
–
17/83
18/82
19/81
20/80
–
2
CN
33/67
35/65
37/63
–
30/70
33/67
36/64
40/60
2
3
–
tion of the intermediates. The formation of Lewis acidic silylium
[
[
[
[
[
BF
BF
BF
BF(CN)
B(CN)
4
3
2
]
–
species of the type [Me Si–X–LA] (X = halogen or CN ) in the
(CN)]–
3
]–
reactions of Me
3
Si–X with Lewis acids has been described previ-
(CN)
2
]–
[
52–58]
ously,
catalysis.
and silylium ions have been isolated and used in
3
]–
[59–63]
4
A good measure for the intrinsic stability of the [BF4–n(CN) ]–
n
ions is the Lewis acidity of the parent Lewis acid BF (CN) (n =
3
–n
n
–
[46]
0
–3), for example, BF for the [BF ] ion.
The Lewis acidity
3
4
[
46–51]
can be expressed by the fluoride ion affinity (FIA)
or (for
this specific problem) the cyanide ion affinity (CIA). Here, the
Lewis acidity combines the strength of a Lewis acid LA(g) with
the energy that is released upon the binding of a fluoride or
cyanide ion in the gas phase: LA(g) + X → [LA–X]– + ΔH298
–
(g)
(g)
[
FIA = –ΔH298 (X = F) and CIA = –ΔH298 (X = CN)]. According to
our DFT computations (Table 11), the Lewis acidity of
BF3–n(CN) (n = 1–3) increases strongly compared with that of
n
BF . For example, the FIA values increase along BF (77.4, cf.
3
3
[
47]
experimental values of 71–76)
107.1) < B(CN)3 (120.1 kcal mol ). Therefore, cyanidation
< BF (CN) (92.4) < BF(CN)
2
2
–1
(
greatly increases the thermodynamic stability of all cyanidobo-
–
rates relatively to the perfluorinated [BF ] ion, or B(CN) repre-
4
3
sents a very strong Lewis acid, which is much stronger than
BF3.
Table 11. Calculated gas-phase fluoride ion affinities and cyanide ion affinities
–
1
(
CIA) of all BF3–n(CN)
n
Lewis acids [kcal mol ].
3
Scheme 7. Proposal for catalytic cycles (T = Me Si, LA = Lewis acid). Top: LA
generates the bis(silylated) nitrilium ion ([T–CN–T] ), which acts as the cata-
ΔH298
ΔG298
+
–
]–
BF
BF
BF(CN)
B(CN)
3
+ F → [BF
4
–77.4
–92.4
–107.1
–120.1
–48.5
–60.8
–77.7
–96.0
–68.1
–83.9
–98.1
–112.1
–38.9
–50.1
–67.1
–85.8
lyst; middle and bottom: the LA–NC–T adduct acts as a Lewis acidic catalyst
–
(CN)]–
2
(CN) + F → [BF
3
–
and attacks one fluorine atom of [BF4–n(CN)
n
] either at the Si (middle) or the
–
]–
2
+ F → [BF
2
(CN)
2
LA center (bottom).
–
–
3
3
+ F → [BF(CN) ]
+ CN → [BF
3
(CN) + CN → [BF (CN)
–
(CN)]–
BF
BF
3
Notably, the catalytic cycles in Scheme 7 are proposals based
on the assumption that Lewis acids in the presence of a strong
–
]–
2
2
2
]–
–
BF(CN)
B(CN)
2 3
+ CN → [BF(CN)
+ CN → [B(CN)
–
–
donor (such as Me SiCN) form adducts (e.g., LA–NC–SiMe ) that
3
4
]
3
3
diminish their acidity compared with those of the free acids.
–
–
The situation of this formally simple F /CN substitution is
Furthermore, the CIA values supports this finding as
they increase considerably along BF3 (48.5) BF (CN)
rather complex, and the amount of catalyst and Me SiCN, the
Lewis acidities of the different catalysts, the countercation of
BF4–n(CN) ] , and the increasing amount of generated Me SiF
in each cycle must all be considered. Additionally, the in situ
3
<
2
–1
(
60.8) < BF(CN) (77.7) < B(CN) (96.0 kcal mol ); however, the
2 3
–
[
n
3
CIAs are much smaller than the FIAs.
As demonstrated experimentally, Lewis acidic catalysts en-
hance the fluoride abstraction. Through the addition of a Lewis
acid, the S 1-type substitution at Me SiCN and the fluoride ab-
formed BF3–n(CN) species can also act as Lewis acids. Last but
n
–
–
not least, all F /CN substitutions reactions should be regarded
as equilibrium reactions that depend on temperature and pres-
sure.
N
3
straction should be enhanced by the formation of a catalyst of
the type [Me Si–CN–LA]. As shown in Scheme 7, three different
3
types of catalytic cycle can be proposed. They can even super-
+
impose with each other. The intermediately formed Me Si
3
(solv.)
Conclusions
can act as
0
a Lewis acid as well as BF3–n(CN)n (n =
–
–
–
–3), which is formed upon formal F abstraction. Hence, a gen- The utilization of Lewis acids enhances the F /CN exchange in
eralized catalyst can be formulated as [LA1–CN–LA2] (with different types of [BF4–n(CN)n]– salts upon the addition of
+
+
LA1 = Me Si and LA2 = Me Si and all other considered Lewis Me SiCN. It was possible to apply very mild conditions (ambient
3
3
3
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org
1180
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
temperature, low pressure) when small amounts of Lewis acidic the removal of the dichloromethane in vacuo, a white solid was
obtained. The solid was washed three times with diethyl ether
catalysts were added. Depending on the reactions conditions
(
5 mL) and dried at 50 °C in vacuo to afford [nBu N][BF(CN) ]
4 3
and the catalyst, selective, high-yielding routes to pure
–
–
–
–
(1.063 g, 3.03 mmol, 84 %), m.p. 62 °C. C19H N BF (350.33): calcd.
36 4
[
BF (CN)] , [BF (CN) ] , [BF(CN) ] , and [B(CN) ] salts were
3 2 2 3 4
1
C 65.14, H 10.36, N 15.99; found C 65.49, H 10.51, N 16.29. H NMR
300 K, CD CN, 300.13 MHz): δ = 0.97 (t, 12 H, CH ), 1.35 (m, 8 H,
–
–
found. The best results were obtained when the F /CN substi-
tution reactions were performed in ionic liquids such as
(
3
3
13
1
CH CH ), 1.61 (m, 8 H, CH CH N), 3.09 (m, 8 H, NCH ) ppm. C{ H}
3
2
2
2
2
[
nPr HN][BF (CN) ] or BMIm[BF4–n(CN)n].
3 4–n n
NMR (300 K, CD CN, 75.47 MHz): δ = 13.9 (s, 4 C, CH ), 20.4 (m, 4
3
3
C, CH CH ), 24.4 (s, 4 C, NCH CH ), 59.4 (m, 4 C, NCH ), 127.9 [dq,
2
3
2
2
2
1
2
11
JC,B = 75 Hz, J = 37 Hz, 3 C, BF(CN) ] ppm. B NMR (300 K,
CD CN, 96.29 MHz): δ = –17.9 [d, 1 B, BF(CN) ] ppm. F NMR (300 K,
C,F
3
1
9
3
3
CD CN, 282.40 MHz): δ = –211.7 [q, 1 F, BF(CN) ] ppm. IR (25 °C,
3
3
Experimental Section
ATR, 32 scans): ν˜ = 536 (w), 737 (m), 802 (w), 903 (s), 937 (m), 959
(m), 989 (w), 1039 (s), 1080 (w), 1111 (w), 1171 (w), 1323 (w), 1362
(w), 1381 (w), 1473 (m), 2214 (w), 2875 (m), 2935 (m), 2964 (m)
General: A detailed description of all used chemicals, drying proce-
dures, analytical devices, catalytic studies, and so forth can be found
–
1
in the Supporting Information. Here, we have only listed the most cm . Raman (25 °C, 6 mW, five scans): ν˜ = 220 (2), 258 (5), 297 (1),
efficient synthesis routes.
447 (1), 485 (1), 505 (1), 533 (1), 591 (2), 880 (1), 906 (4), 925 (1),
1
1
032 (1), 1054 (2), 1109 (4), 1170 (1), 1323 (4), 1348 (1), 1448 (4),
General Procedure for the Synthesis of [nBu N][BF (CN) ] (n =
4
4–n
n
483 (1), 2216 (10), 2736 (1), 2876 (7), 2898 (4), 2938 (8), 2967 (5)
2
–4) in the Presence of a Catalyst: [nBu N][BF ] (1 equiv.) and a
–1
4
4
cm .
Lewis acid (ca. 5 mol-%) were added to a Schlenk flask. Then,
Me SiCN (10 equiv.) was added under an argon atmosphere with a
3
[nBu N][B(CN) ]: [nBu N][BF ] (497 mg, 1.51 mmol), GaCl (5 mol-
4
4
4
4
3
syringe. According to the solubility of the Lewis acid, a suspension
or solution formed. The reaction mixture was stirred at ambient
temperature or under reflux for several hours. The exact conditions
can be found in the following synthesis instructions. The reaction
mixture quickly turned brown, especially during the syntheses of
%
, 13 mg, 0.07 mmol), and Me SiCN (1.50 g, 15.1 mmol) were
3
heated under reflux under an argon atmosphere for 3 h. After the
removal of the CH Cl , the obtained light yellow solid was dried at
2
2
50 °C in vacuo to afford [nBu N][B(CN) ] (425 mg, 1.19 mmol, 79 %),
4 4
m.p. 83 °C. C H N B (357.34): calcd. C 67.22, H 10.15, N 19.60;
found C 66.43, H 9.96, N 19.00. H NMR (300 K, CDCl , 300.13 MHz):
δ = 1.03 (t, 12 H, CH ), 1.44 (m, 8 H, CH CH ), 1.62 (m, 8 H, CH CH N),
3.12 (m, 8 H, NCH ) ppm. B NMR (300 K, CDCl , 96.29 MHz): δ =
20 36 5
[
nBu N][BF(CN) ] and [nBu N][B(CN) ]. For the subsequent workup,
1
4
3
4
4
3
the excess Me SiCN and formed Me SiF were removed in vacuo.
3
3
3
3
2
2
2
The resulting oil was mixed with distilled water and aqueous H O2
11
2
2
3
(30 wt.-%). The suspension was stirred at 80 °C for several hours
1
13
1
–38.2 [s, J = 71 Hz, 1 B, B(CN) ] ppm. C{ H} NMR (300 K, CDCl ,
B,F 4 3
and then cooled to ambient temperature, the suspension was fil-
tered, and the solid was washed twice with water. The product was
75.47 MHz): δ = 13.5 (s, 4 C, CH ), 19.5 (m, 4 C, CH CH ), 23.6 (s, 4
3
2
3
1
C, NCH CH ), 58.7 (m, 4 C, NCH ), 122.5 [q + sept, J = 71 Hz,
2
2
2
C,B
extracted from the remaining solid with CH Cl , and the solution
1
2
2
JC,B = 23 Hz, 4 C, B(CN) ] ppm. IR (25 °C, ATR, 32 scans): ν = 534
4
˜
was filtered. The organic phase was dried with anhydrous Na SO ,
2
4
(w), 735 (m), 802 (w), 885 (m), 932 (s), 966 (m), 991 (m), 1036 (m),
061 (w), 1111 (w), 1169 (w), 1244 (w), 1323 (w), 1360 (w), 1381
m), 1456 (w), 1473 (s), 2222 (w), 2877 (m), 2935 (m), 2964 (m)
which was removed by filtration. The dichloromethane was evapo-
rated with a rotary evaporator. The obtained product was dried
in vacuo for several hours to yield white or light yellow solids of
1
(
–1
cm . Raman (25 °C, 6 mW, five scans): ν = 258 (1), 486 (1), 521 (1),
597 (1), 884 (1), 906 (1), 930 (1), 1035 (1), 1058 (1), 1110 (2), 1132
˜
[
nBu N][BF (CN) ], [nBu N][BF(CN) ], or [nBu N][B(CN) ].
4 2 2 4 3 4 4
(
(
(
1), 1171 (1), 1309 (1), 1324 (1), 1449 (2), 1462 (1), 1485 (1), 2222
10), 2739 (1), 2871 (3), 2877 (3), 2897 (2), 2900 (2); 2918 (3), 2938
4), 2968 (2), 2984 (2), 3009 (1) cm .
[
7
nBu N][BF (CN) ]: [nBu N][BF ] (491 mg, 1.49 mmol), FeCl (20 mg,
4 2 2 4 4 3
mol-%), and Me SiCN (1.58 g, 1.59 mmol) were stirred under an
3
–
1
argon atmosphere at ambient temperature for 3 h. After purification
as described in the general procedure, the obtained white solid was
dried at 50 °C in vacuo to yield [nBu N][BF (CN) ] (400 mg,
[
EMIm][BF(CN) ]: EMIm[BF ] (0.739 g, 3.73 mmol), [Ph C][BF ]
3
4
3
4
4
2
2
(3.4 mol-%, 43 mg, 0.13 mmol), and Me SiCN (3.67 g, 37 mmol)
3
1
.17 mmol, 69 %), m.p. 48 °C. C H N BF (343.31): calcd. C 62.97,
18 36 3 2
were stirred under an argon atmosphere at ambient temperature
for 20 h. The excess Me SiCN and any Me SiF were removed in
vacuo to afford a light brown oily residue, which was suspended in
aqueous H O (4 mL, 40 mmol, 30 wt.-%), and the suspension was
stirred at 70 °C for 1 h. After the H O solution cooled to ambient
temperature, butyl acetate (20 mL) was added. The resulting mix-
ture was transferred into centrifuge tubes. After centrifugation
1
H 10.57, N 12.24; found C 62.58, H 10.65, N 12.35. H NMR (300 K,
CDCl , 300.13 MHz): δ = 0.99 (t, 12 H, CH ), 1.41 (m, 8 H, CH CH ),
1
CDCl , 96.29): δ = –7.2 [t, J = 42 Hz, 1 B, BF (CN) ] ppm. C{ H}
NMR (300 K, CDCl , 75.47 MHz): δ = 13.3 (s, 4 C, CH ), 19.4 (m, 4 C,
CH CH ), 23.6 (s, 4 C, NCH CH ), 58.6 (m, 4 C, NCH ) ppm. F NMR
3
3
3
3
3
2
1
1
.61 (m, 8 H, CH CH N), 3.13 (m, 8 H, NCH ) ppm. B NMR (300 K,
2 2 2
1
13
1
2 2
3
B,F
2
2
2
2
3
3
19
2
3
2
2
2
1
(300 K, CDCl , 282.40 MHz): δ = –153.1 [q, J = 42 Hz, 2 F, BF (CN) ]
3 B,F 2 2
(2000 rpm, 2 min), the supernatant layer was separated. The butyl
ppm. IR (25 °C, ATR, 32 scans): ν˜ = 550 (w), 621 (w), 633 (w), 737
acetate was removed with a rotary evaporator. The obtained light
yellow oil was washed three times with diethyl ether (5 mL) and
dried at 70 °C in vacuo to afford EMIm[BF(CN) ] (0.694 g, 3.17 mmol,
8
4
1
(m), 798 (w), 839 (w), 879 (s), 916 (m), 939 (m), 945 (m), 1007 (s),
1
1
049 (s), 1074 (s), 1107 (m), 1171 (w), 1242 (w), 1321 (w), 1360 (w),
–
1
3
383 (w), 1473 (m), 2210 (w), 2877 (m), 2937 (m), 2964 (m) cm .
5 %). C H N BF (219.03): calcd. C 49.35, H 5.06, N 37.97; found C
9 11 5
Raman (6 mW, five scans): ν˜ = 175 (2), 260 (4), 393 (1), 486 (1), 632
1
8.84, H 5.03, N 37.27. H NMR (300 K, CD CN, 300.13 MHz): δ =
3
(2), 877 (2), 900 (2), 1031 (1), 1056 (2), 1064 (2), 1109 (4), 1129 (2),
.46 (t, 3 H, CH ), 3.82 (s, 3 H, NCH ), 4.16 (q, 2 H, CH ), 7.32 (m, 1
3
3
2
1
2
322 (3), 1363 (1), 1448 (5), 1463 (3), 2211 (8), 2745 (1), 2879 (8),
1
1
H, EtNCH), 7.38 (m, 1 H, MeNCH), 8.40 (s, 1 H, NCHN) ppm. B NMR
–
1
897 (5), 2925 (8), 2939 (10), 2972 (6) cm .
1
9
(300 K, CD CN, 96.29 MHz): δ = –17.9 [d, 1 B, BF(CN) ] ppm.
F
3
3
[
nBu N][BF(CN) ]: [nBu N][BF ] (1.189 g, 3.6 mmol), [Ph C][BF ]
NMR (300 K, CD CN, 282.40 MHz): δ = –211.6 [q, 4 F, BF(CN) ] ppm.
4
3
4
4
3
4
3
3
(
3.6 mol-%, 44 mg), and Me SiCN (3.55 g, 36 mmol) were stirred
IR (25 °C, ATR, 32 scans): ν˜ = 596 (w), 621 (s), 646 (m), 702 (w), 746
3
under an argon atmosphere at ambient temperature for 19 h. After
(m), 837 (m), 899 (s), 939 (s), 957 (m), 1045 (s), 1167 (s), 1336 (w),
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org
1181
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1
390 (w), 1429 (w), 1452 (w), 1468 (w), 1572 (m), 2214 (w), 2991
CH Cl . The organic phase was dried with anhydrous Na SO and
2
2
2
4
–
1
(w), 3120 (w), 3157 (w) cm .
filtered. The dichloromethane was evaporated with a rotary evapo-
rator. The obtained product was dried in vacuo for several hours to
yield a colorless to yellow oil of [nPr NH][BF (CN) ], [nPr NH]-
[BMIm][B(CN) ]: [BMIm][BF ] (1.05 g, 4.65 mmol), GaCl (7 mol-%,
4 4 3
3
2
2
3
6
0 mg, 0.34 mmol), and Me SiCN (4.6 g, 46 mmol) were heated
3
[
BF(CN) ], or [nPr NH][B(CN) ].
3
3
4
under reflux under an argon atmosphere for 2 h. The excess
Me SiCN and any Me SiF were removed in vacuo to afford a dark
[nPr
NH][BF
(CN)
]: [nPr
NH][BF
] (0.57 g, 2.47 mmol), FeCl
SiCN (2.48 g, 25 mmol) were
stirred at ambient temperature for 23 h. The product was dried in
vacuo to afford [nPr NH][BF (CN) ] (363 mg, 1.48 mmol, 60 %) as a
colorless oil, m.p. 7 °C. C11 BF (245.12): calcd. C 53.90, H 9.05,
N 17.14; found C 53.23, H 9.11, N 16.64. H NMR (300 K, CDCl
300.13 MHz): δ = 1.01 (t, 9 H, CH ), 1.73 (m, 6 H, CH CH ), 3.06 (m,
6 H, NCH ), 7.39 (br, 1 H, NH) ppm. B NMR (300 K, CD Cl
96.29 MHz): δ = –7.3 [t, JB,F = 41 Hz, 1 B, BF ] ppm. F NMR
(300 K, CD Cl , 282.40 MHz): δ = –154.1 [q, B,F = 41 Hz, 2 F,
BF (CN)
] ppm. IR (25 °C, ATR, 32 scans): ν˜ = 552 (w), 631 (w), 754
(m), 868 (s), 939 (m), 1009 (s), 1053 (s), 1088 (s), 1387 (w), 1462 (m),
.17 (t, 2 H, NCH ), 7.33 (s, 1 H, BuNCH), 7.34 (s, 1 H, MeNCH), 8.44 1473 (m), 2210 (w), 2806 (w), 2885 (w), 2943 (w), 2976 (m), 3061 (w)
3
3
3
2
2
3
4
3
brown oily residue, which was suspended in aqueous H O (7 mL,
(5 mol-%, 23 mg, 0.14 mmol), and Me
3
2
2
7
0 mmol, 30 w%), and the solution was stirred at 90 °C for 1 h.
After the H O2 suspension cooled to ambient temperature, butyl
3
2
2
2
acetate (50 mL) was added. The resulting mixture was transferred
into centrifuge tubes. After centrifugation (2000 rpm, 2 min), the
supernatant layer was separated. The butyl acetate was removed
with a rotary evaporator, and the product was dried at 100 °C in
vacuo to afford [BMIm][B(CN) ] (1.00 g, 3.95 mmol, 85 %). C H N B
H
22
N
3
2
1
,
3
3
2
3
11
,
2
2
2
1
19
(CN)
2
J
4
12 15 6
2
1
(254.10): calcd. C 56.72, H 5.95, N 33.07; found C 56.58, H 6.33, N
2
2
1
3
1
4
2.74. H NMR (300 K, CD CN, 250.13 MHz): δ = 0.97 (t, 3 H, CH CH ),
2
2
3
2
3
.37 (m, 2 H, CH CH ), 1.87 (m, 2 H, CH CH ), 3.94 (s, 3 H, NCH ),
2
3
2
2
3
2
11
–1
(s, 1 H, NCHN) ppm. B NMR (300 K, CDCl , 80.25 MHz): δ = –38.4
cm .
3
1
13
1
[s, JB,F = 71 Hz, 1 B, B(CN) ] ppm. C{ H} NMR (300 K, CD CN,
4
3
[
nPr NH][BF(CN) ]: [nPr NH][BF ] (0.50 g, 2.16 mmol), PCl (8 mol-
3 3 3 4 5
6
1
2.90 MHz): δ = 13.07 (s, 1 C, CH ), 19.15 (s, 1 C, CH CH ), 31.56 (s,
3 2 3
%
, 41 mg, 0.20 mmol), and Me SiCN (2.18 g, 22 mmol) were stirred
3
C, CH CH ), 36.39 (s, 1 C, NCH ), 49.94 (s, 1 C, NCH ), 122.3 [q +
2
2
3
2
at ambient temperature for 23 h. The product was dried in vacuo
1
sept, J = 71, 24 Hz, 4 C, B(CN) ], 122.5 (s, 1 C, BuNCH), 123.7 (s,
1
B,C
4
to afford [nPr NH][BF(CN) ] (304 mg, 1.21 mmol, 56 %) as a yellow
3
3
C, MeNCH), 135.0 (s, 1 C, NCHN) ppm. IR (25 °C, ATR, 32 scans):
oil. C H N BF (252.14): calcd. C 57.16, H 8.79, N 22.22; found C
1
2 22 4
ν˜ = 623 (s), 650 (m), 746 (m), 841 (m), 930 (s), 964 (m), 976 (m),
1
5
7.00, H 8.40, N 21.93. H NMR (300 K, CDCl , 300.13 MHz): δ = 1.01
3
1
1
109 (m), 1165 (s), 1242 (m), 1338 (w), 1369 (w), 1383 (w), 1429 (w),
464 (m), 1572 (m), 1726 (m), 2222 (w), 2877 (w), 2937 (w), 2964
(
t, 9 H, CH ), 1.73 (m, 6 H, CH CH ), 3.01 (m, 6 H, NCH ), 7.25 (br, 1
3 2 3 2
11
1
H, NH) ppm. B NMR (300 K, CD Cl , 96.29 MHz): δ = –17.7 [d, J =
4
–
1
2
2
B,F
(w), 3099 (w), 3118 (w), 3153 (w) cm .
19
4 Hz, 1 B, BF(CN) ] ppm. F NMR (300 K, CD Cl , 282.40 MHz): δ =
3 2 2
1
Li[B(CN) ]: Li[BF ] (5 g, 53 mmol) and GaCl (470 mg, 2.67 mmol, –212.3 [q, JB,F = 44 Hz, 1 F, BF(CN)
5
atmosphere. Then, Me SiCN (50 mL, 375 mmol) was added with a
syringe. During the addition, the solution became warm, and in-
tense gas evolution was observed. The reaction mixture was heated
under reflux for 4 h and then cooled to ambient temperature, and
Me SiCN (20 mL, 160 mmol) was added. The reaction mixture was
heated under reflux for a further 13 h, and the excess Me SiCN and
any Me SiF were removed in vacuo. The remaining brown solid was
dissolved in H O (50 mL) and aqueous H O (10 mL, 100 mmol, 30
wt.-%), and the solution was stirred at 70 °C for 2 h. After the solu-
tion was cooled to ambient temperature, it was filtered, and all
volatile compounds were removed in vacuo. The product was ex-
tracted with CH CN in a Soxhlet extractor for several hours. After
the removal of the solvent in vacuo, the product was recrystallized
from tetrahydrofuran (thf) to afford Li[B(CN) ] (4.1 g, 33.6 mmol,
6
] ppm. IR (25 °C, ATR, 32 scans):
4
4
3
3
mol-%) were added to a 100 mL Schlenk flask under an argon ν˜ = 754 (m), 822 (w), 895 (s), 939 (m), 959 (m), 984 (m), 1049 (s),
1387 (w), 1460 (m), 1473 (m), 2214 (w), 2885 (w), 2943 (w), 2976
3
–
1
(m), 3074 (w) cm .
[
nPr NH][B(CN) ]: [nPr NH][BF ] (3.31 g, 14.3 mmol), GaCl (5 mol-
3 4 3 4 3
%
, 138 mg, 0.78 mmol), and Me SiCN (15.0 g, 151 mmol) were
3
3
heated under reflux for 6 h. The product was dried in vacuo to
3
afford [nPr NH][B(CN) ] (2.91 mg, 11.2 mmol, 78 %) as an orange
3
4
3
oil. C H N B (259.16): calcd. C 60.25, H 8.56, N 27.02; found C
1
3 22 5
2
2 2
1
5
1
9.56, H 8.47, N 25.86. H NMR (300 K, CD Cl , 300.13 MHz): δ =
2 2
.01 (t, 9 H, CH ), 1.71 (m, 6 H, CH CH ), 3.00 (m, 6 H, NCH ), 6.67
3
2
3
2
11
(br, 1 H, NH) ppm. B NMR (300 K, CD Cl , 96.29 MHz): δ = –38.3
2 2
[
9
1
s, 1 B, B(CN) ] ppm. IR (25 °C, ATR, 32 scans): ν˜ = 754 (m), 845 (w),
4
3
30 (s), 966 (m), 980 (m), 1041 (w), 1093 (w), 1387 (w), 1460 (m),
–1
473 (m), 2224 (w), 2883 (w), 2943 (w), 2976 (m), 3088 (w) cm .
4
3 %). LiC N B (121.82): calcd. C 39.44, H 0.00, N 45.99; found C
K[B(CN) ]: [nPr NH][B(CN) ] (2.0 g, 7.72 mmol) in CH Cl (20 mL)
4 3 4 2 2
4
4
1
1
3
8.97, H 0.17, N 45.15. B NMR (300 K, D O, 96.29): δ = –38.2 [s, and KOH (433 mg, 7.72 mmol) in water (distilled, 4 mL) were stirred
2
1
13
1
JB,C = 71 Hz, 1 B, B(CN) ] ppm. C{ H} NMR (300 K, D O, 62.90 MHz):
at ambient temperature for 30 min. Subsequently, all volatile com-
pounds were removed in vacuo to afford a beige solid. The residue
4
2
1
δ = 122.8 [q + sep, J = 72, 24 Hz, 4 C, B(CN) ] ppm. IR (25 °C,
ATR, 32 scans): ν = 856 (m), 947 (s), 986 (m), 2262 (w) cm . Raman
BC
4
–1
was washed twice with CH
Cl
(15 mL). The product was recrystal-
] (952 mg, 6.18 mmol, 80 %) as
B (153.98): calcd. C 31.20, N 36.39;
found C 37.17, N 36.26. 11B NMR (300 K, D O, 96.29 MHz): δ = –38.2
˜
2
2
(
(
25 °C, 12 mW, five scans): ν˜ = 193 (6), 201 (2), 523 (1), 526 (1), 950 lized from water to afford K[B(CN)
4
–
1
1), 988 (1), 2263 (10) cm .
a white crystalline solid. KC N
4 4
2
General Procedure for the Synthesis of [nPr NH][BF4–n(CN)n]
3
[
9
s, 1 B, B(CN) ] ppm. IR (25 °C, ATR, 64 scans): ν˜ = 831 (w), 851 (s),
4
(
n = 2–4): [nPr NH][BF ] (1 equiv.) and the Lewis acid (ca. 5 mol-%)
–1
3
4
37 (s), 974 (m), 1001 (m), 2233 (w) cm . Raman (25 °C, 6 mW, five
were added to a Schlenk flask. Then, Me SiCN (10 equiv.) was added
–1
3
scans): ν˜ = 174 (1), 497 (1), 528 (1), 946 (1), 2236 (10) cm .
under an argon atmosphere with a syringe. The reaction mixture
was stirred at ambient temperature or under reflux for several
hours. The exact conditions can be found in the following synthesis
introductions. During the reactions, the solutions quickly turned
brown. For the workup, the excess Me SiCN and formed Me SiF
Acknowledgments
3
3
Financial support by the Lonza Group Ltd. (grant to K. S.) is
were removed in vacuo. The resulting oil was mixed with distilled
water and aqueous H O (30 wt.-%). The mixture was stirred at gratefully acknowledged. The authors also thank the Deutsche
2
2
ambient temperature for 15 h, and the product was extracted with Forschungsgemeinschaft (DFG) for financial support within the
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org
1182 © 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
29] J. Harloff, M. Karsch, A. Schulz, A. Villinger, Eur. J. Inorg. Chem. 2014, 896–
Priority Program (Material Synthesis Near Room Temperature,
SPP 1708).
9
07.
30] D. Williams, B. Pleune, J. Kouvetakis, M. D. Williams, R. A. Andersen, J.
Am. Chem. Soc. 2000, 122, 7735–7741.
[
[
[
[
31] B. H. Hamilton, C. J. Ziegler, Chem. Commun. 2002, 842–843.
32] C. Nitschke, M. Köckerling, Z. Anorg. Allg. Chem. 2009, 635, 503–507.
33] C. Nitschke, M. Köckerling, Inorg. Chem. 2011, 50, 4313–4321.
34] D. Freudenmann, S. Wolf, M. Wolff, C. Feldmann, Angew. Chem. Int. Ed.
Keywords: Ionic liquids · Lewis acids · Borates ·
Cyanidoborates · Fluoridoborates
2
011, 50, 11050–11060; Angew. Chem. 2011, 123, 11244.
35] Q. Zhang, K. De Oliveira Vigier, S. Royer, F. Jérôme, Chem. Soc. Rev. 2012,
1, 7108–7146.
[
[
[
[
[
[
1] G. Wittig, P. Raff, Z. Naturforsch. B 1951, 6, 225–226.
2] G. Wittig, P. Hornberger, Z. Naturforsch. B 1951, 6, 225.
3] E. Bessler, Z. Anorg. Allg. Chem. 1977, 430, 38–42.
4] J. Emri, B. Györi, J. Chem. Soc., Chem. Commun. 1983, 1303.
5] B. Györi, J. Emri, I. Fehér, J. Organomet. Chem. 1983, 255, 17–28.
6] B. F. Spielvogel, F. U. Ahmed, M. K. Das, A. T. McPhail, Inorg. Chem. 1984,
[
[
4
36] H. Weingärtner, Angew. Chem. Int. Ed. 2008, 47, 654–670; Angew. Chem.
2008, 120, 664–682.
[37] I. Krossing, J. M. Slattery, C. Daguenet, P. J. Dyson, A. Oleinikova, H. We-
ingärtner, J. Am. Chem. Soc. 2006, 128, 13427–13434.
[38] H. Xue, R. Verma, J. M. Shreeve, J. Fluorine Chem. 2006, 127, 159–176.
[39] D. Zhao, Z. Fei, R. Scopelliti, P. J. Dyson, Inorg. Chem. 2004, 43, 2197–205.
[40] P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed. 2000, 39, 3772–3789;
Angew. Chem. 2000, 112, 3926.
2
3, 3263–3265.
7] E. Bernhardt, G. Henkel, H. Willner, Z. Anorg. Allg. Chem. 2000, 626, 560–
68.
[
[
[
5
8] M. Berkei, E. Bernhardt, M. Schürmann, M. Mehring, H. Willner, Z. Anorg.
Allg. Chem. 2002, 628, 1734.
[41] V. Armel, J. M. Pringle, M. Forsyth, D. R. Macfarlane, D. L. Officer, P.
Wagner, Chem. Commun. 2010, 46, 3146–3148.
9] E. Bernhardt, M. Berkei, H. Willner, M. Schürmann, Z. Anorg. Allg. Chem.
2
003, 629, 677–685.
10] E. Bernhardt, M. Finze, H. Willner, Z. Anorg. Allg. Chem. 2003, 629, 1229–
234.
11] T. Küppers, E. Bernhardt, H. Willner, H. W. Rohm, M. Köckerling, Inorg.
Chem. 2005, 44, 1015–1022.
[42] N. Madria, T. A. Arunkumar, N. G. Nair, A. Vadapalli, Y.-W. Huang, S. C.
Jones, V. P. Reddy, J. Power Sources 2013, 234, 277–284.
[43] W. R. Pitner, M. Uerdingen, 2009, DE102009037300A1.
[44] D. Kuang, P. Wang, S. Ito, S. M. Zakeeruddin, M. Grätzel, J. Am. Chem. Soc.
2006, 128, 7732–7733.
[
[
[
[
[
[
1
12] M. Finze, E. Bernhardt, M. Berkei, H. Willner, J. Hung, R. M. Waymouth,
Organometallics 2005, 24, 5103–5109.
13] M. Neukirch, S. Tragl, H.-J. Meyer, T. Küppers, H. Willner, Z. Anorg. Allg.
Chem. 2006, 632, 939–944.
[45] E. Bessler, J. Goubeau, Z. Anorg. Allg. Chem. 1967, 352, 67–76.
[46] I. Krossing, I. Raabe, Chem. Eur. J. 2004, 10, 5017–5030.
[47] J. C. Haartz, D. H. McDaniel, J. Am. Chem. Soc. 1973, 95, 8562–8565.
[48] H. D. B. Jenkins, I. Krossing, J. Passmore, I. Raabe, J. Fluorine Chem. 2004,
125, 1585–1592.
[49] T. E. Mallouk, G. L. Rosenthal, G. Mueller, R. Brusasco, N. Bartlett, Inorg.
Chem. 1984, 23, 3167–3173.
[50] J. M. Slattery, S. Hussein, Dalton Trans. 2012, 41, 1808–1815.
[51] K. O. Christe, D. A. Dixon, D. McLemore, W. W. Wilson, J. A. Sheehy, J. A.
Boatz, J. Fluorine Chem. 2000, 101, 151–153.
[52] R. Labbow, F. Reiß, A. Schulz, A. Villinger, Organometallics 2014, 33, 3223–
3226.
[53] F. Reiß, S. Axel, A. Villinger, Chem. Eur. J. 2014, 20, 11800–11811.
[54] M. F. Ibad, P. Langer, F. Reiß, A. Schulz, A. Villinger, J. Am. Chem. Soc.
2012, 134, 17757–17768.
14] T. Küppers, M. Köckerling, H. Willner, Z. Anorg. Allg. Chem. 2007, 633,
2
80–284.
15] T. Küppers, E. Bernhardt, C. W. Lehmann, H. Willner, Z. Anorg. Allg. Chem.
007, 633, 1666–1672.
2
[
[
16] E. Bernhardt, H. Willner, Z. Anorg. Allg. Chem. 2009, 635, 2511–2514.
17] J. A. P. Sprenger, J. Landmann, M. Drisch, N. Ignat'ev, M. Finze, Inorg.
Chem. 2015, 54, 3403–3412.
[
[
[
18] S. Ellinger, C. Rijksen, L. Ott, K. Sievert, J. Harloff, A. Schulz, 2014, WO
2
014029833 A1 20140227.
19] A. Schulz, L. Ott, C. Rijksen, S. Ellinger, K. Sievert, J. Harloff, 2014, WO
014029834 A1 20140227.
2
20] C. Zur Täschler, S. Ellinger, P. Franke, J. Harloff, C. Rijksen, C. Täschler, L.
[55] M. F. Ibad, P. Langer, A. Schulz, A. Villinger, J. Am. Chem. Soc. 2011, 133,
21016–21027.
Ott, A. Schulz, K. Sievert, A. Stoffers, 2014, WO 2014167034 A2,
2
0141016.
[56] A. Schulz, J. Thomas, A. Villinger, Chem. Commun. 2010, 46, 3696–3698.
[57] M. Lehmann, A. Schulz, A. Villinger, Angew. Chem. Int. Ed. 2009, 48, 7444–
7447; Angew. Chem. 2009, 121, 7580.
[58] A. Schulz, A. Villinger, Chem. Eur. J. 2010, 16, 7276–7281.
[59] H. F. T. Klare, K. Bergander, M. Oestreich, Angew. Chem. Int. Ed. 2009, 48,
9077–9079; Angew. Chem. 2009, 121, 9241.
[60] L. Olsson, C.-H. Ottosson, D. Cremer, J. Am. Chem. Soc. 1995, 117, 7460–
7479.
[61] S. Duttwyler, C. Douvris, N. L. P. Fackler, F. S. Tham, C. A. Reed, K. K.
Baldridge, J. S. Siegel, Angew. Chem. Int. Ed. 2010, 49, 7519–7522; Angew.
Chem. 2010, 122, 7681.
[
[
[
[
[
[
[
[
21] C. Zur Täschler, S. Ellinger, P. Franke, J. Harloff, C. Rijksen, C. Täschler, L.
Ott, A. Schulz, K. Sievert, A. Stoffers, 2014, WO 2014167035 A1 20141016.
22] K. Sievert, S. Ellinger, J. Harloff, L. Ott, C. Rijksen, A. Schulz, 2014, EP
2
772495 A1 20140903.
23] C. Zur Täschler, K. Sievert, A. Schulz, J. Harloff, S. Ellinger, C. Täschler,
015, WO 2015067404 A1 20150514.
24] C. Zur Täschler, K. Sievert, A. Schulz, J. Harloff, S. Ellinger, C. Täschler,
015, WO 2015067405 A1 20150514.
25] M. Karsch, H. Lund, A. Schulz, A. Villinger, K. Voss, Eur. J. Inorg. Chem.
012, 5542–5553.
26] J. Harloff, M. Karsch, H. Lund, A. Schulz, A. Villinger, Z. Anorg. Allg. Chem.
014, 640, 1330–1341.
27] H. Lund, J. Harloff, A. Schulz, A. Villinger, Z. Anorg. Allg. Chem. 2014, 640,
971–1982.
28] H. Lund, J. Harloff, A. Schulz, A. Villinger, Z. Anorg. Allg. Chem. 2013, 639,
54–764.
2
2
2
[62] H. F. T. Klare, M. Oestreich, Dalton Trans. 2010, 39, 9176–9184.
[63] G. Meier, T. Braun, Angew. Chem. Int. Ed. 2009, 48, 1546–1548; Angew.
Chem. 2009, 121, 1575–1577.
2
1
Received: December 27, 2015
Published Online: February 18, 2016
7
Eur. J. Inorg. Chem. 2016, 1175–1183
www.eurjic.org
1183
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim