Pseudohalide HCN aggregate ions: [N3(HCN)3]?, [OCN(HCN)3]?, [SCN(HCN)2]?and [P(CN·HCN)2]?

Article abstract of DOI:10.1039/d0dt02973d

In the presence of μ-nitridobis(triphenylphosphonium) cation, [PPN]⁺, it was possible to stabilize and isolate [PPN]⁺-salts bearing the highly labile hydrogen cyanide aggregate anions of pseudohalides X (X = N₃, OCN and SC

Full text of DOI:10.1039/d0dt02973d

Dalton
Transactions
View Article Online
View Journal
PAPER
Pseudohalide HCN aggregate ions: [N3(HCN)3]−,
[OCN(HCN)3]−, [SCN(HCN)2]− and [P(CN·HCN)2]− †
Cite this: DOI: 10.1039/d0dt02973d
a
Jörg Harloff, a Axel Schulz,
*
a,b Philip Stoer a and Alexander Villinger
In the presence of μ-nitridobis(triphenylphosphonium) cation, [PPN]+, it was possible to stabilize and
isolate [PPN]+-salts bearing the highly labile hydrogen cyanide aggregate anions of pseudohalides X (X =
N3, OCN and SCN). From a concentrated solution of the [PPN]X salts in HCN, crystals of [PPN][X(HCN)3]
(X = N3, OCN) or [PPN][SCN(HCN)2] could be obtained, when the crystallization was carried out fast and
at low temperatures. The reaction of liquid HCN with the PCO− salt led to formation of dicyanophosphide,
which crystallized as HCN disolvate [P(CN·HCN)2]−. All synthesized salts with hydrogen-bridged pseudo-
halide solvate anions are thermally unstable. Immediate loss of HCN was observed in the crystals outside
the HCN solution. Oligomerization begins at ambient temperature, even in HCN solution. All discussed
species were characterized by means of Raman spectroscopy, single crystal X-ray analysis and quantum-
chemical calculations.
Received 25th August 2020,
Accepted 11th September 2020
DOI: 10.1039/d0dt02973d
rsc.li/dalton
ing from K[FHF].9 Nearly 80 years later, Bozorth determined
the X-ray structure of the potassium salt for the first time.10
Introduction
The term “Pseudohalogen” was already introduced in 1925 by Later on, the solid state structures of higher aggregates of the
Birckenbach and Kellermann and describes heteroatomic type [F(HF)n]− (n = 2–5, 7) were determined.11–18 Furthermore,
molecules that are very similar in their chemical behaviour to the molecular structures of [Cl(HCl)n]− (n = 1–5)19–23 and [Br
halogens.1–5 The most prominent features of pseudohalogen (HBr)n]− (n = 1, 2)21,24,25 were investigated by single crystal
compounds are probably the formation of (i) single negatively X-ray diffraction, while the iodides could not be crystallized so
charged anions, such as X− = CN−, OCN−, SCN−, N3 or OCP−, far. In 1965, Dove synthesized the first hydrogen dipseudoha-
−
(ii) poorly soluble silver or mercury salts (AgX, Hg2X2), dimers lide salt, [Ph4As][H(NCS)2], starting from [Ph4As][HCl2] and
(X–X, e.g. NC–CN) as well as Brønsted acids HX. In addition, two equivalents of KSCN in SO2 (Scheme 1).26 By means of
some pseudohalogens exist as neutral inter(pseudo)halogens X-ray crystallography a slightly bent structure of the anion in
(e.g. BrCN) and form covalent non-metal compounds (e.g. the solid state could be observed.27,28 Dehnicke et al. were able
P(CN)3) as well as salts, containing pseudohalogenido com- to synthesize [Ph4P][H(N3)2] with in situ generated HN3, which
plexes, such as [Si(CN)6]2− 6–8
.
Moreover, some disproportion- was treated with [Ph4P]N3 (Scheme 1).29 Salthouse and
Waddington described the attempted synthesis of [nPr4N][H
ate in basic water solution (e.g. NC–CN + 2 OH− → OCN−
+
CN− + H2O). This concept can be extended to further charac- (CN)2], which led to an oily residue with the characteristics
teristic features with respect to solvate cations ([HnXn−1]+) and (according to IR analysis) of the mono-solvate anion [H
anions ([X(HX)n]−), which can also exist as mixed halogen/ (CN)2]−.30 However, the determination of its molecular struc-
pseudohalogen ions, for example [Y(HX)n]− (X and Y = halide ture is still missing.
and/or pseudohalide).
Less is known about mixed aggregate anions of the type [Y
Solvate anions containing the same halides have been (HX)n]−. Most of these compounds were studied using IR,
known for almost two centuries. It dates back to 1856, when
Edmond Frémy described the synthesis of anhydrous HF start-
aInstitut für Chemie, Universität Rostock, Albert-Einstein-Straße 3a, 18059 Rostock,
Germany. E-mail: axel.schulz@uni-rostock.de
bLeibniz-Institut für Katalyse an der Universität Rostock, Albert-Einstein-Straße 29a,
18059 Rostock, Germany
†Electronic supplementary information (ESI) available. CCDC 2024349–2024353.
For ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/d0dt02973d
Scheme 1 Synthesis of [Ph4As][H(NCS)2],26 [Ph4P][H(N3)2]29 and
[WCC][CN(HCN)n]31 ([WCC] = weakly coordinating cation = [Ph4P] for n
= 2 and [PPN] for n = 3).
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
NMR or MS techniques, which gave information on the bond
strengths of the hydrogen bridges.30,32–38 Recently, the group
of Hasenstab-Riedel reported on the crystal structures of
[Ph4P]- and [PPN]-salts, containing solvate anions with
different halides, such as [X(HCl)4]− (X = Br, I), [I(HCl)3]−, [I
(HBr)2]− and [X(HF)2(HX)]− (X = Br, I), as well as the first struc-
tures of [ClO4(HF)2]− and [Br(HCN)]−, which contain both, a
halide and pseudohalide.23
Just recently, we reported on the synthesis of HCN aggre-
gates of the type [CN(HCN)n]− (n = 2, 3), starting from [WCC]
CN ([WCC] = weakly coordinating cation39,40 = [PPN], [Ph4P])
and liquid HCN at low temperatures.31 This work prompted us
to investigate the HCN solvation of other pseudohalides, such
as N3−, OCN−, SCN− and OCP−. Here, we present the synthesis
of low-temperature stable [PPN]+-salts with [X(HCN)3]− (X = N3,
OCN), [SCN(HCN)2]− and [P(CN·HCN)2]− as HCN-solvate
counter anions, which were investigated by X-ray crystallogra-
phy and low-temperature Raman spectroscopy.
Fig. 1 Ball-and-stick representation of the [N3(HCN)3]− ion in a crystal
of [PPN][N3(HCN)3]. The distorted Y-shaped molecular anion is formed
by three hydrogen bridges (a–c). When short contacts (d and e) are con-
sidered, a chain is formed along the c-axis. The cations are omitted for
clarity. Selected distances are given in Å, heavy atom distances in the
H-bridges: N1–C1 2.97(2), N3–C2 3.02(2), N3–C3 3.033(8); a = 2.12, b =
2.15, c = 2.03. Shortest interanionic distances: d = N4–C3’ 3.391(6), e =
C1–N6’ 3.70(2).
Results and discussion
Synthesis
We started this project with the synthesis of pseudohalide
salts of the type [PPN]X (X = N3, OCN, SCN, OCP), which were
produced using a modified literature reaction.27 Gaseous HCN
was synthesized from NaCN and stearic acid at 80–100 °C
under vacuum.41–44 The [PPN]X salts were cooled to 0 °C and
an excess (approx. 16–19 eq.) of cooled liquid HCN was trans-
ferred onto the salts. This led to dissolution of the pseudoha-
lides when the mixtures were shaken gently, except for the
OCP salt, where formation of a suspension was observed.
Furthermore, in case of the [PPN]OCN and [PPN]OCP, fast pro-
gressing polymerization/oligomerization of the liquid layer was
observed, which led to a darkening and an increase in the vis-
cosity of the phase. In contrast, the HCN mixtures of the [PPN]
N3 and [PPN]SCN remained nearly colorless. Since all these
reactions were carried out without further solvent, the highly
concentrated reaction medium can be considered as a mixture
with the properties of an ionic liquid.6,31 Since we were not
able to obtain crystals from any slowly cooled reaction mixture
in HCN due to the formation of different solvate anions, a
rather unconventional crystallization method was established
as we noticed fast crystallization of the reaction mixture upon
addition of non-polar perfluoropolyether. To this end,
Fomblin YR-1800 perfluoropolyether was placed in a self-built
low-temperature device for crystal picking, cooled with liquid
nitrogen to −60 °C. Subsequently, the HCN mixtures were
cooled until they solidified (for specifications see ESI†) and
Fig. 2 Ball-and-stick representation of the [OCN(HCN)3]− ion in
a
crystal of [PPN][OCN(HCN)3]. Due to disorder (a layer 52% and b layer
48%), two different distorted Y-shaped isomers, which are formed by
three hydrogen bridges (a–c), can be considered. Top: O-bound dimer
2O (2 O–H-bridges and 1 N–H-bridge), bottom: N-bound dimer 2N (2
N–H-bridges and 1 O–H-bridge). In each case a chain is formed when
short contacts (d–e) are considered. The cations are omitted for clarity.
Selected distances are given in Å, heavy atom distances in the H bridges:
(top) O1a–C2 3.06(2), O1a–C3 2.95(2), N1a–C4 2.97(3). a = 2.13, b =
2.06, c = 2.02. (Bottom) N1b–C2 2.95(3), N1b–C3 3.16(3), O1b–C4 2.89
(3). a = 2.00, b = 2.24, c = 1.95. Shortest interionic distances: (both) d =
N4–C3’ 3.379(7), e = C4–N3’ 3.68(1).
then transferred into the ether. Interestingly, in all cases, even [PPN][OCN(HCN)3] (2, Fig. 2) and [PPN][SCN(HCN)2] (3, Fig. 3,
at those low temperatures, release of gaseous HCN was Scheme 2).
observed within one minute, accompanied with the formation
In case of [PPN]OCP (see ESI Fig. S11† for X-ray structure
of X-ray suitable crystals. Structure elucidation unequivocally elucidation), a reaction with the HCN was observed, which led
revealed the formation of [PPN]+-salts containing the pseudo- to the degradation of the phosphaethynolate anion, finally
halide HCN-aggregate anions [PPN][N3(HCN)3] (1, Fig. 1), affording [PPN][P(CN·HCN)2] (4, Fig. 4). This is an exergonic
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
Fig. 4 Left: Ball-and-stick representation of the [P(CN·HCN)2]− ion in
the crystal of [PPN][P(CN·HCN)2]. The bent molecular anion is formed by
two hydrogen bridges (a and b). A centrosymmetric dimer (finally a
layer) is formed when further contacts (c and d) are considered.
Selected distances in Å and selected bond angles in °: P1–C1 1.764(4),
P1–C2 1.768(3), C1–N1 1.158(5), C2–N2 1.157(5), N1–C3a 3.088(7), N2–
C4 3.112(5). a = 2.29, b = 2.17. C1–P1–C2 94.3(2), P1–C1–N1 174.1(4).
Shortest interionic distances: c = C2–N4’ 3.906(5), d = N2–C4’ 3.439(5).
Right: Ball-and-stick (anions) and wireframe (cations) representation of
the molecular structure of [PPN][P(CN·HCN)2] in the crystal along the
b-axis. The [P(CN·HCN)2]− ions form a layer in the a–b-plane. Disorder
(4% probability) is not shown for clarity (see ESI Fig. S10†).
Fig. 3 Left: Ball-and-stick representation of a separated [SCN(HCN)2]−
ion in the crystal of [PPN][SCN(HCN)2], whereas the cations are omitted
for clarity. The distorted L-shaped molecular anion is formed by two
hydrogen bridges (a and b). Selected distances in the H-bridges in Å and
selected bond angles in °: S1–C2 3.452(3), N1–C3 3.060(5). a = 2.18, b =
2.50. C1–S1–C2 95.83(9), C3–N1–S1–C2 78.2(3). Right: Ball-and-stick
(anions) and wireframe (cations) representation of the molecular struc-
ture of [PPN][SCN(HCN)2] in the crystal. View along the a-axis.
rately by X-ray diffraction, the distances of the heavier atoms
(A, B) in the A⋯H–B-bridges will be discussed. All ORTEP rep-
resentations can be found in the ESI.† 1 crystallized in the
monoclinic space group C2/c with four formula units per cell
(Fig. 1).
The unit cell contains well-separated cations and anions.
Distorted Y-shaped, trisolvate anions are formed due to three
C–H⋯N hydrogen bridges. Two solvate molecules coordinate
to the Nα of the central azide (H-bridges a and b) with heavy
atom distances of 3.02(2) and 3.033(8) Å, whereas the third
HCN is linked to the Nγ atom (d(N1⋯C1) = 2.97(1) Å, H-bridge
c in Fig. 1). Interestingly, there are head-to-tail arrangements
between the HCN groups of two adjacent Y-shaped solvate
anions but rather long distances of 3.391(6) (d) and 3.70(2) Å
(e, cf. dvdW(C⋯N) = 3.25 Å)45 are observed. Hence, the formed
anionic strands along the c-axis, might be better explained due
to packing effects in the crystal. Due to the HCN-solvation, the
N3 anion is slightly bent (∡(N–N–N) = 177.5(6)°) and two
slightly different N–N distances (1.173(6) and 1.136(10) Å) were
found.
Scheme 2 Synthesis of [PPN][X(HCN)n] (X = SCN for n = 2; X = N3,
OCN for n = 3).
process according to computations (ΔrG° = −35.8 kcal mol−1
see below, Table S12†). So far, the formation of a possible by-
product, such as formaldehyde (PCO−
HCN [P
,
+
4
→
(CN·HCN)2] + H2CO), could not be observed by NMR analysis.
However, we want to emphasize that all above discussed reac-
tions and crystallizations were carried out several times, always
leading to the discussed products. According to Raman spec-
troscopy, all solvates 1–4 are very labile with respect to moist-
ure and loss of HCN was observed when the compounds were
warmed-up to room temperature. Since HCN is the weakest
acid amongst all considered HX species, always only the [X
(HCN)n]− ions, with the proton attached to the CN− ion, were
observed in accord with computations (Table S12†).
2 crystallizes isostructural to 1 (C2/c) and again well-separ-
ated cations and distorted Y-shaped anions are formed. Due to
the disorder of the central OCN− ion, two isomers have to be
considered. An O-linked isomer 2O (Fig. 2, top) and a
N-connected isomer 2N (Fig. 2, bottom) with an occupation
probability of 52(1)% and 48(1)%, respectively, which are the
Structure elucidation
All crystals were selected at −60 °C under inert conditions (N2 same considering the esd. In 2O, two C–H⋯O hydrogen
stream). A point fume cupboard was used because highly toxic bridges are formed from two HCN groups, which are attached
HCN is released from the reaction mixture. Since the exact to the oxygen of the cyanate ion (a: d(O1a⋯C2) = 3.06(2) and b:
position of the hydrogen atoms cannot be determined accu- d(O1a⋯C3) = 2.95(2) Å) and only one HCN, which is bound to
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
the terminal N atom, forming a C–H⋯N hydrogen bridge (c: Raman spectroscopy
d(N1a⋯C4) = 2.97(3) Å, Fig. 2). In 2N, it is exactly the other way
The Raman data of all compounds were obtained from cooled
single crystals, which were used for the X-ray structure elucida-
tion. It was very difficult to obtain reasonable spectra. In all
cases, no bands for νC–H,HCN vibration of the molecular anions
of 1–4 were observed due to the highly dynamic nature of the
hydrogen-bridged systems, which leads to a broadening of the
bands. Moreover, C–H vibrational bands of the cation were
also found in the relevant area.
around: there are two C–H⋯N (a: d(N1b⋯C2) = 2.95(3) and b:
d(N1b⋯C3) = 3.16(3) Å) and one C–H⋯O hydrogen bridges (c:
d(O1b⋯C4) = 2.89(3) Å, Fig. 2). Again, both isomers form
anionic strands along the c-axis due to head-to-tail arrange-
ments, attributed to packing effects in the crystal (d:
d(C4⋯N3′) = 3.379(7) and e: d(N4⋯C3′) = 3.68(1) Å, Fig. 2), of
two hydrogen cyanides with adjacent [OCN(HCN)3]− anions.
3 crystallizes in the monoclinic space group P21 with two
formula units per cell (Fig. 3, right). The [SCN(HCN)2]− anion
is formed by one C–H⋯N hydrogen bridge (a: d(N1⋯C2) =
3.060(5) Å) and one C–H⋯S hydrogen bridge (b: d(S1⋯C3) =
3.452(3) Å, Fig. 3, left). The HCN, attached to the sulphur of
the thiocyanate anion, is linked nearly orthogonal with ∡
(C2⋯S1–C1) = 95.83(9)°, thus leading to formation of a dis-
torted L-shaped disolvate anion. In contrast to the crystal
structures of 1 and 2, no larger voids are formed by the bulky
[PPN]+ cation. Hence, only separated [SCN(HCN)2]− ions are
found, which are located in a pocket, formed by the surround-
ing [PPN]+ cations (Fig. 3, right).
For the trisolvate species, 1 and 2, broad bands for the νCN
stretching modes were found in the expected region for all
three coordinated HCN molecules (1: 2056(br)/2067, 2: 2060/
2073 cm−1, Table 2; cf. 2097 for pure HCN at 233 K; 2057(br)/
2081 and 2055(br)/2071 cm−1 for [CN(HCN)2]− and [CN
(HCN)3]−, respectively),31 however, no splitting for the in-phase
(symmetric) and out-of-phase (antisymmetric) vibrations could
observed. Furthermore, the antisymmetric stretching vibration
of the N3 ion in 1 was detected at 2077 cm−1, while the sym-
−
metric mode was found at 1332 cm−1 with the largest intensity
(cf. [PPN]N3 1323 cm−1). The cyanide mode, νCN,OCN, was
observed at 2164 and for the SCN− ion at 2073 cm−1. For dicya-
[PPN][P(CN·HCN)2] 4 crystallizes in the monoclinic space
group P21/c. Layers of cations and anions are formed in the
crystal (Fig. 4, right). Structural parameter of the bent, but
planar [P(CN·HCN)2]− in comparison with crystallographic
data of the salts, bearing the unsolvated [P(CN)2]− ion, are
listed in Table 1, indicating that solvation does not signifi-
cantly change these structural parameters. However, as dis-
cussed before, interionic interactions, which arise from hydro-
gen bonding as well as CN head-to-tail arrangements (Fig. 4),
lead to the formation of a layer type anion structure in contrast
to the structural motifs found in unsolvated [Cat][P(CN)2]
([Cat] = Na/K[18]crown-6,46,47 [Ph4P]48). Due to disorder, two
different anion layers are found. In the main layer (a layer with
96%), each nitrogen of the dicyanophosphide anion is
attached to one HCN molecule by one C–H⋯N hydrogen
bridge, which results in the formation of the disolvate anion
[P(CN·HCN)2]− (a: d(N1⋯C3a) = 3.088(7) and b: d(N2⋯C4) =
3.112(5) Å, Fig. 4, left). The [P(CN·HCN)2]− ions are further
linked by very weak van der Waals interactions between two
adjacent anions (c: d(C2⋯N4′) = 3.906(5) and d: d(N2⋯C4′) =
3.439(5) Å. Considering the b layer (4(1)%), with one dis-
ordered HCN moiety, a monosolvate anion [P(CN·HCN)(CN)]−
is formed, whereby the second HCN remains uncoordinated in
the cavity (see ESI Fig. S10†).
nophosphide 4, the symmetric νPC–N,s (antisymmetric νPC–N,as
)
stretching vibration of the phosphorus bound cyanide groups
could be observed at 2105 (2084) cm−1 (cf. [PPh4][P(CN)2]:
2111, 2089 cm−1).48 Additionally, two bands arising from the
coupled CN-vibrations of the two solvated HCN molecules
were assigned to the bands at 2022 (in-phase) and 2025 (out-
of-phase) cm−1, respectively.
Computational results
DFT calculations (pbe0/aug-cc-pVTZ + dispersion corrections)
were performed (see ESI Tables S2–S13†), to obtain additional
structural und thermodynamic data for the isolated pseudoha-
lide-HCN solvates of 1–3 and for the dicyanophosphide HCN-
solvate 4. The calculated bond lengths for the isolated anions
in the gas phase are in good agreement with the observed
bond lengths in all crystallized compounds (cf. ESI
Table S13†). However, the geometry (angles and dihedral
angles) of the solid state structures 1–4 slightly deviates, which
can be attributed to very weak energy potentials, a problem
that was already found for other mixed aggregates of the type
[XHY]− (X, Y = F, Cl, Br, CN).49,50 In agreement with our experi-
Table 2 Selected Raman data obtained from single crystals of 1–4.
Frequencies in cm−1 a
νCN,2HCN-solvate
(sym/anti)
νXYZ,
Table 1 Selected structural data (distance in Å, bond angle in °) of salts,
containing the [P(CN)2]− anion
Compound
νCN,HCN-solvate
pseudohalideg
1
2
3
4
2056b
2060b
—
2067
2077(anti)/1332(sym)c
2164
2073
Cation
d(P–C)
d(C–N)
∡(C–P–C)
2051d/2059e
—
2073
Naa,46
Ka,47
1.76(2)–1.72(2)
1.67(1)–1.69(1)
1.80(1)–1.755(6)
1.764(4)–1.766(3)
1.155
95(1)
2022/2025
2105(sym)/2084(anti)f
1.15(2)–1.14(2)
1.20(1)–1.108(8)
1.157(5)–1.158(5)
95.2(6)
92.7(3)
94.3(2)
a Tentative assignment based on comparison with computed data.
b Broad signal. c Only most intense band given. d Connected to N.
e Connected to S. f P-bounded cyanide. g XYZ for 1: NNN, 2: NCO, 3:
NCS, and 4: PCN.
[Ph4P]48
[PPN]b
a [18]crown-6 salt. b As HCN disolvate.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
mental data and in accord with the gas phase acidities, HCN
represents the weakest acid. That is, protonation of X− (N3,
NCO, NCS and PCO) is endergonic with 6.7–24.1 kcal mol−1
(Table S11†). This means that those isomers are always the
most stable in which Z⋯HCN hydrogen bonds are found (Z =
terminal atoms of X− ion).
Due to the flat potentials as well as many possible connec-
tivities, often a large series of computations, including a
variety of isomers, had to be considered, which will not be dis-
cussed in detail here (see ESI, Tables S2–S10†). For example,
22 different isomers for the [N3(HCN)3]− ion were found. Of
these, the Y-shaped solvate YT1 with three C–H⋯N H-bridges
is the energetically favoured isomer (Fig. 5), which agrees well
with the experimental results (cf. Fig. 1). Already the isomeriza-
tion of only one HCN moiety to HNC (YT2) leads to a rise in
energy of 8.24 kcal mol−1. Furthermore, two additional bent
chain-like isomers BT1 and BT2 were found, which are only
2.97 and 6.26 kcal mol−1 less favoured, respectively.
Interestingly, proton migration during the optimization was
observed (BT2, for further isomers see ESI Table S2†), when
the azide anion is located at the end of the chain. This leads to
formation of hydrogen azide (HN3) and a cyanide ion (CN−),
which is linked by one N–H⋯C and one C–H⋯N hydrogen
bridge. The stepwise formation of the HCN solvate
[N3(HCN)n]− for n = 1–3 (see ESI, Table S3†) were computed to
be exothermic and exergonic, but with less negative values
(ΔnH2°98 : −22.7, −17.4, −13.4; ΔnG2°98 : −16.0, −10.1, −7.1 kcal
mol−1) when the degree of HCN-aggregation increases (n =
1–3).
Fig. 6 Selected computed isomers of the trisolvate ion [OCN(HCN)3]−
with the relative total energies (pbe0/aug-cc-pVTZ, in italics) in kcal
mol−1
.
migration (BT4, for further isomers see ESI Table S4†) was
observed during the optimization process, forming a hydrogen
cyanate and a cyanide ion, which is linked via one N–H⋯C
and one C–H⋯N hydrogen bridge. All reaction energies for the
stepwise formation of both trisolvates YT1 and YT2 were found
to be exothermic and exergonic for n = 1–3 (YT1: ΔnH2°98
:
;
−24.6, −17.8, −12.7; ΔnG°298 = −17.6, −9.8, −6.7 kcal mol−1
YT2: ΔnH2°98 : −24.6, −15.8, −12.5; ΔnG2°98 = −17.6, −9.0,
−4.7 kcal mol−1, Tables S5 and S6†).
18 isomers were found for the disolvate of [SCN(HCN)2]−,
(Table S7†). In contrast to the experimental results, the calcu-
lated isomer with the lowest energy (YD1, Fig. 7) features a
Y-shaped structure, however the experimentally observed bent
L-shaped isomer BD2 is only 0.58 kcal mol−1 higher in energy,
which might be compensated due to crystal packing (cf.
Fig. 3). BD2 exhibits a similar shape as the observed [SCN
(HCN)2]− ion in 3, however, the gas phase structure deviates
slightly from the experimental geometry, especially the C2–S1–
N1–C3 dihedral angle with 179.9° (cf. [SCN(HCN)2]− in 3 with
78.2(3)°). When the SCN− anion is located at the end of the
disolvate HCN⋯HCN, two additional types of structures are
found, depending on whether the hydrogen cyanide is con-
nected by a C–H⋯N (LD1, linear) or C–H⋯S (BD1, L-shaped)
H-bridge. Beside these structures, a C1-symmetric, distorted
trigonal pyramidal-shaped isomer PD1 with two C–H⋯S hydro-
gen bridges was found, which is 4.47 kcal mol−1 less favoured.
42 isomers were found for the [OCN(HCN)3]− trisolvate ion
(Table S4†). As mentioned before, two isomers have to be con-
sidered according to the results from X-ray diffraction. The
energetically preferred isomers (YT1, Fig. 6) exhibits a non-
planar structure and is built by two C–H⋯N and one C–H⋯O
hydrogen bridges, as found for the observed isomer 2N (cf.
Fig. 2). The planar Y-shaped isomer YT2, which shows the
same connectivity as observed for isomer 2O, was found to be
only 2.08 kcal mol−1 less favoured in the gas phase. This
might explain the observation of both isomers in the solid
state (52(1)% for 2O and 48(1)% for 2N according to X-ray crys-
tallography). In addition, further bent chain-like structures
were found, in which the OCN anion either shows a 1,3-O,N-
coordination mode (BT1 and BT2) or is only linked at one term-
inal atom of the OCN− ion (O: BT3 and N: BT4). Again, proton
Fig. 5 Selected computed isomers of the trisolvate ion [N3(HCN)3]−
Fig. 7 Selected computed isomers of the disolvate ion [SCN(HCN)2]−
with the relative total energies (pbe0/aug-cc-pVTZ, in italics) in kcal
with the relative total energies (pbe0/aug-cc-pVTZ, in italics) in kcal
mol−1
.
mol−1
.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
[PPN][OCN(HCN)3],
[PPN][SCN(HCN)2]
and
[PPN][P
(CN·HCN)2], starting from [PPN]X (X = N3, OCN, SCN, OCP)
and HCN. The molecular anions in 1–3 represent the first
members of solvate anions of the type [X(HY)n]−, in which X
and Y are different pseudohalides. In all cases, the HCN sol-
vates were formed since HCN is the weakest acid amongst the
considered pseudohalide acids HX; therefore, no protonation
of the pseudohalide X− was found. These temperature labile
compounds were characterized by X-ray diffraction, Raman
Fig. 8 Selected computed isomers of the disolvate ion [P(CN·HCN)2]−
with the relative total energies (pbe0/aug-cc-pVTZ, in italics) in kcal
spectroscopy
and
quantum
chemical
calculations.
mol−1
.
Experimentally, it was shown that the trisolvates, [N3(HCN)3]−
and [OCN(HCN)3]−, form distorted Y-shaped molecular
anions, while the HCN disolvate [SCN(HCN)2]− shows a dis-
torted L-shaped geometry. Moreover, it was observed that the
central cyanate in the molecular anion in 2 is disordered, so
two different isomers, 2O with two C–H⋯O H-bridges and 2N
with two C–H⋯N H-bridges, have to be considered. The phos-
phaethynolate anion, OCP−, was found to react with HCN,
leading to formation of the dicyanophosphide anion in a
hitherto unknown decomposition process. Its [PPN]+-salt 4
crystallized in form of a HCN disolvate, forming a bent [P
(CN·HCN)2]− anion. In accord with experimental structural
data, the computational results show that a Y-shaped structure
of the [N3(HCN)3]−, [OCN(HCN)3]−, and [SCN(HCN)2]− anion is
always the energetically slightly preferred isomer. However,
due to very flat potentials, distortions from the ideal structure
were found in the solid state, which can be attributed to small
solid state effects (especially for the [SCN(HCN)2]− anion with
an energy difference of ΔErel. = 0.6 kcal mol−1 between the two
energetically lowest lying isomers).
The stepwise formation (n = 1–2) of the energetically favoured
[SCN(HCN)2]− anion YD1 (BD2) was again calculated to be
exothermic and exergonic for all steps (ΔnH2°98 = −19.4 (−19.4),
−14.0 (−13.5); ΔnG°298 : −12.0 (−12.0), −7.5 (−7.2) kcal mol−1
)
and even a possible but not observed aggregation of a third
HCN was found to be energetically favoured (Δ3H2°98 = −11.1,
Δ3G°298 = −4.9 kcal mol−1, Table S8†).
As depicted in Fig. 8, the C2v-symmetric HCN disolvate [P
(CN·HCN)2]−, BD1, was found to be the energetically preferred
isomer. Like the solid-state structure for the anion in 4 (cf.
Fig. 4), the computed species exhibits a planar, bent geometry
and is associated to two HCN molecules by two C–H⋯N hydro-
gen bridges. Two further isomers were found, one bent struc-
ture with Cs-symmetry, BD2 (ΔErel. = +3.6 kcal mol−1), exhibit-
ing an almost linear HCN⋯HCN chain, and a Cs-symmetric,
bisphenoidal isomer, DD1 (ΔErel = +6.27 kcal mol−1), with two
C–H⋯P H-bridges, which form the axis of the pseudo-trigonal
bipyramid with the P-atom in the centre (Fig. 8, Table S10†).
From the computational study, some features can be gener-
alized: (i) the energetically favoured isomer, is the one with the
maximum number of covalent NC–H bonds (of the HCN
solvate molecule) and the largest number of C–H⋯N H-bridges.
(ii) If isonitrile groups are present in the structure, the energy of
the isomer increases significantly. (iii) All anion structures
observed in the X-ray structure elucidations represent the energeti-
cally most favoured isomers from gas phase calculations, except
for the [SCN(HCN)2]− solvate ion, which possesses two more or
less energetically equal isomers YD1 and BD2, with BD2, the experi-
mental structure, being slightly less favoured (ΔErel. = 0.58 kcal
mol−1) at the theoretical method applied. (iv) The theoretically
determined distances of the heavier atoms in the hydrogen bridges
C–H⋯O (d(C⋯O) = 2.66–2.95 Å) are usually shorter than the C–
H⋯N hydrogen bonds (d(C⋯N) = 2.77–3.11 Å), while the C–H⋯S
distances (d(C⋯S) = 3.29–3.41 Å) are the longest (only isomers with
covalent NC–H bonds and isomers without proton migration con-
sidered), in accord with the experimentally determined distances.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
We thank the University of Rostock for access to the high-per-
formance computing facilities, and especially Malte Willert for
his assistance with the queueing system and software installa-
tions. Furthermore, we wish to thank the DFG (SCHU 1170/9-
2), especially the SPP 1708, for financial support.
Notes and references
1 L. Birckenbach and K. Kellermann, Ber. Dtsch. Chem. Ges.,
1925, 58, 786–794.
2 L. Birckenbach and K. Kellermann, Ber. Dtsch. Chem. Ges.,
1925, 58, 2377–2386.
3 L. Birckenbach and K. Huttner, Ber. Dtsch. Chem. Ges.,
1929, 62, 153–163.
4 L. Birckenbach and M. Linhard, Ber. Dtsch. Chem. Ges.,
1929, 62, 2261–2277.
Conclusions
A low temperature crystallization technique in a rather uncon-
ventional crystallization medium (Fomblin YR-1800) was used
to generate and isolate single crystals of [PPN][N3(HCN)3],
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
5 L. Birckenbach and K. Huttner, Z. Anorg. Allg. Chem., 1930, 28 S. Nuzzo, B. Twamley, J. A. Platts and R. J. Baker, Chem.
190, 38–47. Commun., 2016, 52, 13296–13298.
6 J. Harloff, D. Michalik, S. Nier, A. Schulz, P. Stoer and 29 B. Neumüller, F. Schmock and K. Dehnicke, Z. Anorg. Allg.
A. Villinger, Angew. Chem., Int. Ed., 2019, 58, 5452–5456. Chem., 1999, 625, 1243–1245.
7 V. V. Stopenko, A. M. Golub and H. Köhler, Chemistry of 30 J. A. Salthouse and T. C. Waddington, J. Inorg. Nucl. Chem.,
Pseudohalides, Elsevier, Amsterdam, 1986. 1978, 40, 1696–1698.
8 Z. M. Smallwood, M. F. Davis, J. G. Hill, L. J. R. James and 31 K. Bläsing, J. Harloff, A. Schulz, A. Stoffers, P. Stoer and
P. Portius, Inorg. Chem., 2019, 58, 4583–4591.
9 E. M. Frémy, Ann. Chim. Phys., 1856, 3, 5–50.
10 R. M. Bozorth, J. Am. Chem. Soc., 1923, 45, 2128–2132.
A. Villinger, Angew. Chem., Int. Ed., 2020, 59, 10508–10513.
32 J. W. Nibler and G. C. Pimentel, J. Chem. Phys., 1967, 47,
710–717.
11 R. Kruh, K. Fuwa and T. E. McEver, J. Am. Chem. Soc., 1956, 33 M. Meot-Ner, S. M. Cybulski, S. Scheiner and J. F. Liebman,
78, 4256–4258. J. Phys. Chem., 1988, 92, 2738–2745.
12 S. I. Troyanov, I. V. Morozov and E. Kemnitz, Z. Anorg. Allg. 34 D. H. McDaniel and R. E. Valleé, Inorg. Chem., 1963, 2,
Chem., 2005, 631, 1651–1654. 996–1001.
13 S. I. Ivlev, T. Soltner, A. J. Karttunen, M. J. Mühlbauer, 35 F. Y. Fujiwara and J. S. Martin, J. Am. Chem. Soc., 1974, 96,
A. J. Kornath and F. Kraus, Z. Anorg. Allg. Chem., 2017, 643,
1436–1443.
7625–7631.
36 J. C. Evans and G. Y. S. Lo, J. Phys. Chem., 1966, 70, 20–25.
14 D. Mootz, U. Ohms and W. Poll, Z. Anorg. Allg. Chem., 1981, 37 S. A. Chacko, I. H. Krouse, L. A. Hammad and
479, 75–83. P. G. Wenthold, J. Am. Soc. Mass Spectrom., 2006, 17, 51–55.
15 D. Mootz and W. Poll, Z. Anorg. Allg. Chem., 1982, 484, 158– 38 J. A. Salthouse and T. C. Waddington, J. Chem. Soc., 1964,
164.
1, 4664–4666.
16 D. Mootz and W. Poll, Z. Naturforsch., 1984, 39, 290–297.
17 D. Mootz and D. Boenigk, J. Am. Chem. Soc., 1986, 108,
6634–6636.
18 D. Mootz and D. Boenigk, Z. Anorg. Allg. Chem., 1987, 544,
159–166.
39 C. J. Price, H.-Y. Chen, L. M. Launer and S. A. Miller,
Angew. Chem., Int. Ed., 2009, 48, 956–959.
40 L. Mann, E. Hornberger, S. Steinhauer and S. Riedel, Chem.
– Eur. J., 2018, 24, 3902–3908.
41 K. Bläsing, J. Bresien, R. Labbow, A. Schulz and
A. Villinger, Angew. Chem., Int. Ed., 2018, 57, 9170–9175.
19 R. Rathore, S. V. Lindeman, C. J. Zhu, T. Mori,
P. V. R. Schleyer and J. K. Kochi, J. Org. Chem., 2002, 67, 42 R. Labbow, D. Michalik, F. Reiß, A. Schulz and A. Villinger,
5106–5116. Angew. Chem., Int. Ed., 2016, 55, 7680–7684.
20 D. Mootz and A. Deeg, Z. Anorg. Allg. Chem., 1992, 615, 43 A. Hinz, R. Labbow, C. Rennick, A. Schulz and
109–113.
J. M. Goicoechea, Angew. Chem., Int. Ed., 2017, 56, 3911–
21 J. L. Atwood, S. G. Bott, C. M. Means, A. W. Coleman,
3915.
H. Zhang and M. T. May, Inorg. Chem., 1990, 29, 467– 44 P. Günther, R. Mexer and F. Muller-Skjord, Z. Phys. Chem.,
470. 1935, 175, 154–169.
22 D. Mootz and J. Hocken, Z. Naturforsch., 1989, 44, 1239– 45 M. Mantina, A. C. Chamberlin, R. Valero, C. J. Cramer and
1246. D. G. Truhlar, J. Phys. Chem., 2009, 113, 5806–5812.
23 S. Hasenstab-Riedel, P. Voßnacker, S. Steinhauer and 46 W. S. Sheldrick, J. Kroner, F. Zwaschka and A. Schmidpeter,
J. Bader, Chem.
chem.202001864.
24 L. W. Schroeder and J. A. Ibers, Inorg. Chem., 1968, 7, 594–
599.
25 G. W. Driver and I. Mutikainen, Dalton Trans., 2011, 40,
10801–10803.
–
Eur. J., 2020, DOI: 10.1002/
Angew. Chem., Int. Ed. Engl., 1979, 18, 934–935.
47 A. Schmidpeter, G. Burget, F. Zwaschka and
W. S. Sheldrick, Z. Anorg. Allg. Chem., 1985, 527, 17–32.
48 J. F. Binder, S. C. Kosnik, P. B. J. St Onge and
C. L. B. Macdonald, Chem. – Eur. J., 2018, 24, 14644–14648.
49 J. W. Larson and T. B. Mcmahon, Inorg. Chem., 1984, 23,
2029–2033.
26 M. F. A. Dove, Chem. Commun., 1965, 23–24.
27 S. Arlt, J. Harloff, A. Schulz, A. Stoffers and A. Villinger, 50 M. Meot-Ner, S. M. Cybulski, S. Scheiner and J. F. Liebman,
Inorg. Chem., 2019, 58, 5305–5313.
J. Phys. Chem., 1988, 92, 2738–2745.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.