Binary polyazides of cadmium and mercury

Article abstract of DOI:10.1002/chem.201406023

Following our interest in binary element-nitrogen compounds we report here on the synthesis and comprehensive characterization (M.p., IR/Raman, elemental analysis, ¹⁴N/¹³³Cd/¹⁹⁹Hg NMR) of tri-and tetraazido cadmate and mercurate anions [E(N₃)_(2+n)]^n- (E = Cd, Hg; n = 1, 2) in a series of [Ph₄P]⁺ and [PNP]⁺ ([PNP]⁺ = bis(triphenylphosphine)-iminium) salts. The azide/chloride exchange in CH₂Cl₂ as well as the formation of tetrazolate salts in CH₃CN solutions of the polyazido mercurates were investigated. Single crystal X-ray structures of all new compounds, and for comparison [Ph₄P][Cd₂(N₃)₅(H₂O)], were determined. Moreover, the synthesis of anhydrous cadmium(II) azide and its DMSO adduct is presented for the first time. For a better understanding of structure and bonding in E(N₃)₂, [E(N₃)₃]^- and [E(N₃)₄]^2-, theoretical calculations at the M06-2X/aug-cc-pVDZ level were carried out.

Full text of DOI:10.1002/chem.201406023

DOI: 10.1002/chem.201406023
Full Paper
&
Heavy Metal Azides |Hot Paper|
Binary Polyazides of Cadmium and Mercury
Axel Schulz[a, b] and Alexander Villinger*[a]
Abstract: Following our interest in binary element–nitrogen
compounds we report here on the synthesis and compre-
hensive characterization (M.p., IR/Raman, elemental analysis,
the polyazido mercurates were investigated. Single crystal X-
ray structures of all new compounds, and for comparison
[Ph4P][Cd2(N3)5(H2O)], were determined. Moreover, the syn-
thesis of anhydrous cadmium(II) azide and its DMSO adduct
is presented for the first time. For a better understanding of
structure and bonding in E(N3)2, [E(N3)3]ꢀ and [E(N3)4]2ꢀ,
theoretical calculations at the M06-2X/aug-cc-pVDZ level
were carried out.
14N/133Cd/199Hg NMR) of tri- and tetraazido cadmate and mer-
nꢀ
curate anions [E(N3)(2+n)
]
(E=Cd, Hg; n=1, 2) in a series of
[Ph4P]+ and [PNP]+ ([PNP]+ =bis(triphenylphosphine)-
iminium) salts. The azide/chloride exchange in CH2Cl2 as well
as the formation of tetrazolate salts in CH3CN solutions of
Introduction
which frequently detonates spontaneously during crystal
growth.[5]
Like most binary transition metal–nitrogen compounds, mercu-
ry and cadmium azides represent a class of highly endothermic
compounds. The difficulties in the isolation and handling of
such nitrogen-rich compounds arise from their highly endo-
thermic character, leading often to explosive decomposition.[1]
Nevertheless, as early as 1890 Curtius reported on the isolation
of mercury(I) azide Hg2(N3)2 by combining aqueous mercury(I)
nitrate or elemental mercury with hydrazoic acid HN3, while
mercury(II) azide a-Hg(N3)2 was first prepared by Berthelot
et al. in the year 1894 by the reaction of mercury(II) oxide and
HN3 in aqueous solution.[2a,b]
Moreover, the nitridodimercury azide [Hg2N]N3, the azide of
Millon’s base, which features a three-dimensional cationic
[Hg2N]+ framework structure incorporating the azide ions in
the cavities, could be isolated in the reaction of a-Hg(N3)2 with
aqueous ammonia by our group.[2a,5,6]
Numerous organomercury(II) azides RHgN3 including alkyl,
aryl and perfluorinated derivatives are known and were
thoroughly investigated.[7] However, only lately Klapçtke et al.
observed that organomercury(II) azides easily reacted in a 1,3-
dipolar cycloaddition reaction with organonitriles forming tet-
razole rings.[7g] Thereby, the reaction is regioselective and even
non-activated (electron rich) nitriles such as acetonitrile could
be applied. All binary mercury azides were comprehensively
studied by vibrational spectroscopy (IR/Raman),[3a,b,5] spectro-
photometric methods,[4a] X-ray powder diffraction,[3b,5] single-
crystal X-ray diffraction,[4b,5] and thermal analysis.[3c–e,5] On the
other hand, only little is known about polyazido mercurates.
While the tetraazido mercurate anion [Hg(N3)4]2ꢀ is unknown in
literature, Beck et al. described the triazido mercurate anion
[Hg(N3)3]ꢀ in the salt [Ph4P][Hg(N3)3], which could be character-
ized by vibrational methods, however, structural data is not
available.[8]
The first vibrational characterization of both azides could be
carried out by Dehnicke et al. in 1968.[3a] Nowadays, both
mercury azides are more conveniently obtained by the preci-
pitation from aqueous solutions of mercury(I) or (II) nitrate
with sodium azide. After washing and drying in air, Hg2(N3)2
and a-Hg(N3)2 are obtained as pure, colorless microcrystalline
solids in almost quantitative yields.[3a,4] In addition, b-Hg(N3)2,
a second modification of mercury(II) azide, can be prepared by
slow diffusion of aqueous NaN3 solution into a solution of
mercury(II) nitrate which is separated by a layer of aqueous
NaNO3.[3c,4c] Only recently, we reported on the successful
isolation and characterization of this highly labile polymorph,
The first synthesis of cadmium(II) azide dates back to the
year 1898, when Curtius and Rissom reported on the reaction
of cadmium(II) carbonate and excess aqueous HN3 affording
pale yellow crystals, which were characterized as insensitive to
impact (Scheme 1).[9a] Later, Birckenbach and Wçhler et al.
could demonstrate that pure colorless Cd(N3)2 is highly
sensitive to impact and shock.[3e,9b] Recently, Schnick et al. suc-
ceeded in the first vibrational and structural characterization of
cadmium(II) azide.[10] While the former reaction never led to
detonations during the preparation of cadmium(II) azide,[9,3c,10]
Bassiꢀre reported an alternative procedure starting from
cadmium(II) nitrate and sodium azide, which frequently result-
[a] Prof. Dr. A. Schulz, Dr. A. Villinger
Institut fꢀr Chemie, Universitꢁt Rostock
Albert-Einstein-Strasse 3a, 18059 Rostock (Germany)
E-mail: alexander.villinger@uni-rostock.de
[b] Prof. Dr. A. Schulz
Abteilung Materialdesign
Leibniz-Institut fꢀr Katalyse e.V. an der Universitꢁt Rostock
Albert-Einstein-Strasse 29a, 18059 Rostock (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201406023.
Chem. Eur. J. 2015, 21, 3649 – 3663
3649
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ed in spontaneous explosions of the crystal growth solution,[11a]
later confirmed by Reckeweg and Simon.[11b] Contrary to a-
Hg(N3)2, cadmium(II) azide is hygroscopic and slowly decom-
poses in air forming yellow basic products. Nevertheless, an
anhydrous synthesis has not been reported so far.[9a,11a,10] Cad-
mium(II) azide reacts readily with aqueous ammonia, forming
the stable complex Cd(N3)2(NH3)2 which shows no further con-
densation as compared to mercury(II) azide (vide supra).[6] Ac-
cordingly, many adducts of cadmium(II) azide with bases such
as pyridine or ethylendiamine were prepared and structurally
characterized.[9a,6,12] In 1967, Beck et al. reported on the synthe-
sis of the first polyazido cadmate in the salt [Ph4P]
[Cd2(N3)5(H2O)].[8] Since no single-crystal X-ray data is available,
the constitution of the formal pentaazido dicadmate(II) anion
[Cd2(N3)5]ꢀ remains unknown. The first evidence for a triazido
cadmate anion was found by Clegg and Krischner in the struc-
ture of K[Cd(N3)3]·H2O.[13a] However, the former compound
shows several significant short potassium azide contacts and
therefore should rather be considered as K/Cd double salt. In
contrast, the [Cd(N3)3]ꢀ anion in the salt [Me4N][Cd(N3)3] pre-
pared by Mautner et al., forms a three-dimensional “perov-
skite-like” network consisting of Cd(N3)6 octahedra, which are
connected by m1,3-bridging azide units, while the [Me4N]+ cat-
ions reside in the large cages and are only weakly bound by
van der Waals interactions.[14] The only report for a tetraazido
cadmate anion can be found in the work of Krischner et al.,
who described the synthesis of M2[Cd(N3)4] (M=K, Tl) from
aqueous solutions of cadmium carbonate, HN3 and KN3 or TlN3,
respectively.[13b] While Tl2[Cd(N3)4] detonates on impact, the po-
tassium salt is almost insensitive. Interestingly, a slight excess
of TlN3 in the reaction led to the isolation of the formal
Tl8[Cd3(N3)14] salt (=Tl2[Cd(N3)4]·6Cd(N3)2·6TlN3).[15] Both thallium
salts could be structurally characterized and also show octahe-
drally coordinated cadmium(II) cations, surrounded by six azide
ligands.[13a] However, besides these structure determinations
and thermal analyses, the polyazido cadmates have scarcely
been characterized, and vibrational data is completely missing.
ambient temperature under exclusion of light. The complete
consumption of cadmium fluoride during reaction could easily
be monitored by powder X-ray diffraction.[16]
After centrifugation, washing and drying at 60–808C in
vacuo, pure Cd(N3)2 was obtained as colorless microcrystalline
solid in good yields.
Scheme 2. Synthetic routes to Cd(N3)2 and Cd(N3)2·3/2 DMSO.
Crystalline Cd(N3)2 is shock- and friction-sensitive and explo-
des violently on fast heating emitting a bright blue flashlight.
Cadmium(II) azide synthesized by this procedure is stable for
at least two years without noticeable decomposition if stored
under argon in the absence of light. Similar to a-Hg(N3)2,
Cd(N3)2 can be recrystallized from acetonitrile or water,
however, contrary to mercury(II) azide, which also can be ob-
tained solvent-free from DMSO, cadmium(II) azide crystallizes
as colorless DMSO solvate with the formal composition
Cd(N3)2·3/2 DMSO (Scheme 2, Figure 1). Differently to Cd(N3)2,
the DMSO solvate is not sensitive to friction or shock and is
even reasonably stable at 1208C in the molten state (vide
infra).[16]
Synthesis of [M]n[E(N3)(2+n)
]
As illustrated in Scheme 3, two synthetic routes seemed to be
feasible for preparing salts of polyazido cadmate and
mercurate anions [M]n[E(N3)(2+n)] ([M]+ =weakly coordinating
cation=[Ph4P]+, [PNP]+ =[(Ph3P)2N]+; E=Cd, Hg, with n=1,
2): 1) direct reaction of binary cadmium or mercury azide E(N3)2
with the corresponding phosphonium azide [M]N3 in the re-
quired stoichiometry; 2) iodide/azide exchange in the tri- and
tetraiodo cadmates or mercurates [M]n[EI(2+n)] by treatment
with silver azide AgN3. In the latter reaction, it is also possible
to generate the tetraiodo metallate anion [EI4]2ꢀ in situ by reac-
tion of triiodo metallate [M][EI3] salts with the corresponding
iodide salts [M]I.
nꢀ
Scheme 3. Different synthetic routes to [E(N3)(2+n)
2; [M]+ =[Ph4P]+, [PNP]+).
]
salts (E=Cd, Hg; n=1,
Scheme 1. Different synthetic routes to Hg2(N3)2, Hg(N3)2 and Cd(N3)2.
Results and Discussion
Synthesis of Cd(N3)2 and Cd(N3)2·3/2 DMSO
While the procedure starting from binary azides avoids by-
products, which facilitates the isolation of polyazido metallate
salts, the manipulation with Cd(N3)2 and Hg(N3)2 always bears
the risk of serious explosions. On the other hand, in the reac-
tion of iodometallate salts with silver azide, large amounts of
silver iodide precipitate have to be filtered off. However, the
iodide/azide exchange reaction limits the hazards to the
known risks during handling neat AgN3. The required polyiodo
To avoid the presence of water during the synthesis, we decid-
ed to prepare cadmium(II) azide by means of a fluoride/azide
exchange reaction in cadmium(II) fluoride (Scheme 2). The
reaction of excess trimethylsilyl azide Me3SiN3 with freshly
prepared CdF2 in CH3CN proceeded within one day at
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3650
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 1. ORTEP drawings of the coordination sphere of Cd(N3)2·3/2 DMSO
(top), [Ph4P][Cd2(N3)5(H2O)] (middle) and [Ph4P][Cd(N3)3] (bottom) in the crys-
tal. Thermal ellipsoids with 50% probability at 173 K. Disorder not displayed.
Selected bond lengths [ꢂ]: Cd(N3)2·3/2 DMSO (symmetry codes: (i) ꢀx+2,
ꢀy+1, ꢀz; (ii) ꢀx+1, ꢀy+1, ꢀz): Cd1ꢀN1 2.273(1), Cd1ꢀO1 2.316(1), Cd1ꢀ
N4 2.321(1), Cd1ꢀN1i 2.322(1), Cd1ꢀN10ii 2.329(1), Cd1ꢀN7 2.431(1), Cd2ꢀO2
2.249(1), Cd2ꢀN4 2.273(1), Cd2ꢀO3 2.282(1), Cd2ꢀN10 2.283(1), Cd2ꢀN7ii
2.348(1), Cd2ꢀN7 2.513(1), N1ꢀN2 1.207(2), N2ꢀN3 1.141(2), N4ꢀN5 1.196(2),
N5ꢀN6 1.141(2), N7ꢀN8 1.217(2) N8ꢀN9 1.141(2), N10ꢀN11 1.196(2), N11ꢀ
N12 1.153(2); [Ph4P][Cd2(N3)5(H2O)] (symmetry codes: (i) ꢀx+2,ꢀy+1,ꢀz+1;
(ii) x+1,y,z): Cd1ꢀN1 2.267(2), Cd1ꢀN7 2.284(2), Cd1ꢀO1 2.302(2), Cd1ꢀN10
2.338(2), Cd1ꢀN4 2.381(2), Cd1ꢀN10i 2.415(2), Cd2ꢀN13 2.285(2), Cd2ꢀN1ii
2.308(2), Cd2ꢀN7 2.310(2), Cd2ꢀN4i 2.351(2), Cd2ꢀN4ii 2.410(2), Cd2ꢀN10
2.416(2), N1ꢀN2 1.193(3), N2ꢀN3 1.140(3), N4ꢀN5 1.230(3), N5ꢀN6 1.145(3),
N7ꢀN8 1.193(3), N8ꢀN9 1.149(3), N10ꢀN11 1.220(3), N11ꢀN12 1.135(3), N13ꢀ
N14 1.182(3), N14ꢀN15 1.154(3); [Ph4P][Cd(N3)3] (symmetry codes:
(i) ꢀx+2,ꢀy,ꢀz+1; (ii) ꢀx+1,ꢀy,ꢀz+1; (iii) xꢀ1,y,z): Cd1ꢀN4 2.316(2), Cd1ꢀ
N4i 2.316(2), Cd1ꢀN7i 2.357(2), Cd1ꢀN7 2.357(2), Cd1ꢀN1 2.4078(2), Cd1ꢀN1i
2.408(2), Cd2ꢀN4 2.332(2), Cd2ꢀN4ii 2.332(2), Cd2ꢀN1iii 2.347(2), Cd2ꢀN1i
2.347(2), Cd2ꢀN9ii 2.397(2), Cd2ꢀN9 2.397(2), N6ꢀN5 1.150(2), N3ꢀN2
1.147(2), N1ꢀN2 1.203(2), N4ꢀN5 1.205(2), N9ꢀN8 1.177(2), N8ꢀN7 1.179(2).
Figure 2. ORTEP drawings of the coordination spheres of the anions in
[Ph4P][Hg(N3)3] (top) and [PNP][Hg(N3)3] (bottom) in the crystal. Thermal
ellipsoids with 50% probability at 173 K. Selected bond lengths [ꢂ]: [Ph4P]
[Hg(N3)3] (symmetry code: (i) ꢀx+2,ꢀy+1,ꢀz+1; (ii) ꢀx+1,ꢀy+1,ꢀz+1):
Hg1ꢀN1 2.113(2), Hg1ꢀN7 2.097(2), Hg1ꢀN4 2.452(2), Hg1ꢀN4i 2.570(1), Hg1
N1ii 2.966(2), N1ꢀN2 1.213(2), N2ꢀN3 1.143(2), N4ꢀN5 1.199(2), N4ꢀHg1
2.574(2), N5ꢀN6 1.154(2), N7ꢀN8 1.202(2), N8ꢀN9 1.146(2); [PNP][Hg(N3)3]
(symmetry code: (i)ꢀx,ꢀy+1,ꢀz); Hg1ꢀN4 2.077(4), Hg1ꢀN1 2.089(3), Hg1ꢀ
N9 2.467(3), Hg1ꢀN7 2.485(4), N1ꢀN2 1.202(5), N2ꢀN3 1.145(5), N4ꢀN5
1.187(5), N5ꢀN6 1.121(5), N7ꢀN8 1.165(4), N8ꢀN7i 1.165(4), N9ꢀN10 1.170(3),
N10ꢀN9i 1.170(3).
metallate salts can easily be prepared by the reaction of
cadmium(II)- or mercury(II) iodide with the corresponding
phosphonium iodide [M]I in acetone or dichloromethane, re-
spectively. After heating to reflux, precipitation and washing
with n-pentane and drying in vacuo, all polyiodo cadmate and
mercurate salts are obtained as colorless, microcrystalline
solids in almost quantitative yields.[16b]
We found, that in the present study, both synthetic routes
according to Scheme 3 gave almost identical results. However,
while salts of triazido- and tetraazido mercurate anions with
both different phosphonium cations could successfully be
isolated, not all possible combinations of polyazido cadmate
anion and cation were found to be stable in the solid state
(Figures 1–3). Both azido cadmates [Ph4P][Cd(N3)3] and
[PNP]2[Cd(N3)4] were easily prepared as colorless microcrystal-
line solids from [Ph4P][CdI3] or [PNP][CdI3]/[PNP]I, respectively,
with AgN3. Likewise, the reactions of Cd(N3)2 with stoichio-
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3651
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
metric amounts of [M]N3 in polar solvents such as acetonitrile
or DMSO proceeded in good yields.
In contrast, it was impossible to isolate the formal triazido
cadmate [Cd(N3)3]ꢀ bearing the bulky [PNP]+ cation, as well as
the tetraazido cadmate [Cd(N3)4]2ꢀ with the smaller [Ph4P]+
cation (Vion: [Ph4P]+ 464 vs. [PNP]+ 723 ꢂ3),[17] since these solu-
tions always deposited crystals of the corresponding tetra- or
triazido cadmate salts, respectively, almost independently of
the stoichiometry. During many crystallization attempts, only
traces of [Ph4P]2[Cd(N3)4] could be identified by single crystal X-
ray diffraction, however, without any further analytical charac-
terization.[16] Interestingly, besides these tri- and tetraazido
cadmates, the formation of a pentaazido dicadmate anion as
reported in the salt [Ph4P][Cd2(N3)5(H2O)] was not observed.
Therefore, we prepared the latter compound in analogy to the
procedure given in literature from the reaction of [Ph4P]Cl,
NaN3 and cadmium(II) sulfate in water.[8] While all analytical
data compare very well to those reported by Beck et al., we
found that in contrast to the conclusions of the authors, [Ph4P]
[Cd2(N3)5(H2O)] could be recrystallized from hot water without
decomposition, in good yields. However, if the pentaazido di-
cadmate salt was recrystallized from CH3CN, only the triazido
cadmate salt [Ph4P][Cd(N3)3] instead of a possible acetonitrile
adduct was obtained.
Figure 3. ORTEP drawings of the molecular structure of the [E(N3)4]2ꢀ anions
in the crystal (left: [Ph4P]+, right: [PNP]+ salt). Thermal ellipsoids with 50%
probability at 173 K. Disorder not displayed. Selected bond lengths [ꢂ]:
[Ph4P][Cd(N3)4]: Cd1aꢀN4a 2.160(4), Cd1aꢀN7 2.185(3), Cd1aꢀN10a 2.197(7),
Cd1aꢀN1a 2.224(5), N1aꢀN2a 1.179(9), N2aꢀN3a 1.110(8), N4aꢀN5a 1.183(4),
N5aꢀN6a 1.147(8), N7ꢀN8 1.191(3), N8ꢀN9 1.151(3), N10aꢀN11a 1.183(4),
N11aꢀN12a 1.147(8); [PNP][Cd(N3)4]: Cd1ꢀN7 2.08(2), Cd1ꢀN10 2.194(4),
Cd1ꢀN4 2.205(5), Cd1ꢀN1 2.29(2), N1ꢀN2 1.195(7), N2ꢀN3 1.160(7), N4ꢀN5
1.159(8), N5ꢀN6 1.137(9), N7ꢀN8 1.191(7), N8ꢀN9 1.160(7), N10ꢀN11
1.156(8), N11ꢀN12 1.130(8); [Ph4P][Hg(N3)4]: Hg1ꢀN10 2.184(2), Hg1ꢀN7
2.189(2), Hg1ꢀN4 2.294(2), Hg1ꢀN1 2.295(2), N1ꢀN2 1.183(2), N2ꢀN3
1.155(2), N4ꢀN5 1.203(2), N5ꢀN6 1.152(2), N7ꢀN8 1.197(2), N8ꢀN9 1.153(2),
N10ꢀN11 1.202(2), N11ꢀN12 1.151(2); [PNP][Hg(N3)4] (symmetry code:
(i) ꢀx+1,y,ꢀz+3/2): Hg1ꢀN4 2.180(2), Hg1ꢀN4i 2.180(2), Hg1ꢀN1i 2.286(2),
Hg1ꢀN1 2.286(2), N1ꢀN2 1.183(3), N2ꢀN3 1.160(3), N4ꢀN5 1.190(2), N5ꢀN6
1.143(3).
Contrary to the polyazido cadmates, tri- and tetraazido mer-
curates [PNP][Hg(N3)3] and [PNP]2[Hg(N3)4] from the reaction of
Hg(N3)2 with [PNP]N3 in acetonitrile, as well as both [Ph4P]+
mercurate salts, obtained from [Ph4P][HgI3] and AgN3 or
Hg(N3)2 and [Ph4P]N3 in acetonitrile or DMSO could be success-
fully isolated as colorless crystalline solids. In case of the
[Ph4P]+ salts, however, a pronounced influence of solvent po-
larity on the crystallization was observed. While solutions of
acetonitrile and dichloromethane promote the deposition of
[Ph4P][Hg(N3)3] almost independently of the stoichiometry, crys-
talline tetraazido mercurate [Ph4P]2[Hg(N3)4] was only accessible
from polar, very concentrated DMSO solutions. It is interesting
to note, that the reaction of mercury(I) azide with [Ph4P]N3
could also be applied in the preparation of [Ph4P]2[Hg(N3)4],
since Hg2(N3)2 readily disproportionates into Hg(N3)2 and ele-
mental mercury in acetonitrile. It is worthwhile to mention,
that all analytical data of [Ph4P][Hg(N3)3] (M.p., IR/Raman) are in
good agreement with the values found in literature.[8]
as solvent, the crystalline product was also found to be slightly
contaminated by chloride ([Ph4P][Hg(N3)2.94Cl0.06]) as deduced
from a single X-ray structure determination (Figure 4).[16] Inter-
estingly, vibrational spectroscopy (IR/Raman) and melting-
point determination are not completely suited to detect such
small chloride impurities. Only recently, the stability of [Ph4P]N3
and [Ph4P]3[Bi(N3)6] in chlorinated solvents was investigated by
our group, which displayed a reversible reaction for the
Furthermore, we investigated the stability of tetraphenyl-
phosphonium polyazido mercurates in dichloromethane solu-
ꢀ
ꢀ
tions with regard to a Clꢀ/N3 exchange.[16,18] When a highly
Clꢀ/N3 exchange, following a second order rate law.[18c] In any
concentrated solution of [Ph4P]2[Hg(N3)4] was stored at ambi-
ent temperature for one week, depending on the reaction and
crystallization time, species such as [Ph4P][Hg(N3)1.84Cl1.16] and
[Ph4P]2[Hg(N3)1.42Cl2.58], which display a partial substitution of
the azide groups by chloride atoms, could be isolated and
characterized by single crystal X-ray analyses. For [Ph4P]
[Hg(N3)1.84Cl1.16], a melting point of 1398C was determined,
which is only slightly lower compared to pure triazido mercu-
rate ([Ph4P][Hg(N3)3], 1478C). Contrarily, for [Ph4P][Hg(N3)3]
a considerably slower exchange rate was observed indicating
a larger complex stabilization constant. Nevertheless, when the
synthesis of [Ph4P][Hg(N3)3] was performed in dichloromethane
case, to exclude a possible chloride/azide exchange, DMSO
and CH3CN should be used as solvents for chloride-free
synthesis of polyazido metallate salts.
Moreover, we observed the formation of methyltetrazolate
salts as an interesting side reaction in concentrated acetonitrile
solutions of polyazido mercurates bearing the bulky [PNP]+
cation. The diazido methyltetrazolato mercurate salt [PNP]
[Hg(N3)2(CH3CN4)] as CH3CN hemisolvate was isolated in small
amounts as colorless needle-like crystals from concentrated
CH3CN solutions of [PNP]2[Hg(N3)4] after storage for several
days at 58C (Scheme 4). The tetrazolate salt could be recrystal-
lized from DMSO to give the solvent free compound. Both
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3652
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
value of 3258C found by Wçhler et al. (Table 1).[3e] However, it
should be mentioned that the determination of the thermal
detonation point of binary azides is in general strongly
influenced by crystallite size, heating-rate and quantity.[3e,5]
Contrary to binary Cd(N3)2, the adduct compound
Cd(N3)2·3/2 DMSO does not explode, but melts at 1208C. Like-
wise, [Ph4P][Cd(N3)3] and [PNP]2[Cd(N3)4] are thermally robust
and possess similar melting points of 198 and 2018C, respec-
tively, while the salt [Ph4P][Cd2(N3)5(H2O)] even remains solid to
above 2668C in accord with the observation by Beck et al.
(Table 1).[8] For the tri- and tetraazido mercurate salts also
remarkably high melting points in the range 119–1978C are
observed. Thereby, the substitution of one azide ligand in
[PNP][Hg(N3)3] by a methyltetrazolate anion in the structurally
related [PNP][Hg(N3)2(CH3CN4)] obviously shows only little
effect on the melting point (119 vs. 1228C). While Cd(N3)2,
Cd(N3)2·3/2 DMSO and [Ph4P][Cd2(N3)5(H2O)] are hygroscopic and
slowly decompose in air, mercury(II) azide as well as crystalline
tri- and tetraazido cadmate and mercurate salts, respectively,
are neither air, nor moisture sensitive. All compounds are
slightly sensitive to light and should therefore be handled and
stored in absence of sunlight. All azide compounds were fully
characterized by 14N, 133Cd and 199Hg NMR, IR and Raman
spectroscopy, elemental analysis and single-crystal structure
elucidation.[16] Table 1 summarizes the characterization data of
all considered cadmium and mercury azide compounds.
The 14N NMR spectra of all compounds (run in [D6]DMSO at
300 K) show two well-resolved resonances (Table 1) in accord
with literature known values for covalently bound azido
groups (e.g., ꢀ136 (43) and ꢀ253 ppm (480 Hz) for
[Bi(N3)4]ꢀ).[18c,19] The observation of only one set of azide signals
and the absence of a separate Na resonance indicate strong
quadrupole relaxation effects and a rapid ligand exchange on
the NMR timescale.[20] As expected, a medium-sharp signal at
d=ꢀ132–ꢀ133 ppm (Dn1/2 =116–230 Hz) for the Nb atoms of
the cadmium compounds, and ꢀ131 ppm (Dn1/2 =40–51 Hz)
for the Nb atoms in the mercury salts, is observed. Obviously
the shift of the Nb-resonance is only slightly influenced by the
degree of substitution in the series of azide compounds, with
exception of the resonance of binary Hg(N3)2 which is slightly
shifted to higher field ꢀ133 ppm (Dn1/2 =65 Hz; d=ꢀ135, Dn1/
2 =30 Hz in D2O). While the remarkably broadened Na/g-reso-
nances of the cadmium species in the range ꢀ277–ꢀ284 ppm
(Dn1/2 =530–1200 Hz) indicate no identifiable dependence on
the degree of substitution by azide ligands, the influence on
the shift for the Na/g atoms in the mercury compounds is more
pronounced. In the series Hg(N3)2 >[Hg(N3)3]ꢀ >[Hg(N3)4]2ꢀ, the
Na/g resonance is shifted to higher field (d=ꢀ261–ꢀ270 ppm),
while the half-widths of the signals decline (Dn1/2 =690–
276 Hz).
Figure 4. ORTEP drawings of the molecular structure of the anions
2ꢀ
[Hg(N3)2.94Cl0.06] ] ]
ꢀ (top), [Hg(N3)1.84Cl1.16 ꢀ (middle) and [Hg(N3)1.42Cl2.58
(bottom) in the crystal. Thermal ellipsoids with 50% probability at 173 K.
Selected bond lengths [ꢂ]: [Ph4P][Hg(N3)2.94Cl0.06] (symmetry code:
(i) ꢀx+1,ꢀy+2,ꢀz+1; (ii) ꢀx+1,ꢀy+1,ꢀz+1): Hg1ꢀN7 2.106(3), Hg1ꢀN1
2.122(3), Hg1ꢀN4 2.460(2), Hg1ꢀCl1 2.48(2), Hg1ꢀN4i 2.580(2), Hg1ꢀN1ii
2.976(2), N1ꢀN2 1.205(3), N2ꢀN3 1.151(3), N4ꢀN5 1.193(3), N5ꢀN6 1.154(3),
N7ꢀN8 1.191(6), N8ꢀN9 1.140(5). [Ph4P][Hg(N3)1.84Cl1.16] (symmetry code:
(i) ꢀx+1,ꢀy+2,ꢀz+1): Hg1ꢀN1 2.132(7), Hg1ꢀN4 2.1327(19), Hg1ꢀCl1
2.6657(5), Hg1ꢀCl1i 2.7763(5), Cl1ꢀHg1i 2.7763(5), N1ꢀN2 1.209(6), N2ꢀN3
1.154(5), N4ꢀN5 1.195(3), N5ꢀN6 1.147(3); [Ph4P][Hg(N3)1.42Cl2.58]: Hg1ꢀN4
2.22(2), Hg1ꢀN1 2.258(5), Hg1ꢀCl1 2.423(8), Hg1ꢀCl2 2.529(9), N1ꢀN2
1.178(8), N2ꢀN3 1.170(9), N4ꢀN5 1.18(2), N5ꢀN6 1.13(2).
mercurates are neither shock, nor friction sensitive and defla-
grate only gently in a flame. Albeit only small quantities of the
tetrazolate crystals were observed during the crystallization in
acetonitrile solutions, the formal [2+3] cycloaddition reaction
between azido mercurate anions and acetonitrile should
always be considered as possible source for impurities.
Scheme 4. Formation of diazido methyltetrazolato mercurate salt [PNP]
[Hg(N3)2(CH3CN4)].
Physical and spectroscopic data
In contrast to binary cadmium and mercury(II) azide, none of
the polyazido metallate salts [M]n[E(N3)(2+n)] as well as [Ph4P]
[Cd2(N3)5(H2O)] and the DMSO adduct of Cd(N3)2 is sensitive to
friction and shock. While crystalline a- and b-Hg(N3)2 are
known to sublime at temperatures above 1808C on gentle
warming,[5] pure Cd(N3)2 is found to detonate at about 3458C
on slow heating (208Cminꢀ1) which compares well with the
Only negligible differences were found between spectra of
the tri- and tetraazido mercurates for both different counter
ions [Ph4P]+ and [PNP]+, respectively, indicating only weak
cation–anion interactions in solution. Contrary to the 14N data,
the 133Cd NMR spectra display a clear trend dependent on the
degree of substitution. While the resonance of cadmium(II)
azide at ꢀ580 ppm is found at highest field (cf. Cd(N3)2 in D2O:
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3653
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Selected analytical data.[a] For comparison, the data of [Ph4P]N3 and [PNP]N3 are included.[16a]
Cd
M.p. 133Cd NMR 14N NMR[a]
[8C] [ppm] [ppm]
IR nas(N3)[b]
[cmꢀ1
Raman nas(N3)[c]
[cmꢀ1
]
]
Cd(N3)2
Cd(N3)2 · /2 DMSO
345[d] ꢀ580[e]
ꢀ133(200), ꢀ282(1100) 2096(s), 2055(s), 1987(w)
– 2051(s, broad)
2165(5), 2134(5), 2102(7), 2096(7), 2080(7), 2044(9)
2115(8), 2077(2), 2070(2), 2060(1)
3
120
–
[Ph4P][Cd2(N3)5(H2O)] 266
ꢀ540
ꢀ449
ꢀ487
ꢀ419
ꢀ417
ꢀ132(230), ꢀ277(1200) 2099(m), 2071(m), 2052(m), 2024(s), 2134(22), 2073(15), 2055(13), 2045(13), 2032(14)
[Ph4P][Cd(N3)3]
[PNP][Cd(N3)3]
[Ph4P]2[Cd(N3)4]
[PNP]2[Cd(N3)4]
198
ꢀ133(170), ꢀ284(820) 2065(s), 2044(s), 2015(s)
2098(7), 2038(1), 2024(1)
[f]
–
–
ꢀ133(154), ꢀ281(770)
ꢀ133(116), ꢀ281(530)
–
–
–
–
[f]
201
ꢀ133(160), ꢀ280(830) 2072(m), 2037(s)
2073(3), 2041(5)
Hg
M.p.
[8C]
199Hg NMR
[ppm]
14N NMR[a]
[ppm]
IR nas(N3)[b]
[cmꢀ1
Raman nas(N3)[c]
[cmꢀ1
]
]
a-Hg(N3)2
195–200[g]
122
147
119
143
197
214
235
ꢀ1745[h]
ꢀ1420
ꢀ1546
ꢀ1539
ꢀ1407
ꢀ1387
–
ꢀ133(65), ꢀ261(690)
ꢀ131(51), ꢀ267(380)[i]
ꢀ131(47), ꢀ267(428)
ꢀ131(46), ꢀ266(400)
ꢀ131(40), ꢀ270(300)
ꢀ131(40), ꢀ270(276)
ꢀ131(15), ꢀ277(25)
ꢀ131(18), ꢀ276(52)
2089(m), 2044(s)
2059(m), 2038(s), 2026(s)
2038(s), 2006(s)
2071(m), 2041(s), 2021(s), 2000(s)
2049(m), 2007(s)
2050(w), 2007(s)
2127(4), 2094(11), 2071(3)
2062(16), 2042(15), 2030(14)
2077(10), 2045(4), 2008(2)
2062(9), 2036(3)
2054(10), 2020(4), 2010(3), 2001(3)
2050(1), 2026(1), 2014(1)
n.o.[j]
n.o.[j]
[PNP][Hg(N3)2(CH3CN4)]
[Ph4P][Hg(N3)3]
[PNP][Hg(N3)3]
[Ph4P]2[Hg(N3)4]
[PNP]2[Hg(N3)4]
[PNP]N3
2004(m), 1989(s),
1993(s),
[Ph4P]N3
–
[a] Values from [D6]DMSO, in parenthesis Dn1/2 values in Hz. [b] Relative IR intensities denoted as weak (w), medium (m), strong (s). [c] Relative Raman in-
tensities scaled to 100. [d] Detonation. [e] Cd(N3)2 in D2O: 133Cd: d=ꢀ633 ppm; 14N NMR d=ꢀ133(39), ꢀ281 ppm (360). [f] Not isolated. [g] Sublimation
(no decomposition up to 3608C). [h] Hg(N3)2 in D2O: 199Hg NMR d=ꢀ1785 ppm; 14N NMR d=ꢀ135 (30), ꢀ265 ppm (400). [i] 2-Methyltetrazolate:
1H-15N-HMBC d=ꢀ80 (CH3CN2N2), 8.3 ppm (CH3CN2N2). [j] Not observed; not Raman active due to almost ideal D1h symmetry
ꢀ633; CdCl2: ꢀ586 ppm in water),[21a,b] the pentaazido
dicadmate [Cd2(N3)5(H2O)]ꢀ (ꢀ540 ppm), followed by the triazi-
do- [M][Cd(N3)3] ([M]+ =[Ph4P]+: ꢀ449; [PNP]+: ꢀ487 ppm)
and tetraazido cadmates [M]2[Cd(N3)4] ([M]+ =[Ph4P]+: ꢀ419;
[PNP]+: ꢀ417) become more deshielded. The 199Hg NMR spec-
tra of the mercury(II) azides also exhibit always one sharp
signal, which is shifted to low field in the sequence Hg(N3)2
(ꢀ1745), [M][Hg(N3)3] ([M]+ =[Ph4P]+: ꢀ1546; [PNP]+:
ꢀ1539 ppm) and [M]2[Hg(N3)4] ([M]+ =[Ph4P]+: ꢀ1407; [PNP]+:
ꢀ1387 ppm) following the opposite trend as observed in the
14N spectra (vide supra). This trend is in agreement with values
found for the series HgCl2 (ꢀ1560 ppm), [HgCl3]ꢀ (ꢀ1436 ppm)
and [HgCl4]2ꢀ (ꢀ1152 ppm) in water.[22] It is interesting to note,
that the Hg resonance in the tetrazolate substituted com-
pound [Hg(N3)2(CH3CN4)]ꢀ is considerably low-field shifted
(ꢀ1420 ppm) as compared to the corresponding triazido mer-
curate [Hg(N3)3]ꢀ (ꢀ1539 ppm), which might indicate an addi-
tional deshielding by the aromatic tetrazolate ring. However,
the opposite trend upon substitution of azide groups by
methyltetrazolate units was observed by Klapçtke et al. in the
neutral organomercury azides RHgN3 (cf. PhHgN3: ꢀ1292 vs.
PhHg(CH3CN4): ꢀ1324 ppm and CH3HgN3: ꢀ959 vs.
CH3Hg(CH3CN4): ꢀ980 ppm in [D4]methanol).[7f,g] It should be
mentioned that the chemical shift in 133Cd and 199Hg NMR
spectroscopy strongly depends on the concentration and
polarity of the solvent (e.g., HgCl2: ꢀ1560 in D2O vs. ꢀ1502 in
[D6]DMSO; Hg(N3)2: ꢀ1785 in D2O vs. ꢀ1745 ppm in
[D6]DMSO).[21]
stretching mode (ns(N3); Cd: 1367–1331; Hg: 1373–1265 cmꢀ1),
and the deformation mode of the N3-unit (g/d(N3); Cd: 667–
594; Hg: 678–576 cmꢀ1). The existence of more than one azido
ligand results in in-phase (i.p.) and out-of-phase (o.p.) coupling
of vibrational modes in accord with theoretical data (Table 1
and Figure 5).[1,16] Moreover, the combination mode of antisym-
metric and symmetric vibration nas +ns(N3) can be found in the
IR spectra (Cd: 3405–3302; Hg: 3309–3253 cmꢀ1). As illustrated
in Figure 5, the nas(N3) in the series Hg(N3)2 >[Hg(N3)3]ꢀ >
[Hg(N3)4]2ꢀ is shifted to lower wave numbers, thereby indicat-
ing an increasingly ionic bonding situation to the azide
groups, in accord with our calculations (vide infra).[16] In the
analogous cadmium compounds, the trend is less pronounced,
which can be rationalized by the fact, that except of the tetraa-
zido cadmates, which are distorted tetrahedrally coordinated,
all other Cd compounds are six-coordinated in the solid state,
leading to a higher degree of ionicity in the cadmium nitrogen
bonds. It is worth mentioning, that our vibrational data of
binary cadmium(II) azide agree well with the values reported
by Schnick et al.[10b]
Interestingly, the in-phase and out-of-phase N-Cd-N stretch-
ing modes (ni.p./no.p.(N-Cd-N)) in the range 329–248 cmꢀ1 are
observed at considerably lower wave number than in the mer-
cury compounds, which indicates a stronger bonding (higher
degree of covalent bonding) in the mercury(II) azides as com-
pared to the analogous cadmium compounds, in agreement
with our expectations. For the polyazido mercurate salts, only
the in-phase N-Hg-N stretching mode (ni.p.(N-Hg-N)) in the
range 373–331 cmꢀ1 is found, while for binary a-Hg(N3)2 both,
the out-of-phase (no.p.(N-Hg-N); 434 and 420 cmꢀ1) as well as
the in-phase vibrational mode (ni.p.(N-Hg-N)) at 349 cmꢀ1 can
be identified in the Raman spectra. However, while these
The vibrational spectra (ATR-IR/Raman) of cadmium and
mercury compounds feature the presence of azido ligands as
indicated by the antisymmetric stretching mode (nas(N3); Cd:
2165–2024; Hg: 2127–2000 cmꢀ1, Table 1), the symmetric
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3654
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
The closest metal–metal distances are found in the [Ph4P]+
salts, with similar values for the cadmium and mercury
compounds (Cd···Cd 7.436, Hg···Hg 7.482 ꢂ), while the steric
demand of the [PNP]+ cation enforces considerably larger E···E
distances (Cd···Cd 11.49, Hg···Hg 12.10 ꢂ). In contrast, the struc-
tural units in the formal triazido metallate anions [E(N3)3]ꢀ, as
well as the formal pentaazido dicadmate anion [Cd2(N3)5]ꢀ and
the DMSO solvate of Cd(N3)2 (Figure 1, Figure 2 and Figure 6),
are considerably associated, forming one-dimensional chains in
the solid state (Figure 7 and Figure 8).
Figure 6. ORTEP drawing of the asymmetric unit of the [Hg(N3)2(CH3CN4)]ꢀ
anion in the crystal. Thermal ellipsoids with 50% probability at 173 K. Disor-
dered N3 unit not displayed. Selected bond lengths [ꢂ]: Hg1ꢀN4a 2.088(17),
Hg1ꢀN1 2.237(5), Hg1ꢀN7 2.260(3), Hg1ꢀN20i 2.275(4), Hg2ꢀN11 2.102(4),
Hg2ꢀN14 2.123(4), Hg2ꢀN10 2.392(4), Hg2ꢀN17 2.397(4), N1ꢀN2 1.188(6),
N2ꢀN3 1.159(6), N4aꢀN5a 1.199(5), N5aꢀN6a 1.151(5), N7ꢀC1 1.325(5), N7ꢀ
N8 1.356(5), N8ꢀN9 1.286(5), N9ꢀN10 1.360(5), N10ꢀC1 1.327(5), N11ꢀN12
1.200(4), N12ꢀN13 1.151(5), N14ꢀN15 1.199(5), N15ꢀN16 1.139(5), N17ꢀC3
1.323(5), N17ꢀN18 1.350(5), N18ꢀN19 1.303(5), N19ꢀN20 1.357(5), N20ꢀC3
1.324(6), N20ꢀHg1ii 2.275(4). Symmetry codes: (i) xꢀ1,y,z; (ii) x+1,y,z.
Figure 5. Antisymmetric stretching modes (Raman spectra) of the N3 units
(nas N3). Top: Cd(N3)2, Cd(N3)2·3/2 DMSO, [Ph4P][Cd2(N3)5(H2O)], [Ph4P][Cd(N3)3]
and [PNP]2[Cd(N3)4]. Bottom: a-Hg(N3)2, [Ph4P][Hg(N3)3], [PNP][Hg(N3)3], [PNP]
[Hg(N3)2(CH3CN4)], [Ph4P]2[Hg(N3)4] and [PNP]2[Hg(N3)4].
values agree well with the IR data published by Dehnicke
et al., it should be noted that the ni.p.(N-Hg-N) mode was
erroneously assigned by the latter.[3a] As expected, [PNP]
[Hg(N3)2(CH3CN4)] exhibits also the typical antisymmetric
(Raman: nas(N3)=2062(1), 2042(1), 2030 cmꢀ1(1)) as well as
symmetric vibrational modes (Raman: ns(N3)=1388(1), 1322(1),
1278 cmꢀ1(1)) for the azide moieties. In addition, a strong N-N
ring vibration at 1274 (s) (IR, cf. Raman: 1278 cmꢀ1(1)) is
observed.
With the exception of the tetraazido cadmate salts
[M]2[Cd(N3)4], which display the anions in a slightly distorted
tetrahedral ligand sphere with average N-Cd-N angles in the
range 101.6–115.48 (Table 2), all other cadmium azides are
composed of six-coordinated cadmium centers, resembling
slightly distorted octahedral coordination polyhedra with
N-Cd-N angles averaging to the ideal value of 908
(Cd(N3)2·3/2 DMSO 76.4–107.0, [Ph4P][Cd2(N3)5(H2O)] 78.3–101.8,
[Ph4P][Cd(N3)3] 82.6–97.48, cf. 75.7–99.78 in Cd(N3)2).[10b]
Accordingly,
short
Cd···Cd
distances
are
found
(Cd(N3)2·3/2 DMSO 3.585–3.689, [Ph4P][Cd2(N3)5(H2O)] 3.523–
3.697, [Ph4P][Cd(N3)3] 3.637 ꢂ, Figure 1) which lie only slightly
above the sum of the van der Waals radii (SrvdW(CdꢀCd)=
3.20 ꢂ,[23] cf. Cd···Cd in Cd(N3)2 3.689 ꢂ).[10b] Within these struc-
tures, the octahedral units are either connected by m2(1,1)- or
m3(1,1,1)-bridging azide ligands (Scheme 5), leading to common
edge-sharing between two or three octahedra, respectively.
In the structure of Cd(N3)2·3/2 DMSO, these interactions lead
to the formation of tetrameric units consisting of four triply
edge-sharing octahedra. These units are further connected to
the adjacent units by m2(1,1)-bridging azide ligands, resulting in
the formation of one-dimensional chains along [100] (Figure 1
X-ray crystallography
The single crystal X-ray structures of all synthesized cadmium
and mercury azide compounds, except [PNP][Cd(N3)3], which
was found to be inaccessible in the solid state (vide supra),
and, moreover, the structures of the starting materials [PNP]
[CdI3], [PNP]2[CdI4], [Ph4P][HgI3] and [Ph4P]2[HgI4] were
determined.
The crystallographic data are summarized in the Supporting
Information (Tables S1–S8). As depicted in Figure 3, well sepa-
rated [E(N3)4]2ꢀ anions are found in the [Ph4P]+ and [PNP]+
salts of the tetraazido cadmates and mercurates, respectively.
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3655
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Selected structural data from single crystal X-ray determinations. Bond lengths in ꢂ, angles in 8.
Cd
Cd···Cd[a]
X-Cd-X
(X=N, O)
CdꢀO[b] CdꢀN
NaꢀNb
NbꢀNg
Na-Nb-Ng
[c]
[e]
Cd(N3)2
3.689[d]
75.7–99.7
–
2.348 m3(1,1,3)
1.195
1.200, 1.219
1.147
1.146, 1.138
178.4
178.9, 179.3
Cd(N3)2·2/3 DMSO
3.585–3.689 76.4–107.0
[Ph4P][Cd2(N3)5(H2O)] 3.523–3.697 78.3–101.8
2.289
2.302
2.302 m2(1,1), 2.427 m3(1,1,1)
2.285, 2.292 m2(1,1), 2.385 m3(1,1,1) 1.182, 1.193, 1.226 1.154, 1.145, 1.140 177.6, 178.3, 179.7
[Ph4P][Cd(N3)3]
[Ph4P]2[Cd(N3)4]
[PNP]2[Cd(N3)4]
3.637
7.436
11.49
82.6–97.4
101.6–115.4
102.9–110.8
–
–
–
2.351 m2(1,1), 2.377 m2(1,3)
2.191
2.193
1.204, 1.179
1.184
1.175
1.145, 1.177
1.139
1.147
178.9, 179.2
175.2
174.4
Hg
Hg···Hg[a]
N-Hg-N[f,g]
HgꢀNtetrazole
–
–
–
2.331 m2(1,3), eq.
–
–
HgꢀNazid
2.062 m3(1,1,3)
2.105 ax.,[g] 2.517 m2(1,1), eq.
2.083 ax., 2.476 m2(1,3), eq.
2.154 ax.
NaꢀNb
1.217
1.208, 1.199
1.195, 1.168
1.196
1.200,1.193
1.190, 1.183
NbꢀNg
1.141
1.145, 1.154
1.133, 1.168
1.149
1.152, 1.154
1.143, 1.160
Na-Nb-Ng
a-Hg(N3)2
4.377–4.566
3.888/4.161
6.458/6.811
6.706/6.765
7.482
173.0
175.1
[Ph4P][Hg(N3)3]
[PNP][Hg(N3)3]
[PNP][Hg(N3)2(CH3CN4)][h]
[Ph4P]2[Hg(N3)4]
[PNP]2[Hg(N3)4]
158.3/78.9/98.3[e]
159.5/95.0/96.8
174.9, 179.2
173.5, 180.0
174.5
176.0, 176.5
174.8, 177.4
157.1/104.9/97.5
126.9/105.3/105.7
131.6/108.1/104.0
2.186 ax., 2.295 equiv
2.180 ax., 2.286 equiv
12.10
[a] Closest E···E distances. [b] If there are more than one azide, tetrazolate or DMSO ligands, the average value is given. [c] Values taken from Ref. [10b].
[d] Distance between edge-sharing CdN6 polyhedra, an additional distance of 4.184 ꢂ is found between corner-sharing CdN6 polyhedra. [e] mn(k,l,m) denotes
the bridging mode of the N3 unit according to Scheme 5. [f] The angles are listed in the order Nax-Hg-Nax, Neq-Hg-Neq and average of Nax-Hg-Neq. [g] Refers
to the axial or equatorial positions in the distorted tetrahedral ligand sphere (Scheme 6). [h] Disordered N3 unit was not considered for mean values.
Scheme 5. Observed bridging modes of the azide units (E=Cd, Hg).
and Figure 7). Between these polymeric units, only weak
Ng···HDMSO interactions are found all within the sum of the van
der Waals radii. The closest N···H distance amounts to 2.457 ꢂ
(cf. SrvdW(NꢀH) 3.00 ꢂ).[23] The same m2(1,1)- or m3(1,1,1)-bridging
mode, as observed before, is found in the structure of [Ph4P]
[Cd2(N3)5(H2O)], but contrarily, each octahedron is connected to
each of the four neighboring octahedra by m3(1,1,1)-bridging
azide ligands, resulting in a linear oriented double strand
along [100], surrounded by the [Ph4P]+ cations (Figure 1 and
Figure 7). It is interesting to note, that the [Cd2(N3)5(H2O)]ꢀ
anion is isostructural to the analogue tetramethylammonium
pentaazido dizincate anion in the salt [Me4N][Zn2(N3)5(H2O)]
published by Mautner et al.[24] In the solid-state structure of
[Ph4P][Cd(N3)3], besides m2(1,1)-bridging azide ligands, leading to
the formation of a straight assembly of edge sharing octahedra
along [100], an additional m2(1,3)-bridging mode of azide units
between adjacent Cd centers is observed, resulting in an
alternating tilting of the octahedral coordination polyhedra
(Figure 1 and Figure 7).
As illustrated in Table 2, the CdꢀN and NꢀN distances vary
strongly dependent on the coordination mode of the azide
units. Accordingly, the shortest CdꢀN distances are found for
Figure 7. Side views of the one dimensional chains in the crystal. Top:
the single bonded, terminal azide groups in the tetraazido cad-
Cd(N3)2·3/2 DMSO (view along [010]). Middle: [Ph4P][Cd2(N3)5(H2O)] (view
along [001]). Bottom: [Ph4P][Cd(N3)3] (view along [001]). [Ph4P]+ cations not
displayed.
mates [M]2[Cd(N3)4] with average CdꢀN bond lengths of 2.191
([M]+ =[Ph4P]+) and 2.193 ꢂ ([M]+ =[PNP]+), which lie slightly
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3656
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
above the sum of the covalent radii for a cadmium nitrogen
single bond (Srcov(CdꢀN)=2.07 ꢂ), but within the sum of the
ion radii (Srion(CdꢀN)=2.24 ꢂ) for a tetrahedral coordination
sphere.[25,23] The azide units adopt a typical trans-bent structure
with short average NaꢀNb and NbꢀNg distances (1.184 and
1.139 ꢂ in [Ph4P]2[Cd(N3)4], 1.175 and 1.147 ꢂ in [PNP]2[Cd(N3)4])
and Na-Nb-Ng angles which deviate considerably from linearity
([Ph4P]2[Cd(N3)4] 175.2, [PNP]2[Cd(N3)4] 174.48).
In contrast, the CdꢀN length to the terminal azide ligand in
the octahedral coordination sphere of the cadmium center in
the [Cd2(N3)5(H2O)]ꢀ anion is considerably elongated
(2.285(2) ꢂ), in accord with the sum of the ionic radii Srion(Cdꢀ
N) of 2.41 ꢂ for an octahedral coordination sphere.[23] This, to-
gether with the slightly longer NaꢀNb and NbꢀNg distances of
1.182(3) and 1.154(3) ꢂ, respectively, and a flat angle of 177.68
indicate a higher degree of ionic bonding. In the same way,
the CdꢀN bond lengths to the m2 and m3-bridging azide units
are considerably elongated (Cd(N3)2·3/2 DMSO m2(1,1) 2.302, m3(1,1,1)
2.427, [Ph4P][Cd2(N3)5(H2O)] m2(1,1) 2.292, m3(1,1,1) 2.385, [Ph4P]
[Cd(N3)3] m2(1,1) 2.351 ꢂ, cf. m3(1,1,3) 2.348 ꢂ in Cd(N3)2).[10b] The
NaꢀNb and NbꢀNg distances follow the same trend, with the
longest CdꢀN and NꢀN bonds as well as largest Na-Nb-Ng
angles found for the triply m3(1,1,1)-bridging azides (Table 2).
While the terminal coordinated bridging azide units always ex-
hibit NaꢀNb bond lengths, which are longer than the NbꢀNg
bonds, in the [Cd(N3)3]ꢀ anion two almost equal distances of
1.179 and 1.177 ꢂ are observed in the m2(1,3)-bridging azide
ligand (cf. NaꢀNb/g 1.171–1.176 ꢂ in [Me4N][Cd(N3)3]).[14b] The
mean CdꢀO bond length of 2.289 ꢂ in Cd(N3)2·3/2 DMSO, and
the CdꢀO bond to the water molecule in [Ph4P][Cd2(N3)5(H2O)]
with 2.302 ꢂ are considerably longer than the sum of the cova-
lent radii Srcov(CdꢀO)=1.99,[25] but in accord with the sum of
ionic radii Srion(CdꢀO) of 2.30 ꢂ[23] (cf. SrvdW(CdꢀO) 3.10 ꢂ).[23] In
the polyazido cadmate salts, only weak N···H interactions be-
tween the anions and cations ([Ph4P][Cd(N3)3] 2.618, [Ph4P]
[Cd2(N3)5(H2O)] 2.599, [Ph4P]2[Cd(N3)4] 2.439, [PNP]2[Cd(N3)4]
Figure 8. Side views of the one dimensional chains in the crystal. Top: [Ph4P]
[Hg(N3)3] (view along [001]). Middle: [PNP][Hg(N3)3] (view along [010]).
Bottom: [PNP][Hg(N3)2(CH3CN4)] (view along [001]). [Ph4P]+ and [PNP]+
cations not displayed.
SrvdW(Hg···N)=3.10 ꢂ,[23] Figure 8), resulting in a zig–zag chain
along [100]. The Hg···Hg distance amounts to 3.888 ꢂ within
the dimercurate anions, and 4.161 ꢂ between neighboring
dimers (cf. Hg···Hg 4.377–4.566 ꢂ in a-Hg(N3)2).[5] In contrary, in
the [PNP]+ salts of the triazido mercurate and its tetrazolate
derivative, the tetrahedral coordination polyhedra are less as-
sociated. While the mercurate units in [PNP][Hg(N3)3] are con-
2.517 ꢂ), within the sum of the van der Waals radii of SrvdW(Nꢀ nected via shared m2(1,3)-bridging azide ligands (Hg···Hg 6.458
H) 3.00 ꢂ are observed in the solid state.[23] However, the clos-
est intramolecular N···H contacts are found between the water
molecule and the adjacent terminal azide groups in [Ph4P]
[Cd2(N3)5(H2O)], with donor–acceptor distances (DꢀH···A) of
2.879 and 2.919 ꢂ (cf. SrvdW(NꢀO) 3.10 ꢂ,[23] Figure 1).
and 6.811 ꢂ), a similar coordination geometry is induced by
the 1,3-bridging mode of the 2-methyl-tetrazolate units
resulting in the formation of polyanionic chains along [100],
with Hg···Hg distances of 6.706 and 6.765 ꢂ. Obviously, the
bulky [PNP]+ cations induce a considerable amount of strain
within the mercurate units, leading to a stronger separation of
mercury centers.
In analogy to the cadmium compounds, the triazido
mercurate [Hg(N3)3]ꢀ and diazido tetrazolato mercurate
[Hg(N3)2(CH3CN4)]ꢀ anions, respectively, are also aggregated in
the solid state, forming one-dimensional polyanionic chains,
surrounded by the [Ph4P]+ or [PNP]+ cations (Figure 2 and
Figure 8). However, the Hg2+ cations in all azide compounds
adopt a strongly distorted tetrahedral coordination geometry
(Figure 2, Figure 3 and Figure 8).
In accord with our calculations (vide infra), always two short-
er and two clearly longer HgꢀN bond lengths are observed
within the HgN4 coordination spheres, which comply with the
axial and equatorial ligands in a strongly distorted tetrahedral
geometry as depicted in Scheme 6 ([Ph4P][Hg(N3)3] ax. 2.105,
eq. 2.517, [PNP][Hg(N3)3] ax. 2.083, eq. 2.476, [PNP]
[Hg(N3)2(CH3CN4)] ax. 2.154, [Ph4P]2[Hg(N3)4] ax. 2.186, eq. 2.295,
[PNP]2[Hg(N3)4] ax. 2.180, eq. 2.286 ꢂ, cf. ax. 2.062 in a-
Hg(N3)2,[5] Table 2). These values lie slightly above the sum of
the covalent radii (Srcov(HgꢀN)=2.04 ꢂ),[25] but within the sum
of the ionic radii Srion(HgꢀN)=2.42 ꢂ.[23] The average HgꢀN
bond length between the mercury center and the nitrogen
atoms of the m2(1,3)-bridging tetrazolate ligand equals to
In [Ph4P][Hg(N3)3], the triazido mercurate anions are
associated by two m2(1,1)-bridging azide ligands, resulting in the
formation of a centrosymmetric [Hg2(N3)6]2ꢀ dianion, which re-
sembles two tetrahedral coordination polyhedra, sharing one
common edge. These dimers are further connected to adjacent
units via weak intermolecular Hg···N interactions just below the
sum of the van der Waals radii (Hg1ꢀN1ii 2.966 ꢂ, cf.
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3657
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
2.331 ꢂ (cf. 2.113 ꢂ in MeHg(CH3CN4)).[7g] As displayed in
For nitrogen an aug-cc-pVDZ basis set was used and for Cd
and Hg a fully relativistic pseudopotential (Cd: ECP28 MDF, Hg:
ECP60 MDF)[29] and an aug-cc-pVDZ basis set (26s23p14d2f)/
[5s5p4d2f]):[30]
Table 2, the HgꢀN bond lengths increase continuously along
the
series
a-Hg(N3)2 >[Ph4P][Hg(N3)3]>[PNP][Hg(N3)3]>
[Ph4P]2[Hg(N3)4]>[PNP]2[Hg(N3)4], in accord with an increasing
ionic bond situation. Interestingly, in the tri- and tetraazido
mercurates the shortest HgꢀN bonds are always found in the
EðN3Þ2 þ N3 ! ½EðN3Þ3ꢁꢀ
ð1Þ
ð2Þ
ꢀ
corresponding [PNP]+ salts. With the exception of the m2(1,3)
-
½EðN3Þ3ꢁꢀ þ N3 ! ½EðN3Þ4ꢁ
ꢀ
2ꢀ
bridging azide units in [PNP][Hg(N3)3], which are ideal linear
(N-N-N 1808), and feature an average NꢀN bond length of
1.168 ꢂ, the typical trans-bent structure is observed for all
other azide ligands, with consider-
For the E(N3)2 species three configurations (cis-C2v, trans-C2h,
and cis-C2) were considered (Figure 9 and Figure S29 in the
Supporting Information). As displayed in Figure S22 in the
Supporting Information, the potential energy surfaces of all
E(N3)2 species are very flat with respect to the position of the
azido ligands to each other (cis vs. trans).
ably different NaꢀNb and NaꢀNg
bond distances and Na-Nb-Ng
angles which differ considerably
from linearity (Table 2). As with the
analogous cadmium compounds,
only weak N···H interactions are
observed between the anions and
cations in the polyazido mercurate
salts, ([Ph4P][Hg(N3)3] 2.456(2),
([PNP][Hg(N3)3] 2.596(3), [PNP]
Scheme 6. Distorted tetra-
hedral coordination environ-
ment in the mercury azide
compounds (ax=axial, eq=
equatorial).
[Hg(N3)2(CH3CN4)]
[Ph4P]2[Hg(N3)4]
2.44(2),
2.484(2),
[PNP]2[Hg(N3)4] 2.509(3) ꢂ), which
lie within the sum of the van der
Waals radii of SrvdW(NꢀH) 3.00 ꢂ.[23]
Computational data
To shed some light on the structural diversity and bonding of
E(N3)2, [E(N3)3]ꢀ, and [E(N3)4]2ꢀ (E=Cd, Hg) species, a series of
computations at the M06-2X level was carried out, addressing
thermodynamic and structural questions. For this reason the
gas phase thermodynamics of the successive anion formation
Figure 9. Computed structures of E(N3)2 (top: cis-C2, cis-C2v and trans-C2h),
[E(N3)3]ꢀ (s-C3h, s-Cs, y-C1), and [E(N3)4]2ꢀ (S4, D2, and C2 symmetry).
ꢀ
upon adding N3 to E(N3)2 was calculated (Table S32 in the
Supporting Information). Additionally, it was our aim to study
in detail the structural and electronic influence of the azido
ligands. A summary of the structural data as well as the
calculated NBO[26–28] data are given in Table 3 and Table S33 in
the Supporting Information.
While the energies for the minimum structures cis-C2v and
cis-C2 are almost identical, the trans-C2h species is slightly less
favored (DG298 <2 kcalmolꢀ1) and represents a transition state
(NIMAG=1, number of imaginary frequencies). All distances
and angles of these three species differ only slightly, also indi-
cating highly flexible azido ligands with respect to rotation
about the HgꢀN axis (very small rotational barriers), in accord
with our experimental X-ray data (vide infra).
Table 3. Net charges (q) and occupation numbers (occ.) in e.
[a]
E=Cd
E(N3)2
[E(N3)3]ꢀ[b]
[E(N3)4]2ꢀ[c]
q(E)
1.49
ꢀ0.83
0.22
ꢀ0.14
0.55
1.61
ꢀ0.80
0.22
ꢀ0.29
0.38
1.67
ꢀ0.72
0.22
ꢀ0.42
0.32
A very similar situation is found for the anionic complexes
[E(N3)3]ꢀ and [E(N3)4]2ꢀ. For [E(N3)3]ꢀ three different isomers are
found: y-C1, s-Cs and s-C3h (Figure 9), which all lie within 1 kcal
molꢀ1 (relative energies). Interestingly, there is a difference be-
tween the Hg and Cd species for the two non-C3h symmetric
species. In y-C1- and s-Cs-[Hg(N3)3]ꢀ, always one longer (2.25–
2.26 ꢂ) and two shorter HgꢀN distances (2.153–2.163 ꢂ) are
observed along with one large N-Hg-N (140.8–146.78) and two
smaller N-Hg-N angles (106.5–111.68), in accordance with X-ray
data where such preformed Hg(N3)2 units are observed (e.g.,
[PNP][Hg(N3)3] HgꢀNax 2.083, HgꢀNeq 2.476 ꢂ, Nax-Hg-Nax 159.5,
q(Na)
q(Nb)
q(Ng)
occ.(5s)
[a]
E=Hg
E(N3)2
[E(N3)3]ꢀ[b]
[E(N3)4]2ꢀ[c]
q(E)
1.28
ꢀ0.75
0.22
ꢀ0.11
0.88
1.44
ꢀ0.75
0.21
ꢀ0.28
0.59
1.53
ꢀ0.69
0.21
ꢀ0.40
0.46
q(Na)
q(Nb)
q(Ng)
occ.(6s)
[a] cis-C2 isomer. [b] s-C3h isomer. [c] S4 isomer.
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3658
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Neq-Hg-Neq 95.08, vide supra). That is, these species can also be
referred to as formal Hg(N3)2·[M]N3 species. For the analogous
Cd species all CdꢀN bond lengths and N-Cd-N angles are very
similar and close to C3h symmetry (Table S30 in the Supporting
Information).
While for [Hg(N3)4]2ꢀ three different isomers (with C2, D2 and
S4 symmetry) are found, only the C2 and S4 symmetric species
are stable minima on the Cd potential energy surface.
Again, all considered species are very similar in energy
(DG298 <1.2 kcalmolꢀ1) displaying highly flexible azido ligands
with respect to their arrangement around the metal ion
(Table S31 in the Supporting Information).
As expected, NBO analysis clearly indicates ionic bonding
ꢀ
between formal E2+ and N3 ions. However, a closer look at
the computed net charges and the occupation of the 5s (Cd)
and 6s (Hg) orbitals, respectively, reveals small but significant
differences between Hg and Cd species as well as along the
series E(N3)2, [E(N3)3]ꢀ, and [E(N3)4]2ꢀ. Assuming ideal E2+ ions,
empty 5s (Cd) and 6s (Hg) orbitals are expected. As displayed
in Table 3, the computed occupation numbers of the valence
s-type atomic orbitals are quite different from zero (Hg: be-
tween 0.465–0.88 e; Cd: 0.32–0.55 e) strongly decreasing along
E(N3)2 >[E(N3)3]ꢀ >[E(N3)4]2ꢀ. In accordance with these findings,
the computed positive net charges on E increase along this
series (Hg: between 1.28–1.53 e; Cd: 1.49–1.67 e). Two things
can be derived from these NBO data: 1) the (small) covalency
in the EꢀN bond increases from Cd to Hg, but 2) decreases
along E(N3)2 >[E(N3)3]ꢀ >[E(N3)4]2ꢀ. The ionic bonding between
the formal E2+ and the azido ligands becomes clearly visible in
the electron localization function (ELF), which also displays
distorted lone pairs located on the Na atoms (Figure 10 and
Figure 10. Two-dimensional cross-section through one molecule plane (in-
cluding E) of the electron localization function (ELF). Left: Cd; right: Hg spe-
cies; top: cis-C2-E(N3)2, middle: s-C3h-[E(N3)3]ꢀ, and bottom: S4-[E(N3)4].
assumed that solid state effects such as electrostatic interac-
tion between cation and anion stabilizes salts bearing such
dianions. However, the larger energy gain for the Cd species
may explain the larger coordination number (in all cases six
except from [Cd(N3)4]2ꢀ), while the Hg species prefer four with
often quite different HgꢀN distances and a preformed Hg(N3)2
moiety.
Figure S22 in the Supporting Information) as well as
a
stronger core polarization for the Hg species compared to the
Cd species.
The charge distribution in all azides is characterized by alter-
nating net charges along the M(d+)-N(dꢀ)-N(d+)-N(dꢀ) units. Inter-
estingly, the positive net charge on Nb is almost constant in all
species (Hg: 0.21–0.22 e, Cd: 0.22 e), the negative charge on
Na only slightly decreases along E(N3)2 >[E(N3)3]ꢀ >[E(N3)4]2ꢀ
and Cd species>Hg species (Hg: ꢀ0.75–ꢀ0.69 e, Cd: ꢀ0.72–
ꢀ0.83 e), while the terminal Ng atoms feature a significant in-
crease of negative charge along this series (Hg(N3)2: ꢀ0.11,
Conclusion
In conclusion, we present here a comprehensive study of the
[Ph4P]+ and [PNP]+ salts of triazido and tetraazido cadmate
and mercurate anions. All salts are easily accessible by means
of iodide/azide exchange in polyiodo metallates with AgN3, or
by direct reaction of the corresponding binary azide E(N3)2
with [M]N3 in acetonitrile or DMSO. While the solid-state struc-
tures display the Hg2+ cations in a strongly distorted tetra-
hedral coordination environment, the cadmium(II) centers are
octahedrally coordinated, with exception of the tetraazido cad-
mate anion, which builds almost ideal tetrahedral coordination
polyhedra. In accord with our calculations, the bonding be-
tween the metal centers and the azide ligands can be referred
to as dominantly ionic. In the Hg compounds, a considerable
core polarization and a small covalent contribution in the
HgꢀN bonds, decreasing along E(N3)2 >[E(N3)3]ꢀ >[E(N3)4]2ꢀ can
be discussed. Moreover, possible side reactions during the syn-
thesis of polyazido mercurates such as the chloride/azide ex-
change in CH2Cl2, as well as the formation of mercury methyl-
tetrazolate salts by formal [2+3]-cycloaddition with acetonitrile
were observed. While comparative thermodynamic, vibrational
[Hg(N3)3]ꢀ: ꢀ0.25, [Hg(N3)4]2ꢀ
:
ꢀ0.40 e; Cd(N3)2: ꢀ0.14,
[Cd(N3)3]ꢀ: ꢀ0.29, [Hg(N3)4]2ꢀ: ꢀ0.42 e) which is also displayed
by the difference in the 14N NMR shift of the Ng atom (vide
supra, Table 1).
Experimentally we could show, that salts bearing the
ꢀ
[E(N3)3]ꢀ anion can be prepared from neutral E(N3)2 and N3
ions and in agreement with these results the gas phase Gibbs
energies for this reaction [Eq. (1)] are all exergonic decreasing
along E=Cd<Hg (DG298: ꢀ66.6 vs. ꢀ46.5 kcalmolꢀ1, Table S32
in the Supporting Information). However, the next step, the ad-
dition of a second azido ligand leading to the formation of the
[E(N3)4]2ꢀ dianion [Eq. (2)], was computed to be endergonic in
the gas phase with the same trend as discussed before, that is
[Cd(N3)4]2ꢀ (+27.8 kcalmolꢀ1) is the more stable than the anal-
ogous Hg species (+35.9 kcalmolꢀ1). Consequential, it can be
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3659
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
138 cmꢀ1 (10); elemental analysis calcd (%) for C3H9N6O1.5CdS1.5: Cd
35.84; found: Cd 35.77.
(IR/Raman) and NMR data of polyazido cadmates and
mercurates could be obtained for the first time, the structural
elucidation of [Hg(N3)3]ꢀ and [Hg(N3)4]2ꢀ salts fills an open gap
in transition-metal chemistry.
[Ph4P][Cd(N3)3]
Procedure 1: To a stirred suspension of tetraphenylphosphonium
triiodo cadmate [Ph4P][CdI3] (0.416 g, 0.5 mmol) in acetonitrile
(10 mL), silver azide AgN3 (neat, 0.240 g, 1.60 mmol) was added in
one portion at ambient temperature. The yellowish suspension
was stirred for one hour and filtered (F4), resulting in a colorless
solution. The solvent was removed in vacuo and the residue was
dissolved in DMSO (0.6 mL). The colorless solution was allowed to
evaporate over a period of two days resulting in the deposition of
colorless needle-like crystals. The viscous supernatant was removed
by decantation and the crystals were washed with a few drops of
acetone and dried between two filter papers for five minutes.
Drying in vacuo (5 mbar) over P4O10 in a desiccator for 12 h gave
tetraphenylphosphonium triazido cadmate [Ph4P][Cd(N3)3] as color-
less crystals (yield: 50–60%).
Experimental Section
Caution! Covalent azides are potentially hazardous and can de-
compose explosively under various conditions! Especially Cd(N3)2
and Hg(N3)2 are extremely shock-sensitive and can explode violent-
ly upon the slightest provocation. Appropriate safety precautions
(safety shields, face shields, leather gloves, protective clothing)
should be taken when dealing with large quantities.
Cd(N3)2
To a stirred suspension of cadmium(II) fluoride CdF2 (0.451 g,
3.0 mmol) in acetonitrile (5 mL), neat trimethylsilylazide, Me3SiN3
(1.728 g, 15.0 mmol) was added in one portion at ambient temper-
ature. Stirring for 24 h under exclusion of light resulted in a color-
less precipitate and a yellowish supernatant. The solid was collect-
ed by centrifugation (up to 450 g), the supernatant was removed
by decantation and the resulting colorless precipitate was resus-
pended in 2 mL of acetonitrile. The washing procedure was repeat-
ed five times, until the supernatant was colorless. The residue was
dried in vacuo at ambient temperature for one hour and was then
carefully grounded by means of a plastic spatula. Heating for 2 h
at 60–808C in vacuo yielded cadmium(II) azide Cd(N3)2 as a colorless
solid, which was stored under argon at ambient temperature
under exclusion of light (yield: 80–90%). M.p. 3458C (detonation,
heating-rate 208Cminꢀ1 under argon); 133Cd NMR (300 K,
[D6]DMSO, 111.00 MHz): d=ꢀ580 ppm (Dn1/2 =350 Hz); 133Cd NMR
(300 K, D2O, 111.00 MHz): d=ꢀ633 ppm (Dn1/2 =130 Hz); 14N NMR
(300 K, [D6]DMSO, 36.14 MHz): d=ꢀ133 (Nb, Dn1/2 =200 Hz),
ꢀ282 ppm (Ng, Dn1/2 =1100 Hz); 14N NMR (300 K, D2O, 36.14 MHz):
d=ꢀ133 (Nb, Dn1/2 =39 Hz), ꢀ281 ppm (Ng, Dn1/2 =360 Hz); IR
Procedure 2: Cadmium(II) azide Cd(N3)2 (0.049 g, 0.25 mmol) and
tetraphenylphosphonium azide [Ph4P]N3 (0.095 g, 0.25 mmol) were
combined and dissolved in DMSO (0.6 mL) at ambient tempera-
ture. The colorless solution was allowed to evaporate over a period
of two days resulting in the deposition of colorless needle-like crys-
tals. The viscous supernatant was removed by decantation and the
crystals were washed with a few drops of acetone and dried be-
tween two filter papers for five minutes. Drying in vacuo (5 mbar)
over P4O10 in a desiccator for 12 h gave tetraphenylphosphonium
triazido cadmate [Ph4P][Cd(N3)3] as colorless crystals (yield: 50–
60%); M.p. 1988C; 133Cd NMR (300 K, [D6]DMSO, 111.00 MHz): d=
ꢀ449 ppm (Dn1/2 =410 Hz); 14N NMR (300 K, [D6]DMSO, 36.14 MHz):
d=ꢀ133 (Nb, Dn1/2 =170 Hz), ꢀ284 ppm (Ng, Dn1/2 =820 Hz); IR
~
(ATR, 32 scans): n=3369 (w), 3350 (w), 3302 (w), 3079 (w), 3062
(w), 3051 (w), 3020 (w), 2991 (w), 2065 (s), 2044 (s), 2015 (s), 1585
(m), 1574 (w), 1482 (m), 1435 (s), 1334 (m), 1314 (w), 1285 (w),
1277 (w), 1208 (w), 1183 (m), 1162 (w), 1108 (s), 1072 (w), 1026 (w),
995 (m), 978 (w), 940 (w), 930 (w), 849 (w), 836 (w), 751 (m), 720
(s), 685 (s), 667 (m), 651 (m), 644 (m), 618 (m), 606 cmꢀ1 (m);
~
(ATR, 32 scans): n=3405 (w), 2752 (w), 2729 (w), 2674 (w), 2602
~
Raman (784 nm, 50 mW, 258C, 4 acc., 60 s): n=3169 (1), 3144 (1),
(w), 2096 (s), 2055 (s), 1987 (w), 1362 (m), 1313 (w), 1295 (w), 1226
(w), 1188 (w), 667 (m), 653 (m), 611 (w), 594 cmꢀ1 (w); Raman
3064 (6), 3009 (1), 2992 (1), 2956 (1), 2098 (2), 2038 (1), 2024 (1),
1586 (6), 1483 (1), 1441 (1), 1353 (2), 1341 (4), 1305 (2), 1287 (1),
1231 (1), 1189 (2), 1166 (2), 1110 (2), 1100 (3), 1075 (1), 1027 (4),
1001 (10), 933 (1), 753 (1), 727 (1), 681 (2), 616 (2), 531 (1), 476 (1),
450 (1), 395 (1), 294 (2), 266 (3), 253 (3), 198 cmꢀ1 (7); elemental
analysis calcd (%) for C24H20N9CdP: C 49.88, H 3.49, N 21.81; found:
C 49.48, H 3.73, N 21.68.
~
(784 nm, 21 mW, 258C, 5 acc., 60 s): n=2165 (1), 2134 (1), 2102 (1),
2096 (1), 2080 (1), 2044 (1), 1367 (9), 1312 (1), 1295 (1), 665 (1), 651
(1), 617 (1), 609 (1), 595 (1), 314 (1), 271 (2), 213 (4), 181 (5), 151 (6),
117 (10), 102 (6), 77 (6), 68 cmꢀ1 (8).
Cd(N3)2·3/2 DMSO
Cadmium(II) azide, Cd(N3)2 (0.04 g, 0.2 mmol) was dissolved in
DMSO (0.4 mL) at ambient temperature. The colorless solution was
concentrated in vacuo and a few drops of acetone were added
slowly to initiate crystallization. Storage at ambient temperature
for ten hours, resulted in the deposition of needle-like crystals. Re-
moval of supernatant by syringe and drying in vacuo for one hour
gave cadmium(II) azide DMSO solvate Cd(N3)2·3/2 DMSO as colorless
[PNP]2[Cd(N3)4]
Procedure 1: To a stirred suspension of bis-bis(triphenylphos-
phine)iminium tetraiodo cadmate [PNP]2[CdI4] (0.740 g, 0.44 mmol)
in acetonitrile (8 mL), silver azide, AgN3 (neat, 0.288 g, 1.92 mmol)
was added in one portion at ambient temperature. The yellowish
suspension was stirred for one hour and filtered (F4), resulting in
a colorless solution. The solvent was removed in vacuo and the
residue was dissolved in DMSO (0.6 mL). The colorless solution was
allowed to evaporate over a period of two days resulting in the
deposition of large colorless block-like crystals. The viscous super-
natant was removed by decantation and the crystals were washed
with a few drops of acetone and dried between two filter papers
for five minutes. Drying in vacuo (5 mbar) over P4O10 in a desiccator
for 12 h gave bis-bis(triphenylphosphine)iminium tetraazido
cadmate [PNP]2[Cd(N3)4] as colorless crystals (yield: 70–80%).
~
crystals (yield: 70–90%). M.p. 1208C; IR (ATR, 32 scans): n=3382
(w), 3340 (w), 3005 (w), 2915 (w), 2638 (w), 2546 (w), 2451 (w),
2051 (s), 1435 (m), 1405 (m), 1343 (m), 1305 (m), 1293 (m), 1236
(m), 1209 (w), 1023 (s), 1007 (s), 995 (s), 953 (s), 938 (s), 904 (m),
707 (m), 674 (m), 658 (m), 633 (m), 605 cmꢀ1 (m); Raman (784 nm,
~
21 mW, 258C, 2 acc., 60 s): n=3011 (1), 2918 (1), 2115 (1), 2077 (1),
2070 (1), 2060 (1), 1414 (1), 1348 (4), 1311 (2), 1296 (2), 1240 (2),
1210 (1), 1040 (2), 1030 (1), 1010 (2), 1002 (1), 958 (2), 940 (1), 714
(2), 708 (2), 679 (6), 383 (1), 360 (1), 341 (2), 315 (2), 232 (7),
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3660
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Procedure 2: Cadmium(II) azide Cd(N3)2 (0.049 g, 0.25 mmol) and
bis(triphenylphosphine)iminium azide [PNP]N3 (0.290 g, 0.50 mmol)
were combined and dissolved in DMSO (0.6 mL) at ambient tem-
perature. The colorless solution was allowed to evaporate over
a period of two days resulting in the deposition of large colorless
block-like crystals. The viscous supernatant was removed by de-
cantation and the crystals were washed with a few drops of ace-
tone and dried between two filter papers for five minutes. Drying
in vacuo (5 mbar) over P4O10 in a desiccator for 12 h gave bis-bis-
(triphenylphosphine)iminium tetraazido cadmate [PNP]2[Cd(N3)4] as
colorless crystals (yield: 70–80%). M.p. 2018C; 133Cd NMR (300 K,
[D6]DMSO, 111.00 MHz): d=ꢀ417 ppm (Dn1/2 =320 Hz); 14N NMR
(300 K, [D6]DMSO, 36.14 MHz): d=ꢀ133 (Nb, Dn1/2 =160 Hz),
analysis calcd (%) for C24H22N15OCd2P: C 36.38, H 2.80, N 26.52;
found: C 36.41, H 2.77, N 26.49.
[Ph4P][Hg(N3)3]
Procedure 1: To a stirred suspension of tetraphenylphosphonium
triiodo mercurate [Ph4P][HgI3] (0.276 g, 0.3 mmol) in acetonitrile
(10 mL), silver azide, AgN3 (neat, 0.148 g, 0.99 mmol) was added in
one portion at ambient temperature. The colorless suspension was
filtered (F4). Removal of solvent and drying in vacuo gave
tetraphenylphosphonium triazido mercurate [Ph4P][Hg(N3)3] as a
colorless microcrystalline solid (yield: 80–90%).
Procedure 2: To a stirred suspension of mercury(II) azide Hg(N3)
(0.142 g, 0.5 mmol) in acetonitrile (5 mL), tetraphenylphosphonium
azide [Ph4P]N3 (neat, 0.191 g, 0.5 mmol) was added in one portion
at ambient temperature. The colorless, slightly turbid solution was
filtered (F4), resulting in a colorless clear solution. Removal of sol-
vent and drying in vacuo gave tetraphenylphosphonium triazido
mercurate [Ph4P][Hg(N3)3] as a colorless microcrystalline solid (yield:
80–90%). M.p. 1478C; 199Hg NMR (300 K, [D6]DMSO, 89.58 MHz):
d=ꢀ1546 ppm (Dn1/2 =144 Hz); 14N NMR (300 K, [D6]DMSO,
~
ꢀ280 ppm (Ng, Dn1/2 =830 Hz); IR (ATR, 32 scans): n=3360 (w),
3350 (w), 3081 (w), 3057 (w), 3045 (w), 3025 (w), 3012 (w), 2992
(w), 2072 (m), 2037 (s), 1586 (w), 1574 (w), 1480 (w), 1435 (m),
1313 (m), 1288 (m), 1268 (s), 1181 (w), 1157 (w), 1110 (s), 1076 (w),
1025 (w), 998 (m), 929 (w), 865 (w), 850 (w), 838 (w), 797 (w), 757
(w), 747 (m), 741 (m), 720 (s), 690 (s), 670 (m), 664 (m), 645 (w), 615
(w), 548 (s), 532 cmꢀ1 (s); Raman (784 nm, 50 mW, 258C, 4 acc.,
~
120 s): n=3172 (1), 3147 (1), 3060 (9), 3010 (1), 2993 (1), 2956 (1),
2913 (1), 2073 (1), 2041 (1), 1589 (8), 1575 (3), 1482 (1), 1438 (1),
1338 (3), 1279 (1), 1183 (2), 1162 (2), 1111 (4), 1072 (1), 1027 (5),
1004 (10), 932 (1), 854 (1), 798 (1), 743 (1), 726 (1), 665 (3), 616 (2),
551 (1), 527 (1), 490 (1), 398 (1), 362 (1), 329 (2), 299 (1), 280 (1),
267 (2), 246 (3), 237 (3), 227 (1), 193 (2), 170 cmꢀ1 (2); elemental
analysis calcd (%) for C72H60N14CdP4: C 63.70, H 4.45, N 14.44;
found C 63.36, H 4.53, N 14.62.
36.14 MHz): d=ꢀ131 (Nb, Dn1/2 =47 Hz), ꢀ267 ppm (Ng, Dn1/2
=
~
428 Hz); IR (ATR, 32 scans): n=3299 (w), 3253 (w), 3055 (w), 2038
(s), 2006 (s), 1585 (w), 1483 (m), 1434 (m), 1314 (m), 1269 (m), 1183
(m), 1167 (w), 1159 (w), 1108 (s), 1074 (w), 1026 (w), 995 (m), 939
(w), 856 (w), 840 (w), 754 (m), 722 (s), 688 (s), 649 (w), 638 (w), 615
(w), 593 cmꢀ1 (w); Raman (784 nm, 50 mW, 258C, 4 acc., 45 s): n=
~
3171 (1). 3147 (1), 3094 (2), 3066 (2), 3056 (1), 2077 (2), 2045 (1),
2008 (1), 1586 (4), 1575 (1), 1485 (1), 1442 (1), 1435 (1), 1315 (2),
1274 (1), 1266 (1), 1185 (1), 1169 (1), 1159 (1), 1109 (1), 1099 (3),
1075 (1), 1025 (10), 999 (1), 931 (1), 724 (1), 679 (3), 649 (1), 615 (3),
530 (1), 448 (1), 394 (1), 344 (6), 291 (2), 250 cmꢀ1 (5); elemental
analysis calcd (%) for C24H20N9HgP: C 43.28, H 3.03, N 18.93; found:
C 43.54, H 3.32, N 18.61.
[Ph4P][Cd2(N3)5(H2O)]
To a stirred aqueous solution of cadmium(II) azide Cd(N3)2 (pre-
pared by the addition of a solution of sodium azide NaN3 (0.488 g,
7.50 mmol) in water (dist., 5 mL) to a solution of cadmium(II)
sulfate·8/3 hydrate CdSO4·8/3 H2O (0.128 g, 0.5 mmol) in water (dist.,
5 mL)), a solution of tetraphenylphosphonium chloride [Ph4P]Cl
(0.124 g, 0.33 mmol) in water (dist., 3 mL) was added dropwise
with rapid stirring, resulting in a colorless precipitate. The clear su-
pernatant was removed by centrifugation and decantation and the
residue was washed with water (dist., 3 mL). The supernatant was
again removed by centrifugation and decantation. For recrystalliza-
tion, the colorless residue was dissolved in boiling water (dist.,
about 15 mL) and filtered hot (F4). The resulting clear colorless so-
lution was allowed to slowly cool to 58C, resulting in the precipita-
tion of needle-like crystals. Removal of the supernatant by decanta-
tion and drying in air, followed by storage over P4O10 for two days
at ambient temperature (or heating to 808C in vacuo for about
6 h) gave tetraphenylphosphonium pentaazido dicadmate hydrate
[Ph4P][Cd2(N3)5(H2O)] as colorless crystals (yield: 50–60%). M.p.
2668C; 133Cd NMR (300 K, [D6]DMSO, 111.00 MHz): d=ꢀ540 (Dn1/
2 =310 Hz); 14N NMR (300 K, [D6]DMSO, 36.14 MHz): d=ꢀ132 (Nb,
[Ph4P]2[Hg(N3)4]
Procedure 1: To a stirred suspension of mercury(II) azide Hg(N3)
(0.072 g, 0.25 mmol) in DMSO (1 mL), tetraphenylphosphonium
azide [Ph4P]N3 (neat, 0.191 g, 0.5 mmol) was added in one portion
at ambient temperature, resulting in a colorless clear solution.
Slow removal of solvent and drying in vacuo gave bis-tetraphenyl-
phosphonium tetraazido mercurate [Ph4P]2[Hg(N3)4] as a colorless,
crystalline solid (yield: 90–95%).
Procedure 2: To a stirred solution of tetraphenylphosphonium tria-
zido mercurate [Ph4P][Hg(N3)3] (0.120 g, 0.3 mmol) in DMSO (5 mL),
tetraphenylphosphonium azide [Ph4P]N3 (neat, 0.229 g, 0.6 mmol)
was added in one portion at ambient temperature. The colorless
solution was concentrated to an approximate volume of 3 mL in
vacuo. The pale rose solution was allowed to slowly evaporate at
ambient temperature in a desiccator for several hours, resulting in
the deposition of colorless crystals. Removal of solvent by decant-
ation and drying in vacuo gave bis-tetraphenylphosphonium
tetraazido mercurate [Ph4P]2[Hg(N3)4] as colorless crystals (yield:
80–90%).
~
Dn1/2 =230 Hz), ꢀ277 (Ng, Dn1/2 =1200 Hz); IR (ATR, 32 scans): n=
3357 (w), 3310 (w), 3082 (w), 3063 (w), 2099 (m), 2071 (m), 2052
(m), 2024 (s), 1585 (m), 1482 (m), 1435 (m), 1327 (m), 1298 (m),
1278 (m), 1228 (m), 1185 (m), 1164 (w), 1105 (s), 1072 (w), 1027 (w),
996 (m), 942 (w), 929 (w), 921 (w), 857 (w), 847 (w), 761 (m), 754
(m), 721 (s) 687 (s), 649 (w), 636 (w), 626 (w), 619 (w), 611 (w), 604
Procedure 3 (not for preparative purposes): To a stirred colorless
suspension of mercury(I) azide Hg2(N3)2 (0.097 g, 0.2 mmol) in
acetonitrile (5 mL), tetraphenylphosphonium azide [Ph4P]N3 (neat,
0.191 g, 0.5 mmol) was added in one portion at ambient tempera-
ture. The greyish suspension was stirred for one hour and filtered
(F4). The colorless clear solution was concentrated to incipient
crystallization in vacuo and stored at ambient temperature for ten
hours. Removal of solvent by decantation and drying in vacuo
(w), 597 cmꢀ1 (w); Raman (633 nm, 2 mW, 258C, 8 acc., 20 s): n=
~
3166 (1). 3144 (1), 3084 (1), 3062 (3), 3047 (1), 3006 (1), 2989 (1),
2955 (1), 2134 (2), 2073 (1), 2055 (1), 2045 (1), 2032 (1), 1583 (4),
1573 (1), 1479 (1), 1433 (1), 1331 (2), 1303 (2), 1275 (1), 1228 (1),
1183 (1), 1160 (1), 1103 (1), 1093 (3), 1024 (4), 997 (10), 982 (1), 934
(1), 927 (1), 917 (1), 721 (1), 692 (1), 675 (3), 611 (2), 523 (1), 477 (1),
390 (1), 290 (1), 257 (3), 248 (4), 236 (3), 193 cmꢀ1 (6); elemental
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3661
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
gave
bis-tetraphenylphosphonium
tetraazido
mercurate
(w), 3023 (w), 2991 (w), 2050 (w), 2007 (s), 1586 (w), 1573 (w), 1480
(w), 1433 (m), 1313 (m), 1287 (m), 1267 (s), 1180 (m), 1158 (m),
1110 (s), 1074 (m), 1025 (m), 997 (m), 929 (w), 863 (w), 851 (w), 838
(w), 798 (m), 757 (m), 746 (m), 741 (m), 720 (s), 690 (s), 664 (m),
615 (m), 548 (s), 532 cmꢀ1 (s); Raman (784 nm, 0.13 mW, 258C,
[Ph4P]2[Hg(N3)4] as colorless crystals (yield: 30–40%). M.p. 1438C;
199Hg NMR (300 K, [D6]DMSO, 89.58 MHz): d=ꢀ1407 ppm (Dn1/2
=
450 Hz); 14N NMR (300 K, [D6]DMSO, 36.14 MHz): d=ꢀ131 (Nb, Dn1/
~
2 =40 Hz), ꢀ270 ppm (Ng, Dn1/2 =300 Hz); IR (ATR, 32 scans): n=
~
3297 (w), 3171 (w), 3079 (w), 3061 (w), 3011 (w), 2993 (w), 2049
(m), 2007 (s), 1585 (w), 1483 (m), 1435 (s), 1340 (w), 1310 (m), 1265
(m), 1188 (w), 1164 (w), 1104 (s), 1028 (m), 996 (m), 932 (w), 856
(w), 846 (w), 753 (s), 717 (s), 684 (s), 640 (m), 616 (m), 603 cmꢀ1 (w);
6 acc., 15 s): n=3061 (2), 2050 (1), 2026 (1), 2014 (1), 1589 (3), 1576
(1), 1483 (1), 1438 (1), 1319 (1), 1268 (1), 1181 (1), 1163 (1), 1110 (4),
1071 (1), 1028 (4), 1004 (10), 999 (2), 936 (1), 854 (1), 799 (1), 742
(1), 726 (2), 695 (1), 665 (5), 640 (1), 617 (2), 552 (1), 525 (1), 490 (1),
399 (1), 364 (1), 332 (4), 280 (1), 266 (3), 253 (2), 243 (2), 236 (4),
192 (3), 171 cmꢀ1 (2); elemental analysis calcd (%) for
C72H60N14HgP4: C 59.81, H 4.18, N 13.56; found: C 59.29, H 4.20, N
13.96.
~
Raman (1064 nm, 500 mW, 258C, 500 acc.): n=3171 (1), 3146 (1),
3065 (6), 3010 (1), 2993 (1), 2962 (1), 2054 (3), 2020 (1), 2010 (1),
2001 (1), 1587 (8), 1577 (3), 1485 (1), 1439 (1), 1323 (2), 1268 (1),
1191 (2), 1167 (2), 1110 (1), 1098 (2), 1029 (6), 1002 (10), 930 (1),
757 (1), 727 (1), 679 (3), 643 (1), 617 (2), 531 (1), 461 (1), 394 (1),
331 (7), 286 (2), 255 (3), 203 cmꢀ1 (3); elemental analysis calcd (%)
for C48H40N12HgP2: C 55.04, H 3.85, N 16.05; found: C 54.80, H 3.83,
N 15.86.
[PNP][Hg(N3)2(CH3CN4)]
The mother liquor from crystallization of [PNP]2[Hg(N3)4] (obtained
by the reaction of mercury(II) azide and [PNP]N3 in acetonitrile)
was concentrated in vacuo and stored for two days at 58C, result-
ing in the deposition of needle-like crystals. The crystalline residue
was warmed to ambient temperature to partly dissolve the solid.
The supernatant was removed by decantation and the crystals
were dried in vacuo which gave bis(triphenylphosphine)iminium
[PNP][Hg(N3)3]
Mercury(II) azide Hg(N3)2 (0.057 g, 0.20 mmol) and bis(triphenyl-
phosphine)iminium azide [PNP]N3 (0.080 g, 0.21 mmol) were com-
bined and dissolved in acetonitrile (2 mL) at ambient temperature.
The colorless solution was allowed to evaporate in air over
a period of 10 h resulting in the deposition of colorless needle-like
crystals, surrounded by a solidified oil. The crystals were separated,
transferred to a Schlenck tube and dried in vacuo which gave bis-
(triphenylphosphine)iminium triazido mercurate [PNP][Hg(N3)3] as
colorless crystalline solid (yield: 60–70%). M.p. 1198C; 199Hg NMR
(300 K, [D6]DMSO, 89.58 MHz): d=ꢀ1539 ppm (Dn1/2 =640 Hz);
diazido(5-methyl-1H-tetrazol-1-yl)mercurate
acetonitrile
hemisolvate [PNP][Hg(N3)2(CH3CN4)]·0.5CH3CN as colorless crystals.
Solvent-free crystals of [PNP][Hg(N3)2(CH3CN4)] were obtained by
recrystallization from DMSO at ambient temperature. Yields were
not determined.
[PNP][Hg(N3)2(CH3CN4)]·0.5CH3CN: M.p. 1228C (668C loss of
CH3CN); 199Hg NMR (300 K, [D6]DMSO, 89.58 MHz): d=ꢀ1420 ppm
(Dn1/2 =1380 Hz); 1H-15N-HMBC (300 K, [D6]DMSO, 500.13,
50.69 MHz): d=ꢀ80 (CH3CN2N2), 8.3 ppm (CH3CN2N2); 14N NMR
(300 K, [D6]DMSO, 36.14 MHz): d=ꢀ131 (Nb, Dn1/2 =51 Hz),
ꢀ267 ppm (Ng, Dn1/2 =380 Hz); 13C NMR (300 K, [D6]DMSO,
125.8 MHz): d=1.08 (s, CH3CN), 9.52 (s, CH3CN4), 118.0 (s, CH3CN),
14N NMR (300 K, [D6]DMSO, 36.14 MHz): d=ꢀ131 (Nb, Dn1/2
=
~
46 Hz), ꢀ266 ppm (Ng, Dn1/2 =400 Hz); IR (ATR, 32 scans): n=3309
(w), 3279 (w), 3075 (w), 3056 (w), 3043 (w), 3025 (w), 3010 (w),
2990 (w), 2071 (m), 2041 (s), 2021 (s), 2000 (s), 1587 (w), 1481 (m),
1436 (m), 1305 (m), 1297 (m), 1232 (s), 1179 (m), 1160 (m), 1109 (s),
1027 (m), 996 (m), 981 (m), 929 (m), 893 (w), 857 (w), 847 (w), 804
(m), 759 (m), 744 (m), 720 (s), 689 (s), 666 (m), 616 (m), 606 (m),
591 (m), 532 cmꢀ1 (s); Raman (784 nm, 50 mW, 258C, 4 acc., 60 s):
1
3
126.8 (dd, J(13C–31P)=107.2 Hz, J(13C–31P)=1.71 Hz, ipso-Ph), 129.5
(m, meta-Ph), 131.9 (m, ortho-Ph), 133.6 (s, para-Ph), 156.9 ppm (s,
1
CH3CN4); H NMR (300 K, [D6]DMSO, 500.13 MHz): d=2.07 (s, 1.5H,
~
n=3061 (2), 2062 (1), 2036 (1), 1588 (2), 1577 (2), 1483 (1), 1440
CH3CN, 1J(13C–1H)=136 Hz), 2.38 (s, 3H, CH3CN4), 7.52–7.65 (m,
24H, ortho/meta-Ph), 7.70 ppm (m, 6H, para); IR (ATR, 32 scans):
(1), 1341 (2), 1335 (2), 1272 (1), 1181 (1), 1162 (2), 1110 (3), 1030 (4),
1002 (8), 933 (1), 850 (1), 804 (1), 850 (1), 727 (1), 667 (4), 654 (1),
615 (3), 551 (1), 525 (1), 493 (1), 449 (1), 373 (10), 282 (1), 268 (3),
249 (43), 238 (43), 229 (4), 207 cmꢀ1 (3); elemental analysis calcd
(%) for C36H30N13HgP2: C 49.97, H 3.49, N 16.19; found C 49.32, H
3.44, N 16.28.
~
n=3307 (w), 3090 (w), 3078 (w), 3057 (w), 3025 (w), 2992 (w), 2934
(w), 2286 (w), 2253 (w), 2022 (s), 1588 (w), 1574 (w), 1495 (w), 1482
(m), 1435 (s), 1378 (m), 1285 (s), 1251 (s), 1181 (m), 1159 (m), 1110
(s), 1070 (m), 1025 (m), 997 (m), 927 (w), 854 (w), 845 (w), 803 (m),
761 (m), 745 (m), 720 (s), 689 (s), 665 (m), 616 (w), 546 (s), 530 cmꢀ1
~
(s); Raman (784 nm, 21 mW, 258C, 4 acc., 75 s): n=3178 (1), 3148
[PNP]2[Hg(N3)4]
(1), 3064 (3), 3013 (1), 2995 (1), 2937 (1), 2258 (1), 2252 (1), 2052
(1), 2046 (1), 1590 (4), 1577 (2), 1498 (1), 1487 (1), 1441 (1), 1391
(1), 1376 (1), 1326 (1), 1274 (1), 1253 (1), 1241 (1), 1183 (1), 1162
(1), 1112 (4), 1072 (1), 1029 (5), 1003 (10), 928 (1), 848 (1), 804 (1),
761 (1), 748 (1), 725 (1), 700 (1), 691 (21), 665 (5), 648 (1), 617 (5),
552 (1), 527 (1), 495 (1), 439 (1), 402 (1), 370 (1), 331 (5), 283 (1),
270 (2), 249 (4), 239 cmꢀ1 (4); elemental analysis calcd (%) for
C39H34.5N11.5HgP2: C 50.54, H 3.75, N 17.38; found C 50.07, H 3.90, N
17.61.
Mercury(II) azide Hg(N3)2 (0.057 g, 0.20 mmol) and bis(triphenyl-
phosphine)iminium azide [PNP]N3 (0.238 g, 0.41 mmol) were com-
bined and dissolved in acetonitrile (2 mL) at ambient temperature,
resulting in a clear colorless solution. The solution was concen-
trated to incipient crystallization in vacuo (ca. 1 mL) and stored at
ambient temperature followed by cooling at 58C for ten hours,
resulting in the deposition of large colorless block-like crystals. The
supernatant was removed by decantation and the crystals
were dried in vacuo which gave bis-bis(triphenylphosphine)imini-
um tetraazido mercurate [PNP]2[Hg(N3)4] as colorless crystals. Fur-
ther concentration of the mother liquor and storage at 58C for
20 h gave a second crop of crystals (combined yield: 80–90%).
M.p. 1978C; 199Hg NMR (300 K, [D6]DMSO, 89.58 MHz): d=
~
[PNP][Hg(N3)2(CH3CN4)]: M.p. 1228C; IR (ATR, 32 scans): n=3295
(w), 3093 (w), 3077 (w), 3058 (w), 3025 (w), 2992 (w), 2936 (w),
2059 (m), 2038 (s), 2026 (s), 1587 (w), 1574 (w), 1481 (m), 1435 (s),
1363 (m), 1307 (m), 1274 (m), 1244 (s), 1181 (m), 1162 (m), 1108 (s),
1065 (m), 1025 (m), 997 (m), 929 (w), 853 (w), 845 (w), 802 (m), 758
(m), 746 (m), 719 (s), 689 (s), 666 (m), 616 (w), 551 cmꢀ1 (s); Raman
ꢀ1387 ppm (Dn1/2 =366 Hz);
14N NMR
(300 K,
[D6]DMSO,
~
36.14 MHz): d=ꢀ131 (Nb, Dn1/2 =40 Hz), ꢀ270 ppm (Ng, Dn1/2
=
(633 nm, 5 mW, 258C, 5 acc., 20 s): n=3174 (1), 3148 (1), 3065 (6),
3027 (1), 3010 (1), 2993 (1), 2957 (1), 2940 (1), 2062 (1), 2042 (1),
~
276 Hz); IR (ATR, 32 scans): n=3295 (w), 3078 (w), 3056 (w), 3045
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3662
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[11] a) M. Bassiꢀre, Compt. Rend. 1937, 204, 1573–1574; b) O. Reckeweg, A.
Simon, Z. Naturforsch. B 2003, 58, 1097–1104.
2030 (1), 1588 (4), 1576 (1), 1492 (1), 1449 (1), 1441 (1), 1388 (1),
1322 (1), 1278 (1), 1243 (1), 1185 (1), 1166 (1), 1111 (3), 1067 (1),
1030 (5), 1002 (10), 945 (1), 927 (1), 855 (1), 845 (1), 804 (1), 790 (1),
746 (1), 724 (1), 692 (2), 668 (3), 661 (2), 618 (2), 554 (1), 527 (1),
496 (1), 385 (1), 371 (1), 351 (5), 319 (1), 281 (1), 251 (4), 238 cmꢀ1
(4).
[12] a) I. Agrell, Acta. Chem. Scand. 1970, 24, 3575–3589; b) N. Mondal, M. K.
Saha, S. Mitra, V. Gramlich, J. Chem. Soc. Dalton Trans. 2000, 3218–
3221; c) M. A. S. Goher, F. A. Mautner, M. A. M. Abu-Youssef, A. K. Hafez,
A. M.-A. Badr, J. Chem. Soc. Dalton Trans. 2002, 3309–3312; d) M. A. S.
Goher, F. A. Mautner, A. K. Hafez, M. A. M. Abu-Youssef, C. Gspan, A. M.-
A. Badr, Polyhedron 2003, 22, 975–979; e) F. A. Mautner, C. Gspan,
M. A. S. Goher, M. A. M. Abu-Youssef, Monatsh. Chem. 2005, 136, 107–
117; f) A. Kriza, M. Tatucu, J. Coord. Chem. 2011, 64, 195–201.
[13] a) W. Clegg, H. Krischner, A. I. Saracoglu, G. M. Sheldrick, Z. Kristallogr.
1982, 161, 307–313; b) H. Haselmair, H. Krischner, Z. Anorg. Allg. Chem.
1980, 468, 76–77.
Acknowledgements
Dipl. Chem. Alexander Hinz (University Rostock), Dipl. Chem.
Fabian Reiß (University Rostock), and M.Sc. Kati Rosenstengel
(Univeristy Rostock) are acknowledged for helpful advice. Dr.
Dirk Michalik (University Rostock) is acknowledged for NMR
measurements. We thank the DFG (grant SCHU 1170 8-1), the
University of Rostock, and the Leibniz Institute for Catalysis for
financial support.
[14] a) S. Hanna, F. A. Mautner, B. Koppelhuber-Bitschnau, M. A. M. Abu-Yous-
sef, Mater. Sci. Forum 2000, 321–324, 1098–1101; b) Z.-Y. Du, Y.-P. Zhao,
W.-X. Zhang, H.-L. Zhou, C.-T. He, W. Xue, B.-Y. Wang, X.-M. Chen, Chem.
Commun. 2014, 50, 1989–1991.
[15] H. Krischner, C. Kratky, H. E. Maier, Z. Kristallogr. 1982, 161, 225–229.
[16] Full set of experimental data for all considered Cd and Hg azides can
be found in the Supporting Information. Synthesis and full characteriza-
tion of all starting materials including X-ray structure elucidation of
[PNP][CdI3], [PNP]2[CdI4], [Ph4P][HgI3] and [Ph4P]2[HgI4]·CH2Cl2.
[17] D. W. M. Hofmann, Acta Crystallogr. B 2002, 58, 489–493.
[18] a) B. Neumꢃller, F. Schmock, S. Schlecht, K. Dehnicke, Z. Anorg. Allg.
Chem. 2000, 626, 1792–1796; b) R. Haiges, A. Vij, J. A. Boatz, S. Schneid-
er, T. Schroer, M. Gerken, K. O. Christe, Chem. Eur. J. 2004, 10, 508–517;
c) K. Rosenstengel, A. Schulz, A. Villinger, Inorg. Chem. 2013, 52, 6110–
6126.
Keywords: azides · cadmium · cycloaddition · mercury ·
structure elucidation
[1] a) I. C. Tornieporth-Oetting, T. M. Klapçtke, Angew. Chem. Int. Ed. Engl.
1995, 34, 511–520; Angew. Chem. 1995, 107, 559–568; b) W. P. Fehl-
hammer, W. Beck, Z. Anorg. Allg. Chem. 2013, 639, 1053–1082.
[2] a) T. Curtius, Chem. Ber. 1890, 23, 3023–3033; b) M. Berthelot, P. Vieille,
Bull. Soc. Chim. Paris 1894, 11, 744–748.
[19] A. Villinger, A. Schulz, Angew. Chem. Int. Ed. 2010, 49, 8017–8020;
Angew. Chem. 2010, 122, 8190–8194.
[20] a) R. S. Berry, J. Chem. Phys. 1960, 32, 933–940; b) M. Sandstrçm, I. Pers-
son, S. Ahrland, Acta. Chem. Scand. A 1978, 32, 607–625.
[21] a) J. D. Kennedy, W. McFarlane, J. Chem. Soc. Perkin Trans. 2 1977, 1187–
1191; b) R. J. Kostelnik, A. A. Bothner-By, J. Magn. Reson. (1969–1992)
1974, 14, 141–151; c) M. A. Sens, N. K. Wilson, P. D. Ellis, J. D. Odom, J.
Magn. Reson. (1969–1992) 1975, 19, 323–336.
[3] a) D. Seybold, K. Dehnicke, Z. Anorg. Allg. Chem. 1968, 361, 277–283;
b) P. Nockemann, U. Cremer, U. Ruschewitz, G. Meyer, Z. Anorg. Allg.
Chem. 2003, 629, 2079–2082; c) B. L. Evans, A. D. Yoffe, P. Gray, Chem.
Rev. 1959, 59, 515–566; d) L. Wçhler, W. Krupko, Chem. Ber. 1913, 46,
2045–2057; e) L. Wçhler, F. Martin, Angew. Chem. 1917, 30, 33–39.
[4] a) T. R. Musgrave, R. N. Keller, Inorg. Chem. 1965, 4, 1793–1796; b) U.
Mꢃller, Z. Anorg. Allg. Chem. 1973, 399, 183–192; c) F. D. Miles, J. Chem.
Soc. 1931, 2532–2542.
[22] G. Klose, F, Volke, G. Peinel, G. Knobloch, Magn. Reson. Chem. 1993, 31,
548–551.
[23] Holleman Wiberg, Lehrbuch der Anorganischen Chemie, 102. Aufl., Walter
de Gruyter, Berlin, 2007, Anhang IV.
[5] H. Lund, O. Oeckler, T. Schrçder, A. Schulz, A. Villinger, Angew. Chem. Int.
Ed. 2013, 52, 10900–10904.
[24] F. A. Mautner, H. Krischner, C. Kratky, Z. Kristallogr. 1985, 172, 291–296.
[25] P. Pyykkç, M. Atsumi, Chem. Eur. J. 2009, 15, 12770–12779.
[26] J. E. Carpenter, F. Weinhold, J. Mol. Struct. 1988, 169, 41–62.
[27] F. Weinhold, J. E. Carpenter, The Structure of Small Molecules and Ions,
1988, Plenum, pp. 227–236.
[28] F. Weinhold, C. Landis, Valency and Bonding: A Natural Bond Orbital
Donor–Acceptor Perspective, 2005, Cambridge University Press, and ref-
erences therein.
[6] W. Strecker, E. Schwinn, J. Prakt. Chem. 1939, 152, 205–218.
[7] a) A. Perret, R. Perrot, Helv. Chim. Acta 1933, 16, 848–857; b) K. Deh-
nicke, J. Strꢄhle, D. Seybold, J. Mꢃller, J. Organomet. Chem. 1966, 6,
298–300; c) K. Dehnicke, D. Seybold, J. Organomet. Chem. 1968, 11,
227–241; d) A. F. Shihada, K. Dehnicke, J. Organomet. Chem. 1971, 26,
157–160; e) D. J. Brauer, H. Bꢃrger, G. Pawelke, K. H. Flegler, A. Haas, J.
Organomet. Chem. 1978, 160, 389–401; f) T. M. Klapçtke, B. Krumm, R.
Moll, Z. Anorg. Allg. Chem. 2011, 637, 507–514; g) T. M. Klapçtke, B.
Krumm, R. Moll, Eur. J. Inorg. Chem. 2011, 422–428.
[29] D. Figgen, G. Rauhut, M. Dolg, H. Stoll, Chem. Phys. 2005, 311, 227–244.
[30] K. A. Peterson, C. Puzzarini, Theor. Chem. Acc. 2005, 114, 283–296.
[8] W. Beck, W. P. Fehlhammer, P. Pçllmann, E. Schuierer, K. Feldl, Chem. Ber.
1967, 100, 2335–2361.
[9] a) T. Curtius, J. Rissom, J. Prakt. Chem. 1898, 58, 261–309; b) L. Bircken-
bach, Z. Anorg. Allg. Chem. 1933, 214, 94–96.
Received: November 8, 2014
[10] a) F. P. Bowden, K. Singh, Proc. R. Soc. London Ser. A 1954, 227, 22–37;
b) F. Karau, W. Schnick, Z. Anorg. Allg. Chem. 2005, 631, 2315–2320.
Published online on January 22, 2015
Chem. Eur. J. 2015, 21, 3649 – 3663
www.chemeurj.org
3663
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim