Crystal and molecular structures of the pentafluorosulfanyl compounds, SF5X (X = -NC, -CN, -NCO, -NCS and -NCCl2) by X-ray diffraction at low temperature

Article abstract of DOI:10.1002/zaac.200400070

A new method for the synthesis of the unstable pentafluorosulfanyl isocyanide from N-(pentafluorosulfanyl)(dichloromethanimine) was developed, thereby allowing for the study of its spectroscopic data. The structure of pentafluorosulfanyl isocyanide was determined by X-ray crystallography at 113 K. The molecule possesses an almost linear S-N-C arrangement and an unexpectedly long S-N bond. In addition, the structures of pentafluorosulfanyl cyanide, pentafluorosulfanyl isocyanate, pentafluorosulfanyl isothiocyanate, and N-(pentafluorosulfanyl)(dichloro)methanimine were determined by single crystal X-ray crystallography at low temperatures.

Full text of DOI:10.1002/zaac.200400070

Crystal and Molecular Structures of the Pentafluorosulfanyl Compounds, SF5X
(X ؍ -NC, -CN, -NCO, -NCS and -NCCl2) by X-ray Diffraction at Low
Temperature
Jürgen Buschmanna, Dieter Lentz*,b, Peter Lugera, Genivaldo Perpetuoa, Dagmar Preugschatb,c, Joseph S.
Thrasherc, Helge Willnerd, and Hans-Jörg Wölkb
a Berlin, Freie Universität, Institut für Chemie Ϫ Kristallographie
b Berlin, Freie Universität, Institut für Chemie Ϫ Anorganische und Analytische Chemie
c Tuscaloosa, Alabama, The University of Alabama, Department of Chemistry
d Wuppertal, Fachbereich 9 Ϫ Anorganische Chemie Ϫ der Universität
Received March 12th, 2004.
Abstract. A new method for the synthesis of the unstable penta-
fluorosulfanyl isocyanide from N-(pentafluorosulfanyl)(dichloro-
methanimine) was developed, thereby allowing for the study of its
spectroscopic data. The structure of pentafluorosulfanyl isocyanide
was determined by X-ray crystallography at 113 K. The molecule
possesses an almost linear S-N-C arrangement and an unexpectedly
long S-N bond. In addition, the structures of pentafluorosulfanyl
cyanide, pentafluorosulfanyl isocyanate, pentafluorosulfanyl isothi-
ocyanate, and N-(pentafluorosulfanyl)(dichloro)methanimine were
determined by single crystal X-ray crystallography at low tempera-
tures.
Keywords: Crystal Structure; Pentafluorsulfanyl compounds; Iso-
cyanides; Nitriles; Isocyanates; Isothiocyanates
Einkristall-Strukturanalysen der Pentafluorsulfanyl Verbindungen, SF5X
(X ؍ -NC, -CN, -NCO, -NCS and -NCCl2) bei tiefen Temperaturen
Inhaltsübersicht. Zur Synthese des instabilen Pentafluorsulfanyliso-
cyanids wurde eine neue Methode entwickelt. Die Struktur des Mo-
leküls wurde durch Röntgenstrukturanalyse bei 113 K bestimmt.
Das Molekül besitzt eine nahezu lineare S-N-C-Anordnung und
eine unerwartet lange S-N-Bindung. Die Strukturen von Penta-
fluorsulfanylcyanid, Pentafluorsulfanylisocyanat, Pentafluorsul-
fanylisothiocyanat und N-(pentafluorosulfanyl)(dichloro)methan-
imine wurden zusätzlich durch Einkristallröntgenstrukturanalysen
bei tiefen Temperaturen bestimmt.
Introduction
are stable gases at ambient temperature, and their chemistry
and physical properties were extensively studied [1]. Of
the pentafluorosulfanyl pseudohalides, F5S-NCO [3] and
F5S-NCS [3] are known since 1964, whereas others like
F5S-OCN [4], F5S-CN [5] and F5S-NC [6] were synthesized
more recently, and F5S-ONC or F5S-CNO are still un-
known.
The incorporation of highly electronegative substituent
groups, such as pentafluorosulfanyl, into a molecule causes
significant changes in its properties [1, 2]. The SF5 group is
sterically bulky because of the octahedral surrounding
about the sulfur atom which requires bond angles of about
90 degrees. Unlike sulfur hexafluoride, which is extremely
stable, the stability of pentafluorosulfanyl compounds is de-
pendent on the group that formally replaces one of the six
fluorine atoms.
The two known pentafluorosulfanyl halides SF5Cl and
SF5Br are the most important starting materials for the syn-
thesis of pentafluorosulfanyl compounds. Both compounds
* PD Dr. D. Lentz
Fachbereich Chemie, Biologie, Pharmazie
Institut für Chemie Ϫ Anorganische Chemie
Freie Universität Berlin
Fabeckstrasse 34-36
D-14195 Berlin
Germany
As expected F5S-NC is less stable than F5S-CN and in
contrast to CF3NC [7] isomerizes in the gas phase at ambi-
ent temperature even at low pressure [6]. However, accord-
ing to ab initio calculations CF3NC [8] and F5S-NC [9] have
comparable activation energies for the unimolecular iso-
lentz@chemie.fu-berlin.de
1136
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
DOI: 10.1002/zaac.200400070
Z. Anorg. Allg. Chem. 2004, 630, 1136Ϫ1142

Structures of the Pentafluorosulfanyl Compounds
merization to the corresponding nitriles. According to a PE
spectroscopic analysis the isomerization of trifluoromethyl
isocyanides at low pressure starts at 750 K and is about
50 % complete at 1000 K [10]. Whereas the chemistry [11]
and gas phase structure [12] of pentafluorosulfanyl cyanide
were studied, almost nothing is known about the isocyanide
isomer [6].
In this paper, we report on a more efficient synthesis of
pentafluorosulfanyl isocyanide and the single crystal X-ray
structure determinations of the isomeric pair F5S-NC and
F5S-CN. To complete the solid state structural data of
pentafluorosulfanyl pseudohalogens, the compounds F5S-
NCO, F5S-NCS and F5S-NCCl2 were included in this
study.
However, pentafluorosulfanyl isocyanide was formed by
using triisopropyl phosphite and directly removing the
volatile materials under vacuum. But, the material collected
in the trap cooled with liquid nitrogen contains variable
amounts of impurities that are difficult to remove due to
very similar volatility as well as the instability and the high
reactivity of F5S-NC. The major impurities are as follows:
thionyl fluoride, N-(pentafluorosulfanyl)(dichloro)-meth-
aneimine, pentafluorosulfanyl isothiocyanate, silicon tetra-
fluoride, isopropyl fluoride, and isopropyl chloride. The
presence of SOF2 and SiF4 could be suppressed by strictly
avoiding any traces of water and carefully drying the
complete apparatus under vacuum. Whereas the reaction
product isopropyl chloride could be easily separated by
fractional condensation, it is almost impossible to remove
all of the isopropyl fluoride that is due to a side reaction.
The presence of N-(pentafluorosulfanyl)(dichloro)methane-
imine is not due to an incomplete reaction; we found that
even a carefully purified sample of F5S-NC exhibits signals
due to F5S-NϭCCl2 after being in contact to Halocarbon
grease (trifluorovinyl chloride polymer). When a sample of
pentafluorosulfanyl isocyanide is stored at low pressure
(5 hPa) in a gas infrared cell with sodium chloride windows
fixed by Halocarbon wax the signal of F5S-NC (ν(NC)
2080 cmϪ1) disappears completely within several days and
a new band is observed at 1645 cmϪ1, whereas the low
frequency range of the spectrum seems to be virtually un-
changed. This is due to the formation of F5S-NϭCCl2 as
shown by additional analysis of the material by mass spec-
troscopy and 19F NMR spectroscopy.
Results and Discussion
The published multi step pathway for the synthesis of
pentafluorosulfanyl isoycanide outlined in Scheme 1 never
resulted in yields greater than 5 %.Thus we looked for other
methods that would allow its preparation from more readily
available starting materials along with attainment of better
yields. For example, in 1977 Knunyants et al. reacted
N-(trifluoromethyl)(difluoro)methaneimine
with
alkyl
phosphites in an oxidative addition reaction to form phos-
phoranes, which could be isolated and characterized [13].
Elimination of the alkylfluoride and the fluorophosphonic
acid ester yielded trifluoromethyl isocyanide.
2-Fluoropropene turned out to be the impurity that was
the most difficult to separate. Thus, we looked for phos-
phites which would result in less volatile fluoroalkanes.
A commercially available sample of triisodecyl phosphite
(Aldrich) turned out to be far too impure to be useful in
the preparation of pure pentafluorosulfanyl isocyanide.
Tris(2-octyl) phosphite, which could be prepared from 2-
octanol and phosphorus trichloride [14], gave F5S-NC in
higher purity but lower yields as its higher viscosity re-
quired a higher reaction temperature and longer reaction
times. Triphenyl phosphite reacts only very slowly with
F5S-NϭCCl2. Although several P-F containing compounds
were formed, no evidence for the existence of pentafluoro-
sulfanyl isocyanide was found. Therefore, the reaction of
F5S-NϭCCl2 with triisopropyl phosphite is currently the
best method for the preparation of F5S-NC, especially if the
inert 2-fluoropropane does not influence further reactions.
Further problems in the chemistry of pentafluorosulfanyl
isocyanide arise from its isomerization to the thermo-
dynamically more stable cyanide. According to ab initio cal-
culations on the MP2 level of theory and density functional
(B3LYP) calculations the activation energy for a unimolecu-
lar rearrangement is far too high to allow the isomerization
in the gaseous phase at ambient temperature [9]. In accord-
ance with these predictions an isomerization does not occur
in the gaseous phase even on heating as can be seen from
the infrared spectra taken of a gaseous sample at low press-
N-(Pentafluorosulfanyl)(dichloro)methaneimine, which
can be prepared in one step from cyanogen chloride and
pentafluorosulfanyl chloride [3], reacts(Scheme 2) with vari-
ous trialkyl phosphites well below ambient temperature.
However, it was not possible to isolate any phosphorane
intermediate during our attempts although we tried tri-
methyl, triethyl, triphenyl, triisopropyl, triisodecyl, and
tris(2-octyl) phosphites.
According to the NMR spectra, the reaction of F5S-Nϭ
CCl2 with triethyl phosphite or trimethyl phosphite resulted
in a complicated mixture of several fluoro phosphorus com-
pounds. No attempts were made to separate these com-
pounds as on warming no pentafluorosulfanyl isocyanide
could be detected among the volatile products.
Z. Anorg. Allg. Chem. 2004, 630, 1136Ϫ1142
zaac.wiley-vch.de
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1137

J. Buschmann, D. Lentz, P. Luger, G. Perpetuo, D. Preugschat, J. S. Thrasher, H. Willner, H.-J. Wölk
ure (5 hPa), which show no sign of isomerization but a slow
reaction with the Halocarbon  wax. However, a rapid iso-
merization takes place in concentrated solutions in deutero-
dichloromethane above Ϫ45 °C as can be seen from the
19F NMR and 13C NMR spectra, respectively. The typical
resonance of the isocyanide carbon atom at 154.4 ppm dis-
appears whereas the resonance of the nitrile carbon atom
at 96.59 ppm increases. A similar observation can be made
in the low-temperature Raman spectrum. The spectrum re-
mains unchanged between Ϫ150 °C and Ϫ70 °C. After
warming to Ϫ60 °C, the signal of the NC band at
2080 cmϪ1 disappeared completely and new bands at 2233
and 2254 cmϪ1 appeared. After 16 h at ambient tempera-
ture the colorless liquid sample had solidified and turned
red brown. A strong fluorescence prevented any further
study by Raman spectroscopy. The volatile materials were
collected in a trap kept at Ϫ196 °C and identified as penta-
fluorosulfanyl cyanide by comparison of its spectra with an
authentic sample.
Fig. 1 Molecular structures (ORTEP [19]) of pentafluorosulfanyl
isocyanide (left) and pentafluorosulfanyl cyanide (right). Ellipsoids
drawn at a 50 % probability scale. All atoms except F3 lie on a
crystallographic mirror plane at y ϭ 0.25. The atom F3Ј is genera-
ted by the symmetry operation x, 0.5 Ϫ y, z.
The asymmetric unit of both compounds consists of half
a molecule F5S-NC and F5S-CN, respectively. The mol-
ecules possess crystallographic Cs symmetry whereas the
molecular symmetry is almost C4v. All atoms except the flu-
orine atom F3 lie on the crystallographic mirror plane at
y ϭ 0.25.
Crystal Structures
As pentafluorosulfanyl isocyanide and pentafluorosulfanyl
cyanide are highly volatile materials and pentafluorosul-
fanyl isocyanide is unstable towards isomerization above
Ϫ70 °C, crystals of both compounds had to be obtained by
in situ crystallization from the melt in X-ray capillaries di-
rectly on the diffractometer. Due to the instability of penta-
fluorosulfanyl isocyanide several samples were crystallized.
Out of four data collections, two could be used to solve
and refine the structure. Pentafluorosulfanyl isocyanide and
pentafluorosulfanyl cyanide crystallize orthorhombic with
very similar lattice constants. The cell volume of the cyanide
Selected bond lengths and angles are summarized in
Table 1 in comparison to the geometry obtained from ab
initio calculations on the HF and MP2 level of theory,
respectively, and the values of a gas electron diffraction
study of F5S-CN [12]. In both compounds, the sulfur atom
is octahedrally surrounded by five fluorine atoms and one
nitrogen or carbon atom, respectively. Bond angles of cis
and trans substituents are very close to 90° and 180°, and
they are equal within 3 σ. There is no significant difference
in the distances of sulfur fluorine bonds in either the cis or
trans position to the functional group. The bond lengths
and angles of pentafluorosulfanyl cyanide obtained from X-
ray crystallography and a combination of microwave spec-
troscopy and gas electron diffraction [12] show an excellent
agreement and are consistent with the values obtained from
theoretical calculations at different levels of theory. The
situation differs for pentafluorosulfanyl isocyanide.
3
˚
is about 4 A smaller than that of the isocyanide. Systematic
absent reflections leave a choice between the centrosym-
metric space group Pnma (no. 62) or the corresponding
acentric space group Pn21a (no. 33). As the E- value statistic
indicated a centrosymmetric space group, the structures
were solved and refined in the centrosymmetric space
group. The molecular structures and the atomic numbering
schemes are depicted in Figure 1.
Table 1 Structural parameters of pentafluorosulfanyl isocyanide and pentafluorosulfanyl cyanide.
F5S-NC
F5S-CN
X-ray
X-ray
HF6/311ϩϩG* MP2/6-311G* X-ray
ED
HF6/311ϩϩG* MP2/6-311G*
crystal 1
1.710(7)
1.158(9)
crystal 2
1.693(8)
1.168(12)
S-N/S-C
C-N/N-C
S-Fcis
S-Ftrans
S-N-C/S-C-N
1.6550
1.1568
1.6812
1.1918
1.5950
1.5916
180
1.759(3)
1.137(5)
1.765(5) 1.7566
1.1248
1.7509
1.1477
1.6033
1.5926
180
1.522-1.543(5) 1.515-1.540(7) 1.5530
1.543(4)
179.5(6)
1.536-1.548(3) 1.564(6) 1.5571
1.553(2)
179.6(3)
1.546(5)
178.4(10)
178.9(5)
89.5-89.8(4)
89.5-90.6(3)
1.5531
180
1.556(6) 1.5491
180
180
90.1(2)
90
180
180
89.81
90
Ftrans-S-N/Ftrans-S-C 179.9(3)
180
89.44
90
180
89.87
90
180.0(2)
89.8-90.1(2)
89.5-90.5(2)
180
Fcis-S-Ftrans
Fcis-S-Fcis
89.7-90.4(3)
89.6-90.4(3)
89.77
90
1138
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
zaac.wiley-vch.de
Z. Anorg. Allg. Chem. 2004, 630, 1136Ϫ1142

Structures of the Pentafluorosulfanyl Compounds
Whereas the experimental S-F and N-C bond lengths are
similar or even shorter than the theoretical values, which
might be explained by libration effects, the experimental S-
˚
N bond length of 1.711(7) A (crystal 1) and 1.693(7) (crys-
tal 2) seems to be unexpectedly long.
As can be concluded from the structure determinations
˚
of F5S-CCH [16, 17] [S-C 1.736(6) A], F5S-CFϭCF2 [18]
˚
˚
[S-C 1.784(8) A], and F5S-CH2-SF5 [18] [S-C 1.823(6) A],
which show a shrinkage of the sulfur-carbon bond of about
3
˚
0.09 A on going from a sp hybridized carbon atom to a sp
carbon atom, an S-N bond involving a sp hybridized nitro-
gen atom should be shorter than that involving a sp2
hybridized nitrogen atom. Structural data for several com-
pounds containing the F5S-N moiety are available from gas
electron diffraction, microwave spectroscopy and X-ray dif-
fraction [19Ϫ31], with S-N bond lengths ranging from 1.66
Fig. 2 Molecular structure (ORTEP [19]) of pentafluorosulfanyl
isocyanate. Ellipsoids drawn at a 50 % probability scale. Molecule
A left and molecule B right differ mainly by the staggered and
eclipsed conformation, respectively.
˚
to 1.71 A, including F5S-NCO (1.668(6), ED) [19], F5S-Nϭ
Table 2 Structural parameters of pentafluorosulfanyl isocyanate
(molecule A and B), pentafluorosulfanyl isothiocyanate, and
N-(pentafluorosulfanyl)(dichloro)methanimine.
CϭN-CF3 (1.690(7), ED) [20], F5S-NϭSF2 (1.679(9), MW)
Ϫ
[21], F5S-NNO2
(1.681-1.709, X-ray) [22],F5S-NF2
˚
(1.698(2) A, ED) [28], and (F5S)2NX (X ϭ H, F) (S-N
F5S-NCO
F5S-NCS F5S-NϭCCl2
˚
1.685(5) A, ED) [29]. Unusually long S-N bond distances
were observed only in the case of extreme steric hindrance
(F5S)3N (S-N 1.829(6) A, ED) [30] and for F5S-NO2
Molecule A Molecule B
˚
S1ϪN1
1.659(2)
1.208(2)
1.149(2)
1.662(2)
1.212(2)
1.149(2)
1.675(2)
1.210(3)
1.675(2)
1.692(2)
1.258(2)
N1ϪC1
C1ϪO1(S2)
C1ϪCl1
C1ϪCl2
SϪFcis
(1.900(7), ED) [31]. Our X-ray crystal structure determi-
˚
nations of F5S-NϭCCl2 (S-N 1.693(2) A), F5S-NϭCϭS (S-N
1.7034(16)
1.7035 (16)
1.564(2)Ϫ
1.574(2)
1.562(2)
88.33(7)Ϫ
94.17(7)
˚
˚
1.675(2) A), and F5S-NϭCϭO (S-N 1.658(3), 1.662(3) A)
1.552Ϫ
1.572(2)
1.556(1)
89.5Ϫ
93.0(1)
128.7(1)
1.559Ϫ
1.576(2)
1.559(1)
88.5Ϫ
1.560Ϫ
1.577(2)
1.567(2)
89.8Ϫ
˚
gave values between 1.66 and 1.69 A for S-N bond lengths
between an SF5 group and a bent, two-coordinate nitrogen
atom. As unusually long S-N bond lengths had only been
observed in the case of strong steric hindrance such as in
SϪFtrans
N1ϪS1ϪFcis
93.9(1)
93.2(1)
S1ϪN1ϪC1
127.9(1)
172.8(2)
177.98(6)
Ϫ1.24(2)
129.2(2)
173.7(2)
178.25(7)
50.5(2)
127.03(11)
˚
N1ϪC1ϪO1(S2) 172.6(2)
F1ϪS1ϪN1 177.82(6)
F2ϪS1ϪN1ϪC1 56.87(2)
(F5S)3N, an upper limit of 1.66 A should be justified for a
176.03(8)
45.1(2)
S-N bond distance in F5S-NC from comparison with
known experimental data. Furthermore, the theoretical cal-
culations indicate S-N bond lengths of 1.6550 (HF) and
˚
1.6812 (MP2) A, which, however, show a greater depen-
dence on the level of theory than for pentafluorosulfanyl
cyanide. In accordance with the fact that the structure de-
terminations with two different samples gave S-N bond
lengths equal with respect to 3σ the unexpectedly long S-N
distance found in the X-ray structure analysis may be due to
an unknown amount of cyanide in the sample under study.
However, an analysis of the anisotropic displacement par-
ameters of C and N in comparison with those of F5SCN
shows no unusual values as can be seen qualitatively by
regarding the ORTEP of both compounds (Figure 1). Thus
the unusually long S-N distance may be real and explains
the ease of isomerization to the cyanide, although the ab
initio calculations gave no hint of a long S-N distance.
It was not possible to solve this problem by refinement
of a disorder model. An attempt to refine the structure in
the asymmetric space group Pn21a (No. 33) had no effect
on the S-N bond length, resulted in much larger estimated
standard deviations, and does not seem to be appropriate
as the E value statistic indicates a centrosymmetric space
group.
and N-(pentafluorosulfanyl)(dichloro)methanimine are de-
picted in Figures 2, 3 and 4, respectively. Selected bond
lengths and angles are summarized in Table 2. Crystals of
all three compounds were grown in situ from the melt di-
rectly on the diffractometer. All three compounds crys-
¯
tallize triclinic, space group P1. The unit cell contains 4
(F5S-NCO), 2 (F5S-NCS), and 2 (F5S-NϭCCl2) molecules,
respectively. Thus, the asymmetric unit of pentafluorosul-
fanyl isocyanate consists of two molecules that differ mostly
by the N-C-S-F torsion angle.
Molecule A adopts a staggered conformation with a
F2-S1-N1-C1 torsion angle of 56.9(2)°, and molecule B ad-
opts an eclipsed conformation with a F2-S1-N1-C1 torsion
angle of Ϫ1.2(2)°. In both conformers, the F1-S-F angles
are close to 90°. The N-S-F angles to the cis fluorine atoms
vary between 89.5(1)° and 93.0(1)° for molecule A and
88.5(1)° and 93.9(1)° for molecule B. The F1-S-N arrange-
ment is almost linear for both conformers. The S-N dis-
tances and bond lengths and angles of the isocyanate group
are very similar for both molecules. Most remarkably, the
S-F distances show no significance on the position cis or
trans to the isocyanate group, respectively.
The molecular structures of pentafluorosulfanyl isocyan-
ate (molecule A and B), pentafluorosulfanyl isothiocyanate,
Z. Anorg. Allg. Chem. 2004, 630, 1136Ϫ1142
zaac.wiley-vch.de
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1139

J. Buschmann, D. Lentz, P. Luger, G. Perpetuo, D. Preugschat, J. S. Thrasher, H. Willner, H.-J. Wölk
pentafluorosulfanyl isocyanide are very toxic. All reactions
should be carried out under a well-ventilated hood.
F5S-NϭCCl2 [3], F5S-NCO [32], F5S-NCS [32], and F5S-CN [11]
were prepared according to literature methods. Trimethyl phosphite
(Aldrich), triethyl phosphite (Aldrich), and triisopropyl phosphite
(Aldrich) were distilled before use. Triphenyl phosphite (Aldrich)
and triisodecyl phosphite (Aldrich) were used as received. Tris-
(2-octyl) phosphite was prepared from 2-octanol and PCl3 accord-
ing to literature methods [14]. All reactions were carried out under
dry argon by using standard Schlenk and vacuum techniques. Vol-
atile materials were handled in a conventional glass vacuum line,
and amounts were determined by PVT measurements. Moisture
Fig. 3 Molecular structure (ORTEP [19]) of pentafluorosulfanyl
sensitive compounds were handled in a dry box (Braun) under a
dry argon atmosphere. 1H, 13C, 19F, and 31P NMR spectra were
recorded using a JEOL FX 90Q or JEOL LAMBDA 400 instru-
ment with TMS or solvent signals (1H, 13C), CFCl3 (19F), H3PO4
(31P) as standards. Infrared spectra were taken on a Perkin Elmer
883, Nicolet 5SXC, or a Bruker Vector 22 instrument. Raman spec-
tra were recorded on a Bruker RFS 100 FT-Raman spectrometer.
Mass spectra were obtained on a Varian MAT 711 (80 eV).
isothiocyanate. Ellipsoids drawn at a 50 % probability scale.
Fig. 4 Molecular structure (ORTEP [19]) of N-(pentafluorosulfanyl)-
(dichloro)methanimine. Ellipsoids drawn at a 50 % probability
scale.
Fig. 5 Schematic representation of the apparatus for the synthesis
of F5S-NC
In the case of pentafluorosulfanyl isothiocyanate and
N-(pentafluorosulfanyl)(dichloro)-methanimine, only the
staggered conformation with F2-S1-N1-C1 torsion angles
of 50.5(2) and 45.1(2)° are observed in the solid state. The
S-N distances increase on going from F5S-NCO to F5S-
NCS and F5S-NϭCCl2 from 1.659(2) to 1.675(2) and
Pentafluorosulfanyl isocyanide, F5S-NC. The reactions were carried
out in an apparatus like the one shown in Figure 1. In a typical
experiment, the apparatus was evacuated and flame dried until a
constant vacuum below 10Ϫ3 mbar was obtained. The U-traps and
the cold finger were cooled by liquid nitrogen, light petroleum/
liquid nitrogen (Ϫ120 °C), and dry ice/acetone, respectively. The
valve between the reaction flask and the trap was closed, and the
flask was cooled with liquid nitrogen. F5S-NϭCCl2 (200 µL) and
triisopropyl phosphite (200 µL) were added through the septum
using a syringe. To avoid a leak, the valve between the septum and
the reaction flask was closed, and the valve to the U-trap system
was opened. The liquid nitrogen bath below the reaction flask was
replaced by a cold, empty Dewar, and the reaction mixture was
allowed to warm slowly to ambient temperature with stirring. The
contents of the traps were analyzed by gas IR spectroscopy. The
trap kept at Ϫ196 °C contained most of the F5S-NC (yield 10 to
15 %), SOF2 and 2-fluoro-propane, whereas the trap kept at
Ϫ120 °C contained the less volatile materials and some F5S-NC.
To avoid the formation of the very volatile 2-fluoropropane and 2-
chloropropane, a similar experiment using tris(2-octyl) phosphite
was carried out in the apparatus described above. First, tris(2-octyl)
phosphite (500 µL) and then F5S-NϭCCl2 (200 µL) were added to
˚
1.692(2) A, respectively, whereas the S-F distances are al-
most unaffected. The N-C bond distances of the isocyanate
and the isothiocyanate are equal within 2σ, whereas the
methanimine exhibits an expected longer NϭC bond in ac-
cordance with the different hybridization at the carbon
atom. The FϪSϪN bond angles of the three compounds
show only minor variations. The S-N-C bond angle vary
between 127.0(1) and 129.2(2)° for F5S-NϭCCl2 and F5S-
NCS, respectively. All molecules form typical molecular lat-
tices without any unusual intermolecular contacts.
Experimental Section
Caution! Some sulfur fluorine compounds are very toxic.
Phosphorus containing by-products in the synthesis of
1140
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
zaac.wiley-vch.de
Z. Anorg. Allg. Chem. 2004, 630, 1136Ϫ1142

Structures of the Pentafluorosulfanyl Compounds
Table 3 Crystal data and structure refinement for F5S-NC, F5S-CN, F5S-NCO, F5S-NCS and F5S-NϭCCl2.
F5S-NC crystal 1 F5S-NC crystal 2 F5S-CN
F5S-NCO
F5S-NCS
F5S-NϭCCl2
empirical formula
formula weight
temperature/K
CF5NS
153.08
113(2)
CF5NS
153.08
123(2)
CF5NS
153.08
107(2)
CF5NOS
169.08
125(2)
CF5NS2
185.14
120(2)
CCl2F5NS
223.98
207(2)
˚
wavelength/A
0.71069
colorless
5 x 0.5 x 0.5
orthorhombic
Pnma
9.623(4)
5.895(2)
8.278(3)
0.71069
colorless
5 x 0.5 x 0.5
orthorhombic
Pnma
9.659(3)
5.904(2)
8.262(2)
0.71069
colorless
5 x 0.5 x 0.5
orthorhombic
Pnma
9.494(5)
5.848(3)
8.391(4)
0.71069
colorless
5 x 0.5 x 0.5
triclinic
0.71069
colorless
5 x 0.5 x 0.5
triclinic
0.71069
Color
colorless
5 x 0.5 x 0.5
triclinic
Crystal size/mm3
Crystal system
space group
¯
¯
¯
P1
5.097(2)
P1
P1
5.582(3)
˚
a/A
4.851(2)
7.407(2)
7.946(2)
97.47(2)
90.67(2)
91.88(2)
282.90(16)
2
˚
b/A
9.525(4)
10.472(4)
73.27(3)
84.84(3)
84.94(3)
483.9(3)
4
7.665(2)
8.373(6)
73.80(3)
88.79(5)
70.38(2)
323.0(3)
2
˚
c/A
α/deg
β/deg
γ/deg
3
˚
V/A
Z
469.6(3)
4
2.165
471.1(2)
4
2.158
465.9(4)
4
2.183
ρcalcd/(g cmϪ3
)
2.321
0.705
2.173
0.956
2.303
1.348
µ/mmϪ1
0.699
0.697
0.705
θ range/deg
Index ranges
3.25 to 28.01
Ϫ12ՅhՅ12
0ՅkՅ7
Ϫ7ՅlՅ10
1461
3.24 to 27.04
Ϫ12ՅhՅ2
0ՅkՅ7
Ϫ10ՅlՅ10
681
3.24 to 45.08
Ϫ2ՅhՅ18
Ϫ11ՅkՅ1
Ϫ3ՅlՅ16
3873
2.04 to 47.53
Ϫ10ՅhՅ1
Ϫ19ՅkՅ19
Ϫ21ՅlՅ21
7820
2.59 to 29.99
Ϫ6ՅhՅ3
Ϫ10ՅkՅ10
Ϫ11ՅlՅ5
1460
2.95 to 42.51
Ϫ8ՅhՅ10
Ϫ13ՅkՅ14
Ϫ10ՅlՅ15
5073
Reflns collected
Reflns unique
623 (Rint ϭ 0.1357) 551 (Rint ϭ 0.0926) 2053 (Rint ϭ 0.0433) 7162 (Rint ϭ 0.0220) 1429 (Rint ϭ 0.0216) 4661 (Rint ϭ 0.0206)
Data / restraints / parameters 623 / 0 / 46
551 / 0 / 46
1.032
2053 / 0 / 46
1.037
7162 / 0 / 163
1.095
1429 / 0 / 82
1.149
4661 / 0 / 91
1.043
GOF s on F2
1.090
R1, wR2 [I>2sigma()]
R1, wR2 (all data)
ρmax/ρmin/e A
0.0705, 0.1714
0.1101, 0.1920
0.901/Ϫ0.574
0.0622, 0.1448
0.1600, 0.1918
0.552/Ϫ0.501
0.0580, 0.1785
0.1126, 0.2135
0.936/Ϫ0.638
0.0431, 0.1298
0.0515, 0.1360
0.964/Ϫ0.611
0.0327, 0.0869
0.0379, 0.0944
0.530/Ϫ0.293
0.0542, 0.1664
0.0848, 0.1802
1.260/Ϫ0.489
Ϫ3
˚
the flask upon cooling with liquid nitrogen. The traps and the cold
finger were kept at Ϫ196 °C, Ϫ100 °C, and Ϫ78 °C respectively.
The reaction mixture was allowed to warm very slowly with con-
stant stirring while keeping the temperature of the reflux condenser
at Ϫ78 °C to avoid the loss of F5S-NϭCCl2. F5S-NC was collected
in the trap kept at Ϫ196 °C, and this trap contained SOF2 as the
only minor impurity according to the IR spectrum.
direction. Finally the crystal was annealed at a preset temperature.
The crystals were then cooled slowly to the temperature of the
measurement listed in Table 3.
2
The data were Lp corrected and reduced to Fo .The structure solu-
tion was obtained by direct methods (SHELXS-86) [34] for all com-
pounds. The structures were refined by full-matrix least-squares re-
finement based on F2 (SHELXL-97) [35] with anisotropic thermal
displacement parameters. Crystal data, experimental conditions,
and refinement values are listed in Table 3. The resulting exper-
imental bond lengths and angles for the compounds are presented
in Tables 1 and 2 together with the corresponding values from the
ab initio calculations. The molecular structures and the corre-
sponding atomic numbering schemes are depicted in Figures 1, 2,
3, and 4, respectively.
Further details of the crystal structure investigations are available
from the Fachinformationszentrum Karlsruhe, D-76344 Eggen-
stein-Leopoldshafen (Germany), on quoting the depository num-
ber CSD- 413756 SF5NC crystal 1, CSD- 413757 SF5NC crystal
2, CSD-413755 SF5CN, CSD-413818 SF5NCO and CSD-413754
SF5NCS, the name of the authors, and citation of the paper. Full de-
tails (excluding structure factors) have been deposited at the Cam-
bridge Crystallographic Data Centre, 12 Union Rd., Cambridge CB2
1EZ, under the number CCDC-232317 SF5NCCl2. Copies may be
obtained free of charge on application to the Director (Telefax:
Int. ϩ12 23 33 60 33; e-mail: deposit@ccdc.cam.ac.uk).
Crystal structure determinations
F5S-CN, F5S-NC, F5S-NCO, F5S-NCS, and F5S-NϭCCl2 were
condensed into glass capillaries of 0.5 mm diameter and 0.01 mm
wall thickness by using a glass vacuum line. A column of 3Ϫ4 mm
of the liquid was cooled with liquid nitrogen, and the capillary was
flame sealed under vacuum at a length of 30 mm. After warming
to Ϫ78 °C in a dry ice acetone bath, the capillary was mounted on
an insulated arcless goniometer head while maintaining the sample
at Ϫ78 °C. This was especially true in the case of the isocyanide in
order to avoid its decomposition. The samples were mounted in a
stream of cold nitrogen gas of an integrated cooling device [33]
onto a computer-controlled Siemens four-circle single-crystal dif-
fractometer using MoKα radiation with Nb filter. Single crystals
were grown by setting the temperature of the gas stream to a few
degrees below the melting point of each compound, i.e. to
Ϫ127.5 °C (F5S-CN), Ϫ144.0 °C (F5S-NC), Ϫ140.0 °C (F5S-NCO),
Ϫ143 °C (F5S-NCS) and -53 °C (F5S-NϭCCl2), respectively,
(mp ϭ Ϫ123.0 °C for F5S-CN; Ϫ140.0 °C for F5S-NC; Ϫ136.0
for F5S-NCO; Ϫ138 °C for F5S-NCS; Ϫ48 °C for F5S-NϭCCl2,
respectively). The major part of the sample column was molten
from its lower side using a coaxial coil of heating wire. By using
an electronic control device for the heat output of the coil, the
phase borderface was very slowly moved in the opposite, downward
Acknowledgement. We thank the Deutsche Forschungsgemeinschaft
and the Freie Universität Berlin (FPS Topologische und elektroni-
sche Eigenschaften von Molekülverbindungen) for financial sup-
port. D. P. gratefully acknowledges a research scholarship by the
DFG, a Ph. D. scholarship was provided by the DAAD to G. P.
Z. Anorg. Allg. Chem. 2004, 630, 1136Ϫ1142
zaac.wiley-vch.de
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
1141

J. Buschmann, D. Lentz, P. Luger, G. Perpetuo, D. Preugschat, J. S. Thrasher, H. Willner, H.-J. Wölk
[22] M. E. Sitzmann, R. Gilardi, R. J. Butcher, W. M. Koppes, A.
References
G. Stern, J. S. Thrasher, N. J. Trivedi, Z.-Y. Yang, Inorg. Chem.
2000, 39, 843.
[23] S. G. Bott, M. Clark, J. S. Thrasher, J. L. Atwood, J. Crys-
tallogr. Spectrosc. Res. 1987, 17, 187.
[24] R. Gilardi, J. L. Flippen-Anderson, C. George, Acta Crrystal-
logr. Sect. C 1991, 47, 442.
[25] M. E. Sitzmann, R. Gilardi, J. Fluorine Chem. 1993, 63, 203.
[26] R. Damerius, D. Leopold, K. Seppelt,Z. Anorg. Allg. Chem.
1989, 578, 110.
[27] D. Preugschat, J. S. Thrasher, Z. Anorg. Allg. Chem. 1996,
622, 1411.
[28] J. Haase, H. Oberhammer, W. Zeil, O. Glemser, R. Mews,
Z. Naturforsch. 1971, 26a, 1333.
[29] A. Waterfeld, H. Oberhammer, R. Mews, Angew. Chem. 1982,
94, 389; Angew. Chem. Suppl. 1982, 834; Angew. Chem. Int.
Ed. Engl. 1982, 21, 354.
[1] R. D. Verma, R. L. Kirchmeier, J. M. Shreeve, Adv. Inorg.
Chem. 1994, 41, 125.
[2] D. Lentz, K. Seppelt, in Chemistry of Hypervalent Compounds,
K. Akida (Ed.), Wiley-VCH, New York 1999.
[3] C. W. Tullock, D. D. Coffman, E. L. Muetterties, J. Am. Chem.
Soc. 1964, 86, 357.
[4] A. Schmuck, K. Seppelt, Angew. Chem. 1987, 99, 138; Angew.
Chem. Int. Ed. Engl. 1987, 26, 134.
[5] O. Lösking, H. Willner, Angew. Chem. 1989, 101, 1283; An-
gew. Chem. Int. Ed. Engl. 1989, 28, 1255.
[6] J. S. Thrasher, K.V. Madappat, Angew. Chem. 1989, 101, 1284;
Angew. Chem. Int. Ed. Engl. 1989, 28, 1256.
[7] D. Lentz, J. Fluorine Chem. 1984, 24, 523.
[8] J. B. Moffat, Chem. Phys. Lett. 1978, 55, 125.
[9] K. V. Madappat, Dissertation, The University of Alabama,
Tuscaloosa, Alabama 1993.
[30] M. R. Choudhury, J. W. Harrell, Jr., J. B. Nielsen, J. S.
Thrasher, J. Chem. Phys. 1988, 88, 5353; J. B. Nielsen, J. S.
Thrasher, K. D. Robinson, S. G. Bott, H. Zhang, J. L. At-
wood, P. Zylka, H.Oberhammer, unpublished work.
[31] N. Lu, J. S. Thrasher, S. von Ahsen, H. Willner, D. Hnyk,
H. Oberhammer, D. Lentz, Abstracts of Papers, 224th ACS
National Meeting, Boston, MA, August 18-22, 2000 (2002),
FLUO-044.
[10] H. Bock, R. Dammel, D. Lentz, Inorg. Chem. 1984, 23, 1535.
[11] J. Jacobs, S. E. Ulic, H. Willner, G. Schatte, J. Passmore, S. V.
Sereda, T. S. Cameron, J. Chem. Soc., Dalton Trans. 1996, 383.
[12] J. Jacobs, G. S. McGrady, H. Willner, D. Christen, H. Ober-
hammer, P. Zylka, J. Mol. Struct. 1991, 245, 275.
[13] A. F. Gontar, N. A. Til’kunova, E. G. Bykhovskaya, I. L.
Knunyants I. L. Izv. Akad. Nauk SSSR Ser. Kim. 1977, 10,
2379; Bull. Acad. Sci. USSR Div. Chem. Sci. 1977, 26, 2212.
[14] S. E. Fry, N. J. Pinta, J. Org. Chem. 1984, 49, 4877.
[15] ORTEP3 for Windows Ϫ L. J. Farrugia, J. Appl. Crystallogr.
1997, 30, 565.
[32] J. S. Thrasher, J. L. Howell, A. F. Clifford, Chem. Ber. 1984,
117, 1707; J. S. Thrasher, M. Clark, J. B. Nielsen, C. Alvarado,
M. T. Anderson, Inorg. Synth. 1992, 29, 38.
[33] H. Dietrich, H. Dierks, Messtechnik (Braunschweig) 1970,
78, 184.
[16] A. G. Csaszar, K. Hedberg, R. J. Terjeson, G. L. Gard, Inorg.
Chem. 1987, 26, 955.
[34] SHELXS86 Ϫ Program for Crystal Structure Solution. G. M.
Sheldrick, Institut für Anorganische Chemie der Universität,
Tammanstrasse 4, D-3400 Göttingen, Germany, 1986;
SHELXS86 Ϫ G. M. Sheldrick, in “Crystallographic Comput-
ing 3,” G. M. Sheldrick, C. Krüger, R. Goddard, Eds.; Oxford
University Press: pp. 175Ϫ189, 1985.
[17] P. Zylka, D. Christen, H. Oberhammer, H. J. Mol. Struct.
1991, 249, 285.
[18] I. Weis, H. Oberhammer, G. L. Gard, R. Winter, K. Seppelt,
J. Mol. Struct. 1992, 269, 197.
[19] H. Oberhammer, K. Seppelt, R. Mews, J. Mol. Struct. 1983,
101, 325.
[20] E. O. John, H. G.Mack, H. Oberhammer, R. L. Kirchmeier,
J. M. Shreeve, Inorg. Chem. 1993, 32, 287.
[21] R. S. White, S. R. Bailey, J. D. Graybeal, J. S. Thrasher, M.
H. Palmer,J. Mol. Spectrosc. 1988, 129, 243.
[35] SHELX97 [Includes SHELXS97, SHELXL97, CIFTAB] Ϫ
Programs for Crystal Structure Analysis (Release 97-2). G. M.
Sheldrick, Institut für Anorganische Chemie der Universität,
Tammanstrasse 4, D-3400 Göttingen, Germany, 1998.
1142
 2004 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim
zaac.wiley-vch.de
Z. Anorg. Allg. Chem. 2004, 630, 1136Ϫ1142