Unprecedented conformational variability in main group inorganic chemistry: The tetraazidoarsenite and -antimonite salts A+[M(N 3)4]- (A = NMe4, PPh4, (Ph3P)2N; M = As, Sb), five similar salts, five different anion structures

Article abstract of DOI:10.1021/ic3021996

A unique example for conformational variability in inorganic main group chemistry has been discovered. The arrangement of the azido ligands in the pseudotrigonal bipyramidal [As(N₃)₄]^- and [Sb(N₃)₄]^- anions theoretically can give rise to seven different conformers which have identical MN₄ skeletons but different azido ligand arrangements and very similar energies. We have now synthesized and structurally characterized five of these conformers by subtle variations in the nature of the counterion. Whereas conformational variability is common in organic chemistry, it is rare in inorganic main group chemistry and is usually limited to two. To our best knowledge, the experimental observation of five distinct single conformers for the same type of anion is unprecedented. Theoretical calculations at the M06-2X/cc-pwCVTZ-PP level for all seven possible basic conformers show that (1) the energy differences between the five experimentally observed conformers are about 1 kcal/mol or less, and (2) the free monomeric anions are the energetically favored species in the gas phase and also for [As(N₃)₄]^- in the solid state, whereas for [Sb(N₃)₄]^- associated anions are energetically favored in the solid state and possibly in solutions. Raman spectroscopy shows that in the azide antisymmetric stretching region, the solid-state spectra are distinct for the different conformers, and permits their identification. The spectra of solutions are solvent dependent and differ from those of the solids indicating the presence of rapidly exchanging equilibria of different conformers. The only compound for which a solid with a single well-ordered conformer could not be isolated was [N(CH₃) ₄][As(N₃)₄] which formed a viscous, room-temperature ionic liquid. Its Raman spectrum was identical to that of its CH₃CN solution indicating the presence of an equilibrium of multiple conformers.

Full text of DOI:10.1021/ic3021996

Article
pubs.acs.org/IC
Unprecedented Conformational Variability in Main Group Inorganic
Chemistry: the Tetraazidoarsenite and -Antimonite Salts A+[M(N3)4]−
(A = NMe4, PPh4, (Ph3P)2N; M = As, Sb), Five Similar Salts, Five
Different Anion Structures
Ralf Haiges,* Martin Rahm, and Karl O. Christe*
Loker Hydrocarbon Research Institute and Department of Chemistry, University of Southern California, Los Angeles, California
90089, United States
S
* Supporting Information
ABSTRACT: A unique example for conformational variability in inorganic main group
chemistry has been discovered. The arrangement of the azido ligands in the pseudotrigonal
bipyramidal [As(N3)4]− and [Sb(N3)4]− anions theoretically can give rise to seven different
conformers which have identical MN4 skeletons but different azido ligand arrangements
and very similar energies. We have now synthesized and structurally characterized five of
these conformers by subtle variations in the nature of the counterion. Whereas
conformational variability is common in organic chemistry, it is rare in inorganic main
group chemistry and is usually limited to two. To our best knowledge, the experimental
observation of five distinct single conformers for the same type of anion is unprecedented.
Theoretical calculations at the M06-2X/cc-pwCVTZ-PP level for all seven possible basic
conformers show that (1) the energy differences between the five experimentally observed
conformers are about 1 kcal/mol or less, and (2) the free monomeric anions are the
energetically favored species in the gas phase and also for [As(N3)4]− in the solid state,
whereas for [Sb(N3)4]− associated anions are energetically favored in the solid state and possibly in solutions. Raman
spectroscopy shows that in the azide antisymmetric stretching region, the solid-state spectra are distinct for the different
conformers, and permits their identification. The spectra of solutions are solvent dependent and differ from those of the solids
indicating the presence of rapidly exchanging equilibria of different conformers. The only compound for which a solid with a
single well-ordered conformer could not be isolated was [N(CH3)4][As(N3)4] which formed a viscous, room-temperature ionic
liquid. Its Raman spectrum was identical to that of its CH3CN solution indicating the presence of an equilibrium of multiple
conformers.
recently, the synthesis and structural characterization of
[PPh4][As(N3)4] and [PPh4][Sb(N3)4] have also been re-
ported.10
INTRODUCTION
■
Conformational variability in inorganic main group chemistry is
rare and the number of observable conformers is usually limited
to two.1 In the course of our systematic study of group 15
tetraazidoanions, we have surprisingly discovered five(!) differ-
ent conformers for these anions which all have the same MN4
skeletons but differ in the arrangements of their azido ligands.
During the past two years, we have carried out an experimental
and theoretical study of the different conformers which are
summarized in this paper.
EXPERIMENTAL SECTION
■
Caution! While the compounds described in this manuscript are quite stable
and no explosion occurred during the course of this study, many polyazides
are extremely shock-sensitive and can explode violently upon the slightest
provocation. As a result, these materials should be handled only on a very
small scale using appropriate safety precautions.11 Ignoring safety
precautions may lead to serious injuries!
Polyazido compounds are of general interest for energetic
materials and have been the focus of numerous experimental
studies in recent years.2 Thus, the following binary group 15
polyazides have previously been synthesized: P(N3)3,3a P(N3)5,3a
[P(N3)6]−,3b,c As(N3)3,3d,e,4 [As(N3)4]−,5 As(N3)5,6 [As-
(N3)4]+,5 [As(N3)6]−,3f,5 Sb(N3)3,3g,4 [Sb(N3)4]−,5 [Sb-
(N3)5]2−,7 Sb(N3)5,6 [Sb(N3)4]+,5 and [Sb(N3)6]−,5,6 Bi-
Materials and Apparatus. All reactions were carried out in Teflon-
FEP ampules that were closed by stainless steel valves. Volatile materials
were handled in a Pyrex glass vacuum line. Nonvolatile materials were
handled in the dry nitrogen atmosphere of a glovebox. Raman spectra
were recorded in the Teflon-FEP reactors or in heat-sealed Pyrex glass
melting point capillaries in the range of 4000−80 cm−1 on a Bruker
Equinox 55 FT-RA spectrophotometer using a Nd:YAG laser at 1064
nm with power levels less than 100 mW. Infrared spectra were recorded in
(N3)3,2,8,9 [Bi(N3)4]−,2,8 [Bi(N3)5]2−,2 and [Bi(N3)6]3−
2,8
although only [P(N3)6]−,3c As(N3)3,4 [As(N3)6]−,3f,5 Sb(N3)3,4
[Sb(N3)5]2−,7 [Sb(N3)6]−,6,7 [Bi(N3)4]−,2,8 [Bi(N3)5]2−,2 and
Received: October 9, 2012
Published: December 7, 2012
2,8
[Bi(N3)6]3−
have been structurally characterized. Very
© 2012 American Chemical Society
402
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
the range 4000−400 cm−1 on a Midac, M Series, FT-IR spectrometer
using KBr pellets. The pellets were prepared inside the glovebox using
an Econo mini-press (Barnes Engineering Co.). NMR spectra were
recorded at 298 K on Bruker AMX 500 (14N, ν0 = 36.13 MHz) and
Varian VNMRS-600s (15N, ν0 = 60.79 MHz) spectrometers using
(CD3)2CO and CD3CN solutions in 5 mm glass tubes that were sealed
with a J. Young valve. Neat CH3NO2 (δ = 0.00 ppm) was used as
external reference. Differential Thermal Analysis (DTA) curves were
recorded with a purge of dry nitrogen gas on an OZM Research
DTA552-Ex instrument using the Meavy 2.2.0 software. The heating
rate was 5 °C/min, and the sample size 20−30 mg. SbF3 (Ozark
Mahoning) was used without further purification. AsF3 (Advanced
Research Chemicals) and Me3SiN3 (Aldrich) were purified by fractional
condensation prior to use. Solvents were dried by standard methods and
freshly distilled prior to use. NMe4N3, PPh4N3, and (Ph3P)2NN3
((Ph3P)2N = bis(triphenylphosphoranylidene)ammonium) were pre-
pared by ion-exchange from NMe4Cl, PPh4Cl and (Ph3P)2NCl,
respectively.12
490 (0.3), 455 (3.4), 424 (0.7), 360 (0.7), 342 (1.1), 281 (0.1sh), 270
(2.0), 265 (0.1sh), 240 (2.9), 231 (0.1sh), 200 (1.7) cm−1; IR (KBr)
3077 (vw), 3054 (w), 3022 (vw), 2992 (vw), 2105 (vs), 2082 (vs), 2056
(s), 2024 (vs), 1825 (vw), 1772 (vw), 1587 (mw), 1573 (w), 1482
(mw), 1436 (vs), 1315 (s), 1291 (s), 1270 (vs), 1246 (s), 1183 (m),
1158 (w), 1116 (vs), 1073 (w), 1025 (w), 997 (m), 932 (vw), 851 (vw),
796 (w), 757 (m), 748 (m), 724 (vs), 693 (vs), 616 (w), 547 (s), 533
(vs), 527 (s), 502 (s), 451 (mw), 449 (mw), 440 (w), 436 (w), 429 (w),
424 (m), 419 (m), 415 (m) cm−1.
Preparation of A[Sb(N3)4] (A = NMe4, PPh4, (Ph3P)2N). Mixtures
of AN3 (1.00 mmol) and SbF3 (1.00 mmol) were loaded into a Teflon-
FEP ampule, followed by the addition of 4 mL of acetonitrile and
Me3SiN3 (4.0 mmol) in vacuo at −196 °C. The resulting mixtures were
then allowed to warm to ambient temperature over a period of about
30 min. After 10 h at room temperature, clear, colorless solutions were
obtained. The temperature was lowered to −20 °C, and all volatile
material was pumped off. After additional pumping for 24 h at ambient
temperature crystalline colorless solids were obtained. Weights expected
for 1.00 mmol of [NMe4][Sb(N3)4], [PPh4][Sb(N3)4], and
[(Ph3P)2N][Sb(N3)4]: 0.364 g, 0.629 g, and 0.828 g; weights found:
0.355 g, 0.634 g, and 0.835 g, respectively.)
[NMe4][Sb(N3)4]. DTA decomposition onset: 204 °C (explosion);
15N NMR (CD3CN, 298 K): δ = −137.8 ppm (τ1/2 = 11 Hz) (Nβ),
−251.6 ppm (τ1/2 = 11 Hz) (Nγ), −337.2 ppm (τ1/2 = 1 Hz) (NMe4);
14N NMR (CD3CN, 298 K): δ = −137.5 ppm (shoulder next to
CH3CN) (Nβ), −251 ppm (τ1/2 = 270 Hz) (Nγ), −337.1 ppm (τ1/2 = 4
Hz) (NMe4); Raman (75 mW, −40 °C): 3035 (3.2), 2983 (2.7), 2958
(1.6), 2942 (1.1), 2926 (2.7), 2897 (0.5), 2819 (0.9), 2254 (0.5), 2100
(3.6), 2067 (2.5), 2040 (2.0), 2001 (0.2), 1995 (0.2), 1912 (0.2), 1485
(0.5), 1451 (2.0), 1420 (0.5), 1336 (3.0), 1336 (3.0), 1282 (1.8), 1221
(0.5), 1179 (0.5), 1078 (0.5), 957 (0.9), 932 (0.5), 921 (0.5), 760 (1.4),
760 (1.4), 654 (1.4), 614 (0.5), 606 (0.5), 462 (0.9), 401 (10.0), 379
(2.0), 307 (2.3), 288 (1.6), 236 (1.8), 227 (1.6), 227 (1.6), 180 (3.2),
180 (3.2), 102 (6.4) cm−1, IR (KBr): 3017 (m), 2979 (w), 2035 (vs),
1645 (w), 1624 (w), 1488 (s), 1408 (m), 1321 (m), 1275 (w), 956 (s),
947 (s), 418 (m), 396 (w) cm−1.
Preparation of A[As(N3)4] (A = NMe4, PPh4, (Ph3P)2N). A sample
of AN3 (1.00 mmol) was loaded into a Teflon-FEP ampule, followed by
the addition of 4 mL acetonitrile in vacuo at −196 °C. The mixture was
warmed to ambient temperature to form a clear solution. After cooling
to −196 °C, AsF3 (1.00 mmol) and Me3SiN3 (4.0 mmol) were
condensed into the vessel. The mixture was then allowed to warm to
ambient temperature over a period of about 30 min. After 10 h at room
temperature, a clear, colorless solution was obtained. The temperature
was lowered to −20 °C, and all volatile material was pumped off. After
additional pumping for 24 h at ambient temperature a clear colorless
liquid (A = NMe4) or crystalline colorless solids (A = PPh4, (Ph3P)2N)
were obtained. Weights expected for 1.00 mmol of [NMe4][As(N3)4],
[PPh4][As(N3)4], and [(Ph3P)2N][As(N3)4]: 0.317 g, 0.582 g, and
0.782 g; weights found: 0.328 g, 0.575 g, and 0.789 g, respectively.
[NMe4][As(N3)4]. DTA decomposition onset: 75 °C (explosion); 14
N
NMR (CD3CN, 298 K): δ = −137.4 ppm (shoulder next to CH3CN)
(Nβ), −243 ppm (τ1/2 = 240 Hz) (Nγ), −337.2 ppm (τ1/2 = 4 Hz)
(NMe4); Raman (50 mW, −40 °C): 3035 (3.6), 2982 (5.5), 2956 (2.7),
2924 (3.6), 2896 (0.9), 2872 (0.1), 2819 (0.9), 2130 (5.5), 2093 (2.7),
2014 (0.9), 1451 (1.8), 1332 (2.7), 1270 (0.9), 951 (0.9), 754 (1.8), 672
(2.7), 573 (0.1), 550 (0.0+), 538 (0.0+), 465 (10.0), 420 (2.7), 293
(1.8), 247 (1.8) cm−1; IR (KBr): 3014 (m), 2085 (m), 2035 (vs), 2023
(vs), 2011 (s), 1487 (s), 1408 (w), 1400 (w), 1257 (vw), 956 (m), 948
(m), 845 (vw), 802 (mw), 418 (s) cm−1.
[PPh4][Sb(N3)4]. DTA decomposition onset: 202 °C; 14N NMR
(CD3CN, 298 K): δ = −137.6 ppm (shoulder next to CH3CN) (Nβ),
−253 ppm (τ1/2 = 220 Hz) (Nγ); Raman (60 mW, 20 °C): 3173 (0.5),
3147 (0.5), 3086 (2.5), 3063 (10.0), 3013 (1.0), 2960 (0.5), 2099 (3.0),
2072 (2.5), 2056 (1.0), 2037 (1.0), 2025 (1.0), 1589 (4.5), 1578 (1.5),
1485 (0.5), 1442 (1.0), 1329 (2.5), 1276 (1.0), 1192 (1.5), 1170 (1.0),
1102 (2.5), 1074 (0.5), 1029 (3.5), 1003 (9.5), 948 (0.5), 766 (0.5), 727
(1.0), 683 (2.0), 662 (1.0), 618 (2.0), 527 (0.5), 418 (9.5), 392 (2.5),
322 (2.5), 296 (1.5), 257 (3.0), 234 (2.0), 206 (3.0) cm−1, IR (KBr):
3078 (vw), 3058 (vw), 2144 (w), 2099 (s), 2071 (vs), 2050 (s), 2036
(vs), 2023 (vs), 1635 (w), 1586 (w), 1483 (mw), 1443 (m), 1439 (m),
1319 (m), 1267 (m), 1187 (w), 1164 (w), 1107 (ms), 1025 (vw), 995
(mw), 856 (vw), 764 (m), 723 (ms), 690 (m), 640 (vw), 616 (w), 603
(vw), 590 (vw), 569 (vw), 548 (vw), 525 (s), 430 (mw) cm−1.
[(Ph3P)2N][Sb(N3)4]. DTA decomposition onset: 210 °C; 14N NMR
(CD3CN, 298 K): δ = −137.4 ppm (shoulder next to CH3CN) (Nβ),
−255 ppm (τ1/2 = 250 Hz) (Nγ); Raman (75 mW, −40 °C): 3177 (0.4),
3147 (0.5), 3061 (10.0), 3013 (0.8), 2962 (0.4), 2126 (0.4), 2100 (1.3),
2081 (1.0), 2055 (0.8), 2039 (0.2), 1591 (4.6), 1577 (1.3), 1440 (0.1),
1337 (1.0), 1276 (0.1), 1266 (0.1), 1186 (0.7), 1166 (0.6), 1114 (2.0),
1030 (2.5), 1002 (7.9), 730 (0.2), 666 (1.5), 618 (1.0), 492 (0.1), 417
(2.7), 391 (1.1), 371 (0.4), 314 (0.5), 281 (0.4), 270 (1.1), 239 (1.5),
201 (0.6) cm−1; IR (KBr) 3077 (vw), 3058 (w), 3040 (vw), 3022 (vw),
2992 (vw), 2151 (vw, sh), 2142 (w), 2098 (s), 2079 (s), 2058 (s), 2037
(vs), 1588 (mw), 1573 (w), 1481 (m), 1437 (s), 1321 (s), 1304 (s),
1291 (s), 1270 (s), 1262 (s, sh), 1188 (w, sh), 1181 (mw), 1161 (w, sh),
1158 (w), 1115 (s), 1074 (vw), 1026 (w), 997 (mw), 925 (vw), 850 (w),
794 (w), 758 (m), 747 (m), 726 (s, sh), 724 (s), 693 (s), 665 (vw), 654
(vw), 616 (vw), 603 (vw), 588 (vw), 546 (s), 533 (s), 504 (s), 427 (w),
411 (w) cm−1.
[PPh4][As(N3)4]. DTA decomposition onset: 87 °C (explosion); 14
N
NMR ((CD3)2CO, 298 K): δ = −136.9 ppm (τ1/2 = 19 Hz) (Nβ), −246
ppm (τ1/2 = 180 Hz) (Nγ); 14N NMR (CD3CN, 298 K): δ = −137.6
ppm (shoulder next to CH3CN) (Nβ), −243 ppm (τ1/2 = 320 Hz) (Nγ);
Raman (40 mW, 20 °C): 3170 (0.4), 3145 (0.9), 3090 (2.2), 3066
(10.0), 3010 (0.4), 3010 (0.4), 2116 (2.6), 2089 (0.9), 2060 (0.1), 2049
(0.1), 2020 (0.9), 1587 (3.9), 1576 (0.9), 1484 (0.0), 1443 (0.4), 1437
(0.4), 1324 (0.9), 1276 (0.9), 1188 (0.9), 1167 (0.9), 1111 (0.9), 1101
(1.7), 1073 (0.0), 1027 (2.6), 1001 (7.4), 987 (0.4), 940 (0.4), 933
(0.0), 757 (0.0), 727 (0.4), 682 (2.2), 651 (0.4), 617 (0.9), 535 (0.0),
528 (0.0), 466 (3.9), 428 (0.9), 393 (0.4), 334 (0.9), 298 (0.9), 276
(1.7), 267 (1.3), 254 (2.6), 239 (1.3), 200 (2.6), 200 (2.6), 160 (5.7)
cm−1; IR (KBr): 3086 (w), 3061 (w), 2203 (vw), 2145 (w), 2112 (s),
2085 (vs), 2045 (vs), 2036 (vs), 2017 (vs), 1984 (s), 1973 (ms), 1584
(mw), 1574 (w), 1481 (m), 1440 (ms), 1437 (s), 1434 (ms), 1394 (w),
1359 (vw), 1341 (w), 1315 (mw), 1259 (m), 1250 (ms), 1185 (w), 1164
(w), 1160 (w), 1108 (s), 1023 (w), 994 (m), 853 (vw), 801 (w), 759
(ms), 754 (ms), 724 (s), 690 (s), 648 (w), 615 (w), 602 (vw), 571 (vw),
528 (vs), 464 (mw), 446 (w), 443 (w), 440 (w), 435 (mw), 424 (ms)
cm−1.
[(Ph3P)2N][As(N3)4]. DTA decomposition onset: 85 °C; 14N NMR
(CD3CN, 298 K): δ = −137.2 ppm (shoulder next to CH3CN) (Nβ),
−241 ppm (τ1/2 = 200 Hz) (Nγ); Raman (75 mW, −40 °C): 3176 (0.4),
3148 (0.6), 3058 (10.0), 3014 (0.7), 2994 (0.6), 2961 (0.4), 2108 (1.9),
2085 (0.7), 2058 (1.9), 2023 (0.1), 1591 (4.7), 1577 (1.4), 1486 (0.1),
1442 (0.3), 1326 (0.7), 1282 (0.9), 1260 (0.3), 1187 (0.7), 1166 (0.9),
1113 (2.1), 1076 (0.1), 1030 (2.6), 1002 (7.3), 940 (0.3), 859 (0.1), 800
(0.1), 757 (0.1), 729 (0.4), 683 (0.1), 666 (2.0), 618 (1.0), 550 (0.1),
Crystal Structure Determinations. The single crystal X-ray
diffraction data of [PPh4][As(N3)4] and [PPh4][Sb(N3)4] were
collected on a Bruker SMART APEX 3-circle platform diffractometer,
equipped with an APEX CCD detector with the χ-axis fixed at 54.74°,
403
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
and using Mo Kα radiation (Graphite monochromator) from a fine-
focus tube. The diffractometer was equipped with an LT-3 apparatus for
low-temperature data collection using controlled liquid nitrogen boil off.
A complete hemisphere of data was scanned on ω (0.3°) with a run time
of 10-s per frame at a detector resolution of 512 × 512 pixels using the
SMART software package.13 A total of 1271 frames were collected in
three sets and a final set of 50 frames, identical to the first 50 frames, was
also collected to determine any crystal decay. The frames were then
processed on a PC, running Windows 2000 software, by using the
SAINT software package13 to give the hkl files corrected for Lp/decay.
The absorption correction was performed using the SADABS
program.13 The structures were solved by the Patterson Method using
the SHELX-90 program and refined by the least-squares method on F2,
SHELXL-97 incorporated in SHELXTL Suite 6.12 for Windows
NT/2000.13 All non-hydrogen atoms were refined anisotropically.
The single crystal X-ray diffraction data of [(Ph3P)2N][As(N3)4],
[NMe4][Sb(N3)4], and [(Ph3P)2N][Sb(N3)4] were collected on a
Bruker SMART APEX DUO 3-circle platform diffractometer, equipped
with an APEX II CCD detector with the χ-axis fixed at 54.74°, and using
Mo Kα radiation (TRIUMPH curved-crystal monochromator) from a
fine-focus tube. The diffractometer was equipped with an Oxford
Cryostream 700 apparatus for low-temperature data collection. A
complete hemisphere of data was scanned on ω (0.5°) with a run time of
1-s per frame at a detector resolution of 512 × 512 pixels using the BIS
software package.14 A total of 1464 frames was collected in four sets. The
frames were then integrated using the SAINT algorithm14 to give the hkl
files corrected for Lp/decay. The absorption correction was performed
using the SADABS program.14 The structures were solved by the Direct
Method and refined on F2 using the Bruker SHELXTL Software
Package.14 All non-hydrogen atoms were refined anisotropically.
ORTEP drawings were prepared using the ORTEP-3 for Windows
V2.02 program.15 Further crystallographic details can be obtained from
the Cambridge Crystallographic Data Centre (CCDC, 12 Union Road,
Cambridge CB21EZ, U.K. (Fax: (+44) 1223-336-033; e-mail: deposit@
ccdc.cam.ac.uk) on quoting the deposition no. CCDC 844411−
844415.).
Computational Methods. All calculations were performed using
the recent hybrid meta exchange-correlation density functional M06-
2X16 in Gaussian 09, revision A.02.17 M06-2X is a reliable general
purpose density functional theory (DFT) functional for main-group
chemistry, with a mean average deviation of ∼2.2 kcal/mol, as
demonstrated by several benchmarks.16,18−20 We have recently verified
its usefulness by accurately predicting the structures of some bismuth
azides and iodine and bromine fluorides, for which other DFT methods,
such as B3LYP, are known to perform poorly.2
In this work nitrogen atoms were described by the cc-pVTZ basis
set,21 whereas the heavier arsenic and antimony centers were treated by
the cc-pwCVTZ-PP basis set.22 The latter is based on a small-core
relativistic Stuttgart pseudopotential that accurately recovers both scalar
and spin−orbit relativistic effects of inner core electrons. It treats the
valence and subvalence shells with a flexible triple-ζ Gaussian
contraction to capture core−valence correlation effects.
the empirical scaling required is unusually large, typically ∼0.911 for the
asymmetric azide stretching region (2000−2200 cm−1) and ∼0.915 for
the entire spectral region (0−2200 cm−1). Similarly large scaling factors
were necessary both for calculations on gas-phase systems, as well as
those described with implicit solvation through the SCRF-PCM
approach. These scaling factors differ significantly from the 0.98 value
initially recommended for the M06-2X functional.16 For comparison we
also calculated the spectra of several antimony and arsenic tetraazido
anions at the B3LYP/cc-pwCVDZ-PP level. The B3LYP functional
provided comparable results, albeit with an average scaling factor of
∼0.95 for the azide antisymmetric stretching region (see the Supporting
Information).
RESULTS AND DISCUSSION
■
Syntheses of the Arsenic and Antimony Polyazides.
Anionic polyazides, A+ [M(N3)x+y
]
y−‑can be prepared by the
y
reaction of ionic azides A+N3 with neutral polyazido
−
compounds, M(N3)x.2,6,11 The neutral polyazido compounds
can be obtained from the corresponding covalent fluorides, MFx,
and Me3SiN3. In this work, salts with the tetraazidoarsenites,
[As(N3)4]−, and -antimonites, [Sb(N3)4]−, were prepared in a
one-step, quantitative synthesis from AsF3 and SbF3, respectively,
by reaction with Me3SiN3 and a stoichiometric amount of the
corresponding ionic azide AN3 in acetonitrile solution (eq 1).
CH3CN
MF + AN3 + 3Me3SiN3 ⎯⎯⎯⎯⎯⎯→⎯ A[M(N3)4] + 3Me3SiF
3
M = As, Sb; A = NMe4, PPh4, (Ph3P)2N)
(1)
This resulted in rapid and complete fluoride-azide exchange
and easy product separation. All tetraazidopnictogen(+III)
compounds were isolated as colorless and room-temperature
stable materials. While tetramethylammonium tetraazidoarsen-
ite, NMe4[As(N3)4], was a clear, colorless, viscous, ionic liquid at
room temperature, all the other tetraazides were colorless
crystalline solids. Because of the relatively high ionicity of their
azido ligands and the presence of large counterions which
diminish shock propagation, these compounds are less sensitive
and explosive than the neutral parent compounds, M(N3)3, and
during the course of our study could be manipulated at room
temperature without explosions. The five crystalline
tetraazidopnictogen(+III) salts were also investigated by DTA.
The tetraazidoantimonites were found to be thermally more
stable than the corresponding arsenites. While [NMe4][As-
(N3)4], [PPh4][As(N3)4], and [(Ph3P)2N][As(N3)4] were
found to have a decomposition onset at 75 °C, 87 °C, and
85 °C, respectively, [NMe4][Sb(N3)4], [PPh4][Sb(N3)4], and
[(Ph3P)2N][Sb(N3)4] started to decompose at higher temper-
atures of 204 °C, 202 °C, and 210 °C, respectively. This indicates
that the nature of the counterion has little or no influence on the
thermal stability of the tetraazidopnictogen(+III) compounds.
Crystal Structures. Crystal structures were determined for
[PPh4][As(N3)4], [(Ph3P)2N][As(N3)4], [NMe4][Sb(N3)4],
[PPh4][Sb(N3)4], and [(Ph3P)2N][Sb(N3)4]. Selected crystallo-
graphic data for these structures are listed in Table 1.
Relative energies of anions in solutions were evaluated through the
self-consistent reaction field approach, as implemented in the Solvent
Model Density (SMD)23 polarizable continuum (PCM) model in
Gaussian 09. Some systems were also optimized with PCM, in which
cases the smaller cc-pwCVDZ-PP basis set for arsenic and antimony and
the corresponding cc-pVDZ basis set for nitrogen were used. Only
minor changes from the gas-phase structures were observed in these
cases.
The Hessian matrix was evaluated for all structures, to ensure that
only minima on the potential energy surface were considered, and to
provide vibrational frequencies. We have found only limited data on the
accuracy of M06-2X for predicting vibrational frequencies. Predicted
harmonic theoretical frequencies inherently differ from the observed
fundamental frequencies, which include anharmonicity and mode−
mode couplings. Therefore, empirical scaling of the predicted
frequencies is necessary before direct comparisons can be made with
the observed values. However, although we observe good agreement
between our calculated spectra and experiment in general, we note that
[PPh4][As(N3)4]. Our X-ray crystal structure of [PPh4][As-
(N3)4] is in good agreement with the previously published
structure10 and shows the presence of well isolated [As(N3)4]−
anions and [PPh4]+ cations. The closest As···N and N···N
contacts between neighboring anions are 5.60 Å and 3.31 Å,
respectively. Figure 1A depicts the structure of the [As(N3)4]−
anion in its [PPh4]+ salt. Selected bond lengths and angles of the
anion are listed in Table 2. This structure approximates most
404
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
Table 1. Crystallographic Data for [PPh4][As(N3)4], [(Ph3P)2N][As(N3)4], [NMe4][Sb(N3)4], [PPh4][Sb(N3)4], and
[(Ph3P)2N][Sb(N3)4]
compound
empirical formula
formula weight
temperature, K
crystal system
space group
a, Å
[PPh4][As(N3)4]
C24H20AsN12P
582.41
[(Ph3P)2N][As(N3)4]
C36H30AsN13P2
781.58
[NMe4][Sb(N3)4]
C4H12N13Sb
364.02
[PPh4][Sb(N3)4]
C24H20N12PSb
629.24
[(Ph3P)2N][Sb(N3)4]
C36H30N13P2Sb
828.42
163(2)
100(2)
100(2)
183(2)
100(2)
monoclinic
C2
triclinic
triclinic
monoclinic
P21/n
triclinic
P1
̅
P1
̅
P1
̅
16.723(5)
7.161(2)
11.702(3)
90
10.223(10)
10.642(10)
16.875(16)
95.375(12)
94.219(13)
97.37(2)
7.8125(1)
9.0472(1)
9.9280(2)
96.332(1)
109.999(1)
98.009(1)
644.15(2)
2
12.751(1)
7.3433(8)
27.749(3)
90
10.3792(4)
10.6043(4)
16.7458(7)
95.1460(1)
94.9940(1)
100.5720(1)
1794.4(1)
2
b, Å
c, Å
α, deg
β, deg
109.853(4)
90
91.411(2)
90
γ, deg
V, Å3
1318.0(6)
2
1806.0(6)
2
2597.4(5)
4
Z
ρcalc g/cm3
1.468
1.445
1.877
1.609
1.533
F(000)
592
800
356
1256
836
index ranges
−20 ≤ h ≤ 21,
−9 ≤ k ≤ 9,
−14 ≤ l ≤ 10
1.389
−13 ≤ h ≤ 13,
−14 ≤ k ≤ 14,
−22 ≤ l ≤ 22
1.083
−11 ≤ h ≤ 10,
−12 ≤ k ≤ 12,
−14 ≤ l ≤ 14
2.151
−15 ≤ h ≤ 12,
−8 ≤ k ≤ 9,
−34 ≤ l ≤ 34
1.163
−14 ≤ h ≤ 14,
−15 ≤ k ≤ 15,
−23 ≤ l ≤ 23
0.906
μ, mm−1
crystal size, mm
no. of rflns. collect.
no. of indep. rflns.
Rint
0.15 × 0.09 × 0.05
4625
0.34 × 0.28 × 0.25
32448
0.51 × 0.31 × 0.31
16956
0.22 × 0.06 × 0.05
15017
0.20 × 0.19 × 0.09
29102
2846
8950
3898
5205
10971
0.0310
0.0314
0.0194
0.0713
0.0440
no. of params.
R1, wR2 [I > 2σ(I)]
R1, wR2 (all data)
(Δ/ρ)min/max, e Å−3
173
469
168
343
469
0.0464, 0.0816
0.0667, 0.0890
0.434/−0.229
0.0273, 0.0645
0.0353, 0.0676
0.449/−0.368
0.0123, 0.0327
0.0126, 0.0329
0.763/−0.618
0.0487, 0.0785
0.0916, 0.0928
0.856/−0.486
0.0349, 0.0681
0.0544, 0.0746
0.504/−0.735
Figure 1. ORTEP plot for the anion in [PPh4][As(N3)4] (A) and [(Ph3P)2N][As(N3)4] (B). The predicted, but unobserved, sterically active lone
valence electron pair has been added to the ORTEP plot for a better understanding of the observed geometry. Thermal ellipsoids are drawn at the 30%
probability level.
closely that of the total energy minimum of the free undisturbed
[As(N3)4]− anion (As-1, see below).
directly in the crystal structure, its approximate location can be
deduced from the observed geometry of the MN4 skeleton, that
is, from the orientation of the plane defined by the Nα atoms of
the axial ligands and the central atom. When the lone pair domain
is located on the C2 axis, the plane defined by the two axial Nα
atoms and the central atom, will bisect the equatorial MN2 angle.
This is the case for the [PPh4]+ salt of [As(N3)4]−, predicting the
presence of a lone pair domain on the C2 axis.
If the formal lone valence electron pair at the arsenic atom is
included, the structure can be derived from a trigonal bipyramid
with the lone pair domain in an equatorial position which
because of its greater space requirement compresses the axial
N4−As−N4′ angle. In an ideal pseudo trigonal bipyramidal
arrangement, the plane defined by the axial ligands and the lone
pair bisects the angle between the equatorial ligands. Although
the lone pair domain on the central atom cannot be observed
[(Ph3P)2N][As(N3)4]. The crystal structure of the [(Ph3P)2N]+-
salt of the [As(N3)4]− anion also consists of isolated [As(N3)4]−
405
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
Table 2. Selected Skeletal Bond Lengths [Å] and Angles [deg] for the Anions in A[M(N3)4]
A = [PPh4]; M = As
A = [(Ph3P)2N]; M = As
A = [NMe4]; M = Sb
A = [PPh4]; M = Sb
A = [(Ph3P)2N]; M = Sb
M−N1
M−N4
M−N7
M−N10
1.888(4)
2.129(4)
1.888(4)
2.129(4)
90.82(16)
97.2(2)
1.911(1)
2.201(1)
1.905(1)
2.047(1)
83.42(5)
95.49(5)
87.19(5)
88.62(5)
167.62(5)
84.25(5)
−7.4
2.1149(9)
2.2616(9)
2.1109(9)
2.2635(9)
82.48(4)
96.26(4)
79.61(4)
81.67(3)
156.11(4)
84.70(4)
6.1
2.078(4)
2.254(4)
2.071(4)
2.244(4)
84.12(16)
95.10(19)
77.96(17)
83.90(16)
154.96(16)
80.46(16)
−4.5
2.096(2)
2.240(2)
2.082(2)
2.280(2)
87.10(9)
93.22(8)
80.84(9)
82.14(7)
160.29(8)
83.03(8)
5.2
N1−M−N4
N1−M−N7
N1−M−N10
N4−M−N7
N4−M−N10
N7−M−N10
83.03(16)
83.03(16)
170.7(3)
90.82(16)
0.3
a
α
a
Angle between the plane defined by N4, As and N10 and the plane bisecting N1−As−N7. Positive values indicate a rotation toward N7, resulting in
a rotation of the lone pair toward N1.
Figure 2. Chain structure of the [Sb(N3)4]− anion in [NMe4][Sb(N3)4]. Thermal ellipsoids are shown at the 50% probability level. Selected bond
lengths [Å] and angles [deg]: Sb−N4 2.262(1), Sb−N6″ 3.176(1), Sb−N10 2.263(1), Sb−N10′ 2.957(1) Sb″−N6 3.172(1), N4−Sb−N6″ 70.1(1),
N10−Sb−N10′ 79.4(1), Sb−N10−Sb′ 114.5(1), Sb−N4−N5 120.4(1), Sb−N6″−N5″ 131.4(1).
and [(Ph3P)2N]+ ions. Even though [(Ph3P)2N]+ is larger than
[PPh4]+, the [As(N3)4]− anions in [(Ph3P)2N][As(N3)4] are
packed closer to each other and the closest As···N and N···N
contacts between neighboring anions are 3.78 Å and 3.34 Å,
respectively. The compound crystallizes in the triclinic space
domain, and the two equatorial N3 ligands are oriented toward
this ligand. The N1−As−N7 angle between the equatorial azido
ligands is not bisected by the axial plane defined by N4, As, and
N10. Instead, the plane is rotated to the side of N1 by about 7°.
This leads to the conclusion that in [(Ph3P)2N][As(N3)4] the
center of the lone pair domain of arsenic is slightly tilted toward
N7. The calculated energy difference between the two observed
[As(N3)4]− conformations is only 0.4 kcal/mol, and the
observation of two different structures can be attributed to
minor differences in their respective lattice energies.
group P1. The structure of the [As(N ) ]− anion in its
̅
3
4
[(Ph3P)2N]+-salt is depicted in Figure 1B, and selected bond
lengths and angles are listed in Table 2. The structures of the
anions in [PPh4][As(N3)4] and [(Ph3P)2N][As(N3)4] are both
derived from a trigonal bipyramid.
However, the azido ligands are oriented differently in the two
salts. In the [PPh4]+-salt (Figure 1A), both axial azido ligands are
oriented toward the lone pair domain. One equatorial N3 group is
pointing above the equatorial plane and the second one is
pointing below the plane. In the [(Ph3P)2N]+-salt (Figure 1B),
only one of the axial azido ligands is oriented toward the lone pair
[NMe4][Sb(N3)4]. This compound crystallizes in the triclinic
space group P1. Because of the smaller size of the [NMe ]+
̅
4
counterion, the [Sb(N3)4]− anions are in close proximity to each
other and form two Sb···N bridges with two neighboring anions
which are shorter than the sum of the van der Waals (vdW) radii
resulting in a chain structure (Figure 2). The chains are made up
406
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
Figure 3. ORTEP plot of the isolated [Sb(N3)4]− units in [NMe4][Sb(N3)4] (A) and [PPh4][Sb(N3)4] (B) including the bridging nitrogen atoms from
two neighboring [Sb(N3)4]− units. The red labels t and μx,y indicate the ligand types, terminal and bridging, respectively. Bridging bonds to neighboring
units are drawn in red. Thermal ellipsoids are drawn at the 30% probability level. Again, the predicted lone pair domain has been added to the ORTEP
plots for illustrative purposes.
from alternating planar four-membered Sb2N2 parallelograms,
with two μ1,1-bridging N3 ligands and chair-type eight-membered
Sb2(N3)2 rings with two μ1,3-bridging N3 ligands.
the antimony atom is increased to seven, and the observed ligand
arrangement is an unusual motif for CN 7. In principle it is
a strongly distorted trigonal bipyramid with two additional
N-bridges coming in from above and below the equatorial plane
and from the left of the plane formed by N10, Sb, and N4. The
two bridges enhance the repulsion effect of the lone pair thus
further compressing the axial and equatorial N−Sb−N angles.
Selected bond length and angles for the [Sb(N3)4]− unit in
[NMe4][Sb(N3)4] are listed in Table 2.
It should be noted that according to the official nomenclature
conventions, the coordination number is determined exclusively
by the number of ligands, and, therefore, sterically active lone pairs
are not included in the coordination number. In this paper, the
stericallyactive lone pairs have beenincluded inthe CNtofacilitate
the rationalization of the observed geometriesand the term“formal
coordination number” has been used to describe this situation.
[PPh4][Sb(N3)4]. Our X-ray crystal structure of [PPh4][Sb-
(N3)4] is in good agreement with the previously published
structure.10 As in the [N(CH3)4]+ salt, the [Sb(N3)4]− units are
in close proximity to each other and form two Sb···N bridges,
with two neighboring anions which are shorter than the sum of
the vdW radii. While in the [NMe4]+-salt the two azido bridges,
one μ1,1 and one μ1,3, involve the two axial ligands and result in a
chain structure, in the [PPh4]+-salt both bridges are μ1,3 bridges
and involve one axial and one equatorial N3 group. The
[Sb(N3)4]− units are interconnected through the azido bridges
resulting in a helical chain structure (Figure 4) for the anion in
[PPh4][Sb(N3)4] which are aligned along the [010] direction of
the crystal.
The parallelograms result from two shorter (2.263(3) Å) and
two longer (2.957(1) Å) Sb−N bonds and resemble those found
in the Sb(N3)3 and [PPh4][Bi(N3)4] “Mitsubishi-type” struc-
tures.2,4,8 The Sb−N bonds in the eight-membered rings have a
length of 2.262(1) Å and 3.172(1) Å. A similar chain structure
has previously been observed for the anion in [PPh4][Bi(N3)4]2,8
and the (Py)2Bi(N3)3 adduct.8 However, the structures of these
bismuth compounds are quite different as they do not contain the
eight-membered rings. In each isolated individual [Sb(N3)4]−
unit of the chain, the formal lone valence electron pair at the
antimony atom is sterically active. The location of the lone pair
domain in [NMe4][Sb(N3)4] can be deduced from the following
analysis. The close environment around each antimony atom
consists of two terminal N3 ligands (N1 and N7), one μ1,1-
bridging N3 ligand (N10), one μ1,3-bridging N3 ligand (N4), and
the sterically active Sb lone valence electron pair. This
arrangement is derived from a pseudotrigonal bipyramid with
the two terminal N3 ligands and the formal lone valence electron
pair in the three equatorial positions, and the two bridging N3
groups occupying the two axial positions (Figure 3A). Both
axial N3 groups are bent away from the formal lone pair but are
not coplanar; the dihedral angle between the two planes formed
by N11, N10, N4 and N5, N6, N10 is 80.2(5)°. Because of
the increased repulsion from the formal lone pair and the two
long Sb−N bridges coming in from the same direction, the axial
N4−Sb−N10 bond angle is compressed from the ideal 180°
to 156.1(4)°. The equatorial N1−Sb−N7 angle is not bisected
by the axial plane defined by N4, Sb, and N10. Instead, the
plane is rotated to the side of N7 by about 6°. The ligand
arrangement in [Sb(N3)4]− resembles somewhat that of the
monomeric EBiN4 skeleton in [PPh4][Bi(N3)4] with one
equatorial azido group pointing up and the other one pointing
down and the two axial N3 groups being tilted away from the
formal lone pair.2,8
As in the [NMe4]+-salt, the basic geometry of the [Sb(N3)4]−
unit in the [PPh4]+-salt is derived from a pseudotrigonal
bipyramid with one terminal N3 ligand (N7), one μ1,3 bridging
N3 ligand (N1) and the sterically active formal lone pair being
located at the three equatorial positions, and a terminal (N4) and
a μ1,3 bridging (N10) N3 group occupying the two axial positions
(Figure 3B). The axial N−Sb−N bond angle is compressed from
the ideal 180° to 155.0(2)° by the more space demanding lone
pair and two Sb···N bridges. While the basic geometry of the
SbN4 skeleton is the same for [NMe4][Sb(N3)4] and [PPh4]-
[Sb(N3)4], the arrangement of the ligands is different. In the
When the two long nitrogen bridges from the two neighboring
−
[Sb(N3)]4 units and the formal lone pair at the Sb atom are
included in the coordination sphere, the coordination number of
407
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
Figure 4. Part of the chain structure of the [Sb(N3)4]− anion in [PPh4][Sb(N3)4] in the [010] direction of the crystal. Thermal ellipsoids are shown at
the 30% probability level.
Figure 5. Dimeric structure of the [Sb(N3)4]− anion in the crystal of [(Ph3P)2N][Sb(N3)4]. Thermal ellipsoids are shown at the 50% probability level.
Selected bond lengths [Å] and angles [deg]: Sb−N10 2.280(2), Sb−N12′ 3.024(2), N10−Sb−N12′ 81.6(1), Sb−N10−N11 112.9(1), Sb−N12′−
N11′ 145.0(1).
[NMe4]+-salt, one equatorial N3 group is pointing up and the
other one is pointing down while both axial N3 groups are tilted
away from the lone pair domain, are not coplanar, and are
bridging (Figure 3A). In the [PPh4]+-salt, both equatorial N3
groups are pointing below the equatorial plane, and both axial N3
groups are tilted toward the lone pair. The two bridging azido
groups involve one equatorial (N1) and one axial (N10) ligand
(Figure 3B). In addition, both axial N3 groups are virtually
coplanar with an angle of less than 1° between the planes defined
by N11, N10, N4 and N5, N4, N10. Again, the equatorial N1−
Sb−N7 angle is not bisected by the axial plane defined by N4, Sb,
and N10. Instead, the plane is rotated to the side of N1 by about
5°. This results in a slight rotation of the antimony lone pair
domain toward N7. This different arrangement results in the
formation of a helical chain anion structure.
[(Ph3P)2N][Sb(N3)4]. The [(Ph3P)2N]+-salt crystallizes in the
triclinic space group P1 with two formula units in the unit cell
̅
(Z = 2). Again, the [Sb(N3)4]− anions are not isolated units
but form isolated Sb···N bridged dimers. In each dimer, two
[Sb(N3)4]− units, that are related through a center of symmetry,
are linked through two μ1,3-azido bridges (Figure 5). These N3-
bridges together with the two Sb atoms form an eight-membered
[Sb2(N3)2] ring in a flattened chair conformation. While each Sb
atom in the [NMe4]+ and [PPh4]+-salts forms two Sb···N bridges
resulting in a polymeric anion structure with a “formal coordina-
tion number” of seven, in the [(Ph3P)2N]+-salt each Sb atom
forms only one Sb···N bridge, resulting in a dimeric structure
with a “formal coordination number” of six. The shortest Sb···N
distances between different dimers are 6.965 Å. The formation
of dimeric anion units is probably due to the higher steric
demand of the [(Ph3P)2N]+-counterion compared to [PPh4]+
Figure 6. ORTEP plot of the anion in [(Ph3P)2N][Sb(N3)4] including
the bridging nitrogen atom N12′ from the neighboring [Sb(N3)4]− unit.
Thermal ellipsoids are drawn at the 30% probability level. The red labels
t and μ1,3 indicate terminal and μ1,3 bridging ligands, respectively. The
bridging bond Sb−N12′ is 3.024(2) Å long. The unobserved formal
lone pair has been added for illustrative purposes.
and renders the formation of a polymeric anion chain less
favorable. Again, the geometry of the SbN4 skeleton in the
[(Ph3P)2N]+-salt is derived from a pseudotrigonal bipyramid
with two terminal N3 ligands and the sterically active formal lone
valence electron pair occupying the three equatorial positions
408
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
Figure 7. Calculated gas-phase geometries and symmetries of the major conformers of (a) [As(N3)4]− and (b) [Sb(N3)4]−. Bond lengths are given in Å.
(Figure 6). One terminal and the bridging N3 group occupy the
two axial positions and are virtually coplanar. The angle between
the two planes formed by N5, N4, Sb and N11, N10, Sb is less
than 1°. The axial N4−Sb−N10 bond angle is compressed from
the ideal 180° to 160.3(1)°, and the equatorial N1−Sb−N7 angle
is compressed to 93.2(1)° from the ideal 120°. The bridging axial
N3 ligand points toward the formal lone pair while the terminal
one points away from it. Of the two equatorial ligands, one points
up and one points down. Thus, the ligand arrangement in the
[(Ph3P)2N]+-salt is again different from those observed in the
other salts. The axial plane defined by N4, Sb, and N10 is rotated
to the side of N7 by about 5° and does not bisect the equatorial
N1−Sb−N7 angle. This results in a slight rotation of the
antimony lone pair domain toward N1.
of M) skeleton, the arrangement of the azido ligands in each
anion is completely different. To our best knowledge, such a
great variability of conformers is unprecedented in inorganic
main group chemistry. In the following theoretical section we
have analyzed each conformer in detail to rationalize the
multitude of observed structures, and to facilitate the analysis of
the solid-state and solution vibrational spectra. The observation
of extraordinary conformational variability for these anions can
be attributed to the high fluxionality of the azido ligands as
demonstrated by the presence of many low-frequency torsional
modes (see Supporting Information). The big surprise was our
ability to isolate and characterize, in spite of their very small
energy differences, each conformer in pure form in the solid state.
Computational Results. To quantify the fluxionality of the
tetraazido-antimony and -arsenic anions, we calculated the gas-
phase energies and structures of all the seven possible major
conformations (Figure 7) at the M06-2X/cc-pwCVTZ-PP level
of theory, that is, all combinations of axial ligands pointing to the
“front and back”, and equatorial ligands pointing “up and down”.
Conformers 1−5 were optimized from their closest experimental
X-ray crystal structure analogues, whereas the starting geometries
for the experimentally unobserved conformers 6 and 7 were
obtained through rotation of one axial azido ligand of conformers
4 and 2, respectively. This was followed by solution energy
calculations at the SMD-PCM-M06-2X/cc-pwCVTZ-PP level.
When the additional long nitrogen bridge to N12′ of the
second [Sb(N3)4]− unit of the dimer is included in the coordina-
tion sphere, the “formal coordination number” is six. This bridge
is pointing above the equatorial plane formed by N1−Sb−N7 of
the [Sb(N3)4]− unit (Figure 6). Selected bond lengths and angles
for the [Sb(N3)4]− unit in [(Ph3P)2N][Sb(N3)4] are summarized
in Table 2.
Conformational Variability. This series of [M(N3)4]−
(M = As and Sb) structures is highly unusual because five
different conformers were found for five different salts. Although
all anions possess the same MN4E (E = formal lone pair domain
409
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
a
Table 3. Relative Energies (in kcal/mol) of All Considered Conformers at the M06-2X/cc-pwCVTZ-PP Level of Theory
−
−
As(N3)4
Sb(N3)4
conformer:
ΔH0
ΔG0
ΔH0
ΔG0
b
1
0.0 (0.0)
0.0 (0.0)
−0.4 (−0.5)
0.4 (0.0)
0.4 (0.2)
2.0 (1.0)
3.9 (1.9)
3.1 (1.1)
0.0 (0.0)
0.0 (0.0)
0.7 (0.4)
c
c
2
3
4
−1.1 (−1.2)
−0.2 (−0.6)
0.1 (−0.1)
0.1 (−0.8)
2.4 (0.4)
−1.3 (−2.2)
−0.7 (−1.2)
0.1 (−0.4)
−1.0 (−2.2)
1.4 (−0.6)
0.4 (−1.4)
0.1 (−0.5)
0.7 (0.2)
b
5
0.7 (−0.4)
2.6 (0.7)
6
7
1.6 (−0.4)
1.7 (−0.2)
43.9 (0.7)
43.7 (2.4)
[3−2]
[3−3]
31.0 (−10.2)
31.6 (−7.8)
a
b
c
Values for acetonitrile (1M, 298 K) are given in parentheses. Two enantiomers. Transition state corresponding to rotation of one axial azido
ligand.
Because of the great similarity of the different conformers,
inherent errors in the DFT method should largely cancel out.
Therefore, we can expect a high precision in relative energies,
considerably better than the method’s already good general
accuracy of ∼2.2 kcal/mol3.
The conformers of [As(N3)4]− are shown in Figure 7a and
their relative energies are listed in Table 3. Not surprisingly, the
conformers are very close in energy. Indeed, As-1−4 are so close
(ΔG0 < ∼0.5 kcal/mol), that we can expect rapid interconversion
in solution. Conformers As-5−7 lie 1.5−2.5 kcal/mol higher in
energy relative to the lowest conformer As-2, and should have a
negligible contribution to the overall population.
In contrast to the equilibrium cases in solution, the single
crystals of each compound contain only one conformer showing
small thermal ellipsoids and no disorder. We note that the two
conformers predicted to have marginally lower energy (As-1 and
As-2) in the gas-phase are the ones also observed experimentally
in the [PPh4]+ and [(Ph3P)2N]+-salts, respectively.
The analogous seven monomeric conformers (Sb-1−7) of
[Sb(N3)4]− are shown in Figure 7b and are closely related to their
arsenic analogues, except Sb-2, which is a transition state of Cs
symmetry. Despite several attempts, no local minimum
corresponding to Sb-2 was found. This is likely due to a very
shallow potential energy surface, and does not affect our
conclusions.
The antimony conformers proved closer in energy compared
to their arsenic analogues, with Sb-2 and Sb-6 being the least
favorable by a slight margin (Table 3). However, because all the
antimony salts proved dimeric or polymeric experimentally, we
also investigated two conformers of the dimer [(Sb(N3)4)2]2−,
[Sb-3−Sb-2] and [Sb-3]2, which resulted after optimization
from our X-ray structures (Figure 8). Presumably, the tendency
of the azidoantimonites to associate arises from the larger
antimony atom seeking higher coordination numbers. Since
dimers and extended chains require sharing of bridging azide
ligands for their formation, conformers 2 and 3 appear ideally
suited for this purpose.
For the considered systems, the entropic cost of dimer forma-
tion, because of lowered vibrational, rotational, and translational
degrees of freedom, is approximately 11−13 kcal/mol at ambient
temperature. In addition, there exists a significant Coulombic
penalty when combining two negatively charged species.
Together, these effects prevent the formation of extended systems
in the gas-phase. However, solvated and solid-state systems
are markedly different, and surrounding solvent molecules and
crystal lattice fields can greatly affect and govern which structures
are formed. Whereas the Gibbs free energy penalty for forming
Figure 8. Calculated gas-phase geometries and symmetries of two
[(Sb(N3)4)2]2− conformers. Bond lengths are given in Å.
dimeric [Sb(N3)4]− units in the gas-phase is a staggering
44 kcal/mol, the existence of such species in solution or solid state
is more likely. Optimization of [Sb-3−Sb-2] in the implicit reaction
field of acetonitrile (SMD-PCM) revealed the relative enthalpy of
binding to be −10 kcal/mol. This value completely counteracts
the entropic penalty, and the free energy of binding is close to zero
(at 1 M and 298 K). In other words, we can expect a mixture of
dimers and monomers in polar solvents. If dimers are present in
solution, it is possible that they could aggregate further into larger
systems, which subsequently could nucleate into crystals. Thus, such
a dynamic case might explain why [Sb(N3)4]− salts crystallize as
various extended systems, whereas [As(N3)4]− salts do not.
Because barriers to azide rotation are small, the interconver-
sion of conformers in solution will be extremely fast, and on the
time scale of molecular vibrations. Therefore, the dynamic
equilibrium is too fast for NMR measurements and might only be
accessible by means of vibrational spectroscopy.
Vibrational Spectra. Raman spectra of the solids and
CH3CN and SO2 solutions and infrared spectra of the solids were
recorded for all compounds (Figures 9−13, Experimental Section,
and Supporting Information). The calculated vibrational spectra
for all free, gaseous conformers are giveninFigures 14−16, and the
Supporting Information.
An analysis of the observed Raman spectra allows the
following conclusions. (1) Since the pseudotrigonal bipyramidal
EMN4 (E = formal lone pair domain of M) skeletons are very
similar for all the conformers, the skeletal modes are not very
useful for distinguishing the different conformers. However, the
frequencies, relative intensities, and number of observed bands of
410
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
Figure 9. Azide antisymmetric stretching regions in the Raman spectra
of liquid [NMe4][As(N3)4] (broad bands due to rapidly exchanging
equilibrium of several conformers) and solid [PPh4][As(N3)4] and
[(Ph3P)2N][As(N3)4] (narrower bands due to well isolated ions with
different ligand arrangements).
Figure 11. Azide antisymmetric stretching region in the Raman spectra
of liquid [NMe4][As(N3)4] (black) and its CH3CN solution at −30 °C
(green).
[M(N3)4]−-salts exhibit very distinct, well resolved, characteristic
patterns. Not surprisingly, the ionic-liquid, [NMe4][As(N3)4],
exhibits broader bands and a spectrum almost identical to that
observed for its CH3CN solution (Figure 11), suggesting the
simultaneous presence of multiple conformers. The presence of
multiple conformers also explains our failure to obtain a crystal
structure for this compound; (2) the Raman bands for the
CH3CN and SO2 solutions are different and much broader
than those of the solids and very similar for [PPh4][As(N3)4]
(Figure 12) and [(Ph3P)2N][As(N3)4] (Supporting Informa-
tion, Figure S22), demonstrating that the anion structures are
no longer governed by the cation and that the solution spectra
are due to rapidly exchanging multiple conformers. It is also
noteworthy that the spectra are solvent dependent, that is, differ
significantly for the CH3CN and SO2 solutions, implying different
solvent interactions and/or conformer equilibria. Because of the
relatively low solubility of these salts, the signal-to-noise ratio of
the solution spectra is somewhat poor but sufficient to demon-
strate the above points; (3) the [PPh4]+ (Figure 13), [NMe4]+
(Supporting Information, Figure S23) and [(Ph3P)2N]+-salts
(Supporting Information, Figure S24) of [Sb(N3)4]− behave
similarly to the arsenic analogues when dissolved in CH3CN
or SO2 but, as expected, their spectra are somewhat different
from those of the As salts. Contrary to [NMe4][As(N3)4], the
[NMe4][Sb(N3)4] salt is solid at room temperature and contains
a well-ordered, single conformer, polymeric anion which, in spite
Figure 10. Azide antisymmetric stretching regions in the Raman spectra
of solid [NMe4][Sb(N3)4] (infinite chains with alternating Sb2N2
parallelograms and eight-membered Sb2(N3)2 rings), [PPh4][Sb(N3)4]
(helical chains with μ1,3 bridges), and [(Ph3P)2N][Sb(N3)4] (dimeric
anion with an eight-membered Sb2(N3)2 ring).
the antisymmetric stretching vibrations of the azido ligands are
strongly influenced by the ligand arrangements and occur in a
region which is not obscured by cation or solvent bands.
Therefore, they are ideally suited for characterizing the different
conformers. As can be seen from Figures 9 and 10, the five solid
411
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
Figure 12. Azide antisymmetric stretching region in the Raman spectra
of [PPh4][As(N3)4]: neat solid (black), CH3CN solution at −30 °C
(green), and SO2 solution at −30 °C (red).
of the presence of two different kinds of azido bridges, gives rise to
a surprisingly simple Raman spectrum (Figure 10 and Supporting
Information, Figure S23); and (4) freezing of the CH3CN
solutions at −80 °C did not dramatically change the spectra from
those recorded for the solutions at −30 °C.
As one might expect from the fact that only the solid
[As(N3)4]− salts contain single, well isolated, monomeric
conformers, the agreement between the observed solid-state
and calculated theoretical data is best for the arsenic compounds.
Figure 14 shows the calculated gas-phase Raman spectrum of
As-1 together with the solid-state spectrum of [PPh4][As(N3)4]
in the azide antisymmetric stretching region, 2000−2140 cm−1.
Four characteristic bands, corresponding to the axial and
equatorial in- and out-of-phase stretching modes, show that
the ions are not markedly perturbed by the presence of a solid
lattice.
Figure 13. Azide antisymmetric stretching region in the Raman spectra
of [PPh4][Sb(N3)4]: neat solid (black), SO2 solution at −30 °C (red),
CH3CN solution at −30 °C (green), frozen CH3CN solution at −80 °C
(blue).
In solution multiple structures are in rapid equilibrium, and
identification of individual conformers is challenging. With the
exception of the axial in-phase vibrations most azide stretching
modes differ only slightly between each conformer. Fortunately,
our calculations show that the corresponding Raman intensities
can vary significantly, thus enabling a more quantitative
fingerprinting of experimental spectra (Figure 15). For example,
a relatively small increase in the contribution from As-2 is in
principle sufficient to explain the major difference in the intensity
patterns of [PPh4][As(N3)4] in acetonitrile and SO2 solutions
(Figure 12). In an effort to model the Raman spectrum of
[PPh4][As(N3)4] in acetonitrile solution, which arises from a
complex equilibrium, nine theoretical spectra (seven conformers
and two enantiomers of As-1 and As-5) were weighted in a
Figure 14. Experimental solid-state [PPh4][As(N3)4] (dotted line) and
theoretical [As(N3)4]− (solid line) Raman spectra in the azide
antisymmetric stretching region.
412
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
ASSOCIATED CONTENT
* Supporting Information
Figures S1−S25 and Tables S1−S29. This material is available
free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: haiges@usc.edu (R.H.), kchriste@usc.edu (K.O.C.).
■
Notes
The authors declare no competing financial interest.
Figure 15. Theoretical Raman intensity patterns of the azide
ACKNOWLEDGMENTS
antisymmetric stretching modes of the seven [As(N3)4]− conformers.
■
This work was supported by the National Science Foundation,
the Office of Naval Research, the Air Force Office of Scientific
Research, the Defense Threat Reduction Agency, and the
Department of Energy. We appreciate the NSF CRIF Grant
1048807 to support an X-ray diffractometer. Computational
work at USC was supported by the University of Southern
California Center for High-Performance Computing and
Communications. M.R. is grateful for financial support from
the Swedish Research Council (VR).
DEDICATION
■
Dedicated to Professor George Olah on the occasion of his 85th
birthday.
REFERENCES
■
(1) Klapoetke, T. M.; Tornieporth-Oetting, I. C. Nichtmetallchemie;
VCH: Weinheim, Germany, 1994; p 283.
Figure 16. Experimental spectrum of a solution of [PPh4][As(N3)4] in
CH3CN (−30 °C, dotted line) and theoretical Raman spectra of the
different [As(N3)4]− conformers (solid line) in the azide antisymmetric
stretching region. The inset shows the predicted populations of the
individual conformers at −30 °C in acetonitrile solution.
(2) Haiges, R.; Rahm, M.; Dixon, D. A.; Garner, E. B.; Christe, K. O.
Inorg. Chem. 2012, 51, 1127 , and references cited therein.
(3) (a) Buder, W.; Schmidt, A. Z. Anorg. Allg. Chem. 1975, 415, 263.
(b) Volgnandt, P.; Schmidt, A. Z. Anorg. Allg. Chem. 1976, 425, 189.
(c) Portius, P.; Fowler, P. W.; Adams, H.; Todorova, T. Z. Inorg. Chem.
Boltzmann distribution at −30 °C (Figure 16). The procedure
successfully explains the observed line broadening and the overall
intensity pattern.
Spectra of the antimony compounds in solution are more
complex because of a higher number of populated monomeric
conformers, polyanions, as well as the presence of bridging
ligands. Consequently, they were not analyzed in further detail.
2008, 47, 12004. (d) Klapotke, T. M.; Geissler, P. J. Chem. Soc., Dalton
̈
Trans. 1995, 3365. (e) Geissler, P.; Klapotke, T. M. Spectrochim. Acta,
̈
Part A 1995, 51, 1075. (f) Klapotke, T. M.; Noth, H.; Schuett, T.;
Warchhold, M. Angew. Chem., Int. Ed. 2000, 39, 2108. (g) Klapotke, T.
̈
̈
̈
M.; Schulz, A.; McNamara, A. J. Chem. Soc., Dalton Trans. 1996, 2985.
(4) Haiges, R.; Vij, A.; Boatz, J. A.; Schneider, S.; Schroer, T.; Gerken,
M.; Christe, K. O. Chem.Eur. J. 2004, 10, 508.
(5) Karaghiosoff, K.; Klapotke, T. M.; Krumm, B.; Noth, H.; Schutt, T.;
̈
̈
̈
CONCLUSIONS
Suter, M. Inorg. Chem. 2002, 41, 170.
■
(6) Haiges, R.; Boatz, J. A.; Vij, A.; Vij, V.; Gerken, M.; Schneider, S.;
Schroer, T.; Yousufuddin, M.; Christe, K. O. Angew. Chem., Int. Ed. 2004,
43, 6676.
A systematic experimental and theoretical study of the structures
of the [NMe4]+, [PPh4]+, and [(Ph3P)2N]+ salts of the
tetraazidoarsenite and tetraazidoantimonite anions demonstrates
that in highly fluxional species possessing numerous minimum
energy structures of very similar energies and low barriers for
their interconversion, subtle changes in solid-state effects can
result in very different, well-defined, ordered anion structures.
Thus, for five similar salts five different structures were experi-
mentally observed in the solid state providing a text book example
for the existence of conformational variability in inorganic main
group chemistry. It is shown that very small changes in packing
energies and cation−anion interactions can give rise to different
anion structures, reminding us of chameleons which can adapt
their colors to small changes in their environment. In solution,
multiple conformers can exist, undergoing facile interconversion
resulting in averaged, solvent dependent fluxional structures.
Therefore, for highly fluxional compounds, such as polyazides,
great care must be exercised with conclusions concerning structural
principles and a preferred structure based on the observed
experimental geometry of a single salt.
(7) Lyhs, B.; Jansen, G.; Blaeser, D.; Woelper, C.; Schulz, S. Chem.
Eur. J. 2011, 17, 11394.
(8) Villinger, A.; Schulz, A. Angew. Chem., Int. Ed. 2010, 49, 8017.
(9) Schulz, S.; Lyhs, B.; Jansen, G.; Blaeser, D.; Woelper, C. Chem.
Commun. 2011, 47, 3401.
(10) Schulz, A.; Villinger, A. Chem.Eur. J. 2012, 18, 2902.
(11) Haiges, R.; Boatz, J. A.; Yousufuddin, M.; Christe, K. O. Angew.
Chem., Int. Ed. 2007, 46, 2869.
(12) Haiges, R.; Schroer, T.; Yousufuddin, M.; Christe, K. O. Z. Anorg.
Allg. Chem. 2005, 631, 2691.
(13) (a) SMART V 5.625, Software for the CCD Detector System; Bruker
AXS: Madison, WI, 2001; (b) SAINT V 6.22, Software for the CCD
Detector System; Bruker AXS: Madison, WI, 2001; (c) SADABS, Software
for the CCD Detector System; Bruker AXS: Madison, WI, 2001;
(d) Sheldrick, G. M. SHELXS-90, Program for the Solution of Crystal
Structure; University of Gottingen: Gottingen, Germany, 1990;
̈
̈
(e) Sheldrick, G. M. SHELXL-97, Program for the Refinement of Crystal
Structure; University of Gottingen: Gottingen, Germany, 1997;
̈
̈
(f) Sheldrick, G. M. SHELXTL 6.12 for Windows NT/2000, Program
413
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414

Inorganic Chemistry
Article
library for Structure Solution and Molecular Graphics; Bruker AXS:
Madison, WI, 2000.
(14) (a) Bruker Instrument Service v2011.4.0.0; Bruker AXS: Madison,
WI, 2011; (b) SAINT V7.68A; Bruker AXS: Madison, WI, 2009;
(c) SADABS V2008/1; Bruker AXS: Madison, WI, 2008; (d) Bruker
SHELXTL V2011.4-0; Bruker AXS: Madison, WI, 2011; (e) Sheldrick,
G. M. Acta Crystallogr. 2008, A64, 112−122.
(15) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(16) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215.
(17) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.;
Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.;
Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. J. A.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N. K., R.; Normand, J.; Raghavachari, K.; Rendell,
A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam,
N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich,
̈
S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.;
Fox, D. J. Gaussian 09, Revision A.02; Gaussian, Inc.: Wallingford, CT,
2009.
(18) Goerigk, L.; Grimme, S. Phys. Chem. Chem. Phys. 2011, 13, 6670.
(19) Valero, R.; Gomes, J. R. B.; Truhlar, D. G.; Illas, F. J. Chem. Phys.
2008, 129, 124710/1.
(20) Zhao, Y.; Truhlar, D. G. J. Chem. Theory Comput. 2011, 7, 669.
(21) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007.
(22) Peterson, K. A.; Yousaf, K. E. J. Chem. Phys. 2010, 133, 174116/1.
414
dx.doi.org/10.1021/ic3021996 | Inorg. Chem. 2013, 52, 402−414