Sila-Substitution of Alkyl Nitrates: Synthesis, Structural Characterization, and Sensitivity Studies of Highly Explosive (Nitratomethyl)-, Bis(nitratomethyl)-, and Tris(nitratomethyl)silanes and Their Corresponding Carbon Analogues

Article abstract of DOI:10.1021/ic902387y

A series of analogous nitratomethyl compounds of carbon and silicon of the formula types Me₃ElCH₂ONO₂ (1a/1b), Me ₂El(CH₂ONO₂)₂ (2a/2b), MeEl(CH ₂ONO₂)₃ (3a/3b), (CH₂) ₄El(CH₂ONO₂)₂ (4a/4b), and (CH ₂)₅El(CH₂ONO₂)₂ (5a/5b) were synthesized [El = C (a), Si (b); (CH₂)₄El = (sila)cyclopentane-1,1-diyl; (CH₂)₅El = (sila)cyclohexane-1,1-diyl]. All compounds were characterized by using NMR, IR, and Raman spectroscopy and mass spectrometry. In addition, the crystal structures of Me₂C(CH₂ONO₂)₂ (2a), (CH₂)₄C(CH₂ONO₂)₂ (4a), Me₂Si(CH₂ONO₂)₂ (2b), and (CH ₂)₅Si(CH₂ONO₂)₂ (5b) were determined by single-crystal X-ray diffraction. The gas-phase structures of the C/Si analogues 1a and 1b were determined by electron diffraction and compared with the results of quantum chemical calculations at different levels of theory. The thermal stabilities of the C/Si pairs 1a/1b-5a/5b were investigated by using DSC. In addition, their friction and impact sensitivities were measured with standard BAM methods. The extreme sensitivities of the silicon compounds 1b-5b compared to those of the corresponding carbon analogues 1a-5a were discussed in terms of the structures of the C/Si analogues and possible geminal Si???O interactions.

Full text of DOI:10.1021/ic902387y

Inorg. Chem. 2010, 49, 4865–4880 4865
DOI: 10.1021/ic902387y
Sila-Substitution of Alkyl Nitrates: Synthesis, Structural Characterization,
and Sensitivity Studies of Highly Explosive (Nitratomethyl)-, Bis(nitratomethyl)-,
and Tris(nitratomethyl)silanes and Their Corresponding Carbon Analogues
†
,†
†
†
‡
Camilla Evangelisti, Thomas M. Klap €o tke,* Burkhard Krumm, Anian Nieder, Raphael J. F. Berger,
‡
,‡
§
,§
Stuart A. Hayes, Norbert W. Mitzel,* Dennis Troegel, and Reinhold Tacke*
†
Department of Chemistry, Ludwig Maximilians University of M u€ nchen, Butenandtstrasse 5-13 (D),
‡
D-81377 M u€ nchen, Germany, Faculty of Chemistry, University of Bielefeld, Universit a€ tsstrasse 25,
D-33615 Bielefeld, Germany, and Institute of Inorganic Chemistry, University of W u€ rzburg, Am Hubland,
§
D-97074 W u€ rzburg, Germany
Received December 2, 2009
A series of analogous nitratomethyl compounds of carbon and silicon of the formula types Me ElCH ONO (1a/1b),
3
2
2
Me El(CH ONO ) (2a/2b), MeEl(CH ONO ) (3a/3b), (CH ) El(CH ONO ) (4a/4b), and (CH ) El(CH ONO )
2 2
2
2
2 2
2
2 3
2 4
2
2 2
2 5
2
(
5a/5b) were synthesized [El = C (a), Si (b); (CH ) El = (sila)cyclopentane-1,1-diyl; (CH ) El = (sila)cyclohexane-
2 4 2 5
1
addition, the crystal structures of Me C(CH ONO ) (2a), (CH ) C(CH ONO ) (4a), Me Si(CH ONO ) (2b), and
,1-diyl]. All compounds were characterized by using NMR, IR, and Raman spectroscopy and mass spectrometry. In
2
2
2 2
2 4
2
2 2
2
2
2 2
(
CH ) Si(CH ONO ) (5b) were determined by single-crystal X-ray diffraction. The gas-phase structures of the C/Si
2 5 2 2 2
analogues 1a and 1b were determined by electron diffraction and compared with the results of quantum chemical
calculations at different levels of theory. The thermal stabilities of the C/Si pairs 1a/1b-5a/5b were investigated by
using DSC. In addition, their friction and impact sensitivities were measured with standard BAM methods. The extreme
sensitivities of the silicon compounds 1b-5b compared to those of the corresponding carbon analogues 1a-5a were
discussed in terms of the structures of the C/Si analogues and possible geminal Si
3 3 O interactions.
3
Introduction
cleavage of the O-NO group is mostly the first chemical
2
step of decomposition, followed by ignition, growth of
deflagration, and transition from deflagration to detonation.
This initial chemical step can be initiated by different external
stimuli, like shock, friction, heat, or electrostatic discharge.
In some circumstances, the ignition can lead directly to
Alkyl nitrates are well-known compounds, and they are
used in many different applications such as drugs, propel-
1,2
lants, or explosives. Since Alfred Nobel tamed nitroglycer-
in with kieselguhr, highly sensitive alkyl nitrates have become
2
manageable for industrial, military, and civil purposes. One
2,3
detonation.
of the routinely used alkyl nitrates is pentaerythrityl tetra-
The first sila-analogues of alkyl nitrates (C/Si exchange)
nitrate (PETN), C(CH ONO ) , as a primary explosive. A
2
2 4
4
were reported in 1964, along with the syntheses of related
derivative also bearing a neopentane backbone is 1,1,1-tris-
nitratomethyl)ethane (also known as metriol trinitrate, 3a;
compounds of the formula type R ElCHClCH ONO (El =
3
2
2
(
4
Si, Sn) and organosiloxane-based nitrates obtained from the
Scheme 1), which is used as an alternative to nitroglycerin in
2
corresponding silanols of the formula type (HO) Si(CHCl-
n
propellant and explosive formulations.
5
CH ONO ) (n = 2, 3). The first synthesis of a silicon ana-
2
2 4-n
The characteristic of an explosive reaction is a fast propa-
gating decomposition via a shock-wave mechanism. In the
case of alkyl nitrate based explosives, the homolytic bond
logue of an alkyl nitrate with a neopentane backbone was
6
described in a patent, where the use of Me SiCH ONO (1b,
3
2
2
Scheme 1) was claimed as an alternative to nitroglycerin for
medical applications, and the explosive behavior of this
*
To whom correspondence should be addressed. E-mail: tmk@cup.
uni-muenchen.de (T.M.K.), mitzel@uni-bielefeld.de (N.W.M.), r.tacke@
uni-wuerzburg.de (R.T.). Phone: (þ49)89-2180-77491 (T.M.K.), (þ49) 521-
(3) (a) Hiskey, M. A.; Brower, K. R.; Oxley, J. C. J. Phys. Chem. 1991, 95,
3955–3960. (b) Akhavan, J. The Chemistry of Explosives; The Royal Society of
Chemistry: Cambridge, 1998; pp 63-67. (c) Klap €o tke, T. M. Chemie der
hochenergetischen Materialien; de Gruyter: Berlin, 2009; pp 87-88.
1
06-6182 (N.W.M.), (þ49) 931-31-85250 (R.T.).
(
1) Boschan, R.; Merrow, R. T.; van Dolah, R. W. Chem. Rev. 1955, 55,
85–510.
2) (a) Meyer, R.; K o€ hler, J.; Homburg, A. Explosives, 5th ed.; Wiley-VCH:
4
(
(
4) Fink, W. U.S. Patent 3,127,431, March 31, 1964.
Weinheim, 2002; pp 216-217. (b) Klap €o tke, T. M. Chemie der hochenergetischen
Materialien; de Gruyter: Berlin, 2009; pp 1-6.
(5) Fink, W. U.S. Patent 3,222,319, December 7, 1965.
(6) Barcza, S. U.S. Patent 3,592,831, July 13, 1971.
r 2010 American Chemical Society
Published on Web 04/30/2010
pubs.acs.org/IC

4
866 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
Scheme 1. Carbon/Silicon Analogues 1a/1b-5a/5b with Various
Numbers of Nitratomethyl Groups Bound to the Central Carbon or
Silicon Atom
here represent a logical extension of related investigations
dealing with sila-drugs and sila-odorants.
17
18
Experimental Section
Caution! Many alkyl nitrates are sensitive and represent
energetic materials; therefore, they must be handled with care!
Since sila-analogues of alkyl nitrates are even more sensitive
toward shock and friction, they must be manipulated very cau-
tiously! During the work with alkyl nitrates, and especially with
their sila-analogues, wearing leather jacket, face shield, steel-
reinforced Kevlar gloves, ear protection, and electrically grounded
shoes is mandatory. Only electrically grounded and metal-free
equipment was used during the syntheses.
Materials. Nitric acid (100%, fuming), Me
SiCH I, Me Si(CH Cl) , acetonitrile (Aldrich), MeC(CH
acetic anhydride (Acros Organics), Me CCH OH (Aldrich and
3
SiCH
2
OH, Me
3
-
2
2
2
2
2 3
OH) ,
3
2
Acros Organics), and silver nitrate (VWR) were used as received.
1
13
1
General Remarks. The H, C{ H}, N, and Si{ H} NMR
14
29
1
spectra were recorded using a Jeol 400 eclipse FT-NMR spectro-
1
meter (compounds 1a/b-5a/b) operating at 400.2 MHz ( H),
1
3
14
29
00.6 MHz ( C), 28.9 MHz ( N), and 79.5 MHz ( Si) or a
1
Bruker DRX-300 NMR spectrometer (compounds 6-11) oper-
ating at 300.1 MHz ( H), 75.5 MHz ( C), and 59.6 MHz
1
13
2
9
1
Si). Chemical shifts (ppm) are given with respect to TMS ( H,
(
13
29
14
C, Si; δ 0) and MeNO ( N; δ 0) as external standards (com-
2
1
pounds 1a/b-5a/b) or relative to internal [D
δ 2.49; [D
5
]DMSO ( H,
]DMSO ( C, δ 39.5; [D ]-
( H, δ 7.24; CDCl ), internal CDCl
C, δ 77.0; CDCl ), or external TMS ( Si, δ 0; [D ]DMSO,
compound was recognized, with possible applications as an
energetic material.
13
6
]DMSO), internal [D
6
6
1
DMSO), internal CHCl
3
3
3
Recently, the extremely impact- and friction-sensitive
silicon analogue of PETN [C(CH ONO ) ), sila-PETN
13
29
(
3
6
1
2
2 4
CDCl ) (compounds 6-11). Analysis and assignment of the H
3
7
1 1 13 1
(
Si(CH ONO ) ], was synthesized. The first studies of its
NMR data of 6-11 were supported by H, H COSY, C, H
13 1
HMQC, and C, H HMBC experiments, and assignment of the
2
2 4
outstanding sensitivity compared to its carbon analogue
PETN were performed and different decomposition path-
ways of the C/Si analogues C(CH ONO ) and Si-
1
3
13
1
C NMR data was supported by DEPT135, C, H HMQC,
1
3
1
and C, H HMBC experiments. Infrared (IR) spectra were
recorded on a Perkin-Elmer Spectrum BXII FT-IR instrument
equipped with a Diamant-ATR Dura Sampler at 25 ꢀC. Raman
spectra were recorded on a Perkin-Elmer Spectrum 2000R
NIR FT-Raman instrument equipped with a Nd:YAG laser
2
2 4
(
CH ONO ) were investigated by using quantum chemical
2
2 4
,8
7
methods.
To study the characteristics and the reasons for the high
instability of (nitratomethyl)silanes, several silicon com-
pounds of this type with various numbers of nitratomethyl
groups and their corresponding carbon analogues, the C/Si
pairs 1a/1b-5a/5b (Scheme 1), were synthesized and char-
acterized, including their friction and impact sensitivities.
The synthetic access to the precursors of the aforementioned
(
1064 nm) with 300 mW at 25 ꢀC, except for compound 3b; the
Raman data of 3b were collected using a Isa Jobin-Yvon T64000
Raman instrument with CCD detector (EEV CCD115-11)
equipped with a Spectra Physics Ar laser (514.5 nm) with
þ
30 mW at 25 ꢀC. Melting, boiling, and decomposition points
were determined by differential scanning calorimetry (DSC;
Perkin-Elmer Pyris 6 DSC, calibrated by standard pure indium
and zinc). Measurements were performed at a heating rate of
(nitratomethyl)silanes, the corresponding (hydroxymethyl)-
and (iodomethyl)silanes, and related compounds with
SiCH X groups (X = Cl, Br, N , SH) has been discussed
2
3
9-16
(17) Recent publications dealing with sila-drugs: (a) B u€ ttner, M. W.;
Burschka, C.; Daiss, J. O.; Ivanova, D.; Rochel, N.; Kammerer, S.; Peluso-
Iltis, C.; Bindler, A.; Gaudon, C.; Germain, P.; Moras, D.; Gronemeyer, H.;
Tacke, R. ChemBioChem 2007, 8, 1688–1699. (b) Tacke, R.; Popp, F.; M €u ller,
B.; Theis, B.; Burschka, C.; Hamacher, A.; Kassack, M. U.; Schepmann, D.;
W €u nsch, B.; Jurva, U.; Wellner, E. ChemMedChem 2008, 3, 152–164.
extensively.
The studies on the sila-explosives reported
(
7) Klap o€ tke, T. M.; Krumm, B.; Ilg, R.; Troegel, D.; Tacke, R. J. Am.
Chem. Soc. 2007, 129, 6908–6915.
8) Liu, W.-G.; Zybin, S. V.; Dasgupta, S.; Klap o€ tke, T. M.; Goddard,
(
(c) Warneck, J. B.; Cheng, F. H. M.; Barnes, M. J.; Mills, J. S.; Montana, J. G.;
W. A., III J. Am. Chem. Soc. 2009, 131, 7490–7491.
Naylor, R. J.; Ngan, M.-P.; Wai, M.-K.; Daiss, J. O.; Tacke, R.; Rudd, J. A.
Toxicol. Appl. Pharmacol. 2008, 232, 369–375. (d) Lippert, W. P.; Burschka, C.;
G €o tz, K.; Kaupp, M.; Ivanova, D.; Gaudon, C.; Sato, Y.; Antony, P.; Rochel, N.;
Moras, D.; Gronemeyer, H.; Tacke, R. ChemMedChem 2009, 4, 1143–1152.
(
(
9) Kobayashi, T.; Pannell, K. H. Organometallics 1991, 10, 1960–1964.
10) Anderson, W. K.; Kasliwal, R.; Houston, D. M.; Wang, Y.;
Narayanan, V. L.; Haugwitz, R. D.; Plowman, J. J. Med. Chem. 1995, 38,
789–3797.
11) Daiss, J. O.; Barth, K. A.; Burschka, C.; Hey, P.; Ilg, R.; Klemm, K.;
Richter, I.; Wagner, S. A.; Tacke, R. Organometallics 2004, 23, 5193–5197.
12) Ilg, R.; Troegel, D.; Burschka, C.; Tacke, R. Organometallics 2006,
5, 548–551.
13) Daiss, J. O.; Burschka, C.; Mills, J. S.; Montana, J. G.; Showell,
G. A.; Warneck, J. B. H.; Tacke, R. Organometallics 2006, 25, 1188–1198.
14) Shimizu, M.; Iwakubo, M.; Nishihara, Y.; Oda, K.; Hiyama, T.
ARKIVOC 2007, 29–48.
15) Troegel, D.; M o€ ller, F.; Burschka, C.; Tacke, R. Organometallics
009, 28, 5765–5770.
16) Apfel, U.-P.; Troegel, D.; Halpin, Y.; Tschierlei, S.; Uhlemann, U.;
3
(
e) Tacke, R.; M €u ller, V.; B €u ttner, M. W.; Lippert, W. P.; Bertermann, R.; Daiss,
J. O.; Khanwalkar, H.; Furst, A.; Gaudon, C.; Gronemeyer, H. ChemMedChem
009, 4, 1797–1802. (f) Johansson, T.; Weidolf, L.; Popp, F.; Tacke, R.; Jurva, U.
(
2
(
Drug Metab. Dispos. 2010, 38, 73–83. (g) Tacke, R.; Nguyen, B.; Burschka, C.;
Lippert, W. P.; Hamacher, A.; Urban, C.; Kassack, M. U. Organometallics 2010,
2
(
2
9, 1652-1660.
(
18) Recent publications dealing with sila-odorants: (a) B u€ ttner, M. W.;
(
Burschka, C.; Junold, K.; Kraft, P.; Tacke, R. ChemBioChem 2007, 8, 1447–
1454. (b) B €u ttner, M. W.; N €a tscher, J. B.; Burschka, C.; Tacke, R. Organome-
tallics 2007, 26, 4835–4838. (c) Tacke, R.; Metz, S. Chem. Biodiversity 2008, 5,
920–941. (d) Metz, S.; N €a tscher, J. B.; Burschka, C.; G €o tz, K.; Kaupp, M.; Kraft,
P.; Tacke, R. Organometallics 2009, 28, 4700–4712. (e) N €a tscher, J. B.;
Laskowski, N.; Kraft, P.; Tacke, R. ChemBioChem 2010, 11, 315–319.
(
2
(
Schmitt, M.; Popp, J.; G o€ rls, H.; Dunne, P.; Venkatesan, M.; Coey, M.;
Rudolph, M.; Vos, J. G.; Tacke, R.; Weigand, W., manuscript in preparation.

Article
Inorganic Chemistry, Vol. 49, No. 11, 2010 4867
β = 5 ꢀC in closed aluminum containers with a hole (1 μm) on
the top for gas release with a nitrogen flow of 5 mL/min. The
reference sample was an closed aluminum container with air.
Friction and impact sensitivities were determined by standard
BAM methods. A B u€ chi GKR-51 apparatus was used for the
bulb-to-bulb distillations of 9-11. The melting points of 7 and 8
were determined with a B u€ chi Melting Point B-540 apparatus
using samples in sealed glass capillaries. The crystal structures
were determined by using an Oxford Xcalibur CCD diffracto-
meter (compounds 2a, 2b, 4a, and 5b) or a Stoe IPDS diffracto-
calculated electron wavelength was assumed to be correct,
and the nozzle-to-plate distances were checked for consistency
with the measured values leading to a small scaling being applied
to the long distance. For 1b, the data reduction was performed
using the previously measured nozzle-to-plate distances, and a
small scaling was applied to the electron wavelength for the data
set recorded at the short distance. The electron wavelengths and
nozzle-to-plate distances are provided as Supporting Informa-
tion, along with other data analysis parameters including the
19
s-limits, weighting points, R factors (R
data correlation values, and the correlation matrix.
The amplitudes of vibration, u, used in the r and rh1 refine-
D G
and R ), scale factors,
meter (compounds 7 and 8) and graphite-monochromated
˚
Mo-K
R
radiation (λ = 0.71073 A). The structures were solved
with SHELXS-97 and were refined with SHELXL-97, imple-
g
2
1
ments, and the distance corrections for curvilinear perpendicular
motion, kh1, were calculated by using the program SHRINK.
2
mented in the program package WinGX and finally checked
2
30,31
23
using Platon. Elemental analyses of the nitratomethyl compounds
were not performed because of their highly sensitive and explo-
sive properties. Mass spectrometric data were obtained from a
Jeol Mstation JMS 700 spectrometer, except for 2b and 3b,
which were too hazardous.
This made use of frequency calculations for 1a and 1b, respec-
tively, at the RI-MP2/TZVPP and HF/6-31G* level of theory
(for details of calculations, see below). The SHRINK input file
for 1b was generated using EasyInp written by K. B. Borisenko,
while that for 1a was generated using a combination of EasyInp
32
Gas Electron Diffraction (GED). Electron scattering intensi-
ties were recorded at room temperature on a combination of
reusable Fuji and Kodak imaging plates using a Balzers KD-G2
and Q2SHRINK, written by A. Zakharov.
Theoretical Calculations. All Hartree-Fock (HF) calcula-
tions were of the restricted type, and the second-order Møller-
Plesset (MP2) calculations made use of the resolution-of-the-
identity (RI) method and the default frozen-core partitioning
2
4
Gas-Eldigraph
(formerly operated in T u€ bingen by H.
2
5
Oberhammer ). This has been equipped with an electron source
built by STAIB Instruments, which was operated at 60 kV (1a)
or 70 kV (1b). The accelerating voltage was monitored using a
3
3
as implemented in Turbomole. The HF calculations using the
Pople-style 6-31G* basis sets were performed by using the
3
4a
0
to 10 V signal generated by the high-voltage supply, propor-
default criteria in Gaussian 03, while those using the def2-
TZVPP (herein shortened to TZVPP) basis set were performed
by using Turbomole.
tional to the variable 0 to 100 kV output, which was stable to
within 0.1 to 0.2 mV during the course of the experiment. The
image plates were scanned using a Fuji BAS 1800 scanner,
yielding digital 16-bit gray scale image data. Further details
about the Bielefeld GED apparatus and the experimental
The theoretical calculations for structure optimization of 2a
and 2b were performed at the B3LYP level of theory (cc-pVDZ
3
4b
basis sets) using Gaussian 03,
and the NBO analyses were
2
6
method are published elsewhere.
performed using the Gaussian 03 nbo5-tool. Instead, the calcu-
lations of the enthalpies of formation were performed at the
CBS-4 M level of theory.
In preparation for data reduction, the long and short nozzle-
to-plate distances were remeasured after recording the short-
distance data and before recording the long-distance data for 1a.
The relative scaling of the two scanning directions was also
recalibrated by using an exposed image plate with two pairs of
pin holes, which was scanned in two orientations, approxi-
mately perpendicular to one another. The data were reduced
General Procedure for the Nitration of the (Hydroxymethyl)-
alkanes and (Hydroxymethyl)cycloalkanes. For the synthesis of
1a, 2a, 4a, and 5a, the respective hydroxymethyl or bis(hydroxy-
methyl) compound [Me CCH OH, Me C(CH OH) , (CH ) -
3
2
2
2
2
2 4
C(CH OH) , (CH ) C(CH OH) ; 1.0 mmol (1.0 mol equiv)]
2
2
2 5
2
2
2
7
to total intensities using Strand et al.’s program PIMAG
version 040827) in connection with a sector curve, which was
was added in small portions at 20 ꢀC to a stirred mixture of nitric
acid (100%; 2.1 and 4.2 mol equiv, respectively) and acetic
anhydride (6.0 mol equiv). [For the synthesis of 3a, the tris-
(hydroxymethyl) compound MeC(CH OH) was added in small
(
based on experimental xenon scattering data and tabulated
scattering factors of xenon. Further data reduction yielding
molecular-intensity curves was performed by using version
2
3
portions at 0 ꢀC to a stirred mixture of nitric acid (100%; 3.3 mol
equiv) and dichloromethane (5 mL).] The reaction mixture was
then stirred at ambient temperature for 1.5 (1a, 2a, 4a, 5a) or 2 h
(3a), followed by the addition of ice-water (5 mL) and pentane
(15 mL). The aqueous phase was separated and extracted with
pentane (2 ꢀ 20 mL), and the combined organic extracts were
washed with water (2 ꢀ 20 mL), neutralized by washing with a
concentrated aqueous solution of sodium hydrogen carbonate,
and then dried over anhydrous sodium sulfate. The solvent was
removed under reduced pressure, and the residue was purified
by distillation (1a) or by removal of the remaining solvent traces
in vacuo (0.01 mbar, 20 ꢀC, 3-6 h) (2a-5a). For further details
of the syntheses, see Table 1.
2
8
2
.4 of the ed@ed program, using the scattering factors of
29
Ross et al. For both compounds the ratio of the electron
wavelength to the nozzle-to-plate distances was checked using
˚
benzene data and the widely accepted r value of 1.397 A for the
a
C-C distance in benzene. In the determination of 1a, the
(
19) (a) http://www.bam.de/index_en.html. (b) Official Journal of the
European Union, Council Regulation (EC) No. 440/2008, 30th of March 2008.
20) Hylands, K. A.; Moodie, R. B. J. Chem. Soc., Perkin Trans. 2 1997,
09–714.
21) Sheldrick, G. M. SHELXL-97, A program for the refinement of
crystal structures; University of G €o ttingen: G €o ttingen, Germany, 1994.
(
7
(
(22) Farrugia, L. J. WinGX, J. Appl. Crystallogr. 1999, 32, 837–838.
(23) Spek, A. L. Platon; Utrecht University: Utrecht, The Netherlands, 1999.
(24) Zeil, W.; Haase, J.; Wegmann, L. Z. Instrum. 1966, 74, 84–88.
(25) Oberhammer, H. In Molecular Structure by Diffraction Methods;
2
3 2 2
,2-Dimethyl-1-nitratopropane, Me CCH ONO (1a). The
crude product was distilled in vacuo (60 ꢀC/50 mbar) to give
Sim, G. A, Sutton, L. E., Eds.; The Chemical Society: London, 1976; Vol. 4,
pp 24-44.
(30) Sipachev, V. A. J. Mol. Struct., THEOCHEM 1985, 121, 143–151.
(31) Novikov, V. P.; Sipachev, V. A.; Kulikova, E. I.; Vilkov, L. V. J. Mol.
Struct. 1993, 301, 29–36.
(
26) Berger, R. J. F.; Mitzel, N. W. Z. Naturforsch. 2009, 64b, 1259-1268.
27) Gundersen, S.; Samdal, S.; Strand, T. G.; Volden, H. V. J. Mol.
(
Struct. 2007, 832, 164–171.
28) Hinchley, S. L.; Robertson, H. E.; Borisenko, K. B.; Turner, A. R.;
(32) Zakharov, A. V. Q2SHRINK, version 1.3; http://edsoftware.
(
sourceforge.net.
Johnston, B. F.; Rankin, D. W. H.; Ahmadian, M.; Jones, J. N.; Cowley,
(33) Turbomole, version 5.7; Ahlrichs, R.; B €a r, M.; H €a ser, M.; Horn, H.;
A. H. Dalton Trans. 2004, 2469–2476.
K €o lmel, C. Chem. Phys. Lett. 1989, 162, 165-169.
(
29) Ross, A. W.; Fink, M.; Hilderbrandt, R. In International Tables for
(34) (a) Frisch, M. J. et al. Gaussian 03, revision C.02; Gaussian, Inc.:
Wallingford, CT, 2004. (b) Frisch, M. J. et al. Gaussian 03, revision D.01;
Gaussian, Inc.: Wallingford, CT, 2004.
Crystallography; Wilson, A. C. J., Ed.; Kluwer Academic Publishers: Dordrecht,
The Netherlands, 1992; Vol. C, pp 245-338.

4
868 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
Table 1. Nitration Procedures, Molar Equivalents of the Nitration Agent HNO
3
, and Reaction Times for the Syntheses of the Carbon-Based Nitratomethyl Compounds
1
a-5a as well as Yields and Melting, Boiling, and Decomposition Points of 1a-5a
compound 1a 2a
HNO /Ac HNO /Ac
3a
4a
HNO /Ac
4.2
1.5
83
5a
nitration agents
3
2
O
3
2
O
HNO
6.3
2.0
86
3
(CH
2
Cl
2
)
3
2
O
HNO /Ac
3
2
O
a
molar equivalents of nitric acid
reaction time [h]
% yield
2.1
1.5
88
-/174
4.2
1.5
82
4.2
1.5
86
-/190
b
b
b
b
b
mp./bp. [ꢀC]
18-23/178
-15/182
23-25/190
a
b
Calculated for 1.0 mol equiv of the starting material. Decomposition.
Table 2. Nitration Procedures (Methods A and B), Molar Equivalents of the Nitration Agent, and Reaction Times for the Syntheses of the Silicon-Based Nitratomethyl
Compounds 1b-5b as well as Yields and Melting, Boiling, and Decomposition Points of 1b-5b
compound
nitration procedure
1b
2b
3b
4b
5b
B
A
A
A
A
4.2
1.5
96
a
molar equivalents of nitration agent
reaction time [h]
% yield
2.0
1.5
86
4.2
1.0
87
6.3
1.5
80
4.2
1.5
94
b
b
b
b
b
b
mp./bp. [ꢀC]
-/85
-10/107
∼ -18/107
∼ -18/97
∼ 4/96
a
b
Calculated for 1.0 mol equiv of the starting material. Decomposition.
1
þ
1
(
(
2
(
(
1
9
6
a 88% yield as a colorless liquid. NMR (CDCl ): H, δ 0.97
222.3 [M þ 2H] . Mp.: 23-25 ꢀC. Bp.: 190 ꢀC (dec.). Sensitivities:
impact >100 J; friction >108 N.
1,1-Bis(nitratomethyl)cyclohexane, (CH ) C(CH ONO ) (5a).
3
1
3
s, 9 H, CCH ), 4.12 (s, 2 H, CCH O); C, δ 26.4 (CCH ), 31.4
3 2 3
14
C ), 82.1 (CCH O); N, δ -41.2. Raman (300 mW): 2968 (vs),
q
2
2 5
2
2 2
910 (vs), 2796 (w), 1629 (w), 1460 (m), 1365 (w), 1280 (s), 1261
m), 1261 (w), 1045 (vw), 975 (vw), 921 (m), 868 (m), 769 (s), 696
The crude product was dried in vacuo (0.01 mbar, 25 ꢀC, 3 h) to
1
give 5a in 86% yield as a colorless liquid. NMR (CDCl ): H,
3
-
1
w), 611 (m), 478 (m), 405 (vw) cm . IR: 2963 (w), 2875 (vw),
13
δ 1.45-1.55 (m, 10 H, CCH C), 4.37 (s, 4 H, CCH O); C, δ
20.9 (γ-CH ), 25.5 (δ-CH ), 29.9 (β-CH ), 37.2 (C ), 74.5
(CCH O); N, δ -43.5. Raman (300 mW): 2946 (vs), 2864 (s),
1635 (w), 1473 (w), 1447 (m), 1388 (vw), 1365 (w), 1283 (s), 1251
(w), 1030 (w), 988 (w), 867 (s), 852 (s), 833 (m), 684 (vw), 606 (m),
475 (m) cm . IR: 2936 (w), 2862 (vw), 1622 (vs), 1455 (w), 1371
(vw), 1316 (vw), 1280 (vs), 1267 (vs), 1078 (vw), 1018 (vw),
988 (m), 930 (m), 930 (w), 843 (vs), 755 (s), 702 (m), 684 (vw),
2
2
621 (vs), 1479 (vw), 1467 (vw), 1279 (vs), 1219 (vw), 1043 (vw),
76 (w), 943 (vw), 924 (vw), 865 (m), 847 (m), 759 (vw), 694 (vw),
38 (vw) cm . MS (DEIþ) [m/e]: 133.1 [M ]. Bp.: 174 ꢀC.
2
4
2
2
q
1
2
-
1
þ
Sensitivities: impact, >100 J; friction, >360 N.
,2-Dimethyl-1,3-dinitratopropane, Me C(CH
-
1
2
2
2
ONO
2
)
2
(2a).
The crude product was dried in vacuo (0.01 mbar, 25 ꢀC, 3 h)
1
to give 2a in 82% yield as a colorless liquid. NMR (CDCl ): H,
3
1
3
-1
þ
δ 1.08 (s, 6 H, CCH
3
), 4.26 (s, 4 H, CCH
2
O); C, δ 22.1 (CCH
O); N, δ -44.5. Raman (300 mW): 2969
vs), 2910 (s), 2808 (vw), 2732 (vw), 1635 (w), 1467 (m), 1403
vw), 1372 (w), 1285 (s), 1032 (w), 982 (vw), 921 (w), 870 (s), 795
w), 693 (w), 608 (m), 475 (m), 408 (vw) cm . IR: 2977 (w), 2900
vw), 1826 (w), 1755 (vw), 1623 (vs), 1477 (w), 1369 (m), 1268
vs), 1230 (w), 1122 (m), 1032 (w), 977 (s), 943 (w), 923 (w), 843
3
),
636 (w) cm . MS (DEIþ) [m/e]: 235.2 [M þ H] . Bp.: 190 ꢀC
1
4
3
4.7 (C ), 76.5 (CCH
q
2
(dec.). Sensitivities: impact >100 J; friction, >120 N.
(
(
(
(
(
(
General Procedure of the Nitration of the (Hydroxymethyl)-
silanes and (Hydroxymethyl)silacycloalkanes. Method A. For
the synthesis of 2b-5b, the respective bis(hydroxymethyl)
-
1
or tris(hydroxymethyl) compound [Me
Si(CH
2
OH)
2
, MeSi-
2
(CH OH) , (CH Si(CH OH) , (CH Si(CH
5
2
3
2
)
4
2
2
2
)
2
OH)
2
; 1.0 mmol
-
1
vs), 756 (s), 718 (w), 703 (w), 637 (w) cm . MS (DEI-) [m/e]:
(1.0 mol equiv)] was added in small portions to a stirred mixture
of nitric acid (100%; 4.2 and 6.3 mol equiv, respectively) and
acetic anhydride (6.0 mol equiv). The reaction mixture was
stirred at ambient temperature for 1 h (2b) or 1.5 h (3b-5b),
followed by the addition of ice-water (5 mL) and pentane
(15 mL). The aqueous phase was separated and extracted with
pentane (2 ꢀ 20 mL), and the combined organic extracts were
washed with water (2 ꢀ 20 mL), neutralized by washing with a
concentrated aqueous solution of sodium hydrogen carbonate,
and then dried over anhydrous sodium sulfate. The solvent was
removed under reduced pressure, and the residue was purified
by removal of the remaining solvent traces in vacuo (0.01 mbar,
25 ꢀC, 2 h). For further details of the syntheses, see Table 2.
Method B. For the synthesis of 1b, the respective iodomethyl
-
1
93.1 [M - H] . Mp.: 18-23 ꢀC. Bp.: 178 ꢀC (dec.). Sensitivities:
impact >100 J; friction >96 N.
-Methyl-2-nitratomethyl-1,3-dinitratopropane, MeC(CH
3a). The crude product was dried in vacuo (0.01 mbar, 25 ꢀC, 6 h)
to give 3a in 86% yield as a colorless viscous liquid. NMR
2
2 2 3
ONO )
(
1
13
(
CDCl
3
): H, δ 1.08 (s, 3 H, CCH
), 34.7 (C ), 76.5 (CCH
300 mW): 2972 (vs), 2910 (m), 2802 (vw), 2738 (vw), 1641 (m),
3
), 4.26 (s, 6 H, CCH
2
O); C,
1
4
δ 22.1 (CCH
(
1
(
(
(
(
3
q
2
O); N, δ -44.5. Raman
467 (m), 1407 (vw), 1374 (w), 1286 (vs), 1036 (w), 994 (vw), 918
-
1
vw), 869 (s), 639 (m), 601 (m), 465 (w), 414 (w) cm . IR: 2962
vw), 2907 (vw), 1825 (vw), 1753 (vw), 1626 (vs), 1473 (w), 1374
w), 1271 (vs), 1124 (m), 1031 (w), 990 (s), 923 (w), 835 (vs), 752
-
1
þ
s), 724 (m), 704 (m) cm . MS (DCIþ) [m/e]: 256.1 [M þ H] .
Mp.: -15 ꢀC (ref 2: -15 ꢀC). Bp.: 182 ꢀC (dec.) (ref 2: 182 ꢀC).
Sensitivities: impact >15 J; friction >108 N.
compound Me SiCH I [1.0 mmol (1.0 mol equiv)] was added at
3 2
0 ꢀC to a stirred solution of silver nitrate (2.0 mol equiv) in
acetonitrile (3 mL). The reaction mixture was then stirred at
ambient temperature for 1.5 h under exclusion of light, followed
by the addition of pentane (10 mL) under vigorous stirring. The
pentane phase was separated, and the extraction procedure with
pentane was repeated twice. The pentane extracts were com-
bined, and the solvent was removed by microdistillation (40 ꢀC,
200 mbar). For further details of the synthesis, see Table 2.
1
,1-Bis(nitratomethyl)cyclopentane, (CH ) C(CH ONO ) (4a).
2 4 2 2 2
The crude product was dried in vacuo (0.01 mbar, 25 ꢀC, 3 h) to give
1
4
1
a in 83% yield as a colorless liquid. NMR (CDCl
.57-1.62 (m, 4 H, β-CH ), 1.67-1.72 (m, 4 H, γ-CH
s, 4 H, CCH O); C, δ 24.9 (γ-CH ), 32.8 (β-CH ), 44.8 (C ), 75.0
3
): H, δ
2
2
), 4.33
13
(
(
2
14
2
2
q
CCH
2
O); N, δ -43.9. Raman (300 mW): 2967 (vs), 2877 (s),
635 (w), 1454 (m), 1384 (vw), 1284 (s), 1239 (w), 1029 (w), 991
vw), 906 (m), 869 (s), 703 (w), 607 (m), 471 (w) cm . IR: 2958 (w),
874 (vw), 1728 (m), 1622 (vs), 1453 (w), 1370 (vw), 1338 (vw), 1272
vs), 1166 (w), 1130 (vw), 1024 (vw), 990 (m), 956 (w), 851 (vs),
55 (s), 700 (w), 668 (vw), 637 (vw) cm . MS (DEIþ) [m/e]:
1
(
2
(
-1
Trimethyl(nitratomethyl)silane, Me
pound 1b was synthesized according to Method B and was
3 2 2
SiCH ONO (1b). Com-
1
isolated in 86% yield as a colorless liquid. NMR (CDCl
3
): H, δ
0.10 (s, 9 H, J( H, Si) = 3.4 Hz, SiCH ), 4.04 (s, 2 H,
3
-
1
2
1
29
7

Article
Inorganic Chemistry, Vol. 49, No. 11, 2010 4869
2
1
J( H, Si) = 2.2 Hz, SiCH
29
13
1
13 29
2
O); C, δ -2.8 ( J( C, Si) =
9-11) in methanol, and the resulting mixture was heated under
reflux for 20 h. The solvent was removed under reduced
pressure, and the residue was purified as described in the
respective protocols given below.
1
13 29
14
2
3 2
6.5 Hz, SiCH ), 66.8 ( J( C, Si) = 24.2 Hz, SiCH O); N,
29
δ -34.3; Si, δ 0.1. Raman (300 mW): 2961 (m), 2904 (vs), 1635
(
w), 1413 (w), 1305 (m), 1248 (vw), 1213 (vw), 874 (w), 826 (m),
-
1
25 (vw), 706 (w), 649 (m), 608 (m), 554 (m) cm . IR: 2961 (vw),
630 (vs), 1436 (vw), 1302 (vs), 1252 (s), 1213 (vw), 977 (vw),
7
1
8
1
Bis(hydroxymethyl)dimethylsilane, Me Si(CH OH) (6). This
2
2
2
compound was synthesized from 9 (9.15 g, 44.8 mmol), metha-
nol (400 mL), and acetyl chloride (563 mg, 7.17 mmol). The
crude product was purified by fractional distillation using a
-1
22 (vs), 774 (s), 754 (s), 706 (m) cm . MS (DCIþ) [m/e]:
þ
49.2 [M ]. Decomp. point: 85 ꢀC. Sensitivities: impact >1 J;
friction >64 N.
Dimethylbis(nitratomethyl)silane, Me
Vigreux column to give 6 in 73% yield as a colorless liquid (3.96 g,
1
2
Si(CH
2
ONO
2
)
2
(2b).
Compound 2b was synthesized according to Method A and
32.9 mmol). Bp.: 65-66 ꢀC/0.2 mbar. NMR ([D ]DMSO): H,
6
3
1
1
δ -0.03 (s, 6 H, SiCH ), 3.17 (d, J( H, H) = 4.3 Hz, 4 H,
3
3 1 1 13
was isolated in 87% yield as a colorless oil. NMR (CDCl
3
):
SiCH O), 3.91 (t, J( H, H) = 4.3 Hz, 2 H, OH); C, δ -6.2
2
1
2
1
29
29
H, δ 0.29 (s, 6 H, J( H, Si) = 3.5 Hz, SiCH
3
), 4.18 (s, 4 H,
(SiCH ), 51.4 (SiCH O); Si, δ -2.7. Anal. Calcd for C H O Si:
3
2
4
12 2
2 1
J( H, Si) = 2.3 Hz, SiCH
29
13
1
13 29
2
O); C, δ -6.0 ( J( C, Si) = 27.7
C, 39.96; H, 10.06. Found: C, 39.9; H, 9.9.
1
13 29
14
3 2
Hz, SiCH ), 63.6 ( J( C, Si) = 26.5 Hz, SiCH O); N, δ -37.6;
2 4 2 2
1,1-Bis(hydroxymethyl)-1-silacyclopentane, (CH ) Si(CH OH)
2
9
Si, δ -1.2. Raman (300 mW): 2971 (m), 2910 (vs), 2853 (w),
638 (vw), 1435 (vw), 1305 (m), 1254 (vw), 1213 (vw), 979 (vw),
(7). This compound was synthesized from 10 (3.15 g, 13.7 mmol),
methanol (150 mL), and acetyl chloride (142 mg, 1.81 mmol).
The crude product was purified by bulb-to-bulb distillation
(85 ꢀC/0.03 mbar), and the resulting colorless liquid was crystal-
lized from acetonitrile (4 mL; slow cooling to -20 ꢀC and
crystallization over a period of 24 h). The product was isolated
by removal of the mother liquor via a syringe and then dried in
vacuo (0.2 mbar, 20 ꢀC, 2 h) to give 7 in 60% yield as a colorless
1
8
2
1
-
1
33 (m), 731 (vw), 671 (m), 646 (vw), 614 (vw), 551 (m) cm . IR:
961 (vw), 2919 (vw), 1724 (vw), 1625 (vs), 1437 (vw), 1405 (vw),
297 (vs), 1250 (s), 1212 (w), 1163 (vw), 1072 (w), 980 (w), 817
-
1
(
vs), 752 (m), 670 (m), 641 (m) cm . Mp.: -10 ꢀC. Decomp.
point: 107 ꢀC. Sensitivities: impact <0.5 J; friction <5 N.
Methyltris(nitratomethyl)silane, MeSi(CH ONO ) (3b). Com-
pound 3b was synthesized according to Method A and was isolated
2
2 3
crystalline solid (1.20 g, 8.24 mmol). Mp.: 37-38 ꢀC. NMR
1
1
in 80% yield as a colorless oil. NMR (CDCl
3
2
): H, δ 0.45 (s, 3 H,
([D
6
]DMSO): H, δ 0.53-0.59 (m, 4 H, β-CH
2
), 1.49-1.54 (m,
O), 3.99 (t,
), 26.8 (γ-CH ),
Si: C, 49.27;
2 1
J( H, Si) = 3.6 Hz, SiCH ), 4.32 (s, 6 H, J( H, Si) = 2.3 Hz,
29
1
29
3
1
1
4 H, γ-CH
2
), 3.27 (d, J( H, H) = 4.4 Hz, 4 H, SiCH
J( H, H) = 4.4 Hz, 2 H, OH); C, δ 7.6 (β-CH
2
3
13
1
13 29
3
1
1
13
SiCH
2
O); C, δ -8.7 ( J( C, Si) = 28.8 Hz, SiCH
3
), 60.7
O); N, δ -40.3; Si, δ -5.4.
Raman (30 mW): 2972 (s), 2921 (vs), 1639 (w), 1434 (w), 1306 (vs),
2
2
1
13 29
14
29
29
(
J( C, Si) = 28.1 Hz, SiCH
2
49.9 (SiCH
H, 9.65. Found: C, 49.3; H, 9.6.
1,1-Bis(hydroxymethyl)-1-silacyclohexane, (CH ) Si(CH OH)
2
2
O); Si, 14.2. Anal. Calcd for C
6
H
14
O
2
1
6
257 (w), 1213 (vw), 989 (w), 839 (vs), 744 (w), 670 (w), 647 (w),
2
5
2
-
1
19 (w), 528 (m) cm . IR: 2926 (vw),1628 (vs), 1433 (w), 1404
(8). This compound was synthesized from 11 (6.04 g, 24.7 mmol),
methanol (370 mL), and acetyl chloride (558 mg, 7.11 mmol). The
crude product was crystallized from acetonitrile (15 mL; slow
cooling to -20 ꢀC and crystallization over a period of 24 h). The
product was isolated by removal of the mother liquor via a
syringe and then dried in vacuo (0.1 mbar, 20 ꢀC, 3 h) to give 8
(vw), 1298 (vs), 1253 (s), 1214 (m), 1098 (w), 986 (m), 811 (vs), 746
-
1
(
s) cm . Mp.: ∼ -18 ꢀC. Decomp. point: 107 ꢀC. Sensitivities:
impact <0.5 J; friction <5 N.
,1-Bis(nitratomethyl)-1-silacyclopentane, (CH ) Si(CH ONO )
2 2
1
2
4
2
(4b). Compound 4b was synthesized according to Method A and
was isolated in 94% yield as a colorless oil. NMR (CDCl
0
1
3
): H, δ
in 70% yield as a colorless crystalline solid (2.76 g, 17.2 mmol).
1
.80-0.86 (m, 4 H, β-CH ), 1.63-1.69 (m, 4 H, γ-CH ), 4.28 (s, 4
Mp.: 39-40 ꢀC. NMR ([D
β-CH ), 1.30-1.38 (m, 2 H, δ-CH
3.25 (d, J( H, H) = 4.4 Hz, 4 H, SiCH
6
]DMSO): H, δ 0.60-0.64 (m, 4 H,
2
2
13 29
2
1
29
13
1
H, J( H, Si) = 2.2 Hz, SiCH O); C, δ 8.2 ( J( C, Si) = 27.3
2
2
), 1.58-1.67 (m, 4 H, γ-CH
2
),
O), 3.99 (t, J( H, H) =
), 24.1 (δ-CH ), 29.5 (γ-CH ),
O); Si, -7.7. Anal. Calcd for C Si: C, 52.45;
H, 10.06. Found: C, 52.1; H, 9.8.
2
1
), 62.4( J( C, Si) = 25.8 Hz, SiCH
N, δ -37.8; Si, δ 13.7. Raman (300 mW): 2942 (vs), 2923 (vs),
861 (m), 1638 (vw), 1453 (w), 1435 (w), 1407 (w), 1303 (m), 1251
13 29
3
1
1
3
1
1
Hz, β-CH
2
), 26.8(γ-CH
2
2
O);
2
14
29
13
4.4 Hz, 2 H, OH); C, δ 7.9 (β-CH
2
2
2
2
9
2
49.6 (SiCH
2
7 16 2
H O
(
(
(
(
w), 1210 (vw), 1194 (vw), 1153 (vw), 1080 (vw), 1017 (vw), 979
vw), 944 (vw), 848 (vs), 705 (w), 655 (w), 611 (vw), 554 (m), 420
vw) cm . IR: 2935 (w), 2863 (vw), 1625 (vs), 1452 (vw), 1431
vw), 1403 (vw), 1297 (vs), 1250 (s), 1212 (w), 1152 (vw), 1077 (m),
General Procedure for the Synthesis of the Bis(acetoxy-
-1
16
methyl)silanes 9-11. The respective bis(chloromethyl)silane
was added in a single portion at ambient temperature to a stirred
suspension of sodium acetate (3 mol equiv) in N,N-dimethyl-
formamide, and the resulting mixture was then stirred under
reflux for 18 h. The solvent was removed by distillation (45 ꢀC/
10 mbar), diethyl ether (200 mL) and water (200 mL) were added
to the residue, the organic layer was separated, and the aqueous
phase was extracted with diethyl ether (2 ꢀ 200 mL) and then
discarded. The combined organic extracts were dried over
anhydrous sodium sulfate, the solvent was removed under
reduced pressure, and the residue was purified by bulb-to-bulb
distillation.
1
031 (w), 1018 (w), 977 (w), 814 (vs), 744 (m), 674 (m), 639 (s)
-1
þ
cm . MS (FABþ) [m/e]: 235.1 [M - H] . Mp.: ∼ -18 ꢀC.
Decomp. point: 97 ꢀC. Sensitivities: impact <0.5 J; friction <5 N.
1,1-Bis(nitratomethyl)-1-silacyclohexane, (CH
2 5 2 2 2
) Si(CH ONO )
(
5b). Compound 5b was synthesized according to Method A and
1
was isolated in 96% yield as a colorless oil. NMR (CDCl
0
1
SiCH O); C, δ 8.0 ( J( C, Si) = 26.8 Hz, β-CH ), 23.8 (δ-
CH
δ -37.2; Si, δ -6.5. Raman (300 mW): 2923 (vs), 2858 (s), 1641
3
): H, δ
), 1.40-1.48 (m, 2 H, δ-CH ),
.67-1.77 (m, 4 H, γ-CH ), 4.24 (s, 4 H, J( H, Si) = 2.2 Hz,
.82-0.88 (m, 4 H, β-CH
2
2
2
1
29
2
13
1
13 29
2
2
1
13 29
14
2
2 2
), 29.0 (γ-CH ), 62.2 ( J( C, Si) = 25.9 Hz, SiCH O); N,
29
2 2 2
Bis(acetoxymethyl)dimethylsilane, Me Si(CH OAc) (9). This
(
(
(
vw), 1450 (w), 1403 (vw), 1304 (m), 1213 (vw), 1105 (vw), 1077
vw), 1008 (vw), 982 (vw), 909 (vw), 838 (m), 796 (w), 666 (m), 608
vw), 553 (m) cm . IR: 2925 (w), 2853 (w), 1718 (vw), 1628 (vs),
compound was synthesized from bis(chloromethyl)dimethylsi-
1
6
lane (9.00 g, 57.3 mmol), sodium acetate (14.1 g, 172 mmol),
and N,N-dimethylformamide (95 mL) to give 9 in 79% yield as a
yellowish liquid (9.29, 45.5 mmol). Bp.: 55 ꢀC/0.2 mbar. NMR
-1
1
1
460 (vw), 1447 (vw), 1434 (vw), 1403 (vw), 1297 (vs), 1251 (m),
210 (vw), 1181 (w), 1101 (vw), 1075 (vw), 993 (w), 977 (w), 910
1
(CDCl ): H, δ 0.09 (s, 6 H, SiCH ), 1.99 (s, 6 H, C(O)CH ), 3.78
3
3
3
-1
13
(
m), 832 (s), 801 (s), 752 (s), 641 (s) cm . MS (FABþ) [m/e]:
(s, 4 H, SiCH
(SiCH O), 171.6 (C(O)CH
Si: C, 47.03; H, 7.89. Found: C, 46.8; H, 7.9.
1,1-Bis(acetoxymethyl)-1-silacyclopentane, (CH Si(CH
(10). This compound was synthesized from 1,1-bis(chloromethyl)-
2
O); C, δ -6.1 (SiCH
3
), 20.6 (C(O)CH
); Si, -7.7. Anal. Calcd for
3
), 55.3
þ
29
2
impact <0.5 J; friction >36 N.
49.2 [M - H] . Mp.: ∼ 4 ꢀC. Decomp. point: 96 ꢀC. Sensitivities:
2
3
8 16 4
C H O
General Procedure for the Synthesis of the Bis(hydroxy-
methyl)silanes 6-8. Catalytic amounts of acetyl chloride were
added dropwise at ambient temperature within 1 min to a stirred
solution of the respective bis(acetoxymethyl)silane (compounds
2
)
4
2 2
OAc)
16
1-silacyclopentane (9.13 g, 49.8 mmol), sodium acetate (12.5 g,
152 mmol), and N,N-dimethylformamide (80 mL) to give 10 in

4
870 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
7
6% yield as a yellowish liquid (8.77 g, 38.1 mmol). Bp.: 85 ꢀC/0.4
Scheme 2. Synthesis of the Carbon-Based Nitratomethyl Compounds
1a-5a, Starting from the Corresponding Hydroxymethyl Derivatives
1
mbar. NMR (CDCl
3
): H, δ 0.59-0.64 (m, 4 H, β-CH
2
),
1.51-1.56 (m, 4 H, γ-CH
), 1.98 (s, 6 H, C(O)CH ), 3.85 (s, 4
2 3
13
H, SiCH O); C, δ 8.9 (β-CH ), 20.5 (C(O)CH ), 26.8 (γ-CH ),
2
2
3
2
29
54.9 (SiCH
2
3
O), 171.7 (C(O)CH ); Si, 14.9. Anal. Calcd for
C
10
H
18
O
4
Si: C, 52.15; H, 7.88. Found: C, 52.0; H, 7.9.
,1-Bis(acetoxymethyl)-1-silacyclohexane, (CH Si(CH
11). This compound was synthesized from 1,1-bis(chloromethyl)-
1
2
)
5
2 2
OAc)
(
16
1
1
8
-silacyclohexane (10.0 g, 50.7 mmol), sodium acetate (12.5 g,
52 mmol), and N,N-dimethylformamide (80 mL) to give 11 in
7% yield as a yellowish liquid (10.8 g, 44.2 mmol). Bp.: 90 ꢀC/0.2
1
mbar. NMR (CDCl ): H, δ 0.67-0.72 (m, 4 H, β-CH ),
3
2
1
(
(
(
.34-1.41 (m, 2 H, δ-CH
s, 6 H, C(O)CH ), 3.87 (s, 4 H, SiCH
C(O)CH ), 24.0 (δ-CH ), 29.3 (γ-CH
C(O)CH3); Si, -7.7. Anal. Calcd for C H O Si: C, 54.07; H,
2
), 1.61-1.69 (m, 4 H, γ-CH
2
), 1.99
), 20.7
2
O), 171.6
13
3
2
O); C, δ 8.5 (β-CH
), 54.0 (SiCH
2
3
2
2
29
11
20 4
8.25. Found: C, 53.7; H, 8.1.
Results and Discussion
Syntheses. The acyclic carbon-based nitratomethyl
Scheme 3. Synthesis of the Bis(hydroxymethyl)silanes 6-8, Starting
from the Corresponding Bis(chloromethyl)silanes
3
3
compounds Me CCH ONO (1a), Me C(CH ONO )
2 2
3
2
2
2
2
3
(
2a), and MeC(CH ONO ) (3a) were synthesized accord-
2 2 3
ing to Scheme 2 by treatment of the corresponding
commercially available hydroxymethyl compounds with
fumic nitric acid. The related cyclic nitratomethyl com-
pounds, (CH ) C(CH ONO ) (4a) and (CH ) C(CH -
2
4
2
2 2
2 5
2
ONO ) (5a), were obtained analogously (Scheme 2; for
2
2
the synthesis of the corresponding hydroxymethyl com-
pounds, (CH ) C(CH OH) and (CH ) C(CH OH) , see
2
4
2
2
2 5
2
2
ref 35). The nitrates 1a-5a were synthesized by treatment
of the corresponding alcohols with an excess of fumic
nitric acid and acetic anhydride at 0 ꢀC, and the reaction
mixtures were then stirred at ambient temperature, fol-
lowed by an aqueous workup. In the case of 3a, the
precursor MeC(CH OH) was treated with an excess of
2
3
fumic nitric acid at 0 ꢀC using dichloromethane as the
solvent, followed by stirring of the reaction mixture at
ambient temperature and subsequent aqueous workup.
Compounds 1a-5a were isolated in 82-88% yield with-
out any byproduct. Several other syntheses of these
3
6
compounds are known but in anticipation of the syn-
theses of their corresponding sila-analogues, this gentle
and non-oxidizing synthetic route was used.
For an alternative synthesis of the (nitratomethyl)-
silanes 1b-3b, the corresponding (iodomethyl)silanes
were used as precursors. Me SiCH I was commercially
For the synthesis of the (nitratomethyl)silanes 1b-5b,
the corresponding (hydroxymethyl)silanes were synthe-
sized as precursors by using the method described in
refs 12 and 15. The syntheses of the bis(hydroxymethyl)-
silanes 6-8 by using this particular synthetic pathway
were not reported before and therefore shall be briefly
described. Compounds 6-8 were synthesized according
to Scheme 3, starting from the respective bis(chloro-
3
2
available, and Me Si(CH I) was prepared from the
2
2 2
commercially available bis(chloromethyl)silane Me Si-
2
37,38
(CH Cl) .
in the literature.
MeSi(CH I) was synthesized as reported
2 3
2
2
15
The silicon-based nitratomethyl compounds 1b-5b
were synthesized analogously to their corresponding
carbon analogues 1a-5a by treatment of the correspond-
ing hydroxymethyl derivatives with an excess of fuming
1
6
methyl)silanes. Thus, treatment of Me Si(CH Cl) ,
2
2
2
(
CH ) Si(CH Cl) , and (CH ) Si(CH Cl) with sodium
2 4 2 2 2 5 2 2
acetate in N,N-dimethylformamide furnished the corre-
sponding bis(acetoxymethyl)silanes 9-11 (76-87% yield),
which upon methanolysis, in the presence of acetyl chloride
as a source for the formation of catalytic amounts of
hydrogen chloride, yielded the corresponding bis(hydroxy-
methyl)silanes 6-8 (60-73% yield).
7
nitric acid and acetic anhydride (Scheme 4). Compounds
1b-3b were additionally prepared by treatment of the
corresponding iodomethyl derivatives with an excess of
silver nitrate at 0 ꢀC using acetonitrile as the solvent,
followed by stirring of the reaction mixture at ambient
temperature and subsequent extraction with pentane
(
(
Scheme 5). In the case of 2b-5b, the first approach
nitration of the corresponding alcohols) was found to
(
35) Domin, D.; Benito-Garagorri, D.; Mereiter, K.; Fr o€ hlich, J.; Kirchner,
K. Organometallics 2005, 24, 3957–3965.
36) Agrawal, J. P.; Hodgson, R. D. Organic Chemistry of Explosives;
John Wiley & Sons: Chichester, 2007, pp 87-123.
(
(37) Roberts, J. D.; Dev, S. J. Am. Chem. Soc. 1951, 73, 1879–1880.
(38) Vivet, B.; Cavelier, F.; Martinez, J. Eur. J. Org. Chem. 2000, 807–811.

Article
Inorganic Chemistry, Vol. 49, No. 11, 2010 4871
Scheme 4. Synthesis of the Silicon-Based Nitratomethyl Compounds
resonances are low-field shifted in both series of com-
pounds. The influence of the alkyl substituents (methyl,
butane-1,4-diyl, and pentane-1,5-diyl) on the proton reso-
1b-5b, Starting from the Corresponding Hydroxymethyl Derivatives
nances of the ElCH O groups of the bis(nitratomethyl)
2
derivatives 2a, 2b, 4a, 4b, 5a, and 5b is small. In both cases
(
1
El = C or Si), the ElCH O H resonances are low-field
2
shifted for the cyclic derivatives (4a, 4b, 5a, and 5b)
compared to the acyclic dimethyl compounds 2a and 2b.
1
3
The C NMR resonances of the C/Si pairs 1a/1b-5a/
b are more influenced by the number of the nitrato-
5
methyl groups than the H resonances. With increasing
1
number of the nitratomethyl moieties, the CH and CH
2
3
resonances exhibit a high-field shift of 3-4 ppm. In
contrast, the central R-carbon atom shows a low-field
shift when increasing the number of nitratomethyl
groups. The ring size of the cyclic carbon compounds 4a
1
3
and 5a exerts only a small influence on the C NMR
resonances of the CH O groups. The ring CH groups of
2
2
Scheme 5. Synthesis of the Silicon-Based Nitratomethyl Compounds
b-3b, Starting from the Corresponding Iodomethyl Derivatives
13
a and 5a are easier to distinguish in the C than in the H
1
4
1
1
3
NMR spectra. The C resonances of the central carbon
atom (R-atom) are strongly affected by the alkyl substit-
uents (methyl, butane-1,4-diyl, and pentane-1,5-diyl).
The exchange of the central carbon by a silicon atom
1
3
shifts the C NMR resonances to higher field, with the
same shifting tendencies as observed in the series of the
carbon compounds 1a-5a. Because of hyperconjugation
3
9
13
and β-silyl effect, the C NMR resonances of the SiCH
groups are high-field shifted by 0 to -10 ppm.
The N NMR resonances of the nitrato groups of the
3
1
4
be the best synthetic method. In contrast, for the synthesis
of 1b the second method turned out to be more advanta-
geous. Compounds 1b-5b were isolated in 80-96%
yield.
carbon compounds 1a-5a are observed in a region be-
1
4
tween -41 and -45 ppm. The N resonances of the
corresponding sila-analogues 1b-5b are shifted to lower
field in a region between -34 and -41 ppm. In both series
of compounds, high-field shifting is observed with increas-
ing number of the nitratomethyl groups. The ring size of
the C/Si pairs 4a/4b and 5a/5b has only a weak influence on
NMR Analyses. The carbon-based nitratomethyl com-
pounds 1a-5a and their corresponding sila-analogues
1
b-5b show quite similar trends in the chemical shifts
of their H NMR resonances. Compared to the acyclic
1
1
4
carbon compounds 1a-3a, their sila-analogues 1b-3b
are high-field shifted in the H NMR spectra. Especially
the N chemical shift, but a high-field shift of 3-4 ppm is
observed for each additional nitratomethyl moiety.
1
29
the resonances of the methyl protons are strongly influ-
enced by the respective R-atoms, the central carbon or
The acyclic silicon compounds 1b-3b show Si NMR
resonances in a region between 0.1 and -5.4 ppm, and
high-field shifting is observed with increasing the number
silicon atom. The β-CH proton resonances are strongly
2
2
9
high-field shifted for the silicon compounds 1b-3b com-
pared to those of their corresponding carbon analogues
of the nitratomethyl groups. The Si resonances of the
silicon compounds 2b, 4b, and 5b are strongly dependent
on the nature of the alkyl substituent (2b, -1.2 ppm; 4b,
13.7 ppm; 5b, -6.5 ppm).
1
a-3a. The resonances of the hydrogen atoms of the ring
systems of the cyclic carbon compounds 4a and 5a differ
from those of the sila-analogues 4b and 5b, respectively.
For 5a only one broad H resonance signal was observed,
Physical Properties. The boiling, melting, and decom-
position points of the C/Si pairs 1a/1b-5a/5b were meas-
ured by differential scanning calorimetry (DSC) (Table 1).
The acyclic carbon compounds 1a-3a have boiling points
in the temperature range 174-183 ꢀC, but 2a and 3a already
start to decompose at 160 ꢀC. For the cyclic carbon
compounds 4a and 5a, no boiling points could be deter-
mined because they start to decompose at 190 ꢀC. For 2a
and 4a, unexpectedly high melting points were observed
(2a, ∼ 20 ꢀC; 4a, 23 ꢀC). Compound 3a has a melting point
1
because of overlapping of the different CH resonances of
2
the ring system, whereas in the case of 5b three clearly
defined multiplet structures were observed. In detail, the
1
δ-CH2 H resonances of 5b are only slightly influenced
compared to those of 5a, whereas the β-CH resonances
2
are strongly high-field shifted and the γ-CH resonances
2
low-field shifted compared to those of 5a. In the case of
1
the C/Si analogues 4a and 4b, the β-CH2 H resonances
2
are not affected by the R-atom (C or Si). However, as
of ∼ -15 ꢀC, and 5a becomes highly viscous upon cooling,
in the case of 5a and 5b, the β-CH protons of 4b
but does not solidify at -18 ꢀC.
2
show strongly high-field shifted resonances compared
to those of the carbon analogue 4a. The H resonances
The silicon compounds 1b-5b decompose violently in
the temperature range 85-107 ꢀC, which is 80-95 ꢀC
1
of the exocyclic CH groups are quite similar for the cyclic
2
C/Si pairs 4a/4b and 5a/5b, that is, they are only slightly
influenced by the R-atom (C or Si). With increasing
the number of the nitratomethyl groups, the proton
(
39) Schraml, J. In The Chemistry of Organic Silicon Compounds;
Rappoport, Z., Apeloig, Y., Eds.; John Wiley & Sons: Chichester, 2001;
Vol. 3, pp 223-339.

4
872 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
lower in temperature compared to the corresponding
carbon analogues 1a-5a (Tables 1 and 2). Interestingly,
the decomposition temperatures of the cyclic silicon
compounds 4b and 5b are slightly lower than those for
the acyclic derivatives 1b-3b. This is in contrast to the
corresponding carbon analogues 1a-5a, which show
slightly higher decomposition points for the cyclic com-
pounds 4a and 5a compared to their acyclic derivatives
1
a-3a. The highest decomposition points of the silicon-
based nitratomethyl compounds 1b-5b were observed
for 2b and 3b (ca. 107 ꢀC). The boiling points of 1b-5b
could not be determined because of prior explosive
decomposition. The silicon compounds 1b-5b have
lower melting points (2b, ∼ -10 ꢀC; 3b, -18 ꢀC; 4b,
-
18 ꢀC; 5b, ∼ 4 ꢀC) than the corresponding carbon
analogues, except for 5b with a melting point of about
4
ꢀC. Only 5b crystallized at 4 ꢀC in a refrigerator, whereas
all the other compounds solidified at -18 ꢀC in the freezer,
except for compound 1b, which remained liquid. The silicon-
based nitratomethyl compounds can be stored at -18 ꢀC
(
1b-5b) and 4 ꢀC (4b, 5b), respectively, without decomposi-
tion but decompose slowly at ambient temperature.
The decomposition of 1b in dry trichloromethane at
2
9
1
2
5 ꢀC, -4 ꢀC, and -18 ꢀC was studied by Si{ H} NMR
spectroscopy as a function of time (Figure 1). The ali-
quots of 1b were stored at 4 ꢀC and -18 ꢀC as pure
substance (in air), and NMR samples were freshly pre-
pared in dry trichloromethane after a period of time. The
NMR spectra at ambient temperature were collected
from one NMR sample prepared in dry trichloro-
methane, which was repeatedly measured after a period
of time. After 10 days of storing 1b at ambient tempera-
ture and air, about 20% (and after 60 days about 50%)
was decomposed. The main decomposition product was
3
9
unambiguously identified as hexamethyldisiloxane, the
formation of which is not understood at this time. At 4 ꢀC
after 24 days, about 20% of 1b was decomposed to give
hexamethyldisiloxane, but no significant decomposition
was observed upon storing at -18 ꢀC over a period of
2
4 days. The NMR samples were taken from pure sub-
stances for the series at 4 ꢀC and -18 ꢀC.
Friction and Impact Sensitivities. The experiments con-
cerning the friction and impact sensitivities of the C/Si
pairs 1a/1b-5a/5b were performed according to BAM
Figure 1. Decomposition of 1b (*) at various temperatures (25 ꢀC, 4 ꢀC,
(
Bundesanstalt f u€ r Materialforschung and -pr u€ fung) stan-
3
and -18 ꢀC) and as a function of time in dry CDCl as monitored by
2
9 1
Si{ H} NMR spectroscopy.
1
9
dards. The parent carbon compounds 1a-5a were
found to be much less sensitive toward impact and
friction compared to their corresponding sila-analogues
tests was observed, which would explain the higher sensi-
tivity of the bis(nitratomethyl) compound 2a compared to
the tris(nitratomethyl) derivative 3a.
1b-5b (Table 3). Even the lowest sensitivity (impact
and friction) observed for the silicon compounds (1b) is
much higher than the highest sensitivity of the discussed
carbon analogues (3a). The impact sensitivities of the
silicon compounds are in the range of the critical value of
measurability of the BAM setup (< 0.5 J).
The number of the nitratomethyl groups is the most
important factor influencing the friction sensitivity of the
compounds studied. The cyclic silicon compounds 4b and
Crystal Structure Analyses. The nitratomethyl com-
2
pounds Me
(CH C(CH
and the silicon-containing precursors (CH
(7) and (CH Si(CH OH) (8) were structurally characteri-
C(CH
ONO
) (4a), and (CH
2
) (2a), Me
Si(CH
ONO
ONO
Si(CH
)
(2b),
) (5b)
OH)
2
2
2
2
2
2
2
)
2
ONO
2
) Si(CH
5
4
2
2
2
)
2
4
2
2
)
2
5
2
2
zed by single-crystal X-ray diffraction. The crystallographic
data for 2a, 2b, 4a, and 5b are given in Table 4. The
molecular structures of these compounds in the crystal are
depicted in Figures 2-5; selected bond lengths and angles
are listed in Tables 5 and 6. For the crystal structure analyses
of the precursors 7 and 8, see the Supporting Information.
The carbon-based nitratomethyl compound 2a crystal-
5
b are less sensitive toward friction than the corresponding
acyclic derivatives 1b-3b, with 4b (five-membered ring)
showing a higher sensitivity toward friction than 5b (six-
membered ring). This might be due to the larger ring strain
in the case of 4b. Compounds 2a and 4a were measured as
liquids, but spontaneous crystallization during the friction
lizes in the monoclinic space group P2 /c, with four
1

Article
Inorganic Chemistry, Vol. 49, No. 11, 2010 4873
Table 3. Friction and Impact Sensitivities Measured for 1a-5a and 1b-5b (BAM Testing)
compound
1a
2a
3a
4a
5a
1b
2b
3b
4b
5b
a
friction [N]
impact [J]
>360
>100
>96
>100
>108
>15
>108
>100
>120
>100
>64
>1
<5
<0.5
<5
<0.5
<5
<0.5
>36
<0.5
b
a
b
Measuring range, 5 N < x < 360 N. Measuring range, 0.5 J < x < 100 J.
Table 4. Crystal Data and Details of the Structure Determinations of 2a, 2b, 4a, and 5b
2
a
2b
4a
5b
formula
M
C
5
H
10
N
2
O
6
C
4
H
10
N
2
O
6
Si
C
7
H
12
N
2
O
6
7 14 2 6
C H N O Si
r
194.15
100(2)
0.71073
monoclinic
P2 /c
0.15 ꢀ 0.18 ꢀ 0.03
7.5819(5)
10.5551(6)
11.5215(7)
106.530(7)
883.93(9)
4
1.459
0.135
408
7.90-52.00
-9 e h e 9
-
-
4902
210.23
140(2)
0.71073
monoclinic
P2 /c
0.1 ꢀ 0.1 ꢀ 0.02
7.6671(3)
11.0136(4)
12.0510(5)
105.407(5)
981.04(7)
4
1.423
0.243
440
7.68-52.00
-9 e h e 9
-13e k e 16
-14 e l e 14
9722
1936 [Rint = 0.0200]
128
1.180
220.19
100(2)
0.71073
monoclinic
P2 /n
0.21 ꢀ 0.17 ꢀ 0.06
7.0159(6)
9.6605(9)
14.7399(9)
91.061(6)
998.86(14)
4
1.464
0.129
464
7.66-52.98
-8 e h e 8
-11 e k e 11
-18 e l e 18
6995
1952 [Rint = 0.0379]
184
0.859
250.29
200(2)
0.71073
T/K
˚
λ/A
crystal system
monoclinic
P2 /c
space group
crystal size/mm
1
1
1
1
0.2 ꢀ 0.2 ꢀ 0.2
10.4017(9)
6.7723(6)
16.6297(14)
94.726(8)
1167.47(17)
4
1.424
0.217
528
7.50-52.00
-12 e h e 12
-8 e k e 8
-14 e l e 20
5355
2254 [Rint = 0.0955]
145
0.990
˚
a/A
˚
b/A
˚
c/A
β/deg
3
˚
V/A
Z
-3
F
calc./g/cm
μ/mm
F(000)
-
1
2θ range/deg
index ranges
13 e k e 12
12 e l e 14
reflections collected
reflections unique
parameters
1726 [Rint = 0.0414]
118
0.959
GoF
R
R
1
/wR
/wR
2
2
[I>2σ(I)]
(all data)
0.0411/0.0756
0.0786/0.0853
þ0.201/-0.172
0.0301/0.0825
0.0406/0.0887
þ0.326/-0.172
0.0329/0.0678
0.0711/0.0751
þ0.151/-0.169
0.0588/0.0814
0.1686/0.1141
þ0.315/-0.291
1
-3
˚
max/min residual electron density/e A
Figure 3. Molecular structure of 2b in the crystal showing the atom
labeling scheme. Thermal ellipsoids are shown at the 50% probability level.
Figure 2. Molecular structure of 2a in the crystal showing the atom
labeling scheme. Thermal ellipsoids are shown at the 50% probability level.
4
1
˚
C-O single bond (1.43 A), with bond lengths of
approximately 1.46 A (2a) and 1.45 A (4a), respectively,
molecules in the unit cell, a calculated density of 1.459 g/
˚
˚
3
cm , and a cell volume of 883.93(9) A . As also observed
3
˚
but almost identical with the CH -O bond lengths ob-
2
for the cyclic derivative 4a (monoclinic, space group P2 /
1
40
3
n, four molecules in the unit cell, density 1.464 g/cm ), the
served for PETN. The structure of the NO group is
2
quite similar to that observed for PETN and gaseous
nitrogen dioxide. The lengths of the N-O bonds in this
moiety are between those of a classic N-O single and
double bond, in good agreement with PETN and gaseous
bond lengths and angles of the C-CH -ONO unit are in
2
2
4
0
good agreement with the data obtained for PETN. Only
the CH -O bonds are slightly longer than a “normal”
2
4
0,41
nitrogen dioxide.
The silicon-based nitratomethyl compound 2b crystal-
(
(
40) Trotter, J. Acta Crystallogr. 1963, 16, 698–699.
41) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry, 5th ed.;
John Wiley & Sons: New York; 1988, pp 321-323.
lizes in the monoclinic space group P2 /c, with four
1

4
874 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
ments this is not a valid comparison. Alternatives are the
use of two-bond radii of Bartell or, to be preferred, the
4
4
4
one-angle radii of Glidewell. The distances to be con-
5
˚
sidered for 2b [2.672(1)/2.690(1) A] and 5b [2.679(3)/
˚
.723(3) A] are surprisingly close to the sum of Glidewell’s
2
values for silicon (1.55 A) and oxygen atoms (1.13 A) at
˚
˚
˚
.68 A. This indicates an absence of pronounced attractive
2
Si O interactions. Such attractive interactions between
3
3 3
geminal silicon (acceptor) and donor atoms (O, N) are
known to have significant or even dominating effects on
molecular structures. Examples are the structures of SiON
4
6
(
(
e.g., F SiONMe ) and SiNN compounds (e.g., F SiN-
3 2 3
4
7
Me)NMe ) ; this effect has been coined the R-effect in
2
silicon chemistry. However, in simple systems containing
SiCO units (e.g., F SiCH OMe), which are more closely
4
8
3
2
related to compounds 2b and 5b, such pronounced inter-
actions have not been found for the structures of the
molecular ground states. The very similar Si-C-O angles
of 2b and 5b [105.55(8)-108.6(2)ꢀ] and the analogous
C-O-C angles of 2a and 4a [106.55(14)-107.09(13)ꢀ]
do also not support the interpretation of our data in terms
of such kind of interactions.
Figure 4. Molecular structure of 4a in the crystal showing the atom
labeling scheme and thermal ellipsoids at the 50% probability level.
NBO analyses were performed at the B3LYP/cc-pVDZ
level of theory (Figures 6 and 7). The resulting data of the
C/Si analogues 2a and 2b show no significant evidence for
intramolecular C
3 3 O and Si 3 3 3 O interactions, respec-
3
tively.
The NBO charge distributions of compounds 2a and 2b
are remarkably different (see Figures 6 and 7). Specifically,
the NBO charge of the quaternary central carbon atom
of 2a shows a negative value of -0.100 e (e = elec-
tron; -0.058 e for C(CH ) ). In contrast, the central
3
4
silicon atom of 2b shows a positive value of 1.754 e
1.782 e for Si(CH ) ). The values for the CH carbon
(
3
4
3
atoms of compound 2a are clearly negative (-0.646 e),
Figure 5. Molecular structure of 5b in the crystal showing the atom
labeling scheme. Thermal ellipsoids are shown at the 50% probability level.
but the CH carbon atoms show only a slightly negative
2
charge (-0.090 e). The corresponding values of the silicon
analogue 2b show strongly negative charges at both the
3
molecules in the unit cell, a cell volume of 981.04(7) A ,
˚
3
and a density of 1.423 g/cm . The cyclic derivative 5b
CH
3
(-1.175 e) and the CH carbon atoms (-0.561 e).
2
The NBO charges of the oxygen and nitrogen atoms of the
nitrato moieties of 2a and 2b show almost no influence of
the carbon/silicon exchange. The same holds true for the
NBO charges of the hydrogen atoms of the C/Si ana-
logues 2a and 2b.
crystallizes in the monoclinic space group P2 /c, with four
1
3
molecules in the unit cell, a cell volume of 1167.47(17) A ,
˚
3
and a density of 1.424 g/cm .
The bond lengths of these compounds are similar to
those of their carbon analogues. The lengths of the
The enthalpies of formation (Δ Hꢀ) of the single mole-
f
CH -O bonds of the silicon compound 2b [1.4512(13)
2
cules 1a-5a and 1b-5b were calculated at the CBS-4 M
level of theory (Table 7). The enthalpy changes of the
˚
and 1.4495(15) A] are marginally shorter than the corre-
sponding bond lengths of the carbon analogue 2a
˚
isodesmic reactions (Δ Hꢀ) shown in Scheme 6 were
r
[
1.457(2) and 1.460(2) A]. Also, the O-NO bonds of
2
obtained by the sum of Δ Hꢀ of the products (1a-5a,
f
the silicon compounds are slightly longer than those of
the carbon analogues. In comparable moieties like the
nitratomethyl groups of the C/Si analogues 2a and 2b the
bond strength is inversely related to the bond length.
Si(CH ) ) minus the sum of Δ Hꢀ of the reactands
3
4
f
(
1b-5b, C(CH ) ).
3
4
Therefore, the O-NO bonds of the silicon compound
2
(44) Bartell, L. S. J. Chem. Phys. 1960, 32, 827–831.
(45) Glidewell, C. Inorg. Chim. Acta 1975, 12, 219–227.
(46) (a) Mitzel, N. W.; Losehand, U. Angew. Chem. 1997, 109, 2897–2899.
Angew. Chem., Int. Ed. Engl. 1997, 36, 2807-2809. (b) Losehand, U.; Mitzel,
N. W. Inorg. Chem. 1998, 37, 3175–3182. (c) Mitzel, N. W.; Losehand, U. J. Am.
Chem. Soc. 1998, 120, 7320–7327. (d) Mitzel, N. W.; Losehand, U.; Wu, A.;
Cremer, D.; Rankin, D. W. H. J. Am. Chem. Soc. 2000, 122, 4471–4482.
are probably slightly weaker than those of the carbon
analogue, but like the situation in methyl nitrate.
4
2
To identify whether there are attractive interactions
between the silicon atoms and the bridging oxygen atoms
of 2b and 5b, one could be tempted to compare the
distance between them with the sum of their van der
(
e) Mitzel, N. W.; Vojinovi ꢀc , K.; Fr €o hlich, R.; Foerster, T.; Rankin, D. W. H.
J. Am. Chem. Soc. 2005, 127, 13705–13713.
47) (a) Mitzel, N. W. Chem.;Eur. J. 1998, 4, 692–698. (b) Vojinovi ꢀc , K.;
4
3
Waals radii. However, in the case of geminal arrange-
(
McLachlan, L. J.; Hinchley, S. L.; Rankin, D. W. H.; Mitzel, N. W. Chem.;Eur.
J. 2004, 10, 3033–3042.
(48) Mitzel, N. W. Z. Naturforsch. 2003, 58b, 759–763.
(
(
42) Cox, A. P.; Waring, S. Trans. Faraday Soc. 1971, 67, 3441–3450.
43) Bondi, A. J. Phys. Chem. 1964, 68, 441–451.

Article
Inorganic Chemistry, Vol. 49, No. 11, 2010 4875
˚
Table 5. Bond Lengths (A) of 2a, 2b, 4a, and 5b in the Crystal
a (El = C)
2
2b (El = Si)
1.4021(13)
1.3927(14)
1.1878(15)/1.2035(14)
1.2154(17)/1.2005(17)
1.4512(13)
4a (El = C)
1.3977(16)
1.3900(16)
1.1948(18)/1.2001(18)
1.1938(18)/1.2022(18)
1.4541(18)
1.4535(18)
1.517(2)/1.514(2)
1.549(2)
5b (El = Si)
1.406(4)
1.407(4)
1.201(4)/1.205(4)
1.192(4)/1.210(4)
1.447(4)
1.446(4)
1.875(4)/1.883(4)
1.852(4)
N1-O1
N2-O2
N1-O3/4
N2-O5/6
C1-O1
C2-O2
El-C1/2
El-C3
1.3908(19)
1.3954(18)
1.2033(19)/1.2096(19)
1.211(2)/1.2106(19)
1.460(2)
1.457(2)
1.4495(15)
1.528(2)/1.523(2)
1.535(2)
1.537(3)
1.8860(12)/1.8892(12)
1.8489(14)
1.8500(13)
El-C4
El-C6
El-C7
1.550(2)
1.850(4)
1.532(5)
1.523(5)
1.513(5)
1.538(5)
C3-C4
C4-C5
C5-C6
C6-C7
1.524(2)
1.514(3)
1.514(2)
Table 6. Bond Angles (deg) of 2a, 2b, 4a, and 5b in the Crystal
a (El = C)
2
2b (El = Si)
4a (El = C)
5b (El = Si)
O3-N1-O4
O5-N2-O6
O1-N1-O3/4
O2-N2-O5/6
N1-O1-C1
N2-O2-C2
El-C1-O1
El-C2-O2
C1-El-C2
C1-El-C3
C1-El-C4
C1-El-C6
C1-El-C7
C2-El-C3
C2-El-C4
C2-El-C6
C2-El-C7
C3-El-C4
C3-El-C6
C3-El-C7
El-C3-C4
C3-C4-C5
C4-C5-C6
C5-C6-C7
El-C6-C5
El-C7-C6
129.01(17)
129.29(16)
118.29(16)/112.70 (16)
118.04(17)/112.67(16)
114.16(13)
113.92(13)
106.55(14)
106.61(14)
112.02(14)
128.75(12)
129.32(13)
118.37(11)/112.88(11)
112.38(12)/118.30(12)
114.40(9)
114.27(10)
106.69(8)
105.55(8)
108.68(5)
129.24(15)
128.66(15)
118.25(14)/112.51(15)
118.42(15)/112.93(15)
113.55(12)
113.63(13)
106.68(13)
107.09(13)
112.37(12)
107.89(14)
129.2(4)
128.7(4)
118.1(4)/112.6(4)
118.6(4)/112.7(3)
114.5(3)
114.1(3)
108.6(2)
106.6(2)
103.77(17)
108.09(18)
106.75(15)
110.16(15)
109.50(6)
106.60(6)
112.50(13)
112.09(13)
107.30(14)
111.34(18)
113.92(18)
110.73(15)
106.73(15)
107.37(6)
109.83(6)
112.62(18)
110.49(15)
114.74(7)
104.47(13)
107.07(17)
111.2(3)
113.6(3)
115.0(3)
114.4(3)
105.84(15)
103.47(16)
102.34(15)
104.94(14)
110.6(3)
Figure 6. NBO chargesof2a, calculated at the B3LYP/cc-pVDZlevelof
theory. All given values are in e (e = electron). The atom labels are shown
in parentheses (see also Figure 2).
Figure 7. NBO chargesof2b, calculatedatthe B3LYP/cc-pVDZ levelof
theory. All given values are in e (e = electron). The atom labels are shown
in parentheses (see also Figure 3).
As can be seen from Table 7, all isodesmic reactions are
exothermic (Δ Hꢀ < 0). In detail, with increasing number
of nitratomethyl groups, the value of xΔ Hꢀ increases.
r
When comparing 1a with 2a and 1a with 3a, a doubling
r

4
876 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
Table 7. Enthalpies of Formation (Δ
to 1a-5a (Scheme 6)
f
3 4 3 4 r
Hꢀ, kcal/mol) for 1a-5a, 1b-5b, C(CH ) , and Si(CH ) and Enthalpy Changes of the Isodesmic Reactions (Δ Hꢀ, kcal/mol) of 1b-5b
reaction
Δ
f
Hꢀ (1b-5b)
Δ
f
Hꢀ (C(CH
3
)
4
)
Δ
f
Hꢀ (1a-5a)
Δ
f
Hꢀ (Si(CH
3
)
4
)
r
Δ Hꢀ
1b f 1a
2b f 2a
3b f 3a
4b f 4a
5b f 5a
-456720.0
-632040.4
-807358.9
-680527.0
-705155.8
-123857.1
-123857.1
-123857.1
-123857.1
-123857.1
-299182.9
-474507.9
-649830.6
-523000.7
-547624.0
-281398.6
-281398.6
-281398.6
-281398.6
-281398.6
-4.4
-9.0
-13.1
-15.2
-9.8
Scheme 6. Isodesmic Reactions of 1b-5b with C(CH
Si(CH
3
)
4
to 1a-5a and
This distance is slightly shorter than the sum of the van
4
der Waals radii (3.04 A). The O1 O4 distance of 2b is
3
0
3
)
4
˚
3
3 3
˚
slightly longer than that of 2a [3.066(1) A] and is also
slightly longer than the sum of the van der Waals radii.
For both compounds, a second intermolecular contact is
0
found between the oxygen atoms O2 and O4 , with
˚
˚
O2 O4 distances of 2.862(2) A (2a) and 2.921(1) A
3 3
3
(
2b), respectively, which are somewhat shorter than the
˚
4
3
sum of the van der Waals radii (3.04 A). In addition, a
third weak interaction between the nitrogen atom N2 and
0
the oxygen atom O4 was found. In the case of 2a, this
˚
distance is 2.921(1) A which is slightly shorter than the
4
sum of the van der Waals radii (3.07 A). In the case of
3
˚
0
˚
b, the N2 O4 distance amounts to 3.123(2) A.
2
3
3 3
Gas Electron Diffraction Studies. The molecular struc-
and a tripling, respectively, of the Δ Hꢀ value is observed.
tures of the C/Si analogues 1a and 1b were determined in
the gas phase by gas electron diffraction (GED), aided by
ab initio calculations using various theoretical models.
The gas-phase structures of 1a and 1b are shown in
Figures 9 and 10.
r
Comparison of the acyclic molecule 2a with the related
cyclic molecule 5a (six-membered ring) reveals similar
Δ Hꢀ values (-9.054 and -9.753 kcal/mol). In contrast,
r
for the cyclic derivative 4a (five-membered ring) a sig-
nificantly higher Δ Hꢀ value (-15.209 kcal/mol) is ob-
r
GED Molecular Models. On the basis of the ab initio
calculations, the molecular models for 1a and 1b were
constructed using overall C symmetry. In addition to this
served. A possible explanation for this phenomenon is the
increased ring strain in the case of 4a.
The intramolecular Si O distances between the sili-
s
assumption of molecular symmetry, local C3v symmetry
was assumed for both the ElMe group (El = C, Si) and
3
3 3
con atoms and the bridging oxygen atoms observed for 2b
˚
and 5b are 2.672(1) and 2.723(3) A, respectively, and this
3
each individual methyl group. Finally, on the basis of the
relatively small scattering intensity for hydrogen, the
hydrogen atoms in the CH bridge were positioned with
is the expected range for geminal silicon and oxygen
atoms. The proximity of these atoms makes it plausible
that the first step in the decomposition reaction is the
highly exothermic formation of an Si-O bond and could
promote the initial step for a chain reaction. Calculations
for the decomposition pathway of Si(CH ONO ) con-
2
equal El-C-H (El = C, Si) and O-C-H angles, while
the C-H distance in the CH bridge was assumed to be
2
equal to that of the methyl groups. Using this combina-
tion of assumptions, the total number of internal coordi-
nates for each compound was reduced from 54 to 14, as
shown in Tables 8 and 9. The atom numbering schemes
used for the molecular models, corresponding to the
descriptions in Tables 8 and 9, are shown in Figures 9
and 10. The only difference between the models for the
two compounds, other than in the descriptions due to the
exchange of carbon for silicon, was that the C-O and
N-O bond lengths were described by an average and
difference for 1a, but were included as independent para-
meters for 1b.
2
2 4
firm this view that the initial step of this process is the
formation of an Si-O bond between the bridging oxygen
atom and the central silicon atom. This is in contrast to
8
the decomposition pathway known for common nitrate
ester explosives, where the homolytic O-NO bond
2
3
cleavage is the initial step. However, it is in line with
the finding of very shallow Si-C-N bending potentials
observed for (aminomethyl)silanes, which can also be
seen as contribution to the increased reactivity of such
4
9
compounds at silicon.
A closer look at the crystal structures of 2a and 2b
provides a potential explanation for the relatively high
melting points of these compounds. Six intermolecular
interactions were found for both compounds, namely,
GED Refinements. The experimental molecular-inten-
sity and radial-distribution curves for 1a and 1b are
shown in Figures 11-14, with the refined difference
curves shown at the bottom of each figure. The refine-
ments were performed along the lines of the SARACEN
0
0
those between the O1-N1-O3/4 and the O1 -N1 -
0
0
50
O3 /4 moieties. The resulting dimers and interactions
are shown in Figure 8. The nitrato moieties of the
molecular pairs are facing each other. In the case of 2a,
method, which places flexible restraints on parameters
that are not well resolved from the GED experiment, with
0
˚
the shortest distance is the O1 O4 contact [2.984(2) A].
3 3
3
(50) (a) Blake, A. J.; Brain, P. T.; McNab, H.; Miller, J.; Morrison, C. A.;
Parsons, S.; Rankin, D. W. H.; Robertson, H. E.; Smart, B. A. J. Phys.
Chem. 1996, 100, 12280–12287. (b) Mitzel, N. W.; Rankin, D. W. H. Dalton
Trans. 2003, 3650–3662.
(
49) Mitzel, N. W.; Vojinovi ꢀc , K.; Foerster, T.; Robertson, H. E.;
Borisenko, K. B.; Rankin, D. W. H. Chem.;Eur. J. 2005, 11, 5114–5125.

Article
Inorganic Chemistry, Vol. 49, No. 11, 2010 4877
0
0
0
Figure 8. Aggregationmotifsofpairs ofmolecules in the crystalsof2a(left) and 2b (right). The contactsO1 O4 , O2 O4 , and N2 O4 are shownas
3
3 3
3 3 3
3 3 3
dashed lines.
Figures 11 and 13) were grouped together with fixed
relative amplitudes. Restraints were also applied to all
refined amplitudes, and for 1b restraint uncertainties of
1
1
0% of the calculated values were applied, while those for
˚
a were set to 10% for distances less than 4.5 A and 20%
˚
for distances greater than 4.5 A. The full lists of intera-
tomic distances, amplitudes of vibration, and distance
corrections for the rh1 refinements, including details of
which amplitudes were tied together, are provided as
Supporting Information.
The refinement of 1a yielded a good fit of the experi-
mental to theoretical intensities for both the r and r
Figure 9. Molecular structure of 1a in the gas phase showing the atom
labeling scheme used for the GED model.
g
h1
structure types, as can be seen from the low R factors (RG)
of 4.7% and 5.2%, respectively. In the case of 1b, the
relatively low vapor pressure at room temperature re-
quired longer exposure times and a higher beam current
than have been found to be optimum, and the data are
correspondingly noisier than for 1a and in previous GED
structure investigations using this combination of appa-
2
6,51
ratus and refinement techniques.
factors of 16.4% and 14.9% for the respective r and r
The resulting R
g
h1
structure types are higher than might be expected; how-
ever, the quality of the fit can also be judged by the
appearance of the molecular-intensity and radial-distribu-
tion curves (Figures 13 and 14). In particular, the residuals
of the molecular-intensity curves can be seen to be ran-
domly distributed, indicating that the higher R factor is
due to a lower signal-to-noise ratio rather than a refine-
ment problem. Curiously, the standard deviations for the
refined parameter values of 1a and 1b (Tables 8 and 9) are
similar to one another, but from the preceding discussion
it would be expected that the uncertainties for 1b should be
larger than those for 1a. However, comparison of the
RDCs (Figures 12 and 14) reveals that the structure of 1b
may be marginally better resolved by GED: While there
are nine distinct peaks in the RDC for both compounds,
the RDC for 1a exhibits only one additional weak
Figure 10. Molecular structure of 1b in the gas phase showing the atom
labeling scheme used for the GED model.
the value of the restraint being taken from theory and the
uncertainty estimated by the convergence of the calcula-
tions (a strict interpretation of the SARACEN method
would require the full number of independent parameters
to be included in the model, but in this case we have made
use of the assumptions mentioned above). All 14 inde-
pendent parameters for both 1a and 1b were therefore
refined as shown in Tables 8 and 9. Seven of these 14
parameters were refined unrestrained for 1a and eight in
the case of 1b, the difference being that the C-O and
N-O distances for 1b were refined independently,
whereas for 1a they were included as an average and a
difference, the latter of which was restrained. Amplitudes
of vibration were also refined for both molecules, but
those corresponding to distances under a single peak
in the respective radial-distribution curve (RDC, see
˚
shoulder at about 3 A, but the RDC for 1b also contains
˚
two large shoulders at about 1.4 and 2.7 A, as well as a
˚
small shoulder at 3.4 A. Thus, the lower signal-to-noise
ratio observed for 1b is expected to be counterbalanced by
(51) Hagemann, M.; Berger, R. J. F.; Hayes, S. A.; Stammler, H.-G.;
Mitzel, N. W. Chem.;Eur. J. 2008, 14, 11027–11038.

4
878 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
Table 8. Details of the Independent Parameters Used in the GED Refinement of 1a, Refined Parameter Values, Calculated Values Obtained at the RI-MP2/TZVPP Level of
a
Theory, and Applied Restraints
parameter description
GED r
g
GED rh1
MP2 r
e
restraint uncertainty
p
p
p
p
p
p
p
p
p
p
p
p
p
p
1
r C-C average
1.539(1)
1.430(2)
0.040(4)
1.211(1)
1.117(2)
109.2(2)
106.3(3)
113.6(3)
132.0(6)
111.0(3)
0.002(4)
4.4(2)
1.538(1)
1.429(2)
0.035(4)
1.210(1)
1.114(2)
108.7(2)
106.9(3)
113.2(3)
128.7(6)
110.8(1)
0.003(4)
3.7(2)
1.525
1.422
0.033
1.208
1.090
109.9
106.9
113.2
130.0
108.3
0.005
2.5
2
r C-O, N-O3 average
r C-O, N-O3 difference
3
0.010
0.010
b
4
r N-O ꢀ2 mean
5
r C-H
6
— C(Me)-C-C(Me)
— C-C-O
7
8
— C-O-N
— O5-N-O8
— H-C-H
9
10
11
12
13
14
1.0
c
r C-C difference
0.005
1.0
1.0
d
3
CMe tilt
e
— O3-N-O difference
4.7(5)
0.008(4)
4.7%
3.4(5)
0.004(4)
5.2%
4.4
0.006
f
r N-O ꢀ2 difference
0.005
G
R factor (R )
a
˚
All distances are in A and angles are in deg. The two sets of parameter values for the GED refinement, r
to accounting for vibrational motion. The r structure is a vibrational average, while the rh1 structure is an approximation to an equilibrium structure, r
g
and rh1, correspond to different approaches
g
e
.
b
c
d
Average of N-O5 and N-O8. C-C(Me) minus C-C18. Defined as a decrease in the C1-C2-C18 angle with respect to that required for C3v
symmetry. O3-N-O5 minus O3-N-O8. N-O5 minus N-O8.
e
f
Table 9. Details of the Independent Parameters Used in the GED Refinement of 1b, Refined Parameter Values, Calculated Values Obtained at the RI-MP2/TZVPP Level of
a
Theory, and Applied Restraints
parameter description
GED r
g
GED rh1
MP2 r
e
restraint uncertainty
p
p
p
p
p
p
p
p
p
p
p
p
p
p
1
r Si-C mean
r C-O
1.883(1)
1.447(5)
1.426(4)
1.208(1)
1.104(3)
110.9(4)
107.1(3)
113.0(4)
131.6(7)
109.9(4)
0.033(3)
2.8(6)
1.882(1)
1.437(4)
1.435(3)
1.206(1)
1.103(3)
110.7(4)
108.3(3)
113.6(4)
131.3(7)
109.5(3)
0.033(3)
1.6(5)
1.879
1.433
1.425
1.206
1.090
111.2
105.4
113.4
130.6
108.0
0.034
1.3
2
3
r N-O3
b
4
r N-O ꢀ2 mean
5
r C-H
0.010
6
— C(Me)-Si-C(Me)
— Si-C-O
7
8
— C-O-N
9
— O5-N-O8
— H-C-H
10
11
12
13
14
1.0
c
r Si-C difference
0.005
1.0
1.0
d
3
SiMe tilt
e
— O3-N-O difference
5.9(5)
0.006(3)
16.4%
5.2(5)
0.005(3)
14.9%
4.6
0.003
f
r N-O ꢀ2 difference
0.005
G
R factor (R )
a
˚
All distances are in A and angles are in deg. The two sets of parameter values for the GED refinement, r
to accounting for vibrational motion. The r structure is a vibrational average, while the rh1 structure is an approximation to an equilibrium structure, r
g
and rh1, correspond to different approaches
g
e
.
b
c
d
Average of N-O5 and N-O8. Si-C18 minus Si-C(Me). Defined as a decrease in the C1-Si-C18 angle with respect to that required for C3v
symmetry. O3-N-O5 minus O3-N-O8. N-O5 minus N-O8.
e
f
an increase in the structural information inherent in its
electron diffraction pattern, compared to 1a.
Discussion of the Gas-Phase Geometries. A selection of
important geometric parameters for 1a and 1b are listed in
Table 10, showing the convergence of the calculated
structures, alongside the experimental GED structures,
the latter being determined using both the r and the r
g
h1
5
2
models for vibrational motion. Comparison of the
calculated and experimental values indicates that the
Me ElCH groups (El=C, Si) are reasonably well des-
3
2
cribed by HF theory, with only small changes in the
values of these parameters when improving the theoreti-
cal treatment by applying MP2 theory. In contrast, HF
(
52) An internuclear distance obtained directly from the GED data, r
the harmonic mean and can be converted to the arithmetic mean, r , using the
þ u /r. Distance
a
, is
g
2
root-mean-squared amplitude of vibration, u: r
g
≈ r
a
corrections, k, are also regularly applied in GED refinements to account
for the ‘shrinkage’ effect. When the vibrational motion is considered to be
harmonic with curvilinear trajectories, the resulting distances and distance
corrections are termed rh1 and kh1, respectively.
Figure 11. Experimental (O), theoretical, and difference (experimental
minus theoretical) molecular-intensity curves for 1a.

Article
Inorganic Chemistry, Vol. 49, No. 11, 2010 4879
Figure 14. Experimental and difference (experimental minus theoretical)
radial-distribution curves for 1b. Prior to Fourier transformation, the
2
molecular-intensity data were multiplied by s exp[-0.00003s /(ZSi - fSi
O
)
3
Figure 12. Experimental and difference (experimental minus theoretical)
radial-distribution curves for 1a. Prior to Fourier transformation, the
˚
O
- f )] and the s-range was extended to 0 and to 36 A by model data.
-1
(
Z
2
molecular-intensity data were multiplied by s exp[-0.00002s /(Z
N
N
- f )-
3
-
1
)] and the s-range was extended to 0 and to 36 A by model data.
˚
(
Z
O
- f
O
Table 10. Selected Structural Parameters for 1a and 1b as Determined Experi-
a
mentally by GED and Calculated by Theory
b
c
c
HF/6-31G*
HF/
MP2/
GED
method
r
e
r
e
r
e
r
g
r
h1
Compound 1a
r C-H
1.085
1.536
1.530
1.439
1.327
1.187
1.179
107.4
116.5
127.9
1.083
1.533
1.528
1.435
1.324
1.179
1.170
107.9
116.7
127.8
1.090
1.526
1.521
1.438
1.405
1.211
1.205
106.9
113.2
130.0
1.117(2) 1.111(3)
1.540(1) 1.539(1)
1.538(3) 1.535(3)
1.450(3) 1.441(3)
1.410(2) 1.408(2)
1.215(2) 1.213(2)
1.207(2) 1.208(2)
106.3(3) 106.9(3)
113.6(3) 114.0(3)
132.0(6) 129.7(7)
r C-C(Me)
r C-C18
r C-O
r N-O3
r N-O5
r N-O8
—
—
—
C-C-O
C-O-N
O5-N-O8
Compound 1b
r C-H
1.086
1.886
1.919
1.439
1.333
1.185
1.179
106.6
116.9
128.0
1.084
1.877
1.913
1.435
1.330
1.177
1.171
107.4
117.1
128.0
1.090
1.870
1.905
1.433
1.425
1.207
1.204
105.4
113.4
130.6
1.104(3) 1.103(3)
1.875(1) 1.874(1)
1.908(3) 1.907(3)
1.447(5) 1.437(4)
1.426(4) 1.435(3)
1.211(2) 1.208(2)
1.205(2) 1.204(2)
107.1(3) 108.3(3)
113.0(4) 113.6(4)
131.6(7) 131.3(7)
r Si-C(Me)
r Si-C18
r C-O
Figure 13. Experimental (O), theoretical, and difference (experimental
minus theoretical) molecular-intensity curves for 1b.
r N-O3
r N-O5
r N-O8
theory appears to have severe limitations with regard to
the treatment of the nitrato group. The terminal N-O
bonds are significantly elongated by applying MP2 the-
ory, by more than 3 pm when the geometries calculated
with identical basis sets are compared. The largest dis-
crepancy between the HF and MP2 geometries of 1a and
—
—
Si-C-O
C-O-N
— O5-N-O8
a
˚
All distances r are in A, angles — are in deg, and the numbers in
parentheses are one standard deviation in the last digit. For an explana-
b
tion of the structure types, r , r , and r , see ref 52. 6-31G** was used
e
g
h1
c
for 1a. def2-TZVPP.
1
b is in the length of the N-O3 bond, which is lengthened
˚
˚
by 0.081 A (1a) and 0.095 A (1b), respectively. Such a
large difference between the HF and MP2 method would
normally cast doubt on the accuracy of the MP2 geome-
try; however, the N-O bond lengths determined by GED
agree very well with those determined by the MP2 method
˚
(1a) and 0.010 A (1b), respectively. There is also a good
agreement for most of the angles, with discrepancies
mostly less than 1ꢀ, the exceptions being the O5-N-O8
angle for 1a and the Si-C-O angle for 1b, which there-
fore should be treated with some caution.
With regard to the increased friction and impact sensi-
tivity of the silicon compound 1b compared to its carbon
analogue 1a, the interesting parameters are the relative
lengths of the N-O3 bonds and the Si-C-O and
C-C-O angles. Both the experimental and calculated
structures indicate a slight weakening of the N-O3 bond
for the silicon compound 1b, with an increase of about
(although the differences between the N-O5 and N-O8
distances were restrained to the MP2 values).
The reliability of the GED results can be assessed by the
agreement between the two models (r and r ) used to
g
h1
describe the vibrational motion, as well as the standard
deviations from the least-squares fit. The agreement
between the two models is reasonably good for all bond
˚
distances, the largest discrepancies for r
being 0.009 A
C-O

4
880 Inorganic Chemistry, Vol. 49, No. 11, 2010
Evangelisti et al.
˚
.02 A for r
0
. The calculated structures of 1a and 1b
N-O3
strated to be a powerful tool for the development of novel
17
18
also indicate a small (ca. 1ꢀ) contraction of the Si-C-O
angle (1b) compared to the C-C-O angle (1a), but the
experimentally established structures indicate a differ-
ence in the opposite direction. This parameter was not
particularly well resolved in the GED structure of 1b, as it
was strongly dependent on the vibrational model. The
observed discrepancy may be due to different dynamic
behavior for 1a and 1b rather than a deficiency in the
theoretical model for describing the static equilibrium
geometries. In any case, the energy difference correspond-
ing to such a small angle contraction would be minimal
drugs and odorants with unique properties. The studies
presented here also clearly demonstrate the high potential of
the sila-substitution concept (C/Si exchange) for the develop-
ment of new silicon-based explosives (in this context see also
refs 7 and 8), with properties that differ significantly from
those of their corresponding carbon analogues.
Acknowledgment. Financial support of this work by the
Ludwig Maximilians University of M u€ nchen (LMU), the
Fonds der Chemischen Industrie (FCI), the European
Research Office (ERO) of the U.S. Army Research Labora-
tory (ARL), and the Armament Research, Development
and Engineering Center (ARDEC) under contract nos.
W911NF-09-2-0018, W911NF-09-1-0120, and W011NF-
and the elongation of the O-NO bond in 1b is the only
2
significant structural difference that might account for its
greater instability.
09-1-0056 is gratefully acknowledged. The authors ac-
Conclusion
knowledge collaborations with Dr. Mila Krupka (OZM
Research, Czech Republic) in the development of new
testing and evaluation methods for energetic materials
and with Dr. Muhamed Sucesca (Brodarski Institute,
Croatia) in the development of new computational codes
to predict the detonation and propulsion parameters of
novel explosives. We are indebted to and thank Drs. Betsy
M. Rice and Brad Forch (ARL, Aberdeen, Proving
Ground, MD) and Dr. Gary Chen (ARDEC, Picatinny
Arsenal, NJ) for many helpful and inspiring discussions
and support of our work. We thank Prof. Dr. Andreas
Kornath and Can D o€ rtbudak, M. Sc., for their support
in measuring the Raman spectra. Special thanks to Dr.
Matthias Scherr, Karin Lux, M. Sc., Dipl.-Chem. Franz
Martin, and Dr. Christian Burschka for collecting the
X-ray data sets and Stefan Huber for measuring the
impact and friction sensitivities.
The silicon-based nitratomethyl compounds Me Si(CH -
2
2
ONO ) (2b), MeSi(CH ONO ) (3b), (CH ) Si(CH ONO )
2 2
2
2
2
2 3
2 4
2
(4b), and (CH ) Si(CH ONO ) ) (5b) were synthesized for the
2 5 2 2 2
first time. In addition, the known derivative Me SiCH ONO
2
3
2
(
1b) was resynthesized. The silicon compounds 1b-5b were
found to be much more sensitive toward friction and im-
pact than their mostly known carbon analogues Me CCH -
3
2
ONO (1a), Me C(CH ONO ) (2a), MeC(CH ONO ) (3a),
2
2
2
2 2
2
2 3
(CH ) C(CH ONO ) (4a, new), and (CH ) C(CH ONO )
2 4 2 2 2 2 5 2 2 2
(
5a). The thermal stabilities of the silicon compounds 1b-5b
are approximately 80-95 ꢀC lower than those of the corre-
sponding carbon analogues 1a-5a. Weak O-NO bonds,
2
strong C-O bonds, and the weaker Si-CH O bond of the
2
silicon compounds are claimed to be responsible for the higher
sensitivities compared to their corresponding carbon ana-
logues. GED studies and calculations at different levels
of theory established the gas phase structures of 1a and 1b
in good agreement with the structural parameters obtained in
single-crystal X-ray diffraction studies of 2a, 2b, 4a, and 5b.
The carbon/silicon switch strategy has already been demon-
Supporting Information Available: Gas electron diffraction
data for 1a and 1b and crystallographic data for 2a, 2b, 4a, 5b, 7,
and 8. This material is available free of charge via the Internet at
http://pubs.acs.org.