Vibrational spectra and electronic structure of germatranols (HO) 4 - N Ge(OCH2CH2)nNR3 - N (R = H; N = 1-3) with transannular Gea?N bonding

Article abstract of DOI:10.1016/j.jorganchem.2012.12.036

Series of germatrane compounds with OH functionality and Gea?N transannular bonding, i.e. 1-germatranol (I), 1,1-quasigermatranol (II) and 1,1,1-hypogermatranol (III) with general formula (HO)_{4 - n}Ge(OCH ₂CH₂)_nNR_{3 - n} (R = H, Me; n = 1-3) are synthetized and their solid state IR spectra are recorded. Equilibrium structures and vibrational spectra of monomeric and centrosymmetric dimeric species of these compounds are predicted by the DFT B3LYP/aug-cc-pVDZ method. The assignment of the vibrational spectra of I-III was carried out based on theoretical IR spectra scaled with factors obtained earlier for halogermatranes to take account of the substantial and systematic discrepancy between Gea?N interatomic distance of the gas phase and theoretical predictions on the one hand and X-ray solid state data on the other. Dimeric models provide a good agreement between experimental and predicted frequencies, including GeOH bending modes, of I, crystalline state of which is formed by dimers, and slightly worse for II, in which some weaker O-Ha?O and N-Ha?O hydrogen bonds are not considered in the dimeric model. However, this model for III is poor since it does not take into account hydrogen bonding of free OH and NH₂ groups. Despite the steady decrease of the Gea?N interatomic distance, corresponding frequencies demonstrate the substantial increase on going from I to II and practically no increase from II to III.

Full text of DOI:10.1016/j.jorganchem.2012.12.036

Journal of Organometallic Chemistry 747 (2013) 62e68
Contents lists available at SciVerse ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Vibrational spectra and electronic structure of germatranols (HO)4
ꢀ n
Ge(OCH CH ) NR
(R ¼ H; n ¼ 1e3) with transannular Ge/N
2
2 n
3 ꢀ n
bonding
I.S. Ignatyev a,*, G.S. Samokhin , T.A. Kochina , V.V. Belyaeva , S.Ya. Khaikin , M. Montejo ,
b
b
c
d
e
e
b
J.J. López González , M.G. Voronkov
Department of Chemistry, Radiochemistry Laboratory, St. Petersburg State University, 199034 St. Petersburg, Russia
Grebenshchikov Institute of Silicate Chemistry RAS, 199034 St. Petersburg, Russia
Favorsky Institute of Chemistry SB RAS, Irkutsk, Russia
Boreskov Institute of Catalysis (St. Petersburg Branch), SB RAS, 199034 St. Petersburg, Russia
Department of Physical and Analytical Chemistry, University of Jaén, Campus “Las Lagunillas”, E-23071 Jaén, Spain
a
b
c
d
e
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 20 November 2012
Received in revised form
Series of germatrane compounds with OH functionality and Ge/N transannular bonding, i.e. 1-
germatranol (I), 1,1-quasigermatranol (II) and 1,1,1-hypogermatranol (III) with general formula (HO)
Ge(OCH CH NR
(R ¼ H, Me; n ¼ 1e3) are synthetized and their solid state IR spectra are
4
ꢀ
n
2
2
)
n
3 ꢀ n
18 December 2012
recorded. Equilibrium structures and vibrational spectra of monomeric and centrosymmetric dimeric
species of these compounds are predicted by the DFT B3LYP/aug-cc-pVDZ method. The assignment of the
vibrational spectra of IeIII was carried out based on theoretical IR spectra scaled with factors obtained
earlier for halogermatranes to take account of the substantial and systematic discrepancy between
Ge/N interatomic distance of the gas phase and theoretical predictions on the one hand and X-ray solid
state data on the other. Dimeric models provide a good agreement between experimental and predicted
frequencies, including GeOH bending modes, of I, crystalline state of which is formed by dimers, and
slightly worse for II, in which some weaker OeH/O and NeH/O hydrogen bonds are not considered in
the dimeric model.
Accepted 19 December 2012
Keywords:
Germatranols
Molecular structure
Electronic structure
Vibrational spectrum
DFT
However, this model for III is poor since it does not take into account hydrogen bonding of “free” OH
and NH groups. Despite the steady decrease of the Ge/N interatomic distance, corresponding fre-
2
quencies demonstrate the substantial increase on going from I to II and practically no increase from II to
III.
Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction
A special interest is involved in derivatives of germatranes con-
taining hydroxyl groups [15e21] due to the ability of hydroxyl
Silatranes and germatranes, that are the molecules with trans-
annular M/N bonding of the general formula XM(OCH CH
CH N five-
membered cycles were subjects of numerous investigations
mainly due to their biological applications [1e4]. Recently, there
appeared publications concerning metallatrane analogs containing
only two cycles, i.e. ocanes [5e12], and just a few investigations
groups toform avarietyofhydrogen bonds with other protondonors
and Lewis acids. The crystal structure of anhydrous 1-germatranol
[HOGe(OCH CH ) N, I] was firstly reported by Voronkov et al. [21].
2 2 3
It consists of centrosymmetrical dimers of germatranol molecules
linked by two hydrogen bonds between axial GeOH groups and
2
2 3
) N
(
X ¼ H, Alk, Hal; M ¼ Si, Ge) containing three MOCH
2
2
equatorial oxygen atoms of the GeOCH
The building block of the crystal structure of the analogous
compound with two hydroxyl groups, i.e. germocane (HO) Ge(OCH
CH NH (II), has also a dimeric structure [22]. However, this mole-
2 2
CH N cycle.
[13,14] were devoted to hypogermatranes containing only one
2
2
MOCH CH N cycle.
2
2
2 2
)
cule has a diol functionality and in contrast to I forms dimers by two
OH groups specific to silanediol dimers [23e25]. These dimers have
free hydroxyl groups which may be exploited to link dimers to form
ribbon-like networks as in other silanediol structures [23,24]. As
*
Corresponding author. Tel.: þ7 812 328 95 95.
E-mail address: ignatiev@ujaen.es (I.S. Ignatyev).
0
022-328X/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jorganchem.2012.12.036

I.S. Ignatyev et al. / Journal of Organometallic Chemistry 747 (2013) 62e68
63
concernstriol analogs of thisclass ofmolecules, i.e. hypogermatranes
containing only one GeOCH CH N cycle (III), to the best of our
2
2
knowledge no data on the structure of this compound exist.
Here we report IR spectra of compounds IeIII and optimized
quantum chemical structures and predicted vibrational spectra of
these molecules in monomeric and dimeric forms. Theoretical re-
sults are used for the interpretation of experimental spectra.
2
. Experimental and theoretical details
-Hydroxy-5-aza-2,8,9-oxa-1-germatricycloundecane (1-germa-
1
tranol, I) 1,1-dihydroxy-5-aza-2,8-dioxa-1-germabicyclooctane (1,1-
quasigermatranol, II) 1,1,1-trihydroxy-5-aza-2-octa-1-germacyclo
pentane (1,1,1-hypogermatranol, III) are obtained from GeO
corresponding 2-hydroxyethylamines (HOC
(n ¼ 0e2) in
:1 ratio inwater media (90 C) without catalysts [26]. IR spectra were
recordedwithaBrukerTENSOR27spectrometer(splitmulltechnique
27]) and a Varian 3100 FT-IR spectrometer (KBr pellets). These are
2
and
2 4
H
)
3 ꢀ nNH
n
ꢁ
1
[
shown in Figs. 1S, 2S and 3S of supplementary material.
The geometries of all of the molecules have been fully optimized
and theoretical vibrational spectra have been calculated by the
B3LYP [28,29] hybrid density functional method with the Dunning
correlation consistent aug-cc-pVDZ basis set [30,31] as they are
implemented in the Gaussian09 [32] suite of programs.
Scaling of the force field was done according to the SQM method
[33,34] using the Molvib program [35,36]. For this, the force field
matrix expressed in Cartesian coordinates was transferred from the
Gaussian output to Molvib, and transformed to natural internal
coordinates.
3
. Results
Equilibrium structures of monomeric germatranols (I, II, and III)
studied are presented in Fig. 1 and corresponding dimeric struc-
tures in Fig. 2. The existing X-ray structural data for I and II are
compared with our predicted geometry parameters in Tables 1
and 3. In I the dimeric model describes all major intermolecular
interactions, however, in dimeric structures II and especially III, the
presence of “free” (not involved in hydrogen bonding) hydroxyl
groups limits the predictive power of models considered.
The replacement of oxygen atoms of OCH
CH N cycles by more
2 2
electronegative OH groups results in the steady increase of the
positive charge on germanium and shortening of Ge/N trans-
2 2
annular bonds. Geometry parameters of OCH CH N cycles practi-
cally do change on going from I to III. Despite this, IR spectra in the
ꢀ
1
1
000e100 cm region differ substantially due to strong mixing of
cyclic modes among themselves and with deformation modes of
GeOH groups. However, theoretical vibrational spectra allowed us
to make an attempt to make a systematic assignment of the
vibrational spectra of germatranols. Observed frequencies below
Fig. 1. Optimized molecular structures and main geometrical parameters (bond dis-
tances in A) of monomeric 1-germatranol (a), 1,1-quasigermatranol (b) and 1,1,1-
hypogermatranol (c).
ꢀ
ꢀ1
1500 cm (Tables 2, 4 and 5) are compared with theoretical har-
monic vibrations. Higher frequencies which include the stretching
of the OH and CH bonds are not listed since the values of predicted
harmonic frequencies are far from experimental values due to the
high anharmonicity of these vibrations. Moreover, samples contain
water and its OH stretching bands also blur this region.
One of the most challenging problems in the assignment of
vibrational spectra of atranes is the location of the M/N stretching
frequencies and attempts to link the values of these frequencies
with the strength of the transannular bond [41e45]. However,
quantum chemical methods (DFT and MP2) systematically over-
estimate M/N interatomic distances. This serious disagreement
between theoretical M/N distances and those obtained from the
solid state diffraction experiment is inherent in all atranes. As was
shown by Dillen [40] the only way to remove this discrepancy is to
use computational models simulating the crystalline state and
including longerange interactions. Due to this difference the pre-
dicted frequencies of the vibrations involving M/N stretching are
substantially lower than experimental solid state IR frequencies.
Ignatyev and Sundius [37] tried to compensate this underestima-
tion of theoretical vibrational frequencies by scaling the corre-
sponding theoretical force constants. Scaling of force constants by
the SQM method [33,34] allows us to minimize also some other
discrepancies between experimental and predicted frequencies,
e.g. CO and MO stretching frequencies. Note, that the assignment
ꢀ
1
of vibrations, especially in the spectral range below 600 cm is
rather ambiguous due to the mixture of modes of OCH
2 2
CH N
cycles.

64
I.S. Ignatyev et al. / Journal of Organometallic Chemistry 747 (2013) 62e68
Table 2
Experimental and theoretical vibrational spectra of HOGe(OCH
2
CH
quencies in cm , IR intensity in km mol ) compared with the scaled theoretical
spectrum of FGe(OCH CH N.
2 3
) N, [I] (fre-
ꢀ
1
ꢀ1
2
2 3
)
IR (unscaled)
n
Assignment
n
(scaled)b 1-F-germatrane
(scaled)b Symm
Monomer Dimera
n
1
482 m 1504 (0)
1503 (0)
d
d
d
d
d
d
d
d
d
d
d
d
r
r
r
r
r
r
r
r
d
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
2
2
2
2
2
2
2
2
2
2
2
2
1484
1480
1477
1459
1458
1457
1376
1375
1361
1353
1343
1342
1275
1273
1266
1241
1240
1229
1164
1162
1484
1479
A
E
1
1
500 (2)
497 (2)
1499 (5)
1498 (6)
1480 (17)
1479 (8)
1479 (7)
1397 (6)
1395 (7)
1457 m 1479 (5)
1460
1459
A
E
1
1
478 (6)
477 (4)
1
1
397 w 1395 (1)
394 (3)
371 w 1378 (12) 1380 (29)
1375
E
1
1362
1352
1341
A
A
E
1
1
1
372 (2)
363 (1)
361 (1)
1373 (11)
1363 (2)
1362 (1)
1272 m 1296 (13) 1298 (29)
2
2
2
2
2
2
2
2
1271
E
1
1
1
1
1
294 (13) 1296 (25)
282 (9)
257 (1)
256 (1)
246 (0)
1282 (20)
1260 (1)
1257 (2)
1248 (1)
1186 (4)
1184 (5)
1133 (289)
1266
1239
A
E
1229
1161
A
E
1
1
162 m 1183 (2)
1181 (3)
102 s
GeOH
dimer)
(
1
129 (135) 1122 (152)
n
n
n
r
OC
OC
OC
CH
1098
1062
1062
1075
1043
1042
1018
972
1097
1057
A
E
1
065 m 1102 (76) 1101 (159)
1
1
101 (72) 1095 (133)
086 (22) 1086 (58)
2
1075
1039
A
E
1
1
042 m 1065 (10) 1066 (23)
063 (9) 1064 (17)
028 m 1025 (60) 1025 (23)
58 (116)
nas NC
nas NC
n CC
3
3
1
1019
925
A
E
9
d
GeOH
(monomer)
9
35 m 932 (15)
31 (17)
06 m 898 (38)
933 (34)
932 (31)
898 (99)
877 (3)
877 (4)
756 (2)
n
n
r
r
r
CC
CC
CH
CH
CH
NC
GeO(H)
GeOH
926
925
897
865
863
748
690
9
9
2
2
2
898
864
A
E
8
8
75,4)
75 (2)
7
35 m 754 (0)
n
n
s
s
3
750
A
Fig. 2. Optimized molecular structures and main geometrical parameters (bond dis-
ꢀ
678 s
686 (150)
715 (272)
638 (235)
tances in A) of the hydrogen bonded dimers of 1-germatranol (a), 1,1-quasigermatranol
636 s
(
b) and 1,1,1-hypogermatranol (c).
(dimer)
5
5
87 (26)
81 (37)
619 (197)
600 (22)
587 (17)
582 (82)
571 (133)
539 (7)
n
as GeO
as GeO
3
654
642
610
600
597
559
454
670
601
E
E
n
3
3.1. 1-Germatranol
6
13 s
621 (92)
13 (74)
594 m 588 (29)
d
d
d
NCC
NCC
NCC
GeO
6
The equilibrium structure of the 1-germatranol [HOGe(OCH
2
587
551
451
A
A
A
5
4
69 m 538 (7)
21w 427 (1)
n
d
s
3
2 3
CH ) N, I] molecule is depicted in Fig. 1a and main geometry pa-
425 (2)
s
NC
Ge/N)
3
rameters obtained in this work are compared with experimental
data and the results of previous theoretical predictions in Table 1.
(n
4
4
21 (2)
18 (1)
422 (1)
419 (2)
318 (19)
313 (12)
297 (0)
276 (27)
258 (34)
200 (94)
235 (17)
208 (5)
156 (7)
143 (3)
97 (12)
d
d
d
d
d
d
d
n
d
d
r
r
d
s
as NC
as NC
GeOC
GeOC
3
3
418
415
317
314
305
290
280
250
223
214
168
147
105
74
418
315
E
E
3
3
3
2
2
32 sh 317 (14)
20 sh 313 (7)
10 m 294 (0)
76 m 257 (14)
60 m 254 (20)
212 (28)
210(20)
182 sh 203 (2)
Table 1
GeO
OCC
OCC
4
313
287
A
E
ꢀ
Experimental and theoretical geometry parameters (A,degree) of 1-germatranol
2 2 3
[HOGe(OCH CH ) N, I].
Experimentb
242 m
Ge/N
GeO4
GeO4
271
216
A
E
Geometry
M05ꢀ2x/
B3LYP/aug-cc-pVDZ
parametersa
6-11 þ G(d,p)
b
Monomer
Dimer
dimer
1
68 m 168 (32)
GeO
GeO
4
4
144
111
E
Ge/N
2.142 (3)
1.779 (3)
1.806 (3)
1.797 (3)
1.787 (3)
1.425(5)
1.529(3)
1.478(2)
178.8(1)
2.381
1.757
1.811
1.787
1.793
2.338
1.791
1.831
1.821
1.815
1.414
1.534
1.472
178.8
2.373
1.777
1.843
1.822
1.818
1.415
1.534
1.471
177.7
1
9
7
43 (3)
2 (4)
2 (25)
GeeO1
GeeO3
GeeO7
GeeO10
OeC, av
CeC, av
NeC, av
GeOC
GeOH
A
(monomer)
a
Only IR active A
u
modes are presented, six inter-monomeric rotational and
ꢀ
1
translational modes with frequencies below 75 cm are not shown.
Scale factors from ref. 37: CO: 0.874, CC: 1.024, GeO: 1.167, GeO(H): 1.000,
GeOH: 1.000, CN: 0.948, GeOC: 1.000,
b
N(1)Ge(1)O(4)
s
d : 2.633, das: 0.971, r: 1.570, OCC: 0.973, HCH:
0.973, CCH: 0.970, NCC: 1.114, GeOC: 1.000, CNC: 0.914.
a
Notations in Figs. 1 and. 2.
Data from ref. [21].
b

I.S. Ignatyev et al. / Journal of Organometallic Chemistry 747 (2013) 62e68
65
Table 3
Table 4
ꢀ
Experimental and theoretical geometry parameters (A, degrees) of (HO)
2
Ge(OCH
2
Experimental and theoretical vibrational spectra of (HO)
(frequencies in cm , IR intensity in km mol ).
2
Ge(OCH
2
CH
2
)
2
NH, [II]
ꢀ1
ꢀ1
CH
2
)
2
NH [II].
Geometry
Experimentb
MP2/
B3LYP/aug-cc-pVDZ
IR (unscaled) Assignment
n
n
(scaled)b
parametersa
6-311þþG(d,p)
b
Monomer
Dimer
Monomer
1502 (6)
Dimera
monomer
1
484 m
1502 (13)
1495 (0)
1489 (14)
1480 (20)
1464 (2)
1391 (24)
1373 (3)
1366 (7)
1336 (2)
1304 (13)
1266 (2)
1255 (5)
1225 (2)
1167 (1)
d
d
d
d
d
d
d
d
d
r
r
r
r
r
d
n
n
n
r
n
d
d
d
n
s
n
r
r
s
n
n
CNH
1488
Ge/N6
GeeO1
GeeO3
GeeO7
GeeO10
O3eC4
O7eC8
C4eC5
2.1246 (5)
1.7994 (5)
1.8044 (5)
1.7969 (5)
1.7735 (5)
1.4288 (8)
1.4219 (9)
1.5169 (11)
1.5238 (10)
1.4789 (9)
1.4739 (9)
84.24 (2)
83.99 (2)
177.53 (2)
85.76 (2)
95.44 (2)
94.02 (3)
2.207
1.799
1.813
1.823
1.805
1.413
1.411
1.528
1.530
1.472
1.470
82.3
81.9
176.3
81.1
101.3
96.1
97.5
2.270
1.806
1.819
1.834
1.816
1.415
1.415
1.534
1.534
1.473
1.473
81.3
2.262
1.827
1.822
1.836
1.788
1.415
1.415
1.535
1.535
1.472
1.472
81.5
1
1
495 (0)
489 (8)
CH
CH
CH
CH
CH
CH
CH
CH
2
2
2
2
2
2
2
2
1474
1470
1463
1447
1375
1358
1351
1319
1290
1250
1241
1209
1151
1
1
1
462 w
452 m
366 w
1479 (11)
1462 (1)
1392 (10)
1
1
374 (2)
367 (3)
1
1
1
1
1
1
348 sh
279 m
262 m
251 w
228 w
151 m
1336 (1)
1306 (6)
1266 (1)
1256 (3)
1225 (1)
1167 (2)
C8eC9
N6eC5
N6eC9
CH
CH
CH
CH
CH
2
2
2
2
2
N6eGeeO3
N6eGeeO7
N6eGeeO1
N6eGeeO10
O1eGeeO3
O1eGeeO7
O1eGeeO10
O3eGeeO7
O3eGeeO10
80.9
80.9
173.7
81.3
169.3
83.3
1146 (197)
GeOHeq (dimer)
1
1
1
115 m
1134 (45)
1111 (95)
1091 (76)
1136 (86)
1109 (156)
1091 (154)
1070 (1)
CN
OC
OC
1106
1068
1055
1043
1032
974
96.5
95.2
097 s
064 vs
96.0
93.1
96.51 (2)
118.15 (3)
119.67 (3)
103.8
127.6
112.4
95.2
126.3
113.4
1
069 (1)
1033 (54)
75 (160)
CH
CC
2
116.9
114.7
1
033 vs
1032 (130)
9
GeOHeq (monomer)
GeOHax (dimer)
GeOHax (monomer)
CC
CNH
CN
a
Atom notations from Figs. 1 and 2.
Data from ref.[22].
1011 s
1004 (216)
b
960 (13)
934 (12)
959
934
916
874
859
833
933 m
933 (25)
923 (240)
890 (4)
877 (5)
832 (53)
928 (89)
892 w
882 sh
816 m
891 (4)
877 (5)
832 (18)
Although in contrast to halogermatrane molecules [37] there are no
more C symmetry axis in germatranol, the axial GeO1(H) bond
collinear to the Ge/N bond) is the shortest among four GeO bonds
atom notations in Fig. 1). However, its predicted bond length is in
CH
CH
2
2
3
(
(
790 (188)
GeOHeq (dimer)
GeO(H) eq
GeO(H) ax
699 vs
666 (142)
654 (105)
619 (119)
593 (24)
580 (3)
709 (155)
630 (400)
617 (170)
594 (39)
579 (7)
722
705
656
626
574
551
458
420
396
377
666 vs
637 vs
606 vs
560 m
a good agreement with the experimental value in contrast to hal-
osilatranes and halogermatranes where theoretical MeHal bond
distances differ substantially from experimental ones [37]. As
compared to the equilibrium structure of trimethylgermanol the
substitution of methyl groups by electronegative oxygen atoms
leads to the substantial shortening of the GeO1(H) bond from 1.830
n
d
d
as GeO
2
NCC
NCC
542 (19)
403 (10)
542 (77)
407 (33)
ns GeO2 (n Ge/N)
458 m
n
s
d
d
s
s
d
d
d
d
s
s
d
Ge/N
GeOHeq (monomer)
NC
4
20(75)
417 (18)
77 (3)
4
14 w
46 m
418 (23)
379 (15)
340 (116)
2
ꢀ
to 1.791 A (B3LYP/aug-cc-pVDZ). The reduction of bond length due
3
GeOC
to the substitution by electronegative elements is well known and
usually is attributed to negative hyperconjugation [38,39]. The
other effect which may influence the length of the axial GeO1(H)
bond is the formation of coaxial Ge/N bonding. In order to sepa-
rate these two effects we have predicted the structure of
3
GeOHax (dimer)
GeOHax (monomer)
GeOC
305 (35)
286 (17)
277 (13)
306
282
279
307 m
282 m
261 m
298 (40)
285 (33)
282 (87)
GeO
GeO
GeO
2
2
2
(H) (dimer)
(H) (monomer)
241 (46)
253
226
HOGe(OMe)
3
at the same theory level. In this molecule the equi-
248
18
239 (7)
214 (2)
GeOHeq (dimer)
GeOHeq (monomer)
ꢀ
librium bond distance is 1.780 A. Thus, small increase of the
GeO1(H) bond length in germatranol may be attributed to Ge/N
bonding. However, the different arrangement of oxygen atoms
around germanium in germatranol and trimethoxygermanol may
also contribute to this effect. The GeO1(H) bond length is in a good
agreement with X-ray experimental data (Table 1). This agreement
improves in the equilibrium structure of the dimer (Fig. 2a). Here
the length of the proton-donating GeO1 bond decreases, while that
of Ge-O3 (containing a proton-accepting oxygen atom) increases.
The length of the GeO7 and GeO10 bonds, not involved in inter-
dimeric hydrogen bonding, practically do not change.
220 (11)
2
GeO
dimer)
GeO
2
(n
Ge/N)
(
195 (20)
d
2
(n
Ge/N)
216
(monomer)
208
176 (1)
157 (4)
190 (47)
173 (8)
125 (6)
87 (3)
d
r
r
d
GeOC
172
151
105
77
181
GeO
GeO
4
106 (0)
4
80 (0)
GeOC
a
u
Only IR active A modes are presented, six intermonomeric rotational and
ꢀ
1
translational modes with frequencies below 75 cm are not shown.
b
Scale factors: CO: 0.874, CC: 1.024, GeO: 1.167, GeOH: 1.000, CN: 0.948, GeOC:
1
.000, GeN: 2.663 OGeO: 1.000, OCC: 0.973, HCH: 0.973, CCH: 0.973, NCC: 0.973,
The Ge/N theoretical interatomic distances in germatranols are
GeOC: 1.000, CNC: 0.914, CNH: 1.000, OGeOH: 0.800.
markedly longer than that obtained by X-ray diffraction [21]
ꢀ
(
Table 1). This difference (ca. 0.2 A) is of the same order as in
XM(OCH
2
CH
2
)
3
N (M ¼ Si, Ge; X ¼ F, Cl) compounds [37]. Note that
FGe(OCH CH ) N Note, that predicted vibrational frequencies of
2
2 3
our B3LYP values are shorter than corresponding M05 ꢀ 2x/6-
the XGe(OCH CH ) N skeleton for X ¼ OH are very close to those for
2
2 3
3
11 þ G(d,p) bond lengths of the previous work [21].
X ¼ F however, due to the absence of the C axis in the former
3
Experimental IR frequencies of 1-OH germatrane (I) are com-
pared with predicted frequencies in Table 2. They contain not only
theoretical harmonic vibrational frequencies, but also frequencies
molecule, degenerate vibrations of the latter molecule are split.
The main contribution of the Ge/N stretching is located in the
ꢀ
1
vibration with the predicted frequency of 212 cm (Table 2). In the
ꢀ
1
of HOGe(OCH
2
CH
2
)
3
N, obtained with scaled theoretical force field
dimeric model it lays slightly lower (200 cm ) in keeping with the
(
scale factors are transferred from ref. [37]) as well as those of
small increase of the Ge/N interatomic distance in the dimer

66
I.S. Ignatyev et al. / Journal of Organometallic Chemistry 747 (2013) 62e68
ꢀ
1
Table 5
assigned to the experimental 636 cm band, while the attribution
of the GeOH bending is not so straightforward. There are two strong
Experimental and theoretical vibrational spectra of (HO)
3
Ge(OCH
2
2 2
CH )NH , [III]
ꢀ
1
ꢀ1
(
frequencies in cm , IR intensity in km mol ).
ꢀ1
IR bands in the 1050e1100 cm frequency range. One of them is
the band assigned to the stretching of the OC bond (1065 cm
which frequency does not change from that of 1-F-germatrane [40].
Thus the 1102 cm band may be assigned to hydrogen bonded
ꢀ1
IR (unscaled) Assignment
n
n
(scaled)b
)
Monomer
Dimera
ꢀ1
1
1
1
1
622 m
1630 (34)
1493 (1)
1484 (5)
1392 (6)
365 (1)
1306 (3)
1265 (7)
1632 (64)
1494 (2)
1483 (8)
1391 (15)
1365 (3)
1306 (7)
d
d
d
d
d
r
r
r
d
n
n
d
d
d
d
d
d
n
r
s
n
s
n
n
n
d
s
s
n
d
s
s
d
s
d
s
d
d
s
r
r
d
NH2
1624
518 m
458 w
391 w
CH
CH
CH
CH
2
2
2
2
1474
1464
1392
1355
1312
1256
1191
d
GeOH (Table 2).
3.2. 1,1-Quasigermatranol
1
1
1
327 w
272 w
CH
CH
CH
2
2
2
1265 (13)
1194 (3)
2 2 2 2
The equilibrium structure of the ocane (HO) Ge(OCH CH ) NH
1
194 (2)
(II) molecule is depicted in Fig. 1b. The length of the axial GeO1(H)
bond slightly increases compared to I. Since bond lengths and NBO
charges demonstrate that there is no significant redistribution of
electron density on going from I to II, this small increase in the
GeO1(H) bond length may be assigned to a shorter Ge/N intera-
tomic distance in II. In the equilibrium structure of centrosym-
metric dimer of II (Fig. 2b) the Ge/N interatomic distance is
slightly shorter than in a monomer and is 0.137 A longer than that
found in the X-ray study [22]. Note, that this difference between
theoretical and experimental Ge/N distance in II (Table 3) is
smaller than in I (0.231 A, Table 1). The dimer of II is formed uti-
lizing two OH groups of each monomer: one hydroxyl group acts as
a proton donor (GeO10H, Fig. 2b) and the other (GeO1H) as an
acceptor. Note, that in the experimental structure GeO10 is the
shortest GeO bond. This cannot be rationalized within the mono-
mer model [22]. However, in the dimeric model it shortens sub-
stantially in accordance with the experimental data. In the crystal
structure [22] dimers are linked by the hydrogen atoms of OH
acceptor groups and NeH/O hydrogen bonds to form polymeric
chains of dimers. However, these types of hydrogen bonds are
substantially weaker than the bonding in a dimer and the vibra-
tional spectrum of the ocane crystal may be described in a first
approximation by the dimeric model. In the dimeric model axial
GeO(H) bonds in which the oxygen atom acts like a proton acceptor
in contrast to the real crystal do not take part in hydrogen bonding
between dimers and this may cause some discrepancies between
theoretical and experimental geometry parameters and the corre-
sponding vibrational frequencies.
1
1
145 w
070 m
1158 (171)
1118 (130)
1074 (47)
992 (372)
1014 (250)
980 (14)
GeOeqH (dimer)
1118 (64)
073 (18)
020 (250)
1010 (41)
83 (56)
79 (131)
OC
NC
GeOH free
CNH
CNH
1090
1060
1026
1013
999
979
979
926
903
1
1
1015 m
9
9
GeOeqH (monomer)
GeOaxH (dimer)
GeOaxH (monomer)
CC
978 sh
929 sh
889 sh
861 sh
782 s
972 (384)
ꢀ
926 (135)
906 (6)
875 (3)
906 (13)
875 (2)
CH
GeOeqH (dimer)
GeOeq (H) ( NH
NH GeOeq (H))
2
865
799 (232)
722 (224)
671 (94)
650 (80)
608 (279)
591 (108)
550 (51)
ꢀ
6
6
89 (138)
65(26)
50 (102)
632 (126)
94 (14)
550 (18)
57 (101)
440 (76)
273 (14)
412 (24)
s
2
)
737
703
689
649
622
536
455
440
474
345
6
2
(n
GeO (H) free
GeOax (H)
GeOeq (C)
NCC
5
87 m
60 sh
5
5
4
GeOHeq (monomer)
GeOH free
Ge/N
GeOC (n Ge/N)
GeOeqH (dimer)
GeOaxH (dimer)
491 m
453 m
358 w
346 w
317 w
279 w
440 (175)
257 (36)
413 (90)
333 (50)
300 (75)
289 (62)
GeO
GeOeqH (monomer)
GeO
GeOaxH (monomer)
4
(dimer)
2
2
2
92 (9)
75 (16)
52 (44)
248 (8)
32 (15)
212 (28)
53 (4)
25 (0)
291
266
254
237
228
196
153
128
67
272 (20)
243 (9)
4
2
47 w
27 w
GeO
GeO
4
2
4
(monomer)
GeOaxH)
2
230 (2)
182 (8)
140 (11)
91 (7)
NH
2
(
s
1
1
GeO
GeO
4
4
7
1 (1)
GeOC
The predicted vibrational spectra of monomers and dimers of II
are compared with IR experimental data in Table 4. In addition to
unscaled IR spectra of the monomer and dimer (only IR active A
u
modes are shown) the scaled spectrum of the monomer is pre-
sented. Scale factors are those used for I (transferred from ref. [37])
a
Only IR active A
u
modes are presented, six intermonomeric rotational and
ꢀ
1
translational modes with frequencies below 75 cm are not shown.
b
Scale factors: CO: 0.874, CC: 1.024, GeO: 1.167, GeOH: 1.00, CN: 0.948, GeOC:
1.00, GeN: 2.663, OGeO: 1.00, OCC: 0.973, HCH: 0.973, CCH: 0.973, NCC: 0.973,
GeOC: 1.00, CNC: 0.914, CNH: 1.120, HNH: 0.950, OGeOH: 0.800.
with one exception. Due to the different symmetry of the GeO
fragment (C3v in I and C2v in II) the definition of the bending co-
ordinates of the GeO group was changed. Therefore, the axial ds
coordinates used in the non-redundant set of vibrational co-
ordinates for the reproduction of Ge/N vibrations cannot be
employed for II. However, the introduction of the redundant GeN
4
(
Table 1). The scaling of the theoretical force field of the monomer
4
ꢀ1
with factors from ref. [37] brings this frequency to 250 cm which
ꢀ
1
is in good agreement with the observed 242 cm band (Table 2).
Note that, in FGe(OCH CH N, the scaled theoretical frequency of
the analogous vibration is 271 cm (experimental Raman band at
2
2 3
)
ꢀ
1
coordinate with scale factor 2.633 used for the
non-redundant set of ref. 26 and putting the scale factor for
to unity, gives practically the same vibrational frequencies.
The main contribution of the GeN bond stretching is in a vibra-
d
s
coordinate in the
ꢀ
1
2
73 cm ). This difference is in agreement with the shorter equi-
d
s
equal
ꢀ
librium Ge/N bond length in F-germatrane (2.284 A [26] vs
2
ꢀ
.338 A).
There are also small differences in the frequencies of these vi-
ꢀ
1
ꢀ1
tion at 403 cm in a monomer (407 cm in the dimer). However,
scaling of the GeN stretching in a way described above gives
brations between the monomeric and dimeric models. However,
larger difference may be traced for vibrations with frequencies in
ꢀ
1
458 cm (Table 4). There exists a band in the experimental IR
spectrum with the same frequency which may be assigned to
ꢀ
1
6
20e570 cm region (assigned as
Certainly, the largest difference between frequencies of mono-
mers and dimers may be observed for the vibrations of the GeOH
group. In the 1-germatranol monomer GeOH and GeOH modes
have frequencies of 958 and 72 cm . In the dimeric model these
n
as GeO
3
and
d
NCC in Table 2).
n
Ge/N. All other frequencies of the cage vibrations in the dimer do
not differ substantially from those calculated for a monomer.
Large shifts of frequencies are observed for vibrations of the
GeOH groups. In addition to the above mentioned high anharmo-
nicity of CH and OH stretching modes, which prevents their com-
parison with theoretical vibrational frequencies, in II the NH
stretching band overlaps the OH stretching region. Therefore, in
d
s
ꢀ1
ꢀ1
frequencies rise up to 1133 and 638 cm due to the formation of
hydrogen bonds between GeOH groups. Both vibrations have high
theoretical IR intensities (Table 2). The
s GeOH mode may be

I.S. Ignatyev et al. / Journal of Organometallic Chemistry 747 (2013) 62e68
67
Table 4 (similar to Table 2) only the vibrational bands with fre-
coincides well with the predicted frequency of the dimer; however
most probably the frequency of this deformation may lie higher
due to the involvement of protons of these groups in inter-dimeric
interactions.
ꢀ
1
quencies below 1500 cm are presented. The predicted frequency
ꢀ
1
of the axial GeOH group (proton acceptor) deformation is 760 cm
with no nearby experimental band. However, in the dimer its value
ꢀ
1
is 1004 cm
1
and it may be assigned to a strong IR band at
011 cm . The dimeric frequency of the deformation of the
equatorial GeOH group (proton donor) may be assigned to the IR
Along with axial and “free” hydroxyl group protons hydrogen
atoms of NeH groups may take part in inter-dimeric bonding to
form polymeric chains. Therefore, the theoretical description of the
ꢀ1
ꢀ1
ꢀ1
band at 1151 cm which may overlap the low intensity
The predicted frequencies of GeOH, in the dimer are in a fair
agreement with their experimental counterparts at 816 and
r
CH
2
band.
700e450 cm frequency range in the dimeric model is rather poor.
ꢀ1
s
Only one strong band (491 cm ) with shoulders at 587, 560, and
ꢀ
1
453 cm is observed in this range. Theory predicts seven modes
with frequencies in this range. Three of them, that are GeOax
(GeO1), GeOeq (GeO3), and NCC, may be assigned to the shoul-
ders at 587 and 560 cm with fair agreement between theoretical
ꢀ1
3
46 cm . However, the assignment of the latter mode (
s
GeOHax) is
n
dubious, since its proton in the crystal is involved in inter-dimeric
hydrogen bonds and the absence of these interactions in the
dimeric model may cause significant errors in the predicted fre-
quency of the torsional mode.
n
d
ꢀ
1
ꢀ
1
and experimental frequencies while the main 491 cm
band
should be assigned to GeO7H (free) as theory predicts high IR
s
The substantial discrepancy exists between the frequency of the
CNH mode in the experimental and theoretical spectra (both
monomer and dimer models). It may be the result of the absence of
the NeH/O hydrogen bonding in these models. Similar to OeH/O
interactions included in the dimeric model, the introduction of
intensity of this mode. There is a large discrepancy between
experimental and theoretical frequencies for this case, but it may be
understandable due to the fact the dimeric model does not describe
their involvement in inter-dimeric hydrogen bonds. Two from three
d
2
remaining bands have a large contribution of NH deformations
these interactions may lead to the increase of the
d
CNH frequency.
which are inadequately described in the dimeric model and the
third belongs to the GeO stretching of “free” GeOH groups which do
not participate in hydrogen bonding in the dimeric model.
3.3. 1,1,1-Hypogermatranol
ꢀ1
The spectral features of III in the 1650e1200 cm region may
also indicate the involvement of NH group in hydrogen bonding.
There are two vibrations with the substantial contribution of the
The structure of the triol member of germatranols studied, i.e.
hypogermatranol containing only one GeOCH CH N cycle (III), as
2
2
2
ꢀ1
well as germanium and silicon analogs with other X substituents,
was not determined experimentally. The predicted structure of III
is depicted in Fig. 1c. Due to the presence of three hydroxyl groups
in the monomer there are different plausible structures of the
dimer. They were analyzed for the case of methylsilanetriol in ref.
Ge/N stretching with predicted frequencies of 273 and 412 cm
for the monomer. However scaling of the Ge/N stretching coor-
dinate by the factor 2.663 (employed for the scaling of theoretical
ꢀ
1
force fields of I and II gives frequencies equal to 474 and 345 cm
(Table 5)). They may be assigned to IR bands 453 and 358 cm
ꢀ
1
.
[
25]. From four structures which correspond to energy minima two
Thus, modes with the largest contribution of the Ge/N may be
ꢀ
1
ꢀ1
have maximum coordination energy, i.e. the one which utilize for
hydrogen bonding only two hydroxyl groups of each molecule, and
the other in which all three hydroxyl groups participate in intra-
dimeric bonding [25]. Two distinct types of the corresponding
hypogermatranol (III) dimeric forms with almost equal energy
were found at the B3LYP/6-31G* level but with the cc-pVDZ basis
both converge to the first type (with only two hydroxyl groups
involved in hydrogen bonding). This structure, presented in Fig. 2c,
is similar to the analogous dimeric form of II, but has completely
free (not involved in hydrogen bonding) hydroxyl groups. Thus,
similar to dimer II, Fig. 2b, molecule III may form multiple
hydrogen bonds not only with protons of hydroxyl groups which
oxygen atoms act as proton acceptors (like in II), but also with free
assigned in I to the 242 cm IR band, in II to the 458 cm band,
ꢀ1
and in III to the 453 cm band. Minor, but considerable con-
tribution of this coordinate is in vibrations that may assigned to IR
ꢀ1
ꢀ1
ꢀ1
bands with 421 cm (I), 560 cm (II), and 358 cm (III). Thus,
although the tendency of the increase of the Ge/N frequency on
n
going from I to II may be related with the shortening of the Ge/N
interatomic distance be traced, it is not straightforward, since these
vibrations are strongly mixed with other vibrations of the
n
Ge(OCH2CH2) N cage.
4. Conclusions
1. Solid state IR spectra of germatranol compounds
(HO)4 ꢀ nGe(OCH CH NR3 ꢀ n (R ¼ H; n ¼ 1e3) with a trans-
annular Ge/N bonding which contain different number of
Ge(OCH CH )N cycles were recorded. The assignment of the
(
non-bonded) hydroxyl groups. Thus, the solid state structure of III
2
2 n
)
may differ substantially from that of II. The comparison of pre-
dicted (monomeric and dimeric forms) and experimental IR
spectra of III (Table 5) and especially GeOH deformations may draw
some light on the plausible solid state structure, however these
conclusions may be quite tentative. Among GeOH bending modes,
the highest frequency in the dimeric model belongs to the defor-
mation of the GeOeqH group (GeO10 at Fig. 2c), which donates its
proton to the hydrogen bond. Its predicted frequency of 1158 cm
is in a good agreement with the IR 1145 cm band. In contrast to
this vibration, the frequency of which rises from 979 cm
a monomer to 1158 cm in the dimer, the second hydroxyl group
non-bonded in the dimer (GeO7H) practically does not change its
frequency. The experimental 1015 cm band may correspond to
the predicted value 992 cm , but most likely this “free” hydroxyl
group (GeO7H in the dimeric model) is involved in hydrogen
bonding in solid state and its frequency may lie above 1050 cm
The predicted frequencies of the axial GeOaxH group (GeO1-proton
acceptor) deformation are 926 cm (monomer) and 972 cm
(
2
2
observed bands was carried out with the help of theoretical
predicted vibrational frequencies obtained at the B3LYP/aug-cc-
pVDZ level of theory. Theoretical force fields scaled with factors
2 2 3
obtained earlier for the FGe(OCH CH ) N molecule were used to
take into account the systematic difference between the gas
phase and theoretical M/N distances (M ¼ Si, Ge) on the one
hand and X-ray solid state transannular bond lengths on the
other.
ꢀ
1
ꢀ1
ꢀ
1
in
ꢀ
1
2. Equilibrium structures of germatrane molecules (OH)
CH
annular Ge/N interatomic distance with n. Vibrations with
a greater contribution of the Ge/N stretching coordinate may
n 2
Ge(OCH
2
)
4 ꢀ nNHn ꢀ 1 (n ¼ 1e3) exhibit the reduction of a trans-
ꢀ
1
ꢀ
1
ꢀ1
be assigned to IR bands at 242 (I), 458 (II), and 453 (III) cm .
ꢀ
1
.
Thus, despite the steady decrease of the Ge/N interatomic
distance, corresponding frequencies demonstrate the sub-
stantial increase on going from I to II and a practically no
increase on going from II to III.
ꢀ1
ꢀ1
ꢀ1
ꢀ1
dimer). The 978 cm
shoulder of the 1015 cm
IR band

68
I.S. Ignatyev et al. / Journal of Organometallic Chemistry 747 (2013) 62e68
3
. The dimeric form of I revealed by the X-ray diffraction study is
an energy minimum on the potential energy surface of
[10] E.K. Lermontova, A.A. Selina, S.S. Karlov, A.V. Churakov, J.A.K. Howard, Yu.
F. Oprunenko, M.Yu. Antipin, J. Sundermeyer, G.S. Zaitseva, J. Organomet.
Chem. 691 (2006) 5710e5724.
2 2 3 2
[HOGe(OCH CH ) ] . Ongoing from monomer to dimer the
[
[
11] S.S. Karlov, E.Kh. Yakubova, E.V. Gauchenova, A.A. Selina, A.V. Churakov,
J.A.K. Howard, D.A. Tyurin, J. Lorberth, G.S. Zaitseva, Z. Naturforsch. 58b (2003)
length of the proton-donating GeO bond decreases, while that
containing a proton-accepting oxygen atom increases.
1165e1170.
12] S.S. Karlov, E.Kh. Lermontova, M.V. Zabalov, A.A. Selina, A.V. Churakov,
4
5
6
7
2 2 2 2 2
. The equilibrium structure of the [(HO) Ge(OCH CH ) NH]
J.A.K. Howard, M.Yu. Antipin, G.S. Zaitseva, Inorg. Chem. 44 (2005) 4879e4886.
dimer is in a good agreement with the structure found in the X-
ray study, although weaker inter-dimeric OH/O and N/H
interactions revealed in solid state are not present in the
dimeric model. In contrast to I, where the Ge/N transannular
distance increases on going from monomer to dimer, in II it
decreases.
[13] M.G. Voronkov, G.S. Samokhin, D.V. Vrazhnov, T.A. Kochina, Russ. J. Gen.
Chem. 82 (2012) 170e171.
[14] I.S. Ignatyev, M.G. Voronkov, T.A. Kochina, D.V. Vrazhnov, Russ. J. Gen. Chem.
80 (2010) 2274e2282.
[15] M.G. Voronkov, Z.A. Ovchinnikov, V.P. Baryshok, Izv. Akad. Nauk SSSR, Ser.
Khim. 4 (1987) 880e882.
[16] M.G. Voronkov, Z.A. Ovchinnikov, V.P. Baryshok, Zh. Obshch. Khim. 57 (1987)
2643e2644.
. There is a significant increase of predicted
d
GeOH and
s
GeOH
[17] V.F. Mironov, T.K. Gar, N.Yu. Khromova, O.D. Frid, Zh. Obshch. Khim. 56 (1986)
638e641.
[18] T.K. Gar, N. Yu Khromova, N.V. Sonina, V.S. Nikitin, M.V. Polyakova,
V.F. Mironov, Zh. Obshch. Khim 49 (1979) 1516e1522.
frequencies on going from monomeric to dimeric models due
ꢀ1
to formation of hydrogen bonds (from 958 to 1133 cm and
ꢀ
1
from 74 to 638 cm in I). These deformation vibrations of
[19] H.C. Chiang, S.M. Lin, C.H. Ueng, Acta Crystallogr. C 48 (1992) 991e993.
hydrogen bonded GeOH groups may be assigned to the
[20] D.H. Chen, H.C. Chiang, Polyhedron 14 (1995) 687e691.
[21] M.G. Voronkov, A.A. Korlyukov, E.A. Zelbst, S.P. Knyazev, I.M. Vasilyev,
E.A. Chernyshev, M.Yu. Antipin, J. Struct. Chem. 51 (2010) 719e724.
ꢀ1
experimental IR bands observed at 1102 and 636 cm for 1-
germatranol.
[22] M.G. Voronkov, A.A. Korlyukov, D.R. Khamitova, M.I. Buzin, G.S. Samokhin, T.
. Similar increase of theoretical
d GeOH and s GeOH frequencies
A. Kochina, J. Struct. Chem., in press.
[
[
23] P.D. Lickiss, Adv. Inorg. Chem. 42 (1995) 147e262.
24] V. Chandrasekhar, R. Boomishankar, S. Nagendran, Chem. Rev. 104 (2004)
is observed in II. For the equatorial GeOH group (proton
ꢀ1
acceptor) these modes are assigned to 1115 and 816 cm
bands, and for the axial (proton donor) group to 1011 and
5847e5910.
[25] I.S. Ignatyev, M. Montejo, F. Partal Ureña, J.J. López González, Chem. Phys. Lett.
ꢀ1
412 (2005) 359e364.
3
46 cm IR bands. However, the assignment of the latter mode
[26] M.G. Voronkov, A.A. Korlyukov, G.S. Samokhin, D.V. Vrazhnov, T.A. Kochina,
may be dubious, due to the absence of inter-dimeric hydrogen
bonds in our model.
. Only one energy minimum was found at the potential energy
Izv. Akad. Nauk SSSR, Ser. Khim. (2012) 987e993.
[27] B.C. Smith, Fundamentals of Fourier Transform Infrared Spectroscopy, CRC
Press, Boca Raton, 2011, p. 97.
[
[
28] A.D. Becke, J. Chem. Phys. 98 (1993) 5648e5652.
29] C.T. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785e789.
3 2 2 2 2
surface of the [(HO) Ge(OCH CH )NH ] dimer, in which two
hydroxyl groups of each monomer form hydrogen bonds (one
as a proton donor and the other as an acceptor) while the third
one remains “free”. Substantial deviations of theoretical fre-
quencies from experimental ones may indicate that a dimeric
model of III is a poor approximation for the description of the
solid state experimental spectrum of III in contrast to dimeric
models of II and, especially, I. Solid state structure may com-
prise a three dimensional network formed by strong hydrogen
bonds.
[30] D.E. Woon, T.H. Dunning, J. Chem. Phys. 98 (1993) 1358e1371.
[31] T.H. Dunning, J. Phys. Chem. A 104 (2000) 9062e9080.
[
32] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson,
H. Nakatsuji, M. Caricato, X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino,
G. Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers,
K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.M. Millam,
M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski,
R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P. Salvador,
J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö. Farkas, J.B. Foresman, J.V. Ortiz,
J. Cioslowski, D.J. Fox, Gaussian 09, Revision B.01, Gaussian Inc., Wallingford,
Appendix A. Supplementary data
2
009.
33] P. Pulay, G. Fogarasi, G. Pongor, J.E. Boggs, A. Vargha, J. Am. Chem. Soc. 105
1983) 7037e7047.
34] G. Fogarasi, X.F. Zhou, P.W. Taylor, P. Pulay, J. Am. Chem. Soc. 114 (1992)
191e8201.
35] T. Sundius, J. Mol. Struct. 218 (1990) 321e326.
[
[
[
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jorganchem.2012.12.036.
(
8
References
[36] T. Sundius, Vib. Spectrosc. 29 (2002) 89e95.
37] I. Ignatyev, T. Sundius, Spectrochim. Acta A 95 (2012) 37e45.
[
[
[
[
1] M.G. Voronkov, Top. Curr. Chem. 84 (1979) 77e135.
2] J.G. Verkade, Coord. Chem. Rev. 137 (1994) 233e295.
3] V.K. Pestunovich, S.V. Kirpichenko, M.G. Voronkov, Silatranes and their tri-
cyclic analogs, in: Z. Rappoport, Y. Apeloig (Eds.), Chemistry of Organic Silicon
Compounds, vol. 2, Wiley, Chichester, 1998, pp. 1447e1537.
[38] A.E. Reed, P.v.R. Schleyer, J. Am. Chem. Soc. 109 (1987) 1362e1373.
[39] B. Rempfer,H. Oberhammer, N. Aunert, J. Am. Chem. Soc.108(1986)3893e3897.
[40] J. Dillen, J. Phys. Chem. A 108 (2004) 4971e4977.
[41] Yu. P. Egorov, M.G. Voronkov, T.B. Lutsenko, G.I. Zelchan, Chem. Heterocyc.
Comp. 2 (1966) 17e24.
[42] M. Imbenotte, G. Palavit, P. Legrand, J. Raman Spectrosc. 14 (1983) 135e137.
[43] M. Imbenotte, G. Palavit, P. Legrand, J.P. Huvenne, G. Fleury, J. Mol. Spectr. 102
(1983) 40e55.
[44] I.S. Ignatyev, A.N. Lazarev, S.G. Shevchenko, V.P. Baryshok, Izvest. Akad. Nauk
SSSR, Ser. Khim. 7 (1986) 1518e1526 (Russ. Chem. Bull. 35 (1986) 1375e1382).
[45] R. Eujen, E. Petrauskas, A. Roth, D.J. Brauer, J. Organomet. Chem. 613 (2000)
86e92.
[
[
[
[
[
[
4] J.K. Puri, R. Singh, V.K. Chahal, Chem. Soc. Rev. 40 (2011) 1791e1840.
5] T.K. Gar, V.F. Mironov, Metalloorg. Khim. 1 (1988) 260.
6] R.Y. Chen, L.Z. Liu, Z.B. Zhang, Heteroat. Chem. 6 (1995) 503e506.
7] E. Lukevics, S. Belyakov, O. Pudova, J. Organomet. Chem. 523 (1996) 41e45.
8] X.T. Feng, S. Cui, R.Z. Cao, L.Z. Liu, Main Group Met. Chem. 20 (1997) 213e216.
9] R. Cea-Olivares, V. Garcia-Montalvo, M.M. Moya-Cabrera, Coord. Chem. Rev.
249 (2005) 859e872.