Silicon-29 NMR spectra of trimethylsilylated alcohols

Article abstract of DOI:10.1135/cccc19970816

²⁹Si NMR spectra of trimethylsilyl (TMS) derivatives of 26 simple alcohols were measured under standardized conditions (i.e., in sufficiently diluted deuteriochloroform solutions). Due to association with the solvent the chemical shifts are in almost all cases larger than those reported earlier for different solutions. This observation is in agreement with the proposed mechanism of steric effects as being due to sterically controlled association with the solvent. The use of chloroform as a solvent enhances steric effects hut at the same time it can reduce small differences due to polar effects in closely related compounds. In the studied class of compounds the gross dependence of the chemical shift on polar effects is not substantially affected by the change of the solvent.

Full text of DOI:10.1135/cccc19970816

816
Kvicalova, Cermak, Blechta, Schraml:
SILICON-29 NMR SPECTRA OF TRIMETHYLSILYLATED ALCOHOLS
1
2
3
4
Magdalena KVICALOVA , Jan CERMAK , Vratislav BLECHTA and Jan SCHRAML
Institute of Chemical Process Fundamentals, Academy of Sciences of the Czech Republic,
165 02 Prague 6-Suchdol, Czech Republic; e-mail: 1 kvicalova@icpf.cas.cz, 2 cermak@icpf.cas.cz,
3 blechta@icpf.cas.cz , 4 schraml@icpf.cas.cz
Received January 16, 1997
Accepted February 14, 1997
29Si NMR spectra of trimethylsilyl (TMS) derivatives of 26 simple alcohols were measured under
standardized conditions (i.e., in sufficiently diluted deuteriochloroform solutions). Due to association
with the solvent the chemical shifts are in almost all cases larger than those reported earlier for dif-
ferent solutions. This observation is in agreement with the proposed mechanism of steric effects as
being due to sterically controlled association with the solvent. The use of chloroform as a solvent
enhances steric effects but at the same time it can reduce small differences due to polar effects in
closely related compounds. In the studied class of compounds the gross dependence of the chemical
shift on polar effects is not substantially affected by the change of the solvent.
Key words: 29Si NMR; Trimethylsilyl derivatives; Alcohols; Polar effects; Steric effects.
Studies of 29Si NMR spectra of trimethylsilyl1 (TMS) and tert-butyldimethylsilyl2
(TBDMS) derivatives of amino acids produced excellent correlation (r = 0.999 for 20
data points) between the 29Si chemical shifts in the two classes of compounds measured
under identical conditions3. The correlation was surprisingly good in the sense that it
involved both O- and N-bonded silicon atoms and covered a large range of the chemi-
cal shifts (δ(TBDMS) 8.1 to 25.8 ppm). Application of this correlation to chemical
shifts in disilyl derivatives of diols showed, however, large deviations even in cases
where no interaction between the two bulky groups was possible4. In order to find the
cause of these deviations, we decided to remeasure chemical shifts of TMS derivatives
of simple alcohols under the conditions (diluted deuteriochlorofom solutions) now con-
sidered standard. Similar reinvestigation of TMS derivatives of several adamantanols5
and steroid alcohols6 has allowed simple interpretation of the results in terms of steri-
cally controlled association with the solvent. In the present paper we report 29Si chemi-
cal shifts of TMS derivatives of simple alcohols (and phenol) and compare them with
the values reported earlier under not so well defined experimental conditions.
EXPERIMENTAL
Two different procedures were used for preparation of TMS derivatives of alcohols. (1) Larger-scale
preparations were performed by refluxing 10–20 g of alcohol with hexamethyldisilazane (HMDS,
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

Short Communication
817
1.25 mol per 1 mol OH) until evolution of ammonia ceased (about 2 days) followed by fractional
distillation. (2) Small-scale derivatizations were carried out by stirring ca 0.5 g of alcohol with sily-
lating reagent (bis(trimethylsilyl)acetamide, BSA, N-(trimethylsilyl)dimethylamine, TMSDMA, and
N-(trimethylsilyl)diethylamine, TMSDEA, chosen to maximize boiling point differences (1.3 mol per
1 mol OH) under dry nitrogen at 60–80 °C for 2 h. Pure TMS derivatives were isolated by fractional
distillation.
The NMR spectra were measured in dry deuteriochlorofom solutions containing 1% (v/v) of hexa-
methyldisilane (HMDSS) as a secondary reference. The concentration of the sample in the measured
solution was reduced until the 13C chemical shift of HMDSS was δ = –2.48 ± 0.02, relative to the
central line of the solvent at 76.99 pm (see ref.3 for details of this standard procedure).
All the NMR spectral measurements were performed on a Varian UNITY-200 spectrometer
(operating at 50.3 MHz for 13C and at 39.7 MHz for 29Si NMR measurements), using the standard
software (APT and INEPT pulse sequences) vnmr 3.2. The spectra were recorded in the temperature
range 22–24 °C. The 29Si NMR spectra were measured by the INEPT with the pulse sequence op-
timized3 for TMS derivatives, i.e., for coupling to 9 protons and coupling constant of 6.5 Hz. Acquisition
(1.0 s) was followed by a relaxation delay of 5 s. During the acquisition period WALTZ decoupling
was used and FID data (8 K) were sampled for the spectral width of 4 000 Hz. Zero filling to 32 K
and a mild exponential broadening were used in the data processing. The 29Si π/2 pulses were at maximum
1
17 µs long whereas H π/2 were 10 µs in a 5 mm switchable probe. The 29Si spectra were referenced to
the line of HMDSS at δ = –19.79. The 13C NMR spectra were measured using the spectral width of 16 000 Hz.
WALTZ decoupling was applied both during acquisition (1 s) and relaxation delay (2–5 s). Zero filling
to 64 K and 1–3 Hz line broadening were used in data processing.
Solvent accessible surface (A) was calculated for oxygen atoms and probe diameter of 100 pm
exactly as described previously6 for A (10) and geometry optimization.
RESULTS AND DISCUSSION
The 29Si NMR experimental results are summarized in Table I. As Figs 1 and 2 illus-
trate, the already discussed7 dependences of the chemical shift on Taft polar σ* con-
stants of substituents R and on the atomic charges on the silicon atoms, Q(Si), are not
substantially affected by the change in the experimental conditions.
With the single exception of the phenol derivative (which is more complex because
of additional effects due to magnetic anisotropy and formation of CT complex with
chloroform), all the 29Si chemical shifts measured under standard conditions (i.e., new
data) are larger than those reported earlier which were measured essentially in neat
liquids. The difference by far exceeds the estimated errors of the measurements (esti-
mated previously7 to be ±0.3 ppm and ±0.02 ppm here)*. This finding is in full agree-
ment with our earlier observations on adamantanol5 and steroid alcohol6 derivatives and
* The difference observed for hexamethyldisiloxane deserves special attention. This compounds is still
sometimes used as a reference despite the fact that already the compilation of Marsmann8 listed for it
a number of chemical shift values ranging from 4.0 to 7.2. Obviously, this chemical shift is subject to
a number of influences (concentration, solvent, and temperature) and the compound should not be
used as a secondary reference.
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

818
Kvicalova, Cermak, Blechta, Schraml:
TABLE I
29Si chemical shifts, δ, in trimethylsilylated alcohols R–O–Si(CH3)3, Taft polar σ* constants of sub-
stituent R and net atomic charges on the silicon atom, Q(Si)
δ(29Si)
R
σ*a
Q(Si) . 103a
olda
new
new – old
CH3
17.20
14.53
13.80
14.00
12.12
12.70
6.20
19.64
17.09
17.06
17.00
14.71
14.57
7.62
2.44
2.56
3.26
3.00
2.59
1.87
1.42
1.45
1.66
1.51
2.97
0.90
0.98
1.68
2.31
1.41
–
0.000
–0.100
–0.115
–0.130
–0.190
–0.150
–0.300
0.385
229.39
229.12
229.10
229.10
228.91
–
CH3CH2
CH3CH2CH2
CH3(CH2)3
(CH3)2CH
c-C6H11
(CH3)3C
228.75
229.46
229.15
229.10
229.10
229.77
230.04
–
Cl(CH2)2
18.36
16.70
16.10
14.30
20.65
21.75
17.40
16.59
18.29
–
19.81
18.36
17.61
17.27
21.55
22.73
19.08
18.90
19.70
14.78
18.74
16.37
17.16
7.34
Cl(CH2)3
0.140
Cl(CH2)4
0.050
Cl(CH2)6
0.006
Cl2CHCH2
CCl3CH2
0.693
0.946
C6H5CH2
0.215
CH2=CHCH2
BrCH2CH2
CH3CH2(CH3)CH
H3CO(CH2)2
(CH3)3CCH2
(CH3)3C(CH2)2
(CH3)3Si
0.170
–
0.360
229.50
–
–0.210
0.240
16.25
14.10
15.60
6.83
2.49
2.27
1.56
0.51
2.30
3.22
–
229.25
229.08
229.10
228.76
229.88
229.18
229.11
229.10
–
–0.165
–0.059
–0.810
–0.260
–0.093
0.033
(CH3)3SiCH2
(CH3)3Si(CH2)2
(CH3)3Si(CH2)3
(CH3)3Si(CH2)4
C6H5
16.10
13.10
–
18.40
16.32
16.97
17.02
19.43
13.90
19.62
3.12
–0.19
–0.012
0.600
a Values taken from ref.7.
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

Short Communication
819
with the theory proposed. In detail, however, the picture is more complex owing to
larger variability of the structure of compounds studied here.
The observed chemical shift is a weighted average of the chemical shifts in the free
compound and in its associate complex with chloroform. The chemical shift δ in the
complex is larger than that in the free compound. The populations of the complex and
21
δ, ppm
17
13
9
5
–0.2
0.2
0.6
1.0
σ*
FIG. 1
29Si chemical shift dependences on polar substituent constants σ* for new (●) and old (❍) data
21
δ, ppm
17
13
9
5
228.9
229.3
229.7
230.1
3
Q(Si) . 10
FIG. 2
29Si chemical shift dependences on the net atomic charge on the silicon atom Q(Si) for new (●) and
old (❍) data
Collect. Czech. Chem. Commun. (Vol. 62) (1997)

820
Kvicalova, Cermak, Blechta, Schraml:
the free compound are given by the concentration and association constant which we
take as proportional to the basicity of oxygen atom and its surface accessible for asso-
ciation with the solvent. In a series of closely related alcohol derivatives, one could
assume that either the oxygen basicity does not vary appreciably (e.g., in secondary
steroid alcohols6) or that it can be approximated by atomic net charge Q(O) on the
oxygen atom. Thus, in the series methyl, ethyl, isopropyl, and tert-butyl derivatives the
solvent-accessible surface A decreases as intuitively expected in the series 4.39, 3.20,
1.94, and 1.44 (104 pm2), respectively, while the net charge on the oxygen atom
becomes more negative in the series –0.33843, –0.34259, –0.34571, and –0.34814 e,
respectively. Obviously, in this short series, the solvent mediated steric effects are, at
least partially, compensated by the increase in the basicity of oxygen. The large
changes in the chemical shifts when going from primary to secondary, and further to
tertiary alcohol derivatives appear to be governed by the charge on the silicon atom
(Fig. 2).
These results point to a drawback of using the standard conditions. When inductive
effects increase the basicity of oxygen and at the same time, decrease the positive
charge on the silicon atom, association with chlorofom decreases the overall sensitivity
of 29Si chemical shifts to such inductive effects.
The authors are grateful to the Grant Agency of the Czech Republic for a Grant No. 203/94/1026
which made this work possible.
REFERENCES
1. Schraml J., Kvicalova M., Schwarzova I., Velisek J.: Magn. Reson. Chem. 32, 591 (1994).
2. Kubec R., Velisek J., Kvicalova M., Cermak J., Schraml J.: Magn. Reson. Chem. 33, 458 (1995).
3. Schraml J.: Prog. Nucl. Magn. Reson. 22, 289 (1990).
4. Kvicalova M., Blechta V., Kobylczyk K., Piekos R., Schraml J.: Collect. Czech. Chem. Commun.
62, 761 (1997).
5. Schraml J., Cermak J.: Collect. Czech. Chem. Commun. 55, 452 (1990).
6. Kasal A., Schraml J., Cermak J.: Magn. Reson. Chem. 32, 394 (1994).
7. Schraml J., Chvalovsky V., Magi M., Lippmaa E.: Collect. Czech. Chem. Commun. 46, 377
(1981).
8. Marsmann H.: NMR Basic Principles and Progress 17, 65 (1981).
Collect. Czech. Chem. Commun. (Vol. 62) (1997)