Precise measurements of 115/117/119Sn NMR frequencies and first observation of isotope-induced chemical shifts 1Δ117/119Sn(15N), 1Δ115/119Sn(15N), and 2Δ117/119</

Article abstract of DOI:10.1016/S0022-2860(01)00729-3

Tris(trimethylstannyl)amine 1 and trimethylsilyl-bis(trimethylstannyl)amine 2 (both ca. 13% ¹⁵N labelled) were studied by ^115/117/119Sn-, ¹⁵N-, and ²⁹Si NMR spectroscopy. Precise measurements of the Sn NMR frequ

Full text of DOI:10.1016/S0022-2860(01)00729-3

Journal of Molecular Structure 602±603 2002) 177±184
www.elsevier.com/locate/molstruc
Precise measurements of 115/117/119Sn NMR frequencies
and ®rst observation of isotope-induced chemical shifts
1
2
1D 117/119Sn 15N), D 115/119Sn 15N), and D 117/119Sn 29Si) in
N-trimethylstannyl aminesq
a,
b
Bernd Wrackmeyer , Martina Vosteen , Wolfgang Storch
b
*
a
È È
Laboratorium fur Anorganische Chemie, Universitat Bayreuth, D-95440 Bayreuth, Germany
È È È
Department Chemie der Universitat Munchen, Butenandt-Straûe 5-13, Haus D, D-81377 Munchen, Germany
b
Received 14 February 2001; accepted 7 March 2001
Abstract
Tris trimethylstannyl)amine 1 and trimethylsilyl-bis trimethylstannyl)amine 2 both ca. 13% 15N labelled) were studied by
115/117/119Sn-, 15N-, and 29Si NMR spectroscopy. Precise measurements of the Sn NMR frequencies of the respective Sn/15N
isotopomers did not reveal any appreciable primary isotope effect, in agreement with the previous results. The standard
frequencies J 119Sn) and J 117Sn), determined for tetramethyltin more than 30 years ago, were found to be very accurate,
1
and Jꢀ115Sn† ˆ 32:718781 MHz was measured in this work. In the 15N NMR spectra, small secondary isotope effects D 117/
1
119Sn 15N) and D 115/119Sn 15N) could be detected for the ®rst time, the 15N nucleus in the heavier isotopomer being more
1
shielded. In the 29Si NMR spectrum of 2, the expected isotope-induced chemical shift D14/15N 29Si) is shown, and in addition
2
the so far unknown effect over two bonds D117/119Sn 29Si) is visible. q 2002 Elsevier Science B.V. All rights reserved.
Keywords: Amines; Tin; Tin±; Nitrogen±; Silicon NMR; Isotope-induced chemical shifts
1. Introduction
115Sn) or larger by a factor .20 117Sn, 119Sn) [1,2].
The nucleus of choice for NMR studies is normally
119Sn, but 117Sn NMR spectra can be measured almost
with the same sensitivity, and there are applications
of 115Sn NMR [3]. For many tin compounds, the
115/117/119Sn resonance frequencies can in principle
be determined with high accuracy, although this is not
routinely done. Frequently the 115/117/119Sn resonance
signals in solutions are somewhat broadened by
exchange processes due to dynamic association
processes, by slow molecular motion and/or by slow
motion of groups attached to the tin nucleus, or by
partially relaxed scalar indirect nuclear spin±spin
coupling of the tin nuclei with quadrupolar nuclei
Of all elements tin possesses the largest number of
stable isotopes from 112Sn 0.96%) to 124Sn 5.94%),
the most abundant being 120Sn 32.85%). Tin is also
unique, since there are three spin-1/2 nuclei, suitable
for NMR experiments, namely 115Sn 0.35%), 117Sn
7.61%) and 119Sn 8.58%), all of which have a recep-
tivity for NMR experiments either close to that of 13C
q
Dedicated to Professor Graham A. Webb on the occasion of his
65th birthday.
* Corresponding author. Fax: 149-921-552-157.
E-mail address: b.wrack@uni-bayreuth.de B. Wrackmeyer).
0022-2860/02/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-2860 01)00729-3

178
B. Wrackmeyer et al. / Journal of Molecular Structure 602±603 :2002) 177±184
119SnMe3
119SnMe3
15N
115/117/119SnMe3
14N
115/117/119SnMe3
15N
15N
117SnMe3
115SnMe3
Me3Sn
SnMe3
SnMe3 Me3Sn
Me3Sn
Me3Sn
SnMe3
15N
115SnMe3
15N
117SnMe3
Me3Sn
SnMe3
Me3Sn
Scheme 1. Relevant isotopomers for NMR studies of tris trimethylstannyl)amine 1.
[1,2]. High precision 115/117/119Sn frequency measure-
ments are available so far only for tetramethyltin,
Me4Sn [4,5], the accepted reference for Sn chemical
shifts. Primary isotope effects on shielding and
coupling constants involving the tin nuclei appear to
be close to zero [6]. Isotope induced chemical shifts of
other nuclei due to the presence of 115Sn, 117Sn or
119Sn have not been noted. In order to get a more
clear picture on these issues, we have determined
the precise frequency of 115Sn in Me4Sn, and we
have been looking for a suitable compound, in
which there are at least three chemically equivalent
tin atoms present, all linked directly to a magnetically
active nucleus, for which the resonance can be
observed within short time, and for which the nuclear
shielding might likely to be affected by small changes
in its surroundings, e.g. by the changes in the
mass of the tin isotopes. The ®nal choice was
tris trimethylstannyl)amine, N SnMe3)3 1, 15N
labelled ca. 13%).
ening of the 13C resonance signal for the central
carbon atom, preventing a highly accurate determina-
tion of relevant frequencies. We refrained from
preparing 13C labelled material, since the outcome
of the measurements seemed to be doubtful, and
part of the expensive preparation is not necessarily a
high-yield procedure.
Considering the long transversal relaxation time of
15N nuclei [8], in the absence of protons attached to
nitrogen giving rise to extremely sharp 15N reso-
nances, partially 15N labelled tris trimethylstannyl)-
amine, N SnMe3)3 1, appeared to be a better
candidate. This compound is readily available by the
quantitative reaction of Me2N±SnMe3 with 15N
labelled ammonia see Section 4). Scheme 1 shows
the most abundant isotopomers relevant to NMR spec-
troscopy of this compound ignoring the methyl
groups). Compound 1 15N labelled or unlabelled)
has been repeatedly studied mainly with respect to
the signs of coupling constants [9±12]. The measure-
ments carried out in this work were aiming for precise
115/117/119Sn resonance frequencies, in order to ®nd
primary isotope effects, and for precise 15N NMR
frequencies in order to detect secondary isotope effects,
arising from the presence of various tin isotopes. In
addition to 1, the 15N labelled N-trimethylsilyl-N,N-
bis trimethylstannyl)amine, Me3Si±N SnMe3)2 2,
available from 1 by the reaction with Me3SiCl [13],
was prepared and studied by tin NMR as well as by
15N- and 29Si NMR. A list of experimental NMR data
for the compounds 1 and 2 is given in Table 1.
2. Results and discussion
2.1. General
Starting to look for a suitable compound, we ®rst
selected tris trimethylstannyl)methane, HC SnMe3)3,
and tetrakis trimethylstannyl)methane, C SnMe3)4
see Ref. [7] for previous NMR studies). The
measurements were not very successful, owing to
the low natural abundance of the 13C nucleus which
made it dif®cult to record the spectra of the isoto-
pomer containing 13C/115Sn within short time.
Prolonged measurement times led to slight broad-
2.2. Tin NMR measurments
115/117/119Sn NMR spectra were measured by the

Table 1
15N-, 119Sn- and 29Si NMR parameters of the compounds 1 and 2. Compound 1 in D8-toluene ca. 15% v/v); compound 2 in C6D6 ca. 15% v/v), both in 5 mm OD) tubes, measured
at 23 ^ 18C; accuracy of chemical shifts d better than ^0.05 ppm d 13C ^ 0.1); coupling constants J 29Si,X) are given in ), J 119Sn,X) in [ ], J 117Sn,X) in { }, and J 115Sn,X) in
, . , with an accuracy better than ^ ˆ 0:03 Hz; data were reproduced by using three different spectrometers. Isotope induced chemical shifts 1D and 2D in ppb ^0.2); a negative
sign denotes a shift to higher frequency with respect to the NMR signal of the respective lighter isotopomer)
No d13
C
d15
N
d119Sn
1D14/15N 119Sn)
2 38.4
1D 115/119Sn 119Sn) 1D117/119Sn 119Sn)
1
2 1.1 [364.5]
2 396.2 [83.68] 87.8 [185.7]
{79.97} ,73.42 . {170.4} ,163.3 .
2 3.2
2 1.7
d29Si
1D14/15N 29Si) 1D 28/29Si 15N)
1D117/119Sn 119Sn) 2D117/119Sn 29Si)
2
2 1.7 MeSn)
[378.0, 4.1] 6.1
2 375.4 [61.43]
{58.70} 6.54)
64.7 [185.7]
7.55)
5.79 [7.55]
{7.21}
2 46.7
2 10.5
2 2.2
2 1.0
2 0.8
MeSi) 54.1) [10.2]

180
B. Wrackmeyer et al. / Journal of Molecular Structure 602±603 :2002) 177±184
Fig. 1. 111.9 MHz 119Sn NMR spectra of tris trimethylstannyl)amine, N SnMe3)3 1 labelled ca. 13% with 15N), after eight transients. A)
Recorded with inverse gated 1H decoupling; the dominant broad central signal represents the 119Sn/14N isotopomer, the broad satellites are due
to the 119/117Sn/14N isotopomer, the sharp doublet due to the 119Sn/15N isotopomer is already visible; B) Recorded with the refocused INEPT
1
pulse sequence and H decoupling [14±17]; the intensity of the broad central line is already somewhat reduced due to loss of magnetisation
during the refocusing delays in the pulse sequence; C) Recorded with the INEPT-HEED pulse sequence [18], by which the magnetisation of
the 119Sn/14N isotopomer decays to zero as a result of additional delays two times 75 ms in the Hahn-echo extension).
1
refocused INEPT pulse sequence with H decoupling
[14±17]. Polarisation transfer was based on the
decoupled 119Sn NMR spectrum here recorded by
inverse gated decoupling in order to suppress the
negative NOE, since g 115/117/119Sn) , 0) shows the
dominant broad central line due to the isotopomer
containing 14N/119Sn Fig.1 A)). In the 119Sn NMR
spectrum recorded by the 1H decoupled refocused
INEPT pulse sequence [14±17], the intensity of the
coupling
constants
2JꢀSn; 1H† ˆ 57 Hzꢀ119Sn†;
54.5 Hz 117Sn) and 50.0 Hz 115Sn). The effect of
the presence of the quadrupolar 14N nucleus on all
tin NMR spectra is readily seen by the inspection of
119Sn NMR spectra of 1 in Fig. 1. The `normal' H
1
Table 2
Experimental resonance frequencies n Sn) for the 1H NMR signal of Me4Si at exact 300.13 MHz) and ratios of resonance frequencies for
1
15N SnMe3)3 and Me4Sn J Sn) values with the H NMR signal of Me4Si at exact 100.000000 MHz)
15N SnMe3) 1
Frequencies
dSna
dSn
Frequencies
Me4Sn
n 119Sn) MHz)
n 117Sn) MHz)
n 115Sn) MHz)
111.9303013
106.9525975
98.2074992
88.78
88.90
88.87
0
0
0
37.290629
35.632256
32.718746
J 119Sn) MHz)b,c
J 117Sn) MHz)b,c
J 115Sn) MHz)b,c
Frequency ratio
1.046541215
1.139732731
1.089047154
Frequency ratio
1.046541642
1.139732831
1.089047117
n 119Sn)/n 117Sn)
n 119Sn)/n 115Sn)
n 117Sn)/n 115Sn)
J 119Sn)/J 117Sn)
J 119Sn)/J 115Sn)
J 117Sn)/J 115Sn)
dSn ˆ {‰ꢀnꢀSn†=3:0013† 2 JꢀSn†Š=JꢀSn†}106:
a
b
c
Recently proposed [32] frequencies for Me4Sn.
Old reference frequencies Me4Sn 1 30% C6D6) ‰Jꢀ119Sn† ˆ 37:290665 MHz [4]; Jꢀ 117Sn† ˆ 35:632295 MHz [4]; Jꢀ115Sn† ˆ
2:718781 MHz this work) give dSn values up to 1 ppm different, but are more consistent d 119Sn ˆ 87:82; d117Sn ˆ 87:81 and d115Sn ˆ
87:83:

B. Wrackmeyer et al. / Journal of Molecular Structure 602±603 :2002) 177±184
181
Fig. 2. 30.4 MHz 15N NMR spectrum of N-trimethylsilyl-N,N-bis trimethylstannyl)amine, Me3SiN SnMe3)2 2 labelled ca. 13 % with 15N),
after four transients. The 117/119Sn satellites are marked by arrows and the 29Si satellites by asterisks. Close to the bottom of the central line are
additional satellite due to 15N±13C couplings across two bonds.
broad central line is already somewhat reduced owing
to the loss of the fast decaying magnetization for the
14N/119Sn isotopomer during the pulse sequence Fig.
1 B)). Finally, the INEPT-HEED [18] HEED ˆ
Hahn echo extended) experiment was applied to elim-
inate the broad central line completely Fig. 1 C)).
The same situation applies of course to 117Sn and
115Sn NMR spectra. In contrast with the broad central
resonance, the frequencies of the sharp doublet owing
to the scalar Sn±15N coupling can always be measured
precisely. In agreement with the previous results [6],
there is no appreciable primary isotope effect on tin
nuclear shielding, as shown in Table 2 by compar-
ison of the respective ratio of the resonance frequen-
cies and the chemical shifts dSn with respect to
Me4Sn.
and the slight broadening of the lines can be toler-
ated.
2.3. 15N NMR spectra
The refocused INEPT pulse sequence with 1H
decoupling was used to record the 15N NMR spectra.
Delays corresponding to assumed coupling constants
15
3Jꢀ N; 1H† ˆ 1:3 Hz for 1 and 1.8 Hz for 2 gave best
results for polarisation transfer, and the 1H decoupling
power 1H decoupling frequency offset set on reso-
nance) was adjusted for low power selective decou-
pling in order to minimize the dielectric heating of the
sample during the acquisition time. This is important
since acquisition times of 15±20 s were necessary to
collect the FID's without truncation.
In 15N NMR spectra of 1, a line width h1=2
0:075 Hz after four scans) could be achieved by
careful shimming of the magnet. This line width
ˆ
The tin NMR spectra of 2 shows the same features
as those of 1, except that there are also 29Si satellites
due to geminal scalar Sn±29Si coupling the number
of relevant isotopomer in Scheme 1 is markedly
increased by 29Si if one of the SnMe3 group is replaced
by a SiMe3 group), and the intensities of 117Sn or 115Sn
satellites are lower than in 1. Thus, longer measure-
ment times were required which led to the slight
broadening of all signals. Therefore, these spectra
were not used to assess any isotope effects with the
0
increased only to h1=2 ˆ 0:115 Hz after 400 transi-
ents. This made it possible to assign the 117/119Sn satel-
lites easy to observe already after a single transient)
as well as the 115Sn satellites with high accuracy. The
119Sn satellites are completely symmetric within the
experimental error) with respect to the central 15N
resonance signal. The 117Sn satellites are shifted
slightly to lower ®eld, and this trend increases for
1
exception of D 14/15N 119Sn) which is a large effect

182
B. Wrackmeyer et al. / Journal of Molecular Structure 602±603 :2002) 177±184
ment of the 29Si satellites gives 1D28/29Si 15N) value in
the same order of magnitude as for other examples
[19,20].
2.4. 29Si NMR spectra
In the 29Si NMR spectra of 2, the central line is
broadened owing to partially-relaxed scalar 29Si±14N
1
coupling. Since the magnitude of J 29Si,14N) of 2 is
1
markedly smaller when compared with J 119Sn,14N)
of 1 or 2 compare the coupling constants with 15N in
Table 1), the broadening is much less dramatic than in
the tin NMR spectra, and isotope effects can be deter-
mined without recording 29Si NMR spectra in which
the central line is eliminated. This is shown in Fig. 3,
2
where a small isotope-induced chemical shift D117/
119Sn 29Si) becomes evident for the ®rst time. On the
high-®eld side, the 117/119Sn satellites are resolved,
whereas they overlap on the low-®eld side. The FID
resolution was 0.03 Hz which excludes any artefact.
The precise value of the coupling constant
1J 29Si,15N) already determined from the 15N NMR
spectrum of 2 see Fig. 2) and con®rmed in this
experiment Fig. 3), allows here for the accurate deter-
1
mination of D 14/15N 29Si).
Fig. 3. 59.6 MHz 29Si NMR spectrum of N-trimethylsilyl-N,N-
bis trimethylstannyl)amine, Me3SiN SnMe3)2 2 labelled ca. 13%
with 15N), after four transients. The 15N satellites asterisks) are
readily assigned knowing the magnitude from the 15N NMR spec-
trum see Fig. 2). The 117/119Sn satellites arrows) are resolved at the
high-®eld side and overlap at the low-®eld side which indicates a
3. Conclusions
It appears that it is worthwhile to look for
secondary isotope effects 1D 117/119Sn X) or 1D 115/
119Sn X), in particular if the chemical shift of the X
nucleus is sensitive to small changes in the surround-
ings. Our knowledge about isotope-induced chemical
shifts is still limited, at least for heavier nuclei [21].
The present work adds some new data, which are in
principle in agreement with expectations the 15N
nucleus in the heavier isotopomer is better shielded).
Clearly, a larger data set is required to establish such a
rule ®rmly for the isotope effects observed here.
Surprisingly, a two bond effect 2D 117/119Sn Si) has
also been found. That means, it might be rewarding
to study compounds with a stannyl group in the
geminal position relative to a nucleus X in more detail
than it has been done before. To the best of our
knowledge, the isotope effects detected in this work
represents the in¯uence of the smallest mass differ-
ence in percent observed so far in NMR spectroscopy.
2
small isotope-induced chemical shift D117/119Sn 29Si).
the 115Sn satellites see Table 1). Clearly, this points to
1
isotope-induced chemical shifts D 117/119Sn 15N) and
1D115/119Sn 15N), observed here for the ®rst time.
In the case of 2, independent of the solvent D8-
toluene or C6D6, and also independent of the condi-
tions i.e. short or very long repetition delays), broad-
ening and even splitting of the 15N resonance signals
were observed already after 8 or 16 transients, most
likely a result of temperature dependence of the 15N
chemical shift. Fig. 2 shows the 15N NMR spectrum of
2 after four transients where the signal is still narrow
117/
0
ꢀh1=2 ˆ 0:10 Hz† and enables one to measure the
119Sn and 29Si satellites with the necessary accuracy.
Again, it was found that a small isotope-induce
1
chemical shift D 117/119Sn 15N) exists. The measure-

B. Wrackmeyer et al. / Journal of Molecular Structure 602±603 :2002) 177±184
183
4. Experimental
Acknowledgements
4.1. General, starting materials, and NMR
spectroscopy
We thank the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie for the
support of this work. B.W. thanks CINVESTAV
Mexico City) for the measurement time at the Bruker
DPX 300 instrument.
Dimethylaminotrimethyltin, Me2N±SnMe3 [22±
24] was prepared according to the literature proce-
dure. Trimethyltin chloride was obtained by the reac-
tion of tetramethyltin with tin tetrachloride 3:1 ratio;
slow mixing without a solvent, heating to 1308C for
2 h; recrystallised from hexane). The 15N-labelled
starting material [15N]ammonium chloride [13.6%
enriched) was a commercial product Chemotrade±
Leipzig), used without further puri®cation.
References
[1] B. Wrackmeyer, in: G.A. Webb Ed.), Reports on NMR
Spectroscopy, vol. 16, Academic Press, London, 1985,
pp. 73±185.
[2] B. Wrackmeyer, in: G.A. Webb Ed.), Reports on NMR Spec-
troscopy, vol. 38, Academic Press, London, 1999, pp. 203±
264.
NMR spectra were measured using Bruker DRX
500, DPX 300 and ARX 250 spectrometers, all
equipped for multinuclear magnetic resonance. The
[3] J.C. Meurice, M. Vallier, M. Ratier, J.-G. Duboudin, M.J.
Petraud, J. Chem. Soc., Perkin Trans. 2 1996) 1331.
[4] A.G. Davies, P.G. Harrison, J.D. Kennedy, R.J. Puddephatt,
T.N. Mitchell, W. McFarane, J. Chem. Soc. A) 1969) 1136.
[5] P.G. Harrison, J.J. Zuckerman, S.E. Ulrich, J. Am. Chem. Soc.
93 1971) 5398.
1
908 pulses for H, 117/119Sn, 29Si and 15N were cali-
brated for each spectrometer, using the actual samples
1 and 2 which were sealed in 5 mm tubes in order to
prevent any changes in the composition. The 908 pulse
for 115Sn was assumed to be same as that for 117Sn on
each spectrometer. Digital resolution of the FIDs were
in general 0.05 Hz or better. All spectra were
processed with a minimum of line broadening
0.01 Hz) or without any line broadening. The condi-
tions for measuring the NMR spectra were adjusted to
correspond close to those reported for ultra-high reso-
lution NMR [25±30].
[6] H.C.E. McFarlane, W. McFarlane, C.J. Turner, Mol. Phys. 37
1979) 1639.
È
[7] W. Biffar, T. Gasparis-Ebeling, H. Noth, W. Storch, B.
Wrackmeyer, J. Magn. Reson. 44 1981) 54.
[8] G.J. Martin, M.L. Martin, J.-P. Gouesnard, 15N NMR spectro-
scopy, in: P. Diehl, E. Fluck, R. Kosfeld Eds.), NMR Ð
Basic Principles and Progress, vol. 18, Springer, Berlin, 1981.
[9] B. Wrackmeyer, H. Zhou, Magn. Reson. Chem. 28 1990)
1066.
[10] E. Kupce, B. Wrackmeyer, J. Magn. Reson. 99 1992) 338.
[11] E. Kupce, B. Wrackmeyer, J. Magn. Reson. 101A 1993)
324.
È
[12] B. Wrackmeyer, J. Weidinger, H. Noth, W. Storch, T. Seifert,
4.2. Syntheses
M. Vosteen, Z. Naturforsch., Teil B 53 1998) 1494.
[13] H.W. Roesky, H. Wietzer, Chem. Ber. 107 1974) 3186.
[14] G.A. Morris, R. Freeman, J. Am. Chem. Soc. 101 1979)
760.
4.2.1. [15N]tris:trimethylstannyl)amine 1
[15N]tris trimethylstannyl)amine 1 was obtained by
the reaction of dry [15N]ammonia gas evolved from
the reaction mixture of [15N]ammonia chloride and
an aqueous solution of NaOH) with Me2NSnMe3.
[15] G.A. Morris, J. Am. Chem. Soc. 102 1980) 428.
[16] G.A. Morris, J. Magn. Reson. 41 1980) 185.
[17] D.P. Burum, R.R. Ernst, J. Magn. Reson. 39 1980) 163.
[18] E. Kupce, B. Wrackmeyer, J. Magn. Reson. 97 1992) 568.
[19] B. Wrackmeyer, G. Kehr, H. Zhou, S. Ali, Magn. Reson.
Chem. 34 1996) 921.
The product
described [31].
1 was puri®ed and isolated as
[20] B. Wrackmeyer, G. Kehr, D. Wettinger, Magn. Reson. Chem.
36 1998) S157.
[21] C.J. Jameson, Isotopes, in: E. Buncel, J.R. Jones Eds.), The
Physical and Biomedical Sciences, vol. 2, Elsevier, New
York, 1991, pp. 1±54.
4.2.2. [15N]N-Trimethylsilyl-N,N-
bis:trimethylstannyl)amine 2
Treatment of [15N]tris trimethylstannyl)amine 1
with an excess of trimethylsilylchloride gave 2 as
described for the non-labelled compound according
to the literature procedure [13].
[22] K. Jones, M.F. Lappert, Proc. Chem. Soc. 1962) 358.
[23] K. Jones, M.F. Lappert, J. Chem. Soc. 1965) 1944.
[24] C.M. Wright, E.L. Muetterties, Inorg. Synth. 10 1976) 137.
[25] A. Allerhand, R.E. Addleman, D. Osman, J. Am. Chem. Soc.
107 1985) 5809.

184
B. Wrackmeyer et al. / Journal of Molecular Structure 602±603 :2002) 177±184
[26] S.R. Maple, A. Allerhand, J. Magn. Reson. 66 1986) 168.
[27] S.R. Maple, J.E. Carson, A. Allerhand, J. Am. Chem. Soc. 111
1989) 7293.
[30] B. Wrackmeyer, E. Kupce, Z. Naturforsch., Teil B 53 1998)
411.
[31] K. Sisido, S. Kozima, J. Org. Chem. 29 1964) 907.
[32] P. Granger, unpublished results; in preliminary form: R.K.
Harris, in: D.M. Grant, R. K. Harris Eds.), Encyclopedia of
NMR, Wiley, Chichester, 5 1996) p. 3301.
[28] E. Kupce, E. Lukevics, J. Magn. Reson. 80 1988) 359.
[29] E. Kupce, E. Lukevics, J. Chem. Soc., Chem. Commun.
1989) 818.