Alkali metal reduction of 2-halogeno- and 2-thiolato-2,3-dihydro-1H-1,3,2-diazaboroles

Article abstract of DOI:10.1039/b106128n

The reduction of 2-bromo-1,3,2-diazaborole tBuNCH=CH(tBu)BBr (4c) with a potassium mirror in 1,2-dimethoxyethane afforded a non-separable 1:2:1 mixture of the compounds tBuNCH=CHN(tBu)BH (5), tBuNCH=CHN(tBu)BOCH₃ (6) and {tBuNCH=CHN(tBu)B}₂O (7). Reaction of 4c with a potassium-sodium alloy in N,N,N′,N′-tetramethylethylenediamine led to 5 as the major product. 1,3,2-Diazaboroles tBuNCH=CHN(tBu)BN-Me₂ (8) and {tBuNCH=CHN(tBu)B}₂ (9) were spectroscopically identified as minor products. The treatment of 4c with potassium-sodium alloy in toluene solution in the presence of [15]crown-5 yielded a 1:1 mixture of 5 and the benzyl derivative tBuNCH=CHN(tBu)BCH₂Ph (12). The same reaction in toluene-d₈ produced the deuterated species 5-d₁ and 12-d₇. tBuNCH=CHN(tBu)BCH₃ (17) and 1,4-diazabutadiene (tBuN=CH)₂ (18) resulted from the treatment of tBuNCH=CHN(tBu)BSCH₃ (15) with potassium-sodium alloy in n-hexane. In contrast to this, compound 9 was obtained as the main product of the reduction of tBuNCH=CHN(tBu)BStBu (16) under similar conditions. The reduction of the 1-bromo-2-tert-butyl-1,2-dihydro[1,3,2]diazaborolo[1,5-a]pyridine (19) smoothly produced the respective diborane(4) derivative (20) which was subjected to X-ray diffraction analysis.

Full text of DOI:10.1039/b106128n

View Article Online / Journal Homepage / Table of Contents for this issue
Alkali metal reduction of 2-halogeno- and 2-thiolato-2,3-dihydro-
1H-1,3,2-diazaboroles
Lothar Weber,*a Markus Schniedera and Peter Lönneckeb
a Fakultät für Chemie der Universität Bielefeld, Universitätsstr. 25, D-33615 Bielefeld,
Germany. E-mail: lothar.weber@uni-bielefeld.de
b Institut für Anorganische Chemie der Universität Leipzig, Johannesallee 29, D-04103 Leipzig,
Germany. E-mail: loenneck@rz.uni-leipzig.de
Received 10th July 2001, Accepted 3rd October 2001
First published as an Advance Article on the web 9th November 2001
The reduction of 2-bromo-1,3,2-diazaborole tBuNCH᎐CH(tBu)BBr (4c) with a potassium mirror in 1,2-
᎐
dimethoxyethane afforded a non-separable 1 : 2 : 1 mixture of the compounds tBuNCH᎐CHN(tBu)BH (5),
᎐
tBuNCH᎐CHN(tBu)BOCH (6) and {tBuNCH᎐CHN(tBu)B} O (7). Reaction of 4c with a potassium–sodium alloy
᎐
᎐
3
2
in N,N,NЈ,NЈ-tetramethylethylenediamine led to 5 as the major product. 1,3,2-Diazaboroles tBuNCH᎐CHN(tBu)BN-
᎐
Me (8) and {tBuNCH᎐CHN(tBu)B} (9) were spectroscopically identified as minor products. The treatment of 4c
᎐
2
2
with potassium–sodium alloy in toluene solution in the presence of [15]crown-5 yielded a 1 : 1 mixture of 5 and the
benzyl derivative tBuNCH᎐CHN(tBu)BCH Ph (12). The same reaction in toluene-d produced the deuterated
᎐
2
8
species 5-d and 12-d . tBuNCH᎐CHN(tBu)BCH (17) and 1,4-diazabutadiene (tBuN᎐CH) (18) resulted from the
᎐
᎐
1
7
3
2
treatment of tBuNCH᎐CHN(tBu)BSCH (15) with potassium–sodium alloy in n-hexane. In contrast to this,
᎐
3
compound 9 was obtained as the main product of the reduction of tBuNCH᎐CHN(tBu)BStBu (16) under similar
᎐
conditions. The reduction of the 1-bromo-2-tert-butyl-1,2-dihydro[1,3,2]diazaborolo[1,5-a]pyridine (19) smoothly
produced the respective diborane(4) derivative (20) which was subjected to X-ray diffraction analysis.
Introduction
The preparation and full characterisation of the first stable
carbene A1 and the first isolable silylene B2 are landmarks in
the chemistry of these otherwise elusive electron-sextet species.
The diagonal relationship of boron and silicon in the periodic
table as well as the concept of isoelectronic compounds has
Scheme 1
motivated us to look for the yet unknown boranide anions C.
monohalogenoboranes.7 Their further reduction to dinuclear
Quantum chemical calculations have indicated that salts with
the anion C should be accessible in a chemical laboratory.3
Moreover, in the homologous series of the elements, B, Al, Ga,
In, and Tl anions based on gallium (D) should be the most
stable. This prediction has recently been verified by the syn-
thesis and X-ray structural characterisation of [K{[18]crown-
mono- and di-anions was recently reported.8,9 Ionic species
such as Mϩ(BR2)Ϫ, however, have never been observed.
In this paper we give an account of the alkali-metal reduction
of selected 1,3,2-diazaboroles and on the synthesis of novel
diborane(4)s.
6}(thf ) ][tBuNCH᎐CHN(tBu)Ga]Ϫ (3) in the research group of
᎐
2
Schmidbaur (Scheme 1).4
Results and discussion
Recently, we developed high-yield syntheses of 2-halogeno-
2,3-dihydro-1H-1,3,2-diazaboroles (E) as the required precur-
sors for the generation of anions such as C.5,6 It is conceivable
that the reduction of compound E to anion C proceeds via
bis(1,3,2-diazaborol-2-yls) as intermediates. Acyclic diborane-
(4)s are accessible by alkali-metal reduction of the respective
Preliminary investigations have shown that reduction of the 2-
halogeno-1,3,2-diazaboroles tBuNCH᎐CHN(tBu)BX (4b: X =
᎐
Cl; 4c: Br; 4d: I) with sodium–potassium alloy in n-hexane
is very sluggish and affords 2-hydro-1,3,2-diazaborole tBu-
NCH᎐CHN(tBu)BH (5) as the only tractable product.10
᎐
DOI: 10.1039/b106128n
J. Chem. Soc., Dalton Trans., 2001, 3459–3464
This journal is © The Royal Society of Chemistry 2001
3459

View Article Online
In order to improve the reducing capability of the alkali
metals we repeated the reduction of 4c in coordinating
solvents such as 1,2-dimethoxyethane (DME) and tetramethyl-
ethylenediamine (TMEDA) (Scheme 2).
2-bromo-1,3,2-diazaborole 4c was complete in less than 2 h and
gave a 1 : 1 mixture of 5 and the 2-benzyl-1,3,2-diazaborole 12.
As it was not possible to separate the materials, compound 12
was independently prepared from 4c and benzyllithium in a
toluene–n-hexane mixture. Diazaborole 12 was obtained as
a colourless solid in 88% yield. For a better understanding, the
reaction was repeated in toluene-d8 and in benzene-d6. Treat-
ment of 4c with the blue solution in toluene-d8 furnished a
1 : 1 mixture of the 2-deutero-1,3,2-diazaborole 5-d1 and the
2-benzyl-1,3,2-diazaborole 12-d7. An analogous reaction in
benzene-d6, however, led to the exclusive production of the
2-protio-1,3,2-diazaborole 5 (Scheme 4).
Scheme 2
The reaction of the 2-bromo-derivative 4c with a potassium
mirror in DME at ambient temperature took 36 h (monitored
by 11B-NMR) and afforded a 2 : 1 : 1 mixture of the 2-methoxy-
1,3,2-diazaborole 6, compound 511 and the diborolyloxane 7.5
Due to their similar volatilities and solubilities, compounds 6
and 7 could not be separated. In order to unambiguously
identify the 2-methoxy derivative 6, this species was synthesised
independently from 4c and sodium methoxide in n-hexane.
The product was isolated as a colourless oil by distillation
(10Ϫ3 Torr, 100–150 ЊC, 66% yield).
The reduction of diazaborole 4c with sodium–potassium
alloy in TMEDA at room temperature came to an end (by
11B-NMR) after 3 d. The major product was the 2-hydro deriv-
ative 5, which was separated in 70% yield by distillation.
The remaining material was an inseparable mixture of three
diazaboroles in equal amounts. Two of these minor products
were spectroscopically identified as compound 8, which was
previously synthesised by reaction of 4c with gaseous dimethyl-
amine,12 and bis(1,3,2-diazaborol-2-yl) 9. As solutions of com-
pound 4c in dimethoxyethane or tetramethylethylenediamine
are stable, the formation of 6 and 8 must involve radical 11,
which results from electron transfer processes (Scheme 3).
Scheme 4
There was no evidence for the formation of 5-d1 or 2-phenyl-
1,3,2-diazaborole-d5 (13). Clearly, the solvent C6D6 did not par-
ticipate as a source of deuterium or of the phenyl substituent,
as was the case in toluene as solvent. In order to rationalize
these results, at least two pathways are conceivable (Scheme 5).
As already depicted in Scheme 3, precursor 4c may be
reduced to radical 11 via radical anion 10. Radical 11 could be
further reduced to the boranide salt 14, which, as a strong
Brønsted base, abstracts a deuteron from the solvent to give
5-d1 and a perdeuterated benzyl anion (path a). The latter might
replace bromide from the starting material 4c to produce 12-d7.
Alternatively, radical 11 could remove a deuterium atom
from the solvent with formation of 5-d1 and a perdeuterated
benzyl radical. Combination of the latter with intermediate 11
would also generate product 12-d7. Usually, such a radical
mechanism involving toluene also affords considerable
amounts of dibenzyl by dimerisation of the benzyl radicals.
Here however, neither dibenzyl, nor its perdeuterated form,
could be detected (1H-, 13C-NMR, MS). The reluctance of
benzene and C6D6 to provide hydrogen or deuterium atoms in
Scheme 3
Hydrogen abstraction from the reaction medium converts
radical 11 into the 2-hydro derivative 5. In TMEDA solutions
the latter process is the main reaction, whereas dimerization of
11 accounts for the formation of 9. Obviously under the con-
ditions employed here the generation of diborane(4) 9 is not
a favourable process. We then tried to improve the reducing
capability of the alkali metal by the addition of crown ether.
Here, as in the case of the gallanide 3,4 the eventual formation
of a boranide of type C was envisaged. A stoichiometric
amount of [15]crown-5 was added to the slurry of the sodium–
potassium alloy in toluene. The solution spontaneously became
dark blue and, after 30 min of stirring at 20 ЊC, most of the
metal went into solution. The reaction of this solution with
the reduction of 4c is in line with the increased dissociation
13
ؒ
ؒ
enthalpy C6H6
H
ϩ C6H5 (110 kcal molϪ1
)
versus
13
ϩ C6H5CH2 (85 kcal molϪ1
)
and the
ؒ
ؒ
C6H5CH3
H
decreased CH acidity of benzene C6H6
H
ϩ ϩ C6H5Ϫ (pKa =
43)14 versus that of toluene C6H5CH3
Hϩ ϩ C6H5CH2Ϫ (pKa
= 41).14 In the absence of further experimental evidence, we
favour path (a) over path (b).
3460
J. Chem. Soc., Dalton Trans., 2001, 3459–3464

View Article Online
Scheme 7
analytically pure yellow solid in 38% yield. It is well known that
fused polyarenes (e.g. naphthalene, phenanthrene, etc.) are
more readily reduced than monocyclic arenes (e.g. benzene).16
Accordingly, the reduction of 2-bromo-1,2-dihydro[1,3,2]-
diazaborolo[1,5-a]pyridine 1917 by sodium–potassium alloy in
n-hexane proceeded smoothly to furnish compound 20 as a
yellow crystalline solid in 64% yield. The only boron con-
taining by-product was the 2-hydro derivative 21, as evidenced
by 11B-NMR spectroscopy (Scheme 8).
Scheme 5
Regarding the facile reduction of various halogenoboranes
to diborane(4)s with sodium–potassium alloy, the behaviour of
4a–d observed here is surprising.
In view of the straightforward desulfurization of imidazo-2-
thione to imidazol-2-ylidene by potassium,15 we also envisaged
desulfurisation processes as viable routes to our goal.
The required diazaborolyl sulfanes 15 and 16 were syn-
thesised by the reaction of 4c with a slight excess of sodium
methylthiolate or potassium tert-butylthiolate, respectively, in
n-hexane. Compounds 15 and 16 are formed in 90 and 82%
yield, respectively, as colourless solids (Scheme 6).
Scheme 8
In contrast to this, the employment of tetramethylethylene-
diamine as solvent or the addition of crown ether to the reac-
tion mixture in toluene or in tetrahydropyran invariantly led
to complete deterioration of the boron heterocycle.
The 11B-NMR shifts of the novel 1,3,2-diazaboroles fit well
into a series of known values. Increased shielding was observed
for the monocyclic 1,2-dihydro-1,3,2-diazaboroles tBu-
NCH᎐CHN(tBu)BX as follows: X = CH (δ 26.2) > CH Ph
᎐
3
2
(δ 25.5), {tBuNCH᎐CHN(tBu)B} (δ 25.2) > StBu (δ 21.9) >
᎐
2
{tBuNCH᎐CHN(tBu)B} O (δ 21.1) > F (δ 20.3) ≅ Cl (δ 20.2) >
᎐
2
H (δ 18.9) > Br (δ 16.2) > CN (δ 12).18
The 11B-NMR signal of the 1,2-dihydro[1,3,2]diaza-
borolo[1,5-a]pyridine derivative 20 (δ 23.3) is slightly more
shielded than that of 9 (δ 25.2). A similar trend is obvious for
other monocyclic 1,2-dihydro-1,3,2-diazaboroles and bicyclic
1,2-dihydro[1,3,2]diazaborolo[1,5-a]pyridines.17 Moreover, it is
noteworthy that the 11B-NMR absorptions of 9 and 20 are con-
siderably shifted to higher field when compared with those
of bis(dimethylamino)bis(pyrrolyl)diborane(4) (δ 37.0)9 and
bis(dimethylamino)bis(indoyl)diborane(4) (δ 36.7).9 The high
field shift in 9 and 20 is obviously caused by the incorporation
of the boron atom in an aromatic ring system. The 11B-NMR
signal of diborane(4)s are significantly broadened due to
1J(11B–10B) couplings with w1/2 = 120 (9) and 180 Hz (20). In
Scheme 6
2-Methylthiolato-1,3,2-diazaborole 15 was allowed to react
for 14 d with sodium–potassium alloy (9 : 91) to afford the
2-methyl-1,3,2-diazaborole 17 in 44% yield, in addition to
diazabutadiene 18 (40%) (Scheme 7).
The reduction of compound 16 by potassium–sodium alloy
(91 : 9) was performed in an ultrasound bath and carefully
followed by 11B-NMR. After 10–12 h, precursor 16 was com-
pletely consumed. The slurry obtained was centrifuged and
from the supernatant solution diborane(4) 9 was isolated as an
J. Chem. Soc., Dalton Trans., 2001, 3459–3464
3461

View Article Online
Table 1 Selected bond lengths (Å) and angles (Њ) for compound 20
lengths are 1.429(2) and 1.420(2) Å. In 1,3,2-diazaboroles, the
BN bond lengths range from 1.395(4) to 1.450(2) Å. There
is a significant difference between the endocyclic bond lengths
N(1)–C(1) [1.420(2) Å] and N(2)–C(6) [1.399(2) Å]. The same
effect has previously been observed in molecule 22 [1.416(2) and
B(1)–B(1A)
B(1)–N(2)
C(1)–C(6)
N(1)–C(5)
C(2)–C(3)
C(4)–C(5)
1.697(3)
1.436(2)
1.350(2)
1.381(2)
1.345(2)
1.344(2)
B(1)–N(1)
N(1)–C(1)
N(2)–C(6)
C(1)–C(2)
C(3)–C(4)
1.452(2)
1.420(2)
1.399(2)
1.425(2)
1.437(2)
1.395(2) Å]. In [tBuNCH᎐CHN(tBu)BCN Cr(CO)5], these
᎐
CN separations are 1.373(6) and 1.389(7) Å, and have been
regarded as multiple bonds.11 The calculated value for a
N(1)–B(1)–N(2)
N(2)–C(6)–C(1)
B(1)–N(1)–C(1)
C(3)–C(4)–C(5)
B(1A)–B(1)–N(1)
C(6)–N(2)–C(7)
103.0(1)
B(1)–N(2)–C(6)
N(1)–C(1)–C(6)
N(1)–C(5)–C(4)
C(2)–C(3)–C(4)
B(1A)–B(1)–N(2)
109.5(1)
Csp –Nsp single bond is 1.45 Å.20 In 20, this is only valid for
N(2)–C(6). The bridging CN bond is lengthened and compares
2
2
110.3(1)
109.7(1)
120.2(1)
123.8(1)
118.9(1)
107.4(1)
121.4(1)
119.9(1)
133.2(1)
2
2
with one of the exocyclic single bonds Nsp –Csp measured in
2,6-Me C H NCH᎐CHN(2,6-Me C H )BH [1.421(3) Å].11 In
᎐
2
6
3
2
6
3
the six-membered ring of 20, the N(1)–C(5) bond length of
1.381(2) Å also shows multiple bonding. As previously
observed in 22, the carbon–carbon distances in the six-
membered ring of 20 alternate. Double bonds are C(4)–C(5)
[1.344(2) Å], C(2)–C(3) [1.345(2) Å], whereas the bond order
of the longer bonds C(1)–C(2) [1.425(2) Å] and C(3)–C(4)
[1.437(2) Å] is close to unity.
Conclusion
In marked contrast to the facile alkali metal reduction of
a variety of acyclic bis(amino)halogenoboranes to the corre-
sponding diborane(4)s, such a process in the chemistry of
2-halogeno-2,3-dihydro-1H-1,3,2-diazaboroles 4b–d is not
favourable. Treatment of 2-bromo derivative 4c with K–Na
alloys in solvents such as DME, TMEDA, and toluene led
to 1,3,2-diazaboroles with substituents at the boron atom
that were abstracted from the respective reaction media.
Predominantly, the formation of the 2-hydro-1,3,2-diazaborole
5 was observed. Interestingly, in toluene and toluene-d8, equi-
molar amounts of 5 and the 2-benzyl derivative 12 were
obtained, which can be tentatively interpreted with the inter-
mediacy of a salt with the sought-after boranide anion
Fig. 1 ORTEP drawing of the structure of compound 20. Thermal
ellipoids are shown at the 50% probability level.
{tBuNCH᎐CHN(tBu)B}Ϫ. The latter would be the boron
᎐
analogue of Schmidbaur’s gallanide anion D. The bi(1,3,2-
diazaborol-2-yl) 9 was formed as a minor product from 4c
and K–Na alloy in TMEDA. A more effective route to this
diborane(4) is based on the K–Na reduction of the 2-tert-
butylthiolato-1,3,2-diazaborole 16 in n-hexane. Such difficulties
were not encountered in the reduction of the 1-bromo-2-tert-
butyl-1,2-dihydro[1,3,2]diazaborolo[1,5-α]pyridine 19, which
smoothly afforded the diborane(4) derivative 20. These
observations may be rationalized by the availabity of low-lying
unoccupied orbitals and reduced aromaticity in the fused
diazaborole system. Similar observations were previously
made with related 2-bromo-benzo-1,3,2-diazaboroles.21 These
diboranes(4) have to be subjected to reduction processes under
varying conditions to eventually accomplish the synthesis of
the as yet elusive metal boranides.
bis(dimethylamino)bis(pyrrolyl)diborane(4) and bis(dimethyl-
amino)bis(indoyl)diborane(4), values of 320 and 400 Hz were
measured.9
In the 1H-NMR spectra of 6, 9, 15 and 16, the resonances for
the ring protons of the HC᎐CH unit were observed in the
᎐
narrow range between δ 6.12 (X = OCH3) and 6.43 (X = StBu),
which is characteristic for monocyclic 1,3,2-diazaboroles of
this type.18 In the 13C{1H}-NMR spectra, the 13C nuclei of the
CH᎐CH unit in tBuNCH᎐CHN(tBu)BX give rise to a singlet in
᎐
᎐
the region between δ 109.9 (X = F) and 114.8 (X = StBu),18
which includes compounds 6 (X = OCH3; δ 111.5), 9 (X =
tBuNCH᎐CHN(tBu)B; δ 113.1), 12 (X = CH Ph; δ 113.2), 15
᎐
2
(X = SCH3; δ 113.4) and 16 (X = StBu; δ 114.8).
X-Ray structural analysis of compound 20
For a full characterization of one of the diborane(4)s, an X-ray
structural analysis of compound 20 was performed. Light
yellow single crystals were grown from n-hexane at Ϫ30 ЊC.
Selected bonding parameters are compiled in Table 1. The
structure (Fig. 1) features a molecule which may be described as
the combination of two planar 1,3,2-diazaborolo[1,5-a]pyridine
fragments via a boron–boron single bond of 1.697(3) Å. This
distance falls into the region of 1.674(4) and 1.723(4) Å, as
determined in Cp(CO)2RuB(NMe2)B(NMe2)Br19 and bis(di-
methylamino)bis(indoyl)diborane(4),9 respectively. Both planes
enclose a dihedral angle of 85.6Њ. The bonding parameters
within the fused rings compare well with recently characterized
2,3-dihydro[1,3,2]diazaborolo[1,5-a]pyridine 22.17
Experimental
General procedures
All manipulations were performed under an atmosphere of
dry argon using standard Schlenk techniques. All solvents
were dried by standard methods and freshly distilled prior to
use. The compounds tBuNCH᎐CHN(tBu)BBr (4c)5 and tBu-
᎐
N–CH᎐C–CH᎐CH–CH᎐CH–N–B–Br (19)17 were prepared
᎐
᎐
᎐
according to literature methods. MeSH, tBuSH, NaSMe, C7D8
and [15]crown-5 were purchased commercially. NMR spectra
were recorded on a Bruker AM Avance DRX 500 (1H, 11B,
13C) using SiMe4 and BF3ؒOEt2 as external standards. The
IR spectra were recorded on Bruker FT-IR Vector 22 instru-
ments, and mass spectra on a VG Autospec sector field mass
spectrometer (Micromass).
The distances B(1)–N(2) [1.436(2) Å] and B(1)–N(1)
[1.452(2) Å] indicate multiple bonding, with B(1)–N(1) being
slightly longer than B(1)–N(2). In 22, the corresponding bond
3462
J. Chem. Soc., Dalton Trans., 2001, 3459–3464

View Article Online
Preparations
Na alloy in 3 ml of toluene-d8 to give a dark blue solution.
Then, a solution of 4c (0.25 g, 0.9 mmol) in toluene-d8 (5 ml)
was added. After 30 min of stirring at ambient temperature,
NMR and mass spectra of the reaction mixture were taken, 1H-
NMR (toluene-d8) 5-d1: δ 1.30 (s, 18H, tBu), 6.34 (s, 2H, NCH);
12-d7: δ 1.26 (s, 18H, tBu), 6.35 (s, 2H, NCH). A singlet at δ 3.47
was attributed to [15]crown-5. 13C{1H}-NMR (toluene-d8)
5-d1: δ 31.8 [s, C(CH3)3], 51.4 [s, C(CH3)3], 114.1 [s, NCH];
12-d7: δ 31.9 [s, C(CH3)3], 53.1 [s, C(CH3)3], 113.1 [s, NCH]. A
singlet at δ 71.0 was assigned to [15]crown-5. 11B{1H}-NMR
(toluene-d8): 5-d1: δ 18.4 (s); 12-d7: δ 25.5 (s). MS/EI (70eV)
5-d1: m/z = 181 (Mϩ, < 5%); 12-d7: m/z = 277 [Mϩ, 31%], 221
Reaction of 4c with potassium in DME. In a flask a mirror of
0.25 g (6.4 mmol) potassium metal was prepared. A solution
of 2-bromo-1,3,2-diazaborole 4c (0.80 g, 3.1 mmol) in 50 ml
of 1,2-dimethoxyethane (DME) was added and stirred for
18 h at room temperature. Solvent and volatile components
were removed in vacuo (10Ϫ3 mbar) to give an oily yellow
residue. Spectroscopically the 1,3,2-diazaboroles 5 (δ 11B =
18.9), 6 (δ 11B = 22.0), and 7 (δ 11B = 21.1) were identified in a
molar ratio of 1 : 2 : 1. Attempts to separate these compounds
failed.
[Mϩ Ϫ (CH ) C᎐CH , 9%].
᎐
3
2
2
tBuNCH᎐CHN(tBu)BOCH (6). A solution of 4c (1.20 g,
᎐
3
4.7 mmol) in n-hexane (30 ml) was slowly added to a chilled
slurry (Ϫ20 ЊC) of sodium methanolate in 10 ml of n-hexane.
The mixture was warmed to ambient temperature and filtered.
Solvents was removed from the filtrate to afford a yellow oil,
which was purified by distillation (10Ϫ3 mbar, 100–150 ЊC).
Compound 6 was isolated as a colourless oil, yield 0.65 g (66%)
(found: C, 62.74; H, 10.73; N, 13.32; C11H23BN2O requires C,
tBuNCH᎐CHN(tBu)BSCH (15).
A sample of sodium
᎐
3
methylthiolate (0.50 g, 7.1 mmol) was combined with a solution
of 4c (1.30 g, 5.0 mmol) in n-hexane (50 ml) and stirred for 10 d
at room temperature. The hexane phase was decanted from
a colourless solid and then concentrated to the beginning
of crystallization. Storing the solution at Ϫ30 ЊC afforded 15
(1.02 g, 90% yield) as colourless crystals (found: C, 58.13;
H, 9.96; N, 12.19; C11H23BN2S requires C, 58.41; H, 10.25; N,
1
62.88; H, 11.03; N, 13.33%). H-NMR (C6D6): δ 1.34 (s, 18H,
tBu), 3.49 (s, 3H, OCH3), 6.12 (s, 2H, NCH). 13C{1H}-NMR
(C6D6): δ 31.4 [s, C(CH3)3], 51.7 [s, C(CH3)3], 55.3 [s, OCH3],
111.5 [s, NCH]. 11B{1H}-NMR (C6D6): δ 22.1 (s). MS/EI
(70eV): m/z = 210 (Mϩ, 50%).
1
12.38%). H-NMR (C6D6): δ 1.50 (s, 18H, tBu), 1.98 (s, 3H,
SCH3), 6.35 (s, 2H, NCH). 13C{1H}-NMR (C6D6): δ 15.6 [s,
SCH3], 31.9 [s, C(CH3)3], 53.9 [s, C(CH3)3], 113.4 [s, NCH].
11B{1H}-NMR (C6D6): δ 23.3 (s). MS/EI (70eV): m/z = 226 [Mϩ,
37%], 114 [Mϩ Ϫ 2 (CH ) C᎐CH , 100%].
᎐
3
2
2
Reaction of 4c with Na–K in TMEDA. K–Na alloy (0.50 g;
58 : 42) was suspended in TMEDA (40 ml) with the aid
of an ultrasound bath to give a dark blue slurry after 30 min.
A solution of 4c (1.04 g, 4.0 mmol) in 20 ml of TMEDA
was added, and the mixture was stirred for 3 d at 20 ЊC. The
solution was decanted and freed from TMEDA at ambient
temperature and 8 mbar, to give a yellow waxy residue. 2-
Hydro-1,3,2-diazaborole 5 (0.50 g, 70%) was separated by dis-
tillation. The oily residue consists of an inseparable mixture
of 7, 8 and 9, as evidenced by NMR spectroscopy and mass
spectrometry.
tBuNCH᎐CHN(tBu)BStBu (16). Potassium tert-butylthio-
᎐
late (0.71 g, 5.5 mmol) was added at 20 ЊC to a solution of 4c
(1.29 g, 5.0 mmol) in n-hexane (60 ml) and vigorously stirred for
12 h. The solution was filtered, and the filtrate was evaporated
to dryness. Compound 16 was obtained as a waxy solid, yield
1.09 g (82%) (found: C, 62.53; H, 10.91; N, 10.47; C14H29BN2S
requires C, 62.68; H, 10.90; N, 10.44%). 1H-NMR (C6D6):
δ
1.44 [s, 9H, tBuS], 1.55 (s, 18H, tBuN), 6.43 (s,
2H, NCH). 13C{1H}-NMR (C6D6): δ 32.8 [s, NC(CH3)3], 34.8
[s, SCCH3], 45.2 [s, SC(CH3)3], 54.5 [s, NC(CH3)3], 114.8 [s,
NCH]. 11B{1H}-NMR (C6D6): δ 21.9 (s). MS/EI (70eV):
m/z = 268 [Mϩ, 16%].
Reaction of 4c with Na–K and [15]crown-5 in toluene. To a
slurry of K–Na alloy (0.20 g, 58 : 42) in toluene (10 ml) a
solution of 2.40 g (10.9 mmol) of [15]crown-5 in toluene (20 ml)
was added to afford a dark blue solution. After 20 min, a solu-
tion of 4c (1.41 g, 5.4 mmol) in toluene (30 ml) was added
dropwise. Stirring at room temperature was continued for
another hour. After the sedimentation of a dark blue solid had
ceased, the yellow solution was decanted and evaporated to
dryness. According to NMR and MS evidence, the residue
(3.10 g) was a 1 : 1 mixture of 1,3,2-diazaboroles 5 and 12 which
could not be separated from the crown ether.
Reaction of 15 with Na–K. K–Na alloy (0.20 g; 91 : 9) was
added to a solution of 15 (0.55 g, 2.4 mmol) in n-hexane
(30 ml), and the slurry was stirred at ambient temperature for
14 d. The course of the reaction was monitored by 11B NMR
spectroscopy. Then, hexane (20 ml) was added to the reaction
mixture and the liquid phase decanted. Solvent was removed
from the clear solution in vacuo (10 mbar) to give 0.40 g of a
yellow solid.
According to NMR spectra, this residue is a mixture of the
2-methyl-1,3,2-diazaborole 17 and 1,2-bis(tert-butyl-imino)-
ethane 18 in a molar ratio of 4 : 1. Sublimation at 40 ЊC and 7
mbar afforded 0.16 g of pure 18 (40%) and 0.21 g (44%) of
yellow solid 17. The spectra of the products are identical with
analytic samples.
tBuNCH᎐CHN(tBu)BCH Ph (12). A solution of benzyl-
᎐
2
lithium (0.70 g, 7.1 mmol) in toluene (20 ml) was combined at
0 ЊC with a solution of 4c (1.31 g, 5.1 mmol) in n-hexane (40 ml)
and stirred for 2 h at 20 ЊC. Solvent and volatile compounds
were removed in vacuo and the remaining solid was dissolved in
n-hexane (80 ml). The solution was filtered and the filtrate was
evaporated to dryness to give compound 12 as a colourless
solid, yield 1.21 g (88%) (found: C, 75.57; H, 10.02; N, 9.63;
{tBuNCH᎐CHN(tBu)B
} (9). A slurry of 0.5 g of K–Na alloy
2
᎐
1
(91 : 9), 16 (0.81 g, 3.0 mmol) and 50 ml of n-hexane was stirred
for 10–12 h at 20 ЊC in an ultrasound bath. At the end of the
reduction (monitered by
C17H27BN requires C, 75.56; H, 10.07; N, 10.37%). H-NMR
(C6D6): δ 1.27 (s, 18H, tBu), 3.05 (s, 2H, CH2), 6.41 (s, 2H,
11B-NMR), the solids were separated
3
NCH), 7.02 (t, JH,H = 7.7 Hz, 1H, p-H-phenyl), 7.13–7.20 (m,
4H, o,m-H-phenyl). 13C{1H}-NMR (C6D6): δ 31.9 [s, C(CH3)3],
53.1 [s, C(CH3)3], 113.2 [s, NCH], 125.0 (s, p-C-phenyl), 128.4
(s, m-C-phenyl), 131.0 (s, o-C-phenyl), 143.2 (s, i-C-phenyl).
11B{1H}-NMR (C6D6): δ 25.5 (s). MS/EI (70eV): m/z = 270 [Mϩ,
100%].
from the liquid phase by centrifugation. The decanted solution
was evaporated to dryness at 10
Ϫ3 mbar to yield 0.41 g (38%) of
compound 9 (found: C, 66.61; H, 11.31; N, 15.49; C20
H40B2N4
requires C, 67.07; H, 11.26; N, 15.64%). 1H-NMR (C6D6):
δ 1.36 (s, 36H, tBu), 6.37 (s, 4H, NCH). 13
C{1H}-NMR (C6D6):
δ 31.7 [s, C(CH3)3], 53.1 [s, C(CH3)3], 113.1 [s, NCH]. 11B{1H}-
NMR (C6D6): δ 25.2 (s, w1/2 = 120 Hz). MS/EI (70eV): m/z = 358
[Mϩ, 58%].
Reaction of 4c with Na–K and [15]crown-5 in toluene-d8.
[15]crown-5 (0.42 g, 1.9 mmol) was added to the slurry of K–
J. Chem. Soc., Dalton Trans., 2001, 3459–3464
3463

View Article Online
Table 2 Crystallographic data for compound 20
Acknowledgements
Empirical formula
Mr
T /K
Space group
C20H28B2N4
346.08
223
Pbcn
This work was financially supported by the Deutsche
Forschungsgemeinschaft, Bonn, Germany and by the Fonds
der Chemischen Industrie, Frankfurt/Main, Germany, which
is gratefully acknowledged.
Crystal system
Orthorhombic
a/Å
b/Å
18.582(5)
6.952(2)
References
c/Å
15.511(5)
2003.8(1)
V/Å3
1 (a) A. J. Arduengo III, R. L. Harlow and M. Kline, J. Am.
Chem. Soc., 1991, 113, 361; (b) A. J. Arduengo III, H. V. R. Dias
and M. Kline, J. Am. Chem. Soc., 1992, 114, 5530; Review:
W. A. Herrmann and C. Köcher, Angew. Chem., 1997, 109, 2256;
W. A. Herrmann and C. Köcher, Angew. Chem., Int. Ed. Engl., 1997,
36, 2162.
2 (a) M. Denk, R. Lennon, R. Hayashi, R. West, A. V. Belyakov,
H. R. Verne, M. Wagner and N. Metzler, J. Am. Chem. Soc., 1994,
116, 2691; (b) Review: B. Gehrhus and M. F. Lappert, J. Organomet.
Chem., 2001, 617–618, 209.
3 A. Sundermann, M. Reiher and W. W. Schoeller, Eur. J. Inorg.
Chem., 1998, 305.
4 E. S. Schmidt, A. Jokisch and H. Schmidbaur, J. Am. Chem. Soc.,
1999, 121, 9758.
5 L. Weber, E. Dobbert, H.-G. Stammler, B. Neumann, R. Boese and
D. Bläser, Chem. Ber. Recueil, 1997, 130, 705.
6 L. Weber, E. Dobbert, R. Boese, M. T. Kirchner and D. Bläser,
Eur. J. Inorg. Chem., 1998, 1145.
7 D. Loderer, H. Nöth, H. Pommerening, W. Rattay and H. Schick,
Chem. Ber., 1994, 127, 1605 and references cited therein.
8 (a) W. J. Grisby and P. Power, Chem. Eur. J., 1997, 3, 368;
(b) A. Moezzi, M. M. Olmstead and P. P. Power, J. Am. Chem. Soc.,
1992, 114, 2715; (c) A. Moezzi, R. A. Bartlett and P. P. Power,
Angew. Chem., 1992, 104, 1075; A. Moezzi, R. A. Bartlett and
P. P. Power, Angew. Chem., Int. Ed. Engl., 1992, 31, 1082.
9 H. Nöth, J. Knizek and W. Ponikwar, Eur. J. Inorg. Chem., 1999,
1931.
10 E. Dobbert, PhD Thesis, University of Bielefeld, Germany, 1999.
11 L. Weber, E. Dobbert, H.-G. Stammler, B. Neumann, R. Boese and
D. Bläser, Eur. J. Inorg. Chem., 1999, 491.
12 L. Weber, M. Schnieder, T. C. Maciel, H. B. Wartig, M. Schimmel,
R. Boese and D. Bläser, Organometallics, 2000, 19, 5791.
13 J. March, Advanced Organic Chemistry, Wiley, New York, 1985,
p. 166.
14 J. March, Advanced Organic Chemistry, Wiley, New York, 1985,
p. 222.
15 N. Kuhn and T. Kratz, Synthesis, 1993, 561.
16 (a) C. Näther, H. Bock, Z. Havlas and T. Hauck, Organometallics,
1998, 71, 4707; (b) H. Bock, Z. Havlas, D. Heß and C. Näther,
Angew. Chem., 1998, 110, 518; H. Bock, Z. Havlas, D. Heβ and
C. Näther, Angew. Chem., Int. Ed., 1998, 37, 502.
17 L. Weber, M. Schnieder, R. Boese and D. Bläser, J. Chem. Soc.,
Dalton Trans., 2001, 378.
18 Review: L. Weber, Coord. Chem. Rev., 2001, 215, 39.
19 H. Braunschweig, M. Koster and R. Wang, Inorg. Chem., 1999, 38,
415.
20 A. N. Chernega, A. V. Ruban, V. D. Romanenko, L. N. Markovskii,
A. A. Korkin, M. Y. Antipin and Y. T. Struchkov, Heteroatom.
Chem., 1991, 2, 229.
Z
4
µ(Mo-Kα)/mmϪ1
0.068
11876
2430 ([R(int) = 0.0391])
2430/175
Reflections collected
Independent reflections
Reflections with I > 2σ(I )/parameters
Final R1, wR2 [I > 2σ(I )] (all data)
0.0429/0.1135 0.0659/0.1295
{tBuN–CH᎐C–CH᎐CH–CH᎐CH–N–B–} (20). A solution
᎐
᎐
᎐
2
of compound 19 (1.62 g, 6.4 mmol) in n-hexane (100 ml) was
combined with 0.52 g of K–Na alloy (58 : 42) and the resulting
slurry was vigorously stirred at room temperature (6 h). After
the sedimentation of the solids had ceased, the light yellow
liquid phase was decanted and the solvent removed in vacuo.
Traces of the 2-hydro derivative 21 were removed by distillation
with a hot air gun. The resulting solid residue was dissolved in
n-hexane (50 ml), filtered, and the filtrate was stored at Ϫ30 ЊC.
Compound 20 separated as light yellow crystals, yield 0.71 g
(64%). Due to the air and moisture sensitivity of the product,
no reliable elemental analyses could be obtained from the
Microanalytical Laboratory of our department. 1H-NMR
(C6D6): δ 1.29 (s, 18H, tBu), 5.54 (t, JH,H = 6.7 Hz, 2H,
CH), 6.08–6.14 (m, 2H, CH), 6.55 (s, 2H, tBuN-CH), 6.81
(d, JH,H = 9.5 Hz, 2H, CH), 6.99 (d, JH,H = 7.0 Hz, 2H, CH).
1H-NMR (CDCl3): δ 1.39 (s, 18H, tBu), 5.65 (t, 3JH,H = 6.3 Hz,
2H, CH), 6.15–6.18 (m, 1H, CH), 6.66 (s, 2H, tBuN–CH), 6.80
3
3
3
3
3
(d, JH,H = 9.4 Hz, 2H, CH), 7.10 (d, JH,H = 6.3 Hz, 2H, CH).
13C{1H}-NMR (CDCl3): δ 32.2 [s, C(CH3)3], 53.5 [s, C(CH3)3],
105.5 [s, tBuN–CH], 105.9 [s, CH], 118.1 [s, CH], 118.2 [s, CH],
128.5 [s, NCH᎐C], 131.5 [s, CH]. 11B{1H}-NMR (CDCl3):
᎐
δ 23.3 [s, w1/2 = 180 Hz]. MS/EI (70eV): m/z = 346 [Mϩ, 100%],
290 [Mϩ Ϫ (CH ) C᎐CH , 16%].
᎐
3
2
2
Crystal structure determination of compound 20
The data were collected with a Siemens CCD SMART100
detector system on a three-axis platform using graphite-
monochromated Mo-Kα radiation. Software for structure
solution/refinement: Bruker AXS SHELXTL Ver. 5.10 DOS/
WIN95/NT, treatment of hydrogen atoms as riding groups in
idealized positions. All other data are summarized in Table 2.
CCDC reference number 167157.
See http://www.rsc.org/suppdata/dt/b1/b106128n/ for crystal-
lographic data in CIF or other electronic format.
21 L. Weber, H. B. Wartig, H. G. Stammler and B. Neumann, Z. Anorg.
Allg. Chem., in press.
3464
J. Chem. Soc., Dalton Trans., 2001, 3459–3464