Triaminoboranes and their metallation to N-lithiotriaminoboranes

Article abstract of DOI:10.1002/1099-0682(200205)2002:5<1132::AID-EJIC1132>3.0.CO;2-U

In order to study the competition between deprotonation, borate formation, and B-N bond cleavage, six triaminoboranes bearing 1-3 NH functions were treated with butyllithium. The 2-anilino-1,3,2-diazaborolidine 1 was cleanly deprotonated, but the resultin

Full text of DOI:10.1002/1099-0682(200205)2002:5<1132::AID-EJIC1132>3.0.CO;2-U

FULL PAPER
Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes[‡]
Ulrike Braun,[a][‡‡] Tassilo Habereder,[a] Heinrich Nöth,*[a] Hans Piotrowski,[a] and
Marcus Warchhold[a]
Keywords: Boranes / Deprotonation / Lithium / Nitrogen / X-ray diffraction
In order to study the competition between deprotonation,
borate formation, and B−N bond cleavage, six triaminobor-
anes bearing 1−3 NH functions were treated with butylli-
borane (3) with LiBu in 1:1 and 1:2 ratios showed that the
deprotonation was accompanied by the formation of anilino-
(butyl)borates. Treatment of the dilithio derivative of 3 with
thium. The 2-anilino-1,3,2-diazaborolidine 1 was cleanly de- Me3SnCl afforded B[N(SnMe3)Ph]3 (23), a result of substitu-
protonated, but the resulting N-lithio compound could not be ent exchange. Tris(2-pyridylamino)borane (4), with a planar
crystallized. However, as a result of slow hydrolysis, the li-
skeleton, reacted with BuLi both by deprotonation and by
thium 2-oxido derivative 7 was isolated in low yield as a di- B−N bond cleavage. The metallated triaminoborane crystal-
meric ether solvate. Double deprotonation of 2-(dimethylam-
ino)-1,3,2-benzodiazaborolidine (9) was accompanied by
B−N bond cleavage. The dilithio derivative reacted with
Me3SiCl to produce 2-dimethylamino-1,3-bis(trimethylsilyl)-
1,3,2-benzodiazaborolidine (11). Similarly, dianilino(diphen-
lized as dimeric Li4B(NR*)3NHR*(THF)3 (18; R* = 2-pyridyl).
Tris(8-quinolylamino)borane (5) can also be triply lithiated
according to 11B and 7Li NMR spectroscopic data. X-ray de-
termination of the structures of compounds 1−3 showed the
absence of N−H···N hydrogen bonding, which seems to be
ylamino)borane (2) was readily deprotonated to the insoluble responsible for the planar structures of compounds 4 and 5.
dilithium compound. B−N bond cleavage also occurred, how-
ever, and the solvates Ph2NLi·dme (13) and Ph2NLi·2py (14)
were isolated. These are dimeric, the former with a planar
Li2N2 ring, the latter possessing an N−Li−N−Li chain struc-
ture. Triply metallated trianilinoborane (15) crystallized as di-
meric B(NLiPh)3·LiCl·(OEt2)3 (15A). Treatment of trianilino-
The planarity of the BN3 units in 3 and 4 was retained in the
trilithiated compounds 15A and 18, as ascertained by X-ray
structure determination.
(
Wiley-VCH Verlag GmbH, 69451 Weinheim, Germany,
2002)
Introduction
(pmdeta ϭ pentamethyldiethylenetriamine),[14,15] were ob-
tained and their structures determined. In comparison with
lithium organylamides[16,17] the tendency of the lithium (di-
organylboryl)amides to associate through LiϪN bridges did
not seem to be significantly altered. Dimetallation of bis-
(tert-butylamino)organylboranes was reported by Meller et
al.,[6] whilst it is only recently that a threefold lithiation of
the borazine (PhBϪNH)3 was successfully achieved, re-
sulting in the compound (PhBNLi)3·(LiPh)3·(THF)3, char-
acterized by X-ray diffraction analyses.[18]
The metallation of aminoboranes R3ϪnB(NHRЈ)n by or-
ganolithium reagents LiR to give (lithioamino)boranes acts
in competition with the formation of aminoborates
Li[R4ϪnB(NHRЈ)n] and/or BϪN bond cleavage. This was
demonstrated by us in 1980,[2] after the first successful
lithiation of a cyclic aminoborane.[3] As expected, (lithioam-
ino)boranes are useful reagents for the syntheses not only of
other metallated aminoboranes[2,4,5Ϫ7] by transmetallation
reactions, but also for syntheses of diborylamines[8Ϫ10] and
BN heterocycles.[5,6,11] So far, only a few structures of (li-
thioamino)boranes have been determined: the ionic [Li(-
OEt2)3][mes2BϭNϭBmes2],[12] the dimeric Li(OEt2)2-
NmesϪBmes2,[7] and the solvent-free trimeric (lithioam-
ino)-9-borabicyclonane.[13] In the presence of chelating li-
gands, monomeric molecular (lithioamino)boranes, such as
tBu2B(tBuN)Li·tmen or (Me3Sn)NϪBC8H14Li·pmdeta
N-Lithiation of triaminoboranes offers a unique chance
to study the influence of electronic and steric effects. One
might expect that the metallation of triaminoboranes of
type (R2N)2BϪ(NHR) would proceed more readily than
that of R2BNHR,[5] as the amino groups make the boron
atom less prone to attack by a nucleophile, thereby promot-
ing the deprotonation of the NH group. On the other hand,
double and triple deprotonation of aminoboranes
B(NHR)3 would probably become more difficult as the
number of deprotonation steps increased, because introduc-
tion of an increasing number of lithium atoms would be
expected to result in highly associated molecular aggregates
through LiϪN coordination. For this reason we synthesized
[‡]
Chemistry of Boron, 250. Part 249: Ref.[1]
[‡‡] Part of the PhD Thesis of U. Braun, University of Munich,
2000.
[a]
Department of Chemistry, University of Munich,
Butenandtstraße 5Ϫ13, 81377 München, Germany
1132
WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002 1434Ϫ1948/02/0505Ϫ1132 $ 20.00ϩ.50/0 Eur. J. Inorg. Chem. 2002, 1132Ϫ1145

Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes
FULL PAPER
a number of new triaminoboranes possessing NHR groups, and tris(8-aminoquinolyl)borane (5) were prepared by the
studied their metallation with BuLi, and characterized their same procedure as shown in Equation (3).
N-lithio derivatives.
Spectra
New Triaminoboranes
The NMR spectra (1H, 11B, 13C, and 14N) of compounds
1Ϫ5 were fully compatible with the suggested structures
Synthesis
In order to study the competition between N-metallation,
BϪN bond cleavage, and borate formation, we first chose
an aminoborane of type (R2N)2BNHR, with only one
NϪH bond. The molecule of choice was 2-anilino-1,3-di-
phenyl-1,3,2-diazaborolidine (1), obtained by treatment of
2-chloro-1,3-diphenyl-1,3,2-diazaborolidine with aniline in
the presence of triethylamine as shown in Equation (1).
(e.g., chemically and magnetically equivalent anilino and
amino heterocyclic groups). As was to be expected for tri-
aminoboranes, the 11B chemical shifts varied very little and
were in the range typical of this class of compounds. The
line widths, however, varied significantly. The sharpest sig-
nal was found for tris(2-pyridylamino)borane (4; h1/2
ϭ
220 Hz) and the broadest for the quinoline derivative 5
(450 Hz). Only a single broad 14N resonance, at δ ϭ Ϫ323,
was found for compound 2, and one at δ ϭ Ϫ292 for com-
pound 3. On the other hand, the two types of N atoms in
compounds 4 and 5 were readily recognizable. The reson-
ances were at δ ϭ Ϫ106 and Ϫ79, respectively, for the het-
erocyclic N atoms, while those of the amino groups were
observed at δ ϭ Ϫ310 and Ϫ327. This was consistent with
15N NMR spectroscopic data of aminoboranes[24] and N-
heterocycles.[25]
(1)
(2)
1
Interestingly, the H resonances for the NH protons of
the anilino groups were found as signals at δ ϭ 4.65Ϯ0.14,
while those for 4 and 5 were observed at δ ϭ 9.85 and 8.15,
respectively. We attribute this to intramolecular NϪH···N
interactions in the latter two compounds, and this was sup-
ported by the NH stretching vibrations found for these two
aminoboranes at 3293 and 3340 cmϪ1. In 1 to 3, these vi-
brations were observed at higher wavenumbers (3384, 3387,
and 3384 cmϪ1). Obviously, the intramolecular NϪH···N
interaction was weaker in 5 than in 4. This is consistent
with the molecular structures of these compounds as deter-
mined by X-ray structure analysis.
(3)
Molecular Structures
The preparation of dianilino(diphenylamino)borane (2)
by aminolysis of B(NMe2)3 with diphenylamine (1:1 ratio),
The triaminoboranes 1Ϫ5 were characterized by X-ray
followed by aminolysis with aniline, was unsuccessful; we structure analysis. Compound 1 crystallizes in the mono-
observed no reaction at reflux temperature between the tris- clinic system, space group I2/a and Z ϭ 8. Figure 1 shows
(dimethylamino)borane and diphenylamine, although sev- its molecular structure. The three BϪN bonds are of equal
˚
eral successful syntheses of unsymmetrically substituted tri- lengths. Their common value (1.435 A) corresponds very
aminoboranes by transamination reactions have been re- well with BϪN bond lengths observed for other triamino-
ported.[19,20] The low basicity of diphenylamine (pKA
ϭ
boranes.[26] All the N atoms and the B atom reside in planar
0.9)[21,22] and steric effects may be responsible for the failure environments, but because of the incorporation of the
of this reaction, in spite of the fact that the resulting di- boron atom into a five-membered ring system a sharp
methylamine can readily be removed to drive the process NϪBϪN bond angle of 107.4(1)° results; this fits nicely for
towards the side of the transamination products. More sur- a five-membered ring system. Consequently, the other two
prising was the observation that LiNPh2 did not react with NϪBϪN bond angles are wider than 120°. Typical for the
ClϪB(NMe2)2 to produce Ph2NϪB(NMe2)2, either in THF structure of 1 is the propeller-like arrangement of the
or in toluene. However, dichloro(diphenylamino)borane re- phenyl groups, with interplanar angles to the BN3 plane of
acted with aniline in the presence of triethylamine accord- 31.6, 41.7, and 52.1°, the last being that made by the anilino
ing to Equation (2) to give Ph2NB(NHPh)2 (2) in 71% yield. group, which is more twisted than the other two phenyl
Trianilinoborane (3) was readily obtained from aniline groups. In the crystal there are neither intra- nor intermol-
and B(NMe2)2,[19] whilst tris(2-pyridylamino)borane (4)[23] ecular NϪH···N hydrogen bonds.
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145
1133

U. Braun, T. Habereder, H. Nöth, H. Piotrowski, M. Warchhold
FULL PAPER
Trianilinoborane (3) crystallizes in the monoclinic sys-
tem, space group P21/c, Z ϭ 4. Its molecular structure is
shown in Figure 3. Although one might expect that the
three BϪN bonds would be of equal lengths, one finds that
this holds only for two of them (B1ϪN1 and B1ϪN2, aver-
˚
age 1.430 A), while the B1ϪN3 bond is slightly but signific-
˚
antly longer [1.442(2) A]. As might be expected, the boron
atom is sp2-hybridized; the sum of its bond angles is 360°.
As was also the case for the triaminoborane 2, we observe
BϪNϪC bond angles wider than the expected 120°. The
three phenyl groups of 3 are arranged in a propeller-like
fashion, and this has the effect that the NϪH groups are
transoid to each other. There are, however, no intermolecu-
lar NϪH···N interactions.
Figure 1. Molecular structure of 1 in ORTEP presentation; selected
bond lengths and bond angles: B1ϪN1 1.436(2), B1ϪN2 1.437(2),
B1ϪN3 1.433(2), N1ϪC9 1.409(2), N2ϪC3 1.402(2), N3ϪC15
˚
1.400(2), N1ϪC1 1.473(2), N2ϪC2 1.471(2), C1ϪC2 1.523(2) A;
C1ϪN1ϪB1 109.0(1), C3ϪN2ϪB1 131.2(1), B1ϪN2ϪC2 108.9(1),
C3ϪN2ϪC2 119.5(1), B1ϪN3ϪC15 127.5(1), N3ϪB1ϪN2
125.1(1), N1ϪB1ϪN2 107.4(1)°; torsion angle: N1ϪC1ϪC2ϪN2
Ϫ26.8°
Crystals of dianilino(diphenylamino)borane (2) are tri-
¯
clinic, space group P1 with Z ϭ 2. Figure 2 shows its mo-
lecular structure. In contrast to those in 1, its BϪN bond
lengths differ significantly. The BϪN bond of the di-
˚
phenylamino group [1.470(3) A] is longer than the BϪN
˚
bonds to the anilino groups (average 1.425 A). Since the
angle between the BN3 plane and the C1N1C2 unit of the
diphenylamino group is at 26.6° only slightly larger than
the interplanar angles to the anilino groups (24.5 and
23.5°), this longer BϪN bond must be the result of the low
basicity of the diphenylamine. Because of the steric require-
ment of the diphenylamino group, the phenyl groups of the
anilino groups are pointing away from the Ph2N substitu-
ent. This results in a cisoidal arrangement for the NϪH
units. The CϪNϪB bond angles of the anilino substituents
are quite open (132.8° and 130.9°). This is not imposed by
steric requirements of the diphenylamino group, as the
same feature is also observed in compound 3.
Figure 3. Structure representation of compound 3 in ORTEP de-
scription; selected bond lengths and bond angles: B1ϪN1 1.428(2),
B1ϪN2 1.431(2), B1ϪN3 1.442(2), N1ϪC1 1.409(2), N2ϪC7
˚
1.414(2), N3ϪC13 1.402(2) A; N1ϪB1ϪN3 119.1(2), N2ϪB1ϪN3
119.5(2),
N1ϪB1ϪN2
121.3(1),
B1ϪN1ϪC1
127.8(1),
B1ϪN2ϪC7 128.2(1), B1ϪN3ϪC13 129.0(1)°; angles between the
planes of the phenyl groups: C1ϪC6/C7ϪC12 74.4, C1ϪC6/
C13ϪC18 107.1, C7ϪC12/C13ϪC18 110.2°
The main difference between tris(2-pyridylamino)borane
(4) and trianilinoborane (3) is that all atoms of the molecule
of 4 lie in or close to a common plane (see Figure 4), the
interplanar angles of the aminopyridine substituents to the
BN3 plane being less than 5.7°. All BϪN bonds are of equal
[26]
˚
lengths (1.432 A), and are typical of triaminoboranes.
Intramolecular NϪH···N bridges are the most likely reason
for the coplanarity of the molecule. On the other hand, the
BϪNϪC bond angles lie between 130.5 and 132.0°. This
certainly makes hydrogen bonds weaker, because a 120°
BϪNϪC bond angle would allow a closer intramolecular
˚
Figure 2. The molecular structure of 2 in the crystal; selected bond
lengths and bond angles: B1ϪN1 1.470(3), B1ϪN2 1.422(3),
B1ϪN3 1.427(3), N1ϪC1 1.330(3), N1ϪC7 1.423(3), N2ϪC13
N···N distance, which is actually 2.809 A. The ‘‘hydrogen
bond’’ bond angles range from 134 to 137°. This geometry
therefore seems to be in better accord with a dipolar inter-
action NϪH(δϩ)···N(δϪ). A similar structure has been ob-
served for the hydrazinoborane B(NHϪNMe2)3, which has
a planar BN6H3 skeleton with the NC2 group perpendicular
˚
1.415(3), N3ϪC19 1.414(3) A; B1ϪN1ϪC1 120.7(2), B1ϪN1ϪC7
123.1(2),
C1ϪN1ϪC7
116.2(2),
B1ϪN2ϪC13
132.8(2),
H2ϪN2ϪC13 113(1), H2ϪN2ϪB1 114(1), B1ϪN3ϪC19 130.9(2),
H3ϪN3ϪC19 116(1), H3ϪN3ϪB1 113(1), N1ϪB1ϪN3 117.5(2),
N2ϪB1ϪN3 125.9(2), N1ϪB1ϪN2 116.6(1)°
1134
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145

Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes
FULL PAPER
to the BN6 plane. The N···N distance is 2.41(3) A,[27] much
shorter than in compound 4.
All BϪNϪC bond angles are wider than 130°, as found
for all B(NHR) units in compounds 1Ϫ5. However, it was
surprising to find that one of the three BϪN bonds of 5 is
˚
˚
0.02 A longer than the other two. This indicates a somewhat
weaker BϪN bond with less π-bonding, and this longer
bond is indeed associated with the substituent that shows
the strongest twisting.
N-Lithiotriaminoboranes
Synthesis
The triaminoboranes mentioned above were metallated
with n-butyllithium under various conditions. In the case of
1, the metallation proceeded smoothly, according to Equa-
tion (4). Monitoring of the deprotonation by 11B NMR was
not advisable, because the shielding of the boron nucleus
Figure 4. The molecular structure of 4 in the crystal; selected atom did not change significantly (from δ11B ϭ 23.8 to 23.0 in
this case). On the other hand, the 11B NMR spectrum ruled
distances and bond angles: B1ϪN1 1.434(2), B1ϪN2 1.432(2),
B1ϪN3 1.431(2), N1ϪC2 1.385(2), N2ϪC7 1.379(2), N3ϪC12
˚
out both the formation of a borate and BϪN bond cleav-
age. In the first of these two cases, a highfield shift should
be observed, in the second case a shift to lower field should
be expected.[28] Attempts to crystallize compound 6 from
toluene failed. Crystals were obtained after addition of pyr-
idine, but these proved to be the hydrolysis product 7 [see
Equation (5)]. Tetramethylethylenediamine seemed to be a
better ligand than pyridine for crystal formation. However,
the ‘‘crystals’’ of 8 proved to be amorphous.[29]
1.378(2), N5···H1 2.08, N4···H3 2.07, N6···H2 2.12 A; B1ϪN1ϪC1
130.5(1), B1ϪN1ϪH1 114(1), B1ϪN2ϪC7 131.3(1), B1ϪN2ϪH2
115(1), H2ϪN2ϪC7 114(1), B1ϪN3ϪC12 132.0(1), B1ϪN3ϪH3
112(1), C12ϪN3ϪH3 116(1), N1ϪB1ϪN2 120.0(1), N1ϪB1ϪN3
119.9(1), N2ϪB1ϪN3 120.2(1)°
Tris(8-quinolylamino)borane (5) crystallizes in the tri-
¯
clinic system, space group P1, Z ϭ 2 (Figure 5). The mo-
lecular planes of the aminoquinoline substituents are not
uniformly twisted relative to the BN3 plane, but by 18.7°
(N1), 33.9°(N3), and 24.4° (N5). Interactions between their
NϪH protons and the N atom of the heterocycle occur not
between adjacent heterocycles as in compound 4, but within
its own heterocyclic system. The N···N distance is, at 2.720
(4)
˚
A, somewhat shorter than in 4. As is to be expected for a
five-membered ‘‘ring’’ (including a hydrogen bond), the
angles at the bridging H atoms are sharp at ca. 100°.
(5)
Double metallation of the benzodiazaborolidine 9 by 2
equiv. of LiBu at Ϫ78 °C occurred readily, only a single 11
B
NMR signal being present after metallation (δ11B ϭ 29.3).
At Ϫ60 °C, however, a white solid separated from the gray-
green suspension, and no 11B NMR signals could any
longer be observed in the solution. Obviously, the product
10 had become insoluble. For this reason it was impossible
to grow single crystals, nor was it possible to obtain crystals
from the solid dissolved in pyridine. On the other hand, the
Figure 5. The molecular structure of 5 in the crystal; selected atom
distances and bond angles: B1ϪN1 1.429(4), B1ϪN3 1.448(4),
B1ϪN5 1.427(4), N1ϪC1 1.409(4), N3ϪC10 1.389(4), N5ϪC19
˚
1.401(4), N2···H1ϪN 2.25, N4···H3ϪN 2.35, N6···H5ϪN 2.25 A;
B1ϪN1ϪC1 130.5(3), B1ϪN3ϪC10 131.2(3), B1ϪN5 C19
131.6(3),
N1ϪB1ϪN3
119.3(3),
N1ϪB1ϪN5
120.8(3),
N3ϪB1ϪN5 119.6(3)°; torsion angle: C1ϪN1ϪC2ϪN2: Ϫ26.8
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145
1135

U. Braun, T. Habereder, H. Nöth, H. Piotrowski, M. Warchhold
It followed from the previous observations that multimet-
FULL PAPER
elemental analysis of the precipitate was almost consistent
with dimetallated 10 as far as the C/N ratio was concerned. allation required a polar solvent, in order to prevent the
In order to show that 10 had indeed been formed in the rapid formation of insoluble products. The triple lithiation
reaction described by Equation (6), the precipitate was of B(NHPh)3 was therefore studied in the presence of ether.
treated with Me3SiCl. As expected, the bis(trimethylsilyl) The reaction, as in Equation (9), proceeded readily at low
derivative 11 was obtained and characterized by NMR temperature. Two 11B NMR signals were found at δ ϭ 35.4
spectroscopy.
and Ϫ3.5, the latter representing a minor component. The
signal at δ ϭ 35.4 was assigned to compound 15. However,
the formation of a precipitate could not be prevented. The
solid, on dissolution in THF, showed the same signals as
the ether solution. Moreover, the 1H NMR spectrum
showed the absence of NH protons, a good indication of
successful triple metallation. The minor component was,
most probably, due to the formation of LiBuB(NHPh)3 or
LiB(NHPh)4.[30] From the solution, compound 15 was isol-
ated as B(NLiPh)3·LiCl·3OEt2 (15A) in 86% yield.[31]
(6)
The triple metallation of 4 was performed in THF as a
solvent, as the solubility of the aminoborane in ether was
rather low. The solution obtained after metallation showed
a single but rather broad 11B NMR signal at δ ഠ 40. This
lowfield shift (δ ϭ 15.8) indicated a reaction according to
Equation (10), since there was also a lowfield shift of δ ϭ
11.2 found for the triple lithiation of compound 4. How-
ever, the product proved to be not 16, but rather its adduct
with LiNHR. Therefore, BϪN cleavage as in Equation (11)
must also have occurred, to produce Li4B(NR*)3(NHR*)-
(7)
(8) (THF)3 (18; R* ϭ 2-pyridyl). Equations (10) to (13) sum-
marize the results. The formation of borate 19 was indicated
by an 11B NMR signal at δ11B ϭ Ϫ9.[5,28]
The lithiation of dianilino(diphenylamino)borane (2) by
BuLi in the presence of dimethoxyethane (DME) produced
a yellow orange solution, the 11B NMR spectrum of which
showed two signals at δ ϭ 28.9 and Ϫ10 in an intensity
ratio of 85:15, provided that the spectrum was measured as
rapidly as possible. We assigned the lowfield signal to the
metallated product 12 and the highfield signal to a product
containing tetracoordinated boron atoms. The orange solu-
tion turned brown within a few hours, with the formation
of a white precipitate. No 11B NMR signals could then be
detected in the supernatant solution. Nevertheless, crystals
separated slowly from the solution, and these were identi-
fied as Ph2NLi·dme (13). Equation (7) describes a formal
decomposition process for 12.
When pyridine was used to stabilize 12, a clear yellow
solution resulted. Its 11B NMR spectrum, with signals at
δ ϭ Ϫ17.6, Ϫ15.4, 0.2, 1.9, 3.0, and 43.2, demonstrated a
much more complex reaction than that observed in the
presence of DME. Obviously, aminoborates such as
Li[(PhLiN)2B(NPh2)Bu] or aminoboraneϪpyridine adducts
such as (PhHN)2(Ph2N)B·py may form besides
BuB(NPh2)(NLiPh). Crystals separated from the solution,
and these proved to be LiNPh2·(py)2 [14; see Equation (8)].
The lithium diphenylamides 13 and 14 were dimeric in the
solid state (vide infra) but had different structures.
Although the composition of 18 might suggest the forma-
tion of a tetraaminoborate, this was not the case, as verified
by the 11B NMR spectrum and the determination of its
molecular structure (vide infra).
(9)
(10)
(11)
(12)
So far, the stoichiometry of the reaction has not been
determined, as LiNPh2 was almost certainly not the main
product. We assume that BϪN bond cleavage of 2 occurred,
most probably at a stage after dimetallation of 2.
(13)
1136
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145

Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes
FULL PAPER
In order to provide more insight into the metallation of
4, reactions in a 1:1 and a 1:2 ratio were studied in addition
to the 1:3 ratio. The observed 11B NMR signals are sum-
marized in Table 1. In all cases, two main products were
formed. The signal at δ11B ϭ 23.3 indicated that the envir-
onment about the tris(2-pyridylamino)borane was not
greatly changed by the N-metallation. It may well be that
the signals represented a mono- and a dilithiated 4. On tri-
metallation, however, a noticeable deshielding indicated sig-
nificant changes in triply metallated 4. Since 11B NMR sig-
nals at δ ϭ Ϫ9.5 were absent in the 1:1 and 1:2 reactions
but one was observed in the 1:3 reaction, it appeared that
this signal resulted from the anion [Bu3BNHR*]Ϫ, as there
was a sufficient supply of BuϪ anions under these condi-
tions, and the observed chemical shift observed was similar
to the value for Na(Et3BNH2) (δ11B ϭ Ϫ9.8).[28] The boron
resonances at δ ϭ Ϫ1.5 and Ϫ3.3 (see Table 1) were consist-
ent with the formation of Li[BuB(NHR*)3] or/and Li-
[BuB(NHR*)2(NLiR*)].[28]
(17)
In contrast to the triple metallation of 4, attempts to
achieve the triple metallation of tris(8-quinolylamino)bor-
ane (5) in THF were futile. No single crystals separated
from the violet-red solution, although a single broad 11B
NMR signal at δ ϭ 32.9 might have been indicative of the
triple metallation. This assumption was supported by a
single Li resonance. However, the H and 13C NMR spec-
7
1
tra were inconclusive.
Structures
The pyridine adduct of lithium 1,3-diphenyl-2-oxido-
1,3,2-diazaborolidine (7) crystallizes in the triclinic space
group P1 with two independent dimeric molecules (Figure 6
¯
depicts only one of the two molecules). Each molecule has
a crystallographic center of inversion, generating four-mem-
bered planar Li2O2 rings. The LiϪO bonds of the rings are
significantly different, with bond lengths of 1.926(4) and
1.894(3) A in molecule 1 and 1.912(4) and 1.966(3) A in
molecule 2. Moreover, the OϪLiϪO and LiϪOϪLi bond
angles are different for the two molecules (97.8° and 82.2°,
and 118.8° and 84.0°, respectively). Two pyridine molecules
per Li ion complete the coordination shell around the Li
centers, with NϪLiϪN bond angles of 96.0° (molecule 1)
and 93.9° (molecule 2). Surprisingly, these angles are more
acute than the OϪLiϪO bond angles, in spite of the fact
Table 1. 11B NMR signals observed in reactions of tris(2-pyridyl-
amino)borane with various ratios of BuLi; the resonances in italics
represent the main species
˚
˚
Reagent ratio
Solvent
1:1
1:2
1:3
1.3
Ϫ3.3
Ϫ3.3
Ϫ9.5
Ϫ9.3
Ϫ0.9 Ϫ1.5 1.4 23.3
Ϫ0.9 1.2 1.4 23.5
1.3 39.6
THF
THF
THF
0
0.6 1.5 36, br.[a] diethyl ether
Ϫ24.6, Ϫ28.6
[a]
The only signal obtained on dissolving the precipitate in THF.
(14)
(15)
(16)
To provide experimental evidence for the existence of
monolithio compound 20, stannylation by Me3SnCl as in
Equation (14) was attempted. The mono(trimethylstannyl)
compound 21 was obtained, but could be characterized
only by 119Sn NMR spectroscopy.[32] However, the reaction
of R*NHϪB(NLiR*)2 as shown in Equation (15) did not
generate compound 22. The solution showed 119Sn and 11
B
NMR signals, as observed in the stannylation of 20, but Figure 6. One of the two dimeric molecules of 7; hydrogen atoms
are omitted for clarity; selected bond lengths and bond angles:
there were additional signals in both NMR spectra.
B1ϪO1 1.320(2), Li1ϪO1 1.926(4), Li1AϪO1 1.894(3), B1ϪN2
Amongst these were 119Sn (δ ϭ 9.2) and 11B resonances
1.467(3), B1ϪN1 1.478(3), N1ϪC3 1.391(3), N1ϪC1 1.466(2),
C2ϪN2 1.455(2), N2ϪC9 1.386(3), Li1ϪN3 2.152(4), Li1ϪN4
(δ ϭ 30.2) that we assigned to tris[2-pyridyl(trimethylstan-
nyl)amino]borane (23), a compound that could be isolated
and structurally characterized by X-ray crystallography.
Obviously, the distannylated molecule 22 had exchanged
Me3Sn groups to give 23 and 21 as represented in Equa-
tion (17).
˚
2.128(4) A; B1ϪO1ϪLi1 127.9(2), B1ϪO1ϪLi1A 146.4(2),
Li1ϪO1ϪLi1A 82.2(2), O1ϪLi1ϪO1A 97.8(2), O1ϪLi1ϪN4
126.4(2),
O1ϪLi1ϪN3
108.3(2),O1AϪLi1ϪN3
117.1(2),
N3ϪLi1ϪN4 95.0(2), O1ϪB1ϪO1A 128.0(2), O1ϪB1ϪN2
127.6(2),
N1ϪB1ϪN2
104.4(2),
C1ϪN1ϪC3
118.0(2),
B1ϪN1ϪC3 129.9(2), B1ϪN1ϪC1 111.1(2), C2ϪN2ϪC9 118.5(2),
B1ϪN2ϪC9 129.9(2), B1ϪN1ϪC2 111.4(2)°
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145
1137

U. Braun, T. Habereder, H. Nöth, H. Piotrowski, M. Warchhold
FULL PAPER
that the N atoms are not part of a ring system. Con- tricoordinated N2 atom is, as might be expected, even
˚
sequently, the OϪLiϪN bond angles are more open, giving shorter (1.950 A). While the Li1ϪN1ϪLi2 bond angle is
rise to Li ions with a rather distorted ‘‘tetrahedral’’ environ- almost tetrahedral, the C14ϪN1ϪC2 bond angle is more
ment.
open, at 119.0(3)°. The tetracoordinated Li2 atom bears
BϪN bonds longer than those in compound 1 are found three pyridine molecules, with an average LiϪN bond
˚
˚
in 7. The lengthening by ca. 0.04 A is due to the negative length of 2.111 A. The phenyl rings at N2 form interplanar
charge at the oxygen atom. On the other hand, the BϪO angles with the N3Li plane of 17.6Ϫ0.7°, and the in-
˚
distance [1.320(3) A] is shorter than those in B(OMe)3, terplanar angle between N2C2 and Li1N3 at 95.6° shows an
(MeBO)3,[33] or B2(OMe)4,[34] and similar to the BϪO dis- almost orthogonal orientation. This allows the pyridine
˚
tances of 1.317(6) A found in [thf·LiOB(mes)2]2 (mes ϭ ring N3 to adopt a staggered conformation not only with
2,4,6-trimethlphenyl)[35] and 1.345 A in [LiOB[CH(Si- the phenyl groups at N2, but also with the plane around
˚
Me3)2]2.[36] The BϪO bond order in 7 is therefore higher atom N1.
than that in methylborates or trimethylboroxine, and this is
also responsible for the longer BϪN bonds.
The molecular structure of dimeric Ph2NLi·dme (13) is
shown in Figure 7. The centrosymmetric molecule has a
four-membered, rhombic Li2N2 ring. Each Li ion is coord-
inated by two oxygen and two nitrogen atoms, with different
LiϪO and LiϪN bond lengths, a sharp LiϪNϪLi bond
angle (80(1)°], and a more open NϪLiϪN bond angle of
99(1)°.
Figure 8. Molecular structure of dimeric 14; selected bond lengths
and bond angles: Li1ϪN1 2.042(6), Li1ϪN2 1.950(7), Li1ϪN3
2.054(7), Li2ϪN1 2.132(7), Li2ϪN4 2.137(7), Li2ϪN5 2.077(7),
Li2ϪN6 2.119(6), C1ϪN2 1.368(4), C7ϪN2 1.393(4), C18ϪN1
˚
1.405(4), C24ϪN1 1.410(4) A; Li1ϪN1ϪLi2 108.1(3),
Li1ϪN1ϪC18 95.3(3), Li1ϪN1ϪC24 121.0(3), Li2ϪN1ϪC18
111.7(3), Li2ϪN1ϪC24 101.8(2), C1ϪN2ϪC7 119.4(3),
C1ϪN2ϪLi1 128.9(3), Li1ϪN2ϪC7 111.3(3), N1ϪLi1ϪN2
139.8(3), N1ϪLi1ϪN3 108.8(3), N2ϪLi1ϪN3 109.1(3),
N5ϪLi2ϪN6 105.2(3), N1ϪLi2ϪN5 122.0(3), N1ϪLi2ϪN6
105.3(3), N4ϪLi2ϪN5 100.3(3), N4ϪLi2ϪN6 104.3(3),
N1ϪLi2ϪN4 118.0(3)°
Figure 7. Molecular structure of the dimeric 13; selected bond
lengths and bond angles: Li1ϪN1 2.07(2), Li1AϪN1 2.14(3),
Li1ϪO1 2.07(2), Li1ϪO2 2.12(2), N1ϪC1 1.41(1), N1ϪC7
˚
1.425(8) A; Li1ϪN1ϪLi1A 80(1), N1ϪLi1ϪN1A 99(1),
The triply lithiated trianilinoborane 15, which crystal-
lized as B(NLiPh)3·LiCl·3OEt2 (15A), once again proved to
be dimeric in the solid state, its core structure containing
eight Li atoms. Figure 9 shows the molecular structure. The
O1ϪLi1O2 79.7(7), O1ϪLi1ϪN1 125(1), Li1ϪN1ϪC7 124.2(5),
C1ϪN1ϪC7 115.8(9)°
While the structure of 13 resembles the structures of di- dimeric molecule is centrosymmetric. Four Li atoms (Li1,
[37]
meric (smes)HNLi·OEt2
[smes ϭ 2,4,6-tris(tert-bu- Li1A, Li3, Li3A) are three-coordinated by one O, N, and
tyl)phenyl] or (Me3Si)[(Me3Si)3Si]NLi[38] as far as the asso- Cl atom each, while the other four Li centers are tetraco-
ciation through LiϪN bonds is concerned, lithium amides ordinated, atoms Li2 and Li2A by two N atoms, one O
with tetracoordinated Li centers are more common.[16] atom, and one Cl atom, and atoms Li4 and Li4A by three
However, a much more interesting structure was found for N atoms and one Cl atom. In order to describe the struc-
dimeric Ph2NLi·(py)2 (14), the molecular structure of which ture of 15A it is best to start from the B(NLiPh)3 unit. The
is shown in Figure 8. There are two kinds of Li ions. Li1 is core atoms of this molecule are shown in Figure 10.
surrounded by a plane of three nitrogen atoms in which the
The most important feature of compound 15A is the
N1ϪLi1ϪN2 bond angle is quite open, at 139.8°. The long- planar BN3 unit, which shows different BϪN bond lengths
˚
est LiϪN bond is that formed with the pyridine nitrogen of 1.429(9), 1.459(9), and 1.509(9) A to atoms N1, N3, and
atom, followed by those with the nitrogen atom of the tetra- N2, respectively. While B1ϪN1 is practically of the same
˚
coordinated Ph2N group (2.042 A), while the bond to the length as in the parent compound 3, the other two BϪN
1138
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145

Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes
FULL PAPER
B1N1C1Li1 plane. However, the N1ϪLi1 distance is
˚
2.961(6) A, the same distance as found for N1ϪLi4A. The
situation at atom N3 is similar to that at N1, but there is
greater torsion, as demonstrated by the angle of 50.8° be-
tween the B1N1N3 and B1N3C13 planes. This has the con-
sequence of weaker BN-π-bonding, as revealed by the
longer B1ϪN3 bond. The distance from N3 to Li3 [1.983(7)
˚
˚
A] is shorter than that to Li2 [2.005(5) A], but the
B1N3C13Li3 atom system lies almost in a plane (B1 being
˚
distant). Finally, the interplanar angle
Ϫ0.192
A
B1N1N2N3/B1N2C7 is 57.6°, and this seems to be a reason
for the long B1ϪN2 bond to which the pentacoordinated
atom N2 contributes. The coordination of Li2 and Li4A
by two N atoms of the trianilidoborane unit is what most
probably gives rise to an N1ϪB1ϪN3 bond angle as wide
as 132.9(3)°.
Figure 9. Molecular structure of compound dimeric 15A; hydrogen
atoms and C atoms of the ether molecules omitted for clarity
As shown in Figure 10, the Cl anion is coordinated to
the Li atoms of the B(NLiPh)3 unit, with distances of 2.564,
˚
2.377, and 2.362 A to atoms Li2, Li3, and Li4, respectively.
The presence of the chloride results in the formation of a
distorted cubane unit in which one corner of the cube is
missing. The Cl ion is surrounded by four Li ions. An un-
usually open coordination at Cl results, and this is best de-
scribed as a strongly distorted trigonal bipyramid with a
coordination partner missing in the equatorial plane.
Atoms Li1, Li1A, Li3, and Li3A coordinate with three li-
gand atoms (O, N, Cl) in a nonplanar fashion.
The triply lithiated 4 crystallizes with one molecule of
LiHNR* as [Li4B(NR*)3HNR*(THF)3]2 (18)2. Its molecu-
lar structure is shown in Figure 11 and the monomeric unit
is displayed in Figure 12. As may be noted, the metallation
of compound 5 leaves the boron atom in a planar environ-
ment, the extra LiHNR* group not being bonded to atom
B1 to generate a tetraamidoborate. One BϪN bond of the
˚
B(NLiR*)3 unit is short, at 1.440(3) A (B1ϪN1), while the
other two are of almost equal lengths [1.476(3) for N2 and
1.467(3) A for N3]. Thus, all BϪN bonds are longer than
those in the parent compound 4. Each N atom of the BN3
unit is coordinated to two Li atoms, but there is only one
Figure 10. The core structure of [B(NLiPh)3·LiCl·3 OEt2]2 (15A)2;
only the ipso-C atoms of the phenyl rings are shown; the Et groups
of the ether molecules have also been removed; selected bond
lengths and bond angles: B1ϪN1 1.429(4), B1ϪN2 1.509(4),
B1ϪN3 1.459(4), Li1 N1 2.021(6), N1ϪLi4A 2.029(5), Li3ϪN3
˚
1.083(7), Li2ϪN3 1.983(7), Li4ϪN2 2.006(5), Li4AϪN2 2.154(5), Li atom (Li3) that bridges two N atoms of the BN3 unit.
N1ϪC1 1.400(3), N2ϪC7 1.364(4), N3ϪC13 1.373(3). Cl1ϪLi2
This gives rise to a small N1ϪB1ϪN3 bond angle of
2.564(5), Cl1ϪLi3 2.377(5), Cl1ϪLi4 2.362(5), Cl1A Li1 2.451(5),
˚
111.8(2)°. The twisting of the CNB planes relative to the
BN3 plane is 24.4, 50.7, and 88.2° (at N1, N3, N2), respect-
ively.
Li1ϪO1 1.901(6), Li2ϪO2 1.910(6), Li3ϪO3 1.925(6) A;
N1ϪB1ϪN3 132.9(3), N1ϪB1ϪN2 113.9(2), N3ϪB1ϪN2
112.1(2), Li1AϪCl1ϪLi3 121.1(2), Li1AϪCl1ϪLi2 130.2(2),
Li4ϪCl1ϪLi1A 64.7(2), Li4ϪCl1ϪLi2 67.5(2), Li4ϪCl1ϪLi3
88.7(2), Li2ϪCl1ϪLi3 69.8(2)°
bonds are significantly elongated. Actually, the long
B1ϪN2 bond corresponds to a single bond between sp2-
type B and N atoms. The reasons for these differences are
the different NϪBϪNϪC torsion angles and the number
of coordinated Li atoms at the N atoms. Thus, the
B1N1N2N3/B1N1C1 interplanar angle is 21.5°, which still
allows for BN-π-bonding. While atom Li1 is part of the
planar environment around atom N1 (sum of bond angles
357.8°), one may note that the atom Li4A has an N1ϪLi4A
Figure 11. The molecular structure of [B(NLiR*)3·LiHNR*
(THF)3]2 (18)2; hydrogen atoms removed, as well as the C atoms
˚
˚
distance of 2.209(5) A and lies 1.934 A above the of the coordinated THF molecules
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145
1139

U. Braun, T. Habereder, H. Nöth, H. Piotrowski, M. Warchhold
FULL PAPER
Figure 12. The asymmetric (monomeric) unit of 18; bond lengths
and bond angles: B1ϪN1 1.440(3), B1ϪN2 1.476(3), B1ϪN3
1.467(3), N1ϪC1 1.353(3), N2ϪC16 1.332(3), N3ϪC6 1.338(3),
Li1ϪN1 2.114(5), Li3ϪN1 2.046(4), Li4ϪN2 2.065(4), Li2ϪN3
Figure 13. ORTEP plot of compound 23; H atoms have been omit-
ted for clarity; selected bond lengths and bond angles: B1ϪN1
1.443(7), B1ϪN2 1.435(6), B1ϪN3 1.457(6), N1ϪSn1 2.109(4),
N2ϪSn2 2.119(4), N3ϪSn3 2.110(3), Sn1···N4 2.717, Sn2···N5
2.703, Sn3···N6 2.672, N1ϪC1 1.396(5), N2ϪC6 1.383(6),
˚
2.067(4) A; N1ϪB1ϪN2 125.1(2), N1ϪB1 N3 111.8(2),
N2ϪB1ϪN3 122.4(2), C1ϪN1ϪB1 123.9(2), C1ϪN1ϪLi3
138.3(2), C1ϪN1ϪLi1 106.8(2), B1ϪN2ϪC16 113.8(2),
B1ϪN2ϪLi4 108.6(2), B1ϪN2ϪLi4A 104.7(2), Li4ϪN2ϪLi4A
78.4(2), Li4ϪN2ϪC16 117.7(2), Li4AϪN2ϪC16 128.5(2),
B1ϪN3ϪC6 122.9(2), B1ϪN3ϪLi2 96.5(2), B1ϪN3ϪLi3 87.6(2),
C6ϪN3ϪLi2 123.3(2), C6ϪN3ϪLi3 125.4(2), Li2ϪN3ϪLi3
91.8(2)°
˚
N3ϪC11 1.385(5) A; N1ϪB1ϪN2 121.2(4), N1ϪB1ϪN3 118.5(4),
N2ϪB1ϪN3 120.3(4), B1ϪN1ϪC1 123.6(4), C1ϪN1ϪSn1
108.2(3),
B1ϪN1ϪSn1
128.2(3),
B1ϪN2ϪC6
123.7(4),
C6ϪN2ϪSn2 107.9(3), B1ϪN2ϪSn2 128.4(3), B1ϪN3 C11
124.0(4), C11ϪN3ϪSn3 107.8(3), B1ϪN3ϪSn3 128.1(39,
N1ϪSn1ϪC16 104.6(2), N1ϪSn1ϪC17 111.5(2), N1ϪSn1ϪC18
113.0(2)°
The LiϪN atom distances in (18)2 span a range from
shielding of the boron atom, the structure of the aminobor-
ane, and the solvent used. In almost all so far known cases,
including the results presented here, there is competition
between deprotonation and borate formation, as well as
BϪN bond cleavage. The last can be impeded if the boron
atoms are part of a ring system, as shown for the triazabor-
adecalin system[3] or for borazines (RBϭNH)3.[18] However,
an exocyclic BϪN bond is comparatively easily cleaved, as
demonstrated by compound 2, while borate formation can
be fully or at least partly suppressed by the use of tBuLi.[41]
In the case of the double deprotonation of the aminoborane
2, one observed the formation of a borate, most probably
Li[BuB(NPh2)(NLiPh)2], in solution besides Ph2NB-
(NLiPh)2. Both compounds might be the source of LiNPh2,
which was formed slowly. Isolated Ph2NLi solvates form
dimers, as is well known for various lithium diorganylam-
ides. However, chain-like dimers of lithium diorganylamides
are rare.[16,17]
Neither borate formation nor BϪN bond cleavage was
observed in the triple lithiation of trianilinoborane (3) in
the presence of ether, unlike in the case of tris(2-pyridylami-
no)borane, in which BϪN cleavage to give LiNHR* oc-
curred. This amide cocrystallized with B(NLiR*)3 in a 1:1
ratio to give (18)2, a cluster with an Li8N8 core. In addition,
(2-aminopyridyl)butylborates were also formed. Neverthe-
less, deprotonation was the dominant reaction both for 3
and for 4.
˚
2.036(5) to 2.114(5) A for the amino nitrogen atoms, and
˚
2.038(4) to 2.155(4) A for the pyridine nitrogen atoms, with
˚
the exception of the Li2ϪN5 distance of 2.286(4) A. The
reason for the long distance is most probably the specific
situation at atom N5, since this atom in the only pyridine
nitrogen atom that coordinates to two Li centers.
The (stannylamino)borane 23 crystallizes in the mono-
clinic system, its molecular structure being shown in Fig-
ure 13. The BϪN bond lengths range from 1.435(6) to
˚
1.457(6) A. These are slightly longer than those in the par-
ent compound 5. All SnϪN bonds to the amino nitrogen
atoms are of equal length. The pyridine nitrogen atoms are
oriented towards the tin atoms of each Me3SnNR* unit.
This may indicate the presence of SnϪN interactions. If
there are any, then these must be quite weak, as the
CϪSnϪC bond angles deviate only slightly from the tetra-
hedral angle.
All amino nitrogen atoms reside in a planar environment,
but the SnNC planes at N1, N2, and N3 are twisted relative
to the BN3 plane by 40.6, 38.1, and 46.1°, respectively.
These torsion angles explain the lengths of the BϪN bonds.
The twisting results from the steric requirement of the tri-
methylstannyl groups and is the most striking difference on
comparison of this structure with the parent compound 4.
As would be expected, the SnϪC bond lengths are in the
[39,40]
˚
typical range (2.12Ϫ2.15 A).
However, as we have seen for compound 18, the
metallated triaminoboranes did not add amide ions to form
Discussion
The deprotonation of aminoboranes bearing RHN tetraaminoborates. These latter are scarce anyhow, and
groups by butyllithium depends on the steric and electronic form readily only when the amido group is derived from a
1140
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145

Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes
FULL PAPER
heterocyclic system, such as the anions of pyrrole,[42] in-
dole,[43] pyrazole,[44,45] or dihydropyridine.[46] In these cases,
the parent triaminoborane is comparatively Lewis-acidic.
This is obviously not the case for the tris(lithioamino)bor-
anes; the negative charge at the nitrogen atoms reduces the
Lewis acidity of the boron center. Therefore, an amide does
not attack at the boron center, but rather forms coordin-
ative LiϪN bonds. The lithium atoms adopt coordination
numbers ranging from 3 to 5. Therefore, in comparison
with lithium amides,[16,17] the borylated lithium amides of-
fer a much more versatile structural chemistry similar to
that of lithiated silylamides[16] and aminophosphanes.[47] In
addition, they are very useful reagents. This has been shown
previously,[5Ϫ10,48] and we will report on this point in a
forthcoming paper describing the synthesis and reactions
of (tBuLiN)B(R)tmp (tmp ϭ 2,2,6,6-tetramethylpiperidino)
(see ref.[4]).
signal at δ ϭ 41.1, corresponding to 2-butyl-1,3-diphenyl-1,3,2-di-
azaborolidine (ratio approximately 97:3). The solid was removed
from the suspension by filtration. A mixture of aniline (4.6 mL,
50 mmol) and triethylamine (7.0 mL) was added at Ϫ78 °C to the
slightly yellow filtrate. After the reaction mixture had been heated
at reflux overnight and after removal of the solid, the filtrate was
concentrated in volume to approximately 50 mL. Crystals separ-
ated from the solution within a few days. Crystallization from tolu-
ene yielded 2 (6.44 g, 71.0%), m.p. 185 °C. C24H22BN3 (362.81):
calcd. C 79.83, H 6.06, N 11.58; found C 78.09, H 6.21, N 11.08.
Trianilinoborane (3):[49] Aniline (5.5 mL, 60 mmol) was added to a
solution of B(NMe2)3 (3.45 mL, 20 mmol) in toluene (70 mL). Gas
evolution started, and became rapid on heating of the solution at
reflux for 5 h. A single 11B NMR signal at δ ϭ 23.7 indicated a
quantitative reaction. Crystals separated within a few hours. Half
of the solvent was removed from the suspension in vacuo, and 6.6 g
of 3 was isolated. Yield: 87%; m.p. 174 °C. NMR in C6D6: δ1H ϭ
3
4
4.83 (br. s, 3 H, NH), 6.75 (dd, JH,H ϭ8.5, JH,H ϭ 1.1 Hz,6 H,
3
4
o-H), 6.80 (tt, JH,H ϭ 7.6, JH,H ϭ 1.2 Hz, 3 H, p-H), 7.06 (t,
3JH,H ϭ 7.5 Hz, 6 H, m-H); δ13C ϭ 119.9, (o-C), 120.8 (p-C), 129.4
(m-C), 145.4 (i-C); δ11B ϭ 23.7 (h1/2 ϭ 214 Hz); δ14N ϭ Ϫ292.3.
Experimental Section
Tris(2-pyridylamino)borane
(4):
2-Aminopyridine
(5.65 g,
General: All experiments required anhydrous conditions. Schlenk
59.4 mmol) and B(NMe2)3 (20.0 mmol) were heated under reflux
techniques were employed, with dinitrogen as a protecting gas. The in toluene (150 mL). Gas evolution (dimethylamine) rapidly
amines are commercially available; the liquids were dried with
either KOH or CaH2 and distilled, and the solids were recrystal-
started, and the amine was trapped at Ϫ78 °C. On cooling of the
solution to ambient temperature, a white precipitate of 4 formed.
lized before use. BuLi was obtained from Chemetall GmbH. NMR: After filtration, the filtrate was cooled to 0 °C. Colorless needles
Bruker ACP 200 (7Li, 11B, 119Sn), JEOL GSX 270, and JEOL EX separated; these were suitable for X-ray structure determination.
400 (1H, 13C, 119Sn) with Me4Si, C6D6, BF3·OEt2, and SnMe4 as
Yield: 5.2 g (90%), m.p. 121 °C. NMR in C6D6: δ1H ϭ 6.36 (ddd,
3
4
standards. Elemental analyses were performed in the microanalyt-
3JH,H ϭ 8.5, JH,H ϭ 5.4, JH,H ϭ 0.8 Hz, 3 H, C5-H), 6.46 (dt,
ical laboratory of this department. X-ray: Siemens P4 diffracto- 3JH,H ϭ 8.3, JH,H ϭ 0.7 Hz, 3 H, C3-H), 6.99 (ddd, JH,H ϭ 7.3,
4
3
meter equipped with a CCD area detector, graphite monochrom- 3JH,H ϭ 7.4, 4JH,H ϭ 1.6 Hz, 3 H, C4r-H), 8.14 (ddm, 3JH,H ϭ 5.3,
ator, and Mo-Kα source or a STOE IPS diffractometer (compounds
7 and 23). Data sets were collected at Ϫ80 °C (Siemens LT2 device).
4JH,H ϭ 1.8 Hz, 3 H, C6-H), 9.85(s, 3 H, NH); δ13C ϭ 113.6 (C3),
114.6 (C5), 137.4 (C4), 147.2 (C6), 159.6 (C2); δ11B ϭ 24.2 (h1/2 ϭ
218 Hz); δ14N ϭ Ϫ105.7 (pyN), Ϫ310.0 (NH). IR (Hostaflon,
2-Anilino-1,3-diphenyl-1,3,2-diazaborolidine (1): Aniline (0.87 mL,
9.5 mmol) and triethylamine (1.32 mL) were added to a stirred so-
lution of 2-chloro-1,3-diphenyl-1,3,2-diazaborolidine[48] (2.43 g,
9.50 mmol) in toluene (80 mL). A precipitate formed. The mixture
was heated under reflux overnight, the solid was removed by filtra-
cmϪ1): ν ϭ 3293 (NH2), 3090, 3059, 3043, 3022 (CH). C16H15BN5
˜
(290.13): calcd. C 57.14, H 5.71, N 26.62; found C 58.19, H 5.25,
N 27.11.
Tris(8-quinolylamino)borane (5): Tris(dimethylamino)borane (1.8
tion (glass frit), and the orange-yellow solution was concentrated mL, 19 mmol) was added to a suspension of 8-aminoquinoline
to approximately 50 mL in vacuo. Needles of 1 separated from the (4.32 g, 30.0 mmol) in toluene (150 mL). The mixture was heated
solution. Yield: 2.7 g, 71%; m.p. 153 °C. NMR (in C6D6): δ1H ϭ under reflux for 30 h. The red solution showed a broad 11B NMR
3.20 (s, 4 H, NCH2), 4.49 (s br., 1 H, NH), 6.47 (d, 3JH,H ϭ 7.3 Hz,
signal at δ ϭ 24.7. After cooling of the solution to 0 °C, red needles
3
2 H, o-CH aniline), 6.66 (t, JH,H ϭ 7.6 Hz, 1 H, p-CH aniline), separated within a few days. Yield: 3.75 g (85.3%), m.p. Ͼ 200 °C.
3
3
3
6.80 (t, JH,H ϭ 7.2 Hz, 2 H, p-H Ph), 6.90 (t, JH,H ϭ 7.5 Hz, 2 NMR in C6D6: δ1H ϭ 6.74 (dd, JH,H ϭ 3.7 Hz, 3 H, C3-H), 6.97
3
H, m-CH aniline), 6.93 (d, JH,H ϭ 7.6 Hz, 4 H, o-CH, Ph), 7.07 (d, br., 3JH,H ϭ 7.8 Hz, 3 H, C5-H), 7.21 (dd, 3JH,H ϭ 7.9, 3JH,H ϭ
(t, JH,H ϭ 8.8 Hz, 4 H, m-CH,Ph); δ13C ϭ 48.04 (CH2), 119.5 (o-
7.9 Hz, 3 H, C6-H), 7.54 (dd, JH,H ϭ 8.4, JH,H ϭ 1.8 Hz, 3 H,
3
3
4
C, Ph), 119.6 (o-C, aniline), 120.2 (p-C, aniline), 120.9 (p-C, Ph), C4-H), 7.72 (d, 3JH,H ϭ 7.5 Hz, 3 H, C7-H), 8.15 (s 3 H, NH), 8.48
3
4
128.8 (m-C, aniline), 129.0 (m-C, Ph), 143.1 (i-C, aniline), 146.6 (i-
C, Ph); δ11B ϭ 23.0 (h1/2 ϭ 300 Hz). C20H20BN3 (313.20): calcd. C 116.6 (C5), 121.4 (C3), 129.2 (C10), 135.7(C9), 136.1, C6), 140.2
76.62, H 6.38, N 13.41; found C 76.31, H 6.67, N 13.31.
(dd, JH,H ϭ 4.0, JH,H ϭ 1.5 Hz, 3 H, C2-H); δ13C ϭ 112.7 (C7),
(C4), 143.4 (C8), 146.9(C2); δ11B (in toluene) ϭ 24.7 (h1/2
ϭ
450 Hz); δ14N ϭ Ϫ74.8 (ring N), Ϫ326.9 (NH). IR (Nujol, cmϪ1):
Dianilino(diphenylamino)borane (2): A solution of diphenylamine
(4.23 g, 25.0 mmol) in toluene (150 mL) was cooled to Ϫ78 °C. A
solution of BuLi in hexane (1.56 , 16 mL) was added with stirring.
˜
ν ϭ 3340 (νNH). C27H21BN6 (439.81): calcd. C 73.58, H 4.77, N
19.07; found C 73.33, H 482, N 18.61.
A voluminous, white precipitate formed immediately, with some 2-[Lithio(phenyl)amino]-1,3-diphenyl-1,3,2-diazaborolidine (6) and
gas evolution. The suspension was allowed to attain ambient tem-
perature. Stirring was continued for additional 3 h. After the mix-
Formation of Lithium 2-Oxido-1,3-diphenyl-1,3,2-diazaborolidine
(7): A solution of compound 1 (0.490 g, 1.56 mmol) in toluene (40
ture had again been cooled to Ϫ78 °C, a solution of BCl3 in toluene mL) was diluted with ether (10 mL), and the mixture was cooled
(1.1 , 22.7 mL, 25.0 mmol) was slowly added. Stirring was con- to Ϫ78 °C. A solution of BuLi in hexane (1.0 mL, 1.56 mmol) was
tinued at ambient temperature overnight. The 11B NMR spectrum
added by syringe, with stirring. The solution turned orange. At
taken after that time showed a signal at δ ϭ 32.1 (2) and a tiny room temperature the solution showed only a single11B NMR sig-
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145
1141

U. Braun, T. Habereder, H. Nöth, H. Piotrowski, M. Warchhold
FULL PAPER
nal at δ ϭ 23.8 (6). The solid that was left on concentration of the Elemental analysis: C 48.12, H 5.88, N 17.92, corresponding to
solution also showed an 11B NMR signal at δ ϭ 23.8 in deuterio-
benzene, which was assigned to 6. Attempts to grow crystals from
this solution failed. However, after addition of pyridine (3 mL) a
few large crystals grew slowly at 0 °C over two weeks. These proved
C8H11.6N2.54 instead of C8H10N3(BLi2) for 10.
2-(Dimethylamino)-1,3-bis(trimethylsilyl)-1,3,2-benzodia-
zaborolidine (11): Me3SiCl (0.34 mL) was added at Ϫ78 °C to 10
(1.35 mmol) in THF (80 mL). The stirred suspension was allowed
to attain ambient temperature. All volatiles were removed from the
gray-green solution in vacuo, and hexane (30 mL) was added to the
oily residue. This resulted in a brown solution and a white insoluble
product (LiCl). The solid was separated by filtration. From the
filtrate all volatile material was removed. The liquid residue was
dissolved in CDCl3 and characterized as 11 by NMR spectroscopy.
NMR: δ1H ϭ 0.55 (s, 18 H, SiMe), 2.67, (s, 6 H, NMe), 6.91 (m,
2 H, C3 H, C6-H), 7.25 (m, C4-H, C5-H); δ13C ϭ 1.9 SiMe, 41.4
(CMeϭ, 114.3, C3, C6), 118.9 (C4, C5), 142.0 (C1,C2); δ11B ϭ
32.0 (h12 ϭ 227 Hz); δ27Si ϭ 4.7.
to be compound 7, as also characterized by X-ray crystallography.
3
7: NMR in C6D6: δ1H ϭ 3.47 (s, 4 H, NCH2), 6.82 (d, JH,H
ϭ
3
8.5 Hz, 2 H, p-CH), 6.98 (t, JH,H ϭ 7.6 Hz, 4 H, o-CH), 7.28 (t,
3JH,H ϭ 7.8 Hz, 4 H, m-CH), 7.42 (t, JH,H ϭ 8.6 Hz, 4 H, p-CH,
3
3
3
py), 7.68 (t, JH,H ϭ 7.6 Hz, 2 H, p-CH, py), 8.63 (d, JH,H
ϭ
4.5 Hz, 4 H, o-H, py); δ13C ϭ 43.0 (CH2), 113.0 (o-C), 117.8 (p-
C), 123.5 (m-C, py), 129.3 (p-C, py), 148.1 (ipso-C), 149.6 (o-C,
py); δ11B ϭ 22.7 (h1/2 ϭ 350 Hz); δ7Li ϭ 1.59. C24H24BLiN4O
(402.2): calcd. C 71.66, H 6.01, N 13.93; found C 68.37, H 6.35,
N 12.07.
Double Metallation of 2-(Dimethylamino)-1,3,2-benzodiazaboroli-
dine (9): The diazaborolidine 9[50] (0.26 g, 1.6 mmol) was added to Double Metallation of Dianilino(diphenylamino)borane and Forma-
a solution of tetramethylethylenediamine (0.46 mL, 3.2 mmol) in tion of LiNPh2: Dimethoxyethane (0.41 g, 4.5 mmol) was added to
tetrahydrofuran (80 mL). After this had been cooled to Ϫ78 °C, a
hexane solution of BuLi (2.1 mL, 1.56 ) was added to the vigor-
compound 2 (1.63 g, 4.50 mmol) in toluene (80 mL). The solution
was cooled to Ϫ78 °C. A hexane solution of LiBu (5.7 mL, 9.0
ously stirred solution. The color of the solution turned from yellow mmol) was added dropwise with stirring. After the yellow/orange
to gray-green. An 11B NMR spectrum at Ϫ78 °C showed only one
solution had been allowed to attain ambient temperature, it showed
signal at δ ϭ 29.3. On warming, a precipitate formed at about Ϫ60 two 11B NMR signals at δ ϭ 28.9 (12) and Ϫ10.0, in an intensity
°C, and no 11B NMR signal was then observed in the supernatant ration of 85:15. Solvent was removed in vacuo, to leave 10 mL of
solution at room temperature. When the solution was kept at Ϫ78
°C and the solvent was removed stepwise in vacuo, no crystals sep- and proved to be LiNPh2·dme (13). The supernatant liquid showed
arated. The solid that was formed at Ϫ60 °C was dried in vacuo.
11B NMR signals at δ ϭ Ϫ0.3 (weak), 30.8 (main signal), and 42.0
a solution. At 0 °C, colorless crystals separated within several days
Table 2. Crystallographic parameters, information about data collection, and structure solution of compounds 1Ϫ5
Compound
1
2
3
4
5
Empirical formula
Formula mass
Crystal size [mm]
Crystal system
Space group
C20H20BN3
313.20
0.10 ϫ 0.20 ϫ 0.20
monoclinic
I2/a
19.015(1)
7.5388(5)
23.557(2)
90
90.446(1)
90
3376.8(4)
8
1.232
0.073
1328
C24H22BN3
363.26
C18H18BN3
287.16
0.20 ϫ 0.30 ϫ 0.30
monoclinic
P21/c
10.0512(8)
19.902(2)
8.3524(7)
90
113.114(1)
90
1536.6(2)
4
1.241
0.074
608
C27H21BN6
440.31
C15H15BN6
290.14
0.1 ϫ 0.2 ϫ 0.7
monoclinic
P21/c
16.249(1)
5.4326(4)
16.331(1)
90
96.031(9)
90
1433.6(2)
4
1.344
0.085
608
0.10 ϫ 0.10 ϫ 0.30
triclinic
0.10 ϫ 0.10 ϫ 0.20
triclinic
¯
¯
P1
P1
˚
a [A]
9.413(1)
9.776(1)
11.435(1)
93.178(2)
107.399(2)
93.886(2)
998.5(2)
2
7.2478(8)
11.622(1)
13.531(2)
89.536(2)
78.233(2)
86.265(2)
1113.4(2)
2
˚
b [A]
˚
c [A]
α [°]
β [°]
γ [°]
3
˚
V [A ]
Z
ρ(calcd.) [Mg/m3]
1.208
0.071
384
1.313
0.080
460
µ [mmϪ1
F(000)
]
Index range
Ϫ24 Յ h Յ 24
Ϫ10 Յ k Յ 8
Ϫ29 Յ l Յ 29
57.74
Ϫ11 Յ h Յ 11
Ϫ12 Յ k Յ 12
Ϫ14 Յ l Յ 14
57.88
Ϫ13 Յ h Յ 6
Ϫ24 Յ k Յ 24
Ϫ10 Յ l Յ 10
58.30
Ϫ9 Յ h Յ 9
Ϫ14 Յ k Յ 15
Ϫ17 Յ l Յ 17
58.00
Ϫ19 Յ h Յ 19
Ϫ6 Յ k Յ 6
Ϫ19 Յ l Յ 19
51.70
2θ [°]
Temp. [K]
188(2)
9133
3221
2630
191(2)
5816
3045
2049
193(2)
9024
2546
2060
193(2)
6528
3420
1593
200(2)
7902
2740
1873
Refl. collected
Refl. unique
Refl. observed (4σ)
R (int.)
0.0206
0.0288
0.0252
0.0301
0.0393
No. of variables
Weighting scheme[a]
x/y
221
0.6340/1.2204
256
0.1100/0.0000
202
0.0640/0.3750
311
0.0480/0.000
259
0.1496/0.000
GOOF
1.032
0.980
1.026
0.843
0.684
Final R (4σ)
Final wR2
Larg. res. peak [e/A ]
0.0410
0.1021
0.143
0.0564
0.1456
0.219
0.0420
0.1067
0.147
0.0659
0.1751
0.291
0.0355
0.0928
0.157
3
˚
[a]
2
2
2
wϪ1 ϭ σ2Fo ϩ (xP)2 ϩ yP; P ϭ (Fo ϩ 2 Fc )/3.
1142
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145

Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes
FULL PAPER
(weak, Bu2BNPh2). In a similar experiment, pyridine was added as
an auxiliary base to the metallated solution of 2. A yellow solution
resulted. At 0 °C crystals separated, and proved on X-ray crystallo-
mL). The solution was cooled to Ϫ78 °C, stirred, and allowed to
react with a hexane solution of BuLi (1.56 , 3.3 mL). After the
solution had attained ambient temperature, the slightly turbid, yel-
graphy to be LiNPh2·2 py (14). The supernatant solution showed low solution showed 11B NMR signals at δ ϭ 1.3 and Ϫ9.5. Solv-
11B NMR signals at δ ϭ Ϫ17.6, Ϫ16.4, 0.2, 1.9, 3.0, and 43.2. The
last signal is most probably due to Bu2BNPh2 (compounds of type
Bu2BϪNR2 show resonances in the range of δ ϭ 51Ϫ45).[18]
ents were evaporated from the clear filtrate in vacuo to leave 30
mL of a solution, which was layered with hexane (20 mL) and kept
at Ϫ20 °C. Within 1 d colorless crystals had separated. These
rapidly became turbid on standing at ambient temperature, and
did so more rapidly in vacuo; no elemental analysis was therefore
performed. The identity of 18 was ascertained by X-ray structure
determination. The same reaction was performed in ether instead
of THF. In this case, more solid material separated. The solid, dis-
solved in THF, showed only a single 11B NMR signal at δ ϭ 35.0
(h1/2 ϭ 1330) corresponding to 18 was observed. The filtrate gave
11B resonances at δ ϭ 1.5, 0.6, Ϫ24.6, and Ϫ28.6.
Triple Metallation of Trianilinoborane: Compound 3 (0.97 g, the
compound contained chloride!) was dissolved in ether (80 mL) and
cooled to Ϫ78 °C, and BuLi in hexane (1.56 , 10.2 mmol, 6.5 mL)
was added with stirring. After the solution had attained ambient
temperature, a white precipitate appeared. The yellow solution
showed an 11B NMR signal at δ ϭ 35.4 (h1/2 ϭ 640 Hz). Solvent
was removed until 30 mL of a solution remained, from which crys-
tals separated within a few hours at 0°. The crystals readily lost
coordinated ether in vacuo; if drying was carried out carefully,
Triple Metallation of Tris(8-quinolylamino)borane: In analogy to
the lithiation of compound 4, compound 5 (0.97 g, 2.2 mmol), dis-
however,
the
composition
of
the
crystals
was
B(NLiPh)3·LiCl·3OEt2 (15A). The 11B NMR spectrum of the solid, solved in THF (90 mL), was metallated at Ϫ78 °C with BuLi in
dissolved in THF, was the same as that of the solution. The NMR
spectroscopic data were consistent with the formation of
hexane (1.56 , 4.2 mL). An intensively violet solution formed. At
ambient temperature, one 11B NMR signal at δ ϭ 32.8 (h1/2
ϭ
B(NLiPh)3·nOEt2. NMR in C6D6: δ1H ϭ 0.92 (t, JH,H ϭ 7.0 Hz,
1240 Hz) and a single 7Li signal at δ ϭ 2.46 were found. All at-
3
18 H, Me), 3.14 (q, 3JH,H ϭ 6.9 Hz, 12 H, OCH2), 6.85 (t, 3JH,H ϭ tempts to induce crystallization by cooling or dissolving the residue
3
7.0 Hz, 3 H, p-H), 6.96 (d, JH,H ϭ 7.5 Hz, 6 H, o-H), 7.23 (t, (after removal of THF and hexane) in ether, toluene, ether/pyridine,
3JH,H ϭ 7.0 Hz), 6 H, m-H); δ13C ϭ 14.6 (Me), 64.9 (OCH2), 117.7
or pyridine failed, only violet oils being observed. A solution of
(p-C), 125.8 (o-C), 128.8 (m-C), 160.0 (CNLi); δ11B ϭ 35.4 (h1/2
ϭ
the oil in [D8]THF revealed more H NMR signals than expected.
1
640 Hz); δ7Li ϭ 1.24. C30H45BClLi4N3O3(569.71): calcd. C 63.19,
Double Metallation of 4 and Formation of Tris(trimethylstannyl-2-
pyridylamino)borane (23): A solution of 4 (0.40 g, 1.4 mmol) in
THF (80 mL) was metallated at Ϫ78 °C with a hexane solution of
H 7.90, N 7.37; found C 62.09, H 7.78, N 7.89.
Triple Metallation of Tris(2-pyridylamino)borane and Formation of
18: Compound 4 (0.490 g, 1.69 mmol) was dissolved in THF (40 LiBu (1.77 mL, 2.80 mmol). The solution at ambient temperature
Table 3. Crystallographic parameters, information about data collection, and structure solution of compounds 7, 13, 14, 15A, 18, and 23
7
13
14
15A
18
23
Empirical formula
Formula mass
Crystal size [mm]
Crystal system
Space group
C48H48B2Li2N8O2
804.451
C32H40Li2N2O4
530.54
C44H40Li2N6
666.70
C30H45BClLi4N3O3 C36H49BLi4N8O4 C24H39BN6Sn3
569.71
696.40
778.548
0.05 ϫ 0.08 ϫ 0.20
monoclinic
P21/n
0.50 ϫ 0.50 ϫ 0.65 0.02 ϫ 0.20 ϫ 0.30 0.10 ϫ 0.30 ϫ 0.30 0.10 ϫ 0.20 ϫ 0.20 0.3 ϫ 0.3 ϫ 0.4
triclinic
triclinic
monoclinic
P21/c
20.986(2)
9.259(1)
20.893(3)
90
113.670(2)
90
3718.3(7)
4
monoclinic
C2/c
triclinic
¯
¯
¯
P1
P1
P1
˚
a [A]
13.354(2)
13.719(2)
13.812(2)
88.87(2)
79.41(2)
60.43(2)
2155.6(5)
2
10.122(4)
10.876(4)
14.756(5)
82.67(1)
74.055(7)
89.148(8)
1548.8(9)
2
13.487(1)
20.448(2)
24.929(2)
90
100.162(1)
90
11.9799(9)
12.954(1)
14.859(1)
80.759(1)
67.421(1)
66.129(1)
1947.0(3)
2
12.692(1)
15.230(1)
15.878(2)
90
93.987(12)
90
˚
b [A]
˚
c [A]
α [°]
β [°]
γ [°]
3
˚
V [A ]
6767.0(9)
8
3061.8(5)
4
Z
ρ(calcd.) [Mg/m3]
1.2394(3)
0.076
848
1.138
0.073
568
1.191
0.070
1408
1.118
0.145
1.188
0.077
1.6890(3)
2.450
µ [mmϪ1
F(000)
]
2432
740
1520
Index range
Ϫ17 Յ h Յ 17
Ϫ18 Յ k Յ 17
Ϫ18 Յ l Յ 15
56.08
Ϫ12 Յ h Յ 7
Ϫ7 Յ k Յ 11
Ϫ10 Յ l Յ 18
57.92
Ϫ26 Յ h Յ 23
Ϫ11 Յ k Յ 11
Ϫ26 Յ l Յ 25
58.26
Ϫ18 Յ h Յ 13
Ϫ26 Յ k Յ 26
Ϫ30 Յ l Յ 30
58.56
Ϫ14 Յ h Յ 14
Ϫ16 Յ k Յ 16
Ϫ20 Յ l Յ 13
57.34
Ϫ15 Յ h Յ 13
Ϫ18 Յ k Յ 18
Ϫ19 Յ l Յ 19
51.84
2θ [°]
Temp. [K]
200(3)
15632
9511
4756
0.0712
554
0.0510/0
0.827
193(2)
1792
1738
554
0.0087
361
0.0472/0
0.769
193(2)
20944
6024
2782
0.1314
469
0.05580/0
0.970
193(2)
19239
6222
3164
0.0692
405
0.07440/0.8627
1.011
193
11340
5919
4242
0.0237
479
0.1022/0.3850
1.061
200(3)
17655
5898
3695
0.0705
316
0.01320/0
0.785
Refl. collected
Refl. unique
Refl. observed (4σ)
R (int.)
No. of variables
Weighting scheme[a] x/y
GOOF
Final R (4σ)
Final wR2
Larg. res. peak [e/A ]
0.0506
0.1060
0.0566
0.1296
0.118
0.0711
0.1189
0.206
0.0613
0.1338
0.287
0.0559
0.1586
0.340
0.0309
0.0498
3
˚
[a]
2
2
2
wϪ1 ϭ σ2Fo ϩ (xP)2 ϩ yP; P ϭ (Fo ϩ 2 Fc )/3.
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145
1143

U. Braun, T. Habereder, H. Nöth, H. Piotrowski, M. Warchhold
FULL PAPER
[10]
showed 11B NMR signals at δ ϭ 1.4, 1.2, Ϫ0.9, and Ϫ3.3, in addi-
tion to a dominant signal at δ ϭ 23.5. While stirring, a solution of
Me3SnCl (0.28 g, 1.4 mmol) in THF was added. A 119Sn NMR
spectrum taken after 2 h showed signals at δ ϭ 30.8, 26.2, and 9.2,
the last signal resulting from 23. The resonance at δ ϭ 26.2 most
probably corresponds to the monostannylation product 21. Two
signals were observed in the 11B NMR spectrum, at δ ϭ 30.9 and
22.9. THF was removed in vacuo. A waxy material was left behind,
and was treated with toluene (20 mL). Crystals of 23 separated on
cooling of the solution to Ϫ20 °C. Yield: 0.33 g (31%), m.p. 87 °C.
NMR in C6D6: δ1H ϭ 0.39 (br., 27 H, Me), 6.18 [br., 3 H, C(5)H],
6.24 [m, 3 H, C(3)H], 7.84 [br., 3 H, C(6)H]; δ13C ϭ Ϫ0.87 (Me),
113.3 (C3), 137.3 (C4), 146.5 (C6); δ11B ϭ 30.9 (h1/2 ϭ 370 Hz);
δ119Sn ϭ 9.2. C24H39BN6Sn3 (777.88): calcd. C 37.02, H 5.01, N
10.80; found C 37.16, H 5.16, B 10.62.
H. Nöth, H. Vahrenkamp, J. Organomet. Chem. 1968, 12,
23Ϫ36.
C. D. Habben, A. Heine, G. M. Sheldrick, D. Stalke, Z. Natur-
[11]
[12]
[13]
forsch., Teil B 1992, 47, 1367Ϫ1369.
R. A. Bartlett, H. Chen, H. V. R. Dias, M. M. Olmstead, P. P.
Power, J. Am. Chem. Soc. 1988, 110, 446Ϫ449.
B. Wrackmeyer, B. Schwarze, J. Weidinger, M. Milius, Z. Nat-
urforsch., Teil B 1997, 52, 431Ϫ434.
P. Paetzold, C. Pelzer, R. Boese, Chem. Ber. 1988, 121, 51Ϫ59.
T. Seifert, W. Storch, M. Vosteen, Eur. J. Inorg. Chem. 1998,
1343Ϫ1349.
[14]
[15]
[16]
´
A. Sapse, P. von Rague-Schleyer, Lithium Chemistry Ϫ A
Theoretical and Experimental Overview, Wiley Interscience
Publ., New York, London, Sydney, 1995.
R. E. Mulvey, Chem. Soc. Rev. 1991, 20, 167Ϫ209.
A. Troll, Ph.D. Thesis, University of Munich, 1999.
W. D. English, A. L. McCloskey, H. Steinberg, J. Am. Chem.
Soc. 1960, 82, 2122Ϫ2124.
R. J. Brotherton, T. Buckman, Inorg. Chem. 1963, 2, 424Ϫ425.
Handbook of Chemistry and Physics, 66th ed., CRC Press, Inc.,
Boca Raton, FL, USA, 1986, D-159.
A. Streitwiese, Jr., C. H. Heathcock, Organische Chemie, Verlag
Chemie, Weinheim, Deerfield Beach, Basel, 1980, p. 1392.
This compound has also been investigated by R. Contreras, M.
J. Grevy, private communication, 1999.
H. Nöth, B. Wrackmeyer, Chem. Ber. 1974, 107, 3070Ϫ3103.
G. J. Martin, M. L. Martin, J.-P. Gousenard, 15N NMR Spec-
troscopy, in: Basic Principles and Progress (Eds.: P. Diehl, E.
Fluck, R. Kosfeld), Springer Verlag, Berlin, Heidelberg, New
York, 1981.
H. Nöth, Z. Naturforsch., Teil B 1983, 38, 692Ϫ695.
H. Nöth, R. Ullmann, H. Vahrenkamp, Chem. Ber. 1973,
106, 1165Ϫ1171.
H. Nöth, B. Wrackmeyer, Nuclear Magnetic Resonance of
Boron Compounds, in: NMR Basic Principles and Progress
(Eds.: P. Diehl, E. Fluck, R. Kosfeld), Springer Publ., Berlin,
Heidelberg, New York, 1978.
[17]
[18]
[19]
X-ray Structure Determinations: Crystals were covered with per-
fluoroether oil Ϫ those that readily lost coordinated solvent in pre-
cooled oil (approximately Ϫ40 °C). The selected crystal was placed
on a glass fiber and mounted on the goniometer head. The
reflections on five sets of frames were used for a preliminary deter-
mination of the unit cell, by use of the program implemented in
SMART.[51] Data were collected in the hemisphere mode, and re-
duced with the SAINT program.[51] For absorption correction the
SADABS[51] program was employed. The structures were solved by
direct methods implemented in the SHELX97 program package.[52]
Non-hydrogen atoms were refined in an isotropic description. Hy-
drogen atoms were placed in calculated positions and included in
the final cycles of refinement by application of a riding model.
One disordered ether molecule was found in the structure of
compound 15A, while compound 18 contains one uncoordinated
THF molecule for each molecule of 18. Tables 2 and 3 summarize
data relevant to crystallography, data collection and refinement.
CCDC-168322 to -168332 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge at www.ccdc.cam.ac.uk/conts/retrieving.html or from the
Cambridge Crystallographic Data Centre, 12, Union Road, Cam-
bridge CB2 1EZ, UK [Fax: (internat.) ϩ 44-1223/336-033; E-mail:
deposit@ccdc.cam.ac.uk].
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
None of the crystals showed reflections at 2θ Ͼ 20°, and all
reflections were weak and blurred even when fairly large crys-
tals were investigated.
Only a very few 11B NMR spectroscopic data for triaminobor-
ates
or
tetraaminoborates
are
known,
e.g.
Me3SiNHB[NH(CN2C3H6)]3 (δ11B ϭ Ϫ4.8) (R. Contreras,
private communication), [BuB(NPhMe)3]Ϫ (δ11B ϭ Ϫ4.2) (M.
Warchhold, private communication, 2001). See also ref.[27]
The origin of LiCl in the compound could be traced back to
impure trianilinoborane obtained in the transamination of
(Me2N)2BCl containing tris(dimethylamino)borane. From so-
lutions of triply metallated pure B(NHPh)3 crystallization
was unsuccessful.
Ϫ
[31]
Acknowledgments
We thank the Fonds der Chemischen Industrie for support of this
work, as well as Chemetall GmbH for support and generous gifts
of butyllithium.
[32]
[33]
B. Wrackmeyer, Ann. Rep. NMR Spectrosc. 1985, 16, 73Ϫ186.
S. H. Bauer, J. Y. Beach, J. Am. Chem. Soc. 1941,
63,1394Ϫ1399.
P. T. Brain, A. J. Downs, P. Maccallum, D. W. H. Rankin, H.
E. Robertson, G. A. Forsyth, J. Chem. Soc., Dalton Trans.
1991, 1195Ϫ1199.
K. J. Weese, R. A. Bartlett, B. D. Murray, M. M. Olmstead, P.
P. Powe r, Inorg. Chem. 1987, 26, 2409Ϫ2413.
G. Beck, P. B. Hitchcock, M. F. Lappert, I. A. MacKinneon,
J. Chem. Soc., Chem. Commun. 1989, 1312Ϫ1314.
M. Westerhausen, W. Schwarz, Z. Anorg. Allg. Chem. 1993,
619, 1053Ϫ1063.
W. Lippert, Ph.D. Thesis, University of Munich, 1999.
T. Seifert, W. Storch, M. Vosteen, Eur. J. Inorg. Chem. 1998,
1343Ϫ1349.
A. Appel, C. Kober, C. Neumann, H. Nöth, M. Schmidt, W.
Storch, Chem. Ber. 1996, 129, 175Ϫ190, and ref. cited therein.
U. Höbel, H. Nöth, H. Prigge, Chem. Ber. 1986, 119, 325Ϫ337.
V. A. Sazanova, L. P. Sorokina, Dokl. Akad. Nauk SSSR 1955,
105, 993Ϫ996.
[34]
[1]
B. Böck, H. Nöth, U. Wietelmann, Z. Naturforsch., Teil B
2001, 56, 659Ϫ670.
[35]
[36]
[37]
[2]
H. Fußstetter, G. Kopietz, H. Nöth, Chem. Ber. 1980, 113,
728Ϫ738.
[3]
I. Geisler, H. Nöth, Chem. Ber. 1973, 106, 1943Ϫ1951.
[4]
U. Braun, Ph.D. thesis, University of Munich, 2000.
[5]
H. Fußstetter, R. Kroll, H. Nöth, Chem. Ber. 1977, 110,
3829Ϫ3842.
[38]
[39]
[6]
D. Fest, C. D. Habben, A. Meller, G. M. Sheldrick, D. Stalke,
F. Pauer, Chem. Ber. 1990, 123, 703Ϫ706.
H. Chen, R. A. Bartlett, M. M. Olmstead, P. P. Power, S. C.
[7]
[40]
Shoner, J. Am. Chem. Soc. 1990, 112, 1048Ϫ1055.
[8]
H. Nöth, H. Vahrenkamp, J. Organomet. Chem. 1969, 16,
[41]
[42]
357Ϫ369.
[9]
W. Beck, W. Becker, H. Nöth, B. Wrackmeyer, Chem. Ber. 1972,
105, 2823 Ϫ2837.
1144
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145

Triaminoboranes and Their Metallation to N-Lithiotriaminoboranes
FULL PAPER
D. J. English, A. L. McCloskey, H. Steinberg, J. Am. Chem.
Soc. 1961, 83, 2122Ϫ2124.
H. Beyer, K. Niedenzu, J. W. Dawson, J. Organomet. Chem.
1962, 27, 4701Ϫ4702.
SMART, SADABS, SAINT, Siemens Analytical Instruments
Madison, Vers. 5.1.
[43]
[49]
[50]
[51]
[52]
V. A. Sazanova, V. I. Karpov, J. Gen. Chem. USSR 1963, 33,
3241Ϫ3243.
[44]
K. Niedenzu, P. M. Niedenzu, K. R. Warner, Inorg. Chem.
1985, 24, 1604Ϫ1606.
[45]
K. Niedenzu, S. Trofimenko, Top. Curr. Chem. 1986, 131,
1Ϫ37.
[46]
M. Warchhold, Ph.D. Thesis, University of Munich, 2001.
G. W. Sheldrick, SHELXL97, University of Göttingen.
Received August 1, 2001
[47]
B. Eichhorn, H. Nöth, T. Seifert, Eur. J. Inorg. Chem. 1999,
2355Ϫ2368.
[I01283]
[48]
T. T. Wang, K. Niedenzu, J. W. Dawson, Inorg. Chem. 1970,
9, 2150Ϫ2152.
Eur. J. Inorg. Chem. 2002, 1132Ϫ1145
1145