Ambident PCN heterocycles: N- and P-phosphanylation of lithium 1,3-benzazaphospholides

Article abstract of DOI:10.1002/chem.200901753

Synthetic and structural aspects of the phosphanylation of 1,3benzazaphospholides 1_L1, ambident benzofused azaphosphacyclopentadienides, are presented. The unusual properties of phospholyl-l,3,2-diazaphospholes inspired us to study the coupling

Full text of DOI:10.1002/chem.200901753

FULL PAPER
DOI: 10.1002/chem.200901753
Ambident PCN Heterocycles: N- and P-Phosphanylation of
Lithium 1,3-Benzazaphospholides
Bhaskar R. Aluri,[a] Sebastian Burck,[b] Dietrich Gudat,*[b] Mark Niemeyer,[c]
Oldamur Holloczki,[d] Laszlo Nyulaszi,*[d] Peter G. Jones,[e] and Joachim Heinicke*[a]
Dedicated to Professor Reinhard Schmutzler on the occasion of his 75th birthday
¼
Abstract: Synthetic and structural as-
pects of the phosphanylation of 1,3-
benzazaphospholides 1Li, ambident
benzofused azaphosphacyclopentadi-
the case of 7 they are isolated as a di-
meric LiCl(THF) adduct. Structural in-
formation was provided by single-crys-
tal X-ray diffraction and solution NMR
spectroscopy experiments. 2D ex-
change spectroscopy confirmed the ex-
ring phosphorus (DH =22 kJmolꢀ1,
¼
G
DS =2 JKꢀ1 molꢀ1) and very low-tem-
perature rotation of the tBu2P group.
Quantum chemical studies give evi-
dence that 2-unsubstituted benzaza-
phospholides prefer N-phosphanyla-
tion, even with bulky chlorophos-
phanes, and that substituents at the 2-
position of the heterocycle are crucial
for the occurrence of P–N rotamers
and for switching to alternative P-sub-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
A
U
istence of two rotaACTHNGUETRNNUmG ers of the amino-
coupling of 1Li with chlorodiazaphos-
pholene 2, which led to the N-substitut-
ed product 3. Reaction of 1Li with
chlorodiphenyl- and chlorodicyclohex-
ylphosphane likewise gave N-phospha-
nylbenzazaphospholes 4 and 5, whereas
with the more bulky di-tert-butyl- and
di-1-adamantylchlorophosphanes, the
diphosphanes 6 and 7 are obtained; in
phosphane 5 at room temperature; var-
iable-temperature NMR spectroscopy
studies of 6 revealed two dynamic pro-
cesses, low-temperature inversion at
stitution, beyond
a threshold steric
bulk, by both P- and 2-position sub-
stituents.
Keywords: ambident reactivity
·
dynamic processes · heterocycles ·
phosphanes · quantum chemistry
Introduction
anions.[4] Their reactivity, however, is controlled by the het-
eroatoms. Computations on gas-phase structures show that
lithium 1,3-azaphospholide is h5-coordinated.[5] In the pres-
ence of donor solvents—for example, THF—lithium coordi-
Azaphospholides with one[1] or more nitrogen atoms[2,3] in
the ring are structurally related to the cyclopentadienide
[a] Dr. B. R. Aluri, Prof. Dr. J. Heinicke
Institut fꢀr Biochemie—Anorganische Chemie
Ernst-Moritz-Arndt-Universitꢁt Greifswald
Felix-Hausdorff-Strasse 4, 17487 Greifswald (Germany)
Fax : (+49)3834-864377
[d] O. Holloczki, Prof. Dr. L. Nyulaszi
Department of Inorganic Chemistry
Technical University of Budapest, 1521 Budapest (Hungary)
E-mail: Nyulaszi@mail.bme.hu
[e] Prof. Dr. P. G. Jones
E-mail: heinicke@uni-greifswald.de
Institut fꢀr Anorganische und Analytische Chemie
Technische Universitꢁt Braunschweig
Hagenring 30, 38106 Braunschweig (Germany)
[b] Dr. S. Burck, Prof. Dr. D. Gudat
Institut fꢀr Anorganische Chemie
Universitꢁt Stuttgart
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901753. It includes 31P and
13C NMR spectroscopic data, determination of activation enthalpy and
entropy for the inversion at the ring phosphorus atom in 6, details on
quantum chemical calculations, and detailed X-ray structural data of
1, 3, 5, and 7.
Pfaffenwaldring 55, 70550 Stuttgart (Germany)
E-mail: gudat@iac.uni-stuttgart.de
[c] Dr. M. Niemeyer
Institut fꢀr Anorganische und Analytische Chemie
Johannes Gutenberg-Universitꢁt Mainz
Duesbergweg 10–14, 55128 Mainz (Germany)
Chem. Eur. J. 2009, 15, 12263 – 12272
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12263

nates preferably at nitrogen, as shown by crystal-structure
analysis of a more easily accessible lithium 1,3-benzazaphos-
pholide THF trisolvate, that exhibits only a marginal upfield
phosphorus resonance (Dd<10 ppm) with respect to the un-
complexed system. Zero-valent Group 6 metal carbonyls
bind at phosphorus.[6] The reactivity towards non-zero-
valent metal and other electrophilic compounds is different
for mono- and di- or triazaphospholides. Whereas the latter
prefer reaction at a basic dicoordinate nitrogen atom,[2,3] the
1,3-azaphospholides[1] and benzazaphospholides[6,7] display
ambident reactivity that was interpreted by simple hard and
soft acid and base (HSAB) considerations. The real behav-
ior is, however, more subtle and depends also on steric fac-
tors, as seen by the different sites of trimethylsilylation in
various 2-substituted 1,3-benzazaphospholides[6a] and in
analogous arsenic heterocycles.[8] Reactions of such anions
with chlorophosphanes are not yet known. Only the sparing-
corresponding benzazaphospholyldiazaphospholenes prefer
ꢀ
ꢀ
N P- or P P-bonded structures, we have extended our
study to include coupling of 1Li with 2-chloro-1,3-dimesityl-
4,5-dimethyl-1,3,2-diazaphospholene (2)[14] and present the
spectroscopic and structural characterization of the product,
together with a computational study of structural aspects
and relative stabilities of N- or P-phosphanyl-substituted
1,3-benzazaphospholes.
Results and Discussion
2,5-Dimethyl-1H-1,3-benzazaphosphole (1) was synthesized
by nickel-catalyzed phosphonylation of commercially avail-
able 2-bromo-4-methylacetanilide and subsequent reductive
cyclization with excess LiAlH4, the shortest route to these
1H-1,3-benzazaphospholes.[15] Lithiation was performed by
using tert-butyllithium at low temperature in THF. Under
these conditions the bulky lithium reagent prefers NH met-
ly characterized N-dichlorophosphanylbenzazaphosACTHUNTGRNEUNGphole,
formed by prolonged reflux of neat 1H-1,3-benzazaphos-
phole in excess PCl3, was reported. Attempts at coupling
with R2PCl failed,[7] whereas di- and triazaphospholes under-
go clean N-substitution without metalation.[9] To shed more
light on the behavior of the P,N-heterocyclic anions and the
nature of phosphanylation products, the reactivity of an
easily accessible lithium benzazaphospholide 1Li towards se-
lected chlorophosphanes was studied.
[7]
alation to addition at the P=C bond. Lithiation by LiNEt2
or lithium diisopropylamide (LDA)[8] is less suitable, as it re-
quires complete removal of Et2NH or iPr2NH to avoid com-
peting reactions with chlorophosphanes. The conversion of 1
with the chloro-1,3,2-diazaphospholene 2 took place with
high selectivity at nitrogen and furnished crystalline benza-
zaphosphol-1-yl-diazaphospholene
3
in good yield
ꢀ
Apart from the regioselectivity problem, N- and P-phos-
phanyl-1,3-benzazaphospholes are of interest because of the
different electronic structure and properties of the 1H- and
3H-1,3-benzazaphosphole sys-
(Scheme 1). The increased polarity and reactivity of the Cl
P bond of 2[14] may have contributed to the clean course of
the reaction.
tems.[10] In 1-phosphanyl-1,3-
benzazaphospholes, the nitro-
gen lone pair is still part of the
delocalized p system, which is
expected to cause weakening of
ꢀ
the P N bond compared with
ordinary aminophosphanes. In
contrast, 3-phosphanyl-1,3-ben-
zazaphospholes featuring a pyr-
amidal phosphorus atom should
be destabilized by a loss of aro-
ꢀ
maticity. Furthermore, the P P
bond strength differs from that
ꢀ
of the P N bond, and also the
P=C and N=C double-bond en-
ꢀ
ergies are different. P P bond
polarization, however, with con-
comitant buildup of negative
partial charge at the ring phos-
Scheme 1. Lithiation of 1 and coupling of 1Li with chlorophosphanes.
phorus atom, might allow for
some residual stabilization by
weak p interactions. An effect
of this type has been observed for 2-phospholyl-1,3,2-diaza-
To find out whether chlorophosphanes generally prefer
substitution of 1Li at nitrogen, or if this is governed by elec-
tronic and/or steric effects, the investigation was extended
to coupling of 1Li with Ph2PCl, Cy2PCl, tBu2PCl, and 1-
Ada2PCl (Cy=cyclohexyl; Ada=adamantyl) as representa-
phospholenes, which display perceptible lengthening of the
[11]
ꢀ
P P bond in connection with an increased bond ionicity
and are consequently highly reactive in insertion reactions
into polar single[12] and multiple bonds.[13] To find out if the
12264
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12263 – 12272

Ambident PCN Heterocycles
FULL PAPER
Table 1. Selected bond lengths [ꢃ] and angles [8] of the benzazaphos-
AHCTUNGTRENNUNG
tives of diaryl- and dialkylchlorophosphanes with varying
steric bulk. The reaction of 1Li and chlorodiphenylphos-
phane did not proceed in a well-defined way (at ꢀ788C)
and provided a mixture of products. NMR spectroscopic
analysis of the crude material, obtained by removal of the
solvent and extracting the residue with diethyl ether, re-
vealed roughly 40 mol% of 4 (by integration of Me proton
signals), accompanied by about the same amount of recov-
ered 1. Smaller 31P signals indicated tetraphenyldiphosphane
and two unknowns each with 50% of the intensity of the
former and phosphorus resonances in the (3R)-benzaza-
[a]
1
3
5
ACHUTGTNNERNUG[7·LiClACHTUNGTRENNUNG(thf)]2
ꢀ
P C2
1.7284(13) 1.719(3)
1.7934(12) 1.783(3)
1.3694(16) 1.394(4)
1.3799(15) 1.415(3)
1.714(2)
1.789(2)
1.386(2)
1.410(2)
1.7622(15)
–
1.8190(16)
1.8041(16)
1.305(2)
1.418(2)
–
ꢀ
P C3A
ꢀ
N C2
ꢀ
N C7A
ꢀ
N PsubN
–
–
1.812(2)
–
ꢀ
N Li
2.135(3)
1.404(2)
ꢀ
C3A C7A
1.4128(15) 1.404(4)
113.20(9) 115.5(2)
112.01(11) 112.7(2)
110.67(9) 111.8(2)
114.60(10) 110.9(2)
89.52(6) 89.03(14)
1.406(3)
N-C2-P
115.30(14) 114.58(12)
112.44(16) 116.89(14)
111.45(14) 107.97(12)
111.66(15) 111.29(13)
89.09(9)
120.99(17) 121.85(14)
123.70(15) 123.19(11)
N-C7A-C3A
P-C3A-C7A
C2-N-C7A
C2-P-C3A
N-C2-CMe
P-C2-CMe
phosphole range (d
ACHTUNGTRENNUNG
Chlorodicyclohexylphosphane reacted, like 2, in an unam-
biguous way. The yield of the product was somewhat lower,
and in the NMR spectra two sets of signals were observed,
but both belong to the N-substitution product 5 and form
two stable rotamers (see below). Thus, the selectivity was
preserved. However, increase of the steric bulk changes this
preference. Chloro-di-tert-butylphosphane and 1Li react
under the same conditions as used in the aforementioned
couplings and with similar yield, but with attack at phospho-
rus and formation of 6. The further increased steric conges-
tion in chlorobis(1-adamantyl)phosphane leads to the same
preference for reaction at phosphorus but very slow conver-
sion. Reaction monitoring by 31P NMR spectroscopy re-
vealed that the major part of the starting materials was con-
verted only after three weeks at room temperature. Whereas
6 was isolated as a colorless oil by high vacuum distillation
(after extraction with hexane from the crude mixture), anal-
ogous workup of the 3-diadamantylphosphanylbenzaza-
phosphole 7 failed and furnished only decomposition prod-
ucts. However, layering of the concentrated solution of the
crude product in THF with hexane led to isolation of a N-
88.45(7)
118.50(11) 121.8(2)
128.30(10) 122.7(2)
PsubN-N-C2 or
Li-N-C2
–
118.05(18) 119.94(13) 124.65(14)
P
subN-N-C7A or
–
–
–
–
–
130.68(19) 127.51(13)
–
Li-N-C7A
PsubP-P-C2
PsubP-P-C3A
–
–
–
–
–
–
123.88(13)
104.97(5)
117.97(5)
ꢀ
ꢀ
P N or P C
(substituent)
N-P-N or C-P-C
(substituent)
1.706(2),
1.707(2)
89.99(11)
1.8549(19), 1.8896(16),
1.8563(19) 1.8934(16)
106.27(8)
–
–
112.09(7)
NBAP-P-N
{or C}
104.62(11),
A
ACHTUNGTRENNUNG
A
105.54(11) 103.48(8)}
ꢀ
[a] P3 P1 2.2025(6) ꢃ.
coordinated LiCl
ACHTUNGTRENNUNG(THF) adduct, the crystalline chloro-bridg-
ed dimer [7·LiCl(thf)]2.
ACHTUNGTRENNUNG
Both N- and P-phosphanyl-1,3-benzazaphospholes are
highly sensitive to moisture and are attacked by any trace of
water with formation of 1 and R2PHO, easily detectable by
31P NMR spectroscopy. Thus, 6 is cleaved to give 1 and
Figure 1. Crystal structure of 1. Ellipsoids represent 50% probability
tBu2PHO (dACHTUNGTRENNUNG
(31P)=73.5 and 61.9 ppm). The strongly in-
ꢀ
ꢀ
levels. Selected bond lengths [ꢃ] and angles [8]: P C2 1.7284(13), P
ꢀ
creased reactivity of the P N bond relative to common ami-
nophosphanes and the easy hydrolysis of the normally non-
ꢀ
ꢀ
ꢀ
C3A 1.7934(12), C2 N 1.3694(16), N C7A 1.3799(15), C3A C7A
1.4128(15); N-C(2)-P 113.20(9), C(2)-P-C(3A) 89.52(6).
ꢀ
polar P P bond suggests weakened and/or strongly polar-
ꢀ
ꢀ
ized or polarizable P N and P P bonds.
Structural aspects: The molecular structures of 3–7 were
confirmed unambiguously by characteristic or full sets of so-
lution NMR spectroscopic data. For more detailed informa-
tion, the crystal structures of 1, 3, 5, and 7 have been deter-
mined. Compound 1 was included to visualize the changes
in bond lengths and angles (Table 1) caused by the N- or P-
substituents. Finally, NMR spectroscopic studies of 5 and 6
addressing the occurrence of rotamers and dynamic behav-
ior in solution were performed.
system (mean deviation 0.03 ꢃ). Bond lengths and angles
are comparable to those in other 1H-1,3-benzazaphos-
AHCTUNGTRENNUNG
pholes.[15a,16] The 2-methyl group is slightly bent towards the
N atom (N-C2-C8 118.50(11)8 versus P-C2-C8 128.30(10)8),
but does not severely hinder access to the nitrogen atom in
the metalation and substitution reactions, except by very
bulky electrophiles.
The benzazaphospholyldiazaphospholene 3 cocrystallizes
with one molecule of deuterobenzene. The molecular struc-
ture (Figure 2) confirms bonding of the diazaphospholene
unit to the nitrogen atom of the azaphosphole ring and
shows that the planarity of the benzazaphosphole is con-
Crystal structures: The molecular structure of compound 1
(Figure 1) shows the expected planar benzazaphosphole ring
Chem. Eur. J. 2009, 15, 12263 – 12272
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12265

D. Gudat, L. Nyulaszi, J. Heinicke et al.
Figure 2. Molecular structure of 3·C6D6 in the crystal. Ellipsoids represent
30% probability levels. Hydrogen atoms and the cocrystallized deutero-
benzene molecule have been omitted for clarity.
Figure 3. Molecular structure of 5 in the crystal. Ellipsoids represent
50% probability levels.
ꢀ
served (average deviation 0.03 ꢃ). The shape of the diaza-
phospholene ring resembles that in other amino-substituted
diazaphospholenes[17] and is not strictly planar in contrast to
genuine diazaphospholenium cations.[14] The phosphorus
atom lies 0.055 ꢃ above the N5-C4-C3-N2 plane towards
with 1 the C N bonds of the azaphosphole ring are similarly
ꢀ
ꢀ
elongated in 3 and 5, and the P C and C C bonds are
slightly shortened. In the anionic 1Li,[6a] however, the
changes with respect to 1 are different. Finally it should be
mentioned that introduction of N-substituents causes slight
elongation of the P-C-N and N-C-C(Me) angles and reduc-
tion of the C-N-C angles (Table 1).
ꢀ
the benzazaphosphole group, and the exocyclic P N bond,
which forms an angle of 718 to the N5-C4-C3-N2 plane,
fixes the diazaphospholene ring like a bisecting roof (N2-P1-
N26-C27 130.5(2)8 versus N5-P1-N26-C27 ꢀ135.5(2)8) over
the benzazaphosphole ring. The
The P-substituted benzazaphoshole 7 crystallized coordi-
nated to LiClACTHUNTGRNEUNG(THF). The molecule (Figure 4) forms an in-
planar mesityl groups are ar-
ranged with dihedral angles of
ꢀ77.1(3) and 65.3(4)8 for C4-
N5-C6-C7 and C3-N2-C15-C16,
ꢀ
respectively. The exocyclic P N
bond (1.812(2) ꢃ) is considera-
ꢀ
bly longer than the P N bonds
within the ring (Table 1) and
also slightly longer than the
ꢀ
comparable
exocyclic
P N
bond of a pyrazolyldiazaphos-
pholene (1.798(2) ꢃ).[17] The
lengthening relative to the exo-
ꢀ
cyclic P N bond in 5 reflects a
local bond weakening in favor
Figure 4. Molecular structure of [7·LiClACTHUNRGTNEUNG(thf)]2 in the crystal. Ellipsoids represent 50% probability levels. Se-
lected bond lengths [ꢃ] and angles [8]: O(91)-Li-N(1) 98.01(13), O(91)-Li-Cl 107.76(13), O(91)-Li-Cl#1
109.92(14), N(1)-Li-Cl 122.22(14), N(1)-Li-Cl#1 120.17(13), Cl#1-Li-Cl 98.36(11), Li#1-Cl-Li 81.64(11); further
data in Table 1.
of
electronic
stabilization
within the diazaphospholene
ring. Geometry and bond labili-
ty resemble those of adducts of
neutral N-heterocyclic carbenes (Lewis base) at N-heterocy-
clic silylenes, germylenes, or stannylenes (here acting as
Lewis acids).[18] Compound 5, depicted in Figure 3, also dis-
plays an essentially planar heterocyclic system (mean devia-
tion 0.03 ꢃ); it bears an electronically indifferent PCy2
group at nitrogen, arranged in a similar bisecting conforma-
tion (C11-P2-N3-C2 ꢀ126.45(15)8 versus C21-P2-N3-C2
version-symmetric dimer with an Li2Cl2 rhombus. The devia-
tions from planarity within the benzazaphosphole ring are
again small (mean deviation 0.03 ꢃ); the configuration at P3
is distorted pyramidal (angle sum 311.48). The dihedral
angles N1-C2-P3-P1 127.45(11)8 and P1-P3-C3A-C7A
ꢀ
ꢀ113.12(10)8 correspond to the direction of the P3 P1 bond
out of the azaphosphole plane, whereas the dihedral angles
around the P–P axis show that one adamantyl group is ar-
ranged nearly eclipsed to C3A (C3A-P3-P1-C11 17.10(8)8)
and the other nearly anti to C2 (C2-P3-P1-C21 163.78(7)8).
ꢀ
ꢀ
123.70(14)8) with the P2 C11 and P2 C21 bonds on differ-
ent sides of the benzazaphosphole plane and directed to-
wards the benzo-end of the bicyclic ring system. Compared
12266
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12263 – 12272

Ambident PCN Heterocycles
FULL PAPER
ꢀ
The P3 P1 bond length (2.2025(6) ꢃ) lies in the normal
and C7[21] and suggest that the exocyclic PCy2 moiety adopts
an orientation in which the phosphorus lone pair remains
close to the plane of the benzazaphosphole ring and points
towards C7, and thus away from the 2-CH3 group (cf.
Scheme 1; this lone-pair orientation may also be held re-
sponsible for the observation of an additional splitting for
the signals of the adjacent carbon nuclei C7A and C6, usual-
ly doublets by coupling with P3, to triplets). In the major
isomer, the signal of C7 appears as a singlet and the methyl
carbon as a doublet of doublets because of large couplings
to both phosphorus atoms (2,3J=27.4, 31.4 Hz), thereby indi-
cating that the orientation of the PCy2 lone pair is now re-
versed. Altogether, these assignments allow us to attribute
the two observable sets of signals to two different rotamers
for which the half-life time is sufficiently large to give rise
to distinguishable sets of NMR spectroscopic signals at
room temperature. This hypothesis is further substantiated
by the appearance of cross peaks that connect the appropri-
ate signals of both rotamers in a 2D-EXSY (EXchange
SpectroscopY) spectrum (see the Supporting Information)
and confirm that both species are in dynamic equilibrium at
room temperature. The slow rate of interconversion suggests
a rotational barrier of considerable height, and evaluation of
relative populations (integration of methyl 1H NMR spectro-
scopic signals indicated a molar ratio of approximately 9:1
at 296 K). This implies that the orientation of the phospho-
rus lone pair towards the 2-methyl group and of the cyclo-
hexyl substituents towards the benzene ring represents the
more stable rotamer. Interestingly, the same orientiation of
range for diphosphanes ((2.217ꢁ0.08) ꢃ[11]), with the P3
ꢀ
ꢀ
C3A bond slightly lengthened and the P3 C2 bond appreci-
ably lengthened relative to that in 1, whereas the N1 C2
ꢀ
bond is markedly shortened. These bond lengths correspond
to the usual (3R)-benzazaphosphole structure. This, together
with the thermal stability of 6, suggests that the sensitivity
of 6 and 7 towards moisture may be caused by an easier po-
larizability compared to normal diphosphanes.
Solution structures: The most characteristic data for estab-
lishing N- or P-phosphanylation by solution spectra of 3–7
are the phosphorus chemical shifts and P,P coupling con-
stants. Compound 4 (d=92.9 (Pring), 40.5 ppm (PPh2), 3J=
7.0 Hz), which could not be isolated in pure form, was un-
ambiguously identified as a 1-diphenylphosphanylbenzaza-
phosphole by the small P,P coupling constant, which was
definitely too small for a one-bond coupling, and the charac-
teristic downfield shift of the ring-phosphorus resonance rel-
ative to 1 (Dd=21.1 ppm), found also in 3 (Dd=18.7 ppm),
5
(major rotamer Dd=13.4 ppm, minor rotamer Dd=
23.4 ppm), and 1-trimethylsilylbenzazaphospholes.[6a] 3-Phos-
phanylated benzazaphospholes display upfield shifted dou-
blets for the ring phosphorus nuclei and large one-bond P,P
coupling constants (6: Dd=ꢀ77.8 ppm, 1JP,P =425 Hz; 7:
1
Dd=ꢀ84.1 ppm, JP, P =418 Hz). The observation of substan-
tial line broadening for both doublets in the room tempera-
ture spectra of 6 and 7 was interpreted as a result of dynam-
ic exchange broadening, which in the case of 6 was con-
firmed by a variable temperature (VT) NMR spectroscopic
the PCy2 group is also present in crystalline 5 and, based on
3
study (see below). The magnitude of the P,P coupling con-
the evaluation of J
3.
ACHTUNGTRENUN(NG P,C) coupling constants, in solutions of
1
ꢀ
stants in 6 and 7 comes close to the value of JP, P in tBu2P
PtBu2 (ꢀ451 Hz[19]), which is presumably a consequence of
both steric bulk (as evidenced by the large C-P-C bond
Low-temperature NMR spectroscopic studies of compound
6: The above-mentioned dynamic behavior of 6 was studied
in more detail by means of VTNMR spectroscopy experi-
ments. Upon lowering the temperature, the 31P signals
become at first broader, then again sharper, and below
ꢀ508C they split into two sets of signals with relative inten-
sities of approximately 10:1. The proton signal of the tBu2P
group, at room temperature a broad doublet, broadened
upon cooling and then decoalesced into two signals of equal
intensity. Below ꢀ508C, all proton signals split further into
two sets of signals with relative intensities of approximately
angles of 112.1(1)8 for [7·LiClACHTNUGTRENUNG(thf)]2) and a high degree of s
[20]
ꢀ
character in the P P bond. Conclusive information on the
constitution in solution is also provided by the characteristic
13C resonances and J
ACTHUNTGRENNG(U P,C) coupling constants. Thus, N- and
P-substituted benzazaphospholes are clearly distinguished
by the chemical shifts of C3A, C7A, and C7. Compared
with the 1H-1,3-benzazaphosphole 1, the signal of C3A is
upfield shifted in (3R)- (Ddꢂꢀ14–15 ppm) and slightly
downfield shifted (Ddꢂ1–2 ppm) in (1R)-benzazaphos-
ACHTUNGTRENNUNGpholes, whereas C7A and C7 are less deshielded in (1R)-
1
(Ddꢂ3–6, 4–10 ppm) than in (3R)-benzazaphospholes (Dd
ꢂ12, 16 ppm). The effect of phosphanylation at N or P on
the chemical shift of C2 is similar, (Dd=10–12 ppm), but
(P,C) coupling remains much larger for the P=C NR
than the RP C=N heterocycles.
10:1. The observed temperature dependence of both H and
31P NMR spectra can be explained by the occurrence of two
dynamic processes, 1) inversion at the ring phosphorus
1
1
ꢀ
the J
N
atom, thereby inducing coalescence of the H NMR spectro-
ꢀ
scopic signals of the two diastereotopic tert-butyl groups
ꢀ
Solutions of 5 display two sets of phosphorus and carbon
signals at room temperature, each with similar chemical
shifts and, except for the occurrence of clearly different
above ꢀ508C, and 2) freezing the rotation around the P P
bond at low temperature, thereby leading to decoalescence
1
of all H and 31P NMR spectroscopic signals into two sets of
values for 3J
T
resonances with unequal intensities, which are attributable
to two distinguishable rotamers. The P-inversion process can
be visualized by the appearance of strong cross peaks, with
intensities comparable to those of the diagonal peaks, in
2D-EXSY spectra recorded below the coalescence tempera-
AHCTUNGTRENNUNG
33.3 Hz), whereas the signal of 2-CH3 appears as a singlet;
both findings hint at a close proximity between the lone pair
Chem. Eur. J. 2009, 15, 12263 – 12272
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12267

D. Gudat, L. Nyulaszi, J. Heinicke et al.
1
ture (Figure 5); line-shape analysis of H NMR spectra re-
are planar cyclodelocalized ring systems with the substituent
in the ring plane and more stable by approximately
11 kcalmolꢀ1 than the P-substituted structures with pyrami-
dal configuration at the ring phosphorus atom.
corded between 223 and 303 K allowed the determination of
the activation parameters for this process as DH =
¼
Extending the aromatic system by attachment of a ben-
zene ring to the 1,3-benzazaphosphole and substitution with
ꢀ
the more bulky di-tert-butylphosphanyl group instead of
PH2 caused a slight destabilization of the N- with respect to
the P-substituted isomers, however, they are still more
stable than the latter (Figure 7). The bulky N-substituent
1
Figure 5. Expansion of a H 2D-EXSY spectrum of 6 (213 K, mixing time
300 ms, only positive contour levels shown). Signals of PACTHUNRGTNEUNG(tBu)2 substitu-
Figure 7. Possible (di-tert-butylphosphanyl)benzazaphosphole isomers
(B3LYP/6-311+G**level, relative energies in kJmolꢀ1).
ents are denoted as 6B1 and 6B2, and 2-CH3 resonances of the distin-
guishable rotamers as 6R1 and 6R2 (tBu signals of individual rotamers
are not yet visible as the temperature is still above coalescence). Off-di-
agonal peaks indicate chemical exchange due to P inversion (6B1/6B2)
ꢀ
causes notable energetic differences of the two minimum
energy rotamers with intersecting conformations and prefer
the orientation of the tert-butyl groups at the free C2 side,
whereas in the case of P-substitution the differences be-
tween the energies of the rotamers are smaller.
or interconversion of rotamers through P P bond rotation (6R1/6R2).
22 kJmolꢀ1 and DS =2 JKꢀ1 molꢀ1. The dynamic exchange
¼
between the two distinguishable rotamers formed after the
Introduction of additional methyl groups in position 2 and
5, as present in 6, resulted in further destabilization of the
N-tBu2P rotamers, and the P-tBu2P-substituted systems
become the more stable isomers. The computed inversion
barrier about the ring phosphorus is 55 kJmolꢀ1, whereas it
is 85.2 kJmolꢀ1 about the exocyclic phosphorus. Although
the computed inversion about the ring phosphorus is higher
than the experimental value, it is still significantly lower
than for phosphole (75.1 kJmolꢀ1).[22] For these rotamers
31P NMR spectroscopic shifts and coupling constants have
also been calculated, and the values obtained are given in
Figure 8. Relative stabilities and coupling constants are in
good agreement with the experimental results. 31P chemical
shifts cannot be obtained with good accuracy by these calcu-
lations, but the trends are correct. Finally, it should be men-
tioned that, quite unexpectedly, the presence of the 2-
methyl group is crucial to determine the regioselectivity of
the phosphanylation with the bulky ligands.
ꢀ
P P bond rotation is frozen is likewise evident from the ap-
pearance of positive exchange peaks between the signals at
d=2.88/2.72 ppm in the 2D-EXSY spectrum (see Figure 5);
determination of activation parameters from line-shape
analysis was in this case prevented as the lack of sufficient
data in the slow-exchange regime prevented the determina-
tion of reliable chemical shifts for the individual rotamers.
To evaluate the effects influencing the relative stability of
N versus P isomers, quantum chemical studies were carried
out. Initially, N- and P-substituted 1,3-azaphosphole deriva-
ꢀ
tives substituted with PH2 groups were investigated
(Figure 6). It was found that the N-substituted derivatives
The reduction of steric bulk at the phosphanyl group re-
sults in destabilization of the P- in favor of the N-substituted
rotamers. This is seen by replacement of the tBu2P with the
iPr2P or the sterically comparable dicyclohexylphosphanyl
Figure 6. Possible phosphanylazaphosphole isomers (relative energies in
kJmolꢀ1, at the B3LYP/6-311+G** level).
12268
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12263 – 12272

Ambident PCN Heterocycles
FULL PAPER
phosphorus cannot be involved in cyclic delocalization. The
reason for the preference for P-phosphanylation in the case
of bulky phosphanyl groups is that the steric encumbrance is
alleviated by means of the relatively long P–P distance com-
pared to the relatively short P–N distance in N-substituted
compounds in the most stable di-tert-butylphosphanyl-2,5-di-
methylbenzazaphosphole isomers (2.248 versus 1.797 ꢃ).
The introduction of a substituent at the 2-position plays a
crucial role in the destabilization of the bulky N-phosphany-
lated benzazaphospholes and is responsible for the reversal
of the stability order because the most suitable conforma-
tion with the phosphanyl substituents directed to the 2-CH
side is destabilized.
Thus, the chemoselectivity of the phosphanylation of ben-
zazaphospholides is governed by the interplay of aromatic
stabilization and steric destabilization of the N-phosphany-
lated benzazaphospholes and the resulting delicate balance
of energetically similar P- and N-substituted isomers. Small
changes in the substitution pattern can affect the preferred
chemoselectivity. Introduction of substituents with switcha-
ble space demand, either in the 2-position or the phosphanyl
group, might allow switching between N- and P-substituted
benzazaphospholes and an associated response in terms of
their different electronic and spectral properties.
Figure 8. Relative energies of conformational isomers of 6 and its N-sub-
stituted counterparts (B3LYP/6-311+G**level, NMR spectroscopic data
calculated at the B3LYP/6-31+G* level).
substituent. For these substituents, the N-substituted rota-
A
Conclusion
phanyl rotamers (Figure 9), although the energy difference
is very small.
1,3-Benzazaphospholides, benzoanellated diheterocyclopen-
tadienides, react with electrophiles at either of the two het-
eroatoms, nitrogen or phosphorus. In a first approximation
this may be explained by the HSAB principle. The detailed
investigation of the reaction with chlorophosphanes shows,
however, a more sophisticated behavior. The substitution at
nitrogen or phosphorus is controlled in a subtle way by elec-
tronic and steric factors. In the absence of steric hindrance,
the phosphanylation takes place at nitrogen, as the planar
N-substituted benzazaphospholes profit energetically from
the formation of a heteroaromatic system, whereas P-substi-
tuted benzazaphospholes with pyramidal phosphorus are in-
capable of full electronic cyclodelocalization. Substituents at
the 2-position of benzazaphospholides, even the small
methyl group, disfavor N-substitution, however, particularly
by bulky groups. The shorter bond of nitrogen than of phos-
phorus to the substituent causes much stronger hindrance
for N- than P-substitution and energetic destabilization. This
hindrance changes also the preferred conformation of the
N-substituent. In 2-unsubstituted benzazaphospholes the
substituents of the phosphanyl group are directed towards
the five-membered ring end of the benzazaphosphole
system, in the case of a 2-substituent towards the opposite
site, the benzo-end. The knowledge of the factors governing
N- or P-substitution of benzazaphospholides presented in
this contribution may help pre-evaluate the role of steric
factors governing the chemoselectivity in substitution reac-
tions of ambident anions.
Figure 9. Relative energies of conformational isomers of diisopropylphos-
phanyldimethylbenzazaphosphole and 5 and its P-substituted counter-
parts (relative energies at the B3LYP/6-311+G**level, in kJmolꢀ1).
The preference of N-substitution by less-bulky phosphanyl
groups is attributable to the aromatic stabilization of these
ꢀ
ꢀ
isomers, and the difference between the P N and P P bond
strengths. In the P-substituted benzazaphospholes, the phos-
phorus atom is pyramidalized, and thus the lone pair of
Chem. Eur. J. 2009, 15, 12263 – 12272
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12269

D. Gudat, L. Nyulaszi, J. Heinicke et al.
2J=22 Hz, 4J=1 Hz; CH-4), 129.1 (s; m-CH), 129.7 (d, 3J=11.2 Hz; Cq-
Experimental Section
4
5), 130.1 (d, J=2.0 Hz; m-CH), 135.3 (d, J=21.0 Hz; o-Cq or i-Cq), 137.0
(d, 5J=2.0 Hz; p-C), 138.0 (d, J=3.6 Hz; o-Cq or i-Cq), 139.6 (d, J=
3.6 Hz; o-Cq or i-Cq), 144.1 ppm (dd, 2J=4.5, 7.7 Hz; Cq-7A); Cq-2 and
General remarks: All reactions were carried out in heat-dried glass ves-
sels under dry nitrogen or argon using Schlenk techniques and freshly
distilled, ketyl-dried solvents. NMR spectroscopy solvents were dried
(C6D6 and [D8]THF over LiAlH4, CDCl3 over P4O10) and recondensed
before use. Me3SiCl was also recondensed under vacuum. Dicyclohex-
Cq-3A were not detected; 31P{1H} NMR (C6D6): d=90.5 (d, 3J
ACHTUNGTRENNUNG(P,P)=
3.1 Hz; 3-P); 86.0 ppm (brs; N2P); MS (EI, 70 eV, 490 K): m/z (%): 512.2
(0.1) [Mꢀ1]+, 351.2 (100.0) [MꢀC9H9NP]+, 163.1 (9.4) [MꢀC22H28N2P]+,
119.1 (2.3) [MꢀC22H26N3P2]+; elemental analysis calcd (%) for
C31H37N3P2·C6H6: C 75.10, H 7.32, N 7.10; found: C 75.05, H 7.56, N 6.96.
ACHTUNGTRENNUNG
yl-,[23] di-tert-butyl-[24] and di-1-adamantylchlorophosphane[25] were pre-
pared by known procedures. Other chemicals were used as purchased.
NMR spectra were recorded using Bruker ARX300 or Avance 300 NMR
spectrometers at 300.1 (1H), 75.5 (13C), and 121.5 MHz (31P) (Greifswald),
or using an Avance400 spectrometer at 400.13 (1H) and 121.9 MHz (31P)
(Stuttgart). Shift references are tetramethylsilane for 1H and 13C and
Compound 4: tBuLi (0.58 mL 1.7m in pentane, 0.99 mmol) was added
dropwise to a solution of 1 (145 mg, 0.89 mmol) in THF (2 mL) at
ꢀ788C. This mixture was stirred for 5 h while slowly warming up to
ꢀ408C, then again cooled to ꢀ788C, and Ph2PCl (0.16 mL, 0.889 mmol)
was added dropwise. After stirring at room temperature overnight, THF
was removed under vacuum and the residue was extracted with diethyl
ether. Evaporation of the solvent gave an orange yellow viscous oil. The
85% H3PO4 (X=40.480747 MHz) for 31P. Coupling constants refer to J-
1
(H,H) in H and J
(P,C) in 13C NMR spectroscopic data unless stated oth-
erwise, and are given as absolute values. Assignments are supported by
additional distortionless enhancement by polarization transfer (DEPT)
measurements and further experiments described in the text. Assignment
numbers are given in Scheme 1. The assignment Cq represents a quater-
nary carbon center. Mass spectra were recorded using a single-focusing
mass spectrometer AMD40 (Intectra). HRMS measurements were car-
ried out in Gçttingen using a double-focusing sector-field instrument
MAT 95 (Finnigan) with electron impact ionization (EI; 70 eV), or a 7 T
Fourier transform ion cyclotron resonance mass spectrometer APEX IV
(Bruker Daltonics) with electrospray ionization (ESI; PFK as reference
substances). Melting points were determined using a Sanyo Gallenkamp
melting-point apparatus in capillaries (uncorrected). Elemental analyses
were performed using a Perkin–Elmer 24000CHN/O analyzer.
31P NMR (C6D6) spectrum displayed a mixture of products that included
3
doublets indicating 4 by typical chemical shift and J
N
(d=40.5 (d, 3J=6.9 Hz; Ph2P), 92.9 ppm (d, 3J=7.0 Hz; 3-P)), content
42 mol% of benzazaphospholes by integration of methyl protons. The
methyl groups of 4 appear in the 1H NMR (CDCl3) spectrum at d=2.12
(s; 5-Me) and 2.77 ppm (dd, J
are superimposed. Further strong signals indicate 1 (d
d(1H)=2.21 (d, 3J
(P,H)=12.1 Hz; 2-Me), 2.29 ppm (s; 5-Me), content
42 mol% of benzazaphospholes) and at d
(31P)=ꢀ14.8 ppm (Ph2P-PPh2).
Weaker singlets are observed at d
(31P)=116.9, 86.7 (w, m; both un-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
known), 17.9 ppm (w; tBuPPh2), and ꢀ5.55 and ꢀ5.74 ppm, each with
half intensity of the Ph2P-PPh2 signal. tBu proton signals of tBuPPh2 are
Compound 1: This compound, 31P
ACTHNUTRGNEN(UG CDCl3): d=71.8 ppm, was prepared as
3
observed at d(1H)=1.15 ppm (d, J
G
reported earlier.[15a] Single crystals
were grown from 1 in a THF/hexane
mixture. Crystal data are compiled in
Table 2.
Table 2. Crystallographic data.
1
3
5
ACHUTGTNNERNUG[7·LiClACHTUNTGERN(NUNG thf)]2
Compound 3: Compound 2[14] (1.16 g,
3.00 mmol) was dissolved in THF
(20 mL) and cooled to ꢀ788C. Then a
formula
Mr
T [K]
l [ꢃ]
C9H10NP
163.15
C37H37D6N3P2
597.70
C21H31NP2
359.41
C66H94Cl2Li2N2O2P4
1156.09
133(2)
173(2)
100(2)
103(2)
solution of 1Li, prepared from
1
0.71073
0.71073
1.54184
1.54184
(0.55 g, 3.37 mmol) and the equimolar
amount of tBuLi solution in THF
(10 mL) at ꢀ788C, was slowly added.
The mixture was allowed to warm to
room temperature and stirred for 1 h.
The solvent was removed under
vacuum and the residue was extracted
with toluene (30 mL). After filtration
through Celite, the solution was con-
centrated under vacuum to about
5 mL and the precipitate was dissolved
by addition of THF (3 mL). Yellow
crystals, formed at +48C, were collect-
ed by filtration and dried under
vacuum to give 3 (1.01 g, 64%). M.p.
1478C; 1H NMR (C6D6): d=1.57 (s,
6H; 2Me), 1.66 (s, 6H; 2Me), 2.00 (s,
crystal system
space group
a [ꢃ]
b [ꢃ]
c [ꢃ]
monoclinic
Pc
8.3117(8)
5.7552(4)
8.8751(8)
90
98.927(4)
90
419.40(6)
2
1.292
0.257
172
0.4ꢄ0.3ꢄ0.3
2.48 to 30.56
ꢀ11%h%11
ꢀ8%k%8
ꢀ12%l%12
6139
monoclinic
P21/c
17.673(3)
9.0335(13)
21.343(4)
90
90.863(14)
90
3407.0(10)
4
1.165
0.156
1264
0.7ꢄ0.45ꢄ0.2
2.21 to 27.00
0%h%22
0%k%11
ꢀ27%l%27
7687
7447
0.0451
27.00 (99.9%)
–
7447/0/379
0.866
R1=0.0624
wR2=0.1611
R1=0.1020
wR2=0.1740
0.961/ꢀ0.491
monoclinic
P21/n
11.6572(3)
10.6438(3)
16.3790(4)
90
94.812(2)
90
2025.09(9)
4
1.179
1.944
776
0.35ꢄ0.15ꢄ0.04
4.48 to 71.14
ꢀ13%h%14
ꢀ13%k%13
ꢀ20%l%19
57680
triclinic
P1
¯
11.1411(6)
11.6873(12)
12.9262(8)
77.481(6)
83.739(6)
66.133(6)
1502.2(2)
1
1.278
2.327
620
0.10ꢄ0.08ꢄ0.03
3.50 to 71.09
ꢀ9%h%13
ꢀ14%k%13
ꢀ15%l%15
28262
a [8]
b [8]
g [8]
V [ꢃ3]
Z
1calcd [Mgmꢀ3
]
m [mmꢀ1
(000)
]
F
A
crystal size [mm3]
q range [8]
index ranges
6H; 2Me), 2.10 (s, 6H; 2Me), 2.31 (s,
3
3H; 5-Me), 2.68 (dd, J
G
collected reflns
independent reflns
4J
ACHTUNGTRENNUNG(P,H)=7.0 Hz, 3H; 2-Me), 6.49 (s,
2495
0.0154
3849
0.0341
5559
0.0291
2H; m-H), 6.73 (s, 2H; m-H), 7.21 (d,
1J=8.6 Hz, 1H; H-6), 7.83 (d, 3J-
ACHTUNGTRENNUNG(P,H)=4.3 Hz, 1H; H-4), 7.94 ppm (d,
R
N
completeness to q [8]
max/min transmission
data/restraints/params
GOF on F2
final R indices
AHCTUNGTRENNUNG
30.00 (99.5%)
0.9270/0.8145
2495/2/107
1.106
R1=0.0338
wR2=0.0836
R1=0.0343
wR2=0.0842
0.597/ꢀ0.184
67.50 (99.9%)
1.00000/0.68127
3849/0/219
1.072
R1=0.0447
wR2=0.1114
R1=0.0478
wR2=0.1140
1.027/ꢀ0.582
67.50 (98.4%)
1.00000/0.75697
5559/0/354
1.033
R1=0.0348
wR2=0.0954
R1=0.0415
wR2=0.1000
0.545/ꢀ0.307
1J=8.6 Hz, 1H; H-7); 13C{1H} NMR
(C6D6): d=10.5 (d, 3J=2.0 Hz; NC-
CH3), 18.9 (s; p-CH3), 19.3 (d, 4J=
3.2 Hz; o-CH3), 20.5 (d, 4J=0.5 Hz; o-
CH3), 20.3 (dd, 2J=28.1 Hz, 3J=
40.1 Hz; 2-Me), 21.0 (5-Me), 118.8 (s;
CH-7), 120.5 (d, 2J=6.3 Hz; N-C=),
125.7 (d, 4J=2.5 Hz; CH-6), 128.7 (dd,
AHCTUNGTRENNUNG
]
12270
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12263 – 12272

Ambident PCN Heterocycles
FULL PAPER
Compound 5: Chlorodicyclohexylphosphane (0.64 mL, 2.897 mmol) was
added dropwise at ꢀ788C to a solution of 1Li, prepared from 1 (450 mg,
2.76 mmol) in THF (5 mL) and tBuLi (2.02 mL, 3.03 mmol, 1.5m in pen-
tane) as described above. After stirring overnight at room temperature
THF was removed under vacuum. The residue was extracted with hexane
yielding a yellow crystalline slush. This mixture was purified by crystalli-
zation from a mixture of THF/hexane to give colorless solid 5 (465 mg,
47%) . NMR spectra at 258C in C6D6 indicate two sets of signals, ratio
by 1H integration of methyl signals approximately 90–85:10–15 mol%.
1H NMR (C6D6): major rotamer: d=0.8–1.72 (5 m, 20H; PCy2), 2.25 (s,
doublets of the product and contamination by small amounts of uncon-
verted 1Li (d
traces of two unknown phosphorus compounds (dCATHNUGTRNENUG
(31P)=56.7 ppm) and Ada2PCl (d(31P)=139.8 ppm), and
ACHUTGTNRENNUG CAHTUNGTRENNUGN
(31P)=18.2, 83.7 ppm).
An attempt at high-vacuum distillation (separate experiment) led to ther-
mal decomposition. Overlayering of a concentrated solution of crude 7 in
THF by hexane gave crystals of [7·LiClACHTNUGTRENUNG(THF)]2, thus allowing a structure
analysis. The crystal data are compiled in Table 2. 1H NMR (C6D6): d=
1.50, 1.58, 1.72–1.85, 1.90–2.10 (m, 30H; ada), 2.22 (s, 3H; 5-Me), 2.84 (d,
ACTHNUTRGNEUNG
3J(P,H)=9.3 Hz, 3H; 2-Me), 7.07 (dd, 3J=7.8 Hz, 4J=1.1 Hz, 1H; H-6),
7.87 (brs, 1H; H-4), 8.06 ppm (d, 3J=8.0 Hz, 1H; H-7); 13C{1H} (DEPT)
3H; 5-Me), 2.69 (m, 2H; PCHa), 2.95 (dd, 3J
3.4 Hz, 3H; 2-Me), 7.09 (dd, 3J=8.6 Hz, 4J=1.6 Hz, 1H; H-6), 7.68 (d,
3J=8.6 Hz, 1H; H-7), 7.80 ppm (dd, 3J(P,H)=4.6 Hz, 4J=1.6 Hz, 1H; H-
4); minor rotamer: d=2.27 (s; 5-Me), 2.62 (d, 3J
(P,H)=14.9 Hz; 2-Me),
7.17 (dd, Jꢂ8 Hz, Jꢂ2 Hz; H-6), 7.80 (superimposed d; H-4), 8.49 ppm
(dd, J=8.7 Hz, J
(P,H)=4.5 Hz; H-7); cyclohexyl superimposed; 13C{1H}
A
R
NMR (C6D6): d=21.37 (s; 5-Me), 21.84 (brd, 2J=25.8 Hz; 2-Me), 25.80
3
(s; CH
G
1J=30.9 Hz; Cq-1), 42.88 (dd, 2J=10.2 Hz, 3J=6.0 Hz; CH2), 67.80 (s;
AHCTUNGTRENNUNG
CH2A
CTHUNGTRENNUNG
AHCTUNGTRENNUNG
3
3
4
CH-6), 133.78 (d, J=5.3 Hz; Cq-5), 136.2 (br; Cq-3A), 156. 7 (br; Cq-7A),
186.3 ppm (vbr; Cq-2); 31P{1H} NMR (C6D6): d=ꢀ12.3 (brd, 1J
ACHTUNGTRENNUNG(P,P)=
3
4
ACHTUNGTRENNUNG
418 Hz; 3-P), 39.4 ppm (brd, 1J
(P,P)ꢂ410 Hz); HRMS of uncoordinated
2
(DEPT) NMR (C6D6): major rotamer: d=20.95 (s; 5-Me), 21.67 (dd, J=
27.4 Hz, 3J=31.4 Hz; 2-Me), 26.23 (s; CH2-d), 26.32 (d, 3J=6.8 Hz; CH2-
g), 26.47 (CH2-g), 28.89 (d, 2J=9.4 Hz; CH2-b), 30.99 (d, 2J=29.5 Hz;
CH2-b), 36.18 (d, 1J=18.7 Hz; CH-a), 115.03 (s; CH-7), 125.70 (d, 4J=
7 (ESI in MeCN): m/z calcd for [C29H39NP2+H]+: 464.26305; found:
464.26285; elemental analysis calcd (%) for C66H94Cl2Li2N2O2P4
(1156.15): C 68.56, H 8.20, N 2.42; found: C 68.35, H 8.43, N 2.26.
2
3
2.4 Hz; CH-6), 129.48 (d, J=21.8 Hz; CH-4), 130.12 (d, J=10.8 Hz; Cq-
Crystal-structure analyses: Selected data collection parameters and other
crystallographic data are summarized in Table 2.
5), 142.95 (d, 1J=36.2 Hz; Cq-3A), 145.79 (dd, 2J=9.3 Hz, J=5.1 Hz; Cq-
1
2
7A), 184.59 ppm (dd, J=47.6 Hz, J=19.3 Hz; Cq-2); minor rotamer: d=
19.90 (d, 2J=29.4 Hz; 2-Me), (5-Me, CH2-d, CH2-g superimposed), 28.65
(d, 2J=8.0 Hz; CH2-b), 31.38 (d, 2J=29.5 Hz; CH2-b), 38.04 (d, 1J=
18.7 Hz; CH-a), 116.92 (d, 3J=33.3 Hz; CH-7), 125.99 (t, 4J=2.6 Hz;
CH-6), 128.04 (d, 2J=21.1 Hz; CH-4), 129.78 (d, 3J=12.9 Hz; Cq-5),
142.38 (d, 1J=34.6, 3J=3.9 Hz; Cq-3A), 150.72 (dd, 2J=24.1, 5.1 Hz; Cq-
7A), 180.75 ppm (dd, 1J=48.7 Hz, 2J=11.8 Hz; Cq-2); 31P{1H} NMR
Data collection and reduction: All measurements were carried out at low
temperature using crystals mounted in inert oil on glass fibers. Com-
pound 1 was measured using a Bruker SMART 1000 CCD; compound 3
using a Siemens P4 diffractometer (both with MoKa radiation); com-
pound 5 and [7·LiClACTHNUTRGNENUG(thf)]2 using an Oxford Diffraction Nova diffractome-
ter with CuKa radiation. Data for 3 were not corrected for absorption ef-
fects; for the other structures, absorption corrections (semiempirical from
equivalents) were applied on the basis of multiscans.
(C6D6): major rotamer: d=61.5 (d, 3J
A
1J
1J
ACHTUNGTRENNUNG
Structure solution and refinement: Structures were solved by direct meth-
ods and refined on F2O by full-matrix least-squares refinement using the
SHELX suite of programs.[27] Hydrogen atoms were included using a
riding model or with rigid methyl groups.
ACHTUNGTRENNUNG(P,P)=4.2 Hz; P-3); trace 73.2 ppm (1); HRMS (ESI in MeOH/NH4):
m/z calcd for C21H31NP2 [M+H]+: 360.20045; found: 360.20042 (low in-
tensity by decomposition).
Compound 6: Di(tert-butyl)chlorophosphane (0.25 mL, 1.46 mmol) was
added dropwise at ꢀ788C to a solution of 1Li, prepared from 1 (200 mg,
1.23 mmol) in THF (2 mL) and tBuLi (0.9 mL, 1.35 mmol, 1.5m in pen-
tane) as described above. The mixture was stirred at room temperature
overnight. THF was removed from the reaction mixture under vacuum,
the residue was extracted with hexane, and hexane was evaporated under
vacuum to give a yellow viscous oil. NMR spectroscopic control (C6D6)
showed that the oil consisted mainly of 6, contaminated by small amounts
of impurities. One crystallized and proved crystallographically identical
to tBu2PHO·LiCl.[26] The oil was separated and purified by high-vacuum
distillation at 608C bath temperature, b.p. around 308C/1.1ꢄ10ꢀ5 mbar,
to give a colorless oil (233 mg, 62%). 1H NMR (C6D6): d=1.12 (brd, 3J-
Exceptions and special features: For 1, the NH hydrogen was refined
freely; the structure was refined as a racemic twin with components 0.56,
0.44(7). For 3, the carbon atoms of the cocrystallized deuterobenzene
molecule were constrained to a regular hexagon.
CCDC-734965 (1), 726723 (3), 734966 (5), and 734967 ([7·LiClACHTUNGTRENNUNG(thf)]2)
contain the supplementary crystallographic data for this paper. These
data can be obtained free of charge from The Cambridge Crystallograph-
ic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Quantum chemical calculations: All calculations were carried out with
the Gaussian 03 program package.[28] Full geometry optimizations were
performed for all examined molecules at the B3LYP/6-311+G** level,[29]
and vibrational frequencies were calculated to establish the nature of sta-
tionary points obtained. 31P NMR spectroscopic shifts (using tetramethyl-
silane and gas-phase PH3 as reference[30]) and P,P coupling constants
were calculated at the B3LYP/6-31+G*//B3LYP/6-311+G** level.
ACHTUNGTRENNUNG AHCTUNGTRNEN(UGN P,H)=
(P,H)=12.4 Hz, 18H; PCMe3), 2.18 (s, 3H; 5-Me), 2.76 (d, 3J
9.2 Hz, 3H; 2-Me), 7.03 (dd, 3J=8.1 Hz, 4J=1.5 Hz, 1H; H-6), 7.72 (brs,
1H; H-4), 8.02 ppm (d, 3J=8.1 Hz, 1H; H-7); 13C{1H} (DEPT) NMR
(C6D6): d=21.28 (s; 5-Me), 21.80 (dd, 2J=25.1 Hz, 3J=3.6 Hz; 2-Me),
31.38 (dd, 2J=13.3 Hz, 3J=5.2 Hz; CMe3), 36.64 (vbrd, J=26.3 Hz;
CqMe3), 124.71 (s; CH-7), 129.23 (d, 2J=18.6 Hz; CH-4), 129.81 (s; CH-
6), 134.16 (d, 3J=6.6 Hz; Cq-5), 135.39 (brd, 1J=9.7 Hz; Cq-3A), 156.59
(brd, 2J=14.8 Hz; Cq-7A), 184.89 ppm (dd, 1Jꢂ25 Hz, 2J=9.7 Hz; Cq-2);
31P{1H} NMR (C6D6): d=ꢀ6.0 (d, 1J
(P,P)=425 Hz; 3-P), 46.4 (brd, 1J-
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG(P,P)=430 Hz); small hydrolysis impurities at 61.9 (tBu2PHO) and
Acknowledgements
73.5 ppm (1); MS (EI, 70 eV, 508C): m/z: calcd (%) for C17H27NP2
(307.16): 308 (2), 307 (10) [M+], 250 (6), 194 (34), 57 (100); HRMS after
short air contact (ESI in MeOH): m/z calcd for [C17H27NP2+O+H+]:
324.16406; found: 324.16406 [M++OH]; (ESI in MeOH, CH3COONa):
m/z calcd for [C17H27NP2+O+Na+]: 346.14601; found: 346.14631.
Financial support and a fellowship (B.R.A.) from the Deutsche For-
schungsgemeinschaft is gratefully acknowledged. We thank B. Witt, W.
Heiden, and M. Steinich for NMR and mass spectra, and Dr. H. Frauen-
dorf and G. Sommer-Udvarnoki (Georg-August-Universitꢁt Gçttingen,
Institut fꢀr Organische und Biomolekulare Chemie) for HRMS measure-
ments. D.G. and L.N. acknowledge the financial support from DFG-HAS
and COST CM0802.
Compound [7]: Di(1-adamantyl)chlorophosphane (2.3 g, 6.83 mmol) was
added dropwise at ꢀ788C to a solution of 1Li, prepared from 1 (1.11 g,
6.80 mmol) in THF (10 mL) and tBuLi (4.7 mL, 7.52 mmol, 1.6m in pen-
tane) as described above, and stirred at room temperature. The solution
turned slowly from yellow to orange. Reaction control showed that after
three weeks at room temperature the major part of the starting materials
had reacted. THF was removed under vacuum to give an orange semi-
crystalline crude product (2.7 g). The 31P NMR spectra displayed broad
[1] J. Heinicke, Tetrahedron Lett. 1986, 27, 5699–5702.
[2] a) A. Schmidpeter in Comprehensive Heterocyclic Chemistry II
(Eds.: R. Katritzky, C. W. Rees, E. F. V. Scriven), Pergamon,
Oxford, 1996, Chapter 4.22; b) A. Schmidpeter in Phosphorus-
Chem. Eur. J. 2009, 15, 12263 – 12272
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12271

D. Gudat, L. Nyulaszi, J. Heinicke et al.
Carbon Heterocyclic Chemistry: The Rise of a New Domain (Ed.: F.
Mathey), Pergamon, Amsterdam, 2001, pp. 363–461; c) R. K.
Bansal, J. Heinicke, Chem. Rev. 2001, 101, 3549–3578.
[18] a) B. Gehrhus, P. B. Hitchcock, M. F. Lappert, J. Chem. Soc. Dalton
Trans. 2000, 3094–3099; b) W. M. Boesfeld, B. Gehrhus, P. B. Hitch-
cock, M. F. Lappert, P. von R. Schleyer, Chem. Commun. 1999, 755–
756; c) F. E. Hahn, L. Wittenbecher, M. Kꢀhn, T. Lꢀgger, R. Frçh-
lich, J. Organomet. Chem. 2001, 617–618, 629–634.
[3] a) L. Szarvas, Z. Bajko, S. Fusz, S. Burck, J. Daniels, M. Nieger, D.
Gudat, Z. Anorg. Allg. Chem. 2002, 628, 2303–2310; b) G. Mꢁrkl, I.
Trçtsch, Angew. Chem. 1984, 96, 899–900; c) J. G. Kraaijkamp,
D. M. Grove, G. van Koten, A. Schmidpeter, Inorg. Chem. 1988, 27,
2612–2617.
[4] a) “Metallocene complexes as catalysts for olefin polymerization”:
Coord. Chem. Rev. 2006, 250, 1–272, edited by H. G. Alt; b) “Metal-
locenes in regio- and stereoselective synthesis”: Top. Organomet.
Chem. 2005, 8, 1–244, edited by T. Takahashi.
[5] T. Veszprꢅmi, J. Mꢆtrai, J. Heinicke, M. K. Kindermann, J. Mol.
Struct. 2001, 559, 189–195.
[6] a) A. Surana, S. Singh, R. K. Bansal, N. Peulecke, A. Spannenberg,
J. Heinicke, J. Organomet. Chem. 2002, 646, 113–124; b) J. Heinicke,
A. Surana, N. Peulecke, R. K. Bansal, A. Murso, D. Stalke, Eur. J.
Inorg. Chem. 2001, 2563–2567.
[19] H. C. E. McFarlane, W. McFarlane, A. J. Nash, J. Chem. Soc. Dalton
Trans. 1980, 240–244.
[20] J. G. Verkade, J. A. Mosbo in Phosphorus-31 NMR Spectroscopy in
Stereochemical Analysis, Methods in Stereochemical Analysis, Vol. 8
(Eds.: J. G. Verkade, L. D. Quin), VCH, Weinheim, 1987, p. 425.
[21] L. D. Quin in Phosphorus-31 NMR Spectroscopy in Stereochemical
Analysis, Methods in Stereochemical Analysis, Vol. 8 (Eds.: J. G. Ver-
kade, L. D. Quin), VCH, Weinheim 1987, p. 391.
[22] A. Dransfeld, L. Nyulaszi, P. v. R. Schleyer, Inorg. Chem. 1998, 37,
4413–4420.
[23] K. Issleib, W. Seidel, Chem. Ber. 1959, 92, 2681–3008.
[24] M. Fild, O. Stelzer, R. Schmutzler in Inorganic Synthesis, Vol. 14
(Eds.: A. Wold, J. Ruff), McGraw-Hill, New York, 1973, pp. 6–7.
[25] J. R. Goerlich, R. Schmutzler, Phosphorus Sulfur Silicon Relat.
Elem. 1995, 102, 211–215.
[26] B. R. Aluri, M. K. Kindermann, P. G. Jones, I. Dix, J. W. Heinicke,
Inorg. Chem. 2008, 47, 6900–6912.
[27] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[28] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
[7] K. Issleib, R. Vollmer, Z. Anorg. Allg. Chem. 1981, 481, 22–32.
[8] J. Heinicke, A. Petrasch, A. Tzschach, J. Organomet. Chem. 1983,
258, 257–269.
[9] A. Schmidpeter, H. Tautz, Z. Naturforsch. B 1980, 35, 1224–1228.
[10] a) L. Nyulꢆszi, G. Csonka, J. Rꢅffy, T. Veszprꢅmi, J. Heinicke, J. Or-
ganomet. Chem. 1989, 373, 49–56; b) L. Nyulꢆszi, G. Csonka, J.
Rꢅffy, T. Veszprꢅmi, J. Heinicke, J. Organomet. Chem. 1989, 373,
57–61; c) T. Veszprꢅmi, L. Nyulꢆszi, J. Reffy, J. Heinicke, J. Phys.
Chem. 1992, 96, 623–626; d) L. Nyulꢆszi, T. Veszprꢅmi, J. Rꢅffy, B.
Burkhardt, M. Regitz, J. Am. Chem. Soc. 1992, 114, 9080–9084;
e) L. Nyulaszi, Chem. Rev. 2001, 101, 1229–1246.
[11] S. Burck, D. Gudat, M. Nieger, Angew. Chem. 2004, 116, 4905–
4908; Angew. Chem. Int. Ed. 2004, 43, 4801–4804.
[12] S. Burck, D. Gudat, M. Nieger, D. Vindus, Eur. J. Inorg. Chem.
2008, 704–707.
AHCTUNGTREGeNNUN ry, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[13] S. Burck, D. Gudat, M. Nieger, Angew. Chem. 2007, 119, 2977–
2980; Angew. Chem. Int. Ed. 2007, 46, 2919–2922.
[14] a) S. Burck, D. Gudat, M. Nieger, Z. Benkç, L. Nyulaszi, D. Szie-
berth, Z. Anorg. Allg. Chem. 2009, 635, 245–252; b) S. Burck, D.
Gudat, K. Naettinen, M. Nieger, M. Niemeyer, D. Schmid, Eur. J.
Inorg. Chem. 2007, 5112–5119; c) D. Gudat, A. Haghverdi, H.
Hupfer, M. Nieger, Chem. Eur. J. 2000, 6, 3414–3425.
[15] a) J. Heinicke, N. Gupta, A. Surana, N. Peulecke, B. Witt, K. Stein-
hauser, R. K. Bansal, P. G. Jones, Tetrahedron 2001, 57, 9963–9972;
b) R. K. Bansal, N. Gupta, J. Heinicke, G. N. Nikonov, F. Saguitova,
D. C. Sharma, Synthesis 1999, 264–269.
[29] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648; b) C. Lee, W. Yang,
R. G. Parr, Phys. Rev. B 1988, 37, 785.
[30] C. van Wꢀllen, Phys. Chem. Chem. Phys. 2000, 2, 2137–2144.
[16] G. Becker, W. Massa, O. Mundt, R. E. Schmidt, C. Witthauer, Z.
Anorg. Allg. Chem. 1986, 540/541, 336–344.
[17] S. Burck, D. Gudat, F. Lissner, K. Naettinen, M. Nieger, T. Schleid,
Z. Anorg. Allg. Chem. 2005, 631, 2738–2745.
Received: June 24, 2009
Published online: October 9, 2009
12272
www.chemeurj.org
ꢂ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 12263 – 12272