Steric promotion of aromatic C-H bond activation in primary benzylamines

Article abstract of DOI:10.1002/(sici)1099-0682(199906)1999:6<1029::aid-ejic1029>3.0.co;2-7

ortho-Palladation of a sterically crowded primary benzylamine, α- phenylneopentylamine, was accomplished in a moderate yield of 50% in the reaction with the weakest of palladation agents (Li₂PdCl₄) under very mild conditions, due to

Full text of DOI:10.1002/(sici)1099-0682(199906)1999:6<1029::aid-ejic1029>3.0.co;2-7

FULL PAPER
Steric Promotion of Aromatic C–H Bond Activation in Primary Benzylamines
Valery V. Dunina,*[a] Lyudmila G. Kuz’mina,[b] Marina Yu. Kazakova,[a] Ol’ga N. Gorunova,[a]
Yury K. Grishin,[a] and Elena I. Kazakova[c]
Keywords: Palladium / Primary benzylamines / NMR spectroscopy / Structure elucidation / Conformation analysis
ortho-Palladation of a sterically crowded primary benzyl-
amine, α-phenylneopentylamine, was accomplished in a
moderate yield of 50% in the reaction with the weakest of
palladation agents (Li2PdCl4) under very mild conditions,
due to a steric promotion of an aromatic C–H bond activation.
The structure of dimer 1a thus formed and the palladacycle
conformation were established on the basis of 1H-NMR
spectroscopy of its mononuclear derivatives with [D5]pyri-
dine (3a) and triphenylphosphane (4a), and an X-ray investi-
gation of the latter.
Despite numerous publications dealing with the direct
cyclopalladation of primary benzylamines, the possibility of
steric promotion of this process still remains unexplored.
The pioneer work of Lewis and co-workers[14] cannot be
considered as conclusive enough because this publication
does not contain any spectral evidence that the isolated (tri-
phenylmethylamine)palladium complexes have a cyclopalla-
dated structure.
Our investigation of the direct ortho-palladation of sec-
ondary arylalkylamines[15Ϫ19] has revealed a phenomenon
of a remarkable steric facilitation of this process. The reac-
tion of Li2PdCl4 with a series of N-methylbenzylamines was
found to be dependent on the volume of substituent R1
(R1 ϭ H, Me, tBu) in the α-benzylic position.
Introduction
Widening the scope of direct intramolecular palladation
reactions is of great importance for further development of
regioselective palladium-mediated organic synthesis.[1][2] In
particular, cyclometallated derivatives of primary amines
seem to be especially promising synthons for the heterocy-
clization processes based on a combined use of the CϪPd
and NϪH bond reactivities.[3Ϫ5] However, after initial un-
successful attempts with the use of tetrachloropalladate,[6]
the direct ortho-palladation reactions of these substrates
proved to be elusive.
Several recent findings have made an important contri-
bution towards a solution to this problem: (i) the use of
palladation agents of higher electrophilicity; and (ii) the re-
alization of the necessity for creating a coordination vac-
ancy at a metal centre. The first point includes a substi-
tution of palladium acetate for commonly used tetrachloro-
palladate salts.[7Ϫ11] The second goal may be simply
achieved by keeping the palladium/ligand ratio equal to 1:1.
It results in the formation of binuclear coordination com-
plexes as reaction intermediates capable of dissociating eas-
ily to form two unsaturated tricoordinated species of high
reactivity.[8][10] At this stage, a polar solvent may facilitate
the dissociative formation of the coordination vacancy that
is required for a subsequent CϪH bond activation. In prin-
ciple, all these problems may be solved simultaneously by
utilizing the methodology involving a halide abstraction
from the intermediate bis(amine) complexes [Pd(HL)2Hal2]
by the action of soluble Agϩ salts.[12][13]
Scheme 1. Steric promotion of aromatic CϪH bond activation in
secondary benzylamines
Thus, the reaction with α-nonsubstituted N-methylben-
zylamine stops at the stage of the simple coordination com-
plex of type 2;[15] α-Me-substituted amine forms ortho-pal-
ladated dimer of type 1 only in refluxing methanol in mod-
erate yield of 47%.[15] In contrast, α-tBu-substituted ana-
logue undergoes the cyclopalladation reactions even at
room temperature in excellent yield of 86%.[18][20] More-
over, activation of sp3 CϪH bond (26%) becomes possible
in the case of the latter sterically crowded benzylamine in
competition with sp2 CϪH bond activation (60%) to give
two possible regioisomers in a ca. 1:3 ratio.[20]
[a]
Department of Chemistry, M. V. Lomonosov Moscow State
University,
119899, V-234, VrobЈyovy Gory, Moscow, Russia
Fax: (internat.) ϩ 7-095/932-8846
E-mail: dunina@org.chem.msu.su
[b]
N. S. Kurnakov Institute of General and Inorganic Chemistry,
Leninsky Prospect, 31, 117907, Moscow, Russia
[c]
Department of Chemistry and Biology, Syktyvkar State Univer-
These findings gave an impetus for studying a similar
steric effect in the CϪH bond activation of primary ben-
zylamines. The role of solvents, particularly acetone, in
these processes is also of interest. Another task is the inves-
sity,
167005, Petrozavodskaya Street, 120, Syktyvkar, Komi Repu-
blik, Russia
Supporting information for this article is available on the WWW
under http://www.wiley-vch.de/home/eurjic or from the author.
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ1948/99/0606Ϫ1029 $ 17.50ϩ.50/0
1029

V. V. Dunina et al.
FULL PAPER
tigation of conformational features of the five-membered
The results obtained in this series of ortho-palladation
1
palladacycles containing a primary amino group using H- reactions with palladium(II) acetate make it evident that
NMR and X-ray studies. These characteristics seem to be polar solvents have some advantages over the non-polar
especially important for utilization of the new kind of palla- ones. It may be a consequence of their strong solvating ef-
dacycles in the processes of chiral recognition (for example, fect that assists in the generation of the reactive three-coor-
see ref.[21]).
dinate intermediate of the [(HL)Pd(OAc)2(Solv)]-type[8] re-
quired for subsequent CϪH bond activation. Solvent ba-
sicity may also be important at the stage of CϪH bond acti-
vation.[24]
Results and Discussion
Cyclopalladation Reactions
The first successes in the palladation of primary benzyl-
amines were achieved with the use of acetone as sol-
vent.[8,11,13] Such a choice seems to be a rather strange one
when the possibility of a metal-assisted interaction between
the carbonyl and NH2 groups is considered. Precedents for
the CϭN bond formation[25] and cleavage[26Ϫ28] at the pal-
ladium matrix are known. These transformations may allow
the reaction to go through ortho-palladation of the more
reactive short-lived ketimine intermediate. This route may
result in the facilitation of the whole process. However, con-
trary to expectations, the use of acetone as solvent in the
palladation of the amine HL1 did not provide any advan-
tages over the use of methanol; moreover, in this case, more
vigorous conditions or increased reaction times are re-
quired. This observation allowed for the exclusion of the
ortho-palladation route via intermediate ketimine ligand
formation (at least for HL1).
To estimate the role of steric effects in the processes of
the intramolecular CϪH bond activation in primary ben-
zylamines, we have chosen α-phenylneopentylamine (HL1)
bearing a bulky tert-butyl group in the α-benzylic position
as the most suitable ligand. In accordance with recent infor-
mation regarding the cyclometalation mechanisms,[22Ϫ24]
the equimolar ligand/Pd ratio was kept under all conditions
to facilitate the formation of the free coordination vacancy
after the first stage of the ligand-to-palladium N-coordi-
nation.
Firstly, a study of the HL1 ortho-palladation was carried
out using palladium(II) acetate, one of the most efficient
reagents due to its high electrophilicity.
The results obtained when using the weak palladation
agent Li2PdCl4 may be considered as the most convincing
evidence for the steric promotion of the CϪH bond acti-
vation: dimer 1a was isolated in a 50% yield in the reaction
performed in MeOH in the presence of sodium acetate ex-
cess at room temperature.
Scheme 2. Influence of conditions on the ortho-palladation of the
primary amine HL1 with palladium(II) acetate
The reaction performed in benzene at room temperature
afforded (after standard AcOϪ/ClϪ metathesis and chroma-
tographic purification) the desired dimeric chloro-bridged
ortho-palladated complex 1a in a 43% yield. In comparison,
the α-methyl-substituted µ-bromo analogue was obtained
previously under similar conditions in a yield of only 17%
after one day, or in the same yield (43%) after three days.[9]
Scheme 3. ortho-Palladation of primary amine HL1 using tetrachlo-
ropalladate
Under these conditions a simple coordination complex
Both enantiomers of the corresponding µ-chloro dimer trans-[Pd(HL1)2Cl2] (2a) was isolated in a yield of ca. 43%
were prepared by Fuchita in higher yields of 57Ϫ63% by (based on the ligand) as a side product. Its 1:2 stoichi-
means of a two-step procedure (including the isolation of ometry differs from the starting Pd/HL1 ratio of 1:1 and
the intermediate µ-acetato dimer) only after heating at 50°C may be explained by a partial PdII reduction. Under the
for a long period of time.[11]
same conditions, primary α-methylbenzylamine forms only
The yield of dimeric complex 1a was increased to 68% by a coordination compound of type 2, which tends to decom-
carrying out the same reaction in acetone at higher tem- pose at increased temperatures.[15] The formation of com-
perature (50°C); however, this procedure requires ca. 50 plex 2a occurs without any diastereoselectivity: a mixture
hours of heating. The best results were achieved when the of nearly equal amounts of (R*,R*)- and (S,R)-dia-
reaction was performed in methanol; dimer 1a was isolated stereomers was isolated in the reaction of racemic HL1 with
in a yield of nearly 70% after heating for 1.5 hours at 50°C Li2PdCl4. One of these diastereomers may be separated by
or after stirring for 24 hours at room temperature. In com- crystallization.
parison, the formation of the α-Me-substituted analogue in
We believe that two main factors determine the easier
a 50Ϫ70% yield required heating at 60°C for four hours.[21] ortho-palladation of the sterically hindered ligands (includ-
1030
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039

Steric Promotion of Aromatic CϪH Bond Activation in Primary Benzylamines
FULL PAPER
ing amine HL1): (i) a large effective volume of the ligand ordination compound 2a. An assignment of their 1H-NMR
must increase an internal energy of the intermediate bi- spectra was based on homonuclear spin-spin decoupling
nuclear coordination compounds due to a set of unfavour- experiments and NOE differential spectroscopy.[31]
able non-bonding interactions,[29] thus stimulating an inter-
The IR spectrum of dimer 1a contains a single intense
mediate dissociation to form reactive three-coordinate spec-
ies; (ii) the same steric effect must result in a weakening of
the PdϪN bond in the coordination intermediate,[30] to
make the palladium(II) centre more electrophilic. Both ef-
fects essentially facilitate the CϪH bond activation.
absorption of the aromatic CϪH bonds out-of-plane defor-
mation at 740 cmϪ1[18,19,32,33] compared to two bands found
in the IR spectra of the free starting amine HL1 and coordi-
nation complex 2a (720, 785 and 705, 750 cmϪ1, respec-
1
tively). The H-NMR spectra of mononuclear adducts 3a
and 4a reveal only four well-resolved single-proton reso-
nances of the ortho-phenylene group in the aromatic field
(δ ϭ 6.0Ϫ7.0), compared to two complicated multiplets of
ortho (δ ϭ 7.00, 4 H) and meta and para protons (δ ϭ 7.27,
6 H) further downfield in the case of coordination com-
plex 2a.
For the subsequent spectral and structural investigations
of the new dimeric complex 1a, its mononuclear derivative
with [D5]pyridine (3a) was generated in situ, and an adduct
with PPh3 (4a) was prepared by a standard µ-chloro
bridge cleavage.
The nine-proton singlet of the tBu group (δ ϭ 1.23Ϫ1.30
for adducts 3a and 4a) unambiguously supports the sugges-
tion that the palladation site is the ortho-position of ben-
zylamine Ph ring. Both the diastereotopic nonequivalence
(∆δ ϭ 0.3Ϫ2.7 ppm) and the coordination shift values
(∆δ ϭ 1.7Ϫ4.5 ppm) observed for the two NH2 protons in
the complexes 2aϪ4a confirm the Pd-coordination of the
primary amino group. Thus, the η1-N-coordination of the
HL1 ligand in complex 2a and η2-C,N-binding of the (L1)Ϫ
ligand in adducts 3a, 4a and also in starting dimer 1a are
evident enough.
The assignment of the signals of the aromatic pallada-
cycle protons is important not only for the confirmation of
the ortho-palladated structure and the geometric configura-
tion of these and related complexes, but also for the utili-
zation of the palladacycle as an internal reference in the
estimation of the stereochemistry of Pd-bonded chiral li-
gands.[34Ϫ39]
Scheme 4. Mononuclear derivatives of the ortho-palladated primary
amine HL1
The identity of the samples of dimer 1a prepared under
all afore-mentioned conditions was confirmed by their
melting points, TLC of complex 1a and the 1H-NMR spec-
tra of its mononuclear derivatives 3a and 4a.
Structure of Complexes 1a؊4a in Solution
The ortho-palladated structure of complexes 1a, 3a and
1
4a was confirmed by the IR and H-NMR spectra (Table
Unfortunately, most publications devoted to the ortho-
1) using a comparison with the spectral data for simple co- palladation of primary benzylamines either disregard any
1
Table 1. H-NMR-spectroscopic data for ligand HL1 and its palladium complexes 2Ϫ4[a]
[b]
Side chain protons
Aromatic protons
H5, 1 H
tBu, 9 H α-CH, 1 H
NH1, 1 H
NH2, 1 H
H6, 1 H
H4, 1 H
H3, 1 H
HL1
0.91 (s)
0.83 (s)
3.69 (s)
1.40 (br. s., 2 H)
7.21Ϫ7.32 (m, 5 H)
7.00 (m, 2 H, ortho-H), 7.27 (m, 3 H, meta- and para-H)
2a[c]
3.73 (dd,
3.11 (br. tr,
3.81 (br. d,
3
3
2JHNH ϭ 11.0,
3JHNCH ϭ 11.2)
2JHNH ϭ 11.0,
3JHNCH ϭ 3.1)
1
JHCNH ϭ 11.2,
JHCNH ϭ 3.1)
2
3a
1.23 (s)
1.30 (s)
3.96 (d,
3.16 (br. d,
5.89 (br. m,
6.04 (A-part,
3J5,6 ϭ 7.7,
4J4,6 ϭ 1.2)
6.75 (B-part,
3J4,5 ϭ 7.4,
4J3,5 ϭ 1.4)
6.98 (C-part,
3J3,4 ϭ 7.6,
4J4,6 ϭ 1.2)
7.01 (D-part,
3J3,4 ϭ 7.6,
4J3,5 ϭ 1.4)
3JHCNH ϭ 5.7)
2JHNH ϭ 10.7)
3JHCNH ϭ 5.7,
2JHNH ϭ 10.7)
4a[d]
4.06 (dd,
3.73 (br. dd,
2JHNH ϭ 10.6,
3JHNP ϭ 3.1)
4.01 (br. m,
6.39 (A-part,
3J5,6 ϭ 7.7,
4J4,6 ϭ 1.3,
JHP ϭ 6.0
6.44 (B-part,
3J4,5 ϭ 7.4,
4J3,5 ϭ 1.5,
JHP ϭ 0.5)
6.83 (C-part,
3J3,4 ϭ 7.4,
4J3,5 ϭ 1.3)
7.00 (D-part,
3J3,4 ϭ 7.7,
4J3,5 ϭ 1.5)
3JHCNH ϭ 5.8,
4JHP ϭ 6.5)
2JHNH ϭ 10.6,
3JHNCH ϭ 5.8,
3JHNP ϭ 5.0)
4a[e,f]
1.28 (s)
4.07 (dd,
3.58 (br. d,
5.18 (br. m,
6.31 (m, 2 H, H5 ϩ H6)
6.78 (m)
6.95 (m)
3JHCNH ϭ 6.0,
4JHP ϭ 7.0)
2JHNH ϭ 10.6,
3JHNP Յ 1.0)
2JHNH ϭ 10.6,
3JHNCH ϭ 6.0,
3JHNP ϭ 4.0)
[a]
[b]
Spectra recorded at 400 MHz in CDCl3 (except where noted), δ in ppm relative to TMS as internal standard, J in Hz. Ϫ
ABCD
system in the case of 3a, ABCDX system (X ϭ 31P) for 4a. Ϫ
Spectrum of individual (R,R)* diastereomer. Ϫ
Signals of PPh3
[c]
[d]
3
[e]
protons (CDCl3): δ ϭ 7.36 (m, 6 H, meta-H), 7.42 (m, 3 H, para-H), 7.73 (m, 6 H, JHP 11.1, ortho-H). Ϫ Spectrum was recorded in
[f]
3
[D6]acetone/CDCl3, 10:1. Ϫ Signals of PPh3 protons: δ ϭ 7.35 (m, 6 H, meta-H), 7.42 (m, 3 H, para-H), 7.70 (m, 6 H, JHP 11.2,
ortho-H), 1.95 (s, 3 H, MeCN).
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039
1031

V. V. Dunina et al.
FULL PAPER
assignments of the aromatic protons[9,13,21] or do so only in rings, respectively. The large high-field value of the H6 res-
1
part.[11] Moreover, several attempts to assign these signals onance in the H-NMR spectrum of adduct 3a (δ ϭ 6.04)
for phosphane adducts[8][10] gave a result which strongly may be considered as an indication of the nearly orthogonal
disagreed with the usual trends reported for the related de- mutual arrangement of the pyridine and palladated phenyl-
[33,42Ϫ44]
rivatives of secondary[7,18,19,40,41] and tertiary
ben- ene rings.[49]
zylamines. This disagreement includes three points: (i) the
In the case of phosphane derivative 4a, the signal of the
most downfield signal was assigned to the H6 proton (near- H6 proton appears at δ ϭ 6.39, with the constant JHP
ϭ
est to the metallation site); (ii) as a result, the long-range 6.0 Hz, which is in close agreement with the values pre-
5
constant JHP of a very high value was ascribed to the H3 viously reported for related derivatives of primary aralkyl-
proton that is far remote from the 31P nucleus; (iii) at the amines (JHP ϭ 5.5Ϫ6.4 Hz).[50][51] For the more distant aro-
same time, the corresponding constant for the H6 proton is matic H5 proton, this value decreases to JHP ϭ 0.5 Hz; it
lacking despite its proximity to the 31P nucleus being closer can only be detected by means of an NMR-spectrum simu-
[45]
˚
than the sum of their van der Waals radii of 3.0 A
(for lation.
˚
example, 2.78 A was found by the X-ray study of a related
The conformation of the palladacycle formed by primary
complex[46]).
benzylamine HL1 could be estimated by means of the
Such a large discrepancy between the spectral character- analysis of the HCαϪNH, HCαϪ31P and HNϪ31P spin-
istics of palladacycles for derivatives of primary benzyla- spin coupling constants. A strong dependence of the ef-
mines on the one hand, and those of secondary and tertiary ficiency of the interaction between the protons of pallada-
amines on the other, forced us to provide unambiguous sup- cycle and the 31P nucleus of Pd-bonded phosphane ligands
port for our version of assignment on the basis of the NOE on the extent of their planarity may be expected from gen-
differential spectra of the phosphane adduct 4a.
eral considerations.[52][53]
A distinction in the assignments of two NH protons
(NHeq and NHax) is not a simple problem. In most reported
1H-NMR spectra of ortho-palladated primary aralkyl-
amines, these resonances were described as broad mul-
tiplets[8,13,21,51] or broad singlets[9][11] without their assign-
ments. In the spectrum of pyridine adduct 3a recorded in
CDCl3, the diastereotopic NH2 proton signals appear as a
broad apparent singlet (∆ν > 20 Hz) and a broad doublet
at δ ϭ 5.89 and 3.14, respectively, with a typical geminal
1
Figure 1. Selected H{1H} NOEs for the phosphane adduct 4a in
CDCl3
2
coupling constant of JNH-NH ഠ 11 Hz (cf. ref.[51,54,55]).
Homonuclear decoupling of adduct 3a showed that a
The irradiation of the low-field H3 proton of the palla- broad downfield signal at δ ϭ 5.89 originates from the NH
dated phenylene ring (δ ϭ 7.00 for the 4a in CDCl3) results proton interacting with the α-methine proton (3JHCNH
ϭ
in an enhancement of ca. 4.5% of the signal for the α- 5.7 Hz). In the case of phosphane adduct 4a, the signals of
benzylic methine proton nearest to H3 (see Figure 1); when NH protons are additionally complicated due to their spin-
1
the ortho-protons of the PPh groups were irradiated (δ ϭ spin coupling with the 31P nucleus. In the H-NMR spec-
7.73, identified on the basis of the constant value of 3JHP ϭ trum of 4a measured in a [D6]acetone/CDCl3 mixture, two
11.1 Hz) the only observed interligand response was that of NH proton signals appear as a broad multiplet at δ ϭ 5.19
3
the H6 proton (δ ϭ 6.39, ca. 2%). Hence, a downfield posi- (3JHNP ϭ 4.0 Hz, JHCNH ϭ 6.0 Hz) and a doublet at δ ϭ
tion of the H3 signal and an upfield location of the H6 res- 3.58 (3JHNP Յ 1.0 Hz) with a geminal coupling constant of
onance of the palladated phenylene group are evident.
2JHNH ϭ 10.6 Hz.
Under these conditions, the α-methine proton of 4a ap-
1
Thus, in the H-NMR spectra of the arylphosphane ad-
ducts of ortho-palladated primary benzylamines, the se- pears as a doublet of doublets at δ ϭ 4.13 with the 3JHCNH
quence of the signals of the phenylene H6ϪH3 protons re- and JHCP constants of 6.0 and 7.0 Hz, respectively. Both
4
mains the same as in the case of their analogues derived of these constants seem to be extremely useful for confor-
from secondary and tertiary amines, namely, from high field mational analysis. The dependence of 1H-31P spin-spin
(H6) to low field (H3). This tendency, as well as the detec- coupling on the position of the α-CH proton (pseudo-axial
tion of the JHP constant for the H6 proton, may be used for or pseudo-equatorial) becomes evident from the spectral
the reliable estimation of the geometric configuration and pattern for α-nonsubstituted palladacycles. Only one of the
for the solution of stereochemical problems by means of diastereotopic protons of the α-CH2 group in phosphane
NOE technique (cf. ref.[39,47,48]) using palladacycles as an adducts of ortho-palladated secondary[7][19] and tertiary
internal reference.
benzylamines[56] shows a coupling with the 31P nucleus
4
A trans(C,Cl) geometry of the mononuclear derivatives (4JHCP ϭ 2.0Ϫ4.2). The average constant value of JHCP
ϭ
of dimer 1a seems to be evident enough from the high field 2.8 Hz reported for the two equivalent (on NMR time
position of the signals of the H6 proton (3a) or the H6 and scale) α-benzylic protons in the more flexible α,N-nonsub-
H5 protons (4a) of the palladated phenylene moiety, caused stituted palladacycles,[50] may be used as a middle point for
4
by the influence of anisotropy of the pyridine or P-phenyl the conformation estimates. The JHCP ϭ 6Ϫ8 Hz reported
1032
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039

Steric Promotion of Aromatic CϪH Bond Activation in Primary Benzylamines
FULL PAPER
for the strictly planar aldimine palladacycles may serve as CarplusϪConroy equation[64] (ca. Ϫ30° and Ϫ150°, respec-
3
the upper boundary point.[21][57]
tively). Hence, nearly equal values of the JHCNH constant
4
The values of the JHCP constant for the α-methine pro- are expected for both of the NH protons of the primary
ton ranging within 6.5Ϫ7.0 Hz in the H-NMR spectra of amino group. Moreover, the CαϪH bond in this confor-
1
complex 4a are close to the above-mentioned upper limit. It mation should be oriented nearly orthogonally to the mean
4
may serve as an indication of its pseudo-equatorial position coordination plane, and the JHCP constant for this proton
which is in agreement with the prediction of a pseudo-axial should be approximately zero (cf. ref.[52,53,65]).
position of a bulky α-tBu group. A comparison of these
All these arguments allow for the exclusion of the δ(S)
conformation of palladacycle in 3a and 4a from the con-
sideration. It should be noted that the opposite (and incor-
rect) conclusion can be made from the analysis of “classic”
characteristics with those for the related derivatives of sec-
ondary[18][41] and tertiary α-tert-butylbenzylamines[44]
(4JHCP ϭ 5.4Ϫ6.1 Hz) allows for a suggestion that a steric
interaction between the α-tBu substituent and the aromatic
Newman projections constructed without regard for the
H3 proton is the main factor controlling the palladacycle
palladacycle flattening (with all dihedral angles equal to
3
conformation. The role of the N-substituents seems not to
60°): nearly equal JHCNH constant values have to be pre-
be very significant in this case, in contrast with the assump-
dicted for both NH protons in the λ(S) conformation and
tion made previously.[58][59]
different values only for the δ(S) skew-envelope. The differ-
3
Unfortunately, no information could be found regarding
ence in the predicted values of the JHCNH constant may
4
the JHCP constant for the α-methine proton of related α-
serve as a basis of the reliable criteria for identification of
Me-substituted palladacycles based on primary arylalkyl-
the axial and equatorial NH protons in the palladacycles
amines; usually this signal was described as a mul-
derived from primary and secondary benzylamines.
tiplet,[8][13] or a broad multiplet.[9][21] If the signal of the α-
As additional support for the axial position of the tBu
methine proton for the ortho-palladated derivative of pri-
group in the palladacycle of λ(S) or δ(R) conformation, the
mary α-(1-naphthyl)ethylamine is really a pure quadruplet
downfield shift of its signal compared to the free ligand
as was reported,[51] the value of this constant must be equal
HL1 should be mentioned (∆δ ϭ 0.4Ϫ0.5 ppm for the com-
to zero. Hence, the palladacycle must adopt the confor-
plexes 3a and 4a). Commonly, such an effect is considered
mation with the α-Me group in an equatorial position,
as a consequence of the metal anisotropy influence.[66][67]
which contradicts the data (4JHCP ϭ 5.8-6.3 Hz) reported
In the case of 3a and 4a complexes, the actual shift calcu-
for related derivatives of the corresponding tertiary amine
lated for one proton must be essentially higher (∆δ up to
containing η1-P-bonded phosphane ligands.[60Ϫ63]
4 ppm) if one keeps in mind that each of the nine protons
A valuable additional source of stereochemical infor-
of the tBu group spends nearly one ninth of all time in the
mation is the HNϪCαH coupling. Newman projections
proximity to the metal centre. Such a high shift may serve
along the NϪCαH bond for two possible conformations of
as an indirect indication for a weak tBu···Pd agostic interac-
the palladacycle, λ(S) and δ(S) (Figure 2), were constructed
tion.[66Ϫ69]
with regard to the real values of the torsion angles taken
from the X-ray study of adduct 4a (see the next section).
All these data support our statement that the pallada-
cycles of dimer 1a and its mononuclear derivatives 3a and
4a exist in solutions predominantly in the λ(S) or δ(R) con-
formation with the equatorial α-CH proton and the axial
tBu group. This result is important for further practical use
of this new chiral matrix, because the conformational rigid-
ity of palladacycles enhances their ability for chiral recog-
nition.[36,52,53,63,70,71]
3
Figure 2. Newman projections of (S)-palladacycle along the
NϪC(α) bond for two possible conformations, λ(S) and δ(S)
The relation between the JHNP constants for the two
NϪH protons of the primary amino group is difficult to
predict. Taking into account the similar dihedral angles
The most important dihedral angles between the CαϪH HaxϪNϪPdϪP and HeqϪNϪPdϪP, found for adduct 4a
and the pseudo-axial and pseudo-equatorial NϪH bonds for in the crystal (110.7° and Ϫ130.9°, respectively), it is con-
the λ(S) conformation are equal to ca. 31° and Ϫ88°, ceivable to suggest that both NH protons may reveal a spin-
respectively, in the crystal of 4a. As a result, an observable spin coupling with the 31P nucleus. A little more efficient
3JHCNH constant can be predicted only for one of the two interaction may be expected for the pseudo-equatorial NH
1
NH protons, namely NHax, if the same λ(S) conformation proton. However, the H-NMR spectrum of 4a reveals the
3
remains in solution. This particular situation takes place in reverse relation, namely, JHNP ϭ 4.0Ϫ5.0 and 1.0Ϫ3.1 Hz
3
the case of adducts 3a and 4a: JHCNH ϭ 5.6Ϫ6.0 Hz was for the NHax and NHeq protons, respectively. Examples of
found for only one downfield resonance of the NH proton. equal[37,38,62,72] and substantially different[36,60,62,73] values
In contrast, in the case of the hypothetical δ(S) confor- of the 4JHP constant for NMe groups may be found among
mation of the palladacycle, the dihedral angles between the the spectra of phosphane complexes with the ortho-palla-
CαϪH and axial and equatorial NϪH bonds should be dated N,N-dimethyl-α-(1-naphthyl)ethylamine, despite a
nearly equal to one another from the point of view of the high conformational stability of this palladacycle. This ob-
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039
1033

V. V. Dunina et al.
FULL PAPER
servation made the JHP values for NH2 and NMe2 groups
less suitable for the conformation analysis.
The structure of coordination complex 2a can shed some
light on the nature of the steric promotion of ortho-palla-
1
dation reactions. Its H-NMR spectrum indicates a restric-
ted rotameric mobility of η1-N-coordinated ligand HL1. A
3
large difference between JHCNH constant values for two
NH protons (11.2 and 3.1 Hz) pointed to the φ3 or φ5 con-
formation of Pd-bonded amine, with transoid orientation of
one of the NH protons relative to the α-methine one. For
the alternative φ1 rotamer, nearly equal and low values of
3JHCNH constant may be predicted[64] from the Newman
projections (Figure 3).
Figure 4. Molecular structure of the acetonitrile solvate of phos-
Figure 3. Newman projections of Pd-coordinated (S)-α-phenyl-
neopentylamine along the NϪC(α) bond for three main rotamers,
φ1, φ3 and φ5
phane adduct 4a
˚
Table 2. Bond lengths [A] for [2-(1-amino-2,2-dimethylpropyl)phenyl-
C,N]chloro(triphenylphosphane)palladium(II)·MeCN solvate 4a
The absence of deshielding effects (caused by the metal
anisotropy[66Ϫ69]) for the tBu protons in the spectra of com-
plex 2a (δ ϭ 0.83 is close to δ ϭ 0.91 for the free ligand
HL1) allows for the differentiation between φ3 and φ5 rot-
americ forms in favour of the latter, where the tBu group is
far remote from the palladium centre.
Therefore, at the stage of monodentate N-coordination,
the bulky HL1 ligand adopts predominantly the φ5 confor-
mation with the Ph ring disposed in proximity to the metal
centre (cf. ref.[74][75]). This conformation seems to be the
most suitable starting state for the subsequent ortho-palla-
dation. In contrast, monodentate N-coordination of the re-
lated α-Me-substituted primary benzylamine in the frame-
work of binuclear µ-acetato coordination complex occurs
in the φ3 conformation with remote transoid position of the
Ph ring with respect to the palladium centre.[10] Such a dif-
ference in the starting rotameric states of N-coordinated α-
tBu- and α-Me-substituted primary benzylamines may con-
tribute to some extent to the steric promotion of the CϪH
bond activation in the former instance.
Pd(1)ϪC(1)
Pd(1)ϪN(1)
Pd(1)ϪP(1)
Pd(1)ϪCl(1)
P(1)ϪC(24)
P(1)ϪC(12)
P(1)ϪC(18)
N(1)ϪC(7)
C(1)ϪC(2)
C(1)ϪC(6)
C(2)ϪC(3)
C(3)ϪC(4)
C(4)ϪC(5)
C(5)ϪC(6)
C(6)ϪC(7)
C(7)ϪC(8)
C(8)ϪC(11)
C(8)ϪC(10)
C(8)ϪC(9)
C(12)ϪC(17)
2.006(4)
2.087(4)
2.256(2)
2.3996(14)
1.817(5)
1.827(4)
1.826(5)
1.496(5)
1.385(6)
1.423(6)
1.395(6)
1.369(7)
1.384(7)
1.387(6)
1.510(6)
1.544(6)
1.526(7)
1.531(7)
1.537(7)
1.377(7)
C(12)ϪC(13)
C(13)ϪC(14)
C(14)ϪC(15)
C(15)ϪC(16)
C(16)ϪC(17)
C(18)ϪC(19)
C(18)ϪC(23)
C(19)ϪC(20)
C(20)ϪC(21)
C(21)ϪC(22)
C(22)ϪC(23)
C(24)ϪC(29)
C(24)ϪC(25)
C(25)ϪC(26)
C(26)ϪC(27)
C(27)ϪC(28)
C(28)ϪC(29)
N(2)ϪC(1Љ)
1.381(6)
1.379(7)
1.356(8)
1.370(8)
1.382(8)
1.383(6)
1.391(6)
1.376(7)
1.365(9)
1.372(9)
1.403(8)
1.379(6)
1.390(6)
1.376(7)
1.371(8)
1.374(8)
1.390(7)
1.129(11)
1.448(11)
C(1Љ)ϪC(2Љ)
terized structurally, namely, that of the α-Me-substituted
analogue (R)-4b[13] and of the 2-phenylglycine methyl ester
derivative (R)-4c.[11] The related derivative of 2-phenylethyl-
amine containing a six-membered palladacycle was also
studied by X-ray diffraction.[10]
Complex 4a exists in the crystal as a racemate; the unit
X-ray Structure Investigation of Phosphane Derivative cell contains two centrosymmetrically related enantiomers
4a
of the mononuclear complex and two solvate molecules of
acetonitrile (Figure 4). ortho-Palladated structure of adduct
Suitable crystals of triphenylphosphane adduct 4a were
grown from a dichloromethane/acetonitrile mixture. The
molecular structure of the complex is presented in Figures
4 and 5; selected bond lengths and angles are given in
Tables 2 and 3, respectively.
Only two other triphenylphosphane adducts of ortho-pal-
ladated primary benzylamines have been previously charac-
1034
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039

Steric Promotion of Aromatic CϪH Bond Activation in Primary Benzylamines
FULL PAPER
[10,11,13]
˚
Table 3. Bond angles [°] for [2-(1-amino-2,2-dimethylpropyl)phenyl- 2.092Ϫ2.089 A for 4b and 4c, respectively).
C,N]chloro(triphenylphosphane)palladium(II)·MeCN solvate 4a
The
shorter value of the PdϪN bond in the structures of com-
plexes 4a؊c, compared to that reported for the related de-
C(1)ϪPd(1)ϪN(1) 80.6(2)
C(11)ϪC(8)ϪC(7)
109.0(4)
112.0(4)
109.4(4)
˚
rivatives of tertiary benzylamines [for example, 2.19(1) A
C(1)ϪPd(1)ϪP(1) 96.00(13) C(10)ϪC(8)ϪC(7)
N(1)ϪPd(1)ϪP(1) 176.38(11) C(9)ϪC(8)ϪC(7)
for N-isopropyl-N,α-dimethylbenzylamine analogue[46]
]
C(1)ϪPd(1)ϪCl(1) 169.37(12) C(17)ϪC(12)ϪC(13) 117.4(4)
N(1)ϪPd(1)ϪCl(1) 89.04(11) C(17)ϪC(12)ϪP(1)
P(1)ϪPd(1)ϪCl(1) 94.39(5)
C(24)ϪP(1)ϪC(12) 103.4(2)
C(24)ϪP(1)ϪC(18) 107.2(2)
C(12)ϪP(1)ϪC(18) 103.0(2)
C(24)ϪP(1)ϪPd(1) 111.8(2)
C(12)ϪP(1)ϪPd(1) 115.42(14) C(16)ϪC(17)ϪC(12) 121.2(5)
C(18)ϪP(1)ϪPd(1) 114.9(2)
C(7)ϪN(1)ϪPd(1) 110.5(3)
may serve as reliable evidence in favour of a tighter coordi-
nation of the primary amino group with the palladium cen-
tre.
The palladium atom in complexes 4a and 4b has nearly
square-planar coordination with a slight tetrahedral distor-
tion: the dihedral angles between the planes (CPdN) and
(PPdHal) fall within the range 2.7Ϫ7.8°.
The palladacycle conformation in adduct 4a may be de-
scribed as a distorted envelope with the nitrogen atom dis-
placed from the plane (PdC1C6C7) of the remaining four
122.2(4)
120.4(3)
C(13)ϪC(12)ϪP(1)
C(14)ϪC(13)ϪC(12) 121.4(5)
C(15)ϪC(14)ϪC(13) 120.2(5)
C(14)ϪC(15)ϪC(16) 119.7(5)
C(15)ϪC(16)ϪC(17) 120.0(6)
C(19)ϪC(18)ϪC(23) 118.6(4)
C(19)ϪC(18)ϪP(1)
C(23)ϪC(18)ϪP(1)
118.3(3)
123.1(4)
C(2)ϪC(1)ϪC(6)
117.1(4)
C(2)ϪC(1)ϪPd(1) 129.7(3)
C(6)ϪC(1)ϪPd(1) 112.9(3)
C(20)ϪC(19)ϪC(18) 121.9(5)
C(21)ϪC(20)ϪC(19) 119.4(6)
C(22)ϪC(21)ϪC(20) 120.4(5)
C(21)ϪC(22)ϪC(23) 120.6(5)
C(18)ϪC(23)ϪC(22) 119.0(5)
C(29)ϪC(24)ϪC(25) 118.6(4)
C(1)ϪC(2)ϪC(3)
C(4)ϪC(3)ϪC(2)
C(3)ϪC(4)ϪC(5)
C(4)ϪC(5)ϪC(6)
C(5)ϪC(6)ϪC(1)
C(5)ϪC(6)ϪC(7)
C(1)ϪC(6)ϪC(7)
N(1)ϪC(7)ϪC(6)
N(1)ϪC(7)ϪC(8)
C(6)ϪC(7)ϪC(8)
122.3(4)
119.9(5)
119.5(5)
121.4(4)
119.8(4)
121.9(4)
118.2(4)
105.1(3)
112.3(4)
117.0(4)
˚
˚
atoms by Ϫ0.5999 A as compared to Ϫ0.5113 and 0.399 A
for 4b and 4c, respectively. As a measure of a palladacycle
twisting, the value of the dihedral angle between the phenyl-
ene ring of the benzylaminate ligand (C6H4) and the mean
coordination plane (PdC1HalNP) may be used. The values
equal to 20.0 and 15.5° for 4a and 4b, respectively, indicate
a more puckered conformation of the α-tBu-substituted
palladacycle.
C(29)ϪC(24)ϪP(1)
C(25)ϪC(24)ϪP(1)
119.5(4)
121.7(4)
C(26)ϪC(25)ϪC(24) 121.0(5)
C(27)ϪC(26)ϪC(25) 120.0(5)
C(26)ϪC(27)ϪC(28) 119.8(5)
C(27)ϪC(28)ϪC(29) 120.5(5)
C(24)ϪC(29)ϪC(28) 120.0(5)
C(11)ϪC(8)ϪC(10) 108.6(5)
C(11)ϪC(8)ϪC(9) 107.9(5)
C(10)ϪC(8)ϪC(9) 109.8(5)
N(2)ϪC(1Љ)ϪC(2Љ)
178.3(11)
The stereochemistry of palladacycle conformation is in
1
agreement with the H-NMR data in solutions, namely, λ
for the (S), and δ for the (R) enantiomer. The bulky α-tBu
4a (and, therefore, that of starting dimer 1a) is supported. substituent assumes a position close to the axial one and
The complex has a trans(C,Cl) geometry, expected from the the α-methine proton is arranged nearly equatorially:
spectral data. The PdϪC and PdϪN bond lengths [2.006(4) C7ϪC8 and C7ϪH7a bonds form angles of 168.4° (an adjac-
˚
and 2.087(4) A, respectively] are similar to those reported ent angle 11.6°) and 66.9°, respectively, with the normal to
for other derivatives of primary benzylamines (2.019 and the mean coordination plane.
Figure 5. The mutual arrangement of two H-bonded (NϪH...Cl) centrosymmetrically related enantiomers of phosphane adduct 4a in
the crystal
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039
1035

V. V. Dunina et al.
FULL PAPER
clear decoupling, NOE experiments and spectral simulation. Ϫ IR
spectra were recorded with a UR-250 spectrometer.
It is probable that a weak agostic interaction between one
of the Me groups of the tBu substituent with the palladium
centre contributes to some extent to a stabilization of this
Solvents: Acetonitrile was distilled under argon from P2O5 and then
from K2CO3; benzene, hexane and ether were dried with the appro-
priate drying agents and then distilled under argon from Na; anhy-
drous MeOH was prepared by distillation from MgOMe; dichloro-
methane was passed through a short Al2O3 column and distilled
under argon; CCl4 was refluxed in the presence of P2O5 and dis-
tilled; acetone of highest purity, and [D1]chloroform (from Aldrich)
conformation: the H10b···Pd distance is 2.802 A, which is
˚
less than the sum of the van der Waals radii for these atoms
(3.1 A[45]). It is noteworthy that similar secondary interac-
˚
tions were previously observed in the structures of ortho-
palladated derivatives of α-tert-butyl-N,N-dimethylbenzyl-
amine, namely, of a dimer (2.91Ϫ2.92 A[76]) and its mono-
˚
˚
were used as received; [D5]pyridine was dried with 4-A molecular
nuclear derivative with pyridine (2.74Ϫ2.85 A[20]).
˚
sieves and distilled from KOH.
The presence of the primary amino group in the pallada-
cycle of 4a results in a specific packing of molecules in the
crystal, based on the hydrogen bonds of the type NϪH···Cl
(see Figure 5). Two enantiomers of the racemic complex are
Reagents: Sodium acetate trihydrate was dehydrated by heating at
400°C; triphenylphosphane was purified by twofold recrystalliza-
tion from a benzene/hexane mixture; hydroxylamine hydrochloride
was used as received; lithium tetrachloropalladate was prepared ac-
bonded in a centrosymmetric dimer by means of two rather cording to a published procedure[6] and carefully dried in vacuo (3
strong hydrogen bonds, with the NH1a···Cl1a distances be-
Torr) over P2O5; palladium(II) acetate was purchased from Aldrich
and used as received.
˚
ing equal to 2.416 A, which is essentially less than the sum
of the van der Waals radii for these atoms (3.3 A[45]). The
˚
Starting Compounds: tert-Butyl phenyl ketone oxime was prepared
by heating of pivalophenone[77] (12.80 g, 79 mmol) with hy-
droxylamine hydrochloride (6.05 g, 87 mmol) in the presence of
AcONa·3H2O (11.8 g, 87 mmol) in aqueous methanol according to
a published procedure;[78] after recrystallization from EtOH, the
oxime was isolated in an 88% yield (17.93 g); m.p. 163Ϫ164°C. Ϫ
1H NMR (CDCl3): δ ϭ 1.15 (s, 9 H, tBu), 7.12 (dd, 3JHH ϭ 7.9 Hz,
4JHH ϭ 1.5 Hz, 2 H, ortho-H of Ph group), 7.33Ϫ7.45 (m, 3 H,
meta-H and para-H of Ph group), 9.25 (br. s, 1 H, OH).
˚
similar NH···Br hydrogen bonds (2.46 A) were also found
in the structure of α-Me-substituted analogue 4b.[13]
In the crystal, the solvate acetonitrile is bonded with the
coordinated chlorine atom by means of a hydrogen bond
involving a methyl hydrogen atom (see Figure 4): The
1
CH2”f···Cl distance of 2.672 A is also less than the sum of
˚
the van der Waals radii for these atoms (3.3 A[45]).
˚
Racemic ␣-tert-Butylbenzylamine (HL1) was prepared by reduction
of tert-butyl phenyl ketone oxime (16.68 g, 0.094 mmol) under con-
Conclusions
ditions previously described[79] in
a yield of 72% (11.06 g,
0.068 mmol); b.p. 108Ϫ109°C/18 Torr; 1H NMR (CDCl3): δ ϭ 0.91
(s, 9 H, tBu), 1.40 (br. s, 2 H, NH2), 3.69 (s, 1 H, α-CH), 7.21Ϫ7.32
(m, 5 H, Ph).
The possibility of a steric promotion of the aromatic
CϪH bond activation in primary benzylamines bearing
only one bulky α-substituent was shown. ortho-Palladation
of α-phenylneopentylamine may be achieved under very
mild conditions, using Li2PdCl4 in methanol at room tem-
perature.
To the best of our knowledge, this work is the first pres-
entation of the completely interpreted H-NMR spectra of
ortho-palladated primary benzylamine. It allows for estab-
lishing that the palladacycle formed adopts the λ(S) or δ(R)
Synthesis of Cyclopalladated Compounds: All reactions were per-
formed under argon in anhydrous solvents. Reactions were moni-
tored by means of TLC on Silufol plates using an ether/heptane
(5:1) mixture as eluent (unless otherwise indicated). In the case of
µ-acetato complex formation, its conversion into the µ-chloro ana-
logue by AcOϪ/ClϪ metathesis (by treatment of a test probe of
reaction mixture with LiCl solution in MeOH) was undertaken be-
fore spotting on the TLC plate. Dimeric complexes obtained were
1
conformation with an axially oriented α-tBu group both in purified (unless otherwise indicated) using flash chromatography
on a “dry” column[80][81] (Silpearl, d ϭ 2.5 cm, h ϭ 3.5 cm) with
the eluentsЈ polarity increasing (ether/heptane mixtures in the ra-
tios from 1:10 up to 10:1 and then pure ether).
the crystal and in solution. It seems reasonable to propose
that such stereochemistry is dictated by unfavourable inter-
actions between a bulky equatorial α-substituent and the
neighbouring phenylene proton in alternative δ(S) or λ(R)
Di-µ-chlorobis[2-{1-amino-2,2-dimethylpropyl}phenyl-C,N]-
conformations. The conformational stability of a new palla-
dipalladium (II) (1a)
dacycle shows a promising utility in a homochiral state as
a matrix for the diverse processes of chiral recognition.
1. ortho-Palladation with Palladium(II) Acetate: i) Palladium ace-
tate (0.206 g, 0.918 mmol) was added to racemic α-tert-butylben-
zylamine HL1 (0.150 g, 0.919 mmol) in methanol (5 mL) and the
reaction mixture was stirred for 24 h at room temp. A solution of
Experimental Section
General: 1H- and 31P-NMR spectra were recorded with a Varian
lithium chloride (0.190 g, 4.48 mmol) in methanol (3 mL) was ad-
ded and the mixture was stirred for 2 h and then filtered. The resi-
due was washed with methanol, combined with the mother liquor,
concentrated in vacuo, and redissolved in dichloromethane (10
VXR-400 spectrometer operating at the frequencies 400 and
1
161.9 MHz for H and 31P nuclei, respectively. The measurements
were carried out at ambient temperature in CDCl3 solutions (unless mL). The solution formed was washed with water to remove inor-
otherwise indicated). The proton chemical shifts are reported in ganic components, dried with sodium sulfate and concentrated to
parts per million relative to TMS as internal standard; the 31P a volume of 2Ϫ3 mL. Flash column chromatography gave the coor-
chemical shifts are given relative to H3PO4 as an external reference. dination complex 2a (from the first fraction) as a pale yellow mi-
The assignment of signals was performed on the basis of homonu- crocrystalline solid in a yield of 19% (0.0443 g, 0.088 mmol) after
1036
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039

Steric Promotion of Aromatic CϪH Bond Activation in Primary Benzylamines
FULL PAPER
recrystallization from a dichloromethane/hexane mixture. From the
second fraction, the µ-Cl dimer 1a was isolated as a lime-yellow
amorphous powder in a yield of 70% (0.195 g, 0.321 mmol) after
precipitation from benzene by hexane.
ual diastereomer 2a see Table 1. Ϫ C22H34Cl2N2Pd (503.85): calcd.
C 52.44, H 6.80, N 5.56; found C 52.73, H 6.73, N 5.76.
1H NMR of mixture of diastereomers (R,R)*/(R,S)-2a (CDCl3):
δ ϭ 0.81 (s, 18 H, tBu), 0.82 (s, 18 H, tBu), 3.15 (br. t, 2 H, NH),
3.19 (br. t, 2 H, NH), 3.67 (br. m, 4 H, NH), 3.74 (m, 4 H, α-CH),
7.04 (m, 8 H, ortho-H of Ph groups), 7.26 (m, 12 H, meta- and
para-H of Ph groups).
For dimer 1a: m.p. 217Ϫ219°C (dec). Ϫ Rf ϭ 0.54. Ϫ IR (nujol):
1
ν˜ ϭ 3213, 3308 cmϪ1 [ν(NH2)]; 740 cmϪ1 [δ(CHar)]. Ϫ H NMR
(CDCl3, isomeric mixture): δ ϭ 1.14 (br. s, 9 H, tBu), 2.87 and 4.03
(br. s, 2 H, NH2), 3.62 and 3.66 (s, 1 H, α-CH), 6.72 (br. s, 1 H,
C6H4), 6.89 (m, 2 H, C6H4), 7.19 (br. s, 1 H, C6H4). Ϫ
C22H32Cl2N2Pd2 (608.26): calcd. C 43.44, H 5.30, N 4.23; found C
43.85, H 5.29, N 4.61.
[2-{1-Amino-2,2-dimethylpropyl}phenyl-C,N]chloro([D5pyridine)]-
palladium(II) (3a) was generated in situ by the addition of 2Ϫ3
drops of [D5]pyridine to a solution of dimer 1a in CDCl3 in an
1
NMR tube. For H-NMR data of complex 3a see Table 1.
For coordination complex 2a: m.p. 171Ϫ173°C, Rf ϭ 0.92.
[2-{1-Amino-2,2-dimethylpropyl}phenyl-C,N]chloro(triphenyl-
phosphane)palladium (II) (4a) was prepared by treatment of dimeric
complex 1a (0.100 g, 0.164 mmol) with triphenylphosphane
(0.086 g, 0.33 mmol) in benzene (15 mL). The reaction mixture was
stirred for 40 min at room temp. and concentrated to dryness.
Recrystallization from dichloromethane/hexane and then from di-
chloromethane/acetonitrile afforded the mononuclear adduct 4a in
ii) The same reaction of palladium acetate (0.225 g, 1.00 mmol) and
amine HL1 (0.163 g, 1.0 mmol) conducted in methanol (5 mL) at
50°C required 1.5 h for completion. After the treatment of the reac-
tion mixture as described above [i)] it afforded the dimer 1a in a
68% yield (0.207 g, 0.340 mmol); m.p. 217Ϫ219°C (dec), Rf ϭ 0.54
(ether/heptane, 5:1); Rf
ϭ 0.44 (benzene/acetone, 7:1). Ϫ
C22H32Cl2N2Pd2: calcd. C 43.44, H 5.30, N 4.23; found C 43.40,
H 5.35, N 4.27. Ϫ Under these conditions the coordination com-
plex 2a was formed in trace quantities.
Table 4. Crystal data, data collection, structure solution and refine-
ment parameters for [2-(1-amino-2,2-dimethylpropyl)phenyl-
C,N]chloro(triphenylphosphane)palladium(II)·MeCN solvate 4a
iii) In a similar way, the reaction of palladium acetate (0.281 g,
1.25 mmol) and amine HL1 (0.204 g, 1.25 mmol) was performed in
acetone (10 mL). After heating at 50°C for 50 h and treatment of
the reaction mixture as described above [i)], dimer 1a was isolated
in a yield of 68% (0.258 g, 0.424 mmol); m.p. 219Ϫ221°C (dec),
Rf ϭ 0.54.
Empirical formula
Formula weight
Colour, habit
C31H34ClN2PPd
607.42
Light yellow, prism
0.09 ϫ 0.12 ϫ 0.45
triclinic
Crystal size [mm]
Crystal system
¯
P1
Space group
Unit cell dimensions:
iv) The interaction of palladium acetate (0.212 g, 0.944 mmol) and
ligand HL1 (0.154 g, 0.943 mmol) in benzene (7 mL) at room temp.
afforded, after the stirring for 24 h and the treatment of reaction
mixture as described above [i)] the dimer 1a in a 43% yield (0.123 g,
0.202 mmol); m.p. 217Ϫ219°C (dec), Rf ϭ 0.54. Under these con-
ditions the coordination complex 2a was isolated in a 41% yield
(0.097 g, 0.193 mmol): m.p. 172Ϫ173°C, Rf ϭ 0.92.
˚
a [A]
10.112(2)
12.245(2)
12.939(3)
90.88(3)
111.45(3)
101.40(3)
1455.0(5)
2
1.386
0.806
˚
b [A]
˚
c [A]
α [°]
β [°]
γ [°]
3
˚
V [A ]
Z
Density (calcd.) [g/cm3]
Absorption coefficient
2. ortho-Palladation with Lithium Tetrachloropalladate(II): Racemic
α-tert-butylbenzylamine (0.150 g, 0.919 mmol) in methanol (20
mL) was added to a mixture of lithium tetrachloropalladate
(0.255 g, 0.973 mmol) and excess of sodium acetate (0.232 g,
2.83 mmol) in methanol (20 mL). The reaction mixture was stirred
at room temp. for 23 h; after overnight storage, it was filtered and
the residue was washed with methanol and then dichloromethane.
The combined organic solutions were concentrated in vacuo and
redissolved in dichloromethane (10 mL). Chromatographic sepa-
ration using a dry column (Silpearl, d ϭ 2.5 cm, h ϭ 5 cm, ether/
hexane from 1:10 up to 10:1 and then pure ether) and recrystalliza-
tion as described above [i)] gave dimer 1a as a limonϪyellow
amorphous powder in a yield of 50% (0.140 g, 0.230 mmol) from
the second fraction. The mixture of diastereomeric coordination
complexes (R,R)*/(R,S)-2a in an approximately 1:1 ratio (1H-NMR
data) was isolated from the first fraction in a 43% yield based on
the ligand (0.100 g, 0.198 mmol); one of diastereomers was sepa-
rated from this mixture by recrystallization from benzene/hexane
and then from dichloromethane/hexane in a yield of 12% (0.0286 g,
0.057 mmol) as a pale yellow microcrystalline solid.
[mmϪ1
]
F(000)
624
Diffractometer
EnrafϪNonius CAD-4
T [K]
293
˚
Radiation λ [A]
graphite-monochromated Mo-Kα
(0.71073)
Scan mode
ω/2θ
Scan width [°]
1.2 ϩ 0.35 tan(θ)
Min/max scan rate [°/min] 2/8
θ range [°]
2.22Ϫ24.97
Index ranges
Ϫ12 Յ h Յ 11, Ϫ14 Յ k Յ 14,
0 Յ l Յ 15
Reflections collected
Independent reflections
Absorption correction
Min. and max.
4696
4486 [R(int) ϭ 0.0326]
empirical (Ψ scan)
0.7365 and 0.9643
transmission
Decay correction
Solution method
Refinement method
none
direct methods (SHELX-86)
full-matrix least-squares on F2
(SHELXL-93)
Hydrogen treatment
all H atoms found in Fourier synthesis
and refined using riding model
Data/restraints/parameters 4486/0/330
For 1a: m.p. 217Ϫ219°C (dec), Rf ϭ 0.54. Ϫ C22H32Cl2N2Pd2
(608.26): calcd. C 43.44, H 5.30, N 4.23; found C 43.51, H 5.55,
N 4.36.
Goodness of fit on F2
0.996
Final R indices [I > 2σ(I)] R1 ϭ 0.0351, wR2 ϭ 0.0850
R indices (all data)
Extinction coefficient
Largest diff. peak and
R1 ϭ 0.0654, wR2 ϭ 0.0941
0.0000(5)
0.583 and Ϫ0.682
For (R,R)*-2a: m.p. 171Ϫ173°C, Rf ϭ 0.92. Ϫ IR (nujol): ν˜ ϭ
3310, 3230 cmϪ1 [br., ν(NH2)], 750, 790 [δ(CHar)]; (CCl4): ν˜ ϭ
3310, 3233 cmϪ1 [br., ν(NH2)]. Ϫ For 1H-NMR data of the individ-
Ϫ3
˚
hole [e ϫ A
]
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039
1037

V. V. Dunina et al.
FULL PAPER
nenko, I. P. Beletskaya, Russ. J. Org. Chem. (Engl. Transl.) 1994,
30, 1497Ϫ1506.
V. V. Dunina, L. G. KuzЈmina, E. B. GolovanЈ, O. N. Goru-
nova, Yu. K. Grishin, J. Organomet. Chem., to be submitted.
J. Albert, J. Granell, J. Minguez, G. Muller, D. Sainz, P. Valegra,
Organometallics 1997, 16, 3561Ϫ3564.
A. D. Ryabov, Chem. Rev. 1990, 90, 403Ϫ424.
A. J. Canty, G. van Koten, Acc. Chem. Res. 1995, 28, 406Ϫ413.
T. Yagyu, S.-i. Aizawa, Sh. Funahashi, Bull. Chem. Soc. Jpn.
1998, 71, 619Ϫ629.
J. Dehand, M. Pfeffer, C. R. Acad. Sci. Paris, Ser. C 1975,
281, 363Ϫ366.
H. Onoue, J. Moritani, J. Organomet. Chem. 1972, 43, 431Ϫ436.
I. Jardine, F. J. McQuillin, Tetrahedron Lett. 1972, 459Ϫ461.
A. G. J. Ligtenbard, E. K. van den Beuken, A. Meetsma, N.
Veldman, W. J. J. Smeets, A. L. Spek, B. L. Feringa, J. Chem.
Soc., Dalton Trans. 1998, 263Ϫ270.
V. V. Dunina, O. A. Zalevskaya, V. M. Potapov, Russ. Chem.
Rev. (Engl. Transl.) 1988, 57, 434Ϫ473 and refs. therein.
A. L. Seligson, W. C. Trogler, J. Am. Chem. Soc. 1991, 113,
2520Ϫ2527.
J. K. M. Sanders, J. D. Mersh, Progr. NMR Spectrosc. 1982,
15, 353Ϫ400.
a 79% yield (0.147 g, 0.260 mmol) as a colourless microcrystalline
MeCN solvate, m.p. 208Ϫ209°C, Rf ϭ 0.30 (Silufol, benzene/ace-
tone, 7:1). Ϫ IR (CCl4): ν˜ ϭ 3238 cmϪ1 (br.), 3260, 3302 cmϪ1
[ν(NH2)]. Ϫ For 1H-NMR data of complex 4a see Table 1. Ϫ
31P{1H} NMR (CH2Cl2): δ ϭ 41.57 (s). Ϫ C29H31ClNPPd·C2H3N
(607.47): calcd. C 61.29, H 5.64, N 4.61; found C 60.82, H 5.70,
N 4.24.
[20]
[21]
[22]
[23]
[24]
X-ray Crystallography for Complex 4a: Details of the X-ray experi-
ment are given in Table 5. The experimental data were collected
with an ENRAF Nonius CAD4 diffractometer using graphite-
monochomatized Mo-Kα radiation. Experimental reflections were
corrected for Lorentz and polarization effects. The hydrogen atoms
were found from a difference Fourier synthesis and included in the
refinement using the “riding model” with Biso equal to 1.5Beq of
the parent atom. The structure was refined in the anisotropic
approximation for the non-hydrogen atoms. A solvent molecule of
acetonitrile was found during the structure solution, which forms
a hydrogen bond with the coordinated chloride ligand of the main
complex. The structure was solved and refined using the SHELXS-
86[82] and SHELXL-93[83] software. Crystallographic data (exclud-
ing structure factors) for the structure reported in this paper have
been deposited with the Cambridge Crystallographic Data Centre
as supplementary publication no. CCDC-104570. Copies of the
data can be obtained free of charge on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK [Fax: int. code ϩ 44-1223/
336-033; E-mail: deposit@ccdc.cam.ac.uk].
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
L. J. Bellami, The Infrared Spectra of Complex Molecules, J.
Wiley, London, New York, 1963, chapter 5.
V. V. Dunina, V. P. Kislyi, N. S. Gulyukina, Yu. K. Grishin, I.
P. Beletskaya, Organomet. Chem. USSR (Engl. Transl.) 1992,
5, 1297Ϫ1305.
[34]
C. E. Barclay, G. Deeble, R. J. Doyle, Sh. A. Elix, G. Salem, T.
L. Jones, S. B. Wild, A. C. Willis, J. Chem. Soc., Dalton Trans.
1995, 57Ϫ65.
[35]
[36]
R. J. Doyle, G. Salem, A. C. Willis, J. Chem. Soc., Dalton Trans.
1997, 2713Ϫ2723.
D. G. Allen, G. M. McLaughlin, G. B. Robertson, W. L.
Steffen, G. Salem, S. B. Wild, Inorg. Chem. 1982, 21,
1007Ϫ1014.
[37]
[38]
Acknowledgments
G. Salem, S. B. Wild, Inorg. Chem. 1983, 22, 4049Ϫ4054.
J. W. L. Martin, J. A. L. Palmer, S. B. Wild, Inorg. Chem. 1984,
23, 2664Ϫ2668.
Q. Jiang, H. Ruegger, L. M. Venanzi, J. Organomet. Chem.
1995, 488, 233Ϫ240.
Y. Fuchita, H. Tsuchiya, Inorg. Chim. Acta 1993, 209, 229Ϫ230.
V. V. Dunina, E. B. GolovanЈ, N. S. Gulyukina, A. V. Buyevich,
Tetrahedron: Asymmetry 1995, 6, 2731Ϫ2746.
V. V. Dunina, E. B. GolovanЈ, V. V. Pedchenko, A. A. Bori-
senko, I. P. Beletskaya, Organomet. Chem. USSR (Engl. Transl.)
1992, 5, 1121Ϫ1129.
V. V. Dunina, E. B. GolovanЈ, Tetrahedron: Asymmetry 1995,
6, 2747Ϫ2754.
V. V. Dunina, M. Yu. Kazakova, Yu. K. Grishin, O. R. Maly-
shev, and E. I. Kazakova, Russ. Chem. Bull. (Engl. Transl.)
1997, N7, 1375Ϫ1384.
A. Bondi, J. Phys. Chem. 1964, 68, 441Ϫ451.
L. G. KuzЈmina, Yu. T. Struchkov, O. A. Zalevskaya, V. V. Du-
nina, V. M. Potapov, J. Gen. Chem. USSR (Engl. Transl.) 1987,
57, 2499Ϫ2507.
B.-H. Aw, S. Selvaratnam, P.-H. Leung, N. H. Rees, W. McFar-
lane, Tetrahedron: Asymmetry 1996, 7, 1753Ϫ1762.
P.-H. Leung, S. Selvaratnam, C. R. Cheng, K. F. Mok, N. H.
Rees, W. McFarlane, J. Chem. Soc., Chem. Commun. 1997,
751Ϫ752.
This work was supported in part by the Russian Foundation for
Basic Research (grant 95-03-09227).
[39]
[40]
[41]
[1]
M. Pfeffer, Pure Appl. Chem. 1992, 64, 335.
[42]
[2]
A. D. Ryabov, Synthesis 1985, 233Ϫ252.
[3]
J. Chengebroyen, M. Pfeffer, C. Sirlin, Tetrahedron Lett. 1996,
37, 7263Ϫ7266.
[43]
[44]
[4]
J. Vicente, J.-A. Abad, A. D. Frankland, M.-C. Ramirez de
Arellano, J. Chem. Soc., Chem. Commun. 1997, 959Ϫ960.
[5]
J. F. Hartwig, Synlett 1997, 329Ϫ340.
[6]
A. C. Cope, E. C. Friedrich, J. Am. Chem. Soc. 1968, 90,
909Ϫ913.
[45]
[46]
[7]
Y. Fuchita, H. Tsuchiya, A. Miyafuji, Inorg. Chim. Acta 1995,
233, 91Ϫ96.
[8]
J. Vicente, I. Saura-Liamas, M. G. Palin, P. G. Jones, J. Chem.
Soc., Dalton Trans. 1995, 2535Ϫ2539.
[47]
[48]
[9]
J. Albert, J. Granell, A. Luque, J. Minguez, R. Moragas, M.
Font-Bardia, X. Solans, J. Organomet. Chem. 1996, 522, 87Ϫ95.
[10]
J. Vicente, I. Saura-Liamas, M. G. Palin, P. G. Jones, M. C. R.
de Arellano, Organometallics 1997, 16, 826Ϫ833.
[11]
Y. Fuchita, K. Yoshinaga, Y. Ikeda, J. Kinoshita-Kawashima,
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
A. J. Deeming, I. P. Rothwell, M. B. Hursthouse, L. New, J.
Chem. Soc., Dalton Trans. 1978, 1490Ϫ1496.
P. W. Clark, S. F. Dyke, J. Organomet. Chem. 1985, 281,
389Ϫ396.
J. Chem. Soc., Dalton Trans. 1997, 2495Ϫ2499.
[12]
A. Avshu, R. D. OЈSullivan, A. W. Parkins, J. Chem. Soc., Dal-
ton Trans. 1983, 1619Ϫ1624.
[13]
J. Vicente, I. Saura-Liamas, P. G. Jones, J. Chem. Soc., Dalton
J. Albert, J. M. Cadena, J. Granell, Tetrahedron: Asymmetry
1997, 8, 991Ϫ994.
Trans. 1993, 3619Ϫ3624.
[14]
B. N. Cockburn, D. V. Howe, T. Keating, B. F. G. Johnson, J.
N. W. Alcock, J. M. Brown, D. I. Hulmes, Tetrahedron: Asym-
metry 1993, 4, 743Ϫ756.
Lewis, J. Chem. Soc., Dalton Trans. 1973, 404Ϫ410.
[15]
V. V. Dunina, O. A Zalevskaya, V. M. Potapov, J. Gen. Chem.
N. W. Alcock, D. I. Hulmes, J. M. Brown, J. Chem. Soc., Chem.
Commun. 1995, 395Ϫ397.
USSR (Engl. Transl.) 1984, 54, 389Ϫ397.
[16]
V. V. Dunina, O. A Zalevskaya, I. P. Smolyakova, V. M. Pota-
R. Annunziata, S. Cenini, F. Demartin, G. Palmisano, S. Tol-
lari, J. Organomet. Chem. 1995, 496, C1ϪC3.
V. V. Dunina, L. G. KuzЈmina, A. G. Parfyonov, Yu. K. Gri-
shin, Tetrahedron: Asymmetry 1998, 9, 1917Ϫ1921.
K. Tani, H. Tashiro, M. Yoshida, Ts. Yamagata, J. Organomet.
Chem. 1994, 469, 229Ϫ236.
pov, J. Gen. Chem. USSR (Engl. Transl.) 1986, 56, 674Ϫ684.
[17]
L. G. KuzЈmina, O. Yu. Burtseva, M. A. Porai-Koshitz, V. V.
Dunina, O. A. Zalevskaya, V. M. Potapov, J. Gen. Chem. USSR
(Engl. Transl.) 1989, 59, 2525Ϫ2534.
[18]
V. V. Dunina, N. S. Gulyukina, E. B. GolovanЈ, I. Yu. Nali-
mova, A. A. Koksharova, I. P. Beletskaya, Organomet. Chem.
USSR (Engl. Transl.) 1993, 6, 36Ϫ46.
P. W. Clark, S. F. Dyke, J. Organomet. Chem. 1984, 276,
421Ϫ430.
[19]
V. V. Dunina, N. S. Gulyukina, I. V. Byakova, Yu. F. Opru-
1038
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039

Steric Promotion of Aromatic CϪH Bond Activation in Primary Benzylamines
FULL PAPER
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[71]
[72]
[73]
[74]
S. Y. M. Chooi, M. K. Tan, P.-H. Leung, K. F. Mok, Inorg.
Chem. 1994, 33, 3096Ϫ3103.
S. Y. M. Chooi, T. S. A. Hor, P.-H. Leung, K. F. Mok, Inorg.
Chem. 1992, 31, 1494Ϫ1500.
P.-H. Leung, A. C. Willis, S. B. Wild, Inorg. Chem. 1992, 31,
1406Ϫ1410.
M. Pabel, A. C. Willis, S. B. Wild, Inorg. Chem. 1996, 35,
1244Ϫ1249.
S. Y. M. Chooi, S.-Y. Siah, P.-H. Leung, K. F. Mok, Inorg.
Chem. 1993, 32, 4812Ϫ4818.
D. C. R. Hockless, P. A. Gugger, P.-H. Leung, R. C. Maya-
dunne, M. Pabel, S. B. Wild, Tetrahedron 1997, 53, 4083Ϫ4094.
H. Gunter, NMR Spectroscopy Ϫ Basic Principles, Concepts,
and Applications in Chemistry, John Wiley, Chichester, 1995, p.
115.
H. Jendralla, Ch. H. Li, E. Paulus, Tetrahedron: Asymmetry
1994, 5, 1297Ϫ1320.
W. Lesueur, E. Solari, C. Floriani, A. Chiesi-Villa, C. Rizzoli,
Inorg. Chem. 1997, 36, 3354Ϫ3362.
M.-C. Lagunas, R. A. Gossage, W. J. J. Smeets, A. L. Spek, G.
van Koten, Eur. J. Inorg. Chem. 1998, 163Ϫ168.
M. Brookhart, M. L. H. Green, J. Organomet. Chem. 1983,
250, 395Ϫ408.
T. C. Jones, A. J. Nielson, C. E. Rickard, Aust. J. Chem. 1984,
37, 2179Ϫ2192.
G. Chelucci, M. A. Cabras, A. Saba, A. Sechi, Tetrahedron:
Asymmetry 1996, 7, 1027Ϫ1032.
H. Doucet, J. M. Brown, Tetrahedron: Asymmetry 1997, 8,
3775Ϫ3784.
S. Gladiali, S. Pulacchini, D. Fabbri, M. Manassero, M. San-
soni, Tetrahedron: Asymmetry 1998, 9, 391Ϫ395.
V. V. Dunina, O. A. Zalevskaya, I. P. Smolyakova, V. M. Pota-
pov, L. G. KuzЈmina, Yu. T. Struchkov, L. N. Reshetova, J. Gen.
Chem. USSR (Engl. Transl.) 1986, 56, 1164Ϫ1175.
L. G. KuzЈmina, Yu. T. Struchkov, V. V. Dunina, O. A. Zalev-
skaya, V. M. Potapov, J. Gen. Chem. USSR (Engl. Transl.) 1987,
57, 599Ϫ609.
V. V. Dunina, L. G. KuzЈmina, M. Yu. Kazakova, Yu. K. Gri-
shin, Yu. A. Veits, E. I. Kazakova, Tetrahedron: Asymmetry
1997, 8, 2537Ϫ2545.
J. H. Ford, C. D. Thompson, C. S. Marvel, J. Am. Chem. Soc.
1935, 57, 2619Ϫ2623.
[75]
[76]
[77]
[65]
[66]
[67]
[68]
[69]
[70]
[78]
[79]
[80]
G. A. Tsatsas, Ann. Chim., Ser. 12 1946, 1, 342Ϫ394.
P. Billon, Ann. Chim., Ser. 10 1927, 7, 314Ϫ384.
J. T. Sharp, I. Gosney, A. G. Rowley, Practical Organic Chemis-
try Ϫ A student handbook of techniques, London Ϫ New York,
Chapman and Hall, 1989, chapter 4.2.2d.
[81]
[82]
[83]
L. M. Harwood, Aldrichim. Acta 1985, 18, 25.
G. M. Sheldrick, Acta Crystallogr. 1990, A46, 467.
G. M. Sheldrick, SHELXL-93 Ϫ Program for the Refinement
of Crystal Structures, University of Göttingen, Germany, 1993.
Received October 27, 1998
J.-M. Valk, T. D. W. Claridge, J. M. Brown, Tetrahedron: Asym-
metry 1995, 6, 2597Ϫ2610.
[I98367]
Eur. J. Inorg. Chem. 1999, 1029Ϫ1039
1039