Markovnikov versus anti-Markovnikov selectivity in the amination of terminal olefins coordinated to platinum(II)

Article abstract of DOI:10.1016/j.inoche.2006.01.014

The products of addition of secondary amines to platinum(II)-coordinated terminal olefins undergo a fast Markovnikov (M)-anti-Markovnikov (anti-M) isomerization process leading, to an equilibrium composition different from that obtained under kinetic control; in basic medium, the amination products undergo intramolecular ring closing reaction which is faster for the anti-M isomers, so decreasing the M/anti-M ratio in the final complexes.

Full text of DOI:10.1016/j.inoche.2006.01.014

Inorganic Chemistry Communications 9 (2006) 500–503
www.elsevier.com/locate/inoche
Markovnikov versus anti-Markovnikov selectivity in the
amination of terminal olefins coordinated to platinum(II)
*
Giuseppe Lorusso, Nicola G. Di Masi, Luciana Maresca , Concetta Pacifico, Giovanni Natile
`
Dipartimento Farmaco-Chimico, Universita degli Studi di Bari, Via E. Orabona 4, I-70125 Bari, Italy
Received 16 January 2006; accepted 23 January 2006
Available online 27 March 2006
Abstract
The products of addition of secondary amines to platinum(II)-coordinated terminal olefins undergo a fast Markovnikov (M)–anti-
Markovnikov (anti-M) isomerization process, leading to an equilibrium composition different from that obtained under kinetic control;
in basic medium, the amination products undergo intramolecular ring closing reaction which is faster for the anti-M isomers, so decreas-
ing the M/anti-M ratio in the final complexes.
Ó 2006 Elsevier B.V. All rights reserved.
Keywords: Amination; Platinum(II) coordinated olefins; Markovnikov/anti-Markovnikov
The addition of nucleophiles to olefins coordinated to
late transition metals is a key step in many organometallic
reactions [1]. In this context, particular attention has been
devoted to the amination of terminal olefins, the interest
being focused on catalytic pathways able to enhance the
regioselectivity of the nucleophilic attack (M versus anti-
M) [2,3].
Selectivity aspects, concerned with the formation of M
and anti-M addition products and subsequent ring closing
reactions with formation of azametallacyclobutane species,
have been widely investigated in the case of platinum(II)
[4–6]; however, differences between kinetically and thermo-
dynamically controlled compositions of the isomeric mix-
tures have never been reported or discussed.
We have investigated the reaction of the cationic com-
plexes [PtCl(g2-olefin)(tmeda)]+ (olefin = propene, 1a,
and styrene, 1b; tmeda = N,N,N0,N0-tetramethylethylen-
ediamine) [7] with dialkylamines (NHMe2, m, and NHEt2,
n) (Scheme 1) in CDCl3 and over the temperature range
243–303 K. The addition reaction was fast in the whole
range of temperatures (occurring within the mixing time
of the reagents) and showed a net kinetic preference for
the M isomer (M/anti-M ratio ꢀ80:20). With time, and
at a rate that increased as the temperature was raised [8],
the solution reached an equilibrium composition which
was in favour of the M isomer in the case of 2an (M/
anti-M ratio P95:5, Fig. 1) and in favour of the anti-M iso-
mer in the case of 2bm and 2bn (M/anti-M ratio ꢀ30:70
and 10:90, respectively; Fig. 2). In the examined systems,
the equilibration between the two forms (M and anti-M)
was a rapid process at temperatures above 263 K, as wit-
nessed by the fast equilibration of solutions having initially
the kinetically controlled composition.
Under basic conditions (addition of a slight excess of
aqueous KOH to a dichloromethane solution of the plati-
num substrate at ambient temperature), the deprotonated
amination products (2 in Scheme 1) undergo a further
transformation in which the aminic nitrogen displaces the
cis chloride forming azaplatinacyclobutane species (3 in
Scheme 1) [4a,4b]. We have found that the M/anti-M ratio
decreases in the conversion of the linear addition product
into the ring closed species. The linear addition product
2an, containing at equilibrium only 5% of anti-M isomer,
yields, upon ring closing reaction, compound 3an for which
the anti-M isomer raises to 25%. Similarly, the linear
*
Corresponding author. Tel.: +39 080 5442774; fax: +39 080 5442230.
E-mail address: maresca@farmchim.uniba.it (L. Maresca).
1387-7003/$ - see front matter Ó 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.inoche.2006.01.014

G. Lorusso et al. / Inorganic Chemistry Communications 9 (2006) 500–503
501
M
Pt
M
R' R'
N
N
Cl
N
N
N
base
-HCl
NHR'2
Pt
R
+NHR'2
-NHR'2
R
N
N
Cl
Pt
2xy
3xy
+NHR'2
-NHR'2
R' R'
R
1x
N
N
Cl
N
N
N
base
-HCl
Pt
NHR'2
Pt
R
R
anti-M
anti-M
x = a (R = CH3), b (R = C6H5), c (R = H); y = m (R' = CH3), n (R' = CH2CH3)
M = Markovnikov; anti-M = anti-Markovnikov
Scheme 1. Summary of the amine addition reaction leading to 2 and of the
following ring closing reaction leading to 3.
Fig. 2. Region of the NMR spectrum of compound 2bm showing the
resonance peaks of the phenyl protons of styrene for the anti-M (7.12,
7.19, and 7.58 ppm for p, m, and o protons, respectively; peaks labelled
with an asterisk) and M isomers (7.40 for o and p protons and 7.57 for m
protons). The signals of the anti-M isomer increase considerably as the
temperature is raised from 243 K (kinetically controlled composition,
spectrum A) to 293 K (thermodynamically controlled composition,
spectrum B).
addition product 2bm, for which at equilibrium the anti-M
isomer is already prevalent (70%), gives, after ring closing
process, compound 3bm for which the anti-M isomer is
the exclusive form. The behaviour of 2an and 2bm can be
explained by assuming a greater propensity of the linear
anti-M addition products to undergo ring closing reaction.
The first outcome of this investigation is a net kinetic
preference for the M addition product that does not appear
to depend very much upon the type of olefin substituent
(methyl or phenyl). A slipping of the olefin in the p-bonded
precursor, which brings the more substituted carbon atom
farther from the metal centre and more exposed to the sol-
vent, can account for such a kinetic preference [5].
In contrast, the thermodynamic stabilities of the M
and anti-M addition products appear to depend very
much upon the nature of the substituents. A methyl
group favours the M isomer, whereas a phenyl group
favours the anti-M form. This could be attributed to a
phenyl stabilization and a methyl destabilization of the
high electron charge localized on the metal-bound carbon
atom.
The new finding of a ready M/anti-M isomerization
appears to contradict a recent paper by Pryadun et al. [9]
reporting a very slow isomerization (months at room tem-
perature) for the transformation of an initially formed anti-
M addition product (cis-[PtCl2{g1–CH(Me)CH2NHEt2}
(PPh3)]) into the M isomer (cis-[PtCl2{g1–CH2CH(Me)N-
HEt2}(PPh3)]). The Pryadun data, however, are susceptible
to a different interpretation. The presence of a strong labi-
lizing ligand (PPh3) trans to the leaving chloride and of
excess diethylamine able to deprotonate the ammonium
group could have induced a fast ring closing reaction with
the consequence that the reported isomerization process
does not concern the linear ammonium ethanide species
(analogous to compounds 2 in Scheme 1) but the ring
closed product (analogous to compounds 3 in Scheme 1)
[10].
Fig. 1. Region of the NMR spectrum of compound 2an showing the
resonance peaks of the CH3 protons of propene and diethylamine for the
anti-M (doublet at 0.95 ppm for propene and triplets at 1.34 and 1.39 ppm
for diethylamine: peaks labelled with an asterisk) and M (doublet at
1.26 ppm for propene and triplets at 1.43 and 1.49 ppm for diethylamine)
isomers. The signals of the anti-M isomer practically disappear as the
temperature is raised from 243 K (kinetically controlled composition,
spectrum A) to 293 K (equilibrium composition, spectrum B; the multiplet
centred at 1.59 ppm belongs to one of the two Pt–CH2– protons).
In the presently investigated systems, the intramolecular
cyclization, via displacement of the cis chloride by the N

502
G. Lorusso et al. / Inorganic Chemistry Communications 9 (2006) 500–503
atom of the 2-aminoethanide chain, has been clearly estab-
lished (see also Supplementary Information). Generally,
either the presence of a labilizing ligand trans to the leaving
chloride [4], and/or of a significant steric bulk on the etha-
nide chain was required (such a steric bulk promoting the
ring closing reaction by Thorpe–Ingold effect [11]) [6].
However, notwithstanding the fact that our addition prod-
ucts lack a labilizing ligand trans to the leaving chloride, a
ready ring closing reaction is observed with propene and
styrene. Furthermore, we could prove that it is possible
to completely remove the substituents from the ethanide
chain and still have the ring closing reaction.
Hence, the ethene derivative [PtCl(g1–CH2CH2
NHEt2)(tmeda)](ClO4), 2cn, was treated with a slight
excess of NHEt2. Under such weakly basic conditions, par-
tial conversion of the linear compound 2cn into the cyclic
compound 3cn took place (Figure S1, Tables S1 and S2).
Therefore, we can conclude that in the systems here consid-
ered neither the trans labilization of the leaving chloride
nor the presence of substituents on the ethanide chain is
necessary for the ring closing reaction to take place. The
main difference between our substrates and those previ-
ously investigated [4,6] is the absence of a formal negative
charge on the aminoethanide compound (deprotonated
compound 2) undergoing cyclization. A greater electrophi-
licity of the metal centre could therefore be an important
aiding factor for the ready occurrence of the ring-closing
process.
Fig. 3. ORTEP drawing of the complex cation [Pt{CH(Ph)CH2NEt2–
+
jC,jN}(tmeda)] (3bn), the ellipsoids enclose 20% probability. Selected
˚
bond distances [A] and angles [°]: Pt–N1 2.084(3), Pt–N2 2.209(3), Pt–N3
2.085(3), Pt–C8 2.053(4), C8–C7 1.522(6), C7–N3 1.516(5), C8–Pt–N1
100.3(1), N1–Pt–N2 83.3(1), N2–Pt–N3 106.0(1), C8–Pt–N3 70.4(1), Pt–
N3–C7 92.0(2), N3–C7–C8 103.5(3), C7–C8–Pt 93.1(2), N1–Pt–N3
170.5(1), C8–Pt–N2 176.4(1).
[C10–C9–C14 = 117.3(4)°] [20,21] and the lengthening of
˚
the trans Pt–N2 bond, which is 0.12 A longer than the
other two Pt–N bonds (the latter Pt–N distances are prac-
tically coincident, no matter if one belongs to a four-mem-
ber and the other to a five-member chelate ring). The
stabilization of such an electron charge by the phenyl ring
has been invoked to explain the thermodynamic preference
for the anti-M addition product in the styrene derivatives.
The steric interaction between substituents on cis ligands
Another outcome of the present investigation is the
greater propensity for the linear anti-M isomer, as com-
pared to the M one, to undergo ring closing reaction.
The anti-M addition product differs from the M isomer
for having the alkyl substituent (methyl or phenyl in the
present investigation) and the amine group on different car-
bons of the ethanide chain. Steric repulsion between the
two groups would set them in an antiperiplanar conforma-
tion. As a consequence, the amine group should be brought
in a synclinal conformation with respect to platinum and
therefore in a suitable position to give chloride substitution
and ring closing reaction. This cooperative effect cannot be
present in the M isomer having the alkyl substituent and
the amine group on the same carbon atom of the ethanide
chain.
[particularly
C9ꢁ ꢁ ꢁC3 = 3.426(8)
and
C6ꢁ ꢁ ꢁC15 =
˚
3.548(8) A] can fully explain the pyramidal distortion of
the platinum coordination centre.
The cyclic species (3) of this work are closely related to
the metallacycles recently proposed by Hartwig for the rho-
dium-catalyzed anti-M amination of vinylarenes [3]. Our
findings prove that the formation of such intermediates is
very likely, and the greater propensity of the anti-M isomer
to undergo ring closing reaction may be the reason for the
anti-M selectivity of the Hartwig reaction.
In conclusion, the present investigation has demon-
strated the following aspects: (i) the kinetic preference for
the Markovnikov isomer in the amination of platinum-
coordinated olefins; (ii) the ready isomerization between
Markovnikov and anti-Markovnikov addition products
taking place at room temperature; (iii) the ease of the ring
closing reaction in systems that lack a formal negative
charge on the metal core; (iv) the faster ring closure of
the linear anti-Markovnikov amination product (as com-
pared to the Markovnikov one) contributing to the
increase of this form in the ring closed products.
The X-ray structure of the species 3bn [12] is reported in
Fig. 3 [13–19]. The cyclobutane metallacycle is rather
strained and ‘‘quasi planar’’; as a consequence, the substit-
uents on adjacent atoms of the aminoethanide chain are
placed in eclipsed positions. Notwithstanding the large
deviation from theoretical values of bond angles, the cyc-
lobutane-metallacycle bond distances are rather normal
and in accord with values found in the analogous
[Pt(CH2CH2NMe2–jC,jN)Cl(PPh3)] complex in which,
however, the ethanide chain had no substituents and the
Acknowledgements
˚
Pt–N bond was ca. 0.04 A longer because of the phosphine
The authors are grateful to the University of Bari (Con-
tribution ex 60%). Dr. Francesco Cannito is gratefully
acknowledged for his assistance in the preparation of the
manuscript.
trans influence [4b]. The accumulation of electron charge
on the platinum-bound carbon atom (C8) is revealed by
the narrowing of the corresponding angle in the phenyl ring

G. Lorusso et al. / Inorganic Chemistry Communications 9 (2006) 500–503
503
[9] R. Pryadun, D. Sukumaran, R. Bogadi, J.D. Atwood, J. Am. Chem.
Soc. 126 (2004) 12414–12420.
Appendix A. Supplementary data
[10] Evidences in favour of a ring closed species rather than a linear
addition product for the complex formulated as [Pt{CH2CH(Me)N-
HEt2}Cl2(PPh3)] in Ref. [9] are found in the reported NMR
parameters for the metallaethanide chain. The values are practically
coincident with those reported for the ring closed compound
[Pt{CH2CH(Me)NMe2–jC,jN}Cl(PPh3)] (Ref. [4a,4b]) and very
close to those of the M and anti-M isomers of our complex 3an.
Also the reported high value of 3J PtH (up to 70 Hz) is typical of ring-
closed species rather than of linear compounds (ca. 40 Hz).
[11] (a) R.M. Beesley, C.K. Ingold, J.F.J. Thorpe, J. Chem. Soc. 107
(1915) 1080–1106;
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.inoche.
2006.01.014.
References
[1] (a) J. Haggin, Chem. Eng. News 71 (1993) 23;
(b) F.A. Cotton, G. Wilkinson, in: Advanced Inorganic Chemistry,
sixth ed., Wiley, New York, 1999;
(c) M. Beller, J. Seayed, A. Tillack, H. Jiao, Angew. Chem., Int. Ed.
45 (2004) 3368–3398.
[2] (a) T.E. Muller, M. Beller, Chem. Rev. 98 (1998) 675–703;
¨
(b) C.K. Ingold, J. Chem. Soc. 119 (1921) 305;
(c) S.W. Smith, M.S. Newman, J. Am. Chem. Soc. 90 (1968) 1253–
1257.
(b) H.M. Senn, P.E. Blo¨chl, A. Togni, J. Am. Chem. Soc. 122 (2000)
4098–4410;
(c) U. Nettekoven, J.F. Hartwig, J. Am. Chem. Soc. 124 (2002) 1166–
1167;
(d) M. Kawatsura, J.F. Hartwig, J. Am. Chem. Soc. 122 (2000) 9546–
9547;
(e) M. Utsonomiya, J.F. Hartwig, J. Am. Chem. Soc. 125 (2003)
14286–14287.
[12] Crystal data for [Pt{(CH3)2NCH2CH2N(CH3)2}{CHPhCH2N-
(CH2CH3)2-jC,jN}]ClO4: C18H34N3O4ClPt, Mw = 587.027, white
˚
parallelepiped, monoclinic, space group P21/c, a = 10.1870(5) A,
˚
˚
b = 15.6315(7) A,
c = 14.1576(7) A,
and
b =
95.348(1)°,
3
V = 2244.6(2) A , Z = 4,
q
calcd = 1.737 g cmꢂ3
, BrukerAXS X8
˚
APEX CCD system equipped with a four-circle Kappa goniometer
˚
and
a 4K CCD detector, radiation Mo Ka, k = 0.71073 A,
2hmax = 80.00°, temperature of measurement 293 2 K, no. of
measured reflections 25,934, no. of independent reflections 12,990,
no. of reflections included in the refinement 8248 (I > 2r(I)), data
reduction and unit cell refinement (package SAINT-IRIX) (Ref. [13])
integrated and corrected for Lorentz, polarization, and absorption
[3] (a) M. Utsonomiya, R. Kuwano, M. Kawatsura, J.F. Hartwig, J. Am.
Chem. Soc. 125 (2003) 5608–5609;
(b) M. Utsonomiya, J.F. Hartwig, J. Am. Chem. Soc. 126 (2004)
2702–2703;
(c) J. Takaya, J.F. Hartwig, J. Am. Chem. Soc. 127 (2005) 5756–
5757.
effects using the program SADABS (Ref. [14]), l = 6.397 mmꢂ1
,
structure solution by direct methods and Fourier methods, SIR2002
(Ref. [15]) and SHELXL97 (Ref. [16]), no. of parameters 380, H
atoms were located by Fourier difference and refined with isotropic
displacement, R1 = 0.0471 and wR2 = 0.0914 for observed reflections,
R1 = 0.0888 and wR2 = 0.1080 for all reflections, refinement residual
[4] (a) A. De Renzi, A. Panunzi, A. Vitagliano, Gazz. Chim. Ital. 108
(1978) 45–50;
(b) A. De Renzi, B. Di Biasio, G. Morelli, A. Vitagliano, Inorg. Chim.
Acta 63 (1982) 233–241;
(c) A. Panunzi, A. De Renzi, R. Palumbo, G. Paiaro, J. Am. Chem.
Soc. 91 (1969) 3879–3883;
(d) E. Benedetti, A. De Renzi, G. Paiaro, A. Panunzi, C. Pedone,
Gazz. Chim. Ital. 102 (1972) 744–754;
(e) A. De Renzi, G. Paiaro, A. Panunzi, V. Romano, Chim. Ind.
(Milan) 55 (1973) 248–251.
electron density 3.432 e Aꢂ3, calculations and molecular graphics
˚
were carried out using PARST97 (Ref. [17]), WinGX (Ref. [18]) and
ORTEP-3 for Windows (Ref. [19]) packages. CCDC – 269917
contains the supplementary crystallographic data for this paper.
These data can be obtained on line free of charge from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB 2 1 EZ
UK; fax +44 1223 336 033; or deposit@ccdc.cam.ac.uk.
[13] Bruker, SAINT-IRIX. Bruker AXS Inc., Madison, Wisconsin, USA,
2003.
[14] G.M. Sheldrick, SADABS, Program for Empirical Absorption
Correction of Area Detector Data, University of Go¨ttingen, Ger-
many, 1996.
[15] M.C. Burla, M. Cavalli, B. Carrozzini, G.L. Cascarano, C. Giaco-
vazzo, G. Polidori, R. Spagna, Sir2002: the program, J. Appl. Cryst.
36 (2003) 1103.
[5] (a) O. Eisenstein, R. Hoffmann, J. Am. Chem. Soc. 103 (1981) 4308–
4320;
(b) C. Hahn, Chem. Eur. J. 10 (2004) 5888–5899.
[6] (a) M. Green, J.K.K. Sahran, I.M. Al-Najjar, Organometallics 3
(1984) 520–524;
(b) J.K.K. Sahran, M. Green, I.M. Al-Najjar, J. Chem. Soc., Dalton
Trans. 5 (1984) 771–777;
(c) M. Green, J.K.K. Sahran, I.M. Al-Najjar, J. Chem. Soc., Dalton
Trans. 7 (1981) 1565–1571.
[7] (a) G. Gervasio, S.A. Mason, L. Maresca, G. Natile, Inorg. Chem. 27
(1986) 2207–2211;
[16] G.M. Sheldrick, Programs for Crystal Structure Analysis (Release 97-
2). Institut fur Anorganische Chemie der Universita¨t, Tammanstrasse
¨
¨
(b) L. Maresca, G. Natile, Comments Inorg Chem. 16 (1994) 95–112,
and quoted references;
(c) F.P. Fanizzi, L. Maresca, G. Natile, C. Pacifico, Gazz. Chim. Ital.
124 (1994) 184;
4, D-3400 Go¨ttingen, Germany, 1998.
[17] (a) M. Nardelli, Comput. Chem. 7 (1983) 95–97;
(b) M. Nardelli, J. Appl. Crystallogr. 28 (1995) 659.
[18] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837–838.
[19] L.J. Farrugia, J. Appl. Crystallogr. 30 (1997) 565.
[20] A. Domenicano, P. Mazzeo, A. Vaciago, Tetrahedron Lett. 13 (1976)
1029–1032.
(d) G. Bandoli, A. Dolmella, F.P. Fanizzi, N.G. Di Masi, L. Maresca,
G. Natile, Organometallics 21 (2002) 4595–4603;
(e) G. Lorusso, G. Boccaletti, N.G. Di Masi, F.P. Fanizzi, L.
Maresca, G. Natile, Eur. J. Inorg. Chem. 24 (2004) 4751–4754.
[8] In our experimental conditions (chloroform solution, mM concen-
trations) the equilibrium was reached in 4 h at 263 K and in almost
the mixing time of the reagents at 298 K.
[21] (a) A.R. Campanelli, A. Domenicano, F. Ramondo, J. Phys. Chem.
A 107 (2003) 6429–6440;
(b) A.R. Campanelli, A. Domenicano, F. Ramondo, I. Hargittai, J.
Phys. Chem. A 108 (2004) 4940–4948.