Highly active and selective platinum(II)-catalyzed isomerization of allylbenzenes: Efficient access to (E)-anethole and other fragrances via unusual agostic intermediates

Article abstract of DOI:10.1021/om100033a

Terminal alkene isomerization reactions can be efficiently catalyzed by Pt^II complexes bearing a chelating diphosphine and an alkyl or, better, aryl moiety under mild experimental conditions. In particular diphosphines, such as dppb, characterized by a large bite angle in conjunction with a pentafiuorophenyl residue coordinated to Pt enable quantitative conversion of the reagent into internal, alkenes within few hours at 50 °C in CHCl₃ as solvent. E/Z selectivity can be as high as 98:2 for allylbenzene, and the catalytic system can be fruitfully applied to the preparation of E fragrances derived by isomerization of substituted, allylbenzene derivatives. The selectivity increases during the progress of the reaction because of a subsequent catalytic step where the Z alkene coordinates to the Pt and is converted into the E isomer. NMR investigation on the catalyst showed formation of agostic Pt...H intermediate species derived by insertion of the substrate into the Pt-aryl bond followed by β-hydride elimination. Formation of such agostic species is promoted by the steric hindrance imparted by the diphosphine characterized by a large bite angle. Kinetic studies and DFT calculations on the possible agostic intermediates shed light on their structure and enable the formulation of a possible catalytic mechanism.

Full text of DOI:10.1021/om100033a

Organometallics 2010, 29, 1487–1497 1487
DOI: 10.1021/om100033a
Highly Active and Selective Platinum(II)-Catalyzed Isomerization of
Allylbenzenes: Efficient Access to (E)-Anethole and Other Fragrances via
Unusual Agostic Intermediates
Alessandro Scarso,*,† Marco Colladon,† Paolo Sgarbossa,‡ Claudio Santo,†
Rino A. Michelin,‡ and Giorgio Strukul*,†
†
ꢀ
Dipartimento di Chimica, Universita Ca’ Foscari di Venezia, Dorsoduro 2137, 30123 Venezia, Italy, and
Dipartimento di Processi Chimici dell’Ingegneria, Universita di Padova, Via Marzolo 9, 35131 Padova, Italy
‡
ꢀ
Received January 12, 2010
Terminal alkene isomerization reactions can be efficiently catalyzed by PtII complexes bearing a
chelating diphosphine and an alkyl or, better, aryl moiety under mild experimental conditions. In
particular diphosphines, such as dppb, characterized by a large bite angle in conjunction with a
pentafluorophenyl residue coordinated to Pt enable quantitative conversion of the reagent into
internal alkenes within few hours at 50 °C in CHCl3 as solvent. E/Z selectivity can be as high as 98:2
for allylbenzene, and the catalytic system can be fruitfully applied to the preparation of E fragrances
derived by isomerization of substituted allylbenzene derivatives. The selectivity increases during the
progress of the reaction because of a subsequent catalytic step where the Z alkene coordinates to the
Pt and is converted into the E isomer. NMR investigation on the catalyst showed formation of agostic
Pt H intermediate species derived by insertion of the substrate into the Pt-aryl bond followed by
3 3 3
β-hydride elimination. Formation of such agostic species is promoted by the steric hindrance
imparted by the diphosphine characterized by a large bite angle. Kinetic studies and DFT calcu-
lations on the possible agostic intermediates shed light on their structure and enable the formulation
of a possible catalytic mechanism.
Introduction
synthesis mediated by chiral Rh catalyst developed by
Noyori,7 and (iii) the industrial isomerizations of allyl ben-
zene derivatives such as estragole, eugenol, and safrole to the
corresponding internal alkenes used as fragrances, where
high productivity as well as high selectivity for the E isomer
are important challenges.2 In the latter process, E products
are marketed, while Z isomers are characterized by an
unpleasant taste and in some cases a toxic character, requir-
ing a lower than 1% content for human use. At present,
tedious purification steps are required to isolate the E isomer,
thus causing an increasing interest in the development of
highly selective isomerization processes in order to decrease
production costs as well as avoid the stoichiometric use of
bases. Currently, industrial isomerization of estragole to
E-anethole is performed with KOH at high temperatures
(200 °C) with moderate yields of about 56% and E/Z selec-
tivities of 82:18, while most fragrances reported above are
still obtained by extractions from natural sources with an
overall annual production of several million metric tons.8
The catalysts developed for terminal alkene isomerization
are numerous. Heterogeneous ones are preferred for indus-
trial applications and rely on solid bases whose activity is
related to the ability to deprotonate the substrate, thus
providing a carbanion isomerizing into thermodynamically
Alkene isomerization is an important atom-efficient reac-
tion fulfilling both sustainability criteria and widespread
applications in industry.1 In fact, it is mainly used for the
synthesis of commodities such as linear olefins in the SHOP
process or adiponitrile preparation from butadiene and
HCN, where two isomerization steps are involved, or in
vinylnorbornene isomerization to provide ethylidenenorbor-
nene, a useful comonomer for elastomer production.2
This reaction is also largely employed in the synthesis of
pharmaceuticals and fine chemicals such as fragrances.3-5
Noteworthy examples are (i) 6-methyl-6-hepten-2-one iso-
merization in BASF vitamin A synthesis,6 (ii) the asymmetric
allylamine isomerization from myrcene in the (-)-menthol
*Corresponding authors. E-mail: alesca@unive.it; strukul@unive.it.
Fax: þ39-0412348517.
(1) van Leeuwen P. W. N. M. Homogeneous Catalysis; Kluwer Aca-
demic Publishers: Dordrecht, 2004; pp 101-105.
(2) Parshall G. W.; Ittel S. D. Homogeneous Catalysis, 2nd ed.; Wiley:
New York, 1992; pp 9-24.
(3) Cornils B.; Herrmann W. A. Applied Homogeneous Catalysis with
Organometallic Compounds; Wiley: New York, 2002; pp 980-991.
(4) (a) Bauer K.; Garbe D.; Surburg H. Common Fragrance and
Flavour Materials, 4th ed.; Wiley: New York, 2001; pp 171-226. (b) Kraft
P.; Swift K. A. D. Current Topics in Flavour and Frangrance Research;
Swift, K. A. D., Ed.; VHCA, Z€urich, Switzerland, and Wiley-VCH Veralg
GmbH, Weinheim, Germany, 2008.
(7) Inoue, S.-I.; Takaya, H.; Tani, K.; Otsuka, S.; Sato, T.; Noyori, R.
J. Am. Chem. Soc. 1990, 112, 4897–4905.
(8) Sharma, S. K.; Srivastava, V. K.; Jasra, R. V. J. Mol. Catal. A
(5) Chapuis, C.; Jacoby, D. Appl. Catal. A 2001, 221, 93–117.
(6) Pommer, H.; Nurrenbach, A. Pure Appl. Chem. 1975, 43, 527–552.
2006, 245, 200–209.
r
2010 American Chemical Society
Published on Web 02/17/2010
pubs.acs.org/Organometallics

1488 Organometallics, Vol. 29, No. 6, 2010
Scheme 1. Terminal Alkene Isomerization Reaction via (A) 1,2-Hydrogen Shift and (B) 1,3-Hydrogen Shift
Scarso et al.
more stable internal alkenes.9 Such solid catalysts are usually
active at high temperatures (>200 °C) and are critically affec-
ted by the presence of water. A different approach applies to
homogeneous isomerization catalysts, generally transition
metal complexes operating via two possible main pathways
(Scheme 1):2 (A) hydride addition elimination mechanism
(1,2-hydrogen shift) and (B) π-allyl mechanism (1,3-hydro-
gen shift). By virtue of the π-allyl intermediate, mechanism B
may have a dramatic effect on the E/Z ratio, providing high
amounts of the E isomer, which is favored both kinetically
and thermodynamically.10
experimental conditions, mediated by the electron-poor
cationic PtII complexes shown in Chart 1, together with
some mechanistic investigations and DFT calculations that
evidenced for the first time the involvement of agostic species
as key intermediates in the isomerization process.
Results and Discussion
The Catalysts. Most of the complexes tested as catalysts in
the isomerization reaction have already been reported else-
where (see Experimental Section). 1, 2, 3, 5, and 6 are new
complexes, and they have been synthesized in good yields by
halide abstraction from the corresponding Pt-Cl derivatives
according to the procedure reported in the Experimental
Section. As shown in Chart 1, all catalysts are cationic PtII
(or PdII in one case) soft Lewis acids containing a dipho-
sphine (or two trans monophosphines in 5) in which the
P-Pt-P bite angle was varied by changing the length of the
aliphatic chain linking the two phosphorus atoms, along
with an alkyl or aryl ligand (substituted or unsubstituted), to
modify both the electronic and steric properties of the com-
plex and finally a labile and easily displaceable water molecule.
Isomerization of Allylbenzene. The activities of the cata-
lysts in the selective isomerization of allyl-benzene to the
corresponding E and Z isomers (Scheme 2), chosen as a
model substrate, are reported in Table 1. It should be
emphasized that the isomerization of this substrate as well
as the other ones reported in this work can be carried out in
air with no need for purified, dry solvents. Complex 1
showed interesting activity toward allyl benzene with almost
complete conversion after 24 h with an E/Z ratio of 87:13,
which remained almost constant during the reaction pro-
gress. Substitution of the methyl residue with a more elec-
tron-withdrawing trifluoromethyl group like in 2 caused a
substantial increase in activity and, more interestingly, a
marked increase in selectivity for the E isomer, with the E/Z
value growing to 95:5 during the reaction progress.
Metal centers capable of catalyzing alkene isomerization
span from Ti2 and Fe2 to softer Lewis acidic ones such as
Rh,11 Ru,8,12 Ir,13 Pd,14,15,12b and Ni.16 The latter class, albeit
more expensive, is generally more productive.
During our endeavor on PtII oxidation reactions with
H2O2 using electron-poor complexes as catalysts17 we obser-
ved partial alkene isomerization during epoxidation that
prompted us to investigate more deeply the reaction. Herein
we report the results observed in the selective isomerization
of alkenes for the achievement of fragrances under mild
(9) (a) Kishore, D.; Kannan, S. Green Chem. 2002, 4, 607–610.
(b) Srivastava, V. K.; Bajaj, H. C.; Jasra, R. V. Catal. Commun. 2003, 4,
543–548. (c) Kishore, D.; Kannan, S. J. Mol. Catal. A 2004, 223, 225–230.
(d) Kishore, D.; Kannan, S. App. Catal. A 2004, 270, 227–235. (e) Kishore,
D.; Kannan, S. J. Mol. Catal. A 2006, 244, 83–92.
(10) Tuner, M.; Jouanne, J. V.; Brauer, H.-D.; Kelm, H. J. Mol.
Catal. 1979, 5 (425), 433–447.
(11) (a) Morrill, T. C.; D’Souza, C. A. Organometallics 2003, 22,
1626–1629. (b) Yan, Y.; Zhang, X.; Zhang, X. J. Am. Chem. Soc. 2006, 128,
16058–16061. (c) Tan, K. L.; Bergman, R. G.; Ellman, J. J. A. Am. Chem.
Soc. 2002, 124, 13964–13965.
(12) (a) Schmidt, B. Chem. Commun. 2004, 742–743. (b) Sharma, S. K.;
Srivastava, V. K.; Pandya, P. H.; Jasra, R. V. Catal. Commun. 2005, 6, 205–
209. (c) Hanessian, S.; Giroux, S.; Larsson, A. Org. Lett. 2006, 8, 5481–
5484. (d) Courchay, F. C.; Sworen, J. C.; Ghiviriga, I.; Abboud, K. A.;
Wagener, K. B. Organometallics 2006, 25, 6074–6086. (e) Yue, C. J.; Liu, Y.;
He, R. J. Mol. Catal. A 2006, 259, 17–22. (f) Krompiec, S.; Kuꢁznik, N.;
Urbala, M.; Rzepa, J. J. Mol. Catal. A 2006, 248, 198–209. (g) Donohoe,
T. J.; O'Riordan, T. J. C.; Rosa, C. P. Angew. Chem. 2009, 121, 1032–1035.
Angew. Chem., Int. Ed., 2009, 48, 1014-1017; (h) Erdogan, G.; Grotjahn,
D. B. J. Am. Chem. Soc. 2009, 131, 10354–10355.
The increase of the E/Z value during the reaction progress
is observed also with catalysts 4b and 4bPd and might be
indicative of either an evolution of the catalytically active
species or the existence of a consecutive reaction. The pre-
sence of an aryl residue on the complex as in 3 induces a
further enhancement of catalytic activity with complete
conversion in only 5 h and an E/Z value of 95:5. Again,
replacement of the aryl residue with a more electron-with-
drawing pentafluorophenyl residue as in 4e led to an even
better catalyst with complete conversion in only 2 h and
E/Z selectivity of 97:3.
(13) Baxendale, I. R.; Lee, A.-L.; Ley, S. V. J. Chem. Soc., Perkin
Trans. 1 2002, 1850–1857.
(14) Jennerjahn, R.; Piras, I.; Jackstell, R.; Franke, R.; Wiese, K.-D.;
Beller, M. Chem.;Eur. J. 2009, 15, 6383–6388.
(15) Kim, I. S.; Dong, G. R.; Jung, Y. H. J. Org. Chem. 2007, 72,
5424–5426.
(16) Zhang, J.; Gao, H.; Ke, Z.; Bao, F.; Zhu, F.; Wu, Q. J. Mol.
Catal. A 2005, 231, 27–34.
(17) (a) Colladon, M.; Scarso, A.; Sgarbossa, P.; Michelin, R. A.;
Strukul, G. J. Am. Chem. Soc. 2006, 128, 14006–14007. (b) Colladon, M.;
Scarso, A.; Sgarbossa, P.; Michelin, R. A.; Strukul, G. J. Am. Chem. Soc.
2007, 129, 7680–7689. (c) Pizzo, E.; Sgarbossa, P.; Scarso, A.; Michelin,
R. A.; Strukul, G. Organometallics 2006, 25, 3056–3062.

Article
Organometallics, Vol. 29, No. 6, 2010 1489
Chart 1. PtII Complexes Investigated As Catalysts for Alkene Isomerization
Scheme 2. Allyl Benzene Isomerization to internal E and Z
Alkenes Mediated by PtII Catalysts
Table 1. Catalytic Isomerization of Allyl Benzene with PtII
Catalysts: Catalyst Screeninga
catalyst
time (h) yield (%) E:Z (%)
none
16
1
3
24
2
4
24
1
3
5
24
2
24
2
24
2
24
2
4
0
8
1, [(dppb)Pt(CH3)(H2O)]OTf
2, [(dppb)Pt(CF3)(H2O)] OTf
3, [(dppb)Pt(C6H5)(H2O)] OTf
90:10
87:13
86:14
75:15
81:19
95:5
94:6
95:5
95:5
89:11
62:38
93:7
80:20
80:20
92:8
92:8
97:3
83:17
94:6
96:4
20
96
11
41
99
19
85
99
22
30
99
14
91
16
99
99
6
Spurred by the interesting results observed with complex
4e, we decided to explore the effect of different ligands on the
activity and selectivity of the reaction. 4a showed low activity
and selectivity, while both properties increased with 4b,
characterized by a larger bite angle. Notably, the flexibility
of the diphosphine ligand seems to positively influence acti-
vity, as observed by comparison with complex 4c, a more
rigid analogue of 4b, the latter showing a higher catalytic
activity. An even larger bite angle (complex 4e) caused a
further increase in activity. While Lewis acidity seems a
common requirement with respect to epoxidation of term-
inal alkenes with the same class of complexes,17 an opposite
trend is observed with respect to the diphosphine bite angle,
as in epoxidation larger bite angles reduce activity because of
steric congestion around the vacant site arising from water
displacement. It is interesting to notice that while activity is
strongly influenced by the bite angle of the diphosphine,
much lower is the influence on stereoselectivity (see complex
4a vs 4d). The cis coordination of the bidentate ligands is
another important issue as observed by comparison with
complex 5, bearing two trans triphenyl phosphine ligands,
showing very low catalytic activity. Substitution with a more
bulky electron-withdrawing group as in 6 caused a dramatic
decrease in activity and also a very low selectivity (see 4b)
probably due to a difficult approach of the alkene to the
complex. The PdII complex 4b-Pd, structurally analogous to
PtII 4b, unexpectedly showed a lower activity. Dimeric
species 7, where upon dissociation only one coordination
site is available for alkene coordination, provided the iso-
merized products with very low activity but good selectivity.
4a, [(dppm)Pt(C6F5)(H2O)]OTf
4b, [(dppe)Pt(C6F5)(H2O)]Otf
4c, [(S,S-Chiraphos)Pt(C6F5)(H2O)]OTf
4d, [(dppp)Pt(C6F5)(H2O)]OTf
4e, [(dppb)Pt(C6F5)(H2O)]OTf
4b-Pd, [(dppe)Pd(C6F5)(H2O)]OTf
24
24
24
5
27
21
22
5
5, trans-[(PPh3)2Pt(C6F5)(H2O)OTf
6, [(dppe)Pt(C8H3F6)(H2O)]OTf
7, [(dppb)Pt(μ-OH)]2(BF4)2
51:49
95:5
95:5
24
25
a Experimental conditions: [sub]0 = 0.83 M, cat. 2% mol, solvent
chloroform 0.5 mL, T 50°C. Yield determined by GC analysis; dr and
product assignment determined by 1H NMR analysis.
With 4e as the best catalyst at hand, we investigated the
effect of catalyst loading and solvent on the model isomer-
ization reaction. In Table 2 are reported the results for
experiments performed with decreasing amounts of complex
4e. With a loading of <1% mol the reaction is complete in
4 h still with good selectivity, while with 0.2% activity is
obviously negatively affected, while selectivity at high
conversion does not appear to be significantly influenced.

1490 Organometallics, Vol. 29, No. 6, 2010
Scarso et al.
Table 2. Catalytic Isomerization of Allyl Benzene with
[(dppb)Pt(C6F5)(H2O)](OTf) 4e: Optimization of the
Experimental Conditionsa
conditions
time (h)
yield (%)
TON
E:Z (%)
sub/cat. 50:1, CHCl3
sub/cat. 108:1, CHCl3
2
2
4
4
52
4
24
4
24
4
99
69
99
10
95
50
99
34
91
73
99
50
75
107
50
475
54
107
37
97:3
84:16
95:5
83:17
92:8
80:20
98:2
61:39
67:33
72:28
82:18
sub/cat. 500:1, CHCl3
sub/cat. 108:1, DCE
sub/cat. 108:1, THF
sub/cat. 108:1, toluene
98
79
107
Figure 1. Allyl benzene isomerization with 4e (2% mol) at 50 °C,
[sub] = 0.83 M. (]) (1E)-prop-1-en-1-ylbenzene, (b) (1Z)-prop-
1-en-1-ylbenzene, (O) allylbenzene, ([) selectivity to (1E)-prop-
1-en-1-ylbenzene.
24
a Experimental conditions: [sub]0 =0.81 M, solvent 0.5 mL, T 50°C.
Yield determined by GC analysis; dr and product assignment deter-
mined by 1H NMR analysis.
Table 3. Substrate Scope in the Catalytic Isomerization of
Terminal Alkenes with 4e in CHCl3 at 50 °Ca
Scheme 3. Isomerization of 4-Methylpentene with 4e in
CHCl3 at Room Temperature
is characterized by better activities and comparable selecti-
vities to other well-studied homogeneous systems.12
As already evidenced by the above substrates, the catalyst
is sensitive to the presence of donor heteroatoms; in fact allyl
acetate, allyl imidazole, and allyl ethyl ether were unreactive
because of competitive coordination by the heteroatoms on
the PtII center. Only allyl phenyl ether was active due to the
intrinsically lower Lewis basic character of the oxygen atom
of the substrate, as observed for all the other methoxy aryl
substrates investigated, although it unexpectedly led to
complete stereoselection for the Z isomer18 (Table 3).
Isomerization of terminal acyclic alkenes is very fast with
4e even at room temperature, but the reaction does not stop
at the 2-alkene products, as double-bond migration occurs
until thermodynamic distribution19 is reached, leading to a
complex mixture of products as illustrated in Scheme 3.
Analogously isomerization of 2-allylcyclohexanone did not
show the formation of 2-alkenes; only the conjugated enone
was observed, indicative of sequential double-bond migra-
tion (entry 6, Table 3). These results clearly evidence that the
activity of PtII complexes toward alkenes20 is better for ter-
minal olefins, but it is good also for internal21 and sterically
hindered ones.
a Experimental conditions: [sub]0 = 0.81 M, 4e 1% mol, solvent
chloroform 0.5 mL, T 50°C. Yield determined by GC analysis; diaster-
eoselective ratio (E/Z) and product assignment determined by 1H NMR
analysis.
Reactivity is enhanced in aromatic solvents and decreased in
polar THF probably because of solvent competition for co-
ordination to PtII, while selectivity is positively influenced by
chlorinated solvents with an E/Z ratio up to 98:2 in dichloro-
ethane (DCE).
Scope of the Reaction. In order to investigate the scope of
the reaction, we tested catalyst 4e in CHCl3 with several
terminal alkenes (Table 3). In particular the class of allyl
benzene derivatives that are industrially important sub-
strates for the production of fragrances4 showed good per-
formance with activities higher than known homogeneous
and heterogeneous systems. The selectivity for estragole
(allyl-p-methoxybenzene, entry 2) as substrate is one of the
highest ever reported with good conversions under much
milder experimental conditions with respect to known ex-
amples.12f Similarly, isomerization of eugenol (entry 4),
3,4-dimethoxyallylbenzene (entry 3), and safrole (entry 5)
The reaction profiles in the isomerization of allylbenzene
with the complexes reported in Table 1 were analyzed, and an
example is reported in Figure 1 for complex 4e. As can be
seen, after a short induction time, both isomerized alkenes
start being produced with an initially constant selectivity.
Then, as the substrate is depleted, the Z product, after
(18) Brett Runge, M.; Mwangi, M. T.; Bowden, N. B. J. Organomet.
Chem. 2006, 691, 5278–5288.
(19) Grimme, S.; Steinmetz, M.; Korth, M. J. Org. Chem. 2007, 72,
2118–2126.
(20) Sivaramakrishna, A.; Su, H.; Moss, J. R. Organometallics 2007,
26, 5786–5790.
(21) Pt(II)-catalyzed 1-hexene isomerization as side reaction:
Karshtedt, D.; Bell, A. T.; Don Tilley, T. J. Am. Chem. Soc. 2005,
127, 12640–12646.

Article
Table 4. Energy Values for the Different Alkene Isomers
Calculated for Allylbenzene and Safrole
Organometallics, Vol. 29, No. 6, 2010 1491
the presence of any Pt-H species, which usually have reso-
nances at very high fields (δ<0 ppm). At the same time, the 31
P
resonances typical of 4e tend to decrease after a few catalytic
cycles (5 min, 2% yield in internal alkenes), and the formation
of a new species is observed, characterized by the presence of
two P resonances (18.3 ppm, 1JP-Pt 5266 Hz, and 14.9 ppm,
1JP-Pt 3156 Hz) both without Pt-F coupling constants
indicative of the loss of the pentafluoro phenyl residue
(Figure 2A). At the same time, GC-MS analysis showed the
presence of traces of pentafluoroaryl-alkenes, confirming the
cleavage of the Pt-C6F5 bond and coupling of the fluorinated
organic residue with the substrate. After 40 min only the new
species was present, which remained unchanged until reaction
reached completeness. With p-methoxy-allylbenzene a similar
behavior occurred with disappearance of 4e signals and for-
mation of the new species similar to the one observed with
allylbenzene (Figure 2B).
Spectroscopic analysis allowed describing such species as
unusually stable PtII-alkyl complexes with Pt H agostic
3 3 3
1
interactions,22 justified by the exceptionally high JP-Pt
coupling constant23 and forced by the steric hindrance
imparted in this case by the dppb ligand. Attempts to observe
1
agostic interactions on H NMR in the upfield region of
the spectrum at very low temperature (down to 168 K)22 were
unsuccessful probably because of overlap of such signals
with those belonging to the substrate and products of the
isomerization reactions present in much higher concentra-
tions. This weak interaction arises when sterically congested
ligands with large bite angles surround Pt-alkyl species,
bringing the H atom in close contact with the Pt metal center.
Interestingly, upon increasing benzene ring substitution as
in 3,4-dimethoxyallylbenzene, eugenol, and safrole as sub-
strates, two new species were present, both characterized by
reaching a maximum concentration, is converted into the
E isomer in a consecutive isomerization process. This beha-
vior is typical even if it can be more or less evident depend-
ing on the individual PtII catalyst and explains why in
Tables 1-3 high selectivities to the E isomer occur only at
high conversions.
1
one large (∼5000 Hz) and one small (∼3000 Hz) JP-Pt in
27:73, 25:75, and 45:55 ratios, respectively (Figure 2C, D,
and E). The chemical shifts and coupling constants observed
for such species are substrate sensitive and were found to be
fully in agreement with the formation of diphosphine plati-
Some DFT calculations on allylbenzene and safrole and
their respective isomers have been carried out in order to
evaluate the thermodynamic composition at the equilibrium
of the different alkenes. The energy values at three different
temperatures are reported in Table 4. As can be seen, a
moderate decrease in energy difference is observed with
temperatures between the terminal alkenes and the internal
ones, while the latter show little difference. At 50 °C, i.e., the
temperature of the catalytic experiments, and for both sub-
strates considered the E-2 isomers are about 6.6 kcal/mol
more stable with respect to the corresponding terminal
alkene and 3.2-3.4 kcal/mol more stable with respect to
the corresponding Z-2 isomers. This corresponds to an
approximate population of 0% terminal, 99% E-2, and
1% Z-2 alkene, indicating that in Figure 1 the initial selec-
tivity corresponds to a kinetic control on the terminal to
internal isomerization step, while the final selectivity app-
roaches the thermodynamic one.
In Situ NMR Studies. In order to shed light on the reaction
pathway and understand the nature of possible reaction
intermediates, we performed some NMR experiments where
1H and 31P spectra of the system were recorded during the
reaction progress under conditions identical to those repor-
ted in Table 1 for the catalytic experiments in order to
monitor product formation as well as catalyst speciation.
With complex 4e as the catalyst and allylbenzene as the
substrate the 1H NMR spectrum of the system did not show
num-alkyl complexes with agostic Pt H interaction.22
3 3 3
The rate of catalyst evolution during the reaction progress
and the presence of agostic intermediate species as resting
states are most likely induced by the steric hindrance im-
parted by the ligand. This also explains the reactivity differ-
ences observed for homologous complexes 4a-e (Table 1),
which differ in the bite angle of the diphosphine on going
from dppm to dppb.
Agostic Pt H species are detected also for other cata-
3 3 3
lysts bearing the dppb diphosphine ligand as in 1, 2, and 3.
An analogous 31P NMR investigation was performed on
catalysts 1, 2, and 3, and in all cases agostic species (Scheme
4) were present as minor species, while their formation was
slower compared to 4e depending on the alkyl (aryl) ligand
according to the order -CH3<-CF3<-C6H5<-C6F5, in
agreement with the effects on Lewis acidity17c and catalyst
activity already observed in Table 1. The reactivity order
observed for catalysts bearing different alkyl- and aryl-
coordinated moieties seems in general agreement with the
(22) (a) Carr, N.; Dunne, B. J.; Mole, L.; Orpen, A. G.; Spencer, J. L.
J. Chem. Soc., Dalton Trans. 1991, 863–871. (b) Carr, N.; Mole, L.; Orpen,
A. G.; Spencer, J. L. J. Chem. Soc., Dalton Trans. 1992, 2653–2662.
(c) Spencer, J. L.; Mhinzi, G. S. J. Chem. Soc., Dalton Trans. 1995, 3819–
3824.
(23) Mullen, C. A.; Gagne, M. R. J. Am. Chem. Soc. 2007, 129,
11880–11881.

1492 Organometallics, Vol. 29, No. 6, 2010
Scarso et al.
Scheme 4. Proposed Isomerization Pathway
2136 Hz), even though for 3 (1757 Hz) the Pt-PC-trans is
smaller probably due to enhanced back-bonding ability.
After reaction completion the remaining PtII species
2þ
observed for 1, 2, 3, and 4e was mainly [Pt(dppb)(μ-OH)]2
(δ 3.3 ppm, JP-Pt 3530 Hz), formed by interaction with
trace water, which is only sparingly active toward terminal
allyl benzene isomerization (25% of conversion with E/Z
95:5 after 24 h in the same experimental conditions as in
Table 1).
Possible Reaction Pathway. The reaction profile shown in
Figure 1 clearly indicates that in the isomerization of allyl-
benzene with 4e the decay of substrate is independent of its
concentration, allowing the determination of the reaction
rate very accurately. This behavior is general, although the
extent of the induction time, which in the case of 4e is very
short, can vary depending on catalyst nature and concentra-
tion. In order to gain more insight into the mechanism, the
effect of 4e concentration in the same reaction was deter-
mined. This is reported in Figure 3 and shows that the
reaction is first-order on initial catalyst concentration. No
reaction is observed in the absence of catalyst, and for
concentrations above 2 ꢀ 10-2 M the catalyst becomes
incompletely soluble in the reaction medium.
Figure 2. 31P spectra of agostic intermediate species arising from
4e and (A) allylbenzene, (B) p-methoxy-allylbenzene, (C) eugenole,
(D) 3,4-dimethoxy-allylbenzene, and (E) safrole.
strength of the Pt-C bond, as confirmed by the increas-
ing Pt-PC-trans coupling constants (1 1849 Hz, 2 1904 Hz, 4e

Article
Organometallics, Vol. 29, No. 6, 2010 1493
Figure 3. Effect of the concentration of 4e on the rate of iso-
merization of allylbenzene. Experimental conditions: [sub]0 =
0.83 M, solvent chloroform 0.5 mL, T 50 °C.
Figure 4. ORTEP drawing of the cationic portion of the com-
plex cis-[(dppb)Pt(C6F5)(H2O)] SO3CF3 H2O. Selected bond
These observations lead to the following rate expression,
evidencing the first-order dependence of the rate on initial
catalyst concentration and the zero-order dependence on
substrate concentration:
3
3
lengths (A) and angles (deg): Pt-P(1) 2.212(3), Pt-C(29)
2.059(9), Pt-P(2) 2.312(3), Pt-O(1) 2.139(6); P(1)-Pt-P(2)
93.8(1), C(29)-Pt-O(1) 87.2(3), C(29)-Pt-P(1) 90.9(3), O(1)-
Pt-P(2) 88.1(2), O(1)-Pt-P(1) 176.7(2), C(29)-Pt-P(2)
175.1(3).
rate ¼ k½Ptꢁ0
In Scheme 4 a possible reaction pathway is suggested
accounting for all the above-mentioned spectroscopic and
kinetic evidence. The catalyst coordinates the alkene, which
becomes electron poor and induces the migration of the C6F5
residue, leading to overall insertion of the alkene and for-
mation of Pt-alkyl species. The latter is then subjected to
β-hydride elimination with formation of diphosphine Pt-
hydride and release of a pentafluoro phenyl vinyl byproduct
(seen by GC-MS). These initial steps account for the forma-
tion of the actual catalyst and justify the induction time. The
same Pt-H species is obtained independently of the alkyl/
aryl ligand present on the metal, albeit with different rates
(see above). The produced hydride is coordinatively unsatu-
rated and promptly reacts with the alkene followed by rapid
hydride migration to the double bond and formation of a Pt-
alkyl species that is constrained by the large bite angle24 of
dppb, eventually showing the agostic interaction.25 Here we
can distinguish two possible agostic intermediates: one aris-
ing from hydride attack to the alkene C2 and formation of an
unproductive agostic intermediate A and one derived by
hydride attack to C1 leading to species B. The latter slowly
transforms into species C (the rate-determining step) by
dissociation of the weak agostic interaction followed by
rotation around the Pt-C bond and formation of a new
agostic contact with a hydrogen bound to C3. Species C
undergoes β-hydride elimination and formation of Pt-H
species and the isomerized alkene, which is quickly replaced
by the incoming terminal alkene that binds more strongly to
PtII complexes.26 Through ab initio and DFT calculations
applied to Pt(PR3)2(H)(propene)þ complexes (R = H, F,
CH3) Creve et al. demonstrated that insertion of the alkene in
Scheme 5. Proposed Isomerization Pathway for Z to E Alkene
Interconversion
(24) Dierkes, P.; van Leeuwen, P. W. N. M. J. Chem. Soc., Dalton
Trans. 1999, 1519–1529.
(25) Mole, L.; Spencer, J. L.; Carr, N.; Orpen, A. G. Organometallics
1991, 10, 49–52.
(26) (a) Crabtree R. H. The Organometallic Chemistry of the Transi-
tion Metals; John Wiley & Sons: New York, 1988. (b) Collman J. P.;
Hegedus L. S.; Norton J. T.; Finke R. G. Principles and Applications of
Organotransition Metal Chemistry; University Science Books: Mill Valley,
CA, 1987.
the Pt-H bond and β-elimination processes are characteri-
zed by small activation barriers, up to 2.4 kcal/mol. More-
over the same study underlined the preference for hydride
migration to C2, leading to linear alkyl complexes, rather

1494 Organometallics, Vol. 29, No. 6, 2010
Scheme 6. Energy Values for Minimized Agostic Species A-D Involved in Schemes 4 and 5
Scarso et al.
than attack on C1, which provides branched Pt-alkyl
species.27 This general view is consistent with the rate law
and the spectroscopic and reactivity data reported above and
with the essential features of the classical 1,2-hydrogen shift
mechanism shown in Scheme 1A.
On this basis, the slow step of the reaction is suggested to
be the transformation of species B into C, where the PtII
metal center releases the agostic interaction with C1-H and
after rotation around the Pt-alkyl bond it makes agostic
contact with C3-H. Experimental evidence based on NMR
dynamic experiments on a PdII complex bearing an isopropyl
observed experimentally in this work (Figure 1A) arises from
kinetic control of the transformation of B into the two
possible C species, one where the alkyl residue with agostic
interaction is pro-E and one pro-Z.
What pushes the selectivity toward the E isomer approach-
ing thermodynamic equilibrium in the case of complex 4e
and similar species is a consecutive isomerization of the (Z)-
2-alkene when the concentration of the starting substrate
approaches zero (Figure 1). Preferential coordination of
the (Z)-2-alkene is favored by the steric hindrance present
in the catalyst (Table 1) and helps to increases the selectivity
of the process toward the E isomer. As depicted in Scheme 5,
the Z alkene recoordinates to Pt-H species, favored by
reduced steric requirements, and hydride attack to the alkene
occurs. Two possible pathways are now present depending
on attack to C3 or C2. In the first case an agostic interaction
occurs with C2-H (C(Z)), and its equilibration with the
analogous C(E) occurs by simple interchange of an agostic
residue with a Pt H agostic interaction demonstrated
3 3 3
that such rotation can have activation barriers as high as
ca. 9 kcal/mol.28 As a consequence, initial E/Z selectivity
(27) Creve, S.; Oevering, H.; Coussens, B. B. Organometallics 1999,
18, 1967–1978.
(28) Tempel, D. J.; Brookhart, M. Organometallics 1998, 17, 2290–
2296.

Article
Organometallics, Vol. 29, No. 6, 2010 1495
interaction between methylene C2 hydrogens, which requires
a simple small rotation of the C2-C3 bond of the reagent.
The more stable C(E) species then converts to Pt-H species
with an E-coordinated alkene, which is displaced by an
incoming Z alkene. A similar process is possible from the
original Pt-H species if attack of the hydride occurs on C2
(Scheme 5). In agreement with this view is also the solvent
effect exhibited in Table 2. Where competition exists with
more coordinating solvents, enrichment in the E isomer via
consecutive coordination of the (Z)-2-alkene becomes more
and more difficult. However, where the steric hindrance of
the (Z)-2-alkene is not compatible with the steric congestion of
the catalyst (i.e., with 3,4-dimethoxyallylbenzene, eugenol, and
safrole as substrates) the first isomerization process is much
slower and the consecutive Z to E interconversion becomes
negligible, leaving the selectivity unaffected.
isomerization process that occurs through coordination of
the Z alkene to Pt and hydride attack on C3 rather than on
C2, leading to CZ agostic species that are in thermodynamic
equilibrium with CE. Calculations provided also quantitative
support to the preferred coordination of the Pt-hydride
species, derived by elimination of the fluorinated alkene, to
terminal alkene, followed by Z and eventually by E products.
In fact, binding energies calculated for the three possible
alkenes are terminal -18.74, Z-16.54, and E-15.00 kcal/mol,
in agreement with the selectivity profile observed experimen-
tally during the reaction progress (Figure 1).
Conclusion
The catalytic system here reported has the potential to be a
viable protocol for the synthesis of a variety of fragrances via
the catalytic isomerization of the corresponding terminal
alkenes, as it shows good yields within hours and, in some
cases, excellent E/Z selectivities with low (<1% mol) catalyst
loading under mild experimental conditions, in air and with
ordinary solvents. In particular the results observed in the
isomerization of estragole (99% yield, E/Z 97:3) at 50 °C in
12 h are, to our knowledge, the best thus far reported in terms
of overall activity and selectivity under the mildest experi-
mental conditions. It is worth noting that despite the well-
known affinity of PtII complexes for terminal alkenes, this is
the first example of an isomerization reaction catalyzed by
this transition metal. Moreover a pathway involving agostic
species has been clearly identified here for the first time as
demonstrated by 31P investigation. This is made possible by
the existence of unusual sterically hindered PtII-alkyl species
Crystal Structure Determination of 4e and DFT Calcula-
tions on Possible Intermediates. In order to substantiate the
reaction pathways reported in Schemes 4 and 5 and shed light
on the nature of the Pt H agostic intermediates present as
3 3 3
resting states in the catalytic cycle, several unsuccessful
attempts were made to crystallize such species out of the
reaction mixture. Nevertheless, a reliable starting point
concerning ligand conformation was obtained from crystals
of 4e grown from dichloromethane/hexane, whose structure
was solved (Figure 4). An interesting and somehow unex-
pected feature is the -(CH2)4- alkyl chain being not puck-
ered but completely above (or below) the coordination plane.
This positions two aromatic rings of the ligand pointing
almost perpendicularly under the coordination plane, while
the other two remain in equatorial position, acting as arms
that limit the space occupied by the perfluoro-aromatic
moiety and water. The value found for the P1-Pt-P2 bite
angle (93°) is in line with that proposed by van Leeuwen for
dppb in Rh complexes1,29 and is sufficiently large to induce
agostic interactions in the catalytically active PtII species.
With these geometrical data as a starting point for energy
minimizations, we performed a series of DFT calculations
aiming at comparing the energies of the different agostic
intermediates30 A, B, C, and D (Scheme 6) to find support for
the mechanisms reported above (Schemes 4 and 5).
capable of making the agostic Pt H interchange the slow
3 3 3
step of the isomerization process and by the moderate
aptitude of this metal to promote β-hydride elimination with
respect to other transition metal centers such as Rh, Ru, Pd,
or Ni. Although obvious intermediates in classical 1,2-
hydrogen shift via hydride addition/elimination involving
β-hydride elimination, agostic species in isomerization have
never been identified before.
Examples of asymmetric isomerization of simple alkenes
are rare,31 while only a few more data are available for
isomerization of double bonds close to functional groups
(i.e., allylic alcohols).32 The very high selectivity toward the
E alkene observed with the present system in some cases
relies mainly on a consecutive process allowing only the
Z alkene to enter the coordination sphere of the catalyst
and be converted into the E isomer. The strong ligand effect
on catalytic activity and product selectivity observed with
complexes 4 augurs well for the development of an asym-
metric version of the PtII-catalyzed alkene isomerization,
which is currently underway in our laboratory.
Comparison between agostic species A and B derived
from allylbenzene shows that B is more stable by about 1 to
1.5 kcal/mol, while the energy difference is about half for the
more sterically demanding safrole. Both species are present
in the catalytic cycle before the rate-determining step, and
the calculated energies are in agreement with the experimen-
tal spectroscopic observation of only one agostic species
being practically seen for allylbenzene (B) and two for safrole
and other disubstituted allylbenzene derivatives (A and B
with small energy differences) as reported in Figure 2. In
principle, an energy difference of 1 kcal gives rise to an
isomer population of ∼83:17, likely leading to the possibility
to practically observe spectroscopically only one species.
Further comparison of the energies for species C and D
shows that the latter are generally less stable as expected
considering that the aromatic residue in D is closer to the
aromatic rings of the chelating diphosphine, while in C
species it is placed on an opposite position with respect to
the Pt atom. This observation supports a Z to E consecutive
Experimental Section
Reagents and Materials. General Procedures. 1H NMR and
31P{1H} NMR spectra were recorded at 298 K, unless otherwise
stated, on a Bruker AVANCE 300 spectrometer operating at
300.15 and 121.50 MHz, respectively. δ values in ppm are
relative to SiMe4 and 85% H3PO4. 19F{1H} NMR spectra were
(31) Chen, Z.; Halterman, R. L. J. Am. Chem. Soc. 1992, 114, 2276–
2277.
(29) (a) Sakakura, T. J. Organomet. Chem. 1984, 267, 171.
(b) Abatjoglou, A. G.; Billig, E.; Bryant, D. R. Organometallics 1984, 3, 923.
(30) Lein, M. Coord. Chem. Rev. 2009, 253, 625–634.
(32) (a) Tanaka, K.; Qiao, S.; Tobisu, M.; Lo, M. M.-C.; Fu, G. C.
J. Am. Chem. Soc. 2000, 122, 9870–9871. (b) Tanaka, K.; Fu, G. C. J. Org.
Chem. 2001, 66, 8177–8186.

1496 Organometallics, Vol. 29, No. 6, 2010
Scarso et al.
2
recorded at 298 K on a Bruker AC200 spectrometer operating at
188.25 MHz. δ values in ppm are relative to CFCl3. All reactions
were monitored by 1H NMR.
1JPt-P = 1581 Hz, JP-P = 17 Hz, PC-trans). FT-IR (ν, CsI):
283.34 (Pt-Cl str.).
[Pt(Ph)(OH2)(dppb)](OTf), 3. The procedure is similar to that
followed for complex 1 starting from [PtCl(Ph)(dppb)] (0.14 g,
0.19 mmol) and 0.43 mL of an acetone solution of AgOTf
(0.20 mmol). Yield: 0.12 g, 80.7%. Anal. Calcd for C35H35F3O4-
P2PtS: C, 48.56; H, 4.07; S, 3.70. Found: C, 48.39; H, 3.89; S,
3.87. 1H NMR (δ, acetone-d6): 6.62-7.91 (m, Ar), 3.12 (m,
CH2), 3.00 (m, CH2), 1.93 (m, CH2), 1.60 (m, CH2). 31P{1H}
Substrates. All alkenes used as substrates are commercial
products (Aldrich) and were used without further purification.
Synthesis of the Complexes. All work was carried out with the
exclusion of atmospheric oxygen under a dinitrogen atmosphere
using standard Schlenk techniques. Solvents were dried and
purified according to standard methods. Substrates were puri-
fied by passing through neutral alumina and stored in the dark
at low temperature. AgOTf, triphenylphosphine, and dppb were
commercial products and were used without purification. The
complexes [PtCl(Ph)(COD)],33 [Pt(C6H3(CF3)2)2(dppe)],34 [Pt-
Cl(Me)(dppb)],35 [PtCl(CF3)(dppb)],36 [PtCl(C6F5)(PPh3)2],37
[Pt(C6F5)(OH2)(P-P)](OTf) (with P-P = dppm 4a, dppe 4b,
(S,S)-chiraphos 4c, dppp 4d, and dppb 4e),17 [Pd(C6F5)(OH2)-
1
2
NMR (δ, acetone-d6): 24.55 (d, JPt-P = 1757 Hz, JP-P
=
14 Hz, PC-trans); 8.41 (d, 1JPt-P=4654 Hz, 2JP-P=14 Hz, PO-trans).
19F{1H} NMR (δ, acetone-d6): -80.66 (s, OTf).
[Pt(C6F5)(OH2)(PPh3)2](OTf), 5. To a solution of [PtCl-
(C6F5)(PPh3)2] (0.13 g, 0.14 mmol) in wet dichloromethane
(20 mL) and acetone (5 mL) at room temperature was added
0.35 mL of an acetone solution of AgOTf (0.16 mmol). The
suspension was stirred overnight; then the solid AgCl was
filtered off and the solution was concentrated. Upon treatment
with diethyl ether, a white solid was obtained, filtered off, and
dried under vacuum. Yield: 0.13 g, 86.3%. Anal. Calcd for
C43H32F8O4P2PtS: C, 49.01; H, 3.06. Found: C, 48.97; H, 3.02.
1H NMR (δ, CDCl3): 7.30-7.55 (m, Ar). 31P{1H} NMR (δ,
39
(dppe)](OTf) (4b-Pd),38 and [Ptdppb(OH)]2(BF4)2 (7) were
synthesized following procedures reported in the literature.
Elemental analyses were performed at the Department of Ana-
lytical, Inorganic and Organometallic Chemistry of the
ꢀ
Universita di Padova.
New Complexes. [Pt(Me)(OH2)(dppb)](OTf), 1. To a solu-
tion of [PtCl(Me)(dppb)] (0.16 g, 0.24 mmol) in wet dichloro-
methane (30 mL) was added 0.54 mL of an acetone solution of
AgOTf (0.25 mmol). The suspension was stirred for 2 h; then the
solid AgCl was filtered off, and the solution was concentrated.
Upon treatment with n-hexane, a white solid was obtained,
filtered off, and dried under vacuum. Yield: 0.12 g, 61.2%. Anal.
Calcd for C30H33F3O4P2PtS: C, 44.83; H, 4.14; S, 3.99. Found:
C, 44.39; H, 4.29; S, 3.87. 1H NMR (δ, acetone-d6): 7.55-7.91
(m, Ar), 1.59 (m, CH2), 2.03 (m, CH2), 2.80 (m, CH2), 3.07 (m,
1
CDCl3): 22.28 (m, JPt-P = 2721 Hz); 19F{1H} NMR (δ,
CDCl3): -80.88 (s, OTf), -122.55 (d, 3JPt-F = 427 Hz, 3JF-F
=
22 Hz, o-F), -165.62 (t, 3JF-F=20 Hz, p-F), -166.85 (m, m-F).
[PtCl(C6H3(m,m-CF3)2)(dppe)]. To a solution of [Pt(C6H3-
(CF3)2)2(dppe)] (0.20 g, 0.20 mmol) in dichloromethane (20 mL)
and methanol (5 mL) was added 0.015 mL of acetyl chloride
(0.20 mmol). The solution was stirred overnight and then
concentrated. The product was purified by chromatography
on silica gel (diethyl ether/n-hexane, 9:2, Rf =0.49). Yield: 0.11
g, 66.6%. Anal. Calcd for C34H27ClF6P2Pt: C, 48.50; H, 3.23.
Found: C, 48.39; H, 3.29. 1H NMR (δ, CDCl3): 7.20-7.97 (m,
Ar), 2.49 (m, CH2), 2.22 (m, CH2). 31P{1H} NMR (δ,CDCl3):
40.43 (s, 1JPt-P =1765 Hz, PC-trans), 39.63 (s, 1JPt-P =3996 Hz,
PO-trans). 19F{1H} NMR (δ, CDCl3): -65.40 (s, CF3).
CH2), 0.29 (m, 2JPt-H = 49 Hz, 3Jcis-P-H = 2 Hz, 3Jtrans-P-H
7 Hz, CH3). 31P{1H} NMR (δ, acetone-d6): 30.20 (m, 1JPt-P
=
=
1849 Hz, 2JP-P = 14 Hz, PC-trans); 12.00 (m, 1JPt-P = 4651 Hz,
2JP-P=14 Hz, PO-trans). 19F {1H} NMR (δ, acetone-d6): -80.68
(s, OTf).
[Pt(CF3)(OH2)(dppb)](OTf), 2. The complex was synthesized
as for 1 starting from [PtCl(CF3)(dppb)] (0.15 g, 0.21 mmol) and
0.46 mL of an acetone solution of AgOTf (0.21 mmol). Yield:
0.10 g, 57.8%. Anal. Calcd for C30H30F6O4P2PtS: C, 42.01; H,
[Pt(C6H3(m,m-CF3)2)(OH2)(dppe)](OTf), 6. The complex
was synthesized as for 1 starting from [PtCl(C6H3(CF3)2)-
(dppe)] (0.065 g, 0.072 mmol) and 0.17 mL of an acetone
solution of AgOTf (0.079 mmol). Yield: 0.059 g, 78.4%. Anal.
Calcd for C35H29F9O4P2PtS: C, 43.17; H, 3.00; S, 3.29. Found:
C, 43.39; H, 2.89; S, 3.37. 1H NMR (δ, acetone-d6): 7.35-8.01
(m, Ar), 2.97 (m, CH2), 2.66 (m, CH2). 31P{1H} NMR (δ,
1
3.53; S, 3.74. Found: C, 42.12; H, 3.39; S, 3.87. H NMR (δ,
acetone-d6): 7.58-8.02 (m, Ar), 3.16 (m, CH2), 2.83 (m, CH2),
2.10 (m, CH2), 1.63 (m, CH2). 31P {1H} NMR (δ, acetone-d6):
19.29 (m, JPt-P = 1904 Hz, JF-P = 55 Hz, JP-P = 21 Hz,
1
3
2
acetone-d6): 49.87 (s, 1JPt-P =1863 Hz, PC-trans), 33.15 (s, 1JPt-P
=
1
P
C-trans); 6.09 (m, JPt-P = 4321 Hz, PO-trans). 19F {1H} NMR
4385 Hz, PO-trans). 19F{1H} NMR (δ, acetone-d6): -65.12 (s,
CF3), -80.58 (s, OTf).
(δ, acetone-d6): -30.76 (dd, 2JPt-F = 483 Hz, 3Jcis-P-F = 8 Hz,
3Jtrans-P-F =55 Hz, CF3), -80.61 (s, OTf).
Isomerization Reactions. These were carried out in air in a 2
mL vial equipped with a screw-capped silicone septum to allow
sampling. Stirring was performed by a Teflon-coated bar driven
externally by a magnetic stirrer (700 rpm). Constant tempera-
ture was maintained by water or alcohol circulation through an
external jacket connected with a thermostat. Typically, the
proper amount of catalyst (0.016 mmol, 2% mol) was placed
in the vial, followed by the solvent (1.0 mL); subsequently the
vial was thermostated. After stirring for 10 min, the substrate
(0.83 mmol) was added and time was started. All reactions were
[PtCl(Ph)(dppb)]. To a solution of [PtCl(Ph)(cod)] (0.10 g,
0.12 mmol) in dichloromethane (20 mL) at room temperature
was added 0.11 g (0.25 mmol) of dppb. The solution was stirred
for 1 h, then was concentrated and treated with n-hexane, giving
a white solid, which was filtered off, washed with hexane, and
dried under vacuum. Yield: 0.16 g, 92.3%. Anal. Calcd for
C34H33ClP2Pt: C, 55.63; H, 4.53. Found: C, 55.67; H, 4.42. 1H
NMR (δ, CD2Cl2): 6.51-7.80 (m, Ar), 2.69 (m, CH2), 2.37 (m,
CH2), 2.00 (m, CH2), 1.59 (m, CH2). 31P{1H} NMR (δ,CD2Cl2):
17.51 (m, 1JPt-P = 4293 Hz, 2JP-P = 17 Hz, PCl-trans), 9.09 (m,
1
monitored by GC analysis and H NMR by periodically sam-
pling directly from the reaction mixtures with a microsyringe.
Quenching of the samples by adding an excess of LiCl was
performed prior to NMR analysis.
(33) Clark, H. C.; Manzer, L. E. J. Organomet. Chem. 1973, 59, 411–
428.
(34) Eaborn, C.; Odell, K. J.; Pidcock, A. J. Chem. Soc., Dalton
Computational Details. The calculations were performed
with the ADF2007.140 package at the BLYP41 level using the
Trans. 1978, 357–368.
(35) Romeo, R.; D’Amico, G. Organometallics 2006, 25, 3435–3446.
(36) Zanardo, A.; Michelin, R. A.; Pinna, F.; Strukul, G. Inorg.
Chem. 1989, 28, 1648–1653.
(37) Al-Fawaz, A.; Aldridge, S.; Coombs, D. L.; Dickinson, A. A.;
Willock, D. J.; Ooi, L.-L.; Light, M. E.; Coles, S. J.; Hursthouse, M. B.
J. Chem. Soc., Dalton Trans. 2004, 4030–4037.
(40) (a) te Velde, G.; Bickelhaupt, F. M.; van Gisbergen, S. J. A.;
Fonseca Guerra, C.; Baerends, E. J.; Snijders, J. G.; Ziegler, T. Chem-
istry with ADF. J. Comput. Chem. 2001, 22, 931–967. (b) Fonseca Guerra,
C.; Snijders, J. G.; te Velde, G.; Baerends, E. J. Theor. Chem. Acc. 1998, 99,
391. (c) ADF2007.01, SCM, Theoretical Chemistry; Vrije Universiteit:
Amsterdam, The Netherlands, http://www.scm.com.
(38) Casado, A. L.; Espinet, P.; Gallego, A. M.; Martı
´
nez-Ilarduya,
J. M. Chem. Commun. 2001, 339–340.
(39) Gavagnin, R.; Cataldo, M.; Pinna, F.; Strukul, G. Organome-
tallics 1998, 17, 661–667.
(41) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098–3100. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.

Article
Organometallics, Vol. 29, No. 6, 2010 1497
TZ2P basis set in the frozen core approximation. Scalar
relativistic effects were taken into account by the zeroth-
order regular approximation (ZORA).42 The geometry optimi-
zations were performed without any symmetry constraint,
followed by analytical frequency calculations to confirm that
a minimum had been reached. Cartesian coordinates and en-
ergies of stationary points are reported in the Supporting
Information.
Acknowledgment. The authors thank MIUR, Uni-
ꢁ
ꢀ
versita Ca’ Foscari di Venezia, and Universita degli Studi
di Padova for financial support. G.S. thanks Johnson-
Matthey for the loan of platinum. L. Sperni is gratefully
acknowledged for GC-MS analysis.
Supporting Information Available: Molecular modeling co-
ordinates for agostic species A-D and the alkene isomers
reported in Table 4 are reported together with a cif file for
complex 4e. This material is available free of charge via the
Internet at http://pubs.acs.org.
(42) Van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys.
1994, 101, 9783.