however, because the ruthenium catalyst can be very easily
separated from the cationic iron complex after the metathesis
reaction. The catalyst is extracted using ether in which the
iron complexes are not soluble. The separation of the catalyst
has appeared as a non-trivial problem in the myriad of
decomplexation of 1c, whereas it is not available in pure
form by metathesis of the iron-free derivative 1d (Scheme
1). The same situation arises in the p-xylene series 2a-g.
1
0
+
publications dealing with the RCM of organic substrates
Scheme 1. Selectivity Effected by the CpFe Moiety in the
Metathesis Reactionsa
11
and has been the special subject of very recent papers. After
metathesis of the iron complexes, separation of the ruthenium
catalyst is followed by the easy observation of its methylene
1
signal at 19 ppm vs SiMe
4
in H NMR, as reported by Grubbs
2 2 3 2
for the less active complex [Ru(dCH )Cl (PCy ) ] in which
the benzylidene ligand has been replaced by the methylene
12
group coming from the terminal olefin. An excellent green
alternative is to carry out the RCM reaction of the neutral,
iron-free aromatic compound 3d in the ionic liquid 1-butyl-
6
3-imidazolium hexafluorophosphate ([bmin]PF ) at 80 °C,
which gives a 75% yield of 3e after extraction of the neutral
a
methathesis product using ether.13
(i) 5% [Ru(dCH )Cl (PCy ) ], CH Cl , rt, 1 h; (ii) hν (vis),
2
2
3 2
2
2
MeCN, 1 equiv PPh
3
; (iii) 10% [Ru(dCH
2
)Cl
2 3 2 2 4 2
(PCy ) ], C H Cl ,
The case of toluene, p-xylene, mesitylene, and tetra-
methylbiphenyl leading to triple allylation at each benzylic
carbon is distinct from the others for which only two allyl
groups are found in the benzylic position (Table 1). The
RCM reaction rapidly forms a cyclopentene ring in benzylic
position, and the third allyl group is left, which is not the
case when only two allyl groups are present. Thus, the third
allyl group can potentially give rise to cross metathesis (CM)
in the case of the mono- and polymethylaromatics except
8
0 °C, 3 days; (iv) 5% [Ru(dCH
2
)Cl (PCy )
2
3 2
], rt, 3 days.
+
Interestingly, the iron-free arene 3d resulting from CpFe -
8b
mediated nonallylation of mesitylene in 3a gives the RCM
reaction selectively since its CM reaction is slow, again
probably for steric reasons. Both the RCM and RCM + CM
products are of interest. The RCM products are hetero-
bifunctional. For instance, the RCM products provided from
3b and 4b are heterobifunctional dendritic cores. The RCM
+ CM products will have potential in materials chemistry
since linear polymers should be available from p-xylene by
ADMET and dendritic polymers should be accessible from
mesitylene. This aspect of the research line is under study.
In conclusion, we have provided, from very simple arenes,
a general and facile method of synthesis of a new class of
organometallic and organic polycyclic cyclopentenyl aryl
8a
8c
when methyl groups are adjacent (6af6b and 8af8b in
5
+
-
6
which Co ) η -CpCo , PF ). This possibility has been
scrutinized in the example of toluene. Its iron-free, triallylated
derivative 1d gives competitive RCM and CM reaction, and
no selectivity is observed. It is only possible to isolate the
organic bis-arene product 1g resulting from concomitant
RCM and CM of 1d. On the other hand, CM is considerably
1
4
+
slowed in the CpFe complex 1b, probably for steric reasons,
+
whereas RCM proceeds readily. Thus, the RCM product 1c
is selectively obtained with the iron complex 1b at 20 °C;
the corresponding free arene 1e can be obtained by photolytic
derivatives. The CpFe moiety not only provides a straight-
forward polyallylation method under routine conditions,
providing new, remarkable topologies, but also plays a key,
useful role in the metathesis process for the separation of
(
11) (a) Maynard, H. D.; Grubbs, R. H. Tetrahedron Lett. 1999, 40,
1
5
the products and the selectivity.
4
137-4140. (b) Paquette, L. A.; Schloss, J. D.; Efremoy, F.; Gallou, F.;
Mendez-Andino, J.; Yang, J. Org. Lett. 2000, 2, 1259-1261. (c) Ahn, Y.
M.; Yang, K.; Georg, G. I. Org. Lett. 2001, 3, 1411-1413.
Acknowledgment. We thank Dr. J. Ruiz for samples of
a and 8a and Drs. J. Juiz and S. Nlate (Bordeaux) for fruitful
(12) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996,
2
1
18, 100-110.
13) Buijsman, R. C.; Van Vuuren, E.; Sterrenburg J. G. Org. Lett. 2001,
, 3785-3787.
14) (a) Newkome, G. R.; Moorefield, C. N.; V o¨ gtle, F. Dendritic
Molecules: Concepts, Syntheses and PerspectiVes; VCH: New York, 1996.
b) Sunder, A.; Heinemann, J.; Frey, H. Chem. Eur. J. 2000, 6, 2499-
506.
15) All the new complexes have been suitably characterized by H and
(
discussions.
3
(
Supporting Information Available: Experimental pro-
1
13
cedures and spectroscopic data ( H and C NMR spectra
and MALDI TOF mass spectra) for all the new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(
2
1
(
13
C NMR (including the distinct endo and exo positions of the new
8
c
complexes 6c and 8c in which the directionality is expected ) and by their
molecular peak in the MALDI TOF mass spectrum. Spectroscopic and
analytical data are provided in Supporting Information.
OL0172875
Org. Lett., Vol. 4, No. 4, 2002
653