Role of the gem-difluoro moiety in the tandem ring-closing metathesis-olefin isomerization: Regioselective preparation of unsaturated lactams

Article abstract of DOI:10.1021/jo0525635

Careful selection of the metathesis catalyst, solvent, and reaction conditions allows for the efficient and regioselective synthesis of isomeric fluorinated and nonfluorinated lactam derivatives II and III from precursor amides I through a ring-closing me

Full text of DOI:10.1021/jo0525635

Role of the gem-Difluoro Moiety in the Tandem Ring-Closing
Metathesis-Olefin Isomerization: Regioselective Preparation of
Unsaturated Lactams
Santos Fustero,*,†,‡ Mar´ıa Sa´nchez-Rosello´,† Diego Jime´nez,† Juan F. Sanz-Cervera,†,‡
Carlos del Pozo,† and Jose´ Luis Acen˜a‡
Departamento de Qu´ımica Orga´nica, UniVersidad de Valencia, E-46100 Burjassot, Spain, and Laboratorio
de Mole´culas Orga´nicas, Centro de InVestigacio´n “Principe Felipe”, E-46013 Valencia, Spain
santos.fustero@uV.es
ReceiVed December 13, 2005
Careful selection of the metathesis catalyst, solvent, and reaction conditions allows for the efficient and
regioselective synthesis of isomeric fluorinated and nonfluorinated lactam derivatives II and III from
precursor amides I through a ring-closing metathesis (RCM) reaction or a tandem RCM-isomerization
protocol, respectively. The presence of the gem-difluoro moiety in the starting materials exerts a pivotal
effect by directing the isomerization step, making the overall tandem transformation a regioselective
process. The scope, limitations, and synthetic usefulness of this protocol are also discussed.
Introduction
In the past decade, the ring-closing metathesis reaction (RCM)
has emerged as a powerful synthetic tool for the creation of
carbon-carbon bonds, allowing for the synthesis of medium-
and large-sized unsaturated heterocyclic systems.1 This develop-
ment has been made possible mostly as a result of the discovery
of new, more stable ruthenium catalysts,2 the three most
common of which are shown in Figure 1.
FIGURE 1. The most common metathesis catalysts.
Although the RCM reaction is, in general, a clean process,
ruthenium catalysts 1 and 2 occasionally produce what in some
cases may be construed as undesired side reactions. Thus, the
desired reaction products are sometimes accompanied by
variable amounts of other compounds that arise as a result of
isomerization of the newly created double bond.3,4 For example,
in their study of the RCM mechanism of N,N-diallyltosylamide
catalyzed by several ruthenium complexes, Dixneuf and co-
workers found that the reaction conditions and the nature of
the catalyst had a great effect on the final mixture.5
Several reports indicate that the decomposition products of
Ru-metathesis catalysts (generally Ru-hydride complexes) may
be responsible for these side reactions and specifically for alkene
isomerization.6 In addition, Grubbs et al. recently demonstrated
† Universidad de Valencia.
‡ Centro de Investigacio´n “Principe Felipe”.
(4) During the preparation of this manuscript Grubbs and co-workers
published a paper that describes a method for the suppression of undesired
isomerization side reactions that are sometimes associated with olefin
metathesis: Hong, S. H.; Sanders, D. P.; Lee, C. W.; Grubbs, R. H. J. Am.
Chem. Soc. 2005, 127, 17160-17161.
(1) (a) Philips, A. J.; Abell, A. D. Aldrichimica Acta 1999, 32, 75-89.
(b) Schultz-Fademrecht, C.; Deshmukh, P. H.; Malagu, K.; Procopiou, P.
A.; Barrett, A. G. Tetrahedron 2004, 60, 7515-7524. (c) Deiters, A.; Martin,
S. F. Chem. ReV. 2004, 104, 2199-2238.
(2) Grubbs, R. H. In Handbook of Metathesis; Wiley-VCH Verlag Gmbh
and Co. KGaA: Weinheim, 2003; Vols. 1-3.
(3) Review: Schmidt, B. Eur. J. Org. Chem. 2004, 1865-1880 and
references therein.
(5) Bassetti, M.; Centola, F.; Se´meril, D.; Bruneau, C.; Dixneuf, P. H.
Organometallics 2003, 22, 4459-4466.
(6) Fu¨rstner, A.; Ackermann, L.; Gabor, B.; Goddard, R.; Lehmann, C.;
Mynott, R.; Stelzer, F.; Thiel, D. Chem. Eur. J. 2001, 7, 3236-3253.
10.1021/jo0525635 CCC: $33.50 © 2006 American Chemical Society
Published on Web 03/09/2006
2706
J. Org. Chem. 2006, 71, 2706-2714

RegioselectiVe Preparation of Unsaturated Lactams
SCHEME 1. Tandem “Isomerization-RCM” and Tandem
“RCM-Isomerization” Reactions
examples of the application of the tandem RCM-isomerization
protocol to date.12
Concurrently, the development of new synthetic methodolo-
gies for the preparation of nitrogen heterocyclic compounds is
of great importance in organic synthesis, as these substances
constitute the basic structural elements of many potentially
bioactive compounds. Among nitrogen heterocycles, lactams
have attracted considerable interest, mostly because of their
usefulness in drug discovery. One approach that has been applied
successfully to the design of biologically active compounds is
the inclusion of such compounds in peptidic chains. Thus, for
instance, Freidinger lactams have been used extensively for the
preparation of conformationally restricted peptidomimetics.13 In
addition, lactams are present in the backbone of several natural
products, including bengamides, which have shown important
cytotoxic properties.14
that the second generation ruthenium catalyst 2 evolves into a
ruthenium hydride when heated in benzene.7 Despite the
widespread perception that these side reactions constitute a
general problem, it is also true that they have been used
advantageously by several authors in specific synthetic strate-
gies. Thus, a growing number of newly discovered catalytic
processes mediated by Ru catalysts 1-3 that do not involve
olefin metathesis and therefore possess even greater synthetic
versatility have appeared in the literature.8,9
From a synthetic point of view, the tandem “isomerization-
RCM” and the tandem “RCM-isomerization” reactions might
constitute the most useful and selective nonmetathetic trans-
formations mediated by these ruthenium catalysts. The combi-
nation of these two strategies allows for the preparation of
different types of carbo- and heterocyclic systems in a very
efficient and simple manner (Scheme 1).
Whereas several examples of the application of the first
protocol are already known,10 the alternative RCM-isomeriza-
tion is much less common. Two recent examples of this
approach have been Snapper’s and Schmidt’s independent use
of tandem RCM-isomerization reactions for the synthesis of
cyclic enol ethers,11 which would otherwise be far more difficult
to obtain. It is important to point out, however, that in both
cases the applied protocols require the presence of an additive
to promote the transformation of the Ru-carbene into an Ru-
hydride complex, which is ultimately responsible for the
isomerization processes. It is also worth noting that apart from
these two preparations of enol ethers, there are no other
Unsaturated lactams constitute especially versatile building
blocks because the presence of the double bond allows for
further functionalization, including but not limited to epoxidation
and dihydroxylation. Strategies involving RCM reactions have
already been applied successfully to the synthesis of several
unsaturated lactam derivatives.15 The incorporation of fluorine
atoms in these compounds would constitute an interesting
structural variation, as their introduction is known to impart
unique biological properties to the parent molecule.16 Several
methodologies specifically regarding the preparation of fluori-
nated lactams have been described in the literature,17 but to the
best of our knowledge only one example of their preparation
through an RCM reaction has been described thus far.18
In this paper, we report on new examples of the tandem
RCM-isomerization reaction that allow for the efficient prepa-
ration of fluorinated as well as nonfluorinated, unsaturated
lactam derivatives. The pivotal role of the gem-difluoro moiety
in directing the isomerization step is also discussed. The novelty
here is that, in contrast with previously reported methods,11 our
protocol for the tandem sequence RCM-isomerization does not
require the use of any additives to generate the ruthenium
hydride species, which are believed to be responsible of the
regioselective isomerization reaction. We also discuss the scope,
limitations, and synthetic usefulness of this new protocol.
(12) Snapper and co-workers described one example of the preparation
of a tosyl enamide by means of his RCM-isomerization protocol (see ref
11a).
(7) Hong, S. H.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2004,
126, 7414-7415.
(13) (a) Freidinger, R. M.; Veber, D. F.; Perlow, D. S.; Brooks, J. R.;
Saperstein, R. Science 1980, 656-658. (b) Freidinger, R. M.; Perlow, D.
S.; Veber, D. F. J. Org. Chem. 1982, 47, 104-109. (c) For a review, see:
Cluzeau, J.; Lubell, W. D. Biopolymers (Peptide Sci.) 2005, 80, 98-150.
(14) See, for example: (a) Adamczeski, M.; Quin˜oa`, E.; Crews, P. J.
Org. Chem. 1990, 55, 240-242. (b) Fernandez, R.; Dherbomez, M.;
Letourneux, Y.; Nabil, M.; Verbist, J. F.; Biard, J. F. J. Nat. Prod. 1999,
62, 678-680. (c) Thale, Z.; Kinder, F. R.; Bair, K. W.; Bontempo, J.;
Czuchta, A. M.; Versace, R. W.; Phillips, P. E.; Sanders, M. L.; Wattanasin,
S.; Crews, P. J. Org. Chem. 2001, 66, 1733-1741.
(15) (a) Fu, G. C.; Grubbs, R. H. J. Am. Chem. Soc. 1992, 114, 7324-
7325. (b) Fu, G. C.; Nguyen, S. T.; Grubbs, R. H. J. Am. Chem. Soc. 1993,
115, 9856-9857. (c) Huwe, C. M.; Kiehl, O. C.; Blechert, S. Synlett 1996,
65-66. (d) Rutjes, F. P.; Schoemaker, H. E. Tetrahedron Lett. 1997, 38,
677-680. (e) Miller, S. J.; Blackwell, H. E.; Grubbs, R. H. J. Am. Chem.
Soc. 1996, 118, 9606-9614.
(16) Baasner, B.; Hagemann, H.; Tatlow, J. C. Houben-Weyl Methods
in Organic Chemistry, Georg Thieme Verlag: Sttutgart, 1999; Vol. E10a,
Organo-Fluorine Compounds.
(8) For a recent revision of nonmetathetic reactions mediated by Grubbs’
catalysts, see: Alcaide, B.; Almendros, P. Chem. Eur. J. 2003, 9, 1258-
1262.
(9) For the deprotection of allylic amines through isomerization reactions
with Grubbs’ catalysts, see: (a) Alcaide, B.; Almendros, P.; Alonso, J. M.;
Aly, M. F. Org. Lett. 2001, 3, 3781-3785. (b) Alcaide, B.; Benito, P.;
Alonso, J. M.; Luna, A. Synthesis 2005, 668-672. For the Kharasch addition
of chloroform across olefins mediated by ruthenium catalysts, see: (c)
Tallarico, J. A.; Malnick, L. M.; Snapper, M. L. J. Org. Chem. 1999, 64,
344-345.
(10) For representative examples, see: (a) Fu¨rstner, A.; Thiel, O. R.;
Ackermann. L.; Schanz, H.-J.; Nolan, S. P. J. Org. Chem. 2000, 65, 2204-
2207. (b) Kinderman, S. S.; van Maarseveen, J. A.; Schoemaker, H. E.;
Hiemstra, H.; Rutjes, F. P. J. T. Org. Lett. 2001, 3, 2045-2048. (c) Arisawa,
M.; Terada, Y.; Nakagawa, M.; Nishida, A. Angew. Chem., Int. Ed. 2002,
41, 4732-4734. (d) Huang, J.; Xiong, H.; Hsung, R. P.; Rameshkumar,
C.; Mulder, J. A.; Grebe, T. P. Org. Lett. 2002, 4, 2417-2420.
(11) (a) Sutton, A. E.; Seigal, B. A.; Finnegan, D. F.; Snapper, M. L. J.
Am. Chem. Soc. 2002, 124, 13390-13391. (b) Schmidt, B. J. Org. Chem.
2004, 69, 7672-7687. For a recent example of the application of the RCM-
isomerization protocol, see: Bressy, C.; Menant, C.; Piva, O. Synlett 2005,
577-582.
(17) (a) Uneyama, K.; Yanagiguchi, T.; Asai, H. Tetrahedron Lett. 1997,
38, 7763-7764. (b) Nagashima, H.; Isono, Y.; Iwamatsu, S. J. Org. Chem.
2001, 66, 315-319. (c) Weber, K.; Gmeiner, P. Synlett 1998, 885-887.
(18) Marhold, M.; Buer, A., Hiemstra, H.; van Maarseveen, J. H.; Haufe,
G. Tetrahedron Lett. 2004, 45, 57-60.
J. Org. Chem, Vol. 71, No. 7, 2006 2707

Fustero et al.
TABLE 1. Regioselective Preparation of Fluorinated Lactams 5
and 6
SCHEME 2. Initial Formation of Seven-Membered Lactams
5a and 6a
entry
4
methoda
R
yield (%) (5) yield (%) (6)
1
2
3
4
5
6
7
8
9
4a
4b
4c
4d
4a
4b
4c
4d
4e
4f
A
A
A
A
B
B
B
B
B
B
CH2Ph
95 (5a)
89 (5b)
70 (5c)
82 (5d)
73 (6a)
93 (6b)
72 (6c)
95 (6d)
70 (6e)
91 (6f)
Results and Discussion
p-MeOC6H4
CH2CO2Et
c-C6H11
As part of our ongoing project directed toward the preparation
of fluorinated nitrogen heterocycles,19 we planned to use a ring-
closing metathesis reaction as the key step in the preparation
of new fluorinated lactams. Our initial approach for the
preparation of seven-membered fluorinated lactams 5 was quite
simple: as starting materials for the RCM reaction we synthe-
sized several diolefinic tertiary amides 4 (Scheme 2), which
were easily prepared either through N-allylation of the corre-
sponding secondary fluorinated amides or reaction of 2,2-
difluoro-4-pentenoic acid with secondary allylic amines.20
We expected amides 4 to afford ꢀ-lactams 5 as major reaction
products under standard RCM conditions, and indeed, in an
initial assay, the RCM of 4a took place easily in 2 h with catalyst
2 in refluxing dichloromethane. However, to our surprise, the
expected reaction product 5a was only a minor constituent of
the crude reaction mixture, appearing together with an isomer
in a 1:3 ratio. Fortunately, these two products were easily
separated with standard chromatographic techniques, and the
major isomer was subsequently identified as the enamide 6a
(Scheme 2). Interestingly, the unsaturated lactam 7, resulting
from a double bond isomerization toward the opposite side of
the bond, was not observed at all, which seemed to reveal an
intriguing influence of the CF2 group in the regioselectivity of
the process.
The formation of lactam 6a, which arises from an isomer-
ization reaction following the RCM, prompted us to optimize
the reaction conditions in order to achieve the regioselective
formation of either amides 5 or 6. We first studied the influence
of the solvent, carrying out the reaction at room temperature in
the presence of catalyst 2 (10 mol %). In solvents such as
dichloromethane, tetrahydrofuran, and chloroform the reaction
was very fast. In fact, it was completed in only 10 min at room
temperature, with the expected RCM product 5 as the only
identified (with the aid of GC-MS) and isolated reaction
product. Similar results were obtained with carbon tetrachloride
and trifluorotoluene. In these cases, however, longer reaction
times of up to 2 h were required to reach complete conversion.
Interestingly, a different situation occurred when toluene was
CH2Ph
p-MeOC6H4
CH2CO2Et
c-C6H11
p-FC6H4
o-MeOC6H4
10
a Method A: refluxing dichloromethane (2 × 10-2 M); 1 (5-10 mol
%); 5-6 h. Method B: refluxing toluene (2 × 10-2 M); 2 (5-10 mol %),
1-2 h.
used as solvent; after 2 h, a nearly 1:1 mixture of both the
metathesis and the isomerized products was obtained.21
We next turned our attention to the combined influence of
temperature and catalyst. After testing several combinations of
temperatures, catalysts, and solvents, we found that when the
reaction was carried out in the presence of Grubbs’ catalyst 1
(5-10 mol %)20 in refluxing dichloromethane, ꢀ-lactams 5 were
formed exclusively as a result of a direct RCM reaction with
no significant concurrent isomerization (Table 1, Method A,
entries 1-4). In sharp contrast, when catalyst 2 (5-10 mol %)20
was used in refluxing toluene, amides 4a-f underwent the
tandem RCM-isomerization sequence to furnish the corre-
sponding seven-membered enamides 6a-f as the exclusive
reaction products in good to excellent yields (Table 1, Method
B, entries 5-10). We also found that the overall process could
be carried out in two steps. For example, when lactam 5a,
prepared with Method A, was heated in toluene with catalyst
2, an isomerization reaction took place smoothly, affording
lactam 6a in 95% yield, with no need for any additives to
promote the isomerization. Moreover, in the case of these seven-
membered rings the RCM proceeds much faster than the
isomerization reaction, thus preventing the formation of ring-
contracted products that would arise if the RCM reaction took
place on the isomerized olefin.22,23
(21) The different behavior of toluene as solvent had already been
observed by Fu¨rstner in some RCM processes. He hypothesized that
competing interactions between the mesityl group and aromatic solvents
might reduce the stabilizing effect of the intramolecular π-π stacking with
the benzylidene substituent; as a result, the catalyst would have more
conformational freedom and therefore could display a different reactivity.
See ref 6.
(22) Double bond isomerization prior to the RCM has been observed in
several reports. See ref 10.
(23) The preparation of nonfluorinated analogues to 6 has been described
through a tandem isomerization-RCM of enynes, the inverse process of
that described here. See ref 10d.
(19) (a) Fustero, S.; Piera, J.; Sanz-Cervera, J. F.; Catala´n, S.; Ram´ırez
de Arellano, C. Org. Lett. 2004, 6, 1417-1420. (b) Fustero, S.; Bartolome´,
A.; Sanz-Cervera, J. F.; Sa´nchez-Rosello´, M.; Soler, J. G.; Ram´ırez de
Arellano, C.; Simo´n-Fuentes, A. Org. Lett. 2003, 5, 2523-2526. (c) Fustero,
S.; Salavert, E.; Sanz-Cervera, J. F.; Piera, J.; Asensio, A. Chem. Commun.
2003, 844-845.
(20) See Supporting Information for experimental details.
2708 J. Org. Chem., Vol. 71, No. 7, 2006

RegioselectiVe Preparation of Unsaturated Lactams
SCHEME 3. Preparation of Seven-Membered Lactams 9,
10, and 12
answering these questions, we considered that the relatively high
temperature used in the process might have led to the generation
of a ruthenium hydride species from catalyst 2, thereby causing
the formation of Ru hydride, which could in turn be responsible
of the observed isomerization.26 We thus proposed the mech-
anism shown in Scheme 4 as a plausible explanation for our
results.
To test this hypothesis and also to answer our other two
questions, we performed a monitored experiment and found that
when amide 4a was heated in an NMR tube in deuterated
toluene in the presence of catalyst 2, after 10 min at 80 °C a
1:1 mixture of amides 5a and 6a were the only products
1
observed in the H NMR spectrum. Since compound 6a is the
only product that was isolated after 2 h of heating under these
conditions (Table 2, entry 5), this indicates that the cyclization
of dienic amides 4 occurs rapidly, with lactam 5 being formed
first. The ruthenium hydride generated in the reaction conditions
could then add to the olefinic moiety of lactam 5, giving rise to
the isomeric intermediates A and B (Scheme 4). â-Elimination
of the appropriate hydride would lead to the final isomerized
amides. Intermediates A and B each contain two different kinds
of suitable hydrogen atoms that can undergo the elimination
process (H2 and H4 for A; H1 and H3 for B). The elimination of
hydrogen H2 or H3 on intermediates A and B would cause the
formation of transition states (TSs) E and F, respectively, which
in turn would again revert to the starting lactam 5 (Scheme 4).
In contrast, the elimination of hydrogen H4 on intermediate A
would lead to TS C, the formation of which is favored as a
result of the stabilization of the positive partial charge in the R
carbon to the nitrogen, which in turn is caused by the lone
nitrogen electron pair. Finally, the elimination of hydrogen H1
on intermediate B would cause the formation of TS D, which
would be destabilized due to the fact that the partial positive
charge on the C atom contiguous to the gem-difluoro moiety
should be destabilized by its electron withdrawing effect. Thus,
the formation of TS D would be disfavored in comparison to
that of TS C. Since TSs C and D would lead to the formation
of lactams 6 and 7, respectively, this mechanism, which takes
into account the destabilizing influence of the gem-difluoro
group on TS D as compared to the stabilizing effect of the N
atom on TS C, would explain why lactams 7 are not observed
as reaction products. Finally, the amide 5 formed from TSs E
and F would once again be subjected to the catalytic cycle,
forming at the end of the process the most thermodynamically
stable product 6. The formation of this compound is therefore
favored by the high temperature used in the reaction. At this
point we decided to carry out ab initio optimization of the
conformers of compounds 5a, 6a, and 7 at HF/6-31G* level of
theory. Our results showed that compound 6a is 1.8 kcal‚mol-1
more stable than 5a, which at first glance seems to point to a
thermodinamic control of the reaction product. However, our
calculations also showed that compound 7 has a very similar
energy compared to compound 5a (7 is 1.7 kcal‚mol-1 less
stable than 6a and 0.1 kcal‚mol-1 more stable than its isomer
5a). Despite this similarity in energy content, however, 7 was
never detected in our reactions, even though we carefully
monitored their evolution by means of GC-MS and NMR
techniques. This fact seems to point out that the absence of 7
We next turned our attention to the preparation of the
corresponding nonfluorinated lactam derivatives not only to
extend the scope of this RCM-isomerization protocol24 but also
to compare their behavior with that of their fluorinated
counterparts under similar reaction conditions. Thus, when we
applied our strategy (Method B) to the nonfluorinated amide 8,
we observed the formation of the expected lactam 9 in 67%
yield along with another isomeric lactam 10 (29%), which was
easily separated by means of flash chromatography. This latter
compound arises from the double bond isomerization toward
the opposite side (Scheme 3). It is also interesting to note that
the more stable R,â-unsaturated lactam 11 was not detected in
the crude mixture. As we had done with their fluorinated
counterparts, we also performed this synthesis using a two-step
sequence to check the results of the olefin isomerization under
the conditions set out by us. Thus, the nonisomerized RCM
product 12 was easily prepared as described by Guibe´ and co-
workers.25 Treatment of lactam 12 with catalyst 2 in refluxing
toluene afforded a mixture of lactams 9 and 10 in a 2:1 ratio
with a yield of 79%. Similar results were also obtained with
lactam 12 when the ruthenium hydride complex [RuHCl(CO)-
(PPh3)3], a typical isomerization catalyst, was used instead of
the second generation catalyst 2, giving in this case lactams 9
and 10 in a 3:1 ratio with 80% yield. This result seems to support
our hypothesis that an Ru hydride species resulting from the
heating of catalyst 2 causes the isomerization process, as
discussed above.
These results led us to pose three questions concerning the
overall process, namely, (i) what the possible mechanism of
the tandem protocol is, (ii) what role the gem-difluoro moiety
plays in the observed regioselectivity of the isomerization
process, and (iii) how the isomeric lactams are formed. In
(24) The synthesis of nonfluorinated cyclic enamides via copper-catalyzed
intramolecular vinylation of amides was described very recently. However,
this method does not afford rings with more than seven members, and it is
greatly affected by the basicity of the substituent on the nitrogen atom.
Our method thus circumvents the problems of this approach. See: Hu, T.;
Li, C. Org. Lett. 2005, 7, 2035-2038.
(26) Ru-hydride species are known to catalyze olefin isomerization: (a)
Wakamatsu, H.; Nishida, M.; Adachi, N.; Mori, M. J. Org. Chem 2000,
65, 3966-3970 and references therein. (b) Krompiec, S.; Pigulla, M.;
Szczepankiewicz, W, Bieg, T.; Kudnik, N.; Leszczynska-Sejda, K., Kubicki,
M.; Borowiak, T. Tetrahedron Lett. 2001, 7095-7098.
(25) Vo-Thanh, G.; Voucard, V.; Sauriat-Dorizon, H.; Guibe´, G. Synlett
2001, 37-40.
J. Org. Chem, Vol. 71, No. 7, 2006 2709

Fustero et al.
SCHEME 4. Possible Mechanism for Formation of Isomerized Lactams
TABLE 2. Further Tandem RCM-olefin Isomerization Results
6a is also the most stable of the three possible isomers. This
result would thus agree with our proposed mechanism for the
observed selectivity in the formation of 6a.27
This proposed mechanism would also explain why nonflu-
orinated amides 8 afford a mixture of isomerized lactams 9 and
10 when treated with catalyst 2 in hot toluene. Thus, amides 8
would first cyclize to lactam 12, the reaction of which with a
ruthenium hydride gives rise to the corresponding intermediates
to A and B. In this case, the absence of the gem-difluoro moiety
would cause TS D not to be destabilized; the result would thus
be the observed mixture of lactams 9 and 10. Since the presence
of the nitrogen atom still favors TS C over D, this could explain
why amide 9 is the major reaction product, as it arises from the
former TS. This mechanism also offers a possible explanation
of why the R,â-unsaturated amides 11 are never formed, as the
electron-withdrawing effect of the carbonyl group would
disfavor the corresponding transition state with a partial positive
charge on the contiguous carbon atom (Scheme 4).
In summary, and remembering that the isomeric lactams 7
(Scheme 2) and 11 (Scheme 3) were never detected in our
reactions, we can conclude that both the CF2 group and the
carbonyl exert a decisive influence on the regioselectivity of
the process. In this case, the CF2 group seems to behave like a
protected carbonyl group, thus hindering the formation of 7 in
a manner similar to that in which the formation of nonfluorinated
derivatives such as 11 are also hindered.28
We then extended our study to the preparation and isomer-
ization of lactams with different ring sizes. Thus, when amide
13 was subjected to the tandem sequence (Method B), it afforded
the isomerized δ-lactam 14 exclusively and in good yield
(Scheme 5). Once again, no R,â-unsaturated δ-lactam 15 was
detected, probably because of the electron withdrawing effect
of the carbonyl group.
We decided to prepare the corresponding difluorinated
analogue 17 by means of a strategy different from that used for
the ꢀ-lactams 5a and 6a because of the ease with which
a Yields are given in parentheses.
might be due to an electronic effect caused by the presence of
the CF2 group, and hence the lack of formation of 7 must be
caused by kinetic control, even though the isomerization product
(27) A theoretical study of the mechanism of these reactions is currently
underway in our laboratory and will be published in due course.
2710 J. Org. Chem., Vol. 71, No. 7, 2006

RegioselectiVe Preparation of Unsaturated Lactams
SCHEME 5. Preparation of Six-Membered Lactams 14 and
17
SCHEME 6. Preparation of Eight-Membered Fluorinated
Lactams 19 and 21
SCHEME 7. Preparation of Eight-Membered Lactams 23,
24, and 25
precursor 4a can be obtained. Thus, treatment of amide 4a with
the ruthenium hydride catalyst [RuHCl(CO)(PPh3)3] regiose-
lectively afforded, as expected, its enamidic isomer 16 in 71%
yield, which was then cyclized under standard RCM conditions
to yield exclusively the desired δ-lactam 17 in 66% yield
(Scheme 5).
We also tested our tandem protocol on the preparation of
eight- and nine-membered cyclic enamides (Schemes 6 and 7).
Although the RCM-isomerization sequence was efficient in
the preparation of the eight-membered lactams 19 (65%) and
23 (57%), the yields were somewhat lower than those of the
previously obtained lactams. Additionally, in contrast with the
formation of the six- and seven-membered rings described
above, we also detected the formation of variable amounts of
other lactam derivatives in these cases by means of GC-MS
analysis. In the case of the eight-membered fluorinated lactams,
the main product 19 was accompanied by small amounts of two
other compounds, namely, the isomeric ς-lactam 20 (<5%) and
the seven-membered lactam 6a (6%) (Scheme 6). Compound
6a, which was separated easily by means of flash chromatog-
raphy, results from a competitive tandem isomerization-RCM
sequence.
amounts of other products identified as isomeric eight- (24,
11%), seven- (9 + 10, 22%), and six-membered lactams (14,
4%)29 were also formed (Scheme 6). In this sense, the two-step
sequence constitutes a useful alternative for the preparation of
amides 19 and 23. Thus, when difluorinated amide 18 was
treated with catalyst 1 (10 mol %) (Method A), lactam 21 was
isolated as the only reaction product. When, in turn, compound
21 was heated in toluene in the presence of catalyst 2, it afforded
a 9:1 mixture of lactams 19 and 6a, respectively, in 68% overall
yield. However, the best results for the preparation of lactam
19 from its isomer 21 were obtained when the ruthenium hydride
catalyst [RuHCl(CO)(PPh3)3 (10 mol %)] was used, with lactam
19 as the only reaction product being formed in 80% yield
(Scheme 6).
Although the yield of the fluorinated lactam 19 is still
synthetically useful, the results obtained for its nonfluorinated
counterpart were less satisfactory. Thus, when our RCM-
isomerization protocol was applied to compound 22, GC-MS
analysis indicated that besides the cyclic enamide 23, significant
In the same way, lactam 25 was obtained through treatment
of amide 22 with first generation Grubbs’ catalyst 1 in 60%
yield (Method A). When compound 25 was heated in toluene
with catalyst 2, the desired isomeric lactam 23 was isolated in
56% yield, together with a mixture of other amides, which were
easily separated by means of flash column chromatography
(Scheme 7).
(28) To confirm this hypothesis, we decided to study the isomerization
process on the tertiary amide N,N-dibenzyl-2,2-difluoro-4-pentenamide. We
found that when this compound was treated with catalyst 2 in conditions
similar to those in Method B, only the corresponding homodimer resulting
from a cross metathesis was observed. Moreover, when hydride catalyst
RuHCl(CO)(PPh3)3 was used, the starting amide was recovered unchanged,
without a trace of double bond isomerization, which illustrates the reluctance
of the CF2-allyl grouping to undergo isomerization processes:
Much less interesting from a synthetic point of view were
the results obtained in the synthesis of nine-membered lactams.
(29) In this case, δ-lactam 14 would arise from the result of a double
isomerization in the nonfluorinated precursor 22. In contrast, no six-
membered lactam was detected in the case of fluorinated derivatives, which
indicates that the double bond isomerization catalyzed by ruthenium catalyst
2 does not take place on the difluoroolefinic chain.
J. Org. Chem, Vol. 71, No. 7, 2006 2711

Fustero et al.
In this case, regardless of whether the starting amides contained
the gem-difluoro moiety or not, complex mixtures of nine-,
eight-, seven-, and six-membered lactams were obtained (15%,
36%, 16%, and 3% and 0%, 26%, 23%, and 19% yields for
fluorinated and nonfluorinated derivatives, respectively, as
determined through GC-MS). Evidently, and in contrast with
smaller rings, the formation of the nine-membered ring is
disfavored, a circumstance that facilitates double bond isomer-
izations prior to the RCM, which in turn promotes the formation
of complex mixtures of lactams with smaller ring sizes.
Finally, to extend the scope of this method, we decided to
apply our tandem RCM-isomerization protocol (Method B) to
some compounds that had previously been obtained by other
authors through standard RCM methods (Table 2). Thus, the
tandem protocol was tested first on amide (()-26 (entry 1),30
which led to the isolation of the protected Freidinger lactam
(()-27 in a regioselective fashion, albeit in moderate yield
(51%).
The next two examples (entries 2 and 3) illustrate the different
behavior of two N,N-diallylamines. Thus, while diallyltosyla-
mide 28 provided the expected isomeric N-tosyl-2,3-dihydro-
pyrrole 2910b in good yield when subjected to Method B, the
phenylalanine methyl ester derivative 30 did not afford the
expected derivative 31.31
The following two entries in Table 2 (entries 4 and 5) show
the results for the application of our metathesis-isomerization
sequence to the synthesis of seven-membered cyclic enol ethers
from acyclic allyl ethers. Cyclic enol ether (()-33 had already
been synthesized by Snapper and co-workers in 54% yield using
their tandem RCM-isomerization protocol with catalyst 2 in
refluxing CH2Cl2 under a 95:5 N2/H2 atmosphere.11a With our
protocol, the same compound (()-33 was obtained in 60% yield
from (()-32 simply by reacting it with ruthenium catalyst 2 in
refluxing toluene. Once again, the reaction conditions of our
Method B probably cause the generation of a ruthenium hydride
from catalyst 2, thus allowing the tandem RCM-isomerization
reactions to be performed conveniently and in the absence of
additives.32
When our RCM-isomerization protocol was applied to the
dienic system 34 (entry 5), a mixture of two isomerization
products 35 and 36 resulted. The fact that the enol ether 35 is
the major product proves once again the decisive influence of
the heteroatom on the isomerization step.
The last two examples involve the synthesis of fluorinated
compounds (R)-38 and 40. Our group had previously reported
on the use of substrate (R)-37 (entry 6) as starting material for
the preparation of bicyclic difluorinated oxazolidinones by
means of RCM with catalyst 1.33 Here, in contrast, when we
refluxed a toluene solution of (R)-37 in the presence of catalyst
2, a regioselective double bond isomerization took place after
the RCM, thus affording compound (R)-38 in excellent yield
(85%).
Finally, we studied the application of our protocol to the
preparation of CF3-substituted piperidines. In this context,
Bonnet-Delpon and co-workers very recently published a
straightforward route to this class of compounds through
metathesis cyclization of unsaturated trifluoromethylamines with
the first generation catalyst 1.34 As in previous examples, both
the metathesis product (arising from the regular RCM reaction
with catalyst 1) and the acyclic diene 39 were easily transformed
into the final isomerized tetrahydropiperidines 40 in good yield
through treatment with catalyst 2 in refluxing toluene35 (Table
2, entry 7).
The outcome of all of the reactions depicted in Table 2 can
be easily explained, since the result of the isomerization process
is conditioned either by the presence of the gem-difluoro moiety,
in which case the product of the isomerization toward the
opposite side of the double bond is formed exclusively, or by
the heteroatom, which favors the isomerization toward the
heteroatom itself. When both structural features are present
simultaneously, they influence the isomerization reaction in the
same direction, thus making these processes highly regioselec-
tive.
Conclusion
In conclusion, we have developed several new examples of
our tandem RCM-isomerization protocol that prove its useful-
ness in the synthesis of isomerized lactams of various ring sizes.
The process is efficient in the preparation of five- to eight-
membered rings, in which the RCM step is very fast because
of the favorable ring size formation and thus occurs well before
the isomerization. The only limitation to the method involves
the formation of nine-membered rings. In this case, the less
favored RCM step occurs more slowly, allowing the isomer-
ization processes to take place first, which leads to the formation
of mixtures of products with several ring sizes. Still, our results
lead us to the conclusion that the second-generation Grubbs’
catalyst 2 in refluxing toluene is likely to generate a ruthenium
hydride species in situ, without the need for additives, which
(30) Hoffmann, T.; Waibel, R.; Gmeiner, P. J. Org. Chem. 2003, 68,
62-69.
(31) In this case, when substrate 30 was subjected to the tandem protocol
conditions [Ti(OiPr)4 must be used as additive in this case because of the
presence of a free amine group], pyrrol 41 was the only product isolated,
arising from the aromatization of the isomerized product in the workup
procedure. The same results were obtained when we started from the 2,5-
dihydropyrrol 42 previously synthesized by RCM in a two-step sequence
as described by Yang et al. See: Yang, Q.; Xiao, W.-J.; Yu, Z. Org. Lett.
2005, 7, 871-874. It is assumed that the basicity of the nitrogen atom plays
an important role in the nature of the final product. The presence of an
electron-withdrawing group in the nitrogen atom prevents the dehydroge-
nation step towards the corresponding pyrrole.
(33) Fustero, S.; Navarro, A.; Pina, B.; Garc´ıa Soler, J.; Bartolome´, A.;
Asensio, A.; Simo´n, A.; Bravo, P.; Fronza, G.; Volonterio, A.; Zanda, M.
Org. Lett. 2001, 3, 2621-2624.
(34) Magueur, G.; Legros, J.; Meyer, F.; Oure´vitch, M.; Crousse, B.;
Bonnet-Delpon, D. Eur. J. Org. Chem. 2005, 1258-1265.
(35) Similar results were also obtained in a two-step sequence in which
the initially formed, nonisomerized metathesis product was treated with
catalyst 2 in refluxing toluene:
(32) In contrast, when we treated difluorinated compound 4a in refluxing
toluene as in Schmidt’s protocol (NaOH/i-PrOH),11b only 30% of the
isomerization product 6a was observed even after continuous heating of
the reaction flask for 1 week.
2712 J. Org. Chem., Vol. 71, No. 7, 2006

RegioselectiVe Preparation of Unsaturated Lactams
(m, 5H); 13C NMR (CDCl3, 75.5 MHz) δ 34.7 (t, 2JCF ) 26.5 Hz),
45.5 (t, 4JCF ) 4.9 Hz), 52.5 (t), 116.5 (t, 1JCF ) 247.8 Hz), 124.1
(t, 3JCF ) 6.0 Hz), 125.2 (d), 127.8 (d), 127.8 (d), 128.7 (d), 136.0
may, in turn, be responsible for the isomerization reaction. We
have also demonstrated the crucial influence exerted by the gem-
difluoro moiety in directing the isomerization step, making the
tandem RCM-isomerization a regioselective and hence syn-
thetically useful process.
2
(s), 164.4 (t, JCF ) 28.8 Hz); 19F NMR (CDCl3, 282.4 MHz) δ
-101.3 (t, JFH ) 16.0 Hz, 2F); HRMS (EI+) calcd for C13H13F2-
NO (M+) 237.0965, found 237.0956. Anal. Calcd for C13H13F2-
NO: C, 65.81; H, 5.52; N, 5.90. Found: C, 65.83; H, 5.52; N,
5.80.
Experimental Section
General Procedure for Preparation of Cyclic Enamides 6.
Method B. This method is analogous to Method A, except that
here catalyst 2 is used and the solvent is toluene under reflux. With
this catalyst (5 mol % for 6a-c; 10% mol for 6d-f), the reactions
took 2-3 h to complete.
General Procedure for Preparation of Amides 4. A catalytic
amount of dimethylformamide (DMF) (0.1 mL) was added to a
solution of 2,2-difluoro-4-pentenoic acid (5.15 mmol) in dry
dichloromethane (17 mL). The solution was cooled with an ice bath,
after which oxalyl chloride (2.0 M in DCM, 5.15 mmol) was added
dropwise. The resulting brown solution was stirred for 1 h at room
temperature. The reaction was then cooled again to 0 °C, after which
Et3N (10.3 mmol), the corresponding primary amine (5.7 mmol),
and a catalytic amount of DMAP were added simultaneously. Upon
completion of the addition, the ice bath was removed and the
reaction mixture was stirred for 5-6 h at room temperature. The
crude mixture was then diluted with dichloromethane (20 mL) and
washed with saturated ammonium chloride (3 × 15 mL). The
organic layer was dried over anhydrous sodium sulfate and filtered,
and the solvents were removed under reduced pressure. The
resulting brown oil was purified by means of flash chromatography
to afford the corresponding secondary amide. A solution of the
secondary amide in dry DMF (13 mL) was then added to a
suspension of NaH (3.73 mmol) in dry DMF (4 mL) at 0 °C. After
the resulting solution stirred for 15 min, allyl bromide (9.3 mmol)
was added dropwise. The ice bath was then removed, and the
reaction mixture was stirred at room temperature until TLC analysis
indicated that the starting material was no longer present (3-4 h).
The DMF was removed under reduced pressure, and the resulting
yellow residue was treated with water (20 mL) and extracted with
ethyl acetate (3 × 20 mL). The organic layer was dried over
anhydrous sodium sulfate and filtered, and the solvents were
removed under reduced pressure. After purification of the residue
by means of flash chromatography, amides 4 were obtained.
N-Allyl-N-benzyl-2,2-difluoro-4-pentenamide (4a). Flash chro-
matography [n-hexanes-EtOAc (10:1)] (Rf ) 0.40) afforded 4a
1-Benzyl-3,3-difluoro-1,3,4,5-tetrahydro-2H-2-azepinone (6a).
Flash chromatography [n-hexanes-EtOAc (7:1)] (Rf ) 0.20)
1
afforded 6a as a brown oil (73% yield). H NMR (CDCl3, 300
MHz) δ 2.13-2.20 (m, 2H), 2.39-2.54 (m, 2H), 4.64 (s, 2H), 5.38
(dt, J ) 8.6, 6.2 Hz, 1H), 5.85 (dt, J ) 8.9, 1.2 Hz, 1H), 7.21-
3
7.23 (m, 5H); 13C NMR (CDCl3, 75.5 MHz) δ 20.3 (t, JCF ) 5.6
Hz), 38.4 (t, 2JCF ) 24.2 Hz), 51.5 (t), 116.7 (t, 1JCF ) 249.3 Hz),
117.5 (d), 127.8 (d), 128.0 (d), 128.7 (d), 128.9 (d), 136.0 (s), 164.4
(t, 2JCF ) 29.6 Hz); 19F NMR (CDCl3, 282.4 MHz) δ -96.9 (t, JFH
) 15.5 Hz, 2F); HRMS (EI+) calcd for C13H13F2NO (M+) 237.0965,
found 237.0962. Anal. Calcd for C13H13F2NO: C, 65.81; H, 5.52;
N, 5.90. Found: C, 65.72; H, 5.60; N, 5.99.
Synthesis of Seven-Membered, Nonfluorinated Lactams 9 and
10. These were prepared from 8 as starting material by means of
the general procedure described above for compounds 6 (Method
B). Flash chromatography [n-hexanes-EtOAc (5:1)] afforded
1-benzyl-2,3,4,5-tetrahydro-1H-2-azepinone (9) (Rf ) 0.25) (67%
yield) and 1-benzyl-2,3,6,7-tetrahydro-1H-2-azepinone (10) (Rf )
0.10) (29% yield), both as brown oils. Data for 1-benzyl-2,3,4,5-
1
tetrahydro-1H-2-azepinone (9): H NMR (CDCl3, 300 MHz) δ
1.97-2.10 (m, 4H), 2.52-2.56 (m, 2H), 4.57 (s, 2H), 5.27 (dt, J
) 9.0, 5.6 Hz, 1H), 5.83 (dt, J ) 9.0, 1.4 Hz, 1H), 7.17-7.24 (m,
5H); 13C NMR (CDCl3, 75.5 MHz) δ 26.1 (t), 26.5 (t), 35.9 (t),
50.4 (t), 117.8 (d), 127.3 (d), 127.8 (d), 128.5 (d), 129.6 (d), 137.6
(s), 174.1 (s); HRMS (EI+) calcd for C13H15NO (M+) 201.1154,
found 201.1127. Data for 1-benzyl-2,3,6,7-tetrahydro-1H-2-
1
azepinone (10): H NMR (CDCl3, 300 MHz) δ 2.03-2.07 (m,
1
as a pale yellow oil (60% overall yield). H, 13C, and 19F NMR
2H), 3.25-3.26 (m, 2H), 3.42-3.47 (m, 2H), 4.56 (s, 2H), 5.55-
5.56 (m, 2H), 7.19-7.27 (m, 5H); 13C NMR (CDCl3, 75.5 MHz)
δ 28.4 (t), 35.9 (t), 45.3 (t), 49.7 (t), 120.9 (d), 128.4 (d), 128.0
(d), 128.6 (d), 129.0 (d), 137.8 (s), 173.6 (s); HRMS (EI+) calcd
for C13H15NO (M+) 201.1154, found 201.1129.
showed the presence of two rotamers around the amide bond in a
1
1.4:1 ratio. H NMR (CDCl3, 300 MHz) δ 2.84-2.98 (m, 2H),
3.80 (d, J ) 5.8 Hz, 1H, minor rotamer), 3.95 (d, J ) 5.8 Hz, 1H,
major rotamer), 4.51 (s, 1H, major rotamer), 4.65 (s, 1H, minor
rotamer), 4.99-5.24 (m, 4H), 5.60-5.86 (m, 2H), 7.13-7.28 (m,
Synthesis of Six-Membered Nonfluorinated and Fluorinated
Lactams 14 and 17. 1-Benzyl-1,2,3,4-tetrahydro-2-pyridinone
(14). Prepared from 13 with the general procedure described above
for compounds 6 (Method B). Flash chromatography [n-hexanes-
EtOAc (2:1)] (Rf ) 0.15) afforded 14 as a brown oil (82% yield).
Its spectroscopic data matched those previously reported.36
N-Benzyl-2,2-difluoro-N-[(E/Z)-1-propenyl]-4-pentenamide (16).
A solution of ruthenium hydride catalyst [RuHCl(CO)(PPh3)3] (10
mol %) in dry toluene was added via cannula to a solution of 4a
(0.3 mmol) in toluene (10 mL). The resulting dark brown solution
was heated at reflux for 3 h. The solvents were removed under
reduced pressure, and the residue was purified by means of flash
chromatography [n-hexanes-EtOAc (10:1)] (Rf ) 0.40), affording
amide 16 as a complex mixture of amide rotamers and E/Z isomers
(71% yield). The compound 16 thus obtained was used immediately
in the subsequent cyclization step.
2
5H); 13C NMR (CDCl3, 75.5 MHz) δ 39.5 (t, JCF ) 23.9 Hz),
4
4
47.9 (t), 48.2 (t), 48.9 (t, JCF ) 6.3 Hz), 49.7 (t, JCF ) 6.3 Hz),
1
118.1 (t), 118.5 (t, JCF ) 255.9 Hz), 118.7 (t), 121.1 (t); 127.4
(d), 127.6 (d), 127.7 (d), 128.0 (d), 128.3 (t, 3JCF ) 5.2 Hz), 128.7
2
(d), 131.3 (d), 132.8 (d), 135.9 (s), 136.13 (s), 163.2 (t, JCF
)
29.0 Hz), 163.4 (t, 2JCF ) 29.3 Hz); 19F NMR (CDCl3, 282.4 MHz)
δ -98.1 (t, JFH ) 18.0 Hz, 2F, minor rotamer), -99.1 (t, JFH
)
18.0 Hz, 2F, major rotamer); HRMS (EI+) calcd for C15H17F2NO
(M+) 265.1278, found 265.1286.
General Procedure for Preparation of E-Lactams 5. Method
A. A solution of ruthenium catalyst 1 (5 mol % for 5a-c; 10 mol
% for 5d) in dry dichloromethane was added via cannula to a
solution of amides 4 (0.8 mmol) in dichloromethane (2 × 10-2
M). The resulting dark brown solution was heated at reflux until
TLC indicated that the starting material was no longer present
(usually after 5-6 h). The solvents were removed under reduced
pressure, and the residue was purified by means of flash chroma-
tography.
1-Benzyl-3,3-difluoro-1,2,3,4-tetrahydro-2-pyridinone (17).
Prepared from 16 with the general procedure described above for
compounds 6 (Method B) except that here catalyst 2 was used in
refluxing dichloromethane. Flash chromatography [n-hexanes-
EtOAc (5:1)] (Rf ) 0.20) afforded 17 as a white solid (66% yield).
1-Benzyl-3,3-difluoro-1,3,4,7-tetrahydro-2H-2-azepinone (5a).
Flash chromatography [n-hexanes-EtOAc (7:1)] (Rf ) 0.10)
1
afforded 5a as a pale brown oil (95% yield). H NMR (CDCl3,
300 MHz) δ 2.80 (tq, J ) 16.1, 2.1 Hz, 2H), 3.82-3.84 (m, 2H),
4.63 (s, 2H), 5.53-5.58 (m, 1H), 5.62-5.68 (m, 1H), 7.15-7.27
(36) Ojima, I.; Korda, A.; Shay, W. R. J. Org. Chem. 1991, 56, 2024-
2030.
J. Org. Chem, Vol. 71, No. 7, 2006 2713

Fustero et al.
3
2
Mp 72-74 °C. 1H NMR (CDCl3, 300 MHz) δ 2.82 (tq, J ) 17.1,
2.0 Hz, 2H), 4.69 (s, 2H), 5.03-5.09 (m, 1H), 5.99 (dt, J ) 7.9,
1.7 Hz, 1H), 7.18-7.29 (m, 5H); 13C NMR (CDCl3, 75.5 MHz) δ
NMR (CDCl3, 75.5 MHz) δ 25.8 (t, JCF ) 4.9 Hz), 34.1 (t, JCF
) 23.3 Hz), 48.0 (t), 48.2 (t), 49.0 (t, 4JCF ) 6.3 Hz), 49.8 (t, 4JCF
1
) 6.3 Hz), 115.4 (t), 118.1 (t), 118.7 (t), 119.5 (t, JCF ) 255.0
2
3
31.8 (t, JCF ) 25.3 Hz), 49.6 (t), 101.8 (t, JCF ) 5.2 Hz), 112.1
Hz), 127.4 (t), 127.6 (d), 127.7 (d), 128.0 (d), 128.7 (d), 128.7 (d),
1
2
(t, JCF ) 246.7 Hz), 127.8 (d), 128.1 (d), 128.3 (d), 128.9 (d),
131.4 (d), 133.0 (d), 136.0 (s), 136.2 (s), 136.7 (d), 163.4 (t, JCF
135.5 (s), 160.2 (t, 2JCF ) 29.9 Hz); 19F NMR (CDCl3, 282.4 MHz)
δ -105.1 (dt, JFH ) 17.2, 2.6 Hz, 2F); HRMS (EI+) calcd for
C12H11F2NO (M+) 223.0809, found 223.1077. Anal. Calcd for
C12H11F2NO: C, 64.57; H, 4.97; N, 6.27. Found: C, 64.44; H, 4.79;
N, 6.06.
) 29.3 Hz), 163.6 (t, JCF ) 29.3 Hz); 19F NMR (CDCl3, 282.4
2
MHz) δ -109.1 (t, JFH ) 17.3 Hz, 2F, minor rotamer), -110.0 (t,
JFH ) 17.3 Hz, 2F, major rotamer); HRMS (EI+) calcd for C16H19F2-
NO (M+) 279.1435, found 279.1467.
Synthesis of Eight-Membered Fluorinated and Nonfluori-
nated Lactams 19, 21, and 23. 1-Benzyl-3,3-difluoro-1,2,3,4,5,6-
hexahydro-2-azocinone (19). Prepared from 18 with the general
procedure described above for compounds 6 (Method B). GC-
MS analysis showed a mixture of compounds 19, 20, and 6a in a
13:1:1.2 ratio. Flash chromatography [n-hexanes-EtOAc (5:1)] (Rf
2,2-Difluoro-5-hexenoic Acid. A 1 M solution of the Grignard
derivative from either 4-bromo-1-butene in THF (13 mmol) was
added to a solution of diethyl oxalate (11.8 mmol) in THF (8 mL)
at -78 °C. After stirring for 1 h, the reaction mixture was quenched
with saturated aqueous NH4Cl, and the aqueous layer was extracted
with ethyl acetate (3 × 20 mL). The combined organic layers were
washed with saturated aqueous NaCl (3 × 20 mL) and dried over
anhydrous Na2SO4. Filtration and evaporation of solvents gave the
corresponding crude R-keto ester, which was purified by means of
flash chromatography.
1
) 0.30) afforded 19 as a yellow oil (65% yield). In the H NMR
spectrum some signals appear broadened because of conformational
equilibrium. 1H NMR (CDCl3, 300 MHz) δ 1.19-2.03 (br m, 6H),
4.35 (br s, 1H), 4.78 (br s, 1H), 5.35 (q, J ) 8.2 Hz, 1H), 5.89 (d,
J ) 7.5 Hz, 1H), 7.22-7.27 (m, 5H); 13C NMR (CDCl3, 75.5 MHz)
Next, Deoxofluor (10.2 mmol) and ethanol as catalyst (1.2 mmol)
were added to a solution of the R-keto ester (6 mmol) in dry
dichloromethane (19 mL) at 0 °C. The reaction mixture was stirred
for 14 h and then quenched with saturated aqueous NaHCO3 and
extracted with dichloromethane (3 × 20 mL). The combined organic
layers were washed with aqueous HCl (3 × 15 mL) and brine (3
× 15 mL), dried over anhydrous sodium sulfate, and filtered, after
which the volatiles were removed under reduced pressure to afford
the intermediate difluorinated ester. This was, in turn, dissolved in
dry THF (18 mL) and treated with water (5 mL) and LiOH‚H2O
(18 mmol) in an ice bath. After stirring for 5 h, the reaction mixture
was acidified with an aqueous solution of HCl and extracted with
ethyl acetate (3 × 20 mL). The organic layers were washed with
brine (3 × 20 mL), dried over anhydrous Na2SO4, and filtered,
after which the filtrate was distilled under reduced pressure to afford
the pure carboxylic acid in 85% yield. Bp 30-33 °C (10-2 Torr);
1H NMR (300 MHz, CDCl3) δ 2.08-2.24 (m, 4H), 4.96-5.06 (m,
2H), 5.69-5.78 (m, 1H), 10.22 (br s, 1H); 13C NMR (75.5 MHz,
3
2
δ 20.7 (t, JCF ) 4.6 Hz), 23.6 (t), 33.0 (t, JCF ) 25.3 Hz), 51.7
(t), 119.6 (t, 1JCF ) 251.7 Hz), 125.5 (t, 4JCF ) 2.0 Hz), 126.7 (d),
2
127.8 (d), 128.5 (d), 128.9 (d), 135.5 (s), 164.6 (t, JCF ) 27.9
Hz); 19F NMR (CDCl3, 282.4 MHz) δ -112.2 (dd, JFF) 245.7
Hz, JHF ) 13.8 Hz, 1F), -99.6 (d, JFF ) 240.5 Hz, 1F); HRMS
(EI+) calcd for C14H15F2NO (M+) 251.1122, found 251.1129.
1-Benzyl-3,3-difluoro-1,2,3,4,5,8-hexahydro-2-azocinone (21).
Prepared from 18 with the general procedure described above for
compounds 5 (Method A). Flash chromatography [n-hexanes-
EtOAc (7:1)] (Rf ) 0.15) afforded 25 as a yellow oil (70% yield).
1H NMR (CDCl3, 300 MHz) δ 2.08-2.22 (m, 2H), 2.30-2.37 (m,
2H), 3.83 (d, J ) 7.2 Hz, 2H), 4.57 (s, 2H), 5.69-5.83 (m, 2H),
7.19-7.28 (m, 5H); 13C NMR (CDCl3, 75.5 MHz) δ 19.9 (t, JCF
3
2
4
) 6.3 Hz), 35.0 (t, JCF ) 25.3 Hz), 42.5 (t, JCF ) 5.5 Hz), 51.0
(t), 118.4 (t, 1JCF ) 249.5 Hz), 127.8 (d), 128.4 (d), 128.7 (d), 129.4
2
(d), 132.8 (d), 136.3 (s), 165.5 (t, JCF ) 28.8 Hz); 19F NMR
(CDCl3, 282.4 MHz) δ -104.3 (br s, 2F); HRMS (EI+) calcd for
C14H15F2NO (M+) 251.1122, found 251.1109. Anal. Calcd for
C14H15F2NO: C, 66.92; H, 6.02; N, 5.57. Found: C, 66.74; H, 6.15;
N, 5.42.
3
2
CDCl3) δ 25.5 (t, JCF ) 4.6 Hz), 33.5 (t, JCF ) 23.0 Hz), 115.7
1
2
(t, JCF ) 250.4 Hz), 116.2 (t), 135.5 (d), 168.7 (t, JCF ) 33.6
Hz). 19F NMR (282.4 MHz, CDCl3) δ -107.9 (t, JFH ) 16.0 Hz,
2F). HRMS (EI+) calcd for C6H8F2O2 (M+) 150.0942, found
150.0452.
1-Benzyl-1,2,3,4,5,6-hexahydro-2-azocinone (23). Prepared from
22 with the general procedure described above for compounds 6
(Method B). GC-MS analysis showed the presence of compounds
23, 24, 9, 10, and 14 in a 14:3:3:2:1 ratio. Flash chromatography
[n-hexanes-EtOAc (4:1)] (Rf ) 0.20) afforded pure 23 as a brown
General Procedure for Preparation of Difluorinated Amides
18 and 22. A catalytic amount of dimethylformamide (DMF) (0.1
mL) was added to a solution of the corresponding R,R-difluoro-
ω-alkenoic acid (5.15 mmol) in dry dichloromethane (17 mL). The
solution was cooled to 0 °C with an ice bath, after which oxalyl
chloride (2.0 M in DCM, 5.15 mmol) was added dropwise. The
resulting brown solution was stirred for 1 h at room temperature.
The reaction was then cooled a second time to 0 °C, and then Et3N
(10.3 mmol), allyl benzylamine (5.7 mmol), and a catalytic amount
of DMAP were added simultaneously. After the addition was
completed, the ice bath was removed, and the reaction mixture was
stirred for 5-6 h at room temperature. The crude mixture was then
diluted with DCM (20 mL) and washed three times with saturated
ammonium chloride (2 × 15 mL). The organic layer was dried
over anhydrous sodium sulfate and filtered; the solvents were then
removed under reduced pressure. The resulting brown oil was
purified by means of flash chromatography to afford the corre-
sponding tertiary amides 18 and 22.
1
oil (57% yield). H NMR (CDCl3, 300 MHz) δ 1.40 (br s, 2H),
1.61-1.77 (m, 4H), 2.43-2.47 (m, 2H), 4.57 (s, 2H), 5.34 (q, J )
8.0 Hz, 1H), 5.87 (d, J ) 7.9 Hz, 1H), 7.19-7.24 (m, 5H); 13C
NMR (CDCl3, 75.5 MHz) δ 23.1 (t), 23.7 (t), 24.6 (t), 30.0 (t),
49.0 (t), 125.9 (d), 127.3 (d), 128.3 (d), 128.4 (d), 128.6 (d), 137.1
(s), 173.7 (s); HRMS (EI+) calcd for C14H17NO (M+) 215.1310,
found 215.1304.
Acknowledgment. The authors thank the Ministerio de
Educacio´n y Ciencia (BQU2003-01610) and the Generalitat
Valenciana (GR03-193 and GV05/079) for financial support.
M.S.R. and D.J. express their thanks for predoctoral fellowships,
and C.P. expresses his gratitude for a Ramo´n y Cajal contract.
N-Allyl-N-benzyl-2,2-difluoro-5-hexenamide (18). Flash chro-
matography [n-hexanes-EtOAc (15:1)] (Rf ) 0.30) afforded 18
Supporting Information Available: Experimental procedures
and characterization data for compounds 5c-g and 6b-e, their
precursors, and compounds 26-42, as well as the results for ab
initio energy calculations of compounds 5a, 6a, and 7. This material
is available free of charge via the Internet at http://pubs.acs.org.
1
as a colorless oil (60% yield). H, 13C, and 19F NMR showed the
presence of two rotamers around the amide bond in a 1.4:1 ratio.
1H NMR (CDCl3, 300 MHz) δ 2.24-2.28 (m, 4H), 3.81 (d, J )
6.0 Hz, 2H, minor rotamer), 3.97 (d, J ) 6.0 Hz, 2H, major
rotamer), 4.52 (s, 2H, major rotamer), 4.66 (s, 2H, minor rotamer),
4.93-5.21 (m, 4H), 5.61-5.83 (m, 2H), 7.13-7.29 (m, 5H); 13C
JO0525635
2714 J. Org. Chem., Vol. 71, No. 7, 2006