Synthesis of [2H,13C]-labelled diallylmalonates - Useful probes for the study of transition metal-catalysed 1,6-diene cycloisomerisation

Article abstract of DOI:10.1002/1099-0690(200105)2001:9<1635::aid-ejoc1635>3.0.co;2-c

Reliable synthetic routes to isotopically labelled dimethyl [¹³C,²H]diallylmalonates (1) have been developed. Dimethyl diallylmalonates readily undergo transition-metal catalysed cycloisomerisation and the isotopically labelled substrates are expected to be of general utility in the study of the mechanism of such reactions. In the synthetic routes developed, the ²H labelled allyl chains were introduced by reduction of propargyl functionality with Schwartz reagent. The regioand stereo-selectivity of such reactions allowing highly selective labelling strategies. The ¹³C labelled allyl chains were introduced via Pd-catalysed allylic alkylation of ¹³C- or ²H-labelled dimethyl allylmalonate anions using [3-¹³C1]allyl benzoate as the allylic electrophile. Using these methods, dimethyl [1,7-(E,E)-²H₂]hept-1,6-dienyl-4,4-dicarbox ylate (5), dimethyl [2,6-²H₂]hept-1,6-dienyl-4,4-dicarboxylate (6), dimethyl [1,7-(Z,Z)-²H₂]hept-1,6-dienyl-4,4-dicarboxylate (7), dimethyl [1,1,2,6,7,7-²H₆]hept-1,6-dienyl-4,4-dicarboxylate (8), dimethyl [1,3-¹³C₁,5,7-¹³C₁] hept-1,6-dienyl-4,4-dicarboxylate (9), dimethyl [1,3-¹³C₁,6,-²H₁]hept 1,6-dienyl-4,4-dicarboxylate (10), and dimethyl [1,3-¹³C₁,7-(E)-²H₁] hept-1,6-dienyl-4,4-dicarboxylate (11) have been prepared.

Full text of DOI:10.1002/1099-0690(200105)2001:9<1635::aid-ejoc1635>3.0.co;2-c

FULL PAPER
2
13
Synthesis of [ H, C]-Labelled Diallylmalonates ؊ Useful Probes for the Study
of Transition Metal-Catalysed 1,6-Diene Cycloisomerisation
[a]
[a]
Katharine L. Bray and Guy C. Lloyd-Jones*
Keywords: Allylmalonates / Isotopic labelling / Cycloisomerisation
Reliable synthetic routes to isotopically labelled dimethyl
labelled dimethyl allylmalonate anions using [3-13
benzoate as the allylic electrophile. Using these methods, di-
1
C ]allyl
1
3
2
[
C, H]diallylmalonates (1) have been developed. Dimethyl
diallylmalonates readily undergo transition-metal catalysed
cycloisomerisation and the isotopically labelled substrates dimethyl [2,6- H
2
methyl [1,7-(E,E)- H
2
]hept-1,6-dienyl-4,4-dicarboxylate (5),
2
2
]hept-1,6-dienyl-4,4-dicarboxylate (6), di-
methyl [1,7-(Z,Z)- H ]hept-1,6-dienyl-4,4-dicarboxylate (7),
2
2
are expected to be of general utility in the study of the mech-
anism of such reactions. In the synthetic routes developed,
2
6
dimethyl [1,1,2,6,7,7- H ]hept-1,6-dienyl-4,4-dicarboxylate
2
13
13
the H labelled allyl chains were introduced by reduction of
(8), dimethyl [1,3-
oxylate (9), dimethyl [1,3-13C
carboxylate (10), and dimethyl [1,3-
dienyl-4,4-dicarboxylate (11) have been prepared.
1 1
C ,5,7- C ]hept-1,6-dienyl-4,4-dicarb-
2
propargyl functionality with Schwartz reagent. The regio-
and stereo-selectivity of such reactions allowing highly se-
1 1
,6- H ]hept-1,6-dienyl-4,4-di-
13
2
1 1
C ,7-(E)- H ]hept-1,6-
1
3
lective labelling strategies. The C labelled allyl chains were
introduced via Pd-catalysed allylic alkylation of C- or 2H-
13
Introduction
cellent (94%) yield. Radetich and RajanBabu[6] have de-
veloped the use of cationic complexes ‘[(L)M(allyl)][OTf]’
Cycloisomerisation reactions of dienes, enynes, and di- (L ϭ triarylphosphane, M ϭ Pd or Ni) as effective pro-
ynes represent a powerful method for the synthesis of car- catalysts for cycloisomerisation. When M ϭ Ni, 2 (R ϭ Me)
bocyclic and heterocyclic rings with regio- and stereocon- is obtained in 92% yield and when M ϭ Pd, a mixture of 2
[
1]
trol. The first report of the transition metal catalysed (Rh) (22%) and 4 (78%) are obtained in overall 91% yield. Heu-
cycloisomerisation of 1,6-dienes was by Shaw et al. in mann and Moukhliss[ have demonstrated the potential of
7]
[
2]
[3Ϫ8]
1
971. Since then a number of other transition metal
enantiomerically pure N,N-ligands for Pd-catalysed cyclo-
[
9]
and lanthanide complexes have also been reported for this isomerisation by achieving promising enantioselectivity in
reaction. Of note is the reliable chemistry of malonates conversion of 1 (R ϭ Et) to 3 (up to 60% ee).
which has made dialkyl hept-1,6-dienyl-4,4-dicarboxylates
A knowledge of the mechanism(s) involved in cycloiso-
1
(i.e. diallylmalonates) popular substrates for the study and
merisation would obviously be essential in understanding
the selectivities induced by the various catalysts outlined
above as well as being useful in the development of new
catalysts. However, as far as we are aware, to date, only the
preliminary studies of Grigg et al.,[ have been reported.
Nonetheless, for conversion of 1 into 2 it is obvious that at
least one hydride migration is involved and that conversion
development of 1,6-diene cycloisomerisation reac-
[
3,5,6,7]
tions
(Scheme 1, R is usually Me or Et).
3]
(
direct or indirect) of 1 into 3 or 4 involves at least two.
Scheme 1. Cycloisomerisation of the 1,6-diene 1 (R ϭ Me, Et etc.)
to cyclopentenes 2, 3, and 4
To gain information about the mechanism it is therefore
imperative that the hydride migration(s) be tracked between
2
Important features in the cycloisomerisation of 1 are con- substrate and product. H-labelling is an efficient method
trol of regioselectivity (2 versus 3 versus 4) and stereochem- by which this may be achieved and thus by consideration
istry (in 2 and 3). Consequently, the discovery of novel cata- of the three basic mechanisms that have been proposed
and design of ligands that induce enantio- we devised a series of H-labelling patterns for 1 (R ϭ Me).
selectivity have attracted much recent interest. For ex- Furthermore, since the symmetry of 1 can potentially pose
ample, Itoh et al., have reported a range of Ru-based problems in determining from where and to where the hy-
catalysts which are highly effective for cycloisomerisation dride has migrated, we also designed a number of 1 C-label-
with excellent regiocontrol. Thus, by use of [Ru(COD)Cl]n ling strategies. Since cycloisomerisation of 1 to 2, 3, or 4
as pro-catalyst, 1 (R ϭ Me) is cycloisomerised in iPrOH at involves only H-migrations, 13C-labelling allows distinction
[3,5,6]
[
5,6,7]
2
lytic systems
[
7]
[
5]
3
9
0 °C to give exclusively the exo-methylene isomer 2 in ex- of one allyl chain (in 1 and products) from the other and
also allows the origin and destination of intermolecular hy-
a]
[
School of Chemistry, University of Bristol,
dride migrations to be deduced. Our initial strategies for
CantockЈs Close, Bristol BS8 1TS, UK
Fax (internat.) ϩ44Ϫ117/929 8611
the synthesis of such 13C-labelled substrates were based on
[10]
E-mail: guy.lloyd-jones@bris.ac.uk
‘routine’ malonate chemistry leading to an aldehyde. We
Eur. J. Org. Chem. 2001, 1635Ϫ1642

WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/0509Ϫ1635 $ 17.50ϩ.50/0
1635

K. L. Bray, G. C. Lloyd-Jones
FULL PAPER
had some difficulties in finding conditions for this sequence equivalents of reagent per alkyne to be reduced. By using
and furthermore, the aldehyde unexpectedly failed to un- this method and then quenching intermediate 13 with D O
2
2
dergo the Wittig chemistry planned for installation of the (99.9% D) we obtained [(E,E)- H ]-5 in 80% yield,
2
13
C label. Consequently we began to explore alternative Scheme 2.
strategies. Herein we describe these studies and report in
full the synthesis and characterisation of symmetrically la-
2
2
2
2
belled [(E,E)- H ]-5, [ H ]-6, [(Z,Z)- H ]-7, [ H ]-8, and
2
2
2
[11]
6
13
[
C ]-9 and isotopically desymmetrised
doubly labelled
2
13
2
13
2
substrates [ C , H ]-10 and [ C -(E)- H ]-11.
1
1
1
1
These labelled probes, used individually and in concert,
have now proved of utility in the study of the mechanism
2
of the Pd-catalysed conversion of 1 to 2, 3, and 4[12] and we
Scheme 2. The synthetic routes employed to generate H-labelled
dimethyl hept-1,6-dienyl-4,4-dicarboxylates 5, 6, 7, and 8
expect that they will prove of use in other systems.
Careful H, H, and 13C NMR analysis indicated that
the deuterium location was exclusively at the double bond
termini [C(1,7)] and with an (E)-geometry Ϫ thereby con-
firming that we had attained the high levels of regioselectiv-
ity and stereoselectivity required. However, the level of deu-
1
2
Results and Discussion
2
Synthesis of Symmetrical H-Labelled Substrates
2
Our strategy for the incorporation of H-labels in the al- terium incorporation was not perfect (Ͼ 96% D ). This was
2
[
13]
lyl groups of 1,
lay in the regio- and stereo-selective re- ascribed to the extreme sensitivity of the intermediate 13
duction of propargylic to allylic functionality. Hydrozir- towards adventitious water (i.e. unlabelled). The near-per-
conation of dimethyl dipropargyl malonate (12) using fect regioselectivity obtained through hydrozirconation was
Schwartz reagent, followed by protolysis of the terminal al- exploited in the synthesis of the internally deuterated alkene
2
lylic CϪZr bond (with H O) gave 1 without competing isomer [ H ]-6 (Ͼ 99% D ) by use of deuterated Schwartz
2
2
2
over-reduction of the intermediate alkenyl-zirconocene (13) reagent, Cp Zr(Cl)D.
2
or reduction of the ester functionality.[
14]
However, we
By deuteriumϪproton exchange of 12 with MeOD (cat.
found that with stoichiometric reagent, conversion was in- NaOMe) we obtained [ H ]-14 which on hydrozirconation,
complete. We initially suspected that the malonate unit was then quenching with H O, gave [(Z,Z)- H ]-7 with very
2
2
2
2
2
acting as an efficient Zr-chelator and thereby sequestering high deuterium incorporation (Ͼ 99% D ). The room-tem-
2
1
2
reagent. To combat this we added Cp ZrCl (1 equivalent) perature H NMR 400 MHz spectrum of [Z,Z- H ]-7 in
first and then the Schwartz reagent. However, this had no CDCl is complicated by molecular tumbling at the H-res-
2
2
2
2
3
effect and the incomplete reduction traced to the quality of onance frequency causing the three spin states to inter-
the commercial reagent.[ Reactions were therefore per- change at a sufficient rate to interfere with the multiplicity
15]
3
formed by titration with Schwartz reagent until no substrate of the JDH (ca. 2.6 Hz, trans) scalar coupling. Con-
could be detected by TLC. Although reagent quality varied sequently the internal CHϭC signal becomes rather com-
from batch to batch, typically we required ca. 1.2Ϫ1.5 plex and not a double, triple (1,2,1) triplet (1,1,1) as ex-
1636
Eur. J. Org. Chem. 2001, 1635Ϫ1642

Useful Probes for the Study of Transition Metal-Catalysed 1,6-Diene Cycloisomerisation
FULL PAPER
pected. Switching to a lower field (270 MHz), or increased
However, to our dismay, all Wittig-type reagents that we
temperature (55 °C) or more viscous solvent (C D ) had could derive from MeI (our intended source of the 13C la-
6
6
2
only marginal effects. The very high H incorporation is bel) completely failed to methylenate 16 (Ǟ 1) or 17 (Ǟ 18)
2
[18]
supported by the absence of a singlet adjacent to the up- under a range of conditions. Attempted methylenation of
field shifted CHD [1:1:1] triplet at δ ϭ 119.0 [C(1,7)] in the ketone 19,[ generated from 2 through oxidative cleavage
19]
[5]
13
[20]
C NMR spectrum.
under Sharpless conditions,
also failed. It thus appears
2
The combination of deuterated alkyne [ H ]-14 with deu- that such reactions are incompatible with the malonate res-
2
terated Schwartz reagent facilitated synthesis of the all-deu- idue. Indeed, we could find no published examples of Wit-
2
terated alkene isomer [ H ]-8 (Ͼ 96% D ) by employing a tig-type olefination of an aldehyde with malonate func-
6
6
2
D O quench. As with [(E,E)- H ]-5, the slightly lower deu- tionality at the β-position. We therefore turned our atten-
terium incorporation in [ H ]-8 arises through adventitious tion to methods by which we could attach a pre- C-la-
2
2
2
13
6
water effecting nondeuterated protolysis of the alkenyl-zir- belled allyl chain to a malonate unit. A Pd-catalysed allylic
[
21]
13
conocene intermediates.
alkylation reaction
employing a C-labelled allylic elec-
and a malonate anion seemed a reasonable
method. Even though this would generate a 1:1 mixture of
trophile[
22]
Synthesis of 13C-Labelled Substrates
Our first strategy for the introduction of a single 13C-
13
13
[1- C ]- and [3- C ]-labelling patterns which complicates
1 1
1
the H NMR spectra, vide infra, it does provide the advant-
label at the terminus of one of the allyl chains in 1 was
age of gaining 13C-labelling at two sites in 1 and 4 and at
based on introduction of the label at the last step of the
13
synthesis via methylenation of aldehyde 16 with a 13C-la-
four sites in 2 and 3. We elected to prepare a C-allylic
belled Wittig reagent. The δ-bromo analogue of 16 has been
reported as a substrate for radical cyclisation by Malacria
[
10a]
et al.
Based on this communication, we ozonolysed di-
methyl allylmalonate (15), protected the resultant aldehyde
[
16,17]
as a dimethyl acetal, allylated, and then deprotected
with aq. TFA
[
17c]
to give 16 Ϫ Scheme 3. Analogous pro-
pargylation instead of allylation afforded 17 and propargyl-
ation of 15 afforded 18. A test reaction confirmed that 18
could be selectively reduced to 1.
13
Scheme 3. Routes attempted in the generation of C-labelled di-
methyl hept-1,6-dienyl-4,4-dicarboxylate (1). Conditions: i, O
MeOH (88%); ii, (MeO) CH, cat. TsOH·H O (79%); iii, allyl brom-
ide, NaH, THF (50%); iv, TFA, CH Cl (74%); v, range of Wittig-
type methylenation conditions; vi, propargyl bromide, NaH, THF
75%); vii, Cp Zr(Cl), CH Cl ; viii, NaIO , CCl , MeCN, H O, cat.
RuCl ·H 0 (79%)
3
,
3
2
2
2
1
3
Scheme 4. Synthetic routes employed for the preparation of C-
1
3
2
(
2
2
2
4
4
2
labelled dimethyl hept-1,6-dienyl-4,4-dicarboxylate 9 and C, H-
labelled 10 and 11
3
2
Eur. J. Org. Chem. 2001, 1635Ϫ1642
1637

K. L. Bray, G. C. Lloyd-Jones
FULL PAPER
ester since this should be readily isolable and shelf stable hyde 20 reacted almost instantly with the ylide at Ϫ68 °C
and could then be coupled to either dimethyl malonate or (the bright yellow colour is discharged), if the reaction is
to dimethyl allyl malonate (15) under the standard Pd-cata- then quenched the yield of allyl benzoate is only 32%.
lysed allylic alkylation conditions.[ So that the allylic elec- Changing solvent to toluene or ether gave no real improve-
trophile would not be inconveniently volatile, but could also ment and no other identifiable products are isolated. How-
be readily distilled if necessary, we decided to prepare a ever, if the reaction is conducted at Ϫ10 °C and then al-
21]
1
3
benzoate ester. Furthermore, since C-labelled formalde- lowed to warm to ambient temperature before quenching,
hyde (or paraformaldehyde) is rather expensive, we chose to the yield is much improved (67%). We assume that the col-
label C(3) using a Wittig reaction (with phosphonium salt lapse of what is presumably the betaine intermediate is slow.
13
derived from CH I) rather than C(1) using a vinyl Using these conditions and generating the ylide from
3
[
22a]
13
13
anion.
[Ph P CH ][I] we obtained [3- C ]allyl benzoate (21) (Ͼ
3 3 1
The synthesis of the aldehyde (20, Scheme 4) required for 99% 13C) in good yield. Unfortunately, [3,3- H ]allyl benzo-
the Wittig reaction proved straightforward. Solketal (the ate (22) prepared analogously from [Ph P CD ][I] (Ͼ 98%
2
2
13
3
3
1
,2-monoacetonide of glycerol) was converted into its D ) was obtained with low deuterium incorporation (70%
3
[24]
benzoate ester (71%), the acetal hydrolysed (92%) and then D , 29% D , 1% D ).
2
1
0
the diol oxidatively cleaved with periodate to give 20 in 59%
With 13C-labelled allyl benzoate in hand, preparation of
yield. Using [Ph PCH ][I] as ylide precursor, it took some the 13C-labelled diallylmalonates 9, 10, and 11 was readily
3
3
while to find a method for the successful methylenation of achieved. Using Pd dba ·dba and 1,1Ј-bis(diphenylphos-
2
3
[
23]
2
0. The first trace of allyl benzoate (Ͻ 5%) was observed phanyl)ferrocene (dppf) as ligand in THF (1 mol-% Pd), the
when we employed KH with 18-crown-6 in THF. Switching sodium salt of dimethyl malonate reacted rapidly with [3-
13
13
to KHMDS in THF gave a marked improvement and isol-
C ]-21 to give [1,3- C ]-labelled allylmalonate 23 in 88%
1 1
able quantities of allyl benzoate were obtained. However, yield. Selective reductions of dimethyl propargylmalonate,
2
2
conditions were found to be crucial since even though alde- as Scheme 2, gave [ H ]-24 and [(E)- H ]-25. The anions of
1
1
Figure 1. A: The 1H NMR (300 MHz, CDCl
spectrum of the terminal alkene protons. Spectra B and C: observed and simulated sub-spectra of dimethyl [1,3- C
) spectrum of dimethyl hept-1,6-dienyl-4,4-dicarboxylate (1) and simulated (‘sim’) sub-
13 13
3
1
1
,5,7- C ]hept-1,6-
1
3
2
dienyl-4,4-dicarboxylate (9, 400 MHz) and dimethyl [1,3- C
1 1
,7-(E)- H ]hept-1,6-dienyl-4,4-dicarboxylate (11, 300 MHz)
1638
Eur. J. Org. Chem. 2001, 1635Ϫ1642

Useful Probes for the Study of Transition Metal-Catalysed 1,6-Diene Cycloisomerisation
FULL PAPER
2
3, 24, and 25 were treated, again under Pd-catalysis, with In contrast to 16, aldehyde 20 reacted rapidly with Ph Pϭ
3
13
13
[
3- C ]-21 to give doubly labelled products [ C ]-9, CH Ϫ even at Ϫ68 °C. However, only when the ylide was
1
2
2
1
3
2
13
2
[
C , H ]-10 and [ C ,(E)- H ]-11 in moderate to good generated from KHMDS and then the betaine warmed to
1
1
1
1
yields.
20 °C did this yield the desired allyl benzoate. Nonetheless,
13
under these optimised conditions, good yields of C-la-
1H NMR Spectra of Labelled Substrates
13
belled allyl benzoate [ C ]-21 are obtained and this can be
1
2
smoothly coupled to malonate units via Pd-catalysed allylic
Although quantitation of the H-label incorporation was
13
1
alkylation reactions. The use of a symmetrical C-labelled
Pd-π-allyl intermediate provides the advantage that the 13C-
label is introduced at two sites [C(1) and C(3)] in 1 Ϫ al-
though the resulting mutually exclusive 50% incorporation
performed by integration of H NMR signals (and sup-
ported by MS analyses, H NMR and C NMR), the in-
2
13
2
corporation of the various isotopic labels in [(E,E)- H ]-5,
2
2
1
2
2
13
13
2
[
[
H ]-6, [(Z,Z)- H ]-7, [ H ]-8, [ C ]-9, [ C , H ]-10, and
2 2 6 2 1 1
1
3
2
1
at each site does slightly complicate H NMR spectra. In
C -(E)- H ]-11 causes considerable changes in the
H
1
1
2
parallel we have developed high yielding H-labelling
NMR spectra as compared to unlabelled 1.
2
In particular the mutually exclusive 50% 13C-incorpora-
methods which thereby allow preparation of single ( H) and
13
2
mixed ( C, H) label systems with defined regiochemistry
and alkene geometry.
tion in the allyl chain, arising through essentially symmet-
rical [Pd-π- C-allyl] intermediates, results in complicated
mixtures. To confirm the isotopic labelling patterns in the
non-symmetrically labelled compounds [ C ]-9, [ C , H ]-
0, and [ C (E)- H ]-11, we simulated their H NMR
spectra. The following simplifications and methods were
employed: a) δ , δ , and J values were extracted (dir-
ectly or by simulation) from NMR analysis of unlabelled 1;
13
In summary, six isotopically labelled variants of 1 are
readily prepared and are useful probes in the study of the
mechanism of catalysed cycloisomerisation reactions.[
For example, Pd-catalysed cycloisomerisation of 1 by neut-
1
3
13
2
2
1
1
12]
13
2
1
1
1 1
ral[
26]
and cationic chloropalladium nitrile complexes
[7]
H
C
H,H
(
[(MeCN) PdCl ] and [(MeCN) PdCl][OTf] respectively)
2
2
3
2
1
1
gives 3 with good regioselectivity. Use of [ H ]-6, which af-
b) J
values were extracted directly from H NMR spec-
2
C,H
2
n
n
fords [ H ]-3, reveals that these two catalysts are mechan-
2
tra, and ‘normal’ J
ployed;
and J
values (n Ͼ 1) were em-
C,H
C,C
istically distinct Ϫ Scheme 5.[
27]
Additionally, Pd0 com-
[
25]
c) JX,D were predicted by using the approxi-
mation JX,D ഠ [(JX,H)/6.5];[ d) isotope shifts were ob-
25]
plexes of 1,6-dienes have recently attracted attention since
[
28]
they can provide useful sources of ‘‘naked metals’’
and
tained by iteration and e) we treated any carbon with 50%
C incorporation as two separate species, one with 100%
C and one with 0%. The simulated spectra correlated qu-
ite well with the observed H NMR spectra, see for example
II
1
1
3
3
the NMR characterisation of analogous Pd complexes of
[27]
the diene 1 is greatly faciliated by isotopic substitution.
Furthermore, the 13C-labelled allyl benzoate [ C ]-21 may
13
1
1
1
3
prove of utility in other synthetic strategies including metal-
allyl chemistry.
sub-spectra for the terminal alkene protons in [ C ]-9 and
2
1
3
2
[
C ,(E)- H ]-11 Ϫ Figure 1.
1 1
Conclusions
Experimental Section
We have developed reliable and practical synthetic routes
for the synthesis of six isotopically labelled variants of 1 General: Solvents and reagents were purified by standard proced-
2
2
2
2
(
[
namely, [(E,E)- H ]-5, [ H ]-6, [(Z,Z)- H ]-7, [ H ]-8, ures. Anhydrous solvents were purchased from Fluka or Aldrich
2 2 2 6
13 13
1
3
13
2
13
2
3
and used as received. CH I (Ͼ99% C) was purchased from Pro-
C ]-9, [ C , H ]-10, and [ C -(E)- H ]-11) in a state of
2
1
1
1
1
mochem and from Aldrich, MeOD (99% D) and D
were purchased from Aldrich and from Goss Scientific,
Cp Zr(Cl)D was purchased from Fluka. Material supplied by Ald-
rich was found to be completely inactive. KHMDS (0.5  in tolu-
2
O (99.9% D)
high purity. The initial synthetic strategy involved introduc-
tion of the 13C-labels via a Wittig-type methylenation of
2
aldehyde 16. However, this failed to react with a variety of
ylides Ϫ even at reflux in THF. Consequently we changed
strategy and utilised α-benzoyloxy aldehyde 20 (Scheme 4).
13
ene) was purchased from Fluka. [PPh
3 3
CH ][I] was prepared in
essentially quantitative yield by collection of the crystalline mat-
erial deposited from reaction of stoichiometric quantities of PPh
3
1
3
3
and CH I in toluene. Dimethyl allylmalonate [dimethyl but-3-ene-
1
,1-dicarboxylate] was purchased from Lancaster Synthesis. A mix-
ture of dimethyl propargylmalonate [dimethyl but-3-yne-1,1-di-
carboxylate] and dimethyl (dipropargyl)malonate [dimethyl hept-
1
,6-dienyl-4,4-dicarboxylate] was prepared by reaction of dimethyl
malonate with NaH and excess propargyl bromide in THF. The
two components were separated column chromatography. α-
Benzoyloxy acetaldehyde was prepared according to a literature
procedure.[ When appropriate, reactions were carried out under
nitrogen using standard Schlenk techniques. Ϫ NMR experiments
were performed on JEOL instruments (Delta 270, Lambda 300,
29]
2
GX400, and Eclipse 400). NMR spectra were recorded in CDCl
3
Scheme 5. Deployment of [ H
2
]-6 to demonstrate mechanistic dif-
1
or in CHCl
ual CHCl
3
. H NMR spectra were referenced internally to resid-
(δ ϭ 7.27) or to TMS (δ ϭ 0.00). H NMR spectra were
ferences between neutral and cationic chloropalladium catalysts for
the regioselectiove cycloisomerisation of 1 to 3
2
3
Eur. J. Org. Chem. 2001, 1635Ϫ1642
1639

K. L. Bray, G. C. Lloyd-Jones
FULL PAPER
referenced internally to CDCl
3
(δ ϭ 7.27, ca. 1% added to CHCl
3
). 171.2 (CO
2
); δ
D
(46 MHz, CHCl
3
): 5.14 (br. s); m/z (CI): 215
13
C NMR spectra were referenced to CDCl
3
(δ ϭ 77.0). in 13C [MH ] (23%), 187 (61), 155 (33), 143 (30), 83 (100), 57 (20).
ϩ
NMR spectra of labelled compounds, when appropriate, satellites
are reported and are denoted as ‘sat’. Spectral simulation ( H) was
2
Dimethyl [1,1,2,6,7,7- H
Following the procedure for 5, 14 (185 mg, 0.88 mmol) was treated
with Cp Zr(Cl)D (576 mg, 2.22 mmol) and quenched with D O to
give 8 (159 mg, 83%) as a colourless oil. Ͼ97% 4,4Ј-D ; Ͼ 96%
): 2.64 (s, 4 H,
): 36.7 (CH ), 52.3
), 118.8 (sp, JCD 24.6, CDϭCD ), 131.7 (t, JCD
), 171.2 (CO ); δ (46 MHz, CHCl ): 5.13 (s, 2 D,
), 5.68 (s, 1 D, CDϭCD
] (5), 187 (19), 159 (44), 142 (30), 127 (39), 111 (40), 97
6
]Hept-1,6-dienyl-4,4-dicarboxylate (8):
1
performed using g NMR software. Coupling constants extracted in
such a manner are reported as Јsimul.Ј. Ϫ Mass spectra were ob-
tained using both CI and EI sources on a Fisons ЈMicromass
AutospecЈ mass spectrometer. Ϫ Flash column chromatography:
Merck silica gel 60. Ϫ TLC: 0.25 mm, Merck silica gel 60 F254
2
2
2
1
5
,5,5Ј,5Ј-D
CH ), 3.72 (s, 6 H, CH
), 57.6 (CE
3.83, CDϭCD
4
( H NMR). Ϫ δ
H 3
(400 MHz, CDCl
2
3
); δ (100 MHz, CDCl
C
3
2
1
1
(
CH
3
2
2
2 4 4
visualising at 254 nm or with acidic (H SO ) aq. KMnO solution
2
2
2
D
3
(ca. 2%).
ϩ
ϩ
CDϭCD
218 [M
2
2
); m/z (CI ): 219 [MH ] (2%),
2
ϩ
Dimethyl [1,7- H
2
]Hept-1,6-diynyl-4,4-dicarboxylate (14): NaOMe
(17 mg, 0.3 mmol) was added to a stirred solution of dimethyl hept-
(67), 85 (65), 71 (74), 57 (100).
1
,6-diynyl-4,4-dicarboxylate (368 mg, 1.77 mmol) in MeOD
2
Dimethyl [3- H
1
]But-3-ene-1,1-dicarboxylate (24): Following the
3
3
(5 cm ). After 0.5 h the mixture was concentrated in vacuo to 1 cm
procedure for 5, dimethyl but-3-yne-1,1-dicarboxylate (302 mg,
.78 mmol) was treated with Cp Zr(Cl)D (697 mg, 2.69 mmol) and
quenched with H O to give 24 (207 mg, 67%) as a colourless oil.
3
and fresh MeOD (5 cm ) was added. The cycle was repeated three
times after which the mixture was concentrated in vacuo, D
1
2
2
O
2
3
(2 cm ) was added and then the mixture extracted with CH
2
Cl
2
(3
1
3
Ͼ 99% D ( H NMR). Ϫ δ
H
(300 MHz, CDCl
3
): 2.65 (d, 2 H, J
), 3.74 (s, 6 H, CH ),
(75 MHz, CDCl ): 32.7 (CH CDϭ
), 117.5 (ϭCH ), 133.6 (t, CD 23.6,
(46 MHz, CHCl ): 5.80 (s); m/z (CI ):
3
3
ϫ 10 cm ). The combined extracts were washed with D
dried with anhydrous MgSO and concentrated in vacuo to afford
the title compound (254 mg, 68%) as a white solid. Ͼ99% D
NMR). Ϫ δ (400 MHz, CDCl ): 3.00 (s, 4 H, CH ), 3.77 (s, 6 H,
ϫ CH ); δ (100 MHz, CDCl ): 22.6 (CH ), 53.1 (CH ), 56.4
CE ), 71.5 (t, JCD 38.4, CD), 77.8 (t, JCD 7.7, CH CCD), 169.0
CO ); δ (46 MHz, CHCl ): 2.04 (s); m/z (CI): 211 [MH ] (10%),
2
O (2 cm ),
3
7
.5, CH
2
CDϭ), 3.46 (t, 1 H, J 7.5, CHCH
); δ
), 52.5 (HCE
2
3
4
5
.05Ϫ5.15 (m, 2 H, ϭCH
2
C
3
2
1
2
( H
1
)
, 51.4 (CH
3
2
2
J
H
3
2
ϩ
CDϭCH
1
2
ϩ
), 169.3 (CO
2
); δ
D
3
2
(
(
2
1
3
C
3
2
3
74 [MH ] (34%), 149 (11), 142 (73), 114 (15), 110 (29), 88 (67),
4 (110).
1
2
2
2
8
ϩ
2
D
3
2
02 (13), 197 (12), 187 (52), 178 (42), 170 (42), 170 (40), 151 (46),
23 (64), 119 (100), 109 (62), 105 (44).
Dimethyl [4-(E)- H
the procedure for 5, dimethyl but-3-yne-1,1-dicarboxylate (403 mg,
.37 mmol) was treated with Cp Zr(Cl)H (850 mg, 3.29 mmol) and
quenched with D O to give 25 (238 mg, 58%) as a colourless oil.
1
]But-3-ene-1,1-dicarboxylate (25): Following
2
2
2
Dimethyl [1,7-(E,E)- H
stirred mixture of dimethyl hept-1,6-diynyl-4,4-dicarboxylate
2
]Hept-1,6-dienyl-4,4-dicarboxylate (5): To a
2
1
3
Ͼ 94% D ( H NMR). Ϫ δ
H
(300 MHz, CDCl
3
): 2.67 (ddd, 2 H, J
CHϭ), 3.47 (t, 1 H, J 7.4, HCCH ), 3.74 (s,
), 5.12 (dt, 1 H, 3Jtrans 17.0, J 1.5, CHD), 5.77 (dt, 1 H,
3
3
(150 mg, 0.72 mmol) in CH
2
Cl
2
(5 cm ) was added Cp
2
Zr(Cl)H
4
3
7
.4, 7.1, J 1.5, CH
H, CH
2
2
(516 mg, 2.00 mmol) at 0 °C. After 80 min at 0 °C the mixture was
4
6
3
quenched by the addition of D
through a pad of silica, eluted with CH
concentrated in vacuo. Purification by column chromatography us-
ing hexane/ethyl acetate (5:1) as the eluent afforded 5 as a colour-
2
O (2 cm ). The solution was filtered
3
3
J 7.1, trans 17.0, CHϭCHD); δ
CHϭ), 51.3 (HCE ), 52.4 (CH ), 117.3 (t, JCD 24.8, CHD),
33.7 (CHϭCHD), 169.2 (CO ); δ (46 MHz: CHCl ): 5.00 (s); m/
z (CI): 174 [MH ] (20), 149 (16), 142 (40), 125 (28), 115 (18), 114
24), 97 (26), 71 (36), 61 (16), 59 (14), 57 (52).
J
C 3
(75 MHz, CDCl ): 32.8
3
Cl
2 2
(3 ϫ 5 cm ) and then
1
(
CH
2
2
3
1
2
D
3
ϩ
1
less oil (124 mg, 80%). Ͼ 96% D
2
( H NMR). Ϫ δ
H
(400 MHz,
), 5.10
(
3
CDCl
d, 2 H, 3Jtrans 16.0, CHϭCHD), 5.64 (dt, 2 H, 3Jtrans 16.0, J 8.0,
CHϭCHD); δ (100 MHz, CDCl ): 36.9 (CH CHϭ), 52.4 (CH ),
), 118.9 (t, JCD 23.8, CHϭCHD), 132.1 (CHϭCHD),
3
): 2.64 (d, 4 H, J 8.0, CH
2
CHϭ), 3.72 (s, 6 H, CH
3
3
1
3
1 3
[3- C ]Allyl Benzoate (21): A 0.5  solution of KN(SiMe )
11.4 cm , 5.7 mmol) was added dropwise to a suspension of
PPh P CH ][I] (2.31 g, 5.70 mmol) in THF (30 cm ) at Ϫ10 °C
3 3
and the bright yellow suspension was then stirred for 2 h. Benzoyl-
oxyacetaldehyde (850 mg, 5.18 mmol) in THF (2 cm ) was then
added, the reaction mixture allowed to warm to room temp. and
then stirred for 3 h. After quenching with water (10 cm ), the mix-
ture was extracted with diethyl ether (3 ϫ 20 cm ). The combined
(
3
(
[
C
3
2
3
1
3
3
5
1
7.6 (CE
71.2 (CO
2
ϩ
2
D 3
); δ (46 MHz, CHCl ): 5.15 (s); m/z (CI): 215 [MH ]
3
ϩ
ϩ
ϩ
(
(
42%), 214 [M ] (45), 213 [M Ϫ H] (27), 212 [M Ϫ D] (21), 199
28), 197 (29), 172 (25), 140 (100).
3
2
3
Dimethyl [2,6- H
2
]Hept-1,6-dienyl-4,4-dicarboxylate (6): Following
the procedure for 5, dimethyl hept-1,6-diynyl-4,4-dicarboxylate
206 mg, 0.99 mmol) was treated with Cp
.77 mmol) and quenched with H O to give 6 (144 mg, 68%) as a
( H NMR). Ϫ δ (400 MHz, CDCl
CDϭ), 3.72 (s, 6 H, CH ), 5.10 (br. s, 4 H, ϭ
(100 MHz, CDCl ): 36.8 (CH CDϭ), 52.4 (CH ), 57.6
), 119.0 (ϭCH ), 131.9 (t, JCD 23.8, CDϭCH ), 171.2 (CO );
(46 MHz, CHCl ): 5.68 (s); m/z (EI): 214 [M ] (3%), 183 (6),
extracts washed with brine (10 cm
3
), dried with anhydrous magnes-
(
2
2
Zr(Cl)D (716 mg, ium sulfate, and concentrated in vacuo to yield a yellow oil. Puri-
2
fication by column chromatography using hexane/ethyl acetate
1
colourless oil. Ͼ 99% D
.64 (s, 4 H, CH
CH ); δ
2
H
3
): (20:1) as the eluent afforded 21 (556 mg, 67%) as a colourless oil.
2
2
3
Ϫ δH (300 MHz, CDCl ): 4.83 (ddt, 2 H,
CH CHϭ), 5.25 (dddt, 1 H,
3
J 5.6,
J 1.5,
3
JCH 11.2,
Jcis 10.4,
4
J 1.5,
J 1.5,
3
1
2
3
4
4
2
C
3
2
3
2
JCH 159.6,
1
13
1
155.5, 2J 1.5, 3Jtrans 17.3, J 1.5,
(CE
2
2
2
2
CHH), 5.41 (dddt, 1 H,
J
CH
ϩ
13
CHH), 6.04 (ddt, 1 H, J 5.6, 3Jcis 10.4, 3J
3
δ
1
D
3
trans
17.3, CHϭ),
72 (26), 154 (24), 140 (100), 123 (18).
7.41Ϫ7.50 (m, 2 H, m-CHaryl), 7.53Ϫ7.61 (m, 1 H, p-CHaryl),
.04Ϫ8.10 (m, 2 H, p-CHaryl); δ (75 MHz, CDCl ): 65.5
), 128.3 (m-CHaryl), 129.6 (o-CHaryl),
8
C
3
2
Dimethyl [1,7-(Z,Z)- H
2
]Hept-1,6-dienyl-4,4-dicarboxylate (7): Fol-
1
3
(
2 2
CH CHϭ), 118.2 (ϭ CH
lowing the procedure for 5, 14 (206 mg, 0.99 mmol) was treated
with Cp Zr(Cl)H (657 mg, 2.55 mmol) and quenched with H O to
give 7 (136 mg, 74%) as a colourless oil. Ͼ 99% D ( H NMR). Ϫ
CHϭ), 3.72 (s, 6
1
13
1
1
8
32.6 (d, JCC 70.2, CHϭ C), 132.9 (p-CHaryl), 135.3 (i-Caryl),
2
); m/z (EI): 163 [M ] (17%), 118 (8), 105 (100), 86 (32),
4(46), 77 (60).
2
2
ϩ
66.2 (CO
1
2
3
δ
H
(400 MHz, CDCl
H, CH
), 5.10 (2 H, 3
CHϭCHD); δ (100 MHz, CDCl
), 118.9 (t, 1
CD 23.8, CHϭCHD), 132.1 (CHϭCHD),
3 2
): 2.57 (d, 4 H, J 7.3, CH
13
3
1
Jcis 10.3, CHϭCHD), 5.52Ϫ5.61 (m, 2 H, Dimethyl [2,4- C ]But-3-ene-1,1-dicarboxylate (23): To a stirred
3
C
3
): 36.9 (CH
2
CHϭ), 52.4 (CH
3
), suspension of sodium hydride (49 mg, 2.02 mmol) in THF (4 cm )
5
7.6 (CE
2
J
was added dimethylmalonate (0.21 mL, 1.84 mmol) at 0 °C. After
1640
Eur. J. Org. Chem. 2001, 1635Ϫ1642

Useful Probes for the Study of Transition Metal-Catalysed 1,6-Diene Cycloisomerisation
FULL PAPER
1
3
2
1
5 min a solution of dppf (15 mg, 1.5 mol-%) and Pd
2
·dba
3
·dba
Dimethyl
[1,3- C
1
,7-(E)- H
1
]Hept-1,6-dienyl-4,4-dicarboxylate
3
(
(
17 mg, 1 mol-% Pd) in THF (1 cm ) was added, followed by 21 (11): Following a procedure analogous to that outlined for 23, 21
300 mg, 1.84 mmol). After 2 h the mixture was quenched by the
(215 mg, 1.32 mmol) was treated with 25 (228 mg, 1.32 mmol) to
give 11 (216 mg, 76%) as a colourless oil. Ͼ 94% D ( H NMR). Ϫ
δ (300 MHz, CDCl ): 2.64 (ddd, 4 H, J 7.4, J 1.1, JDH simul. 0.2,
H 3
3
1
addition of 2.8  NH
4
Cl aq. (10 cm ) and extracted with ethyl acet-
ate (3 ϫ 20 cm ). The combined extracts were washed with brine
10 cm 3), dried with anhydrous magnesium sulfate, and concen-
trated in vacuo to yield a brown oil. Purification by column chro-
3
3
4
4
13
3
3
4
(
CH
CH
2
2 2 2
ϭCHCH CCH , and ddt J 7.4, JCH 6.7, J 1.1 CH CHϭ
1
3
3
4
1
13
3
2
3
, ddt J 7.4, J 1.1 JCH 132.0, CH
2
CHϭ, and dddd, JCH
4
4
DH simul. 0.2, 13CH
matography using hexane/ethyl acetate (6:1) as the eluent gave 23 5.0, J 7.4, J 1.1,
J
2
ϭCHCH ), 3.72 [s, 6 H,
2
2
3
4
1
(284 mg, 88%) as a colourless oil. Ϫ δ
H
(300 MHz, CDCl
3
): 2.65
C(CO
2
CH
3
)
2
13
], 5.11 (dddt, 1 H, Jsimul. 1.8, Jcis simul. 9.9, J 1.1, JCH
158.0, CHϭ CHH, and dddt Jsimul. 1.8, JCH 12.7, Jcis simul. 9.9,
3
3
4
1
2
3
3
(
1
dddt, 2 H, J 7.3, 7.3, JCH 6.8, J 1.2, CH CHϭ, and dddt, JCH
2
31.7, J 7.3, 7.3, J 1.2, 13CH
3
4
CHϭ), 3.47 (t, 1 H, J 2.0, HCCH
3
,
4
J 1.1, CH
), 5.08 (dddd, 3Jtrans simul. 16.8, J 1.1, CHϭ CHH, and dddt Jsimul. 1.8, 3Jtrans si-
3
13
1
2
2
13
2
2
CHϭCHH), 5.12 (dddt, 2 H, JCH 154.0, Jsimul. 1.8,
2
3
4
13
2
and dt, JCH 15.2, J 2.0, HC CH
2
), 3.72 (s, 6 H, CH
CH 10.3, ϭCHHcis, and dddd,
H, 2J 1.2, J 10.3, J 1.2,
3
4
3
J
1
J
CH
3
mul. 16.8, 3JCH 7.6, J 1.1, CH
4
13
CHϭCHH, and ddt JDH simul. 0.3,
2
2
58.3, J 1.2, J 10.3, J 1.2, ϭ CHHcis), 5.12 (dddd, H, J 1.2, J 3Jtrans simul. 16.8, 4J 1.1, CHϭCHD), 5.65 (dddt, 2 H, 2JCH simul.
2
3
4
13
2
1
1
1
4
3
1
2
3
3
3
3
13
7.1, J 1.2, JCH 7.3, ϭCHHtrans, and dddd, JCH 154.4, J 1.2, J
2
0.4, Jtrans simul. 16.80, Jcis simul. 9.9, J 7.4, CHϭ CH , and dddt
4
13
3
2
CH simul. 5.0, 3
cis simul. 9.9, 3
3
13
7.1, J 1.2, ϭ CHHtrans), 5.77 (ddt, 1 H, J 17.1, 10.3, 7.3, CHϭ
J
J
J
trans simul. 16.8, J 7.4, CH
2
CHϭ,
1
CC 33.0 and 42.3, and ddt, J 7.4, 3Jtrans simul. 16.8, 3JDH simul. 1.5, CHϭCDH); δ
3
CH
2
); δ
CH
CHϭ), 51.4 (d, 1
HCE CH ), 52.5 (CH
C
(100 MHz, CDCl
3
): 32.9 (sat:
J
C
1
3
13
3
CHϭCHD, 13CH
2
J
CC 33.0, HCE
2
CH
2
), 51.4 (d,
J
CC 3.1,
), 133.9
CHϭ), 169.3
); m/z (EI ): 172 [M ] (14%), 155 (66), 140 (33), 123 (57), 95
84), 84 (100), 71 (38), 57 (64).
(75 MHz, CDCl
3
): 36.6 (CH
2
2
13
CHϭ, CH
2
CHϭ
CHϭ
), 119.0 (t, JCD 19.3, CHϭCHD), 132.1 (d, JCC 69.6, CHϭ
1
13
13
13
2
2
3
), 117.7 (sat: JCC 70.0, CHϭ CH
2
CH
CH
2
) 52.4 (CH ), 57.7 (CE
3
2
), 119.3 (CHϭ CH
2
,
CH
2
1
13
1
13
1
1
(
(
(
d, JCC 70.0, CHϭ CH
CO
2
), 133.9 (d, JCC 42.3 CH
2
2
ϩ
ϩ
13
), 132.1 (CHϭCHD), 132.3 (d, 1J CC 42.9, CH
13
2
CH
(CO ); δ
183 (41), 172 (8), 155 (66), 151 (40), 123 (43), 95 (38), 84 (100), 79
85), 63 (29), 57 (16).
2
2
CHϭ),171.2
ϩ
2
D
(46 MHz, CHCl ): 5.14 (s); m/z (EI): 215 [MH ] (33%),
3
13
13
Dimethyl [1,3- C
Following a procedure analogous to that outlined for 23, 21
256 mg, 1.57 mmol) was treated with 23 (272 mg, 1.57 mmol) to
give 9 as a colourless oil (64 mg, 19%). Ϫ δ (400 MHz, CDCl )::
and ddt,
1 1
,5,7- C ]Hept-1,6-dienyl-4,4-dicarboxylate (9):
(
(
H
3
3
3
4
13
Acknowledgments
G. C. L.-J. thanks the Zeneca Strategic Research Fund, Pfizer Ltd.
and Lancaster Synthesis for generous support. K. L. B. thanks the
2
.64 (4 H, JCH 6.7, J 7.4, J 1.1, E
2
C(CH
2
13
CH CH
2
)
2
1
3
4
13
1
J
CH 132.0, J 7.4, J 1.1, CH
2
CHCH
2
CE
2
CH
2
, and dddt, JCH
32.0, 3JCH 5.0, J 7.4, J 1.1, 13CH
2
4
CE
CH
13
CH , and dddt, JCH 6.7,
3
1
J
2
2
2
3
3
4
13
13
CH 5.0, J 7.4, J 1.1, CH
2
CE
2
3
2
CHϭ CH
2
3
); 3.72 (s, 6 H, University of Bristol for a postgraduate studentship. We are very
2
4
CH
3
), 5.12 (2 H, dddt, Jsimul. 1.8, Jcis 9.9, J 1.1, JCH 12.7, CHϭ
grateful to Dr. Martin Murray (University of Bristol) for help with
NMR experiments.
CHH, and dddt JCH 158.0, 2
1
Jsimul. 1.8,
3
J
simul. 9.9, J 1.1, CHϭ
4
13
CHH), 5.13 (2 H, dddt, JCH 154.0, Jsimul. 1.8, 3Jtrans simul, 16.8,
1
2
4
13
2
simul. 1.8, 3
trans simul. 16.8, 3
J 1.1, CHϭ CHH, and dddt,
J
J
J
CH
[1]
[2]
[1a]
Reviews: B. M. Trost, Acc. Chem. Res. 1990, 23, 34Ϫ42. Ϫ
.6, J 1.1, CHϭCHH); 5.65 (2 H, dddt, Jcis simul. 9.9, 3Jtrans simul.
4
3
7
[1b]
B. M. Trost, M. J. Krische, Synlett 1998, 1Ϫ15.
6.8, 3J 7.4, 2
J
CH 5.0, CH
13
CHϭCH
, and dddt, 3
); δ
CHϭ), 52.4 (CH
3
J
cis simul. 9.9,
1
2
2
A. Bright, J. F. Malone, J. K. Nicholson, J. Powell, B. L. Shaw,
3
3
2
13
J
trans simul. 1.8, J 7.4, JCH 0.4, CH CH
2
C
(100 MHz, CDCl
3
):
J. Chem. Soc., Chem. Commun. 1971, 712Ϫ713.
[3] [3a]
6.9 (sat: 1JCC 33.8 and 43.0, 13CH
R. Grigg, T. R. B. Mitchell, A. Ramasubbu, J. Chem. Soc.,
3
2
1
), 57.6 (dd,
[3b]
1
3
13
Chem. Commun. 1979, 669Ϫ670. Ϫ
R. Grigg, T. R. B. Mit-
J
CC 33.8, JCC 3.1, CE
2
), 119.2 (sat: JCC 69.2, CHϭ CH
CHϭ), 132.2 (d,
2
13
), 132.2
chell, A. Ramasubbu, J. Chem. Soc., Chem. Commun. 1980,
d, 1JCC 43.0, CH
13
1
JCC 69.2,CHϭ CH
),
(
2
2
[3c]
2
7Ϫ28. Ϫ
Ramasubbu, R. M. Scott, J. Chem. Soc., Perkin Trans. 1
984, 1745Ϫ1754.
R. Grigg, J. F. Malone, T. R. B. Mitchell, A.
1
2
71.2 (CO ).
1
13
2
Dimethyl [1,3- C
Following a procedure analogous to that outlined for 23, 21
1
,6- H
1
]Hept-1,6-dienyl-4,4-dicarboxylate (10):
[4]
E. Schmitz, R. Urban, G. Zimmermann, J. Prakt. Chem. 1976,
318, 185; E. Schmitz, U. Hench, D. Habisch, J. Prakt. Chem.
1976, 318, 471.
Y. Yamamoto, N. Ohkoshi, M. Kameda, K. Itoh, J. Org. Chem.
1999, 64, 2178Ϫ2179.
(
(
(
175 mg, 1.08 mmol) was treated 24 (186 mg, 1.08 mmol) to give 10
100 mg, 43%) as a colourless oil. Ͼ 99% D ( H NMR). Ϫ δ
300 MHz, CDCl
[
[
5]
6]
1
H
3
4
3
): 2.64 (dt, 4 H, JDH simul. 1.1, J 1.1, CH
2
3
CDϭ
B. Radetich, T. V. RajanBabu, J. Am. Chem., Soc. 1998, 120,
and ddt, JCH 132.0, J 7.4, J 1.1, 13CH
1
3
4
CHϭ, and ddt, J 7.4,
,
2
3
8
007Ϫ8008.
3
4
13
3
J
CH 6.7, J 1.1, CH
2
CHϭ CH
2
, and ddt, JDH simul. 1.1, JCH 5.0,
[7]
[8]
A. Heumann, M. Moukhliss, Synlett 1998, 1211Ϫ1212.
4
2
2
3
2
13
2 2 2 3
J 1.10, CH CE CH CDϭ), 3.72 (s, 6 H, CH ), 5.11 (ddt, 2 H,
Note that cycloisomerisation is not always a desirable reaction
Ϫ for example, it recently was reported to be a seriously com-
peting side reaction (43%) in a Ru-catalysed ring-closing meta-
thesis of N-tosyl diallyl amine. See: A. Fürstner, M. Liebl, C.
W. Lehmann, M. Picquet, R. Kunz, C. Bruneau, D. Touchard,
P. H. Dixneuf, Chem. Eur. J. 2000, 6, 1847Ϫ1857.
3
4
1
J
J
J
J
simul. 1.8, JHD simul. 1.5, J 1.1, CDϭCHH, and dddt, JCH 158.0,
3
4
13
2
simul. 1.8, Jcis simul. 9.9, J 1.1, CHϭ CHH, and dddt, Jsimul. 1.8,
3
4
13
2
cis simul. 9.90, JCH 12.7, J 1.1, CH CHϭCHH), 5.12 (ddt, 2 H,
3
4
1
simul. 1.8, JDH simul. 2.6, J 1.1, CDϭCHH, and dddt, JCH 154.0,
2Jsimul. 1.8, Jtrans simul. 16.8, J 1.1, CHϭ CHH, and dddt, Jsimul.
3
4
13
2
[9] [9a]
W. E. Piers, P. J. Shapiro, E. E. Bunel, J. E. Bercaw, Synlett
.8, 3Jtrans simul. 16.8, 3JCH 7.6, J 1.1, CH
4
13
CHϭCHH), 5.65 (dddt,
[9b]
1
1
2
1990, 74Ϫ84. Ϫ
Chem. Soc. 1991, 113, 6268Ϫ6270. Ϫ
Hoberg, J. Am. Chem. Soc. 1992, 114, 3123Ϫ3126. Ϫ
Negishi, T. Takahashi, Acc. Chem. Res. 1994, 27, 124Ϫ130. Ϫ
[9f]
K. S. Knight, R. M. Waymouth, J. Am.
[9c]
H, 2
CH simul. 0.4, 3J 7.4, 3
cis simul. 9.9, 3
G. A. Molander, J. O.
[9d]
J
J
Jtrans simul. 16.8, CHϭ
E.-I.
1
3
, and dddt, JCH 5.0, 3J 7.4, 3
2
cis simul. 9.9, 3
CH
CH
2
J
Jtrans simul. 16.8,
1
3
[9e]
2
CHϭ), δ
), 119.2 (CHϭ CH
t, JCD 69.6, CDϭCH ), 132.0 (CHϭ CH
CO ); δ (46 MHz, CHCl ): 5.69 (s); m/z (EI): 215 [MH ] (2%),
C
(75 MHz, CDCl
3
): 36.9 (CH
2
CE
2
), 52.4 (CH
3
),
U. M. Dzhemilev, Tetrahedron 1995, 51, 4333Ϫ4346. Ϫ
J. Christoffers, R. G. Bergman, J. Am. Chem. Soc. 1996, 118,
1
3
13
5
(
(
9.7 (CE
2
2
, CDϭCH
2
13
,
CH
2
13
CHϭCH
2
), 131.9
[
9g]
1
4715Ϫ4716. Ϫ
30, 201Ϫ207.
S. Thiele, G. Erker, Chem. Ber./Recueil 1997,
2
2
,
2
CH CHϭ), 171.2
1
ϩ
2
D
3
[10] [10a]
P. Denn, L. Fensterbank, M. Malacria, Tetrahedron Lett.
1
8
83 (3), 172 (2), 155 (7), 151 (6), 140 (4), 123 (5), 105 (4), 95 (6),
6 (64), 84 (100).
[10b]
1
998, 39, 833؊836. ؊
D. Stein, R. Samy, R. Nouguier, D.
[10c]
Crich, M. P. Bertrand, J. Org. Chem. 1997, 62, 275Ϫ286. Ϫ
Eur. J. Org. Chem. 2001, 1635Ϫ1642
1641

K. L. Bray, G. C. Lloyd-Jones
FULL PAPER
[
20]
A. M. Bernard, P. P. Piras, P. Toriggia, Synthesis 1990,
P. J. H. Carlsen, T. Katsuki, V. S. Martin, B. Sharpless, J. Org.
Chem. 1981, 46, 3936Ϫ3938.
5
27Ϫ529.
[
[
[
11]
12]
13]
[21]
[22]
G. C. Lloyd-Jones, Synlett 2001, 161Ϫ183.
J. Tsuji, Palladium Reagents and Catalysts, Innovations in Or-
K. L. Bray, G. C. Lloyd-Jones, unpublished results
As far as we are aware, there has only been one report of the
ganic Synthesis, Wiley, Chichester, UK, 1995.
As far as we are aware, there have been no reports of the use
1
3
deliberate preparation of deuterated-allyl malonates: diethyl
of C-labelled allylic electrophiles in Pd-catalysed allylic al-
13 13
2
[
1,7-(Z,Z)- H
2
]hept-1,6-diene-4,4-dicarboxylate and diethyl [4-
kylation in the literature, nor have any [1- C ]-, [2- C
1
1
]-, or
]-Allylic elec-
]Prop-2-en-1-ol: H.
Badawi, P. Lorencak, K. W. Hillig, M. Imachi, R. L. Kuczkow-
2
13
13
(
Z)- H
1
]but-3-ene-1,1-dicarboxylate were prepared as a mix-
[3- C ]-labelled allylic esters been prepared. [ C
1
1
[22a] 13
ture by Fe-mediated reaction between dimethyl malonate and
trophiles have been reported:
[1- C
1
2
[13a]
ethyl 3-[(Z)- H
1
]propen-1-yl carbonate:
Y. Xu, B. Zhou, J.
[
22b]
13
Org. Chem. 1987, 52, 974Ϫ977. Additionally, diethyl [4,4,2,2-
ski, J. Mol. Struct. 1987, 162, 247Ϫ254. Ϫ
[1- C
1
]-1-Chlo-
2
H
H
4
]but-3-ene-1,1-dicarboxylate and diethyl [4,4,3,3,2,2-
6
roprop-2-ene: W. Kirmse, V. Zellmer, B. Goer, J. Am. Chem.
2
[22c]
13
]but-3-ene-1,1-dicarboxylate have been isolated as side
Soc. 1986, 108, 4912Ϫ4917. Ϫ
[2- C ]-1-Bromoprop-2-
1
2
products from reaction of [1,1,3,3- H
4
]-1,3-dibromopropane
ene: G. A. Olah, C. L. Jeuell, A. M. White, J. Am. Chem. Soc.
1969, 91, 3961Ϫ3962.
2
and [1,1,2,2,3,3- H
6
]-1,3-dibromopropane with diethyl malon-
J. A. van Zee, A. L. Kwiram, J. Am. Chem. Soc. 1990,
12, 5012Ϫ5018.
[
13b]
[23]
ate:
1
Using BuLi as base and conducting the reaction in THF gave
no trace of allyl benzoate (Ͻ 1%) even though all of 20 was
consumed. Use of sodium rather than lithium containing bases
[
[
[
14]
15]
16]
J. Schwartz, J. A. Labinger, Angew. Chem., Int. Ed. Engl. 1976,
1
S. L. Buchwald, S. J. Lamaire, R. B. Nielsen, B. T. Watson,
Organic Synth. 1992, 71, 77Ϫ82.
Removal of the dimethyl acetal protecting group proved very
troublesome. A number of methods were tested, including
2 4
H SO /wet silica-gel (see reference [10b]), oxalic acid/wet silica-
5, 333Ϫ339.
3 2 2
(NaHMDS, THF; NaHCO aq./CH Cl (see: J. Pitlik, G.
Batta, F. Sztaricskai, Liebigs Ann. Chem. 1992, 895Ϫ898) or
NaH, THF) made no improvement. No identifiable side prod-
1
13
31
ucts ( H, C, P NMR) were obtained and we suspect that
irreversible deprotonation of 20 leads to decomposition of the
enolate to give volatile products arising by fragmentation.
[24]
2 2
Quenching the reaction with D O instead of H O had no effect
gel (see reference [17a]), wet silica-gel/CH
17a]), and TsOH/CH Cl (see reference [17b]). However, under
these conditions, either there was no deprotection or complete
2 2
Cl (see reference
on the deuterium incorporation. However, material balance in-
[
2
2
dicated that the mol of 20 not converted into 22 tallied well
1
with the amount of H incorporation at C(3). Furthermore,
2 2
decomposition. Eventually we found that TFA in CH Cl (see
reference [17c]) effected very clean deprotection and gave alde-
2
there was no H incorporation at C(1). This suggests that de-
1
protonation of 20 is irreversible and that the liberated H be-
hyde 16 in 74% yield.
[
[
17] [17a]
comes redistributed by exchange of protons between the re-
sulting the phosphonium salt and the ylide. See: H. J.
Bestmann, Chem. Ber. 1962, 95, 58Ϫ63.
F. Huet, A. Levchevallier, M. Pellet, J. M. Conia, Synthesis
[
17b]
1
978, 63Ϫ65. Ϫ
M. Mikolajczyk, P. Balczewski, Synthesis
17c]
[
1
987, 659Ϫ661.
R. A. Ellison, E. R. Lukenbach, C.-W.
[
[
[
[
25]
26]
27]
M. Hesse, H. Meier, B. Zech, Spectroscopic Methods in Organic
Chemistry, Stuttgart, Germany, 1997.
K. L. Bray, I. J. S. Fairlamb, G. C. Lloyd-Jones, Chem. Com-
mun. 2001, 187Ϫ188.
K. L. Bray, I. J. S. Fairlamb, P. J. Kaiser, G. C. Lloyd-Jones, P.
A. Slatford, unpublished results.
Chiu, Tetrahedron Lett. 1975, 499Ϫ502.
18]
The following sets of conditions (all in THF) were tested with
1
6 without more than a trace (Ͻ 1%) of desired diene 1 being
3 3
detectable by TLC: a) with [PPh CH ][I], BuLi, Ϫ78 °C to
room temp. or KOtBu Ϫ78 °C to room temp., or NaH, room
temp. or KH, 18-crown-6, 0 °C to room temp.; b) with
28] [28a]
Ph
2
P(O)CH
3
, NaH, 0 °C or KOtBu, Ϫ78 °C to room temp.;
J. Krause, G. Cestaric, K.-J. Haack, K. Seevogel, W. Storm,
[28b]
c) with (EtO)
2
P(O)CH
3
, NaH, 0 °C or KOtBu, Ϫ78 °C to
K.-R. Pörschke, J. Am. Chem. Soc. 1999, 121, 9807Ϫ9823.
room temp.
M. G o´ mez Andreu, A. Zapf, M. Beller, Chem. Commun. 2000,
[
19]
For example, refluxing a THF solution of the ketone 19 with
two equivalents of the ylide generated from [PPh CH ][I] and
in press.
[29]
3
3
S. Hashiguchi, Y. Maeda, S. Kishimoto, M. Ochiai, Hetero-
BuLi gave no trace (Ͻ 1%) of the desired alkene 2. Even though
the red colour of the ylide was fully discharged after 24 h at re-
flux.
cycles 1986, 24, 2273Ϫ2283.
Received September 25, 2000
[O00469]
1642
Eur. J. Org. Chem. 2001, 1635Ϫ1642