Mn(III)-based oxidative free-radical cyclizations of alkenyl Meldrum's acids

Article abstract of DOI:10.1016/S0040-4020(01)01054-7

Oxidative cyclization of unsaturated Meldrum's acids can be carried out at temperatures as low as -30°C. The rate-determining step is cyclization of the enolate to the alkene (11 to 14 and 15) rather than enolization, which is the rate-determining step with dimethyl 4-pentenylmalonate (1). While cyclization of 1 gives mainly cyclopentanes, cyclization of Meldrum's acids provides a versatile route to cyclohexenes.

Full text of DOI:10.1016/S0040-4020(01)01054-7

TETRAHEDRON
Pergamon
Tetrahedron 58 +2002) 25±34
MnꢀIII)-based oxidative free-radical cyclizations of alkenyl
Meldrum's acids
Barry B. Sniderp and Rachel B. Smith
Department of Chemistry MS015, Brandeis University, Waltham, MA 02454-9110, USA
Received 24 September 2001; accepted 1 November 2001
AbstractÐOxidative cyclization of unsaturated Meldrum's acids can be carried out at temperatures as low as 2308C. The rate-determining
step is cyclization of the enolate to the alkene +11 to 14 and 15) rather than enolization, which is the rate-determining step with dimethyl
4-pentenylmalonate +1). While cyclization of 1 gives mainly cyclopentanes, cyclization of Meldrum's acids provides a versatile route to
cyclohexenes. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
cyclizes to give a ,9:1 mixture of cyclopentanemethyl
radical 5 and cyclohexyl radical 6. Reaction of 5 with
Cu+OAc)2 gives Cu+III) intermediate 4 which undergoes
oxidative elimination to give methylenecyclopentane 7
and ligand transfer to give lactone 8. A similar oxidation
converts 6 to cyclohexene 9. Reaction in AcOH +558C, 28 h)
affords a 2.5:1 mixture of 8 and 7. Reaction in EtOH +608C,
156 h) is much slower, but yields a 2:3 mixture of 8 and 7,
while reaction in DMSO +758C, 68 h) forms a 1:10 mixture
of 8 and 7.
We have extensively developed Mn+OAc)3-based oxidative
free-radical cyclizations of unsaturated carbonyl and
dicarbonyl compounds as a versatile procedure for generat-
ing a radical that can cyclize by enolization and oxidation of
the enolate. Oxidation of the cyclic radical by Cu+OAc)2
generates the least substituted alkene regiospeci®cally.1
Enolization of dimethyl 4-pentenylmalonate +1) to give
manganese enolate 2 is the slow, rate-determining step in
the oxidative cyclization with Mn+OAc)3 +Scheme 1).2±4
Rapid loss of Mn+II) from 2 generates radical 3, which
We thought the cyclization of alkenyl Meldrum's acid 10
might occur more rapidly and at lower temperatures since
Meldrum's acid is more acidic than dimethyl malonate by
,8.5 pKa units in either H2O or DMSO.5 Selective mono-
alkylation of Meldrum's acid is dif®cult; the dialkylated
product usually predominates.6,7 Fortunately, 10 can be
easily prepared by a modi®cation of the literature pro-
cedure.8
Ethylenediammonium
diacetate-catalyzed
Knoevenagel condensation of Meldrum's acid and
4-pentenal for 1 h at 258C followed by addition of
BH3´NHMe2 and stirring for 18 h affords 69% of 10. This
one-pot reaction was used with the appropriate aldehyde or
ketone to prepare all the cyclization substrates in 50±70%
yield.9
2. Results and discussion
Oxidative cyclization of 10 as a 0.1 M solution in EtOH with
2 equiv. of Mn+OAc)3´2H2O and 1 equiv. of Cu+OAc)2´H2O
for 10 min at 258C provides 8% of methylenecyclopentane
13, 65% of cyclohexene 16,10 and 4% of cyclohexene 17
+Scheme 2). Oxidation of cyclohexyl radical 15 with Cu+II)
selectively removes the least hindered proton to form
mainly cyclohexene 16.1a Similar results are obtained in
AcOH, with 3 equiv. of KOAc added to minimize
Scheme 1. Oxidative cyclization of malonate 1.
Keywords: cyclisation; manganese; copper; radical reactions; cyclo-
hexenes.
p
Corresponding author. Tel.: 11-781-736-2550; fax: 11-781-736-2516;
e-mail: snider@brandeis.edu
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020+01)01054-7

26
B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
hydrolysis of the Meldrum's acid, except that 5±10% of the
lactone analogous to 8 is obtained instead of 13.
and form enolates that have a large oxidation potential.4a
Formation of the enolate is the rate-determining step with
a-substituted b-keto esters, which are less acidic and form
enolates that have a lower oxidation potential.
Bausch's studies of the oxidation potential of the enolates of
dimethyl malonate and Meldrum's acid in DMSO support
this interpretation.11 The pKas of Meldrum's acid and
dimethyl malonate in DMSO are 7.3 and 15.9, respectively,
indicating that enolization of Meldrum's acid is favored by
11.8 kcal mol21 +Eqs. +1) and +2)). On the other hand the
oxidation potentials of the enolates are 1.16 and 0.77 V,
respectively, indicating that it is 9 kcal mol21 easier to
oxidize the enolate of dimethyl malonate. Combining
these two reactions establishes that formation of the radical
from Meldrum's acid should be favored by 2.8 kcal mol21
.
Scheme 2. Oxidative cyclization of Meldrum's acid 10.
ꢀ1†
ꢀ2†
As expected, oxidative cyclization of 10, with a half life in
EtOH of ,5 min at 258C and 2 h at 2308C, is much faster
than that of the less acidic dimethyl ester 1, which requires
156 h at 608C for complete reaction. More surprisingly, 10
gives predominantly products derived from cyclohexyl
radical 15, while 1 gives mainly products derived from
cyclopentanemethyl radical 5 suggesting that these two
reactions are mechanistically distinct.
Loss of a proton from 1 to give enolate 2 should be slow,
with rapid loss of Mn+II) to give acyclic radical 3. On the
other hand, loss of a proton from 10 to give enolate 11
should be fast. However, loss of Mn+II) to give acyclic
radical 12 will be slow because the oxidation potential of
11 is large, so that cyclization of 11 to 14 and 15 is the
rate-determining step. Since these two cyclizations are
mechanistically distinct, there is no contradiction in the
preferential formation of cyclopentanemethyl radical 5
from 1 and cyclohexyl radical 15 from 10.
The oxidative cyclization of either the cis or trans isomer of
18 to provide 33% of hexenylcyclopentane 19 and 21% of
cyclohexene 20 requires 8 h at 258C with a half-life of 2 h.
The change in half life from ,5 min with 10 to 2 h with 18
clearly indicates that the double bond participates in the
rate-determining step. Loss of a proton from 10 to give
Mn+III) enolate 11 should be rapid and reversible. Loss of
Mn+II) from enolate 11 to form radical 12 cannot be
occurring because this reaction would proceed at the same
rate with 11 and the enolate derived from 18 +Scheme 3).
Selectivity for cyclohexene 16 can be increased by carrying
out the cyclization with only 0.05 equiv. of Cu+OAc)2´H2O.
Under these conditions we obtain only 0.5% of methylene-
cyclopentane 13, 74% of cyclohexene 16 and 3% of cyclo-
hexene 17. At low Cu+II) concentration, radicals 14 and 15
have longer life times and equilibrate prior to oxidation,
presumably by opening to acyclic radical 12. Under thermo-
dynamic control, the cyclohexyl radical 15 should be
strongly preferred. Even though radical 12 is not formed
from enolate 11, Bausch's studies indicate that 12 is more
stable than 3, so that equilibration of 14 and 15 via 12 should
occur if trapping of the radical is slow.12
Scheme 3.
Direct cyclization of the alkene of 11 with the Mn+III)
enolate to give cyclic radicals 14 and 15 is the slow, rate-
determining step in the oxidative cyclization of these
Meldrum's acid derivatives, while formation of enolate 2
is the slow, rate-determining step in the oxidative cycliza-
tion of 1. We have previously observed analogous shifts in
mechanism as a function of the acidity of the dicarbonyl
compound and the oxidation potential of the enolate. Direct
cyclization of the enolate is the rate-determining step with
a-unsubstituted b-keto esters, which are relatively acidic
To summarize, these mechanistic studies clearly indicate
that the oxidative cyclization of very acidic alkenyl
Meldrum's acids 10 and 18 is quite different from that of
4-pentenylmalonate diester 1. Formation of enolate 11 is
rapid while formation of enolate 2 is slow. Since the half
life for oxidative cyclization of 18 is .25 times slower than
for the cyclization of 10, the double bond must be involved
in the rate determining step, which is probably the cycliza-
tion of 11 to give cyclic radicals 14 and 15 with loss of
Mn+II) without the intermediacy of acyclic radical 12

B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
27
+Eq. +3)). Although this cyclization does not ®t into standard
mechanistic schemes, the kinetic data clearly indicate that
radical 12 is not an intermediate. It is tempting to suggest
that the alkene coordinates to Mn+III) or Cu+II) prior to
cyclization. However, coordination of alkenes to these
metals in high oxidation states is unlikely.
ꢀ3†
Scheme 4.
are also obtained indicating that the hindered tertiary radical
reacts with Cu+II) slowly so that hydrogen abstraction
from EtOH is competitive. With only 0.05 equiv. of
Cu+OAc)2´H2O, 35cis +28%) and 35trans +14%) are the
major products, with only traces of 34cis and 34trans. No
saturated products are formed in AcOH, but the overall
yields are lower. Hydrogenation of 34 over Pd/C or Rh/
Al2O3 gives predominantly dihydro 33, indicating that
cleavage to the Meldrum's acid anion and an allylic organo-
metallic, which is hydrogenated, is faster than hydrogena-
tion of the double bond14 +Scheme 5).
Oxidation of radicals 14 and 15 proceeds by formation of
Cu+III) intermediates with a second order rate constant of
approximately 106 M21 s21. If 1.0 equiv. of Cu+OAc)2 is
used the radicals are trapped rapidly so that the ratio of
products re¯ects the ratio of radicals formed in the cycliza-
tion. If only 0.05 equiv. of Cu+OAc)2 is used the radicals
have a longer lifetime and can equilibrate by reversible ring
opening to acyclic radical 12.12 This process will lead to
higher concentrations of the more stable cyclohexyl radical
15, which will lead to increased yields of 16 and 17 at the
expense of methylenecyclopentane 13.
Oxidative cyclization of 21 with 2 equiv. of Mn+OAc)3´
2H2O and only 0.05 equiv. of Cu+OAc)2´H2O for 10 min
at 258C proceeds analogously to give 70% of cyclohexene
22, 8% of cyclohexene 23, and 1% of the methylenecyclo-
pentane analogous to 13. However, oxidative cyclization of
248c with 2 equiv. of Mn+OAc)3´2H2O and 1 equiv. of
Cu+OAc)2´H2O requires 2.5 h at 258C and gives 49% of
methylenecyclopentane 25 and 23% of ethoxymethylcyclo-
pentane 26. No cyclohexene was observed; ,4% of
the cyclohexene was obtained with only 0.05 equiv. of
Cu+OAc)2´H2O, which should favor equilibration of the
radicals. Severe 1,3-diaxial interactions between
the geminal dimethyl group and the carbonyl groups of
the Meldrum's acid make the formation of cyclohexyl
radical 27 kinetically and thermodynamically disfavored.
The hydrogen in Cu+III) intermediate 29 is very hindered.
Therefore, ligand transfer with solvent to give 26 competes
with oxidative elimination to give the expected methylene-
cyclopentane 2513 +Scheme 4).
Oxidative cyclization of 30 is rapid, giving 43% of methyl-
enecyclohexane 31 and 23% of methylcyclohexene 32,
both of which are derived from the tertiary cyclohexyl
radical. Oxidative cyclization of 33 with 2 equiv. of
Mn+OAc)3´2H2O and 2 equiv. of Cu+OAc)2´H2O affords
exclusively the tertiary cyclopentanemethyl radical, which
is oxidized by Cu+OAc)2 to give 29% of 34cis and 21% of
34trans. Saturated products 35cis +7%) and 35trans +3%)
Scheme 5.
Cyclization of cyclohexene 36 affords a mixture of 37 and
38 and compounds that appear to be dimers. The cyclization
should be slow since the cyclohexene must adopt an
unfavorable conformation with a pseudoaxial side chain.
To prevent dimerization, we added 36 over 12 h to the

28
B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
Scheme 8.
somewhat lower than those of the comparable 4-alkenyl
Meldrum's acids. Cyclizations of 5-alkenyl Meldrum's
acids will be slower since they have a more negative entropy
of cyclization, while dimerization and other side reactions
will occur at the same rate +Scheme 9).
Scheme 6.
oxidants. Under these conditions we obtain 20% of 37,10
42% of 38, and no dimers. The dimerization was investi-
gated in detail with methyl Meldrum's acid +39), which
gives 21% of C±C dimer 40,15a 38% of C±O dimer 41,
and 10% of 42, resulting from addition of 41 to methylene
Meldrum's acid formed by oxidation of 3915 +Scheme 6).
Attempted tandem cyclization of 43 yields 5% of
methylenecyclopentane 44 and 71% of cyclohexene 45.
Cyclohexyl radical 46eq must ¯ip to 46ax with an axial
allyl side chain for tandem cyclization to occur. 1,3-Diaxial
interactions make this energetically unfavorable, so that it is
slower than oxidation of 46eq to 45, even if only 0.05 equiv.
of Cu+OAc)2´H2O is used +Scheme 7).
Scheme 9.
In conclusion, we have developed a facile oxidative
cyclization of unsaturated Meldrum's acid derivatives that
can be carried out at temperatures as low as 2308C. The
rate-determining step is cyclization of the enolate to the
alkene +11 to 14 and 15) rather than enolization, which is
the rate-determining step with dimethyl 4-pentenylmalonate
+1). While cyclization of 1 gives mainly cyclopentanes,
cyclization of Meldrum's acid derivatives provides a
versatile route to cyclohexenes that should be broadly
applicable in organic synthesis.
3. Experimental
3.1. General
Scheme 7.
NMR spectra were recorded at 400 MHz in CDCl3 unless
otherwise indicated. Chemical shifts are reported in d and
Tandem oxidative cyclization of 47 provides 67% of per-
hydroindane 49 as the only product. The second cyclization
of cyclohexyl radical 48 is fast since a fused, rather than a
bridged, ring system is formed and cyclization can therefore
occur from either chair conformer of 48 +Scheme 8).
coupling constants in Hz. IR spectra are reported in cm21
.
Most of these compounds did not give parents by CI mass
spectroscopy. The highest mass peak was usually M-58
+acetone) resulting from the facile fragmentation of the
Meldrum's acid.16
Cyclizations of 5-alkenyl Meldrum's acids can also be
successfully carried out to give cycloheptenes and unsatu-
rated cyclohexanes. Oxidative cyclization of 50 affords 37%
of methylenecyclohexane 51 and 19% of cycloheptene 52.
Oxidative cyclization of 53 provides 33% of propenylcyclo-
hexane 54. As expected, the yields of these cyclizations are
3.2. Preparation of alkenyl Meldrum's acids
3.2.1. 2,2-Dimethyl-5-ꢀ4-pentenyl)-1,3-dioxane-4,6-dione
ꢀ10). Ethylene diammonium diacetate +EDDA) +248 mg,
1.4 mmol) was added to a solution of Meldrum's acid

B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
29
+1.188 g, 8.5 mmol) and 4-pentenal +460 mg, 5.5 mmol) in
absolute EtOH +6.6 mL) at rt and the resulting solution was
stirred for 1 h. BH3´NH+CH3)2 +326 mg, 5.5 mmol) was
added to the reaction, which was stirred for 17.5 h. Water
+35 mL) and 5% aqueous HCl solution +3 mL) were added
to the reaction, which was extracted with CH2Cl2
+3£20 mL). The combined organic layers were dried
+MgSO4) and concentrated to give 1.47 g of crude product.
Flash chromatography +9:1 hexanes/EtOAc containing
0.5% HOAc) gave 805 mg +69%) of pure 10 as a white
28.8, 26.5 +2C), 26.4. The 1H NMR spectrum is identical to
that previously reported.8c
3.2.5.
2,2-Dimethyl-5-ꢀ1,4-dimethyl-4-pentenyl)-1,3-
dioxane-4,6-dione ꢀ30). Reaction of EDDA +217.2,
1.2 mmol), Meldrum's acid +691 mg, 4.8 mmol) and
5-methyl-5-hexen-2-one +539 mg, 4.8 mmol) in absolute
EtOH +6 mL) followed by reduction with BH3´NH+CH3)2
+295 mg, 5.0 mmol) and workup and chromatography as
described earlier gave 634 mg +55%) of pure 30 as a color-
1
1
solid: mp 48±508C; H NMR 5.80 +ddt, 1, Jˆ17.1, 10.4,
less oil which solidi®ed upon standing: mp 52±538C; H
6.1 Hz), 5.04 +d, 1, Jˆ17.1 Hz), 4.99 +d, 1, Jˆ10.4 Hz),
3.52 +t, 1, Jˆ4.9 Hz), 2.17±2.10 +m, 4), 1.79 +s, 3), 1.76
+s, 3), 1.61±1.53 +m, 2); 13C NMR 165.5 +2C), 137.7, 115.2,
104.8, 46.1, 33.5, 28.4, 26.9, 26.1, 25.6. This compound
has previously been synthesized by reaction of allyltri-
methylsilane with 6,6-dimethyl-5,7-dioxaspiro[2.5]octane-
4,8-dione.17
NMR 4.73 +br s, 1), 4.70 +br s, 1), 3.50 +d, 1, Jˆ3.1 Hz),
2.62±2.59 +dtq, 1, Jˆ3.1, 7.0, 6.7 Hz), 2.17±1.99 +m, 2),
1.78±1.72 +m, 2), 1.76 +s, 3), 1.75 +s, 3), 1.73 +s, 3), 1.13 +d,
3, Jˆ6.7 Hz); 13C NMR 165.4, 164.7, 145.2, 110.5, 104.6,
50.5, 35.9, 33.5, 31.3, 28.2, 27.3, 22.2, 16.3. Anal. Calcd for
C13H20O4: C, 64.98; H, 8.39. Found: C, 64.74; H, 8.43.
3.2.6. 2,2-Dimethyl-5-ꢀ1,5-dimethyl-4-hexenyl)-1,3-diox-
ane-4,6-dione ꢀ33). Reaction of EDDA +144 mg,
0.8 mmol), Meldrum's acid +571 mg, 4.0 mmol) and
6-methyl-5-hepten-2-one +496 mg, 3.9 mmol) in absolute
EtOH +5.1 mL) followed by reduction with BH3´NH+CH3)2
+246 mg, 4.2 mmol) and workup and chromatography as
described earlier gave 534 mg +53%) of pure 33 as a color-
3.2.2.
5-ꢀcis-4-Decenyl)-2,2-dimethyl-1,3-dioxane-4,6-
dione ꢀ18cis). Reaction of EDDA +88 mg, 0.49 mmol),
Meldrum's acid +421 mg, 2.9 mmol), and cis-4-decenal
+301 mg, 1.96 mmol) in absolute EtOH +4.0 mL) followed
by reduction with BH3´NH+CH3)2 +127 mg, 2.1 mmol)
and workup and chromatography as described earlier
gave 160 mg +54%) of pure 18cis as an oil: 1H NMR
5.44±5.31 +m, 2), 3.51 +t, 1, Jˆ5.2 Hz), 2.15±1.95 +m, 6),
1.79 +s, 3), 1.76 +s, 3), 1.57±1.49 +m, 2), 1.37±1.26 +m, 6),
0.89 +t, 3, Jˆ7.0 Hz); 13C NMR 165.6 +2C), 131.0, 128.4,
104.8, 46.1, 31.5, 29.3, 28.5, 27.2, 27.0, 26.9, 26.5, 26.3,
22.6, 14.1.
1
less oil: H NMR 5.11 +t, 1, Jˆ7.3 Hz), 3.47 +d, 1, Jˆ
3.1 Hz), 2.64±2.58 +m, 1), 2.11±1.96 +m, 2), 1.75 +s, 3),
1.74 +s, 3), 1.71±1.57 +m, 2), 1.68 +s, 3), 1.61 +s, 3), 1.11
+d, 3, Jˆ7.3 Hz); 13C NMR 165.6, 164.8, 132.2, 123.7,
104.6, 50.4, 33.6 +2C), 28.2, 27.4, 26.2, 25.7, 17.7, 16.4.
3.2.7. 5-ꢀ3-Cyclohexenylmethyl)-2,2-dimethyl-1,3-diox-
ane-4,6-dione ꢀ36). Reaction of EDDA +72 mg,
0.4 mmol), Meldrum's acid +251 mg, 1.7 mmol), and
3-cyclohexenecarboxaldehyde +179 mg, 1.6 mmol) in
absolute EtOH +2 mL) followed by reduction with
BH3´NH+CH3)2 +101 mg, 1.7 mmol) and workup and chro-
matography as described earlier gave 226 mg +56%) of pure
36 as a white solid: mp 113±1148C; 1H NMR 5.70±5.62 +m,
2), 3.51 +t, 1, Jˆ6.1 Hz), 2.18±1.69 +m, 8), 1.81 +s, 3), 1.77
+s, 3), 1.36±1.26 +m, 1); 13C NMR 165.9 +2C), 127.0, 125.5,
104.8, 43.6, 33.0, 31.3, 31.2, 28.5, 28.4, 26.7, 24.8. Anal.
Calcd for C13H18O4: C, 65.53; H, 7.61. Found: C, 65.07; H,
7.65.
3.2.3. 2,2-Dimethyl-5-ꢀ1-methyl-4-pentenyl)-1,3-dioxane-
4,6-dione ꢀ21). Reaction of EDDA +203 mg, 1.1 mmol),
Meldrum's acid +844 mg, 5.9 mmol) and 5-hexen-2-one
+548 mg, 5.6 mmol) in absolute EtOH +7.0 mL) followed
by reduction with BH3´NH+CH3)2 +342 mg, 5.8 mmol) and
workup and chromatography as described earlier gave
490 mg +39%) of pure 21 as a white solid, followed by
230 mg of 73% pure 21 contaminated with the product of
condensation of Meldrum's acid with acetone derived from
hydrolysis of Meldrum's acid. The overall yield of 21 was
1
52%: mp 64±658C; H NMR 5.78 +ddt, 1, Jˆ10.4, 17.1,
6.7 Hz), 5.02 +dd, 1, Jˆ1.8, 17.1 Hz), 4.97 +dd, 1, Jˆ1.8,
10.4 Hz), 3.44 +d, 1, Jˆ2.4 Hz), 2.63±2.57 +m, 1), 2.21±
2.12 +m, 1), 2.09±2.00 +m, 1), 1.78±1.61 +m, 2), 1.74 +s,
3), 1.73 +s, 3), 1.10 +d, 3, Jˆ6.7 Hz); 13C NMR 165.5, 164.7,
138.0, 115.1, 104.7, 50.5, 33.4, 32.7, 31.9, 28.2, 27.4, 16.3.
Anal. Calcd for C12H18O4: C, 63.70; H, 8.02. Found: C,
63.68; H, 8.10.
3.2.8. 2-ꢀ2-Propenyl)-4-pentenal ꢀ43a). Pyridine +5.2 mL,
64 mmol) was added to a solution of CrO3 +3.19 g,
31.9 mmol) in CH2Cl2 +25 mL) at 08C. The mixture was
stirred for 5 min at this temperature and allowed to warm
to 258C over 15 min. 2-+2-Propenyl)-4-penten-1-ol18
+574 mg, 4.6 mmol) in CH2Cl2 +2 mL) was added by
cannula. The reaction was stirred at 258C for 8 h and
decanted. The ¯ask was washed with three portions of
ether. The combined organic layers were washed sequen-
tially with 5% aqueous NaOH, 5% aqueous HCl, satd
CuSO4, water, and satd NaHCO3, dried +MgSO4) and
concentrated to give 469 mg +83%) of crude aldehyde 43a
which was used without puri®cation.
3.2.4.
2,2-Dimethyl-5-ꢀ2,2-dimethyl-4-pentenyl)-1,3-
dioxane-4,6-dione ꢀ24). Reaction of EDDA +154 mg,
0.85 mmol), Meldrum's acid +737 mg, 5.1 mmol), and 2,2-
dimethyl-4-pentenal +90%, 425 mg, 3.4 mmol) in absolute
EtOH +13.6 mL) followed by reduction with BH3´NH+CH3)2
+221 mg, 3.7 mmol) and workup and chromatography as
described earlier gave 555 mg +68%) of pure 248c as a
1
waxy solid: mp 38±408C; H NMR 5.88 +ddt, 1, Jˆ10.3,
17.7, 7.5 Hz), 5.07 +br d, 1, Jˆ10.3 Hz), 5.05 +br d, 1,
Jˆ17.7 Hz), 3.42 +t, 1, Jˆ4.1 Hz), 2.16 +d, 2, Jˆ4.1 Hz),
2.04 +d, 2, Jˆ7.5 Hz), 1.84 +s, 3), 1.77 +s, 3), 0.95 +s, 6); 13C
NMR 166.1 +2C), 134.7, 117.7, 104.5, 46.7, 43.1, 36.5, 33.2,
3.2.9. 2,2-Dimethyl 5-ꢀ2-propenyl-4-pentenyl)-1,3-diox-
ane-4,6-dione ꢀ43). Reaction of EDDA +165 mg,
0.91 mmol), Meldrum's acid +985 mg, 6.84 mmol) and
crude aldehyde 43a +469 mg, 3.79 mmol) in absolute

30
B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
EtOH +45 mL) followed by reduction with BH3´NH+CH3)2
+272 mg, 4.6 mmol) and workup and chromatography as
described earlier gave 432 mg +38% for 2 steps) of pure
+115 mL, 0.69 mmol) in absolute EtOH +1.6 mL), followed
by reduction with BH3´NH+CH3)2 +41 mg, 0.69 mmol) and
workup and chromatography as described above gave
1
1
43 as a colorless oil: H NMR 5.77 +ddt, 2, Jˆ17.1, 10.4,
110 mg +60%) of pure 53 as an oil: H NMR 5.39±5.27
6.7 Hz), 5.03 +d, 2, Jˆ17.1 Hz), 5.02 +d, 2, Jˆ10.4 Hz), 3.68
+t, 1, Jˆ6.1 Hz), 2.15±2.02 +m, 7), 1.75 +s, 3), 1.71 +s, 3);
13C NMR 165.7 +2C), 136.5 +2C), 116.8 +2C), 104.8, 44.3,
38.3 +2C), 34.5, 30.9, 28.6, 26.5.
+m, 2), 3.49 +t, 1, Jˆ5.2 Hz), 2.14±1.97 +m, 6), 1.79 +s,
3), 1.76 +s, 3), 1.49±1.39 +m, 4), 0.95 +t, 3, Jˆ7.6 Hz);
13C NMR 165.6 +2C), 132.0, 128.6, 104.8, 46.1, 29.6,
28.4, 27.0, 26.7, 26.6, 26.2, 20.5, 14.3.
3.2.10. 3-Ethenyl-3-methyl-6-heptenal ꢀ47a). 3-Methyl-2-
hepten-1-ol19 +1.726 g, 13.7 mmol) in THF +17 mL) was
added by cannula to a slurry of NaH +60%, 658 mg,
16.5 mmol) in THF +9 mL) and the resulting solution was
stirred for 30 min. +E)-+Carboxyvinyl)trimethylammonium
betaine20 +2.3 g, 17.8 mmol) was added and the reaction was
heated at re¯ux for 15.5 h. The reaction was allowed to cool
and water +60 mL) and ether +20 mL) were added. The
layers were separated and the aqueous layer was washed
with ether +2£25 mL), acidi®ed to pH 1 and extracted
with ether +3£75 mL). The organic layers were dried and
concentrated to give 1.02 g +38%) of 3-+3-methyl-2,6-hepta-
dienyloxy)-propenoic acid. Bulb to bulb distillation
+160±1808C, 4 mm Hg) gave 748 mg +78%) of 47a,
which was used without further puri®cation: 1H NMR
9.74 +t, 1, Jˆ3.4 Hz), 5.85 +dd, 1, Jˆ17.7, 11.0 Hz), 5.78
+ddt, 1, Jˆ17.7, 9.8, 6.7 Hz), 5.13 +d, 1, Jˆ11.0 Hz), 5.02
+d, 1, Jˆ17.7 Hz), 5.01 +dd, 1, Jˆ17.7, 1.5 Hz), 4.95 +d, 1,
Jˆ9.8 Hz), 2.41 +dd, 1, Jˆ3.4, 15.0 Hz), 2.33 +dd, 1, Jˆ3.4,
15.0 Hz), 2.03±1.97 +m, 2), 1.53±1.48 +m, 2), 1.16 +s, 3).
3.3. Oxidative cyclization of alkenyl meldrum's acids
3.3.1. Oxidative cyclization of 10. A solution of Cu+OAc)2´
H2O +23 mg, 0.11 mmol) in EtOH +0.8 mL) was heated to
re¯ux to ensure complete dissolution and cooled to 258C.
Mn+OAc)3´2H2O +61 mg, 0.22 mmol) was added and N2
was bubbled through the resulting solution for 15 min.
Meldrum's acid 10 +24 mg, 0.11 mmol) in 0.2 mL of
EtOH was added by cannula and the addition ¯ask was
rinsed with EtOH +0.1 mL) which was also added to the
reaction. The reaction was stirred for 10 min and quenched
with several drops of 10% aqueous NaHSO3 solution and
3 mL of H2O. The reaction was extracted with CH2Cl2
+3£10 mL), dried +MgSO4) and concentrated to give
21 mg of a crude mixture of 13, 16, and 17. Puri®cation
by ¯ash chromatography +9:1 hexanes/EtOAc) gave
0.5 mg of 13, followed by 3.5 mg of a 1:7 mixture of 13
and 16, 11 mg of a 2:20:1 mixture of 13, 16, and 17 and
3 mg of a 3:1 mixture of 16 and 17. The overall yield of 13
was 8%, of 16 was 63% and of 17 was 5%.
3.2.11. 5-ꢀ3-Ethenyl-3-methyl-6-heptenyl)-2,2-dimethyl-
1,3-dioxane-4,6-dione ꢀ47). Reaction of EDDA +118 mg,
0.65 mmol), Meldrum's acid +750 mg, 5.2 mmol) and
crude aldehyde 47a +660 mg, 4.3 mmol) in absolute EtOH
+10 mL) followed by reduction with BH3´NH+CH3)2
+269 mg, 4.6 mmol) and workup and chromatography as
described earlier gave 585 mg +49%) of pure 47 as an oil:
1H NMR 5.80 +ddt, 1, Jˆ10.4, 17.1, 6.7 Hz), 5.70 +dd, 1,
Jˆ17.7, 11.0 Hz), 5.05 +dd, 1, Jˆ11.0, 1.2 Hz), 5.00 +dd, 1,
Jˆ17.7, 1.2 Hz), 4.96 +dd, 1, Jˆ17.1, 1.2 Hz), 4.91 +dd, 1,
Jˆ10.4, 1.2 Hz), 3.52 +d, 1, Jˆ4.9 Hz), 2.07±1.95 +m, 4),
1.78 +s, 3), 1.76 +s, 3), 1.46±1.37 +m, 4), 1.01 +s, 3); 13C
NMR 165.4 +2C), 146.0, 139.2, 114.0, 112.6, 104.7, 46.2,
39.9, 39.3, 36.6, 28.4, 28.3, 26.8, 22.1, 21.5.
Data for 8,8-dimethyl-1-methylene-7,9-dioxaspiro[4.5]de-
1
cane-6,10-dione +13): H NMR 5.20 +s, 1), 5.06 +d, 1, Jˆ
1.2 Hz), 2.65 +tt, 2, Jˆ1.2, 7.3 Hz), 2.47 +t, 2, Jˆ6.7 Hz),
2.11 +tt, 2, Jˆ7.3, 6.7 Hz), 1.81 +s, 3), 1.72 +s, 3); 13C NMR
+partial) 170.0, 154.5, 110.1, 104.9, 37.5, 34.1, 25.6.
Data for 3,3-dimethyl-2,4-dioxaspiro[5.5]undec-8-ene-1,5-
dione +16): H NMR 5.90±5.83 +m, 1), 5.77±5.71 +m, 1),
1
2.66±2.63 +m, 2), 2.27±2.22 +m, 2), 2.17 +br t, 2, Jˆ6.1 Hz),
1.77 +s, 3), 1.74 +s, 3); 13C NMR 169.6 +2C), 125.7, 122.3,
104.8, 47.3, 30.7, 30.5, 29.6, 28.4, 21.6. This compound has
been previously made by a Diels±Alder reaction between
methylene Meldrum's acid and butadiene.10
Partial data for 3,3-dimethyl-2,4-dioxaspiro[5.5]undec-7-
ene-1,5-dione +17) was obtained from the mixture: 1H
NMR 6.26 +dt, 1, Jˆ9.8, 3.7 Hz), 5.55 +br d, 1, Jˆ
9.8 Hz), 1.97 +tt, 2, Jˆ6.1, 3.3 Hz), 1.77 +s, 3).
3.2.12. 5-ꢀ5-Hexenyl)-2,2-dimethyl-1,3-dioxane-4,6-dione
ꢀ50). Reaction of EDDA +86 mg, 0.48 mmol), Meldrum's
acid +455 mg, 3.2 mmol) and 5-hexenal +310 mg,
3.2 mmol), prepared by Collins' oxidation of 5-hexen-1-
ol, in absolute EtOH +4 mL) followed by reduction with
BH3´NH+CH3)2 +196 mg, 3.3 mmol) and workup and chro-
matography as described earlier gave 407 mg +57%) of pure
55 as a white solid: mp 448C; 1H NMR 5.80 +ddt, 1, Jˆ10.4,
17.1, 6.7 Hz), 5.00 +dd, 1, Jˆ17.1, 1.2 Hz), 4.95 +dd, 1,
Jˆ10.4, 1.2 Hz), 3.50 +t, 1, Jˆ5.0 Hz), 2.14±2.07 +m, 4),
1.79 +s, 3), 1.76 +s, 3), 1.53±1.39 +m, 4); 13C NMR 165.5
+2C), 138.4, 114.6, 104.8, 46.1, 33.3, 28.7, 28.4, 26.9, 26.5,
26.0. Anal. Calcd for C12H18O4: C, 63.70; H, 8.02. Found: C,
63.64; H, 8.13.
3.3.2. Oxidative cyclization of 18cis. Oxidative cyclization
of 18cis +99 mg, 0.35 mmol) with Mn+OAc)3´2H2O +188
mg, 0.7 mmol) and Cu+OAc)2´H2O +71 mg, 0.35 mmol) in
EtOH +3.5 mL) for 8 h followed by workup and chroma-
tography as described above gave 33% +33 mg) of 19
followed by 22% +21 mg) of 20.
Data for 1-+1E-hexenyl)-8,8-dimethyl-7,9-dioxaspiro[4.5]-
decane-6,10-dione +19): 1H NMR 5.56 +dt, 1, Jˆ15.6,
6.7 Hz), 5.28 +dd, 1, Jˆ15.6, 9.2 Hz), 3.28±3.22 +m, 1),
2.38±2.24 +m, 2), 2.05±1.82 +m, 6), 1.66 +s, 6), 1.32±1.19
+m, 4), 0.83 +t, 3, Jˆ6.7 Hz); 13C NMR 172.4, 169.5, 135.5,
126.9, 104.6, 59.0, 58.1, 38.5, 33.3, 32.0, 31.2, 29.7, 28.5,
25.5, 22.0, 13.8.
3.2.13. 2,2-Dimethyl-5-ꢀcis-5-octenyl)-1,3-dioxane-4,6-
dione ꢀ53). Reaction of EDDA +25 mg, 0.14 mmol),
Meldrum's acid +100 mg, 0.69 mmol), and cis-5-octenal

B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
31
Data for 3,3-dimethyl-7-pentyl-2,4-dioxaspiro[5.5]undec-8-
ene-1,5-dione +20): 1H NMR 5.83±5.78 +m, 1), 5.65 +dd, 1,
Jˆ10.4, 1.8 Hz), 3.00 +m, 1), 2.31±2.09 +m, 4), 1.73 +s, 3),
1.72 +s, 3), 1.48±1.19 +m, 8), 0.84 +t, 3, Jˆ7.3 Hz); 13C
NMR 171.1, 165.7, 126.6, 125.3, 104.8, 52.3, 42.2, 31.6,
31.25, 31.20, 29.6, 28.9, 26.8, 22.4, 22.3, 13.9.
previously gave 71 mg of crude product. Puri®cation by
¯ash chromatography +14:1 hexanes/EtOAc) gave 4.6 mg
+6%) of pure 31 followed by 43 mg of a 3:2 mixture of 31
and 32 +an overall yield of 43% of 31 and 23% of 32).
Data for 3,3,7-trimethyl-10-methylene-2,4-dioxaspiro[5.5]-
undecane-1,5-dione +31): mp 85.1±86.18C; H NMR 4.87
1
3.3.3. Oxidative cyclization of 21. Oxidative cyclization of
21 +116 mg, 0.5 mmol) with Mn+OAc)3´2H2O +275 mg, 1.0
mmol) and Cu+OAc)2´H2O +5 mg, 0.03 mmol) in EtOH
+5 mL) for 10 min followed by workup and chromatography
as described above gave 1 mg +1%) of the methylenecyclo-
pentane, followed by 89 mg of a 9:1 mixture of 22 and 23,
respectively, giving an overall yield of 1 mg +1%) of the
methylenecyclopentane, 80 mg +70%) of 22 and 8.6 mg
+8%) of 23.
+d, 1, Jˆ1.8 Hz), 4.71 +d, 1, Jˆ1.2 Hz), 2.79 +dd, 1, Jˆ14.0,
1.8 Hz), 2.54 +dd, 1, Jˆ14.0, 1.2 Hz), 2.48±2.39 +m, 2),
2.21±2.15 +m, 1), 1.93 +dddd, 1, Jˆ13.4, 12.8, 12.8,
4.3 Hz), 1.74 +s, 3), 1.73 +s, 3), 1.72±1.63 +m, 1), 0.95 +d,
3, Jˆ6.7 Hz); 13C NMR 171.1, 166.0, 141.7, 111.5, 104.7,
55.8, 41.5, 39.3, 33.1, 30.3, 30.0, 28.7, 17.7. Anal. Calcd for
C13H18O4: C, 65.53; H, 7.61. Found: C, 65.03; H, 7.54.
Partial data for 3,3,8,11-tetramethyl-2,4-dioxaspiro[5.5]-
undec-8-ene-1,5-dione +32) were determined from the
1
mixture: H NMR 5.54 +br s, 1), 2.80-2.71 +m, 1), 2.45±
Partial data for 4,8,8-trimethyl-1-methylene-7,9-dioxa-
spiro[4.5]decane-6,10-dione: H NMR 5.18 +s, 1), 5.02 +s,
1
2.34 +m, 1), 2.26±2.15 +m, 2), 0.98 +d, 3, Jˆ6.7 Hz).
1), 1.13 +d, 3, Jˆ6.7 Hz).
3.3.6. Oxidative cyclization of 33. Oxidative cyclization of
33 +102 mg, 0.40 mmol) with Mn+OAc)3´2H2O +214 mg,
0.8 mmol) and Cu+OAc)2´H2O +160 mg, 0.8 mmol) in
EtOH +4 mL) for 13 h followed by workup as described
above and puri®cation by chromatography +12:1 hexanes/
EtOAc) gave 4 mg of pure 34cis, followed by 25 mg of a
83:17 mixture of 34cis and 35cis, 21 mg of a 2:4:1:1 mixture
of 34cis, 34trans, 35cis and 35trans, 10 mg of a 1:9 mixture
of 35trans and 34trans, and 2 mg of pure 34trans, +an over-
all yield of 29% of 34cis, 21% of 34trans, 7% of 35cis, and
3% of 35trans.
Data for 3,3,11-trimethyl-2,4-dioxaspiro[5.5]undec-8-ene-
1
1,5-dione +22): H NMR 5.87 +dt, 1, Jˆ7.3, 2.4 Hz), 5.67
+ddt, 1, Jˆ7.3, 4.9, 2.4 Hz), 2.81 +ddt, 1, Jˆ17.7, 5.5,
2.4 Hz), 2.50±2.41 +m, 2), 2.19±2.12 +m, 2), 1.75 +s, 3),
1.74 +s, 3), 0.99 +d, 3, Jˆ6.7 Hz).
Partial data for 3,3,11-trimethyl-2,4-dioxaspiro[5.5]undec-
1
7-ene-1,5-dione +23): H NMR 6.23 +dt, 1, Jˆ3.7, 9.8 Hz),
5.51 +dt, 1, Jˆ2.3, 9.8 Hz), 2.30±2.24 +m, 2).
3.3.4. Oxidative cyclization of 24. Oxidative cyclization of
24 +87 mg, 0.36 mmol) with Mn+OAc)3´2H2O +250 mg,
0.94 mmol) and Cu+OAc)2´H2O +93 mg, 0.47 mmol) in
EtOH +4.7 mL) for 2.5 h followed by workup as described
above gave 75 mg of crude product. Puri®cation by ¯ash
chromatography +15:1 hexanes/EtOAc) gave 43 mg +49%)
of 25 as a white solid followed by 20 mg +23%) of 26 as an
oil.
Data for cis-4,8,8-trimethyl-1-+1-methylethenyl)-7,9-dioxa-
spiro[4.5]decane-6,10-dione +34cis): H NMR; 4.98 +s, 1),
1
4.94 +s, 1), 3.48 +dd, 1, Jˆ11.6, 8.5 Hz), 2.91 +ddq, 1,
Jˆ11.0, 9.2, 6.7 Hz), 2.33 +dddd, 1, Jˆ12.9, 12.2, 11.6,
4.8 Hz), 2.08 +dddd, 1, Jˆ14.9, 9.2, 7.9, 4.8 Hz), 1.96
+dddd, 1, Jˆ12.9, 8.5, 7.9, 4.3 Hz), 1.81 +dddd, 1, Jˆ14.9,
12.2, 11.0, 4.3 Hz), 1.73 +s, 3), 1.70 +s, 3), 1.67 +s, 3), 1.04
+d, 3, Jˆ6.7 Hz); 13C NMR 171.2, 141.9, 115.5, 105.1, 63.9,
58.4, 48.8, 31.2, 29.9, 29.6, 28.3, 22.0, 15.3, +one carbonyl
carbon not observed).
Data for 3,3,8,8-tetramethyl-1-methylene-7,9-dioxaspiro-
[4.5]decane-6,10-dione +25): mp 59.6±60.38C; H NMR
1
5.16 +s, 1), 5.00 +s, 1), 2.49 +s, 2), 2.31 +s, 2), 1.80 +s, 3),
1.69 +s, 3), 1.20 +s, 6); 13C NMR 169.9 +2C), 153.7, 110.8,
104.7, 59.4, 49.7, 49.4, 38.7, 30.1, 28.5 +2C), 27.6. Anal.
Calcd for C13H18O4 C, 65.53; H, 7.61. Found: C, 65.17; H,
7.60.
A 1D NOESY experiment with irradiation of H1 at d 3.48
showed NOEs to H4 at d 2.91 establishing the cis stereo-
chemistry, the two H2 protons at d 2.33 and 1.96, a vinyl
proton at d 4.94 and the methyl group at d 1.70. A 1D
NOESY experiment with irradiation of H4 at d 2.91 showed
NOEs to H1 at d 3.48 and the two H3 protons at d 2.08 and
1.81.
Data for 1-+ethoxymethyl)-3,3,8,8-tetramethyl-7,9-dioxa-
spiro[4.5]decane-6,10-dione +26): H NMR 3.62 +dddd, 1
1
Jˆ12.8, 11.0, 6.1, 6.1 Hz), 3.53 +dd, 1, Jˆ9.1, 6.1 Hz),
3.40 +dq, 1, Jˆ9.2, 7.3 Hz), 3.38 +dq, 1, Jˆ9.2, 7.3 Hz),
3.36 +dd, 1, Jˆ11.0, 9.1 Hz), 2.11 +d, 1, Jˆ13.4 Hz), 2.07
+d, 1, Jˆ13.4 Hz), 1.94 +dd, 1, Jˆ12.8, 12.8 Hz), 1.73 +s, 3),
1.68 +s, 3), 1.59 +dd, 1, Jˆ12.8, 6.1 Hz), 1.25 +s, 3), 1.17 +s,
3), 1.11 +t, 3, Jˆ7.3 Hz); 13C NMR 172.3, 170.2, 104.3,
69.6, 66.5, 55.8, 52.8, 52.1, 43.8, 40.6, 30.0, 29.93, 29.88,
28.0, 14.9.
Data for trans-4,8,8-trimethyl-1-+1-methylethenyl)-7,9-
dioxaspiro[4.5]decane-6,10-dione +34trans): H NMR 4.93
+d, 1, Jˆ1.2 Hz), 4.83 +s, 1), 3.52 +dd, 1, Jˆ12.2, 6.1 Hz),
2.83 +ddq, 1, Jˆ9.8, 6.4, 7.3 Hz), 2.32±2.13 +m, 2), 1.92 +br
ddd, 1, Jˆ12.2, 6.4, 6.1 Hz), 1.74 +s, 3), 1.72 +s, 3), 1.71 +s,
3), 1.67±1.54 +m, 1), 1.13 +d, 3, Jˆ7.3 Hz).
1
A 1D NOESY experiment with irradiation of H1 at d 3.52
showed an NOE to Me4 at d 1.13, but no NOE to H4 at d
2.83 indicating that this is the trans isomer, and to the two
H2 protons at d 1.92 and 2.32±2.13, a vinyl proton at d 4.83
and the methyl group at d 1.72.
3.3.5. Oxidative cyclization of 30. Oxidative cyclization of
30 +74 mg, 0.29 mmol) with Mn+OAc)3´2H2O +161 mg,
0.6 mmol) and Cu+OAc)2´H2O +59 mg, 0.3 mmol) in EtOH
+3 mL) for 5 min followed by workup as described

32
B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
A similar cyclization of 33 +80.5 mg, 0.32 mmol) with
Mn+OAc)3´2H2O +170 mg, 0.64 mmol) and Cu+OAc)2´H2O
+3 mg, 0.02 mmol) in EtOH +3.2 mL) for 9 h followed by
workup as described above gave 44.3 mg of crude cyclized
products containing mainly 35cis and 35trans. The crude
mixture was hydrogenated over 10% Pd on carbon
+44.5 mg, 0.18 mmol) in EtOH +1.5 mL) to remove any
34. The reaction was ®ltered through Celite, concentrated
and puri®ed by chromatography +15:1 hexanes/EtOAc)
giving 5 mg of pure 35cis, followed by 29 mg of a 3:2
mixture of 35cis and 35trans, and ®nally ,1 mg of pure
35trans, +an overall yield of 28% of 35cis and 14% of
35trans).
167.1, 130.6, 125.9, 104.7, 63.8, 47.9, 39.3, 36.1, 34.3, 33.6,
29.9, 28.2; HRMS +DCI/NH3) calcd for C13H16O4´NH4
+MNH41) 254.1404, found 254.1392.
3.3.8. Dimerization of 2,2,5-trimethyl-1,3-dioxane-4,6-
dione ꢀ39). Mn+OAc)3´2H2O +340 mg, 1.27 mmol) was
added to
a
solution of Cu+OAc)2´H2O +127 mg,
0.64 mmol) in EtOH +4.3 mL). Methyl Meldrum's acid
+39) +200 mg, 1.27 mmol) in EtOH +2 mL) was added by
cannula. The reaction was stirred for 10 h followed by
workup as previously described to give 157 mg of crude
material. Puri®cation by ¯ash chromatography +15:1±5:1
hexane/EtOAc) gave 73 mg +38%) of 41, followed by
42 mg +21%) of 40, and 29 mg +10%) of 42.
Data for cis-4,8,8-trimethyl-1-+1-methylethyl)-7,9-dioxa-
spiro[4.5]decane-6,10-dione +35cis): mp 74.7±75.48C; H
1
Data for 5-+3-ethoxy-2-methyl-1,3-dioxopropoxy)-2,2,5-
trimethyl-1,3-dioxan-2,4-dione +41): H NMR 4.25 +dq, 1,
1
NMR 2.84 +ddq, 1, Jˆ11.6, 8.3, 6.7 Hz), 2.67 +ddd, 1,
Jˆ9.8, 9.8, 9.8 Hz), 2.12±1.97 +m, 2), 1.90±1.62 +m, 3),
1.76 +s, 3), 1.75 +s, 3), 1.07 +d, 3, Jˆ7.3 Hz), 0.90 +d, 3,
Jˆ6.7 Hz), 0.89 +d, 3, Jˆ6.7 Hz); 13C NMR 171.3, 165.8,
105.6, 61.7, 61.1, 48.6, 31.7, 30.6, 30.4, 29.3, 29.1, 22.1,
21.7, 16.2.
Jˆ11.0, 7.0 Hz), 4.23 +dq, 1, Jˆ11.0, 7.0 Hz), 3.59 +q, 1,
Jˆ7.3 Hz), 1.92 +s, 3), 1.87 +s, 3), 1.82 +s, 3), 1.48 +d, 3,
Jˆ6.7 Hz), 1.31 +t, 3, Jˆ7.0 Hz); 13C NMR 169.5, 168.5,
164.6, 164.3, 107.9, 73.0, 61.9, 45.1, 28.7, 28.5, 23.5, 13.9,
13.2; IR 1758, 1736, 1647, 1265; HRMS +DCI/NH3) calcd
for C13H18O8´NH4 +MNH41) 320.1346, found 320.1345.
A 1D NOESY experiment with irradiation of H1 at d 2.67
shows NOEs to H4 at d 2.84 establishing the cis stereo-
chemistry, one of the protons at d 2.12±1.97, protons at d
1.86, d 1.83-1.62, and the methyl groups at d 0.90 and d
0.89.
Data for 2,2,20,20,5,50-hexamethyl-[5,50-di-1,3-dioxane]-
4,40,6,60-tetrone +40): mp 196±1978C +lit.15a mp 197±
1
1988C); H NMR 2.08 +s, 6), 1.82 +s, 6), 1.79 +s, 6); 13C
NMR 167.8 +4 C), 107.2 +2C), 53.1 +2C), 29.2 +2C), 28.7
+2C), 22.2 +2C). Dimerization of 39 with PhI+OAc)2 gave
1
Data for trans-4,8,8-trimethyl-1-+1-methylethyl)-7,9-dioxa-
spiro[4.5]decane-6,10-dione +35trans): 1H NMR 2.73±2.65
+m, 2), 2.17 +dddd, 1, Jˆ12.6, 7.6, 7.6, 2.2 Hz), 2.04 +dddd,
1, Jˆ12.3, 7.2, 7.2, 2.1 Hz), 1.88±1.75 +m, 2), 1.74 +s, 6),
1.60±1.51 +m, 1), 1.11 +d, 3, Jˆ6.7 Hz), 0.93 +d, 3,
Jˆ6.7 Hz), 0.84 +d, 3, Jˆ6.7 Hz).
45% of 41 with an identical H NMR spectrum.15a
Data for 42: 1H NMR 4.25±4.15 +m, 2), 3.94 +dd, 1, Jˆ4.6,
3.7 Hz), 2.94 +dd, 1, Jˆ14.7, 3.7 Hz), 2.77 +dd, 1, Jˆ14.7,
4.6 Hz), 1.91 +s, 3), 1.90 +s, 3), 1.85 +s, 3), 1.82 +s, 3), 1.76
+s, 3), 1.59 +s, 3), 1.29 +t, 3, Jˆ7.0 Hz); 13C NMR 171.1,
169.9, 165.1, 164.9, 164.5 +2C), 108.2, 105.2, 73.2, 62.4,
52.9, 44.8, 29.9, 28.9, 28.7, 28.6, 25.5, 23.3, 21.3, 13.7; IR
1756, 1637, 1265; HRMS +DCI/NH3) calcd for
C20H26O12´NH4 +MNH41) 476.1792, found 476.1768.
3.3.7. Oxidative cyclization of 36. Oxidative cyclization of
36 +126 mg, 0.53 mmol in 20 mL of EtOH) by slow addition
by syringe pump to Mn+OAc)3´2H2O +285 mg, 1.1 mmol)
and Cu+OAc)2´H2O +106 mg, 0.53 mmol) in EtOH +5.3 mL)
over 12 h and stirring of the resulting solution for 9 h
followed by workup as described above and puri®cation
by chromatography +16:1 hexanes/EtOAc) gave 6.9 mg of
pure 37, followed by a 35:65 mixture of 37 and 38
+50.4 mg), and 19.8 mg of pure 38 +an overall yield of
20% of 37 and 42% of 38).
3.3.9. Oxidative cyclization of 43. Oxidative cyclization of
43 +113 mg, 0.45 mmol) with Mn+OAc)3´2H2O +241 mg,
0.90 mmol) and Cu+OAc)2´H2O +180 mg, 0.90 mmol) in
EtOH +4.5 mL) for 2.5 h followed by workup as described
above gave 111 mg of crude cyclized product. Puri®cation
by ¯ash chromatography +15:1 hexanes/EtOAc) gave 6 mg
+5%) of pure 44 followed by 80 mg +71%) of 45.
Data for 20,20-dimethylspiro[bicyclo[2.2.2]oct-5-ene-2,50-
[1,3]dioxane]-40,60-dione +37): mp 78.5±79.08C +lit.10
Data for 8,8-dimethyl-1-methylene-3-+2-propenyl)-7,9-
dioxaspiro[4.5]decane-6,10-dione +44): 1H NMR 5.81
+ddt, 1, Jˆ17.7, 9.8, 6.7 Hz), 5.17 +s, 1), 5.05 +dd, 1,
Jˆ17.7, 1.9 Hz), 5.03 +s, 1), 5.02 +br d, 1, Jˆ9.8 Hz), 2.72
+dd, 1, Jˆ15.2, 7.3 Hz), 2.62 +ttt, 1, Jˆ7.3, 7.3, 6.7 Hz),
2.45 +dd, 1, Jˆ12.8, 7.3 Hz), 2.33 +ddt, 1, Jˆ15.2, 11.0,
3.1 Hz), 2.27±2.16 +m, 3), 1.81 +s, 3), 1.70 +s, 3); 13C
NMR 170.3, 169.7, 153.6, 136.5, 116.2, 110.3, 104.9,
58.2, 42.4, 40.3, 38.9, 38.7, 30.2, 27.8.
1
78.7±79.18C); H NMR 6.48 +t, 1, Jˆ7.3 Hz), 6.12 +t, 1,
Jˆ7.3 Hz), 3.02±2.98 +m, 1), 2.83±2.78 +m, 1), 2.22 +dd,
1, Jˆ12.8, 2.4 Hz), 2.01 +dt, 1, Jˆ12.8, 2.9 Hz), 1.85±1.74
+m, 2), 1.82 +s, 3), 1.68 +s, 3), 1.33±1.17 +m, 2); 13C NMR
168.90, 168.88, 136.3, 128.8, 105.1, 54.4, 39.7, 32.5, 30.4,
29.2, 28.0, 22.3, 22.0. The 1H NMR spectrum is identical to
that previously reported.10
Data for 20,20-dimethylspiro[bicyclo[3.2.1]oct-2-ene-7,50-
1
[1,3]dioxane]-40,60-dione +38): H NMR 5.81±5.77 +m, 1),
Data for 3,3-dimethyl-8-+2-propenyl)-2,4-dioxaspiro[5.5]-
undec-8-ene-1,5-dione +45): 5.80±5.70 +m, 3), 5.10±5.02
+m, 2), 2.84 +br d, 1, Jˆ17.5 Hz), 2.50±2.36 +m, 2), 2.22±
2.05 +m, 3), 1.84 +dd, 1, Jˆ13.4, 11.6 Hz), 1.77 +s, 3), 1.73
+s, 3).
5.64 +t, 1, Jˆ7.3 Hz), 2.95 +t, 1, Jˆ5.5 Hz), 2.68 +dd, 1,
Jˆ12.8, 7.3 Hz), 2.58±2.42 +m, 2), 2.36 +dd, 1, Jˆ13.8,
1.5 Hz), 2.28±2.23 +m, 1), 2.16 +br d, 1, Jˆ18.3 Hz),
1.88±1.67 +m, 1), 1.84 +s, 3), 1.72 +s, 3); 13C NMR 168.8,

B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
33
3.3.10. Oxidative cyclization of 47. Oxidative cyclization
of 47 +75 mg, 0.27 mmol) with Mn+OAc)3´2H2O +145 mg,
0.54 mmol) and Cu+OAc)2´H2O +5 mg, 0.03 mmol) in EtOH
+2.7 mL) for 4 h followed by workup as described
previously gave 63 mg of crude product. Puri®cation by
¯ash chromatography +13:1 hexanes/EtOAc) gave 50 mg
+67%) of pure cis-octahydro-2,2,7a0-trimethyl-30-methyl-
ene-spiro[1,3-dioxane-5,50-[5H]indene]-4,6-dione +49): mp
105.0±105.78C; 1H NMR 4.87 +br s, 1), 4.85 +br s, 1), 2.50±
2.41 +m, 3), 2.20 +ddd, 1, Jˆ14.0, 12.8, 4.9 Hz), 1.99±1.81
+m, 4), 1.97 +ddd, 1, Jˆ11.9, 11.6, 9.5 Hz), 1.73 +s, 6), 1.66
+ddd, 1, Jˆ14.0, 4.9, 4.3 Hz), 1.32 +ddd, 1, Jˆ11.9, 7.9,
3.1 Hz), 0.99 +s, 3); 13C NMR 171.6, 168.4, 155.5, 107.2,
104.6, 48.6, 47.3, 39.3, 35.6, 33.3, 30.2, 29.2, 29.0, 28.9,
28.5, 28.0. Anal. Calcd for C16H22O4: C, 69.04; H, 7.97.
Found: C, 68.28; H, 7.97.
171.7, 167.3, 130.4, 129.2, 104.7, 54.0, 46.7, 34.1, 29.5,
29.1, 27.2, 24.4, 20.2, 17.8.
Acknowledgements
We are grateful to the NIH for generous ®nancial support
+GM-50151).
References
1. For reviews see: +a) Snider, B. B. Chem. Rev. 1996, 96, 339±
363. +b) Melikyan, G. G. Org. React. 1997, 49, 427±675.
+c) Melikyan, G. G. Aldrichim. Acta 1998, 31, 50±64.
+d) Snider, B. B. In Manganese(III)-Mediated Radical
Reactions; Renaud, P., Sibi, M., Eds.; Radicals in Organic
Synthesis; Wiley-VCH: Weinheim, 2001; Vol. 3, pp 198±
218 Chapter 2.3.
A 1D NOESY experiment with irradiation of the vCH2
group at d 4.85±4.90 showed NOEs to H3 and H3a at d
2.5±2.4, H4 at d 2.0±1.8 and the methyl group at d 0.99.
A 1D NOESY experiment with irradiation of the methyl
group at d 0.99 showed NOEs to H3a at d 2.5±2.4, and
hydrogens at d 2.0±1.8, 1.66 and 1.32. The intense NOE
between the methyl group and H3a establishes that the ring
fusion is cis.
2. +a) Curran, D. P.; Morgan, T. M.; Schwartz, C. E.; Snider,
B. B.; Dombroski, M. A. J. Am. Chem. Soc. 1991, 113,
6607±6617. +b) Snider, B. B.; Merritt, J. E.; Dombroski,
M. A.; Buckman, B. O. J. Org. Chem. 1991, 56, 5544±
5553. +c) Snider, B. B.; McCarthy, B. A. J. Org. Chem.
1993, 58, 6217±6223.
3. For related oxidations with Fe+III) and Ce+IV) see:
+a) Citterio, A.; Cerati, A.; Sebastiano, R.; Finzi, C.; Santi,
R. Tetrahedron Lett. 1989, 30, 1289±1292. +b) Baciocchi, E.;
Belli Paolobelli, A.; Ruzziconi, R. Tetrahedron 1992, 48,
4617±4622.
3.3.11. Oxidative cyclization of 50. Oxidative cyclization
of 50 +94 mg, 0.42 mmol) with Mn+OAc)3´2H2O +223 mg,
0.84 mmol) and Cu+OAc)2´H2O +168 mg, 0.84 mmol) in
EtOH +4.2 mL) for 4 h followed by workup as described
previously gave 78 mg of crude product containing
some dimer. Puri®cation by ¯ash chromatography +16:1
hexanes/EtOAc) gave 22 mg of pure 51, followed by
5 mg of a 83:16 mixture of 51 and 52, and 17 mg of
80±90% pure 52 +an overall yield of 37% of 51 and 19%
of 52).
4. For mechanistic studies of related substrates see: +a) Snider,
B. B.; Patricia, J. J.; Kates, S. A. J. Org. Chem. 1988, 53,
2137±2143. +b) Baciocchi, E.; Giese, B.; Farshchi, H.;
Ruzziconi, R. J. Org. Chem. 1990, 55, 5688±5691.
5. +a) Pearson, R. G.; Dillon, R. L. J. Am. Chem. Soc. 1953, 75,
2439±2443. +b) Arnett, E. M.; Harrelson, Jr., J. A. J. Am.
Chem. Soc. 1987, 109, 809±812. +c) Arnett, E. M.; Harrelson
Jr, J. A. Gazz. Chim. Ital. 1987, 117, 237±243. +d) Bordwell,
F. G. Acc. Chem. Res. 1988, 21, 456±463.
Data for 3,3-dimethyl-7-methylene-2,4-dioxaspiro[5.5]un-
decane-1,5-dione +51): H NMR 5.16 +s, 1), 4.85 +s, 1),
1
2.46 +t, 2, Jˆ5.8 Hz), 2.23 +t, 2, Jˆ6.5 Hz), 2.00 +tt, 2,
Jˆ6.5, 6.2 Hz), 1.78 +s, 3), 1.75±1.68 +m, 2), 1.73 +s, 3),
13C NMR 167.8 +2C), 144.1, 114.3, 105.1, 56.4, 35.3, 32.3,
29.2, 28.1, 26.8, 21.2.
6. +a) Chen, B. C.; Lue, P. Org. Prep. Proc. Int. 1992, 24, 185±
188. +b) Kore, A. R.; Mane, R. B.; Salunkhe, M. M. Bull. Soc.
Chim. Belg. 1995, 104, 643±645. +c) Shing, T. K. M.; Li,
L.-H.; Narkunan, K. J. Org. Chem. 1997, 62, 1617±1622.
7. For reviews of Meldrum's acid chemistry see: +a) McNab, H.
Chem. Soc. Rev. 1978, 7, 345±358. +b) Chen, B.-C. Hetero-
cycles 1991, 32, 529±597.
Data for 3,3-dimethyl-2,4-dioxaspiro[5.6]dodec-8-ene +52):
1H NMR 5.87 +dt, 1, Jˆ11.0, 5.5 Hz), 5.64±5.58 +m, 1),
2.76 +d, 2, Jˆ8.0 Hz), 2.35 +dt, 2, Jˆ5.5, 6.1 Hz), 2.25
+m, 2), 1.94±1.87 +m, 2), 1.74 +s, 3), 1.73 +s, 3); 13C NMR
170.6 +2C), 132.8, 123.6, 104.7, 51.2, 35.7, 33.1, 29.1, 28.9,
28.7, 19.8.
8. +a) Hrubowchak, D. M.; Smith, F. X. Tetrahedron Lett. 1983,
24, 4951±4954. +b) Trost, B. M.; Gerusz, V. J. J. Am. Chem.
Soc. 1995, 117, 5156±5157. +c) Pattenden, G.; Teague, S. J.
Tetrahedron 1997, 43, 5637±5652.
9. This route cannot be used with 2,6-dimethyl-5-heptenal and
3,7-dimethyl-6-octenal because the initially formed Knoeve-
nagel products undergo intramolecular Diels±Alder reaction:
+a) Takano, S.; Satoh, S.; Ogasawara, K. Heterocycles 1985,
23, 41±44. +b) Tietze, L. F.; Kiedrowski, G. V.; Fahlbusch,
K.-G.; Voss, E. Org. Synth. 1990, 69, 31±37.
3.3.12. Oxidative cyclization of 53. Oxidative cyclization
of 53 +36 mg, 0.14 mmol) with Mn+OAc)3´2H2O +76 mg,
0.28 mmol) and Cu+OAc)2´H2O +1.2 mL of 0.12 M solution
in EtOH, 0.14 mmol) for 8 h followed by workup as
described previously gave 25 mg of crude product. Puri®ca-
tion by ¯ash chromatography +9:1 hexanes/EtOAc) gave
12 mg +33%) of 90% pure 3,3-dimethyl-1E-propenyl-2,4-
dioxaspiro[5.5]undecane-1,5-dione +54): 1H NMR 5.57
+dq, 1, Jˆ15.0, 6.7 Hz), 5.23 +ddd, 1, Jˆ15.0, 9.2,
1.5 Hz), 2.81 +ddd, 1, Jˆ12.1, 9.2, 4.3 Hz), 2.09±1.61 +m,
7), 1.70 +s, 3), 1.66 +s, 3), 1.62 +d, 3, Jˆ6.1 Hz), 1.39
+ddddd, 1, Jˆ12.1, 12.1, 12.1, 3.8, 3.8 Hz); 13C NMR
10. Products 16 and 37 have previously been prepared by Diels±
Alder reaction of methylene Meldrum's acid with butadiene
and cyclohexadiene, respectively: +a) Zia-Ebrahimi, M.;
Huffman, G. W. Synthesis 1996, 2, 215±218. +b) Buzinkai,
J. F.; Hrubowchak, D. M.; Smith, F. X. Tetrahedron Lett.
1985, 26, 3195±3198. +c) Brown, R. F. C.; Eastwood, F. W.;
McMullen, G. L. Aust. J. Chem. 1977, 30, 179±193.

34
B. B. Snider, R. B. Smith / Tetrahedron 58 (2002) 25±34
11. Bausch, M. J.; Guadalupe-Fasano, C.; Gostowski, R.;
Selmarten, D.; Vaughn, A. J. Org. Chem. 1991, 56, 5640±
 Â
5642. For related studies see: Nedelec, J.-Y.; Lachaise, I.;
15. For related dimerizations see Ref. 4a and +a) Yan, J.; Zhong,
L.-R.; Chem, Z.-C. J. Org. Chem. 1991, 56, 459±461.
+b) Citterio, A.; Santi, R.; Fiorani, T.; Strologo, S. J. Org.
Chem. 1989, 54, 2703±2712. +c) Peek, R.; Streukens, M.;
Thomas, H. G.; Vanderfuhr, A.; Wellen, U. Chem. Ber.
1994, 127, 1257±1262.
Nohair, K.; Paugam, J. P.; Hakiki, M. Bull. Soc. Chim. Fr.
1995, 132, 843±849. +c) Weber, M.; Fischer, H. Helv. Chim.
Acta 1998, 81, 770±780.
12. For equilibration of analogous cyanoacetates see: Curran,
D. P.; Chang, C.-T. J. Org. Chem. 1989, 54, 3140±3157.
13. We observed small amounts of ligand transfer products in
other reactions, but were not able to characterize them since
the cyclopentanemethyl radical is usually a minor product.
16. Egger, H. Monatsh. Chem. 1967, 98, 1245±1255.
17. Bambal, R.; Kemmitt, R. D. W. J. Chem. Soc., Chem.
Commun. 1988, 734±735.
18. Yamamoto, Y.; Iwasa, M.; Sawada, S.; Oda, J.-i. Agric. Biol.
Chem. 1990, 54, 3269±3274.
È
19. Hiyama, T.; Tsukanaka, M.; Nozaki, H. J. Am. Chem. Soc.
1974, 96, 3713±3714.
20. Vogel, D. E.; Bu
Èchi, G. H. Org. Synth. 1988, 66, 29±36.
14. +a) Nilsson, Y. I. M.; Andersson, P. G.; Backvall, J.-E. J. Am.
Chem. Soc. 1993, 115, 6609±6613. +b) Bricout, H.;
Carpentier, J.-F.; Mortreux, A. Tetrahedron Lett. 1997, 38,
1053±1056.