Regioselectivity in the palladium-catalyzed addition of carbon nucleophiles to dihydropyran derivatives

Article abstract of DOI:10.1021/jo961973d

The regioselectivity of Pd-catalyzed malonate or sulfonylacetate addition to dihydropyran derivatives is highly dependent upon the substitution pattern of the dihydropyran nucleus and is governed by electronic, rather than steric factors. In certain instances, subtle steric features also play a role in controlling regioselectivity by altering the conformation of the intermediate η³-allyl Pd-complex.

Full text of DOI:10.1021/jo961973d

J . Org. Chem. 1997, 62, 1257-1263
1257
Regioselectivity in th e P a lla d iu m -Ca ta lyzed Ad d ition of Ca r bon
Nu cleop h iles to Dih yd r op yr a n Der iva tives
Marc-Raleigh Brescia, Yvonne Class Shimshock, and Philip DeShong*
Department of Chemistry and Biochemistry, The University of Maryland College Park, Maryland 20742
Received October 22, 1996
The regioselectivity of Pd-catalyzed malonate or sulfonylacetate addition to dihydropyran derivatives
is highly dependent upon the substitution pattern of the dihydropyran nucleus and is governed by
electronic, rather than steric factors. In certain instances, subtle steric features also play a role in
controlling regioselectivity by altering the conformation of the intermediate η3-allyl Pd-complex.
Sch em e 1
Palladium-catalyzed transformations play a major role
in the arsenal of synthetic chemistry.1,2 Of particular
interest is the Pd(0)-catalyzed alkylation of allylic alcohol
derivatives by stabilized carbon nucleophiles. This reac-
tion is widely applied in organic synthesis because the
alkylation generally proceeds with retention of configu-
ration at the least hindered site of the allylic starting
material.3-5 During the course of our synthesis of
pseudomonic acid A6 (1), our strategy was to introduce
the C-8 side chain7 of dihydropyran 2 via Pd-catalyzed
alkylation of allylic benzoate 3 as indicated in Scheme
1. In the alkylation, it was anticipated that the nucleo-
phile would be introduced at C-8 (C-3)7 regioselectively
since the alkyl side chain had a larger A-value than the
corresponding methoxy substituent.8 However, alkyla-
tion did not occur at C-3 of dihydropyran 3, the least
hindered site to produce pyran 5; rather it occurred
regiospecifically at C-5 yielding diester 4 (Scheme 2).
Previous studies of Pd-catalyzed alkylation in carbo-
hydrate derivatives by Baer,9 Dunkerton,10 and Curran11
in addition to other heteroatom-substituted systems1b
indicated that both steric and electronic factors had a role
in controlling the regioselectivity of alkylation of dihy-
dropyran derivatives. For example, Baer and Hanna
reported that the sole alkylation adduct of the reaction
of pyranoside 6 and diethylmalonate was diester 7.9
Baer’s results, taken in conjunction with our findings in
the pseudomonic acid series, suggested that electronic
factors played a more prominent role in influencing the
regioselectivity of the Pd-catalyzed alkylations than had
been previously recognized. However, the limited num-
Sch em e 2
X Abstract published in Advance ACS Abstracts, February, 1; 1997.
(1) (a) Godleski, S. A. In Comprehensive Organic Synthesis; Trost,
B. M., Ed.; Permagon Press: New York, 1991; Vol. 4, pp 585-661. (b)
For a discussion of the regioselectivity of Pd-catalyzed alkylation in
allylic systems with a heteroatom directly attached to the allyl system,
see: Godleski, S. A. In Comprehensive Organic Synthesis; Trost, B.
M., Ed.; Permagon Press: New York, 1991; Vol. 4, pp 615-635, and
references cited therein.
(2) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457-2483.
(3) Trost, B. M. Tetrahedron 1977, 33, 2615-2649.
(4) Trost, B. M. Pure Appl. Chem. 1979, 51, 787-800.
(5) Trost, B. M. Acc. Chem. Res. 1980, 13, 385-393.
(6) Class, Y. J .; DeShong, P. Tetrahedron Lett. 1995, 36, 7631-7634.
(7) Pseudomonic acid numbering. We shall use pyran numbering
throughout the remainder of the manuscript.
(8) Carey, F. A.; Sundberg, R. J . Advanced Organic Chemistry;
Plenum Press: New York, 1990; pp 130-139.
(9) Baer, H. H.; Hanna, Z. S. Can. J . Chem. 1981, 59, 889-906.
(10) Dunkerton, L. V.; Serino, A. J . J . Org. Chem. 1982, 47, 2812-
2814.
(11) It was recognized that the relative stereochemistry of the two
substituents on dihydropyran 2 would also play a role in controlling
the regiochemistry of nucleophilic attack; however, we anticipated that
the general features of the addition would not be altered by the relative
stereochemistry of the substituents. See also: Curran, D. P.; Suh, Y.-
G. Carbohydrate Res. 1987, 171, 161-191.
ber of dihydropyran derivatives studied under these
conditions precluded a comprehensive understanding of
the relative importance of steric and electronic factors
in the regioselectivity of alkylation.6,9-11 Accordingly, a
systematic study of the Pd-catalyzed reaction of dihy-
dropyran derivatives was undertaken to assess the
importance of substituents and their relative stereo-
chemistry in the alkylation process. As described below,
this study demonstrated that electronic characteristics
of substituents is the dominant determinant in control-
S0022-3263(96)01973-1 CCC: $14.00 © 1997 American Chemical Society

1258 J . Org. Chem., Vol. 62, No. 5, 1997
Brescia et al.
Sch em e 4
F igu r e 1.
Sch em e 3
chosen to serve as models for the allylic benzoates
employed in the synthesis of pseudomonic acid deriva-
tives (3 in Scheme 1). Methyl was chosen as the C-6
substituent because its A-value is known precisely.8
Since the A-value of the methyl group is more than twice
the size of the methoxy group it was anticipated that
alkylation should occur regioselectively at C-3 in this
series of compounds, assuming that steric factors are
more important than electronic factors in controlling the
regiochemistry of alkylation. If, on the other hand,
electronic factors are more important than steric interac-
tions, one anticipates that C-5 alkylation products would
be favored. The experimental results shown in Scheme
4 demonstrate conclusively that C-5 alkylation, the
products of electronic control of the regioselectivity,
predominated, in analogy with our results in the
pseudomonic acid series and Baer and Hanna’s findings9
(Scheme 2). It is particularly noteworthy that the same
regioselectivity was observed in the alkylation irrespec-
tive of the relative configuration of the two substituents.
The slight differences in the regioselectivity of the
alkylation of benzoates 13 and 16 of 5:1 vs 9:1, respec-
tively, can be attributed to conformational preferences
adopted by the intermediate Pd-complex and will be
discussed below.
In the final model system investigated, allylic ben-
zoates 19 and 21 lacking the anomeric methoxy group
at C-2 were subjected to alkylation (Scheme 5). On the
basis of the hypothesis, it was anticipated that lack of
the C-2 methoxy group would remove the electronic bias
of the system and favor C-3 alkylation due to steric
control of the alkylation. In fact, this was the case with
the cis-benzoate 19, which gave diester 20 as the exclu-
sive product of alkylation. However, alkylation of trans-
benzoate 21 gave a 1:7 ratio of diesters 22 and 23 in
which the C-5 regioisomer predominated. Our prediction,
as well as the results from other systems,9-11 was that
C-3 alkylation should be the major product of alkylation,
even in trans-benzoate 21. For example, Curran ob-
served regioselective C-3 alkylation in the Pd-catalyzed
allylic alkylation of trans-acetate 24 under similar condi-
tions (Scheme 5).11 However, in Curran’s example, it is
unclear whether coordination of the Lewis basic ester
moiety could have altered the “inherent” regioselectivity
of attack by the nucleophile on the π-allyl Pd intermedi-
ate. Benzoate 21, a compound that has the identical
relative configuration of substituents as Curran’s acetate
24 and must proceed through a common Pd-complex (at
ling the regioselectivity of alkylation in unsymmetrically
substituted allylic benzoates such as dihydropyrans 3,
13, and 16, respectively. In the absence of the “ano-
meric”-like substituent at C-2, however, substituents on
the ring also affect the conformation of the π-allyl Pd-
complexes and, ultimately, the regioselectivity of alky-
lations.
Our hypothesis is that alkylation at C-5 in dihydro-
pyrans 3 and 6 is strongly favored due to decreased
electrophilicity at C-3. The reasoning for this prediction
is that the π-allyl Pd-complex has resonance contributors
with cationic character at C-3 and C-5, respectively (A-
C, Figure 1). The presence of two inductively electron-
withdrawing oxygens at C-2 decreases the contribution
of resonance form C to the hybrid and results in increased
cationic character at C-5, resonance form B.
The first set of dihydropyran derivatives studied were
diastereomeric benzoates 8 and 11, respectively. Upon
the basis of both steric and electronic considerations, it
was anticipated that alkylation would occur at C-5, and
this expectation was borne out as summarized in Scheme
3. Irrespective of the relative configuration of the allylic
benzoate, the major product arose from alkylation at C-5.
Under a standard set of alkylation conditions,12 trans-
allylic benzoate 8 gave a 9:1 ratio of diesters 9 and 10,
while cis-allylic benzoate 11 gave exclusively diester 12.
The regioselectivity and stereospecificity of each alkyla-
tion product was confirmed by homonuclear decoupling
experiments. The selectivity is greater with cis-allylic
benzoate 11 because the incoming diethyl malonate anion
must approach the π-allyl Pd intermediate from the same
face as the methoxy group at C-2 (vide infra).
In a second series of allylic derivatives, benzoates 13
and 16 were subjected to the alkylation protocol (Scheme
4). These benzoates have a methyl group at C-6 in
addition to an anomeric methoxy group at C-2 and were
(12) A standard protocol for the alkylation was employed for these
studies in an effort to minimize concerns about the regioselectivity
arising from differences in the reaction conditions or catalysts involved.
In some instances, the yields of adduct suffered from lack of optimiza-
tion of reaction conditions.

Addition of Carbon Nucleophiles to Dihydropyran Derivatives
J . Org. Chem., Vol. 62, No. 5, 1997 1259
Sch em e 5
avoiding eclipsing interactions with the pseudoaxial
methyl group (Scheme 6).
The situation was altered dramatically for alkylation
of benzoate 21 (Scheme 7). In this case, displacement of
the benzoate by Pd(0) provided complex 31 with the
methyl group and the Pd having a syn-relationship. (For
the sake of consistency, the enantiomer of 21 is drawn
in this scheme). The methyl group sterically interacts
with Pd and its ligands in both conformations 31A and
31B, respectively, resulting in a Pd-complex where Pd
lies toward C-3 as in complex 32A. This “distortion” of
the π-allyl Pd-complex can be represented as a σ,η2-
complex 32B.14 Addition of malonate to the “distorted”
Pd-complex 32 occurred in an SN2′-like fashion regio-
selectivity at C-5 to afford dihydropyran 23.
Additional support for the proposal that “distorted”
π-allyl Pd-complexes are responsible for the high regio-
selectivity was provided by the experiments shown in
Scheme 8. Treatment of cis-benzoate 19 under Stille
conditions14d with Pd(0)-catalyst and PhSnMe3 gave a 3:2
mixture of dihydropyrans 33 and 34. Arylation of the
π-allyl Pd-complex is known to occur by transmetalation
of the aryl group from tin to Pd followed by reductive
elimination. Benzoate 19 gave symmetrical π-allyl
Pd-complex 30; therefore, arylation was expected to occur
nonselectively at either end (C-3 or C-5) of the allyl
Pd-complex.
Arylation of unsymmetrical Pd-complex 31, on the
other hand, must occur with high regioselectivity at the
end of the allylic system distal to the sterically demand-
ing methyl substituent. In this case, the arylation results
supported the proposal of a “distorted” intermediate
Pd-complex, and a 15:1 mixture of dihydropyrans 35 and
36, respectively, was obtained (a 4 to 2 ratio of 33 and
34 were also present).
The results summarized above indicated that both
electronic and conformational factors exert significant
control of the regioselectivity of functionalized pyran
derivatives. In particular, substituents on the same face
of the pyran as the metal and its ligands result in
“distortion” of the π-allyl Pd intermediate and lead to
high regioselectivity in subsequent coupling reactions.
We are currently assessing whether these factors also
affect the regioselectivity of cyclohexane and cyclopentane
derivatives. Application of this strategy to the synthesis
of pyran-based natural products is underway, and results
will be reported in due course.
least with regard to relative stereochemistry of Pd and
C-6 substituents), provided, however, the regioisomeric
product (Scheme 5). As shown previously, it has been
demonstrated that Pd-catalyzed alkylation of cis-ben-
zoate 27 with either sulfonyl anion or malonate gave
dihydropyrans 28 and 29, respectively, with virtually
complete regioselectivity.6
Exp er im en ta l Section
All reagents were distilled, recrystallized, or chromato-
graphed prior to use unless otherwise noted. Tetrahydrofuran
(THF) and diethyl ether (Et2O) were distilled from sodium/
benzophenone ketyl while methylene chloride (CH2Cl2) and
dimethylformamide (DMF) were distilled from calcium hy-
dride. Glassware used in the reactions described below was
dried overnight in an oven at 120 °C. All reactions were run
under an atmosphere of N2.
We attribute the significant difference in regioselec-
tivity of alkylation between diastereomeric benzoates 19
and 21 to distortion in the respective π-allyl Pd inter-
mediates as outlined in Schemes 6 and 7. Treatment of
cis-benzoate 19 with Pd(0) catalyst resulted in formation
of π-allyl Pd intermediate 30. This complex could exist
in two conformations, 30A and 30B, where 30A predomi-
nates. A survey of X-ray crystallographic structures of
π-allyl Pd-complexes in the Cambridge Crystallographic
Database indicated that the substituents syn to Pd on
the ring favor the pseudoequatorial conformation, pre-
sumably to avoid steric interactions with the bulky
ligands on Pd.14 Attack of malonate anion onto Pd-
complex 30A should occur regioselectivity at C-3 thus
(13) The survey consisted of all ten structures of Pd-complexes in
which a π-allyl Pd-complex of a cyclic structure had been determined.
For example, see BEXHIV10 or FIDVIX in the Cambridge Crystal-
lographic Database.
(14) The structure of π-allyl Pd-complexes in these reactions remains
controversial. The role of η3 vs σ,η2-complexes in alkylation reactions
has been discussed by several authors. See: (a) Trost, B. M.; Verho-
even, T. R. J . Org. Chem. 1976, 41, 3215-3216. (b) Fiaud, J . C.;
Malleron, J . L. Tetrahedron Lett. 1981, 22, 1399-1402. (c) Akermark,
B.; Hansson, S.; Krakenberger, B.; Vitagliano, A.; Zetterberg, K.
Organometallics 1984, 3, 679-682. (d) Del Valle, L.; Stille, J . K.;
Hegedus, L. S. J . Org. Chem 1990, 55, 3019-3023.

1260 J . Org. Chem., Vol. 62, No. 5, 1997
Brescia et al.
Sch em e 6
Sch em e 7
Gas chromatography was performed using a gas chromato-
graph equipped with a flame ionization detector and a 25-m
capillary column coated with crosslinked methyl silicone. All
compounds for which elemental analysis was not obtained
were >95% pure as judged by gas chromatographic and NMR
spectral analysis unless otherwise noted. 1H-NMR spectra of
these substances are included in the Supporting Information.
Gen er a l P r oced u r e for th e Syn th esis of Diester s 4, 9,
10, 12, 14, 15, 17, 18, 20, 22, 23, a n d 29. Sodium hydride
(60% dispersion in oil, 0.15 g, 0.38 mmol) was washed with 2

Addition of Carbon Nucleophiles to Dihydropyran Derivatives
Sch em e 8
J . Org. Chem., Vol. 62, No. 5, 1997 1261
× 2 mL of hexane and once with 2 mL of THF. To a
suspension of oil-free NaH in 5 mL of THF was added diethyl
malonate (770 µL, 0.38 mmol). After 15 min of stirring, the
resulting diethyl malonate anion solution was added via
cannula to a solution of allylic benzoate (0.56 mmol), Pd(PPh3)4
(45 mg, 0.039 mmol), and triphenylphosphine (100 mg, 0.39
mmol) in 5 mL of THF. The resulting orange mixture was
heated at 66 °C for 48 h. The reaction mixture was then
separated between 10 mL of distilled H2O and 10 mL of Et2O.
The aqueous phase was extracted with 3 × 10 mL of Et2O.
The combined organic layers were dried over MgSO4 and
concentrated in vacuo. Purification of the residue by flash
chromatography (20 mm, 12 cm, 5-15% Et2O/hexane) gave
the diesters as a colorless oil. The ratio of diesters was
determined by gas chromatography and confirmed by 1H-NMR.
Both the major and minor diesters were subjected to homo-
nuclear decoupling to determine their regiochemistry. Only
the major diester in each reaction was fully characterized (vide
infra), since the minor diester could not be successfully
separated from the major diester.
Gen er a l P r oced u r e14d for th e Syn th esis of Dih yd r o-
p yr a n s 33-36. Phenyltrimethyltin (0.22 mL, 0.90 mmol) was
added to a yellow mixture of allylic benzoate (130 mg, 0.60
mmol), Pd(dba)2 (17 mg, 0.030 mmol), and LiCl (76 mg, 1.8
mmol) in 10 mL of DMF. The mixture was degassed and
stirred at room temperature for 72 h. The reaction mixture
was then separated between 75 mL of distilled H2O and 75
mL of Et2O. The organic layer washed with an additional 75
mL of distilled H2O, dried over MgSO4 and concentrated in
vacuo. Purification of the residue by flash chromatography
(20 mm, 16 cm, 0-10% Et2O/hexane) gave 63 mg (61%) of a
yellow oil. The ratio and regiochemistry of the products was
determined in the same manner as the diesters (vide supra).
IR (CCl4) 2956 (s), 2931 (s), 2856 (s), 1756 (s), 1737 (s), 1462
(m). 1H-NMR (CDCl3) 0.00-0.10 (m, 6), 0.75-0.88 (m, 9),
1.15-1.30 (m, 9), 1.50-2.03 (m, 3), 2.65-2.85 (m, 1), 3.35-
3.50 (m, 3), 3.78-4.31 (m, 6), 4.64-4.96 (m, 1), 5.64-5.79 (m,
1), 5.90-6.02 (m, 1).
Alcohols 40 and 46 have been prepared by Achmatowicz,15
however, we prepared 40 and 46 using methodology developed
in the DeShong lab (Schemes 9 and 10, respectively).6 Under
standard Mitsunobu conditions,16 40 was converted to benzoate
8, while 46 gave benzoate 16. Alternatively, 40 was treated
with n-BuLi and BzCl to afford benzoate 11, while 46 was
converted to benzoate 13. Finally, benzoates 13 and 16 were
reduced under Gray conditions17 to afford benzoates 19 and
21, respectively (Scheme 11).
Ben zoa te 8.18,19 Benzoate 8 was obtained in 67% yield. Mp
49.5-50.5 °C (hexane:Et2O). TLC Rf ) 0.54 (3:2 hexane:
EtOAc). IR (CCl4) 3065 (w), 2995 (m), 2931 (m), 2896 (m), 2826
(m), 1722 (s), 1455 (s), 1272 (s). 1H-NMR (CDCl3) 3.45 (s, 3),
3.96 (dd, 1, J ) 1.4, 13.0), 4.23 (dd, 1, J ) 2.8, 13.0), 4.92 (d,
1, J ) 3.1), 5.17 (ddd, 1, J ) 1.4, 2.8, 5.1), 6.06 (dd, 1, J ) 3.1,
10.3), 6.18 (dd, 1, J ) 5.1, 10.3), 7.50-8.08 (m, 5). 13C-NMR
(CDCl3) 55.6, 61.3, 63.8, 94.1, 125.1, 128.3, 129.7, 130.0, 130.8,
133.0, 166.1. HRMS (EI) Calcd for C13H14O4: 234.0892 (M+).
Found: 234.0897.
Diester 9. This compound was obtained as a 9 to 1 ratio
of diesters 9 and 10. TLC Rf ) 0.35 (3:1 hexane:EtOAc). IR
(15) Achmatowicz J r., O.; Buckowski, P.; Szechner, B.; Zierchowska,
A.; Zamojski, A. Tetrahedron 1971, 27, 1973-1996.
(16) Mitsunobu, O. Synthesis 1981, 1-28.
(17) Rolf, D.; Gray, G. R. J . Am. Chem. Soc. 1982, 104, 3539-3541.
(18) Bock, K.; Pederson, C. Acta. Chem. Scand. 1971, 25, 2757-
2764.
(19) Achmatowicz, O., J r.; Bukowski, P. Rocz. Chem. 1973, 47, 99-
113.
(20) Dyong, I.; Weigand, J .; Thiem, J . Liebigs. Ann. Chem. 1986, 4,
577-786.
Diester 4. Diester 4 was obtained in 67% yield as a mixture
of three diastereomers. TLC Rf ) 0.45 (6:1 hexane:EtOAc).

1262 J . Org. Chem., Vol. 62, No. 5, 1997
Brescia et al.
Sch em e 9
Sch em e 10
(CCl4) 3056 (w), 2981 (m), 2931(m), 2888 (m), 1756 (s) 1738
(s), 1450 (m), 1394 (m), 1369 (m). 1H-NMR (CDCl3) 1.24 (t, 6,
J ) 7.1), 2.68 (ddd, 1, J ) 3.4, 5.5, 10.0), 3.38 (s, 3), 3.53 (d, 1,
J ) 10.0), 3.63 (d, 1, J ) 12.0) 4.00, (dd, 2, J ) 3.4, 12.0), 4.15
(q, 4, J ) 7.1), 4.71 (d, 1, J ) 2.7), 5.76 (dd, 1, J ) 2.7, 10.2),
5.96 (dd, 1, J ) 5.5, 10.2). 13C-NMR (CDCl3) 14.0, 33.4, 53.5,
55.3, 59.5, 61.6, 94.7, 127.7, 129.3, 168.1. HRMS (EI) Calcd
for C13H20O6: 272.1260 (M+). Found: 272.1248.
Ben zoa te 11.18,19 Benzoate 11 was obtained in 64% yield.
TLC Rf ) 0.60 (3:2 hexane:EtOAc). IR (CCl4) 3063 (w), 2988
(m), 2931 (s), 2888 (s), 2831 (s), 1725 (s), 1606 (m), 1475 (m),
1450 (s). 1H-NMR (CDCl3) 3.45 (s, 3), 3.85-4.01 (m, 2) 4.87
(br s, 1), 5.49-5.78 (m, 1), 5.85-5.92 (dm, 1, J ) 10.0), 6.02-
6.08 (dm, 1, J ) 10.0), 7.36-8.03 (m, 5). 13C-NMR (CDCl3)
55.8, 60.0, 65.4, 95.2, 128.4, 129.0, 129.1, 129.7, 129.8, 133.2,
165.9. HRMS (EI) Calcd for C13H14O4: 234.0892 (M+).
Found: 234.0905.
Sch em e 11
Diester 12. Diester 12 was obtained in 87% yield. TLC
Rf ) 0.38 (3:1 hexane:EtOAc). IR (CCl4) 3050 (w), 2988 (s),
2931 (s), 2906 (s), 2888 (s), 2825 (m), 1756 (s) 1738 (s), 1463
(s), 1444 (s), 1388 (s), 1369 (s). 1H-NMR (CDCl3) 1.21 (t, 3, J
) 7.1), 2.98-3.07 (m, 1), 3.18 (d, 1, J ) 9.1), 3.34 (s, 3), 3.63-
3.73 (m, 2), 4.13, (q, 4, J ) 7.1), 4.75-4.77 (m, 1), 5.67-5.74
(dm, 1, J ) 10.3), 5.81-5.88 (dm, 1, J ) 10.3). 13C-NMR
(CDCl3) 13.9, 34.0, 52.9, 55.2, 60.2, 61.5, 95.0, 122.0, 130.2,
167.5. HRMS (EI) Calcd for C13H20O6: 272.1260 (M+).
Found: 272.1273.
Ben zoa te 13. Benzoate 13 was obtained in 67% yield. Mp
47.5-49.5 °C (hexane/Et2O). TLC Rf ) 0.55 (3:1 hexane:
EtOAc). IR (CCl4) 3063 (w), 2994 (s), 2963 (m), 2938 (s), 2856
(m), 2831 (m), 1719 (s), 1600 (m), 1450 (s), 1400 (s), 1375 (s),
1339 (s), 1318 (s). 1H-NMR (CDCl3) 1.33 (d, 3, J ) 6.5), 3.49
(s, 3), 3.97 (dq, 1, J ) 2.5, 6.5), 5.04 (d, 1, J ) 1.2), 5.21-5.23
(m, 1), 5.94-5.97 (dm, 1, 10.1), 6.10-6.14 (dm, 1, J ) 10.1),
7.37-7.41 (m, 2), 7.50-7.54 (m, 1), 8.04-8.07 (m, 2). 13C-NMR
(CDCl3) 16.6, 55.1, 66.5, 69.6, 98.0, 127.1, 128.4, 129.8, 129.9,
132.7, 133.1, 166.2. HRMS (EI) Calcd for C14H16O4: 248.1049
(M+). Found: 248.1048.
) 6.5, 6.5), 5.10 (s, 1), 5.27 (d, 1, J ) 6.5), 5.89 (d, 1, J ) 10.2),
6.01 (d, 1, J ) 10.2), 7.39-7.43 (m, 2), 7.52-7.55 (m, 1), 8.00-
8.02 (m, 2). 13C-NMR (CDCl3) 18.5, 54.9, 70.1, 71.3, 97.0,
125.9, 129.2, 129.4, 129.6, 130.2, 133.1, 165.9. HRMS (EI)
Calcd for C14H16O4: 248.1049 (M+). Found: 248.1050.
Diester 17. This compound was obtained in 95% yield as
a 9 to 1 ratio of diesters 17 and 18. TLC Rf ) 0.19 (9:1 hexane:
EtOAc). IR (CCl4) 3050 (w), 2975 (s), 2931 (s), 2906 (s), 2875
(s), 2825 (s), 1756 (s), 1738 (s), 1444 (m), 1400 (s), 1369 (s).
1H-NMR (CDCl3) 1.22 (t, 6, J ) 6.9), 1.32 (d, 3, J ) 6.6), 2.67
(ddd, 1, J ) 3.4, 4.3, 9.9), 3.38 (s, 3), 3.43 (d, 1, J ) 9.9), 3.91
(dq, 1, J ) 3.4, 6.6), 4.15 (q, 4, J ) 6.9), 4.83 (s, 1), 5.76 (d, 1,
J ) 10.3), 5.89 (dd, 1, J ) 4.3, 10.3). TLC Rf ) 0.41 (3:1
hexane:EtOAc). 13C-NMR (CDCl3) 14.0, 20.2, 39.0, 52.0, 53.9,
55.0, 61.4, 61.6, 69.3, 95.4, 127.7, 167.8, 168.2. HRMS (EI)
Calcd for C14H22O6: 286.1416 (M+). Found: 286.1403.
Ben zoa te 19. Benzoate 19 was obtained in 86% yield. TLC
Rf ) 0.25 (9:1 hexane:EtOAc). IR (CCl4) 3063 (m), 2981 (s),
2956 (s), 2913 (s), 2875 (s), 2819 (m), 1720 (s), 1700 (s), 1613
(w), 1594 (w), 1450 (s), 1419 (s). 1H-NMR (CDCl3) 1.29 (d, 3,
J ) 6.4), 3.83 (dq, 1, J ) 2.2, 6.4), 4.15-4.32 (m, 2), 5.19-5.21
(m, 1), 6.02-6.10 (m, 2), 7.38-7.55 (m, 3), 8.07 (d, 2, J ) 7.2).
13C-NMR (CDCl3) 16.4, 65.5, 66.9, 71.9, 122.6, 128.1, 129.6,
129.9, 132.1, 132.8, 166.4. HRMS (EI) Calcd for C13H14O3:
218.0943 (M+). Found: 218.0952.
Diester 20. Diester 20 was obtained in 85% yield. TLC Rf
) 0.27 (9:1 hexane:EtOAc). IR (CCl4) 3038 (w), 2981 (s), 2963
(s), 2938 (m), 2906 (m), 2869 (m), 2825 (m), 1756 (s), 1738 (s),
1463 (m), 1444 (m), 1369 (s). 1H-NMR (CDCl3) 1.19 (d, 3, J )
6.9), 1.24 (t, 6, J ) 7.1), 2.69-2.76 (dm, 1, J ) 10.0), 3.55 (d,
1, J ) 10.0), 3.72 (dd, 1, J ) 3.3, 11.9), 3.85 (d, 1, J ) 11.9),
4.09-4.23 (m, 5), 5.71-5.73 (m, 2). 13C-NMR (CDCl3) 14.1, 20.9,
34.2, 54.9, 61.4, 61.5, 66.2, 70.8, 124.6, 134.0, 168.0. HRMS
(EI) Calcd for C13H20O5: 257.1389 (M - H+). Found: 257.1402.
Ben zoa te 21. Benzoate 21 was obtained in 86% yield. TLC
Rf ) 0.32 (9:1 hexane:EtOAc). IR (CCl4) 3069 (w), 3044 (m),
Diester 14. This compound was obtained in 92% yield as
a 5 to 1 ratio of diesters 14 and 15. TLC Rf ) 0.33 (3:1 hexane:
EtOAc). IR (CCl4) 3050 (w), 2981 (s), 2931 (s), 2906 (s), 2875
(s), 2825 (m), 1762 (s), 1738 (s), 1663 (w), 1469 (s), 1444 (s),
1369 (s). 1H-NMR (CDCl3) 1.14 (d, 3, J ) 6.9), 1.20 (t, 6, J )
6.8), 2.95-2.98 (m, 1), 3.36 (s, 1), 3.47 (d, 1, J ) 9.3), 3.99 (q,
4, J ) 6.8), 4.15 (br q, 1, J ) 6.9), 4.97 (br s, 1), 5.66-5.69
(dm, 1, J ) 10.3), 5.84-5.88 (dm, 1, J ) 10.3). 13C-NMR
(CDCl3) 13.9, 17.2, 37.6, 52.0, 54.4, 61.4, 69.0, 97.2, 128.4,
129.6, 168.3. HRMS (EI) Calcd for C14H22O6: 286.1416 (M+).
Found: 286.1419.
Ben zoa te 16.20 Benzoate 16 was obtained in 95% yield.
IR (CCl4) 3094 (w), 3063 (m), 2988 (s), 2938 (s), 2875 (s), 2825
(s), 1725 (s), 1606 (m), 1494 (m), 1469 (s), 1450 (s), 1406 (s).
1H-NMR (CDCl3) 1.35 (d, 3, J ) 6.5), 3.46 (s, 3), 3.99 (dq, 1, J

Addition of Carbon Nucleophiles to Dihydropyran Derivatives
J . Org. Chem., Vol. 62, No. 5, 1997 1263
2981 (s), 2938 (s), 2869 (m), 2831 (s), 1725 (s), 1606 (m), 1450
(s), 1375 (s). 1H-NMR (CDCl3) 1.28 (d, 3, J ) 6.5), 3.77 (dq, 1,
J ) 6.5, 6.5), 4.13-4.25 (m, 2), 5.26-5.28 (m, 1), 5.81-5.84
(dm, 1, J ) 10.3), 5.91-5.94 (dm, 1, J ) 10,3), 7.41 (t, 2, J )
7.6), 7.53 (t, 1, J ) 7.6), 8.02 (d, 2, J ) 7.6). 13C-NMR (CDCl3)
18.1, 64.6, 71.1, 72.0, 124.4, 128.3, 129.6, 129.7, 130.1, 133.0,
166.0. HRMS (EI) Calcd for C13H14O3: 218.0943 (M+).
Found: 218.0935.
Diester 23. This compound was obtained in 87% yield as
a 1 to 7 ratio of diesters 23 and 24. TLC Rf ) 0.18 (9:1 hexane:
EtOAc). IR (CCl4) 3044 (w), 2981 (s), 2938 (s), 2906 (m), 2875
(m), 2831 (m), 1756 (s), 1731 (s), 1456 (m), 1438 (s), 1394 (s).
1H-NMR (CDCl3) 1.13-1.24 (m, 9), 2.59-2.64 (m, 1), 3.46 (d,
1, J ) 6.7), 3.69 (dq, 1, J ) 6.1, 6.1), 3.97-4.05 (m, 2), 4.08-
4.21 (m, 4), 5.66-5.79 (m, 2). 13C-NMR (CDCl3) 13.9, 14.0,
18.1, 40.0, 53.8, 61.2, 61.5, 62.5, 70.4, 123.9, 127.6, 168.0.
HRMS (EI) Calcd for C13H20O5: 257.1389 (M - H+). Found:
257.1402.
Diester 29. Diester 29 was obtained in 43% yield. TLC Rf
) 0.24 (9:1 hexane:EtOAc). IR (CCl4) 3038 (w), 2956 (s), 2931
(s), 2090 (w), 2856 (m), 1756 (s), 1738 (s), 1463 (m), 1369 (m),
1256 (s). 1H-NMR (CDCl3) 0.03 (s, 6), 0.83 (s, 9), 1.15 (d, 3, J
) 6.2), 1.25 (t, 6, J ) 7.2), 1.44-1.51 (m, 1), 1.73-1.79 (m, 1),
2.72-2.77 (dm, 1, J ) 10.0), 3.52 (d, 1, J ) 10.0), 3.67-3.72
(dm 1, J ) 11.7), 3.79-3.82 (dm, 1, J ) 11.7), 3.96-4.01 (m,
1), 4.09-4.22 (m, 5), 5.75-5.78 (m, 2). 13C-NMR (CDCl3) -4.8,
-4.3, 14.0, 14.1, 23.5, 34.3, 44.9, 54.7, 61.3, 61.4, 65.5, 65.9,
72.0, 124.9, 132.7, 168.3. HRMS (FAB) Calcd for C21H37O6Si:
415.2516 (M - H+). Found: 415.2531.
(CDCl3) 21.2, 41.5, 70.3, 71.4, 126.8, 128.0, 128.1, 128.5, 131.9,
141.3. HRMS (EI) Calcd for C12H14O:174.0966 (M+). Found:
174.0965.
Data for 34: TLC Rf ) 0.41 (9:1 hexane:EtOAc). 1H-NMR
(CDCl3) 1.14 (d, 3, J ) 6.1), 3.21-3.23 (m, 1), 3.52 (dq, 1, J )
2.8, 6.1), 4.29-4.32 (m, 2), 5.74-5.78 (dm, 1, J ) 10.2), 5.89-
5.93 (dm, 1, J ) 10.2), 7.17-7.25 (m, 3), 7.29-7.33 (m, 2).
Dihydropyran 34 could not be separated completely from 33
so it was not further characterized.
Dih yd r op yr a n 35. This mixture of dihydropyrans was
obtained in 65% yield as a 4:2:15:1 ratio of 33-36, respectively.
TLC Rf ) 0.45 (9:1 hexane:EtOAc). IR (CCl4) 3088 (w), 3063
(s), 3031 (s), 2981 (s), 2931 (s), 2906 (s), 2875 (s), 2850 (s), 2819
(s). 1H-NMR (CDCl3) 1.31 (d, 3, J ) 6.8), 3.28-4.30 (m, 1),
3.81 (dd, 1, J ) 4.2, 11.1), 3.95 (dd, 1, J ) 4.2, 11.1) 4.27-4.30
(m, 1), 5.81-5.87 (m, 2), 7.20-7.24 (m, 2), 7.27-7.32 (m, 3).
13C-NMR (CDCl3) 20.8, 40.9, 69.1, 69.8, 126.5, 126.9, 128.1,
128.3, 131.8, 142.7. HRMS (EI) Calcd for C12H14O: 174.1044
(M+). Found: 174.1038.
Ack n ow led gm en t. We thank Dr. Yiu-Fai Lam and
Ms. Caroline Homonnay-Preyer for their assistance in
obtaining NMR and mass spectral data. We also thank
Professors J ames White and Dennis Curran for provid-
ing spectral data of dihydropyran 2 and related deriva-
tives. The generous financial support of the University
of Maryland and an unrestricted grant from Lederle
Laboratories is acknowledged.
Dih yd r op yr a n s 33 a n d 34. This compound was obtained
in 65% yield as a 3 to 2 ratio of dihydropyrans 33 and 34. Data
for 33: TLC Rf ) 0.45 (9:1 hexane:EtOAc). IR (CCl4) 3088 (w),
3063 (s), 3031 (s), 2981 (s), 2969 (s), 2931 (s), 2875 (s), 2850
(s), 2825 (s), 1606 (s). 1H-NMR (CDCl3) 1.29 (d, 3, J ) 6.7),
3.45 (dd, 1, J ) 9.8, 11.1), 3.61-3.66 (m, 1), 4.09-4.13 (m, 1),
4.28-4.31 (m, 1), 5.80-5.84 (dm, 1, J ) 10.2), 5.86-5.90 (dm,
1, J ) 10.2), 7.19-7.23 (m, 3), 7.30-7.34 (m, 2). 13C-NMR
Su p p or tin g In for m a tion Ava ila ble: 1H-NMR spectra of
compounds 4, 9, 12-14, 17, 19-21, 23, 29, and 33-35 (14
pages). This material is contained in libraries on microfiche,
immediately following this article in the microfilm version of
the journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
J O961973D