Tealhedron Lacrs. Vol. 34. No. 9, PP. 1429-1432 1993

Rmted in Great Britain

cuModa9/93 s6.oo+M) PCXgMlXmPWSLtd

smREos-SYNTBESES OFTEREEWTERENT CLASSES OF NEOLlGNANS FROM TEE SAME STARTJNG IWTERLU!S

Thomas A EngIer,* Dong (Donna) Wei and MichaeI A btavic Department of caemhy, university of KaMa& Lawrence3 Kansas 66045-0046

ABSl7tAC7? Lewis Acid catalyzed reactions of 2alkoxy-~l,4-ben=@m- withstyrenescanbe manipulated to form either the burchellin or guianin neo* or 4-allyl-2-aryM-methyl-2J- dih)&ObUlZOfUlXUlS.

Neolignans are naturally occur&g dimers of pro~lphenols.’ More than Meen diEerent Qpes of

neolignan frameworks have been isolated, in&ding those found in burchellin, l, guiani~ 2, and kadsurenone,

4; a related framework is 3 which has not been obsetved in a natural product to our knowledge. Interest in the

neolignans arises from .their widespread ouxence and diverse bMogi4A activuy..‘~ we have reparted

syntheses of the carbon frameworks 4 and 5 starting with reactions of styrenes with Zalkoxy-, and Zalktxy-6

alkyl-1,4benzoquinones.* We now report a new method to sekWive& fom eompowIcLF with jitmeworks I, 2 or

3 via re&ions of 2-aUca*y-54Uyl.I,~.3

Reactions of styrenes 6 with quinones 74 promoted by excess Ti(IV), either as TicI, or a reagent

prepared from mixtures of TiC14:Ti(OiPr)42d , produced cyclobutanes 8 and 9 with the former as the major, if

not exclusive, product (Scheme I and Table, entries l-8). The formation of 8 was unexpecM based on previous

mechanistic hypothesis.* However, the SnC14-promoted reactions of 6 with 7 at -78 ‘c gave either the

cyclobutane 9 (entries 9-l&13,14) or, in the case of 6a with 7b, the bicyclic adduct l3a (Ar=C6H5, entry 12).

Products g, 9 and l3a were all isolated as single diastere~merr?~ Upon warming the SrKQpromoted

reactions of 6b/c with 7b to -30 “c (entries 15 16), mixtures of the keto-enol tautomers 15 and 16 were found

which may result from rearrangement of 9 via 14. Methylation of the mixunes ga& 17 [Ar=3,4-(OCH#!t&

64%j’ and burchelhn, 1(80%), respectively.6

Treatment of cyclobutanes 8 and 9 with CP&02H in C!F5CH20H or THP at room temperature or above

effected their rearrangement to 12 ’ (in 4568% yiel&) and l3a (in 3974% yields), mly, presumably via

intermediates 10 and 11. Small amounts of quinones 7r/b (932%) were also found resulting ikom fragmentation

of 10/11. The dihydrobcn&urans 32 were formed as a mixture of tmns and cis isomers with the former

predaminating(~12:1).Tbebicyclicproducrr~were~~u~diarterecrmerr6aodwenevidently

1429

1430

Table. Lewis Acid Promoted Reactions of Styrenes 6 with %AUy&akoxy-1,4-benzoquinone~ 7.

1 6a 2 6a 3 6b

: 6b 6b

6 6c 7 6c 8 6c 9 6a 10 6b 11 6c 12 6a 13 6b 14 6c 15 6h 16 6c

78 7s 7s 7a 7a 7a 70 7b 7a 70 7a 7b 7b 7b 7b 7b

TiCI (i) TiCI (4J

‘Ti(IV)-solid’ (4.5)’ $1 TiQ,:Ti(OiPr), (2) 2:l lX14tTi(OiPr)4 (3) 3:l ‘IQ:Ti(OPr)4 (2) ‘Ti(IV’)-solid”c (2.3)’ ‘Ti(IV)-solid~ (1.5)’ . _

=&a g- .

s&l; (1) SKI.4 (1.5) SnQ4 (2)

SnQ4 (2)

sncl, (1) SnQ4 (1)

-78 -78 -78 -78 -78 -78 -78 -78 -78 -78 -78

-7860 -78 -78

-78+30

_ _ dd’(&&)

82 la HlM

97 (f:O) 3 70 (1:O) - 3 83 (1:O) - - 72 (4~1) - - 76 (3:l) - - 83 (1:O) - -

92 (5.7:1) - - 91 (2S:l) - - 34 (0:l) - - 7&(0:1) - - 87 (OA) - -

80 &) 71 - - 16

86 (OA) - - - 78

-7&-30 - - 69

a) With respect to quinone. b) Isolated yields. c) Indicates that none of this product was isolated, d) An oE- white precipitate prepared by mix& TiCI and Ti(OiPr), is a 5:l proportion in CH2Cl2 at r.t.; the solid is separated and dried under vacuum. e) By weight with respect to quinone.

produced from 11 by intramolecular attack of the carbocation center on the en01 ether moiety followed by

dealkylation. In contrast, at -30 “c the related intermediates 14 underwent C-O bond formation and dealkylation

to give U/16. In 14, coordination of the Sn(iV) to the enol ether oxygen apparently lowers the nuckophihcity

of the carbon-carbon double bond. Methylation of l3a with CH#K$O~ in acetone Save l3b (in 75-8996 yield),

one of which [Ar-3,4-(OCH,O&H~ has been previously converted to the natural product guiak, 2.s6

One rationale for the selective formation of 8,9 and l3 involves an quiliirium between coordination

of the Lewis acid to 7 at either the C-l or G4 carbonyl to @ve 18 and 29, re.spectively (Schemes II and III).

The former may be favored due to chelation effects, however, the C-4 oxygen is more basic. Then, with SnCa,

as catalyst, cycloaddition of l& with the styrene rtereoselectively gives the O-stabilized carbocation 1)

(thermodynamic control, Scheme If); the aryl group of the styrene would be expected to occupy an endo position

with respect to the pentadienyl carbocation moiety in the cycloaddition.s3 CkavaSe of the Cl/G7 bond then

gives the benzylic carbocation 14 and, at -78 “C, intramolecular addition of the cationic carbon to the “tin-

enolate” moiety occurs to yield 9. The tram relationship between the atyl and methyl groups on the cyclobutane

ringht9wouldbeexpeckdforstericreasons and since the bond between C-5 and C-6 is established in the

initial cycloaddition and remains intact during the reatrangement of 19 to 9, the relative stereochemistry of the

remaining stereogenic centers follows from the cycloaddition. The formation of l3 in reaction of 6a with 7b

results from debenzylation of 19?” In reactions of&/c, debenzylation of 19 does not compete with the ckavage

of the Gl/C-7 bond to @ve 14 (and then 9) due to stabilization of the earbocation center in 14 by the aromatic

lk, R-H 12b, R&Ha

14 R 15 16 17

ForM7,XandRerethesameesinstwtingS/7.

&home II Scheme III

ring substituents. With excess Ti(lV) as catalyst, the product expect4 hm cycbaddition of Ti(lV)quinone

campkx29togivetlandthen2tand8isfound(Schemem). Thismaybeduetoscveralfactors. han

quiliirium mixture of 18b and 2Q, the gcbaddition across G3/GS oftbe former with the styrene may be slower

thancycloadditionaaossGUC6oftbelatterduetorteric~importdbytbeGS~gmup. Thus,

formation of 21 may be kiuetically determined. Alternatively, the Ti(IV) may bind to the r-face of the quinone

~~bothG~andG3/c5ofthe~p~tocycloaddition2’1andthehttcrir~rlawerdueto~c

hindrance to bond formation between C-8 of the styrene and G5 of the compkx A third possibility is that a

q&one-[Ti(IV)h compl& my be im~hvd and ~oaddition acrms G2KX is again fhored for s&c

reasonsP

1432

ACWOWLEZIGMENZS. This work was supported financially by the National Institutes of Health, the National Science Foundation, the Alfred P. Qoan Foundation and Eli JiJly and Company. We thank Professor Steve Angle for helpful disct&ons, spectral data and shating the results of his experiments prior to publication.

REFERENCES AND NOTES.

1. a) Gottlieb, 0. R.; Yoshida, hf. In “Natural Products of Woody Plants I” Rowe, J. W., E&, Springer- Verlag: Berlin, 1989, p. 439. b) Gottlieb, 0. R. ??vg cha. Org. Nat. prod 1978,35,1. c) Ponpipom, hf. M.; Bugianti,, R. L, Brooker, D. R.; Yue, B. -Z, Hwang, S.-B.; Shen. T.-Y. 1 Mcd cha 1987,30, 136. d) Hirano, T.; Wakasugi, A, Oohara, M.; Oka, Ic; Sashida, Y. Plunra Mcd l991,57,331. e) Hattori, M.; Hada, S.; Watahiki, A.; Jhara, W.; Shu, Y.-Z.; Kakiuchi, N.; Mizuno, T.; Namba, T. Char. wuunl. Bug, 1986,34,3885. f) See also, Agrawal, P. K.; Thakur, R. S. Org. Msg. Rts. 1985,23,389.

2. a) Engler, T. k; Combrink, K D.; Ray, J. E. J. Am Cha. Sue. 1988,210. 7931. b) Engler, T. A, Jxtavic, M. k, Reddy, J. P. ibid. 1991, 113, 5068. c) Engler, T. k, Letavic, M. k; Combrink, K. D.; Takusagawa, F. J. Org. Chem. l!J90,55,5810. d) Engler, T. A, Letavic, M. k; Takusagawa, F. Tetmhedron Let% 1992, in press.

3. For leading references to previous synthetic efforts, see reference 2 and a) Btichi, G.; Mak, C. -P. J. Am Chtvt Sot. 1977, 99,8073. b) Angle, S. R.; Turnbull, K. D. ibid. 1999,112, 3698. c) Mottlock, S. V.; Seckington, J. K; Thomas, E. J. J. Chem. Sot. Per&in Zh~ns X1988,2305. d) Shizuri, Y.; Shigemori, H.; Suyama, Ic, Nakamura, Ic; Okuno, Y.; Ohkubo, M.; Yamamura, S. In “Studies in Natural Products Chemistry, Vol. 8” Atta-ur-Rahman, Ed.; Elsevier: Amsterdam, 1991; p. 159. e) Gates, B. D.; Dalidow@ P.; Tebben, k, Wang, S.; Swenton, J. S. J. Org. Chem, 1992,57, 2135.

4. Prepared in ~54% yield by Tic14 catalyzed reaction of 2-methoxy- and 2-benzykuty-1,4benzoquinone~ with allyltrimethylsilane; see, Hosomi, A; Sakurai, H. Tetyhedmn L&t. 1977,nW1.

5. All compounds were characterized by high field H (1508 MHz) and C (125 MHz) NMR, JR and/or exact mass. AU ‘H and 13C resonances were assigned by ‘H- H decoupling and heteronu&ar multiple bond correlation (HMBC) experiments and the stereochemistry in cyclobutanes 81’9 was unambiguously determined by ‘H-‘H NOE experiments.ti

6. The high field 13C and ‘H spectra of W and l/l7 were consistent with data reported previously. For W, see refereirces 3b/c. For l/l% see references lb/f and, a) Wenkert, E., Gottlieb, H. E.; Gottheb, 0. R, Pereira, M. 0. da S.; Form& M. D. phycochaisay 1976, IS, 1547. b) Filho, R B.; Figliuolo, R; Got&b, 0. R. ibid. 1980,19,659. c) Gottlieb, 0. R, DaSihu, M. L, Ferreira, Z. S. ibid. l975,14,1825. d) Aiia, C J.; Fernandes, J. B.; Gottlieb, 0. R.; Maia, J. G. S. ibti. 1975,14,1597. We also thank Pr&ssor Angle for copies of additional spectra of compounds related to l3, and l/l7.

7. a) Hendrickson, J. B.; Singh, V. J. char Sot; Chuu CMmun. 1983,837. b) corcOran, R C, Ma, J. J. Am w Sot. 1991,113,8973.

S. a) Poll, T.; Metter, J. 0.; Hebnchen, G.Angetv. than Int. Fd A??& l985,24,1l2. b) Schultz, A G.; Lee, H. Tcnahcdron Lett. W&33,4397.

9. Of course, other mechanistic possiiilities exist. For example, direct alkylation of the the Lewis acid- quinone complau may give 14 or 22 without proceed@ through 19121. As SW by Swenw% the same I stacking interactions that favor the endo aryl orientation in the cycloaddition may also favor this orientation in a direct alkylation reaction. Although the resuhs do not discount alternate mechanirms, the cycloaadition mechanism nicely explains the facts and is useful as a working model for predict& purposes.

10. During the preparation of this manusuip~ Professor Steve Angle informed us that biyclo [3-U] octenedknms related to l3b reamtnge to benzofuranoids related to 17 upon treatment with TiQ4 at -78 Oc and warming to room temperature. See also, reference 3a

(Received in USA 7 October 1992; accepted 17 November 1992)