Stereocontrol between remote atom centers in acyclic substrates. Anti addition of hydride to 1,5-, 1,6-, and 1,7-hydroxy ketones

7964

J . Org. Chem. 1998, 63, 7964-7981

Ster eocon tr ol betw een Rem ote Atom Cen ter s in Acyclic

Su bstr a tes. An ti Ad d ition of Hyd r id e to 1,5-, 1,6-, a n d 1,7-Hyd r oxy

Keton es^†

Han-Cheng Zhang, Bruce D. Harris, Michael J . Costanzo, Edward C. Lawson,

Cynthia A. Maryanoff, and Bruce E. Maryanoff *

The R. W. J ohnson Pharmaceutical Research Institute, Spring House, Pennsylvania 19477

Received J uly 9, 1998

For conformationally unconstrained, acyclic organic compounds, the control of stereogenic centers

at remote positions of a chain, that is, at a distance of four or more atom centers, remains a

challenging problem in asymmetric synthesis. We report on our studies of 1,5, 1,6, and 1,7

diastereoselectivity in hydride reductions of acyclic hydroxy amino ketones and related compounds,

which were sparked by our discovery of high 1,5 diastereocontrol (>10:1) with substrates such as

17 and 23. We have been able to achieve both high 1,5- and 1,6-anti diastereocontrol in the reduction

of 1,5- and 1,6-hydroxy ketone substrates, respectively. However, the level of 1,7-anti diastereo-

control with 1,7-hydroxy ketones was only moderate. More specifically, reduction of 23 to 24 with

R-alpine-hydride or Zn(BH₄)₂in CH₂Cl₂(predominantly) at -78 °C gave high 1,5-anti stereoselec-

tivity (anti/syn ) 10:1 or 13:1, respectively), and reduction of 34 to 35 with R-alpine-hydride (CH₂-

Cl₂) gave high 1,6-anti selectivity (anti/syn ) 12:1, respectively), whereas reduction of 46 to 44

with R-alpine-hydride (CH₂Cl₂) gave only moderate 1,7-anti stereoselectivity (anti/syn ) 3:1).

Results for reductions of 1,5- and 1,6-hydroxy ketone substrates having the N-benzyl structural

subunit replaced (i.e., 27 f 28, 29 f 30, 31 f 32, 52 f 53, 54a f 55a , 54b f 55b, 54c f 55c,

and 56 f 57) clearly indicate that the stereoelectronic character of this subunit plays a critical

role in the attainment of high anti asymmetric induction. Thus, while we obtained exceptionally

high 1,6-anti stereoselectivity in the reduction of the N-mesitylmethyl substrate, 54c, to 1,6-diols

55c (anti/syn ) 22:1) with R-alpine-hydride at -78 °C in CH₂Cl₂, the N-methyl substrate, 54b,

gave a relatively modest anti/syn ratio of 3:1. The diminished anti/syn ratio of 4:1 in the R-alpine-

hydride reduction of methoxy amino ketone 50 to 51 also indicates the importance of the free

hydroxyl group for attaining high 1,6-anti stereoselectivity. To rationalize the high remote anti

stereocontrol in such acyclic systems, we discuss a chelation-controlled mechanism, involving

external hydride addition to a bicyclic metal complex with a coordinated ketone carbonyl (e.g., 33)

vs internal hydride addition to a monocyclic metal complex with an uncoordinated ketone carbonyl

(e.g., 58).

In dealing with conformationally unconstrained, acyclic

organic compounds, the control of stereogenic centers at

remote positions of the chain, that is, at a distance of

four or more atom centers, remains a challenging task.¹

Only during the past 10 years has this difficult problem

in asymmetric synthesis succumbed to a stream of

notable successes. The types of substrates used in

successful transformations have varied greatly, but the

most remarkable cases are those involving substrates

devoid of any conformationally rigidifying structural

features, such as a cyclic array or an alkene group, to

serve as a facilitator for the transfer of asymmetry (the

“stereocommunication” event¹). Indeed, various ex-

amples of high 1,4,^1,21,5,^2o,r,31,6,⁴and 1,7 to 1,14⁵remote

asymmetric induction in “strictly acyclic systems” have

now appeared in the literature, including our own

preliminary results concerning 1,5 and 1,6 diastereocon-

* Author to whom correspondence should be addressed. Fax: 215-

628-4985. E-mail: bmaryano@prius.jnj.com.

^†This paper is dedicated to Prof. Kurt Mislow on the occasion of

his 75th birthday.

(1) Mikami, K.; Shimizu, M. J . Synth. Org. Chem. J pn. 1993, 51,

21-31.

(2) (a) Myers, A. G.; Yang, B. H.; Chen, H.; Mckinstry, L.; Kopecky,

D. J .; Gleason, J . L. J . Am. Chem. Soc. 1997, 119, 6496-6511. (b)

Captain, L. F.; Xia, S.; Liotta, D. C. Tetrahedron Lett. 1996, 37, 4293-

4296. (c) Magure, R. J .; Mulzer, J .; Bats, J . W. Tetrahedron Lett. 1996,

37, 5487-5490. (d) Tomooka, K.; Keong, P.-H.; Nakai, T. Tetrahedron

Lett. 1995, 36, 2789-2792. (e) Sato, T.; Kido, M.; Otera, J . Angew.

Chem., Int. Ed. Engl. 1995, 34, 2254-2256. (f) Molander, G. A.; Haar,

J . P., J r. J . Am. Chem. Soc. 1993, 115, 40-49. (g) Shimizu, M.; Mikami,

K. J . Org. Chem. 1992, 57, 6105-6106. (h) Molander, G. A.; Haar, J .

P., J r. J . Am. Chem. Soc. 1991, 113, 3608-3611. (i) Panek, J . S.; Yang,

J . J . Am. Chem. Soc. 1991, 113, 6594-6600. (j) Trost, B. M.; Urabe,

H. J . Org. Chem. 1990, 55, 3982-3983. (k) Fujisawa, T.; Takemura,

I.; Ukaji, Y. Tetrahedron Lett. 1990, 31, 5479-5482. (l) Corey, E. J .;

Hannon, F. J .; Boaz, N. W. Tetrahedron Lett. 1989, 45, 545-555. (m)

Reetz, M. T.; Steinbach, R.; Westermann, J .; Urz, R.; Wenderoth, B.;

Peter, R. Angew. Chem., Int. Ed. Engl. 1982, 21, 2254-2255. (n) Reetz,

M. T.; Kessler, K.; Schmidtberger, S.; Wenderoth, B.; Steinbach, R.

Angew. Chem., Int. Ed. Engl. 1983, 22, 989-990. (o) Tomooka, K.;

Okioaga, T.; Suzuki, K.; Tsuchihashi, G. Tetrahedron Lett. 1987, 28,

6335-6339. (p) Nishigaichi, Y.; Takuwa, A.; J odai, A. Tetrahedron Lett.

1991, 32, 2383-2387. (q) Nakata, T.; Schmid, G.; Vranesic, B.;

Okigawa, M.; Smithpalmer, T.; Kishi, Y. J . Am. Chem. Soc. 1978, 100,

2933-2935. (r) Still, W. C.; Darst, K. P. J . Am. Chem. Soc. 1980, 102,

7385-7387. (s) Harada, T.; Matsuda, Y.; Imanaka, S.; Oku, A. J . Chem.

Soc., Chem. Commun. 1990, 1641-1643.

(3) (a) Thomas, E. J . Chem. Commun. 1997, 411-418. (b) Evans,

D. A.; Coleman, P. J .; Cote, B. J . Org. Chem. 1997, 62, 788-789. (c)

Hanessian, S.; Yang, H.; Schaum, R. J . Am. Chem. Soc. 1996, 118,

2507-2508. (d) Paterson, I.; Gibson, K. R.; Oballa, R. M. Tetrahedron

Lett. 1996, 37, 8585-8588. (e) Fang, J .-M.; Chang, C.-J . J . Chem. Soc.,

Chem. Commun. 1989, 1787-1788. (f) McNeill, A. H.; Thomas, E. J .

Tetrahedron Lett. 1992, 33, 1369-1373. (g) McNeill, A. H.; Thomas,

E. J . Tetrahedron Lett. 1990, 31, 6239-6243. (h) Pippel, D. J .; Curtis,

M. D.; Du, H.; Beak, P. J . Org. Chem. 1998, 63, 2-3.

(4) (a) Quallich, G. J .; Keavey, K. N.; Woodall, T. M. Tetrahedron

Lett. 1995, 36, 4729-4732. (b) Stanway, S. J .; Thomas, E. J . Tetrahe-

dron Lett. 1995, 36, 3417-3420. (c) Carey, J . S.; Thomas, E. J .

Tetrahedron Lett. 1993, 34, 3935-3938.

Published on Web 10/06/1998

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7965

Sch em e 1

trol in the addition of hydride to hydroxy amino ketones.⁶

It is particularly encouraging for future progress in this

field that these positive results have been associated with

diverse synthetic transformations, such as aldol condensa-

tions,^3b1,4 conjugate additions,^2balkylations,^2adirect

additions to aldehydes and ketones,^2h,s,4a,6molecular

rearrangements,^2dand free radical reactions.^3cA striking

recent development is the 1,13/1,14 stereocontrol in the

aldol condensation of benzaldehyde with chiral acyclic

sulfones to produce only two of four possible diastereo-

mers.^5a,bOther exciting examples include the highly

selective 1,7 asymmetric induction in tin(IV) bromide

promoted condensations of aldehydes with 6-hydroxyall-

ylstannanes (syn/anti ) 95:5),^5cthe 1,6 induction in tin-

(IV) bromide promoted reactions between aldehydes and

6-hydroxy-5-methyl-hex-2-enyl(tributyl)stannanes (anti/

syn ) 93:7),^4bthe 1,5 asymmetric induction in aldol

condensations of methyl ketones (anti/syn ) 95:5),^3band

the 1,4 induction in the alkylation of N-acylated pseu-

doephedrines (99% de).^2aIn the context of remote asym-

metric induction, there are several remarkable examples

of stereocommunication over four or more atom centers

in systems with a cyclic array (e.g., cyclic acetal, oxazo-

line, cyclic boronate), which could well supply a certain

measure of conformational constraint. For these types

of substrates, which are not “strictly acyclic”, high 1,4,⁷

1,5,⁸1,6,⁹and 1,7¹⁰asymmetric inductions have been

realized.

A common feature present in many of the examples of

high remote diastereocontrol seems to be interaction

between a prostereogenic center and a remote stereogenic

center by means of a highly organized intermediate

species, especially one involving formation of a transient

cyclic metal complex that can bring the prostereogenic

center and the remote stereocenter into reasonable prox-

imity. This is illustrated nicely by the work of Magnus

and co-workers, where reacting chiral amino sulfone

amides 1 are presumably disposed as bicyclic metal

chelates 2, bringing the stereogenic element into the

vicinity of the reacting carbanion center to achieve 1,13

asymmetric induction in alkylation with benzaldehyde

(eq 1, n ) 4).^5a,bBesides this type of chelation control,

remote acyclic diastereocontrol has also been effected well

with π-allyltricarbonyl iron complexes^8gand arene tri-

carbonyl chromium complexes.^8lExamples of effective

1,4 and 1,5 asymmetric induction in the absence of

putative chelation control have also appeared in the

literature, such as in alkylation of pseudoephedrines,^2a

in stereoselective addition to alkoxy acetals,^2fand in

C-alkylation of, or 1,4-addition to, Evans-type oxazoli-

dinones^7i,8cor Oppolzer-type camphorsultams.^7b,8m

(5) (a) Magnus, N.; Magnus, P. Tetrahedron Lett. 1997, 38, 3491-

3494. (b) Linnane, P.; Magnus, N.; Magnus, P. Nature 1997, 385, 799-

801. (c) Carey, J . S.; Thomas, E. J . J . Chem. Soc., Chem. Commun.

1994, 283-284. (d) Denmark, S. E.; Marble, L. K. J . Org. Chem. 1990,

55, 1984-1986.

(6) Preliminary communications: (a) Zhang, H.-C.; Costanzo, M. J .;

Maryanoff, B. E. Tetrahedron Lett. 1994, 35, 4891-4894. (b) Zhang,

H.-C.; Harris, B. D.; Maryanoff, C. A.; Maryanoff, B. E. Tetrahedron

Lett. 1996, 37, 7897-7900. (c) For a writeup of an invited lecture, see:

Maryanoff, B. E.; Zhang, H.-C.; Costanzo, M. J .; Harris, B. D.;

Maryanoff, C. A. In Reductions in Organic Synthesis; Abdel-Magid, A.

F., Ed.; American Chemical Society: Washington, DC, 1996; pp 138-

152.

(7) (a) Akhoon, K. M.; Myles, D. C. J . Org. Chem. 1997, 62, 6041-

6045. (b) Oppolzer, W.; Rosset, S.; De Brabander, J . Tetrahedron Lett.

1997, 38, 1539-1540. (c) Frost, C.; Linnane, P.; Magnus, P.; Spyvee,

M. Tetrahedron Lett. 1996, 37, 9139-9142. (d) Tomooka, K.; Nagasawa,

A.; Wei, S.-Y.; Nakai, T. Tetrahedron Lett. 1996, 37, 8899-8902. (e)

Enders, D.; Meiers, M. Angew. Chem., Int. Ed. Engl. 1996, 35, 2261-

2263. (f) Ghosh, A. K.; Onishi, M. J . Am. Chem. Soc. 1996, 118, 2527-

2528. (g) Sibi, M. P.; J i, J . Angew. Chem., Int. Ed. Engl. 1996, 35, 190-

192. (h) Sibi, M. P.; Deshpande, P. K.; J i, J . Tetrahedron Lett. 1995,

36, 8965-8968. (i) Davies, S. G.; Sanganese, H. J . Tetrahedron:

Asymmetry 1995, 6, 671-674. (j) Oppolzer, W.; Moretti, R.; Zhou C.

Helv. Chim. Acta 1994, 77, 2363-2381. (k) Oppolzer, W.; Blagg, J .;

Rodriguez, I.; Walther, E. J . Am. Chem. Soc. 1990, 112, 2767-2772.

(l) Evans, D. A. Aldrichimica Acta 1982, 15, 23-32.

(8) (a) Molander, G. A.; McWilliams, J . C.; Noll, B. C. J . Am. Chem.

Soc. 1997, 119, 1265-1276. (b) Meyers, A. I.; Stoianova, D. J . Org.

Chem. 1997, 62, 5219-5221. (c) Han, Y.; Hruby, V. J . Tetrahedron

Lett. 1997, 38, 7317-7320. (d) Wipf, P.; Takahashi, H. J . Chem. Soc.,

Chem. Commun. 1996, 2675-2676. (e) Kishida, M.; Eguchi, T.;

Kakinuma, K. Tetrahedron Lett. 1996, 37, 2061-2062. (f) Shimano,

M.; Meyers, A. I. J . Org. Chem. 1995, 60, 7445-7455. (g) Ley, S. V.;

Meek, G.; Metten, K.-H.; Pique, C. J . Chem. Soc., Chem. Commun.

1994, 1931-1932. (h) Nicolas, E.; Russell, K. C.; Hruby, V. J . J . Org.

Chem. 1993, 58, 766-770. (i) Meyers, A. I.; Shipman, M. J . Org. Chem.

1991, 56, 7098-7102. (j) Erker, G.; Sosna, F.; Betz, P.; Werner, S.;

Kruger, C. J . Am. Chem. Soc. 1991, 113, 564-573. (k) Tamura, R.;

Watabe, K.; Katayama, H.; Suzuki, H.; Yamamoto, Y. J . Org. Chem.

1990, 55, 408-410. (l) Uemura, M.; Minami, T.; Hirotsu, K.; Hayashi,

Y. J . Org. Chem. 1989, 54, 469-477. (m) Oppolzer, W.; Kingma, A. J .;

Poli, G. Tetrahedron 1989, 45, 479-488. (n) Matsumoto, T.; Matsuda,

F.; Hasegawa, K.; Yanagiya, M. Tetrahedron 1984, 40, 2237-2243.

(9) (a) Frey, L. F.; Tillyer, R.; Caille, A. S.; Xu, F.; Tschaen, D. M.;

Dolling, U.-H.; Grabowski, E. J . J . Abstracts of the 214th National

Meeting of the American Chemical Society, Las Vegas, NV, September

7-11, 1997; ORGN-176. (b) Yamamoto, Y.; Hara, S.; Suzuki, A. Synlett

1996, 883-884. (c) Roush, W. R.; Wada, C. K. J . Am. Chem. Soc. 1994,

116, 2151-2152. (d) Meyers, A. I.; Hanagan, M. A.; Trefonas, L. M.;

Baker, R. J . Tetrahedron 1983, 39, 1991-1999.

Several examples of high remote diastereocontrol in

the addition of nucleophiles to ketones have been

reported.^{2h,s,4a,6,7a,8n,9d,10a-c}For example, Akhoon and

Myles achieved high (>99:1) 1,4 asymmetric induction

in the diastereoselective additions of Grignard reagents

to chiral R-keto acetals 3 (eq 2),^7awherein a bicyclic

magnesium chelate, 6, is believed to be the reactive

species that allows for asymmetric induction in alcohol

formation. Meyers and co-workers^9ddeveloped an inter-

esting case of 1,6 remote asymmetric induction in the

preparation of chiral phthalides by using chiral oxazo-

lines, in which treatment of 7 with Grignard reagents

gave moderate to high (maximum of 90:10) diastereose-

(10) (a) Molander, G. A.; Bobbitt, K. L. J . Am. Chem. Soc. 1993,

115, 7517-7518. (b) Nishide, K.; Shigeta, Y.; Obata, K.; Node, M. J .

Am. Chem. Soc. 1996, 118, 13103-13104. (c) Tamai, Y.; Koike, S.;

Ogura, A.; Miyano, S. J . Chem. Soc., Chem. Commun. 1991, 799-800.

7966 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

lective addition to the ketone (eq 3), and proposed a

chelation-control model, 10, that takes advantage of the

conformational rigidity imparted by the oxazoline ring.

Molander and Bobbitt^10aachieved excellent 1,7 asym-

metric induction in the reduction of chiral keto boronates

in the preparation of enantiomerically enriched (93% ee)

secondary alcohols bearing alkyl substituents that have

little steric or electronic differentiation (eq 4) and sug-

gested a six-membered-ring transition state, 14, involving

chelation of the ketone carbonyl to the boron center.

acyclic 1,n-hydroxy ketones since our discovery of high

1,5 and 1,6 asymmetric induction with anti selectivity

(e.g., Scheme 1).⁶We suggested that these transforma-

tions might occur via a chelation-control mechanism

entailing formation of a bicyclic metal chelate. The

reacting substrate here contains no conformational con-

straint, such as imparted by a cyclic acetal, oxazoline, or

cyclic boronate, to assist in the transfer of asymmetry;

i.e., the substrate is strictly acyclic. Our exciting pre-

liminary findings of high asymmetric induction by a

single remote stereogenic center in these strictly acyclic

substrates have sparked further studies to define the

scope and limitations of the process, in terms of substrate

structure, reducing agents, solvents, and reaction tem-

perature. In this paper, we disclose the results of this

research along with a mechanistic discussion of transi-

tion-state metal chelate models to help rationalize the

observed stereochemical outcomes.

Resu lts a n d Discu ssion

Our entry into the field of remote acyclic stereocontrol

was instigated by a project on analogues of the cardio-

vascular drug nebivolol (15).¹¹There are a total of 10

possible stereoisomers for this structure (4 dl and 2 meso

forms), and the investigational drug substance is a

racemic mixture of the (S,R,R,R) and (R,S,S,S) enanti-

omers, with the latter exhibiting potent, selective â₁-

adrenergic blocking activity.¹¹In the search for nebivolol

analogues, we wanted to limit the number of possible

stereoisomers by eradicating the stereogenic centers in

the chromane units. Therefore, we set out to synthesize

bis-benzofuran analogue 16 (Scheme 1), which is merely

comprised of one dl pair and one meso form. As a

straightforward synthetic approach, we hoped to identify

suitable conditions for controlling the stereochemistry of

the reduction of 17¹²to obtain the desired dl form of 18.

In 1988, when our work commenced, there was little

precedent in the literature to guide us in successfully

effecting such an acyclic 1,5 reduction. In fact, examples

of high asymmetric induction by a singular, remote

stereogenic center bearing a hydroxyl group in nucleo-

philic addition to acyclic hydroxy ketones are un-

common,^2m-p,sparticularly with hydride as the nucleo-

phile.^1,2s,13In a report by Maier et al.,¹⁴the reduction of

δ-hydroxy ketone 19 with LiAlH₄in Et₂O provided the

corresponding 1,5 diols with a anti/syn ratio of just 55:

45. However, we thought at the outset that the diaste-

reoselectivity for the reduction of 17 with LiAlH₄might

be better because of the amine nitrogen atom in the

chain, which could conceivably cooperate with the hy-

droxyl to direct hydride delivery.

Red u ction s of 1,5-Hyd r oxy Keton es. Reduction of

17¹²with LiAlH₄in ether resulted in a negligible degree

of asymmetric induction (dl/meso 18 ) 1:1), and the

bulkier LiAlH(O-t-Bu)₃also showed no selectivity (Table

1, entries 7 and 8). The dl and meso products were

isolated as a mixture, and the anti/syn ratios were

1

determined by H NMR in CDCl₃(at 300 or 400 MHz)

We have been interested in the addition of hydride to

on account of the two diastereotopic benzyl protons of the

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7967

Ta ble 1. Red u ction of 17 to d l-(a n ti)- a n d m eso-(syn )-18

entry

reagent

solvent

dl/meso

1^a

2^c

3^c

4

5

6

7

8

9

10

11

12

13

14

15

16^j

Pd(OH)₂, H₂

NaBH₄

MeOH

THF

CH₂Cl₂

THF

Et₂O

THF

CH₂Cl₂

THF

Et₂O

THF

1:1^b

1:1^d

1:1^f

1:1^d

1:1^f

1:1^d

1:1^f

2:1^d

2.5:1^d

7:1^d

NaBH₄/CeCl₃

NaBH(OAc)₃

Red-Al^e

K-selectride^g

LiAlH₄

LiAlH(O-t-Bu)₃

(i-Bu)₂AlH^h

BH₃‚THF

BH₃‚pyridine

BH₃‚Me₂S

22ⁱ

21ⁱ

Zn(BH₄)₂

20ⁱ

Our results with 17 suggested the intermediacy of a

metal chelate with external hydride delivery, as discussed

by Baker et al.¹⁸for the 1,3 asymmetric reduction of a

γ-hydroxy ketone with LiEt₃BH. Thus, we reasoned that

the stereoselectivity might be enhanced by changing

conditions to favor a more rigid chelate species, for

example, by using a less coordinating solvent and by

lowering the reaction temperature. In follow-up studies

we used 23,¹⁹a simpler, more readily available δ-hydroxy

ketone, as the substrate (Scheme 2; Table 2). The

reduced products were isolated as a mixture, and the

a

b

At 23 °C for 2 h. N-debenzylated products, dl- and meso-16,

were obtained; the ratio was determined by ¹³C NMR. ^cAt 5 °C

for 2 h. The ratio was determined by ¹H NMR. ^e65 wt % in

d

toluene. ^fRatio was estimated by TLC. 1.0 M in THF. 1.0 M

g

h

i

j

in toluene. 0.5 M in THF. Started at -98 °C, then slowly

warmed to 23 °C over 18 h (adapted from ref 6a.).

anti isomer, which fortuitously appeared as a pair of

doublets (so-called “AB quartet”) at δ 3.75 and 3.92 (J )

13.7 Hz), while the two enantiotopic benzyl protons in

the syn isomer appeared as a sharp singlet at δ 3.79.

In an attempt to achieve an excess of anti over syn

diols, we explored 14 additional reducing agents (Table

1). The reactions were conducted at -78 °C, then slowly

warmed to 23 °C over 18 h (unless otherwise noted). Most

of the reactions delivered virtually no diastereoselection

(Table 1). However, with R-alpine-hydride¹⁵(20; entry

16) we obtained a useful level of anti diastereoselectivity

(dl/meso ) 7:1) and were able to isolate analytically pure

dl-18 in 34% yield. With the alpine-hydride reducing

agent, there is an issue of enantioselectivity in the dl diol

because of the chiral reagent’s high enantiomeric enrich-

ment. Interestingly, however, this dl product had virtu-

ally no enantiomeric enrichment, as assessed by 400-MHz

¹H NMR with (S)-(+)-2,2,2-trifluoromethyl-1-(9-anthryl)-

ethanol (in CDCl₃).¹⁶With Zn(BH₄)₂in ether (entry 15)

and lithium thexyl-(R)-limonylborohydride (21) in THF

(entry 14), modest 1,5 asymmetric induction was ob-

served, but none of the reducing agents produced an

excess of meso (syn) over dl (anti) diols. NB-Enantride

(22), which is structurally related to S-alpine-hydride (the

enantiomer of 20) and reported to be more effective for

the asymmetric reduction of ketones,¹⁷was surprisingly

ineffective (entry 13).

1

anti/syn ratios were determined by H NMR in CDCl₃.

The diastereotopic benzyl protons of the anti isomer again

appeared as a pair of doublets (δ 3.70 and 3.99, J ) 13.5

Hz), while the two protons in the syn isomer appeared

as a singlet (δ 3.83).

The following observations from a series of experiments

with 23 are noteworthy (Table 2). (1) When 23 was

reduced with R-alpine-hydride under the conditions

employed for 17, similar diastereoselectivity was achieved

(entry 1). (2) The 1,5 asymmetric induction is indepen-

dent of reagent chirality, as the R and S forms of 20

furnished the same ratio of dl/meso 24 (cf. entries 1 and

2). (3) The anti/syn selectivity increased with decreasing

temperature (entries 3-5). (4) The structure of the

hydride anion plays an important role, with a bulkier

hydride anion leading to a higher ratio of dl/meso 24

(entries 7-9; also see Table 1). (5) Replacement of THF

as the solvent by noncoordinating CH₂Cl₂(the solvent

composition with the THF from the reducing agent was

ca. 90% CH₂Cl₂) significantly enhances 1,5 selectivity (cf.

entries 3 and 7, 11 and 12).

The best selectivity of anti vs syn diols in the reduction

of 23 was realized with Zn(BH₄)₂in CH₂Cl₂at -78 °C

(anti/syn ) 13:1, entry 12; the solvent composition with

the Et₂O from the reducing agent was ca. 85% CH₂Cl₂).

For most of the entries in Table 2, only modest yields

were realized along with recovered starting material, and

the use of excess reducing agent and/or an extended

reaction time at -78 °C did not significantly increase the

chemical conversion. This problem might be connected

with the high proportion of cyclic hemiketal form, relative

to the reducible acyclic keto form, in this substrate (23).

We investigated the issue of enantioselectivity in the

dl diol from 23. Reduction of 23 with R-alpine-hydride

(11) (a) Van Lommen, G.; De Bruyn, M.; Schroven, M. J . Pharm.

Belg. 1990, 45, 355. (b) Van Neuten, L.; Dupont, A. G.; Vertommen,

C.; Goyvaerts, H.; Robertson, J . J . Hum. Hypertens. 1997, 11, 139-

144. (c) Van Bortel, L. M. A. B.; De Hoon, J . N. J . M.; Kool, M. J . F.;

Wijnen, J . A. G.; Vertommen, C. I. M.; Van Nueten, L. G. M. Eur. J .

Clin. Pharmacol. 1997, 51, 379-384. (d) Stoleru, L.; Wijns, W.; van

Eyll, C.; Bouvy, T.; Van Nueten, L.; Pouleur, H. J . Cardiovasc.

Pharmacol. 1993, 22, 183-190. (e) Also, see: Anon. Drugs Future 1991,

16, 964; 1995, 20, 1070; 1998, 22, 1169-1170. (f) Nebivolol hydrochlo-

ride is marketed in Germany by Menarini Company for the treatment

of hypertension under the trade name Nebilet.

(12) We prepared 17 from 2-acetylbenzofuran in four steps using

modified published procedures; see Experimental Section.

(13) Greeves, N. In Comprehensive Organic Synthesis; Trost, B. M.;

Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 8, pp 1-24.

(14) Maier, G.; Roth, G.; Schmitt, R. K. Chem. Ber. 1985, 118, 704-

721.

(17) Midland, M. M.; Kazubski, A. J . Org. Chem. 1982, 47, 2495-

2496.

(15) Krishnamurthy, S.; Vogel, F.; Brown, H. C. J . Org. Chem. 1977,

42, 2534-2536.

(16) Pirkle, W. H.; Hoover, D. J . Top. Stereochem. 1982, 13, 263-

331.

(18) Baker, R.; Cottrell, I. F.; Ravenscroft, P. D.; Swain, C. J . J .

Chem. Soc., Perkin Trans. 1 1985, 2463-2468.

(19) Prepared by a published procedure: Brown, C. L.; Lutz, R. E.

J . Org. Chem. 1952, 17, 1187-1193.

7968 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

Sch em e 2

the peak intensities of N-Me at δ 2.41 (syn) and 2.45

(anti)], with the isomer assignment being based on ¹³C

NMR data in analogy with 24. The loss of high diaste-

reocontrol in the reduction of 31 to 32 clearly indicates

that the steric properties of the N substituent can play

an important role in achieving high 1,5 diastereoselec-

tivity, with the bulkier benzyl group giving much better

anti selectivity.

To rationalize our results for 1,5 diastereoselectivity,

we initially proposed a working model involving a 5,5-

bicyclic chelate structure, such as depicted in 33, in which

the lithium or zinc ion is complexed with the hydroxyl,

amine, and ketone groups and endo attack on the bicyclic

array by the hydride species leads to the anti 1,5-diols

preferentially.^6aThis can account for the effects of

solvent, Zn(II), nitrogen replacement, and NB-enantride

(22). However, since this model does not fit well with

our new results for the reduction of 31, we are inclined

to offer another viewpoint, which is presented later in

some detail (vide infra).

at -98 °C produced dl-24 with no measurable optical

rotation (589 nm). More importantly, there was no

enantiomeric enrichment as determined by 400-MHz ¹H

NMR with (S)-(+)-2,2,2-trifluoromethyl-1-(9-anthryl)-

ethanol,¹⁶in agreement with the aforementioned result

with 17. As a control, we reduced PhC(O)CH₂N(Bn)₂

with R-alpine-hydride at -98 °C to the corresponding

alcohol and found that it had a mere 10% enantiomeric

1

excess (ee) by the chiral H NMR assay.

To determine the impact of the phenyl groups in 23

on the diastereoselectivity, dimethyl 1,5-hydroxy ketone

25 was prepared from commercially available racemic

R-(methylaminomethyl)benzyl alcohol and chloroacetone

(Scheme 3). Reduction of 25 with R-alpine-hydride or

with LiBH₄at -78 °C in CH₂Cl₂went very slowly.

Increasing the reaction temperature to 0 °C gave diols

26 in 84% and 78% isolated yield with anti/syn ratios of

7.8:1 and 4:1, respectively, similar to the ratios obtained

with the corresponding diphenyl system, 23 (Table 2).

The influence of the amine nitrogen (actually the

N-Bn unit) in 23 was then investigated. We replaced

the N-benzyl moiety with carbon or sulfur groups (Scheme

4). Thus, δ-hydroxy ketone 27²⁰was reduced in THF at

-78 °C with R-alpine-hydride or Zn(BH₄)₂to give an

isomeric mixture of 1,5-diols 28 in a ratio of 1.2:1, as

determined by 75-MHz ¹³C NMR (from the intensities of

three pairs of carbon signals: δ 22.04/22.11, 38.71/38.63,

and 74.22/74.07) and 300-MHz ¹H NMR (from integrated

areas for the OH peaks at δ 2.70 and 2.85). Sulfur-

containing δ-hydroxy ketone 29 was prepared from

partial reduction of the commercially available diketone

with NaBH₄. Reduction of 29 to 30²¹with R-alpine-

hydride in THF at -78 °C also gave poor diastereoselec-

tivity with an anti/syn ratio of 1.3:1 as estimated by ¹³C

NMR [from the peak intensities for δ 41.7 (anti CH₂),

42.2 (syn CH₂); 72.4 (anti CH), 73.2 (syn CH)]. This

indicates that divalent sulfur is not effective as a ste-

reodirecting entity in this situation. An oxygen-contain-

ing δ-hydroxy ketone was also prepared²²and subjected

to reduction, but no reaction was observed with R-alpine-

hydride or Zn(BH₄)₂from -78 to 23 °C, probably because

of the high proportion of the cyclic hemiketal form in this

substance (>95% by ¹³C NMR). The nearly 1:1 anti/syn

ratio for the reductions of 27 and 29 certainly underscores

the importance of the amine nitrogen (N-substituted) in

17 and 23 for achieving the high 1,5 asymmetric induc-

tion. We then tested the stereochemical effect of the N

substituent on 1,5 diastereoselectivity. N-methyl δ-hy-

droxy ketone 31, prepared in 69% yield from 2-(methyl-

amino)-1-phenylethanol and phenacyl bromide in the

presence of N,N-diisopropylethylamine, was reduced with

R-alpine-hydride in CH₂Cl₂at -78 °C to afford 32 with

An attempt was then made to achieve an excess of syn

over anti diols possibly via internal hydride delivery.

Thus, δ-hydroxy ketone 23 was reduced with tetra-

methylammonium triacetoxyborohydride, a well-known

stereoselective reducing agent for â-hydroxy ketones by

hydroxyl-directed intramolecular hydride delivery,²³at

23 °C in MeCN-HOAc (1:1) to afford diols 24 in 75%

yield with an anti/syn ratio of 1:1.5 (Table 2; entry 13).

Although the syn vs anti selectivity is low, this is a rare

example in which we obtained more syn than anti diol,

possibly on account of hydroxyl-directed hydride deliv-

ery.

Red u ction s of 1,6- a n d 1,7-Hyd r oxy Keton es. The

unusually high 1,5 asymmetric induction obtained with

17 and 23 encouraged us to prospect for high 1,6 and 1,7

acyclic diastereoselection in related reductions of hydroxy

ketones, presumably via 5,6-, 6,6-, and 5,7-bicyclic che-

lation control. 1,6-Hydroxy ketone 34, readily prepared

from 2-(benzylamino)-1-phenylethanol and 3-chloropro-

piophenone in the presence of N,N-diisopropylethyl-

amine, was reduced to 1,6-diols 35 by using various

reducing reagents (Scheme 5; Table 3). In general, the

reductions of 34 proceeded much more rapidly in CH₂-

Cl₂at -78 °C and in higher yields than did the reductions

of the corresponding 1,5-hydroxy ketone, 23. The best

selectivity for anti vs syn diols was obtained in the

reduction of 34 with R-alpine-hydride (anti/syn ) 12:1,

entry 1). Reduction with Zn(BH₄)₂or with lithium thexyl-

1

an anti/syn ratio of 2.3:1 as estimated by H NMR [from

(20) Huang, R. L.; Williams, P. J . J . Chem. Soc. 1958, 2637-2640.

(21) Valle, G.; Buso, M.; De Lucchi, O. Z. Kristallogr. 1987, 181,

95-98.

(22) Prepared from R-diazoacetophenone and 1-phenyl-1,2-ethanediol

in the presence of boron trifluoride etherate (Mayfield, R. J .; Yates, P.

Org. Prep. Proced. Int. 1971, 3, 201-204). The isolated product existed

almost entirely in the cyclic hemiketal form (based on ¹³C NMR data)

with a ca. 1:1 ratio of diastereomers (based on ¹H NMR data).

(23) Evans, D. A.; Chapman, K. T.; Carreira, E. M. J . Am. Chem.

Soc. 1988, 110, 3560-3578.

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7969

Ta ble 2. Red u ction of 23 to d l-(a n ti)- a n d m eso-(syn )-24^a

entry

reagent

temp (°C)

time (h)

solvent

THF

dl/meso

yield (%)

1

2

3

4

5

6

7

8

9

10

11

12

13

R-alpine-hydride

S-alpine-hydride

R-alpine-hydride

LiEt₃BH^c

LiBH₄

Zn(BH₄)₂

-98 f 0

-78

-40

0

-40

-78

0

16

5.5

5

2

5

26

2

5

26

2

6:1

4:1

3:1

6:1

10:1

8:1

5:1

2:1

4:1

13:1

1:1.5

57

62

17

37

62

40

55

THF

b

CH₂Cl₂

THF^d

CH₂Cl₂

-40

-78

25

d

47

75

Me₄NB(OAc)₃H

MeCN-HOAc

a

b

Ratios were determined by ¹H NMR; yields are for mixtures isolated from preparative TLC. Reaction contains 10-15% THF from

the reducing agent. ^c1.0 M in THF. Reaction contains 10-15% Et₂O from the reducing agent.

d

Ta ble 3. Red u ction of 34 to a n ti a n d syn d iols 35^a

Sch em e 3

entry

reagent

equiv time (h) anti/syn^byield (%)

1

2

3

4

5

6

7

8

9

R-alpine-hydride^c

Zn(BH₄)₂

2.1

2.5

2.1

1.0

3

18

3

1

3

23

12:1^d

7.5:1^d

2:1^d

83

77

74

no rxn

77

69

e

LiBH₄

21^c,f

8:1

22^c,f

L-selectride^c,g

K-selectride^c,g

R-alpine-hydride^c

2:1

1:1

3.2:1

5.9:1

79

e

Zn(BH₄)₂

37^h

a

b

Sch em e 4

Reactions were conducted at -78 °C in CH₂Cl₂. Ratio

determined by HPLC except where noted. ^cReaction contains 10-

15% THF from the reducing agent. Ratio determined by ¹H NMR.

d

^eReaction contains 10-15% Et₂O from the reducing agent. ^f0.5

g

h

M in THF. 1.0 M in THF. Reaction did not go to completion.

protons (“AB quartets”) centered at δ 3.74 (anti) and δ

3.72 (syn), with no overlap of the signals because of a

large chemical shift difference between the pair of

doublets for the anti isomer [∆δ(anti) ) ca. 0.40 ppm]

and a small difference between the pair of doublets for

the syn isomer [∆δ(syn) ) ca. 0.11 ppm]. Since unam-

biguous assignment of the benzyl resonances required the

synthesis of an authentic sample of syn- or anti-35,

commercially available (R)-(+)-3-chloro-1-phenyl-1-pro-

panol was reacted with benzylamine in the presence of

KI and K₂CO₃to give N-benzylamino alcohol 36 (Scheme

6). Attempts to alkylate 36 with (R)-2-chloro-1-phenyl-

ethanol resulted in very poor yields (<10%) of the desired

1,6-diols due to decomposition of the chlorohydrin under

the reaction conditions (method A in Scheme 6). A neat

reaction between 36 and (R)-styrene oxide at 120 °C

afforded in 42% isolated yield of the anti diols 35 (R,R)

Sch em e 5

(R)-limonylborohydride (21) also gave significant selectiv-

ity (anti/syn ) 7.5:1 and 8:1, respectively, entries 2 and

4). It is noteworthy that no reduction was observed with

NB-enantride (22) under the reaction conditions (entry

5). Reductions with LiBH₄and bulky L-selectride gave

the same modest selectivity (anti/syn ) 2:1, entries 3 and

6), while K-selectride gave no selectivity at all (entry 7).

Attempts to carry out the reaction with one mol equiv of

R-alpine-hydride gave a complete reaction with signifi-

cantly reduced selectivity (anti/syn ) 3.2:1, entry 8),

while reduction with one mol equiv of Zn(BH₄)₂gave an

incomplete reaction accompanied by diminished selectiv-

ity (anti/syn ) 5.9:1, entry 9). In any event, such high

diastereoselectivity for a 1,6 reduction in a strictly acyclic

substrate, anti/syn ) 12:1 (entry 1), is quite remarkable!

The diastereomeric ratios for 35 were nicely quanti-

1

(method B, Scheme 6), the H NMR spectrum (CDCl₃) of

which was identical to that of the major product obtained

from the reduction of 34, confirming the assignment. A

better yield for the preparation of authentic 35 was

obtained by using method C (Scheme 6). Thus, (R)-2-

chloro-1-phenylethanol was reacted with benzylamine at

180 °C to give 37 in 80% yield, which was alkylated with

(S)-3-chloro-1-phenyl-1-propanol to give the syn diols 35

(R,S) in 89% isolated yield. The ¹H NMR spectrum

(CDCl₃) for 35 (R,S) was identical to that of the minor

product obtained from the reduction of 34.

A “reversed” 1,6-hydroxy ketone 38 (R enantiomer

only) was prepared from 36 and 2-bromoacetophenone

and reduced with R-alpine-hydride or Zn(BH₄)₂in CH₂-

Cl₂at -78 °C to give 1,6-diols 35 in 75% and 72% yield,

respectively (Scheme 7). The anti/syn ratios of 5:1 and

3:1, respectively, for this 1,6 acyclic system are signifi-

1

tated by HPLC and 400-MHz H NMR. There were two

different pairs of doublets for the aliphatic benzylic

7970 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

Sch em e 6

Sch em e 9

Sch em e 10

Sch em e 7

Sch em e 8

subunit was then prepared (Scheme 9) and reduced with

R-alpine-hydride in CH₂Cl₂at -78 °C to give 1,7-diols

44 in 88% yield. No asymmetric induction was observed

here, with the anti/syn ratio being 1:1.

1,7-Hydroxy ketone 46 was thought to be a more

promising substrate because of its â-hydroxy amino

system. Since an attempt to prepare 46 via N-alkylation

of 2-(benzylamino)-1-phenylethanol with 4-chlorobuty-

rophenone in the presence of N,N-diisopropylethylamine

or KI/K₂CO₃failed, the ketone was protected with eth-

ylene glycol to afford 45 (Scheme 10). N-Alkylation of

2-(benzylamino)-1-phenyl-ethanol with 45 in the presence

of KI/K₂CO₃was followed by cleavage of the 1,3-dioxolane

to give 46. Reduction of 46 in CH₂Cl₂at -78 °C with

R-alpine-hydride or Zn(BH₄)₂afforded 1,7-diols 44 in 84%

and 63% yield, with anti/syn ratios of 3:1 and 1:2,

respectively. The outcome with Zn(BH₄)₂represents an

unusual case favoring the syn isomer, but the degree of

1,7 stereocontrol in either direction is not particularly

impressive. The isomer assignment for 44 is based on

¹H NMR data, in analogy with 35. Thus, mixture 44

displayed two different pairs of doublets for the aliphatic

benzylic protons (“AB quartets”) centered at δ 3.72 (anti)

and δ 3.70 (syn), with no overlap of the signals because

of a large chemical shift difference between the pair of

doublets for the anti isomer [∆δ(anti) ) ca. 0.45 ppm]

and a small difference between the pair of doublets for

the syn isomer [∆δ(syn) ) ca. 0.30 ppm].

cant, albeit not remarkable. As far as 1,6 acyclic stereo-

control is concerned, a 5:1 ratio can be viewed as very

meaningful. By comparison of this result with the

reduction of 34, it seems that the degree of 1,6 diaste-

reoselectivity is dependent on the distance between the

hydroxyl and amine groups, with the â-hydroxy amine

arrangement being preferred over the γ-hydroxy amine

arrangement.

Given the favorable results with the 1,6 system, we

decided to pursue the related 1,7-hydroxy ketone cases.

Hydroxy ketone 39 (R enantiomer) was prepared from

36 and reduced with R-alpine-hydride in CH₂Cl₂at -78

°C to give 1,7-diols 40 in 80% yield (Scheme 8). In this

instance, the anti/syn (dl/meso) ratio was a mere 1.2:1.

1

The dl and meso products were assigned by H NMR, as

the two diastereotopic benzyl protons of the anti (dl)

isomer appeared as a pair of doublets at δ 3.44 and 3.79

(J ) 13.7 Hz) and the two enantiotopic benzyl protons

in the syn (meso) isomer appeared as a singlet at δ 3.65.

1,7-Hydroxy ketone 43 containing a δ-hydroxy amine

Despite our modest results with the 1,7-hydroxy ke-

tones, we can take solace from the impressive results

with a 1,6-hydroxy ketone. Indeed, we have a very

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7971

Sch em e 11

Sch em e 13

Sch em e 14

Sch em e 12

In the 1,5 system, replacement of the amine nitrogen

(actually the N-Bn group) in 1,5-hydroxy ketone 23 with

carbon or sulfur groups resulted in poor diastereoselec-

tivity (Scheme 4). Unfortunately, the corresponding

oxygen-containing 1,5-hydroxy ketone²²did not reduce,

probably because of the high proportion of the cyclic

hemiketal form (six-membered ring) in the substrate

(confirmed by ¹³C NMR). Thus, the oxygen-containing

1,6-hydroxy ketone 52 was synthesized from 1-phenyl-

1,2-ethanediol and 3-chloropropiophenone (Scheme 13),

and ¹³C NMR showed that it did not exist in the cyclic

hemiketal form. Reduction of 52 with R-alpine-hydride

proceeded to completion in 18 h at -78 °C to furnish a

1:1 mixture of syn and anti diols 53, clearly demonstrat-

ing an abolition of anti stereoselectivity on replacement

of N-Bn with O. We sought to evaluate the importance

of the nitrogen substituent on the 1,6 diastereoselectivity.

The benzyl group was removed from 34 to give N-H

analogue 54a , which was reduced with R-alpine-hydride

in CH₂Cl₂at -78 °C to yield 1,6-diols 55a with just a 2:1

anti/syn ratio, as estimated by ¹³C NMR (Scheme 14). The

benzyl group in 34 was then replaced with a small methyl

group, as in 54b. Reduction of 54b with R-alpine-hydride

in CH₂Cl₂at -78 °C yielded 1,6-diols 55b with a

relatively modest anti/syn ratio of 3:1 (estimated by ¹³C

NMR). The isomer assignments for 55a and 55b were

based on ¹³C NMR data in analogy with 35. High 1,6

diastereoselectivity was also lost when the benzyl group

in 34 was replaced by an electron-withdrawing acyl group

(Scheme 15). Intermediate 54a was N-acylated with

benzoyl chloride in the presence of pyridine to give 56,

reduction of which with R-alpine-hydride at -78 °C led

interesting example of high 1,6 asymmetric induction in

an acyclic system, involving hydride reduction of a

prochiral ketone (sp²center) six atoms removed from an

existing stereocenter bearing a hydroxyl group (34 f 35,

Scheme 5). The hydroxyl and intrachain nitrogen in 34

appear to participate in a metal chelate that serves to

direct hydride addition. To test for the importance of a

free hydroxyl, we prepared methoxy amino ketone 50,

via Boc derivative 47, and subjected it to reduction with

R-alpine-hydride at -78 °C in CH₂Cl₂(Scheme 11).

Methoxy amino alcohols 51 were isolated in 80% yield

as a 4:1 mixture of the anti and syn forms, and the isomer

assignment was established by the synthesis of an

authentic sample of anti-51 (Scheme 12). Clearly, me-

thylation of the hydroxyl group in 34 caused a partial

loss of stereoselectivity (anti/syn ) 12:1 for the parent

reaction, 34 f 35). If one considers a 5,6 bicyclic chelate

model as being crucial to the 1,6 diastereoselectivity, then

the methoxy group must be reasonably effective as a

ligand for lithium in this reaction, but not as effective

as a hydroxyl.

1

to a mixture of diols 57. Since the H NMR spectrum of

We then proceeded to assess the influence of the

intrachain nitrogen in 34 on the 1,6 diastereoselectivity.

57 showed broad peaks, probably because of the presence

of slowly interconverting rotamers from the amide

7972 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

Sch em e 15

another viewpoint. Although there was little change in

anti stereoselectivity on replacing the phenyl groups in

23 with methyl groups (viz. 25), replacing the benzyl

group on the nitrogen in 23 with a methyl group (viz.

31) resulted in a large loss in 1,5 diastereoselectivity. We

suggest that a five-membered-ring complex forms with

the ketone not involved in chelation and that reduction

occurs via internal hydride delivery,²³as shown in model

58. Here, the steric interaction between the N-Bn and

ketone carbonyl groups would force the ketone to adopt

a suitable orientation to provide the anti diol. The

difference between 23 and 31 could be attributed to the

N-methyl group having much less influence on the

orientation of the ketone than the larger N-Bn group.

Lower temperatures may also enhance the control of the

orientation of the ketone carbonyl, as noted by the effect

of temperature on the stereoselectivity of the reaction of

23. The results with amine replacement, N-Bn f CH₂

(viz. 27) and N-Bn f S (viz. 29), are consistent with this

scenario. The good donor solvent THF may not favor the

internal hydride delivery as much as the noncoordinating

CH₂Cl₂because of THF complexation to the metal center

and disruption of the anti addition mechanism.

group, the ratio of the two diastereomers was difficult to

determine. Thus, the bis-silyl ether of 57 was prepared,

the amide group was reduced to the corresponding amine

with borane-THF, and the silyl groups were removed

to give a mixture of diols 35, which had an anti/syn ratio

of only 1.5:1, as measured by ¹H NMR (Scheme 15). The

results for replacement of the N-benzyl structural subunit

in these 1,6-hydroxy ketones clearly indicate that the

stereoelectronic character of this subunit plays a key role

in the attainment of high 1,6 asymmetric induction. To

improve the anti stereoselectivity, we reasoned that a

bulkier benzyl group on the nitrogen might be advanta-

geous. Consequently, 54a was N-alkylated with 2,4,6-

trimethylbenzyl chloride in the presence of N,N-diiso-

propylethylamine in THF-DMF at 60 °C to afford 54c,

which was reduced with R-alpine-hydride at -78 °C in

CH₂Cl₂to produce 1,6-diols 55c in 56% yield with an

exceptional anti/syn ratio of 22:1 (determined by 600-

1

MHz H NMR in analogy with 35)!

Mech a n istic Con sid er a tion s. To rationalize our

results for the 1,5-anti diastereoselectivity, we originally

proposed a 5,5-bicyclic chelate structure, such as model

33, in which the lithium or zinc ion is complexed with

the hydroxyl, amine, and ketone groups in 17 or 23 to

achieve high asymmetric induction.^6aIn the conforma-

tionally rigid array, endo attack by the hydride species

would be sterically unfavorable, so that attack is pre-

ferred from the less hindered exo side, leading to the anti

1,5-diols preferentially. When the reaction is performed

in THF, this good donor solvent presumably competes

with the heteroatom groups in 17 or 23 for complexation

of the metal center and thereby counteracts chelate

formation. On the other hand, the relatively noncoordi-

nating CH₂Cl₂and low reaction temperature enhance

chelate participation. With this external hydride delivery

model, a bulkier hydride anion should give a higher ratio

of anti 1,5-diols, as generally observed. The absence of

good anti stereoselectivity with NB-enantride (22) may

be associated with the presence of a coordinating ether

group in the reagent itself, which could inhibit formation

of the necessary chelate complex or contribute to a

reorganization of the reactive complex. The results with

nitrogen atom replacement are consistent with this

scenario in that N-Bn f CH₂(27) and N-Bn f S (29)

gave very little selectivity.

To account for the outstanding 1,6 acyclic diastereo-

control, we originally proposed a 5,6-bicyclic chelate

structure, such as in model 59.^6bThe lithium or zinc ion

would be complexed with the hydroxyl, amine, and ketone

groups in 34 to establish a conformationally rigid array,

and the hydride species would attack the carbonyl in the

six-membered ring from an axial direction to allow for

an equatorial phenyl ring to develop in the transition

state. This leads to the major anti 1,6-diol product. A

model with a trans-like 5,6 ring fusion, where the

nitrogen configuration is inverted, could also be enter-

tained. However, this overly complicates something that

is already rather speculative. In model 59, anti selectiv-

ity would be favored by a bulkier nitrogen substituent,

as we observed in the trend of stereoselectivity for the

reduction of 34 and 54a -c. Significantly, the anti/syn

ratio improved as the substituent varied from hydrogen

to methyl to benzyl to 2,4,6-trimethylbenzyl. Replace-

ment of the N-benzyl group with an oxygen group (viz.

52) abolished selectivity. When the N-Bn in 34 was

replaced by NC(O)Ph (viz. 56), a group with a similar

steric size but with an electron-deficient nitrogen and an

altered coordination site, little selectivity was obtained.

Presumably, in the case of substrates 52 and 56, there

is an absence of coordination by the heteroatom center

in the chain such that the putative bicyclic chelate cannot

form. Although the methoxy group is also a reasonably

However, given the additional data with the reduction

of the N-methyl analogue 31, we are compelled to offer

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7973

good ligand for lithium, it was not as effective for anti

stereocontrol as the hydroxyl (cf. reductions of 34 and

50).

NMR Mech a n istic Stu d ies. We undertook NMR

studies (400 MHz) in an attempt to identify the formation

of well-defined metal chelates with representative hy-

droxy amino ketone substrates. Treatment of 1,5-hy-

droxy ketone 23 with up to 2 mol equiv of LiBF₄, in

increments of 0.5 mol equiv, in THF-d₈at 23 °C showed

significant shifting of proton resonances and a change

in the splitting patterns of the methylene protons R to

nitrogen. However, ¹³C NMR spectra suggested that the

substrate remained in the cyclic ketal form (quaternary

carbon at 95 ppm). Treatment of 1,6-hydroxy ketone 34

with up to 2.0 mol equiv of LiBF₄, in increments of 0.5

mol equiv, in THF-d₈at 23 °C showed minor shifting of

the proton resonances and the eventual appearance of

olefinic signals, suggesting decomposition of the sub-

strate. The 1,6-hydroxy ketone system, 34, was treated

with 1-2 mol equiv of LiBF₄in THF-d₈at 0, -53, and

-83 °C. At -83 °C, a significant change in the splitting

pattern was noted for the methylene protons between the

amine and the hydroxyl carbon, from a single multiplet

to two multiplets. While some of these NMR results

suggest some type of chelation to Li and/or minor

conformational changes, there was no strong evidence for

the formation of bicyclic metal chelates or, for that

matter, stable, highly organized metal complexes.

Some NMR experiments (400 MHz, 2D-COSY) were

performed with 34 and Ti(i-OPr)₄, which we thought

might generate a more stable, characterizable chelate.

With 1 mol equiv of the titanium reagent at -80 °C in

CH₂Cl₂, we observed two new species, possibly stereo-

isomers of a titanium complex, with the following changes

in chemical shifts (34 f 34 + Ti): δ 2.40/2.52 (pair of br

m, CHOHCH₂N) f 2.60-2.65 (m) and 3.22/3.78 (dd/dd);

2.70/3.25 (pair of m, CH₂CH₂N) f 3.15-3.40 (br m); 3.1-

3.4 (br m, CH₂CdO) f 3.15-3.40 (br m); 3.38/3.93 (pair

of d, PhCH₂N) f 3.85/4.10 (pair of d); 4.30 (br s, OH);

4.73 (br s, CHOH) f 5.37/5.50 (pair of d); 7.90 (br s,

o-benzoyl) f 7.68 (m). The greatest change in chemical

shift occurred for the methine at the hydroxyl center (∆δ

of 0.64 and 0.77), which is reasonable for titanium ester

formation. Also, the methylenes next to nitrogen were

shifted downfield to some degree, suggesting coordination

of the nitrogen to titanium. However, the methylene next

to the keto group did not experience a meaningful shift,

suggesting that the carbonyl group is not strongly

coordinated to the metal. The upfield shift of the

o-benzoyl protons suggests that there is an interaction

of the keto group with titanium. The NMR interpretation

is complicated by the presence of two major species,

clearly indicated by the multiple resonances for the

methine; thus, we are not able to interpret these data in

terms of precise structures for the titanium complexes.

The reduction of 1,6-hydroxy ketones was complete

within a couple of hours at -78 °C, while the reduction

of 1,5-hydroxy ketones was not complete even after 24 h

at -78 °C. This difference in reactivity could arise from

the reduction occurring via different pathways. Thus,

reduction of the 1,6-hydroxy ketones could proceed via a

5,6-bicyclic intermediate with the ketone carbonyl acti-

vated toward reduction because of metal complexation,

whereas reduction of the 1,5-hydroxy ketones could

proceed via an internal delivery of hydride with the

ketone carbonyl not activated by complexation in a

bicyclic intermediate.

The lack of stereocontrol for the 1,7-hydroxy ketones

39, 43, and 46 could be attributed to a modest facial

preference for hydride attack in the corresponding 6,6-

and 5,7-bicyclic chelates (see models 60 and 61) or in the

seven-membered-ring, monocyclic complex with the ke-

tone uncoordinated to the metal.

Con clu sion

High stereocontrol (ca. 10-20:1) between remote 1,5

and 1,6 stereocenters was achieved in the reduction of

certain hydroxy amino ketones, namely 17, 23, 25, 34,

and 54c, with R-alpine-hydride or Zn(BH₄)₂. The anti

diastereoselectivity was optimized in a noncoordinating

solvent, such as CH₂Cl₂, and at diminished temperatures,

such as -78 °C. The distance between the hydroxyl and

amino groups had a greater effect on the diastereoselec-

tivity than did the distance between the ketone and

amino groups, with â-hydroxy amino compounds provid-

ing the best results. The stereoelectronic properties of

7974 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

of the hydrotribromide (140 g, 0.028 mol) and 2-pyrrolidinone

(26 mL) was added. The reaction was heated at reflux for 1.5

h, cooled to 10 °C, and filtered. The filtrate was concentrated

in vacuo to a residue that was recrystallized twice from 95%

ethanol (650 mL) to yield 2-(bromoacetyl)benzofuran (257.9 g,

79%) as beige crystals, mp 70-82 °C (lit.²⁴mp 90-91 °C). A

solution of 2-(bromoacetyl)benzofuran (100.0 g, 0.42 mol) in

2-pyrrolidinone was reacted with dibenzylamine (82.5 g, 0.42

mol) according to the procedure of Burger and Deinet,²⁵and

the crude product was partially dissolved in boiling 2-propanol

and filtered. The filtrate was recrystallized from 2.5 L of

2-propanol to afford 71.3 g (48%) of free base as a tan solid,

which was dissolved 2.5 L of ethyl acetate, acidified with HCl

gas, and crystallized to furnish dibenzylaminoacetylbenzofuran

hydrochloride (59.3 g, 36%) as a white solid, mp 196 °C, dec

(lit.²⁵mp 191 °C). A solution of this product (59.0 g, 0.151

mol) in methanol (100 mL) was combined with 20% Pd(OH)₂/C

and hydrogenated on a Parr apparatus at 60 psig for 45 min.

The filtrate was concentrated in vacuo, and the residue was

recrystallized from 2-propanol to afford 2-(1-hydroxy-2-

benzylaminoethyl)benzofuran‚HCl (33.5 g, 74%) as a white

solid, mp 189-191 °C (lit.²⁵mp 187.5 °C). A solution of this

benzofuran (28.2 g, 0.93 mol) and N,N-diisopropylethylamine

(24.0 g, 0.186 mol) in THF (370 mL) was cooled to 5 °C, treated

with 2-(bromoacetyl)benzofuran (23.3 g, 0.098 mol), and slowly

warmed to 23 °C over 16 h with stirring under nitrogen. The

reddish-brown suspension was filtered, and the filtrate was

concentrated in vacuo. The residue was partitioned between

ethyl acetate (400 mL) and 3 N NaOH (100 mL), and the

organic layer was rinsed twice with brine. The organic

solution was dried (K₂CO₃) and concentrated in vacuo. The

residue was recrystallized from methanol to afford 17 (26.3 g,

63%) as tan crystals, mp 116-119 °C. An analytical sample

was purified by flash column chromatography (CHCl₃) and

recrystallized from methanol to give 17 as beige crystals, mp

125-128 °C. ¹H NMR (mixture of two cyclic hemiketals, cis/

trans ) 5:1, and the ketone form) δ 2.60-2.80 (m, 2 H), 3.10-

3.40 (m, 2.4 H), 3.65 (s, 2 H, cis ketal CH₂Ph), 3.89 (d, J )

13.5 Hz, 0.2 H, ketone CH₂Ph), 4.05-4.10 (m, 0.6 H, ketone

CH₂Ph and CH₂CdO), 4.70 (br s, 0.2 H, OH), 4.77 (br s, 1H,

OH), 4.90-5.07 (m, 0.2 H, trans ketal CH-O), 5.44 (dd, J )

11.4, 2.4 Hz, 1H, cis ketal CH-O), 6.71 (s, 0.2 H), 6.75 (s, 1

H), 6.91 (s, 1 H), 7.11-7.70 (m, 13 H). CI-MS m/z 426 (MH⁺).

IR (KBr; no carbonyl absorption because of a ring-chain

equilibrium between the ketone form and two cyclic hemi-

the intervening chain between the hydroxyl and ketone

groups of the substrate played a key role in the 1,5 (cf.

23, 27, 29, and 31) and 1,6 (cf. 34, 52, 54a -c, and 56)

asymmetric induction, with anti diastereoselectivity be-

ing strongly favored by a bulkier N-alkyl substituent.

By way of a dramatic illustration, the reduction of 54c,

possessing an N-trimethylbenzyl group, with R-alpine-

hydride at -78 °C in CH₂Cl₂furnished 55c with an

impressive anti/syn ratio of 22:1. The reduction of 50 to

51 (anti/syn ) 4:1) reflects on the importance of a free

hydroxyl for achieving high 1,6-anti stereoselectivity.

Interestingly, replacement of hydroxyl with a methoxyl

significantly diminished, but did not destroy, anti ste-

reocontrol.

One of the major challenges in synthetic organic

chemistry is achieving good control of stereochemistry

between remote sites in conformationally flexible sys-

tems. There are a limited number of examples of remote

asymmetric induction in acyclic molecules with stereo-

genic and prostereogenic centers separated by distances

of 1,5, 1,6, or greater, especially examples unperturbed

by the presence of other, pre-existing stereogenic fea-

tures. To be sure, no general reactions for the construc-

tion of remote chiral relationships in such acyclic systems

are yet known. We have discovered a new reaction type

that is capable of delivering high 1,5- and 1,6-anti

asymmetric induction in a strictly acyclic array, presum-

ably on the basis of a highly organized metal chelate. The

hydroxyl and amino groups are involved in directing

addition of the hydride reagent preferentially to one face

of the ketone carbonyl. One can reasonably consider that

a bicyclic chelation model may be responsible for the

unusually high anti stereocontrol for the 1,6 acyclic

system. However, this model appears to be a less likely

participant in the 1,5 acyclic system. Despite the absence

of a clear structural understanding of the mechanism for

high 1,5- and 1,6-anti stereocontrol in these reductions,

one can still draw consolation from the exciting stereo-

chemical results.

ketals) ν_max3404 (s), 1471 (s), 1252 (s), 1163 (m), 1042 (s) cm^-1

.

Anal. Calcd for C₂₇H₂₃NO₄: C, 76.22; H, 5.45; N, 3.29.

Found: C, 76.39; H, 5.43; N, 3.22.

Exp er im en ta l Section

Gen er a l P r oced u r es. Unless otherwise noted, materials

were obtained from commercial suppliers and used without

further purification. ¹H and ¹³C NMR spectra (300, 400, or

600 MHz for proton) were obtained in CDCl₃(Me₄Si internal

reference), unless noted otherwise. Chemical ionization mass

spectra were obtained with methane as the reagent gas (1

torr). Analytical TLC was performed with Whatman 250-µm

silica gel plates; preparative TLC was performed with Analtech

1000-µm silica gel GF plates. Flash column chromatography

was conducted with flash column silica gel (40-63 µm), and

column chromatography was conducted with standard silica

gel. Preparative HPLC was performed on a Waters Prep 500A

with silica gel columns. Analytical HPLC was carried out on

a YMC basic column (25 cm × 4.6 mm), with detection at 254

nm by using a Waters 481 UV detector, unless noted otherwise.

Microanalysis was performed by Robertson Microlit Labora-

tories, Inc.

1-(2-Ben zofu r a n yl)-2-[[2-(2-b en zofu r a n yl)-2-h yd r oxy-

eth yl](p h en ylm eth yl)a m in o]-eth a n on e (17). A solution of

2-acetylbenzofuran (219.7 g, 1.37 mol) in 2-pyrrolidinone (157

mL) was heated to reflux while being stirred mechanically

under argon. 2-Pyrrolidinone hydrotribromide (701.6 g, 1.37

mol) was added as a solid portionwise over ca. 40 min. After

10 min, an exotherm occurred, which was controlled by a water

bath. During the exotherm, the reaction turned from dark-

orange-red to white; after the exotherm ceased (ca. 15 min),

reflux was maintained by heating. After 1.5 h, another portion

(R*,R*)-r,r′-[[(P h en ylm et h yl)im in o]b is(m et h ylen e)]-

bis-2-ben zofu r a n m et h a n ol (a n ti-18). (R)-Alpine-hydride

(148 mL, 0.074 mol; 0.5 M in THF) was added dropwise over

1.5 h to a stirred solution of 17 (15.0 g, 0.035 mol) in THF

(300 mL) under argon at -95 to -98 °C. The reaction was

slowly warmed to 0 °C over 18 h and concentrated in vacuo.

The residue was dissolved in ethyl ether (400 mL) and

extracted with aqueous 1 N HCl (200 mL; CAUTION, foam-

ing). An oil separated from the ether layer and was collected

with the acidic aqueous layer. The ether layer was extracted

with 1 N HCl (3 × 100 mL), and the combined acidic potions

(including the oil) were extracted with ethyl ether (2 × 200

mL), basified to ca. pH 11 with 50% aqueous NaOH, and

extracted with ethyl ether (3 × 200 mL). The combined

extracts were washed twice with brine, dried (K₂CO₃), and

concentrated in vacuo to furnish crude 18 as a brown oil that

was a 5:1 mixture of the anti and syn isomers (13.1 g, 87%).

This oil was crystallized twice from methanol at -20 °C to

give the anti isomer of 18 as a white solid (5.07 g, 34%), mp

103-104 °C. ¹H NMR δ 2.97-3.21 (m, 4 H), 3.68 (br s, 2 H,

OH), 3.73 (d, J ) 13.6 Hz, 1 H, CH₂Ph), 3.89 (d, J ) 13.6 Hz,

1 H, CH₂Ph), 4.88 (dd, J ) 8.9, 3.4 Hz, 2 H, CHOH), 6.59 (s,

(24) Shriner, R. L.; Anderson, J . J . Am. Chem. Soc. 1939, 61, 2705-

2708.

(25) Burger, A.; Deinet, A. J . J . Am. Chem. Soc. 1945, 67, 566-

569.

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7975

2 H), 7.12-7.34 (m, 9 H), 7.41 (d, J ) 8.0 Hz, 2 H), 7.49 (d, J

) 7.5 Hz, 2 H). FAB-MS m/z 428 (MH⁺). The enantiomeric

5 °C, treated with phenacyl bromide (44.4 g, 0.022 mol), and

slowly warmed to 23 °C over 18 h with stirring. The resulting

dark-brown suspension was filtered, and the filtrate was

concentrated in vacuo. The residue was partitioned between

CH₂Cl₂and 3 N NaOH. The organic layer was rinsed twice

with brine, dried (K₂CO₃), and concentrated in vacuo. The

residue was dissolved in methanol (600 mL) and acidified with

gaseous HCl. The deposited solid was collected and recrystal-

lized from methanol to afford the HCl salt of 23 (57.5 g, 71%)

as white crystals, mp 189-191 °C, dec. Analytical data were

collected on the free base, 23. ¹H NMR δ 2.22 (t, J ) 11.3 Hz,

1 H), 2.36 (d, J ) 11.2 Hz, 1 H), 2.94 (d, J ) 11.1 Hz, 1 H),

3.03 (d, J ) 11.2 Hz, 1 H), 3.61 (s, 2 H), 5.24 (dd, J ) 2.7, 11.0

Hz, 1 H), 7.29-7.46 (m, 13 H), 7.70 (d, J ) 6.8 Hz, 2 H). ¹³C

NMR δ 59.7, 62.2, 63.1, 71.8, 95.3, 125.7, 126.4, 127.4, 127.7,

128.0, 128.2, 128.4, 128.9, 129.1, 136.6, 139.8, 141.8. ES-MS

m/z 346 (MH⁺). FAB-HRMS calcd for C₂₃H₂₃NO₂+ H⁺,

346.1807; found, 346.1818. The spectral data indicate that 23

is mainly in the cyclic tautomeric form.

2-[Ben zyl-(2-h yd r oxy-2-p h en yleth yl)a m in o]-1-p h en yl-

et h a n ol (24, a n t i/syn ).²⁷A. Typ ica l Alp in e-H yd r id e

Red u ction . To a solution of 23 (75 mg, 0.22 mmol) in CH₂-

Cl₂(5 mL) at -78 °C was added dropwise R-alpine-hydride

(0.46 mmol, 0.5 M in THF) via syringe over 35 min. The

mixture was stirred at -78 °C for 26 h and quenched with

water. The mixture was extracted with CH₂Cl₂, and the

organic layer was washed (water, brine), dried (Na₂SO₄), and

concentrated in vacuo. The residue was purified by prepara-

tive TLC (hexanes/EtOAc, 2.5:1) to give 30 mg (40%) of 24 as

a colorless viscous solid with an anti/syn ratio of 10:1. ¹H NMR

δ 2.71-2.92 (m, 4 H), 3.70/3.99 (pair of d, J ) 13.5 Hz, 1.82

H, anti PhCH₂), 3.83 (s, 0.18 H, syn PhCH₂), 4.70/4.76 (pair

of dd, syn CHOH/anti CHOH, J ) 10.3, 3.4 Hz for the anti

isomer), 7.25-7.36 (m, 15 H). ¹³C NMR δ 59.9, 62.7 (anti),

63.6 (syn), 70.9 (anti), 72.1 (syn), 125.9, 127.6, 127.7, 128.4,

128.6, 129.2, 138.0, 142.2. CI-MS m/z 348 (MH⁺). FAB-

HRMS calcd for C₂₃H₂₅NO₂+ H⁺, 348.1964; found, 348.1976.

A 1.5:1 anti/syn mixture of 24, prepared by reducing 23 with

NaBH₄in methanol at 5 °C (as described for syn-18), gave the

following spectral data. ¹H NMR (400 MHz) δ 2.58-2.91 (m,

4 H, 2 CH₂), 3.57 (br s, 2 H, OH), 3.64/3.92 (pair of d, J ) 13.7

Hz, 0.6 H, anti PhCH₂), 3.83 (s, 0.8 H, syn PhCH₂), 4.63 (dd,

J ) 10.3, 4.6 Hz, 0.8 H, syn CHOH), 4.70 (dd, J ) 10.3 Hz,

3.4 Hz, 1.2 H, anti CHOH), 6.95-7.55 (m, 15 H).

1

ratio of the anti isomer of 18 was determined by 400-MHz H

NMR by using 5 mg of sample and 60 mg of (S)-(+)-trifluoro-

methyl-1-(9-anthryl)ethanol in 1 mL of CDCl₃at 21 °C. Useful

spectral nonequivalence (∆δ ) 0.02) was observed for the

3-position protons of the benzofuran groups at δ 6.552 and

6.572; ratio ) 50:50 ( 5.

(R*,S*)-r,r′-[[(P h en ylm eth yl)im in o]bis(m eth ylen e)]bis-

2-ben zofu r a n m eth a n ol (syn -18). A stirred suspension of 17

(5.00 g, 0.0118 mol) in methanol (300 mL) at 5 °C was treated

with NaBH₄(0.889 g, 0.0235 mol). After 1 h, the reaction was

acidified with 3 N HCl (50 mL), basified to pH 11 with 3 N

NaOH, and concentrated in vacuo. The residue was parti-

tioned between CH₂Cl₂and water and the organic layer was

washed with brine, dried (K₂CO₃), and concentrated in vacuo

to give crude 18 as a 1.3:1 mixture of the syn and anti isomers.

These isomers were separated by preparative HPLC (CH₂Cl₂),

after pre-equilibrating the column with CH₂Cl₂/MeOH/NH₄-

OH (90:9:1), to afford the syn isomer of 18 (1.76 g, 35%) as an

amber oil. ¹H NMR δ 3.05-3.20 (m, 4 H), 3.32 (br s, 2 H, OH),

3.78 (s, 2 H, CH₂Ph), 4.89 (dd, J ) 8.0, 4.9 Hz, 2 H, CHOH),

6.62 (s, 2 H), 7.12-7.30 (m, 9 H), 7.41 (d, J ) 8.0 Hz, 2 H),

7.51 (d, J ) 7.6 Hz, 2 H). FAB-MS m/z 428 (MH⁺).

r,r′-[(Im in o)b is(m et h ylen e)]b is-2-b en zofu r a n m et h a -

n ol (a n ti-16). A solution of anti-18 (4.91 g, 0.115 mol) in

methanol/concentrated NH₄OH (9:1; 80 mL) was combined

with 20% Pd(OH)₂/C (0.323 g) and hydrogenated at 1 atm for

2 h. The mixture was filtered through Nylon 66 (0.45 µm) and

concentrated in vacuo. The residue was recrystallized from

methanol to give anti-16 (2.75 g, 71%) as a white crystalline

solid, mp 123-125 °C. ¹H NMR (DMSO-d₆) δ 1.94 (br s, 1 H,

NH), 2.85-3.04 (m, 4 H, CH₂), 4.72-4.86 (m, 2 H, CHOH),

5.68 (d, J ) 5.3 Hz, 2 H, CHOH), 6.72 (s, 2 H), 7.15-7.32 (m,

4 H), 7.52 (d, J ) 8.0 Hz, 2 H), 7.57 (d, J ) 7.6 Hz, 2 H). CI-

MS m/z 338 (MH⁺). Anal. Calcd for C₂₀H₁₉NO₄: C, 71.20; H,

5.68; N, 4.15. Found: C, 71.07; H, 5.63; N, 4.08.

r,r′-[Im in ob is(m et h ylen e)b is-2-b en zofu r a n m et h a n ol

(syn -16). A solution of the syn-18 (3.62 g, 0.0085 mol) in

methanol and concentrated NH₄OH (9:1; 59 mL) was combined

with 20% Pd(OH)₂/C (0.238 g) and hydrogenated at 1 atm.

After 3 h, another portion of 20% Pd(OH)₂/C (0.238 g) was

added and stirring was continued for 1.5 h. The reaction was

filtered through Nylon 66 (0.45 µm) and concentrated in vacuo.

The residue was recrystallized twice from methanol to give

syn-16 (1.19 g, 41%) as a white crystalline solid, mp 125-127

°C. ¹H NMR (DMSO-d₆) δ 1.95 (br s, NH, 1 H), 2.83-3.06 (m,

4 H, CH₂), 4.70-4.85 (m, 2 H, CHOH), 5.70 (d, J ) 4.9 Hz, 2

H, CHOH), 6.74 (s, 2 H), 7.13-7.33 (m, 4H), 7.52 (d, J ) 7.9

Hz, 2 H), 7.57 (d, J ) 7.5 Hz, 2 H). CI-MS m/z 338 (MH⁺).

Anal. Calcd for C₂₀H₁₉NO₄: C, 71.20; H, 5.68; N, 4.15.

Found: C, 71.23; H, 5.68; N, 4.11.

2-[(2-Hyd r oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-1-

p h en yleth a n on e (23). A solution of 2-bromoacetophenone

(100.0 g, 0.50 mol) in THF (500 mL) was added at 5 °C over

45 min under argon to a stirred solution of dibenzylamine (99.1

g, 0.50 mol) and N,N-diisopropylethylamine (129.8 g, 1.01 mol).

The reaction was slowly warmed to 23 °C over 18 h and

filtered. The filtrate was concentrated in vacuo, and the

residue was recrystallized from 2-propanol to afford the amino

ketone as beige crystals (101.5 g, 67%), mp 81-82 °C (lit.²⁶

mp 81.5-82 °C). A solution of this material (99.4 g, 0.33 mol)

in methanol (2 L) was treated with concentrated HCl (27 mL,

0.33 mol), combined with 20% Pd(OH)₂/C (9.20 g), hydroge-

nated on a Parr apparatus at 60 psig for 1 h, and filtered. The

filtrate was concentrated in vacuo, and the residue was

recrystallized from 2-propanol to afford 73.2 g (85%) of

2-benzylamino-1-phenylethanol‚HCl as white crystals, mp

225-228 °C (lit.¹⁹mp 219.5-221.5 °C). A solution of the free

base of this material (49.1 g, 0.22 mol) and N,N-diisopropyl-

ethylamine (54.8 g, 0.42 mol) in THF (370 mL) was cooled to

B. Typ ica l Zn (BH₄)₂Red u ction .²⁸To a solution of 23

(75 mg, 0.22 mmol) in CH₂Cl₂(5 mL) at -78 °C was added

dropwise Zn(BH₄)₂(0.30 mmol, 0.5 M in Et₂O)²⁸via syringe

over 35 min. The mixture was stirred at -78 °C for 26 h and

quenched with water. The mixture was extracted with CH₂-

Cl₂, and the organic layer was washed (water, brine), dried

(Na₂SO₄), and concentrated in vacuo. The residue was sepa-

rated by preparative TLC (hexanes/EtOAc, 2.5:1) to give 35

mg (47%) of 24 (anti and syn) as a colorless viscous solid with

1

an anti/syn ratio of 13:1 (determined by H NMR).

3-[(2-H yd r oxyp r op yl)(p h e n ylm e t h yl)a m in o]-2-p r o-

p a n on e (25). To a solution of 1-benzylamino-2-propanol (660

mg, 4.0 mmol) and N,N-diisopropylethylamine (1.03 g, 8.0

mmol) in THF (25 mL) was added chloroacetone (555 mg, 6.0

mmol). The mixture was stirred at 60 °C for 20 h and

concentrated in vacuo. The residue was separated by flash

column chromatography (hexanes/EtOAc, 3:1) to afford 530 mg

(60%) of 25 as a colorless viscous oil. ¹H NMR δ 1.10 (d, J )

6.3 Hz, 3 H), 1.35 (s, 3 H), 1.80 (t, J ) 11.0 Hz, 1 H), 2.06 (d,

J ) 10.9 Hz, 1 H), 2.71 (t, J ) 9.2 Hz, 2 H), 3.51 (s, 2 H),

3.99-4.10 (m, 1 H), 7.25-7.36 (m, 5 H). ¹³C NMR δ 18.7, 25.1,

(27) (a) For the R,R enantiomer of the anti isomer of 24, see:

Manickam, G.; Sundararjan, G. Tetrahedron: Asymmetry 1997, 8,

2271-2278. Dubois, L.; Fiaud, J .-C.; Kagan, H. B. Tetrahedron:

Asymmetry 1995, 6, 1097-1104. Trost, B. M.; Van Vranken, D. L.;

Bingel, C. J . Am. Chem. Soc. 1992, 114, 9327-9343. (b) An expansion/

blowup of the N-benzyl methylene region of the ¹H NMR spectrum is

presented in the Supporting Information (see the paragraph at the

end of this paper regarding Supporting Information).

(26) Suzuki, T.; Takamoto, M.; Okamoto, T.; Takayama, H. Chem.

Pharm. Bull. 1986, 34, 1888-1900.

(28) Gensler, W. J .; J ohnson, F.; Sloan, A. D. B. J . Am. Chem. Soc.

1960, 82, 6074-6081.

7976 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

59.1, 61.3, 62.2, 65.0, 93.7, 127.0, 128.3, 129.9, 137.1. CI-MS

m/z 222 (MH⁺). Anal. Calcd for C₁₃H₁₉NO₂‚0.1H₂O: C, 69.99;

H, 8.68; N, 6.28. Found: C, 69.76; H, 8.70; N, 6.19.

128.31 (CH), 144.72 (0.55 C), 144.75 (0.45 C). ES-MS m/z 257

(MH⁺). CI-HRMS calcd for C₁₇H₂₀O₂+ NH₄⁺, 274.1807; found,

274.1793. Remarkably, the ¹H NMR spectrum revealed two

OH signals, one for each isomer, allowing for their quantitation

by integration. However, there was no evidence for a pair of

triplets at δ 3.58 and 3.60, as reported by Neudeck and Schlo¨gl

for a 60-MHz (CDCl₃) spectrum of 28.³⁰Clearly, ¹³C NMR is

more sensitive for identifying the two isomers. We were

unable to assign anti and syn isomers.

3-[(2-H yd r oxyp r op yl)(p h e n ylm e t h yl)a m in o]-2-p r o-

p a n ol (26, a n ti/syn ).²⁹To a solution of 25 (35 mg, 0.16 mmol)

in CH₂Cl₂(7 mL) at -78 °C was added dropwise R-alpine-

hydride (1.1 mL, 0.5 M in THF, 0.55 mmol) via syringe. The

mixture was stirred at -78 °C, and no reaction was observed

after 17 h. The cooling bath was removed, and the reaction

mixture was allowed to warm to 0 °C within 2 h. After stirring

at 0 °C for 1 h, the mixture was quenched with saturated

aqueous NH₄Cl. The mixture was extracted with ethyl

acetate, and the organic layer was washed (water, brine), dried

(Na₂SO₄), and concentrated in vacuo. The residue was sepa-

rated by preparative TLC (hexanes/EtOAc, 1:1) to give 30 mg

(84%) of 26 as a colorless viscous oil with an anti/syn ratio of

7.8:1. ¹H NMR δ 1.02 (d, J ) 6.3 Hz, 6 H), 2.37 (d, J ) 6.4

Hz, 4 H), 3.44 (d, J ) 13.7 Hz, 0.885 H, anti PhCH₂) 3.64 (s,

0.23 H, syn PhCH₂), 3.76-3.83 (m, 3 H, includes d for other

anti PhCH₂), 7.18-7.29 (m, 5 H). ¹³C NMR δ 20.2 (anti), 20.5

(syn), 59.7 (anti), 60.1 (syn), 62.0 (anti), 63.5 (syn), 63.9 (anti),

65.3 (syn), 127.3, 128.4, 128.9, 138.3. CI-MS m/z 224 (MH⁺).

5-Hyd r oxyl-1,5-d ip h en yl-1-p en ta n on e (27). A solution

of 1,5-diphenyl-2,4-pentadiene-1-one²⁰(119.2 g, 0.509 mol) in

a 500 mL of benzene/ethanol (1:1) was treated with chlorotris-

(triphenylphosphine)rhodium(I) (9.40 g, 0.0102 mol) and hy-

drogenated on a Parr apparatus at 50 psig and 50 °C for 1.5

h. (Note: The diol was the major product when the literature

method²⁰was used.) The reaction was cooled to 23 °C,

concentrated in vacuo, diluted with ethyl ether (800 mL), and

filtered. The filtrate was concentrated in vacuo and recrystal-

lized from methanol (300 mL) at 5 °C to give 1,5-diphenyl-1-

pentanone as a tan solid (91.7 g, 76%), mp 46-48 °C (lit.²⁰mp

45-47 °C). N-Bromosuccimide (23.9 g, 0.134 mol) and ben-

zoylperoxide (8.13 g, 0.034 mol) were added to a vigorously

stirred, degassed, refluxing solution of this ketone (32.3 g,

0.134 mol) in CCl₄(150 mL). The reaction was irradiated with

a 150-W incandescent flood lamp for 15 min, cooled to 23 °C,

and filtered. The filtrate was rinsed sequentially with satu-

rated aqueous NaHCO₃, 1 N Na₂S₂O₃, and brine, dried

(MgSO₄), and concentrated in vacuo. The residue was recrys-

tallized from hexanes/EtOAc (9:1) to give 26.6 g (31%) of

5-bromo-1,5-diphenyl-1-pentanone as a white solid (lit.²⁰mp

112-114 °C). This material was reacted with 5% aqueous

KOH in 1,4-dioxane according to the literature method²⁰to

afford 27, which was recrystallized from cyclohexane to afford

white crystals, mp 78-80 °C (lit.²⁰mp 79-80 °C). ¹H NMR

(300 MHz) δ 1.71-2.00 (m, 4 H, 2 CH₂), 2.20 (br s,1 H, OH),

2.90-3.15 (m, 2 H, CH₂CdO), 4.60-4.85 (m, 1 H, CHOH),

5.15-5.25 (m, 0.03 H, ketal OCHOH), 7.15-7.70 (m, 8 H),

7.80-8.05 (m, 2 H). ¹³C NMR (75 MHz) δ 20.3 (CH₂), 38.1

(CH₂), 38.4 (CH₂), 74.2 (CHOH), 125.7 (CH), 127.4 (CH), 127.9

(CH), 128.36 (CH), 128.43 (CH), 132.9 (CH), 136.8 (C), 144.4

(C), 200.0 (CdO). ES-MS m/z 255 (MH⁺). The spectral data

indicate that 27, unlike 17 and 23, exists almost exclusively

(>95%) in the keto form.

2-[(2-Hyd r oxy-2-p h en yleth yl)th io]-1-p h en yleth a n on e

(29).²¹A solution of diphenacyl sulfide (307 mg, 1.13 mmol)

and methanol (31 mL) was cooled to 0 °C and treated with

NaBH₄(11 mg, 0.29 mmol). The mixture was allowed to warm

to 23 °C and was stirred for 24 h (reaction monitored by TLC:

hexanes/EtOAc, 3:1; developed twice; visualized by UV and

phosphomolybdic acid). After being stirred for 24 h, the

mixture was recooled to 0 °C and more NaBH₄(11 mg, 0.29

mmol) was added. The reaction was again warmed to 23 °C

and stirred. After 24 h, the reaction was recooled to 0 °C and

NaBH₄(5.0 mg, 0.13 mmol) was added. The reaction was

warmed to 23 °C and stirred for 6 h. TLC analysis showed

only a trace of starting material and formation of a ca. 1:1

mixture of keto alcohol and diol. The mixture was poured into

30 mL of water at 0 °C, concentrated in vacuo to remove the

methanol, and extracted with ethyl acetate (2 × 60 mL). The

combined organic solution was washed with brine, dried

(MgSO₄), and concentrated in vacuo to provide crude product

(326 mg). The mixture was purified by column chromatogra-

phy by using an ethyl acetate-hexanes gradient (10-35%

EtOAc) to provide pure 29 (104 mg, 34%) as a light-yellow oil.

¹H NMR δ 2.80 (dd, J ) 9.0, 14.0 Hz, 1 H), 2.94 (dd, J ) 3.7,

14.0 Hz, 1 H), 3.26 (s, 1 H), 3.85 (d, J ) 14.6 Hz, 1 H), 3.91 (d,

J ) 14.6 Hz, 1 H), 4.82 (m, 1 H), 7.20-7.40 (m, 5 H), 7.47 (dd,

J ) 7.1, 7.4 Hz, 2 H), 7.59 (t, J ) 7.4 Hz, 1 H), 7.96 (d, J ) 7.1

Hz, 2 H). ¹³C NMR δ 37.3, 41.6, 72.1, 125.7, 127.7, 128.4,

128.5, 128.6, 133.5, 135.1, 142.4, 194.8. CI-MS m/z 273 (MH⁺).

Anal. Calcd for C₁₆H₁₆O₂S: C, 70.56; H, 5.92; S, 11.77.

Found: C, 70.34; H, 5.92; S, 11.75.

r,r′-[(Th io)bis(m eth ylen e)]bis-ben zen em eth a n ol (30).²¹

To a solution of the keto alcohol 29 (40 mg, 0.147 mmol) in

CH₂Cl₂(6.7 mL) stirring at -78 °C was added dropwise

R-alpine-hydride (0.62 mL, 0.309 mmol) over 15 min. The

reaction was stirred at -78 °C and assayed by TLC (25%

hexanes/EtOAc, 3:1; developed twice; visualized with phos-

phomolybdic acid). After 1 h, a trace of 29 remained; after 3

h, the reaction was complete. The mixture was quenched with

saturated aqueous NH₄Cl (2 mL) at -78 °C and allowed to

warm to 23 °C. The aqueous layer was extracted with ethyl

acetate (3 × 10 mL), and the combined extracts were dried

(MgSO₄) and concentrated in vacuo. The residue was purified

by column chromatography by using an ethyl acetate-hexanes

gradient (20-35% EtOAc) to provide the diols 30 (32.4 mg,

81%). The ratio was determined to be ca. 1.3:1 anti/syn by ¹³

C

NMR peak intensities (δ 41.7 for anti CH₂, 42.2 for syn CH₂;

72.4 for anti CH, 73.2 for syn CH). The spectral data agreed

with results already in the literature.^{21 1}H NMR δ 2.70-2.85

(m, 2 H), 2.85-3.00 (m, 2 H), 3.20 (s, 1 H, anti OH), 3.32 (s, 1

H, syn OH), 4.75-4.80 (m, 2 H), 7.20-7.40 (m, 10 H). ¹³C

NMR δ 41.7 and 42.2 (anti/syn isomers), 72.4 and 73.2 (anti/

syn isomers), 125.7, 127.9, 128.4, 142.3, and 142.4 (anti/syn

isomers). CI-MS m/z 239 (MH⁺- 2H₂O).

2-[(2-Hyd r oxy-2-p h en yleth yl)m eth yla m in o]-1-p h en yl-

eth a n on e (31). To a solution of R-(methylaminomethyl)-

benzyl alcohol (3.0 g, 22.0 mmol) and N,N-diisopropylethyl-

amine (3.4 g, 26.3 mmol) in THF (145 mL) was added

3-chloropropiophenone (5.2 g, 26.1 mmol). The mixture was

stirred at 23 °C for 24 h and then diluted with CH₂Cl₂. The

solution was washed with aqueous NH₄Cl (100 mL), dried

(MgSO₄), and concentrated in vacuo. The residue was purified

by flash column chromatography (hexanes/EtOAc, 2:1) to

afford 3.05 g (69%) of 31 as a light-yellow viscous oil. ¹H NMR

δ 2.04-2.24 (m, 2 H), 2.31 (s, 3 H), 2.89-2.99 (m, 2 H), 5.26

1,5-Diph en yl-1,5-pen tan ediol (28, an ti/syn ).^4a,30R-Alpine-

hydride (1.65 mL, 0.5 M in THF, 0.83 mmol) was added

dropwise under argon over 15 min to a stirred solution of 27

(100 mg, 0.39 mmol) in THF (2.5 mL) at -100 °C. The mixture

was slowly warmed to 5 °C over 16 h and concentrated in

vacuo. The residue was partitioned between 1 N HCl and

ethyl ether; the organic layer was dried (Na₂SO₄) and concen-

trated in vacuo. The residue was purified by preparative TLC

(hexanes/EtOAc, 3:1) to give 28 (77 mg, 77%) as a syrupy 1.2:1

mixture of isomers. ¹H NMR (300 MHz, 77 mg/mL) δ 1.10-

1.95 (m, 6 H, 3 CH₂), 2.70 (br s, 1.1 H, OH), 2.85 (br s, 0.9 H,

OH), 4.57 (pair of dd that appears as a t, 2 H, CHOH), 7.10-

7.45 (m, 10 H). ¹³C NMR (75 MHz, 77 mg/mL) δ 22.04 (0.55

CH₂), 22.11 (0.45 CH₂), 38.63 (0.45 CH₂), 38.71 (0.55 CH₂),

74.07 (0.45 CH), 74.22 (0.55 CH), 125.76 (CH), 127.33 (CH),

(29) For the S,S enantiomer of the anti isomer of 26, see: Dubois,

L.; Fiaud, J .-C.; Kagan, H. B. Tetrahedron 1995, 51, 3803-3812.

(30) Neudeck, H.; Schlo¨gl, K. Monatsh. Chem. 1975, 106, 229-259.

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7977

(dd, J ) 2.8, 8.2 Hz, 1 H), 7.30-7.47 (m, 8 H), 7.72 (d, J ) 6.7

Hz, 2 H). ¹³C NMR δ 45.7, 61.9, 65.0, 71.6, 95.5, 125.7, 126.4,

127.4, 127.8, 127.9, 140.0, 142.3. ES-MS m/z 270 (MH⁺). Anal.

Calcd for C₁₇H₁₉NO₂‚0.1H₂O: C, 75.31; H, 7.14; N, 5.17.

Found: C, 74.98; H, 7.19; N, 4.84.

sium carbonate (2.90 g, 21.0 mmol) in acetonitrile (180 mL)

was heated at 75 °C with stirring for 48 h. The reaction

mixture was cooled to 23 °C and filtered; the filtrate was

concentrated in vacuo, and the residue was dissolved in ethyl

acetate. The solution was washed (water, brine), dried (Na₂-

SO₄), and concentrated in vacuo. The residue was purified

by flash column chromatography (CH₂Cl₂/MeOH, 9:1) to afford

2-[(2-Hyd r oxy-2-p h en yleth yl)m eth yla m in o]-1-p h en yl-

eth a n ol (32, a n ti/syn ). Reduction of 31 (85 mg, 0.32 mmol)

with 2.2 mol equiv of R-Alpine-Hydride (CH₂Cl₂, -78 °C, 24

h) and isolation by preparative TLC (CH₂Cl₂/MeOH, 95:5) gave

51 mg (35%) of 32 (60% of 31 was recovered) as a colorless

viscous oil with an anti/syn ratio of 2.3:1 (determined by ¹H

and ¹³C NMR). ¹H NMR δ 2.41 (s, 0.9 H, syn NMe) and 2.45

(s, 2.1 H, anti NMe), 2.52-2.77 (m, 4 H), 4.68-4.73 (m, 2 H),

7.21-7.39 (m, 10 H). ¹³C NMR δ 42.0 (syn), 43.2 (anti), 65.6

(anti), 66.6 (syn), 70.4 (anti), 70.9 (syn), 125.8, 127.5, 128.3,

142.0. ES-MS m/z 272 (MH⁺). Anal. Calcd for C₁₇H₂₁NO₂‚

0.7H₂O: C, 71.86; H, 7.96; N, 4.93. Found: C, 71.99; H, 7.57;

N, 4.48.

3-[(2-Hyd r oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-1-

p h en yl-1-p r op a n on e (34). To a solution of 2-benzylamino-

1-phenylethanol (5.00 g, 22.0 mmol) and N,N-diisopropyleth-

ylamine (4.23 g, 32.7 mmol) in THF (150 mL) was added

3-chloropropiophenone (3.70 g, 22.0 mmol). The mixture was

stirred at 23 °C for 16 h and then diluted with ethyl acetate

(200 mL). The organic solution was washed with water (50

mL) and brine (50 mL), dried (Na₂SO₄), and concentrated in

vacuo. The crude product was purified by flash column

chromatography (hexanes/EtOAc, 2:1) to afford 6.30 g (80%)

of 34 as a colorless viscous oil. ¹H NMR (400 MHz, 2D-COSY)

δ 2.57-2.65 (dd, J ) 13.5, 10.1 Hz, 1 H, CHOHCH₂N), 2.70-

2.75 (dd, J ) 3.5, 13.5 Hz, 1 H, CHOHCH₂N), 2.92-3.00 (m,

1 H, CH₂CH₂N), 3.15-3.23 (m, 3 H, CH₂CH₂N and CH₂CdO),

3.58 (d, J ) 13.7 Hz, 1 H, PhCH₂N), 3.93 (d, J ) 13.7 Hz, 1 H,

PhCH₂N), 4.71 (dd, J ) 3.5, 10.1 Hz, 1 H, CHOH), 7.22-7.37

(m, 10 H), 7.49 (dd, J ) 7.7 Hz, 2 H, m-benzoyl), 7.60 (dd, J )

7.3 Hz, 1 H, p-benzoyl), 7.92 (d, J ) 7.5 Hz, 2 H, o-benzoyl).

¹³C NMR δ 36.3, 49.3, 59.0, 63.0, 69.8, 125.8, 127.1, 127.2,

127.3, 128.0, 128.2, 128.3, 128.4, 128.5, 128.8, 133.1, 136.7,

138.2, 142.0, 199.2. ES-MS m/z 360 (MH⁺). Anal. Calcd for

2.15 g (76%) of 36 as a colorless tacky solid. [R]²⁴+29.2 (c

D

1.0, MeOH). ¹H NMR δ 1.82-2.00 (m, 2 H), 2.88-3.01 (m, 2

H), 3.85 (s, 2 H), 4.94 (dd, J ) 3.2, 8.5 Hz, 1 H), 7.25-7.38 (m,

10 H). ¹³C NMR δ 36.6, 47.0, 53.2, 74.8, 125.4, 126.9, 127.6,

128.1, 128.5, 128.6, 128.9, 137.4, 144.5. CI-MS m/z 242 (MH⁺).

Anal. Calcd for C₁₆H₁₉NO‚0.03CH₂Cl₂: C, 78.95; H, 7.88; N,

5.75. Found: C, 79.19; H, 7.48; N, 6.11.

Syn th esis of Au th en tic a n ti Diols 35 (R,R). A neat

mixture of 36 (36 mg, 0.15 mmol) and (R)-styrene oxide (108

mg, 0.90 mmol) was stirred at 120 °C for 2 h. The crude

mixture was cooled and purified by preparative TLC (CH₂Cl₂/

MeOH, 20:1) to afford 23 mg (42%) of 35 (anti, R,R) as a

colorless viscous oil. ¹H NMR δ 1.81-2.05 (m, 2 H), 2.57-

2.80 (m, 3 H), 2.90-2.99 (m, 1 H), 3.53 (d, J ) 13.2 Hz, 1 H),

3.93 (d, J ) 13.2 Hz, 1 H), 4.73-4.81 (m, 2 H), 7.23-7.39 (m,

15 H). CI-MS m/z 362 (MH⁺). The ¹H NMR spectrum was

identical with that of the major product obtained from the

reduction of ketone 34

(R )-r-[2-[(P h e n y lm e t h y l)a m in o ]m e t h y l]b e n ze n e -

m eth a n ol ((R)-37).³²(R)-2-Chloro-1-phenylethanol (5.33 g,

34.0 mmol, Aldrich) and benzylamine (8.53 g, 79.6 mmol) were

heated to 180 °C for 2 h. The crude mixture was cooled and

purified by flash column chromatography (CH₂Cl₂/MeOH/

concentrated NH₄OH, 92:7:1) to afford 6.20 g (80%) of (R)-37

as a colorless viscous oil. ¹H NMR δ 2.74 (dd, J ) 3.3, 8.8 Hz,

1 H), 2.91 (dd, J ) 8.6, 3.6 Hz, 1 H), 3.81 (d, J ) 3.6 Hz, 2 H),

4.72 (dd, J ) 5.3, 3.5 Hz, 1 H), 7.24-7.39 (m, 10 H). ¹³C NMR

δ 53.4, 56.4, 71.7, 125.7, 127.0, 127.4, 128.1, 128.3, 139.8, 142.4.

CI-MS m/z 228 (MH⁺). Anal. Calcd for C₁₅H₁₇NO: C, 79.27;

H, 7.54; N, 6.17. Found: C, 79.19; H, 7.48; N, 6.11.

Syn th esis of Au th en tic syn Diols 35 (R,S). A mixture

of (R)-37 (1.0 g, 4.4 mmol), (S)-3-chloro-1-phenylpropanol (1.5

g, 8.8 mmol, Aldrich), potassium iodide (1.5 g, 9.0 mmol), and

potassium carbonate (1.2 g, 8.7 mmol) in DMF (29 mL) was

heated at 160 °C for 14 h. The reaction was cooled and

concentrated in vacuo. The residue was dissolved in ethyl

acetate and washed sequentially with saturated aqueous

ammonium chloride (2 × 10 mL), water (2 × 10 mL), and brine

(2 × 10 mL). The organic layer was dried (MgSO₄) and

concentrated. The residue was purified by flash column

chromatography (hexanes/EtOAc, 4:1) to afford 1.42 g (89%)

of 35 (syn, R,S) as a colorless viscous oil. ¹H NMR δ 1.91-

1.98 (m, 2 H), 2.69 (d, J ) 6.5 Hz, 2 H), 2.75-2.92 (m, 2 H),

3.66 (d, J ) 13.3 Hz, 1 H), 3.78 (d, J ) 13.3 Hz, 1 H), 4.77 (t,

J ) 6.5 Hz, 1 H), 4.86 (t, J ) 6.5 Hz, 1 H), 7.28-7.37 (m, 15

H). ES-MS m/z 362 (MH⁺). The ¹H NMR spectrum was

identical with that of the minor product obtained from the

reduction of ketone 34.

C

₂₄H₂₅NO₂: C, 80.20; H, 7.01; N, 3.90. Found: C, 79.80; H,

7.04; N, 3.86.

r-[2-[(2-Hyd r oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-

eth yl]ben zen em eth a n ol (35, a n ti/syn ). To a solution of 34

(24 mg, 0.067 mmol) in dry CH₂Cl₂(3 mL) at -78 °C was added

dropwise R-alpine-hydride (0.5 M in THF, 0.28 mL, 0.14 mmol)

via syringe. The mixture was stirred at -78 °C for 3 h,

quenched with water, allowed to warm to 23 °C, and extracted

with ethyl acetate. The organic layer was washed (water,

brine), dried (Na₂SO₄), and concentrated in vacuo. The residue

was purified by preparative TLC (hexanes/EtOAc, 4:1) to give

diol 35 as a mixture of anti and syn isomers. The diastereo-

meric ratio for 35, anti/syn ) 12:1 in this case, was nicely

quantitated by HPLC [YMC Basic column, 25 cm × 4.6 mm;

0.1% TFA-water (A)/0.1% TFA-acetonitrile (B); 1 mL/min;

gradient, 0-5 min 80%A/20%B, 5-30 min 50% A/50%B, 30-

(R)-2-[(3-Hyd r oxy-3-p h en ylp r op yl)(p h en ylm eth yl)a m i-

n o]-1-p h en yleth a n on e (38). To a solution of 36 (405 mg,

1.7 mmol) and N,N-diisopropylethylamine (260 mg, 2.0 mmol)

in THF (11 mL) was added 2-bromoacetophenone (340 mg, 1.7

mmol). The reaction mixture was stirred at 23 °C for 24 h

and diluted with ethyl acetate. The solution was washed with

water (20 mL) and brine (20 mL), dried (Na₂SO₄), and

concentrated in vacuo. The crude product was purified by

flash column chromatography (hexanes/EtOAc, 2.3:1) to afford

35 min 80%A/20%B; t_R) 25.1 min for the syn isomer, t_R

)

25.8 min for the anti isomer, t_R) 27.3 min for hydroxy ketone

1

34; λ ) 210 nm] and by 400-MHz H NMR. ¹H NMR δ 1.81-

2.07 (m, 2 H), 2.57-3.00 (m, 4 H), 3.54/3.94 (pair of d, J )

13.2 Hz, 1.85 H, anti PhCH₂), 3.66/3.77 (pair of d, J ) 13.2

Hz, 0.15 H, syn PhCH₂), 4.73-4.87 (m, 2 H), 7.23-7.39 (m,

15 H).^{27b 13}C NMR δ 35.6, 51.7 (syn), 52.7 (anti), 59.2 (syn),

59.5 (anti), 62.7(syn), 62.9 (anti), 70.9, 73.5 (syn), 74.5 (anti),

125.5, 125.6, 127.1, 127.2, 128.3, 128.5, 128.8, 129.3, 131.2,

137.8, 142.2, 144.4. CI-MS m/z 362 (MH⁺). Anal. Calcd for

580 mg (96%) of 38 as a colorless viscous oil. [R]²⁴+27.1 (c

D

0.39, CHCl₃). ¹H NMR δ 1.88-1.94 (m, 2 H), 2.92-2.97 (m, 2

H), 3.74 (d, J ) 13.1 Hz, 1 H), 3.94 (d, J ) 13.1 Hz, 1 H), 3.98

(d, J ) 8.3 Hz, 2 H), 4.95-4.99 (m, 1H), 7.19-7.45 (m, 12 H),

7.55 (dd, J ) 7.2 Hz, 1 H), 7.86 (d, J ) 7.6 Hz, 2 H). ¹³C NMR

δ 35.6, 53.0, 58.3, 59.0, 74.2, 125.6, 126.9, 127.6, 127.9, 128.2,

128.6, 129.4, 133.4, 135.8, 137.7, 144.8, 197.4. ES-MS m/z 360

C

₂₄H₂₇NO₂‚0.5H₂O: C, 77.81; H, 7.62; N, 3.78. Found: C,

77.80; H, 7.42; N, 3.59.

(R )-r-[2-[(P h e n y lm e t h y l)a m i n o ]e t h y l]b e n z e n e -

m eth a n ol (36).³¹A solution of (R)-(+)-3-chloro-1-phenylpro-

panol (2.00 g, 11.7 mmol, Aldrich), benzylamine (3.76 g, 35.1

mmol), potassium iodide (3.49 g, 21.0 mmol), and potas-

(32) Remuzon, P.; Soumeillant, M.; Dussy, C.; Bouzard, D. Tetra-

hedron 1997, 53, 17711-17726. Chini, M.; Crotti, P.; Macchia, F. J .

Org. Chem. 1991, 56, 5939-5942.

(31) For the racemic mixture of 36, see: Matyus, P.; Zara-Kaczian,

E.; Boros, S.; Bocskei, Z. J . Heterocycl. Chem. 1996, 33, 583-590.

7978 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

(MH⁺). Anal. Calcd for C₂₄H₂₅NO₂: C, 80.19; H, 7.01; N, 3.90.

Found: C, 79.66; H, 7.01; N, 3.91.

4.63 (t, J ) 5.3 Hz, 1 H), 7.22-7.42 (m, 7 H), 7.53 (t, J ) 7.3

Hz, 1 H), 7.85 (d, J ) 7.5 Hz, 2 H). ¹³C NMR δ 23.7, 37.6,

54.3, 58.4, 58.8, 73.7, 125.7, 127.0, 127.2, 128.0, 128.3, 128.4,

131.1, 145.2, 198.2. ES-MS m/z 374 (MH⁺). Anal. Calcd for

C₂₅H₂₇NO₂‚0.3H₂O: C, 79.25; H, 7.35; N, 3.70. Found: C,

79.63; H, 7.33; N, 3.58.

(R)-3-[(3-Hyd r oxy-3-p h en ylp r op yl)(p h en ylm eth yl)a m i-

n o]-1-p h en ylp r op a n on e (39). To a solution of 36 (425 mg,

1.76 mmol) and N,N-diisopropylethylamine (252 mg, 1.95

mmol) in THF (12 mL) was added 3-chloropropiophenone (300

mg, 1.78 mmol). The reaction mixture was stirred at 23 °C

for 24 h and diluted with ethyl acetate. The solution was

washed (water, brine), dried (Na₂SO₄), and concentrated in

vacuo. The residue was purified by flash column chromatog-

raphy (hexanes/EtOAc, 2:1) to afford 400 mg (61%) of 39 as a

colorless viscous oil. ¹H NMR δ 1.83-1.95 (m, 2 H), 2.70-

3.09 (m, 4 H), 3.19 (t, J ) 7.2 Hz, 2 H), 3.54 (d, J ) 13.2 Hz,

1 H), 3.84 (d, J ) 13.2 Hz, 1 H), 4.86 (dd, J ) 3.6, 8.0 Hz, 1

H), 7.28-7.40 (m, 10 H), 7.44 (t, J ) 7.4 Hz, 2 H), 7.55 (t, J )

7.2 Hz, 1 H), 7.90 (d, J ) 7.2 Hz, 2 H). ¹³C NMR δ 34.7, 36.1,

48.9, 53.0, 59.1, 75.1, 125.4, 126.8, 127.4, 128.0, 128.1, 128.5,

129.2, 133.1, 136.5, 137.6, 144.7, 199.8. ES-MS m/z 374 (MH⁺).

Anal. Calcd for C₂₅H₂₇NO₂‚0.2H₂O: C, 79.63; H, 7.33; N, 3.72.

Found: C, 79.83; H, 7.44; N, 3.38.

4-[(2-Hyd r oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-1-

p h en ylbu ta n ol (44, a n ti/syn ). Reduction of 43 (68 mg, 0.18

mmol) with R-alpine-hydride (2.5 mol equiv, CH₂Cl₂, -78 °C,

2.5 h) and isolation by flash column chromatography (CH₂Cl₂/

MeOH, 99:1) gave 60 mg (88%) of 44 as a colorless viscous oil;

anti/syn ) 1:1 (quantitated by ¹H NMR). ¹H NMR δ 1.52-

1.83 (m, 4 H), 2.45-2.73 (m, 4 H), 3.50/3.94 (pair of d, J )

13.4 Hz, 1.00 H, anti CH₂Ph), 3.55/3.84 (pair of d, J ) 13.4

Hz, 1.00 H, syn CH₂Ph), 4.5628-4.75 (m, 2 H), 7.24-7.35 (m,

15 H). ¹³C NMR δ 23.3 (syn), 23.5 (anti), 37.1 (syn), 37.5 (anti),

54.0 (anti), 54.1 (syn), 58.7 (syn), 58.9 (anti), 62.6 (syn), 62.8

(anti), 69.8 (anti), 69.9 (syn), 74.0, 125.8, 127.2, 127.3, 127.8,

128.2, 128.3, 128.4, 129.2, 137.8, 142.2, 144.8. ES-MS m/z 376

(MH⁺). Anal. Calcd for C₂₅H₂₉NO₂‚1.0H₂O: C, 76.31; H, 7.95;

N, 3.56. Found: C, 76.30; H, 7.60; N, 3.26.

4-[(2-Hyd r oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-1-

p h en ylbu ta n on e (46). In a flask equipped with a Dean-

Stark trap, a solution of 4-chlorobutyrophenone (3.00 g, 16.4

mmol), ethylene glycol (1.56 g, 25.1 mmol), and p-toluene-

sulfonic acid monohydrate (312 mg, 1.64 mmol) in benzene (55

mL) was stirred at 90 °C for 18 h. The reaction mixture was

cooled to 23 °C and diluted with ethyl acetate (100 mL). The

solution was washed with 10% Na₂CO₃(30 mL), water (30 mL),

and brine (30 mL), then dried (Na₂SO₄), and concentrated in

vacuo to give 3.70 g (99%) of ketal 45 as a yellow solid.^{34 1}H

NMR δ 1.81-1.91 (m, 2 H), 2.04 (t, J ) 6.4 Hz, 2 H), 3.53 (t,

J ) 6.8 Hz, 2 H), 3.77 (t, J ) 6.8 Hz, 2 H), 4.02 (t, J ) 6.8 Hz,

2 H), 7.29-7.37 (m, 3 H), 7.44 (d, J ) 7.6 Hz, 2 H). ¹³C NMR

δ 27.1, 37.8, 45.1, 64.5, 110.0, 125.6, 128.0, 128.2, 142.3. Anal.

Calcd for C₁₂H₁₅ClO₂: C, 63.58; H, 6.67. Found: C, 63.34; H,

6.65. A solution of 2-benzylamino-1-phenylethanol (1.02 g, 4.5

mmol), 45 (2.07 g, 9.1 mmol), potassium iodide (1.50 g, 9.0

mmol), and potassium carbonate (1.24 g, 9.0 mmol) in DMF

(90 mL) was stirred at 90 °C for 24 h. The mixture was cooled

to 23 °C and diluted with ethyl acetate (100 mL). The solution

was washed (water, brine), dried (Na₂SO₄), and concentrated

in vacuo. The residue was treated with 5% aqueous HCl in

THF at 23 °C for 24 h. The mixture was neutralized with 1 N

NaOH, extracted with CH₂Cl₂(100 mL), dried (MgSO₄), and

concentrated in vacuo. The residue was purified by flash

column chromatography (hexanes/EtOAc, 4:1) to afford 500 mg

(30%) of 46 as a colorless viscous oil. ¹H NMR δ 1.88-2.10

(m, 2 H), 2.55-2.75 (m, 4 H), 2.89-3.00 (m, 2 H), 3.51 (d, J )

13.4 Hz, 1 H), 3.91 (d, J ) 13.4 Hz, 1 H), 4.69 (dd, J ) 4.2, 9.6

Hz, 1 H), 7.21-7.32 (m, 10 H), 7.44 (dd, J ) 7.6 Hz, 2 H), 7.55

(dd, J ) 7.4 Hz, 1 H), 7.91 (d, J ) 7.2 Hz, 2 H). ¹³C NMR δ

21.2, 35.9, 53.1, 58.4, 62.4, 69.6, 125.8, 127.2, 127.3, 127.9,

128.2, 128.4, 128.5, 128.8, 129.0, 132.9, 136.8, 138.3, 142.1,

199.6. ES-MS m/z 374 (MH⁺). Anal. Calcd for C₂₅H₂₇NO₂‚

0.5H₂O: C, 78.50; H, 7.38; N, 3.66. Found: C, 78.64; H, 7.09;

N, 3.45.

3-[((R)-3-Hyd r oxy-3-p h en ylp r op yl)(p h en ylm eth yl)a m i-

n o]-1-p h en ylp r op a n ol (40, a n ti/syn ). Reduction of 39 (25

mg, 0.067 mmol) with R-alpine-hydride (2.1 mol equiv, CH₂-

Cl₂, -78 °C, 2 h) and isolation by preparative TLC (CH₂Cl₂/

MeOH, 9:1) gave 20 mg (80%) of 40 as a viscous oil, anti/syn

1

) 1.2:1 (determined by H NMR). ¹H NMR δ 1.75-1.98 (m, 4

H), 2.52-2.83 (m, 4 H), 3.44/3.79 (pair of d, J ) 13.1 Hz, 1.10

H, anti CH₂Ph), 3.65 (s, 0.90 H, syn CH₂Ph), 4.75-4.81 (m, 2

H), 7.17-7.40 (m, 15 H). ¹³C NMR δ 35.2, 51.7, 58.7, 73.7,

125.4, 126.6, 127.1, 127.4, 127.9, 128.2, 128.5, 128.7, 128.8,

131.8, 137.6, 144.6. ES-MS m/z 376 (MH⁺). Anal. Calcd for

C

₂₅H₂₇NO₂‚0.3H₂O: C, 78.72; H, 7.83; N, 3.68. Found: C,

78.33; H, 7.43; N, 3.45.

4-(P h en ylm eth yl)a m in o-1-p h en ylbu ta n ol (42). Sodium

borohydride (1.2 g, 32.1 mmol) was added to a solution of

3-chloropropiophenone (5.69 g, 31.2 mmol) in THF (310 mL)

at 0 °C. The reaction was stirred for 1 h at 0 °C and quenched

with saturated aqueous NH₄Cl (10 mL) and water (100 mL).

The solution was concentrated in vacuo and extracted with

CH₂Cl₂. The organic solution was dried (MgSO₄) and concen-

trated in vacuo. The residue was purified by flash column

chromatography (hexanes/EtOAc, 4:1) to afford 4.5 g (78%) of

41 as a pale-yellow viscous oil.^{33 1}H NMR δ 1.73-2.10 (m, 4

H), 3.50-3.3.58 (m, 2 H), 4.69 (t, J ) 5.6 Hz, 1 H), 7.25-7.38

(m, 5 H). ¹³C NMR δ 28.8, 36.0, 44.9, 73.8, 125.7, 127.6, 128.5,

144.2. Anal. Calcd for C₁₀H₁₃ClO‚0.5H₂O: C, 65.05; H, 7.10;

N, 19.20. Found: C, 65.28; H, 7.20; N, 18.88. A solution of

41 (5.5 g, 29.8 mmol), benzylamine (4.8 g, 44.8 mmol),

potassium iodide (9.9 g, 59.6 mmol), and potassium carbonate

(8.2 g, 59.3 mmol) in DMF (300 mL) was stirred at 90 °C for

72 h. The reaction mixture was cooled to 23 °C and diluted

with CH₂Cl₂(100 mL). The solution was washed with water

(30 mL), dried (Na₂SO₄), and concentrated in vacuo. The

residue was purified by flash column chromatography (hex-

anes/EtOAc, 4:1) to afford 3.2 g (42%) of 42 as a colorless

viscous oil. ¹H NMR δ 1.58-1.96 (m, 4 H), 2.60-2.67 (m, 1

H), 2.68-2.83 (m, 1 H), 3.78 (s, 2 H), 4.66 (dd, J ) 3.2, 5.0 Hz,

1 H), 7.18-7.37 (m, 10 H). ¹³C NMR δ 26.8, 39.2, 49.1, 53.6,

73.5, 125.5, 126.6, 127.1, 127.4, 127.5, 128.0, 128.3, 128.5,

Red u ction of 1,7-Hyd r oxy Keton e 46 to 44. Reduction

of 46 (25 mg, 0.067 mmol) with R-alpine-hydride (2.5 mol

equiv, CH₂Cl₂, -78 °C, 2.5 h) and isolation by preparative TLC

(hexanes/EtOAc, 3:1) gave 21 mg (84%) of 44; anti/syn ) 3:1

(determined by ¹H NMR). Reduction of 46 with Zn(BH₄)₂(4.0

mol equiv, CH₂Cl₂, -78 °C, 22 h) gave a 63% TLC isolated

yield of 44; anti/syn ) 1:2 (¹H NMR).

(2-H yd r oxy-2-p h en ylet h yl)p h en ylm et h yl Ca r b a m ic

Acid 1,1-Dim eth yleth yl Ester (47). 2-Benzylamino-1-phe-

nylethanol (3.0 g, 13.2 mmol) was dissolved in acetonitrile (150

mL), and the solution was treated with N,N-diisopropylethy-

lamine (4.6 mL, 26.4 mmol). The solution was cooled with

stirring to 0 °C, when the starting material separated. The

suspension was treated with di-tert-butyl carbonate (3.17 g,

14.52 mmol), and the resulting mixture was warmed to 23 °C.

145.93. ES-MS m/z 256 (MH⁺). FAB-HRMS calcd for C₁₇H₂₁

-

NO + H⁺, 256.1701; found, 256.1706.

2-[(4-Hyd r oxy-4-p h en ylbu tyl)(p h en ylm eth yl)a m in o]-1-

p h en yleth a n on e (43). To a solution of 42 (1.3 g, 5.1 mmol)

and N,N-diisopropylethylamine (0.8 g, 6.2 mmol) in THF (34

mL) was added 2-bromoacetophenone (1.2 g, 6.0 mmol). The

mixture was stirred at 23 °C for 24 h and diluted with CH₂-

Cl₂. The solution was washed with aqueous NH₄Cl (100 mL),

dried (MgSO₄), and concentrated in vacuo. The residue was

purified by flash column chromatography (hexanes/EtOAc, 2:1)

to afford 1.0 g (53%) of 43 as a light-yellow viscous oil. ¹H

NMR δ 1.58-1.80 (m, 4 H), 2.68-2.81 (m, 2 H), 3.86 (s, 2 H),

(33) Figadere, B.; Chaboche, C.; Franck, X.; Peyrat, J .-F.; Cave, A.

J . Org. Chem. 1994, 59, 7138-7141.

(34) Purchase, C. F.; Goel, O. P. J . Org. Chem. 1991, 56, 457-459.

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7979

After the mixture was stirred for 18 h, TLC analysis (5%

MeOH/CH₂Cl₂, ninhydrin) showed completion of the reaction.

The reaction was diluted with ether (200 mL), and the mixture

was washed sequentially with 0.1 N HCl (2 × 50 mL), water

(2 × 30 mL), and brine (1 × 30 mL). The organic layer was

dried (MgSO₄) and concentrated in vacuo to give 47 (4.54 g,

100% crude yield). The product solidified on standing; it was

recrystallized from hot hexanes and dried in vacuo at 45 °C

for 12 h to give 47 (3.66 g, 85%) as white crystals, mp 77-78

°C. ¹H NMR (DMSO-d₆) δ 1.33/1.36 (s, 9H, Boc rotamers),

3.05-3.35 (m, 2H), 4.30-4.53 (m, 2H), 4.81 (m, 1H), 5.51 (m,

1H), 7.15-7.40 (m, 10 H). ¹³C NMR (DMSO-d₆) δ 29.7, 51.8/

53.0 (Boc rotamers), 55.5/56.0 (Boc rotamers), 72.9/73.2 (Boc

rotamers), 80.5, 127.7, 128.6, 128.8, 128.9, 129.8, 130.1, 140.2/

140.5 (Boc rotamers), 145.4, 156.7. CI-MS m/z 328 (MH⁺).

Anal. Calcd for C₂₀H₂₅NO₃: C, 73.37; H, 7.70; N, 4.28.

Found: C, 73.23; H, 7.60; N, 4.42.

(2-Met h oxy-2-p h en ylet h yl)p h en ylm et h yl Ca r b a m ic

Acid 1,1-Dim eth yleth yl Ester (48). An oven-dried 50-mL

pointed flask was charged with a NaH oil dispersion (34 mg,

60% dispersion, 0.84 mmol), and the dispersion was washed

with pentane (2 × 2 mL). The resulting solid was suspended

in anhydrous ether (2 mL), and the suspension was cooled with

stirring to 0 °C. The stirred suspension was treated dropwise

with a solution of 47 (250 mg, 0.763 mmol) and dimethyl

sulfate (0.08 mL, 0.84 mmol) in ether (2 mL). The reaction

was allowed to warm to 23 °C. After the mixture was stirred

for 4 h, TLC analysis (EtOAc/hexanes, 1:3; phosphomolybdic

acid) showed incomplete reaction. The mixture was cooled to

0 °C and treated sequentially with dimethyl sulfate (0.08 mL,

0.84 mmol) and the 60% NaH oil dispersion (34 mg, 0.84

mmol). After the mixture was stirred for an additional 18 h,

TLC analysis showed completion of the reaction. The reaction

was cooled to 0 °C and diluted with anhydrous ether (20 mL).

The reaction was quenched with saturated aqueous NH₄Cl (10

mL). The organic phase was washed (water, brine), dried

(MgSO₄), and concentrated to give a quantitative yield of crude

product, which was purified by column chromatography (hex-

anes/EtOAc, 9:1) to yield 48 (222.5 mg, 86%) as a colorless oil.

¹H NMR δ 1.43/1.48 (s, 9 H, Boc rotamers), 3.23 (s, 3 H), 3.25-

3.50 (m, 2 H), 4.30-4.60 (m, 3 H), 7.10-7.40 (m, 10 H). ¹³C

NMR δ 28.3, 50.9/52.1 (Boc rotamers), 52.7/53.5 (Boc rotam-

ers), 56.9, 79.7, 82.4/83.1 (Boc rotamers), 126.5, 126.8, 126.9,

127.6, 127.7, 128.3, 138.3/138.7 (Boc rotamers), 139.9, 155.8.

CI-MS m/z 342 (MH⁺). Anal. Calcd for C₂₁H₂₇NO₃: C, 73.87;

H, 7.97; N, 4.10. Found: C, 73.97; H, 8.00; N, 4.12.

â-Meth oxy-N-(p h en ylm eth yl)ben zen eeth a n a m in e (49)‚

TF A. A stirred solution of Boc-amine 48 (898 mg, 2.63 mmol)

in CH₂Cl₂(3.7 mL) at 0 °C was treated dropwise with

trifluoroacetic acid (3.7 mL, 48.0 mmol). The reaction was

warmed to 23 °C and stirred for 1 h, at which time TLC

analysis (EtOAc/hexanes, 35:65; ninhydrin stain) showed

complete reaction. The mixture was concentrated to dryness

and azeotropically treated with CH₂Cl₂(3 × 5 mL). The oily

residue was further dried on a vacuum pump and then

crystallized from ether/hexanes. The solid was dried in vacuo

for 18 h at 45 °C to provide the TFA salt of 49 (605 mg, 65%)

as a white crystalline solid, mp 85-87 °C. ¹H NMR δ 2.95

(dd, J ) 12.8, 10.1 Hz, 1 H), 3.03 (dd, J ) 12.8, 3.2 Hz, 1 H),

3.19 (s, 3 H), 4.12 (s, 2 H), 4.52 (dd, J ) 10.1, 3.2 Hz, 1 H),

7.20-7.50 (m, 10 H). ¹³C NMR δ 51.1, 51.9, 56.5, 78.9, 126.5,

128.7, 128.8, 129.1, 129.4, 129.9, 130.5, 136.9. CI-MS m/z 242

(MH⁺). FAB-HRMS calcd for C₁₆H₁₉NO + H⁺, 242.1544;

found, 242.1538. Anal. Calcd for C₁₆H₁₉NO‚CF₃CO₂H: C,

60.80; H, 5.67; N, 3.94. Found: C, 60.80; H, 5.75; N, 3.89.

3-[(2-Meth oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-1-

p h en ylp r op a n on e (50). To a solution of unrecrystallized 49‚

TFA (1.94 g, 4.4 mmol) in THF (28 mL) being stirred at 0 °C

was added N,N-diisopropylethylamine (3.4 mL, 19.5 mmol) and

3-chloropropiophenone (783 mg, 4.6 mmol). The mixture was

allowed to warm to 23 °C. After the mixture was stirred for

48 h, TLC analysis (MeOH/CH₂Cl₂,1:9; ninhydrin stain) showed

a trace of starting material and product formation. The

mixture was concentrated, and the residue was diluted with

ethyl acetate. The solution was washed with 0.01 N HCl (3 ×

10 mL) and brine (2 × 10 mL). The organic layer was dried

(Na₂SO₄) and concentrated to dryness. Although chromato-

graphic purification of the crude product proved difficult, it

was eventually purified by column chromatography with an

ethyl acetate-hexanes gradient (5-20% EtOAc) to give 50

(521 mg, 32%) as an orange oil. ¹H NMR δ 2.68 (dd, J ) 13.9,

5.0 Hz, 1 H) 2.90 (dd, J ) 13.9, 7.2 Hz, 1 H), 2.95-3.15 (m, 4

H), 3.19 (s, 3 H), 3.69 (d, J ) 13.8 Hz, 1 H), 3.83 (d, J ) 13.8

Hz, 1 H), 4.25 (dd, J ) 7.2, 5.0 Hz, 1 H), 7.20-7.35 (m, 10 H),

7.42 (dd, J ) 7.6, 7.3 Hz, 2 H), 7.53 (t, J ) 7.3 Hz, 1 H), 7.85

(d, J ) 7.6 Hz, 2 H). ¹³C NMR δ 36.8, 50.1, 56.7, 59.7, 61.1,

82.8, 126.9, 127.5, 128.0, 128.1, 128.2, 128.5, 128.6, 128.8,

132.9, 136.9, 139.4, 141.1,199.7. CI-MS m/z 374 (MH⁺). FAB-

HRMS calcd for C₂₅H₂₇NO₂+ H⁺, 374.2120; found, 374.2114.

3-[(2-Meth oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-1-

p h en ylp r op a n ol (51, a n ti/syn ). A solution of methoxy

ketone 50 (32.5 mg, 0.087 mmol) in CH₂Cl₂(4 mL) at -78 °C

was treated dropwise with R-alpine hydride (0.37 mL, 0.185

mmol) over 15 min. After being stirred for 3 h, the reaction

was quenched with saturated aqueous NH₄Cl. The mixture

was warmed to 23 °C, and the layers were separated. The

organic layer was dried (Na₂SO₄) and concentrated to dryness.

The residue was purified by preparative TLC (hexanes/EtOAc,

1:3) to provide diols 51 (26.4 mg, 80%). ¹H NMR δ 1.77-1.95

(m, 2 H), 2.50-2.55 (m, 1 H), 2.72-2.80 (m, 1 H), 2.87-3.00

(m, 2 H), 3.24 (s, 3 H), 3.47 (d, J ) 13.2 Hz, syn, 0.2 H), 3.61

(d, J ) 13.3 Hz, anti, 0.8H), 3.88 (d, J ) 13.3 Hz, anti, 0.8 H),

4.03 (d, J ) 13.2 Hz, syn, 0.2 H), 4.30 (dd, J ) 3.4, 9.1 Hz, 1

H), 4.80 (dd, J ) 2.6, 9.1 Hz, 1 H), 7.22-7.37 (m, 15 H). The

anti/syn ratio was determined to be 4.0:1 by HPLC (YMC basic

25 cm × 4.6 mm; mobile phase 0.1% TFA-water (A) and 0.1%

TFA-acetonitrile (B) with a linear ratio of 65% A/35% B; 1

mL/min; λ ) 210 nm; injection volume, 5 µL; t_R) 25.7 min

(syn isomer), 27.0 min (anti isomer)). The data were in line

with those for the authentic anti isomer.

(S)-(2-Hyd r oxy-2-p h en yleth yl)p h en ylm eth yl Ca r ba m -

ic Acid 1,1-Dim eth yleth yl Ester ((S)-47). (S)-2-Chloro-1-

phenylethanol (5.33 g, 34.0 mmol, Aldrich) and benzylamine

(3.64 g, 34.0 mmol) were stirred at 180 °C for 2 h and then

cooled to 23 °C. The crude product, obtained as for (R)-37,

was purified by flash column chromatography (CH₂Cl₂/MeOH/

NH₄OH, 93:6:1) to afford 6.20 g (80%) of (S)-37 as a colorless

viscous oil. The ¹H NMR spectrum was identical with that

for (R)-37. To a solution of (S)-37 (1.23 g, 5.4 mmol) and N,N-

diisopropylethylamine (1.40 g, 10.8 mmol) in acetonitrile (60

mL) was added di-tert-butyl carbonate (1.3 g, 6.0 mmol). The

mixture was stirred at 23 °C for 21 h and diluted with CH₂Cl₂

(100 mL). The solution was washed with saturated aqueous

NH₄Cl (20 mL), water (20 mL), and brine (20 mL) and then

dried (Na₂SO₄) and concentrated in vacuo. The residue was

purified by flash column chromatography (CH₂Cl₂/MeOH/NH₄-

OH, 97:2:1) to afford 1.60 g (98%) of (S)-47 as a colorless

viscous oil. ¹H NMR δ 1.48 (s, 9 H), 3.30-3.54 (m, 2 H), 4.15-

4.46 (m, 2 H), 4.85-4.90 (m, 1 H), 7.18-7.30 (m, 10 H). ¹³C

NMR δ 28.3, 52.4, 55.5, 74.0, 125.7, 127.2, 128.5, 142.3, 157.9.

Anal. Calcd for C₂₀H₂₅NO₃‚0.8H₂O: C, 70.28; H, 7.85; N, 4.10.

Found: C, 69.98; H, 7.47; N, 4.22.

(S)-(2-Meth oxy-2-p h en yleth yl)p h en ylm eth yl Ca r ba m -

ic Acid 1,1-Dim eth yleth yl Ester ((S)-48). To a cooled

solution (0 °C) of (S)-47 (600 mg, 1.8 mmol) in ethyl ether (20

mL) was added sodium hydride (50 mg, 2.0 mmol), followed

by methyl iodide (320 mg, 2.3 mmol). The reaction mixture

was stirred at 23 °C for 1 h and then quenched with water (5

mL). The solution was extracted with ethyl acetate (50 mL),

dried (Na₂SO₄), and concentrated in vacuo. The residue was

purified by flash column chromatography (hexanes/EtOAc,1.5:

1) to afford 200 mg (32%) of (S)-48 as a colorless viscous oil.

¹H NMR δ 1.48 (s, 9 H), 3.23 (s, 3 H), 3.28-3.48 (m, 2 H),

4.37-4.49 (m, 3 H), 7.19-7.34 (m, 10 H). ¹³C NMR δ 28.4,

52.3, 53.6, 57.0, 79.8, 126.7, 127.7, 128.4, 142.3, 157.9. FAB-

HRMS calcd for C₂₁H₂₈NO₃+ H⁺, 342.2069; found, 342.2040.

â-Me t h oxy-N -(p h e n ylm e t h yl)b e n ze n e e t h a n a m in e ‚

TF A ((S)-49). To a cooled solution (0 °C) of (S)-48 (160 mg,

0.5 mmol) in CH₂Cl₂(2 mL) was added trifluoroacetic acid (3.0

g, 26.3 mmol). The reaction mixture was stirred at 0 °C for 1

7980 J . Org. Chem., Vol. 63, No. 22, 1998

Zhang et al.

h and concentrated in vacuo to afford 165 mg (98%) of (S)-49

as a white solid. ¹H NMR δ 2.95-3.12 (m, 2 H), 3.18 (s, 3 H),

4.17 (s, 2 H), 4.47 (dd, J ) 2.8, 10.2 Hz, 1 H), 7.23-7.39 (m,

10 H). ¹³C NMR δ 51.4, 52.1, 56.5, 78.6, 126.4, 128.9, 129.8.

FAB-HRMS calcd for C₁₆H₁₉NO + H⁺, 242.1547; found,

242.1545.

128.7, 133.3, 142.4, 199.5. ES-MS m/z 270 (MH⁺). An analyti-

cally pure sample was obtained by flash column chromatog-

raphy (hexanes/EtOAc, gradient of 90:10 to 50:50). Anal.

Calcd for C₁₇H₁₉NO₂‚0.1H₂O: C, 75.31; H, 7.14; N, 5.17.

Found: C, 75.23; H, 7.00; N, 5.25.

3-[(2-Hyd r oxy-2-p h en yleth yl)m eth yla m in o]-1-p h en yl-

p r op a n on e (54b). To a solution of R-(methylaminomethyl)-

benzyl alcohol (1.06 g, 7.0 mmol) and N,N-diisopropylethy-

lamine (1.81 g, 14.0 mmol) in THF (46 mL) was added

3-chloropropiophenone (1.42 g, 8.4 mmol). The mixture was

stirred at 23 °C for 5 h and filtered. The filtrate was

concentrated in vacuo, the residue was dissolved in ethyl

acetate, and the solution was washed (0.01 N HCl, brine). The

organic solution was treated with 14 mL of 1 N HCl in ether

and 40 mL of ether. The mixture was cooled to 0 °C, and the

solvent was decanted. The sticky residue was triturated with

ether and dissolved in 0.01 N HCl (15 mL). The aqueous

solution was washed with ether (2 × 10 mL), basified with 1

N NaOH (14 mL), and extracted with ethyl acetate (2 × 30

mL). The organic solution was washed with brine (2 × 15 mL),

dried (Na₂SO₄), and concentrated in vacuo to afford 1.78 g

(90%) of 54b as a colorless solid. ¹H NMR δ 2.39 (s, 3 H), 2.53-

2.56 (m, 2 H), 2.83-2.92 (m, 1 H), 3.05-3.21 (m, 3 H), 4.72

(dd, J ) 5.0, 8.8 Hz, 1 H), 7.24-7.60 (m, 8 H), 7.97 (d, J ) 7.3

Hz, 2 H). ¹³C NMR δ 36.4, 42.1, 52.5, 66.1, 69.5, 125.9, 127.4,

128.1, 128.3, 128.7, 133.2, 136.9, 142.1, 199.2. CI-MS m/z 284

(MH⁺). An analytically pure sample was obtained by flash

column chromatography (hexanes/EtOAc, 70:30). Anal. Calcd

for C₁₈H₂₁NO₂: C, 76.30; H, 7.47; N, 4.95. Found: C, 75.98;

H, 7.36; N, 4.75.

3-[(2-Hyd r oxy-2-p h en ylet h yl)(2,4,6-t r im et h ylb en zyl)-

a m in o]-1-p h en ylp r op a n on e (54c). A mixture of 54a (455

mg, 1.69 mmol), 2,4,6-trimethylbenzyl chloride (514 mg, 3.05

mmol), and N,N-diisopropylethylamine (873 mg, 6.77 mmol)

in THF-DMF (4:1, 18 mL) was stirred at 50 °C for 30 h. The

mixture was concentrated in vacuo, and the residue was

separated by flash column chromatography (hexanes/EtOAc/

Et₃N, 6:1:0.2; quickly to avoid the decomposition of the

product on the column) to give 387 mg (57%) of 54c as a

colorless viscous solid. ¹H NMR δ 2.22 (s, 3 H), 2.31 (s, 6 H),

2.53-2.69 (m, 2 H), 2.89-2.99 (m, 1 H), 3.05-3.24 (m, 3 H),

3.59 (d, J ) 12.5 Hz, 1 H), 3.85 (d, J ) 12.5 Hz, 1 H), 4.63 (dd,

J ) 3.2, 10.0 Hz, 1 H), 6.77 (s, 2 H), 7.21-7.36 (m, 5 H), 7.42

(t, J ) 7.4 Hz, 2 H), 7.55 (t, J ) 7.3 Hz, 1 H), 7.86 (d, J ) 7.4

Hz, 2 H). ¹³C NMR δ 20.2, 20.8, 36.6, 47.1, 49.8, 53.2, 57.3,

62.9, 70.2, 125.8, 127.2, 128.0, 128.1, 128.5, 128.8, 129.0, 129.2,

131.1, 133.0, 136.5, 136.7, 137.5, 142.1, 199.4. ES-MS m/z 402

(MH⁺). Free base 54c was converted to an HCl salt in ether

by adding 1 N HCl in ether (colorless solid). The resulting

colorless solid was collected and triturated with ether. Anal.

Calcd for C₂₇H₃₁NO₂‚1.0HCl‚0.4H₂O: C, 72.84; H, 7.43; N, 3.15;

Cl, 7.96. Found: C, 72.98; H, 7.41; N, 3.12; Cl, 8.15.

3-[(2-Meth oxy-2-p h en yleth yl)(p h en ylm eth yl)a m in o]-1-

p h en ylp r op a n ol (a n ti-51). A solution of (S)-49 (202 mg, 0.9

mmol), potassium iodide (300 mg, 1.8 mmol), potassium

carbonate (250 mg, 1.8 mmol), and (S)-3-chloro-1-phenylpro-

panol in DMF was refluxed for 18 h. The mixture was

concentrated in vacuo and diluted with ethyl acetate (50 mL).

The solution was washed with saturated aqueous ammonium

chloride (10 mL), water (10 mL), and brine (10 mL). The

organic layer was dried (MgSO₄) and concentrated in vacuo.

The residue was purified by flash column chromatography

(hexanes/EtOAc, 9:1) to afford 248 mg (75%) of anti-51 as a

colorless viscous oil. [R]²⁴+28.5 (c 1.0, CHCl₃). ¹H NMR δ

D

1.73-1.95 (m, 2 H), 2.49-2.55 (m, 1 H), 2.71-2.78 (m, 1 H),

2.89-2.98 (m, 2 H), 3.23 (s, 3 H), 3.60 (d, J ) 13.4 Hz, 1H),

3.87 (d, J ) 13.3 Hz, 1 H), 4.30 (dd, J ) 3.2, 9.0 Hz, 1 H), 4.80

(dd, J ) 2.5, 9.2 Hz, 1 H), 7.21-7.39 (m, 15 H). ¹³C NMR δ

35.5, 53.9, 56.7, 59.7, 61.3, 75.1, 82.1, 125.7, 126.7, 127.8, 128.5,

130.9, 138.3, 140.4, 145.2. Anal. Calcd for C₂₅H₂₉NO₂: C,

79.97; H, 7.79; N, 3.74. Found: C, 79.79; H, 7.70; N, 3.56.

This material was used to assign the anti isomer in the anti/

syn mixture of 51 from the reduction of 50.

3-(2-Hydr oxy-2-ph en yleth oxy)-1-ph en ylpr opan on e (52).

A solution of 1-phenyl-1,2-ethanediol (1.3 g, 9.4 mmol) in CH₂-

Cl₂(50 mL) was cooled to 0 °C, and sodium hydride (230 mg,

9.4 mmol) was added in portions. After the mixture was

stirred at 0 °C for 30 min, 3-chloropropiophenone (1.6 g, 9.5

mmol) was added and the reaction was stirred at 23 °C for 1

h. The reaction was quenched with saturated aqueous am-

monium chloride (10 mL) and diluted with CH₂Cl₂(100 mL).

The solution was washed (water, brine), dried (MgSO₄), and

concentrated in vacuo. The residue was purified by flash

column chromatography (hexanes/EtOAc, 2:1) to afford 400 mg

(16%) of 52 as a white solid. ¹H NMR δ 3.28 (t, J ) 6.0 Hz, 2

H), 3.49 (t, J ) 9.4 Hz, 1 H), 3.67 (dd, J ) 3.2, 6.9 Hz, 1 H),

3.90-4.06 (m, 2 H), 4.90 (d, J ) 9.0 Hz, 1 H), 7.25-7.40 (m, 5

H), 7.47 (dd, J ) 7.9 Hz, 2 H), 7.58 (dd, J ) 7.4 Hz, 1 H), 7.97

(d, J ) 7.2 Hz, 2 H). ¹³C NMR δ 38.5, 66.1, 72.5, 76.6, 126.2,

127.7, 128.1, 128.4, 128.7, 133.3, 136.8, 140.2, 198.4. Anal.

Calcd for C₁₇H₁₈O₃: C, 75.54; H, 6.72. Found: C, 75.34; H,

6.71.

3-(2-Hyd r oxy-2-p h en yleth oxy)-1-p h en ylp r op a n ol (53,

a n ti/syn ). Reduction of 52 (49 mg, 0.18 mmol) with R-alpine-

hydride (2.2 mol equiv, -78 °C, 18 h) and isolation by flash

column chromatography (hexanes/EtOAc, 7:3) provided 40 mg

(82%) of 53 as a colorless oil; anti/syn ) 1:1 (quantitated by

¹H NMR). ¹H NMR (DMSO-d₆) δ 1.57-1.85 (m, 2 H, CH₂Ph),

3.30-3.56 (m, 4 H), 4.56-4.62/4.63-4.69 (2 m, 0.50 H each),

5.14/5.32 (2 d, J ) 4.5 Hz, 0.50 H each), 7.20-7.37 (m, 10 H).

¹³C NMR (DMSO-d₆) δ 29.8, 69.2, 71.2, 73.1, 78.0, 127.4, 128.0,

128.4, 128.7, 129.6, 129.7, 145.0, 147.8. Anal. Calcd for

3-(2-H y d r o x y -2-p h e n y le t h y la m in o )-1-p h e n y lp r o -

p a n ol (55a , a n ti/syn ). Reduction of 54a (81 mg, 0.30 mmol)

with R-alpine-hydride (3.1 mol equiv, CH₂Cl₂, -78 °C, 4.5 h)

and isolation by preparative TLC (CH₂Cl₂/MeOH/NH₄OH, 90:

9:1) gave 70 mg (86%) of 55a as a colorless viscous oil; anti/

C

₁₇H₂₀O₃‚0.1H₂O: C, 74.39; H, 7.44. Found: C, 74.06; H, 7.69.

3-(2-H y d r o x y -2-p h e n y le t h y la m in o )-1-p h e n y lp r o -

1

syn ) 2:1 (quantitated by H NMR; quantitated and assigned

p a n on e (54a ). To a solution of 2-amino-1-phenylethanol (0.96

g, 7.0 mmol) and N,N-diisopropylethylamine (1.81 g, 14.0

mmol) in THF (30 mL) cooled with an ice bath was added

dropwise a solution of 3-chloropropiophenone (1.20 g, 7.14

mmol) in THF (16 mL). The reaction mixture was stirred at

23 °C for 5 h, at which time TLC indicated that reaction was

complete. A copious white precipitate formed. The mixture

was stored at 5 °C overnight, and the solids were collected

and washed with ether (4 × 30 mL). The solids were

partitioned between 1 N NaOH (30 mL) and CH₂Cl₂(80 mL).

The organic solution was washed (water, brine), dried (Na₂-

SO₄), and concentrated in vacuo to afford 0.63 g (33%) of 54a

as a colorless solid. ¹H NMR δ 2.72 (dd, J ) 9.2, 12.2 Hz, 1

H), 2.96 (dd, J ) 3.6, 12.3 Hz, 1 H), 3.01-3.22 (m, 4 H), 4.72

(dd, J ) 3.3, 9.1 Hz, 1 H), 7.26-7.39 (m, 10 H), 7.48 (dd, J )

7.7 Hz, 2 H), 7.58 (dd, J ) 7.2 Hz, 1 H), 7.96 (d, J ) 7.5 Hz, 2

H). ¹³C NMR δ 38.7, 44.1, 57.2, 71.6, 125.8, 127.5, 128.0, 128.4,

by ¹³C NMR). ¹H NMR δ 1.76-1.98 (m, 2 H), 2.72-3.01 (m, 4

H), 4.78-4.92 (m, 2 H), 7.16-7.42 (m, 10 H). ¹³C NMR δ 36.9

(syn), 37.2 (anti), 47.5 (syn), 47.6 (anti), 56.5 (syn), 56.7 (anti),

72.0 (syn), 72.1 (anti), 74.7 (syn), 74.8 (anti), 125.5, 125.7,

127.1, 127.7, 128.2, 128.4, 142.1, 144.4. CI-MS m/z 272 (MH⁺).

Anal. Calcd for C₁₇H₂₁NO₂‚0.7H₂O: C, 71.91; H, 7.96; N, 4.94.

Found: C, 71.99; H, 7.83; N, 4.98.

3-[(2-Hyd r oxy-2-p h en yleth yl)m eth yla m in o]-1-p h en yl-

p r op a n ol (55b, a n ti/syn ). Reduction of 54b (85 mg, 0.30

mmol) with R-alpine-hydride (2.1 mol equiv, -78 °C, 4.5 h)

and isolation by preparative TLC (CH₂Cl₂/MeOH, 92:8) gave

70 mg (82%) of 55b as a colorless viscous oil; anti/syn ) 3:1

(quantitated by ¹H NMR; quantitated and assigned by ¹³C

NMR). ¹H NMR δ 1.82-1.95 (m, 2 H), 2.39 (syn), 2.40 (anti)

(2 s, 3 H), 2.44-2.88 (m, 4 H), 4.78-4.91 (m, 2 H), 7.26-7.38

(m, 10 H). ¹³C NMR δ 35.1 (syn), 35.2 (anti), 42.0 (syn), 42.4

(anti), 55.7 (syn), 55.9 (anti), 66.0 (anti), 66.2 (syn), 70.3, 73.8

Remote Stereocontrol in Acyclic Substrates

J . Org. Chem., Vol. 63, No. 22, 1998 7981

(syn), 74.4 (anti), 125.5, 125.8, 127.2, 127.5, 128.2, 128.3, 142.1,

144.5. CI-MS m/z 286 (MH⁺). Free base 55b was converted

to an HCl salt in ether by adding 1 N HCl in ether (colorless

solid). Anal. Calcd for C₁₈H₂₃NO₂‚1.0HCl‚0.4H₂O: C, 65.70;

H, 7.60; N, 4.26, Cl, 10.77. Found: C, 65.95; H, 7.47; N, 4.14;

Cl, 10.92.

4 H), 4.51 and 4.80 (br s, 1 H), 4.97 (br s, 1 H), 7.17-7.63 (m,

15 H). ES-MS m/z 376 (MH⁺). Since the complex ¹H NMR

would not permit the assay of stereoisomers, benzamide 57

was reduced to the corresponding amine 35. Thus, tert-

butyldimethylsilyl chloride (55 mg, 0.36 mmol) was added to

a solution of 57 (61 mg, 0.16 mmol), 4-(dimethylamino)pyridine

(15 mg, 0.12 mmol), and triethylamine (0.07 mL, 0.50 mmol)

in CH₂Cl₂(1.2 mL). The reaction mixture was stirred at 23

°C for 18 h and then diluted with CH₂Cl₂(50 mL). The

solution was washed with water (10 mL), dried (Na₂SO₄), and

concentrated in vacuo. The residue was dissolved in THF (0.25

mL), borane-THF (0.4 mL, 0.4 mmol) was added, and the

mixture was refluxed for 6 h. Aqueous K₂CO₃(0.6 mL, 0.24

mmol) was added, and the mixture was refluxed for 1 h. The

mixture was cooled and diluted with ethyl acetate (50 mL).

The solution was washed (water, brine), dried (Na₂SO₄), and

concentrated in vacuo. The residue was dissolved in CH₂Cl₂

(1 mL) and treated with tetrabutylammonium fluoride (0.65

mL, 0.65 mmol). After 1 h, the mixture was diluted with CH₂-

Cl₂and washed with water. After being dried (MgSO₄) and

concentrated in vacuo, the residue was purified by preparative

TLC (CH₂Cl₂/MeOH/NH₄OH, 90:9:1) to afford 30 mg (51%) of

35 as a colorless viscous oil with an anti/syn ratio of 1.5:1 (by

¹H NMR). ¹H NMR was consistent with the structural

assignment and compared favorably with earlier data for anti/

syn mixtures of 35 (vide supra).

3-[(2-Hyd r oxy-2-p h en ylet h yl)(2,4,6-t r im et h ylben zyl)-

a m in o]-1-p h en ylp r op a n ol (55c, a n ti a n d syn ). Reduction

of 54c (80 mg, 0.20 mmol) with R-alpine-hydride (2.1 mol

equiv, CH₂Cl₂, -78 °C, 4.5 h) and isolation by preparative TLC

(hexanes/EtOAc, 4:1) gave 45 mg (56%) of 55c as a colorless

viscous oil; anti/syn ) 22:1 (determined by 600-MHz ¹H NMR).

¹H NMR δ 1.83-1.88 (m, 1 H), 1.99-2.05 (m, 1 H), 2.33 (s, 3

H), 2.40 (syn), 2.42 (anti) (2s, 6 H), 2.64-2.67 (m, 2 H), 2.79

(dd, J ) 10.3, 13.3 Hz, 1 H), 2.93 (m, 1 H), 3.62 and 3.64 (d, J

) 12.6 Hz, anti; d, J ) 12.7 Hz, syn; 1 H), 3.89 and 3.95 (d, J

) 12.7 Hz, syn; d, J ) 12.6 Hz, anti; 1 H), 4.69 (dd, J ) 3.8

and 8.6 Hz, 1 H), 4.79 (d, J ) 8.8 Hz, 1 H), 6.93 (s, 2 H), 7.26-

7.37 (m, 10 H). CI-MS m/z 404 (MH⁺). Free base 55c was

converted to an HCl salt in ether by adding 1 N HCl in ether

(colorless solid). Anal. Calcd for C₂₇H₃₃NO₂‚1.0HCl‚0.4H₂O:

C, 72.51; H, 7.84; N, 3.13; Cl, 7.93. Found: C, 72.76; H, 7.54;

N, 2.93; Cl, 8.06.

N-(2-H yd r oxy-2-p h en ylet h yl)-N-(3-oxo-3-p h en ylp r o-

p yl)ben za m id e (56). To a cooled (0 °C) solution of 54a (500

mg, 1.9 mmol) in CH₂Cl₂(37 mL) was added benzoyl chloride

(300 mg, 1.9 mmol) dropwise. The reaction mixture was

stirred at 0 °C for 1 h and then treated with pyridine (150

mg, 1.9 mmol). After being stirred at 0 °C for 4 h, the mixture

was washed with water (50 mL) and brine (50 mL), dried

(MgSO₄), and concentrated in vacuo. The crude product was

purified by flash column chromatography (hexanes/EtOAc, 2:1)

to afford 550 mg (79%) of 56 as a colorless viscous oil. ¹H NMR

δ 2.51 (br s, 2 H), 3.16-3.60 (m, 3 H), 3.78-3.98 (m, 1 H),

4.70 and 4.90 (br s, 1 H), 7.28-7.62 (m, 12 H), 7.82 (d, J ) 7.3

Hz, 1 H), 8.10 (d, J ) 7.5 Hz, 2 H). ¹³C NMR δ 36.2, 43.3,

56.6, 58.4, 72.1, 72.5, 127.4, 127.9, 128.5, 128.9, 130.0, 130.6,

131.0, 135.0, 138.8, 144.8, 172.9, 199.9, 200.9. ES-MS m/z 374

(MH⁺). Anal. Calcd for C₂₄H₂₃NO₃‚0.4H₂O: C, 75.73; H, 6.30;

N, 3.68. Found: C, 75.98; H, 6.18; N, 3.53.

Ack n ow led gm en t. We thank Prof. Lanny Liebe-

skind for helpful discussions and Paul Lobben for

technical assistance. We thank Gregory Leo for NMR

data and NMR studies on metal ion complexation. We

thank Rekha Shah for several HPLC analyses. We

thank Douglas Alves-Santana, Richard Dunphy, and

J ane Xiang for mass spectral data. Part of this work

was performed in the J anssen Research Foundation,

Spring House, PA.

Su p p or tin g In for m a tion Ava ila ble: Proton NMR spectra

for 24, 35, and 44 with a focus on the N-benzyl proton signals

used to quantitate anti and syn isomers (5 pages). This

material is contained in libraries on microfiche, immediately

follows this article in the microfilm version of the journal, and

can be ordered from the ACS; see any current masthead page

for ordering information.

N-(2-Hyd r oxy-2-p h en yleth yl)-N-(3-h yd r oxy-3-p h en yl-

p r op yl)ben za m id e (57, a n ti/syn ). Reduction of 56 (86 mg,

0.23 mmol) with 2.2 mol equiv of R-alpine-hydride (-78 °C,

CH₂Cl₂, 20 h) and isolation by flash column chromatography

(hexanes/EtOAc, 1:1) gave 61 mg (70%) of 57 as a colorless

viscous oil. ¹H NMR (T ) 353 K) δ 1.89 (br s, 2 H), 3.50 (br s,

J O981341M

Article Doi

Stereocontrol between remote atom centers in acyclic substrates. Anti addition of hydride to 1,5-, 1,6-, and 1,7-hydroxy ketones

DOI: 10.1021/jo981341m

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jo981341m

Full text of DOI:10.1021/jo981341m

Products guided by the article

R&D Labs maybe for 185376-67-4

Relevant to this article

Hot Product