Bernhard
Breit
*,
Peter
Demel
and
Antje
Gebert
Institut für Organische Chemie und Biochemie, Albert-Ludwigs-Universität Freiburg, Albertstr. 21, 79104 Freiburg, Germany. E-mail: bernhard.breit@organik.chemie.uni-freiburg.de; Fax: 0049-761-2038715; Tel: 0049-761-2036051
First published on 11th November 2003
Regioselective and diastereoselective hydroformylation of mono- and 1,3-disubstituted allylic alcohol o-DPPB esters is described. The products represent synthetically important anti-aldol retrons.
We herein report on the first regio- and diastereoselective hydroformylation of 1,3-disubstituted allylic alcohol derivatives 3 (R′ = Me) with the aid of the catalyst-directing o-DPPB group. Extension of this methodology towards hydroformylation of mono-substituted allylic alcohols 3 (R′ = H) is also described.
In the course of the hydroformylation of an allylic alcohol derivative 3 two selectivity issues appear, regio- and diastereoselectivity, which have to be controlled simultanously.
We were pleased to find that hydroformylation of allylic o-DPPB ester 5 (R = Bn) proceeded smoothly at 30 °C and 20 bar syngas pressure with a rhodium catalyst loading of 0.7 mol% to give the anti-aldehyde 6 as the major product in good regio- and diastereoselectivity‡ (Table 1, entry 1). However, the reaction rate was too low. Increasing the catalyst loading to 1.2 mol% reduced the reaction time to 7 d (entry 2). Further improvement was achieved when the pressure was increased to 30 and finally 40 bars (entries 3–5). However, going to higher reaction temperatures resulted in a significantly reduced diastereoselectivity (entry 6) and gave a significant amount of the undesired elimination product 8. Hence, best results were achieved employing 1.8 mol% catalyst loading, 40 bar syngas and 30 °C in toluene for about 44 h (Table 1, entry 5). These conditions allowed a 90% conversion of o-DPPB ester 5 to furnish anti-aldehyde 6 in a regioselectivity (6 : 7) of 90 : 10 and a diastereomer ratio (anti-6 : syn-6) of 98 : 2.
Entry | [Rh]a (mol%) | p CO/H2/bar | T/°C | Time | Conv.b,c (%) | 8 b (%) | rsb (6 : 7) | dr (6)b (anti : syn) |
---|---|---|---|---|---|---|---|---|
a [Rh] = [Rh(CO)2acac]/1.67 P(OPh)3. b Determined from NMR analysis of the crude reaction product. c Chemoselectivity towards aldehyde formation was 100% in all cases. | ||||||||
1 | 0.7 | 20 | 30 | 13 d | 83 | 5 | 87 : 13 | 89 : 11 |
2 | 1.2 | 20 | 30 | 7 d | 90 | 7 | 90 : 10 | 93 : 7 |
3 | 1.2 | 30 | 30 | 4 d | 89 | <2 | 90 : 10 | 98 : 2 |
4 | 2.0 | 30 | 30 | 47 h | 92 | <2 | 89 : 11 | 94 : 6 |
5 | 1.8 | 40 | 30 | 44 h | 90 | <2 | 90 : 10 | 98 : 2 |
6 | 2.0 | 40 | 40 | 46 h | 90 | 10 | 89 : 11 | 90 : 10 |
7 | 2.0 | 40 | 20 | 48 h | 88 | <2 | 90 : 10 | 98 : 2 |
With these optimized conditions in hand we looked at the dependence of the selectivity parameters as a function of the nature of the substituent R at the controlling stereogenic center. Thus, excellent diastereoselectivity was obtained for primary alkyl substituents (Table 2, entries 1, 2, 4). Somewhat reduced diastereoselectivities were noted for secondary alkyl substituents (entries 5, 6), whereas regioselectivity stays in most cases in the order of 9 : 1. However, for an isopropyl-substituted derivative the best regioselectivity (98 : 2, entry 5) was observed.
Entry | Major product | Conv.a,b (%) | 8 a (%) | rsa (6 : 7) | dr (6)a (anti : syn) |
---|---|---|---|---|---|
a Determined from NMR analysis of the crude reaction product. b Chemoselectivity towards aldehyde formation was 100% in all cases. | |||||
1 |
![]() |
90 | 3 | 91 : 9 | 95 : 5 |
2 |
![]() |
90 | <2 | 90 : 10 | 98 : 2 |
3 | from cis-5 (R = Bn) | 40 | 39 | 61 : 39 | 58 : 42 |
4 |
![]() |
90 | 5 | 92 : 8 | 92 : 8 |
5 |
![]() |
91 | 3 | 98 : 2 | 87 : 13 |
6 |
![]() |
95 | 7 | 90 : 10 | 91 : 9 |
In order to learn about the influence of double bond geometry on the regio- and stereochemical outcome of the title reaction we studied hydroformylation of cis-configured allylic-o-DPPB ester 5 (R = Bn). Surprisingly, chemo-, regio- and diastereoselectivity were significantly lower compared to the trans-alkenic system 5 (see Table 2, entries 2 and 3).
Interestingly, when monosubstituted allylic o-DPPB esters were used a similar regio- and stereodirecting effect of the catalyst-directing o-DPPB group was observed (Scheme 1, Table 3). Thus, regioselectivities up to 86 : 14 (10 : 11) and diastereoselectivities‡ up to 95 : 5 (anti-10 : syn-10) were found (Table 3) with formation of the anti-aldol retron 10 as the major product. Interestingly, aldehyde 10 represents a potentially valuable building block for the synthesis of polypropionates.
![]() | ||
Scheme 1 Conditions: (i) 1.8 mol% [Rh(CO)2acac], 3 mol% P(OPh)3, H2/CO (1 : 1) 40 bar, toluene, 30 °C, 46 h. |
In conclusion, o-DPPB-directed regio- and diastereoselective hydroformylation of 1,3-disubstituted and mono-substituted allylic o-DPPB esters 3 could be achieved. This methodology gives access to the anti-aldol retron which is difficult to reach through aldol chemistry directly. Thus, directed hydroformylation may become a synthetically attractive alternative to established aldol7 or allylmetal8 chemistry for the construction of anti-aldol retron type products.
We thank DFG, the Fonds of the Chemical Industry and the Krupp Foundation (Krupp Award for young university teachers to BB) for financial support.
Footnotes |
† Electronic supplementary information (ESI) available: experimental. See http://www.rsc.org/suppdata/cc/b3/b311378g/ |
‡ The anti-stereochemical relation of 6
(R = Bn) and 10
(R =
i-Pr) could be determined upon chemical transformation into the benzylidene acetals 13 and 14, respectively, employing the following reaction sequence: i LiAlH4, ether, 0 °C; ii PhCH(OMe)2, TsOH cat., CH2Cl2, rt. Inspection of coupling constants as well as NOESY data allowed assignment of relative configuration. |
This journal is © The Royal Society of Chemistry 2004 |