J. Am. Chem. Soc.

1984, 106, 5031-5033 503 1

Table I N-(L-aminoacy1)diaminoethanesstabilized RNA double helices

to a greater extent than did the D-amino acid derivatives, while
ee of L
yield, %” isomer, %’
the D isomers afforded single-stranded polymers greater protection
against hydrolysis by ribonuc1eases.l6 Similarly, L-lysyl-L-amino
polymer 1O:lC 5:lC 1O:lC 5:Ic acid dipeptides stabilized double-helical complexes more than did
the corresponding L-lysyl-D-amino acid diastereomer^.'^ These
noncovalent interactions appeared to depend upon the presence
of a diammonium salt, since there was little discrimination when
the nucleic acid interacted with the D- or L-amino acid alone.
a Yields are reported as the percentage of mononucleotide units acy- Some details of the reaction reported here argue against its
lated by the imidazolide of N-(3,5-dinitrobenzoyl)-D~-alanine.’The direct involvement in the early evolution of the chirality of nu-
enantiomeric excess (ee) of the polymer-bound N-(3,5-dinitro- cleosides and amino acids. In particular, the presence of a N-
benzoy1)alanine is defined as the % L isomer minus the % D isomer.
Uncertainties are 2~3%. ‘The ratio of reactants, (imidazolide):(nu-
protecting group precludes its participation in a recursive mech-
cleotide units). anism for peptide-bond formation. Moreover, we have observed
that when the activated amino acid carries no N-protecting group,
dient.s The amino acid derivative was detected by its absorbance the selectivity is reversed, and it is the Pamino acid that is selected
at 254 nm, and the acylation yield was determined from the by D p~lynucleotides.~*’~ The structural features responsible for
chromatographic peak area using a molar absorptivity (254 nm) these stereochemical results are now under investigation.
of 14.6 X lo3 M-l cm-l (measured in 0.1 M NH40Ac, pH 4.7,
containing 20% acetonitrile). The enantiomers of the purified Acknowledgment. We thank Professor W. H. Pirkle for the
dinitrobenzoylalanine were separated by HPLC on a chiral sta- gift of a chiral HPLC column and the National Aeronautics and
tionary phase consisting of one enantiomer of l-amino-1-(6,7- Space Administration (Grant NAGW-493) for financial support.
dirnethylnaphth- l-yl)-2-methylpropane linked to a 5 - ~ msilica Registry No. Poly(A), 24937-83-5;poly(I), 30918-54-8;poly(U),
matrix.6 The derivatives were eluted with a mobile phase of 4:l 27416-86-0;poly(C), 3081 1-80-4;imidazolide of N-(3,5-dinitro-
waterlmethanol containing 1 g/L K H C 0 3 and 20 g/L K2S04 benzoyl)-DL-alanine, 91 157-76-5.
and were detected by their absorbance at 254 nm. Relative
amounts of the D and L enantiomers were determined from the (15) (a) Gabbay, E.; Kleinman, R. J . Am. Chem. SOC. 1967, 89,
chromatographic peak areas. Table I shows the percent of nu- 7123-7125. (b) Gabbay, E. J.; Kleinman, R.; Shimshak, R. R. Ibid. 1968,
cleoside hydroxyl groups that were acylated and the ee of the L 90, 1927-1928. (c) Gabbay, E. J. Ibid. 5257-5263. (d) Gabbay, E. J.;
isomer that was incorporated into each polymer. The reactant Kleinman, R. W. Biochem. J . 1970, 117, 247-256.
(16)Gabbay, E. J.; Kleinman, R.; Shimshak, R. R. Biopolymers 1968, 6,
N-( 3,5-dinitrobenzoyl)-~~-alanine was shown to be racemic (ee 993-996.
*
= 0 3%) by HPLC on the chiral stationary phase. (17) Usher, D. A,; Profy, A. T.; Needels, M. C., unpublished results.
Previously, polynucleotides have been acylated by the imida-
zolides of N-acetyl-L-amino acids and L-amino acid trifluoro-
acetate^,^-^ but in these experiments, no attempt was made to Very High 1,2-Asyrnrnetric Induction in the Reaction of
compare the behavior of D- and L-amino acid derivatives. We Allyl-9-BBN with Certain Imines. Evidence for a
recently reported that aminoacylation of the “internaln 2’-hydroxyl Stereoelectronic Effect To Enhance the Cram
group of a dinucleoside monophosphate (D-inosinylyl-D-inosine) Selectivity
by the imidazolide of racemic N-(tert-butoxycarbony1)alanine
resulted in the formation of excess L ester.* Since aminoacyl Yoshinori Yamamoto,* Toshiaki Komatsu, and
imidazoles are believed to acylate polyribonucleotides principally Kazuhiro Maruyama
at the 2’-hydroxyl groups along the ribose b a ~ k b o n e , ~the~’
preferential incorporation of (dinitrobenzoy1)-L-alanineinto the Department of Chemistry, Faculty of Science
polymers is consistent with the earlier work. The stereoselectivity Kyoto University, Kyoto 606, Japan
appears to be somewhat lower for the pyrimidine polymers than
for the purine polymers, possibly reflecting the smaller amount Received April 18, 1984
of secondary structure in the former. The overall yields of esters The discovery of new methods for 1,2- and 1,3-asymmetric
(Table I) were consistent with the reported relative rates of induction in acyclic systems has been of keen interest in synthetic
polymer acylation by N-acetylamino acid imidazolides, which and theoretical organic chemistry.l Especially, the Cram/
decreased in the order poly(U) > poly(A) > p ~ l y ( C ) . ~Ste--~ anti-Cram problem has been one of the longstanding concerns,
reoselective diastereomer formation by amino acids has been and unfortunately the Cram selectivity of ordinary aldehydes
reported previously,“’-’4 but there appears to be no prior case having an a-chiral center is generally no so high (eq l).2-4 We
involving the aminoacylation of a nucleic acid.
have discovered that the Cram selectivity is remarkably enhanced
Derivatives of amino acids have been shown to interact ste-
in the reaction of imines with allyl-9-BBN (eq 2). This finding
reoselectively with polynucleotides. Gabbay et aI.l5 found that
and elucidation of the enhancement provide a conceptual advance
in the Cram/anti-Cram problem.
( 5 ) Profy, A. T.;Lo, K.-M.; Usher, D. A. Nucleic Acids Res. 1983, 11,
1617-1632. The low Cram selectivity in the reaction of allyllic organo-
(6)The column and the composition of the eluant were provided by Pro-
fessor w. H. Pirkle.
(7) White, W. E., Jr.; Lacey, J. C., Jr.; Weber, A. L. Biochem. Biophys. (1) For example: Heathcock, C. H. In “Comprehensive Carbanion
Res. Commun. 1973, 51, 283-291. Chemistry”; Durst, T., Buncel, E., Eds.; Elsevier: New York, 1984. Evans,
(8) Weber, A. L.; Fox, S. W. Biochim. Biophys. Acta 1973,319, 174-187. D. A,; Nelson, J. V.; Taber, T. R. “Topics in Stereochemistry”; Wiley-In-
(9) Weber, A. L.; Lacey, J. C., Jr. J. Mol. Euol. 1975, 6, 309-320. terscience: New York, 1982;Vol 13, p 1. “Asymmetric Synthesis”; Morrison,
(10) Herlinger, H.; Kleimann, H.; Ugi, I. Justus Liebigs Ann. Chem. 1967, J. D., Ed.; Academic Press: New York, 1983; Vol 2. Masamune, S. In
706, 37-46. ‘Organic Syntheses Today and Tomorrow”; Trost, B. M., Hutchinson, R.,
(1 1) Cervinka, 0.; Budilova, J. Collect. Czech. Chem. Commun. 1967, 32, Eds.; Pergamon Press: New York, 1981;p 197. Hoffmann, R. W. Angew.
2383-2386. Chem., Int. Ed. Engl. 1982, 21, 5 5 5 . Yamamoto, Y . ;Maruyama, K. Het-
(12) (a) Doty, P.; Lundberg, R. D. J . Am. Chem. SOC. 1956, 78, erocycles 1982, 18, 357.
4810-4812. (b) Matsuura, K.; Inoue, S.;Tsuruta, T. Mukromol. Chem. 1965, (2) Excellent l,2-and 1.3-asymmetric induction through chelation control
85, 284-286. (c) Tsuruta, T.; Inoue, S.; Matsuura, K. Biopolymers 1967, 5, has been realized in the nucleophilic addition of organometallics to a- and
31 3-319. @-alkoxysubstituted aldehydes. Reetz, M. T.; Kesseler, K.; Schmidtberger,
(13) Blair, N. E.; Bonner, W. A. Origins Life 1980, 10, 255-263. S.; Wenderoth, B.; Steinbach, R. Angew. Chem., Int. Ed. Engl. 1983, 22, 989.
(14) (a) Imanishi, Y.; Ohnishi, H.; Hashimoto, Y.Inr. 1. Biol. Macromol. Reetz, M. T.;Jung, A. J . Am. Chem. SOC.1983, 105, 4833. Still, W. C.;
1981, 3, 97-104. (b) Hashimoto, Y.; Imanishi, Y. Biopolymers 1981, 20, Schneider, J. A. Tetrahedron Lett. 1980, 21, 1035. Still, W. C.; McDonald,
507-524. J. H. Ibid. 1980, 21, 1031.

0002-786318411506-5031$01.50/0 0 1984 American Chemical Society

 5032 J. Am. Chem. SOC.,Vol. 106, No. 17, 1984 Communications to the Editor

Ph x C/H tw’-
L
. H L
I01 M=Li,Mg.B. OH
Si,... 1 3 - 4

Ph A/w E (1) H I
R’
OH 8
2
-2:1

H Me
AR‘ Ph antl-Cram addllion
-------*
NH 11
N
R‘/ 10
‘R’
3 and L may be conceivable, though it is a sort of 1,2-axial-
equatorial interaction. Therefore, we examined the reaction of
12 with allyl-9-BBN. The Cram product (13) was obtained very
I- Pr i-

4
up t o 100~0

metallic compounds with ordinary chiral aldehydes having no 13

ability to be chelated is presumably due to the fact that the
selectivity is determined only by the conventional steric factor a t
the chiral center. Since the a-chiral center goes to the equatorial
position of 5, the steric influence of ligand (L) does not reach the
L
I 14
predominantly; 13:14 = 9 2 9 (eq 3).1° Here again, the confor-
mation of the chiral center is probably fixed as shown in 15. The
L L
I R I
Me 6
5
M=B, A I , Mg. Z n , Li, Ti, Z r , ,..

chiral center. If the chiral center goes to the axial position, the 16
selectivity must depend upon both the original steric factor of the 15
chiral center and the steric influence of L. We have disclosed allyl group attacks from the less hindered side (11 and 15). The
that the reaction of imines with allylic organometallic compounds transition state (16) is destabilized due to the steric repulsion
proceeds through such a transition state (6), presumably due to between the methyl group (and/or phenyl group) and L.“
the stereoelectronic effect of the imine group.5 Therefore, we Normally, the 1,3-diaxial interaction is more stronger than the
examined the reaction of imines having a-chiral centers with allylic 1,2-axial-equatorial interaction. Accordingly, the selectivity of
organometallics. The results are summarized in Table I. entry 2 is 10076, while that of 12 is 92%.’*
Even allylmagnesium chloride produced fairly good Cram se-
lectivity in the imine reaction (entry 3 vs. 6). Allyl-9-BBN gave (6) The structure was determined as follows. The Cram product 1 was
converted into the anti-Cram amine 7. Similarly, the anti-Cram isomer 2
the Cram product6 either exclusively or very predominantly (entries
1 and 2). This remarkable enhancement may be explained as
follows. The conformation of the chiral center is fixed as shown
in 8 owing to the steric influence of L, and the allyl group attacks
from the less hindered side (8 and 10). The transition state, 9, la1 H2/Pd-C,’ ibl 1-mthyl-2-fluoro dl
is highly destabilized owing to the steric repulsion between the pyridinium soif/Ei,N.8 (cl LiN3. 7
methyl group and the 9-BBN ring proton^.^ Id1 LiAlH,, lei R’-C-R2/MgS04 ( f l LiAlH4
The steric interaction between R group and L (6)is a sort of
1,3-diaxial interaction. A similar interaction between the R’ group 1
was transformed into the Cram amine. The reduction products of the ho-
(3) Midland has reported that the reduction of a-phenylpropionaldehyde moallylamines were compared with authentic samples thus obtained.
with LiBH(sec-Bu), proceeds with an exceptionally high Cram selectivity. (7).The double bond must be reduced prior to the displacement with N3-.
The direct reaction of 1 was accompanied with significant amounts of the
Midland, M. M.; Kwon, Y. C. J. Am. Chem. SOC.1983, 105, 3725. 8-elimination product.
(4) Heathcock has found an excellent Cram selectivity in the Lewis acid (8) Hojo, K.; Kobayashi, S.;Soai, K.; Ikeda, S.; Mukaiyama, T. Chem.
mediated reaction of silyl ketene acetals with ordinary aldehydes. Heathcock, k r r . mi,635.
C. H.; Flippin, L. A. J. Am. Chem. SOC. 1983,105, 1667. (9) The bridgehead proton (‘H”) of 9-BBN ring covers the six-membered
(5) Yamamoto, Y.; Komatsu, T.; Maruyama, K., paper presented at the ring. The plane of the C-R bond of imine groups and the plane of the C-“H”
Fifth International Symposium on Boron Chemistry, Swansea, Wales, July bond intersects with nearly orthogonal angle. See also: Yamamoto, Y.;
1983, and at the Eighth International Symposium on Synthesis in Organic Yatagai, H.; Maruyama, K. J. Am. Chem. SOC.1981, 103, 3229.
Chemistry, Cambridge, England, July 1983. The manuscript will be submitted (10) The structure of 13 and 14 was determined as described in ref 6.
shortly. The reaction of crotyl-9-BBN with certain imines provides the erythro (4S)-5-Methyl-4-hexanol,prepared by the literature procedure (Hoffmann,
homoallylamines predominantly, excluding the possibility of a boat transition R. W.; Herold, T. Chem. Eer. 1981, 114, 375), was converted into (4R)-5-
state. The coordination of boron to the nitrogen is essential to induce the methyl-4-hexylamine. Reduction of 13 with Pd(OH), on carbon gave the
reaction, which indicates no participation of an open transition state. corresponding (4R)-5-methyl-4-hexylamine.
 J. Am. Chem. SOC.1984,106, 5033-5034 5033

Table 1. High Cram Selectivity of Imine Reactions' electrochemical reduction of .COzproceeds with large overpo-
tentials2 and generally yields formate. Electrocatalytic reductions
imine, RCH=NR' allylorganometal Cram (3): of COz to yield formate have been observed with supported Pd,
entry R R' M anti-Cram (4) as well as through the coupling of formate dehydrogenase with
1 PhCH(CH3) n-Pr 9-BBN 96:4 methyl~iologen.~ Our efforts have focussed upon utilizing
2 PhCH(CH3) i-Pr 9-BBN 1oo:o transition-metal complexes to promote reduction of C 0 2 to CO
3 PhCH(CH,) n-Pr MgCl 84:16 via (1).

entry
4 PhCH(CH3)

aldehyde
i-Pr MgCl

allylorganometal
70:30
Cram (1):
anti-Cram (2)
CO,(g) + 2 6 + 2H+(aq) - CO(g) + HzO(l) (1)

5 PhCH(CH3)CHO 9-BBN 55:45 CO(PC)~ was deposited onto pyrolytic graphite or carbon cloth
6 MgCl 60:40 surfaces either by adsorption from THF/Co(Pc) solutions or by
7 SiMelb 70:30 droplet evaporation of THF/Co(Pc) solutions. Controlled po-
"All reactions were carried out on a 1-mmol scale at -78 OC under
tential electrolysis of such modified carbon cloth electrodes at -1.0
N 2 and quenched at 0 OC. Total isolated yields were in a range of V vs. SSCE in aqueous solution (pH 5.0, 0.05 M citrate buffer,
88-98%. The Cramanti-Cram ratio was determined by 'H N M R Eo'(COZ/CO) = -0.65 V vs. SCE5) under 1 atm of COz(g)
analysis and/or GLPC (THEED, lo%, 2 m). bTiC14 was useed as a produced CO(g) as the major carbon-containing species. The
Lewis acid. catalytic nature of the reaction has been confirmed by formation
of over 105 molar equiv of C O per molar equiv of electroactive
The present findings suggest that the Cram/anti-Cram problem catalyst (Table I).
of carbonyl groups might be solved by a similar approach." Typical coulometric experiments (Table I) for potentials from
Further work along this line is now under active investigation. -0.95 to -1.2 V vs. SCE indicate that 5 5 4 0 % of the charge passed
can be accounted for as C O formation and 35-30% detected as
(1 1) There may be a question that the bad interactions depicted in 9 and Hzrimplying overall coulometric efficiencies of 90-95% for the
16 can be avoided by rotating the carbon so that the hydrogen is in the position catalytic reaction of Co(Pc) with COz/HzO solutions. Although
of the methyl in 9 and the phenyl in 16. Inspection with a Dreiding model spot tests indicate the presence of oxalate and formate, as pre-
clearly indicates that such conformations are destabilized by the steric re-
pulsion between the 9-BBN ring and the phenyl group in 9 and between the viously reported for a similar Co(Pc)/graphite system at more
9-BBN ring and the methyl group in 16. cathodic operating potentials: we observe that these species are
(12) (a) For addition of allylboronatesto Schiff bases, see: Hoffmann, R. present in only trace amounts, and that the major carbon-con-
W.; Eichler, G.; Endesfelder, A. Liebigs Ann. Chem. 1983,2000. (b) When taining product is gaseous CO.
alkylorganometallics,such as BuCu-BF, and Bu2CuLi.BF,, were utilized, the
Cram/anti-Cram selectivity was low (-4:l). This is reasonable since the Neither of the first two reported reduction potentials for Co(Pc)
six-membered cyclic transition state is not involved in this reaction. For the in D M F s ~ l u t i o n-0.40
, ~ and -1.40 V vs. SCE, correspond with
reaction of imines with RCuaBF,, see: Wada, M.; Sakurai, Y . ;Akiba, K . the potentials a t which we detect the onset of CO(g) production
Tetrahedron Lett. 1984, 25, 1079. in aqueous media (-0.9 V vs. SCE). Furthermore, cyclic volt-
(13) An oxonium salt of aldehydes may take a trans geometry (17). If
ammograms for the Co(Pc)O/- couple are found to be identical
under 1 atm of COz or 1 atm of Ar for THF/Co(Pc) solutions
as well as for C/Co(Pc) surfaces in aqueous media. This evidence
17
seems to preclude initial binding of COz to Co(Pc)- as a viable
pathway unless there is only an extremely weak Co(Pc)-COz
so, the R group may go to the axial position as described above. In fact, the interaction.
reaction of a-phenylpropionaldehyde with allyl-9-BBN in the prwence of The aqueous reduction of Co(Pc)/graphite surfaces in the
Et30*BF4- produced 1 and 2 in a ratio of 7:3 (cv. entry 5). We are also
investigating the Lewis acid mediated reaction of acetals bearing an a-chiral absence of COz yields two proton-coupled reductions which appear
center, the results of which will be published soon. at -0.58 and -0.95 V vs. S C E at p H 5.0. Over a range of pH
1.5-5.5 we observe a positive shift of Eo'for th: first reduction
wave of 59 mV/pH unit. The second reduction wave is quasi-
Catalytic Reduction of C 0 2 at Carbon Electrodes reversible, and scan rates of 5 V/s yield reversible behavior for
this couple. Interestingly, association of the first electrochemical
Modified with Cobalt Phthalocyanine wave with the one-electron Co(Pc)O/- couple implies that the
Charles M. Lieber and Nathan S. Lewis*
Department of Chemistry, Stanford University (2) (a) Russell, P. G.; Kovac, N.; Srinivasan, S.; Steiiberg, M. J . Elec-
Stanford, California 94305 rrochem. Soc. 1977, 124, 1329-1338 and references cited therein. (b) Eggins,
B. R.; McNeill, J. J . Electrounul. Chem. 1983, 148, 17-24. (c) Canfield, D.;
Received April 20, 1984 Frese, K. W., Jr. J. Electrochem. SOC.1983, 130, 1772-1773.
(3) (a) Stadler, C. J.; Chao, S.;Summers, D. P.; Wrigliton, M. S.J . Am.
We report the electrocatalytic reduction of aqueous solutions Chem. SOC.1983, 105, 6318-6320. (b) Ibid. 1984, 106, 2723-2725. (c)
Stadler, C. J.; Chao, S.;Wrighton, M. S. Ibid., 1984, 106, 3673-3675. (d)
of COz(g) to CO(g). The reaction occurs on carbon electrodes Parkinson, B. A.; Weaver, P. Nature (London), in press.
modified by adsorption of cobalt phthalocyanine, Co(Pc). The (4) Co(Pc) was obtained from Eastman Kodak Co. and was purified by
COz reduction can be achieved within 300 mV of the thermo- sublimation. C02(g) concentrations were monitored with an Orion Model
dynamic COz/CO redox potential, and essentially the only car- 95-02 electrode; CO(g) and H2(g) analyses were performed using a Carle
Model 197-B gas chromatograph in the standard factmxy configuration
bon-containing product is CO(g). In contrast, Co(Pc) dissolved (columns and thermistor at 60 OC, hydrogen-transfer catalyst at 570 OC;
in homogeneous solution yields poor stability and low catalytic retention times, H2, 1.5 min, CO, 10.7 min). Oxalate and formate were
efficiency for COz activation. Thus, in addition to an energy- detected using standard qualitative spot test reagents,6a and quantitative
efficient activation of COz(g), this system demonstrates the ef- determination of oxalate was performed polarographically by determination
of Eu'+. Cyclic voltammetric data (100-500 mV/s) were used to determine
fectiveness of chemical modification of electrodes in suppressing the coverage of electroactivecatalyst on the electrode surface; such coverages
deleterious decomposition pathways during electrocatalysis. ranged from 4 X lo-" to 40 X IO-" mol/cm2 and were generally somewhat
The abundance of COz as a one-carbon precursor has evoked less than the total amount of catalyst deposited by the droplet evaporation
considerable interest in its catalytic transformations.' Direct technique.
(5) Randin, J. P. 'Encyclopedia of Electrochemistry of the Elements";
Bard, A. J., Ed.; Marcel Dekker: New York, 1976; Vol. VIII, p 172.
(1) (a) Eisenberg, R.; Hendricksen. D. E. Ado. Cutal. 1979,28, 79-172. (6) (a) Meshitsuka, S.; Ichikawa, M.; Tamaru, K. J. Chem. SOC.,Chem.
(b) Kolomnikov, I. S.;Grigoryan, M. Kh. Russ. Chem. Reu. (Engl. 7'run.d.) Commun. 1974,158-159. (b) Takahashi, K.; Hiratsuka, K.; Sasaki, H. Chem.
1978, 47, 334-353. (c) Denise, B.; Sneeden, R. P. A. CHEMTECH 1982, Lett. 1979, 305-308. (c) Ibid. 1977, 1137-1 140.
12, 108-112. (d) Darensbourg, D. J.; Kudaroshi, R. A. Adu. Organomet. (7) Clack, D. W.; Hush, N. S.;Woolsey, I. S. Inorg. Chim. Acta 1976, 19,
Chem. 1983, 22, 129-168. 129-132.

0002-7863/84/ 1506-5033$01.50/0 0 1984 American Chemical Society