Asymmetric autocatalysis triggered by carbon isotope (13 C/12C) chirality

Article abstract of DOI:10.1126/science.1170322

Many apparently achiral organic molecules on Earth may be chiral because of random substitution of the 1.11% naturally abundant 13C for 12C in an enantiotopic moiety within the structure. However, chirality from this source is experimentally difficult to discern because of the very small difference between 13^C and 12^C. We have demonstrated that this small difference can be amplified to an easily seen experimental outcome using asymmetric autocatalysis. In the reaction between pyrimidine-5-carbaldehyde and diisopropylzinc, addition of chiral molecules in large enantiomeric excess that are, however, chiral only by virtue of isotope substitution causes a slight enantiomeric excess in the zinc alkoxide of the produced pyrimidyl alkanol. Asymmetric autocatalysis then leads to pyrimidyl alcohol with a large enantiomeric excess. The sense of enantiomeric excess of the product alcohol varies consistently with the sense of the excess enantiomer of the carbon isotopically chiral compound.

Full text of DOI:10.1126/science.1170322

REPORTS
24. C. A. Wilson, R. K. Grubbs, S. M. George, Chem. Mater.
36. A. H. Simmons, C. A. Michal, L. W. Jelinski, Science 271,
84 (1996).
37. Y. Termonia, Macromolecules 27, 7378 (1994).
38. N. Du et al., Biophys. J. 91, 4528 (2006).
39. T. Lefèvre, M-.E. Rousseau, M. Pézolet. Biophys. J. 92,
2885 (2007).
TEM (EFTEM), and time-of-flight secondary ion mass
spectroscopy (TOF-SIMS) data. We also thank J. Woltersdorf,
M. Becker, and J. Liu from the Max Planck Institute of
Microstructure Physics in Halle, Germany, for EFTEM and
Raman measurements. We appreciate TOF-SIMS
measurement by M. Krause and help and support in
arranging the mechanical properties measurements and
TOF-SIMS measurement by M. Petzold from the
Fraunhofer Institute for Mechanics of Materials in Halle,
Germany. This work has received financial support from
the Bundesministerium für Bildung und Forschung
(grant FKZ 03X5507) and Deutsche Forschungsgemeinschaft
(grants BR 3370/3-1 and Mi 390/17-1).
17, 5625 (2005).
25. A. Fernández, L. R. Scott, Phys. Rev. Lett. 91, 018102 (2003).
26. F. Cordier, S. Grzesiek, J. Mol. Biol. 317, 739 (2002).
27. F. R. N. Gurd, P. E. Wilcox, Adv. Protein Chem. 11, 311
(1956).
28. F. H. Arnold, B. L. Haymore, Science 252, 1796 (1991).
29. J. T. Kellis, R. J. Todd, F. H. Arnold, Nat. Biotechnol. 9,
994 (1991).
40. Y. Liu, Z. Shao, F. Vollrath, Nat. Mater. 4, 901 (2005).
41. A. L. Let, D. E. Mainwaring, C. Rix, P. Murugaraj,
J. Non-Cryst. Solids 354, 1801 (2008).
30. K. H. Gustavson, Nature 182, 1125 (1958).
31. K. S. J. Rao, G. V. Rao, Mol. Cell. Biochem. 137, 61 (1994).
32. H. Sun, H. Li, R. A. Weir, P. J. Sadler, Angew. Chem. Int.
Ed. 37, 1577 (1998).
42. Z. Shao, R. J. Young, F. Vollrath, Int. J. Biol. Macromol.
24, 295 (1999).
43. R. E. Fornes, R. W. Work, N. Morosoff, J. Polym. Sci.
Polym. Phys. Ed. 21, 1163 (1983).
33. T. Handel, W. F. DeGrado, J. Am. Chem. Soc. 112, 6710
(1990).
44. G. V. Guinea, M. Elices, J. Pérez-Rigueiro, G. R. Plaza,
J. Exp. Biol. 208, 25 (2005).
Supporting Online Material
34. Generally, in our ALD process, ZnO deposition has shown
relatively poor self-limiting features. At longer exposure
times, the process shows parasitic chemical vapor
deposition (high film thickness with relatively poor
uniformity). Therefore, in case of ZnO, an exposure of
only 5 s was applied.
35. It has been observed that in the case of the
metal-infiltrated silks, the contraction ratio produced
by humidity up to 100% or temperature up to 70°C
is in an almost negligible range (<3%).
45. D. Porter, F. Vollrath, Adv. Mater. 21, 487 (2009).
46. We thank W. Lee from the Korea Research Institute of
Standards and Science in Daejeon, Korea; H. Han from
the Department of Material Science and Engineering,
Pohang University of Science and Technology in Pohang,
Korea; and G.-M. Kim from the Department of Physics,
Martin-Luther-University Halle-Wittenberg in Merseburg,
Germany, for their helpful discussions on physical and
chemical interpretations of mechanical properties,
WAXS, NMR spectra, Raman spectra, energy-filtered
www.sciencemag.org/cgi/content/full/324/5926/488/DC1
Materials and Methods
SOM Text
Figs. S1 to S11
Tables S1 to S4
References
6 November 2008; accepted 11 March 2009
10.1126/science.1168162
solely by virtue of carbon isotope substitution
present a challenging problem. To the best of our
knowledge, there have been no reports in which
carbon isotopically substituted chiral compound in-
duces enantioenrichment in an asymmetric reaction.
Here, we show that asymmetric autocatalysis
with amplification of enantioenrichment (16, 17)
can be successfully applied to the discrimination
between enantiomers that differ solely in substitu-
tion of carbon isotopes. Asymmetric autocatal-
ysis has enormous power to recognize molecular
chirality (18–22). In the reaction between pyrimidine-
5-carbaldehyde 4 and diisopropylzinc (i-Pr2Zn), the
zinc alkoxides of the isotopically substituted chiral
alcohols 1 to 3 cause a slight enantiomeric excess
(ee) in the zinc alkoxide of pyrimidyl alkanol 5.
After autocatalytic amplification of this enantio-
meric excess, alkanol 5 accumulates an enhanced
ee with a sense that varies consistently with that
of the abundant enantiomer of the isotopically
substituted chiral compound 1, 2, or 3 (Fig. 2).
The isotopically chiral compounds 1 to 3 were
synthesized as shown in Fig. 3 (23). First, methyl-
Asymmetric Autocatalysis Triggered by
Carbon Isotope (13C/12C) Chirality
Tsuneomi Kawasaki, Yukari Matsumura, Takashi Tsutsumi, Kenta Suzuki,
Masateru Ito, Kenso Soai*
Many apparently achiral organic molecules on Earth may be chiral because of random
substitution of the 1.11% naturally abundant 13C for 12C in an enantiotopic moiety within the
structure. However, chirality from this source is experimentally difficult to discern because of the
very small difference between 13C and 12C. We have demonstrated that this small difference can be
amplified to an easily seen experimental outcome using asymmetric autocatalysis. In the reaction
between pyrimidine-5-carbaldehyde and diisopropylzinc, addition of chiral molecules in large
enantiomeric excess that are, however, chiral only by virtue of isotope substitution causes a slight
enantiomeric excess in the zinc alkoxide of the produced pyrimidyl alkanol. Asymmetric
autocatalysis then leads to pyrimidyl alcohol with a large enantiomeric excess. The sense of
enantiomeric excess of the product alcohol varies consistently with the sense of the excess
enantiomer of the carbon isotopically chiral compound.
he limits of knowledge in recognizing groups, although the enantiomeric excess arising
and understanding chirality are continu- from this source may be minute (Fig. 1).
T
ally pushed back by innovative methods
There are reports on enantioselective reactions 13C-methylphenyl methanol 1 was prepared. To
(1–4). Although the theory of a tetrahedral asym- (11, 12) or on a kinetic resolution (13) induced by restrict chiral effects to the influence of the car-
metric carbon atom with four differing substitu- hydrogen isotope (H/D) replacement, and the use bon isotopic chirality, both enantiomers of 1 were
ents has been understood since the 19th-century of hydrogen isotopes to explore enzymatic mech- prepared from the same chiral ligand (–)-8 as
work of van’t Hoff and LeBel (5–9), molecules anisms is well known (14). In a polymer, isotope shown in Fig. 3, route I. The enantiomers could
for which this characteristic arises solely from chirality (H/D) has been shown to cooperatively be synthesized by direct asymmetric dimethyl-
carbon isotopic substitution are rare (10). It is not control macromolecular helical handedness (15). zinc addition to acetophenone using sulfonamide
known whether a reactive difference between en- However, the effect of carbon isotope substitution (–)-8 (24) as a chiral ligand. (R)-Alcohol 1 was
antiomers can arise solely from a difference in is expected to be far smaller. Thus, enantioselective synthesized using 13C-labeled acetophenone 6 as
carbon isotopic substitution. Such an observation reactions induced by molecules that are chiral the substrate. On the other hand, (S)-1 was
has been experimentally inaccessible. Considering
that the natural abundance of 13C is ~1.11% on
Earth, many organic molecules may be chiral aris-
ing from carbon isotope differences in enantiotopic
Department of Applied Chemistry, Tokyo University of Science,
Kagurazaka, Shinjuku-ku, Tokyo 162-8601, Japan.
*To whom correspondence should be addressed. E-mail:
soai@rs.kagu.tus.ac.jp
Fig. 1. Generation of chirality by carbon isotope substitution.
492
24 APRIL 2009 VOL 324 SCIENCE www.sciencemag.org

REPORTS
prepared from the 13C-labeled dimethylzinc and Finally, phenyl-1,2,3,4,5,6-13C6-phenylmethanol 3 trifluoromethylphenylacetic acid (MTPA). Be-
unlabeled acetophenone 7 using the same chiral was synthesized by a diphenyl(1-methylpyrrolidin- cause the chemical shift differences for the pair of
ligand (–)-8. In route II, asymmetric centers of 2-yl)methanol (DPMPM)–catalyzed asymmetric normal and 13C-labeled groups can be observed,
tert-alcohol 1 were introduced by asymmetric aryl transfer reaction (27, 28) to the13C-substituted the integration of these peaks shows the diaste-
epoxidation (25) to the 13C-labeled allyl alcohol 9, benzaldehyde 12 to afford 13 with a chiral stereo- reomeric excess of these MTPA esters, that is, the
followed by reduction of the tosylate and the genic center. Reductive removal of the bromide enantiomeric excess of the synthesized carbon
epoxide, which afforded the enantiomers of the substituent gave the isotopically substituted en- isotopic chiral alcohols.
desired tertiary alcohol 1.
antiomers of diphenylmethanol 3. The reduc-
We found that the isotopic chirality in alcohols
Next, (S)- and (R)-1-methyl-13C-oxy-3- tion of the aryl bromide via the corresponding 1 to 3 was successfully discriminated by asym-
methoxy-2-propanol 2 were synthesized from highly polar boronic acid facilitated separation metric autocatalysis. The results of asymmetric
(R)- and (S)-glycidyl methyl ether 11 using 13C- of unreacted compounds in each step. The en- autocatalysis triggered by the carbon isotopically
labeled sodium methoxide. The stereoinversion antiomeric excess of the synthesized alcohols chiral alcohols are shown in Table 1. First of all, we
reaction (26) of (S)-2 and alcoholysis also afforded 1 and 2 could be determined from proton nu- investigated methyl-13C-methylphenyl methanol
the opposite (R)-2. Thus, both enantiomers 2 could clear magnetic resonance (1H-NMR) spectra 1, which is a chiral tertiary alcohol formed by the
be prepared from (R)-11 as the chiral substrate. of their diastereomeric esters of a-methoxy-a- substitution of the 13C-methyl group. In series A1,
asymmetric autocatalysis was performed in the
presence of (R)-1 with 89% ee; the synthesis of
the isotopically substituted alcohol proceeded via
route I, including an asymmetric dimethylzinc
addition using the same chiral ligand (–)-8, and
production of (S)-5 with 92% ee was observed
(entry 1). In turn, when (S)-1 was used as the
chiral trigger, (R)-5 with 96% ee was obtained in
95% yield (entry 2). These relationships of
absolute configurations between 1 and the result-
ing 5 are highly reproducible (entries 3 to 8).
Additional eight experiments of autocatalytic re-
action performed using (R)-1 (table S1, entries 1 to
4) and (S)-1 (entries 5 to 8), have also strongly
supported the stereochemical correlations between
(R)- and (S)-alcohol 1 and pyrimidyl alkanol (S)-
and (R)-5 (23). These results using carbon isotopic
chiral tert-alcohol 1, whose enantiomers were syn-
thesized from the same chiral source, clearly indi-
cate that the direction of asymmetric induction is
strongly correlated to the carbon isotopic chirality
in compound 1.
Using the single enantiomer (–)-8 in synthe-
sizing both (R)- and (S)-alcohols 1 excluded the
possibility that residual contamination from 8
could have induced the enantiomeric excess of
the product alkanol 5. If trace amounts of 8 in the
added sample of alcohol 1 were responsible for
the asymmetric induction, we would expect com-
Fig. 2. Discrimination of carbon isotopic chirality by asymmetric autocatalysis.
pound 5 to exhibit the same absolute configura-
Fig. 3. Synthetic routes to
carbon isotope-modified
chiral compounds. a)
Ti(Oi-Pr)4, ligand (–)-8,
toluene or ether; b) di-
isopropyl tartrate (DIPT),
Ti(Oi-Pr)4,t-BuOOH, CH2Cl2;c)
p-toluenesulfonic anhydride,
4-(dimethylamino)pyridine,
pyridine; d) LiAlH4,ether;e)
Na, 13CH3OH; f) Ph3P, di-
isopropyl azodicarboxylate,
(p-NO2)C6H4CO2H, tetrahy-
drofuran (THF); g) K2CO3,
MeOH; h) diphenyl(1-
methylpyrrolidin-2-
yl)methanol (DPMPM),
(p-Br)C6H4B(OH)2, di-
ethylzinc, toluene then
recrystallization; i) chlorotrimethylsilane, Et3N, THF; j) n-BuLi, B(OEt)3, THF; k) diethylzinc, toluene then H2O; l) tetra-n-butylammonium fluoride, THF.
www.sciencemag.org SCIENCE VOL 324 24 APRIL 2009
493

REPORTS
tion regardless of the choice of (R)- or (S)-1 as a ty: Chiral alcohol (R)-1 acts as an initiator of (S)-1 each 8 times (table S1, entries 9 to 16 and
potential trigger. Instead, as confirmed in Table 1, (S)-alkanol 5 formation (entries 9, 11, 13, and 17 to 24). The results show consistent stereo-
entries 1 to 8, and table S1, entries 1 to 8, (R)- and 15), and (S)-1 leads to (R)-5 (entries 10, 12, 14, chemical correlations without any exception (23).
(S)-alkanols 5 were obtained reproducibly from and 16, respectively). When other samples of
adding isotopically chiral (S)- and (R)-alcohols 1, (R)- and (S)-tert-alcohol 1, which were prepared arising from the isotopically chiral compound, we
respectively, ruling out any role of trace 8. in different vessels, were used as chiral inducers, conducted the autocatalytic reaction induced by
To check for chiral influences other than that
In series A2 of Table 1, the enantiomers of the same stereochemical outcomes were ob- the achiral alcohol 1, which was synthesized by
tert-alcohols 1 were synthesized via route II. In tained. To establish reproducibility, we carried route II using unlabeled 9. Previously, we have
this series, the induction sense was the same as out an additional 16 asymmetric autocatalysis reported spontaneous absolute asymmetric synthe-
observed in series A1, with high reproducibili- experiments, using isotopically chiral (R)- and sis in the autocatalytic pyrimidine-5-carbaldehyde/
diisopropylzinc system (29, 30). That is, under
achiral conditions, enantioenriched (S)- or (R)-
Table 1. Chiral discrimination of the carbon isotopically chiral alcohols using asymmetric autocatalysis.
alkanol 5 could be obtained by the reaction of
aldehyde 4 and i-Pr2Zn in combination with asym-
metric autocatalysis. The stochastic behavior of
the formation of (S)- and (R)-alkanol 5 was ob-
served during 84 trials (30). Thus, we considered
that if there is no chiral factor, which can induce
the chirality in asymmetric autocatalysis, the for-
mation of both (S)- and (R)-alkanol should be
observed. On the other hand, if there were some
chiral factors in the reaction, predominant forma-
tion of (S)- or (R)-pyrimidyl alkanol 5 correlated
to the chiral factor should be observed. When the
autocatalysis was performed in the presence of
unlabeled achiral compound 1 prepared using
D-(–)–diisopropyl tartrate (DIPT) in the ep-
oxidation step, enantioenriched (S)- and (R)-
alkanol 5 were obtained in three and five instances
from eight attempts, respectively. In addition,
achiral compound 1, prepared by using tartrate
with the opposite absolute configuration as a
chiral ligand, initiated the formation of enan-
tiomers of 5 with equal frequency, that is, four
occurrences of S and four of R. Detailed experi-
mental results of asymmetric autocatalysis in the
presence of nonlabeled achiral alcohol 1 are given
in table S2 and supporting text (23). Considering
the 24 reproducible asymmetric induction results
obtained using isotopically chiral 1 (Table 1, en-
tries 9 to 16, and table S1, entries 9 to 24), these
observations of both enantiomers of 5 during 16
trials with achiral 1 confirm the absence of any
chiral contaminants in the synthesized sample of
1 that can influence the asymmetric autocatalysis.
In series B of Table 1, sec-alcohol 2 was used
to trigger asymmetric autocatalysis. In the pres-
ence of (S)-alcohol 2, enantioselective i-Pr2Zn
addition to pyrimidine-5-carbaldehyde 4 led to
(S)-pyrimidyl alkanol 5 with 97% ee (entry 17).
On the other hand, the opposite enantiomer (R)-2
induced the production of (R)-5 with 93% ee
(entry 18). When we repeated the experiments,
(S)- and (R)-sec-alcohol 2 induced the formation
of (S)- and (R)-alkanol 5 with the same sense of
stereochemistry, respectively (entries 19 to 21).
To exclude any chiral influence other than that
of carbon isotopic chirality, the (R)-alcohol 2
used as the chiral initiator in entries 22 to 24 was
prepared from the opposite enantiomer (S)-2
by the stereoinversion reaction; the same stereo-
chemical correlation as before was observed in
formation of enantioenriched (R)-5. The repro-
ducibility was further established by an addi-
tional 14 asymmetric autocatalysis experiments
13C/12C chiral
alcohol (ee)
Pyrimidyl alkanol 5
ee
Entry
Isolated yield
Configuration
Series A1*†‡
1
2
3
4
5
6
7
8
(R)-1 (89% ee)
(S)-1 (93% ee)
(R)-1 (89% ee)
(S)-1 (93% ee)
(R)-1 (89% ee)
(S)-1 (93% ee)
(R)-1 (89% ee)
(S)-1 (93% ee)
94
95
80
94
92
96
88
96
88
94
93§
93
S
R
S
R
S
R
S
R
96
97
[not determined]
92
Series A2†‡||
9
10
11
12
13
14
15
16
(R)-1 (86% ee)
(S)-1 (86% ee)
(R)-1 (86% ee)
(S)-1 (90% ee)
(R)-1 (86% ee)
(S)-1 (84% ee)
(R)-1 (86% ee)
(S)-1 (90% ee)
94
87
93
98
85
93
88
93
94
95
94
93
92
88
89
90
S
R
S
R
S
R
S
R
Series B‡¶
17
(S)-2 (>95% ee)
(R)-2 (>95% ee)
(S)-2 (>95% ee)
(R)-2 (>95% ee)
(S)-2 (>95% ee)
(R)-2 (>95% ee)
(R)-2 (>95% ee)
(R)-2 (>95% ee)
92
86
89
89
90
90
92
78
97
93
92
76
83
95
89
89
S
R
S
R
S
R
R
R
18
19
20
21
22#
23#
24#
Series C‡**
25
(S)-3 (99% ee)
(R)-3 (96% ee)
(S)-3 (99% ee)
(R)-3 (96% ee)
(S)-3 (99% ee)
(R)-3 (96% ee)
(S)-3 (99% ee)
(R)-3 (96% ee)
96
96
73
80
84
87
85
80
95
92
94
92
94
92
70
76
S
R
S
R
S
R
S
R
26
27
28
29
30
31
32
*Carbon isotopically chiral initiator 1 was synthesized via route I in Fig. 3.
0.05:1.05:2.15. Aldehyde 4 and i-Pr2Zn were added in three separate portions.
†The molar ratio of 1:4:i-Pr2Zn =
‡Forty-six additional runs were carried out
in this way, further confirming the invariance of the relationship between the handedness of alcohol 1–3 and that of product
5, but with a variation in the absolute value of the product ee. In series A1 (8 further runs) with alcohol 1 of 89% ee (R) or
93% ee (S), the range of product ee was 79 to 94%; for series A2 (16 further runs) with alcohol 1 varying in ee between 84
and 90%, the range of product ee was between 32 and 95%; for series B (14 further runs) with alcohol 2 (R) or (S) of >95%
ee, the range of product ee was between 63 and 88%, and for series C (8 further runs) with alcohol 3 of 99% ee (S) or 96% ee
(R), the range of ee of the product 5 was between 63 and 88%. See table S1 for the full data.
§The enantioenrichment of
product alkanol 5 amplifies according to the addition of aldehyde 4 and i-Pr2Zn: For example, in this entry the step-by-step
measured ee values of alkanol 5 during three consecutive reactions were 5% ee (after the initial dropwise addition of 4 and
i-Pr2Zn), 32% ee (after the second addition), and then 93% ee (after the third addition).
alcohol 1 was synthesized via route II in Fig. 3. ¶The molar ratio of 2:4:i-Pr2Zn = 0.04:1.8:3.74. Aldehyde 4 and i-Pr2Zn
were added in four separate portions. The ee values of 2 were determined from 1H-NMR of its MTPA-ester.
#(R)-Alcohol 2
used as chiral initiator was synthesized by a stereoinversion reaction from (S)-2. **The molar ratio of 3:4:i-Pr2Zn =
||Carbon isotopically chiral
0.025:0.75:1.525. Aldehyde 4 and i-Pr2Zn were added in four separate portions. The ee of 3 was assigned based on the
measured ee of 13.
494
24 APRIL 2009 VOL 324 SCIENCE www.sciencemag.org

REPORTS
18. T. Shibata et al., J. Am. Chem. Soc. 120, 12157
using isotopically chiral (S)- and (R)-2 eight
The neglected carbon isotopic chirality of many
(1998).
and six times, respectively (table S1, entries 25 organic compounds on Earth—a characteristic
to 32 and 33 to 38). The results show the same that has largely eluded discrimination using con-
stereochemical correlations without any excep- temporary methods—can thus be discriminated
19. K. Soai et al., J. Am. Chem. Soc. 121, 11235 (1999).
20. T. Kawasaki et al., J. Am. Chem. Soc. 128, 6032 (2006).
21. D. G. Blackmond, C. R. McMillan, S. Ramdeehul,
A. Schorm, J. M. Brown, J. Am. Chem. Soc. 123, 10103
(2001).
tion (23).
by asymmetric autocatalysis, which is a highly
Finally, in series C of Table 1, (S)- and (R)- sensitive reaction for recognizing and amplify-
phenyl-1,2,3,4,5,6-13C6-phenylmethanol 3 were ing the extremely small chiral influence be-
used as chiral initiators of asymmetric autocatal- tween 12C and 13C. The method described above
ysis. When (S)-3 was used as the chiral trigger, may expand the scope of research on carbon iso-
enantioselective addition of i-Pr2Zn to pyrimidine- tope chirality in organic molecules in nature (33).
5-carbaldehyde 4 afforded (S)-pyrimidyl alkanol Natural enantiomeric excesses in this class of
5 with 95% ee in 96% yield (entry 25). In con- carbon isotopically chiral compounds may be
trast, the reaction in the presence of (R)-3 gave very low, however. The possible role in asym-
(R)-alkanol 5 with 92% ee (entry 26). These cor- metric autocatalytic reactions of such compounds
relations between the 13C isotope chirality and with low ee remains to be clarified.
the configuration of the produced alkanol 5 also
22. D. Lavabre, J.-C. Micheau, J. R. Islas, T. Buhse, Top. Curr.
Chem. 284, 67 (2008).
23. Materials, methods, and additional experimental results
of asymmetric autocatalysis are available as supporting
material on Science Online.
24. O. Prieto, D. J. Ramón, M. Yus, Tetrahedron Asymmetry
14, 1955 (2003).
25. T. Katsuki, K. B. Sharpless, J. Am. Chem. Soc. 102, 5974
(1980).
26. O. Mitsunobu, M. Yamada, Bull. Chem. Soc. Jpn. 40,
2380 (1967).
27. K. Soai, S. Niwa, Chem. Rev. 92, 833 (1992).
28. G. Zhao, X. Li, X. Wang, Tetrahedron Asymmetry 12, 399
(2001).
References and Notes
showed strong reproducibility, with (S)- and (R)-3
inducing the production of (S)- and (R)-5, re-
spectively (Table 1, entries 27 to 32, and table S1,
entries 39 to 46) (23).
1. K. Nakanishi, N. Berova, R. W. Woody, Circular
29. K. Soai et al., Tetrahedron Asymmetry 14, 185 (2003).
30. T. Kawasaki, K. Suzuki, M. Shimizu, K. Ishikawa, K. Soai,
Chirality 18, 479 (2006).
Dichroism: Principles and Applications (Wiley,
New York, 2000).
2. J. Haesler, I. Schindelholz, E. Riguet, C. G. Bochet, W. Hug,
Nature 446, 526 (2007).
31. M. M. Green, C. Khatri, N. C. Peterson, J. Am. Chem. Soc.
115, 4941 (1993).
In these enantioselective reactions, the initial
formation of the zinc alkoxide of the isotopically
chiral alcohol tips the enantiomeric balance of
the initial i-Pr2Zn addition to aldehyde 4; thus, a
small enantiomeric excess of the zinc alkoxide of
5 is induced. After this step, the subsequent auto-
catalytic amplification of the small enantiomeric
excess causes the zinc alkoxide of 5 to accumu-
late at enhanced ee, with an absolute configura-
tion tied to that of the 13C-labeled chiral alcohol
(1, 2, or 3) used to trigger the process. Thus, the
extremely small chiral effect generated by the
substitution of the carbon isotope must be re-
sponsible, not for the amplification of the enan-
tiomeric excess in the asymmetric autocatalysis,
but for the enantioselection observed, that is, for
the enantiomer produced in excess in the forma-
tion of the initial zinc alkoxide intermediate and
therefore in the final production of 5.
Only minute energy differences can be asso-
ciated with the isotopically substituted enantiomer
of 1, 2, or 3 in causing the tiny enantiomeric ex-
cess of the alkoxide that triggers the asymmetric
autocatalysis arising from the dialkylzinc addi-
tion to 4 (Fig. 2). It has been recognized that such
minute energy differences are not interpretable
with the tools available to structural theory (15, 31).
In the initial reaction stage, the minute energy
differences have a tiny effect, which is then
cooperatively amplified to the final large enantio-
3. L. A. Nafie, T. B. Freedman, Enantiomer 3, 283 (1998).
4. L. D. Barron, Nature 238, 17 (1972).
5. K. Mislow, Top. Stereochem. 22, 1 (1999).
6. M. Tsukamoto, H. B. Kagan, Adv. Synth. Catal. 344, 453
(2002).
32. T. Shibata, S. Yonekubo, K. Soai, Angew. Chem. Int. Ed.
38, 659 (1999).
33. B. Barabás, L. Caglioti, K. Micskei, C. Zucchi, G. Pályi,
Orig. Life Evol. Biosph. 38, 317 (2008).
34. This work was supported by a Grant-in-Aid for
Scientific Research from The Ministry of Education,
Culture, Sports, Science and Technology (MEXT). The
authors are grateful to the reviewers for helpful
discussions.
7. I. Weissbuch, L. Addadi, M. Lahav, L. Leiserowitz, Science
253, 637 (1991).
8. A. Eschenmoser, Science 284, 2118 (1999).
9. J. R. Cronin, S. Pizzarello, Science 275, 951 (1997).
10. M. A. Evans, J. P. Morken, J. Am. Chem. Soc. 124, 9020
(2002).
Supporting Online Material
www.sciencemag.org/cgi/content/full/1170322/DC1
Materials and Methods
SOM Text
Figs. S1 to S11
Schemes S1 to S10
11. H. Pracejus, Tetrahedron Lett. 7, 3809 (1966).
12. I. Sato, D. Omiya, T. Saito, K. Soai, J. Am. Chem. Soc.
122, 11739 (2000).
13. A. Horeau, A. Nouaille, K. Mislow, J. Am. Chem. Soc. 87,
4957 (1965).
14. D. Arigoni, E. L. Eliel, Top. Stereochem. 4, 127 (1969).
15. M. M. Green et al., Angew. Chem. Int. Ed. 38, 3139
(1999).
Tables S1 and S2
16. K. Soai, T. Shibata, H. Morioka, K. Choji, Nature 378,
29 December 2008; accepted 3 March 2009
Published online 26 March 2009;
10.1126/science.1170322
767 (1995).
17. I. Sato, H. Urabe, S. Ishiguro, T. Shibata, K. Soai,
Angew. Chem. Int. Ed. 42, 315 (2003).
Include this information when citing this paper.
A Global View of the
Lithosphere-Asthenosphere Boundary
Catherine A. Rychert* and Peter M. Shearer
meric excess. It is therefore impossible to disclose a The lithosphere-asthenosphere boundary divides the rigid lid from the weaker mantle
structural reason for the difference in frequency of and is fundamental in plate tectonics. However, its depth and defining mechanism are not
these initial enantiotopic face selectivities. The well known. We analyzed 15 years of global seismic data using P-to-S (Ps) converted phases and
“cooperation–amplification” effect, which has been imaged an interface that correlates with tectonic environment, varying from 95 T 4 kilometers
discussed regarding the structural effect of hy- beneath Precambrian shields and platforms to 81 T 2 kilometers beneath tectonically altered
drogen deuterium chiral substitution in a helical regions and 70 T 4 kilometers at oceanic island stations. High-frequency Ps observations
polymer (15), finds reactive analogy in this auto- require a sharp discontinuity; therefore, this interface likely represents a boundary in
catalytic system. The product handedness is en- composition, melting, or anisotropy, not temperature alone. It likely represents the
tirely predictable, but the absolute value of the ee lithosphere-asthenosphere boundary under oceans and tectonically altered regions, but
varies from run to run. This observation indicates it may constitute another boundary in cratonic regions where the lithosphere-asthenosphere
a stochastic influence in the initiation phase bi- boundary is thought to be much deeper.
ased by the chiral additive. The locally induced
enantiomeric excess of zinc alkoxide is then am-
plified by the normal asymmetric autocatalytic
process (32).
apping the depth and character of the to be a challenge. Global surface-wave studies
lithosphere-asthenosphere boundary with (1–4) image rigid lithospheres that increase in
existing seismic methods has proven thickness from oceans to continents at broad
M
www.sciencemag.org SCIENCE VOL 324 24 APRIL 2009
495