Approaching sub-ppm-level asymmetric organocatalysis of a highly challenging and scalable carbon–carbon bond forming reaction

Article abstract of DOI:10.1038/s41557-018-0065-0

The chemical synthesis of organic molecules involves, at its very essence, the creation of carbon–carbon bonds. In this context, the aldol reaction is among the most important synthetic methods, and a wide variety of catalytic and stereoselective versions have been reported. However, aldolizations yielding tertiary aldols, which result from the reaction of an enolate with a ketone, are challenging and only a few catalytic asymmetric Mukaiyama aldol reactions with ketones as electrophiles have been described. These methods typically require relatively high catalyst loadings, deliver substandard enantioselectivity or need special reagents or additives. We now report extremely potent catalysts that readily enable the reaction of silyl ketene acetals with a diverse set of ketones to furnish the corresponding tertiary aldol products in excellent yields and enantioselectivities. Parts per million (ppm) levels of catalyst loadings can be routinely used and provide fast and quantitative product formation in high enantiopurity. In situ spectroscopic studies and acidity measurements suggest a silylium ion based, asymmetric counteranion-directed Lewis acid catalysis mechanism.

Full text of DOI:10.1038/s41557-018-0065-0

Articles
https://doi.org/10.1038/s41557-018-0065-0
Approaching sub-ppm-level asymmetric
organocatalysis of a highly challenging and
scalable carbon–carbon bond forming reaction
Han Yong Baeꢀ ꢀ1, Denis Höfler1, Philip S. J. Kaib1, Pinar Kasaplar1, Chandra Kanta De1, Arno Döhring1,
Sunggi Lee1, Karl Kaupmees2, Ivo Leitoꢀ ꢀ2 and Benjamin Listꢀ ꢀ1*
The chemical synthesis of organic molecules involves, at its very essence, the creation of carbon–carbon bonds. In this context,
the aldol reaction is among the most important synthetic methods, and a wide variety of catalytic and stereoselective ver-
sions have been reported. However, aldolizations yielding tertiary aldols, which result from the reaction of an enolate with a
ketone, are challenging and only a few catalytic asymmetric Mukaiyama aldol reactions with ketones as electrophiles have been
described. These methods typically require relatively high catalyst loadings, deliver substandard enantioselectivity or need
special reagents or additives. We now report extremely potent catalysts that readily enable the reaction of silyl ketene acetals
with a diverse set of ketones to furnish the corresponding tertiary aldol products in excellent yields and enantioselectivities.
Parts per million (ppm) levels of catalyst loadings can be routinely used and provide fast and quantitative product formation in
high enantiopurity. In situ spectroscopic studies and acidity measurements suggest a silylium ion based, asymmetric counter-
anion-directed Lewis acid catalysis mechanism.
hiral tertiary aldols are encountered in a variety of natural Results and discussion
products and pharmaceuticals (Fig. 1a)1. Despite the con- Optimization of catalyst and silyl ketene acetal. For our initial
C
siderable advances in catalytic asymmetric aldol reactions of studies, the addition of silyl ketene acetal 2a to 2-acetonaphthone
aldehydes2–6, only a few examples using ketones as electrophiles are (1a) was chosen as a model reaction (Fig. 2). Indeed, with 0.1mol%
known so far7–9. Thermodynamics typically favour the correspond- of chiral disulfonimide (DSI) catalyst C-127,30,31, no desired product
ing retro-aldol reaction10–13, while the lower reactivity of ketones was detected even after prolonged reaction times (24h). Full con-
compared to aldehydes further challenges the reaction kinetically. sumption of the ketone was observed, but instead of the desired
Importantly, however, the energy that is released in the tautom- aldol product, only silyl enol ether 4 was obtained. In sharp con-
erization of the initial enol product of the reverse aldol reaction trast, with 0.1mol% of IDPi catalyst C-2, tertiary aldol product 3aa
to the ketone is absent in the corresponding Mukaiyama variant. was formed within 3min and with an enantiomeric ratio (e.r.) of
Therefore, the forward Mukaiyama aldol reaction14–16 is favoured, 69.5:30.5.
even with ketones. However, the enantiofacial differentiation of
With these results in hand, an increase in enantioselectiv-
ketones is still much more challenging than that of aldehydes17. The ity through further synthetic manipulations of the four aryl sub-
Denmark group disclosed the first Mukaiyama aldol reaction of stitutents at the 3,3′-positions of the chiral 1,1′-bi-2-naphthol
ketones catalysed by chiral bis-N-oxides acting as a Lewis base17,18
.
(BINOL) scaffold of the catalyst was pursued (Table 1, for detailed
This method requires the use of highly reactive trichlorosilyl eno- results on the optimization of solvents and silyl ketene acetals, see
lates of methyl acetate as the nucleophile (Fig. 1b, top). Shibasaki SupplementaryTable1). BesidesthereportedcatalystsC-2and C-321,
and Kanai developed an enantioselective Mukaiyama aldol reac- several new IDPi catalysts were synthesized (C-4, C-5, C-6, C-7
tion of trimethylsilyl enolates to ketones in the presence of a chi- and C-8) and applied to the model reaction. Gratifyingly, elongat-
ral copper(i) fluoride–phosphine complex as the catalyst (Fig. 1b, ing the alkyl chains at the 3,5-positions of the four aryl substituents
bottom)19. Furthermore, the group of Campagne reported that increased the enantioselectivity, ultimately leading to an e.r. of 97:3
chiral 2,2′-bis(di-p-tolylphosphino)-1,1′-binaphthyl (Tol-BINAP)– with catalyst C-8 (entries 4 to 7).
copper(ii) trifluoromethanesulfonate catalyses the vinylogous
In all cases, the catalytic activities were extremely high and
the reaction was completed within 30min with 1.0mol% of cata-
Mukaiyama aldol reaction of ketones20.
Despite these advances, a more practical and general proto- lyst. Moreover, the catalyst loading can be reduced to 0.05mol%
col for the Mukaiyama aldol reaction of ketones was envisioned (500ppm) without deterioration of yield and enantioselectivity
exploiting the extraordinary reactivity of our recently developed (97:3 e.r., entry 8). After screening of several different silyl ketene
imidodiphosphorimidate (IDPi) catalysts21–25. In recent years, our acetals (2a–2d), tert-butyldimethylsilyl (TBS) or triethylsilyl (TES)
group has made progress in the field of silylium-based, asymmet- proved to be the most suitable silyl groups in terms of reactivity and
ric counteranion-directed Lewis acid catalysis (silylium ACDC)26 enantioselectivity (entries 8 to 11).
of Mukaiyama-type27–29 reactions. These transformations invari-
ably require the use of aldehydes or imines as electrophiles, while Ketone scope and synthetic utility. Under the optimized reac-
ketones have remained elusive.
tion conditions, a variety of ketones (1b–1v) were subjected
1Max-Planck-Institut für Kohlenforschung, Mülheim an der Ruhr, Germany. 2University of Tartu, Institute of Chemistry, Tartu, Estonia. *e-mail: list@kofo.mpg.de
NAtuRe CHemIStRY | www.nature.com/naturechemistry
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NaTure CHeMIsTry
Articles
a
OH
Me OH
Me
OH
Me
CO2H
Me OH
N
O
OMe
HO2C
CHO
O
O
Me
H
OH
N
OH
Mycarose
O
O
OH
Me
Citramalate
O
O
OH
Me
HO
O
CO2Me
CO2H
Me OH
O
Me
HO
O
Dicrotaline
O
O
O
OH
Omacetaxine mepesuccinate
(Homoharringtonine)
Vineomycinone B2
Mevalonolactone
OH
b
Denmark (2002) ref.17
O
HO Me
L* =
Me Me
Ph
OMe
SiCl3
OMe
O
L* (10 mol%)
96%, 91:9 e.r.
O
O
O
+
HO Me
Cy
HO Me
+
+
n-Bu
N
N
n-Bu
R1
R2
O
O
O
O
–
–
OMe
Ph
OMe
t-Bu
t-Bu
91%, 66:34 e.r.
87%, 55:45 e.r.
Shibasaki and Kanai (2006) ref.19
L* (4 mol%)
L* =
N(n-Bu)2
CuF(PPh3)
3 •2EtOH
SiMe3
O
O
O
HO Me
HO Me
(2.5 mol%)
O
+
PCy2
Cy2P
Fe
R1
R2
Ph
OMe
93%, 96:4 e.r.
Cy
OMe
(EtO)3SiF (200 mol%)
PhBF3K (10 mol%)
OMe
89%, 88.5:11.5 e.r.
Fig. 1 | Representative examples of tertiary aldol-containing natural products and catalytic asymmetric routes to these moieties. a, Natural products
containing tertiary aldols. b, Previously reported asymmetric Mukaiyama aldol reactions of ketones as electrophiles, using trichlorosilyl enolates of methyl
acetate catalysed by a chiral Lewis base (top) and trimethylsilyl enolates catalysed by a chiral copper(i) fluoride–phosphine complex (bottom). Cy,
cyclohexyl, L*, chiral ligand.
TBS
C-1: (S)-DSI
(0.1 mol%)
C-2: (S,S)-IDPi
(0.1 mol%)
O
O
TBSO Me
O
TBS
O
Me
+
OMe
Et2O (0.2 M),
20 °C
Et2O (0.2 M),
20 °C
OMe
1a
2a
4
3aa
24 h
>99% yield
<3 min
>99% yield
69.5:30.5 e.r.
Ar
Ar
Ar
CF3
O
O
SO2
N
P
P
NH
SO2
Ar
=
O
O
NH
N
CF3
Tf
Tf
Ar
Ar
Ar
C-1: (S)-DSI
C-2: (S,S)-IDPi
Fig. 2 | Reactivity comparison of DSI C-1 and IDPi C-2 in the mukaiyama aldol reaction of ketone as electrophile under identical reaction conditions.
Reactions were performed with ketone 1a (0.5ꢀmmol), silyl ketene acetal 2a (2.0ꢀequiv.) and catalyst (C-1 or C-2, 0.1ꢀmol%) in Et2O (2.5ꢀml, 0.2ꢀM) at 20ꢀ°C.
The yield was determined after chromatographic purification and e.r. was determined by HPLC analysis. Tf, SO2CF3, TBS, SiMe2t-Bu.
to this protocol (Table 2). Remarkably, a loading of 0.005– successfully transformed into desired products 3c and 3l
0.05 mol% (50–500 ppm) with catalysts C-4, C-7 or C-8 in Et2O at (e.r. = 94:6 and 86:14), respectively.
–20 °C proved to be sufficient to provide the desired products
To illustrate the synthetic utility of our method, derivatizations
(3b–3v) in quantitative yields (up to >99%) and with excellent of the chiral tertiary aldols 3 were performed. Due to the high vola-
e.r. (up to >99:1). Besides (hetero)aryl-alkyl ketones (1b–1k), tility of silyl ketene acetal 2 and the extremely low catalyst loading,
alkyl-alkyl ketones (1m–1o) were also converted with good to analytically pure product 3 could be easily obtained after remov-
excellent e.r. (88:12 to 95:5). Alkynyl ketone 1p gave an excellent ing the volatiles under reduced pressure. The clean reaction profile
e.r. of 95:5. Interestingly, benzylideneacetone-type α,β-unsatu- of products allowed direct conversion of the material without the
rated ketones 1q–1v were highly reactive and generally gave need for further chromatographic purification. Several function-
good 1,2- versus 1,4-regioselectivity (5:1 to 10:1) and excellent alized chiral tertiary alcohols, such as β-silyloxy aldehyde 5 (89%,
enantioselectivity (e.r. = 97.5:2.5 to >99:1). Ketones other than 95:5 e.r.), diols 6 (two steps, 92%, 95:5 e.r.) and 7 (88%, 95:5 e.r.),
methyl ketones, ethyl ketone 1c and cyclic ketone 1l were also carboxylic acid 8 (91%, 95:5 e.r.), a product of transesterification
NAtuRe CHemIStRY | www.nature.com/naturechemistry
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NaTure CHeMIsTry
Articles
Table 1 | Optimization of the IDPi catalysts and the silyl ketene acetals 2
Ar
Ar
O
O
R3SiO Me
SiR3
O
O
O
(S,S)-IDPi
N
Me
OMe
P
P
+
OMe
O
O
Et2O (0.2 M),
NH
N
–20 °C
Tf
Tf
Ar
Ar
1a
2a–2d
3aa–3ad
Me
Me
Me
Me
Me
Me
Ar
=
Me
Me
C-3
C-4
C-5
C-6
C-7
C-8
entry
Catalyst (mol%)
SiR3
time (h)
Yield (%)
e.r.
1
C-2 (1.0)
TBS (SiMe2t-Bu, 2a)
TBS (2a)
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
12
99 (3aa)
99 (3aa)
99 (3aa)
99 (3aa)
99 (3aa)
99 (3aa)
99 (3aa)
99 (3aa)
99 (3ab)
99 (3ac)
67 (3ad)
75:25
79.5:21.5
89:11
2
C-3 (1.0)
3
C-4 (1.0)
TBS (2a)
4
5
C-5 (1.0)
TBS (2a)
91:9
C-6 (1.0)
TBS (2a)
95.5:4.5
95.5:4.5
97:3
6
7
C-7 (1.0)
TBS (2a)
C-8 (1.0)
TBS (2a)
8
9
10
11a
C-8 (0.05ꢀ=ꢀ500ꢀppm)
C-8 (0.05ꢀ=ꢀ500ꢀppm)
C-8 (0.05ꢀ=ꢀ500ꢀppm)
C-8 (0.05ꢀ=ꢀ500ꢀppm)
TBS (2a)
97:3
TMS (SiMe3, 2b)
TES (SiEt3, 2c)
TIPS (Sii-Pr3, 2d)
12
89:11
12
95.5:4.5
69:31
24
Reactions were performed with ketone 1a (0.5ꢀmmol), silyl ketene acetal 2 (2.0ꢀequiv.) and catalyst (0.05–1.0ꢀmol%) in Et2O (2.5ꢀml, 0.20ꢀM) at –20ꢀ°C. All yields are those of isolated products and e.r. was
determined by HPLC analysis. TBS, tert-butyldimethylsilyl; TMS, trimethylsilyl; TES, triethylsilyl; TIPS, triisopropylsilyl. aReaction was performed at 24ꢀ°C.
(9, 84%, 95:5 e.r.), and amide 10 (two steps, 83%, 95:5 e.r.) were materials and solvent were recharged and subjected to the aldol
successfully obtained. In addition, chiral tertiary aldols 3p and 3u reaction under the same conditions without the addition of more
were quantitatively transformed to product 3n, without erosion of catalyst (batch B; see Supplementary Table 6). A third portion of
e.r. (see Supplementary Sections 7 and 8 for details).
substrate and reagent was added (batch C; see Supplementary
Table 7), ultimately leading to a total conversion of 95.2% (82%
Towards extremely low catalyst loadings. To further investigate yield: total TON ~911,000). It is noteworthy that ppm-level load-
the performance of IDPi catalyst C-8, very low catalyst loading32 ings in organocatalysis have previously been reported, for instance,
experiments were implemented on a preparative scale. Deca-grams in silylative kinetic resolutions of racemic alcohols33 and in asym-
of products 3i (>99%, 20.5g) and 3v (92%, 20.1g) were quantita- metric transfer hydrogenations34. To the best of our knowledge,
tively obtained in very high enantioselectivities (95:5 e.r. and 96:4 however, approaching sub-ppm level loadings (<1ppm, Fig. 3b,c)
e.r.) using 25ppm or 2.8ppm of the catalysts, respectively (Fig. 3a; of an organocatalyst in a challenging carbon–carbon bond forming
see Supplementary Section 12 for analytical results). Substrate 1u reaction with high enantioselectivity is unprecedented.
was further explored17. Due to the existence of a background reac-
In contrast to the formidable catalytic activity of our IDPi cata-
tion (Supplementary Fig. 2), conditions suitable for obtaining excel- lyst, DSI C-1 (0.1mol%) provided a mixture of 1,4-addition product
lent e.r. in these large-scale syntheses were first identified (at 5°C 11 with tertiary aldol 3u (47:53 e.r., 54% isolated yield, 11/3u=1.5:1;
with 8.0M; see Supplementary Table 3 for full data). Initially, we see Supplementary Fig. 5). In addition, the silyl enol ether of start-
fixed the amount of ketone 1u at 1.00g and lowered the catalyst ing ketone 1u (compound 12, 4%) was also generated (Fig. 3e).
loadings from 8ppm (0.10mg) to 2.8ppm (0.037mg). In all cases,
full conversion of 1u and quantitative yield of tertiary aldol prod- Investigation of the reaction mechanism. To gain insights into
uct 3u (~90% yield) with >95:5 e.r. was observed. Excitingly, with the reaction mechanism, real-time analytical investigations were
30.0g of substrate, only 1.1mg of IDPi C-8 successfully produced conducted using in situ Fourier transform infrared (FT-IR) spec-
59.2g of the enantioenriched product (86% corrected yield, turn- troscopy (Fig. 4a)35,36. The absorption peak of the carbonyl group
over number (TON) ~307,000, Fig. 3a and Supplementary Table 4). of ketone 1a at 1,688cm−1 was monitored during the reaction in the
Ketone 1u was partially soluble at the early stage of the reaction, and presence of 0.1mol% of IDPi catalyst C-6 at 20°C (step A). After
slowly dissolved during the course of the reaction, finally leading to 7min, silyl ketene acetal 2a (1,656cm−1) was added to the reaction
a homogeneous solution (Fig. 3d).
mixture (step B). The reaction proceeded rapidly and was completed
1
We also investigated the repeated addition of starting materi- in less than 3min (>99% conversion by H NMR spectroscopy) to
als after consumption of the initial reaction mixture with 2.8ppm give aldol 3aa (Fig. 4a, left, 1,744 cm−1) as a single product with an
catalyst loading (batch A, Fig. 3b; see Supplementary Table 5). On e.r. of 95:5. As previously mentioned, DSI C-1 did not catalyse the
completion of the reaction and removal of the volatiles, starting desired transformation but exclusively gave silyl enol ether 4 instead
NAtuRe CHemIStRY | www.nature.com/naturechemistry
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NaTure CHeMIsTry
Articles
Table 2 | Substrate scope for the mukaiyama aldol reaction of ketone 1 with silyl ketene acetal 2a using different IDPi catalysts, and
synthetic utilities of product 3
TBS
TBSO R2
O
O
(S,S)-IDPi (500 ppm)
O
+
R1
R2
R1
OMe
Et2O (0.2 M), –20 °C
*
OMe
2a
1b–1v
3b–3v
O
O
O
TBSO Et
TBSO Me
TBSO Me
O
TBSO
[Gram scale]
Me
OMe
OMe
OMe
OMe
C-7 (50 ppm: 0.48 mg)
0.25 M, 20 °C, 12 h:
99% (1.92 g), 95:5 e.r.
MeO
MeO
3b
3c
3d
C-8: 99%, 96:4 e.r.
3e
C-7: 99%, 98:2 e.r.
C-8: 99%, 94:6 e.r.
C-7: 99%, 95:5 e.r.
O
O
O
O
O
TBSO Me
TBSO Me
TBSO Me
TBSO Me
TBSO Me
MeO
Me
OMe
OMe
OMe
OMe
OMe
Ph
t-Bu
MeO
MeO
Me
3f
3g
3h
3i
3j
C-8: 96%, 92:8 e.r.
C-8: 99%, 92:8 e.r.
C-8: 99%, 97:3 e.r.
C-8: 95%, 99:1 e.r.
C-8: 99%, 95:5 e.r.
[Gram scale]
C-8 (100 ppm: 1.5 mg)
2.0 M, 0 °C, 3 d:
O
O
O
TBSO
TBSO Me
TBSO Me
O
TBSO Me
S
*
OMe
OMe
OMe
99% (2.47 g), 95:5 e.r.
C-8 (50 ppm: 0.76 mg)
2.0 M, 0 °C, 7 d:
Ph
OMe
3k
3l
C-8: 93%, 86:14 e.r.
3m
3n
C-8: 99%, 97:3 e.r.
C-7: 99%, 95:5 e.r.
C-4: 94%, 88:12 e.r.
90% (2.25 g), 95:5 e.r.
O
O
O
O
O
TBSO Me
TBSO Me
TBSO Me
TBSO Me
TBSO Me
O
O
Me
OMe
OMe
OMe
OMe
OMe
Ph
3o
3p
3q
3r
3s
Cl
Me
C-4: 95%, 90:10 e.r.
C-7: 96%, 95:5 e.r.
C-7: 87%a 97.5:2.5 e.r.
C-8: 84%a 99:1 e.r.
C-8: 82%a 98:2 e.r.
O
O
O
[Gram scale]
[Gram scale]
TBSO Me
O
TBSO Me
TBSO Me
OMe
OMe
OMe
C-8 (50 ppm: 0.66 mg)
2.0 M, 0 °C, 6 h:
88%a (2.02 g), 97:3 e.r.
C-8 (50 ppm: 0.60 mg)
8.0 M, 25 °C, 3 h:
99%b (2.15 g), 97:3 e.r.
Me
3t
3u
3v
C-8: 78%a 99:1 e.r.
C-8: 91%a 97.5:2.5 e.r.
C-8: 99%b >99:1 e.r.
Me
TBSO Me
O
O
TBSO Me
TBSO Me
Ph
OH
Me
MeO
MeO
S
OMe
H
Conditions:
3n, quant.
95:5 e.r. and 97.5:2.5 e.r.
(i) Et3Si-H (10 equiv.), Pd/C (10 wt%),
MeOH, 20 °C, 1 h
7
5
80%, 95:5 e.r.
89%, 95:5 e.r.
(ii) DIBAL-H (1.1 equiv.), CH2Cl2,
–78 °C, 1.5 h
(ix)
(ii)
From
(i)
3p and 3u
From 3i
From 3k
(iii) LiAlH4 (1.2 equiv.), THF, 0 °C, 6 h
(iv) n-Bu4NF (2.0 equiv.), THF, 0 °C, 3 h
(v) CH2Cl2/CF3CO2H (50:1, vol/vol),
0 °C to r.t., 12 h
TBSO R2
R1
TBSO Me
HO Me
O
MeO
MeO
CO2H
S
(viii)
(iii) then (iv)
OH
OMe
From 3i
From 3k
3
(vi) NH2Me (5.0 equiv.), THF, 70 °C, 12 h
(vii) Zn4(OCOCF3)6O (10 wt%),
EtOH, 75 °C, 12 h
No chromatography
8
6
91%, 95:5 e.r.
92%, 95:5 e.r.
From 3ab, 3ad, 3e, 3k and 3u From 3m
(viii) LiOH (5.0 equiv.), water (10 equiv.),
MeOH/THF (vol/vol 1:1), 65 °C, 3 h
then HCl (10% aq.), pH = 7
(ix) MeMgI (2.2 equiv.), CH2Cl2,
0 °C, 24 h
(iv) (v)
O
O
HO R2
O
HO Me
HO Me
(vii)
(vi)
Me
S
R1
Ph
OEt
N
From 3u'
OMe
From 3k'
H
9
3'
Quant. (prone to retro-aldol)
10
83%, 95:5 e.r.
84%, 95:5 e.r.
Reactions were performed with ketone 1 (0.5ꢀmmol), silyl ketene acetal 2a (2.0ꢀequiv.) and IDPi catalyst (0.05ꢀmol%ꢀ=ꢀ500 ppm) in Et2O (2.5ꢀml, 0.20ꢀM) at –20ꢀ°C for 12–72ꢀh. Gram-scale reactions were
performed with ketone 1 (1.0ꢀg), silyl ketene acetal 2a (1.1 equiv.) and IDPi catalyst (50 or 100ꢀppm) in Et2O under the indicated conditions. All yields are those of isolated products and e.r. was determined
by HPLC analysis. aCorrected yields of 1,2-adducts are indicated (9–22% of 1,4-adducts were observed, see Supplementary Section 11 for details). bNo 1,4-adduct was observed.
NAtuRe CHemIStRY | www.nature.com/naturechemistry
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NaTure CHeMIsTry
Articles
a
IDPi
TBS
(2.8 or 25 ppm)
TBSO R2
O
O
O
+
R1
R2
R1
OMe
Et2O
(2.0 or 8.0 M)
OMe
1
3
2a (1.1 equiv.)
(Decagram scale)
O
TBSO
Me
O
TBSO
Me
Me
O
TBSO
Me
MeO
MeO
OMe
Ph
OMe
Ph
OMe
From 1u (30.0 g)/C-8 (2.8 ppm: 1.1 mg)
From 1i (10.0 g)/C-8 (25 ppm: 2.1 mg)
From 1v (10.0 g)/C-8(2.8 ppm: 0.34 mg)
8.0 M, 5 °C, 13 d
2.0 M, 10 °C, <0.5 d
8.0 M, 24 °C, 2 d
3i
3v
3u
92% (20.1 g)a, TON 329,000, 96:4 e.r.
86% (59.2 g)b, TON 307,000, 95:5 e.r.
>99% (20.5 g), TON 40,000, 95:5 e.r.
b
Batch A
O
TBS
C-8
O
O
TBSO Me
1.9 x 10-4 mmol: 0.37 mg
Batch B
Batch C
+
Ph
Me
Ph
OMe
OMe
2a (1.1 equiv.)
Et2O, 5 °C
1u
3u
Batch ppm of C-8
1u (mmol)/2 (mmol)/Et2O (ml)/C-8 (mg)
Time (days)
Total conv. (1u, %)/yield (3u, %)
Total TON
e.r.
A
B
2.8
1.4
68.4 (10.0 g)/75.2 (14.2 g)/8.55 (8.0 M)/0.37
5
~100/91 (20.8 g)
99.1/89 (40.7 g)
325,000
636,000
95:5
95:5
A + 68.4 (10.0 g)/75.2 (14.2 g)/8.55/0
A + 15
C
0.9
B + 68.4 (10.0 g)/75.2 (14.2 g)/8.55/0
B + 60
d
95.2/82 (56.3 g)
911,000
95:5
c
100
80
60
40
20
0
1 day
13 days
Batch A:2.8 ppm of C-8 (10.0 g of 1u)
2.8 ppm of C-8 (30.0 g of 1u)
Batch B: 1.4 ppm of C-8
Batch C: 0.9 ppm of C-8
3u (59.2 g)
1u
0
5
10 15 20 25 30 35 40 45 50 55 60
Time (days)
e
MeO2C
Ph
Using DSI C-1 (0.1 mol% = 1,000 ppm)
1u (6.84 mmol)
O
OTBS
TBSO Me
OTBS
Me
+
+
Ph
OMe
Ph
2a (1.1 equiv.)
Et2O (8.0 M)
5 °C, 24 h:
11
3u, 47:53 e.r.
54% (11/3u = 1.5:1)
12
4%
Fig. 3 | mukaiyama aldol reaction of ketones on a preparative scale with catalyst loadings between 0.9 and 25ꢁppm. a, Reaction details and results.
b, Repeated addition of starting materials. c, Reaction progress of the aldol reaction of 1u to 3u catalysed by C-8 (0.9–2.8ꢀppm). d, Images of crude reaction
mixtures with thin-layer chromatography (TLC) analyses (left and middle images: after 1ꢀday and 13ꢀdays; TLC eluent: acetone/hexanesꢀ=ꢀ1:20, left lane:
1u/centre lane: co-spot/right lane: mixture) and dried product 3u (right image: 250ꢀml round-bottom flask). e, Reaction outcome using DSI C-1 (0.1ꢀmol%)
as catalyst. aNo 1,4-adduct was observed. bThe corrected yield of 1,2-adduct is indicated. TON, turnover number.
(Fig. 4a, right, 1,751cm−1)37. The remarkably different reaction out- ketene acetal 2a to give the equivalent of a silylium ion pair 13 as the
comes may be explained by considering counteranion basicities. catalytically relevant Lewis acid (Fig. 4c). Silylium transfer onto the
We measured pKa values of several different chiral Brønsted acid ketone then furnishes an oxocarbenium ion pair, the fate of which is
catalysts (developed in our research group) with reference to some dependent on the counteranion; that is, the DSI anion is sufficiently
achiral acids in acetonitrile (Fig. 4b, see Supplementary Table 10 for basic to deprotonate the oxocarbenium ion and furnishes the cor-
pKa values of our chiral acid organocatalysts and the correspond- responding enol silane (pKa of DSI C-1=8.4). In contrast, the IDPi
ing references). In line with our previous investigations21,23,25, we anion is only a very weak base, enabling the desired Mukaiyama
propose that our IDPi catalyst is a precatalyst, which is silylated by pathway to occur (pKa of IDPi C-3=4.5, in acetonitrile).
NAtuRe CHemIStRY | www.nature.com/naturechemistry
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NaTure CHeMIsTry
Articles
a
TBS
O
Step A
Step B
O
O
TBSO Me
C-6 or C-1
2a (2 equiv.)
(0.1 mol%)
Me
OMe
+
Et2O
20 °C
1a
3aa
4
Using C-6 (IDPi): <3 min, >99% yield, 95:5 e.r.
Using C-1 (DSI): 0%
0%
9 h, 88% yield
0.100
3aa (using C-6)
B
1,744 cm–1
4 (using C-1)
1,751 cm–1
0.100
B
0.000
0.050
A
A
0.000
1,750
08:20:00
1,700
1,750
00:26:00
1,700
1,650
1,650
00:00:00
00:03:00
1,600
b
C-3 (IDPi)
C-1 (DSI)
8.4
4.5
4.1
pKa
(in CH3CN)
Stronger
BA
11.0
HCl
10.6
8.5
6.4
5.5 5.1
0.7
0.3
(C6F5SO2)2NH
CF3SO3H
4-NO2C6H4SO3H
(CF3SO2)2NH
HBr
HI
Picric acid
4-CH3C6H4SO3H
c
O
O
O
P
N
O
–
N
R1
R2
P
Tf
N
O
1
TBS
+
TBS
O
TBS
O
Tf
O
R2
δ+
OMe
R1
OMe
O
2a
2a
O
N
P
O
O
–
P
IDPi
N
Tf
TBS
N
+
Tf
O
O
O
N
Active silylium LA
equivalent (13)
P
N
O
P
Me
OMe
–
O
N
Tf
Tf
TBSO R2
O
+
O
TBS
TBS
O
R2
R1
OMe
3
R1
OMe
Fig. 4 | Study of the reaction mechanism. a, Analytical results of the in situ FT-IR investigation of the catalytic Mukaiyama aldol reaction of ketone 1a with
silyl ketene acetal 2a using IDPi C-6 (left) and DSI C-1 (right) as catalysts. b, Measured pKa values of chiral acids (DSI C-1 and IDPi C-3) and reported pKa
values of achiral acids in acetonitrile (see Supplementary Section 9 for details). BA, Brønsted acid; LA, Lewis acid. c, Proposed reaction mechanism of the
IDPi-catalysed Mukaiyama aldol reaction.
Conclusions
only extremely low catalyst loadings are generally required, with
In conclusion, we have developed a broadly applicable Mukaiyama 50–500ppm being sufficient to obtain the desired products quan-
aldol reaction of ketones 1 with silyl ketene acetals 2. In the pres- titatively, typically in less than 12 h. We also report preparative-
ence of IDPi catalysts, a variety of (hetero)aryl-alkyl and alkyl- scale reactions that require only single-digit ppm loadings and
alkyl ketones are converted into the corresponding tertiary in one case, even less than 1ppm. Although the reactions under
aldols in almost quantitative yields (up to 99% yield) and with these extreme conditions require purified reagents and extended
very high to excellent enantioselectivities (e.r. up to 99:1). The reaction times (2–80days), to our knowledge the catalyst loadings
developed method is operationally simple and scalable, and the reported are unprecedented in asymmetric carbon–carbon bond
obtained products can be readily derivatized. We also show that forming reactions.
NAtuRe CHemIStRY | www.nature.com/naturechemistry
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NaTure CHeMIsTry
Articles
22. Xie, Y. et al. Catalytic asymmetric vinylogous Prins cyclization: a highly
methods
diastereo- and enantioselective entry to tetrahydrofurans. J. Am. Chem. Soc.
138, 14538–14541 (2016).
23. Lee, S., Kaib, P. S. J. & List, B. Asymmetric catalysis via cyclic, aliphatic
oxocarbenium ions. J. Am. Chem. Soc. 139, 2156–2159 (2017).
24. Liu, L. et al. Catalytic asymmetric [4+2]-cycloaddition of dienes with
aldehydes. J. Am. Chem. Soc. 139, 13656–13659 (2017).
25. Gatzenmeier, T., Kaib, P. S. J., Lingnau, J. B., Goddard, R. & List, B. Te
catalytic asymmetric Mukaiyama–Michael reaction of silyl ketene acetals with
α,β-unsaturated methyl esters. Angew. Chem. Int. Ed. 57, 2464–2468 (2018).
26. Mahlau, M. & List, B. Asymmetric counteranion-directed catalysis: concept,
defnition, and applications. Angew. Chem. Int. Ed. 52, 518–533 (2013).
27. García-García, P., Lay, F., García-García, P., Rabalakos, C. & List, B.
A powerful chiral counteranion motif for asymmetric catalysis. Angew. Chem.
Int. Ed. 48, 4363–4366 (2009).
28. Ratjen, L., Garcia-Garcia, P., Lay, F., Beck, M. E. & List, B. Disulfonimide-
catalyzed asymmetric vinylogous and bisvinylogous Mukaiyama aldol
reactions. Angew. Chem. Int. Ed. 50, 754–758 (2011).
29. Tap, A., Blond, A., Wakchaure, V. N. & List, B. Chiral allenes via
alkynylogous Mukaiyama aldol reaction. Angew. Chem. Int. Ed. 55,
8962–8965 (2016).
30. van Gemmeren, M., Lay, F. & List, B. Asymmetric catalysis using chiral,
enantiopure disulfonimides. Aldrichimica Acta 47, 3–13 (2014).
31. James, T., van Gemmeren, M. & List, B. Development and applications of
disulfonimides in enantioselective organocatalysis. Chem. Rev. 115,
9388–9409 (2015).
32. Giacalone, F., Gruttadauria, M., Agrigento, P. & Noto, R. Low-loading
asymmetric organocatalysis. Chem. Soc. Rev. 41, 2406–2447 (2012).
33. Park, S. Y., Lee, J.-W. & Song, C. E. Parts-per-million level loading
organocatalysed enantioselective silylation of alcohols. Nat. Commun. 6,
7512 (2015).
General procedure for the 30g scale catalytic Mukaiyama aldol reaction of
ketones. Ketone 1u (30.0g, 205mmol, 1equiv.) was placed in a fame-dried
J. Young Schlenk fask, equipped with a Tefon-coated magnetic stirring bar. IDPi
C-8 (1.1mg, 5.7×10−4 mmol, 2.8ppm) and Et2O (8.0M, 25.6ml) were added at
25°C and stirred for 30min. Te reaction mixture was cooled to –78°C and silyl
ketene acetal 2a (42.5g, 226mmol, 1.1equiv.) was slowly added. Te reaction
mixture was stirred for 60min at –10°C, then for 60min at 0°C, and fnally stirred
for 13days at 5°C. Organic volatiles (boiling points of volatiles: Et2O=34.6°C; silyl
ketene acetal 2a=62°C/12mbar) were evaporated in vacuo to aford the desired
tertiary aldol product 3u (59.2g, 86% yield, 95:5 e.r.).
Full experimental details and the characterization of new compounds are
provided in the Supplementary Information.
Data availability. The data that support the findings of this study are available
from the corresponding author upon request.
Received: 2 November 2017; Accepted: 11 April 2018;
Published: xx xx xxxx
References
1. List, B. et al. A Catalytic enantioselective route to hydroxy-substituted
quaternary carbon centers: resolution of tertiary aldols with a catalytic
antibody. J. Am. Chem. Soc. 121, 7283–7291 (1999).
2. Nelson, S. G. Catalyzed enantioselective aldol additions of latent enolate
equivalents. Tetrahedron: Asymmetry 9, 357–389 (1998).
3. Mahrwald, R. Diastereoselection in Lewis-acid-mediated aldol additions.
Chem. Rev. 99, 1095–1120 (1999).
4. Machajewski, T. D. & Wong, C.-H. Te catalytic asymmetric aldol reaction.
Angew. Chem. Int. Ed. 39, 1352–1375 (2000).
34. Xu, W. et al. Metal-templated design: enantioselective hydrogen-bond-driven
catalysis requiring only parts-per-million catalyst loading. J. Am. Chem. Soc.
138, 8774–8780 (2016).
35. Zhang, Z. & List, B. Kinetics of the chiral disulfonimide-catalyzed
Mukaiyama aldol reaction. Asian J. Org. Chem. 2, 957–960 (2013).
36. Zhang, Z. et al. Asymmetric counteranion-directed Lewis acid organocatalysis
for the scalable cyanosilylation of aldehydes. Nat. Commun. 7, 12478 (2016).
37. Song, J. J. et al. N-Heterocyclic carbene-catalyzed silyl enol ether formation.
Org. Lett. 10, 877–880 (2008).
5. Mahrwald, R. Modern Aldol Reactions (Wiley-VCH, Weinheim, 2004).
6. Mahrwald, R. Modern Methods in Stereoselective Aldol Reactions (Wiley-VCH,
Weinheim, 2013).
7. Cozzi, P. J., Hilgraf, R. & Zimmermann, N. Enantioselective catalytic
formation of quaternary stereogenic centers. Eur. J. Org. Chem. 2007,
5969–5994 (2007).
8. Hatano, M. & Ishihara, K. Recent progress in the catalytic synthesis of
tertiary alcohols from ketones with organometallic reagents. Synthesis 2008,
1647–1675 (2008).
9. Adachi, S. & Harada, T. Catalytic enantioselective aldol additions to ketones.
Eur. J. Org. Chem. 2009, 3661–3671 (2009).
Acknowledgements
10. Elliott, M. L., Urban, F. J. & Bordner, J. Synthesis and absolute confguration
of (R)- and (S)-ethyl 3-(4-oxocyclohex-2-enyl)propionate. J. Org. Chem. 50,
1752–1755 (1985).
Support from the Max Planck Society, the Deutsche Forschungsgemeinschaft (Leibniz
Award to B.L. and Cluster of Excellence RESOLV, EXC 1069) and the European Research
Council (Advanced Grant ‘C–H Acids for Organic Synthesis, CHAOS’) is acknowledged.
The authors thank J.L. Kennemur for her suggestions during the preparation of this
manuscript, P. Gupta for his assistance on the preparation of artwork, the technicians of
our group, and the members of our NMR, MS and HPLC departments for their excellent
service. The work of K.K. and I.L. was supported by grant IUT20-14 from the Estonian
Ministry of Education and Research. This paper is dedicated to Prof. T. Mukaiyama in
celebration of his 90th birthday (Sotsuju).
11. Rodriguez, M. J. Process for performing retro-aldol reactions. US patent
5,677,423 (1996).
12. Matovic, R., Ivkovic, A., Manojlovic, M., Tokic-Vujosevic, Z. & Saicic, R. N.
Ring closing metathesis/fragmentation route to (Z)-confgured medium ring
cycloalkenes. Total synthesis of ( )-periplanone C. J. Org. Chem. 71,
9411–9419 (2006).
13. Hatano, M., Takagi, E. & Ishihara, K. Sodium phenoxide−phosphine oxides
as extremely active Lewis base catalysts for the Mukaiyama aldol reaction
with ketones. Org. Lett. 9, 4527–4530 (2007).
Author contributions
H.Y.B. developed the reaction and investigated the substrate scope, derivatizations of
the aldol products, and implemented in situ FT-IR study. D.H. first observed the high
activity of IDPi catalysts in the described reaction. The IDPi catalysts were developed by
H.Y.B., P.S.J.K. and B.L. H.Y.B., P.S.J.K., P.K. and S.L. synthesized the IDPi catalysts used
in this study. H.Y.B., C.K.D. and A.D. investigated large-scale and low-catalyst loading
experiments. K.K. and I.L. measured pKa values of acid catalysts. B.L. designed and
oversaw the project. H.Y.B. and B.L. wrote the manuscript.
14. Kobayashi, S., Fujishita, Y. & Mukaiyama, T. Te efcient catalytic
asymmetric aldol-type reaction. Chem. Lett. 19, 1455–1458 (1990).
15. Kan, S. B. J., Ng, K. K. H. & Paterson, I. Te impact of the Mukaiyama aldol
reaction in total synthesis. Angew. Chem. Int. Ed. 52, 9097–9108 (2013).
16. Matsuo, J.-I. & Murakami, M. Te Mukaiyama adol reaction: 40 years of
continuous development. Angew. Chem. Int. Ed. 52, 9109–9118 (2013).
17. Denmark, S. E. & Fan, Y. Catalytic, enantioselective aldol additions to
ketones. J. Am. Chem. Soc. 124, 4233–4235 (2002).
18. Denmark, S. E., Fan, Y. & Eastgate, M. D. Lewis base catalyzed,
enantioselective aldol addition of methyl trichlorosilyl ketene acetal to
ketones. J. Org. Chem. 70, 5235–5248 (2005).
Competing interests
The authors declare no competing interests.
19. Oisaki, K., Zhao, D., Kanai, M. & Shibasaki, M. Catalytic enantioselective
aldol reaction to ketones. J. Am. Chem. Soc. 128, 7164–7165 (2006).
20. Moreau, X., Bazán-Tejeda, B. & Campagne, J.-M. Catalytic and asymmetric
vinylogous Mukaiyama reactions on aliphatic ketones: formal asymmetric
synthesis of taurospongin A. J. Am. Chem. Soc. 127, 7288–7289 (2005).
21. Kaib, P. S. J., Schreyer, L., Lee, S., Properzi, R. & List, B. Extremely
active organocatalysts enable a highly enantioselective addition of
allyltrimethylsilane to aldehydes. Angew. Chem. Int. Ed. 55, 13200–13203
(2016).
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41557-018-0065-0.
Reprints and permissions information is available at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to B.L.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
NAtuRe CHemIStRY | www.nature.com/naturechemistry
© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.