Further insights into the organocatalytic reaction of 2,2-dimethyl-1,3-dioxan-5-one with α-silyloxy aldehydes

Article abstract of DOI:10.1016/j.tetlet.2016.10.042

Proline-catalysed reactions of dihydroxyacetone isopropylidene acetal, 1, with enantiopure α-silyloxy aldehydes 2/4/6/8 afford 90–95% yields of cross-aldol products (only one stereoisomer in each case), provided that 15 ± 10 equiv of H₂O are pr

Full text of DOI:10.1016/j.tetlet.2016.10.042

Accepted Manuscript
Further insights into the organocatalytic reaction of 2,2-dimethyl-1,3-dioxan-5-
one with α-silyloxy aldehydes
Dani Sánchez, Héctor Carneros, Alejandro Castro-Alvarez, Enric Llàcer, Ferran
Planas, Jaume Vilarrasa
PII:
S0040-4039(16)31355-7
DOI:
http://dx.doi.org/10.1016/j.tetlet.2016.10.042
TETL 48212
Reference:
Tetrahedron Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
1 August 2016
29 September 2016
12 October 2016
Please cite this article as: Sánchez, D., Carneros, H., Castro-Alvarez, A., Llàcer, E., Planas, F., Vilarrasa, J., Further
insights into the organocatalytic reaction of 2,2-dimethyl-1,3-dioxan-5-one with α-silyloxy aldehydes, Tetrahedron
Letters (2016), doi: http://dx.doi.org/10.1016/j.tetlet.2016.10.042
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Tetrahedron Letters
1
Further insights into the organocatalytic reaction of 2,2-dimethyl-
1,3-dioxan-5-one with -silyloxy aldehydes
Dani Sánchez, Héctor Carneros, Alejandro Castro-Alvarez, Enric Llàcer, Ferran Planas, Jaume
Vilarrasa*
Organic Chemistry Section, Facultat de Química, Universitat de Barcelona, Av. Diagonal 645, 08028 Barcelona, Catalonia, Spain
A R T I C L E I N F O
____________________________
A B S T R A C T
________________________________________________________________________
Article history:
Proline-catalysed reactions of dihydroxyacetone isopropylidene acetal, 1, with enantiopure
-silyloxy aldehydes 2/4/6/8 afford 90–95% yields of cross-aldol products (only one stereoisomer
in each case), provided that 15±10 equiv of H2O are present in the medium. Individual reaction
steps have been monitored by NMR spectroscopy to gain insight into the cause of these remark-
able results. It is confirmed that proline 'protects' 2 as oxazolidinone 2-ox, but it means that
proline—the catalyst—is trapped by 2; several equiv of H2O are required to avoid it. Moreover,
H2O mediates prolyl group exchanges and goes against the dehydration of aldols. For the first
time, we detected by NMR one 'wanted' intermediate (an enamine of the adduct).
Received ..................
Revised .....................
Accepted ..................
Available online ........
Keywords:
© 2016 Elsevier Ltd. All rights reserved
Organocatalysis
Cross-aldol reactions
Chiral -silyloxy aldehydes
The proline-catalysed reaction of the natural product
dihydroxyacetone and its derivatives, especially of 2,2-
dimethyl-1,3-dioxan-5-one (dihydroxyacetone isopropyl-
idene acetal, 1), with various aldehydes has been often used
for the preparation of polyols and monosaccharides;1,2 the
pioneers were the groups of Barbas, MacMillan, Enders
and Cordova.1 It is a remarkable direct cross-aldol-like
reaction (formation of -hydroxy ketones or ketols), as the
enamine of a ketone (1) reacts with an aldehyde, whereas
self-aldol products are minor. The challenge is that both
carbonyl compounds have -hydrogens. Most of these
reactions have been performed with achiral aldehydes or D-
glyceraldehyde derivatives, affording acceptable or good
yields and stereoselectivities.1,2 More recently, Hanessian3
and Enders groups4 have reported two outstanding examples
with chiral -silyloxy aldehydes (Scheme 1).
We were interested in going further: to widen the scope
of aldehydes and reaction conditions reported by Enders et
al.;4 to gain insight into the features of -OR aldehydes and
dihydroxyacetone derivatives; and to apply the improved
conditions to the preparation of pentitol analogues (instead of
hexitols1) that may give natural aminosaccharide antibiotics by
standard replacement of appropriate OH groups for azide ion.
Scheme 1. Reaction of 1 with -silyloxy aldehydes. Precedents.
To our delight, in a preliminary experiment with 1, tert-
butyldiphenylsilyloxy propanal 2 and (S)-proline (Pro),
hydroxyketone 3 was formed more rapidly than in the
examples of Scheme 1, in 10:1 v/v DMSO/H2O. In fact, the
conversion of 2 into 3 was complete within 15 h.5 Only one
stereoisomer was observed (that shown in Scheme 2, top);
the isolated yield was 91%.5 In the experiment we had used 25
equiv of H2O! Although the beneficial effect of several equiv
of H2O on Pro-mediated aldol reactions is well known,6 such a
large excess of H2O should be very detrimental for the
formation of any enamine and/or iminium salt. We soon
found that the addition of 5 to 25 equiv of H2O was ideal in
this particular case.7,8
———
*Corresponding author. Tel.: +34 934021258, fax: +34 933397878.
E-mail address: jvilarrasa@ub.edu

2
Tetrahedron Letters
With 15 mol % of Pro (1/2/Pro/H2O in 3 : 1 : 0.15 : 15
Using (R)-proline, the diastereomer of 3 drawn in
Scheme 2 (1,3,4,5-tetrahydroxypentan-2-one derivative 3')
was exclusively formed. The OH/OR groups are in a 3,4-
anti-4,5-syn relationship, according to their NMR spectra
(which agree with those reported1–4 for related compounds).
Moreover, the enantiomers of 3 and 3' were readily prepared
by using the enantiomer of 2. Chiral HPLC (see Supple-
mentary Data) confirmed the stereopurity of the products.
We examined the scope of the reaction using other -
silyloxy aldehydes (4, 6 and 8). Complete conversions to
stereopure anti adducts (see Scheme 2) were also obtained,
with various silyl groups,9 although the reaction times had
to be lengthened in some cases.
ratio), for 48 h, the reaction was also complete. Thus, it is
not necessary to use 30 mol % of Pro. As expected, the
lower the amount of catalyst the lower the reaction rate, but
nothing else. With only 1.2 equiv of 1 and 10 mol % of Pro
(1.2 : 1 : 0.10 : 10) the reaction was slower (66% of
conversion within two days, ca. 100% after one week), as
also expected, but the stereoselectivity was maintained.
To discard that some stereoisomers formed in minute
amounts were not detected by NMR or HPLC, we prepared
the stereoisomers of 3 shown in Scheme 3 by equilibration
at room temperature via the enolate, in the presence of a
weakly basic tertiary amine such as 1,4-diazabicyclo[2.2.2]-
octane (DABCO, 20 mol %), to prevent retro-aldol
reactions. Thus, 3 was converted into 3'' (1:3 ratio, after 48
h in DMSO-d6, 1:1.5 after 7 days in CDCl3). We separated
the mixture and confirmed by NMR and HRMS the structure
of 3''. Similarly, 3' was converted into 3''' (1:2.8 ratio after
7 days in CDCl3). The J34 values of 3'' and 3''' were much
lower than those of their respective precursors. Calculations
to understand why these equilibria are shifted toward the
syn isomers, as well as to explain the differences between
the coupling constants, are given as Supplementary Data. In
summary, the 3,4-anti isomers (consequence of the aldol
reaction) can be partially isomerised to their 3,4-syn analogues.
These syn isomers were absent from the aldol reactions.
Scheme 3. Isomerisation of anti-aldols 3 and 3' to their syn-aldols.
Such a stereopurity is astonishing. A partial racemisation
of 2/4/6/8 in such polar media may be expected (as well as
self-aldol adducts from partially racemised aldehydes via
their enamines).10 However, we demonstrated that -OR
aliphatic aldehydes gave rise to oxazolidinone tautomers
rather than enamine tautomers when they are treated with
equimolar amounts of Pro.11 We have now determined Keq ≈
200 ± 20 for the equilibrium of Scheme 4. Moreover, we
confirmed that -silyloxy aldehydes and 0.3 equiv of Pro
did not racemise at all in DMSO-d6 or in DMSO-d6 with 25
equiv of D2O, for 24 h.
Scheme 2. Direct cross-aldol reaction between 1 and 2/4/6/8 in DMSO-d6
with 15–25 equiv of H2O (indicated in parentheses) and 0.3 equiv of Pro.
Reaction times for complete conversion are given for a 3:1 ratio of 1 to the
aldehydes. Relevant 1H NMR spectral data in CDCl3 are included.

Tetrahedron Letters
3
significant amount of free 2. With 50 equiv of D2O, the
hydrate of 2 became the major compound (and nothing else
was detected, that is, 2-ox and 2 disappeared).14 Thus, it is
confirmed that without adding ≥ 10 equiv of H2O the catalyst
(Pro) is mostly trapped by -OSiR3 aldehydes, so the enamine
of 1 can hardly be formed. It is also noted that the formation
of hydrates, which may have a negative effect, is not
significant under the reaction conditions.
The fact is that the reaction of 2/4/6/8 with Pro to form
bicyclic exo-oxazolidinones (e.g. 2-ox) is so favoured that the
solid (Pro) is solved rapidly in anhydrous DMSO (stored over
MS) and also in DMF, despite the fact that only up to 0.3
equiv of H2O can be generated, as only 0.3 equiv of Pro is
used. In our case, Pro was stored in a desiccator over P4O10 to
ensure that no additional moisture was introduced. In other
words, the added water is not necessary to solubilise Pro.
Addition of several equiv of H2O is required to avoid that
the catalyst (Pro) almost completely disappears from the
medium by reaction with -silyloxy aldehydes, which,
obviously, are in excess with regard to the catalyst.12
Not all the merit belongs to Pro. Part of the success has to
be attributed to the special features of 1. In fact, the parallel
reaction performed with cyclohexanone (instead of 1) and 2,
plus 0.3 equiv of Pro and 25 equiv of H2O for 15 h, gave the
corresponding aldol(s) with a conversion of only 38%. With
10 equiv of cyclohexanone and with 10 equiv of H2O, after
48 h of reaction the outcome was only 51%; this fact can be
explained by the low concentration of cyclohexanone
enamine in the medium rather than by a low reactivity.
We examined the reaction of ketone 1 with Pro alone
(Scheme 6). After mixing them one-to-one in anhydrous
DMSO-d6, enamine 1-en was the only new product
observed in the 1H and HSQC NMR spectra. Oxazolidinone
Scheme 4. Reaction of 2 with (S)-proline. Tautomers of 2-ox such as
enamines and zwitterions were not detected by 1H NMR.
1
1-ox was not detected by H NMR even after overnight
accumulation. The Keq for 1 + Pro = 1-en + H2O, determined
from the integration of appropriate signals, is 2.5 (2.5 ±
0.5). By contrast, cyclohexanone (and most ketones) do not
form enamines in the same ratio; for example, we estimate
Keq(en) ≈ 0.006 for the formation of the enamine from
cyclohexanone and Pro. It may explain why ketone 1 (but
not standard ketones) is efficient in the cross-aldol reactions
here examined: at least there is a detectable concentration
of enamine 1-en in the medium.
-Silyloxy aldehydes are special indeed. -Alkoxy
aldehydes do not behave identically. 4-Methoxybenzyloxy
derivative 10, chosen as representative, was also fully
converted into the aldol adducts, within 40 h, but in this case
the formation of C5-epimers as impurities was observed,13
more in the presence of Pro than of (R)-Pro (Scheme 5). An
explanation is that the multiple equilibria1,2 involved in the
formation of the oxazolidinones are not so shifted to
oxazolidinone 10-ox as in the parallel equilibria involving
2/4/6/8. A scheme is given as Supplementary Data.
Scheme 6. Dioxanone 1 and Pro give enamine 1-en rather than
oxazolidinone 1-ox.
Mixing 1 and 2-ox may give rise to the 1 + 2-ox = 1-en +
2 equilibrium, an exchange of the 'prolyl group' between an
enamine and an oxazolidinone (Scheme 7, top). This
equation is the subtraction of the formation equilibrium of
2-ox and that of 1-en. The K value would be ≈2.5 ± 0.5 /
200 ± 20 ≈ 0.013 ± 0.002.15 In practice, this equilibrium
cannot be established as 1-en and 2 shall react from the
very beginning. We have performed such an experiment.
We prepared 2-ox in DMSO-d6 (in the presence of MS) and
added 3 equiv of 1, without water. Nothing happened.
However, when this reaction was repeated with 10 equiv of
H2O, aldol 3 appeared quickly; it was the major species
Scheme 5. Reaction of 1 with 10 in the presence of proline.
Nevertheless, the question is to know which species
predominate(s) when 15±10 equiv of H2O is added to the
DMSO solutions of the oxazolidinones 2-ox/4-ox/6-ox/8-ox.
To a solution of 2-ox in DMSO-d6 (prepared by mixing
equimolar amounts of 2 and Pro, in the presence of crushed
4 Å MS and filtering under Ar) we added 10 equiv of D2O.
This gave rise to the appearance of a small aldehyde signal
(2), that is, to the reversal of Scheme 4, as expected. No other
intermediates, such as hemiaminals (2-ha) or any enamine
tautomer or conformer (2-en), were detected. With 25 equiv
of D2O, similar molar amounts of 2-ox, 2 (H1 9.50, 3J = 0.8)
and the hydrate of 2 (H1 4.67, 3J = 4.5) were noted. In other
words, it is necessary to add > 10 equiv of H2O to observe a
1
noted by H NMR after few hours and the predominant
product after 15 h (as in Scheme 2). Thus, the role of water
may be viewed as a mediator or initiator of the exchange of
the prolyl group between 1 and 2-ox.

4
Tetrahedron Letters
The presence of several equiv of H2O is not necessary to
mentary Data. Moreover, in another experiment we added 1
to a DMSO-d6 solution such as that of Figure 1. Nothing
significant occurred. However, it was sufficient to pour 0.5
equiv of D2O into the NMR tube to observe a prolyl
exchange: decrease of 3-en, rise of 1-en and 3 (1 + 3-en =
1-en + 3). If drops of D2O were added all the intermediates
were hydrolysed, as expected. The direct hydrolysis and the
exchange reaction mediated by H2O are two faces of the
same coin, but may be viewed as independent.
cause the hydrolysis of the intermediates within brackets of
Scheme 7 (where other possible tautomers are not included
to save space). In DMSO, the last hydrolysis is so shifted
toward the final product that moisture and/or part of the 0.3
equiv of H2O produced from the reactions of Pro with 1 and
2 are sufficient for the turnover of Pro. These intermediates
behave as dehydrating agents. Exchange reactions may also
participate in the hydrolysis of these intermediates.16
In summary, fully stereoselective cross-aldol reactions
between ketone 1 (via its enamine 1-en) and enantiopure
aldehydes 2/4/6/8 are reported, expanding the reactivity of 1
already known. Besides, different stereoisomers of aldol 3, a
representative member of the series, have been prepared and
characterised for the first time. As aldehydes 2/4/6/8 fully trap
the catalyst (Pro) to give the corresponding bicyclic exo-
oxazolidinones, 15±10 equiv of H2O must be added to shift the
equilibrium position of Scheme 4 to the left otherwise there is
no sufficient catalyst available for the formation of enamine 1-
en (which is obviously not favoured by the presence of H2O);
this has been assumed by several authors, but it is now shown
by NMR. Such a huge excess of H2O is not necessary for the
last step, i.e. for the hydrolysis of the intermediate detected
here for the first time, 3-en, since it is very prone per se to
hydrolysis. Trace amounts of H2O favour the exchanges of the
prolyl group and some amounts of H2O militate against aldol
condensations (dehydration of the aldol). In short, the excellent
yields are owing to four favourable effects of H2O counter-
acting two unfavourable effects. Final corollary: trace amounts
or much lower amounts of H2O (and/or of Pro) would be
required if the equilibrium of Scheme 8 was more shifted to
the right, viz. using analogues of 1 with a higher tendency to
give enamines and using aldehydes with a slightly lower
tendency to give oxazolidinones than the silyloxy derivatives
examined here, as we hope to demonstrate in due course.
Scheme 7. Transfer of the prolyl group between 1 and bicyclic
oxazolidinone 2-ox, followed by the reaction of 1-en with 2.
It was hard to detect by NMR any of the intermediates
within brackets of Scheme 7.17 However, when we treated 1
and 2 with 120 mol % of Pro under very anhydrous
conditions (DMSO-d6, 4 Å MS, CaH2), two enamines were
clearly observed by 1H NMR, HSQC and HMBC, viz. 1-en
and 3-en (Figure 1). Other tautomers of 3-en (the regio-
isomeric enamine, oxazolidinones and zwitterion 3-zw) were
not detected. Mixing 1 and Pro (1:1) under these dehydrating
conditions, to first form sufficient amounts of 1-en, followed
by addition of 2, provided identical spectra, but the appear-
ance of dehydrated dimer of 1 (self-condensation) and
dehydrated aldol as by-products was often observed by NMR.
Scheme 8. Plausible equilibria between analogues of 1 and Pro-
protected aldehydes (bicyclic oxazolidinones) other than 2/4/6/8.
Acknowledgments
Grants CTQ2012-39230 and 2015-71506-R (MINECO-
FEDER) are acknowledged. D.S. and A.C.A. held PhD
studentships from our University (UB) and Chile Govern-
ment, respectively; currently they have a fellowship from
the Fundació Cellex. H.C. has a PhD studentship
(CTQ2012). E.L. initiated these studies in 2010 as a PhD
student of the UB. E.P. was an undergraduate student (TFG
2015). Inspiration from articles of Enders and Hanessian
groups mentioned in the introduction is recognised; J.V.
dedicates this paper to these colleagues. Personal comments
from Prof. D. E. Ward (Ref. 10) are also acknowledged.
Figure 1. Reaction of 1, 2 and Pro in the presence of dehydrating agents
(1H NMR spectrum, from 5.90 to 4.90 ppm)
By addition of 10 equiv of D2O to the NMR tube (of
Figure 1) hydrolysis of 3-en took place, as expected,
whereas the product (deuterated 3) appeared. The NMR
signals of 1-en and 2-ox also disappeared, although not so
irreversibly and/or rapidly as the signals of 3-en. See Supple-

Tetrahedron Letters
5
Supplementary data
Blackmond, D. G. J. Am. Chem. Soc. 2007, 129, 15100–15101;
for excellent, representative reviews, see: (h) Gruttadauria, M.;
Giacalone, F.; Noto, R. Adv. Synth. Catal. 2009, 351, 33–57; (i)
Rah, M.; Singh, V. K. Chem. Commun. 2009, 6687–6703; (j)
Mase, N.; Barbas, C. F. Org. Biomol. Chem. 2010, 8, 4043–4050;
(k) Bhowmick, S., Mondal, A.; Ghosh, A.; Bhowmick, K. C.
Tetrahedron: Asymmetry 2015, 26, 1215–1244.
Supplementary data associated with this article (experi-
mental details, NMR spectra, complementary schemes and
DFT calculations at different levels) can be found, in the
online version, at http://dx.doi.org/ 10.1016/j._tetlett.2015...
7. When we performed the reaction of 1 and 2 without adding H2O,
after 15 h at rt the conversion was 40%. With 3 equiv of H2O (with
regard to 2) the conversion reached 85%. With 5, 15, and 25 equiv
of H2O, 100%. With 30 equiv, 85%. With 35 equiv, 75%. With 50
equiv, 40%. Thus, in this case there is a plateau between 5 and 25
equiv of H2O. Using 5 equiv and 15 equiv of H2O, the initial
reaction rates were similar (48% conversions after 4 h).
8. We only consider here homogeneous solutions, in which the
concentration of H2O can affect some steps. We do not deem
biphasic systems ("on water" or "under water", where the
enamine is found in the organic layer formed by non-polar
solvents and/or large excesses of carbonyl compounds floating on
the water phase), catalysts with long hydrophobic chains that may
trap enamines, etc. For excellent essays, see: (a) Brogan, A. P.;
Dickerson, T. J.; Janda, K. D. Angew. Chem., Int. Ed. 2006, 45,
8100–8102; (b) Hayashi, Y. Angew. Chem., Int. Ed. 2006, 45,
8103–8104; (c) Blackmond, D. G.; Armstrong, A.; Coombe, V.;
Wells, A. Angew. Chem., Int. Ed. 2007, 46, 3798–3800.
9. Organocatalytic reactions of 1 with 2/4/6/8 have not been
reported (exhaustive SciFinder search). Reactions of 1 with
BnOCH2CHO are known, however: (a) Cordova, A.; Zou, W.;
Dziedzic, P.; Ibrahem, I.; Reyes, E.; Xu, Y. Chem. Eur. J. 2006,
12, 5383–5397; (b) Ibrahem, I.; Zou, W.; Xu, Y.; Cordova, A.
Adv. Synth. Catal. 2006, 348, 211–222; (c) Grondal, C.; Enders,
D. Tetrahedron 2006, 62, 329–337; (d) Ibrahem, I.; Cordova,
A. Tetrahedron Lett. 2005, 46, 3363–3367: (e) Ref. 5.
10. That Pro may racemise chiral -substituted aldehydes is not a
surprise. See, e.g., the first article of the three famous papers of
R. B. Woodward et al. on the synthesis of erythromycin (J. Am.
Chem. Soc. 1981, 103, 3210–3213).
11. Sánchez, D.; Castro-Alvarez, A.; Vilarrasa, J. Tetrahedron Lett.
2013, 54, 6381–6384.
12. Many authors explained that H2O suppresses the formation of
unproductive species (oxazolidinones). Some pointed out that
H2O increases the catalyst concentration. See Refs. 6h–j. In the
present case, it is compulsory to add H2O.
13. In contrast to 2/4/6/8, aldehyde 10 racemises in the presence of
Pro, in DMSO at room temperature. The epimer of 11 turned
out to be the enantiomer of 11'. The epimer of 11' was ent-11.
14. This hydrate was not observed in DMSO-d6 by addition of only
15 equiv of D2O, unless the medium was more ionic: only
when we added 5 equiv of anhydrous LiBr, the percentage of
hydrate in relation to aldehyde was 1:15.
15. The result was 0.023 when measured indirectly by exchange
reactions (see Supplementary Data). By contrast, for standard
ketones such as cyclohexanone, we estimate that the corres-
ponding exchange equilibria have K < 3·10–5.
References and notes
1. For reviews, see: (a) Mlynarski, J.; Gut, B. Chem. Soc. Rev.
2012, 41, 587–596: (b) Enders, D.; Narine, A. A. J. Org.
Chem. 2008, 73, 7857–7870, and refs. 9 and 18 therein; (c)
Markert, M.; Mahrwald, R. Chem. Eur. J. 2008, 14, 40–48.
2. For very recent papers, see: (a) Enders, D.; Terteryan, V.;
Palecek, J. Synthesis 2010, 2979–2984; (b) Albrecht, L.; Jiang,
H.; Dickmeiss, G.; Gschwend, B.; Hansen, S. G.; Jørgensen,
K. A. J. Am. Chem. Soc. 2010, 132, 9188–9196; (c) Yang, H.;
Mahapatra, S.; Cheong, P. H.-Y.; Carter, R. G. J. Org. Chem.
2010, 75, 7279–7290; (d) Pearson, A. J.; Panda, S.
Tetrahedron 2011, 67, 3969–3975; (e) Doyagüez, E. G.;
Corrales, G.; Garrido, L.; Rodríguez-Hernández, J.; Gallardo,
A.; Fernández-Mayoralas, A. Macromolecules 2011, 44, 6268–
6276; (f) Pearson, A. J.; Panda, S. Org. Lett. 2011, 13, 5548–
5551; (g) Ma, G.; Bartoszewicz, A.; Ibrahem, I.; Cordova, A.
Adv. Synth. Catal. 2011, 353, 3114–3122; (h) Fanton, J.;
Camps, F.; Castillo, J. A.; Guérard-Hélaine, C.; Lemaire, M.;
Charmantray, F.; Hecquet, L. Eur. J. Org. Chem. 2012, 203–
210; (i) Palyam, N.; Niewczas, I.; Majewski, M. Synlett 2012,
23, 2367–2370, and preceding articles; (j) Cieplak, M.;
Ceborska, M.; Cmoch, P.; Jarosz, S. Tetrahedron: Asymmetry
2012, 23, 1213–1217; (k) Enders, D.; David, S.; Deckers, K.;
Greb, A.; Raabe, G. Synthesis 2012, 44, 3483–3488; (l)
Doyaguez, E. G.; Fernández-Mayoralas, A. Tetrahedron 2012,
68, 7345–7354; (m) Halperin, S. D.; Britton, R. Org. Biomol.
Chem. 2013, 11, 1702–1705; (n) Miura, D.; Fujimoto, T.;
Tsutsui, A.; Machinami, T. Synlett 2013, 24, 1501–1504; (o)
Dibello, E.; Brovetto, M.; Seoane, G.; Gamenara, D.
Tetrahedron Lett. 2013, 54, 5895–5897; (p) Leczycka, K.;
Chaciak, B.; Cieplak, M.; Cmoch, P.; Jarosz, S. Carbohydr.
Res. 2015, 403, 98–103; (q) Miura, D.; Machinami, T. Modern
Res. Catal. 2015, 4, 20–27; (r) for an application to the
synthesis of palmerolide C, see: Florence, G. J.; Wlochal, J.
Chem. Eur. J. 2012, 18, 14250–14254.
3. Hanessian, S.;Mi, X. Synlett 2010, 761–764.
4. Enders, D.; David, S.; Deckers, K.; Greb, A.; Raabe, G. Synthesis
2012, 44, 3483–3488.
5. The reactions of 1 (0.6 M) with 2 (0.2 M, concentrations in
DMSO/H2O, 10:1 v/v, that is, around 25 equiv of H2O), were
complete after stirring overnight and gave exclusively (TLC,
HPLC, 1H NMR) the indicated stereoisomer, starting from
chiral methyl lactate (≥ 98% ee). Apart from the excess of 1, the
only impurity in the final reaction mixture was its dimer (15%),
see: Enders, D.; Grondal, C. Angew. Chem., Int. Ed. 2005, 44,
1210–1212.
6. For the beneficial effect of several equiv of H2O on Pro-
mediated reactions, see: (a) Nyberg, A.; Usano, A.; Pihko, P.
M. Synlett 2004, 1891–1896 (a case with 10 equiv, the highest
yield but poor ee); (b) Pihko, P.; Pihko, P. M.; Laurikainen, K.
M.; Usano, A.; Nyberg, A. I.; Kaavi, J. A. Tetrahedron 2006,
62, 317–328; (c) See Ref. 2d and ref. 14 therein; (d) see Ref. 2b
(1, 5 equiv of H2O, 68–90% yields); (e) Hayashi, Y.; Aratake,
S.; Itoh, T.; Okano, T.; Sumiya, T.; Shoji, M. Chem. Commun.
2007, 957–959 (best conditions: 3 equiv of H2O); (f) also see:
Torii, H.; Nakadai, M.; Ishihara, K.; Saito, S.; Yamamoto, H.
Angew. Chem., Int. Ed. 2004, 43, 1983–1986; for a dual role of
H2O, see: (g) Zotova, N.; Franzke, A.; Armstrong, A.;
16. The exchange of the prolyl group between these intermediates
and the carbonyl compounds yet present in the medium,
mediated by trace amounts of H2O, is very shifted to the right.
See: Isart, C.; Burés, J.; Vilarrasa, J. Tetrahedron Lett. 2008,
49, 5414–5418. Also see the main text below.
17. This is general for -substituted ketones. For aldehydes and
cyclic ketones, all these types of intermediates have been charac-
terised in different solvents. (a) Seebach, D.; Beck, A. K.; Badine,
D. M.; Limbach, M.; Eschenmoser, A.; Treasurywala, A. M.;
Hobi, R.; Prikoszovich, W.; Linder, B. Helv. Chim. Acta 2007,
90, 425–471. (b) Haindl, M. H.; Hioe, J.; Gschwind, R. M. J. Am.
Chem. Soc. 2015, 137, 12835–12842, and refs. cited therein.

Graphical abstract
Further insights into the organocatalytic reaction of 2,2-dimethyl-1,3-dioxan-5-one with
-silyloxy aldehydes
Dani Sánchez, Héctor Carneros, Alejandro Castro-Alvarez, Enric Llàcer, Ferran Planas, Jaume Vilarrasa*

Tetrahedron Letters
Highlights
Further insights into the organocatalytic reaction of 2,2-
dimethyl-1,3-dioxan-5-one with -silyloxy aldehydes
Dani Sánchez et al., TL 2106
__________________________________________________
__________________________________________________
_____________
· Excellent yield and selectivity of proline-catalysed reactions
with -OSiR3 aldehydes
· Several aldols (pentitol derivatives) were characterised for the
first time
· H2O has many roles, not only the hydrolysis of protected
aldehydes (oxazolidinones)
· Possible negative (2) and positive (4) effects of D2O, in
DMSO-d6, have been examined
· A 'wanted' reaction intermediate, an enamine of the aldol, has
been detected by NMR