Spiro-fused carbohydrate oxazoline ligands: Synthesis and application as enantio-discrimination agents in asymmetric allylic alkylation

Article abstract of DOI:10.3762/bjoc.12.18

In the present work, we describe a convenient synthesis of spiro-fused D-fructo- and D-psico-configurated oxazoline ligands and their application in asymmetric catalysis. The ligands were synthesized from readily available 3,4,5-tri-O-benzyl-1,2-O-isoprop

Full text of DOI:10.3762/bjoc.12.18

Spiro-fused carbohydrate oxazoline ligands: Synthesis and
application as enantio-discrimination agents in asymmetric
allylic alkylation
Jochen Kraft1, Martin Golkowski2 and Thomas Ziegler*1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2016, 12, 166–171.
1Institute of Organic Chemistry, University of Tuebingen, Auf der
Morgenstelle 18, 72076 Tuebingen, Germany and 2Department of
Pharmacology, University of Washington, 1959 NE Pacific St, Box
357280, Seattle, WA 98195, USA
doi:10.3762/bjoc.12.18
Received: 17 November 2015
Accepted: 18 January 2016
Published: 29 January 2016
Email:
Thomas Ziegler* - thomas.ziegler@uni-tuebingen.de
Associate Editor: M. Rueping
* Corresponding author
© 2016 Kraft et al; licensee Beilstein-Institut.
License and terms: see end of document.
Keywords:
asymmetric catalysis; carbohydrates; oxazolines; palladium; spiro
compounds
Abstract
In the present work, we describe a convenient synthesis of spiro-fused D-fructo- and D-psico-configurated oxazoline ligands and
their application in asymmetric catalysis. The ligands were synthesized from readily available 3,4,5-tri-O-benzyl-1,2-O-isopropyl-
idene-β-D-fructopyranose and 3,4,5-tri-O-benzyl-1,2-O-isopropylidene-β-D-psicopyranose, respectively. The latter compounds
were partially deprotected under acidic conditions followed by condensation with thiocyanic acid to give an anomeric mixture of
the corresponding 1,3-oxazolidine-2-thiones. The anomeric 1,3-oxazolidine-2-thiones were separated after successive benzylation,
fully characterized and subjected to palladium catalyzed Suzuki–Miyaura coupling with 2-pyridineboronic acid N-phenyldiethanol-
amine ester to give the corresponding 2-pyridyl spiro-oxazoline (PyOx) ligands. The spiro-oxazoline ligands showed high asym-
metric induction (up to 93% ee) when applied as chiral ligands in palladium-catalyzed allylic alkylation of 1,3-diphenylallyl acetate
with dimethyl malonate. The D-fructo-PyOx ligand provided mainly the (R)-enantiomer while the D-psico-configurated ligand gave
the (S)-enantiomer with a lower enantiomeric excess.
Introduction
The design of new chiral ligands for stereo-differentiating metal pure natural products and therefore, nearly ideal starting materi-
catalysts that enable asymmetric syntheses is still a highly als for ex-chiral pool syntheses. In addition, carbohydrates have
active field of research in organic chemistry, for there is a con- significantly gained attention as ligands for metal complexes
tinuously growing demand for enantiomeric pure building that enable asymmetric catalysis over the past decades, and an
blocks for pharmaceuticals, agrochemicals or flavors. Carbo- array of highly efficient privileged ligands for metal-catalyzed
hydrates are inexpensive and easy to obtain enantiomerically enantioselective syntheses have been derived from carbo-
166

Beilstein J. Org. Chem. 2016, 12, 166–171.
hydrates so far [1-6]. However, the enantioselective construc- interesting structural features in terms of ligand–metal bite
tion of C–C bonds, especially of tertiary carbon stereocenters, angles and shielding of the palladium center by the carbo-
remains an ongoing challenge. Over the last decades though, hydrate scaffold from one specific side [17]. For example, we
transition metal-catalyzed reactions like the asymmetric allylic concluded from the crystallographic data that the orientation of
alkylation (Tsuji–Trost reaction) have evolved into one of the the OBn group at C-3 in D could have a major impact on the
more powerful tools for synthesizing such tertiary stereocenters “shielded” side of the metal center and thus, also on the stereo-
[7,8]. As a benchmark test for selectivity, the palladium-cata- selectivity of the palladium-catalyzed allylic addition of
lyzed asymmetric addition of dimethyl malonate to 1,3- dimethyl malonate to 1,3-diphenylallyl acetate. Encouraged by
diphenylallyl acetate was often used in the literature for testing these aspects and in order to further our investigations in asym-
the scope of carbohydrate derived ligands for this purpose metric catalysis with spiro-fused oxazoline ligands, we devised
[9-13]. For instance, Kunz and Gläser have demonstrated the a new convenient synthetic strategy. Herein we present a new
stereo-differentiating potential of carbohydrate ligands in this straightforward synthesis of spiro-fused D-fructo- and D-psico-
type of reaction where their gluco-PHOX ligand, derived from configurated PyOx ligands and their application in palladium-
glucosamine, resulted in a high enantiomeric excess of up to catalyzed asymmetric alkylation.
98% [14].
Results and Discussion
Recently, Vidal et al. reported on a spiro-bis(isooxazoline) We started our synthesis from readily available 3,4,5-tri-O-
ligand A (Figure 1) [15] prepared via 1,3-dipolar cycloaddition benzyl-1,2-O-isopropylidene-β-D-fructopyranose (1) [18] and
of 2,6-pyridinedicarbonitrile N,N-dioxide to acetyl protected 3,4,5-tri-O-benzyl-1,2-O-isopropylidene-β-D-psicopyranose (2)
exo-glucal. The performance of ligand A in asymmetric cataly- [19], respectively (Scheme 1). Deprotection of the isopropyl-
sis was then tested in the palladium-catalyzed allylic addition of idene group under acidic conditions gave the corresponding
dimethyl malonate to 1,3-diphenylallyl acetate which, however, diols 3 [20] and 4 as anomeric mixtures. Condensation of the
afforded the desired allylic substitution product only in traces. latter with thiocyanic acid in a Ritter-type reaction according to
The authors attributed the inefficiency of ligand A to the a slightly modified procedure described by Tatibouët et al. [21]
cleavage of the spiro moiety resulting in the formation of ther- gave an anomeric mixture of 1,3-oxazolidin-2-thiones 5 and 6,
modynamically more stable aromatic isoxazole B (Figure 1).
respectively. The anomers of 5 and 6 could not be separated by
standard column chromatography though. Thiocarbamate 5 was
As part of our ongoing research program towards the design of previously mentioned in the literature [21], but due to the fast
carbohydrate derived ligands for metal-catalyzed asymmetric anomerisation and the relative instability of 1,3-oxazolidin-2-
syntheses [16] we recently described the preparation of spiro- thiones it was never characterized. However, in our hands,
fused oxazoline ligands of type C from D-fructose [17]. We anomers could easily be separated by chromatography after
could demonstrate that our oxazoline ligands, in contrast to A, benzylation of 5 and 6 with BnBr and NaH to give the corre-
were able to form air and moisture-stable palladium complexes sponding benzylated sulfanyloxazolines 7 and 8 which were air
of type D upon reaction with Pd(cod)Cl2 (Figure 1). Single and moisture stable (see Supporting Information File 1 for full
crystal X-ray diffraction of these Pd complexes revealed some experimental details). Nevertheless, it should be noted that these
Figure 1: Pd-catalyzed cleavage of spiro-bis(isooxazoline) ligand A to isoxazole B and Pd-complex D prepared from spiro-oxazoline C.
167

Beilstein J. Org. Chem. 2016, 12, 166–171.
Scheme 1: Synthesis of 2-benzylsulfanyl-1,3-oxazolines 7 and 8.
compounds slowly start to decompose after 2 weeks at −28 °C
under an atmosphere of nitrogen.
Next, benzylsulfanyloxazolines 7 and 8 were subjected to
copper-assisted palladium-catalyzed Suzuki–Miyaura-type
cross coupling with commercially available 2-pyridineboronic
acid N-phenyldiethanolamine ester 11 as boron source
(Scheme 2) [22-24]. 2-Pyridineboronic acid 11 was chosen due
to its increased nucleophilicity compared to other boron sources
like pinacol boronic esters or MIDA boronates [25,26]. The
cross coupling proceeded smoothly in THF and gave ligands 9
and 10 in good yields (Scheme 2). It is noteworthy that the low
yield (32%) of 10a is due to the instability of its spiro-fused
oxazoline moiety during chromatography on silica gel but not to
any unwanted side reactions during the cross coupling. Even
small traces of HCl usually present in CDCl3 as impurity
resulted in full decomposition of 10a in seconds.
With ligands 9 and 10 in hand, we moved on to the application
in asymmetric catalysis. As a model system, the Pd-catalyzed
allylic alkylation of dimethyl malonate (13) to rac-1,3-
diphenylallyl acetate (12) was chosen (Scheme 3).
Scheme 2: Pd-catalyzed cross coupling of benzylsulfanyloxazolines 7
and 8.
As mentioned above, this reaction was often used as a bench- In all cases investigated here, the alkylated product 14 was iso-
mark for new chiral ligands and examined in great detail lated after purification by chromatography and its enantiomeric
[9-14,27,28].
excess was determined via chiral HPLC using a Reprosil chiral-
Scheme 3: Palladium catalyzed allylic substitution.
168

Beilstein J. Org. Chem. 2016, 12, 166–171.
NR column. The absolute configuration was assigned by com- entry 10). Similar to ligand 9b, the β-configurated D-psico-
parison of the optical rotation values with literature data [29] ligand 10b leads to a somewhat lower enantiomeric excess of
which are based on the chemical correlation method leading to (S)-14 of 9% (Table 1, entry 11).
(+)-(S)-2-phenylsuccinate [30] and by X-ray structure determi-
nation of (R,E)-3,5-diphenylpent-4-enyl camphor-10-sulfonate The stereoselectivity of the Pd-catalyzed allylic substitution can
[31]. Thus, a positive optical rotation value refers to the (R)-en- be explained via a model for the proposed transition state
antiomer, whereas a negative value belongs to the (S)-enantio- (Scheme 4). As a consequence of the spiro-fused carbohydrate
mer. In addition, the absolute configuration was independently moiety at the oxazoline ring, exo (15x and 17x) and endo (15n
determined by 1H NMR in the presence of the optically active and 17n) diastereomers of the palladium complexes can be
NMR shift reagent (+)-Eu(hfc)3 [29]. All synthesized ligands distinguished. It is also reasonable to assume that exo/endo
were active precatalysts for the allylic substitution, as can be isomers 15 and 17 undergo fast allyl rotation via a η3–η1–η3
seen in Table 1.
isomerization mechanism and thus, exist in a dynamic equilib-
rium which is approximately ten to hundred times faster than
alkylation [7]. Therefore, four reaction pathways are possible,
but only two lead to the observed stereoselectivties. We assume,
that the nucleophilic attack occurs at the allyl terminus trans to
the oxazoline ring, which is in accordance with previously re-
ported findings in allylic substitution using PyOx ligands
[32,33]. If the nucleophile attacks from the (Si)-face in the
fructo-configurated complex 15n product 14 must have the (S)-
configuration which is, however, contrary to the observed
steroselectivity. Therefore, we suggest 15x to be the preferred
isomer for attack by the nucleophile from the (Re)-face which
leads to the η2-complex 16 with (R)-configuration.
Table 1: Pd-catalyzed allylic alkylation using ligands 9 and 10.
Entry
Ligand
Solvent
Yielda
eeb
1
9a
CH2Cl2
CH2Cl2
MePh
74%
80%
68%
80%
80%
62%
43%
traces
traces
67% (R)
9% (R)
76% (R)
59% (R)
71% (R)
88% (R)
93% (R)
n.d.
2
9b
3
9a
4
9a
MeCN
CH2Cl2
MePh
5c
6c
7d
8
9a
9a
9a
MePh
10a
10a
10a
10b
MePh
9
CH2Cl2
n.d.
10e
11
Cl(CH2)2Cl 43%
MePh 56%
59% (S)
9% (S)
In the case of D-psico-ligand 10a, the (S)-enantiomer was found
to be the major enantiomer. That leads to the assumption that
the endo complex 17n is the predominant species attacked by
the nucleophile from the (Si)-face which results in η2-complex
18 exhibiting (S)-configuration. Although the exo complexes
are assumed to be thermodynamically more stable than the endo
isomers, complex 17x is disfavoured, because of the steric
repulsion of the phenyl group in the allylic substrate and the
OBn group at C-3 of the D-psicose moiety. This assumption
also explains our observation, that the D-psico-configurated
ligand 10a requires higher temperatures for formation of the
endo complex 17n and thus, results in a slower reaction rate and
in a lower enantioselectivity compared to ligand 9a.
aIsolated yield after chromatography. bDetermined by chiral HPLC.
cT = 0 °C. dT = −20 °C. eT = 50 °C.
The asymmetric allylic alkylation was carried out in the pres-
ence of 5 mol % [PdCl(C3H5)]2 and 11 mol % chiral ligands 9
and 10, respectively. The D-fructo-configurated ligands 9a and
9b showed preparative yields for (R)-14 in the range of 74–80%
(Table 1, entries 1 and 2). The α-anomer 9a showed a signifi-
cant higher selectivity (67% ee) than the β-anomer 9b (9% ee).
In order to investigate the solvent and temperature effects of the
reaction we conducted further studies with the more selective
ligand 9a. While the reaction proceeded smoothly in all tested
solvents, the enantiomeric excess increased in toluene (76% ee,
Table 1, entry 3), whereas acetonitrile had a negative effect on
selectivity (Table 1, entry 4). Lowering the temperature in-
creased the selectivity of the reaction as well (Table 1, entries
5–7). Conducting the reaction at a temperature of −20 °C
yielded (R)-14 with an enantiomeric excess of 93%. Surprising-
ly, when the D-psico-configurated ligand 10a was used in tolu-
ene or methylene chloride under conditions identical to those
used with ligand 9a, only traces of the alkylated product 14
could be obtained. To our delight, however, ligand 10a was
active in 1,2-dichloroethane at 50 °C and gave the opposite en-
antiomer (S)-14 with an enantiomeric excess of 59% (Table 1,
Conclusion
In summary, we have synthesized four spiro-fused carbo-
hydrate ligands in the D-fructo- and D-psico-series via a
straightforward synthetic route. The key steps in our synthesis
were a Ritter type condensation reaction of partially benzyl-pro-
tected D-fructose and D-psicose derivatives with thiocyanic
acid to afford the corresponding 1,3-oxazolidine-2-thiones, and
the cross coupling of the latter under modified Suzuki–Miyaura
conditions. The prepared ligands were shown to be active
precatalysts for the asymmetric allylic alkylation of 1,3-
diphenylallyl acetate with dimethyl malonate. The D-fructo-
configurated ligands provided the (R)-enantiomer with up to
169

Beilstein J. Org. Chem. 2016, 12, 166–171.
Scheme 4: Proposed transition state of allylic substitution.
93% ee, whereas the D-psico-configurated ligands gave the (S)- Acknowledgements
enantiomer in a somewhat lower selectivity (up to 59% ee). The Financial support of this work by the Deutsche Forschungs-
stereochemical outcome of the reaction could be explained by a gemeinschaft (DFG) is gratefully acknowledged. We also thank
proposed transition state of the allylic substitution. Further Dr. Dorothee Wistuba and her team for recording the mass
insights into the reaction mechanism of the allylic substitution spectra, Dr. Markus Kramer and his team for measuring the
using similar PyOx ligands are currently under investigation NMR spectra, Petra Krüger for performing the elemental
and will be published elsewhere.
analyses, Dr. Jochen Neumaier for the help with HPLC analy-
sis and Dr. Gregor Lemanski for helpful discussions.
Supporting Information
References
1. Boysen, M. M. K., Ed. Carbohydrates – Tools for Stereoselective
Synthesis; Wiley-VCH: Weinheim, Germany, 2013.
doi:10.1002/9783527654543
Supporting Information File 1
Experimental procedures, analytical data and copies of
NMR spectra.
2. Lehnert, T.; Özüduru, G.; Gruel, H.; Albrecht, F.; Telligmann, S. M.;
Boysen, M. M. K. Synthesis 2011, 2685–2708.
doi:10.1055/s-0030-1260143
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-12-18-S1.pdf]
170

Beilstein J. Org. Chem. 2016, 12, 166–171.
3. Benessere, V.; Del Litto, R.; De Roma, A.; Ruffo, F. Coord. Chem. Rev.
2010, 254, 390–401. doi:10.1016/j.ccr.2009.05.001
4. Steinborn, D.; Junicke, H. Chem. Rev. 2000, 100, 4283–4317.
doi:10.1021/cr9903050
30.Hayashi, T.; Yamamoto, A.; Hagihara, T.; Ito, Y. Tetrahedron Lett.
1986, 27, 191–194. doi:10.1016/S0040-4039(00)83974-X
31.von Matt, P.; Llyod-Jones, G. C.; Minidis, A. B. E.; Pfaltz, A.; Macko, L.;
Neuburger, M.; Zehnder, M.; Rüegger, H.; Pregosin, P. S.
Helv. Chim. Acta 1995, 78, 265–284. doi:10.1002/hlca.19950780202
32.Chelucci, G.; Medici, S.; Saba, A. Tetrahedron: Asymmetry 1997, 8,
3183–3184. doi:10.1016/S0957-4166(97)00425-4
5. Diéguez, M.; Pàmies, O.; Claver, C. Chem. Rev. 2004, 104,
3189–3215. doi:10.1021/cr0306889
6. Castillión, S.; Claver, C.; Díaz, Y. Chem. Soc. Rev. 2005, 34, 702–713.
doi:10.1039/b400361f
33.Nordström, K.; Macedo, E.; Moberg, C. J. Org. Chem. 1997, 62,
1604–1609. doi:10.1021/jo961490+
7. Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395–422.
doi:10.1021/cr9409804
8. Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921–2943.
doi:10.1021/cr020027w
License and Terms
9. Diéguez, M.; Jansat, S.; Gomez, M.; Ruiz, A.; Muller, G.; Claver, C.
Chem. Commun. 2001, 1132–1133. doi:10.1039/b101567m
10.Mata, Y.; Diéguez, M.; Pàmies, O.; Claver, C. Adv. Synth. Catal. 2005,
347, 1943–1947. doi:10.1002/adsc.200505192
This is an Open Access article under the terms of the
Creative Commons Attribution License
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
11.Diéguez, M.; Pàmies, O.; Claver, C. Adv. Synth. Catal. 2005, 347,
1257–1266. doi:10.1002/adsc.200505013
12.Diéguez, M.; Pàmies, O.; Claver, C. J. Organomet. Chem. 2006, 691,
2257–2262. doi:10.1016/j.jorganchem.2005.11.024
13.Khiar, N.; Navas, R.; Fernández, I. Tetrahedron Lett. 2012, 53,
395–398. doi:10.1016/j.tetlet.2011.11.043
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
14.Gläser, B.; Kunz, H. Synlett 1998, 53–54. doi:10.1055/s-1998-3124
15.Goyard, D.; Telligmann, S. M.; Goux-Henry, C.; Boysen, M. M. K.;
Framery, E.; Gueyrard, D.; Vidal, S. Tetrahedron Lett. 2010, 51,
374–377. doi:10.1016/j.tetlet.2009.11.028
(http://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
16.Kraft, J.; Schmollinger, D.; Maudrich, J.; Ziegler, T. Synthesis 2015, 47,
199–208. doi:10.1055/s-0034-1379473
doi:10.3762/bjoc.12.18
17.Kraft, J.; Ziegler, T. Carbohydr. Res. 2015, 411, 56–63.
doi:10.1016/j.carres.2015.04.009
18.Kaji, E.; Kurimoto, E.; Saiga, R.; Matsuura, A.; Harada, K.; Nishino, T.
Heterocycles 2005, 66, 453–468. doi:10.3987/COM-05-S(K)54
19.Nakajima, N.; Matsumoto, M.; Kirihara, M.; Hashimoto, M.; Katoh, T.;
Terashima, S. Tetrahedron 1996, 52, 1177–1194.
doi:10.1016/0040-4020(95)00975-2
20.Campbell, M. M.; Heffernan, G. D.; Lewis, T. Carbohydr. Res. 1994,
251, 243–250. doi:10.1016/0008-6215(94)84289-2
21.Tatibouët, A.; Lawrence, S.; Rollin, P.; Holman, G. D. Synlett 2004,
1945–1948. doi:10.1055/s-2004-830886
22.Liebeskind, L. S.; Srogl, J. Org. Lett. 2002, 4, 979–981.
doi:10.1021/ol0200091
23.Leconte, N.; Pellegatti, L.; Tatibouët, A.; Suzenet, F.; Rollin, P.;
Guillaumet, G. Synthesis 2007, 857–864. doi:10.1055/s-2007-965932
24.Silva, S.; Tardy, S.; Routier, S.; Suzenet, F.; Tatibouët, A.;
Rauter, A. P.; Rollin, P. Tetrahedron Lett. 2008, 49, 5583–5586.
doi:10.1016/j.tetlet.2008.07.023
25.Berionni, G.; Maji, B.; Knochel, P.; Mayr, H. Chem. Sci. 2012, 3,
878–882. doi:10.1039/C2SC00883A
26.Lennox, A. J. J.; Lloyd-Jones, G. C. Chem. Soc. Rev. 2014, 43,
412–443. doi:10.1039/C3CS60197H
27.Sprinz, J.; Kiefer, M.; Helmchen, G.; Reggelin, M.; Huttner, G.;
Walter, O.; Zsolnai, L. Tetrahedron Lett. 1994, 35, 1523–1526.
doi:10.1016/S0040-4039(00)76748-7
28.Steinhagen, H.; Reggelin, M.; Helmchen, G. Angew. Chem. 1997, 109,
2199–2202. doi:10.1002/ange.19971091921
Angew. Chem. Int. Ed. Engl. 1997, 36, 2108-2110.
doi:10.1002/anie.199721081
29.Ramillien, M.; Vanthuyne, N.; Jean, M.; Gherase, D.; Giorgi, M.;
Naubron, J.-V.; Piras, P.; Roussel, C. J. Chromatogr. A 2012, 1269,
82–93. doi:10.1016/j.chroma.2012.09.025
171