N-Alkyl oxazolidines as stereocontrol elements in asymmetric Diels-Alder cycloadditions of 9-substituted anthracene derivatives

Article abstract of DOI:10.1039/b610055d

Chiral 9-oxazolidinyl anthracene derivatives have been prepared as single diastereoisomers by condensation of 9-anthraldehyde with the appropriate N-alkyl amino alcohol. Asymmetric Diels-Alder cycloadditions of these with N-methyl maleimide proceeds in go

Full text of DOI:10.1039/b610055d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
N-Alkyl oxazolidines as stereocontrol elements in asymmetric Diels–Alder
cycloadditions of 9-substituted anthracene derivatives†
Harry Adams, Ramadan A. Bawa and Simon Jones*
Received 13th July 2006, Accepted 15th September 2006
First published as an Advance Article on the web 2nd October 2006
DOI: 10.1039/b610055d
Chiral 9-oxazolidinyl anthracene derivatives have been prepared as single diastereoisomers by
condensation of 9-anthraldehyde with the appropriate N-alkyl amino alcohol. Asymmetric Diels–Alder
cycloadditions of these with N-methyl maleimide proceeds in good yield and in good
diastereoselectivity, the sense of which may be controlled by judicious choice of the N-alkyl group.
comparable levels of diastereoselectivity to those as previously
Introduction
observed. Previous attempts at using the 2-(anthr-9-yl)oxazolinyl
The use of chiral auxiliaries for achieving stereocontrol has been
anthracene have shown that the stereogenic centre in this species
popularised through the work of groups such as Evans and
lies too far away from the reactive centre to exert any degree of
stereocontrol in the Diels–Alder cycloaddition.5 Therefore, it was
proposed that if another stereogenic centre was installed closer to
the anthracene ring, it should enhance the selectivity. Thus, the
2-(anthr-9-yl)oxazolidine derivatives 3, 4 and 5 were proposed as
appropriate candidates for investigating this aspect. Each of these
would allow us to probe the role of the nitrogen substituent in
this reaction, with oxazolidine 3 having the potential for hydrogen
bonding, while the other analogues 4 and 5 have protecting groups
with very different steric properties.
Oppolzer, and is now widely used in the academic and indus-
trial organic chemistry communities.1 The plethora of synthetic
transformations that may be performed on the auxiliary–substrate
adduct is primarily determined by the group that is attached to
the auxiliary. Thus, Evans’ auxiliary is usually attached to an
acid chloride through formation of an acyl bond and the range
of synthetic transformations that can be performed may then
be considered as being dependant upon formation of this type
of bond. Thus, enolate alkylation, aldol addition and conjugate
addition can all be performed in excellent diastereoselectivity
and good yield.2 However, addition of Grignard or alkyl lithium
reagents can often result in complex mixtures, including addition
at the oxazolidinone carbonyl group.3 If a different attachment
strategy could be employed that makes a different type of bond,
then other synthetic transformations not normally amenable to
regular auxiliaries would become available.
Based on this approach, research by this group and others
has described the use of chiral 9-substituted anthracenes such
as 9-methoxyethyl anthracene 1 and 9-trifluoromethoxyethyl
anthracene 2 as a new class of chiral auxiliary.4 Here diastereose-
lective addition of an alkene occurs across the 9 and 10 positions
of the anthracene moiety, the selectivity of which is controlled
by the stereogenic centre pendant from the anthracene. Cleavage
of the auxiliary from the substrate can then be achieved through
photolysis and/or flash vacuum pyrolysis. In previous studies we
and others had to resort to introduction of the stereogenic centre
of the anthracene moiety by asymmetric synthesis. Although this
method gave the desired targets in good enantiomeric excess,
we were interested in developing methods to install stereogenic
elements directly. This should greatly simplify the preparation of a
suitable auxiliary, however the stereogenic elements should be able
to control the diastereoselectivity of the alkene cycloaddition with
There is good precedent for 1,3-oxazolidines controlling stere-
oselective transformations of adjacent sp2 hybridised atoms. 1,3-
Oxazolidines were first used for stereoselective functionaliza-
tion of alkenes by Abdallah et al. in the palladium-catalysed
cyclopropanation of an ephedrine-derived oxazolidine in good
yield and excellent diastereoselectivity (>90%).6 Since then,
other stereoselective transformations include conjugate addi-
tion of cuprates,7 dihydroxylation,8 and intra-molecular bromo-
lactonisation.9 However, the use of chiral oxazolidines to control
the stereochemistry in Diels–Alder cycloadditions is rather lim-
ited. While chiral dienophiles bearing oxazolidines have been used
as stereocontrol elements in Diels–Alder cycloadditions showing
high levels of diastereoselectivity,10 only a single report of the use
of chiral oxazolidinyl dienes has been reported in the synthesis of
hydroxylated piperidines, although the stereoselectivity observed
was rather modest.11
Department of Chemistry, University of Sheffield, Dainton Building, Brook
Hill, Sheffield, UK. E-mail: simon.jones@sheffield.ac.uk; Fax: +44 (0)114
222 9346; Tel: +44 (0)114 222 9346
† Electronic supplementary information (ESI) available: Copies of NMR
spectra for all compounds, experimental procedures for the preparation of
compounds 8, 9, 10 and 11, and X-ray crystallographic information. See
DOI: 10.1039/b610055d
4206 | Org. Biomol. Chem., 2006, 4, 4206–4213
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
In this work we report the approaches to the synthesis of
oxazolidines 3–5 and evaluation in Diels–Alder reactions with
N-methyl maleimide.
Results
Condensation between anthracene 9-carboxaldehyde and (S)-
phenylalaninol furnished the 9-anthrylimino alcohol 6 in an
excellent yield of 96% (Scheme 1). This imino alcohol did not
cyclise to form the desired oxazolidine 3 using a range of acidic
catalysts including Amberlyst, ZnCl2, AlCl3, TMSCl and HCl.
In all cases, starting material was returned and in some cases
9-anthraldehyde was also observed. This probably results from
a series of competing reactions, whereby the rate of cyclisation
is slow compared to the rate of hydrolysis and/or iminium ion
formation. As an alternative approach, 9-[4-(S)-benzyloxazolinyl]
anthracene 7, prepared using a previously reported procedure,5
was subjectedto catalytichydrogenation to reduce the double bond
to obtain the target oxazolidine. However, only starting material
was obtained. Attempts to hydrogenate oxazolidine 5 (see later)
also proved unsuccessful.
Scheme 2 Preparation of oxazolidine 4. Reagents and conditions: (a)
SOCl2, MeOH, rt, overnight; (b) ClCO2Me, NaHCO3, H2O, 0 ◦C→rt,
overnight; (c) LiAlH4, THF, reflux, overnight; (d) anthracene 9-carbox-
aldehyde, MgSO4, THF, rt, 6 h.
The relative stereochemistry around C2 of the oxazolidine ring
was confirmed by single crystal structure and 1H NMR studies (see
below) showing that the ring closure led to the syn isomer with
the anthryl, benzyl and methyl groups all adopting equatorial
positions. This preference is the same as has been previously
observed,8,9 being attributed to the greater thermodynamic sta-
bility of the cis isomer in such five-membered systems.12
The corresponding N-benzyl derivative 5 was prepared via
a similar procedure. Condensation of benzaldehyde with (S)-
phenylalaninol in dry methanol followed by reduction using
sodium borohydride formed the target aminoalcohol 11 in ex-
cellent yield (99%). This was then reacted with anthracene 9-
carboxaldehyde in dry THF in the presence of MgSO4 as before,
however, only 30% conversion was observed after 48 h stirring
at rt, probably due to the increased steric hindrance caused by
the benzyl group. Thus, this reaction was repeated using a Lewis
acid, Mg(OTf)2, in dry CH2Cl2, reaching completion within 48 h
at rt yielding an 85 : 15 mixture of diastereoisomers that was
recrystallised to provide the target oxazolidine 5 in 80% yield as a
single diastereoisomer (Scheme 3).
Scheme 1 Attempted preparation of oxazolidine 3. Reagents and condi-
tions: (a) MgSO4, THF, rt, 4 h; (b) see text; (c) H2, Pd/C, EtOAc or H2,
Pd/C, HOAc, rt, 8–24 h.
Attention was then turned to the N-alkyl oxazolidines and in the
first instance the N-methyl derivative. It was envisioned that the
use of these compounds would lead to the in situ formation of an
iminium ion, increasing the electrophilicity of the carbon centre,
thus facilitating nucleophilic addition of the alcohol. N-Methyl-
(S)-phenylalaninol 10 required for this synthesis was prepared
in a three-step synthetic route starting with treatment of (S)-
phenylalanine with thionyl chloride in dry methanol to form
(S)-phenylalanine methyl ester hydrochloride 8 in excellent yield
(97%). This was reacted with methyl chloroformate forming the
N-methoxyformyl-(S)-phenylalanine methyl ester 9, which was
further reduced to the N-methyl-(S)-phenylalaninol, both steps
proceeding once again in excellent yield (90%). The target oxazo-
line 4 was then easily prepared via a condensation reaction between
anthracene 9-carboxaldehyde and N-methyl-(S)-phenylalaninol
giving a single diastereoisomer of the target compound in excellent
yield (Scheme 2).
Scheme 3 Preparation of oxazolidine 5. Reagents and conditions: (a)
MeOH, rt, 1 h; (b) NaBH4, 0 ◦C→rt, 2 h; (c) anthracene 9-carboxaldehyde,
Mg(OTf)2, CH2Cl2, rt, 48 h.
The relative stereochemistry of the major isomer around the C2
of the oxazolidine ring was confirmed by single crystal structure,
again revealing that ring closure took place to provide the syn
isomer as in the case of the formation of N-methyloxazolidine
4. As before, all of the substituents adopt equatorial positions
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 4206–4213 | 4207
©

View Article Online
minimising 1,3-diaxial interactions, however, the N-benzyl group
now occupied a position over the plane of the anthracene ring
probably as a result of p–p interactions. The slightly reduced
diastereomeric ratio probably results from increased steric in-
teractions of the close proximity of two benzyl groups and the
anthracene ring system.
The N-methyl oxazolidine 4 showed slightly reduced reactivity
in the Diels–Alder cycloaddition with 1 eq. of N-methyl maleimide
in toluene at 80–85 ◦C, giving 60% conversion after 2 h, compared
to 9-methoxyethyl anthracene 1, which gave 100% conversion
under the same conditions. The cycloaddition gave a mixture of
two diastereoisomers in a ratio of 85 : 15, the ratio of which did
not vary with either time or temperature (Scheme 4).
Alder reaction between the N-methyl oxazolidine 4 and N-methyl
maleimide was irreversible and thus kinetically controlled.
The preferred conformation of the oxazolidine 4 would most
likely be such that the aminal hydrogen atom lies parallel to
the anthracene ring system, thus avoiding peri-interactions of
the nitrogen and oxygen substituents of the oxazolidine with the
anthracene ring (Fig. 1). Approach of the dienophile to either
face occurs such as to minimise electrostatic repulsion between the
carbonyl of the N-methyl maleimide and either nitrogen or oxygen
atom in an analogous fashion to the model already generally
accepted for this type of cycloaddition.4c,e,i Approach A is then
favoured since the other face experiences greater steric repulsion
from the large N-methyl group compared to the electrostatic
repulsion from the oxygen atom (Fig. 1).
In order to elucidate the structure of diastereisomer 13, attempts
were made to prepare both cycloadducts 12 and 13 via another
synthetic route. The product of the Diels–Alder reaction between
anthracene 9-carboxaldehyde and N-methyl maleimide rac-14 was
condensed with N-methyl-(S)-phenylalaninol 10 in the presence
of MgSO4 leading to the formation of a mixture of oxazolidines
(Scheme 5). However, the 1H NMR spectrum of the crude product
revealed the formation of two diastereoisomers 15 and 16 in a
60 : 40 ratio, which were not the same as those observed when
the N-methyl oxazolidine 4 was reacted with N-methyl maleimide
in toluene, implying the formation of a product with different
stereochemistry around C2 of the oxazolidine ring. The sensitivity
Scheme 4 Diels–Alder addition of N-methyl oxazolidine 4. Reagents and
conditions: (a) 1.0 eq., N-methyl maleimide, toluene, 80–85 ◦C.
Although the conversion to product could be improved by
heating at 90–95 ◦C for 4 h (89%) with no change in diastere-
oselectivity, no improvement in yield or selectivity was obtained
at room temperature with or without Lewis acid-catalysis using
Cu(OTf)2–BEN (N,N-dibenzylidene ethylene diamine).
The major diastereoisomer 12 was isolated by diffusion re-
crystallisation from CH2Cl2–petroleum ether 40–60 ◦C and its
structure was determined by single crystal X-ray diffraction
1
and H NMR studies (see later). As in the case of the parent
oxazolidine 4, all ring substituents occupied equatorial positions.
After recrystallisation, the mother liquor was found to be enriched
with the minor isomer 13 and was thus stirred with an excess of
N-methyl maleimide in toluene for 2 h at 80–85 ◦C showing no
change in the diastereomeric ratio demonstrating that the Diels–
Scheme 5 Preparation of diastereoisomers 15 and 16. Reagents and
conditions: (a) N-Methyl maleimide, toluene, 80–85 ◦C, 24 h; (b) MgSO4,
THF, rt, 19 h.
Fig. 1 Rationale for the selectivity observed in the Diels–Alder cycloaddition of N-methyl oxazolidine 4 with N-methyl maleimide.
4208 | Org. Biomol. Chem., 2006, 4, 4206–4213 This journal is The Royal Society of Chemistry 2006
©

View Article Online
Table 1 Interconversion of diastereoisomers 15 and 16 to diastereoiso-
mers 12 and 13
Ratio (%)a
Entry
Time (day)
12
13
15
16
1
2
3
4
5
6
7
8
0
7
9
13
15
21
24
33
0
16
16
29
37
45
48
53
0
16
22
32
35
37
38
38
60
49
46
28
21
13
9
40
19
16
10
7
5
4
7
2
a Ratio was calculated from the integrals of appropriate signals in the
1H NMR spectrum.
Scheme
6
Formation of aldehyde (−)-14 and condensation with
of the crude product to silica gel chromatography prevented the
isolation of these diastereoisomers as it decomposed to form
the starting aldehyde rac-14. Deactivating the silica gel using
triethylamine also led to decomposition. However, attempted
separation of these isomers by recrystallisation from CH2Cl2–
petroleum ether 40–60 ◦C at rt gave the thermodynamically
favoured diastereoisomer 12.
N-methyl phenylalaninol. Reagents and conditions: (a) SiO2, CH2Cl2–H2O
(9 : 1; v/v), rt, 18 h; (b) N-methyl-(S)-phenylalaninol 10, MgSO4, THF, rt,
19 h; (c) 12 days, CDCl3.
This interconversion was studied further using in situ NMR
studies. The mixture of diastereoisomers 15 and 16 from the
reaction between rac-14 and N-methyl-(S)-phenylalaninol were
left in a CDCl3 solution and analysed over a period of time
(Table 1). Isomers 15 and 16 slowly reverted to a 1 : 1 mixture
of isomers 12 and 13, supporting the thought that the outcome
of the reaction between the adduct rac-14 and N-methyl-(S)-
phenylalaninol 10 was kinetically controlled, forming the trans
oxazolidine adducts 15 and 16, which slowly interconverted to the
thermodynamically more stable cis cycloadducts 12 and 13 over a
prolonged period of time.
Fig. 2 Structures depicting N-methyl oxazolidine NMR spin system.
signals with large and small coupling constants. However, in the
case of the major isomer 12, the protons H5a and H5b were seen in
the 1H NMR spectrum as two apparent triplets or double doublets
with large coupling constants. A molecular modelling study was
carried out on the adducts 12 and 15 to measure the dihedral
angles between hydrogen atoms (H4 and H5a; H5b in both systems)
(Fig. 2, Table 2).13
Themolecular modelling onaddu◦ct 15 revealed that thedihedral
angle between H4 and H5b was 117.2 , whereas between H4 and H5a
was 7.3◦. According to the Karplus correlation, the corresponding
coupling constants for such angles should be small and big
respectively (Table 2). This is in fact what was observed in the
1H NMR spectrum of the crude material. Likewise, according to
the data obtained from the single crystal structure of the major
isomer 1◦2, the dihedral angle between H4 and H5b was found to
be 142.3 , whereas between H4 and H5a was 21.8◦. These values
Since the poor selectivity observed in this reaction is likely to
be a consequence of the N-methyl phenylalaninol 10 effectively
acting as a resolving agent, a single diastereoisomer of product
would probably be obtained if enantiomerically pure aldehyde
14 was used. Therefore, pure diastereoisomer 12 was stirred for
18 h with silica gel in CH2Cl2–H2O (9 : 1; v/v). This produced
the enantiomerically pure aldehyde adduct (−)-14 in 73% yield
that was involved in a condensation reaction with N-methyl-
(S)-phenylalaninol 10 in dry THF in the presence of MgSO4
giving 78% conversion to the corresponding oxazolidine adduct
15 (Scheme 6).
1
The H NMR spectrum of the crude material revealed that
a single diastereoisomer was formed which was the same as the
major isomer formed when the rac-14 was reacted with N-methyl-
(S)-phenylalaninol 10. This diastereoisomer 15 interconverted
to isomer 12 after 14 days standing in CDCl3 and diffusion
recrystallisation from CH2Cl2–petroleum ether 40–60 ◦C gave a
53% yield of this isomer.
Table 2 Observed and calculated bond angles and coupling constants for
diastereoisomers 12 and 15
◦
◦
a
b
Isomer
12
HC–CH2
hcalc
/
hobs
/
J
pred/Hzc
J
obs/Hzd
H4–H5a
H4–H5b
H4–H5a
H4–H5b
11.9
21.8
142.3
—
7–9
4–6
8–9
1–2
7.8
7.8
13.4
1.2
Further confirmation of the cis configuration of the oxazolidine
ring came from comparison of the observed and calcuated dihedral
angles with the predicted magnitude of the coupling constants.
Thus, when both aldehydes rac-14 and (−)-14 were condensed
135.9
7.3
117.2
15
—
1
a Calculated using molecular modelling. b Measured from crystal structure.
c Predicted range from the Karplus correlation. d Observed values from the
1H NMR spectra.
with N-methyl-(S)-phenylalaninol 10, the H NMR spectrum of
the crude diastereoisomers in both cases revealed the appearance
of the two protons H5a and H5b (Fig. 2), as two double doublet
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 4206–4213 | 4209
©

View Article Online
correlated well with the calculated ones and correspond to large
coupling constants according to the Karplus prediction.
Likewise, the N-benzyl oxazolidine 5 was involved in a ther-
mal Diels–Alder reaction with N-methyl maleimide in toluene,
showing slightly increased reactivity towards the Diels–Alder
cycloaddition with N-methyl maleimide (81% conversion) after
2 h stirring at 80–85 ◦C, when compared to the reactivity of the N-
methyl counterpart 4 (60% conversion) under the same conditions.
However, this time formation of a single diastereoisomer 17 was
observed with the opposite sense of stereoselection as the N-
methyl counterpart, as confirmed by single crystal X-ray analysis
(Scheme 7).
Fig. 3 Structure depicting N-benzyl oxazolidine NMR spin system.
ring. This effect is relayed to the benzyl stereodirecting group
through 1,2-tortional interactions, ultimately leading to a high
energy transition state. Unable to react via this mode of approach,
the oxazolidine ring undergoes a ring opening–closure process to
generate the much less stable trans-oxazolidine. This once again
has one face of the anthracene blocked by the N-benzyl group and
incurs interaction of this group with the deforming anthracene
ring, but now does not suffer the same 1,2 tortional interactions
since the benzyl stereodirecting group is axial. Cycloaddition
followed by epimerisation of the oxazolidine ring cannot occur
since this would lead to the opposite sense of stereochemistry of
the cycloaddition reaction.
Unfortunately it was not possible to access the stereoisomers
of this reaction by condensation of anthracene 9-carboxaldehyde
cycloadduct rac-14 and N-benzyl-(S)-phenylalaninol 11 in the
presence of Mg(OTf)2, presumably due to the steric effects due
to the bulky amino alcohol being used.
Scheme 7 Diels–Alder cycloaddition of N-benzyl oxazolidine 5 with
N-methyl maleimide. Reagents and conditions: (a) toluene, 90–95 ◦C, 4 h.
Conclusion
We have demonstrated that chiral oxazolidines may be quickly and
selectively prepared from readily available N-alkyl amino alcohols
and that this class of compound undergoes highly stereoselective
cycloadditions. The sense of diastereoselection of this reaction
depends critically upon the N-alkyl group employed. Although the
level of diastereocontrol is excellent and shows that oxazolidines
may be used effectively to control the outcome of Diels–Alder
cycloadditions, their use in our auxiliary chemistry is limited,
since they appear to easily decompose to afford the corresponding
aldehydes.
Both the anthryl moiety and N-benzyl group adopted equatorial
conformations to the oxazolidine ring, while the other benzyl
group was enforced to orientate axially to the oxazolidine ring.
Most importantly, the stereochemistry around C2 changed from S
to R configuration, which was rather unusual in such ring systems
as the cis configuration is usually more thermodynamically
stable. Additionally, the coupling constants for the oxazolidine
ring system followed that of the minor isomer formed in the
cycloadditions of N-methyl oxazolidine 4 (Fig. 3, Table 3).
The stereoselectivity of this reaction was believed to follow
the same model as for the N-methyl oxazolidine (Fig. 4), but
since the N-benzyl group is much larger it effectively shuts off
cycloaddition via approach B, exposing only one face to react,
the stereochemistry of which was once again determined using
an electrostatic repulsion model. However, a consequence of this
mode of approach is that the pseudo axial N-benzyl group now
suffers severe steric interaction with the deforming anthracene
Experimental
All solvents used were freshly dried over sodium except CH2Cl2
which was dried over LiAlH4. Et3N was distilled over KOH. Glass-
ware was flame dried and cooled in vacuo before use and all
reactions were carried out under nitrogen unless otherwise stated.
TLC was carried out using Merk aluminium TLC sheets (silica
gel 60 F254). Visualisation of the TLC plates was carried out using
a UV lamp or by dipping in KMnO4 then exposure by heating.
Flash column chromatography was carried out with Fluorochem
Limited Silica Gel 40–63u 60A.
Table 3 Observed and calculated bond angles and coupling constants for
diastereoisomer 17
◦
◦
a
b
Isomer
HC–CH2
hcalc
/
hobs
/
J
pred/Hzc
J
obs/Hzd
Melting points were measured on a Gallenkamp apparatus and
are uncorrected. Specific rotations were performed on an Optical
Activity LTD. AA-10 automatic polarimeter at 589 nm (Na D-line)
and measured at 20◦C unless otherwise stated. [a]D values are given
in 10−1 deg cm2 g−1. All infrared spectra were recorded on a Perkin
Elmer Spectrum RX/FT-IR system with a DuraSamplIR II ATR
17
H4–H5a
H4–H5b
10.0
114.2
22.5
98.2
7–9
0–2
6.4
4.2
a Calculated using molecular modelling. b Measured from crystal structure.
c Predicted range from the Karplus correlation. d Observed values from the
1H NMR spectra.
4210 | Org. Biomol. Chem., 2006, 4, 4206–4213
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Fig. 4 Model for the diastereoselectivity of oxazolidine 5.
1
accessory. 250 MHz H NMR and 62.5 MHz 13C NMR were
CH2OH), 7.28–7.47 (9H, m, ArCH), 7.89–8.00 (4H, m, ArCH),
=
carried out on an AC-250 supported by an Aspect 200 data system.
Residual proton signals from the deuteriated solvents were used
as references [chloroform (1H 7.25 ppm, 13C 77 ppm)]. Coupling
constants were measured in Hz. Mass spectra were recorded on a
Micromass Autospec M spectrometer. Elemental microanalysis
was performed using a Perkin Elmer 2400 CHNS/0 Series II
elemental analyzer.
8.43 (1H, s, ArCH), 9.03 (1H, s, N CH); dC(62.5 MHz; CDCl3;
Me4Si) 38.8 (CH2), 66.4 (CH2O), 76.0 (CH), 124.9 (CH), 125.2
(CH), 126.4 (CH), 126.5 (CH), 128.6 (C), 128.7 (CH), 128.8 (CH),
129.1 (CH), 129.8 (CH), 131.2 (C), 138.7 (C), 162.2 (C); m/z (EI)
339.1620 (M+. C24H21NO requires 339.1623), 248 (100%), 191 (17),
178 (18), 91 (74), 77 (12).
Data collected were measured on a Bruker Smart CCD area
detector with Oxford Cryosystems low temperature system. Single
crystals of compounds 4, 5, 12 and 17 were grown as described,
mounted in inert oil and transferred to the cold gas stream of
the diffractometer. In all cases, the relative stereochemistry of
the compounds was recorded and cross referenced to the known
stereogenic centre.
(2S,4S)-2-(9-Anthracenyl)-3-methyl-4-benzyl oxazolidine 4
A solution of (S)-N-methylphenylalaninol 1015 (0.627 g, 3.8 mmol)
in dry THF (30 cm3) was added dropwise at room temperature
to a stirred mixture of anthracene-9-carboxaldehyde (0.782 g,
3.8 mmol) and MgSO4 (0.5 g, 4.86 mmol) in dry THF (20 cm3).
The resulting mixture was stirred for a further 6 h at rt, filtered
through Celite and the solvent was removed to give a yellow
solid of the target oxazolidine 4 (1.325 g, 99%). A sample was
recrystallised from toluene for analytical purposes. Mp 204–209 ◦C
(from toluene); (found: C, 84.9; H, 6.45; N, 3.85. C25H23NO
requires C, 84.95; H, 6.6; N, 4.0%); [a]2D6 +18.0 (c 0.5 in CHCl3);
[N-(9-Anthracenylmethylene)-(S)-phenylalaninol 6
A solution of anthracene-9-carboxaldehyde (0.567 g, 2.75 mmol)
in dry THF (20 cm3) was added dropwise at room temperature
to a stirred mixture of (S)-phenylalaninol14 (0.415 g, 2.75 mmol)
and MgSO4 (0.5 g, 4.86 mmol) in dry THF (10 cm3). The resulting
mixture was stirred for 4 h at room temperature, filtered through
Celite and the solvent was removed to obtain a yellow solid of the
title compound 6 (0.896 g, 96%). A sample of the hydroxyl imine
was recrystallised from EtOAc for analytical purposes. Mp 178–
180 ◦C (from EtOAc); (found: C, 84.7; H, 6.2; N, 4.01. C24H21NO
requires C, 84.9; H, 6.2; N, 4.1%); [a]2D5 −70.0 (c 0.5 in CHCl3);
mmax/cm−1 1624 (C C); dH(250 MHz; CDCl3; Me4Si) 2.23 (3H, s,
=
CH3), 2.98–3.26 (3H, m, PhCH2 and CHN), 4.15–4.26 (2H, m,
CH2O), 6.24 (1H, s, NCHO), 7.30–7.49 (9H, m, ArCH), 7.99–8.03
(2H, m, ArCH), 8.49 (1H, s, ArCH), 8.75–8.79 (2H, m, ArCH);
dC(62.5 MHz; CDCl3; Me4Si) 36.1 (CH3), 37.9 (CH2), 66.6 (CH),
70.9 (CH2), 94.7 (CH), 124.8 (CH), 124.9 (CH), 125.6 (CH), 126.7
(CH), 126.9 (C), 128.6 (CH), 129.1 (CH), 129.5 (CH), 129.8 (CH),
131.5 (C), 138.1 (C); m/z (EI) 354.1851 (M+. C25H23NO requires
354.1858), 353 (10%), 262 (100), 232 (25), 178 (35), 91 (30).
mmax/cm−1 1642 (C N); dH(250 MHz; CDCl3; Me4Si) 2.14 (1H,
=
br s, OH), 3.05 (1H, dd, J 13.6 and 9.3, CHHPh), 3.18 (1H, dd, J
13.6 and 4.3, CHHPh), 3.46–3.97 (1H, m, CHN), 4.02 (2H, br s,
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 4206–4213 | 4211
©

View Article Online
Crystal structure determination of compound 4‡
left stirring for a further 4 h at 90–95 ◦C, cooled to room temper-
ature and the solvent was removed under reduced pressure to give
the target compound as two diastereoisomers in a ratio of (85 :
15) (0.246 g, 89% conversion). A sample was recrystallised from
CH2Cl2–petroleum ether 40–60 ◦C (diffusion recrystallisation) to
afford a white solid of the major diastereois◦omer 12. Mp 215–
217 ◦C (from CH2Cl2–petroleum ether 40–60 C); [a]2D5 +25.0 (c 1
Crystal data. C25H23NO, M = 353.44, monoclinic, a =
3
˚
˚
18.188(3), b = 7.0490(10), c = 17.257(2) A, U = 1891.7(5) A ,
T = 150(2) K, space group C2, Z = 4, l(Mo–Ka) = 0.075 mm−1,
10741 reflections measured, 2328 unique (Rint = 0.0406) which
were used in all calculations. The final wR(F2) was 0.0943 (all
data).
in CHCl3); mmax/cm−1 1692 (C O); dH(250 MHz; CDCl3; Me4Si)
=
2.41 (3H, s, CH3), 2.54 (3H, s, CH3), 2.93 (1H, dd, J 13.5 and 8.2,
CHHPh), 3.05 (1H, dd, J 8.4 and 3.3, CHHPh), 3.15–3.33 (2H,
m, PhCHC(O)CH and CHN), 3.72 (1H, d, J 8.4, C(O)CH), 3.92
(1H, app t, J 7.8, CHHO), 4.15 (1H, app t, J 7.8, CHHO), 4.69
(1H, d, J 3.3, PhCHCH), 5.84 (1H, s, NCHO), 7.03–7.39 (11H,
m, ArCH), 7.53–7.60 (1H, m, ArCH), 8.61 (1H, d, J 7.9, ArCH);
dC(62.5 MHz; CDCl3; Me4Si) 24.1 (CH3), 37.4 (CH2), 37.6 (CH3),
46.8 (CH), 48.0 (CH), 49.2 (CH), 49.8 (C), 66.4 (CH), 69.3 (CH2),
94.0 (CH), 123.4 (CH), 124.1 (CH), 125.1 (CH), 125.9 (CH),
126.1 (CH), 126.3 (CH), 126.6 (CH), 126.8 (CH), 128.1 (CH),
128.5 (CH), 128.7 (C), 129.4 (CH), 137.8 (C), 138.4 (C), 138.9
(C), 141.9 (C), 175.8 (C), 176.7 (C); m/z (ES) 465.2168 (MH+.
C30H29N2O3 requires 465.2178).
(2S,4S)-2-(9-Anthracenyl)-3,4-dibenzyl oxazolidine 5
A solution of N-benzylphenylalaninol 1116 (0.76 g, 3.15 mmol) in
dry CH2Cl2 (10 cm3) was added dropwise at room temperature to a
stirred mixture of anthracene-9-carboxaldehyde (0.5 g, 2.43 mmol)
and Mg(OTf)2 (0.08 g, 0.248 mmol) in dry CH2Cl2 (10 cm3). The
resulting mixture was stirred for 48 h at room temperature, filtered
through Celite and the solvent was removed to give a yellow solid
as a mixture of diastereoisomers (85 : 15) which was recrystallised
from CH2Cl2–petroleum ether 40–60 ◦C to provide a single isomer
of the desired oxazolidine 5 (0.91 g, 85%). Mp 101–103 ◦C (from
CH2Cl2–petroleum ether 40–60 ◦C); (found: C, 86.8; H, 6.6; N,
3.2. C31H27NO requires C, 86.7; H, 6.3; N, 3.3%); [a]2D5 −22.0 (c 1
1
in CHCl3); mmax/cm−1 1491 and 1452 (C C); dH(250 MHz; CDCl3;
Selected H NMR signals for the minor diastereoisomer 13.
dH(250 MHz; CDCl3; Me4Si) 3.66 (1H, d, J 7.9, C(O)CH), 4.66
(1H, d, J 3.4, PhCHCH), 5.82 (s, 1H).
=
Me4Si) 2.79 (1H, dd, J 13.4 and 9.5, CHHPh), 2.89 (1H, dd, J
13.4 and 4.3, CHHPh), 3.36–3.50 (1H, m, CHN), 3.62 (1H, d,
J 13.7, CHHPh), 3.75 (1H, d, J 13.7, CHHPh), 4.15 (1H, app
t, J 7.7, CHHO), 4.22 (1H, app t, J 7.7, CHHO), 6.48 (1H, s,
NCHO), 6.86–6.99 (4H, m, ArCH), 7.12–7.76 (10H, m, ArCH),
7.97 (2H, d, J 7.9, ArCH), 8.42 (1H, s, ArCH), 8.91 (2H, d, J 8.2,
ArCH); dC(62.5 MHz; CDCl3; Me4Si) 40.2 (CH2), 55.1 (CH2),
65.6 (CH), 71.5 (CH2), 93.5 (CH), 124.7 (CH), 124.8 (CH), 125.1
(CH), 125.3 (CH), 125.5 (CH), 125.8 (CH), 126.4 (CH), 126.7
(CH), 127.6 (CH), 128.0 (CH), 128.6 (CH), 128.7 (CH), 129.0
(CH), 129.1 (CH), 129.3 (CH), 129.9 (CH), 131.5 (C), 131.6 (C),
138.2 (C), 138.7 (C); m/z (EI) 429.2091 (M+. C31H27NO requires
429.2093), 429 (9%), 338 (93), 269 (6), 252 (7), 206 (8), 178 (28),
157 (8), 91 (100).
Crystal structure determination of compound 12‡
Crystal data. C31H30Cl2N2O3, M = 549.47, monoclinic, a =
3
˚
˚
7.6834(10), b = 10.5554(14), c = 16.299(2) A, U = 1319.4(3) A ,
T = 150(2) K, space group P21, Z = 2, l(Mo–Ka) = 0.283 mm−1,
11779 reflections measured, 3123 unique (Rint = 0.0224) which
were used in all calculations. The final wR(F2) was 0.0766 (all
data).
(3aS,9aS)-3a,4,9,9a-Tetrahydro-4-[(2R,4S)-3,5-dibenzyl-
2-oxazolidinyl]-2-methyl-4,9[1ꢀ,2ꢀ]-benzeno-1H-benz[f ]isoindole-
1,3(2H)-dione 17
Selected 1H NMR signals for the minor diastereoisomer.
dH(250 MHz; CDCl3; Me4Si) 3.14 (1H, dd, J 12.2 and 4, CHHPh),
3.92 (1H, dd, J 8.5 and 3.6, CHHO), 4.40 (1H, dd, J 8.5 and 6.4,
CHHO), 6.87 (1H, s, NCHO), 8.50 (1H, s, ArCH).
N-Methyl maleimide (0.129 g, 1.160 mmol) was added in one
portion as a solid at 90–95 ◦C to a stirred solution of oxazolidine
5 (0.5 g, 1.16 mmol) in toluene (10 cm3). The resulting mixture
was left stirring for 4 h at 90–95 ◦C, cooled to room temperature
and the solvent was removed under reduced pressure to give a
single diastereoisomer of the target cycloadduct (0.47 g, 75%).
A sample was recrystallised from CH2Cl2–petroleum ether 40–
60 ◦C (diffusion recrystallisation) to afford a white solid of the
title compound 17. Mp 204–207 ◦C from CH2Cl2–petroleum ether
40–60 ◦C); (found: C, 79.7; H, 5.9; N, 5.1. C36H32N2O3 requires C,
80.0; H, 6.0; N, 5.2%); [a]2D5 +66.1 (c 1 in CHCl3); mmax/cm−1 1688
Crystal structure determination of compound 5‡
Crystal data. C31H27NO, M = 429.54, triclinic, a = 7.284(3),
3
˚
˚
b = 9.462(3), c = 18.426(6) A, U = 1144.2(7) A , T = 293(2) K,
space group P1, Z = 2, l(Mo–Ka) = 0.074 mm−1, 12849 reflections
measured, 5114 unique (Rint = 0.0593) which were used in all
calculations. The final wR(F2) was 0.1422 (all data).
(3aR,9aR)-3a,4,9,9a-Tetrahydro-4-[(2S,4S)-5-benzyl-3-methyl-2-
oxazolidinyl]-2-methyl-4,9[1ꢀ,2ꢀ]-benzeno-1H-benz[f ]isoindole-
1,3(2H)-dione 12
=
=
(C O), 1454, 1428 and 1376 (C C); dH(250 MHz; CDCl3; Me4Si)
2.50 (3H, s, CH3), 2.81 (1H, dd, J 13.4 and 10.0, CHHPh), 3.01
(1H, dd, J 13.4 and 5.1, CHHPh), 3.10 (1H, dd, J 8.3 and 3.1,
PhCHC(O)CH), 3.60 (1H, d, J 8.3, C(O)CH), 3.63–3.71 (3H, m,
CH2Ph and CHN), 3.83 (1H, dd, J 8.3 and 4.2, CHHO), 4.15
(1H, dd, J 8.3 and 6.4, CHHO), 4.77 (1H, d, J 3.1, PhCHCH),
6.55 (1H, s, NCHO), 7.00–7.04 (2H, m, ArCH), 7.06–7.46 (14H,
m, ArCH), 8.10 (1H, dd, J 7.0 and 1.8, ArCH), 8.17 (1H, dd, J 6.7
and 2.2, ArCH); dC(62.5 MHz; CDCl3; Me4Si) 24.3 (CH3), 38.4
(CH2), 46.6 (CH), 48.3 (CH), 49.8 (CH), 50.3 (CH2), 56.6 (CH2),
N-Methyl maleimide (0.083 g, 0.75 mmol) was added in one
portion as a solid at 90–95 ◦C to a stirred solution of oxazolidine 4
(0.177 g, 0.5 mmol) in toluene (3 cm3). The resulting mixture was
‡ CCDC reference numbers 614833 (4), 614834 (5), 614835 (12) and 614836
(17). For crystallographic data in CIF or other electronic format see DOI:
10.1039/b610055d
4212 | Org. Biomol. Chem., 2006, 4, 4206–4213
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
64.5 (CH), 67.8 (CH2), 90.9 (CH), 123.5 (CH), 125.0 (CH), 125.1
(CH), 125.7 (CH), 126.2 (CH), 126.7 (CH), 127.0 (CH), 127.3
(CH), 128.0 (CH), 128.2 (CH), 128.4 (CH), 128.6 (CH), 129.2
(CH), 137.1 (C), 137.9 (C), 138.8 (C), 139.1 (C), 139.4 (C), 141.9
(C), 175.7 (C), 176.3 (C); m/z (ES) 541.2508 (MH+. C36H33N2O3
requires 541.2491).
Acknowledgements
We would like to thank the Committee of Higher Education in
Libya and The Libyan Arab People’s Bureau for a scholarship
(R.A.B.).
References
Crystal structure determination of compound 17
1 (a) S. Jones, J. Chem. Soc., Perkin Trans. 1, 2002, 1–21; (b) P. O’Brien,
J. Chem. Soc., Perkin Trans. 1, 2001, 95–113.
Crystal data. C36H32N2O3, M = 540.64, monoclinic, a =
3
2 D. J. Ager, I. Prakash and D. R. Schaad, Chem. Rev., 1996, 96, 835–
˚
˚
10.3383(14), b = 9.6773(13), c = 14.613(2) A, U = 1390.0(3) A ,
T = 150(2) K, space group P2(1), Z = 2, l(Mo–Ka) = 0.082 mm−1,
15882 reflections measured, 3373 unique (Rint = 0.0471) which
were used in all calculations. The final wR(F2) was 0.0935 (all
data).
875.
3 (a) S. Jones and H. C. Norton, Synlett, 2004, 338–340; (b) R. A. J.
Wybrow, N. G. Stevenson and J. P. A. Harrity, Synlett, 2004, 140–
142.
4 (a) S. Jones and J. C. C. Atherton, Tetrahedron: Asymmetry, 2001,
12, 1117–1119; (b) J. C. C. Atherton and S. Jones, Tetrahedron Lett.,
2001, 42, 8239–8241; (c) J. C. C. Atherton and S. Jones, J. Chem. Soc.,
Perkin Trans. 1, 2002, 2166–2173; (d) J. C. C. Atherton and S. Jones,
Tetrahedron Lett., 2002, 43, 9097–9100; (e) A. Sanyal and J. K. Snyder,
Org. Lett., 2000, 2, 2527–2530; (f) M. S. Corbett, X. Liu, A. Sanyal and
J. K. Snyder, Tetrahedron Lett., 2003, 44, 931–935; (g) K. L. Burgess,
N. J. Lajkiewicz, A. Sanyal, W. Yan and J. K. Snyder, Org. Lett., 2005,
7, 31–34; (h) A. Snyder, Q. Yuan and J. K. Snyder, Tetrahedron Lett.,
2005, 46, 2475–2478; (i) K. L. Burgess, M. S. Corbett, P. Eugenio, N. J.
Lajkiewicz, X. Liu, A. Sanyal, W. Yan, Q. Yuan and J. K. Snyder,
Bioorg. Med. Chem., 2005, 13, 5299–5309.
5 S. Jones and R. A. Bawa, Tetrahedron, 2004, 60, 2765–2770.
6 H. Abdallah, R. Gree and R. Carrie, Tetrahedron Lett., 1982, 23, 503–
506.
7 (a) P. Mangeney, A. Alexakis and J. F. Normant, Tetrahedron Lett.,
1983, 24, 373–374; (b) J. Berlan, J. Besace, E. Stephan and P. Gresson,
Tetrahedron Lett., 1985, 26, 5765–5768; (c) J. Berlan, D. Besace, G. Prat
and J. Pourcelot, J. Organomet. Chem., 1984, 264, 399–408.
8 L. Colombo, C. Gennari, G. Poli and C. Scolastico, Tetrahedron Lett.,
1985, 26, 5459–5462.
3a,4,9,9a-Tetrahydro-4-carboxaldehyde-2-methyl-4,9[1ꢀ,2ꢀ]-
benzeno-1H-benz[f ]isoindole-1,3(2H)-dione 14
N-Methyl maleimide (1.621 g, 14.60 mmol) was added in one
portion as a solid at 80–85 ◦C to a stirred solution of 9-
anthraldehyde (2 g, 9.7 mmol) in toluene (20 cm3). The resulting
mixture was left stirring for 24 h at 80–85 ◦C, cooled to room
temperature and the solvent was removed under reduced pressure
to give the target compound. The crude was purified by column
chromatography (20% EtOAc–petroleumether 40–60◦C)toafford
a white solid of the title compound 14 2.356 g (77%); mp 243–
248 ◦C; (found: C, 75.9; H, 4.7; N, 4.3. C20H15NO3 requires C,
75.70; H, 4.8; N, 4.4%); mmax/cm−1 1772 (C O), 1724 (C O), 1691
=
=
=
=
(C O), 1457 and 1432 (C C); dH(250 MHz; CDCl3; Me4Si) 2.44
(3H, s, CH3), 3.26 (1H, dd, J 8.6 and 3.4, PhCHC(O)CH), 3.67
(1H, d, J 8.6, C(O)CH), 4.72 (1H, d, J 3.4, PhCHCH), 7.05–7.27
(6H, m, ArCH), 7.35–7.40 (1H, m, ArCH), 7.54–7.61 (1H, m,
ArCH), 10.81 (1H, s, CHO); dC(62.5 MHz; CDCl3; Me4Si) 24.4
(CH3), 45.9 (CH), 47.5 (CH), 48.0 (CH), 57.7 (C), 122.9 (CH),
123.4 (CH), 124.7 (CH), 125.4 (CH), 126.8 (CH), 127.2 (CH),
127.5 (CH), 136.4 (C), 138.2 (C), 139.1 (C), 141.2 (C), 175.7 (C),
176.2 (C), 200 (C); m/z (ES) 318.1125 (MH+. C20H16NO3 requires
318.1130).
9 C. Agami, F. Couty, L. Hamon and O. Venier, J. Org. Chem., 1997, 62,
2106–2112.
10 (a) N. Casamitjana, M. Amat, N. Llor, M. Carreras, X. Pujol, M.
Montserrat Fernandez, V. Lopez, E. Molins, C. Miravitlles and J. Bosch,
Tetrahedron: Asymmetry, 2003, 14, 2033–2039; (b) S. G. Pyne and J.
Safaei-G, J. Chem. Res., 1996, (S)160–161; (c) S. G. Pyne, J. Safaei-G,
D. C. R. Hockless, B. W. Skelton, A. N. Sobolev and A. H. White,
Tetrahedron, 1994, 50, 941–956; (d) H. Hoffmann, M. Bolte, B. Berger
and D. Hoppe, Tetrahedron Lett., 1993, 34, 6537–6540; (e) S. G. Pyne,
B. Dikic, P. A. Gordon, B. W. Skelton and A. H. White, Aust. J. Chem.,
1992, 46, 73–93; (f) S. G. Pyne, B. Dikic, P. A. Gordon, B. W. Skelton
and A. H. White, J. Chem. Soc., Chem. Commun., 1991, 1505–1506;
(g) W. R. Roush, A. P. Essenfeld, J. S. Warmus and B. B. Brown,
Tetrahedron Lett., 1989, 30, 7305–7308.
11 A. Hussain and P. B. Wyatt, Tetrahedron, 1993, 49, 2123–2130.
12 C. Agami and T. Ritzk, Tetrahedron, 1985, 41, 537–540.
13 Using CAChe Workstation Pro 6.1.12 for Windows (Fujitsu Ltd). The
calculations were carried out by using PM3 methods.
14 R. G. Kostyanovsky, I. M. Gella, V. I. Markov and Z. E. Samojlova,
Tetrahedron, 1974, 30, 39–45.
(3aR,9aR)-3a,4,9,9a-Tetrahydro-4-carboxaldehyde-2-methyl-
4,9[1ꢀ,2ꢀ]-benzeno-1H-benz[f ]isoindole-1,3(2H)-dione 14
The title compound was obtained by stirring cycloadduct 12
(0.100 g, 0.216 mmol) in a solution of CH2Cl2–H2O (5 cm3; 9 :
1, v/v) for 18 h at rt in the presence of silica gel (approx. 50 mg).
The reaction was filtered, evaporated and purified by column
chromatography (20% EtOAc–petroleumether 40–60◦C)toafford
a white solid of the title compound 14 (0.050 g, 73%). [a]2D5 −48.0
(c 0.5 in CHCl3).
15 A. Karim, A. Mortreux, F. Petit, G. Buono, G. Peiffer and C. Siv,
J. Organomet. Chem., 1986, 317, 93–104.
16 S. R. Hitchcock, G. P. Nora, C. Hedberg, D. M. Casper, L. S. Buchanan,
M. D. Squire and D. X. West, Tetrahedron, 2000, 56, 8799–8807.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 4206–4213 | 4213
©