Stereoisomerism in 3-[N-(2-acetoxypropanoyl)-N-acylamino]-quinazolin-4(3H)-ones, enantioselective acylating agents

Article abstract of DOI:10.1039/b005814i

The title compounds diacylaminoquinazolinones (DAQs) are enantioselective acylation agents for amines and a detailed study of their stereostructures was undertaken with the aim of understanding how this enantioselectivity arises. The N-N bond in these DAQs is a chiral axis. Even where both N-acyl groups are (S)-2-acetoxypropanoyl, the N-N bond is still a chiral axis because in the most stable conformation of the planar imide moiety, one exolendo orientation of the carbonyl groups is much preferred over the alternative (endolexo) as revealed by NMR spectroscopy. A conformational preference within the 2-acetoxypropanoyl grouping accounts for the presence of a single exolendo conformation in solution for some of these DAQs (see above) but an interconverting exolendo? endolexo mixture for others. Where a single exolendo conformation is present in solution, evidence is presented that this closely resembles the X-ray determined crystal structure. A mechanism for the second acylation step to form these DAQs is proposed, which involves preliminary O-acylation of the 3-(monoacylamino)quinazolinone. The Royal Society of Chemistry 2000.

Full text of DOI:10.1039/b005814i

View Article Online / Journal Homepage / Table of Contents for this issue
Stereoisomerism in 3-[N-(2-acetoxypropanoyl)-N-acylamino]-
quinazolin-4(3H)-ones, enantioselective acylating agents
Abdullah G. Al-Sehemi, Robert S. Atkinson,* John Fawcett and David R. Russell
1
Department of Chemistry, Leicester University, Leicester, UK LE1 7RH
Received (in Cambridge, UK) 19th July 2000, Accepted 19th September 2000
First published as an Advance Article on the web 21st November 2000
The title compounds diacylaminoquinazolinones (DAQs) are enantioselective acylation agents for amines and a
detailed study of their stereostructures was undertaken with the aim of understanding how this enantioselectivity
arises. The N–N bond in these DAQs is a chiral axis. Even where both N-acyl groups are (S)-2-acetoxypropanoyl,
the N–N bond is still a chiral axis because in the most stable conformation of the planar imide moiety, one
exo/endo orientation of the carbonyl groups is much preferred over the alternative (endo/exo) as revealed by NMR
spectroscopy. A conformational preference within the 2-acetoxypropanoyl grouping accounts for the presence of
a single exo/endo conformation in solution for some of these DAQs (see above) but an interconverting exo/endo
endo/exo mixture for others. Where a single exo/endo conformation is present in solution, evidence is presented that
this closely resembles the X-ray determined crystal structure. A mechanism for the second acylation step to form
these DAQs is proposed, which involves preliminary O-acylation of the 3-(monoacylamino)quinazolinone.
3-(Diacylamino)quinazolinones (DAQs), e.g. 1, are highly
chemoselective acylating agents for primary amines in the
presence of secondary amines and for the less hindered of two
secondary amines (Scheme 1).1
conformation is an exo/endo arrangement for the two carbonyl
groups and for e.g. DAQ 3, exo/endo endo/exo interconver-
sion occurs at room temperature at a restricted enough rate to
give rise to broadening of the acetyl methyl signals in its NMR
spectrum.
At Ϫ83 ЊC in [2H6] acetone solution, signals from the two
acetyl methyl groups have separated into two singlets of equal
intensity and, significantly, the methylene protons of the ethyl
substituent become non-equivalent because the N–N bond is
now a chiral axis.2
The chemoselectivity exhibited by e.g. DAQ 1 towards amines
(Scheme 1) has a stereoselective counterpart, the preferred
reaction of an enantiopure DAQ with one enantiomer of a
racemic amine giving rise to kinetic resolution of the amine2
(Scheme 2).
Scheme 1
When the two N-acyl groups in the DAQ are not identical,
the N–N bond becomes a chiral axis with the two planes con-
taining the quinazolinone and imide moieties orthogonal to
one another. The barrier to N–N bond rotation is sufficiently
high to allow separation of diastereoisomers (atropisomers)
when there is a chiral centre also present in the DAQ. Thus,
the stereostructures of both (racemic) diastereoisomers of
DAQ 2 have been identified by X-ray crystallography determin-
ations: interconversion between them takes place on heating in
toluene (∆G‡ = 121 kJ molϪ1) by rotation around the N–N
bond.2
There is also a significant, albeit smaller, barrier to stereo-
isomerism within the imide group in these DAQs. The preferred
Scheme 2
In this paper,3 the stereoisomerism in DAQs bearing at least
one N-2-acetoxypropanoyl group is examined with a view to
understanding the origin of the enantioselectivity in kinetic
resolution of racemic amines using these DAQs.
DOI: 10.1039/b005814i
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421
This journal is © The Royal Society of Chemistry 2000
4413

View Article Online
Results and discussion
3-Aminoquinazolinone 5 was prepared from -valine by the
route shown in Scheme 3 in an overall yield of 46% without the
need for chromatographic separation at any stage.
Fig. 1 The molecular structure of 7a. H atoms on chiral centres are
shown with open bonds, all other H atoms have been omitted for clarity.
Scheme 3 Reagents: i, KNO2, HOAc; ii, SOCl2; iii, methyl anthran-
ilate; iv, NH2NH2, EtOH, 147 ЊC; v, ButMe2SiCl, imidazole; vi, (S)-
CH3CH(OAc)COCl, pyr. CH2Cl2.
Acylation of 3-aminoquinazolinone 5 with (S)-2-acetoxy-
propanoyl chloride gave the 3-monoacylaminoquinazolinone
(MAQ) 6. Even a chromatographically purified sample of
MAQ 6 showed small additional signals in its NMR spectrum,
consistent with the presence of a minor N–N bond rotamer
(ratio ~10:1).4 Reaction of MAQ 6 with 2-methylpropanoyl
chloride in dichloromethane containing pyridine at room tem-
perature for three days gave a mixture of four diastereoisomeric
DAQs 7a–d in order of their elution in chromatography using a
chromatotron (Scheme 4).
NMR spectra of these DAQs (see below) suggest that exo/endo
conformations are also those preferred in solution.
It is noteworthy that in the crystal structure of DAQ 7d, the
2-methylpropanoyl carbonyl group has an exo-conformation
whereas in those of DAQs 7a–c, this group has an endo-
orientation. The conformations around the (Q)C2–C(Me,H,-
OSi) bonds in 7b and 7d are similar and have the C–OSi bond
out of the plane of the Q-ring by 74Њ and 82Њ respectively. The
conformations around the (Q)C2–C(Me,H,OSi) bonds in 7a
and 7c are likewise similar, but differ from those around 7b and
7d in having the C–OSi bond close to the plane of the Q-ring
(21Њ and 14Њ respectively).
Scheme 4 Reagents and conditions: i, PriCOCl, pyr. CH2Cl2, 3 days;
ii, BuLi, THF; iii, (S)-CH3CH(OAc)COCl.
After flash chromatography to remove residual MAQ 6, the
bulk of the major DAQ diastereoisomer 7b was separated by
crystallisation. Pure samples of 7a, 7c and 7d were then
obtained by chromatotron chromatography. X-Ray crystal
structures of all the DAQs were carried out and reveal their
stereostructures as shown3 (see also Fig. 1).†
Conversion of MAQ 8 into DAQs 7a (4%) and 7c (34%)
was also possible by formation of its N-lithium salt using
butyllithium followed by addition of (S)-2-acetoxypropanoyl
chloride.
These crystal structures confirm that, at least in the crystal-
line form, an exo/endo conformation for the carbonyl groups
of the imide moiety is preferred. This preference has been
previously found in N-alkyl-substituted acyclic imides5 and in
most other crystal structures of DAQs we have examined.2 The
Diastereoisomers 7a and 7d apparently arise by epimeris-
ation at the (S)-2-acetoxypropanoyl chiral centres and it was
assumed that the (Q)C2-attached chiral centre was stereostable.
However, in the X-ray crystal structure determination of DAQ
¯
7a, the space group P1 indicates that the molecule was racemic.
The bulk of DAQ 7a was obtained as an oil and only a very
small number of small crystals were produced from this oil on
standing: one of these crystals was used for the structure
determination. The oily material which remained failed to
produce any more crystalline material even after setting aside
for several months and it was also found to have a significant
rotation [α]D = ϩ126.8 (c 0.5, CDCl3), so the extent of racemis-
ation is believed to be small: the enantiopurity of these DAQs
7a–d appears to be high based on the high enantioselectivities
obtained in their reactions with amines.3
† The crystal structures of 7b–d have been reported previously and are
drawn so as to simulate the conformations they take up in the crystal in
each case. The crystal structure of 7a was not previously reported and is
reproduced here along with the simulated drawing for comparison with
7b–d. CCDC reference number 207/485. See http://www.rsc.org/
suppdata/p1/b0/b005814i/ for crystallographic files in .cif format.
4414
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421

View Article Online
Faster exo/endo–endo/exo interconversion [explanation (a)
above] is unlikely because the NMR spectrum of DAQ 7b
remains unchanged even when run at Ϫ40 ЊC with no visible
broadening of the signals: the chemical shifts δ of the
CH3CH(OAc) protons in 7a and 7b (see below) do not support
(a).
Although at this point it is difficult to exclude (b) above, it
is not clear what property it is that 7a and 7b share which
would lead to both preferring an exo/exo conformation in
solution since they have different configurations both at their
N–N chiral axes and their 2-acetoxypropanoyl chiral centres
and, at least in the crystalline state, different preferred con-
formations around the (Q)C2–C(Me,H,OSi) bonds (see above).
The chemical shifts δ for the CHOAc protons in DAQs 7a
and 7b are 6.06 and 6.00 respectively compared with δ 5.5
for the corresponding proton in MAQ 6. Deshielding of the
CHOAc protons in 7a and 7b is ascribed to the proximity in
solution of the neighbouring 2-methylpropanoyl carbonyl
group in the exo/endo conformer. It is clear from the crystal
structures of 7a and 7b that the CHOAc proton is also close to
the carbonyl of the 2-methylpropanoyl group. The conclusion
to be drawn is that there is predominantly one conformation in
solution for the imide moieties in DAQs 7a and 7b [see (c)
above] and that these are similar to the preferred conformations
in the respective crystal structures.
For DAQs 7c and 7d, their NMR spectra at Ϫ50 ЊC
reveal that the more abundant imide conformation in both
cases is also the exo/endo with the CH(OAc)CH3 proton signals
deshielded at δ 6.08 and 6.01 respectively (see above) by the
2-methylpropanoyl carbonyl as for DAQs 7a and 7b. Thus, the
major imide conformation for DAQ 7c in solution is the same
as that in the crystal structure but for DAQ 7d it is the minor
conformation in solution which corresponds to that found in
the crystal structure.
Fig. 2
DAQs 7b and 7c differ only in the configuration of their
N–N axes. Heating pure DAQ 7b in boiling deuterochloro-
form for 48 h converted it into a 1:1 mixture of 7b–7c as shown
by NMR spectroscopy. Likewise, heating 7d for 13 h under the
same conditions gave a 6:1 mixture of 7a–7d. Both these
equilibrations are assumed to take place by rotation around the
N–N bond. As expected, there was no interconversion between
7a and 7c or between 7b and 7d, i.e. epimerisation at the
2-acetoxypropanoyl centre does not take place under these
conditions.
Conformations of imide moieties of DAQs 7a–d in solution
The NMR spectra of DAQs 7a–d present a dichotomy in
appearance for the CHOAc and CHPri signals of 7a and 7b,
which are sharp, and those for 7c and 7d, which are broadened
(Fig. 2).
The origin of these exo/endo–endo/exo equilibrium positions
in DAQs 7a–d is still under investigation but it is clear that it
must involve interactions between the substituents on the chiral
centre and the quinazolinone when these are cis (i.e. in the endo/
exo conformation): these interactions are adverse in DAQs 7a
and 7b but less so in DAQs 7c and 7d.
It seemed likely that these signals for DAQs 7c and 7d are
broadened because of exo/endo endo/exo interconversion in
the imide moiety which is becoming slow on the NMR time-
scale (cf. the temperature-dependent NMR spectrum for DAQ
3 above). In support of this conclusion, the broadened quartet
for CHOAc in the NMR spectrum of DAQ 7d at 27 ЊC (δ 5.73)
separated at Ϫ50 ЊC into a broadened singlet and a broadened
quartet (δ 6.08 and 5.57, ratio 2.9:1, respectively) which we
assigned to CHOAc protons in exo- and endo-COCH(OAc)-
CH3 groups respectively (see below) (exo = 2-acetoxypropan-
oylcarbonyl cis to Q). Likewise, the broadened quartet at δ 5.71
in the NMR spectrum of DAQ 7c, run at room temperature,
separated into a broadened quartet and a broad singlet at
δ 6.01 and 4.97, ratio 2.7:1 respectively when the spectrum
was run at Ϫ50 ЊC.‡
The torsional angles ꢀ between the C᎐O and C–O bonds
᎐
of the –NC(᎐O)–C(–OAc)H(CH ) grouping in nine crystal
᎐
3
structures we have which contain 2-acetoxypropanoyl as an
N-substituent, lie between 19Њ and 39Њ for eight of them. If
there is a strong preference for retention of this torsional angle
in solution, i.e. this conformation is much preferred over others,
then in the alternative endo/exo forms of e.g. DAQs 7b and 7c
(Scheme 5), a sterically repulsive interaction between the methyl
group (MeCHOAc) and the (Q)C2-substituent will result for 7b
endo/exo but not for 7c endo/exo. Consequently a displacement
of the equilibrium wholly onto the 7b exo/endo side can be
accounted for, whereas 7c will be an interconverting mixture of
exo/endo and endo/exo forms. An analogous explanation nicely
accounts for the differences in the NMR spectra of DAQs 7a
and 7d.
There are a number of possible explanations for the sharp-
ness of the corresponding signals in DAQs 7a and 7b:
(a) the interconversion in solution between exo/endo and
endo/exo forms is faster than for DAQs 7c and 7d;
The conclusion that DAQs 7a and 7b are present in solution
essentially in one exo/endo conformation whereas DAQs 7c and
7d are present as two rapidly interconverting exo/endo–endo/exo
conformations is supported by an analogous correlation for the
NMR spectra of DAQs 13 and 14 and an X-ray crystallo-
graphic structure determination of DAQ 13.6 These two com-
pounds were prepared in the usual way by diacylation of the
corresponding 3-aminoquinazolinones 9 and 10 with (S)-2-
acetoxypropanoyl chloride (for 13) and rac-2-acetoxypropanoyl
chloride (for 14) (Scheme 6).
Conversion of MAQ 11 into DAQ 13 is significantly
faster than the corresponding reaction of MAQ 6 with
2-methylpropanoyl chloride and pyridine and the degree of
epimerisation at the 2-acetoxypropanoyl centre, giving any
meso-compound (see below) is less.
(b) the conformation in solution is exo/exo, notwithstanding
the exo/endo conformation in solution;§
(c) in solution the exo/endo conformation is significantly
more stable than the endo/exo.
‡ For simplicity, we have assumed that for DAQ 7d the major imide
conformer is the exo/endo but, in view of the broadening of the signal at
δ 6.08, it may be in equilibrium with a small concentration of another
conformation (exo/exo?). Likewise for DAQ 7c, the smaller broadened
signal at δ 4.97 is assumed to be that from the endo/exo conformer
possibly in equilibrium with another minor conformer.
§ In principle, the endo/endo conformation is also possible but this is
probably the least stable of the three possible planar imide conformers
and is neglected (it is excluded in any case for the same reasons as the
exo/exo).
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421
4415

View Article Online
Scheme 7
2-substituent of the latter. Thus, DAQs 13 and 14 each exist
in solution preferentially in a single exo/endo conformation for
the same reasons that DAQs 7a and 7b do, in spite of the
fact that in 13 and 14 the two acyl groups on nitrogen are
identical.
The NMR spectra of DAQs 13 and 14 do not support exo/
exo conformations (Scheme 7) for them in solution (see earlier
discussion for DAQs 7a and 7b above). Although the two (S)-2-
acetoxypropanoyl groups in such exo/exo conformations would
be diastereotopic¶ the magnitude of the chemical shift differ-
ences (see above for CH3CHOAc quartets) is more compatible
with single exo/endo conformations i.e. 13a exo/endo as above.
Scheme 5
meso-DAQ 13b
There are two possible meso-compounds as a result of epimeris-
ation at one chiral centre in DAQ 13 (Scheme 8). From our
Scheme 6 Reagents: i, (S)-CH3CH(OAc)COCl, pyr. CH2Cl2; ii, rac-
CH3CH(OAc)COCl, pyr. CH2Cl2.
As in the X-ray crystal structures of DAQs 7a–d, the pre-
ferred conformations for the imide moiety in DAQs 13 is exo/
endo and, for the two 2-acetoxypropanoyl groups, the torsion
angles ꢀ have magnitudes of Ϫ25.2Њ (endo) and Ϫ20.8Њ (exo)
(see above).
The two CH3CHOAc signals in the NMR spectrum of DAQ
13 appear as widely separated sharp quartets (δ 5.01 and 5.97)
with the lower field signal characteristically deshielded by the
neighbouring endo carbonyl group; there was no change in the
appearance of the spectrum when run at Ϫ50 ЊC in CDCl3. A
similarly wide separation of chemical shift (δ 4.87 and 6.0) was
present for the corresponding signals of DAQ 14.
It is important to point out that, unlike the exo/endo and
endo/exo conformations of DAQ 3 which are enantiomeric,
those for e.g. DAQ 13 (13a exo/endo and 13a endo/exo; Scheme
7) are diastereoisomeric and, in principle, different in energy.
[These two conformations for the imide as in DAQ 1 are inter-
converted by rotation around the N–N bond and thus the N–N
bond is not a chiral axis in this molecule if 13a exo/endo
13a endo/exo is occurring.]
Scheme 8
earlier discussion, the stability of the meso-DAQ diastereo-
isomer 13b would be expected to be greater than 13c since both
imide conformations in the latter suffer from the previously
described adverse steric interactions as indicated. The NMR
If, as previously discussed for DAQs 7a–d, the same small
torsion angle ꢀ between the C᎐O and C–O bonds is preferred
᎐
for DAQ 13 in solution as well as in the crystalline form,
then the equilibrium position greatly in favour of 13a exo/
endo becomes clear because 13a endo/exo suffers from the
same adverse steric interaction between the methyl in the
2-acetoxypropyl group cis to the quinazolinone and the methyl
¶ If the two carbonyl groups in 13a exo/exo (see Scheme 6) were to be
replaced in turn by e.g. thiocarbonyl groups, the relationship between
the two resulting molecules would be a diastereoisomeric one. Hence
the two (S)-2-acetoxypropanoyl groups are diastereotopic (see R. S.
Atkinson, Stereoselective Synthesis, Wiley, Chichester, 1995, p.15).
4416
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421

View Article Online
Scheme 9
spectrum of a minor component isolated by chromatography
which an analogous enamine 19 was postulated as an inter-
mediate (Scheme 10).
of the crude product from acylation of MAQ 11 (Scheme 6)
shows, besides the two CH3CHOAc quartets from DAQ 13a,
two broadened quartets at δ 5.47 and 5.59, possibly from the
same proton in both meso-diastereoisomers 13b and 13c. A
pure sample of presumably meso-isomer 13b was isolated by
chromatotron chromatography and found to have zero rotation.
A 1:1 ratio of exo/endo–endo/exo diastereoisomers of DAQ
13b (and 13c) would be anticipated|| with interconversion
between the two conformations becoming fast on the NMR
timescale (as for DAQs 7c and 7d) and hence giving rise to
broadening of the quartet signals as in the NMR spectrum
above.
Scheme 10
Deacylation of O-acylimidate 15 by protonation on the imid-
ate nitrogen (cf. acylation to give 16) would reform MAQ 6,
which would have undergone some epimerisation at the 2-acet-
oxypropanoyl chiral centre if interconversion with enamine
17 had occurred.
It is noteworthy that the two major diastereoisomers DAQ 7b
and 7d produced have the same configurations for their
N–N chiral axes and that the transition state 21a leading to
their formation (Scheme 11) would have the quinazolinone
Formation of DAQs 7a–7d: mechanism of epimerisation in 7a
and 7d
Acylation of MAQ 6 with 2-methylpropanoyl chloride and
pyridine is a slow reaction requiring three days and some MAQ
6 remains unreacted even after this time. When DAQs 7b and 7d
were re-submitted to the acylation conditions with 2-methyl-
propanoyl chloride, pyridine and dichloromethane for 2 days,
they were both recovered unchanged which indicated that
epimerisation occurred in the acylation of MAQ 6 itself.
In fact, the MAQ 6 recovered from the attempted diacylation
reaction was a mixture of diastereoisomers resulting from some
epimerisation at the 2-acetoxypropanoyl centre as shown by the
appearance in its NMR spectrum of a separated singlet signal
just upfield of the CH3CHOCOCH3 signal of the starting
MAQ 6 (δ 2.21 and 2.24).
The tendency for amides to react with acid chlorides on
oxygen will be augmented in MAQ 6 because the quinazolinone
ring will hinder attack by the sp2-hybridised exocyclic nitrogen
on the acid chloride both sterically and, because of its electron-
withdrawing character, electronically. Consequently, the revers-
ibly formed product would be expected to be the O-acylimidate
15 (Scheme 9).†† Reversible protonation–deprotonation of the
imidate 15 would account for epimerisation at the 2-acetoxy-
propanoyl centre. Further acylation of this O-acylimidate 15
with 2-methylpropanoyl chloride on the now more accessible
and more basic nitrogen would deliver the N-acyliminium
species 16 whose deacylation leads to DAQs 7a–d.
Scheme 11
2-substituent in what is believed to be a preferred conformation
(cf. conformations of the Q2-substituent in DAQs 7b and 7d in
the crystal structures;‡‡ see earlier), with attack by the acid
chloride on the lone pair in 21a (rather than 21b).
Summary
A precedent for the formation of the enamine 17 is the con-
version of MAQ 18 into the pyrazolo[1,5-c]quinazoline 20 in
Acylation of MAQ 6 with 2-methylpropanoyl chloride and
pyridine gives a mixture of four DAQ diastereoisomers 7a–d
|| The exo/exo conformation might be expected to be favoured for DAQ
13c but a pure sample of this compound was not isolated.
†† It is possible that initial acylation takes place on the Q-carbonyl
oxygen and the intramolecular transfer to form imidate 15 occurs.
‡‡ If the conformation of the Q2-substituent in the crystal structures of
DAQs 7a and 7c were used in 21a, the predicted result would be the
same.
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421
4417

View Article Online
resulting from the axial chirality of the N–N bond together
with partial epimerisation at the 2-acetoxypropanoyl chiral
centre. Diastereoisomers 7a and 7b, 13 and 14 show an
unexpected preference in solution for one of the two exo/endo
conformations of the imide moiety. The preferred exo/endo
conformation, which corresponds in each case to that present
in the X-ray determined crystalline structures of these DAQs, is
itself believed to arise from a preferred conformation within
the NCOCH(OAc)CH3 grouping which gives rise to an adverse
steric effect in the alternative endo/exo conformation. In
diastereoisomers 7c and 7d, this unfavourable interaction is
absent and an equilibrium mixture of exo/endo and endo/exo
conformers is present in each case.
The presence of a single conformation for these DAQs in
solution makes it more likely that this will be the reacting con-
formation with amines. Since these DAQs react highly enantio-
selectively with chiral primary and secondary amines giving
kinetic resolution, the findings in this paper will be invaluable in
attempts to derive a transition state model to account for the
sense of enantioselectivity and hence to make rational changes
to maximise the degree of this enantioselectivity.
solution washed with water (3 × 25 cm3), dried and the solvent
removed under reduced pressure to give 3-amino-2-[(S)-1-
hydroxy-2-methylpropyl]quinazolin-4(3H)-one as pale
4
a
yellow oil [α]D = Ϫ24 (c 1.2, CHCl3) which solidified on standing
to give a colourless solid mp 138–140 ЊC (from ethanol) (11 g,
82%) (Found: C, 62.1; H, 6.6; N, 17.7. C12H15N3O2 requires C,
61.8; H, 6.5; N, 18.0%) (Found: MHϩ 233.1164. C12H14N3O2,
requires MHϩ 233.1164); νmax/cmϪ1 3300w, 1690 and 1600;
δH 0.82 and 1.1 (6H, 2 × d, J 6.9, CH3CHCH3), 2.3 (1H, m,
CH3CHCH3), 3.98 (1H, d, J 8.2, OH), 4.75 (2H, s, NH2), 4.91
(1H, dd, J 8.2 and 3.5, CHOH), 7.4 [1H, ddd, J 8.2, 7.0 and 1.3,
6-H(Q)], 7.6 [1H, ddd, J 8.2, 7.0 and 1.3, 7-H(Q)], 7.65 [1H,
dd, J 8.2 and 1.3, 8-H(Q)] and 8.16 [1H, d, J 8.2, 5-H(Q)];
δC 15.5 and 20.5 (CH3CHCH3), 32.6 (CH3CHCH3), 73.5
(CHOH), 120.4 [CCO(Q)], 126.9, 127.2, 127.5 and 134.9 [4 ×
CH(Q)], 146.2 [CN᎐C(Q)], 158.7 [C᎐N(Q)] and 162.5 [CO(Q)];
᎐
᎐
m/z (%) 233 (MHϩ, 23), 216 (34), 190 (100), 175 (78) 145 (48),
132 (73) and 84 (74).
(S)-3-Amino-2-(1-tert-butyldimethylsilyloxy-2-methylpropyl)-
quinazolin-4(3H)-one 5
The 3-aminoquinazolinone 4 above (6 g, 26 mmol) was dis-
solved in N,N-dimethylformamide (DMF) (15 cm3), tert-
butyldimethylsilyl chloride (TBDMS-Cl) (7.4 g, 49 mmol)
and imidazole (4.4 g, 64 mmol) were added and the solution
stirred at room temperature for two days. Light petroleum
(30 cm3) was then added, the solution washed with water, dried
and evaporated under reduced pressure. Crystallisation of the
resulting white solid gave the desired 3-aminoquinazolinone 5
(8.5 g, 96%), mp 126–128 ЊC (from light petroleum) (Found:
C, 61.9; H, 8.4; N, 12.0. C18H29N3O2Si requires C, 62.0; H, 8.4;
N, 12.1%); νmax/cmϪ1 3336m and 1675m; [α]D = Ϫ95 (c 1.3,
CHCl3); δH Ϫ0.12 and 0.1 (6H, 2 × s, CH3SiCH3), 0.94 and
1.2 (6H, 2 × d, J 6.6, CH3CHCH3), 0.97 [9H, s, (CH3)3CSi], 2.85
(1H, m, CH3CHCH3), 4.65 (1H, d, J 6.6, CHOSi), 5.86 (2H,
s, NH2), 7.59 [1H, ddd, J 8.2, 7.0 and 1.6, 6-H(Q)], 7.82 [1H,
ddd, J 8.2, 7.0 and 1.0, 7-H(Q)], 7.85 [1H, dd, J 8.2 and 1.6,
8-H(Q)], 8.39 (1H, dd, J 8.2, and ∼1.0, 5-H(Q)]; δC Ϫ4.8 and
Ϫ4.5 (CH3SiCH3), 18.9 [(CH3)3C], 19.2 and 19.9 (CH3CHCH3),
26.1 [(CH3)3CSi], 31.4 (CH3CHCH3), 82.9 (CHOSi), 120.0
[CCO(Q)], 126.7, 127.2, 128.0 and 134.2 [4 × CH(Q)], 146.6
Experimental
For general experimental details see ref. 7.
Preparation of 3-amino-2-[(S)-1-hydroxy-2-methylpropyl]quin-
azolin-4(3H)-one 4
2-Hydroxy-3-methylbutanoic acid (13 g) (prepared from L-
valine8) was dissolved in dry ether (20 cm3) and freshly distilled
acetyl chloride (15 cm3) added with stirring. After setting
aside overnight, the ether and unreacted acetyl chloride were
evaporated under reduced pressure to leave 2-acetoxy-3-methyl-
butanoic acid as a cloudy oil (17 g) δH 1.02 and 1.05 (6H, 2 × d,
J 6.6, CH3CHCH3), 2.16 (3H, s, CH3CO), 2.25 (1H, m,
CH3CHCH3), 4.9 (1H, d, J 6.2, CHO) and 10.15 (1H, s, CO2H).
This acid (13 g) was converted to its acid chloride by dissolving
in dry ether (40 cm3) adding two drops of N,N-dimethyl-
formamide and then thionyl chloride (12 cm3) dropwise with
stirring. After setting aside overnight, ether and unreacted
thionyl chloride were removed under reduced pressure to give
the acid chloride. To a briskly stirred solution of this acid
chloride (13 g) in dry ether (300 cm3) was added methyl anthran-
ilate (23.2 cm3; 2.2 eq.) and the thick white precipitate which
formed was stirred for a further 1 h. After setting the reaction
mixture aside overnight, the white solid was filtered, washed
well with ether and the combined filtrates washed successively
with hydrochloric acid (2 M, 5 × 50 cm3), saturated aqueous
sodium hydrogen carbonate and saturated brine then dried and
the solvent removed by evaporation under reduced pressure to
give an oil which solidified over 12 h. Methyl (S)-N-(2-acetoxy-
3-methylpropanoyl)anthranilate was obtained as a colourless
solid (17 g, 86%) mp 44–46 ЊC (from ethanol) (Found: Mϩ
293.1263. C15H19NO5, requires Mϩ 293.1263); [α]D = Ϫ125.6
(c 1.2, CHCl3); νmax/cmϪ1 3320m, 1700m, 1610s and 1590m;
δH 0.89 and 0.92 (6H, 2 × d, J 5.7, CH3CHCH3), 2.21 (3H, s,
CH3CO), 2.31 [1H, m, (CH3)2CH], 3.78 (3H, s, OCH3), 5.15
(1H, d, J 4.0, HCO), 7.1 [1H, ddd, J 8.2, 7.0 and 1.0, 5-H(Ar)],
7.54 [1H, ddd, J 8.5, 7.0 and 1.6, 4-H(Ar)], 8.02 [1H, dd, J 8.2
and 1.6, 6-H(Ar)], 8.76 [1H, dd, J 8.5 and 1.0, 3-H(Ar)]; δC 17.2
and 19.2 (CH3CHCH3), 21.1 (CH3CHCH3) 31.1 (CH3CO),
52.6 (OCH3), 78.6 (HCO), 115.2 (CCO2Me), 120.7, 122.6, 131.2
and 135.0 [4 × CH(Ar)], 141.0 [CNH(Ar)] and 168.8, 169.2 and
170.6 (3 × CO); m/z (%) 293 (Mϩ, 41), 178 (46), 151 (100) and
119 (35). The foregoing anthranilate was dissolved in ethanol
(25 cm3) and heated with hydrazine (14.0 cm3) in a closed steel
container at 147 ЊC for 24 h. After cooling, the bulk of the
solvent was removed under reduced pressure and the residue
dissolved in dichloromethane (40 cm3), the dichloromethane
[CN(Q)], 154.3 [C᎐N(Q)], and 160.1 [CO(Q)]; m/z (%) 347
᎐
(Mϩ, 9.0), 232 (70), 290 (100), 275 (54), 259 (42), 247 (58), 232
(54), 218 (59), 203 (52), 187 (47), 75 (29) and 59 (43).
General procedure for the mono-N-acylation of 3-aminoquinazol-
inones
To the 3-aminoquinazolinone dissolved in dry dichloromethane
(2 cm3 gϪ1) containing dry pyridine (1.5 mol equiv.) was added
the acid chloride (1.5 mol equiv.) dropwise with stirring. After
stirring for 12 h at room temperature, more dichloromethane
(~50 cm3) was added and the solution washed with saturated
aqueous sodium hydrogen carbonate, then water, dried and the
solvent removed under reduced pressure.
General procedure for the preparation of 3-(diacylamino)quin-
azolinones (DAQs) from 3-(monoacylamino)quinazolinones
(MAQs)
To a solution of the MAQ (1 mol equiv.), prepared as described
above, in dichloromethane (2 cm3 gϪ1) containing dry pyridine
(1.5 mol equiv.), was added the acid chloride (2–3 mol equiv.)
dropwise over 10 min and the mixture stirred and heated under
reflux for 2–4 days with exclusion of moisture, monitoring the
disappearance of the starting MAQ by TLC. After cooling,
additional dichloromethane (40 cm3) was added and the solu-
tion washed with aqueous sodium hydrogen carbonate, then
water, dried and the dichloromethane removed under reduced
4418
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421

View Article Online
pressure. The bulk of the residual pyridine was removed using
chromatography on silica using light petroleum–ethyl acetate
an oil pump and the product purified by flash chromatography.
(3:1) as eluent to give the racemic MAQ 6 as a colourless oil
(0.35 g, 88%) (Rf 0.26). Examination of the product by NMR
spectroscopy showed it to be a 1:1 mixture of diastereoisomers
with signals in addition to those in the spectrum above at δ 1.63
(3H, d, J 6.7, CH3CHOAc) and 2.21 (3H, s, OCOCH3). Partial
separation was achieved by re-chromatography using a chrom-
atotron and light petroleum–ethyl acetate (5:1) as eluent to give
one fraction containing a 5:1 mixture of diastereoisomers with
the major diastereoisomer identical to that obtained previously.
3-(S)-2-Acetoxypropanoylamino-2-[(S)-1-tert-butyldimethyl-
silyloxy-2-methylpropyl]quinazolin-4(3H)-one 6
The general procedure for monoacylation was followed using
3-aminoquinazolinone 5 (1 g, 2.9 mmol), pyridine (0.23 g, 2.9
mmol), dichloromethane (4 cm3) and (S)-2-acetoxypropanoyl
chloride (0.34 g, 4.4 mmol) and the mixture stirred for 24 h at
room temperature. The brown oil obtained after work-up as
described above was purified by column chromatography on
silica using light petroleum–ethyl acetate (3:1) as eluent to give
the MAQ 6 as a colourless oil (1.15 g, 88%) (Rf 0.26) (Found:
Preparation of 3-[N-(2-acetoxypropanoyl)-N-(2-methyl-
propanoyl)amino-2-[(S)-1-tert-butyldimethylsilyloxy-2-methyl-
propyl]quinazolin-4-(3H)-ones (7a–d)
MHϩ 462.2425. C23H35N3O5Si requires MHϩ 462.2424); νmax
/
cmϪ1 3260br, 1745s, 1698s and 1610s; [α]D = ϩ10.6 (c 1.6,
CHCl3); δH 0.0 and 0.14 (6H, 2 × d, J 6.6, CH3CHCH3), 0.94
(9H, s, (CH3)3CSi), 1.72 (3H, d, J 6.9, CH3CHOAc), 2.04 (1H,
m, 7 lines, (CH3)CH), 2.24 (3H, s, CH3COO), 4.44 (1H, d, J 6.6,
CHOSi), 5.50 (1H, q, J 6.9, CH3CHOAc), 7.51 [1H, ddd, J 8.2,
7.0 and 1.6, 6-H(Q)], 7.76 [1H, dd, J 8.2 and 1.6, 8-H(Q)], 7.80
[1H, ddd, J 8.2, 7.0 and 1.6, 7-H(Q)], 8.26 [1H, d, J 8.2, 5-H(Q)]
and 8.65 (1H, s, NH); minor N–N bond rotamer (observable
signals) 1.70 (3H, d, J 6.6, CH3CHOAc), 2.2 (3H, s, OCOCH3),
4.47 (1H, br d, J 5.0, CHOSi) and 8.78 (1H, s, NH). From
comparison of signals at δ 1.72 and 1.70 the ratio of rotamers
was ~10:1; δC Ϫ4.8 and Ϫ4.5 (2 × CH3), 14.6 [C(CH3)3], 18.2,
18.6, 19.5 and 21.4 (4 × CH3), 26.2 [(CH3)3CSi], 60.8 and 70.7
(2 × CH), 121.2 [CCO(Q)], 127.4, 127.7, 128.3 and 135.3
A modification of the general procedure for diacylation was
followed using MAQ 6 (2 g, 4.34 mmol), dichloromethane
(5 cm3), pyridine (0.69 g, 8.7 mmol) with 5 drops of DMF and
2-methylpropanoyl chloride (0.924 g, 8.9 mmol) and stirring
continued for five days at room temperature. A pale yellow oil
(3.2 g) was obtained on work-up, and TLC of the crude product
mixture showed the presence of five major products. Flash
column chromatography over silica gel with light petroleum–
ethyl acetate (5:1) as eluent gave a colourless oil (~1.9 g)
containing four compounds with Rf values in the range 0.58–
0.48 (see below).
Further elution with light petroleum–ethyl acetate (1:1) as
eluent gave MAQ 6 as a colourless oil (0.1 g, 5%) (Rf 0.26, 3:1
light petroleum–ethyl acetate), identical with the starting
material but comprising a 1.9:1 mixture of diastereoisomers
by NMR from comparison of the signals at δH 1.63 and 1.72
(see above).
[4 × CH(Q)], 146.7 [CN᎐C(Q)], 156.8 [C᎐N(Q)], 160.1 [CO(Q)],
᎐
᎐
170.1 and 175 (2 × CO); m/z (%) (FAB) 462 (MHϩ 89), 404 (63),
307 (22), 275 (22), 216 (20), 154 (100), 136 (93).
A solution of the oil obtained in the first fraction in metha-
nol (2 cm3) deposited crystals on setting aside to give the major
diastereoisomer 7b (0.55 g, 24%) (Rf 0.55, 5:1 light petroleum–
ethyl acetate) as colourless crystals, mp 110–112 ЊC (from
methanol) (Found: MHϩ, 532.2842. C27H41N3O6Si requires
MHϩ 532.2842) (Found: C, 60.9; H, 7.8; N, 7.9. C27H41N3O6Si
requires C, 60.5; H, 7.8; N, 7.9%); νmax/cmϪ1 1740s, 1705s, 1607s
and 1265s; δH Ϫ0.23 and 0.27 (6H, 2 × s, CH3SiCH3), 0.82 and
1.0 [6H, 2 × d, J 6.6, CH3CHCH3), 0.93 [9H, s, (CH3)3CSi], 1.12
and 1.25 (6H, 2 × d, J 6.6, CH3CHCH3), 1.60 (3H, d, J 6.9
CH3CHOAc), 2.08 (3H, s, CH3CO2), 2.19 [1H, m, 7 lines
(CH3)2CHCHOSi], 2.76 [1H, h, J 6.6, (CH3)2CHCO], 4.45
(1H, d, J 6.6, CHOSi), 6.00 (1H, q, J 6.9, CHOAc), 7.52 [1H,
ddd, J 8.2, 6.7 and 1.6, 6-H(Q)] and 7.75 [1H, dd, 8.2 and 1.6,
8-H (Q)], 7.83 [1H, ddd, J 8.2, 6.7 and 1.6, 7-H(Q)], 8.23 [1H,
dd, J 8.2 and 1.6, 5-H(Q)]; δC Ϫ3.7 and Ϫ2.5 (CH3SiCH3), 16.8,
17.0, 20.0, 20.1, 20.7 and 20.9 (6 × CH3), 19.2 [(CH3)3CSi], 26.5
[(CH3)3CSi], 32.5 and 33.4 [2 × (CH3CHCH3)], 71.5 and 76.3
(2 × CH), 121.4 [CCO(Q)], 127.8, 128.0, 128.4 and 135.9
3-[(2-Methylpropanoyl)amino]-2-[(S)-1-tert-butyldimethyl-
silyloxy-2-methylpropyl]quinazolin-4(3H)-one 8
The general procedure for monoacylation was followed using
3-aminoquinazolinone 5 (2 g, 5.8 mmol), pyridine (0.68 g,
8.6 mmol), dichloromethane (4 cm3) and 2-methylpropanoyl
chloride (0.73 g, 6.9 mmol). The brown oil obtained on work-up
was triturated with ethyl acetate–light petroleum and the solid
obtained crystallised to give the title 3-[(2-methylpropanoyl)-
amino]quinazolinone 8 as colourless crystals (1.9 g, 79%),
mp 116–118 ЊC (from light petroleum) (Rf 0.38, 5:1 light
petroleum–ethyl acetate); [α]D = ϩ27 (c 2.1 CHCl3) (Found:
MHϩ 418.2526. C22H36N3O3Si requires MHϩ 418.2526);
δH (mixture of N–N bond rotamers), major rotamer Ϫ0.02 and
0.16 (6H, 2 × s, CH3SiCH3), 0.94 and 1.04 (6H, 2 × d, J 6.6,
CH3CHCH3), 0.97 (9H, s, (CH3)3C), 1.36 and 1.39 [6H, 2 × d,
J 6.9, (CH3)2CHCO], 2.08 (1H, 7 lines CH3CHCH3), 2.73 [1H,
h, J 6.9, (CH3)2CHCO], 4.45 (1H, d, J 7.0, CHOSi), 7.53 [1H,
ddd, J 8.2, 7.0 and 1.6, 6-H(Q)], 7.75 [1H, ddd, J 8.2, 7.0 and
1.6, 7-H(Q)], 7.82 [1H, dd, J 8.2 and 1.6, 8-H(Q)], 8.21 (1H, s,
NH) and 8.31 [1H, d, J 8.2, 5-H(Q)]; δC Ϫ4.9 and Ϫ4.4
(CH3SiCH3), 18.6 [C(CH3)3], 19.4, 19.7, 33.2 and 34.5 (4 ×
CH3), 26.2 [(CH3)3C], 121.4 [CCO(Q)], 127.5, 128.8 and 135.1
[4 × CH(Q)], 146.6 [CN᎐C(Q)], 157.1 [C᎐N(Q)], 160.3 [CO(Q)]
᎐
᎐
and 170.7, 172.2 and 179.7 (3 × CO); m/z (%) (FAB) 532 (MHϩ,
89), 474 (100), 404 (68), 275 (71), 232 (48) and 187 (49). An
X-ray structure determination on crystals obtained from
methanol, showed that the DAQ 7b has (S)-configurations
for both chiral centres and an (R)-configuration for the N–N
bond.3
[CH(Q)], 146.8 [CN᎐C(Q)], 160.4 [CO(Q)] and 174.1 (CO)
᎐
(CHOSi missing); minor rotamer (observable signals) 0.04 and
0.12 (6H, 2 × s, CH3SiCH3), 2.3 (1H, 7 lines CH3CHCH3) and
4.65 (1H, d, J 7.0, CHOSi); from comparison of the signals at
δ 4.45 and δ 4.65, the ratio of N–N bond rotamers was 6:1;
m/z (%) (FAB) 418 (MHϩ, 100), 360 (77), 275 (23) and 216 (31).
Evaporation of the methanol after separation of the bulk of
DAQ 7b above and chromatography of the residue using a
chromatotron with light petroleum–ethyl acetate (18:1) as
eluent gave first DAQ 7a as a colourless gum (0.12 g, 5%)
(Rf 0.58, 5:1 light petroleum–ethyl acetate) (Found MHϩ,
532.2843. C27H41N3O6Si requires MHϩ 531.2843); δH Ϫ0.01
(6H, s, CH3SiCH3), 0.94, 1.20 and 1.64 (15H, 5 × d, J 6.7,
2 × CH3CHCH3 and CH3CHOAc), 1.00 [9H, s, (CH3)3CSi)],
1.87 [1H, m, (CH3)2CH], 2.11 (3H, s, CH3CO2), 2.51 [1H, h,
J 6.7, (CH3)2CHCO], 4.53 (1H, d, J 2.1, CHOSi), 6.06 (1H,
q, J 6.7 CHOAc), 7.52 [1H, ddd, J 8.3, 6.4 and 1.6, 6-H(Q)],
7.80 [1H, dd 8.3 and 1.6, 8-H (Q)], 7.84 [1H, ddd, J 8.3, 6.4
Reaction of 3-aminoquinazolinone 5 with racemic 2-acetoxy-
propanoyl chloride
Using the same procedure above, monoacylation of 3-amino-
quinazolinone 5 (0.3 g, 0.86 mmol) was carried out with
pyridine (0.102 g, 1.3 mmol), dichloromethane (2 cm3) and
racemic 2-acetoxypropanoyl chloride (0.195 g, 1.3 mmol). The
brown oil obtained after work-up was purified by column
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421
4419

View Article Online
and 1.6, 7-H(Q)] and 8.23 [1H, dd, J 8.3 and 1.6, 5-H(Q)];
δC Ϫ5.1 and Ϫ2.2 (CH3SiCH3), 14.9, 16.4, 20.3, 20.5 and 21.3
(6 × CH3), 18.8 [(CH3)3CSi], 26.3 (CH3)3CSi), 32.9 and 33.5
[2 × (CH3)2CH], 72.0 and 73.5 (2 × CH), 121.1 [CCO(Q)],
A crystal grown from methanol was shown by X-ray struc-
ture determination to have an (S)-configuration at the Q-2
chiral center and an (R)-configuration at the 2-acetoxy-
propanoyl chiral center and (R)-configuration for the N–N
chiral axis.3
127.5, 127.6, 128.7 and 135.8 [4 × CH(Q)], 146.9 [CN᎐C(Q)],
᎐
157.6 [C᎐N(Q)], 160.8 [CO(Q)] and 171.2, 172.6 and 180.5
᎐
(3 × CO); m/z (%) (FAB) 532 (MHϩ, 45), 474 (51), 404 (54), 275
(56), 232 (32) and 187 (32). A small number of crystals were
obtained from the gum on setting aside and were separated by
trituration with ethanol. An X-ray structure determination
showed the DAQ 7a was racemic with S(R) at Q-2 and R(S) at
2-acetoxypropanoyl centres and an S(R)-configuration for the
N–N axis. The gum obtained after removal of the ethanol after
separation of the crystals by trituration failed to crystallise
further even on standing for several months, and had [α]D =
ϩ126.8 (c 0.5 CDCl3).
Reaction of MAQ 8 with (S)-2-acetoxypropanoyl chloride
mediated by butyllithium
To a flame-dried 2-necked flask equipped with 3-way tap and
a septum cap and stirring bar was added solution of MAQ 8
(0.3 g, 0.72 mmol) in dry THF (1 ml). The flask was cooled
to Ϫ78 ЊC under an atmosphere of nitrogen and a solution of
butyllithium (0.6 cm3 of a 1.6 M solution in hexane, 1.2 mmol)
in THF (1 ml) was added with stirring via a syringe followed by
(S)-2-acetoxypropanoyl chloride (0.23, 1.4 mmol) also drop-
wise via syringe, with stirring throughout. After 1 h at Ϫ78 ЊC,
the mixture was allowed to warm to ambient temperature and
water was added. Most of the THF was removed under reduced
pressure and the residue extracted with dichloromethane (3 × 5
cm3). The combined organic extracts were washed with brine,
dried and evaporated under reduced pressure. Flash chromato-
graphy of the residue on silica with light petroleum–ethyl
acetate (10:1) as eluent gave DAQ 7a (Rf 0.58, 5:1 light
petroleum–ethyl acetate) (0.016 g, 4%) identical with that
isolated above. Further elution with the same solvent gave the
DAQ diastereoisomer 7c (Rf 0.52, 5:1 light petroleum–ethyl
acetate) (0.131 g, 34%) identical with that isolated previously.
Further elution with the same solvent gave unreacted starting
material 6 (0.152 g) identical with that isolated previously.
Further elution with the same solvent mixture gave more of
the major DAQ 7b (Rf 0.55, 5:1 light petroleum–ethyl acetate)
as colourless crystals (0.25 g, overall yield 35%).
Further elution with the same solvent mixture gave DAQ 7c
(Rf 0.52, 5:1 light petroleum–ethyl acetate) as a colourless oil
(0.161 g, 7%), which solidified on standing. Recrystallisation
gave a colourless solid, mp 106–108 ЊC (from methanol) (Found
MHϩ 532.2842. C27H42N3O6Si requires MHϩ, 532.2842); νmax
/
cmϪ1 1740s, 1705s, 1605s and 1245s; δH Ϫ0.11 and 0.04 (6H,
2 × s, CH3SiCH3), 0.92 and 1.01 (6H, 2 × d, J 6.7, CH3-
CHCH3), 0.94 [9H, s, (CH3)3CSi], 1.17 and 1.28 [6H, 2 × d,
J 6.7, (CH3)2CHCO], 1.54 (3H, d, J 6.7 CH3CHOAc), 1.82 [1H,
m, (CH3)2CHCOSi], 2.11 (3H, s, CH3CO2), 2.91 [1H, br h,
(CH3)2CH], 4.49 (1H, br s, CHOSi), 5.71 (1H, br q, J 6.7
CHOAc), 7.50 [1H, ddd, J 8.0, 6.7 and 1.6, 6-H(Q)], 7.76 [1H,
dd, J 8.0 and 1.6, 8-H(Q)], 7.82 [1H, ddd, J 8.0, 6.7 and 1.3,
7-H(Q)] 8.24 [1H, dd, J 8.0 and 1.3, 5-H(Q)]; variable-
temperature NMR studies at 400 MHz showed maximum
broadening of the quartet signal at δ 5.71 occurred at 0 ЊC, and
at Ϫ50 ЊC two signals at 4.97 (br q) and 6.01 (br s) were
obtained (ratio 1:2.7); δC Ϫ4.6 and Ϫ3.6 (CH3SiCH3), 15.3,
17.1, 19.6, 20.7, 20.7 and 21.0 (6 × CH3), 18.9 [(CH3)3CSi], 26.3
[(CH3)3CSi), 32.7 and 34.4 [2 × (CH3)2CH], 70.6 and 74.4
(2 × CH), 121.3 [CCO(Q)], 127.8, 128.0, 128.5 and 135.8
3-[(S)-2-Acetoxypropanoylamino]-2-methylquinazolin-4(3H)-
one 11
Using the general procedure for monoacylation, 3-amino-
quinazolinone 9 (1 g, 5.7 mmol), pyridine (0.9 g, 11.4 mmol),
dichloromethane (3 cm3) and (S)-2-acetoxypropanoyl chloride
(1.72 g, 11.4 mmol) were stirred for 24 h at room temperature.
After work-up the yellow oil obtained was purified using
column chromatography on silica with light petroleum–ethyl
acetate (1:1) as eluent to give MAQ 11 (1.12 g, 68%) (Rf = 0.36)
as a colourless oil (1.12 g, 68%) (Found: MHϩ 290.1141.
C14H15N3O4 requires MHϩ 290.1141); νmax/cmϪ1 3390m 1740s,
1692s and 1612s; δH (mixture of N–N bond rotamers) major
rotamer 1.65 (3H, d, J 6.6, CH3CHOAc), 2.22 (3H, s, CH3CO),
2.48 (3H, s, CH3), 5.42 (1H, br q, J 6.6, CHOAc), 7.41 [1H, br
m, 6-H(Q)], 7.59 [1H, d, J 8.0, 8-H(Q)], 7.73 [1H, br m,
7-H(Q)], 8.13 [1H, d, J 8.0, 5-H(Q)] and (1H, s, NH); δC 17.7,
21.3 and 21.6 (3 × CH3), 70.2 (CHOAc), 120.8 [CCO(Q)],
[4 × CH(Q)], 146.6 [CN᎐C(Q)], 156.9 [C᎐N(Q)], 160.4 [CO(Q)],
᎐
᎐
170.4, 171.7 and 179.4 (3 × CO); m/z (%) (FAB) 532 (MHϩ,
45), 474 (51), 404 (54), 275 (56), 232 (32) and 187 (32). A
crystal structure determination showed that DAQ 7c has (S)-
configurations for both chiral centres and an (S)-configuration
for the N–N chiral axis.3
Further elution with the same solvent mixture gave DAQ 7d
(Rf 0.48, 5:1 light petroleum–ethyl acetate) as a viscous, colour-
less oil which crystallized on standing (0.60 g, 26%), mp 68–
70 ЊC (from methanol) (Found MHϩ 532.2843. C27H42N3O6Si
requires MHϩ, 532.2843); νmax/cmϪ1 1740s, 1707s, 1610s and
1245s; δH Ϫ0.12 and 0.12 (6H, 2 × s, CH3SiCH3), 0.52 and 0.96
(6H, 2 × d, J 6.7, CH3CHCH3), 0.94 [9H, s, (CH3)3CSi], 1.23
and 1.27 (6H, 2 × d, J 6.7, CH3CHCH3), 1.56 (3H, d, J 6.9
CH3CHOAc), 2.09 (3H, s, CH3COO), 2.92 [1H, m, (CH3)2-
CHCOSi], 3.06 [1H, br h, (CH3)2CH], 4.34 (1H, d, J 4.6,
CHOSi), 5.73 (1H, br q, J 6.7 CHOAc), 7.52 [1H, ddd, J 8.3, 6.9
and 1.6, 6-H(Q)], 7.75 [1H, dd, J 8.3 and 1.6, 8-H(Q)], 7.82 [1H,
ddd, J 8.3, 6.9 and 1.3, 7-H(Q)], 8.25 [1H, dd, J 8.3 and 1.3,
5-H(Q)]; variable-temperature NMR studies at 400 MHz
showed maximum broadening of the quartet signal at δ 5.73 for
the CHOAc signal occurred at 0 ЊC, and at Ϫ50 ЊC two signals
at δ 5.57 (br q) and 6.08 (br s) were obtained (ratio 1:2.9);
δC Ϫ4.3 and Ϫ3.2 (CH3SiCH3), 16.2, 17.8, 19.9, 20.3, 20.7 and
20.9 (6 × CH3), 19.0 [(CH3)3CSi], 26.4 [(CH3)3CSi], 32.5 and
34.4 [2 × (CH3)2CH], 70.3 and 75.9 (2 × CH), 121.3 [CCO(Q)],
127.1, 127.2, 127.6 and 135.5 [4 × CH(Q)], 147.3 [CN᎐C(Q)],
᎐
155.5 [CN(Q)], 160.5 [CO(Q)] and 170.7 and 171.4 (2 × CO);
δH minor rotamer (observable signals) 1.62 (3H, d, J 6.6,
CH3CHOAc), 2.20 (3H, s, CH3CO), 2.52 (3H, s, CH3) and (1H,
s, NH); m/z (%) (FAB) 290 (MHϩ, 48) and 154 (100). From
comparison of the signals at δ 2.22 and 2.20, the ratio of N–N
bond rotamers was ~1.2:1.
3-{Bis[(S)-2-acetoxypropanoyl]amino}-2-methylquinazolin-
4(3H)-one 13
Using the general procedure for diacylation, amide 11 (1 g, 3.46
mmol), pyridine (0.55, 6.9 mmol) and distilled (S)-2-acetoxy-
propanoyl chloride (1.04 g, 6.9 mmol) were stirred continuously
in dichloromethane (3 cm3) for three days at room temperature.
The brown oil obtained was purified by column chromato-
graphy on silica using light petroleum–ethyl acetate (1:1) as
eluent to give the DAQ 13 as a colourless oil. Re-chromato-
graphy using a chromatotron and light petroleum–ethyl acetate
(3:1) as eluent gave (S,S)-DAQ 13 as a colourless oil (0.6 g,
43%) which solidified on standing and crystallised to give
colourless solid, mp 124–126 ЊC (from ethanol) (Found: MHϩ
127.9, 128.0, 128.4 and 135.8 [4 × CH(Q)], 146.6 [CN᎐C(Q)],
᎐
156.4 [C᎐N(Q)], 160.3 [CO(Q)], 170.0, 171.7 and 178.7
᎐
(3 × CO); m/z (%) (FAB) 532 (MHϩ, 100), 474 (78), 404 (60),
275 (64), 232 (44) and 187 (53).
4420
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421

View Article Online
404.1457. C19H21N3O7 requires MHϩ 404.1458); νmax/cmϪ1
1745s, 1700s and 1613s; δH 1.40 and 1.36 (6H, 2 × d, J 6.7,
2 × CH3CH), 2.09 and 2.12 (6H, 2 × s, 2 × CH3CO2), 2.36 (3H,
s, CH3), 5.01 and 5.97 (2H, 2 × q, J 6.7, 2 × CHOAc), 7.50 [1H,
ddd, J 8.3, 7.0 and 1.3, 6-H(Q)], 7.68 [1H, dd, J 8.3 and 1.3,
8-H(Q)], 7.81 [1H, ddd, J 8.3, 7.0 and 1.3, 7-H(Q)] and 8.19
[1H, dd, J 8.3 and 1.3, 5-H(Q)]; δC 16.4, 16.8, 20.7, 20.9 and
21.3 (5 × CH3), 68.9 and 71.7 (2 × CHOAc), 120.8 [CCO(Q)],
CHOAc), 7.47 [1H, ddd, J 8.2, 7.0 and 1.3, 6-H(Q)], 7.73 [1H,
dd, J 8.2 and 1.3, 8-H(Q)], 7.81 [1H, ddd, J 8.2, 7.0 and 1.3,
7-H(Q)] and 8.19 [1H, dd, J 8.2 and 1.3, 5-H(Q)]. There was no
change in the 1H NMR spectrum (at 400 MHz) at temperatures
down to Ϫ50 ЊC. δC 16.5, 16.9, 20.7, 20.4, 32.3 and 22.9 (6 ×
CH3), 30.6 (CH3CHCH3), 68.9 and 71.5 (2 × CHOAc), 120.8
[CCO(Q)], 127.5, 127.5, 128.2 and 135.9 [4 × CH(Q)], 147.3
[CN᎐C(Q)], 160.7 [CN(Q)], 163.5 [CO(Q)] and 170.5, 170.7,
᎐
127.7, 127.8, 127.9 and 136.2 [4 × CH(Q)], 147.2 [CN᎐C(Q)],
171.4 and 172.5 (4 × CO); m/z (%) (FAB) 432 (MHϩ, 100), 372
᎐
156.1 [CN(Q)], 160.0 [CO(Q)] and 170.2, 171.1, 171.3 and 172.0
(4 × CO); m/z (%) (FAB) 404 (MHϩ, 48), 290 (40) and 154 (100).
A crystal grown from methanol was shown by an X-ray struc-
ture determination to have (S)-configurations at both chiral
centres.6
Variable-temperature NMR studies at 400 MHz were carried
out on DAQ 13 at different temperatures (Ϫ50, Ϫ25, 0 and
27 ЊC) but there was no change in the spectrum.
Further elution with the same solvent gave a mixture of
meso-DAQ diastereoisomers 13b and possibly 13c as a colour-
less oil (0.06 g, 4%) (Rf 0.53, 1:1 light petroleum–ethyl acetate)
(Found: MHϩ 404.1457. C19H21N3O7 requires MHϩ 404.1458);
νmax/cmϪ1 1745s, 1700s and 1613s; δH (major diastereoisomer)
1.55 (6H, d, J 6.7, CH3CHOAc), 2.10 (6H, s, 2 × CH3CO2), 2.53
[3H, s, CH3(Q)], 5.47 (2H, br q, J 6.7, 2 × CH3CHOAc), 7.48
[1H, ddd, J 8.0, 6.5 and 1.1, 6-H(Q)], 7.66 [1H, dd, J 8.0 and
1.1, 8-H(Q)], 7.79 [1H, ddd, J 8.0, 6.5 and 1.1, 7-H(Q)] and 8.25
[1H, dd, J 8.0 and 1.1, 5-H(Q)]; δC 16.9, 20.7 and 21.2
(5 × CH3), 70.1 (2 × CHOAc), 121.1 [CCO(Q)], 127.6, 127.8,
(40), 318 (78) and 230 (50).
Crystal structure determination of 7a
Crystal data. C27H41N3O6Si, M = 531.72, triclinic, space group
¯
P1, a = 10.781(2), b = 11.994(3), c = 12.330(3) Å, α = 79.69(2),
β = 81.95(2), γ = 78.00(3)Њ, V = 1525.6(6) Å3, T = 200 K, Z = 2,
µ(Mo-Kα) = 0.118 mmϪ1 colourless block, crystal dimensions
0.39 × 0.34 × 0.28 mm. Data were measured on a Siemens P4
diffractometer with graphite monochromated Mo-Kα radiation
(λ = 0.7107 Å) using an ω scan technique. Three standard reflec-
tions monitored every 100 scans showed no significant variation
in intensity, the reflections were corrected for Lorentz and
polarisation effects. 3388 data were measured (2.7 < θ < 21Њ),
with 3265 independent reflections (merging Rint = 0.059). No
corrections for absorption or crystal decay were required.
The structures were solved by direct methods and refined by
full-matrix least squares on F2 using the program SHELXL-97.
All hydrogen atoms were included in calculated positions
(C–H = 0.96 Å) using a riding model. Full matrix least squares
based on F2 gave R1 = 0.0699, wR2 = 0.149 for all data, for 319
parameters.
128.1 and 135.9 [4 × CH(Q)], 147.0 [CN᎐C(Q)], 155.2
᎐
[C᎐N(Q)], 160.0 [CO(Q)] and 170.5 and 171.6 (4 × CO); minor
᎐
diastereoisomer (observable signals) 2.03 (6H, s, 2 × CH3CO2),
2.46 [3H, s, CH3(Q)], 5.59 (2H, br q, J 6.7, 2 × CH3CHOAc)
and 8.20 [1H, dd, J 8.0 and 1.1, 5-H(Q)]. From the integration
of the signals at δ 5.47 and 5.59 the ratio of major:minor dias-
tereoisomers is 2.5:1; m/z (%) (FAB) 404 (MHϩ, 61), 290 (25)
and 154 (45). Re-chromatography using a chromatotron and
light petroleum–ethyl acetate (5:1) as eluent gave a pure sample
of the major meso-diastereoisomer 13b as colourless oil (8 mg).
Acknowledgements
We thank the Ministry of Education (Saudi Arabia) for funding
(to A. G. Al-S.).
References
1 R. S. Atkinson, E. Barker and M. J. Sutcliffe, Chem. Commun., 1996,
1051.
2 R. S. Atkinson, E. Barker, P. J. Edwards and G. A. Thomson,
J. Chem. Soc., Perkin Trans. 1, 1996, 1047.
3 Preliminary communication: A. G. Al-Sehemi, R. S. Atkinson,
J. Fawcett and D. R. Russell, Chem. Commun., 2000, 43.
4 E. Barker, PhD Thesis, University of Leicester, 1995; see also,
G. A. El-Hiti, Spectrosc. Lett., 1999, 32, 671.
5 cf. E. A. Noe and M. Raban, J. Am. Chem. Soc., 1975, 77, 5871.
6 A. G. Al-Sehemi, R. S. Atkinson, J. Fawcett and D. R. Russell,
Tetrahedron Lett., 2000, 41, 2243.
7 R. S. Atkinson, S. Ulukanli and P. J. Williams, J. Chem. Soc., Perkin
Trans. 1, 1999, 2121.
8 A. N. Van Draanen, S. Arseniyadis, M. T. Crimmins and
C. H. Heathcock, J. Org. Chem., 1991, 56, 2499.
9 R. S. Atkinson, P. J. Edwards and G. A. Thomson, J. Chem. Soc.,
Perkin Trans. 1, 1994, 3209.
Racemic 3-[bis(2-acetoxypropanoyl)amino]-2-isopropylquin-
azolin-4(3H)-one 14
3-Aminoquinazolinone 10 (1.5 g, 7.4 mmol), pyridine (1.15 g,
14.7 mmol) and racemic 2-acetoxypropanoyl chloride (3 g, 20
mmol) in dichloromethane (5 cm3) were set aside at room tem-
perature for 4 days. The yellow oil (1.9 g) obtained on work-up
used in the general procedure was purified by column chrom-
atography on silica using light petroleum–ethyl acetate (2:1)
as eluent to give racemic DAQ 14 as colourless crystals (1.8 g,
56%), mp 114–116 ЊC, [α] = Ϫ1 (c 1, CHCl3) (Found: MHϩ
432.1771. C21H25N3O7 requires MHϩ 432.1771); δH 1.16 and
1.39 (6H, 2 × d, J 6.6, CH3CHCH3), 1.35 and 1.62 (6H, 2 × d,
J 6.9, 2 × CH3CHOAc), 2.1 and 2.16 (6H, 2 × s, 2 × CH3CO2),
3.4 (1H, h, J 6.6, CH3CHCH3), 4.87 and 6.0 (2H, 2 × q, J 6.9,
J. Chem. Soc., Perkin Trans. 1, 2000, 4413–4421
4421