Synthesis of SB-214857 using copper catalysed amination of arylbromides with L-aspartic acid

Article abstract of DOI:10.1055/s-2001-16780

An efficient synthesis of a 1,4-benzodiazepine was accomplished using copper catalysed amination of an arylbromide with the soluble bis-tetrabutylammonium salt of L-aspartic acid as a key step.

Full text of DOI:10.1055/s-2001-16780

LETTER
1423
Synthesis of SB-214857 using Copper Catalysed Amination of Arylbromides
with L-Aspartic Acid
Jean-Baptiste Clement, Jerome F. Hayes,* Helen M. Sheldrake, Peter W. Sheldrake, Andrew S. Wells
GlaxoSmithKline Pharmaceuticals, Old Powder Mills, Nr Leigh, Tonbridge, Kent, TN11 9AN, UK
E-mail: hayesj00@sbphrd.com
Received 11 May 2001
of L-aspartic acid dimethyl ester with diphenyliodonium
Abstract: An efficient synthesis of a 1,4-benzodiazepine was ac-
hexafluorophosphate in dichloromethane using copper (I)
complished using copper catalysed amination of an arylbromide
iodide as catalyst and were very pleased to isolate 2 in
with the soluble bis-tetrabutylammonium salt of L-aspartic acid as a
76% yield and >95% ee (Scheme 2). Interestingly use of
copper (II) acetate as catalyst resulted in significant race-
mization (70%ee).8
key step.
Key words: copper catalysis, aromatic amination, amino acids,
benzodiazepines, SB-214857
The product 2 was converted to 1 via the route illustrated
(Scheme 2). A key feature involved the Mannich reaction
of 2 with bis(ethoxymethyl)methylamine 39 to form
quinazoline 4 in 53% yield.10 Hydrolysis of the quinazo-
line ring to form intermediate 5 was accomplished using a
65:35 mixture of methanol: 0.14 M hydrochloric acid and
the crude product 5 was cyclised with sodium methoxide
in methanol to give 1. The overall conversion from 4 to 1
was 50% (40% ee). Use of hydrogen chloride in dry meth-
anol for the conversion of 4 to 5 resulted in a very slow re-
action (less than 20% conversion within 5 days) but the
resulting product 1 was obtained with 86% ee.
Lotrafiban (SB-214857) is a potent GPIIb/IIIa receptor
antagonist and therefore inhibits platelet aggregation and
possibly thrombosis.1 The route of manufacture of SB-
214857 involves enzymatic resolution of the racemate of
the 1,4-benzodiazepine ester 1 to obtain the chiral benzo-
diazepine nucleus.2 In order to avoid this resolution we ac-
tively pursued a chiral synthesis of 1. The most direct
approach would involve as a key step, amination of a suit-
able arene with an L-aspartic acid derivative (Scheme 1).
Despite recent significant developments in the field of ar-
omatic amination3,4 to date there are no examples in the
literature of the N-arylation of aspartic acid. The likely in-
stability of aspartic acid at very high temperatures would
probably make the established but harsh Ullmann type
amination conditions5 unsuitable. There were two reports
in the literature however which encouraged us to pursue
this amination approach. We noted with interest a paper
by Scherrer and Beatty6 which described very mild condi-
tions for the copper catalysed amination of diphenyliodo-
nium-2-carboxylate with 2,3-dimethyl aniline. We
envisaged that the exceptional leaving group ability of
the aryl iodonium group7 allowed these mild conditions
to be employed. We therefore attempted the reaction
The problems associated with this route (racemisation or
slow quinazoline ring hydrolysis) led us to pursue alterna-
tive conditions which would allow amination of 2-bro-
mobenzaldehyde or a similar analogue. We anticipated
that conversion of the aminated product to benzodiaz-
epine 1 would be less problematical. We noticed with in-
terest the work published by D. Ma where certain amino
acids were reacted with aryl iodides or bromides in the
presence of palladium(II)/copper (I) iodide catalyst and
bases.11 He subsequently reported a developed system
using only copper(I) iodide and potassium carbonate.12
In both papers it is mentioned that the reaction only suc-
ceeds for -amino acids bearing hydrophobic substituents
Scheme 1
Synlett 2001, No. 9, 1423–1427 ISSN 0936-5214 © Thieme Stuttgart · New York

1424
J.-B. Clement et al.
LETTER
Scheme 2 Reagents: a) Ph2I+ PF6 , CuI, (20mol%), Pri2NH, CH2Cl2, r.t., 2 days, 76%. b) HBF4 Et2O, 3, MeCN, 53%. c) MeOH, HCl.
d) NaOMe, MeOH, 50%.
Table 1 Copper Catalysed Amination of Aryl Halides with Amino Acids
Conditions: a) CuI, 0.2 eq., MeCN, Bu4NOH (40% aqueous solution), b) MeOH, AcCl, reflux, 2h.
aIsolated yield after esterification unless otherwise indicated. bDetermined by chiral HPLC. c0.1Eq. of CuI was used.
dIsolated yield of the free acid. eIsolated as pyrrolidinone 6. fIsolated as ester 7.
(exemplified by valine, phenylalanine and proline) and soluble in dimethylformamide, dimethylacetamide or ac-
that those with hydrophilic substituents (specifically etonitrile. Significantly this salt reacted with iodobenzene
glutamic acid, serine or glycine) gave no product under in the presence of copper(I)iodide to form N-phenyl
these conditions. Arylation of aspartic acid was not men- aspartic acid in either dimethylacetamide or acetonitrile.
tioned. We felt that this lack of reactivity was due to insol- Esterification of the product using HCl in methanol gave
ubility of the amino acids. We therefore looked for a way 2 in 37% yield after chromatography and 89% ee. In order
to solubilise them in dipolar aprotic solvents and were in- to determine the generality of the procedure we solubi-
terested to find that treatment of aspartic acid with tet- lised several other amino acids as their tetrabutyammoni-
rabutylammonium hydroxide gave a salt which was um salts and found that with copper catalysis they also
Synlett 2001, No. 9, 1423–1427 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Synthesis of SB-214857 using Copper Catalysed Amination
1425
reacted with a variety of aryl halides (Table 1). Yields of In an attempt to improve the efficiency of this amination
the N-arylated products were modest using these unopti- process we screened several copper catalysts for the reac-
mised conditions but we had demonstrated for the first tion between 2-bromoanisole and aspartic acid (Table 2).
time that by solubilising amino acids with hydrophilic This screen indicated that catalytic efficiency was influ-
substituents as their tetrabutylammonium salts, they could enced mainly by the oxidation state of the copper where
be arylated under mild conditions using copper catalysis. Cu(0) > Cu(I) > Cu(II) and that the optimal catalyst sys-
No racemisation of the products derived from L-valine tem was Raney copper (19 mol%)mixed with copper (I)
(entry 2) or L-proline (entry 5) was detected. Glutamate oxide (1 mol%).
derivative 6 suffered some racemisation and was obtained
with 50% ee (entry 4). Interestingly there was no signifi-
cant difference in yield between aryl halides containing
electron donating and electron withdrawing substituents
We then applied the optimum conditions found (entry 5)
and the copper(I)iodide conditions (entry 9) to the amina-
tion of the aryl bromides 10 and 11 which were required
for a facile synthesis of benzodiazepine 114 (Table 3). The
(compare entry 7 with 10 and entry 11 with 12). This is
most useful conditions in terms of yield and optical purity
consistent with a single electron transfer mechanism oper-
were found in entry 1. This product was subsequently es-
ating.13 2-Bromobenzonitrile (entry 7) gave slightly high-
terified with hydrogen chloride in methanol to give 13 in
40% yield and 90% ee.
er yields than the iodide (entry 8).
It is striking that Raney copper copper I oxide causes
significantly more racemisation than copper(I)iodide but
the reasons for this are not clear. A possible mechanism
for racemisation could involve reversible -hydride elim-
ination from aspartic acid when complexed to copper.15–17
Table 2 Amination of 2-Bromoanisole with L-Aspartic acid bis-tetrabutylammonium salt
using Copper Catalysis
Conditions: a) L-aspartic acid, 1.0eq.; tetrabutylammonium hydroxide (40% aqueous solution),
2.0eq.; copper catalyst, MeCN, reflux, 18h
aYields were measured by HPLC solution assay against a reference standard.
Synlett 2001, No. 9, 1423–1427 ISSN 0936-5214 © Thieme Stuttgart · New York

1426
J.-B. Clement et al.
LETTER
Table 3 Copper Catalysed Amination of Protected 2-bromobenzaldehyde derivatives with L-Aspartic Acid
Conditions: a) L-aspartic acid, tetrabutylammonium hydroxide, 1.8 eq., catalyst, MeCN, reflux, b)MeOH,
AcCl, reflux
aHPLC solution assay against a reference standard. bEE = 93% after 1h, 79% after 3h and 55% after 18h.
Scheme 3 Conditions: a) MeNH2 in ethanol b) Pd-C, MeOH, H2, c) stir in MeOH, rt, 3 days.
Completion of the synthesis to 1 is illustrated (Scheme 3).
The imine 14 was formed by reaction of 13 with methy-
lamine in methanol and in situ hydrogenation over 10%
palladium on charcoal gave the amine 15. After stirring
for 4 days in methanol at room temperature this cyclised
to the required benzodiazepine 1. Racemic benzodiaz-
epine 1 (which is 10 times less soluble than the single
enantiomer in methanol) precipitated and was filtered off
to leave 1 (79% from 13, 99% ee) in solution. Conversion
of 1 to SB-214857 has been described in a recent publica-
tion.18
Acknowledgement
The authors wish to thank Kaye Cattanach for assistance with chiral
HPLC.
References and Notes
(1) Nichols, A.J.; Vasko, J.A.; Koster, P.F.; Valocik, R.E.; and
Samanen, J.M. GPIIb/IIIa Antagonists as Novel M
Drugs:Potential Therapeutic Agents. In Cellular
Adhesion:Molecular Definition to Therapeutic Potential;
Metcalf, B.W.; Dalton, B.J.; Poste, G. Eds.; Plenum Press:
New York 1994, pp213-237.
In conclusion we have shown that amino acids containing
hydrophilic substituents can be N-arylated under mild
conditions using copper catalysis once solubilised as their
tetrabutylammonium salts. Raney copper containing 5%
copper(I)oxide is a very effective catalyst but it causes
significant racemisation of aspartic acid. Copper(I)iodide
is a less efficient catalyst but it does not cause racemisa-
tion to the same extent. Using this methodology, benzodi-
azepine 1 an advanced intermediate to Lotrafiban was
prepared from bromobenzene or more conveniently from
2-bromobenzaldehyde.
(2) Wells, A.S WO 9829561. Chem. Abstr. 1998, 129, 121708.
(3) a) Harwig, J.F. Synlett 1997, 4, 329-340; b) Buchwald, S.L.
Acc. Chem. Res. 1998, 31, 805 - 818.
(4) Chan, D.M.T.; Monaco, K.L.; Wang, R-P. and Winters, M.P.
Tetrahedron Lett. 1998, 39, 2933.
(5) Hager, F.D Org. Synthesis, Col. Vol. 1, 2nd Edn. 1948, 544.
(6) Scherrer, R.A. and Beatty, H.R. J. Org. Chem. 1980, 45, 2127.
(7) Varvoglis, A. Hypervalent Iodine in Organic Synthesis In Best
Synthetic Methods; Katritzky, A.R., Meth-Cohn, O and Rees,
C.W., Eds., Academic Press: London 1997, pp4-7.
(8) After completion of this work Kang et al published their
results on aromatic amination of hypervalent iodine
compounds with simple amines; Kang, S-K.; Lee, S-H. and
Lee D. Synlett 2000, 7, 1022.
Synlett 2001, No. 9, 1423–1427 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Synthesis of SB-214857 using Copper Catalysed Amination
1427
(9) Heaney, H and Papageorgiou, G. Tetrahedron 1996, 52, 3473.
(10) Use of 1,3,5-trimethylhexahydro-1,3,5-triazine as electrophile
resulted in only a 25% conversion to 4 leaving a significant
amount of unreacted starting material.
(11) Ma, D. and Yao, J. Tetrahedron Asymmetry 1996, 7, 3075.
(12) Ma, D.; Zhang, Y.; Yao, J.; Wu, S.and Tao, F. J. Am. Chem.
Soc. 1998, 120, 12459-12467.
a volume of 15 mL by atmospheric distillation. Further
acetonitrile (15 mL) was added and distilled off and this
procedure repeated until the boiling point of the
distillate = 81 °C. Copper(I)iodide (286 mg, 1.51 mmol) and
the imine 10 (1.80 g, 7.52 mmol) were then added at room
temperature to the resultant dried solution and the reaction
mixture was heated under reflux for 2 h. under argon. HPLC
solution assay against a reference standard indicated a 55%
yield of 12. (Waters Symmetry C-18, 4.6 250mm. Flow rate
1.0 mL/min. Eluant A: 0.1% methanesulfonic acid in water;
Eluant B: 0.1% methanesulfonic acid in acetonitrile. Gradient
90% to 50% A over 15 min, 50% to 90% A over 1 min, 90%
A over 14 min.; uv detection at 254 nm). The reaction mixture
was concentrated to dryness in vacuo and the resultant oil was
dissolved in methanol (20 mL). Acetyl chloride (2.7 mL,
37.60 mmol) was added carefully with vigorous stirring and
the reaction mixture was then heated under reflux for 2 h. The
mixture was then concentrated to dryness in vacuo and the
residue was dissolved in ethyl acetate (60 mL) and washed
with ammonia (40 mL of an 8% aqueous solution) and water
(40 mL) and then was dried (MgSO4). The solvent was
removed in vacuo and the residue was chromatographed on
silica gel using ethyl acetate : hexane 1:3 as eluant. The
product was obtained as a yellow oil (0.80 g, 40%); HRMS,
Found M+ 265.0950. C13H15NO5 requires 265.0950. IR 1740,
1650, 1580, 1510 and 1430 cm-1, 1H NMR (CDCl3): 2.91
(dd, ABX, J = 6.2, 16.0 Hz, 1H), 2.97 (dd, ABX, J = 6.2, 16.1
Hz, 1H), 3.71 (s, 3H), 3.78 (s, 3H), 4.67 (m, 1H), 6.72 (d,
J = 8.5 Hz, 1H), 6.79 (dt, J = 0.8, 7.5 Hz, 1H), 7.41 (dt,
J = 1.4, 8.0 Hz, 1H), 7.51 (dd, J = 1.6, 7.7 Hz, 1H), 8.78 (d,
J = 8.2 Hz, 1H) and 9.83 (s, 1H).
(13) Lindley, J. Tetrahedron,1984, 40, 1433.
(14) 2-Bromobenzaldehyde, which gave a complex mixture under
the amination reaction conditions, was readily converted to
the imine 10 by treatment with tert-butylamine in acetonitrile.
(15) We established that under the amination reaction conditions it
is the bis-tetrabutylammonium salt of aspartic acid that
racemises and not the product since prolonged exposure (18 h)
of the bis-tetrabutylammonium salt of aspartic acid with
Raney copper-copper(I)oxide followed by reaction with
bromobezene for 30 min. gave 2 after esterification (20% ee)
Under the same conditions the bis-tetrabutylammonium salt of
N-phenylaspartic acid did not racemise.
(16) Teplyakov, A.V. and Bent, B.E. J. Am. Chem. Soc.,1995, 117,
10076-10087.
(17) Buelow, M.T. and Gellman, A.J. J. Mol. Catal. 1998, 131,
55-70.
(18) Andrews, I.; Atkins, R.; Bellingham, R.F.;Breen, G.; Carey,
J.S.; Etridge, S.K.; Hayes, J.F.; Hussain, N.; Morgan, D.O.;
Share, A.; Smith, S.; Walsgrove, T.C. and. Wells, A.S.
Tetrahedron Lett. 2001, 42, 4915.
(19) Experimental procedure for preparation of N-(2-
formyphenyl)aspartic acid dimethylester from aryl
bromide 10. To a suspension of L-aspartic acid (1.00 g, 7.52
mmol) in acetonitrile (20 mL) was added tetrabutyl-
ammonium hydroxide (8.85 mL of a 40% aqueous solution,
13.53 mmol) at room temperature. The resultant clear solution
was concentrated to dryness in vacuo and was then redis-
solved in acetonitrile 30 mL. The solution was concentrated to
Article Identifier:
1437-2096,E;2001,0,09,1423,1427,ftx,en;D12601ST.pdf
Synlett 2001, No. 9, 1423–1427 ISSN 0936-5214 © Thieme Stuttgart · New York