Synthesis of Lapatinib via direct regioselective arylation of furfural

Article abstract of DOI:10.1016/j.tetlet.2014.09.039

A new synthesis of Lapatinib, an orally active drug for breast cancer, is described. The synthesis involves a palladium catalyzed regioselective arylation of furfural with 6-bromo-N-(3-chloro-4-((3-fluorobenzyl)oxy)phenyl)quinazolin-4-amine. This key step replaces an atom inefficient Suzuki cross coupling reaction used in a previously disclosed route and significantly shortens the synthesis.

Full text of DOI:10.1016/j.tetlet.2014.09.039

Tetrahedron Letters 55 (2014) 6007–6010
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis of Lapatinib via direct regioselective arylation of furfural
a
a
b
Greg Erickson a, Jiasheng Guo a, , Mike McClure , Mark Mitchell , Marie-Catherine Salaun ,
⇑
Andrew Whitehead b,
⇑
a Global API Chemistry, Product Development, GlaxoSmithKline, Research Triangle Park, NC 27709, United States
b 2nd Generation and Process Robustness, Product Development, GlaxoSmithKline, Currabinny, Carrigaline, County Cork, Ireland
a r t i c l e i n f o
a b s t r a c t
Article history:
A new synthesis of Lapatinib, an orally active drug for breast cancer, is described. The synthesis involves a
palladium catalyzed regioselective arylation of furfural with 6-bromo-N-(3-chloro-4-((3-fluoroben-
zyl)oxy)phenyl)quinazolin-4-amine. This key step replaces an atom inefficient Suzuki cross coupling
reaction used in a previously disclosed route and significantly shortens the synthesis.
Ó 2014 Elsevier Ltd. All rights reserved.
Received 1 August 2014
Revised 4 September 2014
Accepted 8 September 2014
Available online 16 September 2014
Keywords:
Lapatinib
Direct arylation
C–H activation
Pivalic acid
Palladium acetate
Lapatinib (Tykerb, 1), a dual ErbB1/ErbB2 tyrosine kinase inhib-
itor, is an orally active drug discovered by GlaxoSmithKline for the
treatment of patients with advanced or metastatic breast cancer.1–3
O
OHC
2
90%
(EtO)3CH/EtOH,
NH4NO3, reflux
Cl
O
HN
O
(EtO)2HC
O
S
O
O
NH
N
F
1) nBuLi/THF or DME, -30 oC
2) B(OiPr)3, -30 oC
3) AcOH, -30 oCto r.t.
4) H2O
N
52%
HO3S
H O
1
2
2
Cl
O
Lapatinib (ditosylate monohydrate)
HN
N
+
B(OH)2
O
OHC
I
N
F
A previously disclosed route of synthesis is detailed in Scheme 1
and the key step in the synthesis is the Suzuki cross coupling of
5-formyl-2-furanylboronic acid (3) and iodoquinazoline (4).4
Although this reaction is very high yielding it is atom inefficient
as it produces a significant amount of iodide and boron wastes.
Furthermore the synthesis of the boronic acid (3) is lengthy and
requires low temperature.
3
4
87%
10% Pd/C, Et3N, EtOH
60 oC
Cl
NH2
O
MeO2S
1) iPr2NEt,
6
NaBH(OAc)3, THF, 30 oC
HN
OHC
O
N
F
1
2) TsOH
80%
N
5
⇑
Corresponding authors. Tel.: +1 919 483 1076 (J.G.), +353 21 4512 470 (A.W.).
E-mail addresses: Jiasheng.X.Guo@gsk.com (J. Guo), Andrew.J.Whitehead@gsk.
Scheme 1. Previous disclosed route of synthesis.
com (A. Whitehead).
http://dx.doi.org/10.1016/j.tetlet.2014.09.039
0040-4039/Ó 2014 Elsevier Ltd. All rights reserved.

6008
G. Erickson et al. / Tetrahedron Letters 55 (2014) 6007–6010
(9) were the optimal substrates albeit significant amounts of C3
5
and C4-arylated regio-isomers (10 and 11) along with desbromo
(12), homocoupling (13), and 3,5-bis-arylated furfural (14) side
products were formed (Scheme 3). It was also found that the best
conversions were achieved when furfural was used in excess. In
this case no additional solvent was required. The use of PdCl2/
Cy3P/KOAc in DMF at 110 °C gave similar results to the Doucet cat-
alyst system.
Cl
O
HN
N
+
X
O
N
F
R
7
The Fagnou protocol14 was then investigated. Heating a mixture
of 9, furfural (2–10 equiv), Pd(OAc)2, Cy3P, PivOH (0.3 equiv), and
K2CO3 in DMF at 130 °C resulted in poor conversion to 5 along with
many side products. The reaction was sluggish and 9 was not com-
pletely consumed even after very long reaction time (24 h). None
of the other solvents screened (DMAc, diethyl carbonate, Me-THF,
cyclopentyl methyl ether, toluene, trifluorotoluene, xylene, and
DMSO) gave better results. However, when furfural was used as
the solvent, the reaction proceeded very smoothly at 130 °C and
9 was completely consumed in 2–3 h.
8
R= CHO,CH2OH, CH2OAc,CH(OEt)2
X= Br, Cl
Scheme 2. Retrosynthetic analysis.
We envisioned that the Suzuki coupling could be replaced by a
single step direct arylation of a suitably functionalized furan (7)
and quinazoline (8) derivatives as detailed in Scheme 2.
We then focused on the optimization of the direct arylation of
furfural under various conditions. Pd(OAc)2 gave the best results
among the catalysts evaluated, which include Pd(OAc)2, PdCl2,
Pd(OH)2/C, Pd(Cy3P)2Cl2, and Pd2(dba)3.
If successful, this direct arylation approach would not only sig-
nificantly shorten the synthesis of Lapatinib but also eliminate
both the undesirable iodide and boron wastes, remove the need
of pre-preparing the boronic acid and is therefore more atom-eco-
nomical and environmentally friendly.
A number of ligands were screened and it was found that the
ligands played a key role in reaction rate and regioselectivity
(Chart 1). The reaction with Me(t-Bu)2PÁHBF4 was the fastest and
gave the highest yield among the ligands screened. The reaction
with Cy3PÁBHF4 was slightly slower than Me(t-Bu)2PÁHBF4 but sig-
nificantly faster and gave much higher yields than other ligands.
The reaction proceeded either very slowly or did not proceed with
some very bulky ligands such as t-Bu3P and strong electron-donat-
ing ligands such as (o-MeOPh)3P. Prolonged reactions gave more
desbromo side product. More undesired regioisomers were formed
with bidentate ligands such as DPPM and BDPP.
The effect of various additives and bases was also studied.
Stronger, weaker, and bulkier acids (including diacids) were evalu-
ated. Pivalic acid was found to be the most efficient acid for the
arylation, which is consistent with the literature report.14 Use of
pivalic acid is crucial for a successful regioselective arylation. The
reaction did not proceed or was extremely slow in the absence of
PivOH. The amount of PivOH also has a significant impact on the
reaction rate and regioselectivity. The reaction was significantly
slower with <0.2 equiv of PivOH while significantly faster with
increased amounts of PivOH (>1 equiv). However, reaction with
more than 1 equiv of PivOH also generated more undesired regioi-
somers. The optimal amount was identified as 0.3–0.5 equiv of
PivOH.
Formation of an aryl–aryl bond through direct C–H activation
has been intensively investigated and significant progress in this
area has been achieved.5–10 Direct arylation has also been applied
to aromatic heterocycles including furan derivatives.11 Control of
the regioselectivity in arylation of substituted furans remains a
challenge due to the presence of several C–H bonds. GlaxoSmithK-
line previously reported the first palladium catalyzed (5 mol %
PdCl2, 10 mol % Cy3P, and KOAc and Bu4NBr in DMF at 110 °C) direct
arylation at the C-5 position of furfural with aryl iodides or aryl bro-
mides.12 Doucet described low catalyst loading ligand-free palla-
dium-catalyzed (0.1 mol % Pd(OAc)2, KOAc, DMAc, 150 °C) direct
C-5 arylation of furfural with aryl bromides.13 Fagnou has devel-
oped a protocol for aryl–aryl bond formation via C–H activation
with palladium catalyst (2 mol % Pd(OAc)2, 4 mol % Cy3PÁBHF4 and
additive pivalic acid) and applied the protocol to direct C-5 arylation
of furfural with aryl bromides.14 Fagnou has also reported direct C-5
arylation of furfural using aryl bromides and Pearlman’s catalyst.15
Copper-catalyzed direct arylation of furfural at the C-5 position with
aryl bromides has also been reported.16
In the first instance it was necessary to determine what was the
optimal substitution on both furan (7) and quinazoline (8) which
would best facilitate the directed arylation. Doucet’s catalyst con-
ditions were used due to their simplicity and after extensive
screening it was found that furfural (2) and bromo-quinazoline
Cl
Cl
Cl
O
O
O
OHC
HN
N
HN
HN
N
O
2
O
0.1-0.5 mol%Pd(OAc)2,
F
F
N
N
F
N
KOAc, 140 ºC 12h
50% conversion
OHC
N
+
12 (10%)
Cl
10 (8.9%)
11 (14.5%)
+
5
Cl
Cl
O
(54%)
O
O
N
N
HN
N
HN
N
NH
N
Cl
Br
F
O
N
F
F
F
N
N
NH
HN
N
9
O
F
OHC
N
O
13 (6%)
Cl
14 (1.6%)
Scheme 3. Direct arylation of furfural with 9 under ligandless/low catalyst loading conditions.

G. Erickson et al. / Tetrahedron Letters 55 (2014) 6007–6010
6009
PCy2
PCy2
i-Pr
P
P
P
P
i-Pr
P
i-Pr
Xphos
Cy-John-Phos
DPPE
DPPM
cataCXium A
P
P
O
.HBF4
P
P
P
P
(t-Bu)2NpP.HBF
.HBF4
4
Cp3P.HBF4
DPE-Phos
BDPP
Chart 1. Impact of Ligands on Direct Arylation of furfural with 9. (a) Conditions: 9 (1 equiv), furfural (27.7 equiv), Pd(OAc)2 (0.02 equiv), PivOH (0.5 equiv); ligand
(0.04 equiv); 115 °C.
Base also played a key role in the reaction. Significant by-prod-
ucts were observed with organic bases such as Et3N and Hunig’s
base and minimal conversion was observed with Li2CO3, Na2CO3,
and CaCO3. The reaction proceeded significantly faster with
K2CO3 than other bases. However, the chloride in 9 also reacted
with furfural once the bromide in 9 had been completely con-
sumed to form a bis-furfural coupling side product (15).17
while the reaction did not proceed at lower temperature (<90 °C).
The optimal temperature range was identified as 110–120 °C.
Other additives/co-catalysts such as Ag2CO3, Ag2O, CuBr, and
Cu(OAc)2 and phase transfer catalysts gave no improvement in
either the reaction rate or the regioselectivity.
Although the reaction with Me(t-Bu)2PÁHBF4 gave the highest
yield, 15 was also formed as a side product. Finally, the best reac-
tion conditions were identified as 2 mol % Pd(OAc)2, 4 mol % Cy3-
H
O
PÁBHF4, 0.5 equiv PivOH, and 2 equiv KOAc in
5 volume of
furfural at 110–120 °C for 4–7 h under inert atmosphere. The
desired product 5 was formed in around 75–80% reaction yield
under the optimized conditions. The work-up and isolation process
was straightforward. The reaction mixture was diluted with ace-
tone, cooled to room temperature, and filtered to remove potas-
sium salts. The product was crystallized by adding water to the
filtrate to give 5 in 63–65% isolated yield with >98% purity.18
The complete synthesis is then illustrated in Scheme 4. Conden-
sation of 2-amino-5-bromobenzoic acid (16) with formamide
under mild conditions afforded 6-bromo-4(1H)-quinazolinone
(17) in 94% yield. Conversion of 17 to 6-bromo-4-chloroquinazo-
line (18) by treatment with POCl3 followed by in situ displacement
with aniline (19) gave 9 in 93% yield. Regioselective arylation then
subsequent reductive amination with NaBH(OAc)3 followed by for-
mation of a tosylate salt afforded Lapatinib in 44% overall yield.
In summary, we have developed an elegant high yielding new
route for the synthesis of Lapatinib. The new route is 5-steps
shorter than the previously disclosed route, significantly more
atom efficient and eliminates a hazardous waste stream.
O
O
HN
OHC
N
F
O
N
15
It was very difficult to control the formation of 15 unless the reaction
was closely monitored near the completion and cooled down imme-
diately after the reaction was complete. Similarly, 15 was also formed
when KHCO3, Cs2CO3, K3PO4, K2HPO4, and CsOPiv were used. When
KOAc or CsOAc was used, the reaction was slower than K2CO3 as
the base and 15 was not formed even if the reaction mixture was
heated overnight after all bromide in 9 had been consumed.
The reaction temperature is also a very important factor for the
reaction. Higher temperature (>130 °C) resulted in the formation of
higher amounts of 15 (even when KOAc was used as the base)

6010
G. Erickson et al. / Tetrahedron Letters 55 (2014) 6007–6010
Formamide
Supplementary data
O
O
130 o
C
Br
Br
OH
NH2
NH
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2014.09.
039.
N
94%
16
17
POCl3
References and notes
Cl
Cl
N
O
1. Burris, H. A., III Oncologist 2004, 9, 10–15.
2. Lackey, K. E. Curr. Top. Med. Chem. 2006, 6, 435–460.
Br
+
3. Petrov, K. G.; Zhang, Y.-M.; Carter, M.; Cockerill, G. S.; Dickerson, S.; Gauthier, C.
A.; Guo, Y.; Mook, R. A., Jr.; Rusnak, D. W.; Walker, A. L.; Wood, E. R.; Lackey, K.
E. Bioorg. Med. Chem. Lett. 2006, 16, 4686–4691.
4. (a) McClure, M. S.; Osterhout, M. H.; Roschangar, F.; Sacchetti, M. J. PCT Int. WO
02/02552.; (b) Parry, P. R.; Bryce, M. R.; Tarbit, B. Org. Biomol. Chem. 2003, 1,
1447–1449; (c) Thames, S. F.; Odom, H. C. J. Heterocycl. Chem. 1966, 490.
5. Chen, X.; Engle, K. M.; Wang, D. H.; Yu, J. Q. Angew. Chem., Int. Ed. 2009, 48,
5094–5115.
H2N
N
F
19
18
93%
Cl
O
6. Alberico, D.; Scott, M. E.; Lautens, M. Chem. Rev. 2007, 107, 174–238.
7. Campeau, L. C.; Stuart, D. R.; Fagnou, F. Aldrichim. Acta 2007, 40, 35–41.
8. McGlacken, G. P.; Bateman, L. M. Chem. Soc. Rev. 2009, 38, 2447–2464.
9. Ackermann, L.; Vicente, R.; Kapdi, A. R. Angew. Chem., Int. Ed. 2009, 48, 9792–
9826.
10. Verrier, C.; Lassalas, P.; Théveau, L.; Quéguiner, G.; Trécourt, F.; Marsais, F.;
Hoarau, C. Beilstein J. Org. Chem. 2011, 7, 1584–1601.
11. Seregin, I. V.; Gevorgyan, V. Chem. Soc. Rev. 2007, 36, 1173.
12. McClure, M. S.; Glover, B.; McSorley, E.; Millar, A.; Osterhout, M. H.;
Roschangar, F. Org. Lett. 2001, 3, 1677–1680.
HN
Br
N
F
N
O
9
Pd (OAc)2, Cy3P.HBF4
KOAc, 115 oC
63%
H
ˇ
13. (a) Dong, J. J.; Roger, J.; Pozgan, F.; Doucet, H. Green Chem. 2009, 11, 1832–
O
1846; (b) Battace, A.; Lemhadri, M.; Zair, T.; Doucet, H.; Santelli, M.
Organometallics 2007, 26, 472–474.
Cl
14. (a) Lafrance, M.; Fagnou, K. J. Am. Chem. Soc. 2006, 128, 16496–16497; (b)
Lafrance, M.; Lapointe, D.; Fagnou, K. Tetrahedron 2008, 64, 6015–6020; (c)
Liégault, B.; Lapointe, D.; Caron, L.; Vlassova, A.; Fagnou, K. J. Org. Chem. 2009,
74, 1826–1834; (d) Schipper, D. J.; Fagnou, K. Chem. Mater. 2011, 23, 1594–
1600; (e) Gorelsky, S. I.; Lapointe, D.; Fagnou, K. J. Org. Chem. 2012, 77, 658–
668; (f) Tan, Y.; Hartwig, J. J. Am. Chem. Soc. 2011, 133, 3308–3311; (g)
Wakioka, M.; Nakamura, Y.; Wang, Q.; Ozawa, F. Organometallics 2012, 31,
4810–4826; (h) Lapointe, D.; Markiewicz, T.; Whipp, C. J.; Toderian, A.; Fagnou,
K. J. Org. Chem. 2011, 76, 749–759.
O
HN
N
OHC
O
N
F
5
15. Parisien, M.; Valette, D.; Fagnou, K. J. Org. Chem. 2005, 70, 7578–7584.
16. Zhao, D.; Wang, W.; Yang, F.; Lan, J.; Yang, L.; Gao, G.; You, J. Angew. Chem., Int.
Ed. 2009, 48, 3296–3300.
1)
6, iPr2NEt, NaBH(OAc)3
2) TsOH
80%
17. The structure of 15 was tentatively assigned by LC–MS/MS.
18. Typical procedure for the synthesis of 5 via direct regioselective arylation of
furfural. A reaction vessel was charged with 9 (50 g, 109.0 mmol), Cy3PÁHBF4
(1.606 g, 4.36 mmol), Pd(OAc)2 (0.489 g, 2.18 mmol), KOAc (21.4 g,
218.0 mmol), and PivOH (5.57 g, 54.5 mmol). The mixture was purged with
nitrogen. Furfural (250 ml) was added. The mixture was purged with nitrogen
for 10 min, then heated to 110–115 °C and stirred at 110–115 °C until the
reaction was complete. The reaction mixture was cooled to 70 °C and treated
with acetone (600 ml). The mixture was further cooled to 30 °C and filtered to
remove the undesired solids. The filtrate was heated to 55 °C. Water (250 ml)
was added at a rate maintaining the temperature above 45 °C. The mixture was
then cooled to 40 °C. The resultant slurry was stirred at 40 °C for 2 h and then
cooled to 30 °C over 1.5 h. The solids were filtered, washed twice with acetone/
water (1:1, 150 ml), and dried at 60 °C in a vacuum oven to give 5 as a yellow
crystalline solid (32.5 g, 63% yield). 1H NMR (500 MHz, DMSO-d6) d 5.26 (s, 2H),
7.18 (m, 1H), 7.28 (d, J = 8.8 Hz, 1H), 7.30–7.36 (m, 2H), 7.38 (d, J = 3.8 Hz, 1 H),
7.47 (m, 1H), 7.71 (m, 1H), 7.73 (d, J = 3.8 Hz, 1H), 7.83 (d, J = 8.8, 1H), 7.99 (d,
J = 2.5 Hz, 1H), 8.26 (dd, J = 8.8, 1.50 Hz, 1H), 8.58 (s, 1H), 8.93 (s, 1H), 9.67 (s,
1H), 10.06 (s, 1H); 13C NMR (125 MHz, DMSO-d6) d 69.4, 109.7, 114.0 (d,
J = 21.8 Hz), 114.2, 114.7 (d, J = 20.9 Hz), 115.2, 119.4, 121.1, 122.6, 123.3 (d,
J = 2.7 Hz), 124.5, 126.2, 128.7, 129.4, 130.5 (d, J = 8.2 Hz), 132.8, 139.6 (d,
J = 7.3 Hz), 149.9, 150.1, 152.0, 155.2, 157.6, 157.7, 162.2 (d, J = 244.3 Hz),
177.7. The analytical data are identical to those of an authentic sample.
Cl
O
HN
N
O
S
O
O
NH
N
F
HO3S
H2O
1
2
Lapatinib (ditosylate monohydrate)
Scheme 4. New route for the synthesis of Lapatinib.
Acknowledgments
We thank Ping Xiong, Byron Johnson Daniel Reynolds, and Anna
Vogt for analytical support.