Full text of DOI:10.1021/op100090j

Organic Process Research & Development 2010, 14, 1464–1468
Novel Preparation of H1 Receptor Antagonist Fexofenadine
Jun Huang,†,‡ Wen Wang,† and Li-Xin Wang*,†
Key Laboratory of Asymmetric Synthesis and Chirotechnology of Sichuan ProVince, Chengdu Institute of Organic Chemistry,
Chinese Academy of Sciences, Chengdu 610041, P.R. China, and Graduate School of Chinese Academy of Sciences,
Beijing 100039, P.R. China
Abstract:
for the same preparation while using 2-(4-bromophenyl)-2-
methyl-propanenitrile as the starting material, which provided
an easy separation of the final product since the nitrile
intermediate is easily recrystallized. Ethyl 2-methyl-2-p-tolyl-
propanoate,14 2-(4-bromophenyl)acetonitrile,15 and methyl 2-(4-
formylphenyl)-2-methylpropanoate16 have also been used as
starting material to avoid the Friedel-Crafts reaction. With the
exception of the above two strategies, biooxidation of the methyl
group of terfenadine is also an effective method to obtain
fexofenadine.3,17
A novel synthetic route for the preparation of H1 receptor
antagonist fexofenadine is described. The synthetic route started
from the para-substituted aromatic derivative of methyl 4-(cya-
nomethyl)benzoate, 2, and gave fexofenadine in 26.0% overall yield
via six steps. The whole process featured a method wherein
fexofenadine could be obtained in excellent quality without ortho-
or meta-unpurified regioisomers.
Summarily, Friedel-Crafts acetylation is a highly convenient
and simple strategy for the preparation of fexofenadine or related
intermediates; however, the process lacks selectivity and is of
limited use in commercial production.18 Therefore, it is still an
industrial request to obtain more qualified fexofenadine, espe-
cially without (or with less) ortho and meta impurities in a
commercially challenging, less costly, and more environmen-
tally friendly manufacturing and easily scaled up process.
As a part of our unceasing interest in active pharmaceutical
ingredients (API) and related intermediates,19,20 we report herein
a novel synthetic route for the preparation of fexofenadine. The
solely para-positioned raw material methyl 4-(cyanomethyl)-
benzoate, 2, is first devised as the starting material, and high-
quality fexofenadine was efficiently obtained in 26.0% overall
yield with convenient reactions and commercially available
chemicals. The whole scheme, strategies, and synthetic route
are depicted in Scheme 1.
Introduction
R,R-Dimethyl-4-[1-hydroxy-4-[4-(hydroxydiphenylmethyl)-
1-piperidinyl]butyl] benzeneacetic acid (fexofenadine, Figure
1) is a nonsedating-type H1 receptor antagonist and a useful
antihistaminic drug,1,2 which was launched by Hoechest Marion
Roussel in 1996. Compared with its analogue terfenadine,
fexofenadine has fewer systemic side effects because of its lower
permeability into central nervous system tissues.3,4
The synthesis of fexofenadine was first disclosed in U.S.
Patent 4,254,129,5 starting from ethyl R,R-(dimethylphenyl)ac-
etate and 4-chlorbutyroyl via Friedel-Crafts reaction, followed
by condensation with R,R-diphenylpiperidine methanol to form
a keto ester intermediate; subsequent reduction and hydrolysis
afforded fexofenadine. After that, many methods have been
developed to prepare this molecule.6 Schroeder7 reported a new
method using Friedel-Crafts reaction of methyl R,R-(dimet-
hyphenyl)acetate with succinic anhydride to introduce the para
substituent. Natalini8 developed a new route based on the
conversion of R-halo-alkylarylketone into the corresponding
substituted 2-arylalkanoic ester via a crucial Giordano’s pro-
cedure catalyzed by ZnBr2. Although improved Friedel-Crafts
reactions were also reported by Krauss,9 Wu,10 and Richard,11
Friedel-Crafts reactions unavoidably produced a mixture of
ortho, meta, and para regioisomers, which made the separation
and purification relatively difficult for stricter quality require-
ments. To circumvent the Friedel-Crafts reaction, another
strategy is using 1,4-disubstituted or para-positioned phenyl
derivatives. Kawai12 reported an alternative route starting from
commercially available 4-bromophenylacetic acid. In this
process, a costly Pd(0) catalyst and environmental unfriendly
mercury oxide were involved. Graziano13 reported another
analogous route starting from R,R-dimethyl-(4-bromophenyl)-
acetic acid methyl ester, and HgO was replaced by costly PtCl2
to avoid the byproduct. Milla2b reported an improved process
Results and Discussion
Our design strategy lies in the following facts: unwanted
isomers mainly come from the poorly selective Friedel-Crafts
reaction, and separation cost comes mainly from the separation
of the unwanted regioisomers. To overcome these disadvan-
tages, we devised a novel synthetic route and chose the
commercially available solely para-positioned phenyl derivative
2 as starting material, which is depicted in Scheme 1. Accord-
ingly, methyl benzoate derivative 2 was converted to methyl
4-(2-cyanopropan-2-yl)benzoate, 3, by methylation. The key
step in our process is the condensation of 3 with γ-butyrolac-
* Corresponding author. Telephone: (+86)-28-8525-5208. E-mail: wlxioc@
cioc.ac.cn.
† Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences.
‡ Graduate School of Chinese Academy of Sciences.
Figure 1. Structure of fexofenadine.
1464
•
Vol. 14, No. 6, 2010 / Organic Process Research & Development
10.1021/op100090j  2010 American Chemical Society
Published on Web 10/04/2010

Scheme 1. Novel preparation of fexofenadine
tone, 4, in the presence of a strong base and afforded the primary
alcohol 6a via the desired key intermediate 5 which was easily
purified by simple extraction and acid-base neutralization.
Primary alcohol 6a was easily converted to the key intermediate
6b which was condensed with azacyclonol 7 to give nitrile 8.
Finally, fexofenadine was obtained after reduction with NaBH4
and in situ hydrolysis in HCl.
Methyl 4-(2-cyanopropan-2-yl)benzoate 3 was easily pre-
pared from methyl benzoate derivative 2 with Me2SO4 in the
presence of a phase transfer catalyst, e.g. tetrabutylammonium
bromide, at low temperature (0-3 °C) in almost quantitative
yield (96%). The temperature of this conversion is crucial;
higher temperatures led to the hydrolysis of the ester in alkaline
conditions.
The key step of the whole process is the preparation of
ketone 6a. Ketone 6a and its structural analogues are the key
intermediates for the preparation of fexofenadine in the reported
methods which mainly focused on the construction of these
interesting structures. Among them, Friedel-Crafts reaction5,7,18
was one of the straightforward methods to manufacture these
key intermediates in spite of the unavoidable formation of ortho
and meta isomers. Except Friedel-Crafts reaction, coupling of
methyl 2-(4-bromophenyl)-2-methylpropanoate and 3-butyne-
1-ol in the presence of Pd(0)/Cu2Br2 and addition of 2-(1,3-
dioxolan-2-yl)ethyl magnesium bromide to 4-subsituted benzalde-
hyde14,16 were also used. In our strategy, a decarboxylative
condensation between methyl benzoate derivative 3 and γ-bu-
tyrolactone 4 was first employed to introduce the key intermedi-
ate 6a via salt 5 which is water-soluble and easily to be
separated and purified by simple extraction and acid-base
neutralization. Methyl benzoate derivative 3 was transformed
(1) (a) Slater, J. W.; Zechnich, A. D.; Haxby, D. G. Drugs 1999, 57,
31–47. (b) Markham, A.; Wagstaff, A. J. Drugs 1998, 55, 269–274.
(2) (a) Honig, P. K.; Woosley, R. L.; Zamani, K.; Conner, D. P.; Cantilena,
L. R. Clin. Pharmacol. Ther. 1992, 52, 231–238. (b) Zaragoza, F. J. M.
U.S. Patent 2003/0166682A1, 2003.
(3) Ouarradi, A. E.; Arnaud, I. S.; Buisson, D. Tetrahedron 2008, 64,
11738–11744.
(4) Monahan, B. P.; Ferguson, C. L.; Killeavy, E. S.; Lloyd, B. K.; Troy,
J.; Cantilena, L. R. J. Am. Med. Assoc. 1990, 264, 2788–2790.
(5) Carr, A. A.; Dolfini, J. E.; Ohio, C.; Wright, G. J. U.S. Patent
4,254,129, 1981.
(6) Reviews see: (a) Liu, B.; Rong, R.-B.; Yuan, H.; Yu, Z.-Y. Huaxue
Yu ShengwuGongcheng 2005, 2, 54–56. (b) Hu, C.-D.; Zhang, J.
Guangzhou Huagong. 2009, 37, 30–31.
(18) Selected examples: (a) Ambra, T. E. D. U.S. Patent 5,589,487, 1996.
(b) Ambra, T. E. D. U.S. Patent 5,663,412, 1997. (c) Ambra, T. E. D.
U.S. Patent 5,750,703, 1998. (d) Krauss, R. C.; Strom, R. M.;
Scortichini, C. L.; Kruper, W. J.; Wolf, R. A.; Wu, W. W.; Carr, A.
A.; Hay, D. A.; Rudisill, D. E.; Panzone, G. U.S. Patent 6,242,606B1,
2001. (e) Krauss, R. C.; Strom, R. M.; Scortichini, C. L.; Kruper,
W. J.; Wolf, R. A.; Wu, W. W.; Carr, A. A.; Hay, D. A.; Rudisill,
D. E.; Panzone, G. U.S. Patent 6,147,216, 2000. (f) Ambra, T. E. D.
U.S. Patent 2002/0007068A1, 2002. (g) Ma, L.; Gan, M.-Y.; Chen,
F.-Q.; Chen, C.; Lu, W. CN Patent 101182306A, 2008. (h) Ram-
achandra, R. D.; Narayanrao, K. R.; Gopaldas, G. M.; Ramdas, B. D.
WO Patent 2005/019175A1, 2005. (i) Ayers, T. A.; Brown, P. W.
U.S. Patent 2003/0022919 A1, 2003.
(7) Collin, S.; Ryan, H.; Richard, C. WO Patent 02/102776A1, 2002.
(8) Giacomo, B. D.; Coletta, D.; Natalini, B.; Ni, M.-H.; Pellicciari, R.
Farmaco 1999, 54, 600–610.
(9) Krauss, R. C.; Strom, R. M.; Scortichini, C. L.; Kruper, W. J.; Wolf,
R. A.; Wu, W. W.; Carr, A. A.; Hay, D. A.; Rudisill, D. E.; Panzone,
G. U.S. Patent 6,348,597, 2001.
(10) Lv, B. H.; Yang, X. Y.; Wu, F. H. Chin. J. Med. Chem. 2004, 14, 96.
(11) Richard, K. C. H.; Michelea, K. WO Patent 93/21156 A1, 1993.
(12) (a) Kawai, S. H.; Hambalek, R. J.; Just, G. J. Org. Chem. 1994, 59,
2620–2622. (b) Ambra, T. E. D.; Pilling, G. M. WO Patent 97/
23213A1, 1997.
(13) Graziano, C.; Giuseppe, B.; Domenico, M. EP Patent 1260505A1,
2002.
(19) (a) Wang, L.-X.; Shen, J.-F.; Tang, Y.; Chen, Y.; Wang, W.; Cai,
Z.-G.; Du, Z.-J. Org Process Res. DeV. 2007, 11, 487–489. (b) Li,
Y.; Xu, X.-Y.; Pan, S.-Y.; Wang, L.-X.; Xia, C.-Q. Chin. J. Chem.
2009, 27, 433–436.
(14) Fang, Q. K.; Senanayake, C. H.; Wilkinson, H. S.; Wald, S. A.; Li,
H. Tetrahedron Lett. 1998, 39, 2701–2704. Sun, Y.; Tang, L. H.; Tao,
L.; Qin, X. H. Chin. J. Pharm. 2006, 37, 439.
(20) (a) Wang, L.-X.; Tang, Y.; Chen, Y.; Tian, F. EP Patent 1942110A1,
2008. (b) Wang, L.-X.; Tang, Y.; Chen, Y.; Tian, F. WO Patent
2007028337 A1, 2007. (c) Wang, L.-X.; Tang, Y.; Chen, Y.; Tian, F.
U.S. Patent 2008249311 A1, 2008. (d) Wang, L.-X.; Tang, Y.; Chen,
Y.; Tian, F. CN Patent ZL200510060721.2, 2007; (e) Wang, L.-X.;
Tang, Y.; Chen, Y.; Tian, F. CN Patent ZL200510061230.X, 2007;
(f)Wang,L.-X.;Tang,Y.;Chen,Y.;Tian,F.CNPatentZL200510061231.4,
2007; (g) Wang, L.-X.; Peng, H. Q.; Du, Z. J.; Lan, H. F. Chinese
Patent CN 200610027596.X, 2006.
(15) Patel, S.; Waykole, L.; Repic, O.; Chen, K.-M. Synth. Commun. 1996,
26, 4699–4710.
(16) (a) Senanayake, C. H.; Fang, Q. K.; Wilkinson, S. H. U.S. Patent
6,147,217, 2000. (b) Miroslav, V.; Andrej, B.; Miroslav, K.; Ivan, V.
WO Patent 2004/043922 A1, 2004.;(c) Venkateswaran, S. C.; Singh,
J. P.; Lalit, W. WO Patent 2007/135693A2, 2007.
(17) Mazier, C.; Jaouen, M.; Sari, M.-A.; Buisson, D. Bioorg. Med. Chem.
Lett. 2004, 14, 5423.
Vol. 14, No. 6, 2010 / Organic Process Research & Development
•
1465

Table 1. Condensation of 3 with 4 in various basesa
entry
base
solvent
Tb [°C]
yieldc [%]
1
2
3
4
5
6
NaH
benzene
80
80
45
32
35
55
63
61
Na
benzene
NaNH2
NaH
NaH
NaH
benzene
toulene
80
110
110
80
toluene/DMF ) 10:1
toluene/DMEd ) 10:1
a Experimental conditions: mixture of 3 (0.1 mol), 0.125 mol base, and 0.12 mol γ-butyrolactone 4 in solvent was mechanically stirred at the designed temperature.
b Reaction temperature of 3 with 4. c Isolated yield of 6a. d DME ) 1,2-dimethoxyethane.
Table 2. Coupling of 6b-d with azacyclonol 7 in the presence of various bases and solvents
entry
X
solvent
base
T [°C]
yielda 8 [%]
yielda 8a [%]
1
2
3
4
5
6
7
Br (6b)
Br (6b)
Br (6b)
Br (6b)
Br (6b)
Cl (6c)
OTs (6d)
THF
Na2CO3
NaHCO3
Et3N
60
60
25
25
25
60
60
0
5
32
63
55
48
55
83
74
31
5
THF
DMF
THF
THF
acetone
acetone
Et3N
DIPA
NaHCO3
Et3N
5
12
6
a Isolated yield.
to the salt 5 in the presence of base followed by decarboxylation
to give the desired intermediate 6a. Initially, sodium, sodium
amide, and NaH were investigated, and NaH was recommended
in terms of the yield of 6a (Table 1). The reaction performed
in toluene21 may give a better yield. In this process, we found
that the solubility of the formed target sodium 5 in toluene was
not ideal, and 1,2-dimethoxyethane (DME) was added as
cosolvent; the solubility and stirring state of the reactants were
improved, and 61% yield was obtained.22 Therefore, in larger-
scale processing, the cosolvent of toluene and DME was used.
After the condensation was completed, some water was added
to dissolve the product 5, and the organic phase which contained
the unreacted 3 and byproduct (mainly 4-(2-cyanopropan-2-
yl)benzoic acid) may be easily separated and reused by simple
extraction with toluene. The aqueous phase was adjusted to
weak basicity with 1 N HCl, heated to 60 °C, and hydrolyzed
to the desired intermediate 6a which could be used directly in
the next step. During the process of hydrolysis, the pH value
should be controlled below 10 to avoid hydrolysis of nitrile.
The crude 6a could easily transform to bromide 6b, chloride
6c, or sulfonic ester 6d.
Bromide 6b was coupled with azacyclonol 7 in the presence
of base and suitable solvent. The base is crucial for the coupling
reaction to avoid the undesired cyclization byproduct 8a, and
the ratio of 8 to 8a greatly changes with different bases. Several
bases have been studied and the results are shown in Table 2.
Coupling product 8 was mainly formed (entries 4-5), when
organic bases such as Et3N or diisopropylamine (DIPA) used.
By contrast, inorganic bases such as Na2CO3 and NaHCO3
mainly afforded unwanted cyclization 8a (entries 1, 2), and
during a larger scale-up processing, large volume of CO2
foamed fiercely. Solvent also greatly affects the formation of 8
and 8a when Et3N used as base. Product 8 was dominated in
THF (entry 4), whereas 8 was almost equal to 8a in DMF (entry
3). Chloride 6c and sulfonic esters 6d could also react with
azacyclonol 7 while giving relatively lower yields (entries 6-7).
The recommended scale-up reaction conditions were found to
be Et3N as base, THF as solvent, and 6b as substrate at room
temperature.
(21) The original process used benzene as solvent.
(22) DMF was also added as auxiliary solvent, the solubility and stirring
state of the reactants were adequately improved, and the yield was
increased up to 63%. The amount of DMF used may be adjusted as
needed, and this process was scaled up to a 20-kg scale without
observation of safety hazards. Before further industrial testing, we were
advised by the OPRD reviewers to recheck related cases to find that
DMF/NaH is a known process safety hazard (http://www.crhf.org.uk/
incident101.html). We immediately informed our partner to re-evaluate
this step and cease the scale-up attempt.
Finally, excellent quality of target molecule fexofenadine 1
without unwanted regioisomers may be easily obtained after
1466
•
Vol. 14, No. 6, 2010 / Organic Process Research & Development

one-pot-wise hydrolysis and reduction in 81% yield as is well
documented.2b
2H, J ) 12 Hz, Ar), 8.11 (d, 2H, J ) 12 Hz, Ar). 13C NMR
(75 MHz, CDCl3): δ ) 25.6, 28.6, 37.2, 47.9, 67.64, 125.2,
125.5, 130.0, 130.6, 147.2, 172.7, 192.2. IR (cm-1): 3062, 2990,
2237, 1749, 1679, 1604, 1570, 1481, 1452, 1409, 1370,
1016,950, 857, 812, 683. HR-MS (ESI) Calcd for
C15H15NNaO3: 280.0944. Found: 280.0941.
Conclusions
In conclusion, we have devised and reported a novel
synthetic route for the preparation of fexofenadine of excellent
quality. The solely para-positioned fexofenadine and its salt were
obtained in excellent quality and at acceptable raw material cost
without ortho or meta unpurified regioisomers. The whole
process and all of the procedures are now under processing
evaluation in a 20-kg scale reaction in China Huahai Pharma-
ceuticals Co., Ltd.
2-(4-(4-Hydroxybutanoyl)phenyl)-2-methylpropaneni-
1
trile (6a).2b Isolated yield: 63% (two steps). H NMR (300
MHz, CDCl3, TMS): δ ) 1.73 (s, 6H, CH3), 1.96-2.06 (m,
2H, CH2), 3.13 (t, 2H, J ) 6 Hz, COCH2), 3.75 (t, 2H, J ) 6
Hz, OCH2), 7.58 (d, 2H, J ) 6 Hz, Ar), 8.01 (d, 2H, J ) 6 Hz,
Ar).
Procedure for 6b,c. The solution of the crude 6a (160 g)
in 1.5 L of CH2Cl2 and 40% HBr or 36% HCl (500 mL) was
mechanically stirred at 25 °C for 8 h. The mixture was diluted
with 2.5 L of CH2Cl2 and 2.5 L of water. The organic phase
was separated and washed with brine (2 × 1 L). After removal
of the solvent, the residue was recrystallized in ethanol to
provide white solid 6b,c.
2-(4-(4-Bromobutanoyl)phenyl)-2-methylpropanenitrile
(6b)2b (X ) Br). Yield: (160 g), 53% for three steps from 3.
1H NMR (300 MHz, CDCl3, TMS): δ ) 1.75 (s, 6H, CH3),
2.27-2.35 (m, 2H, CH2), 3.18 (t, 2H, J ) 6 Hz, COCH2), 3.45
(t, 2H, J ) 6 Hz, BrCH2), 7.59 (d, 2H, J ) 12 Hz, Ar), 8.00
(d, 2H, J ) 12 Hz, Ar).
2-(4-(4-Chlorobutanoyl)phenyl)-2-methylpropanenitrile
(6c)8 (X ) Cl). Yield: (130 g), 55% for three steps from 3. 1H
NMR (300 MHz, CDCl3, TMS): δ ) 1.75 (s, 6H, CH3),
2.21-2.27 (m, 2H, CH2), 3.18 (t, 2H, J ) 6 Hz, COCH2), 3.68
(t, 2H, J ) 9 Hz, ClCH2), 7.59 (d, 2H, J ) 6 Hz, Ar), 8.01 (d,
2H, J ) 6 Hz, Ar).
4-(4-(2-Cyanopropan-2-yl)phenyl)-4-oxobutyl-4-methyl-
benzenesulfonate (6d) (X ) OTs). 4-Methylbenzene-1-sulfo-
nyl chloride (260 g, 1.37 mol) in 1.5 L of CH2Cl2 was added
dropwise to the solution of the crude 6a (160 g) and Et3N (350
g, 3.46 mol) in 1.5 L of CH2Cl2. The mixture was mechanically
stirred at 25 °C for 8 h, diluted with 2.5 L of CH2Cl2 and 2.5
L of water. The organic phase was separated and washed with
brine (2 × 1 L). After removal of the solvent, the residue was
recrystallized in ethanol to provide white solid 6d (205 g). Yield:
54% for three steps from 3. 1H NMR (300 MHz, CDCl3, TMS):
δ ) 1.75 (s, 6H, CH3), 2.08-2.14 (m, 2H, CH2), 2.40 (s, 3H,
CH3), 3.05 (t, 2H, J ) 6 Hz, COCH2), 4.15 (t, 2H, J ) 6 Hz,
ClCH2), 7.30 (d, 2H, J ) 6 Hz, Ar), 7.57 (d, 2H, J ) 6 Hz,
Ar), 7.76 (d, 2H, J ) 6 Hz, Ar), 7.92 (d, 2H, J ) 6 Hz, Ar).
2-(4-(4-(4-(Hydroxydiphenylmethyl)piperidin-1-yl)bu-
tanoyl)phenyl)-2-methylpropanenitrile (8).7 Diphenyl(piperi-
din-4-yl)methanol 7 (340 g, 1.2 mol) and triethylamine (120 g,
1.2 mol) were added to a solution of 6b (290 g, 1 mol) in 1 L
of THF. The mixture was mechanically stirred at 25 °C for
8 h. After filtration of excessive diphenyl(piperidin-4-yl)metha-
nol and evaporation of the solvent, the crude solid was
recrystallized in ethyl acetate/petroleum ether (v/v, 1:2) to
provide white solid 8 (300 g, 0.63 mol). Yield: 63%. 1H NMR
(300 MHz, CDCl3, TMS): δ ) 1.49-1.52 (4H, m, CH2), 1.75
(6H, s, CH3), 1.99-2.04 (4H, m, CH2, CH), 2.45-2.50 (4H,
m, NCH2), 2.99-3.07 (4H, m, NCH2, CH2CO), 7.15-7.60 and
7.96-8.00 (14H, m, Ar).
Experimental Section
Unless otherwise noted, reagents were purchased from
commercial suppliers and used without further purification. 1H
and 13C NMR spectra were recorded on a Bruker-300 (300/75
MHz) spectrometer using CDCl3, DMSO-d6 as solvent, and
TMS as internal standard. The IR spectra were performed on a
NICOLET MX-1E FT-IR instrument. ESI mass spectra were
performed on a Finnigan LCQDECA, and high-resolution MS
spectral data were recorded on a Bruker Daltonics Bio TOF.
Methyl 4-(2-cyanopropan-2-yl)benzoate (3).19 NaOH (3
kg, 75 mol) was dissolved in 10 L of water; after it cooled to
room temperature, 20 g of PTC (tetrabutylammonium bromide)
and methyl 4-(cyanomethyl)benzoate 2 (1.05 kg, 6 mol) were
then added in a suitable reactor. After further cooling to 0-3
°C, dimethyl sulfate (2.65 kg, 21 mol) was added dropwise at
0-3 °C. The mixture was adequately stirred for 4 h, water (5
L) and CH2Cl2 (3 × 2 L) were then added. The organic phase
was separated and washed with saturated aq solution of NH4Cl
and concentrated under reduced pressure. Colorless solid 3 (1.17
kg, 5.76 mol) was obtained. Yield: 96%. 1H NMR (300 MHz,
CDCl3, TMS): δ ) 1.70 (s, 6H, CH3), 3.88 (s, 3H, CH3), 7.52
(d, 2H, J ) 6 Hz, Ar), 8.01 (d, 2H, J ) 6 Hz, Ar). 13C NMR
(75 MHz, CDCl3) δ ) 28.8, 37.2, 52.0, 123.7, 125.1, 129.6,
130.0, 146.1, 166.1.
2-(4-(4-Hydroxybutanoyl)phenyl)-2-methylpropaneni-
trile (6a). To a stirred solution of methyl 4-(2-cyanopropan-
2-yl) benzoate 3 (200 g, 1 mol) and sodium hydride (30 g, 1.25
mol, 60% mineral oil dispersion) in 1 L solvent (Vtoluene/DME
)
10:1), was added dropwise to the solution of dihydrofuran-
2(3H)-one 4 (104 g, 1.2 mol) in 500 mL of toluene for 1 h at
80 °C. The mixture was refluxed for a further 4 h. Then 2 ×
500 mL water was added to extract salt 5. After phase
separation, the aqueous solution was extracted with toluene, and
the organic phase was combined and recycled. The aqueous
solution was adjusted to pH ) 9-10 with 1 N HCl, and then
hydrolyzed at 60 °C for 4 h and extracted with CH2Cl2 (2 ×
500 mL). The combined organic layer was washed with
saturated aq solution of NH4Cl (500 mL) and concentrated under
reduced pressure. After removal of the solvent, the crude oil
liquid product 6a (165 g, HPLC: 79.8%) was obtained and used
for the next step without further purification.
2-Methyl-2-(4-(2-oxo-tetrahydrofuran-3-carbonyl)phe-
nyl)propanenitrile (Free Ketone 5). Colorless solid. 1H NMR
(300 MHz, CDCl3, TMS): δ ) 1.75 (s, 6H, CH3), 2.54-2.93
(m, 2H, CH2), 4.45-4.59 (m, 3H, COCHCO, OCH2), 7.62 (d,
Vol. 14, No. 6, 2010 / Organic Process Research & Development
•
1467

2-(4-(Cyclopropanecarbonyl)phenyl)-2-methylpropaneni-
trile (8a).2b 1H NMR (300 MHz, CDCl3, TMS): δ ) 1.04-1.10
(m, 2H, CH2), 1.23-1.28 (m, 2H, CH2), 2.64 (s, 6H, CH.3),
2.64-2.69 (m, 1H, COCH), 7.59 (d, 2H, J ) 6 Hz, ArH), 8.04
(d, 2H, J ) 6 Hz, ArH).
HPLC: 99.7%, single impurity <0.1% and total impurities
<0.3%). Yield: 81%. H NMR (300 MHz, DMSO): δ )
1
1.35-1.39 (m, 2H), 1.45 (s, 6H, 2 × CH3), 1.65-1.77 (m, 6H,
CH2), 2.80-2.88 (m, 5H, CH2N, CH), 3.28-3.32 (m, 2H,
CH2N), 4.52 (1H, m, CHOH), 5.62 (s, 1H, OH), 7.12-7.17
(m, 2H, Ar), 7.25-7.31 (m, 8H, Ar), 7.52 (d, 4H, J ) 7.64
Hz, Ar). 13C NMR (75 MHz, CDCl3): δ ) 20.36, 23.73, 26.44,
35.85, 41.16, 45.48, 45.48, 51.59, 55.77, 71.35, 78.22, 125.20,
125.64, 125.68, 126.05, 127.89, 143.50, 144.04, 146.66, 177.56.
r,r,-Dimethyl-4-[1-hydroxy-4-[4-(hydroxydiphenylmethyl)-
1-piperidinyl]butyl]benzeneacetic Acid (1), Fexofenadine.8
NaBH4 (140 g, 3.7 mol) and KOH (2.8 kg, 50 mol) were added
to a solution of 8 (480 g, 1 mol) in solvent (5 L of ethanol: 500
mL of water). The resulting mixture was mechanically stirred
at 20 °C for 1 h and refluxed for 12 h. After the mixture was
adjusted to pH ) 5-6 by 2 N hydrochloric acid, the solvent
was removed in vacuum. The residue was then diluted with
water (5 L) and extracted with ethyl acetate (3 × 3 L). The
combined organic phase was washed with brine (5 L) and dried
over anhydrous Na2SO4. Evaporation of the solvent afforded
crude 1. The crude solid was recrystallized from the mixture
of water and ethanol (v/v, 2:1) to give white solid 1 (410 g,
Acknowledgment
This project was financially supported by Huahai Pharma-
ceuticals Co., Ltd., Zhejiang, China.
Received for review March 31, 2010.
OP100090J
1468
•
Vol. 14, No. 6, 2010 / Organic Process Research & Development