Selective nosylation of 1-phenylpropane-1,3-diol and perchloric acid mediated Friedel-Crafts alkylation: Key steps for the new and straightforward synthesis of tolterodine

Article abstract of DOI:10.1055/s-2008-1067084

We have developed a new and straightforward synthesis of racemic tolterodine [N,N-diisopropyl-3-(2-hydroxy-5-methyl-phenyl)-3-phenylpropanamine]. The synthesis involves selective nosylation on the primary alcohol of 1-phenylpropane-1,3-diol using 4-nitrob

Full text of DOI:10.1055/s-2008-1067084

PAPER
1841
Selective Nosylation of 1-Phenylpropane-1,3-diol and Perchloric Acid
Mediated Friedel–Crafts Alkylation: Key Steps for the New and
Straightforward Synthesis of Tolterodine
Synthesis of
T
olter
a
odine thlia A. De Castro, Hakjune Rhee*
Department of Chemistry and Applied Chemistry, Hanyang University, Ansan Sa-3-Dong 1271, Kyunggi-Do 426-791, Korea
Fax +82(31)4073863; E-mail: hrhee@hanyang.ac.kr
Received 6 February 2008; revised 13 March 2008
impelled us to study thoroughly the structure of toltero-
dine. We considered the retrosynthetic analysis shown in
Scheme 1. We reported recently that 1-phenylpropane-
1,3-diol could be produced easily from the reduction of
Abstract: We have developed a new and straightforward synthesis
of racemic tolterodine [N,N-diisopropyl-3-(2-hydroxy-5-methyl-
phenyl)-3-phenylpropanamine]. The synthesis involves selective
nosylation on the primary alcohol of 1-phenylpropane-1,3-diol us-
ing 4-nitrobenzenesulfonyl chloride, subsequent diisopropylamine ethyl benzoylacetate (3) in excellent yield.4 Thus, we
substitution, and Friedel–Crafts alkylation using aqueous perchloric
thought that converting the hydroxy functionality into a
acid.
better leaving group would easily facilitate diisopropyl-
Key words: drugs, diol, selective sulfonation, nucleophilic substi-
tution, Friedel–Crafts alkylation
amine substitution and subsequent Friedel–Crafts alkyla-
tion to form tolterodine in a straightforward manner.
Tolterodine is the drug of choice for most patients for the
treatment of urinary urge incontinence resulting from an
overactive bladder as it has fewer side effects.1 Through
the years there have been myriad ways published for its
racemic2 and asymmetric synthesis.3 Mostly these meth-
ods made used of coumarin or hydrocoumarin derivatives
as starting material or intermediates. Most of these meth-
ods employed protection and deprotection techniques.
This makes the routes lengthy, hence requiring more re-
agents and solvents. In addition, some of these routes en-
tail longer reaction times and an inert environment while
others utilize harsh conditions and hazardous chemicals
such as diisobutylaluminum hydride, lithium aluminum
hydride, methyl iodide, among others. There are also
shorter routes, however, they gave lower yields or enan-
tioselectivities whereas others require tedious workup.
For asymmetric synthesis, the use of chiral auxiliaries3a,c
or expensive metal catalysts such as rhodium2b,3b,d is im-
practical due to the cost incurred. Recently, we published
a new route for the synthesis of tolterodine that also em-
ployed a protection–deprotection technique via a two-
phase reaction.4 This route is simple and short, however,
we have found a new route that is even simpler and elim-
inates the use of protecting groups. It is straightforward
and has very short reaction times making it advantageous
over other known procedures. We believed that this new
method is economical and strategically feasible for indus-
trial application. Thus, we wish to report here our new ap-
proach towards the synthesis of rac-tolterodine.
OH
LG'
N
N
1
rac-tolterodine
O
O
LG'
OEt
LG
3
2
Scheme 1 Retrosynthetic analysis of rac-tolterodine
We first considered chlorination of 1-phenylpropane-1,3-
diol using 2,4,6-trichloro-1,3,5-triazine and N,N-
dimethylformamide5 followed by Friedel–Crafts alkyla-
tion using graphite6 in chlorobenzene. The reactions pro-
ceeded quite well. Chloride is a better leaving group than
the hydroxy group thus we expected diisopropylamine
substitution to be smooth, however, we failed to obtain
our desired product. This scenario prompted us to consid-
er selective functionalization of the primary alcohol to
make it a better leaving group (Scheme 2). Commonly,
selective sulfonation of a diol is carried out using 4-tolu-
enesulfonyl chloride with various catalysts.7 Thus, we
first tried the sulfonation of diol 4 with 4-toluenesulfonyl
chloride. The yield with or without catalyst is relatively
the same and both conditions require long reaction times
so, we preferred doing it without the catalyst. The long re-
action times prompted us to consider another reagent for
Our aim was to develop a method of synthesizing toltero-
dine without using the protection–deprotection strategy
SYNTHESIS 2008, No. 12, pp 1841–1844
x
x.
x
x
.
2
0
0
8
Advanced online publication: 16.05.2008
DOI: 10.1055/s-2008-1067084; Art ID: F03308SS
© Georg Thieme Verlag Stuttgart · New York

1842
K. A. De Castro, H. Rhee
PAPER
functionalization. There have been reports on the consid- ification. Nucleophilic reaction occurs rapidly due to the
erable higher reactivity of 4-nitrobenzenesulfonate (nosy- higher reactivity of nosylate as leaving group as afore-
late) as a leaving group, which results in improved mentioned. This scenario prompted us to modify our strat-
subsequent nucleophilic displacement.8 We have proven egy. Since we identified that the problem was the
these ourselves in our previous study.4 Thus we tried it for nucleophilicity of the phenolic hydroxy group under basic
selective primary alcohol functionalization as well. As far conditions and the greater ability of nosylate to leave, we
as we know this is the first attempt at using 4-nitroben- decided to perform the diisopropylamine substitution
zenesulfonyl chloride (NsCl) for the selective sulfonation prior to the Friedel–Crafts alkylation (Scheme 3, path B).
of primary alcohols, such as diol 4. The reaction proceed- Substitution proceeded efficiently in three hours giving
ed smoothly in one hour giving 5a in 85% yield.9
compound 7 in 86% yield. This substitution reaction is
very quick compare to the known procedures.2,3 This is at-
tributed to the leaving group ability of the nosylate moi-
ety. The use of 4-nitrobenzenesulfonyl chloride is better
than 4-toluenesulfonyl chloride because the former re-
quires much shorter reaction times (3 h) than the latter (52
h). Furthermore, the used of tosylated alcohol 5b gave
only the diisopropylamine substitution product 7 in only
65% yield. This further supports the reactivity of nosylate.
O
O
OH
NaBH4, MeOH
OEt
OH
15 min, r.t.
98%
3
4
RCl, Et3N
CH2Cl2
The last step was the Friedel–Crafts alkylation. Initially
we tried our procedure using iron(III) chloride hexahy-
drate, but unfortunately the reaction was unsuccessful and
simply gave the starting material, compound 7. The pres-
ence of two electron-donating groups, both capable of co-
ordinating with the metal, suppressed the reaction. We
thought that increasing the amount of iron(III) chloride
hexahydrate to more than one equivalent would work, but
it did not. We tried another stronger acid catalyst, 12-
tungstophosphoric acid11 (H3PW12O40); however, it also
gave the starting material, compound 7. This strengthens
our speculation that a metal catalyst will not work in a
compound having two electron-donating groups. There
has been a study using the superacid trifluoromethane-
sulfonic acid to ionize initially the alcohol in amino alco-
hols to generate dicationic intermediates.12 However, this
acid is very expensive. Thus, we considered using a strong
inorganic acid, like aqueous perchloric acid, which is rel-
atively stronger than sulfuric acid. In addition, there has
been reports in the literature on the effectivity of perchlo-
OH
OR
5a: R = Ns, 1 h, 0–5 °C, 85%
5b: R = Ts, 24 h, r.t., 79%
Scheme 2
Next, we tried Friedel–Crafts alkylation using iron(III)
chloride hexahydrate10 to facilitate C–C bond formation
giving compound 6 in a short period (Scheme 3, path A).
Unfortunately, we were unsuccessful in obtaining rac-
tolterodine from diisopropylamine substitution even for a
prolonged period. Based on our 1H NMR analysis, it gave
a cyclic product instead. We presumed that under basic
conditions the phenolic hydroxy group was readily depro-
tonated and acted as nucleophile to the carbon to which
the nosylate group was attached, undergoing cyclic ether-
path A
OH
.
OH
p-cresol, FeCl3 6H2O
CH2Cl2
ONs
ONs
reflux, 30 min
58%
5a
6
HN(i-Pr)2
reflux, 3 h
86%
(i-Pr)2NH
reflux
path B
p-cresol
70% aq HClO4
OH
OH
.
CH2Cl2
N
N
r.t., 1 h
87%
7
rac-tolterodine
Scheme 3
Synthesis 2008, No. 12, 1841–1844 © Thieme Stuttgart · New York

PAPER
Synthesis of Tolterodine
1843
3-Hydroxy-3-phenylpropyl 4-Nitrobenzenesulfonate (5a)
ric acid in dichloromethane at room temperature for orga-
nocatalytic Friedel–Crafts alkylations.13 Using this acid in
excess protonates the amine moiety, which removes its
nucleophilicity while the benzylic alcohol group readily
reacts with p-cresol to give the coupling product, rac-
tolterodine. This reaction proceeded cleanly and quickly
at room temperature. All reactions described used simple
aqueous workup and purification methods to give the de-
sired products.
1-Phenylpropane-1,3-diol (4, 0.6215 g, 4.084 mmol) was dissolved
in CH2Cl2 (8.2 mL). NsCl (0.9503 g, 4.288 mmol) was slowly added
using a solid dropping funnel while the mixture was stirred in an ice
bath (0–5 °C). Et3N (0.85 mL, 6.126 mmol) was finally added and
the mixture was continuously stirred at 0–5 °C for 1 h. The mixture
was immediately filtered through a thin pad of silica gel (EtOAc).
The mixture was concentrated using a rotary evaporator. It was fur-
ther purified using column chromatography (silica gel, hexane–
EtOAc, 2:1) to give 5a (1.172 g, 85%) as a white solid; mp 83–85
°C.
In conclusion, we have developed a new route for toltero-
dine without using the common protection–deprotection
strategy. The used of 4-nitrobenzenesulfonyl chloride for
selective functionalization of the primary alcohol and
aqueous perchloric acid for Friedel–Crafts alkylation are
new ideas. These enable us to obtain rac-tolterodine in a
straightforward manner. The chiral resolution of rac-
tolterodine to (R)-tolterodine using L-tartaric acid has
been achieved using our previous reported method.4 In ad-
dition, all our steps require simple, inexpensive, and
readily available reagents as well as simple purification
methods giving an overall yield of rac-tolterodine of 62%.
IR (KBr): 3559, 1534, 1357, 1323, 1179, 1107, 1073 cm–1.
1H NMR (300 MHz, CDCl3): d = 2.04 (m, 2 H), 2.48 (s, 1 H), 4.14
(m, 1 H), 4.35 (m, 1 H), 4.76 (t, J = 6.2 Hz, 1 H), 7.28 (m, 5 H), 8.06
(d, J = 8.7 Hz, 2 H), 8.36 (d, J = 9.0 Hz, 2 H).
13C NMR (75 MHz, CDCl3): d = 38.0, 69.2, 70.3, 124.8, 125.9,
128.3, 128.9, 129.5, 141.7, 143.5, 151.0.
HRMS (EI): m/z [M]+ calcd for C15H15NO6S: 337.0620; found:
337.0620.
3-Hydroxy-3-phenylpropyl 4-Toluenesulfonate (5b)7c
1-Phenylpropane-1,3-diol (4, 1.439 g, 9.458 mmol) was dissolved
in CH2Cl2 (18.9 mL). TsCl (1.803 g, 9.458 mmol) was slowly added
We believed that this route is a better alternative for while the mixture was stirred at r.t. Et3N (0.85 mL, 6.126 mmol)
was finally added and the mixture was continuously stirred for 24 h.
tolterodine synthesis especially from an economic stand-
The mixture was washed with H2O (3 × 30 mL). The combined or-
point. The reduction of a b-keto ester with sodium boro-
ganic phase was dried (anhyd MgSO4), filtered, and concentrated
hydride in methanol solvent and the general application of
using a rotary evaporator. The mixture was further purified using
column chromatography (silica gel, hexane–EtOAc, 3:1) to give 5b
(2.280 g, 79%) as a pale yellow oil.
Friedel–Crafts alkylation using perchloric acid are under
investigation.
1H NMR (300 MHz, CDCl3): d = 1.95 (s, 1 H), 2.03 (m, 2 H), 2.46
(s, 3 H), 4.06 (m, 1 H), 4.29 (m, 1 H), 4.81 (t, J = 6.7 Hz, 1 H), 7.31
(m, 7 H), 7.80 (d, J = 8.5 Hz, 2 H).
1H and 13C NMR spectra were obtained using a Varian 300 spec-
trometer (300 and 75 MHz respectively) with TMS as internal stan-
dard. IR spectra were recorded on a Bio-Rad FTS 6000 FT-IR
spectrophotometer as KBr pellets. Uncorrected melting points were
determined with a Gallenkamp melting point apparatus. HRMS
were obtained on a JMS 700 spectrometer. Analytical TLC was
conducted on E. Merck 60 F254 aluminum-backed silica gel plates
(0.2 mm). Developed plates were visualized using UV light or 2.0%
phosphomolybdic acid stain. Flash column chromatography was
performed using Merck silica gel 60 (230–400 mesh).
3-(2-Hydroxy-5-methylphenyl)-3-phenylpropyl 4-Nitroben-
zenesulfonate (6)
3-Hydroxy-3-phenylpropyl 4-nitrobenzenesulfonate (5a, 1.199 g,
3.555 mmol) and FeCl3 6 H2O (0.0961 g, 0.3555 mmol) were dis-
solved in CH2Cl2 (14.6 mL). p-Cresol (1.538 g, 14.22 mmol) was
then added and the mixture was refluxed vigorously for 30 min. The
mixture was allowed to cool to r.t. and was quenched with H2O (15
mL) for 25 min. The layers were separated. The organic phase was
washed with H2O (2 × 15 mL). It was dried (anhyd MgSO4), fil-
tered, and concentrated using a rotary evaporator. The mixture was
further purified using column chromatography twice (silica gel,
hexane–EtOAc, 5:1 and hexane–EtOAc, 8:1) to give 6 (0.8896 g,
59%) as a gummy yellow solid.
1-Phenylpropane-1,3-diol (4)4
Ethyl benzoylacetate (3, 25.80 mL, 149.9 mmol) was dissolved in
MeOH (300 mL) and NaBH4 (17.02 g, 449.8 mmol) was added
slowly to the mixture at r.t.; the mixture was stirred for 15 min. All
of the MeOH was evaporated and EtOAc was added (50 mL). The
mixture was shaken with supersaturated brine soln (30 mL) and the
mixture was filtered to remove the solid. The layers were separated
and the aqueous layer was extracted using EtOAc (50 mL as many
times necessary). The organic phases were combined, dried (anhyd
MgSO4), filtered, and concentrated using rotary evaporator. The
mixture was further purified using short column chromatography
(silica gel, hexane–EtOAc, 2:1) to give 4 (22.36 g, 98%) as a color-
less oil.
IR (KBr): 3487, 1534, 1349, 1311, 1183, 1092 cm–1.
1H NMR (300 MHz, CDCl3): d = 2.21 (s, 3 H), 2.41 (m, 2 H), 4.11
(t, J = 6.3 Hz, 2 H), 4.32 (t, J = 7.8 Hz, 1 H), 6.56 (d, J = 7.2 Hz, 1
H), 6.85 (d, J = 7.2 Hz, 1 H), 6.86 (s, 1 H), 7.20 (m, 5 H), 8.00 (d,
J = 8.7 Hz, 2 H), 8.30 (d, J = 8.4 Hz, 2 H).
13C NMR (75 MHz, CDCl3): d = 21.0, 33.6, 40.0, 70.5, 116.1,
124.6, 126.9, 128.1, 128.4, 128.6, 128.9, 129.0, 129.4, 130.4, 141.7,
142.7, 150.8, 151.1.
1H NMR (300 MHz, CDCl3): d = 1.95 (m, 2 H), 2.79 (s, 1 H), 3.23
(s, 1 H), 3.84 (m, 2 H), 4.94 (dd, J = 8.3, 6.0 Hz, 1 H), 7.30 (m, 5 H).
HRMS (FAB): m/z [M]+ calcd for C22H21NO6S: 427.1090; found:
1H NMR (300 MHz, CDCl3): d (upon D2O exchange) = 1.96 (m, 2
H), 3.83 (t, J = 5.5 Hz, 1 H), 4.79 (s, 2 H), 4.94 (dd, J = 8.5, 4.1 Hz,
1 H), 7.31 (m, 5 H).
427.1085.
3-(Diisopropylamino)-1-phenylpropan-1-ol (7)14
3-Hydroxy-3-phenylpropyl 4-nitrobenzenesulfonate (5a, 3.501 g,
10.38 mmol) was refluxed with i-Pr2NH (29.30 mL) for 3 h. The
mixture was allowed to cool to r.t. and was filtered using CH2Cl2 for
washing. The filtrate was washed with H2O and brine alternately
several times until the salt was completely removed. All the organic
13C NMR (75 MHz, CDCl3): d = 40.7, 61.6, 74.5, 125.9, 127.8,
128.8, 144.6.
Synthesis 2008, No. 12, 1841–1844 © Thieme Stuttgart · New York

1844
K. A. De Castro, H. Rhee
PAPER
(2) (a) Gage, J. R.; Cabaj, J. E. WO 199829402, 1998.
phase was combined, dried (anhyd MgSO4), filtered, and concen-
trated using a rotary evaporator. The product was further dried un-
der vacuum to give 7 (2.102 g, 86%) as a yellow oil.
1H NMR (300 MHz, CDCl3): d = 1.06 (d, J = 6.6 Hz, 6 H), 1.16 (d,
J = 6.9 Hz, 6 H), 1.66 (m, 1 H), 1.84 (m, 1 H), 2.79 (m, 2 H), 3.20
(m, 2 H), 4.90 (dd, J = 9.9, 2.5 Hz, 1 H), 7.32 (m, 5 H).
(b) Botteghi, C.; Corrias, T.; Marchetti, M.; Paganelli, S.;
Piccolo, O. Org. Process Res. Dev. 2002, 6, 379.
(c) Kumar, Y.; Prasad, M.; Misra, S. WO 2003014060,
2003. (d) Kankan, R. N.; Rao, D. R. WO 2005061432,
2005. (e) Srinivas, K.; Srinivasan, N.; Reddy, K. S.;
Ramakrishna, M.; Reddy, C. R.; Arunagiri, M.; Kumari, R.
L.; Venkatarman, S.; Mathad, V. T. Org. Process Res. Dev.
2005, 9, 314. (f) Venkataraman, S.; Mathad, V. T.; Reddy,
K. S.; Srinivasan, N.; Reddy, C. R.; Arunagiri, M.; Kumari,
R. L. US 20060094904, 2006. (g) Toplak, R. WO
2006066931, 2006. (h) Turchetta, S.; Ciambecchini, U.;
Massardo, P. US 20060079716, 2006.
13C NMR (75 MHz, CDCl3): d = 18.8, 21.9, 35.3, 44.2, 47.5, 76.3,
125.8, 127.0, 128.4, 145.5.
rac-Tolterodine
3-(Diisopropylamino)-1-phenylpropan-1-ol (7, 2.084 g, 8.855
mmol) was dissolved in CH2Cl2 (4.00 mL) and p-cresol (0.9576 g,
8.855 mmol) was added to the soln followed by 70% aq HClO4
(5.35 mL). The mixture was stirred at r.t. for 3 h. Ice was added to
the mixture and it was neutralized with aq NaOH; the mixture was
quenched for 30 min. The layers were separated and the organic lay-
er was washed with brine and distilled H2O. The organic phases
were combined, dried (anhyd MgSO4), filtered, and concentrated
using a rotary evaporator. The product was further dried under vac-
uum to give the product (2.507 g, 87%) as a gummy yellow solid.
(3) (a) Anderson, P. G.; Schink, H. E.; Osterlund, K. J. Org.
Chem. 1998, 63, 8067. (b) Colombo, L.; Rossi, R.; Castaldi,
G.; Allegrini, P. CA 2502640, 2005. (c) Hedberg, C.;
Anderson, P. G. Adv. Synth. Catal. 2005, 347, 662.
(d) Chen, G.; Tokunaga, N.; Hayashi, T. Org. Lett. 2005, 7,
2285. (e) Ulgheri, F.; Marchetti, M.; Piccolo, O. J. Org.
Chem. 2007, 72, 6056.
(4) De Castro, K. A.; Ko, J.; Park, D.; Park, S.; Rhee, H. Org.
1H NMR (300 MHz, CDCl3): d = 1.07 (d, J = 6.3 Hz, 6 H), 1.13 (d,
J = 6.9 Hz, 6 H), 2.11 (s, 3 H), 2.36 (m, 2 H), 2.72 (m, 2 H), 3.23
(m, 2 H), 4.49 (dd, J = 11.1, 3.9 Hz, 1 H), 6.53 (br s, 1 H), 6.83 (m,
2 H), 7.23 (m, 1 H), 7.32 (d, J = 4.5 Hz, 4 H).
13C NMR (75 MHz, CDCl3): d = 19.7, 20.2, 21.0, 33.3, 39.5, 42.2,
48.2, 118.4, 126.4, 128.0, 128.5, 128.8, 128.9, 129.6, 132.7, 144.9,
153.5.
Process Res. Dev. 2007, 5, 918.
(5) De Luca, L.; Giacomelli, G.; Porcheddu, A. Org. Lett. 2002,
4, 553.
(6) Sereda, G. A. Tetrahedron Lett. 2004, 45, 7265.
(7) (a) Martinelli, M. J.; Nayyar, N. K.; Moher, E. D.; Dhokte,
U. P.; Pawlak, J. M.; Vaidyanathan, R. Org. Lett. 1999, 1,
447. (b) Choudary, B. M.; Chowdari, N. S.; Kantam, M. L.
Tetrahedron 2000, 56, 7291. (c) Martinelli, M. J.;
Vaidyanathan, R.; Pawlak, J. M.; Nayyar, N. K.; Dhokte, U.
P.; Doecke, C. W.; Zollars, L. M. H.; Moher, E. D.; Khau, V.
V.; Kosmrlj, B. J. Am. Chem. Soc. 2002, 124, 3578.
(d) Das, B.; Reddy, V. S.; Reddy, M. R. Tetrahedron Lett.
2004, 45, 6717. (e) Fazaeli, R.; Tangestaninejad, S.; Aliyan,
H. Can. J. Chem. 2006, 84, 812.
Acknowledgment
K.A.D.C. greatly appreciates the Korean Ministry of Education
through the second stage of BK21 project for Hanyang University
graduate program for the financial support. She also thanks Mr.
Jong Moon Park for the help in the spectral characterization.
(8) Kolb, H. C.; Van Nieuwenhze, M. S.; Sharpless, K. B.
Chem. Rev. 1994, 94, 2483.
(9) The flash column chromatography may be avoided by using
crystallization. This is for industrial consideration.
(10) Iovel, I.; Mertins, K.; Kischel, J.; Zapf, A.; Beller, M.
Angew. Chem. Int. Ed. 2005, 44, 3913.
(11) Heydari, A.; Khaksar, S.; Akbari, J.; Esfandyari, M.;
Pourayoubi, M.; Tajbakhsh, M. Tetrahedron Lett. 2007, 48,
1135.
(12) Li, A.; Kindelin, P. J.; Klump, D. A. Org. Lett. 2006, 8,
1233.
(13) Li, D.; Guo, Y.; Ding, Y.; Xiao, W. Chem. Commun. 2006,
799.
References
(1) (a) Nilvebrant, L. Life Sci. 2001, 68, 2549. (b) Chancellor,
M. B.; Appell, R. A.; Sathyan, G.; Gupta, S. K. Clin.
Therapeut. 2001, 23, 753. (c) Kaur, K.; Aeron, S.;
Bruhaspathy, M.; Shetty, S. J.; Gupta, S.; Hegde, L. H.;
Silamkoti, A. D. V.; Mehta, A.; Chugh, A.; Gupta, J. B.;
Sarma, P. K. S.; Kumar, N. Bioorg. Med. Chem. Lett. 2005,
15, 2093.
(14) Razzetti, G.; Mantegazza, S.; Rossi, R.; Allegrini, P. EP
20060823, 2006.
Synthesis 2008, No. 12, 1841–1844 © Thieme Stuttgart · New York