Expeditious process improvement for the synthesis of RWJ-333966

Article abstract of DOI:10.1021/op0601565

We improved chemical processes for synthesizing RWJ-333966 1 and obtained this compound over five steps in 40% overall yield.

Full text of DOI:10.1021/op0601565

Organic Process Research & Development 2006, 10, 1227−1230
Expeditious Process Improvement for the Synthesis of RWJ-333966
Michel Guillaume*
Chemical Process Research, Johnson & Johnson Pharmaceutical Research and DeVelopment, Turnhoutseweg, 30,
2340 Beerse, Belgium
Table 1
Abstract:
We improved chemical processes for synthesizing RWJ-333966
1 and obtained this compound over five steps in 40% overall
yield.
iPrOH
(L/mol)
filtration
(°C)
yield
(%)
Introduction
1
0.8
25
0
40
82
RWJ-333966 1 is a selective vasopressin 1a receptor
antagonist that presents utility in rodent models, predictive
of clinical efficacy for the treatment of anxiety disorders. It
is well tolerated after multiple daily administrations and does
not adversely affect learning and memory.1 Preclinical studies
were initiated to support clinical proof-of-principle and
clinical profile regarding sedation, motor impairment, and
tolerance. Development was halted before first-in-human
dose, after a proper GMP synthesis had been devised
(Scheme 1).
noticed no difference in reactivity. In a second series of trials,
we used other solvents (acetone and MIK), but we observed
lower conversion, longer reaction times, and/or formation
of impurities as compared with DMA. In a third series of
experiments (entries 5 and 6), we optimized reaction condi-
tions using 1 L/mol DMA instead of 2 L/mol and 1 equiv of
K2CO3 instead of 2 equiv. The reaction was carried out at
80 °C during 5-7 h.
Addition of water generated an oily precipitate, and we
therefore extracted the product with toluene (2 L/mol).
Having noticed that the next step (BOC deprotection of 5 to
6) worked well in dichloromethane with 5 equiv of trifluoro-
acetic acid, we tried the reaction directly on the toluene
extract and successfully obtained the bis-trifluoroacetate of
63 as a white crystalline compound with 82% yield. The
fourth step (6 to 8) was originally performed by adding
diphenyl isothiocyanate 7 to secondary amine 6. Because of
the limited commercial availability of 7, it was decided to
prepare this reagent in situ from diphenylamine and thio-
carbonyldiimidazole (Scheme 2).
The reaction was performed in dichloromethane, and it
was assumed that no additional base was required as far as
released imidazole could neutralize the first equivalent of
trifluoroacetic acid, the second equivalent being neutralized
by the second imidazole formed in the reaction. The
assumption was right: the reaction worked indeed well
without adding a base. Surprisingly, the reaction proceeded
well at 25 °C, whereas it is mentioned in literature that
heating is often necessary for the second addition of amine.4
After washing the reaction mixture with water, adding
ethanol, and distilling off the dichloromethane, product 8
was crystallized with 84% yield.
Results and Discussion
In the first step, 2 was protected with Boc-anhydride (1
equiv). Using more then 1 equiv led to the corresponding
unwanted di-Boc derivative. Originally, the reaction was
performed in THF, and compound 3 was obtained after a
tedious workup. Therefore, three other solvents were
screened: dichloromethane, toluene, and isopropanol. In the
latter case, the product nicely precipitated, and we further
optimized the process by reducing the amount of solvent and
working at 0 °C. We obtained 3 with 82% yield (Table 1).
The second step (3 to 5) involves a nucleophilic substitu-
tion of protected amino alkyl bromide 4 on the benzimid-
azolone 3. Results are presented in Table 2. The reaction
was originally performed in DMF using Cs2CO3 as base.
Because of the limited stability of DMF and because it
contains dimethylamine in various quantities, we wanted to
change it. We also wanted to use a cheaper base2 and to
obtain a crystalline compound. In a first series of trials
(entries 1 and 2), we used N,N-dimethyl acetamide (DMA,
2 L/mol) instead of DMF and K2CO3 instead of Cs2CO3 and
(3) The first equivalent of acid is obviously bound tightly to the NH of the
secondary amine. On the contrary, the second equivalent of CF3COOH
(pKa ∼0.5) is expected to bind only weakly with benzimidazolone (pKa
) 0.76, simulated with ACD software). Anyway, titration results consist-
ently indicate the presence of 2 equiv of acid.
* To whom correspondence should be addressed. E-mail: mguillau@
prdbe.jnj.com.
(1) Urbanski, M. J.; Gunnet, J. W.; Demarest, K. J. WO 02/0555 14, 2002.
(2) Comparison of commercial (Acros) prices indicates a huge difference: 665
EUR/kg for Cs2CO3 and 23 EUR/kg for K2CO3!
(4) Grzyb, J. A.; Shen, M.; Yoshina-Ishii, Ch.; Chi, W.; Brown, R. S.; Batey,
R. A. Tetrahedron 2005, 61, 7153.
10.1021/op0601565 CCC: $33.50 © 2006 American Chemical Society
Published on Web 10/21/2006
Vol. 10, No. 6, 2006 / Organic Process Research & Development
•
1227

Scheme 1
Table 2
3
4
K2CO3
(equiv)
entry
(equiv)
(equiv)
solvent, conditions
conversion
1
2
3
4
5
6
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
1
1
DMA, 2 L/mol, 80 °C, 2 h
DMA, 2 L/mol, 80 °C, 2 h
acetone, 2 L/mol, 56 °C (rfx.), 1 h
MIK, 2 L/mol, 56 °C (rfx.), 16 h
DMA, 1 L/mol, 80 °C, 5 h
DMA, 1 L/mol, 80 °C, 7 h
>99%
>99%
33%
>99% + impurities
>99%
>99%
Scheme 2
amide. Finally we performed the reaction with methylamine
(MeNH2). With 5 equiv in MeOH/H2O, the reaction worked
well, but still 11% (LC area%) of the amide remained after
1 h at 45 °C. At the same temperature, but with 10 equiv of
MeNH2, the product still contained 5% of the amide after
1.5 h. Having extracted the product, we evaporated the
solvent and noticed the reclosure of the amide. To avoid this,
we added HCl/H2O after the reaction to precipitate the
product. We obtained 90% active yield, but it contained
unfortunately MeNH2‚HCl as well. We screened several acids
to selectively isolate the product, but our attempts were
unsuccessful. Therefore, compound 1 was chromatographed
after base liberation. After the chromatography (RP, aceto-
nitrile/water/NH4OAc), we extracted the product with dichloro-
methane and evaporated the solvent. We noticed however
some product degradation as outlined in Scheme 4.
The last step (8 f 1) was by far the most difficult one
(Scheme 3).5
First, we tried the hydrolysis of the phthalimide moiety
with HCl (in water and in water/EtOH). The SM disappeared,
but a typical smell indicated that the thiourea had been
hydrolyzed. Then, we tried ethylenediamine (2 equiv) in
s-BuOH at 90 °C. After 30 min, some product was formed,
but LC/MS analysis indicated a competition with aminolysis
of thiourea. Then we used hydrazine (2 equiv) in EtOH. The
conversion was not complete, and another 1 equiv was added.
A precipitate was formed which contained the product but
with insufficient purity, even after recrystallization from
MeOH. With KOH (5 equiv) in EtOH the first imide
carbonyl was cleaved to the amide after 30 min, but then,
intermediate 9 was cleaved at the thiourea instead of the
Through the formation of the fumarate salt,6 we were able
to get rid of this limited amount of impurities (up to 3%).
At lab scale, we synthesized about 70 g of DS in total. The
project was discontinued before reaching the pilot plant scale.
(5) For the deprotection of phtalimide derivatives, see references in: Greene,
Th. W.; Wuts, P. G. M. ProtectiVe Groups in Organic Synthesis, 3rd ed.;
Wiley: 1999; pp 564-566.
(6) The following acids were tried as well: succinic acid, maleic acid, mandelic
acid, fumaric acid, citric acid, lactic acid, and diisopropylidene keto gulonic
acid.
1228
•
Vol. 10, No. 6, 2006 / Organic Process Research & Development

Scheme 3
Scheme 4
In conclusion, we developed an efficient and scalable
1H NMR (DMSO-d6, TMS, 400 MHz) δ (ppm): 1.44
(s,9H), 1.68 (dd, J1 ) 11.96, J2 ) 1.89 Hz, 2H), 2.20 (qd,
J1 ) 12.55, J2 ) 4.41 Hz, 2H), 2.87 (m, 2H), 4.10 (d, J )
11.83 Hz, 2H), 4.29-4.38 (m, 1H), 6.94-7.00 (m, 3H),
7.15-7.21 (m, 1H), 10.83 (s,1H). MS (ESI) m/z: 318 (MH+).
Anal. Calcd for C17H23N3O3: C, 64.33; H, 7.30; N, 13.24.
Found: C, 64.25; H, 7.36; N, 13.45.
method to synthesize compound 1 over five steps in 40%
overall yield. Unfortunately, due to its amphiphilic nature
(amine and thiourea on the same molecule), the product
showed only limited stability and must be chromatographed
before its salt is formed.
2-[3-(2-Oxo-3-piperidin-4-yl-2,3-dihydrobenzoimidazol-
1-yl)propyl]isoindole-1,3-dione Ditrifluoroacetate (6‚
2CF3COOH). Bromopropylphtalimide 4 (13.4 g, 0.05 mol,
1 equiv) and potassium carbonate (6.9 g, 0.05 mol) were
added to a solution of 3 (15.8 g, 0.05 mol) in 50 mL of
DMA. The reaction mixture was heated to 80 °C and stirred
for 6 h. Water was added (100 mL, 2 L/mol) followed by
toluene (100 mL, 2 L/mol). The organic layer was separated
and dried over Na2SO4 (50 g). One-third of the solvent was
partially evaporated and cooled to 20 °C. Trifluoroacetic acid
(28.5 g, 0.25 mol) was added to the residue, maintaining
the temperature below 60 °C. The mixture was stirred at 60
°C during 4 h and cooled to 40 °C over 30 min. The mixture
was seeded at 40 °C, cooled to 20 °C over 2 h, and then
stirred for 16 h. The trifluoroacetic salt of 6 was filtered,
washed with toluene, and dried in a vacuum at 40 °C during
4 h. Yield: 25.8 g (82%).
1H NMR (DMSO-d6, TMS, 400 MHz) δ (ppm): 1.88 (d,
J ) 12.38 Hz, 2H), 2.00-2.09 (q, J ) 7.2 Hz, 2H), 2.49-
2.61 (m, 2H), 3.07-3.17 (m, 2H), 3.45 (d, J ) 12.38 Hz,
2H), 3.63 (t, J ) 7.08 Hz, 2H), 3.91 (t, J ) 7.08 Hz, 2H),
4.50-4.60 (m, 1H), 6.96-7.09 (m, 2H), 7.22-7.30 (m, 1H),
7.30-7.37 (m, 1H), 7.77-7.85 (m, 4H), 8.55 (m,1H), 8.79
(m, 1H). MS (ESI) m/z: 405 (MH+). Acid titration for
C23H24N4O3‚2C4H2F6O4: 100,7 w/w %. Anal. Calcd for
C23H24N4O3‚2C4H2F6O4: C, 51.22; H, 4.11; N, 8.85.
Found: C, 50.48; H, 4.09; N, 8.87.
Experimental Section
General Procedures. Commercially available solvents
and reagents were used without further purification. 1-Pip-
eridin-4-yl-1,3-dihydrobenzoimidazol-2-one was obtained
from a commercial source.
1H NMR spectra were recorded on a Bruker Avance 400
MHz spectrometer, with TMS as internal standard in DMSO-
d6. MS spectra were determined with a Waters mass
spectrometer with a UPLC technique. Elemental analysis
were carried out on a Carlo Erba EA1110.
Trifluoroacetic and fumaric contents were measured by
potentiometric acid titration on a Metrohm 716 DMS Titrino
apparatus with a Ag/AgCl electrode.
Preparative HPLC was performed on a Merck system with
UV detection at 260 nm through a Prochrom, KR100-10 RP
18 column. The separation was carried out by isocratic
elution, using aqueous acetonitrile and ammonium acetate
in the ratio 50:50. The flow rate was 500 mL/min.
4-(2-Oxo-2,3-dihydro-benzoimidazol-1-yl)-piperidine-
1-carboxylic Acid tert-Butyl Ester (3). A suspension of
1-piperidin-4-yl-1,3-dihydrobenzoimidazol-2-one 27 (217 g,
1 mol) in isopropanol (800 mL, 0.8 L/mol) was heated to
30 °C. Di-tert-butyl dicarbonate (218 g, 1 mol) was added,
neat, keeping the temperature below 50 °C during addition.
The reaction mixture was stirred at 40 °C during 1 h, cooled
to 20 °C, and stirred 16 h at this temperature. The mixture
was cooled to 0 °C and stirred for 1 h. The product was
collected by filtration and dried in a vacuum oven at 50 °C.
Yield: 260 g (82%).
4-{3-[3-(1,3-Dioxo-1,3-dihydroisoindol-2-yl)propyl]-2-
oxo-2,3-dihydrobenzoimidazol-1-yl}piperidine-1-carbo-
thioic Acid Benzhydrylamide (8). 1,1-Diphenylmethyl-
amine (37.2 g, 0.22 mol) was added to a solution of
thiocarbonyldiimidazole (39.2 g, 0.22 mol, 1.1 equiv) in
(7) From our own production facility. This compound is also available in gram
quantities from major fine chemical suppliers.
Vol. 10, No. 6, 2006 / Organic Process Research & Development
•
1229

dichloromethane (200 mL, 1 L/mol)8 maintaining the tem-
perature below 35 °C. The reaction mixture was stirred at
20 °C for 1 h, and 6‚2CF3COOH (126.4 g, 0.2 mol) was
added portionwise keeping the temperature below 35 °C. The
reaction mixture was stirred at 20 °C for 1 h, and water was
added (400 mL, 2 L/mol). The organic layer was separated,
and the solvent was two-thirds evaporated. Ethanol (1 L, 5
L/mol) was added to the residue. The mixture was further
evaporated until a temperature of 80 °C was reached. The
heterogeneous mixture was cooled to 20 °C, stirred for 4 h,
cooled to 0 °C, and stirred for a 1 h. 8 was filtered and dried
under a vacuum at 40 °C. Yield: 106 g (84%)
1H NMR (DMSO-d6, TMS, 400 MHz) δ (ppm): 1.77 (d,
J ) 9.73 Hz, 2H), 1.99-2.08 (q, J ) 7.08 Hz, 2H), 2.20-
2.32 (m, 2H), 3.16 (t, J ) 12.38 Hz, 2H), 3.63 (t, J ) 7.08
Hz, 2H), 3.89 (t, J ) 7.08 Hz, 2H), 4.49-4.57 (m, 1H),
4.96 (d, J ) 12.83 Hz, 2H), 7.02 (m, 2H), 7.12 (d, J ) 7.08
Hz, 1H), 7.17 (d, J ) 8.4 Hz, 1H), 7.23 (d, J ) 7.52 Hz,
1H), 7.26-7.38 (m, 10H), 7.79-7.85 (m, 4H), 8.43 (d, J )
8.4 Hz, 1H). MS (ESI) m/z: 630 (MH+). Anal. Calcd for
C37H35N5O3S: C, 70.57; H, 5.60; N, 11.12. Found: C, 70.32;
H, 5.50; N, 11.08.
stirred for awhile and filtered as 1‚HCl. Yield: 48 g (96%,
76% purity, this corresponds to an active yield of 72%)
Another quantity of 1‚HCl was obtained by adding
another 50 mL of concentrated HCl and stirring overnight
at room temperature. Yield: 21.2 g (43%, 41.5% purity, this
corresponds to an active yield of 18%). Total active yield )
90%.
After base liberation 1 was purified by preparative
chromatography. The resulting aqueous solution was ex-
tracted with dichloromethane, and the organic layer was
carefully evaporated under a vacuum below 35 °C.
Salt formation: 1 (21.6 g, 0.043 mol) and fumaric acid
(5.4 g, 1.1 equiv) were dissolved in ethanol (435 mL) at
reflux, cooled to room temperature, and filtered. 1‚C4H4O4
was dried under a vacuum at 40 °C. Yield: 21.6 g (81%,
96.4% purity, this corresponds to an active yield of 78%).
1H NMR (DMSO-d6, TMS, 400 MHz) δ (ppm): 1.75 (d,
J ) 9.51 Hz, 2H), 1.91 (q, J ) 7.2, 2H), 2.20-2.32 (m,
2H), 2.80 (t, J ) 7.3 Hz, 2H), 3.14 (t, J ) 12.44 Hz, 2H),
3.89 (t, J ) 6.77 Hz, 2H), 4.49-4.6 (m, 1H), 4.93 (d, J )
12.81 Hz, 2H), 6.41 (s, 2H), 6.99-7.10 (m, 2H), 7.13 (d, J
) 7.32 Hz, 2H), 7.24-7.37 (m, 11H), 8.47 (d, J ) 8.78 Hz,
1H). MS (ESI) m/z: 500 (MH+). Acid titration for C29H33N5-
OS‚C4H4O4: 99.1 w/w%. Anal. Calcd for C29H33N5OS‚
C4H4O4: C, 64.37; H, 6.06; N, 11.37. Found: C, 64.16; H,
5.88; N, 10.85.
4-[3-(3-Aminopropyl)]-2-oxo-2,3-dihydrobenzoimid-
azol-1-yl-piperidine-1-carbothioic Acid Benzhydrylamide
(E)-2-Butenedioate (1‚C4H4O4). Phthalimide 8 (63 g, 0.1
mol) was dissolved in methanol (400 mL, 4 L/mol), and
methylamine 40% in water (85 mL, 1 mol) was added. The
reaction mixture was warmed to 45 °C and stirred for 1.5 h.
After cooling to 10 °C with an ice bath, HCl conc. (85 mL)
was added until pH ∼3. The heterogeneous mixture was
Received for review August 1, 2006.
OP0601565
(8) Based on compound 6‚2CF3COOH.
1230
•
Vol. 10, No. 6, 2006 / Organic Process Research & Development