Development of a scalable process for 1-(3,5-dichlorophenyl)-5-iodo-3- methyl-(4-methylbenzyl)-1H-imidazo[1,2-a]imidazol-2-one: A key intermediate for the synthesis of LFA-1 inhibitors

Article abstract of DOI:10.1021/op049813o

A safe, robust, chromatography-free and reproducible process for the multi-kilogram synthesis of vinyl iodide 2, a key intermediate for the synthesis of LFA-1 inhibitors, was developed and implemented at the pilot plant. Execution of the above process all

Full text of DOI:10.1021/op049813o

Organic Process Research & Development 2005, 9, 137−140
Development of a Scalable Process for 1-(3,5-Dichlorophenyl)-5-iodo-3-methyl-
(4-methylbenzyl)-1H-imidazo[1,2-a]imidazol-2-one: A Key Intermediate for the
Synthesis of LFA-1 Inhibitors
Rogelio P. Frutos,* Magnus Eriksson, Xiao-Jun Wang, Denis Byrne, Richard Varsolona, Michael D. Johnson,
Lawrence Nummy, Dhileepkumar Krishnamurthy,* and Chris H. Senanayake
Department of Chemical DeVelopment, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road/P.O. Box 368,
Ridgefield, Connecticut 06877-0368, U.S.A.
Abstract:
iodination of 3 with NIS at ambient temperature. This
iodination gave a 4:1 mixture of 2 and 5 (eq 1) that was
A safe, robust, chromatography-free and reproducible process
for the multi-kilogram synthesis of vinyl iodide 2, a key
intermediate for the synthesis of LFA-1 inhibitors, was devel-
oped and implemented at the pilot plant. Execution of the above
process allowed us to support preclinical activities in the LFA-1
program.
Introduction
Small molecules that block the protein-protein interaction
of cell adhesion molecules such as LFA-1 (lymphocyte
function-associated antigen) and ICAM-1 (intracellular cell
adhesion molecule) are potential drugs for the treatment of
immune disorders such rheumatoid arthritis, psoriasis, and
Crohn’s disease.1 Accordingly, SAR studies by our Discovery
department aimed at identifying potent LFA-1 inhibitors led
to the selection of 1H-imidazo[1,2-R]imidazol-2-ones such
as 1 (Figure 1) as promising clinical candidates.2 As a result,
the need arose for the preparation of multi-kilogram quanti-
ties of these compounds in order to support preclinical
activities. Since the original Discovery route suffered from
a number of limitations that made it unsuitable for scale-up,
process research was initiated to develop a suitable and
scaleable synthetic route toward these compounds. A major
drawback of the original route was the synthesis of inter-
mediate 2 (Figure 1) by means of a nonregioselective
iodination. Herein, we report the development of a scaleable
process for the regioselective synthesis of key intermediate
2 for its multi-kilogram production in the pilot plant.
separated by silica gel chromatography. Evaluation of
different reaction conditions failed to give 2 as a single
product, but by reducing the temperature to 0 °C and
quenching the reaction at 60% conversion, the ratio of 2 to
5 was improved to approximately 11:1 along with traces of
4 (3-5%). However, the isolation of 2 still required a
difficult and labor intensive purification by silica gel chro-
matography.
To circumvent the regioselectivity problems associated
with the direct iodination of 3, we envisioned the synthesis
of intermediate 6 (with the correct regiochemistry and the
same oxidation state as 2) and transformation of the lactam
carbonyl group into the required vinyl iodide (Figure 2).
Figure 2.
Bis-lactam 6 was prepared from 73 according to our
previously described synthesis (Scheme 1).4 Dehydration
reagents other than Ph3P/CCl4, such as Burgess reagent or
p-TsCl/Et3N, failed to give any of the desired 8.
With a practical route to intermediate 6 now established,
we turned our attention to the transformation of bis-lactam
6 into iodide 2. Early attempts at implementing Wiemer’s5
procedure for the TMSI (or TMSCl/NaI) promoted trans-
Figure 1.
(1) Springer, T. A. Nature 1990, 346, 425.
Results and Discussion
As mentioned above, in the original Discovery route
compound 2 was obtained by means of a nonregioselective
(2) Wu, J.-P.; Emeigh, J.; Gao, D. A.; Goldberg, D. R.; Kuzmich, D.; Miao,
C.; Potocki, I.; Qian, K. C.; Sorcek, R. J.; Jeanfavre, D. D.; Kishimoto, K.;
Mainolfi, E. A.; Nabozni, G.; Peng, C.; Reilly, P.; Rothlein, R.; Sellati, R.
H.; Woska, J. R.; Chen, S.; Gunn, J. A.; O’Brien, D.; Norris, S. H.; Kelly,
T. A. J. Med. Chem. 2004, 5356 and references therein.
* To whom correspondence should be addressed. E-mail: rfrutos@
rdg.boehringer-ingelheim.com.
(3) (a) Yee, N. K. Org. Lett. 2000, 2, 2781. (b) Frutos, R. P.; Stehle, S.; Nummy,
L.; Yee, N. Tetrahedron: Asymmetry 2001, 12, 101.
10.1021/op049813o CCC: $30.25 © 2005 American Chemical Society
Published on Web 01/15/2005
Vol. 9, No. 2, 2005 / Organic Process Research & Development
•
137

Scheme 1a
Table 1. Synthesis of 2 under different conditions
product
%
yield
(of 2)
reagents
and solvent
T
ratios
entry
1
(°C) time
(2:3:6:9)
TMSCl (3 equiv)
NaI (3 equiv)
H2O (0.7-1 equiv)
CH2Cl2
TMSCl (4 equiv)
NaI (4 equiv)
H2O (2 equiv)
CH2Cl2
TMSCl (6 equiv)
NaI (4.5 equiv)
H2O (4.5 equiv)
CH3CN
TMSI (4.4 equiv)
H2O (2.9 equiv)
CH3CN
25 2-5 h 75:12:<1:1 70-75
25 20 min 75:12:<1:1 70-75
a Reagents and conditions: (a) (i) Ethyl isocyanoacetate, CH2Cl2, rt, 2.5 hr;
(ii) Ph3P, CCl4, Et3N, rt, 12 h (81%); (b) Me3Al, Ph3PO, toulene, 25 °C (91%).
Scheme 2a
2
3
4
-15 80 h
-7 40 h
90:5:3:2
86:5:1:8
82
82
a Reagents and conditions: (a) KN(TMS)2, (EtO)2POCl, THF, -20 °C (92%);
(b) NaI, TMSCl, H2O, CH2Cl2 (70-75%).
efficiency of the reaction. For example, it was not obvious
why the reaction required water to proceed efficiently, and
the means by which compound 3 formed during the reaction
were not clear. Moreover, the reaction mixture turned red/
brown suggesting the formation of I2, which was quenched
upon addition of sodium thiosolfate. These issues needed to
be understood better to facilitate the identification of the best
possible reaction conditions that would maximize the yield
of the desired product 2 and eliminate side product formation.
Reports on the use of the TMSCl/NaI/water system for
the synthesis of vinyl iodides via hydroiodination of alkynes
suggest this system is a good way to generate HI in situ
under mild conditions.7 Consequently, a possible mechanism
for the transformation of 9 into 2 consistent with our
observations and previous reports on the use of the TMSCl/
NaI/H2O system for the generation of HI is shown in Scheme
3. Accordingly, protonation of 9 followed by iodide addition
to the resulting cation 10 would result in 11, which upon
elimination of diethyl phosphate would afford the desired
vinyl iodide 2. In the presence of excess HI under prolonged
reaction times, however, iodide 2 could be protonated to give
12, which would result in the formation of 3 upon elimination
of I2.
In agreement with the proposed mechanism is the obser-
vation that a significant amount of the proto-deiodination
side product 3 is obtained when iodide 2 is resubmitted to
the reaction conditions. Furthermore, when D2O is added to
the reaction mixture instead of H2O, 3 is obtained with
deuterium incorporation at the position formerly occupied
by the iodide atom (eq 2).
With the above mechanism in mind, we looked at
systematically modifying the reaction parameters (solvent,
reaction time, temperature, order of addition, etc.) in order
formation of ketone-derived vinyl phosphates into vinyl
iodides to our lactam-derived phosphate 9 resulted in low
and irreproducible yields of 2. After careful experimentation,
however, we discovered that addition of 1 equiv of water to
the reaction mixture resulted in good and consistent yields
of iodide 2.4 Accordingly, vinyl iodide 2 was synthesized
reproducibly in multi-kilogram amounts using the approach
shown in Scheme 2.6
Our original4 experimental conditions for the transforma-
tion of 9 into 2 involved addition of TMSCl (3 equiv), NaI
(3 equiv), and water (1 equiv) to a solution of phosphate 9
in dichloromethane to give an approximately 6:1:1 mixture
of 2, 3, and unreacted 9 along with traces of 6 after 2-5 h,
resulting in a 70-75% isolated yield of 2 after chromatog-
raphy (Table 1, entry 1). Prolonged reaction times resulted
in larger amounts of 3, and stopping the reaction earlier
minimized the amount of 3 but resulted in a larger amount
of unreacted 9. In addition, under apparently identical
reaction conditions, the time it took for the reaction to reach
the desired product distribution varied considerably. As a
result, the reaction had to be monitored closely by either
HPLC or TLC and quenched at the optimum time to give
the above-mentioned ratio of products.
To our knowledge there are no published reports on the
mechanism for this transformation that could shed light into
further improvement of this process. There were a few issues
that needed to be investigated in order to improve the
(4) Frutos, R. P.; Johnson, M. Tetrahedron Lett. 2003, 44, 6509.
(5) Lee, K.; Wiemer, D. F. Tetrahedron Lett. 1993, 34, 2433.
(6) Alternative procedures for the multi-kilogram production of 6 have been
developed. For additional information on the subject, see Wang, X.-j.; Zhang,
L.; Xu, Y.; Krishamurthy, D.; Varsolona, R.; Nummy, L.; Frutos, R. P.;
Byrne, D.; Chung, J. C.; Farina, V.; Senanayake, C. Tetrahedron Lett. 2005,
46, 273.
(7) Kamiya, N.; Chikami, Y.; Ishii, Y. Synlett 1990, 675 and references therein.
138
•
Vol. 9, No. 2, 2005 / Organic Process Research & Development

Scheme 3. Proposed mechanism for the formation of 2 and
3 from 9
Ultimately, major improvements were realized by lower-
ing the reaction temperature and changing the solvent from
dichloromethane to acetonitrile. Under the improved condi-
tions, the HI catalyzed proto-deiodination pathway of 2 to 3
was minimized and 2 was obtained in a higher yield and
with an improved impurity profile. Accordingly, using the
TMSCl/NaI system in acetonitrile at -15 °C (Table 1, entry
3), the formation of 2 was more favorable although the
reaction took a long time to reach completion. However, the
reaction time could be shortened by using commercially
available TMSI (4.4 equiv) instead of NaI/TMSCl to
overcome the poor solubility of NaI and accelerate the
reaction (Table 1, entry 4). Also, the amount of water was
adjusted (2.8 equiv) once again to maximize the yield and
minimize the formation of impurities. Hence, although the
reaction still took a considerably long time to reach comple-
tion (40 h), it was reproducible and robust. Accordingly,
under the above conditions the 2:3:9:6 product distribution
was a much improved 86:5:1:8. Most importantly, the
improved impurity profile of the reaction now allowed for
the isolation of iodide 2 by crystallization in 82% yield and
>98% purity eliminating the need for silica gel chromatog-
raphy. This third generation process was then successfully
implemented at our pilot plant for the synthesis of multi-
kilogram amounts of 2.
to drive the reaction to completion, minimize the formation
of 3, eliminate the need for chromatography, and make the
reaction conditions more suitable for pilot-plant implementa-
tion. Along these lines, some improvements were obtained
by reversing the order of addition and charging a dichloro-
methane solution of phosphate 9 into a premade mixture of
TMSCl and NaI in dichloromethane. Also, the amounts of
TMSCl and NaI were increased to 4 equiv as opposed to 3,
and the amount of water was increased to 2 equiv relative
to 9. In addition, powdered NaI was utilized to minimize
solubility issues associated with particle size. Under these
modified conditions, the yield and product distribution did
not change significantly, but the time it took for the reaction
to reach the above product distribution was consistent (Table
1, entry 2). However, the reaction was much faster and had
to be quenched after approximately 20 min at ambient
temperature to avoid excessive formation of 3. As a result,
it was difficult to monitor the reaction progress, but the
reaction time, product ratios, and yield were reproducible,
permitting the implementation of these conditions in fixed
equipment at the pilot plant for the synthesis of multi-
kilogram amounts of 2.
Conclusion
To summarize, based on our proposed mechanism for the
formation of 2 from 9 (Scheme 3), we developed a safe,
robust, chromatography-free, and reproducible process for
the multi-kilogram synthesis of vinyl iodide 2 at the pilot
plant. Implementation of the above process allowed us to
support the drug substance needs of toxicology and formula-
tion in the LFA-1 program.
Experimental Section
Unless otherwise specified, all reactions were carried out
in oven-dried glassware under an atmosphere of nitrogen.
NMR spectra were recorded on a Bruker DPX-400 NMR
spectrometer. Shifts are reported in ppm relative to trimethyl-
silane; coupling constants (J) are reported in hertz, refer to
apparent peak multiplicities, and may not necessarily be true
coupling constants. The commercially available starting
materials were used as received without further purification,
and all solvents were dried by standard methods prior to use.
All melting points are recorded using a Fisher-Johns melting
point apparatus and are uncorrected.
3-(4-Bromobenzyl)-1-(3,5-dichlorophenyl)-5-iodo-3-
methyl-1H-imidazo[1,2-a]imidazol-2-one (2): Phosphate 9
(45.7 g. 78.7 mmol), MeCN (240 mL), and water (3.97 mL,
221 mmol) were charged sequentially into a 1 L flask, and
the solution was cooled to -15 °C. TMSI (49.3 mL, 347
mmol) was added slowly over 30 min maintaining the
internal temperature below -7 °C. The resulting yellow
suspension was stirred for 40 h. The mixture was diluted
with toluene (125 mL) maintaining the temperature at -7
°C, and the resulting cold suspension was quenched with 5
M NaOH (63 mL). The organic layer was concentrated by
We also investigated the use of commercially available
aqueous solutions of HI; however, the use of these solutions
gave poor results resulting in the formation of mostly
unreacted 9 and lactam 6, which most likely forms from the
hydrolysis of 9.
Vol. 9, No. 2, 2005 / Organic Process Research & Development
•
139

distillation, and additional toluene (315 mL) was charged
followed by aqueous 0.5 M Na2SO3 (188 mL). The aqueous
layer was discarded, and the organic layer was washed with
aqueous 2% NaCl (250 mL). The toluene solution was
filtered through a pad of layered Celite 545 (26 g), silica
gel 60 (83 g) and active carbon (Darco G-60, 10.2 g), and
the cake washed with additional toluene (2 × 125 mL). The
combined toluene layer was concentrated under reduced
pressure to an oil and azeotroped with IPA (2 × 125 mL) to
remove residual toluene. The residue was heated to 75 °C
and diluted with IPA (70 mL) to afford a 34% w/w solution
of 2. The solution was cooled to room temperature (4 °C/
hour), diluted with water (38 mL), and stirred for 2 h. The
solid product was collected by filtration, washed with 40%
w/w aqueous 2-propanol, and dried to afford 37.3 g of 2
N, 9.00. Found: C, 48.89; H, 3.02; N, 8.81. HRMS
(APCI): calcd for C19H15N3O2Cl2Br, 465.9719; found,
467.9703.
[4-(4-Bromobenzyl)-1-(3,5-dichlorophenyl)-4-methyl-
5-oxo-imidazolidin-2-ylideneamino]acetic Acid Ethyl Es-
ter (8): Ethyl isocyanatoacetate (0.287 mL, 2.56 mmol) was
added dropwise to a stirred solution of 7 (1.0 g, 2.49 mmol)
and dichloromethane (5 mL) at room temperature. The
mixture was stirred for 10 min at room temperature, and the
urea intermediate formed as a white precipitate. Triphenyl-
phosphine (1.31 g, 4.98 mmol), triethylamine (0.69 mL, 4.98
mmol), and carbon tetrachloride (0.48 mL, 4.98 mmol) were
added to the stirred suspension. The mixture was then stirred
at ambient temperature for 12 h. Aqueous workup (1 N HCl,
dichloromethane, MgSO4) afforded a yellow oil. Flash
chromatography (silica gel, 4:1 hexanes/ethyl acetate v/v)
afforded 906 mg (71%) of product as a white solid: mp
1
(82%) as an off-white solid (>95% purity). H NMR: δ
7.54 (d, J ) 1.8, 2H), 7.24-7.29 (m, 3H), 6.96 (s, 1H), 6.78
(d, J ) 8.4, 2H), 3.53 (d, J ) 14, 1H), 3.24 (d, J ) 14, 1H),
1.92 (s, 3H). 13C NMR: δ 174.3, 148.8, 135.8, 135.4, 134.2,
132.0, 131.8, 131.0, 127.4, 122.2, 120.6, 68.4, 56.7, 42.0,
22.2. HRMS (APCI): calcd for C19H14N3OCl2BrI, 575.8736;
found, 575.8742.
1
103-105 °C; H NMR (400 MHz, CDCl3) δ 1.31 (t, J )
7.1 Hz, 3H), 1.52 (s, 3H), 2.95 (d, J ) 12.9 Hz, 1H), 2.98
(d, J ) 12.9 Hz, 1H), 4.05-4.13 (m, 3H), 4.23 (m, 2H),
6.57 (d, J ) 1.6 Hz, 2H), 7.04 (d, J ) 8.2 Hz, 2H), 7.37 (m,
3H); 13C NMR (CDCl3, 100 MHz) δ 14.1, 23.7, 42.9, 44.2,
61.7, 70.4, 120.9, 125.6, 129.4, 130.8, 131.9, 133.2, 134.8,
136.1, 151.1, 169.6, 181.5. Anal. Calcd for C21H20BrCl2N3O3:
C, 49.15; H, 3.93; N, 8.19. Found: C, 49.46; H, 3.92; N,
7.96.
Phosphoric Acid 5-(4-Bromobenzyl)-7-(3,5-dichloro-
phenyl)-5-methyl-6-oxo-6,7-dihydro-5H-imidazo[1,2-a]-
imidazol-3-yl Ester Diethyl Ester (9): Potassium bis-
(trimethylsilyl)amide (265 mL of a 0.5 M solution in toluene,
133 mmol) was added dropwise to a stirred solution of 6
(51.5 g, 110.3 mmol) and diethyl chlorophosphate (23.9 mL,
165 mmol) at -20 °C. The mixture was stirred at -20 °C
for 1 h. Aqueous workup (aqueous NH4Cl, ethyl acetate,
MgSO4) afforded an oil. Flash chromatography (silica gel,
hexanes/ethyl acetate 2:1 v/v) afforded 61.2 g (92%) of
product 9 as a yellow oil: 1H NMR (400 MHz, CDCl3) δ
1.44 (t, J ) 7.1 Hz, 6H), 1.86 (s, 3H), 3.26 (d, J ) 13.9
Hz), 3.34 (d, J ) 13.9 Hz, 1H), 4.33 (m, 4 H), 6.50 (s, 1H),
6.84 (d, J ) 8.2 Hz, 2H), 7.24-7.28 (m, 3H), 7.58 (d, J )
1.6 Hz, 2H).
3-(4-Bromobenzyl)-1-(3,5-dichlorophenyl)-3-methyl-
1,6-dehydro-imidazo[1,2-a]imidazole-2,5-dione (6): Method
A: Toluene (450 mL) was added to 76.9 g of a mixture of
ester 8 (47.1 g, 91.7 mmol) and triphenylphosphine oxide
(29.2 g, 105 mmol), and the resulting solution was cooled
to -10 °C. Trimethylaluminum (46 mL of a 2 M solution
in toluene, 92 mmol) was added dropwise keeping the
temperature at or below 0 °C, and the mixture was then
allowed to reach ambient temperature. The mixture was
stirred at ambient temperature for 2 h, and more trimethyl-
aluminum (27.6 mL of a 2 M solution in toluene, 55.2 mmol)
was added in two portions at 2 h intervals. The mixture was
placed over an ice bath and slowly quenched with 1 N HCl
(360 mL). The organic portion was separated, and the
aqueous portion was extracted with toluene (200 mL). The
combined organic portions were washed with water and
concentrated under reduced pressure to afford an orange oil.
Flash chromatography (silica gel, hexanes/ethyl acetete 4:1
v/v) afforded 38.1 g (89%) of product as an oil that solidified
upon standing: mp 52-54 °C; 1H NMR (400 MHz, CDCl3)
δ 1.84 (s, 3H), 3.24 (d, J ) 13.8 Hz, 1H), 3.43 (d, J ) 13.8
Hz, 1H), 4.18 (d, J ) 21.9 Hz, 1H), 4.30 (d, J ) 21.9 Hz,
1H), 6.95 (d, J ) 8.3 Hz, 2 H), 7.29 (d, J ) 1.8 Hz, 2H),
7.33 (t, J ) 1.8 Hz, 1H), 7.38 (d, J ) 8.3 Hz, 2H); 13C
NMR (100 MHz, CDCl3) δ 21.5, 40.8, 61.3, 65.1, 122.3,
122.6, 128.5, 131.0, 132.0, 132.5, 132.7, 135.5, 154.6, 174.3,
174.9. Anal. Calcd for C19H14BrCl2N3O2: C, 48.85; H, 3.02;
Acknowledgment
The authors wish to thank Dr. V. Farina and Professors
Scott Denmark and Peter Wipf for insightful comments and
suggestions.
Received for review October 12, 2004.
OP049813O
140
•
Vol. 9, No. 2, 2005 / Organic Process Research & Development