An efficient process for the large-scale synthesis of a 2,3,6-trisubstituted indole

Article abstract of DOI:10.1021/op300303p

The efficient synthesis of a key trisubstituted indole intermediate 1 is described. The synthetic route required the use of an aryl Grignard reagent which was not commercially available, and the large-scale formation of this fragment and the thermal evalu

Full text of DOI:10.1021/op300303p

Article
pubs.acs.org/OPRD
An Efficient Process for the Large-Scale Synthesis of a 2,3,6-
Trisubstituted Indole
Anthony D. Alorati,* Andrew D. Gibb,* Peter R. Mullens, and Gavin W. Stewart
Global Process Chemistry, Merck Sharp & Dohme Ltd, Hertford Road, Hoddesdon, Hertfordshire EN11 9BU, United Kingdom
ABSTRACT: The efficient synthesis of a key trisubstituted indole intermediate 1 is described. The synthetic route required the
use of an aryl Grignard reagent which was not commercially available, and the large-scale formation of this fragment and the
thermal evaluation for this step is presented. The key step in the sequence was a Truce−Smiles rearrangement to provide an
advanced ketone intermediate which, upon reduction, cyclized to the desired indole 1. Design of experiment (DoE) optimization
of this reduction is also presented. In total >50 kg of target indole 1 were synthesized in 55% overall yield over five steps using
this new route.
INTRODUCTION
RESULTS AND DISCUSSION
■
■
During the course of an ongoing program, 50 kg of a key
intermediate of trisubstituted indole 1 were required. Two
previous multikilogram campaigns had been demonstrated,
employing a three-step sequence to generate known indole 1
(Scheme 1).1 However, the high cost and long lead times of
reactants 2 and 5 meant that this route would not be viable for
large-scale preparation of 1 within the desired time frame.
It was therefore apparent that a new route to 1 was required
that would address these issues. The indole motif is prevalent in
a large number of target molecules, and as such there is
extensive coverage in the literature of synthetic routes towards
these compounds.2,3 Two of the most direct routes were
screened for applicability towards the synthesis of 1. The first
route investigated was the Larock method4,5 via alkyne 7 which
would be expected to provide the desired regiochemical
outcome (Scheme 2).
The synthetic scheme for our approach is shown in Scheme 4.
SNAr reaction to 17 and rearrangement to 18 proved to be both
selective and high yielding under the published DMF/
potassium carbonate conditions. Investigations therefore
focused on the synthesis of the phenolic ketone fragment 8
and the final hydrogenation/ring closure to afford target indole
1.
2-Cyclohexyl-1-(2-methoxy-phenyl)ethanone 15 For-
mation. To expedite the synthesis it was necessary to form the
Grignard reagent of 2-bromoanisole 12 in-house. The potential
for uncontrolled exothermic activity is high unless the system is
well characterised thermally. Thus, thermal evaluation of the
system was carried out using ARSST and RC-1 calorimetry.
Typically an initial charge of halide is limited and a safe and
reliable initiation procedure developed. Thus elemental
magnesium insertion to bromide 12 was initiated via catalytic
charge of DIBAL-H (2.5 mol %).10 Crucially DIBAL-H both
dried the solvent and afforded reliable and rapid activation of
Mg. A more commonly used Mg activator is 1,2-dibromo-
ethane, however this compound exhibited a lag time of up to 1
h at 50 °C. Halogen-metal exchange with 12, using iPrMgCl or
iPrMgCl.LiCl, proved unsuccessful with only partial conversion
achieved even at elevated temperatures. Moreover, any excess
Grignard used to drive the exchange reactions to completion
complicated downstream chemistry.
Optimization of the procedure in an RC-1 (1 L jacketed
flask) showed initiation occurred readily at <40 °C, judged by
appearance and rate of temperature rise (RC-1 run in both
adiabatic and isoperibolic modes). The adiabatic temperature
rise for the system with an initial 10 mol % of bromide 12 (10
vol THF vs Mg) was calculated at 28 °C. The extent of
Grignard 14 formation was readily monitored by sampling,
quenching into water and analyzing by HPLC. For bulk
implementation, the charge of 12 was reduced to an initial 8
mol % and the DIBAL-H charged at 25 °C. The resulting
exotherm raised the batch temperature to approximately 60 °C.
However, synthesis of substituted alkyne starting material 7
on scale would be potentially problematic. There were also
issues with the sourcing of starting materials, and additionally,
the key sp3−sp carbon−carbon bond formation to generate 7
proved to be difficult. When 7 was prepared on lab scale,
however, the Larock cyclisation via intermediate 9 did indeed
give the desired 2-phenoxy regioselectivity in approx 9:1 ratio
1
as determined by H NMR experiments.
The second approach to the preparation of 1 is by formation
of imine 10 using aniline 6 and ketone 8, followed by an
intramolecular Heck reaction. This method is also well-
described in the literature and has been shown to work with
a variety of ortho-substituted halo-anilines.6 Unfortunately,
initial screening of this reaction gave modest conversions at
best, and subsequently, both approaches were discontinued.
Attention was then switched to a recent publication by the
Snape group7 who had demonstrated a novel synthesis of a 2-
substituted indole using a Truce−Smiles rearrangement8,9 as a
key step (Scheme 3). Initial lab-scale experiments showed this
route to be promising on gram quantities, so attention was
focused towards optimization and scale-up of this chemistry.
Herein we report the successful preparation of >50 kg of 1 via
this type of approach.
Received: October 25, 2012
Published: November 15, 2012
© 2012 American Chemical Society
1947
dx.doi.org/10.1021/op300303p | Org. Process Res. Dev. 2012, 16, 1947−1952

Organic Process Research & Development
Article
Scheme 1. Synthesis used in first deliveries for 1
Scheme 2. Direct approaches to 2,3,6-trisubstituted indoles
Scheme 3. General scheme for the Snape approach to indoles
The remaining bromide 12 charge in THF solution was added
to the batch at such a rate as to maintain the temperature
between 50 and 60 °C. In a single batch 60 kg of 12 was
processed to generate Grignard 14 (>98% conversion) and
used directly in the next step.
Commercially available cyclohexylacetic acid 11 was
converted to the corresponding Weinreb amide11 13 by
formation of the CDI adduct and addition of N,O-
dimethylhydroxylamine hydrochloride. Complete conversion
to 13 was achieved using 1.3 equiv CDI and 1.5 equiv
hydroxylamine. Alternative and potentially cheaper method-
ologies, e.g. addition of the Grignard to the ester or the acid
chloride, were examined but were not competitive within the
short development time available. Initially the THF solution of
11 was added to a cooled slurry of CDI in THF. This in turn
was added to a slurry of N,O-dimethylhydroxylamine hydro-
chloride in THF and aged overnight. Addition of triethylamine
as a base gave a more rapid and reliable conversion, eliminating
variable reaction times which resulted from low solubility of
N,O-dimethylhydroxylamine hydrochloride in THF. An MTBE
extraction and aqueous wash sequence then afforded a solution
of 13 in 92% assay yield.
Upon addition of Grignard 14 to the solution of 13, >98%
conversion to desired ketone 15 was achieved. A maximum
charge of 1.2 mol equiv of 14 was employed and any excess was
minimized to reduce downstream anisole levels. Quench of this
1948
dx.doi.org/10.1021/op300303p | Org. Process Res. Dev. 2012, 16, 1947−1952

Organic Process Research & Development
Article
Scheme 4. Synthetic route via Truce-Smiles rearrangement to 2,3,6-trisubstitued indole 1
Scheme 5. Fate of residual THF through synthetic sequence
reaction mixture with aqueous HCl and addition of heptane (1
vol.) afforded clear separations during the work up. After a
wash with aqueous sodium hydrogen carbonate, the organic
phase assayed for an overall 93% yield of 15 (from 11).
2-Cyclohexyl-1-(2-hydroxy-phenyl)ethanone 8 forma-
tion. A comprehensive screen of demethylation conditions for
conversion of 15 to 8 was undertaken incorporating a variety of
solvents, Lewis and Brϕnsted acids. The use of common iodide
(NaI and LiI) and bromide (LiBr and KBr) sources with
dipolar aprotic and high boiling protic solvents largely proved
ineffective up to 120 °C. Utilizing MgI2 (0.6 equiv) in THF at
50 to 60 °C gave complete and clean deprotection of 15 in <4
h, with the coordinating ortho methoxy group accelerating the
demethylation. However, the bulk availability of this reagent
was limited and cost prohibitive.
Ultimately the screen identified HBr as a reagent for
additional optimization. Thus, the final method developed for
demethylation used 2 equiv of concentrated aqueous HBr in 3
vol of methane-sulfonic acid. With the reaction at 70 °C,
greater than 98% conversion was achieved in <3 h. Additionally
no significant degradation was observed in reaction mixtures
aged at 90 °C overnight. However it was demonstrated that
significantly increased quantities of HBr or additional water in
the reaction can result in a build up of aldol-type impurities.
As a consequence of demethylation on bulk scale, 300 mol of
methyl bromide gas (bp 4 °C) were produced. Due to its
extensive use as a fumigating and agricultural chemical,
considerable literature exists relating to its handling and
disposal. Scrubbing the headspace/exhaust with 20% aqueous
ethanolamine has been well-defined by the Pfizer group12 and
this was employed.
The major impurities in 8 were phenol and 2-methoxyphenol
derived from the original Grignard charge but neither of these
compounds proved problematic in down stream chemistry.
However, residual THF in the methoxy ketone input stream
proved to be an issue. Ring-opening of residual THF under
demethylation conditions afforded dibromobutane (Scheme 5).
This in turn quantitatively alkylated the subsequent Smiles
rearranged product 18 to form 19, necessitating the control of
THF to <2 mol % by distillation prior to demethylation.
4-[1-Cyclohexyl-2-(2-hydroxy-phenyl)-2-oxo-ethyl]-3-
nitrobenzoic Acid Methyl Ester 19. Formation of aryl ether
17 proceeded via SNAr reaction of phenol 8 and trisubstituted
fluoro aromatic 16 followed by an in situ Truce−Smiles
rearrangement to form nitro phenol 18. Initial conditions
screened for formation of 17 and 18 centered around those
reported7 and included solvents (DMF, DMAc, DMSO),
temperature, concentration, type and equivalents of base
1949
dx.doi.org/10.1021/op300303p | Org. Process Res. Dev. 2012, 16, 1947−1952

Organic Process Research & Development
Article
Scheme 6. Indole and hydroxyl indole via reductive ring closure
(potassium carbonate and potassium tert-butoxide). Superior
yields (91% assay yield) were obtained using 10 vol of either
DMF or DMSO with an excess of potassium carbonate at room
temperature (typical reaction time was 3−4 h). For benchtop
screening on milliliter scale, magnetic stirring was employed.
Employing batches of potassium carbonate from different
bulk suppliers led to a variation in reaction profile and assay
yield. Particle size measurements revealed no significant
difference in primary particle size, 20 μm, or agglomerate size
(100−150 μm). However, batches that contained soft
agglomerates that readily broke up in the reaction conditions,
performed well, whereas the presence of hard agglomerates led
to stalled reactions that failed to rearrange to 18. To break
down these agglomerates and potentially increase the surface
area heat/cool cycles of potassium carbonate in DMF were
attempted before reagent addition but were unsuccessful.
Finally, we resorted to in-house pin milling to ensure a supply
of nonagglomerated base which led to high assay yields. On
scale, the SNAr13 reaction was complete after 2.5 h at 23 °C,
and the product fully rearranged after aging overnight at 37 °C
(87% assay yield). Product 18 was isolated after an extractive
workup to furnish a solution stream in THF or MTBE for the
next step.
improve solution stream performance with carbon treatment or
exposure to silica were unsuccessful.
In an effort to drive reactions to completion, fit-for-purpose
DoE designs, two-level 1/2 factorial, were run to examine the
following critical factors: temperature (40−75 °C), pressure
(10−70 psi), quantity of acetic acid (0−50% by vol) and
catalyst loading (10−20 wt %). A working model was obtained
that gave an excellent correlation between the predicted and
actual results (Figure 1). The DoE indicated the prominent
The cheaper methyl-4-chloro-3-nitrobenzoate analogue of 16
was briefly investigated but led to a poor purity profile and
moderate yield. With 16, the SNAr step proceeded rapidly, and
the one-pot reaction behaved as a distinct two-stage sequential
reaction. However, with the methyl-4-chloro-3-nitrobenzoate,
the SNAr and rearrangement steps proceeded in parallel, leading
to increased impurities and crossover reactions.
3-Cyclohexyl-2-(2-hydroxy-phenyl)-1H-indole-6-car-
boxylic Acid Methyl Ester 1. Preliminary experimentation
highlighted the formation of N-hydroxy indole 20 (Scheme 6),
formed via ring closure of the intermediate hydroxylamine, as a
significant problem and resulted in stalled reactions and poor
purity profiles. Hence, a screen of appropriate solvents and
hydrogenation catalysts was performed to identify optimal
conditions. Using an Endeavor hydrogenation apparatus, a
screen was carried out followed by DoE analysis to rapidly
assess critical parameters once a hit was established. During this
development it became apparent that the purity of 18 greatly
impacted the reaction profile.
Initial hydrogenation screens using the crude reaction
streams gave poor results (<1:3 ratio of 1:20), with type 91
Pd on carbon (10% Pd, 50% wet) in a range of common
organic solvents. A broad screen of other Pt and Pd catalysts
also gave gross mixtures of 1 and 20 with minimal additional
conversion to 1 after resubjecting the reaction mixtures to fresh
catalyst. Continued screening on chromatographed 18
identified acetic acid as a key additive to ensure >95%
conversion to 1. However, replicating this with crude input
streams gave, at best, 50% conversion to desired 1. Attempts to
Figure 1. DX6-generated schematic for actual vs model-predicted
conversion to indole.
effects to be high levels of acetic acid and higher temperatures
to achieve good conversion to 1. Higher pressures (>50 psi)
generally decreased conversion to 1, although this was
complicated with other two-factor interactions. By incorporat-
ing just the main effects from the fit-for-purpose design, the
final bulk conditions were derived by reducing volumes for
throughput and considering the boiling point of THF to be <60
°C. Thus, final conditions were 35 psi, 60 °C, 20 wt % catalyst
(type 91 10% Pd/C) with 50% acetic acid in THF (5 vol).
Three batches on approximately 25-kg scale each afforded
>50:1 desired indole 1 vs hydroxyindole 20, in 85% assay yield.
On completion, the catalyst was removed by filtration and the
indole crystallized by concentration followed by addition of n-
heptane to minimize liquor losses. In total, 52.5 kg of indole 1
was isolated by filtration as a white solid. Three batches gave an
1950
dx.doi.org/10.1021/op300303p | Org. Process Res. Dev. 2012, 16, 1947−1952

Organic Process Research & Development
Article
average isolated yield of 77% in high purity with mother liquor
losses between 5 and 8%.
50 °C, and sampled by HPLC to confirm complete
consumption of bromoanisole. The Grignard solution was
filtered and transferred into a separate vessel, rinsed with THF
(55 kg), and held at ∼35 °C overnight. In total, 67.8 kg of 12
was made in 99% yield as a 46% w/w THF solution.
CONCLUSION
■
A practical, scalable, and high-yielding synthesis of target indole
1 has been developed, starting from commercially available raw
materials. This concise and efficient approach has been
demonstrated on >50 kg scale, and key safety issues have
been addressed to afford indole 1 in 55% overall yield in seven
chemical transformations. The entire sequence can be tele-
scoped, significantly reducing cycle times, and extends the
scope of the Truce−Smiles rearrangement.
2-Cyclohexyl-1-(2-methoxy-phenyl)ethanone (15). To
a vessel containing the MTBE solution of 13 was charged 1.1
equiv of the Grignard solution 14 (wrt to 13) over 40 min,
maintaining internal temperature between 20 and 25 °C. A
second charge of ∼0.1 equiv Grignard 14 was sufficient to
achieve >97% conversion. The batch was cooled to 5 °C and
carefully quenched with 2 M HCl (225 L), keeping the reaction
temperature <25 °C. Heptane (41 kg) was added and agitated
for 10 min, and the two phases settled. The lower aqueous layer
was separated, and the organics were washed with a 3.5%
NaHCO3 (185 kg) and a final water wash (180 kg). The batch
was distilled from 480 to 140 L to solvent switch into heptane
and dry the batch. In total 55.9 kg of product was isolated in
99% yield as a 45% w/w heptane solution (55.9 kg of 15 in
total of 124.0 kg) and a water content of 10.8 μg water in 100
EXPERIMENTAL SECTION
■
General. Starting materials were obtained from commercial
suppliers and were used without further purification. HPLC
analyses were performed on Agilent series 1100 using C-18
reverse phase methods eluting acetonitrile and 0.1% H3PO4
(aq). Assay yields were obtained using analytical standards
obtained by chromatography or recrystallisation. Isolated yields
refer to yields corrected for purity on the basis of HPLC assays
using purified standards. NMR spectra were obtained at 400
MHz for 1H and 100.6 MHz for 13C. All coupling constants are
reported in Hertz (Hz).
1
μL (Karl−Fischer titration). H NMR (400 MHz, CDCl3) δ
7.65 (1H, d, 8.0 Hz), 7.45 (1H, t, 9.0 Hz), 7.0 (1H, t, 8.0 Hz),
6.95 (1H, d, 7.5 Hz), 3.90 (3H, s), 2.85 (2H,d, 7.0 Hz), 1.95−
1.88 (1H, m), 1.75−1.62 (5H, m), 1.32−1.16 (3H, m), 1.05−
0.94 (2H, m). 13C NMR (400 MHz, CDCl3) δ 203.1, 158.6,
132.9, 129.9, 129.3, 120.63, 111.4, 55.4, 51.3, 34.3, 33.3, 26.9,
26.34. HRMS (ES) Calcd for C15H21O2 (MH+) 233.1542.
Found 233.1535.
2-Cyclohexyl-N-methoxy-N-methylacetamide (13). A
slurry of CDI (54.5 kg) in THF (98.1 kg) was cooled to 5 °C.
A solution of cyclohexylacetic acid (36.8 kg) in THF (32.7 kg)
was added over 75 min, maintaining the internal temperature
between 5 and 7 °C. The reaction mixture was then aged for 15
min at 7 °C and conversion to CDI adduct monitored via a
methanol/triethylamine solution quench. In a separate vessel a
slurry of N,O-dimethylhydroxylamine hydrochloride (37.9 kg)
in THF (65.4 kg) was made. The CDI adduct solution was
added over 22 min; the internal temperature was kept below
below 25 °C. The reaction was aged at 23 °C overnight. After
sampling, the reaction was incomplete, and three portions of
Et3N (14.0 kg) were added over 7.5 h. The batch was aged
overnight at 25 °C to afford complete conversion to the
Weinreb amide. The reaction was quenched via addition of
TBME (136.3 kg), heptane (25.2 kg), and 2 M HCl solution
(184 L). The lower aqueous layer was cut, and the organics
were washed with a 2 M HCl solution (74 L). The organics
were then washed with a 3.5% NaHCO3 solution (74 L). The
organics were diluted with heptane (25.2 kg) and afforded a
final water wash (74 kg) before being dried via distillation to a
final batch volume of 187 L and a water content of 62 μg water
in 100 μL (Karl−Fischer titration). In total 44.2 kg of product
was isolated in 92% yield as a 30% w/w MtBE solution (149.8
kg).
(2-Methoxylphenyl)magnesium Bromide (14). Magne-
sium (8.6 kg) and THF (77.0 kg) were charged to a clean dry
vessel. The agitator was started with minimum agitation. The
contents of the vessel were heated to 30 °C, and DIBAL-H, 1
M in heptane (0.8 L), was charged to the batch followed by an
initial charge of the 2-bromoanisole (4.0 kg). Almost
immediately an exotherm was observed, indicating reaction
initiation. The batch was cautiously warmed to 50−55 °C and
sampled by HPLC to confirm complete consumption of 2-
bromoanisole. A solution of 2-bromoanisole (56.0 kg) in THF
(160 kg) was slowly added over 100 min, maintaining the
internal temperature vessel batch above 60 °C but not at an
aggressive reflux. The batch was aged for 20 min, cooled to 40−
2-Cyclohexyl-1-(2-hydroxy-phenyl)ethanone (8). To a
45% w/w solution of 2-cyclohexyl-1-(2-methoxy-phenyl)-
ethanone 15 in heptane (61 kg, 120 mol) was added
methane-sulfonic acid (123 kg). The mixture was heated to
50 °C at <500 mbar to remove residual solvent by distillation.
At atmospheric pressure concentrated aqueous HBr was added
over at least 30 min in a temperature range of 65−75 °C. After
aging for 150 min the reaction was cooled to 20 °C, and a
vacuum purge cycle removed any residual methyl bromide. In a
separate vessel a mixture of MTBE (83 kg), heptane (19 kg),
and water (111 kg) was charged, followed by the reaction
mixture added at a rate to maintain the temperature below 30
°C. The lower acidic aqueous layer was separated and the
organic phase washed sequentially with water (84 kg), 5% w/v
sodium bicarbonate solution (90 kg), and finally water (84 kg).
The organic phase was distilled to dry, using heptane as
necessary to achieve a final residual water content of <10 mol %
vs substrate. Afforded 52.4 kg of 8 as an orange solution in
1
heptane in 98% assay yield. H NMR (400 MHz, CDCl3) δ
12.49 (1H, s) 7.77 (1H, d, 8.0 Hz), 7.45 (1H, t, 7.5 Hz), 7.00
(1H, d, 8.0 Hz), 6.90 (1H, t, 7.5 Hz), 2.85 (2H, d, 7.0 Hz),
1.96−1.90 (1H, m), 1.78−1.65 (5H, m), 1.31−1.24 (3H, m),
1.17−1.05 (2H, m). 13C NMR (400 MHz, CDCl3) δ 206.7,
162.6, 136.2, 130.2, 119.7, 118.7, 118.5, 45.9, 34.9, 33.4, 26.18,
26.12. HRMS (ES) Calcd for C14H19O2 (MH+) 219.1385.
Found 219.1382.
4-[1-Cyclohexyl-2-(2-hydroxy-phenyl)-2-oxo-ethyl]-3-
nitrobenzoic Acid Methyl Ester (18). To a suspension of
potassium carbonate (34 kg) and methyl-4-fluoro-3-nitro-
benzoate 16 (22 kg), in DMF (158 kg) was added a solution
of 8 (24.4 kg) in DMF (20 kg) over 30 min. This mixture was
aged at RT for approx 60 min, then water (2.1 kg) and
additional potassium carbonate (15.4 kg) were added, and the
resulting mixture was aged at 35 °C overnight. In a separate
vessel a quench mixture of water (130 kg), conc HCl (40 kg),
1951
dx.doi.org/10.1021/op300303p | Org. Process Res. Dev. 2012, 16, 1947−1952

Organic Process Research & Development
Article
Martin, S. W.; Gentles, R. G. PCT Int. Appl. WO/2007/092000, 2007.
(f) Apito, E.; Habermann, J.; Narjes, F.; Rico Ferreira, M. R.;
Stansfield, I. PCT Int. Appl. WO/2007/129119, 2007 (g) Stansfield,
I.; Koch, U.; Habermann, J.; Narjes, F. PCT Int. Appl. WO/2008/
075103, 2008. (h) Conte, I.; Haberman, J.; Mackay, A.; Narjes, F.;
Ricco Ferreira, M. R.; Stansfield, I. UK Pat. Appl. GB 2451184 A,
2009. (i) Conte, I.; Habermann, J.; Mackay, A.; Narjes, F.; Rico
Ferreira, M. R.; Stansfield, I. U.S. Pat. Appl. Publ. US 2009048239,
2009.
(2) Recent reviews of indole syntheses: (a) Beller, M. Adv. Synth.
Catal. 2008, 350, 2153. (b) Humphrey, G.; Kuethe, J. Chem. Rev.
2006, 106 (7), 2875.
(3) Stuart, D.; Bertrand-Laperie, M.; Burgess, K.; Fagnou, K. J. Am.
Chem. Soc. 2008, 49 (130), 16474.
(4) Roschanger, F.; Liu, J.; Estanove, E.; Dufor, M.; Rodriguez, S.;
Farina, V.; Hickey, E.; Hossain, A.; Jones, P.-J.; Lee, H.; Lu, B. Z.;
Varsolona, R.; Schroder, J.; Beaulieu, P.; Gillard, J.; Senanayake, C. H.
Tetrahedron Lett. 2008, 49, 363.
(5) Zeni, G.; Larock, R. Chem. Rev. 2004, 104, 2285.
(6) Knapp, J. M.; Zhu, J. S.; Tantillo, D. J.; Kurth, M. J. Angew. Chem.,
Int. Ed. 2004, 43, 4526.
(7) Snape, T. J. Synlett 2008, 17, 2689.
(8) Truce, W. E.; Ray, W. J.; Norman, O. L.; Eickemeyer, D. B. J. Am.
Chem. Soc. 1958, 80, 3625.
(9) (a) Kimbaris, A.; Cobb, J.; Tsakonas, G.; Varvounis, G.
Tetrahedron 2004, 60, 8807. (b) Mitchell, L. H.; Barvian, N. C.
Tetrahedron Lett. 2004, 45, 5669. (c) Randal, E. W.; McKennon, M. J.
Tetrahedron Lett. 2000, 41, 4541. (d) Padwa, A.; Filipkowski, M. A.;
Kline, D. N.; Murphree, S. S.; Yeske, P. E. J. Org. Chem. 1993, 58,
2061. (e) Madaj, E. J.; Snyder, D. M.; Truce, W. E. J. Am. Chem. Soc.
1986, 108, 3466. (f) Wubbels, G. G.; Halverson, A. M.; Oxman, J. D. J.
Am. Chem. Soc. 1980, 102, 4848. (g) Snyder, D. M.; Truce, W. E. J.
Am. Chem. Soc. 1979, 101, 5432. (h) Truce, W. E.; Brand, W. W. J.
Org. Chem. 1970, 35, 1828. (i) Crowther, G. P.; Hauser, C. R. J. Org.
Chem. 1968, 33, 2228.
(10) Tilstam, U.; Weinmann, H. Org. Process Res. Dev. 2002, 6, 906.
(11) Miao., C. K.; Sorcek., R.; Jones, P.-J. Tetrahedron Lett. 1993, 34,
2259.
(12) Hettenbach, K.; am Ende, D. J.; Leeman, K.; Dias, E.;
Kasthurikrishnan, N.; Brenek, S. J.; Ahlijanian, P. Org. Process Res. Dev.
2002, 6, 407.
(13) It was possible to isolate 17 via water-induced precipitation from
the DMF reaction mixture, whereas the rearranged product 18 was
only ever observed as an oil. Although an isolation and purification at
this point may have been beneficial, the two-step “one-pot” process
was run for convenience and throughput.
(14) Narjes, F.; Crescenzi, B.; Ferrara, M.; Habermann, J.; Colarusso,
S.; Ferreira, M. R. R.; Stansfield, I.; Mackay, A. C.; Conte, I.; Ercolani,
C.; Zaramella, S.; Palumbi, M.-C.; Meuleman, P.; Leroux-Roels, G.;
Giuliano, C.; Fiore, F.; Di Marco, S.; Baiocco, P.; Koch, U.; Migliaccio,
G.; Altamura, S.; Laufer, R.; De Francesco, R.; Rowley., M. J. Med.
Chem. 2011, 54, 289.
and MTBE (50 kg) was charged. The reaction mixture was
added to the quench mixture over approx 1 h and the bottom
aqueous phase separated. The main organic phase was washed
with 1 M HCl (50 kg) and water (50 kg) to remove residual
DMF. The solution was concentrated and solvent exchanged to
afford 79.2 kg of 18 as a 42.5% w/w solution in THF in 87%
assay yield.
Methyl 4-(2-(2-cyclohexylacetyl)phenoxy)-3-nitroben-
1
zoate methyl (17): H NMR (400 MHz, CDCl3) δ 8.63
(1H, d, 3 Hz), 8.11 (1H, d, 8 Hz), 7.76 (1H, d, 7 Hz), 7.53
(1H, t, 6 Hz), 7.37 (1H, t, 6 Hz), 7.04 (1H, d, 8 Hz), 6.87 (1H,
d, 9 Hz), 3.97 (3H, s), 2.79 (2H, d, 7 Hz), 1.91−1.82 (1H, m),
1.61−1.58 (5H, m), 1.27−1.13 (3H, m), 0.96−0.83 (2H, m).
13C NMR (400 MHz, CDCl3) δ 200.9, 164.5, 154.1, 151.9,
140.0, 135.1, 133.3, 132.6, 130.5, 127.5, 126.1, 125.0, 121.1,
118.3, 52.6, 50.5, 34.0, 33.0, 26.1, 26.0. HRMS (ES) Calcd for
C22H24NO6 (MH+) 398.1604. Found 398.1590.
4-(1-Cyclohexyl-2-(2-hydroxyphenyl)-2-oxoethyl)-3-nitro-
1
benzoate (18): H NMR (400 MHz, CDCl3) δ 12.30 (1H, s),
8.43 (1H, s), 8.21 (1H, d, 9.0 Hz), 7.96−7.90 (2H, m), 7.49
(1H, t, 7.5 Hz), 6.98−6.90 (2H, m), 5.25 (1H, d, 8 Hz), 2.29
(1H, m), 1.86−1.65 (4H, m), 1.29−0.96 (6H, m). 13C NMR
(400 MHz, CDCl3) δ 205.2, 164.6, 163.1, 150.4, 137.2, 136.5,
133.3, 130.5, 130.2, 125.5, 119.9, 119.4, 118.6, 67.9, 52.75, 51.1,
43.3, 32.4, 30.0, 26.1, 26.0, 25.6. HRMS (ES) Calcd for
C22H24NO6 (MH+) 398.1604. Found 398.1607.
3-Cyclohexyl-2-(2-hydroxy-phenyl)-1H-indole-6-carboxylic
Acid Methyl Ester (1). A suspension of 20% Pd/C (5.1 kg, 50%
wet type), THF solution of 18 (59.4 kg), and acetic acid (66
kg) was stirred, inerted, and subjected to hydrogen headspace
pressure of 2.3 bar. The mixture was heated to 60 °C overnight
until hydrogen uptake ceased. The mixture was cooled and
clarified via solka floc. Three batches were combined for
crystallisation and distilled to 140 L. Heptane (550 L) was
charged slowly to crystallise the product and minimise liquor
loss. Indole 1 was isolated by filtration, washed with heptane/
DCM (2 × 40 kg, 1:1 w/w), and then dried at 60 °C in vacuo,
affording 1 as a white, crystalline solid in 74% yield, 98.5 LCAP,
97.0 LCWP. Data are as reported in the literature.14
AUTHOR INFORMATION
Corresponding Author
*Anthony_alorati@merck.com (A.D.A.); Andrew_gibb@
merck.com (A.D.G.).
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank Simon Hamilton for high-resolution MS and Michael
Moore for additional support.
■
REFERENCES
■
(1) (a) Oka, T.; Ikegashira, K.; Hirashima, S.; Yamanaka, H.; Noji, S.;
Niwa, Y.; Matsumoto, Y.; Sato, T.; Ando, I.; Nomura, Y. PCT Int.
Appl. WO/2005/080399, 2005. (b) Hudyma, T. W.; Zheng, X.; He,
F.; Ding, M.; Bergstrom, C. P.; Hewawasam, P.; Martin, S. W.;
Gentles, R. G. PCT Int. Appl. WO/2006/020082, 2006. (c) Conte, I.;
Ercolani, C.; Narjes, F.; Pompei, M.; Rowley, M.; Stansfield, I. PCT
Int. Appl. WO/2006/046030, 2006. (d) Ikegashira, K.; Oka, T.;
Hirashima, S.; Noji, Satoru, Y.; Hiroshi, H.; Adachi, T.i; Tsuruha, J.;
Doi, S.; Hase, Y.; Noguchi, T.; Ando, I.; Ogura, N.; Ikeda, S.;
Hashimoto, H. J. Med. Chem. 2006, 24, 6950. (e) Hudyma, T. W.;
Zheng, X.; He, F.; Ding, M.; Bergstrom, C. P.; Hewawasam, P.;
1952
dx.doi.org/10.1021/op300303p | Org. Process Res. Dev. 2012, 16, 1947−1952