C.A. Dvorak, J. Liang, N.S. Mani et al.
Tetrahedron Letters 67 (2021) 152843
leaving group, such as a tosylate, was also met with failure. A
reductive Beckmann was also attempted on oxime 8 (Dibal-H at
0 °C) however this produced multiple products as was not perused.
This led us to explore synthesizing the seven-membered ring
much earlier in the synthetic sequence. Diazo compounds are
known to react with ketones in a Tiffeneau-Demjanov-type ring
expansion reaction to give a one carbon ring expansion when the
ketone substrate is cyclic [9]. The inserting carbon on the diazo
moiety carries with it any substituents during the formation of
the new N + 1 ring. Therefore, a diazoacetophenone would directly
provide the correct 1,3-diketo intermediate required for the fused
pyrazole system in question. However, this strategy would require
the reaction between diazomethane and the corresponding ben-
zoyl chloride and we wished to avoid the necessity of having to
prepare diazomethane on scale. We consequently chose to exam-
ine the use of ethyl diazoacetate, which is commercially available
and a more stable diazo substrate. The reaction between N-Boc-
4-piperidone and ethyl diazoacetate proceeds smoothly to provide
the corresponding b-keto ester 11 [10]. An examination of 11
reveals that all three carbon atoms required for the pyrazole hete-
rocycle are contained within the b -keto ester subunit. Therefore,
we sought to explore the utility of 11 as a central starting point
in the construction of fused pyrazole-azepine heterocycles
Scheme 3.
To this end, reaction of 11 with a slight excess of hydrazine in
hot ethanol gave pyrazole 12 in excellent yield. Treatment of 12
with benzyl bromide provided a mixture of benzylated compounds
in a ca. 81% overall yield. The mixture was a ~ 1:1 ratio of the N and
O benzylated isomers that were not separable on flash silica gel
chromatography. Gratifyingly, when this mixture was subjected
to N-phenyltriflamide, the desired O-triflate 14 could be isolated
cleanly away from the mixture containing the O benzyl derivative.
No attempts were made to recycle this O-benzylated side product.
The sequence leading to triflate 14 was ultimately improved by
changing the order of triflate formation and benzylation. As shown
in Scheme 4, pyrazole 12 was first treated with N-phenyltriflamide
in pyridine to provide pyrazole triflate 15 in excellent yield. The
resulting reagent byproduct from N-phenyltriflamide after the
transfer of a single trifluorosulfinyl moiety was found to interfere
with subsequent steps. This reagent byproduct could be removed
by simply washing the mixture with a mild base, such as aqueous
potassium carbonate, allowing for the isolation of the desired tri-
flate 15 in sufficient purity to proceed on in the sequence. The alky-
Fig. 1. JNJ-18038683.
Scheme 1. Non-selective Pyrazole Synthesis.
possible enamines were being formed leading to two different
fused pyrazole-azepine isomers. One potential solution would be
to take advantage of a symmetrical ketone to build the pyrazole
prior to a ring expansion event to afford the azepine ring. The Beck-
mann rearrangement of a cyclohexanone oxime to give a caprolac-
tam would be well suited to explore a late stage ring expansion
strategy. There have also been reports of a Dibal-H mediated
reductive Beckmann rearrangement [7] that would allow for the
direct formation of the corresponding secondary amine. To exam-
ine this in detail Scheme 2, 1,4-dioxaspiro[4.5]decan-8-one, a pro-
tected 1,4-dione, was subjected to the identical three step one-pot
sequence to construct the 4-chloroaryl pyrazole in modest yield
affording fused heterocycle 6. Alkylation of 6 with benzyl bromide
produced the fully elaborated fused pyrazole 7 along with traces of
a benzyl regioisomer that was easily removed by flash silica gel
column chromatography. The oxime 8 was then prepared by treat-
ing the ketal with aqueous HCl to remove the protecting group fol-
lowed immediately by the addition of hydroxylamine
hydrochloride to the resulting ketone in a solution of pyridine.
The oxime 8 precipitated directly from the reaction and was thus
isolated by a simple filtration.
lation of the pyrazole heterocycle had already exhibited
a
preference for reactivity at the N1 nitrogen and with the triflate
in place, similar regiochemistry was observed as the benzyl group
was selectively installed on the N1 nitrogen of 15 with benzyl bro-
mide and potassium tert-butoxide to give key intermediate 14.
Subjecting oxime 8 to strong acids (HCl; H2SO4, TFA) to attempt
a Beckmann rearrangement resulted in generating the correspond-
ing fully aromatic aniline or phenol analogue of 9 via a dehydra-
tion/aromatization sequence that is known as the Semmler-Wolff
reaction [8]. The oxime was poorly soluble in most organic solvents
and attempts to convert the oxime hydroxyl group into a suitable
Scheme 2. Synthesis of Oxime 8.
Scheme 3. Selective Synthesis of Key Pyrazole Triflate.
2