Synthesis of Nitrogen Heterocycles via Photochemical Ring Opening of Pyridazine N-Oxides

Article abstract of DOI:10.1021/acs.orglett.6b02562

A photochemical method for the direct synthesis of 1H-pyrazoles from pyridazine N-oxides was developed. This chemistry features a regioselective approach to nonsymmetrically substituted pyridazine N-oxides. Herein, we highlight the first strategic use of photoinduced ring-opening reactions of 1,2-diazine N-oxides for the preparative synthesis of nitrogen heterocycles.

Full text of DOI:10.1021/acs.orglett.6b02562

Letter
pubs.acs.org/OrgLett
Synthesis of Nitrogen Heterocycles via Photochemical Ring Opening
of Pyridazine N‑Oxides
Maribel Portillo, Michael A. Maxwell, and James H. Frederich*
Department of Chemistry and Biochemistry, Florida State University, Tallahassee, Florida 32306, United States
S
* Supporting Information
ABSTRACT: A photochemical method for the direct synthesis of 1H-pyrazoles from pyridazine N-oxides was developed. This
chemistry features a regioselective approach to nonsymmetrically substituted pyridazine N-oxides. Herein, we highlight the first
strategic use of photoinduced ring-opening reactions of 1,2-diazine N-oxides for the preparative synthesis of nitrogen
heterocycles.
fueled interest inexamining thephotochemistry ofheterocycle N-
hemists noted as early as 1901 that oxidation of diazines
Cincreased their sensitivity to UV light. This observation oxides decades later.2 In 1968, Buchardt discovered that
1
photolysis of 3,6-diphenylpyridazine N-oxide (1) afforded a
mixture of the parent pyridazine, 2,5-diphenylfuran (2), and 3-
benzoyl-5-phenylpyrazole (3, Scheme 1).3 This intriguing
reaction proceeds through (Z)-diazoenone A, whichis transiently
generatedfrom1uponexposuretoUVlight. Thus, heterocycles 2
and 3 are formed via secondary reactions of intermediate A.3b
InvestigationsbyBuchardt focusedonthephotochemistry of1.
In this system, rearrangement to A was the main photoreaction,
enabling 3 to be isolated in 75% yield. Conversely, studies by
Ogata and Igeta showed that photolysis of alkylpyridazine N-
oxides resulted in photodeoxygenation.4 This disparate reactivity,
together with a marked absence of methods for the regioselective
oxidation of nonsymmetrically substituted pyridazines, has
hindered the utility of this chemistry. Indeed, to the best of our
knowledge, the Buchardt rearrangement of 1 is the only example
of a productive photoinitiated ring-opening reaction involving
pyridazine N-oxides.5,6
Scheme 1. Photochemistry of Pyridazine N-Oxides
The photochemistry of 1 captured our attention because both
the ring-opening process and the secondary reactions of A appear
well suited for applications in heterocycle synthesis. Further
development of this chemistry, however, required a strategy to
suppress deoxygenation. Encouraged by Sigwalt’s report that
electron-donating substituents slowed the photoreduction of
Received: August 30, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.6b02562
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Table 1. Regioselective Suzuki Cross-Coupling of 3,6-
Dichloropyridazine N-Oxide
Scheme 2. Photochemical Isomerization of 3-Aryl-6-
pyrrolidinylpyridazine N-Oxides
a
a
b
entry
[Pd] (mol %)
time (h)
Ar (product)
Ph (8a)
yield (%)
c
1
2
3
4
5
6
d
d
d
3
3
3
3
3
5
5
5
5
2
93, 88
1
Tol (8b)
99
97
92
94
88
70
72
0.5
2
1-naphthyl (8c)
4-OMeC6H4 (8d)
3,4-benzodioxyl (8e)
4-FC6H4 (8f)
0.5
1
,
e
f
7
8
9
4
2,4-(Cl)2C6H3 (8g)
4-NMe2C6H4 (8h)
4-pyridinyl (8i)
4
,
g
2
73
a
Reaction conditions: 3,6-dichloropyridazine N-oxide (1.00 mmol),
Ar-B(OH)2 (1.1 equiv), aq K3PO4 (2 M, 1.5 mL·mmol−1 halide),
PdCl2dppf (3−5 mol %), THF (0.3 M), 65 °C. Isolated yield after
purification. 20.0 mmol scale. Reaction run in dioxane (0.3 M) at 95
b
c
d
e
f
g
°C. 1.5 equiv Ar-B(OH)2. 1.2 equiv Ar-B(OH)2. 3,6-Dipyridinyl
product isolated in 8% yield.
Table 2. Optimization and Control Experiments for the
Photochemical Isomerization of 9a to 7a
a
Isolated yields are reported and are an average of two experiments.
Furan side product was isolated in 5−9% yield.
b
pyridine N-oxides, we anticipated that perturbing the electronics
of 1 would enable us to retain the N-oxide group.7 A set of
preliminary reactions supported this hypothesis (4 → 5, Scheme
1). Thus, we targeted structure 6 with the expectation that adding
a heteroatom substituent to the pyridazine nucleus would
enhance the desired photochemistry. If successful, this design
would allow us to harness the photoinduced ring opening of 6 in a
general way while concurrently establishing a new route to 1H-
pyrazoles of significant utility (e.g., 7).8,9 Our development of this
method, which features a regioselective approach to substituted
pyridazine N-oxides, is reported in this paper.
The first objective of our investigation was to establish a route
to scaffold 6. As alluded to above, direct oxidation of the
corresponding pyridazines was problematic. In our hands, this
strategygeneratedbothN-oxideregioisomers. Tocircumventthis
obstacle, we focused instead on executing site-selective reactions
on 3,6-dichloropyridazine N-oxide.10 Noncatalyzed SNAr reac-
tionsofthisheterocycleareknown;however, regioselectivitiesare
modest, and site-selectivity is dependent on the structure of the
nucleophile.11 Alternatively, selective cross-coupling reactions of
polychlorinated diazines have been described by Hu, Villemin,
and Buchwald.12,13 Inspired by these studies, we envisioned using
a two-stage process to convert 3,6-dichloropyridazine N-oxide to
6. Accordingly, a regioselective Pd-catalyzed Suzuki reaction
a
b
c
entry
modification
none
ratio 7a:11:12
7a yield (%)
1
28:1:1
no reaction
7:2:1
93
0
2
ambient light
35 °C
THF (0.5 M)
THF (0.1 M)
MeCN
3
65
d
4
20:2:1
51:1:2
ND
69
5
93
e
6
47
7
MeCN (0.1 M)
MeOH
47:0:1
ND
92
e
8
60
9
MeOH (0.1 M)
dioxane (0.1 M)
CH2Cl2
24:2:1
26:4:1
dec
93
81
10
11
ND
f
d
12
hν = 254 nm
hν = 300 nm
11:2:1
8:2:1
40
f
13
67
a
Reactions carried out at 65 °C unless otherwise indicated.
b
c
1
Determined by H NMR of unpurified reaction mixtures. Isolated
d
e
yield after purification. Unreacted 9a observed. Multiple side
f
products observed along with 7a, 11, and 12. Photolysis carried out
in dioxane (0.1 M). ND = not determined.
B
DOI: 10.1021/acs.orglett.6b02562
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
Scheme 3. Scope of Heteroatom Groups at the C6 Position
a
b
c
Isolated yields are reported and are an average of two experiments. 19% of furan side product was observed. Furan side product was formed in 5−
d
e
f
9% yield. Photodeoxygenation was observed. Irradiated at 350 nm for 14 h. The mass balance was starting material.
reacted with unactivated and electron-rich arenes to selectively
Scheme 4. Multigram-Scale Synthesis of hNK3 Receptor
Antagonist 17
a
affordC3-arylated products 8a−einexcellentyield(entries1−5).
Higher catalyst loading (5 mol %) was needed to achieve good
yields using electron-deficient boronic acids (entries 6 and 7).
Nitrogen-containing partners also reacted slowly, although good
yields of 8h−i were obtained using 5 mol % of PdCl2dppf in
dioxane at 95 °C (entries 8 and 9). Notably, the C6-arylated
isomer of 8 was not detected in these reactions.
Having reduced to practice a robust synthesis of 3-aryl-5-
chloropyridazine N-oxides (8), we were positioned to test our
central hypothesis. Accordingly, 8a was reacted with pyrrolidine
and K2CO3 in THF at 65 °C to furnish 3-phenyl-6-
pyrrolidinylpyridazine N-oxide (9a) in 94% yield.18 To our
delight, photolysis of 9a did not result in deoxygenation. Instead,
pyrazole 7a was cleanly formed along with minor amounts of
furan11 andallene 12. ConsistentwiththemechanisminScheme
1, these photoproducts are generated from diazo intermediate 10,
a reactive species we observed by NMR but could not isolate.19
Thus, 11and12likelyariseviaacompetingcarbenepathwayfrom
decomposition of 10 (e.g., A → 2, Scheme 1). Upon further
experimentation, we found that these side products were virtually
eliminatedbyexposingasolutionof9ainTHF(0.3 M)to350nm
UV light at 65 °C for 2 h. This optimized protocol furnished 7a in
93% yield.
Salient results from our optimization studies are summarized in
Table 2. UV light was found to be essential to the reaction (entry
2). In general, increasing the reaction temperature from 35 to 65
°C was required to promote formation of 7a relative to 11 and 12
(entry 3). THF was the most versatile solvent for this reaction;
however, excellent results were achieved by irradiating more
dilute solutions of 9a (0.1 M) in MeOH, MeCN, or dioxane
(entries 6−10). Conversely, CH2Cl2 was incompatible with this
chemistry (entry 11). We also obtained inferior results using
shorter wavelength UV light (entries 12 and 13). Importantly,
a
Reagents and conditions: (a) 4-fluorophenylboronic acid (1.1 equiv),
2 M aq K3PO4, PdCl2dppf (5 mol %), THF (0.3 M), 65 °C, 2 h, 85%;
(b) 15 (2.0 equiv), K2CO3 (3.0 equiv), THF (0.3 M), 65 °C, 24 h,
92%; (c) hv (350 nm), THF (0.3 M), 65 °C, 5 h, 90%.
would first be used to install the desired aryl group, thereby
allowing a range of heteroatom nucleophiles to be added in a
second operation.
With this strategy in mind, we set out to identify a catalyst
system capable of discriminating between the two aryl−chloride
bonds in 3,6-dichloropyridazine N-oxide. Using 1.1 equiv of
phenylboronic acid as a cross-coupling partner, we screened Pd/
phosphine ligand combinations reported for Suzuki reactions of
nitrogen heterocycles.14,15 The majority of catalyst systems we
testedprovidedundesiredmixturesofC3-arylatedproduct8aand
1. The exception was PdCl2dppf, which furnished 8a as the main
product.16 After further exploration of the reaction parameters,
wedeterminedthatcatalyticPdCl2dppf(3mol%)incombination
with aq K3PO4 and THF afforded 8a in 93% yield (Table 1, entry
1).17
As illustrated in Table 1, this protocol was compatible with a
range of arylboronic acids. Thus, 3,6-dichloropyridazine N-oxide
C
DOI: 10.1021/acs.orglett.6b02562
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
these reactions were carried out under ambient atmosphere using
untreated solvents and inexpensive borosilicate glassware.
With optimized reaction conditions in hand, the initial scope of
the photoinduced isomerization reaction was explored. In
general, the composition of the aryl substituent had little impact
on the reaction outcome (Scheme 2). Sterically demanding
substituents (7c) as well as both electron-rich (7d and 7e) and
electron-deficient (7f, and 7g) arenes were equally well tolerated
in the reaction. Moreover, N-oxides containing basic nitrogen
functional groups afforded excellent yields of the corresponding
pyrazoles (7h and 7i). Taken together, these results indicate that
inductive effects of C3-aryl groups are muted when a heteroatom
substituent is present in structure 9.
We also investigated the scope of C6 heteroatom groups in the
photoreaction. As illustrated in Scheme 3, a family of pyridazine
N-oxides (9j−y) was prepared by reacting 8a with nitrogen,
oxygen, and sulfur nucleophiles. When reacted under optimal
photochemical conditions, a range of tertiary and secondary
amines were well tolerated (7j−r), including those containing
sulfur (7l) and both unprotected amine (7m) and alcohol (7r)
functional groups. Azetidine (7n) and aniline (7o) derivatives
were also compatible with this chemistry. Similarly, photo-
substrates 9s−y containing aryl ethers reacted smoothly to give
the corresponding pyrazole esters 7s−x in 60−89% yield. The
lone exception in this group was phenol derivative 9y, which
affordedpyrazole7y in43%yieldalongwiththeparentpyridazine
derived from photodeoxygenation. Notably, pyridazine N-oxides
harboring sulfur groups at C6 failed to react to any appreciable
extent.
AUTHOR INFORMATION
Corresponding Author
■
*E-mail: frederich@chem.fsu.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by Florida State University. We
thank Dr. Ron Clark for assistance with X-ray crystallography. We
also thank Professors Igor Alabugin (FSU) and Gregory Dudley
(FSU) for access to chemicals and equipment.
REFERENCES
(1) Wacker, L. Justus Liebigs Ann. Chem. 1901, 317, 375.
(2) Albini, A.; Alpegiani, M. Chem. Rev. 1984, 84, 43.
■
(3) (a) Kumler, P. L.; Buchardt, O. J. Am. Chem. Soc. 1968, 90, 5640.
(b) Tomer, K. B.; Harrit, N.; Rosenthal, I.; Buchardt, O.; Kumler, P. L.;
Creed, D. J. Am. Chem. Soc. 1973, 95, 7402.
(4) (a) Ogata, M.; Kano, K. Chem. Commun. 1967, 1176. (b) Tsuchiya,
T.; Arai, H.; Igeta, H. Tetrahedron Lett. 1969, 10, 2479. (c) Tsuchiya, T.;
Arai, H.; Igeta, H. J. Chem. Soc., Chem. Commun. 1972, 550.
(5) N-Alkylpyrroles are formed in 7−15% yield via the photoreaction of
6-methylpyridazine N-oxide in the presence of amines: Tsuchiya, T.;
Arai, H.; Igeta, H. Chem. Pharm. Bull. 1973, 21, 1516.
(6) The ring opening of pyridazinium salts with KOH is known:
Kanchupalli, V.; Joseph, D.; Katukojvala, S. Org. Lett. 2015, 17, 5878.
(7) Streith, J.; Sigwalt, C. Bull. Chem. Soc. Fr. 1970, 1157.
(8) Fustero, S.; Sanchez-Rosello, M.; Barrio, P.; Simon-Fuentes, A.
Chem. Rev. 2011, 111, 6984.
(9) For biologically active molecules containing the 3-aryl-5-
acylpyrazole scaffold, see: (a) Hoveyda, H. R.; Roy, M.; Blanc, S.;
Noel, S.; Salvino, J. M.; Ator, M. A.; Fraser, G. Bioorg. Med. Chem. Lett.
2011, 21, 1991. (b) Dymock, B. W.; Barril, X.; Brough, P. A.; Cansfield, J.
E.; Massey, A.; McDonald, E.; Hubbard, R. E.; Surgenor, A.; Roughley, S.
D.; Webb, P.; Workman, P.; Wright, L.; Drysdale, M. J. J. Med. Chem.
2005, 48, 4212.
(10) Tsukamoto, Y.; Kudo, N.; Kaneko, T.; Komai, T.; Nakagawa, H.
Pat. Appl. WO2006101058A1, Sep 28, 2006.
To highlight the synthetic utility of this method, we aimed to
apply our strategy to the synthesis of a drug-type molecule.
Toward this end, we became aware of a report by Cephalon that
identified 3-aryl-5-acylpiperazinyl pyrazoles as potent antagonists
of the G protein-coupled human tachykinin NK3 receptor.9a The
pharmacologyofthisscaffoldreportedlymimicsthatofosanetant,
a compound evaluated in human clinical trials by Sanofi for the
treatment of schizophrenia.
̈
(11) (a) Sako, S. Chem. Pharm. Bull. 1962, 10, 956. (b) Ochiai, M.;
Okada, T.;Morimoto, A.;Kawakita, K. J. Chem. Soc., PerkinTrans. 11976,
1988.
(12)(a)Lin, S.;Liu,Z.;Hu,Y. J. Comb.Chem.2007,9,742.(b)Villemin,
D.; Jullien, A.; Bar, N. Tetrahedron Lett. 2007, 48, 4191. (c) Smith, S. M.;
Buchwald, S. L. Org. Lett. 2016, 18, 2180.
(13) For a review on regioselective cross-couplings of polyhalogenated
heterocycles, see: Rossi, R.; Bellina, F.; Lessi, M. Adv. Synth. Catal. 2012,
354, 1181.
(14)(a)Kudo, N.;Perseghini, M.;Fu, G. C. Angew. Chem., Int. Ed. 2006,
45, 1282. (b) Billingsley, K. L.; Anderson, K. W.; Buchwald, S. L. Angew.
Chem., Int. Ed. 2006, 45, 3484.
A three-step synthesis of hNK3 antagonist 17 (IC50 = 240 nM)
is described in Scheme 4. Accordingly, a large-scale (20.0 mmol)
cross coupling of 3,6-dichloropyridazine N-oxide and 4-
fluorophenylboronic acid furnished 8f in 85% yield. Subsequent
amidationof8fwithpiperazine15generatedphotosubstrate16in
92% yield. Photolysis of 16 (THF, 0.3 M) smoothly afforded
pyrazole 17inhigh yieldafter 5h. Whencarried out on 10.0mmol
scale, this reaction produced 3.42 g of 17 (90% yield) as a
crystalline solid.
In summary, we have developed a distinct photochemical
reaction for the synthesis of pyrazoles. This approach highlights
the utility of photoinduced ring-opening reactions of pyridazine
N-oxides for the first time. In contrast to existing methods,8 the
photochemistry described here offers a scalable, modular, and
exceedingly mild way to prepare 3-aryl-5-acyl-1H-pyrazoles.
Efforts to adapt this chemistry to continuous flow and further
apply the photochemical ring opening of 1,2-diazine N-oxides to
complex synthetic targets are ongoing.
(15) See the Supporting Information for more details.
(16) Regiochemistry of 8a and 9a was confirmed by X-ray
crystallography. These data have been deposited at the Cambridge
Structural Database as entries 1500539 (8a) and 1500541 (9a). The
regiochemistry of 8b−i was inferred by formation of the pyrazoles 7b−i.
(17) Resubmission of 8a to these reaction conditions afforded 1
quantitatively after 3 h. Thus, we speculate that the N-oxide group,
together with use of the bulky ferrocene ligand, directs the reaction to the
C3 position via combined steric effects.
(18) Compound 9a exhibited strong absorbance bands at 275 nm (π→
π*) and 358 (n→π*). See the Supporting Information for UV−vis data.
(19) An NMR experiment following the thermal conversion of 10 and
to 7a is available in the Supporting Information.
ASSOCIATED CONTENT
■
S
* Supporting Information
TheSupportingInformationisavailablefreeofchargeontheACS
Publications website at DOI: 10.1021/acs.orglett.6b02562.
Optimization data, detailed synthetic procedures, charac-
terization data, and 1H and 13C NMR spectra (PDF)
D
DOI: 10.1021/acs.orglett.6b02562
Org. Lett. XXXX, XXX, XXX−XXX