Photocatalytic Alkylation of Pyrroles and Indoles with α-Diazo Esters

Article abstract of DOI:10.1021/acs.orglett.9b02612

This article describes the photoalkylation of electron-rich aromatic compounds with diazo esters. C-2-alkylated indoles and pyrroles are obtained with good yields even though the photocatalyst loading is as low as 0.075 mol %. For EWG-substituted substrates, the addition of a catalytic amount of N,N-dimethyl-4-methoxyaniline is required. Both EWG-EWG- and EWG-EDG-substituted diazo esters are suitable as alkylating agents. The reaction selectivity and mechanistic experiments suggest that carbenes/carbenoid intermediates are not involved in the reaction pathway.

Full text of DOI:10.1021/acs.orglett.9b02612

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Photocatalytic Alkylation of Pyrroles and Indoles with α‑Diazo Esters
Łukasz W. Ciszewski,† Jakub Durka,†,‡ and Dorota Gryko*,†
†Institute of Organic Chemistry, Polish Academy of Sciences, Kasprzaka 44/52 01-224 Warsaw, Poland
‡Department of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-664 Warsaw, Poland
S
* Supporting Information
ABSTRACT: This article describes the photoalkylation of
electron-rich aromatic compounds with diazo esters. C-2-
alkylated indoles and pyrroles are obtained with good yields
even though the photocatalyst loading is as low as 0.075 mol %.
For EWG-substituted substrates, the addition of a catalytic
amount of N,N-dimethyl-4-methoxyaniline is required. Both
EWG−EWG- and EWG−EDG-substituted diazo esters are
suitable as alkylating agents. The reaction selectivity and
mechanistic experiments suggest that carbenes/carbenoid intermediates are not involved in the reaction pathway.
Scheme 1. Reactivity of Diazo Reagents toward Indoles
ive membered heteroaromatic rings are common struc-
Ftural motifs in pharmaceuticals, agrochemicals, and
functional dyes.1 Indoles and pyrroles exhibiting biological
activity often possess at least one alkyl substituent attached to
the aromatic ring at either the C2 or C3 position.
Consequently, mild and regioselective methods for C(sp2)−
C(sp3) bond formation are of high importance.
Classic methods for C−H alkylation of electron-rich
heteroaromatic compounds such as Friedel−Crafts-type
reactions or metalation followed by reaction with electrophiles
are suitable for functionalization of simple derivatives, as they
often require conditions that are not compatible with many
functional groups. Current procedures employing transition-
metal catalysis broaden the scope, yet due to the character of a
catalyst, anhydrous conditions, and elevated temperatures are
often required.2 For indole derivatives, reactions usually lead to
less challenging C3 alkylation products if no directing group is
present.3
In recent years, photoredox catalysis emerged as a facile
synthetic tool4 for the formation of C−C bonds and also for
alkylation of electron-rich heteroarenes.5 These reactions rely
on ruthenium or iridium complexes which under light
irradiation reduce electron-deficient alkyl halides to radicals
that easily react with electron-rich heteroaromatic compounds.
Utilization of Au(I) photocatalyst allows similar reactions, yet
this process inherently suffers from low selectivity due to
instability of radicals.6 Very recently, Glorius et al. reported
visible-light-induced C2 alkylation of heteroaromatic com-
pounds using pyridinium salts as radical precursors.7 However,
due to the utilization of electron-donor−acceptor complexes,
the scope of the method is limited to benzene-fused
heteroarenes.
the C2 position2b,3 because the directing group is present at
the indole nitrogen atom (Scheme 1B).9 We have recently
reported that under light irradiation diazo esters can act as
alkylating agents toward in situ formed enamines,10 and a later
similar approach was also utilized for enantioselective
alkylation of 2-acylimidazoles by Meggers et al.11 We wondered
whether and how the use of photoredox catalysts would change the
reactivity of diazo reagents toward electron-rich heteroarenes.
Herein, we report a new photocatalytic method for C2
alkylation of heteroarenes with diazo compounds (Scheme
1C). The reaction selectivity supported by mechanistic studies
suggests that carbene/carbenoid intermediates are not involved
in the reaction pathway.
Along this line, thermolysis of ethyl diazoacetate (EDA, 2)
in neat N-methylindole (1) gives C3-alkylated product
resulting from the ring opening of the corresponding
cyclopropane intermediates (Scheme 1A).8 As more stable
entities, metal carbenoids insert into the indole C−H bond at
Received: July 25, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b02612
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
We initiated our studies by performing the model reaction of
EDA (2) with N-methylindole (1) catalyzed by Ru(bpy)3Cl2
under light irradiation. Product 3 selectively alkylated at C2
position formed in 30% yield (Table 1). Background
s−1 to 4.89 × 107 s−1, suggesting that the protonated form A of
diazo ester may indeed be the actual quencher of the excited
catalyst. As a consequence, after extrusion of the nitrogen
radical B is generated, which reacts with N-methylindole (1) to
3+
give radical C. Subsequent oxidation with Ru(bpy)3 allows
a
Table 1. Background Experiments
for the recovery of the catalyst. Deprotonation of cation D
furnishes the desired alkylated product. The alternative
pathway would involve carbene intermediates, but both
photolysis of benzyl diazoacetate (254 nm) and the reaction
with N-methylindole (1) in the presence of benzophenone (as
a triplet sensitizer) yield only a mixture of O−H insertion,
Wolff rearrangement products, and traces of C3 alkylation
product, suggesting that in our case that their presence is not
an option.
In the next step, we optimized the reaction conditions for
the model reaction. Common organic dyes or metal complexes
broadly used in photoredox catalysis were tested, but only
Ru(bpy)3Cl2 was able to catalyze the model reaction (details
can be found in SI). The catalyst loading could be as low as
0.075 mol %; however, for the scope and limitations studies it
was kept at 0.2 mol % level to ensure activity of the catalyst at
longer reaction times. The use of protic solvents with the
addition of water assured the formation of product 3 with the
best result being obtained in a mixture of MeOH/H2O (10:1
v/v). With the optimal ratio of N-methylindole (1) to EDA
(2) being 4:1, the reaction yielded product in 76% yield. An
excess of heteroaromatic compound suppresses the formation
of dialkylated product. The process could be scaled up to 5
mmol, furnishing product with only slightly diminished yield
(68%) after 26 h; in this case, excess starting material was
recovered (Scheme 3).
entry
PC
solvent
yield (GC) (%)
1
2
Ru(bpy)3Cl2
none
Ru(bpy)3Cl2
Ru(bpy)3Cl2
CH3CN
CH3CN
CH3CN
CH3CN
30
nr
nr
nr
b
3
c
4
a
Reaction conditions: 1 (1.25 mmol, 5 equiv), 2 (0.25 mmol, 1
equiv), Ru(bpy)3Cl2 (2.5 mol %), CH3CN (2.5 mL), blue LED
b
c
irradiation, 8 h. No light irradiation. Reaction mixture was not
degassed.
experiments revealed that the exclusion of ether the photo-
catalyst or light irradiation stopped the reaction completely
(entries 2 and 3). Importantly, the reaction required anaerobic
conditions (entry 4).
The proposed mechanism for the alkylation of indoles is
shown in Scheme 2. We assumed that the crucial step of this
Scheme 2. Plausible Mechanism for the Alkylation of
Indoles
With the optimized conditions in hand, we conducted scope
and limitations studies. To this end, a series of α-diazo esters
was synthesized and tested in a reaction with N-methylindole
(1) (Scheme 3).
Scheme 3. Reaction of N-Me-indole with Diazo Esters*
photochemical alkylation involves single-electron reduction of
diazoethyl acetate, but a detailed mechanistic overview
required further exploration. The addition of TEMPO, a
free-radical scavenger, halted the reaction completely, proving
that radical species are indeed involved. The Stern−Volmer
analysis indicated that only EDA (1) quenches luminescence
of the Ru catalyst.
This result supports generation of electron-deficient radical
B via single-electron reduction of ethyl diazoacetate (2).
However, EDA (2) reduction potential (see the SI) is too low
to be accessible by any form of the photocatalyst used (*Ru2+/
Ru3+ E= −0.78 V vs SCE),12 but in protic solvents, particularly
in the presence of water, diazo compounds are in equilibrium
with their protonated form (A);13 as a result, their reduction
potential should increase.14 In fact, in aqueous conditions
(same as the reaction medium) the emission-quenching rate
constant of ethyl diazoacetate increases from kq = 1.26 × 107
*
Reaction conditions: 1 (1 mmol, 4 equiv), diazo compound (0.25
mmol, 1 equiv), Ru-cat: Ru(bpy)3Cl2 (0.5 μmol, 0.2 mol %), MeOH/
a
H2O (10:1) 2.75 mL, blue LED irradiation, 4.5−8 h. 5 mmol scale,
b
c
73% of 1 was recovered. Determined by GC, Determined by NMR.
d
e
No Ru(bpy)3Cl2 added. MeCN instead of MeOH.
B
DOI: 10.1021/acs.orglett.9b02612
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
In contrast to our previous reports,10 disubstituted diazo
reagents reacted equally well with the yield and selectivity
depending on the substitution pattern and their absorption
characteristics. In general, acceptor−acceptor (EWG−EWG)
diazo compounds were more reactive toward indole 1 than
acceptor−donor (EWG−EDG) substituted compounds. Along
this line, ethyl diazo(diethylphosphono)acetate proved the
most reactive, giving alkylated derivative 6 in 81% yield. α-
Diazo ketones−ethyl 2-diazoacetoacetate and diethyl (1-diazo-
2-oxopropyl)phosphonate did not afford corresponding
products as prior alkylation the carbonyl group transformed
into acetal or hemiacetal, hence changing the character of the
diazo reagent. This, however, does not mean that all EWG−
EDG disubstituted diazo compounds are unreactive under
developed conditions, and ethyl 2-diazo-3-hydroxy-3-phenyl-
propanoate furnished the corresponding product 7 in 54%
yield.
Scheme 4. Scope and Limitation Studies: Heteroarenes*
Intriguingly, diethyl (cyano(diazo)methyl)phosphonate and
methyl phenyldiazoacetate underwent selective alkylation at
the C3 position, suggesting a possible parallel pathway. In
2018, Davies and co-workers showed that phenyl diazoacetate
absorbing at approximately 450 nm (DCM) under blue light
irradiation reacts with N-methylindole (1) to give C−H
insertion products exclusively at the C3 position as a result of
presumed carbene insertion.15 We observed that in a mixture
of MeOH/H2O the corresponding product 8 (C3 re-
gioisomer) forms but is accompanied by methyl 2-methoxy-
2-phenylacetate, resulting from carbene insertion into the O−
H bond of methanol. Changing the reaction media to MeCN/
H2O and adding Ru photocatalyst allowed us to suppress side
reactions; as a consequence, the yield of product 8 increased
up to 44%. Similarly, the carbene might be also involved in
reaction with diethyl (cyano(diazo)methyl)phosphonate as
both Ru-catalyzed and uncatalyzed alkylation gave the same
product 9. Additionally, under blue light irradiation diethyl
(cyano(diazo)methyl)phosphonate in MeOH photodecom-
posed to a mixture of diethyl cyanomethylphosphonate (66%)
and diethyl ((cyanomethoxy)methyl)phosphonate (11%)
resulting from the reaction of the corresponding carbene
with MeOH (see SI). These results corroborate that once the
diazo compound absorbs within the wavelength region of the light
used for irradiation, the regioselectivity of the alkylation reaction
alters from C2 to C3, as a consequence of a different operating
mechanism.
In the next step, various indole and pyrrole derivatives were
tested (Scheme 4). The method worked equally well for
unprotected both indoles and pyrroles, giving products 10−12,
16, and 19 in decent yields accompanied by small amounts of
other regioisomers (see SI). Mild reaction conditions allowed
us to functionalize even substrates bearing the fragile
cyclopropyl group at the C3 position without ring opening
being observed. Electron-deficient heteroarene 14 remained
intact under the developed conditions, presumably due to
increased oxidation potential of intermediate C inaccessible by
Ru photocatalyst.
*
Reaction conditions: heteroarene (1 mmol, 4 equiv), diazo
compound (0.25 mmol, 1 equiv), Ru(bpy)3Cl2 (0.5 μmol, 0.2 mol
%), MeOH/H2O (10:1) 2.75 mL, blue LED irradiation, 4.5−8 h.
a
b
c
Determined by GC. Determined by NMR. 2 equiv of heteroarene
was used.
hormone melatonin, giving the corresponding product 18 in
91% yield, even though the excess of the starting material was
reduced 2-fold. N-Methylpyrrole afforded the corresponding
alkylated products 20 and 21 in good yields, but once the −Me
substituent was replaced with the electron-withdrawing phenyl
group the yield diminished substantially.
Puzzled by the poor reactivity of heteroarenes with
diminished electron density, N-Boc-indole (24) and N-Boc-
pyrrole, we wondered whether Ru(bpy)3 is a sufficiently
strong electron acceptor for the oxidation of type C radical to
the respective cation D (Scheme 2). If not, the catalytic cycle
cannot close. Therefore, a variety of redox-active additives
enabling closing of the catalytic cycle were tested (see the SI).
The addition of a catalytic amount (10 mol %) of 4-methoxy-
N,N-dimethylanilnie (25) facilitated the reaction for indole
and pyrrole derivatives with diminished electron density. The
quenching rate constant (kq = 1.46 × 109 s−1) of the
fluorescence of the Ru catalyst by aniline (25) was 2 orders of
magnitude higher than for EDA (2), indicating that indeed a
different mechanistic pathway should operate in this case
3+
2+
(Scheme 5). We assumed that the excited state of Ru(bpy)3
oxidizes amine 25 to the radical cation generating Ru(bpy)3
+
Not surprisingly, the substitution pattern on the phenyl ring
has a substantial impact on the reaction regioselectivity. While
the reaction of 4-methoxy-1-methylindole yielded C2-alkylated
product 15 with high selectivity, the 5-methoxy derivative
furnished a mixture of C4 and C2 regioisomers 16a,b and
17a,b in accordance with increased electron densities on the
corresponding positions in the starting material. The method
was also suitable for functionalization of the sleep-regulating
species. The reduced catalyst is able to generate radical B,
which reacts with heteroaromatic substrate giving radical C′.
The final step involves hydrogen atom transfer between radical
cation of the amine 25 and radical C′, which results in
formation of the product.
As a consequence, a series of electron-deficient heteroarenes
alkylated at C2 position was obtained. N-Boc-indole (24)
which was not reactive under standard conditions after the
C
DOI: 10.1021/acs.orglett.9b02612
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
high yield (72%), while for N-phenylpyrrole the yield increased
almost 2.3-fold. Notably, even electron-rich N-
(dimethylamino)pyrrole afforded product 32 in decent yield.
In conclusion, a new photocatalytic method for C2
alkylation of indoles and pyrroles has been developed. The
method requires unprecedentedly low catalyst loading (0.075
mol %), tolerates a variety of functional groups in both
heteroaromatic substrates and diazo compounds, and is easily
scalable. The addition of N,N-dimethyl-4-methoxyaniline (25)
enables the synthesis of alkylated derivatives even from
electron-deficient indoles and pyrroles.
Scheme 5. Effect of 4-Methoxy-N,N-dimethylanilnie (28)
Addition on the Plausible Mechanism
Mechanistic studies corroborate the proposed reaction
pathways involving radical species. However, for diazo
compounds exhibiting strong absorption within the wavelength
region of the light used for irradiation, the regioselectivity of
the alkylation reaction alters from C2 to C3.
ASSOCIATED CONTENT
* Supporting Information
■
S
addition of amine (25) furnished corresponding product 14 in
60% yield (Scheme 6). Other indole derivatives bearing
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.9b02612.
Scheme 6. Alkylation of Indole and Pyrrole Derivatives in
the Presence of Aniline 25*
Optimization details, mechanistic experiments, exper-
imental procedures, and characterization data for all new
compounds (PDF)
Models for the 3d printed photoreactor (ZIP)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: dorota.gryko@icho.edu.pl.
ORCID
Dorota Gryko: 0000-0002-5197-4222
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
̈
We thank Prof. B. Konig from the University of Regensburg,
Faculty of Chemistry and Pharmacy, Institute of Organic
Chemistry, Regensburg, Germany, for a photoreactor setup
with an aluminum cooling block, which served as a model for
our 3D printed photoreactor. Financial support for this work
was supported by the Ministry of Science and Higher
Education (Ł.W.C., grant no. 0205/DIA/2016/45) and the
Foundation for Polish Sciences (D.G., Grant No. FNP TEAM
POIR.04.04.00-00-4232/17-00).
*
Reaction conditions: heteroarene (1 mmol, 4 equiv), diazo
compound (0.25 mmol, 1 equiv), 28 (25 μmol, 10 mol %),
Ru(bpy)3Cl2 (0.5 μmol, 0.2 mol %), MeOH/H2O (10:1) 2.75 mL,
blue LED irradiation, 4.5−8 h. Other isomer yields determined by
NMR. Other isomer yields determined by GC.
a
REFERENCES
b
■
(1) Lednicer, D. Strategies for Organic Drug Synthesis and Design;
John Wiley & Sons, Inc.: Hoboken, 2008.
electron-withdrawing substituents are also suitable substrates
for this reaction. 5-Bromoindole provides an almost equimolar
mixture of C2 26a and C4 26b alkylated derivatives, while C2-
substituted product 27 mainly formed from the corresponding
N-methyl derivative. The regioselectivity increased for indoles
bearing a strong electron-withdrawing cyano group; regardless
of the substitution pattern, C2 product 28 was solely obtained.
In addition, alkylation of L-tryptophan methyl ester gave 67%
of the corresponding alkylated derivative 29.
(2) (a) Seregin, I. V.; Gevorgyan, V. Chem. Soc. Rev. 2007, 36, 1173.
(b) Li, Y.-P.; Li, Z.-Q.; Zhu, S.-F. Tetrahedron Lett. 2018, 59, 2307.
̈
(c) Sambiagio, C.; Schonbauer, D.; Blieck, R.; Dao-Huy, T.;
Pototschnig, G.; Schaaf, P.; Wiesinger, T.; Zia, M. F.; Wencel-
Delord, J.; Besset, T.; Maes, B. U. W.; Schnu
2018, 47, 6603.
̈
rch, M. Chem. Soc. Rev.
(3) (a) Gibe, R.; Kerr, M. A. J. Org. Chem. 2002, 67, 6247. (b) Li, Z.;
Shi, Z.; He, C. J. Organomet. Chem. 2005, 690, 5049.
(4) (a) Prier, C. K.; Rankic, D. A.; MacMillan, D. W. C. Chem. Rev.
̈
2013, 113, 5322. (b) Hari, D. P.; Konig, B. Chem. Commun. 2014, 50,
The addition of amine 25 was particularly beneficial for
pyrroles where a substantial increase in yields was observed.
Alkylation of N-Boc-protected pyrrole yielded product 23 in
6688. (c) Romero, N. A.; Nicewicz, D. A. Chem. Rev. 2016, 116,
10075. (d) Douglas, J. J.; Sevrin, M. J.; Stephenson, C. R. J. Org.
Process Res. Dev. 2016, 20, 1134. (e) Staveness, D.; Bosque, I.;
D
DOI: 10.1021/acs.orglett.9b02612
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Stephenson, C. R. J. Acc. Chem. Res. 2016, 49, 2295. (f) Yoon, T. P.
Acc. Chem. Res. 2016, 49, 2307. (g) Boubertakh, O.; Goddard, J. P.
Eur. J. Org. Chem. 2017, 2017, 2072. (h) Buzzetti, L.; Crisenza, G. E.
M.; Melchiorre, P. Angew. Chem., Int. Ed. 2019, 58, 3730.
(5) (a) Sun, A.; McAtee, R.; McClain, E.; Stephenson, C. Synthesis
2019, 51, 1063. (b) Tucker, J. W.; Narayanam, J. M. R.; Krabbe, S.
W.; Stephenson, C. R. J. Org. Lett. 2010, 12, 368. (c) Furst, L.;
Matsuura, B. S.; Narayanam, J. M. R.; Tucker, J. W.; Stephenson, C.
R. J. Org. Lett. 2010, 12, 3104. (d) Jang, H.-L.; Kim, H. T.; Cho, E. J.;
Joo, J. M. Asian J. Org. Chem. 2015, 4, 1386.
(6) McCallum, T.; Barriault, L. Chem. Sci. 2016, 7, 4754.
(7) James, M. J.; Strieth-Kalthoff, F.; Sandfort, F.; Klauck, F.;
Wagener, F.; Glorius, F. Chem. - Eur. J. 2019, 25, 8240.
(8) Jackson, R. W.; Manske, R. H. Can. J. Res. 1935, 13b, 170.
(9) (a) Iagafarova, I. E.; Vorobyeva, D. V.; Peregudov, A. S.; Osipov,
S. N. Eur. J. Org. Chem. 2015, 2015, 4950. (b) Liu, X. G.; Zhang, S. S.;
Wu, J. Q.; Li, Q.; Wang, H. Tetrahedron Lett. 2015, 56, 4093.
(c) Wang, C.; Maity, B.; Cavallo, L.; Rueping, M. Org. Lett. 2018, 20,
3105. (d) Kuvshinova, S. S.; Smolyakov, A. F.; Vorobyeva, D. V.;
Osipov, S. N.; Loginov, D. A. Mendeleev Commun. 2018, 28, 359.
(e) Ghorai, J.; Chaitanya, M.; Anbarasan, P. Org. Biomol. Chem. 2018,
16, 7346. (f) Dhole, S.; Chiu, W.; Sun, C. Adv. Synth. Catal. 2019,
361, 2916.
́
(10) (a) Rybicka-Jasinska, K.; Ciszewski, Ł. W.; Gryko, D. Adv.
́
Synth. Catal. 2016, 358, 1671. (b) Rybicka-Jasinska, K.; Shan, W.;
Zawada, K.; Kadish, K. M.; Gryko, D. J. Am. Chem. Soc. 2016, 138,
́
15451. (c) Rybicka-Jasinska, K.; Orłowska, K.; Karczewski, M.;
Zawada, K.; Gryko, D. Eur. J. Org. Chem. 2018, 2018, 6634.
(11) Huang, X.; Webster, R. D.; Harms, K.; Meggers, E. J. Am. Chem.
Soc. 2016, 138, 12636.
(12) Juris, A.; Balzani, V.; Belser, P.; von Zelewsky, A. Helv. Chim.
Acta 1981, 64, 2175.
(13) Albery, W. J.; Bell, R. P. Trans. Faraday Soc. 1961, 57, 1942.
(14) In general, protonation lowers both LUMO and HOMO
energies, thus increasing the reduction potential in comparison with
nonprotonated compound. The effect of protonation on reduction
potential was observed for aromatic hydrocarbons: (a) Given, P. H.;
Peover, M. E. J. Chem. Soc. 1960, 385. For aromatic nitro
compounds, see: (b) Cadle, S. H.; Tice, P. R.; Chambers, J. Q. J.
Phys. Chem. 1967, 71, 3517.
(15) Jurberg, I. D.; Davies, H. M. L. Chem. Sci. 2018, 9, 5112.
E
DOI: 10.1021/acs.orglett.9b02612
Org. Lett. XXXX, XXX, XXX−XXX