Site-Selective C-H Alkylation of Piperazine Substrates via Organic Photoredox Catalysis

Article abstract of DOI:10.1021/acs.orglett.9b04456

Piperazine-containing compounds serve as one of the most important classes of compounds throughout all fields of chemistry. Alas, current synthetic methods have fallen short of providing a general method for the synthesis of highly decorated piperazine fragments. Herein, we present a site-selective approach to the C-H functionalization of existing piperazine compounds using photoredox catalysis. This manifold relies on the predictable differentiation of electronically distinct nitrogen centers within the piperazine framework, granting access to novel C-alkylated variants of the starting piperazines.

Full text of DOI:10.1021/acs.orglett.9b04456

Letter
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Site-Selective C−H Alkylation of Piperazine Substrates via Organic
Photoredox Catalysis
†
†
‡
‡
Joshua B. McManus, Nicholas P. R. Onuska, Matthew S. Jeffreys, Nicole C. Goodwin,
,
†
and David A. Nicewicz*
†Department of Chemistry, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-3290, United
States
‡
GlaxoSmithKline, 1250 South Collegeville Road, Collegeville, Pennsylvania 19426, United States
*
S Supporting Information
ABSTRACT: Piperazine-containing compounds serve as one
of the most important classes of compounds throughout all
fields of chemistry. Alas, current synthetic methods have fallen
short of providing a general method for the synthesis of highly
decorated piperazine fragments. Herein, we present a site-
selective approach to the C−H functionalization of existing
piperazine compounds using photoredox catalysis. This
manifold relies on the predictable differentiation of electroni-
cally distinct nitrogen centers within the piperazine framework,
granting access to novel C-alkylated variants of the starting
piperazines.
n an effort to optimize “drug-like” physicochemical
synthetic limitations are hindering access to these more
elaborate targets.
I
properties, medicinal chemists often turn to nitrogen-
containing heterocycles as integral core structures. Piperazines,
containing two differentiable nitrogens and an unsubstituted,
saturated carbon skeleton, were found to be the third-most
prevalent such heterocycle in current pharmaceuticals behind
To date, the development of a general method for the
construction of C-substituted piperazine cores has yet to be
achieved. Classically, these compounds are synthesized via the
reduction of 2,5-diketopiperazines generated via peptide
1
2−4
only piperidines and pyridines (Figure 1). Despite the
cyclodimerizations or by the multicomponent Ugi reaction.
However, these methods require tedious multistep procedures
that can offer only a subset of the desired substituted
piperazines. Other methods developed for accessing sub-
stituted piperazine or morpholine-type products have tradi-
tionally relied on de novo syntheses of the target heterocycle via
5
−10
innovative cyclization strategies.
While current methods for functionalizing existing piper-
azines typically rely on harsh or substrate-specific reaction
11−13
conditions,
recent findings by our laboratory (Figure 2A)
focusing on the alkylation of photoredox-generated α-carbamyl
radicals (1) to give elaborate, protected secondary amine
products (2) served as inspiration for the development of a
targeted approach to the mild, site-selective, and atom-
economical synthesis of substituted piperazine compounds
from their more readily available unsubstituted precursors
1
4
(
Figure 2B). A single example has been disclosed using this
15
Figure 1. Representative piperazine-containing pharmaceuticals.
type of strategy in a photoredox setting; however, we believe
expanding on such a strategy would grant direct access to a
variety of piperazines bearing additional functionality from
inexpensive but structurally related starting materials. This
approach would be advantageous from a medicinal chemistry
ubiquity of piperazine-containing therapeutics, structural
diversity of these compounds along the piperazine’s carbon
1
framework is remarkably limited. This observation provokes
the query of whether structurally diverse piperazines represent
a class of inconsequential compounds or whether current
Received: December 12, 2019
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b04456
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Scope of the Site-Selective Piperazine Alkylation
d
Reaction
Figure 2. (A) Mechanism for the alkylation of carbamates and (B)
site-selective piperazine alkylation.
perspective, as the vast array of existing bioactive piperazines
could be further functionalized via predictable late-stage
modification of the piperazine core.
To develop a synthetically viable strategy as outlined in
Figure 2, we believed that sufficient electronic differentiation of
the two nitrogen atoms of a piperazine substrate would be
necessary for a site-selective reaction to proceed. We
hypothesized that a photoinduced electron transfer (PET)
event oxidizing a differentially N,N′-disubstituted piperazine
substrate would form a nitrogen-centered cation radical at the
more electron-rich locus. Single-electron oxidation at the more
electron-rich nitrogen center followed by deprotonation and
subsequent spin-center shift would yield α-amino radical 1
(
Figure 2A). These versatile reactive intermediates have been
utilized to introduce varied functionality α to a nitrogen
1
6
center. Herein,l we describe a site-selective approach to
direct C−H alkylation of piperazine compounds through the
electronic differentiation of the two nitrogen centers within the
heterocycle.
Adapting conditions from work previously developed and
1
4
disclosed by our lab, we combined the piperazine substrate
with an acridinium photocatalyst (A or B) and the electrophilic
coupling partner, methyl vinyl ketone (MVK). Utilizing the
conditions displayed in Scheme 1, we were able to establish
general reactivity of piperazine substrates by coupling N,N′-
diBoc-piperazine with MVK to from adduct 3 in excellent
a
b
c
d
Catalyst A. Catalyst B. Reaction performed at 0 °C. Isolated
yields and regioselectivities are averages of two trials.
shows only modest selectivity for α-functionalization favoring
the Boc-carbamate, the differentiation between two similar
carbamate functional groups demonstrates the promising
chemoselectivity of this approach.
Further exploration into substituent effects on nitrogen
revealed that modest selectivity can be achieved between a
number of carbamate functional groups using this method (4−
7 and 9). In addition to carbamates, trimethylacetamide was
also tolerated under these conditions to form 8 in a 1.5:1 ratio,
again favoring reactivity proximal to the Boc-protected
nitrogen. Benzamide derivatives 10−12 could also be smoothly
converted to the desired alkylation adducts in good isolated
yields and excellent regioselectivity in the case of 10.
1
4
yield. While this result showed proof of concept, neither
nitrogen center on N,N′-diBoc-piperazine could be differ-
entiated from a stereoelectronic perspective, following the
alkylation event. This prevents downstream transformations of
3
via the selective deprotection of a single Boc group. Because
of this limitation, we sought to apply this method to
piperazines bearing differentiated functionality or orthogonal
protecting groups on the two nitrogen centers.
To this end, we were intrigued to find that compound 4
could be accessed from N-Boc-N′-Cbz-piperazine in a 1.7:1
mixture of regioisomers and in good yield. While this example
B
DOI: 10.1021/acs.orglett.9b04456
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Moreover, 10 could be scaled up from 0.3 to 1.0 mmol without
notable changes in yield or rr (see the Supporting Information
for details). This ability to electronically differentiate between
two common nitrogen protecting groups underscores the
utility of this method. Such a protocol allows for a stepwise
procedure for accessing synthetically valuable piperazine
building blocks.
Next, we sought to probe the compatibility of more electron-
rich nitrogen centers compared to those derived from
carbamates and amides. It was determined that arylamine
derivatives were well-suited for this reaction and that
compounds 13−16 could be formed with excellent levels of
regiocontrol. Interestingly, we observed a reversal in
regioselectivity for these substrates with alkylation occurring
preferentially proximal to the more electron-rich arylamine. A
similar effect is observed for strongly withdrawing sulfonamide
protecting groups with alkylation occurring exclusively at the
site proximal to the N-Boc group, albeit in lower yields (17 and
performed to measure the relative electron density at both
nitrogen centers in the neutral (q) substrate along with the
•
+
cation radical intermediate (q ) (see the Supporting
Information for details). The site selectivity was then predicted
by measuring the change in the NPA values of the individual
nitrogen atoms between the neutral species and the
corresponding radical cation (Δq), wherein the nitrogen
atom that undergoes the most drastic change in electron
17,18
density was predicted to be the principal site of alkylation.
Figure 3 illustrates this phenomenon for a set of substrates, and
1
8). Lastly, this method was applied to the late-stage
functionalization of a derivative of the commercial antibiotic,
ciprofloxacin (19), in fair yield and as a single regioisomer.
During the development of this procedure, it was discovered
that the regioselectivity of alkylation was strongly dependent
on reaction temperature. Thus, when the reaction mixture was
cooled from 29 to 0 °C, the regioselectivity for this
transformation could be enhanced, increasing the rr for 10
from 10:1 to 21:1 and that for 15 from 9:1 to 20:1 (see the
Supporting Information for details); these trials did not affect
which regioisomer forms as the major product.
In addition to MVK, several other Michael acceptors could
be coupled to the piperazine substrates in an analogous
fashion. Scheme 2 displays these Michael acceptors, now
a
Scheme 2. Scope of the Radical Acceptors
Figure 3. Natural population analysis of selected piperazine substrates
with the correlation between piperazine electronics and site
selectivity.
a
Isolated yields and regioselectivities are averages of two trials.
in most cases, the major product isomer was successfully
predicted using this type of NPA analysis. This predictive
model was effective even for piperazines with seemingly
electronically indistinguishable groups on nitrogen (4 and 8).
Additionally, site selectivity could be predicted regardless of
the site of alkylation, with accurate predictions for products
featuring either proximal (4 and 8−10) or distal (13, 15, and
19) functionalization to the N-Boc moiety. The sole outliers in
this study were compounds 9 and 17. Computational analysis
bearing withdrawing groups such as sulfones (20), amides
(
21), and nitriles (22). Interestingly, acceptors 20−22 gave
enhanced regioselectivity when compared with that with MVK,
with acceptors 20 and 22 giving the desired products in
excellent yields and as single regioisomers.
To further understand the driving force behind the site
selectivity observed in Schemes 1 and 2, DFT natural
population analysis (NPA) was utilized as a means to predict
product outcome. To effectively utilize NPA to predict the
major product for these reactions, computational analysis was
•+
of 9 revealed that the majority of charge density was localized
on the fluorene moiety of the Fmoc group, rather than the
piperazine, leading to product prediction that does not
C
DOI: 10.1021/acs.orglett.9b04456
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
correlate with experimental evidence. As for sulfonamide-
protected adduct 17, alkylation adjacent to the sulfonamide is
predicted as the major product. We postulate that the α-
sulfonamidoyl radical is indeed formed preferentially to the
expected α-carbamyl radical; however, due to known
decomposition pathways of α-sulfonamidoyl radicals, forma-
tion of the predicted C−H alkylation adduct is stymied,
resulting in the minor distal alkylation adduct being the only
1
9
observed product. This would also account for the
diminished yields of these substrates.
We expected that our computational results should reflect
not only the identity of the major product but also the overall
regioselectivity (rr) of the transformation. We were pleased to
observe that when the log of the regioisomeric ratio [log(rr)]
was plotted against the log of the difference between calculated
Δq values [log(ΔΔq)], a linear relationship was observed (see
the Supporting Information for computation details). The
graph in Figure 3 shows a plot of this calculated linear
relationship relative to an N-Boc group at 29 °C. Substrates
plotted in the blue portions of the graph exhibit high selectivity
under these reaction conditions with compounds on the right
favoring functionalization adjacent to N-Boc and those to the
left favoring alkylation opposite to that group. This relationship
suggests that regioselectivity is governed solely by the cation
radical character on each nitrogen and has little to do with
steric considerations or the pK of the proton α to the radical
a
cation. It is worth noting that these calculations predicted that
1
9 would be functionalized with a 48:1 rr, favoring alkylation
Figure 4. Selectivity patterns in alkylated piperazine substrates.
Isolated yields and regioselectivities are averages of two trials.
adjacent to the N-aryl group. For 19, we were able to detect
only a single regioisomer matching the predicted major
product.
When utilizing N-phenyl-N-[(trifluoromethyl)sulfonyl]-
acrylamide (21) as the radical acceptor, the observed
regioselectivities in these reactions also correlated well with
calculated Δq values. The differing slope of this line (cf. Figure
groups. This procedure showed a dependence on the reaction
temperature where site selectivity could be enhanced by
decreasing reaction temperatures. Natural population analysis
was also utilized to establish a predictive model for site
selectivity, complementing experimental results. It is our hope
that the synthetic utility of this transformation eases the access
to more structurally diverse piperazine compounds.
3
) relative to that of the plot generated using MVK as an
acceptor reflects differences in the electronic properties of
these two acceptors. In future iterations of this analysis, the
inclusion of a second parameter or set of parameters designed
to account for Michael acceptor electronics would allow for the
unification of these data into a single curve, via multivariate
analysis.
Lastly, we sought to explore how preexisting substitution
patterns along the piperazine core affect the site selectivity of
this process. Compounds 23 and 24 were prepared as single
regioisomers and diastereomers, with their structures con-
firmed by X-ray crystallography (Figure 4A). These crystal
structures illustrate the importance of an axial relationship
among substituents about the piperazine core in systems with
amide and carbamate protecting groups on the nitrogen
centers. We posit that the selectivity of these reactions is
ASSOCIATED CONTENT
Supporting Information
■
*
S
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.orglett.9b04456.
1
Experimental procedures, X-ray data, and supporting H
1
3
and C NMR spectra (PDF)
Accession Codes
CCDC 1957034−1957036 contain the supplementary crys-
tallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by
emailing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
dictated via a Fu
preinstalled methyl groups are situated in a pseudoaxial
̈
rst−Plattner-type transition state in which the
1,3
orientation to minimize buildup of A strain with the adjacent
14,20,21
carbonyl group (Figure 4B).
Although 24 was formed in
only 16% yield, we believe this is in accordance with the results
discussed thus far as we would expect selectivity to occur
adjacent to the N-Boc center; however, if these sites are
congested by preexisting functionality, we expect to see poor
conversion to an adduct resembling 24.
In conclusion, we have developed a method for the site-
selective C−H alkylation of piperazine compounds, tolerant of
a variety of orthogonal and versatile nitrogen protecting
AUTHOR INFORMATION
Corresponding Author
■
*
E-mail: nicewicz@.unc.edu.
ORCID
Nicole C. Goodwin: 0000-0002-6093-7175
David A. Nicewicz: 0000-0003-1199-9879
D
DOI: 10.1021/acs.orglett.9b04456
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors acknowledge generous funding for this project
from GlaxoSmithKline (GSK). The authors also thank Dr. Josh
Chen (The University of North Carolina at Chapel Hill) for
collection and elucidation of X-ray crystal structures, Grazyna
Graczyk-Millbrandt (GSK) for assistance with NMR studies,
and Shuping Dong (GSK) for aiding with analytical separation
techniques.
REFERENCES
■
(
(
5
1) Vitaku, E.; Smith, D. T.; Njardarson, J. T. J. Med. Chem. 2014,
7, 10257−10274.
2) Dinsmore, C. J.; Beshore, D. C. Tetrahedron 2002, 58, 3297−
3
312.
(
3) Borthwick, A. D. Chem. Rev. 2012, 112, 3641−3716.
(
4) Stucchi, M.; Cairati, S.; Cetin-Atalay, R.; Christodoulou, M. S.;
Grazioso, G.; Pescitelli, G.; Silvani, A.; Yildirim, D. C.; Lesma, G. Org.
Biomol. Chem. 2015, 13, 4993−5005.
(
(
5) Hsieh, S.-Y.; Bode, J. W. Org. Lett. 2016, 18, 2098−2101.
6) Luescher, M. U.; Bode, J. W. Angew. Chem. 2015, 127, 11034−
1
1038.
(
7) Luescher, M. U.; Vo, C.-V. T.; Bode, J. W. Org. Lett. 2014, 16,
236−1239.
8) Ruider, S. A.; Mu
013, 52, 11908−11911.
9) Zhai, H.; Borzenko, A.; Lau, Y. Y.; Ahn, S. H.; Schafer, L. L.
Angew. Chem., Int. Ed. 2012, 51 (49), 12219−12223.
10) Yar, M.; McGarrigle, E. M.; Aggarwal, V. K. Angew. Chem., Int.
Ed. 2008, 47, 3784−3786.
11) Chen, W.; Ma, L.; Paul, A.; Seidel, D. Nat. Chem. 2018, 10,
65−169.
12) Payne, P. R.; Garcia, P.; Eisenberger, P.; Yim, J. C.-H.; Schafer,
L. L. Org. Lett. 2013, 15, 2182−2185.
13) Jovel, I.; Prateeptongkum, S.; Jackstell, R.; Vogl, N.;
Weckbecker, C.; Beller, M. Chem. Commun. 2010, 46, 1956−1958.
14) McManus, J. B.; Onuska, N. P. R.; Nicewicz, D. A. Generation
1
(
2
̈
ller, S.; Carreira, E. M. Angew. Chem., Int. Ed.
(
(
(
1
(
(
(
and Alkylation of α-Carbamyl Radicals via Organic Photoredox
Catalysis. J. Am. Chem. Soc. 2018, 140, 9056−9060.
(
15) McNally, A.; Prier, C. K.; MacMillan, D. W. C. Science 2011,
3
(
2
(
34, 1114−1117.
16) Nakajima, K.; Miyake, Y.; Nishibayashi, Y. Acc. Chem. Res.
016, 49, 1946−1956.
17) Margrey, K. A.; McManus, J. B.; Bonazzi, S.; Zecri, F.;
Nicewicz, D. A. J. Am. Chem. Soc. 2017, 139, 11288−11299.
18) Ohkubo, K.; Mizushima, K.; Iwata, R.; Fukuzumi, S. Chem. Sci.
011, 2, 715−722.
19) Zhang, H.; Hay, E. B.; Geib, S. J.; Curran, D. P. J. Am. Chem.
Soc. 2013, 135, 16610−16617.
20) Chow, Y. L.; Colon, C. J.; Tam, J. N. S. Can. J. Chem. 1968, 46,
821−2825.
21) Hart, D. J. J. Am. Chem. Soc. 1980, 102, 397−398.
(
2
(
(
2
(
́
E
DOI: 10.1021/acs.orglett.9b04456
Org. Lett. XXXX, XXX, XXX−XXX