Unexpected transformations of 3-(bromoacetyl)coumarin provides new evidence for the mechanism of thiol mediated dehalogenation of α-halocarbonyls

Article abstract of DOI:10.1016/j.tetlet.2017.01.082

The mechanism for the thiol mediated dehalogenation of α-halogenated carbonyls has remained an unresolved problem, despite its ongoing application in synthetic organic chemistry. Nakamura and co-workers first proposed that net dehalogenation occurs via sequential nucleophilic substitutions, while Israel and co-workers concluded that the rate at which dehalogenation occurred suggested that dehalogenation proceeds in a single concerted step. In this study, we investigated the debromination and nucleophilic substitution of 3-(bromoacetyl)coumarin with a variety of thiophenols, whose electron donating or withdrawing natures resulted in large variations in the degree of nucleophilic substitution and dehalogenation products, respectively. Results from these experiments, in addition to an unexpected formation of thioether containing dibenzo[b,d]pyran-6-ones from a Robinson annulation, has provided new evidence for this disputed mechanism.

Full text of DOI:10.1016/j.tetlet.2017.01.082

Accepted Manuscript
Unexpected transformations of 3-(bromoacetyl)coumarin provides new evi-
dence for the mechanism of thiol mediated dehalogenation of α-halocarbonyls
Faith N. Magwenzi, Setshaba D. Khanye, Clinton G.L. Veale
PII:
S0040-4039(17)30113-2
DOI:
http://dx.doi.org/10.1016/j.tetlet.2017.01.082
TETL 48585
Reference:
Tetrahedron Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
21 December 2016
15 January 2017
23 January 2017
Please cite this article as: Magwenzi, F.N., Khanye, S.D., Veale, C.G.L., Unexpected transformations of 3-
(bromoacetyl)coumarin provides new evidence for the mechanism of thiol mediated dehalogenation of α-
halocarbonyls, Tetrahedron Letters (2017), doi: http://dx.doi.org/10.1016/j.tetlet.2017.01.082
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Tetrahedron Letters
Unexpected transformations of 3-(bromoacetyl)coumarin provides new evidence for
the mechanism of thiol mediated dehalogenation of α-halocarbonyls
Faith N. Magwenzi,a Setshaba D. Khanye,b Clinton G. L. Vealea,
aFaculty of Pharmacy, Rhodes University, Grahamstown, 6140, South Africa
bDepartment of Chemistry, Rhodes University, Grahamstown, 6140, South Africa
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
The mechanism for the thiol mediated dehalogenation of α-halogenated carbonyls has remained
an unresolved problem, despite its ongoing application in synthetic organic chemistry.
Nakamura and co-workers first proposed that net dehalogenation occurs via sequential
nucleophilic substitutions, while Israel and co-workers concluded that the rate at which
dehalogenation occurred suggested that dehalogenation proceeds in a single concerted step. In
this study, we investigated the debromination and nucleophilic substitution of 3-
(bromoacetyl)coumarin with a variety of thiophenols, whose electron donating or withdrawing
natures resulted in large variations in the degree of nucleophilic substitution and dehalogenation
products, respectively. Results from these experiments, in addition to an unexpected formation
of thioether containing dibenzo[b,d]pyran-6-ones from a Robinson annulation, has provided new
evidence for this disputed mechanism.
Available online
Keywords:
Reductive dehalogenation
Thiophenol
α-Halocarbonyl
Coumarin
2009 Elsevier Ltd. All rights reserved.
The unique chemical characteristics of both synthetic and
naturally occurring coumarin derivatives has resulted in extensive
investigation regarding their potential applications as
agrochemicals, pharmaceuticals, fluorescent dyes and
cosmetics.1–5 Accordingly, we were interested in incorporating
this scaffold into our antimalarial ethanone-α-thioether
programme.6 We reasoned that coumaryl-3-ethanone-α-thioethers
(1) could be prepared through the apparently facile nucleophilic
substitution of 3-(bromoacetyl) coumarin (2) with an appropriate
thiophenol in acetone (Scheme 1a). A similar transformation has
previously been reported using ethanol as a solvent,7 which we
opted to avoid due to our experience where excess ethanol could
possibly compete as a nucleophile. However, we did not
anticipate significant challenges in generating a small library of
compounds. Our first attempt at this reaction (Table 1, entry a1)
with unsubstituted thiophenol (3a) resulted in low conversion to
the corresponding thioether 1a, while unexpectedly a significant
reductive dehalogenation of α-haloketones is not an unknown
phenomenon,14 we noted that this transformation is still not
entirely understood.
amount of compound
2
underwent
a
net reductive
dehalogenation, forming 3-acetylcoumarin (4) with very little
detectable α-bromoketone remaining.
Chemoselective reductive dehalogenation of organo-halide
compounds has persistently evoked the interest of synthetic
organic chemists, not only for environmental applications, but
also for the controlled manipulation of chemical scaffolds.8
Consequently, the selective reduction of α-haloketones has also
proven to be a synthetic transformation of general utility which
Scheme 1. (a) Our proposed scheme for the synthesis of various thioethers.
(b) Reductive dehalogenation pathway as described by Nakamura and co-
workers (c) Reductive dehalogenation mechanism proposed by Israel and co-
workers
routinely recurs in the literature.9–13 While, thiol assisted
———
 Corresponding author. Tel.: +27-46-603-8096; fax: +27-603-7506; e-mail: c.veale@ru.ac.za

2
Nakamura and co-workers15 noted that heating α-
chlorodeoxybenzoin (5) with excess sodium methanethiolate in
ethanol unexpectedly formed deoxybenzoin (6) in near
quantitative yield, while repeating the reaction at 0 °C gave rise
to α-(methylthio)deoxybenzoin (7, Scheme 1b). Furthermore,
they determined that compound 6 could also be formed upon
heating 7 with additional methanethiolate, leading to the
reasonable conclusion that the mechanism for net reductive
dehalogenation of 5 occurs first by thioether formation, followed
by a second nucleophilic attack by excess thiolate on the
thioether sulfur. They further observed that displacement of
methanethiolate was slower using thiophenol, which they
reasoned was due to the lower nucleophilicity of thiophenol.
Israel and co-workers16 also reported a rapid and unexpected
reductive dehalogenation of 14-iodo-N-(trifluoroacetyl)-
daunorubicin in the presence of alkanethiols or thiophenol (2 eq.)
with K2CO3 in ethanol at room temperature. They further
demonstrated the same phenomenon occurred in quantitative
yields with α-iodoacetophenone (8), while repetition with either
α-bromo or chloro acetophenones resulted in formation of the
corresponding thioethers. They concluded from their study, that
the respective halogen reactivity influences the course of the
reaction toward either substitution or dehalogenation.
Furthermore, they hypothesized that the rate at which
dehalogenation occurred precluded the pathway proposed by
Nakamura and co-workers leading them to offer an alternative
mechanistic explanation, in which dehalogenation occurs in a
single discreet step (Scheme 1c).
In order to gain further mechanistic insight, several
experiments were repeated with lower thiophenol equivalents
(1.2 eq. entries a2–d2, f2 and h2). Here we generally observed
similar levels of dehalogenation, and in instances where lower
dehalogenation was seen, this corresponded with a greater
proportion of α-bromoketone remaining rather than thioether
formation. Of greatest significance in this series was entry d2
where dehalogenation increased in the presence of lower
equivalents of para-amino thiophenol. Importantly, no reaction
was observed when no thiophenol was added (Entry j).
Table 1.a
Entry
Thiophenol
1b
2b
4b
a1c
21
3
2
23
77
74
a2d
b1
b2
17
15
3
9
80
76
c1
c2
46
9
0
54
56
35
We reasoned that apart from halogen reactivity, the primary
differences between the mechanisms of Nakamura and Israel lie
in the relative ability of the thiol to act as a nucleophile, therefore
requiring two equivalents of thiophenol (Nakamura) or a proton
donor, leading to enol formation, requiring one thiophenol
equivalent (Israel). Hudson and Klopman had previously
demonstrated that electron withdrawing and donating groups
decrease and increase nucleophilicity of thiophenols,
respectively.17 Accordingly, we embarked on an exploratory
study to further understand the role that the electronic
environment of thiophenol would have on the reductive
dehalogenation of 2.
d1
d2
88
48
0
0
12
52
e1
e2
85
5
10
NA
NA
NA
f1
f2
6
5
0
94
73
22
g1
g2
0
12
88
NA
NA
NA
h1
h2
14
17
1
7
85
76
Our first group of experiments (Table 1, entries b1–e1)
focused on increasing the electron density on the thiophenol
sulfur using electron donating substituents. Initially, a negligible
influence on conversion was observed using the weakly electron
donating para-methyl group (Entry b1), while thiophenols
containing para-methoxy, amino and dimethylamino moieties
(Entries c1–e1), which are able to donate electron density
through resonance, reversed this trend. In particular, the
increased nucleophilicity of the para-amino and dimethylamino
thiophenols resulted in satisfying conversion to the respective
thioethers (1d and 1e), coupled with low levels of
dehalogenation.
i1
i2
0
0
100
NA
NA
NA
j
None
0
100
0
aReagents and conditions: Compound 2 (100 mg, 0.374 mmol, 1 eq.),
acetone (10 mL), thiophenol (2 or 1.2 eq.), reflux, 5 h. bPercent
conversion determined by 1H NMR integrals. cThiophenol (2 eq.).
dThiophenol (1.2 eq.).
Overall a general pattern was observed, where strongly
electron withdrawing substituents on the thiophenol ring resulted
in greater reductive dehalogenation compared to strongly electron
donating substituents. In the context of Nakamura’s nucleophilic
displacement theory which cited the lower nucleophilicity of
thiophenol as a factor contributing to decreased dehalogenation,
one would expect that stronger nucleophiles such as 3d and 3e
would display superior dehalogenation, while conversely, para-
nitro thiophenol possessing far less electron density at the sulfur,
should not dehalogenate to such an extent. Furthermore, this
mechanism requires a second equivalent of thiophenol to
complete dehalogenation. Therefore, in experiments conducted
with 1.2 eq. of thiophenol a greater proportion of thioether would
be expected in the reaction mixture, having been unable to
undergo the second nucleophilic attack. However, our results
indicated that similar levels of dehalogenation were achievable
with fewer equivalents of thiophenol, a result more congruent
Accordingly, we then turned our attention to the influence that
electron withdrawing groups, and the consequent reduction in
sulfur electron density, might have on the relative conversion of
2. In contrast to entries d1 and e1, the presence of a strongly
electron withdrawing para-nitro group (Entry f1) resulted in
significant conversion to 4, with no trace of the starting α-
bromoketone and only a small quantity of thioether 1f. Similarly,
para-chloro thiophenol (Entry g1) resulted in significant
dehalogenation, with no detectable thioether 1g. Interestingly,
however, was the presence of a low yet still appreciable quantity
of unreacted starting material 2, indicating, that the capacity for
reductive dehalogenation did not necessitate an increase in
nucleophilic character. The reaction, with meta-chloro thiophenol
(Entry h1) resulted in a similar degree of dehalogenation, this
time with the formation of a small amount of thioether 1h, while
ortho-chloro thiophenol (Entry i1) resulted in complete
dehalogenation.

Table 2.a Yields of 1d and 5 a-e
with the proposed mechanism of Israel and co-workers, and as
mentioned, lower levels of dehalogenation did not correspond
with increased thioether being present, but rather, the presence of
unreacted starting material. Finally, in the case of entry d2,
lowered equivalents of a demonstrably good nucleophile resulted
in increased dehalogenation, further suggesting that the
dehalogenation and substitution mechanisms are unrelated in this
system.
Compound
Structure
Yieldb (%)
quant.
1d
Having struggled to initiate bromine substitution with several
thiophenols, we were curious as to the effect that stoichiometric
quantities of K2CO3 would have on this reaction, possibly
allowing the desired nucleophilic displacement to outcompete
dehalogenation (Scheme 2). We first attempted this modified
reaction with the best performing nucleophile from the previous
study, para-amino thiophenol. Here the addition of K2CO3
resulted in complete conversion to thioether 1d, seemingly
suppressing the small levels of dehalogenation observed in the
absence of base. Accordingly, this reaction was repeated using
several thiophenols (3a, f–i) which had displayed limited
capacity as nucleophiles.
1H NMR analysis of the crude reaction mixtures offered no
clear evidence as to the presence of the α-bromoketone, nor the
substitution and dehalogenation products. However, purification
of the respective crude mixtures, led to the isolation of a series of
unreported C-8 thiophenol functionalised dibenzo[b,d]pyran-6-
ones in low yields (5a–e, Table 2). Related dibenzo[b,d]pyran-6-
one scaffolds are found in several pharmaceutically interesting
natural products,18 find important applications in physical
chemistry,19 and have been the subject of a number of recent
synthetic procedures.20–24
10
4
5a
5b
5c
5d
5e
13
12
13
aReagents and conditions: Compound 2 (100 mg, 0.374 mmol, 1 eq.),
acetone (10 mL), thiophenol (2 eq.), K2CO3 (2 eq.) reflux, 5 h.
bIsolated yields.
Koelsch and Sundet previously reported the formation of 7-
hydroxy-9-coumarinyl-6H-benzo[c]chromen-6-one via a Michael
type condensation between two subunits of 4 when heated at
reflux in an ethanol solution containing piperidine.25 This
allowed us to propose a mechanism for the formation of 5a–e as
depicted in Scheme 2. Firstly, the addition of base facilitates
nucleophilic substitution of the bromine atom, leading to the
formation of an α-thiocarbonyl (1). Subsequent Robinson
annulation between the α,β-unsaturated system of the 3-
acylcoumarin and an enolate formed from acetone gives
compound 6. The electrophilicity of the α,β-unsaturated system is
enhanced through chelation between the potassium cations and
the 1,3-dicarbonyl system. Following addition of acetone, the
strong electron withdrawing nature of the thiophenol moiety
renders the adjacent methine protons slightly acidic, thereby
facilitating deprotonation and subsequent enolate formation. The
resultant enolate is then free to nucleophilically attack the
carbonyl leading to cyclisation. This is followed by dehydration
to complete the Robinson annulation, and subsequent
aromatisation results in the formation of phenol, as reported by
Koelsch and Sundet, to yield compound 5. This final
aromatisation step was described by Koelsch and Sundet as a
spontaneous intramolecular aldolization and dehydrogenation, in
addition to the dehydration step. While the mechanism involved
here is currently uncertain, it is likely an air facilitated oxidation.
Moriuchi and co-workers.26 have explored a vanadium catalysed
oxidative aromatisation of 2-cyclohexenones in the presence of
atmospheric oxygen and a source of bromide. They also showed
that this reaction could occur in low yield without the catalyst,
which may provide a clue as to what is occurring here.
Three further control experiments were conducted, where 1)
thiophenol was omitted, 2) TEA replaced K2CO3 and 3) the
reaction was performed with methyl ketone 4 instead of 2.
Interestingly, in our hands, we did not observe the formation of
the corresponding dibenzo[b,d]pyran-6-one in any of these
reactions. We believe these observations tentatively support the
notion that potassium chelation enhances the electrophilicity of
the α,β-unsaturated system and that the electron withdrawing
nature of the thiophenol facilitates the second enolate formation.
Additionally, we did not observe cyclization in the presence of
para-amino thiophenol, which we hypothesize is due to its
electron donating capabilities which hinders methine
deprotonation. While this transformation is potentially useful it
also provides additional insight into the mechanism of
debromination observed here.
The study described herein, was used to further develop the
understanding of the thiol mediated dehalogenation of α-bromo
carbonyls. A previous related study conducted by Israel and co-
workers proposed that removal of iodine from an α-iodocarbonyl
preferentially occurs in a single step, which does not require a
strongly nucleophilic thiol species. However, this mechanism is
incongruent with the observations of Nakamura and co-workers ,
and more recently Liu and co-workers,14 who both showed that
removal of chlorine from α-chlorocarbonyls occurs first by
nucleophilic substitution. These seemingly contrasting
mechanisms can be plausibly explained in the context of
Pearson’s hard and soft acid base theory.27 The thiophenol sulfur
is considered a ‘soft’ base which would preferably interact with
the soft iodine, facilitating single step dehalogenation, whereas,
sulfur is less likely to interact with the harder chlorine, therefore
preferentially resulting in nucleophilic substitution as the initial
step.

4
Acknowledgments
Financial support for this study was provided by Rhodes
University. FNM gratefully acknowledges financial support from
the Rhodes University Henderson Foundation.
Supplementary data
Supplementary data relating to experimental procedures and
spectral data can be found in the online version.
References
(1) Calcio Gaudino, E.; Tagliapietra, S.; Martina, K.;
Palmisano, G.; Cravotto, G. RSC Adv. 2016, 6, 46394–
46405.
(2) Qian, R.; Tong, H.; Huang, C.; Li, J.; Tang, Y.; Wang, R.;
Lou, K.; Wang, W. Org. Biomol. Chem. 2016, 14, 5007–
5011.
(3) Hoeksema, H.; Johnson, J. L.; Hinman, J. W. J. Am. Chem.
Soc. 1955, 53, 6710–6711.
(4) Belluti, F.; Perozzo, R.; Lauciello, L.; Colizzi, F.;
Kostrewa, D.; Bisi, A.; Gobbi, S.; Rampa, A.; Bolognesi,
A. L.; Recanatini, M.; R. B.; Scapozza, L.; Cavalli, A. J.
Med. Chem. 2013, 56, 7516–7526.
(5) Zhao, H.; Neamati, N.; Hong, H.; Mazumder, A.; Wang, S.;
Sunder, S.; Milne, G. W. A.; Pommier, Y.; Burk Jr., T. R. J.
Med. Chem. 1997, 40, 242–249.
(6) Svogie, A. L.; Isaacs, M.; Hoppe, H. C.; Khanye, S. D.;
Veale, C. G. L. Eur. J. Med. Chem. 2016, 114, 79–88.
(7) Lin, W.; Cao, X.; Yuan, L.; Ding, Y. Chem. Eur. J. 2010,
16, 6454–6457.
Scheme 2. Proposed reaction pathway for the formation of compounds 5a-e
(8) Alonso, F.; Beletskaya, I. P.; Yus, M. Chem. Rev. 2002,
Evidence obtained in this study demonstrated that a mixture of
substitution and debromination products were obtained in
varying degrees depending on the electronic environment of the
respective thiophenol, and that dehalogenation was not related to
thiol nucleophilicity. Of particular importance was the
observation that similar levels of dehalogenation were observed
with lowered equivalents of the least nucleophilic thiophenols.
102, 4009–4091.
(9) Borah, H. N.; Boruah, R. C.; Sandhu, J. S. J. Chem. Soc.,
Chem. Commun. 1991, 154–155.
(10) Wang, Y. C.; Yan, T. H. Chem. Commun. 2000, 545–546.
(11) Molander, G. A.; Hahn, G. J. Org. Chem. 1986, 51, 1135–
1138.
(12) Narayanam, J. M. R.; Tucker, J. W.; Stephenson, C. R. J. J.
Am. Chem. Soc. 2009, 131, 8756–8757.
(13) Choi, H. Y.; Chi, D. Y. Org. Lett. 2003, 5, 411–414.
(14) Dong, W.-L.; Cai, W.-X.; Wu, R.; Li, Z.-M.; Zhao, W.-G.;
Liu, X.-H. Phosphorus Sulfur Silicon Relat. Elim. 2016,
191, 980–983.
(15) Ōki, M.; Funakoshi, W.; Nakamura, A. Bull. Chem. Soc.
Jpn. 1971, 44, 828–832.
(16) Seshadri, R.; Pegg, W. J.; Israel, M. J. Org. Chem 1981, 46,
2596–2598.
(17) Hudson, R. F.; Klopman, G. J. Chem. Soc 1962, 1062–
1067.
This suggests that the mechanism of debromination observed
here mirrors that of Israel’s de-iodination, whereby soft sulfur
interacts with the relatively soft bromine. Additionally, strong
electron withdrawing groups present on the reductive thiophenols
decreases the pKa, thereby facilitating proton donation and enol
formation, ultimately resulting in dehalogenation. Equally, an
increase of thiol nucleophilicity either through electron donating
groups, or the introduction of base, favours nucleophilic
displacement of bromine, rather than dehalogenation.
In conclusion, we have gained new insight into an unresolved
problem in chemistry, relating to the mechanism of thiol induced
dehalogenation of α-haloketones. We observed significant
changes in the degrees of dehalogenation and/or nucleophilic
substitution of α-bromoketone 2 in the presence of variably
substituted thiophenols. Furthermore, similar levels of
dehalogenation were observed with lowered equivalents of
thiophenol. These reactants were chosen to influence their
respective electronic environments and as such provided
evidence that the mechanisms of nucleophilic substitution and
dehalogenation are not necessarily related, but can rather be
influenced by subtle variations in the electronic environments of
the respective reactants. This explanation also provides a feasible
explanation for the differences observed by Nakamura and Israel,
respectively. With respect to 3-(bromoacetyl) coumarin,
reductive dehalogenation occurs in a single step in the presence
of softer bases rather than requiring an initial nucleophilic
substitution. Finally, the addition of stoichiometric quantities of
K2CO3, resulted in small, yet isolatable quantity of new
thiophenol containing dibenzo[b,d]pyran-6-ones.
(18) Hussain, I.; Nguyen, V. T. H.; Yawer, M. A.; Dang, T. T.;
Fischer, C.; Reinke, H.; Langer, P. J. Org. Chem. 2007, 72,
6255–6258.
(19) Fletcher, S. P.; Dumur, F.; Pollard, M. M.; Feringa, B. L.
Science 2005, 310, 80–82.
(20) Suárez-Meneses, J. V.; Oukhrib, A.; Gouygou, M.;
Urrutigoïty, M.; Daran, J.-C.; Cordero-Vargas, A.; Ortega-
Alfaro, M. C.; López-Cortés, J. G. Dalt. Trans. 2016, 45,
9621–9630.
(21) Dai, J.-J.; Xu, W.-T.; Wu, Y.-D.; Zhang, W.-M.; Gong, Y.;
He, X.; Zhang, X.-Q.; Xu, H.-J. J. Org. Chem. 2015, 80
(80), 911–919.
(22) Yang, W.; Wang, J.; Wei, Z.; Zhang, Q.; Xu, X. J. Org.
Chem. 2016, 81, 1675–1680.
(23) Ramirez, N. P.; Bosque, I.; Gonzalez-Gomez, J. C. Org.
Lett. 2015, 17, 4550–4553.
(24) Tibrewal, N.; Pahari, P.; Wang, G.; Kharel, M. K.; Morris,
C.; Downey, T.; Hou, Y.; Bugni, T. S.; Rohr, J. J. Am.
Chem. Soc. 2012, 134, 18181–18184.
(25) Koelsch, C. F.; Sundet, S. A. J. Am. Chem. Soc. 1950, 72,
1844–1845.

(26) Moriuchi, T.; Kikushima, K.; Kajikawa, T.; Hirao, T.
Tetrahedron Lett. 2009, 50 (52), 7385–7387.
(27) Pearson, R. G. J. Am. Chem. Soc. 1963, 85, 3533–3539.

6
Highlights

This work provides new insight into the disputed
mechanism of thiol mediated dehalogenation of α-
halogenated carbonyls


Dehalogenation can be manipulated by alteration of
thiophenol substitution
A new synthesis of unknown thiolated benzopyranones
has been discovered

Graphical Abstract
To create your abstract, type over the instructions in the
template box below.
Unexpected transformations of 3-
Leave this area blank for abstract info.
(bromoacetyl)coumarin provides new
evidence for the mechanism of thiol mediated
dehalogenation of α-halocarbonyls
Faith N. Magwenzi, Setshaba D. Khanye, Clinton G. L. Veale
Fonts or abstract dimensions should not be changed or altered.