Organosulfur compounds are arguably among the most important functionalities in current
organic chemistry.[1] Particularly, sulfides have shown exciting properties in materials science,[2]
[3] as bioactive natural or synthetic compounds,[4] and in metabolic studies of sulfoxide and sulfone drugs.[5,6] Despite the prodigious variety of methodologies described in the literature to obtain
sulfides, the reduction of sulfoxides is still of common interest to the synthetic
community.[7]
[8] Scheme [1] summarizes the general outcome of the most traditional methodologies used for this
purpose: 1) Catalytic hydrogenations, which usually involve the use of transition
metal complexes, and high-pressure hydrogen;[9] 2) Use of boron[10] and silicon hydrides,[11] which also need a transition metal catalyst; 3) Deoxygenations, where electrophiles
are generally used to activate the sulfoxide forming a sulfonium salt, which uses
a nucleophilic scavenger to afford the corresponding sulfide and the oxygenated electrophile;[12] 4) Electrochemical methods, which are believed to involve electrophilic activation
and an electron transfer process that induces the loss of oxygen;[13] and 5) Photocatalytic processes, which are thought to proceed via radical fragmentation
of the sulfoxide.[14]
[15]
Scheme 1 Methodologies commonly used in sulfoxide reduction to sulfides
Although those methodologies are helpful, all of them still have some critical issues
to solve. Notably, using homogenous metal catalysts, usually high cost, is inconvenient
for large-scale preparations, in addition to the incompatibilities with other functional
groups susceptible to reduction. Also, when using transition metals, there is always
the risk of the product having trace metals, which is undesirable in the pharmaceutical
industry.[16] On the other hand, the use of silicon and boron-based hydrides, as well as the use
of electrophilic reagents, inescapably need chromatographic purification processes.
Finally, most of those methodologies must be carried out at low temperatures because
of the stability of the electrophilic reagents or to increase the chemoselectivity.
The reduction of sulfoxides takes on an even greater value when it leads to sulfides
with outstanding biological activity. This function is present in molecules such as
the benzimidazole family, with important pharmacological activities[17] principally in human and veterinary medicine to treat helminthic infections.[18] Among this family, some examples of the sulfoxide-sulfide pair of significant pharmacological
interest are ricobendazole-albendazole, oxfendazole-fenbendazole and triclabendazole
sulfoxide-triclabendazole (Figure [1]). Both the sulfides and the respective sulfoxides in Figure [1] (belonging to the family of benzimidazoles) present mainly anti-helminthic activity
and have been studied as potential antitumor agents.[19]
Figure 1 Different bioactive couples of sulfoxide-sulfide benzimidazole for helminthic treatments
Considering that a common limitation in the pharmaceutical market of sulfoxides is
chirality, sometimes this weakness can be circumvented by using the respective sulfide
as a prodrug when oxidation to the respective sulfoxide takes place in vivo. Albendazole is a representative case of this behavior, undergoing rapid oxidation
to chiral ricobendazole and subsequent oxidation to the respective sulfone, considered
inactive.[5]
Studies of previously mentioned bioactive molecules containing the sulfide functional
group are developed, especially studies focused on new activities[20] and resistance to benzimidazoles and their metabolites.[21]
[22]
[23]
[24]
[25] On this basis, several teams have studied the reduction of sulfoxides, proposing
alternative methods, such as electrochemical approaches[26] or heterogeneous catalysis.[27] Unfortunately, even if the new procedures provide high chemoselectivity, the use
of chromatographic purification and access to specialized equipment or catalysts are
still substantial limitations. Consequently, we believe that developing new sulfoxide
reduction methodologies is particularly important for the organic chemistry community.
Even more so, if these methodologies are characterized by their robustness, easy purification,
and treatment of the products, if they are scalable, use low-cost materials, and are
performed metal-free.
In 2020, we reported a new methodology for the chemoselective reduction of sulfoxides
through the combined use of oxalyl chloride and trimethoxybenzene (TMB).[28] The electrophilic reagent allows the formation of an intermediate sulfonium salt.
The electrophilic chlorine is subsequently trapped by a nucleophilic, oxidizable functional
group (TMB), leading to the formation of TMB-Cl in the reaction medium, as shown in
Scheme [2].
Scheme 2 Sulfoxide reduction using the couple (COCl)2/TMB
Even if this method has the most significant scope described to date, the by-product
TMB-Cl is the main disadvantage of this methodology since, in some cases, it complicates
the purification process of the target sulfide. On the other hand, chromatographic
purification is mandatory. Based on this, we have visualized new synthetic approaches
that generate a volatile by-product making the reaction attractive from a practical
and economic point of view because facile purification and workup proceed with excellent
yields.[29]
Herein, we describe a new approach where the sulfoxide reduction is performed by dissolving
it in anhydrous DCM and adding 1.5 equivalents of triethylsilane, followed by slow
addition of oxalyl chloride under an inert atmosphere (N2). Once gas evolution has ended after adding oxalyl chloride, the starting sulfoxide
has generally been completely consumed.
Table 1 Solvent Optimization
|
|
Entry
|
Solvent
|
Time (min)
|
Yield (%)
|
Purity (%)
|
|
1
|
Et2O
|
60
|
95
|
89
|
|
2
|
CH2Cl2
|
15
|
96
|
96
|
|
3
|
THF
|
180
|
100
|
87
|
|
4a
|
CH2Cl2
|
15
|
97
|
95
|
a The experiment was carried out without N2 atmosphere.
To identify the optimal solvent, we carried out the reaction in diethyl ether and
THF, in addition to CH₂Cl₂, and evaluated the yields and product purity, as summarized
in Table [1]. The reaction proceeded successfully in all solvents tested, even under open-vessel
conditions (Table [1], entry 4). However, the best yield-to-purity ratio was achieved in CH₂Cl₂ under
a nitrogen atmosphere (entry 2). If the substrate is not soluble in CH₂Cl₂, the reaction
can be performed in ether solvents (entries 1 and 3), although this typically requires
longer reaction times and additional purification steps.
The above procedure reduces different sulfoxides in excellent yields, generally from
the evaporation at a reduced pressure of the solvent and volatile by-products of the
crude reaction. Thus, ricobendazole is reduced to albendazole-HCl in 98 ± 1.5% isolated
yield with a purity of 96 ± 1.0% determined by qNMR. Although the hydrochloride is
of interest due to its better solubility (which increases bioavailability) than the
neutral free form;[30] if neutral albendazole is needed, a simple workup with a saturated aqueous NaHCO3 solution allows obtaining albendazole in high yields and purity (98% and 99%, respectively)
as determined by qNMR. Likewise, this procedure enables various sulfides, including
derivatives with alkyl substituents, aryl-aryl, and alkyl-aryl (compounds 2–12 in Figure [2]), among which chlorbenside (5) stands out as an essential sulfide with acaricidal activity obtained in quantitative
yield. The important and well-known sulfoxide Omeprazole is reduced in excellent yield
to the corresponding sulfide 6. It is also important to highlight some limitations. For instance, only a partial
reduction is observed when strongly electron-withdrawing groups are substituting the
sulfur (compound 4 in Figure [2]), probably due to the lower nucleophilicity of the sulfoxide oxygen, which leads
to complex formation of the intermediate chlorosulfonium salt and, consequently, lower
conversion. Nevertheless, the unreacted starting material can be quantitatively recovered
without any change. Heterocyclic sulfoxides (compounds 7 and 9) were also reduced in excellent yields; however, when using sulfoxides that react
via thionium ions, the yield significantly drops.[31]
Figure 2 Selected scope of sulfide products in the sulfoxide reduction through the couple
(COCl)2/Et3SiH
We then turned our attention to evaluating the chemoselectivity of the reaction. To
this end, we employed esters and aromatic iodides, which showed no reactivity under
the reaction conditions; compound 10 was isolated in 94% yield. Terminal alkynes exhibited slight reactivity and compound
11 was obtained in 72% yield; however, small amounts of unidentified by-products were
also observed. Finally, aldehydes proved to be fully compatible, as compound 12 was isolated in excellent yield (92%).
Finally, due to the difficulty in obtaining starting materials containing both functional
groups, we conducted competition experiments to evaluate the chemoselectivity of the
reaction against ketones, sulfones, and aryl bromides. In all cases, only the reduction
of the sulfoxide was observed, and the competing compounds were isolated in yields
exceeding 88%, demonstrating the excellent compatibility of our method with these
functional groups (Scheme [3]).
Scheme 3 Competition experiments under standard conditions
Scheme 4 Mechanism proposal
On the other hand, GC-MS analysis of the crude reaction mixtures revealed the formation
of Et₃SiCl (see Supporting Information) as a by-product. This observation, along with
the clear formation of HCl (products 2, 6, and 7), led us to propose the reaction mechanism shown in Scheme [4]. The proposed mechanism builds upon our previously described pathway and the confirmed
formation of the chlorosulfonium salt.[28] In this mechanism, hydride acts as a chloride scavenger, generating HCl. The chloride
(sulfonium counterion) then behaves as a nucleophile. This increases the nucleophilicity
of the Si–H bond, ultimately leading to the formation of the Et₃SiCl by-product. Further
in-depth studies on the reaction mechanism are currently underway in our laboratory
and will be reported in due course.
In summary, we have developed a methodology for reducing sulfoxides to sulfides through
the Et3SiH/(COCl)2 couple. The reaction proceeds with excellent yields without metal catalysts and generally
with an easy purification step (evaporation of volatiles). Additionally, it was possible
to obtain albendazole hydrochloride from ricobendazole and other sulfides that demonstrate
the wide scope of the reaction.
Methyl 5-(Propylthio)-2-benzimidazolecarbamate Hydrochloride (Albendazole-HCl, 2);
Typical Procedure
To an oven-dried 250 mL round-bottom, single-necked flask (Quickfit ground-glass stopper
24/29), provided with an egg-shaped Teflon-coated stirring bar (3.5 cm) was added
ricobendazole (1; 3.00 g, 10.13 mmol, 1 equiv.). Subsequently, the flask was equipped with a three-way
stopcock with septa and connected through a chemically resistant hose to a vacuum/N2 manifold line (Figure [3]A).
Figure 3 Reduction of ricobendazole (1) to albendazole-HCl (2): A) single-neck flask equipped with a three-way stopcock and connected through a
chemically resistant hose to a vacuum/N2 manifold; B) Suspension of ricobendazole connected to mineral oil bubbler leading
to sodium bicarbonate solution; C) Crude after oxalyl chloride addition.
Three evacuation cycles replace the system atmosphere with N2 (stopcock in position 1). Next, the three-way stopcock was closed (stopcock in position
3), and the connection to the manifold was replaced with a link to a mineral oil bubbler,
which leads to a saturated aqueous sodium bicarbonate solution (30 mL) contained in
a 50 mL Erlenmeyer flask (Figure [3]B). With the stopcock in position 2, the DCM was added through the septa (68.0 mL)
in four portions via a 20 mL plastic syringe equipped with a 21-gauge stainless-steel
needle. The white suspension obtained (Figure [3]B) was stirred at 250 rpm and kept at rt (20 °C). Et3SiH (2.48 mL, 15.2 mmol, 1.5 equiv.) was added through the septum via a 5 mL plastic
syringe equipped with a stainless-steel needle in a single portion. The stirring was
increased to 750 rpm, and oxalyl chloride (1.35 mL, 15.2 mmol, 1.5 equiv.) was added
in approximately 1 min via a 2 mL plastic syringe with a stainless-steel needle through
the septa. While the addition proceeded, an intense bubbling was observed, the homogenization
of the solution and progressive evolution of the reaction mixture to orange color
were also observed. The color persisted when the addition was complete (Figure [3]C). Once the bubbling stopped (after approximately 50 seconds), the stopcock was
guided to position 3. After 15 min of stirring at rt, TLC revealed that the starting
material had been completely consumed (Figure [4]A). The elution was realized using EtOAc, Rf
= 0.26 and 0.58 for ricobendazole (1) and albendazole-HCl (2), respectively; and visualized under 254 nm UV-lamp.
Figure 4 Ending of the reaction: A) TLC showing total consumption of ricobendazole (1: ricobendazole;
2: a mixture of ricobendazole + crude reaction; 3) crude reaction. B) Final aspect
of the reaction.
An additional stirring for 20 min was allowed to elapse, the time during which the
precipitation of the product and slight yellowing of the reaction were observed (Figure
[4]B). Next, the stirring was stopped, all the connections removed, and the crude (Figure
[5]A) was concentrated at reduced pressure in a rotary evaporator (500 mbar and 45 °C
bath temperature). Once the solvent was removed, one portion of acetone (15.0 mL)
was added, and the crude was carried out to dryness again. First, at 500 mbar and
45 °C bath temperature for 5 min, then 30 mbar and 45 °C bath temperature for 30 min.
Later, one portion of CH2Cl2 (15.0 mL) was added, and the crude was carried out to dryness again. First, at 500
mbar and 45 °C bath temperature for 5 min, then 30 mbar and 45 °C bath temperature
for 30 min. Both processes were repeated three times (this to drag all volatile by-products).
After this process, the solid became much lighter and had a drier appearance (Figure
[5]B). The weight of the flask revealed that 3.20 g of albendazole-HCl (2; 99% yield and 97% purity) was obtained as a stable and slightly yellow amorphous
solid in good purity without additional treatment; mp 175–177 °C.
IR (ATR): 3033, 2963, 2870, 1751, 1636, 1578, 1489, 1439, 1312, 1238, 1088, 922, 756,
733, 594, 501 cm–1.
1H NMR (DMSO-d
6, 400 MHz): δ = 7.55–7.59 (m, 2 H), 7.33 (dd, J = 8.4, 1.8 Hz, 1 H), 3.86 (s, 3 H), 2.93 (t, J = 7.2 Hz, 2 H), 1.56 (q, J = 7.3 Hz, 2 H), 0.96 (t, J = 7.3 Hz, 3 H).
13C NMR (DMSO-d
6, 101 MHz): δ = 152.9, 144.4, 131.6, 130.7, 128.5, 125.4, 113.9, 113.4, 53.6, 35.4,
21.9, 13.1.
HRMS (ESI): m/z [M – Cl]+ calcd for C12H16N3O2S+: 266.0958; found: 266.0965.
Figure 5 Removal of volatile by-products. A) Appearance of the reaction crude after the DCM
washes; B) Final appearance of albendazole-HCl.
Albendazole
If neutral albendazole is needed, a simple workup with a saturated NaHCO3 aqueous solution allows to obtain albendazole.
The albendazole-HCl (2; 3.20 g, 11.2 mmol) was poured into an Erlenmeyer containing aq NaHCO3 (941 mg, 11.2 mmol in 34 mL of H2O) and stirred for 6 h at rt and left in the fridge at 0 °C overnight. Then, the solid
was filtered and washed with cold H2O (26 mL). The solid was dried at 45 °C and 20 mbar for 2 h, affording 2.91 g (98%
yield and 99% purity) of albendazole as a pale yellow solid; mp 205–206 °C .
IR (neat): 3005, 2958, 1620, 1558, 1442, 1265, 1091 cm–1.
1H NMR (DMSO-d
6, 400 MHz): δ = 11.71 (br s, 2 H), 7.42 (s, 1 H), 7.34 (d, J = 8.2 Hz, 1 H), 7.10 (d, J = 8.2, Hz, 1 H), 3.75 (s, 3 H), 2.85 (t, J = 7.1 Hz, 2 H), 1.44–1.59 (m, 2 H), 0.94 (t, J = 7.3 Hz, 3 H).
13C NMR (DMSO-d
6, 100 MHz): δ = 154.8, 147.8, 126.8, 124.1 1, 115.8, 114.2, 52.5, 36.7, 22.1, 13.1
(C=O not observed).
HRMS (ESI): m/z [M + H]+ calcd for C12H17N3O2S+: 266.0958; found: 266.0965.
