[1,3]/[1,4]-sulfur atom migration in β-hydroxyalkylphosphine sulfides

Article abstract of DOI:10.3762/bjoc.16.11

β-Hydroxyalkylphosphine sulfides undergo [1,3]- or [1,4]-sulfur atom phosphorus-to-carbon migration in the presence of Lewis or Br?nsted acids. The direction of sulfur atom migration depends on the type of acid used for the reaction. In the presence of a

Full text of DOI:10.3762/bjoc.16.11

[1,3]/[1,4]-Sulfur atom migration in
β-hydroxyalkylphosphine sulfides
Katarzyna Włodarczyk1, Piotr Borowski2 and Marek Stankevič*1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2020, 16, 88–105.
1Department of Organic Chemistry, Faculty of Chemistry, Marie
Curie-Skłodowska University in Lublin, 33 Gliniana St., 20-614 Lublin,
Poland and 2Department of Chromatographic Methods, Faculty of
Chemistry, Marie Curie-Skłodowska University in Lublin, 3 Marie
Curie-Skłodowska Sq., 20-031 Lublin, Poland
doi:10.3762/bjoc.16.11
Received: 28 September 2019
Accepted: 10 January 2020
Published: 21 January 2020
Email:
Associate Editor: J. A. Murphy
Marek Stankevič* - marek.stankevic@poczta.umcs.lublin.pl
© 2020 Włodarczyk et al.; licensee Beilstein-Institut.
* Corresponding author
License and terms: see end of document.
Keywords:
Brønsted and Lewis acids; DFT; mechanistic studies; rearrangement;
stereochemistry; sulfur atom migration
Abstract
β-Hydroxyalkylphosphine sulfides undergo [1,3]- or [1,4]-sulfur atom phosphorus-to-carbon migration in the presence of Lewis or
Brønsted acids. The direction of sulfur atom migration depends on the type of acid used for the reaction. In the presence of a
Brønsted acid, mainly [1,3]-rearrangement is observed, whereas a Lewis acid catalyzes the [1,4]-sulfur migration. To gain insight
into the mechanism of these transformations, the stereochemistry of these rearrangements have been tested, along with the conduc-
tion of some control experiments and DFT calculations.
Introduction
Among all organic reactions, rearrangements are an exciting ([1,2]-rearrangement) [38-42], and [2,3]-sigmatropic rearrange-
class of transformations where unusual or even unexpected ments of propargyl and allylphosphinites [43-46].
products can be obtained. Many rearrangements are of practical
use in synthetic organic chemistry, including Beckmann [1-4], Other [1,2]-rearrangements include α,β-epoxyphosphonate
Claisen [5-8], pinacol [9-12], Wagner–Meerwein [13-16], opening [47], β-halo-α-hydroxyphosphonate rearrangement
Curtius [17-20], Hofmann [21-24], Overmann [25-28] rear- [48], phosphirene rearrangement [49], or the unusual phos-
rangements, and many others. In organophosphorus chemistry, phonate–phosphinate rearrangement of dimethyl phosphonates
rearrangements are less developed transformations, however, [50]. Other [1,3]-rearrangements of organophosphorus com-
some examples can be found in the literature (Scheme 1). These pounds include phosphoenolpyruvate formation from phospho-
include the famous Arbuzov rearrangement [29-32] as well as nopyruvate [51], benzylphosphonium salt formation from the
phospha-Fries ([1,3]-rearrangements) [33-37], phospha-Brook corresponding o-methylaryl-substituted precursors [52], or the
88

Beilstein J. Org. Chem. 2020, 16, 88–105.
Scheme 1: Arbusov, phospha-Fries, and phospha-Brook rearrangements.
formation of ketophosphonates from vinylphosphates [53]. Depending on the structure, the formation of either the phos-
[1,4]-Rearrangements include that of o-phosphorus-substituted phaindane 3 or the benzophosphorinane 2 skeleton was ob-
benzyl carbanions [34], phosphorus group migration in O-phos- served. This method could also be performed in a stereoselec-
phorylated 1,4-benzodiazepines [54], or phosphoryl group car- tive manner, given that corresponding chiral substrates were
bon-to-oxygen transfer [55]. The common feature of every rear- used for the cyclization. The chiral compounds suitable for
rangement presented above is the cleavage of one single bond cyclization could be obtained by desymmetrization of phos-
between phosphorus and either carbon or oxygen atom, while phine sulfides (Scheme 3) [58].
the multiple bond remains intact.
Herein, we present the results concerning an unusual transfor-
mation of β-hydroxyalkylphosphine sulfides, which undergo
[1,3]- or [1,4]-rearrangement in the presence of an acid,
yielding the corresponding β/γ-mercaptoalkylphosphine oxides.
In this rearrangement, a sulfur atom transfers from phosphorus
to carbon, whereas the phosphorus–carbon bonds remain intact.
Scheme 3: The synthesis of P-stereogenic β-hydroxyalkylphosphine
sulfides.
Depending on the type of acid, sulfur atom migration may occur
at the β- or the γ-carbon atom.
Results and Discussion
In order to gain insight into the cationic cyclization of
In our previous papers [56,57], we described the intramolecular β-hydroxyalkylphosphine sulfides, a set of substrates 6–25
cationic cyclization of β-hydroxyalkylphosphine oxides was prepared from dimethylphenylphosphine sulfide 4
(Scheme 2).
(Table 1).
Scheme 2: Cyclization of 1a and 1b under acidic conditions.
89

Beilstein J. Org. Chem. 2020, 16, 88–105.
Table 1: Synthesis of substrates for this study.
entry
R
R’
compound
yield, %
dr
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Me
Et
iPr
H
H
H
H
H
6
7
8
9
77
90
75
84
71
89
91
92
76
84
85
84
86
88
78
70
72
88
88
78
51:49
57:43
55:45
61.5:38.5
60:40
57:43
52:48
62.5:37.5
58:42
51:49
53:47
t-Bu
Ph
cyclohexane
Et
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
H
Me
Me
Me
Me
Me
Me
Me
Me
Et
Ph
n-Bu
n-Pr
iPr
t-Bu
t-BuCH2
Me
55:45
53:47
Et
iPr
iPr
–(CH2)4–
–(CH2)5–
–(CH2)6–
n-Bu
n-Bu
All compounds were obtained in good yields, both from alde- Unexpectedly, when the reaction was performed under
hydes and ketones. For aldehydes and unsymmetrically substi- conditions developed originally for β-hydroxyalkylphosphine
tuted ketones, the formation of the products as diastereomeric oxides [56,57], this led to the formation of cyclic phosphine
mixtures could be observed. Unfortunately, the diastereomeric oxides 3 and 26 rather than phosphine sulfides. It may be
excess of the reaction was low in each case.
assumed that hydrolysis of the P=S bond occurred under
these reaction conditions and, probably, racemization
In order to analyze the reactivity of β-hydroxyalkylphosphine of the phosphorus center in the case of nonracemic
sulfides under acidic conditions, the cyclization of 8 and 19 was β-hydroxyalkylphosphine sulfides. Indeed, the attempted
attempted (Scheme 4).
cyclization of (SP)-19 under standard reaction conditions led
Scheme 4: Cyclization of 8 and 19 in the presence of H3PO4.
90

Beilstein J. Org. Chem. 2020, 16, 88–105.
to the formation of the completely racemic phosphine oxide 3 the ability of Lewis acids to catalyze the cyclization of
(Scheme 5).
β-hydroxyalkylphosphine sulfides (Table 2).
The above results clearly show that phosphoric acid was inap- As it turned out, the sulfides derived from aldehydes 7–9 and 11
propriate for the cyclization of nonracemic β-hydroxyalkyl- failed to undergo a rearrangement (Table 2, entries 1–4). On the
phosphine sulfides, and therefore, an alternative methodology other hand, compounds derived from ketones 12, 14–17, and
had to be developed for the synthesis of chiral organophos- 20–25 underwent a ‘clean’ reaction under the conditions. How-
phorus compounds possessing either phosphaindane or ever, instead of cyclization, a phosphorus-to-carbon sulfur
benzophosphorinane skeletons. As such, it was decided to test migration took place, affording the corresponding mercap-
Scheme 5: Cyclization of (SP)-19 in the presence of H3PO4.
Table 2: Reaction of β-hydroxyalkylphosphine sulfides with Lewis-acidic AlCl3.
entry
1
substrate
product (yield)
no reaction
7
8
2
3
4
no reaction
no reaction
no reaction
9
11
91

Beilstein J. Org. Chem. 2020, 16, 88–105.
Table 2: Reaction of β-hydroxyalkylphosphine sulfides with Lewis-acidic AlCl3. (continued)
5
12
29 (74%)
6
mixture
13
7
14
30 (94%)
31 (81%)
32 (74%)
33 (74%)
34 (96%)
8
15
9
16
10
17
11
18
12
19
13
20
37 (59%)
38 (59%)
14
21
15
22
39 (40%)
92

Beilstein J. Org. Chem. 2020, 16, 88–105.
Table 2: Reaction of β-hydroxyalkylphosphine sulfides with Lewis-acidic AlCl3. (continued)
16
23
40 (60%)
17
24
41 (88%)
42 (60%)
18
25
aConversion based on 31P NMR spectroscopic analysis of the crude mixture.
toalkylphosphine oxides 29–33 and 37–42 with satisfying yields in accordance with the shifts of this group in other phosphine
(Table 2, entries 5, 7–10, and 13–18). This sulfur atom phos- oxides [61-63]. Moreover, the two signals at 2.91–2.98 and
phorus-to-carbon migration in phosphine sulfides has previ- 3.16–3.24 ppm, respectively, corresponded to a hydrogen atom
ously been mentioned only once in the literature [59]. Another bonded to a sulfur atom [64,65].
example of such a migration has been reported by Creary and
Innocencio for functionalized α-hydroxythiophosphonic acid Even more conclusive was the analysis of the 13C NMR spec-
diesters [60].
tra of 12 and 29 (Figure 2).
What is more interesting, the sulfur atom did not simply replace In the 13C NMR spectra of compound 12, the signal of the ipso-
oxygen at the β-carbon atom: in every case, the mercapto group carbon atom could be found at 133.1 and 133.2 ppm, respec-
was placed at the γ-carbon atom, suggesting that the reaction tively (a mixture of two diastereomers), with 1J coupling con-
followed a more complex mechanism. All reactions tested stants of 76.9 and 78.1 Hz, respectively. In 29, the signal of the
proceeded with the formation of an additional chirality center at same carbon atom could be found at 133.6 and 134.2 ppm, re-
the γ-carbon atom in the alkyl substituent, and therefore, almost spectively, with 1J coupling constants of 95.8 and 93.9 Hz, re-
all products were obtained as mixtures of epimers. Interestingly, spectively, values typical for arylphosphine derivatives
compound 18 underwent isobutene elimination rather than rear- possessing a P=O bond [61,63,66]. Similarly, shifts and cou-
rangement under the reaction conditions, whereas compound 19 pling constants of CH3 groups bound to phosphorus (17.2 ppm/
led to the formation of a mixture of three products, with 70.3 Hz and 17.7 ppm/70.3 Hz) and α-CH2 groups (35.4 ppm/
alkenylphosphine sulfide 35 being the major compound.
69.7 Hz and 36.0 ppm/69.7 Hz) indicated the presence of a P=O
bond within the molecule. The biggest difference, however, was
The structures were assigned based on NMR analysis as all seen for the C–X bond. For 12, the signals of the tertiary
compounds failed to give single crystals for X-ray analysis. The carbinolic carbon atom were found at 73.1 ppm/4.3 Hz and
presence of an oxygen atom attached to the phosphorus atom in 73.3 ppm/4.3 Hz, respectively. For 29, the same signals were
the products could be deduced based on 1H and 13C NMR anal- found at 41.2 ppm/11.0 Hz and 41.6 ppm/10.7 Hz, respectively.
ysis. In the 1H NMR spectra of phosphine sulfide 12, the chemi- The shifts indicated the presence of a C–S and not a C–O bond
cal shifts of the signals belonging to the aromatic protons in the carbon skeleton [64,65], and the coupling constant value
differed significantly more compared to the appropriate phos- was typical for an γ-carbon atom present in a phosphine deriva-
phine oxide. In 29 (a mixture of two diastereomers), the same tive [67-69].
chemical shifts were remarkably closer to each other (Figure 1).
In this regard, 31P NMR analysis provides very little informa-
In addition, signals of methyl groups attached to phosphorus tion about the reaction mechanism. For substrates 6–25, the
could be found at 1.70 and 1.71 ppm, respectively, which was 31P NMR shifts were in the range of 32.66–39.33 ppm, which
93

Beilstein J. Org. Chem. 2020, 16, 88–105.
Figure 1: 1H NMR spectra of compounds 12 and 29.
was the typical range for aryldialkylphosphine sulfides [70,71]. could be isolated as the main products, and both 13 and 18
The signal of aryldialkylphosphine oxides are usually found afforded product mixtures (Table 3, entries 7 and 11). General-
around 30–50 ppm [61,72,73], and hence the overlapping ly, the rearrangement proceeded efficiently and most of the
signals make this technique inappropriate for the analysis of the products were isolated in good yields except from compounds
reaction products.
possessing cyclopentyl and cycloheptyl fragments, which
afforded the β-mercaptophosphine oxides in low yields.
The outcome of the above-discussed reaction may suggest that
the substrate underwent water elimination followed by intramo- In order to gain insight into the reaction mechanism when using
lecular carbon–sulfur bond formation, and that the formed Brønsted acid, it was decided to perform the reactions at differ-
cationic intermediates underwent hydrolysis during aqueous ent temperatures and to monitor the reaction mixture at differ-
workup, leading to the observed products. It was assumed that a ent times (Table 4).
similar sulfur atom migration should occur in the presence of
Brønsted acid under dry conditions. Therefore, a set of phos- When the reaction was performed at 55 °C, the conversion of
phine sulfides prepared above was subjected to the reaction substrate 20 to the corresponding product 54 reached 93% after
using a H2SO4/AcOH mixture (Table 3).
1 hour (Table 4, entry 1). Apart from this product, the presence
of alkenylphosphine sulfide 59 was detected in the reaction
It was found that the outcome of each reaction depended on the mixture. Prolonged heating led to the consumption of alkene 59
structure of the substrate. β-Hydroxyalkylphosphine sulfides and its transformation to 54. For the reaction performed at
6–9 and 11, derived from aldehydes, underwent estrification 25 °C, the principle observations remained the same except that
reactions, yielding the corresponding acetates 43–45, 47, and 48 the reaction speed was remarkably lower. Nevertheless, these
(Table 3, entries 1–5). On the other hand, most of the substrates experiments allowed to conclude that the rearrangement
derived from ketones 12–16 and 18–25 underwent intramolecu- proceeded via a two-step mechanism, with the first step being
lar sulfur migration with the formation of mainly β-mercap- hydroxy group removal followed by intramolecular formation
toalkylphosphine oxides in very good yields (Table 3, entries 6, of a sulfur–carbon bond. The second step seemed to be slow,
8–10, and 12–18). Only in the case of compounds 16 and 21, and the intermediate carbocation could be trapped in the form
the corresponding γ-mercaptoalkylphosphine oxides 32 and 38 of an alkenylphosphine sulfide.
94

Beilstein J. Org. Chem. 2020, 16, 88–105.
Figure 2: 13C NMR spectra of compounds 12 and 29.
The comparison of the reactivity of the same substrate It appeared that the sulfides derived from aldehydes 8 and 9
under two different reaction conditions clearly showed failed to undergo transformation to the corresponding
that distinct reaction mechanisms were responsible for alkenylphosphine sulfides when treated with a MsCl/NEt3 mix-
the substrate’s transformation under Lewis and Brønsted acid ture. Instead, the corresponding mesylates 60 and 61 were iso-
catalysis, but both seemed to include the preliminary formation lated as the sole products (Scheme 6). The compounds derived
of alkenylphosphine sulfide. This was shown for the Brønsted from ketones 12, 16, and 19 underwent elimination reactions to
acid-mediated reaction (Table 4) but remains to be proved the β,γ-alkenylphosphine sulfides 62, 64, and 65 under these
for Lewis acid use. It was then decided to prepare conditions. For sulfide 12, a mixture of two β-alkenylphos-
substrates possessing double bonds at the β position phine sulfides 62 and 63 (a mixture of E- and Z-isomer) was ob-
(Scheme 6).
tained under each reaction condition tested.
95

Beilstein J. Org. Chem. 2020, 16, 88–105.
Table 3: Reaction of β-hydroxyalkylphosphine sulfides with Brønsted-acidic H2SO4.
entry
1
substrate
product (yield)
6
7
8
9
43 (83%)
44 (95%)
2
3
4
47 (34%)a (25%)
5
11
6
7
8
12
13
14
50 (87%)
mixture of products
51 (88%)
96

Beilstein J. Org. Chem. 2020, 16, 88–105.
Table 3: Reaction of β-hydroxyalkylphosphine sulfides with Brønsted-acidic H2SO4. (continued)
9
52 (95%)
15
10
16
32 (79%)
11
mixture of products
18
12
19
53 (69%)
54 (89%)
13
20
14
21
38 (91%)
55 (26%)
15
22
16
23
56 (96%)
57 (27%)
58 (37%)
17
24
18
25
aYield of the product as isolated as a mixture with substrate. bt = 96 h.
97

Beilstein J. Org. Chem. 2020, 16, 88–105.
toalkylphosphine oxides in good yields. Surprisingly, the reac-
tivity of the mesylates 60 and 61 appeared to be completely dif-
ferent from the parent β-hydroxyalkylphosphine sulfides 8 and
9, which failed to undergo a rearrangement (see Table 2, entries
2 and 3). The difference must therefore have arisen from the
different C–O bond dissociation energies of the alcohols and
mesylates.
Table 4: Control experiments under Brønsted acid catalysis.
Assuming that the Lewis acid coordinated to either hydroxy
oxygen atom or to one of the oxygen atoms of the mesyl group,
the activation energies for C–O bond dissociation are ca.
34.6 kcal/mol for the secondary alcohol 8, ca. 23.3 kcal/mol for
the mesylate 60, and ca. 25.0 kcal/mol for the tertiary alcohol
20, as estimated by DFT calculations. The value for mesylate
roughly fitted the range according to which the reaction may
proceed at room temperature. For mesylate 61, with a tert-butyl
group at the β-carbon atom, the initial dissociation of the C–O
bond must have been followed by CH3 group migration before
the observed sulfur transfer took place.
entry
T, °C
T, h
yield of 54a, % yield of 59a, %
1
2
3
4
5
6
7
55
55
55
25
25
25
25
1
2
24
1
2
4
93
94
7
6
100
17
30
56
100
0
83
70
44
0
24
aConversion based on NMR analysis of the crude reaction mixture.
Analogous attempts were made for compounds 62–65
Once all the compounds had been prepared, it was decided to (Scheme 8).
test the reactivity of mesylates 60 and 61 in the presence of
AlCl3 (Scheme 7).
A mixture of alkenylphosphine sulfides 62 and 63 afforded the
same γ-mercaptoalkylphosphine oxide 29 in good yield in the
The reaction of these compounds with aluminum chloride presence of AlCl3. Similarly, sulfide 64 afforded the phosphine
proceeded smoothly, leading to the corresponding γ-mercap- oxide 32, possessing a mercapto group at the γ-carbon atom,
Scheme 6: Synthesis of the alkenylphosphine sulfides used in study.
Scheme 7: The reaction of mesylate compounds with Lewis-acidic AlCl3.
98

Beilstein J. Org. Chem. 2020, 16, 88–105.
phosphine sulfide 20, possessing two ethyl groups at the β-car-
bon atom. First, the Brønsted acid-catalyzed rearrangement was
investigated by molecular modeling (Scheme 9).
The first step of the rearrangement involved the protonation of
20, which afforded intermediate I, where the proton inserted be-
tween the sulfur and the oxygen atom. This intermediate
appeared to be stable and failed to undergo further transformat-
ion, according to calculations. The reason of this behavior was
most probably the cyclic structure of the intermediate. Howev-
er, if the intermediate I can undergo a second protonation
process, intermediate III forms through transition state II.
Water elimination from III proceeded with slight increase in
energy, affording intermediate IV, which may undergo two dif-
ferent transformations: The first transformation involves disso-
ciation of a water molecule, which leads to the formation of the
dication IX through transition state VIII. Taking into account
that all energy barriers were well below 10 kcal/mol, it seemed
that this transformation was reversible, and therefore the dica-
tion IV may have entered an alternative pathway, which
involved the rotation of the molecule around the Cα–Cβ bond
(activation barrier: +5.4 kcal/mol, stabilization: −3.4 kcal/mol),
which was feasible under the reaction conditions. The formed
Scheme 8: The reaction of alkenylphosphine sulfides with AlCl3.
under the same reaction conditions. Finally, sulfide 65 under- dication V underwent an intramolecular nucleophilic substitu-
went a transformation into oxide 53, possessing an SH group at tion through transition state VI, which proceeded with a small
the β-carbon atom, under the reaction conditions. The latter activation barrier, leading to the cyclic intermediate VII and
result was slightly different from the two previous ones, as here, further to the rearranged product 54.
the sulfur atom migrated to the β and not the γ-carbon atom.
Moreover, the parent alcohol 19 underwent mainly a dehydra- The tentative mechanism discussed above does not explain the
tion reaction and the formation of α,β-alkenylphosphine sulfide formation of the γ-mercaptoalkylphosphine sulfides from
35 in the presence of AlCl3.
β-hydroxyalkylphosphine sulfides 16 and 21 and, to some
extent, 8 and 11. In these cases, DFT calculations allowed to
All of these experiments suggested that the transformation of propose a slightly different mechanism (see Supporting Infor-
β-hydroxyalkylphosphine sulfides in the presence of Lewis acid mation File 1). In this case, a water molecule, formed after pro-
most probably proceeded through the initial formation of an tonation of the OH group, facilitated proton transfer from the γ-
intermediate alkenylphosphine sulfide, which then underwent to β-carbon atom via a process similar to the elimination/elec-
an intramolecular reaction leading to a formal [1,4]-rearrange- trophilic addition to alkenes pathway.
ment product. It seemed that the position of the double bond
was crucial for the rearrangement, which had to be situated be- The computational results discussed above showed that the
tween the β- and the γ-carbon atom. In this position, there was crucial point for sulfur atom migration was the formation of a
no coupling with the thiophosphoryl group, and the double tertiary carbocation at the β-carbon atom, which was quite
bond retained its nucleophilic character. This allowed the coor- straightforward under these acidic conditions. To understand
dination of an AlCl3 molecule to the C=C bond, which then be- the completely different selectivity in Lewis acid-catalyzed re-
came susceptible to an intramolecular nucleophilic attack by a arrangements, DFT was again applied to the reaction between
sulfur atom. For α,β-alkenylphosphine sulfide 35, the conjuga- 20 and AlCl3 (Scheme 10).
tion of the double bond with thiophosphoryl fragment made the
activation by AlCl3 impossible, which prevented the rearrange- Compound 20 underwent a complexation with an AlCl3 mole-
ment.
cule, leading to intermediate I, and the overall transformation
proceeded with remarkable stabilization. In the next step, disso-
To gain detailed insights into the possible reaction mechanism, ciation of the C−O bond occurred through transition state II,
DFT calculations were performed on model β-hydroxyalkyl- finally leading to alkene III. The activation energy for the C−O
99

Beilstein J. Org. Chem. 2020, 16, 88–105.
Scheme 9: Rearrangement of 20 in the presence of Brønsted acid. The calculated energies next to the arrows are reported in kcal/mol.
Scheme 10: Rearrangement of 20 in the presence of Lewis acid. The calculated energies next to the arrows are reported in kcal/mol.
100

Beilstein J. Org. Chem. 2020, 16, 88–105.
bond cleavage in 20, a tertiary alcohol, was similar to the value remarkable activation energy (31.0 kcal/mol), which was fol-
calculated for the mesylate 60 (23.3 kcal/mol) and was lower lowed by slight stabilization of the formed γ-carbanion (−6.1
than that for β-hydroxyalkylphosphine sulfide 8, formally a sec- kcal/mol). The latter should then undergo intramolecular C–S
ondary alcohol (34.6 kcal/mol).
bond formation, followed by hydrolysis, finally leading to the
rearranged product.
One could consider an alternative mechanism for the formation
of III including the six-membered transition state VII. In this It was then decided to investigate the stereoselectivity of the
case, proton migration would occur from the γ-carbon atom to acid-catalyzed rearrangement using enantiomerically enriched
chloride, resulting in the formation of HCl, Al(OH)Cl2, and phosphine sulfides. Therefore, the chiral substrates (SP)-60,
intermediate III. However, all attempts to monitor this particu- (SP)-65, (SP)-19, and (SP)-20 were prepared from
lar transition state failed.
dimethylphenylphosphine sulfide 4 using sparteine chemistry
(Scheme 11).
The formed intermediate III underwent a reaction with a second
AlCl3 molecule, which associated with the double bond,
yielding IV in a slightly exothermic process. The formed inter-
mediate possessing an activated C=C bond underwent intramo-
lecular carbon–sulfur bond formation through the transition
state V. In this case, the reaction proceeded solely at the γ-car-
bon atom, and all attempts directed at the discovery of the tran-
sition state where the sulfur attack occurred at the β-carbon
atom failed. It was possible that the electron density of the C=C
double bond was slightly shifted towards the β-carbon atom due
to possible interaction with the P=S fragment, and coordination
of an AlCl3 molecule additionally activated the γ-carbon atom
for a sulfur attack.
Regarding the correctness of the proposed mechanism, two
issues associated with the reactivity of 17 and 19, respectively,
in the presence of AlCl3 must be discussed. The first is the for-
mation of the adduct with tert-butyl methyl ketone, and the
second is the formation of the adduct with acetone. For 17, the
formation of β,γ-alkenylphosphine sulfide was not possible,
Scheme 11: The synthesis of chiral substrates for rearrangement
reactions.
whereas the transformation of 19 afforded mainly the α,β-unsat- First, the enantioenriched alcohol (SP)-17, mesylate (SP)-60,
urated compound 35. DFT analysis of the C–O bond cleavage and alkenylphosphine sulfide (SP)-65 were subjected to the
in the adducts analogous to I revealed two different pathways. reaction with Lewis-acidic AlCl3 (Scheme 12).
A complex of 19 with AlCl3 underwent simultaneous C–O bond
cleavage and α-hydrogen (but not γ-hydrogen) atom abstraction, Rearrangement of the tertiary alcohol (SP)-17 proceeded very
which led to 35, with activation barrier of +24.8 kcal/mol and a efficiently, affording the corresponding product in high yield.
stabilization energy of −16.1 kcal/mol. Yet, at the moment, this Unfortunately, the stereoselectivity of the reaction appeared to
preference remains to be further investigated.
be quite low due to the significant racemization of the phos-
phorus center under the applied conditions. On the other hand,
On the other hand, the complex of 17 and AlCl3 underwent the treatment of (SP)-60 with Lewis acid led to the formation of
C–O bond cleavage and γ-hydrogen atom abstraction, with an chiral γ-mercaptoalkylphosphine oxide 46 in good yield and
activation barrier of +16.5 kcal/mol, which led to the with only slight decrease in enantiomeric excess. The transfor-
corresponding alkene, with a stabilization energy of mation of alkenylphosphine sulfide (SP)-65 under similar reac-
−13.5 kcal/mol. The loss of AlCl3/H2O afforded the β,γ- tion conditions led to the formation of β-mercaptoalkylphos-
alkenylphosphine sulfide, which then underwent coordination phine oxide 53, albeit in low yield and with a complete lack of
of the second AlCl3 molecule, with an overall slight stabiliza- stereoselectivity.
tion (+6.0/−8.5 kcal/mol). The adduct analogous to IV under-
went methyl group migration to the β-carbon atom, along with The results presented above show that, at least for mesylates,
the formation of a covalent Al–C bond. The last step required a the sulfur atom migration may have been a stereospecific
101

Beilstein J. Org. Chem. 2020, 16, 88–105.
In the next step, the enantiomerically enriched β-hydroxyalkyl-
phosphine sulfides (SP)-19 and (SP)-20 were subjected to the re-
arrangement in the presence of Brønsted-acidic H2SO4
(Scheme 13).
It was found that phosphine sulfide (SP)-17 underwent efficient
rearrangement, affording a mixture of three compounds:
β-mercaptoalkylphosphine oxide 66, its S-acylated derivative
67, and S-acylated γ-mercaptoalkylphosphine oxide 68. Com-
pound 66 was obtained with complete stereoselectivity, where-
as the enantiomeric excess of both 67 and 68 was higher, 78%
ee and 76% ee, respectively. The phosphine sulfides (SP)-19
and (SP)-20 underwent rearrangement to the corresponding
chiral β-mercaptoalkylphosphine oxides 53 and 54 in good
yields and with complete stereoselectivity as compared to
the rearrangement of (SP)-17. Based on this, it can be con-
cluded that sulfur atom migration occurred via a highly stereo-
specific pathway through an intramolecular substitution mecha-
nism.
Scheme 12: The reaction of (SP)-60 and (SP)-65 with AlCl3.
process. What was more striking, the complete racemization of Here, the stereochemical differences in the outcome of the rear-
(SP)-65 and the extended racemization of (SP)-17 occurred rangement of (SP)-17 must be noticed. In the presence of Lewis-
under the reaction conditions. This suggested that the rearrange- acidic AlCl3, the rearrangement led effectively to the corre-
ment may have proceeded either via two different pathways, or sponding γ-mercaptoalkylphosphine oxide 33, albeit with a
one of the reaction components may have influence the rate of remarkable deterioration of enantiomeric excess. The rearrange-
racemization. In the case of (SP)-17, the postulated dissociation ment in the presence of Brønsted-acidic H2SO4 afforded only
of AlCl3⋅OH may have generated traces of HCl, which may minor amounts of S-acylated γ-mercaptoalkylphosphine oxide
have react with the cyclic intermediate, leading to a pentavalent 68, but the enantiomeric excess was even higher than in the
compound. The latter would have been very susceptible to substrate. This suggested that the [1,3]-rearrangement was
Berry or turnstile pseudorotation. This, in turn, should have led stereospecific in the presence of Brønsted acid, whereas the
to a remarkable deterioration of the enantiomeric excess of the [1,4]-rearrangement proceeded with additional partial enrich-
product. The same might be postulated for the rearrangement of ment of one enantiomer. The latter may have occurred through
alkenylphosphine sulfide 65, which should form a pentavalent the preferential acylation of one of two diastereoisomers under
intermediate equally well.
the reaction conditions where the second chirality center
Scheme 13: Reaction of chiral β-hydroxyalkylphosphine sulfides with Brønsted acid.
102

Beilstein J. Org. Chem. 2020, 16, 88–105.
at the phosphorus atom influenced the ratio of the acylation
reaction.
Supporting Information
Supporting Information File 1
With chiral β-mercaptoalkylphosphine sulfides in hand, it was
decided to check their ability to undergo cyclization in the pres-
ence of AlCl3 (Scheme 14).
Experimental procedures, compound characterization data,
a mechanism for the rearrangement of 16 in the presence of
Brønsted acid assessed by computational methods, and
coordinates of computed compounds.
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-16-11-S1.pdf]
Supporting Information File 2
Copies of NMR spectra of all pure compounds and their
mixtures.
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-16-11-S2.pdf]
Funding
Financial support from the National Science Centre (research
grant no. 2012/07/E/ST5/00544) is kindly acknowledged.
Scheme 14: Attempted cyclization of enantiomerically enriched 53 and
46.
ORCID® iDs
Piotr Borowski - https://orcid.org/0000-0003-4462-6913
These results differed remarkably from the reactivity of the
analogous hydroxyalkylphosphine sulfides, which exclusively
underwent sulfur atom migration. The nature of this difference
is currently a subject of further research in our laboratory.
Nevertheless, the above results opened a new pathway for the
synthesis of bicyclic organophosphorus compounds possessing
a chirality center at phosphorus. Moreover, the formation of the
bicyclic compounds proceeded under mild reaction conditions
and with a high degree of stereoselectivity.
Marek Stankevič - https://orcid.org/0000-0003-2917-6774
References
1. Lindsay, A. C.; Sperry, J. Tetrahedron 2017, 73, 4355–4362.
doi:10.1016/j.tet.2017.05.093
2. Maronna, M. M.; Kruissink, E. C.; Parton, R. F.; Tinge, J. T.;
Hoelderich, W. F. Catal. Commun. 2017, 90, 23–26.
doi:10.1016/j.catcom.2016.11.014
3. Naik, P. N.; Alkobati, N. A. H.; Kusurkar, R. S. ARKIVOC 2015, No. vii,
362–376. doi:10.3998/ark.5550190.p009.291
4. Kotha, S.; Ravikumar, O.; Majhi, J. Beilstein J. Org. Chem. 2015, 11,
1503–1508. doi:10.3762/bjoc.11.163
5. Zhang, B.; Wang, X.; Cheng, C.; Sun, D.; Li, C. Angew. Chem., Int. Ed.
2017, 56, 7484–7487. doi:10.1002/anie.201704086
6. Shymanska, N. V.; Pierce, J. G. Org. Lett. 2017, 19, 2961–2964.
doi:10.1021/acs.orglett.7b01185
7. Chan, C.-K.; Chen, Y.-H.; Tsai, Y.-L.; Chang, M.-Y. J. Org. Chem.
2017, 82, 3317–3326. doi:10.1021/acs.joc.7b00108
8. Ricard, S.; Sanapo, G. F.; Rahem, N.; Daoust, B. J. Org. Chem. 2016,
81, 5066–5073. doi:10.1021/acs.joc.6b00512
9. Gómez-Martínez, M.; Baeza, A.; Alonso, D. A. ChemCatChem 2017, 9,
1032–1039. doi:10.1002/cctc.201601362
10.Wu, H.; Wang, Q.; Zhu, J. Angew. Chem., Int. Ed. 2016, 55,
15411–15414. doi:10.1002/anie.201609911
11.Rao, C. N.; Khan, F. A. Org. Biomol. Chem. 2015, 13, 2768–2775.
doi:10.1039/c4ob02423k
12.Rudolf, O.; Rouchal, M.; Lyčka, A.; Klásek, A. Helv. Chim. Acta 2013,
96, 1905–1917. doi:10.1002/hlca.201300074
13.Korda, A.; Pakulski, Z.; Cmoch, P.; Gwardiak, K.; Karczewski, R.
Tetrahedron 2017, 73, 1740–1744. doi:10.1016/j.tet.2017.02.024
14.Nishiyama, Y.; Yokoshima, S.; Fukuyama, T. Org. Lett. 2016, 18,
2359–2362. doi:10.1021/acs.orglett.6b00789
Conclusion
In conclusion, in the present paper, we describe the rearrange-
ment of a series of β-hydroxyalkylphosphine sulfides. Both ex-
perimental and theoretical analysis of the reaction mechanism
were performed. Depending on the structure of the initial
alcohol, the acid used (Brønsted or Lewis), and the reaction
conditions (temperature, time), the acid-promoted transformat-
ions of the alcohols may lead to two different products: β- or
γ-mercaptoalkylphosphine oxides. The formation of γ-mercap-
toalkylphosphine oxides proceeded with the formation of an ad-
ditional chirality center at the γ-carbon atom. Attempted rear-
rangements of P-stereogenic β-hydroxy- and β-mesyloxyalkyl-
phosphine sulfides in the presence of either Brønsted or Lewis
acid led to the corresponding rearranged products with variable
degree of stereoselectivity. It was found that the P-stereogenic
β-/γ-mercaptoalkylphosphine oxides underwent intramolecular
cyclization to the corresponding bicyclic phosphine oxides with
high stereoselectivity and high yield under mild reaction condi-
tions.
103

Beilstein J. Org. Chem. 2020, 16, 88–105.
15.Shen, Z.; Pan, X.; Lai, Y.; Hu, J.; Wan, X.; Li, X.; Zhang, H.; Xie, W.
Chem. Sci. 2015, 6, 6986–6990. doi:10.1039/c5sc02485d
16.Fu, J.-G.; Ding, R.; Sun, B.-F.; Lin, G.-Q. Tetrahedron 2014, 70,
8374–8379. doi:10.1016/j.tet.2014.08.072
17.Zhou, H.; Hong, J.; Huang, J.; Chen, Z. Asian J. Org. Chem. 2017, 6,
817–820. doi:10.1002/ajoc.201700248
18.Milbeo, P.; Moulat, L.; Didierjean, C.; Aubert, E.; Martinez, J.;
Calmès, M. J. Org. Chem. 2017, 82, 3144–3151.
doi:10.1021/acs.joc.7b00122
19.He, X.; Cao, C.; Liang, J.; Li, X.; Zhang, T.; Meng, F. Synlett 2017, 28,
386–390. doi:10.1055/s-0036-1588905
20.Zhang, D.; Zheng, H.; Wang, X. Tetrahedron 2016, 72, 1941–1953.
doi:10.1016/j.tet.2016.02.060
21.Miyamoto, K.; Yamashita, J.; Narita, S.; Sakai, Y.; Hirano, K.; Saito, T.;
Wang, C.; Ochiai, M.; Uchiyama, M. Chem. Commun. 2017, 53,
9781–9784. doi:10.1039/c7cc05160c
40.Pallikonda, G.; Santosh, R.; Ghosal, S.; Chakravarty, M.
Tetrahedron Lett. 2015, 56, 3796–3798.
doi:10.1016/j.tetlet.2015.04.073
41.Hayashi, M.; Nakamura, S. Angew. Chem., Int. Ed. 2011, 50,
2249–2252. doi:10.1002/anie.201007568
42.Hammerschmidt, F.; Schmidt, S. Eur. J. Org. Chem. 2000, 2239–2245.
doi:10.1002/1099-0690(200006)2000:12<2239::aid-ejoc2239>3.0.co;2-
v
43.Hasanov, H. H.; Ivanov, I. K.; Christov, V. C. Heteroat. Chem. 2017,
28, e21357. doi:10.1002/hc.21357
44.Gomes, F.; Fadel, A.; Rabasso, N. J. Org. Chem. 2012, 77,
5439–5444. doi:10.1021/jo300790m
45.Himbert, G.; Ruppmich, M.; Knöringer, H. J. Chin. Chem. Soc. 2003,
50, 143–151. doi:10.1002/jccs.200300020
46.Demay, S.; Lotz, M.; Polborn, K.; Knochel, P. Tetrahedron: Asymmetry
2001, 12, 909–914. doi:10.1016/s0957-4166(01)00123-9
47.Kost, D.; Sprecher, M. S. Tetrahedron Lett. 1975, 16, 4483–4486.
doi:10.1016/s0040-4039(00)91099-2
22.Huang, H.; Yang, Q.; Zhang, Q.; Wu, J.; Liu, Y.; Song, C.; Chang, J.
Adv. Synth. Catal. 2016, 358, 1130–1135.
doi:10.1002/adsc.201501071
48.Fischer, G. W.; Schneider, P. J. Prakt. Chem. 1977, 319, 399–407.
doi:10.1002/prac.19773190308
49.Haber, S.; Schmitz, M.; Bergsträßer, U.; Hoffmann, J.; Regitz, M.
Chem. – Eur. J. 1999, 5, 1581–1589.
23.Jevtić, I.; Došen-Mićović, L.; Ivanović, E. R.; Ivanović, M. D. Synthesis
2016, 48, 1550–1560. doi:10.1055/s-0035-1561405
24.Yoshimura, A.; Middleton, K. R.; Luedtke, M. W.; Zhu, C.;
Zhdankin, V. V. J. Org. Chem. 2012, 77, 11399–11404.
doi:10.1021/jo302375m
doi:10.1002/(sici)1521-3765(19990503)5:5<1581::aid-chem1581>3.0.c
o;2-2
25.Hennum, M.; Odden, H. H.; Gundersen, L.-L. Eur. J. Org. Chem. 2017,
846–860. doi:10.1002/ejoc.201601446
50.Qian, R.; Roller, A.; Hammerschmidt, F. J. Org. Chem. 2015, 80,
1082–1091. doi:10.1021/jo502567j
26.Sharif, S. A. I.; Calder, E. D. D.; Harkiss, A. H.; Maduro, M.;
Sutherland, A. J. Org. Chem. 2016, 81, 9810–9819.
doi:10.1021/acs.joc.6b01881
27.Kotha, S.; Ravikumar, O. Tetrahedron Lett. 2016, 57, 1994–1996.
doi:10.1016/j.tetlet.2016.03.087
28.Rajasekaran, P.; Ansari, A. A.; Vankar, Y. D. Eur. J. Org. Chem. 2015,
2902–2913. doi:10.1002/ejoc.201500129
29.Jasiak, A.; Mielniczak, G.; Owsianik, K.; Koprowski, M.; Krasowska, D.;
Drabowicz, J. J. Org. Chem. 2019, 84, 2619–2625.
doi:10.1021/acs.joc.8b03053
30.Piekutowska, M.; Pakulski, Z. Tetrahedron Lett. 2007, 48, 8482–8486.
doi:10.1016/j.tetlet.2007.09.162
31.Matveeva, E. V.; Odinets, I. L.; Kozlov, V. A.; Shaplov, A. S.;
Mastryukova, T. A. Tetrahedron Lett. 2006, 47, 7645–7648.
doi:10.1016/j.tetlet.2006.08.050
51.Bowman, E.; McQueney, M.; Barry, R. J.; Dunaway-Mariano, D.
J. Am. Chem. Soc. 1988, 110, 5575–5576. doi:10.1021/ja00224a054
52.Solomon, S. A.; Allen, L. K.; Dane, S. B. J.; Wright, D. S.
Eur. J. Inorg. Chem. 2014, 1615–1619. doi:10.1002/ejic.201301203
53.An, Y.-Z.; An, J. G.; Wiemer, D. F. J. Org. Chem. 1994, 59,
8197–8202. doi:10.1021/jo00105a042
54.Gololobov, Y. G.; Krasnova, I. Y.; Barabanov, S. V.; Khrustalev, V. N.;
Andronati, S. A.; Pavlovsky, V. I. Tetrahedron Lett. 2014, 55,
4879–4882. doi:10.1016/j.tetlet.2014.07.010
55.Sokolsky, A.; Smith, A. B., III. Org. Lett. 2012, 14, 4470–4473.
doi:10.1021/ol3019709
56.Włodarczyk, K.; Stankevič, M. Tetrahedron 2016, 72, 5074–5090.
doi:10.1016/j.tet.2016.06.070
57.Włodarczyk, K.; Borowski, P.; Drach, M.; Stankevič, M. Tetrahedron
2017, 73, 239–251. doi:10.1016/j.tet.2016.12.008
58.Muci, A. R.; Campos, K. R.; Evans, D. A. J. Am. Chem. Soc. 1995,
117, 9075–9076. doi:10.1021/ja00140a028
32.Bhattacharya, A. K.; Thyagarajan, G. Chem. Rev. 1981, 81, 415–430.
doi:10.1021/cr00044a004
33.Korb, M.; Lehrich, S. W.; Lang, H. J. Org. Chem. 2017, 82, 3102–3124.
doi:10.1021/acs.joc.7b00030
59.Chen, C. H.; Brighty, K. E. Tetrahedron Lett. 1980, 21, 4421–4424.
doi:10.1016/s0040-4039(00)92189-0
34.Berchel, M.; Salaün, J.-Y.; Couthon-Gourvès, H.; Haelters, J.-P.;
Jaffrès, P.-A. Dalton Trans. 2010, 39, 11314–11316.
doi:10.1039/c0dt00880j
35.Jayasundera, K. P.; Watson, A. J.; Taylor, C. M. Tetrahedron Lett.
2005, 46, 4311–4313. doi:10.1016/j.tetlet.2005.04.104
36.Taylor, C. M.; Watson, A. J. Curr. Org. Chem. 2004, 8, 623–636.
doi:10.2174/1385272043370717
60.Creary, X.; Inocencio, P. A. J. Am. Chem. Soc. 1986, 108, 5979–5983.
doi:10.1021/ja00279a052
61.Stankevič, M.; Pisklak, J.; Włodarczyk, K. Tetrahedron 2016, 72,
810–824. doi:10.1016/j.tet.2015.12.043
62.Hockless, D. C. R.; McDonald, M. A.; Pabel, M.; Wild, S. B.
J. Organomet. Chem. 1997, 529, 189–196.
doi:10.1016/s0022-328x(96)06641-7
37.Legrand, O.; Brunel, J. M.; Constantieux, T.; Buono, G. Chem. – Eur. J.
1998, 4, 1061–1067.
63.Rajendran, K. V.; Gilheany, D. G. Chem. Commun. 2012, 48, 817–819.
doi:10.1039/c1cc14856g
doi:10.1002/(sici)1521-3765(19980615)4:6<1061::aid-chem1061>3.0.c
o;2-v
64.Simian, H.; Robert, F.; Blank, I. J. Agric. Food Chem. 2004, 52,
306–310. doi:10.1021/jf035008h
38.Zhang, F.; Jiang, M.; Liu, J.-T. Tetrahedron Lett. 2017, 58, 1871–1874.
doi:10.1016/j.tetlet.2017.04.002
65.Jin, L.; Wang, J.; Dong, G. Angew. Chem., Int. Ed. 2018, 57,
12352–12355. doi:10.1002/anie.201807760
39.Kondoh, A.; Aoki, T.; Terada, M. Chem. – Eur. J. 2017, 23, 2769–2773.
doi:10.1002/chem.201605673
104

Beilstein J. Org. Chem. 2020, 16, 88–105.
66.Copey, L.; Jean-Gérard, L.; Framery, E.; Pilet, G.; Robert, V.;
Andrioletti, B. Chem. – Eur. J. 2015, 21, 9057–9061.
doi:10.1002/chem.201501324
67.Dong, J.; Liu, L.; Ji, X.; Shang, Q.; Liu, L.; Su, L.; Chen, B.; Kan, R.;
Zhou, Y.; Yin, S.-F.; Han, L.-B. Org. Lett. 2019, 21, 3198–3203.
doi:10.1021/acs.orglett.9b00922
68.Hu, G.; Chen, W.; Fu, T.; Peng, Z.; Qiao, H.; Gao, Y.; Zhao, Y.
Org. Lett. 2013, 15, 5362–5365. doi:10.1021/ol402672e
69.Binyamin, I.; Meidan-Shani, S.; Ashkenazi, N. Beilstein J. Org. Chem.
2015, 11, 1332–1339. doi:10.3762/bjoc.11.143
70.Woźnicki, P.; Korzeniowska, E.; Stankevič, M. J. Org. Chem. 2017, 82,
10271–10296. doi:10.1021/acs.joc.7b01767
71.Bruzik, K. S.; Stec, W. J. J. Org. Chem. 1990, 55, 6131–6135.
doi:10.1021/jo00312a018
72.Chen, Q.; Yan, X.; Du, Z.; Zhang, K.; Wen, C. J. Org. Chem. 2016, 81,
276–281. doi:10.1021/acs.joc.5b02308
73.Pietrusiewicz, K. M.; Zabłocka, M.; Monkiewicz, J. J. Org. Chem. 1984,
49, 1522–1526. doi:10.1021/jo00183a009
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(https://creativecommons.org/licenses/by/4.0). Please note
that the reuse, redistribution and reproduction in particular
requires that the authors and source are credited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(https://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.16.11
105