Mild stereoselective formation of tri- and tetrasubstituted olefins by regioselective ring opening of 1,1-disubstituted vinyl oxiranes with dialkyl dithiophosphates

Article abstract of DOI:10.1016/j.tetlet.2014.04.028

Presented in this Letter is a mild new method to stereoselectively ring open 1,1-disubstituted vinyl oxiranes using dialkyl dithiophosphate nucleophiles. This new reaction proceeds in toluene at room temperature to afford tri- and tetrasubstituted allylic alcohols with high Z-selectivity. The syn relationship between the alcohol and the newly incorporated dithiophosphate moiety presented an opportunity to evaluate the anionic hopping properties of the ring opening product. Treatment of the ring opening product with base results in a facile phosphate hopping followed by a selective thiolate 3-exo-trig instead of a 5-exo-tet cyclization.

Full text of DOI:10.1016/j.tetlet.2014.04.028

Tetrahedron Letters 55 (2014) 3232–3234
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Mild stereoselective formation of tri- and tetrasubstituted olefins
by regioselective ring opening of 1,1-disubstituted vinyl oxiranes
with dialkyl dithiophosphates
⇑
Boying Guo, Edon Vitaku, Jon T. Njardarson
Department of Chemistry and Biochemistry, University of Arizona, 1306 E. University Blvd., Tucson, AZ 85721, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Presented in this Letter is a mild new method to stereoselectively ring open 1,1-disubstituted vinyl oxir-
anes using dialkyl dithiophosphate nucleophiles. This new reaction proceeds in toluene at room temper-
ature to afford tri- and tetrasubstituted allylic alcohols with high Z-selectivity. The syn relationship
between the alcohol and the newly incorporated dithiophosphate moiety presented an opportunity to
evaluate the anionic hopping properties of the ring opening product. Treatment of the ring opening
product with base results in a facile phosphate hopping followed by a selective thiolate 3-exo-trig instead
of a 5-exo-tet cyclization.
Received 27 March 2014
Revised 8 April 2014
Accepted 9 April 2014
Available online 21 April 2014
Keywords:
Vinyl oxirane
Dithiophosphoric acid
Ring opening
Z-Selective
Ó 2014 Elsevier Ltd. All rights reserved.
Regioselective
One of our research programs has been focused on the catalytic
ring expansions of strained rings such as vinyl oxiranes,1 which we
demonstrated could be selectively converted to 2,5-dihydrofuran
products using Cu(hfacac)2.2 The main competing reaction path-
way we needed to suppress during our vinyl oxirane investigations
was a facile 1,2-hydride shift, which is readily triggered using a
variety of Lewis and Brønsted acids. In our continued quest to
develop new useful methods for forming heterocycles, we have
recently begun investigating and exploiting the useful heteroatom
hopping behavior of dialkyl dithiophosphates. Recently, we
reported a scalable one pot synthesis of 3,6-dihydro-2H-thiopyrans
from b-dithiophosphate carbonyl precursors that were accessed
via a Michael addition step.3 Motivated to find a research question
that would allow us to overlap these two programs, we speculated
if vinyl oxiranes could be selectively ring opened with a dithio-
phosphate nucleophile. Such a reaction would present the result-
ing alcohol with the possibility of participating in an anionic
hopping cascade to form a heterocyclic product. Toward that end,
we chose to investigate what product(s) would form when a vinyl
oxirane was treated with a dithiophosphoric acid (Fig. 1).4 Two
Brønsted acid products and four dialkyl dithiophosphate displace-
ment products were envisioned as the most likely outcomes, with
the (Z)-1,4-addition product of most value to our anionic hopping
studies.
S
P
R
O
R
R
S
OR
OR
O
P
S
OR
OR
HS
R
R
O
hydride shift ring expansion
1,4-addition (E)
OH
S
Cu(hfacac)2
S
P
R
OR
OR
S
P
P
OR
OR
OH
OH
S
OR
OR
S
S
OH
R
1,4-addition (Z)
R
1,2-addition
O
Figure 1. Vinyl oxirane ring opening challenge.
There are only few reported studies on the ring opening of vinyl
oxiranes with sulfur nucleophiles, most of which employ alkyl- or
aryl thiols in the presence of a Lewis acid or base (thiolate).5 Major-
ity of these reactions favor direct ring opening C2-of the vinyl
oxirane (1,2-additions). There are even fewer examples of 1,4-
selective ring openings using sulfur nucelophiles, all of which
favored E-olefin products.6 Although dialkyl dithiophosphates have
never been used to ring open vinyl oxiranes specifically, they have
been shown to be competent as nucleophiles to open simple non-
conjugated oxiranes.7 Cyclohexene oxide type substrates reveal
that these addition reactions are stereoselective (anti) while
oxirane nucleophilic attacks at C1 are favored for acyclic ones.
To answer the ring opening challenge presented in Figure 1, we
first synthesized vinyl oxirane 1 from the corresponding vinyl
⇑
Corresponding author. Tel.: +1 520 626 0754; fax: +1 520 621 8407.
E-mail address: njardars@email.arizona.edu (J.T. Njardarson).
http://dx.doi.org/10.1016/j.tetlet.2014.04.028
0040-4039/Ó 2014 Elsevier Ltd. All rights reserved.

B. Guo et al. / Tetrahedron Letters 55 (2014) 3232–3234
3233
ketone using Matteson’s procedure.8 We were delighted to learn
Table 2
Vinyl oxirane substrate scope
that treatment of 1 with diethyl dithiophosphoric acid in toluene
at room temperature resulted in clean conversion to a 15:1 ratio
of allylic alcohols 2 and 3, wherein the Z-isomer (2) was the major
product. Remarkably, this mildly acidic sulfur nucleophile not only
added exclusively at C4 (1,4-addition), while suppressing the com-
peting hydride shift pathway, but it also afforded the desired (Z)-
1,4-addition product. When the similarly acidic diethyl phosphoric
acid (O instead of S) was employed, hydride shift product 4 was
isolated as the major product. Treatment of the dithiophosphoric
acid with excess base, such as triethylamine (shown), or pre-made
metal salts resulted in no ring opening of 1. Clearly both the pres-
ence of sulfur and the acidic hydrogen are essential for the success
of this new selective ring opening reaction.
With these exciting ring opening results in hand, we next
decided to understand how solvents and phosphate substituents
would impact reaction efficiency and selectivity (Table 1). These
studies revealed that toluene is an excellent solvent for this reac-
tion, with benzene and chlorinated solvents also being suitable.
Absence of solvent (entry 8) does not impact efficiency or C4-
addition preference, but completely erodes the stereoselectivity.
Running the reaction at lower temperatures (entry 4) resulted in
improved stereoselectivity although accompanied by a significant
increase in reaction time. Changing the phosphate alkyl group from
ethyl to either methyl (entry 1) or isopropyl (entry 2) did not
change much apart from some loss of efficiency when a diisopropyl
substituent was employed.
With optimal reaction conditions identified, we investigated
seven additional vinyl oxirane substrates (Table 2). Large groups
at C2 result in excellent Z-selectivity (entries 1, 2, and 4). This
selectivity erodes a bit when the vinyl group contains an internal
substituent (entry 3) and all selectivity is lost when the vinyl group
is terminally substituted (entry 5). Interestingly, when the pure
E- and Z-isomer products from entry 5 are left standing in either
chloroform or toluene solutions at room temperature, both equili-
brate over time to an almost identical 2.5:1 Z/E-mixture.9 At this
point, the mechanism of this equilibration process is not known.10
When the vinyl oxiranes are not 1,1-disubstituted (entries 6–7) all
C4-addition selectivity is lost in favor of C2 (1,2-addition). For
butadiene monoxide (entry 8) there is no preference for 1,2- or
1,4-addition, with the 1,4-addition product of predominantly
E-configuration.
Entry Starting material
Product(s)
Yield
(%)
Ph
S
P
Ph
OEt
OEt
1
87
85
85
HO
S
O
Z/ E = 15:1
Ph
Ph
S
O
P
OEt
OEt
2
HO
Ph
S
Z/ E = 17:1
Ph
S
P
OEt
OEt
3
HO
S
O
Z/E = 6:1
only Z
Ph
S
P
Ph
HO
4
5
81
82
O
OEt
OEt
S
S
O
P
OEt
C5H11
HO
S
OEt
C5H11
Z/E = 1.5:1
H
S
P
O
O
O
OEt
OEt
S
S
6
7
71
83
C7H15
HO
HO
H
C7H15
H
H
S
P
OEt
Ph
OEt
Ph
OH
H
S
S
P
OH
H
OEt
S
P
OEt
8
75
OEt
OEt
+
S
1:1
Z/E = 1:13
8-membered ring phosphate transfer followed by a thiolate cycli-
zation to form a 3,6-dihydro-2H-thiopyran could be realized. We
were confident that the seven membered ring S to O mediated
phosphate transfer from 5 to 6 was a reasonable proposition,11
but less confident whether a 5-exo-tet cyclization to form 2,5-
dihydrothiofuran 8 or a 3-exo-trig cyclization pathway to form
vinyl thiirane 7 would follow.12
To our surprise, treatment of allylic alcohol 2 with sodium
hydride in THF at 40 °C smoothly transformed it to a single new
product, identified as vinyl thiirane 7.13 Clearly, the soft sulfur
nucleophile (sodium thiolate 6) prefers to undergo a 3-exo-trig
instead of a 5-exo-tet cyclization.
Having developed a new route to access substrates like 2, with
high Z-selectivity, we were in a position to evaluate their anionic
hopping behavior (Scheme 2). We had reported, for the
corresponding Z-homoallylic alcohol thiophosphate,4 that an
Table 1
Vinyl oxirane ring opening optimizations
What is the origin of the high Z-selectivity for 1,1-disubstituted
vinyl oxiranes (entries 1–4, Table 2)? We believed we can make
strong argument, based on the results presented in Scheme 1, that
Ph
Ph
Ph
S
P
S
OR
OR
S
P
OR
OR
P
S
HS
OR
OR
O
+
S
Table 1
OH
OH
1
2
3
S
P
Ph
Ph
Ph
OEt
S
OEt
OEt
S
P
HS
OEt
P
S
Entry
R
Solvent
T (°C)
t (h)
2:3
Yield (%)
O
OEt
OEt
toluene, rt
+
S
1
2
3
4
5
6
7
8
Methyl
Isopropyl
Ethyl
Ethyl
Ethyl
Ethyl
Ethyl
Ethyl
Toluene
Toluene
Toluene
Toluene
Benzene
CH2Cl2
rt
rt
rt
0
rt
rt
rt
rt
5
8
5
20
5
0.3
0.5
3
13:1
18:1
15:1
29:1
14:1
11:1
11:1
1.1:1
85
64
87
76
83
75
72
80
87%
OH
OH
1
1,4-selective
(Z-selective)
2
3
15
:
1
O
P
S
P
O
OEt
OEt
HO
CHCl3
No solvent
OEt
HS
2 eq. Et3N
toluene, rt
OEt
toluene, rt
Ph
NO
reaction
64%
1
4
1.2 equiv of freshly distilled dialkyl dithiophosphates were used for each experi-
ment. Olefin isomer ratios were determined by integrating crude 1H NMR. Yields
shown are isolated yields.
hydride shift
Scheme 1. Initial thiophosphate ring opening results.

3234
B. Guo et al. / Tetrahedron Letters 55 (2014) 3232–3234
Ph
(1.2 equiv) in toluene was added dropwise. The reaction mixture
Ph
S->O
was stirred at room temperature until all of the substrate was con-
sumed according to thin layer chromatography (TLC) analysis. Tol-
uene was removed under reduced pressure, and the crude product
was purified using silica gel flash column chromatography.
S
P
NaH, THF
40 °C
78%
S
P
phosphate
transfer
OEt
OEt
OEt
OEt
S
S
OH
ONa
2
5
Acknowledgments
Ph
Ph
Ph
S
P
We are grateful to the National Science Foundation (CHE-
1266365) and The University of Arizona for financial support.
CuL2
3-exo-trig
S
EtO
O
EtO
S
8
7
SNa
ONLY
product
5-exo-tet
6
Supplementary data
Scheme 2. Anionic hopping—unexpected cyclization results.
Supplementary data (experimental procedures are provided as
well as spectral data) associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2014.04.
028.
Assisted ring opening
Non-assisted
ring opening
Intermediate
Ion Pair
OR
OR
H
S
P
S
R
R
O
References and notes
S
P
O
HO
OR
OR
S
1. (a) Ilardi, E. A.; Njardarson, J. T. J. Org. Chem. 2013, 78, 9533–9540; (b)
Njardarson, J. T. Synlett 2013, 787–803; (c) Mack, D. J.; Njardarson, J. T. ACS
Catal. 2013, 3, 272–286.
S
P
S
P
intra- vs.
intermolecular
OR
R
OR
HS
HS
OR
OR
2. (a) Batory, L. A.; McInnis, C. E.; Njardarson, J. T. J. Am. Chem. Soc. 2006, 128,
16054–16055; (b) McGrath, N. A.; Bartlett, E. S.; Sittihan, S.; Njardarson, J. T.
Angew. Chem., Int. Ed. 2009, 48, 8543–8546; (c) Brichacek, M.; Batory, L. A.;
Njardarson, J. T. Angew. Chem., Int. Ed. 2010, 49, 1648–1651; (d) Brichacek, M.;
Batory, L. A.; McGrath, N. A.; Njardarson, J. T. Tetrahedron 2010, 66, 4832–4840;
(e) Mack, D. J.; Njardarson, J. T. Angew. Chem., Int. Ed. 2013, 52, 1543–1547; (f)
Mustard, T. J.; Mack, D. J.; Njardarson, J. T.; Cheong, P. H.-Y. J. Am. Chem. Soc.
2013, 135, 1471–1475.
Figure 2. Mechanistic speculations.
this new reaction is not preceding via a free carbocation as
otherwise the facile hydride shift to form 4 would take place. We
cannot completely rule out a carbocation intermediate, as there
is a possibility that a tight ion pair could be formed between the
carbocation and the dithiolate (Fig. 2). Clearly the carbocation
character at C2 is of great significance for the outcome of this
reaction, with 1,4-addition favored when there is a C2 carbon sub-
stituent present (entries 1–5, Table 2) and 1,2-addition when there
is none (entries 6–8, Table 2). Alternate scenarios include assisted
and non-assisted (direct 1,4-addition) ring openings. We envision
two different assisted ring opening scenarios, wherein both involve
hydrogen bonding between the oxirane atom and the dialkyl
dithiophosphoric acid, but these scenarios differ in terms of
whether the nucleophilic ring opening is intramolecular (one dial-
kyl dithiophosphoric acid involved) or intermolecular (two dialkyl
dithiophosphoric acids involved). The fact that the preferred sol-
vents for this reaction are non-polar and that running the reaction
in the absence of solvent (other competing hydrogen bonds with
reagents) leads to loss of selectivity (Table 1) lends support to
these assisted ring opening scenarios. Presently, we are not able
to convincingly discriminate between these various mechanistic
scenarios.
In summary, we have developed a new mild method for the ring
opening vinyl oxiranes using dialkyl dithiophosphate nucleophiles.
For 1,1-disubstituted vinyl oxiranes the nucleophile adds exclu-
sively at the terminal carbon (C4), and stereoselectively (Z-selec-
tive) when the olefin contains no terminal substituents. Vinyl
oxiranes that do not contain a C2-substituent undergo an alternate
ring opening pathway. We have also shown for the first time that a
product of this reaction can be converted in high yield to a vinyl
thiirane via an anionic hopping-cyclization cascade.
3. Li, F.; Calabrese, D.; Brichacek, M.; Lin, I.; Njardarson, J. T. Angew. Chem., Int. Ed.
2012, 51, 1941–1983.
4. We recently demonstrated that the corresponding vinyl oxetanes ring opened
selectively: Guo, B.; Njardarson, J. T. Chem. Commun. 2013, 10802–10804.
5. (a) Corey, E. J.; Park, H.; Barton, A.; Nii, Y. Tetrahedron Lett. 1980, 21, 4243–
4246; (b) Goldbach, M.; Jakel, E.; Schneider, M. P. J. Chem. Soc., Chem. Commun.
1987, 1434–1435; (c) Ponce, A. M.; Overman, L. E. J. Am. Chem. Soc. 2000, 122,
8672–8676; (d) Degl’Innocenti, A.; Capperucci, A.; Cerreti, A.; Pollicino, S.;
Scapecchi, S.; Malesci, I.; Castagnoli, G. Synlett 2005, 3063–3066; (e)
Yoshimura, Y.; Yamazaki, Y.; Saito, Y.; Takahata, H. Tetrahedron 2009, 65,
9091–9102; (f) Li, Z.; Watkins, E. B.; Liu, H.; Chittiboyina, A. G.; Carvalho, P. B.;
Avery, M. A. J. Org. Chem. 2008, 73, 7764–7767.
6. (a) Dormann, K. L.; Brückner, R. Angew. Chem., Int. Ed. 2007, 46, 1160–1163; (b)
Kang, S.-K.; Ryu, H.-C.; Hong, Y.-T.; Kim, M.-S.; Lee, S.-W.; Jung, J.-H. Synth.
Commun. 2001, 31, 2365–2371; (c) Ji, M.; Choi, H.; Park, M.; Kee, M.; Jeong, Y.
C.; Koo, S. Angew. Chem., Int. Ed. 2001, 40, 3627–3629.
7. (a) Kudelska, W.; Michalska, M. Tetrahedron 1986, 42, 629–636; (b) Mebah, J. M.
D.; Mieloszynski, J. L.; Paquer, D. Phosphorus, Sulfur Silicon Relat. Elem. 1993, 78,
215–222; (c) Michalska, M.; Kudelska, W.; Pluskowski, J.; Koziol, A. E.; Lis, T. J.
Chem. Soc., Perkin Trans. 1 1994, 979–983; (d) Li, Z.; Zhou, Z.; Li, K.; Wang, L.;
Zhou, Q.; Tang, C. Tetrahedron Lett. 2002, 43, 7609–7611; (e) Castro, J.; Cuevas, J.
C.; fiandor, J. M.; Fraile, T.; Heras, F. G.; Ruiz, J. R. Bioorg. Med. Chem. Lett. 2002,
12, 1371–1374; (f) Zhou, Z.; Li, Z.; Quanyong, W.; Liu, B.; Li, K.; Zhao, G.; Zhou,
Q.; Tang, C. J. Organomet. Chem. 2006, 691, 5790–5797.
8. Sadhu, K. M.; Matteson, D. S. Tetrahedron Lett. 1986, 27, 795–798.
9. An interesting racemization of similar allylic phosphorothioate esters has been
reported: Han, X.; Zhang, Y.; Wu, J. J. Am. Chem. Soc. 2010, 132, 4104–4106.
10. We had wondered whether a [3,3]-sigmatropic equilibration pathway would
be
dithiophosphate products from Table 2 (entries 1, 3 and 4) for 12 h in toluene.
For entries and (trisubstituted olefins) we did not observe any
isomerization to the E-olefin isomer. Interestingly, in the case of entry 3
(tetrasubstituted olefin) we observed an isomerization of pure Z-olefin
a possibility. Prompted by a reviewer’s suggestion we refluxed three
1
4
a
product to a 1.4:1 Z/E-olefin mixture. It is important to note that in none of
these isomerization experiments did we observe any other products.
11. Tetrahydrothiophenes can be formed using phosphorothioate esters in the
absence of a competing olefin: Robertson, F. J.; Wu, J. J. Am. Chem. Soc. 2012,
134, 2775–2780.
12. The only report related to such a competition scenario are treatments of cis/
trans-1,4-dihalobutenes with Na2S in MeOH affording varying rations of 2,5-
dihydrothiophene and 3,4-epithio-1-butene: Everhardus, R. H.; Grafing, R.;
Brandsma, L. Recl. Trav. Chim. Pays-Bas 1976, 95, 153–154.
Representative experimental conditions for vinyl oxirane
opening
13. Interestingly, our group has developed a mild copper catalyzed method to
convert vinyl thiiranes, such as 7, to 2,5-dihydrothiophenes (8): Rogers, E.;
Batory, L. A.; McInnis, C. E.; Njardarson, J. T. J. Am. Chem. Soc. 2007, 129, 2768–
2769.
To a flame-dried round bottomed flask charged with a magnetic
stir bar was introduced toluene followed by vinyl oxirane (1 equiv),
at which point freshly distilled diethyl dithiophosphoric acid