Journal of the American Chemical Society
Article
decomposition of Ni−arenoxo intermediate C. Probably due
to this difficulty, the substrate generality of this reaction turned
out to be narrow (Scheme 1). For instance, the reactions of
the aryl exchange reaction by using Ni(cod)2/dcypt in toluene
at 150 °C (Figure 3B). Delightfully, the reaction furnished aryl
sulfide 4Ab in 48% yield together with decarbonylative
etherification product 12 in 53% yield. The reverse reaction
was also conducted, generating only a tiny amount of 11A.
This suggested that the reaction is no longer reversible with
11A. In contrast, when 1A was subjected to these conditions
instead of 11A, we confirmed that the reaction was completely
reversible (details in the SI).
Scheme 1. Substrate Scope of the Reaction Using 1A and
a
6A
Encouraged by this result, we optimized the reaction
conditions by using 11A and 3d (Table 2). The use of
a
Table 2. Optimization of Reaction Conditions Using 11A
Zn 60
mol %
Zn(OAc)2
50 mol %
Na3PO4
yield of
yield of
b
a
b
Conditions: 1A or 6A (0.20 mmol), 3 (2.0 equiv), Ni(OAc)2 (10
entry
2.0 equiv
4Ad/%
12/%
mol %), dcypt (15 mol %), Zn (60 mol %), Zn(OAc)2 (50 mol %),
Na3PO4 (2.0 equiv), toluene (0.80 mL), 150 °C, and 24 h.
1
2
3
4
5
23
26
30
46
49
66
20
4
33
54
54
81
○
○
○
○
○
○
○
○
electron-rich (3c) as well as electron-deficient (3d) aromatic
esters with 1A gave the corresponding aryl sulfides 4 in low
yields. Even when we changed sulfide donor 1A to electron-
deficient 6A, the yield of 4Ae did not improve. In order to
solve this issue, we sought an alternative protocol capable of
accelerating the turnover-limiting decomposition of E
selectively.
c
d
c
6
a
Conditions: 11A (0.20 mmol), 3d (2.0 equiv), Ni(cod)2 (10 mol
%), dcypt (15 mol %), toluene (0.80 mL), 150 °C, and 12 h.
Determined by H NMR analysis. Zn (1.0 equiv). 24 h.
b
c
d
1
To this end, we focused on employing 2-pyridyl sulfide as a
sulfide donor (Figure 3A). We previously developed a
Ni(cod)2/dcypt afforded 4Ad in 23% yield and 12 in 20% yield
(Table 2, entry 1). When 11A was used as a sulfide donor, the
addition of Zn(OAc)2 and Na3PO4 was not effective, but the
addition of only Zn improved the yields of 4Ad and 12 (Table
2, entries 2−4). Increasing the amount of Zn slightly improved
the yield (Table 2, entry 5).19 Finally, a longer reaction time
gave the best result (Table 2, entry 6).
With the optimized conditions in hand, we then evaluated
the substrate scope of the present reaction (Scheme 2). First,
we examined the scope of sulfide groups. This reaction
successfully transferred primary, secondary, and tertiary alkyl
sulfide groups to give the corresponding aryl sulfides in
moderate yields (4Bb, 4Cb, and 4Db). Of note, for methyl
sulfide 4Bb synthesis, the present method is more advanta-
geous than typical cross-coupling methods using methanethiol
because methanethiol is a gaseous (bp = 6 °C) and strongly
odorous reagent, often associated with cumbersome handling.
Other alkyl sulfides bearing CF3 (4Eb), cyclobutyl (4Fb, 4Ff),
and arenes (4Gb) were synthesized in acceptable yields.
Nitrogen substituents on sulfide groups were compatible with
the present reaction (4Hb, 4Ib, and 4Ig). The scope of arenes
was next investigated. It was revealed that para and meta
substituents did not effect on this reaction significantly (4Aa
and 4Ah); however, ortho substituents inhibited the reaction
due to steric repulsions with the catalyst (4Ai). With respect to
the electronic nature of substituents, both electron-donating
and electron-withdrawing groups at the para position were
applicable to the present reaction. Aryl sulfides bearing methyl
(4Aj), methoxy (4Ad), phenoxy (4Ak), and dioxole (4Al)
were obtained in moderate yields. Compared to Morandi’s
recent method that was mainly effective for the synthesis of
aryl sulfides with electron-withdrawing groups,14 it is note-
worthy that our present reaction can also synthesize highly
Figure 3. (A) 2-Pyridyl sulfide for a funneling strategy. (B) Forward
reaction and reverse reaction of 11A. See the SI for details.
decarbonylative synthesis of diarylethers, where we found
that only the reductive elimination of 2-azinyl−O bonds from a
Ni/dcypt complex was facile but that of other aryl−O bonds
was difficult.17,18 Accordingly, the use of 2-pyridyl sulfide can
“funnel” this reversible catalytic cycle into desired products 4
by the selective enhancement of the reductive elimination of
intermediate E. To support this funneling strategy, we first
subjected 2-pyridyl sulfide 11A, which can be synthesized from
pyridine-2-thiol with alkyl bromides, with aromatic ester 3b to
10336
J. Am. Chem. Soc. 2021, 143, 10333−10340