Organic Letters
Letter
We initially examined various Li, Na, K, and Cs bases and
solvents to confirm their suitability for the coupling reaction of
4-iodoacetophenone (1a) with NN-H in the presence of a
tetrakis(triphenylphosphine)palladium(0) [Pd(PPh3)4] cata-
lyst (10 mol %) for 1−2 h at 70 °C (Table 1). All trials in
Table 2. Catalytic Activity of Various Palladium Complexes
and Ligands for the Reaction of 1a with NN
Table 1. Cross-Coupling Reaction of 1a with NN-H
a
entry
catalyst
Pd(PPh3)4
additive
yield (%)
1
2
3
4
5
6
7
8
none
63
62
60
65
79
83
73
7
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd2(dba)3·CHCl3
Pd2(dba)3·CHCl3
PdCl2
Pd(PPh3)2Cl2
Pd(OAc)2
Pd-PEPPSI-iPr
none
dppe (10 mol %)
dcype (10 mol %)
BINAP (10 mol %)
dppf (10 mol %)
XPhos (20 mol %)
SPhos (20 mol %)
none
XPhos (20 mol %)
dppf (10 mol %)
none
a
entry
base
solvent
yield (%)
b
1
2
3
4
5
6
7
8
NaOMe
NaOtBu
KOtBu
LiHMDS
NaOMe
NaHMDS
NaOtBu
KOtBu
MeOH
tBuOH
tBuOH
THF
THF
THF
0
b
0
b
0
9
31
b
b
0
10
11
12
13
14
0
b
b
trace
b
trace
b
0
XPhos (20 mol %)
none
trace
b
b
THF
THF
trace
trace
0
0
b
b
none
b
a
b
9
LiHMDS
NaH
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
trace
Isolated yield based on NN-H. Monitored by TLC.
10
11
12
13
14
15
16
21
b
NaOMe
NaHMDS
NaOtBu
NaOtBu
KOtBu
0
respectively). Tris(dibenzylideneacetone)dipalladium(0)−
chloroform adduct [Pd2(dba)3·CHCl3] and various divalent
palladium complexes, PdCl2−bis(triphenylphosphine) [Pd-
(PPh3)2Cl2], PdCl2−[1,1′-bis(diphenylphosphino)ferrocene
(dppf)], [Pd(OAc)2]−[2-dicyclohexylphosphino-2′,4′,6′-trii-
sopropylbiphenyl (Xphos)], and PdCl2−[1,3-bis(2,6-
diisopropylphenyl)imidazol-2-ylidene](3-chloropyridyl) [Pd-
PEPPSI-iPr], showed almost no catalytic activity (entries 8,
10−13, respectively). It is noteworthy that the yield of the
corresponding NN-Ar in this catalytic process (entry 6) was
higher than those obtained upon similar treatment with
transition metal NN nucleophiles such as NN-Au(PPh3)6 and
The acceleration effect of XPhos was also observed by time-
dependent changes in the concentration of the starting
material NN-H, which can be monitored at λmax maxima of
549 nm by means of UV−vis spectrometry (Figure S1). The
initial rate constants kobs (kobs = −d[NN-H]/dt) were
determined to be 5.50 × 10−4 s−1 for the Pd(PPh3)4 (10
mol %)−XPhos (20 mol %) catalyst and 3.42 × 10−4 s−1 for
the Pd(PPh3)4 catalyst (10 mol %) (Figure 1), which indicated
that the reaction with XPhos proceeds 1.6 times faster than
that without the ligand. The linear time dependence of the
concentration of NN-H in the reaction with the Pd(PPh3)4−
XPhos catalyst (Figure 1a and Figure S1) indicated that the
addition effect of XPhos is ascribed to the acceleration of the
rate-determining step in the catalytic process instead of the
inhibition of the deactivation of the catalytically active species.
To gain information about the reaction mechanism, we
monitored the coupling reaction of NN-H and 1a under the
optimized condition by means of electron spin resonance
(ESR) spectroscopy (Figure 1b). The 10-line ESR signal for
NN-H [|aN| (hyperfine coupling of 14N nuclei) = 0.74 mT, and
|aH| (hyperfine coupling of 1H nucleus) = 0.36 mT],6 observed
at the initial stage of the reaction, was gradually converted into
a five-line signal (|aN| = 0.75 mT) that was assigned to 1P. In
spite of the lack of detection of ESR signals assigned to NN-Na
during the catalytic reaction, the formation of a stoichiometric
29
63
40
53
c
b
Cs2CO3
0
a
b
c
Isolated yield based on NN-H. Monitored by TLC. With 5 mol %
Pd(PPh3)4.
the search for suitable conditions for generating the NN radical
anion using a typical combination of base and soluble solvents,
such as NaOMe and MeOH, NaOtBu and tBuOH, LiHMDS
and THF, and NaOtBu and THF, resulted in no reaction or
gave trace amounts of the desired compound. This was
probably due to fast decomposition or disproportionation of
the unstable NN radical anions (entries 1−8). Notably,
NaOtBu and KOtBu provided good results when used in
combination with toluene (entries 13−15). The successful
results can be explained by the low solubility of NaOtBu and
KOtBu in toluene, which keeps the concentration of the NN
anion low during the catalytic process. This specific
equilibration would successfully inhibit the inevitable dis-
proportionation of NN in this particular system. Cs2CO3
afforded a trace amount of the product probably due to its
weak basicity for the generation of the radical anion (entry 16).
The catalytic activity of various palladium complexes was
examined for the coupling reaction of 1a with NN-H (1 equiv)
in the presence of NaOtBu (1 equiv) in toluene (Table 2).
Pd(PPh3)4 was the best catalyst among the catalysts examined
(entries 1−7). The addition of 1,1′-bis(diphenylphosphino)-
ferrocene (dppf), 2-dicyclohexylphosphino-2′,4′,6′-triisopro-
pylbiphenyl (XPhos), or 2-dicyclohexylphosphino-2′,6′-dime-
thoxybiphenyl (SPhos) effectively increased the product yield
(entry 5, 6, or 7, respectively), while the addition of other
phosphines, 1,2-bis(diphenylphosphino)ethane (dppe), 1,2-
bis(dicyclohexylphosphino)ethane (dcype), or 2,2′-bis-
(diphenylphosphino)-1,1′-binaphthyl (BINAP), did not lead
to any significant additive (or ligand) effect (entries 1−4,
B
Org. Lett. XXXX, XXX, XXX−XXX