Organic Letters
Letter
PdNP formation was studied with easily accessible phenyl-
(trimethylsilyl)methanone (4) and bis(dimethyl(phenyl)silyl)-
methanone (5) as photoactive reagents (see the Supporting
Information (SI)). Acyl silane 4 exhibits UV/vis absorption in
the λ = 350−450 nm region (λmax = 418 nm), whereas the UV/
vis absorption of bis(silyl) ketone 5 is red-shifted to the λ =
425−575 nm region (λmax = 544 nm) (see the SI for UV/vis
spectra). For nanoparticle preparation, acyl silane 4 or bis(silyl)
ketone 5 was diluted in dimethylformamide (DMF), and
Pd(OAc)2 was added under an argon atmosphere. The molar
ratio of the silyl ketones with respect to Pd(OAc)2 was set as
40:1. The reaction mixture was then irradiated with LEDs at λ
= 420, 462, or 520 nm under vigorous stirring. When acyl silane
4 was used, irradiation at λ = 420 nm for 30 min provided a
brown solution, indicating the formation of PdNPs. Indeed,
transmission electron microscopy (TEM) revealed the
formation of small nanoparticles with a mean diameter of 5.2
0.9 nm (Figure 1a). The oxidation state of the PdNPs was
irradiation of Pd(OAc)2 in DMF for 3 h, documenting the
importance of both the silyl ketone and light for successful NP
preparation.11 Irradiation (λ = 420 nm) of a mixture of 4,
Pd(OAc)2, (2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO), and
DMF for 30 min provided benzoyl-TEMPO derived from
trapping of benzoyl radicals by TEMPO, showing that
homolytic Norrish Type I C−Si bond cleavage had occurred.
This indicates that the metal salt reduction process likely
follows a radical pathway.1 However, Pd salt reduction by a
siloxycarbene intermediate generated by a light-induced [1,2]-
Brook rearrangement currently cannot be ruled out.1,12
To test the catalytic activity of these PdNPs, they were used
as catalysts in Suzuki−Miyaura cross-coupling of phenylboronic
acid (6a) with iodobenzene (7a) to form biphenyl (8aa).8 The
activity of Pd@5PVP (1.9 0.5 nm) was investigated first, and
8aa was obtained in 83% isolated yield after 3 h at 50 °C with
0.5 mol % catalyst loading (Table 1, entry 1). Surprisingly, the
Table 1. Suzuki−Miyaura Coupling with Pd@PVP Catalysts
a
and Recycling Experiments
b
entry
Pd@PVP
cycle
time (h)
yield of 8aa (%)
1
2
3
4
5
Pd@5PVP
Pd@4PVP
Pd@4PVP
Pd@4PVP
Pd@4PVP
Pd@4PVP
Pd@4PVP
Pd@PVP
1
1
2
3
4
1
1
1
3
83
93
93
83
98
94
92
90
1
1
1.5
1.5
1
Figure 1. TEM images of the prepared PdNPs: (a) Pd@4 and (b)
Pd@5.
c
6
d
7
f
4
8
0.5
investigated by X-ray photoelectron spectroscopy (XPS), which
showed the characteristic binding energies for Pd0 (Pd 3d5/2 at
335.1 eV and Pd 3d3/2 at 340.4 eV), proving the complete
photochemical reduction of Pd ions to Pd0 (see the SI).9 When
bis(silyl) ketone 5 was used, a dark-violet reaction mixture
resulted after 1 h of irradiation at λ = 462 or 520 nm. TEM
analysis revealed in both cases that small nanoparticles were
formed. Under irradiation with blue light, PdNPs with a mean
diameter of 1.9 0.6 nm were obtained, and irradiation with
green light resulted in PdNPs with a similar diameter of 1.9
0.5 nm (Figure 1b).
Notably, the beginning of PdNP formation under LED
irradiation was observed rapidly after 5−10 min using 4 and 5.
The in situ-formed “naked” PdNPs prepared with ketones 4
(Pd@4) and 5 (Pd@5) showed high stability in DMF.
However, after evaporation of the solvent, aggregation to
larger particles was noted. Therefore, the PdNPs were further
stabilized by addition of commercially available PVP (molar
mass = 10 000 g/mol).10 The PdNP−polymer hybrids Pd@
4PVP and Pd@5PVP could be readily isolated by solvent
evaporation without aggregation. Dynamic light scattering
(DLS) measurements indicated that these PdNPs were coated
with a thin PVP layer (∼2.0 nm for Pd@4PVP and ∼1.8 nm for
Pd@5PVP).
a
Reaction with 0.2 mmol of iodobenzene and 0.3 mmol of
b
phenylboronic acid. Isolated yields after column chromatography.
c
Reaction with 1 mmol of iodobenzene and 1.5 mmol of phenyl-
d
f
boronic acid. Reaction with 0.1 mol % Pd@4PVP. The Pd@PVP
catalyst was prepared according to a literature prcedure.8d,15
larger Pd@4PVP (5.2
0.9 nm) showed higher catalytic
activity, and 8aa was obtained in 93% yield after 1 h under
otherwise identical conditions (Table 1, entry 2).13 Possible
reasons for the lower reactivity of Pd@5PVP might be stronger
adsorption of reactive intermediates on the NP surface or
deactivation of the catalyst by fragments derived from the
bis(silyl) ketone 5.13 Notably, Pd@4PVP was readily recycled.
After completion of the reaction, the EtOH/H2O solvent was
evaporated, and 8aa was extracted with 3:1 pentane/CH2Cl2.
Pd@4PVP was not soluble in pentane/CH2Cl2 and could be
dissolved in EtOH after extraction of 8aa. By means of this
simple protocol, product 8aa was obtained in 83−98% yield
over four cycles (Table 1, entries 2−5).
Leaching experiments were also conducted. After 0.5 h, the
Pd@4PVP catalyst was precipitated, and the Pd content of the
supernatant was analyzed with inductively coupled plasma mass
spectrometry (ICP-MS) (see the SI). The Pd content was
found to be 0.1 ppm. When this “leached Pd” was applied as
the catalyst, no biphenyl was formed after 1 h at 50 °C, and
only a 30% yield of biphenyl was obtained after 3 days. Both
experiments indicate that Pd@4PVP rather than leached Pd is
the catalyst in this reaction.8,14
To document the necessity of light for nanoparticle
preparation, the reactions with 4 and 5 were repeated in the
dark. As expected, formation of PdNPs was not observed after
30 min of stirring in these cases. With 5, however, NP
formation was observed after more than 30 min of stirring in
the dark. Furthermore, no nanoparticles were generated upon
B
Org. Lett. XXXX, XXX, XXX−XXX