Preparation of Secondary Phosphine Oxide Ligands through Nucleophilic Attack on Imines and Their Applications in Palladium-Catalyzed Catellani Reactions

Article abstract of DOI:10.1002/ejic.201600181

Several new amino-type secondary phosphine oxide (SPO) pre-ligands (3a–3h) that contain P–N bonds were synthesized and characterized. SPOs 3a–3h can tautomerize to phosphinous acids (PA, 3a–3h) as genuine ligands. The formation of SPOs 3a–3h occurred first through nucleophilic attack on the imine carbon atom, then by the addition of RPCl₂(R = Ph, Cy, tBu, or iPr), and work-up under acidic conditions. The P–N bond in the newly prepared SPOs is evident from the crystal structures of SPOs 3d and 3h. Reactions of SPOs 3f, 3g, or 3h with Pd(COD)Cl₂(COD = cyclooctadiene) yielded palladium complexes 6f, 6g, or 6h. In these crystal structures, PAs 3f, 3g, and 3h act as didentate ligands through P and N donors. Intriguingly, the reaction of SPO 3f with Pd(COD)Cl₂also gave rise to palladium complexes 7fa, 7fb, and 8f. The crystal structures of 7fa and 7fb show that the palladium atom is chelated by PA 3f and coordinated by a phosphine-like ligand fragmented from SPO 3f through C–N bond dissociation. Finally, the syntheses of carbazole derivatives were pursued in Catellani reactions with SPOs 3g and 3h as pre-ligands. A mechanism is proposed to account for the catalytic reaction (see the Supporting Information). The optimized conditions for Suzuki reactions using selected SPO ligands are also reported.

Full text of DOI:10.1002/ejic.201600181

DOI: 10.1002/ejic.201600181
Full Paper
Pd Catalysts
Preparation of Secondary Phosphine Oxide Ligands through
Nucleophilic Attack on Imines and Their Applications in
Palladium-Catalyzed Catellani Reactions
Chan-Yu Hu,[a] Ya-Qian Chen,[a] Guan-Yu Lin,[a] Ming-Kai Huang,[a] Yu-Chang Chang,[a] and
Fung-E Hong*[a]
Abstract: Several new amino-type secondary phosphine oxide 3h act as didentate ligands through P and N donors. Intrigu-
(SPO) pre-ligands (3a–3h) that contain P–N bonds were synthe- ingly, the reaction of SPO 3f with Pd(COD)Cl2 also gave rise to
sized and characterized. SPOs 3a–3h can tautomerize to phos- palladium complexes 7fa, 7fb, and 8f. The crystal structures of
phinous acids (PA, 3a–3h) as genuine ligands. The formation of 7fa and 7fb show that the palladium atom is chelated by PA
SPOs 3a–3h occurred first through nucleophilic attack on the 3f and coordinated by a phosphine-like ligand fragmented from
imine carbon atom, then by the addition of RPCl2 (R = Ph, Cy, SPO 3f through C–N bond dissociation. Finally, the syntheses of
tBu, or iPr), and work-up under acidic conditions. The P–N bond carbazole derivatives were pursued in Catellani reactions with
in the newly prepared SPOs is evident from the crystal struc- SPOs 3g and 3h as pre-ligands. A mechanism is proposed to
tures of SPOs 3d and 3h. Reactions of SPOs 3f, 3g, or 3h with account for the catalytic reaction (see the Supporting Informa-
Pd(COD)Cl2 (COD = cyclooctadiene) yielded palladium com- tion). The optimized conditions for Suzuki reactions using se-
plexes 6f, 6g, or 6h. In these crystal structures, PAs 3f, 3g, and lected SPO ligands are also reported.
Introduction
metal-catalyzed cross-coupling reactions have also been evalu-
ated and reported elsewhere.[6,4d,4f–4u]
For many years, the various shapes and functions of trisubsti-
tuted phosphines have been the favorite choice for ligand-as-
sisted transition-metal-catalyzed reactions.[1] Nevertheless,
these types of phosphines are mostly vulnerable to oxidation,[2]
which lessens their bonding capacities toward low-valence tran-
sition metals.[3] This severe drawback limits their extensive ap-
plication as authentic soft ligands towards soft metals in coordi-
nation. It is also not a desirable character for the long-term
storage of some precious phosphine compounds. One way to
resolve this obstacle is to allow phosphine to be partially oxi-
dized and maintain its coordinating capacity until it is needed.
Phosphinous acid [PA, P(OH)R1R2], a tautomeric form of its cor-
responding air-stable secondary phosphine oxide [SPO, O=
P(H)R1R2] in solution, is thus a potential candidate and has the
capacity to act as a legitimate phosphine ligand in the coordi-
nation of transition metals (Scheme 1).[4] Since the first synthe-
sis of di-n-alkylphosphine oxides in 1952, various kinds of SPOs
have been prepared in the past several decades and their stabil-
ities towards air and moisture have also been validated.[5,4q]
Moreover, their capabilities as effective ligands in transition-
Scheme 1. Process for the conversion of a secondary phosphine oxide (SPO)
to its corresponding tautomeric form, phosphinous acid (PA), and the onward
formation of a PA-coordinated metal complex.
Besides its commonly formulated simple O=P(H)R1R2 format,
recently designed secondary phosphine oxides come in various
shapes and forms. Pfaltz and Helmchen reported an authentic
bidentate ligand that was formed by incorporating an imine
moiety into the SPO framework [Figure 1(a)].[7] Recently, Acker-
mann demonstrated that some N–P–N-type secondary phos-
phine oxides (or heteroatom-substituted secondary phosphine
oxides, HASPOs) could behave as excellent ligands in various
[a] Department of Chemistry, National Chung Hsing University,
250 Kuo-Kuang Road, Taichung 40227,
Taiwan
E-mail: fehong@dragon.nchu.edu.tw
http://www.nchu.edu.tw/~chem/fehong.htm
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/ejic.201600181.
Figure 1. Two categories of secondary phosphine oxides.
1 © 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 0000, 0–0

Full Paper
Scheme 2. Catellani-type Heck and Suzuki reactions.
catalytic reactions [Figure 1(b)]. By combining with a suitable more deeply into the mechanism of this type of reactions and
palladium salt, these HASPOs worked well for Suzuki–Miyaura thereby improving the catalytic performances.[17] As studies on
reactions as well as amination reactions, even reactions starting SPO-assisted Catellani-type reactions are rare, it is of interest to
from inactive aryl chlorides.[8] Upon deprotonation, Cramer et us to explore the roles played by SPO-type phosphine ligands
al. showed that this N–P–N-type SPO could act as hemilabile in this type of reactions.
ligand that binds Ni0 and AlIII in the asymmetric Ni-catalyzed
Herein, we report the preparation of several new amino-
based secondary phosphine oxides, which contain P–N bonds
hydrocarbamoylation of alkenes.[9]
Recently, phosphines containing P–N bonds in their scaffolds in their frameworks. Their capacities as mono- or bidentate li-
have attracted much attention. It is believed that the phos- gands in several palladium-catalyzed Catellani reactions were
phorus atom in this type of phosphines is significantly electron- evaluated. The structures of several fascinating SPO-ligated pal-
richer than those without the P–N bond, a characteristic that is ladium complexes with unique bonding modes will be dis-
beneficial to the oxidative addition process in the commonly cussed as well.
seen cross-coupling catalytic cycles.[10]
In addition, cross-coupling reactions are well-known proc-
esses for their efficiencies in joining two separate components,
Results and Discussion
in particular, for the formation of carbon–carbon bonds from
Preparation of Amino Secondary Phosphine Oxides 3a–3h
the corresponding aryl or alkyl reactants.[11] Nevertheless, these
methods are limited by their scopes in only combining two
substances at a time. For further incorporation of a third sub-
stance to the coupled compound, more step(s) must be taken.
It would be advantageous if three reactants could be coupled
in the same reaction, for example, in a one-pot reaction.
One of the most attractive multiple-component reactions
was reported in the mid-1980s by Italian chemist M. Catellani
and was later coined as the “Catellani reaction”.[12] At first
glance, the reaction seems not to have many differences from
the conventional cross-coupling reaction (Scheme 2). Neverthe-
less, this type of reaction deliberately promotes the C–H activa-
tion process of Ar–I at its ortho-position and leads to the re-
placement of ortho-hydrogen atoms by an alkyl or aryl group,
which indeed is a fascinating achievement. As seen, C–H activa-
tion has been the target for extensive studies recently.[13] The
Catellani reaction can readily achieve C–H activation of the
ortho-hydrogen atom(s) of aryl halides (mostly aryl iodide). It is
worth noting that norbornene plays an indispensable role as
co-catalyst to promote the targeted route for the reaction. De-
termined by the final elementary step, Catellani developed the
corresponding Catellani-type Heck and Suzuki reactions, which
result in the production of the targeted products after C–H acti-
vation processes.[14]
Several amino secondary phosphine oxides, 3a–3h, were pre-
pared and characterized by spectroscopic methods (Scheme 3,
Figure 2). The general procedure for the preparation of the
compounds is presented as follows. Initially, imines 1 were pre-
pared from each corresponding aldehyde and amine. Subse-
quently, treatment of imines 1 first through nucleophilic attack
on the imine carbon atom, followed by the addition of PhPCl2
thus led to the formation of each matching phosphinamines (2
or 2′ for disubstituted products). Finally, conversion of 2 into
each of the corresponding phosphinous acids 3′ was achieved
through hydrolysis in acidic media. In solution, a tautomeric
equilibrium can exist between the secondary phosphine oxide
3 and its phosphinous acid 3′ under ambient conditions. Nor-
In their early works, Catellani et al. did not use any ligand as
an auxiliary, and the amount of palladium catalyst used was up
to 20 %.[14b,15] Later, Lautens and co-workers employed some
of well-chosen ligands for the reactions and achieved better
performances.[16] Nevertheless, the role played by conventional
ligands such as phosphines remains doubtful in this type of
reactions. Ever since the disclosure of the original Catellani reac-
tion, there have been many interesting reports aimed at delving
Scheme 3. Preparation of amino secondary phosphine oxides. Substituents R
on the four positions could be: phenyl, 3-methoxyphenyl, 2-methoxyphenyl,
4-methoxyphenyl, pyridyl, 2-methyl-1H-imidazolate, thiazole, tert-butyl, iso-
propyl, cyclohexyl.
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
2
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
mally, the former is more favorable in the absence of highly O(2)–P(1)–N(1), O(2)–P(1)–C(19), and N(1)–P(1)–C(19) are
electron-withdrawing substituent(s). The distinctly large cou- 117.42(8)°, 113.62(8)°, and 108.63(8)°, respectively. Clearly, a
pling constant (JP, H = 450–600 Hz) indicates the presence of a pseudo-tetrahedral environment surrounds the phosphorus
P(–H)(=O) fragment in the secondary phosphine oxide.
atom. The major structural features of 3d are also true for 3h.
Reactions of Amino Secondary Phosphine Oxide 3f with
Palladium Salts
The fact that imidazole moieties exhibit coordination capacity
towards palladium through the N(imidazole) site has been dem-
onstrated elsewhere.[6c] Thereby, compound 3f (or 3g, 3h) is
expected to have the potential to act as a bidentate ligand
through both the P(phosphine) and N(imidazole) sites.
Here, 3f was selected to react with various palladium sour-
ces, PdX2 (X = Cl, Br), Pd(COD)X2 (COD = cyclooctadiene),
PdCl2(CH3CN)2, or Pd(OAc)2, and the processes were monitored
by 31P NMR spectroscopy (Scheme 4). The disappearance of the
distinctly large coupling constant, which is caused by the P–H
bond in the SPO form, is strong evidence for the formation of
the phosphinous acid (PA) form and that the compound acts as
a phosphine ligand in the palladium complex. The generation
of the cis (or trans) form of bis(phosphine) ligands coordinated
to the palladium complex, 4f_cis (or 4f_trans), was expected
as the initial and major product by employing PdX2 (X = Cl, Br)
as the palladium sources.[6] In addition, the formation of 5f is
presumed when using Pd(OAc)2 as the palladium source. To our
delight, the crystal structure of 6f was able to be determined
by X-ray diffraction methods (Figure 4). As shown in Figure 4,
the SPO 3f indeed behaves as a bidentate ligand through both
the P(phosphine) and N(imidazole) coordinating sites. Similar
results were also observed for the reactions of 3g and 3h with
PdCl2 [or Pd(COD)Cl2]. The major structural features of 6g and
6h are analogous to that of 6f. Indeed, 3g and 3h also behave
as bidentate ligands (Figure 4).
Figure 2. Secondary phosphine oxides, 3a–3h, and chlorophosphine 2x.
It was found that the hydrolysis of each 2 species indeed led
to the formation of each corresponding SPO 3a–3h except for
2x. The existence of a P–Cl bond in 2x is evident from the
determination of its crystal structure by X-ray diffraction meth-
ods (see the Supporting Information, Figure°S1). To our delight,
Moreover, two fascinating isomeric forms of palladium com-
the crystal structures of 3d and 3h were resolved (Figure 3). plexes, 7fa and 7fb, were also obtained from crystal-growing
The existence of a double bond between P(1) and O(2) in 3d is processes. The crystal structures of 7fa and 7fb revealed that
evident from its short bond, 1.4839(13) Å. The bond angles of 3f indeed acts as a PN bidentate ligand through both the
Figure 3. ORTEP drawings of 3d (left) and 3h (right).
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
3
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 4. Reactions of 3f with Pd(COD)Cl2 and Pd(OAc)2. The compounds highlighted by boxes were characterized by single-crystal X-ray diffraction methods,
whereas those in brackets are presumably reaction intermediates or metastable compounds, which could not be characterized.
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
4
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
P(phosphine) and N(imidazole) sites, like those shown in com-
plex 6f, yet in a much more complicated manner. The structure
of 7fa shows that there is another fragmented phosphinous
ligand, P(OH)(tBu)(NHPh), from 3f that coordinates to palladium
as well (Figure 5). Notably, there is intramolecular hydrogen
bonding between these two oxygen atoms. The palladium
atom resides in a square-planar environment. The process for
the generation of 7fa (or 7fb) is remarkable and worthy of fur-
ther examination. Presumably, a bis-3f-coordinated palladium
dichloride complex 4f_cis (or 4f_trans) was initially formed,
then one of the coordinated ligands of 4f_cis (or 4f_trans) was
fragmented into 2-[chloro(3-methoxyphenyl)methyl]-1-methyl-
1H-imidazole plus one phosphine ligand, 1-tert-butyl-1-
hydroxy-N-phenylphosphanamine, by attack on the N–C bond
of the coordinated 3f by the released proton and chloride. Here,
the released proton presumably was released during the forma-
tion of intramolecular hydrogen bonding between the two PA-
type ligands in 7fa. It is quite unusual that the strong N–C bond
of 3f is broken rather than the presumably weak P–N bond. The
rather similar structure of 7fb compared to that of 7fa was
resolved by single-crystal X-ray diffraction methods (Figure 5).
The only difference lies in the conformation of the monoden-
tate ligand P(OH)(tBu)(NHPh). Unfortunately, the quantities of
7fa and 7fb were not sufficient to execute other spectroscopic
methods than the above-mentioned single-crystal X-ray diffrac-
tion methods.
Figure 5. ORTEP drawings of 7fa (top) and 7fb (bottom).
Another quite unique palladium complex, 8f, was also ob-
Figure 4. ORTEP drawings of 6f (top), 6g (middle), and 6h (bottom).
Eur. J. Inorg. Chem. 0000, 0–0 www.eurjic.org
tained from crystal-growing processes. The crystal structure of
5
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 6. ORTEP drawing of 8f.
8f (Figure 6) reveals that it is a Pd dimer, albeit without a direct phosphine ligands, triphenylphosphine and 1,2-bis(diphenyl-
Pd–Pd bond. Rather, the two Pd atoms are bridged by two de- phosphino)ethane (dppe), were selected to combine with
protonated P(tBu)(OH)(O)– ligands. The tert-butylphosphonous Pd(OAc)2 as catalyst precursors as well as the pre-formed palla-
acid, P(tBu)(OH)2, is believed to derive from the cleavage of the dium complex 6h, for this catalytic process. As already revealed
P–N bond of 3f during hydrolysis, which is different from the in the crystal structure of 6h, 3h could act as a P∩N bidentate
previous case, that is, the formation of 7f, where the N–C bond ligand through both the P(phosphine) and N(imidazole) sites in
of 3f was broken. The core structure of the compound is a six- the coordination of the palladium atom. Therefore, the ratio of
membered ring, which is more like a boat form than a chair Pd/ligand is preset at ca. 1:1 for this type of ligands. By contrast,
form.[18] The two chloride atoms are absent from the starting triphenylphosphine and 3b are monodentate ligands, and the
palladium source, PdCl2. Still, each palladium atom remains in Pd/ligand ratio is designated as 1:2. The reaction conditions are
its initial valence state, +2. Two sets of intramolecular hydrogen as follows: 2-iodotoluene (0.2 mmol), N-(2-bromophenyl)aceta-
bonding can be clearly seen. The surrounding environment of mide (0.4 mmol), base (0.8 mmol), norbornene (0.42 mmol),
Pd(1) is of square-planar geometry. This is also true for Pd(2). palladium salt (10 mol-%), ligand (11–22 mol-%), dimethylacet-
Indeed, the structure of 8f presents a quite unique feature of amide (DMA; 2 mL), which were allowed to react at 135 °C for
its own; two sets of peaks in the 31P NMR spectra corresponding a preset reaction time of 20 h. As shown in Entry 1 (Table 1), a
to four phosphorus atoms are observed in a ratio of 1:1.
modest conversion was obtained, even in the absence of any
phosphine ligand, which is in contradiction to Catellani's previ-
ous observation.[19] In general, the efficiency of the reaction
Application of Selected Complexes in the Preparation of
Carbazoles
Table 1. Catellani reaction for the formation of 1-methyl-9H-carbazole by em-
ploying various ligands.[a]
Catellani et al. reported a fascinating process for the synthesis
of bioactive carbazole derivatives in a one-pot manner. The
process utilizes palladium-catalyzed sequential C–C and C–N
formations from ortho-substituted iodoarenes and N-acetylated
ortho-bromoanilines in the presence of norbornene.[19] Later,
they claimed that the addition of triphenylphosphine (PPh3)
was beneficial for this reaction to achieve significant conversion
when using acetamide as one of the reactants. By using the
selected SPOs 3 as pre-ligands, results of the optimized reaction
conditions for palladium-catalyzed Suzuki–Miyaura reactions
are reported along with results for Heck reactions (in the Sup-
porting Information).
Entry
Ligand
Yield [%][b]
10
11
12
1
–
3b
3h
39.6 (26.4, 25.9[c]
46.4 (36.4)
50.9 (41.1)[f]
54.6 (38.6)
50.5 (37.3)
33.6 (23.5)
46.6 (35.9)
67
)
20.1
14.7
9.0[f]
16.9
13.9
14.0
15.2
–
21.6
33.5
21.2[f]
25.1
33.6
26.9
36.3
–
2[d]
3[e]
4
6h
5[d]
6[e]
7[g]
PPh3
dppe
PPh3
[a] Conditions: 2-iodotoluene (0.2 mmol), N-(2-bromophenyl)acetamide
(0.4 mmol), norbornene (0.42 mmol), K2CO3 (0.8 mmol), Pd(OAc)2 (10 mol-%),
DMA (2 mL), 135 °C, 20 h. [b] Conversion rate was determined by 1H NMR
spectroscopy. Isolated yields are shown in parentheses. [c] DMF was used. [d]
Pd/L = 1:2. [e] Pd/L = 1:1. [f] Reaction for 1 h. [g] 2-Iodotoluene (1.1 equiv.),
N-(2-bromophenyl)acetamide (1.0 equiv.), norbornene (0.25 equiv.), Pd(OAc)2
(5 mol-%), PPh3 (10 mol-%), K2CO3 (2.2 equiv.) in DMF at 105 °C, 48 h.
We wondered whether these newly prepared SPOs could be
utilized as effective ligands for this catalytic system. Thereby,
some of the ligands, 3b and 3h, along with two well-known
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
6
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 5. Palladium-catalyzed sequential C–C and C–N formations by Catellani-type reactions.
using bidenate ligands seems to be better than those utilizing thus speeding up the overall rate of the catalytic cycle. The fact
monodentate ones (Entries 2, 4, 6 vs. 5). There is not much that poor performance was observed in the absence of any
difference when using the pre-formed catalyst precursor ligand indicates that these amino secondary phosphine oxides
against the in situ method (Entry 2 vs. 4). Overall, the applica- are indeed beneficial to the catalytic performance (Entry 4). Li-
tion of these newly prepared SPOs in this catalytic system are gand 3c was assumed to have the potential to act as a biden-
slightly better than those without them. Note that a much tate ligand through the phosphine and pyridyl coordinating
shorter reaction time was needed than that in Catellani's reac- sites. Nevertheless, even worse performance was observed
tion. As shown, the reaction is almost complete within 1 h (En- when using the combination of Pd(OAc)2 and 3c in a 1:1 ratio
try 3). No further improvement in yield for the targeted com- as the catalytic precursor (Entry 3). Therefore, it is more likely
pound could be observed, although there was some alteration that 3c behaves as a monodentate ligand in this case.
in the yields of some side products, 11 and 12 (Scheme 5).
Table 2. Suzuki–Miyaura coupling reactions of 4-bromoanisole and phenyl-
A mechanism is proposed to explain the experimental results
for the formation of the three products observed (Supporting
Information). The routes for the production of 10 and 11 are
proposed in accord with Catellani's work. The observation of 12
is of significant interest to us as there is only the initial acetam-
ide but no iodobenzene fragment incorporated in the structure.
There are two catalytic cycles (cycles 1 and 2) as can be seen
in section S3 in the Supporting Information that might compete
for the consumption of acetamide. Thereby, the reaction was
purposely designed by initially setting the ratio of 2-iodotolu-
ene/N-(2-bromophenyl)acetamide equal to 1:1 or 2:1. In the
former case, only the formation of carbazole 10 was observed.
There was no trace amount of 12 detected. By contrast, both
carbazoles 10 and 12 were found in the latter case. Thereby, it
is safe to say that the rate for the formation of carbazole 10 is
much faster than the generation of 12. It is only after the com-
plete consumption of iodobenzene in the major catalytic cycle
that the reaction of N-(2-bromophenyl)acetamide started to re-
act with Pd0 and led to the production of 12.
boronic acid, employing ligands 3a–3c.[a]
Entry
Ligand
NMR conv. [%]
1
2
3
4
3a
3b
3c
–
38
> 99
52 (41)[b]
25
[a] Conditions: 4-bromoanisole (1.0 mmol), phenylboronic acid (1.5 mmol),
KOH (5 mmol), Pd(OAc)2 (1.0 mol-%), ligand (2.0 mol-%), THF (1 mL), 60 °C,
2 h. [b] [Pd]/3 = 1:1.
In addition, the ratio of Pd(OAc)2/3b was changed from 1:1
to 1:3 intentionally. Not much difference was observed
(Table 3). Therefore, the ratio of Pd(OAc)2/3b was set at 1:2
thereafter.
Table 3. Suzuki–Miyaura coupling reactions of 4-bromoanisole and phenyl-
boronic acid, employing various Pd/L ratios.[a]
Entry
Pd/L
NMR conv. [%]
1
2
3
1:1
1:2
1:3
> 99
> 99
> 99
[a] Conditions: 4-bromoanisole (0.5 mmol), phenylboronic acid (0.75 mmol),
KOH (2.5 mmol), Pd(OAc)2 (1 mol-%), 3b (1–3 mol-%), THF (1 mL), 60 °C, 1 h.
Optimization of Palladium-Catalyzed Suzuki–Miyaura
Reactions of Phenylboronic Acid and Various Bromides by
Using 3a, 3b and 3c as Ligands
As is generally accepted, the starting active catalytic precur-
sor in the Suzuki–Miyaura cross-coupling catalytic cycle is a ze-
As demonstrated repeatedly, the catalytic performance is al- rovalent palladium species. Nevertheless, the most commonly
ways better for an active catalyst precursor, ligand-coordinated used palladium sources are PdII owing to their resistance to air
palladium complex, with a ratio of 1:2 for monodentate ligand/ in long-term storage. Pd(OAc)2 is prone to be reduced more
palladium salt. Initially, the catalytic performances of various readily to zerovalent Pd in the presence of electron-rich phos-
combinations of palladium sources and ligands were investi- phines than halogenated palladium salts, PdX2. In addition, its
gated. The results for the Suzuki–Miyaura coupling reactions availability, relatively cheap cost, ease of handling, and most of
using 4-bromoanisole as the substrate are summarized in all, its efficiency in cross-coupling reactions makes Pd(OAc)2 a
Table 2. As shown, 3b stands out as the best ligand among all favorite choice among the various palladium sources. Here, the
three SPOs with the combination of Pd(OAc)2/ligand = 1:2 (En- screening of various available palladium sources was pursued.
try 2). It is reasonable to assume that the large steric hindrance Interestingly, Pd(OAc)2 as well as all halogenated palladium
of the tBu group in 3b assists the rate of reductive elimination, compounds had excellent performances with short reaction
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
7
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
times of 1.0 h (Table°4, Entries 1–4). Unexpectedly, relatively the optimized reaction solvent as its performance is excellent
poor performance was observed for the zerovalent palladium and it does not undergo self-coupling, such as in the case of
source, Pd2(dba)3 (Entry 5; dba = dibenzylideneacetone). As DMF. After screening these influential factors, it was found that
shown, at least a minimum of 30 min was required for the re- the optimal reaction conditions are as follows: using 3b/
duction of the divalent palladium salt to the zerovalent active Pd(OAc)2 as the catalyst precursor combined with a KOH/THF
species. Once the active species is formed, the completion of system, which are allowed to react at 60 °C for 1 h.
the reaction can be carried out within 30 min.
Table 6. Suzuki–Miyaura coupling reactions of 4-bromoanisole and phenyl-
boronic acid, employing various solvents.[a]
Table 4. Suzuki–Miyaura coupling reactions of 4-bromoanisole and phenyl-
boronic acid, employing various palladium sources.[a]
Entry
Solvent
NMR conv. [%]
Entry
Pd source
NMR conv. [%]
1.0h
1
2
3
4
5
6
THF
DMF
dioxane
toluene
DMSO
hexane
> 99
> 99
92
20
7
0.5 h
2.0 h
1
2
3
4
5
Pd(OAc)2
Pd(COD)Cl2
[Pd(allyl)Cl]2
PdBr2
0
0
0
0
0
> 99
> 99
> 99
> 99
84
> 99
> 99
> 99
> 99
> 99
0
Pd2(dba)3
[a] Performed under the same reaction conditions as given in the footnote
of Table 5 except for the solvent employed.
[a] Conditions: 4-bromoanisole (1.0 mmol), phenylboronic acid (1.5 mmol),
KOH (5 mmol), palladium salt (1.0 mol-%), 3b (2.0 mol-%), THF (1 mL), 60 °C.
Conclusions
The participation of base in Suzuki–Miyaura coupling reac-
tion is essential for the success of the catalysis. It is also one of
the most unpredictable factors. Although the role played by
base has been examined theoretically by computational means,
careful laboratory examination is unavoidable. Thereby, the im-
pact of employing various bases in the reaction was investi-
gated. As shown in Table 5, KOH (or K3PO4) in THF is the most
effective among all the bases screened (Entries 1 and 2). Moder-
ate performance was observed when using NaOtBu as base (En-
try 3), whereas, the other bases, including K2CO3, KOtBu, and
NaOH, performed poorly (Entries 4–6). Unexpectedly, one of the
commonly used bases, Cs2CO3, had no effect on the catalytic
performance (Entry 7). As shown in Entry 8, no conversion was
observed without the presence of base. KOH is chosen as the
base thereafter as its performance is even better than that of
K3PO4 at a lower reaction temperature.
Several new air- and moisture-stable amino secondary phos-
phine oxide pre-ligands (SPOs, 3a–3h) containing P–N bonds
were prepared and applied in palladium-catalyzed Catellani re-
actions. Eight crystal structures, including those of the two SPOs
3d and 3h as well as of the three PA-chelated PdII complexes 6f,
6g, and 6h were reported. Interestingly, two isomeric palladium
complexes, 7fa and 7fb, with fascinate structures were ob-
tained from the reaction of 3f with Pd(COD)Cl2. Complexes 7fa
and 7fb are isomers, and the palladium atom in each isomer is
coordinated by a newly formed PA ligand of P(OH)(tBu)(NHPh)
through C–N bond dissociation of 3f. As to the crystal structure
of 8f, the two bridging ligands, P(tBu)(OH)(O)–, could be frag-
ments resulting from C–N and P–N bond dissociations of 3f.
The Catellani-type reactions for the productions of carbazoles
also yielded valuable organic compounds through unexpected
routes. By using 4-bromoanisole as the substrate, the selected
SPO ligands were employed to find the optimized conditions
for palladium-catalyzed Suzuki and Heck reactions (see the Sup-
porting Information).
Table 5. Suzuki–Miyaura coupling reactions of 4-bromoanisole and phenyl-
boronic acid, employing various bases.[a]
Entry
Base
NMR conv. [%]
1
2
3
4
5
6
7
8
KOH
K3PO4
NaOtBu
K2CO3
KOtBu
NaOH
Cs2CO3
–
> 99
> 99
48
16
12
6
Experimental Section
General: All reactions were carried out under nitrogen by using
standard Schlenk techniques or in a nitrogen-flushed glovebox.
Freshly distilled solvents were used. All processes for separations of
the products were performed by Centrifugal Thin Layer Chromatog-
raphy (CTLC, Chromatotron, Harrison model 8924) or column chro-
matography. GC–MS analysis was performed with an Agilent 5890
gas chromatograph (Restek Rtx-5MS fused silica capillary column:
30 m, 0.25 mm, 0.5 μm) with an Agilent® 5972 mass-selective de-
tector. Routine 1H NMR spectra were recorded with a Varian-400
spectrometer at 399.756 MHz. The chemical shifts are reported in
ppm relative to the internal standard TMS (δ = 0.0 ppm). 31P and
13C NMR spectra were recorded at 161.835 and 100.529 MHz, re-
spectively. The chemical shifts for the former and the latter are
reported in ppm relative to the internal standards H3PO4 (δ =
0.0 ppm) and CHCl3 (δ = 77 ppm), respectively. Mass spectra were
recorded with a JOEL JMS-SX/SX 102A GC/MS/MS spectrometer.
0
0
[a] Conditions: 4-bromoanisole (0.5 mmol), phenylboronic acid (0.75 mmol),
base (2.5 mmol), Pd(OAc)2 (1 mol-%), 3b (2 mol-%), THF (1 mL), 60 °C, 1 h.
For most of cross-coupling reactions, the generation of ionic
intermediate(s) is proposed during the catalytic cycle. Thereby,
polar solvents are always a favorite choice as the solubility of
these species will greatly affect the catalytic performance. As
shown in Table 6, the combination of KOH with either THF, 1,4-
dioxane, or DMF leads to excellent results in 1.0 h (Entries 1–3).
Toluene and DMSO performed poorly (Entries 4 and 5). Even
worse performance was observed when using the nonpolar sol-
vent hexane (Entry 6). From these results, THF was chosen as Electrospray-ionization high-resolution mass spectra (ESI-HRMS)
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
8
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
were recorded with a Finnigan/Thermo Quest Mat 95 XL mass spec-
trometer. The deviations of the experimental values from the theo-
Spectroscopic Data for 3c: 1H NMR (CDCl3): δ = 7.96 (d, JP, H
=
476 Hz, 1 H, P-H), 7.68–7.27 (m, 9 H), 7.21–7.13 (m, 2 H), 6.65 (dd,
retical predictions in the elemental analysis for compounds were J = 6.4, 2.4 Hz, 3 H), 4.12 (d, JP, H = 14.2 Hz, 1 H), 0.99 (s, 9 H) ppm.
presumably caused by encapsulated solvents, CH2Cl2, THF, H2O.
13C NMR (CDCl3): δ = 159.7, 151.6, 148.5, 135.7, 132.7, 131.7, 130.6,
Therefore, HRMS of these compounds have been provided to vali- 128.6, 123.1, 122.0, 116.8 (d, JP, C = 110.0 Hz, 1 C, Ph), 113.0 (16 C, 2
date the identities of them (see Figure°S5).
Ph, py), 66.6 (1 C), 35.6 (1 C, tBu), 26.8 (3 C, tBu) ppm. 31P NMR
(CDCl3): δ = 22.4 (d, JP, H = 476 Hz) ppm. HRMS (EI): calcd. for
C
Synthesis and Characterization of 3a–3h: Benzaldehyde (10.2 mL,
100 mmol) (or picolinaldehyde, 9.5 mL, 100 mmol), aniline (9.1 mL,
100 mmol), and H2O (25 mL) were added into a 250 mL round-
22H25N2OP [M+] 364.1705; found 364.1705. C22H25N2OP (364.43):
calcd. C 72.51, H 6.91, N 7.69; found C 69.06, H 6.90, N 6.82.
bottomed flask. The solution was then stirred at 25 °C for 3 h. The Spectroscopic Data for 3d: 1H NMR (CDCl3): δ = 8.51 (d, J = 4.0 Hz,
resulting organic layer was extracted with ethyl acetate and saline
1 H, Ar), 7.90 (d, J = 8.0 Hz, 1 H, Ar), 7.66 (td, J = 7.7, 2.0 Hz, 1 H,
solution. Then, it was dried with anhydrous MgSO4. A dark-yellow
Ar), 7.45 (d, J = 8.0 Hz, 2 H, Ar), 7.25–7.10 (m, 5 H, Ar), 6.92 (d, J =
solid 1a (17.2 g, 95.0 mmol, 95 %) or 1e (17.3 g, 95.0 mmol, 95 %) 8.0 Hz, 1 H, Ar), 6.88 (s, 1 H, Ar), 6.87 (d, JP, H = 512 Hz, 1 H, P-H),
was obtained after the solvent had been removed under vacuum.
Without further purification, 1a (1.08 g, 6.0 mmol) or 1e (1.08 g,
6.0 mmol) was placed into a 250 mL round-bottomed flask with a
side arm and magnetic stirrer. The air in the flask was removed and
replaced by nitrogen. THF (15 mL) was added and the solution was
cooled to –78 °C. Then, tBuLi (3.6 mL, 7.2 mmol) was added slowly
to the solution along the flask wall. The mixed solution was then
stirred for 1 h before it was warmed up to 25 °C. Meanwhile, PhPCl2
6.81 (dd, J = 8.0, 2.4 Hz, 1 H, Ar), 6.23 (d, J = 12.4 Hz, 1 H, benzylic),
3.74 (s, 3 H, OCH3), 0.87 (d, JP, H = 17.6 Hz, 9 H, tBu) ppm. 13C NMR
(CDCl3): δ = 159.7, 159.5, 149.2, 142.3, 140.2, 136.6, 129.7, 129.68,
129.6, 128.9, 126.4, 123.5, 122.5, 122.2, 115.9, 113.6 (17 C, 2 Ph, py),
73.3 (1 C, benzylic), 49.5 (s, OMe), 33.7 (d, JP, C = 88.0 Hz, 1 C, tBu),
24.6 (3 C, tBu) ppm. 31P NMR (CDCl3): δ = 38.8 (d, JP, H = 512 Hz) ppm.
HRMS (EI): calcd. for C23H27N2O2P [M+] 394.1810; found 394.1819.
C23H27N2O2P (394.45): calcd. C 70.03, H 6.90, N 7.10; found C 69.50,
(0.36 mL, 3.0 mmol) or tBuPCl2 (0.48 g, 3.0 mmol) dissolved in THF H 6.70, N 6.66.
(2 mL) was prepared for addition to the flask. The solution was
cooled to –78 °C again before PhPCl2 or tBuPCl2 in THF was slowly
Spectroscopic Data for 3e: 1H NMR (CDCl3): δ = 7.34–7.28 (m, 2 H,
Ar), 7.20 (t, J = 4.8 Hz, 1 H, Ar), 6.86 (d, J = 7.6 Hz, 1 H, Ar), 6.83 (dd,
JP, H = 447 Hz, 1 H, P-H), 6.81 (s, 1 H, Ar), 6.77 (dd, J = 5.4, 3.2 Hz, 1
H, Ar), 6.53 (d, J = 5.4 Hz, 2 H, Ar), 4.61 (d, J = 6.4 Hz, 1 H, Ar), 4.04
(d, J = 6.0 Hz, 1 H, benzylic), 3.77 (s, 3 H, OMe), 1.08 (d, JP, H = 16.8 Hz,
9 H, tBu), 1.00 (s, 9 H, tBu) ppm. 13C NMR (CDCl3): δ = 159.2, 150.9,
141.9, 132.2, 128.8, 121.0, 114.6, 112.7, 111.7 (12 C, 2 Ph), 66.7 (1 C,
benzylic), 55.1 (1 C, OMe), 34.9 (1 C, tBu), 32.1 (d, JP, C = 69.9 Hz, 1
C, tBu), 27.0 (3 C, tBu), 23.5 (3 C, tBu) ppm. 31P NMR (CDCl3): δ =
48.9 (d, JP, H = 447 Hz) ppm. HRMS (EI): calcd. for C22H32NO2P [M+]
373.2171; found 373.2178. C22H32NO2P (373.47): calcd. C 70.75, H
8.64, N 3.75; found C 68.8, H 8.70, N 3.09.
added. After stirring for an additional 1 h, presumably compound
2a (or 2e) was formed. Without further purification, 3
N HCl was
added to the solution at 0 °C, which was stirred for another 1 h.
The organic layer was extracted twice with ethyl acetate. Then, it
was dried with anhydrous MgSO4. A dark-yellow solid 3a (0.16 g,
0.45 mmol, 15 %) was obtained by column chromatography. The
yields of 3b and 3c were 25 % (0.26 g, 0.75 mmol) and 33 % (0.36 g,
0.99 mmol), respectively. A similar procedure was carried out for the
preparation of 3e except for starting with 3-methoxybenzaldehyde
rather than benzaldehyde. The yield of 3e was 21 % (0.24 g,
0.63 mmol). As for the preparation of 3f, the nucleophile was the
deprotonated 1-methyl-1H-imidazole (by nBuLi), and the yield of 3f
was 17 % (0.20 g, 0.51 mmol). For the preparation of 3d, the nucleo-
phile was the deprotonated 2-bromopyridine, and the yield was
only 3 % (0.035 g, 0.09 mmol). Similar procedures might be carried
out for the syntheses of 3g and 3h except for using deprotonated
imidazole and thiazole as nucleophiles. The yields of 3g and 3h
were 11 % (0.116 g, 0.316 mmol) and 18 % (0.660 g, 1.78 mmol),
respectively.
Spectroscopic Data for 3f: 1H NMR (CDCl3): δ = 7.50 (d, J = 7.6 Hz,
2 H, Ar), 7.13 (m, 2 H, Ar), 7.05 (m, 3 H, Ar), 6.85 (s, 1 H, Ar), 6.79 (s,
1 H, Ar), 6.78 (s, 1 H, Ar), 6.77 (d, JP, H = 502 Hz, 1 H, P-H), 6.67 (dd,
J = 5.6, 4.8 Hz, 1 H, Ar), 6.45 (d, J = 9.2 Hz, 1 H, benzylic), 3.66 (s, 3
H, OMe), 3.45 (s, 3 H, NMe), 0.94 (d, JP, H = 18.0 Hz, 9 H, tBu) ppm.
13C NMR (CDCl3): δ = 159.3, 146.3, 141.0, 138.1, 131.3, 129.0, 128.5,
127.2, 126.5, 121.8, 121.1, 114.7, 113.8 (15 C, 2 Ph, imidazole), 57.1
(1 C, benzylic), 55.1 (1 C, -OMe), 34.4 (d, JP, C = 85.9 Hz, 1 C, tBu),
32.9 (1 C, NMe), 23.6 (3 C, tBu) ppm. 31P NMR (CDCl3): δ = 49.3 (d,
Spectroscopic Data for 3a: 1H NMR (CDCl3): δ = 7.91 (d, JP, H
=
J
P, H = 502 Hz) ppm. HRMS (EI): calcd. for C22H28N3O2P [M+] 397.1910;
474 Hz, 1 H), 7.66–7.61 (m, 1 H), 7.46–7.43 (m, 2 H), 7.34–7.24 (m,
10 H), 6.53–6.51 (dd, J = 6.4, 2.0 Hz, 2 H), 4.07 (d, JP, H = 6.80 Hz, 1
H, P-H), 0.93 (s, 9 H) ppm. 13C NMR (CDCl3): δ = 150.9, 139.9, 132.1,
found 397.1925. C22H28N3O2P (397.46): calcd. C 66.48, H 7.10, N
10.57; found C 67.13, H 7.35, N 10.59.
Spectroscopic Data for 3g: 1H NMR (CDCl3): δ = 7.47 (d, J = 8.0 Hz,
2 H), 7.26–7.25 (m, 2 H), 7.19–7.11 (m, 5 H), 7.08–7.05 (m, 2 H), 6.82
(s, 1 H), 6.79 (d, JP, H = 500.1 Hz, 1 H, P-H), 6.51 (d, J = 9.2 Hz, 1 H,
benzylic), 3.48 (s, 3 H, NCH3), 0.95 (d, J = 17.6 Hz, 9 H, tBu) ppm.
13C NMR (CDCl3): δ = 146.3 (d, J = 3.1 Hz), 140.8 (d, J = 2.7 Hz),
136.4 (d, J = 1.9 Hz), 131.3 (d, J = 4.6 Hz), 129.5, 128.4, 128.0, 127.6,
132.0, 130.8, 130.7, 128.7, 128.6, 128.3, 127.9, 112.9 (d, JP, C
=
113.3 Hz, 1 C, Ph), 112.8 (17 C, Ph), 66.7 (1 C), 34.8 (1 C, tBu), 26.9
(3 C, tBu) ppm. 31P NMR (CDCl3): δ = 22.4 (d, JP, H = 474 Hz) ppm.
HRMS (EI) calcd. for C23H26NOP [M+]: 363.1752; found: 363.1750.
Elemental analysis calcd. (%) for C23H26NOP (363.44): C 76.01, H
7.21, N 3.85; found: C 75.66, H 6.64, N 3.55.
127.1, 126.4, 121.1, 56.6 (d, J = 4.5 Hz, benzylic), 34.4 [d, JP, C
=
Spectroscopic Data for 3b: 1H NMR (CDCl3): δ = 7.33–7.23 (m, 8
H), 6.83 (d, JP, H = 447 Hz, 1 H, P-H), 6.54–6.52 (dd, J = 6.4, 2.4 Hz, 2
H), 4.08 (s, 1 H), 1.18 (d, JP, H = 17.99 Hz, 9 H), 0.96 (s, 9 H) ppm. 13C
NMR (CDCl3): δ = 150.9, 140.1, 132.2, 128.3, 127.8, 127.1, 114.0 115.0
(12 C, Ph), 112.6, 66.7 (1 C), 34.8 (1 C, tBu), 32.1 (d, JP, C = 71.4 Hz, 1
C, tBu), 26.9 (3 C, tBu), 23.4 (3 C, tBu) ppm. 31P NMR (CDCl3): δ =
48.4 (d, JP, H = 447 Hz) ppm. HRMS (EI): calcd. for C21H30NOP [M+]
343.2065; found 343.2060. C21H30NOP (343.45): calcd.C 73.44, H
8.80, N 4.08; found C 72.70, H 8.55, N 3.68.
85.4 Hz, C(CH3)3], 32.8 (1 C, NCH3), 23.5 (d, J = 1.6 Hz, CH3) ppm.
31P NMR (CDCl3): δ = 49.5 (d, JP, H = 511.3 Hz) ppm. HRMS (EI): calcd.
for C21H26N3OP [M+] 367.1813; found 367.1806. IR (ATR): ν = 2330
˜
(br., P–H), 1489 (P=O) cm–1. C21H26N3OP (367.43): calcd. C 68.65, H
7.13, N 11.44; found C 66.66, H 6.45, N 9.69.
Spectroscopic Data for 3h: 1H NMR (CDCl3): δ = 7.85 (d, J = 3.2 Hz,
1 H), 7.36 (t, J = 1.6 Hz, 1 H), 7.33–7.31 (m, 2 H), 7.25–7.23 (m, 5 H),
7.19–7.17 (m, 3 H), 6.74 (d, JP, H = 502.9 Hz, 1 H, P-H), 6.62 (d, J =
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
9
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
10.0 Hz, 1 H, benzylic), 1.05 (d, J = 17.6 Hz, 9 H, tBu) ppm. 13C NMR
(CDCl3): δ = 170.0, 141.9, 140.4, 137.1, 130.6, 129.1, 128.2, 127.6,
127.5, 126.4, 119.6, 63.5 (1 C, benzylic), 33.8 [d, JP, C = 86.6 Hz,
δ = 109.4 ppm. HRMS (ESI–): calcd. for C20H23Cl2N2OPPdS [M – H+]
545.9681; found 544.9604.
Spectroscopic Data for 8f: 1H NMR (CDCl3): δ = 7.80 (s, 1 H, Ar),
7.51 (d, J = 8.0 Hz, 2 H, Ar), 7.32 (t, J = 7.6 Hz, 2 H, Ar), 7.19–7.11
(m, 2 H, Ar), 6.88 (s, 1 H, Ar), 6.86 (s, 1 H, Ar), 6.77 (d, J = 8.0 Hz, 1
H, Ar), 6.70 (s, 1 H, Ar), 6.01 (d, J = 18.8 Hz, 1 H, benzylic), 3.72 (s, 3
C(CH3)3], 23.3 (1 C, CH3) ppm. 31P NMR (CDCl3): δ = 47.8 (d, JP, H
=
496.2 Hz) ppm. HRMS (EI): calcd. for C20H23N2OPS [M+] 370.1269;
found 370.1264. IR (ATR): ν = 2366 (br., P–H), 1491 (P=O) cm–1
.
˜
C20H23N2OPS (370.45): calcd. C 64.84, H 6.26, N 7.56; found C 64.36,
H 5.93, N 7.23.
3
H, OMe), 3.68 (s, 3 H, NCH3), 0.85 (d, JP, H = 16.0 Hz, 9 H, tBu), 0.83
3
(d, JP, H = 16.0 Hz, 9 H, tBu) ppm. 13C NMR (CDCl3): δ = 159.7 (s),
Reactions of 3 with Palladium Salts: Under nitrogen, a 20 mL
Schlenk tube with magnetic stirrer was charged with THF (2.0 mL),
Pd(COD)Cl2 (0.029 g, 0.10 mmol), and 2 mol-equiv. of 3f (0.079 g,
0.20 mmol). The solution was stirred at 60 °C for 16 h before it
was cooled to room temperature. Subsequently, the solvent was
completely removed under reduced pressure. Then, the residue was
dissolved in CH2Cl2 (1.0 mL) and purified by column chromatogra-
phy. The target compound 6f was obtained in 92.2 % yield (0.053 g,
0.092 mmol). Crystals of 6f with suitable size for X-ray diffraction
were obtained after placing purified 6f under a two-layered solu-
tion (diethyl ether/CH2Cl2) at 4 °C for a few days. The identity of 6f
was determined by spectroscopic means as well as by X-ray diffrac-
tion methods. Similar procedures for the preparation of 6g and 6h
were carried out. The yields of 6g (pale-yellow solid) and 6h (bright-
yellow solid) were 62 % (0.0336 g, 0.0620 mmol) and 95 % (0.0520 g,
0.0950 mmol), respectively. Crystals of 7fa and 7fb were picked out
from the results of the crystal-growing process. Since there were
no sufficient amounts of compounds 7fa and 7fb left for further
examination, the spectroscopic analyses of these two compounds
are not available. A similar procedure was carried out for the reac-
tion of 3f (0.079 g, 0.20 mmol) with Pd(COD)Cl2 (0.029 g, 0.10 mmol)
in 1,4-dioxane at 100 °C for 6 h. The resulting product 8f was further
purified by column chromatography, initially using a mixed solvent
(hexanes/ethyl acetate, 1:1) and later pure ethyl acetate. The yield
of 8f was 36.8 % (0.046 g, 0.037 mmol). Suitable crystals for X-ray
crystallographic studies were obtained from a two-layered solution
(diethyl ether/CH2Cl2).
149.0 (s), 146.9 (s), 142.9 (s), 129.8 (s), 129.3 (s), 129.1 (s), 128.7 (s),
124.9 (s), 121.0 (s), 119.8 (s), 114.5 (s), 112.2 (s), 63.9 (d, 2JP, C = 8.9 Hz,
1 C, benzylic), 55.1 (s, 1 C, OMe), 42.4 (d, JP, C = 50.4 Hz, 1 C, tBu),
2
40.4 (d, JP, C = 54.2 Hz, 1 C, tBu), 33.3 (s, 1 C, NCH3), 27.1 (d, JP, C
=
5.3 Hz, 3 C, tBu), 25.6 (d, JP, C = 5.3 Hz, 3 C, tBu) ppm. 31P NMR
(CDCl3): δ = 106.1 (d, JP, P = 13.8 Hz), 121.9 (d, JP, P = 11.5 Hz) ppm.
HRMS (ESI+): calcd. for C52H74N6O8P4Pd2 [M + H+] 1246.2588; found
1247.2621. C52H74N6O8P4Pd2·C4H10O (1320.33): calcd. C 50.88, H
6.40, N 6.36; found C 51.81, H 6.38, N 5.10.
2
General Procedure for Catellani Reactions: A nitrogen-filled
20 mL brown Schlenk tube charged with a magnetic stirrer was
filled with N-(2-bromophenyl)acetamide (0.4 mmol, 0.086 g),
Pd(OAc)2 (0.02 mmol, 5.0 mg), 3h (0.022 mmol, 8.14 mg), K2CO3
(0.8 mmol, 0.111 g), and norbornene (0.42 mmol, 0.040 g). This was
processed in a fume hood, and then the tube was capped with a
tight seal. The air was pumped out of the tube and replaced by
nitrogen. This process was repeated three times. Subsequently, 2-
iodotoluene (0.2 mmol, 0.0254 mL) and DMA (2.0 mL) were added
under nitrogen. The solution was then heated in an oil bath at
135 °C for 20 h. After that, it was cooled to room temperature, and
the resulting products were extracted with water and diethyl ether.
The extracted solution was filtered through a flash column and con-
centrated. Further purification was carried out by centrifugal thin
layer chromatography. Finally, the conversion rates of the particular
products were analyzed by their distinct chemical shifts using 1H
NMR spectroscopy. Similar procedures were performed when using
different ligands such as 3g, 6h, PPh3, and dppe.
Spectroscopic Data for 6f: 1H NMR (CDCl3): δ = 7.66 (d, J = 2.0 Hz,
1 H, Ar), 7.62 (t, J = 2.4 Hz, 1 H, Ar), 7.42–7.39 (m, 2 H, Ar), 7.35–
7.31 (m, 3 H, Ar), 7.23 (t, J = 8.0 Hz, 1 H, Ar), 7.02 (d, J = 1.6 Hz, 1
H, Ar), 6.89 (dd, J = 7.6, 2.0 Hz, 1 H, Ar), 6.78 (dd, J = 8.4, 0.8 Hz, 1
H, Ar), 5.86 (d, J = 24.0 Hz, 1 H, benzylic), 3.89 (s, 3 H, OMe), 3.81
General Procedure for Suzuki–Miyaura Cross-Coupling Reac-
tions: Suzuki–Miyaura cross-coupling reactions were performed ac-
cording to the following procedures. The four reactants, Pd(OAc)2,
ligand, boronic acid, and base were placed in a suitable, oven-dried
Schlenk flask. It was evacuated for 0.5 h and backfilled with nitro-
gen gas before adding solvent and aryl halide through a rubber
septum. The aryl halides, being solids at room temperature, were
added prior to the evacuation/backfill cycle. The flask was sealed
with a rubber septum, and the solution was stirred at the required
temperature for the designated time. Then, the reaction mixture
was diluted with ethyl acetate (3 mL) and the cooled solution
poured into a separating funnel. The mixture was washed with
3
(s, 3 H, NCH3), 0.91 (d, JP, H = 18.8 Hz, 9 H, tBu) ppm. 13C NMR
(CDCl3): δ = 160.4 (s), 145.7 (s), 144.3 (s), 138.2 (s), 130.1 (s), 130.0
(s), 128.8 (s), 127.9 (s), 121.6 (s), 119.6 (s), 116.1 (s), 111.8 (s), 67.1
2
(d, JP, C = 9.95 Hz, 1 C, benzylic), 56.1 (s, 1 C, OMe), 40.6 (d, JP, C
=
2
52.7 Hz, 1 C, tBu), 34.4 (s, 1 C, NCH3), 26.0 (d, JP, C = 5.03 Hz, 3 C,
tBu) ppm. 31P NMR (CDCl3): δ = 112.9 ppm. HRMS (ESI–): calcd. for
C22H28Cl2N3O2PPd [M – H+] 573.0331; found 572.0265.
Spectroscopic Data for 6g: 1H NMR (CDCl3): δ = 7.73 (d, J = 1.6 Hz,
1 H), 7.50–7.48 (m, 2 H), 7.45–7.39 (m, 4 H), 7.38–7.32 (m, 4 H), 7.00
(d, J = 1.6 Hz, 1 H), 5.94 (d, J = 24.0 Hz, 1 H, benzylic), 3.82 (s, 3 H,
NCH3), 0.91 (d, JP, H = 18.8 Hz, 9 H, tBu) ppm. 13C NMR spectroscopic
data for 6g are not available owing to its poor solubility in most
aqueous NaOH (1.0 M, 5 mL), and the aqueous layer was extracted
with ethyl acetate (2 × 5 mL). The combined organic layers were
washed with brine and dried with anhydrous magnesium sulfate.
The dried organic layer was concentrated in vacuo. The residue was
purified by column chromatography to give the desired product.
commonly used deuterated solvents. 31P NMR (CDCl3):
δ =
112.4 ppm. HRMS (ESI–): calcd. for C21H26Cl2N3OPPd [M – H+]
543.0225; found 542.0128.
X-ray Crystallographic Studies: Suitable crystals of 2x, 3d, 3h, 6f,
6g, 6h, 7fa, 7fb, and 8f were sealed in thin-walled glass capillaries
under nitrogen and mounted on a Bruker AXS SMART 1000 diffrac-
tometer. Intensity data were collected in 1350 frames with increas-
ing ω (width of 0.3° per frame). The absorption correction was
based on symmetry-equivalent reflections by using the SADABS
program. The space-group determination was based on a check of
the Laue symmetry and systematic absences, and was confirmed
by using the structure solution. The structure was solved by direct
Spectroscopic Data for 6h: 1H NMR (CDCl3): δ = 8.57 (d, J = 4.0 Hz,
1 H), 7.61 (d, J = 3.6 Hz, 1 H), 7.49–7.40 (m, 9 H), 7.34 (t, J = 7.2 Hz,
1 H), 6.27 (d, J = 24.8 Hz, 1 H, benzylic), 0.84 (d, JP, H = 18.8 Hz, 9 H,
tBu) ppm. 13C NMR (CDCl3): δ = 169.3 (d, J = 4.5 Hz), 146.1, 145.1
(d, J = 6.3 Hz), 137.9 (d, J = 2.8 Hz), 130.2, 129.2 (d, J = 4.5 Hz),
127.7, 127.6, 127.2, 71.5 (d, J = 12.9 Hz, benzylic), 40.1 [d, J =
53.2 Hz, C(CH3)3], 25.9 (d, J = 4.6 Hz, CH3) ppm. 31P NMR (CDCl3):
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
10
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1732–1734; e) W. B. Farnham, R. A. Lewis, R. K. Murray, K. Mislow, J. Am.
Chem. Soc. 1970, 92, 5808–5809; f) T. L. Emmick, R. L. Letsinger, J. Am.
Chem. Soc. 1968, 90, 3469–3465; g) R. H. Williams, A. Hamilton, J. Am.
Chem. Soc. 1955, 77, 3411–3412; h) R. H. Williams, A. Hamilton, J. Am.
Chem. Soc. 1952, 74, 5418–5420; i) G. M. Kosolapoff, R. M. Watson, J. Am.
Chem. Soc. 1951, 73, 4101–4102.
a) D.-F. Hu, C.-M. Weng, F.-E. Hong, Organometallics 2011, 30, 1139–1147;
b) L.-Y. Jung, S.-H. Tsai, F.-E. Hong, Organometallics 2009, 28, 6044–6053;
c) C.-H. Wei, C.-E. Wu, Y.-L. Huang, R. G. Kultyshev, F.-E. Hong, Chem. Eur.
J. 2007, 13, 1583–1593.
methods using a SHELXTL package. All non-H atoms were located
from successive Fourier maps, and hydrogen atoms were refined by
using a riding model. Anisotropic thermal parameters were used for
all non-H atoms, and fixed isotropic parameters were used for H
atoms. CCDC 956234 (for 2x), 956235 (for 3d), 1401620 (for 3h),
1401712 (for 6f), 1401621 (for 6g), 1401622 (for 6h), 956238 (for
7fa), 1401713 (for 7fb), and 958952 (for 8f) contain the supplemen-
tary crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre.
[6]
[7]
[8]
a) M. Liniger, B. Gschwend, M. Neuburger, S. Schaffner, A. Pfaltz, Organo-
metallics 2010, 29, 5953–5958; b) G. Helmchen, A. Pfaltz, Acc. Chem. Res.
2000, 33, 336–345.
a) L. Ackermann, A. R. Kapdi, S. Fenner, C. Korhaaꢀ, C. Schulzke, Chem.
Eur. J. 2011, 17, 2965–2971; b) L. Ackermann, R. Born, Angew. Chem. Int.
Ed. 2005, 44, 2444–2447; Angew. Chem. 2005, 117, 2497–2500.
P. A. Donets, N. Cramer, J. Am. Chem. Soc. 2013, 135, 11772–11775.
a) R. Ghosh, N. N. Adarsh, A. Sarkar, J. Org. Chem. 2010, 75, 5320–5322;
b) L. Ackermann, H. K. Potukuchi, A. Althammer, R. Born, P. Mayer, Org.
Lett. 2010, 12, 1004–1007; c) C. M. So, C. P. Lau, F. Y. Kwong, Org. Lett.
2007, 9, 2795–2798; d) D. Benito-Garagorri, E. Becker, J. Wiedermann, W.
Lackner, M. Pollak, K. Mereiter, J. Kisala, K. Kirchner, Organometallics
2006, 25, 1900–1913.
Acknowledgments
[9]
[10]
We thank the Ministry of Science and Technology of the ROC
(grant MOST 104-2119-M-005-004) for financial support.
Keywords: Secondary phosphine oxides · Ligand design ·
Catellani reactions · C–H activation · Palladium · P ligands
[11]
a) Q. Liang, P. Xing, Z. Huang, J. Dong, K. B. Sharpless, X. Li, B. Jiang, Org.
Lett. 2015, 17, 1942–1945; b) J. Magano, S. Monfette, ACS Catal. 2015, 5,
3120–3123; c) C. Allolio, T. Strassner, J. Org. Chem. 2014, 79, 12096–
12105; d) N. Oger, M. d'Halluin, E. Le Grognec, F.-X. Felpin, Org. Process
Res. Dev. 2014, 18, 1786–1801; e) N. Miyaura, A. Suzuki, Chem. Rev. 1995,
95, 2457–2483.
a) M. Catellani, G. P. Chiusoli, A. Mari, J. Organomet. Chem. 1984, 275,
129–138; b) M. Catellani, G. P. Chiusoli, C. Peloso, Tetrahedron Lett. 1983,
24, 813–816.
a) Y. Cheng, C. Bolm, Angew. Chem. Int. Ed. 2015, 54, 12349–12352;
Angew. Chem. 2015, 127, 12526–12529; b) J. Li, L. Ackermann, Angew.
Chem. Int. Ed. 2015, 54, 8551–8554; Angew. Chem. 2015, 127, 8671–8674;
c) Y. Gao, Y.-B. Huang, W.-Q. Wu, K.-F. Huang, H.-F. Jiang, Chem. Commun.
2014, 50, 8370–8373; d) D. I. Chai, P. Thansandote, M. Lautens, Chem.
Eur. J. 2011, 17, 8175–8188; e) X. Chen, K. M. Engle, D.-H. Wang, J.-Q. Yu,
Angew. Chem. Int. Ed. 2009, 48, 5094–5115; Angew. Chem. 2009, 121,
5196–5217; f) R. G. Bergman, Nature 2007, 446, 391–393; g) K. Godula,
D. Sames, Science 2006, 312, 67–72; h) J. A. Labinger, J. E. Bercaw, Nature
2002, 417, 507–514; i) C. Jia, T. Kitamura, Y. Fujiwara, Acc. Chem. Res.
2001, 34, 633–639; j) A. E. Shilov, G. B. Shul'pin, Chem. Rev. 1997, 97,
2879–2932.
a) Catellani-type Suzuki reactions: M. Catellani, E. Motti, M. Minari, Chem.
Commun. 2000, 157–158; b) Catellani-type Heck reactions: M. Catellani,
F. Frignani, A. Rangoni, Angew. Chem. Int. Ed. Engl. 1997, 36, 119–122;
Angew. Chem. 1997, 109, 142–145.
a) E. Motti, F. Faccini, I. Ferrari, M. Catellani, R. Ferraccioli, Org. Lett. 2006,
8, 3967–3970; b) M. Catellani, S. Deledda, B. Ganchegui, F. Hénin, E. Motti,
J. Muzart, J. Organomet. Chem. 2003, 687, 473–482; c) M. Catellani, E.
Motti, B. Simone, Org. Lett. 2001, 3, 3611–3614; d) M. Catellani, E. Motti,
L. Paterlini, J. Organomet. Chem. 2000, 593–594, 240–244; e) M. Catellani,
F. Cugini, Tetrahedron 1999, 55, 6595–6602.
a) Z. Qureshi, H. Weinstabl, M. Suhartono, H. Liu, P. Thesmar, M. Lautens,
Eur. J. Org. Chem. 2014, 4053–4069; b) A. Rudolph, N. Rackelmann, M.
Lautens, Angew. Chem. Int. Ed. 2007, 46, 1485–1488; Angew. Chem. 2007,
119, 1507–1510.
a) E. Motti, N. D. Cá, D. Xua, S. Armani, B. M. Aresta, M. Catellani, Tetrahe-
dron 2013, 69, 4421–4428; b) D. A. Candito, M. Lautens, Angew. Chem.
Int. Ed. 2009, 48, 6713–6716; Angew. Chem. 2009, 121, 6841–6844; c) C.
Blaszykowski, E. Aktoudianakis, D. Alberico, C. Bressy, D. G. Hulcoop, F.
Jafarpour, A. Joushaghani, B. Laleu, M. Lautens, J. Org. Chem. 2008, 73,
1888–1897; d) P. Thansandote, D. G. Hulcoop, M. Langer, M. Lautens, J.
Org. Chem. 2009, 74, 1673–1678; e) D. I. Chai, M. Lautens, J. Org. Chem.
2009, 74, 3054–3061; f) D. A. Candito, M. Lautens, Org. Lett. 2010, 12,
3312–3315.
[1] a) J. F. Hartwig, Organotransition Metal Chemistry, from Bonding to Cataly-
sis, University Science Books, New York, 2010; b) R. H. Crabtree, The Orga-
nometallic Chemistry of the Transition Metals, 5th ed., John Wiley & Sons
Inc., Hoboken, 2009, chapter 4; c) W. Tang, X. Zhang, Chem. Rev. 2003,
103, 3029–3070; d) J. Hassan, M. Sévignon, C. Gozzi, E. Schulz, M. Lemaire,
Chem. Rev. 2002, 102, 1359–1470; e) N. G. Andersen, B. A. Keay, Chem.
Rev. 2001, 101, 997–1030; f) C. A. Bessel, P. Aggarwal, A. C. Marschilok,
K. J. Takeuchi, Chem. Rev. 2001, 101, 1031–1066.
[12]
[13]
[2] S. T. Diver, in Encyclopedia of Reagents for Organic Synthesis (Eds.: L. A.
Paquette), Wiley, New York, 1995, vol. 7, pp. 5014–5016.
[3] R. G. Parr, R. G. Pearson, J. Am. Chem. Soc. 1983, 105, 7512–7516.
[4] a) P. Sutram, A. Igau, Coord. Chem. Rev. 2016, 97–116; b) H. Landert, F.
Spindler, A. Wyss, H.-U. Blaser, B. Pugin, Y. Ribourduoille, B. Gschwend, B.
Ramalingam, A. Pfaltz, Angew. Chem. Int. Ed. 2010, 49, 6873–6876;
Angew. Chem. 2010, 122, 7025–7028; c) L. Ackermann, H. K. Potukuchi,
A. R. Kapdi, C. Schulzke, Chem. Eur. J. 2010, 16, 3300–3303; d) A. J. Bloom-
field, J. M. Qian, S. B. Herzon, Organometallics 2010, 29, 4193–4195; e) T.
Achard, L. Giordano, A. Tenaglia, Y. Gimbert, G. Buono, Organometallics
2010, 29, 3936–3950; f) C. Li, A. Christiansen, M. Garland, D. Selent, R.
Ludwig, A. Spannenberg, W. Baumann, R. Franke, A. Börner, Eur. J. Org.
Chem. 2010, 2733–2741; g) L. Ackermann, R. Vicente, N. Hofmann, Org.
Lett. 2009, 11, 4274–4276; h) D. X. Yang, S. L. Colletti, K. Wu, M. Song,
G. Y. Li, H. C. Shen, Org. Lett. 2009, 11, 381–384; i) H. Xu, K. Ekouekovi,
C. Wolf, J. Org. Chem. 2008, 73, 7638–7650; j) K. L. Billingsley, S. L. Buch-
wald, Angew. Chem. Int. Ed. 2008, 47, 4695–4698; Angew. Chem. 2008,
120, 4773–4776; k) L. Ackermann, Synlett 2007, 4, 507–526; l) J. Bigeault,
L. Giordano, I. de Riggi, Y. Gimbert, G. Buono, Org. Lett. 2007, 9, 3567–
3570; m) R. Lerebours, C. Wolf, J. Am. Chem. Soc. 2006, 128, 13052–13053;
n) B. Hoge, P. Garcia, H. Willner, H. Oberhammer, Chem. Eur. J. 2006, 12,
3567–3574; o) J. Bigeault, L. Giordano, G. Buono, Angew. Chem. Int. Ed.
2005, 44, 4753–4757; Angew. Chem. 2005, 117, 4831–4835; p) C. Wolf,
R. Lerebours, Org. Lett. 2004, 6, 1147–1150; q) G. Y. Li, W. J. Marshall,
Organometallics 2002, 21, 590–591; r) G. Y. Li, J. Org. Chem. 2002, 67,
3643–3650; s) T. Appleby, J. D. Woollins, Coord. Chem. Rev. 2002, 235,
121–140; t) G. Y. Li, US Patent 6124462, 2000; WO01040147, 2001;
WO01079213, 2001; WO02000574, 2002; u) G. Y. Li, Angew. Chem. Int.
Ed. 2001, 40, 1513–1516; Angew. Chem. 2001, 113, 1561–1564; v) G. Y.
Li, G. Zheng, A. F. Noonan, J. Org. Chem. 2001, 66, 8677–8681; w) P. L.
Polavarapu, F. Wang, J. Org. Chem. 2000, 65, 7561–7565; x) D. M. Roudhill,
R. P. Sperline, W. B. Beaulieu, Coord. Chem. Rev. 1978, 26, 263–279; y) B.
Walther, Coord. Chem. Rev. 1984, 60, 67–105.
[14]
[15]
[16]
[17]
[5] a) A. J. Bloomfield, S. B. Herzon, Org. Lett. 2012, 14, 4370–4373; b) T. M.
Shaikh, C.-M. Weng, F.-E. Hong, Coord. Chem. Rev. 2012, 256, 771–803; c)
B. Kurscheid, L. Belkoura, B. Hoge, Organometallics 2012, 31, 1329–1334;
d) W. B. Beaulieu, T. B. Rauchfuss, D. M. Roundhill, Inorg. Chem. 1975, 14,
[18]
Structurally related compounds were reported; for details, see: a) Y.-Y.
Chang, F.-E. Hong, Tetrahedron 2013, 69, 2327–2335; b) L.-Y. Jung, S.-H.
Tsai, F.-E. Hong, Organometallics 2009, 28, 6044–6053; c) R. B. Bedford,
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
11
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
S. L. Hazelwood, M. E. Limmert, J. M. Brown, S. Ramdeehul, A. R. Cowley,
S. J. Coles, M. B. Hursthouse, Organometallics 2003, 22, 1364–1371; d)
A. M. Z. Slawin, M. Wainwright, J. D. Woollins, J. Chem. Soc., Dalton Trans.
2001, 2724–2730; e) A. M. Z. Slawin, M. Wainwright, J. D. Woollins, New
J. Chem. 2000, 24, 69–71; f) D. Matt, F. Ingold, F. Balegroune, D. Grand-
jean, J. Organomet. Chem. 1990, 399, 349–360.
[19] N. D. Cá, G. Sassi, M. Catellani, Adv. Synth. Catal. 2008, 350, 2179–2182.
Received: February 23, 2016
Published Online: ■
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
12
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Pd Catalysts
C.-Y. Hu, Y.-Q. Chen, G.-Y. Lin,
M.-K. Huang, Y.-C. Chang,
F.-E. Hong* ........................................ 1–13
Preparation of Secondary Phos-
phine Oxide Ligands through Nu-
cleophilic Attack on Imines and
Their Applications in Palladium-Cat-
alyzed Catellani Reactions
Several new secondary phosphine and 8f, with fascinating bonding
oxide (SPO) type ligands, 3a–3n, were modes were characterized by X-ray dif-
prepared. In addition, some SPOs as fraction methods. Some of these li-
well as coordinated palladium com- gands were chosen to assist in palla-
plexes of ligand 3f, that is, 7fa, 7fb, dium-catalyzed Catellani reactions.
DOI: 10.1002/ejic.201600181
Eur. J. Inorg. Chem. 0000, 0–0
www.eurjic.org
13
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim