Influence of electronically and sterically tunable cinnamate ligands on the spectroscopic, kinetic, and thermodynamic properties of bis(triphenylphosphine) palladium(0) olefin complexes

Article abstract of DOI:10.1021/om200849n

A detailed study of the influence of electronic and steric characteristics of cinnamic acid esters on the spectroscopic, kinetic, and thermodynamic properties of bis(triphenylphosphine)palladium(0) cinnamic acid ester complexes is presented (51 different new complexes included). These complexes show a dynamic behavior on the NMR spectroscopic time scale. Therefore, the rotational barriers of the olefin about the metalolefin bond as well as the dissociation entropy and enthalpy of the olefin and the dissociation mechanism could be determined. These findings are interpreted together with the NMR spectroscopic, IR spectroscopic, and X-ray structural data (7 new structures included) concerning the influence of the different olefin ligands on the complex properties by means of Hammett plots. DFT calculations were performed to support the mechanistic conclusions.

Full text of DOI:10.1021/om200849n

Article
pubs.acs.org/Organometallics
Influence of Electronically and Sterically Tunable Cinnamate Ligands
on the Spectroscopic, Kinetic, and Thermodynamic Properties of
Bis(triphenylphosphine)palladium(0) Olefin Complexes
§
Magnus R. Buchner,† Bettina Bechlars,‡ Bernhard Wahl, and Klaus Ruhland*,∥
‡
†WACKER-Institut fur Siliciumchemie, Lehrstuhl fur Anorganische Chemie, and §Lehrstuhl fur Anorganische Chemie mit
̈
̈
̈
Schwerpunkt Neue Materialien, Technische Universitat Munchen, Lichtenbergstraße 4, 85747 Garching bei Munchen, Germany
∥Lehrstuhl fur Chemische Physik und Materialwissenschaften, Universitat Augsburg, Universitatsstraße 1, 86135 Augsburg, Germany
̈
̈
̈
̈
̈
̈
S
* Supporting Information
ABSTRACT: A detailed study of the influence of electronic and steric
characteristics of cinnamic acid esters on the spectroscopic, kinetic, and
thermodynamic properties of bis(triphenylphosphine)palladium(0) cin-
namic acid ester complexes is presented (51 different new complexes
included). These complexes show a dynamic behavior on the NMR
spectroscopic time scale. Therefore, the rotational barriers of the olefin
about the metal−olefin bond as well as the dissociation entropy and
enthalpy of the olefin and the dissociation mechanism could be
determined. These findings are interpreted together with the NMR
spectroscopic, IR spectroscopic, and X-ray structural data (7 new
structures included) concerning the influence of the different olefin
ligands on the complex properties by means of Hammett plots. DFT
calculations were performed to support the mechanistic conclusions.
The electronic influence was assigned via the Hammett param-
eter σp, while the steric demand was determined via the A
values, which are compiled in Table 1.
INTRODUCTION
■
Zerovalent group 10 complexes are among the most applied
catalyst classes in modern synthetic chemistry. Nevertheless,
only relatively simple complexes (e.g., tetrakis(triphenyl-
phosphine)palladium) are usually used. Even though detailed
studies concerning the electronic properties of distinct ligand
systems have been performed, no examinations have been
conducted on subtle changes in the electronic and steric
properties of cinnamate ligands to date. The focus of this
examination is the correlation of directly measurable spectro-
scopic, kinetic, and thermodynamic properties with the electronics
and sterics of the ligands.
Table 1. Hammett and A2 Values
+/−
+/−
σP
σP
σI
σ°R
σ°R
A
OMe
Me
H
−0.27
−0.17
0
−0.78
−0.31
0
0.27
−0.05
0
−0.42
−0.13
0
−1.02
−0.25
0
Me
Et
1.70
1.75
2.15
4.50
iPr
tBu
Cl
0.23
0.54
0.78
0.11
0.74
1.28
0.46
0.42
0.65
−0.18
0.08
0.15
−0.36
CF3
NO2
0.6
The analyzed systems, summarized in Figure 1, were chosen
due to their broad tunability. In these complexes the electronic
RESULTS AND DISCUSSION
Synthesis. The cinnamic acid ester ligands used were
■
synthesized in case of phenol esters and the tert-butyl ester3,4
either by reaction of the appropriate acid chloride with the
alcohol in pyridine at 100 °C5,6 or by deprotonation of the
alcohol with sodium hydride and subsequent salt metathesis
with the acid chloride in nearly quantitative yield. The other
alkyl esters were obtained by refluxing the cinnamic acid in the
corresponding alcohol in the presence of catalytic amounts of
H2SO4 in quantitative yield.7
Figure 1. Investigated cinnamic acid complexes: R1, R2 = OMe, Me, H,
i
t
Cl, CF3,NO2; R3 = Me, Et, Pr, Bu.
properties of the double bond can be adjusted by the substi-
tuent in a position para1 to the olefin (R1). By variation of the
ester group a fine tuning of the electronics (R2) of the olefin
and the steric demand2 (R3) of the ligand can be performed.
Received: September 9, 2011
Published: January 10, 2012
© 2012 American Chemical Society
588
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Scheme 1. Synthetic Routes to Bis(triphenylphosphine)palladium Cinnamates (R1 = OMe, Me, H, Cl, CF3, NO2)
Figure 2. ORTEP13 representation of PdPhCl−PhNO2 (top left), PdPhNO2−PhOMe (top right), PdPhNO2−Ph (bottom left), and PdPhNO2−
PhNO2 (bottom right). Thermal ellipsoids are given at the 50% probability level. Hydrogen atoms are omitted for clarity.
The corresponding palladium complexes were all synthesized
by reaction of Pd(C3H5)Cp8 with 2 equiv of triphenylphos-
phine to generate a bis(triphenylphosphine)palladium(0) frag-
ment9,10 in situ, via reductive elimination, in analogy to the
synthesis of bis(tri-tert-butylphosphine)palladium(0).11 Subse-
quent addition of olefin gives the desired complexes in quan-
titative yield in this simple one-pot reaction (Scheme 1). The
nitro cinnamic acid complexes were isolated as deep red, air-
stable solids, while all other palladium complexes are orange to
yellow solids which decompose in air within minutes. All com-
plexes, with the exception of nitro and trifluoromethyl cinnamic
acid esters, decompose slowly, in the presence of oxygen within
seconds, in solution. The sterically less shielded methyl ester
complexes decompose within minutes in solution at room tem-
perature. However, the degradation is slow enough at −25 °C
for sufficient spectroscopic analysis. Despite the lability in
solution all complexes are stable at ambient temperature in the
solid state under an Ar atmosphere.
diffusion of pentane into a toluene solution of the appropriate
complex (Figure 2).
Table 2 compares the most important structural data of the
four complexes.
The five atoms P1/P2, C1/C2, and Pd are almost in one
plane in a trigonal geometry (olefin/P1/P2). The bond length
Pd−C1 is always shorter than the bond length Pd−C2. At the
same time, the bond length P2−Pd is always longer than the
bond length P1−Pd, in accordance with what one would
predict on the basis of trans influence. The C1−C2 double
bond is elongated through the coordination to the Pd center.
No indication is found by X-ray analysis for an interaction of
the carbonyl function with the metal center. Although the
carbonyl oxo atom is slightly directed toward the Pd center, the
Pd−O distance is not shorter than the sum of van der Waals
radii of Pd and oxygen. No simple influence of the substituents
R1 and R2 on the structure parameters can be concluded from
the X-ray data.
These structures are compared to the structures of three
chelating complexes in Figure 3 determined by us.
In particular, the P1−Pd−P2 angle can be fixed at smaller
angles in comparison to those for the bis(triphenylphosphine)
X-ray Structures. Single crystals of PdPhCl−PhNO2,
PdPhNO2−PhOMe, PdPhNO2−Ph, and PdPhNO2−PhNO2
suitable for X-ray diffraction analysis could be grown by slow
589
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Table 2. Structural Data of the Pd Complexes Discussed in the Text
C1C2
C1−Pd
(Å)
C2−Pd
(Å)
P1−Pd−P2 C1−Pd−P1 C2−Pd−P2
C3O
C2−C3
(Å)
C3−O
(Å)
name
(Å)
P1−Pd (Å) P2−Pd (Å)
(deg)
(deg)
(deg)
(Å)
PdPhNO2−PhNO2
PdPhNO2−Ph
PdPhCl−PhNO2
PdPhNO2−
PhOMe
1.411(5)
1.422(2)
1.427(3)
1.413(5)
2.125(3)
2.136(2)
2.137(2)
2.132(3)
2.157(3)
2.151(2)
2.166(2)
2.140(3)
2.3288(8)
2.3184(4)
2.3249(6)
2.3216(8)
2.3416(9)
2.3310(4)
2.3449(6)
2.3334(8)
106.82(3)
111.29(1)
107.47(2)
112.31(3)
100.7(1)
103.86(4)
99.34(6)
103.13(9)
114.48(9)
106.20(4)
114.85(7)
105.87(9)
1.208(4)
1.202(2)
1.205(3)
1.202(4)
1.448(5)
1.467(2)
1.442(3)
1.460(4)
1.404(4)
1.382(2)
1.398(3)
1.377(4)
dppePdPhNO2−iPr 1.410(6)
2.131(4)
2.148(2)
2.124(2)
2.108(4)
2.126(2)
2.128(2)
2.293(1)
2.3117(9)
2.3062(2)
2.308(1)
2.3022(8)
2.2952(5)
85.82(4)
96.37(3)
104.63(2)
117.5(1)
113.83(6)
107.32(6)
118.2(1)
110.58(6)
108.83(6)
1.220(5)
1.216(3)
1.204(3)
1.452(6)
1.469(3)
1.459(2)
1.355(5)
1.362(3)
1.381(3)
dpppPdPhNO2−Et
1.422(4)
dppbPdPhNO2−
PhOMe
1.422(4)
Figure 3. ORTEP13 representation of dppePdPhNO2−iPr (top left), dpppPdPhNO2−Et (top right), and dppbPdPhNO2−PhOMe (bottom).
Thermal ellipsoids are given at the 50% probability level. Hydrogen atoms are omitted for clarity.
complexes through the chelation and also the configuration of
the two remaining phenyl substituents at each P atom is
influenced. This causes a change in the Pd−C and Pd−P bond
lengths. The smaller the bond angle P1−Pd−P2, the shorter
the Pd−C2 bond length in comparison to Pd−C1, while the
bonds Pd−P2 and Pd−P1 are about equally long in all cases.
The chelating complexes will be important to gain more insight
into the thermal behavior of the complexes and the dissociation
mechanism of olefin ligand later in this paper.
bis(triphenylphosphine)palladium(0) complexes of this study.
No simple correlation was found between the CO stretching
frequencies and the Hammett parameters σP, σI, and σ°R or the
steric A parameter of R1, R2, and R3, either for the complexes or
for the free ligands. Despite this lack of correlation, the CO
stretching frequency shifts significantly to lower wavenumbers
by 10−44 cm−1 upon coordination to the metals in comparison
to the free ligand in all 48 cases, indicating a decrease in bond
order for the carbonyl bond on the coordination of the olefinic
double bond and pointing to the conclusion that mesomeric
structure c in Scheme 2 plays a role in the electronic
description of the complexes. No simple correlation was
IR Spectroscopy. As a diagnostic indicator for the bond-
ing situation in the Pd complexes, we measured the CO
stretching frequency of the ester carbonyl function in the 48
590
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Scheme 2
found between this shift and the Hammett parameters of
R1 or R2.
NMR Spectroscopy. All bis(triphenylphosphine)palladium
cinnamic acid ester complexes except for the 4-nitrocinnamic
1
acids show line broadening of the signals in H, 13C, and 31P
NMR spectroscopy due to the rotation of the olefin about the
metal−olefin bond axis, which will be discussed in detail below.
To get more insight into the electronic situation of the
coordinated olefin, we have evaluated the development of the
chemical shifts Δδ of protons and carbons in the Ar−C1H1
C2H2−C3O moiety relative to the free ligand as well as
the chemical shift δ of the two phosphorus atoms as a func-
tion of the substitution pattern of the cinnamic ester rests
Figure 4. Change in chemical shift ΔδH on coordination of the olefinic
proton H1 as a function of σp of R1 for different values of R2 (black)
and R3 (gray).
2
2
R1, R2, and R3. Additionally the JPP and JCP as well as the
3
3JHH and the JPH coupling constants were used as dia-
gnostic criteria to get more insight into the electronic situ-
ation of the complex.
in a and c and in both cases are dependent on the Hammett
1H NMR. The olefinic protons are both shifted to high field
parameter of R1.
in 1H NMR spectroscopy upon coordination of the olefin (ΔδH
3
Signals for both H1 and H2 show a JHH coupling to each
1
2
negative). The Hammett plots of ΔδH and ΔδH show
decisively different influences of the substituents R1, R2, and
R3 on the change in chemical shifts of the olefinic protons and
allow us to understand in greater detail the electronic situation
of the [cinnamic acid ester]Pd(PPh3)2 complex (Figures 4 and 5).
For H2 a linear correlation is found between the Hammett
parameter of R1 and the chemical shift ΔδH. The slope of the
Hammett plot ρ is very similar in all eight cases (ρ = −0.53
0.04, squared correlation coefficients between 0.96 and 0.98).
other with coupling constants of ca. 10 Hz (independent of the
substitution pattern of the olefin), which are significantly
smaller than those observed for the noncoordinated olefins
with about 16 Hz (also independent of the substitutent pattern
at the cinnamic ester). This is in accordance with the X-ray
results, which indicate a distortion of the planar geometry of
the double bond by forcing all substituents of the olefin away
from the metal. The coupling constants indicate a H−C
C−H torsion angle of ca. 145°, determined via the Karplus
equation, in accordance with the X-ray results.12 Coupling is
also observed to both phosphorus nuclei via the palladium with
coupling constants of ca. 6 Hz for the H2 proton and ca. 7 Hz
for the H1 proton. The similarity of both coupling constants is
presumably caused by the dynamic character of the complexes.
Therefore, both olefinic protons show a multiplicity of a
doublet of pseudotriplets (dpt) and not the expected doublet of
doublets of doublets (ddd). With exception of the 4-nitro-
cinnamate complexes, coupling could only be observed at low
temperatures (246 K). No significant change of these cou-
pling constants was observed as a function of the substitution
R2 and R3 only slightly influence the chemical shift ΔδH with
2
more electron withdrawing groups R2 having a stronger
influence than less electron withdrawing groups.
For H1 a decisively different behavior is found and must be
divided into two areas.
For σp values of R1 lower than about 0.05, the Hammett plot
develops similarly to the case of H2 with the slope ρ = −0.44
0.07 taking into account all eight plots. For σp values of R1
higher than about 0.05 the slope of the Hammett plot decreases
dramatically to about −0.17 0.02, becoming less dependent
on R1.
We conclude from these data that mesomeric structures b
and c in Scheme 2 (Michael-analogous interaction) play a role
in the electronic description of the complexes, in particular for
electron-releasing substituents R1 (although the main con-
tribution will be by mesomeric structure a in accordance with
the well-known Dewar−Chatt−Duncanson model). In meso-
meric structure a the ligand electronics at H1 and C1 is less
perturbated in comparison to the free ligand than in b and c
(thus, Δδ is small for high Hammett parameters at position 1).
For position 2 (H2 and C2) the electronic situations are similar
pattern of the olefinic ligand, either in the complexes or in the
3
free ligands in the case of JHH
.
13C NMR. The same trends concerning the shifts between
coordinated and free ligand of C1 and C2 are present in 13C
NMR spectroscopy. The signals arising from the olefinic carbon
nuclei are shifted upfield in comparison to those for the
noncoordinated ligand. Due to the dynamic behavior no
coupling constants could be determined at ambient temper-
ature, except for the 4-nitro and CF3 cinnamates, and in some
591
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Figure 5. Change in chemical shift ΔδH on coordination of the olefinic proton H2 as a function of σp of R1 for different values of R2 (black) and
R3 (gray).
cases the olefinic carbon signals were not observable at all at
ambient temperature. At low temperature both carbon atoms
show a coupling to the two phosphorus atoms with a cis
coupling constant of ca. 5 Hz for C2 and 3 Hz for C1 and
a trans coupling constant of ca. 18 Hz for C2 and 25 Hz
for C1. Therefore, both signals are observed as a doublet of
doublets.
counterbalanced by olefin-to-metal electron donation (meso-
1
2
meric structure a in Scheme 2), ΔδP and ΔδP are almost
constant. For Hammett parameters higher than 0.1, where
electron donation via R1 is inefficient and thus metal-to-olefin
1
2
back-donation dominates (b and c in Scheme 2), ΔδP and ΔδP
show a linear dependence on the Hammett parameter with ρ =
0.75 for P1 and ρ = 1.5 for P2 (Figures 8 and 9), because now
electron density is provided partially by the P atoms to the Pd
center rather than by the olefin ligand.
The change of chemical shift of C1 on coordination supports
the relevance of mesomeric structures b and c in Scheme 2
(Figure 6). As in the case of H1, at negative Hammett
This effect is more pronounced for P2 (trans to C1) than for
P1. Also, the nitro group in the case of P2 shows a decisive
influence in the R2 position on the chemical shift change of
P2 in contrast to P1 (Scheme 2).
1
parameters there is observed a linear decrease of ΔδC , while at
1
positive Hammett parameters ΔδC remains almost constant.
Again, and in analogy to H2, the behavior of C2 is different.
2
Here for ΔδC a linear relation to the Hammett parameters is
To determine exact coupling constants and observe all
signals, the rotation about the olefin−palladium bond had to be
frozen. Therefore, spectra of all dynamic complexes were
recorded at 246 K (in case of R1 = CF3 at 278 K). Due to
cooling, the dynamic behavior was inhibited and all signals
could be observed. At low temperature the 2JPP coupling shows
a linear dependence in a Hammett plot for R1, as shown in
Figure 10 with ρ = −13.00 0.26 (squared correlation coeffi-
cients between 0.97 and 0.99). If R at the phenol becomes
more electron-withdrawing (R2) or less sterically crowded (R3),
found throughout the complete region with ρ = −6.83 0.26
(Figure 7, squared correlation coefficients between 0.96
and 0.99).
The JCP coupling constants between the carbonyl carbon
and the two phosphorus atoms are in a range between 2 and
4 Hz with no significant dependence on the substitution
pattern of the cinamic acid ester. The size indicates that a
direct interaction of this carbon with the metal center
2
resulting in a J coupling (about 20 Hz in the trans case) is
2
not detectable.
the JPP value decreases continuously.
31P NMR. Further support for the description in Scheme 2 is
found by 31P NMR spectroscopy. Here concerning the Δδ
values (measured relative to free PPh3 as external standard)
Since the P−Pd−P bond angle remains almost constant in
2
the 48 complexes at about 110°, the behavior of the JPP
coupling constants mainly reflects the electron density at the Pd
center, which decreases with more electron-withdrawing sub-
stituents at the cinnamic acid ester ligand due to more pro-
nounced metal-to-olefin back-donation.
1
behavior complementary to H and 13C NMR spectroscopy is
observed (Figures 8 and 9). For Hammett parameters lower
than 0.1, in which case metal-to-olefin back-donation is
592
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Figure 6. Change in chemical shift ΔδC on coordination of the olefinic carbon C1 as a function of σp of R1 for different values of R2 (black) and
R3 (gray).
Figure 7. Change in chemical shift ΔδC on coordination of the olefinic carbon C2 as a function of σp of R1 for different values of R2 (black) and
R3 (gray).
Rotational Barriers for the Cinnamate Ligands. 1H NMR.
The aromatic signals of the two distinguishable PPh3 ligands
are only resolved properly below 270 K (vide supra), and
upon warming to temperatures up to 343 K, the signals broaden
and combine to only one signal in the center of the two signals at
low temperature due to rotation about the metal−olefin bond
(Figure 11; at temperatures higher than 343 K dissociation comes
into play as well, shifting the centerd signal to lower field).
593
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Figure 8. Change in chemical shift ΔδP of P1 relative to free PPh3 as a function of σp of R1.
Figure 9. Change in chemical shift ΔδP of P2 relative to free PPh3 as a function of σp of R1.
In 31P NMR spectrospcopy the same temperature-dependent
rotation can be detected (Figure 12).
The rotational barriers of the cinnamates about the olefin−
palladium axis can be determined via line shape analysis14 and
are compiled in Table 3. The rotational barrier increases with
the electron deficiency of the olefin. An overall linear trend is
observed for σp of R1 (Figure 13).
electron-releasing groups R1 while for electron-withdrawing groups
b and c must be less important to unimportant (Figure 14).
At low temperatures a distinct dependence of the rotation
rate is only found for substituents R1 with electron-releasing
properties (σp < 0), while the rotational rate for substituents
with electron-withdrawing properties shows a much less pro-
nounced dependence on the Hammett parameter σp (Scheme 2).
At 311 K an approximate linear development over the complete
range of Hammett parameters is found, approaching the reaction
constant ρ = −3.17 (Scheme 2).
Using the basic Hammett description
log k /k = σ
p
R
H
1
Dissociation Thermodynamics. H NMR. No sharp signals
we find further evidence for a decisive contribution of mesomeric
structures b and c (in which rotation is facilitated) in the case of
are obtained upon heating beyond the coalescence point of ca. 320 K.
594
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
2
Figure 10. JPP as a function of the Hammett parameter σp of R1 with different R2 (black) and R3 (gray).
1
Figure 11. H NMR spectra of PdPh−PhMe as a function of the temperature (aromatic signals of the PPh3 ligands).
This is caused by the partial dissociation of the ligand,
which leads to an additional equilibrium, causing a shift of
the olefinic signals in H NMR spectroscopy to lower field
The equilibrium constant for the observed dissociation/
coordination equilibrium in toluene depicted in Figure 15 is
given by
1
upon higher temperature and a second line-broadening effect
(Figure 15). Above ca. 360 K the compounds start to de-
compose slowly.
K = [B][C]/[A]
595
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Figure 12. 31P NMR spectra of PdPh−PhMe as a function of the temperature.
Table 3. Rotational Barriers Determined by Line Shape
Analysis
R2
EA (olefin) (kJ/mol)
PdPhOMe−PhOMe
PdPhMe−PhOMe
PdPh−PhOMe
PdPh−PhMe
PdPh−Ph
PdPh−PhNO2
PdPh−Me
0.96
0.97
0.93
0.93
0.94
0.91
0.99
0.99
0.98
0.95
0.93
0.98
49.8 5.9
58.1 5.7
54.2 8.5
57.6 6.9
57.7 8.1
66.4 12.6
60.3 3.6
61.8 3.7
60.5 5.0
63.7 8.3
68.4 13.6
70.8 5.8
PdPh−Et
PdPh−tBu
PdPhCl−PhOMe
PdPhCF3−PhOMe
PdPhNO2−PhOMe
Figure 13. Rotational barriers as a function of σp of R1, R2 = OMe.
The chemical shift of the olefin in the pure complex is δA,
and that of the noncoordinated ligand is δB at ambient tem-
perature. The value of δA was temperature-corrected by in-
dependent measurements of the undissociated complex at
lower temperatures (Figure 15; Supporting Information).
The dissociation enthalpies and entropies can then be ob-
tained from a van’t Hoff plot and are given in Table 4 together
with the quality of the linearity R2.
The dissociation enthalpy increases with electron deficiency
of the double bond, as expected, since the metal-to-olefin back-
bonding and thus the bond strength increases (Figure 16). The
dissociation entropy increases as well with increasing Hammett
parameter, as shown in Figure 17. We explain this by the fact
that at low Hammett parameters the rotation of the olefin is
less frozen than for high Hammett parameters (as supported by
As shown in the Supporting Information, this results in
2
K = (1 − X ) [A ]/(2X (1 + X ))
A
0
A
A
XA = [A]/([A]+[B] + [C])
with the average position of the signal14 in 1H NMR spectroscopy
δ = (n δ + n δ )/(n + n )
A A
B B
A
B
and thus
δ − δ
B
δ − δ
A
B
XA =
δ − δ
B
2 −
δ − δ
A
B
596
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Table 4. Dissociation Enthalpies and Entropies Extracted
from van’t Hoff Plots
R2
ΔH (kJ/mol)
ΔS (J/(mol K))
PdPhOMe−PhOMe
PdPhMe−PhOMe
PdPh−PhOMe
PdPh−PhMe
PdPh−Ph
PdPh−PhNO2
PdPh−Me
0.96
0.99
0.96
0.99
0.99
0.96
0.99
0.99
0.99
0.99
0.83
0.99
0.97
56.3 5.6
62.8 2.7
66.6 7.2
73.6 2.1
71.3 3.2
66.6 7.1
77.2 3.2
85.3 4.0
72.1 1.3
80.6 1.7
92.2 24.2
137.3 7.8
183.8 17.2
111 17
125
132 21
161
8
6
153 10
127 21
174 10
194 12
Figure 14. Plot using the basic Hammett equation concerning the
rotation about the metal−cinnamate ligand axis at different temper-
atures.
PdPh−Et
PdPh−iPr
162
189
4
5
PdPh−tBu
PdPhCl−PhOMe
PdPhCF3−PhOMe
PdPhNO2−PhOMe
dppePdPhNO2−iPr
dpppPdPhNO2−Et
dppbPdPhNO2−PhOMe
206 71
317 23
436 51
the activation energies for the rotation). Thus, for high Hammett
parameters on dissociation this rotation in addition to the
translational degrees of freedom is set free, resulting in a greater
dissociation entropy at higher Hammett parameters.
no dissociation up to 390 K
no dissociation up to 390 K
no dissociation up to 390 K
In both cases (enthalpies and entropies) the Hammett plot
supports the importance of mesomeric structures b and c in
Scheme 2 for the pathway of dissociation, if σp is used as the
fitting parameter, and this will be further supported by the DFT
calculations below. An overall linear behavior is found if for
−CF3 and −NO2 the σP− values are chosen instead of the σP
values.
to the high-field region of the spectrum (Figure 12). This is due
to the formation of an equilibrium with bis(triphenylphosphine)-
palladium, formed through dissociation of olefin. In the tem-
perature range in which no line broadening occurs, the signals in
the 31P NMR spectra are shifted to lower field with increasing
temperature (Figure 18) and the 2Jpp coupling constants increase.
When line broadening starts, the signals move toward each other
and the coupling constant decreases as expected (Figure 19).
31P NMR. The two signals in 31P NMR start to broaden at ca.
270 K and then transit through coalescence at ca. 320 K and
finally re-form a new sharp signal at ca. 330 K. However, this
signal broadens again at higher temperature and massively shifts
1
Figure 15. H NMR spectra of PdPh−PhMe as a function of the temperature (olefinic proton signals).
597
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Figure 18. 31P NMR shifts of PdPh−PhMe as a function of the
temperature.
Figure 16. Dissociation enthalpies as a function of σp of R1 (gray),
R2 = OMe. If for R1 = −CF3 and R1 = −NO2 σP− instead of σP is used,
an overall linear relation is found (black).
2
Figure 19. 31P NMR JPP values of PdPh−PhMe as a function of the
temperature.
Figure 17. Dissociation entropies as a function of σp of R1, R2 = OMe.
If for R1 = −CF3 and R1 = −NO2 σP− instead of σP is used, an overall
linear relation is found (black).
Scheme 3
Dissociation and Rotation Mechanism. On the basis of the
collected data we can draw a more detailed picture for both the
dissociation and the rotation mechanism of the cinnamate
ligand (Scheme 3).
We propose that both pathways are connected insofar as the
cinnamate ligand rotates to a position perpendicular to the P−
Pd−P plane and this is accompanied by an increase in the bond
angle P−Pd−P for both processes, although the reaction
pathways might not be identical in both cases. If the thermal
energy is large enough, from this reaction coordinate the ligand
can be dissociated. If not, the ligand only rotates about the
ligand−metal axis.
This hypothesis is supported by the fact that the complexes
with chelating bis(phosphine) ligands do not show any disso-
ciation of the cinammate ligand up to 390 K, since in this case
the two P donors cannot adopt the necessary 180° angle P−
Pd−P. Also, the rotation about the cinammate−metal axis in
the chelate complexes is decisively slower than for the
bis(triphenylphosphine) complexes, and the most restricted
dppe complex (Table 2, Figure 3) shows the smallest rotational
rates, followed by the dppp complex and the dppb complex
(Table 5).
598
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
Table 5. Rotational Rates for Several Complexes at Various
Temperatures
rotational rate (Hz)
temp dppePdPhNO2− dpppPdPhNO2- dppbPdPhNO2- PdPhNO2−
(K)
iPr
−Et
−PhOMe
PhOMe
292
301
311
321
3.7
3.9
2.3
3
2.7
3.3
3.9
2.5
3.5
4.1
9.1
3.4
14.7
DFT Calculations. To support our mechanistic proposal, we
performed DFT calculations for PdPh−Ph at the B3PW91/
LanL2DZ level. We employed the frozen-coordinate function
on P−Pd−P at different stages (lowest energy: 114.3, 120, 130,
140, 160, 180°) and calculated the lowest energy structure
connected with the frozen angle. At 150° the transition state for
the rotation of the cinnamic acid ester ligand could be located.
At 157°, according to the calculations, the transition state for
the dissociation is found (both transition states were verified by
examining the imaginary vibration mode; in the case of TSrot
the rotation of the olefin was accompanied by a bending of the
P−Pd−P angle, and in the case of TSdiss the bending of the P−
Pd−P angle was accompanied by an increased distance of the
olefin from the metal center; Figure 20).
Figure 21 shows the development of calculated enthalpies for
the frozen structures at different P−Pd−P angles and mirrors
nicely the qualitative picture given in Scheme 3. An approxi-
mate dissociation enthalpy of 43 kJ/mol (experimental 71 kJ/mol)
and rotational barrier EA of 29 kJ/mol (experimental: 58 kJ/mol)
is found.
Figure 21. Calculated enthalpies for different P−Pd−P angles (#:
transition state).
This is accompanied by a rotation of the coordinated double
bond toward 45° relative to the P−Pd−P axis, as depicted in
figure 22, and supports perfectly the proposed mechanism in
Scheme 3.
In the course of increasing the P−Pd−P angle, the C1−C2
distance decreases continuously to the value of the free
cinnamate ligand (1.358 Å; fFigure 23).
CONCLUSION
■
Examining 51 new cinnamic acid ester complexes of Pd(0), we
were able to work out a detailed picture of the electronic
interaction between the double bond of the ligand and the
Figure 20. Calculated transition states for the rotation (left) and the dissociation (right) of the olefin ligand.
599
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
1
a Bruker AMX 400 instrument. H NMR (400 MHz) and 13C NMR
(100 MHz) chemical shifts are given relative to the solvent signal for
CDCl3 (7.26 and 77.2 ppm), C6D6 (7.15 and 128.1 ppm), and C7D8
(2.09 and 20.4 ppm);41,42 19F NMR (377 MHz) used neat CFCl3 and
31P NMR (162 MHz) used 85% H3PO4 as external standards. FAB-MS
analysis was carried out on a Finnigan MAT-90 instrument in a
p-nitrobenzyl alcohol matrix under bombardment with ionized xenon.
ESI-MS measurements were conducted on a Finnigan LCQ isntru-
ment in acetonitrile. Microanalytical data were obtained in the
microanalytical laboratory of the Technische Universitat Munchen.
̈
̈
ATR FT-IR spectroscopy was done with a Thermo Scientific Nicolet
380 Smart Orbit instrument. X-ray single-crystal parameters were
obtained as follows: single crystals were stored under perfluorinated
oil, transferred into a Lindemann capillary, fixed, and sealed. Pre-
liminary examination and data collection were carried out on an area
detection system (APEX II, κ-CCD) at the window of a rotating anode
(Bruker AXS, FR591) and graphite-monochromated Mo Kα radiation
(λ = 0.710 73 Å). Raw data were corrected for Lorentz, polarization,
and (arising from the scaling procedure) latent decay and absorption
effects. The structures were solved by a combination of direct methods
and difference Fourier syntheses. All non-hydrogen atoms were refined
with anisotropic displacement parameters, whereas all hydrogen atoms
were refined with isotropic displacement parameters. Full-matrix least-
Figure 22. Calculated rotational angle of the coordinated double bond
relative to the P−Pd−P axis at different P−Pd−P angles.
2
squares refinements were carried out by minimizing P(Fo − Fc2)2
with the SHELXL-97 weighting scheme.43 The final residual electron
density maps showed no remarkable features. Neutral atom scattering
factors for all atoms and anomalous dispersion corrections for non-
hydrogen atoms were taken from ref 44.
General Procedure for the Synthesis of Phenol Cinnama-
tes. Route I. A 0.50 g amount of acid chloride and 1 equiv of alcohol
were dissolved in 2.5 mL of pyridine and heated to 100 °C for 1 h. The
warm reaction solution was poured into a mixture of 175 mL of 0.5 M
hydrochloric acid and 100 g of crushed ice and extracted three times
with diethyl ether, in the case of 4-nitrocinnamic acid with ethyl
acetate. The combined organic layers were washed twice with satu-
rated NaHCO3 solution and once with brine, dried over Na2SO4, and
filtered, and the solvent was removed under reduced pressure. The
spectroscopic properties of known substances were in accord with the
literature.
Figure 23. Calculated C1−C2 distance as a function of P−Pd−P
angle.
Route II. A 2.49 mmol amount of the alcohol was added to 1 equiv
(2.49 mmol, 59.7 mg) of NaH in 5 mL of thf at 0 °C, and the mixture
was warmed to ambient temperature. After 15 min 1 equiv (2.49 mmol)
of the appropriate acid chloride was added and stirring continued
overnight. The reaction mixture was quenched with 15 mL of water and
extracted three times with Et2O. The combined organic layers were
washed twice with saturated NaHCO3 solution and once with brine,
dried over Na2SO4, and filtered, and the solvent was removed under
reduced pressure to give the anticipated ester.
General Synthetic Procedure for the Complexes. A 30.0 mg
portion (141 μmol) of (η3-allyl)(η5-cyclopentadienyl)palladium(II)
and 74.0 mg (282 μmol) of triphenylphosphine were dissolved in
toluene and added to 141 μmol of the ligand, and the mixture was
stirred for 1 h. After removal of the solvent in vacuo, the residue was
stirred with pentane overnight. The solvent was decanted off and the
residue washed once with pentane to give the product in micro-
analytical grade.
DFT Calculations. The DFT calculations were performed using
the program suite Gaussian 03.45 All molecular geometries were fully
optimized. The frozen-coordinate function implemented in the
program suite was used for the P−Pd−P angle. The DFT method
used includes Becke’s three-parameter hybrid exchange functional in
combination with the correlation functional of Perdew and Wang
(B3PW91). Geometry calculations and frequency calculations were
performed using the valence double-ζ LANL2DZ basis set. Transition
states were proven to be one-dimensional saddle points by frequency
analysis.
metal center. The rotational kinetics of the cinnamic acid ester
as well as the dissociation thermodynamics of this ligand were
determined as a function of the Hammett parameters of the
substituents at the cinnamic acid ester. A mechanism for the
rotation and the dissociation is proposed which connects the
two processes. The mechanism could be supported by DFT
calculations.
EXPERIMENTAL SECTION
■
General Procedures. All manipulations and experiments were
performed under argon using standard Schlenk techniques and in a
glovebox filled with argon unless otherwise stated. Pentane was dried
and degassed using a two-column drying system (MBraun),15 benzene
and toluene were distilled from sodium, and pyridine was distilled
from potassium hydroxide. All solvents were stored under an argon
atmosphere. CDCl3 was used as received from Deutero GmbH;
benzene-d6 and toluene-d8 were dried and degassed by stirring over
sodium potassium alloy, purified by condensation, and stored under
argon.16 [Pd(C3H5)Cl]2 was used as received from ABCR. Tri-
phenylphosphine, cinnamic acid, phenol, 4-methoxyphenol, 4-nitro-
phenol, and p-cresol were purchased from Merck and 4-methoxycin-
namic acid, 4-methylcinnamic acid, and 4-nitrocinnamic acid from
Aldrich and used without further purification. Pd(C3H5)(C5H5),8 the
acid chlorides, and alkyl cinnamates were synthesized according to
literature procedures.7,17−40
1H NMR, 13C NMR, and 31P NMR measurements were performed
on a Bruker Avance 400 spectrometer and 19F NMR measurements on
600
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601

Organometallics
Article
(24) Novokreshchennykh, V. D.; Mochalov, S. S.; Shabarov, Y. S.
J. Org. Chem. USSR 1982, 18, 262−268.
ASSOCIATED CONTENT
* Supporting Information
■
S
(25) Zheng, J.; Shen, Y. Synth. Commun. 1994, 24 (14), 2069−2073.
Text, tables, and figures giving detailed characterization data for
all new compounds, NMR chemical shifts of the complexes,
IR ν(CO) data for the complexes and the free ligand,
calculation procedure for the dissociation equilibrium con-
stants, van’t Hoff plots for the dissociation equilibrium of the
complexes, an Arrhenius plot for the rotation of the olefin in
the complexes, and general experimental data for the X-ray
analysis. This material is available free of charge via the Internet
at http://pubs.acs.org.
́
(26) Garcıa, H.; Iborra, S.; Miranda, M. A.; Primo, J. Heterocycles
1985, 23 (8), 1983−1989.
(27) Walter, R. Ber. Dtsch. Chem. Ges. A/B 1925, 58 (10), 2303−
2310.
(28) Belozwetow. J. Gen. Chem. USSR 1966, 36, 1212.
(29) Cevasco, G.; Thea, S. J. Org. Chem. 1994, 59 (21), 6274−6278.
(30) Lohar, J. M.; Dave, J. S. Mol. Cryst. Liq. Cryst. 1983, 103, 143−
153.
(31) Nagy, O. B.; Reuliaux, V.; Bertrand, N.; Mensbrugghe, A. V. D.;
Leseul, J.; Nagy, J. B. Bull. Soc. Chim. Belg. 1985, 94, 1055−1074.
(32) Pardin, C.; Pelletier, J. N.; Lubell, W. D.; Keillor, J. W. J. Org.
Chem. 2008, 73 (15), 5766−5775.
(33) Charette, A. B.; Janes, M. K.; Lebel, H. Tetrahedron: Asymmetry
2003, 14 (7), 867−872.
(34) Bairwa, R.; Kakwani, M.; Tawari, N. R.; Lalchandani, J.; Ray, M.;
Rajan, M.; Degani, M. S. Bioorg. Med. Chem. Lett. 2010, 20 (5), 1623−
1625.
AUTHOR INFORMATION
Corresponding Author
*E-mail: klaus.ruhland@physik.uni-augsburg.de.
■
ACKNOWLEDGMENTS
■
We thank Stefan Rudel for his synthetic work within his
research practical. For financial support we thank the
WACKER Chemie AG.
(35) Okutome, T.; Kawamura, H.; Taira, S.; Nakayama, T.;
Nunomura, S.; Kurumi, M.; Sakurai, Y.; Aoyama, T.; Fujii, S. Chem.
Pharm. Bull. 1984, 32 (5), 1854−865.
(36) Masllorens, J.; Moreno-Manas, M.; Pla-Quintana, A.; Roglans,
A. Org. Lett. 2003, 5 (9), 1559−1561.
REFERENCES
■
(37) Gerig, J. T.; McLeod, R. S. Can. J. Chem. 1975, 53 (4), 513−
518.
(1) Hansch, C.; Leo, A.; Taft, R. W. Chem. Unserer Zeit 1991, 91 (2),
165−195. Shorter, J. Chem. Unserer Zeit 1985, 6, 197−208.
(2) Hirsch, J. A. In Topics in Stereochemistry; Allinger, N. L., Eliel, E.
L., Eds.; Wiley: New York, 1967; Vol. 1.
(3) Gillespie, K. M.; Sanders, C. J; O’Shaughnessy, P.; Westmoreland,
I.; Thickitt, C. P.; Scott, P. J. Org. Chem. 2002,
67 (10), 3450−3458.
(4) Wright, S. W.; Hageman, D. L.; Wright, A. S; McClure, L. D.
Tetrahedron Lett. 1997, 38 (42), 7345−7348.
(5) Letcher, R. M.; Yue, T.-Y.; Chiu, K.-F.; Kelkar, A. S.; Cheung,
K.-K. J. Chem. Soc., Perkin Trans. 1 1998, 19, 3267−3276.
(6) Mustafa, A.; Hishmat, O. H. J. Am. Chem. Soc. 1957, 79 (9),
2225−2230.
(7) Becker, H. G. O. Autorenkollektiv. Organikum, 21st ed.; Wiley-
VCH: Weinheim, Germany, 2001.
(38) Skraup, S.; Beng, E. Ber. Dtsch. Chem. Ges. A/B 1927, 60 (4),
942−950.
́ ́ ́
(39) Concellon, J. M.; Perez-Andres, J. A.; Rodríguez-Solla, H. Angew.
Chem., Int. Ed. 2000, 39 (15), 2773−2775.
(40) Parrish, J. P.; Jung, Y. C.; Shin, S. I.; Jung, K. W. J. Org. Chem.
2002, 67 (20), 7127−7130.
(41) Gottlieb, H. E.; Kotlyar, V.; Nudelman, A. J. Org. Chem. 1997,
62 (21), 7512−7515.
(42) Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.; Gottlieb, H. E.;
Nudelman, A.; Stoltz, B. M.; Bercaw, J. E.; Goldberg, K. I.
Organometallics 2010, 29 (9), 2176−2179.
(43) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64 (1), 112−
122.
(44) Fuess, H.; Hahn, T.; Wondratschek, H.; U. Muller., Shmueli, U.;
̈
(8) Tatsuno, Y.; Yoshida, T.; Otsuka, S. Inorg. Synth. 1990, 28, 343−
Prince, E.; Authier, A.; Kopsky, V.; Litvin, D. B.; Rossmann, M. G.;
Arnold, S. H. E.; McMahon, B. Complete Online Set of International
Tables for Crystallography; Springer-Verlag: Berlin, 2007; Vols. A−G.
(45) Frisch, M. J. et al. Gaussian 03, Revision B.01; Gaussian, Inc.,
Pittsburgh, PA, 2003.
346.
(9) Norton, D. M.; Mitchell, E. A.; Botros, N. R.; Jessop, P. G.; Baird,
M. C. J. Org. Chem. 2009, 74 (17), 6674−6680.
(10) Buchner, M. R.; Bechlars, B.; Ruhland, K. Manuscript in
preparation.
(11) Yoshida, T.; Otsuka, S. Inorg. Synth. 1990, 28, 113−119.
(12) Karplus, M. J. Am. Chem. Soc. 1963, 85 (18), 2870−2871.
(13) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(14) Crabtree, R. H. The Organometallic Chemistry of the Transition
Metals, 4th ed.; Wiley: New York, 2005. Complete line shape analysis
was performed using WinDNMR Version 7.1.11, which is available
from http://www.chem.wisc.edu/areas/recih/plt/windnmr.htm.
(15) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15 (5), 1518−1520.
(16) Armarego, W. L. F.; Chai, C. L. L. Purification of Laboratory
Chemicals, 5th ed.; Butterworth-Heinemann,: Oxford, U.K., 2003.
(17) Kamigata, N.; Satoh, M.; Fukushima, T. Bull. Chem. Soc. Jpn.
1990, 63 (7), 2118−2120.
(18) Tsuge, O.; Sone, K.; Urano, S.; Matsuda, K. J. Org. Chem. 1982,
47 (26), 5171−5177.
(19) Phillips, W. M.; Currie, D. J. Can. J. Chem. 1969, 47 (17), 3137−
3146.
(20) Huang, Z.-Z.; Tang, Y. J. Org. Chem. 2002, 67 (15), 5320−5326.
(21) Hashimoto, T.; Shiomi, T.; Ito, J.-i.; Nishiyama, H. Tetrahedron
2007, 63 (52), 12883−12887.
(22) Parrish, J. P.; Dueno, E. E.; Kim, S.-I.; Jung, K. W. Synth.
Commun. 2000, 30 (15), 2687−2700.
(23) Imashiro, R.; Seki, M. J. Org. Chem. 2004, 69 (12), 4216−4226.
601
dx.doi.org/10.1021/om200849n | Organometallics 2012, 31, 588−601