Chasing the agostic interaction in ligand assisted cyclometallation reactions of palladium(II)

Article abstract of DOI:10.1039/c7dt03525j

A 500 MHz NMR study of the reaction between 1-tetralone oxime and PdCl₄^2- in CD₃OD shows resonances attributable to a potential agostic intermediate prior to the formation of the insoluble cyclopalladated product which itself was characterised by X-ray crystallography. Calculated structural, spectroscopic, QTAIM, NBO and NCI analysis results obtained from density functional theory (DFT) calculations give a full description of the putative agostic intermediate [PdCl₂(1-tetralone oxime)] (1) which is shown to include a previously unrecognised π-electron density donation from the aromatic ring to the metal in close proximity to the agostic carbon atom. Changing the (N)-OH donor to (N)-OMe does not effect the magnitude of these interactions. (N)-OH and (N)-OMe acetophenone imines in which the aromatic ring has the potential to rotate show similar agostic and π-electron donation to the alicyclic ring counterparts. 1-Tetralone which coordinates to the metal by a Pd-O bond that is much weaker than the Pd-N complexes has a slightly stronger agostic component and slightly weaker π-electron donation than the oxime counterpart.

Full text of DOI:10.1039/c7dt03525j

View Article Online
View Journal
Dalton
Transactions
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: A. J. Nielson, J. A.
Harrison, A. Sajjad, P. Schwerdtfeger, N. Rees and K. Christensen, Dalton Trans., 2017, DOI:
10.1039/C7DT03525J.
This is an Accepted Manuscript, which has been through the
Volume 45 Number
1 7 January 2016 Pages 1–398
Royal Society of Chemistry peer review process and has been
accepted for publication.
Dalton
Transactions
Accepted Manuscripts are published online shortly after
An international journal of inorganic chemistry
www.rsc.org/dalton
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-9226
PAPER
Joseph T. Hupp, Omar K. Farha et al.
Efficient extraction of sulfate from water using
framework
a Zr-metal–organic
rsc.li/dalton

Please do not adjust margins
Dalton Transactions
Page 1 of 13
View Article Online
DOI: 10.1039/C7DT03525J
Journal Name
ARTICLE
Chasing the agostic interaction in ligand assisted cyclometallation
reactions of palladium (II)†
M. Arif Sajjad,a Kirsten E. Christensen,b Nicholas H. Rees,b Peter Schwerdtfeger,c John A.
Harrisona*and Alastair J. Nielsona*
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
A 500MHz NMR study of the reaction between 1-tetralone oxime and PdCl42- in CD3OD shows resonances attributable to a
potential agostic intermediate prior to the formation of the insoluble cyclopalladated product which itself was
characterised by X-ray crystallography. Calculated structural, spectroscopic, QTAIM, NBO and NCI analysis results
obtained from density functional theory (DFT) calculations give a full description of the putative agostic intermediate
[PdCl2(1-tetralone oxime)] (1) which is shown to include a previously unrecognised π-electron density donation from the
aromatic ring to the metal in close proximity to the agostic carbon atom. Changing the (N)–OH donor to (N)–OMe does
not effect the magnitude of these interactions. (N)–OH and (N)–OMe acetophenone imines in which the aromatic ring has
the potential to rotate show similar agostic and π-electron donation to the alicyclic ring counterparts. 1-Tetralone which
coordinates to the metal by a Pd–O bond that is much weaker than the Pd–N complexes has a slightly stronger agostic
component and slightly weaker π-electron donation than the oxime counterpart.
www.rsc.org/
intercepting them and investigating the main factors influencing
their formation, stability, and reactivity. Any knowledge gained
from such studies should improve understanding of the important
Introduction
Transition metal catalyzed C−H bond funcꢀonalisations are now
widely used in organic synthesis and are crucial to the development
of pharmaceuticals, natural products and materials.1 In this regard,
cyclopalladation reactions2 are very useful as site selectivity can be
controlled using pre-existing functional groups that position C−H
groups for activation.3 Up until recently these functional groups
have involved strongly coordinating ligand groups such as nitrogen
donors which lead to cyclopalladated complexes in which the
metal-carbon bond is less prone to subsequent functionalization
due to the thermodynamic stability imparted by the formation of
the cyclometallate ring system.4 However, it is now recognised that
weakly coordinating donor groups such as carboxylic acids or
ketones form less-stable cyclometallates in which the metal-carbon
bond is more reactive4 and this advance has allowed the activation
and post fuctionalisation to be made catalytic.5
steps involved in catalytic C−H functionalization processes and help
develop more efficient methodologies.6
The key activation step in these ligand-directed C–H bond
functionalizations is the weakening of the C–H bond and this is
achieved by forming an agostic interaction7 which lengthens the
bond by donation of σ-bond electron density to the metal. Removal
of the hydrogen can then be assisted by the presence of a proton
8
accepting ligand such as in the case of Pd(OAc)2 or having a base
present such as NaOAc.9 Associated with the agostic interaction is
the anagostic interaction in which no perturbation of the C-H bond
occurs.10
With the increased use of cyclometallates in the synthesis of
organic compounds, understanding the nature of the C–H bond
activation process is of prime importance. We present here, the
results of a high-field NMR study for a cyclopalladation reaction
where we search for the agostic intermediate and then carry out
DFT studies which gives a complete description of the agostic
Whereas most C−H functionalization methodologies are based on
one-pot approaches that bypass isolation or even detection of
reaction intermediates, there are potentially significant benefits in
interaction and uncovers unexpected complexity for it.
A
preliminary account of some of this work has been
communicated.11
Results and discussion
NMR spectral study of the reaction of 1-tetralone oxime and
2-
PdCl4
The 1-tetralone oxime ligand was used in the reaction with Li2PdCl4
as in earlier work on cyclopalladation reactions using this ligand as a
model system for the modification of steroidal precursors, we were
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 2 of 13
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Journal Name
anagostic interaction that was not available in the earlier work is
that the 1JC−H coupling constant is little changed from the free
ligand (1JC−H values 160.4 and 160.1 Hz respectively).
The spectral characteristics of the anagostic interaction for H8
remain the main event when the temperature is kept at 0o C but
there is slow formation of small amounts of an intermediate
identified by an upfield shift of H8 as well as a soluble
cyclopalladated complex where H8 is not present. Warming the
solution to 35o C increases the proportion of these two species at
the expense of the anagostic intermediate and the very insoluble
chloro-bridged cyclopalladated complex that is the isolated product
of the reaction, crystallises out in the NMR tube. The 13C NMR
spectrum is consistent with the 3-component system indicated by
the 1H NMR spectrum.
1
The 35o C H NMR spectrum is broad, but recooling to 0o C gives
much better peak resolution (fig. 1) and also slows the
transformation of the components. Based on the reported evidence
that agostic hydrogens which lie in or about the coordination plane
1
show an upfield shift in the H NMR spectra compared to anagostic
hydrogens which lie above the coordination plane,7 the H8-doublet
Scheme 1. Postulated intermediates identified by 1H NMR in the spectrum at δ = 7.91 might be assignable to an agostic
spectroscopy at 60 MHz in the reaction of Li2PdCl4 and 1-tetralone hydrogen. Although the resonance lies in a similar position to the
oxime in CD3OD.9
free ligand H8 resonance (δ = 7.87), the other resonances for the
ligand are not those of the free ligand (compare figs. 5 and 6 with
able to monitor by H NMR spectroscopy when anagostic, agostic fig. 1 in the ESI†) and the 13C NMR characteristics are different. This
and soluble cyclopalladated intermediates appeared to be present9 region of the H NMR spectrum corresponds to the poorly resolved
1
1
(scheme 1). At the apparent agostic stage, an intermediate of the region for the upfield shifted H8 in the earlier work.9 However, the
form [PdCl2(1-tetralone oxime)] was isolated but the complex was 1JC−H coupling constant obtained from the CLIP-HSQC13 500 MHz
unstable towards cyclopalladation even in the solid state so that no spectrum is similar to the free ligand value, and does not show the
metrics could be obtained by X-ray crystallography. In addition, this lowering found for complexes containing agostic hydrogens that are
type of complex was utilised in the study as it was expected to in fixed in-plane positions.7 This suggests that the agostic C–H bond
provide a ‘pure’ agostic interaction that did not include acetate as a if present, may move about the coordination plane to some degree
co-ligand which could influence the nature of the interaction by and that the H8 position in the spectrum could be a result of time
providing a Lewis base close by that could lengthen the C–H bond.8
averaging whereby it experiences diminishing regions of equatorial
The original investigation of the reaction between 1-tetralone current density.14
1
1
oxime and Li2PdCl4 was carried out in CD3OD by H NMR at 60 MHz Cooling the 0o solution down to -60o C does not change the JC–H
using a single scan spectrometer where the resolution was coupling constant to any meaningful extent indicating that the
1
insufficient to obtain well defined peaks.9 In the present H NMR apparent time-averaging is still present at this lower temperature.
investigation, spectra were obtained at 500 MHz in an attempt to However, the intensity of the H8-resonance at δ = 7.91 attributed
gain more detailed information about the apparent agostic
intermediate that was identified in the earlier study by an upfield
shift of the H8 proton to the vicinity of the free-ligand H8
resonance.9
1
The 500 MHz H NMR spectrum and 13C NMR spectrum obtained
at 0o C after mixing the components at 0o C showed that there was
only one major component present (see figs. 1-4 in the ESI†) as
found in the earlier study. The 1H NMR spectrum at 500 MHz shows
the reported downfield shift for the C8-hydrogen of 2.60 ppm
relative to the free ligand which is indicative of an anagostic
interaction whereby the nitrogen atom is coordinated to the metal
and H8 lies above the coordination plane.10 With analogy to other
anagostic complexes arising from the use of PdCl42-, the complex
would be formulated as the anion [PdCl3(1-tetralone oxime)]−.12 In
the 13C NMR spectrum, the C8 resonance shifts from 125.0 ppm in
the free ligand to 128.6 ppm when coordination of the ligand to Pd Fig. 1 Aromatic region of the 1H NMR spectrum for the reaction
is indicated by a shift in the C1 (C=N bond carbon) resonance between PdCl42- and 1-tetralone oxime in CD3OD mixed at 0o C,
position from 155.5 to 164.0 ppm. A further indication of the warmed to 25o C and recooled to 0o C.
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 3 of 13
Journal Name
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
to the agostic H8 does decrease and two small peaks that are forms a distorted boat-like configuration whereas on the right-hand
present around δ = 9.5 in the 0o C spectrum (see fig. 1) increase in side there are two boat-like configurations arising from positional
size. A DEPT 135 13C NMR spectrum shows there are 8 aromatic C–H differences at C22 and C23 occuring from disorder at these
carbons present for these two species in addition to those found for positions. The Pd−Cl bonds are asymmetric with the bond lying
the anagostic, agostic and soluble cyclopalladated intermediates.
trans to C9 being longer than the one lying trans to nitrogen [bond
Restoring the temperature to 0o C then gives the original spectral lengths 2.474(2) and 2.329(1) Å respectively]. The Pd−C and Pd−N
characteristics for these peaks at this temperature and the same bond lengths are similar to a range of other C,N-cyclometallates in
features were produced on cooling and reheating the solution general and the bond angles associated with the atoms in the
several times to the same temperatures. The position of the δ = 9.5 coordination plane are also unspectacular.
resonances for H8 is well downfield compared to H8 for the free
Coordination of the Pd centres to the two Cl ligands results in an
ligand suggesting that hydrogen is again experiencing an above- overall ‘butterfly’ arrangement for the molecule (fig. 7 of the ESI†)
plane, anagostic environment that is different to that found for H8 and in this case, there are several weak hydrogen bonds potentially
after mixing the components. The low-temperature study would present for some of the hydrogen atom placements. The (N)–OH
then suggest that there is some form of equilibria between the group hydrogens and the adjacent Cl14 ligand have H····Cl
apparent agostic component and the two δ = 9.5 anagostic separations of 3.199 and 2.595 Å with the latter representing a
components but we have not investigated this aspect further at this well-developed interaction and on the other side of the molecule,
stage.
the aromatic ring C8 and C26-hydrogens have H····Cl separations of
2.820 and 2.840 Å.
X-ray crystal structure determination of the insoluble
cyclometallated complex
DFT description of the agostic interaction in [PdCl2(1-tetralone
In our earlier work on the reaction of Li2PdCl4 and 1-tetralone oxime)] (1)
oxime the cyclopalladated product of the reaction was
characterised by analytical data, NMR and IR spectroscopy but we To obtain information about a putative agostic interaction in the
were unable to obtain a crystalline sample. In the present reaction of 1-tetralone oxime and PdCl42-, a full computational
investigation, the NMR tube was not used in a spinning mode and characterisation of the agostic complex [PdCl2(1-tetralone oxime)]
under these conditions, the product that formed in the tube was (1) was carried out. The dispersion-corrected PBE-D3 functional was
found to be crystalline.
used in the calculations as the inclusion of dispersion for weak
interactions is regarded as important.15 In previous computational
An X-ray crystal structure shows the molecule is an asymmetric
chloro-bridged dimer with C,N-cyclopalladated 1-tetralone oxime work on anagostic complexes of rhodium using a variety of 1-
ligands at either end of the molecule in a mirrored coordination tetralone derivatives,10 we carried out some functional testing and
mode and a regular square planar coordination geometry about settled on the PBE-D3 functional after finding that it performed as
each palladium centre (fig. 2). On both sides of the molecule, the well as BP8616 which conforms to the uniform electron gas (UEG)
plane made up by the aromatic rings lie slightly out of this limit17 and is known to perform well for agostic interactions18 as
coordination plane by 10.7 and 4.8o for the left and right hand sides well as anagostic interactions.19 Keeping these points in mind, and
respectively. The alicyclic ring on the left-hand side of the molecule
to give comparability to studies we have carried out on anagostic
interactions, we have continued to employ PBE-D3 for the agostic
interactions in this work.
As a strategy for uncovering the nature of the interactions in the
present study, we have computed the structures of the various
agostic complexes and then compared the structural features in
terms of the important metrics. Then, in relationship to bench
synthesis protocol, we have computed spectroscopic data to give an
initial indication of the change in properties and followed on with
in-depth analysis of the interactions using QTAIM and NBO analysis.
We are not aware of a previous complete computational study like
this for an aromatic ring agostic interaction involved in a
Fig. 2 X-ray crystal structure of the insoluble cyclometallated cyclometallation reaction. This type of approach to understanding
complex retrieved from the NMR monitored reaction. Thermal underlying fundamentals greatly reduces the need for bench
ellipsoids represent 50% probability. Selected bond lengths, syntheses.10,20
separations (Å) and angles (o): Pd(13)−N(11) 2.012(5), Pd(13)–C(9) The computed structure of [PdCl2(1-tetralone oxime)] (1) (fig. 3;
1.983(6), Pd(13)–Cl(14) 2.474(2), Pd(13)–Cl(34) 2.329(1), C(1)–N(11) selected metrics in table 1) shows that the oxime ligand coordinates
1.286(8), N(11)–O(12) 1.382(7), Cl(14)····H-O(32) 2.60, Cl(14)····H- to the metal with the C8–H bond electron density lying trans to the
O(12) 3.20, Cl(34)····H-C(8) 2.82, Cl(34)····H-C(28) 2.84; N(11)– Cl2 ligand. This orientation arises when the ligand is rotated about
Pd(13)–Cl(34) 173.8(2), N(11)–Pd(13)–Cl(14) 95.5(2), C(9)–Pd(13)– the Pd–N bond so that the (N)–OH hydrogen forms a close
Cl(14) 174.8(2), C(9)–Pd(13)–N(11) 80.3(2), C(9)–Pd(13)–N(11) approach to the Cl2 ligand [(N)–OH····Cl distance, 2.092 Å] which is
95.5(2), Pd(13)–Cl(34)–Pd(33) 93.7(5), Pd(13)–Cl(14)–Pd(33)86.7(5), representative of a weak hydrogen bond. At the same time, the
Cl(14)– Pd(13)–Cl(34) 88.4(5), Pd(13)–N(11) –O(12) 125.3(4).
alicyclic ring portion of the ligand has a boat-like conformation
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 4 of 13
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Journal Name
5
8a
6
8
HO
2.092
N
2.240
7
Cl2
H
Pd 1.819
(a)
(b)
2.865
Cl1
(b)
Fig. 4 Space-filled models for 1 looking (a) underneath the agostic
interaction; (b) above the agostic interaction.
(a)
For the C8–H bond agostic approach in 1, coordination of the
ligand is such that the Pd····C and Pd····H separaꢀons are 2.240 and
1.819 Å respectively (fig. 3b) when the H atom sits more below the
coordination plane than the C atom sits above it (figs. 3c and 4). A
major feature of the interaction is the lengthening of the C−H bond
to 1.152 Å compared with the free-ligand value of 1.081 Å. This
lengthening is well documented for agostic C−H bonds and is
(c)
Fig. 3 (a) The energy minimised structure of [PdCl2(1-tetralone
oxime)] (1); (b) Aromatic ring numbering and important separations
(Å) for 1; (c) ball and stick diagram showing the NPdCl2 coordination
plane extended (red line) into the C(8)−H bond.
determined by the strength of the interaction.4b,f A second feature
of the interaction is the deformation of the C8−H bond out of the
plane of the aromatic ring, as shown by the H7−C7−C8−H8 torsion
angle of -27.2o. It can be seen that if the H atom did not bend away
from the metal, an unacceptable clash would develop and so if the
nitrogen is to coordinate to the metal, the H atom has no
alternative but to bend away. Both the lengthening and the
deformation have been shown to occur by X-ray crystallographic
determinations of complexes containing agostic interactions
involving aromatic rings21 and both are features of Ru, Rh and Ir
pincer complexes and more recently a Co pincer, containing
aromatic C−H agostic bonds.22 A third feature seen in agostic
complex 1 is that the C7−C8 and C8−C8a bonds lengthen in
comparison to the individual adjacent C−C bonds whereas in the
free ligand, the two bond lengths are shorter. However, the various
C−C−C bond angles are all very similar [see the comparisons for the
bond lengths and angles in fig. 9 of the ESI†].
which is very similar to that found in the free ligand (fig. 8 of the
ESI†). A similar ring conformation was found by computation for the
anagostic approach in [RhCl(CO)2(1-tetralone oxime)].10 With this
conformation and the (N)–OH····Cl hydrogen bonding component in
the present agostic complex, the Pd–N–C bond rotation angle is
very flat at just 14.4o and the benzene ring then lies at a plane
angle of 34.5o to the palladium coordination plane (Ar/CP angle in
table 1).
Table 1 Selected metrics and spectroscopic data for
complexes 1−5
Calculated spectroscopic data for 1 (table 1) indicate that in the 1H
NMR spectrum there is an up field shift of 1.54 ppm for the agostic
H8 compared with the free ligand. This shift is known to arise from
anisotropy associated with the metal centre23 and was used to
indicate the presence of an agostic interaction in the present
cyclopalladation reaction9 but it has more recently been shown that
the extent of the shift cannot be related to the interaction strength
as the equatorial current density about the metal centre can vary.14
It should be noted that this calculated upfield shift of 1.54 ppm is
for a static interaction in the gas phase at 0 K and does not relate to
1
2
3
4
5
Bond Length (Å)a
Pd-N
Pd-Cl(cis)
Pd-Cl(trans)
C-Hb
2.015
2.272
2.293
1.152
1.081
2.053
2.279
2.280
1.151
1.090
2.011
2.272
2.294
1.152
1.090
2.047
2.278
2.280
1.148
1.090
2.088c
2.260
2.252
1.164
1.090
Separations (Å)a
Pd····H
1.819
2.240
2.865
1.841
2.224
2.899
1.822
2.228
2.887
1.853
2.217
2.931
1.786
2.206
2.846
Pd····C
Cl····H
Angles (o)
Pd-N-C
119.9
95.2
54.0
116.7
93.2
55.7
120.8
94.4
54.6
117.6
92.2
56.6
114.3
94.4
53.9
Pd····H-C
Pd····C-H
Dihedrals (o)
Cl-Pd-N=C
Ar/CP angle
N-C-C-C torsiond
C-H deform
14.3
34.5
15.1
-27.2
37.2
51.4
14.7
-26.0
13.7
36.1
17.9
-27.1
37.7
54.6
18.5
-25.4
27.3
42.6
11.1
-28.0
1
the position of H8 determined by the earlier experimental H NMR
study. Similarly, in the calculation, spectroscopic characteristics that
relate to the interaction strength, are associated with C−H bond
electron density depletion and for 1 there is a significant decrease
in the 1JC−H coupling constant in comparison to the free ligand
(103.2 and 139.4 Hz respectively) and the calculated C–H stretching
frequency shifts to lower energy by 670 cm-1 (νC-H, 2462.8 cm-1 in 1,
3133.3 cm-1 for the free ligand).
Spectral Data
b, e
)
νC-H (cm-1
2462.8
3133.3
670.6
6.59
2473.0
3132.9
659.9
5.27
2468.5
3134.8
666.3
6.75
2498.5
3134.5
636.0
5.47
2346.8
3119.9
773.1
4.75
δ H8 (ppm)b, e
8.13
8.22
8.04
8.18
8.33
+1.54
103.2
139.4
36.2
+2.95
105.9
139.5
33.6
+1.29
103.0
139.0
36.0
+2.71
106.6
140.2
33.6
+3.58
103.8
141.1
37.3
1JC-H (Hz)b, e
a Valuesto 3 decimal places consistent with level of
QTAIM analysis (selected data in table 2) shows that a bond path
and a bond critical point (bcp) occur for the metal and H8 and there
is a centrally located ring critical point (rcp) (see fig. 5a) which are
all features common to agostic interactions. Other features typical
optimisation, SCF convergence and XC grids. b Values in
bold are for the free ligand. c Pd-O bond length.
d N−C1−C8a−C8 torsion angle e Values in italics indicate the
difference between the complex and free ligand.
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 5 of 13
Journal Name
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Table 2 Selected QTAIMa and NBOb data for complexes 1
−5
contributions within the molecule could also be affecting these
parameters. Keeping this point in mind, and as a descriptive starting
point for the later work presented, the C8 atomic charge becomes
significantly more negative than in the free ligand [q(C) values, -
0.084e and 0.001e respectively]; the total energy changes very little
[E(C) values, -40.1 and -38.1 respectively]; the dipolarisation
becomes larger [M(C) values, 0.142 and 0.115 respectively] and the
atomic volume decreases significantly [v(C) values, 69.1 and 82.3
respectively]. However, one of the purposes of the present study is
to try and ascertain what the agostic effect is on these parameters
1
2
3
4
5
Pd····X bond path
Pd····H
0.0820
0.2631
Pd····C
0.0820
0.2107
Pd····H
0.0828
0.2569
Pd····C
0.0822
0.2001
Pd····C
0.0864
0.2636
ρ(bcp)
∇
2ρ(bcp)
C-H bcp
ρ(bcp)
0.2333
0.2810
0.1059
0.0126
0.2346
0.2810
0.1005
0.0126
0.2338
0.2810
0.1076
0.0132
0.2361
0.2810
0.0997
0.0133
0.2286
0.2810
0.1024
0.0096
ε(bcp)
H-Atomic basin
Charge, q(H)
0.056
-0.631
0.099
36.0
0.044
-0.633
0.108
37.2
0.057
-0.632
0.101
36.1
0.044
-0.629
0.110
37.4
0.050
-0.634
0.105
36.5
Total Energy, E(H)
Dipolarisatn, M(H)
Volume, V(H)
C-Atomic Basin
Charge, q(C)
Pd-Atomic Basin
Charge, q(Pd)
NBO donations:
C-Hσ to
-0.084
0.641
-0.070
0.627
-0.086
0.642
-0.072
0.629
-0.066
0.659
Pd-Clσ*(trans)
Pd-Clσ*(cis)
58.6
10.6
57.7
9.5
56.9
8.9
53.7
8.0
64.4
9.3
C(7)-C(8)π to
Pd-Clσ*(trans)
Pd-Clσ*(cis)
20.3
6.5
18.9
6.0
20.3d
6.8d
19.1d
6.5d
17.9
5.4
C(8)-C(8a)σ to
Pd-Clσ*(trans)
Pd-Clσ*(cis)
1.4
0.4
1.7
0.5
1.4
0.4
1.7
0.5
1.3
0.4
a Atomic units: ρ(bcp) = e/Bohr3;∇2ρ(bcp) = e/Bohr5; ε,
dimensionless; q = e. E = Hartree; M = e*bohr,V =1/bohr3
b 2nd order perturbation energy E(2) values in kcal mol-1.
c Values in bold are for the free ligand. d Values for the C8-
C8aπ donations.
of agostic interactions are the positive values for the electron
density [ρ(bcp), 0.0820] and Laplacian of the electron density
[∇2ρ(bcp), 0.2631] which would indicate a purely electrostatic
interaction but the negative energy density [H(bcp) value -0.0210]
and a value above 1 for |V|/G (|V| = absolute potential energy
density, G = kinetic energy density) which are more reliable
indicators than the Laplacian,24 indicate partial covalency. A 2D
contour plot of the Laplacian of the electron density (fig. 5b) shows
this partial covalency as an area of accumulated charge density
along the C−H bond in the direcꢀon of the Pd to C−H bond vector.
In addition, the C−Hσ bond electron density is depleted with
respect to the free ligand [ρ(bcp) values 0.2333 and 0.2810
respectively] and the ellipticity is very large in the direction of the
bond path [ε(bcp) value, 5.3365]. It has been noted that bond
ellipticities for agostic interactions at the M….H bcp are large and
could provide a qualitative measure of the delocalization within the
QTAIM framework.25 However, it has been shown that the agostic
carbon can successfully compete for the bond path to the metal
with the reason so far being indetermined26 so that direct
comparisons are not always possible (see later).
(a)
(b)
Fig. 5 (a) QTAIM molecular graph showing bond paths between
H8···Pd and OH···Cl separations for 1. (b) 2D contour plot of the
Laplacian of electron density for the C−H and metal interaction for
1.
and in particular to account for any electrostatic contribution to the
agostic bonding component (see later). In this regard it is noted that
the agostic interaction has been regarded as purely coulombic.27
For the present complex, in terms of overall atomic basin charges,
the agostic H-atom would form a weak repulsive interaction with
the metal [q(H) and q(M) values, 0.056e and 0.641e] and the C-
atom a weak attraction [q(C) value, -0.084e]. However, because of
the anisotropic charge distribution about the C−H bond, the
accumulation of electron density towards the metal will mean that
consideration of the atomic basin charges alone would
overestimate the repulsion between the metal and H so that
multipole terms would need to be included. If the H atom did not
bend away from the metal, not only would there be an
unacceptable electrostatic situation but it would come much closer
to the metal centre than the current 1.819 Å.
NBO analysis28 gives a valence bond approach to bonding and can
be used to give a quantitative measure of the C−H bond electron
delocalisation in an agostic interaction.25 Use of the second order
perturbation energy, E(2), derived from the Natural Bond Orbitals,
then allows a quantitative assessment of the orbital nature of an
agostic interaction that is not available from the molecular or
atomic orbitals. It should be noted that the energies obtained from
molecular orbitals are meaningless which has led to the description
Turning to the QTAIM atomic basin characteristics, with the
agostic bond developed and electron density from the C−H bond
donated towards the metal, the charge on H8 is slightly more
positive than in the free ligand [q(H) values, 0.056e and 0.040e
respectively] as might be expected when the C−Hσ bond electron
density is slightly depleted; the total energy becomes more
negative [E(H) values, -0.631 and -0.609 respectively]; the
dipolarisation becomes slightly less positive [M(H) values, 0.099 and
0.126 respectively] and the atomic volume decreases significantly
[v(H) values, 36.0 and 46.5 respectively].
At this stage, it should be noted that care must be taken in not
interpreting the changes in the H-atomic basin as being only due to
development of the agostic approach (ie donation of the C−Hσ
bond electron density to the metal) particularly as other orbital
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 6 of 13
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Journal Name
of using atomic orbitals for analysis as ‘illustrative’ only.28 It should
A major feature of the the NBO analysis is that significant bonding
also be noted that NBO energies cannot be compared with other components that do not involve C–H bond electron density are
energies but are themselves directly comparable. The difficulties present which have not been recognised before. Two of the
associated with the determination of the strength of an agostic aromatic ring bonding molecular orbitals donate to the two Pd–Clσ
interaction have been recorded29 with the actual strengths being antibonding orbitals arising on the Cl atoms that lie trans and cis to
estimated at between 1 and 20 kcal mol-1. 7c, 30
the agostic approach and these are also the antibonding orbitals
The description of the agostic donation provided for 1 by the NBO that the C–Hσ electron density donates into to form the agostic
analysis shows that there are two donations of the C−H electron covalency. One aromatic ring donation involves the π-orbital that
density to the metal and these do not involve metal d orbitals in spans the agostic and ‘outer’ carbons [ C7–C8π to Pd–Clσ*(trans)
this formally d8, Pd2+ system but instead utilise the trans and cis- and Pd–Clσ*(cis) donations, E(2) values, 20.3 and 6.5 kcal mol-1] and
related Pd–Clσ antibonding orbitals [C–Hσ to Pd–Clσ*(trans) and the other and more minor component, involves the σ-orbital that
Pd–Clσ*(cis), E(2) values, 58.6 and 10.0 kcal mol-1 respectively]. Fig. spans the agostic and ‘inner’ carbons [C8–C8aσ to Pd–Clσ*(trans)
6 shows contour plots and spatial images for these orbital and Pd–Clσ*(cis) donations, E(2) values, 1.4 and 0.4 kcal mol-1].
contributions which account for the pure agostic donation in the Complexes in which C–Cσ bond electron density donates to metal
complex. These images also indicate further that if the H8-hydrogen centres have been termed ‘C–C agostic interactions’.31 NBO orbital
did not bend away from the metal, an unacceptable orbital overlap
2D contour plots and 3D spatial images diagrams for the two
interactions involving the aromatic ring C7–C8π orbitals are shown
in Fig. 7. These plots indicate that for both interactions, it is
electron density in the vicinity of C8 that is involved in the donation
to the metal so that there is intimate association in the agostic
complex of the metal centre and the carbon atom that will
eventually form the Pd–Cσ bond in any ensuing cyclometallation
reaction. The two C7–C8π donation E(2) values of 20.3 and 6.5 kcal
mol-1 compare with the (N)–OH···Cl weak hydrogen bonding
component (H···Cl separation 2.092 Å) in the complex of 23.8 kcal
mol-1.
In regard to the two donations, it should be noted that in aromatic
ring systems, σ and π orbitals do not overlap and this is the case for
the C8–Hσ and C7–C8π orbital systems so that even though the two
donate to the same metal-based orbitals, there is no realistic
synergic effect developed and the donations are separate entities.
However, the NBO analysis shows there is a small overlap of the
C7–C8π orbital with the C8–Hσ* orbital (carbon end) [E(2) value, 3.7
kcal mol-1] so that synergy is not entirely lacking.
Fig. 6 NBO orbital overlap 2D contour and 3D spatial overlap images
for 1. Top, C–Hσ to Pd–Clσ*(trans) donation [E(2) value, 58.6 kcal
mol-1]. Bottom, C–Hσ to Pd–Clσ*(cis) donation [E(2) value, 10.6 kcal
mol-1].
condition would most probably result, especially in the case of the
C–Hσ to Pd–Clσ*(trans) orbital donation. The NBO analysis also
shows there are two small-energy back donations from the metal to
the C–H antibonding orbital [LP (lone pair) Pd to C–Hσ* E(2) values,
2.9 and 4.9 kcal mol-1 respectively] which are known features
associated with agostic interactions7b and represent a type of weak
hydrogen bond that is repulsive in terms of the electrostatics [q(H)
and q(Pd) values 0.056e and 0.0641e respectively] but attractive in
terms of the weak covalent back donation. There are also slight
contributions to the C–H antibonding orbital from other lone pairs
(LP’s) on the metal, one on the cis–Cl ligand and Pd–Cl bonding (BD)
components [E(2) values range from 0.15 to 0.68 kcal mol-1 (see full
details in data 3 of the ESI†) The interaction involving the C–H NBO
orbitals for the complex thus provides for much stronger donation
to the metal than for back-donation from it. Other work has shown
that back donation can become strong in some complexes and
provide a dominant ‘hydrogen bonding’ situation.25
Fig. 7 NBO orbital overlap 2D contour and 3D spatial overlap images
for (1). Top, C7–C8π to Pd–Clσ*(trans) donation [E(2) value, 20.3
kcal mol-1]. Bottom, C7–C8π to Pd–Clσ*(cis) donation [E(2) value,
1.4 kcal mol-1].
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 7 of 13
Journal Name
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Although the NBO energies of the π-donations are only about one between the C7–H atom with the cis-Cl ligand (distance 3.048 Å)
third that of the agostic donations, they still provide a ‘clip’ that and the (N)–OH oxygen atom with an adjacent alicyclic ring
would appear to help anchor the agostic approach and there is also
hydrogen (distance 2.640 Å) both of which show light blue-green
the electrostatic component, albeit weak in comparison [q(C) and iso-surfaces indicating weak attractive interactions.
q(M) values, -0.084e and 0.641e], to consider. For donations that
As a final feature of the complete analysis of the agostic
increase the complexity but more importantly provide an assistance interaction in 1 we include the QTAIM charges for the aromatic ring
to the agostic interaction, we have previously introduced the term of the 1-tetralone oxime free ligand and the agostic complex so that
‘syndetic’ from the Greek word syndekos meaning ‘serving to a comparison can be made as to the effect of coordination in the
unite‘, ‘to connect‘ or ‘bind together’.11 The agostic interaction is agostic mode. For the aromatic ring, this is an important
thus not fixed by C–H bond electron density donation to the metal consideration as electrophilic aromatic substitution involving
alone and it appears that it can be a more complex interaction for Wheland character has been proposed for the metallation process
aromatic rings than previously thought. It can be seen that subsequent to the agostic approach.7 The Pd····C separaꢀon in 1 is
aromaticring π-orbitals adjacent to the Pd centre, are intimately, also quite close (2.240 Å) and the NBO analysis showed there is an
albeit weakly, involved with the metal and this has consequence for intimate association of the metal and a C–Cπ orbital close to the C8
the formation of the metal carbon bond in any ensuing carbon atom (fig. 7). For the carbon atoms of the aromatic ring (fig.
cyclometallation reaction.
9), the agostic carbon (C8) becomes much more negative than in
the free ligand [q(C) values, -0.084e and 0.001e respectively] and
the two carbons ortho to this position become more positive,
particularly the C7 carbon which involves the syndetic C–Cπ
donation where q(C) changes from -0.014e to 0.014e. The carbon
para to C8 also becomes more positive, but to a much smaller
degree than the two ortho positions and there are similar small
increases in positive charge from both meta positions. These
features would appear to be associated more with electron
donation out of the aromatic ring when the agostic and syndetic
donations take place than any Wheland-type character. The
aromatic ring hydrogens reflect this situation, with H5 and H6
becoming slightly more positive compared to the free ligand, H7,
which involves the syndetic C–Cπ donation, becomes much more
Although not necessary for the agostic interaction identification,
we have included a non-covalent-interaction (NCI) analysis for 1 to
give for the first time, the characteristics for this type of weak
interaction. The use of the NCI approach is useful in weak
interaction situations where QTAIM analysis does not give a bond
path due to a bcp and ring critical point (rcp) coalescing32 but in the
present case the bcp and rcp are well separated. This approach
allows the nature of weakly interacting molecular regions to be
resolved from 3D NCI iso-surface plots which are based on colour
gradients where red indicates destabilising interactions, blue for
stabilising interactions and green for the very weak delocalised but
stabilising interactions. The depth of the particular colour is
associated with how strong the interaction is.33
The iso-plot for 1 (fig. 8) shows that the agostic interaction is positive [q(H) values, 0.079e and 0.014e] and the agostic hydrogen
indicated by an outer orange disc of depletion. This type of outer with its charge the least effected [q(H) values, 0.056e and 0.040e].
ring has been interpreted as a small area of steric crowding34 and is The small increase for H8 may be a result of Pd to C–Hσ* back
also seen for the Pd–Cl bonds but not the Pd–N bond which donation where the two major contributions have E(2) values of 4.9
contains the smaller N atom. A blue disc for the interaction and the and 2.9 kcal mol-1.
Pd–Cl bonds does not appear in this plot. The N–OH····Cl interaction
which the NBO analysis shows is weaker than the agostic
interaction is also revealed as a lighter blue iso-surface.
Features of the NCI plot not revealed by bond paths and bcp’s in
the QTAIM analysis are the weak intramolecular interactions
(a)
(b)
Fig. 8 NCI isosurface plot of complex 1. The sign of the second
density Hessian eigenvalue (λ2) times the electron density is shown
Fig. 9 QTAIM atomic charges for aromatic ring carbon atoms [q(H)
by the colour map.
values in e]. (a) 1-tetralone oxime. (b) complex 1.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 8 of 13
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Journal Name
∇2ρ(bcp) is >0, the electron density, p is locally depleted35]. The
ellipticity is now lower [ε(bcp) value, 2.2016, 5.3365 for 1] and –
V/G more negative (-1.3165; -1.2419 for 1). For the C–H bond
electron density, there is little change for the two complexes when
the complexes and ligands are compared [ρ(bcp) differences,
0.0464 and 0.0477 for 2 and 1 respectively].
The effect of structural changes made to the ligand.
(a) (N)-OMe complex 2
A change was made to the nature of the ligand donor N-atom to
remove the (N)–OH···Cl hydrogen bonding that is present in 1 and
thus allow the ligand to be unconstrained by this component. In
(N)–OMe complex 2 (scheme 2) where the H-bonding is removed,
there would be an apparent clash of the OMe group with the
coordination plane Cl2 atom with the result that the ligand rolls
slightly upwards so that the plane made by the coordination atoms
now drops down the C8–H bond [see the visualisation of this in fig.
8 in the ESI†]. This change can be seen in the metrics where the Pd–
N bond becomes significantly longer than for (N)–OH complex 1
(2.053 and 2.015 Å respectively) and the Cl2–Pd-C=N dihedral
increases to 37.2o (14.4o in 1) even though the alicyclic ring section
For the atomic basin properties, with the small decrease in the
electron density at the bcp in comparing 2 with 1, the H-atomic
charge is now very similar to the free ligand and there are only
small changes to the other parameters which would appear to be
insignificant. For the C-atomic basin, where the free ligands of both
complexes have similar atomic basin properties, for the complexes,
the atomic charge becomes less negative [q(C) values, -0.070 and -
0.084 for 2 and 1 respectively] and when this occurs the carbon
moves slightly closer to the metal in 2 compared to 1 (2.224 and
2.240 Å respectively) which suggests that electrostatics do not play
a major role for the positioning of this atom. The energies are the
same, but the dipolarisation and the volume become smaller which
would fit with the expected trends.
Turning to the NBO characteristics, as the Pd–N bond length has
increased (2.053 and 2.015 Å in 2 and 1 respectively) there is little
change in the strength of the donation from N to the cis-Pd–Cl
antibonding orbital but for the trans case, significant lowering
occurs [LP(1) N to BD* Pd–Cl2 values 96.3 and 105.4 kcal mol-1]. In
this situation, where the ligand has rolled up a little and where the
Pd····H distance in the agostic interaction has lengthened (1.841
and 1.819 Å in 2 and 1 respectively) and the Pd····C distance has
decreased (2.224 and 2.240 Å in 2 and 1 respectively), the C–H
bond electron density donation to the metal NBO energetics
becomes marginally smaller [C–Hσ to Pd–Cl σ*(trans) and Pd-Cl
σ*(cis) E(2) values 57.7 and 9.5 kcal mol-1 in 2; 58.6 and 10.0 kcal
mol-1in 1]. Also, the donations from the metal to the C–Hσ
antibonding orbital show little change as do the syndetic aromatic
ring π-orbital donations to the metal (see data 3 of the ESI†). This
would indicate that the ligand positional change has little effect on
the agostic interaction in terms of orbital overlaps.
MeO
2.224
Cl2
Pd 1.841
2.899
Cl1
Scheme 2. Structure and important separations (Å) for 2.
of the ligand shows very little change (e.g. N–C1–C8a–C8 dihedral
angles 14.7 and 15.1o for 2 and 1 respectively).
The plane the aromatic ring makes with the coordination plane
(Ar/CP angle in table 1) now increases markedly from 34.5o in 1 to
51.4o in 2. In this case, the Pd····H separaꢀon increases significantly
to 1.841 Å (1.819 Å in 1) and the Pd····C separation becomes smaller
(2.224 and 2.240 Å for 2 and 1 respectively) so that the effect of the
rolling action is to take the H atom away from the metal and push
the C atom towards it. The C–H bond lengthening shows little
change (1.151 and 1.152 Å in 2 and 1 respectively] as does the C–H
bond deformation angle out of the arene ring plane (-26.0o in 2, -
27.2o in 1).
(b) Acetophenone oxime complexes 3 and 4
To gain further insight into the newly identified syndetic π-
donations, we have looked at analogues of 1 and 2 in which the
alicyclic ring is not present (scheme 3). In this situation, the
aromatic ring is free to rotate which could potentially allow the
metal greater access to the aromatic ring π-electron density. This
The calculated spectroscopic data for 2 show an up field shift in
the 1H NMR spectrum of 2.95 ppm for H8 compared with the free
ligand as the hydrogen moves further away from the metal centre
and possibly into a new anisotropy region14 in comparison to 1 (1.54
ppm up field shift) and the 1JC–H coupling constant is slightly larger
(105.9 Hz for 2, 103.2 Hz for 1; the free ligand values are essentially
the same at 139.5 and 139.4 Hz respectively) indicating that there is
a very slight weakening of the agostic interaction in terms of the
strength of the C–H bond. The calculated C–H stretching frequency
reflects this situation with a smaller shift to lower energy (660 cm-1
compared to 670 cm-1 for 1).
(3) R = H
(4) R = Me
(3) R = H
In terms of the QTAIM characteristics, there is a bcp with the
same electron density value as for 1 [ρ(bcp), 0.0820] but the
(4) R = Me
Laplacian of the electron density which gives
a better
Scheme 3. Structure of acetophenone oxime complexes 3 and 4.
representation at the flat surface at the bcp, has a lower value
(∇2ρ(bcp), 0.2107; 0.2631 for 1) [in QTAIM methodology, when
8 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 9 of 13
View Article Online
DOI: 10.1039/C7DT03525J
Journal Name
ARTICLE
type of ligand is the more usual aromatic ring system on which
cyclometallation reactions are carried out for directed C–H bond
functionalisations.4
Pd–Clσ*(trans) and Pd–Clσ*(cis) E(2) values 56.9, 8.9 kcal mol-1 c.f.
58.6, 10.0 kcal mol-1 for 3 and 1 respectively and the aromatic ring
syndetic π-donation NBO E(2) values are essentially the same (20.3,
The computed structure for [PdCl2(acetophenone oxime)] (3) is 6.8; and 20.3, 6.5 kcal mol-1 for 3 and 1 respectively). However, with
shown in fig. 10 where it is compared with the structure of 1. the unconstrained aromatic ring in 3, the donation changes to the
Comparing the agostic bond metrics of 3 with alicyclic ring complex inner π-orbital system which is the equivalent of the C8–C8aπ -
1, the Pd····H separation distance is very similar (1.822 and 1.819 Å orbital system of tetralone oxime complex 1 rather than the C7–
respectively) but the Pd····C separation is slightly shorter (2.228 and C8π system.
2.240 Å respectively). Inspection of the other structural details In a further study of the effect of a rotatable aromatic ring without
(table 1) shows that they are very similar, including the aromatic an (N)–OH····Cl ‘lock’, the structure of the (N)–OMe analogue,
ring/coordination plane angles (Ar/CP angles, 36.1 and 34.5o [PdCl2(acetophenone methoxy-oxime)] complex
4 was also
respectively), the C–H bond deformations (-27.1 and -27.2o calculated. In this case, the Pd····H and Pd····C separaꢀons (1.853
respectively) and the C–H bond length (1.152 Å in both complexes; and 2.217 Å) are very similar to those of [PdCl2(1-tetralone
1.090 and 1.081 Å in the respective free ligands). Overall, the C–H methoxy-oxime)] complex 2 (1.841 and 2.224 Å) as are the relevant
bonds in the two agostic complexes position in an almost identical angles and dihedrals for the two complexes shown in table 1. Also
manner and this is not dependant on whether the aromatic ring has similar are the QTAIM and NBO parameters shown in table 2. In
the potential to rotate. The calculated NMR and IR data shows that particular, the NBO analysis indicates the agostic donation is
the agostic interaction is equally well developed in 3 than 1 with the slightly smaller in complex 4 than in 2 [C–Hσ to Pd–Clσ*(trans) and
1JC–H coupling constant difference to the free ligand essentially Pd–Clσ*(cis) E(2) values, 53.7 and 8.0; 57.7 and 9.5 respectively]
unchanged (33.0 and 33.2 Hz respectively] as are the calculated C–H and the syndetic aromatic ring π-donations which also switch to
stretch shifts to lower energy in comparison to the free ligand the inner π-orbital system, are very similar [C8–C8aπ to Pd–Cl
values (666.3 and 670.6 cm-1).
σ*(trans) and Pd–Clσ*(cis) E(2) values, 19.1, 6.5 and 18.9, 6.0 kcal
Inspection of the QTAIM data shows that at the Pd····C bcp the mol-1 for 4 and 2 respectively].
electron density is equally well developed [e.g. ρ(bcp) values are
both 0.2810, H(bcp) values, -0.0311 and -0.0379; -V/G 1.4013 and (c) [PdCl2(1-tetralone)] complex 5
1.5150 respectively for 3 and 1 respectively] and there is similar
As the approach to catalytic activation of aromatic C–H bonds
electron density in the C8–H bonds and the ellipticities do not differ involves a variety of ligand coordinated directing groups,3 in which
much (compare the ρ(bcp) and ε(bcp) values for 3 and 1 in table 2]. less-stable cyclometallates lead to better postfunctionalisation,4 we
In the NBO analysis, the agostic donation is slightly smaller [C–Hσ to have also looked at the non-derivatised 1-tetralone itself as a
ligand. In complex 5 (scheme 4) the oxygen atom coordinates to the
Pd centre with a significantly longer bond (Pd–O bond length 2.088
Å) than the oxime derivatives [Pd–N bond lengths 2.015 and 2.053
Å respectively for the (N)–OH and (N)–OMe derivatives 1 and 2].
[PdCl2{acetophenone (N)–OH}] (3) [PdCl2{1-tetralone (N) –OH }] (1)
5
Scheme 4. Structure of 1-tetralone oxime complex 5.
With oxygen as the coordinating atom, the effect on the agostic
interaction in 5 (fig. 10) is to bring H8 closer to the metal centre
(Pd····H separation 1.786 Å) than for the oxime ligands [Pd····H
[PdCl2{acetophenone (N)-OMe}] (4)
[PdCl2(1-tetralone)] (5).
separations 1.819 and 1.841 Å for the (N)–OH and (N)–OMe
derivatives respectively]. This is also the case for the C8-carbon
where the Pd····C distance closes to 2.206 Å in comparison to 2.240
and 2.224 Å for oxime complexes 1 and 2. Overall, the C–Hσ
bonding component lies more squarely over the coordination plane
than for the nitrogen donor ligands (fig. 8 of the ESI†). The other
important structural feature pertaining to the C8–H bond is the
Fig. 10 Comparisons of the structures of [PdCl2(acetophenone
oxime)] (3), [PdCl2(1-tetralone oxime)] (1), [PdCl2(acetophenone
methoxy-oxime)] (4) and [PdCl2(1-tetralone)] (5).
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 10 of 13
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Journal Name
deformation out of the plane of the aromatic ring, which at -28.0o is agostic donation and this changes the way in which the agostic
slightly larger than for the oxime complexes and equates to the interaction has been traditionally viewed and as such, has
closest Pd····H separation for all the ligands (1.786 Å). However, the significant implication for C–H bond activation in ligand-assisted
deformation is not much different to (N)–OH ligand complex 1 cyclometallation reactions leading to C–C bond functionalization.
where the Pd····H separation is 1.819 Å.
Further computational studies are underway to assess the
Inspection of the spectroscopic data shows that the agostic relevance of syndetic donations in transition states in catalytic
interaction is better developed with the ketone oxygen donor cycles involving aromatic ring agostic interactions and also in the
present. The 1JC–H coupling constant is the smallest of all the increasing number of ligand coordinated directing groups that lead
previous complexes and the calculated C–H stretch shifts to the to metal-carbon bond formation.
lowest energy (compare the values in table 1). For the NBO analysis,
it is seen that the agostic donation for 5 is slightly larger than for all Experimental
the other ligands [C–Hσ to Pd–Clσ*(trans) and Pd–Clσ*(cis) E(2)
values, 64.4 and 9.33 kcal mol-1 (total 73.7 kcal mol-1); 58.6 and 10.0
kcal mol-1 (total 68.6 kcal mol-1) for 1, the next highest] and the
Materials and general methods
Lithium tetrachloropalladate was prepared by adding two
equivalents of anhydrous lithium chloride in water to a suspension
of palladium chloride in water and stirring the mixture until the
solid dissolved and a dark red solution was obtained. The solution
was filtered, the solvent removed under vacuum at 80o C until a
dark red hygroscopic solid was obtained which was stored under
nitrogen. 1-Tetralone oxime was prepared from 1-tetralone
(Aldrich) by reaction with hydroxylamine hydrochloride and sodium
hydroxide in ethanol. NMR spectra were obtained on a Bruker
Avance III 500 MHz spectrometer. For the NMR reaction studies,
sufficient CD3OD (Aldrich) was added to Li2PdCl4 and 1-tetralone
oxime in separate flasks to dissolve the components and the flasks
were then cooled in dry-ice acetone baths to the required
temperature. The contents were mixed at this temperature and
quickly transferred to an NMR tube cooled to the particular
temperature and the tube was then placed in the spectrometer also
cooled to this temperature. All NMR spectra were run without
spinning the NMR tube.
aromatic ring C7–C8π syndetic donation is slightly lower [C7–C8π
to Pd-Clσ*(trans) and Pd–Clσ*(cis) E(2) values, 17.9, 5.4 kcal mol-1
(total 23.3 kcal mol-1); 20.3, 6.5 kcal mol-1 (total 26.8 kcal mol-1) for
1. Interestingly, the back-donation from the metal to the C8–Hσ*
orbital which can be expected to further weaken that bond, is the
highest of all the complexes [LP Pd to C–Hσ*E(2) value, 8.1 kcal mol-
1
for 5, 5.0 kcal mol-1 for 3, the next highest]. The Pd–O bond is
significantly weaker than the Pd–N bonds in the (N)-donor ligands
[E(2) values, 58.2 and 6.0 kcal mol-1 (total 64.2 kcal mol-1) in 5;
105.4, 17.6 kcal mol-1 (total 123.0 kcal mol-1) for example, in 1]. As
already mentioned, this feature is of importance in the post-
functionalisation of the cyclometallate where the metal-carbon
bond for this type of ligand are found to be more reactive.4,5
Conclusion
The results of the work presented here indicate that a possible
agostic intermediate can be seen in the NMR spectra of the reaction
2-
X-ray single crystal structure determination
of 1-tetralone oxime and PdCl4 in CD3OD solution which gives rise
Low temperature36 single crystal X-ray diffraction data were
to a cyclometallated complex but importantly, this does not show
the lowering of the 1JC–H coupling constant normally associated
with agostic approaches. The computed structure of the agostic
intermediate [PdCl2(1-tetralone oxime)] (1) shows features
associated with the carbon atom making a close approach to the
metal but the H-atom bends away from it. In this situation, the
aromatic ring is inclined at an angle to the metal centre and an NBO
analysis indicates that C–Cπ orbitals donate to the same Pd-Clσ*
orbitals as does the C–Hσ bond electron density. This shows that
the agostic interaction in this type of system is more complex than
previously recognised in that the C–Cπ orbital donation would
appear to add assistance to the stability of the agostic interaction.
The term ‘syndetic donation’ is used to reflect the assistance the C–
Cπ orbital donation component provides. Although the two
components donate to the same metal-based orbitals, there is little
evidence for any meaningful synergy between them. The syndetic
donation is also present where rotatable aromatic rings engage in
the agostic interaction and this is also the case where a less stable
oxygen donor provides the initial coordination of the ligand rather
than the more usual N atom donor. This work presents a full
computational study for an agostic interaction after recognition of
the new complexity that is involved with the syndetic donation. The
study reveals that there can be an intimate association of an
collected using
a
(Rigaku) Oxford Diffraction SuperNova
diffractometer (λ=1.5418 Å). Raw frame data were reduced using
CrysAlisPro and the structures were solved using 'Superflip’37 before
refinement with CRYSTALS38 as per data 1 (CIF) of the ESI†. Full
refinement details are also given in data 1 (CIF) of the ESI†.
Crystallographic data have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication no.
1560179. Copies of the data can be obtained, free of charge via
http://www.ccdc.cam.ac.uk/data_request/cif.
Computational details
DFT PBE-D339 functional based geometry optimizations and
vibrational calculations for the complexes 1-6 were performed
using Gaussian09 (G09).40 A triple-zeta high quality basis set (aug-
cc-pVTZ-PP)41 was employed for Pd; aug-cc-pVTZ42 for the attached
ancillary ligands (Cl atoms), agostic hydrogen and nitrogen
(attached to Pd) and a double-zeta quality basis set (aug-cc-pVDZ)42
was used for the remainder of the atoms. All the structures were
energy-minimised and no imaginary frequencies were found in the
vibrational analysis. For the QTAIM analysis, the input files (.wfx)
were obtained from G0940 and QTAIM calculations were performed
aromatic ring π-orbital with the metal in conjunction with the with the AIMALL software package.43 NBO calculations were
10 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 11 of 13
Journal Name
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
performed with the NBO6.0 package44 and NBOView2.044 was used
to visualize the contours and surfaces of donor-acceptor
interactions. The same procedure was used for the ligand
calculations.
C. Whitwood, I. J. S. Fairlamb and R. N. Perutz,
Organometallics, 2015, 34, 4376–4386.
13 A. Enthart, J. C. Freudenberger, J. Furrer, H. Kessler and B. Luy,
J. Magn. Reson., 2008, 192, 314-22.
14 J. E. Barquera-Lozada, A. Obenhuber, C. Hauf and W. Scherer,
J. Phys. Chem. A, 2013, 117, 4304–4315.
15 (a) J, P. Pedrew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
1996, 77, 3865–3868; (b) E. R. Johnson, I. D. Mackie and G. A.
Dilabio, J. Phys. Org. Chem., 2009, 22, 1127–1135.
16 (a) W. Hieringer, J. Eppinger, R. Anwander and W. A. Herman,
J. Am. Chem. Soc., 2000, 122, 11983–11994; (b) E. Ruiz, D. R.
Salahub and A. Vela, J. Phys. Chem., 1996, 100, 122650–
12276; (c) A. C. Cooper, E. Clot, J. C. Huffman, W.E. Streib, F.
Maseras, O. Eisenstein and K. G. Caulton, J. Am. Chem. Soc.,
1999, 121, 97–106.
17 (a) D. A. Pantazis, J. E. McGrady, M. Besora, F. Maseras and
M. Etienne, Organometallics, 2008, 27, 1128–1134; (b) M.
Etienne, J. E. McGrady and F. Maseras, Coord. Chem. Rev.,
2009, 253, 635–646.
References
1
2
(a) E. M. Beck and M. J. Gaunt in Topics in Current Chemistry:
C-H Activation, Vol. 292. (Eds.: J. Q. Yu, Z. J. Shi, Springer-
Verlag: Berlin, Heidelberg, 2010, pp. 85-121; (b) J.
Yamaguchi, A. D. Yamaguchi and K. Itami, Angew. Chem., Int.
Ed., 2012, 51, 8960−9009; (c) J. Wencel-Delord, T. Dröge, F.
Liu and F. Glorius, Chem. Soc. Rev., 2011, 40, 4740−4761.
See for example: (a) M. Pfeffer, Pure Appl. Chem., 1992, 64
335−342; (b) I. Omae, Coord. Chem. Rev., 2004, 248
,
,
995−1023; (c) A. D. Ryabov, Synthesis, 1985, 233−252; (d) J.
Dupont, M. Pfeffer and J. Spencer, Eur. J. Inorg. Chem., 2001,
1917−1927; (e) R. B. Bedford, Chem. Commun., 2003,
1787−1796; (f) J. Dupont, C. S. Consorti and J. Spencer,
Chem. Rev., 2005, 105, 2527–2571; (g) J. P. Djukic, J. B.
Sortais, L. Barloy and M. Pfeffer, Eur. J. Inorg. Chem., 2009,
817–853; (h) M. Albrecht, Chem. Rev., 2010, 110, 576−623;
(i) L. Cuesta and E. P. Urriolabeitia, Comments Inorg. Chem.,
2012, 33, 55−85.
18 D. A. Pantazis, J. E. McGrady, F. Maseras and M. J. Etienne,
Chem. Theory Comput., 2007, 3, 1329–1336.
19 C. Lepetit, J. Poater, M. E. Alikhani, B. Silvi, Y. Canac, J.
Contreras-Garcia, M. Sola and R. Chauvin, Inorg. Chem.,
2015, 54, 2960–2969.
20 M. Lein, J. A. Harrison, A. J. Nielson, Dalton Trans. 2011
10731–10741.
,
21 W. Yao, O.Eisenstein and R. H. Crabtree, Inorg Chim. Acta,
1997, 254, 105–111.
22 For pincer ligand complexes, see the references in
Angew. Chem., Int. Ed; 2016, 55, 3045–3048.
23 (a) R. G. Miller, R. D. Stauffer, D. R. Fahey and R. Parnell, J.
Am. Chem. Soc., 1970, 92, 1511–1521; (b) J. F. Van Baar, K.
3
(a) X. Chen, K. M. Engle, D.-H. Wang and J.-Q. Yu, Angew.
Chem., Int. Ed., 2009, 48, 5094−5115; (b) T. W. Lyons and M.
S. Sanford, Chem. Rev., 2010, 110, 1147−1169; (c) O.
Baudoin, Chem. Soc. Rev., 2011, 40, 4902−4911; (d) J.
Wencel-Delord, T. Dröge, F. Liu and F. Glorius, Chem. Soc.
Rev., 2011, 40, 4740−4761; (e) A. F. M. Noisier and M. A.
Brimble, Chem. Rev., 2014, 114, 8775−8806. (f) D. E.
Stephens, J. Lakey-Beitia, A. C. Atesin, T. A. Atesin, G. Chavez,
Vrieze and P. J. Stufkens, J. Organometallic Chem., 1975, 85
249–263.
,
24 (a) I. Rozas, I. Alkorta and J. Elguero, J. Am. Chem. Soc., 2000,
122, 11154–11161; (b) W. Koch, G. Frenking, J. Guass, D.
Cremer and J. R. Collins, J. Am. Chem. Soc., 1987, 109, 5917–
5934.
25 T. S. Thakur and G. R. Desiraju, J. Mol. Struct. Theochem.,
2007, 810, 143–154.
26 M. Jablonski, J. Phys. Chem. A, 2015, 119, 4993−5008.
27 M. P. Mitoraj, A. Michalak and T. Ziegler, Organometallics,
2009, 28, 3727−3733.
28 F. Weinhold and C. Landis in Valency and bonding: A Natural
Bond Orbital Donor-Acceptor Perspective, Cambridge
University Press, Cambridge, England, 2005.
H. D. Arman, and O. V. Larionov, Catalysis, 2015,
167−175.
K. M. Engle, T.-S. Mei, M. Wasa and J.-U. Qu, Acc. Chem Res.,
5,
4
5
2012, 45, 788−802.
(a) G. Rouquet and N. Chatani, Angew. Chem., Int. Ed.
52, 11726−11743; and refs cited therein; (b) I. P. Beletskaya
and A. V. Cheprakov, J. Organometallic Chem., 2004, 689
, 2013,
,
4055−4082; (c) M. Albrecht, Chem. Rev., 2010, 110, 576−623.
B. Vabre, F. Deschamps and D. Zargarian, Organometallics,
6
7
2014, 33, 6623−6632.
(a) M. Lein, Coord. Chem. Rev., 2009, 253, 625–634; (b) M.
Brookhart, M. L. H. Green and G. Parkin, Proc. Natl. Acad. Sci.
U.S.A., 2007, 104, 6908–6914; (c) W. Scherer and G. S.
McGrady, Angew. Chem., Int. Ed., 2004, 43, 1782–1806; (d)
A. Vigalok, O. Uzan, L. J. W. Shimon, Y. Ben-David, J. M. L.
Martin and D. Milstein, J. Am. Chem. Soc., 1998, 120, 12539–
29 P. Margl, L. Deng and T. Ziegler, Organometallics, 1998, 17
933–946.
,
30 F. A. Cotton and A. G. Stanislowski, J. Am. Chem. Soc.,
96, 5074–5082.
1974,
31 B. G. Harvey and R. D. Ernst, Eur. J. Inorg. Chem., 2017,
1205–1226.
12544; (e) J. J. Schneider, Angew. Chem., Int. Ed., 1996, 35
,
32 A. H. Pakiari and K. Eskandari, J. Mol. Struct. Theochem., 2007,
806, 1–7.
1068–1076; (f) R. H. Crabtree, Angew. Chem., Int. Ed., 1993,
32, 789–805; (g) M. Brookhart and M. L. H. Green, J.
Organometallic Chem., 1983, 250, 395–408.
D. L. Davies, S. M. A. Donald and S. A. MacGregor, J. Am.
Chem. Soc., 2015, 127, 13754−13755.
A. J. Nielson, J. Chem. Soc. Dalton Trans., 1981, 205-211
33 B. V. Pandivan, P. Deepa and P. Kolandaivel, Phys. Chem.
Chem. Phys., 2015, 17, 27496–27508
.
8
9
34 J. Contreras-García, W. Yang and E. R. Johnson, J. Phys. Chem. A.
2011, 115, 12983−12990.
35 P. L. A. Popelier, Coord. Chem. Rev., 2000, 187, 169–189
.
10 J. A. Harrison, A. J. Nielson, M. A. Sajjad, G. C. Saunders
and P. Schwerdtfeger, Eur. J. Inorg. Chem., 2016, 64–77 and
references therein.
36 J. Cosier and A. M. Glazer, J. Appl. Cryst., 1986, 19, 105-107.
37 L. Palatinus and G. Chapuis, J. Appl. Cryst., 2007, 40, 786–
790.
38 (a) P. Parois, R. I. Cooper and A. L. Thompson, Chem. Cent. J.,
2015, 9, 1–30; (b) R. I. Cooper, A. L. Thompson and D. J., Watkin,
J. Appl. Cryst., 2010, 43, 1100–1107.
39 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys.,
2010, 132, 154104−154119.
40 Gaussian 09, Revision E.01, M. J. Frisch, G. W. Trucks, H. B.
11 M. A. Sajjad, K. E. Christensen, N. H. Rees, P. Schwerdtfeger,
J. A. Harrison and A. J. Nielson, Chem Commun., 2017, 53
4187–4190.
,
12 (a) Z. Ke and T. R. Cundari, Organometallics, 2010, 29, 821–
834; (b) Y. J. Wu, S. Q. Huo, J. F. Gong, X. L. Cui, L. Ding, K.
L. Ding, C. X. Du, Y. H. Liu and M. P. Song, J. Organometallic
Chem., 2001, 637-639, 27–46; (c) J. Milani, N. E. Pridmore, A.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins

Please do not adjust margins
Dalton Transactions
Page 12 of 13
View Article Online
DOI: 10.1039/C7DT03525J
ARTICLE
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G.
Journal Name
Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji,
M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F.
Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N.
Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A.
Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega,
J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C.
Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A.
J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K.
Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.
Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian, Inc.,
Wallingford CT, 2009.
41 K. A. Peterson, D. Figgen, M. Dolg and H. Stoll, J. Chem. Phys.,
2007, 126, 124101–124112.
42 T. H. Dunning Jr., J. Chem. Phys., 1989, 90, 1007–1023.
43 AIMAll (version 16.08.17), T. A. Keith, TK Gristmill Software,
Overland Park KS, USA, 2016. http://aim.tkgristmill.com.
44 E. D. Glendening, J. K. Badenhoop, A. E. Reed, J. E. Carpenter, J.
A. Bohmann, C. M. Morales, C. R. Landis and F. Weinhold, NBO,
version 6.0, Theoretical Chemistry Institute, University of
Wisconsin, Madison, WI, 2013. http://nbo6.chem.wisc.edu/.
12 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 13 of 13
Dalton Transactions
View Article Online
DOI: 10.1039/C7DT03525J
Chasing the agostic interaction in ligand assisted cyclometallation
reactions of Palladium (II)
M. Arif Sajjad, Kirsten E. Christensen, Nicholas H. Rees, Peter Schwerdtfeger, John A. Harrison and Alastair J.
Nielson
Computations show that a possible aromatic ring agostic intermediate seen by NMR
spectroscopy in a cyclometallation reaction by palladium (II) can involve C=Cπ electron
density close to the agostic cabon being donated to the metal.
≡