Photoelectron imaging and spectroscopy of MI2- (M = Cs, Cu, Au): Evolution from ionic to covalent bonding

Article abstract of DOI:10.1021/jp103173d

We report a combined experimental and theoretical investigation of MI ₂^- (M = Cs, Cu, Ag, Au) to explore the chemical bonding in the group IA and IB diiodide complexes. Both photoelectron imaging and low-temperature photoelectron spectroscopy are applied to MI₂ ^- (M = Cs, Cu, Au), yielding vibrationally resolved spectra for CuI₂^- and AuI₂^- and accurate electron affinities, 4.52 ± 0.02, 4.256 ± 0.010, and 4.226 ± 0.010 eV for CsI₂, CuI₂, and AuI₂, respectively. Spin-orbit coupling is found to be important in all the diiodide complexes and ab initio calculations including spin-orbit effects allow quantitative assignments of the observed photoelectron spectra. A variety of chemical bonding analyses (charge population, bond order, and electron localization functions) have been carried out, revealing a gradual transition from the expected ionic behavior in CsI₂^- to relatively strong covalent bonding in AuI₂^-. Both relativistic effects and electron correlation are shown to enhance the covalency in the gold diiodide complex.

Full text of DOI:10.1021/jp103173d

1
1244
J. Phys. Chem. A 2010, 114, 11244–11251
-
Photoelectron Imaging and Spectroscopy of MI (M ) Cs, Cu, Au): Evolution from Ionic
2
†
to Covalent Bonding
Yi-Lei Wang, Xue-Bin Wang, Xiao-Peng Xing,§,| Fan Wei, Jun Li,* and
‡
§,|
‡
,‡
,
⊥
Lai-Sheng Wang*
Department of Chemistry & Key Laboratory of Organic Optoelectronics and Molecular Engineering of
Ministry of Education, Tsinghua UniVersity, Beijing 100084, China, Department of Physics, Washington State
UniVersity, 2710 UniVersity DriVe, Richland, Washington 99354, Chemical & Materials Sciences DiVision,
Pacific Northwest National Laboratory, MS K8-88, P.O. Box 999, Richland, Washington 99352, and
Department of Chemistry, Brown UniVersity, ProVidence, Rhode Island 02912
ReceiVed: April 8, 2010; ReVised Manuscript ReceiVed: April 28, 2010
-
We report a combined experimental and theoretical investigation of MI
2
(M ) Cs, Cu, Ag, Au) to explore
the chemical bonding in the group IA and IB diiodide complexes. Both photoelectron imaging and low-
-
temperature photoelectron spectroscopy are applied to MI
2
(M ) Cs, Cu, Au), yielding vibrationally resolved
and accurate electron affinities, 4.52 ( 0.02, 4.256 ( 0.010, and 4.226 ( 0.010
, respectively. Spin-orbit coupling is found to be important in all the diiodide
-
-
spectra for CuI
eV for CsI , CuI
2
and AuI
2
2
2
, and AuI
2
complexes and ab initio calculations including spin-orbit effects allow quantitative assignments of the observed
photoelectron spectra. A variety of chemical bonding analyses (charge population, bond order, and electron
localization functions) have been carried out, revealing a gradual transition from the expected ionic behavior
-
-
in CsI
2 2
to relatively strong covalent bonding in AuI . Both relativistic effects and electron correlation are
shown to enhance the covalency in the gold diiodide complex.
number of theoretical studies for detailed spectral assignments.1
8-20
I. Introduction
The importance of spin-orbit (SO) coupling for the assignment
The chemical bond is the most important concept in modern
chemistry. Traditionally chemical bonds are classified as
covalent or ionic in molecular systems. The group IB halides
are usually considered as ionic compounds. However, because
19
of the PES spectra has been pointed out by Misha et al., whose
calculations including SO effects yielded quantitative agreement
with the experimental data reported in ref 17.
In the current study, we present a joint experimental and
1
of the strong relativistic effects, gold has been found to display
theoretical investigation on the coinage metal diiodide com-
chemistry very different from its lighter congeners.2,3 Specifi-
cally, significant covalent bonding character has been increas-
-
plexes, MI
2
(M ) Cu, Ag, Au), emphasizing on the evolution
-
of chemical bonding from Cu to Au. The ionic CsI
2
complex
4
ingly found in gold-containing compounds, from the well-
is included as a reference. Both photoelectron imaging and low-
-
5-7
known Au(CN)
2
complex
to the simple gold oxide and
-
temperature PES data have been obtained for MI
2
(M ) Cs,
8
sulfide molecules. A question arises if the gold halide molecules
may in fact contain covalent characters. Because iodine pos-
sesses the lowest electron affinity among the halide elements
and it is very close to that of gold, we suspect that significant
covalent bonding may exist between Au and I. Therefore, in
-
Cu, Au). The AgI
2
complex could not be studied experimen-
tally using the electrospray source due to the low solubility of
AgI. Vibrationally resolved PES spectra have been obtained for
-
-
CuI
2 2
and AuI at 266 nm, as well as rich electronic structure
information at 193 nm using vibrationally cold anions. Density
the current article we investigate the evolution of chemical
functional theory (DFT) and ab initio calculations are carried
-
bonding in the coinage iodide complexes, MI
2
(M ) Cu, Ag,
-
out on the four MI
2
species to help spectral interpretation and
-
Au) using the more ionic CsI
2
complex as a reference.
chemical bonding analyses. As expected, SO coupling effects
are found to be important for all the diiodide complexes. Our
theoretical analyses reveal typical ionic bonding behavior in
9
Gold halides form interesting solid materials, and the electronic
structures of gaseous gold halide molecules have been investigated
1
0-12
in a number of previous computational studies,
in particular,
-
-
-
CsI
CuI
2
and significant covalent character in AuI
2
, while that in
1
3-16
regarding the relativistic effects in these molecules.
The
(X ) Cl, Br, I) have been
investigated previously using photoelectron spectroscopy (PES)
-
2
and AgI is found to be between the ionic and covalent
2
-
gold dihalide complexes AuX
2
bonding.
1
7
and mass spectrometry, with an emphasis on the production of
divalent gold(II) species, AuX , upon electron detachment from
II. Experimental Methods
2
-
The MI
2
(M ) Cu, Au, Cs) complexes were produced using
-
-
the AuX
2
anions. The PES data on AuX
2
have stimulated a
electrospray ionization (ESI), and the photodetachment experi-
ments were performed using two methods: photoelectron
imaging21 and low-temperature PES with a magnetic-bottle
†
Part of the “Klaus M u¨ ller-Dethlefs Festschrift”.
To whom correspondence should be addressed, junli@tsinghua.edu.cn
*
22
-
and Lai-Sheng_Wang@brown.edu.
photoelectron analyzer. The AgI
complex was omitted in
2
‡
Tsinghua University.
Washington State University.
Pacific Northwest National Laboratory.
Brown University.
the experiment due to the low solubility of AgI.
A. Photoelectron Imaging. The photoelectron imaging
experiment was carried out on a room temperature ESI-PES
§
|
⊥
1
0.1021/jp103173d  2010 American Chemical Society
Published on Web 05/12/2010

Investigation of MI2-
J. Phys. Chem. A, Vol. 114, No. 42, 2010 11245
apparatus23 by replacing the original magnetic-bottle photoelec-
tron analyzer with a velocity-map imaging system, details of
which have been described elsewhere. About 1 mM solutions
of CsI and AuI in water/methanol mixed solvents (1/3 volume
for completeness, even though no experimental data were
presented. The scalar relativistic (SR) and SO coupling effects
were taken into account by the zero-order-regular approximation
(ZORA).28 Geometries were optimized at the SR-ZORA level
and single-point energy calculations were performed with
inclusion of the SO effects. Vibrational frequency calculations
were carried out to verify that the species were minima on the
potential energy surface. The B3LYP hybrid functional was also
2
1
ratio) or CuI in pure acetonitrile were used as electrospray
-
solutions to generate the respective MI
2
complexes. Anions
from the ESI source were directed into an ion trap via a rf-only
ion guide. After a trapping time of 0.1 s, the ions were pulsed
out into a time-of-flight mass spectrometer and directed into
the center of the photoelectron imaging system, where they were
detached by a linearly polarized laser beam at 266 nm. The
nascent electron cloud was accelerated by a high voltage pulse
applied to the imaging electrodes and was projected onto a
phosphor screen behind a set of microchannel plates. The
positions of the photoelectrons on the phosphor screen were
recorded by a CCD camera and accumulated as a photoelectron
-
used to optimize the MI2 anions with the Gaussian 03
program.29 The basis sets used are the same as in the ab initio
MOLPRO calculations described below.
To compare with the experimental PES spectra, we performed
-
high-level ab initio calculations for MI
2
(M ) Cs, Cu, Ag,
30
Au) using the MOLPRO 2008 program. In these calculations
we used both the CCSD(T) (coupled-cluster with single and
31
double and perturbative triple excitations) and CASSCF
complete-active-space self-consistent field) methods. The
5
image. The usual accumulation time was about 10 laser shots
32
(
at a 10 Hz repetition rate. Since the laser polarization was
parallel to the surface of the microchannel plates and the
phosphor screen detector, the original shape of the three-
dimensional electron clouds could be rebuilt through inverse
Abel transformation from the recorded two-dimensional images,
which were performed using the BASEX program.24 The
electron binding energy spectra and photoelectron angular
distributions were obtained from the rebuilt three-dimensional
images. The electron kinetic energy resolution of our imaging
system is about 2.5% for 1 eV electrons, as calibrated from the
-
geometries of MI
2
were optimized with CCSD(T); single-point
energies of the ground and excited states of the neutrals were
calculated at the geometry of the anions at the same level. The
electron binding energies corresponding to one-electron transi-
-
tions from the closed-shell ground states of the MI
the ground and excited states of the MI
using a CASSCF/CCSD(T)/SO approach that was used previ-
2
anions to
2
neutrals were obtained
-
7
ously for M(CN)
2
(M ) Cu, Ag, Au). In this approach the
SO splittings were treated as a perturbation to the scalar-
relativistic state energies and were calculated on the basis of
CASSCF wave functions with the diagonal matrix elements
replaced by the CCSD(T) energies. The SO coupling effect was
included by using a state-interacting method33 with SO pseudo-
potentials. The valence orbitals, the 6s orbital for Cs, 3d and
-
- 21
2
66 and 355 nm spectra of Br and I .
B. Low-Temperature Photoelectron Spectroscopy. The
low-temperature PES experiment was carried out using a
magnetic-bottle apparatus coupled to an ESI source and a
22
cryogenically controlled ion trap. Anions from the ESI source
were guided by a series of rf-only quadrupole devices into a
Paul trap, which is connected to the coldfinger of a closed cycle
He refrigerator (10-350 K). The ions were accumulated and
4
s for Cu, 4d and 5s for Ag, 5d and 6s for Au, and 5p for I,
were included in the active space during the CASSCF calcula-
tions. As a result, we correlated 11 valence electrons in 7 active
orbitals, denoted as CASSCF(11,7) for CsI
for MI (M ) Cu, Ag, Au). The basis sets used were
12s11p5d3f2g)/[8s8p5d3f2g] for Cs, cc-pVTZ-PP for Cu, Ag,
2
and CASSCF(21,12)
cooled in the ion trap before being pulsed into the extraction
-
2
zone of a time-of-flight mass spectrometer. The MI
2
species
34
(
were mass-selected and decelerated before being detached by
93 nm photons (6.424 eV) from an ArF excimer laser or 266
3
5
36
and Au, and aug-cc-pVTZ-PP for I, respectively. The
effective core potentials with 46, 10, 28, 60, and 28 core
electrons were used for Cs, Cu, Ag, Au, and I, respectively.
For CuI
the calculated VDEs related to detaching electrons from the
Cu 3d δ orbital. Considering the necessity of balancing the
1
nm photons (4.661 eV) from a Nd:YAG laser in the interaction
zone of a magnetic-bottle photoelectron analyzer operated at a
34,36,37
-
2
, we noted larger than usual deviations (∼0.2 eV) of
20 Hz repetition rate with the ion beam on and off on alternating
laser shots for background subtraction. Photoelectrons were
collected at near 100% efficiency by the magnetic bottle and
analyzed in a 5.2 m long electron flight tube. Time-of-flight
photoelectron spectra were collected and converted to kinetic
g
1
0-x 1+x
3
3
d
4s configurations involving the highly contracted Cu
d orbitals,38 a larger basis set with diffuse functions,
35,37
-
aug-cc-pVTZ-PP,
was used for Cu in the CCSD(T) single-
energy spectra calibrated with the known spectra of I and
-
point energy calculations, which led to excellent results.
ClO
2
. The electron binding energy spectra reported were
obtained by subtracting the kinetic energy spectra from the
respective detachment photon energies. The kinetic energy
resolution (∆E/E) of the magnetic-bottle electron analyzer was
IV. Experimental Results
-
Figure 1 shows the photoelectron images of MI
Au, Cs) recorded at 266 nm and room temperature. The
photoelectron images for CuI
displaying one spectral band with similar electron kinetic
energies and angular distributions. The image of CsI is very
2
different, displaying much lower electron kinetic energies with
isotropic angular distributions. The different angular distributions
2
(M ) Cu,
2
% (i.e., 20 meV for 1 eV electrons).
-
-
2
and AuI
2
are similar, both
III. Computational Methods
-
The theoretical investigations were performed using both DFT
and wave function-based ab initio methods. The DFT calcula-
tions were performed using the generalized gradient approxima-
tion (GGA) with the PBE exchange-correlation functional25
implemented in the Amsterdam Density Functional (ADF
between the photoelectron images of the coinage diiodides and
-
CsI
2
suggest that they possess different electronic structures
2
009.01) program.26 The uncontracted Slater basis sets with the
and different highest occupied molecular orbitals (HOMOs)
from which an electron is detached. The electron binding energy
spectra obtained from the images display vibrational fine features
quality of triple-ꢀ plus two polarization functions (TZ2P) were
2
10
used, with the frozen core approximation applied to [1s -4d ]
2
6
2
6
2
14
-
for Cs and I, [1s -3p ] for Cu, [1s -4p ] for Ag, and [1s -4f ]
with a rather broad Franck-Condon envelope for CuI
2
and
for Au, respectively.27 The AgI
-
complex was also calculated
-
-
2
2 2
AuI . The electron binding energies of CsI are much higher.

1
1246 J. Phys. Chem. A, Vol. 114, No. 42, 2010
Wang et al.
-
Figure 2. Photoelectron spectra of MI
2
(M ) Cs, Cu, Au) at 12 K
-
Figure 1. Photoelectron images of MI
(
2
(M ) Cs, Cu, Au) at 266 nm
and 193 nm. The short bars represent the positions of the calculated
vertical detachment transitions.
inset), photoelectron spectra obtained from the images (gray curves),
and 266 nm photoelectron spectra at 12 K (black curves). The double
arrow represents the laser polarization. The vertical dot lines indicate
-
-
the 0-0 transitions resolved in the 12 K spectra for CuI
2
and AuI
2
-
-
CuI
2
and AuI
2
(Table 1), respectively, from the clearly
and adiabatic detachment energy estimated from the 266 nm spectrum
-
observed 0-0 transitions. The ADE for CsI
from the onset of the 12 K spectrum to be 4.52 ( 0.02 eV
(Table 1).
The 193 nm spectra for MI
shown in Figure 2. Eight well-resolved peaks, labeled X and
2
was estimated
-
of CsI
2
at 12 K (not shown).
TABLE 1: Experimental Adiabatic Detachment Energies
(M ) Cs, Cu, Au) and the Observed
Totally Symmetric Vibrational Frequencies for the Ground
-
-
(
ADEs) for MI
2
(M ) Cu, Au, Cs) at 12 K are
2
State of CuI
2
and AuI
2
-
A-G, were observed for CuI
2
, whereas six peaks (X and A-E)
ADE (eV)a
Vib. Freq. (cm-1
)
spanning roughly the same energy range were resolved for
-
-
AuI
2
2
. Six peaks (X and A-E) were observed for CsI , with
-
CsI
CuI
AuI
2
4.52 ( 0.02
4.256 ( 0.010
4.226 ( 0.010
-
-
the first three peaks (X, A, B) very closely spaced. The vertical
detachment energies (VDEs) for all observed peaks are given
in Table 2 and compared with theoretical calculations. The MI
complexes are superhalogens, all with extremely high ADEs.
The current 193 nm spectrum for AuI
resolved than the previous spectrum taken at room temperature.
The current ADE of 4.226 ( 0.010 eV for AuI
266 nm and 12 K is much improved relative to the value of
2
160 ( 15
180 ( 15
b
2
-
2
a
Also represent the electron affinities for the corresponding
neutral MI species. A less accurate value of 4.18 ( 0.07 was
39
b
2
-
2
at 12 K is better
reported previously (ref 17).
17
-
2
measured at
In addition to the broad ground state band, the onset of a second
-
band is observed in the 266 nm spectrum of CsI
2
.
-
4.18 ( 0.07 eV measured previously at 193 nm and room
We have also obtained the PES spectra at 12 K for MI
M ) Cu, Au, Cs) using the low-temperature magnetic-bottle
2
temperature.17 The spectra of CuI
-
and CsI
-
have not been
2
2
(
22
-
-
reported previously. However, they show some similarity to the
PES spectra of CuBr
apparatus. The 266 nm spectra at 12 K for CuI
2
and AuI
2
-
-
40,41
2
2
and NaI , respectively.
are superimposed on the spectra obtained with the imaging
method at room temperature in Figure 1 (black thick curves).
Because of the spectral cutoff associated with the magnetic-
V. Theoretical Results
We have optimized the structures and found that the MI
(M ) Cs, Cu, Ag, Au) anions are all linear with a Σ
-
bottle apparatus, only a rising tail with an onset around 4.5 eV
2
-
1
+
was obtained for CsI
2
at 266 nm at 12 K (not shown). The
g
ground
effects of vibrational cooling at 12 K were dramatic. Due to
the elimination of vibrational hot bands, the 12 K spectra were
significantly sharpened. In particular, a single vibration progres-
state and D∞h symmetry, consistent with previous calculations.
These structures are confirmed to be minima, and the results of
the vibrational analysis at the PBE level are given in Table 3.
Also given in Table 3 are the optimized PBE, CCSD(T), and
B3LYP bond distances and normalized vibrational frequencies
(NVF) to evaluate the I-M-I bonding strengths. The Au-I
distance at the CCSD(T) level is in excellent agreement with a
previous calculation at a similar level of theory.19 We have also
computed the vibrational frequencies for the neutral final states
-
-
sion was clearly revealed for CuI
2 2
and AuI with frequencies
-1
of 160 ( 15 and 180 ( 15 cm , respectively (Table 1). Because
of the elimination of vibrational hot bands, the adiabatic
detachment energy (ADE) for the ground state transition or the
electron affinity of the corresponding neutral can be determined
accurately, as marked by the dashed line in Figure 1. The
obtained values are 4.256 ( 0.010 and 4.226 ( 0.010 eV for
2 2
for AuI and CuI , and the calculated symmetric stretching

Investigation of MI2-
J. Phys. Chem. A, Vol. 114, No. 42, 2010 11247
TABLE 2: Observed and Calculated Vertical Detachment
TABLE 3: Calculated PBE, CCSD(T), and B3LYP Bond
-
-
Energies (VDEs) for MI
2
(M ) Cs, Cu, Ag, Au) and
Lengths and PBE Vibrational Frequencies for MI
Cs, Cu, Ag, Au)
2
(M )
Final-State Assignments for M ) Cs, Cu, Au
VDE (eV)
expta
calcd
vibrational
frequency (cm-1) NVF (cm
)
-1 a
R(M-I) (Å)
feature
final state
CCSD(T) B3LYP PBE
σ
u
σ
g
π
u
σ
u
σ
g
π
u
-
2
2
4.55 4.55 (4.63)b
4.63 4.68
4.72 4.72
4.89 4.92
5.55 5.49
CsI
2
X
A
B
C
D
E
X
A
B
C
D
E
73% [ Σ1/2u], 27% [ Π1/2u]
2
-
-
100% [ Π3/2u
]
CsI
2
3.563
2.430
2.622
2.575
3.620 3.585 96 63
2.458 2.482 255 125 (138) 62 172 125 42
2.660 2.679 196 113 48 162 113 39
2.619 2.611 196 141 (159) 52 196 141 52
13 85 63 11
2
b
100% [ Π3/2 g
]
CuI
AgI
AuI
2
2
2
-
-
52% [ Π1/2 g], 48% [ Σ1/2 g
]
]
2
2
2
2
b
73% [ Π1/2u], 27% [ Σ1/2u
]
2
2
52% [ Σ1/2 g], 48% [ Π1/2 g
5.77 5.68
4.29 4.29 (4.47)
a
-
2
b
The normalized vibrational frequencies (NVF) were computed
by assuming that all the metal atoms have the same mass as Au.
The frequencies for the neutrals are listed in parentheses. The
geometry optimizations and frequency analysis are carried out with
CuI
AgI
AuI
2
100% [ Π3/2 g
]
2
2
98% [ Π1/2 g ], 2% [ Σ1/2 g
]
4.73 4.76
5.02 5.03
5.16 5.12
5.33 5.29
5.65 5.55
5.84 5.71
6.24 6.09
6.28
b
2
2
99% [ Σ1/2 g], 1% [ Π1/2 g
]
2
100% [ Π3/2u
]
2
2
spin-orbit coupling.
52% [ Π1/2u], 48% [ Σ1/2u
]
2
100% [ ∆5/2 g
]
2
2
F
G
98% [ ∆3/2 g], 2% [ Π3/2 g
]
]
2
2
52% [ Σ1/2u], 48% [ Π1/2u
2
2
9
9
9
8% [ Π3/2 g], 2% [ ∆3/2 g]
2 2
7% [ Π1/2 g], 3% [ Σ1/2 g
]
]
6.40
7.44
4.78
5.23
5.27
5.40
5.86
6.21
7.76
8.01
8.37
8.48
2
2
6% [ Σ1/2 g], 4% [ Π1/2 g
-
2
2
100% [ Π3/2 g
]
2
2
9
1
5
9
5
1
5
5
9
9
1% [ Π1/2 g], 9% [ Σ1/2 g
]
2
00% [ Π3/2u
]
2
2
2% [ Σ1/2u], 48% [ Π1/2u
]
]
]
2
2
1% [ Σ1/2 g], 9% [ Π1/2 g
2
2
2% [ Π1/2u], 48% [ Σ1/2u
2
00% [ ∆5/2 g
]
2
2
7% [ ∆3/2 g], 43% [ Π3/2 g
]
]
2
2
7% [ Π3/2 g], 43% [ ∆3/2 g
2
2
7% [ Π1/2 g], 3% [ Σ1/2 g
]
]
2
2
7% [ Σ1/2 g], 3% [ Π1/2 g
8.87
-
2
2
b
2
X
A
B
C
D
E
99% [ Π3/2 g], 1% [ ∆3/2 g
]
4.28 4.28 (4.51)
2
2
94% [ Π1/2 g], 6% [ Σ1/2 g
]
4.91 4.84
5.20 5.21
5.43 5.40
5.72 5.73
6.28 6.15
6.16
2
100% [ Π3/2u
]
2
2
52% [ Σ1/2u], 48% [ Π1/2u
]
]
2
2
91% [ Σ1/2 g], 9% [ Π1/2 g
2
2
52% [ Π1/2u], 48% [ Σ1/2u
]
2
1
5
5
8
9
00% [ ∆5/2 g
]
2
2
2% [ Π3/2 g], 48% [ ∆3/2 g
]
]
]
6.93
7.89
8.01
8.82
Figure 3. Orbital interactions from fragment MO analysis for MI -
2
2
1% [ ∆3/2 g], 49% [ Π3/2 g
2
2
2
(M ) Cs, Au). The highest singly or doubly occupied orbitals are
7% [ Π1/2 g], 13% [ Σ1/2 g
-
2
2
labeled with arrows. The orbitals of the I · · · I fragment have one-
0% [ Σ1/2 g], 10% [ Π1/2 g
]
electron vacancy (illustrated with a diagonal grain pattern). The I-I
a
b
-
-
The uncertainty for the measured VDEs is (0.02 eV. The first
distance in the I· · ·I fragment is taken as the same as in the CsI
2
-
vertical electron detachment energies (VDE
Au) were aligned with the experimental data for easier comparison
for the higher VDEs. The theoretical VDE given in parentheses
1
) of MI
2
(M ) Cs, Cu,
anion. The energy levels are shifted to align all the nonbonding orbitals
for the convenience of examining orbital interactions without the
electrostatic effects. The 5s and 5p orbitals of Cs and Au are not
counted.
1
were computed without including SO effects for the neutral and
anionic ground states. The SO effects were included in the calcula-
tions of excitation energies for the neutral species. The higher
VDEs were obtained by adding the experimental first VDE to the
neutral excitation energies calculated at the geometry of the anions.
We present the energy levels of the Kohn-Sham orbitals in
Figure 3, which provides a qualitative understanding of the
-
orbital interactions between M and I
2
and demonstrates the
-
-
ionic bonding in CsI
2
2
versus the covalent bonding in AuI .
frequencies are in good agreement with the experimental
observations for CuI
in Table 2 for MI
The isocontour surfaces of the Kohn-Sham molecular orbitals
2
and AuI
2
. The calculated VDEs are given
-
-
of MI
splittings due to SO coupling are shown in Figure 5; the σ
(or σ and π ) MOs can mix via SO interactions. The atomic
2
(M ) Cs, Cu, Ag, Au) are shown in Figure 4. The
2
(M ) Cs, Cu, Ag, Au). The first VDE in
g
and
each case was calculated as the total energy difference between
the neutral and anionic ground states at the anion geometry.
VDEs for the excited states were computed by adding the first
VDE to the excitation energies of the neutral species at the anion
geometry. The calculated first VDE, given in parentheses in
Table 2, was overestimated by ∼0.1-0.2 eV relative to the
π
g
u
u
charges and M-I bond orders have been analyzed using several
different theoretical formalisms, as presented in Table 4. Figure
-
6
(
displays the electron localization functions (ELF) of MI
2
42
M ) Cs, Cu, Ag, Au).
experimental values. To facilitate comparison with the experi-
VI. Comparison between Experiment and Theory
-
ment in the cases of MI
2
(M ) Cs, Cu, Au), the first VDE
-
was aligned with the experimental value and the higher VDEs
were shifted by the differences between the calculated and
experimental first VDE, as given in Table 2 and plotted in Figure
The electronic structures of MI
2
can be understood from
-
orbital interactions or charge transfer between M and I
to the familiar valent MO patterns of the diatomic molecules,
such as F or I , the 2σ , 2σ , 1π , 1π valent manifold formed
2
. Similar
2
(vertical bars).
2
2
g
u
g
u

1
1248 J. Phys. Chem. A, Vol. 114, No. 42, 2010
Wang et al.
Figure 4. The isocontour surfaces of the Kahn-Sham orbitals for the
-
MI
2
complexes (isocontour ) 0.03 au). The 5s and 5p orbitals of Cs
and Au are not counted.
from the I 5p orbitals are still present in the MI -
systems and
2
-
and AuI -
are split according to the nature of the I · · ·(M)· · ·I interactions.
Figure 5. The SO splittings of the SR-ZORA MOs of CsI
2
2
.
The 5s and 5p orbitals of Cs and Au are not counted, consistent with
Figures 3 and 4.
As shown in Figure 3, when Cs is inserted between the two
-
iodine atoms, the ionic interactions between Cs and I
2
primarily
lead to charge transfer from the Cs 6s orbital to the ligands and
the I 5p manifold is only split by 0.9 eV. However, when Au
is inserted between the two I atoms, the Au 6s and 5dz2 orbitals
TABLE 4: Theoretical Atomic Net Charges on M and M-I
-
Bond Orders in MI
2
(M ) Cs, Cu, Ag, Au)
Charge Population
interact with the σ
stabilization of 2σ
Similarly, the Au 5dπ orbitals interact with the π
leading to the stabilization of 1π and destabilization of 2π
g
ligand orbital, giving rise to significant
and destabilization of 4σ (Figure 3).
ligand orbitals,
NPA
Hirshfeld
Voronoi
MDC-q
AIM
g
g
g
Cs
0.95
0.46
0.48
0.24
0.40
-0.01
0.03
0.43
-0.03
0.05
0.65
-0.08
-0.03
-0.18
0.83
0.31
0.23
g
g
.
Cu
Ag
Au
The significant orbital interactions between the Au s-d hybrids
and the I 5p manifold are clear from the MO isocontour surfaces
shown in Figure 4. Consequently, the I 5p manifold is split by
about 4 eV in energy (Figure 3).
-0.10
-0.10
-0.08
Bond Order
Wiberg
Mayer
G-J
N-M(3)
Figure 5 shows that there are significant SO effects for the I
5
p and Au 5d orbitals, where the splittings of the spinors
Cs-I
0.05
0.43
0.41
0.51
0.33
0.88
0.77
0.79
0.19
0.50
0.49
0.55
1.05
1.00
0.99
0.99
qualitatively reveal the respective contributions of the I 5p and
Au 5d orbitals (Figure 4). The energy levels of the spinors for
CsI
Cu-I
Ag-I
Au-I
-
2
are in qualitative agreement with the PES spectral pattern.
The assignments of the PES spectra based on the SO-split states
are given in Table 2; the calculated VDEs are also compared
with the experimental spectra as short vertical bars in Figure 2.
those by Misha et al. reported previously with the inclusion of
19
-
SO coupling. We see that the HOMO of CsI
of the I ligands, whereas in the three coinage diiodides the
HOMO is the π orbital of the I ligands with significant
antibonding interaction with the metal d orbital. The different
2
u
is the σ orbital
Overall, the calculated VDE patterns are in excellent agreement
2
-
with the experimental data for all three complexes, MI
2
(M )
g
2
Cs, Cu, Au). The slightly inferior agreement for the high-lying
excited states is a result of the limited active space employed
-
HOMOs for CsI
2
and the coinage diiodides are consistent with
during the expensive CASSCF calculations. The current as-
the different angular distributions for the ground-state transitions
observed in the photoelectron images (Figure 1). Further
-
signments and calculated VDEs for AuI
2
are consistent with

Investigation of MI2-
J. Phys. Chem. A, Vol. 114, No. 42, 2010 11249
bond lengths in the coinage diiodide complexes are within the
range of the M-I covalent single-bond lengths (2.45, 2.61, and
2
.57 Å for Cu, Ag, Au, respectively) estimated by the self-
45
consistent covalent radii derived by Pyykk o¨ . The optimized
Au-I distance of 2.58 Å at CCSD((T) level is consistent with
-
the experimental distances in [AuI
(
2
]
units found in crystals
2.529, 2.5604, and 2.7713 Å).4
6-48
The PBE-optimized Cs-I
distance is 0.06 Å shorter than the sum of the covalent radii.
The M-I bond strengths can be measured by the vibrational
frequencies of the totally symmetric stretching modes because
it is not influenced by the mass of the central metal atom. Indeed,
-
-
g 2 2
the σ vibrational frequency of AuI is larger than that of CsI ,
indicating that the Au-I bonds are much stronger than the Cs-I
bonds. Considering the larger mass of Au than that of Ag, Cu,
and Cs, we also calculated the normalized vibrational frequen-
cies (NVF) by assuming that all the metal atoms have the same
mass as Au. Both the σ
g
vibrational frequencies and NVFs
indicate that the M-I bond strength increases as Cs-I , Ag-I
<
Cu-I < Au-I, which agrees well with the calculated Wiberg
bond orders listed in Table 3. Note the above analyses nicely
confirm the experimental observation, which shows that the
-
symmetric stretching frequency of AuI
2
is slightly higher than
-
that of CuI
2
.
-
The charge population and bond order analyses for MI
2
(
M ) Cs, Cu, Ag, Au) given in Table 4 reveal that all the
methods predict much less positive charge on the group
IB-metal atoms than on Cs, indicating that Cu, Ag, and Au
-
have covalent character in contrast to the ionic bonding in CsI
2
.
-
The AuI
2
complex is especially interesting: All the charge
population formalisms except natural population analysis (NPA)
predict negative charges on the Au atom. This slightly negative
-
charge on Au is consistent with the strong s-d hybridization
Figure 6. The electron localization functions (ELFs) for MI
Cs, Cu, Ag, Au) calculated with the PBE functional.
2
(M )
-
of Au and the much stronger covalent bonding in the AuI
2
complex than in its lighter congeners. Except the N-M(3) bond
order, which includes ionic interactions, all the calculated bond
orders are larger for M-I (M ) Cu, Ag, Au) than for the Cs-I
bond, indicating that the Cs-I bond is ionic whereas the coinage
complexes have more covalent character, with Au-I having
the strongest covalent bonding.
examination shows that the orbital orderings of CuI -
are also different. In particular, the 4d σ and δ MOs in CuI
are higher lying, whereas they are relatively lower lying in
and AuI
-
2
2
-
g
g
2
-
-
AuI
2
. Therefore, detachment from the δ
g
MO in CuI
2
was
observed at 193 nm, corresponding to the bands E and F in
The electron localization functions (ELFs) shown in Figure
-
-
Figure 2. But the δ
g
MO of AuI
2
possesses higher electron
6
reflect the electron pair density in the MI
2
complexes, which
binding energies and was not accessible using the 193 nm
provide further insight into the nature of the bonding between
M and I. In covalent bonding the forces binding the nuclei are
exerted by an increase in the charge density shared mutually
between the two atoms, whereas in ionic bonding the two atoms
are bound by increased cationic or anionic charges localized in
the region of each nucleus. As shown in Figure 6, there is no
electron pair density between the Cs and I atoms, as expected
from their ionic bonding. Significant electron pair density exists
between the M-I (M ) Cu, Ag, Au) bonds, increasing from
Cu to Au and suggesting relatively strong covalent character.
19,43,44
detachment laser. As pointed out previously,
SO coupling
is very important to be included for Au- and I-containing
compounds in order to achieve good agreement with experi-
mental data.
VI. Discussion
1
0
The group IA elements possess a ns (n - 1)d configuration
10
1
compared to (n - 1)d ns for the group IB elements. The filled
n - 1)d shell gives rise to distinct chemistry and electronic
10
(
structures for group IB compounds through s-d hybridization,
relative to the predominantly ionic bonding in group IA
compounds. Particularly, the strong relativistic effects in gold
The different bonding nature between the groups IA and IB
compounds can also be glimpsed from the orbital interaction
-
1
analysis (Figure 3). In CsI
2
, the Cs 6s orbital is much higher
1
-
enhance the s-d hybridization, resulting in significant σ-co-
in energy than the frontier 5p orbitals of I · · ·I , resulting in an
electron transfer from Cs to I · · ·I and the formation of an ionic
compound [(I )Cs (I )]. Their orbital interactions are weak, and
the energy levels of the I 5p manifold change very little from
2 2
the fragments to the CsI complex. The AuI complex is quite
different. Because the energy of the Au 6s orbital is much closer
to that of the frontier 5p orbitals of I · · ·I , the strong orbital
-
-
valent bonding between Au and I in AuI
2
.
-
-
+ -
The M-I bond lengths in CsI
2
are much longer than those
by about 1 Å (Table 3). In the group IB complexes,
-
in AuI
2
-
-
the M-I bond lengths increase from Cu to Ag and then decrease
from Ag to Au due to the relativistic effects in Au. The PBE
M-I bond lengths increase by 0.20 Å from Cu to Ag but
decrease by 0.07 Å from Ag to Au. The fact that the M-I bond
lengths of the group IB complexes are smaller than that of Cs
by about 1 Å is consistent with the atomic radii. All the M-I
1
-
interactions lower the occupied 2σ
cupied 4σ
of Au can hybridize with the 6s orbital, promoting orbital
g
orbital and lift the unoc-
orbital significantly. The directional 5dz2 orbitals
g

1
1250 J. Phys. Chem. A, Vol. 114, No. 42, 2010
Wang et al.
overlap to form the Au-I covalent bonds. The 3d orbitals
of Cu are more tightly contracted in the radial distribution
while the energy gap between Ag 4d and 5s is considerably
large, thus the s-d hybridization is relatively unfavorable
for Cu and Ag than for Au, as confirmed by the charge
(5) Dietz, O.; Ray o´ n, V. M.; Frenking, G. Inorg. Chem. 2003, 42, 4977.
(
6) Zaleski-Eigierd, P.; Patzschke, M.; Pyykk o¨ , P. J. Chem. Phys. 2008,
28, 224303.
7) Wang, X. B.; Wang, Y. L.; Yang, J.; Xing, X. P.; Li, J.; Wang,
1
(
L. S. J. Am. Chem. Soc. 2009, 131, 16368.
(8) Zhai, H. J.; B u¨ rgel, C.; Bonacic-Koutecky, V.; Wang, L. S. J. Am.
Chem. Soc. 2008, 130, 9156.
population, bond order, and ELF analyses. Therefore, the
(
9) S o¨ hnel, T.; Hermann, H.; Schwerdtfeger, P. J. Phys. Chem. B 2005,
-
series of MI
2
(M ) Cs, Cu, Ag, Au) complexes investigated
1
09, 526.
-
here provide a gradual transition from ionic bonding in CsI
to covalent bonding in AuI
It is also interesting to compare the covalent bonding in AuI
with that of AuO
2
(10) Schulz, A.; Hargittai, M. Chem.sEur. J. 2001, 7, 3657.
(11) S o¨ hnel, T.; Brown, R.; Kloo, L.; Schwerdtfeger, P. Chem.sEur.
J. 2001, 7, 3167.
12) Schwerdtfeger, P.; Krawczyk, R. P.; Hammerl, A.; Brown, R. Inorg.
Chem. 2004, 43, 6707.
(13) Schwerdtfeger, P.; McFeaters, J. S.; Liddell, M. J.; Hru sˇ a´ k, J.;
Schwarz, H. J. Chem. Phys. 1995, 103, 245.
(14) Seth, M.; Cooke, F.; Schwerdtfeger, P.; Heully, J. L.; Pelissier, M.
J. Chem. Phys. 1998, 109, 3935.
15) Brown, J. R.; Schwerdtfeger, P.; Schroder, D.; Schwarz, H. J. Am.
Soc. Mass Spectrom. 2002, 13, 485.
-
2
.
-
2
(
-
-
2
or AuS
2
, which were reported recently to
possess linear structures with multiple covalent bonding char-
8
-
-
acter. The valent MOs of AuI
2 2
are similar to those in AuO
-
or AuS
2
. However, the latter is two electrons fewer and the
MO is
antibonding, which cancels out the bonding interaction of the
2
π
g
MO is half-filled. As seen in Figure 4, the 2π
g
(
-
-
(16) Mishra, S. Phys. Chem. Chem. Phys. 2008, 10, 3987.
1π
g
MO. Because the 2π
there is net π bonding in these two molecules, giving rise to
multiple bonding character. However, both the 2π and 1π MOs
and they cancel each other, leaving
g 2 2
MO is half-filled in AuO or AuS ,
(
17) Schr o¨ der, D.; Brown, R.; Schwerdtfeger, P.; Wang, X. B.; Yang,
X.; Wang, L. S.; Schwarz, H. Angew. Chem., Int. Ed. 2003, 42, 311.
18) Dai, B.; Yang, J. Chem. Phys. Lett. 2003, 379, 512.
(19) Mishra, S.; Vallet, V.; Domcke, W. ChemPhysChem 2006, 7, 723.
(20) Mishra, S. J. Phys. Chem. A 2007, 111, 9164.
(21) Xing, X. P.; Wang, X. B.; Wang, L. S. J. Chem. Phys. 2009, 130,
74301.
22) Wang, X. B.; Wang, L. S. ReV. Sci. Instrum. 2008, 79, 073108.
(23) Wang, L. S.; Ding, C. F.; Wang, X. B.; Barlow, S. E. ReV. Sci.
Instrum. 1999, 70, 1957.
24) Dribinski, V.; Ossadtchi, A.; Mandelshtam, V. A.; Reisler, H. ReV.
Sci. Instrum. 2002, 73, 2634.
25) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. ReV. Lett. 1996, 77,
3865.
g
g
(
-
are fully occupied in AuI
only a single σ bond between Au and I. Just like in Au(CN)
2
-
2
,
-
the covalent bonding imparts significant stability to AuI
conclusion is consistent with the fact that AuI
complex in solution and, unlike AuBr
undergo disproportionate reactions.
2
. This
is a highly stable
or AuCl -
, it does not
0
-
2
(
-
2
2
4
9
(
VII. Conclusions
(
A series of metal diiodide complex, MI -
(M ) Cs, Cu, Ag,
2
(
26) ADF 2009.01, SCM, Theoretical Chemistry, Vrije Universiteit,
Au), have been studied both experimentally and theoretically
to examine the metal-iodine chemical bonding. Photoelectron
imaging and low-temperature photoelectron spectroscopy were
used to obtain vibrational and electronic structure information,
yielding accurate electron affinities, 4.52 ( 0.02, 4.256 ( 0.010,
Amsterdam, The Netherlands (http://www.scm.com). (a) te Velde, G.;
Bickelhaupt, F. M.; Baerends, E. J.; Guerra, C. F.; van Gisbergen, S. J. A.;
Snijders, J. G.; Ziegler, T. J. Comput. Chem. 2001, 22, 931. (b) Guerra,
C. F.; Snijders, J. G.; te Velde, G.; Baerends, E. J. Theor. Chem. Acc. 1998,
9
9, 391.
(27) van Lenthe, E.; Baerends, E. J. J. Comput. Chem. 2003, 24, 1142.
(28) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
9, 4597.
and 4.226 ( 0.010 eV for CsI
and well-resolved detachment transitions to the ground and
excited states of the neutral MI complexes. Spin-orbit coupling
2 2 2
, CuI , and AuI , respectively,
9
(29) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
2
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03,
ReVision A.1; Gaussian, Inc.: Wallingford, CT, 2003.
is found to be important in these systems. Theoretical vertical
detachment energies computed using the CASSCF/CCSD(T)/
SO approach allow quantitative assignments of all spectral
transitions. Chemical bonding analyses, including charge popu-
lations, bond orders, electron localization functions, and orbital
interactions reveal a transition from typical ionic bonding
-
-
character in CsI
2 2
to relatively strong covalent bonding in AuI .
Acknowledgment. The experimental work was supported by
the National Science Foundation (CHE-1036387) and by the
U.S. Department of Energy (DOE), Office of Basic Energy
Sciences, Division of Chemical Sciences, Geosciences and
Biosciences, and partly performed at the W. R. Wiley
Environmental Molecular Sciences Laboratory, a national
scientific user facility sponsored by DOE’s Office of Biologi-
cal and Environmental Research and located at Pacific
Northwest National Laboratory, which is operated for the
DOE by Battelle. The theoretical work was supported by
NKBRSF (2006CB932305, 2007CB815200) and NSFC
(
30) Werner, H.-J. MOLPRO, Version 2008.1, a package of ab initio
programs; see: http://www.molpro.net.
(
31) (a) Purvis, G. D.; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910. (b)
Scuseria, G. E.; Janssen, C. L.; Schaefer, H. F. J. Chem. Phys. 1988, 89,
7382.
(
32) (a) Roos, B. O.; Taylor, P. R.; Siegbahn, P. E. M. Chem. Phys.
980, 48, 157. (b) Roos, B. O. Int. J. Quantum Chem. 1980, S14, 175.
33) Berning, A.; Schweizer, M.; Werner, H.-J.; Knowles, P. J.; Palmieri,
P. Mol. Phys. 2000, 98, 1823.
34) Lim, I. S.; Schwerdtfeger, P.; Metz, B.; Stoll, H. J. Chem. Phys.
005, 122, 104103.
1
(
(
(
20933003) in China. The calculations were performed using
2
the DeepComp 7000 Supercomputer at the Computer Net-
work Information Center, Chinese Academy of Sciences and
the Shanghai Supercomputing Center.
(
(
35) Peterson, K. A.; Puzzarini, C. Theor. Chem. Acc. 2005, 114, 283.
36) Peterson, K. A.; Shepler, B. C.; Figgen, D.; Stoll, H. J. Phys. Chem.
A 2006, 110, 13877.
37) Figgen, D.; Rauhut, G.; Dolg, M.; Stoll, H. Chem. Phys. 2005, 311,
27.
(
2
References and Notes
(
38) Hay, P. J. J. Chem. Phys. 1977, 66, 4377.
(
(
(
1) Pyykk o¨ , P. Chem. ReV. 1988, 88, 563.
(39) (a) Gutsev, G. L.; Boldyrev, A. I. Chem. Phys. 1981, 56, 277. (b)
2) Schwarz, H. Angew. Chem., Int. Ed. 2003, 42, 4442.
3) Pyykk o¨ , P. Angew. Chem., Int. Ed 2004, 43, 4412.
Gutsev, G. L.; Boldyrev, A. I. AdV. Chem. Phys. 1985, 61, 169.
(40) Wang, X. B.; Wang, L. S.; Brown, R.; Schwerdtfeger, P.; Schr o¨ der,
D.; Schwarz, H. J. Chem. Phys. 2001, 114, 7388.
(4) Wang, L. S. Phys. Chem. Chem. Phys. 2010, DOI: 10.1039/C003886E.

Investigation of MI2-
J. Phys. Chem. A, Vol. 114, No. 42, 2010 11251
(
41) Wang, X. B.; Ding, C. F.; Wang, L. S.; Boldyrev, A. I.; Simons,
J. J. Chem. Phys. 1999, 110, 4763.
42) (a) Becke, A. D.; Edgecombe, K. E. J. Chem. Phys. 1990, 92, 5397.
b) Savin, A.; Flad, J.; Preuss, H.; Jepsen, O.; Andersen, O. K.; von
Schnering, H. G. Angew. Chem. 1992, 104, 186.
43) Huang, W.; Bulusu, S.; Pal, R.; Zeng, X. C.; Wang, L. S. J. Chem.
Phys. 2009, 131, 234305.
44) Huang, W.; Pal, R.; Wang, L. M.; Zeng, X. C.; Wang, L. S. J. Chem.
Phys. 2010, 132, 054305.
45) (a) Pyykk o¨ , P.; Riedel, S.; Patzschke, M. Chem.sEur. J. 2005,
1, 3511. (b) Pyykk o¨ , P.; Atsumi, M. Chem.sEur. J. 2009, 15, 186. (c)
Pyykk o¨ , P.; Atsumi, M. Chem.sEur. J. 2009, 15, 12770.
(46) Wang, H. H.; Montgomery, L. K.; Geiser, U.; Porter, L. C.; Carlson, K. D.;
Ferraro, J. R.; Williams, J. W.; Cariss, C. S.; Rubinstein, R. L.; Whitworth, J. R.;
Evain, M.; Novoa, J. J.; Whangbo, M.-H. Chem. Mater. 1989, 1, 140.
(47) Kaska, W. C.; Mayer, H. A.; Elsegood, M. R. J.; Horton, P. N.;
Hursthouse, M. B.; Redshaw, C.; Humphrey, S. M. Acta Crystallogr., Sect.
E 2004, 60, m563.
(
(
(
(48) Angle, C. S.; Woolard, K. J.; Kahn, M. I.; Golen, J. A.; Rheingold,
A. L.; Doerrer, L. H. Acta Crystallogr., Sect. C 2007, 63, m231.
(
(
49) Cotton, F. A.; Wilkinson, G. AdVanced Inorganic Chemistry, 5th
(
ed.; Wiley & Sons: New York, 1988; p 948.
1
JP103173D