UV photodissociation of η5-C5H5NiNO: An excited-state Jahn-Teller distortion produces a cartwheeling NO

Article abstract of DOI:10.1021/jp105026n

The 225 nm photodissociation of cyclopentadienylnickel nitrosyl was studied using velocity-mapped ion imaging with 1 + 1′ REMPI detection of the NO (X(2)Π_1/2,3/2, v″ = 0) photofragment. The product recoil energy and angular distributions were measured for selected rotational states of NO. The NO product displays two speeds, a slow product peaked at the center of the ion image and a fast anisotropic product that has an inverted rotational population. In rotational states above J″ = 40.5, an even faster anisotropic NO photofragment appears, most likely because the metal-containing dissociation partner emerges in a lower electronic state, increasing the available energy. The μ-v-j vector correlations were measured and are consistent with the orientation μ∥v⊥ j. The observed vector correlations arise from an excited-state Jahn-Teller distortion of the parent, a distortion that bends the Ni-NO coordinate prior to dissociation.

Full text of DOI:10.1021/jp105026n

1
0922
J. Phys. Chem. A 2010, 114, 10922–10928
5
UV Photodissociation of η -C H NiNO: An Excited-State Jahn-Teller Distortion Produces a
5
5
Cartwheeling NO
‡
†
§
Amber L. Peden, Ryan D. Kieda, Kelsey A. Breck, Joseph R. Basore, Caleb A. Kent, and
Jeffrey A. Bartz*
Department of Chemistry, Kalamazoo College, 1200 Academy Street, Kalamazoo, Michigan 49006
ReceiVed: June 1, 2010; ReVised Manuscript ReceiVed: August 27, 2010
The 225 nm photodissociation of cyclopentadienylnickel nitrosyl was studied using velocity-mapped ion
2
imaging with 1 + 1′ REMPI detection of the NO (X Π1/2,3/2, V′′ ) 0) photofragment. The product recoil
energy and angular distributions were measured for selected rotational states of NO. The NO product displays
two speeds, a slow product peaked at the center of the ion image and a fast anisotropic product that has an
inverted rotational population. In rotational states above J′′ ) 40.5, an even faster anisotropic NO photofragment
appears, most likely because the metal-containing dissociation partner emerges in a lower electronic state,
increasing the available energy. The µ-v-j vector correlations were measured and are consistent with the
orientation µ|v⊥ j. The observed vector correlations arise from an excited-state Jahn-Teller distortion of the
parent, a distortion that bends the Ni-NO coordinate prior to dissociation.
1
. Introduction
Nitric oxide is ubiquitous in biological systems, with roles
observed through measurements of changes in nuclear position
with EXAFS,10 then by observing structural changes in the
11
pentamethyl analog by low-temperature X-ray diffraction, and
in neurotransmission, vasodilation, and the immune response.
Consequently, the role of metal-NO bonding, in biological
systems and nonbiological systems, is an area of intense
interest. As a ligand, NO is versatile because it can bind to
metal centers in a variety of ways. As a cation, NO , it is
isoelectronic with CO, with sp hybridization about the N. In
this case NO binds to a metal center in a linear geometry with
a 180° bond angle. As an anion, NO , the N is considered sp
hybridized and binds to a metal with a bent geometry. Nitric
oxide can bind side-on to a metal center, as is seen in the
12
also by changes in low temperature infrared spectra. The
2
photoinduced linkage isomerization results in a side-on η
coordination in the photoexcited Cp*NiNO (Cp* ) pentam-
1
13
ethylcyclopentadiene). The Ni-NO bond angle is 92° in the
+
metastable excited state.
Density Functional Theory calculations of the ground and
excited states indicate that there are two metastable states for
CpNiNO, the one observed by low-temperature X-ray diffraction
-
2
and a second with the NO rotated so that the O is bonded to
14-16
Ni, forming a CpNiON linkage isomer.
The transition state
2
coordination of NO to copper in nitrile reductase. The binding
2
between the linear Ni-NO ground state and the η metastable
mode affects the reactivity of NO, consequently studying
systems with dynamic interconversion between coordination
geometries is important to understand the factors that drive
reactivity.
15
state has a 137° Ni-NO bond.
Gas phase photodissociation of CpNiNO was studied near
4
50 nm, with [2 + 2] resonance-enhanced multiphoton ioniza-
17
tion (REMPI) for detecting NO in a mass spectrometer. The
dissociation populated a broad distribution of NO rotational
states. The authors suggest that at 450 nm excitation the NO
distribution results from a bent Ni-NO transition state. Pho-
todissociation of CpNiNO near 225 nm, using 1 + 1 REMPI
A volatile metal nitrosyl, cyclopentadienylnickel nitrosyl (η5-
5 5
C H NiNO or CpNiNO, I), is a model for the interconversion
between linear and bent metal-NO bonding. In the ground state,
CpNiNO has a linear Ni-NO bond and C5V symmetry.3 The
highest occupied molecular orbital (HOMO) and the lowest
unoccupied molecular orbital (LUMO) are doubly degenerate
in CpNiNO and the related pentamethyl derivative. In matrix
isolation experiments, photoexcitation of CpNiNO results in a
lowered wavenumber for ν(NO), indicating a change in Ni-NO
bonding.8
Because the LUMO in CpNiNO is doubly degenerate,
excitation of an electron into one of the two degenerate LUMOs
will result in a Jahn-Teller distortion of the excited state. A
photoexcited distortion of the Ni-NO bond angle was first
,4
+
to generate NO for detection, yielded an NO rotational
distribution fit by a Boltzmann temperature.
18
5-7
If photoexcited CpNiNO indeed undergoes a Jahn-Teller
distortion in the gas phase, then the NO photofragment should
show a negative vector correlation between the velocity (v) and
the angular momentum (j) vectors. A negative v-j vector
correlation would be consistent with the observed Ni-NO bend
in condensed phase. The negative v-j vector correlation would
also indicate that the observed changes in CpNiNO bonding
occur through motions in the excited state rather than through
a back reaction of solvent-caged CpNi and NO.
Here we report the velocity-mapped ion imaging of NO from
the photodissociation of CpNiNO at 225 nm. In a velocity-
mapping instrument, the initial velocity of the photoproduct is
mapped onto an x-y point on a two-dimensional detector. Thus,
velocity map imaging gives full speed and angular information
about the photofragments. In the present study of CpNiNO
,9
*
To whom correspondence should be addressed. Tel: (269) 337-7021.
Fax: (269) 337-7508. E-mail: jbartz@kzoo.edu.
‡
Current Address: Department of Chemistry, University of Wisconsins
Madison, Madison, Wisconsin 53706.
†
Current Address: Department of Chemistry, Indiana University, Bloom-
ington, Indiana 47401.
§
Current Address: Department of Chemistry, University of North
Carolina at Chapel Hill, Chapel Hill, North Carolina 27599.
1
0.1021/jp105026n  2010 American Chemical Society
Published on Web 09/24/2010

5
UV Photodissociation of η -C5H5NiNO
J. Phys. Chem. A, Vol. 114, No. 41, 2010 10923
2
24-227 nm probe beam that was separated from the dye
fundamental by two dichroic mirrors. During the experiment, a
photoelastic modulator (Hinds PEM 90, I/CF50) changed the
linear polarization of the probe beam. To generate the ionization
pulse, the more intense beam from the second Nd:YAG was
sent through a delay line on the laser table, then focused into a
2
Raman shifter containing ca. 25 psi of H . After recollimating
the beam, a Pellin-Broca prism separated the 308-nm first anti-
Stokes beam from the other wavelengths, creating the ionization
beam. The probe and ionization beams were spatially overlapped
on the laser table. In the interaction region, the dissociation beam
and the probe/ionization beams counterpropagated, all perpen-
dicular to the time-of-flight axis. The delay between the
dissociation and probe pulses was typically 50 ns. In some
experiments, only one dye laser was used, so the tunable
Figure 1. Schematic representation of the velocity-mapping time-of-
flight mass spectrometer. Three laser pulses, dissociation, probe, and
ionization, intersect the molecular beam at right angles to both the
molecular beam and the time-of-flight chamber.
photodissociation, use of linearly polarized dissociation and
probe lasers and fits of the resulting ion images determined the
vector correlations. The vector correlations reveal the role of
an excited-state Jahn-Teller distortion in the photofragmentation
process.
2
24-227 nm beam did both the dissociation and the probe steps.
In any one experiment, the linearly polarized probe laser was
scanned across the Doppler profile of a rovibronic transition
0-30 times. At each laser position of the scan, 20-30 images
2
were collected for one probe polarization, the photoelastic
modulator changed the polarization, and 20-30 images were
collected for the perpendicular probe polarization. Each image
presented here is the sum of 12 000 individual images.
2
. Experimental Methods
Cyclopentadienylnickel nitrosyl was prepared by previously
+
reported procedure and transferred to a Schlenk flask by vacuum
Once the NO ions are generated, they are accelerated by
19
methods. In an experiment, the compound was treated by three
freeze-pump-thaw cycles and then immersed in 40-50 °C
water. High-purity He carrier gas (∼600 Torr) was flowed over
the sample. The transfer line to the experiment was heated to
the same temperature by water circulation. Since CpNiNO has
the cylindrical ion optics into a 40 cm flight tube. They strike
a multichannel plate detector coupled to a phosphor screen
(
Burle 1340FM). The front of the multichannel plates was gated
+
with a <1 µs pulse at -400 V, timed to detect the NO ions.
The detector speed scaling factor was determined through
20
a vapor pressure of 30 Torr at 60 °C, the pressure of CpNiNO
photodissociation of NO
tion of the NO, similar to a process described by Hradil et al.
In considering the conservation of energy in the NO photo-
dissociation, the speed-scaling factor was found to be 9.44 m
2
at 355 nm with state-selective detec-
2
2
in the sample was assumed to be less than 20 Torr.
The sample was introduced into a velocity-mapping time-
of-flight mass spectrometer (Figure 1) with a design similar to
2
2
1
-
1
-1
that described by Eppink and Parker. The source chamber is
pumped by a diffusion pump (Varian VHS-6) backed by a
mechanical pump (Alcatel 2020A). The sample and carrier gas
are introduced to the experiment through a solenoid valve
s
pixel . Each ion image from the phosphor screen was
collected by a CCD camera (Cooke SensiCam QE) and
transferred to the laboratory computer running LabVIEW
(
National Instruments).
(General Valve Series 9). The resulting molecular beam travels
Timing of the lasers and the detectors was synced to the phase
through a skimmer (Beam Dynamics) into the flight tube, which
is pumped by a turbomolecular pump (Pfeiffer TPU240). The
skimmed beam passes through a 1.5 mm hole in the repeller of
the ion optics (Jordan TOF Products), where it interacts with
the lasers. In these experiments, the repeller is held at 1250
VDC, extractor at 915 VDC, and the grid at ground. The
of the photoelastic modulator. A digital delay generator (BNC
55) sent signals to the lasers and a home-built timing box.
5
The timing box used a signal from the digital delay generator,
indicating that the Nd:YAG Q-switches should be fired, to set
a delay for the desired linear polarization exiting the photoelastic
modulator. A second digital delay generator (SRS DG535),
synced to the output of the probe laser, controlled the timing
of the detector and camera acquisition.
Density Functional Theory (DFT) calculations were per-
formed in Gaussian 98W using B3LYP and the LANL2DZ basis
set.23
-
6
operating pressure in the source chamber is ∼5 × 10 Torr
-6
and in the flight tube is <1 × 10 Torr.
The laser system for these experiments consisted of two Nd:
YAG lasers, two dye lasers, and a Raman shifter. One Nd:YAG-
pumped dye laser generated a dissociation beam. The other Nd:
YAG and dye laser made both a probe beam that excited a
rovibronic transition in NO [A(V′ ) 0) r X(V′′ ) 0)], and an
ionization beam that ionized excited NO molecules in a 1 + 1′
REMPI scheme.
3
. Results
Initial two-color experiments on the photodissociation of
CpNiNO around 225 nm have identified a bimodal speed
distribution after 1 + 1′ REMPI of the nascent NOsa slow
product peaked at the center of the ion image and a faster
product that is anisotropic in the laboratory frame. The fraction
of the ion image that comes from the anisotropic product varies
with the rotational quantum number, J′′, of the NO product.
Figure 2 shows the results of allowing the tunable 224-227
nm beam to both dissociate CpNiNO and excite the nascent
NO [A(V′ ) 0) r X(V′′ ) 0)]. In Figure 2, the rotational state
ranges from J′′ ) 21.5 to 33.5. Little of the fast anisotropic
NO product is seen at rotational states below J′′ ) 20.5, although
fast anisotropic NO products are detected for rotational states
up to at least J′′ ) 46.5.
The 225-nm dissociation beam was generated by frequency-
doubling the 450-nm output of a dye laser (Radiant Narrow-
Scan), pumped by the 355-nm third harmonic of one Nd:YAG
laser (Innolas Spitlight 600). The 225-nm dissociation beam was
separated from the fundamental by two dichroic mirrors.
Between experiments, a half-wave plate changed the linear
polarization of the dissociation beam. The probe and ionization
beams are created with the 355-nm third harmonic of the second
Nd:YAG laser (Quanta-Ray LAB-190), which was split 70:30
by a dichroic beamsplitter. The less intense beam pumped a
second dye laser (Sirah Cobra Stretch). The second dye laser
fundamental was frequency-doubled to generate a tunable

1
0924 J. Phys. Chem. A, Vol. 114, No. 41, 2010
Peden et al.
Figure 3. Ion images from the 225 nm photodissociation of η5-
5 5
C H NiNO detected by 1 + 1′ REMPI through the Q11 + P21 (33.5)
rovibronic transition of the NO photofragment. The labels on the left
indicate whether the linear polarization of the fixed 225-nm dissociation
laser is either perpendicular (double-ended vertical arrow) or parallel
(
.) to the time-of-flight axis. The labels on the top indicate whether
the linear polarization of the tunable 224-227 nm probe laser is either
perpendicular (double-ended vertical arrow) or parallel (.) to the time-
of-flight axis.
5
Figure 2. Ion images from the photodissociation of η -C
5 5
H NiNO near
TABLE 1: Bipolar Moments, Physical Description of the
Bipolar Moments, And the Physical Limits of the Bipolar
Moments Studied Herein
2
25 nm detected through state-selective 1 + 1′ REMPI of the NO
photofragment. The ion images appear on the left and the row sums of
the images on the right. Both show the appearance of a fast anisotropic
NO product at higher rotational quantum numbers, J′′, detected through
the Q11 + P21 transition at (a) J′′ ) 21.5, (b) J′′ ) 26.5, and (c) J′′ )
a
bipolar moment
physical description
physical limits
2
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
0
(20)
(02)
(22)
(22)
(42)
µ-v velocity anisotropy
µ-j rotational alignment
v-j correlation
quadratic µ-v-j correlation
quartic µ-v-j correlation
+1 to -1/2
+1 to -1/2
+1 to -1/2
+1/2 to -1
+1 to -1/2
3
3.5. In each image, a single laser around 225 nm, polarized
2
0
perpendicular to the time-of-flight axis, provides both the dissociation
and probe pulse. In the false color images, blue shows the highest ion
intensity and black indicates zero ions.
0
0
2
0
2
0
a
Adapted from Dixon.24
To probe the vector correlation in the photodissociation and
NO product, a third laser was added to the experiment. The
three-color experiment had three individual laser pulses: a 225-
nm dissociation beam to dissociate CpNiNO, a state-selective
tunable 224 - 227 nm probe beam to excite the nascent NO
fragment, and a 308 nm beam to ionize the excited NO by 1 +
correlations in the photodissociation event. The bipolar moments
2
2
studied here include ꢀ (20), the µ-v velocity anisotropy; ꢀ (02),
0
0
2
0
the µ-j rotational alignment; ꢀ (22), the v-j correlation; ꢀ (22),
0
0
2
the quadratic µ-v-j correlation; and ꢀ (42), the quartic µ-v-j
0
correlation. The ion images were analyzed using the fitting
1
′ REMPI. The 225-nm dissociation beam was detuned from
program fimage, which fits the ion images to two velocity
distributions, one, a Gaussian distribution with no anisotropy
and a second, a displaced Gaussian distribution with laboratory-
frame anisotropy defined by the bipolar harmonics.25 For each
rovibronic transition studied, a set of four images was fit,
corresponding to the four permutations of the linear polarizations
of the dissociation and probe lasers. Figure 4 shows the synthetic
images created from the forward convolution fitting of the data
in Figure 3. The synthetic data recreate the shape and relative
intensities well.
any NO (A r X) transitions to limit the background from the
dissociation laser; 225 nm light is capable of dissocia-
ting the parent and ionizing the NO product by [1 + 1] REMPI.
The probe laser pulse energy was decreased to eliminate signal
from just the probe laser or from only the probe and ionization
beams. The dissociation and probe lasers had independently
controlled linear polarizations, either parallel (double-ended
vertical arrow) or perpendicular (.) to the detector plane. Figure
3
shows typical results for the 225-nm photodissociation of
CpNiNO with independent control of the linear polarizations
of the dissociation and probe beams. The data in Figure 3 are
from probing the Q11 + P21(33.5) transition of NO. In Figure
Table 2 summarizes the fits of several NO rovibronic
transitions. All of the transitions fitted arise from the Q11 + P21
or Q22 + R12 branches of the rovibronic spectrum. In NO, the
2
3
, the ion image is strongly affected by changing the linear
Λ-doublet components of the Π ground state are labeled Π(A′)
26
polarization of the dissociation and of the probe beams.
or Π(A′′). In the high-J limit, the Λ-doublets Π(A′) and Π(A′′)
are symmetric and antisymmetric, repectively, with respect to
reflection in the plane of rotation. The Q11 + P21 and Q22 + R12
transitions reported in Table 2 arise from the Π(A′′) Λ-doublet.
Several P11 transitions were also studied, arising from the Π(A′)
Λ-doublet of NO. The assignments of the Λ-doublets are based
The ion images were analyzed by the semiclassical bipolar
24
harmonic scheme first described by Dixon. Nestorov et al. have
extended Dixon’s method to extract bipolar moments from ion
25
imaging experiments. Briefly, the cylindrical symmetry of the
dissociation event is described in terms of bipolar harmonics.
The weighting coefficients of the bipolar harmonics, the bipolar
moments, are constrained to physically allowed values as shown
in Table 1. The bipolar moments characterize the vector
27
on the energy level diagram of Lahmani et al., confirmed by
28
the parity assignments of Geuzebroek et al., and the Π(A′)
26
and Π(A′′) assignments of Alexander et al. The P11 images

5
UV Photodissociation of η -C5H5NiNO
J. Phys. Chem. A, Vol. 114, No. 41, 2010 10925
Figure 4. Synthetic ion images of the 225 nm photodissociation of
5
η -C
5
H
5
NiNO based on a forward convolution fit of the ion images in
Figure 3. The synthetic images show similar qualitative featuresstwo
speed distributions, a slower isotropic distribution peaked at zero and
a faster anisotropic distribution.
Figure 5. Ion images from the 225 nm photodissociation of η5-
C
5
H
5
NiNO showing a faster anisotropic NO product appearing at J′′ >
4
0.5. As the rotational state changes from (a) J′′ ) 39.5, to (b) J′′ )
4
3.5, and finally to (c) J′′ ) 46.5, the NO photofragments have more
showed the similar qualitative behavior under the change in
linear polarization of the probe and dissociation beams, however
in all cases, the intensities of the fast anisotropic product in the
rotational and more translational energy. In these images, the 225 nm
dissociation laser is polarized perpendicular to the time-of-flight axis
and the probe laser is polarized parallel to the time-of-flight axis. The
highest ion intensity is white and the lowest is black.
P
11 transitions were much weaker. Subsequently, the fimage
program fit the data poorly. The fast anisotropic product detected
for the P11 transitions are half as intense as the Q11 + P21 or
Q22 + R12 transitions, as determined by comparing the intensity
of the fast anisotropic product to the slower isotropic products.
The difference in intensity indicates that the NO photofragment
is more likely to exit in the Π(A′′) than in the Π(A′) Λ-doublet.
At higher rotational quantum numbers (J′′ > 40.5), a faster
anisotropic NO photoproduct appears in the ion images. Figure
5
shows the NO velocity images for three rovibronic transitions,
Q
11 + P21(39.5), Q11 + P21(43.5), and Q11 + P21(46.5). In Figure
Figure 6. Depiction of the doubly degenerate lowest-occupied
1
6
5
(a) the anisotropic velocity distribution is peaked at 1200 m/s
molecular orbital (LUMO) labeled 7e by Boulet et al. The LUMO
1
for the Q11 + P21(39.5) transition. In Figure 5(b), with the probe
laser tuned to the Q11 + P21(43.5) transition, an additional
dissociation route, generating a faster NO product, is evident.
When the probe laser is tuned to the Q11 + P21(46.5) transition,
the peaks of the two anisotropic NO products in Figure 5(c)
are 1000 and 1700 m/s.
has an antibonding interaction between the Ni metal center and NO
ligand. Populating either of these two orbitals will generate a
Jahn-Teller instability that will induce a bend in the Ni-NO
coordinate.
Promoting an electron to one of the two degenerate 7e
1
orbitals in CpNiNO results in a Jahn-Teller instability. As a
result, the Ni-NO bond bends. Dissociating from a bent
geometry produces the inverted rotational state distribution for
the fast NO photoproduct seen here. Similar inverted NO
product state distributions have been observed in the photodis-
4
. Discussion
The appearance of a fast anisotropic distribution at J′′ > 20.5
in Figure 2 is consistent with a prompt, parallel dissociation
along the Ni-NO coordinate. Boulet et al. have assigned the
sociation of CH
3
ONO and other nitrites, although in nitrites the
ground state is bent.2
9-36
The bent nitrite ground state is mapped
2
25 nm excitation to the transition from 8a
1
to the doubly
1
6
degenerate 7e
1
LUMO (Figure 6). Under C5V symmetry, the
onto the excited state where an impulse along the R-NO
coordinate generates a product state distribution peaked at high
J”. Schinke termed the mapping of the bent ground state
geometry into an inverted rotational state distribution, the
transition dipole for the E′′ r A′ transition is aligned along the
symmetry axis of the molecule, parallel to the Ni-NO bond.
The ion images in Figures 2, 3, and 5 are consistent with the
parallel assignment.
3
7
rotation-reflection principle. In CpNiNO, the rotational
TABLE 2: Average Velocities of the Isotropic and Anisotropic Speed Distributions, the Bipolar Moments, and the Molecular
Frame Polarization Parameters Found in the Anisotropic Speed Distribution Listed by Rovibronic Transition in NO
2
2
5
[
A ( Σ, W′ ) 0) r X ( Π
Ω
, W′′ ) 0)] Following 225 nm Photodissociation of η -C
5
H
5
NiNO
2
2
0
2
2
(2)
transition
<v>iso
<v>aniso
ꢀ
0
(20)
ꢀ
0
(02)
ꢀ
0
(22)
ꢀ
0
(22)
ꢀ
0
(42)
1
Re[a (||, ⊥)]
Q
Q
Q
Q
11 + P21 (33.5)
11 + P21 (33.5)
22 + R12 (35.5)
11 + P21 (39.5)
1000
1000
1100
1200
1400
1400
1300
1300
0.98
1
0.86
0.99
+1
-0.32
-0.34
-0.21
-0.17
-1/2
-0.22
-0.26
-0.15
-0.12
-1/2
0.21
0.35
0.18
0.20
+1/2
-0.30
-0.35
-0.25
-0.21
-1/2
0.04
0.01
0.09
0.05
0
a
b
physical limit: µ|v⊥j
a
b
2
4
T
h
e
b
i
p
o
l
a
r
m
o
m
e
n
t
w
a
s
fi
x
e
d
a
t
t
h
e
p
h
y
s
i
c
a
l
l
i
m
i
t
.
A
d
a
p
t
e
d
f
r
o
m
D
i
x
o
n
.

1
0926 J. Phys. Chem. A, Vol. 114, No. 41, 2010
Peden et al.
5
Figure 7. Internal motion in the parent η -C
5
H
5
NiNO corresponds to
the vectors and vector correlation observed in the laboratory frame.
The transition dipole (µ) is aligned with the symmetry axis and the
initially linear Ni-NO. As the Ni-NO bond bends, motion around
the N-O center-of-mass is asymptotically linked with rotation in NO,
shown as the perpendicular vector j (.) in the diagram. Breaking the
Ni-NO bond results in the NO velocity (v) aligned preferentially with
the transition dipole (µ) and perpendicular to the angular momentum
Figure 8. Symmeterized experimental, fit, and difference images show
2
the role that ꢀ
0
(42) plays in making an accurate fit of the ion images.
Image (a) is the symmeterized version of Figure 3(b), using a different
false color scheme to accentuate the dip in intensity along the vertical
axis. Image (b) is a synthetic image based on the full parameter fit of
(
j). The bipolar moments in the limiting case of µ|v⊥j are indicated,
24
adapted from Dixon.
the experimental image. Image (c) is a fit of the experimental image
2
excluding the quartic µ-v-j bipolar moment, ꢀ
0
(42). Images (d)-(f)
product state distribution for the fast NO product is inverted
although the ground state is linear in the CpNi-NO coordinate,
appearing to violate the rotation-reflection principle. One
explanation is that CpNiNO photodissociation is prompt but is
not direct. As such, the Jahn-Teller distortion is faster than
the dissociation of the Ni-NO bond. The time averaged
are differences between the images (a)-(c) as indicated. The difference
image scales show positive differences in red-orange and negative
differences in blue. The similarities in difference images (d) and (f)
2
show that ꢀ
0
(42) has a significant contribution that is evident in both
the experimental (a) and fitted (b) images.
3
8
correlation function of Yang and Bersohn was used with the
As shown in Figure 7, the angular momentum vector in the
ejected NO (.) is perpendicular to the plane defined by the
bent Ni-NO moiety. The measured bipolar moments appear
in Table 2. The vector correlations indicate that the general
2
0
traditional anisotropy parameter, ꢀ [) 2ꢀ (20)], and the moment
3
9
of inertia of CpNiNO to estimate an excited-state lifetime of
CpNiNO on the order of hundreds of femtoseconds. The Ni-NO
-
1 12
bending frequency in the CpNiNO ground state is 490 cm ,
corresponding to a vibrational period of 68 fs, a sufficiently
short time to permit the Ni-NO bond to bend prior to
dissociation. Similar excited-state behavior, starting from a linear
ground state and resulting in a bent excited state, has been seen
in ICN.4
vector correlations are µ|v⊥j, however the only bipolar moment
2
that reaches the theoretical limit is ꢀ
0
(20), the velocity anisot-
ropy. The velocity anisotropy is large because the parent
molecule dissociates promptly, has relatively high moment of
inertia, and is cooled in a molecular beam.
0-44
2
The bipolar moment ꢀ
0
(42), quartic µ-v-j correlation, is
The isotropic peak at the center of the ion image is assumed
to be the dissociation of vibrationally hot CpNiNO ground state.
Photodissociation experiments with a 450 nm or a 280 nm
dissociation beam produced only the slower isotropic product.
The lower energy dissociation photons (450 or 280 nm) must
have sufficient energy to dissociate CpNiNO in an RRKM
fashion, but insufficient energy to access the prompt dissociating
state.
The combination of bending and breaking the Ni-NO bond
results in strong vector correlations between and among velocity,
v, angular momentum, j, and transition dipole, µ, as is seen in
Table 2. The NO photoproduct emerges with a cartwheel-like
motion, cartwheels that are strongly oriented in the laboratory
frame. The strong vector correlations begin with internal motions
in CpNiNO, depicted in Figure 7, that correlate with the
orientations of µ, v, and j. In Figure 7, the values of the five
bipolar moments fit here are indicated for the limiting case of
a parallel electronic transition, prompt dissociation, and a
cartwheeling motion of the NO. Because the CpNiNO undergoes
a parallel transition and the molecule dissociates promptly, the
NO velocity vector is along the symmetry axis and parallel to
the polarization of the dissociation laser. After excitation but
before dissociation, the Ni-NO bond bends, an internal motion
associated with the angular momentum vector in the free NO.
2
0
important in fitting the data, even though the value of ꢀ (42)
can be determined by other bipolar moments. Costen and Hall
have shown how using fewer fit parameters in ICN dissociation
43,45
accurately determines the vector correlation.
The justification
for using fewer parameters in the fit comes from considering
(2)
the molecular frame polarization parameter, Re[a (||, ⊥)],
1
(
1
2)
1
2
0
1
7
2
0
12
35
2
0
Re[a (||, ⊥)] ) 2
√
6[5
ꢀ (02) - ꢀ (22) - ꢀ (42)
(1)
]
which is sensitive to the interference of multiple dissociative
states.46 In the high-J limit, the contribution of the molecular
frame polarization parameter should vanish. Table 2 shows the
values of Re[a(
2)
(||, ⊥)] calculated from eq 1, which are nearly
1
zero for each transition studied. Because the molecular frame
polarization parameter approaches zero, the fits are overdeter-
2
2
mined because the values of ꢀ
calculate the expected value of ꢀ
necessary for accurate fits of the images.
0
0
(02) and ꢀ (22) can be used to
2
2
0
0
(42). Even so, ꢀ (42) is
A parallel set of forward convolution fits using fimage were
performed to test whether ion images, such as Figure 8(a), were
reasonably fit with fewer parameters. The first set of fits included
2
ꢀ
0
(42) [Figure 8(b)] while the second set [Figure 8(c)] did not.

5
UV Photodissociation of η -C5H5NiNO
J. Phys. Chem. A, Vol. 114, No. 41, 2010 10927
Figure 9. The possible orientations of the singly occupied NO π*
orbital, in the high-J limit, shown relative to the plane of rotation. NO
has an unpaired electron that populates one of two π* orbitals. When
the unpaired electron occupies the π* orbital in the plane of rotation,
it is labeled Π(A′). When the unpaired electron occupies in the π*
orbital perpendicular to the plane it is labeled Π(A′′).
Figure 10. A plot of the relative energies of the lowest two unoccupied
5
orbitals in η -C
5
H
5
NiNO as the Ni-NO bond bends. At 180° the
orbitals are doubly degenerate. As the Ni-NO bond initially bends,
the π* orbital in the Ni-NO plane, colored green and yellow for clarity,
is lower in energy. As the Ni-NO bond angle goes below 100°, the
out-of-plane π* orbital (red and blue) stabilizes as the in-plane π* orbital
rises in energy. The in-plane π* orbital is asymptotically connected to
the Π(A′) Λ-doublet in NO and the out-of-lane π* is asymptotically
For each experimental transition in Table 2, either the full fit,
2
2
with ꢀ
0
0
(42), or the partial fit, without ꢀ (42), yielded similar
values for the other bipolar moments and the product speeds.
connected to the Π(A′′) Λ-doublet. In the photodissociation of
2
2
5
One slight difference emerged in the ꢁ values wherein the ꢁ
η -C
5
H
5
NiNO, the NO photofragment has a higher population in the
Π(A′′) Λ-doublet with the unpaired electron out of the plane of rotation.
The energies of the in- and out-of-plane orbitals predict that the Ni-NO
bond is highly bent during the dissociation process.
values for the partial fits were at least 10% larger.
2
The full parameter fit, with ꢀ
0
(42), and the partial fit, without
2
ꢀ
0
(42), show differences as the resulting synthetic images are
compared. Figure 8(d) is a difference image between the
2
0
complete fit, including ꢀ (42), and the partial fit. In Figure 8(d)
shows how the initially degenerate LUMO changes as a function
of Ni-NO angle. As the Ni-NO bond bends, the π* orbital
that is asymptotically connected to the Π(A′) Λ-doublet is in
the Ni-NO plane. Conversely, the π* orbital that is asymptoti-
cally connected to the Π(A′′) Λ-doublet is perpendicular to the
Ni-NO plane. As the Ni-NO angle changes from linearity,
the in-plane π* orbital lowers in energy, reaching a minimum
at 120°. The out-of-plane π* orbital energy lowers slightly, as
the Ni-NO bond bends. At an Ni-NO angle of 100°, the
energies of both the in-plane and out-of-plane π* orbitals cross.
At angles below 100°, the energy of the in-plane π* orbital
increases sharply as out-of-plane orbital, the one that asymptoti-
cally leads to the Π(A′′) Λ-doublet in NO, decreases in energy.
Slight bends in the Ni-NO bond will not result in a NO
photofragment in the Π(A′′) Λ-doublet, however extreme
bending motions will.
a positive difference is a red-orange color and a negative
difference is a blue color. The difference between the experi-
mental image 8(a) and the complete fit 8(b) appears in Figure
(e) and the difference between the experimental image 8(a)
and the partial fit 8(c) appears in Figure 8(f). It is clear in Figure
(e) that the synthetic image based on the forward convolution
fits is not a perfect reproduction of the experimental data, but
the differences between the experiment and synthetic images
are much more significant in the partial fits, when ꢀ
neglected. In comparing the intensities and patterns of difference
images 8(d), 8(e), and 8(f), the two that have the most similarity
are 8(d) and 8(f) showing that the difference in contribution of
(42) is the same in the fit and experimental images. The dip
in intensity along the vertical bisector of the symmeterized ion
image [Figure 8(a)] gives a visual clue that ꢀ (42) is necessary
0
to fit the data. The intensity dip is a general feature of the ion
images with both the dissociation and probe laser polarizations
perpendicular to the detector axis for transitions that have a high
fraction of Q character. The bipolar moment ꢀ
order bipolar harmonic that better fits the dip. Although in the
8
8
2
0
(42) is
2
ꢀ
0
2
The most surprising result is the appearance of NO photo-
products with higher translational energy and higher rotational
energy above J′′ ) 40.5. Fits to the ion image from the Q11
+
2
0
(42) adds a higher
P
21 (J′′ ) 46.5) transition in Figure 5(c) give similar bipolar
moments to those of the lower NO rotational states. As a result,
it is assumed that the dissociation pathway to produce the faster
NO products at higher J′′ states is similar to that of the other
anisotropic NO products. Since one of the photoproducts, CpNi,
contains a metal, there are low-lying CpNi electronic states that
affect the energy available to NO. A TDDFT calculation on
(
2)
high-J limit, the contribution of Re[a
1
(||, ⊥)] vanishes, and as
a result, ꢀ2(02) and ꢀ2
(22) can be used to calculate the expected
(42), but inclusion of ꢀ (42) is necessary to fit the
0
0
2
0
2
value of ꢀ
0
ion images fully.
The nascent NO exits the dissociation in a preferred orienta-
tion of the unpaired electron. In the high-J limit, the two
Λ-doublets of the Π ground state of NO correspond to π*
-1
CpNi finds the lowest excited state is 310 cm above the ground
2
state. The difference in translational energy release between the
fastest products in Figure 5, parts (a) and (c), is around 300
orbitals that are in the plane or perpendicular to the plane of
rotation, labeled Π(A′) and Π(A′′), respectively, as shown in
Figure 9. The strong Q11 + P21 and Q22 + R12 transitions
reported in Table 2 arise from the Π(A′′) Λ-doublet. The weaker
-1
cm , accounting for the gain in translational energy. The
Ni-NO bond bending that results in higher NO rotation must
facilitate a curve crossing in the CpNiNO to another dissociating
surface. On the second dissociating surface, the CpNi fragment
has lower electronic energy, which increases the energy available
to partition into photofragment recoil.
P
11 transitions are poorly fit and originate from the Π(A′)
Λ-doublet. The ejected NO has memory of the photodissociation
event because the NO shows a preference for Π(A′′) over Π(A′).
A series of single-point DFT calculations were performed to
understand why the NO shows a preference for having the
unpaired electron out of the plane of NO rotation. In the
calculations, both the Ni-NO and Cp-Ni-N bonds were bent
to keep the center of mass along the Cp-Ni axis. Figure 10
5
. Conclusions
Photodissociation of the organometallic compound, CpNiNO,
at 225 nm generates an NO photofragment from bent excited

1
0928 J. Phys. Chem. A, Vol. 114, No. 41, 2010
Peden et al.
state arising from a Jahn-Teller distortion. Forward convolution
fits to the photoejected NO show strong vector correlations
between the transition dipole, velocity, and angular momentum.
The initially linear Ni-NO bond bends significantly prior to
dissociation, consistent with changes in bonding observed in
the condensed phase. Although DFT and TDDFT calculations
give some information on why the excited state bends and shows
a preference for the Π(A′′) Λ-doublet a more exhaustive
computational study of the excited state behavior of CpNiNO
would give a more detailed explanation for the observations
shown here and a better understanding of metal-NO bonding
in general.
(16) Boulet, P.; Chermette, H.; Weber, J. Inorg. Chem. 2001, 40, 7032–
7
039.
(
17) Georgiou, S.; Wight, C. A. Chem. Phys. Lett. 1986, 132, 511–515.
(
(
18) Georgiou, S.; Wight, C. A. J. Chem. Phys. 1988, 88, 7418–7423.
19) King, R. B., Organometallic Syntheses; Academic: New York, 1965;
Vol. 1, pp 169-171.
20) Jolly, P. W., The Organic Chemistry of Nickel; Academic: New
York, 1974.
(
(
21) Eppink, A. T. J. B.; Parker, D. H. ReV. Sci. Instrum. 1997, 68,
3
477–3484.
(22) Hradil, V. P.; Suzuki, T.; Hewitt, S. A.; Houston, P. L.; Whitaker,
B. J. J. Chem. Phys. 1993, 99, 4455–4463.
23) Frisch, M. J., et al. Gaussian 98 (ReVision A.5.4), Gaussian:
Pittsburgh, PA, 1998.
(
(
(
24) Dixon, R. N. J. Chem. Phys. 1986, 85, 1866–1879.
25) Nestorov, V. K.; Hinchliffe, R. D.; Uberna, R.; Cline, J. I.; Lorenz,
Acknowledgment. This work was supported by the National
Science Foundation (Awards CHE-0420928 and CHE-0314745).
A.P. acknowledges summer support from the Hutchcroft Fund.
R.K. received summer support from the Heyl Foundation. K.B.
acknowledges support from the Howard Hughes Medical
Institute through an Undergraduate Science Education Program
grant to Kalamazoo College (Award 52005128). The authors
thank Professor Joseph Cline for providing the fimage program.
K. T.; Chandler, D. W. J. Chem. Phys. 2001, 115, 7881–7891.
(
(
26) Alexander, M. H. et al. J. Chem. Phys. 1988, 89, 1749–1753.
27) Lahmani, F.; Lardeux, C.; Solgadi, D. Chem. Phys. Lett. 1986, 129,
2
4.
(
28) Geuzebroek, F. H.; Tenner, M. G.; Kleyn, A. W.; Zacharias, H.;
Stolte, S. Chem. Phys. Lett. 1991, 187, 520–6.
29) Winniczek, J. W.; Dubs, R. L.; Appling, J. R.; McKoy, V.; White,
M. G. J. Chem. Phys. 1989, 90, 949–963.
30) Suter, H. U.; Br u¨ hlmann, U.; Huber, J. R. Chem. Phys. Lett. 1990,
71, 63–67.
31) Yin, H.-M.; Sun, J.-L.; Li, Y.-M.; Han, K.-L.; He, G.-Z.; Cong,
S.-L. J. Chem. Phys. 2003, 118, 8248–8255.
32) Br u¨ hlmann, U.; Dubs, M.; Huber, J. R. J. Chem. Phys. 1987, 86,
1249–1257.
(
(
1
(
References and Notes
(
(
1) Richter-Addo, G. B.; Legzdins, P.; Burstyn, J. Chem. ReV. 2002,
02, 857–1270, Thematic Issue on Nitric Oxide Chemistry.
2) Tocheva, E. I.; Rosell, F. I.; Mauk, A. G.; Murphy, M. E. P. Science
004, 304, 867–870.
3) Cox, A. P.; Thomas, L. F.; Sheridan, J. Nature 1958, 181, 1157–
158.
1
2
1
(
33) Br u¨ hlmann, U.; Huber, J. R. Z. Phys. D: At., Mol. Clusters 1987,
(
7
2
1
2
, 1–8.
(
34) McCoustra, M. R. S.; Hippler, M.; Pfab, J. Chem. Phys. Lett. 1992,
(
00, 451–458.
(
35) D’Azy, O. B.; Lahmani, F.; Lardeux, C.; Solgadi, D. Chem. Phys.
985, 94, 247–256.
36) Br u¨ hlmann, U.; Huber, J. R. Chem. Phys. Lett. 1988, 143, 199–
03.
(
(
4) Cox, A.; Brittain, A. Trans. Faraday Soc. 1970, 66, 557–562.
5) Field, C.; Green, J.; Mayer, M.; Nasluzov, V.; Rosch, N.; Siggel,
(
M. Inorg. Chem. 1996, 35, 2504–2514.
6) Li, X.; Tse, J.; Bancroft, G.; Puddephatt, R.; Tans, K. Inorg. Chem.
996, 35, 2515–2523.
(
(
(
(
37) Schinke, R. J. Chem. Phys. 1986, 85, 5049–5060.
38) Yang, S.; Bersohn, R. J. Chem. Phys. 1974, 61, 4400–4407.
39) Karunatilaka, C.; Subramanian, R.; Pedroza, D.; Idar, D. J.;
1
(
7) Green, J.; Underwood, C. J. Organomet. Chem. 1997, 528, 91–94.
8) Crichton, O.; Rest, A. J. Chem. Soc., Chem. Commun. 1973, 1973,
(
Kukolich, S. G. J. Phys. Chem. A 2007, 111, 6191–6196.
40) Nadler, I.; Mahgerefteh, D.; Reisler, H.; Wittig, C. J. Chem. Phys.
4
07–407.
9) Crichton, O.; Rest, A. J. J. Chem. Soc., Dalton Trans. 1977, 986–
(
(
1
985, 82, 3885–3893.
9
93.
10) Chen, L. X.; Bowman, M. K.; Wang, Z.; Montano, P. A.; Norris,
J. R. J. Phys. Chem. 1994, 98, 9457–9464.
11) Fomitchev, D. V.; Furlani, T. R.; Coppens, P. Inorg. Chem. 1998,
7, 1519–1526.
12) Schaiquevich, P. S.; Güida, J. A.; Aymonino, P. J. Inorg. Chim.
Acta 2000, 303, 277–281.
13) Fomitchev, D. V.; Furlani, T. R.; Coppens, P. Inorg. Chem. 1998,
7, 1519–1526.
14) Boulet, P.; Buchs, M.; Chermette, H.; Daul, C.; Furet, E.; Gilardoni,
F.; Rogemond, F.; Schlaepfer, C. W.; Weber, J. J. Phys. Chem. A 2001,
05, 8999–9003.
15) Boulet, P.; Buchs, M.; Chermette, H.; Daul, C.; Gilardoni, F.;
Rogenmond, F.; Schlaepfer, C. W.; Weber, J. J. Phys. Chem. A 2001, 105,
991–8998.
(
(41) O’Halloran, M. O.; Joswig, H.; Zare, R. N. J. Chem. Phys. 1987,
87, 303–313.
(42) Black, J. F.; Waldeck, J. R.; Zare, R. N. J. Chem. Phys. 1990, 92,
3519–3538.
(43) Costen, M. L.; North, S. W.; Hall, G. E. J. Chem. Phys. 1999,
111, 6735–6749.
(44) Marinelli, W. J.; Sivakumar, N.; Houston, P. L. J. Phys. Chem.
2002, 88, 6685–6692.
(
3
(
(
3
(
(45) Costen, M. L.; Hall, G. E. Phys. Chem. Chem. Phys. 2007, 9, 272–
287.
1
(
46) Rakitzis, T. P.; Hall, G. E.; Costen, M. L.; Zare, R. N. J. Chem.
(
Phys. 1999, 111, 8751–8754.
8
JP105026N