Reactions of laser-ablated osmium and ruthenium atoms with nitrogen. Matrix infrared spectra and density functional calculations of osmium and ruthenium nitrides and dinitrides

Article abstract of DOI:10.1021/jp993338s

Laser-ablated osmium and ruthenium atoms were reacted with nitrogen molecules; the products were isolated in solid argon and nitrogen and identified by infrared spectroscopy. Both MN and NMN nitrides are observed, and estimates for the triatomic bond angles are made using nitrogen and ruthenium isotopic data. The growth of NOsN on annealing in solid argon suggests that osmium atoms insert into the dinitrogen triple bond at cryogenic temperatures, allowing a lower limit of ～473 kJ/mol to be estimated for the average Os-N bond energy in NOsN. The force constants for MN and NMN (M= Fe, Ru, Os) were calculated using all available isotopic data; force constants increase moving down the metal group, and diatomic MN force constants are larger than those for the corresponding NMN triatomic molecules. DFT calculations for the ruthenium and osmium nitrides give reasonable agreement with experiment. Bonding analyses for these molecules show . that the M-N bonds are largely nonpolar with bond orders in the range 2.5-3.0. Several metal dinitrogen complexes are also observed and assignments are proposed.

Full text of DOI:10.1021/jp993338s

1152
J. Phys. Chem. A 2000, 104, 1152-1161
Reactions of Laser-Ablated Osmium and Ruthenium Atoms with Nitrogen. Matrix Infrared
Spectra and Density Functional Calculations of Osmium and Ruthenium Nitrides and
Dinitrides
Angelo Citra and Lester Andrews*
Department of Chemistry, UniVersity of Virginia, CharlottesVille, Virginia 22901
ReceiVed: September 17, 1999; In Final Form: NoVember 15, 1999
Laser-ablated osmium and ruthenium atoms were reacted with nitrogen molecules; the products were isolated
in solid argon and nitrogen and identified by infrared spectroscopy. Both MN and NMN nitrides are observed,
and estimates for the triatomic bond angles are made using nitrogen and ruthenium isotopic data. The growth
of NOsN on annealing in solid argon suggests that osmium atoms insert into the dinitrogen triple bond at
cryogenic temperatures, allowing a lower limit of ∼473 kJ/mol to be estimated for the average Os-N bond
energy in NOsN. The force constants for MN and NMN (M) Fe, Ru, Os) were calculated using all available
isotopic data; force constants increase moving down the metal group, and diatomic MN force constants are
larger than those for the corresponding NMN triatomic molecules. DFT calculations for the ruthenium and
osmium nitrides give reasonable agreement with experiment. Bonding analyses for these molecules show
that the M-N bonds are largely nonpolar with bond orders in the range 2.5-3.0. Several metal dinitrogen
complexes are also observed and assignments are proposed.
Introduction
(4%) at 5-8 mmol/h for 60-90 min, and with undiluted 14N2
and 14N2 + 15N2 for 30 min at the same rate. FTIR spectra were
recorded with 0.5 cm-1 resolution on a Nicolet 550 instrument.
Matrix samples were successively warmed and recooled, and
more spectra were collected; the matrix was subjected to broad-
band photolysis with a medium-pressure mercury arc (Philips,
175 W) with globe removed (240-580 nm) at different stages
in the annealing cycles.
Osmium and ruthenium have an extensive nitrido and
dinitrogen chemistry. Species of the form OsNX52- and OsNX4
-
(X ) Cl, Br, I) are known, and they undergo a variety of
reactions, including the abstraction of sulfur from thiocyanate
and substitution by phosphine ligands.1 These species contain
strong Os-N bonds with significant multiple bond character,
as shown by the bond lengths and stretching frequencies. The
analogous ruthenium nitrides have been characterized.1 Various
dinitrogen complexes are also known for both metals, the most
notable of these is [Ru(NH3)N2]2+ which was the first stable
dinitrogen complex to be isolated.2 The analogous osmium
complex [Os(NH3)N2]2+ is a useful synthetic intermediate.3 The
interaction between the metal surfaces and nitrogen has also
been studied with regard to their effectiveness in ammonia
synthesis.4-6
In this work, osmium and ruthenium nitrides and dinitrogen
complexes are formed by the reaction of the respective metal
atoms with dinitrogen and are isolated in argon and nitrogen
matrixes at cryogenic temperatures. The reaction products are
characterized using infrared spectroscopy and density functional
theory, and the results are discussed within the context of the
previous matrix isolation study of iron nitrides in this labora-
tory.7
Results and Discussion
Infrared spectra and density functional theory (DFT) calcula-
tions of osmium and ruthenium dinitrogen reaction products will
be presented in turn. The osmium and ruthenium nitride spectra
in the frequency ranges 1140-860 cm-1 and 1000-800 cm-1
are shown in Figures 1 and 2, respectively. The first spectrum
in Figure 1 shows osmium in a pure nitrogen isotopic mixture
14N2 + 15N2 after the first annealing cycle, and the remaining
spectra show the annealing behavior for similar samples diluted
in argon matrixes. The spectra in Figure 2 for ruthenium in
mixed isotope nitrogen matrixes clearly show the natural
ruthenium isotopic structure. Absorptions due to the osmium
and ruthenium dinitrogen complexes are observed above 2000
cm-1 and are shown in Figures 3 and 4, respectively, for osmium
and ruthenium with 14N2 + 15N2. Figures 5 and 6 show the
corresponding 14N2 argon matrix experiments with osmium and
ruthenium, respectively. All absorptions and isotopic counter-
parts are listed in Tables 1-4.
Experimental Section
The technique for laser ablation and FTIR matrix investigation
has been described previously.8,9 Osmium (E-VAC Products,
99.98%, hot pressed) and ruthenium (Goodfellow Metals,
99.9%) metal targets were mounted on a rotating (1 rpm)
stainless steel rod. The Nd:YAG laser fundamental (1064 nm,
10 Hz repetition rate, 10 ns pulse width, 40-50 mJ pulses) was
focused on the target through a hole in the CsI cryogenic
window (maintained at 7-8 K by a Displex refrigerator). Metal
atoms were co-deposited with 14N2 and 14N2 + 15N2 in argon
Density functional theory in the Gaussian 98 program was
employed to calculate structures and frequencies for ruthenium
and osmium nitride product molecules to verify assignments
and to provide information on their ground-state properties.10
The B3LYP functional, 6-311+G(3d) basis set for nitrogen, and
Los Alamos ECP plus DZ basis set for ruthenium and osmium
were used in all calculations.11-13 The calculated geometries,
relative energies, and calculated frequencies are given in Tables
5 and 6, respectively.
10.1021/jp993338s CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/25/2000

Laser-Ablated Osmium and Ruthenium Atoms
J. Phys. Chem. A, Vol. 104, No. 6, 2000 1153
Figure 1. Infrared spectra in the 1140-860 cm-1 region for laser-ablated osmium atoms co-deposited with nitrogen: (a) after 1 h deposition with
pure 14N2 + 15N2 (43% + 57% mixture), (b) after annealing to 25 K with pure 14N2 + 15N2 (43% + 57% mixture), (c) after 1 h deposition with 2%
14N2 + 2% 15N2 in argon, (d) after annealing to 25 K, (e) after annealing to 40 K, (f) after annealing to 45 K. The 943 cm-1 band that grows on
annealing is probably due to (N2)xOsO2.
Figure 2. Infrared spectra in the 1000-950 cm-1 and 840-800 cm-1 regions with 0.125 cm-1 resolution for laser-ablated ruthenium atoms co-
deposited with 14N2 + 15N2 after annealing to 25 K.
OsN and RuN. A sharp, intense band observed at 1049.6
cm-1 on deposition of osmium with pure nitrogen grows on
early annealing but decreases markedly on photolysis (not
shown). The 15N2 counterpart for this band at 1016.6 cm-1
defines an isotopic ratio of 1.03246, in close agreement with
the harmonic diatomic ratio of 1.03247, calculated using the
average atomic mass of osmium. No new bands are observed
when osmium is deposited with an equal 14N2 + 15N2 mixture,
indicating that only one nitrogen atom participates in the
vibrational mode (Figure 1). These bands are due to dinitrogen
complexed OsN in a nitrogen matrix. In argon, a weak band is
observed at 1130.3 cm-1, with a 15N2 counterpart at 1094.8 cm-1
that is present initially but disappears on annealing. A second
band grows in at 1052.2 cm-1 with a 15N2 counterpart at 1019.2
cm-1. Both bands have diatomic ratios, and are assigned to
isolated and complexed OsN, respectively.
frequencies are in excellent agreement with the ratios predicted
for harmonic RuN, all being within 0.1% of those values. When
ruthenium atoms are deposited with 15N2, an analogous set of
bands is observed in the range 955.8-951.3 cm-1. Figure 2a
shows the spectrum for the mixed 14N2 + 15N2 sample containing
both features and no intermediate bands. The nitrogen isotopic
ratios for different combinations of ruthenium isotopes have
been calculated and compared to the corresponding values
expected for the harmonic diatomic molecule RuN calculated
using the appropriate masses, and all are found to agree within
0.1%. No intermediate bands are observed when a 14N2 + 15N2
mixture is used, indicating the participation of only one nitrogen
atom, and these bands are assigned to the diatomic molecule
RuN.
Analogous bands due to Ru14N and Ru15N are observed at
984.0 and 954.8 cm-1 in argon matrixes. These are almost
identical to the frequencies in solid nitrogen but are broad and
no isotopic structure can be resolved. These bands are not
present initially but grow in during annealing. Both OsN and
RuN have been studied in detail by Ram, et al. using Fourier
transform emission spectroscopy and high level ab initio
When ruthenium atoms are deposited with pure nitrogen, a
group of bands is observed in the range 984.8-980.5 cm-1
,
the most intense band being 981.5 cm-1 (not shown). These
bands are present on deposition, grow during early annealing,
and decrease on photolysis, and are due to seven isotopes of
ruthenium in natural abundance. The relative intensities of the
bands is exactly what is expected for a mode involving only
one ruthenium atom. The calculated ratios for each pair of
2
2
calculations, and are found to have ∆ and Σ+ ground states,
with the spectroscopic constants ro ) 1.618 Å, ωe ) 1148.0
cm-1, ωexe ) 5.5 cm-1 (OsN) and ro ) 1.574 Å, ωe ) 1122.0

1154 J. Phys. Chem. A, Vol. 104, No. 6, 2000
Citra and Andrews
Figure 3. Infrared spectra in the 2150-2010 cm-1 region for laser-ablated osmium atoms co-deposited with pure 14N2 + 15N2: (a) after 20 min
deposition, (b) after annealing to 25 K, (c) after annealing to 30 K, (d) after annealing to 35 K, (e) after 25 min photolysis, (f) after annealing to
37 K.
Figure 4. Infrared spectra in the 2230-2050 cm-1 region for laser-ablated ruthenium atoms co-deposited with pure 14N2 + 15N2: (a) after 30 min
deposition, (b) after annealing to 25 K, (c) after 30 min photolysis, (d) after annealing to 30 K, (e) after annealing to 35 K.
cm-1, ωexe ) 6.5 cm-1 (RuN).14,15 The gas-phase vibrational
frequencies are much higher than the matrix values, and this is
attributed to complexation of OsN and RuN by dinitrogen, to
give (NN)xOsN and (NN)xRuN. Only bands due to the most
complexed form of the molecule are expected in a nitrogen
matrix where the molecules are surrounded by nitrogen mol-
ecules, but additional absorptions may be observable in argon
due to the uncomplexed and partially complexed molecules. This
is the case for the osmium system, the 1130.3 cm-1 band being
only 7 cm-1 below the (anharmonic) gas-phase frequency for
OsN. The isolated OsN absorptions disappear on annealing and
bands corresponding to the fully complexed molecule grow in,
which are very close to the nitrogen matrix values, as has been
observed in several metal nitride systems.16-19 Vibrations due
to the ligands should also be observable, but unfortunately the
2200-2300 cm-1 region where dinitrogen ligands are expected
is congested and no assignment can be made.
The results of the DFT calculations for the doublet states of
these molecules are summarized in Tables 5 and 6. In both cases
the initial calculations did not give the lowest energy states,

Laser-Ablated Osmium and Ruthenium Atoms
J. Phys. Chem. A, Vol. 104, No. 6, 2000 1155
Figure 5. Infrared spectra in the 2200-2000 cm-1 region for laser-ablated osmium atoms co-deposited with 3% 14N2 in argon: (a) after 2 h
deposition, (b) after annealing to 25 K, (c) after 25 min photolysis, (d) after annealing to 30 K, (e) after annealing to 35 K, (f) after annealing to
40 K, (g) after annealing to 45 K.
Figure 6. Infrared spectra in the 2170-2030 cm-1 region for laser-ablated ruthenium atoms co-deposited with 4% 14N2 in argon: (a) after 1 h
deposition, (b) after annealing to 25 K, (c) after 30 min photolysis, (d) after annealing to 30 K, (e) after annealing to 35 K, (f) after annealing to
40 K.
TABLE 1: Infrared Absorptions (cm-1) Observed Following the Co-deposition of Laser-Ablated Osmium Atoms with Nitrogen
at 7-8 K
14N2
15N2
14N2 + 15N2
isotopic ratio
assignment
2119.7
2118.2
2109.9
2091.8
1061.2
1049.6
908.5
903.4
901.1
521.1
518.2
2049.1
2047.6
2039.4
2022.1
1028.2
1016.6
880.6
875.7
874.1
505.5
502.5
2119.7, 2093.5, 2071.8, 2059.1, 2049.1
2118.2, 2091.8, 2071.8, 2056.4, 2047.6
2109.4, 2039.4
2091.8, 2022.1
1061.2, 1028.2
1049.6, 1016.6
908.5, 892.4, 880.6
903.4, 887.4, 875.5
901.1, 885.5, 874.1
1.03445
1.03448
1.03457
1.03447
1.03209
1.03246
1.03168
1.03163
1.03089
1.03086
1.03124
Os(NN)5 (site)
Os(NN)5
Os(NN)x (X<4)
Os(NN)x (X<4)
OsN (site)
OsN
NOsN (site)
NOsN (site)
NOsN
Os(NN)5 (site)
Os(NN)5
521.1, 513.7, 505.5
518.2, 511.1, 502.5
and different initial orbital occupations were used in subsequent
calculations to find the ground states for OsN and RuN which
DFT predicts to be 2∆ and 2Σ+, respectively, in agreement with
experiment.14,15 The calculated bond lengths of 1.621 and 1.581
Å are also in excellent agreement with the experimental values,
being too long by only 0.003 and 0.007 Å, respectively. The
calculated harmonic frequencies are higher than the gas-phase
values, and must be scaled by 0.942 and 0.945, which are

1156 J. Phys. Chem. A, Vol. 104, No. 6, 2000
Citra and Andrews
TABLE 2: Infrared Absorptions (cm-1) Observed following
the Co-deposition of Laser-Ablated Osmium Atoms with
Nitrogen/Argon Mixtures at 7-8 K
effect. The force constant is lowest for FeN, increasing by ∼21%
moving down each row in the periodic table. The increase in
bond strength moving down the group is consistent with the
fact that 5d metals form stronger bonds than the 3d or 4d metals
in the same group with hydrides, phosphines, and other ligands.21
The force constants for 192OsN and 102RuN have also been
calculated using the gas-phase harmonic frequencies, and are
included in Table 7. These are larger than the matrix values,
which means that the M-N bond is significantly weakened
when additional N2 molecules bind to the metal center. No
experimental data is available for isolated FeN, only complexed
FeN was observed in the previous matrix isolation study of the
iron nitrides. On the basis of the osmium and ruthenium results,
a decrease of 20-30% would be expected in the force constant
for (N2)xFeN relative to FeN.
14N2
15N2
14N2 + 15N2
isotopic ratio assignment
2186.8
a
a
2138.9 2067.6 2138.9, 2067.6a
1.03448
1.03441
Os(NN)5
Os(NN)4
Os(NN)2
Os(NN)3
Os(NN)
OsN
2106.0
2083.2 2013.9 2083.1, 2033.8, 2013.9
2070.0
2044.2 1976.2 2044.2, 1976.2
1130.3 1094.8 1130.3, 1094.8
1052.2 1019.2 1052.2, 1019.2
900.4 872.9 900.4, 884.7, 872.9
a
a
a
a
1.03441
1.03243
1.03238
1.03150
(NN)xOsN
NOsN
a Spectra were too congested to observe these bands.
reasonable scale factors for the B3LYP functional, basis set,
and ECP used.20 It is clear from these results that DFT
reproduces the experimental results very effectively. The
calculated parameters for two higher lying doublet states of OsN
and RuN are also included in Tables 5 and 6 which are in
reasonable agreement with those calculated using high level ab
initio methods.14,15
NOsN and NRuN. When osmium atoms are deposited with
pure nitrogen, a weak band is observed at 901.2 cm-1 that grows
in strongly on annealing and is diminished on photolysis (not
shown). A less intense matrix site at 908.3 cm-1 is diminished
on annealing but grows in on photolysis. The 15N2 counterparts
for these bands are at 880.4 and 873.8 cm-1, respectively, and
give almost identical isotopic ratios of 1.03136 and 1.03169
for the different matrix sites. These are lower than the diatomic
value of 1.03216 indicating a greater participation of the metal
atom in the mode relative to the diatomic molecule. When the
It is interesting to compare the force constants for OsN, RuN,
and FeN, which are shown in Table 7. These were calculated
using the matrix frequencies for the complexed molecules
(NN)xMN as discussed above. It is still valid to calculate the
complexed MN force constant using the equation for a harmonic
diatomic molecule provided that the mechanical coupling
between the MN group and the rest of the molecule is weak.
This is found to be the case for OsN where the isotopic ratios
are almost identical for isolated OsN in argon (1.03243) and
fully complexed OsN in nitrogen (1.03246), both being close
to the harmonic diatomic value of 1.03247. This is also apparent
from the lack of a secondary isotopic effect for the bands due
to (NN)xOsN and (NN)xRuN in pure nitrogen matrixes. Thus,
changes in frequencies as the molecules are complexed by
dinitrogen are due to an electronic effect and not a mechanical
experiment is repeated using a 14N2
+
15N2 mixture, an
intermediate band is observed at 885.6 cm-1 with a weaker
matrix site at 892.2 cm-1, which are more intense than either
of the pure isotope bands (Figure 1a). The intensity pattern for
this triplet of peaks is appropriate for a vibrational mode that
involves two equivalent nitrogen atoms, and NOsN is the most
logical assignment for these bands. The fact that the intermediate
band is observed shows that the molecule is formed through
reaction 1, rather than reaction 2, since no isotopically scrambled
14N15N is present to form 14NOs15N by direct insertion. This is
reasonable considering the abundance of OsN and nitrogen
atoms in the pure nitrogen matrix, the latter being evident from
TABLE 3: Infrared Absorptions (cm-1) Observed following the Co-deposition of Laser-Ablated Ruthenium Atoms with
Nitrogen at 7-8 K
14N2
15N2
14N2 + 15N2
isotopic ratio
assignment
2151.2
2149.3
2147.5
2147.0
2003.0
1932.2
1874.9
1657.7
984.8
983.6
983.1
982.5
982.0
981.5
980.5
838.4
837.8
837.1
836.5
835.3
833.4
832.7
832.1
831.5
830.2
499.0
2079.6
2077.6
2076.1
2075.5
1937.2
1868.1
1841.8
1603.3
955.8
b
954.0
953.4
952.9
952.4
951.3
b
2150.0, 2121.6, 2099.5, 2087.0, 2078.4
1.03443
1.03451
1.03439
1.03445
1.03397
1.03431
1.01797
1.03393
1.03034
Ru(NN)5 site
Ru(NN)5
Ru(NN)5 site
Ru(NN)5 site
-
2003.0, 1992.9, 1948.8, 1937.2
a
1874.9, 1841.8
1657.7, 1649.4, 1613.0, 1603.3
984.8, 955.8
983.6, b
983.1, 954.0
982.5, 953.4
982.0, 952.9
981.5, 952.4
980.5, 951.3
N3
NO
N3
96RuN
98RuN
1.03050
1.03052
1.03054
1.03055
1.03069
99RuN
100RuN
101RuN
102RuN
104RuN
838.4, b
837.8, b
837.1, b
836.5, b
N99RuN (site)
N100RuN (site)
N101RuN (site)
N102RuN (site)
N104RuN (site)
N99RuN
b
b
b
b
835.3, b
810.1
809.4
808.7
808.1
806.8
486.2
833.4, 820.3, 810.1
832.7, 819.6, 809.4
832.1, 819.0, 808.7
831.5, 818.3, 808.1
830.2, 817.1, 806.8
499.0, 493.4, 486.2
1.02876
1.02879
1.02894
1.02896
1.02900
1.02633
N100RuN
N101RuN
N102RuN
N104RuN
Ru(NN)5
a Spectra were too congested to observe these bands. b Bands could not be resolved.

Laser-Ablated Osmium and Ruthenium Atoms
J. Phys. Chem. A, Vol. 104, No. 6, 2000 1157
TABLE 4: Infrared Absorptions (cm-1) Observed following
the Co-deposition of Laser-Ablated Ruthenium Atoms with
Nitrogen/Argon at 7-8 K
dinitrogen molecule and break the strong triple bond in a cold
argon matrix. This is a significant result, as it shows that the
reaction does not require actiVation energy and is thermody-
namically faVorable. Since the temperature in the matrix is low,
the entropy contribution to the free energy change is negligible
compared to the enthalpy change. Given that the bond enthalpy
in N2 is 945 kJ/mol,23 the average Os-N bond enthalpy in
NOsN must be g473 kJ/mol. The high affinity of osmium for
dinitrogen shown here parallels the high activity of the metal
as a catalyst in ammonia synthesis.6 It could be argued that
osmium reacts with dinitrogen in two steps, rather than a direct
insertion. The first step would be to form OsN and N, which
would then combine to give NOsN. If the OsN and N were
effectively trapped together in the argon cage then little
14NOs15N would be expected in the mixed experiment using
14N2 + 15N2. However, for the first step to be thermodynamically
favorable, the OsN bond strength would have to exceed the N2
bond strength, and this is not expected to be the case as is
discussed below. This reaction can occur on deposition with
energetic metal atoms, but not on annealing with cold metal
atoms.
The growth of NOsN on annealing in solid argon and the
near agreement of NOsN frequencies in solid argon and nitrogen
suggests that the NOsN molecule may be ligated by dinitrogen
but to a lesser degree than (NN)xOsN in dinitrogen. However,
dinitrogen ligation of the NOsN insertion product may provide
thermodynamic assistance for the insertion reaction 2.
The 14/15 isotopic ratios of 1.03136 and 1.03115 for these
bands in N2 and N2/Ar can be used in conjunction with the
appropriate G matrix elements to estimate the bond angle for
the bent symmetrical molecule.24,25 If the observed bands are
14N2
15N2
14N2 + 15N2
isotopic ratio
assignment
2148.8 2077.3
2126.5 2055.7
2119.3 2048.8
2091.2 2021.5
2077.6 2008.6
2034.6 1967.1 2034.6, 1967.1
984.0 954.8 984.0, 954.8
a
a
a
a
a
1.03442
1.03444
1.03441
1.03448
1.03435
1.03431
1.03058
1.02907
1.02899
Ru(NN)5
Ru(NN)4
Ru(NN)4 (site)
Ru(NN)3
Ru(NN)2
Ru(NN)
RuN
831.9 808.4 831.9, 818.7, 808.4
830.5 807.1 830.5, 817.4, 807.1
N102RuN
N104RuN
a Spectra were too congested to observe these bands.
the band at 1657.4 cm-1 due to N3.16,22
OsN + N f NOsN
Os + N2 f NOsN
(1)
(2)
Analogous bands are observed at 900.3, 884.7, and 872.9
cm-1 when osmium is co-deposited with 14N2 + 15N2 in argon,
very close to the pure nitrogen values (as was found for OsN).
However, the annealing behavior for these bands in argon is
different from that in pure nitrogen (Figure 1). The bands are
not present initially, but grow in during annealing. The
intermediate band also grows in, but not to the same extent,
and after the final annealing cycle the intensity ratio is
approximately 3:1:3. This implies that reaction 2 is dominant,
but that reaction 1 also occurrs to a small extent. These
observations demonstrate that osmium atoms insert into the
TABLE 5: Calculated States, Geometries, and Relative Energies for Osmium Nitride Species
energy
(kJ/mol)
molecule
OsN
state
S2
geometry (Å, deg)
frequencies (cm-1) [intensities (km/mol)]
2
∆
0
+70
0.7503
0.7500
0.7755
r(Os-N): 1.6208
r(Os-N): 1.6123
r(Os-N): 1.6797
r(Os-N): 1.6825,
NOsN: 105.1°
r(Os-N): 1.7000
NOsN: 111.8°
r(Os-N): 1.7305,
NOsN: 123.5°
r(Os-N): 1.639
r(Os-Os): 2.412
NOsOs: 103.1°
φ(NOsOsN): 109.6°
1219.1 [39]
1230.0 [36]
1036.3 [28]
2Σ+
2Π
+160
+110
NOsN
1A1
1121.6 [6], 1005.0 [39], 457.5 [0]
3B1
5A2
1A
+184
+244
-205a
2.0000
6.0001
1053.8 [4], 891.6 [109], 396.1 [0]
977.2 [27], 567.3 [2], 269.5 [2]
NOsOsN
1172.7 [95], 1159.4 [7], 243.3 [4], 198.4 [0], 167.6 [17], 136.0 [3]
a Relative to twice the energy of OsN (2∆).
TABLE 6: Calculated States, Geometries, and Relative Energies for Ruthenium Nitride Species
energy
(kJ/mol)
molecule
RuN
state
S2
geometry (Å, deg)
frequencies (cm-1) [intensities (km/mol)]
2Σ+
0
+62
0.7500
0.7557
0.7519
r(Ru-N): 1.5810
r(Ru-N): 1.6386
r(Ru-N): 1.6280
r(Ru-N): 1.6864,
NRuN: 107.2°
r(Ru-N): 1.687
NRuN: 113.0°
r(Ru-N): 1.724,
NRuN: 116.5°
r(Ru-N): 1.812
r(Ru-Ru): 2.746
RuNRu: 98.5°
1187.4 [150]
1028.3 [29]
1024.8 [156]
2Π
2
∆
+76
NRuN
1A1
3B1
5A1
3B2
+364
1041.2 [10], 856.0 [100], 409.2 [6]
+351
+396
-234a
2.0002
6.0005
2.0004
927.0 [7], 897.2 [4], 261.1 [0]
984.7 [42], 378.2 [5], 249.2 [15]
Ru2N2
807.7 [15], 780.4 [4], 597.7 [88], 481.6 [0], 362.3 [12], 235.3 [34]
φ(NRuRuN): 155.6°
a Relative to twice the energy of RuN (2Σ+).

1158 J. Phys. Chem. A, Vol. 104, No. 6, 2000
Citra and Andrews
TABLE 7: Calculated Force Constants (nm-1) for MN,
lower limits of 121° and 110°, respectively, very similar to the
values for NRuN and NOsN. There is enough isotopic data
available to calculate the force constants for these molecules,
using the best estimates for the bond angles and the appropriate
atomic masses to calculate the G matrix elements,28 and the
results are shown in Table 7. All resolvable bands due to the
different isotopes for iron and ruthenium in solid nitrogen were
used, and stated force constants are averages over two (NFeN)
and five (NRuN) values calculated using different isotopes. The
metal isotopes for osmium could not be resolved and the average
atomic mass for osmium was used. Considering also the
uncertainty in the bond angle, the force constants are expected
to be in greatest error for NOsN. The force constants are less
than those for the corresponding diatomic molecules but show
the same trend, increasing down the metal group.
(N2)xMN, and NMN (M ) Fe, Ru, Os)
NMN
metal
iron
ruthenium
osmium
MN fM-N
fM-N
fM-N/M-N
575
554
588
722
58
115
115
702 (913)a
851 (1013)b
a Force constant calculated using the gas-phase data in ref 14. b Force
constant calculated using the gas-phase data in ref 15.
due to the ν3 mode then upper limit estimates of 120° and 126°
are derived for NOsN in nitrogen and argon matrixes, respec-
tively. This is consistent with the large number of NMN
molecules characterized, which have similar angles.7,16-19 The
bent geometries for most of these molecules agree with the
predictions of a Walsh-like treatment of AB2 triatomics where
the central atom is a transition metal, and molecules with less
than 19 valence electrons are bent.26 The isotopic ratio for the
terminal atoms in the ν3 mode of a triatomic molecule gives an
upper limit on the bond angle.24,25 The isotopic ratio for the
central atom would give a lower limit on the angle, but
unfortunately the osmium isotopes in NOsN could not be
resolved. The bands in the nitrogen matrix are sharper and their
positions measured with greater accuracy, so the angle calculated
using these numbers is considered to be more reliable. Con-
sidering the results for iron and ruthenium (see below), the best
estimate for the NOsN bond angle is 115 ( 5°.
These force constants may be used in conjunction with the
spectroscopic parameters derived previously for OsN and
RuN14,15 to estimate bond energies. The ωe and ωexe values for
a diatomic molecule can be used to estimate De analytically if
a Morse potential is assumed.29 In this way, values of 722 and
579 kJ/mol are estimated for De(OsN) and De(RuN). The force
constant calculated for NOsN is 15.2% lower than that for OsN,
and if it is assumed that the bond energies scale the same way,
the bond dissociation energy in NOsN is estimated to be 613
kJ/mol, well above the lower limit of 473 kJ/mol predicted
earlier. The force constants for OsN and NOsN in nitrogen,
where both are complexed by dinitrogen, were used in making
the comparison, since there is no force constant for NOsN with
which to compare that for OsN in the gas phase. If the force
constant treatment is applied to the ruthenium data, a bond
dissociation energy of 485 kJ/mol is estimated for each bond
in NRuN. This upper limit is only slightly above the lower limit
of ∼473 kJ/mol required for a spontaneous insertion reaction,
and so direct insertion by ruthenium into the dinitrogen bond
is not likely on annealing. It is clear from this discussion that
the Os-N bond energy in NOsN is well above the lower limit
of ∼473 kJ/mol required for the insertion of osmium into the
dinitrogen bond to be thermodynamically favorable. The Ru-N
bond energy in NRuN is estimated to be much closer to this
lower limit, and the available experimental data indicates that
it is too low to provide a driving force for the spontaneous
insertion by ruthenium into the dinitrogen bond.
When ruthenium is co-deposited with pure 14N2, a group of
five bands is observed in the range 833.4-830.2 cm-1 with the
same characteristic intensity pattern observed for RuN, indicat-
ing that only one ruthenium atom is involved in the mode. These
bands correspond to the five heaviest isotopes of ruthenium,
the bands due to other isotopes not being observed. Weaker
matrix sites to these bands are observed in the range 838.4-
835.3 cm-1, analogous to the higher frequency matrix site
observed for NOsN. The 15N counterparts for the dominant
matrix site are in the range 810.1-806.8 cm-1, and when the
experiment is repeated with a 14N2 + 15N2 mixture intermediate
bands are observed in the range 820.3-817.1 cm-1 to give an
overall intensity pattern that is appropriate for a mode that
involves two equivalent nitrogen atoms. The most straightfor-
ward assignment for these bands is to NRuN, formed by the
addition of nitrogen atoms to RuN molecules. Unlike NOsN,
both metal and nitrogen isotopic data are available to estimate
both lower and upper limits for the bond angle, respectively.24,25
Using of all the available isotopic data allows upper and lower
limits of 117° and 109° to be determined for the N-Ru-N
bond angle. The average value of 113° represents a reliable
estimate of the bond angle owing to cancellation of anharmonic
effects.
Bands due to NRuN are observed in argon matrixes at 831.9/
830.5, 818.7/817.4, and 808.4/807.1 cm-1 where bands due to
the 104Ru and 102Ru isotopes can be resolved. However these
bands are weak and broad, and the greater uncertainty in their
positions would lead to an inferior bond angle estimate
compared to the nitrogen matrix data where the bands were
sharp and intense. On annealing the argon matrix, the bands
due to 14NRu14N, 15NRu15N, and 14NRu15N grow slightly but
the same proportion, indicating that the molecule is being formed
by the addition of a nitrogen atom to RuN rather than the direct
insertion of ruthenium into a dinitrogen molecule.
The results of the DFT calculations for the triatomics are
given in Tables 6 and 7. NOsN is found to have an 1A1 ground
state with a bond angle of 105.1°, in reasonable agreement with
the estimated experimental value. The calculated ν3 frequency
is significantly lower than the ν1 frequency, and has a much
higher intensity. This is in agreement with experiment where
the ν1 mode could not be observed, and the intermediate band
in the mixed isotopic experiment is red shifted from the midpoint
of the pure isotope bands. The ν3 frequency is higher than the
experimental value, and a scale factor of 0.897 is required, which
is reasonable for the density functional and basis set used.17
3
The results for NRuN are more ambiguous. The B1 state is
lowest in energy, and the bond angle of 113.0° is in excellent
agreement with the value of 112.8° calculated from the isotopic
data. However, a large negative frequency is calculated for this
state and so it must be rejected. The triplet states 3A1, 3A2, and
3B2 were calculated by altering the MO occupations, but these
are all more than 65 kJ/mol higher in energy than the 3B1 state.
1
3
The A1 state for NRuN is only 13 kJ/mol higher than the B1
state, and the calculated geometry is in reasonable agreement
with the experimental value. The calculated frequencies for this
state are also reasonable, the ν3 mode has a much higher
The iron dinitride NFeN was observed previously in this
laboratory and the bond angle was calculated to be 115 ( 5°
using nitrogen and iron isotopic data, which gave upper and

Laser-Ablated Osmium and Ruthenium Atoms
J. Phys. Chem. A, Vol. 104, No. 6, 2000 1159
TABLE 8: NBO Bonding Analysis for Osmium and Ruthenium Nitride Products
molecule
bond
populationa
%Mb
%nsc
%npc
%(n-1)dc
%Nb
%2sc
%2pc
99.3
OsN (2∆) R
Os-N (π)
Os-N (π)
1.00 (0.00)
1.00 (0.01)
Os-N (σ)
1.00 (0.00)
1.00 (0.00)
Os-N1 (“π”)
1.89 (0.16)
1.69 (0.27)
2.00 (0.00)
1.89 (0.16)
1.69 (0.27)
Ru-N (π)
1.00 (0.00)
1.00 (0.02)
Ru-N (σ)
1.00 (0.00)
1.00 (0.00)
Ru-N1 (“π”)
1.96 (0.26)
1.93 (0.51)
2.00 (0.00)
1.92 (0.19)
1.00 (0.00)
52.9
43.5
1.00 (0.01)
44.3
44.3
2.00 (0.00)
36.0
33.9
58.3
36.0
33.9
1.00 (0.00)
50.3
41.1
1.00 (0.02)
48.0
48.0
2.00 (0.00)
41.4
47.2
64.0
52.9
0.0
36.5
42.3
0.0
0.0
0.0
0.6
30.2
0.0
0.0
0.0
3.2
12.3
0.0
3.2
12.3
0.0
0.0
0.7
1.3
0.0
0.0
0.0
0.4
0.2
0.0
0.4
0.0
100.0
62.9
0.6
100.0
100.0
0.0
63.7
85.8
100.0
63.7
85.8
0.0
100.0
56.0
0.1
100.0
100.0
0.0
67.9
88.7
100.0
53.8
100.0
47.1
56.5
69.2
55.7
55.7
100.0
64.0
66.1
41.7
64.0
66.1
100.0
49.7
58.9
98.6
52.0
52.0
100.0
58.6
52.8
36.0
62.3
47.1
0.0
17.4
57.7
0.0
0.0
99.3
82.2
17.6
99.4
99.4
0.0
81.8
99.2
100.0
81.8
99.2
0.0
99.5
82.7
10.2
99.5
99.5
0.0
85.1
99.5
99.3
85.1
Os-N (σ)
OsN (2∆) â
Os-N (π)
81.9
Os-N (π)
0.0
0.0
NOsN (1A1)
Os-N1 (“σ”)
Os-N1 (“σ+π”)
Os-N2 (“π”)
Os-N2 (“σ”)
Os-N2 (“σ+π”)
RuN (2Σ+) R
Ru-N (π)
58.3
33.1
1.9
0.0
33.1
1.9
50.3
0.0
43.3
50.8
0.0
41.7
17.7
0.1
0.0
17.7
0.1
49.7
0.0
17.1
49.2
0.0
100.0
99.5
89.3
99.3
Ru-N (σ)
RuN (2Σ+) â
Ru-N (π)
Ru-N (π)
0.0
0.0
NRuN (1A1)
Ru-N1 (“σ”)
Ru-N1 (“σ+π”)
Ru-N2 (“π”)
Ru-N2 (“σ”)
64.0
31.7
11.1
0.0
36.0
14.5
0.1
0.0
14.5
37.7
45.8
a The number in parentheses is the population of the corresponding antibond. b The total contribution of the atom’s orbitals to the bond.
c Hybridization of the bond at the atom.
intensity and is lower in frequency than the ν1 mode. The
calculated frequency is only slightly higher than the experimental
value, and the scale factor for this mode is 0.97. This is much
higher than that for NOsN which is unexpected but does not
necessarily mean it is wrong, just that DFT is an approximate
and reactions between two open shell species do not usually
require activation energy. The absence of dimerization reactions
between the nitrides can be explained by the fact that these
molecules are heavily ligated in argon and nitrogen matrixes,
and the combination reaction would also involve a rearrange-
ment of the surrounding dinitrogen ligands. The more tightly
the ligands in these complexes are bound, the more hindered
the combination reaction will be. The strength of the bonding
in these complexes is of course unknown, but can be inferred
from the red shift in metal nitride frequency when the isolated
molecule is complexed by dinitrogen. For example, the yttrium
and rhodium nitrides YN and RhN were only weakly complexed
by dinitrogen, as shown by red shifts of only 29 and 29.7 cm-1
between the isolated diatomic molecules in argon and their
dinitrogen complexed forms.18,30 By contrast, the Os-N fre-
quency is red shifted by 78.1 cm-1, and the tighter binding in
this complex that this would suggest may be enough to
completely hinder the dimerization reaction.
Bonding Analyses. The bonding in the diatomic and triatomic
molecules has been examined using the Natural Bond Orbitals
(NBO),31 and the results are summarized in Table 8. Both OsN
and RuN have triple bonds, consisting of one σ and two π bonds.
However, the force constants calculated earlier show that the
bond strengths are quite different despite having the same bond
orders. NOsN is predicted to consist of two triple bonds, each
formed from a sigma type bond, a pi type bond, and a third
bond that is a mixture of σ and π components. NRuN is similar,
except that there is only one sigma/pi mixed bond to shared
between both bonds, to give a bond order of two and a half
rather than three. The M-N bonds are not very polarized toward
either atom and have no significant ionic character, in agreement
with studies of other transition metal and lanthanide nitrides.32,33
Os(NN)n and Ru(NN)n. When osmium is co-deposited with
pure nitrogen, intense bands are observed at 2119.7 and 2118.2
cm-1. These bands decrease slowly on annealing but are
destroyed on photolysis. The 15N2 counterparts are at 2049.1
and 2047.6 cm-1, giving isotopic ratios of 1.03445, and 1.03448
which are typical of dinitrogen stretching modes. These are
clearly due to an osmium dinitrogen complex in different matrix
sites. When osmium is co-deposited with a 14N2 + 15N2 mixture,
5
treatment. The lowest quintet state, A2, is 45 kJ/mol higher
than the 3B1 state which is not unreasonably high, but must be
rejected since the ν3 frequency and intensity is unrealistically
low. In summary, the ground states of both NOsN and NRuN
1
are found to be A1, although further calculations on these
molecules would be useful.
Os2N2 and Ru2N2. It could be argued that the bands assigned
above to NOsN and NRuN are due to dimers of the diatomic
molecules, since the observation of the scrambled isotopic
molecule in the mixed 14N2
+
15N2 experiment would be
consistent with the combination of two diatomic molecules.
However, this assignment must be rejected in light of the
annealing behavior of the bands in an argon matrix. If a dimer
were formed via the combination of OsN molecules, then a 1:2:1
intensity pattern would be observed. There is no way to reconcile
assignment of the observed bands to (OsN)2 with the low
intensity of the intermediate bands relative to the pure isotope
bands in argon. The growth of the analogous bands in the
ruthenium nitride system on annealing is consistent with the
formation of (RuN)2, but the ruthenium isotopic data clearly
show that only one ruthenium atom is present.
Both cyclic and open structural isomers with the formulas
Os2N2 and Ru2N2 have been calculated using DFT, and the
lowest energy geometries for each metal are given in Tables 5
and 6. It is interesting that these molecules are not observed
despite the abundance of diatomic molecules and the calculated
energy changes for the dimerization reactions (-205 kJ/mol for
OsN, -234 kJ/mol for RuN). This suggests that either the
diatomic molecules are not very mobile or that the reaction is
an activated process that cannot occur at such low temperatures.
Low mobility of isolated species is unlikely to be the cause
considering the number of aggregate species that are observed
in so many other metal nitride and oxide systems. A significant
activation energy would also be surprising since OsN and RuN
have doublet ground states, ie they are both open shell species,

1160 J. Phys. Chem. A, Vol. 104, No. 6, 2000
Citra and Andrews
weaker intermediate bands are observed at 2094.8/2091.8,
2071.8, and 2059.1/2056.4 cm-1 (Figure 3). The presence of
what appears to be three intermediate bands that are weaker
than the pure isotope absorptions suggests that four dinitrogen
units are bound to the metal in a tetrahedral geometry.34 Weaker
bands are observed in the range 2200-2125 cm-1 that are
probably due to the mixed isotope symmetric stretching modes,
though specific assignments cannot be made with certainty. A
much weaker band is observed at 518.2 cm-1 that tracks with
the high-frequency bands; the 15N2 counterpart for this band at
502.5 cm-1 gives an isotopic ratio of 1.03124, which is
appropriate for an M-N2 stretching mode of a dinitrogen
complex. One intermediate band is observed in the mixed
isotopic experiment at 511.1 cm-1 to give a 1:2:1 intensity
cm-1 (B, Figure 5), and 2186.8 cm-1 (F, Figure 5) on annealing,
but the 15N2 counterparts or intermediate components could not
be determined due to spectral congestion. However, Os(NN)3
and Os(NN)4 must be present, and the bands B and D,
respectively, are tentatively assigned to these species. No
assignment is suggested for the band at 2186.8 cm-1. If it tracked
with the 2138.9 cm-1 band, it could be due to the out-of-phase
stretching mode of the axial ligands in Os(NN)5, but this is not
the case.
When ruthenium is co-deposited with pure nitrogen, intense
bands are observed at 2151.2, 2149.3, 2147.5 and 2147.0 cm-1
which are due to a ruthenium dinitrogen complex in different
matrix sites. The 15N2 counterparts for these bands are at 2079.6,
2077.6, 2076.1, and 2075.5 cm-1 respectively giving isotopic
ratios that are typical for dinitrogen complexes. These bands
decrease on annealing and photolysis. When the experiment is
pattern.
A better assignment for these bands is to
Os(N2)5, which is expected to be the most highly coordinated
osmium dinitrogen complex based on the previous matrix
isolation study of the iron dinitrides, the known carbonyl
chemistry of the iron group, and the 18 electron rule.7,35-37 Two
infrared active dinitrogen vibrations are expected for a trigonal
bipyramidal five coordinate complex; the out of phase stretch
of the axial ligands, a2′′, and the degenerate stretching mode of
the equatorial ligands, e′. It is questionable whether the 2056.4
cm-1 band is actually an intermediate component, and if this is
ignored then the observed nonbinomial intensity pattern is
consistent with the degenerate stretching mode of the equatorial
ligands in a complex with D3h symmetry.34 The second mode
expected for this geometry is not observed, and this may be
due to an interaction between the complex and the host matrix
or an inherently low intensity. If this assignment is correct, then
the 1:2:1 intensity pattern observed in the mixed isotopic
experiment near 500 cm-1 would be due to the out of phase
Os-N stretching modes for the axial ligands, which may have
relatively high intensity at the expense of the axial N-N
stretching mode intensities. Indirect support for this assignment
comes from a theoretical study of the iron dinitrogen complexes
repeated with a 14N2
+
15N2 mixture what look like three
intermediate bands are observed at 2121.6, 2099.5, and 2087.0
cm-1 which are less intense than either of the pure isotope bands
(Figure 4). The lowest frequency component at 2087.0 cm-1 is
probably not an intermediate component, and these bands are
assigned to the degenerate stretching mode in Ru(NN)5,
analogous to osmium. The second dinitrogen stretching mode
expected for this molecule is not observed, but a band that gives
a 1:2:1 triplet in the mixed isotopic experiment is observed at
499.0 cm-1 that tracks with the high-frequency bands, also
analogous to the osmium dinitrogen system.
When ruthenium is co-deposited with nitrogen in argon
(Figure 6), prominent bands are observed at 2034.6 (A), 2077.6
(B), 2091.2 (C), 2126.5/2119.3 (D), and 2148.8 cm-1 (E), with
15N2 counterparts at 1967.1, 2008.6, 2053.0, and 2077.3 cm-1
.
No intermediate bands are observed for the 2034.6/1967.1 cm-1
pair, and these bands are assigned to Ru(NN). The spectra are
too confused to be able to identify intermediate components
for the other bands, but it is quite likely that these correspond
to Ru(NN)2, Ru(NN)3, Ru(NN)4, and Ru(NN)5 as the latter
species absorbs at essentially the same frequency in argon as
in pure nitrogen.
Fe(NN)1-5.
38 The a2′′ dinitrogen stretching mode in Fe(NN)5 is
predicted to have less than one-seventh the intensity of the e′
mode, and is only slightly higher in intensity than the a2′′ metal-
dinitrogen stretching mode. It is likely that the results are similar
for the other metals in the iron group, and only a small error in
the calculation or a small perturbation by the surrounding matrix
would be required to alter the intensities of the high and low
frequency a2′′ modes such that only the lower a2′′ mode is
observable.
This discussion provides a reasonable interpretation of the
dinitrogen complex observations, but without further information
many of these assignments are only tentative, and a more
thorough study of the osmium and ruthenium dinitrogen
complexes will be required before more definitive assignments
can be made.
Conclusions
When osmium is co-deposited with nitrogen in argon, bands
are observed at 2083.1 cm-1 (C, Figure 5) and 2044.2 cm-1
Laser-ablated osmium and ruthenium atoms were reacted with
nitrogen molecules to form the diatomic and triatomic nitrides
MN and NMN and numerous dinitrogen complexes M(NN)n.
Osmium is obserVed to insert into the dinitrogen bond on
annealing in the cold matrix, which shows that Os will be a
good nitrogen fixation catalyst and allows a lower limit on the
average Os-N bond enthalpy in NOsN to be established,
namely, g0.5 × D(NtN). DFT calculations were effective in
reproducing the experimental observations for the nitride and
dinitrides, but further investigation of the dinitride molecules
will be useful. Comparison of the force constants in the iron
group for MN and NMN show that the bonds get stronger
moving down the group from iron to ruthenium to osmium, the
diatomic force constants are greater than those for NMN by
∼18% for ruthenium and osmium. Complexation of the diatomic
molecules by dinitrogen reduces the force constants in OsN and
RuN considerably. Examination of the MN and NMN molecules
using NBO analyses show that the nitrides are all covalently
(A, Figure 5), with 15N2 counterparts at 2013.9 and 1976.2 cm-1
.
Both bands are diminished on annealing and on photolysis. The
2083.1/2013.9 cm-1 pair has an intermediate band at 2033.8
cm-1 in the mixed isotopic experiment with approximately twice
the intensity of the pure isotope peaks, indicating that two
dinitrogen units are involved in the vibration. These bands are
therefore assigned to Os(NN)2. No intermediate bands are
observed for the 2044.2/1976.2 cm-1 pair in the mixed isotope
experiment, and these bands are assigned to the mono dinitrogen
complex Os(NN). An intense band grows in strongly during
annealing at 2138.9 cm-1 (E, Figure 5), with a 15N2 counterpart
at 2067.6 cm-1. The spectra are too congested to be able to
identify the intermediate components in the mixed isotopic
experiment, but these bands must correspond to the highest
coordination complex and are therefore assigned to Os(NN)5,
blue shifted by 19 cm-1 with respect to the nitrogen matrix
value. Other bands grow in at 2106.0 cm-1 (D, Figure 5), 2070.0

Laser-Ablated Osmium and Ruthenium Atoms
J. Phys. Chem. A, Vol. 104, No. 6, 2000 1161
(14) Ram, R. S.; Lievin, J.; Bernath, P. F. J. Chem. Phys. 1999, 111,
bound with triple bonds, except for NRuN which has Ru-N
bond orders of 2.5.
3449.
(15) Ram, R. S.; Lievin, J.; Bernath, P. F. J. Chem. Phys. 1998, 109,
6329.
(16) Andrews, L.; Bare, W. D.; Chertihin, G. V. J. Phys. Chem. A 1997,
101, 8417.
(17) Zhou, M.; Andrews, L. J. Chem. Phys. Chem. A 1998, 102, 9061.
(18) Citra, A.; Andrews, L. J. Phys. Chem. A 1999, 103, 3410.
(19) Andrews, L. J. Electron. Spectrosc. 1998, 97, 63.
(20) Bytheway, I.; Wong, M. W. Chem. Phys. Lett. 1998, 282, 219.
(21) Collman, J. P.; Hegedus, L. Principles and Applications of Organo
Transition Metal Chemistry; University Science Books: Menlo Park, CA,
1980.
Dinitrogen complexes are observed for both metals, and on
the basis of the mixed isotopic frequency and intensity patterns
the most highly coordinated complexes in pure nitrogen are
probably Os(NN)5 and Ru(NN)5, analogous to the pentacarbonyl
complexes already known for these metals.
Acknowledgment. We gratefully acknowledge N.S.F. sup-
port for this research under Grant CHE 97-00116.
(22) Tain, R.; Facelli, J. C.; Michl, J. J. Phys. Chem. 1988, 92, 4073.
(23) Huber, K. P.; Herzberg, G. Constants of Diatomic Molecules; van
Nostrand Reinhold: New York, 1979.
(24) Allavena, M.; Rysnik, R.; White, D.; Calder, V.; Mann, D. E. J.
Chem. Phys. 1969, 50, 3399.
(25) Green, D. W.; Ervin, K. M. J. Mol. Spectrosc. 1981, 88, 51.
(26) Van Zee, R. J.; Hamrick, Y. M.; Li, S. Weltner, W., Jr. J. Phys.
Chem. 1992, 96, 7247.
(27) Andrews, L.; Souter, P. F.; Bare, W. D.; Liang, B. J. Phys. Chem.
A 1999, 103, 4649.
(28) Wilson, E. B., Jr.; Decius, J. C.; Cross, P. C. Molecular Vibrations,
Dover Publications: New York, 1955.
(29) Herzberg, G. Spectra of Diatomic Molecules, D. Van Nostrand,
Princeton, NJ, 1950. The present upper limit estimate for De(OsN) is above
a lower limit value (616 kJ/mol) calculated by Lievin using the MRCI
method described in ref 14.
(30) Chertihin, G. V.; Bare, W. D.; Andrews, L. J. Phys. Chem. A 1998,
102, 3697.
(31) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88,
899.
(32) Neuhaus, A.; Veldkamp, A.; Frenking, G. Inorg. Chem. 1994, 33,
5278.
(33) Willson, S.; Andrews, L. J. Phys. Chem. A 1999, 103, 1311.
(34) Darling, J. H.; Ogden, J. S. J. Chem. Soc., Dalton. Trans. 1972,
No 3, 2496.
(35) Calderazzo, F.; L’ Eplattenier, F. Inorg. Chem. 1967, 6, 1220.
(36) Rushman, P.; van Buuren, G. N.; Shiralian, M.; Pomeroy, R. K.
Organometallics 1983, 2, 693.
References and Notes
(1) Cotton, S. A. Chemistry of Precious Metals; Blackie Academic and
Professional: 1997.
(2) Pell, S.; Mann, R. H.; Taube, H.; Armor, J. N. Inorg. Chem. 1974,
13, 479.
(3) Love, J. L.; Robinson, W. T. Inorg. Chem. 1972, 11, 1662.
(4) Dietrich, H.; Geng, P.; Jacobi, K.; Ertl, G. J. Chem. Phys. 1996,
104, 375.
(5) Matsushima, T. Surf. Sci. 1988, 197, L287.
(6) King, D. A.; Woodruff, D. P. The Chemical Physics of Solid
Surfaces; Elsevier: New York, 1993; Vol. 6.
(7) Chertihin, G. V.; Andrews, L.; Neurock, M. J. Phys. Chem. 1996,
100, 14609.
(8) Burkholder, T. R.; Andrews, L. J. Chem. Phys. 1991, 95, 8697.
(9) Chertihin, G. V.; Andrews, L. J. Chem. Phys. 1995, 99, 6356.
(10) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y., Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Lui, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.;
Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98, Revision A.6;
Gaussian, Inc.: Pittsburgh, PA, 1998.
(37) Collman, J. P, Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books: Mill Valley, CA, 1987.
(38) Duarte, H. A.; Salahub, D. R.; Haslett, T.; Moskovits, M. Inorg.
Chem. 1999, 38, 3895.
(11) Lee, C.; Yang, E.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(12) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639.
Krishnan, R.; Binkley, J. S.; Seegar, R.; Pople, J. A. J. Chem. Phys. 1980,
72, 650.
(13) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.