Reactions of laser-ablated U atoms with (CN)2: Infrared spectra and electronic structure calculations of UNC, U(NC)2, and U(NC)4 in solid argon

Article abstract of DOI:10.1039/c4cc09946j

Reactions of laser-ablated U atoms with (CN)₂ produce UNC, U(NC)₂, and U(NC)₄ as the major products, identified from their Ar matrix infrared spectra and precursors partially and fully substituted with ¹³C and ¹⁵N. Mixed isotopic multiplets substantiate product stoichiometries. Band positions and quantum chemical calculations verify the isocyanide bonding. This journal is

Full text of DOI:10.1039/c4cc09946j

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Reactions of laser-ablated U atoms with (CN)2:
infrared spectra and electronic structure
calculations of UNC, U(NC)2, and U(NC)4 in solid
argon†
Cite this: Chem. Commun., 2015,
51, 3899
Received 12th December 2014,
Accepted 29th January 2015
Yu Gong,a Lester Andrews,*a Benjamin K. Liebov,a Zongtang Fang,b
DOI: 10.1039/c4cc09946j
Edward B. Garner, IIIb and David A. Dixon*b
www.rsc.org/chemcomm
Reactions of laser–ablated U atoms with (CN)2 produce UNC, U(NC)2, coworkers with a variety of substituents.5–7 The local environment of
and U(NC)4 as the major products, identified from their Ar matrix uranium for the isolated complex is different in the gas phase as
infrared spectra and precursors partially and fully substituted with investigated by computations and the noble gas matrix as studied by
13C and 15N. Mixed isotopic multiplets substantiate product stoichio- infrared spectroscopy of the ligand stretching mode as compared to
metries. Band positions and quantum chemical calculations verify cyanide compounds such as [Cp*2U(CN)3][NEt4]3.6 Thus it is reason-
the isocyanide bonding.
able for the NC/CN attachment to the U centers to be different in
these cases. So far, the only known uranium isocyanide examples
Cyanogen, (CN)2, is a pseudohalogen with two important where the bare CN ligand is nitrogen bound to uranium are
differences: (1) the C–C bond energy, 136.7 Æ 1.6 kcal molÀ1
,
CH3UNC and CH2QU(H)NC prepared by reactions of laser ablated
is at least double any of the halogen molecules,1 and (2) the CN uranium atoms and acetonitrile in argon matrices.8 We report here
radical can bond at either C or N, leading to cyanides and the formation of UNC, U(NC)2, and U(NC)4 complexes from the U
isocyanides. Different metal centers may prefer one type of and cyanogen reaction isolated in argon matrices at 4 K and
2
¨
bonding over the other. Pyykko and coworkers proposed that supporting theoretical calculations.
uranium bonding to CN is not favorable because both are s
Argon matrix infrared spectra from the reaction of laser ablated
donors and p acceptors. They hypothesized that the isocyanide uranium atoms and 1.0% cyanogen are illustrated in Fig. 1, which
arrangement is preferred because the N is a better p donor than shows the major absorptions after sample deposition at 4 K,
the C. They reported electronic structure calculations that ultraviolet (UV) irradiation, and annealing to 30 K in trace (a).
showed that it is slightly more exothermic to replace 2 F in Details of the experiments are given in the Supporting material.
UF6 with 2 NC than with 2 CN groups and that the stretching The sharp band at 2054.2 cmÀ1 observed in all of our spectra is due
frequencies are about 200 cmÀ1 higher for the cyano ligand to the precursor isomer CNCN produced by photolysis from
than for the isocyano ligand. Other theoretical investigations of the emission plume.9 Four major uranium dependent product
uranyl-CN/NC bonding have been reported.3,4 Sonnenberg, bands were observed on deposition at 2040.6, 2028.5, 2020.0
et al. predicted that the tetra coordinated ion [UO2(NC)4]2À is and 1989.4 cmÀ1 (Table 1). It is important to consider the effect
more stable than the CN isomer, and the reverse holds for the of (CN)2 concentration, using 0.3%, for example. The first band at
penta-coordinated analogs.4 These results suggest that the U–CN 2040.6 cmÀ1 is clearly stronger relative to the other bands; the
and U–NC bonding have comparable energies, which depend 2028.5 and 2020.0 cmÀ1 bands track in relative intensity just as they
intimately on the local environment.
did using 1% reagent, and the lower band at 1989.4 cmÀ1 is
There are a number of ligand supported cyanido complexes markedly weaker than the other product absorptions. Annealing
of uranium most notably from the work of Ephritikhine and had the same band sharpening effect as described for Fig. 1, and
photolysis again clearly increased the lower bands. In addition,
annealing to 20 K also revealed growth of the CN radical at
2044.2 cmÀ1 most likely from combination of atoms produced by
laser plume photolysis. These spectra show that there are products
of three different stoichiometries, which give rise to the first, the
second and third, and the lowest band system, respectively.
The isotopically substituted cyanogen spectra in Fig. 1 provide the
identification for these three products. The first band at 2040.6 cmÀ1
shifted to 1999.4 cmÀ1 using (13CN)2 with the 1.02061 frequency ratio
a University of Virginia, Department of Chemistry, Charlottesville, VA, 22904-4319,
USA. E-mail: lsa@virginia.edu
b The University of Alabama, Department of Chemistry, Tuscaloosa, AL, 35487-0336,
USA. E-mail: dadixon@ua.edu
† Electronic supplementary information (ESI) available: Experimental details.
NBO Population Analysis with the B3LYP Functional for the lowest energy states
of U(NC)n for n = 1 to 4. Total energies and geometries (x, y, z Cartesian
coordinates in Å) for the lowest energy states of U(NC)n and U(CN)n for n = 1 to
4. See DOI: 10.1039/c4cc09946j
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 3899--3902 | 3899

View Article Online
Communication
ChemComm
Fig. 1 Infrared spectra of laser ablated U and (12CN)2 and 13C substituted (CN)2 reaction products codeposited in argon at 4 K and after UV photolysis and
annealing to 30 K to optimize product yield. (a) 1% (12CN)2, (b) 1% (12,13CN)2, (c) 1% (12CN)2 and 1% (13CN)2 mixed, (d) 1% (13CN)2. The * denotes the CNCN
photoisomer of cyanogen.
Table 1 Experimental and DFT (B3LYP) calculated isocyanide N–C stretches (in cmÀ1) for NC bonded to U from reactions with cyanogen. Calculated
intensities in parentheses in km molÀ1
Molecule
UNC expt
(
12C14N)2
(
13C14N)2
(
12C14N)2
+
14N13C12C14N + (13C14N)2
(
12C14N)2 + (13C14N)2
2040.6
1999.4
2040.6, 1999.4
2040.6
1999.4
4UNC calc.
2080.6s (493)
2038.7s
2080.6, 2038.7
2080.6
2038.7
U(NC)2 expt
2028.5
2020.0
1988.1
1979.4
2028.5, 2024.7
1988.1, 2020.0
1983.3, 1979.4
2028.5
1988.1
2020.0
1979.4
5U(NC)2 calc.
2072.3b2 (681)
2084.0a1 (171)
2031.2b2
2044.3a1
2072.3, 2036.8
2031.0, 2084.0
2080.9, 2044.3
2072.3
2031.0
2084.0
2044.3
U(NC)3 expt
2015
1975
4U(NC)3 calc.
2072.0e (1453)
2088.7a1 (54)
2030.0e
2046.8a1
2029.9, 2039.3
2033.9, 2072.8
2079.7, 2084.4
U(NC)4 expt
1989.4
1949.8
1989.1, 1959.7
1968, 1954
1949.6
1989.4
1959.7
1949.8
3U(NC)4 calc.
2045.6b (927)
2046.4e (2100)
2086.0a (0)
2004.2b
2004.9e
2045.3a
2004.4, 2004.9
2025.3, 2065.9
2010.2, 2045.8
2046.4, 2080.6
2004.5
2017.0
2046.0
2074.3
and exhibited no intermediate mixed isotopic components.
The next pair of bands at 2028.5 and 2020.0 cmÀ1 with
Analogous behaviour was found in the (C14,15N)2 reaction (not (12CN)2 maintain constant relative intensities and arise from
shown) with the pure 15N component at 2008.8 cmÀ1 and ratio the same species, which absorb at 1988.1 and 1979.4 cmÀ1 with
1.01583. These observations show that this vibration involves (13CN)2 and define the 1.02032 and 1.02051 frequency ratios. It
one C and one N atom. This 2040.6 cmÀ1 frequency for UNC may is significant that no new intermediate bands are observed in
be compared with the 2029.7 cmÀ1 value reported for CH3UNC.8 the experiment with mixed reagents, trace (c), which for this
3900 | Chem. Commun., 2015, 51, 3899--3902
This journal is ©The Royal Society of Chemistry 2015

View Article Online
ChemComm
Communication
product is the sum of the pure isotopic spectra. However, using
the scrambled reagent, trace (b) with the (12CN)(13CN) reagent
also present, these bands plus two new bands are observed at
2024.7 and 1983.3 cmÀ1, almost at the median of the first band
pairs. The latter two bands are due to the mixed isotopic
product species (12CN)U(N13C) and show that U undergoes a
concerted reaction with cyanogen. Analogous mixed isotopic
bands were observed for (C14N)U(15NC). Therefore the pair of
bands at 2028.5 and 2020.0 cmÀ1 with (12CN)2 are due to the
symmetric and antisymmetric UN–C stretching modes of
(12CN)–U–(N12C) with two equivalent isocyanide ligands.
The situation is different for the last band at 1989.4 cmÀ1
which contains one new band at 1959.7 cmÀ1 in trace (c) and
two additional new bands in trace (b) at 1968 and 1954 cmÀ1
.
This shows that two precursor molecules are involved in this
product molecule. The weak band at 2015 cmÀ1 has the proper
13C isotopic counterpart to be due to an N–C mode, but we
cannot obtain any mixed isotopic data for this band. On the
basis of the calculations described below, we tentatively assign
Fig. 2 Spin densities of the lowest energy uranium isocyanides for 1 to
4 NC groups with a contour level of 0.02.
the weak 2015 cmÀ1 band to U(NC)3. Finally, we can assign the 5U(NC)2 is predicted to be 52 cmÀ1 above the experimentally
1989.4 cmÀ1 band to U(NC)4 where we are able to observe three observed value with the higher intensity. The weaker a1 band
of the mixed isotopic components. The band at 1991.9 cmÀ1 for the symmetric stretch is predicted to be at a higher
could be assigned as the matrix site splitting of the 1989.4 cmÀ1 frequency of 2084 cmÀ1, 56 cmÀ1 above the weaker experi-
band or to the b mode of 3U(NC)4 in S4 symmetry as discussed mental band at 2028 cmÀ1. The higher energy 5U(CN)2 structure
below based on the computational results.
has two weak computed bands at 2213 and 2215 cmÀ1 which
In order to better understand the energetics of the various are not consistent with the observed spectral features. The CN
structures identified in the matrix and to assist in the IR spectral stretches for the U–CN bonding arrangement increase with
assignments, density functional theory (DFT) with the B3LYP increasing numbers of CN in contrast to the decrease predicted
hybrid exchange–correlation functional10 was used with the aug- for the isocyanide bonding arrangement and the decrease
cc-pVTZ basis set11 on C and N and the ECP60MWB effective observed experimentally. The ground state is 7s15f3 with about
core potential and ECP60MWB_SEG [10s,9p,5d,4f,3g] basis set 0.5 e donated to the 6d from the NBO populations. The 0.54 e is
on U.12 The calculations were done in the spin unrestricted spin polarized with 0.41 e a spin and 0.14 e b spin.
4
formalism. As a test of the functional, calculations on the states
The ground state U(NC)3 in C3V symmetry is more stable than
of UCN and UNC were also performed with the PBE and PW91 the lowest energy 4U(CN)3 structure by 18 kcal molÀ1. The 4U(NC)3 is
functionals which gave the same energy ordering and similar predicted to have an intense e antisymmetric stretch band, 57 cmÀ1
frequencies. The B3LYP calculated frequencies are shown in higher than the one tentatively assigned at 2015 cmÀ1. The higher
Table 1. The Gaussian 09 program system was used for the energy a1 symmetric stretch is predicted to have a much lower
DFT calculations.13 The relative energetics are given in the intensity (factor of B25). The ground state is 5f3 with about 0.5 e
Supporting material as are the complete list of frequencies.
donated to the 6d from the NBO populations. The 0.54 e in the d is
The lowest energy structure for addition of a single ligand to essentially spin paired so it does not contribute to the spin state.
U is 4UNC with 6UCN 29 kcal molÀ1 higher in energy. The Weak bands at 2219 cmÀ1 (195 km molÀ1) for the e mode and
intense UN–C stretching mode of 4UNC is predicted to be 40 cmÀ1 2221 cmÀ1 (17 km molÀ1) for the a mode are predicted for 4U(CN)3.
above the experimentally observed value consistent with the fact
that the calculated values are harmonic and the experimental ones 3U(CN)4 structure in C2v (Btetrahedral) symmetry by 30 kcal molÀ1
The 3U(NC)4 in S4 symmetry is more stable than the lowest energy
.
contain an anharmonic component. The 6UCN is predicted to have A very intense degenerate e stretch band at 2046 cmÀ1 with an
a much weaker UC–N stretching mode at a higher frequency of additional intense band of b symmetry within 1 cmÀ1 is predicted
2169 cmÀ1 (12 km molÀ1), and there is no observed absorption in for 3U(NC)4 consistent with the assignment of the experimental band
this region. The electron configuration of the 4UNC ground state is at B1990 cmÀ1. The difference between the calculated and experi-
7s25f3 (see Table S1, ESI† for the NBO populations.14) There is mental values is 56 cmÀ1, consistent the differences given above.
about 0.3 e spin paired on the 6d. The spin densities are shown in Three stretching bands for 3U(CN)4 are predicted to be at
Fig. 2. The details of the spin populations are given in the B2220 each with an intensity of B210 km molÀ1. The electron
Supporting material.
configuration is essentially f2 with 2.41 e in the 5f (the excess is
5
The lowest energy structure with two ligands is U(NC)2 in spin paired) and 1.0 e spin paired in the 6d.
C2v symmetry with 5U(CN)2, 11 kcal molÀ1 higher in energy.
The geometry results in the ESI† show that the CN bond
Note that the UNC bonding arrangement is not quite linear (see distance is shorter in the cyanide compounds and longer in the
Fig. 2). The intense UN–C antisymmetric b2 stretching mode of isocyanides consistent with the higher CN stretch in the
This journal is ©The Royal Society of Chemistry 2015
Chem. Commun., 2015, 51, 3899--3902 | 3901

View Article Online
Communication
ChemComm
former. Coupled with this is the fact that the U–C bond Notes and references
distance is 0.12 to 0.15 Å longer than the U–N bond distance
1 Y.-R. Luo, Comprehensive Handbook of Chemical Bond Energies, CRC
showing that the U is interacting more strongly with the
isocyanide than with the cyanide. In addition, the positive
charge on the U (Supporting material) in the isocyanides is
larger than on the cynanides except for n = 1 where they are
comparable. Thus there is more negative charge on the NC in the
isocyanides than on the CN in the cyanides. The isocyanides are
highly polarized with a charge of about À1.0 e on the N and a
positive charge of 0.28 to 0.37 on the C. In contrast, the charges
on the C and N are both about equal and negative for the
cyanides. Thus the electrostatic and dipolar interactions are
larger in the isocyanides than in the cyanides. As a consequence,
the isocyanide bonding interactions are larger than in the
cyanides and the isocyanides are more stable with shorter U–N
bond lengths and longer NC bond distances. The prior results2–4
are focused on the addition of CNÀ/NCÀ to U(VI) in contrast to
the current work which focuses on the addition of CNÀ/NCÀ to
U(I) to U(IV) with no other ligands present. The differing results
for whether isocyanide or cyanide bonding to U is preferred
clearly depend on the charge on the U and the nature of any
other ligands that are present.
The laser ablated uranium atom reaction with cyanogen
during condensation with excess argon at 4 K produces three
major products which are identified as the isocyanides UNC,
U(NC)2 and U(NC)4 from experimental mixed isotopic spectra,
the spectral region of the absorptions, and B3LYP product
energy and frequency calculations. The analysis of the charges
and the geometries shows that the isocyanides bond more
strongly to the U in the +I to +IV oxidation states than the
cyanides when no other ligands are present.
Press, Taylor and Francis Group, 2007.
¨
2 M. Straka, M. Patzschke and P. Pyykko, Theor. Chem. Acc., 2003, 109,
332–340.
3 C. Clavaguera-Sarrio, S. Hoyau, N. Ismail and C. J. Marsden, J. Phys.
Chem. A, 2003, 107, 4515–4525.
4 J. L. Sonnenberg, P. J. Hay, R. L. Martin and B. E. Bursten, Inorg.
Chem., 2005, 44, 2255–2262.
´
5 J.-C. Berthet, P. Thuery and M. Ephritikhine, Chem. Commun., 2007,
604–606.
´
´
6 J. Maynadie, N. Barros, J. C. Berthet, P. Thuery, L. Maron and
M. Ephritikhine, Angew. Chem., Int. Ed., 2007, 46, 2010–2012.
´
7 J.-C. Berthet, P. Thuery and M. Ephritikhine, Organometallics, 2008,
27, 1664–1666.
8 H. G. Cho and L. Andrews, Organometallics, 2012, 31, 535–544.
9 M. E Jacox and W. E. Thompson, J. Chem. Phys., 2007, 126, 054308.
10 (a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652; (b) C. Lee,
W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys.,
1988, 37, 785–789.
11 R. A. Kendall, T. H. Dunning, Jr. and R. J. Harrison, J. Chem. Phys.,
1994, 96, 6796–6806.
12 (a) http://www.theochem.unistuttgart.de/pseudopotentials/index.en.
html; (b) X. Cao, M. Dolg and H. Stoll, J. Chem. Phys., 2003, 118,
487–496; (c) X. Cao and M. Dolg, THEOCHEM, 2004, 673, 203–209.
13 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson,
H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino,
G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark,
J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith,
R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski,
G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,
O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian
09, Revision C.01, Gaussian, Inc., Wallingford CT, 2010.
14 F. Weinhold and C. R. Landis, Valency and Bonding: A Natural Orbital
Perspective, Cambridge University Press, Cambridge, UK, 2005;
A. E. Reed, L. Curtiss and F. Weinhold, Chem. Rev., 1988, 88,
899–926; E. D. Glendening, C. R. Landis and F. Weinhold,
J. Comput. Chem., 2013, 34, 1429–1437; E. D. Glendening,
J. K. Badenhoop, A. E. Reed, J. E. Carpenter, J. A. Bohmann,
C. M. Morales, C. R. Landis and F. Weinhold, NBO6 program,
Theoretical Chemistry Institute, University of Wisconsin, Madison,
WI, 2013; http://nbo6.chem.wisc.edu/.
We gratefully acknowledge financial support from DOE
Grant No. DE-SC0001034 to LA. DAD acknowledges the Department
of Energy, Office of Basic Energy Sciences, Chemical Sciences,
Geosciences, and Biosciences, Heavy Element Program for support
via a subcontract from Argonne National Laboratory. DAD also
thanks the Robert Ramsay Chair Endowment, University of
Alabama, for support.
3902 | Chem. Commun., 2015, 51, 3899--3902
This journal is ©The Royal Society of Chemistry 2015