Vasiliy Goncharov, Jiande Han, Leonid Kaledin, Probing actinide electronic structure using fluorescence and multi-photon ionization spectroscopy Vasiliy Goncharov, Jiande Han, Leonid Kaledin, and Michael Heaven Department of Chemistry DoE

Motivation In principle, computational quantum chemistry methods may be used to predict the properties of hazardous radioactive materials. Computational prediction is desirable as this could greatly reduce the need for difficult and expensive laboratory studies. Computational methods for treating heavy element compounds are being developed but they need to be tested against reliable data for gas phase molecules. There is a critical need for such experimental data, and this information can be obtained from studies of less hazardous Th and U compounds.

Key issue for actinide chemistry Role of the f-electrons in bonding and electronic structure. Challenges for experiment Open f- and d- shells result in high densities of electronic states. Refractory species  high temperatures. Challenges for theory Strong relativistic effects. Large numbers of electrons.

Examples of the difficulties in comparing experiment and theory for actinide compounds Calculations for UO2 have yielded predictions for the ground state of (5ffu)2 3S-g, (5ffu) (5fdu) 3Hg or 3S-g and (5ffu)(7ssg) 3Fu Data for UO2 obtained in rare gas matrices show anomalously large guest-host interactions DG1/2 = 914 cm-1 (Ne), 776 cm-1 (Ar) (Zhou, Andrews, Ismail & Marsden JPCA, 104, 5495 (2000)) Most high level theoretical calculations yield an ionization energy for UO2 above 6 eV. Electron impact measurements yield 5.4 eV

Ionization energies are often used in the determination of bond energies Uncertainties in the IE’s propagate through the thermodynamic data base

Simple electronic structure model for ionic actinide compounds M.O. theory does not provide an easily interpretable zeroth-order picture Ligand field theory works well for lanthanides - is it suitable for actinides? Basic concept - Mn+ perturbed by Ln- Successful if the f-orbitals are very compact R. W. Field, Ber. Bunsenges. Phys. Chem. 86, 771 (1982)

Example of LFT applied to CeO 1 (Field, Linton et al.) 2 2 3 Jf=7/2 3 3 4 4 Ce2+(4f6s) Ce2+(4f6s)O2- 1 0- 2 1 3 Jf=5/2 3 2 2 Ja W=Ja W=Ja-1 W=Ja-2

Spatial extent of lanthanide (Nd) and actinide (U) atomic orbitals N. Edelstein, J. Alloy. Comp. 223, 197 (1995)

Electronic spectroscopy of UO (Kaledin, McCord & Heaven JMS 164, 27 (1994)) Ground state U2+(5f37s,5I4)O2- X(1)W=4 First excited state (294 cm-1) U2+(5f27s2,3H4)O2- X(1)W=4 LFT prediction with no adjustable parameters, X(1)W=1 Low-lying states can be interpreted using an adjustable parameter LFT model. Is this meaningful?

Spectroscopic studies of actinide oxides using multi-photon ionization techniques Resonance enhanced multi-photon ionization (REMPI) Photo-ionization efficiency curves (PIE) Mass-analyzed threshold ionization (MATI) Zero kinetic energy (ZEKE) photoelectron spectroscopy

Multi-Photon Ionization Processes M+ + e- hv1 hv2 REMPI ZEKE PIE MATI Pulsed electric field

ZEKE & Mass-selected REMPI spectrometer Microchannel plate (cation detection) MO+ Metal target Skimmer hv2 Einzel lens Pulsed valve Grids MO+He Microchannel plate (electron detection) Vaporization laser hn1

Spectroscopy of ThO+ - a simple test case Theoretical expectations - single unpaired electron outside of a closed shell metal ion core Ground state: Th3+(7s)O2-, X2S+ First excited state manifold: Th3+(6d)O2-, 2D, 2P, 2S+

REMPI Spectrum for ThO reveals new vibronic transitions Low-resolution scan Wavelength, nm

Rotationally resolved 1-0 band of the F’(0+)-X(0+) transition

Photoionization of ThO Intermediate resonance at 28578.8 cm-1 IP= 53260(5) cm-1 6.6035(6) eV e=21.4 V/cm Threshold at 24653.5 cm-1 ThO+ ion signal Literature value 6.1(1) eV Second Photon Energy /cm-1

PFI-ZEKE Spectra of ThO, X 2+ (v+= 0) state X 2+, v+ = 0, N+ scanned W = 0+ ThO O, v’ = 0, J’ fixed W = 0+ ThO X, v” = 0, J”  X 2+ state, v+ = 0 Bo+ = 0.3450(6) cm-1 IE(ThO)=53253.8(2) cm-1

PFI-ZEKE Spectra for ThO+ Rotational structure of ThO+ A, =5/2, v=0 ThO+ A, v+=0 Broken lines show zero-field energies 2 A’, v=0 1 ThO X, v=0

Spectroscopic data for ThO+ State To /cm-1 (Theory1) T /cm-1 (This Work) B /cm-1 e /cm-1 exe /cm-1 X 2+  = 0  = 1  = 2  = 6  = 7 0, {52020} 0, {IE = 53253.8(2)} 950.0(1) 1895.3(1) 5627.0(1) 6547.2(5) 0.3450(6) 0.3439(5) 0.3434(5) 0.3409(10) –— 954.97(6) 2.45(3) 1 23/2  = 3  = 4 2477 2933.7(1) 3846.2(1) 5656.8(1) 6554(1) 0.3373(8) 0.337(1) 0.3379(13) 917.1 2.35 1 25/2 5886 5814.4(1) 6729.9(1) 0.3410(2) 0.340(1) [915.5(2)] 1 21/2  = 5 9167 7404.1(1) 8303.6(1) 9198.5(1) 10088.7(2) 11855.0(2) 0.3365(11) 0.3354(10) 0.3334(6) 0.3330(7) 0.333(2) 904.22(2) 2.339(3) 1. Rajni Tyagi, PhD Thesis, OSU 2005. Advisor, R. M. Pitzer

Ionization makes the Th-O bond weaker but stiffer IE(Th)=6.3067 eV Th+ + O IE(ThO)=6.6027 eV Hence, the ThO+ bond is weaker D0+ IE(Th) D0-D0+=0.296 eV Th + O but its vibrational frequency is higher e /cm-1 ThO 895.77 ThO+ 954.97 D0 IE(ThO) and B(ThO+)>B(ThO)

Avoided curve crossings are responsible for the anomalous relationship between the bond energy and molecular constants Th+ + O must correlate with this limit on dissociation Energy /cm-1 X2S+ R/Å Th3+(7s)O2- approximate configuration at equilibrium

Photoionization spectroscopy of UO U2+(5f37s, 5I4)O2-  UO*  U3+(5f3, 4I4.5)O2- IE (electron impact) = 5.6(1) eV Goncharov & Heaven, RH01

He I Photoelectron Spectrum of UO/UO2 Vapor

Low-lying states of UO+: What to expect? 11392 cm-1 (Ref. 1) 4I6.5 8024 cm-1 (Ref. 1) U3+ [5f3]: the lowest energy term–4I: 4I5.5 4265 cm-1 (Ref. 1) 4I4.5 0 cm-1 (Ref. 1) (Ref. 2) 4I7.5 9796 4I6.5 7251 4600 – 4800 cm-1 U3+ [5f3] + O2-[2p6]: 4I: 3991 4I5.5 600 – 700 cm-1 4I4.5 0 cm-1 W: 7.5 6.5 5.5 4.5 3.5 2.5 1.5 0.5 1 Jean BLAISE and Jean-François WYART, Selected Constants Energy Levels and Atomic Spectra of Actinides. 2 L. A. Kaledin et. al. Journal of Molecular Spectroscopy 164, 27-65 (1994)

PFI-ZEKE Spectra of UO, X(1)4.5(v+= 0) state X(1)4.5, v+=0, J+ scanned UO [19.453]3(v’ = 0) fixed UO X(1)4(v” = 0)  X(1)4.5(v+=0) Bo+ = 0.3467(7) cm-1 IE(UO)=6.031065(25) eV e- Impact1: 5.6(1) eV Theory / Best Value (CASSCF)2: 6.040 eV 1  E. G. Rauh and R. J. Ackermann, J. Chem. Phys. 60, 1396 (1974). 2  J. Paulovic, L. Gagliardi, J. Dyke, K. Hirao, J. Chem. Phys. 122, 144317 (2005).

PFI-ZEKE Spectra of UO, [1.132] (1)2.5 state (v+=0) X(1)4.5, v+=1 (1)3.5, v+=0 UO+ X(1)4.5, v+=0 scanned UO [19.453]3(v’ = 0) fixed UO X(1)4(v” = 0)  [1.132] (1)2.5 state Bo+ = 0.3364(3) cm-1 To=1132.42 cm-1

UO+ Energy Levels Diagram: 0 – 5200 cm-1 range (only v+ = 0 & 1 levels are shown, all states originate from U3+[5f3]O2- configuration )

Comparison of the experimentally obtained data to theoretical calculations for UO+ State This Work PFI-ZEKE (2006) LFT Calculations L. Kaledin (1994/2006) MCSCF/CI Rajni&Pitzer (2005) MCSCF/VCI Krauss& Stevens (1983) U3+ energy levels X(1)4.5 0/0 0 {66%4I9/2+8%4H9/2} X 4I4.5[5f3] (1)3.5 764.93(20) 633/757 582 {47%4H7/2+15%4I7/2…} 1319  (1)2.5 1132.42(20) 696/1129 856 {20%4 5/2+17%45/2…) 1895 (1)1.5 1284.4(5) 580/1281 1076 {17%4 3/2+8%43/2…) 2094 (1)0.5 1325.5(8) 695/1328 3296 (1)5.5 4177.83(20) 3991/4163 3744 {69%4 I11/2+11%4H11/2…) 2563 4265 4I5.5 [5f3] (2)4.5 4758.45(20) 4601/4773 4180 {36%4H9/2+15%4I9/2…} 3599 (2)3.5 5161.96(20) 4770/5121 4045 (2)2.5 5219.37(20) 4744/5293 (3)3.5 4982.44(20) 4287 {58%4H7/2[5f27s]}

Trends in bond length and vibrational frequency change resulted Comparison of the experimentally obtained data to theoretical calculations for UO+ Spectroscopic data, UO+ This Work Krauss&Stevens AREP-MCSCF-SO Paulovič et. al. /CASSCF ANO-RCC basis set Rajni&Pitzer MCSCF-SO X, e /cm-1 911.9(2) 92530 912  X, re /Å 1.798(5) 1.842 1.802 1.812 Trends in bond length and vibrational frequency change resulted from photo-ionization of ThO and UO ThO (Ref. 1) ThO+ (Ref. 2) Difference UO (Ref. 3) UO+ This work X, re /Ǻ 1.840 1.804(3) 0.036(3) 1.8383 1.798(5) 0.040(5) X, e /cm-1 895.77 954.97(6) 59.2(1) 846.5(6) 911.9(2) 65.4(8) 1 Edvinsson, G.; Selin, L.-E.; Aslund, N., Arkiv. Fysik. 30, 283-319 (1965). 2 V. Goncharov and M. C. Heaven, J. Chem. Phys. (2006). 3 L. A. Kaledin et. al. Journal of Molecular Spectroscopy 164, 27-65 (1994)

m(UO)=3.363 D m(NdO)=3.31 D Tongmei Ma et al. WF09 Indication that LFT may be viable for actinides provided by recent measurements of the dipole moment and magnetic g-factor for UO m(UO)=3.363 D m(NdO)=3.31 D Tongmei Ma et al. WF09

Electronic spectroscopy of UO2 Points of interest Ground state configuration: (5ff) 2, (5ff)(5fd), or (5ff)(7ss) ? Ionization energy: theory >0.5 eV above experiment Gagliardi. et al (JPCA 105, 10602, 2001) conclude that the experimental values are wrong Origin of the anomalous vibrational frequency matrix shifts. 914 cm-1 (Ne) vs. 776 cm-1 (Ar)

Is the ground state configuration U(5f2)O2 or U(5f7s)? 2F7/2 3u 5ff5fd 3H 4g 5g 6g 4u 5ff7ss or 3u 3F 2F5/2 2u Lowest energy for W=Ja f fs

Calculations for UO2 by Chang & Pitzer (2002) Spin-orbit SCF-CI using relativistic core potentials 3H4 5f2 3D 3F 5f7s

First observation of the electronic spectrum of gas phase UO2 UO2+ Ion signal UO2++e- v2 (fixed) 2g v1 3u 2u X 3F

Visible range spectra for UO2 show progressions of bending vibrational levels Note that odd-Dv transitions were not observed. This is consistent with a linear structure in both the ground and excited states.

Rotational resolution could not be achieved using a laser linewidth of 0.06 cm-1 The rotational structure was not resolved. Surprising as the rotational constant should be around 0.16 cm-1. Checks for power broadening and fragmentation yielded negative results - The congestion is a property of UO2 Ground state configuration determined from vibronic structure

Delayed ionization of UO2 at energies just above threshold Bond dissociation energy of UO2 (7.85 eV)exceeds the IP Mixing of UO2++e- with highly excited levels of UO2 lengthens the lifetime. t=220 ns Decay of UO2+ interferes with MATI detection

Photo-ionization of UO2 X 3F(2u) v1=31838 cm-1 v2 UO2++e- 3F(2g) Photo-ionization of UO2 I.P.=49424(20) cm-1 (6.128(3) eV) UO2+ Ion signal

Ionization potentials (eV) for UO2: Comparison with previous determinations and theory Reference Method UO2 This work MATI, PIE 6.128(3) Rauh & Ackerman ‘74 Electron Impact 5.4(1) Capone et al. ‘99 Gagliardi et al. ‘01 CASPT2 6.17 Zhou et al. ‘00 B3LYP 6.3 Majumdar et al. ‘02 MP2 6.05 Rajni & Pitzer ‘05 SOCI 5.7 Calculations for the energies of low-lying excited states are still not converged - see Fleig et al. JCP 124, 104106 (2006)

Conclusions Spectroscopic studies of the low-lying states of actinide compounds yield interpretable data. ZEKE is particularly well suited for mapping the low energy electronic structure. Relativistic quantum chemistry is making good progress, but there is a long way to go. Calculations for the simplest molecules are still very challenging. Preliminary indications are that LFT yields meaningful insights for ionic compounds. Ionization energies for refractory actinide compounds require systematic re-evaluation.

Thanks to - DoE Michael Duncan (UGA) Fredric Merkt (ETH Zurich) Robert Field (MIT) Russ Pitzer (OSU) Laura Gagliardi (U. of Geneva) Björn Roos (Lund) DoE

Anomalous Behavior of Matrix Isolated UO2 Jin Jin and Chris Lue

Calculated vibrational frequencies for gas phase UO2 Zhou, Andrews, Ismail & Marsden, JPC A, 104, 5495 (2000) (spin-free, relativistic DFT) 3Fu v(su)=931 cm-1 3Hg v(su)=814 cm-1 Large vibrational matrix effect is attributed to a re-ordering of the electronic states caused by a strong interaction between UO2 and Ar

Bruce Bursten

Matrix Isolation Apparatus used to Study UO2 in Solid Ar Cold Mirror (12K) ABLATION LASER Configuration for sample preparation

Dispersed fluorescence spectrum obtained using 266 nm excitation (Nd/YAG 4th harmonic)

Band positions for matrix isolated UO2 and comparison of observed low-lying energy levels with theoretical predictions

The emission spectrum is consistent with transitions that terminate on the low-lying 5f7s states, but is this still the lowest energy configuration? u Non-radiative relaxation g 5f7s u g 5f2

Laser excitation spectrum Detection band Fluorescence intensity Dispersed fluorescence Gas phase transition - 366.9 nm

Conclusions for UO2 in Ar Emission and excitation spectra indicate that 5f7s is the lowest energy configuration for UO2 in solid Ar The anomalous effect of Ar on the vibrational frequency may be due to host-induced state mixing Studies of the UO2-Ar van der Waals complex will be used to probe the issue of incipient chemical bond formation.

Electronic Spectrum of UO Low resolution Note the dramatic effect of isotopic substitution on this spectrum. Bright states are mixed with many dark states. Isotopic substitution reorders the perturbations

IR Studies of UO2 Isolated in Rare Gas Matrices Green, Gabelnick et al. ANL, 1973 and 1980 UO2 trapped in solid Ar: v(sg)=776.10 v(su)=765.45 v(pu)=225.2 Andrews et al. UVA, 1993 and 2000 UO2 trapped in solid Ar: v(su)=776.0 UO2 trapped in solid Ne: v(su)=914.8 D=138.8 cm-1

Photoionization of atomic uranium I.P.=49959(1) cm-1 f 3s2 4I9/2 (literature value: 49958.4(5) cm-1) U++e- e=343 V/cm e=21 V/cm v2 f 3s2p 5I5 v1=17908.17 cm-1 f 3ds2 5L6

Angular momentum selection rules are relaxed by field-induced mixing of the Rydberg series Spectrum for excitation via J’=7

PFI-ZEKE Spectra for ThO+ Rotational structure of ThO+ X2+, v+=1 ThO+ X, v+=1 Broken lines show zero-field energies 2 O, v=0 1 ThO X, v=0

Observed Energy Level Structure for UO2 Low-lying excited state at 360 cm-1 cannot be explained by the 5f2 3H ground state assumption. The results are in good agreement with the structure expected for 5f7s (first predicted by Zhou et al. JCPA, 104, 5495 2000)