Computational and Theoretical Chemistry 1135 (2018) 28–33

Contents lists available at ScienceDirect

Computational and Theoretical Chemistry

journal homepage: www.elsevier.com/locate/comptc

All-electron triple zeta basis sets for the actinides T

a a,1 a,b,⁎ a a,2
A.Z. de Oliveira , C.T. Campos , F.E. Jorge , I.B. Ferreira , P.A. Fantin
a
Departamento de Física, Universidade Federal do Espírito Santo, 29060-900 Vitória, ES, Brazil
b
Unidade Acadêmica de Física, Universidade Federal de Campina Grande, 58429-900 Campina Grande, PB, Brazil

A R T I C LE I N FO A B S T R A C T

Keywords: Non-relativistic and Douglas-Kroll-Hess (DKH) segmented all-electron basis sets of valence triple zeta quality
All-electron triple zeta basis sets plus polarization functions (TZP) for the actinides are developed. To describe accurately the properties that
Segmented contraction depend on a good description of the electrons far away from the nuclei, the corresponding augmented sets (ATZP
Actinides and ATZP-DKH) are reported as well. For the actinide monoxides, bond distances, dissociation energies, natural
DKH2-B3LYP calculations
charges and populations of the valence orbitals of the actinides, and bond indices are calculated with the B3LYP/
Actinide monoxides
TZP-DKH procedure. For Am and No, the B3LYP/ATZP-DKH static mean dipole polarizabilities are also com-
puted. To assess the performance of these small size all-electron basis sets, comparison with theoretical and
experimental data reported previously in the literature is done. These sets must be also helpful on calculations of
properties involving simultaneously core and valence electrons.

1. Introduction
Among the actinide elements, only Th, U, and Pu are found in the
nature, whereas Pa is produced from radioactive decay of uranium and
Np, Am, Cm, Bk, and Cf are formed from transmutation reactions [1]. In
contrast, the heavier actinides are formed when lighter elements are
bombarded by neutrons. Some of the main applications of the early
actinides are in the production of electrical energy and as catalysts in
chemical reactions.
Due to the various difficulties (radioactivity, toxicity, and in-
stability) that arises studying actinide compounds in laboratory, theo-
retical models become a promising alternative to understand the
chemistry of the actinides. Consequently, the number of theoretical
works about actinides has increased considerably in the last ten years
(see, e.g., Refs. [2–12] and references therein). However, the quantity
of complications that appear in a theoretical treatment of an actinide
compound is not so different of that to carry out an experimental
measure. Now, differently of the lanthanides, as the 5f orbital energy of
the actinides is close to those of the 6p, 6d, and 7s orbitals, the amount
of accessible oxidation states augments [13–15]. Besides, the relativistic
effects as well as the electron correlation corrections are paramount
[16] and they must be included in the calculations. To surpass these
difficult many theoretical approaches were purposed. Among them, we
can cite the effective core potential (ECP) [17], which is known to re-
duce the computational time of the calculations because only the
valence electrons are taken into account. Certainly, this approach is
limited to evaluate properties that do not involve core electrons like
EPR, Mössbauer and X-ray absorption spectra, and electron density
analysis. An alternative to overcome these limitations is to use the
density functional theory (DFT) along with a compact all-electron basis
set coupled to the Douglas-Kroll-Hess (DKH) [18–20] Hamiltonian or to
zeroth order regular approximation (ZORA) [21–23].
Differently from what happens for the light elements, the number of
all-electron basis sets for the actinides found in the literature is sig-
nificantly smaller. It is worth mentioning the ZORA basis sets of Slater-
type functions (STFs) and the relativistic contracted basis sets available
in the ADF [24] and PRIRODA-04 codes [25], respectively. The atomic
natural orbital (ANO) basis sets [26] suitable for DKH-CASPT2 calcu-
lations. We must also cite the hierarchical sequences of correlation
consistent basis sets developed by Peterson et al. [7,27] and by Dyall
[28] to be used with the DKH and Dirac-Hartree-Fock Hamiltonians,
respectively. The efficient use of the STFs is restricted to non-hybrid
functional and to a limited number of codes, while the generally con-
tracted ANO basis sets are appropriate for accurate ab initio calculations
of small molecules. The computational efficiency of the generally con-
tracted basis sets are significantly reduced when they are used in
electronic structure codes written for segmented contracted sets. On the
other hand, a general contracted basis set program can handle a seg-
mented set without any loss of efficiency, because the latter set is a
special case of the first one. For medium and large size molecules,

⁎
Corresponding author at: Departamento de Física, Universidade Federal do Espírito Santo, 29060-900 Vitória, ES, Brazil.
E-mail address: Jorge@cce.ufes.br (F.E. Jorge).
1
Permanent address: Instituto Federal do Espírito Santo, Campus Aracruz, Avenida Morobá, 248, Bairro Morobá, 29192-733 Aracruz, ES, Brazil.
2
Permanent address: Instituto Federal do Espírito Santo, Campus Guarapari, Alameda Francisco Vieira Simões, 720, Aeroporto, 29216-795 Guarapari, ES, Brazil.

https://doi.org/10.1016/j.comptc.2018.05.010
Received 1 March 2018; Received in revised form 16 May 2018; Accepted 16 May 2018
Available online 17 May 2018
2210-271X/ © 2018 Elsevier B.V. All rights reserved.
A.Z. de Oliveira et al.
where the use of integral screening becomes very important to reduce
the computational cost of the calculations, the screening works more
efficiently for segmented than for generally contracted basis sets [29].
Segmented all-electron relativistically contracted (SARC) basis sets for
the actinides to be used with the second-order DKH (DKH2) and ZORA
Hamiltonians were constructed by Pantazis and Neese [30]. The un-
contracted (29s20p16d12f) and contracted [21s13p10d7f] SARC basis
sets contain 253 and 159 functions, respectively.
Non-relativistic and relativistic segmented all-electron basis sets of
valence triple zeta quality plus polarization functions (TZP) and the
corresponding augmented sets (ATZP) for the elements from H to La
and from Hf to Ac were generated by Jorge et al. [31–39]. These sets
are compact enough to have a good balance between computational
cost and accuracy and, consequently, they have been used with success
on property calculations of atoms, molecules, and clusters [40–43].
Thus, the main objective of this article is to extend the TZP, ATZP, TZP-
DKH, and ATZP-DKH basis sets for the elements from Th to Lr. For the
actinide monoxides (AnO), bond lengths, dissociation energies, natural
charges and populations of the valence orbitals of the actinides, and
bond indices are calculated at the DKH2-DFT/TZP-DKH level of theory.
In addition, ATZP-DKH mean dipole polarizabilities of Am and No are
reported. Comparison with theoretical and experimental data published
previously in the literature is carried out.
2. TZ, TZP, ATZP, TZP-DKH, and ATZP-DKH basis set generations
2.1. TZ basis set for Th-Lr
The used procedure in this work to construct the TZ basis set for the
actinides is similar to that for the elements from Hf to Ac [37,38],
namely: Firstly, a subroutine containing the improved generator co-
ordinate Hartree-Fock (HF) method [44] is employed to generate the
uncontracted basis set of each element. Next, another subroutine, that
optimizes all exponents of the uncontracted basis set, is used to improve
the total HF energy of Th-Lr. To carry out these optimization processes,
the ground states of Th (6d27s2, 3F), Pa (5f26d17s2, 4K), U (5f36d17s2,
5
L), Np (5f46d17s2, 6L), Pu (5f67s2, 7F), Am (5f77s2, 8S), Cm (5f76d17s2,
9
D), Bk (5f97s2, 6H), Cf (5f107s2, 5I), Es (5f117s2, 4I), Fm (5f127s2, 3H),
Md (5f137s2, 2F), No (5f147s2, 1S), and Lr (5f147s27p1, 2P) and the
ATOM-SCF code [45] are used. After some tests, we verify that it is
necessary to add one more diffuse Gaussian function to describe cor-
rectly the 6d orbitals of the elements from Pa to No. In the optimization
processes of these functions, the elements from Pa to No, whose 6d
orbitals are not occupied in the ground state, an excited state arising
from electronic configuration 5fn−16d17s2 (5f146d17s2 for Lr) is taken
into account. It should be mentioned here that for Th, the smallest
Gaussian function exponents of f symmetry are optimized considering
the 5f16d17s2 electronic configuration. From Pa through No, an excited
state arising from electronic configuration 5fn−17s27p1 (6d17s27p1 for
Th) is used to optimize the most diffuse function of the p symmetry. The
HF method is employed to generate all these functions. At the end of all
these processes the numbers of primitive Gaussian functions of the
uncontracted basis set for the elements Th and Lr (22s15p13d7f) and for
Pa-No (22s15p14d7f) are 181 and 186, respectively. Finally, a sub-
routine developed by our research group and coupled to the ATOM-SCF
program is employed to find the best segmented contraction scheme for
each atom. For that, the criterion of the smallest difference between the
energies obtained with the uncontracted and contracted basis sets of
each element is used. The contraction schemes as well as the TZ set for
Th and Lr [10s6p6d4f] and for Pa-No [10s6p7d4f] can be found in the
webpage http://qcgv.ufes.br/.
2.2. TZP basis set for Th-Lr
To generate the TZP basis set for the actinides, 2g polarization
functions are added to the TZ set and, then, they are optimized using
29
Computational and Theoretical Chemistry 1135 (2018) 28–33
the restricted open-shell second-order Mϕller-Plesset (ROMP2) pertur-
bation theory and a subroutine coupled to the Gaussian 09 code [46].
This program is also used in the molecular calculations. In the ROMP2
calculations, the frozen core approximation and spherical Gaussian
functions are employed. The number of correlated electrons in the
ROMP2 calculations of each element is given in parenthesis: Th (4), Pa
(5), U (6), Np (7), Pu (8), Am (9), Cm (10), Bk (11), Cf (12), Es (13), Fm
(14), Md (15), No (16), and Lr (17). The TZP basis sets for Th and Lr
[10s6p6d4f2g] and for Pa-No [10s6p7d4f2g] have 104 and 109 func-
tions, respectively.
2.3. ATZP basis set for Th-Lr
To have a good description of the electrons far from the nuclei, it is
necessary to add to the TZP set diffuse functions. These functions are
indispensables on calculations of electron affinity, polarizabilities, op-
tical rotation, and electronic circular dichroism. With this in mind,
diffuse functions of s, p, d, and f symmetries are added to the TZP basis
sets and, then, they are optimized considering the anion ground state
HF energies. Next, the ROMP2 minimum energy criterion is applied to
the ground states of the anions to choose the g diffuse functions to be
added to the latter set. Consequently, the augmented TZP (ATZP) basis
set for the actinides constructed in this work has the following final
forms: [11s7p7d5f3g] for Th and Lr and [11s7p8d5f3g] for Pa-No.
2.4. TZP-DKH and ATZP-DKH basis sets for Th-Lr
For molecules containing heavy elements, the relativistic effects are
significant, so, they must be taken into account on calculations of any
chemical or physical property. A method widely used to estimate the
scalar relativistic effects of atomic and molecular systems is the DKH
[18–20]. De Jong et al. [47] verified that if the contraction coefficients
of the cc-pVXZ (X = D, T, Q, and 5) basis sets of the elements H, He,
B–Ne, Al–Ar, and Ga–Br were re-optimized with the DKH Hamiltonian,
undesirable results of some compounds could be avoided. So, with the
objective to include the scalar relativistic effects, the TZP basis set
contraction coefficients for H-La and Hf-Ac [31–33,37,38] were re-
contracted at the DKH level. This set was nominated TZP-DKH [37–39].
For the actinides, the DKH2 Hamiltonian [48] and the Gaussian 09
code [46] are employed to reoptimize the contraction coefficients of
TZP. It should be mentioned here that the exponents and contraction
schemes of the non-relativistic basis set are not altered in this proce-
dure. Even so, the radial changes of the wave function caused by core
orbital contractions and valence orbital expansions are included. For
the elements from Th to Lr, Brummelhuis et al. [49] showed that the
DKH2 Hamiltonian is variational.
A similar process was employed by Pantazis and Neese [30] to de-
velop relativistic basis sets for the actinides. Initially, ROHF calcula-
tions along with the series αl χ−i (i = 1, 2, 3, … and χ = 2.20, 2.40,
2.50, and 2.60 for l = s, p, d, and f) were used to construct the un-
contracted basis set. Then, the DKH2 and ZORA Hamiltonians were
employed to determine the contraction coefficients of the SARC basis
sets.
The ATZP-DKH basis set for the actinides is obtained adding the
diffuse functions reported in Section 2.3 to the TZP-DKH set. It should
be emphasized that the only difference between the ATZP and ATZP-
DKH basis set is the contraction coefficients.
3. Results and discussion
As previously stated, to handle theoretically compounds containing
actinides relativistic effects become essentials. Therefore, only the re-
lativistic basis sets generated in this work (TZP-DKH and ATZP-DKH) in
conjunction with the DKH2 Hamiltonian will be used henceforth.
The success of the B3LYP functional [50,51] on atomic and mole-
cular property calculations is well known to all. This functional has
A.Z. de Oliveira et al. Computational and Theoretical Chemistry 1135 (2018) 28–33

Table 1 Table 2
Natural bond orbital analysis of the actinide monoxides from DKH2-B3LYP/ Equilibrium bond lengths (Å) of the actinide monoxides.
TZP-DKH calculations.
Multiplicity DKH2- DKH2- ECP-B3LYPc DKH-CASSCF-
qAna WBIb nAnc B3LYPa B3LYPb CASPT2d

ThO 1.18 1.47 7 s(1.90)5f(0.06)6d(0.87) ThO 1 1.845 1.843 1.833 1.863

PaO 1.11 1.54 7 s(1.92)5f(1.48)6d(0.55) PaO 2 1.793 1.811 1.812 1.811
UO 1.19 1.38 7 s(0.99)5f(3.09)6d(0.80) UO 5 1.831 1.851 1.843 1.837
NpO 1.19 1.36 7 s(0.99)5f(4.19)6d(0.69) NpO 6 1.830 1.842 1.836 1.837
PuO 1.33 1.15 7 s(0.98)5f(5.20)6d(0.55) PuO 7 1.832 1.827 1.830 1.818
AmO 1.28 1.18 7 s(0.92)5f(6.39)6d(0.44) AmO 8 1.820 1.839 1.836 1.800
CmO 1.15 1.38 7 s(0.95)5f(7.08)6d(0.87) CmO 9 1.840 1.849 1.842 1.835
BkO 1.17 1.27 7 s(1.27)5f(7.92)6d(0.67) BkO 8 1.834 1.841 1.835 −
CfO 1.11 1.32 7 s(0.96)5f(9.24)6d(0.72) CfO 7 1.821 1.832 1.822 −
EsO 1.11 1.21 7 s(0.94)5f(10.32)6d(0.65) EsO 6 1.830 1.823 1.822 −
FmO 1.32 1.12 7 s(0.05)5f(12.06)6d(0.61) FmO 3 1.898 1.901 1.850 −
MdO 1.27 1.24 7 s(0.64)5f(12.73)6d(0.38) MdO 2 1.914 1.911 1.898 −
NoO 1.31 1.20 7 s(0.47)5f(13.86)6d(0.39) NoO 1 1.921 1.937 1.923 −
LrO 1.34 1.21 7 s(1.00)5f(13.97)6d(0.71) LrO 2 1.855 1.873 1.871 −

a a
Charge (e) of the actinide atom. This work. TZP-DKH all-electron basis set generated in this work (for Th-Lr)
b and in Ref. [39] (for O).
Wiberg bond index (e).
b
c
Populations (e) of the actinide valence orbitals. This work. SARC (for Th-Lr) and TZP-DKH (for O) all-electron basis sets
taken from Refs. [30] and [39], respectively.
c
been successfully used by us and by other authors to study electronic Results from Ref. [4]. For the actinides, the Cologne–Stuttgart small-core
structures of systems containing sixth row elements as well as actinides relativistic effective core potential along with the valence basis set
(see, e.g., Refs. [3,4,30,37,38] and references therein). Apparently, it [10s9p5d4f3g] were used, while for oxygen the aug-cc-pVTZ basis set was
seems the best choice for the present work. Initially, the DKH2-B3LYP/ applied.
d
Results from Ref. [56]. For the actinides and oxygen the ANO-VTZP all-
TZP-DKH procedure is used to determine the actinide monoxide equi-
electron basis sets were used.
librium structures and, then, the dissociation energies. Next, atomic
charges and Wiberg bond indices (WBIs) [52] are evaluated from the
3.2. Bond length
natural bond orbital analysis [53]. We recall that the Wiberg bond
index is an estimate of the number of covalent bonds of a chemical
As the AnO molecules include all actinide atoms, it is important to
bonding. For the elements Am and No, static mean dipole polariz-
test the performance of the TZP-DKH basis set proposed in this work on
abilities are computed as well.
bond distance calculations. For the early actinide monoxides there are
innumerous bond length results obtained at the different levels of
theory reported in the literature [4,54–56], whereas for the late acti-
3.1. Natural bond orbital analysis
nide monoxides this number is reduced [4].
To assess the reliability of our results showed in Table 2, initially,
The natural atomic charges (qAn) and valence orbital populations
they are compared with the only experimental bond lengths available in
(nAn) of the actinides as well as the WBIs obtained at the DKH2-B3LYP/
the literature, namely: 1.84018613 (microwave spectroscopy) [57] for
TZP-DKH level of theory for the actinide monoxides are displayed in
ThO and 1.8383 Å (high-resolution photoelectron spectroscopy) [58]
Table 1.
for UO. Both our results underestimate the experimental values by
From the magnitudes of the atomic charges and WBIs, one can note
0.005 and 0.007 Å, respectively. Considering that for ThO, the CASPT2
that the chemical bonds of the actinide monoxides have similar ionic
bond distance [56] is overestimated by 0.02 Å, the errors obtained by us
and covalent characters and that the atomic charges and WBIs are be-
can be consider normal for actinide monoxides. Kovács et al. [3] ver-
tween 1.1 and 1.4 e and 1.1 and 1.6 e, respectively. Another remarkable
ified that the spin-orbit effects on the bond lengths of the early actinide
characteristic of these compounds is that those containing the late ac-
monoxides are always ≤0.007 Å.
tinides are in general more ionic than those containing the early ones.
From Table 2, one can note that the trend of the DKH2-B3LYP/TZP-
On the other hand, when the bond indices are considered, the opposite
DKH bond lengths is different from that observed for the actinide ionic
occurs, i.e., the oxides of early actinides have larger covalent char-
radii [57], which is known as actinide contraction and that is analogous
acters.
to the lanthanide contraction occurring for the 4f atoms. This effect
An analysis of the actinide valence orbital populations gives us in-
happens only in the bonds that have a high ionic character. We see in
formation about the contribution of these orbitals to the chemical bond.
the last section that for the actinide monoxides, the covalent and ionic
Some general trends can be noted when we go from ThO to LrO: The 6d
contributions to the chemical bond are equivalents. However, due to
population decreases while the 7f increases. Similar tendencies were
the high ionic character of the bonding (see Ref. [10] and references
previously observed for other actinide molecules [15]. There is an
therein) occurring in the actinide trichlorides, the actinide contraction
abrupt fall from PaO to UO due to the 7s orbital to lose an electron,
is clearly manifested, i.e., one can observe a reduction of the bond
which is spread among the valence orbitals of uranium and oxygen. In
distances along the actinide series from Th to Lr.
general, the occupation numbers of the 7s orbitals are close to 1 or 2.
Some peculiar features of the actinide monoxide bond lengths can
These results could be interpreted as lone electrons or lone pairs, re-
be understood using the data of the natural bond orbital analysis
spectively. In contrast, the 7s populations in FmO, MdO, and NoO are
compiled in Table 1. Initially, at the DKH2-B3LYP/TZP-DKH level, we
reduced to 0.05, 0.64, and 0.47 e, respectively. As the numbers of
observe the short bond distance of PaO when compared with the other
electrons in the actinide 6d orbitals of these compounds increase to
compounds. From Table 1, one verifies that the bond in PaO has the
0.61, 0.38, and 0.39 e (cf. Table 1), this is an indication that these
largest covalent character with the largest 5f contribution to the bond.
orbitals participate in chemical bonding.
As the radius of the 5f orbital is smaller than the 6d one, it can be one of
the reasons for the short bond distance in PaO. We can also note that
the population of the 7s orbital is doubly occupied in ThO and PaO,

30
A.Z. de Oliveira et al. Computational and Theoretical Chemistry 1135 (2018) 28–33

while for most actinide monoxides it is singly occupied. The fully filled with each other. Comparing with the experimental (731 kJ/mol for
7s subshell has a smaller radius than a half filled one. It can certainly be NpO) [12] and estimated (460 kJ/mol for EsO) [60] values, one verifies
another cause of the shortening of the bonds in ThO and PaO. Another that the SARC results present the worst agreements.
point to be discussed here is the systematic increase of the DKH2- In addition, we note that the DKH2-B3LYP/TZP-DKH dissociation
B3LYP/TZP-DKH bond lengths going from CfO to NoO, which is at- energies decrease from ThO to AmO, an abrupt increase occurs in CmO
tributed to the filling of the π* and, then, the σ* antibonding molecular and, then, it is followed by another gradual diminution from BkO to
orbitals. Finally, the shorter bond distance in LrO possibly is due to NoO while, for LrO, sudden rise happens again. For the early actinide
nonbonding 5f orbital energy of this compound to be about 0.1 Hartree monoxides in gas phase, similar trend is observed considering the ex-
more stable than the corresponding ones obtained for the other mon- perimental dissociation energies [12,59], while for the monoxides of
oxides. the late actinides, the tendency of the results predicted by Haire’s
Bond lengths obtained with the DKH2-B3LYP method and with the promotion model [60] is the same of that obtained by us. These finds
SARC (for Th-Lr) [30] and TZP-DKH (for O) [39] all-electron basis sets, reinforce the idea that the dissociation energy results reported in this
with the Cologne-Stuttgart small-core relativistic effective core poten- work are reliable.
tial along with the valence basis set [10s9p5d4f3g] (117 functions) for In summary, with exceptions of ThO, CmO, and LrO, the general
the actinides and the aug-cc-pVTZ basis set for oxygen, and with the characteristic of the DKH2-B3LYP/TZP-DKH dissociation energies of
DKH-CASSCF-CASPT2 method along with the ANO-VTZP all-electron the actinide monoxides is to decrease along the row. To understand
basis sets for the actinides and oxygen are included in Table 2. The these results, it is necessary to take into account the stabilities of both
agreements among the bond distances calculated with the four theo- molecules and products. Before doing that, it is important to recall that
retical models are in general very good. For example, comparing the there is a general trend between bond strength and bond length as well
results computed with the TZP-DKH and SARC all electron basis sets at as between bond strength and vertical dissociation energy. A larger
the same level of theory, the largest difference occurs for UO and is bond distance in general is related to a weaker bond and vice versa. The
equal to 0.02 Å. It is an indicative of the accuracy that can be achieved large bond strength of LrO is due to the strong sigma bond formed by
with our smaller basis set in bond length calculations. the single 7p electron of lawrencium and the 2p electrons of oxygen. It
explains why this monoxide has larger dissociation energy and there-
3.3. Dissociation energy fore shorter bond length than those of NoO and MdO. Table 2 shows
that the bond lengths of UO and BkO are greater than that of AmO by
Table 3 shows the relativistic dissociation energies for the actinide 0.01 Å. Even so, the latter monoxide has dissociation energy smaller
monoxides obtained with the B3LYP hybrid functional in conjunction than those of the uranium and berkelium monoxides. The reason for
with all-electron and valence basis sets. For SARC [30] and TZP-DKH, this is that the ground state electronic configuration of Am (5f77s2) is
the counterpoise corrections [58] are included in the dissociation en- very stable.
ergy calculations in order to eliminate the basis set superposition errors
(BSSEs). The results reported in Ref. [4] also consider these corrections. 3.4. Polarizability
From Table 3, one can see that in general one result agrees well with
all other corresponding theoretical values, indicating that the TZP-DKH It is known that for the actinides the spin-orbit effects on polariz-
basis set works well in predicting not only bond lengths but also dis- ability calculations are not neglected [61]. Besides, the only DKH po-
sociation energies of actinide monoxides. Large differences (∼40 kJ/ larizability results found by us in the literature are for Am and No. With
mol) between dissociation energies calculated with the TZP-DKH and the purpose to test the performance of the relativistic augmented basis
SARC basis sets are found only for NpO and EsO. In contrast, for these set generated in this work, we decided to calculate the polarizabilities
two monoxides, the TZP-DKH and ECP-B3LYP results agree very well of these two elements and, then, to compare them with the available
results.
Table 3 The DKH2-B3LYP/ATZP-DKH mean dipole polarizability of 126.1
Dissociation energies (kJ/mol) of the actinide monoxides. a.u. for Am compares satisfactorily with the DKH2-CASPT2/
26s23p17d13f5g3h result of 116 a.u [26]. On the other hand, for No,
Multiplicity DKH2-B3LYPa DKH2-B3LYPb ECP-B3LYPc
the DKH2-B3LYP/ATZP-DKH (114.0 a.u.) and DKH2-CCSD(T)/
ThO 1 909 919 914 37s34p25d16f3g2h (115.6 a.u.) [61] results agree very well. At the
PaO 2 862 849 861 DKH2-B3LYP level of theory, the polarizability of No evaluated with a
UO 5 750 744 811 very large uncontracted all-electron basis set (37s34p25d16f3g2h) is
NpO 6 749 705 768
PuO 7 599 605 598
only 2.7 a.u. smaller than ours. It looks like that the ATZP-DKH basis set
AmO 8 501 499 492 is able to give reliable polarizability results.
CmO 9 687 686 684
BkO 8 581 567 537 4. Conclusions
CfO 7 522 515 479
EsO 6 399 360 403
FmO 3 352 334 369 For Th-Lr, compact non-relativistic (TZP and ATZP) and relativistic
MdO 2 324 312 341 (TZP-DKH and ATZP-DKH) segmented all-electron basis sets are gen-
NoO 1 267 275 290 erated in this work. It is worth mentioning here that in DFT calculations
LrO 2 582 581 583 of medium and large size molecules, they are more efficient than gen-
a erally contracted sets and that, differently from the relativistic ECP
This work. TZP-DKH all-electron basis set generated in this work (for Th-Lr)
valence basis sets, they are also able to compute properties that depend
and in Ref. [39] (for O). BSSE counterpoise corrections are included in the
of a good description of the core and valence electrons of the atoms that
calculations.
b
This work. SARC (for Th-Lr) and TZP-DKH (for O) all-electron basis sets constitute the molecular systems. For a better performance, we re-
taken from Refs. [30] and [39], respectively. BSSE counterpoise corrections are commend to use them with the similar quality sets reported previously
included in the calculations. for the elements H-La and Hf-Ac [31–39].
c
Results from Ref. [4]. For the actinides, the Cologne–Stuttgart small-core Below we list some specific conclusions obtained from our results:
relativistic effective core potential along with the valence basis set
[10s9p5d4f3g] were used, while for oxygen the aug-cc-pVTZ basis set was – The natural bond orbital analysis carried out in this work using the
applied. DKH2-B3LYP/TZP-DKH procedure shows that the bonds of the

31
A.Z. de Oliveira et al.
actinide monoxides have similar covalent and ionic characters. For
the actinide atoms of these compounds, one verifies that (i) the 5f
populations increase from ThO to LrO, (ii) the 7s populations are in
general equal to 2 or 1, and (iii) the 6d populations in general de-
crease from ThO to AmO and from BkO to NoO, whereas the max-
imum values occur for ThO and CmO.
– Now, considering the DKH2-B3LYP/TZP-DKH bond distances of the
actinide monoxides, one can note that PaO presents the shortest one
and that there is an increase from CfO to NoO followed by an abrupt
fall in LrO. The oscillations that happen in the bond lengths along
the actinide row can be understood through the results of the nat-
ural bond orbital analysis.
– For ThO, CmO, and LrO, the DKH2-B3LYP/TZP-DKH dissociation
energies present maximum values, whereas for the other monoxides
a downward trend along the actinide row is observed.
– Comparing with the results obtained at higher levels of theory, the
DKH2-B3LYP/TZP-DKH procedure shows to be reliable in polariz-
ability calculations of actinide monoxides.
Finally, we can note that the basis sets developed by us besides
producing accurate results for the actinide monoxides, they have
smaller sizes than the SARC sets [30] and are compact enough to be
competitive with the relativistic ECP valence basis set used in Ref. [4].
The non-relativistic and relativistic basis sets of TZ quality con-
structed by our research group for the elements H-La and Hf-Lr can be
found in different formats at the website http://qcgv.ufes.br/.
Acknowledgement
We would like to acknowledge the financial support of CNPq,
CAPES, and FAPES (Brazilian Agencies).
References
[1] T. Gray, The Elements: A Visual Exploration of Every Known Atom in the Universe,
Black Dog & Leventhal Publishers, New York, 2009.
[2] G. Schreckenbach, G.A. Shamov, Theoretical actinide molecular science, Acc.
Chem. Res. 43 (2010) 19–29.
[3] A. Kovács, R.J.M. Konings, J.K. Gibson, I. Infante, L. Gagliardi, Quantum chemical
calculations and experimental investigations of molecular actinide oxides, Chem.
Rev. 115 (2015) 1725–1759.
[4] A. Kovács, P. Pogány, R.J.M. Konings, Theoretical study of bond distances and
dissociation energies of actinide oxides AnO and AnO2, Inorg. Chem. 51 (2012)
4841–4849.
[5] M.V. Ryzhkov, B. Delley, Electronic structure of predicted endohedral fullerenes
An@C40 (An = Th–Md), Compt. Theor. Chem. 1013 (2013) 70–77.
[6] N.S. Mosyagin, A.V. Zaitsevskii, L.V. Skripnikov, A.V. Titov, Generalized relativistic
effective core potentials for actinides, Int. J. Quantum Chem. 116 (2016) 301–315.
[7] K.A. Peterson, Correlation consistent basis sets for actinides. I. The Th and U atoms,
J. Chem. Phys. 142 (2015) 074105.
[8] F. Kias, F. Talbi, A. Elkechai, A. Boucekkine, C-F bond breaking by bare actinide
monocations in the gas phase: a relativistic DFT study, Compt. Theor. Chem. 1118
(2017) 133–143.
[9] Y.-X. Zhonga, Y.-R. Guo, Q.-J. Pan, A theoretical probe of high-valence uranium and
transuranium silylamides: structural and redox properties, Chem. Phys. Lett. 646
(2016) 75–80.
[10] A. Kovács, R.J.M. Konings, D. Szieberth, B. Krámos, Study of the An-Cl bond con-
traction in actinide trichlorides, Struct. Chem. 25 (2014) 991–996.
[11] D. Wang, W.F. van Gunsteren, Z. Chai, Recent advances in computational actinoid
chemistry, Chem. Soc. Rev. 41 (2012) 5836–5865.
[12] J.K. Gibson, Role of atomic electronics in f-element bond formation: bond energies
of lanthanide and actinide oxide molecules, J. Phys. Chem. A 107 (2003)
7891–7899.
[13] V. Vetere, P. Maldivi, B.O. Roos, C.J. Adamo, Covalent vs electrostatic interactions
in rare earth systems: a comparative study of U(III), U(IV), and U(V) and Nd(III)
bonding properties by DFT and CAS-PT2 approache, J. Phys. Chem. A 113 (2009)
14760–14765.
[14] T.W. Hayton, Metal–ligand multiple bonding in uranium: structure and reactivity,
Dalton Trans. 39 (2010) 1145–1158.
[15] I. Kirker, N. Kaltsoyannis, Does covalency really increase across the 5f series? A
comparison of molecular orbital, natural population, spin and electron density
analyses of AnCp3 (An = Th–Cm; Cp = η5-C5H5), Dalton Trans. 40 (2011)
124–131.
[16] W.A. de Jong, R.J. Harrison, J.A. Nichols, D.A. Dixon, Fully relativistic correlated
benchmark results for uranyl and a critical look at relativistic effective core
32
Computational and Theoretical Chemistry 1135 (2018) 28–33
potentials for uranium, Theor. Chem. Acc. 107 (2001) 22–26.
[17] M. Dolg, Effective core potentials, in: J. Grotendorst (Ed.), Modern Methods and
Algorithms of Quantum Chemistry, vol. 3, second ed., John von Neumann Institute
for Computing, Jülich, 2000.
[18] M. Douglas, N.M. Kroll, Quantum electrodynamical corrections to the fine structure
of helium, Ann. Phys. 82 (1974) 89–155.
[19] B.A. Hess, Applicability of the no-pair equation with free-particle projection op-
erators to atomic and molecular structure calculations, Phys. Rev. A 32 (1985)
756–763.
[20] B.A. Hess, Relativistic electronic-structure calculations employing a two-component
no-pair formalism with external-field projection operators, Phys. Rev. A 33 (1986)
3742–3748.
[21] E. van Lenthe, E.J. Baerends, J.G. Snijders, Relativistic total energy using regular
approximations, J. Chem. Phys. 101 (1994) 9783–9792.
[22] E. van Lenthe, J.G. Snijders, E.J. Baerends, The zero-order regular approximation
for relativistic effects: the effect of spin–orbit coupling in closed shell molecules, J.
Chem. Phys. 105 (1996) 6505–6516.
[23] C. van Wüllen, Molecular density functional calculations in the regular relativistic
approximation: method, application to coinage metal diatomics, hydrides, fluorides
and chlorides, and comparison with first-order relativistic calculations, J. Chem.
Phys. 109 (1998) 392–399.
[24] G. te Velde, F.M. Bickelhaupt, E.J. Baerends, C.F. Guerra, S.J.A. Van Gisbergen,
J.G. Snijders, T. Ziegler, Chemistry with ADF, J. Comput. Chem. 22 (2001)
931–967.
[25] D.N. Laikov, Y.A. Ustynyuk, PRIRODA-04: a quantum-chemical program suite. New
possibilities in the study of molecular systems with the application of parallel
computing, Russ. Chem. Bull. 54 (2005) 820–826.
[26] B.O. Roos, R. Lindh, P.-Å. Malmqvist, V. Veryazov, P.-O. Widmark, New relativistic
ANO basis sets for actinide atoms, Chem. Phys. Lett. 409 (2005) 295–299.
[27] R. Feng, K.A. Peterson, Correlation consistent basis sets for actinides. II. The atoms
Ac and Np–Lr, J. Chem. Phys. 147 (2017) 084108.
[28] K.G. Dyall, Relativistic double-zeta, triple-zeta, and quadruple-zeta basis sets for the
actinides Ac–Lr, Theor. Chem. Acc. 117 (2007) 491–500.
[29] F. Jensen, Atomic orbital basis sets, WIREs Comput. Mol. Sci. 3 (2013) 273–295.
[30] D.A. Pantazis, F. Neese, All-electron scalar relativistic basis sets for the actinides, J.
Chem. Theory Comput. 7 (2011) 677–684.
[31] P.L. Barbieri, P.A. Fantin, Gaussian basis sets of triple and quadruple zeta valence
quality for correlated wave functions, F.E. Jorge, Mol. Phys. 104 (2006)
2945–2954.
[32] S.F. Machado, G.G. Camiletti, A. Canal Neto, F.E. Jorge, Raquel S. Jorge, Gaussian
basis set of triple zeta valence quality for the atoms from K to Kr: application in DFT
and CCSD(T) calculations of molecular properties, Mol. Phys. 107 (2009)
1713–1727.
[33] C.T. Campos, F.E. Jorge, Triple zeta quality basis sets for atoms Rb through Xe:
application in CCSD(T) atomic and molecular property calculations, Mol. Phys. 111
(2013) 167–173.
[34] P.A. Fantin, P.L. Barbieri, A. Canal Neto, F.E. Jorge, Augmented Gaussian basis sets
of triple and quadruple zeta valence quality for the atoms H and from Li to Ar:
applications in HF, MP2, and DFT calculations of molecular dipole moment and
dipole (hyper)polarizability, J. Mol. Struct. (Theochem) 810 (2007) 103–111.
[35] G.G. Camiletti, A. Canal Neto, F.E. Jorge, S.F. Machado, Augmented Gaussian basis
sets of double and triple zeta valence qualities for the atoms K and Sc-Kr: appli-
cations in HF, MP2, and DFT calculations of molecular electric properties, J. Mol.
Struct. (Theochem.) 910 (2009) 122–125.
[36] L.S.C. Martins, F.A.L. de Souza, G.A. Ceolin, F.E. Jorge, R.C. de Berrêdo,
C.T. Campos, Augmented Gaussian basis sets for the elements K, Sc-Kr, Rb, and Y-
Xe: application in HF, MP2, and DFT calculations of molecular electric properties,
Compt. Theor. Chem. 1013 (2013) 62–69.
[37] L.S.C. Martins, F.E. Jorge, S.F. Machado, All-electron segmented contraction basis
sets of triple zeta valence quality for the fifth-row elements, Mol. Phys. 113 (2015)
3578–3586.
[38] C.T. Campos, A.Z. de Oliveira, I.B. Ferreira, F.E. Jorge, L.S.C. Martins, Segmented
all-electron Gaussian basis sets of double and triple zeta qualities for Fr, Ra, and Ac,
Chem. Phys. Lett. 675 (2017) 1–5.
[39] F.E. Jorge, A. Canal Neto, G.G. Camiletti, S.F. Machado, Contracted Gaussian basis
sets for Douglas-Kroll-Hess calculations: estimating scalar relativistic effects of some
atomic and molecular properties, J. Chem. Phys. 130 (2009) 064108.
[40] A. Antusek, M. Sulka, Ab initio calculations of NMR shielding of Sc3+, Y3+ and
La3+ ions in the water solution and 45Sc, 89Y, 138La and 139La nuclear magnetic
dipole moments, Chem. Phys. Lett. 660 (2016) 127–131.
[41] A.S. Novikov, D.M. Ivanov, M.S. Avdontceva, V.Y. Kukushkin, Diiodomethane as a
halogen bond donor toward metal-bound halides, Cryst. Eng. Comm. 19 (2017)
2517–2525.
[42] S.E. Galembeck, G.F. Caramori, A. Misturini, L.C. Garcia, R.P. Orenha, Metal–ligand
bonding situation in ruthenophanes containing mMultibridged cyclophanes,
Organometallics 36 (2017) 3465–3470.
[43] F.E. Jorge, L.G.M. de Macedo, Dipole (hyper)polarizabilities of neutral silver clus-
ters, Chin. Phys. B 25 (2016) 123102.
[44] E.V.R. de Castro, F.E. Jorge, J.C. Pinheiro, Accurate Gaussian basis sets for second-
row atoms and ions generated with the improved generator coordinate Hartree-
Fock method, Chem. Phys. 243 (1999) 1–7.
[45] S.J. Chakravorty, G. Corongiu, J.R. Flores, V. Sonnad, E. Clementi, V. Carravetta,
I. Cacelli, Modern Techniques in Computational Chemistry MOTECC-89, ESCOM,
Leiden, 1989.
[46] 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.
A.Z. de Oliveira et al.
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, 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, D.J. Fox, Gaussian 09, Revision A.02, Gaussian Inc.,
Wallingford, CT, 2009.
[47] W.A. de Jong, R.J. Harrison, D.A. Dixon, Parallel Douglas-Kroll energy and gra-
dients in NWChem: estimating scalar relativistic effects using Douglas-Kroll con-
tracted basis sets, J. Chem. Phys. 114 (2001) 48–53.
[48] G. Jansen, B.A. Hess, Revision of the Douglas-Kroll transformation, Phys. Rev. A 39
(1989) 6016–6017.
[49] R. Brummelhuis, H. Siedentop, E. Stockmeyer, The ground-state energy of re-
lativistic one-electron atoms according to Jansen and Hess, Doc. Math. 7 (2002)
167–182.
[50] A.D. Becke, Density-functional thermochemistry. III. The role of exact Exchange, J.
Chem. Phys. 98 (1993) 5648–5652.
[51] C. Lee, W. Yang, R.G. Parr, Development of the Colle-Salvetti correlation-energy
formula into a functional of the electron density, Phys. Rev. B 37 (1988) 785–789.
[52] K.B. Wiberg, Application of the pople-santry-segal CNDO method to the cyclopro-
pylcarbinyl and cyclobutyl cation and to bicyclobutane, Tetrahedron 24 (1968)
1083–1096.
[53] A.E. Reed, L.A. Curtiss, F. Weinhold, Intermolecular interactions from a natural
33
Computational and Theoretical Chemistry 1135 (2018) 28–33
bond orbital, donor-acceptor viewpoint, Chem. Rev. 88 (1988) 899–926.
[54] B.B. Averkiev, M. Mantina, R. Valero, I. Infante, A. Kovács, D.G. Truhlar,
L.J. Gagliardi, How accurate are electronic structure methods for actinoid chem-
istry? Theor. Chem. Acc. 129 (2011) 657.
[55] I. Infante, A. Kovács, G. La Macchia, A.R.M. Shahi, J.K. Gibson, L.J. Gagliardi,
Ionization energies for the actinide mono- and dioxides series, from Th to Cm:
theory versus experiment, J. Phys. Chem. A 114 (2010) 6007–6015.
[56] A. Kovács, R.J.M. Konings, Computed vibrational frequencies of actinide oxides
AnO0/+/2+ and AnO20/+/2+ (An = Th, Pa, U, Np, Pu, Am, Cm), J. Phys. Chem. A
115 (2011) 6646–6656.
[57] N.M. Edelstein, J. Fuger, J.J. Katz, L.R. Morss, Summary and comparison of prop-
erties of the actinide and transactinide elements, in: N.M. Edelstein, J. Fuger, L.R.
Morss (Eds.), The Chemistry of the Actinide and Transactinide Elements, vol. 3,
Springer, Dordrecht, The Netherlands, 2006, pp. 1753−1835.
[58] S.F. Boys, F. Bernardi, The calculation of small molecular interactions by the dif-
ferences of separate total energies. Some procedures with reduced errors, Mol. Phys.
19 (1970) 553–566.
[59] R.J.M. Konings, L.R. Morss, J. Fuger, Thermodynamic properties of actinides and
actinide compounds, in: N.M. Edelstein, J. Fuger, L.R. Morss (Eds.), The Chemistry
of the Actinide and Transactinide Elements, vol. 4, Springer, Dordrecht, The
Netherlands, 2006, pp. 2113−2224.
[60] R.G. Haire, High-temperature vaporization of transplutonium oxides, J. Alloys
Compd. 213 (214) (1994) 185–190.
[61] C. Thierfelder, P. Schwerdtfeger, Effect of relativity and electron correlation in
static dipole polarizabilities of ytterbium and nobelium, Phys. Rev. A 79 (2009)
032512.