Essential role of heterocyclic structure of N-alkylated 2-pyrrolidone derivatives for recycling uranium from spent nuclear fuels

Article abstract of DOI:10.1246/bcsj.20200061

In a simple and versatile reprocessing method for recycling U and Pu from spent nuclear fuels, cyclic amides like Nalkylated 2-pyrrolidone derivatives (NRPs) are exclusively employed. However, there has been no convincing rational to explain why such a heterocyclic structure is required. To answer this question, we employed N-cyclohexyl-2-pyrrolidone (NCP) and N-cyclohexylformamide (NCF) as cyclic and acyclic monodentate amides, and focused on the following 3 topics in this study; (1) structural chemistry of their uranyl dinitrato complexes, (2) precipitation behavior of UO22+ from HNO3(aq) by using these amides, and (3) their chemical stability in HNO3(aq) simulating the reprocessing conditions for spent nuclear fuels. Fundamental coordination chemistry of UO2(NO3)2(L)2 (L = NCP, NCF) was found to be common to both L, regardless of the presence or absence of the pyrrolidone ring. Furthermore, both L exhibit comparable capability in precipitation of UO22+ from HNO3(aq). The most critical difference between NCP and NCF was found in their chemical stability in HNO3(aq), where NCF was gradually decomposed through acid-catalyzed hydrolysis, while NCP remained intact for at least 4 h. In conclusion, the pyrrolidone ring of NRPs plays an important role to sterically protect the carbonyl C from nucleophilic hydrolysis which initiates the amide C(=O)N bond cleavage.

Full text of DOI:10.1246/bcsj.20200061

Advance Publication Cover Page
Essential Role of Heterocyclic Structure of N-Alkylated 2-Pyrrolidone Derivatives
for Recycling Uranium from Spent Nuclear Fuels
Tsubasa Inoue, Hiroyuki Kazama, Satoru Tsushima, and Koichiro Takao*
Advance Publication on the web April 22, 2020
doi:10.1246/bcsj.20200061
© 2020 The Chemical Society of Japan
Advance Publication is a service for online publication of manuscripts prior to releasing fully edited, printed versions. Entire
manuscripts and a portion of the graphical abstract can be released on the web as soon as the submission is accepted. Note that the
Chemical Society of Japan bears no responsibility for issues resulting from the use of information taken from unedited, Advance
Publication manuscripts.

Essential Role of Heterocyclic Structure of N-Alkylated 2-Pyrrolidone Derivatives
for Recycling Uranium from Spent Nuclear Fuels
Tsubasa Inoue,1 Hiroyuki Kazama,1 Satoru Tsushima,2,3 and Koichiro Takao1*
1Laboratory for Advanced Nuclear Energy, Institute of Innovative Research, Tokyo Institute of Technology, 2-12-1 N1-32, O-okayama,
Meguro-ku, 152-8550 Tokyo, Japan.
2Institute of Resource Ecology, Helmholtz-Zentrum Dresden-Rossendorf, Bautzner Landstraße 400, 01328 Dresden, Germany.
3Tokyo Tech World Research Hub Initiative (WRHI), Institute of Innovative Research, Tokyo Institute of Technology, 2-12-1, O-okayama,
Meguro-ku, 152-8550 Tokyo, Japan.
E-mail: ktakao@lane.iir.titech.ac.jp
Koichiro Takao
Born in Hiroshima, Japan, in 1980. During undergraduate and graduate courses, he worked for uranium coordination
chemistry, especially for its exotic oxidation state U(V), and received Ph.D. in 2006 from TokyoTech, under supervision
by Prof. Yasuhisa Ikeda. After working at TokyoTech (posdoc, 2006-2008) and Helmholtz-Zentrum Dresden-Rossendorf
(Germany, posdoc, 2008-2010), and Seikei University (assistant professor in Tsubomura Laboratory, 2010-2012), he was
appointed as an assistant professor in Research Laboratory for Nuclear Reactors, TokyoTech, and then promoted to an
associate professor in 2015. His research interests are coordination and solution chemistry of actinides and relevant elements in nuclear fuel
cycle.
Abstract
In a simple and versatile reprocessing method for
materials, especially in reprocessing process for spent nuclear
fuels to recycle them as a next generation of nuclear fuels.2-4
There are several reprocessing methods proposed so-called
REDOX, PUREX, UREX, and also THOREX, where
formation of an extractable uranyl nitrato complex,
UO2(NO3)2.(L)2 (L = monodentante ligand) is employed to
make solvent extraction of U selective and efficient.
recycling U and Pu from spent nuclear fuels, cyclic amides like
N-alkylated 2-pyrrolidone derivatives (NRPs) are exclusively
employed. However, there have been no convincing rationales
why such a heterocyclic structure is required. To answer this
question, we employed N-cyclohexyl-2-pyrrolidone (NCP) and
N-cyclohexylformamide (NCF) as cyclic and acyclic
monodentate amides, and focused on the following 3 topics in
this study; (1) structural chemistry of their uranyl dinitrato
complexes, (2) precipitation behavior of UO22+ from HNO3(aq)
by using these amides, and (3) their chemical stability in
HNO3(aq) simulating the reprocessing process for spent nuclear
fuels. Fundamental coordination chemistry of UO2(NO3)2(L)2
(L = NCP, NCF) were found to be common to both L,
regardless of the presence or absence of the pyrrolidone ring.
Previously,
we
have
discovered
that
N-cyclohexyl-2-pyrrolidone (NCP, Fig. 1) has a potential to
2+
precipitate UO2 selectively and efficiently from HNO3(aq),
where UO2(NO3)2(NCP)2 is formed as a sparingly soluble
6
microcrystalline compound.5, After this finding, we have
engaged ourselves to develop a brand-new precipitation-based
simple reprocessing method for spent nuclear fuels by using
8
N-alkylated 2-pyrrolidone derivatives (NRPs, Fig. 1).7, By
substituting the cyclohexyl group of NCP with wide variety of
linear and branched alkyl chains, the ligand structures have
Furthermore, both
L exhibit comparable capability in
2+
precipitation of UO2 from HNO3(aq). The most critical
difference between NCP and NCF was found in their chemical
stability in HNO3(aq), where NCF was gradually decomposed
through acid-catalyzed hydrolysis, while NCP remained intact
for at least 4 h. In conclusion, the pyrrolidone ring of NRPs
plays an important role to sterically protect the carbonyl C from
nucleophilic hydrolysis which initiates the amide C(=O)−N
bond cleavage.
been further designed and optimized to afford desired
10
precipitation yield and/or selectivity.9,
More recently, we
have introduced a concept of coordination polymer to further
2+
improve the efficiency and selectivity of UO2 precipitation
for development of a simple and versatile reprocessing method
for various nuclear fuel cycles.11-14 The similar
precipitation-based reprocessing concept especially towards a
thorium fuel cycle was also proposed by an Indian group.
Firstly tetraalkyl urea derivatives were employed as
precipitants,15 and later they expanded type of ligand to dialkyl
amides.16
Keywords: uranyl nitrate, reprocessing, structural
chemistry, chemical stability, protection
Introduction
Uranium exhibits rich redox chemistry by taking broad
oxidation states ranging from III to VI. Its most stable
oxidation state is hexavalent, where U6+ strongly tends to be
hydrolyzed to form UO22+; a uranyl ion.1 Understanding
2+
Fig. 1. Schematic structures of NCP, NRP, and NCF.
coordination chemistry of UO2
and neighboring early
−
actinides like Th and Pu in the presence of NO3 is the most
essential in separation and purification of these nuclear fuel
In our previous studies, cyclic amides like NRPs were
10, 17-20
extensively employed.9,
Although the effect of

N-substituent has been well investigated, importance of the
presence of its heterocyclic structure has not been fully
elucidated. The reprocessing process of spent nuclear fuels are
always taken place in the presence of a few molar HNO3.
Therefore, we wonder that organic substances like NRPs are
always suffered from a risk of oxidation and/or related
decomposition reactions by HNO3. In the former time, our
colleagues reported radiolytic behavior of NRPs in HNO3(aq),
where bond cleavages at N−CH2 and C(=O)−CH2 were
predominantly observed.7, 21 However, the role and necessity of
the pyrrolidone ring in NRPs are still not clear.
Å, c = 9.7106(5) Å, α = 113.604(8)°, β = 93.657(7)°, γ =
108.779(8)°, V = 620.23(9) Å3, Z = 1, T = 113 K, Dcalcd = 1.950
g·cm−3, μ = 66.051 cm−1, GOF = 1.100, R (I > 2σ) = 0.0291, wR
(all) = 0.0586. IR (cm−1, diamond ATR): 2930(s), 2855(s),
1604(s, C=O stretching), 1515(s), 1496(m), 1468(m), 1422(w),
1283(s), 1190(w), 1032(w) 925(s, ν3, asymmetric stretching of
[O≡U≡O]2+), 812(w), 747(m), 696(w), 648(w), 578(w),
543(w), 444(w). Anal. calcd for C20H34N4O10U: C, 32.97; H,
4.70; N, 7.69. Found: C, 33.32; H, 4.70; N, 7.74.
UO2(NO3)2(NCF)2: crystallographic data: C14H26N4O10U,
fw = 648.41, triclinic, P-1 (#2), a = 5.4458(9) Å, b =
6.5628(10) Å, c = 14.902(2) Å, α = 96.478(7)°, β = 100.070(7)°,
In this work, we employed N-cyclohexylformamide (NCF,
Fig. 1) as an acyclic monodentate ligand to compare it with
NCP. To clarify the role and importance of the heterocyclic
structure of NRPs, the cyclic and acyclic monodentate amides
γ = 95.131(7)°, V = 517.79(14) Å3, Z = 1, T = 113 K, Dcalcd
=
2.079 g·cm−3, μ = 78.985 cm−1, GOF = 1.040, R (I > 2σ) =
0.0401, wR (all) = 0.0829. IR (cm−1, diamond ATR): 3321(bs),
3081(w), 2936(s), 2854(s), 1635(s, C=O stretching), 1549(m),
1476(m), 1365(m), 1317(m), 1065(w), 1042(w), 931(s, ν3,
asymmetric stretching of [O≡U≡O]2+), 811(w), 749(bw),
542(w), 489(w), 428(w). Anal. calcd for C14H26N4O10U: C,
25.93; H, 4.04; N, 8.64. Found: C, 26.13; H, 4.07; N, 8.44.
Precipitation Behavior of UO2(NO3)2(L)2 in HNO3(aq)
(L) were compared in terms of structural chemistry of
2+
UO2(NO3)2(L)2, precipitation behavior of UO2
from
HNO3(aq), and chemical stability in HNO3(aq). Previously, our
group has twice reported molecular and crystal structures of
UO2(NO3)2(NCP)2 in triclinic P-1 space group.6, 9 In contrast,
2+
we have also clarified that the analogous PuO2 complex,
PuO2(NO3)2(NCP)2, crystallizes in monoclinic P21/c,17 despite
similar chemistry usually expected in these actinyl ions. Such a
contrast between actinides at the same oxidation state let us to
hypothesize that there could be polymorphs of
AnO2(NO3)2(NCP)2 (An = U, Pu). Therefore, we have also
investigated here whether UO2(NO3)2(NCP)2 again crystallizes
in P-1, or shows polymorphs.
(L
=
NCP, NCF).
Uranyl nitrate hexahydrate and
60%HNO3(aq) were together diluted in distilled water to
prepare a stock solution of 1.5 M UO22+ in 3.0 M HNO3(aq). To
aliquots of this solution (2 mL each) in 15 mL Pyrex glass
centrifuge tubes, NCP or NCF were added in different molar
ratios towards UO22+, followed by vigorous agitation. The
centrifuge tubes were stored at 298 K in a thermostat water
2+
bath for 1 h. After centrifugation, concentrations of UO2
remaining in the supernatants (Csup) were determined together
with that in the stock solution (Cini) by ICP-AES (PerkinElmer
Optima 3000). The precipitation yield (PYL, %) of UO2 and
effectiveness of L in UO22+ recovery (EL, %) were evaluated as
Experimental
2+
Materials. Caution! All the isotopes in natural uranium
are α emitters and also contain their radioactive daughters.
Therefore, standard precautions for handling radioactive
materials should be followed.
follows.
Uranium trioxide was dissolved in a mixture of 60%
HNO3 and distilled water to prepare a stock solution of UO2
PYL = (Cini – Csup)/Cini × 100
EL = 2 PYL/(CL/Cini) × 100
(1)
(2)
2+
(0.500 M) in 3.0 M HNO3(aq). This solution (100 μL) was
mixed with ethanol (50 μL, Wako Pure Chemical) and
N-cyclohexylformamide (NCF, 12.6 μL, 100 μmol, Aldrich).
Yellow crystals of UO2(NO3)2(NCF)2 suitable for X-ray
structure determination deposited from this reaction mixture
through slow evaporation of solvents at room temperature
within 3 days. By substituting NCF to NCP (17.9 μL, 100 μmol,
Tokyo Chemical Industry Co., Ltd.), UO2(NO3)2(NCP)2 was
also obtained as yellow crystals in a similar manner.
Characterization of UO2(NO3)2(L)2 (L = NCP, NCF).
A single crystal of UO2(NO3)2(L)2 was mounted on a Kapton
capillary. Intensity data were collected using an imaging plate
area detector on Rigaku RAXIS RAPID with graphite
monochromated Mo-Kα radiation (λ = 0.71075 Å). The
structure was solved by direct method (SIR92),18 and expanded
using Fourier techniques. All non-hydrogen atoms were
anisotropically refined by SHELXL 2017/1.19 Hydrogen atoms
were refined as riding on their parent atoms with Uiso(H) =
1.2Ueq(C). The final cycle of the full-matrix least-squares
refinement on F2 was based on the observed reflections and
parameters and converged with the unweighted and weighted
agreement factors, R and wR. All calculations were performed
using the CrystalStructure crystallographic software package.20
UO2(NO3)2(L)2 was further characterized by IR spectroscopy
(JASCO FT/IR-4700 equipped with a diamond ATR apparatus),
elemental analysis (Yanako CHN coder MT-6) and powder
XRD (Bruker D2PHASER, Cu-Kα radiation).
where CL denotes the nominal concentration of loaded L to
precipitate UO22+. The coefficient 2 in the numerator of Eq. (2)
is based on the stoichiometry of L in UO2(NO3)2(L)2.
Stability Assessment of L (= NCP, NCF) in HNO3(aq).
Chemical stability of L in HNO3(aq) at 50°C was assessed by
1H NMR spectroscopy (JEOL JNM-ECX400P, 1H: 399.78
MHz). Prior to starting experiment, all conditions of the NMR
instrument like shimming and sample temperature were
preliminarily prepared. L (0.50 M) was injected into a 3.0 M
HNO3(aq) with 10 vol% D2O in the NMR tube heated to 50°C.
After vigorous shaking within several seconds, the sample tube
was again loaded into the NMR instruments. After D-lock and
gradient shimming, the NMR kinetic experiment was started.
From injecting L to recording the first spectrum, it took less
than 5 min. The 1H NMR spectrum at 50°C was recorded every
30 min. For NCF, its chemical stability was also studied at
[HNO3]
= 1.0 M and 5.0 M. In all NMR samples,
p-toluenesulfonic acid (TsOH) was added as an internal
standard for peak integrals. To theoretically investigate
chemical stability of NCP and NCF, quantum chemical
calculations of these complexes were performed through
structure optimization and consecutive Mulliken population
analysis at B3LYP level utilizing triple-zeta basis set by using
ORCA program version 4.021
UO2(NO3)2(NCP)2: crystallographic data: C20H34N4O10U,
fw = 728.54, triclinic, P-1 (#2), a = 8.6265(5) Å, b = 8.7530(4)

Results and Discussion
trans-geometry in the equatorial plane. Both UO2(NO3)2(L)2
show quite similar structural parameters in terms of U≡Oyl
(1.77-1.78 Å), U−OL (2.34-2.35 Å), and U−ONO3 (2.53-2.55 Å)
bond distances and O≡U≡O (180°), C=O−U (142-148°), and
N−C=O (123-125°) bond angles. These structural features well
agree with observation in the IR spectra. The characteristic IR
Molecular and Crystal Structures of UO2(NO3)2(L)2.
In both NCP and NCF systems, crystals deposited from the
ethanol solutions exhibited typical yellow color of UO22+. CHN
elemental analyses clearly indicated that these yellow crystals
had the common chemical composition UO2(NO3)2(L)2 (L =
NCP, NCF). Note that our main focus here is structural
chemistry of these complexes. Therefore, the deposited crystals
were directly subjected to the following characterization
experiments without estimation of detailed synthetic yields.
Instead, yields of these compounds from HNO3(aq) will be
discussed in the next section.
2+
absorption bands arising from asymmetric stretching of UO2
(ν3) were detected at 925 cm−1 for UO2(NO3)2(NCP)2 and 931
cm−1 for UO2(NO3)2(NCF)2. Furthermore, C=O stretching
(νC=O) of L was also found at 1604 cm−1 and 1635 cm−1 for
UO2(NO3)2(NCP)2 and UO2(NO3)2(NCF)2, respectively. Both
νC=O bands are redshifted compared to those of free L,
2+
Molecular and crystal structures of UO2(NO3)2(L)2 (L =
NCP, NCF) were determined by single crystal X-ray diffraction.
Prior to discussing structural details of UO2(NO3)2(L)2, the
open question about reproducibility of the triclinic crystal
structure or possible polymorphism of UO2(NO3)2(NCP)2 (see
Introduction) has to be resolved first. As a result, it was
clarified that UO2(NO3)2(NCP)2 again crystallized in triclinic
P-1. Furthermore, all the lattice constants determined here (a =
8.6265(5) Å, b = 8.7530(4) Å, c = 9.7106(5) Å, α =
113.604(8)°, β = 93.657(7)°, γ = 108.779(8)° at 113 K) are
nearly identical to those we reported previously (a = 8.627(1) Å,
b = 8.748(1) Å, c = 9.707(1) Å, α = 113.61(1)°, β = 93.73(1)°, γ
= 108.74(1)° at 113 K)9 despite different experimental runs.
This is also the case in a comparison between these cell
parameters at different temperatures (a = 8.6794(10) Å, b =
8.8490(10) Å, c = 9.7523(10) Å, α = 113.86(1)°, β = 93.31(1)°,
γ = 109.06(1)° at 293 K).6 Consequently, crystallization of
UO2(NO3)2(NCP)2 in triclinic P-1 from the ethanolic mother
liquor was confirmed to be reproducible and not to exhibit any
thermally-induced polymorphs at 113-293 K. As mentioned
above, PuO2(NO3)2(NCP)2 recrystallized from ethanol
(monoclinic P21/c)22 shows different crystal structure from
UO2(NO3)2(NCP)2 despite the same oxidation states of actinide
elements which usually exhibit similar coordination chemistry.
It is still not very clear at this moment why the difference in the
crystal structures of AnO2(NO3)2(NCP)2 occurs between
different actinides in hexavalence.22 However, there seems to
be no critical reasons why UO2(NO3)2(NCP)2 refuses the
monoclinic system similar to that observed in
PuO2(NO3)2(NCP)2 analogue. As a matter of fact, the powder
XRD pattern (Fig. S2, Supporting Information) clearly reveals
that both triclinic and monoclinic phases are present in
indicating κ-O coordination to UO2 as demonstrated by the
X-ray structure determination. These characteristics are also
typically observed in other UO2(NO3)2(L)2 complexes with
10, 12, 23
NRP derivatives9,
ligands.24-27
and with other monodentate
In UO2(NO3)2(NCF)2, hydrogen bonding interactions
have been found between the neighboring complexes to donate
−
amide >NH (N(1)−H(1A)) in one side to terminal O of NO3
(O(5)) in the other side each other as shown in Fig. 3, making
ladder-like [UO2(NO3)2(NCF)2]n hydrogen bond polymer in the
crystal structure. Such an interaction is feasible with NCF
because of presence of the >NH group, while there is little
chance for NCP to form such an interaction due to lack of
strongly polarized H atoms.
Interestingly, a tilt angle of the coordinated NO3− from the
uranyl equatorial plane in UO2(NO3)2(NCF)2 (14.21°, Fig. 2(e))
is greater than that in UO2(NO3)2(NCP)2 (5.74°, Fig. 2(b)), and
also those of other reported UO2(NO3)2(NRP)2 complexes
(0.56-4.70°, see Table S1).10 In the crystal structure of
−
UO2(NO3)2(NCF)2 (Fig. 4), contacts between Oyl and NO3 of
the neighboring complex (Oyl···ONO3 = 2.92 Å, Oyl···N = 2.96
Å) are slightly shorter than the sum of van der Waals radii (rw;
2 rw(O) = 3.04 Å, rw(O) + rw(N) = 3.07 Å),28 indicating some
non-covalent interactions are present in these short contacts. As
−
center N of NO3 is positively charged in its resonance
structure, it may show a Lewis π-acidity to accept a Lewis
−
basic group like O. Recently, such a behavior of NO3 ,
so-called “π-hole bonding”, has been pointed out by Bauzá and
co-workers through database survey, and corroborated by
theoretical computations to reveal the donor-acceptor orbital
2+
interactions.29 The Lewis basicity of Oyl in UO2 is usually
considered to be very weak, and rarely results in an interaction
UO2(NO3)2(NCP)2 as precipitated directly from 3.0
M
with a strong Lewis acid like B(C6F5)3.30-32 Taking into account
−
HNO3(aq) (vide infra), where the monoclinic phase occurs
predominantly. Therefore, UO2(NO3)2(NCP)2 may crystallize in
different P-1 and P21/c phases. At this moment, we still do not
understand a reason why UO2(NO3)2(NCP)2 preferentially
recrystallizes in the triclinic phase from ethanolic mother
liquors.
UO2(NO3)2(NCF)2 also crystallizes in triclinic P-1 with
lattice constants a = 5.4458(9) Å, b = 6.5628(10) Å, c =
14.902(2) Å, α = 96.478(7)°, β = 100.070(7)°, γ = 95.131(7)°. It
is not very surprising that there are no similarity in the cell
parameters with UO2(NO3)2(NCP)2 (a = 8.6265(5) Å, b =
8.7530(4) Å, c = 9.7106(5) Å, α = 113.604(8)°, β = 93.657(7)°,
γ = 108.779(8)°).
that the Lewis π-acidity of NO3 is not very strong, Lewis
acid-base interactions in the short contacts currently observed
should be weak. A packing effect through crystallization of
UO2(NO3)2(NCF)2 is considered to afford the short contacts
between Oyl and NO3− in consequence.
Another striking difference of UO2(NO3)2(NCF)2 from
UO2(NO3)2(NCP)2 can be observed in flip angles of N−C−OL
mean
plane
from
the
uranyl
equatorial
plane.
UO2(NO3)2(NCF)2 shows smaller flip angle (30.6°) compared
to UO2(NO3)2(NCP)2 (56.9°) and also compared to other
reported UO2(NO3)2(NRP)2 complexes (56.5-86.4°, see Table
S1 in Supporting Information).10 As a result, the cyclohexyl
groups of NCF do not largely deviate from the uranyl
equatorial plane (Fig. 2(e)). In contrast, those of NCP are
significantly lifted by tilting the pyrrolidone rings, and
superimposed over Oyl (Fig. 2(b)). This trend is also common
in other UO2(NO3)2(NRP)2 complexes reported so far.10 In NCP
and other NRPs, there should be significant steric hindrance
Molecular structures and selected structural parameters of
UO2(NO3)2(L)2 (L = NCP, NCF) are shown in Fig. 2 and Table
1, respectively. In both compounds, center
U atom is
surrounded by two O atoms at the apical positions (Oyl) and by
−
six additional
O atoms of bidentate NO3 (ONO3) and
2+
−
monodentate L (OL) in the equatorial plane of UO2 to afford
hexagonal bipyramidal coordination polyhedron. A set of two
NO3 and two L molecules in each complex makes a
between coordinated NO3 and a methylene moiety vicinal to
the carbonyl group of the pyrrolidone ring (Fig. 2(c)). This
steric demand results in the steeper flip angle of the pyrrolidone
−

ring. In this regards, an intramolecular C−H···Oyl hydrogen
bond seems to occur between C(5)−H(5) and O(1) of
UO2(NO3)2(NCP)2 (see hydrogen bond parameters in Table 1).
This contact (H(5)···O(1): 2.67 Å) is slightly shorter than rw(H)
+ rw(O) = 2.72 Å,28 and still meets criteria of C−H···O
hydrogen bonds.33 Such an interaction was also found in other
UO2(NO3)2(L)2 complex (L = N,N-diisobutyl-N’,N’-diisopropyl
urea) by Kannan et al.16 On the other hand, such a steric effect
is not very significant in UO2(NO3)2(NCF)2 due to the absence
of pyrrolidone ring (Fig. 2(f)). Actually, there seems to be slight
−
collision between formyl H(1) and O(4) of NO3 to promote
−
tilting NO3 in the uranyl equatorial plane and to allow the
Lewis acid-base interaction between NO3− and Oyl found in Fig.
4.
−
Fig. 4. Lewis acid−base interactions between Oyl and NO3 of
UO2(NO3)2(NCF)2. Short contact distances are written in Å. Symmetry
operations: (i) 2−x, −y, 2−z; (ii) 1+x, y, z; (iii) 3−x, −y, 2−z; (iv) x−1, y,
z; (v) 1−x, −y, 2−z.
Table 1 Selected Structure Parameters of UO2(NO3)2(L)2 (L = NCP,
NCF)a
bond distances /Å
U≡Oyl
U−OL
U(1)−O(1)
U(1)−O(2)
U(1)−O(3)
U(1)−O(4)
C(1)−O(2)
C(1)−N(1)
1.774(3)
2.349(3)
2.536(3)
2.529(3)
1.263(5)
1.323(5)
U(1)−O(1)
U(1)−O(2)
U(1)−O(3)
U(1)−O(4)
C(1)−O(2)
C(1)−N(1)
1.766(4)
2.337(4)
2.530(4)
2.551(4)
1.253(8)
1.303(9)
U−ONO3
C=O
amide C−N
bond angles /°
O≡U≡O
O(1)−U(1)−O(1)i
O(1)−U(1)−O(2)
C(1)−O(2)−U(1)
N(1)−C(1)−O(2)
180.0
O(1)−U(1)−O(1)ii
O(1)−U(1)−O(2)
C(1)−O(2)−U(1)
N(1)−C(1)−O(2)
180.0
O≡U−OL
88.11(12)
147.6(3)
125.1(4)
87.7(2)
141.9(4)
123.3(6)
C=OL−U
Fig. 2. Molecular structures of UO2(NO3)2(L)2 (L = NCP (a, b, c),
NCF(d, e, f)) in different views. Thermal ellipsoids in panels (a), (b),
(d), and (e) are drawn in 50% probability level. In panels (a) and (d),
hydrogen atoms are omitted for clarity. Panels (c) and (f) show CPK
models along linear [O=U=O]2+, where C atoms in pyrrolidone ring and
formamide moieties are highlighted in light-green. Symmetry
operations: (i) 2−x, −y, −z; (ii) 2−x, −y, 2−z.
N−C=OL
dihedral angles /°
N(2)−U(1)−O(2)−C(1)
U(1)−O(2)−C(1)−N(1)
64.61
67.76
N(2)−U(1)−O(2)−C(1)
U(1)−O(2)−C(1)−N(1)
47.09
141.28
flipe angle of N−C−OL mean plane from uranyl equatorial plane /°
56.9
tilt angle of NO3− from uranyl equatorial plane /° b
5.74
30.6
14.21
hydrogen bond parameters
D−H /Å
C(5)−H(5)
H(5)···O(1)
0.98
2.67
N(1)−H(1A)
H(1A) ···O(5)iii
N(1) ··O(5)iii
0.88
2.12
H···A /Å
D···A /Å
C(5)···O(1)
3.47
2.96
D−H···A /°
short contacts /Å
C(5)−H(5)···O(1)
139.4
N(1)−H(1A) ··O(5)iii
159.2
c
−
O(1)···O(4)iv
O(1)···N(2)iv
2.917
2.963
c
−
a Symmetry operations: (i) 2−x, −y, −z; (ii) 2−x, −y, 2−z; (iii) x−1, y+1,
z, (iv) x−1, y, z. b Derived from procedure shown in Fig. S3 (Supporting
Information). c Not available.
Fig. 3. Ladder−like hydrogen bond polymer of UO2(NO3)2(NCF)2.
Symmetry operations: (i) 2−x, −y, 2−z; (ii) x−1, y+1, z; (iii) 1−x, 1−y,
2−z; (iv) x+1, y−1, z; (v) 3−x, 1−y, 2−z.
Precipitation Behavior of UO2(NO3)2(L)2 in HNO3(aq).
From the X-ray crystallographic study, it was revealed that
2+
NCF interacts with UO2 as a monodentate ligand in the
−
presence of NO3 to form UO2(NO3)2(NCF)2 in a similar
manner to NCP, although several unique inter- and
intramolecular interactions were found in its molecular and

crystal structures. This fact allows us to expect that addition of
show similar coordination chemistry with UO2(NO3)2 which is
commonly found in the wide class of UO2(NO3)2(L)2. Also, it
was confirmed that NCF is still active in precipitating UO2
from HNO3(aq), even though the efficiency is not as high as
that using NCP. Therefore, chemical fundamentals to form
UO2(NO3)2(L)2 are common to both NCP and NCF.
2+
NCF affords efficient precipitation of UO2 from HNO3(aq)
like in the case of NCP.5, 7
2+
2+
In order to assess how efficiently UO2 is precipitated
from HNO3(aq), NCF in different loading amount was mixed to
3.0 M HNO3(aq) dissolving 1.5 M UO22+. The precipitation
yield (PYNCF, %) and effectiveness (ENCF, %) were plotted in
When UO2(NO3)2(NCF)2 precipitate was left in 3.0 M
HNO3(aq) at room temperature overnight, remarkable gas
evolution from the sample mixture was observed. Although we
did not perform chemical analysis of the evolved gas, it can be
speculated that NCF has been decomposed under the presence
of HNO3. In contrast, no visual change was observed in the
mixture of UO2(NO3)2(NCP)2 precipitate with 3.0 M HNO3(aq)
under the same condition, when NCP was used to precipitate
UO22+. Therefore, the chemical stability appears to be critically
different between NCF and NCP.
2+
Fig. 5 as functions of molar ratio between loaded L and UO2
initially present in the 3.0 M HNO3(aq). Consequently, it was
2+
confirmed that NCF also has a potential to precipitate UO2
from 3.0 M HNO3(aq). With an increase in NCF loading level,
PYNCF has gradually increased, and reached up to 79.1% at
[NCF]/[UO22+] = 2.0, which is the stoichiometric ratio from
UO2(NO3)2(NCF)2 thoroughly characterized in the former
section. Furthermore, ENCF stayed in the range from 72.1% to
78.2%, indicating that the loaded NCF was not fully utilized to
form UO2(NO3)2(NCF)2 precipitation.
1
To assess the chemical stability of NCF in HNO3(aq), H
The same experimental procedure was also repeated for
NCP to check reproducibility of its performance in the UO2
NMR spectra of 3.0 M HNO3(aq) dissolving 0.5 M NCF were
recorded at 50°C every 30 min. The initial concentration of
NCF in this experiment was lower than any of those in the
precipitation experiments of Fig. 5, while it was selected for
safety reason to avoid uncontrollable gas evolution in the NMR
instrument. Additionally, the test temperature was also higher
than that in the last section. The temperature employed here
(50°C) was optimal to study the chemical stability of NCF
within the realistic duration of experiment, because
decomposition of NCF at room temperature takes overnight as
2+
precipitation. The obtained data were also plotted in Fig. 5. As
a result, PYNCP were always higher than PYNCF, and reached
99.9% at [NCP]/[UO22+] = 2.0, where ENCP was 99.9%. Even at
the lower loading of NCP ([NCP]/[UO22+] = 1.5), precipitation
of UO22+ was nearly quantitative (PYNCP 73.0%, ENCP 97.3%) to
confirm reproducibility of the high performance of NCP in the
UO22+ precipitation from HNO3(aq).5, 7
As we have previously argued in the cases of NRPs,7, 8 the
difference between NCF and NCP observed in this study would
be ascribed to the difference in their hydrophobicity. To make
such a discussion more quantitative, the logarithmic partition
coefficient (log Po/w) of organic substance in a 1-octanol/water
biphasic system was employed as a measure of hydrophobicity.
From the actual distribution experiment of L in 1-octanol/water
biphasic system, log Po/w of NCF and NCP have been estimated
to be 1.04 and 2.39, respectively. As a result, NCF with lower
1
mentioned above. The obtained H NMR spectra are shown in
Fig. 6. In the initial spectrum, a broad singlet centered at 7.9
ppm can be assigned to >NH group, which is hydrogen bonded
in the aqueous sample solution to give such remarkable
broadening. The formyl H was observed at 7.51 ppm as a
singlet. The methine H in the cyclohexyl group was detected at
3.18 ppm as a broad singlet. The signals appeared in the high
field region from 0.6 ppm to 1.4 ppm are assigned to remaining
methylene H atoms of the cyclohexyl group. The latter two sets
of 1H signals described above are broadened by chair-boat
conformation exchange.
In Fig. 6, intensity of each signal of NCF decreased with
the elapse of time. Simultaneously, new signals have appeared
at 7.74 ppm, 6.80 ppm, and 2.65 ppm, which are assigned to
formyl group of formic acid, ammonium and methine groups of
protonated cyclohexylamine, respectively. The signal at 6.80
ppm is split into triplet by spin-spin coupling with parent 14N (I
= 1) with JHN = 50 Hz. Although some difference can be found
in the higher field region of cyclyhexyl methylene groups, they
are too crowded and too complicated to extract any information
about NCF decomposition. In summary, it is clarified that NCF
is hydrolyzed into cyclohexylamine and formic acid in 3.0 M
HNO3(aq).
2+
hydrophobicity resulted in lower performance for the UO2
precipitation from HNO3(aq) than that of NCP with higher
hydrophobicity. Note that no significant decomposition of both
ligands have been detected in the current experimental time
2+
window at 25°C. Therefore, the precipitation behavior of UO2
observed here are not affected by degradation of the ligands
(vide infra).
Fig. 5. Precipitation yield of UO2(NO3)2(L)2 and effectiveness of L (=
2+
NCP, NCF) as functions of molar ratio of loaded L against UO2
Initial condition: 1.5 M UO22+ in 3.0 M HNO3(aq).
.
Fig. 6. 1H NMR spectra of 0.50 M NCF in 3.0 M HNO3(aq) at 50°C at
different elapsed time. Asterisks: signals of TsOH as an internal
reference.
Chemical Stability of L in HNO3(aq). Through the
discussion in the previous sections, we have explored similarity
and difference between NCF and NCP. Although there are
minor differences in the crystal structures, both compounds

From the peak integrals, concentrations of NCF and
cyclohexylamine in each spectrum were estimated, and plotted
against elapsed time in Fig. 7. Note that formic acid cannot be
quantified correctly because of its further reaction with HNO3
to finally give CO2, H2O, and NOx.34-37 Furthermore, the same
experiments were also performed in 1.0 M and 5.0 M
HNO3(aq) (Figs. S4 and S5, Supporting Information). The
results obtained in these different [HNO3] conditions were also
displayed in Fig. 7. With increasing [HNO3] from 1.0 M to 3.0
M, the hydrolysis of NCF got to be enhanced, while this
reaction was slightly suppressed with further increase in
[HNO3] from 3.0 M to 5.0 M, implying that NCF is regenerated
under higher [HNO3]. This seems to be unusual at a glance.
However, as both hydrolysis of the amide bond and its
formation through condensation are reversibly catalyzed by H+,
a set of these forward and backward reactions of NCF tends to
be equilibrated in the actual sample solution. However, the
hydrolytic decomposition of NCF is actually predominant as
shown in Scheme 1 because of the successive reaction of
resulting formic acid with HNO3. Partial pressures of gaseous
final products like CO2 and NOx in Scheme 1 cannot be
controlled in the current experiments. Furthermore, equilibria
of NOx are too complicated to be dealt with in this reaction
system. Therefore, it is difficult to extract further details about
decomposition kinetics of NCF from Fig. 7.
will induce positive polarization of the carbonyl C, enhancing
electrophilicity of this moiety. Nevertheless, NCP has not been
hydrolyzed at 50°C in 3.0 M HNO3(aq) at least for 4 h.
One can immediately realize that nucleophilic attack of
H2O to the carbonyl C is sterically hindered by the pyrrolidone
ring. We have further considered whether or not there are any
electronic contributions arising from such a heterocyclic
structure. After structure optimization of each ligand by the
ORCA program,21 Mulliken population analysis has been
performed. As a result, gross atomic charge on the carbonyl C
of NCP is +0.19, which is slightly greater than that of NCF
(+0.17). Therefore, it is difficult to find strong correlation
between effective charge of the carbonyl C and tendency
towards hydrolysis shown in Scheme 1. This implies that an
electronic effect is less important to understand the difference
in chemical stability between NCF and NCP. Thus, the essential
role of the pyrrolidone ring can be ascribed to steric protection
of carbonyl C from nucleophilic attach by H2O.
Conclusion
In this study, NCP and NCF were selected respectively as
cyclic and acyclic monodentate amide ligands (L) to be
compared, and aimed to clarify an essential role of the
pyrrolidone ring of NRPs under the precipitation-based nuclear
fuel reprocessing conditions bearing a few molar HNO3.
Although there are several minor differences were detected in
molecular and crystal structures, fundamental coordination
chemistry of UO2(NO3)2(L)2 were found to be basically
common to both NCP and NCF, regardless of the presence or
absence of the pyrrolidone ring. In terms of precipitation
2+
behavior of UO2 from HNO3(aq), NCF is lower in the
precipitation yield and its effectiveness compared to NCP
presumably due to its lower hydrophobicity. However, NCF is
still feasible to be utilized for this objective. The most critical
difference between NCP and NCF was observed in their
chemical stability in HNO3(aq), where NCF was gradually
decomposed through acid-catalyzed hydrolysis, while NCP is
durable to remain unchanged for at least 4 h.
Summarizing all the information obtained in this study,
we can conclude that the pyrrolidone ring plays an important
role to sterically protect carbonyl C of its amide moiety from
nucleophilic attack by H2O. It should be emphasized that any
cleavages at the N−CH2 and C(=O)−CH2 bonds of NCP
observed under the γ-ray irradiation39 have not been detected in
the current chemical stability assessment. Such a difference
suggests that the radiolytic decomposition of NCP undergoes
unique reaction pathways through direct bond breaking by the
γ-ray and/or indirect effect by radiolytically induced hydroxy
radicals, not being observable in ordinary chemical mechanism.
Fig. 7. Decomposition behavior of NCF in different HNO3(aq) at 50°C
with elapse of time.
Scheme 1. Acid-catalyzed hydrolytic decomposition of NCF in
HNO3(aq).
We thank Prof. Emer. Yasuhisa Ikeda for his helpful
advice. This work contains the results of “Development of
Advanced Reprocessing System Based on Use of Pyrrolidone
Derivatives as Novel Precipitants with High Selectivity and
Controllability” and “Fundamental Study on Simple
Reprocessing Method for Spent Thorium Fuels by Using
Uranium-Selective Precipitant” entrusted to Tokyo Institute of
Technology by MEXT, Japan. This work was also supported by
World Research Hub Initiative (WRHI) Program of Institute of
Innovative Research, Tokyo Institute of Technology. Quantum
chemical calculations were performed at the Center for
Information Services and High-Performance Computing (ZIH)
at the Technische Universität Dresden, Germany, using the
library program ORCA.
The chemical stability of NCP was also studied at 50°C in
3.0 M HNO3(aq) in a similar manner to NCF. As shown in Fig.
S6 (Supporting Information), the H NMR spectrum remained
unchanged even after 4 h, indicating that NCP is much more
chemically durable than NCF. This must be exclusively due to
the presence of the pyrrolidone ring. Actually, the carbonyl O
of the pyrrolidone group may interact with H+ as we reported
recently, where crystalline anhydrous H+ adducts of pyrrolidone
derivatives were isolated from HNO3(aq).13, 14, 38 This effect
1

Supporting Information
IR spectra of free L and UO2(NO3)2(L)2, powder XRD
patterns of UO2(NO3)2(L)2 as precipitated from 3.0
23 N. Koshino, M. Harada, M. Nogami, Y. Morita, T.
Kikuchi, Y. Ikeda, Inorg. Chim. Acta, 2005, 358, 1857.
24 C. E. F. Rickard, Acta Crystallogr., Sect. E: Struct. Online,
2001, 57, m446.
25 A. E. Bradley, C. Hardacre, M. Nieuwenhuyzen, W. R.
Pitner, D. Sanders, K. R. Seddon, R. C. Thied, Inorg.
Chem. 2004, 43, 2503.
M
−
HNO3(aq), tilt angles of NO3 from uranyl equatorial plane of
UO2(NO3)2(L)2 (L = NCF, NRPs) and schematic drawing for its
evaluation, 1H NMR spectra of NCF in 1.0 M and 5.0 M
1
HNO3(aq), and H NMR spectra of NCP in 3.0 M HNO3(aq)
are presented in the Supporting Information. This material is
available free of charge on J-STAGE.
26 J. G. Reynolds, A. Zalkin, D. H. Templeton, Inorg Chem
1977, 16, 3357.
Crystallographic data reported in this manuscript have
been deposited with Cambridge Crystallographic Data Centre
as supplementary publication no. CCDC-1987422 and 1987423.
Copies of the data can be obtained free of charge via CCDC
Website.
27 K.-A. Hughes, P. C. Burns, Acta Crystallogr., Sect. C:
Cryst. Struct. Commun. 2002, 59, i7.
28 A. Bondi, J. Phys. Chem. 1964, 68, 441.
29 A. Bauzá, A. Frontera, T. J. Mooibroek, Nat. Commun.
2017, 8, 14522.
30 S. Fortier, T. W. Hayton, Coord. Chem. Rev. 2010, 254,
197.
References
1
2
3
L. R. Morss, N. M. Edelstein, J. Fuger, The Chemistry of
the Actinide and Transactinide Elements. 4th ed.,
Springer, Dordrecht, The Netherlands, 2011.
M. Benedict, T. H. Pigford, H. W. Levi, Nuclear
Chemical Engineering. 2nd ed., McGraw-Hill, United
States, 1981.
Introduction of Thorium in the Nuclear Fuel Cycle Short-
to Long-term Considerations (NEA No. 7224). OECD
Nuclear Energy Agency, 2015.
I. Kumari, B. V. R. Kumar, A. Khanna, Nucl. Eng. Des.
2020, 358, 110410.
T. R. Varga, M. Sato, Z. Fazekas, M. Harada, Y. Ikeda, H.
Tomiyasu, Inorg. Chem. Commun. 2000, 3, 637.
T. R. Varga, A. C. Bényei, Z. Fazekas, H. Tomiyasu, Y.
Ikeda, Inorg. Chim. Acta, 2003, 342, 291.
31 P. L. Arnold, A. F. Pecharman, E. Hollis, A. Yahia, L.
Maron, S. Parsons, J. B. Love, Nat. Chem. 2010, 2, 1056.
32 M. J. Sarsfield, M. Helliwell, J. Am. Chem. Soc. 2004,
126, 1036.
33 T. Steiner, Chem. Commun. 1997, 0, 727.
34 R. F. Bradley, C. B. Goodlett, Denitration of Nitric Acid
Solutions by Formic Acid. Savannah River Laboratory,
Aiken, South Carolina, 1972.
35 H. Nakamura, I. Yamaguchi, M. Kubota, J. Nucl. Sci.
Technol. 1978, 15, 760.
36 M. Kubota, I. Yamaguchi, H. Nakamura, J. Nucl. Sci.
Technol. 1979, 16, 426.
37 W. Li, W. Duan, T. Sun, C. Liu, J. Wang, J. Chen, J.
Radioanal. Nucl. Chem. 2017, 314, 221.
38 H. Kazama, S. Tsushima, K. Takao, Cryst. Growth Des.
2019, 19, 6048.
39 M. Nogami, Y. Sugiyama, T. Kawasaki, M. Harada, Y.
Kawata, Y. Morita, T. Kikuchi, Y. Ikeda, Sci. China Chem.
2012, 55, 1739.
4
5
6
7
8
N. Koshino, M. Harada, Y. Morita, T. Kiikuchi, Y. Ikeda,
Prog. Nucl. Energy, 2005, 47, 406.
Y. Morita, K. Takao, S.-Y. Kim, Y. Kawata, M. Harada,
M. Nogami, K. Nishimura, Y. Ikeda, J. Nucl. Sci. Technol.
2009, 46, 1129.
9
Y. Ikeda, E. Wada, M. Harada, T. Chikazawa, T. Kikuchi,
H. Mineo, Y. Morita, M. Nogami, K. Suzuki, J. Alloys
Compd. 2004, 374, 420.
10 K. Takao, K. Noda, Y. Morita, K. Nishimura, Y. Ikeda,
Cryst. Growth Des. 2008, 8, 2364.
11 K. Takao, Y. Ikeda, H. Kazama, Ener. Proc. 2017, 131,
157.
12 H. Kazama, S. Tsushima, Y. Ikeda, K. Takao, Inorg.
Chem. 2017, 56, 13530.
13 K. Takao, H. Kazama, Y. Ikeda, S. Tsushima, Angew.
Chem. Int. Ed. 2019, 58, 240.
14 K. Takao, J. März, M. Matsuoka, T. Mashita, H. Kazama,
S. Tsushima, RSC Adv. 2020, 10, 6082.
15 B. G. Vats, D. Das, B. Sadhu, S. Kannan, I. C. Pius, D. M.
Noronha, M. Sundararajan, M. Kumar, Dalton Trans.
2016, 45, 10319.
16 S. Kannan, M. Kumar, B. Sadhu, M. Jaccob, M.
Sundararajan, Dalton Trans. 2017, 46, 16939.
17 S.-Y. Kim, K. Takao, Y. Haga, E. Yamamoto, Y. Kawata,
Y. Morita, K. Nishimura, Y. Ikeda, Cryst. Growth Des.
2010, 10, 2033.
18 A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi,
J. Appl. Crystallogr. 1993, 26, 343.
19 G. M. Sheldrick, Acta Crystallogr. 2015, C71, 3.
20 CrystalStructure 4.2.7, Crystal Structure Analysis
Package; Rigaku Corporation, Tokyo, Japan, 2017.
21 F. Neese, WIREs Comput. Mol. Sci. 2018, 8, e1327.
22 S.-Y. Kim, K. Takao, Y. Haga, E. Yamamoto, Y. Kawata,
Y. Morita, K. Nishimura, Y. Ikeda, Cryst. Growth Des.
2010, 10, 2033.