Search strategy for new efficient organophosphorus extractants for concentrating radionuclides

Full text of DOI:10.1134/S0012500808100054

ISSN 0012-5008, Doklady Chemistry, 2008, Vol. 422, Part 2, pp. 260–264. © Pleiades Publishing, Ltd., 2008.
Original Russian Text © I.G. Tananaev, A.A. Letyushov, A.M. Safiulina, I.B. Goryunova, T.V. Baulina, V.P. Morgalyuk, E.I. Goryunov, L.A. Gribov, E.E. Nifant’ev,
B.F. Myasoedov, 2008, published in Doklady Akademii Nauk, 2008, Vol. 422, No. 6, pp. 762–766.
CHEMISTRY
Search Strategy for New Efficient Organophosphorus
Extractants for Concentrating Radionuclides
Corresponding Member of the RAS I. G. Tananaeva, A. A. Letyushova, A. M. Safiulinab,
I. B. Goryunovac, T. V. Baulinac, V. P. Morgalyukc, E. I. Goryunovc,
Corresponding Member of the RAS L. A. Gribovb, Corresponding Member of the RAS E. E. Nifant’evc,
and Academician B. F. Myasoedovb
Received June 30, 2008
DOI: 10.1134/S0012500808100054
Long-term systematic studies of actinide and lan- tive extraction constant) of uranyl nitrate and nitric
thanide extraction from acid solutions have implied that acid, on the one hand, and the sums of the electronega-
tivities of substituents at the phosphorus atom in this
series of compounds, on the other. From calculations,
the extraction ability of MNOPCs decreases with
increasing substituent electronegativity.
bidentate neutral organophosphorus compounds
(BNOPCs) are the most efficient extractants for these
metals [1]. Naturally, new, even more efficient extract-
ants are expected to exist among these compounds. In
our opinion, the most promising strategy for choosing
such extractants is a systematic theoretical search for
BNOPCs with high complexing abilities to 4f and 5f
elements followed by experimental verification of the-
oretical inferences.
Although electronegativities can be determined
autonomously, the above correlation unfortunately
cannot be applied to BNOPCs, which are of the most
practical importance [1]. In this context, another
approach to the problem was proposed on the basis
of quantum-chemical calculations of molecular elec-
trostatic fields (MESFs); we have successfully used
this approach [6]. Not only did this strategy reveal
the calculation parameters that can explain distinc-
tions in reactivities of organophosphorus com-
pounds, but it also helped us to create a complex
quantum-chemical approach to predicting the extrac-
tion ability of these compounds, including transu-
ranic elements.
One major task of this search is to determine the
relationship between the extraction ability and structure
of organophosphorus compounds. Correlations of
extraction constants with some features of structure
families were used for this purpose, for example, corre-
lations with the electronegativity [2], Taft constant σ∗
[3], and Kabachnik constant σp [4] (on the assumption
that the substituent effect is additive). In particular, it
was found for various types of monodentate neutral
organophosphorus compounds (MNOPCs)—phos-
phates, phosphonates, phosphinates, and phosphine
oxides—that the reactive site (P=O group) and the com-
position of the complex (the ratio element : ligand)
remain unchanged, while the extraction ability consid-
erably increases in this series. Rozen et al. [5] proposed
a correlation between lnKex values (where Kex is effec-
Here, we use MESF quantum-chemical calculations
to choose N-diphenylphosphoryl-N'-alkylureas as the
most effective BNOPCs for concentrating and separat-
ing 4f and 5f elements; we also compare theoretical pre-
dictions with experiments.
Concept of the strategy. Complexing between an
organic reagent and a metal ion (with the desolvation of
the reacting species being ignored) can be divided into
two stages. First, the metal cation and the ligand group
(LG) of the reagent come closer to each other. Second,
they directly react at distances close to chemical bond
lengths; this reaction involves electron exchange and
the generation of the relevant complex. It is believed
that the removed substituents only insignificantly alter
the electronic state of the atoms in the LG of the
reagent, which should not considerably affect its reac-
tivity to the given metal atom.
a Frumkin Institute of Physical Chemistry
and Electrochemistry, Russian Academy of Sciences,
Leninskii pr. 31, Moscow, 119991 Russia
b Vernadsky Institute of Geochemistry and Analytical
Chemistry, Russian Academy of Sciences, ul. Kosygina 19,
Moscow, 119991 Russia
c Nesmeyanov Institute of Organoelement Compounds,
Russian Academy of Sciences, ul. Vavilova 28,
Moscow, 119991 Russia
260

SEARCH STRATEGY FOR NEW EFFICIENT ORGANOPHOSPHORUS EXTRACTANTS
261
Table 1. Comparison of the MESF parameters calculated for N-diphenylphosphoryl-N'-n-alkylureas Ph2P(O)NHC(O)NHR (II)
and (N-n-alkylcarbamoylmethyl)diphenylphosphine oxides RNHC(O)CH2P(O)Ph2 (III)
Oxygen–oxygen dis-
Torsion angles be-
Relative charges q on oxygen
tances between
P=O and C=O groups,
Å
MESF zone
width, Å
tween P=O
atoms in LGs, au
and C=O groups, deg
R
(II)
(III)
(II)
(III)
(II)
(III)
(II)
(III)
P=O
C=O
P=O
C=O
n-C5H11
n-C6H13
n-C7H15
n-C8H17
n-C9H19
n-C10H21
36
39
40
38
38
39
60
56
51
56
49
47
–0.638
–0.639
–0.638
–0.638
–0.638
–0.638
–0.339
–0.340
–0.340
–0.339
–0.339
–0.339
–0.628
–0.626
–0.624
–0.626
–0.623
–0.623
–0.318
–0.315
–0.313
–0.315
–0.312
–0.311
3.26
3.24
3.22
3.25
3.23
3.21
3.53
3.50
3.50
3.51
3.49
3.54
7.87
8.20
8.75
8.13
7.85
7.82
6.42
6.65
6.80
6.95
6.90
6.90
Note. The bond lengths in the P=O and C=O groups in compounds II and III are equal to 1.51 and 1.23 Å, respectively.
Considering the first stage, one finds that a complex sion angles, less strongly bonded complexes are
formed.
electrostatic field is generated in electrically neutral
organic reagent molecules due to the specific distribu-
tion of negative electron charges in a confined region
around atomic nuclei. When such a neutral organic
molecule approaches the metal cation, their MESFs
interact. This interaction induces the polarization of the
electronic shells of organic ligand molecules, either
enhancing the reaction or hindering or even inhibiting
it [7]. The MESFs of the reagents are characterized by
positive and negative regions with the relevant poten-
tials. Overlapping of different-sign MESFs brings mol-
ecules closer to each other; overlapping of similar
MESFs induces repulsion. Molecular attraction
strengthens with an increasing product of the absolute
values of potentials at points in the overlap region.
Here, we assume that there are no competitive reac-
tions (e.g., the protonation of electron-donating atoms
of the LG) or solvation shells. This allows us to apply
the quantum-chemical approach directly to the reaction
between the metal ion and organic reagent and to
explain, using the chosen parameters, the different
reactivities of various types of reagents.
Calculation procedure. The LEV program (devel-
oped at the Vernadsky Institute of Geochemistry and
Analytical Chemistry) was used for quantum-chemical
simulations. Initial conformations of the test structures
that approached the real conformations as maximally as
possible and ensured the highest reactivities were cre-
ated using molecular imaging means. The torsion angle
between the LGs (P=O and C=O) was nullified; that is,
the test compounds were initially endowed with an
ability to form bidentate complexes.
The conformations were calculated using the
MNDO semiempirical method. The first stage included
geometry optimization and a search for the most stable
conformations with a minimal energy; as a result, three-
dimensional images of optimized models were created.
The spatial arrangement of atoms, electron densities,
effective atomic charges, bond lengths, bond angles,
and torsion angles were also calculated at this stage. At
the second stage, MESFs were calculated and MESF
distribution maps were designed, including geometry
optimization data.
Knowledge of the MESF distribution can help in
studying differences between interactions at the first
and second complexing stages. A series of control
parameters responsible for noticeable differences in
reactivities can be distinguished. For example, the
higher the effective negative charge of the electron-
donating atom of the LG, the stronger the interaction
between this group and the metal ion and, thus, the
higher the reactivity of the compound.
In cases where there are two LGs in the reagent mol-
ecule, each containing an electron-donating atom (e.g.,
P=O and C=O groups), an additional control parameter
appears, namely, the torsion angle between these
groups. When this angle is null (that is, when both LGs
are in one plane) or small, the organic reagent can form
Choice of models. Structural modification of
compounds that have already demonstrated their
strongly bonded bidentate complexes. With greater tor- high efficiency seemed the most logical strategy for
DOKLADY CHEMISTRY Vol. 422 Part 2 2008

262
TANANAEV et al.
Table 2. Comparison of uranium(VI) distribution ratios DU
in extraction from 2.81 M HNO3 to 0.05 M solutions of com-
pounds IIb–IIf in chloroform as a function of radical length
R (phase ratio: 1 : 1)
existence of two phenyl radicals at the phosphorus atom
in a molecule of a potential extractant is the optimal
structural variant to ensure the maximal extraction effi-
ciency.
As regards variation of substituents at the nitrogen
atom, recent studies [11] showed that the CMPOs gen-
erated by the replacement of the N,N-dibutyl group in
DU
Compound
R
62
53
70
44
52
(IIb)
(IIc)
(IId)
(IIe)
(IIf)
n-C6H13
n-C7H15
n-C8H17
n-C9H19
n-C10H21
molecule I by an NHAlk group (where Alk = CnH2n + 1
)
are not inferior to compound I in actinide extraction
efficiency from nitric acid solutions and are even
slightly superior to it in synthetic availability. With
account for this, the BNOPC structures containing a
terminal NHAlk moiety are preferred from the practical
standpoint for the design of potential organophospho-
rus extractants.
Notably, the CH2 spacer between the phosphoryl
and carbonyl complexing groups in a CMPO molecule
does not ensure electronic interactions between these
groups. Therefore, the replacement of the methylene
spacer by a divalent radical seems, in fact, the only vari-
ant of structural modification of CMPOs that is poten-
tially capable of radically improving the complexing
properties: this replacement would ensure delocaliza-
tion of the electron system of the complexing fragment,
which in turn would enhance the efficiency of cation
binding. An imide group –NH– can be one such bridg-
ing radical; as a result, the carbamoylmethylphosphine
oxide structure would transform into the N-phosphory-
lurea structure. To achieve the best result, these urea
molecules should retain the substituents at the phos-
phorus atom and terminal nitrogen atom that were rec-
ognized as optimal for CMPOs.
Table 3. Uranium(VI) distribution ratios DU as a function
of HNO3 concentration in extraction to 0.05 M solutions
of compounds I and IId in chloroform (phase ratio: 1 : 1)
DU
[HNO3], mol/L
(I)
(IId)
0.58
2.81
6.52
1.0
1.3
1.5
2.6
70
72
searching among BNOPCs for new types of potential
extractants for 4f and 5f elements. At present, car-
bamoylmethylphosphine oxides (CMPOs) are the
best organophosphorus extractants for actinide and
lanthanide extraction from nitric acid solutions [8, 9].
Of them, (N,N-dibutylcarbamoylmethyl)diphenylphos-
phine oxide Ph2P(O)CH2C(O)NBu2 (I) has the highest
activity [8]; in particular, it is used in the Russian ver-
sion of the TRUEX process.
Fundamentally, there are three variants of structural
modification of this compound: replacement of substit-
uents at the phosphorus atom, variation of substituents
at the nitrogen atom, and replacement of the methylene
group by another spacer.
To test this idea, we carried out MESF quantum-
chemical calculations for both N-diphenylphosphoryl-
N'-n-alkyl(C5–C10)ureas (IIa–IIf) and their closest ana-
logues in the CMPO series, chosen to be references:
[N-n-alkyl(C5–C10)carbamoylmethyl]diphenylphosphine
oxides IIIa–IIIf.
O
O
O
O
R
H
R
H
C6H5
C6H5
C6H5
C6H5
C N
C N
P NH
P CH2
(IIa–IIf)
(IIIa–IIIf)
Systematic investigations of the effect of substitu-
ents at the phosphorus atom on the extraction ability of
CMPOs were successfully performed under the super-
vision of Academicians M.I. Kabachnik and B.F. Mya-
soedov at the end of the last century in the course of
optimization of the TRUEX process. Anomalous aryl
strengthening was observed in the CMPO series; this
effect, first discovered in the study of extraction proper-
ties of methylenediphosphine dioxides, consisted of the
following: the replacement of alkyls at the phosphorus
atom in a BNOPC molecule by more electronegative
a: R = n-C5H11, b: R = n-C6H13, c: R = n-C7H15,
d: R = n-C8H17, e: R = n-C9H19, f: R = n-C10H21
Because all other structural fragments in both types
of compounds were identical apart from the bridge
between the phosphoryl and carbonyl groups, all dis-
tinctions revealed by the calculations can be completely
assigned to the replacement of the methylene spacer in
CMPOs (III) by an imide spacer in phosphorylureas
(II). The results are summarized in Table 1.
The quantum-chemical calculations allowed us to
aryl substituents considerably improves the extraction compare the most important parameters for the reactiv-
characteristics of the compound, despite the attendant ity of BNOPCs. Such parameters are the effective
decrease in basicity [10]. Thus, in all probability, the charges on the oxygen atoms of the phosphoryl and car-
DOKLADY CHEMISTRY Vol. 422 Part 2 2008

SEARCH STRATEGY FOR NEW EFFICIENT ORGANOPHOSPHORUS EXTRACTANTS
263
bonyl groups, the torsion angles between these groups, CMPOs IIIa–IIIf differ considerably: the MESF zone
and the MESF zone width around the P=O and C=O width in CMPOs III is 6.77 Å, whereas in ureas II, it is
groups of the models.
close to 8.10 Å.
The results of the MESF quantum-chemical calcula-
tions also imply that the complexing and extraction
properties of N-diphenylphosphoryl-N'-n-alkyl(C5–
C10)ureas IIa–IIf should be far higher than those for
We can infer that the replacement of the methylene
by an imide spacer in the model compounds renders the
structure more rigid because of the decrease in the tor-
sion angle between the P=O and C=O groups. While
this torsion angle for CMPOs IIIa–IIIf is about 53°, for
compounds IIa–IIf it is close to 38°. As a result, the
distances between the phosphoryl and carbonyl oxygen
atoms in phosphorylureas II are shorter than in related
CMPOs III, which should evidently improve the com-
plexing ability of such ureas.
Average relative negative charges q on phosphoryl
and carbonyl oxygen atoms for CMPOs III do not
exceed –0.625 and –0.314 au, respectively; for phos-
phorylureas II, the respective values are slightly higher:
–0.638 and –0.339 au.
related
[N-n-alkyl(C5–C10)carbamoylmethyl]diphe-
nylphosphine oxides IIIa–IIIf due to the decreased tor-
sion angles between the P=O and C=O groups, closer
values of effective charges q on the oxygen atoms of the
LG, and considerably increased width of the MESF
zone. Of the aforementioned ureas, compounds with
the n-alkyl length ranging from hexyl to octyl should
have the highest efficiency.
To experimentally verify the above inference, we
synthesized a set of phosphorylureas (IIa–IIf) using a
recent three-stage one-pot process [12] and proceeding
While the P=O and C=O atomic distances in the test from low-cost commercially available reagents
classes of compounds are almost identical, the MESF (Scheme 1). This process produces target compounds
zone widths in ureas IIa–IIf and the corresponding with high yield (>90%) and high purity (at least 97%).
SO2Cl2, ≈20°C
Ph2PCl
[Ph2P(O)Cl]
CCl4
NaOCN, [MgCl2], ≈20°C
RNH2, ≈20°C
[Ph2P(O)NCO]
Ph2P(O)NHC(O)NHR
MeCN
IIa–IIf
Scheme 1.
Uranium(VI) extraction by 0.05 M solutions of N'-n-alkylureas for separating 4f and 5f elements,
1
compounds IIb–IIf from 2.81 M HNO3 to chloroform inferred from MESF quantum-chemical simulations,
and implies the potential of this strategy for finding
showed that, indeed, the highest distribution ratios in
this set of compounds are achieved for ureas with N'-n-
alkyl radicals C6–C8; N'-n-octylurea IId has the maxi-
mal efficiency (Table 2).
logDEu
2.0
The ability to extract uranium(VI) from nitric acid
solutions of various strengths for this urea is far higher
than for the most effective CMPO extractant (diphe-
nyl(N,N-dibutylcarbamoylmethyl)phosphine oxide I);
this effect is most prominent for high acid concentra-
tions (Table 3).
1.5
1.0
IId
0.5
0
IIId
Moreover, the predicted advantage of phosphoryl-
ureas (II) over CMPOs in extraction ability is actually
observed not only for actinides but also for lanthanides:
in ability to extract europium(III) from nitric acid solu-
tions, phosphorylurea IId exceeds structurally related
phosphine oxide IIId by more than two orders of mag-
nitude (figure).
0
1
2
3
4
5
6
[HNO3], mol/L
–0.5
–1.0
–1.5
–2.0
To summarize, the experimental check completely
validates the high efficiency of N-diphenylphosphoryl-
Extraction of indicator europium(III) amounts by 0.1 M
solutions of Ph P(O)NHC(O)NH-n-C H (IId) and
Ph P(O)CH C(O)NH-n-C H (IIId) in chloroform vs.
2
8 17
1
Urea IIa is not sufficiently soluble in chloroform for obtaining
solutions of the required concentration.
2
2
8 17
HNO concentration (phase ratio: 1 : 1).
3
DOKLADY CHEMISTRY Vol. 422 Part 2 2008

264
TANANAEV et al.
3. Taft, R.W., Gurka, D., Joris, L., et al., J. Am. Chem. Soc.,
new, improved types of extractants for solving various
engineering and analytical problems.
1969, vol. 91, no. 17, pp. 4801–4808.
4. Mastryukova, T.A. and Kabachnik, M.I., Usp. Khim.,
1969, vol. 38, no. 10, pp. 1751–1782.
ACKNOWLEDGMENTS
5. Rozen, A.M., Klimenko, N.M., Krupnov, B.V., and Niki-
forov, A.S., Dokl. Akad. Nauk SSSR, 1986, vol. 287,
no. 4, pp. 915–919.
6. Sudarushkin, S.K., Morgalyuk, V.P., Tananaev, I.G.,
et al., Izv. Akad. Nauk, Ser. Khim., 2006, no. 4, pp. 609–
615.
We are grateful to S.K. Sudarushkin for help in
quantum-chemical calculations and O.I. Artyushin for
supplying us with a sample of phosphine oxide IIId.
This work was supported by the Council for Grants
of the President of the Russian Federation for Support
of Leading Scientific Schools (grant no. NSh–
6602.2006.3), the Russian Foundation for Basic
Research (project nos. 05–03–08017 and 05–03–
32908), and the Russian Academy of Sciences through
Program no. 8 “Synthesis of New Chemicals and Mate-
rials Design”).
7. Gribov, L.A. and Mushtakova, S.P., Kvantovaya khimiya
(Quantum Chemistry), Moscow: Gardarika, 1999.
8. Myasoedov, B.F., Chmutova, M.K., Kochetkova, N.E.,
et al., Solvent Extr. Ion Exch., 1986, vol. 4, no. 1, pp. 61–
81.
9. Horwitz, E.P., Kalina, D.G., Diamond, H., et al., Solvent
Extr. Ion Exch., 1985, vol. 3, nos. 1/2, pp. 75–109.
10. Kabachnik, M.I., Myasoedov, B.F., Mastryukova, T.A.,
et al., Izv. Akad. Nauk, Ser. Khim., 1996, no. 11,
pp. 2624–2630.
REFERENCES
1. Rozen, A.M. and Krupnov, B.V., Usp. Khim., 1996,
11. Artyushin, O.I., Sharova, E.V., Odinets, I.L., et al., Izv.
Akad. Nauk, Ser. Khim., 2004, no. 11, pp. 2395–2403.
vol. 65, no. 11, pp. 1052–1079.
2. Rozen, A.M. and Nikolotova, Z.I., Zh. Neorg. Khim., 12. Goryunov, E.I., Nifant’ev, E.E., and Myasoedov, B.F.,
1964, vol. 9, no. 7, pp. 1725–1743.
RF Patent 2 296 768, Byull. Izobret., 2007, no. 10.
DOKLADY CHEMISTRY Vol. 422 Part 2 2008