Zirconium(IV)-Benzene Phosphonate Coordination Polymers: Lanthanide and Actinide Extraction and Thermal Properties

Article

pubs.acs.org/IC

Zirconium(IV)−Benzene Phosphonate Coordination Polymers:

Lanthanide and Actinide Extraction and Thermal Properties

Vittorio Luca,*^,†Juan J. Tejada,^†Daniel Vega,^†Guilhem Arrachart,^‡and Cyrielle Rey^‡

^†Comision

Aires, Argentina

^‡Institut de Chimie Sep

F-30207 Bagnols sur Cez

́ ́ ́

Nacional de Energía Atomica, Centro Atomico Constituyentes, Avenida General Paz 1499, San Martin 1650, Buenos

́

arative de Marcoule, ICSM UMR 5257, CEA-CNRS-UM-ENSCM, Site de Marcoule, Bat

̂

iment 426 BP 17171,

̀

e, France

S

* Supporting Information

ABSTRACT: Coordination polymers with diﬀerent P/(Zr + P) molar ratios were

prepared by combining aqueous solutions of Zr(IV) and benzenephosphonate

derivatives. 1,3,5-Benzenetrisphosphonic acid (BTP) as well as phosphonocarbox-

ylate derivatives in which carboxylate substitutes one or two of the phosphonate

groups were chosen as the building blocks. The precipitates obtained on combining

the two solutions were not X-ray amorphous but rather were indicative of poorly

ordered materials. Hydrothermal treatment did not alter the structure of the

materials produced but did result in improved crystalline order. The use of HF as a

mineralizing agent during hydrothermal synthesis resulted in the crystallization of at

least three relatively crystalline phases whose structure could not be determined

owing to the complexity of the diﬀraction patterns. Gauging from the similarity of the

diﬀraction patterns of all the phases, the poorly ordered precipitates and crystalline

materials appeared to have similar underlying structures. The BTP-based zirconium

phosphonates all showed a higher selectivity for lanthanides and thorium compared with cations such as Cs⁺, Sr²⁺, and Co²⁺.

Substitution of phosphonate groups by carboxylate groups did little to alter the pattern of selectivity implying that selectivity in

the system was entirely determined by the −POH group with little inﬂuence from the −COOH groups. Samples with the highest

phosphorus content showed the highest extraction eﬃciencies for lanthanide elements, especially the heavy lanthanides such as

Dy³⁺and Ho³⁺with separation factors of around four with respect to La³⁺. In highly acid solutions (4 M HNO₃) there was a

pronounced variation in extraction eﬃciency across the lanthanide series. In situ, nonambient diﬀraction was performed on

ZrBTP-0.8 loaded with Th, Ce, and a complex mixture of lanthanides. In all cases the crystalline Zr₂P₂O₇pyrophosphate phase

was formed at ∼800 °C demonstrating the versatility of this structure.

of solvents and extractants. Furthermore, the deployment of

solid-phase extraction technologies is straightforward compared

with liquid−liquid extraction.

1. INTRODUCTION

The separation or partitioning of trivalent lanthanides from

trivalent minor actinides and subsequent transmutation of the

latter in fast reactors represents an important option in the

move to a more sustainable fuel cycle for nuclear ﬁssion energy,

since the potential beneﬁts are reduced repository heat load and

improved energy utilization¹as well as greatly reduced long-

term waste radiotoxicity.²

The major approaches for the separation of trivalent minor

actinides from other ﬁssion product isotopes have been liquid−

liquid extraction, otherwise known as solvent-extraction, and

pyroelectrochemical separations.^2,3Even though the latter is a

dry process not involving the use of solvents, it is the former

that is at a more advanced stage of development and

deployment. Solid-phase extraction is an alternative to

liquid−liquid extraction that in relative terms has been little

contemplated. It has the advantage of not requiring the organic

solvents that represent a signiﬁcant ﬁre/explosion hazard within

nuclear installations and not having to deal with third-phase

formation that is another undesirable aspect of solvent-

extraction along with the need to manage the decomposition

Because of the similarity of their chemistry the trivalent

lanthanides and minor actinides are notoriously diﬃcult to

separate. For the same reasons intralanthanide extractions are

also diﬃcult to accomplish and often require hundreds of

laborious extraction cycles.⁴Since the predominant oxidation

state of lanthanides and minor actinides is generally three under

environmental conditions they are classiﬁed according to

Pearson as hard ions. They are thus most likely to form strong

complexes with hard ligands containing highly electronegative

elements such as F⁻, SO₄²⁻, CO₃²⁻, PO₄³⁻, and OH⁻.⁵

Organophosphorus reagents such as TBP, DTPA, CMPO,

and 2-ethylhexyl phosphoric acid (HDEHP) are therefore

among the most useful extractants in solvent extraction systems

and form the basis of many solvent extraction processes for the

separation of lanthanides from actinides.⁶⁻⁹

Received: April 17, 2016

A

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

The key element of a solid-phase extraction technology is the

selectivity of the solid-phase extractant for the target elements

or radionuclides. The literature is indeed replete with

descriptions of interesting solid materials with potentially

interesting adsorption properties, although relatively few of

these have been studied in great detail. For instance, extensive

commercial use is made of porous materials such as zeolites for

applications in the extraction of heavy metal ions and the

separation and storage of gases as well as in catalysis. Zeolite-

like mixed-coordination silicotitanates and related materials

have applications for the selective adsorption of radio

cesium.^10,11Functionalized mesoporous metal oxides prepared

by anchoring of organic molecules such as organosilanes and

organophosphonates to the surfaces of mesoporous materials

prepared through supramolecular self-assembly oﬀer consid-

erable tailorability, since there is scope to vary both the organic

and inorganic component and the porosity of the oxide. The

use of high surface area supports is generally considered a

means of maximizing the number of surface active sites and

therefore capacity. Indeed our group and others have shown

considerable potential of such functionalized mesoporous

oxides for the adsorption of lanthanides and actinides.¹²⁻²⁰

One of the limitations of such functionalized mesoporous oxide

materials, however, is the potential fragility of the coupling

between metal oxide and the functional organic group. Another

limitation is the fact that the loading of functional organic

groups on the preassembled metal oxide is generally low, hence,

compromising the capacity of the adsorbent. The coassembly of

a functional open framework metal oxide in the presence of

surfactants is an approach that can be realized in one pot and

has the potential to achieve higher loading of organic. Such

coassembly or direct synthesis has been undertaken and

extensively reported in the case of functionalized mesoporous

materials.²¹

Extraction chromatographic resins, or solvent impregnated

resins as they are sometimes also called, are similar to the

functionalized mesoporous materials in the sense that an extract

molecule (that is generally not miscible with water) is

incorporated within a porous host substrate.²²In this case

macroreticular polymeric resins have been the popular choice,

and the incorporation of the extractant molecules usually

occurs via impregnation utilizing a suitable solvent. Since the

extractant molecule is not covalently bound within the matrix

there is scope for it to be leached.²³Such extraction

chromatographic resins containing organophosphorus reagents

have shown utility for small scale analytical applications.²⁴⁻²⁶

Yet another class of materials that has seen signiﬁcant

interest in more recent times consists of the coordination

polymers that result from the coordination of metal cations by

multifunctional organic ligands to form compounds that extend

continuously in one, two, and three dimensions as in the case of

metal organic frameworks (MOFs).²⁷MOFs generally have

ordered three-dimensional porous structures, and while the vast

majority of reported coordination polymers and MOFs are

macroscopic crystalline materials, they can also be amor-

phous.²⁸The hexacyannoferrates can be regarded as traditional

members of the coordination polymer family, and their ion-

exchange properties are well-documented and have been

extensively explored. That the couplings of group IV metal

cations and organic moieties can yield polymer-like porous

materials was demonstrated by Clearﬁeld and others as early as

the mid-1990s.^29,30Nowadays the one-step hydrothermal

reaction between reactive metals or organometallic precursors

and complexing polyfunctional organic molecules with or

without the use of a template to form a coordination polymer

network is yielding an ever-expanding array of new and exciting

compounds.³¹If these coordination polymers consist of

crystalline open frameworks then they are known as MOFs.

Like mesoporous metal oxides generated through supra-

molecular self-assembly, MOFs can have enormous surface

areas. Their utility for the adsorption and separation of gases is

being studied intensively, and impressive capacities and

selectivities for diﬀerent gases have been demonstrated.

Given the aforementioned aﬃnity between phosphonate

ligands and trivalent lanthanides and actinides it is natural to

consider the use of coordination polymers involving organo-

phosphorus ligands possessing multiple binding sites. Indeed

group (IV) metal organophosphonates based on the α-

zirconium phosphate motif have been extensively studied for

many decades and have shown utility for the separation of

lanthanides and actinides.³²⁻³⁴Particularly attractive are the

compounds generated with Zr(IV), since they are highly

insoluble in both acidic and alkaline solutions. Of the thousands

of coordination polymers and MOFs that have been prepared

to date, only relatively few involve the coupling of Zr(IV) with

phosphonate ligands to give highly insoluble compounds.³⁵⁻³⁸

Insofar as the ability of such compounds to extract actinides is

concerned, such materials have only relatively recently begun to

be studied. For instance, Carboni et al. studied the adsorption

of UO₂on UiO MOFs at a pH of 2.5.³⁹

2+

We recently focused our attention⁴⁰on the development of

coordination polymers prepared through the coordination of

Zr(IV) to the trisphosphonic acid amino trismethylene

phosphonate (ATMP) under hydrothermal conditions. These

X-ray amorphous materials showed very high aﬃnities for the

trivalent lanthanides and interesting intralanthanide separation

factors. Intralanthanide separations are increasing in importance

owing to the desire to utilize various lanthanides in advanced

technologies.

Subsequently we reported on the preparation and extraction

properties of similarly amorphous compounds formed between

zirconium(IV) and [1-hydroxy-2-(1H-imidazol-1-yl)ethane-1,1-

diyl]bis(phosphonic acid) (zoledronic acid).⁴¹These zirco-

nium-zoledronate coordination polymers displayed surprisingly

similar selectivities to the zirconium-ATMP compounds despite

the inclusion of the imidazolate functionality. Additionally,

certain compositions were shown to have excellent capability

for separating the actinide thorium from complex mixtures of

lanthanides, actinides, and ﬁssion product elements particularly

when granulated with polyacrylonitrile. The quantitative

extraction of thorium from a complex mixture or lanthanide

elements could have important implications in rare earth

minerals processing as it can be an important radioactive

contaminant.⁴²

It was also demonstrated that the zirconium-zoledronate

coordination polymers, once loaded with an actinide, could also

easily be converted to phosphate ceramics with the NZP

structure that are known as highly durable host phases for

actinides. Such a cradle-to-grave strategy has been discussed as

long ago as the 1970s⁴³and has been applied in relation to the

use of crystalline silicotitanates, where it was demonstrated that

highly Cs-selective inorganic ion exchangers could be converted

directly into stable pollucite ceramics⁴⁴and mesoporous

zirconium titanates.¹⁴

In the present contribution we explored the possibility of

coupling zirconium(IV) to 1,3,5-benzene trisphosphonic acid

B

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Methyl 3,5-Bromobenzoate. ¹H NMR (400 MHz, CDCl₃):

δ(ppm) 8.12 (d, 2H_ar), 7.87 (t, 1H_ar), 3.95 (s, 3H, CH₃); ¹³C NMR

(100 MHz, CDCl₃): δ(ppm) 164.5 (CO), 138.3−123.1 (C_ar), 52.8

(CH₃).

(BTP) ligands and ligands in which the 1,3,5-trisubstituted

benzene ring carboxylic acid functions systematically replace

one (BDP) or two (BMP) of phosphonic acid groups. Such a

replacement in addition to structural nuances might be

expected to modulate the basicity of PO groups and the

acidity of P(O)OH protons as well as impose steric eﬀects such

as to inﬂuence metal ion extraction eﬃciency.²⁶The rigidity of

the 1,3,5-trisubstituted benzene could yield framework

structures that would have greater structural order as well as

higher thermal and chemical stability than achievable with more

ﬂexible ligands.

2.3. General Procedure for Arbusov Reaction. The Arbusov

reaction was performed using an adapted literature protocol.⁴⁵Brieﬂy,

a suspension of appropriate bromo derivatives (dimethyl 5-

bromoisophthalate; methyl 3,5-bromobenzoate, or 1,3,5-tribromoben-

zene) (8−20 mmol) in 1,3-diisopropylbenzene (20−50 mL) was

heated to 180 °C for 15 min under nitrogen allowing the solubilization

of the bromo derivatives. The solution was cooled to room

temperature, and nickel(II) bromide (2−5 mmol) was added under

nitrogen. The reaction mixture was heated to 180 °C, and triethyl

phosphite (30−90 mmol) was added dropwise over 8 h. The resulting

mixture was maintained overnight under reﬂux, under nitrogen. Excess

of triethyl phosphite and the 1,3-diisopropylbenzene was distilled oﬀ

under reduced pressure. The residues were then puriﬁed by ﬂash

chromatography on silica gel using CH₂Cl₂as eluent, followed by

CH₂Cl₂/CH₃OH (95/5).

2. EXPERIMENTAL SECTION

2.1. Synthesis of Benzene Phosphonates. 1,3,5-Trisubstituted

benzene ring with one phosphonic and two carboxylic acid groups

(BMP: Benzene MonoPhosphonic acid); two phosphonic and one

carboxylic acid groups (BDP: Benzene DiPhosphonic acid); or three

phosphonic acid group (BTP: Benzene TriPhosphonic acid) were

obtained by a catalytic Arbusov reaction from the corresponding

bromide compound and triethyl phosphite in the presence of

nickel(II) bromide followed by a hydrolysis reaction of the methyl

carboxylate and phosphonate groups. See reaction scheme in the

Supporting Information section.

Dimethyl 5-(Diethoxyphosphoryl)isophthalate. ¹H NMR (400

MHz, CDCl₃): δ(ppm) 8.77 (s, 1H_ar), 8.56−8.59(d, 2H_ar), 4.09 (m,

4H, CH₂O), 4.05 (s, 6H, CH₃O), 1.28 (t, 6H, CH₃); ¹³C NMR (100

MHz, CDCl₃): δ(ppm) 165.3 (CO), 136.8−129.4 (C_ar), 62.6

(CH₂O), 52.6 (CH₃O), 16.3(CH₃); ³¹P NMR (162 MHz, CDCl₃):

δ(ppm) 15.65.

Methyl 3,5-Bis(diethoxyphosphoryl)benzoate. ¹H NMR (400

MHz, CDCl₃): δ(ppm) 8.59 (d, 2H_ar), 8.39(M, 1H_ar), 4.14 (m, 8H,

CH₂O), 3.95 (s, 3H, CH₃O), 1.32 (t, 12H, CH₃); ¹³C NMR (100

MHz, CDCl₃): δ(ppm) 165.3 (CO), 138.7−129.3 (C_ar), 62.6

(CH₂O), 52.6 (CH₃O), 16.3(CH₃); ³¹P NMR (162 MHz, CDCl₃):

δ(ppm) 15.44.

1

Hexaethylbenzene-1,3,5-tris(phosphonate). H NMR (400 MHz,

CDCl₃): δ(ppm) 8.38(m, 3H_ar), 4.15 (m, 12H, CH₂O), 1.33 (t, 24H,

CH₃); ¹³C NMR (100 MHz, CDCl₃): δ(ppm) 138.3−129.7 (C_ar), 62.6

(CH₂O), 16.3(CH₃); ³¹P NMR (162 MHz, CDCl₃): δ(ppm) 15.32.

2.4. General Procedure for Hydrolysis Reaction. A suspension

of appropriate phosphonate derivatives (dimethyl 5-(diethoxyphos-

phoryl)isophthalate, methyl 3,5-bis(diethoxyphosphoryl)benzoate, or

hexaethylbenzene-1,3,5-tris(phosphonate) (15 mmol) in hydrochloric

acid solution (20%, 50 mL) were reﬂuxed for 12 h. The acidic solution

was evaporated dryness under reduced pressure. Then the residue was

dissolved in a minimum of distilled water, and freeze-drying aﬀorded

ﬁnal compounds as solid.

5-Phosphonoisophthalic Acid: Benzene Monophosphonic Acid).

¹H NMR (400 MHz, D₂O): δ(ppm) 8.43(s, 1H_ar), 8.35−8.31(d,

2H_ar); ¹³C NMR (100 MHz, D₂O): δ(ppm) 168.4 (CO), 135.8-

130.3(C_ar); ³¹P NMR (162 MHz, D₂O): δ(ppm) 11.88. HRMS (ESI-)

Calcd. Mass for C₈H₆O₇P: 244.9851, Found 244.9854.

5-Bromoisophthalic acid was treated by thionyl chloride and a

methanolic solution to provide dimethyl 5-bromoisophthalate, which

was then engaged in a catalytic Arbusov reaction. The resulting

dimethyl 5-(diethoxyphosphoryl)isophthalate compound was treated

by a hydrochloric acid solution (HCl 20%) providing the 5-

phosphonoisophthalic acid, which was called BMP.

3,5-Dibromobenzoic acid was treated by thionyl chloride and a

methanolic solution to provide methyl 3,5-dibromobenzoate, which

was then engaged in a catalytic Arbusov reaction. The resulting methyl

3,5-bis(diethoxyphosphoryl)benzoate compound is treated by a

hydrochloric acid solution (HCl 20%) providing the 3,5-diphospho-

nobenzoic acid, which was called BDP.

1,3,5-Tribromobenzene was engaged in a catalytic Arbusov reaction

providing the hexaethylbenzene-1,3,5-tris(phosphonate) compound,

which was treated by a hydrochloric acid solution (HCl 20%) yielding

the benzene-1,3,5-tri(phosphonic) acid, which was called BTP.

2.2. General Procedure for Benzoate Compounds. A solution

of 3,5-dibromobenzoic acid or 5-bromoisophthalic acid (30 mmol) in

SOCl₂(10−15 mL) was reﬂuxed for 5 h under nitrogen atmosphere.

Excess SOCl₂was removed under reduced pressure, and then the

residue was reﬂuxed in methanol (10−15 mL) for 12 h. The solvent

was evaporated, and the residue was ﬁltrated and washed by methanol

to aﬀord in a quantitative yield the ﬁnal compounds as white solids.

Dimethyl 5-Bromoisophthalate. ¹H NMR (400 MHz, CDCl₃):

δ(ppm) 8.4(s, 2H_ar), 8.18 (s, 2H_ar), 3.96 (s, 6H, CH₃); ¹³C NMR (100

MHz, CDCl₃): δ(ppm) 165.1 (CO), 136.3−122.9 (C_ar), 51.5

(CH₃).

1

3,5-Diphosphonobenzoic Acid: Benzene Diphosphonic Acid). H

NMR (400 MHz, D₂O): δ(ppm) 8.40(m, 2H_ar), 8.22(t, 1H_ar); ¹³C

NMR (100 MHz, D₂O): δ(ppm) 169.1 (CO), 136.6−130.2(C_ar);

³¹P NMR (162 MHz, D₂O): δ(ppm) 12.27. HRMS (ESI-) Calcd.

Mass for C₇H₇O₈P₂: 280.16, Found 280.9619.

Benzene-1,3,5-tri(phosphonic acid): Benzene Triphosphonic

Acid). ¹H NMR (400 MHz, D₂O): δ(ppm) 8.11(m, 3H_ar); ¹³C

NMR (100 MHz, D₂O): δ(ppm) 135.2- 132.2(C_ar); ³¹P NMR (162

MHz, D₂O): δ(ppm) 12.94. HRMS (ESI-) Calcd. Mass for C₆H₈O₉P₃:

316.9381, Found 316.9363.

2.5. Synthesis of Zr(IV)-Benzene Phosphonate Compounds.

The coordination polymers were prepared by ﬁrst dissolving the

benzene phosphonate (BMP, BDP, or BTP) in water using a magnetic

stirrer. In some cases 4 M HNO₃or hydroﬂuoric acid was added to the

stirred BMP, BDP, or BTP solution. Once the dissolution was

complete, an appropriate volume of 0.5 M ZrOCl₂·8H₂O solution was

added while stirring thoroughly. The mixture was then loaded into a

polytetraﬂuoroethylene-lined autoclave, which was heated in a fan-

forced oven up to ∼160 °C for several days. At the end of the reaction

the autoclave was quenched, and the product was isolated by ﬁltration

followed by washing with milli-Q water. In addition to using water as

the solvent, numerous materials with variable phosphorus mole

C

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

fractions were also prepared in tetrahydrofuran and EtOH. These

materials are being referred to as generation I (Gen I). A series of

compounds was synthesized hydrothermally using a ﬁxed P/(Zr + P)

mole ratio but with variable amounts of HF. The HF was used to

mineralize the Zr and ensure it remains in solution during the

hydrothermal reaction. These materials with variable F/Zr molar ratio

are called generation II (Gen II) and generation III (Gen III) for the

more crystalline materials.

Granulation of the Zr phosphonates was achieved as previously

described.¹⁶Brieﬂy, to a 6 wt % solution of polyacrylonitrile (PAN) in

dimethyl sulfoxide (DMSO) was added suﬃcient ZrBTP-x to give a 60

wt % loading of the compound in PAN. The solid was dispersed in the

PAN-DMSO solution by agitation and sonication until a smooth slurry

was obtained. Beads were generated by adding droplets of the solution

to a stirred water bath.

2.6. X-ray Powder Diﬀraction. X-ray diﬀraction (XRD) measure-

ments were performed on a Panalytical Empyrian diﬀractometer using

Cu Kα radiation and PIXCel^3Ddetector in addition to automatic

divergence slit, beta ﬁlter, and 0.04 rad Soller slits on the incident and

diﬀracted beam side. Variable-temperature measurements were

performed using an Anton Paar HTK16N high-temperature strip

heater chamber.

in milligrams per gram of adsorbent, and k is the rate constant (g/mg

min).

Quantiﬁcation of the elements in solution was achieved using a

Bruker S2 Picofox total external reﬂection X-ray ﬂuorescence

spectrometer (TXRF). To 990 μL of each solution to be analyzed

was added 10 μL of an internal standard (Ga), and then 10 μL of this

solution was pipetted onto a quartz analysis plate. Uncertainties

relating to the TXRF analyses for the heavy elements concerned were

determined by repeated measurement of diﬀerent samples of the same

concentration. These uncertainties were estimated to be of the order

of 5%.

3. RESULTS

3.1. X-ray Powder Diﬀraction. In Figure 1 are shown the

X-ray powder diﬀraction patterns of a series of ZrBTP-x

2.7. Solid-State P-31 Magic-Angle Spinning Nuclear Mag-

netic Resonance. The ³¹P CP MAS NMR spectra were recorded on

a Bruker 400 ultrashield VS spectrometer with a 12 kHz MAS spinning

rate using Bruker MAS probes (4 mm rotors). The typical contact

time was 2−3 ms, and the typical recycle delay was 5−30 s. The ³¹P

MAS NMR data were acquired using a single-pulse (Bloch decay)

1

experiment with and without strong H decoupling (B_1H≈ 80 kHz)

during acquisition. The measured ³¹P chemical shifts were indirectly

referenced to the primary ³¹P reference of 85% H₃PO₄(δ_iso= 0.0

ppm) via a secondary solid reference of ammonium dihydrogen

phosphate (NH₄)H₂PO₄.

2.8. Thermal Analysis. Thermogravimetric analysis-diﬀerential

thermal analysis-mass spectrometry (TGA-DTA-MS) analyses were

performed with a Setaram Setsys Evolution 16 Instrument under

ﬂowing air heated at 5 °C/min, which was coupled to a Hiden

QGA300 Instrument for gas analysis. TGA-diﬀerential scanning

caloriimetry (DSC) analyses were accomplished on a TA Instruments

Q600 instrument in ﬂowing air.

2.9. Nitrogen Porosimetry. Measurements of nitrogen adsorp-

tion−desorption isotherms were performed on a Micromeritics ASAP

2420 instrument.

Figure 1. X-ray powder diﬀraction patterns of Zr(IV)-BTP-x materials

(a) ZrBTP-0.7 precipitate, (b−e) ZrBTP-0.6, ZrBTP-0.7, ZrBTP-0.8,

and ZrBTP-0.9, (f) ZrO₂precipitate from ZrOCl₂·8H₂O.

2.10. Measurement of Sorption Properties. Adsorption

properties were measured using mixed cation solutions in 0.10 M

HNO₃. Solution A contained Cs⁺, Sr²⁺, La³⁺, Ce³⁺, Pr³⁺, Nd³⁺, Gd³⁺,

Dy³⁺, Ho³⁺, Co²⁺, and Th⁴⁺at a concentration of ∼0.10 mmol/L (∼20

mg/L) for a total ionic strength of ∼1 mmol/L. Solution B was

prepared by a 10-fold dilution of solution A so that the total ionic

strength was 0.10 mmol/L and individual cation concentrations were

∼1−2 mg/L.

materials synthesized by combining varying proportions of

ZrOCl₂·8H₂O and BTP in water such that a precipitate formed

spontaneously. The spontaneous formation of a precipitate

attests to the very low solubility of the Zr(IV)-BTP

compounds. The precipitates were then reacted hydrothermally

(160−170 °C). The XRD pattern of the precipitate obtained

from composition ZrBTP-0.7 (Figure 1a) showed that it was

not totally X-ray amorphous possessing two very broad

reﬂections at 11.5 and 22.4° 2θ. Scherrer analysis indicates

that the particle size for these materials was ∼2 nm. XRD

patterns of the hydrothermally treated suspensions having

diﬀerent values of x are shown in Figure 1b−e. Compared to

the non-hydrothermally treated precipitate, the hydrothermally

treated ZrBTP-0.7 phase was clearly more crystalline (Figure

1b). As the mole fraction of phosphorus increased (x) the

crystallinity seemed to improve, with the ZrBTP-0.8 phase

appearing to be the most crystalline phase possessing a series of

broad reﬂections at 11.5, 20.0, and 22.4, and 28.1° 2θ. These

materials are therefore clearly more crystalline than the

previously reported Zr-ATMP⁴⁰and Zr-zoledronate⁴¹coordi-

nation polymers that showed no X-ray reﬂections. Also shown

The distribution coeﬃcient (K_d), which is simply a mass-weighted

partition coeﬃcient, was calculated according to the following

equation:

(C_i− C_f)

V

M

K_d=

×

C_f

Here C_iand C_fare the initial and ﬁnal concentrations, respectively, M

is the mass of adsorbent used in grams, and V is the volume of solution

treated in milliliters. Unless otherwise stated the volume-to-mass (mL/

g) ratio used for the experiments was set at 120. Kinetics data were

treated using the pseudo-second-order rate law as described by Ho et

al.^46,47

dq_t

= k(q_e− q_t)²

dt

Here q_tis the amount of target ion sorbed in milligrams per gram of

adsorbent at time t, q_eis the amount of metal ion sorbed at equilibrium

D

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

in Figure 1f is the pattern of a precipitate obtained by the direct

precipitation of hydrated zirconia through the addition of

NaOH solution to a solution of ZrOCl₂·8H₂O following by the

same hydrothermal treatment. The XRD pattern indicates the

formation of monoclinic (m) and tetragonal zirconia (t) of very

small particle size. The particle sizes for these phases as

determined by Scherrer analysis were between 10 and 15 nm.

Given that the main reﬂections of the either monoclinic or

tetragonal zirconia phases present in the precipitate at 24.1 (m),

28.0 (m) 30.2 (t), 31.3 (m), 34.2 (m), 50.3 (t), and 59.9° (t) 2θ

could not be clearly discerned in the pattern of any of the

ZrBTP-x phases it would appear that zirconia is not present, at

least in signiﬁcant concentrations. Also arguing against the

existence of ZrO₂-like Zr−O−Zr bonding in these compounds

is the fact that the bands of ZrO₂(530, 620, and 725 cm⁻¹) are

not evident in the Fourier transform infrared (FTIR) spectra of

this series of compounds (Figure S1 of Supporting Information

section).⁴⁸

It was observed in additional experiments in which HNO₃

was added to the precipitate prior to hydrothermal treatment

that somewhat more crystalline materials could be produced

(see Figure S2 of Supporting Information section) by

increasing the acidity of the suspension. It was also observed

(Figure S3 Supporting Information section) that a change of

solvent from water to ethanol and other solvents preserved the

essential features of the XRD patterns for all compositions,

although increased broadening was observed in some cases.

Cation binding experiments to be described later in this

paper indicated that the best-performing materials were the

ZrBTP-0.8 and ZrBTP-0.9 phases, and therefore initial eﬀorts

focused on optimizing these materials.

With the aim of reducing the possibility of precipitating

zirconia and increase crystallinity, a series of syntheses were

undertaken in which increasing amounts of HF were added to

the ZrBTP-0.8 precursor composition to mineralize any

zirconium oxyhydroxide that might form under the hydro-

thermal conditions thus maintaining Zr(IV) in solution. The

XRD patterns of these materials are shown in Figure 2. It can

be observed that as the F/Zr molar ratio in the hydrothermal

preparations was increased the products became increasingly

crystalline. Presumably this is a result of the mineralization of

oxyhyroxide zirconium species in solution. What is apparent

from this series of XRD patterns is that regardless of the

amount of HF used, the phase or phases being generated were

always similar diﬀering only in the degree of order. This would

seem to include even the precipitate that was directly generated

by addition of BTP to ZrOCl₂solution (Figure 1a).

For the remainder of this paper, and as a matter of

convenience, materials prepared without the use of HF and

giving XRD patterns such as in Figure 2a are referred to as Gen

I, those prepared with the use of small amounts of HF and

giving XRD patterns such as those of Figure 2b,c are referred to

as Gen II, and those giving XRD patterns of crystalline

materials such as in Figure 2d,e as Gen III.

It is also very important to note that in none of these

patterns could ZrO₂phases be identiﬁed. In fact, close

inspection of the XRD patterns of Figure 2b−e suggests the

presence of at least two phases characterized by both relatively

broad (indicated by ↓) and sharp reﬂections.

Figure 2. X-ray powder diﬀraction patterns of materials synthesized as

a function of the F/Zr molar ratio. (a) ZrBTP-0.8−0, (b) ZrBTP-0.8−

9.4, (c) ZrBTP-0.8−18.7, (d) ZrBTP-0.8−24.0, (e) ZrBTP-0.8−38.8.

Scherrer analysis of the sharpest patterns indicated a particle

size in the range of 100−300 nm for all phases. This was more

than an order of magnitude larger than was obtained for the

materials prepared in the absence of mineralizing agent.

3.2. Nitrogen Porosimetry. The adsorption−desorption

isotherms for the series of the Gen I ZrBTP-x samples prepared

without the use of HF and having the diﬀraction patterns of

Figure 1 are shown in Figure 3a. For this series of samples all

isotherms were essentially type II that are indicative of

nonporous or macroporous materials with wide pore-size

distributions. The Zr-BTP-0.7 phase had the highest BET

surface area (500 m²/g) and Zr-BTP-0.6 and ZrBTP-0.8

present similar surface area (279 and 247 m²/g, respectively).

The ZrBTP-0 sample, or in other words the ZrO₂precipitate,

had a relatively low BET surface area (80 m²/g). As the value of

P/(Zr + P) increased so too did the extraction capacity of the

material (see later).

The adsorption−desorption isotherms for the series of

ZrBTP-0.8-f samples synthesized using increasing amounts of

HF corresponding to the diﬀraction patterns of Figure 2 are

shown in Figure 3b. It is apparent that for ZrBTP-0.8−0,

ZrBTP-0.8−24, and ZrBTP-0.8−38.8 the isotherms are clearly

of Type II, while for ZrBTP-0.8−9.4 and ZrBTP-0.8−18.7 the

isotherms are distinctly of Type IV with a type H2 hysteresis

loop. The surface areas measured by nitrogen porosimetry

follow a similar pattern as a function of HF content as

determined by small-angle X-ray scattering measurements (see

Figure S4 of Supporting Information section).

From Figure 4b it can be observed that as the amount of

ﬂuoride included in the synthesis increased, the samples

became more crystalline, and the surface areas decreased.

Type IV isotherms were observed for Zr-BTP0.8−9.4 and Zr-

BTP0.8−18.7 and signal that a signiﬁcant change in porous

texture occurred in these samples (Figure 3b).

Attempts to index the series of sharp reﬂections were not

successful mainly because a unique unit cell could not be

encountered. In part this could be because it is likely that all

reﬂections are not due to a single phase.

For the ZrBTP-0.8-f compositions, increasing Zr/F and

therefore crystallinity trends with decreasing surface area as

E

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 3. Nitrogen adsorption−desorption isotherms of (a) ZrBTP-x samples with variable x and prepared without HF and (b) ZrBTP-0.8-HF using

increasing amounts of HF in the range from 0 to 39. Closed symbols correspond to the adsorption branch; open symbols indicate the desorption

branch.

Figure 4. BET surface areas of (a) ZrBTP-x series prepared without HF and (b) ZrBTP-0.8-HF using increasing amounts of HF.

Figure 5. Backscattering electron images of crystalline ZrBTP-0.8−24 at low (a) and high (b) magniﬁcation.

expected. This in turn resulted in reduced adsorption capacity

for cation extraction as also expected.

3.3. Scanning Electron Microscopy. Scanning electron

microscopy could not be used to observe images of the particles

of the less crystalline Gen I samples ZrBTP-0.8−0, ZrBTP-0.8−

9.4 and ZrBTP-0.8−18.7 due to the small grain size. In this case

TEM was found to be more appropriate.

For more crystalline samples of ZrBTP-0.8 prepared using

HF (Figure 2e) Scherrer analysis yielded particle diameters in

the 20−50 and 150−200 nm ranges. This appears consistent

with what was observed in SEM images where spherical

particles can be observed at high magniﬁcation (Figure 5b). For

less crystalline samples clear particle morphologies could not be

observed using SEM.

For the more crystalline samples such as ZrBTP-0.8−24

having well-deﬁned diﬀraction patterns and Type II nitrogen

adsorption isotherms, discrete roughly spherical particles with

diameters of the order of 50 nm were observed (Figure 5).

Particles of this dimension should give surface areas of 30 m²/g

assuming a density of ∼4 cm³/g. This value of area is close to

what is actually measured. Such particle diameters were also

roughly in line with what was measured by the Scherrer

analysis. In other words, to ﬁrst approximation, the surface area

of these materials is due to external particle surfaces.

In some cases elongated nanoparticles of approximately the

same dimensions were observed. This particle size was entirely

consistent with the aforementioned Scherrer analysis. At the

length scale being probed no other morphologies were evident

suggesting that the sample could be monophasic. The average

of numerous EDAX analyses over wide and narrow areas of the

sample gave an average phosphorus mole ratio (P/(Zr + P)) of

0.82 for the ZrBTP-0.8 sample of Figure 1a. This value is very

close to the value expected based on initial reactant mole ratios

conﬁrming that all the phosphorus that was included in the

F

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 6. TEM images of ZrBTP-0.8 (a) GEN I and (b) GEN III (ZrBTP-0.8−23.8).

Figure 7. ³¹P MAS NMR of ZrBTP-x and ZrBDP-x samples. The diﬀerent resonances are labeled 1−4 for convenience.

precursor solution was incorporated into the ﬁnal product. This

was also the case for sample of Figure 1b.

those published by Brunet et al.⁴⁹for microporous materials

formed from aluminum and 4,4-biphenyldiphosphonic

(BPDP). These aromatic polyphosphonates bear some

similarities to the BTP except that the phosphonate groups

are on adjacent carbons rather than alternate carbons as in

BTP. In the work by Brunet et al., no speciﬁc assignments were

made in the ³¹P MAS NMR, which like the present materials

also contained four or more resonances.

The MAS NMR spectrum of solid BTP acid shows a group

of six relatively sharp resonances having isotropic shifts ranging

from 18.7 to 10.7 ppm (Figure 7a). The spectrum is indicative

of six chemically distinct P sites that are likely distinguished in

terms of the degree of protonation.

When the phosphonate ligands complex Zr(IV) as in the

precipitation of ZrBTP-0.7 to ZrBTP-0.9 and ZrBDP-0.7 to

ZrBDP-0.9 (Figure 7) four discrete broad isotropic shifts (1−4)

are observed. These four resonances are shifted signiﬁcantly

upﬁeld from that of the free acids. The chemical shifts span a

range from −4 ppm for resonance one to 14 ppm for resonance

four. There was no dependence of the spectra on the nature of

the solvent used for the syntheses. The reduction in the

number of resonances, the broadening, and the upﬁeld shifts

are indicative of signiﬁcant changes in the phosphonate

bonding as might be expected on complexation with Zr(IV).

Similar general trends, including the chemical shift spread, are

consistent with what was observed in the Zr-zoledronate⁴¹and

Zr-ATMP⁴⁰coordination polymers reported recently. The

increased broadening of the spectra of the ZrBTP-x samples

3.4. Transmission Electron Microscopy Imaging.

Figure 6 shows the TEM images of Gen I and III materials.

For the Gen I materials prepared in the absence of HF (Figure

6a) the grain size was very ﬁne and only just observable. It is

estimated that the particles were of the order of 10 nm in

diameter. Such particles would be expected to yield surface

areas of at least 150 m²/g, which were observed (247m²/g for

Zr-BTP-0.8−0).

For the Gen III materials, two particle morphologies were

observed in the TEM images including round particles and

elongated particles (Figure 6b). These particles appeared to

have relatively narrow size distributions. Analysis of the

occasional round particle that could be found isolated from

the other particles showed them to be quite crystalline.

3.5. Solid-State Phosphorus-31 Magic-Angle Spinning

Nuclear Magnetic Resonance. Phosphorus-31 is an S = 1/2

nucleus, has a high isotopic abundance, high sensitivity, and

shows a large chemical shift range. Despite the fact that

phosphorus is an important element in many biomaterials and

that it is increasingly important in the context of the design and

preparation of open framework materials, there exists a

surprising paucity of systematic solid-state ³¹P NMR work on

phosphonate compounds.

Single-pulse solid-state ³¹P MAS NMR spectra were recorded

of the diﬀerent ZrBTP-x and ZrBDP-x samples along with the

free acids (Figure 7). These spectra are superﬁcially similar to

G

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 8. Intensities of the various ³¹P resonances as a function of phosphorus content for ZrBTP-x (left) and ZrBDP-x (right). The lines are only

intended to serve as a guide.

relative to BTP itself is consistent with a loss of order. The

spectra of the Zr-benzene phosphonates, however, diﬀer in two

important respects relative to the other mentioned systems.

First the diﬀerence between the center of mass of the spectra of

the free acids and the Zr-zoledronate and Zr-ATMP polymers

was minimal, whereas the center of mass of the Zr-benzene

phosphonates was shifted signiﬁcantly upﬁeld relative to the

free acids. Second, in the Zr-zoledronate and Zr-ATMP systems

new upﬁeld resonances were observed, as the phosphorus

content increased, that were attributed to the generation of

relatively uncondensed and deprotonated phosphonate groups.

These additional upﬁeld shifts were assigned in a manner

consistent with available information on ³¹P MAS NMR,

especially the work of Josse et al.⁵⁰

advanced multipulse sequences that can provide information on

the phosphonate connectivity.⁵¹

3.6. Thermal Analysis. The samples prepared in this study

were subjected to thermal analysis primarily to assess their

stability. The TGA-DTA of the ZrBTP-0.8-H₂O as a

representative of the family of ZrBTP-0.8 prepared in water

solvent is shown in Figure 9a. The weight loss that commences

at ∼50 °C and ends at ∼150 °C was due to the loss of adsorbed

water as shown by the real-time mass spectrum recorded of the

products released at the reactor outﬂow during the heating

ramp. Following the loss of the weakly bound water a gradual

weight loss then occurs between ∼455 and 608 °C. Thereafter

a steep weight loss occurs that is due to decomposition of the

BTP with the only decomposition product being CO₂. No

phosphorus was observed at this point in the mass spectrum.

In contrast to the ZrBTP-x phases, the ZrBDP-x phases

displayed a more progressive weight loss over the temperature

range from 450 to 850 °C, at which point a steep loss was

observed (Figure 9c). The mass spectral measurements

demonstrated that the gradual weight loss was due to the

breakdown of organic groups with emission of CO and CO₂.

Once again, no phosphorus was detected at any temperature

suggesting that all phosphorus is retained in the product until at

least 1200 °C. The fact that CO is generated in addition to CO₂

may be because carboxylate groups are being broken down in

addition to the aromatic ring, which seems to only produce

CO₂.

The striking diﬀerence between the ³¹P MAS NMR of the

ZrBTP-x and Zr-zoledronate system is that in the ZrBTP-x

system there is generally little or no change in either the

number or the frequency of the four isotropic shifts observed as

the mole ratio of phosphorus is increased. What changes as a

function of phosphorus content is the relative intensities of the

various resonances.

The ³¹P MAS NMR spectra of the ZrBDP-x (and also the

ZrBMP-x see Figure S5 of Supporting Information section)

materials are remarkably similar to those of the ZrBTP-x

samples in that they contain four resonances at similar

frequencies with all being displaced upﬁeld relative to the

BDP by approximately the same amount. Although the spectra

of all the ZrBDP-x samples contain four resonances for the

ZrBDP-0.7 sample resonance 4 is barely discernible and was

not ﬁtted. As in the case of the ZrBTP-x system the relative

intensities of the resonances varies as a function of phosphorus

content. Spectral ﬁtting facilitated the measurement of the

relative intensities of these resonances that are plotted in Figure

8.

The plots of Figure 8 show that the variation in relative

intensities of resonances 2−4 as a function of phosphorus

content is similar for both BTP- and BDP-derived materials, all

increasing with phosphorus content. However, resonance one

decreases for ZrBTP, whereas it increases for ZrBDP as a

function of x. It would however be imprudent to place too great

a signiﬁcance on this relatively minor diﬀerence between the

three systems. In fact what seems remarkable is the similarity of

the diﬀerent systems and the invariance in the trends in

intensities of the various resonances as a function of x. It is

suspected that this has to do with the rigidity of the ligands and

the distance between phosphonate and carboxylate groups. To

advance beyond this interpretation would require the use of

Gomez-Alcantara et al.⁵²reported the preparation of lamellar

microporous aluminum phosphonates from 4-(4-

phosphonophenoxy)phenyl phosphonic acid (PPPA) and

observed decomposition at temperatures between 400 and

600 °C. The diﬀerence in total weight loss between ZrBTP-0.8

and ZrBDP-0.8 with similar phosphorus mole fraction (P/(Zr +

P) is really due to the increasing organic content of the latter

sample for a given total mass.

3.7. In Situ Variable-Temperature X-ray Diﬀraction. In

situ variable-temperature XRD was undertaken in air of ZrBTP-

0.8 in an attempt to shed light on the mechanism of its thermal

decomposition (Figure 10) and to diﬀerentiate possible phases

in terms of their thermal stabilities.

The pattern at 200 °C is identical to that recorded at 25 °C

and remains relatively unchanged until at least 500 °C. What is

striking about this series of patterns is the disappearance at 700

°C of the reﬂections marked by the heart-shaped symbol that

were observed in the range of 25−600 °C to leave a reduced set

of sharp reﬂections marked by ⧫. This series of sharp

reﬂections could be well-indexed on a face-centered cubic

H

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 10. In situ variable-temperature X-ray powder diﬀraction

patterns of ZrBTP-0.8.

As the temperature was increased beyond 800 °C a phase

transition occurred with coalescence of all phases into a single

cubic (Pa3) pyrophosphate structure (ZrP O ). In fact the

̅

2

7

pattern at 900 °C appears relatively uncontaminated, and

successful Rietveld analysis could be undertaken giving a unit

cell parameter of 8.2849 Å (see Figure S6 of Supporting

Information section).

3.8. Cation Binding. Screening of the poorly crystalline

Gen I ZrBTP-x extractants with diﬀerent mole ratios of

phosphorus was initially performed using the high ionic

strength mixed cation solution (1 mmol/L) with nominal

concentration of individual cations of 20 mg/L in 0.1 M HNO₃

(solution A), and the data are shown in Figure 11. The use of

Figure 11. Extraction of cations of diﬀerent ionic radius by poorly

crystalline ZrBTP-0.8, ZrBTP-0.7, ZrBTP-0.6, ZrBTP-0.5, prepared in

the absence of HF from mixed cation solutions with each cation at

nominal 20 mg/L in 0.10 M HNO₃(Solution A). The V/M ratio was

100.

mixed cation solutions was justiﬁed on the basis that

applications of the materials at the back-end of the present or

future nuclear fuel cycles or rare-earth mineral processing will

always encounter mixtures of all LNs. The data are ordered

according to decreasing ionic radii except for the case of Th⁴⁺(r

= 1.13 Å), which is placed at the end of the horizontal axis for

clarity. The data show that, for ZrBTP-0.8 and ZrBTP-0.7, the

Figure 9. Thermal analysis in air and mass spectra, respectively, of (a,

b) ZrBTP-0.8, (c, d) ZrBDP-0.8, (e, f) ZrBMP-0.8.

cell. By deduction it appears that the patterns indicated by both

the heart-shaped symbol and ⧫ are from discrete relatively

crystalline phases. In other words, the synthesized ZrBTP-x

materials contain at least two relatively crystalline phases. Also

present in the low-temperature patterns are a series of broader

reﬂections that could be from a third phase.

order of increasing selectivity is Cs⁺< Sr²⁺= Co²⁺< LN³⁺

<

Th⁴⁺or in other words the selectivity depends to ﬁrst order on

cation valence and cation polarizability as in a typical

I

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Hofmeister series.⁵³As the mole fraction of phosphorus in the

adsorbents decreased (ZrBTP-0.6 and ZrBTP-0.5) so too did

the overall K_dvalues. This suggests that the cation binding

capacity in these materials is driven principally by the

availability of phosphonate groups. To test this hypothesis it

is expected that the FTIR of lanthanide- and thorium-loaded

extractants would show shifts in the vibrational spectrum of the

phosphonate groups as was observed for the Zr(IV)-

zoledronate system.⁴¹

the materials is essentially determined by the phosphonate

binding groups.

For low ion strength solutions (total cation concentration of

∼0.1 mmol/L) in which individual cation concentrations were

between 1 and 2 mg/L (Figure 13), the dependence of K_don

Experiments were conducted to ascertain the extraction

eﬃciency of Zr(IV)-polyphosphonates prepared from the

diﬀerent polyphosphonate ligands in the absence of HF (Gen

I materials). In Figure 12a are shown the distribution

Figure 13. Extraction of cations of diﬀerent ionic radius by Gen I

ZrBTP-0.8, ZrBDP-0.8, and ZrBMP-0.8 from low ionic strength mixed

cation solutions with each cation at nominal 2 mg/L in 0.10 M HNO₃

(solution B). Elements are ordered according to decreasing ionic radii

except for Th.

cation radius for Gen I materials was similar as obtained using

the higher ion strength solutions. That is, strong sorption of

trivalent lanthanides with much weaker sorption of monovalent

(Cs⁺) and divalent cations (Sr²⁺and Co²⁺) obtained.

Unlike the situation of extraction from the high ionic

strength solutions (solution A), extraction of all lanthanides

from the low ion strength solution (solution B) was

quantitative for ZrBTP-0.8, ZrBDP-0.8, and ZrBMP-0.8. This

diﬀerence in lanthanide extraction on total cation concentration

is therefore likely to be due to the reduced availability of free

phosphonic acid groups in the latter material despite the fact

that the overall phosphorus mole fraction was constant.

To evaluate the extraction eﬃciency of the Gen II (ZrBTP-

0.8−9.4, ZrBTP-0.8−24) and Gen III (ZrBTP-0.8−38)

materials shown in the XRD patterns of Figure 2, batch

contact experiments were conducted using the low ion strength

solutions. The data of Figure 14 show that all materials

extracted lanthanides more or less quantitatively across the

series. Another striking feature of these data is that for these

Gen II and III materials there is virtually no adsorption of Cs⁺

and very little adsorption of Sr²⁺.

It is well-appreciated that at very low ionic strengths, K_dis

close to being a true equilibrium concentration between cations

in solution and binding sites on the adsorbent surface. In this

study the lower concentration limit was essentially set by the

detection limit of the measurement technique (TXRF). As

concentration is increased additional equilibria come into play

between lanthanides and availability of the diﬀerent surface

binding sites with possibly diﬀerent binding energies. This is

manifest by the dependence of K_don total cation concentration

in solution of the survey measurements.

Figure 12. Extraction of cations of diﬀerent ionic radius by GEN I (a)

ZrBTP-8, ZrBDP-0.8, and ZrBMP-0.8 (b) ZrBTP-0.6, ZrBDP-0.6, and

ZrBMP-0.6 from mixed cation solutions with each cation at nominal

20 mg/L in 0.10 M HNO₃(solution A). Elements are ordered

according to decreasing ionic radii except for Th.

coeﬃcients measured for ZrBTP-0.8, ZrBDP-0.8, and

ZrBMP-0.8 prepared in the absence of HF. In general the

extraction eﬃciency (K_d) measured using solution A was

reduced as the number of phosphonic acid groups per benzene

unit was reduced. Because these experiments were conducted

with mixed cation solutions having relatively high ion strength

with nominal individual cation concentrations of 20 mg/L and

at ﬁxed volume-to-mass ratio, the reduction in K_dcould reﬂect

the reduced capacity of the material as the contribution of

organic to the total mass of each sample increased. Put another

way, although ZrBTP-0.8, ZrBDP-0.8, and ZrBMP-0.8 all have

the same overall phosphorus molar ratio (P/(Zr + P)) this

series of compounds has decreasing numbers of phosphonic

acid groups per unit mass, and therefore reduced capacity is not

unexpected. That the overall dependence of K_dvalues on ionic

radii was relatively constant irrespective of the substitution of

carboxylate groups for phosphorus suggests that the behavior of

In an attempt to determine the total capacity of the most

eﬀective Gen I ZrBTP-0.8 materials we ﬁrst measured the

Ce(III) adsorption isotherm (Figure 15) in 0.10 M HNO₃. The

isotherm was of L-type,⁵⁴and adsorption did not reach a

plateau even up to the highest equilibrium cation concen-

trations used. The isotherm is ﬁtted relatively well by Toth,

J

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

measured for Gd³⁺and Ho³⁺as individual cations under similar

experimental conditions (25 °C and 0.10 M HNO₃). For the

Gen I ZrBTP-0.8 material, both isotherms show a similar

appearance at low equilibrium cation concentrations (Figure

15b). However, unlike the case of Ce³⁺adsorption on Gen I

materials, a deﬁnite plateau is reached at ∼100 mg/g adsorbent

for these materials at equilibrium cation concentrations

between 50 and 100 mg/L for both Gd³⁺and Ho³⁺. The

data for the Gd³⁺shows in addition a tendency for a second

upswing greater than 100 mg/L suggesting perhaps that a

contribution from a second phase comes into play at higher

cation concentrations. In the case of the Gen III materials a

marked reduction in uptake is observed up to equilibrium

cation concentrations of 150 mg/L followed by what appears to

be an upswing.

Because the applications being considered here will

ultimately involve complex acidic mixtures of cations, and

recognizing that the selectivity is evaluated through measure-

ment of the pseudoequilibrium constant K_dand that this is a

function of the equilibrium concentration/ionic strength

among other things, we undertook to measure adsorption

isotherms of the individual cations from the complex mixtures

(Figure 16).

Figure 14. Distribution coeﬃcients for cations of diﬀerent ionic radius

by Gen II (ZrBTP-0.8−9.4, ZrBTP-0.8−24) and Gen III crystalline

Zr(IV) organophosphonates (ZrBTP-0.8−38 from), and mixed cation

solutions with each cation at nominal 2 mg/L in 0.10 M HNO₃

(solution B). Elements are ordered according to decreasing ionic

radius except for Th.

Figure 16. Multiequilibrium adsorption isotherms for ZrBTP-0.8 for

mixed cation solutions with each cation at nominal concentration 20

mg/L in 0.10 M HNO₃.

At low equilibrium concentrations strong uptake is observed

from most cations in accordance with previously described

measurements of capacity. However, as the equilibrium

concentration for each cation is increased above ∼1 mg/L it

is observed that Dy³⁺and Ho³⁺are particularly strongly

extracted compared to the other lanthanides. In fact, thorium is

quantitatively extracted at all equilibrium concentrations being

considered. What these data show is that since the capacity for

Th⁴⁺, Dy³⁺, and Ho³⁺is signiﬁcantly higher than for the other

species the adsorbent would be progressively enriched in these

elements. This paves the way for the eﬃcient extraction of

these elements from complex solutions. It should be remarked

that Cs⁺, Sr²⁺, and Co²⁺were all very poorly extracted at all

equilibrium concentrations.

Figure 15. Adsorption isotherm in 0.10 M HNO₃for (a) Ce³⁺on

poorly crystalline Gen I ZrBTP-0.8 material and (b) Gd³⁺and Ho³⁺on

Gen I and Gen III materials.

The extraction eﬃciency of both the ZrBTP-0.8 and ZrBTP-

0.8 granulated with PAN in 4 M HNO₃was measured in

solutions of increasing HNO₃concentrations. The data

obtained in 4 M HNO₃solutions are shown in Figure 17a.

Even in concentrated nitric acid solution the present materials

are capable of signiﬁcant extraction of lanthanides with respect

to Co²⁺, which is only very weakly extracted. Moreover, within

experimental uncertainty there is a smooth increase in

Langmuir, and Freundlich models. The value taken as the

capacity of this material for Ce(III) depends on the model that

is considered best. Clearly in the given situation the decision is

somewhat arbitrary, although it appears that the capacity is

somewhere greater than 100 mg/g.

To compare the performance of the poorly crystalline Gen I

and crystalline Gen III materials, adsorption isotherms were

K

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

Figure 17. Extraction parameters for ZrBTP-0.8 and granulated PAN-ZrBTP-0.8 measured in 4 M HNO₃and low ionic strength solutions (0.2 mM)

displayed in terms of (a) percentage extraction, (b) K_d, and (c) separation factor.

extraction eﬃciency as a function of lanthanide atomic number.

For a solid extractant the ratio of the K_dvalues can be

considered the distribution ratio or separation factor S. The

data of Figure 17c shows that there is a steady increase in S

from ∼1 to 10 as a function of increasing lanthanide atomic

number. Under the given conditions therefore the separation

factor between heavy and light lanthanides is easily suﬃcient for

a plausible light-heavy lanthanide separation and probably also

an eﬃcient intralanthanide separation.

The separation factor between Th⁴⁺and the trivalent LNs is

particularly large suggesting that Th⁴⁺could easily be separated

from a very complex mixture. The PAN-granulated ZrBTP-0.8

system would therefore seem to oﬀer intralanthanide separation

comparable to those achievable using alkaline diphosphonic

acid extractants only with the associated advantages of solid-

phase extractant.⁵⁵

3.9. Thermal Transformation of Cation-Loaded Ad-

sorbents. To address the possibility of disposition of the

radiolanthanide and/or actinide-loaded zirconium phospho-

nates, a range of the Gen I and II materials were saturated with

Th⁴⁺and Ce³⁺. Samples were also prepared by saturation with

mixtures of lanthanides, and these were then heated to diﬀerent

temperatures both statically and dynamically in the heating

chamber of the diﬀractometer.

Figure 18. Rietveld analysis of ex situ X-ray powder diﬀraction pattern

of Th-saturated ZrBTP-0.8 heated directly to 1200 °C in air.

In separate experiments, the poorly ordered Gen I ZrBTP-

0.8 and Th-loaded ZrBTP-0.8 samples were heated in situ on

the high-temperature stage of an X-ray diﬀractometer.

Typically, data were taken every 100 °C with each temperature

being attained using a 10 °C/min ramp. The data (Figure 19)

show that the pyrophosphate forms abruptly from poorly

ordered material at between 700 and 800 °C without any

intermediate phase implying a ﬁrst-order transition.

The variation in unit cell volume as a function of temperature

is shown in Figure 20 and shows that, whereas the unit cell

volume for the Th-saturated materials undergoes a nonlinear

thermal expansion, the cell volume of the as-prepared material

remained relatively constant.

Rietveld analysis of a sample of the poorly ordered Th-

saturated ZrBTP-0.8 (Gen I) heated directly to 1200 °C in air

is shown in Figure 18. Multiple SEM-EDX measurements gave

the stoichiometry of the heated sample as approximately

Th_0.23Zr_0.77P_1.79O_6.5. The diﬀraction pattern could be identiﬁed

as that of the ZrP₂O₇pyrophosphate and could be reﬁned on a

cubic cell Pa3 with an a-parameter of 8.3250 Å. During the

̅

reﬁnement Th was allowed to occupy the Zr site according to

the determined stoichiometry. During reﬁnement this

occupancy changed only a little from the initial value. The

determined a-parameter was somewhat larger than that

obtained for the ZrP₂O₇without Th included, which gave a

value of 8.2849 Å.

Additional experiments with Ce-saturated ZrBTP-0.8 (Gen

I) also yielded the pyrophosphate structure with composition

Ce_0.14Zr_0.86P_1.76O_6.33having unit cell volumes similarly enlarged

with respect to the as-prepared ZrBTP-0.8 materials as was the

case for the Th-saturated ZrBTP-0.8 materials. Such enlarged

L

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

special techniques have had to be developed to overcome this

problem.⁵⁶One such technique involves the layering in a small

vial of two solutions on top of each other, one containing the

metal and the other the ligand. Slow diﬀusion of one solution

into the other allows slow crystallization to occur. Although this

technique has been used in certain cases to produce very small

quantities of materials suitable for single-crystal X-ray

diﬀraction, it is not suitable for producing suﬃcient material

for evaluating properties or for producing the quantities of

materials necessary for real-world applications or even to enable

sorption studies to be undertaken.

To the author’s knowledge there have been only three

reports of crystalline MOFs prepared from BTP. These have

involved the following metals including Cu,⁴⁵Pb,⁵⁷and Fe.⁵⁸

However, X-ray amorphous Zr phosphonates based on other

phosphonate ligands have been reported, and the extraction of

heavy elements such as Hg, Cd, Cu, and Pb has been

investigated⁵⁹including by the author and his colleagues.^40,41

Using the rigid BTP ligand herein it has been possible to

prepare ZrBTP compounds with X-ray powder patterns ranging

from those indicative of poorly ordered nanocrystalline

materials to those indicative of extremely well-ordered

nanocrystalline materials depending on whether or not a

mineralizing agent such as HF is used. Not surprisingly, it has

been found here that the most eﬀective ZrBTP extractants are

the least crystalline Gen I materials with x in the range 0.7−0.9

having particle diameters ∼30 nm and a decrease in the surface

area from 500 to 30 m²/g. That said, the somewhat more

crystalline mixed-phase Gen II materials prepared using HF and

having slightly larger particle sizes and surface areas exceeding

50 m²/g, also appear to function reasonably well. In contrast,

the highly crystalline Gen III materials showed relatively low

sorption capacities. Microanalytical measurement and other

data have conﬁrmed that the very ﬁne-grained Gen I materials

are obtained with the same overall stoichiometry of the reaction

mixture. From the XRD results it appears that the poorly

crystalline Gen I materials have the incipient structure of the

phases that subsequently become more crystalline as HF is

included as a mineralizing agent. Unfortunately, the multiphase

nature of the crystalline phases obtained has precluded crystal

structure solution of these materials within the ﬁeld of

compositional space being investigated. Clearly, other techni-

ques will need to be employed to prepare more crystalline

materials suitable for single-crystal or ab initio structure

determination from powders.

Although the nanocrystalline Gen I materials appear to have

similar structural motifs to the microcrystalline Gen II and III

phases it would appear that particle packing as well as

phosphonic acid packing within and between nanoparticles is

high.

Various options could be considered for the structural motif

of the Gen I materials. One possibility could be that the

materials are simply nanoparticles of zirconia or oxyhydroxides

capped with polyphosphonates. However, zirconia nano-

particles can be ruled out, since neither XRD or FTIR data

support this hypothesis. Polyphosphonate-capped, poorly

crystalline zirconium oxyhydroxide can also be discounted if

the resemblance to the crystalline phases that are produced on

addition of HF is accepted and considering that these phases do

no match any known material.

Figure 19. Selected in situ X-ray diﬀraction patterns at diﬀerent

temperatures of Th-saturated ZrBTP-0.8 and corresponding Rietveld

reﬁnements where appropriate.

Figure 20. Unit cell volume determined from the Rietveld reﬁnement

for Gen I ZrBTP-0.8 material heated to diﬀerent temperatures before

and after being saturated with thorium.

unit cell volumes are consistent with replacement of the smaller

Zr⁴⁺by the larger trivalent lanthanide and Th⁴⁺when in the

same coordination.

When the Gen I ZrBTP-0.8 phase was saturated with the

range of cations used in the adsorption experiments the only

phase produced was the pyrophosphate (see Figure S10 of

Supporting Information). Thus, it would appear that the

pyrophosphate structure is always favored at least for ZrBTP-

0.8 composition (80 mol % P). This might not necessarily be

the case for other compositions. Further experimental work is

needed to address the phase composition of the phosphate

ceramics produced as a function of phosphorus mole fraction of

the cation-loaded ZrBTP-x phase.

4. DISCUSSION

It is well-known that most coordination polymers prepared

from tetravalent cations are highly insoluble such that rapid

precipitation occurs when solutions of the metal and ligand are

combined. Thus, recrystallization is usually impossible, and

Regardless of the particle sizes and the phases involved in the

Gen I−III materials, similar cation selectivities are observed for

all samples. From the similarity in the selectivity of all of the

M

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

ZrBTP-x materials and the X-ray amorphous materials

previously prepared from ATMP⁴⁰and zoledronic acid,⁴¹it

can be concluded that sorption properties of the zirconium

polyphosphonates are primarily determined by the acidity of

the phosphonate groups and are little inﬂuenced by the detailed

architecture of the coordination polymer. Indeed, in the case of

the previously reported Zr(IV)-zoledronate materials, perturba-

tions of the P−O stretching frequencies in the FTIR spectra

after sorption of Th⁴⁺convincingly linked the sorption

properties to the phosphonate groups. Similar observations of

modulated phosphonate group stretching frequencies have also

been made for the present materials (see Figure S11 of

Supporting Information section). The existence of bands at

1385 cm⁻¹in the FTIR spectra of Th⁴⁺saturated samples also

suggests that the nitrate anion remains in the coordination

sphere during sorption. In addition, it has also been shown that

progressive replacement of one (ZrBDP-x) and two phospho-

nates (ZrBMP-x) of the benzene ring by carboxylate does little

to modify the underlying selectivity and only serves to reduce

capacity. Since the carboxylate group appears not to modify to

any great extent the selectivity it might be surmised that the

carboxylate group is involved in framework building but not

cation binding. This once again implicates phosphonate groups

in cation binding.

One potentially very important result of this study that

requires further investigation is the tendency of the ZrBTP-x

materials to display a much higher capacity for the extraction of

heavy lanthanides such as Dy³⁺and Ho³⁺compared with light

lanthanides at high equilibrium cation concentrations (Figure

15). At high HNO₃concentrations the increase in selectivity

and hence extractability as a function of lanthanide atomic

number is particularly pronounced. Such data suggest that in a

column conﬁguration, the present materials should show

excellent prospects for the selective extraction of these two

industrially important lanthanides from complex solutions

including those containing other lanthanides. Studies of the

column performance of granulated versions of the ZrBTP

materials are ongoing and will be reported in due course.

In studies of lanthanide adsorption on diﬀerent aluminosi-

licates including zeolites⁶⁰and clay minerals⁶¹K_dvalues exhibit

a general overall increase across the lanthanide series.

Superimposed on this overall increase in K_dwith decreasing

ionic radius is observed a dip at Nd³⁺, Gd³⁺, and Er³⁺

corresponding to the one-fourth, one-half, and three-fourths

ﬁlled 4f subshell. This has been dubbed a “tetrad eﬀect” and

consists of a subtle decrease in K_devery fourth element in the

lanthanide series. Such an eﬀect was best accounted for

theoretically by Kawabe in 1992.⁶²Since the sorption data

presented in this study have not included all lanthanides, it is

not possible to categorically rule out such a subtle eﬀect here.

However, at the ionic strengths investigated a progressive

increase in K_dvalues across the lanthanide series has been

clearly observed, and separation factors between heavy and light

lanthanides range from 2 to 10. At high ionic strengths and

lower HNO₃concentrations a more abrupt discontinuity in the

adsorption of Dy³⁺and Ho³⁺compared to the other

lanthanides was observed. Under such conditions, separation

While most coordination polymers/MOFs that have been

reported to date are macroscopic crystalline materials generally

having high surface areas, examples of nanoparticulate and X-

ray amorphous MOFs have been reported. For instance, Oh et

al.²⁸reported the solvent-initiated spontaneous formation of

nanoparticulate coordination polymers based on Schiﬀ bases. It

has also been demonstrated that crystalline zeolite imidazolate

frameworks (ZIFs) such as ZIF-8 can be amorphized relatively

easily via the application of pressure to yield X-ray amorphous

materials with similar short-range order to the crystalline

counterparts. Such amorphous ZIFs can be advantageous from

an applications point of view. The idea has been ventured that

crystalline ZIFs can be used for the incorporation of guest

molecules such as CO₂and I₂and that subsequent collapse of

the framework through the application of pressure can be used

to entrap these molecules.⁶³

This is related to the thrust of our work over more than a

decade during which we have focused on the development of

materials for the selective extraction of ionic species and the

subsequent conversion of the nanostructured adsorbent into a

crystalline phase in which the extracted species is irreversibly

bound. In this regard our data show that the lanthanide- and

actinide-loaded adsorbents are always converted to Zr

pyrophosphate structures (ZrP₂O₇). As phosphate ceramics

are generally regarded as being very insoluble many studies

have addressed the use of phosphate ceramics as matrices for

the immobilization of radioactive actinide and ﬁssion product

elements. In this regard Itoh et al.⁶⁴have shown that even Cs

can be converted to highly stable phosphate ceramics at 700 °C

having Cs leach rates less than 1 × 10⁻¹¹g cm⁻²day⁻¹in 0.1

mol/L HCl at 100 °C far exceeding those of borosilicate glass.

ZrP₂O₇is one of the most widely studied members of the

pseudo cubic A^IVM^VO₇extensive family of materials (M = P,

2

V, As) in which a great variety of ions can be found, in

particular, that occupy the A site (Si, Ge, Sn, Pb, Ti, Zr, Hf, Mo,

W, Re, Ce, U), while the M site can accommodate elements

such as P, V, and As.⁶⁵

The ideal (usually high-temperature) parent structure (space

group symmetry Pa31, Z″ = 4) is built of corner-connected

AO₆octahedra and nominally linear M₂O₇(pyrophosphate,

pyrovanadate, or pyroarsenate) units (in turn built of two

corner-connected MO₄tetrahedra), as shown in (Figure S9 of

the Supporting Information). Thus, it is obvious that the

structure is quite versatile in terms of the compositions.

The present study explored only the phosphorus-rich region

of the Zr−P−F phase diagram. Further work is needed to

explore the compositional ﬁeld in greater detail and to focus on

the crystallization of single-phase materials that are amenable to

structure solution by either single-crystal diﬀraction or ab initio

methods and more precise determination of structure−function

relationships.

The classical extractants for lanthanide separations include

the likes of alkyl phosphoric and phosphonic acids (e.g., 2-

ethylhexyl phosphoric acid, HDEHP), amines (e.g., Cyannex

301), carboxylic acids (e.g., Versatic 10), and so forth.⁴Each of

these have been investigated extensively and have shown

eﬃciency in both intralanthide and actinide−lanthanide

separations in both solvent extraction and extraction chromato-

graphic form. Recently, Wei et al.⁶⁶investigated HDEHP-based

extraction chromatography in which the support phase

comprised polymer-loaded silica spheres. These materials

were relatively ineﬀective at HNO₃concentrations greater

than one giving K_dvalues less than 100 for trivalent Ce, Gd, Eu,

factors between Dy and Ho and other lanthanides (K_d(Dy)

/

K_d(M)) are in the range of 4−6 and would seem to be quite

useful in terms of the separation of these two lanthanides. In

addition we observed very rapid kinetics for the Gen I materials

with maximum sorption occurring within seconds to minutes,

which is for practical purposes instantaneous.

N

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

and Am. The relatively small diﬀerneces in K_dbetween

lanthanides and the actinide suggested that a practical

intralanthanide or lanthanide−actinide would not be achiev-

able.

ASSOCIATED CONTENT

■

S

* Supporting Information

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/acs.inorg-

chem.6b00954.

On the other hand Silbernagel⁶⁷recently reported the

preparation of zirconium phosphate-phosphonates that have

shown interesting lanthanide extraction properties. The Zr(IV)-

BTP hybrids of the present study seem to signiﬁcantly

outperform those materials, as eﬃcient extractions can be

achieved even up to 4 M HNO₃. These are acid concentrations

relevant to nuclear fuel reprocessing raﬃnates as well as in

mining liquors.

FTIR data, additional XRD patterns, small-angle X-ray

scattering data, polyhedral structure representations

(PDF)

AUTHOR INFORMATION

■

Corresponding Author

*E-mail: vluca@cnea.gov.ar.

Notes

The present ZrBTP-x materials therefore would seem to oﬀer

interesting prospects for lanthanide-major actinide and

lanthanide-minor actinide separations especially at the back-

end of the nuclear fuel cycle. Also extremely relevant is the data

presented that indicated that intralanthanide separations can be

accomplished with the present materials. The ability to separate

a particular lanthanide, especially a heavy lanthanide from

within the series using an operationally simple approach with

few unit operations, might prove to be invaluable, as expanded

technological applications result in increased demand for heavy

lanthanides, in particular.⁶⁸

We are presently focusing on the column deployment of the

nanostructured ZrBTP-x materials especially in the granular

form utilizing a polyacrylonitrile binding matrix as well as better

deﬁning the chemical stability of the phosphate ceramics

derived through thermal treatment of the lanthanide and

actinide-loaded nanostructured Zr phosphonates.

Although the most eﬀective materials prepared here are

nanometer-sized particles that do not lend themselves easily to

use as stationary phases in packed columns, we show here that

good initial results can be achieved through the incorporation

of the ZrBTP-x materials in sub-millimeter size polyacrylonitrile

matrices for use in columns. This work will be reported in due

course.

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

The authors would like express their sincere appreciation to S.

Maynadie of the Institut de Chimie Sep

́

arative de Marcoule for

acquisition of the ³¹P MAS NMR data. Gratitude is due to

Antonio Ciagla for assistance with some of the adsorption

experiments. Major funding for this work was provided by the

́

́ ́

Comision Nacional de Energia Atomica.

REFERENCES

■

(1) Davidson, D.; Hunter, I. Beneﬁts of an Integrated Fuel Cycle on

Repository Eﬀective Capacity. Waste Management SymposiumGlobal

Accomplishments in Environmental and Radioactive Waste Management;

Tucson, AZ, 2006.

(2) Kessler, G. Sustainable and Safe Nuclear Fission Energy

Technology and Safety of Fast and Thermal Nuclear Reactors; Springer:

Heidelberg, Germany, 2012; p 243.

(3) Malmbeck, R.; Nourry, C.; Ougier, M.; Soucek, P.; Glatz, J. P.;

Kato, T.; Koyama, T. Energy Procedia 2011, 7, 93−102.

(4) Xie, F.; Zhang, T. A.; Dreisinger, D.; Doyle, F. Miner. Eng. 2014,

56, 10−28.

(5) Wood, S. A. Chem. Geol. 1990, 82, 159−186.

(6) Gelis, A. V.; Vandegrift, G. F.; Bakel, A.; Bowers, D. L.; Hebden,

A. S.; Pereira, C.; Regalbuto, M. Radiochim. Acta 2009, 97, 231−232.

(7) Lumetta, G. J.; Gelis, A. V.; Braley, J. C.; Carter, J. C.; Pittman, J.

W.; Warner, M. G.; Vandegrift, G. F. Solvent Extr. Ion Exch. 2013, 31,

223−236.

5. CONCLUSIONS

Solid extractants consisting of poorly crystalline Zr(IV)-

coordination polymers based on BTP and carboxylate-

substituted derivatives have been prepared. The use of HF as

a mineralizer resulted in crystalline mixed-phase compounds

whose structure could not be resolved. Of the materials studied,

the poorly crystalline ZrBTP-x phases prepared in the absence

of ﬂuoride showed superior performance in the extraction of

trivalent lanthanides and tetravalent Th even in greater than 3

M HNO₃. Under such highly acidic conditions, the compounds

remained stable and displayed enhanced selectivity for heavy

lanthanides, suggesting that they might be useful for actual

applications. In addition to excellent selectivity for the

mentioned trivalent elements the ZrBTP-x compounds showed

extremely fast kinetics and good capacities. It remains to be

seen if appropriate choice of conditions could result in materials

suitable for the extraction of trivalent actinides from radio

lanthanides. Work addressing column dynamics of granulated

materials and the extraction of radiolanthanides and actinides

from Purex-type raﬃnates is underway. Thermal treatment of

the lanthanide and Th-loaded extractants results in formation of

highly crystalline pyrophosphates with the general formula

A^IVP₂O₇and where Zr, Th, and the lanthanides occupy the A

site. Such materials are expected to be highly insoluble and

therefore suitable for radioactive waste repository disposition

after suitable conditioning.

(8) Mincher, B. J.; Schmitt, N. C.; Case, M. E. Solvent Extr. Ion Exch.

2011, 29, 247−259.

(9) Nash, K. L. J. Alloys Compd. 1997, 249, 33−40.

(10) Zheng, Z.; Philip, C. V.; Anthony, R. G.; Krumhansl, J. L.;

Trudell, D. E.; Miller, J. E. Ind. Eng. Chem. Res. 1996, 35, 4246−4256.

(11) Zheng, Z.; Gu, D.; Anthony, R. G.; Klavetter, E. Ind. Eng. Chem.

Res. 1995, 34, 2142−2147.

(12) Feng, X.; Fryxell, G. E.; Wang, L.-Q.; Kim, A. Y.; Liu, J.;

Kemner, K. M. Science 1997, 276, 923−926.

(13) Griffith, C. S.; De Los Reyes, M.; Scales, N.; Hanna, J. V.; Luca,

V. ACS Appl. Mater. Interfaces 2010, 2, 3436−3446.

(14) Luca, V.; Bertram, W. K.; Widjaja, J.; Mitchell, D. R. G.; Griffith,

C. S.; Drabarek, E. Microporous Mesoporous Mater. 2007, 103, 123−

133.

(15) Sizgek, G. D.; Sizgek, E.; Griffith, C. S.; Luca, V. Langmuir 2008,

24, 12323−12330.

(16) Sizgek, G. D.; Griffith, C. S.; Sizgek, E.; Luca, V. Langmuir 2009,

25, 11874−11882.

(17) Fryxell, G. E.; Wu, H.; Lin, Y.; Shaw, W. J.; Birnbaum, J. C.;

Linehan, J. C.; Nie, Z.; Kemner, K.; Kelly, S. J. Mater. Chem. 2004, 14,

3356−3363.

(18) Fryxell, G. E.; Lin, Y.; Fiskum, S.; Birnbaum, J. C.; Wu, H.;

Kemner, K.; Kelly, S. Environ. Sci. Technol. 2005, 39, 1324−1331.

O

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

Article

(19) Florek, J.; Chalifour, F.; Bilodeau, F.; Larivier

Adv. Funct. Mater. 2014, 24, 2668−2676.

(20) Shusterman, J. A.; Mason, H. E.; Bowers, J.; Bruchet, A.; Uribe,

E. C.; Kersting, A. B.; Nitsche, H. ACS Appl. Mater. Interfaces 2015, 7,

20591−20599.

(21) Zhu, Y.-P.; Yuan, Z.-Y. Mesoporous Organic-Inorganic Non-

Siliceous Hybrid Materials - Basic Principles and Promising Multi-

functionality; Springer: Heidelberg, Germany, 2015.

(22) Braun, T.; Ghersini, G. Extraction Chromatography; Elsevier:

Amsterdam, 1975.

́

e, D.; Kleitz, F.

(54) Limousin, G.; Gaudet, J.-P.; Charlet, L.; Szenknect, S.; Barthes,

V.; Krimissa, M. Appl. Geochem. 2007, 22, 249−275.

(55) Chiarizia, R.; Mcalister, D. R.; Herlinger, A. W. Sep. Sci. Technol.

2005, 40, 69−90.

(56) Batten, S. R.; Neville, S. M.; Turner, D. R. Coordination Polymers

- Design, Analysis and Application; RSC Publishing, 2009; p 10.

(57) Tang, S.-F.; Lv, X.-X.; Li, L.-J.; Wang, C.; Zhao, X.-B. Inorg.

Chem. Commun. 2014, 39, 51−55.

(58) Pramanik, M.; Bhaumik, A. Chem. - Eur. J. 2013, 19, 8507−8514.

(59) Shah, B.; Chudasam, U. Desalin. Water Treat. 2012, 38, 276−

284.

(23) Zhang, A.; Wei, Y.; Kumagai, M. J. Radioanal. Nucl. Chem. 2005,

(60) Gladysz-Plaska, A.; Majdan, M.; Pikus, S.; Lewandowski, W. J.

Colloid Interface Sci. 2007, 313, 97−107.

(61) Coppin, F.; Berger, G.; Bauer, A.; Castet, S.; Loubet, M. Chem.

Geol. 2002, 182, 57−68.

265, 409−417.

(24) Horwitz, E. P.; Chiarizia, R.; Dietz, M. L.; Diamond, H.; Nelson,

D. M. Anal. Chim. Acta 1993, 281, 361−72.

(25) Horwitz, E. P.; Dietz, M. L.; Nelson, D. M.; Larosa, J. J.;

Fairman, W. D. Anal. Chim. Acta 1990, 238, 263−271.

(26) Mokhodoeva, O. B.; Myasoedova, G. V.; Zakharchenko, E. A.

Radiochemistry 2011, 53, 35−43.

(62) Kawabe, I. Geochem. J. 1992, 26, 309−335.

(63) Chapman, K. W.; Sava, D. F.; Halder, G. J.; Chupas, P. J.;

Nenoff, T. M. J. Am. Chem. Soc. 2011, 133, 18583−18585.

(64) Itoh, K.; Nakayama, S. J. Mater. Sci. 2002, 37, 1701−1704.

(65) Withers, R. L.; Tabira, Y.; Evans, J. S. O.; King, I. J.; Sleight, A.

W. J. Solid State Chem. 2001, 157, 186−192.

(27) Hong, M.-C.; Chen, L. Extraction Chromatography; Wiley, 2009.

(28) Oh, M.; Mirkin, C. A. Nature 2005, 438, 651−654.

(29) Poojary, D. M.; Vermeulen, L. A.; Vicenzi, E.; Clearfield, A.;

Thompson, M. E. Chem. Mater. 1994, 6, 1845−1849.

(30) Alberti, G.; Casciola, M.; Costantino, U.; Vivani, R. Adv. Mater.

1996, 8, 291−303.

(66) Wei, W.; Kumagai, M.; Takashima, Y.; Modolo, G.; Odoj, R.

Nucl. Technol. 2000, 132, 413−423.

(67) Silbernagel, R.; Martin, C. H.; Clearfield, A. Inorg. Chem. 2016,

55, 1651−1656.

(31) Kitagawa, S.; Matsuda, R. Coord. Chem. Rev. 2007, 251, 2490−

2509.

(68) Hatch, G. P. Elements 2012, 8, 341−346.

(32) Cahill, R.; Shpeizer, B.; Peng, G. Z.; Bortun, L.; Clearfield, A.

Separation of f Elements, Proc. Am. Chem. Soc. Symp. 1995, 165−76.

(33) Burns, J. D.; Shehee, T. C.; Clearfield, A.; Hobbs, D. T. Anal.

Chem. 2012, 84, 6930−6932.

(34) Burns, J. D.; Clearfield, A.; Borkowski, M.; Reed, D. T.

Radiochim. Acta 2012, 100, 381−387.

(35) Amicangelo, J. C.; Leenstra, W. R. Inorg. Chem. 2005, 44, 2067−

2073.

(36) Costantino, F.; Donnadio, A.; Casciola, M. Inorg. Chem. 2012,

51, 6992−7000.

(37) Leenstra, W. R.; Amicangelo, J. C. Inorg. Chem. 1998, 37, 5317−

5323.

(38) Lin, X.-Z.; Yuan, Z.-Y. Eur. J. Inorg. Chem. 2012, 2012, 2661−

2664.

(39) Carboni, M.; Abney, C. W.; Liu, S.; Lin, W. Chem. Sci. 2013, 4,

2396−2402.

(40) Veliscek-Carolan, J.; Hanley, T. L.; Luca, V. Sep. Purif. Technol.

2014, 129, 150−158.

(41) Luca, V.; Hanna, J. V. Hydrometallurgy 2015, 154, 118−128.

(42) Zhu, Z.; Pranolo, Y.; Cheng, C. Y. Miner. Eng. 2015, 77, 185−

196.

(43) Jenkins, I. L. Hydrometallurgy 1979, 5, 1−13.

(44) Su, Y.; Balmer, M. L.; Wang, L.; Bunker, B. C.; Nyman, M.;

Nenoff, T.; Navrotsky, A. MRS Online Proc. Libr. 1999, 556, 77−84.

(45) Kong, D.; Zon, J.; McBee, J.; Clearfield, A. Inorg. Chem. 2006,

45, 977−986.

(46) Ho, Y.-S. J. Hazard. Mater. 2006, 136, 681−689.

(47) Ho, Y. S.; McKay, G. Adsorpt. Sci. Technol. 2002, 20, 797−815.

(48) Fernandez Lopez, E.; Sanchez Escribano, V.; Panizza, M.;

Carnasciali, M. M.; Busca, G. J. Mater. Chem. 2001, 11, 1891−1897.

(49) Brunet, E.; Cerro, C.; Juanes, O.; Rodriguez-Ubis, J.; Clearfield,

A. J. Mater. Sci. 2008, 43, 1155−1158.

(50) Josse, S.; Faucheux, C.; Soueidan, A.; Grimandi, G.; Massiot, D.;

Alonso, B.; Janvier, P.; Laib, S.; Pilet, P.; Gauthier, O.; Daculsi, G.;

Guicheux, J. m.; Bujoli, B.; Bouler, J.-M. Biomaterials 2005, 26, 2073−

2080.

(51) Hanna, J. V.; Smith, M. E. Solid State Nucl. Magn. Reson. 2010,

38, 1−18.

́

(52) Gomez-Alcantara, M. d. M.; Cabeza, A.; Moreno-Real, L.;

Aranda, M. A. G.; Clearfield, A. Microporous Mesoporous Mater. 2006,

88, 293−303.

(53) Kunz, W. Speciﬁc Ion Eﬀects; World Scientiﬁc, 2009.

P

DOI: 10.1021/acs.inorgchem.6b00954

Inorg. Chem. XXXX, XXX, XXX−XXX

Article Doi

Zirconium(IV)-Benzene Phosphonate Coordination Polymers: Lanthanide and Actinide Extraction and Thermal Properties

DOI: 10.1021/acs.inorgchem.6b00954

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.inorgchem.6b00954

Full text of DOI:10.1021/acs.inorgchem.6b00954

Products guided by the article

R&D Labs maybe for 1192028-87-7

Relevant to this article

Hot Product