The influence of linker geometry in bis(3-hydroxy-N-methyl-pyridin-2-one) ligands on solution phase uranyl affinity

Article abstract of DOI:10.1002/chem.201002372

Seven water-soluble, tetradentate bis(3-hydroxy-N-methyl-pyridin-2-one) (bis-Me-3,2-HOPO) ligands were synthesized that vary only in linker geometry and rigidity. Solution-phase thermodynamic measurements were conducted between pHa 1.6 and pHa 9.0 to dete

Full text of DOI:10.1002/chem.201002372

DOI: 10.1002/chem.201002372
The Influence of Linker Geometry in Bis(3-hydroxy-N-methyl-pyridin-2-one)
Ligands on Solution Phase Uranyl Affinity**
Gꢀza Szigethy and Kenneth N. Raymond*[a]
Abstract: Seven water-soluble, tetra-
dentate bis(3-hydroxy-N-methyl-pyri-
din-2-one) (bis-Me-3,2-HOPO) ligands
were synthesized that vary only in
linker geometry and rigidity. Solution-
phase thermodynamic measurements
were conducted between pH 1.6 and
pH 9.0 to determine the effects of
these variations on proton and uranyl
cation affinity. Proton affinity decreas-
es by introduction of the solubilizing
triethylene glycol group as compared
to unsubstituted reference ligands.
Uranyl affinity was found to follow no
discernable trends with incremental
geometric modification. The butyl-
linked 4li-Me-3,2-HOPO ligand exhib-
ited the highest uranyl affinity, con-
AHCTUNGTRENNsGUN istent with prior in vivo decorporation
results. Of the rigidly-linked ligands,
the o-phenylene linker imparted the
best uranyl affinity to the bis-Me-3,2-
HOPO ligand platform.
Keywords: analytical methods · lig-
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
Introduction
overload treatment.[6–8] Applying rationally designed ligands
towards actinide science is of particularly great interest as
nuclear power becomes an attractive carbon-free energy al-
ternative, and the associated risk of environmental and bio-
logical exposure to actinides increases.
The approach of rational ligand design towards targeted ion
chelation seeks to match ligand traits with the geometric
and thermodynamic preferences of a target ion.[1] Modifi-
A
Unique to the early actinides is the class of linear dioxo
cations (actinyls, AnO2n+) adopted by uranium, neptunium,
and plutonium in their +5 or +6 oxidation states. Actinyls
pose an unusual challenge to the practice of rational ligand
design, because while they display hard Lewis acidity typical
of the f-elements, they maintain near-linearity in all their
known complexes, with the largest distortions of just over
118 occurring only in the presence of very sterically demand-
ing ligands.[9] Unlike most transition-metal dioxo cations, lig-
character of coordinating atoms, ligand denticity, and geo-
metric pre-organization. Optimization of these design
characteristics is observed in bacterial ferric ion shuttling
mechanisms,[2,3] but is also central to synthetic compounds
applied in many different selective extraction/removal appli-
cations, such as heavy-metal decorporation,[4,5] and iron-
[a] Dr. G. Szigethy, Prof. Dr. K. N. Raymond
Chemical Sciences Division
Lawrence Berkeley National Laboratory
Berkeley, CA 94720 (USA)
and
AHCTUNGTREGaNNNU nds coordinate to actinyl cations almost exclusively in a
plane perpendicular to the An=O bonds with deviations
from this behavior again only occurring in sterically congest-
ed complexes.[10] The equatorial coordination plane displays
little to no orbital-dictated directionality, with observed co-
ordination geometries ranging from trigonal to hexagonal
planar, depending on ligand sterics and chelating ability.[11–14]
Additionally, the terminal oxo atoms of the actinyl cations
(especially in their +6 oxidation state) display poor Lewis
basicity, although they have been observed to interact with
metal cations in the solid state.[11,13,15–19]
Department of Chemistry
University of California at Berkeley
Berkeley, CA 94720-1460 (USA)
Fax: (510)486-5283
E-mail: raymond@socrates.berkeley.edu
[**] Part 62 in the series “Specific Sequestering Agents for the Actinides.”
For Part 61, see reference [34].
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem201002372. It contains experi-
mental procedures for the synthesis of bis-Me-3,2-HOPO ligands, X-
ray diffraction collection and refinement details, conformational anal-
The uranyl cation (UO22+) is the most common actinyl
species studied because uranium is the second most abun-
dant naturally occurring actinide as well as the one most fre-
quently employed in nuclear power generation.[20] Addition-
ysis of [UO2(L9)
tra, and speciation diagrams for uranyl titrations.
ACHTUNGTRENNUNG(dmso)], reversibility analyses, UV/Vis titration spec-
1818
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1818 – 1827

FULL PAPER
ally, the uranyl cation is the most stable form for uranium in
oxidizing environments or in vivo. Several different ap-
proaches have been used to design uranyl-selective ligands;
rigid polypyrrole macrocycles have utilized their pre-ar-
ranged geometry to form planar uranyl chelates,[21] apically-
oriented molecules have introduced Lewis acids in the vicin-
ity of one uranyl oxo moiety,[22,23] while some fold-over mac-
rocycles have achieved both these goals simultaneously.[24,25]
Significant effort has been dedicated towards the develop-
ment of efficient uranyl cation extraction agents for both an-
alytical and industrial applications, but these studies typical-
ly address low pH conditions typical of industrial metal-sep-
aration technologies.[26,27] However, the rational develop-
ment of biological decorporation agents necessarily requires
studying the efficacy of uranyl chelation at in vivo pH
ranges, which are higher than those of industrial/remedia-
tion applications. Examples of biologically relevant investi-
gations include that of Czerwinski and co-workers that ad-
dresses uranyl speciation with the naturally occurring sidero-
phore desferrioxamine (DFO),[28] as well as those by Durbin
and co-workers that examine the in vivo decorporating abili-
ty of polybidentate catecholamides and their structural ana-
logues towards the uranyl and transuranic cations.[29–32]
These last studies revealed that 4- or 5-carbon linear linkers
provided optimal decorporating ability to polybidentate lig-
linked ligands explored, the 3,4-thiophene, o-phenylene,
a,a’-m-xylene, and 1,8-fluorene linkers (Figure 1) showed
the most favorable uranyl coordination geometries; the first
two contain very short linkers, while the last two contain the
longest linkers investigated.[34] These results raise the ques-
tion of whether favorable geometry (optimized in short and
long linker lengths) or intramolecular hydrogen bonding
(optimized in short linkers) is most important in determin-
ing uranyl affinity.
Solution-phase titrations were carried out with bis-Me-
3,2-HOPO ligands to determine and compare their uranyl
affinities. Unfortunately, the bis-Me-3,2-HOPO ligands pre-
viously synthesized are not soluble enough in water to
enable such measurements, prompting the design and syn-
thesis of water-soluble bis-Me-3,2-HOPO ligands L2H2–L8H2
(Figure 2). These ligands incorporate linkers that are struc-
turally analogous to previous ligands, but which also contain
a methyl-protected triethyleneglycol moiety to promote
water solubility, enabling quantification of their uranyl affin-
ities.
ACHTUNGTRENNUNGands, but the detailed thermodynamic rationale for this
result is relatively unexplored.
Recent work in our group has focused on optimizing the
equatorial geometric agreement between the uranyl cation
and bis(3-hydroxy-N-methyl-pyridin-2-one) (bis-Me-3,2-
HOPO) ligands utilizing linker molecules of varying geome-
tries.[33–35] The Me-3,2-HOPO moiety is a bidentate, structur-
al analogue to the catecholamide group that binds through
hard Lewis basic oxygen atoms and is thus ideal for chelat-
ing the strong Lewis acidic actinide ions. While short linkers
best preserve intramolecular hydrogen bonds typical of cate-
cholamide ligands,[35] structural comparisons of the uranyl
complexes against that with Pr-Me-3,2-HOPO (L1H,
Figure 1) illustrated that the butyl-linked ligand 4li-Me-3,2-
HOPO provided the most favorable geometry for uranyl
chelation of the linearly-linked, tetradentate ligands (nli-
Me-3,2-HOPO, in which “n” is the number of carbon atoms
in the linker and “li” stands for “linear”).[35] Of the rigidly
Figure 2. Water-soluble bis-Me-3,2-HOPO ligands.
Results and Discussion
Ligand design and synthesis: Water solubility was imparted
to the bis-Me-3,2-HOPO ligands by introduction of the
methyl-protected triethyleneglycol moiety (3,6,9-trioxa-
decane) referred to hereon as PEG. This solubilizing group
carries no charge in the pH range investigated here, and
thus does not influence the charge of the resultant ligands at
varying pH. Uranyl affinity for polyether coordination is
conveniently poor compared to other neutral Lewis bases,
as evidenced by the anhydrous conditions necessary for syn-
thesizing crown-ether inclusion complexes of the uranyl
cation.[36] Thus, coordinative interference by the PEG group
is unlikely in these titration studies.
However, the location of the PEG substitution is of great
concern because it can feasibly affect uranyl affinity in two
ways: 1) by inductively influencing the HOPO ring electron-
ics, thereby affecting proton and uranyl affinity of the lig-
AHCTUNGTREGaNNUN nds, and 2) by introducing an additional steric influence, re-
sulting in a solution-phase structure significantly different
Figure 1. Mono- and bis-bidentate Me-3,2-HOPO ligands previously stud-
ied.[33–35]
from that seen in prior crystallographic analyses.[34,35] The
Chem. Eur. J. 2011, 17, 1818 – 1827
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1819

K. N. Raymond and G. Szigethy
Me-3,2-HOPO moiety is inert towards further substitution,
relegating PEG substitution to the linker moieties. Due to
the extended electronic conjugation in ligands containing
3,4-thiophene and o-phenylene linkers (L6H2 and L7H2),
electronic effects on the Me-3,2-HOPO moiety are un-
the aromatic linkers. Thioalkyl substitution on the 3,4-thio-
phene linker in bis-Me-3,2-HOPO ligands is known to result
(solv)}2] dimer formation,[33]
in sterically induced [{UO2(L)2ACHTGNUTERNUNNG a
behavior that ligand L8H2 will certainly duplicate. In con-
trast, the PEG groups in L7H2 are meta to the amide nitro-
gen atoms, and the ligand is thus expected to maintain pla-
narity upon uranyl chelation, with the resultant conjugation
allowing for a potentially increased inductive influence by
the PEG substitution. To investigate the presence/severity of
such an effect, the propoxy-substituted ligand L9H2 and its
uranyl complex were synthesized (Figure 3). Crystallograph-
ic parameters for this structure are listed in Table 1.
ACHTUNGTRENNUNG
ACHTUNGTRENNUNGands containing the linear nli- and a,a’-m-xylene backbones
(L2H2 through L5H2 and L8H2), because the methylene link-
ers are not conjugated to the Me-3,2-HOPO moieties.
In addition to electronic influences, the potential steric
consequences of PEG substitution become significant when
the backbone is either small or contains only non-ideal posi-
tions upon which to introduce the PEG group. Specifically,
steric interference with the amide linkers of the Me-3,2-
HOPO moiety is of great concern because intramolecular
hydrogen bonding involving the amide proton is a significant
interaction in HOPO and more generally in catecholamide
ligands.[37] Of particular concern is substitution on the 3,4-
thiophene linker, which is relegated only to the 2- and 5-po-
sitions, the steric consequence of which is the formation of
dimers of the form [{UO2(L)2ACHTNUGRTENUNG(solv)}2] as opposed to the
mononuclear uranyl species seen with the unsubstituted 3,4-
thiophene-bis-Me-3,2-HOPO ligand.[33] Because the avail-
able substitution positions on the 2li linker are adjacent to
the linking amides and cannot avoid some influence on the
proton and/or uranyl affinity in L2H2, the PEG moiety was
introduced at the same position on the 3li, 4li, and 5li link-
ers to make affinity comparisons at least internally con-
Figure 3. Ligand L9H2 and two views of the crystal structure of its uranyl
complex [UO2(L9)
ACHTNUTRGNE(NUG dmso)]. Thermal ellipsoids are drawn at the 50%
probability level. Hydrogen atoms and non-coordinating solvent are
omitted for clarity. Oxygen atoms are red, carbons gray, nitrogens blue,
sulfur yellow, and uranium silver.
ACHTUNGTRENNUNG
tains perhaps the most ideal PEG location of the ligands in-
vestigated. Firstly, the PEG moiety is located four carbon
atoms from the linking amide nitrogen atoms, ensuring mini-
mal steric effects, and secondly, the PEG group is electroni-
cally isolated from the HOPO moieties, owing to the pres-
ence of the methylene spacers between the phenyl ring and
the HOPO amides. Thus, it is expected that the behavior of
ligand L8H2 in solution will most closely resemble that of its
unsubstituted analogue compared to the other PEG-bis-Me-
3,2-HOPO ligands.
The crystal structure of [UO2(L9)
ACTHNUTRGNE(UNG dmso)] displays the ex-
pected mononuclear speciation seen with other bis-Me-3,2-
HOPO ligands,[34,35] with a pentagonal-planar uranyl coordi-
nation geometry provided on four points by (L9)2ꢀ, and a
fifth site occupied by a DMSO oxygen. Significantly, the
propoxy substituents are situated away from the amide link-
ers, so it can be assumed that PEG groups in L7H2 will not
impart significant steric influence on the resultant uranyl
Absent from the backbones illustrated in Figure 2 is the
1,8-fluorene moiety that was shown to provide a very good
geometric match to the uranyl cation coordination prefer-
ꢀ
ꢀ
complex. The equatorial U Ophenol and U Oamide bond
lengths are 2.34(1) and 2.42(2) ꢁ, respectively, which are
ꢀ
ences, while displaying a degree of ligand pre-organiza-
statistically identical to the similar equatorial U O bond
tion.[34] Unfortunately, the tremendous insolubility imparted
on bis-Me-3,2-HOPO ligands by the fluorene backbone
cannot be undone by solubilizing group substitution due to
synthetic impracticality and the natural electrophilic substi-
tution behavior of fluorene and its structural analogue di-
benzofuran; thus a soluble 1,8-diaminofluorene-linked bis-
Me-3,2-HOPO ligand was not pursued.
lengths in the uranyl complex with the unsubstituted o-
phenylene-Me-3,2-HOPO ligand.[34] As Figure 3 illustrates,
the backbone aryl group is not coplanar with the HOPO
moieties due to a torsion in the linking amide moieties. The
ꢀ
Cring Namide torsion angles of 156 and 1678 presumably re-
lieve a potential close contact between the amide protons in
the metal-chelating form that is observed to a similar extent
in the unsubstituted uranyl complex (torsions of 148 and
1578),[34] as well as in the EuIII complex with the structurally
similar o-phenylene-1,2-HOPO ligand (torsion angles of
157, 159, and 1698).[38] Thus, while there are slight deviations
Structural consequence of PEG Substitution in L7H2: The
geometric and electronic consequences of PEG substitution
on the 3,4-thiophene and o-phenylene linkers must be ad-
dressed owing to the extended conjugation possible through
from coplanarity in the [UO2(L9)
ACTHNUTRGNE(UNG dmso)] crystal structure,
1820
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1818 – 1827

Ligand effects
FULL PAPER
ACTHNUTRGNEUNG
Table 1. Crystallographic parameters for [UO2(L9)(dmso)].[a]
Solution thermodynamics: In the presence of dissolved
metal ion (Ma+) and protonated ligand (LHn, in which L is
a ligand with n removable protons), a pH-dependent metal–
ligand complex of general formula MmLlHh forms according
to the equilibrium shown in Equation (1). The relative
amount of each species in solution is determined by Equa-
tion (2), the rearrangement of which provides the standard
formation constant notation of logbmlh [Eq. (3)]. The logbmlh
value describes a cumulative formation constant, but a step-
wise formation constant (logK) can be calculated from
logbmlh values by using Equation (4). When addressing pro-
tonation constants, the stepwise formation constants are
commonly reported as dissociation constants (ꢀlogK, or
pKa).
formula
Mr
crystal system
space group
habit
C28H34N4O11SU·0.25H2O
877.19
orthorhombic
Pbca
plate
color
red
a [ꢀ]
b [ꢀ]
c [ꢀ]
a [8]
13.451(2)
15.839(2)
28.445(4)
90
b [8]
90
g [8]
90
V [ꢁ3]
crystal size [mm3]
T [K]
6060.3(16)
0.35ꢂ0.10ꢂ0.05
115(2)
8
1.923
5.494
Z
1calcd [gcmꢀ3
]
]
bmlh
mMaþ þ lLbꢀ þ hHþ
M L H
ðmaꢀlbþhÞþ
mcalcd [mmꢀ1
ƒ!
ð1Þ
ƒ
m
l
h
q min/max [8]
total reflections
data/restr./param.
3.29/26.36
27837
5978/0/410
3428
m
l
h
ð2Þ
ð3Þ
½MmLlHhꢁ ¼ bmlhð½Mꢁ ½Lꢁ ½Hꢁ Þ
FACHTUNGTRENNUNG(000)
m
l
h
log bmlh ¼ log fð½MmLlHhꢁÞ=ð½Mꢁ ½Lꢁ ½Hꢁ Þg
min/max transmission
0.324
R1 [I>2s(I)][b]
0.0304
0.0738
1.040
wR2ACHTUNGTRENNUNG
(all data)[b]
log K01n ¼logfð½LHnꢁÞ=ð½LHnꢀ1ꢁ½HꢁÞg ¼
log fðb01n=b01ðnꢀ1Þg ¼ log b01nꢀlog b01ðnꢀ1Þ
GOF[b]
ð4Þ
[a] CCDC-783329 ([UO2(L9)
ACTHNUGRTENUNG(dmso)]) contains the supplementary crystal-
lographic data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via
Because bis-Me-3,2-HOPO ligands are diprotic and re-
quire deprotonation for metal chelation, their affinity for
any metal is necessarily pH dependent (Scheme 1). Al-
though the most common pH for biological applications is
www.ccdc.cam.ac.uk/data_request/cif.
[b] R1=SjjFo jꢀjFc jj/SjFo j ;
wR2=[S[w(Fo2ꢀFc2)2]/S[w(Fo2)2]]1/2; GOF=[Sw(jFo jꢀjFc j)2/(nꢀm)]1/2
.
the overall structure is similar
to that of its unsubstituted ana-
logue (see Supporting Informa-
tion).
To assess the potential elec-
tronic effect of PEG substitu-
tion on the o-phenylene linker
in L7H2 the 1H NMR reso-
Scheme 1. Intramolecular hydrogen bonding in bis-Me-3,2-HOPO ligands upon deprotonation/metalation.
ACHTUNGTRENNUNG
that of blood serum, or pH 7.4, it is also desirable to investi-
gate the uranyl affinity of bis-Me-3,2-HOPO ligands in
acidic and caustic media, as such conditions are encountered
in industrial or storage tank waste remediation applica-
tions.[39,40]
ACHTUNGTRENNUNG
The amide proton chemical shifts of free L9H2 are 0.13 ppm
upfield of that in o-phenylene-Me-3,2-HOPO, suggesting
some electronic donation into the aryl linker upon alkoxy
substitution. However, these same resonances are only
0.06 ppm upfield of the unsubstituted ligand once coordinat-
ed to the uranyl cation. The similarity in these values sug-
gests that the inductive effects of alkoxy substitution on the
3- and 4-positions of the o-phenylene linker are minor, espe-
cially upon uranyl chelation. Additionally, the amide chemi-
cal shifts in L7H2 are 0.01 ppm shifted from L9H2, indicating
a near-identical inductive effect. The crystallographic and
NMR comparisons thus indicate that the PEG substitution
in L7H2 should have a relatively insignificant impact on the
resultant uranyl complex geometry either through steric or
inductive processes.
Proton titrations/affinity: Protonation constants of ligands
L2H2–L8H2 were determined using potentiometric titrations
at about 150–200 mm ligand concentration. As a point of
comparison for ligands L2H2–L5H2, the protonation con-
stants for the previously reported 2li- and 4li-Me-3,2-
HOPO ligands[35] were also measured, but were performed
spectrophotometrically at about 50 mm ligand concentration
due to their lower solubility. A 5% starting DMSO concen-
tration in all protonation titrations was required for consis-
tency with uranyl titration conditions. Data were analyzed
as described in the Experimental Section and the resultant
pKa values are listed in Table 2. The sum of the pKa values
Chem. Eur. J. 2011, 17, 1818 – 1827
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1821

K. N. Raymond and G. Szigethy
Table 2. pKa values of bis-Me-3,2-HOPO ligands.
ity of the second amide group to the initially deprotonated
phenolate group as shown in Figure 4 (middle), which ap-
plies in both the unsubstituted and PEG-functionalized nli-
Me-3,2-HOPO ligands. Similarly, the drop in pKa1 upon
PEG-functionalization is most likely caused by hydrogen-
bond donation from the PEG-amide nitrogen, as illustrated
in Figure 4 (right). This last stabilization from the PEG-
amide group in ligands L2H2–L5H2, while unintended, can be
considered roughly constant due to the identical location of
the PEG amide group in the linear backbone linkers. Thus,
the electronic and geometric effects of linker length can be
considered to be the prominent trend observed in titrations
with PEG-nli-Me-3,2-HOPO ligands.
n[a]
pKa1
pKa2
SpKa
L1H
–
2
4
2
3
4
5
2
2
5
6.12[41]
5.82(3)
6.01(1)
5.10(6)
5.37(2)
5.25(4)
5.52(7)
4.91(8)
5.09(3)
5.70(6)
–
–
2li-Me-3,2-HOPO
4li-Me-3,2-HOPO
6.68(3)
7.02(4)
6.45(1)
6.72(2)
6.60(3)
6.75(2)
6.22(1)
6.29(1)
6.75(2)
12.50(4)
13.03(4)
11.55(6)
12.09(3)
11.85(5)
12.27(7)
11.13(8)
11.38(3)
12.45(6)
L2H2
L3H2
L4H2
L5H2
L6H2
L7H2
L8H2
[a] n=number of carbon atoms between amide nitrogen atoms.
The pKa values for ligands L6H2–L8H2 cannot be com-
pared against their PEG-free versions due to the poor solu-
bility of the PEG-free analogues in aqueous media. Howev-
er, the PEG moiety in ligands L6H2 and L7H2 is expected to
have some inductive influence on proton affinity, since in
(SpKa) is also listed, which corresponds to the logb012 for-
mation constant and is a general measure of ligand acidity,
with lower SpKa values indicating a more acidic ligand.
One trend visible in Table 2 is that linearly-linked ligands
with shorter linkers (smaller n) have lower SpKa values than
those with longer linkers. This general trend is seen in lig-
1
free L6H2 and L7H2 the amide H NMR resonances are shift-
ed upfield by 0.29 and 0.08 ppm, respectively, as compared
to their unsubstituted analogues.[34] This inductive influence
is obviously much larger in L6H2 than in L7H2, but the steric
consequences of 2,5-disubstitution on the thiophene linker
are known to be significant,[33] so the inductive effect is pre-
sumed to be of minor importance in ligand L6H2 upon
uranyl chelation. In contrast, L8H2 is expected to experience
very little steric or electronic influences by PEG moiety in-
clusion, and indeed has a SpKa very similar to that of L5H2,
which is also an n=5 ligand.
A
4li-Me-3,2-HOPO (DSpKa ꢂ0.5). A second trend is that
substitution of the PEG moiety on the linear linkers by an
amide linker also lowers the SpKa compared to the unsubsti-
tuted ligands, as evidenced by comparing 2li-Me-3,2-HOPO
against L2H2 (n=2) and 4li-Me-3,2-HOPO against L4H2
(n=4). This trend is primarily caused by a drop in pKa1
(DpKa1,max ꢂ0.7), but also affects the pKa2 values (DpKa2,max
ꢂ0.4). The drop in SpKa upon PEG substitution (DSpKa/
PEGꢂ1.1) is more significant than that accompanying in-
cremental shortening of the linker lengths (average DSpKa/n
ꢂ0.23). However, in both these trends, it can be assumed
that the decrease in pKa arises from enhanced stabilization
of the deprotonated phenolate oxygen, since the electronics
of the HOPO moiety remain roughly constant in ligands
L2H2–L8H2.
The intramolecular hydrogen-bond stabilization of pheno-
lates by ortho amide protons are well established and is il-
lustrated in Figure 4 (left) as it applies to Me-3,2-HOPO
moieties.[37] However, the trends observed in Table 2 suggest
that additional hydrogen-bond interactions become signifi-
cant upon shortened linker length (smaller n) or PEG
moiety introduction. The drop in pKa1 associated with short-
ened linker length can be explained by an increased proxim-
Uranyl titrations/affinity: The uranyl affinities of ligands
L2H2 through L8H2 were determined by performing spectro-
photometric titrations. The poor solubility of the uranyl
complexes necessitated 5 mm analyte concentrations as well
as a 5% starting DMSO concentration to assist in solvating
the neutral uranyl complexes at neutral pH. Unfortunately,
uranyl titrations with ligands 2li-Me-3,2-HOPO, 4li-Me-3,2-
HOPO, and L9H2 are impossible, even at these low concen-
trations, owing to slow and reversible precipitation of a neu-
tral [UO2(L)ACTHNUGTRNEUNG(solv)] species near neutral pH. The bidentate
ligand Pr-Me-3,2-HOPO was found to be sufficiently soluble
for spectrophotometric titrations at these concentrations.
Metal-to-ligand ratios used in the titrations were those ob-
served in the crystal structures of the uranyl complexes with
unsubstituted Me-3,2-HOPO ligands, namely 1:1 for bis-Me-
3,2-HOPO ligands and 1:2 for Pr-Me-3,2-HOPO. Three in-
dependent titrations were measured between pH 2.4 and
11.0 except for cases in which reversibility analysis indicated
a point in the titrations beyond which the complexes under-
went an irreversible chemical change. Because the uranyl
complexes routinely form below pH 4, two strong acid titra-
tions (pH 3.0 to 1.6) were carried out for each ligand.
The uranyl titration spectra with PEG-substituted bis-Me-
3,2-HOPO ligands generally exhibited similar absorption
spectra in the range of 250–400 nm, with a large absorption
maximum in the 330–350 nm range, and often with a subtle
shoulder at longer wavelengths. These absorption features
Figure 4. Intramolecular stabilization of deprotonated Me-3,2-HOPO lig-
ACHTUNGTRENNUNGands.
1822
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1818 – 1827

Ligand effects
FULL PAPER
closely resemble those of the free ligands, and are thus at-
tributed to p!p* transitions. Titrations with L6H2 exhibited
the most unusual spectrum investigated, with a broad ab-
sorption feature in the 275–300 nm range, at which the other
bis-Me-3,2-HOPO ligands display absorption minima. Inde-
pendent of absorption features, however, absorption
maxima with all ligands red shift upon increasing pH, corre-
sponding to deprotonation/metallation of the ligand. Al-
though the uranyl complexes with bis-Me-3,2-HOPO ligands
are red or orange in the solid state and in concentrated solu-
tion, no significant absorption transition was observed in
wavelengths above 400 nm at 5 mm concentrations.
tal complex deprotonation). To maintain consistency across
the various bis-Me-3,2-HOPO ligands, titration data were
truncated before the onset of the [UO2(L)Hꢀ2]2ꢀ species.
For most ligands, this data truncation occurred at about
pH 9.0. Complete titration figures are provided in the Sup-
porting Information.
Uranyl titrations with all Me-3,2-HOPO ligands displayed
a rapid increase in intensity between pH 2 and pH 4, indicat-
ing deprotonation of the ligand and complexation of the
uranyl cation. For tetradentate ligands (with the exception
of L6H2), this result was refined as the formation of a
[UO2(L)
simultaneous deprotonation of both Me-3,2-HOPO moie-
ties, but is reasonable considering the similarity between
ACHTUNGTREN(NUNG solv)] complex. This assumption requires the
Most uranyl titrations were reversible through the highest
pH ranges of the titrations (typically pH 11.0–11.4), indicat-
ing that no unforeseen chemical changes occur in the metal–
ligand complex in solution (Figure 5). In contrast, titrations
AHCTUNGTRENNUNG
pKa1 and pKa2 of these ligands and because uranyl chelation
will drive deprotonation of a chelating ligand at lower
pH.[42] The UV/Vis spectra shifted again around neutral pH,
which was refined as the formation of a [UO2(L)(OH)]ꢀ
species (Scheme 2). This partial hydrolysis occurs near neu-
Scheme 2. Formation of a [UO2L(OH)]ꢀ species. Uranyl oxo atoms are
omitted for clarity.
tral pH, because the fifth equatorial coordination position
on the uranyl is known to be occupied by solvent in bis-Me-
3,2-HOPO complexes,[34,35] and thus hydroxide coordination
does not necessitate ligand displacement. A similar partial
hydrolysis event is observed in the uranyl complex with the
potentially hexadentate desferrioxamine B (DFO), in which
a [UO2ACHTNUGRTNEUNG
(DFO)(OH)]ꢀ complex was observed to form at neu-
tral pH.[28]
Because 2,5-disubstituted thiophene-Me-3,2-HOPO lig-
(solv)}2] complexes,[33] such di-
ACHUTNGRENUNaG nds form dimeric [{UO2(L)2ACHTUGNTRENNUGN
meric species were required in the refinement of titration
data with L6H2. Refinement of the strong acid titration data
was attempted two ways, assuming either that a mononu-
clear species [UO2(L6)H0/1]0/+ forms before a [{UO2(L6)2-
Figure 5. Representative reversibilities of uranyl titrations with bis-Me-
3,2-HOPO ligands through pH 11: L3H2, (top, reversible) and L6H2,
(bottom, irreversible).
AHCTUNGTERG(NNUN solv)}2] dimer, or that the dimer forms in a concerted fash-
ion, but only the latter strategy led to stable data refine-
ment. This behavior suggests that the substitution on the
thiophene ligand inhibits mononuclear complex formation
even at the low concentrations used, which prior crystallo-
graphic studies led us to expect.[33] Because both uranyl cat-
with fully conjugated ligands L6H2 and L7H2 exhibited poor
reversibility when titrations were taken to pH 11.0
(Figure 5). It was found that running the titrations only up
to pH 8.5 and 9.0, respectively, restored reversibility in the
uranyl titrations. The cause of this irreversibility is unknown,
but corresponds roughly to additional deprotonation of or
continued hydroxide introduction to the [UO2(L)(OH)]ꢀ
complex to form a [UO2(L)Hꢀ2]2ꢀ species (in which the ad-
ditional Hꢀ1 represents hydroxide coordination or incremen-
ions in the solid state [{UO2(L)2ACHTNUGRTENUNG(solv)}2] dimer crystal struc-
ture displayed pentagonal planar coordination with one site
occupied by solvent, the titration data were refined such
that the [{UO2(L6)
CHTUNGTRENNUNG
2A
hydrolysis events, forming first
{UO2(L6)(OH)}]ꢀ species, followed by a [{UO2(L6)(OH)}2]2ꢀ
a
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Chem. Eur. J. 2011, 17, 1818 – 1827
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1823

K. N. Raymond and G. Szigethy
each ligand displaying different
pKa values and uranyl complex-
ation behavior, a species-inde-
pendent method is needed to
compare uranyl affinities.
A
metric frequently used for this
purpose is that of pM (for this
study pUO2) in which pUO2 =
2+
2+
ꢀlog[UO2 free], and “UO2
free” refers to solvated uranyl
ion free of complexation by
ligand or hydroxide. Using this
metric, higher pUO2 values cor-
respond to a lower concentra-
tion of uncomplexed uranyl in
solution. The pUO2 values are
Scheme 3. Proposed speciation of uranyl titrations with L6H2. Uranyl oxo atoms are omitted for clarity.
species (Scheme 3). The onset of titration irreversibility at
pH 8.0
coincides
with
hydrolysis
beyond
the
[{UO2(L6)(OH)}2]2ꢀ species and is consistent with the onset
of irreversibility in the uranyl titrations with L7H2, which
was observed upon hydrolysis/deprotonation beyond the for-
mation of the [UO2(L7)(OH)]ꢀ species.
The speciation of uranyl complexes with bidentate L1H in
solution is necessarily very different from those with L2H2–
L8H2. Because L1H has only one proton, L1 binds the uranyl
ACTHNUTRGNEUNG
cation at very low pH to first form a [UO2(L1)(solv)x]+ com-
Scheme 4. Proposed speciation of uranyl titrations with L1H2. Uranyl oxo
atoms are omitted for clarity.
plex, which is the major species until pH 5.5, when the
2A
[UO2(L1)
CHTUNGTRENNUNG(solv)] complex, observed in crystal structure anal-
ysis,[35] becomes the dominant
species. This complex first
undergoes the expected partial
Table 3. Logbmlh values for uranyl titrations with Me-3,2-HOPO ligands.
logb11ꢀ2
logb11ꢀ1
logb110
logb12ꢀ1
logb120
logb22ꢀ1
logb22ꢀ1
logb220
hydrolysis
to
form
L1H[a]
L2H2
L3H2
L4H2
L5H2
ꢀ2.50(6)
–
10.64(3)
12.5(1)
12.6(1)
13.9(1)
13.42(7)
–
11.48(4)
-
17.91(2)
–
–
–
–
–
–
–
–
–
–
–
–
[UO2(L1)2(OH)]ꢀ, then experi-
ences one more hydrolysis
below pH 10. This last hydroly-
sis product could be refined as
–
–
–
–
–
–
–
6.30(8)
5.86(6)
6.97(6)
5.64(4)
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
[a]
L6H2
L7H2
L8H2
15.74(5)
–
–
22.47(8)
–
–
28.1(1)
–
–
either
a
[UO2(L1)Hꢀ2]ꢀ or
6.6(1)
5.38(5)
13.04(1)
12.89(1)
[UO2(L1)2Hꢀ2]2ꢀ species, with
near negligible changes in the
other formation constants. The
former corresponds to displace-
ment of one ligand upon coor-
[a]
[a] In addition to low pH data (pH 3 to pH 1.6), only forward (acid to base) titration data was used due to ob-
served irreversibility.
dination of a second hydroxide, while the latter corresponds
to either deprotonation of a coordinated ligand or the intro-
duction of hydroxide to the uranyl coordination plane with-
out ligand displacement. The lack of acidic protons on coor-
dinated Me-3,2-HOPO moieties rules out the possibility of
additional deprotonation. However, the coordination of ad-
ditional hydroxide would crowd the uranyl coordination
plane, and the low concentrations and 1:2 UO2/L1H ratios
used in the titrations suggest that ligand displacement to
form a [UO2(L1)(OH)2]ꢀ species at high pH is the more
likely speciation in these titrations (Scheme 4).
calculated by using standard conditions of [UO22+]=10ꢀ6
m
and [L]=10ꢀ5 m (L:UO22+ =10), and thus the minimum
pUO2 value is 6.0, at which no metal complexation occurs.
While typically reported at physiological pH, pUO2 can be
calculated at any pH once proton and uranyl affinities are
known; pUO2 values at pH 2.5, 7.4 and 8.5 are listed for
each Me-3,2-HOPO ligand in Table 4.
The pUO2 values in Table 4 reveal that the uranyl affinity
of bis-HOPO ligands does indeed vary with changes in
ligand geometry. However, the affinities do not follow grad-
ual trends of the sort observed in Table 2, in which linker
length and degree of conjugation affected ligand proton af-
finity in an incremental fashion. In contrast, small changes
in linker length and geometry resulted in large changes in
Titration data analysis using the speciation models de-
scribed above yielded the formation constants listed in
Table 3. Because logbmlh values are species dependent, with
1824
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1818 – 1827

Ligand effects
FULL PAPER
[a]
Table 4. Calculated pUO2 values for Me-3,2-HOPO ligands.
finity is presumably due to a favorable geometric agreement
between the ligand and the uranyl coordination preferences.
The pUO2 of L7H2 is also higher than that of L2H2 (both n=
2), indicating that ligand rigidity and appropriate preorgani-
zation favor high uranyl affinity, as expected. In contrast,
the pUO2 of L8H2 is significantly lower, despite the favor-
pH 2.5
pH 7.4
pH 8.5
L1H
7.98(3)
7.0(1)
6.6(1)
8.0(1)
7.1(1)
6.01(1)
7.62(4)
6.55(6)
2+
14.70(4)
14.63(8)
14.24(7)
15.39(7)
14.44(6)
13.39(3)
14.97(9)
14.00(3)
15.96(3)
15.75(8)
15.34(6)
16.43(9)
15.16(4)
14.48(2)
16.1(1)
L2H2
L3H2
L4H2
L5H2
L6H2
L7H2
L8H2
AHCTUNGTREGaNNNU ble conformational parameters observed in crystal struc-
tures of the unsubstituted [UO2(m-xy-Me-3,2-HOPO)ACHTUNGTRENNUNG(dmf)]
structure.[34]
14.87(5)
The high pUO2 of L1H, which rivals that of L4H2 at
pH 2.5 and L2H2 at high pH, can be understood by consider-
ing that uranyl chelation by L1H at low pH requires a single
deprotonation event, while chelation of bis-Me-3,2-HOPO
ligands requires two, making (L1)ꢀ coordination relatively
favorable at low pH. At high pH, the speciation model with
L1H also reveals that the relatively high affinity is accompa-
nied by ligand displacement in favor of hydroxide ion. Thus,
although L1H exhibits a high pUO2, the individual Pr-Me-
3,2-HOPO ligands display significantly less affinity towards
the uranyl cation than the bis-Me-3,2-HOPO ligands, as the
chelate effect would lead one to expect.
[a] pUO2 =ꢀlog[UO2 free], [UO22+]=1 mm, [L]=10 mm.
pUO2, with no noticeable correlation with the physical met-
rics gleaned from crystal-structure analysis.[34] Common to
all ligands, however, is a dramatic rise in pUO2 (about 7 log
units) between pH 2.5 and 7.4. It is not possible to credit
this pUO2 increase entirely to a rise in ligand affinity, be-
cause concentrations of uranyl hydrolysis products
[(UO2)m(OH)n](2mꢀn)+ increase upon increased pH, lowering
2+
[UO2 free] independent of ligand identity or efficacy. How-
ever, because Me-3,2-HOPO ligands require deprotonation
for metal chelation, it is reasonable to suspect that the ma-
jority of this rise in pUO2 comes from a change in ligand af-
finity upon complete deprotonation at higher pH. This as-
sumption is also supported by the UV/Vis titration spectra,
which exhibit significant shifts in lmax and absorption intensi-
ty upon increasing pH from 2.5 to 7.4, suggesting increased
ligand chelation of the uranyl cation.
Conclusion
Seven PEG-substituted bis-Me-3,2-HOPO ligands were syn-
thesized as water-soluble, structural analogues to ligands
subjected to prior crystallographic analysis.[34,35] Solution
thermodynamic measurements demonstrated that linker
length and geometry have a measurable effect on the proton
affinity of bis-Me-3,2-HOPO ligands, but that uranyl affinity
of these ligands does not correlate with trends in proton af-
finity. Significantly, the PEG-4li-Me-3,2-HOPO ligand L4H2
bound most strongly to the uranyl cation, which supports
earlier observations from in vivo chelation studies and crys-
tallographic studies.[31,35] The o-phenylene linker in L7H2 im-
parted the second highest uranyl affinity, rivaling that of
L4H2, and had a higher affinity than the similarly sized L2H2.
Because these results are generally independent of ligand
acidity, it can be concluded that ligand geometry imparted
by the linker structure (and perhaps ligand preorganization
in the case of L7H2) is responsible for the observed uranyl
affinities.
One aspect the uranyl titrations and the resultant pUO2
values described above do not address is the matter of selec-
tivity. In most applications associated with biological remov-
al of actinides, high binding constants are necessary, but
high selectivity is also desirable if an administered drug is to
be effective. Such selectivity is also necessary in industrial
extraction processes, so further evaluation of bis-Me-3,2-
HOPO ligands will require titration measurements against
biologically relevant ions such as Ca2+ and Zn2+, as well as
other actinides.
Focusing on ligands L2H2–L5H2, there is no trend in
uranyl affinity upon incremental increase in linker length at
any pH. Because ligand pKa decreases as linker length
shortens, L2H2 would be expected to bind most strongly to
the uranyl cation. However, L2H2 displays the second high-
est affinity, with L4H2 displaying the strongest uranyl affinity
at both low and high pH. The ligand L4H2 has a higher SpKa
than both L2H2 and L3H2, so the observed uranyl affinity
must be due to geometry effects. This high affinity is con-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
with the superior ability for 4li-Me-3,2-HOPO to chelate ac-
tinides in vivo.[30,35]
Of the rigidly-linked ligands L6H2–L8H2, L6H2 is the most
acidic, and yet displays the poorest uranyl affinity at all pH.
This may be due to the inability to form monomeric uranyl
complexes. The low pUO2 of L6H2 at higher pH is presumed
to be a geometric effect caused by the amide rotation neces-
sitated by substitution at the 2- and 5-positoins of the thio-
phene ring.[33] Unfortunately, such distortions are un-
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
substituted 3,4-pyrrole-Me-3,2-HOPO ligands remains
question for future study.
a
Ligand L7H2 shows the most favorable pUO2 of the rigidly
linked ligands, approaching that of L4H2. Because minor
pKa changes are apparently not a significant factor in deter-
mining uranyl affinity in bis-Me-3,2-HOPO ligands, this af-
Chem. Eur. J. 2011, 17, 1818 – 1827
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1825

K. N. Raymond and G. Szigethy
pHab.[48] Data from three independent titrations consisting of one for-
ward and one reverse titration were analyzed separately, with the forward
and backward runs refined separately. Both analysis programs utilized
nonlinear least squares regression to determine formation constants.
Experimental Section
Titration solutions and equipment: Corning high-performance combina-
tion glass electrodes (response to [H+] was calibrated before each titra-
tion[43]) were used together with either an Accumet pH meter or a Met-
rohm Titrino to measure the pH of the experimental solutions. Metrohm
autoburets (Dosimat or Titrino) were used for incremental additions of
acid or base standard solution to the titration cell. The titration instru-
ments were fully automated and controlled using LabVIEW software.[44]
Titrations were performed in 0.1m KCl supporting electrolyte under posi-
tive Ar gas pressure. The temperature of the experimental solution was
maintained at 258C by an externally circulating water bath. UV/Vis spec-
tra for incremental titrations were recorded on a Hewlett–Packard 8452a
spectrophotometer (diode array). Solid reagents were weighed on a Met-
rohm analytical balance accurate to 0.01 mg. All titration solutions were
prepared using distilled water that was purified by passing through a Mil-
lipore Milli-Q reverse osmosis cartridge system and degassed by boiling
for 1 h, while being purged under Ar. Carbonate-free 0.1m KOH was
prepared from Baker Dilut-It concentrate and was standardized by titrat-
ing against potassium hydrogen phthalate using phenolphthalein as an in-
dicator. Solutions of 0.1m HCl were similarly prepared and were stand-
Values for the hydrolysis product of the uranyl cation were taken from a
recent literature publication.[49] Wavelengths between 250–400 nm were
typically used for data refinement of spectrophotometric titrations. The
number of absorbing species to be refined upon was determined by
factor analysis within the pHab program suite.[48] Reversibility of spectro-
photometric titrations was determined by comparison of the species- and
concentration-independent value Av (absorbanceꢂvolume) at selected
wavelengths for the forward and reverse titrations. Speciation diagrams
were generated using HYSS2[50,51] titration simulation software and the
protonation and metal complex formation constants were determined by
potentiometric and spectrophotometric titration experiments.
Acknowledgement
We would like to thank Drs. Rebecca Abergel and Trisha Hoette for as-
sistance with the titrations measurement and data treatment. This re-
search is supported by the Director, Office of Science, Office of Basic
Energy Sciences, and the Division of Chemical Sciences, Geosciences,
and Biosciences of the U.S. Department of Energy at LBNL under Con-
tract No. DE-AC02–05CH11231.
ACHTUNGTRENNUNGardized by titrating against sodium tetraborate decahydrate to Methyl
Red endpoint. Ligand stock solutions were made by dissolving a weighed
amount of ligand accurate to 0.01 mg in DMSO in a volumetric flask.
These stock solutions were frozen when not in use to prevent ligand de-
composition. A stock uranyl solution in 1.2 wt% nitric acid was pur-
chased from Aldrich (4.22 mm) and used as received. UO2:L ratios were
controlled by careful addition of uranyl and ligand solutions to the titra-
tion vessel.
[1] ꢃ. Ansoborlo, B. Amekraz, C. Moulin, V. Moulin, F. Taran, T.
Bailly, R. Burgada, M.-H. Hengꢄ-Napoli, A. Jeanson, C. Den
Auwer, L. Bonin, P. Moisy, C. R. Chim. 2007, 10, 1010–1019.
[2] S. I. Mꢅller, M. Valdebenito, K. Hantke, Biometals 2009, 22, 691–
695.
[3] M. A. Fischbach, H. Lin, D. R. Liu, C. T. Walsh, Nature Chem. Bio.
2006, 2, 132–138.
[4] O. Andersen, Chem. Rev. 1999, 99, 2683–2710.
[5] D. F. Back, ꢃ. Bonfada, M. Manzoni de Oliveira, E. S. Lang, J.
Inorg. Biochem. 2007, 101, 709–714.
Spectrophotometric and potentiometric titrations: All titrations were per-
formed with about 5% starting concentration of DMSO added to the
KCl solution to promote the solvation of protonated ligands and their
neutral uranyl complexes. Spectrophotometric titrations were carried out
in the presence of 10–20 equivalents (as compared to ligand concentra-
tion) of NH4Cl, MES, and HEPES buffers in order to dampen the pH
change between incremental additions of titrant. Each addition of acid or
base was followed by an equilibration period before pH and absorbance
data were collected. For potentiometric titrations this delay was 300 s
and for spectrophotometric titrations was 600 s for free ligand and 600–
1200 s for titrations in the presence of UO22+. Spectra were recorded be-
tween 250–550 nm. The UV-silent region above 420 nm was monitored
for baseline drift as an indication of precipitated material.
[6] M. D. Cappellini, P. Patoneri, Annu. Rev. Med. 2009, 60, 25–38.
[7] R. C. Hider, Biochem. Soc. Trans. 2001, 30, 751–754.
[8] G. J. Kontoghiorghes, K. Neocleous, A. Kolnagou, Drug Saf. 2003,
26, 553–584.
Potentiometric titrations were carried out at 150–200 mm ligand concen-
trations. Spectrophotometric titrations for protonation constants of 2li-
Me-3,2-HOPO and 4li-Me-3,2-HOPO were carried out at about 50 mm
ligand concentrations using a titration cell with a 1 cm path length quartz
cuvette attached. Spectrophotometric titrations for uranyl affinity were
performed at about 6 mm complex concentration using a 6.6 cm path
length quartz-windowed titration cell. All mid-pH titrations were repeat-
ed a minimum of three times and were run forwards and backwards
(from acid to base and reverse) within limits set by the reversibility of
the titration.
[9] J. Maynadiꢄ, J.-C. Berthet, P. Thuery, M. Ephritikhine, Chem.
Commun. 2007, 486–488.
[10] J.-C. Berthet, P. Thuery, J.-P. Dognon, D. Guillaneux, M. Ephriti-
khine, Inorg. Chem. 2008, 47, 6850–62.
[11] C. J. Burns, D. L. Clark, R. J. Donohoe, P. B. Duval, B. L. Scott,
C. D. Tait, Inorg. Chem. 2000, 39, 5464–5468.
[12] N. N. Rammo, K. R. Hamid, T. K. Ibrahim, J. Alloys Compd. 1994,
210, 319–324.
[13] M. J. Sarsfield, M. Helliwell, J. Raftery, Inorg. Chem. 2004, 43,
3170–3179.
Spectrophotometric uranyl titrations of L2H2–L8H2 were performed at ap-
proximately 5 mm concentrations using a 6.6 cm path length quartz-win-
dowed titration cell with a 5% starting DMSO concentration necessary
for complete salvation of neutral uranyl species near neutral pH. Three
independent titrations, each consisting of a forward (acid to base) and re-
verse (base to acid) titration, were measured between pH 2.4 and 11.0,
except for cases in which reversibility analysis indicated a point in the ti-
trations beyond which the complexes underwent an irreversible chemical
change. Low pH titrations were performed from pH 3 down to pH 1.6,
and electrodes were calibrated as described in a previous publication.[45]
[14] D. H. Templeton, A. Zalkin, H. Ruben, L. K. Templeton, Acta Crys-
tallogr. C 1985, 41, 1439–1441.
[15] P. Thuꢄry, B. Masci, M. Takimoto, T. Yamato, Inorg. Chem.
Commun. 2007, 10, 795–799.
[16] W. Zhang, J. Zhao, Inorg. Chem. Commun. 2006, 9, 397–399.
[17] B. Masci, P. Thuꢄry, CrystEngComm 2007, 9, 582–590.
[18] J. A. Danis, M. R. Lin, B. L. Scott, B. W. Eichhorn, W. H. Runde,
Inorg. Chem. 2001, 40, 3389–3394.
[19] P. Thuꢄry, B. Masci, Dalton Trans. 2003, 2411–2417.
[20] I. Grenthe, J. Drozdzynski, T. Fujino, E. C. Buck, T. E. Albrecht-
Schmitt, S. F. Wolf in The Chemistry of the Actinide and Transacti-
nide Elements Vol. 1, 3rd ed. (Eds.: L. R. Morss, N. M. Edelstein, J.
Fuger), Springer, Dordrecht, The Netherlands, 2006, pp. 253–698.
[21] J. L. Sessler, D. Seidel, A. E. Vivian, V. Lynch, B. L. Scott, D. W.
Keogh, Angew. Chem. 2001, 113, 611–614; Angew. Chem. Int. Ed.
2001, 40, 591–594.
Titration data treatment: Potentiometric data were analyzed using Hy-
perquad[46,47] software. The data from at least three independent titra-
tions, each consisting of one forward (pH 3 to pH 10) and one reverse
(pH 10 to pH 3) titration were analyzed separately. These analyses result-
ed in similar values, so were refined together to provide better fit statis-
tics. Spectrophotometric titration data were analyzed using the program
1826
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 1818 – 1827

Ligand effects
FULL PAPER
[22] T. S. Franczyk, K. R. Czerwinski, K. N. Raymond, J. Am. Chem. Soc.
1992, 114, 8138–8146.
[38] A. DꢆAlꢄo, J. Xu, E. G. Moore, C. J. Jocher, K. N. Raymond, Inorg.
Chem. 2008, 47, 6109–6111.
[23] P. H. Walton, K. N. Raymond, Inorg. Chim. Acta 1995, 240, 593–
601.
[39] Ed.: N. R. Council), National Research Council, Washington, DC,
2000, p. 142.
[24] P. L. Arnold, A. J. Blake, C. Wilson, J. B. Love, Inorg. Chem. 2004,
43, 8206–8208.
[40] Ed.: N. R. Council), National Research Council, Washington, DC,
2001, p. 198.
[25] P. L. Arnold, D. Patel, A. J. Blake, C. Wilson, J. B. Love, J. Am.
Chem. Soc. 2006, 128, 9610–9611.
[41] J. Xu, P. W. Durbin, B. Kullgren, K. N. Raymond, J. Med. Chem.
1995, 38, 2606–2614.
[26] J. N. Mathur, M. S. Murali, K. L. Nash, Solv. Extr. Ion Exch. 2001,
19, 357–390.
[42] Z. Szabꢇ, I. Grenthe, Inorg. Chem. 2000, 39, 5036–5043.
[43] P. Gans, B. OꢆSullivan, Talanta 2000, 51, 33–37.
[27] A. P. Paiva, P. Malik, J. Radioanal. Nucl. Chem. 2004, 261, 485–496.
[28] L. Mullen, C. Gong, K. R. Czerwinski, J. Radioanal. Nucl. Chem.
2007, 273, 683–688.
[29] P. W. Durbin, Health Phys. 2008, 95, 465–492.
[30] P. W. Durbin, B. Kullgren, S. N. Ebbe, J. Xu, K. N. Raymond, Health
Phys. 2000, 78, 511–521.
[44] LABVIEW, version 8.0; National Instruments Corp.: 11500
Mopac Expwy, Austin, TX 78759–3504.
[45] A. R. Johnson, B. OꢆSullivan, K. N. Raymond, Inorg. Chem. 2000,
39, 2652–2660.
[46] P. Gans, A. Sabatini, A. Vacca, Talanta 1996, 43, 1739–1753.
[47] P. Gans, A. Sabatini, A. Vacca, HYPERQUAD2003, University of
Leeds and University of Florence, Leeds, U.K., and Florence, Italy,
2003.
N
[31] P. W. Durbin, B. Kullgren, J. Xu, K. N. Raymond, Radiat. Prot.
Dosim. 1998, 79, 433–443.
[32] A. E. V. Gorden, J. Xu, K. N. Raymond, Chem. Rev. 2003, 103,
4207–4282.
[48] P. Gans, A. Sabatini, A. Vacca, Ann. Chim. 1999, 89, 45–49.
[49] P. L. Brown, Radiochimica Acta 2002, 90, 589–593.
[50] L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini, A. Vacca,
Coord. Chem. Rev. 1999, 184, 311–318.
[33] G. Szigethy, K. N. Raymond, Inorg. Chem. 2009, 48, 11489–11491.
[34] G. Szigethy, K. N. Raymond, Inorg. Chem. 2010, 49, 6755–6765.
[35] J. Xu, K. N. Raymond, Inorg. Chem. 1999, 38, 308–315.
[36] P. Thuꢄry, N. Keller, M. Lance, J. D. Figner, M. Nierlich, New J.
Chem. 1995, 19, 619–625.
[51] P. Gans, HySS2; University of Leeds, Leeds, U. K., 2000.
[37] T. M. Garrett, M. E. Cass, K. N. Raymond, J. Coord. Chem. 1992,
25, 241–253.
Received: August 17, 2010
Published online: January 5, 2011
Chem. Eur. J. 2011, 17, 1818 – 1827
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1827