Equilibrium and structure of the Al(III)-ethylenediamine-N,N′-bis(3- hydroxy-2-propionate) (EDBHP) complex. A multi-method study by potentiometry, NMR, ESI MS and X-ray diffraction

Article abstract of DOI:10.1039/b517192j

The equilibrium and structure of the complex formed by Al(iii) and ethylenediamine-N,N′-bis(3-hydroxy-2-propionate) (EDBHP^2-) have been studied using pH-potentiometry, ¹H and ²⁷Al NMR, ESI MS and single crystal X-ray diffraction methods. The EDBHP ligand is a strong Al-binder in aqueous solution for pH between 4 and 8 and for c_Al = c_EDBHP ≥ 0.1 mmol dm^-3. The dominating complex identified by ESI MS and potentiometry is a neutral dimer, Al₂L ₂(OH)₂, with logβ_22-2 = 14.16 ± 0.03. In the solid Al₂(EDBHP)₂(OH)₂· 2H₂O the Al(iii) ions are connected through a double hydroxo bridge. Both four-dentate organic ligands are coordinated terminally through two carboxylate groups and two N-donors forming three five-membered chelate rings. The hydroxyl groups of the ligand EDBHP remain protonated and are not coordinated to the aluminium ions. The structure and composition of the dimer are very likely the same in solution and the solid state. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b517192j

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
ꢀ
Equilibrium and structure of the Al(III)-ethylenediamine-N,N -bis
(
3-hydroxy-2-propionate) (EDBHP) complex. A multi-method study
by potentiometry, NMR, ESI MS and X-ray diffraction†
a
a
b
b
a
a
R o´ bert J o´ szai, Imola Kerekes, Igarashi Satoshi, Kiyoshi Sawada, L a´ szl o´ Z e´ k a´ ny and Imre T o´ th*
Received 5th December 2005, Accepted 10th March 2006
First published as an Advance Article on the web 6th April 2006
DOI: 10.1039/b517192j
ꢀ
The equilibrium and structure of the complex formed by Al(III) and ethylenediamine-N,N -bis
2
−
1
27
(
3-hydroxy-2-propionate) (EDBHP ) have been studied using pH-potentiometry, H and Al NMR,
ESI MS and single crystal X-ray diffraction methods. The EDBHP ligand is a strong Al-binder in
−
3
aqueous solution for pH between 4 and 8 and for cAl = cEDBHP ≥ 0.1 mmol dm . The dominating
complex identified by ESI MS and potentiometry is a neutral dimer, Al (OH) , with logb22−2
(EDBHP) (OH) O the Al(III) ions are connected through a double
2
L
2
2
=
14.16 ± 0.03. In the solid Al
2
2
2
·2H
2
hydroxo bridge. Both four-dentate organic ligands are coordinated terminally through two carboxylate
groups and two N-donors forming three five-membered chelate rings. The hydroxyl groups of the ligand
EDBHP remain protonated and are not coordinated to the aluminium ions. The structure and
composition of the dimer are very likely the same in solution and the solid state.
Introduction
several metal ions, Cu(II), Ni(II), Co(II), Pb(II), Cd(II), Zn(II), and
Fe(III), using potentiometry and UV-VIS spectrophotometry, but
5
EDTA-type ligands are widely used in industry and analytical
chemistry. These ligands are quite stable, i.e. they do not decom-
pose easily in natural systems and when they do it is only slowly.
The accumulation of these chelating ligands in natural water is
anticipated to disturb the ecosystem by causing metal deficiency.
Therefore, there is an urgent need to develop biodegradable
analogues of aminopolycarboxylates in order to decrease the
no data are available for Al(III) in the literature.
Aluminium(III) usually forms very stable complexes with O-
6
donor ligands. In the presence of charged anchor group(s), like
carboxylate(s), the alcoholic hydroxyl groups can be co-ordinated
to Al(III) via deprotonation of the hydroxyl-group even in the acidic
pH-range. These species tend to form multinuclear complexes.
7
Modern mass-spectrometry, especially using electro-spray ion-
ization techniques, is becoming a very powerful tool for studying
1
environmental impact ofthese compounds. Several biodegradable
EDTA-derivatives have been synthesised and characterised, e.g.
8
coordination chemistry in solutions, but further studies are
succinic acid and the malonic acid derivatives of ethylenediamine
needed to explore the required instrument conditions, as well as
the advantages and limitations of this method for quantitative
analysis.
2–4
are good metal binders, but further work is required to find
suitable environmental-friendly ligands.
ꢀ
Ethylenediamine-N,N -bis(3-hydroxy-2-propanoic acid) (H
2
ED-
2−
+
We report here results for the Al(III)–EDBHP –H system. The
aim of this study is to elucidate the equilibrium and the structure
BHP) (see Fig. 1) forms stable complexes with alkaline earth and
transition metal ions. Dvorakova and her co-workers have studied
the acid–base and complex-forming features of this ligand with
1
both in solution and for the solids using pH-potentiometry, H
27
and Al NMR, electrospray ionization mass spectroscopy and
single-crystal X-ray diffraction methods.
Results and discussion
ꢀ
Ethylenediamine-N,N -bis(3-hydroxy-2-propanoic acid) (H
2
ED-
BHP or H L) (Fig. 1) undergoes consecutive protonation/
2
deprotonation steps, n = 1, 2, 3 and 4:
ꢀ
Fig. 1 Ethylenediamine-N,N -bis(3-hydroxy-2-propanoic acid) (H
2
ED-
2
−
+
n−2
L
+ nH ꢀ H
n
(1)
(2)
BHP). The letters a, b and c indicate three different (covalently C-bonded)
H-sites of the ligand.
n−2
2−
+ n
b
n
= [H
n
L
]/[L ][H ]
The titration curves for the ligand show two well-separated steps.
a
The Z vs pH curves (Fig. S1, see ESI†) can be described by these
protonation steps, see eqn (1) and (2):
Department of Inorganic and Analytical Chemistry, University of Debrecen,
L
Hungary. E-mail: imretoth@delfin.unideb.hu
b
Department of Chemistry, Faculty of Science, Niigata University, Japan
+
−
c
H
− ([H ] − [OH ])
†
Electronic supplementary information (ESI) available: [DETAILS]. See
Z
L
=
(3)
DOI: 10.1039/b517192j
c
L
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3221–3227 | 3221

View Article Online
+
ꢀ
where c
H
is the total concentration of H ions; the equilibrium
be considered as b and b , although the two sets of triplet signals
are overlapping over the whole pH-range. (Spectra recorded at
360 MHz show similar features, but slight differences clearly
indicate that the spectra are not identical under different fields,
therefore indicating chemical non-equivalence of the H-atoms, i.e.
+
+
−pH
concentration of the H ion is [H ] = 10 ; c
L
is the total
is the ionic
−
logKw+pH
concentration of the ligand; [OH ] = 10
, K
w
product of water. The calculated constants are logb
1
= logK
) = 15.43 ± 0.05 (K , K
the stepwise protonation constants, logK
= 6.25 ± 0.02). The
third and fourth protonation constants, do not play any role
in this study, logK = 1.7 ± 0.4 could
= 2.3 ± 0.2 and logK
also be estimated from the curves. These values are in reasonable
1
=
9.18 ± 0.01 and logb
2
= log(K
1
K
2
1
2
are
ꢀ
ꢀ
ꢀ
2
a and a ; b and b ; c and c should be considered.) Line width values
are changing as well. The broadening is most pronounced at pH ∼
−
2−
3
4
8.7–9.8, i.e. in a region where HL and L co-exist in comparable
concentration. We cannot give a quantitative description of the
spectral change, but it should be attributed to a less symmetric
structure at high pH compared to a lower one. An obvious
reason for the lowered symmetry and/or lowered fluxionality is the
formation of H-bond(s) with the hydroxyl group(s). Fig. 3 shows
two possibilities: (a) five membered ring(s) involving the N-atom
and the neighbouring alcoholic OH-group, and (b) six-membered
ring(s) consisting of the carboxylate group and the neighbouring
agreement with those available in the literature (logK
1
= 9.20,
◦
logK
2
= 6.07, logK
3
= 2.1 and logK
4
= 1.6 measured at 20 C in
−
3
5
0
.1 mol dm KNO ).
3
1
The effect of the pH on the ligand can also be detected by H
NMR through the well known dependence of chemical shift on
pH, i.e. an up-field shift can be attributed to deprotonation. The
change in shift is the smallest for site a since it is relatively far from
the protonation sites, and larger for both b and c sites. b and c sites
are almost equally sensitive to both deprotonation steps.
−
alcoholic OH-group (the HL has only one protonated N-group,
but the symmetry is not lowered, because of the fast proton transfer
from one N atom to another). Dissociation of proton(s) from the
Moreover, another feature of the spectra also deserves attention.
Magnetically non-equivalent H
A
and H
B
protons are in all of
2
H EDBHP might be required for the H-bond(s) formation, but
the –CH groups in the symmetric ligand itself because of the
2
the rate of a rearrangement might also be pH-dependent. Hence,
both the strength and the fluxionality of the intramolecular ring(s)
might be affected by pH-change. A detailed study of the dynamics
of the supposed H-bonded structures is beyond the scope of this
paper.
The complex formation in the Al(III)–EDBHP –H system has
been followed by pH-potentiometry. A typical ZAl–pH curve is
shown in Fig. 4.
proton couplings. Additional splitting should be observed if a
decrease of the symmetry results in chemical non-equivalence of
ꢀ
ꢀ
ꢀ
the –CH -groups (a and a ; b and b ; c and c ). As one can see in
2
Fig. 2, the line shape for of all three proton signals changes with
pH. The most dramatic change appears for site c, i.e. for the H-
atoms of the ethylene-group. The apparent singlet signal at pH =
2
−
+
4
.33 (d ≈ 3.65 ppm, 1) splits to a quite complicated multiplet of c
ꢀ
ꢀ
and c for pH > 7.01 (d ≈ 3.1 ppm, 2). The a and a sites appear
already at pH = 4.33 having two sharp, overlapping sets of AB
doublets, 3. The separation of these signals, 4 is almost complete
at pH = 8.7, and poor at pH = 10.51, 5. Site b protons should also
+
−
c
H
− ([H ] − [OH ])
Z
Al
=
(4)
c
Al
1
= 10 mmol dm−3 in D
Fig. 2 500 MHz H NMR spectra of EDBHP at different pH values. c
L
2
O. (See Fig. 1 for a, b and c and the text for numbers 1–5).
3
222 | Dalton Trans., 2006, 3221–3227
This journal is © The Royal Society of Chemistry 2006

View Article Online
to merge these two models have failed. Unfortunately, there is
no easy way to distinguish between the models using chemical
10
analogy. One can consider the multidentate EDBHP as an EDTA
analogue, therefore the formation of a 1 : 1 complex can be
expected. However, the bridging ability of the hydroxycarboxylic
acids envisages the formation of bi- or polynuclear complexes, i.e.
one can argue for the model with the dimer. At the same time, both
of the models show clearly that the EDBHP ligand can strongly
suppress the formation of Al(III)-hydroxo complexes up to pH ≈ 8
for our experimental conditions, thus this ligand is a good Al(III)-
binder in the range of pH = 4–8. Fig. 5 shows typical distribution
curves using both models. A remarkable feature of the diagrams
is that the fraction of Al(III) decreases from 1 to 0 over a very
narrow range of pH (3.5 to 4.8). One can expect this behaviour
from primary data; see Fig. 4, showing a decrease of ZAl by almost
3
in this region. The sharp change in the curve could be attributed
to the almost simultaneous coordination of the ligand and the
hydrolysis of Al(III). The increased acidity of water molecules in
6,7
the AlL
x
2
y
(H O) system is well documented in the literature.
Fig. 3 Possible intramolecular H-bonded rings of EDBHP: (a) five-
2−
−
membered rings in L ; (b) six-membered rings in HL .
Fig. 4
ZAl vs. pH according to eqn (4). Titration curve of the
2−
+
−3
−3
Al(III)–EDBHP –H system at cAl = 5.00 × 10 mol dm and c =
L
.00 × 10− mol dm . Symbols (᭺) represent measured points, the line is
3
−3
6
calculated by the fitted constants, see Fig. 5a.
where cAl is the total concentration of Al(III). (ZAl gives the ratio
between the averaged number of bonded protons in the system
and Al(III).)
2
−
+
Equilibria in the Al(III)–EDBHP –H system can be repre-
sented by the following general equation:
pAl + qL2− + rH ꢀ Al
3
+
+
+3p−2q+r
q r
L H
(5)
(6)
p
Fig. 5 Distribution of Al(III)-containing species for cAl = cAl = 5 ×
−
3
−3
1
0
mol dm with logb110 = 9.3, logb22−2 = 14.16 and logb11−2 = −2.71,
+
3p− 2q+r
3+
p
2−
q
+ r
b
pqr = [Al
p
L
q
H
r
]/[Al ] [L ] [H ]
27
symbols (᭺) represent the fraction of Al(III) measured by Al NMR (a);
9
and with logb
= 5.86 and logb11−2 = −2.76 (b).
The fitting procedure by the PSEQUAD program has resulted
in two, more or less equally good, models. The best fit can
be attributed to a model consisting of a neutral, AlLH−1 and
11−1
0
It is worth comparing the stability constants of Al–EDBHP
complexes for the monomeric species to the constants of
−
an anionic, AlLH−2 , species with logb11−1 = 5.86 ± 0.03 and
−
11
logb11−2 = −2.76 ± 0.10, respectively. The other model, with
AlEDTA (logb110 = 16.7 and logb11−1 = 10.9). The value
of logb110 = 9.3 ± 0.1 is smaller by 7 orders of magnitude
practically equally good-fitting parameters, consists of a cationic
+
0
−
monomer, AlL and a dominating neutral dimer, Al
2
L
2
H
−2 and
than the stability constant of AlEDTA , in accordance with
expectations for a four-dentate EDBHP versus a six-dentate
EDTA . However, the value of pK
−
2−
AlLH−2 with logb110 = 9.3 ± 0.1, logb22−2 = 14.16 ± 0.03, and
+
−
4−
logb11−2 = −2.71 ± 0.10, respectively. The AlL and AlLH−2 are
minor species formed over narrow pH-ranges, therefore the larger
uncertainties of logb110 and logb11−2 can be accepted. All attempts
1
= 3.4 in the case of the
+
+
EDPHP system (i.e. for AlL + H
2
O ꢀ AlL(OH) + H ) shows
+
−
the larger acidity of AlL compared to AlEDTA (pK
1
= 4.8),
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3221–3227 | 3223

View Article Online
27
2−
+
= 5.00 × 10−3 mol dm−3
Fig. 6 104.2 MHz Al NMR spectra of Al(III)–EDBHP –H system at different pH values, cAl = c
L
.
moreover the AlL(OH) can be further hydrolysed (AlL(OH) +
Al
Al
2
L
(OH)
2
(OH)
2
·2H
2
O (Fig. 7, Tables 1 and 2). The double bridged
(OH) O is similar to
·2H
−
+
4+
H
2
O ꢀ AlL(OH)
2
+ H ) with pK
2
≈ 8.3. This process is not
2
2
structural motif of Al
2
2
L
2
2
2
−
known for Aledta(OH) . In fact, the Al(III)-binding capacity
of EDPHP is very good, and it can be attributed to the high
overall stability of the 11−1 and 11−2, or 22−2 species. It can
12
be visualized by the typical distribution curves shown in Fig. 5.
3
Model calculation (for cAl = c
L
= 0.1 mmol dm− and involving
the most important aluminium-hydroxo species) clearly shows that
EDBHP can also suppress the formation of binary Al(III)-hydroxo
complexes in such a diluted solution.
NMR measurements support the equilibrium model(s). How-
1
ever, H NMR spectra (see Fig. S2 in ESI†) serve only as
‘fingerprints’, being very complicated for both the free ligand and
the Al–ligand complex, but the appearance of new signals among
the signals of the free ligand, and the change of spectra with pH
is in accordance with the pH-potentiometry.
2
7
Al NMR measurements have been found especially useful
for following the sharp decline of the free Al(III). Fig. 6 shows
the sharp peak of Al(III) at 0 ppm and only one other broad
signal at about 25–30 ppm up to pH = 7. The position of this
broad signal indicates an octahedral coordination of the donor
2 2 2
Fig. 7 ORTEP view of the complex, Al (EDBHP) (OH) at the 50%
probability level. The asterisked atoms (*) are related to unasterisked ones
by an inversion center (−x, −y, −z). There are three independent dimers
in the unit cell and only one of them is shown in here.
13
atoms of amino-carboxylate ligands. Unfortunately, integration
of these broad signals is too uncertain, which is why these values
cannot be quantitatively compared to the distribution diagram.
For similar reasons there is no chance to detect minor hydrolyzed
Al(III)-species that might be formed. However, the decrease of the
measured intensity for the Al NMR signal of Al(III) at d = 0
follows the fraction of the free aluminium ion calculated from
the pH-potentiometry quite well (see Fig. 5a, (᭺) symbols). This
Table 1 Crystal data and structure refinement parameters for
Al
2
(EDBHP)
2
(OH)
2
·2H
2
O
Empirical formula
Formula weight
Crystal size/mm
Crystal system
16 2 4 16
C H34Al N O
592.42
2
7
0.26 × 0.24 × 0.12
Triclinic
¯
P1
Space group
a/A˚
11.292(1)
11.959(2)
14.636(2)
73.682(7)
88.008(7)
82.015(6)
1878.5(4)
3
finding strongly supports the adequacy of the model. Increasing
b/A˚
c/A˚
−
the pH further, the signal of Al(OH)
4
(not presented in Fig. 6) at
◦
8
0 ppm appears, showing the release of Al(III) from the EDBHP
a/
b/
◦
complex in accordance with the calculated speciation.
◦
c /
The issue of the real stoichiometry for the AlLH−1 (i.e. hydrolysis
of Al(III) or deprotonation of the hydroxyl group) and the dilemma
of monomer/dimer in solution have not been unequivocally
answered by potentiometry and NMR. On the other hand, the
X-ray structure of selected single crystals gives a definite answer
to these two questions. There is an interesting dimeric structure of
V/A˚
Z
3
−
3
Dcalc/g cm
1.571
2.01
−
1
l/cm
R
1
0.0885
0.2662
wR
2
3
224 | Dalton Trans., 2006, 3221–3227
This journal is © The Royal Society of Chemistry 2006

View Article Online
Table 2 Selected structural data of Al
2
(EDBHP)
2
(OH)
2
·2H
2
O and sev-
in negative ionization mode gives a peak at m/z = 591.1, see Fig. 8.
4+
eral complexes with Al
2
(OH)
2
core
−
The peak can be attributed to the formula, Al
2
C
16
O
14
N
4
Cl . In fact
−
this flying anion, Al
2
L
2
(OH)
2
Cl , can be formed from the neutral
a
b
c
d
Parameters/ligand
Al(1)–O(7)/A˚
Al(1 )–O(7*)/A
Al–N/A˚
EDBHP
ida
nta
heidi
0
−
dimer, Al
2
L
2
(OH)
2
accommodating one Cl . The advantage of
e
e
the chloride is that the calculated isotope distribution gives strong
support for the assignation, as shown in the insert of Fig. 8.
1.858(7)
1.84(1)
2.042(5)
1.853/1.849
1.847/1.847
2.081/2.083
1.816
1.878
2.073
1.898
1.797
2.056
ꢀ
˚
2
.08(1)
1.87(1)
.89(1)
Al–O (carb)/A˚
1.884/1.873
1.889/1.880
—
1.904
1.891
1.909
—
1.889
1.891
—
1
—
—
Al–O(aq)/A˚
1.884/1.876
1.884
ꢀ
◦
Al(1)–O(H)–Al(1 )/
101.6(3)
102.2/102.2
100.6
102.7
a
This work. O(7) and O(7*) are O-atoms of the bridging OH-groups
according to the numbering on Fig. 7. The figures are averages of slightly
different values of three molecules in the unit cell. For further details see
b
c
d
e
2 2
ESI.† Ref. 14 and ref. 15. Ref. 16. Ref. 15. Notice that in the Al (heidi)
d
2 2 2
complex an Al (O–CH –) unit exists.
14,15
2−
16
the Al
2
(OH)
2
(imda)
2
and Al
2
(OH)
2
(nta)
2
complexes. (imda:
Fig. 8 ESI MS spectrum of an aqueous sample consisting of cAlCl3
=
imino-diacetate, nta: nitrilo-triacetate.) The importance of l-OH
−
3
c
EDBHP = 0.001 mol dm , the pH is adjusted by KOH to about 5.5. The
bridges in the formation of oligomeric aluminium species has
17
insert shows the simulated pattern of the different Cl-isotopomers.
recently been reported by Casey et al. The EDBHP ligands
are coordinated in a four-dentate, non-bridging fashion, both
form three five-membered chelate rings, and the hydroxyl oxygen
atoms of the 3-hydroxy-2-propionil groups keep the protons
and these groups are not coordinated to the Al(III) ions. The
lack of interaction between the hydroxyl group of EDBHP
and Al(III), in contrast to other hydroxycarboxylate-systems, e.g.
The conclusion of this multi-method study is that aluminium(III)
forms stable complex(es) with EDBHP ligands in solution for pH
−3
between 4 and 8 with cAl = cEDBHP ≥ 0.1 mmol dm . The domi-
nating complex is a neutral dimer, Al
2
L
2
(OH) , although a model
2
with monomer species could also be fit to the pH-potentiometric
data. Al(III) ions have distorted octahedral geometry in the
4
−
7,18
[Al
3
(H−1Citrate)
3
(OH)] complexes and [Al
2
(heidi)
2
(OH
2
)] (H
2
solid Al
2
(EDBHP)
2
(OH)
2
·2H
2
O. Two Al(III) ions are connected
15
heidi = hydroxyethyliminodiacetic acid), indicates the impor-
tance of the relative arrangements of the anchor acetate and
the alcoholic HO-group. The 3-hydroxy-2-propionate group of
EDBHP is not coordinated through both O-donors, although it
could result in a five-membered ring, because of the unfavourable
geometry. Additional water molecules are not coordinated to
the inner sphere of the Al(III)-ions; the two water molecules are
through a double hydroxo bridge. Both four-dentate organic lig-
ands are coordinated terminally through two carboxylate groups
and two N-donors forming three five-membered chelate rings. The
hydroxyl groups of the ligand EDBHP remain protonated and
are not coordinated to the aluminium ions. The structure and
composition of the dimer are very likely the same in solution
and the solid state.
crystalline waters. Selected structural data of Al
and several complexes with an Al (OH) entity are collected in
Table 1. The figures for the core of the complexes are very similar
for our compound and Al (OH) (imda) , but no large differences
2
(OH)
2
L
2
·2H
2
O
2
2
Experimental
2
2
2
Preparation of solutions and crystals of
can be seen in either the case of nta or heidi.
Al (EDBHP) (OH) ·2H O
2
2
2
2
There is one question remaining, namely, whether the dimer is
already formed in solution or if it just appears in the solid phase.
Certain evidence for the dominance of the dimer in solution with
the same composition as in the solid would probably mean that
the structure of the solid is retained in solution. However, one can
guess that the driving force for dimer formation from its monomer
is an increase of the total concentration during evaporation of
the solvent, or just that the dimer has the lowest solubility in
water or in the water–ethanol mixture. The ESI MS method
should give information about the stoichiometry of the species
in solution, although the above-mentioned dilemma of the change
in concentration cannot be avoided. The best chance of retaining
the composition of the solution during the ionization procedure
in the ESI MS equipment arises for the case of inert metal–
ligand systems. Al(III) is relatively inert for water exchange and for
−
3
0
.2000 mol dm AlCl
3
was prepared from a weighed quantity of
99.99% Al-wire (Aluminium Work, Ajka, Hungary) dissolving it in
20
3
equivalents of HCl as described elsewhere, therefore no excess
−
3
acid was present in the stock solution. 0.2 mol dm
K EDBH was
2
prepared from solid H EDBH (Tosoh Co. Ltd., Japan), dissolving
2
it in 2 equivalents of carbonate free KOH solution and standard-
ized by pH-potentiometry, see above. Single crystals were grown
−
3
−3
from an aqueous solution (cAl = 5.0 × 10 mol dm , cAl/c
at pH ≈ 5.5 with no supporting electrolyte) after diffusion of
6% ethanol layered above the sample in a glass tube with 3 mm
L
= 1
9
diameter. The crystallization of the colourless crystals started after
a few weeks, and only very few small crystals could be harvested.
pH-potentiometry
18
◦
multidentate ligand exchange. Therefore, Al(III)–ligand systems
The titrations were carried out at 25 C and constant ionic
19
−3
are good models for ESI MS studies. The MS spectrum recorded
strength, in 1 mol dm KCl in an argon atmosphere. The pH
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3221–3227 | 3225

View Article Online
measurement was performed with a high precision pH meter,
Radiometer PHM 84 equipped with a Metrohm 6.0234.100
were refined anisotropically, using weighted full-matrix least
squares on F . Hydrogen atoms attached to hydroxyl oxygen
2
21
combination electrode. The electrode was calibrated as pH =
atoms and water ones were located in a difference map and other
hydrogen atoms were placed in calculated positions with isotropic
thermal parameters, and were not refined. All calculations were
+
−
log[H ], therefore the derived constants were stoichiometric
constants. The titrant was added with an automatic burette,
Radiometer ABU 80. The solutions were stirred at constant speed.
The titration of the free ligand was started from pH = 10.5 with
26
performed with the teXsan crystallographic software package of
Molecular Structure Corporation.
−
3
0
.2231 mol dm HCl. The concentration of the K
2
EDBHP stock
CCDC reference number 602726.
For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b517192j
−
3
solution (c = 0.220 mol dm ) and the protonation constants of
the ligand were evaluated using the SUPERQUAD program. The
Al(III)–ligand samples (v
22
3
−3
−3
0
= 20.0 cm , cAl = 5.00 × 10 mol dm ,
c
L
/cAl = 1 and 1.2) at pH = 3–10.5 were titrated with 0.2215 M
Acknowledgements
KOH solution. The required equilibration time after addition of
KOH to achieve constant pH values was less than 10 min in the
ranges of pH < 5.2 and pH > 8, and substantially longer (up to over
We acknowledge OTKA T 038296 and K63388 for the financial
support, and the Tosoh Co. Ltd. for donating the EDBHP ligand.
R. J. is grateful to the Sasaki Foundation for an Environmen-
tal Sciences scholarship. The authors also thank Dr S a´ ndor
K e´ ki and his research group at the Department of Applied
Chemistry, Debrecen University for ESI-MS measurements, Dr
Mikhail Maliarik for valuable discussions and Dr David Lawrence
1
h) in the non-buffered region close to pH = 7. These data values
were not involved in the equilibrium calculations. The protonation
constants of the ligand measured in independent titrations, and
the stability constants of Al(III)–hydroxo species were kept at
fixed values during the refinement of the stability constant for the
23
9
studied Al(III)–EDBHP complexes by the PSEQUAD program
(
Link o¨ ping University) for linguistic correction of this manuscript.
(
See also Table S1 in ESI†). The uncertainties of the constants
were estimated values referring to reproducibility, rather than (the
smaller) uncertainty of the program.
References
1
D. R. Williams, Chem. Br., 1998, 34(1).
NMR measurements
2 R. Takahashi, N. Fujimoto, M. Suzuki and T. Endo, Biosci., Biotech-
nol., Biochem., 1997, 61, 1957.
3 D. Schowanek, T. C. J. Feijtel, C. M. Perkins, F. A. Hartman, T. W.
Federle and R. J. Larson, Chemosphere, 1997, 34(11), 2375.
4
1
5
00/360 MHz H NMR spectra were recorded on a Bruker Avance
1
13
5
00/Bruker Avance 360 spectrometer with H/ C dual/QNP
T. C. J. Feijtel and D. Schowanek, Environ. Sci. Technol., 2001, 35,
probe in 5 mm tubes. Typical acquisition parameters: pulse length
1
765.
◦
5
ls (45 ), pulse repetition time 3 s, spectral window 4.5 kHz,
5
E. Dvorakova, B. Kopecka, J. Majer, M. Struhar and V. Springer, Chem.
Zvesti, 1972, 26, 316.
size 32 K, number of scans 32, D O-lock, the chemical shift
referred to DSS as external standard. 104.23 MHz Al NMR
spectra were measured by a Bruker AM 400 instrument using
2
2
7
6 P. Rubini, A. Lakatos, D. Champmartin and T. Kiss, Coord. Chem.
Rev., 2002, 228, 137.
7
¨
L.-O. Ohman and S. Sj o¨ berg, Inorg. Chim. Acta, 1983, 79, 215; L.-O.
a 10 mm BB probe. Acquisition parameters: pulse length 10 ls
O¨ hman and S. Sj o¨ berg, J. Chem. Soc., Dalton Trans., 1983, 39, 1851;
◦
¨
L.-O. Ohman, Inorg. Chem., 1988, 27, 2565.
(
30 ), pulse repetition time 0.2 s, spectral window 120 kHz, size
8
9
V. B. Di Marco, G. G. Bombi, M. Ronaldo and P. Traldi, Rapid
Commun. Mass Spectrom, 2006, 20, 710.
8
0
K, number of scans 2000, D
2
O-lock, the chemical shift referred to
in 0.01 mol dm HCl (d = 0 ppm), as external
−
3
−3
.01 mol dm AlCl
3
L. Z e´ k a´ ny and I. Nagyp a´ l, PSEQUAD, a Comprehensive Program for
the Evalution of Potentiometric and/or Spectrophotometric Data in
Computation Methods for Determination of Formation Constants, ed.
by D. J. Legget, Plenum Press, New York, 1985.
standard.
Electrospray ionisation mass spectrometry
1
0 Model calculations indicate some differences between the two models
+
−
4
] at a narrow pH-range of 8–9, but the
both in [H ] and [Al(OH)
differences are very small and seem to be hardly measurable by
ESI MS measurements were performed using a Bruker BioTOF II
ESI-TOF instrument equipped with a Cole Palmers 74900 Series
pump (sample flow rate 2 ll min ) in the negative ionization
mode at 2000 V corona, 4500 V endplate, 4000 V cylinder, 2000 V
capillary enter and 120 V capillary exit voltage. The temperature
27
potentiometry or Al NMR.
−
1
11 G. Anderegg, Critical Stability Constants of EDTA Complexes, IUPAC
Chemical Data Series No. 14. Pergamon, 1977; A. E. Martell and
R. M. Smith, Critical Stability Constants, Plenum Press, New York,
1
974.
◦
of the N
2
gas used for desolvation was 100 C. The frequency of
12 I. Puigdomenech, in MEDUSA (Windows Interface to the MS-DOS
Version of INPUT, SED and PREDOM Fortran Program Drawing
Chemical Equilibrium Diagrams). Program is available at http://www.
kemi.kth.se/medusa/ or http://web.telia.com/∼u15651596/, 1997.
data collections was 2 GHz. The sample consisted of cAlCl3 = c
L
=
−
3
0
.001 mol dm at pH ∼ 5.5.
1
1
3 J. W. Akitt, Prog. NMR Spectrosc., 1989, 21, 1.
4 P. S. Petrosyants, M. A. Malyarik and A. B. Ilyukhin, Zh. Neorg. Khim.,
X-Ray diffraction
The diffraction intensities of a colourless platelet crystal
1
995, 40, 769.
1
1
5 W. Schmitt, P. A. Jordan, R. K. Henderson, G. R. Moore, C. E. Anson
and A. K. Powell, Coord. Chem. Rev., 2002, 228, 115.
6 Giovanni C. Valle, G. Giorgio Bombi, Benedetto Corain, Mose´
Favarato and P. Zatta, J. Chem. Soc., Dalton Trans., 1989, 8,
1513.
of Al (EDBHP) (OH) O) were collected with graphite
2
2
2
·2(H
2
monochromated Mo-K
a
radiation using a Rigaku RAXIS-IV
◦
Imaging Plate diffractometer at 23 C to a maximum 2h value of
◦
17 W. H. Casey, M. M. Olmstead and B. L. Phillips, Inorg. Chem., 2005,
6
5.1 . The data were corrected for Lorentz and polarisation effects.
4
4, 4888.
24
The structure was solved by heavy-atom Patterson methods and
1
8 A. Bodor, I. B a´ nyai, L. Z e´ k a´ ny and I. T o´ th, Coord. Chem. Rev., 2002,
228, 163.
2
5
expanded using Fourier techniques. All non-hydrogen atoms
226 | Dalton Trans., 2006, 3221–3227
3
This journal is © The Royal Society of Chemistry 2006

View Article Online
1
2
2
2
2
9 V. B. Di Marco, G. G. Bombi, M. Tubaro and P. Traldi, Rapid Commun.
Mass Spectrom., 2003, 17, 2039.
24 PATTY: P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman,
R. de Gelder, R. Israel and J. M. M. Smits, Technical Report of the
Crystallography Laboratory, University of Nijmegen, The Netherlands,
1994.
25 DIRDIF94 P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman,
R. de Gelder, R. Israel and J. M. M. Smits, The DIRDIF-94
program system, Technical Report of the Crystallography Laboratory,
University of Nijmegen, The Netherlands, 1994.
0 A. Bodor, I. T o´ th, I. B a´ nyai, Z. Szab o´ and G. T. Hefter, Inorg. Chem.,
2
000, 39, 2530.
1 H. M. N. H. Irving, M. G. Miles and L. D. Pettit, Anal. Chim. Acta,
967, 38, 475.
2 P. Gans, A. Sabatini and A. Vacca, Inorg. Chim. Acta, 1983, 79,
19.
1
2
3 C. F. Baes and R. E. Mesmer, The Hydrolysis of Cations, Wiley, New
York, 1976. p. 112.
26 in ‘teXtan’, teXsan: Crystal Structure Analysis Package, Molecular
Structure Corporation, 1985 & 1999.
This journal is © The Royal Society of Chemistry 2006
Dalton Trans., 2006, 3221–3227 | 3227