J ournal of Chemical and Engineering Data, Vol. 43, No. 6, 1998 1073
Sch em e 1. Syn th etic Rou te for P r ep a r a tion
Acyclic P olyeth er Dica r boxylic Acid s (1-9)
solid. Their melting points, 1H NMR spectra, and IR
spectra are listed in Table 1.
P oten tiom etr ic Titr a tion . Protonation constants and
stability constants of synthesized organic ligands in this
study were determined in 90 vol % methanol. Table 2
shows the first (log âH L) and second protonation constant
2
(log âH L), respectively, for each ligand. Upon the elonga-
2
tion of the ethylene glycol unit connected between the two
benzene rings, both protonation constants gradually in-
creased. Stability constants for the complexation of alka-
line earth metal ions were measured and are listed in Table
3. For most metal ions, the order of stability constant is
Ca2+ > Sr2+ Z Ba2+ . Mg2+. As the number of ethylene
glycoloxy units decreases, the stability constants toward
the calcium ion gradually decrease. It was reported that
when a ligating agent complexes with a specific metal ion,
major factors that influence the stability constant are
radius of metal ion, oxidation state of metal ion, species of
donor atoms, solvation effect, anion effect, and dipolar
interaction between the organic ligand and metal ion
(Vo¨gtle and Weber, 1985). So, the calcium selectivity
indicates that when the pseudo-cyclic conformation of the
organic ligand is carried out, the length of the monoeth-
ylene glycol unit is sufficient to fit with the calcium ion to
give a maximized ion-dipolar interaction and electrostatic
interaction. The stability constants based on both ef-
ficiency and selectivity toward an alkaline earth cation are
influenced by the introduction of a lipophilic group, C4H9.
The lipophilic ligand shows better conformation when a
metal ion is introduced. This is probably attributed to a
micelle effect with which the cyclic formation takes place
much more easily. From the distribution diagram, L2- was
observed to exist at over pH 6, indicating two hydrogens
are completely dissociated. The distribution species of
metal ion (M), complex (LM), and LMH-1 (MLOH), which
combines with the OH anion, give information on 1:1
maximum complexation in the pH range 4-6. At pH over
10, basic conditions, most of the metal complexes were
observed to exist as MLOH.
to 0.001 pH unit. The glass combination electrode was an
ORION Ross Electrode model 81-02. The filling solution
was made of 3 M TMACl in 90 vol % methanol (90 mL
MeOH/100 mL solution). The measurement was under-
taken at (25 ( 0.1) °C under a nitrogen atmosphere. The
pH meter was standardized against oxalate buffer solutions
before titration. Oxalate buffer solution (pH 3.73) were
made by ammonium oxalate monohydrate and oxalic acid
in 90 vol % methanol. Succinate buffer solutions (pH 6.73)
were made from succinic acid and sodium succinate salt
in 90 vol % methanol. Tetramethylammonium chloride (0.1
M) was used to adjust the ionic strength. A glass combina-
tion electrode, a microburet with 5.0 mL capacity, and 50
mL of 1 mM ligand in 90 vol % methanol were introduced
into a thermostated titration vessel equipped with a
magnetic stirring bar. The concentration of titrant, tet-
ramethylammonium hydroxide, was 20.3 mM. Titrations
were conducted in triplicate. Protonation and stability
constants were calculated by the program BEST (Martell,
1992).
Extr a ction P r oced u r es. Solvent extraction was fol-
lowed from the procedures reported by Bartsch et al. (1992).
Advantages of this method that it is known to speed the
experiment and it uses a small amount of organic ligand.
For competitive extractions of alkaline earth metal chloride
solution, a microextraction technique was used. An aque-
ous solution of alkaline earth metal chloride (2.00 mL,
0.125 M with 0.2 M cesium hydroxide for pH adjustment)
and 2.00 mL of 0.01 M organic ligand in chloroform in a
10 mL centrifuge tube were mixed by vortex mixer for 5
min. After centrifuging, the equilibrium pH of the upper
aqueous layer was measured. Then a 1.00 mL sample of
the chloroform layer was taken and stripped with 1.00 mL
of 0.1 M HCl solution. The concentration of metal ion in
the aqueous phase was determined by an atomic absorption
spectrometer.
Two-P h a se Extr a ction . To further prove the influence
of the lipophilicity and the polyether length of the organic
ligand on metal ion complexation, solvent extraction ex-
periments with all ligands except 1, 4, and 7 (see Table 4)
for alkaline earth metal ions were carried out. Solvent
extraction using 1, 4, and 7, which do not contain lipophilic
side chains, could not be carried out with this experiment
because of their low solubility in chloroform. A profile for
stripped metal ion concentrations by HCl aqueous solution
after extracted with the organic ligand 2 which performs
the best complexation behavior among nine organic ligands
in chloroform, is shown in Figure 1. The extracted metal
ion concentration was observed to increase with pH in-
crease. Calcium ion was selectively extracted at about 40-
50% from the source phase. Table 4 shows the selectivity
order of extracted metal ions as Ca2+ > Sr2+ Z Ba2+
.
Mg2+, indicating the same propensity of calcium ion
selectivity over other alkaline earth metal ions as observed
from potentiometric titration experiment. Hence the ex-
traction ability of an organic ligand is strongly relate to
that of the stability constant for complexation. The total
loaded amount of metal ions is over 90% in all cases.
Especially, ligand 2 shows the best extracting behavior to
give 99.4 of loading percentage. In Table 4, the selectivity
of calcium ion for 2, 5, and 8 in which butyl groups are
branched was observed to be somewhat better than that
for 3, 6, and 9 with tetradecyl group substituted. Walkow-
iak et al. (1992) have reported that when the pendent side
Resu lts a n d Discu ssion
Syn th eses. The synthetic route for the preparation of
acyclic polyether dicarboxylic acids is shown in Scheme 1.
Bisphenol was used as a starting material, which can be
prepared from the reaction of catechol with bis(2-chloro-
ethyl) ether in NaOH solution. Reaction of bisphenol with
2.5 equiv of 2-bromoalkanoic acid in the presence of 5 mol
of sodium hydride as a base provided the desired products
in over 90% yield. Attempted reactions with potassium
hydride instead of sodium hydride were found to be less
effective. Recrystallization from diethyl ether gave a white