A new carrier for selective removal of heavy metal ions from aqueous solutions through bulk liquid membranes

Article abstract of DOI:10.1002/ejoc.200400222

The carrier-mediated transport of heavy metal ions through bulk liquid membranes has been examined: toxic Hg²⁺, Cd²⁺ and Pb ²⁺ ions were studied, along with Cu²⁺ ions for comparative purposes. The ability of a n

Full text of DOI:10.1002/ejoc.200400222

FULL PAPER
A New Carrier for Selective Removal of Heavy Metal Ions from Aqueous
Solutions through Bulk Liquid Membranes
Nicoletta Spreti,[a] Lucia Brinchi,[b] Raimondo Germani,[b] Maria Vincenza Mancini,[a] and
Gianfranco Savelli*[b]
Keywords: Carriers / Lipophilic polyamines / Liquid membranes / Toxic heavy metal ions
The carrier-mediated transport of heavy metal ions through
bulk liquid membranes has been examined: toxic Hg2+, Cd2+
and Pb2+ ions were studied, along with Cu2+ ions for compar-
ative purposes. The ability of a new carrier, 2,2Ј-bis(p-oct-
yloxybenzyl)diethylenetriamine (bis-pODET), to complex
and transport all the selected metal ions is reported. Differing
affinities of the carrier for the different metal ions and the
different experimental conditions required for their release
into the receiving phase allowed the selective separation of
while the inefficacy of the separation of Cd2+/Pb2+ and Hg2+
/
Cu2+ mixtures was for two different reasons: (i) the carrier is
able to extract the metal ions with similar levels of ability,
and (ii) the carrier metal ion complexes require the same
acidity of the receiving phase to release the metal ions. The
capability of the carrier to transport Hg2+ efficiently in con-
secutive cycles is also reported: over 90% of the metal ions
were transferred into the receiving phase for three consecut-
ive processes.
equimolar binary mixtures. For Hg2+/Cd2+ and Hg2+/Pb2+ ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
mixtures, two different separation methods were performed,
Germany, 2004)
Introduction
Moreover, the carrier must have a suitable lipophilic-hydro-
philic balance so as both to be soluble in the membrane
and at the same time to be able to reach aqueous/organic
interfaces, and long-chained ligands have been successfully
used to transport copper() ions across a chloroform mem-
brane.[17]
We have recently reported the ability of 1,1,7,7-tetraethyl-
4-(tetradecyl)diethylenetriamine (TE14DT) to act as a car-
rier for Cu2ϩ, Cd2ϩ, Hg2ϩ and Pb2ϩ ions.[18,19] While cap-
able of extracting all the selected metal ions from the source
phase into the membrane, TE14DT also displayed a re-
markable efficacy to transport only Cu2ϩ and Cd2ϩ. In the
case of Hg2ϩ, lack of transport was due to the high stability
of the Hg2ϩϪcarrier complex in the organic phase, and
modifications of some external parameters (i.e., the pH of
the receiving phase) did not allow the release of the metal
ion.
Environmental contamination by heavy metal ions re-
flects both natural sources and human industrial
processes,[1Ϫ3] and their effects impact on several organ sys-
tems and biochemical activities.[4,5] The necessity to reduce
the amount of heavy metal ions to acceptable levels in was-
tewater streams, and subsequent possible reuse of these me-
tal ions, has resulted in increasing interest in the develop-
ment of new methods for their selective removal from mix-
tures.
Liquid membrane technology, coupled with suitable
selective carriers, provides a possible means by which to ac-
complish such separations.[6Ϫ9]
Bulk liquid membrane (BLM) systems are useful for
screening carriers and they are widely used to study ion
transport from aqueous compartments to others through
organic membranes. Crown ethers and acyclic polyethers,
with different donor atoms, as well as other complexing
molecules, have been extensively employed to effect trans-
port of some of the most toxic heavy metal ions, such as
mercury(),[10,11] lead()[12Ϫ15] and cadmium().[15,16]
This paper presents
a
new carrier: 2,2Ј-bis(p-
octyloxybenzyl)diethylenetriamine (bis-pODET). Its chelat-
ing region is very similar to that of TE14DT, since three N
atoms act as complexing agents. On the other hand, it pos-
sesses two alkyl chains and aromatic moieties and has a
pincer-type structure (Scheme 1).
Its transport capability was evaluated as a function of the
nature of the membrane and of the pH of the receiving
phase, and optimal experimental conditions for single-metal
ion transport were found. Selectivity tests were also per-
formed, and two different separation methods of binary
mixtures are presented.
[a]
Dipartimento di Chimica, Ingegneria Chimica e Materiali,
`
Universita di L’Aquila,
Via Vetoio Coppito, 67100 L’Aquila, Italy
[b]
CEMIN, Centro di Eccellenza Materiali Innovativi
`
Nanostrutturati, Dipartimento di Chimica, Universita di
Perugia,
Via Elce di Sotto 8, 06123 Perugia, Italy
Fax: (internat.) ϩ 39-075-5855538
E-mail: savelli@unipg.it
Eur. J. Org. Chem. 2004, 3865Ϫ3871
DOI: 10.1002/ejoc.200400222
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3865

N. Spreti, L. Brinchi, R. Germani, M. V. Mancini, G. Savelli
FULL PAPER
For Hg2ϩ, partition experiments were also carried out
with variation of [Hg2ϩ] in the source phase and constant
values of carrier. The two sets of experiments, with [car-
rier] ϭ1 ϫ 10Ϫ3  and 5 ϫ 10Ϫ4 , were both consistent
with a ratio of 3:2 in the association of Hg2ϩ to the carrier.
Single Ion Transport
To choose the proper organic phase for selectivity studies,
single ion transport experiments were performed. Table 2
reports the percentages of M2ϩ in the source phase and
receiving phase at different times. A similar decrease in the
amount of M2ϩ in the source phase was observed in the
early hours of the transport processes, both in dichloro-
methane and in chloroform, and the ability of the carrier
to extract all the metal ions seems very similar. Moreover,
in dichloromethane no significant differences in metal ion
extraction were found after 24 and 48 h, the percentages of
metal ions in the source phase ranging from 20 to 34.
Transfer of M2ϩ in the chloroform membrane seemed to
be quite different. In fact, after 48 h, ca. 10% of Cd2ϩ and
Pb2ϩ remained in the source phase, while the percentages
of Hg2ϩ and Cu2ϩ were 35 and 51, respectively, indicating
a potential metal ion screening capability of bis-pODET in
the chloroform membrane.
Scheme 1
Results and Discussion
Partition Measurements
Furthermore, only Cd2ϩ and Pb2ϩ ions were efficiently
transferred into the receiving phase in both membranes (see
Table 2), and the chloroform membrane showed better re-
sults and so was then chosen for the next experiments.
As reported in the Introduction, the structure of bis-pO-
DET is quite similar to that of TE14DT, with three N
atoms representing the chelating region. Like TE14DT, this
new carrier was able to transport Cd2ϩ ions, and its efficacy
was practically the same.[19] However, different patterns
were found for Cu2ϩ and Pb2ϩ ions, indicating that not only
the nature of the chelating atoms but also many other fac-
tors can influence the carrier-mediated transport. In fact,
The ability of the carrier bis-pODET to extract the metal
ions from the source phase to the membrane was investi-
gated with two different organic solvents, chloroform and
dichloromethane, by the procedure reported in the Exp.
Sect. Carrier extraction results, along with spontaneous me-
tal ion partitions, are reported in Table 1.
Table 1. Mol percentages of M2ϩ extracted from the source phase
to the organic phase (conditions: 25 °C; source phase: 3 mL of 5
ϫ 10-3  MCl2 in 0.15  acetate buffer pH ϭ 4.68; membrane: 3
mL of 1 ϫ 10-3  carrier in the organic solvent)
Membrane
Cu2ϩ
Cd2ϩ
Pb2ϩ
Hg2ϩ
TE14DT was an excellent carrier for Cu2ϩ [18]
while the
,
slow diffusion of Pb2ϩϪcarrier complex through the mem-
brane was probably the cause of the lack of transport, most
of the Pb2ϩ remaining in the source phase.[19] On the other
hand, in the presence of bis-pODET, after 48 h, 58% Pb2ϩ
had been transported, while Cu2ϩ was rapidly extracted
CHCl3
CH2Cl2
bis-pODET in CHCl3
bis-pODET in CH2Cl2
Ͻ 1
Ͻ 1
18
Ͻ 1
Ͻ 1
16
Ͻ 1
Ͻ 1
19
Ͻ 1
11
22
27
27
20
35
For the charge balance in the membrane, M2ϩ and into the chloroform, but its release into the receiving phase
M2ϩϪcarrier complex partitioning in the organic phase was negligible.
should occur as ion pairs. In the case of Hg2ϩ, we have
Similar behaviour (i.e., fast decrease of metal ion concen-
already shown that the main accompanying ion in the par- tration in the source phase and slight release in the receiving
tition between source phase and membrane is ClϪ.[19]
one) was also observed for Hg2ϩ (11% after 48 h). Further-
In the absence of the carrier, only Hg2ϩ partitioned in more, Hg2ϩ diffused through the chloroform spontaneously
the dichloromethane membrane, and its diffusion from the in the absence of the carrier, and its percentage in the receiv-
source phase to the receiving phase has already been re- ing phase after 48 h was ca. 30% (data not shown). This is
ported.[19]
probably due to formation of stable complexes between bis-
Furthermore, Table 1 shows that the carrier was able to pODET and both Cu2ϩ and Hg2ϩ ions and, in order to
extract all the metal ions and that only slight differences improve their transport, variations of the experimental con-
between the two organic phases were observed, except in ditions seemed to be necessary.
the case of Hg2ϩ: the higher percentage in dichloromethane
Among the external factors involved in transport pro-
with respect to chloroform (35% vs. 22%) was probably due cesses, the pH of the receiving phase could play a crucial
to its spontaneous partitioning (11%) in the former mem- role. Transport in these systems is indeed pH-driven: the
brane.
metal ions in the source phase are exchanged by the carrier
3866
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org Eur. J. Org. Chem. 2004, 3865Ϫ3871

Selective Removal of Heavy Metal Ions from Aqueous Solutions
FULL PAPER
Table 2. Mol percentages of M2ϩ in the receiving (R) and source (S) phases at 25 °C
Time [h]
bis-pODET[a]
TE14DT[b]
CH2Cl2
CHCl3
Cd2ϩ Pb2ϩ
CH2Cl2
Cd2ϩ
Cu2ϩ
Hg2ϩ
Cu2ϩ
Pb2ϩ
Hg2ϩ
Cu2ϩ
Cd2ϩ
Pb2ϩ
Hg2ϩ
R
S
R
S
R
S
R
S
R
S
R
S
R
S
R
S
R
S
R
S
R
S
R
S
4
8
24
48
1
1
2
3
59 15 60 10 58
53 33 45 16 52
52 55 15 36 29
Ͻ 1 55
46
11 37
11 35
Ͻ 1 48 10 64
Ͻ 1 40 29 48 17 38
6
59
1
5
43 23 72
36 44 52 19 70
7
78 Ͻ 1 94
5
41
3
1
2
4
93 11 34
92 14 28
90 15 26
1
1
36 41 27 25 23 11 26 78 19 53 37
34 62 20 38 21 12 24 89 72 19
51 75
7
58 12
8
[a]
Conditions: source phase: 8 mL of 5 ϫ 10Ϫ3  MCl2 in 0.15  acetate buffer pH ϭ 4.68; membrane: 20 mL of 1 ϫ 10Ϫ3  carrier in
the organic solvent; receiving phase: 8 mL of 3.2 ϫ 10Ϫ3  HCl (pH ϭ 2.5). Refs.[18,19] Conditions: source phase: 8 mL of 5 ϫ 10Ϫ3
[b]
 MCl2 in 0.15  acetate buffer pH ϭ 4.68; membrane: 20 mL of 1 ϫ 10Ϫ3  carrier in the organic solvent; receiving phase: 8 mL of
0.1  HCl (pH ϭ 1).
for protons originating from the receiving one. A buffered
aqueous solution was therefore used as source phase to
minimize pH variation due to Hϩ counter transport. The
release of the metal ions into the stripping phase depended
on the protonation of atoms in the chelating region, so
when Hg2ϩ and Cu2ϩ concentrations had levelled off, small
amounts of concentrated HCl were added in the stripping
phase and, despite the system becoming heterogeneous, the
concentrations of the two metal ions rose significantly.
Experiments with more concentrated HCl solution as re-
ceiving phase (pH ϭ 0.3) were then carried out, and per-
centages/time profiles are shown in Figure 1. Data relating
to transport at pH ϭ 2.5 are also displayed.
The extraction curves in the early hours of the process
are very similar, independently of the pH of the receiving
phase, but the more efficient release of the metal ions al-
lowed further extraction. Notwithstanding the improve-
ment in the transport and the low amount of metal ions in
the membrane (ca. 15%) for both Cu2ϩ and Hg2ϩ, bis-pO-
DET was far more effective for Hg2ϩ than for Cu2ϩ, the
percentages in the receiving phase after 48 h being 80% and
46%, respectively.
Simultaneous Transport of Hg2؉/M2؉ Mixtures
On the basis of single ion transport experiments, different
operational conditions could potentially allow sequential
separation of Hg2ϩ/Cd2ϩ and Hg2ϩ/Pb2ϩ mixtures. In-
itially, with a pH ϭ 2.5 receiving phase, Cd2ϩ and Pb2ϩ
should be efficiently transported, with Hg2ϩ stored in the
membrane as metal ionϪcarrier complex. Then, after the
removal of the concentrated Cd2ϩ (or Pb2ϩ) solution, the
addition of a new aqueous phase at pH ϭ 0.3 should allow
Figure 1. Mol percentages of Hg2ϩ (top) and Cu2ϩ (bottom) vs.
time in the source phase (filled symbols) and receiving phase (open
symbols) with 3.2 ϫ 10Ϫ3  HCl, pH ϭ 2.5 (squares) and 0.5 
HCl, pH ϭ 0.3 (circles) as receiving phase; [bis-pODET] ϭ 1 ϫ
10Ϫ3  in chloroform; [MCl2] ϭ 5 ϫ 10Ϫ3  in 0.15  acetate
buffer, pH ϭ 4.68; T ϭ 25 °C
the release of all Hg2ϩ ions present in the membrane.
Equimolar mixtures of Cd2ϩ (or Pb2ϩ) and Hg2ϩ were
therefore used as source phase for these experiments, and
[M2ϩ
]
was 5 ϫ 10Ϫ3 , as in the single ion transport
tot
experiments. The carrier concentration was doubled, since
the affinity of bis-pODET for Hg2ϩ was much higher than
for Cd2ϩ (and Pb2ϩ) and the amount of the carrier was
not enough to transport Cd2ϩ and Pb2ϩ ions through the tracted from the source phase, and, as expected, only Cd2ϩ
membrane (data not shown). Results relating to Hg2ϩ/Cd2ϩ ions were transported in the receiving phase (ca. 90%).
mixtures are shown in Figure 2. In the first part of the pro- Then, Hg2ϩ ions, stored in the membrane, were quickly
cess (0Ϫ48 h), both Hg2ϩ and Cd2ϩ were efficiently ex- transferred into the new aqueous stripping phase at pH ϭ
Eur. J. Org. Chem. 2004, 3865Ϫ3871
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3867

N. Spreti, L. Brinchi, R. Germani, M. V. Mancini, G. Savelli
FULL PAPER
Figure 3. Mol percentage of Hg2ϩ(circles) and Pb2ϩ (squares) vs.
time in the source phase (filled symbols) and receiving phase (open
symbols) in the simultaneous transport; [bis-pODET] ϭ 1 ϫ 10Ϫ3
 in chloroform; [PbCl2] ϭ [HgCl2] ϭ 2.5 ϫ 10Ϫ3  in 0.15 
acetate buffer, pH ϭ 4.68; receiving phase: 0.5  HCl, pH ϭ 0.3;
T ϭ 25 °C
Figure 2. Mol percentages of Hg2ϩ (circles) and Cd2ϩ (squares) vs.
time in the source phase (filled symbols) and receiving phase (open
symbols) in simultaneous transport; [bis-pODET] ϭ 2 ϫ 10Ϫ3  in
chloroform; [CdCl2] ϭ [HgCl2] ϭ 2.5 ϫ 10Ϫ3  in 0.15  acetate
buffer, pH ϭ 4.68; T ϭ 25 °C; see text for the experimental pro-
cedure
Finally, the separation of an Hg2ϩ/Cu2ϩ mixture was at-
0.3. In this way, sequential separation of Cd2ϩ and Hg2ϩ tempted. Both metal ions were efficiently transported by
was carried out.
use of the more acidic solution as receiving phase, and so
Similar experiments were performed with equimolar mix- only a much higher carrier affinity for either of the metal
tures of Pb2ϩ and Hg2ϩ, but lower selectivity was obtained. ions could allow selective transport.
After 48 h, only 67% of Pb2ϩ had been transferred into the
Experiments were carried out with 1 ϫ 10Ϫ3  and 2 ϫ
receiving phase, with more than 25% remaining in the mem- 10Ϫ3  carrier. A very fast decrease in Hg2ϩ concentration
brane. When the receiving phase was replaced with the in the source phase was observed and, after the first
more acidic one, besides the quantitative transport of Hg2ϩ 8Ϫ10 h, all the metal ions had been extracted by the carrier.
ions, the total release of Pb2ϩ present in the organic phase Conversely, the extraction of Cu2ϩ ions was dependent on
was observed. Moreover, a further increase in carrier con- carrier concentration, being more rapid with 2 ϫ 10Ϫ3

centration decreased the efficiency of the whole process: bis-pODET (41% and 18% in the source phase after 8 h
both metal ions were extracted more rapidly, but their re- with 1 ϫ 10Ϫ3  and 2 ϫ 10Ϫ3  carrier, respectively). In
lease into the stripping phases was strongly inhibited. Metal both experiments, however, 24Ϫ48 h were necessary to re-
ion/carrier ratio therefore also plays an important role: as move all Cu2ϩ ions from the source phase. Therefore, bis-
already reported, an excess carrier concentration either has pODET showed a greater affinity for Hg2ϩ than Cu2ϩ ions.
no significant effect[20,21] or causes a slight decrease in
transport efficiency.[11,18,22]
On the other hand, the release of both metal ions into
the receiving phase was more effective with the lower carrier
In view of the high affinity of bis-pODET for Hg2ϩ with concentration. After 24 h, quantitative transport of Hg2ϩ
respect to Cd2ϩ and Pb2ϩ ions observed in these last experi- was observed, while the amount of Cu2ϩ transferred into
ments, a different separation method was used. The use of the stripping phase reached 60% and remained constant in
a more acidic solution as receiving phase should allow the the next hours. A similar Hg2ϩ/Cu2ϩ ratio, but lower re-
fast transport of Hg2ϩ, most of the competitive ions re- spective percentages (80% for Hg2ϩ and 50% for Cu2ϩ),
maining in the source phase. In this way, there is a ‘‘real were obtained in the stripping phase with 2 ϫ 10Ϫ3  car-
competition’’ between Cd2ϩ (or Pb2ϩ) and Hg2ϩ ions for rier.
the carrier. Figure 3 displays the actual separation of Pb2ϩ
Hg2ϩ mixture (the results with Cd2ϩ/Hg2ϩ ions were very play an important role in increasing the selectivity of Hg2ϩ
similar). transport. In the presence of a smaller amount of the car-
/
On the basis of these results, carrier concentration could
After 10 h, the stripping phase contained ca. 80% Hg2ϩ rier, stronger competition between the metal ions should be
and only 10% Pb2ϩ (or Cd2ϩ), while in the source phase achieved, due to more rapid transport of the metal ion
there remained ca. 10% Hg2ϩ and more than 75% Pb2ϩ (or showing the greater affinity for the carrier. Experiments
Cd2ϩ). Hence, from a starting equimolar binary solution, with 5 ϫ 10Ϫ4  bis-pODET were then carried out. More
we obtained (in a few hours) an Hg2ϩ/Pb2ϩ (or Cd2ϩ) ratio than 70% of Hg2ϩ ions was transported by the carrier, while
of ca. 8:1 in the receiving phase and the reciprocal in the both Cu2ϩ extraction and release proved to be inhibited,
source.
although 30% of the Cu2ϩ had been transferred to the strip-
3868
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org Eur. J. Org. Chem. 2004, 3865Ϫ3871

Selective Removal of Heavy Metal Ions from Aqueous Solutions
FULL PAPER
ping phase after 24 h and the Hg2ϩ/Cu2ϩ ratio was only
slightly increased.
Results obtained with the other receiving phase (pH ϭ
0.3) are reported in Table 4. In this case the transport of
Thus, notwithstanding that bis-pODET exhibited a Cu2ϩ was promoted and, after 48 h, most of the metal ion
higher affinity for Hg2ϩ than for Cu2ϩ ions, no effective had been delivered into the stripping phase. It is noteworthy
mixture separation was accomplished under the experimen- that the Cu2ϩ/Cd2ϩ and Cu2ϩ/Pb2ϩ ratios in the receiving
tal conditions used, and therefore alternative methods phase decreased as the process was in progress. In fact, on
would appear necessary.
short timescales, a sharp decrease in Cu2ϩ concentration in
The separation of Hg2ϩ from metal ion mixtures is a the source phase was observed, along with a rapid rise in
challenging problem in this field of chemistry, and some the receiving one; concentrations then tended to a limiting
highly efficient and selective carriers have been reported in value and to level off. On the other hand, Pb2ϩ and Cd2ϩ
the literature.[10,11,23] However, the competitive metal ions ions were gradually transported with a consequent lowering
are often insoluble in the source phase[11] or are not ex- of selectivity.
tracted by the carrier,[23] so their transport is negligible. In
our system all the competitive ions are transported by the
carrier and the experimental conditions can be modified de-
pending on the nature of the contaminated aqueous solu-
tions.
Table 4. Mol percentages of M2ϩ in the receiving (R) and source
(S) phases in simultaneous transports (Cu2ϩ/Cd2ϩ and Cu2ϩ/Pb2ϩ
at 25 °C)
Time [h][a]
Cu2ϩ/ Cd2ϩ
Cu2ϩ Cd2ϩ
Cu2ϩ/Pb2ϩ
Cu2ϩ Pb2ϩ
R
S
R
S
R
S
R
S
Simultaneous Transport of Cu2؉/Cd2؉ and Cu2؉/Pb2؉
Selectivity tests were also performed with binary equimo-
lar mixtures of Cu2ϩ and Cd2ϩ or Pb2ϩ. The HCl concen-
tration in the receiving phase could also be an useful means
to control selectivity in these transports. Cu2ϩ ions, as pre-
viously seen for Hg2ϩ, were extracted by the carrier, but,
for their release, the more acidic receiving phase had to be
used.
Two different sets of experiments in which the pH of the
stripping phase was varied (i.e., 2.5 and 0.3) were then per-
formed. The carrier affinity for Cu2ϩ was always much
higher than for Cd2ϩ or Pb2ϩ, and in order to allow the
transport of these ions across the membrane, the concen-
tration of the carrier was doubled when a pH ϭ 2.5 solution
was used as receiving phase (Table 3).
2
4
8
24
48
11
16
32
61
71
68
53
35
22
12
5
5
9
22
39
89
86
80
73
56
13
23
42
78
85
55
35
18
8
1
2
6
22
39
95
88
80
58
38
4
[a]
Conditions: source phase: 8 mL of a solution of 2.5 ϫ 10Ϫ3

CuCl2 and 2.5 ϫ 10Ϫ3  CdCl2 (or PbCl2) in 0.15  acetate buffer
pH ϭ 4.68; membrane: 20 mL of 1 ϫ 10Ϫ3  carrier in chloroform;
receiving phase: 8 mL of 0.5  HCl (pH ϭ 0.3).
Simultaneous Transport of Pb2؉/Cd2؉
Simultaneous transport of Cd2ϩ and Pb2ϩ was then
examined. Both metal ions were effectively transported ac-
ross the membrane in the receiving phase at pH ϭ 2.5, and
the ability of bis-pODET to complex more efficiently, and
therefore to transport, either of the metal ions was evalu-
ated. The data reported in Table 5 clearly indicate that no
selectivity between Cd2ϩ and Pb2ϩ was observed, the per-
centages being very similar to those of the single ion trans-
port (see Table 2). The use of two differently concentrated
HCl solutions as receiving phase did not improve carrier
selectivity, since either no significant effect or a decrease in
the transport of both metal ions were observed in receiving
phases at pH ϭ 3.3 and pH ϭ 0.3, respectively.
Table 3. Mol percentages of M2ϩ in the receiving (R) and source
(S) phases in simultaneous transports (Cu2ϩ/Cd2ϩ and Cu2ϩ/Pb2ϩ
at 25 °C)
Time [h][a]
Cu2ϩ/ Cd2ϩ
Cu2ϩ Cd2ϩ
Cu2ϩ/Pb2ϩ
Cu2ϩ Pb2ϩ
R
S
R
S
R
S
R
S
2
4
8
24
48
1
1
2
3
5
48
23
9
8
5
7
67
42
23
21
15
1
1
1
1
1
55
36
18
12
6
3
57
47
34
24
18
22
39
49
56
13
23
41
53
Carrier Recycling
[a]
Conditions: source phase: 8 mL of a solution of 2.5 ϫ 10Ϫ3

CuCl2 and 2.5 ϫ 10Ϫ3  CdCl2 (or PbCl2) in 0.15  acetate buffer
pH ϭ 4.68; membrane: 20 mL of 2 ϫ 10Ϫ3  carrier in chloroform;
receiving phase: 8 mL of 3.2 ϫ 10Ϫ3  HCl (pH ϭ 2.5).
As previously reported, a strongly acidic aqueous solu-
tion must be used to allow the release of Hg2ϩ ions from
the membrane into the receiving phase, in which bis-pO-
The extraction of Cu2ϩ from both mixtures was very fast DET and Hg2ϩ formed a very stable complex. Conse-
and, after 8 h, 90% of the metal ion had been accumulated quently, the three N atoms of the chelating region were
in the organic phase (only ca. 10% in the aqueous phases). highly protonated and the use of the carrier for further
A moderate rise in Cd2ϩ and Pb2ϩ concentration in the transports could be compromised. On the other hand, be-
receiving phase was observed, with over 50% of the metal sides the high efficiency and selectivity of bis-pODET for
ions being transported. These results are not very satisfac- Hg2ϩ ions, the capability to perform many consecutive pro-
tory, since 30% of the Cd2ϩ or Pb2ϩ, in addition to Cu2ϩ
is stored in the membrane.
,
cesses should increase its potential applicability in Hg2ϩ re-
moval from polluted aqueous solutions.
Eur. J. Org. Chem. 2004, 3865Ϫ3871
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3869

N. Spreti, L. Brinchi, R. Germani, M. V. Mancini, G. Savelli
FULL PAPER
Table 5. Mol percentages of M2ϩ in the receiving phase in simultaneous transports (Cd2ϩ/Pb2ϩ as function of the pH of the receiving
phase at 25 °C)
Time [h][a]
Cd2ϩ
pH ϭ 3.3
Pb2ϩ
pH ϭ 3.3
pH ϭ 2.5
pH ϭ 0.3
pH ϭ 2.5
pH ϭ 0.3
4
24
48
16
73
77
21
66
73
7
13
14
7
44
55
10
42
46
8
24
28
[a]
Conditions: source phase: 8 mL of a solution of 2.5 ϫ 10Ϫ3  CdCl2 and 2.5 ϫ 10Ϫ3  PbCl2 in 0.15  acetate buffer pH ϭ 4.68;
membrane: 20 mL of 1 ϫ 10Ϫ3  carrier in chloroform; receiving phase: 8 mL of HCl at different concentrations.
Experiments to evaluate the capability of the carrier to transport was therefore examined, and two different separ-
transport Hg2ϩ efficiently for several cycles were therefore ation procedures were optimized.
carried out, both aqueous phases being removed and re-
For Cd2ϩ/Pb2ϩ and Hg2ϩ/Cu2ϩ mixtures, the inefficacy
placed with fresh solutions after the first transport process. of separation was due to two different reasons: (i) the car-
Figure 4 reports the profile of µmols of Hg2ϩ in the source rier is able to extract the metal ions with similar degrees of
phase and receiving phase observed over three consecutive ability, and (ii) the carrierϪmetal ion complexes require the
cycles.
same acidity of the receiving phase to release the metal ions.
Finally, the capability of the carrier to transport Hg2ϩ
efficiently over consecutive cycles increases its potential ap-
plicability to selective removal of mercury from waste-
water solutions.
Experimental Section
Membrane ‘‘Uptake’’: A glass test tube, fitted with a screw cap, was
filled with both the source phase (3 mL, 5 ϫ 10Ϫ3  MCl2) and
the organic membrane (3 mL, 1 ϫ 10Ϫ3  carrier). After a vigorous
shake, the organic layer was slowly stirred (250 rpm) for 12Ϫ18 h.
Quantitative determinations of the metal ions in the aqueous phase
were then performed.
Transport Experiments: The glass transport cell used in these stud-
ies has already been reported in an earlier publication.[19] An or-
ganic solution (20 mL, 1 ϫ 10Ϫ3  carrier) was used as liquid mem-
brane separating an aqueous solution (source phase) containing the
metal salt or a mixture of metal salts (8 mL, 5 ϫ 10Ϫ3  MCl2),
and a receiving phase (8 mL, 3.2 ϫ 10Ϫ3  or 0.5  HCl). The
liquid membrane phase was slowly stirred (250 rpm) with the aid
of a Teflon-coated magnetic bar. The metal ion concentrations in
both aqueous phases were monitored as a function of time.
Figure 4. µmols of Hg2ϩ in the source phase (filled squares) and
receiving phase (open squares) in three consecutive cycles; [bis-pO-
DET] ϭ 1 ϫ 10Ϫ3  in chloroform; [HgCl2] ϭ 5 ϫ 10Ϫ3  in 0.15
 acetate buffer, pH ϭ 4.68; receiving phase: 0.5  HCl, pH ϭ 0.3;
T ϭ 25 °C.
Over 90% of the Hg2ϩ was transported in every process
and only in the last cycle did the rate of the metal ion trans-
port seem to be slower than in the first ones.
Metal Ion Analysis: The concentrations of the metal ions in both
partition and single ion transport experiments were measured by
the sensitive spectrophotometric method described in an earlier
publication.[19] A diode array Hewlett Packard 8452 A spectropho-
tometer was used for the quantitative determinations. Graphite fur-
nace atomic absorption spectrophotometers for the measurements
of metal ion concentrations in competitive experiments were used:
PerkinϪElmer Model Zeeman 3030 for Cu2ϩ, Cd2ϩ and Pb2ϩ and
Varian-Spectra 200 with a GTA100 graphite furnace for Hg2ϩ de-
terminations. Each reported result is given as the arithmetic mean
of three separate experiments; reproducibly was confirmed as Ϯ
5% or better.
Conclusion
The transport capability of a new carrier, bis-pODET, for
the selective removal of heavy metal ions is reported. Cad-
mium, lead and mercury were chosen for this study, owing
to their high toxicity, along with copper for comparative
purposes.
Single ion transport experiments showed that the carrier
was able to complex and transfer all the metal ions from
the source phase to the receiving phase. On the other hand,
bis-pODET displayed different affinities for the selected me-
tal ions, in the order Hg2ϩ Ͼ Cu2ϩ Ͼ Cd2ϩ ഠ Pb2ϩ, and
the release of the metal ions from the membrane was depen-
Materials: Metal ion salts, 1,5-diphenylthiocarbazone, 4-hydroxy-
benzaldehyde, octyl bromide and diethylenetriamine were analyti-
cal grade, purchased from Aldrich and used as received. Solvents,
purchased from CarloϪErba, were ACS-certified quality. 1H NMR
dent on the pH of the receiving phase. Selective metal ion spectra were recorded with a Bruker instrument at 200 MHz;
3870  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org
Eur. J. Org. Chem. 2004, 3865Ϫ3871

Selective Removal of Heavy Metal Ions from Aqueous Solutions
FULL PAPER
chemical shifts (δ ppm) are referred to internal standard (CH3)4Si.
Melting points were measured with a Büchi 510 melting point ap-
paratus and are uncorrected. IR spectra were recorded with a Para-
then transferred to a separating funnel, treated with water (200 mL)
and extracted twice with diethyl ether. The organic layer was
washed with NaOH (10%) and finally with water until neutrality
gon 500 FT-IR PerkinϪElmer spectrophotometer. GLC was car- was reached. The organic solution, after drying with Na2SO4, was
ried out with an HP 5890 (FID) instrument with a 30 m ϫ 0.53 mm
capillary column (HP-5, 1.5-µm film thickness) or a 15 m ϫ
0.53 mm capillary column (HP-INNOVAX, 1 µm film thickness)
with on-column injection. All column chromatography was carried
concentrated in a rotary evaporator, giving a colourless oil, which
solidified on a water/ice bath. The yield of this product was 70%
(5.67 g). H NMR (CDCl3): δ ϭ 0.86 (t, 6 H, 2 CH3), 1.13Ϫ1.42
1
(m, 20 H, 10 CH2), 1.58Ϫ1.76 (m, 7 H, 2 CH2 ϩ 3 NH), 2.67 (s,
out on silica gel from Riedel-de Hae¨n (0.032Ϫ0.063 mm), and with 8 H, 2 N(CH2)2N], 3.66 (s, 4 H, 2 NϪCH2ϪAr), 3.90 (t, 4 H, 2
petroleum ether/Et2O (8:2) as eluent. TLC was carried out on silica
gel (ALUGRAM SIL G/UV254) with petroleum ether/Et2O (8:2)
as eluent.
CH2ϪOϪAr), 6.81 (d, 4 H, Ar), 7.15 (d, 4 H, Ar) ppm.
C34H57N3O2 (539.8): calcd. C 75.64, H 10.64, N 7.78; found C
75.60, H 10.59, N 7.82.
Synthesis of the Carrier
Acknowledgments
Preparation of p-Octyloxybenzaldehyde: Freshly recrystallized
(from water) and dried 4-hydroxybenzaldehyde (0.24 mol, 29.31 g),
anhydrous potassium carbonate (0.29 mol) and acetonitrile
(400 mL) were placed in a 1-L round-bottomed flask fitted with a
reflux condenser, nitrogen inlet, mechanical stirrer, and pressure-
equalizing dropping funnel. The reaction mixture was heated to
reflux and stirred. As reflux was reached, octyl bromide (0.23 mol,
44.42 g) was added dropwise over 1 h. The mixture was then heated
at reflux for 12 h. After the mixture had cooled to room tempera-
ture, it was transferred to a separating funnel (2 L), treated with
water (500 mL), and extracted with petroleum ether. The organic
layer was washed twice with NaOH (10%) and finally with water
until neutrality was reached. The organic solution, after drying
with Na2SO4, was concentrated in a rotary evaporator, giving a
pale yellow oil. The yield of this oil was 98% (52.82 g), and the
purity by GLC 98%. No treatment of the oil followed. 1H NMR
(CDCl3): δ ϭ 0.88 (t, 3 H, CH3), 1.01Ϫ1.75 (m, 10 H, 5 CH2),
1.80Ϫ1.95 (m, 2 H, CH2), 4.06 (t, 2 H, CH2O), 7.00 (d, 2 H, Ar),
7.85 (d, 2 H, Ar), 9.01 (s, 1 H, COH) ppm. C15H22O2 (234.3): calcd.
C 76.88, H 9.46, found, C 76.90, H 9.50.
The authors are thankful to Mr. Marcello Centofanti for his techni-
cal assistance in atomic absorption spectrophotometer mercury de-
termination. Support of this work by the Ministero dell’Istruzione,
`
Universita e Ricerca, Rome (COFIN 2003) is gratefully acknowl-
edged.
[1]
R. A. Goyer, ‘‘Toxic Effects of Metals’’ in: Casarett and Doull’s
Toxicology: The Basic Science of Poisons (Ed.: C. D. Klassen),
McGraw-Hill, New York, 1996.
L. Friberg, G. F. Nordberg, C. Nordman (Eds.) in: Handbook
[2]
on the Toxicology of Metals, Elsevier, Amsterdam, 1986, vol. 2.
H. G. Seiler, H. Sigel (Eds.), in Handbook on Toxicity of Inor-
[3]
ganic Compounds, Marcel Decker, New York, 1988.
R. A. Goyer, M. G. Cherian (Eds.), in Toxicology of Metals:
[4]
Biochemical Aspects, Springer-Verlag, Heidelberg, 1995.
B. Sarkar (Ed.) in: Biological Aspects of Metals and Metal-
[5]
Related Diseases, Raven Press, New York, 1983.
R. D. Noble, J. D. Way, in Liquid Membranes: Theory and Ap-
[6]
plications, American Chemical Society, Washington D. C.,
1987.
R. M. Izatt, J. S. Bradshaw, J. D. Lamb, R. L. Brueing, in
[7]
Liquid membranes: Chemical Applications (T. Araki, H. Tsu-
kube, Eds.), CRC Press, Boca Raton, 1990.
Preparation of 2,2Ј-Bis(p-octyloxybenzylimine)diethylamine [NЈ-(p-
Octyloxybenzylidene)-N-{2-[(p-octyloxybenzylidene)amino]ethyl}-
1,2-ethanediamine]: Diethylenetriamine (0.086 mol, 8.87 g) was ad-
ded to a solution of p-octyloxybenzaldehyde (0.178 mol, 41.71 g)
in chloroform (200 mL) in a 500-mL round-bottomed flask fitted
with a DeanϪStark separator. The mixture was heated to reflux
while the water/chloroform azeotrope was distilled. Once distil-
lation had finished, a reflux condenser was substituted for the sep-
[8]
J. M. Lehn, in Supramolecular Chemistry. Concepts and Per-
spectives, VCH, Weinheim, 1995.
[9]
R. M. Izatt, J. Inclusion Phenom. Macrocyclic Chem. 1997,
29, 197Ϫ220.
[10]
M. Shamsipur, M. H. Mashhadizadeh, G. Azimi, Sep. Purif.
Technol. 2002, 27, 155Ϫ161.
[11]
A. Safavi, E. Shams, J. Membr. Sci. 1998, 144, 37Ϫ43.
[12]
J. S. Kim, M. H. Cho, J. H. Cho, J. H. Lee, R. A. Bartsch, Y.
˚
arator, and molecular sieves (4 A; 15 g) were added. After the sys-
I. Lee, I. H. Kim, Talanta 2000, 51, 99Ϫ105.
J. S. Kim, M. H. Cho, S. C. Lee, J. H. Pang, J. H. Lee, A. Ohki,
tem had been heated at reflux for 12 h and cooled to room tem-
perature, molecular sieves were removed by filtration. The organic
solution was concentrated in a rotary evaporator to give a yellow
semisolid. This was crystallized from petroleum ether, and recrys-
tallized from methanol to give, after drying under vacuum, a crys-
talline white solid. The yield of this solid was 80% (36.86 g). M.p.
48Ϫ49 °C. 1H NMR (CDCl3): δ ϭ 0.88 (t, 6 H, 2 CH3), 1.20Ϫ1.45
(m, 20 H, 10 CH2), 1.69Ϫ1.82 (m, 4 H, 2 CH2), 2.97 (t, 4 H, 2
NϪCH2), 3.71 (t, 4 H, 2 ϭNϪCH2), 3.90 (t, 4 H, 2 CH2ϪO-Ar),
6.80 (d, 4 H, Ar), 7.57 (d, 4 H, Ar), 8.22 (s, 2 H, 2 CHϭN) ppm.
[13]
Talanta 1999, 49, 69Ϫ75.
[14]
H. Parham, M. Shamsipur, J. Membr. Sci. 1994, 95, 21Ϫ27.
[15]
S. Upadhyay, V. A. Varghese, L. Mishra, S. N. Singh, S. C.
Srivastava, J. Membr. Sci. 1994, 93, 149Ϫ156.
[16]
M. Akhond, M. Shamsipur, J. Membr. Sci. 1996, 117,
221Ϫ226.
[17]
F. M. Menger, J.-J. Lee, J. Org. Chem. 1993, 58, 1909Ϫ1916.
L. Brinchi, P. Di Profio, R. Germani, G. Savelli, N. Spreti, Eur.
J. Org. Chem. 2002, 930Ϫ937.
L. Brinchi, R. Germani, M. V. Mancini, G. Savelli, N. Spreti,
[18]
[19]
Eur. J. Org. Chem. 2004, 1330Ϫ1335.
Preparation of 2,2Ј-Bis(p-octyloxybenzyl)diethylenetriamine {NЈ-(p-
Octyloxybenzyl)-N-[2-(p-octyloxybenzylamino)ethyl]-1,2-ethane-
diamine}: NaBH4 (0.035 mol) was added in small aliquots to a
stirred mixture of 2,2Ј-bis(p-octyloxybenzylimine)diethylamine
(0.015 mol, 8.04 g) in methanol (200 mL), in a round-bottomed
flask fitted with a reflux condenser. The mixture was heated at
reflux for 10 h and allowed to cool to room temperature. It was
[20]
M. B. Gholivand, S. Khorsandipoor, J. Membr. Sci. 2000,
180, 115Ϫ120.
[21]
A. Safavi, E. Shams, Talanta 1999, 48, 1167Ϫ1172.
[22]
A. Safavi, F. Peiravian, E. Shams, Sep. Purif. Technol. 2002,
26, 221Ϫ226.
[23]
Q. Li, Q. Liu, X. Wei, Talanta 1996, 43, 1837Ϫ1842.
Received April 2, 2004
Eur. J. Org. Chem. 2004, 3865Ϫ3871
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3871