Three-phase transport studies of benzimidazole ligands bearing hydrophobic N-alkyl substituents: Transport of metal(II) ions (Ni, Cu, Zn, Cd) in single- and multiple-ion experiments

Article abstract of DOI:10.1039/a705387h

New ligands having two benzimidazole units, or one benzimidazole and one carboxylic acid unit, the two functions linked by a chain, and bearing hydrophobia N-alkyl substituents have been tested as transfer agents for metal(II) ions (Ni, Cu, Zn, Cd) using

Full text of DOI:10.1039/a705387h

View Article Online / Journal Homepage / Table of Contents for this issue
Three-phase transport studies of benzimidazole ligands bearing
hydrophobic N-alkyl substituents: transport of metal(II) ions (Ni, Cu,
Zn, Cd) in single- and multiple-ion experiments†
a
b
b
,a
Craig J. Matthews, Troy A. Leese, Derek Thorp and Joyce C. Lockhart*
a
Department of Chemistry, University of Newcastle Upon Tyne, UK NE1 7RU
Zeneca Specialties, PO Box 42, Hexagon House, Blackley, Manchester, UK M9 8ZS
b
New ligands having two benzimidazole units, or one benzimidazole and one carboxylic acid unit, the two
functions linked by a chain, and bearing hydrophobic N-alkyl substituents have been tested as transfer agents for
metal() ions (Ni, Cu, Zn, Cd) using three-phase transport techniques. The transport was pH-dependent, so all
comparisons were made with buffered feed and receiving phases at fixed pH. Variations in the nature and number
[
single and double N-alkyl-substituted bis(benzimidazole) ligands] of N-alkyl substituents, linker chain and its
donor atoms were explored and resultant transport trends are discussed. While single-ion transport results were
ambivalent (singly alkylated 2-oxapropane derivatives being the best extractants for copper, and the analogous
2
-thiapropane ligands for zinc), in multiple-ion experiments copper was always transported in preference to zinc.
Even for ligands with a longer linking chain (based on 2,5-dioxaoctane) the order for extraction of metal() ions
was always Cu > Zn > Cd and the transport of Ni was negligible. From the benzimidazole ligands screened, those
having a carboxylic acid moiety reveal the highest metal-ion selectivity.
The measurement of transport rates through bulk liquid mem-
branes is a well established technique and has been used exten-
FO
OR
1
sively over the last twenty years to screen new and novel carriers
(
transfer agents) with respect to selectivity. The transport of
L
Organic membrane
ML
substrates through membranes has received much attention
2–6
because of the potential applications of the technique. It has
been used to model biological membrane-transport systems, to
develop ion-selective electrodes, to separate metal ions and to
treat waster-water flows. The transport of a single substrate
species through a bulk liquid membrane by a suitable carrier
M
M
Aqueous feed
Aqueous receive
(
L) is depicted in Scheme 1. The substrate (M) is initially
located in the aqueous feed in the carrier (L) in the organic
membrane with no substrate residing in the receiving phase.
The rate at which the substrate is transported through the bulk
liquid membrane is expressed as J , the number of moles of
substrate transported per 24 h × 10 . The factors that influence
m
7
L + M
ML
Ke = [ML]/[M][L]
the transport rate Jm and which have been altered during this
study are the carrier and metal ion.
FO
OR
Scheme 1 Representation of carrier-facilitated diffusion of a single
substrate across an organic membrane. Aqueous ‘feed’ and aqueous
receive’ refer to the adopted industrial nomenclature for in and out
phases respectively; FO and OR represent the interfacial areas between
the aqueous feed–organic membrane and organic membrane–aqueous
receive phases respectively
In this work a range of new bis(benzimidazole) (3f, 3h–10f,
0h, Table 1, Scheme 2) and benzimidazole carboxylic acid (11–
5, Table 1, Scheme 3) ligands bearing hydrophobic N-alkyl
1
1
‘
substituents have been screened for potential metal-ion selectiv-
7
ity, complementing earlier work on N-octadecylbenzimidazoles
7
1
and 2. Rushton and co-workers assumed that passive trans-
port applied for fully alkylated bis(benzimidazole) carriers.
However, our preliminary results indicated an effect of pH on
transport of metal() ions. Thus the comparisons are all made
at fixed pH although benzimidazoles of different pK are
involved. To obviate the problems of comparison between dif-
ferent ligands the pH (receiving and feed phases) was buffered
using a sodium acetate buffer. Lindoy and co-workers have
adopted this approach when screening Schiff-base-derived
macrocycles for metal-ion selectivity. Sodium cations and acet-
ate anions present in the receive and source (feed) phases are
believed to have little or no effect on the transport rates or the
metal-ion selectivity. Benzimidazole ligands show no affinity for
sodium cations, especially in the presence of transition-metal
ions, and acetate is known to be a poorer co-transported anion
than fluoride. The only transport considered, therefore, is
between the benzimidazole ligand and the metal nitrate salt
introduced into the feed solution. In the impending Discussion
section direct comparisons are made only between the benz-
imidazole ligands developed in this study.
11
8,9
10
Experimental
Reagents and solvents used were of commercially available
reagent grade quality. The alcohols 2-hexyldecanol, 2-octyl-
dodecanol and isooctadecanol were received from Zeneca
Specialties and were of industrial grade quality. 3,6,9-
Trioxaundecanedioic acid was supplied by Hoescht chemical
company. Dimethylformamide (dmf) was distilled over calcium
†
Supplementary data available: relative differences in transport rates.
For direct electronic access see http://www.rsc.org/suppdata/dt/1998/
9/, otherwise available from BLDSC (No. SUP 57310, 3 pp.) or the
7
RSC Library. See Instructions for Authors, 1998, Issue 1 (http://
www.rsc.org/dalton).
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88
79

View Article Online
Table 1 N-Alkylated hydrophobic bis(benzimidazole) and benzimidazolecarboxylic acid ligands 3f, 3h–10f, 10h, 11–15
a
3 a,b
4 a,b
N,NЈ-dialkyl
N-Alkyl
Carboxylic acid
X
R
R
3
4
5
6
7
8
9
f
f
f
f
f
f
f
3h
4h
5h
6h
7h
11
—
—
—
12
13
14
15
S
S
O
O
(CH ) CH
(CH ) CH
2 5
2
7
3
3
3
3
3
3
3
3
(CH ) CH
(CH ) CH
2
9
2 7
(CH ) CH
(CH ) CH
2 5
2
7
(CH ) CH
(CH ) CH
2
9
2 7
O
(CH ) CH(CH )CH C(CH )
CH(CH )CH C(CH )
2
2
3
2
3
3
3
2
3
3
b
b
8h
OCH CH O
OCH CH OCH CH O
OC H O-o
(CH ) CH
(CH ) CH
2
2
2
7
3
3
3
2
5
3
3
3
9h
10h
(CH ) CH
(CH ) CH
2
2
2
2
2
7
2 5
1
0f
(CH ) CH
(CH ) CH
2 5
6
4
2
7
a
5
6
b
1
2
For N-alkylated hydrophobic benzimidazolecarboxylic acid ligands (11–15): R = CH CH , R = (CH ) CH (Scheme 3). R = R = H, except for
2
3
2
3
3
1
2
compounds 8f and 8h for which R = R = CH (Scheme 2).
3
R1
R1
R2
O
S
N
N
CH2 CH R3
R4
X
X
+
+
2 CH3
O
R2
N
H
N
H
O
(i ), (ii )
R1
R2
R1
R2
R1
R2
R1
R2
N
N
N
N
N
N
N
X
N
H
R4
R4
R4
R3
R3
R3
Half alkylated bis-benzimidazole
Fully alkylated bis-benzimidazole
Scheme 2 Synthesis of N-alkylated hydrophobic bis(benzimidazole) ligands. (i) Dry dimethylformamide, dry powdered K CO , 80 ЊC, 72 h;
2
3
(
ii) flash chromatography
NH2
performed on a UNICAM 701 optical emission spectro-
photometer or at Zeneca Specialties. All three-phase transport
experiments were performed in triplicate in an apparatus simi-
+
2 HO2C
X
CO2H
NH
7
lar to that described previously; the reproducibilities for three-
(
+/–)
phase metal-ion transport are estimated as ±15ꢀ respectively.
(
i ), (ii ), (iii )
R6
In comparisons of J
error are represented as approximately equivalent by the sign
≈). Blank experiments (no carrier present in the organic mem-
m
, values that lie within the boundaries of
R5
(
N
N
brane) were performed to determine membrane leakage; in each
X
CO2H
nϩ
case the amount of leakage was regarded as negligible (M ca.
<
0.05 ppm). Metal-ion concentrations in aqueous solutions
(
+/–)
were determined using ICP techniques. 1,3-Bis(benzimidaz-
ol-2-yl)-2-thiapropane,
pane,
hexane
12a
R6
1,3-bis(benzimidazol-2-yl)-2-oxapro-
R5
1
2a
1,6-bis(5,6-dimethylbenzimidazol-2-yl)-2,5-dioxa-
+
12b
12c
and 1,2-bis(benzimidazol-2-ylmethyoxy)benzene
12d
were prepared via the Phillips
Bis(N-stearylbenzimidazol-2-yl)-2-oxapropane 1 and 1,3-bis-
5,6-dimethyl-N-stearylbenzmidazol-2-yl)-2-oxapropane 2 have
condensation reaction. 1,3-
N
N
N
N
X
(
7
been reported elsewhere and N-2-(±)-ethylhexylamino-o-
phenylenediamine was prepared according to a method
described by Zeneca Specialties.
(
+/–)
(+/–)
R6
R6
R5
R5
Scheme 3 Synthesis of N-branched alkylbenzimidazolecarboxylic
acids. (i) Finely ground reagents; (ii) molten mass heated for 2 h with
efficient stirring; (iii) flash chromatography [35–70 µm silica, eluent
CH Cl –CH OH (10:1)]
Synthesis
1,9-Bis(benzimidazol-2-yl)-2,5,8-trioxanonane. To 3,6,9-tri-
oxaundecanedioic acid (11.56 g, 0.05 mol) was added o-phenyl-
enediamine (11.26 g, 0.10 mol); the two reagents were finely
ground and intimately mixed. The mixture was stirred and
heated to 140–160 ЊC in a molten fluid state for 4 h and cooled to
room temperature to produce a brown glass-like solid which
2
2
3
hydride and used immediately. Melting points were determined
with a Gallenkamp apparatus and are uncorrected. Proton and
13
C NMR spectra were run on Bruker WP200, WB300 and
AMX500 spectrometers; chemical shifts (δ) are reported in
parts per million (ppm) relative to deuteriated solvents as
internal standards. Merck silica gel 60F254 plates for thin-layer
chromatography were developed in the solvent systems stated.
Elemental analysis was performed on a Carlo Erba 1106 Ele-
mental Analyser; fast atom bombardment (FAB) and electron
impact (EI) mass spectra were obtained on a Kratos MS80 RF
instrument. Inductively coupled plasma (ICP) analysis was
3
was extracted with portions of hot ethanol (5 × 50 cm ). The
ethanolic solution was refluxed with decolourising charcoal for
1 h, filtered whilst hot and reduced in vacuo to yield a pale
yellow oil. The oil was dissolved in dichloromethane (3 × 200
3
3
cm ) and washed with water (5 × 150 cm ) until the orange
colouration was no longer evident in the aqueous washings. The
organic extracts were dried over anhydrous magnesium sulfate,
8
0
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88

View Article Online
filtered and slowly reduced in volume in vacuo until the first
signs of precipitation. The resultant solution was kept at Ϫ20 ЊC
overnight which resulted in a white precipitate forming. Yield
previously dried under vacuum at 60 ЊC was added finely
ground dry potassium carbonate (21.2 g, 0.15 mol) under a
nitrogen atmosphere in freshly distilled dry dimethylformamide
3
6
.12 g (36ꢀ), m.p. 141 ЊC (Found: C, 65.5; H, 5.7; N, 14.9.
(500 cm ). The suspension was gently warmed to 80 ЊC for 4 h
1
C H N O requires C, 65.6; H, 6.05; N, 15.3ꢀ). H NMR [200
and 2-hexyldecyl tosylate (62.7 g, 0.16 mol) was added slowly
over 2 h. The suspension was stirred for 72 h at 80 ЊC under
a nitrogen atmosphere then cooled to room temperature and
filtered through Celite to yield an orange solution which was
concentrated in vacuo to yield a golden oil. The oil was diluted
20
22
4
3
MHz, (CD ) SO]: δ 3.71–3.83 (m, 8 H, CH CH OCH CH ),
3
2
2
2
2
2
5,6
4
.81–4.84 (s, 4 H, C H N CH O), 7.25–7.27 (m, 2 H, H of
7 5 2 2
4,7
aryl, AAЈBBЈ, in slow exchange), 7.57–7.61 (m, 4 H, H of
aryl, AAЈBBЈ in slow exchange) and 12.6 (br s, 2 H, NH).
3
(
(
FAB) mass spectrum; m/z (assignment, relative intensity) 367
with water (500 cm ), extracted into dichloromethane (5 × 500
ϩ
3
3
M ϩ H , 40ꢀ). EI mass spectrum: m/z (assignment, relative
cm ) and washed with water (5 × 300 cm ). The combined
organic extracts were dried over anhydrous magnesium sulfate,
filtered and reduced in volume in vacuo to yield a golden oil.
Thin-layer chromatography of the crude oil indicated the pres-
ence of two major alkylated products. The oil was purified via
flash chromatography [35–70 µm silica; eluent ethyl acetate–
light petroleum (b.p. 60–80 ЊC) (40:60)] to yield the fully alkyl-
ated product as a pale yellow oil. Yield (8.54 g, 16ꢀ) (Found: C,
ϩ
intensity) 366.1686 (M , 95ꢀ); C H N O requires 366.1692.
20
22
4
3
The hydrophobic alkyl tosylates (toluene-p-sulfonates) were
13
prepared according to the general procedure outlined below.
2
-Hexyldecyl tosylate. To a stirred solution of 2-hexyl-
decanol (196.2 g, 0.81 mol) in cold dry pyridine (previously
dried over molecular sieves) was added a slight excess of
toluene-p-sulfonyl chloride (158.2 g, 0.83 mol) in small portions
ensuring the temperature did not exceed Ϫ5 ЊC. The solution
was then allowed to stir for 6 h at Ϫ10 ЊC in an ice–salt bath and
then overnight in a cold room. The reaction mixture was
7
7.9; H, 10.6; N, 7.2; S, 4.1. C H N S requires C, 77.6; H, 10.6;
48 78 4
1
N, 7.5; S, 4.3ꢀ). H NMR (200 MHz, CDCl ): δ 0.76–0.87 [m,
3
1
2 H, 4 × CH (R)], 1.15–1.23 [m, 48 H, alkyl (R)], 1.84–1.98
3
[
br m, 2 H, CH, (R)], 3.91–3.94 [d, 4 H, NCH (R)], 4.05 (s, 4 H,
2
4–6
3
poured over an ice (3 l) and concentrated HCl (700 cm ) mix-
CH S), 7.15–7.37 (m, 6 H, H of aryl) and 7.58–7.67 (m, 2 H,
H of aryl). C NMR (50 MHz, CDCl ): δ 150.8 (C ), 142.6
(
4
2
ture ensuring the pH ≈ 1 with Congo Red indicator paper, at a
temperature <10 ЊC. The mixture was then stirred for 1 h and
allowed to settle, upon which the tosylated alcohol separated to
the top of the aqueous layer to yield a pale yellow oil. The
aqueous waste was removed and the oil extracted into dichloro-
7
13
2
3
9
8
5
6
4
7
C ), 135.9 (C ), 122.7 (C ), 122.2 (C ), 119.8 (C ), 110.1 (C ),
8.6 (CH S), 38.6 (NCH ), 35.6–14.0 (remaining C atoms of
2
2
hydrophobic alkyl chains). EI mass spectrum: m/z 742.5980
ϩ
(
M , 30ꢀ); C H N S requires 742.5947.
48 78 4
3
methane (450 cm ). The organic extracts were washed with 0.1
3
3

HCl (3 × 200 cm ), saturated brine (3 × 200 cm ) and finally
1-Benzimidazolyl-3-(N-2-hexyldecyl)benzimidazolyl-2-thia-
3
distilled water (3 × 200 cm ). The organic extracts were dried
with magnesium sulfate, filtered and reduced in volume in vacuo
to yield a viscous pale yellow oil. Yield 272.5 g, 85ꢀ (Found: C,
propane 3h. The half-alkylated bis-benzimidazole product was
isolated from the column as a pale yellow oil. Yield 14.54 g,
40ꢀ. (Found: C, 73.8; H, 8.5; N, 10.4; S, 5.9. C H N S
3
2
46
4
1
6
8
2
9.7; H, 10.6; S, 8.1. Calc. for C H O S: C, 69.7; H, 10.2; S,
requires C, 74.1; H, 8.9; N, 10.8; S, 6.2ꢀ). H NMR (200 MHz,
23
40
3
1
.1ꢀ). H NMR (200 MHz, CDCl ): δ 0.66–0.71 [m, 6 H,
CDCl ): δ 0.60–0.72 [m, 6 H, 2 × CH (R)], 1.04–1.11 [m, 24 H,
3
3
3
× CH (R)], 1.01–1.05 [m, 24 H, alkyl (R)], 1.35–1.42 [m, 1 H,
alkyl (R)], 1.79–1.82 [m, 1 H, CH (R)], 3.60 (s, 2 H, CH S), 3.77
3
2
CH (R)], 2.27 (s, 3 H, aryl-CH ), 3.71–3.74 [d, 2 H, OCH2
(s, 2 H, CH S), 3.88–3.96 [m, 2 H, NCH (R)], 7.04–7.21 (m, 5
3
2
2
4
Ј,5,5Ј,6,6Ј
4,7
(
R)], 7.12–7.18 (d, 2 H, aryl) and 7.59–7.63 (d, 2 H, aryl). EI
H, H
of aryl), 7.46–7.49 (m, 2 H, H of aryl), 7.51–7.70
7Ј 13
ϩ
mass spectrum: m/z 241.2522 (M Ϫ C H O S, 30ꢀ); C H O
(m, 1 H, H of aryl) and 12.98 (br s, 1 H, NH. C NMR (50
7
7
2
16 33
2Ј
2
9Ј
9
8
requires 241.2531.
MHz, CDCl ): δ 151.6 (C ), 150.6 (C ), 141.5 (C ), ‡(C ), ‡(C ),
3
8Ј
5,6
5Ј
6Ј
4Ј
1
‡
1
35.8 (C ), 123.5 (C ), 122.9 (C ), 122.5 (C ), 119.5 (C ),
2
-Octyldodecyl tosylate. From 2-octyldodecanol (241.8 g,
4 7 7Ј
(C ), ‡(C ), 110.6 (C ), 48.9 (CH S), 38.7 (NCH ) and 32.0–
2
2
0
.81 mol) and toluene-p-sulfonyl chloride (158.2 g, 0.83 mol) in
4.3 (remaining C atoms of hydrophobic alkyl chains). FAB
mass spectrum: m/z 519 (M ϩ H , 100ꢀ). EI mass spectrum:
m/z 518.3424 (M , 30ꢀ); C H N S requires 518.3443.
the same manner as described above a viscous golden oil was
obtained. Yield 273.0 g, 74ꢀ (Found: C, 72.3; H, 9.5; S, 6.1.
ϩ
ϩ
32
46
4
1
Calc. for C H O S: C, 71.6; H, 10.7; S, 7.1ꢀ). H NMR (200
27
48
3
MHz, CDCl ): δ 0.52–0.64 [m, 6 H, 2 × CH (R)], 0.91–0.99 [m,
1,3-Bis(N-2-octyldodecylbenzimidazol-2-yl)-2-thiapropane 4f.
From the preparation using 1,3-bis(benzimidazol-2-yl)-2-thia-
propane (20.5 g, 0.07 mol) and 2-octyldodecyl tosylate (72.9 g,
0.16 mol) the crude oil was purified via flash chromatography
[35–70 µm silica; eluent ethyl acetate–light petroleum (b.p. 60–
3
3
3
2 H, alkyl (R)], 1.28–1.42 [m, 1 H, CH, (R)], 2.17 (s, 3 H, aryl-
CH ), 3.62–3.65 [d, 2 H, OCH (R)], 7.05–7.09 (d, 2 H, aryl),
3
2
and 7.49–7.57 (d, 2 H, aryl). EI mass spectrum: m/z 297.3152
ϩ
(
M Ϫ C H O S, 30ꢀ); C H O requires 297.3157.
7 7 2 20 41
8
0 ЊC)] to give a viscous golden oil. Yield 6.54 g, 11ꢀ (Found:
Isooctadecyl tosylate. From isooctadecanol (179.1 g, 0.81
C, 78.4; H, 10.5; N, 7.0; S, 3.9. C H N S requires C, 78.6; H,
1
5
6
94
4
mol) and toluene-p-sulfonyl chloride (158.2 g, 0.83 mol) in the
same manner as described above was obtained a viscous golden
oil. Yield 212.0 g, 62ꢀ (Found: C, 70.5; H, 9.1; S, 7.4. Calc. for
1
1.1; N, 6.6; S, 3.8ꢀ). H NMR (200 MHz, CDCl ): δ 0.70–0.88
3
[
m, 12 H, 4 × CH (R)], 1.19–1.27 [m, 64 H, alkyl (R)], 1.85–
3
1
4
2
1
.97 [br, m, 2 H, CH (R)], 3.92–3.96 [d, 4 H, NCH (R)], 4.06 (s,
1
2
C H O S: C, 70.7; H, 10.4; S, 7.6ꢀ); H NMR (200 MHz,
4–6
25
44
3
H, CH S), 7.16–7.34 (m, 6 H, H of aryl) and 7.64–7.72 (m,
H, H of aryl). C NMR (50 MHz, CDCl ): δ 150.8 (C ),
2
CDCl ): δ 0.58–1.36 [m, 35 H, alkyl (R)], 2.13 (s, 3 H, aryl-CH ),
7
13
2
3
3
3
3
7
(
.55–3.72 [m, 2 H, OCH (R)], 7.02–7.06 (d, 2 H, aryl) and
9
8
5
6
4
2
42.6 (C ), 135.9 (C ), 122.8 (C ), 122.1 (C ), 119.8 (C ), 110.1
.47–7.51 (d, 2 H, aryl). EI mass spectrum: m/z 269.2832
7
(
C ), 48.6 (CH S), 38.6 (NCH ) and 32.4–14.3 (remaining C
ϩ
2
2
M Ϫ C H O S, 30ꢀ); C H O requires 269.2844.
7
7
2
18 37
atoms of hydrophobic alkyl chains). (FAB) mass spectrum; m/z
N-Alkylation of the bis-benzimidazoles to yield the half and
fully alkylated hydrophobic bis-benzimidazoles (3f, 3h–10f,
0h) was carried out according to the general procedure out-
ϩ
ϩ
8
55 (M ϩ H , 25ꢀ). (EI) mass spectrum: m/z 854 (M , (5ꢀ);§
C H N S) requires 854.7199.
56
94
4
1
lined below. The singly and the doubly N-alkylated products
were separated from a one-pot reaction mix using flash chrom-
‡
Owing to the prototropy of the unalkylated benzimidazole unit these
carbons are in a state of fast exchange and their signals are not
observed at room temperature on this NMR time-scale.
§ Fast atom bombardment and EI mass spectrometry were obtained on
a Kratos MS80 RF instrument. Owing to the high molecular mass of
the fully alkylated ligands 4f and 6f–10f an accurate mass to four
decimal places was unobtainable.
atography (35–70 µm silica; fully alkylated ligands R ≈ 0.85 and
f
half-alkylated ligands R ≈ 0.35) with the relevant eluent.
f
1
,3-Bis(N-2-hexyldecylbenzimidazol-2-yl)-2-thiapropane 3f.
To 1,3-bis(benzimidazol-2-yl)-2-thiapropane (20.1 g, 0.07 mol)
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88
81

View Article Online
1
-Benzimidazolyl-3-(N-2-octyldodecyl)benzimidazolyl-2-
1-Benzimidazolyl-3-(N-2-octyldodecyl)benzimidazolyl-2-
thiapropane 4h. The half-alkylated bis(benzimidazole) product
was isolated from the column as a pale yellow oil. Yield 17.0 g,
2ꢀ (Found: C, 76.1; H, 10.1; N, 9.4; S, 5.6. C H N S requires
C, 75.2; H, 9.5; N, 9.8; S, 5.6ꢀ). H NMR (200 MHz, CDCl ):
oxapropane 6h. The half-alkylated bis(benzimidazole) product
was isolated from the column as a pale yellow oil. Yield 15.3 g,
46ꢀ (Found: C, 76.4; H, 10.4; N, 9.5. C H N O requires C,
4
36
54
4
36 54
4
1
1
77.4; H, 9.8; N, 10.0ꢀ). H NMR (200 MHz, CDCl ): δ 0.59–
3
3
δ 0.77–0.83 [m, 6 H, 2 × CH (R)], 1.14–1.22 [m, 32 H, alkyl
0.69 [m, 6 H, 2 × CH (R)], 0.83–1.07 [m, 32 H, aryl (R)], 1.50–
3
3
(
R)], 1.79–1.84 [m, 1 H, CH (R)], 3.71 (s, 2 H, CH S), 3.88 (s, 2
1.60 [m, 1 H, CH (R)], 3.93–3.97 [m, 2 H, NCH (R)], 4.64 (s, 2
2
2
4
Ј,5,5Ј,6,6Ј
of
7Ј
H, CH S), 3.99–4.07 [m, 2 H, NCH (R)], 7.15–7.35 (m, 5 H,
H
1
H, CH O), 4.68 (s, 2 H, CH O), 6.97–7.11 (m, 5 H, H
2
2
2
2
4
Ј,5,5Ј,6,6Ј
4,7
4,7
of aryl), 7.40–7.62 (m, 2 H, H of aryl), 7.63–7.78 (m,
aryl), 7.14–7.41 (m, 2 H, H of aryl), 7.51–7.64 (m, 1 H, H of
aryl), 13.03 (br s, 1 H, NH); C NMR (50 MHz, CDCl ) δ
7Ј
13
13
H, H of aryl) and 12.96 (br s, 1 H, NH). C NMR (50 MHz,
3
2Ј
2
9Ј
9
8
2Ј
2
9Ј
8
9
CDCl ): δ 151.6 (C ), 150.6 (C ), 141.5 (C ), ‡(C ), ‡(C ), 135.8
151.1 (C ), 150.4 (C ), 141.8 (C ), ‡(C ), ‡(C ), 135.5 (8Ј), 123.4
5,6 5Ј 6Ј 4Ј 7Ј 4
3
8
Ј
5,6
5Ј
6Ј
4Ј
4
(
C ), 123.5 (C ), 122.9 (C ), 122.5 (C ), 119.5 (C ), ‡(C ),
(C ), 123.3 (C ), 122.6 (C ), 119.6 (C ), 110.6 (C ), ‡(C ),
7
7Ј
7
‡
(C ), 110.6 (C ), 48.9 (CH S), 38.7 (NCH ) and 32.0–14.3
‡(C ), 66.9 (CH O), 65.2 (CH O), 48.5 (NCH ), 32.0–14.3
2
2
2
2
2
(
remaining C atoms of hydrophobic alkyl chains). EI mass
(remaining C atoms of hydrophobic alkyl chains). EI mass
ϩ
ϩ
spectrum: m/z 574.4062 (M , 30ꢀ); C H N S requires
spectrum: m/z 558.4252 (M , 30ꢀ); C H N O requires
36
54
4
36 54
4
5
74.4069.
558.4298.
1
,3-Bis(N-2-hexyldecylbenzimidazol-2-yl)-2-oxapropane 5f.
1,3-Bis(N-2-isooctyldecylbenzimidazol-2-yl)-2-oxapropane 7f.
From the preparation using 1,3-bis(benzimidazol-2-yl)-2-oxa-
propane (15.2 g, 0.055 mol) and added 2-isooctyldecyl tosylate
(52.6 g, 0.12 mol), a crude oily product was purified via flash
chromatography [35–70 µm silica; eluent ethyl acetate–light
petroleum (b.p. 60–80 ЊC)] to give a viscous golden oil 7f. Yield
1.8 g, 4ꢀ (Found: C, 78.7; H, 11.5; N, 7.3. C H N O requires
From the preparation using 1,3-bis(benzimidazol-2-yl)-2-oxa-
propane (16.4 g, 0.06 mol) and 2-hexyldecyl tosylate (53.5 g,
.13 mol), the crude oil was purified via flash chromatography
35–70 µm silica; eluent ethyl acetate–light petroleum (b.p. 60–
0 ЊC)] to give a viscous golden oil 5f. Yield 9.38 g, 22ꢀ (Found:
C, 78.9; H, 9.0; N, 7.2. C H N O requires C, 79.3; H, 10.8; N,
0
[
8
48
78
4
52 86
4
1
1
7
.7ꢀ). H NMR (200 MHz, CDCl ): δ 0.75–0.85 [m, 12 H,
C, 79.7; H, 11.1; N, 7.2ꢀ). H NMR (200 MHz, CDCl ): δ 0.62–
3
3
4
× CH (R)], 1.14–1.22 [m, 48 H, alkyl (R)], 1.89–1.97 [br m, 2
1.43 [m, 68 H, alkyl (R)], 1.66–1.79 [br m, 2 H, CH (R)], 3.75–
4.09 [br m, 4 H, NCH (R)] 4.91 (s, 4 H, CH O), 7.19–7.30 (m, 6
3
H, CH (R)], 3.93–3.97 [d, 4 H, NCH (R)], 4.84 (s, 4 H, CH O),
2
2
7
2
2
4–6
4–6
7
13
7
.14–7.29 (m, 6 H, H of aryl) and 7.67–7.73 (m, 2 H, H of
H, H of aryl) and 7.62–7.77 (m, 2 H, H of aryl). C NMR
13
2
9
2 9 8
aryl). C NMR (50 MHz, CDCl ): δ 150.1 (C ),. 142.6 (C ),
(50 MHz, CDCl ): δ 150.2 (C ), 142.5 (C ), 135.9 (C ), 123.2
3
3
8
5
6
4
7
5
6
4
7
1
35.9 (C ), 123.2 (C ), 122.3 (C ), 120.3 (C ), 110.3 (C ), 65.1
(C ), 122.3 (C ), 119.9 (C ), 110.3 (C ), 65.1 (CH O), 47.6
2
(
CH O), 48.7 (NCH ) and 35.6–14.0 (remaining C atoms of
(NCH ) and 35.6–14.0 (remaining C atoms of hydrophobic
2
2
2
ϩ
hydrophobic alkyl chains). EI mass spectrum: m/z 726.6230
(
alkyl chains). FAB mass spectrum: m/z 784 (M ϩ H , 25ꢀ).
ϩ
ϩ
M , 10ꢀ); C H N O requires 726.6176.
EI mass spectrum: m/z 783 (M , 5ꢀ);§ C H N O requires
48
78
4
52 86
4
7
82.6801.
1
-Benzimidazolyl-3-(N-2-hexyldecyl)benzimidazolyl-2-oxa-
propane 5h. The half-alkylated bis(benzimidazole) product was
1-Benzimidazolyl-3-(N-2-isooctyldecyl)benzimidazolyl-2-oxa-
isolated from the column as a pale yellow oil. Yield 14.0 g,
propane 7h. The half-alkylated bis(benzimidazole) product was
isolated from the column as a pale yellow oil. Yield 15.3 g, 52ꢀ
(Found: C, 76.3; H, 10.1; N, 9.8. C H N O requires C, 76.9; H,
4
7
6ꢀ (Found: C, 76.8; H, 9.5; N, 11.4. C H N O requires C,
32
46
4
1
6.5; H, 9.2; N, 11.1ꢀ). H NMR (200 MHz, CDCl ): δ 0.77–
3
34 50
4
1
0
.87 [m, 6 H, 2 × CH (R)], 1.13–1.24 [m, 24 H, alkyl (R)],
9.5; N, 10.6ꢀ). H NMR (200 MHz, CDCl ): δ 0.47–1.18 [m, 34
3
3
1
4
H
.80–1.99 [m, 1 H, CH (R)], 3.70–3.74 [d, 2 H, NCH (R)],
H, alkyl (R)], 1.75–1.86 [m, 1 H, CH (R)], 3.54–3.86 [br m, 2 H,
NCH (R)], 4.72 (s, 2 H, CH O), 4.74 (s, 2 H, CH O), 7.03–7.13
2
.77 (s, 2 H, CH O), 4.82 (s, 2 H, CH O), 7.14–7.27 (m, 5 H,
2
2
2
2
2
4
Ј,5,5Ј,6,6Ј
4,7
4Ј,5,5Ј,6,6Ј
4,7
of aryl), 7.58–7.66 (m, 2 H, H of aryl), 7.67–7.71
(m, 5 H, H
of aryl), 7.28–7.51 [br m, 2 H, H of aryl],
7Ј 13
7Ј
13
(
m, 1 H, H of aryl) and 13.08 (br s, 1 H, NH). C NMR
7.60–7.63 (m, 1 H, H of aryl) and 13.10 (br s, 1 H, NH). C
2
Ј
2
9Ј
9
2Ј 2 9Ј
(
50 MHz, CDCl ): δ 151.1 (C ), 150.4 (C ), 141.8 (C ), ‡(C ),
NMR (50 MHz, CDCl ): δ 151.3 (C ), 150.6 (C ), 141.9 (C ),
3
9 8 8Ј 5–6 5Ј 6Ј
3
8
8Ј
5,6
5Ј
6Ј
‡
(C ), 135.5 (C ), 123.3 (C ), 123.3 (C ), 122.6 (C ), 119.6
‡(C ), ‡(C ), 135.7 (C ), 123.3 (C ), 123.3 (C ), 122.7 (C ),
4Ј 4 7 7Ј
4
Ј
4
7
7Ј
(
C ), ‡(C ), ‡(C ), 110.6 (C ), 66.9 (CH O), 65.2 (CH O),
119.6 (C ), ‡(C ), ‡(C ), 110.6 (C ), 65.9 (CH O), 65.6 (CH O),
2 2
2
2
4
8.5 (NCH ) and 32.0–14.3 (remaining C atoms of hydro-
47.7 (NCH ) and 32.0–14.3 (remaining C atoms of hydrophobic
alkyl chains). EI mass spectrum: m/z 530.3917 (M , 5ꢀ);
2
2
ϩ
phobic alkyl chains). (FAB) mass spectrum: m/z 519 (M ϩ
H , 100ꢀ). EI mass spectrum: m/z 502.3688 (M , 30ꢀ);
ϩ
ϩ
C H N O requires 530.3985.
34
50
4
C H N O requires 502.3672.
32
46
4
1
,6-Bis(N-2-hexyldecyl-5Ј,6-dimethylbenzimidazol-2-yl)-2,5-
1
,3-Bis(N-2-octyldodecylbenzimidazol-2-yl)-2-oxapropane 6f.
dioxahexane 8f. From the preparation using 1,6-bis(5,6-
dimethylbenzimidazol-2-yl)-2,5-dioxahexane (3.34 g, 8.84
mmol) and added 2-hexyldecyl tosylate (7.96 g, 20.08 mmol),
the crude oil was purified via flash chromatography (35–70 µm
silica; eluent methanol–dichloromethane) to give a viscous
golden oil 8f. Yield 1.32 g, 18ꢀ (Found: C, 78.7; H, 11.3; N, 6.8.
From the preparation using 1,3-bis(benzimidazol-2-yl)-2-oxa-
propane (16.7 g, 0.06 mol) and added 2-octyldodecyl tosylate
64.0 g, 0.14 mol), the crude oil was purified via flash chrom-
atography [35–70 µm silica; eluent ethyl acetate–light petroleum
b.p. 60–80 ЊC)] to give a viscous golden oil 6f. Yield 10.2 g, 20ꢀ
(
(
(
1
1
Found: C, 79.7; H, 11.5; N, 7.0. C H N O requires C, 80.1; H,
1.3; N, 6.7ꢀ). H NMR (200 MHz, CDCl ): δ 0.76–0.81 [m, 12
C H N O requires C, 78.4; H, 11.0; N, 6.8ꢀ). H NMR (200
56
94
4
54 90
4
2
1
MHz, CDCl ): δ 0.80–0.89 [m, 12 H, 4 × CH (R)], 1.15–1.24
3
3
3
H, 4 × CH (R)], 1.09–1.17 [m, 64 H, alkyl (R)], 1.87–1.98 [br
[m, 48 H, alkyl (R)], 1.96–2.07 [br m, 2 H, CH (R)], 2.34–2.36
(d, 12 H, aryl-CH ), 3.57 (s, 4 H, OCH ), 3.98–4.02 [d, 4 H,
3
m, 2 H, CH (R)], 3.93–3.97 [d, 4 H, NCH (R)], 4.81 (s, 4 H,
2
3
2
4–6
7
CH O), 7.11–7.25 (m, 6 H, H of aryl) and 7.63–7.70 (m, 2 H,
H of aryl). C NMR (50 MHz, CDCl ): δ 150.2 (C ), 142.5
C ), 135.9 (C ), 123.2 (C ), 122.3 (C ), 119.9 (C ), 110.3 (C ),
5.1 (CH O), 48.7 (NCH ) and 35.6–14.0 (remaining C atoms
NCH (R)], 4.78 (s, 4 H, C H N CH O), 7.02 (s, 2 H, H of aryl)
2 7 2 2 2
4 13
2
7
13
2
and 7.46 (s, 2 H, H of aryl). C NMR (50 MHz, CDCl ):
3
2 9 8 5 6
3
9
8
5
6
4
7
(
6
δ 149.7 (C ), 140.9 (C ), 134.5 (C ), 132.0 (C ), 130.8 (C ), 120.0
4 7
(C ), 110.4 (C ), 69.2 (C H N CH O), 66.1 (CH O), 48.3
7 2 2 2 2
2
2
of hydrophobic alkyl chains). FAB mass spectrum: m/z 839
(NCH ) and 35.6–14.0 (remaining C atoms of hydrophobic
alkyl chains). EI mass spectrum: m/z 826 (M , 60ꢀ);§
2
ϩ
ϩ
ϩ
M ϩ H , 5ꢀ). EI mass spectrum: m/z 838 (M , 5ꢀ);§ C H -
56
94
N O requires 838.7428.
4
C H N O requires 826.7064.
54 90
4
2
8
2
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88

View Article Online
2
9
1
-(N-2-Hexyldecyl)-5,6-dimethylbenzimidazolyl)-6-(5,6-
NMR (50 MHz, CDCl ): δ 149.5 (C ), 148.3 (aryl), 142.3 (C ),
3
8
5
6
4
dimethylbenzimidazolyl)-2,5-dioxahexane 8h. The half-alkylated
bis(benzimidazole) product was isolated from the column as a
136.0 (C ), 123.1 (C ), 122.6 (aryl), 122.2 (C ), 120.2 (C ), 115.2
(aryl), 110.3 (C ), 65.02 (C H N CH O), 48.4 (NCH ) and 38.3–
7
7
4
2
2
2
pale yellow oil. Yield 1.82 g, 34ꢀ (Found: C, 75.2; H, 9.6; N,
14.1 (remaining C atoms of hydrophobic alkyl chains). EI
1
ϩ
9
.2. C H N O requires C, 75.7; H, 9.7; N, 9.3ꢀ). H NMR
mass spectrum: m/z 819, (M , 100ꢀ);§ C H N O requires
38
58
4
2
54 82
4
2
(
200 MHz, CDCl ): δ 0.85–0.91 [m, 6 H, 2 × CH (R)], 1.24 [br
818.6438.
3
3
s, 24 H, alkyl (R)], 1.78–2.05 [br m, 1 H, CH (R)], 2.34 (s, 6 H,
un-N-alkylated aryl CH ), 2.40–2.44 (d, 6 H, N-alkylated
1-(Benzimidazol-2-ylmethoxyl)-2-(N-2-hexyldecyl)benzimid-
azolylmethoxybenzene 10h. The half-alkylated bis(benzimid-
azole) product was isolated from the column as a pale yellow oil.
Yield 1.86 g, 31ꢀ (Found: C, 77.5; H, 8.5; N, 9.4. C H N O
3
aryl CH ), 3.75–3.89 (m, 4 H, CH O), 4.07–4.11 [d, 2 H, NCH
3
2
2
(
R)], 4.85 (s, 2 H, CH O), 4.99 (s, 2 H, C H N CH O), 7.02 (s, 1
2 7 3 2 2
7Ј 4,7 4Ј
H, H of aryl), 7.25 (br s, 2 H, H of aryl), 7.46 (s, 1 H, H of
aryl) and 13.03 (br s, 1 H, NH). C NMR (50 MHz, CDCl ): δ
1
1
1
38
50
4
2
1
3
1
requires C, 76.7; H, 8.5; N, 9.4ꢀ); H NMR (200 MHz, CDCl ):
3
3
2
Ј
2
9Ј
9
8
8Ј
51.3 (C ), 149.9 (C ), 140.8 (C ), ‡(C ), ‡(C ), 134.3 (C ),
32.5 (C ), 131.4 (C ), 130.7 (C ), 119.7 (C ), ‡(C ), ‡(C ),
δ 0.64–0.72 [m, 6 H, 2 × CH (R)], 1.05–1.15 [br m, 24 H, alkyl
3
5,6
5Ј
6Ј
4Ј
4
7
(R)], 1.87–2.06 [br m, 1 H, CH (R)], 3.93–3.96 [d, 2 H, NCH2
(R)], 5.30 (s, 2 H, C H N CH O), 5.32 (s, 2 H, C H N CH O),
7
Ј
10.6 (C ), 71.5 (C H N CH O), 69.0 (C H N CH O), 67.5
7
3
2
2
7
3
2
2
7
5
2
2
7
5
2
2
(
CH O), 65.3 (CH O), 65.2 (CH O), 48.5 (NCH ) and 32.0–
6.70–6.86 (m, 1 H, aryl), 6.87–6.99 (m, 2 H, aryl), 7.03–7.27 (m,
2
2
2
2
1
4.3 (remaining C atoms of hydrophobic alkyl chains). EI
6 H, aryl), 7.41–7.58 (br m, 1 H, aryl), 7.60–7.66 (br m, 1 H,
ϩ
13
mass spectrum: m/z 602.4630 (M , 20ꢀ); C H N O requires
aryl), 7.68–7.72 (m, 1 H, aryl) and 13.38 (br s, 1 H, NH).
C
38
58
4
2
2Ј
2
6
02.4560.
NMR (50 MHz, CDCl ): δ 150.7 (C ), 150.2 (C ), 149.6 (aryl),
3
9Ј 9 8 8Ј
1
47.5 (aryl), 142.9 (C ), ‡(C ), ‡(C ), 135.5 (C ), 124.6
5,6 5Ј 6Ј
1
,9-Bis(N-2-hexyldecylbenzimidazol-2-yl)-2,5,8-trioxanonane
(C ), 124.6 (C ), 123.5 (C ), 122.8 (aryl), 122.7 (aryl), 119.9
4Ј 4 7 7Ј
9
f. From the preparation using 1,9-bis(benzimidazol-2-yl)-2,5,8-
(C ), ‡(C ), ‡(C ), 115.6 (aryl), 110.6 (C ), 65.9 (C H N -
7 5 2
trioxanonane (3.17 g, 8.66 mmol) and added 2-hexyldecyl
tosylate (7.54 g, 19.01 mmol), the crude oil was purified via flash
chromatography (35–70 µm silica; eluent methanol–dichloro-
methane; 1:10); to give a viscous golden oil 9f. Yield 0.64 g, 8ꢀ
CH O), 65.3 (C H N CH O), 69.3 (CH O), 48.7 (NCH ) and
2 7 5 2 2 2 2
31.9–14.1 (remaining C atoms of hydrophobic alkyl chains) EI
ϩ
mass spectrum: m/z 594.3910 (M , 5ꢀ); C H N O requires
38
50
4
2
594.3934.
(
Found: C, 77.1; H, 10.8; N, 6.4. C H N O requires C, 76.6;
52 86 4 3
1
H, 10.6; N, 6.9ꢀ). H NMR (200 MHz, CDCl ): δ 0.82–0.85 [m,
4-[N-(±)-2-Ethylhexylbenzimidazolyl]-3-thiabutanoic acid 11.
To thiadiglycolic acid [2Ј-thiobis(acetic acid)] (5.12 g, 34.09
mmol) was added N-(±)-2-ethylhexylamino-o-phenylene-
diamine (3.75 g, 17.03 mmol). The mixture was finely ground
and stirred with heating to 140–160 ЊC in a molten fluid
state for 4 h, then cooled to room temperature to produce a
brown glassy solid which was extracted with portions of hot
3
1
2 H, 4 × CH (R)], 1.17–1.24 [m, 48 H, alkyl (R)], 1.98–2.07
3
[
br m, 2 H, CH (R)], 3.50–3.60 (m, 8 H, OCH ), 4.0–4.11 [d, 4
2
H, NCH (R)], 4.79 (s, 4 H, C H N CH O), 7.17–7.29 (m, 6 H,
H
MHz, CDCl ): δ 150.6 (C ), 142.3 (C ), 135.9 (C ), 122.9 (C ),
1
2
7
4
2
2
4
–6
7
13
of aryl) and 7.67–7.72 (m, 2 H, H of aryl). C NMR (50
2 9 8 5
3
6
4
7
22.1 (C ), 120.1 (C ), 110.2 (C ), 70.3 (C H N CH O), 69.6
7 4 2 2
3
(
CH O), 66.1 (CH O), 48.5 (NCH ) and 38.2–14.1 (remaining
ethanol (5 × 50 cm ). The ethanolic solution was refluxed with
2
2
2
C atoms of hydrophobic alkyl chains). FAB mass spectrum: m/z
decolourising charcoal for 1 h, filtered whilst hot and reduced in
ϩ
ϩ
8
16 (M ϩ H , 100ꢀ). EI mass spectrum: m/z 814 (M , 10ꢀ);§
volume in vacuo to yield a pale brown oil. The oil was dissolved
3
C H N O requires 814.6700.
in dichloromethane (3 × 200 cm ), washed with water (5 × 150
52
86
4
3
3
cm ) and dried over anhydrous magnesium sulfate. The organic
1
-Benzimidazolyl-9-(N-2-hexadecylbenzimidazolyl)-2,5,8-tri-
extracts were filtered and reduced in vacuo to yield a brown oil.
Thin-layer chromatography of the crude oil indicated the
presence of two major products [the desired N-alkylated benz-
oxanonane 9h. The half-alkylated bis(benzimidazole) product
was isolated from the column as a pale yellow oil. Yield 0.62 g,
1
7
0
1
4
2ꢀ (Found: C, 73.2; H, 9.2; N, 9.6. C H N O requires C,
imidazole acid R ≈ 0.3 and the unwanted fully condensed N-
36
54
4
3
f
1
3.2; H, 9.2; N, 9.5ꢀ). H NMR (200 MHz, CDCl ): δ 0.74–
alkylated bis(benzimidazole) R ≈ 0.9]. The oil was purified via
3
f
.82 [m, 6 H, 2 × CH (R)], 1.13–1.28 [br m, 24 H alkyl (R)],
flash chromatography (35–70 µm silica; eluent CH Cl –
3
2
2
.93–2.05 [br m, 1 H, CH (R)], 3.60–3.72 (m, 8 H, CH O), 4.01–
CH OH, 10:1) to yield the N-alkylated benzimidazole acid as a
2
3
.05 [d, 2 H, NCH (R)], 4.83 (s, 2 H, C H N CH O), 4.91 (s, 2
colourless oil. The oil was triturated with diethyl ether to yield a
pure white solid 11.¶ Yield 1.08 g, 19ꢀ; m.p. 108 ЊC (Found: C,
64.6; H, 7.6; N, 8.5. C H N O S requires C, 64.6; H, 7.8; N,
2
7
5
2
2
4
Ј,5,5,6,6Ј
H, C H N CH O), 7.04–7.29 (m, 5 H, H
of aryl), 7.31–
7
5
2
2
4,7
7Ј
7
1
.43 [br m, 2 H, H of aryl], 7.64–7.70 (m, 1 H, H of aryl) and
2.80 (br s, 1 H, NH). C NMR (50 MHz, CDCl ): δ 152.1
1
8
26
2
2
1
3
1
8.4ꢀ). H NMR (200 MHz, CDCl ): δ 0.90–0.98 [m, 6 H,
3
3
2Ј
2
9Ј
9
8
8Ј
5,6
(
C ), 150.7 (C ), 141.9 (C ), ‡(C ), ‡(C ), 135.8 (C ), 123.2 (C ),
2 × CH (R)], 1.19–1.79 [m, 8 H, alkyl (R)], 1.97–2.17 [br m, 1
3
5Ј
6Ј
4Ј
4
7
7Ј
1
22.4 (C ), 122.1 (C ), 119.9 (C ), ‡(C ), ‡(C ), 110.4 (C ), 70.7
H, CH (R)], 3.30 (s, 2 H, CH CO ), 4.12–4.16 (d, 2 H, NCH ),
2
2
2
4
–6
(
C H N CH O), 70.5 (C H N CH O), 69.3 (CH O), 67.4
4.30 (s, 2 H, C H N CH ), 7.28–7.42 (m, 3 H, H of aryl),
7.81–7.89 (m, 1 H, H of aryl) and 11.33 (br s, OH). C NMR
7
5
2
2
7
5
2
2
2
7
4
2
7
2
13
(
CH O), 65.7 (CH O), 48.6 (NCH ) and 31.9–14.1 (remaining
2
2
2
2
9
C atoms of hydrophobic alkyl chains). FAB mass spectrum: m/z
(50 MHz, CDCl ): δ 172.1 (C᎐O), 150.6 (C ), 139.4 (C ), 135.1
3
8 5 6 4 7
ϩ
ϩ
5
1
91 (M ϩ H , 70ꢀ). EI mass spectrum: m/z 590.4190 (M ,
00ꢀ); C H N O requires 590.4196.
(C ), 123.6 (C ), 123.1 (C ), 118.9 (C ), 110.3 (C ), 48.3
(C H N CH ), 40.0 (NCH ), 36.4 (CH CO ) and 30.9–10.9
36
54
4
3
7
4
2
2
2
2
2
(
remaining C atoms of alkyl chains). FAB mass spectrum: m/z
ϩ
1
,2-Bis(N-2-hexyldecylbenzimidazol-2-ylmethoxy)benzene
335 (M ϩ H , 100ꢀ).
1
0f. From the preparation using 1,2-bis(benzimidazol-2-
The benzimidazole carboxylic acid ligands 12–15 were pre-
ylmethoxy)benzene (3.72 g, 10.02 mmol) and added 2-hexyl-
decyl tosylate (8.15 g, 20.56 mmol), a crude oil was purified via
flash chromatography (35–70 µm silica; eluent methanol–
dichloromethane; 1:10) to give a viscous golden oil 10f. Yield
pared according to the general procedure outlined above.
4-[N-(±)-2-Ethylhexylbenzimidazolyl]-3-oxabutanoic acid 12.
To diglycolic acid [2-oxybis(acetic acid)] (4.43 g, 33.04 mmol)
was added N-(±)-2-ethylhexylamino-o-phenylenediamine (3.64
g, 16.53 mmol). The oil was purified via flash chromatography
(35–70 µm silica; eluent methanol–dichloromethane; 1:10) to
give a viscous colourless oil which was triturated with diethyl
2
.2 g, 27ꢀ (Found: C, 79.8; H, 10.2; N, 6.8. C H N O
54 82 4 2
1
requires C, 79.2; H, 10.1; N, 6.8ꢀ). H NMR (200 MHz,
CDCl ): δ 0.71–0.82 [m, 12 H, 4 × CH (R)], 1.01–1.24 [br s, 48
3
3
H, aryl (R)], 1.87–2.02 [br m, 2 H, CH (R)], 3.99–4.03 [d, 4 H,
NCH (R)], 5.28 (s, 4 H, C H N CH O), 6.79–6.97 (m, 2 H,
2
7
4
2
2
13
aryl), 7.04–7.13 (m, 2 H, aryl) and 7.15–7.21 (m, 6 H, aryl).
C
¶ No attempt has been made to separate the (ϩ) and (Ϫ) isomers.
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88
83

View Article Online
ether to yield a pure white solid.¶ Yield 1.98 g, 38ꢀ; m.p. 105 ЊC
Found: C, 68.3; H, 8.5; N, 8.8. C H N O requires C, 67.9;
(
Table 2 Transport from single and binary mixtures of metal() ions
18
26
2
3
1
H, 8.2; N, 8.8ꢀ). H NMR (200 MHz, CDCl ): δ 0.86–0.94 [m,
3
Ligand
Cu
Zn
Cd
Cu
Zn
6
H, 2 × CH (R)], 1.16–1.42 [m, 8 H, alkyl (R)], 1.96–2.02 [br
3
m, 1 H, CH (R)], 4.13 (s, 2 H, CH CO ), 4.13–4.17 (d, 2 H,
1
3f
4f
5
6
3
144.6
165.2
143.5
92.8
209.7
125.8
127.9
131.1
188.8
233.8
275.8
81.8
27.5
65.3
58.6
23.8
18.9
10.4
10.4
19.5
16.5
21.4
171
168.1
137
99.8
104.1
294
303.2
422.4
355
68.7
23.8
25.2
27.8
10
11
13.6
3.7
2
2
4
–6
NCH ), 5.13 (s, 2 H, C H N CH ), 7.26–7.41 (m, 3 H, H of
aryl), 7.77–7.83 (m, 1 H, H of aryl) and 12.05 (br s, OH).
NMR (50 MHz, CDCl ): δ 172.0 (C᎐O), 149.1 (C ), 138.5 (C ),
34.2 (C ), 123.1 (C ), 122.4 (C ), 118.3 (C ), 109.8 (C ), 66.0
2
7
4
2
2
7
13
C
f
2
9
3
f
88.6
8
5
6
4
7
1
h
231.1
226.6
302.2
247.1
327
(
C H N CH ), 62.4 (CH CO ), 47.6 (NCH ) and 38.9–10.0
4h
5h
6h
7
4
2
2
2
2
2
(
3
1
remaining C atoms of alkyl chains). FAB mass spectrum: m/z
19 (M ϩ H , 60ꢀ). EI mass spectrum: m/z 318.1958 (M ,
ϩ
ϩ
66.1
76.6
4.7
4.7
7
h
431.6
5ꢀ); C H N O requires 318.1943.
18
26
2
3
pH 5.05, NaO CMe buffer. Columns 2–4; single-ion transport, feed
2
II
Ϫ3
solution, [M ] = 0.1 mol dm . Columns 5 and 6; transport from binary
mixtures, feed solution, [M ] = 0.2 mol dm . pH receive 3.42.
7
-[N-(±)-2-Ethylhexylbenzimidazolyl]-3,6-dioxaheptanoic
II
Ϫ3
acid 13. To 3,6-dioxaoctanedioic acid (4.85 g, 27.23 mmol) was
added N-(±)-2-ethylhexylamino-o-phenylenediamine (3.00 g,
1
3.36 mmol). The oil was purified via flash chromatography
(
(
C H N CH ), 64.2 (CH CO ), 48.6 (NCH ) and 39.9–10.8
7 4 2 2 2 2 2
(
35–70 µm silica; eluent methanol–dichloromethane, 1:10) to
remaining C atoms of alkyl chains). FAB mass spectrum:
give a viscous colourless oil.¶ Yield 1.65 g, 34ꢀ (Found: C, 66.6;
H, 8.2; N, 7.5. C H N O requires C, 66.3; H, 8.3; N, 7.7ꢀ). H
NMR (200 MHz, CDCl ): δ 0.84–0.92 [m, 6 H, 2 × CH (R)],
ϩ
ϩ
m/z 411 (M ϩ H , 35ꢀ). EI mass spectrum: m/z 410.2186 (M ,
0ꢀ); C H N O requires 410.2206.
1
20
30
2
4
2
24
30
2
4
3
3
1
3
.15–1.40 [m, 8 H, alkyl (R)], 1.94–2.01 [br m, 1 H, CH (R)],
Results
.74 (s, 4 H, OCH ), 4.14 (s, 2 H, CH CO ), 4.14–4.17 (d, 2 H,
2
2
2
4
–6
2ϩ
2ϩ
NCH ), 4.97 (s, 2 H, C H N CH ), 7.24–7.39 (m, 3 H, H of
aryl), 7.77–7.83 (m, 1 H, H of aryl) and 10.05 (br s, OH).
NMR (50 MHz, CDCl ): δ 173.0 (C᎐O), 150.2 (C ), 140.0 (C ),
1
The hydrophobic benzimidazoles transported Cu , Zn and
2ϩ
2
7
4
2
2
7
13
C
Cd at transport rates [or flux (J )] shown in Tables 2–7. The
m
2
9
transport rates and potential metal-ion selectivity of the hydro-
phobic benzimidazole ligands are considered in terms of
hydrophobicity and branching, N-alkyl substitution and the
nature (by systematic alteration of the bridge length using add-
itional ether donor atoms and methylene spacer groups) of the
linking bridge under conditions of multiple-ion transport (i.e.
more than one metal present).
3
8
5
6
4
7
35.0 (C ), 123.7 (C ), 123.0 (C ), 119.3 (C ), 110.5 (C ), 70.7
(
C H N CH ), 69.3 (OCH ), 69.2 (OCH ), 64.2 (CH CO ),
7
4
2
2
2
2
2
2
4
8.4 (NCH ) and 39.7–10.8 (remaining C atoms of alkyl
2
ϩ
chains). FAB mass spectrum: m/z 363 (M ϩ H , 85ꢀ). EI mass
spectrum: m/z 362.2217 (M , 90ꢀ); C H N O requires
3
ϩ
20
30
2
4
62.2206.
Three-phase single-ion transport
1
0-[N-(±)-2-Ethylhexylbenzimidazolyl]-3,6,9-trioxadecanoic
acid 14. To 3,6,9-trioxaundecanedioic acid (7.52 g, 33.86 mmol)
was added N-(±)-2-ethylhexylamino-o-phenylenediamine (3.61
g, 16.37 mmol). The oil was purified via flash chromatography
2ϩ
The ligands 1, 3f, 3h–6f, 6h and 7h were found to transport Cu
(
2ϩ
Jm, 88.5–327.0) and Zn (Jm 66.1–275.8) effectively (Table 2,
columns 2–4) under the conditions of single-ion transport (feed
= 0.1 mol dm , pH = 5.05 with NaO CMe
buffer). Significant transport of Cd is only effected by the
(
35–70 µm silica; eluent methanol–dichloromethane, 1:10) to
2ϩ
Ϫ3
phase; [M ]
total
2
give a viscous colourless oil.¶ Yield 2.46 g, 37ꢀ (Found: C, 66.1;
2ϩ
1
H, 8.6; N, 7.0. C H N O requires C, 65.0; H, 8.4; N, 6.9ꢀ). H
22
34
2
5
fully alkylated thioether bridged ligands 3f (J , 65.3) and 4f (J ,
m
m
NMR (200 MHz, CDCl ): δ 0.79–0.87 [m, 6 H, 2 × CH (R)],
2ϩ
3
3
58.6). Transport of Ni by the series of ligands 1, 3f, 3h–6f, 6h
and 7h was found to be negligible which is probably kinetic in
1
3
.18–1.34 [m, 8 H, alkyl (R)], 1.88–2.10 [br m, 1 H, CH (R)],
.55–3.72 (m, 8 H, OCH ), 4.06–4.09 (d, 2 H, NCH ), 4.13 (s, 2
2ϩ
2
2
origin due to the slow substitution of [Ni] (aq).
4
–6
H, CH CO ), 4.85 (s, 2 H, C H N CH ), 7.20–7.32 (m, 3 H, H
aryl), 7.70–7.77 (m, 1 H, H of aryl) and 10.05 (br s, OH).
NMR (50 MHz, CDCl ): δ 173.1 (C᎐O), 150.5 (C ), 140.4 (C ),
1
2
2
7
7
4
2
2
13
C
9
Hydrophobicity. In considering the effect of increasing hydro-
phobicity, ligands 1, 3f, 3h–6f, 6h and 7h were examined under
single-ion transport conditions (Table 2, columns 2–4). These
ligands contain comparable ether and thioether linking bridges
but possess alkyl chains of differing hydrophobicity. Increasing
the degree of hydrophobicity (stearyl < hexadecyl < octadecyl
2
3
8
5
6
4
7
35.1 (C ), 123.5 (C ), 122.8 (C ), 119.4 (C ), 110.4 (C ), 71.0
(
C H N CH ), 70.6 (OCH ), 70.4 (OCH ), 69.6 (OCH ), 69.4
7
4
2
2
2
2
2
(
OCH ), 69.0 (OCH ), 64.7 (CH CO ), 48.3 (NCH ) and 39.7–
2 2 2 2 2
1
0.8 (remaining C atoms of alkyl chains). FAB mass spectrum:
ϩ
ϩ
m/z 407 (M ϩ H , 40ꢀ). EI mass spectrum: m/z 406.2458 (M ,
0ꢀ). C H N O requires 406.2468.
<
isooctadecyl) has little or no effect on metal-ion selectivity as
4
22
34
2
5
expected. The hydrophobic alkyl chains, lacking potential co-
ordinating sites, are believed to anchor the ligand to the organic
membrane and are remote from the ligand’s co-ordinating
environment. It is apparent that there is no clear trend linking
an increase in ligand hydrophobicity to an increase in the trans-
port rate under the conditions of single-ion transport and using
chloroform as the organic membrane.
7
-[2-N-(±)-2-Ethylhexylbenzidazolyl]-catechol-o,o-acetic
acid 15. To catechol-o,o-diacetic acid (5.35 g, 23.65 mmol) was
added N-(±)-2-ethylhexylamino-o-phenylenediamine (2.60 g,
1
1.83 mmol). The oil was purified via flash chromatography
(
35–70 µm silica; eluent methanol–dichloromethane, 1:10) to
give a viscous colourless oil which was triturated with diethyl
ether to yield a pure white solid.¶ Yield 1.31 g, 27ꢀ; m.p. 138 ЊC
Nature of alkyl substitution (single or double N-alkylation).
The number of alkyl groups on the bis(benzimidazole) ligands
(3f, 3h–6f, 6h) has a considerable impact on metal-ion selectiv-
(
7
Found: C, 70.2; H, 7.4; N, 6.8. C H N O requires C, 70.2; H,
24
30
2
4
1
.4; N, 6.8ꢀ). H NMR (200 MHz, CDCl ): δ 0.82–0.91 [m, 6
3
2ϩ
2ϩ
2ϩ
H, 2 × CH (R)], 1.23–1.52 [m, 8 H, alkyl (R)], 1.98–2.12 [br m,
ity and the transport rate of flux (J ) of Cu , Zn and Cd in
3
m
1
5
H, CH (R)], 4.21–4.25 (d, 2 H, NCH ), 4.65 (s, 2 H, CH CO ),
single-ion transport experiments (Table 2, columns 2–4). A
2
2
2
.46 (s, 2 H, C H N CH ), 6.68–7.03 (m, 4 H, aryl), 7.22–7.37
trend for the metal-ion transport rates (J ) for these ligands is
7
4
2
2
m
13
(
m, 3 H, aryl), 7.76–7.81 (m, 1 H, aryl) and 12.05 (br s, OH).
C
identified in Table 3. The half (singly) alkylated ligands have
2
2ϩ
NMR (50 MHz, CDCl ): δ 171.6 (C᎐O), 149.1 (C ), 149.0
Cu fluxes (JCu2ϩ) 2–3 times higher than their fully (doubly)
3
9
8
(
(
aryl), 146.7 (aryl), 139.3 (C ), 134.7 (C ), 124.1 (aryl), 123.5
C ), 121.9 (C ), 119.1 (C ), 113.8 (aryl), 110.7 (C ), 66.4
alkylated analogues. This phenomenon is also observed for the
5
6
4
7
2ϩ
transport of Zn by the half-alkylated thioether-bridged lig-
8
4
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88

View Article Online
Table 3 Trend observed in metal-ion selectivity and transport rate
Jm) from single-ion transport experiments due to the influence of alkyl
substitution
Table 5 Trend observed in metal-ion selectivity and transport rate (Jm)
from multiple-ion transport experiments due to the influence of alkyl
substitution
(
Hydrophobic bis(benzimidazole) ligands
Hydrophobic bis(benzimidazole) ligands
Metal ion
Thioether bridged
Ether bridged
Metal ion
Thioether bridged
Ether bridged
2ϩ
2ϩ
a
a
Cu
Zn
Cd
h > f
h > f
f > h
h > f
f > h
f > h
Cu
h > fb
h > fa
2
ϩ
2ϩ
Zn
f > h
f > h
2
ϩ
a
b
This trend is also observed in single-ion transport experiments.
Reverse of trend observed in single-ion transport experiments.
h = Half-alkylated, f = fully alkylated.
Table 4 Trend observed in metal-ion selectivity from single-ion trans-
petitive experiments between pairs of cations have shown sup-
pression of transport of one metal ion relative to its single-ion
port experiments due to the influence of the linking bridge
7,14
rate.
In the competition experiments described here the
Hydrophobic bis-benzimidazoles
2ϩ
transport rate of Cu by the fully alkylated ligands remained
approximately the same as in single-ion experiments but the
N-Alkylation
Ether bridged
Thioether bridged
2ϩ
Zn transport rate was dramatically suppressed. However, the
half-alkylated ligands displayed an increase in the transport of
2
ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
N-Alkyl
N,NЈ-Dialkyl
Cu > Zn ӷ Cd
Zn > (≈) Cu ӷ Cd
2ϩ 2ϩ
2
ϩ
2ϩ
Zn > Cu ӷ Cd
Cu > (≈) Zn ӷ Cd
2ϩ
2ϩ
Cu and a suppression of the Zn transport rate to an almost
negligible level. The increase in JCu2ϩ probably reflects the use of
a more concentrated binary feed solution i.e. an increase in
2ϩ
ands 3h and 4h. These ligands promote Zn fluxes (JZn2ϩ) twice
those of their fully alkylated analogues 3f and 4f but the reverse
2ϩ
Ϫ
3
2ϩ
total [M ] and [NO ]. The relative differences in Cu and
2ϩ
situation arises for the ether-bridged ligands; (J 2ϩ) 5f, 6f > 5h,
Zn transport rates arising from experiments conducted in
isolation and competition are shown in SUP 57310. The half-
alkylated ligands exhibited a superior discriminatory behaviour
Zn
2ϩ
6
h. Although the transport of Cd by the hydrophobic bis-
(
benzimidazole) ligands is relatively low, the trend cannot be
2ϩ
2ϩ
2ϩ
ignored. The fully alkylated ligands promote a higher Cd flux
for the transport of Cu and Zn compared to their fully
alkylated analogues which can be clearly seen in Table 2. The
trend observed in single-ion transport was slightly altered under
conditions of competition (Table 5), the half-alkylated ether
(
JCd2ϩ) than do their half-alkylated analogues under the condi-
tions of single-ion transport.
2ϩ
Nature of linking bridge. The effect of alkyl substitution on
metal-ion selectivity is also associated with the nature of the
linking bridge. The half-alkylated ligands containing an ether
and thioether ligands having a higher Cu flux (J 2ϩ) than the
Cu
2ϩ
analogous fully alkylated ligands whereas the Zn flux (JZn2ϩ
)
was higher for the fully alkylated ligands.
2ϩ
bridge (5h–7h) have a higher Cu flux (JCu2ϩ) in comparison to
2ϩ
the Zn fluxes (JZn2ϩ), i.e. JCu2ϩ > JZn2ϩ (Table 2). This phe-
nomenon is reversed for the fully alkylated analogues (1, 5f and
Nature of linking bridge. The fully alkylated ether- and
thioether-bridged bis(benzimidazole) ligands displayed com-
2ϩ
2ϩ
6
f), i.e. JZn2ϩ > JCu2ϩ. Concerning the thioether-bridged ligands
a complete reversal in trend arises; half-alkylated ligands 3h and
parable metal-ion selectivity for Cu over Zn in binary mix-
2ϩ
tures. For the half-alkylated ligands a higher selectivity for Cu
2
ϩ
2ϩ
4
h have higher Zn fluxes JZn2ϩ > JCu22ϩ whereas the fully alkyl-
over Zn was displayed, and the oxygen-bridged ligands 5h, 6h
ϩ
ated ligands 3f and 4f have higher Cu fluxes JCu2ϩ > JZn2ϩ. The
fully alkylated thioether-bridged ligands 3f and 4f display a
marginal affinity for transporting Cd JCd = 65.3 and J 58.6
and 7h were best (Table 2, last two columns).
2ϩ
2ϩ
Multiple-ion transport: tertiary mixtures
respectively. The remaining ligands of the series showed little
Since a direct comparison of transport rates under conditions
of competition is the one of consequence in any future appli-
cation, tertiary mixtures containing an aqueous buffered equi-
2ϩ
propensity to transport Cd ; J 2ϩ 10.4–27.5. The trend is
Cd
identified in Table 4. Overall, the single-ion transport results
suggested copper and zinc selectivity might be established with
the half-alkylated ether-bridged (5h–7h) and fully alkylated
bridged (1, 5f, 6f) ligands respectively, irrespective of the degree
of hydrophobicity.
2ϩ
2ϩ
2ϩ
molar solution of Cu , Zn and Cd were also examined as a
2ϩ
feed solution (source phase; [M ]
= 0.1 ϩ 0.1 ϩ 0.1 = 0.3
total
Ϫ3
mol dm ) to determine the ultimate selectivity of the half, fully
alkylated ligands (1, 2, 5f, 5h, 8f, h–10f,h) and the N-alkylated
benzimidazolecarboxylic acids 12–15. The linking bridges of
these ligands were systematically altered using additional ether
donor atoms and methylene spacer groups to effect a possible
Three-phase multiple-ion transport in binary mixtures
Pairwise comparisons of transport rate were made for the series
2ϩ
2ϩ
2
ϩ
increase in the Zn , Cd transport rates and selectivities, thus
omitting ligands 3f, 3h, 4f, 4h, 6f, 6h and 7h from this study.
Only one type of branched alkyl chain was used to supply
hydrophobicity (hexadecyl). The effort in synthesizing ligands
with a range of hydrophobic chains seemed futile following dis-
coveries that the variation of hydrophobicity had no effect on
metal-ion selectivity or flux in a chloroform organic membrane.
To ensure this observation still holds using a tertiary feed of
equimolar metal ions, the fully alkylated straight-chain ligands
1 and 2 were screened as a control. Variation in aromatic substi-
tution was also considered for ligands 2 and 8. The following
results are described in terms of the nature of the linking bridge
and (N- and aromatic) alkyl substitution.
of ligands with a buffered feed solution equimolar in both Cu
2ϩ
2ϩ
Ϫ3
and Zn ([M ]
= 0.1 ϩ 0.1 = 0.2 mol dm ), shown in Table
total
2
, last two columns.
Hydrophobicity. As in the single-ion experiments no definitive
relationship between hydrophobicity and transport rate was
observed. However, results for the binary mixtures (Table 2, last
two columns) suggested a slight trend, although not apparent in
every case, of a decrease in transport rate by fully alkylated
2ϩ
2ϩ
bis(benzimidazole) ligands for Cu and Zn , while the trend is
reversed for the half-alkylated ligands.
Nature of alkyl substitution (single or double N-alkylation).
No reversal in metal-ion selectivity between half (h) and full (f)
alkyl substitution was found, contrary to the finding with
Nature of alkyl substitution (single and double N-alkylation).
As is clear from Table 6, the trend in metal-ion selectivity
in tertiary mixtures for ligands 8f, 8h–10f, 10h was unaffected
f
h
2ϩ
single-ion transport experiments; thus for L and L , Cu
>
2ϩ
Zn , where L = bis(benzimidazole) ligand. Frequently, com-
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88
85

View Article Online
selectivity for copper were dramatically increased in com-
Table 6 Influence of alkyl substitution on the trend in metal-ion
parison to those for the half (8h–10h) and fully alkylated (1, 2,
2ϩ
2ϩ
2ϩ
2ϩ
selectivity for Cu , Zn and Cd in tertiary mixtures of metal() ions
8f–10f) ligands. The increase in Cu (JCu2ϩ) flux from 176.7 for
ligand 12 reached a plateau with increasing bridge length, at a
Jm value of 310–340 for 13–15; this coincided with a suppres-
Ether-bridged
bis-benzimidazole
JCu2ϩ
JZn2ϩ
JCd2ϩ
Selectivity trend
2ϩ
2ϩ
sion of the Zn (JZn2ϩ) and Cd JCd2ϩ fluxes, 2.4–3.8 and 0.84–
3.7 respectively.
2
ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
5
5
8
8
0h
0f
h
f
h
f
271.2
190.4
163.4
164.5
72.4
124.4
79.8
3.5
12.4
9.0
5.5
8.8
35.8
17.5
23.4
1.3
8.3
Cu > Zn > Cd
2ϩ 2ϩ
1
Cu > Zn > Cd
2
ϩ
2ϩ
15.0
31.5
32.4
46.1
10.1
17.0
Cu > Cd > Zn
2ϩ 2ϩ
Cu > Cd > Zn
2ϩ 2ϩ
Discussion
1
1
Cu > Cd > Zn
2ϩ 2ϩ
Cu > Cd > Zn
The present study has revealed that variation in the hydro-
phobicity of the N-alkyl substituent of a benzimidazole had
little or no effect on the metal-ion selectivity of the benzimid-
azole in chloroform three-phase transport. However, the choice
of alkyl substituent has important implications in the hydro-
metallurgical industry, where the degree of hydrophobicity
often governs the choice and amount of modifier that is added
with the ligand to the extracting medium (to obviate surface
effects and prevent leaching of the extracting medium). It had
been hoped that the addition of an alkyl chain to the secondary
nitrogen of a benzimidazole might also provide a more rigid
ligand backbone which could enhance metal-ion selectivity
because a more exact fit of the metal cation with the ligand
would then be required. Previously, alkyl chains with amide and
carbonyl functional groups have been used to provide benz-
imidazoles with hydrophobicity and extra co-ordinating sites in
2
ϩ
2ϩ
9
9
h
f
Cu > Zn > Cd
2ϩ 2ϩ
86.7
Cu > Zn > Cd
7b
an attempt to improve metal-ion selectivity. The amide car-
bonyl groups supplied were found to be non-co-ordinating and
these substituted alkyl chains less effective than unsubstituted
7c
alkyl chains for metal-ion transport.
Single-ion transport measurements permitted (Table 2) com-
parison of structural features of the bis-benzimidazoles with
their efficacy in metal-ion transport. A subtle alteration in the
number and type of donor atoms in the bridge, number of
methylene spacer groups in the bridge, optimum alkyl chain
length and the effect of aromatic ring substitution could
increase or decrease the rate at which the ligand transported a
particular metal ion. Once the optimum structural features have
been established for a ligand which successfully transports the
metal ion of interest, the ability to transport the metal ion in
competition with other metal ions must be tested in order to
ascertain if any genuine selectivity has been engineered. It
seemed, following establishment of single-ion transport results,
Fig. 1 Transport profile of alkylated benzimidazolecarboxylic acid
2
ϩ
2ϩ
2ϩ
ligands 12–15 for Cu , Zn and Cd from multiple-ion transport
2
ϩ
Ϫ3
experiments. Feed phase; [M ]total = 0.3 mol dm , pH 5.05, NaO CMe
2
7
buffer. Bars represent Jm (moles transported 24 h) × 10
by the single vs. double N-alkylation. The enhancement in
2ϩ
Cu transport rate (JCu2ϩ) previously observed for the singly
alkylated ligands 3h–7h (Table 2) was not apparent for the lig-
ands 8f, 8h–10f, 10h (Table 6) when the bridge length was
increased. Indeed, the fully alkylated ligands 8f–10f were found
to exhibit higher metal-ion fluxes compared to their half-
alkylated analogues 8h–10h; the exception is 5f which has a
lower copper flux (JCu2ϩ; 190.4 < 271.2) than that of ligand 5h.
This phenomenon was previously observed with binary equi-
2ϩ
2ϩ
that the optimum structural features to remove Cu and Zn
selectively from a mixture of metal ions (multiple ion) were
displayed by the half (5h–7h) and the fully alkylated (1, 5f,
6
f) ether-bridged ligands respectively; the thioether-bridged
2ϩ 2ϩ
ligands (3f, 3h, 4f, 4h) displayed comparable Cu and Zn
2ϩ
2ϩ
molar mixtures of Cu and Zn (Table 2). Aromatic substitu-
tion with methyl groups also enhanced the transport rate and
transport rates and offered no real discrimination between the
2ϩ
two metal ions (Table 2). The transport of Cd by the series of
ligands 1, 3f, 3h–6f, 6h and 7h was negligible: transport was
only effected by the fully alkylated thioether-bridged ligands 3f
and 4f.
metal-ion selectivity, cf. ligand 2 [J 2ϩ, 236.7; JZn2ϩ, 2.2; JCd2ϩ
0
,
Cu
.9] > ligand 1 [JCu2ϩ, 146.3; JZn2ϩ, 20.0; JCd2ϩ, 9.9]. As outlined
earlier, the use of straight alkyl chains (ligands 1 and 2) as
opposed to branched alkyl chains does not alter the metal-ion
selectivity.
The trend in selectivity under the conditions of single-ion
transport may be viewed in terms of the bridge length, donating
atom (hard or soft) and the geometrical constraints provided by
the ligands 1, 3f, 3h–6f, 6h and 7h. In terms of size, the ionic
2
ϩ
Nature of the linking bridge. The selectivity of Cu
>
2ϩ
2ϩ
Zn > Cd for the series of ligands held as the bridge length
was increased through the addition of methylene spacer groups
and ether donor atoms (Table 6). However, the overall selectiv-
15
2ϩ
2ϩ
2ϩ
radii are in the sequence Cd (0.97), Zn (0.74) and Cu
(
2ϩ
2ϩ
2ϩ
0.70 Å) while Cd is soft and Cu and Zn are borderline.
2ϩ
The size of the metal ion may partly explain the lack of Cd
2ϩ
ity for copper decreased. In general, the flux of Cu
was suppressed as the bridge length was increased whereas the
(JCu2ϩ
)
transport by the series of ligands which have a CH XCH2
2
(
X = S or O) linking bridge. The bridge length is apparently too
2ϩ
2ϩ
flux of Zn (JZn2ϩ) and Cd (JCd2ϩ) was slightly raised. The
sequence of metal-ion selectivity displayed by the series of
ligands is represented in Table 7.
2ϩ
small for effective transport of a relatively large Cd ion in
comparison to the smaller metal ions Cu and Zn which have
comparable ionic radii. In general the softer Cd prefers soft
thioether donating atoms whereas Cu and Zn are borderline
and usually prefer ether and nitrogen donating atoms (hard–
soft acid–base theory). The presence of a soft thioether donor
atom from the linking bridge may be responsible for the slight
transport of Cd by ligands 3f and 4f. The high transport rate
2ϩ
2ϩ
2ϩ
2ϩ
2ϩ
N-Alkylated benzimidazolecarboxylic acids. It is evident that
the benzimidazolecarboxylic acids (for which a transport profile
is shown in bar graph form in Fig. 1) are extremely selective for
16
2ϩ
2ϩ
2ϩ
2ϩ
Cu over Zn and Cd . The transport rate and metal-ion
8
6
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88

View Article Online
Table 7 Sequence of metal-ion selectivity displayed by ligands 1, 2, 5f, 5h, 8f, 8h–10f, 10h
5
h
>(≈)
2
>
>
>(≈)
>(≈) 20.0
>(≈) 8f
>(≈) 31.5
5f
>(≈) 8f
>(≈) 8h
>(≈)
1
>
>
10f
>
>
>
>
>
>
9f
86.7
8f
5.5
5f
8.3
>(≈) 9h
>(≈) 10h
JCu2ϩ
JZn2ϩ
JCd2ϩ
271.2 >(≈) 236.7
190.4 >(≈) 164.5 >(≈) 163.4 >(≈) 146.3
124.4
>(≈) 79.8 >(≈) 72.4
1
0f
35.8
0f
46.1
>
>
>
>
9f
1
>(≈) 9h
>(≈) 17.5
>
>
5f
12.4
>
>
>
>
8h
9.0
9h
>(≈) 10h
>
>
>
>
5h
>
>
>
>
2
23.4
10h
32.4
>(≈) 8.8
3.5
5h
2.2
2
1
>
>
9f
17.0
>(≈) 8h
>(≈) 15.0
>(≈)
1
10.1
>(≈) 9.9
1.3
0.9
(
≈) Within the boundaries of experimental error ±15ꢀ.
2
ϩ
2ϩ
2ϩ
of Cu and Zn is effected by the presence of an ether donat-
the ether donating atom from the linking bridge to both Cu
2ϩ
ing atom in the linking bridge of 1, 5f, 5h–6f, 6h and 7h, and is
more pronounced for Cu with the combination of a second-
ary arylamine functionality (half-alkylated ligands 5h–7h).
Behr et al. have described how the selectivity of a system
between two cations needs to be measured in competition and
not isolation. The present study reinforces this analysis, since
the potential Cu and Zn selectivity predicted from the
single-ion transport rates alone is not found for the mixtures
and Zn . The copper and zinc were five-co-ordinate in a
pseudo-trigonal-bypyramidal geometry with halide anions
completing the remaining co-ordination sites. The thioether-
bridged ligands displayed a strikingly different co-ordination
mode to Zn since the Zn was found to be in its ‘preferred’
four-co-ordinate tetrahedral environment. The benzimidazole
2ϩ
17
2ϩ
2ϩ
2
ϩ
2ϩ
ligand provided an N donor set only; the thioether donating
2
atom of the linking bridge did not participitate in the co-
ordination of the central zinc atom, and the remaining co-
ordination sites were filled by halide anions. If such a rigid
tetrahedral co-ordinating environment were to be maintained in
solution this could aid selectivity. Ligands which impose a
defined geometric array of donor atoms may induce the select-
ive binding of a particular metal ion. Nickel(), copper() and
iron() prefer octahedral/square planar, square planar and
octahedral geometrical environments respectively whilst zinc()
has a complete d electronic configuration and is not subject to
any crystal field stabilisation energy effects or preferred co-
ordinating environments. Therefore, a rigid tetrahedral donor
environment could co-ordinate Zn in preference to other
transition metals. The solid-state structural comparison of the
thioether- and ether-bridged ligands may suggest stronger
bonding to Cu and Zn with ligands possessing ether donat-
ing atoms in the linking bridge rather than thioether donating
atoms. It is probable that the smaller ether donating atom of the
linking bridge can co-ordinate more readily to Cu or Zn in
comparison to the larger thioether donating atom which is
apparently sterically hindered and is unable to provide a co-
ordination site for the central zinc atom when constrained.
2ϩ
2ϩ
(
comparisons in Table 2). The order of selectivity, Cu > Zn ,
is maintained throughout the series of ligands 1, 3f, 3h–6f, 3h
and 7h irrespective of the nature of alkyl substitution and the
linking bridge (S or O donor), but the observed selectivity in
competitive experiments was accompanied by a fluctuation of
the transport rates for Cu and Zn relative to their single-ion
transport rates. The transport of Zn by the series of ligands
was dramatically suppressed. The transport of Cu was main-
tained for the fully alkylated ligands (1, 3f–6f) and increased for
the half-alkylated ligands (3h–7h). The increased anion concen-
tration in the mixed-ion solution is probably relevant. The
changes in transport rate induced a superior selectivity for Cu
over Zn ; those ligands containing ether donating atoms (5h–
2ϩ
2ϩ
2ϩ
2ϩ
10
2ϩ
20
2ϩ
2ϩ
7
h) were best.
2ϩ
2ϩ
Undoubtedly, the unalkylated secondary nitrogen of the
bis(benzimidazole) ligands (5h–7h) is the rationale for the
2ϩ
striking difference in Cu transport rates in mixtures for these
ligands compared to the f analogues. This phenomenon may be
related to other features of the ligand co-ordination, namely
hydrogen bonding, relative dissociation rates for the metal
chelates, and orientation of the ligand (and/or complex) at the
interface in transport cells. The ‘free’ NH might enhance the
2ϩ
2ϩ
2ϩ
The enhancement in Cu transport rates occurring for the
rate of transport by hydrogen bonding to the counter anion
half-alkylated ligands possessing a CH XCH2 (X = S or O)
2
Ϫ
(
NO ) in the fashion which is displayed in the solid state in
bridge is diminished as the bridge length is increased for the
series of ligands 1, 2, 5h, 5f, 8f, 8h–10f, 10h (Table 6). The
suppression of this effect (nature of alkyl substitution) with
increasing bridge length demonstrates that the geometrical,
structural, and donor atom properties of the linking bridge are
the most important facets of a bis(benzimidazole) ligand when
considering metal-ion selectivity. While the overall selectivity of
3
12,18
various crystal structures,
which would be unavailable to the
fully alkylated ligands, devoid of free NH. Further, the hydro-
philic nature of the unalkylated half of the bis-benzimidazole
could facilitate its orientation towards the metal ion at the
organic–aqueous interface, while leakage of ligand to the
aqueous medium is prevented by the anchoring of the molecule
to the organic phase via the highly hydrophobic alkyl chain
connected to the opposite half of the ligand.
In considering the effect of the hydrophobic chains allusion
can be made to some biological situations. With carrier
molecules like valinomycin a hydrophobic ‘coat’ of methyl
groups from e.g. Val, Ala, Ile and Leu apparently enables the
molecule to transport enclosed potassium ions through mem-
branes. In lipids a much longer linear alkyl chain acts as an
interface but is not itself mobile. The branched alkyl chains
chosen here lack the ability of the lipid chain to stack and are
likely to be more disordered. Since no attempt was made to
separate the optical and other isomers (possibly also congeners)
of the alkyl groups used here, the compounds are unlikely to be
homogeneous, but they do have a strategic ‘coat’ of methyl
groups to assist transport.
2ϩ
2ϩ
2ϩ
Cu > Zn > Cd for the series of ligands holds as the bridge
length increases, nevertheless as the bridge length increases the
2ϩ
selectivity for larger metal ions such as Cd increases whereas
the selectivity for smaller metal ions decreases.
The results of metal-ion selectivity for the entire series of
ligands can be summarised for three-phase transport. In gen-
2ϩ
2ϩ
2ϩ
eral Cu > Zn > Cd ; metal-ion selectivity diminishes as
bridge length increases but the trend in selectivity holds.
The transport of a metal ion through a bulk liquid mem-
brane is believed to be dominated by three major rate process-
12b,c,21
es:
the net rate at which the metal ion [and its anion(s)]
crosses the feed phase and associates with the ligand, the rate at
which the ligand–metal complex crosses the organic membrane
and the rate at which the cation can dissociate itself from the
ligand into the receive phase. Therefore at one, a combination,
2ϩ
2ϩ
The co-ordinating environments provided by the ligands with
a CH XCH (X = S or O) linking bridge have previously been
or all of these stages the Cu is ‘winning’ against Zn . The
ligands 1, 2, 3f, 3h–6f, 6h and 7h which possess a short bridge
2
2
18,19
2ϩ
examined crystallographically.
In the solid state, ether-
length have discriminated against Cd . Although as the bridge
bridged ligands provided the expected N O donor set from the
tertiary nitrogen atoms of the benzimidazole heterocycle and
length increases for the series of ligands 1, 2, 5h, 5f, 8f, 8h–10f,
10h the discrimination changes in favour of Cd , Cu remains
2
2ϩ
2ϩ
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88
87

View Article Online
the ‘winning’ metal ion. This phenomenon may be attributed
8 G. S. Walpole, J. Chem. Soc., 1914, 105, 2501.
22
2ϩ
9 R. C. Weast (Editor), Handbook of Chemistry and Physics, 62nd
edn., CRC Press, Boca Raton, FL, 1981, pp. D139–D141.
0 N. A. Bailey, D. E. Fenton, S. J. Kitchen, T. H. Lilley, M. G.
Williams, P. A. Tasker, A. J. Leong and L. F. Lindoy, J. Chem. Soc.,
Dalton Trans., 1991, 627.
to the kinetic lability (and thermodynamic stability) of Cu
2ϩ
2ϩ
in comparison to those of Zn and Cd .
1
The N-alkylated benzimidazolecarboxylic acids (12–15)
proved to be the most selective ligands for the extraction of
2ϩ
Cu in three-phase transport (Fig. 1). Introduction of a longer
11 J. D. Lamb, J. J. Christensen, S. R. Izatt, K. Bedke, M. S. Astin and
R. M. Izatt, J. Am. Chem. Soc., 1980, 102, 3399; J. J. Christensen,
J. D. Lamb, S. R. Izatt, S.E. Starr, G. C. Weed, M. S. Astin and
R. M. Izatt, J. Am. Chem. Soc., 1978, 100, 3219; R. M. Izatt, R. L.
Bruening, G. A. Clark, J. D. Lamb and J. J. Christensen, J. Membr.
Sci., 1986, 28, 77.
2ϩ
bridge reduced the overall Cu selectivity and could be viewed
in the terms described above, but in general there remains a
2ϩ
high selectivity for Cu throughout the series. It is probable
that the benzimidazolecarboxylic acids do not behave as neutral
carriers and may participitate in counter-current transport:
metal ions transported in one direction and protons in the
1
2 (a) H. P. Berends and D. W. Stephan, Inorg. Chim. Acta, 1984, 93,
173; (b) J. C. Lockhart, W. Clegg, M. N. S. Hill and D. J. Rushton,
J. Chem. Soc., Dalton Trans., 1990, 3541; (c) W. Clegg, J. C.
Lockhart and F. H. Musa, J. Chem. Soc., Dalton Trans., 1986, 47;
23
opposite direction. The transport process for carriers bearing
carboxylate functionalities is known to be extremely pH
dependent; a change in pH of the buffered regime may cause
dramatic changes in transport rates and even alter metal-ion
(
d) M. A. Phillips, J. Chem. Soc., 1928, 2393.
3 T. M. Handyside, Ph.D. Thesis, University of Newcastle upon Tyne,
982.
1
1
1
1
1
1
1
24
selectivities.
4 R. M. Izatt, D. V. Dearden, P. R. Brown, J. S. Bradshaw, J. D. Lamb
and J. J. Christensen, J. Am. Chem. Soc., 1983, 105, 1785.
5 R. D. Shannon and C. D. Prewitt, Acta Crystallogr., Sect. B, 1969,
Acknowledgements
We thank the EPSRC and Zeneca Specialties for financial
support (CASE award to C. J. M.).
2
5, 925.
6 R. G. Pearson, J. Am. Chem. Soc., 1963, 85, 3533; J. Chem. Educ.,
963, 43, 581.
7 J. P. Behr, M. Kirch and J. M. Lehn, J. Am. Chem. Soc., 1985, 107,
41.
1
2
8 C. J. Matthews, W. Clegg, M. R. J. Elsegood, L. Horsburgh and
References
J. C. Lockhart, Inorg. Chem., 1996, 35, 7563.
1
B. C. Pressman, Annu. Rev. Biochem., 1976, 45, 501; G. Spach
Editor), Physical Chemistry of Transmembrane Ion Motions,
Elsevier, Amsterdam, 1983; J. M. Lehn, Pure Appl. Chem., 1979, 51,
19 C. J. Matthews, N. C. Martin, W. Clegg, S. L. Heath, M. N. S.
Hill and J. C. Lockhart, Inorg. Chem., in the press.
20 G. B. Bates and D. Parker, Tetrahedron Lett., 1996, 37, 267.
21 J. C. Lockhart, J. Chem. Soc., Dalton Trans., 1988, 1161.
22 F. A. Cotton and G. Wilkinson in Advanced Inorganic Chemistry,
5th edn., Wiley, New York, 1988, pp. 1288 and 1289.
23 K. Kimura, H. Sakamoto, M. Yoshinaga and T. Shono, J. Chem.
Soc., Chem. Commun., 1983, 978; T. M. Fyles, C. A. McGavin and
D. E. Thompson, J. Chem. Soc., Chem. Commun., 1982, 924;
S. Shinkai, H. Kinda, T. Sone and O. Manabe, J. Chem. Soc., Chem.
Commun., 1982, 125.
24 S. Shinkai, H. Kinda, Y. Araragi and O. Manabe, Bull. Chem. Soc.
Jpn., 1983, 56, 559.
(
9
3
87; E. L. Cussler and D. L. Evans, Sep. Purif. Methods, 1974, 3,
99; J. S. Shultz, in Synthetic Membranes: Science, Engineering and
Applications, eds. P. M. Bumgay, H. Lonsdale and M. N. de Pinho,
D. Reidel, Dordrecht, 1986, pp. 523–566.
J. M. Lehn, Struct. Bonding (Berlin), 1973, 16, 1.
W. Simon, W. E. Morf and P. C. Meier, Struct. Bonding (Berlin),
2
3
1
973, 16, 113.
4
R. M. Izatt and J. J. Christensen (Editors), Synthetic Multidentate
Macrocyclic Compounds, Academic Press, New York, 1978.
Y. A. Ovchinnikov, Eur. J. Biochem., 1979, 94, 321.
5
6
J. K. Wright, R. Seckler and P. Overath, Annu. Rev. Biochem., 1986,
5
, 225.
7
(a) D. J. Rushton, Ph.D. Thesis, University of Newcastle Upon
Tyne, 1989; (b) K. I. Kinnear, J. C. Lockhart and D. J. Rushton,
J. Chem. Soc., Dalton Trans., 1990, 1365; (c) J. C. Lockhart and
D. J. Rushton, J. Chem. Soc., Dalton Trans., 1991, 2633.
Received 25th July 1997; Paper 7/05387H
8
8
J. Chem. Soc., Dalton Trans., 1998, Pages 79–88