Bifunctional cyclam-based ligands with phosphorus acid pendant moieties for radiocopper separation: Thermodynamic and kinetic

DOI: 10.1002/chem.201405777

Full Paper

&

Macrocyclic Ligands

Bifunctional Cyclam-Based Ligands with Phosphorus Acid Pendant

Moieties for Radiocopper Separation: Thermodynamic and Kinetic

Studies

[

a]

[a]

[b]

[a]

Monika Paﬄrovµ, Jana Havlí cˇ kovµ, Aneta Pospí sˇ ilovµ, Miroslav Vetrík, Ivana Císa rˇ ovµ,

[

c]

[c]

[b]

[a]

Holger Stephan, Hans-Jürgen Pietzsch, Martin Hrub y´ , Petr Hermann, and Jan Kotek*

Abstract: Two macrocyclic ligands based on cyclam with

trans-disposed N-methyl and N-(4-aminobenzyl) substituents

in 1m aq. HClO at 258C for H L1 and H L2, respectively).

4 2 4

This combination of properties is convenient for selective

copper removal/purification. Thus, the title ligands were

employed in the preparation of ion-selective resins for

radiocopper(II) separation. Glycidyl methacrylate copolymer

beads were modified with the ligands through a diazotisation

reaction. The separation ability of the modified polymers

as well as two methylphosphinic (H L1) or methyl-

2

phosphonic (H L2) acid pendant arms were synthesised and

4

investigated in solution. The ligands form stable complexes

with transition metal ions. Both ligands show high thermo-

dynamic selectivity for divalent copper over nickel(II) and

zinc(II)—K(CuL) is larger than K(Ni/ZnL) by about seven

orders of magnitude. Complexation is significantly faster for

6

4

was tested with cold copper(II) and non-carrier-added Cu

in the presence of a large excess of both nickel(II) and

zinc(II). The experiments exhibited high overall separation

efficiency leading to 60–70% recovery of radiocopper with

high selectivity over the other metal ions, which were

originally present in 900-fold molar excess. The results

showed that chelating resins with properly tuned selectivity

of their complexing moieties can be employed for radio-

copper separation.

the phosphonate ligand H L2, probably due to the stronger

4

coordination ability of the more basic phosphonate groups,

which efficiently bind the metal ion in an “out-of-cage”

complex and thus accelerate its “in-cage” binding. The rate

II

of Cu complexation by the phosphinate ligand H L1 is com-

2

parable to that of cyclam itself and its derivatives with non-

coordinating substituents. Acid-assisted decomplexation of

the copper(II) complexes is relatively fast (t =44 and 42 s

1

=

2

6

1

Introduction

Some copper radioisotopes, for example, Cu (half-life t

=

/2

1

+

64

+

ꢀ

3

.33 h, b emitter), Cu (t₁_/₂=12.7 h, b and b emitter) and

6

7

ꢀ

The investigation of thermodynamically highly stable and

kinetically inert complexes (often formed by macrocyclic li-

Cu (t₁_/₂=61.8 h, b emitter), exhibit a great potential for

both diagnosis (positron emission tomography) and therapy

[

1–3]

ꢀ

[9,19–21]

64

gands)

has been stimulated by their applications in several

(b particles of intermediate energy).

Among them, Cu

areas, for example, as contrast agents in magnetic resonance

has the greatest importance due to its theranostic character.

All of these isotopes are usually produced by proton irradiation

[

4–8]

imaging

and for labelling of biomolecules with metal radio-

[9–18]

61

isotopes for both diagnostic and therapeutic purposes.

in cyclotrons. Isotope Cu is usually produced by irradiation of

6

4

61

[22]

67

a natural zinc target ( Zn(p,a) Cu reaction), and Cu by the

68

67

Zn(p,2p) Cu process, while the most common method to

64

produce Cu is irradiation of an isotopically enriched nickel

[

a] M. Paﬄrovµ, J. Havlí cˇ kovµ, Dr. I. Císa ˇr ovµ, Prof. P. Hermann, Prof. J. Kotek

64

Department of Inorganic Chemistry

target ( Ni(p,n) Cu reaction). In all cases, separation of copper

radionuclides from a large excess of the Zn or Ni target materi-

Faculty of Science, Universita Karlova (Charles University)

Hlavova 2030, 12840 Prague 2 (Czech Republic)

E-mail: modrej@natur.cuni.cz

64

al is required and, in the case of isotopically enriched Ni tar-

gets, recovery of the nickel isotope is essential to production

economics. For the production of non-carrier-added (NCA)

copper radioisotopes, simple ion-exchange techniques are the

[

b] A. Pospí sˇ ilovµ, M. Vetrík, Dr. M. Hrub y´

Institute of Macromolecular Chemistry

The Academy of Sciences of the Czech Republic

HeyrovskØho nµm eˇ stí 2, 16206 Prague 6 (Czech Republic)

[23]

most common. Recently, Dirks et al. described the separation

of Cu from a Ni target by a solid-phase extraction technique

and demonstrated that macroscopic amounts of Zn do not

reduce the retention of Cu. However, the chemical composi-

[c] Dr. H. Stephan, Dr. H.-J. Pietzsch

Helmholtz-Zentrum Dresden-Rossendorf

Institute of Radiopharmaceutical Cancer Research

Bautzner Landstraße 400, 01328 Dresden (Germany)

[24]

tion of the resin was not published. Therefore, research fo-

cused on new ligands for copper(II) to tune the properties of

Supporting information for this article is available on the WWW under

http://dx.doi.org/10.1002/chem.201405777.

Chem. Eur. J. 2015, 21, 4671 – 4687

4671

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

6

4

both ligands and their complexes is a vital part of coordination

tested for separation of Cu from the expected metal-ion

II

chemistry. Ligands with very high selectivity for Cu over

impurities Ni and Zn , to show that preparation and utilisation

of such selective chelating resins in radiochemistry is feasible.

II

Ni /Zn are desired for radiomedicinal purposes.

Some time ago, we started investigations of cyclam-based li-

[

25–36]

gands containing phosphorus acid pendant arms,

which

are suitable for complexation of the first-row transition metal Results and Discussion

ions. Optimal properties are exhibited especially by ligands

Synthesis

having six donor atoms (i.e., with two coordinating pendant

arms). Because of their strong complexing ability, phosphorus

acid groups seem to significantly accelerate the complexation

reaction because, taking into account the common complexa-

tion mechanism of macrocyclic chelators, they form rather

stable initial “out-of-cage” complexes that further rearrange to

Selective 1,8-disubstitution of the cyclam ring by two different

groups can be achieved by stepwise alkylation of cyclam–form-

aldehyde or cyclam–glyoxal bis-aminals followed by alkaline

[37–40]

hydrolysis of the resulting bis-quaternary intermediates.

In

such alkylation reactions, cyclam–formaldehyde bis-aminal 1 is

more reactive and quaternisation steps proceed fast. However,

although several mono-quaternary intermediates have been

isolated, and compound 1 has been successfully employed

for synthesis of some non-symmetrically substituted

“

in-cage” complexes (i.e., the central metal ion is coordinated

by all nitrogen atoms of the macrocyclic unit). It has been ob-

served that fully N-substituted cyclam derivatives (i.e., only

tertiary amino groups are present) have significantly different

properties; in particular, their complexes show significantly

lower kinetic inertness than those of ligands having one or

[38,41–43]

cyclams,

in our case, only mixtures of mono- and bis-

quaternised aminals were isolated after reaction with the first

alkylating agent: using methyl

iodide or p-nitrobenzyl bromide

gave the same result. Therefore,

the required unsymmetrical bis-

quaternary salt was always con-

taminated

by

the

sym-

metrical one originating from

the first reaction step. Thus, sym-

metrically substituted 1,8-di-

methyl or 1,8-bis(p-nitrobenzyl)

cyclams were isolated as by-

products after hydrolysis of the

crude (impure) bis-quaternary

salts (Supporting Information,

Scheme S1), and isolation of the

product 5 required chromato-

graphic separation. One of the

by-products, namely, 1,8-bis(p-ni-

trobenzyl)cyclam, was structural-

ly characterised by single-crystal

X-ray diffraction (Supporting In-

formation, Figure S1).

Scheme 1. Structures of ligands mentioned in the text.

For the above reasons,

cyclam–glyoxal cis-bis-aminal 2

was employed in the quaternisa-

[33,39]

more secondary amino groups. Previously, we have shown that

the ligand 1,8-H te2p-Me (Scheme 1) forms a copper(II) com-

tion reactions (Scheme 2).

In this case, monoquaternary

salts can be isolated in pure form, because the difference in re-

activity between the first and second alkylation steps is signifi-

cant.

4

2

plex with very high selectivity, and acid-assisted dissociation of

the complex is faster than that observed for non-methylated

[

35]

derivative 1,8-H te2p.

The reaction of a monomethyl quaternary aminal with p-ni-

trobenzyl bromide is extremely slow, and therefore the reverse

strategy was used. First, the mono(p-nitrobenzyl) quaternary

4

Thus, we consider that analogous ligands and their conju-

gates can be employed for selective copper(II) separation from

other metal ions, which can be coupled with the possibility to

recover copper(II) by using strong acids. Herein, we report on

coordinating properties of 4-methyl-11-(p-aminobenzyl)cyclam-

based ligands with two methylphosphinic or methylphosphon-

ic acid pendant arms (H L1 and H L2, respectively; Scheme 1).

[33]

salt 3 was isolated, and a suspension thereof in acetonitrile

was allowed to react with a large excess of methyl iodide for

two weeks at room temperature to afford bis-quaternary salt

4, the identity of which was confirmed by X-ray diffraction

(Supporting Information, Figure S2). Surprisingly, in the second

alkylation step, bromide/iodide anion exchange was observed

2

4

The ligands were conjugated with a hydrophilic polymer and

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4672

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

2 4

Scheme 2. Synthesis of H L1 and H L2.

and was also confirmed by elemental analysis of the bulk ma-

terial.

(strong cation-exchange resin, elution with pyridine) in oily/

glassy forms.

II

The kind of base used in the next, hydrolytic step is crucial;

when NaOH or KOH was used, elimination of p-nitrobenzyl

group was observed, and it was followed by further reactions

and, according to TLC, several (unidentified) by-products were

formed besides the required disubstituted cyclam 5. In these

cases, isolation of amine 5 required chromatography, and the

yields of isolated product were rather low (ca. 50%). This

problem was solved by using LiOH, with which the hydrolysis

proceeded without side reactions, and the required product 5

was isolated almost quantitatively after a simple extraction.

Compound 5 is a crystalline solid, and its identity was un-

ambiguously confirmed by single-crystal X-ray diffraction

On mixing ligands H L1 and H L2 with solutions of Cu salts,

2

4

fast complexation proceeds with an immediate colour change

II

to deep green. The Cu complexes of both ligands were found

to be stable towards isomerisation, since neither UV/Vis spec-

tral changes nor new TLC spot(s) were observed after heating

slightly acidic, neutral or alkaline aqueous solutions of the

complexes to reflux for 2 d or allowing the stock solutions of

the complexes to stand at room temperature for several

months.

Thermodynamic properties

(Supporting Information, Figure S3). Although the molecule of

Thermodynamic properties of H L1 and H L2 and their com-

2

4

5

is partly disordered, it is clear that the macrocycle adopts an

plexes were studied by potentiometric titration. Protonation

constants of the free ligands are in the expected range of

values reported for related compounds (Table 1 and Support-

ing Information, Table S1), and the corresponding distribution

diagrams are shown in Figure 1. Starting from fully deprotonat-

[

1,44]

unusual geometry related to Bosnich conformation V.

In further reaction steps, phosphorus acid pendant arms

were introduced by Mannich-type reactions between amine 5,

formaldehyde and hypophosphorous acid (leading to com-

pound 6 after purification by ion-exchange chromatography)

or triethyl phosphite (affording isolated intermediate ester 7,

which was hydrolysed through trimethylsilyl bromide transes-

terification to give compound 8). The aromatic nitro group of

compounds 6 and 8 was then reduced to an amino group by

sodium sulfide and Raney nickel-catalyzed reduction with

borohydride, respectively. Ligands H L1 and H L2 were isolated

2

ꢀ

4ꢀ

ed species L1 and L2 , the first two protonation steps,

which are characterised by logK values of 10.32 and 9.87 to

h

ꢀ

3ꢀ

give HL1 and H L1 species, and 11.43 and 11.42 to give HL2

2

ꢀ

and H L2 , respectively, correspond to protonation of the

2

amino groups of the macrocycle. The doubly protonated

cyclam skeleton is locked in a stable conformation due to the

[1]

presence of intramolecular hydrogen bonds. In cyclam itself

and in several of its derivatives, such hydrogen bonds even

2

4

in zwitterionic form after ion-exchange chromatography

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4673

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

[46]

gous processes in related compounds. Further protonation

[

a]

Table 1. Stepwise protonation constants logK

h

of H

2

L1 and H

4

L2 (0.1m

of H L2 occurs on

a

distant aromatic amino group

4

KNO , 258C) and comparison with those of related ligands.

3

+

(

logK (H L2 )=3.61) and corresponds well with the analogous

5

[

25]

[25]

[47]

[45]

h

H

2

L1

H

4

L2

H

4

te2p

H

4

te2p-Me

2

cyclam

tmc

H

4

teta

+

process on the phosphinate ligand (logK (H L1 )=3.44). The

3

1

2

3

4

5

6

10.32 11.43

9.87 11.42 26.41

3.44 7.46 6.78

–

11.47

12.17

7.20

6.33

1.52

–

11.4

10.28

1.6

2.1

–

9.36

9.02

2.54

2.25

–

10.7

10.13

4.10

3.27

–

last protonation constant found for the phosphonate ligand

[

b]

2+

(

logK (H L2 )=1.32) describes protonation of some of re-

6 6

maining amino groups of the macrocycle. Further protonation

constants were not determined, because they are not

accessible by potentiometry. The assignment of protonation

sites described above and values of the protonation constants

–

6.25 5.36

3.61 1.15

1.32

-

–

h h

[a] K =[H L]/{[H]·[Hhꢀ1L]}. [b] Overall value for association of two protons.

1

31

were fully confirmed by the dependence of H and P NMR

shifts on pH (Supporting Information, Figure S5 and Table S1).

Some minor discrepancies between the protonation constants

obtained by NMR spectroscopy and potentiometry (Supporting

Information, Table S1) can be attributed to less precise calibra-

tion of the small combined glass electrode used in the NMR

experiment (an electrode for direct pH measurement in NMR

tubes was used, and it was calibrated by a standard three-

point calibration with pH 4, 7 and 10 buffers) and uncontrolled

ionic strength in the NMR experiments.

II

To reach full equilibrium in the M –ligand mixtures, the sys-

Figure 1. Distribution diagrams of H

2

L1 (left) and H

4

L2 (right); c

L

=0.004m.

tems were studied by an out-of-cell method. The mixtures

were equilibrated for 5 d (Cu, Zn) or six weeks (Ni). Stability

constants of complexes of the title ligands are compiled in

Table 2 and Table S2 of the Supporting Information, and corre-

[

35]

lead to a reverse order of consecutive protonation constants.

The first deprotonation of H L species stabilised by intramolec-

2

ular hydrogen bonds opens the locked macrocycle conforma-

tion, and the second deprotonation proceeds much more

easily due to the greater flexibility of the HL species. This leads

to a decrease in the solution abundance of the HL species and

consequently to a change in the formal order of the corre-

sponding protonation constants if the maximal abundance of

the HL species is just one-third or lower, as is graphically

shown in Figure S4 of the Supporting Information. If the maxi-

mal abundance of the HL species is lower than 33.33%, the

[

a]

Table 2. Stability constants logK(LM) and the corresponding stepwise

[a]

protonation constants logKh11 of

H

2

4

L1 and H L2 complexes with

selected metal ions (0.1m KNO , 258C).

3

Constant

H

2

L1

H

4

L2

II

Ni

Cu

Zn

Ni

Cu

Zn

logK(LM)

logK111(HLM)

logK211(H

logK₃₁₁(H LM)

11.09

5.4

–

18.25

3.87

–

11.02

5.48

–

16.09

7.62

5.0

23.29

6.90

5.14

3.95

16.26

7.35

5.20

5.05

2

LM)

abundance curves of the H L/HL pair cross each other at a pH

–

6.1

2

3

(

logK₂value) higher than that of the crossing point of the

[

h

a] Kh11 =[H LM]/{[H]·[Hhꢀ1LM]}.

abundance curves of the HL/L pair (logK value). This results in

1

the order logK (HL)<logK (H L). For H L2, this phenomenon

a

2

4

(

Figure 1) leads to very close values of the first two consecu-

sponding distribution diagrams are shown in Figure 2. The

II

tive protonation constants. The overall basicity over these two

highest stability was found for Cu complexes, in accordance

steps logb (Supporting Information, Table S1) is lower for the

with the general Irving–Williams trend. Both ligands show very

2

II

phosphinate ligand H L1 (20.19) compared to the phospho-

high selectivity for Cu ions, as can be seen from a comparison

2

II

nate ligand H L2 (22.85) as well as to acetate ligands (e.g.,

of stability constants of Cu complexes with those of Ni and

4

[

45]

II

H teta, 20.83) due to the electron-withdrawing character of

Zn complexes; the differences are greater than seven orders

4

[

46]

II

the phosphinate moieties.

Comparing bis(methylphospho-

of magnitude. For a given ligand, stabilities of the Ni and Zn

nate) ligands, the basicity of the ring amino groups of H L2 is

complexes are very similar.

4

[

25]

slightly lower than that of H te2p-Me (logb =23.64) due to

For the H L1 complexes, the protonation constant of the

4

2

+

the presence of an electron-withdrawing benzyl group. How-

ever, amine basicity of these two phosphonate derivatives with

[M(HL1)] species is comparable to logK₃of the aromatic

II

amino group of the free ligand only for the Cu complex. For

II

tertiary amino groups is much lower than that of H te2p

the Ni and Zn complexes, these logK values are higher by

4

111

[

25]

(

logb =26.41) with generally more basic secondary amino

about two orders of magnitude. Thus, one can suggest that

equilibrium between deprotonated and monoprotonated spe-

2

[47]

groups, similar to difference between cyclam (logb =21.68)

2

[47]

II

and its N,N’,N’’,N’’’-tetramethyl derivative tmc (logb =18.38).

cies in the Ni /Zn –H L1 systems includes not only protonation

2

Two further protonations of the phosphonate ligand occur

of the distant primary aromatic amino group, but also partial

protonation of ring amino groups; otherwise, there would be

on the phosphonate moieties, and the corresponding logK

h

ꢀ

+

values for formation of H L2 and H L2 are 7.46 and 6.25, re-

no reason for such high logK₁₁₁values of the [M(HL1)]

3

4

spectively; these values conform with those reported for analo-

species. Therefore, full “in-cage” complexation (i.e., binding by

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4674

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

II

Table 3. Comparison of stability constants of complexes of M with H

2

L1,

H

4

L2 and related ligands.

[

a]

Ligand

logbNiL logbCuL

logbZnL pCu

H

2

L1

L2

te2p

11.09

16.09

–

18.25

23.29

23.94 (pc), 24.08

11.02

16.26

17.16

7.63

8.71

8.94 (pc), 9.01

(trans)

8.13 (pc), 8.63

ABn[48]

[b]

[

c]

[c]

(

trans)

[

32]

[b]

[32]

[b]

H

4

te2p

21.99

15.55

25.40 (pc), 26.50

20.35

17.56

[

c,26]

[c]

(

trans)

(trans)

8.74 (pc)

[

b]

[b]

H

4

te2p-

24.03 (pc)

[

35]

Me

cyclam

2

[

47]

22.2

8.65

20.00

28.1

18.3

21.74

15.2

10.4

16.62

11.81

8.55

9.03

[

47]

tmc

teta

[

47]

H

4

II

[

(

[

a] pCu is the negative logarithm of free Cu concentration in solution

cCu =c

L

=0.004m, pH 7.4). [b] pc=pentacoordinate kinetic isomer.

c] trans=trans-O,O’ octahedral thermodynamic isomer.

[46]

sufficient to compare the stabilities of the complexes,

because the ligands also differ significantly in proton affinity,

and metal-ion complexation and protonation are concurrent

II

Figure 2. Distribution diagrams of metal-containing species in the M –H

2

L1

II

processes. Therefore, the ꢀlog[Cu ]=pCu parameter was also

(

left) and M –H

middle) and Zn (bottom).

4

L2 (right) systems; c

L

=c

M

=0.004m; M=Ni (top), Cu

calculated under the conditions used in potentiometry (c =

Cu

c =0.004m) and at physiological pH 7.4. The stabilities of the

L

II

Cu complexes of the novel ligands are comparable/only

all nitrogen atoms of the macrocycle) in protonated species at

slightly lower than those of other cyclam derivatives but still

sufficient for possible in vivo radiocopper applications. The ob-

served high selectivity for divalent copper can be employed in

the removal of copper from mixtures of metal ions, for

example, in separation/purification of radiocopper (see below).

+

acidic pH is suggested only for the [Cu(HL1)] complex.

For complexes of the phosphonate ligand H L2, the first two

4

protonations of the complexes should take place on phospho-

[

26,34,35]

nate groups.

The resulting doubly protonated species

are very probably “in-cage” complexes for all three metal ions

ꢀ

(

a PO H group can be firmly bound to central metal

3

Formation kinetics of the copper(II) complexes

[

26,30,35]

ions),

as almost all corresponding logK values (5.0–7.6,

h

Table 2) are lower than the analogous ones found for the free

ligand (7.46 and 6.25, Table 1). Such a decrease is attributable

to the electron-withdrawing effect of the coordinated metal

ion and it indirectly confirms the expected coordination of

a phosphonate pendant arm. Similarly to the phosphinate

ligand (see above), the next protonation of the complex spe-

Since the complexation rate is a central point for possible

radiochemical applications, formation kinetics of copper(II)

complexes was studied in detail. Formation of copper(II) com-

plexes of the title ligands was monitored directly by UV/Vis

spectroscopy in the pH range 2.4–7.2 (chloroacetate, acetate,

phosphate and 2-(N-morpholino)ethanesulfonic acid (MES) buf-

fers). At pH<4.5, conventional measurements were employed.

At pH>4.5, the stopped-flow technique was used, since the

complexation reactions were very fast. The formation kinetics

cies probably occurs fully on the distant aromatic amino group

II

only for the Cu complex. Further protonation of [M(H L2)] to

2

+

II

[

M(H L2)] is significantly easier for the Ni and Zn complexes

3

II

and, compared to the analogous protonation of the free

ligand, it points again to partial protonation of the amino

groups of the macrocycle in these complexes, similar to that

correspond to a first-order process with respect to Cu , as the

f

II

dependence of k on Cu concentration at constant pH is

obs

linear (Supporting Information, Figure S7). A negligible role of

the coordinating anions in acetate or phosphate buffers on the

complexation process was confirmed by the independence of

suggested above for H L1.

The high complexation selectivity for Cu over Ni and Zn

2

II

f

can be clearly seen from distribution diagrams of the M –H L1/

k_o_b_sfrom buffer concentration at constant pH (Supporting In-

2

H L2 systems (Figure 2) and from summary complexation dia-

formation, Figure S8). When acetate and phosphate buffers

were used at similar pH values, the experiments gave similar

results with expected trends and showed no difference be-

tween both buffers (Supporting Information, Figure S8). Coor-

dination ability of chloroacetate and MES buffers was also ex-

4

grams shown in Figure S6 of the Supporting Information. Full

II

“

in-cage” complexation of Cu is complete at a pH value of

II

about 4, whereas complexations of Ni and Zn start at this pH

and are complete at a pH value of about 7 (H L2) or even

4

[49]

higher (H L1).

pected to be negligible. The complexation reactions were

monitored under pseudo-first-order conditions with a large

2

II

The stability constants of the M –H L1/H L2 complexes are

2

4

II

compared with those of complexes of related ligands in

Table 3. However, values of stability constants alone are not

excess of Cu ions, and the observed rate constants were divid-

II

ed by Cu concentration to obtain second-order rate constants

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4675

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

II

Scheme 3. Proposed mechanism of complexation of the Cu –H

2

L1 system.

f

Figure 3. pH dependence of second-order rate constant k

2

for complexation

L2 (empty triangles)

systems (t=25.08C, I=0.1m KCl). Lines represent the best fits to Equation

II

2 4

reactions in the Cu –H L1 (full diamonds) and Cu –H

II

(

1) for the Cu –H

2 4

L1 system and Equation (2) for the Cu –H L2 system with

values from Table 4.

f

k . The k values are compiled in Tables S3 and S4 of the Sup-

2

porting Information and shown in Figure 3. The overlay of the

f

pH dependence of k and the distribution diagrams of the free

2

ligands (Supporting Information, Figure S9) clearly shows that

several differently protonated species must be taken into ac-

count for treatment of the formation kinetics data. For the

II

ꢀ

Cu –H L1 system, the HL1 species must be included in the fit-

2

ting (although its abundance is negligible in the studied pH

f

range), since k is still increasing with pH even in the pH range

2

5.5–7, in which the ligand is present to approximately 100% as

the zwitterionic H L1 form. Therefore, the data were fitted by

2

testing different combinations of reacting species, with ligand

protonation constants fixed to the values found by potentiom-

etry (Table 1). The best results were obtained by using Equa-

II

Scheme 4. Proposed mechanism of complexation of the Cu –H

4

L2 system.

II

tions (1) and (2) for the Cu –H L1 and Cu –H L2 sys-

2

4

tems, respectively, where K are stepwise protonation

h

Table 4. Rate constants corresponding to differently protonated ligand species

constants of the free ligands and species reactivities

II

during formation of the Cu –H

2

L1 and Cu –H

4

L2 complexes and comparison with

f

are characterised by k(H L) constants. It corresponds

h

those of related ligands.

to reaction mechanisms shown in Schemes 3 and 4,

and contributions of individual species to the overall

[

26]

[51]

Rate constant

H

2

L1

H

–

4

L2

H

4

te2p

H

Me

4

te2p- cyclam

tmc

[

35]

2

reactivity are shown in Figures S10 and S11 of the

ꢀ1

3

[

s mol dm ]

f

Supporting Information. Rate constants k(H L) corre-

h

f

6

6 [52]

5

k(HL)

1.72(9)10

–

1.0510 ,

2.910

-

sponding to differently protonated ligand species are

listed in Table 4.

6 [53]

1.810 ,

6

[54]

8.010

f

6

5

6

[52]

k(H

2

L)

2.98(16)

3.66(17)10 1.9710 2.810 0.135,

[

53]

f

þ

0.39,

kðHLÞ þ kðH LÞ ꢁ K ꢁ ½H ꢂ

f

2

[54]

k₂¼

ð1Þ

0.076

þ

2

K ꢁ ½H ꢂ þ K ꢁ K ꢁ ½H ꢂ

f

3

2

3

k(H

3

L)

–

9.1(7)10

11.7(9)

1.3810 2.810

0.17 0.04

–

4

f

þ

f

þ

kðH LÞ þ kðH LÞ ꢁ K ꢁ ½H ꢂ þ kðH LÞ ꢁ K ꢁ K ꢁ ½H ꢂ

f

2

3

4

3

4

^k2 ^¼

þ

2

þ

3

1

þ K ꢁ ½H ꢂ þ K ꢁ K ꢁ ½H ꢂ þ K ꢁ K ꢁ K ꢁ ½H ꢂ

3

4

3

4

5

ð2Þ

comparable to those of other ligands with two phosphonate

[

26]

[35]

pendant arms (H te2p and parent H te2p-Me , Supporting

4

2

Figure 3 shows that phosphonate ligand H L2 binds cop-

Information, Figure S12). The complexation proceeds via analo-

gous protonated species with similar reactivities of the individ-

ual species (Table 4). In contrast, the reactivity of phosphinate

4

per(II) faster than phosphinate ligand H L1 by 1–2 order(s) of

2

magnitude, with increasing difference on going to higher pH.

Compared to other related systems reported in the literature,

ligand H L1 is similar (Supporting Information, Figure S12) to

2

the overall reaction rate of the phosphonate ligand H L2 is

that of N-benzylcyclam and, based on data published previous-

4

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4676

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

[

51]

[52–54]

ly, to those of tmc and cyclam itself

(for formulae, see

do not affect the reaction significantly compared with the

simple ligands. This supports reports made on complexation of

Scheme 1). N-Benzylcyclam was chosen as a simple model de-

rivative for comparison; it emulates our bifunctional derivatives

by having a related side group but no coordinating pendant

arms.

64

NCA Cu with phosphonate-containing cross-bridged cyclam

64

derivatives claiming fast and quantitative Cu binding even at

[55]

room temperature. Anyway, reactivity of the monoprotonat-

ed species points to the fact that the presence of weakly coor-

dinating phosphinate pendant arms somewhat accelerates “in-

Efficiency of copper(II) incorporation by these ligands was

6

4

also tested radiochemically at pH 5.5. Cu labelling is faster in

comparison to the UV/Vis experiments, and this is caused by

cage” complex formation with H L1 compared to that with

2

64

II [50]

a large excess of the ligands over Cu . The phosphonate

tmc; however, if the different ligand basicities are taken into

account, the overall reactivities of the two ligands are similar

(Supporting Information, Figure S12).

6

4

II

ligand H L2 forms Cu complexes at ambient temperature

4

very rapidly (radiochemical yield >95% within 10 min). Similar

results were obtained for cyclam and N-benzylcyclam. Howev-

er, under these conditions, complexation with the phosphinate

ligand H L1 requires a longer time; only about 60% of H L1 is

labelled with Cu after 10 min. Full complexation requires

0 min at ambient temperature. However, the labelling efficacy

of this ligand was significantly increased at higher tempera-

2

Dissociation kinetics of the copper(II) complexes

64

The acid-assisted dissociation of copper(II) complexes of H L1

9

2

and H L2 was investigated in 0.025–4.89m ^H^C^l^O4 and

4

64

(

H,Na)ClO with an ionic strength of 5.0m at room and elevat-

ture. Almost quantitative Cu labelling of H L1 was observed

at 508C within 10 min.

4

2

d

ed temperatures. The experimentally observed k values are

obs

For ligand H L2, in the studied pH range, the most reactive

compiled in Tables S5 and S6 of the Supporting Information.

The data of complexes of both ligands were successfully fitted

by Equation (3) corresponding to the mechanism shown in

Scheme 5. The results are listed in Table 5, and the best fits are

shown in Figure 4.

4

2ꢀ

species is doubly protonated H L2 , in which both protons

2

are bound to the nitrogen atoms of the macrocycle (see

above) and the phosphonate pendant arms are fully deproton-

ated. The reactivity of this species fully governs the overall rate

of complex formation above a pH value of about 5 (Support-

ing Information, Figure S11). On protonation of the phospho-

d

þ

k

þ k

ꢁ K

ꢁ ½H ꢂ

d

0

1

H

k_o_b_s

¼

ð3Þ

ꢀ

þ

nate moieties affording H L2 and H L2, the reactivity gradual-

1 þ K

H

ꢁ ½H ꢂ

3

4

ly decreases and, for the neutral zwitterionic H L2, is similar to

4

that of the analogous zwitterionic form of phosphinate ligand

H L1 (i.e., doubly protonated macrocycle with each of pendant

2

arms bearing a charge of 1ꢀ; see Table 4). Contrary to phos-

phonate ligand H L2, the most reactive species of phosphinate

4

ꢀ

ligand H L1 in the studied pH range is HL1 , which drives the

2

overall reaction above about pH 4.5 (Supporting Information,

Figure S10). The monoprotonated species is generated much

more easily for H L1 than for H L2 due to the significantly

II

2

Scheme 5. Proposed mechanism of acid-assisted dissociation Cu –H L1 and

2

4

II

4

Cu –H L2 complexes.

lower logK value of the ring amino group (9.87 vs. 11.42, re-

2

spectively; Table 1) and uncharged nature of H L1. Therefore,

2

ꢀ

HL1 can significantly contribute to the overall complexation

The number of protons in kinetically active species is obvi-

ously different in the two complexes (Scheme 5) due to the

rate even in the slightly acidic region. The corresponding rate

f

constant k(HL) is somewhat higher than that reported for tmc

protonatable phosphonate groups in H L2. According to distri-

4

and is comparable to that of cyclam itself (Table 4).

bution diagrams (Figure 2), protonation of the distant amino

The results clearly show that the nature of pendant arms is

decisive for the rate of formation of copper(II) complexes of

the macrocyclic ligands. The highest overall reaction rate was

observed for ligands containing phosphonate group(s), which

have higher complexation ability than phosphinate group(s).

Considering the general mechanism of complex formation by

macrocycle ligands, whereby formation of weak “out-of-cage”

complexes proceeds the final “in-cage” coordination, a higher

abundance of the “out-of-cage” species should accelerate over-

all complex formation. Therefore, stronger affinity of the phos-

phonates to metal ions should lead to a more abundant

group of the aminobenzyl unit is expected for systems with

d

both ligands, and thus probably mono- (n=1, k ) and diproto-

0

d

II

nated ( k ) species are kinetically important for the Cu –H L1

1

2

d

complex, and tri- (n=3, k ) and tetraprotonated ( k ) are kinet-

0

1

II

ically decisive for the Cu –H L2 complex. Although a simplified

equation with k =0 is commonly sufficient for good fit of

4

d

0

II

experimental data (e.g., as was used for the Cu –H te2p-Me

4

2

[35]

complex),

inclusion of this particular rate constant

significantly improved fits by lowering the observed sum of

squared deviations by 20–30%. However, this rate constant

was calculated with relatively large uncertainty (Table 5).

“

out-of-cage” complex and to an increase in effective copper(II)

The dissociation rates and their temperature dependences

are very similar for both studied systems. Activation parame-

ters (obtained from linearisations shown in Supporting Infor-

mation, Figures S13–S15) of the proton-assisted dissociation

concentration in the vicinity of the macrocyclic unit; this

accelerates the rate-determining “in-cage” binding. Weakly

coordinating pendant groups such as N-methylphosphinates

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4677

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

d

step (characterised by k ) are comparable with those reported

1

II

[35]

for the Cu –H te2p-Me complex

and indicate the same

4

2

decomplexation mechanism. The values of the protonation

constants describing the equilibrium between the dissociating

species (logK =pK ꢃ0.7–0.8, Table 5) indicate initial protona-

H

A

tion of the phosphorus acid moiety. Such values also support

the estimated number of protons in the reactive species (see

above). Both studied complexes exhibit lower inertness to-

II

wards acid-assisted decomplexation than the Cu –H te2p-Me

4

2

[35]

complex and much lower inertness than those of the ligands

[26,48]

with secondary amino groups.

Decomplexation half-lives

of complexes of the title ligands are compared with those of

complexes of selected related ligands in Table 6.

The results show that copper(II) complexes of N-tetrasubsti-

tuted cyclam derivatives are generally much more kinetically

labile than complexes of less N-substituted cyclam-based

ligands with secondary amino groups. Therefore, the title bi-

functional ligands may be convenient for applications in which

fast decomplexation is required, for example, separation of

copper(II) ions (see below).

Synthesis and characterisation of ion-selective resins

The selectivity of the studied ligands towards copper(II) com-

plexation and relatively low kinetic inertness can be potentially

employed for separation of copper(II) from mixtures of other

metal ions due to its selective complexation. Separation/re-

moval of copper(II) ions from other metal ions is of interest, for

example, in the treatment of Wilson’s disease (metabolic

disorder causing overload of the human body with divalent

d

II

Figure 4. Dependence of kobs for acid-assisted decomplexation of Cu –H

2

L1

L2 (bottom) complexes at T=25, 35, 45 and 558C (empty

diamonds, full triangles, empty squares and full circles, respectively);

II

(top) and Cu –H

4

I=5.0m (H,Na)ClO

4

. Lines represent the best fits to Equation (3) for values

[57–61]

[62–65]

copper),

in the preparation of copper radioisotopes

from Table 5.

II

Table 5. Kinetic parameters of acid-assisted dissociation of Cu –H

2

L1 and Cu –H

4

L2 complexes.

Rate/equilibrium

constants

Temperature [8C]

Activation/thermodynamic

parameters

25

35

45

55

II

Cu –H

2

L1

d

ꢀ1

ꢀ3

ꢀ2

ꢀ3

ꢀ2

ꢀ1[a]

k

0

1

[s

]

4.33(95)10

1.83(12)10

7.4(2.4)10

1.55(44)10

8.29(51)10

1.0(1.0)10

a

E =29.2(13.0) kJmol

°

ꢀ1[b]

DH =26.6(13.0) kJmol

°

ꢀ1

ꢀ1[b]

DS =ꢀ200(42) JK mol

d

ꢀ2

ꢀ1

ꢀ1[a]

3.95(22)10

1.48(7)10

a

E =57.1(1.5) kJmol

°

ꢀ1[b]

DH =54.5(1.5) kJmol

°

ꢀ1

ꢀ1[b]

DS =ꢀ95(5) JK mol

DH=16.9(8.8) kJmol

ꢀ

1

3

ꢀ1[c]

K

H

[mol dm ]

4.6(2.0)

0.66

6.5(2.6)

0.81

5.3(2.1)

0.72

9.9(3.4)

1.00

ꢀ

1

DS=69(28) JK mol

logK

H

II

4

Cu –H L2

d

ꢀ1

ꢀ3

ꢀ2

ꢀ1[a]

k

0

1

[s

]

8.03(77)10

1.65(12)10

2.76(52)10

5.15(97)10

a

E =49.6(1.8) kJmol

°

ꢀ1[b]

DH =47.0(1.8) kJmol

°

ꢀ1

ꢀ1[b]

DS =ꢀ127(6) JK mol

d

ꢀ1

ꢀ2

ꢀ1

ꢀ1[a]

k

[s

1.81(6)10

4.08(18)10

7.07(42)10

1.34(6)10

a

E =53.4(2.3) kJmol

°

ꢀ1[b]

DH =50.8(2.3) kJmol

°

ꢀ1

ꢀ1[b]

DS =ꢀ108(7) JK mol

DH=9.5(14.9) kJmol

ꢀ

1

3

ꢀ1[c]

K

H

[mol dm ]

5.1(1.9)

0.71

2.68(87)

0.43

5.3(3.2)

0.72

6.1(3.1)

0.79

ꢀ1

DS=43(48) JK mol

logK

H

°

[a] Arrhenius model: [lnk=ꢀ(E

a

/RT)+lnA]. [b] Eyring model: [ln(k/T)=ꢀ(DH /RT)+(DS /R)+ln(k

B

/h)]. [c] lnK

H

=ꢀ(DH/RT)+(DS/R).

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4678

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

They were treated with the monosodium salt of resorcinol in

II

Table 6. Comparison of dissociation half-lives of Cu –H

2 4

L1 and Cu –H L2

[76]

analogy to the reaction with phenol to afford polymer-im-

mobilised 1-hydroxy-3-alkoxybenzene moieties as highly acti-

vated, passive azo coupling components (Scheme 6). Sodium

resorcinolate was used instead of resorcinol to increase the nu-

cleophilicity and facilitate reaction with the polymeric epox-

complexes with those of complexes of related ligands (258C, cH+ =1m,

I=5m (H,Na)ClO ).

4

Ligand

t

1

=

2

H

2

4

L1

L2

te2p-Me

44 s

42 s

2.5 min

9.1 min

19.7 min

20.9 min

13

1

ides. Successful reaction was confirmed by solid-state C{ H}

NMR spectra, in which aromatic carbon signals of the product

[

35]

2

N-benzylcyclam

[

a][26]

H

4

te2p (pc)

were found to be equivalent to the resorcinol content of

[

56]

ꢀ1

cyclam

0

.974 mmolg .

[a] pc=pentacoordinated kinetic isomer.

The polymer-immobilised resorcinol was further used as

a passive azo coupling component for diazotised macrocyclic

ligands. This synthetic strategy was modified from previously

published immobilisation of chelating agents on soluble

and in technological applications (e.g., removal of copper(II)/

[

66–74]

[77]

other ions from wastewater).

polymers containing phenolic moieties. The progress of the

To test our ligands in radiocopper purification, the ligands

were covalently immobilised on hydrophilic, macroporous

polymer beads (Scheme 6). The rate of adsorption of metal

ions on polymer beads generally depends on the rate of com-

plexation by immobilised ligands and on transport phenom-

ena, especially on the diffusion rate and the accessibility of the

chelating groups. Thus, macroporous hydrophilic beads were

chosen, because they ensure both rapid diffusion and good ac-

cessibility of the complexing groups. If a rapidly complexing

chelator is covalently attached to a macroporous matrix,

uptake of metal ions from aqueous solution typically proceeds

reaction was obvious due to the development of an intense

orange colour caused by formation of azo-group chromo-

31

1

phores; it was confirmed by IR spectroscopy and P{ H} solid-

state NMR spectroscopy. The IR spectrum showed significant

ꢀ

1

increase in OH signal intensity at 3100–3600 cm , correspond-

ing to successful loading of OH-rich macrocyclic ligands

(Supporting Information, Figure S16). Other signals typical for

the ligand (P=O etc.) could not be detected due to overlap

ꢀ

1

with GMA-resorcinol signals in the region 500–1500 cm . The

13

1

change in the C{ H} solid-state NMR spectra was not signifi-

31

1

cant; however, the P{ H} solid-state NMR spectrum clearly

showed the appearance of broad signals centred at 21 and

18 ppm for resin-H L1 and resin-H L2 product materials,

[

58]

within a few minutes in a batch experiment. We chose poly-

(

glycidyl methacrylate-co-ethylene dimethacrylate) (GMA)

2

4

[

75]

beads as a starting material

containing reactive epoxide

respectively. Ligand contents in the polymers, determined by

ꢀ

1

groups that are suitable for further synthetic transformations.

copper(II) chelation capacity, were 65 and 44 mmolg

for

resin-H L1 and resin-H L2, respectively, which are sufficient for

2

4

the intended use in radionuclide separation.

The advantage of this simple two-step immobilisation is visi-

bility of the reaction progress (formation of orange azo dye,

Scheme 6) as well as a lack of charged components except the

ligands themselves on the bead surface in the relevant pH

range, so that the beads do not behave as a non-selective

electrostatic ion exchanger. Other coupling conditions were

also tested. Direct reaction of the aromatic primary amino

groups of the bifunctional ligand with epoxide groups gave an

unsatisfactory yield, whereas aminolysis of a chlorosulfonated

polymer support with the aromatic primary amino groups of

the ligand left many free sulfo groups (formed after complete

hydrolysis of the polymers), and thus the final product be-

haved as a non-selective strong cation exchanger rather than

a selective chelating resin.

The azo-dye-loaded macrocyclic ligands were tested for

sorption of trace amounts of copper(II) in the presence of

large excess of zinc(II) and nickel (II) ions (ca. 400–500 molar

excess), which is a model for the intended application, that is

workup of cyclotron targets for radiocopper production by

separating trace amounts of radiocopper from irradiated

nickel/zinc targets. As nickel(II)/zinc(II) concentrations, typical

values obtained after dissolution of cyclotron-irradiated target

ꢀ

3

II

discs were chosen (ꢃ10 mgcm ), and c(Cu ) was about

.02 mgcm to ensure a reasonable sensitivity of detection by

atomic absorption spectroscopy (AAS). The copper concentra-

ꢀ

3

0

Scheme 6. Schematic representation of modification of macroporous

polymer beads by ligands H

2

L1 and H

4

L2.

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4679

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

2

+

Table 7. Relative concentrations (c(M ) in % of the original amount) of metal ions found in supernatants after metal loading, washing and acid-assisted

ꢀ

3

desorption. The nickel(II)/zinc(II) concentrations in the primary loading solution were about 10 mgcm , and the copper(II) concentration in primary load-

ꢀ

3

ing solution was about 0.02 mgcm

.

Supernatant/parameter of sorption

pH 3

pH 5

Cu/Ni binding

Cu/Zn binding

Cu/Ni binding

Cu/Zn binding

c(Cu)

c(Ni)

c(Cu)

c(Zn)

c(Cu)

c(Ni)

c(Cu)

c(Zn)

resin-H

2

L1

supernatant after metal loading

washing, 1st cycle

washing, 2nd cycle

67.0

5.2

3.0

22.2

2.6

24.8

93.3

6.4

0.3

0.0

0

72.5

4.2

2.7

18.4

2.2

20.6

95.3

4.4

0.3

0.0

0

4.7

6.9

7.1

72.4

8.9

81.3

91.1

8.3

0.5

0.0

0

4.3

7.9

6.9

72.3

8.6

80.9

93.9

5.7

0.4

0.0

0

desorption, 1st cycle

desorption, 2nd cycle

total bound metal

[

a]

3

4

selectivity coefficient Cu/M

1.710

1.210

4.410

5.410

resin-H

4

L2

supernatant after metal loading

washing, 1st cycle

washing, 2nd cycle

57.8

5.3

3.8

29.8

3.4

33.2

91.8

7.7

0.5

0.0

0

63.6

4.9

3.5

25.6

2.4

28.0

94.6

5.1

0.3

0.0

0

6.8

7.7

7.8

68.5

9.1

77.6

98.2

1.2

0.5

0.0

0

6.3

9.9

8.2

55.3

20.3

75.6

93.6

6.0

0.4

0.0

desorption, 1st cycle

desorption, 2nd cycle

total bound metal

[

a]

3

4

selectivity coefficient Cu/M

1.910

1.810

3.210

4.710

2

+

2+

[a] Selectivity coefficient is defined as {[Cu

]

adsorbed/[Cu

]

solution}/{[M

]

adsorbed/[M

]solution}.

tion was kept as low as possible to mimic cyclotron target

workup, in which radiocopper is present only in traces. The

molar amount of metal-binding sites of the complexing resin

was about twice the total amount of copper(II) in solution.

Two solution acidities were chosen (pH 3.0 and 5.0). At the

lower pH, copper(II) should be almost completely complexed

according to the distribution diagrams, whereas the other

metal ions should not be bound at all, and at the higher pH

isotopes) by these ligands. Such a pH is commonly used for

labelling of biomolecules with copper radioisotopes. Thus,

this finding may also be exploitable for selective labelling of

radiopharmaceuticals, in which impurities in solutions of

copper radionuclides disturb the labelling process.

64

nat

Separation of Cu/ Cu from Ni ions

II

Ni and Zn ions start to be bound by the title ligands

To further demonstrate possible utilisation of the resins in sep-

6

4

II

(

Figure 2). Weakly chelating chloroacetate (pH 3) and acetate

pH 5) buffers were chosen to prevent hydrolysis and/or pre-

arating Cu from non-radioactive divalent metals (Ni and Zn ),

64

experiments with an NCA [ Cu]CuCl solution were carried

2

cipitation of metal hydroxides, especially at pH 5.0. The metal

salts were 1) adsorbed by the resin in the buffered solution,

out. The protocol followed the procedure described above for

64

non-radioactive sorption experiments. The Cu isotope was

6

4

64

2) the excess of the solution was washed out by washing with

produced via the Ni(p,n) Cu reaction. Separation of Cu/

nat

64

pH 3.0 buffer (two washings, to remove any non-specifically

bound metal ions) and 3) metal ions were desorbed from the

resin with 1m aqueous hydrochloric acid (two portions). In

each supernatant, the metal content was determined by AAS.

An overview of the analytical data is given in Table S7 of the

Supporting Information, and the results are listed in Table 7.

In all cases, only trace amounts of nickel or zinc were ad-

sorbed on the sorbents at both pH 3.0 and 5.0 (concentrations

were at the detection limit of the analytical method used; see

Supporting Information, Table S7). The two modified resins

gave similar results: at pH 3.0, copper(II) is adsorbed only

partly (20–30%), but at pH 5, about 80% of copper(II) is ad-

sorbed. Copper desorption by hydrochloric acid was found to

be nearly quantitative (from copper balance); this is a prerequi-

site for reuse of the sorbent. On going from pH 3.0 to 5.0, the

selectivity coefficients increase by more than one order of

magnitude. Thus, a pH of about 5 seems to be ideal for the

most selective binding of divalent copper (and copper radio-

Cu from Ni utilises the well-known anion-exchange meth-

[78]

odology described previously.

To study selectivity of the

II

sorption process, a defined amount of natural Ni was added

64

to carrier-free [ Cu]CuCl solution. Metal sorption was per-

2

formed at pH 5.5, and the efficiency of sorption and desorption

processes was studied by radioactivity measurements (Tables 8

and 9), and was independently determined by inductively cou-

pled plasma mass spectrometry (ICP-MS) after complete decay

64

II

of the Cu radioactivity (i.e., the found Cu concentrations cor-

responds to traces of the non-radioactive copper present in

the starting solutions; Supporting Information, Tables S8 and

S9). However, ICP-MS measurements are rather qualitative, as

concentrations of metal ions in several cases lay at the edge of

detection limit (ppb) of the method.

6

4

II

For adsorption experiments, the solution containing Cu

II

and excess of natural Ni was added to the studied resin and

the suspension was incubated for 30 min. After this time, resin

was separated from the supernatant and rinsed with buffer so-

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4680

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

and ca. 16 for H L2 for each metal ion). Copper(II) ions are

6

4

Table 8. Balance of Cu radioactivity [MBq] (decay-corrected instrument

6

4

quantitatively complexed above a pH of about 3.5 (c =c , mil-

L

M

reading) after adsorption of [ Cu]CuCl

2

on resins.

limolar range). Ligand H L2 binds copper(II) ions very fast at

4

Material

resin-H

51.53

40.91 (79%)

10.54 (21%)

33.10 (64%)

7.25

2

L1

resin-H

52.18

44.36 (85%)

7.75 (15%)

37.67 (72%)

6.38

4

L2

a pH of about 5, whereby no significant complexation of

zinc(II) and nickel(II) ions occurs; copper(II) complexation by

total activity added to resin

resin, after adsorption

supernatant, after adsorption

resin, after washing (three times)

washing solutions

II

H L1 over Ni and Zn is also very selective, but the Cu com-

2

plexation rate is about two orders of magnitude slower than

that of H L2 and is comparable to that of cyclam and its deriv-

4

atives without coordinating pendant arms. The copper(II) com-

plexes of both ligands are moderately kinetically inert towards

acid-assisted dissociation (half-lives: 44 s (H L1) and 42 s (H L2)

2

4

6

4

Table 9. Balance of Cu radioactivity [MBq] (decay-corrected instrument

in 1m aq. HClO at 258C). However, their decomplexation rates

4

6

4

reading) after desorption of [ Cu]CuCl

aq. HCl.

2

from the studied resins using 1m

are much higher than those of complexes of other cyclam

derivatives and are among the highest described till now.

Therefore, the current studies have shown the great potential

of N’,N’’’-dialkyl-N,N’’-bis(methylphosphinic/methylphosphonic

acid) cyclam derivatives as building blocks of resins for selec-

tive copper(II) complexation. This was documented by prepar-

ing polymer beads modified by the title ligands through an

azo coupling reaction and studying their performance in the

separation of copper(II) from other metal ions. After complexa-

Material

resin-H

2

L1

4

resin-H L2

starting activity

33.10

7.23

25.78 (50%)

1.93

5.13

30.91 (93%)

37.67

7.43

30.06 (58%)

0.60

6.74

36.80 (98%)

resin, after 15 min desorption

supernatant, after 15 min desorption

resin, after washing (three times)

scrubbing solutions

[

a]

[a]

[b]

b]

total desorbed activity

[

a] Based on total activity loaded to resin (Table 8). [b] Based on activity

II

tion and washing off the other metal ions, the Cu ion can be

retained by resin.

released from the resin by elution with a strong acid due to

the kinetic lability of the complexes. This concept operates

well even at the very low concentrations relevant to working

with copper radionuclides, as documented by model separa-

lution (pH 5.5, three times). For desorption experiments, the

copper-loaded resin was rinsed once with 1m aqueous HCl,

and after centrifugation, again three times with 1m aqueous

HCl. After each step, radioactivities of resin and scrubbing

solutions were measured (Tables 8 and 9).

64

II

tion of NCA Cu from large excess of Ni . This proof-of-con-

cept study shows that proper tuning of ligand and polymer

properties can lead to selective resins for purification of

copper and, after appropriate modification of the chelator,

other metal radionuclides for biological/medical applications.

64

Balances of Cu activities showed retention of about 70%

of radioactivity on the resin and nearly quantitative recovery of

this amount after repeated washing. In the first washing cycle

(

1m aq. HCl), about 50–60% of total starting activity was isolat-

Experimental Section

ed (Table 9). The two resins behave similarly; however, the

phosphonate-based resin-H L2 shows slightly better para-

General

4

meters than phosphinate-containing resin-H L1. In both cases,

2

Cyclam (1,4,8,11-tetraazacyclotetradecane, Chematech), hypophos-

phorous acid (50% aq. solution, Fluka), triethyl phosphite (Fluka),

II

very good separations of radiocopper from Ni were achieved:

II

almost all Cu was adsorbed in the presence of a 900-fold

D O (99.95% D, Chemtrade) and Dowex 50 (100–200 mesh, Fluka)

2

II

[

37]

excess of Ni , and sorption of Ni was negligible under these

conditions (Supporting Information, Tables S8 and S9); this is

an important property for the recovery of expensive isotopical-

were used without further purification. Cyclam formaldehyde 1

and glyoxal 2 bis-aminals and monoquaternary salt 3 were

synthesised according to literature methods. N-Benzylcyclam was

prepared according to the published procedure. Paraformalde-

hyde was filtered from an old formaldehyde solution and dried

over P O in vacuo. NMR spectra were recorded on a Varian S300

NMR spectrometer operating at 299.9 MHz ( H), 75.4 MHz ( C) and

21.4 MHz ( P), a Varian

99.9 MHz ( H), 100.6 MHz ( C) and 161.9 ( P) or on Bruker

Avance (III) 600 spectrometer operating at 600.2 MHz ( H) and

[33]

[

79]

64

ly pure Ni. However, exact quantification of ICP-MS results is

II

problematic due to the very low content of Cu , close to the

2

5

detection limit of the method.

1

13

3

1

UNITY

1

3

INOVA 400 spectrometer operating at

1

13

31

Conclusion

1

3

1

13

Two new bifunctional cyclam-based macrocyclic ligands with

two coordinating phosphorus acid pendant arms and p-amino-

benzyl side group were synthesised by simple approaches em-

ploying glyoxal bis-aminal protection of cyclam. The results

showed that the title ligands are among the most suitable for

selective copper(II) complexation. The copper(II) complexes are

150.9 MHz ( C); internal references for H and C were TMS for

CDCl solutions and tBuOH for D O solutions, and external refer-

ence for P was 85% aq. H PO . The H and P NMR spectra of the

3

2

31

1

31

3

4

ligands at different pH were measured on the Varian S300 NMR

spectrometer in water; pH was read by a freshly calibrated com-

bined glass electrode directly in the NMR tube. The electrode was

calibrated with commercial buffer solutions (pH 4, 7 and 10). Dilute

aq. KOH and aq. HCl were used for pH adjustment. An insert tube

^t^h^e^r^m^o^d^yⁿ^a^mⁱ^c^a^l^l^y^s^t^a^b^l^e⁽^l^o^g^bCuL =18.3 and 23.3 for H L1

2

13

1

and H L2, respectively) and are formed with a high selectivity

with 85% aq. H PO was used as reference. Solid-state C{ H} and

3 4

4

31

1

over nickel(II) and zinc(II) complexes (logb_M_Lis ca. 11 for H L1

P{ H} NMR spectra were acquired at 11.7 T on a Bruker Avance III

2

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4681

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

1

3

+

1

HD 500 US/WB NMR spectrometer in 4 mm ZrO rotors. The

C

19.29 (19.63); ESI-MS (+ve): 350.0 ([M+H] , calcd 350.3); H NMR

(600.2 MHz, CDCl ): d=1.19–1.24 (m, 4H, CH CH CH ), 1.69 (s, 1H,

2

cross-polarise (CP) magic-angle spinning (MAS) NMR spectra were

measured at a spinning frequency of 10 kHz and a contact time of

ms with a repetition delay of 5 s. The P MAS NMR spectra were

3

2

CH NHCH ), 1.85 (s, 1H, CH NHCH ), 2.31 (s, 3H, NCH ), 2.40–2.50

(m, 4H, CH CH CH ), 2.54–2.60 (m, 4H, CH CH CH ), 2.61–2.65 (m,

2 2 2 2 2 2

2

3

31

1

measured at a spinning frequency of 11 kHz with a repetition delay

4H, NCH CH N), 2.68–2.76 (m, 4H, NCH CH N), 3.59 ppm (s, 2H,

2

13

3

of 10 s. External standards were used to calibrate C NMR (glycine,

NCH₂arom.), 7.49 ppm (d, 2H, arom., J_H_H=8.4 Hz), 8.11 (d, 2H,

3

1

3

13

1

76.03 ppm) and

P NMR chemical shift scales (CaHPO₄,

arom., J =8.4 Hz); C{ H} NMR (150.9 MHz, CDCl ): d=25.9, 26.0

HH 3

ꢀ

0.6 ppm). The elemental analyses were carried out at the Institute

(s, each 1C, CH CH CH ); 29.8 (s, 1C, NCH ); 43.1, 47.5, 47.6, 49.7 (s,

2 2 2 3

each 1C, CH CH CH ); 51.5, 53.3, 54.6, 58.0 (s, each 1C, NCH CH N);

2 2 2 2 2

of Macromolecular Chemistry of the Academy of Sciences of the

Czech Republic (Prague). For potentiometric titrations, metal ni-

trates (recrystallised from deionised water) were used, and their

stock solutions were standardised by titration with Na H edta ac-

58.4 (s, 1C, NCH arom.); 123.5 (s, 2C, arom.); 129.4 (s, 2C, arom.);

2

147.0 (s, 1C, arom.); 147.2 ppm (s, 1C, arom.).

2

4

1

-Methyl-11-(p-nitrobenzyl)-1,4,8,11-tetraazacyclotetradecane-

,8-bis(methylenephosphinic acid) (6): Amine 5 (2.00 g, 5.6 mmol)

[79]

cording to the recommended procedure. The stock solution of

nitric acid (ꢃ0.03m) was prepared from recrystallised KNO₃on

cation-exchange resin (Dowex 50). Carbonate-free KOH stock solu-

tion (ꢃ0.2m) was standardised against potassium hydrogenphtha-

and 50% aq. hypophosphorous acid (7.50 g, 0.11 mol, 20.0 equiv)

were mixed in water (20 mL). The mixture was heated to 408C

with stirring, and solid paraformaldehyde (0.60 g, 3.0 equiv) was

added in small portions over 60 min. The mixture was further

heated at this temperature for 5 d. Excess paraformaldehyde was

then filtered off. After concentration to a small volume under re-

duced pressure, the reaction mixture was purified on a strong

late, and the HNO solution against the KOH solution. The analyti-

3

cal-grade chemicals employed in the kinetic studies were pur-

chased from Lachema (CuCl ·2H O, ZnCl , NiCl ·6H O, KCl, KOH,

2

H PO , CH COOH, ClCH COOH) or Fluka (MES buffer). The freshly

3

4

3

2

prepared solutions of CuCl , ZnCl and NiCl were standardised by

chelatometry. Throughout the paper, pH means ꢀlog[H ].

+

2

cation-exchange resin (Dowex 50, H form, 300 mL). The excess of

[79]

+

hypophosphorous acid was eluted with water (1000 mL) and the

product 6 was collected by elution with conc. aq. HCl/H O (1/1

2

Synthesis

v/v). The fractions containing pure product were combined and

the solvents were evaporated to dryness to afford compound 6 as

a yellow oil (2.60 g). ESI-MS (+ve): 506.9 ([M+H] , calcd 506.2),

Perhydro-3a-methyl-8a-(p-nitrobenzyl)-3a,5a,8a,10a-tetraaza-

pyrenium diiodide (4): Compound 3 (10.00 g, 21.5 mmol) was sus-

pended in anhydrous acetonitrile (100 mL). After a few minutes of

stirring, an excess of methyl iodide (10.40 g, 73.3 mmol, 3.2 equiv)

was added. The mixture was stirred at room temperature for 7 d.

+

ꢀ

1

(

1

2

ꢀve): 504.7 ([MꢀH] , calcd 504.2); H NMR (600.2 MHz, D O, pD=

2

.1): d=1.72–1.76 (m, 4H, CH CH CH ), 2.28 (s, 3H, NꢀCH ), 2.56–

2

3

.89 (m, 16H, NCH CH N, CH CH CH ), 3.80 (s, 2H, NCH arom.),

2

1

7

.00 (d, 1H, NCH PO H, J_P_H=510.2 Hz), 7.05 (d, 1H, NCH PO H,

2 2 2 2

J_P_H=510.2 Hz), 7.61 (d, 2H, arom., J =7.8 Hz), 8.26 ppm (d, 2H,

arom., J =7.8 Hz); C{ H} NMR (150.9 MHz, D O, pD=1.1): d=

After this time,

a new portion of methyl iodide (10.40 g,

1

3

HH

7

3.3 mmol, 3.2 equiv) was added and the resulting suspension was

3

13

1

HH

2

stirred again at room temperature for a further 7 d. The yellow pre-

cipitate was collected by filtration, washed with anhydrous acetoni-

1

4

1

8.9, 19.4 (s, each 1C, CH CH CH ); 43.3 (s, 1C, NCH ); 46.0, 47.6,

2 2 2 3

8.1, 48.5 (s, each 1C, CH CH CH ); 49.8, 51.3, 51.5, 52.6 (s, each

2

trile (40 mL) and dried in vacuo to give 4·H O as a dark yellow

2

1

C, NCH CH N); 56.5 (d, 1C, NCH PO H, J =103 Hz); 56.9 (d, 1C,

2

1

2

CP

solid (12.04 g, 87%). Elemental analysis found (calcd for

NCH PO H; J_C_P=104 Hz); 58.7 (s, 1C, NCH arom.); 123.4 (s, 2C,

2

C H I N O ·H O, M =645.3): C 37.43 (37.22), H 4.83 (5.15), N 10.76

2

0

31 2

5

2

r

arom.); 130.8 (s, 2C, arom.); 145.9 (s, 1C, arom.); 147.1 (s, 1C,

+

(

10.85), I 38.65 (39.33); ESI-MS (+ve): 372.8 ([Mꢀ2I] , calcd 373.3);

31

1

arom.); P{ H} NMR (121.4 MHz, D O, pD=1.1): d=21.2 (s);

1

2

H NMR (600.2 MHz, [D ]DMSO): d=1.75–1.89 (m, 4H, CH CH CH ),

6

2

1.7 ppm (s).

2

.863.64 (m, 12H, CH CH CH , NCH CH N), 3.68 (s, 3H, NCH ), 4.29–

2 2 2 2 2 3

4

1

6

-Methyl-11-(p-aminobenzyl)-1,4,8,11-tetraazacyclotetradecane-

4

.37 (m, 4H, NCH CH N), 4.91 (s, 2H, NCH arom.), 4.91 (d, 1H,

2 2 2

3

,8-bis(methylenephosphinic acid) (H L1): Method 1: Compound

J_H_H=13.8 Hz, NCHN), 5.06 (d, 1H, J_H_H=13.2 Hz, NCHN), 7.86 (d,

2

3

(0.50 g) and sodium sulfide (4.82 g, 19.7 mmol, 20.0 equiv) were

2

H, J_H_H=8.4 Hz, arom.), 8.34 ppm (d, 2H, J_H_H=8.4 Hz, arom.);

1

3

1

dissolved in water (20 mL) and the solution was heated at 508C for

d. After concentration to a small volume under reduced pressure,

the residue was purified on a silica gel column (60 mL). The

column was eluted with iPrOH/conc. aq. NH /H O (7/3/3), and the

C{ H} NMR (150.9 MHz, [D ]DMSO): d=18.5, 18.7 (s, each 1C,

6

3

CH CH CH ); 39.2, 46.2, 46.4, 49.2 (s, each 1C, CH CH CH ); 50.7 (s,

2

1

2

1

C, NCH ); 50.9, 59.6, 60.1, 63.7 (s, each 1C, NCH CH N); 63.8 (s,

3 2 2

C, NCH arom.); 74.9 (s, 1C, NCHN); 75.1 (s, 1C, NCHN); 123.8 (s,

3

2

excess of sodium sulfide and other impurities were eluted first, fol-

lowed by H L1. Fractions containing the ligand (checked by

C, arom.); 132.9 (s, 2C, arom.); 135.0 (s, 1C, arom.); 148.7 ppm (s,

C, arom.).

2

31

P NMR) were combined and concentrated under reduced pres-

sure. The residue was purified on a weak cation-exchange resin

4

-Methyl-11-(p-nitrobenzyl)-1,4,8,11-tetraazacyclotetradecane

(

5): Lithium hydroxide (5.00 g, 0.12 mol, 10.0 equiv) was added to

a suspension of compound 4 (7.00 g, 10.8 mmol) in methanol

200 mL) and the mixture was stirred at room temperature for 2 d.

+

(Amberlite CG50, H form, 300 mL). Impurities were eluted with

water (500 mL), and the product H L1 was collected by using 3%

2

(

aq. HCl as eluent (500 mL). The fractions containing pure product

were combined, and the solvent was evaporated to dryness in

vacuo. The residue was further purified (to remove residual ammo-

The mixture was then concentrated to dryness under reduced

pressure and the residue redissolved in water (200 mL). The solu-

tion was extracted with dichloromethane (4300 mL). The organic

phases were combined, dried with anhydrous Na SO and volatile

substances were evaporated under reduced pressure to dryness.

The dark brown oil of crude product 5 was dissolved in a small

amount of dichloromethane and impurities were filtered off. The

clear solution was evaporated under pressure and dried in a desic-

cator over P O to give 5·0.4H O as a slowly crystallising red-brown

+

nia) on a strong cation-exchange resin (Dowex 50, H form,

2

4

3

00 mL) by elution with water followed by 5% aq. pyridine. Pyri-

dine-containing fractions were combined and evaporated to dry-

ness to afford H L1·2.5H O as a yellow oil (350 mg, 68%).

2

Method 2: Nickel(II) chloride hexahydrate (117 mg, 0.49 mmol,

0.1 equiv) was dissolved in water (15 mL) under argon atmosphere.

After few minutes, sodium borohydride (188 mg, 4.9 mmol,

1.0 equiv) was added. A solution of compound 6 (2.50 g, 5.0 mmol)

2

5

2

oil (3.80 g, 98%). Elemental analysis found (calcd for

C H N O ·0.4H O, M =356.7): C 60.88 (60.61), H 9.34 (8.99), N

1

8

31

5

2

r

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4682

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

in water/ethanol (1/1, 50 mL) alkalised to pHꢃ8 with NaOH was

added into the suspension of nickel metal. After 2 h, another por-

tion of sodium borohydride (0.95 g, 5.0 equiv) was added, and the

reaction mixture was stirred at room temperature under an argon

atmosphere for 24 h. The reaction mixture was vacuum-filtered

24 h. After concentration to a small volume under reduced pres-

sure, the reaction mixture was dissolved in anhydrous acetonitrile

(50 mL), and the resulting solution was poured into water (400 mL)

with stirring. The mixture was evaporated to dryness under re-

duced pressure and the residue was dissolved in a small amount

of water/conc. aq. HBr (1/1, three times). The residue was dissolved

in conc. aq. HBr and product 8 was precipitated on addition of eth-

anol, collected by filtration, washed with ethanol and dried under

reduced pressure in a desiccator over P O to give 8·3HBr·2H O as

through a bed of silica gel and the product (H L1) was eluted with

2

H O/EtOH (1/1 v/v). The filtrate was concentrated to a small

2

volume in vacuo and the residue was purified on a strong cation-

+

exchange resin (Dowex 50, H form, 300 mL). Impurities were

2

5

2

eluted with water (500 mL) and the product was collected with 5%

aq. pyridine. The pyridine-containing fractions were combined and

evaporated to dryness to afford H L1·2.5H O as a yellow oil (1.00 g,

a pale brown solid (2.80 g, 77% based on 5). Elemental analysis

found (calcd for C H N O P ·3HBr·2H O, M =816.3): C 29.62

20

37

5

8

2

r

(29.43), H 5.62 (5.43), N 8.08 (8.58), P 7.59 (7.48), Br 27.39 (29.37);

2

+

ꢀ

4

0%).

ESI-MS (+ve): 538.2 ([M+H] , calcd 538.2), (ꢀve): 536.7 ([MꢀH] ,

1

calcd 536.2); H NMR (600.2 MHz, D O, pD=1.1): d=1.70–1.73 (m,

2

Elemental analysis found (calcd for C H N O P ·2.5H O, M =

20

39

5

4

2

r

4

H, CH CH CH ), 2.27 (s, 3H, NCH ), 2.53–2.92 (m, 16H, NCH CH N,

2 2 2 3 2 2

5

20.5): C 46.52 (46.15), H 8.82 (8.52), N 13.43 (13.45), P 12.18

2

ꢀ

CH CH CH ), 3.70 (d, 4H, NCH P, J_P_H=9.1 Hz) 3.81 (s, 2H, NCH₂

2 2 2 2

(

11.90); ESI-MS (ꢀve): 474.3 ([MꢀH] , calcd 474.5); MS spectra are

3

1

arom.), 7.61 (d, 2H, arom., J =7.8 Hz), 8.24 ppm (d, 2H, arom.,

HH

shown in Figure S17 (Supporting Information); H NMR (600.2 MHz,

3

13

1

J_H_H=7.8 Hz); C{ H} NMR (150.9 MHz, D O, pD=1.1): d=18.7, 19.8

2

D O, pD=7.5): d=1.94–2.82 (m, 4H, CH CH CH ), 2.86 (s, 3H,

2

(

s, each 1C, CH CH CH ); 43.1 (s, 1C, NCH ); 46.3, 47.6, 47.7, 47.8 (s,

2

3

NCH ), 2.99–3.58 (m, 16H, NCH CH N, CH CH CH ), 4.28 (dd, 2H,

3

2

1

each 1C, CH

2

CH

2

CH

2

1

); 49.2, 50.1, 51.0, 52.6 (s, each 1C, NCH

2

CH

54.3 (d, 2C, NCH P, JCP =142 Hz);

58.4 (s, 1C, NCH arom.); 123.6 (s, 2C, arom.); 130.7 (s, 2C, arom.);

2

N);

NCH arom.), 6.47 (d, 1H, NCH PO H, J =511.9 Hz), 6.89 (d, 2H,

2

PH

3

1

3

P, JCP =140 Hz); 54.7 (d, 2C, NCH

2 2

arom., J =8.4 Hz), 7.31 (d, 2H, arom., J =8.4 Hz), 7.32 ppm (d,

H, NCH PO H, J =511.9 Hz); C{ H} NMR (150.9 MHz, D O, pD=

2 2 PH 2

.5): d=22.8, 22.9 (s, each 1C, CH CH CH ); 41.8, 42.4 (s, each 1C,

CH CH CH ); 47.5 (d, 1C, NCH PO H, J_C_P=42 Hz); 47.8 (s, 1C,

HH

1

13

1

7

31

1

45.6 (s, 1C, arom.); 146.8 ppm (s, 1C, arom.); P{ H} NMR

2

1

(121.4 MHz, D O, pD=1.1): d=12.9 ppm (s).

2

1

4

1

-Methyl-11-(p-aminobenzyl)-1,4,8,11-tetraazacyclotetradecane-

,8-bis(methylenephosphonic acid) (H L2): Nickel(II) chloride

NCH ); 50.5 (d, 1C, NCH PO H, J =42 Hz); 52.3, 53.5, (s, each 1C,

3

2

CP

CH CH CH ); 54.2, 54.6, 55.9, 56.6 (s, each 1C, NCH CH N); 57.0 (s,

4

2

hexahydrate (66 mg, 0.28 mmol, 0.1 equiv) was dissolved in water

10 mL). After few minutes, sodium borohydride (88 mg, 0.9 mmol,

.0 equiv) was added. A solution of compound 8 (2.00 g, 2.5 mmol)

1

7

C, NCH arom.); 117.1 (s, 2C, arom.); 118.5 (s, 2C, arom.); 133.9 (s,

2

31

(

1

C, arom.); 148.9 (s, 1C, arom.); P NMR (121.4 MHz, D O, pD=

2

1

.5): d=22.4 (dt, J_P_H=552 Hz); 26.6 ppm (dt, J =552 Hz); NMR

PH

in water/ethanol (1/1, 40 mL) was alkalised to pHꢃ8 with NaOH

and added to the suspension of nickel metal. After 2 h, an excess

of sodium borohydride (0.46 g, 5.0 equiv) was added and the reac-

tion mixture was stirred at room temperature under an argon at-

mosphere for 24 h. The reaction mixture was vacuum-filtered

spectra are shown in Figures S18–S20 (Supporting Information).

4-Methyl-11-(p-nitrobenzyl)-1,4,8,11-tetraazacyclotetradecane-

1,8-bis(methylenephosphonic acid diethyl ester) (7): Compound

5 (1.95 g, 5.5 mmol) and triethyl phosphite (10.00 g, 60.2 mmol,

11.0 equiv) were mixed and the mixture was heated to 658C with

through a bed of silica gel and the product was eluted with H O/

2

stirring. Solid paraformaldehyde (0.80 g, 3.2 equiv) was added in

small portions over 60 min. The mixture was heated at this temper-

ature for 5 d in a closed flask. The excess paraformaldehyde was fil-

tered off and the reaction mixture was purified on a strong cation

EtOH (1/1 v/v). The eluate was concentrated to a small volume

under reduced pressure, and the residue was purified on a strong

+

cation-exchange resin (Dowex 50, H form, 300 mL). Impurities

were eluted with water (500 mL) and the product was collected

with 5% aq. pyridine (500 mL). The pyridine-containing fractions

were evaporated to dryness to afford H L2·3H O as a yellow oil

+

exchange resin (Dowex 50, H form, 300 mL). The excess of triethyl

phosphite was eluted with water/ethanol (1/1 v/v, 1000 mL) and

4

2

the product was collected by using EtOH/conc. aq. NH (5/1) as the

3

(900 mg,

65%).

Elemental analysis found (calcd

for

eluent (500 mL). The fractions containing pure product were com-

C H N O P ·3H O, M =561.6): C 42.74 (42.78), H 8.08 (8.08), N

2

0

39

5

6

2

r

bined and the solvent was evaporated to dryness to afford 7 as

+

1

5

2.58 (12.47), P 11.57 (11.03); EIS-MS (ꢀve): 506.2 ([MꢀH] , calcd

+

a yellow oil (2.94 g). ESI MS (+): 651.0 ([M+H] , calcd 651.3);

06.2); MS spectra are shown in Figure S21 (Supporting Informa-

1

H NMR (600.2 MHz, [D ]DMSO): d=1.13–1.25 ppm (m, 12H,

6

1

tion); H NMR (600.2 MHz, D O, pD=5.3): d=1.75–2.50 (m, 4H,

2

POCH CH ), 1.53–1.57 (m, 4H, CH CH CH ), 2.14 (s, 3H, NCH ), 2.32–

2

3

2

3

CH CH CH ), 2.60 (s, 3H, NCH ), 2.67–3.30 (m, 16H, NCH CH N,

2

3

2

.68 (m, 16H, CH CH CH , NCH CH N) 2.75–2.84 (m, 4H, NCH P),

2 2 2 2 2 2

CH CH CH ), 3.44 (s, 2H, NCH arom.), 4.10 (s, 2H, NCH PO H ), 4.17

2

3

2

3

.63 (s, 2H, NCH arom.), 3.89–4.02 (m, 8H, POCH CH ), 7.62 (d, 2H,

2

3

(

2

s, 2H, NCH PO H ), 6.80 (d, 2H, arom., J =7.8 Hz), 7.21 ppm (d,

H, arom., J =8.4 Hz); C{ H} NMR (150.9 MHz, D O, pD=5.3):

HH 2

3

13

1

2 3 2 HH

arom., J =8.4 Hz), 8.17 ppm (d, 2H, arom., J =8.4 Hz); C{ H}

NMR (150.9 MHz, [D ]DMSO): d=16.8 (d, 4C, POCH CH , J_C_P

HH

3

13

1

3

=

6

2

3

d=22.2, 22.3 (s, each 1C, CH CH CH ); 41.1 (s, 1C, NCH ); 47.0,

2

3

6

Hz); 24.0, 24.3 (s, each 1C, CH CH CH ); 43.1 (s, 1C, NCH ); 49.5

2 2 2 3

4

7.1, 47.3, 49.5 (s, each 1C, CH CH CH ); 49.8 (d, 1C, NCH PO H ,

2 2 2 2 3 2

1

(

d, 1C, NCH P, J =59 Hz); 50.6 (d, 1C, NCH P, J =60 Hz); 51.0,

2 CP 2 CP

1

J_C_P=88 Hz); 50.8 (d, 1C, NCH PO H , J =97 Hz); 50.3, 54.0, 54.8,

2

3

2

CP

5

1.7, 52.1, 52.2 (s, each 1C, CH CH CH ); 52.4, 52.5, 54.6, 55.1 (s,

2 2 2

5

5.3 (s, each 1C, NCH CH N); 56.2 (s, 1C, NCH arom.); 116.8 (s, 2C,

2 2 2

each 1C, NCH CH N); 57.8 (s, 1C, NCH arom.); 61.6 (d, 4C,

2

arom.); 118.7 (s, 2C, arom.); 133.1 (s, 1C, arom.); 146.8 ppm (s, 1C,

2

POCH CH , J =10 Hz); 123.5 (s, 2C, arom.); 130.2 (s, 2C, arom.);

46.8 (s, 1C, arom.); 149.2 ppm (s, 1C, arom.); P{ H} NMR

2

3

CP

31

arom.); P NMR (121.4 MHz, D O, pD=5.3): d=20.9 (s), 21.2 ppm

31

1

2

1

(s); NMR spectra are shown in Figures S22–S24 (Supporting Infor-

(

121.4 MHz, [D ]DMSO): d=21.4 (s), 22.6 ppm (s).

6

mation).

4

1

-Methyl-11-(p-nitrobenzyl)-1,4,8,11-tetraazacyclotetradecane-

,8-bis(methylenephosphonic acid) (8): Tetraester 7 (2.92 g) was

X-ray diffraction

dissolved in anhydrous acetonitrile (100 mL) and the flask was

wrapped in aluminium foil. Trimethylsilyl bromide (13.70 g,

Single crystals of bis-quaternary salt 4·EtOH were prepared by dif-

fusion of ethanol vapour into an aqueous solution of 4 (saturated

in hot water). Single crystals of 1,8-bis(p-nitrobenzyl)cyclam and

9

0.0 mmol, 20.0 equiv) was added dropwise over 60 min and the

reaction mixture was stirred in the dark at room temperature for

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4683

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

compound 5 were obtained by slow evaporation of ethanolic solu-

tions of the corresponding compounds. Selected crystals were

mounted on a glass fibre in random orientation and the diffraction

data were acquired at 150(1) K (Cryostream Cooler, Oxford Cryosys-

tems) with Mo_K_aradiation (l=0.71073 ). In the cases of 1,8-bis(p-

nitrobenzyl)cyclam and 4·EtOH, data were collected on a Nonius

Kappa CCD diffractometer (Enraf-Nonius) and analysed with the

Table 10. Selected crystallographic data.

Parameter

1,8-bis(p-nitroben-

zyl)-

4·EtOH

5

cyclam

formula

C

24

H

34

N

6

O

4

C

22

H

37

I

2

N

5

O

3

C

18 31 5 2

H N O

M

r

470.57

673.37

yellow prism yellow,

irregular

orthorhombic

P2

349.48

[80]

HKL DENZO program package. The structures were solved by

direct methods and refined by full-matrix least-squares techniques

colour, habit

yellow prism

[81]

[82]

by using SIR92 and SHELXL97 programs. The diffraction data

of compound 5 were collected on an ApexII CCD diffractometer

and analysed by using the SAINT V8.27B (Bruker AXS Inc., 2012)

program package. The structure was solved by direct methods

crystal system

space group

a []

b []

c []

triclinic

ꢀ

P1

monoclinic

P2 /c

1

1 1 1

2 2

7.4190(3)

7.7962(3)

21.2206(6)

93.533(2)

90.674(2)

99.092(2)

1209.39(8)

2

12.2995(4)

19.2331(6)

10.9595(3)

90

96.9690(10)

90

2573.40(14)

4

1.738

9.0984(7)

9.7822(6)

21.9209(15)

90

1951.0(2)

4

1.190

0.080

3819

[

83]

(

SHELXS97 ) and refined by full-matrix least-squares techniques

[82]

a [8]

b [8]

SHELXL97 ). For compound 4·EtOH, absorption correction by

[84]

Gaussian integration was applied. In general, all non-hydrogen

atoms in positions with full occupancy were refined anisotropically.

Although hydrogen atoms could usually be found in the electron

difference map, they were fixed in original (those bound to nitro-

gen atoms) or theoretical (those belonging to carbon atoms) posi-

tions by using a riding model with U (H)=1.2U (X) to keep the

g [8]

3

V [ ]

Z

ꢀ3

1_c_a_l_c_d[gcm

m [mm

]

1.292

0.090

ꢀ

1

]

2.476

5910

5304

unique reflns

obsd reflns

5482

3957

eq

3230

number of refined parameters low.

[

I>2s(I)]

In the case of 1,8-bis(p-nitrobenzyl)cyclam, two centrosymmetric

molecules are present in the unit cell, that is, two half-molecules

form the independent unit. For one of them, a disorder of the

macrocyclic part was found. It was best modelled with two possi-

ble arrangements of the macrocycle with relative occupancy 90:10

R; R’ [I>2s(I)]

0.0649; 0.0425

0.1208; 0.1109

0.0320;

0.0788;

0.0655

0.1366;

0.1314

0.0273

wR; wR’ [I>2s(I)]

0.0585;

0.0569

(

Supporting Information, Figure S1). Disordered parts of the macro-

cycle in both positions share the same (undisordered) nitrobenzyl

side group. Atoms belonging to part of the macrocycle with

higher occupancy were refined in anisotropic mode, and those

with lower occupancy were refined isotropically. For compound

II

weeks (Ni –H L1/H L2 systems) to establish full equilibrium. Then,

2

4

the potential at each titration point (tube) was determined with

a freshly calibrated electrode. For the Zn –H L2 system, the equilib-

rium was established relatively quickly at each titration point

II

4

·EtOH, the independent unit consists of one formula unit, that is,

II

(

ꢄ4 min), and thus Zn –H L2 was studied by conventional titra-

4

one bis-quaternary cation, two iodide anions and an ethanol mole-

cule of solvation. The benzyl group was found be twisted in two

positions sharing the same pivot methylene group. Both possibili-

ties have approximately the same occupancy (refined ratio 52:48),

and the corresponding disordered atoms were refined isotropically

tions. The titrations were carried out in a thermostatted vessel at

2

5.0ꢅ0.18C at a constant ionic strength of I(KNO )=0.1m. The

3

concentration of ligands in the titration vessel was about 0.004m.

The metal:ligand ratio was 1:1 in all cases. The initial volume was

about 1 cm for out-of-cell titrations and about 5 cm for the con-

ventional ones. The measurements were taken with an excess of

HNO added to the initial mixture. The mixtures were titrated with

stock KOH solution in the pH range 1.7–7.1 for out-of-cell

3

(

Supporting Information Figure S2). The ethanol molecule of solva-

tion was located in the electron density map, but it was heavily

3

disordered and the irregularity could not be sufficiently modelled

1

(

the presence of ethanol was confirmed by H NMR spectroscopy).

(

25 points per each titration) and 1.8–11.9 for conventional

40 data points per titration) titrations. Titrations for each system

[85]

Therefore, the ethanol molecule was squeezed by using PLATON.

The number of squeezed electrons corresponded well with the

electron count of one ethanol molecule. For compound 5, one

molecule forms the independent unit. A part of the macrocycle

was disordered and best fitted with relative occupancy 80:20 (Sup-

porting Information, Figure S3). Atoms in positions with higher oc-

cupancy were refined anisotropically, and those with lower occu-

pancy were refined isotropically.

were carried out four times. An inert atmosphere was ensured by

a constant flow of argon saturated with water vapour. The con-

stants with their standard deviations were calculated by using the

[86]

OPIUM program package and are concentration constants de-

h

fined as b =[H L]/{[H] ·[L]}; they can be transformed into the pro-

h

tonation constants logK_h(logK =logb₁and in general

1

logK (H L)=logb ꢀlogb ). Note that logK =pK of correspond-

h

(hꢀ1)

h

A

Selected experimental data are listed in Table 10. CCDC 1019313

ing protonated species. The concentration stability constants b

are defined as b_h_l_m=[H LM ]/{[H] ·[L]·[M] }. The ionic product of

h l m

hlm

h

l

m

(

1,8-bis(p-nitrobenzyl)cyclam), 1019312 (4·EtOH) and 1019311 (5)

contain the supplementary crystallographic data for this paper.

These data can be obtained free of charge from The Cambridge

Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_

request/cif.

water used in the calculations was pK =13.78.

w

Kinetic measurements

In general, experiments were run according to the published pro-

[26,35]

cedures.

The experiments on the kinetics of formation were

Potentiometry

carried out under pseudo-first-order conditions with fiftyfold

In general, the experiments followed a procedure published else-

excess of CuCl (c =5 mm, c =116 mm for H L1, c =100 mm for

H L2). The acidity/basicity was controlled by chloroacetate, acetate,

4

2

Cu

L

2

L

[32]

II

where. The equilibria in M –H L1/H L2 systems were established

2

4

slowly, and therefore an out-of-cell technique was used with wait-

phosphate (conventional UV/Vis experiments) or MES (stopped-

flow experiments) buffers (0.01m). Influence of the buffer anions

II

ing times of 5 d (Cu –H L1/H L2 and Zn –H L1 systems) and six

2

4

2

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4684

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

on the complexation reaction was checked by measurement of k_o_b_s

at various buffer concentrations (range 0.006–0.03m) at a constant

pH of 3.18 (Supporting Information, Figure S8). All experiments

were carried out at 25.0ꢅ0.18C and I=0.1m (KCl). A conventional

diode-array spectrophotometer (Shimadzu UV-2401PC) was used

for measurements in the pH range in which the reaction was slow

(0.9 mL). All following reactions were carried out at ꢀ5 to 08C. The

solution of sodium nitrite was added to the solution of a ligand.

After diazotisation for 10 min, the solution of urea was added, the

solution was left to stand for 5 min and then added to the stirred

suspension of the resorcinol-modified polymer in aq. sodium car-

bonate. After 2 h, the modified polymer was centrifuged off,

washed with water (150 mL) and methanol (50 mL) and dried in

air. Typical yield was 200 mg.

II

(

pH 2.4–5.0 for Cu –H L1 and pH 2.4–4.1 for Cu –H L2). A stopped-

2 4

flow spectrometer (Applied Photophysics SX.18MV) was used at

II

the higher pH (pH 5.2–7.3 for Cu –H L1 and pH 4.5–7.0 for Cu –

2

resin-H L1 (GMA-phosphinate): Elemental analysis:

P

0.40%

2

H L2). The complexation reaction was monitored directly by means

ꢀ1

4

(129 mmolg , corresponds to 65 mmol of the ligand per 1 g of

of the increase of the CT band (l=310 nm) of the Cu–L com-

13

1

resin); C{ H} CP-MAS NMR (125.8 MHz): d=20.0 (aliph.), 46.8

f

plexes. The k value was obtained by fitting the data with the

obs

(aliph.), 53.2 (aliph.), 66.0 (aliph.), 104.5 (arom.), 130.5 (arom.), 160.0

equation A(t)=A [1ꢀexp(ꢀk t)]. An example of the data fitting

31

1

obs

(arom.), 178.0 (C=O), P{ H} CP-MAS NMR (202.4 MHz): d=21 ppm

br).

is shown in Figure S25 of the Supporting Information.

(

+

The dissociation kinetics were studied in the [H ] range of 0.025–

resin-H L2 (GMA-phosphonate): Elemental analysis: P: 0.27%

4

.89m at I=5.0m (H,Na)ClO , with a starting concentrations of

ꢀ1

4

(87 mmolg , corresponds to 44 mmol of the ligand per 1 g of the

II

Cu –H L1 and Cu –H L2 complexes of 7.4 and 5.9 mm, respectively

13

1

2

4

resin); C{ H} CP-MAS NMR (125.8 MHz): d=18.5 (aliph.), 46.0

in the temperature range 25–558C. The reactions were monitored

by means of the decrease in intensity of the CT band (l=310 nm)

of the complexes with time. The kinetic measurements were car-

ried out with a conventional diode-array spectrophotometer (Shi-

madzu UV-2401PC, temperature-controlled by Peltier thermoelec-

(

1

aliph.), 54.5 (aliph.), 65.5 (aliph.), 103.0 (arom.), 129.7 (arom.), 160.0

arom.), 177.5 ppm (C=O); P{ H} CP-MAS NMR (202.4 MHz): d=

8 ppm (br).

31

1

II

d

Separation of copper(II) from Ni /Zn

tric heating to (25–55)ꢅ0.18C). The k_o_b_svalues were calculated

from the experimental data by using a single-exponential model

The chelating polymer resin (resin-H L1 or resin-H L2, 20 mg) was

2

4

A(t)=A exp(ꢀk_o_b_st). An example of data fitting is shown in Fig-

0

added to a mixture of stock solutions of the metal ions (0.700 mL;

see below for composition) and buffer (0.700 mL, pH 3.0 or 5.0; see

below for composition). The suspension was shaken for 1 h at

room temperature and centrifuged. The supernatant was analysed

for metal content by AAS. The pellet was redispersed in washing

buffer (pH 3.0, 1.4 mL, 1st wash; see below for composition), incu-

bated for 20 min at room temperature and centrifuged, and the

supernatant was taken for AAS. Washing with the buffer (pH 3.0,

ure S26 (Supporting Information). Fitting of pH and temperature

dependences of the formation and dissociation rate constants was

[87]

2

performed by using a MicroMath Scientist program with the 1/y

weighting scheme.

Formation experiments with simple ligands were also performed in

6

4

the presence of Cu. The ligands were radiolabelled by adding 1–

6

4

2

MBq of [ Cu]CuCl to 10 mg of H L1, H L2, cyclam and N-benzyl-

2 2 4

cyclam dissolved in 100 mL of 0.05m MES/NaOH buffer (pH 5.5).

The radiolabelling yield as function of time was determined by

radio-TLC. Radio-TL chromatograms were scanned with a radioiso-

1

.4 mL) was repeated once more (2nd wash, the supernatant was

taken for AAS analysis of metal contents). Desorption of the chelat-

ed copper was performed by incubation of the loaded polymer

with aq. HCl (1m, 1.4 mL) for 20 min at room temperature followed

by centrifugation (the supernatant was taken for AAS analysis of

metal contents, 1st desorption). The desorption step was repeated

once more (2nd desorption).

tope thin-layer analyzer (Rita Star, Raytest). For H L1, H L2: silica

2

4

gel plates (Merck, F254) as stationary phase, 1m aq. triethylammo-

64

nium hydrogencarbonate/methanol (3/1, v/v) as eluent;^C^u^C^l2^,

6

4

64

^Rf ⁼⁰^; Cu-L1, R =0.6; Cu-L2, R =0.65. Cyclam, N-benzylcyclam:

f

iTLC-SA strips (Agilent) as stationary phase, 0.1m aq. ethylenedia-

The Ni+Cu stock solution was prepared by dissolution of nickel(II)

chloride hexahydrate (810 mg, 3.4 mmol) and copper(II) sulfate

pentahydrate (1.6 mg, 0.0064 mmol) in water, and the total volume

was adjusted with water to 10.0 mL (c =0.04 mgmL , c =

2

tion of zinc(II) chloride (464 mg, 3.4 mmol) and copper(II) sulfate

pentahydrate (1.6 mg, 0.0064 mmol) in water, and the total volume

was adjusted with water to 10.0 mL (cCu =0.04 mgmL , cZn

22 mgmL ). The chloroacetate buffer stock solution (pH 3.0) was

prepared by dissolution of chloroacetic acid (8.03 g, 85 mmol) in

water (ca 25 mL), titration with aq. NaOH (50 wt.%) to pH 3.0 and

adjustment of total volume to 50.0 mL (cbuffer =1.7m). The acetate

buffer stock solution (pH 5.0) was prepared by dissolution of acetic

acid (5.10 g, 85 mmol) in water (ca 25 mL), titration with aq. NaOH

(50 wt.%) to pH 5.0 and adjustment of total volume to 50.0 mL

(cbuffer =1.7m). The wash buffer (pH 3.0) was prepared by dilution

of the chloroacetate buffer stock solution having pH 3.0 with water

(1/1 v/v).

64

minetetraacetic acid (EDTA) as mobile phase; free Cu as Cu–

6

4

64

EDTA, R =0.9–1.0; Cu–cyclam and Cu–N-benzylcyclam, R =0).

f

ꢀ1

Cu

Ni

ꢀ1

0 mgmL ). The Zn+Cu stock solution was prepared by dissolu-

Synthesis of polymer-supported macrocyclic ligands

Poly(glycidyl methacrylate-co-ethylene dimethacrylate) macropo-

rous beads (60:40 w/w, 20–40 mm, 8.0 g, 32 mmol epoxide

ꢀ

1

=

[75]

ꢀ1

groups)

in solution of 1,3-benzenediol (resorcinol, 16.0 g,

1

45 mmol) and sodium hydroxide (2.9 g, 73 mmol) in N,N-dimethy-

lacetamide (30 mL) were heated at 1008C with stirring by anchor

stirrer for 5 h. The resorcinol-modified polymer (GMA-resorcinol)

was washed subsequently with methanol (250 mL), 5% aq. acetic

acid (200 mL) and methanol (300 mL) and dried in air (yield 8.2 g).

13

1

Elemental analysis found: C 56.35, H 7.96, N 1.32; C{ H} CP-MAS

NMR (125.8 MHz): d=19.8 (aliph.), 45.5 (aliph.), 56.5 (aliph.), 63.0

(

aliph.), 104.5 (arom.), 130.5 (arom.), 159.9 (arom.), 179.0 ppm (C=

13

O). The content of resorcinol was calculated on the basis of

solid-state NMR data and was found to be 0.974 mmolg

C

ꢀ1

.

The polymer with immobilised resorcinol (200 mg) was added to

solution of sodium carbonate (2.7 g, 25 mmol) in water (60 mL).

Ligand H L1 or H L2 (80 mg, ꢃ0.15 mmol) was dissolved in aq. HCl

Separation of radiocopper(II) from nickel

Intensity of radioactive decay was determined by using an Isomed

2000 activimeter (MED, Nuclear-Medizintechnik GmbH, Dresden,

Germany). ICP-MS was performed on a Sektorfeld-ICP-Mass spec-

2

4

(

6m, 2.0 mL). Sodium nitrite (11.4 mg, 0.17 mmol) was dissolved in

water (0.9 mL). Urea (27 mg, 0.45 mmol) was dissolved in water

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org 4685

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

trometer Element2 (ThermoFisher Scientific) after complete decay

of radioactivity. Aq. HCl (30%, Suprapur) was purchased from

Merck-Millipore, and MES from Applichem. Each step was per-

formed under metal-free conditions. Double-distilled water was

used for dilution of HCl (30%) as well as for preparation of the

MES buffer solution.

[8] P. Hermann, J. Kotek, V. Kubí cˇ ek, I. Luke sˇ , Dalton Trans. 2008 , 3027–

3047.

[

9] C. J. Anderson, M. J. Welch, Chem. Rev. 1999 , 99 , 2219–2234.

[

10] W. A. Volkert, T. J. Hoffmann, Chem. Rev. 1999 , 99 , 2269–2292.

11] S. Liu, Chem. Soc. Rev. 2004 , 33 , 445–461.

12] S. Liu, D. S. Edwards, Bioconjugate Chem. 2001 , 12 , 7–34.

13] M. J. Welch, Handbook of Radiopharmaceuticals: Radiochemistry and Ap-

plications (Eds.: C. S. Redvanly), Wiley, Hoboken, 2003.

6

4

Starting [ Cu]CuCl solution in 0.01m aq. HCl was prepared as de-

2

[78]

scribed previously. To 2.59 mL of this solution, a defined amount

[14] W. P. Li, L. A. Meyer, C. J. Anderson, Top. Curr. Chem. 2005, 252, 179–

of nickel was added (10 mL of the stock solution of NiSO contain-

192.

4

ꢀ

1

ing 2.6 mgmL of Ni), resulting in a final radiocopper solution

[15] T. J. Wadas, E. H. Wong, G. R. Weisman, C. J. Anderson, Chem. Rev. 2010 ,

110 , 2858–2902.

16] B. M. Zeglis, J. S. Lewis, Dalton Trans. 2011 , 40 , 6168–6195.

17] C. S. Cutler, H. M. Hemmkens, N. Sisay, S. Huclier-Markai, S. S. Jurisson,

Chem. Rev. 2013 , 113 , 858–883.

18] E. W. Price, C. Orvig, Chem. Soc. Rev. 2014 , 43 , 260–290.

19] S. V. Smith, J. Inorg. Biochem. 2004 , 98 , 1874–1901.

[20] X. Sun, C. J. Anderson, Methods Enzymol. 2004 , 386 , 237–261.

[21] T. J. Wadas, E. H. Wong, G. R. Weisman, C. J. Anderson, Curr. Pharm. Des.

2007, 13, 3–16.

ꢀ1

with Ni concentration of 10000 mgL , which was used in the sorp-

[

tion experiments.

The resins were pretreated by the following procedure: 1 mg of

resin was incubated in aq. HCl (1m) for 24 h. After separation of

the resin by centrifugation, the content of Ni/Cu/Zn in the super-

natant was determined by ICP-MC to determine threshold/back-

ground level. Subsequently, the resin was washed three times with

MES buffer (0.1m, pH 5.5).

[

6

4

[22] S. Thieme, M. Walther, S. Preusche, J. Rajander, H.-J. Pietzsch, O. Solin, J.

Steinbach, Appl. Radiat. Isot. 2013 , 72 , 169–176.

Adsorption of Cu was studied by following procedure: 20 mL of

[

6

4

Cu]CuCl solution (the starting activity was 51.5 and 52.2 MBq for

2

resin-H L1 and resin-H L2, respectively) was added to MES buffer

980 mL, 0.1m, pH 5.5). These solutions were added to 1 mg of

resin (pretreated with HCl). After shaking at room temperature for

0 min and subsequent centrifugation, the supernatant was re-

[

23] a) M. Sadeghi, M. Amiri, P. Rowshanfarzad, M. Avila, C. Tenreiro, Radio-

chim. Acta 2008, 96, 399–402; b) R. Schwarzbach, K. Zimmermann, P.

Bläuenstein, A. Smith, P. A. Schubiger, Appl. Radiat. Isot. 1995 , 46 , 329–

2

4

(

336; c) S. V. Smith, D. J. Waters, N. DiBartolo, Radiochim. Acta 1996, 75,

3

65–68.

moved and its radioactivity was determined. In addition, the resin

was further washed three times with MES buffer (1 mL, 0.1m,

pH 5.5).

[24] C. Dirks, B. Scholten, S. Happel, A. Zulauf, A. Bombard, H. Jungclas, J.

Radioanal. Nucl. Chem. 2010 , 286 , 671–674.

[

25] J. Kotek, P. Vojtí sˇ ek, I. Císa rˇ ovµ, P. Hermann, P. Jure cˇ ka, J. Rohovec, I.

Luke sˇ , Collect. Czech. Chem. Commun. 2000 , 65 , 1289–1316.

[

26] J. Kotek, P. Lubal, P. Hermann, I. Císa rˇ ovµ, I. Luke sˇ , T. Godula, I. Svobo-

dovµ, P. Tµborsk y´ , J. Havel, Chem. Eur. J. 2003 , 9 , 233–248.

[

6

4

Desorption of Cu was performed in the following way: 1m aq.

HCl (1 mL) was added to the loaded resin. After shaking at room

temperature for 15 min and subsequent centrifugation, the super-

natant was removed and the resin was washed three times with

27] I. Svobodovµ, P. Lubal, P. Hermann, J. Kotek, J. Havel, J. Inclusion

Phenom. Macrocyclic Chem. 2004, 49, 11–15.

1

m aq. HCl (1 mL). Supernatants were further analysed to deter-

[28] I. Svobodovµ, P. Lubal, P. Hermann, J. Kotek, J. Havel, Microchim. Acta

2004, 148, 21–26.

mine radioactivity level and by ICP-MS.

[

29] P. Lubal, J. Male cˇ ek, P. Hermann, J. Kotek, J. Havel, Polyhedron 2006 , 25 ,

884–1892.

30] S. Füzerovµ, J. Kotek, I. Císa rˇ ovµ, P. Hermann, K. Binnemans, I. Luke sˇ ,

Dalton Trans. 2005 , 2908–2915.

31] T. Vitha, J. Kotek, J. Rudovsk y´ , V. Kubí cˇ ek, I. Císa rˇ ovµ, P. Hermann, I.

Luke sˇ , Collect. Czech. Chem. Commun. 2006 , 71 , 337–367.

32] I. Svobodovµ, P. Lubal, J. Plutnar, J. Havlí cˇ kovµ, J. Kotek, P. Hermann, I.

Luke sˇ , Dalton Trans. 2006, 5184–5197.

33] J. Plutnar, J. Havlí cˇ kovµ, J. Kotek, P. Hermann, I. Luke sˇ , New J. Chem.

2008, 32, 496–504.

1

Acknowledgements

We thank Dr. K. Lang (Institute of Inorganic Chemistry, Acade-

my of Science of the Czech Republic) for the use of a stopped-

flow apparatus, Ing. O. Policianova (Institute of Macromolecular

Chemistry, Academy of Science of the Czech Republic) for

measuring of solid-state NMR spectra, and Dr. M. Kubeil, C.

Jentschel and R. A. Nadar (Helmholtz-Zentrum Dresden-Ros-

sendorf) for carrying out the radiochemical investigations. This

work was supported by the Grant Agency of the Czech Repub-

lic (grants P207/11/1437 and P304/12/0950).

34] J. Havlí cˇ kovµ, H. Medovµ, T. Vitha, J. Kotek, I. Císa rˇ ovµ, P. Hermann,

Dalton Trans. 2008, 5378–5386.

35] I. Svobodovµ, J. Havlí cˇ kovµ, J. Plutnar, P. Lubal, J. Kotek, P. Hermann,

Eur. J. Inorg. Chem. 2009 , 3577–3592.

[36] L. M. P. Lima, R. Delgado, J. Plutnar, P. Hermann, J. Kotek, Eur. J. Inorg.

Chem. 2011 , 527–538.

[

37] G. Royal, V. Dahaoui-Gindrey, S. Dahaoui, A. Tabard, R. Guilard, P. Pul-

lumbi, C. Lecomte, Eur. J. Org. Chem. 1998 , 1971–1975.

Keywords: copper · ion-selective resins · kinetics · macrocyclic

ligands · radiochemistry

[38] C. Bucher, G. Royal, J. M. Barbe, R. Guilard, Tetrahedron Lett. 1999 , 40 ,

315–2318.

2

[

39] J. Kotek, P. Hermann, P. Vojtí sˇ ek, J. Rohovec, I. Luke sˇ , Collect. Czech.

Chem. Commun. 2000 , 65 , 243–266.

[

40] E. H. Wong, G. R. Weisman, D. C. Hill, D. P. Reed, M. E. Rogers, J. S.

Condon, M. A. Fagan, J. C. Calabrese, K. C. Lam, I. A. Guzei, A. L. Rhein-

gold, J. Am. Chem. Soc. 2000 , 122 , 10561–10572.

[

1] M. Meyer, V. Dahaoui-Ginderey, C. Lecomte, R. Guilard, Coord. Chem.

Rev. 1998 , 178–180 , 1313–1405.

[

2] L. F. Lindoy, Adv. Inorg. Chem. 1998 , 45 , 75–125.

3] R. Delgado, V. FØlix, L. M. P. Lima, D. W. Price, Dalton Trans. 2007 , 2734–

[41] C. Bucher, E. Duval, J.-M. Barbe, J.-N. Verpeaux, C. Amatore, R. Guilard,

C. R. Acad. Sci. Paris, Ser. IIc: Chim. 2000, 3, 211–222.

2745.

[

4] The Chemistry of Contrast Agents in Medical Magnetic Resonance Imag-

ing, 2nd ed. (Eds.: A. Merbach, L. Helm, É. Tóth), Wiley, Hoboken, 2013.

5] Topics in Current Chemistry, Vol. 221 (Ed.: W. Krause), Springer, Heidel-

berg, 2002.

[42] H. Kurosaki, C. Bucher, E. Espinosa, J.-M. Barbe, R. Guilard, Inorg. Chim.

Acta 2001 , 322 , 145–149.

[43] R. Kolinski, Pol. J. Chem. 1995, 69 , 1039–1945.

[44] B. Bosnich, C. K. Poon, M. Tobe, Inorg. Chem. 1965 , 4 , 1102–1108.

[45] G. Anderegg, F. Arnaud-Neu, R. Delgado, J. Felcman, K. Popov, Pure

Appl. Chem. 2005, 77, 1445–1495.

[

6] S. Aime, M. Botta, E. Terreno, Adv. Inorg. Chem. 2005 , 57 , 173–237.

7] P. Caravan, Chem. Soc. Rev. 2006 , 35 , 512–523.

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4686

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

[

46] I. Luke sˇ , J. Kotek, P. Vojtí sˇ ek, P. Hermann, Coord. Chem. Rev. 2001, 216–

17, 287–312.

47] NIST Standard Reference Database 46 (Critically Selected Stability Con-

stants of Metal Complexes), Version 7.0, 2003.

[68] P. K. Tapaswi, M. S. Moorthy, S. S. Park, C.-S. Ha, J. Solid State Chem.

2014 , 211 , 191–199.

[69] Z. Q. Wang, M. Wang, G. H. Wu, D. Y. Wu, A. G. Wu, Dalton Trans. 2014 ,

43 , 8461–8468.

2

[

48] J. Blahut, I. Císa rˇ ovµ, P. Hermann, J. Kotek, unpublished results.

49] N. E. Good, G. D. Winget, W. Winter, T. N. Connolly, S. Izawa, R. M. M.

Singh, Biochemistry 1966 , 5 , 467–477.

50] A. Rçhrich, R. Bergmann, A. Kretzschmann, S. Noll, H.-J. Pietzsch, J.

Steinbach, H. Stephan, J. Inorg. Biochem. 2011 , 105 , 821–832.

51] R. Buxtorf, T. A. Kaden, Helv. Chim. Acta 1974 , 57 , 1035–1042.

52] M. Maeder, Y.-M. Neuhold, G. Puxty, P. King, Phys. Chem. Chem. Phys.

[70] M. R. Awual, I. M. M. Rahman, T. Yaita, M. A. Khaleque, M. Ferdows,

Chem. Eng. J. 2014 , 236 , 100–109.

[71] Y. H. Zhai, D. Yang, X. J. Chang, Y. W. Liu, Q. He, J. Sep. Sci. 2008 , 31 ,

1195–1200.

[72] M. Mureseanu, N. Cioatera, I. Trandafir, I. Georgescu, F. Fajula, A. Galar-

neau, Microporous Mesoporous Mater. 2011 , 146 , 141–150.

[73] S. Malkondu, A. Kocak, M. Yilmaz, J. Macromol. Sci. Pure Appl. Chem.

2009 , 46 , 745–750.

[

2

003 , 5 , 2836–2841.

53] A. P. Leugger, L. Hertli, T. A. Kaden, Helv. Chim. Acta 1978 , 61 , 2296–

306.

[

[74] S. Goubert-Renaudin, M. Etienne, Y. Rousselin, F. Denat, B. Lebeau, A.

2

Walcarius, Electroanalysis 2009, 21, 3–5.

ˇ

[

54] M. Kodama, E. Kimura, J. Chem. Soc. Dalton Trans. 1977 , 1473–1478.

55] a) D. J. Stigers, R. Ferdani, G. R. Weisman, E. H. Wong, C. J. Anderson,

J. A. Golen, C. Moore, A. L. Rheingold, Dalton Trans. 2010 , 39 , 1699–

[75] F. Svec, J. Hradil, J. Coupek, J. Kµlal, Angew. Makromol. Chem. 1975, 48,

135–143.

ˇ

[76] D. Horµk, J. Straka, J. Stokr, B. Schneider, T. B. Tennikova, F. Svec, Polymer

1991 , 32 , 1135–1139.

1701; b) R. Ferdani, D. J. Stigers, A. L. Fiamengo, L. Wei, B. T. Y. Li, J. A.

Golen, A. L. Rheingold, G. R. Weisman, E. H. Wong, C. J. Anderson,

Dalton Trans. 2012 , 41 , 1938–1950 ; c) Y. Guo, R. Ferdani, C. J. Anderson,

Bioconjugate Chem. 2012 , 23 , 1470–1477 ; d) M. Jiang, R. Ferdani, M.

Shokeen, C. J. Anderson, Nucl. Med. Biol. 2013 , 40 , 245–251 ; e) D. Zeng,

Q. Ouyang, Z. Cai, X.-Q. Xie, C. J. Anderson, Chem. Commun. 2014 , 50 ,

[77] M. Hrub y´ , J. Ku cˇ ka, M. Novµkovµ, H. Mackovµ, M. Vetrík, Appl. Radiat.

Isot. Appl. Radiat. Isotop. 2010, 68, 334–339.

[78] S. Thieme, M. Walther, H.-J. Pietzsch, J. Henniger, S. Preusche, P. Mäding,

J. Steinbach, Appl. Radiat. Isot. 2012 , 70 , 602–608.

[79] a) R. P rˇ ibil, Analytical Applications of EDTA and Related Compounds, Per-

gamon Press, Oxford, 1972; b) G. Schwarzenbach, H. Flaschka, Com-

plexometric Titrations, Methuen, London, 1969.

43–45.

[

56] L.-H. Chen, C.-S. Chung, Inorg. Chem. 1988 , 27 , 1880–1883.

57] Z. Y. A. Yu, R. A. Wu, M. H. Wu, L. Zhao, R. B. Li, H. F. Zou, Colloids Surf. B

[80] a) Z. Otwinowski, W. Minor, HKL Denzo and Scalepack Program Pack-

age; Nonius BV: Delft, The Netherlands, 1997; b) Z. Otwinovski, W.

Minor, Methods Enzymol. 1997, 276, 307–326.

2

010 , 78 , 222–228.

ˇ

[58] M. Skodovµ, J. Kucka, M. Vetrík, J. Skopal, Z. Walterovµ, O. Sedlµcek, P.

ˇ

ˇ ˇ

ˇ

´

Stepµnek, J. Mattovµ, P. Pouckovµ, P. Urbµnek, M. Hruby, React. Funct.

[81] A. Altomare, M. C. Burla, M. Camalli, G. Cascarano, C. Giacovazzo, A.

Guagliardi, G. Polidori, J. Appl. Crystallogr. 1994, 27, 435–435.

[82] G. M. Sheldrick, SHELXL97, Program for Crystal Structure Refinement

from Diffraction Data: University of Gçttingen, Gçttingen, 1997.

[83] G. M. Sheldrick, SHELXS97, Program for Crystal Structure Solution from

Diffraction Data: University of Gçttingen, Gçttingen, 1997.

Polym. 2003, 73, 1426–1431.

59] K. F. Murray, D. Lam, K. V. Kowdley, Exp. Opinion Pharmacother. 2003 , 4 ,

[

2239–2251.

60] G. Crisponi, V. M. Nurchi, D. Funni, C. Gerosa, S. Nemolato, G. Faa,

Coord. Chem. Rev. 2010 , 254 , 876–889.

61] L. Cai, X. K. Li, Y. Song, M. G. Cherian, Curr. Med. Chem. 2005, 12, 2753–

[84] P. Coppens, Crystallographic Computing (Eds.: F. R. Ahmed, S. R. Hall,

C. P. Huber), Munksgaard, Copenhagen, 1970, pp. 255–270.

[85] A. L. Spek, PLATON, A Multipurpose Crystallographic Tool, Utrecht Uni-

versity, Utrecht, 2005.

[86] a) M. K y´ vala, I. Luke sˇ , International Conference Chemometrics 1995 (Par-

dubice, Czech Republic) July 3–7, 1995, p. 63; b) M. K y´ vala, P. Lubal, I.

Luke sˇ , IX. Solution Equilibria Analysis with the OPIUM Computer Pro-

gram; Spanish-Italian and Mediterranean Congress on Thermodynamics

of Metal Complexes (SIMEC 98) (Girona, Spain), June 2–5, 1998. The

full version of the OPIUM program is available (free of charge) on

http://www.natur.cuni.cz/~kyvala/opium.html.

2

763.

62] P. Burke, O. Golovko, J. C. Clark, F. I. Aigbirhio, Inorg. Chim. Acta 2010 ,

63 , 1316–1319.

3

63] D. W. McCarthy, R. E. Shefer, R. E. Klinkowtein, L. A. Bass, W. H. Marge-

neau, C. S. Cutler, C. J. Anderson, M. J. Welch, Nucl. Med. Biol. 1997 , 24 ,

35–43.

[

64] C. M. Jeffery, S. V. Smith, A. H. Asad, S. Chan, R. I. Price, Routine Produc-

tion of Copper-64 Using 11.7 MeV Protons, Vol. 1509 (Eds.: M. A. AvilaRo-

driguez, J. P. ONeil, T. E. Barnhart, D. W. Dick, J. M. Koziorowski, S. E. Lapi,

J. S. Lewis), Amer Inst Physics, Melville, USA, 2012, pp. 84–90.

65] C. Alliot, N. Michel, A. C. Bonraisin, V. Bosse, J. Laize, C. Bourdeau, B. M.

Mokili, F. Haddad, Radiochim. Acta 2011 , 99 , 627–630.

[

[87] Scientist for Windows, Version 2.01, Micromath Inc., Salt Lake City, 1995.

66] M. R. Awual, M. Ismael, T. Yaita, S. A. El-Safty, H. Shiwaku, Y. Okamoto, S.

Suzuki, Chem. Eng. J. 2013 , 222 , 67–76.

67] B. Shah, T. Chakrabarky, V. K. Shahi, U. Chudasama, Sep. Sci. Technol.

Received: October 22, 2014

2013 , 48 , 2382–2390.

Published online on February 3, 2015

Chem. Eur. J. 2015, 21, 4671 – 4687

www.chemeurj.org

4687

ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Article Doi

Bifunctional cyclam-based ligands with phosphorus acid pendant moieties for radiocopper separation: Thermodynamic and kinetic

DOI: 10.1002/chem.201405777

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.201405777

Full text of DOI:10.1002/chem.201405777

Products guided by the article

R&D Labs maybe for 74199-16-9

Relevant to this article

Hot Product