Introduction

Macrocyclic amino-carboxylate ligands are widely used in clinical chemistry including medical diagnosis and therapeutic applications involving metal complexes. The ligands consisting of polyaza-macrocyclic rings and pendant arms show high selectivity if the size of the ring and the radius of the metal ion match perfectly. Moreover, the nature of the pendant arm attached to the N-atoms of the macrocycle can be used to tune the physico-chemical properties (stability and selectivity) of the metal complexes. In case of bifunctional ligands, there is also a conjugated biological vector (peptide, antibody etc.), i.e. a covalently bonded entity in the molecule, responsible for the selective targeting of the complexes in living organisms [1].

The most widely studied complexes are the MRI-contrast agents [1], where the macrocyclic ligands are 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetate (DOTA) and its derivatives, binding paramagnetic Gd(III) ions, while the 1,4,7-triazacyclononane-1,4,7-triacetate (NOTA) with a smaller cavity forms extremely stable complexes with smaller ions like Fe(III) or Ga(III) [2]. The interest in [Al(NOTA)] has recently renewed due to a new concept by D’Souza et al. [3–5], using [Al(NOTA)] as a 18F-isotope carrier in positron emission tomography (PET). The preparation of [Al(NOTA)(18F)]− is reported to be faster and simpler compared to the standard procedure for making 18FDG (18fluoro-deoxyglucose) used routinely in PET imaging. Targeted PET agents were developed recently using Al complexes formed with bifunctional NOTA-derivatives; very promising results (both in sensitivity and in resolution) have been reported in animals [3-5]. Radiochemists and medical researchers usually focus on radiochemical yields and in vivo stability, although one should know more about the chemical characteristics of the [Al(NOTA)] complex itself. In this paper, we deal with the physico-chemical characterization of [Al(NOTA)], by investigating and describing the complex stability constant and its dissociation in acidic and alkaline solutions using pH-potentiometry, 1H- and 27Al-NMR spectroscopy. Preliminary results on the identification of [Al(NOTA)(F)]− mixed complex by 19F-NMR are also reported.

Experimental section

Materials and solution preparation

The stock solutions of 0.2 M AlCl3, 0.1 M NaF and 0.03 M NOTA were prepared from solid AlCl3·6H2O, purchased from Reachim, NaF purchased from Sigma-Aldrich and solid NOTA, synthesized earlier in our group [6, 7].

Analysis

Concentrations of the AlCl3 and NOTA solutions were determined by complexometric titration [8] and pH-potentiometry (see below), respectively.

Sample preparation

Owing to the slow formation of macrocyclic complexes, the stability constant of [Al(NOTA)] was determined by the out-of-cell method. 15 samples containing 3 mM NOTA and 3 mM AlCl3 in 0.15 M NaCl were kept sealed at 37 °C in the pH-range 2–3, for 4 months. 3 separate samples of identical composition (at pH 2.0, 2.5 and 3.0) were followed by 1H- and 27Al-NMR to determine the time needed for complete equilibration at all pH values.

For investigating the kinetic properties of [Al(NOTA)] in both acidic and basic solutions, 1 mM [Al(NOTA)] in 1 M HCl and 0.05–1 M NaOH were used.

Samples used to determine the properties of [Al(NOTA)(OH)]− ternary complex using pH-potentiometric titrations contained 4 mM of [Al(NOTA)], 0.15 M of NaCl and 6.6 mM of HCl. Those used for investigating complex protonation contained 4 mM of [Al(NOTA)], 0.15 M of NaCl with starting pH set to 1.

The NMR samples were prepared by mixing Al(III), NOTA and F– in equimolar ratio (4 mM of AlCl3, NaF and NOTA each, in acetate buffer at pH = 4) and heated at 100 °C for various time periods following the literature procedure of Dijkgraaf et al. [5]. Their 19F-NMR spectra were recorded after an additional 24 h. The pH was adjusted by stepwise addition of 0.2 M NaOH and HCl solutions. An attempt to prepare [Al(NOTA)(F)]– was also undertaken by using solvent mixture (1:1 water:ethanol) as described by D’Souza et al. [3] for the preparation of single crystals using stable Na19F (AlCl3, NOTA and F– solutions were mixed in water and the appropriate amount of ethanol added lastly).

NMR experiments

400 MHz 1H-, 376.5 MHz 19F- and 104.3 MHz 27Al-NMR spectra were collected using a Bruker DRX 400 (9.4 T) NMR spectrometer. Calibration was performed with the signal of TMS for 1H-spectra (0 ppm), 10 mM slightly basic NaF for 19F-spectra (0 ppm) and that of 10 mM AlCl3 for 27Al-spectra (0 ppm). The temperature was kept at 25 ± 0.1 °C with a Bruker VT-1000 thermo controller. As all samples were made in H2O, closed capillaries of D2O were inserted into each sample and used for deuterium lock. The spectra were analyzed with the Bruker WinNMR software package.

Potentiometry

Titrations were carried out with a Metrohm 702SM Titrino titration workstation with the use of a Metrohm-6.0233.100 combined electrode. The titrated solutions (6.00 mL in total) were thermostated at 25 °C. The samples were stirred with a magnetic stirrer and N2 gas was bubbled through the solutions before and during the procedure to avoid any interaction with air. The titrations were performed in the pH-range of 2.2–12.05. The pH-electrode was calibrated with KH-phthalate (pH 4.005) and borax (pH 9.177) buffers. [H+] concentrations were calculated from the measured pH values with the method proposed by Irving et al. [9]. Therefore, the equilibrium constants in this paper are stoichiometric constants. A 0.02 M HCl solution (in 0.15 M NaCl) was titrated with 0.2 M NaOH and the difference between the measured and calculated pH values were used to calculate the [H+] from the pH values determined in the titration experiments. Equilibrium constants were calculated using the PSEQUAD program [10].

Results and discussion

Stability of [Al(NOTA)]

In order to characterize the Al(III)-NOTA system, first the protonation constants of NOTA3− were measured using standard pH-potentiometric technique in 0.15 M NaCl (to mimic the conditions present in biological fluids). The ligand protonation constants could be defined as follows (Eq. 1):

$${\text{L}}_{{{\text{i}} - 1}} + {\text{H}}^{ + } \rightleftharpoons {\text{H}}_{{\text{i}}} {\text{L}}$$
(1)
$$K_{{\text{i}}}^{{\text{H}}} = \frac{{[{\text{H}}_{{\text{i}}} {\text{L}}]}}{{[{\text{H}}_{{{\text{i}} - 1}} {\text{L}}][{\text{H}}^{ + } ]}}{{i = 1,2,3,4}}$$
(2)

The protonation constants of NOTA were calculated using the titration data and are shown in Table 1. These values are in a good agreement with the data obtained by Martell [2]. To the best of our knowledge, the stability constant of [Al(NOTA)] has not been determined yet. The formation equilibrium between NOTA3− and Al(III)ion could not be studied by means of direct pH-potentiometry, because of the slow formation reaction of the complex in acidic medium. Hence, the so-called “out-of-cell” technique was used to determine the stability constant.

Using an estimated value for the stability constant of [Al(NOTA)] (lgK ~ 17), 15 separate samples were prepared within the pH-range of 2–3, changing the quantity of added strong acid. Having no information about the formation rate of [Al(NOTA)], three NMR samples (being in the pH range of 2–3) were also prepared to follow the process in time by using 1H- and 27Al-NMR techniques (Fig. 1). NMR spectra were recorded several times until no further changes could be observed. The time required for this was 4 months, after which the 15 “out of cell” samples were opened and their equilibrium pH values were measured. These data were used for equilibrium calculations with PSEQUAD.

Table 1 Protonation constants of the NOTA ligand and equilibrium constants of [Al(NOTA)] (25 °C, 0.15 M NaCl)
Full size table
Fig. 1
figure 1

104.3 MHz 27Al-NMR spectra of the Al(III)-NOTA system (cAl(III) = 2.88 mM, cH3NOTA = 2.99 mM, final pH = 2.79) as function of time

Full size image

Typical spectral changes can be seen in Fig. 1. The signal intensity of free Al(III) at 0 ppm ion decreased, while that of [Al(NOTA)] at 49 ppm increased with time. No further changes were observed after 4 weeks. Equilibration time was longer at lower pH-values. Therefore, the samples were aged for 4 months before opening and measuring their equilibrium pH-values. 1H-NMR spectra showed mostly the free ligand signals after one week, which were replaced with those of the complex, with a more complicated pattern in the aged sample, in accordance with the conversion seen in the 27Al spectra. When the equilibrium was reached, the pH-values for each separate sample were measured and their 1H- and 27Al-NMR spectra were also recorded (Fig. 2).

Fig. 2
figure 2

104.3 MHz 27Al-(top) and 400 MHz 1H-NMR (bottom) spectra of the Al(III)-NOTA system (cAl (III) = 2.88 mM, cH3NOTA = 2.99 mM) at different pH values (aging time: 4 months)

Full size image

The signals in the 1H-NMR spectra are quite difficult to assign, but the complex formation process is clearly shown in the range of 2.8–4.0 ppm with the increase of pH. This phenomenon can easily be followed with 27Al-NMR: the signal intensity of the aqua complex [Al(H6O)]3+ at 0 ppm decreases, while that of [Al(NOTA)] at 49 ppm increases with pH. The 27Al-NMR signal of [Al(NOTA)] is unexpectedly narrow, which can be explained by the compact and symmetric (octahedral) structure of the complex as evidenced by X-ray crystallography [11]. The stability constant of [Al(NOTA)] was calculated from the pH-metric data using PSEQUAD. The standard deviation of the obtained value as well as the fitting parameter decreases significantly when the formation of a monoprotonated species ([Al(HNOTA)]+) was also taken into account. This indicates that the protonation of the complex also occurs under these circumstances, although the lgK H(AlNOTA)  = 1.9 ± 0.3 shows large standard deviation and it can be considered as just an upper limit. The 27Al-NMR shift of the complex does not change with pH at our experimental conditions, pointing to a minor impact of protonation on the overall structure of the complex. The data points gained from 27Al-NMR adhere more to distribution curves calculated without including a protonation constant (see Fig. 3), although the deviations are higher at lower pH. Stability and protonation constants of the complexes are defined by Eqs. 3 and 4:

$$K_{\text{AlNOTA}} = \frac{{ [ {\text{Al(NOTA)]}}}}{{ [ {\text{Al}}^{ 3+ } ] [ {\text{NOTA}}^{{ 3 { - }}} ]}}$$
(3)
Fig. 3
figure 3

Distribution curves of Al(III)-containing species in Al(III)-NOTA system (cAl3+ = 3 mM; cNOTA = 3.05 mM). Lines correspond to pH-potentiometric results, the dashed line is the sum of fractions that contain NOTA-bound Al(III) species [Al(NOTA)] and [Al(HNOTA)]+, data points obtained from 27Al-NMR are marked with spots (free Al3+) and squares (NOTA-bound Al(III)-species)

Full size image
$$K_{\text{AlHNOTA}} = \frac{{ [ {\text{Al(HNOTA)}}^{ + } ]}}{{ [ {\text{Al(NOTA)][H}}^{ + } ]}}.$$
(4)

The square brackets mean equilibrium concentrations. One can conclude that the formation constant of [Al(NOTA)], lgK [Al(NOTA)] = 17.9(1), is quite high compared to that of [Gd(NOTA)] (lgK [Gd(NOTA)] = 13.7) [12]. The stability of [Al(EDTA)]−, having 2 N and 4 O donor atoms is also smaller (lgK = 16.1) compared to [Al(NOTA)] with 3 N and 3 O donors. This could be related to the macrocyclic effect and the size-match between the ligand cavity and the ionic radius: i.e. Al3+ fits very well into the cavity of the NOTA3− ligand.

In order to inspect if the model based on the pH-potentiometric results was adequate, 27Al-NMR signal intensities of Al(III) and [Al(NOTA)] measured at different pH values are shown on the Al(III)-distribution diagram. The results are shown in Fig. 3, along with calculated curves based on the pH-potentiometric data.

No new signal assignable to [Al(HNOTA)]+ was observed due to protonation, the chemical shift is likely insensitive to ligand protonation. Altogether, the results seen in Fig. 3 strongly “validate” the model based on pH-potentiometry. The obtained protonation constant (see Table 1), although having a relatively large error, is similar to those usually observed for macrocyclic metal complexes [13–15].

Considering the formation of the mixed complex [Al(EDTA)(OH)]2− [16, 17] (lgK mix = 5.9), a titration was performed in the alkaline region in order to detect the mixed [Al(NOTA)(OH)]− complex. In fact, a proton release process was observed under basic conditions, which can be explained by the formation of the proposed ternary complex, as defined in Eqs. 5 and 6:

$$[{\text{Al}}({\text{NOTA}})] + {\text{H}}_{2} {\text{O}} \rightleftharpoons {\text{[Al(NOTA)(OH)]}}^{ - } + {\text{H}}^{ + }$$
(5)
$$K_{\text{AlNOTA(OH)}} = \frac{{ [ {\text{Al(NOTA)(OH)}}^{ - } ][{\text H^{ + }} ]}}{{ [ {\text{Al(NOTA)]}}}}.$$
(6)

The value of lgK AlNOTA(OH), determined to be −12.2(1) is 6 orders of magnitude lower than the similar constant for [Al(EDTA)(OH)]2−. Moreover, this ternary complex cannot be studied as a real equilibrium species since it slowly dissociates to [Al(OH)4]− and NOTA3−, although the titration curve could be described quite well with the above constant (in a “fast” titration: waiting time after addition of each dose of titrant NaOH was about 40 s). Considering the stability and protonation constants obtained for [Al(NOTA)], as well as the hydrolysis constants of Al(III) (lgK 11 = −5.48; lgK 12 = −9.9; lgK 13 = −15.6; lgK 14 = −23.0) ion, distribution diagrams were calculated and plotted (Fig. 4).

Fig. 4
figure 4

Distribution of Al(III)-containing species in Al(III)-NOTA systems in a broad pH-range, made with- (right), and without (left) including Al(III) hydrolysis constants, cAl = 3 mM; cNOTA = 3.05 mM. The [Al(NOTA)(OH)]− complex is a metastable species (see text)

Full size image

In spite of the high thermodynamic stability of the [Al(NOTA)] complex, the tetrahydroxido aluminate (III) complex is formed above pH 8. This is a slow process compared to the equilibration time applied during the pH-potentiometric titration, which explains why the distribution curves shown in the left side of Fig. 4 seem to be consistent with the “fast” titration data (real time titrations). If enough time is given for these system to equilibrate, the species found in the acidic region would be the same, while the speciation at pH>7.0 is expected to change dramatically (as shown in the right part of Fig. 4). The transformation of the metastable system (shown left) to the equilibrium state (shown right), i.e. the dissociation kinetics of [Al(NOTA)] were further studied by 27Al-NMR and are discussed in the next part of the paper.

Kinetic study

The high kinetic inertness of metal complexes used in living systems is one of the most important, if not the most important criterion for safe application. The dissociation of inert complexes is negligibly slow regardless of thermodynamic stability, which prevents in vivo release of (toxic) metal ions. For this reason, the kinetic properties of [Al(NOTA)] were investigated in detail. To the best of our knowledge, no such published data is available in the literature.

The in vivo dissociation of metal complexes (ML) can occur when there is either a competition reaction between the chelated metal ion and endogenous ones such as Zn2+ or Cu2+ for the ligand, or between the endogenous ligands for the metal ion in the complex. In other words, the presence of these competing components defines the thermodynamic condition set for the dissociation of ML. The dissociation mechanisms of metal complexes do not differ fundamentally from each other, the reaction most often proceeds via proton- and metal-assisted pathways. In the case of macrocyclic complexes, proton-assisted dissociation is generally the most effective path for dissociation, while metal ions and small endogenous anions (carbonate, citrate and phosphate) may sometimes play a role in accelerating the dissociation of ML complexes of open-chain ligands [18].

In our case, the kinetic inertness of [Al(NOTA)] was investigated both in acidic and basic conditions, at pH-values where the complex is thermodynamically unstable. In these cases, the kinetic inertness of [Al(NOTA)] can be characterized by the rate constants of H+ and OH− ion-catalyzed dissociation reactions, where protonated and ternary hydroxido complexes are formed as intermediates respectively (Eqs. 7 and 8):

$$[{\text{Al}}({\text{NOTA}})] + {\text{H}}^{ + } \rightleftharpoons [{\text{Al}}({\text{HNOTA}})]^{ + }$$
(7)
$$[{\text{Al}}({\text{NOTA}})] + {\text{OH}}^{ - } \rightleftharpoons [{\text{Al}}({\text{NOTA}})({\text{OH)}}]^{ - }$$
(8)

The reactions were examined under pseudo-first order conditions, ensured by the high NaOH and H+ concentration. The observed rate constants, k obs, were fitted to Eq. 9, where k obs is a pseudo-first order rate constant and [Al(NOTA)]t is the total concentration of the complex.

$$- \frac{{{\text{d[Al(NOTA)]}}_{\text{t}} }}{{{\text{d}}t}} = k_{\text{obs}} [ {\text{Al(NOTA)]}}_{\text{t}}$$
(9)

In order to obtain data for the acid-catalyzed dissociation, which was assumed to be quite slow, the reaction was investigated at high acid concentration. Fig. 5 only shows a very small signal (about 1.5 % of the overall intensity) of Al(III) at 0 ppm in 1.0 M HCl, even after 16 days, while the half-life of “gold standard” [Gd(DOTA)]− was found to be a few hours in this medium [18] indicating an immeasurably slow rate of the acid catalyzed dissociation.

Fig. 5
figure 5

27Al-NMR spectrum of 1 mM [Al(NOTA)] in 1 M HCl taken 16 days after the components were mixed

Full size image

Usually, the first step of the proton-assisted dissociation of ML complexes formed with polyamino-polycarboxylate ligands, like NOTA, is the protonation of a pendant arm of the coordinated ligand. The proton is transferred in the second step to a nitrogen donor atom of the macrocyclic ring, followed by the dissociation of the complex. It seems that the compact structure of [Al(NOTA)] strongly hampers this proton transfer between the donor atoms, resulting in a very high inertness under acidic conditions.

Knowing the high affinity of Al(III) towards forming the tetrahydroxide [Al(OH)4]− complex, the dissociation of [Al(NOTA)] was also studied under basic conditions. Four NMR samples were prepared containing 1 mM of [Al(NOTA)] in 0.05–0.5 M NaOH, and their 27Al-NMR spectra were recorded regularly during the next 3 days.

The dissociation reactions were followed by observing the changes in the signal intensity of [Al(NOTA)] over time (Fig. 6). As it is seen in Fig. 6, the dissociation of [Al(NOTA)] is relatively slow: the time required for equilibration in 0.5 M NaOH was more than 3 days shown by the decrease in the intensity of the [Al(NOTA)] signal at 48 ppm. The course of dissociation was followed with the signal intensity change of [Al(NOTA)].

Fig. 6
figure 6

Time-dependence of 27Al-NMR spectra of 1 mM [Al(NOTA)] in 0.5 M NaOH, with the inset box indicating the decrease in [Al(NOTA)] signal intensity as a function of time

Full size image

Using the intensity/time data pairs, the first order rate constants were calculated for each NaOH concentration. According to the data presented in Fig. 7, reaction rates were found to be directly proportional to NaOH concentration.

Fig. 7
figure 7

Dependence of the pseudo first-order rate constants, k obs, for the hydroxide-assisted dissociation of [Al(NOTA)] on the hydroxide concentration (25 ºC, 1.0 M NaCl)

Full size image

Considering the result that [Al(NOTA)] is dominantly present as [Al(NOTA)(OH)]− ternary complex in this NaOH concentration range (mentioned earlier as an “intermediate”), the following reaction pathways can be assumed: [Al(NOTA)(OH)]− dissociates spontaneously (Eq. 10), indicated by the intercept of Fig. 7 and via a hydroxide-assisted reaction, which is characterized by the rate constant k OH (Eq. 11):

$$[ {\text{Al(NOTA)(OH)]}}^{ - } \mathop{\longrightarrow}\limits^{{{\text{k}}_{ 0} }}{\text{ [Al(OH)}}_{ 4} ]^{ - } + {\text{NOTA}}^{ 3- }$$
(10)
$$[ {\text{Al(NOTA)(OH)]}}^{ - } + {\text{OH}}^{ - } \mathop{\longrightarrow}\limits^{{{\text{k}}_{\text{OH}} }}{\text{ [Al(OH)}}_{ 4} ]^{ - } + {\text{NOTA}}^{ 3- } .$$
(11)

The rate law in alkaline medium can be expressed as follows:

$$k_{\text{obs}} = k_{0} + k_{1} [{\text{OH}}^{ - } ]$$
(12)

Fig. 7 shows the best fit of the k obs values to the Eq. 12. The values obtained for k 0 and k 1 are (2.0 ± 0.1) × 10−6 s−1 and (6.8 ± 0.5) × 10−6 M−1s−1, respectively.

Because of the occurrence of various dissociation pathways characterizing the decomplexation of agents used in vivo, direct comparison of their kinetic inertness is difficult. Therefore, the half-lives (t 1/2) of dissociation under physiological conditions (pH = 7.4) are frequently used for comparison. The kinetic data obtained from our investigations show that decomplexation is accelerated by the formation of the ternary hydroxido complex. However, the proton-assisted dissociation of [Al(NOTA)] does not play a significant role. Using the rate constants characterizing the hydroxide-assisted dissociation, the t 1/2 value of [Al(NOTA)] was calculated and found to be 94 h at physiological pH. Considering that the half-life of excretion is about 1.5 h for most contrast agents, the application of [Al(NOTA)] as a carrier for F− ions may rightly be considered safe.

Preliminary results of 19F-NMR study of ternary complex formation

As it was indicated earlier [4, 5], [Al(NOTA)] is known to be capable of carrying the radioactive 18F-isotope that has a key role in PET investigations. Laverman and his coworkers developed a method for radiolabeling the Al(III) complex of an Octreotide derivative of NOTA, in which an acetate arm has been modified (amidated) to link the peptide [19]. This method uses high temperature (100 °C for 10–15 min) for the complexation in acetate buffer, since the formation rate of [Al(NOTA)] is relatively low under these conditions. The radiolabeled compound ([Al(18F)(NOTA)-octreotide]) was found to be intact in human serum after 4 h and no in vivo dissociation was detectable in mice after 30 min. In fact, this procedure provides a promising agent for PET investigations, but no information on the thermodynamic stability or kinetic inertness of the Al(III)-F-NOTA ternary complex, which, available so far, would be useful in predicting its long term in vivo behavior.

The radiolabeling experiments are performed at very low concentrations (at about 10−9 M). Therefore, the procedures cannot be directly adapted for mM concentrations used in 19F-NMR. However, we wanted to detect the formation of [Al(NOTA)(F)]− mixed complex both by fluoride selective electrode and by 19F-NMR. Attempts were made to follow the formation of the mixed complex between [Al(NOTA)] and F−, using both direct potentiometric titration and "out of cell” technique (see Supporting Information). In all cases, the potentials recorded by F−-selective electrode in both the presence and absence of (preprapared) [Al(NOTA)] complex were identical, indicating that the expected complex formation does not occur under these circumstances (25 °C, 0.15 M NaCl). An obvious explanation of this phenomenon can be the high rigidity of [Al(NOTA)]. Since the coordination sphere of the Al(III) ion is saturated with the donor atoms of NOTA, coordination of one further F− ion to the inner sphere requires the dissociation of a carboxylic oxygen, which is an extremely slow process proven by the kinetic experiments in HCl. Therefore, in order to reach equilibrium, the samples were heated for 10 min at 100 °C as described in the literature [4], and their 19F-NMR spectra were recorded after cooling to room-temperature. Only the signal of free F− was detectable after this treatment, which means no detectable formation of the ternary complex, i.e. the formation is quite slow even at 100 °C. It should be noted, that the behavior of [Al(EDTA)F]2− could easily be followed by fluoride selective electrode. [20]

Fig. 8 shows the dependence of 19F-NMR spectra in samples with equimolar Al(III), NOTA and fluoride on heating time, recorded 24 h afterwards. In the sample without heating, there is a new signal, most likely the signal of the ternary complex at −47 ppm among the signals of Al(III) fluorido complexes (−35 to 36 ppm) [21]. In the sample heated for 10 min, the signal of the free F− ion appears and the signals of Al(III) fluorido complexes decrease, together with a slight increase in intensity of the mixed complex. Further treatment of the samples causes complete decomposition of the mixed complex to parent AlF2 + complex and F− ion, shown by the signals at −35 ppm and −5 ppm, respectively. (The signal at −5 ppm is the time averaged signal of F− and HF at pH 4.8 in acetate buffer.)

Fig. 8
figure 8

376.5 MHz 19F-NMR spectra of samples containing Al(III), NOTA and F− in equimolar amounts after heating in pH = 4 acetate buffer for varying times at 100 °C

Full size image

The formation of [Al(NOTA)(F)]− was almost 100 % in 50 % ethanol after 15 min of heating according to the 19F-NMR spectrum (Fig. 9). The chemical shift was −48 ppm, slightly different from the −47 ppm measured in water likely due to solvent effect. The role of ethanol could be related to the decreased basicity of the ring nitrogen donor atoms in NOTA [22]; resulting in faster complex formation due to weaker competition between the H+ ion and Al(III) for the donor atoms of the macrocycle. A detailed study of formation kinetics for metal-macrocyclic complexes in mixed solvents are in progress in our laboratory.

Fig. 9
figure 9

376.5 MHz 19F-NMR spectrum of a sample containing Al(III), NOTA and F− in equimolar amounts after heating for 15 min at 100 °C in 50 % ethanol

Full size image

Our preliminary results show that complex formation kinetics are fairly complicated in the Al(III)—NOTA3−—H+ —F− four-component system, further experiments should be performed to describe and understand the formation kinetics of the mixed complex.

Conclusions

The formation kinetics of [Al(NOTA)] in acidic solutions is fairly slow. Therefore, equilibrium constants have been evaluated from pH-potentiometric data measured in “batch” samples aged for four months. 1H- and 27Al-NMR experiments were used to check the time course of equilibration and validate the equilibrium model consisting of [Al(NOTA)] and [HAl(NOTA)]+ in acidic solution, with equilibrium constants lgK Al(NOTA = 17.9(1) and lgK HAl(NOTA)  = 1.9(3). There is a metastable mixed hydroxido complex detected in alkaline solution, [Al(NOTA)(OH)]− with lgK AlNOTA(OH) = −12.2(1).

The dissociation of [Al(NOTA)] in acidic solutions is immeasurably slow, only about 1.5 % of the complex is converted to Al(III) in 16 days in 1 M HCl. The dissociation in alkaline solution follows the rate law k obs = k 0 + k 1[OH−] with k 0 = (2.0 ± 0.1) × 10−6 s−1 and k 1 = (6.8 ± 0.5) × 10−6 M−1s−1. Using these kinetic parameters the half-life of [Al(NOTA)] dissociation was calculated to be t 1/2 = 94 h at physiological pH.

The formation of the ternary [Al(NOTA)(F)]− via direct reaction of [Al(NOTA)] and F− cannot be detected using fluoride selective electrode and/or 19F-NMR spectroscopy. The kinetics is fairly complicated in the Al(III)—NOTA3-—H+—F− four component system. However a new signal in the 19F-NMR spectrum at −48 ppm might be assigned to the ternary complex, which was formed in water–ethanol mixture with high yield.