Table 1 shows the performances of the ⁹⁹Mo/^99mTc generators based on magnesium ⁹⁹Mo-molybdate compounds prepared in this research. Results are divided into three series, in line with the type of magnesium salts used in the generator preparation: magnesium chloride hexahydrate (series A), magnesium nitrate hexahydrate (series B) and magnesium sulfate hexahydrate (series C). The generator performances were compared with those advised by the Pharmacopoeia for the ^99mTc eluates used with medical purposes: ⁹⁹Mo breakthrough less than 0.015%, a minimum percentage of 95% for the radiochemical purity, a chemical purity less than 10 ppm for aluminium and pH values between 4.5 and 7.5 [15].

⁹⁹Mo breakthrough percentages of less than 0.015% were only obtained in the matrices prepared from: (a) 0.5 mol/L MgCl₂·6H₂O (series A) and b) 1 mol/L MgNO₃·6H₂O solutions (series B) using ammonium molybdate solutions at pH of 7 and a Mg:Mo molar ratio of 1:2 (Figure 1). However, ^99mTc elution efficiencies of theses generators were less than 48%, except for the matrix B7. On the other hand, the highest elution efficiencies (>70%) were obtained in the generators prepared from 0.1 mol/L Mg(NO₃)₂·6H₂O solutions, at a Mg:Mo molar ratio of 0.2:1 and ammonium molybdate solutions at pH values between 4.5 and 10, however under these conditions, ⁹⁹Mo breakthrough of the eluates were more than 0.7%, apart from the matrix B7. ^99mTc eluates of the matrices prepared preferably with Mg:Mo molar ratio of 2:1 presented radiochemical purity of more than 95% and, in general, those made from MgCl₂·6H₂O solutions which satisfy the eluate pH values fixed by the Pharmacopoeia: between 4.5 and 7.5, while the eluate pH values obtained from the matrices formed with MgSO₄·6H₂O were the more acid, between 1 and 3. The average elution volume of all the generators studied ranged between 2 and 3.5 mL and all ^99mTc eluates had an Al content of less than 10 ppm. It is important to note that a high ⁹⁹Mo breakthrough percentage in the eluates entails the presence of Mg²⁺ in solution.

The Mo and Mg content in the generators is directly connected with: the Mg:Mo molar ratio, the type and concentration of magnesium salt used during matrix synthesis and matrix washing before irradiation. Thus the highest (75-50%) and lowest (18–7%) Mo percentages, and vice versa for Mg content, were recorded in the washed matrices and those prepared from MgSO₄·6H₂O solutions at Mg:Mo molar ratio of 2:1 respectively.

Matrix washing caused a decrease of the ^99mTc elution efficiencies and Mg percentage in the matrix, while an increase in magnesium salt concentration (series C) induced a drop in the ^99mTc elution efficiency and acidification of ^99mTc eluates. The addition order of magnesium salt and ammonium molybdate solutions, and ammonium molybdate pH in the matrix process preparation (series B) did not cause meaningful changes in ⁹⁹Mo/^99mTc generator performance.

^99mTc eluates produced by the generators prepared from MgCl₂·6H₂O solutions (series A) mostly attained the pH values established by the Pharmacopoeia: between 4.5 and 7.5, while higher acid eluates were obtained in the matrices synthesized from MgSO₄·6H₂O solutions. When the Mg proportion was higher than Mo in the Mg:Mo molar ratio, ⁹⁹Mo breakthrough percentage increased in the ^99mTc eluates and the Mo percentages in the matrix decreased. All ^99mTc eluates of series A were colorless, those prepared from MgNO₃·6H₂O solutions at pH 10 or adding the ammonium molybdate solutions to magnesium salts (series B) presented a yellow coloration, while some of series C eluates showed a yellow coloring or a blue precipitate, in fact only the ^99mTc eluates obtained from generator prepared with 0.1 mol/L MgSO₄·6H₂O solutions were colorless.

Crystalline phases identified by XRD (see Table 1 and Figure 2a) showed that the type of magnesium salt used in preparing generator matrices determines their chemical composition. In accordance with these data, Mg-Mo compounds prepared from MgCl₂·6H₂O, Mg(NO₃)₂·6H₂O and MgSO₄·6H₂O solutions are mainly constituted of: (a) NH₄MgCl₃·6H₂O, MoO₃ and NH₃(MoO₃)₃; (b) amorphous compounds and unidentified crystalline phases and (c) NH₄MgCl₃·6H₂O, (NH₄)₂Mg(SO₄)₂·6H₂O and amorphous compounds, respectively. These results are congruent with the thermogravimetric and infrared spectra shown in Figures 2b which present a characteristic pattern for each magnesium salt used. The thermal decomposition step multiples of different matrices prove the presence of compound mixtures. In series A, it is possible to identify five main causes for weight-loss, firstly water elimination of ammonium magnesium chloride hydrate (∼116 °C), later the transformation of NH₄MgCl₃ in Mg(OH)Cl and NH₄Cl (160–170 °C), after NH₄Cl decomposition (∼220 °C), the formation of MgO from Mg(OH)Cl (350–550 °C) and finally the decomposition of MoO₃ (770–800 °C) [8,16]. Three weight-losses are evident in the series B at 214, 330 y 770 °C which could be derived from NH₃Cl, Mg(OH) and MoO₃ decomposition respectively considering that amorphous materials and the unidentified phases (Figure 2b) are constituted by Mo, Mg, NO₃⁻, NH₃ and Cl and making an analogy with the compounds formed in series A. In the case of series C, the weight-loss is fixed by the NH₄MgCl₃·6H₂O, (NH₄)₂Mg(SO₄)₂·6H₂O and amorphous molybdenum compound decomposition (see Table 1 and Figure 2b): dehydratation (96–130 °C), elimination of [NH₄⁺] in NH₄MgCl₃ (167 °C) and (NH₄)₂Mg(SO₄)₂ (451 °C), decomposition of NH₄Cl (∼237 °C), formation of MgO (300–400 °C) and decomposition of MoO₃ (756 °C) [16].

The effect of magnesium salt on forming different magnesium-molybdenum compounds during generator matrices preparation was also demonstrated by infrared analysis shown in Figure 2c. The matrix spectra prepared using MgCl₂*6H₂O, Mg(NO₃)₂*6H₂O and MgSO₄*6H₂O solutions have similar troughs in the 3500-1200 cm⁻¹ region, but with differences in band intensities.

In this region, ammonium and water displays strong broad N-H and O-H stretching bands between 3500 and 3300 cm⁻¹ and bands at 1405 and 1640 cm⁻¹ respectively [8,17-18]; while significant differences in intensities and wavenumber values and of each matrix were noted in fingerprint region (1200–400 cm⁻¹). All matrices spectra exhibit characteristic absorption bands of Mo-O-Mo vibration at 960, 910 cm⁻¹ and N-H bonds at 1075 and 1220 cm⁻¹ as well as a band at 620 cm⁻¹ possibly originated from Mg vibrations [8,18-19]. Only the matrix prepared from Mg(NO₃)₂*6H₂O presented broad bands at around 470 cm⁻¹ attributed to Mg-O bonds [19], and that from MgSO₄*6H₂O showed characteristic bands assigned to [SO₄²⁻] (628, 700, 1075, 1130, 1287 cm⁻¹) and Mo-O vibrations at 792 and 880 cm⁻¹ whereas that from MgCl₂*6H₂O presented peaks 760 and 545 cm⁻¹ assigned to the Mo-O vibrations and possibly to Mg-Cl bonds respectively [8,18-19].

The X-ray diffraction patters and thermograms of a typical matrix washed and unwashed are shown in Figure 3a and 3b. The diffractograms show that the washed matrix before irradiation is constituted only by MoO₃ and the unwashed one by a mixture of MoO₃ and NH₄MgCl₃*6H₂O. These data match with thermograms of the washed and unwashed matrix which have typical patterns of pure and mixed compounds respectively. Thus, washed matrices before irradiation cause soluble compounds to be removed, mainly those containing ammonium and magnesium in the matrix and conversion of the molybdenum compounds in MoO₃. It is important to note that this behavior is independent of the type of magnesium salt used in preparing the matrix (Table 1).

The effect of matrix washing on morphology is shown in Figure 3c. The unwashed matrices present a crystalline phase soaked in an amorphous material, and the washed matrices only have the crystalline phase, constituted by rods of different thicknesses and length.

The performance of the ⁹⁹Mo/^99mTc generators based on magnesium-molybdate compounds depends upon matrix preparation and treatment methods. The latter has to be an insoluble precipitate to avoid ⁹⁹Mo leakage, whilst simultaneously allowing ^99mTc release, and have a high Mo content that enables the use of low specific activities of ⁹⁹Mo (2.5 Ci/g) in the generator and a good thermal and radiation stability.

Magnesium molybdate is fairly soluble in water [20-21] and literature has reported obtaining magnesium molybdate precipitates, mainly used in catalysis, applied suitable thermal treatments and magnesium and molybdate concentrations. For example, Yoon et al. have reported magnesium molybdates precipitation from Mg(NO₃)₂·6H₂O and (NH₄)₃Mo₇O₂₄·4H₂O solutions and calcining at different temperatures (200–650 °C) [19]. Ozeki et al. prepared magnesium molybdate solids by concentrating a mixed solution of 0.1 mol/L sodium molybdate and 4.5 mol/L magnesium chloride [20]. Amber et al. obtained MgMoO₄·H₂O from 0.1 mol/L Na₂MoO₄ and 0.5 mol/L MgCl₂ solutions at pH 6 treated at 155 °C for 3 days [22].

In this work, insoluble magnesium molybdate compound precipitation was favored by adjusting pH and irradiation, and not by thermal treatment. Under these experimental conditions, magnesium molybdate compounds obtained were mainly polymetalates salts such as xNH₄MgCl₃·yMoO₃, and polymolybdates [NH₄Mo₅O₁₅(OH)] (see Figure 2 and Table 2). Assuming that compound formation is the result of three steps, firstly the formation of ammonium molybdates according to reaction (1): (1) Mo O 3 + 2 N H 4 OH ↔ ( N H 4 ) Mo O 4 + H 2 O

The molybdate ion generally exists as MoO₄²⁻ in alkaline or neutral solutions (pH > 6) while polymolybdate ions such as [Mo₇O₂₄]⁶⁻, [Mo₈O₂₆]⁴⁻, [Mo₃₆O₁₁₂]⁸⁻ are formed in acid solutions [19,23-27]: (2) xMo O 4 2 − + y H + ↔ M o x O z ( OH ) 8 x − y − 2 z ( 2 x − y ) − + ( y − 4 x − z ) H 2 O

Thus the ammonium molybdate solutions prepared at pH 10 and 7 contain simply MoO₄^2- ions and those at pH 4.5 a mixture of polymolybdate ions where the predominant species is probably the [Mo₇O₂₄]⁶⁻ ion [18]. In the second step, ammonium molybdates react with magnesium solutions to form magnesium-molybdates solutions. Considering these solutions pH values can vary between 8 and 4.3, the magnesium molybdates can be constituted by MoO₄²⁻ (pH > 6) or polymolybdates (pH < 6) according to: (3) ( N H 4 ) Mo O 4 + MgX ↔ MgMo O 4 + N H 4 X (4) ( N H 4 ) ( 2 x − y ) M o x O z ( OH ) 8 x − y − 2 z + ( 2 x − y ) MgX ↔ M g ( 2 x − y ) M o x O z ( OH ) 8 x − y − 2 z + ( 2 x − y ) N H 4 X X = C l − , N O 3 − or S O 4 2 − ,

In the last step, the magnesium molybdates are induced to precipitate by adjusting pH values of the solutions to be between 1.9 and 0.3. At pH < 2, literature has reported the presence of very large polymolybdate species like [Mo₃₆O₁₁₂]⁸⁻ [23,25] or MoO₃·2H₂O precipitation [19,28-29]. X-ray diffraction data (see Table 2 and Figure 2) suggests polymetalates formation such as xNH₄MgCl₃·yMoO₃ or polymolybdates according to: (5) MgMo O 4 + 2 N H 4 X + 2 HCl ↔ N H 4 MgC l 3 · Mo O 3 · H 2 O + N H 4 X (6) uMgMo O 4 + vN H 4 X + wHCl ↔ u NH 4 MgC l 3 · ( N H 4 ) ( 2 x − y ) M o x O z ( OH ) 8 x − y − 2 z + ( v − u − 2 x − y ) N H 4 X (7) M g ( 2 x − y ) Mo x O z ( OH ) 8 x − y − 2 z + vN H 4 X + wHCl ↔ ( 2 x − y ) N H 4 MgC l 3 · ( N H 4 ) ( 2 x − y ) M o x O z ( OH ) 8 x − y − 2 z + v − ( 2 x − y ) N H 4 X

The Cl⁻ ion usually displaces NO₃⁻ and SO₄²⁻ ions [30] when magnesium nitrate and sulfate are employed in preparing magnesium molybdates; for that reason ammonium magnesium chlorides (NH₄MgCl₃) were present in all the series studied (see Table 2 and Figure 2), however mixtures of NH₄MgCl₃ and (NH₄)₂Mg(SO₄)₂ were also identified in matrices prepared from magnesium sulfates.

An excess of molybdenum favors polymolybdates and formation of amorphous phases whereas a surplus of magnesium the presence of ammonium magnesium salts and crystalline phases. Thus, the crystallinity degree of the compounds contained in the matrix is closely attached to ⁹⁹Mo/^99mTc generator performances. For example amorphous matrices presented the best ^99mTc elution efficiencies (series B) while the crystalline (series A) presented lower ⁹⁹Mo breakthrough (see Table 1, Figure 1). Assuming that amorphous materials also consist of molybenum oxides or polymolybdates and that the oxides and hydrous oxides of Mo(VI) exhibit cation exchange properties and show little or no anion exchange character even in acid solution [31] and ammonium magnesium salts have no adsorption properties, so the separation mechanism of the ⁹⁹Mo and ^99mTc in the generators can be explained by free diffusion of ^99mTcO₄⁻ ion inside the matrix because the ^99mTcO₄⁻ anion produced in the generator is not adsorbed in the matrix and can be removed from the chromatographic column by elution with isotonic saline solution, leaving the ⁹⁹Mo inside. In accordance with this argument, a crystalline matrix acts as a molecular sieve preventing ^99mTc mobility and causing generator efficiency decrease. Whereas a flexible random network (amorphous) increases generator efficiency and radiochemical purity because the matrix is more elastic but simultaneously harder and more resistant to mechanical breakdown and more difficult to dissolve. The low ^99mTc eluate radiochemical purities obtained in some generators can be explained by Tc(VII) reduction caused by the presence of insoluble species of polymolybdates, which are strong oxidizing agents [18].

Inorganic materials are susceptible to irradiation-induced amorphization producing particularly volume changes in crystalline or amorphous phases. The main concern with large differential volume changes is that it may affect atomic bonding, local coordination, and the pathways for ion exchange, all of which can impact the release rates of radionuclides [32]. Thus the matrix amorphization caused by its irradiation could be linked to the high ⁹⁹Mo breakthrough obtained in generators for which matrices are mainly formed by amorphous compounds such as the series B and C.

Magnesium ⁹⁹Mo-molybdate compounds were formed from magnesium and molybdate solutions. The molybdate solutions were prepared from MoO3 natural pellets, previously heated to 650 °C for 1 h and dissolved in 2 mol/L NH₄OH at a MoO₃:2NH₄OH molar ratio [8]. The pH of the formed ammonium molybdates was adjusted by adding 4 mol/L HCl and converted into magnesium molybdate by reacting with magnesium solutions. Magnesium molybdates pH were also adjusted using 4 mol/L HCl. The resulting solids were dried for 2 days using an infrared lamp and crushed in an agate mortar. One portion of magnesium molybdate precipitate was placed on a funnel to be washed using 200 mL of distilled water and the washed and unwashed solids were dried for 1 day at 40 °C in a stove. The dried magnesium molybdate were irradiated for 2 h at a neutron fluence of about 1.61 × 10¹³ n cm⁻²s⁻¹ in the Triga Mark III Reactor (Mexico). After irradiation, about 1 g of magnesium ⁹⁹Mo-molybdate (∼4.9 MBq/g) were added into a glass column (12 mm × 70 mm) containing a bed of 1 g acid alumina. The column was finally washed with 20 mL of saline solution [8,17-18]. The magnesium molybdate compounds were synthesized in duplicate at different conditions; where parameters such as magnesium salts and concentrations (MgCl₂·6H₂O, Mg(NO₃)₂·6H₂O, MgSO₄·6H₂O), Mo:Mg molar ratios, ammonium and magnesium molybdates pH and the addition order of magnesium and molybdenum solutions were evaluated (see Table 2).

The generators were eluted with 6 mL of 0.9% NaCl every 24 h for 1 week and the following parameter of the ^99mTc eluates were determined: ^99mTc elution efficiency, ⁹⁹Mo breakthrough, ^99mTc elution profile, ^99mTc radiochemical purity, pH eluate and aluminium concentration. The ^99mTc elution efficiency and the ⁹⁹Mo breakthrough were calculated from the ^99mTc and ⁹⁹Mo activities measured in a CRC-10R Capintec dose calibrator and a GeHp solid state detector (Canberra 7229P) coupled to a PC-multichannel analyzer (ACUSPECT-A, Canberra, Australia). The radiochemical purity of the ^99mTc eluate was determined by paper chromatography using 1 CHR (Whatman®) paper as solid phase and 85% methanol as mobile phase. The ^99mTcO₄⁻R_f was 0.66–0.72. Aluminium and magnesium concentrations in ^99mTc eluates were determined by the aluminon and Eriochrome Black T methods [18,33]. The eluate pH values were determined by pH paper.

Magnesium-molibdate compounds were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), infrared spectrometry, thermogravimetry and neutron activation analysis. The X-ray diffraction patterns were obtained on a Siemens D500 diffractometer for 1 h and scanned from 2.5° to 70° with steps of 0.02°. SEM imaging was performed by Philips SL30. Digital images were obtained at 5,000×, 3,000×, 1,000× and 500× magnifications in randomly selected fields. The infrared measurements were taken on a Nicole Mgna-IR™ spectrometer 550 with the samples pressed in KBr pellets. The thermogravimetric analyses were performed using a Phillips unit at a heating rate of 10°/min under a nitrogen atmosphere [8,18]. Molybdenum and magnesium concentrations were determined by neutron activation. The procedure described in previous works was applied for molybdenum and in the case of magnesium, 50 mg of each magnesium molybdate and MgO, used as reference material, were irradiated in the Triga Mark II reactor at a neutron fluence of about 1.65 × 10¹² n cm⁻²s⁻¹ for 15 s. Magnesium was determined by the 843.4 keV γ-ray of ²⁷Mg by means of a HPGe detector at a counting time of 100 s [13].