An innovative process for fission based 99Mo production has been developed under Isotope Technologies Dresden (ITD) GmbH (former Hans Wälischmiller GmbH (HWM), Branch Office Dresden), and its functionality has been tested and proved at the Pakistan Institute of Nuclear Science and Technology (PINSTECH), Islamabad. Targets made from uranium aluminum alloy clad with aluminum were irradiated in the core of Pakistan Research Reactor-1 (PARR-1). In the mean time more than 50 batches of fission molybdenum-99 (99Mo) have been produced meeting the international purity/pharmacopoeia specifications using this ROMOL-99 process. The process is based on alkaline dissolution of the neutron irradiated targets in presence of NaNO3, chemically extracting the 99Mo from various fission products and purifying the product by column chromatography. This ROMOL-99 process will be described in some detail.
The present sources of molybdenum-99 (99Mo;
The first chemical process for separation of fission 99Mo was described by the Brookhaven group, USA [
Another small-scale production process for fission 99Mo was proposed by the Rossendorf group in which natural uranium as uranium oxide was used as target material [
Based on their own long-term experiences and considering international achievements in 99Mo production, scientists of the Radio-Isotope department of the former Rossendorf institute ZfK designed a new process for fission-based 99Mo production named ROMOL-99 [ The dissolution of the UAlx/Al-clad targets shall be performed in a mixture of NaOH/NaNO3 without H2 generation, under reduced pressure conditions. The Xe shall be trapped cryogenically after passing a gas treatment line. The NH3 generated in the dissolving process shall be separated prior to Mo separation. The radioiodine shall be separated prior to 99Mo-separation as well. During dissolving process nitrite is generated which shall be eliminated prior to the 99Mo separation.
Process flow scheme of the ROMOL-99 process.
The basic parameters of this process has been developed with modern nonradioactive analytical techniques by the IAF-Radioökologie GmbH Dresden, while the active testing and optimization of the process has been carried out at PINSTECH Islamabad under supervision of the German scientists. In this paper the chemical process of the ROMOL-99 technology will be described in some detail.
All chemicals were purchased from E. Merck (Germany) and were of guaranteed reagent grade (GR) or analytical reagent (AR) grade. Al2O3 (90 active acidic for column chromatography, 70–230 mesh ASTM) was used. Silver-coated alumina was freshly prepared at institute. Organic anion-exchange resin was purchased from BioRad, USA.
The non-radioactive development work was performed using uranium-free Al-plates (purchased from PINSTECH) having the same composition as the material used for the original targets.
Tracer experiments were performed using 131I tracer activities which were taken from the PINSTECH routine 131Iodine production, and the 99Mo tracer was taken from the routine PAKGEN 99mTc generator production (99Mo imported from South Africa).
Qualified HEU/Al alloy clad with high purity aluminum target plates [
Gamma ray spectroscopy high-purity Ge detector (Canberra Series 85 multichannel analyzer) was used to determine the activity balance during all process steps and for the determination of radionuclide impurities in the final 99Mo product. This concerns mainly 131I and 103Ru, 132Te. Beta counting of 89Sr and 90Sr was done by a liquid scintillation analyzer (Tri-Carb 1900 TR, Packard Canberra Company) after separation by ion-exchange and precipitation with the aid of carrier. Contamination of alpha emitters was done with the
The final 99Mo product was dispensed and assayed by means of a calibrated ionization chamber. Radioactivity concentration (MBq/cm3) was calculated by dividing the total product activity by the final volume of the product solution. All required nuclear data were taken from NuDat 2.5 [
When dissolving the target plates in the solvent, 3 M NaOH/4 M NaNO3, 3 reactions must be considered leading to different reaction products as follows:
Under the conditions that
The solvent volume needed for the process is determined by the solubility of the sodium aluminate (NaAlO2) which is 2.1 M/L corresponding to 57 g/L Al. Furthermore, the Na concentration should be kept as high as possible, in order to reach safely the saturation concentration for the precipitate Na2U2O7. A high nitrate concentration is needed for avoiding the formation of hydrogen, while the viscosity of the solution should be suitable for easy filtration. We found a composition of 3 M NaOH/4 M NaNO3 as most suitable for the dissolving process.
The dissolving process is strong exothermic (close to 600 kcal are generated for dissolving 100 g Al), and in addition the dissolving speed increases with the second power of the temperature. Thus, the reaction is self-accelerating. Following the experiences collected in Dresden (IAF) and PINSTECH, the control of the dissolving process is easy and safely possible by short heating and cooling pulses. With these techniques one can easily adjust the dissolving temperature at around 70–80°C. Furthermore, the process can be performed at slightly reduced pressure conditions (see Figure
Temperature gain during controlled dissolving process (a) and the pressure situation during the dissolving process (b).
Since the iodine shall be removed from the process solution using a silver-coated column material, the NH3 is recommended to be eliminated because it has potential to influence the efficiency of the iodine removal at the Ag-coated column. The simplest way to separate the NH3 is the distillation from strong basic solution. Preliminary experiments have shown that 150–200 mL distilled volume is sufficient. This volume can be distilled off from the target solution within about 20 minutes. In the production runs, the distilled NH3 is trapped in 5 N H2SO4 solution.
The precipitate that is formed during the dissolving process is composed of mainly 2 components: the Na-diuranate and in addition the nonsoluble hydroxides, oxides or carbonates of several alloying metals of the Al-matrix that are coprecipitated together with the Na2U2O7. Based on analytical data of the Al-matrix material used for the target preparation, the following quantities for the precipitate should be expected (Table
Composition of the target material and the related composition of the precipitate after the dissolving process.
Alloying element | Content (% of Al-weight) | Content in (mg) for 3 targets | Precipitated species | Precipitate in (mg) for 3 targets |
---|---|---|---|---|
Fe | 0.14 | 128,8 | Fe(OH)2 | 207.14 |
Mn | 0.002 | 1,84 | Mn(OH)2 | 3.03 |
Si | 0.01 | 9,20 | SiO2 | 19.33 |
Ca | 0.16 | 147,2 | CaCO3 | 950.43 |
C | 0.1 | 92 | C | 92.00 |
U | 5.16 | 5100 | Na2U2O7 | 6814.85 |
Assuming a density of the precipitate of 4.4 g/cm3 (based on ~30% porosity) and the uranium in the form of Na2U2O7 × 6 H2O, one would obtain a precipitate volume of ~2.37 cm3, which corresponds to a filter bed thickness of
The target element uranium after dissolution must be present exclusively in the chemical form of Na2U2O7 because it is well known that uranium species of lower oxidation stage absorb 99Mo and consequently lower the production yield. Dissolving the same targets alone in NaOH or KOH (without NaNO3) [
As shown from the crystallographic analysis, the target element uranium was found after our ROMOL-99 dissolving process straight as sodium diuranate (Na2U2O7) in the precipitate (Figure
X-ray diffraction patterns of the uranium precipitate that show only the reflections of Na2U2O7 and excluded other uranium species to be present. This X-ray diffractogram was performed by TU Dresden/Geologie.
The time needed for filtration is mainly determined by the surface area and the porosity of the used filter plate and the filter cake, the viscosity of the solution and the filtration pressure. The filter plate consists of a 3 mm thick metallic (INOX) sinter plate with a porosity of ~30
In order to minimize the risk of iodine release in later production steps and waste, the adsorption on silver is the most promising approach for trapping the radioiodine before the 99Mo is separated. During the production process, we have to deal with 132I (2.3 h half-life, daughter of 132Te), 133I (20.8 h half-life), and 131I (8.02 d half-life). Freshly prepared silver-coated Al2O3 material has shown to be the most appropriate material; this material has been prepared according to the Wilkinson et al. method [
Gamma spectrum of the Ag-coated Al2O3 column after passing the filtrated target solution. The measurement was done from large distance. Only gamma signals from iodine radionuclides and Ru could be identified.
For the main separation step—the separation of the 99Mo from the process solution after iodine removal—Al2O3 column chromatography has been selected. Molybdate is adsorbed from weak HNO3-acid media at Al2O3 (this principle is used in the 99Mo/99mTc-generator technology). Thus, the strong basic process solution needs to be acidified. This is not an easy step, since the Al-concentration is high. An anion-exchange process, as it is used in cases were only NaOH or KOH is involved for dissolving the targets under H2 generation, is not possible due to the high
As said before, we also have nitrite in the system, which is recommended to be destroyed. Simultaneously with the acidification process the nitrite is reduced with urea under nitrogen formation according to
The reaction gases of this acidification process and in addition a slight carrier gas flow release also remain volatile iodine species (from iodine residue and decay of Te-parent nuclides) and radio Xe (mainly from 133I-decay). The radio iodine is retained in a gas adsorption trap filled with Ag-IONEX. This is a Zeolite exchange material that adsorbs at
When the nitrogen formation and iodine release is finished, the solution is cooled down to room temperature and is ready for the Al2O3 column process.
The separation of the 99Mo from the acidified target solution is achieved via anion exchange chromatography using week acid Al2O3 as column material. The adsorption efficiency for Mo depends mainly on the salt concentration and the acidity of the solution and not so much on the absorber material itself. For the optimization of the column parameters, the control of the free acidity in the FEED solution played an important role. Due to the high salt concentration (especially that of Al3+), a direct pH-measurement is not possible. A potentiometric titration did not show the required precision (see Figure
Potentiometric titration of 1 mL of the acidified target solution after nitrite destruction diluted to 100 mL with 0.100 N NaOH.
The safe and better is to dilute a sample of the solution by a factor 1 : 100 with distilled water. This solution could perfectly undergo a pH measurement with an ordinary glass electrode. The pH determined in this way was always in the region of 2.2 < pH < 2.6, which means that the free acid concentration in the original FEED solution under practical conditions was in the range of about 0.15 < [H+] < 0.7 M.
Under practical conditions, the volume of the loading solution (FEED) is around 6 L (for ~100 g target material). After the loading process, the column shall be washed with 0.5–1.0 L of 0.5 M HNO3 500 mL water and then with 1500 mL 0.01 M NH4OH. The 99Mo is then eluted with up to 2000 mL of 1 M NH4OH solution. One obtains a raw 99Mo product of already ≫99% radionuclide purity.
In order to define the optimal Al2O3 column parameters one needs to consider the adsorption capacity of the exchange material Al2O3, the selectivity related to the separation from radioactive contaminations, the possible and needed loading- and elution speed which is relevant for the duration of the process.
The capacity of Al2O3 for Mo adsorption is known to be ~30 mg Mo/g Al2O3 column material. In test experiments using 50 mL of model-target solutions containing a Mo-concentration of 20–33 mg Mo/L and columns with 2 g Al2O3 column material (0.7 × 5.6 cm column dimension) using a flow rate of 7 cm/min corresponding to 2.7 mL/min the Mo could be adsorbed with an average yield of >90%. Thus, in these experiments only a small fraction (1.5–2.5%) of the capacity of the exchanger has been utilized. This corresponds to 0.5–0.8 mg Mo/g Al2O3. Based on this data one would need for processing of 3 target plates theoretically 140 g of Al2O3 corresponding to 152 mL Al2O3 for the separation process.
For defining the column dimensions one needs to find a compromise between the needed amount of the Al2O3 material and reasonable high applicable elution speed. Furthermore one has to consider losses due to irreversible bound Mo with increasing Al2O3 quantities. Assuming the following practical conditions: the total volume of liquids that has to pass the column is ~11 L composed from 5.8 L acidified target solution, 3.3 L wash solutions, and 2.0 L elution volume, the linear filtration speed is 7 cm/min (50 mL/min for loading and eluting and 80 mL/min washing), the diameter of the column shall be 5 cm,
one obtains a volume flow speed of 137 mL/min under practical conditions.
Considering the previous determined 140 g or 152 mL Al2O3 absorber material, one would obtain an absorber bed height of 7.8 cm. If one increases the dimensions by at least a factor 2 for compensating not optimal conditions, the length of the Al2O3 column becomes 15.6 cm filled with 304 mL Al2O3 absorber material.
During the commissioning, it has been demonstrated that a 250 g alumina column bed is acceptable which corresponds to a column bed volume of about 275 mL. Table
Optimal parameters for the alumina column process.
Al2O3 column | About 50 mm × 145 mm, bottom G 3 frit |
Al2O3 weak acid material | 250 g, size ~60 |
Loading process: | FEED volume ~5.8 L, 0.15 < [H+] < 0.7 M, ~120 min |
(1) Wash process | 1000 mL 0.5 M HNO3, time about 15 min |
(2) Wash process | 500 mL water, time about 7 min |
(3) Wash process | 1000 mL 0.01 M NH3, time about 15 min |
Liquid waste volume: | 8300 mL |
Elution | 1000 mL 1-2 M NH3, time about 30 min |
Alumina column process: | Total time about 3 h. |
Using the parameters shown in Table
Elution profile for 99Mo elution from the Al2O3 column with 1 M NH3 using the parameters shown in Table
The Al used in the target material contains some quantities of Si. It is well known that Si forms very unpleasant nonsoluble Mo Si-species which may cause dramatic losses in the 99Mo yield. Certain limited quantities of Mo-carrier can help solving this problem. The other way around would be to elute the Al2O3 column with higher-concentrated NH3 (2 M instead of 1 M) or with NaOH.
Molybdenum in its anionic form
The dimensions of a suitable DOWEX-1 column and its operation parameters are determined in a similar way as demonstrated for the Al2O3 column. For a column of about 26 × 120 mm, a linear flow speed of 13 cm/min is the maximum. If the volume of the Mo solution is 2000 mL one would need theoretically 54.6 g of the ion exchange resin. DOWEX-1 in the dry form. Considering the density of 0.65 g/mL resin, this would give an 84 mL volume of the resin. For rinsing the column, 4 bed volumes are required which correspond to 340 mL. Table
Optimal parameters for the DOWEX-1 column process.
DOWEX-1 column | About 26 mm × 120 mm, bottom G 3 frit |
Loading process: | 99Mo-solution in 1 M NH3 volume ~1.0 L, ~15 min |
(1) Wash process | 170 mL water, time about 3 min |
(2) Wash process | 170 mL water, time about 3 min |
Liquid waste volume: | ~1.4 L (depending on FEED and wash volume) |
Elution | 200 mL of 1 M (NH4)2CO3 solution, time about 10 min |
DOWEX column process: | Total time about 35–40 min |
99Mo elution profile from the DOWEX-1 column with 1 M (NH4)2CO3-solution using the parameters shown in Table
The purification step at the DOWEX column delivers 200 mL of the 99Mo molybdate in 1 M (NH4)2CO3 solution. In the following step this eluted solution is evaporated to the dryness in a special evaporator, with condenser. During evaporation, the (NH4)2CO3 is being decomposed; thus no additional salts are introduced into the final configured [99Mo] molybdate solution.
The residue is redissolved in the desired volume of diluted NaOH solution forming the final product solution [99Mo]Na2MoO4. This final product solution is then transferred into a corresponding plastic vial and transferred into the hot cell 3 for further processing, precise measurement and distribution.
A sublimation step (at 1000°C) is foreseen as an additional reserve for improving purity, if required. In this case the residue after evaporation shall be redissolved in diluted HNO3 or NH4OH.
Careful studies have been performed to obtain a full picture on the behavior of the 99Mo, of the most important impurities in 99Mo preparations, for other fission products, and for the target element itself. On one hand, tracer activities of 99Mo and 131I have been used, and after having optimized the separation conditions the same full protocol has been applied to study the separations technology with weak irradiated original target material (activity level ~4 GBq). Figure
Gamma spectra of samples from the precipitate, filtrate, FEED solution (filtrate after iodine removal), from the final product illustrating the different separation and purification steps (for more details see text).
The upper spectrum has been taken from a small fraction of the filter cake (precipitate) followed by the spectrum from a sample from the filtrate. It is clearly seen that only few gamma lines are left in the filtrate, which correspond to 99Mo, its daughter 99mTc, the radioiodine’s, and some fractions of Ru. The strongest signals in the precipitate (239Np, and 140La) are not seen in the filtrate solution. As said before the filtrate is passed through a silver-coated Al2O3 column prior to the acidification process. Thus comparing the spectra 2 and 3, one clearly sees that iodine is missing in the FEED solution (see also Figure
In total the precipitate collects more than 60% of the radioactivity formed in the nuclear process in the chemical form of hydroxides, oxides, or carbonates of the fission products. This corresponds to the fission products of Ba and Sr, the rear earth elements and actinides, Zr/Nb. Te and Sb are nearly quantitatively collected in the precipitate. The results of a corresponding tracer experiment, are summarized in Table
Radioactivity distribution between precipitate and filtrate after dissolving irradiated nat.U-targets of original composition.
Nuclide | Precipitate (%) | Solution (%) |
---|---|---|
239Np | 100 |
|
132Te | 100 | <3 |
143Ce | 100 |
|
141Ce | 100 |
|
140Ba | 100 |
|
140La | 100 |
|
131I | 0.60 | 99.39 |
103Ru | 20–40 | 60–80 |
99Mo | 0.57 | 99.43 |
Note: Ru behaves in different experiments differently, thus these data provide just an estimate.
The most important impurities have been followed up quantitatively throughout the process as good as gamma-spectroscopy could do under practical conditions with limited measuring time. The results are summarized in Table
Radioactivity distribution of 99Mo and the most important impurities during the process.
99Mo | 103Ru | 132Te | 131I | |
---|---|---|---|---|
Filter cake | <0.5% | 19.20% | 96.30% | n.d. |
|
|
|
|
|
Ag-column | n.d. | 22.70% | 1.00% | >98% |
FEED | >99% | 66.90% | 1.60% | <2% |
Al2O3 column | 10.6% | 0.10% | 1.10% | n.d. |
Al-waste | n.d. | 75.40% | n.d. | n.d. |
DOWEX column | 0.003% | 0.06% | n.d. | n.d. |
Final | 86.5% | <0.001% | <0.001% | <0.001% |
The % values relates to the individual content of the specified nuclide and not to 99Mo.
(n.d.: not detected).
The 132Te is nearly quantitatively coprecipitated. The highest 132Te-content in the filtrate was 3.7% of the original quantity. About half of this fraction is retained at the silver column, the other half fraction at the Alumina column. In the filtrate and wash solutions from the Alumina column the 132Te could not be detected any more (with the applied spectrometric parameters).
The iodine is nearly quantitatively retained at the silver column. The small fraction that is passing the silver column is then distributed throughout the system, mainly in the waste solutions through the washing procedures of the columns. Summarizing, the separation and purification process is efficient and easy.
The 99Mo Production Facility (MPF) is installed at PINSTECH Phase-1 building near reactor hall of PARR-1. The technical realization of the ROMOL-99 process in a semiautomated separation facility has been carried out by ITD Dresden GmbH (former Hans Wälischmiller (HWM) GmbH, Branch office Dresden). The main working areas of this facility are the Hot Cell complex (3 Hot Cells), interim liquid storage tanks, charcoal filter beds for iodine retention, xenon delay tanks, and the operator and service areas interconnected with it. Additionally, there are the so-called lock rooms through which the activated targets, the final product, and solid and liquid wastes are moved. Still, there are areas for personnel, preparation of reagents, storage, dosimetry, measurement, and decontamination. Further equipment in other rooms or buildings, which participate in the 99Mo production, is the existing equipment of the main exhaust system with filter chamber, Secomak blowers and the main exhaust blower. The spent target material (loaded filter plates enclosed in screw shut cans) is stored in Spent Fuel bay of PARR-1. The solid low-radioactive wastes (spent ion-exchange columns, tubes, interconnections, and other one-way materials) are stored, while decayed radioactive liquid waste is cementized in the radioactive waste management Group building.
More than 50 commercial batches of fission based 99Mo using ROMOL-99 process have been successfully completed. After the evaporation step, the residue is dissolved in the desired volume of diluted NaOH solution forming the final product solution [99Mo]Na2MoO4. This final product solution is then transferred to the PAKGEN 99mTc generator production site at PINSTECH. These generators are then distributed to the 35 nuclear medical centers in Pakistan. The performance of these generators is comparable to that of generators produced from imported fission 99Mo. The quality of the 99Mo preparations produced at PINSTECH corresponds to the required international standard (Table
Quality parameters of fission 99Mo produced at PINSTECH meeting international standard.
Gamma | 131I | ≤5 × 10−2 MBq/GBq 99Mo |
103Ru | ≤5 × 10−2 MBq/GBq 99Mo | |
Beta | 89Sr | ≤6 × 10−4 MBq/GBq 99Mo |
90Sr | ≤6 × 10−5 MBq/GBq 99Mo | |
Alpha | ≤1 × 10−6 MBq/GBq 99Mo | |
Other gamma | ≤1 × 10−1 MBq/GBq 99Mo |
The ROMOL-99 process allows dissolving UAlx/Al clad dispersion targets under reduced pressure conditions without generation of hydrogen at temperatures between 70 and 80°C. The technology implements the separation of NH3 and radio-iodine prior to the 99Mo separation. Generated nitrite is safely destroyed during the acidification process by urea to N2. The technical realization of the ROMOL-99 process in a semiautomated separation facility has been carried out by ITD Dresden GmbH (former Hans Wälischmiller (HWM) GmbH, Branch Office Dresden). More than 50 commercial batches of fission-based 99Mo using the ROMOL-99 process have been successfully completed at PINSTECH. PAKGEN 99mTc generators were prepared by using this locally produced high purity fission 99Mo and distributed to 35 nuclear medical centers in Pakistan. The performance of these generators is comparable to that of generators produced from imported fission 99Mo.
The authors declare that they have no conflict of interests.