Open Access

Production of novel diagnostic radionuclides in small medical cyclotrons

  • Mateusz Adam Synowiecki1,
  • Lars Rutger Perk1Email author and
  • J. Frank W. Nijsen2
EJNMMI Radiopharmacy and Chemistry20183:3

Received: 1 November 2017

Accepted: 24 January 2018

Published: 20 February 2018


The global network of cyclotrons has expanded rapidly over the last decade. The bulk of its industrial potential is composed of small medical cyclotrons with a proton energy below 20 MeV for radionuclides production. This review focuses on the recent developments of novel medical radionuclides produced by cyclotrons in the energy range of 3 MeV to 20 MeV. The production of the following medical radionuclides will be described based on available literature sources: Tc-99 m, I-123, I-124, Zr-89, Cu-64, Ga-67, Ga-68, In-111, Y-86 and Sc-44. Remarkable developments in the production process have been observed in only some cases. More research is needed to make novel radionuclide cyclotron production available for the medical industry.


Radionuclide productionCyclotronTechnetium-99 mRadioiodineZirconium-89Copper-64Gallium-67Gallium-68Yttrium-86Scandium-44


High-purity radionuclides are a key element in the development of radiopharmaceuticals for applications in nuclear medicine. Radionuclides are used in diagnostic and therapeutic radiopharmaceuticals. Often, they can be imaged by Single Photon Emission Computed Tomography (SPECT) or Positron Emission Tomography (PET). Over the last few decades, research groups in both the private and public sector conducted numerous development studies of new radiotracers. This resulted in a constantly evolving radiopharmaceutical market trying to satisfy the needs of the medical society. Many failed introductions of new radiopharmaceuticals resulted from the inability to stand out from the products that are currently available. This is due to little added value or a lack of clinical relevance. Those that are successful in manifesting their benefits then face another barrier. Zimmermann et al. described that the road of successful new radiopharmaceuticals is paved with failures because of the many industrial and regulatory constraints (Zimmermann, 2013). The clinically used radionuclides should satisfy requirements such as:
  • Physical properties suitable for its application (half-life (T1/2), decay mode, emission energy);

  • Chemical properties suitable for labeling, with high radiochemical yields and a high radionuclidic purity;

  • An acceptable trade-off between the dose received by the patient and the desired effect (image quality or treatment);

  • A reasonable price range.

On the other hand, from a radiopharmaceutical industry point of view, a radionuclide selected for production must meet certain criteria to be of interest:
  • The production efficiency is acceptable in terms of equipment and personnel burden;

  • The starting materials are not too expensive; the supply is secured and remains stable in the long term;

  • It offers a high molar activity;

  • It is expected to be carrier-free or no-carrier added in order to limit the toxicity;

  • Radionuclide production and the post-processing chemistry can be simplified, with a preference for automation (because of Good Manufacturing Practice (GMP), radiation safety and costs concerns);

  • Its half-life is suitable for logistics;

  • There is sufficient demand on the market.

It often happens that the above criteria from both sides are not met, which causes the radionuclide to never reach widespread application. In addition to that, other implications (such as political or economical changes) may discourage the production of certain radionuclides. This is typically the case when government allocated investments or decisions influence the distribution of sources of the radionuclide, e.g. by phasing out a nuclear reactor (Krijger et al., 2013). Currently most hospital radio-pharmacies derive their radionuclides from three types of sources: nuclear research reactors, radionuclide generators and cyclotron facilities (Table 1). It is worth mentioning that the generators still need a reactor or cyclotron source to produce the parent radionuclide. There are also other, much less common ways of producing radionuclides by using linear accelerators (Mang'era et al., 2015), Van De Graaf accelerators (Jones, Robinson Jr., and McIntyre, 1984) or lasers (Bychenkov, Brantov, and Mourou, 2014).
Table 1

Common types of radionuclide sources


Nuclear Reactors



Principle of production

Target material inserted in the neutron flux field undergoes fission or neutron activation transmuting into radionuclide of interest

Long-lived parent radionuclide decays to short-lived daughter nuclide of interest. Daughter nuclide elution follows in pre-determined cycles

Target material irradiation by charged particle beams. Inducing nuclear reactions that transmute the material into radionuclide of interest

Transmutation base



p, d, t, 3He, α or heavy ion beams


- Production of neutron rich radionuclides, mostly for therapeutic use

- High production efficiency

- Centralized production: one research reactor able to supply to large regions or in some cases globally

- Available on site, no need for logistics

- Mostly long shelf life

- Easy to use

- Limited radioactive waste: returned to manufacturer after use

- Production of proton rich elements used as β+ emitters for PET scans

- Decentralized production allows for back-up chains

- High uptime

- High specific activity in most cases

- Small investment in comparison to nuclear reactor

- Little long-lived radioactive waste


- Extremely high investment cost

- High operational costs

- Considerable amounts of long-lived radioactive waste

- Long out-of-service periods

- Trouble to back-up in case of unforeseen downtime

- Demanding logistics, often involving air transport

- Public safety concerns

- Non-proliferation treaty concerns

- Supplies in cycles according to possible elution frequency; in-house use must be timed accordingly

- Trace contaminants of long-lived parent nuclide in eluted product

- Regional network of cyclotrons and complex logistics needed for short-lived produced radionuclides

- Radionuclide production limited depending on installed beam energy

This review article focuses on recent developments in the production of novel diagnostic radionuclides by using small medical cyclotrons. Some of the presented radionuclides can potentially be used in theranostic applications, i.e. a combined diagnostic and therapeutic effect within single application of the radiopharmaceutical.

Cyclotron produced radionuclides

Classical cyclotron produced radionuclides are defined in this article as produced by very well-established technologies using mostly liquid or gas targets (Fig. 1). Typically, they are produced routinely on an almost daily basis and therefore present the bulk of targetry solutions offered by the cyclotron manufacturers. Among these classical radionuclides are: 18F, 13N, 11C and 15O. These four PET radionuclides are also commonly referred to as “standard” radionuclides in literature and are mostly produced using low energy medical cyclotrons. 123I and 111In are two of the cyclotron produced radionuclides used in SPECT studies that can be regarded as classical. However, these are produced by intermediate energy cyclotrons. The novel cyclotron produced radionuclides are defined as all others.
Fig. 1

Radionuclides used in nuclear medicine diagnostics

Production technology may already be well-established, like 64Cu or 124I, but what makes them interesting is that they are mostly either produced on an irregular basis or in singular centers across a large area and that their clinical application is not yet established. Most of them are produced using solid target systems with some of them having a proven history of robustness, while others are still under development or require further optimization. Typically, novel radionuclides have relatively longer half-lives which allows for shipment to distant users. The following novel cyclotron produced radionuclides will be discussed (Fig. 1): 99mTc, 124I, 89Zr, 64Cu, 67Ga, 68Ga, 86Y, 44Sc. Also, the classical radionuclides of 123I and 111In are discussed since there are possibilities to use small medical cyclotrons for their production.

Types of cyclotrons

Small medical cyclotrons (SMC) normally have a proton energy below 20 MeV (Table 2). In the literature, they are referred to as medical cyclotrons, PET cyclotrons or small-sized cyclotrons. They offer proton beam currents typically in the range of 60–100 μA. These accelerators are mostly installed in hospitals, universities and small-scale industrial radionuclide production plants. Many of these small medical cyclotrons have been installed over the last two decades and the number is still increasing. In 2008, almost 700 cyclotrons were installed worldwide (IAEA, Cyclotron Produced Radionuclides: Principles and Practice, 2008). Only seven years later, according to Goethals et al. (Goethals and Zimmermann, 2015), that number has increased to 1218 cyclotrons whereof approximately 1000 are SMCs (Table 2). Most of the SMCs are located in the developed countries, although newly industrialized developing countries are rapidly increasing their cyclotron base. The majority of these SMCs have been manufactured by four companies in order of market share (Schaffer et al., 2015): General Electric Healthcare (GE Healthcare), Ion Beam Applications (IBA), Siemens and Advanced Cyclotron Systems Incorporated (ACSI). There are several other companies producing small or intermediate energy cyclotrons that are developing steadily (Schmor, 2010). Typically, SMCs accelerate protons only, with some of them also capable of accelerating deuterons at half of the specified energy of protons. Their purpose is the production of medical radionuclides for in-house use, research and commercial purposes. In general, they are focused on the production of short-lived standard radionuclides for PET studies.
Table 2

Distinction of cyclotron types (Goethals and Zimmermann, 2015)

Cyclotron type

Energy Range (MeV)

Approximate number

Typical location

Small medical cyclotron (SMC)

< 20 MeV


- hospitals

- universities

- local commercial plants

Intermediate energy cyclotron

20–35 MeV


- regional commercial plants

- research institutes

High energy cyclotron

> 35 MeV


- research institutes

- cancer proton therapy centers

aExcluding proton therapy cyclotrons

Cyclotrons delivering protons of an energy between 20 and 35 MeV are considered intermediate energy cyclotrons or medium cyclotrons. Besides the main proton beam capacity, they usually offer a deuteron beam, and few of them offer an α beam. These machines tend to be located at bigger radiopharmaceutical commercial plants or research institutes. Obviously, they can also be used to produce the classical PET radionuclides, but their main purpose is the production of classical SPECT and novel PET radionuclides, or parent nuclides for generators.

Cyclotrons of particle energies above 35 MeV are considered high energy cyclotrons. They are scattered around the globe in well-regarded research institutes. These machines are tailored to specific research needs and can be designed to accelerate many kinds of particles: protons, deuterons, tritium, alpha and heavy ion beams. Many unique novel radionuclides can be produced, especially parent radionuclides for the generators, like 68Ge or 82Sr. High energy cyclotrons are also installed in large clinical cancer centers for proton beam therapy.

Cyclotron targetry

The most important irradiation parameters determining the formation of a radioactive product are the beam flux, energy and irradiation time as well as number of target nuclei, nuclear reaction cross section and half-life of the produced radioisotope (Krijger et al., 2013; Qaim, 2017). Standard radionuclides are mostly produced in either gas or liquid targets. These targets are very easy to use and require no manual handling for routine production activities. The fluidic nature of the irradiated element makes rapid heat exchange possible. Provided water cooling has the potential to remove the heat produced by the intensity of the beam up to the maximum specification of the cyclotron. Often, helium cooling is used for the beam window, i.e. the target element through which the beam enters the isolated target material within the target body. Just after End of Bombardment (EOB), the unloading process is performed by pressurized inert gas that pushes the produced radioactivity through small bore tubing to the automatic synthesis module. Therefore, the gas and liquid target methods are safe, reliable and fast.

Solid targets

For novel radionuclides, solid targets are often used. Solid targets present a number of difficulties when compared with gas and liquid targets:
  1. 1.

    Often, expensive enriched target materials must be used to produce radionuclides with a high purity;

  2. 2.

    The heat conductivity is much lower, which can lead to overheating problems. Care must be taken to optimize the cooling system and beam parameters to avoid melting of the target. Usually, an inclined target design is proposed with an angle of a few degrees between the incident beam and material layer. This ensures that the entering beam power heat is dissipated over a bigger target area (Fig. 2);

  3. 3.

    The target material layer should be a few hundred μm thick and placed on the back plate made of highly conductive metal. This process requires material deposited isotropically and in a pure form (or sometimes using various oxides). Typically, electrodeposition or powder pressing must be used, followed by sintering. Other methods include powder rolling, laser plating and forming high melting point alloy (Qaim Syed, 2011; Stolarz et al., 2015);

  4. 4.

    Solid state targets require more manual handling which causes a higher radiation hazard for personnel, especially when the target must be retrieved shortly after irradiation. Automation of the delivery line is possible with pneumatic systems;

  5. 5.

    Solid target material in general requires more complex chemical separation steps and further recycling steps.

Fig. 2

Examples of solid targets. a COSTIS (Compact Solid Target Irradiation System) IBA Nirta target, b Custom developed Zirconium-89 target, images courtesy of Roel Mooij, BV Cyclotron VU, Amsterdam, NL c Custom developed inclined solid target, black line depicts beam direction from the cyclotron, cross section visualizes cooling channels (IAEA, Cyclotron Produced Radionuclides: Principles and Practice, 2008)

To overcome overheating problems, low beam currents are used which results in low production yields. Therefore, the beam time is increased to reach useful amounts of radionuclides. Together with chemical and technical challenges, this makes the whole solid target production process time consuming, labor intensive, hazardous in terms of radiation safety and thus expensive.

Many of the solid targets for production of novel radionuclides had to be developed and optimized in-house, which resulted in various and incomparable performances and run parameters when collected between the centers. Design possibilities of those targets often encountered problems related to space limitation caused by cyclotron or target shielding, vault dimensions or technical installations. Nowadays, most cyclotron manufacturers have standardized compact solid target systems available for more popular novel radionuclides. These targets offer easier maintenance, have high automation capabilities and are supplied as a plug-and-play system.

Custom targets have the potential to reach higher yields in comparison to the standard off the shelf compact target (systems). To our opinion this is due to the goal of the custom target developers design their target to be better optimized for their specific conditions of usage and to maximize the yields. On the other hand cyclotron manufacturers with their compact targets, aside for maximizing yields, aim to assure high level of safety, easiness of handling for personnel and introducing more replaceable materials. This situation may create a “trade-off” dilemma for a radionuclide production plant operating a SMC which technology to decide on for a new solid targetry based production.

Liquid targets for radiometals production

A recent development in the production of radiometals involves the use of solution targets (i.e. liquid targets), where the target material is dissolved in an aqueous solution in the form of salts. This method removes the drawbacks that are mentioned for solid targets and offers the following advantages: simple handling, on-demand availability, a faster production process without dissolution and adjustable use of enriched material depending on the needs for a single production, which greatly reduces the cost (Alves et al., 2017). However, the production of radiometals using liquid targets comes with a number of disadvantages, most noticeable the significantly lower production yields. To optimize these methods, certain problems need to be addressed before large-scale implementation.

One of them is the target pressure build-up during irradiation. This effect depends on the beam intensity and concentration of the target material. The irradiation of aqueous solutions leads to water radiolysis and the creation of ions and free radicals of hydrogen, oxygen and hydroxyl groups, which in turn leads to rapid gas evolution in the target chamber (H2 and O2). This effect is further promoted by the introduction of certain types of salt cations and anions. Those effects were studied and minimized in a recently published patent invented by DeGrado et al. (DeGrado, Pandey, and Byrne, 2017). In particular cases, the introduction of strong nitric acid proves to minimize the gas evolution in the target because the nitric acid acts as a free radical scavenger. Another solution is the introduction of a backpressure regulator which keeps the in-target pressure at a stable level during irradiation.

The authors also point out the importance of carefully selecting salt constituents as certain pairs of metallic cations and acidic anions affect the gas evolution in different ways. Additionally, strong acids used in targetry may result in corrosion of the targets. Other experiments using silver or aluminum targets demonstrated the presence of fine particles, leading to clogging of the transfer lines (Hoehr et al., 2012) and thus requiring frequent target maintenance. It is recommended to use heat-resistant and chemically inert materials (such as niobium and tantalum) as target materials. Optimization of target material, metallic salt composition and concentration of acid are all needed to prevent the formation of precipitates. Favorably, the addition of nitric acid also eliminates the precipitation of salts within target (DeGrado et al., 2017).

Pandey et al. show high ratios (approximately 85–90%) of enriched material recycling for liquid targets (Pandey et al., 2014a), which is a highly desirable and financially advantageous factor for considering the use of solution targets.

In general, the liquid target radiometal production technology is not yet mature and requires more studies to optimize production yields. For small facilities and/or facilities at remote locations, the recent liquid target developments allow for in-house production of radionuclides with small, shelf-shielded cyclotrons.

Developments in radionuclide production

Below, the current status and recent developments in the production of a number of novel radionuclides will be presented.


Technetium-99m (99mTc) is still undeniably the most commonly used radionuclide in the world. Its share in all nuclear medicine procedures is approximately 80%, with an estimated 40 million patient preparations per year worldwide (OECD-NEA, 2017). Until recently, almost all of the 99mTc used in nuclear medicine was produced by radioactive decay of 99Mo confined in a generator. 99mTc success in the clinic is based on advantageous physical properties such as: a moderate half-life (T1/2 = 6.0 h), and low energy 99% γ emission of single peak of 140.5 keV. Moreover, other beneficial factors for its widespread use are: a low price, availability, and the fact that it is a generator product. Its parent radionuclide, 99Mo, is dominantly produced by a nuclear fission process of 235U(n,f)99Mo in a few aging high neutron flux nuclear reactors running on Highly Enriched Uranium (HEU). The threat of possible shortages in the near future due to imminent closure of those facilities encouraged many research groups to investigate alternative methods for the production of 99mTc without the use of a nuclear reactor (Table 3). Such studies are supported by the International Atomic Energy Agency (IAEA) through the project: “Accelerator based Alternatives to Non-HEU production of Mo-99/Tc-99m” (Accelerator-based Alternatives to Non-HEU production of Mo-99/Tc-99m, 2011).
Table 3

99mTc production pathways



Currently available

Status and further development

235U(n,f)99Mo → 99mTc

Reactor + generator


Well-established, availability will shrink with phasing out of nuclear research reactors


Small medical or intermediate energy cyclotron

In Canada

Possible worldwide implementation with decreasing nuclear reactor capacity

96Zr(α,n)99Mo → 99mTc

α beam cyclotron + generator


Will not be implemented. Method not competitive, α-beam required, low yields

100Mo(γ,n)99Mo → 99mTc

LINAC+ generator


Under development in Canada, USA and the Netherlands

98Mo(n,γ)99Mo → 99mTc

Reactor + generator

In USA and Japan

Auxiliary method used in nuclear reactors (Blaauw et al., 2017; van der Marck, Koning, and Charlton, 2010). Availability will shrink with phasing out of nuclear research reactors




Theoretically feasible. Further research required

The concept of producing sufficient amounts of 99mTc from enriched 100Mo targets using a cyclotron was known from the early 70’s (Beaver and Hupf, 1971). It was not developed further, since the current capacity of fission produced 99Mo for usage in handy 99Mo/99mTc generators was sufficient for years to come. In addition, the nuclear reactor network was still expanding. In the 90’s, several researchers again started the investigation of the nuclear data for proton induced reactions for molybdenum radionuclides, most notably 100Mo(p,2n)99mTc. The results renewed interest in the production of 99mTc using cyclotrons and set the optimum production energy using protons in the range of 13.5–17 MeV, with a recommendation to avoid higher energies resulting in the production of inseparable impurities such as 98Tc, 97Tc and 96Tc (Manenti et al., 2014; Qaim et al., 2014; Takacs et al., 2016).

Target design became the next milestone for successful cyclotron production of 99mTc, resulting in thick molybdenum coatings for solid targets able to withstand prolonged irradiation of high beam currents accompanied by an easy recovery of 99mTc and recycling of 100Mo after end of beam (Stolarz et al., 2015; Hanemaayer et al., 2014). The worldwide leading group for the targetry, beam optimization and ion source studies are undoubtedly Canadian researchers gathered under the auspices of the TRIUMF centre (Schaffer et al., 2015; Hanemaayer et al., 2014; Buckley, 2013; Benard et al., 2014). For the production of technetium-99m, they used solid targets of enriched molybdenum-100 coatings which resulted in a production yield of up to approximately 513 MBq/μAh. This can lead to 350 GBq of 99mTc on an ACSI TR-19 cyclotron (6.9 h beam time, 300 μA beam, cyclotron beam energy lowered to 18 MeV protons) and about 170 GBq on GE PETtrace 880 (6 h beam, modified to 130 μA, 16.5 MeV protons). In the case of 18 MeV proton energy irradiations, which are slightly above the recommended energy range for 99mTc production, the target contained approximately 95% of 99mTc with 99.5% radionuclidic purity at EOB. The separation efficiency of 99mTc from a molybdenum target, which can be easily automated, was proved to be in the range of 80% - 90%. With the radiochemical purity of 99.7% of the final pertechnate [99mTc]TcO4 solution, the same USP quality requirements (> 95%) as for 99Mo/99mTc generator produced pertechnate are met. The final efficiency of 100Mo recycling is above 90%. Other research groups have produced data on similar experiments that concur with these results (Das et al., 2016; Rovais et al., 2016).

Such capacity can fulfill a daily demand for a large area, thus encouraging the Canadian government to co-finance and support The Canadian National Cyclotron Network project, run by ACSI, which will cover 100% of Canadian needs for 99mTc (ACSI, n.d.). The project was scheduled to be finalized before closure of the Chalk River nuclear reactor, once one of the leading suppliers of 99Mo, now operating as a back-up plant until its planned closure in March 2018. The project is ongoing, facing several delays caused by regulatory approval issues and logistics considerations (Brown, 2016), but some Canadian nuclear medicine patients are already scanned with cyclotron produced 99mTc.

An interesting alternative pathway of cyclotron production of 99Mo has been theoretically examined by a group from the Italian National Institute for Nuclear Physics (Pupillo et al., 2015). They use an α-beam on an enriched Zirconium-96 target, thus inducing a 96Zr(α,n)99Mo nuclear reaction. The above pathway does not produce sufficient amounts of 99Mo for commercial use. However, it has the advantage of an extremely high specific activity in the range of 106 TBq/g (current large generators have a specific activity of up to 370 TBq/g). A limiting factor for the development of this pathway is the unavailability of high current α beams.

Alternative methods can utilize linear accelerators or lasers. The production of molybdenum-99 by utilizing electron induced brehmsstrahlung photons produced in linear accelerators is of particular interest. These targets require the same molydbenum-100 enriched material, but the nuclear reaction pathway is 100Mo (γ,n)99Mo. Recent experiments in this field confirmed that this production pathway is feasible (Mang'era et al., 2015) and a steady supply of 99Mo is expected to be provided soon by the already commissioned 35 MeV, 40 kW Linear Accelerator (LINAC) at the Canadian Light Source. Similar projects based on the same principle are under development or consideration in other countries. NorthStar Medical Radionuclides, USA, is currently finalizing the 99Mo production method from the neutron capture 98Mo(n,γ)99Mo using an old nuclear reactor at the University of Missouri Research Reactor Center (MURR). The company plans to expand its capacity by building a single site LINAC farm, comprised of up to 16 LINACs capable of producing 99Mo by a photonuclear reaction on enriched 100Mo targets. The project is estimated to cover half of the USA demand for 99mTc (NorthStar Medical Radioisotopes Receives $11.75 Million from National Nuclear Security Administration, 2015; Harvey, Isensee, Moffatt, and Messina, 2014). The Lighthouse Consortium, Netherlands, is gathering companies and research institutes under the leadership of ASML, to achieve their goal of building an electron accelerator facility for 99Mo production from enriched 100Mo targets. It is assumed that by 2021 the facility will be able to cover the capacity of the HFR nuclear reactor in Petten (Lighthouse: productie medische isotopen vanaf 2021, 2017). In Japan, considerations on which technology to rely on for securing the future 99mTc supply are ongoing. The country already has a greatly developed particle accelerator infrastructure, allowing to diversify between production from linear accelerators, a vast cyclotron network of appropriate energies or even by using the Japan Proton Accelerator Research Complex (J-PARC) spallation facility (Nakai et al., 2014; Fujiwara et al., 2017).

Another alternative is the use of direct laser light that generates proton or electron fluxes further impinging on final targets. The disadvantage of this method is that the particle fluxes are not mono-energetic, which complicates yield calculations for side products. For this reason, a laser facility would require careful optimization of the laser target-nuclear target setup as a whole, to obtain a minimum number of co-produced impurities. Theoretical work in this field has been done with promising results (Bychenkov et al., 2014), yielding 300 GBq of 99mTc with 0.12% radionuclidic impurities in a 6 h irradiation assuming the utilization of the future International Coherent Amplification Network (ICAN) laser concept. However, in comparison to earlier mentioned advancements of particle accelerators, current laser-based 99mTc production research and infrastructure is not developed enough for short-term considerations.

Radioiodines I-123, I-124

Radioiodines have a long history of usage with changes in favoring different radionuclides throughout the last century (Silberstein, 2012). Today, the widely used 131I therapeutic agent (T1/2 = 8.02 d, 100% β, 90% βav 192 keV, 82% γ 364 keV) is produced in nuclear reactors. 131I is also used for SPECT imaging thanks to its low energy gamma emission.

Iodine-123 is the second most widely used radioiodine and also the second most used imaging agent after 99mTc. Its popularity comes from its availability, perfect γ energy for imaging and appropriate half-life for metabolic studies (Park, 2002). 123I (T1/2 = 13.2 h, 100% EC, 83% γ 158 keV) and 124I (T1/2 = 4.18d, 100% EC+β+, 12% β+av 687 keV, 11% β+av 975 keV, 63% γ 603 keV) are cyclotron produced and broadly used for all kind of applications (Silberstein, 2012; Koehler et al., 2010). 123I is considered a classical radionuclide because of its widespread availability, well-established production method and routine production. Yet it remains an expensive endeavor since the production route utilizes proton irradiations of enriched xenon-124 gas (50,000 $/l as of 2008 (Kakavand et al., 2008)) in closed systems. This method employs two parallel nuclear reactions pathways and the following decay to 123I: 124Xe(p,2n)123Cs → 123Xe → 123I and 124Xe(p,pn)123Xe → 123I. It also requires an incident proton energy range between 20 and 30 MeV, which applies to a much smaller group of intermediate energy range cyclotrons. Such a produced final 123I radionuclide has an excellent radionuclidic purity of 99.9%. It is included in this review due to the availability of an alternative production pathway using enriched Tellurium targets (see later in this section).

Iodine-124 has seen its rise in the first decade of the twenty-first century. Although having a disadvantage of a high energy prompt γ which results in a high patient radiation dose and complicated dose calculations. It emits β+, thus this isotope of iodine is fit for PET scans (Silberstein, 2012). Additional Auger electron emission gives it the capability to be named a theranostic agent. Iodine-124 production data is vast since many alternative nuclear production pathways exist, encompassing a big range of reactions based on: 123-126Te(p,xn)124I, 123,124Te(d,xn)124I, 121,123Sb(α,xn)124I and 123Sb(3He,2n)124I nuclear reactions. This irradiation data has been extensively evaluated (Koehler et al., 2010; Braghirolli et al., 2014; Azizakram et al., 2016; Aslam et al., 2010), with the result being an agreement found on the best approach to produce 124I by means of an 124Te(p,n)124I reaction in the range of 14–7 MeV proton energy, with a yield reaching 21 MBq/μAh and minimum 125I impurities.

Both 123I and 124I have the great advantage of proton reaction cross section having a nuclear threshold below 10 MeV on Tellurium targets (Soppera, Bossant, and Cabellos, 2017a). This gives the possibility to produce them with most cyclotrons. The developments in the last 20 years focused on the design of more efficient solid target systems (Kakavand et al., 2008; Mahunka et al., 1996; Al-Yanbawi and Al Jammaz, 2007; Nagatsu et al., 2011; Qaim et al., 2003; Poniger et al., 2012). TeO2 based thin layer targets can be used for the production of both iodines using the same target, by changing only the tellurium target isotope to match the reactions 123Te(p,n)123I or 124Te(p,n)124. The use of 123I and 124I remains limited because the raw material cost of enriched tellurium (recyclable) remains high: $10–50/mg (Fonslet and Koziorowski, 2013) which translates directly to the radionuclide price.


Zirconium-89 (T1/2 = 78.4 h, 100% EC+β+, 23% β+av 396 keV, 99% γ 909 keV) has been proven very useful in drug development, for instance for new antibody therapies. This is because of its:
  1. 1)

    long half-life that is suitable for studying the biodistribution of long-circulating proteins and antibodies,

  2. 2)

    reproducible applicability in chelating chemistry and conjugation with monoclonal antibodies (mAbs) used in ImmunoPET studies (Vosjan et al., 2010; Ikotun and Lapi, 2011; Rice et al., 2011),

  3. 3)

    balanced physical properties, i.e. sufficiently small β+ energy to maintain good image resolution and acceptable patient dose levels.


Zirconium-89 was associated with a low radionuclidic purity of the recovered product from Yttrium-89 target material and chelation chemistry (Holland, Sheh, and Lewis, 2009). The research group of the Vrije Universiteit (VU) University Medical Center in Amsterdam (NL) simplified the method which resulted in reproducible radiolabeling encouraging the worldwide development of the production and use of 89Zr (Vosjan et al., 2010).

89Zr can be produced in cyclotrons by 3 nuclear reaction pathways:
  1. 1)


  2. 2)


  3. 3)



The great advantage of producing 89Zr from the latter two reactions is the fact that naturally occurring Yttrium is 100% composed of the 89Y radionuclide (thus natY = 89Y). This significantly reduces the cost and availability of the target material in comparison to other radionuclides that require highly enriched materials. The first reaction (natSr(α,xn)89Zr) requires a rarely available α-beam and is also prone to producing impurities from a differentiated strontium isotopic composition (unless enriched) so this pathway is limited to theoretical considerations.

The second reaction (89Y(d,2n)89Zr), although proven to give high yields both theoretically and experimentally (Sadeghi, Enferadi, and Bakhtiari, 2012; Tang et al., 2016), requires a relatively high energy deuteron beam since its reaction threshold starts at 5.9 MeV and peaks in the range of 13–17 MeV (Soppera, Bossant, and Cabellos, 2017b) which excludes most common small medical cyclotrons. Although the GE PETtrace 800 series is capable of producing a deuteron beam with 8.4 MeV and the IBA Cyclone 18/9 is capable of producing deuterons of 9 MeV, it is too low to yield reasonable amounts of 89Zr. Therefore, this pathway is reserved mostly for research centers possessing high energy, multiple beam type cyclotrons.

The third reaction (89Y(p,n)89Zr) is the only way for a small medical cyclotron facility to provide 89Zr. With radiochemical methods developed, recent worldwide research focused on target development to increase the production yields. One of the challenges is the limitation of beam energy, since above the reaction threshold of 13.08 MeV, production of long-lived 88Zr(T1/2 = 83 d) occurs via the (p,2n) reaction, which is an impurity inseparable from the final product. Dabkowski et al. (Dabkowski et al., 2015) used a compact IBA COSTIS target, designed to produce 89Zr, to show that above the reaction threshold of 11.6 MeV there is a co-production of another contaminant 88Y (T1/2 = 107 d) via 89Y(p,pn)88Y. The yttrium atoms must be separated in the following chemistry process minimizing this amount. Below 10 MeV proton beam energy, the authors reported a yield of insufficient quantities of 89Zr. This leaves a narrow beam energy window for practical use with small medical cyclotrons. For 11.6 MeV proton beams, the yields were about 14–16 MBq/μAh with maximum beam currents of 30 μA in 3–3.5 h irradiation time. This gives approx. 1.4 GBq of 89Zr at EOB with a high radionuclidic purity.

In the case of custom made targets, it has been reported that one can achieve higher yields than this, while maintaining high radionuclidic purity of the final product (Sadeghi et al., 2012; Tang et al., 2016; Alfuraih et al., 2013). Custom made yttrium target developments published in recent years (Siikanen et al., 2014; Ellison et al., 2016), which encompass welding yttrium foils, modifications in yttrium foil thicknesses and water cooling, report applicable designs capable of delivering yields of up to 49 MBq/μAh with maximum currents of 45 μA, and a 90% separation efficiency. This gives about 1.4–2.2 GBq of radionuclidic pure 89Zr in 1–2 h beam time, thus effectively shortening the irradiation times on the cyclotron or increasing the production capacity.

Production of 89Zr using solid targets with a small medical cyclotron might be challenging because of workspace limitations. Therefore, Pandey et al. (Pandey et al., 2016) and DeGrado et al. (DeGrado et al., 2017) have studied a few cases of 89Zr production via proton irradiation of liquid targets, filled with a Y(NO3)3/HNO3 solution, and corresponding chemistry. So far, their development has a much lower yield than the solid target technologies, reporting 4.4 MBq/μAh with a 40 μA beam current irradiated for 2 h. This means approx. 370 MBq of 89Zr activity with > 99% radionuclidic purity which is sufficient for a very small range of applications (preclinical or a few patients on-site) and in agreement with previous work on liquid target production of 89Zr (Oehlke et al., 2015).


Copper-64 (T1/2 = 12.7 h, 42.5% EC, 18% β+av 278 keV, 39% βav 190 keV, 0.5% γ 1346 keV) has seen its rise starting in the late 90’s and continuing in the first decade of the twenty-first century. It is one of the commonly used (mostly in the USA) non-standard radionuclides in PET imaging thanks to several advantages (Ikotun and Lapi, 2011). Its half-life allows for shipping to distant centers, its low positron energy and little γ emissions make the image resolution high, and high-yield and high-purity production methods are developed. Moreover, 64Cu has an unexplored theranostic potential, since its positron emission is accompanied with β and abundant auger electron emission. Uses of chelated 64Cu complexes are many, such as small molecules, peptides and mABs. However, the latest developments of 89Zr possess the potential to take over these applications because of a longer half-life that better matches the biological half-life of the mAbs and it is cheaper to produce.

Production of 64Cu is very well established. It is most commonly produced by cyclotrons utilizing a proton beam impinging on an enriched nickel-64 solid target reacting through a 64Ni(p,n)64Cu channel. The reaction threshold is 2.5 MeV and the highest yields are achieved in the proton energy range of 10–15 MeV, well within the energy range of a small medical cyclotron. There are commercially available solid targets, based on electroplating 64Ni on gold discs, which dominate the 64Cu production market. In optimal situations, this method can produce up to 185 GBq (Ikotun and Lapi, 2011) after a long beam time of up to 8 h and currents up to 40 μA with various target thicknesses and proton energies applied, but typically lower yields are enough to supply local area customers for a few days (Avila-Rodriguez, Nye, and Nickles, 2007; Matarrese et al., 2010; Szelecsényi, Blessing, and Qaim, 1993). With radionuclide production secured, recent developments in 64Cu focus on successful automation of the radiochemistry (Poniger et al., 2012; Matarrese et al., 2010; Ohya et al., 2016) and developments of reliable tracers carrying 64Cu (Jalilian and Osso, 2017).

The above method, even though well-established and giving enough high purity material, is not as common as one would expect. It has a number of disadvantages (Alves et al., 2017; Jalilian and Osso, 2017):
  • 64Ni has a very low abundance (0.95%), which requires significant enrichment before irradiations can be performed, making it an expensive method;

  • It requires on-site electroplating equipment and recycling technology;

  • Maintaining the suitable quality of electroplating (surface roughness) during long electroplating processes (6 - 48 h) requires careful handling (Rajec et al., 2010);

  • Long irradiation times are needed occupying the accelerator;

  • The operators can receive a high radiation dose unless an automated process is developed.

Thus, other methods were searched for in order to simplify the whole process. One of the options involved replacing the production pathway with 67Zn(p,α)64Cu reaction, which was investigated by Szelecsényi et al. (Szelecsényi et al., 2014). Even though it was proven that this method is feasible for the commercial production of 64Cu, it offers radiochemical purity challenges and poses the same concerns regarding high enrichment of the solid target material.

The latest most interesting development, concerns producing 64Cu in a liquid target as described by Alves et al. (Alves et al., 2017). Such a change would allow for a faster, safer and simplified operation with automated loading and transfer to an accompanied automated chemistry module. The group conducted a series of experiments with enriched 64Ni dissolved in the novel IBA liquid target made of niobium. After an irradiation of 5 h, 4.6 GBq of 64Cu at EOB was produced, which translates to a yield of 0.14 MBq/(μAh·mg). The authors claim that this can easily be improved by increasing the enriched 64Ni concentration and beam time to achieve up to 25 GBq of 64Cu. The subsequent automated purification gave radiochemically pure 64CuCl2 with a 77% decay corrected yield. This amount could satisfy local needs.

Gallium-68 and Gallium-67

Over the past several years, Positron Emission Tomography (PET) imaging agents labeled with gallium-68 (68Ga) have undergone a significant increase in clinical utilization. 68Ga is conveniently produced from a germanium-68/gallium-68 (68Ge/68Ga) generator. Besides gallium-68, two other gallium radionuclides are useful in nuclear medicine applications, namely 66Ga and 67Ga. Only 67Ga, besides 68Ga, is used routinely in clinical applications. 66Ga (T1/2 = 9.5 h, 44% EC+β+, 51% β+av 1904 keV, 37% γ 1039 keV, 23% γ 2751 keV) is rarely used because of its high positron energy which lowers the resolution of images combined with a high patient dose as it emits multiple γ rays of energies above 1 MeV. Currently, sufficient base of evaluated nuclear data exists for the production of these radionuclides (Aslam, Amjed, and Qaim, 2015).

68Ga (T1/2 = 67.7 min, 100% EC+β+, 88% β+av 836 keV, 3% γ 1077 keV) became the radionuclide of particular interest. Its half-life is relatively short, but it is often irrelevant since the majority of 68Ga is available at nuclear medicine departments from 68Ge/68Ga generators (Germanium has a half-life of 271 days) and the short half-life lowers the patient dose. Together with well-known radiochemistry and kit-based preparation methods, 68Ga development accelerated, finding application mostly in peptide-based tracers, antibodies and small research molecules (Jalilian, 2016). It is also a possible candidate for further theranostic use. Many 68Ga-labelled tracers are used in clinical trials and some are already approved by local authorities: 68Ga-DOTATATE (approved by the US Food and Drug Agency in the USA) and SomaKit TOC™ (approved by the European Medicines Agency) (Brief, 2016; Applications, 2016). For neuroendocrine tumor imaging in the UK, Gallium-68 based PET scans were recently advised as a replacement for 111In based SPECT scans (Kalsy and Vinjamuri, 2016).

Radiochemically pure Germanium-68 is produced via several alternative routes, mostly using a nat,69Ga(p,xn)68Ge nuclear reaction on higher proton energy cyclotrons of > 20 MeV, capable of delivering beam intensities of up to a few hundred microamperes for up to a few weeks of irradiation time (IAEA, 2010). This requires a robust and complex solid target design that can handle a high heat transfer for prolonged periods, as well as reliable purification of the material once the irradiation is complete. The rising need for 68Ga radionuclide fuels developments in this area (Fitzsimmons and Mausner, 2015a; Fitzsimmons and Mausner, 2015b; Sounalet et al., 2014; Bach et al., 2013).

68Ge/68Ga generators have worldwide coverage. Still, small cyclotron production of 68Ga is possible by bombardment by protons on an enriched 68Zn or natZn target (Engle et al., 2012) since the nuclear reaction of 68Zn(p,n)68Ga has a high cross section of up to 1 b in the energy range of 11–14 MeV. However, this method cannot compete with the 68Ge/68Ga generators when it comes to issues such as:
  • complex solid target handling and development;

  • the need for enriched material;

  • acceptance of inseparable impurities of 66Ga and 67Ga.

The comparison looks different when it comes to direct medical cyclotron productions using liquid targets, allowing for fast, small scale on-site productions of radiochemically pure 68Ga. Such a method, based on a 68Zn(p,n)68Ga reaction, would be of particular benefit for hospital cyclotron facilities with spare capacity to add a 68Ga-tracer to their routine operations. To achieve a high purity of 68Ga from direct irradiation of enriched 68Zn, the proton beam energy must be optimized since the threshold for producing the 67Ga contaminant from competing 68Zn(p,2n)67Ga is 12 MeV, so in the peak area of the 68Ga production cross-section. Works toward achieving liquid target technology were initiated by Pandey et al., the same group that was mentioned before for the development of Zr-89 liquid target (Pandey et al., 2014b). They proposed a dissolution of enriched Zn68 target material in nitric acid. The authors claim that they achieved a cheap method for producing small quantities of radiochemically pure 68Ga for 2–4 patients in a beam time of approx. 1 h. Furthermore, a recent study using a novel IBA target was carried out by Alves et al. (Alves et al., 2017), the same research publication that described the possibilities of liquid target 64Cu production. The results look very promising. A 45 min irradiation of 30 mg/ml of 68Zn with a 45 μA proton beam yielded 6 GBq of 68Ga, which can be translated to achieving batches of 40 GBq of 68Ga (pre-purification) by optimization of concentration and beam parameters. This can create a viable alternative for solid target and generator produced 68Ga. Besides general advantages of using a solution target instead of solid target, no long lived 68Ge impurities were found in the final product. Authors of both above mentioned works also motivate their studies with the need to prevent 68Ge breakthrough that occurs in generators, but this argument seems less relevant since today’s modern generators have included mechanisms that nullify that effect (Roesch, 2012a).

Gallium-67 (T1/2 = 78.3 h; 100% EC; 39% γ 93 keV, 21% γ 185 keV, 17% γ 300 keV) is a less widely used SPECT imaging agent, but has a potential for broader usage. Its most notably used tracer worldwide is 67Ga citrate in various inflammatory studies leading to the detection of tumors and infections (Cwikla et al., 1999; Jalilian et al., 2009). 67Ga could also become a therapeutic agent in the future due to the emission of Auger electrons.

Several possible pathways for the production of 67Ga exist from proton to deuteron (Tárkányi et al., 2004) and alpha beams. The most commonly used method is proton irradiation of solid targets of either enriched or natural zinc, inducing a 68Zn (p,2n)67Ga reaction (reaction threshold 10.3 MeV) and a 67Zn(p,n)67Ga reaction (reaction threshold 1.8 MeV) (Aslam et al., 2015; Asad et al., 2014). 68Zn(p,2n)67Ga requires a higher energy cyclotron as the reaction cross-section peaks around 21 MeV. It will inevitably lead to the co-production of large amounts of 68Ga, creating the need to wait until this impurity level is reduced to acceptable limits. 50 GBq was reportedly produced in one batch with this method (Qaim, 2012). Since 67Ga is the longest lived of gallium radionuclides and waiting time for removal of impurities is applied, this method can also use natural zinc targets, considerably reducing costs and employing both reaction pathways. Such a method was presented by Martins et al. (Martins and Osso, 2013) where the authors investigated a new 67Ga purification technique, resulting in radionuclidic pure (99.9%) 67Ga production yields of 40 MBq/μAh with a 26 MeV proton beam on a natural zinc solid target. One must also consider unavoidable losses connected to the extraction efficiency (72%) and decay period of 3 days.

Channel 67Zn(p,n)67Ga offers an alternative for use in small medical cyclotrons that do have enough proton beam energy. To make it efficient, one must consider expensive enrichment since the natural abundance of 67Zn is very low (4%). Staying below the 13 MeV production threshold for 66Ga will not require unnecessary decay-out time for the by-products. Even though this method is theoretically feasible, it is not commercially implemented. To our opinion, this is due to the fact that the long half-life of 3.3 days allows for an easy distribution of this nuclide. A single intermediate energy cyclotron may easily cover regional needs, so implementing a time-consuming, solid, enriched target 67Ga production process on a small medical cyclotron has little economical justification. Perhaps with further increasing demand for this nuclide, small scale production sites will be needed.


Indium-111 (T1/2 = 67.31 h; 100% EC; 94% γ 245 keV, 91% γ 171 keV) is one of the classical radionuclides used in SPECT studies, thanks to its favorable low-energy and high-intensity emissions, and half-life suitable for in vivo studies. Possible medical uses are enormous (Lahiri, Maiti, and Ghosh, 2013), with monoclonal antibodies labeling, blood cell labeling for migration studies, tumor imaging, diabetes studies and more. 111In can also be considered an Auger electron therapy agent (Qaim, 2012). 111I is not only used in nuclear medicine, but also in material science(Lahiri et al., 2013).

Currently, the most common production route for 111In is via proton irradiation of highly enriched Cadmium-112 solid targets, through reaction pathway 112Cd(p,2n)111In. It is carried out by intermediate energy cyclotrons with protons of an energy range of 25 MeV, with the peak reaction rate around 20–22 MeV (cross section approx. 1000 mb). This method offers high yields (248 MBq/μAh) with co-produced impurities of 112Cd and 112Sn, which are later separated. Batches of 50 GBq are reportedly produced on a regular basis (Qaim, 2017).

Despite being a well-established classical cyclotron produced radionuclide, it can be produced using novel methods employing SMC. From this point of view, only nuclear reaction 111Cd(p,n)111In is feasible (Alipoor, Gholamzadeh, and Sadeghi, 2011). It peaks at 15 MeV with approximately 800 mb reaction cross section. A theoretical yield of 67.5 MBq/μAh is expected, which is 4 times less than the one used in the intermediate energy range cyclotrons. An advantage is that up to 20 MeV no considerable co-production of isotopic impurities is expected. However, no literature on such experimental production attempts were found. It is doubtful that this method would find enough interest among small medical cyclotron users, as the radionuclide is readily available in sufficient amounts.


Yttrium-86 (T1/2 = 14.7 h, 100% EC+β+; 12% β+av 535 keV, 6% β+av 681 keV, 83% γ 1077 keV) can be considered a very special PET radionuclide. If it was not for the widely used therapeutic agent 90Y (T1/2 = 2.67d, 100% βav 934 keV), 86Y would probably not receive any special attention. Its positrons have relatively low intensity and high energy for imaging in proper resolution. However, there is a big advantage of being the surrogate of the therapeutic nuclide – it can be used as a biodistribution imaging agent with the use of the same tracers, with already implemented labeling methods. This is especially important as 90Y is a pure β emitter, offering only very limited SPECT imaging due to brehmsstrahlung effect. A 86Y/90Y labeled mixture administered to the patient is a good example of a theranostic application, allowing simultaneous therapy with PET imaging (Rösch, Herzog, and Qaim, 2017).

Zaneb et al. (Zaneb et al., 2015) has theoretically proven that 86Y is best produced on small medical cyclotrons using a 86Sr(p,n)86Y nuclear reaction. It is predicted that the yield for this reaction is for 371 MBq/μAh assuming 100% enrichment of the solid target and irradiation by 14 MeV protons. The main competitive method against it is 88Sr(p,3n)86Y, which offers an almost 3-fold higher integral yield (1005 MBq/μAh), but requires an energy range of 43–33 MeV. The advantage of the latter process is that 88Sr is a much more abundant stable isotope of strontium (82.5% compared to 9.7% for 86Sr), which can translate into a lower cost of enriched materials. However, the high energy range still requires a rare high energy cyclotron, and there are serious considerations with impurity levels of 87mY, 87mY, and 85Y. The two factors combined render the second method uninteresting.

Production via the 86Sr(p,n)86Y reaction can give a product of high radionuclidic purity. A highly present by-product is 86mY (T1/2 = 48 min), but it decays to 86Y which turns out to simply increase nuclide production after EOB and should not be considered an impurity. Target post irradiation cooling and post processing of the nuclide considerably lower the amount of 86mY at the End of Synthesis (EOS). Reischl et al. (Reischl, Roesch, and Machulla, 2002) irradiated a 95.6% enriched solid 86SrCO3 target with 15.1 MeV protons which resulted in average production yields of 48 MBq/μAh with a radionuclidic purity of > 99% at EOB. The later separation by electrolysis (1 h) gives separation yields of 97%, no carrier-added. Using a 10 μA beam over 2.5 h, this results in 1 GBq of purified 86Y. Yoo et al. (Yoo et al., 2005) achieved the same separation yield and radionuclidic purity using a very similar setup. They had the same enrichment of 86Sr and used electrolysis for separation (with different parameters taking up to 3 h). A different target was used, namely SrO, and 2 μA, 14.5 MeV proton beams for 2 h resulted in much higher yields of 166 MBq/μAh. A few years later, a study by Avila-Rodriguez et al. (Avila-Rodriguez, Nye, and Nickles, 2008) revealed a faster separation method using filtration that takes only 20 min. They used several 11 MeV proton beams of 10 μA for 2 h using a 97% enriched strontium 86SrCO3 target. Even though this method is faster, it has a lower separation efficiency of 88%, results in carrier presence in the order of tens of ppm’s and has a lower radionuclidic purity of 97% at EOB (excluding 86mY) with 2.5% 87mY. The yield was 44 MBq/h – so in the same range as Reischl et al. (Reischl et al., 2002) – which confirms the superiority of SrO targets used by Yoo et al. (Yoo et al., 2005). 86Sr was fully recyclable in all cases.

The above mentioned works result in delivering an approximate amount of 1 GBq of 86Y labeled tracer which, in conjunction with 90Y therapy agents, could satisfy the need for a few patients.

An experiment is reported on 86Y production using liquid targets. Oehlke et al. (Oehlke et al., 2015) irradiated a 0.9 ml Niobium-body Havar-window liquid target with a 13 MeV proton beam from a small medical cyclotron. The solution contained natSr (NO3)2 salt dissolved in ultrapure water. Beam conditions were 4.6 μA in average, running for 1 h. The separation using DGA resin was 99% efficient. The resulting yield was 1.44 MBq/μAh, considerably lower than the above mentioned solid target results, but it is worth mentioning that natural strontium was used with a 10 times lower content of 86Sr than enriched material. By doing a simplified estimation, using an enriched material could give us an increase in yield of 10 times to 14 MBq/μAh. Small medical cyclotrons can have 2–3 higher target volumes than reported here – a further increase of the target volume could provide another yield increase. There is also room left to adjust the concentration of the salt. In conclusion, a liquid target method is a possible alternative for easier production of Yttrium-86 on a small scale.


One of the very recent radionuclides of interest for PET imaging is 44Sc (T1/2 = 3.97 h, 100% EC+β+, 94% β+av 632 keV,100% γ 1157 keV). Due to the chemical similarity of Sc3+ cation to Lu3+ and Y3+, DOTA complexes labeled with those radionuclides show very similar properties in vivo. This makes it a possible diagnostic surrogate for therapeutic tracers containing 177Lu or 90Y. Currently, [68Ga]Ga-DOTA-complexes are used for this purpose (Krajewski et al., 2013). 44Sc shows superiority over Gallium-68, given that the routine production method is well established. It does not require an expensive 68Ge/68Ge generator, it has a half-life suitable for logistics, a lower positron energy and slightly higher β+ branching ratio. Moreover, 47Sc (T1/2 = 3.35 d, 100% β, 69% βav 143 keV, 31% βav 204 keV, 68% γ 159 keV) is a therapeutic radionuclide of rising interest, which creates possibilities for a theranostic agent by applying a matched pair of 44Sc/47Sc. 44Sc can also be used independently in peptide based imaging, as well as antibody labeling and small protein labeling (Hernandez et al., 2014).

Multiple methods for the production of 44Sc have been investigated. Experiments were done with an α- beam at 29 MeV (Szkliniarz et al., 2016), a deuteron beam at 16 MeV (Alliot et al., 2015) and an innovative method employing a 44mSc/44Sc in vivo generator (Huclier-Markai et al., 2014). In the latter case, the in vivo generator utilizes the existence of the metastable state of 44mSc (otherwise treated as impurity) that decays by internal transition (98.8%) to 44Sc, which later emits positrons used for imaging. This way, the long half-life of 44mSc (T1/2 = 58.6 h) allows for longer pharmacokinetic studies, which is especially important when dealing with mABs. Additionally, 44mSc is closer with its half-life to the therapeutic isotope 44Sc (2.44 vs 3.35 d), making it possible to monitor drug metabolism over a longer period of time. Huclier-Markai et al. (Huclier-Markai et al., 2014) prove that the recoil energy of the decay of the metastable state, which is only 271 keV above its daughter state of 44Sc, is too low (0.89 eV) for the isotope to leave the chelator molecule, thus assuring stability of the complexes used.

Another method to acquire 44Sc is the 44Ti/44Sc generator (Roesch, 2012b; Filosofov, Loktionova, and Rösch, 2010). There, the parent nuclide of 44Ti (60.6 y, 100% EC, 93% γ 68 keV, 96% γ 78 keV) decays by electron capture to 44Sc for possible elution. The concept is based indirectly on the usage of cyclotrons, as the 44Ti is produced by a 45Sc(p,2n)44Ti reaction. An especially advantageous factor of this method is that 45Sc is the only naturally occurring stable isotope of scandium; therefore, there is no need to acquire expensive enriched material for irradiations. The peak cross section for this reaction is around 20 MeV proton energy (Soppera et al., 2017a), but the 44Ti production method is inappropriate for utilization by SMCs as production of very long-lived 44Ti would require a relatively high proton beam current, which is typically not possible with SMCs. For this reason, Filosofov et al. [104]produced 185 MBq of 44Ti on an intermediate energy cyclotron with proton energies of 25 MeV and a beam current of 200 μA. The final proposed generator allowed them to elute 180 MBq of Scandium on a weekly basis, which was continued for a year. The authors report a negligible amount of 44Ti breakthrough in the order of 10− 5 Bq. However, additional post processing steps are required to obtain a 44Sc solution. Additionally, long elution intervals and little amounts of final product are a certain limitation of the method, which can be countered only by intensive irradiations that could increase the amount of 44Ti.

In the case of small medical cyclotrons the common production route is via the 44Ca(p,n)44Sc nuclear reaction by using a solid target irradiation of highly enriched 44CaCO3 powder. The reaction cross section maximum reaches approx. Eight hundred mb for 11 MeV protons. 44mSc (T1/2 = 58.6 h, 99% IT, 87% γ 271 keV) is an impurity of concern, although the energy of the γ does not carry a risk to patients or the PET image resolution. Its long half-life creates the effect of the in-vivo generator mentioned before. The reaction 44Ca(p,n)44mSc cross-section reaches almost 100 mb and peaks at 13 MeV.

Krajewski et al. (Krajewski et al., 2013) presented an optimization study on the 44Ca(p,n)44Sc reaction pathway which reports 9 MeV to be the best optimized proton energy, giving > 99% radionuclidic purity with only 0.09% 44mSc content in the finished product. Other impurities of 43Sc, 46Sc, 47Sc and 48Sc were negligible or not detectable at all. 4 h irradiation of 20 mg of enriched [44Ca]CaCO3 with 12,4 μA beam resulted in 2.5 GBq of 44Sc. The further separation yield was 70%, calcium content measured by ICP-MS analysis was below 1 ppm. Further labeling yield with DOTA-TATE was > 98%, and the target recycling efficiency was only 60%. Other research groups have reached higher separation efficiencies of 80% (Valdovinos et al., 2015), and high radiochemical yields with different tracers above 90% (Hernandez et al., 2014), but their use of natural calcium as a target limits the production yield and radionuclidic purity, which in turn is not of use in clinical applications.

Recently published work by Meulen et al. (Meulen et al., 2015) presented a more efficient process and confirmed that the low energy range gave the best results, with an 11 MeV proton beam on 10 mg of 97% enriched 44CaCO3 target. With an improved target design, they were able to irradiate with 50 μA, hence using a shorter time of 90 min, reaching 1.9 GBq of 44Sc at EOB (note the amount of irradiated material is halved in comparison to the above study (Krajewski et al., 2013)). Moreover, the group reports a faster and very efficient separation method with 98% efficiency with < 1 ppm metallic contaminant levels. Radionuclidic purity was 99% and further radiolabelling efficiencies with DOTANOC were > 98%. The study also showed great improvement in the recycling of enriched calcium reaching 98% of initial enriched material recovery, significantly reducing the operational costs in comparison to work of Krajewski et al. (Krajewski et al., 2013). The required proton energy is low enough for almost all small medical cyclotrons, provided that the use of the solid target is accepted. The disadvantage of the method is certainly the cost of 44Ca. Its natural abundance is only 2.1%, therefore 97% enriched 44CaCO3 costs about $15/mg (Meulen et al., 2015). Fortunately, the material is highly recyclable.

44Sc was also investigated by Oehlke et al. for the production in a liquid target (Hoehr et al., 2014). Thirteen MeV proton beam irradiated natural calcium nitrate Ca (NO3)2 in dissolved in ultrapure water, in a relatively small volume target of 0.9 ml. The authors report a maximum of 28 MBq at EOB with a 20 μA beam current for 1 h, which may seem disappointing from a clinical perspective. However, a number of alterations could prove the method to be viable such as: increasing the beam time, target volume and concentration, and most importantly, a high enrichment of 44Ca could elevate the produced activity to GBq level.

Less common radionuclides in small medical cyclotron energy range

There are many other non-standard radionuclides that can be produced using small medical cyclotrons, but currently the usage of them is limited, either because of a lack of developed applications, scarce nuclear data, or significant investment needs. In a recent, impressive work by Qaim (Qaim, 2017), a list of such nuclides is presented. Less common diagnostic purpose radionuclides of increasing interest are especially: 51Cr, 45Ti, 76Br, 90Nb or 94mTc.

Therapeutic radionuclides are mostly produced in nuclear reactors and, due to their nature, are most often β emitters. These are, in general, neutron rich elements. However, there are also pathways possible for the production of therapeutic radionuclides using small medical cyclotrons such as 67Cu, 186Re, 103Pd and 225Ac (Qaim, 2017), but those nuclides are reported to be very expensive (Zimmermann, 2013).


This review presented the developments in small medical cyclotron production of radionuclides used for diagnostics over the past few years (Table 4). Some are increasingly being used in nuclear medicine, i.e. 68Ga and 89Zr. Some have a fast clinical introduction, taking over the field already occupied by previously implemented methods such as: cyclotron produced 99mTc over previous molybdenum generator supplied 99mTc; 89Zr over 64Cu in antibody labeling mostly due to a better suited half-life and no enrichment. And lastly, some novel radionuclides, like 86Y or 44Sc, present promising data for production implementation using SMCs, while gaining interest in clinical applications.
Table 4

Comparison of the properties of presented radionuclides

Radionuclide (in order of appearance)

Imaging procedure



Eγ or Eav. β (keV)

Feasible SMC (< 20 MeV) nuclear reaction

Target type




6.43 h

γ 99%




513 MBq/μAh(Benard et al., 2014)



13.2 h

γ 83%




No data



4.18 d

β+ 12% β+ 11% γ 63%






21 MBq/μAh (Braghirolli et al., 2014)



78.4 h

β+ 23% γ 99%






49 MBq/μAh (Siikanen et al., 2014)

4.4 MBq/μAh (Pandey et al., 2016)



12.7 h

β+ 18% β 39% γ 0.5%







304 MBq/μAh (Qaim, 2017)

0.14 MBq/(μAh·mg) (Alves et al., 2017)



67.7 min

β+ 88% γ 3%






No data

1.5 MBq/(μAh·mg) (Alves et al., 2017)



78.3 h

γ 39% γ 21% γ 17%







No data



67.3 h

γ 94% γ 91%





67.5 MBq/μAh (Alipoor et al., 2011)



14.7 h

β+ 12% β+ 6% γ 83%







166 MBq/μAh (Yoo et al., 2005)

1.44 MBq/μAh (Oehlke et al., 2015)



3.97 h

β+ 94% γ 100%






25 MBq/μAh (Meulen et al., 2015)

1.4 MBq/μAh (Hoehr et al., 2014)

Certainly, there are still many concerns for commercializing these novel cyclotron produced radionuclides. Apart from regulatory, financial, pharmaceutical and chemical concerns, there are also technical concerns. These include low production yields, costs of enriched materials, and improvement of solid target methodology, which is time consuming. In our opinion, more work should be published that encompasses production and optimization of non-standard radionuclides using standardized types of compact solid target systems per specific radionuclide. We would like to encourage the commercial compact solid target system users to share their production results and parameters. An alternative method employs the use of liquid targets for this purpose. A number of studies were reported in this paper on the production of novel radionuclides using liquid targets, but so far, the yields reached were mostly too low for commercial implementation. Addressing this issue still requires extensive research in the field of liquid targets used for the production of novel radionuclides.



Advanced Cyclotron Systems Incorporated


Compact Solid Target Irradiation System


End of bombardment

GE Healtcare: 

General Electric Healthcare


Good manufacturing practice


Highly enriched uranium


International Atomic Energy Agency


Ion Beam Applications


International Coherent Amplification Network


Japan Proton Accelerator Research Complex


Linear accelerator


Monocolonal antibodies


University of Missouri Research Reactor Center


Positron emission tomography


Small medical cyclotron


Single photon emission computed tomography


Vrije Universiteit Amsterdam



We would like to thank Roel Mooij, Chief engineer and Manager R&D Cyclotron and Targetry from BV Cyclotron VU for providing us materials about custom made solid targetry for Zirconium-89 production.


No funding received

Availability of data and materials

The dataset(s) supporting the conclusions of this article is(are) available in the IAEA - Nuclear Data Section Live Chart of Nuclides repository:

Authors’ contributions

Substantial contributions to the conception or design of the work: MAS, LRP, JFWN, Acquisition of data: MAS, LRP, Analysis and interpretation of data for the work: MAS, LRP, JFWN, Drafting of the work: MAS, Critical revision of the work for important intellectual content: LRP, JFWN, Final approval of the version to be published: MAS, LRP, JFWN, Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved: MAS, LRP, JFWN. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable

Consent for publication

Not applicable

Competing interests

J.F.W. Nijsen is co-founder and chief scientific officer (0.5 FTE) of Quirem Medical BV. He owns shares in the company Quirem Medical BV and holds several patents on his name as inventor.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors’ Affiliations

Radboudumc, Radboud Translational Medicine B.V, Nijmegen, The Netherlands
Radboudumc, Dept. of Radiology and Nuclear Medicine, Nijmegen, The Netherlands


  1. Accelerator-based Alternatives to Non-HEU production of Mo-99/Tc-99m 2011; IAEA coordinated research project]. Available from:
  2. ACSI. CycloTec™ and the National Cyclotron Network: meeting Canada's medical isotope needs. Available from:
  3. Alfuraih A, et al. Optimization of Zr-89 production using Monte Carlo simulations. J Radioanalytical Nucl Chem. 2013;296(2):1025–9.View ArticleGoogle Scholar
  4. Alipoor Z, Gholamzadeh Z, Sadeghi M. Data evaluation acquired Talys 1.0 code to produce in-111 from various accelerator-based reactions. Int J Mod Phys E Nucl Phys. 2011;20(5):1307–24.View ArticleGoogle Scholar
  5. Alliot C, et al. Cyclotron production of high purity (44m,44) sc with deuterons from (44) CaCO3 targets. Nucl Med Biol. 2015;42(6):524–9.View ArticlePubMedGoogle Scholar
  6. Alves F, et al. Production of copper-64 and gallium-68 with a medical cyclotron using liquid targets. Mod Phys Lett A. 2017;32(17):21.View ArticleGoogle Scholar
  7. Al-Yanbawi S, Al Jammaz I. Standardized high current solid tellurium-124 target for cyclotron production of the radionuclides iodine-123,124. Radiochim Acta. 2007;95(11):657–61.View ArticleGoogle Scholar
  8. Applications, A.A., Advanced accelerator applications announces European Commission approval of SomaKit TOC™. 2016.Google Scholar
  9. Asad AH, et al. Excitation functions of natZn (p,x) nuclear reactions with proton beam energy below 18 MeV. Appl Radiat Isot. 2014;94:67–71.View ArticlePubMedGoogle Scholar
  10. Aslam MN, Amjed N, Qaim SM. Evaluation of excitation functions of the 68,67,66Zn (p,xn)68,67,66Ga and 67Zn (p,α)64Cu reactions: validation of evaluated data through comparison with experimental excitation functions of the natZn (p,x)66,67Ga and natZn (p,x)64Cu processes. Appl Radiat Isot. 2015;96:102–13.View ArticlePubMedGoogle Scholar
  11. Aslam MN, et al. Evaluation of excitation functions of proton and deuteron induced reactions on enriched tellurium isotopes with special relevance to the production of iodine-124. Appl Radiat Isot. 2010;68(9):1760–73.View ArticlePubMedGoogle Scholar
  12. Avila-Rodriguez MA, Nye JA, Nickles RJ. Simultaneous production of high specific activity 64Cu and 61Co with 11.4 MeV protons on enriched 64Ni nuclei. Appl Radiat Isot. 2007;65(10):1115–20.View ArticlePubMedGoogle Scholar
  13. Avila-Rodriguez MA, Nye JA, Nickles RJ. Production and separation of non-carrier-added 86Y from enriched 86Sr targets. Appl Radiat Isot. 2008;66(1):9–13.View ArticlePubMedGoogle Scholar
  14. Azizakram H, et al. An overview of I-124 production at a medical cyclotron by ALICE/ASH, EMPIRE-3.2.2 and TALYS-1.6 codes. Appl Radiat Isot. 2016;112:147–55.View ArticlePubMedGoogle Scholar
  15. Bach HT, et al. Improving the survivability of Nb-encapsulated Ga targets for the production of Ge-68. Nucl Inst Methods Phys Res Sec B Beam Interact Mater Atoms. 2013;299:32–41.View ArticleGoogle Scholar
  16. Beaver JE, Hupf HB. Production of Tc-99m on a medical cyclotron - feasibility study. J Nucl Med. 1971;(11):12, 739.Google Scholar
  17. Benard F, et al. Implementation of multi-curie production of Tc-99m by conventional medical cyclotrons. J Nucl Med. 2014;55(6):1017–22.View ArticlePubMedGoogle Scholar
  18. Blaauw M, et al. Estimation of 99Mo production rates from natural molybdenum in research reactors. J Radioanalytical Nucl Chem. 2017;311(1):409–18.View ArticleGoogle Scholar
  19. Braghirolli AMS, et al. Production of iodine-124 and its applications in nuclear medicine. Appl Radiat Isot. 2014;90:138–48.View ArticlePubMedGoogle Scholar
  20. Brief, A.R.S.N., FDA approves first gallium 68Ga-Dotatate radiopharmaceutical tracer for PET imaging of neuroendocrine tumors. 2016.Google Scholar
  21. Brown, C. Will new isotope sources be ready in time? CMAJ : Can Med Assoc J 2016 ; [252-252]. Available from: [Cited 188 4].
  22. Buckley K. Cyclotron production of Tc-99m. Cyclotrons2013. Proc 20th Int Conf Cyclotrons Appl. 2013:482–6.Google Scholar
  23. Bychenkov VY, Brantov AV, Mourou G. Tc-99m production with ultrashort intense laser pulses. Laser and Particle Beams. 2014;32(4):605–11.View ArticleGoogle Scholar
  24. Cwikla JB, et al. 67Ga SPECT in detection of infection and inflammation. Nucl Med Rev Cent East Eur. 1999;2(2):69–73.PubMedGoogle Scholar
  25. Dabkowski AM, et al. Optimization of cyclotron production for Radiometal of zirconium 89. Acta Phys Pol A. 2015;127(5):1479–82.View ArticleGoogle Scholar
  26. Das MK, et al. Production and separation of Tc-99m from cyclotron irradiated mo-100/natural targets: a new automated module for separation of Tc-99m from molybdenum targets. J Radioanalytical Nucl Chem. 2016;310(1):423–32.View ArticleGoogle Scholar
  27. DeGrado TR, Pandey MK, Byrne J. Solution target for cyclotron production of radiometals: Google Patents; 2017.Google Scholar
  28. Ellison PA, et al. Spot-welding solid targets for high current cyclotron irradiation. Appl Radiat Isot. 2016;118:350–3.View ArticlePubMedPubMed CentralGoogle Scholar
  29. Engle JW, et al. Very high specific activity Ga-66/68 from zinc targets for PET. Appl Radiat Isot. 2012;70(8):1792–6.View ArticlePubMedPubMed CentralGoogle Scholar
  30. Filosofov DV, Loktionova NS, Rösch F. A 44Ti/44Sc radionuclide generator for potential application of 44Sc-based PET-radiopharmaceuticals: Radiochimica Acta International journal for chemical aspects of nuclear science and technology; 2010. p. 149.Google Scholar
  31. Fitzsimmons JM, Mausner L. Production scale purification of Ge-68 and Zn-65 from irradiated gallium metal. Appl Radiat Isot. 2015a;101:60–4.View ArticlePubMedGoogle Scholar
  32. Fitzsimmons JM, Mausner L. Development of a production scale purification of Ge-68 from irradiated gallium metal. Radiochim Acta. 2015b;103(2):117–23.View ArticleGoogle Scholar
  33. Fonslet J, Koziorowski J. Dry distillation of radioiodine from TeO2 targets. Appl Sci Basel. 2013;3(4):675–83.View ArticleGoogle Scholar
  34. Fujiwara M, et al. Production of medical (99m) Tc isotope via photonuclear reaction. Phys Part Nuclei. 2017;48(1):124–33.View ArticleGoogle Scholar
  35. Goethals PE, Zimmermann RG. Cyclotrons used in Nuclear Medicine World Market Report & Directory. 2015th ed; 2015.Google Scholar
  36. Hanemaayer V, et al. Solid targets for Tc-99m production on medical cyclotrons. J Radioanalytical Nucl Chem. 2014;299(2):1007–11.View ArticleGoogle Scholar
  37. Harvey JT, Isensee GH, Moffatt SD, Messina GP. NorthStar progress towards domestic Mo99 production. Washington, DC: presented at Topical Meeting on Molybdenum-99 Technological Development, June 2014; 2014.Google Scholar
  38. Hernandez R, et al. Sc-44: an attractive isotope for peptide-based PET imaging. Mol Pharma. 2014;11(8):2954–61.View ArticleGoogle Scholar
  39. Hoehr C, et al. Radiometals from liquid targets: 94mTc production using a standard water target on a 13MeV cyclotron. Appl Radiat Isot. 2012;70(10):2308–12.View ArticlePubMedGoogle Scholar
  40. Hoehr C, et al. (SC)-S-44g production using a water target on a 13 MeV cyclotron. Nuc Med Biol. 2014;41(5):401–6.View ArticleGoogle Scholar
  41. Holland JP, Sheh YC, Lewis JS. Standardized methods for the production of high specific-activity zirconium-89. Nucl Med Biol. 2009;36(7):729–39.View ArticlePubMedPubMed CentralGoogle Scholar
  42. Huclier-Markai S, et al. Optimization of reaction conditions for the radiolabeling of DOTA and DOTA-peptide with 44m/44Sc and experimental evidence of the feasibility of an in vivo PET generator. Nucl Med Biol. 2014;41:e36–43.View ArticlePubMedGoogle Scholar
  43. IAEA. Production of long lived parent Radionuclides for generators: 68Ge, 82Sr, 90Sr and 188W, in IAEA radioisotopes and radiopharmaceuticals series. Vienna: International Atomic Energy Agency; 2010.Google Scholar
  44. IAEA, Cyclotron Produced Radionuclides: Principles and Practice. Technical report series no.465. Vienna: International Atomic Energy Agency; 2008.Google Scholar
  45. Ikotun OF, Lapi SE. The rise of metal radionuclides in medical imaging: copper-64, zirconium-89 and yttrium-86. Future Med Chem. 2011;3(5):599–621.View ArticlePubMedGoogle Scholar
  46. Jalilian AR. An overview on Ga-68 radiopharmaceuticals for positron emission tomography applications. Iran J Nucl Med. 2016;24(1):1–10.Google Scholar
  47. Jalilian AR, Osso J. The current status and future of theranostic Copper-64 radiopharmaceuticals. Iran J Nucl Med. 2017;25(1):1–10.Google Scholar
  48. Jalilian AR, et al. Evaluation of [67Ga] citrate in the detection of various microorganism infections in animal models. Iran J Nucl Medi. 2009;17(2):34–41.Google Scholar
  49. Jones SC, Robinson GD Jr, McIntyre E. Tandem van de Graaff accelerator production of positron labeled radiopharmaceuticals for routine clinical use. Int J Appl Radiat Isot. 1984;35(8):721–9.View ArticlePubMedGoogle Scholar
  50. Kakavand T, et al. Computer simulation techniques to design Xenon-124 solid target for iodine-123 production. Iran J Radiat Res. 2008;5(4):207–11.Google Scholar
  51. Kalsy N, Vinjamuri S. Should we stop offering indium-111 octreotide scans in favour of gallium-68 PET-CT scans in the UK? Nucl Med Commun. 2016;37(12):1221–2.View ArticlePubMedGoogle Scholar
  52. Koehler L, et al. Iodine-124: a promising positron emitter for organic PET chemistry. Molecules. 2010;15(4):2686–718.View ArticlePubMedGoogle Scholar
  53. Krajewski S, et al. Cyclotron production of sc-44 for clinical application. Radiochim Acta. 2013;101(5):333–8.View ArticleGoogle Scholar
  54. Krijger GC, et al. The necessity of nuclear reactors for targeted radionuclide therapies. Trends Biotechnol. 2013;31(7):390–6.View ArticlePubMedGoogle Scholar
  55. Lahiri S, Maiti M, Ghosh K. Production and separation of in-111: an important radionuclide in life sciences: a mini review. J Radioanalytical Nucl Chem. 2013;297(3):309–18.View ArticleGoogle Scholar
  56. Lighthouse: productie medische isotopen vanaf 2021. 2017; Available from:
  57. Mahunka I, et al. Iodine-123 production at a small cyclotron for medical use. J Radioanalytical Nucl Chem Lett. 1996;213(2):135–42.View ArticleGoogle Scholar
  58. Manenti S, et al. The excitation functions of mo-100(p,x) mo-99 and mo-100(p,2n) Tc-99m. Appl Radiat Isot. 2014;94:344–8.View ArticlePubMedGoogle Scholar
  59. Mang'era K, et al. Processing and evaluation of linear accelerator-produced mo-99/Tc-99m in Canada. J Radioanalytical Nucl Chem. 2015;305(1):79–85.View ArticleGoogle Scholar
  60. Martins PD, Osso JA. Thermal diffusion of Ga-67 from irradiated Zn targets. Appl Radiat Isot. 2013;82:279–82.View ArticleGoogle Scholar
  61. Matarrese M, et al. Automated production of copper radioisotopes and preparation of high specific activity [64Cu]cu-ATSM for PET studies. Appl Radiat Isot. 2010;68(1):5–13.View ArticlePubMedGoogle Scholar
  62. Meulen NPvd, et al. Cyclotron production of 44Sc: from bench to bedside. Nulc Med Biol. 2015;42:745–51.View ArticleGoogle Scholar
  63. Nagatsu K, et al. Fully automated production of iodine-124 using a vertical beam. Appl Radiat Isot. 2011;69(1):146–57.View ArticlePubMedGoogle Scholar
  64. Nakai K, et al. Feasibility studies towards future self-sufficient supply of the <sup>99</sup>mo-<sup>99m</sup>Tc isotopes with Japanese accelerators. Proc Japan Acad Ser B. 2014;90(10):413–21.View ArticleGoogle Scholar
  65. NorthStar Medical Radioisotopes Receives $11.75 Million from National Nuclear Security Administration. 2015; Available from:
  66. OECD-NEA. 2017 medical isotope supply review: 99Mo/99mTc market demand and production capacity projection 2017–2022: The Supply of Medical Radioisotopes; 2017.Google Scholar
  67. Oehlke E, et al. Production of Y-86 and other radiometals for research purposes using a solution target system. Nucl Med Biol. 2015;42(11):842–9.View ArticlePubMedGoogle Scholar
  68. Ohya T, et al. Efficient preparation of high-quality 64Cu for routine use. Nucl Med Biol. 2016;43(11):685–91.View ArticlePubMedGoogle Scholar
  69. Pandey M, et al. Cyclotron production of Ga-68 using a solution target. J Nucl Med. 2014a;55:1.View ArticleGoogle Scholar
  70. Pandey MK, et al. Cyclotron production of (68) Ga via the (68) Zn (p,n) (68) Ga reaction in aqueous solution. Am J Nucl Med Mol Imaging. 2014b;4(4):303–10.PubMedPubMed CentralGoogle Scholar
  71. Pandey MK, et al. Improved production and processing of Zr-89 using a solution target. Nucl Med Biol. 2016;43(1):97–100.View ArticlePubMedGoogle Scholar
  72. Park HM. I-123: almost a designer radioiodine for thyroid scanning. J Nucl Med. 2002;43(1):77–8.PubMedGoogle Scholar
  73. Poniger SS, et al. Automated production of 124I and 64Cu using IBA Terimo and Pinctada metal electroplating and processing modules. AIP Conf Proc. 2012;1509(1):114–9.View ArticleGoogle Scholar
  74. Pupillo G, et al. Accelerator-based production of mo-99: a comparison between the mo-100(p,x) and Zr-96(alpha,n) reactions. J Radioanalytical Nucl Chem. 2015;305(1):73–8.View ArticleGoogle Scholar
  75. Qaim SM. The present and future of medical radionuclide production: Radiochimica Acta; 2012. p. 635.Google Scholar
  76. Qaim SM. Nuclear data for production and medical application of radionuclides: present status and future needs. Nucl Med B. 2017;44:31–49.View ArticleGoogle Scholar
  77. Qaim SM, et al. Some optimisation studies relevant to the production of high-purity 124I and 120gI at a small-sized cyclotron. Appl Radiat Isot. 2003;58(1):69–78.View ArticlePubMedGoogle Scholar
  78. Qaim SM, et al. Evaluation of excitation functions of mo-100(p,d+pn) mo-99 and mo-100 (p,2n) Tc-99m reactions: estimation of long-lived Tc-impurity and its implication on the specific activity of cyclotron-produced Tc-99m. Appl Radiat Isot. 2014;85:101–13.View ArticlePubMedGoogle Scholar
  79. Qaim Syed M. Development of novel positron emitters for medical applications: nuclear and radiochemical aspects: Radiochimica Acta International journal for chemical aspects of nuclear science and technology; 2011. p. 611.Google Scholar
  80. Rajec P, et al. Preparation and characterization of nickel targets for cyclotron production of 64Cu. J Radioanalytical Nucl Chem. 2010;286(3):665–70.View ArticleGoogle Scholar
  81. Reischl G, Roesch F, Machulla H-J. Electrochemical separation and purification of yttrium-86. 2002;90:225–8.Google Scholar
  82. Rice SL, et al. The next generation of positron emission tomography radiopharmaceuticals in oncology. Semin Nucl Med. 2011;41(4):265–82.View ArticlePubMedPubMed CentralGoogle Scholar
  83. Roesch F. Maturation of a key resource - the germanium-68/gallium-68 generator: development and new insights. Curr Radiopharm. 2012a;5(3):202–11.View ArticlePubMedGoogle Scholar
  84. Roesch F. Scandium-44: benefits of a long-lived PET radionuclide available from the (44) Ti/(44) sc generator system. Curr Radiopharm. 2012b;5(3):187–201.View ArticlePubMedGoogle Scholar
  85. Rösch F, Herzog H, Qaim SM. The beginning and development of the Theranostic approach in nuclear medicine, as exemplified by the radionuclide pair (86) Y and (90) Y. Pharmaceuticals. 2017;10(2):56.View ArticlePubMed CentralGoogle Scholar
  86. Rovais MRA, et al. Assessment of the direct cyclotron production of Tc-99m: an approach to crisis management of Tc-99m shortage. Appl Radiat Isot. 2016;112:55–61.View ArticlePubMedGoogle Scholar
  87. Sadeghi M, Enferadi M, Bakhtiari M. Accelerator production of the positron emitter zirconium-89. Ann Nucl Energy. 2012;41:97–103.View ArticleGoogle Scholar
  88. Schaffer P, et al. Direct production of 99mTc via 100Mo (p,2n) on small medical cyclotrons. Phys Procedia. 2015;66:383–95.View ArticleGoogle Scholar
  89. Schmor, P.W. Review of cyclotrons used in the production of radioisotopes for biomedical applications. in Proceedings of Cyclotrons 2010. 2010. Lanzhou.Google Scholar
  90. Siikanen J, et al. A solid target system with remote handling of irradiated targets for PET cyclotrons. Appl Radiat Isot. 2014;94:294–301.View ArticlePubMedGoogle Scholar
  91. Silberstein EB. Radioiodine: the classic Theranostic agent. Semin Nucl Med. 2012;42(3):164–70.View ArticlePubMedGoogle Scholar
  92. Soppera NDE, Bossant M, Cabellos O. JANIS book of proton induced cross sections, comparison of evaluated and experimental data from ENDF/B-VII.1, JENDL/HE-2007, PADF-2007, TENDL-2015 and EXFOR: OECD NEA Data Bank; 2017a.Google Scholar
  93. Soppera NDE, Bossant M, Cabellos O. JANIS book of deuteron induced cross sections: Comparison of evaluated and experimental data from ENDF/B-VII.1, TENDL-2015 and EXFOR; 2017b.Google Scholar
  94. Sounalet T, et al. Strontium-82 and future Germanium-68 production at the ARRONAX facility. Nuclear Data Sheets. 2014;119:261–6.View ArticleGoogle Scholar
  95. Stolarz A, et al. Molybdenum targets produced by mechanical reshaping. J Radioanalytical Nucl Chem. 2015;305(3):947–52.View ArticleGoogle Scholar
  96. Szelecsényi F, Blessing G, Qaim SM. Excitation functions of proton induced nuclear reactions on enriched 61Ni and 64Ni: possibility of production of no-carrier-added 61Cu and 64Cu at a small cyclotron. Appl Radiat Isot. 1993;44(3):575–80.View ArticleGoogle Scholar
  97. Szelecsényi F, et al. Excitation function of (p, α) nuclear reaction on enriched 67Zn: possibility of production of 64Cu at low energy cyclotron: Radiochimica Acta; 2014. p. 465.Google Scholar
  98. Szkliniarz K, et al. Production of medical sc radioisotopes with an alpha particle beam. Appl Radiat Isot. 2016;118:182–9.View ArticlePubMedGoogle Scholar
  99. Takacs S, et al. Benchmark experiment for the cross section of the mo-100(p,2n) Tc-99m and mo-100(p,pn) mo-99 reactions. Nucl Instrums Methods Phys Res Sec B Beam Interact Mater Atoms. 2016;375:60–6.View ArticleGoogle Scholar
  100. Tang Y, et al. A simple and convenient method for production of Zr-89 with high purity. Appl Radiat Isot. 2016;118:326–30.View ArticlePubMedGoogle Scholar
  101. Tárkányi F, et al. Excitation functions of deuteron induced nuclear reactions on natural zinc up to 50 MeV. Nucl Inst Methods Phys Res Sec B Beam Interact Mater Atoms. 2004;217(4):531–50.View ArticleGoogle Scholar
  102. Valdovinos HF, et al. Separation of cyclotron-produced sc-44 from a natural calcium target using a dipentyl pentylphosphonate functionalized extraction resin. Appl Radiat Isot. 2015;95:23–9.View ArticleGoogle Scholar
  103. van der Marck SC, Koning AJ, Charlton KE. The options for the future production of the medical isotope 99Mo. Eur J Nucl Med Mol Imaging. 2010;37(10):1817–20.View ArticlePubMedPubMed CentralGoogle Scholar
  104. Vosjan MJWD, et al. Conjugation and radiolabeling of monoclonal antibodies with zirconium-89 for PET imaging using the bifunctional chelate p-isothiocyanatobenzyl-desferrioxamine. Nat Protocols. 2010;5(4):739–43.View ArticlePubMedGoogle Scholar
  105. Yoo J, et al. Preparation of high specific activity 86Y using a small biomedical cyclotron. Nucl Med Biol. 2005;32(8):891–7.View ArticlePubMedGoogle Scholar
  106. Zaneb H, et al. Nuclear model analysis of excitation functions of proton induced reactions on Sr-86, Sr-88 and Zr-nat: evaluation of production routes of Y-86. Appl Radiat Isot. 2015;104:232–41.View ArticlePubMedGoogle Scholar
  107. Zimmermann RG. Why are investors not interested in my radiotracer? The industrial and regulatory constraints in the development of radiopharmaceuticals. Nucl Med Biol. 2013;40(2):155–66.View ArticlePubMedGoogle Scholar


© The Author(s) 2018