Unambiguous analytical separation and improved syntheses of [11C]acetate and [11C]acetoacetate Authors: Karsten Bamminger, Lukas Nics, Marcus Hacker, Cécile Philippe, Marius Ozenil Published: 2025 ABSTRACT AbstractBackground[11C]Acetate and [11C]acetoacetate are PET radiotracers widely used to assess oxidative metabolism and ketone body utilization, respectively. This study aimed to establish robust, high-yield syntheses of both tracers using a GE TRACERlab FX2 C module, with an emphasis on improving radiochemical purity (RCP), radiochemical yield (RCY), optimizing operational parameters, and developing accurate quality control methods. [11C]Acetate was synthesized via Grignard carboxylation using [11C]CO2 and purified with a cartridge-based system. [11C]Acetoacetate was produced via in-loop [11C]CO2 carboxylation of a lithium enolate precursor, followed by semi-preparative reversed-phase HPLC purification. Quality control was performed by reported ion-exchange chromatography (IEX-HPLC) and novel reversed-phase HPLC (RP-HPLC). A systematic literature review was conducted to evaluate prior quality control methods for [11C]acetoacetate.ResultsOmission of helium flow during [11C]CO2 trap bake-out significantly improved activity recovery from the [11C]CO2 trap (from 63 to 89%) and reduced release time (from 4.8 to 3.1 min). [11C]Acetate and [11C]acetoacetate were synthesized with mean isolated activities of 30.2 GBq and 3.24 GBq and mean RCPs of 96.9% and 97.1%, respectively. The final formulations met all European Pharmacopoeia criteria. While the widely-used IEX-HPLC method failed to differentiate [11C]acetate from [11C]acetoacetate, the newly-developed RP-HPLC method enabled unambiguous separation. Literature analysis revealed that most published studies on [11C]acetoacetate likely overlooked [11C]acetate as a radiochemical impurity due to insufficient analytical separation.ConclusionsReliable synthesis protocols for [11C]acetate and [11C]acetoacetate were established on the GE TRACERlab FX2 C, with significant improvements in [11C]CO2 handling and product purification. Inclusion of the proposed RP-HPLC method enables a more accurate and specific assessment of RCP compared to IEX-HPLC alone and should be considered for a valid quality control of [11C]acetoacetate.Supplementary InformationThe online version contains supplementary material available at 10.1186/s41181-025-00412-0.Keywords: Carbon-11, Acetate, Acetoacetate, Quality control, Purification, Separation, HPLC Background [11C]Acetate is a widely utilized PET tracer for imaging cellular metabolism, particularly fatty acid synthesis and oxygen consumption, across various tissues, including the heart and tumors (Grassi et al. 2012). Numerous production protocols have been developed and systematically compared in recent studies (Wenz et al. 2022). From a radiochemical perspective, all approaches are based on the fixation of [11C]CO2 by a methylmagnesium halide, followed by aqueous hydrolysis (Scheme 1). Nevertheless, these methods differ in specific operational details, including the choice of halide, type of reactor (classic reactor versus in-loop synthesis), and purification strategy (cartridge-based purification versus distillation). Reaction scheme for the radiosynthesis of [11C]acetate through carboxylation of the Grignard reagent with cyclotron-produced [11C]CO2 The European Pharmacopoeia monograph on [11C]acetate outlines a high-performance liquid chromatography (HPLC) method for determining the radiochemical purity (RCP) of the tracer (European Directorate for the Quality of Medicines (EDQM). Monograph 1920 - Sodium Acetate ([1-11C]) Injection. European Pharmacopoeia 112 Monograph 1920 -Sodium Acetate ([1-11C]) Injection 2023). Briefly, [11C]acetate is eluted from a strongly basic anion-exchange column using 0.1 mol/L NaOH. However, specific radiochemical impurities that this method determines are not mentioned in the monography. More commonly, the RCP of [11C]acetate preparations is determined by HPLC on either C18 columns (Mock et al. 2011; Tang et al. 2013; Felicini et al. 2010) or the Aminex HPX-87H column (Kruijer et al. 1995; Bars et al. 2006; Kang et al. 2007; Maurer et al. 2019). The latter has been highlighted for its ability to detect the impurities [11C]bicarbonate, [11C]acetone, and [11C]tert-butanol. [11C]Acetoacetate is predominantly used in PET imaging studies investigating ketone body metabolism in the brain (Nugent et al. 2014) and heart (St-Pierre et al. 2024). Although its application is less widespread compared to [11C]acetate, the number of published radiosynthetic procedures and clinical studies has steadily increased over recent years. [11C]Acetoacetate is typically synthesized via [11C]CO2 carboxylation of the lithium enolate of acetone (Scheme 2), with two notable recent exceptions reported in the literature (Kaur et al. 2024; Qu et al. 2020). Reaction scheme for the radiosynthesis of [11C]acetoacetate via carboxylation of a lithium enolate intermediate, generated from isopropenyl acetate and MeLi, with cyclotron-produced [11C]CO2. Additionally, the reaction yielding [11C]acetate as byproduct in the synthesis of [11C]acetoacetate is depicted Considering that the radiosynthesis of [11C]acetoacetate involves methyllithium (MeLi) as reagent and [11C]CO2 as synthon, [11C]acetate is an expectable byproduct in [11C]acetoacetate formulations (Scheme 2). This side reaction was already observed experimentally by Straatman et al. in the first reported [11C]acetoacetate synthesis (Straatman et al. 1974), which follows the same labeling chemistry as contemporary procedures (Qu et al. 2020; Tremblay et al. 2007; Solingapuram Sai et al. 2017; Prenen et al. 1990). Straatman et al. noted that ‘Any excess of methyl lithium is a problem in this application as it leads to acetic acid as a final product’. and ‘It was also found that use of an excess of methyl lithium in the generation of the enolate anion led to labeled 11C-acetic acid as determined by alumina chromatography’. (Straatman et al. 1974). In 1974, HPLC analysis was not widely available, so the authors tested their [11C]acetoacetate preparations using traditional chemical assays for impurities. The product was adsorbed onto a neutral alumina column, which was sequentially washed with water and then pH 4 acetate buffer to discriminate [11C]acetoacetate from [11C]acetate. Later reports of [11C]acetoacetate have relied exclusively on radio-HPLC to the determine radiochemical purity (Qu et al. 2020; Tremblay et al. 2007; Solingapuram Sai et al. 2017), a method that has since become the standard in modern quality control of carbon-11 PET tracers (Monograph 1920 2023; Monograph 1617 2023; Monograph 1917 2023; Monograph 1924 2023; Monograph 2462 2023). In short, [11C]acetoacetate is eluted with 5 mmol/L H2SO4 from an ion-moderated partition chromatography column (IEX-HPLC), i.e., Aminex HPX-87H. To this end, the aforementioned IEX-HPLC method has been the most commonly applied approach for determining the RCP of [11C]acetoacetate (Qu et al. 2020; Tremblay et al. 2007; Solingapuram Sai et al. 2017) and is also widely applied for [11C]acetate (Kruijer et al. 1995; Bars et al. 2006; Kang et al. 2007; Maurer et al. 2019). However, to the best of our knowledge, the capability of this method to distinguish between [11C]acetate and [11C]acetoacetate has not yet been systematically evaluated. A high RCP is pivotal for ensuring diagnostic accuracy of any radiotracer. For carbon-11 PET tracers, radio-HPLC is typically used as the sole measure of RCP, and considerable effort is put on decreasing the quality control time in order to increase the useable amount of tracer activity (Nics et al. 2018). It may be tempting to assume high RCP based solely on a sharp-eluting radio-HPLC peak. However, the validity of an RCP measurement critically depends on using an analytical method, or a combination of methods, that can clearly separate the PET tracer from expected or known radiochemical impurities. If a radiochemical impurity exhibits a markedly different biodistribution from the intended radiotracer, its presence may become apparent during image analysis. This includes scenarios such as [18F]F− in [18F]FDG (unexpected bone uptake) (Iagaru et al. 2009) and [99mTc]TcO4− in 99mTc-labeled tracers (unexpected thyroid uptake) (Vallabhajosula et al. 2010). In cases where this difference in biodistribution is less obvious, as with [11C]acetate versus [11C]acetoacetate (Croteau et al. 2023), the chances of detecting radiochemical impurities from imaging data are greatly reduced. This underscores the critical importance of careful RCP determination in such cases. During the establishment of a production process for [11C]acetoacetate and [11C]acetate intended for preclinical studies, we (i) adapted existing literature procedures to achieve reliable production with an RCP exceeding 95%, and (ii) identified that the commonly used IEX-HPLC method is incapable of distinguishing [11C]acetoacetate from [11C]acetate. This finding prompted the development of an analytical RP-HPLC method capable of separating and accurately quantifying both compounds. Methods Materials Tetrahydrofuran (THF) (anhydrous, ≥ 99.9%, inhibitor-free, Merck, #401,757), methyl lithium (MeLi) (1.6 M in Et2O, Merck, #197,343), methylmagnesium chloride (MeMgCl) (3.0 M in THF, Merck, #189,901), 0.9% aqueous NaCl (B. Braun, #3,505,731), aqueous ammonia (25%, Merck, #1.05422), phosphoric acid (Merck, #1.60305), ammonium phosphate (Merck, #101,126), sodium 1-octanesulfate (Merck, #74,885), sulfuric acid (Merck, #1.12080), acetonitrile (ACN) (Merck, #100,030), LiI (anhydrous, Merck, 8.18287), molecular sieves (4Å, Merck, #1.05739), NaOH (Honeywell, #S5881), NaCl (Honeywell, #71,379), polypropylene (PP) tubing (100 cm, 1.0 mm ID, Vygon, #1155.10), ETFE line (Tefzel™ tubing, 1/16″ OD × 0.03″ ID, #1528XL), check valves (CODAN R-Lock, #16.5250), 1 mL syringes (PP/PE, oil-free, B. Braun, #9166017 V), Millex-GV filters (0.22 µm, PVDF, Merck, #SLGV033RS), and vented glass vials (sterile, evacuated DRN 4370 TechneVials 25 mL, Curium), target gas N2 + 1% O2 (Air Liquide), Helium (6.0, Messer) and Argon (5.0, Messer) were used as received. Lithium acetoacetate (Merck, #A8509) and sodium acetate (Merck, #241,245) were used as reference standard for the HPLC analyses. Isopropenyl acetate (Sigma-Aldrich, #117,781) was stored over molecular sieves in a glass crimp vial. Aqua ad iniectabilia (B. Braun, #3,505,715) and ethanol (Merck, #1.00986) was used for module washing, cartridge conditioning and reactions. Milli-Q water (Merck) was used for preparation of mobile phases. PS-H+ cartridges (125 mg, CHROMAFIX®, Macherey-Nagel, #731,867), PS-Ag+ cartridges (125 mg, CHROMAFIX®, Macherey-Nagel, #731,866) and the strong anion exchange (SAX) cartridge (600 mg, Maxi-Clean™, S*Pure, #SP-5122471) were preconditioned before use as detailed below. In its standard configuration the GE TRACERlab FX2 C (General Electric Medical System) does not allow direct utilization of [11C]CO2. A bypass line for [11C]CO2 was constructed by placing a three way Teflon valve (VWR, #BOHLF731-06) between V29 and the adjacent ascarite trap. The CO2 trap contained 300 mg of 4 Å 80/10 molecular sieves (Alltech, #5624) and 200 mg of Shimalite-Ni (Shimadzu, #221-27,719). For semi-preparative purification after radiosynthesis the GE TRACERlab FX2 C synthesis module was paired with a S1122 pump (Sykam), a BlueShadow UV detector (KNAUER Wissenschaftliche Geräte). A range of semi-preparative HPLC columns were evaluated for their suitability to purify [11C]acetoacetate, including: Synergi Fusion-RP (250 mm × 10 mm, 4 µm, Phenomenex, #00G-4424-N0), SUPELCOSIL ABZ + (250 mm × 10 mm, 5 µm, Merck, #59,179), Nucleodur PolarTec (250 mm × 10 mm, 5 µm, Macherey-Nagel, #762,223.100), ProntoSIL 120 C18 ace-EPS (250 mm × 5.0 mm, 5 µm, BISCHOFF Analysentechnik u. -geräte, #N2550F18APSJ) and µBondapak C18 (300 mm × 7.8 mm, 10 µm, Waters, #WAT084176). Analytical quality control was conducted using a 1260 Infinity HPLC system (Agilent Technologies) equipped with a GABI Nova radioactivity detector (Elysia Raytest), in combination with either an Aminex HPX-87H column (300 mm × 7.8 mm, Bio-Rad, #1,250,140) or an Aqua C18 analytical column (250 mm × 4.6 mm, 5 µm, Phenomenex, #00G-4299-E0). Synthesis of [11C]acetate An oven-dried 11 mL crimp-top glass vial was equipped with an Argon-filled balloon and cooled to room temperature. The vial was charged with anhydrous THF (1000 µL) and 3 M MeMgCl in THF (500 µL). The synthesis module was prepared according to previously established cleaning protocols (Pichler et al. 2019). The line connecting V1 and the reactor was disconnected from the reactor and thoroughly dried with He from V18. V1 was filled with 2 mL water, V5 with 2 mL 0.9% NaCl, and V6 with 15 mL water. A SAX cartridge, conditioned with 20 mL EtOH, 20 mL 0.9% NaCl and 20 mL water was placed on the SPE position of the module between V11 and V12. Two PS-H+ cartridges were connected and conditioned with 10 mL EtOH and 10 mL water. Two PS-Ag+ cartridges were connected and conditioned with 10 mL EtOH and 10 mL water. Together with other components, the cartridges were assembled to a single-use synthesis unit, which was connected to the synthesis module (see Fig. 1). Flow diagram of the TRACERlab FX2 C module with changes for the production of [11C]acetate highlighted in red and in the white box. Unused module parts are visually faded. Two check valves are used for the connection of the single-use synthesis unit to the [11C]CO2 trap, to ensure its protection from water originating from V1. This adapted diagram is displayed with kind permission of GE Healthcare The CO2 trap was heated to 400 °C while closed (i.e., without He flow) for 10 min and subsequently cooled to below 40 °C with pressurized air. Shortly before transfer of [11C]CO2 to the synthesis module, the polypropylene (PP) loop was preloaded with 50 µL of the diluted Grignard reagent using a 1 mL syringe as indicated in Fig. 1. Approximately 120 GBq of [11C]CO2 were produced via the 14N(p,α)11C nuclear reaction in a high pressure N2/O2 gas target using a GE PETtrace 860 cyclotron (General Electric Medical System). [11C]CO2 was transferred to the synthesis module using target gas and trapped on the CO2 trap. Subsequently, the [11C]CO2 was transferred from the heated CO2 trap (400 °C) to the single use synthesis unit with a He flow of 10 mL/min. Upon plateauing of the signal from the external radioactivity detector, the contents of the PP loop were flushed with 2 mL of water from V1 through a sequence of ion exchange cartridges (PS-H⁺, PS-Ag⁺, and SAX) to the waste bottle. The SAX cartridge was then washed with 15 mL water via V6 and eluted with 2 mL 0.9% aqueous NaCl from V5 into the product collection vial, which was filled in advance with 5 mL 0.9% aqueous NaCl. The product was sterile filtered inline using a Millex-GV 0.22 µm PVDF membrane filter into a sterile, vented, evacuated glass vial. The time list of the synthesis sequence is shown in Table S1 in the Supporting Information. Synthesis of [11C]acetoacetate The lithium enolate reagent was prepared following a previously reported procedure (Qu et al. 2020), with minor modifications. An oven-dried 20 mL crimp-top glass vial was purged with argon and cooled to − 78 °C using an acetone/dry ice bath. Methyl lithium in diethyl ether (1.6 M, 3.9 mL, 6.24 mmol) was added, followed by the dropwise addition of isopropenyl acetate (0.34 mL, 3.12 mmol) via syringe. The reaction mixture was stirred for 1 h at − 78 °C and subsequently diluted with THF (4.4 mL). The synthesis module was prepared according to previously established cleaning protocols (Pichler et al. 2019). For semi-preparative HPLC, a Synergi Fusion-RP column was employed, using 50 mM ammonium dihydrogen phosphate buffer (pH 2.1) as the mobile phase at a flow rate of 5 mL/min. The [11C]CO2 bypass line was connected to port 1 of the HPLC six-port valve using a 60 cm ETFE line equipped with a Luer-lock check valve to ensure the [11C]CO2 trap remains protected from the HPLC eluent (Fig. 2). Flow diagram of the TRACERlab FX2 C with changes for the production of [11C]acetoacetate highlighted in red and in the white box. Unused module parts are visually faded. This adapted diagram is displayed with kind permission of GE Healthcare The CO2 trap was preconditioned by heating to 400 °C under closed conditions (without He flow) for 10 min, followed by rapid cooling to below 40 °C using pressurized air. Immediately prior to [11C]CO2 transfer, the freshly prepared lithium enolate reagent was loaded into a 5 mL HPLC stainless steel injection loop via a dried 15 cm Tefzel line (1/16″ finger-tight fitting with Luer lock connector) from port 1 toward the loop waste outlet. Approximately 120 GBq of [11C]CO2 was produced via a high-pressure N2/O2 gas target on a GE PETtrace 860 cyclotron. The [11C]CO2 was transported to the synthesis module via target gas flow and adsorbed onto the CO2 trap. At a flow rate of 10 mL/min, the [11C]CO2 was thermally desorbed from the trap at 400 °C and directed into the HPLC loop. Upon plateauing of [11C]CO2 release, the six-port HPLC valve was switched to the ‘inject’ position to initiate the semi-preparative HPLC separation. The fraction corresponding to [11C]acetoacetate (tR = 4.7–5.5 min, Fig. S1) was sterile-filtered inline using a Millex-GV PVDF membrane filter into a sterile, vented, evacuated glass vial. To achieve physiological pH and osmolality, 1.3 mL of 0.5 mol/L NaOH containing 1% NaCl was added per minute of peak-cut via a sterile filter. A representative semi-preparative chromatogram of the [11C]acetoacetate synthesis is shown in Fig. S1. The time list of the synthesis sequence is shown in Table S2 in the Supporting Information. Quality control of tracer formulations Radiochemical purity was determined using an Agilent 1260 HPLC System consisting of Quat Pump, MWD VL UV detector and GABI Nova radioactivity detector. Ion-Exchange Chromatography (IEX-HPLC) was used for [C]acetate and [11C]acetoacetate Radio-HPLC was performed on a 1260 Infinity HPLC system equipped with an Aminex HPX-87H column. The mobile phase consisted of 5 mmol/L sulfuric acid (H2SO₄), delivered at a flow rate of 0.8 mL/min. UV detection was carried out at 220 nm. Reversed-Phase Chromatography (RP-HPLC) was used for [11C]acetoacetate The 1260 Infinity HPLC system was coupled to an Aqua C18 column. The mobile phase was delivered isocratically consisting of 100 mmol/L ammonium phosphate buffer (pH 2.1) containing 5 mmol/L sodium 1-octanesulfonate. The flow rate was set to 1.0 mL/min, with UV detection at 220 nm. Osmolality was determined using a vapor pressure osmometer (VAPRO, Wescor). pH measurements were performed with a pH meter 913 (Metrohm) equipped with a Biotrode electrode (Merck 6.0224.100). The gamma emission peak was assessed using an Elysia Raytest miniGITA Dual system. Sterility and endotoxin testing of the product formulation was conducted by MPL Mikrobiologisches Prüflabor GmbH. Residual solvent concentration was assessed by gas chromatography (GC) using an Intuvo 9000 GC system (Agilent Technologies) configured with an Intuvo SSI Guard Chip (Agilent Technologies, #G4587-60,565) and an Ultra Inert Intuvo GC column module (60 m × 0.32 mm ID, 1.80 µm, Agilent Technologies, #123-1364UI-INT), as outlined in the Supplementary Information (Table S3). In short, helium was used as the carrier gas at a flow rate of 4.0 mL/min from 0 to 4.0 min, followed by an increase to 6.5 mL/min for the remainder of the run. The column temperature program was as follows: initial temperature of 50 °C held until 0.5 min, followed by a linear ramp (20 °C/min) to 130 °C by 4.5 min, and a subsequent ramp (60 °C/min) to 170 °C by 5.2 min, which was maintained until the end of the run. The flame ionization detector (FID) was operated at 300 °C, with a hydrogen fuel flow rate of 30 mL/min. Details are described in the Supporting Information. Under these conditions, ACN and THF exhibited retention times of 4.1 and 4.9 min, respectively. Concentration limits for ACN and THF were set in accordance with the European Pharmacopoeia (European Directorate for the Quality of Medicines (EDQM). Monograph 1920 - Sodium Acetate ([1-11C]) Injection. European Pharmacopoeia 112 Monograph 1920 -Sodium Acetate ([1-11C]) Injection 2023) at 410 ppm and 720 ppm, respectively. Aqueous multicomponent certified reference material standards (UL Solutions) in concentrations of 10 ppm, 75 ppm and 250 ppm were used for calibration (Fig. S2). Limits of quantification (LOQ) were determined to be 52.2 ppm for ACN and 68.0 ppm for THF, while limits of detection (LOD) were 17.2 ppm and 22.5 ppm, respectively. LOD and LOQ were estimated ‘Based on the Standard Deviation of a Linear Response and a Slope’ according to ICH guideline Q2(R2). Results Synthesis of [11C]acetate The described radiosynthesis provided 30.2 ± 9.9 GBq of [11C]acetate with a RCP of 96.9 ± 2.8% (n = 11). The decay-corrected RCY was 35.8 ± 9.9% after an average synthesis time of 10 min and a success rate of 100%. Product quality met the European Pharmacopoeia requirements for [11C]acetate (European Directorate for the Quality of Medicines (EDQM). Monograph 1920 - Sodium Acetate ([1-11C]) Injection. European Pharmacopoeia 112 Monograph 1920 -Sodium Acetate ([1-11C]) Injection 2023) with respect to solvent residues (ACN < 50 ppm, THF < 15 ppm), osmolality (290 ± 7 mosmol/kg), pH (6.07 ± 0.29), radionuclidic purity (Emax = 526 ± 9 keV), endotoxins (< 0.99 EU/mL) and sterility. Synthesis of [11C]acetoacetate The described radiosynthesis provided 3.24 ± 0.81 GBq of [11C]acetoacetate with an RCP of 97.1 ± 1.3% (n = 7). The decay-corrected RCY was 4.67 ± 1.36% after an average synthesis time of 15 min, with a synthesis success rate of 78%. Product quality met the European Pharmacopoeia requirements for [11C]acetate (European Directorate for the Quality of Medicines (EDQM). Monograph 1920 - Sodium Acetate ([1-11C]) Injection. European Pharmacopoeia 112 Monograph 1920 -Sodium Acetate ([1-11C]) Injection 2023) with respect to solvent residues (ACN < 25 ppm, THF < 100 ppm), osmolality (261 ± 23 mosmol/kg), pH (6.60 ± 0.55), radionuclidic purity (Emax = 523 ± 12 keV), endotoxins (< 0.99 EU/mL) and sterility. Discussion Optimizations of [11C]acetate synthesis The synthesis procedure of [11C]acetate developed herein is inspired by the in-loop method described by Soloviev and Tamburella (Soloviev and Tamburella 2006). Limited modification in the wiring of the TRACERlab FX2 C module allowed to incorporate the single use sterile PP loop reactor to the module’s flowpath, providing a fully-automated [11C]acetate radiosynthesis for a widely available module. In initial [11C]acetate syntheses, the [11C]CO2 trap was flushed with He (50 mL/min) while heating the trap during module preparation, as this is recommended by the module manufacturer to remove moisture (Healthcare 2024). However, under these conditions, [11C]CO2 release from the trap to the loop was highly inefficient, resulting in only low yields of [11C]acetate. Post synthesis activity distribution analysis revealed that most of the activity was found on the [11C]CO2 trap (~ 40%) and the SAX cartridge (~ 20%). We noticed a significantly more efficient release of [11C]CO2 from the trap, if the trap is not flushed with He during heating as part of the module preparation (Figs. 3 and 4). Hence, subsequent [11C]acetate syntheses were performed without flushing the [11C]CO2 trap, as detailed in the Methods part. Omitting He flow during the pre-synthesis bake-out of the [11C]CO2 trap resulted in a significant increase in activity recovery, from 63 ± 21 to 89 ± 5% (p = 0.008, two-sided t-test, n = 8), and a significant reduction in release time upon heating initiation, from 4.8 ± 1.2 to 3.1 ± 1.3 min (p = 0.027, two-sided t-test, n = 8). This improvement may be attributed to more efficient and uniform heating of the trap in the absence of helium flow. Exemplary synthesis data from a [11C]acetate production run employing a pre-run [11C]CO2 trap bake-out under continuous helium flow (50 mL/min). Approximately 50% of the initially trapped [11C]CO2 was released within 3.5 min after heating to 400 °C, indicating low release efficiency under these conditions Exemplary recorded synthesis data from a [11C]acetate production run employing a pre-run [11C]CO2 trap bake-out without continuous helium flow. Upon heating the trap to 400 °C, approximately 80% of the initially trapped [11C]CO2 was released within 1 min, contributing to an improved radiochemical yield compared to Fig. 3 The absolute yield (mean 30.2 GBq) of [11C]acetate consistently exceeded the quantities required for preclinical studies. Therefore, further optimization of radiochemical yield was not prioritized. Instead, emphasis was placed on the reliability and robustness of the synthesis protocol, which was demonstrated by a 100% synthesis success rate (n = 11). Optimizations of [11C]acetoacetate synthesis Following the previously reported synthesis of [11C]acetoacetate employing cartridge-based purification (Qu et al. 2020), we observed insufficient RCP of the final product (Fig. S5), evidenced by the presence of residual [11C]CO2 and additional radioactive byproducts. Many of the cartridge-based radiosynthesis procedures report RCPs just above or around the 95% threshold (Supporting Information, Table S4), suggesting that reaching a reliable RCP > 95% has been found challenging also for other radiochemistry centers. To address this limitation, we implemented semi-preparative high-performance liquid chromatography (HPLC) to improve product purity. However, the inherently low lipophilicity of [11C]acetoacetate poses a considerable challenge for purification using reversed-phase HPLC. To identify a suitable stationary phase, we evaluated several columns available in semi-preparative format, including SUPELCOSIL ABZ⁺, Nucleodur PolarTec, ProntoSIL 120 C18 ace-EPS, and µBondapak C18. Among these, the polar-embedded C18 stationary phase Synergi Fusion provided adequate retention of [11C]acetoacetate, enabling reliable separation from radioactive impurities (Supporting Information, Fig. S1). The in-loop approach offers several advantages that render it particularly well-suited for the routine production of PET radiopharmaceuticals. The confined reaction environment minimizes the volume of organic solvents required, which is beneficial for downstream purification, particularly when using semi-preparative reversed-phase HPLC columns with limited solvent tolerance. This not only enhances chromatographic performance but also reduces the risk of column overloading. Additionally, connecting the loop to an inert gas source provides a controlled chemical environment conducive to the stability and reactivity of sensitive precursors, such as lithium enolate anions. The compact design and low dead volume of in-loop systems also promote efficient reaction kinetics, potentially leading to improved radiochemical yields. In-loop methodologies have proven reliable for [11C]CO2 fixation reactions, as demonstrated in the synthesis of [11C]acetate, [11C]PHNO (Pfaff et al. 2019) and [11C]WAY-100635 (Krasikova et al. 2009). To the best of our knowledge, however, no previously reported synthesis of [11C]acetoacetate has employed an in-loop [11C]CO2 carboxylation, as described in this work. Following the in-loop [11C]CO2 carboxylation, the majority of the radioactivity remained as unreacted [11C]CO2. In an effort to improve the RCY, the reaction loop was subsequently loaded with undiluted precursor, omitting tetrahydrofuran (THF) as a co-solvent. However, this modification resulted in the formation of side products, ultimately reducing the RCP of the final compound. Based on previous reports indicating that lithium iodide (LiI) enhances the stability of lithium β-ketocarboxylates (Berton et al. 2017), 1 equivalent of LiI was added to the precursor solution prior to loop loading, with the aim of improving the RCY. Despite this intervention, no significant increase in RCY was observed. Further optimization was not pursued, as the achieved yields were deemed sufficient to support both preclinical animal studies and potential future human applications. Of the nine acetoacetate productions performed according to our adapted protocol, two failed for following reasons: In the first case, the operator unintentionally added too much NaOH containing 1% NaCl, which resulted in a pH of 9.2 of the final formulation. In the second case, the steel loop was clogged during transfer of [11C]CO2 from the trap. An RP-HPLC method to discriminate between [11C]acetate and [11C]acetoacetate During method development, we found that the previously employed IEX-HPLC method was unable to adequately separate [11C]acetate from [11C]acetoacetate (see Figs. 5 and S6). Consequently, we consider this method unsuitable for accurate determination of the RCP of [11C]acetoacetate, particularly given the potential formation of [11C]acetate as a byproduct. To address this limitation, we propose an alternative RP-HPLC method (see Figs. 5 and S7) for RCP determination, offering several key advantages: (i) clear discrimination between [11C]acetate and [11C]acetoacetate; (ii) a significantly shorter analysis time, which is critical given the short half-life of carbon-11; and (iii) the possibility of applying a gradient elution program, enabling a more representative assessment of RCP compared to the isocratic conditions typically used in ion-exchange chromatography. Comparison of radiochromatograms of [11C]acetate and [11C]acetoacetate when applying the widely-used IEX-HPLC method versus the RP-HPLC method. In the IEX-HPLC method both radiotracers elute with the same retention time. In the RP-HPLC both radiotracers elute baseline separated, thus allowing for a clear discrimination Analysis of [11C]acetoacetate literature In an analysis of [11C]acetoacetate literature, detailed in the Supporting Information (Fig. S8, Tables S4, and S5), we identified six publications detailing the synthesis and quality control of [11C]acetoacetate and 21 publications focusing on [11C]acetoacetate’s imaging applications. Although [11C]acetoacetate was first described already in 1974, half of the publications are dated to the year interval 2017–2025, highlighting the increasing popularity of this tracer for imaging applications (Fig. S8). Of the five publications detailing the synthesis and quality control of [11C]acetoacetate and employing [11C]CO2 carboxylation, only two early publications (Straatman et al. 1974; Prenen et al. 1990) considered the carbon-11 labeled carboxylic acid [11C]acetic acid (when using MeLi) or [11C]pentanoic acid (when using BuLi) as a radioactive impurity. The other three of these five publications used a radio-HPLC analysis using the Aminex HPX-87H with 5 mM H2SO4 as mobile phase as sole method for determining [11C]acetoacetate’s RCP and hence will not be able to detect [11C]acetate as radiochemical impurity. Of the 21 publications focusing on [11C]acetoacetate’s imaging applications 13 did not describe or reference the method used for determining the RCP of [11C]acetoacetate. All of the remaining eight publications referenced publications using radio-HPLC analysis with Aminex HPX-87H and 5 mM H2SO4 as mobile phase as sole method for determining [11C]acetoacetate’s RCP. Conclusively, contamination of [11C]acetoacetate batches with [11C]acetate would have been unnoticed in the majority of [11C]acetoacetate batches across different PET centers. Conclusion In this study, we successfully established robust and reproducible synthesis protocols for [11C]acetate and [11C]acetoacetate using a GE TRACERlab FX2 C module. We identified that omitting helium flow during the pre-synthesis bake-out of the [11C]CO2 trap significantly improved the release efficiency of [11C]CO2. For [11C]acetoacetate a semi-preparative reversed-phase HPLC purification step was introduced to ensure product purity. Importantly, we demonstrated that the commonly used IEX-HPLC method is incapable of discriminating [11C]acetoacetate from [11C]acetate, a known byproduct of the synthesis, thereby compromising the accuracy of RCP assessments in existing literature. To address this critical limitation, we developed an RP-HPLC method that reliably resolves [11C]acetoacetate from [11C]acetate. These findings underscore the importance of carefully tailored analytical methods for quality control of carbon-11 PET radiopharmaceuticals, particularly when structurally close impurities with similar pharmacokinetics are expected. The presented methods provide a solid foundation for reliable preclinical and clinical production of [11C]acetate and [11C]acetoacetate. Supplementary Information Supplementary Material 1. Acknowledgements We thank Thomas Zenz (Medical University of Vienna) for maintaining the cyclotron and the radiosynthesis module, Dmitry Soloviev (University of Glasgow) for scientific discussion and troubleshooting, Eva-Maria Patronas for support with the GC analytics and Stefan Riese (GE Healthcare) for the right to display modified flow diagrams of the TRACERlab FX2 C in this publication. Abbreviations Acetonitrile Gas chromatography High performance liquid chromatography Ion-exchange high performance liquid chromatography Tetrahydrofuran Positron emission tomography Polypropylene Radiochemical conversion Radiochemical purity Radiochemical yield Reversed-phase high performance liquid chromatography Author contributions KB and MO performed the radiosyntheses as well as quality control and analyzed the data. LN, MH and CP supervised as well as supported data analysis. KB and MO drafted the manuscript. All authors critically revised the manuscript and all authors approved the final manuscript. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data availability Data generated or analysed during this study are included in this published article and its supplementary information files. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Karsten Bamminger, Email: karsten.bamminger@meduniwien.ac.at. Marius Ozenil, Email: marius.ozenil@meduniwien.ac.at. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement Data generated or analysed during this study are included in this published article and its supplementary information files.