[¹⁸F]Trifluoroiodomethane – Enabling Photoredox‐mediated Radical [¹⁸F]Trifluoromethylation for Positron Emission Tomography

Abstract

The development of new tracers for positron emission tomography (PET) is highly dependent on the available synthetic tools for their radiosynthesis. Herein, we present the radiosynthesis and application of [¹⁸F]trifluoroiodomethane – the first reagent for broad scope radical [¹⁸F]trifluoromethylation chemistry in high molar activity. CF₂ ¹⁸FI can be prepared from [¹⁸F]fluoroform with 67±5 % AY and >99 % RCP. Its synthetic utility is demonstrated by the radiosynthesis of previously unprecedented ¹⁸F‐labeled α‐trifluoromethyl ketones and ¹⁸F‐labeled trifluoromethyl sulfides, important motifs that are present in a range of bioactive compounds. Both protocols are Ru‐ and photo‐mediated and proceed under mild reaction conditions. They show good functional group tolerance evidenced by the respective reaction scopes and make use of easily obtainable starting materials. The products can be isolated in 8.3–11.1 GBq/μmol (starting from ca. 5 GBq [¹⁸F]fluoride). The applicability to PET tracer synthesis is shown by the radiolabeling of bioactive compounds, such as derivatives of probenecid and febuxostat. In a broader context, this work opens the door to the utilization of radical [¹⁸F]trifluoromethylation chemistry for the radiolabeling of PET tracers in high molar activity.

The radiosynthesis of [¹⁸F]trifluoroiodomethane is described. CF₂ ¹⁸FI serves as building block for radical [¹⁸F]trifluoromethylation reactions in high molar activity, as exemplified by the Ru‐ and photoredox‐mediated synthesis of a range of ¹⁸F‐labeled α‐trifluoromethyl ketones and trifluoromethyl sulfides. This opens new opportunities for the synthesis of [¹⁸F]CF₃‐containing tracers for positron emission tomography.

Positron emission tomography (PET) is a non‐invasive imaging technique commonly used for the diagnosis of a plethora of diseases and of increasing importance for drug discovery.[ ¹ , ² ] It makes use of radiotracers, i.e., radiolabeled compounds, which interact with target structures in vivo. Among the positron‐emitting radionuclides used for PET, ¹⁸F has a prominent position because of its favorable characteristics, i.e., the amenable half‐life (t_1/2 = 109.8 min) and low positron energy (E_βmax = 0.63 MeV),[ ³ , ⁴ ] and its prevalence in pharmaceutical compounds.[ ⁵ , ⁶ ] It is produced in medical cyclotrons and generally obtained as [¹⁸F]fluoride in low quantities (nmol range). Therefore, existing strategies for the incorporation of ¹⁹F cannot be easily translated to ¹⁸F‐labeling.[ ⁷ , ⁸ ] Particularly challenging is the ¹⁸F‐labeling of highly abundant CF₃‐groups.[ ⁹ , ¹⁰ , ¹¹ ] Approaches using [¹⁸F]fluoride often require the synthesis of specific precursors[ ¹² , ¹³ , ¹⁴ ] (if possible at all) limiting the accessible chemical space. In addition, they often result in low molar activity (A_m) stemming from ¹⁸F‐¹⁹F‐isotopic exchange. The use of ¹⁸F‐labeled building blocks, namely [¹⁸F]fluoroform,[ ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ ] [¹⁸F]CuCF₃,[ ¹⁷ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ ] and [¹⁸F]Ruppert‐Prakash reagent,[ ²⁷ , ²⁸ ] has proven to present a viable solution to these problems by enabling the direct [¹⁸F]trifluoromethylation of often broadly available precursors via nucleophilic and transmetalation chemistry in high molar activity. Consequently, the expansion of this strategy to the exploration of other reactivities, e.g., radical trifluoromethylation, is highly desirable and bears the potential to unlock access to previously inaccessible radiochemical space.

In the past two decades, photoredox catalysis has become an indispensable tool in modern organic chemistry[ ²⁹ , ³⁰ , ³¹ , ³² , ³³ ] and can enable radical chemistry under mild reaction conditions. As a result, it has attracted the attention of radiochemists and found application for the synthesis of PET tracers.[ ³⁴ , ³⁵ , ³⁶ , ³⁷ ] Despite significantly contributing to recent developments in radical trifluoromethylation chemistry,[ ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ ] photoredox‐mediated radical [¹⁸F]trifluoromethylation remains unprecedented. The exploration of this strategy is highly dependent on the synthesis of a suitable ¹⁸F‐labeled reagent that can enable the appropriate reactivity. We envisioned the ideal reagent to be:

1. accessible in <30 min to ensure reasonable activity yield (AY) and allow for enough time for further synthesis and purification of the final product,

2. obtainable in high molar activity (A_M) to ensure usage for the synthesis of PET tracers for the imaging of low‐density targets, and

3. usable in multiple solvents to ensure application to a variety of reactions (Scheme 1a

   Scheme 1.

   

   [a] Requirements for a [¹⁸F]CF₃‐containing reagent suitable for broadly applicable photoredox‐mediated [¹⁸F]trifluoromethylation chemistry; [b] Previously evaluated synthetic route via halogen exchange; [c] Synthesis of [¹⁸F]trifluoroiodomethane from [¹⁸F]fluoroform.

   ).

Two reagents theoretically capable to enable photoredox‐mediated radical [¹⁸F]trifluoromethylation chemistry have been previously reported.[ ⁴² , ⁴³ ] However, their comparably low AY (5–6 %) and A_m (0.24–1.13 GBq/μmol) and the need for high performance liquid chromatography (HPLC) purification render them unattractive for such applications.

We speculated that [¹⁸F]trifluoroiodomethane (CF₂ ¹⁸FI) would be a suitable reagent. Its non‐radioactive isotopologue is a widely used reagent in radical trifluoromethylation chemistry[ ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵¹ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ , ⁵⁹ , ⁶⁰ ] indicating its synthetic potential for applications in radiochemistry. In addition, CF₃I is gaseous which allows purification by distillation. This reduces the synthesis time dramatically and ensures the isolation in a solvent of choice. More than two decades ago, efforts directed toward the synthesis of CF₂ ¹⁸FI ^[61] and the analogous CBrF₂ ¹⁸F ^[62] tried to establish a synthesis via halogen exchange using [¹⁸F]fluoride and CF₂I₂ or CBr₂F₂, respectively (Scheme 1b). However, the simultaneous formation of at least equal quantities of [¹⁸F]fluoroform, inseparable from CF₂ ¹⁸FI by distillation, could not be overcome. Consequently, the isolation and application of CF₂ ¹⁸FI has never been demonstrated. Although unknown for the non‐radioactive isotopologue, we surmised that instead an access from CHF₂ ¹⁸F in the presence of base and a suitable iodine source would be an attractive synthetic route likely providing the reagent with high A_M (Scheme 1c).

In this article, we disclose the radiosynthesis of CF₂ ¹⁸FI in 25 minutes with good AY and high A_M. In addition, we demonstrate its application to the photoredox‐mediated synthesis of ¹⁸F‐labeled α‐trifluoromethyl ketones and trifluoromethyl sulfides.

We commenced our studies with the synthesis of CF₂ ¹⁸FI 2. An initial experiment using potassium bis(trimethylsilyl)amide (KHMDS) as base and iodine as I⁺‐source revealed the general feasibility of the access from CHF₂ ¹⁸F 1 (Table 1, entry 1). However, the majority of 1 was converted to [¹⁸F]2,2,2‐trifluoro‐1‐dimethylamino‐1‐ethanol, a common by‐product resulting from the reaction with DMF. ^[15] Changing the order of addition revealed that adding the base as the last reagent enabled the formation of 2 in a cleaner fashion (entries 2 and 3). Varying the ratio of I₂ and KHMDS led to improved formation of 2 but it was found to be impossible to exceed 30 % radiochemical conversion (RCC) reproducibly (entry 4). A solvent screen showed no desired reactivity in toluene but significantly higher conversion to 2 (86±7 % RCC) when using THF (entries 5 and 6). The screening of other I⁺‐sources, N‐iodosuccinimide (NIS) and 1,3‐diiodo‐5,5‐dimethylhydantoin (DIH), showed no improved reaction performance (entries 7 and 8). Since using THF can generally not lead to the formation of [¹⁸F]2,2,2‐trifluoro‐1‐dimethylamino‐1‐ethanol and 1 is added to the reaction mixture by distillation, we concluded that, for practical purposes, it would be desirable to add 1 as the last reagent, i.e., trap 1 in a mixture of I₂ and KHMDS in THF. Despite initially leading to lower formation of 2 (43 % RCC, entry 9), increasing the amount of KHMDS led to formation of 2 in quantitative RCC (entry 10). Finally, by distillation with a He‐flow of 30 mL/min at 50 °C, 2 could be isolated in 67±5 % AY (n=35) and >99 % RCP.

Table 1.

Optimization of the synthesis of [¹⁸F]trifluoroiodomethane.^[a]

En‐try	KHMDS [μmol]	I+‐source [μmol]	sol‐vent	added last	Consmp. 1 [b]	2 [c]
1	100	I2 (100)	DMF	I2	88±4 %	7±1 %[d]
2	100	I2 (100)	DMF	base	19±8 %	16±5 %
3	100	I2 (100)	DMF	1	0 %	n.d.[d]
4	300	I2 (100)	DMF	base	50±36 %	30±31 %
5	300	I2 (100)	PhMe	base	100 %	n.d.
6	300	I2 (100)	THF	base	95±5 %	86±7 %
7	300	NIS (100)	THF	base	70 %	n.d.[e]
8	300	DIH (100)	THF	base	80±1 %	75±1 %[d]
9	300	I2 (100)	THF	1	48 %	43 %[e]
10	300	I2 (200)	THF	1	100 %	100 %[e] (67±5 %)[f]

[a] Standard reaction conditions: 1 (25–50 MBq), KHMDS, I⁺‐source, solvent (900 μL total), 20 °C, 5 min. [b] Consumption of 1. [c] Radiochemical conversion (RCC) based on HPLC analysis, average ± standard deviation, n=3. [d] n=2. [e] n=1. [f] Activity yield (AY), n=35, using 0.4–1.8 GBq 1.

With CF₂ ¹⁸FI in hand, we sought out to explore its synthetic potential using photoredox‐mediated chemistry. Despite the occurrence in a range of bioactive compounds,[ ⁶³ , ⁶⁴ , ⁶⁵ ] the synthesis of ¹⁸F‐labeled α‐trifluoromethyl ketones remains unprecedented. Inspired by numerous reports describing the synthesis of these motifs from silyl enol ethers or enolates via radical trifluoromethylation,[ ⁴⁵ , ⁶⁶ , ⁶⁷ , ⁶⁸ , ⁶⁹ , ⁷⁰ , ⁷¹ , ⁷² , ⁷³ , ⁷⁴ , ⁷⁵ , ⁷⁶ , ⁷⁷ ] we surmised that CF₂ ¹⁸FI would be suitable to address this synthetic challenge. Upon optimization of reaction parameters, it was found that triisopropylsilyl (TIPS) enol ether 3 a could be converted to 4 a in 94 % RCC. The reaction is enabled by photoredox catalyst [Ru(bpy)₃]Cl₂⋅6H₂O (1 μmol) and tetramethylethylenediamine (TMEDA) as base and proceeds under blue light irradiation at 20 °C in five minutes. Key to the successful reaction design was the choice of the appropriate base and silyl group (for more information, see Supporting Information). The evaluation of the reaction scope showed that the reaction is applicable to a wide range of substrates (Scheme 2). The presence of several electron‐withdrawing substituents, such as fluoro‐, bromo‐, iodo‐, nitrile‐, and trifluoromethylether‐groups in ortho‐, meta‐, para‐position are well tolerated delivering the respective products in 22–96 % RCC (4 b‐4 g). Likewise, substrates bearing electron‐donating substituents, such as methyl‐ and methoxy‐groups, delivered products in 89–97 % RCC (4 h and 4 i).

Scheme 2.

Scope of ¹⁸F‐labeled α‐trifluoromethyl ketones. Standard reaction conditions: substrate (50 μmol), [Ru(bpy)₃]Cl₂⋅6H₂O (1 μmol), TMEDA (50 μmol), [¹⁸F]trifluoroiodomethane (25–50 MBq), MeCN (600 μL), 5 min, 20 °C, blue light. Radiochemical conversion based on HPLC analysis, average ± standard deviation, TIPS=triisopropylsilyl. [a] 10 min reaction time. [b] n=2.

Remarkably, the radiolabeling of quarternary carbon centers (4 j and 4 k) proceeded in 25–83 % RCC. Gratifyingly, we found the strategy to also be applicable to aliphatic ketones (4 l‐4 n, 35–61 % RCC). In this case, the performance could not be further enhanced by using triethyl silyl (TES) enol ethers. ^[45] Multiple aliphatic and aromatic heterocyclic substrates, incl. a tetrahydrothiopyranone, piperidone, and benzothiophene, could be functionalized as well in 35–82 % RCC (4 m – 4 o). To demonstrate the utility of the protocol for the synthesis of PET tracers, derivatives of bioactive compounds probenecid and febuxostat were tested. To our delight, the corresponding products were formed in 40–61 % RCC (4 p and 4 q). In addition, 4 q was radiolabeled using a full batch of CF₂ ¹⁸FI (800±220 MBq, starting from 4.8±0.8 GBq [¹⁸F]fluoride) for which the previously optimized reaction conditions could be readily adapted. The product was then isolated with semi‐preparative HPLC in 6.3±1.4 % radiochemical yield (RCY), 4.9±1.1 % AY, 8.3±3.3 GBq/μmol A_m, and >99 % radiochemical purity (RCP). This clearly shows that the presented protocol is suitable for routine PET tracer productions. It is worth noting that PET tracer productions often start with higher activities of [¹⁸F]fluoride (50–500 GBq) which further increases the molar activity, as previously demonstrated. ^[16]

Apart from enabling the access to previously inaccessible ¹⁸F‐labeled structural motifs like α‐trifluoromethyl ketones, we considered that CF₂ ¹⁸FI could also be used to facilitate the access to previously reported motifs. For instance, trifluoromethyl sulfides are important motifs for drug discovery. Their radiosynthesis has already been described using different approaches.[ ¹⁹ , ²⁸ , ⁷⁸ , ⁷⁹ , ⁸⁰ ] However, all known protocols have in common that they use rather high quantities of precursor (typically 40–100 μmol, with one exception ^[79] ) and do not provide the sulfides in ideal molar activity (max. 2.5 GBq/μmol) ^[28] . In addition, they are limited to either aryl‐ or alkyl‐substituted substrates. Upon optimization of reaction parameters (see Supporting Information), we found that CF₂ ¹⁸FI can be used to convert thiol 5 a to ¹⁸F‐labeled trifluoromethyl sulfide 6 a in 93±3 % RCC (Scheme 3). Once again, the reaction is enabled by photoredox catalyst [Ru(bpy)₃]Cl₂⋅6H₂O (1 μmol) and the appropriate base, in this case 1,1,3,3‐tetramethylguanidine (TMG), and proceeds under blue light irradiation at 20 °C. Remarkably, the reaction is highly efficient and is finished in only 1 minute while requiring only 5 μmol precursor. In addition, the direct use of thiols for the synthesis, without prefunctionalization, has not been reported before. An experiment using a full batch of CF₂ ¹⁸FI (900±320 MBq, starting from 5.3±1.6 GBq [¹⁸F]fluoride) allowed the isolation of 6 a in 80±6 % radiochemical yield (RCY), 72±5 % AY, 11.1±4.7 GBq/μmol A_m, and >99 % radiochemical purity (RCP). The evaluation of the reaction scope showed that the presence of electron‐withdrawing (6 b and 6 c) and –donating substituents (6 d and 6 e) in ortho‐, meta‐, and para‐position are equally well tolerated and delivered the product in 65–98 % RCC. While the presence of an alcohol group led to a reduced RCC (6 f, 25±6 %), carboxylic acids do not hamper the reaction performance at all (6 g, 75±3 %). Different heterocycles, incl. a pyridine and benzothiazole, could be functionalized as well, delivering the products in 95–100 % RCC (6 h and 6 i). Lastly, it was found that the protocol could also be applied to aliphatic compound 6 j (36±12 % RCC). This makes it the first protocol that can functionalize both aromatic and aliphatic substrates.

Scheme 3.

Scope of ¹⁸F‐labeled trifluoromethyl sulfides. Standard reaction conditions: substrate (5 μmol), [Ru(bpy)₃]Cl₂⋅6H₂O (1 μmol), TMG (5 μmol), [¹⁸F]trifluoroiodomethane (25–50 MBq), MeCN (600 μL), 1 min, 20 °C, blue light. Radiochemical conversion based on HPLC analysis, average ± standard deviation. [a] n=2.

In summary, we have developed a protocol for the synthesis and isolation of [¹⁸F]trifluoroiodomethane. CF₂ ¹⁸FI can be used as reagent for photoredox‐mediated [¹⁸F]trifluoromethylation chemistry. Its synthetic utility was demonstrated by the radiosynthesis of previously unprecedented ¹⁸F‐labeled α‐trifluoromethyl ketones and trifluoromethyl sulfides. Both protocols are operationally simple, possess a good functional group tolerance, and allow the isolation of products in high molar activity. The applicability to routine PET tracer synthesis has been demonstrated by the isolation of 4 q, a derivative of bioactive compound febuxostat. We expect this work to set the foundation for the development of more methods using photoredox‐mediated [¹⁸F]trifluoromethylation with CF₂ ¹⁸FI.

Supporting Information

The authors have cited additional references within the Supporting Information.[ ⁸¹ , ⁸² , ⁸³ , ⁸⁴ , ⁸⁵ , ⁸⁶ , ⁸⁷ , ⁸⁸ , ⁸⁹ , ⁹⁰ , ⁹¹ , ⁹² , ⁹³ , ⁹⁴ , ⁹⁵ ]

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Veth L., Windhorst A. D., Vugts D. J., Angew. Chem. Int. Ed. 2025, 64, e202416901. 10.1002/anie.202416901

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.