Chemical reactivity of the 18F electrophilic reagents from the 18O(p,n)18F gas target systems

Nuclear Copyright

Medicine & Biology, Vol. 0 1996 Elsevier Science

23, pp. 559-565, Inc.

ISSN

1996

0969~8051/96/$15.00 SSDI 0969-8051(95)02073-X

+ 0.00

ELSEVIER

Chemical Reactivity of the 18F Electrophilic Reagents from the lHO(p,n)‘“F Gas Target Systems Allyson Bishop,’ N. Satyamurthy,‘* Gerald Bida and Jorge R. Barrio’ 'DIVISION OF NUCLEAR MEDICINE, DEPARTMENT OF MOLECULAR AND MEDICAL PHARMACOLOGY, UCLA SCHOOL OF MEDICINE ANDTHE LABORATORY OF STRUCTURAL BIOLOGY AND MOLECULAR MEDICINE (DOE),LOS ANOELES,CA 90095, USA; 'BIOMEDICAL RESEARCH FOUNDATION OF NORTHWEST LOUISIANA, 1505 KINGS HIGHWAY, SHREVEPORT, LA 71103,USA

ABSTRACT.

fluorinating agents derived A comprehensive evaluation of the reactivity of 18F electrophilic from the ‘sO(p,n)‘8F reaction conducted in target bodies made of aluminum, silver, copper, nickel, and gold-plated copper is reported. Two representative electrophilic reactions, namely addition across a double bond and substitution on an aromatic ring (fluorodemercuration and fluorodestannylation), were tested with the 18F activity generated in the [‘80]0, target systems. Identical reactions were also conducted with analogous nonradioactive fluorinating agents as control experiments. The products of all these reactions were analyzed by NMR spectroscopy. Results of these experiments clearly showed that ‘*F activities recovered from aluminum, silver, and copper target bodies were better suited for radiochemical syntheses, irrespective from a gold-plated target was suitable only for addition of the irradiation protocols employed. The “F activity reactions. Further, the fluorodemercuration reaction essentially failed with the single-step irradiation technique conducted in nickel and gold-plated targets. In contrast, the fluorodestannylation reaction was quite reagents recovered from all the target bodies and irradiation protocols. NUCL MED successful with the “F BIOL 23;5: 559-565, 1996.

KEY WORDS.

[‘*F]F,,

[18F]OFz,

Fl uorodemercuration,

Fluorodestannylation,

6-[“F]fluoro-L-dopa,

Aluminum

target

INTRODUCTION The production of 18F electrophilic species via the prolific 1sO(p,n)18F nuclear reaction (15) is steadily gaining popularity (1, 3, 6, 11, 19, 20). Two major irradiation protocols used with this reaction are (a) a single-step approach involving the irradiation of [‘“O]O,/F, mixtures (3) and (b) a two-step technique that entails irradiation of neat [lRO]oxygen followed by an ‘sF recovery procedure (11). Until recently, the identities of the electrophilic fluorinated species generated in the [‘80]oxygen targets have been solely inferred by chemical reactivity (3, 11, 15, 16). However, reactions of the single-step ‘*F electrophilic fluorinating agents met with limited success when they were extended to aromatic fluorinations involving the synthesis of 6-[18F]fluoro-L-dopa (21). For the first time, we have utilized a multinuclear NMR spectroscopic approach and conclusively established all the major fluorinated species generated in oxygen gas target systems made of aluminum, silver, copper, nickel, and gold-plated copper (1, 4). For instance, the [‘*O]oxygen/fluorine-irradiated target gases contained a mixture of reactive fluorinated species, of which F, was the predominant constituent. Products such as OF, and FONO, (fluorine nitrate) were also recovered in minor amounts from the oxygen gas targets. It is, however, not clear if these minor components were responsible for the observed differences in the reactivities of the “F activity derived from the oxygen gas target systems (21). Thus, to utilize efficiently the ‘*F activities in various radiochemical syntheses, an understanding of the role of fluorinated and other oxygen

*Author for correspondence. Accepted 11 September 1995.

species recovered from different target bodies in chemical reactions is required. Reported herein are our efforts in this regard. A major attention was focused on the most promising and innovative of the target body materials, namely aluminum (1). Because of its low activation properties, ease of fabrication, and efficient release of the 18F activities, we were keen to demonstrate its success in making chemically reactive 18F species and, hence, the emphasis of experiments was on this target body material.

EXPERIMENTAL The products from all reactions were analyzed by NMR spectroscopy. The ‘H (360.14 MHz), 19F (338.87 MHz), l19Sn (134.30 MHz), and ‘99Hg (64.50 MHz) NMR spectra were recorded on a Bruker AM-360 WB spectrometer. The 19F, ‘19Sn, and 19’Hg NMR spectra were obtained in the broad-band proton decoupled mode. The ‘H chemical shifts were referenced to internal tetramethylsilane or external sodium, 2,2-dimethyl-2-silapentane-5~sulfonate, whereas the 19F chemical shifts were referenced to external C,F, (for products of addition reaction) or CFCl, (for products of substitution reaction). Tetramethyltin and dimethylmercury were used as external standards for ‘19Sn and 199Hg NMR, respectively. All NMR samples were analyzed at room temperature, and the chemical shifts are given in ppm. The proton irradiations of [180]oxygen at 13.5 MeV and 10.3 MeV (on-target beam energies) were conducted on a CS-22 (The Cyclotron Corp.) and a RDS 112 (Siemens/CTl) cyclotron, respectively. The “‘Ne (d,a)“F reaction for the production of [lRF]F, was performed at 10.7 MeV (on-target beam energy) on the CS-22 cyclotron. The target bodies used in this investigation generally followed the design described recently (22). Other relevant details

560

A. Bishop

in this regard are described elsewhere (1, 4). The [‘*O]oxygen single-step (3) and two-step (1, 11) irradiation protocols have been previously described. The [‘80]oxygen single-step method provided, on average, 85% [18F]F, + 15% [‘8F]OF, while the two-step process yielded 100% [18F]F, (aluminum target) and 93% [‘*F]F, + 5% [18F]OF, + 2% [‘sF]FONO, (gold-plated target) (1, 4). The production of [‘*F]acetyl hypofluorite utilized the NaOAc * 3HzO method (2).

Electrophilic Addition of “F Activity Recovered from the [“OIOxygen Targets to 3,4,6-tri-O-acetyl-D-gh.md (TACj,

et al.

Quantitation of Ozone in the Irradiated OxygenTarget Gases All target bodies before use with fluorine were utilized for this experimentation. The targets were filled with oxygen (-230 psi; 1602 was substituted for “Oz due to the expense of the latter) and irradiated with protons (typical beam conditions: lo-30 p,A for lo-60 min). The target gases were then bubbled into buffered 1M KI solution (pH = 7.0; phosphate buffer) and titrated against a standard solution of sodium thiosulfate (0.0098 N) to determine the concentration of ozone (5). An ozone concentration of 5-10 ppm was observed for all target systems investigated.

1)

The fluorination of TAG (1) was carried out with [‘sF]F,, [“F]F, (85%) + [18F]OF, (15%) and [“F]AcOF. The fluorinating agent was bubbled into TAG in Freon (10 mL) at room temperature, with a 2 to 1 ratio of glucal to fluorinating agent. The Freon was bubbled off with a gentle flow of nitrogen and the residue was hydrolyzed with 1N HCl (1 mL) at 120°C for 30 min. The product was then purified with an ion-retardation resin/alumina column/C-l8 SepPak arrangement (2). After the decay of the ‘*F radiolabel, the product was analyzed by 19F NMR. The product mixture constituted essentially 2-fluoro-2-deoxy-D-glucose (2) [2-FDG: 19F NMR 6 + 32.74 (a-anomer)and +32.90 (p-anomer)] and 2-fluoro-2-deoxy-Dmannose (3) [2-FDM: 6 + 38.15 (a-anomer) and +56.67 (panomer)]. These 19F chemical shift values agree with the literature data (13). A similar reaction was also conducted with nonradiolabeled F,, OF,, and AcOF for comparison purposes.

Effect of Ozone on the Mercury and Tin(G) Fluorodopa Precursors

(4)

Neat irradiated oxygen (15 p,A x 30 min) was bubbled into the fluorodopa precursors in CDCl, (4) or CFCl, (6) and the residual solution analyzed with NMR. In each experiment, the entire target load of proton irradiated oxygen was bubbled into 150 pmol of either the mercury or tin fluorodopa precursor in 10 mL of solvent at room temperature. In the experiments with the tin precursor, the Freon was then evaporated and the residue reconstituted in CDCl, for ‘H and l19Sn NMR analysis. The 199Hg NMR spectra were recorded for the mercury samples in CDCl,. These spectra were compared with the corresponding standards at the same concentration, as a concentration dependence of ‘99Hg chemical shifts was observed. In addition, ozone, prepared using a Tesla coil, was bubbled into fluorodopa precursor solutions, as described above. The resulting products were analyzed by appropriate NMR techniques. The concentration of ozone generated with the Tesla coil was also determined by the iodometric titration technique.

Aromatic Electrobhilic Substitution Reactions with 18F Activity Ret overed jkm [‘sO]Oxygen f&s Targets 1

The 18F activities recovered from the [‘*O]Oz single-step and two-step irradiation processes were passed through a cartridge of NaOAc * 3HzO and the exiting [‘*F]acetyl hypofluorite (2) was reacted with a mercury derivative, namely N-(trifluoroacetyl)-3,4-dimethoxy-6-trifluoroacetoxymercurio-L-phenylalanine ethyl ester (4) in chloroform (9). The resulting intermediate was hydrolyzed with HI. The crude product upon purification by semi-preparative HPLC yielded pure 6-[“FlfluoroL-dopa (5). The chemical and radiochemical purities of the product were verified by analytical HPLC (282 nm UV and radioactivity detectors). The ‘H and 19F NMR spectra of this product, obtained after ‘*F decay, were identical to the reported data (9). FLUORODEMERCURATION

METHOD.

FLUORODESTANN~ATION METHOD. In this method, a trimethyltin derivative, N-formyl-3,4-di-r-butoxy carbonyloxy-6-(trimethylstannyl)-L-phenylalanine ethyl ester (6) was fluorinated with the irradiated target gas effluent ([18F]Fz or [‘*F]F, + [“F]OF, mixture) directly or with [18F]AcOF and the resulting 6-[‘*F]fluoro intermediate was hydrolyzed with HBr to yield 6-[“Flfluoro-L-dopa, as described previously (10). After the decay of 18F (-24 h), the product was analyzed by ‘H and 19F NMR. The data were found to conform with the literature results (10). As control experiments, AcOF was generated from 1% Fz (in oxygen or argon) and reacted with the mercury (4) and tin (6) precursors, as described above. These precursors were also reacted with 1% OF, (in oxygen or argon). The products isolated from all these reactions were analyzed by ‘H and 19F NMR.

RESULTS

AND

DISCUSSION

Two different types of electrophilic reactions, namely an addition reaction in a carbohydrate system and a substitution reaction in an aromatic system, were chosen to demonstrate the usefulness of the among such recovered ‘*F species and also to assess differences species derived from various target bodies. The substrate 3,4,6-triO-acetyl-D-glucal (1) was chosen for the addition reactions, while arylmercury and aryltin reagents were elected for the substitution reactions. The proportion of [18FJoxygen difluoride in the singlestep proton irradiation of [‘*O]O,/F, averaged 15%, whereas the two-step method in the gold-plated target provided 5% (1, 4). In a given reaction, the contributions of the minor constituent [‘*F]OFz, a reactive gas, in the presence of highly reactive [“F]F,, would be extremely difficult to evaluate. Hence, the selected substrates were also reacted with standard gas mixtures of 1% OF, or 1% Fz in argon and oxygen, separately, and the results were compared with those of the irradiated target gases before any conclusions were drawn. Further, all the products obtained by these reactions were characterized by NMR spectroscopic techniques rather than by chromatographic analysis.

I. Use of “F Actiwity Target in Electrophilic Identification ties recovered

from [“O]Oxygen Addition Reactions

and characterization from a single-step

of reactive ‘*O(p,n)‘sF

[18F]fluorinated reaction based

speon its

Reactivity

of 18F Reagents

from

“O(p,n)“F

561

Gas Targets

(1) F2,OF2

or AcOF

(2) H+ 1

FAG) FIG. 1. Reaction

of 3,4,6-tri.O.acetyl.D.glucal

chemical reactivity has previously been reported (3, 11, 15, 16). These preliminary studies involved the reaction of the “F activity with the triacetyl glucal 1. Based on 19F NMR studies (3), it was shown that the ratio of 2-fluorodeoxyglucose (2) (2-FDG) to 2-fluorodeoxymannose (3) (2-FDM) obtained in this reaction was quite similar to that obtained with the fluorinating agent derived from the neon gas target system. Thus, it was tentatively concluded that the fluorinating agent recovered from the oxygen single-step reaction might be F,. However, our direct investigation of the singlestep target effluents conclusively demonstrated the presence of OF, (1). These results demonstrate that characterization of fluorinating species based on chemical reactivity alone should be approached cautiously. On the other hand, it has been previously shown that the ratios of 2-FDG to 2-FDM resulting from the reaction with TAG are sensitive to the nature of the fluorinating agents (17). Hence, before characterizing any of the irradiated target gases, the reaction of TAG with F,, OF,, and AcOF was conducted (Fig. l), the intermediates hydrolyzed, and the final products analyzed by “F NMR to quantitate the resulting fluoro sugars (Table 1). The ratios of 2-FDG to 2-FDM observed for the reactions of F, and AcOF (Table 1) were identical to the values reported previously (2). It is, however, quite interesting to note that the ratios obtained for OF, and F, are also nearly equal. Thus, it is impractical to use the reaction of TAG for the differentiation of F, from OF,. Incidentally, this explains the previous observation of similar 2-FDG/Z-FDM ratios for the reaction of TAG with the electrophilic fluorinating agents generated in the [lHO]oxygen single-step target (85% [“F]Fz + 15% [“F]OF,) and “me(d,ol)“F reactions (100% [“F]F,) (3). It is advantageous to moderate the highly reactive F, into acetyl

TABLE 1. Effect of Fluorinating 2.FDGl2.FDM”

Agent on the Percentage

Percentage Fluorinating

agent

2-FDG 94 95 73 70

AcOF’

fOFC dF,” OF,

through

NaOAc

* 3H,O

67

of

of 2.FDM 6 5 27 30 33

a In these experiments, the fluorinating agent was bubled into TAG in Freon and the product after acid hydrolysis analyzed by 19F NMR to determine the relanve percentages of Z-FM3 and 2-FDM. h Prepared by passing 1% F, in 0, through NaOAc . 3H,O (2). ’ Prepared by passing 1% F, in argon through NaOAc e 3H,O. d 1% In oxygen.

2

3

(2-FDG)

(Z-FDM)

with fluorinating

agents.

hypofluorite, a mild fluorinating agent for “F radiolabeling processes (7). However, an analogous possibility for the conversion of OF, into AcOF is not known. Further, the effect of 15% OF, present in the [“0]02 single-step irradiated target gas on NaOAc - 3H,O used for the production of gaseous AcOF (2) is unknown. Hence, 1% OF, in argon was passed through a NaOAc -3H,O cartridge, and the effluent gas was reacted with TAG in Freon. The 19F NMR analysis of the product after hydrolysis indicated a ratio of 2-FDG/2-FDM, similar to that obtained with OF,, and different from that observed with AcOF in the same solvent (Table 1). These results suggest that OF, probably does not produce AcOF upon reaction with NaOAc * 3H,O. Nevertheless, the presence of OF, as a minor constituent in the single-step irradiated oxygen target gas does not seem to have any measurable effect on the ratios of 2-FDG/Z-FDM as evidenced by the data provided in Table 2. Further, little differences in yields and product ratios were observed between the lRF activities recovered from target bodies made of different metals, at least from the standpoint of the reaction with TAG. As expected, the ratio of 2-FDG/Z-FDM for the two-step [‘*O]oxygen method, which produces >99% [18F]Fz in an aluminum target body (l), was similar to that of the 2oNe(d,ol)18F reaction (Table 2). In any event, the data given in Table 2 clearly indicate that the 85% [“FjFz present in the singlestep [“O]O, target gases could be converted into [‘aF]AcOF.

II. Use of 18F Activity from [“O]Oxygen @s Target in Aromatic Electrophilic Substitution Reactions FLUORODEMERCURATION REACTION. The “F activities recovered from the [180]02 single-step and two-step methods were tested for the suitability of aromatic substitution reaction with an arylmercury derivative, namely N-(trifluoroacetyl)-3,+dimethoxy-&trifluoroacetoxymercurio-L-phenylalanine ethyl ester (4) (9). This mercury

TABLE 2. 19F NMR Percentages of 2.FDG/Z#FDM Produced from the Fluorination of TAG in Freon with Irradiated Gases” Percentage

Irradiation protocol [ “O]O,

Target Single-step

Aluminum Gold plated

body

2.FDG

copper

C ower

P41($2 Tg;-step e ,a

Aluminum Aluminum

d Irradiated target gases were passed through (2) before reacting with TAG.

a cartridge

96 96 96 95 93 of NaOAc

of

2.FDM 4 : 5 7 . 3H,O

562

A. Bishop

et al.

(1) [18FjAcOF .

(2) HI

4

5 FIG.

2. Fluorodemercuration

reaction-synthesis

precursor, used in the synthesis of 6-[‘sF]fluoro-L-dopa, is quite sensitive to the nature of the fluorinating species, and only acetyl hypofluorite has been shown to be effective for this transformation (Fig. 2) (9). As a logical step, this fluorodemercuration reaction was attempted with the activity obtained from a single-step oxygen gas target since the conversion of this 18F activity into [‘8F]AcOF has been demonstrated with the TAG reaction (Table 2). Accordingly, various target bodies were utilized for the single-step production of ‘*F activity, which in turn was converted into [‘8F]AcOF and reacted with the mercury derivative of L-dopa (4). The radiochemical yields obtained for 6-[“F]fluoro+dopa are provided in Table 3, wherein quite contrasting results are clearly discerned. For example, aluminum, silver, and copper target bodies provided ‘*F activities from which expected radiochemical yields for 5 were obtained. In stark contrast, the gold-plated and the nickel targets nearly failed to provide any 6-[‘8F]fluoro-L-dopa. A similar result has previously been noted (21). In any event, the presence of 15% [‘sF]OFz in the single-step irradiated target gases were not totally detrimental to the fluorodemercuration reaction as evidenced by the data for the aluminum, silver, and copper target bodies (Table 3). Analogously, the two-step [‘sOloxygen irradiations in an aluminum target yielded “F activities usable for 6-[‘sF]fluoro-L-dopa production, whereas that from gold-plated target responded poorly (Table 3). A clear explanation for this observation is unknown at this stage. However, an attempt has been made to rationalize this behavior of the gold-plated and the nickel target bodies and is presented later in this discussion. FLUORODESTANNYLATIONREACTION. Unlike the mercuryprecursor reaction described above, the fluorodestannylation reaction tested with N-formyl-3,4-di-t-butoxycarbonyloxy-6-(trimethyl~ stannyl)-L-phenylalanine ethyl ester was quite insensitive to the irradiation protocols, target body material, and the nature of the fluorinating agents. Excellent radiochemical yields of 6-[‘sF]fluoroL-dopa were realized with this tin precursor for all target bodies and irradiation conditions. Complete details in this regard have recently been reported from this laboratory (10).

of 6-[ “F]fluoro+dopa.

Further, AcOF (from F, in oxygen) and OF, (in oxygen) were reacted with the mercury and tin dopa precursors (see Experimental). The same reactions were conducted with the same fluorinating agents (1%) in an inert gas (argon) as controls. No significant differences in yields or product distributions were perceptible between the use of oxygen and argon as diluents for the fluorinating agents. This strongly suggests that these reactions are rather insensitive toward oxygen. This conclusion is partly supported by the data presented in Table 2. For example, the 18F activity from the deuteron irradiation that was present in an atmosphere of neon gave a ratio of 2-FDG/2-FDM that was rather similar to those of the oxygen gas irradiation reactions. The mercury precursor 4, however, failed to yield 6-fluoro-L-dopa with OF, diluted either with oxygen or argon. This is in conformity with the previous observation that the mercury precursor is quite selective in its reaction with fluorinating agents (9). In contrast, the tin precursor 6 reacted with OF, with ease, irrespective of the diluent gas (10). The data in Table 3 clearly indicate the suitability of single-step ‘sF activities generated in aluminum, silver, and copper target bodies for the fluorodemetallation reaction. On the other hand, 18F activities recovered from both gold-plated and nickel target bodies provided only very poor yields of 6.[“Flfluorodopa. Such conspicuous differences perhaps stem from the variations in oxygen specie(s) that are generated in these target bodies. Irradiation of oxygen is known to produce ozone (8, 12) and it could probably be a prime candidate responsible for the failure of the demercuration reaction. To test this hypothesis, experiments were performed to observe the effect of ozone on the mercury precursor. TABLE 3. Synthesis mercuration Method

Irradiation protocol”

Target

[“O]O,

Single-step

[“O]O,

Two-step

III. Interaction of the Oxygen Species Generated in the Target C&es with Mercury (4) and Tin (6) Precursors Exclusion of oxygen during most organic reactions is required since various substrates and reagents are quite sensitive to it (18). As the single-step oxygen gas target activity is present in oxygen, the reactions that utilize such irradiated gases should be appropriately selected. The electrophilic addition and the aromatic fluorodematallation reactions chosen to characterize the fluorinating agents derived from the oxygen gas targets were first tested for their sensitivity toward oxygen. Standard gas mixtures of 1% F, in oxygen as well as 1% OF, in oxygen were prepared and reacted with TAG.

of 6.[ “F]Fluoro+dopa

Aluminum Silver Copper Nickel Gold-plated copper Aluminum Gold-plated copper

body

Using

Average radio. chemical yieldb (%)

Fluorode-

Number experiments averaged

11 8 8 1

2 2 2 2


11

3

5

a The irradiated target gases were passed through NaOAc . 3H,O for the production of [‘sF]AcOF (2). h At EOB. Literature (9) yield for this reaction usmg *‘Ne(d,cu)‘sF is 11% (EOB).

4

cartridge activities

of

Reactivity

of ‘sF Reagents

from

‘~O(~,IX)‘~F

Gas Targets

563

h

-1320.0

I

a)

-1260

-1280

-1300

4

-1316.6

-1320

-1340

-1360

-1380

-1360

-1260

-1280

-1300

PPM

I,

I

-1240

r

I

r

II

-1280

I

-1320

-1340

-1360

-1380

-1400

PPM

I

-1320

1

’

I

I

-1360

c

I

1

-1400

-1240

-1320

-1360

I

-1400

PPM

PPM FIG. 3. ‘99Hg NMR of the derivative oxygen in an aluminum target; (d)

-1280

4. (a) After after reaction

reaction with ozone; with proton&radiated

Ozone generated with a Tesla coil was bubbled into a solution of the mercury precursor 4 in CDCl, (ratio of 0, to mercury precursor = 1:lO) and the “‘Hg NMR of the resultant product is shown in Fig. 3a. A comparison of this spectrum with that of the standard mercury precursor (Fig. 3b) strongly suggests a chemical transformation of the precursor due to the reaction of 0,. A similar result was also obtained when a NaOAc * 3HzO cartridge was interposed between the ozone generator and the mercury precursor solution, suggesting that 0, is not decomposed by sodium acetate. However, when ozone was passed through a heated metal tube and the exiting gas bubbled into the mercury precursor solution, no decomposition of the starting material was observed via ly9Hg NMR. This is primarily due to the metal catalyzed thermal decomposition of ozone (14). This experiment also proves that molecular oxygen alone would not decompose the mercury precursor, for the decomposition of ozone yields only Oz. In any case, these results indicate the sensitivity of the mercury precursor toward 0,. A series of experiments was also performed to determine the effect of ozone, generated in the target, on the mercury precursor 4. An aluminum and a nickel target were chosen in this regard. These target systems were kept completely free of fluorine. Pure [i60]oxygen gas was proton-irradiated (5.5-7.5 PAh), and the target gas was bubbled into a solution of the mercury precursor 4 in CDCl,. Representative 199Hg NMR spectra for the products obtained from identical reactions of the mercury precursor 4 with oxygen species generated in the aluminum and nickel targets are shown in Fig. 3. Interestingly, a comparison of Fig. 3b (standard) and Fig. 3c indicates that the oxygen species generated in the aluminum target did not affect the mercury derivative. However, the corresponding product from the nickel target reaction exhibited a downfield shift of 4.7 ppm (Fig. 3b vs. Fig. 3d), again suggesting a chemical transformation of the mercury precursor. It is surprising, however, that while comparable amounts (5-10 kmol) of 0, are formed (see Experimental) in all the target bodies,

(b) standard; (c) after reaction oxygen in a nickel target.

with

proton-irradiated

the reactivity of the irradiated oxygen differs considerably. Perhaps the irradiated oxygen gas from nickel and gold-plated targets may contain other oxygen species (e.g., i Ag 0,) along with ozone, and these species are responsible for the chemical transformation of the mercury precursor. It can only be speculated that this process probably also occurs in the carrier fluorine-added nickel and gold-plated targets, and the mercury precursor is, thus, consumed, causing poor yields of 6-[‘*F]fluoro-L-dopa (Table 3). The chemical identity of the mercury derivative transformed by the oxygen species was not further investigated since good radiochemical yields of 6-[‘“F]fluoro-L-dopa were obtained with other target bodies (Table 3). In stark contrast, the tin dopa precursor 6 was highly insensitive to ozone. Even when a 4 molar excess of 0, was bubbled into a solution of 6 in Freon, no decomposition of the tin precursor was observed, as evidenced by ‘H and ““Sn NMR (Fig. 4). Further, [‘60]oxygen irradiated in nickel or other target bodies had little effect on the tin precursor.

CONCLUSIONS The chemical reactivity of the i8F electrophilic species generated via the ‘sO(p,n)‘sF reaction depends, to some extent, upon the material of construction of the target bodies. Of the electrophilic reactions tested, addition across a double bond and aromatic fluorodestannylation were quite insensitive toward the nature of the target body from which the i8F activity was recovered or the irradiation protocols. On the other hand, the fluorodemercuration reaction essentially failed with a gold-plated target body. Hence, caution is warranted in the choice of material for the construction of the “O(p,n)‘“F electrophilic target system. In any event, because of the consistent ‘sF recovery (1) and suitability of the activity for substitution and addition reactions along with its low nuclear activation properties and ease of fabrication, aluminum target bodies

564

A. Bishop

et al.

-25.2

b)

-25.2

I

100

I

I

50

0

I

-50

I

I

-100

-150

PPM FIG.

4.

119Sn

NMR

of the

stannyl

derivative

are perhaps the most appropriate for the production of [“FIF,. Target bodies made of aluminum have been in routine use in our laboratories for nearly 5 years for the production of ‘*F electrophilic reagents (1). We thank the staff of the Biomedical Cyclotron for their technical support. This work was supported in part by Department of Energy Grunt DE-FC0387-ER606 J 5, NJH Grant PO J -NS- J 5654, and donations from the Jennifer Jones Simon and Ahmunson Foundations.

References 1. Bida G. T., Hendry G. O., Bishop A. J. and Satyamurthy [Fluorine-18]FZ production via low energy proton irradiation 1810, plus F,. Proc. IVth International Workshop on Targeay Chemistry. PSI Villigen, Swtzerland (Edited by Weinrelch 2. Bida G. T., Satyamurthy N. and Barrio J. R. (1984) The Z-[‘sF]fluoro-2-deoxy-D-glucose using glycals: a reexamination. Med. 25, 1327. 3. Bida 0. T., Satyamurthy N., Wieland B. and Ruth T. J. production via low energy proton bombardment of [‘sO]O,

N. (1991) of [oxygen-

and Target R.), p. 130. synthesis of J. Nwl. (1987) F-18 + F,. Proc.

6. (a)

After

reaction

with

ozone;

(b)

standard.

2nd Workshop on Targetry and Target Chemistry. (Edited by Ruth T. J., McQuarrie S. A., H&s 4. Bishop Lab&d Ph.D.

A. J. (1994)

Prodwtion,

Identification,

Electrophilic

Fhmimting University

Agents from of California,

Dissertation,

Heidelberg, F.), p. 19.

Germany

and Chmacteri7ation of IRF Low Volume Target Systems. Los Angeles, California.

5. Boyd A. W., Wilis C. and Cyr R. (1970) New determination of stoichiometry of the iodometric method for ozone analysis at pH 7.0. Anal. Chem. 42, 670. 6. Chirakal R, Adams R. M., Firnau G., Schrobilgen G. J., Coates G. and Garnett E. S. (1995) Electrophilic ‘*F from a Siemens 11 MeV protononly cyclotron. Nucl. Med. Biol. 22, 11 I. 7. Guillaume M., Luxen A., Nebeling B., Argentini M., Clark J. C. and Pike V. W. (1991) Recommendations for fluorine-18 production. Appl. Radiat. Isot. 42, 749. 8. Kircher J. F., McNulty J. S., McFarling J. L. and Levy A. (1960) Irradiation of gaseous and liquid oxygen. Rad. Res. 13, 452. 9. Luxen A., Perlmutter M., Bida G. T., Van Moffaert Satyamurthy N., Phelps M. E. and Barrio J. R. (1990) tomated production of 6-[‘aF]fluoro-L-dopa for human Appl. R&at. lsot. 41, 275. 10. Namavari (1992)

M., Bishop Regioselective

A., Satyamurthy N., radiofluorodestannylation

Bida

G., Cook J. S., Remote, semiaustudies with PET. G. and with

Barrio [“F]F,

J. R. and

Reactivity of “F Reagents from “O(p,n)“F

Gas Targets

565

[“F]CH,COOF: A high yield synthesis of 6-[‘sF]fluoro-L-dopa. A@. R&at. Isot. 43, 989. 11. Nickles R. J., Daube M. E. and Ruth T. J. (1984) An ‘so2 target for the production of [‘sF]F,. Int. .J. Appl. R&at. Isot. 35, 117. 12. Parks N. J., Peek N. F. and Goldstein E. (1975) The synthesis of 13Ns labeled atmosphere gases via proton irradiation of a high pressure oxygen target. Int. J. Appl. R&at. ht. 26, 683. 13. Phillips L. and Wray V. (1971) Stereospecific electronegative effects: Part 1. The 19F nuclear magnetic resonance spectra of deoxyfluoro-Dglucopyranoses. J. Chem. Sot. (B) 1618. S. D. and Zaikov G. E. (1984) Ozxw and ItsReactions with 14. R;tzumovskii Organic Compounds. Elsewer, Amsterdam. 15. Ruth T. J. and Wolf A. P. (1979a) Absolute cross sections for the proreaction. Radiochim. Acta 26, 21. duction of ‘sF via the ‘sO(~,n)‘8F 16. Ruth T. J. and Wolf A. P. (1979b) “Small” accelerator, radionuclide and radiopharmaceutlcal production. IEEE Trans. Nucl. Sci. NS-26, I,

17.

I”.

Satyamurthy

N., Bida

G. T., Padgett

H. C. and

Barrio

J. R. (1985)

On

the stereoselectivity glycals: The synthesis 4, 489.

of fluorine and acetyl of 2-deoxy-2-fluorohexoses.

hypofluorite additions to J. Carbohydr. Chem.

18.

Shriver D. F. (1969) The Manipulation Graw Hill, New York.

19.

Solin Label.

20.

Sunderland trophilic presynaptic

21.

Sunderland J. J., DeJesus diolabeling of pre-synaptic from an ll-MeV proton

22.

Wieland B. W., Bida G. T., Padgett H. C. and Hendry G. 0. (1990) Current status of CT1 target systems for the production of PET radiochemicals. Proc. 3rd Workshop on Targemy and Target Chemistry. Vancouver, Canada (Edited by Ruth T. J.), p. 34.

0. and Compd.

of Air-Sensitive

Bergman J. (1986) Production Radiopharm. 23, 1202.

of ‘“F-F,

Compounds. from

Mc‘*Oz.

J.

J. J., DeJesus 0. T., Solin 0. and Nickels R. J. (1991) ElecF-18 from an 1 l-MeV proton cyclotron for the radiolabeling of dopaminergic tracers. 1. Label. Compd. Radiophann. 30, 92. 0. T., Solin 0. and Nickles R. J. (1990) Radopaminergic tracers with electrophilic F-18 cyclotron. J. Nucl. Med. 31 (Suppl), 901.