Download Impact of Target-Gas Purity on F-18 Form in 18F-F2 Production with Neon/Fluorine Target and more Study notes Nuclear medicine in PDF only on Docsity!
The production of ‘8F-F2of high specific activity
using the @°Ne(d,a)'8Freaction has been of considerable
interest in light ofits use in the synthesis of 2-['8F] flu
oro-2-deoxy-D-glucose Q8FDG), a tracer that has been
used successfully in the quantitative measurement of
regionalbrainglucosemetabolism(1). Moreover,itsuse
in clinicalresearchrequiresa highlyreliableproduction
method.
We have recently described a simplified system for
‘8F-F2production that eliminates the need for a high
vacuum system and the handling of pure fluorine gas (2).
This system, although manually operated, eliminated
problems of direct handling of radioactive targets and
allowed for the development of a remotely operated
synthetic method for ‘8FDG(3). A major factor in the
simplificationof thissystemwasthe useof a commer
cially available mixture of 1%fluorine in neon,which
was subsequently mixed with research-grade neon to
ReceivedNov. 26, 1979;revisionacceptedFeb.22, 1980. For reprintscontact:Alfred P. Wolf, PhD, Dept. of Chemistry, BrookhavenNatjonal Laboratory, Upton, NY I 1973.
provide a target gas with the required amount of carrier
‘9F2.The target, which typically contained 0.1% F2 (‘—-‘
60 zmol), and this ‘8F-F2production system have been
used routinely for over a year with no difficulty to pro
duce ‘8FDGfor human neurological studies.
Recently, however, the production of F- 18 as ‘8F-F
decreased rapidly with a concomitant dramatic increase
in the production of F-l8-labeled, chemically inert
gaseous material(s)*. The serious consequences of this
problem to the reliability of the production of ‘8F-F2and
of labeled compounds requiring its use, necessitated an
investigation of the factors influencing the chemical form
of F-18 produced from the 20Ne(d,a)'8F reaction in the
presence of carrier F2. Additionally, it was felt that er
ratic behaviorof this target systemreportedby other
laboratories reinforced the needfor a thorough investi
gationof itscauses.The factorsneedingto beidentified
and reported on in this paper include: (a) assay of target
gases for contaminants that react with F2, resulting in
the production of large quantities of chemically inert
F-18-labeledproducts at the expenseof ‘8F-F2;@(b)de
scription of the analytical methods for assaying target
758 THE JOURNAL OF NUCLEAR MEDICINE
TheEffectof Target-GasPurityontheChemicalFormof F-18during18F-F
ProductionUsingthe Neon/FluorineTarget
Gerald T. Bide, Richard L. Ehrenkaufer, Alfred P. Wolf, Joanna S. Fowler, Robert R. MacGregor, and Thomas J. Ruth
Brookhaven NatlonalLaboratory, Upton, New York
Irradiation of gas mixtures of F2/Ne (@°Ned.@18F)which containedpercent
levels (>0.1 % ) of N2,C02, or CF4resultedIn the productionof unacceptable1ev
cia of F@18-labeledNF3andCF4at the expenseof 18F-F2.Analyticalgaschromato
graphicmethodshavebeendevisedto determIflecontaminantlevelsin the target
gas as well as in the productsarisingfrom them. Commercialmixturesof I % f2/
Na,pureF2,andneonhavebeenanalyzedfor contaminants(N2,02, CO,C02,and
CF4) andfoundto vary widely In the levels of these Impuritiesfrombatchto batch.
TheN2levóisin the I % F2/NemixturesvarIedfrom 0.039to 0.49%, andthe CO
levelefrom 0.028 to 0.13%. No detectable impurities were found In the neon (Re
search Purity), but F2 was found to contain “11% CF4. Reproducibly high yields
of 18F-F2are obtainedif the levels of N2, C02, and CF4 In the final target gas mix
turfare<0.01% andcarrierF2is‘@‘0.1%. HydrocarbonsandCOWerenotdetect
ed In our gas mixtures, but would also be expected to decrease yields of ‘F-F2.
,I Nucl Med 21: 758—762, 1980
determined determined sh. 0.028 none detected 0. none detected none
0. 0.
(dateSampleGas Source received)designation % N2 % 02 % CO2 % CF
BASIC SCIENCES
RADIOCHEMISTRYAND RADIOPHARMACEUTICALS
gas mixtures; (c) the identification and quantitation of
the inert F-I 8-labeled products from the Ne/F2 target;
and (d) conditions and specifications for a reproducible
and effective gaseous target composition for ‘8F-F
production.
MATERIALS AND METHODS
Irradiation conditions, targetry, and gas handling system. All irradiations were performed at the BNL 60-in. (I.5-m) cyclotron. TheInconel600two-porttargetandflow-throughloadingsystem describedpreviously were used(2). The 23-MeV deuteron beam from the cyclotron wasdegradedto 14.0MeV beforeentering the target gas.Sufficient target gas (“—25atm) is usedto degradethe beambelowthresholdunlessotherwiseindicated.Irradiations were carriedout at constantdoseanddoserate—namely,10 @iAfor 10 mm (theoretical F-l8 yield: 50 mCi) (4)—with typical F-18 re coveriesbeing 40-50% of the theoretical value. Although the In conel target wasusedin thesestudiesasa matter of convenience, production targets must havehighly polishedpure nickel surfaces wherever exposedto the F2/Ne mixture (2). The contentsof the target after irradiationwereanalyzedby purging them through a seriesof traps containing KI, sodalime, and charcoal as previously described (2), with modifications to accommodate various sampling vesselsas described in the fol lowing section. Although one can postulate other fluorine-con taming compounds that would be hydrolyzed in water or would oxidize KI, we support our identification of the predominant chemical form recovered from the target as F2 by its chemical reactivity with 3,4,6-tri-O-acetyl-D-glucal to producethe difluoro adducts (5). Work is in progressto develop new analytical tech
niques for F2and the minor reactive gaseousproducts. Target gases. FLOURINE IS A HIGHLY TOXIC AND REACTIVEGAS.THEREADERISDIRECTEDTOA SE
RIESOFARTICLES(6,7)ONTHEMANIPULATIONAND
HANDLING OF GASEOUS FLUORINE IN ORDER TO
BECOME FULLY AWARE OF THE HAZARDS IN HAN
DLING THIS DANGEROUS GAS.
The variouscommercialF2/Ne mixtures are identified for referencein Table 1. Undiluted F2 (3.4 atm)t and undilutedNe (Research Grade)t were also obtained commercially. Target-gas analyses.Gas chromatographic analysis of target gases(before and after irradiation) was used to identify and quantitate N2, 02, CO, CO2. CH4, and CF4 contaminants. Sam piesof target gaswereanalyzeddirectlyfrom the target,aswell asafter removal of fluorine, to ensureagainst the introduction of artifacts due to the possiblereactivity of the fluorine with the columnmaterials.Analyseswere performedusingan analyzer equipped with a thermal-conductivity detector and digital pro cessorto quantitatemasspeakareas.Samplesof the target gas werecollectedin flow-throughgasbulbsfitted with Burrellseals for purposesof analytical screening.Sampleswere withdrawn in gaslight syringes and injected onto the column. Conditions were asfollows (contaminant gas,column characteristics,temperature, He flow rate, retentiontime): for oxygen,molecularsieve5A (60—80mesh),6 ft X 1/@in., 50°,21 cc/mm, 3.0 mm; for nitrogen, molecular sieve,50°,21 cc/mm, 5.7 mm; for carbon monoxide, molecularsieve,50°,50 cc/mm, 9.4 mm; for carbondioxide, silica gel (30—60mesh),6 ft X ‘/@in., 75°,31 cc/mm, 10.6mm;andfor carbontetrafluoride,silicagel, 35°,21 cc/mm, 6.6 mm. Whenabsolutevalueswererequiredandwhentheexclusionof air from the sample was necessary,gaseswere collected in cali bratedglasssamplingloops.The loopswereattachedon-lineto the column, the dead-volumegassweptout, and the samplepassed
1% F2/Ne Matheson(12/77) (^) (A) 0.040 Sh.t 0.068 none detected 1% F2/Ne Matheson(4/78) (^) (B) 0.046 sh. not not
1 % F2/Ne Matheson (8/79) (^) (C) 0.
2% F2/Ne Homemade 1% F2/Ne@ Homemade'
(D)
(E)
0. none detected 1% F2/Ne@ Matheson(1/79) (^) (F) 0.49 0. detected nonedetected none 11.0••
Neon1 Matheson (^) (G) none detected
none detected not
none detected F2 AirProducts^ (H) not•@ not determined determined determined
. 1 % F2/Ne mixtures were purchased in size 3F gas tanks at a pressure of 35 atm. t Appeared as a shoulder on the neon peak and was too small to be integrated. @ These “homemade―mixtures were prepared using Air ProductsF2(Tank H) and MathesonResearch Purity neon (TankG).
IIInthepreparationofTankE,thefluorinewaspassedthrot4@a liquid-N2traptoremoveCF4andCO2.
SAll premixed gases were prepared with ResearchPurftyneon except Tank F, which was inadvertentlysupplied with purified neon. I Research Purity neon.
.. Not directly determined but can be estimated from analysis of D or E.
Volume 21, Number 8 759
TABLE1. IDENTIFICATION,SOURCE,AND PURITYANALYSISOF TARGETGASES
OFIRRADIATEDGASESRunTABLE2. ANALYSES
No. Target gas%18@4@2% ‘8F-NF3^ % 18F-CF
COMPOSITiONBEFORE TABLE3.TARGET GAS
ANIMPURECOMMERCIALAND AFTERIRRADIATIONOF
MIXTURE(TANKF) GAS
PLUSNEON(TANKG)
afterTarget-gasirradiation Amountbefore Amount irradiationcomposition(zmol) (zmol)
afterirradiationirradiationGas^ Amount beforeAmount
(mmol)(mmol)
BASIC SCIENCES
RADIOCHEMISTRYAND RADIOPHARMACEUTICALS
1 TankF(12.7atm) 29 54.1 16. 2 TankF(1.6atm) 29 50.7 20. Neon(24.3atm) 3 SameasRun2 20 62.2 17. 4 TankC(1.2atm) 98 1.8 0. Neon (24.3 atm) 5 TankD (1.02atm) 15 0 85 Neon (25.5 atm) 6 TankE(1.O2atm) 93 0 7 Neon (25.0 atm)
. Pressure represents ‘-@‘@‘90%of thick-target conditions.
CO2 (or some radiolytic decomposition products from
them) are the source of nitrogen and carbon atoms in the
NF3 and CF4.
More specifically, the mass analysis of NF3 produced
as compared with N2 consumed, and of CF4 produced as compared with CO2 consumed, showed that 37.0% of the N2 consumed was accounted for as NF3, and 34.0% of the CO2 consumed was accounted for as CF4. The fluorine balance, also performed, indicated that most (85.5%) of the F2 consumed was present in the gases NF and CF4. As additional evidence that CO2 (or a radio lytic product, such as CO) can serve as a carbon source for ‘8F-CF4formation during irradiation, CO2 was de liberately added to an otherwise pure 1% F2/Ne mixture
(Tank C). This resulted in 87% of the F-18 products
being in the form of ‘8F-CF4.It should be reemphasized
that Tank C consistently gave a high yield of ‘8F-F2on irradiation, with very little labeled inert gas being
formed. The results of this experiment are summarized
in Table 4 and show that the formation of CF4 is ac
companied by a decrease in the amount of CO2 and F2,
with approximately 2 millimolësof F2consumed for each
millimoleof CF4produced.Studiesarecurrentlyunder
way to determine the amounts of other radiolysis prod
ucts (such as CO and 02) from this reaction. Note that whereas only about two-thirds of the F2 was converted
to CF4, the ‘8F-CF4accounted for 87% of the radioac
tivity.
Gas radiochromatography on the inert gas mixture
revealed a minor (1—3%)peak having a retention time
between that of N2 and 02 on Porapak Q. This radio
isotope was isolated and subjected to gamma-ray spec
troscopy. The only photon present below 1500 keV had
an energy 1294 keV and this corresponds to a transition
in the decay of Ar-41. This argon is probably the product
of the (d,p) reaction on the Ar-40 (99.60% natural
abundance) that was present as a trace contaminant in
the 1% F2/Ne (Tank F). We note that the reported
half-life of Ar-41 is 1.83 hr, which is 109.8 m, the ac
cepted half-life of F-l8 (12).
Gas samples taken immediately after bombardment
showed short-lived inert gaseous activity having a t1,i
of <1 mm. While these nuclides were not conclusively
identified, they appear to be Ne-l8 and Ne-19 (t112 =
1.7 and 17 sec, respectively). These short-lived activities
did not interfere with our studies, since analysis was not
TABLE4. C02/F2 TARGETGAS ANALYSIS
BEFOREAND AFTERIRRADIATION
N2 164 90
Co2 32.7 5.
F2 168.6 50.
CF4 none detected 9. NF3 nonedetected 54. Ne 79X103 79X
. Target gas consisted of 5.2 atm 1 % F2/Ne premixed gas fromTankF andpressurizedupto 25.0 atm withneonina targetvolumeof 78.3 ml. Numbersrepresentaveragevalues for threerunseach withs.d. <10%.
C02t CF4t F2t
none detected 0.
. See text for experimental details. t Analyzed by gas chromatography. @ Analyzed by titration with sodium thiosulfate.
Volume 21, Number 8 761
BIDA, EHRENKAUFER, WOLF, FOWLER, MACGREGOR, AND RUTH
made before end of bombardment + 30 mm.
CONCLUSION
Small quantities of N2 and C02, which do not react
with F2 under normal conditions, become highly reactive with F2undertheradiationconditionsof F-l 8 produc
tion. The present study had as its aim the identification
of factors contributing to the formation of gaseous F-
compounds during F-I 8 production. A detailed study of
the mechanisms of NF3 and CF4 formation under ra
diation conditions is currently under way here.
This work shows that irradiation of a 0.1% F2/Ne
mixture that contains <0.01% ofN2, CO2. and CF4 gives
excellent yields of ‘8F-F2during production conditions.
This level of purity should be easily met by commercially
available mixturesof 1% F2/Ne and neonwhen the
standards for Research Purity are maintained by the
supplier. However, we suggest communication with the supplier during procurement of gases and gas mixtures
for the F2/Ne target, to emphasize the importance of
eliminating contaminants. It is especially important that
nitrogen contamination be minimized, since this gas
cannotbeaseasilyremovedasCO2or CF4. Moreover,
all other sources of nitrogen and carbon (air, pump oil,
etc.) in the system should be removed or avoided. The
F-18-labeled gaseous compounds will, of course, depend
on the contaminants in the F2 and Ne, and it is possible
that contaminants other than those we have studied may
be encountered in other batches of gases. These may lead
to other F-l 8-labeled products at the expense of ‘8F-F
because of the highly efficient hot-atom and thermal
reactions of fluorine atoms and ions with the contami
nants and their products of radiolysis. In addition, ab
sorbed water and water vapor must be avoided in any
part of the target and gas-delivery system so that the
‘8FF2is not converted to H18F. The presence of oxygen
at the levels we report here does not adversely affect
production. Further studies are in progress on the reac
tion of fluorine atoms with oxygen.
FOOTNOTES
- We use the term “inert gases― to refer to those compounds that did not reactwith water,potassiumiodide,sodalime, or olefinsunder the conditionswedescribehere. t Air Products Co., special order. t MathesonGas ProductsCo. ‘Hewlett-Packard5830A I Hitachi RMU 7
I Hewlett-PackardModel 7620A 4* Princeton Gamma-Tech ft Canberra
ACKNOWLEDGMENTS
The authorsare grateful to Dr. David Christman for running the massspectraandto Drs. BarclayJonesand DavidSchlyerfor helpful discussions.They also thank Robert Carciello, Clarence Barrett, DonaldWarner,ConradKoehler,andRichardBeckerfor suggestions andassistancewith technicalproblemsencounteredduring this work. The workwasperformedat BrookhavenNationalLaboratoryunder a contract with the U.S. Departmentof Energyand supportedby its Officesof BasicEnergySciencesandofEnvironmentalandBiomedical Research,and by NIH Grant No. 9 ROI NS-15380.Supportfor one of us(RLE) wasprovidedasa subcontractto BNL from theUniversity ofCalifornia at LosAngeles,NIH Grant No. 5 RO1 GM-24839 (Dr. David KuhI, Principal Investigator).
REFERENCES
I. REIVICH M, KUHL D, WOLF AP, et al: The [‘8F1fluoro deoxyglucosemethod for the measurementof local cerebral glucoseutilization in man. Circ Res 44:127-137, 1979
2. CASELLA VR, Ino, T, WOLF AP, Ct al: Anhydrous F-18- labeled elemental fluorine for radiopharmaceutical prepa ration. J Nuci Med 21:750—757, 1980 3. FOWLERiS, KARLSTROMK, KOEHLERC, et al: A hotcell for the synthesisof labelled organic compounds. Trans Am Nude Soc In press
- CASELLA VR, IDo T, WOLF AP: Production of anhydrous ‘8Ffor nuclearmedicine.J Label Cmpd Radiopharm XIII: 209, 1977 (abst) 5. IDo T, WAN C-N, CASELLA V. et al: Labeled 2-deoxy-D- glucose analogues. ‘8F-labeled 2-deoxy-2-fluoro-D-glucose, 2-deoxy-2-fluoro-D-mannose, and I 4C-2-deoxy-2-fluoro D-glucose. J Label Cmpd Radiopharm XIV: 175-183, 1978
- Maiheson Gas Data Book, Rutherford, Ni, Matheson Co., Inc.,197l,pp2ll—2l 7. Chemical Hazards Bulletin, No. C-16, Fluorine, New York, American Insurance Association 8. STOCKLIN G, CACACE F, WOLF AP: Radiogaschromato graphic ‘4C-und 1‘C-markierteraliphatischer Kohlenwas serstoffe und amine. Z Anal Chem 194:406-416, 1963
- WELCH Mi, WITHNELL R, WOLF AP: An automatic GLPC aparatus for the analysisof organic compoundslabelled with short-lived radioisotopes.Chem Instr 2:177-188, 1969 10. MASSONNE J: Gas chromatographic separation of nitrogen tifluoride from carbon tetrafluoride on Porapak Q. Analysis of nitrogen trifluoride. Fresenius'Z Anal Chem 235:341-344, 1968 II. BEATTIE WH: Mass spectral intensities of inorganic fluo rine-containing compounds.Appl Spectroscopy 29:334-337, I 975 12. Table ofisotopes, 7th Edition, Lederer CM, Shirley V5, eds, New York, iohn Wiley and Sons, I 978
762 THE^ JOURNAL^ OF^ NUCLEAR^ MEDICINE
