Deli e able D10.4
Repo on ion sou ce e iciency
o a 11C PET ISOL beam
This p ojec has ecei ed unding om he Eu opean Union’s Ho izon 2020 esea ch and
inno a ion p og amme unde g an ag eemen No 101008571 (PRISMAP). This
documen e lec s only he iew o he au ho (s). The Agency is no esponsible o any
use ha may be made o he in o ma ion i con ains.
Deli e able D10.4
ii
P ojec Ac onym
PRISMAP
P ojec Ti le
The Eu opean medical iso ope p og amme: P oduc ion o high pu i y
iso opes by mass sepa a ion
G an Ag eemen No.
101008571
Topic
INFRAIA-02-2020: In eg a ing Ac i i ies o S a ing Communi ies
P ojec s a da e
01 May 2021
Na u e
Repo
Dissemina ion le el
Public
Due da e
M52
Da e o deli e y
M52
Lead pa ne
MedAus on
Con ibu ing pa ne
CERN
Au ho s
N. Gambino, M. O solic, L. Penescu, C. Schmi ze (MedAus on)
Re iewe s
T.E. Cocolios (KULeu en), S. Ro he (CERN), M. Manzola o (INFN)
Poin o Con ac
C. SCHMITZER
Ins i u ion
MedAus on
E-mail
Claus.schmi ze @medaus on.a
Phone
+4366480878839
© PRISMAP 2021. This wo k is licensed unde a C ea i e Commons
A ibu ion-NonComme cial-NoDe i a i es 4.0 In e na ional License.
Deli e able D10.4
iii
Re ision His o y
Ve sion
Da e
Au ho
Commen
0.1
09.07.2025
M. O solic, L. Penescu, C. Schmi ze
Fi s d a
0.2
01.08.2025
M. O solic, L. Penescu, C. Schmi ze
Final d a o e iew
0.3
14.08.2025
S. Ro he, M. Manzola o, T.E. Cocolios
In e nal e iew
0.4
21.08.2025
M. O solic, L. Penescu, C. Schmi ze
Re ised e sion
0.9
25.08.2025
V. Gob y, K. Leu gen
Re iew and o ma ing
0.91
28.08.2025
M. Manzola o, K. Leu gen
Ve sion o inal e iew and app o al by
he PRISMAP gene al assembly (GA)
1.0
29.08.2025
K. Leu gen
Final e sion, app o ed by he GA
Deli e able D10.4
i
Con en s
Abb e ia ions, Pa icipan sho names
Abb e ia ions
Pa icipan sho names
Lis o Figu es ii
Lis o Tables ii
Summa y 1
1. Con ex 2
2. Mo i a ion 2
2.1 Limi a ions in cu en clinical se up 2
2.2 Expec ed a ge ou pu 3
2.3 Po en ial o e all in eg a ion 3
3. Supe nanogan es esul s 4
3.1 Conclusion and possible use o Supe nanogan 6
4. VADIS MK7 es esul s 6
4.1 Sou ce es ing wi h noble gas mix u e 7
4.2 Sou ce uning o ope a ion wi h [C-12]CO2 8
4.3 E iciency measu emen wi h [C-13]CO2 11
5. COMIC es esul s 11
6. Conclusion and ou look 14
Re e ences 15
Deli e able D10.4
Abb e ia ions, Pa icipan sho names
Abb e ia ions
COMIC
Compac Mic owa e and Coaxial
EBIS
Elec on Beam Ion Sou ce
ECR
Elec on Cyclo on Resonance
FEBIAD
Fo ced Elec on Beam Induced A c Discha ge
ISOLDE
Iso ope Sepa a o Online De ice
LPSC
Labo a oi e de Physique Suba omique e de Cosmologie
PET-CT
Posi on Emission Tomog aphy – Compu ed Tomog aphy
RF
adio equency
SPS
Spa k Plasma Sin e ing
TESIS
Tubula Elec on S ing Ion Sou ce
VADIS
Ve sa ile A c Discha ge Ion Sou ce
Pa icipan sho names
CERN
Eu opean o ganisa ion o nuclea esea ch
NPL
Na ional Physical Labo a o y
PSI
Paul Sche e Ins i u
CEA
Commissa ia à l’éne gie a omique e aux éne gies al e na i es
IST-ID
Associação do Ins i u o Supe io Técnico pa a a IST-ID In es igação e
Desen ol imen o
DTU
Danma ks Tekniske Uni e si e
CHUV
Cen e hospi alie uni e si ai e audois
GANIL
G and Accélé a eu Na ional d’Ions Lou ds
SCK CEN
S udiecen um oo Ke nene gie / Cen e d'é ude de l'éne gie nucléai e
ARRONAX
G oupemen d’in é ê public ARRONAX
ESS
Eu opean spalla ion sou ce ERIC
TUM
Klinikum ech s de Isa de echnischen Uni e si ä München
KULeu en
Ka holieke Uni e si ei Leu en
MedAus on
En wicklungs- und Be iebsgesellscha MedAus on GmbH
SCIPROM
SCIPROM Sà l
MUI
Medizinische Uni e si ä Innsb uck
ILL
Ins i u Max on Laue - Paul Lange in

Deli e able D10.4
i
JRC
JRC -Join Resea ch Cen e- Eu opean Commission
NCBJ
Na odowe Cen um Badań Jąd owych
GSI
GSI Helmhol zzen um  Schwe ionen o schung GmbH
LU
La ijas Uni e si ā e
INFN
Is i u o Nazionale di Fisica Nuclea e
UiO
Uni e si e e i Oslo
Deli e able D10.4
ii
Lis o Figu es
Figu e 1: CO2 o C4+ ionisa ion e iciency wi h espec o CO2/He gas injec ion 5
Figu e 2: CO2 o C4+ ionisa ion e iciency wi h espec o CO2/A gas injec ion 5
Figu e 3: E ec o u bo pump powe on he CO₂ o C4⁺ ionisa ion e iciency, o he bes a io o He o
CO2 iden i ied in Figu e 1. The o he sou ce se ings a e he ones gi en in Table 2. 6
Figu e 4: Ionisa ion e iciency o noble gas mix u e (helium, neon, a gon, k yp on, xenon). Ope a ing
pa ame e s: ca hode empe a u e: 2020 °C, ion sou ce magne : 27.5 mT, anode ol age: 250 V 7
Figu e 5: To al ion cu en as a unc ion o anode ol age a di e en gas injec ion alues wi h noble gas
mix u e. No e: Al hough he eed injec ion comp ises i e noble gases in equal p opo ions (20 %
each), his a io shi s inside he ion sou ce because he low leak a e allows ligh e a oms o en e
mo e easily han hea ie ones. The ac ual composi ion is helium 46 %, neon 21 %, a gon 15 %,
k yp on 10 %, and xenon 8 %. Ope a ional pa ame e s: ca hode empe a u e: 2020 °C, ion sou ce
magne : 30 mT 8
Figu e 6: CO2 o CO+ ionisa ion e iciency wi h CO2 gas as a unc ion o he ca hode empe a u e.
Ope a ing pa ame e s: CO2 gas injec ion o 5.45 x 1014 pps, ion sou ce magne : 30 mT, anode
ol age: 250 V 9
Figu e 7: CO+ cu en as a unc ion o he ion sou ce magne ic ield s eng h. Ope a ing pa ame e s: CO2
gas injec ion o 2.79 x 1014 pps, ca hode empe a u e: 2107 °C, anode ol age: 200 V 9
Figu e 8: Ionisa ion e iciency a io o CO+/CO2+, C+/CO2+ and C+/CO+ as a unc ion CO2 gas injec ion.
Ope a ional pa ame e s: ca hode empe a u e: 2020 °C, ion sou ce magne : 30 mT, anode ol age:
250 V 10
Figu e 9: Ionisa ion e iciency a io o CO+/CO2+, C+/CO2+ and C+/CO+ o he COMIC ion sou ce as a
unc ion o o wa d powe . CO2 gas injec ion o 2.24 x 1016 pps 12
Figu e 10: CO+, CO2+ and C+ cu en as a unc ion o ime o a 200 ms mic owa e pulse. CO2 gas injec ion
o 2.24 x 1016 pps 12
Figu e 11: CO+ cu en as a unc ion bu e gas injec ion o (a) a gon and (b) helium. CO2 gas injec ion o
3.32 x 1016 pps, o wa d powe 20 W 13
Lis o Tables
Table 1: Expec ed gas ou pu om he BN a ge when ope a ed a 1500 °C wi h a con olled oxygen
leak, and beam on a ge as a ailable om he 18 MeV Cyclone KIUBE 3
Table 2: Op imised ion sou ce se ing o C4+ ope a ion 4
Table 3: Ionisa ion e iciencies o CO+, CO2+ and C+ o CO2 gas injec ion a di e en leak p essu e alues,
unco ec ed o backg ound ca bon coming om he ion sou ce ma e ials. Ope a ing pa ame e s:
ca hode empe a u e: 2020 °C, ion sou ce magne : 30 mT, anode ol age: 250 V 10
Table 4: The measu ed alues o CO+ ionisa ion e iciencies. No backg ound was obse ed a mass 29. A
anspo e iciency o 80% is conside ed om he ion sou ce o he Fa aday cup loca ed a e he
sepa a o magne . 11
Summa y
The goal o his esea ch ac i i y ask was o iden i y po en ial solu ions o es ablish adioac i e ion beams in
he o m o C-11 in a clinical ion beam he apy accele a o complex. P e ious s udies [1],[2] p o ided
p o o ypes o C-11 p oduc ion a ge s and concep s o o e all in eg a ion. A c ucial componen is he ion
sou ce o ionize ha sca cely p oduced adioac i e gas.
Th ee ion sou ce op ions ha e been es ed wi h CO2 gas o assess hei pe o mance and gas e iciency in an
a emp o mimic simila condi ions as p o ided by he p oduc ion a ge p o o ype. A 14.5 GHz
Supe nanogan ECR sou ce a MedAus on, a FEBIAD VADIS MK7 a CERN and a 2.4 GHz COMIC ECR sou ce a
The Uni e si y o Manches e we e s udied unde lab condi ions by MedAus on and CERN s a . The
Supe nanogan ion sou ce is a comme cial compac Elec on Cyclo on Resonance (ECR) ion sou ce, whe eas
he FEBIAD and COMIC sou ces o igina e om he ISOL domain, in pa icula elying on he ex ensi e
expe ience accumula ed o e he yea s a he mos ep esen a i e ISOL acili y, ISOLDE.
Measu emen s ha e con i med ha he Supe nanogan ion sou ce demons a es ela i ely low gas e iciency
(obse ed maximum 0.6%). While i did no mee he pa icle ou pu necessa y o he apeu ic applica ions,
i s ill p esen s a iable op ion due o i s ope a ional simplici y. One key ad an age is i s compa ibili y wi h
di ec injec ion in o an exis ing linea accele a o , elimina ing he need o a mass sepa a o o cha ge
b eede —componen s ha ypically lead o addi ional pa icle losses and complexi y.
Despi e alling sho o he equi emen s o ull clinical use, he cu en se up is s ill sui able o gene a ing
diagnos ic ca bon beams a low in ensi y (es ima ed 4x107 pa icles). These can be aluable o applica ions
such as physiological and ange e i ica ion in ea men planning. Howe e , while academically in e es ing,
i s p ac ical use in a clinical se ing is doub ul. The p ima y conce n is i s nega i e impac on wo k low
e iciency, pa icula ly he inc eased in- oom ime equi ed o pa ien s du ing i adia ion p ocedu es.
To imp o e he Supe nanogan’s pe o mance, a possible upg ade in ol es in eg a ing a c yogenic ap,
simila o he one used in he K ion elec on s ing ion sou ce [9]. This would allow o a mo e concen a ed
elease o adioac i e ca bon monoxide h ough pulsed sublima ion in o he plasma chambe , po en ially
inc easing ion p oduc ion e iciency. Howe e , u he s udies a e needed o assess how signi ican ly his
se up could enhance he yield o C4+ ions. In addi ion, issues like beam ep oducibili y and long- e m s abili y
h oughou ea men sessions mus be ca e ully e alua ed.
While obse ed COMIC e iciencies and in ensi ies we e oo low o p o e use ul in he en isaged se up
(obse ed maximum 0.35%), he FEBIAD ion sou ce demons a ed mo e a ou able gas e iciencies. The
FEBIAD sou ce, in pa icula , pe o med well a low gas p essu es (obse ed maximum 5.4%), making i a
s ong candida e o use wi h he p oposed bo on ni ide (BN) a ge . Howe e , a limi a ion o his sou ce is
i s inabili y o e icien ly p oduce high cha ge s a es. As a esul , i s implemen a ion would necessi a e bo h
a mass sepa a o and a cha ge b eede . While he equi ed mass esolu ion is ela i ely modes and hus
echnically achie able, he cha ge b eede i sel would need o be highly capable—able o deli e a ound
1x1011 pa icles pe pulse. This would likely equi e enhancemen s o exis ing EBIS (Elec on Beam Ion Sou ce)
designs o mee hose ou pu le els.
An al e na i e and po en ially mo e obus solu ion could be he use o a c yogenic ap in combina ion wi h
a high-pe o mance elec on s ing ion sou ce, such as he TESIS sys em. TESIS has shown p omising esul s
o p oducing ca bon beams a he equi ed cha ge s a es o he apeu ic pu poses. Since simila sou ces
ha e al eady been success ully in eg a ed wi h c yo aps, his con igu a ion could mee bo h he in ensi y
and quali y demands o ca bon ion he apy.
In conclusion, while he Supe nanogan ion sou ce in i s cu en o m is no sui able o he apeu ic
applica ions due o low pa icle yield, i s simplici y and po en ial o upg ades may jus i y u he esea ch.
Meanwhile, sou ces like FEBIAD and ad anced sys ems like TESIS o e mo e immedia e pa hways owa d
achie ing he necessa y beam quali y and quan i y, pa icula ly when pai ed wi h app op ia e sepa a ion and
b eeding echnologies. Ongoing de elopmen and e inemen o hese sys ems a e c i ical o ensu ing
p ac ical, high-e iciency solu ions o clinical C-11 ion beam he apy.
Deli e able D10.4
2
1. Con ex
One p ominen quali y assu ance limi a ion o pa icle-based adia ion he apy is o e i y he 3-dimensional
loca ion o he B agg peak and hus he deposi ed dose wi hin he pa ien . While he ion beam posi ion and
incoming ene gy can be well de e mined, he pa ien ’s physiology will change o e he cou se o he
ea men which usually las s se e al weeks. As he ion beam dose deposi ion is e ec i ely an implan a ion
p ocess, he use o adioac i e iso opes as p ima y ion beam could enable in- i o moni o ing o he dose
loca ion.
Posi on emi e s like C-11 would hus enable he use o s anda d PET-CTs o di ec co ela ion be ween he
PET signal and he deployed dose dis ibu ion. While his is also possible using s able C-12, using an ac i e
posi on emi e inc eases he signal by an o de o magni ude and esul s in a be e signal o noise a io,
hus imp o ing imaging quali y.
As his me hod can be employed du ing no mal ea men , no ex a machine ime o o he dedica ed imaging
CT sessions a e equi ed. E e y ea men session can be used o ga he new quali y assu ance da a and
igge a necessa y upda e o he ea men plan o ake physiology changes in o accoun . Addi ional online
moni o ing will also help o educe ange unce ain ies and help gua an ee he equi ed sa e y ma gins.
While he i s applica ion would be pos - ea men imaging o p o ide mo e da a on he pa ien ’s ea men
esponse, in he u u e, in- i o online dose e i ica ion will imp o e umou a ge ing and suppo 4-
dimensional ea men planning.
In o de o p o ide a C-11 beam, high yield p oduc ion a ge s ha e o be used [1] and e y e icien
adioac i e ion beam lines a e equi ed. Cu en accele a o anspo e iciency demands a ound 1x1011 C4+
o C6+ pa icles pe second a he sou ce le el o p o ide su icien in ensi y o p o icien ea men . The
sca ce amoun o [C-11]CO gas esul s in igh equi emen s o ion sou ce gas e iciency.
An ex ensi e analysis o he possible echnical solu ions o he implemen a ion o C-11 adioiso opes in an
accele a o -based pa icle he apy cen e was done unde he MEDICIS-PROMED Ho izon 2020 p ojec [2].
In his epo we will p esen he wo k conduc ed a MedAus on, CERN and The Uni e si y o Manches e
on ca bon ion sou ce cha ac e isa ion in iew o gas e iciency o po en ial C-11 adioac i e ion beam lines.
2. Mo i a ion
2.1 Limi a ions in cu en clinical se up
C-12 beams a e gene a ed in pa icle he apy cen es using high-pe o mance ECR ion sou ces. These ion
sou ces gene ally deli e cu en s o o e 100 mic oampe es (µA) in con inuous mode, o , mo e ecen ly,
also in a long-pulse (>100 ms) mode. In bo h cases, howe e , almos all o he gene a ed beam is disca ded,
as he ion beam can only be injec ed in o he synch o on wi hin a ime window o a ew ens o ms. As
s able C-12 is a ailable in su icien quan i ies, hese losses do no pose a p oblem.
Radioac i e pa icle beams p esen a di e en si ua ion. Medical synch o ons a e usually illed once e e y
1-10 seconds. The beam is subsequen ly accele a ed and hen ex ac ed h oughou an ex ac ion spill which
las s up o 8 seconds be o e he nex illing cycle s a s. Radioiso ope p oduc ion is a complica ed and cos ly
p ocess which he e o e calls o e icien use o he sho -li ed adioiso opes. The ime be ween spills should
he e o e be used o accumula e he pa icles o ensu e he in ensi y equi ed o i adia ion. Each s ep mus
achie e high e iciency o ensu e ha he la ges possible p opo ion o he p oduced pa icles is accele a ed.
Howe e , he ion sou ces cu en ly used in pa icle he apy cen es a e no designed o high gas e iciencies
and hus no in insically op imised o he low gas low a es ypically expec ed om a ge ma e ials.
Fu he mo e, hese ion sou ces ha e no s o age capaci y. An EBIS- ype cha ge b eeding sys em was
he e o e p oposed o mee he equi emen s.
Deli e able D10.4
9
Figu e 6: CO2 o CO+ ionisa ion e iciency wi h CO2 gas as a unc ion o he ca hode empe a u e. Ope a ing
pa ame e s: CO2 gas injec ion o 5.45 x 1014 pps, ion sou ce magne : 30 mT, anode ol age: 250 V
Inc easing he ca hode empe a u e om app oxima ely 1,900 °C o 2,100 °C causes he mionic elec on
emission o inc ease, as illus a ed in Figu e 6, in ag eemen wi h he Richa dson–Dushman equa ion 𝑗 ∝
𝑇2𝑒𝛼/𝑇. As he anode d ain cu en (which is a measu e o he he mionic emission cu en ) inc eases, he
CO2 o CO+ e iciency also imp o es. The ise o he elec on densi y, howe e , deepens he elec os a ic
po en ial well in he anode chambe , desc ibed in E o ! Re e ence sou ce no ound. as he inac i e olume.
This is he space in which ions emain apped in he po en ial unnel and can no longe be ex ac ed. As a
esul , he p opo ion o he ac i e olume om which CO⁺ ions can ac ually escape dec eases, and he CO⁺
e iciency eaches a pla eau despi e u he inc ease o he elec on emission.
Figu e 7: CO+ cu en as a unc ion o he ion sou ce magne ic ield s eng h. Ope a ing pa ame e s: CO2 gas injec ion
o 2.79 x 1014 pps, ca hode empe a u e: 2107 °C, anode ol age: 200 V
The VADIS ope a es wi hin a weak axial magne ic ield gene a ed by a solenoid. As he cu en in he ion
sou ce magne inc eases, a shi in CO₂ o CO⁺ e iciency can be obse ed (Figu e 7), which can be explained
by he p esence o wo compe ing mechanisms [5]. On he one hand, he magne ic ield o ces he elec ons

Deli e able D10.4
10
in o spi al o bi s, hus ex ending hei esidence ime in he anode ca i y, which signi ican ly imp o es he
p obabili y o ionisa ion. A he same ime, he inc eased elec on accumula ion in he anode cen e leads o
he o ma ion and deepening o he p e iously men ioned inac i e olume. The p oduced CO⁺ ions emain
apped and can no longe be ex ac ed. The bes ionisa ion e iciency was measu ed a highe magne ic ield
s eng hs.
A highe gas injec ion a es, he ex ac ed cu en sa u a es due o space cha ge limi a ions acco ding o he
Child-Langmui equa ion. This e ec educes he o e all ionisa ion e iciency, as shown in Table 3.
Table 3: Ionisa ion e iciencies o CO+, CO2+ and C+ o CO2 gas injec ion a di e en leak p essu e alues, unco ec ed
o backg ound ca bon coming om he ion sou ce ma e ials. Ope a ing pa ame e s: ca hode empe a u e: 2020 °C,
ion sou ce magne : 30 mT, anode ol age: 250 V
Leak p essu e [mba ]
Injec ed C02 gas (pps)
CO+ [%]
CO2+ [%]
C+ [%]
254
2.76 x 1014
6.5
1.56
0.68
456
4.95 x 1014
4.7
1.74
0.50
952
1.03 x 1015
3.4
1.66
0.34
1454
1.58 x 1015
2.5
1.30
0.25
The da a om Table 3 is p esen ed in Figu e 8 as a ios be ween he di e en ca bon compounds, o illus a e
how he chemical equilib ium e ol es inside he ion sou ce depending on he injec ed gas amoun . I is
obse ed ha he imp o emen o he CO+ ionisa ion e iciency a lowe alues o he injec ed gas is no only
due o an o e all imp o emen o he ionisa ion e iciencies, bu also o an imp o emen o he a io CO/CO2.
The a io C/CO is una ec ed by he amoun o injec ed gas.
Figu e 8: Ionisa ion e iciency a io o CO+/CO2+, C+/CO2+ and C+/CO+ as a unc ion CO2 gas injec ion. Ope a ional
pa ame e s: ca hode empe a u e: 2020 °C, ion sou ce magne : 30 mT, anode ol age: 250 V
These measu emen s led o he ollowing conclusions:
 The ca bon compound wi h he bes ionisa ion e iciency is CO+
Deli e able D10.4
11
 The ion sou ce has be e ionisa ion e iciencies when less gas is injec ed
 The bes se ings o use o he ion sou ce a e:
 Line empe a u e be ween 2050°C and 2100°C
 Anode ol age a 250 V
 Solenoid magne a 6A (which gene a es a magne ic ield es ima ed a 30 mT)Table 4
4.3 E iciency measu emen wi h [C-13]CO2
The goal o he es wi h [C-13]CO2 was o ob ain a eliable measu emen o he ionisa ion e iciency o he
ca bon molecules, because he VADIS has inhe en ly a backg ound o C-12 ions when ope a ed a high
empe a u es.
A e pe o ming anode ol age, magne and line hea ing scans, he se ings iden i ied in sec ion 4.2 we e
con i med as op imal and yielded he ollowing esul s o C-13 (see Table 4), which a e expec ed o also
ep esen he C-11 e iciencies.
Table 4: The measu ed alues o CO+ ionisa ion e iciencies. No backg ound was obse ed a mass 29. A anspo
e iciency o 80% is conside ed om he ion sou ce o he Fa aday cup loca ed a e he sepa a o magne .
Measu emen s a mass 29 a e no a ec ed by any backg ound as he e was no con amina ion obse ed a
his mass du ing he es s wi h noble gases and wi h C-12. An impo an con amina ion (on he o de o µA)
was s ill obse ed a mass 28, o empe a u es abo e 1900 °C. This con amina ion was inc easing du ing he
ope a ion wi h [C-12]CO2. Du ing he es s wi h noble gases, he con amina ion was ound o be negligible
(on he o de o a ew nanoampe es).
5. COMIC es esul s
The 2.45 GHz COMIC (Compac Mic owa e and Coaxial) ion sou ce was designed as an ul a-compac ECR
sou ce a he LPSC in G enoble in close collabo a ion wi h CERN ISOLDE. The ollowing sec ion p o ides an
o e iew o he es s conduc ed a The Uni e si y o Manches e wi h ega d o C-11 compa ibili y.
Leak p essu e [mba ]
Injec ed 13C02 gas (pps)
13CO+ [A]
13CO+ [%]
250
2.71 x 1014
1.87
5.4
500
5.43 x 1014
2.64
3.8
Deli e able D10.4
12
Figu e 9: Ionisa ion e iciency a io o CO+/CO2+, C+/CO2+ and C+/CO+ o he COMIC ion sou ce as a unc ion o o wa d
powe . CO2 gas injec ion o 2.24 x 1016 pps
Figu e 9 shows ha he CO/CO₂ a io inc eases wi h highe mic owa e powe . This is possibly due o he
inc ease in elec on densi y, which is impo an o he neu al dissocia ion o CO₂. Ini ially, elec ons spli
CO₂ p edominan ly in o CO and O (neu al dissocia ion ene gy 5.5 eV), wi h CO subsequen ly being mo e
easily ionised o CO⁺ (ionisa ion ene gy 13.8 eV) [6]. In addi ion, he p opo ion o ho elec ons also inc eases
wi h highe mic owa e powe , which a ou s he di ec ionisa ion o CO₂ o CO⁺ (dissocia i e ionisa ion
ene gy 19.5 eV) [7]. The e o e, he CO/CO₂ a io inc eases un il equilib ium is eached a highe powe s, as
almos all a ailable CO₂ molecules ha e al eady been con e ed. The o ma ion o C⁺, on he o he hand,
emains e y low, because e en highe elec on ene gies a e equi ed. As a esul , bo h C/CO₂ and C/CO
each only low alues.
Figu e 10: CO+, CO2+ and C+ cu en as a unc ion o ime o a 200 ms mic owa e pulse. CO2 gas injec ion o 2.24 x 1016
pps
(a)
Deli e able D10.4
13
Figu e 10 shows he CO₂⁺, CO+ and C+ ion cu en o a 200 ms mic owa e pulse. Fo he CO2+ cu en , a apid
inc ease and o e shoo can be seen, ollowed by a signi ican d op o a s eady pla eau. A e he ini ial
deple ion o CO2 and he di ec p oduc ion o CO2+ s abilises, he s epwise ionisa ion o o he species and
dissocia i e p oduc s akes o e [8]. This esul s in a sho - e m cu en peak o CO₂⁺ ions. Sho ly a e , he
CO₂⁺ cu en d ops again which may ha e se e al causes ha occu simul aneously. Fi s , some o he CO₂ is
no only ionised by elec on collisions, bu also dissocia ed (dissocia ion ene gy 5.5 eV), p oducing CO and
a omic oxygen. This educes he CO₂ concen a ion in he plasma, lea ing ewe molecules a ailable o di ec
ionisa ion o CO₂⁺. Second, CO₂⁺ ions ha ha e al eady been p oduced a e u he con e ed in o CO⁺ o
o he ions by cha ge ans e and collision p ocesses, which u he educes he CO₂⁺ cu en . CO+ and C+
cu en s show a apid inc ease up o a pla eau, in ag eemen wi h expec a ions.
Figu e 11: CO+ cu en as a unc ion bu e gas injec ion o (a) a gon and (b) helium. CO2 gas injec ion o 3.32 x 1016
pps, o wa d powe 20 W
Di e en bu e gases ha e been injec ed in o de o a emp an e iciency boos o CO and CO2 ion
p oduc ion. Resul s o a gon and helium es s a e illus a ed in Figu e 11. The CO⁺ cu en dec eases in bo h
in es iga ed cases wi h inc easing injec ion o bu e gases. When a gon is used, he plasma is only s able
(a)
(b)
Deli e able D10.4
14
wi hin a e y na ow ange, whe eas helium allows o a signi ican ly b oade s able ope a ing window.
Compa ed o ope a ing he sou ce exclusi ely wi h CO₂, he addi ion o bu e gases has an o e all
supp essing e ec on CO⁺ p oduc ion, as he achie able ion cu en in bo h cases emains below he alues
o pu e CO₂ ope a ion. The highes measu ed CO⁺ cu en was app oxima ely 6 µA wi h a CO₂ gas injec ion
o 1.34 × 10¹⁷ pps and a o wa d powe o 50 W. Despi e he o e all lowe CO⁺ cu en yield, howe e , he
bu e gases enable a s able plasma ope a ion a lowe CO₂ gas injec ion a es while s ill main aining good
CO⁺ cu en s. The bes measu ed CO₂ o CO⁺ e iciency was 0.35% wi h an a gon injec ion o 2.24 × 10¹⁶ pps,
a CO₂ injec ion o 4.93 × 10¹⁵ pps, and a o wa d powe o 20 W.
6. Conclusion and ou look
Conduc ed measu emen s ha e con i med a a he low gas e iciency o he Supe nanogan ion sou ce.
Al hough i was no possible o achie e he ea men equi emen s in e ms o numbe o pa icles, i
emains a iable se up due o i s ad an age o simplici y. The ECR ion sou ce can di ec ly injec in o an exis ing
linea accele a o and does no equi e any mass sepa a o o cha ge b eede which by hemsel es ye
educe he ansmi ed pa icle coun . Cu en esul s will only su ice o diagnos ic ca bon beams which can
be used o physiology and ange e i ica ions. Al hough in e es ing om an academic poin o iew i is e y
unlikely ha such a se up would be applied o clinical use sincei will inc ease o e all in- oom ime o he
pa ien in he i adia ion oom. In o de o boos he o e all pe o mance, he Supe nanogan could be
equipped wi h a c yogenic ap as implemen ed in he K ion [9] elec on s ing ion sou ce. This would enable
a highe concen a ion o [C-11]CO ollowed by a pulsed sublima ion elease [9] in o he plasma chambe . I
emains o be s udied how much he C4+ p oduc ion can be boos ed wi h such a se up. Addi ional poin s o
add ess would be ep oducibili y and s abili y h oughou he du a ion o a ea men .
The obse ed gas e iciencies o he o he es ed and op imised sou ces, especially he FEBIAD, show
p omising pe o mances. The FEBIAD also shows a good pe o mance a low gas p essu es which makes i a
sui able candida e o use wi h he en isaged BN a ge . Un o una ely, only low cha ge s a es can be
gene a ed e icien ly which esul s in he need o a dedica ed cha ge b eede and ano he mass sepa a o .
The e iciency o his p ocess was e alua ed in [2], whe e i was ound ha o a combina ion o a ap and
EBIS cha ge b eede , a maximum e iciency o 8% o he con e sion o CO1+ o C6+ can be achie ed. Ye his
e iciency was dec easing owa ds 1% wi h inc easing numbe o C6+ ions pe pulse, due o space cha ge
limi a ions and o he high collec ion ime on he ap compa ed o he s anda d ope a ion (on he o de o
100 ms, ins ead o ew µs). I we assume an a e age e iciency o 5% o his cha ge b eeding s ep, oge he
wi h he VADIS ionisa ion e iciency o abou 5%, we ob ain a o al e iciency o 0.25%, which is below he
measu ed ionisa ion e iciency o Supe nanogan, which does no equi e hese addi ional s eps.
Ano he po en ial solu ion could be a combina ion o a c yogenic ap wi h a high-pe o mance elec on
s ing ion sou ce called TESIS [11] which p omises sui able beam cu en s o he equi ed ca bon cha ge
s a es o ion beam he apy. As simila ion sou ces ha e al eady been equipped and es ed wi h such
c yo aps, such a se up p omises he equi ed beam in ensi ies o ea men .

Deli e able D10.4
15
Re e ences
[1] S. S egemann e al., “A po ous hexagonal bo on ni ide powde compac o he p oduc ion and
elease o adioac i e 11C,” Jou nal o he Eu opean Ce amic Socie y, ol. 41, no. 7, pp. 4086–4097,
July 2021, doi: 10.1016/j.jeu ce amsoc.2020.12.029.
[2] L. Penescu, T. S o a, S. S egemann, J. Pi e s, E. Fio ina, R. Augus o, C. Schmi ze , F. Wenande e
al., "Technical Design Repo o a Ca bon-11 T ea men Facili y", F on ie s in Medicine, ol. 8, 2021,
doi:10.3389/ med.2021.697235
[3] P. Ja din e al., “A om- o-ion ans o ma ion ime in singly cha ged ECRIS”, Nuclea Ins umen s and
Me hods in Physics Resea ch Sec ion B: Beam In e ac ions wi h Ma e ials and A oms, ol. 225, no.
3, pp. 374–382, 2004, doi: 10.1016/j.nimb.2004.06.004.
[4] L. Penescu, R. Ca he all, J. Le y, and T. S o a, “De elopmen o high e iciency Ve sa ile A c
Discha ge Ion Sou ce a CERN ISOLDE”, Re iew o Scien i ic Ins umen s, ol. 81, no. 2, 2010, doi:
10.1063/1.3271245.
[5] F. Maldonado Millan, T. Day Goodac e, and A. Go be g, “Mul iphysics simula ion o a FEBIAD ion
sou ce”, Nuclea Ins umen s and Me hods in Physics Resea ch Sec ion B: Beam In e ac ions wi h
Ma e ials and A oms, ol. 463, pp. 302–304, Jan. 2020, doi: 10.1016/j.nimb.2019.04.078.
[6] M. Tsuji, T. Tanoue, K. Nakano, and Y. Nishimu a, “Decomposi ion o CO2 in o CO and O in a
Mic owa e-Exci ed Discha ge Flow o CO2/He o CO2/A Mix u es”, Chemis y Le e s, ol. 30, no.
1, pp. 22–23, 2001, doi: 10.1246/cl.2001.22.
[7] G. Ju e, “Law ence Be keley Na ional Labo a o , yBe keley, Cali o nia”, XRDS, ol. 17, no. 4, pp. 58–
59, 2011, doi: 10.1145/1961678.1961690.
[8] M. Salahshoo and M. Aslaninejad, “Resonance su ace, mic owa e powe abso p ion, and plasma
densi y dis ibu ion in an elec on cyclo on esonance ion sou ce”, Physical Re iew Accele a o s
and Beams, ol. 22, no. 4, p. 043402, 2019, doi: 10.1103/PhysRe AccelBeams.22.043402.
[9] D.E. Done s, E.D. Done s, E.E. Done s, V.V. Salniko and V.B. Shu o , "P oduc ion o highly cha ged
ion beams K 32+, Xe44+, Au54+ wi h Elec on S ing Ion Sou ce (ESIS) K ion-2 and co esponding basic
and applied s udies", Jou nal o Ins umen a ion, ol. 5, 2010, doi: 10.1088/1748-
0221/5/09/C09001
[10] D.E. Done s, E.D. Done s, E.E. Done s, V.V. Salniko , V.B. Shu o , E.M. Sy esin, "Fo ma ion o high
in ensi e adioac i e ca bon ion beams in elec on s ing ion sou ce", P oceedings o HIAT09,
Venice, I aly, 2009, doi: 10.1088/1748-0221/5/09/C09001
[11] D.E. Done s e al. "Design o a no el ubula elec on s ing ion sou ce (TESIS)", P oceedings o
EPAC08, Genoa, I aly, 2008