Stars

Nuclei

Laboratory

Nuclear

Astrophysics

at CERN n_TOF

Expression of Interest for the CERN 20-year Roadmap of Non-Collider Physics

C. Domingo-Pardo, C. Lederer-Woods

on behalf of the n_TOF Collaboration

INTC Meeting · CERN, Geneva · 4 November 2026

CERN-INTC-2026-014 · INTC-EOI-002

Nuclear

Astrophysics

at CERN n_TOF

Expression of Interest for the CERN 20-year Roadmap of Non-Collider Physics

C. Domingo-Pardo, C. Lederer-Woods

on behalf of the n_TOF Collaboration

INTC Meeting · CERN, Geneva · 4 November 2026

CERN-INTC-2026-014 · INTC-EOI-002

1. Scientific Motivation & Impact
2. Programme Definition
3. Why at CERN: Uniqueness & Synergies
4. Timeline
5. Summary

3

• Sites and evolution: which sources make the heavy elements, and how did their contributions change over Galactic history?
• Stellar conditions: what are the temperatures and neutron densities of the s-, i- and r-processes, and how much does each contribute?
• Presolar grains and stellar spectra: what produces their peculiar abundance patterns?
• Radioactivities and chronometers: how do neutron-induced reactions set their abundances?
• Beyond Pb/Bi: how does nucleosynthesis proceed, and what is the role of fission?

​

Open questions on the origin of the heavy elements

4

(I) Stellar models and presolar grains

• Re-measure the stable isotopes with MACS uncertainties >5% to ≤5%
• Stations: EAR1, EAR2, NEAR

1xEoIs: EoI n_TOF facility upgrades, M.Bacak et al.

​

​

Programme: three physics pillars and the stations that serve them

5

(I) Stellar models and presolar grains

• Re-measure the stable isotopes with MACS uncertainties >5% to ≤5%
• Stations: EAR1, EAR2, NEAR

1xEoIs: EoI n_TOF facility upgrades, M.Bacak et al.

​

(II) s-process conditions in AGB and massive stars, first i-process nuclei

• Measure new branching-point nuclei; extend existing data in neutron energy
• Stations: REAR, DICER, NEAR/CYCLING, n_ACT

2xEoIs: EoI REAR, V. Alcayne et al.; EoI n_ACT, C. Lederer-Woods, A. Mengoni et al.

​

​

Programme: three physics pillars and the stations that serve them

6

(I) Stellar models and presolar grains

• Re-measure the stable isotopes with MACS uncertainties >5% to ≤5%
• Stations: EAR1, EAR2, NEAR

1xEoIs: EoI n_TOF facility upgrades, M.Bacak et al.

​

(II) s-process conditions in AGB and massive stars, first i-process nuclei

• Measure new branching-point nuclei; extend existing data in neutron energy
• Stations: REAR, DICER, NEAR/CYCLING, n_ACT

2xEoIs: EoI REAR, V. Alcayne et al.; EoI n_ACT, C. Lederer-Woods, A. Mengoni et al.

​

(III) i- and r-process nucleosynthesis

• First (n,γ) data on short-lived i- and r-process nuclei
• Stations: n_ACT, ISR + neutron target

3xEoIs: EoI n_TOF facility upgrades, M.Bacak et al.; EoI n_ACT, C. Lederer-Woods, et al.; EoI ISR & Neutron Target, Y. Litvinov et al.

Programme: three physics pillars and the stations that serve them

7

Roadmap coverage of the s-, i- and r-process paths

Adapted from M. Wiedeking et al., Nat. Rev. Phys. (2025), doi:10.1038/s42254-025-00885-7

← Pillar I →

← Pillar II →

← Pillar III →

after 20y roadmap!

8

Pillar I · Presolar grains and stellar models need MACS to ≤5%

Presolar-grain isotope ratios are measured to a few ‰. Stellar models (AGB and massive-stars) have uncertainties of ≤5% → Nuclear data input also has to be accurate to the level of ≤5%

9

Figure courtesy of A. Mengoni

Pillar I · Many measured MACS still exceed 5% uncertainty

10

Only 129 of the 281 s-process isotopes in KADoNiS have MACS (kT = 30 keV) known to ≤5%. The remaining 152 are a priority of the 20-year roadmap.

Pillar I · Only 129 of 281 s-process isotopes are known to ≤5%

MACS uncertainty

11

EAR1

EAR2

Gen.#4 Target

Pillar I · Gen.#4 target and upgraded EAR1/EAR2 to close the gap

RMS < 2%

Roadmap priority: re-measure the ~150 isotopes still above 5% MACS uncertainty (Gen.#4 target, upgraded EARs).

→ EoI n_TOF facility upgrades, M.Bacak et al.

12

Branching-point nuclei can be produced only in very small quantities. Existing TOF data typically cover a few % of the energy range relevant for stellar MACS (kT ≈ 5–100 keV).

Pillar II · Branching-point data cover a narrow energy range

Measurements limited by:

• in-beam backgrounds
• neutron flux

13

DICER concept (LANL) → CERN

→ EoI REAR, V. Alcayne et al.

Pillar II · DICER-type transmission and REAR extend the reach

REAR beamlines

Scarce samples: DICER-type binocular transmission. E_n range limited by γ-flash and backgrounds: REAR low-background beamline.

14

Pillar II · Unmeasured s-process branchings within reach of REAR

A. Tarifeño-Saldivia et al., Phys. Rev. Accel. Beams (2026), doi:10.1103/5d87-m8bn

REAR beamlines

Backward extraction at REAR suppresses the γ-flash (>3× fewer prompt γ) and the high-energy neutron background, extending branching-point measurements in E_n.

15

EoI n_ACT@BDF, C. Lederer-Woods et al.

Pillar II · n_ACT@BDF:

→ enables new data for s-branchings, first i-process cases and galactic radioactivities

→ world leading neutron flux in astrophysical keV range (>10³ x a-NEAR flux )

16

Adapted from M. Wiedeking et al., Nat. Rev. Phys. (2025), doi:10.1038/s42254-025-00885-7

← Pillar III →

Pillar III · i/r-process nuclei are too short-lived for targets

17

Pillar III · Conventional TOF requires long-lived samples

neutron beam

t_1/2 > yr

​

Radioactive target

>10¹⁴ atoms

n_TOF so far:

18

Pillar III · Inverse kinematics: stored ions + neutron target

neutron beam

t_1/2 > yr

​

Radioactive target

>10¹⁴ atoms

neutron target

10⁹ n/cm²

​

n_TOF so far:

New approach:

Stored Radioactive Ion Beam

10⁸ ions, 100 kHz, 10¹³ ions/s

(t_1/2 = s-min-yr)

​

19

Pillar III · Two implementation options at CERN

injection

2 GeV

PSB

​

RIB

​

Neutron

Target

→ EoI ISR & Neutron Target, Y. Litvinov, C.Domingo et al.

compact cyclotron

R. Reifarth, Y. Litvinov et al., Phys. Rev. Accel. Beams 20, 044701 (2017); A. Tarifeño-Saldivia et al., Phys. Rev. Accel. Beams 29, 061601 (2026)

Free-Neutron Target via spallation and ⁹Be(p,xn)

Phys. Rev. Accel. Beams 20, 044701 (2017)

DOI: https://doi.org/10.1103/PhysRevAccelBeams.20.044701

​

​

Phys. Rev. Accel. Beams 29, 061601 (2026)

DOI: https://doi.org/10.1103/5d87-m8bn

21

Impact · s-, i- and r-process paths run far from stability

22

Figures courtesy of I. Dillmann (TRIUMF)

Impact · Measured stellar (n,γ) data stay close to stability

Direct stellar-energy (n,γ) data exist almost only for stable and long-lived nuclei; the i- and r-process paths rely on theory.

23

Figures courtesy of I. Dillmann (TRIUMF)

Impact · ISR + neutron target opens (n,γ) far from stability

​

​

ISR & Neutron Target

ISR & Neutron Target

Stored Radioactive Ion Beam

10⁸ ions, 100 kHz, 10¹³ ions/s

(t_1/2 = s-min-yr)

​

neutron target

10⁹ n/cm²

​

Stored beams on a neutron target reach short-lived s-process branchings, i-process waiting points and selected r-process freeze-out nuclei.

24

Why CERN · Neutron source, stations and ISOLDE on one site

Neutron source

PS beam on a Pb spallation target with a low duty cycle: white spectrum from meV to GeV, high instantaneous flux and high resolution

Stations

EAR1 (resolution), EAR2 (flux), NEAR (activation); n_ACT, REAR and DICER in the future

ISOLDE on site

Radioisotopically pure samples for branching points; high-intensity RIBs; a possible ISR coupled to a neutron target in the same complex

Expertise

High-accuracy cross section methodology, spallation-target design, neutron-beam collimation and RIB production: the prerequisites for the ISR concept

Track record

More than two decades of investment in the facility, detectors and trained personnel

No other facility combines all of these elements on one site.

Run 4

LS4 / Run 5

Beyond

2028

2031

2034

2037

2041

2044

2046

Pillar 1

Precision stable

isotope measurements

(Pillar 1)

²²Ne, ⁸⁶Sr, Ba chain (EAR1)

Ca isotopic chain (EAR1+EAR2+NEAR)

Ongoing stable isotope programme: high accuracy & high resolution

sTED, STAR operational

Pillar 2-3

Radioactive

branching

points, cosmochronology, and selected i-process nuclei

¹³⁵Cs NEAR

¹³⁴Cs EAR2

¹⁰⁷Pd EAR1/EAR2

¹⁸¹,¹⁸²Hf EAR2+n_ACT

¹³⁷Cs CYCLING

s-process branchings EAR2 + NEAR (ISOLDE) + REAR

n_ACT@BDF: s-, i-process cases [BIAS, BEAS stations]

CYCLING installed

n_ACT operational

Pillar 3

i- and r-process nuclei in inverse

kinematics

ISR full

programme

s/i/r-process

LANSCE demo spallation n-target

CRYRING demo cyclotron-driven n-target

CRYRING/TRISR early science

Experimental validation

ISR construction

ISR conceptual design + approval

Pillar 1 — stable isotopes

Pillar 2 — radioactive nuclei

Pillar 3 — inverse kinematics

Key milestone

n_ACT@BDF: s-, i-process cases [BRIS rabbit station]

ISR + Neutron Target operational

Timeline

26

Summary

Objective

Measure the neutron-capture data needed to explain where the heavy elements come from: s-, i- and r-process

Scope

Three pillars: ~150 stable isotopes at <5%; radioactive s-branchings; short-lived i/r nuclei in inverse kinematics

Why CERN

n_TOF flux + resolution, ISOLDE samples and RIBs, n_ACT@BDF, storage-ring know-how: an ecosystem found nowhere else

Needs

Beam time; facility upgrades; n_ACT (~6 MCHF); REAR; ISR + neutron target (cost pending design)

Main challenge

Pillar III: from separate building blocks to a working ISR + neutron target, fundamental demo studies

Status

Pillars I–II ready and running into Run 4; Pillar III relies on n_Act and demonstrator phase (LANSCE, GSI)

27

BACKUP

28

Open questions · Example measurements

Sites and evolution

²²Ne: neutron poison at low metallicity. ^42–48Ca: several sites; ⁴⁸Ca from supernovae or the i-process? ¹⁰⁷Pd: galactic chemical evolution

Stellar conditions

s: ^134,135Cs, ⁹⁴Nb, ⁷⁹Se, ¹⁴⁷Pm, ¹⁶³Ho, ¹⁷¹Tm, ¹⁷⁹Ta, ²⁰⁵Pb. i: ¹³⁷Cs, ¹⁴⁴Ce, ⁹⁰Sr, ⁷⁵Ga, ⁸⁰Ge, ⁸⁵Br. r: ¹³²Sn

Presolar grains and spectra

⁸⁶Sr: ±10% MACS gives ~200‰ on δ(⁸⁸Sr/⁸⁶Sr), half the ~400‰ grain spread. ^134–136Ba, ¹³⁴Cs: δ(^134,138Ba/¹³⁶Ba). ⁹⁴Nb: ⁹⁴Mo in grains

Radioactivities and chronometers

^181,182Hf (¹⁸²Hf–¹⁸²W), ²⁰⁵Pb (²⁰⁵Pb–²⁰⁵Tl), ¹⁰⁷Pd (¹⁰⁷Ag), ¹³⁵Cs/¹³³Cs in the Early Solar System. ⁶⁰Fe, ¹⁸²Hf via double capture at n_ACT

Beyond Pb/Bi and fission

Accurate N_s fixes the r-process residuals N_r = N_⊙ − N_s. s-share of ^206–208Pb, ²⁰⁹Bi sets the r-flow beyond Bi (next slide)

Big-Bang nucleosynthesis

⁷Be(n,α) and ⁷Be(n,p) measured at n_TOF with ISOLDE samples; full cosmological energy range at an ISR (⁷Be: 2.8×10¹⁰ ions/s)

Precise s-process data also test r-process yields, models and their reliability.

s-Process branching points: still a lot to do!

s-process path (blue) with the branching point at ¹⁴⁷Nd, compared with the i-process (green) and r-process (red) paths. Figure from [2], Fig. 2.

~60

branching points can be activated along the s-process path [1, 2]

21

flagged in [1] as priority TOF targets (Käppeler et al. 2011)

Branching flow

Neutron capture competes with β-decay on nuclei with half-lives of days or more.

Stellar probes

Set by temperature, electron and neutron density: ¹³⁴Ba/¹³⁶Ba, Rb, ⁹⁶Zr, ¹⁸²Hf.

Challenge to measure

TOF and activation usually need half-lives above a few years; MACS needed to ~10%.

[1] M. Lugaro, A. Chieffi, in Astrophysics with Radioactive Isotopes, ASSL 453, Springer (2018), Ch. 12, p. 643

[2] M. Lugaro, M. Pignatari, R. Reifarth, M. Wiescher, Annu. Rev. Nucl. Part. Sci. 73, 315 (2023)

Branching points turn s-process abundances into stellar probes.

Branching points (1/4) · From ³⁵S to ⁸⁵Kr

★ Priority TOF candidates (Käppeler et al. 2011), as marked in [1] · T½ terrestrial; → value at 300 MK where given in [1]

n_TOF route: as proposed in the n_TOF post-LS4 White Book · NT = neutron target in a storage ring

[1] M. Lugaro, A. Chieffi, in Astrophysics with Radioactive Isotopes, ASSL 453, Springer (2018), Ch. 12, p. 643

⁸⁵Kr opens already at neutron densities above 5×10⁸ cm⁻³.

Branching points (2/4) · From ⁸⁶Rb to ¹⁴³Pr

★ Priority TOF candidates (Käppeler et al. 2011), as marked in [1] · T½ terrestrial; → value at 300 MK where given in [1]

n_TOF route: as proposed in the n_TOF post-LS4 White Book · NT = neutron target in a storage ring

[1] M. Lugaro, A. Chieffi, in Astrophysics with Radioactive Isotopes, ASSL 453, Springer (2018), Ch. 12, p. 643

The Cs chain spans EAR2, NEAR, CYCLING and n_ACT.

Branching points (3/4) · From ¹⁴⁷Nd to ¹⁷⁷Lu

★ Priority TOF candidates (Käppeler et al. 2011), as marked in [1] · T½ terrestrial; → value at 300 MK where given in [1]

n_TOF route: as proposed in the n_TOF post-LS4 White Book · NT = neutron target in a storage ring

[1] M. Lugaro, A. Chieffi, in Astrophysics with Radioactive Isotopes, ASSL 453, Springer (2018), Ch. 12, p. 643

Short-lived rare-earth branchings need ISR & NT.

Branching points (4/4) · From ¹⁷⁹Hf to ²¹⁰Po

Do not open (T½ drops below days in stars): ¹⁰³Ru, ¹²³Sn, ¹²⁴Sb, ¹⁵⁶Eu, ¹⁶⁰Tb ★ (ISR & NT), ¹⁶¹Tb, ¹⁷⁵Yb, ¹⁹⁸Au, ²⁰⁵Hg

★ Priority TOF candidates (Käppeler et al. 2011), as marked in [1] · T½ terrestrial; → value at 300 MK where given in [1]

n_TOF route: as proposed in the n_TOF post-LS4 White Book · NT = neutron target in a storage ring

[1] M. Lugaro, A. Chieffi, in Astrophysics with Radioactive Isotopes, ASSL 453, Springer (2018), Ch. 12, p. 643

¹⁸¹Hf links s-process branchings to Solar System chronology.

Scientific motivation & impact: s-process branchings

Cases for a high-flux TOF facility (FRANZ)

The i-process …

Scientific motivation & impact: i-process

The r-process …

Scientific motivation & impact: r-process

Why at CERN?

38

Resources

>40

institutes in the n_TOF Collaboration

~6 MCHF

n_ACT capital (Class 3 estimate, 3 stations)

4–5 weeks

beam per radioactive branching case at EAR2

Personnel

• Pillar I: largely within existing Collaboration staffing
• Pillar II: stronger ISOLDE interface; RP-trained staff for sample handling
• Pillar III: effort with host-facility teams; dedicated ISR design & construction staff

Equipment investments

• CYCLING transport system (LS3)
• n_ACT@BDF: ~6 MCHF; no significant operating cost beyond capital
• REAR: see EoI D. Cano Ott et al.
• ISR + neutron target: cost pending technical design

Beam time & operations

• Beam time is the main limit in Run 4–5
• Stable isotopes: several weeks each
• NEAR / n_ACT run parasitically: no TOF beam time
• Facility operation largely covered by CERN institutional support

CERN-led: facility & target, n_ACT@BDF, ISR Collaboration-led: detectors, samples, analysis, personnel

39

Scenarios & priorities

BASELINE

Existing areas + planned upgrades

Needs: EAR1, EAR2, NEAR, CYCLING, GEAR, Gen.#4 target

Delivers

• Pillar I: ~150 stable isotopes towards <5%
• Part of Pillar II: ^134,135Cs, ¹⁰⁷Pd, Hf, …
• High-impact results through Run 5

I

II (part)

ENHANCED

+ n_ACT@BDF and REAR

Needs: n_ACT (~6 MCHF, from ~2032); REAR installed in LS4

Delivers

• Full Pillar II: ¹⁷⁹Ta, ⁹⁴Nb, ¹⁴⁷Pm, ¹⁶³Ho, ⁷⁹Se, ²⁰⁴Tl, …
• First i-process cases: ¹³⁷Cs, ¹⁴⁴Ce, ⁹⁰Sr
• ⁶⁰Fe production/destruction; low-background capture

I

II

III (i)

FULL AMBITION

+ ISR & neutron target

Needs: ISR at HIE-ISOLDE coupled to a free-neutron target

Delivers

• (n,γ) on short-lived s-, i- and r-process nuclei
• First RIB demonstration ~2040
• Systematic programme from 2041

I

II

III

Each step is a separable package with its own EoI; the baseline already delivers Pillar I and part of Pillar II.

40

Complementarity & competition

Alternatives elsewhere

n_TOF / CERN edge

Partners & complementarity

Pillar I

Stable isotopes in s-process

• GELINA (JRC-Geel): high resolution, lower inst. flux
• DANCE (LANSCE), J-PARC ANNRI, CSNS Back-n

• Flux and resolution at once: small σ, small sample masses
• Sub-5% accuracy already demonstrated

• GELINA / DANCE for cross-checks
• Results feed JEFF, ENDF, JENDL; NEA HPRL

Pillar II

s-branchings and first i-process

• DANCE (high-efficiency capture)
• Compact keV activation sources: cannot span 1 keV to several 100 keV

• On-site ISOLDE sample production + PSI / ILL
• NEAR/n_ACT cover the full stellar energy range (n_ACT ≈10³× a-NEAR flux)

• PSI, ILL: radioactive sample production
• Competition limited by technical barriers

Pillar III

Inverse kinematics (short-lived i- and r-process)

• TRISR (TRIUMF-ISAC): future parallel initiative

• Only site combining high-intensity ISOL beams, spallation n-target expertise and low-energy storage rings

• LANSCE, CRYRING@ESR: demonstrator partners
• TRISR: complementary, long term

Theory is no substitute: TALYS, NON-SMOKER, SMARAGD uncertainties for branchings and i-process nuclei reach factors of 2–10, versus the ±5% stellar models need.

41

Feasibility, risks & milestones

Pillar I: stable isotopes

MATURE

• Ready: facility, detectors (C₆D₆, sTED, STAR soon), analysis frameworks
• Limiting factor: beam time, not technology
• Milestones: sTED/STAR and Gen.#4 target operational

Pillar II: branching points

DEMONSTRATED

• Track record: 7 branching nuclei measured in 25 years; ISOLDE samples shown for ¹³⁵Cs, ⁷Be
• Needs: EAR2 optimisation, CYCLING (LS3), GEAR upgrade, approval of n_ACT / REAR / DICER
• Milestones: n_ACT BIAS/BEAS end of Run 4; REAR commissioned 2036

Pillar III: inverse kinematics

HIGHEST RISK / RETURN

• Status: building blocks (neutron target, low-energy storage ring, Wien-filter recoil separation) demonstrated separately
• Main risks: full-system integration; ISR approval and funding at HIE-ISOLDE

Run 4

LANSCE spallation n-target demonstrator

Run 4

CRYRING@ESR full-system proof of principle

LS4 / Run 5

ISR conceptual design + approval

~2040

First radioactive-beam inverse-kinematics measurement

2041+

ISR full s/i/r-process programme

Storage-Ring options for a future “production” facility

42

K. Pak, B. Davids, and Y. K. Kim, Nuclear Engineering and Technology 57, 103392 (2025), arXiv:2312.11859 [physics.acc-ph].

See INTC EoI by C. Domingo-Pardo, Y. Litvinov

A. Tarifeño-Saldivia et al., Phys. Rev. Accel. Beams 29 (2026), https://doi.org/10.1103/5d87-m8bn

Free-Neutron Target via spallation and ⁹Be(p,xn)

Phys. Rev. Accel. Beams 20, 044701 (2017)

DOI: https://doi.org/10.1103/PhysRevAccelBeams.20.044701

​

​

Phys. Rev. Accel. Beams 29, 061601 (2026)

DOI: https://doi.org/10.1103/5d87-m8bn

Why at CERN? Isolde Storage Ring (ISR)

• New ISR design by M. Grieser (2025) → This EoI
• Including Wien-Filters for reaction product detection
• 5.2 m experimental distance

Neutron target

Stray ions due to charge-exchange reactions with residual gas can collide only with the dipoles of the ring (not in the straight sections).

​

Catchers at the dipoles are commonly used to remove these ions, so they are not a problem, both if they are long-lived or short-lived. They can be removed easily. For radioactive beams one adds removable inserts and local shielding, so the activity can be taken out between campaigns.

Realistic examples for neutron-capture at the ring:

η = 1e9 n/cm² → neutron density objective, both spallation/cyclotron

ISR = 1e8 ions → number of stored ions / space-charge limit ISR

f ≈ 100 kHz at ~100 keV/u → revolution frequency in the ring

I = 1e8 x 1e5 = 1e13 ions/s → particle current through the ring

Luminosity = η·I = 1e22 cm⁻²s⁻¹ → reaction luminosity

​

Reaction rate:

R = L · σ · t = 1e22 cm⁻²s⁻¹ × σ[mb] × 1e-27 cm²/mb × 24h·3600s

​

R ≈ 0.86 × σ[mb] counts/day

​

Measuring time for 10% statistical uncertainty:

​

σ = 5 mb → ~4 counts/day → ~23 days → Just one data point may be OK

σ = 10 mb → ~9 counts/day → ~12 days

σ = 50 mb → ~43 counts/day → ~2.3 days (~56 h) → Scan a few points

σ = 100 mb → ~86 counts/day → ~1.2 days (~28 h)

σ = 500 mb → ~430 counts/day → ~5.6 hours → Fully scan the CS in the energy range

​

46

Summary

Goal: quantify the contributions of the s-, i- and r-processes to the heavy elements with laboratory (n,γ) data.

Three pillars

• I · Stable isotopes: 152 of 281 s-process isotopes still have MACS uncertainties >5%; re-measure them to ≤5% at EAR1/EAR2 to interpret presolar grains and constrain AGB and massive-star models.
• II · s-process branchings: new and extended branching-point data with REAR, DICER-type transmission and n_ACT, using samples from ISOLDE, PSI and ILL.
• III · i- and r-process: first (n,γ) data on short-lived nuclei with NEAR-CYCLING and n_ACT, and ultimately in inverse kinematics at an ISR coupled to a neutron target.

47

Summary

Why at CERN

• White neutron spectrum (meV–GeV) from the PS and a Pb spallation target, with high instantaneous flux and high resolution.
• Complementary stations (EAR1, EAR2, NEAR; n_ACT and REAR in the future) cover the full range of sample availability.
• ISOLDE on site: radioisotopically pure samples, high-intensity RIBs and a possible ISR coupled to a neutron target.
• Storage-ring and neutron-target concepts are also pursued at GSI and TRIUMF; CERN is the only site where all components are already co-located.