TiMoFey Accelerator Complex

• TiMoFey is a projected accelerator complex featuring a continuous-wave proton beam on a graphite dump, designed to search for light, feebly interacting particles.
• It employs staged operations with detailed beam flux and luminosity calculations to probe axion-like particles, hidden photons, and millicharged particles.
• The facility integrates complementary decay-vertex and low-threshold scattering detectors, offering competitive sensitivity to sub-GeV new physics.

Searching arXiv for the specified paper and closely related context. TiMoFey is the projected accelerator complex at the Institute for Nuclear Research of RAS in Troitsk, recently included in the Russian National Program “Fundamental Properties of Matter.” It is designed around a continuous-wave proton beam incident on a graphite beam dump and is studied as a multidisciplinary platform for searches for light, feebly interacting particles. In the formulation developed for the facility, the relevant targets are axion-like particles (ALPs), hidden photons, and millicharged particles, with downstream instrumentation configured either to reconstruct visible decays inside a detector volume or to register energy deposits from elastic scattering of particles traversing the detector (Demidov et al., 4 Aug 2025).

1. Accelerator architecture and staged operation

TiMoFey is built around a proton injector consisting of a linear accelerator and a rapid-cycling synchrotron delivering a continuous-wave proton beam to a graphite beam dump located at the center of a 7 m-radius concrete shielding sphere. Two successive operation stages, each of approximately five years, are planned.

At Stage 1, the proton kinetic energy is $T_p = 423\ \mathrm{MeV}$ and the average current is $I_p = 300\ \mu\mathrm{A}$. At Stage 2, the proton kinetic energy is $T_p = 1.3\ \mathrm{GeV}$ and the average current is $I_p = 100\ \mu\mathrm{A}$. The annual live-time is assumed to be $10^7\ \mathrm{s}$, corresponding to an effective duty cycle of $$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$.

The instantaneous proton flux is defined by

$\Phi_p = \frac{I_p}{e}.$

For Stage 1 this gives

$$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$

and for Stage 2

$$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$

The corresponding beam power is estimated as $P = I_p \cdot T_p$, yielding approximately $I_p = 300\ \mu\mathrm{A}$0 for Stage 1 and $I_p = 300\ \mu\mathrm{A}$1 for Stage 2. Under the assumed live-time, the annual proton-on-target exposure is $I_p = 300\ \mu\mathrm{A}$2 POT at Stage 1 and $I_p = 300\ \mu\mathrm{A}$3 POT at Stage 2. These parameters place the facility in a regime where moderate proton energy is combined with substantial integrated exposure.

2. Beam dump, target properties, and luminosity scales

The beam dump target is a graphite cylinder with radius $I_p = 300\ \mu\mathrm{A}$4, length $I_p = 300\ \mu\mathrm{A}$5, and density $I_p = 300\ \mu\mathrm{A}$6. The corresponding number density is

$I_p = 300\ \mu\mathrm{A}$7

The areal density is then

$I_p = 300\ \mu\mathrm{A}$8

Using the Stage 1 proton flux, the instantaneous luminosity is written as

$I_p = 300\ \mu\mathrm{A}$9

with annual integrated luminosity

$T_p = 1.3\ \mathrm{GeV}$0

These quantities are operational rather than collider luminosity measures in the usual sense; they parameterize the proton-dump source term from which mesons, secondary charged states, and hypothetical weakly coupled particles are produced. In the TiMoFey study, the key formulas linking beam current, target density, decay probability, and scattering rates are used directly to derive projected exclusions in model parameter space.

3. Downstream detector configurations

The proposed search program uses two complementary detector systems in the downstream hall. One is optimized for decays of long-lived particles inside a fiducial volume; the other is optimized for weak ionization signatures from traversing millicharged particles.

The multipurpose “Decay-Vertex” detector, denoted Detector A, is placed immediately downstream, at approximately 7 m from the beam dump shielding exit. Its geometry has a $T_p = 1.3\ \mathrm{GeV}$1 cross section and a detector length $T_p = 1.3\ \mathrm{GeV}$2, arranged around a central decay volume. Photon detection is based on a two-section system. The first section is a preradiator of approximately $T_p = 1.3\ \mathrm{GeV}$3, built from alternating thin lead plates and segmented plastic scintillators. It provides photon-to-$T_p = 1.3\ \mathrm{GeV}$4 conversion efficiency of approximately 90%, corresponding to two-photon detection of approximately 80%, and yields two-dimensional position, direction for $T_p = 1.3\ \mathrm{GeV}$5, and partial energy measurement. The second section is a Shashlyk electromagnetic calorimeter composed of lead-scintillator plates read out by wavelength-shifting fibers. Its energy resolution is

$T_p = 1.3\ \mathrm{GeV}$6

which corresponds to approximately $T_p = 1.3\ \mathrm{GeV}$7–$T_p = 1.3\ \mathrm{GeV}$8 at $T_p = 1.3\ \mathrm{GeV}$9, and it reconstructs photon energies from 10 to 800 MeV together with positions and timing. Plastic scintillator layers upstream of the preradiator act as a charged-particle veto. An alternate configuration for $I_p = 100\ \mu\mathrm{A}$0 detection is a $I_p = 100\ \mu\mathrm{A}$1 TPC plus electromagnetic calorimeter for vertexing of lepton pairs.

Detector B is the millicharged-particle scattering detector, located in the same downstream hall adjacent to Detector A. It occupies a $I_p = 100\ \mu\mathrm{A}$2 volume subdivided into 10 planes of $I_p = 100\ \mu\mathrm{A}$3 optically isolated scintillator bars of dimensions $I_p = 100\ \mu\mathrm{A}$4. Each bar is coupled to a PMT, with light yield approximately $I_p = 100\ \mu\mathrm{A}$5. A single electron of 1 keV therefore produces approximately $I_p = 100\ \mu\mathrm{A}$6, and with a threshold of $I_p = 100\ \mu\mathrm{A}$7 the quoted detection efficiency is at least 95%. The detector principle is that a millicharged particle traversing all layers generates coincident single-electron hits, described as a “double-hit” signature, aligned with the beam direction. The recoil-energy threshold is $I_p = 100\ \mu\mathrm{A}$8. Geometric acceptance requires $I_p = 100\ \mu\mathrm{A}$9 to intercept the detector, and background suppression relies on at least two spatially aligned, time-coincident hits plus a plastic veto against cosmic and other charged backgrounds.

4. Axion-like particle searches

The ALP analysis is formulated in terms of gluonic and electroweak couplings,

$10^7\ \mathrm{s}$0

Two benchmarks are considered: a gluon-dominance case with $10^7\ \mathrm{s}$1 and $10^7\ \mathrm{s}$2, and a democracy case with $10^7\ \mathrm{s}$3 (Demidov et al., 4 Aug 2025).

Production proceeds through $10^7\ \mathrm{s}$4 and $10^7\ \mathrm{s}$5 mixing,

$10^7\ \mathrm{s}$6

with

$10^7\ \mathrm{s}$7

and

$10^7\ \mathrm{s}$8

The principal visible signature is $10^7\ \mathrm{s}$9, with partial width

$$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$0

where

$$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$1

The probability that the ALP decays inside Detector A is

$$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$2

The event yield is written as

$$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$3

For both operational stages, projections assume five years of running and $$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$4. The quoted 95% CL reach extends down to $$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$5–$$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$6 for $$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$7–$$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$8, surpassing CHARM, NuCal, and NA62. In the numerical summary based on a background-free 3.84-event 95% CL criterion after five years at each stage, the sensitivity corresponds to $$10^7\ \mathrm{s}/(3.15 \times 10^7\ \mathrm{s}) \simeq 32\%$$9–$\Phi_p = \frac{I_p}{e}.$0 for $\Phi_p = \frac{I_p}{e}.$1–$\Phi_p = \frac{I_p}{e}.$2, with $\Phi_p = \frac{I_p}{e}.$3 reaching down to $\Phi_p = \frac{I_p}{e}.$4. Within the scope of the study, TiMoFey is therefore positioned as a beam-dump facility with substantial reach in the sub-GeV ALP regime.

5. Hidden-photon searches and the leptophobic proxy

The hidden-photon analysis is represented through a leptophobic $\Phi_p = \frac{I_p}{e}.$5-boson proxy. Four production mechanisms are included: $\Phi_p = \frac{I_p}{e}.$6, $\Phi_p = \frac{I_p}{e}.$7, $\Phi_p = \frac{I_p}{e}.$8 treated in chiral perturbation theory, and proton bremsstrahlung $\Phi_p = \frac{I_p}{e}.$9 evaluated with CompHEP cross sections using 8–11 diagrams (Demidov et al., 4 Aug 2025).

The visible decay mode is $$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$0, with width

$$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$1

Hadronic modes, including $$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$2, are incorporated via DarkCast to obtain the total width $$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$3.

For production from neutral pions, the yield is expressed as

$$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$4

where the geometry factor is

$$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$5

The projected exclusion is stated at 95% CL in terms of $$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$6 or $$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$7 versus $$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$8 over the interval 10–500 MeV, with reach well beyond existing NA62, CHARM, NuCal, and PS191 bounds in the 100 MeV domain. In the paper’s numerical summary, TiMoFey probes $$\Phi_p \simeq \frac{300\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 1.9 \times 10^{15}\ \mathrm{p/s},$$9–$$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$0 for $$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$1–$$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$2 after five years at each stage, again under the background-free 3.84-event 95% CL criterion. A plausible implication is that the facility’s moderate proton energy does not preclude competitive hidden-sector sensitivity when meson production, secondary hadronic channels, and downstream decay acceptance are combined.

6. Millicharged particles, neutrino connections, and prospective upgrades

Millicharged particles are introduced through

$$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$3

Production includes $$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$4 in chiral perturbation theory augmented by the millicharge interaction,

$$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$5

together with $$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$6 and $$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$7 scattering channels analogous to those used for the $$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$8-boson case (Demidov et al., 4 Aug 2025).

Detection is based on elastic $$\Phi_p \simeq \frac{100\ \mu\mathrm{A}}{1.6 \times 10^{-19}\ \mathrm{C}} \simeq 0.62 \times 10^{15}\ \mathrm{p/s}.$$9–$P = I_p \cdot T_p$0 scattering in Detector B. The mean number of hits over a path length $P = I_p \cdot T_p$1 is

$P = I_p \cdot T_p$2

with $P = I_p \cdot T_p$3 and $P = I_p \cdot T_p$4. The probability of at least two hits in $P = I_p \cdot T_p$5 is

$P = I_p \cdot T_p$6

for small $P = I_p \cdot T_p$7. The total yield is integrated over $P = I_p \cdot T_p$8 phase space together with the geometric acceptance $P = I_p \cdot T_p$9 and detection efficiency $I_p = 300\ \mu\mathrm{A}$00.

After five years per stage, the 95% CL reach is quoted as $I_p = 300\ \mu\mathrm{A}$01–$I_p = 300\ \mu\mathrm{A}$02 for $I_p = 300\ \mu\mathrm{A}$03 in the MeV–100 MeV range, improving on SLAC-mQ, LSND, BEBC, and SENSEI. In the numerical comparison, the sensitivity reaches $I_p = 300\ \mu\mathrm{A}$04 down to $I_p = 300\ \mu\mathrm{A}$05 for $I_p = 300\ \mu\mathrm{A}$06, closing untested windows above LSND and SENSEI. The detector concept is therefore not limited to decay searches; it is explicitly a low-threshold scattering instrument.

TiMoFey is presented as a multidisciplinary facility rather than as a single-purpose hidden-sector experiment. Its primary mission in the Program framework is coherent elastic neutrino–nucleus scattering from $I_p = 300\ \mu\mathrm{A}$07 decays at rest, and Detector B is also described as sensitive to neutrino-electron scattering, sterile-neutrino searches, and $I_p = 300\ \mu\mathrm{A}$08-nucleus interactions. The same section of the study identifies ALPs, hidden photons, and millicharged particles as dark-matter mediators or candidates and notes potential relevance to models addressing the cosmic-ray positron excess and the 21 cm absorption anomaly.

The study also specifies upgrade paths. A superconducting linac could increase the beam current by a factor of $I_p = 300\ \mu\mathrm{A}$09–$I_p = 300\ \mu\mathrm{A}$10, thereby increasing $I_p = 300\ \mu\mathrm{A}$11 and improving sensitivities. Complementary detectors under consideration include liquid-argon TPCs, high-resolution calorimetry, and time-projection chambers. This suggests a facility concept in which beam power, detector modularity, and a shared downstream hall are intended to support a progressively broadened program in both new-particle searches and neutrino physics.