ForwArd Search ExpeRiment at the LHC

New physics has traditionally been expected in the high-$p_T$ region at high-energy collider experiments. If new particles are light and weakly coupled, however, this focus may be completely misguided: light particles are typically highly concentrated within a few mrad of the beam line, allowing sensitive searches with small detectors, and even extremely weakly coupled particles may be produced in large numbers there. We propose a new experiment, forward search experiment, or FASER, which would be placed downstream of the ATLAS or CMS interaction point (IP) in the very forward region and operated concurrently there. Two representative on-axis locations are studied: a far location, 400 m from the IP and just off the beam tunnel, and a near location, just 150 m from the IP and right behind the TAN neutral particle absorber. For each location, we examine leading neutrino- and beam-induced backgrounds. As a concrete example of light, weakly coupled particles, we consider dark photons produced through light meson decay and proton bremsstrahlung. We find that even a relatively small and inexpensive cylindrical detector, with a radius of $\sim 10\mathrm{cm}$ and length of 5–10  m, depending on the location, can discover dark photons in a large and unprobed region of parameter space with dark photon mass $m_{A^{\prime }}\sim 10–500\mathrm{MeV}$ and kinetic mixing parameter $\epsilon \sim {10}^{-6}-{10}^{-3}$. FASER will clearly also be sensitive to many other forms of new physics. We conclude with a discussion of topics for further study that will be essential for understanding FASER’s feasibility, optimizing its design, and realizing its discovery potential.

Figure 1

Schematic drawings of the LHC ring and the current very forward infrastructure downstream from the ATLAS and CMS interaction points, along with the representative far and near on-axis detector locations for FASER. Note the extreme difference in the transverse and longitudinal scales in the lower figure. Details of the geometry and sample tracks have been taken from Refs. [36, 37, 38]. See the text for details.

Figure 2

Particle multiplicities in 13 TeV $pp$ collisions at the LHC. Left: ${\pi }^0$ and $\eta$ multiplicities from EPOS-LHC [52] (circles), QGSJET-II-04 [53] (squares), and SIBYLL 2.3 [54, 55] (triangles). Right: ${\pi }^0$, ${\pi }^{\pm }$, $\eta$, $\omega$, $\rho$, and $p$ multiplicities from EPOS-LHC [52].

Figure 3

Distribution of ${\pi }^0$ (top) and $\eta$ (bottom) mesons in the $(\theta ,p)$ plane, where $\theta$ and $p$ are the meson’s angle with respect to the beam axis and momentum, respectively. The different panels show results from the simulation codes EPOS-LHC [52] (left), QGSJET-II-04 [53] (center), and SIBYLL 2.3 [54, 55] (right). The total number of mesons is the number produced in one hemisphere ($0<\cos \theta \le 1$) in 13 TeV $pp$ collisions at the LHC with an integrated luminosity of $300{\mathrm{fb}}^{-1}$. The bin thickness is $1/5$ of a decade along each axis. The dashed line corresponds to $p_T=p\sin \theta ={\mathrm{\Lambda }}_{\mathrm{QCD}}\simeq 250\mathrm{MeV}$.

Figure 4

Distribution in the ($\theta$, $p$) plane, where $\theta$ and $p$ are the angle with respect to the beam axis and momentum, respectively, for dark photons produced by ${\pi }^0$ decays (left), $\eta$ decays (center), and proton bremsstrahlung (right), for $A^{\prime }$ parameters $(m_{A^{\prime }},\epsilon )=(20\mathrm{MeV},{10}^{-4})$ (top) and $(100\mathrm{MeV},{10}^{-5})$ (bottom). The right-hand axis indicates the dark photon’s decay length; see Eq. (7). The total number of dark photons is the number produced in one hemisphere ($0<\cos \theta \le 1$) in 13 TeV $pp$ collisions at the LHC with an integrated luminosity of $300{\mathrm{fb}}^{-1}$. The bin thickness is $1/5$ of a decade along each axis. The black dashed, dotted, and dash-dotted lines correspond to $p_{T,A^{\prime }}={\mathrm{\Lambda }}_{\mathrm{QCD}}\simeq 250\mathrm{MeV}$, $m_{A^{\prime }}^2$, and 10 GeV, respectively.

Figure 5

Distribution in the ($\theta$, $p$) plane, where $\theta$ and $p$ are the angle with respect to the beam axis and momentum, respectively, for dark photons that decay in the interval $(L_{\min },L_{\max })=(390\mathrm{m},400\mathrm{m})$ (the far detector location) and are produced by ${\pi }^0$ decays (left), $\eta$ decays (center), and proton bremsstrahlung (right) for $A^{\prime }$ parameters $(m_{A^{\prime }},\epsilon )=(20\mathrm{MeV},{10}^{-4})$ (top) and ($100\mathrm{MeV},{10}^{-5}$) (bottom). The total number of $A^{\prime }\mathrm{s}$ is the number produced in one hemisphere ($0<\cos \theta \le 1$) in 13 TeV $pp$ collisions at the LHC with an integrated luminosity of $300{\mathrm{fb}}^{-1}$. The bin thickness is $1/5$ of a decade along each axis. The dashed and dashed-dotted lines correspond to $p_{T,A^{\prime }}={\mathrm{\Lambda }}_{\mathrm{QCD}}\simeq 250\mathrm{MeV}$ and 10 GeV, respectively. In each plot the right $y$-axis indicates the dark photon’s characteristic decay length $\overline{d}$ [see Eq. (7)]. The angular coverage of the detector is indicated via vertical gray dashed lines.

Figure 6

Left: $N_{\mathrm{sig}}$, the expected number of signal events, for two representative ($m_{A^{\prime }}$, $\epsilon$) points as a function of the distance between the IP and detector, $L_{\max }$, for the near and far detector benchmark design (see text). Right: $N_{\mathrm{sig}}$ for the far detector location as a function of the detector radius $R$.

Figure 7

Same as in Fig. 5, but for the near detector location with $(L_{\min },L_{\max })=(145\mathrm{m},150\mathrm{m})$.

Figure 8

Left: Number of expected events per kilogram of detector mass for the detector at the far location (see text) as a function of the minimal incident neutrino energy $E_{\nu ,\min }$. The red (blue) line corresponds to the total number of CC (single pion production) events induced by neutrinos with energies $E_{\nu }\ge E_{\nu ,\min }$. The plot assumes an integrated luminosity of $300{\mathrm{fb}}^{-1}$. Right: Ratio of the energies of the softer ($E_2$) and harder ($E_1$) tracks from ${\nu }_{\mu }N\to {\mu }^{\pm }{\pi }^{\mp }X$ with $E_{\nu }=100\mathrm{GeV}$ (red histogram) and from $A^{\prime }\to e^{+}{e}^{-}$ pair, assuming $E_{A^{\prime }}\gg m_{A^{\prime }}\gg m_e$, and unpolarized $A^{\prime }$’s (green histogram).

Figure 9

Number of signal events in dark photon parameter space for the far (left) and near (right) detector locations, given an integrated luminosity of $300{\mathrm{fb}}^{-1}$ at the 13 TeV LHC. The different colors correspond to the three production mechanisms: ${\pi }^0\to A^{\prime }\gamma$ (red), $\eta \to A^{\prime }\gamma$ (orange), and proton bremsstrahlung (green). Contours represent the number of signal events $N_{\mathrm{sig}}$. The gray shaded regions are excluded by current experimental bounds. The black stars correspond to the representative parameter-space points of Eq. (8).

Figure 10

Combined 95% C.L. exclusion reach on dark photon parameter space for the far (left) and near (right) detector design benchmarks for an integrated luminosity of $300{\mathrm{fb}}^{-1}$ (solid line) and $3{\mathrm{ab}}^{-1}$ (dashed line). The gray shaded regions are excluded by current experimental bounds, and the colored contours represent projected future sensitivities of LHCb [78, 79], HPS [80], SeaQuest [60], and SHiP [65].

Figure 11

Same as in Fig. 5, but for the off-axis detector location with $(L_{\min },L_{\max })=(90\mathrm{m},100\mathrm{m})$.

Figure 12

Results for the off-axis detector design in dark photon parameter space. The gray shaded regions are excluded by current experimental bounds. Left: Number of signal events given an integrated luminosity of $300{\mathrm{fb}}^{-1}$ at the 13 TeV LHC. The different colors correspond to the three production mechanisms ${\pi }^0\to A^{\prime }\gamma$ (red), $\eta \to A^{\prime }\gamma$ (orange), and proton bremsstrahlung (green). Contours represent the number of signal events $N_{\mathrm{sig}}$. Right: Combined 95% C.L. exclusion reach for an integrated luminosity of $300{\mathrm{fb}}^{-1}$ (solid line) and $3{\mathrm{ab}}^{-1}$ (dashed line), assuming negligible background. The colored contours represent projected future sensitivities of LHCb [78, 79], HPS [80], SeaQuest [60], and SHiP [65].