YAP:Ce – Fast Scintillator & Optical Properties

• YAP:Ce is a Ce-activated oxide perovskite known for its fast (≈28 ns) scintillation response and well-resolved Ce3+ 4f–5d electronic transitions.
• It is synthesized in various forms—single crystals, films, powders, and ceramics—via methods like Czochralski growth and liquid phase epitaxy, affecting defect profiles and optical performance.
• In neutron detection, YAP:Ce’s reduced afterglow and superior neutron-surface behavior yield roughly 20% higher UCN counts compared to ZnS:Ag, despite lower intrinsic light yield.

Yttrium aluminum perovskite activated with cerium, conventionally denoted YAP:Ce or YAlO$_3$:Ce, is a Ce-activated oxide perovskite investigated both as an optical-luminescent material and as a fast scintillator. Within the reported literature, it appears in bulk single crystals, single crystalline films, powders, ceramics, powdered scintillator screens, and $^{10}$B-coated single-crystal neutron detectors. Its documented properties combine a resolved Ce$^{3+}$ 4f–5d electronic structure, defect-sensitive excitonic spectroscopy, a fast scintillation response on the order of tens of nanoseconds in neutron-detector studies, and strong application relevance in ultracold-neutron counting and in-vacuum alpha detection (Bulyk et al., 2024, Babin et al., 2011, Krivos et al., 4 Sep 2025, Krivoš et al., 2024, Meyer et al., 19 Jun 2026).

1. Material identity, forms, and reported physical realizations

YAP:Ce is identified as cerium-doped yttrium aluminum perovskite, with host composition

$\mathrm{YAlO_3}.$

In the cited spectroscopy work on bulk crystals, the Ce concentration in the YAP single crystal was 0.5 mol.%. In the ultracold-neutron powdered-screen study, no numerical cerium concentration is stated; the activator role is conveyed only through the notation “Ce” (Bulyk et al., 2024, Krivos et al., 4 Sep 2025).

The material is reported in several physical forms. A defect-spectroscopy study compares single crystals (SC) grown by the Czochralski method, single crystalline films (SCF) prepared by liquid phase epitaxy (LPE) using PbO-B$_2$O$_3$ flux in a Pt crucible, nanopowders synthesized by coprecipitation with particle size about 27 nm, and ceramics obtained by heating those powders at 1000 $^{\circ}$C. The growth route is treated as physically consequential because the high-temperature melt route favors antisite- and vacancy-related defects, whereas LPE suppresses those intrinsic defects but introduces Pb- and probably Pt-related centers (Babin et al., 2011).

For neutron detection, two YAP:Ce implementations are distinguished. One is a powdered scintillator screen fabricated from a YAP:Ce crystal slab purchased from Epic-Crystal, then ground, size-selected using a 38–70 $\mu$m mesh, rolled onto an optical adhesive layer, and coated with 120 nm $^{10}$B by electron beam evaporation. The fabricated powder layer thickness is reported as about 75 $\mu$m, the screen area as 21.3 \pm 0.5\ \mathrm{cm^2}), and the grain size is listed in one place as < 38 $^{10}$0m, while the fabrication section states that a 38–70 $^{10}$1m mesh was used (Krivos et al., 4 Sep 2025).

The second neutron-detector realization is a single-crystal YAP:Ce absolute UCN detector. In that work, the crystal used for UCN measurements had a $^{10}$2 $^{10}$3B coating on both sides, and in the UCN comparison it was placed directly on a tempered glass window in front of a 51 mm Hamamatsu R7724 PMT, without optical coupling (Krivoš et al., 2024).

In an alpha-detection benchmark for Associated Particle Imaging, YAP:Ce is used as the conventional reference sample in the form of a single-crystal grown by Crytur, with dimensions $^{10}$4 and a $^{10}$5 Al coating to improve light collection (Meyer et al., 19 Jun 2026).

2. Ce$^{10}$6 electronic structure, emission, and thermal quenching

A room-temperature vacuum-UV absorption study establishes the YAP host bandgap as

$^{10}$7

obtained from the absorption edge under a direct-gap assumption. The same paper states that indirect-gap fits were poorer and yielded values lower by about 0.2–0.3 eV. The reported low-lying Ce$^{10}$8 4f $^{10}$9 5d absorption structure comprises a lowest 5d absorption peak at 4.05 eV, a second-lowest 5d level at 237 nm $^{3+}$0, and a third-lowest 5d level at 216 nm $^{3+}$1, with extracted separations $^{3+}$2 and $^{3+}$3. The experimental Ce 5d barycenter is given as 4.76 eV, the theoretical centroid shift as $^{3+}$4, and the crystal-field splitting of the 5d manifold as 1.68 eV experimentally versus 2.30 eV theoretically (Bulyk et al., 2024).

At low temperature, YAP:Ce shows the characteristic broad-band Ce$^{3+}$5 5d $^{3+}$6 4f emission split into transitions to

$^{3+}$7

with the emission shown approximately across 320–440 nm. The zero-phonon line is stated to be unobserved because of very large electron-phonon coupling. Using the midpoint between the absorption and luminescence band maxima, the distance between the lowest 4f and lowest 5d levels is inferred to be 3.76 eV (Bulyk et al., 2024).

The temperature dependence is modeled by

$^{3+}$8

with the fit yielding

$^{3+}$9

for the position of the lowest Ce 5d level below the conduction-band minimum. Thermal quenching of Ce$\mathrm{YAlO_3}.$0 luminescence in YAP:Ce is reported to begin above $\mathrm{YAlO_3}.$1 and is interpreted as thermal ionization of the lowest Ce 5d state into the conduction band, followed by inefficient recapture because electron traps exist. Combining the bandgap, lowest 5d absorption peak, and $\mathrm{YAlO_3}.$2, the Ce 4f ground state is estimated at 2.31 eV above the valence-band maximum from the raw absorption estimate, or 2.6 eV above the valence-band maximum using the midpoint-based estimate (Bulyk et al., 2024).

The same paper identifies a central theoretical discrepancy. Using the standard free-ion Ce$\mathrm{YAlO_3}.$3 5d barycenter reference 6.35 eV, the Dorenbos-style treatment gives a theoretical YAP 5d barycenter of 3.77 eV, substantially below the experimental 4.76 eV. The authors propose that the discrepancy is removed if the host bandgap / conduction-band edge is used as the relevant reference energy for the 5d red shift; for YAP, this yields a lowest 5d energy referred to the bandgap of 4.05 eV, matching the observed lowest Ce absorption band (Bulyk et al., 2024).

A separate defect-spectroscopy paper reports the YAP band gap as

$\mathrm{YAlO_3}.$4

in the context of exciton spectroscopy rather than vacuum-UV edge fitting (Babin et al., 2011). This suggests that within the present literature set, the reported host-gap value depends on the measurement framework and analysis convention.

3. Excitons, antisite defects, and slow components in Ce luminescence

The defect-centered spectroscopy of YAP and YAP:Ce distinguishes among self-trapped excitons (STE), excitons localized at intrinsic defects, excitons localized near impurity centers, and recombination-related bands. In YAP SCF, the dominant intrinsic emission is the $\mathrm{YAlO_3}.$5 band assigned to the STE, with excitation at 7.83 eV in the text and near 7.91 eV in a cited table entry. In YAP SC, where antisite defects are abundant, the dominant 5.63 eV and 4.12 eV emissions are assigned to

$\mathrm{YAlO_3}.$6

and

$\mathrm{YAlO_3}.$7

respectively, that is, excitons localized at an isolated antisite defect and at an antisite defect associated with the nearest oxygen vacancy (Babin et al., 2011).

The same work identifies additional defect-related emissions and impurity centers. In SC, a 5.28 eV band is treated as likely recombination-related rather than a true localized exciton, and a 2.45 eV defect emission is associated with $\mathrm{YAlO_3}.$8-type defects. In SCF, impurity-related centers include a 4.16 eV emission assigned to an exciton localized near a Pt$\mathrm{YAlO_3}.$9 impurity, a 3.63 eV emission assigned to a single Pb$_2$0-based center, and a 3.15 eV emission assigned to a dimer $_2$1 center (Babin et al., 2011).

For YAP:Ce, Ce$_2$2 emission at 4.2 K has two components, 3.30 eV and 3.55 eV. The lowest-energy exciton band exciting Ce emission differs between crystal forms: in YAP:Ce SC the main exciton excitation band is at $_2$3, assigned to an exciton perturbed by Ce$_2$4 associated with antisite defect $_2$5,

$_2$6

whereas in YAP:Ce SCF the main exciton excitation band is at $_2$7 and is assigned to an exciton perturbed by an isolated Ce$_2$8 center,

$_2$9

The absence of antisite defects in SCF and their presence in SC is used as the basis for this assignment (Babin et al., 2011).

The practical significance of these defects lies in scintillation kinetics. The slow component of Ce emission is attributed mainly to tunneling recombination between electron and hole centers created by photostimulated disintegration of regular excitons and photoionization of Ce$_3$0. For excitation energies below 7.5 eV in SC, the slow component can also arise from tunneling recombination involving an antisite-defect-related electron center and a hole Ce$_3$1 center. The same paper concludes that direct defect-to-Ce energy transfer is possible only in a limited region around 7.6–7.7 eV and is not the dominant cause of slow Ce decay (Babin et al., 2011).

Trap-related thermally stimulated processes further connect defect structure to Ce luminescence. In X-irradiated YAP:Ce SC, TSL peaks occur at $_3$2 and $_3$3, assigned to thermal release of electrons trapped at $_3$4-related and $_3$5-related centers, followed by recombination with Ce$_3$6 hole centers. These peaks are absent in the studied powders and ceramics, supporting the conclusion that the antisite-related traps are characteristic of melt-grown SC (Babin et al., 2011).

Defect energetics are treated explicitly by DFT within LSDA using FPLO on 20-atom Pnma supercells. The antisite-pair formation energy is reported as 10.8 eV without a nearby vacancy, 8.1 eV with nearby $_3$7, 10.6 eV with vacancy far away, and 10.2 eV with nearby Ce$_3$8. The total energy of the Ce-doped lattice is lower by 0.25 eV when $_3$9 is near Ce$^{\circ}$0 rather than far away. For the $^{\circ}$1 complex, the Y ion shifts toward the vacancy by about 0.34 $^{\circ}$2. NMR corroborates the presence of antisite-related environments: commercial powder shows an estimated 15–20% concentration of such defective Y sites, whereas commercial YAP crystals show only 2–3%, and nanocrystalline ceramics contain about 80% secondary non-perovskite phase (Babin et al., 2011).

4. Neutron-converter geometry and the role of YAP:Ce in ultracold-neutron detection

In the UCN-detector studies, YAP:Ce is explicitly not the neutron converter. Neutron sensitivity is supplied by a thin $^{\circ}$3B layer deposited on the scintillator surface. UCN capture proceeds through the $^{\circ}$4B reaction,

$^{\circ}$5

producing an alpha and a lithium ion whose energy deposition in YAP:Ce generates scintillation light (Krivos et al., 4 Sep 2025).

The absolute-detector study states the two reaction branches as

$^{\circ}$6

Because UCN are slow, the required converter thickness can be small: $^{\circ}$7 of $^{\circ}$8B is stated to give a capture probability of 96% for $^{\circ}$9 neutrons (Krivoš et al., 2024).

The UCN motivation for choosing YAP:Ce over ZnS:Ag is presented in linked optical and neutron-optical terms. In both UCN papers, YAP:Ce has a short decay time of 28 ns, whereas ZnS:Ag has long and complex decay components extending into several hundred nanoseconds or many microseconds, depending on context. The absolute-detector paper further emphasizes a calculated Fermi potential $\mu$0 for YAP:Ce. The intended consequence is that neutrons below the material optical potential reflect from the crystal surface rather than entering and being lost, while a smooth crystal also supports a more controllable $\mu$1B-coated surface than a powder-based ZnS screen (Krivoš et al., 2024, Krivos et al., 4 Sep 2025).

The powdered-screen implementation and the single-crystal absolute detector differ materially. In the powdered-screen experiment, screens were attached to a tempered-glass window, all screens were coupled to the tempered glass with an optical adhesive, and two identical 51 mm Hamamatsu R774 photomultiplier tubes (PMTs) observed the paired screens symmetrically. In the absolute-detector paper, by contrast, the coated YAP:Ce crystal was placed on tempered glass in front of a 51 mm Hamamatsu R7724 PMT, without optical coupling (Krivos et al., 4 Sep 2025, Krivoš et al., 2024).

5. Scintillation speed, phosphorescence, and measured UCN performance

The powdered-screen study measured YAP:Ce waveforms under 5.5 MeV alpha particles from $\mu$2Am and under UCN capture in the $\mu$3B layer. The decay spectrum was fit with an exponential yielding

$\mu$4

matching the manufacturer value. Relative to ZnS:Ag, the pulse height under UCN excitation was 75% of ZnS:Ag for YAP:Ce and 50% of ZnS:Ag for LYSO:Ce. In the final background-subtracted spectra, the average energy scale among the three scintillators is summarized as

$\mu$5

Thus YAP:Ce gives smaller integrated pulses than ZnS:Ag even while exhibiting much faster timing (Krivos et al., 4 Sep 2025).

A central practical result is YAP:Ce’s reduced afterglow. After exposure to room light for several hours and then storage in a dark box, the YAP:Ce screen had 60% lower count rates than ZnS:Ag after 2 days. The same trend is restated qualitatively: the phosphorescence level reached by ZnS:Ag after 2 days is reached by YAP:Ce in less than 1 day. LYSO:Ce behaved much worse, showing more than 10 times higher count rates than ZnS:Ag after 2 days (Krivos et al., 4 Sep 2025).

The counting measurements in that study used one 10-minute signal run with the UCN gate valve open and one 10-minute background run with the gate valve closed for each scintillator pair. Waveforms were digitized with a CAEN DT5724 100 MS/s, with waveform length 2 $\mu$6s, and integrated over a 200 ns window from the rising edge. ZnS:Ag exhibited retriggering due to its long decay tail, producing excess low-energy counts below about 300 ADU and requiring a threshold treatment with

$\mu$7

YAP:Ce, by contrast, was described as almost void of the re-triggering problem, and its background was subtracted without modification (Krivos et al., 4 Sep 2025).

The UCN counting result in that experiment was

• YAP:Ce: $\mu$8 counts, corresponding to $\mu$9
• ZnS:Ag measured simultaneously with YAP:Ce: $^{10}$0 counts, corresponding to $^{10}$1

with relative uncertainties in areal count density of 2.37% for YAP:Ce and 2.65% for ZnS:Ag. The paper concludes that YAP:Ce detected about 20% more UCN per square centimeter than ZnS:Ag. It also states that both YAP:Ce and LYSO:Ce are viable UCN detectors for high count rate experiments, but that YAP:Ce outperformed LYSO:Ce by every tested metric (Krivos et al., 4 Sep 2025).

The separate absolute-detector study reaches a related but distinct set of results. It measures a neutron counting efficiency for neutrons captured in the $^{10}$2B coating of

$^{10}$3

obtained by coincidence with the $^{10}$4 gamma line in an HPGe detector, using

$^{10}$5

Under direct UCN comparison with a ZnS screen, the corrected count ratio implied

$^{10}$6

meaning that YAP:Ce counted 8% to 28% more UCNs than ZnS after the stated corrections (Krivoš et al., 2024).

That improvement occurred despite severe optical penalties. The same paper reports about 60% light loss caused by 120 nm $^{10}$7B coating and about 50% light loss through one deuterated polystyrene scintillator layer. In waveform comparison, the YAP pulse contained about 40 photoelectrons, whereas ZnS contained about 350 photoelectrons in the tested geometry. The authors’ interpretation is that YAP’s higher UCN count is most likely due not to brighter scintillation but to superior neutron-surface behavior, especially a more uniform $^{10}$8B coating on the smoother crystal surface (Krivoš et al., 2024).

6. Benchmarking beyond UCN, comparative materials context, and limitations

In position-resolved fast-timing alpha-detection studies, YAP:Ce functions as the incumbent comparison material. The reference sample was a single-crystal YAP:Ce with $^{10}$9 dimensions and a $\mu$0 Al coating. Under pulsed X-ray TCSPC, its scintillation response was fit with a rise time of $\mu$1 and two decay components, $\mu$2 (95%) and $\mu$3 (5%). Only about 30% of the total light yield is emitted in the first 10 ns. In the same work, YAP:Ce is assigned refractive index $\mu$4, emission wavelength $\mu$5 in the optical-property table discussion, negligible absorption $\mu$6, and a host band gap of $\mu$7, which places the dopant emission far below the absorption edge and leads to negligible reabsorption in that comparison (Meyer et al., 19 Jun 2026).

The same benchmark finds YAP:Ce to be the brightest of the tested alpha scintillators, with an absolute light yield of $\mu$8. The detected peak is reported as $\mu$9 photoelectrons in a figure caption and 4094 photoelectrons in the summary table. Its measured detector timing resolution is $^{10}$00, whereas the experimentally demonstrated position resolution is better than $^{10}$01, also written elsewhere in the same paper as at least $^{10}$02 or $^{10}$03. The comparison therefore places YAP:Ce as superior in brightness and experimentally demonstrated spatial resolution, but slower than the proposed GaN:Si and ZnO:Ga alternatives (Meyer et al., 19 Jun 2026).

The UCN absolute-detector study provides a complementary heavy-ion-response characterization through Birks’ law. From a $^{10}$04Gd alpha calibration, the $^{10}$05-particle light output is reported as

$^{10}$06

corresponding to 27% of the tabulated alpha energy, and the derived Birks coefficient is

$^{10}$07

This quantifies the heavy-particle quenching that is directly relevant to $^{10}$08B-based neutron detection, where the scintillation signal arises from an alpha and a $^{10}$09Li ion rather than from electrons (Krivoš et al., 2024).

Several limitations recur across the literature. In UCN work, YAP:Ce produces smaller pulses than ZnS:Ag: the UCN pulse height is only 75% of ZnS:Ag in the powdered-screen comparison, and the reduced signal-background separation is explicitly noted as a drawback that may be improved with a more efficient light collection system. The observed performance differences are also said likely to originate from variations in the manufacturing processes and differences in emission spectra, so the measured advantage is not presented as a purely intrinsic material constant independent of detector construction. The same UCN paper therefore retains a role for ZnS:Ag, stating that ZnS:Ag is still a good option when high counting rates are not expected (Krivos et al., 4 Sep 2025).

In comparative host-chemistry terms, YAP:Ce is less thermally robust than LuAP:Ce. The LuAP comparison shows no observable Ce luminescence quenching up to 873 K, whereas YAP:Ce quenches above 650 K because its lowest Ce 5d state lies only 1.27 eV below the CBM. The explanation given is that LuAP combines a slightly larger bandgap with a lower-energy Ce 4f ground state relative to the valence-band maximum, placing the Ce excited states more favorably in the gap (Bulyk et al., 2024).

Taken together, the reported corpus defines YAP:Ce as a fast Ce-activated oxide whose practical identity depends strongly on context. In defect-poor or carefully engineered forms it offers clean Ce luminescence, fast response, negligible self-absorption in some optical geometries, and strong position resolution. In melt-grown single crystals, antisite- and vacancy-related defects introduce slow components and trap-mediated recombination. In UCN instrumentation, YAP:Ce’s combination of 28 ns decay, reduced phosphorescence, and favorable neutron-optical behavior makes it a strong replacement candidate for ZnS:Ag, even though it is optically dimmer in the tested implementations. In fast-timing alpha detection, it remains the conventional benchmark rather than the timing optimum, but it retains a documented advantage in experimentally demonstrated sub-millimeter spatial resolution (Babin et al., 2011, Krivos et al., 4 Sep 2025, Krivoš et al., 2024, Meyer et al., 19 Jun 2026).