Physical property

DDS is a cost-effective organosilicon compound bearing double benzene rings, and some organic dyes can be efficiently dissolved in it. Organosilicon has been demonstrated to have superior radiation resistance relative to conventional PVT- or PS-based scintillators [1, 22-25]. Therefore, it is expected that the PVT-DDS-based PSD organic scintillator will exhibit enhanced radiation stability relative to a pure PVT-based scintillator. The aromatic matrices (PVT and DDS) inherently exhibit specific optical properties: they emit fluorescence primarily in the ultraviolet region (250–380 nm) [26], and their absorption spectra show noticeable absorption starting around 380 nm, increasing sharply below 350 nm [12]. In particular, PVT exhibits excitation and emission peaks at 266 and 310 nm, respectively, whereas DDS shows corresponding peaks at 302 and 353 nm (Fig.S1, Supplementary Materials). To minimize self-absorption by the matrix itself, primary fluorophore PPO is added to absorb the photons from the emission of the matrix and emit photons with a peak emission at 393 nm. To maintain scintillation in the blue-green region of the optical spectrum, which matches well most of the PMT spectrum sensitivity, a further fluorophore dye, namely wavelength-shifter MDAC, is chosen to shift the scintillators emission wavelength to 410–500 nm. As the primary fluorophore, PPO has high solubility in aromatic solvents, whereas MDAC, a secondary dye with high light output (LO), is used for the routine preparation of plastics. The PPO-MDAC dye pair can probably lead to excellent PSD capability of the PVT-DDS-based scintillators from the viewpoint of the basic principle of energy transfer [27]. With 20–40 wt% DDS, the synthesized scintillators are colorless and optically transparent without yellowing and surface dye precipitation to the naked eye, even after six months of preparation (Fig. 1(a)). However, when the DDS content exceeds 40 wt%, microcracks, cracking, and yellowing are observed in the scintillators (Fig. 1(b)-Top) are observed in the scintillators.

Fig. 1Full-screen View

(Color online) #(a) Photographs of the synthesized scintillators (RK₁-RK₅) taken at the time after two, 96 and 176 days of preparation. The scintillators are stored at room temperature in an air atmosphere. (b) Top: The additional synthesized scintillators containing 30 wt% PPO and loaded 30 wt%, 40 wt%, 50 wt%, 60 wt% DDS monomers. Bottom: The additional synthesized scintillators with 50 wt% PPO and doping DDS content ranging from 0 wt% to 30 wt%

In contrast to a pure PVT-based sample (0 wt% DDS), scintillators containing 10–30 wt% DDS and 50 wt% PPO remained optically transparent after 15 days of storage (Fig. 1(b)-bottom), revealing that fluorophore dissolution stability was significantly improved by DDS monomers. This may be explained by the molecular isolation within the network subdomains constructed by such functional monomers, where the spatial separation of fluorophores minimized the intermolecular interactions. Combined with our previous research [21], a conclusion may be drawn here that any small amount of functional monomers that are loaded into the host matrix of scintillator and participate in cross-linked network formation, will always help dye (solute) dispersion and then prevent the dye aggregation.

The mechanical hardness values of the scintillators (RK₁-RK₅) were determined to be 45.5±0.8 HD, 41.4±0.5 HD, 39.7±0.3 HD, 39.3±0.9 HD, and 37.2±0.5 HD, respectively, indicating a general decrease with increasing DDS content. As evidenced in Fig. 2, the 30 wt% DDS composite exhibited no significant spectral alterations in the FTIR spectra compared to the pure PVT matrix (0 wt% DDS), except for a distinct absorption band at 1065.59 cm^-1. This absorption line corresponds to the Si-O-Si vibration. The absence of both new characteristic peaks and peak shifts suggests that the incorporation of DDS does not chemically modify the polymer chain structure but rather establishes a physically compatible blended system. We speculated that during thermal polymerization, chains of low-molecular-weight siloxane derived from DDS monomers were incorporated into the PVT-based cross-linked network. This structural integration, combined with the inherent elasticity of DDS siloxane components, likely contributes to the reduced mechanical hardness. Such a reduction is not necessarily unfavorable in terms of elastic recovery properties. To confirm this, a blunt needle with a diameter of 0.79 mm was used to press the scintillator sample (30 wt% DDS) for one minute. The time evolution of a prepared scintillators (RK₃) in surface morphology is shown in Fig. 3 and Fig.S2 in the Supplementary Materials. The initial indentation produced by pressing exhibited a cylindrical geometry with a depth of 1.5 mm. Remarkably, within 30 s of post-compression, the indentation underwent near-complete recovery to a planar surface. After 60 s, residual deformation or surface extrusion traces were almost imperceptible, confirming that the introduction of DDS imparted elastic recovery properties to the scintillator. This phenomenon may be attributed to the high mobility of siloxane main chains facilitated by the exceptionally low energy barriers for Si-O-Si bond bending and torsional motions [28].

Fig. 2Full-screen View

(Color online) FTIR spectra of the scintillator with 0 wt% DDS and the scintillator with 30wt% DDS (RK₃)

Fig. 3Full-screen View

(Color online) Microscope photograph of the RK₃ sample containing 30 wt% DDS, taken at the time points in the range of 0-240 s as indicated in the pictures

Thermal stability was evaluated using thermal gravimetric analysis (TGA) under a protective nitrogen atmosphere. The temperature was increased at a heating rate of 10 ℃/min from 30 ℃ to 800 ℃. The results are shown in Fig. 4. No mass loss occurred below 120 ℃. The synthesized scintillators exhibited similar thermogravimetric profiles during the first degradation stage (120-260 ℃), where the samples underwent the evolution of inter- or intrachain reactions, generating low-molecular-weight products, including volatile compounds, linear oligomers, and fluorophores (PPO and MDAC). Compared with the 0 wt% DDS scintillator (decomposition temperature ~160 ℃), the scintillators containing DDS started to decompose at a lower temperature, perhaps because of the presence of phase-separated domains created between the DDS and PVT constituents. Above 210 ℃, the thermal degradation rate showed a slight dependence on DDS content in a complex manner requiring future study. In the second stage (260-380 ℃), DDS-containing samples exhibited faster weight loss than pure PVT-based scintillators, indicating reduced thermal stability of DDS short-chain structures. During the third stage (380-800 ℃), degradation continued, with 2–5% solid residue primarily consisting of SiOC carbonaceous substance from mineralization [29].

Fig. 4Full-screen View

(Color online) TGA curves obtained in nitrogen atmosphere for the synthesized scintillators RK₁ − RK₅ and the additional 0 wt% DDS scintillator

Optical performance

The synthesized scintillators utilized PVT and DDS as matrices. Critically, the emission spectra of both PVT and DDS matrices significantly overlap with the absorption spectrum of the fluorophore PPO, enabling efficient Förster resonance energy transfer (FRET) from the matrices to PPO. To simultaneously excite the PVT and DDS matrices, a 270 nm UV source was selected. This wavelength lies well below the 310–330 nm absorption peaks of both the PPO and MDAC fluorophores, thereby minimizing direct fluorophore excitation and allowing us to specifically investigate the resonance energy transfer originating from the excited matrices. Figure 5(a) presents the excitation-emission matrix of the RK₃ scintillator (30 wt% DDS). Under 270 nm excitation, its emission spectrum (scanned from 400–600 nm) shows a distinct 434 nm peak. The excitation spectrum over the range of 250–400 nm was monitored using a 434 nm wavelength. The prominent emission peak at 434 nm signifies the characteristic fluorescence of the MDAC dye. Notably, its presence under 270 nm excitation provides strong evidence for efficient energy transfer from the excited matrices to the MDAC dye.

Fig. 5Full-screen View

(Color online) (a) Excitation-emission spectrum of the scintillator with 30 wt% DDS (RK₃). The excitation spectrum with excitation wavelengths of 250-400.0 nm was recorded at the 434.0 nm emission maximum, while the fluorescence emission spectra were induced by a 270.0 nm-wavelength UV. (b) Emission spectra excited at 270 nm wavelength and (c) ultraviolet-visible transmittance spectra in the wavelength range of 250–800 nm for RK₁-RK₅ samples and the additional synthesized scintillator with 0 wt% DDS

Photoluminescence (PL) spectra excited at 270 nm are shown in Fig. 5(b) and Fig. S3 (Supplementary Materials). Increasing the DDS content promotes π-π stacking of the benzene moieties, enhancing exciplex formation, and resulting in broadened emission peaks with a 10–30 nm red shift. The PL intensity initially increases with DDS concentration but decreases above 30 wt% DDS. At such high concentrations, fluorescence emission is likely dominated by aggregation-caused quenching (ACQ), where dense π-π stacking creates compact molecular proximity. This facilitates electron energy level rearrangement in the matrix, primarily through splitting of ground-state and first excited-state levels. Consequently, the concentration-dependent PL intensity correlates with the scintillator optical transparency, as both phenomena originate from photon loss mechanisms in the aggregated system.

The optical transparency of the scintillators was very high (Fig. 5(c)). All samples exhibited transmittance values between 70% and 87%, which was comparable to that of the commercial scintillator EJ-276 (Fig. S4, Supplementary Materials). Within the visible spectrum, the transmittance initially increased with the DDS content up to 30 wt%. However, further increases in DDS content led to a decrease in transmittance. This non-monotonic behavior was attributed to two factors: self-absorption within the matrix material and aggregation of the dye molecules at higher DDS concentrations. Below 30 wt% DDS, monomers were sparsely dispersed. Most did not participate in thermal polymerization and acted as photon scattering centers, reducing transparency. As DDS approaches 30 wt%, dispersed monomers polymerized into linear-chain segments, enhancing transparency. However, above ~30 wt% DDS, the high density of benzene rings promoted strong π-π stacking interactions that induced fluorophore aggregation [30]. At these concentrations, numerous low-molecular-weight DDS chains embedded within the cross-linked PVT network, creating phase-separated domains with mismatched plastic deformation behaviors. This structural incompatibility generated interfacial shear stresses and micro-cracks at phase boundaries (Fig. 1(b)-Top), causing the observed transparency reduction.

PSD and relative LO

The synthesized scintillators were irradiated with a ²³⁸Pu-Be source to study their PSD performance. The PSD method is based on the different proportions of the prompt and delayed components in scintillation pulses and is used to identify particles. When energetic particles pass through a scintillator, the π electrons of the scintillator molecules are ionized and recombine to produce both singlet and triplet excited states, which promptly de-excite to the lowest singlet (S₁) and triplet states (T₁) [31]. Transition through S₁ to the ground state (S₀) with fluorescence emission on a timescale of 10^-9–10^-8 s yields the prompt component. Direct transition from T₁ to S₀ is spin-forbidden; instead, a bimolecular process of triplet-triplet interaction and annihilation (T₁ + T₁ → S₁ + S₀), namely, the TTA process, occurs kinematically with subsequent de-excitation of the S₁ to S₀ (S₁ → S₀+hν) over a longer timescale (10^-4 s), yielding the delayed component. The T₁ state density depends on the ionizing power of the incident particles. For a given incident energy, heavy charged particles produce more T₁ states per unit path length than light particles, resulting in a more pronounced slow scintillation component. This difference in the ratio of integrated charges between the prompt and delayed components forms the basis of the PSD method.

The PSD capability was evaluated using the charge integration method [32]. For each digital waveform, the integrations over a short-time gate (Q_short, fast component) and a long-time gate (Q_total) yield the PSD parameter: , where Q_short is the charge integral from the waveform start to the prompt peak, and Q_total represents the full-waveform integral. The two-dimensional (2D) plots of PSD versus Q_total are shown in Fig. 6 and Fig. S5 (Supplementary Materials) for the prepared scintillators (RK₁-RK₅, and the 0 wt% DDS sample) two days after synthesis, alongside those for EJ-276. Excellent separation between neutron and gamma-ray signals was observed. The locus with higher PSD values corresponds to neutrons, whereas the locus with lower PSD values corresponds to gamma-rays. Notably, the neutron and gamma bands exhibit distinct curvature, potentially arising from PMT nonlinearity under high-voltage operation (Fig. S6, Supplementary Materials) [33, 34]. The capability of separating neutrons from gamma-rays is quantified by the figure of merit (FOM), defined as . Here, S represents the separation between the gamma and neutron distribution peaks, and σ_γ (σ_n) denotes the full width at half-maximum (FWHM) of the gamma (neutron) peaks in the one-dimensional PSD distribution. These parameters were obtained using Gaussian fitting. Consequently, the PSD FOM depends on two factors: the neutron/gamma peak separation (S) and their respective widths. Increasing the LO reduces the FWHM of both distributions, thereby enhancing the PSD FOM [27].

Fig. 6Full-screen View

(Color online) Two-dimensional PSD distributions of the synthesized RK₁-RK₅ samples and the additional synthesized scintillator with 0 wt% DDS

To quantitatively assess the dependence of the PSD capability on the DDS concentration, we extracted the PSD values within the Q_total range of 0.2–1.2 MeVee (MeVee is the MeV electron equivalent energy) from the 2D distributions. Figure 7(a) plots the calculated FOM values against the DDS concentration, whereas Fig. 7(b) shows the FOM values calculated within each interval of to bins (in units of MeVee) for the prepared scintillators (black dots). As clearly demonstrated in Fig. 7(b), each scintillator achieves superior neutron/gamma separation with an FOM of 1.27 at values down to 0.15 MeVee. The value 1.27 is typically recognized as the FOM threshold for efficient pulse shape discrimination (PSD) [11].

Fig. 7Full-screen View

(Color online) (a) The obtained FOM values evaluated over the energy range of 0.2 to 1.2 MeVee, (b) dependence of the FOM values calculated within the each interval of Q_total-0.1 to Q_total+0.1 bins (in unit of MeVee), where Q_total values are selected in the range of 0.12 to 1.9 MeVee, and (c) relative LO as a function of the DDS concentrations for the prepared scintillators (RK₁-RK₅) and the scintillator containing 0 wt% DDS

The LO of the prepared scintillators was studied using a ¹³⁷Cs source by comparing the Compton edge positions in their pulse height spectra with those of the EJ-200 standard (Fig. S7, Supplementary Materials). As shown in Fig.7(c), the LO relative to EJ-200 exhibits an overall decreasing trend within the calculated uncertainties as the concentration of the DDS monomer increases. Conversely, the neutron/gamma PSD capability, quantified by the FOM, shows a slight concentration-dependent enhancement before reaching a plateau (Fig. 7(a)). This apparent inverse correlation between PSD performance and LO is likely attributed to the formation of a more efficient triplet (T₁) state energy transfer network. This network arises from the gradual substitution of PVT with DDS, which increases the density of benzene rings that are required for triplet energy migration. Furthermore, the enhanced dissolution of fluorophores facilitated by DDS incorporation likely affects the local T₁ excitation density and the efficiency of the TTA process governing the neutron/gamma separation parameter S.

Long-term stability was evaluated through repeated measurements at 96 and 172 days after synthesis (Fig. 7(a) and (c)). After approximately six months of aging, the synthesized scintillators exhibited reductions of 9–16% in FOM and 27–41% in LO relative to their initial values. This degradation highlights a significant challenge. Although the incorporation of DDS offers potential advantages in terms of radiation resistance and PSD performance, the observed aging effects, likely stemming from factors such as fluorophore aggregation and potential matrix instability at higher DDS loadings (as suggested by Fig. 1(b)), or chemical degradation pathways, currently limits the practical long-term utility of these materials. Addressing this stability issue, potentially through optimized fluorophore selection, stabilization additives, or refined matrix compositions, is crucial for realizing the promising PSD performance of PVT-DDS-based scintillators for demanding applications.