Induced ageing of ZnS:Ag microparticles exposed to 13 keV electron beam

Phosphorescent microparticles made of ZnS:Ag were exposed to pulsed electron beams with an energy of 13 keV for periods of time between 30 min and 240 min. An XRD analysis showed no modification of crystalline structure. The average cristalites of ZnS:Ag was 62 nm deduced from SEM imaging. The luminescence spectra showed a decreasing activity with 40% after 30 min of irradiation at a fluence of 5.79 ×1016 electrons/cm2. The broad peak between 445 nm to 480 nm centered aroud 460 nm with a FWHM almost constant aroud 80 nm show no shifting. After a long exposure (over 240 min) and a fluence of 4.60 ×1017 electrons/cm2, the powder suffered a blackening effect attributed to formation of dead layers under electronic excitation combined with increasing of Sulphur vacancies, quantitatively confirmed by EDS analysis, where the proportion of S in ZnS:Ag decreases from 31.42% to 13.75%. Also, the luminescence at this moment dropped to almost 90% under the electron beam effect. The thermal effect could not be correlated with luminescence quenching, which was attributed to the increase in the number of impurities.


Introduction
Used in scintillation-based detection of alpha particles, ZnS:Ag powder is one of the oldest phosphorescent materials.It was used by Crookes [1] to highlight individual scintillations and by Rutherford [2] to discover the atomic nucleus.ZnS powder has been used extensively in fields such as photoelectricity [3], optoelectronics [4] and luminescent devices such as solar cells [5], ultraviolet light emitting diodes [6], optical sensors [7], flat panel displays [8,9] and medicine [10,11].Being characterized by a particularly fast generation of electron-hole pairs, ZnS is also intensively investigated as a photocatalyst for H 2 production [12][13][14] and reduction of CO 2 [15,16].
Rigorous studies [17,18] were undertaken in the early 20th century to understand the physics and chemistry of ZnS powder.With the development and understanding of solid-state physics and using several luminescence techniques such as cathodoluminescence, polarized luminescence, thermoluminescence etc, various models [19,20] were developed to illustrate electronic transitions and the nature of defects responsible for the luminescence behavior in ZnS.There are several models in the literature like donor-acceptor pair emission [21], emission due to clouds of defects [22], excitonic emission [23], localized state emission [24], emission provided by isoelectronic traps [25] or band edge emission [26].Also, the ZnS activated with Ag + ions is a leading scintillator material extensively in neutron, α particles and radon detection [27][28][29][30].Its brilliant blue light emission under excitation with photons, electrons and ions remains one of the most efficient amongst scintillators and phosphors, with a converted visible emission from absorbed incident energy of ∼20% [31] and a light yield of 75 000 photos/MeV [32].When Ag is incorporated into ZnS a self activated emission is observed, which is associated with a complex luminescence center, V Zn -Ag Zn [33,34].Depending on the amount of Ag integrated in the powder, the luminescence peak can increase or decrease, the ideal concentration for maximum luminescence being 3% [35].Above this value, the luminescence begins to decrease, a phenomenon attributed to the concentration quenching effect.Also, the emission peak of ZnS:Ag is slightly shifted compared to pure ZnS.
Several methods for preparing ZnS powders, to be used in the fabrication of ZnS scintillator with good optical properties were proposed such as the hydrothermal method, colloidal processing and wet chemical process methods [36][37][38][39].The luminescence mechanics was studied by several researchers [17,37,38].It is well established that the main luminescence mechanism of ZnS phosphor is radiative transition between deep acceptor (Ag or Cu activator) and donors.The loss of luminescent efficiency can be caused by the degree of imperfection of the crystal, which leads to the increase of non-radiative transitions (i.e.phonos) [39].In the case of intense electron excitation, a dead layer can appear on top in the form of dark dots (i.e.dislocations) correlated with sulfur depletion and the remaining metallic zinc.
When the energy of the incident electrons is greater than the displacement energy of the constituent atoms of the material, additional defects will appear in the powder [40].In our case, electron bombardment of ZnS bonds and destruction of them will lead to Frenkel pairs and new Zn and S vacancies.[41,42].
In this paper, ZnS:Ag powder was tested under vaccum conditions at 10 -4 torr by irradiating it with electron beams of 13 keV.Its morphological structure and luminescence properties stability were investigated.Observing a decrease in luminescence after 30 min, as well as a change in the color of the ZnS:Ag powder, the authors decided to continue exposing the sample up to 240 min until the luminescence will decrease dramatically to see what the exposure limits are.The 13 keV energy of the electron beam was chosen to complement other studies that used lower or higher energy beams and the vacuum conditions to avoid rapid oxidation of the powder [52].

Method
ZnS:Ag powder was exposed to electron beam in an experimental set-up shown in figure 1(a).The powder was commercially purchased, having an average particle size of 8 micrometers while the Cie color coordinates are (0.148; 0.062).Irradiation of the microparticles was carried out by an electron gun consisting of a hollow-anode plasma source and two focusing electromagnetic coils set at 83 and 171 Gauss respectively.The electron gun, described in detail in [54,55], delivers a pulsed electron beam into an interaction chamber found under vacuum conditions of 10 -4 torr achieved by using a turbomolecular pump.The beam was pulsed at a frequency of 46 Hz and the pulse duration was 40 μs.The peak beam current was 1.5 mA measured using a Faraday cup.The signal picked up by the Faraday cup was sent to an oscilloscope by means of a voltage probe through a 50 Ω resistance.Considering the electron beam spot size by 5 mm we can estimate the fluence values for one pulse (7 ×10 11 electrons/cm 2 ) and for the exposure times which are showed on the caption of figures.
The electron beam parameters remain constant throughout the experiment.During irradiation the experimental set-up shown in figure 1(a), was placed inside the vacuum chamber.The blue light emission of excited ZnS:Ag by electron beam of 13 keV can be observe in figure 1(b).The ZnS:Ag powder was positioned on a plate in an oblique position (∼45 degrees) and exposed to the central part of the electron beam with Gaussian profile.The high-resolution imaging of the powder surface was performed by an Apreo S SEM made by Thermo Fisher Scientific.The Everhart-Thornley detector (ETD) was used to work with secondary electrons (SE).For the required images it was used at the parameters of 20 kV and 30 kV, currents of 13 pA and 30 pA, achieving magnifications of 1000× and 5000×.
The luminescence of the ZnS:Ag powder was investigated during irradiation using a spectrometer AvaSpec-ULS2048-USB2 produced by Avantes [56] provided with a UV/VIS grating with 600 l mm −1 , blaze at 300 nm and with a wavelength range from 200 to 850 nm, spectral resolution of ∼0.5 nm and slit size 10 μm.Collection of spectra was realized under a 5 s period integration time.
Energy dispersive spectroscopy (EDS) was carried out using an Octane Elite Super of the Element system from EDAX.The EDS analysis was done on pellets and the tape used is copper.The analysis being done on pellets on a surface smaller than the size of the particles, the integration of the XRD emission signal was done exclusively on the particles, the support and tape used do not interfere during the measurements.The data were collected with the TEAM software.
The temperature was measured with a Lutron TP01, Type K cable probe, 250 °C from Lutron Electronic, the signal collector wire being positioned right in the powder.

XRD analysis
Being a member of II-IV semiconductor family, ZnS presents two main crystalline geometric structures, a sphalerite cubic zincblede(ZB) that crystallizes at lower temperatures (<1020 °C) and wurtzite(W), the hexagonal structure stable at crystallization temperatures > 1020 °C [30,31].Their corresponding band gaps at room temperature are 3.7 eV and 3.8 eV.There are also other hexagonal or rhombohedral shapes, which are typically an effect of twinning or stacking faults of the ZB and W structures and different blend of ZB and W depending on the heating temperature and time.[37,38,57].
The broadening of the peaks of the sample in the XRD spectra shown in figure 2 indicate the crystalline structure.The XRD patterns in the range scan of 25°−60°show five peaks with 2θ values of 28.53°, 33.06°, 47.45°, 56,31°and 59.05°which correspond to (111), ( 200), ( 220), (311) and (222) planes of cubic phase of ZnS matching card number 00-150-9114 from [58] with preferential orientation of (111) plane.The XRD analysis show no modification of the crystalline structure after exposure to a total flux of 4.6 ×1017 electrons/cm 2 .Using Scherrer's equation, where k is the Scherrer shape factor constant (assuming a sphere), λ is the x-ray wavelength (1.54 nm), β is the line broadening at FWHM (Full width at half maximum) in radians and θ being the Bragg angle in degrees (i.e.half of 2θ), an average crystallite size D = 62.20 nm is calculated.No change in crystallite size in any direction was observed after irradiation, due to the fact that the melting point of ZnS is about 1185 °C, and the heating temperature after continuous irradiation for 30 min at a fluence of 5.79 × 10 16 electrons/cm 2 was measured to be only 2 degrees above the ambient temperature (≈ 24 °C).
Although the crystallinity does not seem to be affected by the irradiation, a slight increase in the peaks intensities can still be observed.In general, this increase could indicate more crystallinity in sample [59].In our case, this change could be due to the change in the electron density in the crystallographic positions [60], which could occur after the sulfur dislocation, as we will see below.

Morfological analysis by SEM
Scanning electron microscopy (SEM), a powerful tool to study morphology was used to visualize the microparticles of ZnS:Ag powder, such as figure 3(a).The producer indicate a mean particle size of about 8 μm, but on the SEM an average microparticle diameter of 1.87 μm (figure 3(b)) can be identified using ImageJ software [61], covering a wide range of a few hundreds of nm to 12 μm, thus indicating 2 orders of magnitude larger than average crystallite size.The shapes have predominantly spharelite form.Also, by processing the SEM images with ImageJ, the roughness of the analyzed surface was deduced as it can be seen in figure 3(c).It is known that the increase in roughness could lead to an altered optical properties.The root mean square (rms), Rrms ∼45 μm, and the Roughness Average, Ra ∼18 μm, did not undergo significant changes after prolonged irradiation by 240 min.
After the irradiation of four hours in sessions of 30 min with a a total fluence of 4.6 ×10 17 electrons/cm 2 no significant changes in the surface morphology of the particles could be seen as it show in figures 4(a) and (b).The images was taken for a current 13 pA and voltage 30 kV for the non-irradiated stage (figure 4(a)) and on parameters of current at 25 pA and voltage 20 kV after exposure of 240 min (figure 4(b)).Both pictures was magnified with a 5000 factor.Although no damage to the surface is observed, it can be noted in figures 5(a) and  (b), in the circled regions on the scale of 5 μm, small connections between the smaller particles can be observed as if they were glued.
Other authors [62], irradiating ZnS:Ag powder coated with SiO2 with electron beam at lower energies at 2-5 keV, but with higher exposure of 5-22 C cm −2 (that is, about two orders of magnitude compared to our experiments) found a morphological degradation at the surface level through erosion.However, similar particles without coating at the same parameters did not show morphological changes, although the luminescence decreased dramatically with 55%.
To study possible defect clustering mechanisms in ZnS, other authors [41] used electron and proton beams at fluxes of 10 18 e/cm 2 at energies of the order of MeV, inducing a large concentration of defects and a discoloration of the powder ZnS.After an annealing process, the defects and discoloration disappeared.In an energy regime of the electron beam also of the order of MeV [47], other authors noted an increased crystallinity of the powder.They attributed the effect of stimulating crystal growth to the transfer of kinetic energy of the impinging electrons that rearranged the powder atoms in an orderly manner [63].

EDS spectra
The energy-dispersive spectroscopy (EDS) is the method used to identify the elements in the samples.The reference spectra for ZnS: Ag sample reveal the most intense peaks of Zn and S as espected such as L = 1.01,Kα = 8.62, Kβ = 9.57 for Zn, and K = 2.31 for S, but also small lines of the activators Ag: M = 0.55, Lα = 2.98 and Lβ = 3.15.The presence of the Cu line K = 9.06 is due to the support made of copper on which the sample was placed for analysis.All the peaks mentioned can be seen below in figure 6.
Due to prolonged irradiation, sulfur depletion is found.In the initial stage, the atomic ratio between S and Zn is about 31.42% to 66.58% which means ∼ 0.47 and after an exposure of 4 h totalizing a fluence of 4.6 × 10 17 electrons/cm 2 , the atomic ratio became 13.75% to 86.25%, that is ∼ 0.16.The quantitative differences can be observed in the EDS spectra in the figures 7(a) and (b).The sample for analysis was collected as rigorously as possible from the blackened layer.Quantification rate was done only on sulfur and zinc.This can be correlated with the blackening effect on the powder.

Energy deposition by the electron beam
The total charge on a pulse was measured by integrating the voltage curve for the time produced by the Faraday cup (FC).Acording to table 1. fluence varies between 3.86 × 10 16 electrons/cm 2 to 4.6 × 10 17 electrons/cm 2 and the EB current density per pulse is about 1.5 mA cm −2 as it can be seen in figure 8(a).In the case of our experiment, the surface destroyed by the electron beam that generates the dislocation of sulfur and the appearance of metallized zn can be seen with the naked eye in figure 8(b).
The total energy deposition of the electron beam is given by, Wt = N e * S * E e where N e represent the fluence, S is the irradiated surface (in cm 2 ) and E e is electron beam energy, E e = 2.08 * 10 −15 J (corresponding to 13 keV).
The irradiated mass was taken to be the mass of ZnS:Ag cylinder with the same diameter as the electron beam (5 mm) and the depth of the sample of 2 mm.We did not measure the penetration of the powder of electron beam, but we can infer from the work of an author [39] who irradiated ZnS at energies of 6 keV.He found that the accentuated decomposition of the powder took place in the topomost surface layer at approximately 30 nm and the penetration of the electrons that cause defects in the lattice reached up to 300 nm.

Electroluminescence study
The electroluminescence measurements can be used to examine the defects in semiconductors, both intrinsic and extrinsic.This give insights regarding the impurity and defect energy states even when they are found in low concentrations and consequently is useful to understand defect structures in semiconductors [64].
From previous studies [33,65,67,68] it is known that in ZnS powder there are four types of luminescent point defects: interstitial S, vacancy S, interstitial Zn, vacancy Zn.Based on their energies, the luminescent peaks are inferred to be the interstitial Zn(385 nm), interstitial S(349 nm), vacancy S(427 nm) and vacancy Zn(489 nm).The photoemission of semiconductors is due to the recombination of electrons and holes and is    By inserting Ag into the ZnS crystal lattice, self-activated emission is associated with a complex luminescent center such as VZn-AgZn [33,34,[66][67][68].In our experiments a ZnS:Ag powder with volume of 0,13 cm 3 was exposed to 13 keV electron beam for 30 min, at a fluence of 5.79 × 10 16 electrons/cm 2 .The emission spectra are characterised by broad peaks, arising from allowed transitions between electronic states determined by quantum confinement.
Many studies show that the blue emission spectrum of ZnS:Ag is around 450 nm; however, there are also reports with 460 nm [35] or 470 nm [30], attributed to higher concentrations of defects.
In our case, the broad peak is in the visible light range of 445 to 480 nm, as it can be seen in figure 10(a), centered around 460 nm, with a left peak around 450 nm and a right shoulder with a maximum around 477 nm.It can be attributed to the near band emission.FWHM is almoust constant, around 80 nm.The shape of the emission spectra is preserved throughout all irradiations, without shifting.
Lattice defect or impurities can provide as a recombination center if it is able to get a one type of carrier and further, it can capture the opposite type one, therefore annihilating the pair.The impurities located in deep levels which are not intentionally inserted in the lattice are capable to capture free carriers (electrons in the conduction band and holes in the valence band) produced by excitation during diffusion being in competition with luminescent center, causing them to recombine without photon emission [69].
Another kind of luminescent killer related to impurities is the resonance energy transfer mechanism that removes the energy from the luminescent center [70], and which involves an allowed transition in the sensitizer  and a forbidden transition in the activator without photoconductivity.Also, foreign atoms and molecules can be adsorbed on the surface of the phosphor, which can become luminescence killers and produce a 'dead voltage layer'.However, in the case of our experiment with a vacuum of 10 −4 torr, it is unlikely that such a phenomenon is predominant.
When most of the luminescence centers are excited implying depletion of ground state, the carriers cannot occupy the basic energy level to produce the radiative transition [71,72].This is done when the current density increases.Of major importance in the field emission displays (FEDs) environment, this increase in current density compensates for a sudden drop in energy, but attenuates the brightness.For this reason, it is wanted for a phosphor that has a shorter decay time of activator, than the excitation dwell time [73].Certain authors attribute luminescence saturation to non-radiative Auger recombination involving a donor-acceptor pair in the early stage of decay.[74] In our experiment we constantly monitored current intensity and no significant variations were found; therefore we can consider that this luminescence attenuation mechanism is secondary.Some relatively recent studies [75,76] show that higher concentrations of dopant ions lead to quenching luminescence.A threshold of 3% Ag incorporated in ZnS would be the best option for optical properties, but above this value the intensity decreases due to two possible causes: the sharply generation of Zn vacancy which leads to a massive energy loss through the transmission process and the appearance of Ag 2 S in the sulphide matrix which play a role of non-radiative recombination centers [77].In our experiment, the powder was not subjected to additional Ag enrichment, so this mechanism can be left aside.
There is also a clear correlation between the decrease in luminescence and the increase in temperature [35,74].Some authors [75] attribute the correlation between intensity and temperature to the contribution of two energy loss processes: interaction of excitons with longitudinal acoustic and interaction of excitons with multiple longitudinal optical phonons.A thermoquenching point T 50 , defined as the temperature when the luminescence is halved compared to that measured at room temperature, is found to be, T 50 = 230 °C [39] for a blue emitting ZnS:Ag,Al on 6 KeV and ∼5 μA cm −2 .The non-radiative transition is attributed to depletion of excess energy in the form of phonons in the lattice.In our experiment, if we estimate a total conversion of the deposited energy (table 2) by electron beam into heating temperature using ΔT heat ≈ W t /(m * c), where m is the mass of probe sample and c represents the specific heat of ZnS (c = 4.06 J g −1 K −1 ), a value of ΔT heat ∼ 430 °C (table 2) would be expected for 30 min of irradiation.But measuring the temperature of the powder with a thermocouple we find that only a small fraction of the electron effect contributes to the heating (figure 10(b)).Barely a 2 °C increase was measured during 30 min of exposure.And even if the electron beam was not perfectly aligned with the collector wire of the thermocouple, the signal obtained is much too small to consider the decrease in luminescence due to temperature.
Through this consideration we think the main candidates of the luminescent killer is the increasing of impurities due to irradiation process.
During prolonged bombardment with electrons, ZnS:Ag undergoes a blackening process, called aging.It was investigated in the past [78][79][80] when the authors finding several causes and conditions for this effect: moisture (not our case due to low pressure enviroment), free electrons released from centers through photoconductivity and photochemical decomposition.Although the author [79] assert that vacuum irradiation does not lead to blackening, new data (figure 8(b)) show that this happens.
After a long exposure of a total fluence of 4.6 ×10 17 electrons/cm 2 , realized in session of 30 min of irradiation for 4 h, the luminosity drastically decreased with almoust 90% as it can be seen in figures 11(a) and (b).This can be correlated with the darkening effect.The assertion made by Svitz [79,80] is that through photochemical decomposition the blackening occurs with the formation of a surface dead layer dominated by metallic zinc above the ZnS.
The emission peaks for the unexposed microparticles are similar to the spectra observed by other authors [35,81] who attributes the emissions to Schottky defects and sulfur vacancies.
Also, the blackening can be the effect of the fact that the electrons have much higher energy (3 orders of magnitude) than the binding energy between Zn and S of 6 eV.This can lead to dislocations of the sulphur, with a consequence of the dominance of separated metallized zinc in the upper layer.

Conclusion
The ZnS:Ag powder with a thickness of 2 mm was exposed to a 13 keV electron beam in vacuum conditions of 10 −4 torr.It has a blackening characteristic after a total irradiation with fluence 4.6 x 10 17 electrons/cm 2 .The XRD analysis does not show significant crystallinity changes after irradiation.SEM analysis does not reveal significant modified morphological structures, just only a slight tendency of neighbouring crystallization.The luminescence property of the phosphor decreased by almost 40% after 30 min of irradiation at a fluence of 5.79 ×10 16 .After long exposure on a total fluence of 4.6 ×10 17 electrons/cm 2 , the particles are damaged (at least the layers on the surface) providing just a ∼ 10% luminescence compared to the first use.The quenching luminescence can be attributed to the increase in lattice defects.Also, the aging of ZnS:Ag powder was associated with the darkening of the phosphor which can be explain by sulphur depletion and increasing vacancies and the formation of a metallic zinc dead layer.A depletion of sulphur is confirmed by EDS analysis.
All potential applications in the field of optoelectronics, sensors, life sciences, environmental science, energy and engineering should take into account the aging aspects of ZnS upon prolonged exposure to electron beams and all devices based on this phosphor should be periodically recalibrated.We are very grateful to Professor Emil Pavelescu for his insightful advices and valuable discussion in the framework of PED 734/2022.

Figure 1 .
Figure 1.(a) Sketch of the experimental setup that was placed in the interaction chamber and light emission of the ZNS:Ag powder during irradiation.(b) Blue light emission of ZnS:Ag under the bombardment of electron beam of 13 keV.

Figure 3 .
Figure 3. (a) SEM image of the ZnS: Ag powder (×1000), for a current 25 pA and voltage 20 kV.(b) particle distribution size in the powder.(c) roughness of ZnS: Ag powder after 240 min of 13 keV irradiation.

Figure 5 .
Figure 5. SEM image after 4 h of irradiation:(x5000), at a current 25 pA and voltage20 kV.The encircled areas show adherence between the smaller particles.

Figure 6 .
Figure 6.EDS analysis of the powder.

Figure 7 .
Figure 7. (a) EDS of pristine Zns:Ag evidencing the Zn L and the S K lines; (b) EDS for ZnS:Ag after irradiation for 4 h.
realized through various ways.It can be the direct path of band-band recombination, recombination through shallow traps states or through deep trapped states.In the case of ZnS:Ag powder several path are schematized in figures 9(a) and (b).

Figure 8 .
Figure 8.(a) Current intensity obtaine with a FC; (b) darkening of ZnS:Ag after 4 h of electron beam irradiation.

Figure 10 .
Figure 10.(a) Decreasing luminescence with irradiation time exposure.(b) decreasing intensity with fluence for 30 min of irradiation.

Figure 11 .
Figure 11.(a) Luminescence trend after long exposure; (b) Droping luminescence with fluence and degree of blackening of ZnS:Ag.

Table 1 .
Irradiation time, fluence and energy deposition on long exposure.

Table 2 .
Fluence, energy deposition and espected heating temperature for 30 min of irradiation.