1. Introduction

Mechanoluminescent materials refer to a kind of materials that can emit luminescence upon excitation of external mechanical stimulus, including pressing, squeezing, bending, twisting, rubbing, and so forth [1–4]. Different from other types of luminescence like photo-luminescence and electroluminescence, mechanoluminescence (ML) owns unique advantages such as energy-saving. For example, Jeong et al. introduced a ML composite by embedding zinc sulfide microscopic particles into the polydimethylsiloxane (PDMS) host [5]. Depending on this ML composite, they proved an environmentally friendly wind-driven ML display, opening the possibility of converting the natural abundant energy like the wind to another type of energy such as luminescence that benefits mankind [6–8]. Very recently, Hou et al. prepared a novel interactive mouthguard integrated by several distributed-optical-fiber sensors where ZnS:Cu²⁺/Mn²⁺/Cu⁺ were embedded [9]. Upon stimulation of bite, the ML with different colors from ZnS:Cu²⁺/Mn²⁺/Cu⁺ could be detected. And by combination with machine learning, it is possible to use interactive mouthguards to achieve the operation of computers, smartphones, and even wheelchairs. Considering the potential applications in a broad spectrum of fields, ML has attracted ever increasing attention nowadays [9–15].

According to the documentary records, ML has a very long history of over four hundred years [16]. However, it suffers from a very slow development during the initial period, which is likely to be ascribed to the backward technique, the lack of knowledge, and many others. In 1999, Xu’s group developed several very significant ML materials, including ZnS:Mn²⁺ and SrAl₂O₄:Eu²⁺ [17,18]. Different from previous discoveries, these two materials show excellent reproducible ML, which is a huge improvement. For this reason, it is generally considered that the ML field enters a new era. So far, many kinds of materials that are capable of emitting ML have been designed [19–26]. Nowadays the focus has been mainly put on the following two kinds of mechanoluminescent materials, i.e., self-charged and trap-controlled materials. For the former material, no pre-excitation is needed to trigger the ML. CaZnOS doped with different lanthanides and Mn²⁺ are typical examples [27,28]. They can continue to emit ML under the excitation of an external force without pre-excitation. Unfortunately, limited self-charged mechanoluminescent materials have been reported. By comparison, there are various kinds of trap-controlled materials, as evidenced by the excellent review authored by Zhang et al. [29]. For these materials, pre-light-excitation is necessary for the generation of ML. A typical characteristic of this type of material is that ML will gradually decrease along with the exposure to external force. Realization of these two different kinds of ML mechanisms in one system or host is very interesting and meaningful, which is, however, still challenging. Moreover, the ML wavelength is also of vital importance in practical applications. Nowadays, the ML wavelength has covered the 310∼1700 nm range [30]. Developing ML with a wavelength shorter than 300 nm is very meaningful as it will enable a high signal-to-noise ratio and even background-free ML collections. It is, however, rarely reported, to the best of our knowledge.

In this work, we introduce a new ML material, namely MgF₂:Tm³⁺. Upon stimulation of external force, MgF₂:Tm³⁺ emits bright ML over a wide wavelength range, without the need of pre-excitation. Surprisingly, the ³H₄→³H₆ transition of Tm³⁺ even covers the ultraviolet-C band in the 280-300 nm range with a central peak at 292 nm. It has been further confirmed by the solar blind camera. Moreover, the NIR ML, ascribed to the ³H₄→³H₆ transition of Tm³⁺ peaked at around 800 nm, has also been collected. In addition, after exposure to X-rays, the as-prepared MgF₂:Tm³⁺ presents a much stronger ML than the uncharged sample. After cleaning up at high temperatures to empty the traps, the ML of the same MgF₂:Tm³⁺ returns to the initial level. With the aid of a thermoluminescence test, our material is confirmed to be trap-controlled. It means that both self-powered ML and trap-controlled ML can be achieved in MgF₂:Tm³⁺. In the end, the elastic film is successfully fabricated by incorporating the as-prepared MgF₂:Tm³⁺ into the PDMS, and its ML property is disclosed. The potential application of ML photons is also presented and discussed.

2. Materials and methods

Synthesis of MgF₂:Tm³⁺ and related ML film. All samples were prepared by high-temperature solid-state method. The raw materials are MgF₂ (99.99%, Aladdin) and Tm₂O₃ (99.99%, Aladdin). According to the stoichiometric ratio, the raw materials were weighed and mixed fully. Subsequently, these mixtures were sintered at 1200 °C for 3 h under the N₂/H₂ atmosphere. After cooling down, these samples were ground again and stored carefully for the following characterization and test. For convenient ML measurement, some of the samples was sealed in card protection film (PET, Deli, No. 3819). To prepare the flexible ML film, the as-prepared MgF₂:Tm³⁺ powders were stirred with polydimethylsiloxane (PDMS) in a ratio of 1:1, which were then poured into a mold and dried at 70 °C for 2 h.

Characterization of MgF₂:Tm³⁺. The crystal structure of samples was characterized by using an X-ray diffractometer (XRD, Bruker D8 Discover) with a Cu Kα radiation source (1.5418 Å; cathode voltage: 40 kV; current: 40 mA; scanning range/2θ: 20∼80°). X-ray Photoelectron Spectroscopy (XPS, ESCALAB 250Xi) was used to explore the surface elemental composition and chemical state of samples. The morphology analysis and elemental mapping of MgF₂:Tm³⁺ were performed using scanning electron microscopy (SEM, FEI Nova Nano SEM450).

Optical test of MgF₂:Tm³⁺. The fiber optic spectrometer (Ocean Insight Maya 2000pro) was used to collect the ML spectrum of samples. The transmission of plastic ML film was measured by the UV-NIR spectrometer (U4100). All images obtained during the experiments were captured using a commercial SLR digital camera (EOS 5D Mark III, Canon). The UVC emission of the phosphors was recorded with a solar blind camera (HL4V ASB Mini18 ESY, Suzhou Weiner Optoelectronics Technology Co., Ltd., China).

3. Results and discussion

As presented in Fig. 1(a), all samples of Mg_1-xF₂:xTm³⁺ (x = 0, 0.1%, 0.5%, 1.0%, and 1.5%) present the nearly same XRD patterns. They are consistent with the reference ones (PDF#72-1150) that belong to the tetragonal system. What’s more, there are no other redundant diffraction peaks. All these results support the conclusion that our materials have been successfully prepared with the single rutile-type structure. In this structure, there is only one Mg²⁺ site. As shown in Fig. 1(b), each Mg²⁺ site is surrounded by six [F^-] ligands to form an octahedron. Moreover, each F^- site is surrounded by three Mg²⁺ sites. Combined with a similar ionic radius between Tm³⁺ ion and Mg²⁺ ion, the Tm³⁺ ions are expected to occupy the Mg²⁺ sites.

 

Fig. 1. Characterization of MgF₂:xTm³⁺. (a) XRD patterns of MgF₂:xTm³⁺ (x = 0, 0.1%, 0.5%, 1.0%, and 1.5%); (b) crystal structure of MgF₂; (c), (d) XPS spectra of the representative MgF₂:0.5%Tm³⁺; (e) SEM images and elemental mapping images of Mg, F, and Tm of the representative MgF₂:0.5%Tm³⁺.

Download Full Size | PDF

Moreover, the surface of the samples was analyzed by XPS, and the results are shown in Fig. 1(c) and 1(d). The characteristic binding energies that belong to the Mg, F, and Tm elements’ orbit can be observed. For the partially enlarged XPS result shown in Fig. 1(d), a peak with central binding energy at about 177.6 eV is clearly collected. This is ascribed to the Tm4d orbit, indicating the presence of Tm³⁺ ions in samples [31]. The morphology and element distribution of the representative MgF₂:0.5%Tm³⁺ were also analyzed by SEM and EDS, and the results are presented in Fig. 1(e). As can be seen from this figure, these particles are homogeneous with a size of ∼28 µm. According to the EDS images, the Mg, F, and Tm elements were uniformly distributed over a single MgF₂:0.5%Tm³⁺ particle.

The ML properties of Mg_1-xF₂:xTm³⁺ (x = 0.1%, 0.5%,1.0%, and 1.5%) are then studied in detail. The top panel in Fig. 2(a) schematically shows the experimental setup used for collecting the ML spectra of samples. The left part is the mechanical control unit that can adjust the pressure exerted on the samples. For ease of optical test, the powder samples are fixed in form of plastic film, as presented in the middle part in Fig. 2(a). We also measured the transmittance of plastic film used to fix the samples. As clearly shown in the bottom panel of Fig. 2(a), the film is penetrable to photons longer than 310 nm.

 

Fig. 2. ML properties of the prepared materials. (a) Experimental schematic of ML spectral test. (i) Schematic diagram of the experimental setup used to measure the ML of samples in form of plastic films. (ii) Photograph of the prepared ML plastic film. (iii) Transmittance of the plastic film. (b) ML spectra of MgF₂:xTm³⁺ (x = 0.1%, 0.5%, 1.0%, and 1.5%) under excitation of external force at 30 N; (c) ML intensity of the different emission bands of Tm³⁺ as a function of doping concentration; (d) CIE chromaticity coordinates for ML of MgF₂:xTm³⁺ (x = 0.1%, 0.5%, 1.0%, and 1.5%); (e) image of MgF₂:0.5%Tm³⁺ upon grinding (exposure time: 5 s). (f) ML spectra of MgF₂:0.5%Tm³⁺ under excitation of different forces. (g) ML intensity of MgF₂:0.5%Tm³⁺ as a function of external force.

Download Full Size | PDF

As is well known, dopant concentration has a significant influence on the luminescence intensity for nearly all optical materials. We thus first examine this effect. As presented in Fig. 2(b), all samples could emit ML upon the same external mechanical force at 30 N. And the ML spectral profile of all these samples keep unchanged. Over the 300∼850 nm wavelength range, there are mainly three ML lines, with central emission peaks at 350, 457, and 790 nm. According to the reference, these three lines are mainly attributed to the ¹D₂→³H₆, ¹D₂→³F₄, and ³H₄→³H₆ transitions of Tm³⁺, respectively. The integrated intensities of these three ML lines are then plotted as a function of the doping concentration of Tm³⁺. As presented in Fig. 2(c), the optimal content of Tm³⁺ is 0.5% in molar ratio. Depending on the ML spectra of Mg_1-xF₂:xTm³⁺ (x = 0.1%, 0.5%,1.0%, and 1.5%), the Commission International del’Eclairage (CIE) chromaticity coordinates have also been calculated for all phosphors. As depicted in Fig. 2(d), the coordinates of all samples are located at the junction between the purple and blue areas. This is consistent with the color of the ML image presented in Fig. 2(e) that was captured when grinding the samples. Furthermore, we also investigated the influence of external force on ML intensity. As shown in Fig. 2(f) and Fig. 2(g), the ML intensity increases gradually and linearly with the increase of external force exerted on the samples. This straightforward linear function between ML intensity and imposed external force indicates the possibility of our materials for stress sensing in the future.

As mentioned above, the powder samples are fixed in form of plastic film for ease of optical testing. However, the plastic film is only penetrable to photons longer than 310 nm. It means that the ML light with a shorter wavelength might be overlooked undesignedly, thus probably losing much interesting and useful information. With this in mind, we designed a new ML test system in which a quartz tube was used to wrap the fiber-optics probe, as shown in Fig. 3(a)(i). The quartz tube used in our experiment has excellent transmittance over the whole wavelength range from 200 to 800 nm. By using this ML test system, the ultraviolet emission has been confirmed with the aid of a solar blind camera by grinding the powders directly in the mortar, as presented in Fig. 3(a)(ii). The collected ML spectrum together with the PL spectrum is displayed in Fig. 3(a)(iii). The ML band in the 280-302 nm spectral range with central peak at 292 nm is attributed to the ¹I₆→³H₆ transition of Tm³⁺ [31]. It, however, should be noted that there is a slight difference between the ML and PL spectra. Therefore, it is an important task to construct a theoretical explanation in the future.

 

Fig. 3. ML enhancement and related mechanism. (a) Experimental schematic of ML spectral test, (i) transmittance of plastic film and quartz tube, (ii) schematic diagram of ML experimental setup for quartz tube grinding test, and (iii) comparison of PL and ML spectra of MgF₂:0.5%Tm³⁺. (b) ML enhancement and related mechanism, (i) ML spectra under different conditions, (ii) PersL spectra of MgF₂:0.5%Tm³⁺. Inset shows the PersL image of sample after irradiation (exposure time: 2 s). (iii) TL curves of the irradiated sample under different conditions. Black line: pristine without any pretreatment; Pinkish purple line: after being charged by X-ray; Red line: after continuous ML measurement for 10 minutes.

Download Full Size | PDF

Note that the ML spectra discussed above were all obtained without the need for pre-excitation. It thus can be classified into self-charged ML. On this basis, it is further confirmed that our material also has the feature of trap-controlled ML. Note that no TL signal can be collected for pristine sample, as proved by the black line in Fig. 3(b)(iii). We first collected the ML spectrum of pristine MgF₂:0.5%Tm³⁺ under external force stimulation at 30 N (green line in Fig. 3(b)(i)). Thereafter, this material was exposed to X-ray for 5 minutes. After stoppage of X-ray excitation, MgF₂:0.5%Tm³⁺ shows an evident but very fast persistent luminescence, as presented in Fig. 3(b)(ii). Half an hour later, the ML spectrum of MgF₂:0.5%Tm³⁺ was again collected under the same external force stimulation at 30 N (purple line in Fig. 3(b)(i)). Surprisingly, the ML improved compared with the pristine counterpart. It suggests that in addition to the self-charged property, our material also shows the feature of trap-controlled ML. X-ray exposure is likely to charge the sample and the external force induces the release of ML. To confirm this point, we then performed the thermoluminescence test by monitoring the strongest ML line at 457 nm. As presented in Fig. 3(b)(iii), the luminescence signal could be collected over the temperature range of 303-773 K. It indicates the presence of a continuous distribution of traps for the irradiated MgF₂:0.5%Tm³⁺. After continuous ML measurement for 10 minutes, we again tested the TL curve, as denoted by the red line in Fig. 3(b)(iii). Clearly, most of the trapped carriers have been released. After cleaning up the traps, the sample could emit ML upon the same external force stimulation at 30 N (cyan line in Fig. 3(b)(i)). It is interesting to find that the ML recovered to the pristine state. All these results indicate that pre-exposure to X-ray plays a key role in charging the samples, or in other words, creating more traps in materials. And the force stimulation can induce the release of ML. It should also be emphasized that other excitation light sources had also been used to charge the sample (see Figs. S1 and S2 in Supplement 1). However, no ML enhancement was observed for all conditions.

Depending on the experimental results and existing knowledge, the following plausible explanation is discussed here. For one thing, our material can emit ML in a self-powered way without any pre-excitation. When the luminescent centers Tm³⁺ ions substitute the cationic sites Mg²⁺, there must be some lattice expansion or distortion, accompanied by the formation of defects, for instance, cation vacancy or interstitial anion to maintain charge balance. These defects serve as trap centers to attract electrons or holes. Under stimulation of external force, slight distortion of the crystal lattice may occur, probably leading to the movement of electrons and holes. Subsequently, the electrons and holes recombine and transfer the corresponding recombination energy to the neighboring Tm³⁺ ions, followed by the ML emission from Tm³⁺. Upon excitation of X-rays, more electrons and holes occur and can be captured by the defects. Under stimulation of external force, the electrons trapped in the defects such as fluorine vacancy might be released and captured by the Tm³⁺ sites that have attracted holes. The electron-hole recombines and transfers the corresponding energy to the neighboring Tm³⁺ ions, followed by the ML emission from Tm³⁺. Therefore, the ML intensity has been enhanced on the basis of original self-powered ML.

Finally, we demonstrate the potential applications of the as-prepared MgF₂:0.5%Tm³⁺. The prepared MgF₂:Tm³⁺ powders were put into the Chinese character pattern, which was then sealed by plastic film. As presented at the bottom in Fig. 4(a), the Chinese character pattern shows the bright blue-violet color under stimulation of external force. A ML film was then prepared by embedding MgF₂:0.5%Tm³⁺ into the PDMS host. As presented in the left part of Fig. 4(b), the ML film is semitransparent. When stretching the ML film, it emits blue-colored photons. With this in mind, the ML film was then prepared as the seal of a box. As shown in the right part of Fig. 4(b), when tearing the ML film on the box, the blue-violet luminescence could be observed clearly. It suggests that our materials are promising candidates for anti-counterfeiting applications.

 

Fig. 4. Applications of MgF₂:0.5%Tm³⁺. (a) The pattern of Chinese characters made up of MgF₂:0.5%Tm³⁺ under natural light (top) and external force stimulation (bottom, exposure time: 10 s). (b) Photograph of the transparent ML film made up of MgF₂:Tm³⁺ and PDMS under natural light (left top) and under stretching (left bottom, exposure time: 5 s); diagram of a box sealed by the transparent ML film (top right); photograph taken during tearing the film (bottom right, exposure time: 5 s). (c) Images collected by a solar blind camera during stretching the ML film (see Visualization 1).

Download Full Size | PDF

According to the fore-mentioned analysis, our materials can emit the ML located in the UVC/B region. It has a unique superiority. Due to the absorption by the ozone layer, the UVC part and most of the UCB part of the sunlight cannot reach the earth’s surface. It means that a high signal-to-noise ratio or even background-free detection can be expected by using the UVC/B photons. We then demonstrate this unique superiority. As depicted in Fig. 4(c), the ML signal, during stretching of the ML film, could be identified by a solar blind camera, with a high signal-to-noise ratio. As increasing the stretching force, the ML signal becomes stronger and reaches the maximum at the point of film breakage. These results reveal that our materials have the potential for dynamic stress diagnosis of constructions like bridges and buildings.

4. Conclusions

In summary, we successfully introduce a new ML material, namely MgF₂:Tm³⁺. Upon stimulation of external force, MgF₂:Tm³⁺ emits bright ML over a wide wavelength range from 280 to 850 nm, without the need for pre-excitation. The ¹I₆→³H₆ transition of Tm³⁺, located in the UVC/B region, has also been collected and confirmed by the solar blind camera. In addition, we demonstrate that MgF₂:Tm³⁺ is a type of two-in-one ML material. In addition to the self-recovered ML, MgF₂:Tm³⁺ also presents the property of trap-controlled ML. Specifically, the as-prepared MgF₂:Tm³⁺ presents a much stronger ML than the uncharged sample after exposure to X-rays. After cleaning up at high temperatures to empty the traps, the ML of the same MgF₂:Tm³⁺ returns to the initial level. It means that both self-powered ML and trap-controlled ML can be achieved in MgF₂:Tm³⁺. In the end, the elastic film is successfully fabricated by incorporating the as-prepared MgF₂:Tm³⁺ into the PDMS host, and its ML property is disclosed. The potential application of ML photons, especially over the UVC/B region, is also presented and discussed.