1. Introduction

Semiconductor nanocrystals (SNCs) have drawn much attention for decades because of fascinating light–emitting properties and various potential applications in laser diodes [1], light–emitting diodes [2], photodiodes [3], biological labeling [4], and solar cells [5]. When the size of SNCs is scaled down to a critical value, SNCs begin to exhibit quantum optical and electrical properties, i.e., the emission color varies regularly with the size of SNCs [6, 7]. Such size–scalable light–emitting SNCs are known as semiconductor quantum dots. It has been clarified that the size–dependent emission arises from variation of the band gap with nanocrystal size. Band–gap engineering constitutes one of the most important techniques to adjust light–emitting properties for SNCs. However, the quantum effects also restrain the tunability of light emission colors since these SNCs cannot be scaled down unlimitedly and cannot exceed a critical value. It is thus indispensable to tune the chemical composition of semiconductor quantum dots in order to develop light–emitting SNCs over a broader wavelength region. One typical example is semiconductor alloys that are ternary systems [8], allowing for the adjustment of the band gap by rationally varying the ratio of the dual cations or anions. In addition to the band gap engineering technique, artificially introducing optical defects serves as another efficient route to tune optical properties of SNCs [9]. For instance, by varying surface oxygen vacancy defects, ZnO nanocrystals are able to exhibit tunable emission colors across a wide wavelength range [10]. These defects–induced light–emitting SNCs have found important applications in displays and bio–labeling.

Ga₂O₃ is a well–known transparent conducting oxide with a wide band gap (E_g ~4.9eV), exhibiting high conductivity and intense luminescence [11]. A lot of efforts have been devoted to development of low–cost light–emitting diodes [12], biosensors [13], and catalysts using Ga₂O₃ [14]. Ga₂O₃ is comprised of five phases, i.e., α, β, γ, δ, and ε phases, among which β–Ga₂O₃ is the most stable phase. Apart from the band–edge luminescence in ultraviolet wavelength region [15], the luminescence of Ga₂O₃ is mostly associated with internal defects; as a result, blue [16], blue–green [17], red (cryogenic temperature) emissions have been observed [18]. Introducing multiple defects into Ga₂O₃ and adjusting relative emission intensities of an individual defect has led to white–light emission [19]. Recently, size–tunable phosphorescence was demonstrated in colloidal meta–stable γ–Ga₂O₃ nanocrystals due to size–tunable donor–acceptor pair recombination rate [20]. However, the wavelength tunability is very restrained within 50 nm for these γ–Ga₂O₃ nanocrystals.

In this work, we developed a new class of optical nanocomposites using mesoporous silica particles (MSPs) as platforms to load β–Ga₂O₃ SNCs, and demonstrate novel multi–color light emissions from such materials. In the past decades, MSPs have attracted much attention due to potential applications in biomedicine [21]. MSPs have large surface area and pore volume, and thus can serve as highly robust and tunable delivery platforms for various therapeutic agents. Recently, MSPs have been used to load optical species such as rare–earth ions [22], laser dyes [23], magnetic nanocrystals [24], and metal nanoparticles [25], and demonstrated potential applications in the control of spontaneous emission, lasing, and nonlinearity. The emission spectrum of the nanocomposites developed in this work exhibits interesting features; namely, several narrow peaks are imposed on a broad emission band and the dominant peak regularly shifts by varying the excitation wavelength. As a result, the emission color is tunable over a broad wavelength region ranging from blue, green, yellow, to white. The ease in fabrication and intriguing luminescent properties of these materials make them particularly attractive as light sources.

2. Experimental

MSPs were prepared through a surfactant–mediated method using 16–alkyltrimethylammoium bromide as surfactant and sodium hydroxide as catalyst [26]. MSPs embedded with Ga₂O₃ nanocrystals (denoted as MSGNs hereafter) were fabricated by dispersing MSPs into an ethanol solution of gallium trichloride (GaCl₃), accompanied by baking in ambient atmosphere and subsequent thermal annealing at reducing atmosphere. In a typical procedure, 17.6 mg of GaCl₃ was dissolved in 10 ml of ethanol, followed by addition of 60 mg of MSPs and heated in a water bath (60 °C) for 12 h under vigorous magnetic stirring. The resultant particles were washed three times with water and then baked at 60 °C in a constant oven. The baked particles were firstly annealed in ambient atmosphere at 800 °C for 5 h and then in Ar/H₂ (95%/5%) for 3 h. The photoluminescence was measured in a Hitachi 850 fluorescence spectrophotometer. The x–ray diffraction (XRD) pattern of the powders was recorded on a Rigaku Rint 2500 x–ray diffractometer with Cu Kα radiation (λ = 1.5406Å). The distribution of size and morphology of particles was checked by a JEOL JSM–6700F field–emission scanning electron microscope (FE–SEM). The transmission electron microscope (TEM) observation was performed on a JEOL JEM–2100 TEM with a 200 kV accelerating voltage.

3. Results and discussion

Figures 1(a) and (b) depict the FE–SEM images of mesoporous silica particles before and after impregnation with Ga₂O₃ nanocrystals, respectively. The results indicate that the nanoparticles are monodispersed and free of large clusters, and there is no obvious difference between MSPs and MSGNs. The TEM images show the presence of nanocrystals with size ranging from 3 to 5 nm, as displayed in Fig. 1(c). It is noteworthy to mention that, in comparison to the conventional hydrothermal process [27, 28], the use of mesoporous silica templates gives rather small nanocrystals due to the spatial confinement of pore size. Figure 1(d) shows the X–ray diffraction (XRD) pattern of MSPs (black curve) and MSGNs (red curve), respectively. The diffraction peaks observed in MSGNs can be indexed to amonoclinic structure, well consistent with the reported data of β–Ga₂O₃ (JCPDS Card No. 43–1012).

 

Fig. 1 SEM images of MSPs (a) and MSGNs (b). (c) TEM image of MSGNs, the inset is a high–magnification image. (d) XRD pattern of MSPs (black) and MSGNs (red).

Download Full Size | PDF

MSGNs obtained by annealing in ambient atmosphere exhibits a broad blue emission band centered at λ_em = 450 nm when excited at λ_ex = 250 nm, as presented in Fig. 2(a).The emission intensity of this blue emission band depends on the ratio of Ga to Si, reaching maximum when Ga:Si is 1:1, as shown in Fig. 2(b). In contrast, the annealing in reducing atmosphere resulted in distinct emission profiles, as depicted in Fig. 2(c). For instance, the emission spectrum recorded at an excitation wavelength of λ_ex = 310 nm is comprised of several narrow sub–bands among which the dominant peaks are centered at λ_em = 438, 470, 490, and 525 nm [Fig. 2(c), orange curve]. The emission spectrum can be decomposed into seven Gaussian contributions, locating at λ_em = 385, 402, 418, 438, 470, 490, and 525 nm [Fig. 2(c), gray curves], respectively. The summary of these fit curves [Fig. 2(c), black curve] agrees well the experimental one. These emission peaks correspond to different excitation profiles [Fig. 2c, red, green, blue, and magenta curves], suggesting that the emission profile can be readily adjusted by varying the excitation wavelength. To elucidate this capability, we excited the sample at different wavelength and recorded the emission profile. The results show that the emission profile regularly shifts with the excitation wavelength [Fig. 2(d)]. The spectral positions of the sub emission bands are maintained upon variation of excitation wavelength, but the dominant peak shifts. The dominant peak is located at λ_em = 385 nm for excitation at λ_ex = 250 nm, and shifts to λ_em = 550 nm when excited at λ_ex = 400 nm. Consequently, the emission color is changed with the excitation, as illustrated in Fig. 2(d).

 

Fig. 2 (a) Normalized excitation (black) and emission (red) spectra of MSGNs obtained by annealing in ambient atmosphere. The excitation and emission maxima are located at 250 and 450 nm, respectively. (b) Evolution of the emission spectrum with the molar ratio of Ga to Si for MSGNs obtained by annealing in ambient atmosphere. (c) The emission spectrum (purple) of MSGNs obtained by annealing in reducing atmosphere when excited at λ_ex = 310 nm as well as the excitation spectra when the emission is monitored at λ_em = 438 (red), 470 (green), 490 (blue), and 525 nm (cyan). The emission spectrum is de–convoluted into Gaussian contributions (gray curves). The black curve is the summary of the Gaussian contributions. (d) Evolution of the emission profile when the excitation wavelength ranges from 250 to 400 nm for MSGNs obtained by annealing in reducing atmosphere.

Download Full Size | PDF

Optical emissions centered at different wavelengths have been observed in various Ga₂O₃ nanostructures, and the emission spectral profile largely depends on the synthesis procedure and particle morphology [16–20]. The optical emission is mostly related to internal defects, except that the UV emission is assigned to a band–edge transition. The most well–known defects–related emission in Ga₂O₃ nanostructures occurs at the blue wavelength region, originating from recombination of an electron of a donor formed by oxygen vacancies and a hole on an acceptor formed by gallium vacancies [16]. The blue emission observed in MSGNs obtained by annealing in ambient atmosphere can be well described with this mechanism. This emission can be shifted to the green wavelength region due to difference in fabrication processes. Such emission usually features a single broad band, whereas in specific nanostructures multiple bands have been observed, especially for Ga₂O₃ nanorods and nanowires [15, 19]. For instance, β–Ga₂O₃ nanowires grown at low temperature from a single–source organometallic precursor exhibited several sub emission bands located at 380, 420, 454, 485, and 509 nm [15]. Furthermore, incorporation of nitrogen–related defects led to the appearance of a red emission that contributes to the observation of white–light emission along with the normal blue–green light [19]. The observation of multiple emission peaks implies the existence of multiple energy levels, and it has been suggested that such a phenomenon is associated with donor levels created by oxygen vacancy as well as multiple acceptor levels induced by gallium vacancy or gallium oxide vacancy [19]. Such a multi–level energy scheme is supposed to account for the observation of multiple emission peaks in MSGNs obtained by annealed in reducing atmosphere, as depicted in Fig. 3.

 

Fig. 3 Schematic of multi‒color light emissions in MSGNs. The donor levels are supposed to be created by oxygen vacancies while multiple acceptor levels are induced by gallium vacancies or gallium oxide vacancies, accounting for the observation of multiple peaks in the fluorescence spectrum.

Download Full Size | PDF

To quantitatively characterize the emission colors, we plotted the Commussion Internationale de l'Eclairage (CIE) chromaticity coordinates for MSGNs, as shown in Fig. 4.The (X, Y) coordinates are located in the white light region, ranging from (0.28, 0.35) to (0.24, 0.25), when the sample is excited at high energy from λ_ex = 250 to 290 nm. When the excitation is shifted to the longer wavelength, the (X, Y) coordinates vary from (0.21, 0.21) at λ_ex = 300 nm to (0.33, 0.43) at λ_ex = 400 nm. Correspondingly, the emission color is continuously tunable from blue to yellowish green region. The inset of Fig. 4 illustrates several images of MSGNs for excitations at different wavelength. This continuous tunability of emission colors makes MSGNs particularly interesting as potential light sources.

 

Fig. 4 CIE coordinates of MSGNs. The coordinates are calculated from the emission spectra shown in Fig. 2(c). The insert shows the images of particles excited at different wavelengths. The excitation wavelength corresponds to the CIE coordinates are shown in the right.

Download Full Size | PDF

We have examined the optical properties of bare MSPs, and found no fluorescence therein. Therefore, the above–mentioned emission properties are undoubtedly related to the presence of Ga₂O₃ nanocrystals. In addition, the observation of the above–mentioned emission features is closely related to the reducing atmosphere. Therefore, we can conclude that the reducing atmosphere benefits in formation of multiple defects on Ga₂O₃ nanocrystals. Moreover, the porous channels of MSPs and high specific surface area of Ga₂O₃ nanocrystals facilitate the formation of defects.

4. Conclusions

To conclude, we have fabricated mesoporous silica particles embedded with β–Ga₂O₃ nanocrystals and demonstrated for the first time multi–color light emissions in such materials. The fluorescence spectrum is composed of several narrow sub–peaks where the dominant one is dependent on the excitation wavelength. Consequently, we achieved the tunability of fluorescence over a broad wavelength regime, i.e., the emission color ranges from blue, green, to white light, by simply varying the excitation wavelength. We suggest that donor levels created by oxygen vacancy and multiple acceptor levels induced by gallium vacancy or gallium oxide vacancy account for the observation of multiple emission bands. To further clarify the mechanism responsible for such unusual optical emissions, additional experiments such as low‒temperature emission properties and fluorescence lifetime should be conducted. These studies are well oriented for future research, which is currently underway.