The optical properties of the two monochromic QDs were firstly studied. Fig. 1a and b show the photoluminescence (PL) and absorption spectra of the R-QDs and O-QDs. The FWHM of the R-QDs and O-QDs is about 20.6 and 43 nm, respectively. The positions of dotted lines indicate the PL and absorption peaks. As shown in TEM images, the R-QDs and O-QDs exhibit a cubic morphology with an average size of 13 nm and 12 nm (Fig. 1c and d), respectively. The inset HRTEM images show an interplanar distance of 0.35 nm, which can be assigned to the (111) plane of the cubic phase ZnS.
The optical properties of QDs-silicone thin films made of the monochromic R-QDs and O-QDs with different concentrations are further tested under excitation by a 365 nm laser at 15.88 mW/cm2. Fig. 2a and b show the concentration dependent PL spectra of the QDs and their FWHM is almost constant. Fig. 2c and d exhibit the PL intensity and absolute QE of monochromic QDs-silicone thin films with different concentrations of QDs. With the rise of concentration, the PL intensity of R-QDs silicone thin films increases until the QD concentration reaches 2 mg/mL and then decreases because of concentration quenching. Similar to the variation in PL intensity, the QE of the QDs reaches the highest value of about 85% at the same concentration. The PL intensity and QE of the O-QDs silicone thin films present a similar concentration dependent trend compared to those of the R-QDs based thin films. Differently, the PL intensity and QE of the O-QDs rise quickly till a QD concentration of 4 mg/mL and the maximum values are obtained at the concentration of 10 mg/mL. We infer that it is attributed to the larger Stoke’s shift of the O-QDs than the R-QDs. The maximum QE of the O-QDs silicone thin film is about 76%, which is at the same QD concentration for the highest PL intensity.
Moreover, the 2 mg/mL R-QDs and 10 mg/mL O-QDs based silicone gel films also exhibit the highest PL intensity under different exciting power, as shown in Fig. S1 a and b, respectively. At the above two concentrations, the optical properties of the monochromic QDs are effectively preserved, which weakens the PL quenching caused by the host matrix effect [28,29]. The study helps to find the suitable concentration of the monochromic QDs in silicone film.
To further investigate the concentration influence of QDs in the monochromic QDs based silicone thin films, the time-resolved PL (TRPL) spectra of the thin films with different concentrations are measured and the decay curves are depicted in Figure 3. It is known that PL decay curves can be expressed with a multi-exponential function, as illustrated by equation 1 ,
[Please see the supplementary files section to view the equation.] (1)
where I(t) is the PL intensity at time t. Ai and τi represent the relative amplitude and the excited state lifetime of each exponential component of PL decay, n is the number of decay times. These decay curves, as shown in Fig. 3a and b, can be well fitted by a double exponential function according to equation 1.
The fitting parameters Ai and τi are listed in Table S1 and S2. The amplitude-weighted lifetimes of the R-QDs and O-QDs are selected as their lifetimes (τave) for further investigation. The lifetime can be calculated from the following equation 2  and is listed in Table S1 and S2.
[Please see the supplementary files section to view the equation.] (2)
Fig. 3c and d show the lifetimes of the two monochromic QDs under different concentrations. Both lifetimes increase with the rise of concentrations and the rising rate becomes slower after 1 mg/mL for the R-QDs and 2 mg/mL for the O-QDs, respectively. It indicates that the rise of concentration reduces the distance amongst QDs thus enhances the energy transfer and self-absorption in the monochromic QDs [32,33]. Meanwhile, the increment of the lifetime in O-QDs is more obvious than that in R-QDs, suggesting more energy transfer in O-QDs. However, it appears that the energy transfer does not induce fluorescent quenching of the QDs at low concentration. On the contrary, it may have positive effect on the PL intensity and QEs, as previous shown in Fig. 2.
The optical properties of composite-QDs with different weight ratios of R-QDs:O-QDs were further studied. The PL spectra of the composite-QDs thin films are shown in Fig. 4a. Based on the spectra, the composite-QDs PL peak intensity ratio of 631:605 (nm) is extracted in Fig. 4b. The peak intensity ratio presents a rising increment with the R-QDs percentage, which suggests the energy transfer from O-QDs to R-QDs. Fig. 4c exhibits the overlap between the R-QDs absorption spectrum and the O-QDs emission spectrum. It suggests a high chance of FRET process, in which the O-QDs acts as the donor and R-QDs as the acceptor (shown in Fig. 4d).
Further study focuses on the FRET process in composite-QDs. Fig. 5a presents the effect of R-QDs (acceptor) on the emission kinetics of the O-QDs (donor). The TRPL intensity decreases with the rise of acceptor in the ﬁlm sample (analyzed at peak donor emission wavelength 605nm). Fig. 5b presents the effect of O-QDs (donor) on the emission kinetics of the R-QDs (acceptor). On the contrary, the TRPL intensity increases with the rise of the donor in the ﬁlm sample (analyzed at peak acceptor emission wavelength 631nm). The decay curves in Fig. 5a and b can be fitted with the 2 exponentials, and the detailed amplitudes, lifetime components, and amplitude-weighted lifetimes of the QDs were listed in Table S3. The lifetime of the O-QDs sample was found to be 30.25 ns. When the acceptor R-QDs are introduced, the lifetime of the donor O-QDs decreases (Table S3) owing to the intervene of the energy-transferring channel. The lifetime of the donor becomes shorter with the rise of the acceptor concentration. On the contrary, the lifetime of the R-QDs sample was found to be 13.08 ns. When the donor O-QDs are introduced, the acceptor R-QDs present an increase in lifetime as a result of energy feeding (Table S3) . The calculated results are shown in Fig. 5c, which clearly demonstrates the phenomena.
The FRET process is also investigated by energy transfer efficiency. The efficiency of FRET can be calculated according to lifetime as illustrated by equation 3.
[Please see the supplementary files section to view the equation.] (3)
Where τDA is the donor fluorescence lifetime in the presence of the acceptor, τD is the donor fluorescence lifetime in the absence of acceptor . It shows that τDA is inversely proportional to the energy transfer efficiency. Therefore, as the acceptor-donor ratio increases, the τDA becomes shorter and the energy transfer efficiency increases. Larger energy transfer efficiency reflects higher impact on fluorescence. We further analyze the FRET efficiency of composite-QDs in Fig. 5a. The calculated results are listed in Table 1 and the efficiency reaches 33.2% with the highest proportion of acceptor. Meanwhile, Fig. 5d exhibits the change of FRET efficiency under different ratios of donor-to-acceptor. The FRET efficiency increases with the rise of R-QDs (acceptor) in the composite-QDs and the increment rate of the efficiency is close to that of the R-QDs. It indicates that the increment of energy transfer is sensitive to the increment of the acceptor.
[Please see the supplementary files section to view the table.]
As the best continuity spectrum for LED lighting in orange-red light, the composite-QDs with the 1:10 weight ratio of R-QDs:O-QDs is selected for further study. Fig. 6a exhibits the PL spectra of the composite-QDs silicone thin films with different concentrations of composite-QDs at the same weight ratio of R-QDs:O-QDs (R:O=1:10). Besides the increase of the overall PL intensity, the proportion of the red-light (631 nm) also increases obviously with the rise of the QD concentration, as shown in Fig. 6b. This phenomenon can be attributed to the enhanced FRET with increasing QD concentration. Moreover, the rising rate of the red light becomes slower at higher QD concentration. This could be due to the saturation of the energy transfer (ET) amongst QDs. However, the absolute QEs of the QD-silicone composite thin films exhibits a less than 5% change with different concentrations of composite-QDs, as shown in Fig. 6c. It appears that 1.0-1.5 mg/mL is the most favourable QD concentration for composite-QDs in application, which ensures high QEs with a low spectrum variation.
The TRPL decay curves of the different composite-QDs concentration thin films are shown in Fig. S2. Table S4 lists the amplitudes, lifetime components, and amplitude-weighted lifetimes for the composite-QDs. Their FRET efficiencies are calculated and shown in Table S5. Furthermore, the changes in lifetimes and FRET efficiency with concentration are clearly shown in Fig.6d. In detail, the FRET efficiency exhibits a decrement trend from 22% to 9% with the rise of concentration. Meanwhile, the lifetime recorded at the emission wavelength of the donor O-QDs increases with increasing concentration (orange dots in Fig. 6d). This is similar to the concentration dependent lifetime of the pure O-QDs samples shown in Fig.3. It suggests the existence of the combined effect of FRET and self-absorption (like the monochrome QDs). With the rise of concentration, the enhanced self-absorption leads to the increment of the τDA (the donor fluorescence lifetime in the presence of the acceptor, as shown in Fig. 6d, orange dots), suggesting the inhibition of the FRET between composite-QDs (blue dots in Fig. 6d). At the acceptor R-QDs emission wavelength, the reduced FRET efficiency results in smaller increment of the lifetime in high concentration (red dots in Fig. 6d). As a result, the composite-QDs show relative weak concentration dependent lifetimes and could maintain a stable QE, which benefits the application of composite-QDs in LED applications.
To study the light compensation effect of the composite-QDs in lighting application, WLEDs are fabricated by mixing green-emitting LuAG:Ce phosphor and O-QDs, R-QDs or composite-QDs (R:O=1:10) and packaging the mixture on top of 450 nm emitting GaN chips. Under a driving current of 40 mA, the excitation-luminescence (EL) spectra of the as-prepared WLEDs are illustrated in Fig. 7. The correlated color temperature (CCT) and color coordinates of the WLEDs are shown in Fig. S3 and Table S6. The four WLEDs have almost the same spectra in blue-green light region but different in the orange-red light region. Besides, the LuAG:Ce (only) based WLED shows the lowest color rendering index (CRI) of 48.8 due to the loss of red-orange light region. On the contrary, the composite-QDs based WLED exhibits a broader and flat spectrum in the orange-red light region and the highest CRI of 92.1. Compared to the composite-QDs, the LuAG:Ce (only) and R-QDs based WLEDs present obvious light gap in the orange light region and show great differences in CCT and color coordinates. Although the O-QDs based WLED has the similar CCT and color coordinates with the composite-QDs based WLED, it lacks the red light and thus presents a much lower CRI than that of the composite-QDs. It indicates the promising ability of composite-QDs in enhancing the color quality of WLED.
To further evaluate the experimental results, the luminous efficacy of radiation (LER) was calculated according to the following formula:
[Please see the supplementary files section to view the equation.] (4)
where 683 lm/Wopt is a normalization factor. Wopt, V(λ) and P(λ) are optical power, the human eye sensitivity function and the spectral power density of the light source, respectively [35,36].
The LER results are summarized in Table S6 and similar to previous reports [37,38,39]. According to the results, the LER of the composite-QDs based WLED (sample d) is higher than that of the R-QDs one (sample c) and lower than the O-QDs one (sample b) for the reason that human eyes are more sensitive to the orange light than the red light.