Better color distribution uniformity and higher luminous intensity for LED by using a three-layered remote phosphor structure

Recently, white light-emitting diodes (WLEDs) have gained much recognition among semiconductor light sources and are believed to have the potential to completely replace traditional sources of electrical illumination, such as mercury vapor, fluorescent, and high-pressure sodium lighting. Compared to these conventional lighting technologies, LED is better in energy savings, emits less heat, provides long-life performance that can be over 100,000 operating hours, and is generally associated with a lower maintenance cost [1,2,3]. LED lights are typically constituted by a blue-light–emitting chip coated with a thin layer of yellow-emitting phosphors and silicone compound. The blue light from LED dies passes through the phosphor layers and is partly absorbed and converted into yellow light by the yellow phosphors. This transmitted blue light and yellow light blend together to generate white light. The techniques commonly used to fabricate a LED comprise the freely dispensing and conformal phosphor-coating methods. While these methods are simple, they were found to result in a poor lumen output yield, and also, the phosphor materials became degraded quickly. The light in the dispensing and conformal coating was scattered back to the LED chip and then absorbed, causing a significant light-loss phenomenon and thus leading to poor light extraction efficiency [4, 5]. Moreover, in these traditional packaging methods, the phosphor has a direct interaction with the chip surfaces, and this increases the junction thermal generation during the operation time of an LED device, in turn lowering the performance of phosphor particles and the life span of an LED. Thus, ensuring that the phosphor coating layer is maintained at a distance from the chips can limit the generation of heat. The remote phosphor structure adapted this idea and caught the attention of researchers. This structure can reduce the proportion of light loss caused by deflection of light back to the chip, and thus improves the light transmission and extraction of LEDs [6,7,8]. The issue related to lumen output can be addressed with remote phosphor structure, but it cannot achieve much enhancement in color distribution uniformity [9, 10]. Further, while using a diffuser layer in the LED package decreased the energy efficiency of light transmission, this is a good method to acquire color uniformity [11,12,13]. Other approaches were also proposed to heighten the color properties of remote phosphor structures, such as patterned remote phosphor design [14], self-adaptive coating by slurry coating method combined with self-exposure technique [15], and single bi-color phosphor layer with red and yellow phosphor materials [16], but the luminous efficiency was somehow reduced.

The color rendering index (CRI), the currently used value to evaluate the color quality of an LED, could be enhanced by adding red phosphor into the remote phosphor layer [17]. Generally, the high CRI requires more red emission in the spectral distribution of white light, and this red-phosphor addition can provide more red-light components and achieve a higher CRI. The double-layer (DL) design for remote phosphor structure, which has a film of red-phosphor particles placed above the yellow-phosphor one, was introduced to achieve the promoted CRI. Specifically, it can obtain about 85 CRI when using red phosphor of LiLaO₂:Eu³⁺ with the concentration of more than 20%; however, the luminous flux was dramatically reduced as the concentration of red phosphor increased [18]. Besides, according to the spectral power distribution (SPD) model by Gaussian function, the white light spectrum can be expressed as a tricolor spectrum of blue, green, and yellow, since the yellow phosphor YAG:Ce³⁺ used in LED packages emits light in the spectrum regions of green and yellow colors [19]. Therefore, the green-phosphor layer was added to the remote phosphor structure to replace the red one. The green-yellow remote phosphor configuration can have better luminous flux but lower CRI than the red-yellow geometry [20]. Therefore, it seems difficult to simultaneously achieve color uniformity and high luminous efficacy.

In this study, our experiment involves introducing another remote phosphor structure by combining both red- and green-phosphor layers into the package. This three-layered remote phosphor configuration utilizes the red LiLaO₂:Eu³⁺ and green Zn₂GeO₄:Mn²⁺ phosphors. The green-phosphor layer is placed in between the yellow-phosphor and red-phosphor films. The color parameters and luminescent output of this three-layered (TL) package are calculated and examined with Mie-scattering theory and Beer's law, and then compared to those of DL structures. The obtained figures indicate that, in comparison with those of the DL model, there is a definite improvement in both the color uniformity and luminous efficiency of the TL structure. Therefore, the TL remote phosphor structure could be applied in the fabrication of a higher-quality WLED device that can provide efficient color and luminous performances.

The quality of a LED's chromaticity is generally examined by CRIs, which include the CRI, the most popular parameter in color evaluation. Obviously, a LED with high CRI has better brightness and a higher price. The color rendering indices of both DL and TL models at five different color temperatures are demonstrated in Figure 4. In both structures, the CRI is directly proportional to the CCT; in other words, a higher CCT has better CRI. It is noticed that the DL results in higher CRI at all CCTs. This can be attributed to the stronger red-light intensity provided by the red LiLaO₂:Eu³⁺ phosphor layer. Obtaining high CRI for WLEDs with high CCT is very challenging; therefore, the improvement in this color rendering ability of the DL and TL is noteworthy and important to the development of high-quality WLED devices. The difference between the CRI values of DL and TL structure is not too much: it is about 0.5–1; thus, either structure may be used while innovating in the attempt to carry out CRI improvements to WLEDs. However, if manufacturers aim at obtaining high CRI, it is more advantageous to apply the dual-layer structure.

Fig. 4

CRI is an important parameter by which to measure LED brightness; yet, it does not facilitate an overall evaluation of the color quality offered by an LED device. Moreover, if the LED has an excessively high CRI or brightness, it will cause discomfort to human eyes. Therefore, the color quality scale (CQS) has emerged as an attractive parameter to researchers in examining the color quality of an LED, since it evaluates light color quality via CRI, chromatic coordinate, and observer preferences. Similar to CRI, the CQS in the DL and TL also increases with the upward movement of CCT values, as shown in Figure 5. It is obvious that the TL has better CQS than the DL, regardless of the CCT. The uniformity of chromatic distribution among the red, yellow, and green colors is observed when applying the three-layered configuration, leading to higher color performance for WLEDs. Therefore, it can be said that TL structure is appropriate for boosting the color properties of WLED packages, with excellent CQS.

Fig. 5

Good color quality, or high CRI and CQS, usually leads to inefficiency in lumen output. This has raised a question as to the lumen efficiency of TL phosphor packaging, since it offers high improvement of color quality with impressive CQS results. A series of mathematic expressions pertaining to the light conversion and transmission of dual-layer and TL structures is performed based on the Mie-scattering theory to further analyze the effects of TL on LO, in comparison with the dual-layer.

The transmitted blue light and converted yellow light of DL remote phosphor with a phosphor-film thickness of h is calculated using Eqs (1) and (2), as shown below [22, 23]: (1) PB3=PB0.e−αB22h3,e−αB22h3,e−αB22h3=PB0.e−2αB3h. \matrix{ {P{B_3}} \hfill & { = P{B_0}.{\kern 1pt} {e^{ - {\alpha _{B2}}{{2h} \over 3}}},{e^{ - {\alpha _{B2}}{{2h} \over 3}}},{e^{ - {\alpha _{B2}}{{2h} \over 3}}}} \hfill \cr {} \hfill & { = P{B_0}.{\kern 1pt} {e^{ - 2{\alpha _{B3}}h}}.} \hfill \cr } (2) PY2=12β2PB0αB2−αγ2[e−αγ2h−e−αB2h]e−αγ2h+12β2PB0αB2−αγ2[e−αγ2h−e−αB2h]=12β2PB0αB2−αγ2[e−2αγ1h−e−2αB1h] \matrix{ {P{Y_2}} \hfill & { = {1 \over 2}{{{\beta _2}P{B_0}} \over {{\alpha _{B2}} - {\alpha _{\gamma 2}}}}\left[ {{e^{ - {\alpha _{\gamma 2}}h}} - {e^{ - {\alpha _{B2}}h}}} \right]{e^{ - {\alpha _{\gamma 2}}h}}} \hfill \cr {} \hfill & { + {1 \over 2}{{{\beta _2}P{B_0}} \over {{\alpha _{B2}} - {\alpha _{\gamma 2}}}}\left[ {{e^{ - {\alpha _{\gamma 2}}h}} - {e^{ - {\alpha _{B2}}h}}} \right]} \hfill \cr {} \hfill & { = {1 \over 2}{{{\beta _2}P{B_0}} \over {{\alpha _{B2}} - {\alpha _{\gamma 2}}}}\left[ {{e^{ - 2{\alpha _{\gamma 1}}h}} - {e^{ - 2{\alpha _{B1}}h}}} \right]} \hfill \cr }

Each phosphor layer's thickness of the TL phosphor package is demonstrated as 2h3 {{2h} \over 3} . The computation of transmitted blue light and converted yellow light in the TL can be carried out as: (3) PB3=PB0.e−αB22h3,e−αB22h3,e−αB22h3=PB0.e−2αB3h. \matrix{ {P{B_3}} \hfill & { = P{B_0}.{\kern 1pt} {e^{ - {\alpha _{B2}}{{2h} \over 3}}},{e^{ - {\alpha _{B2}}{{2h} \over 3}}},{e^{ - {\alpha _{B2}}{{2h} \over 3}}}} \hfill \cr {} \hfill & { = P{B_0}.{\kern 1pt} {e^{ - 2{\alpha _{B3}}h}}.} \hfill \cr } (4) PY3′=12β3PB0αB3−αγ3[e−αγ32h3−e−αB32h3]e−αγ32h3+12β3PB0e−αB32h3αB3−αγ3[e−αγ32h3−e−αB32h3]=12β3PB0αB3−αγ3[e−αγ34h3−e−2αB32h3] \matrix{ {P{Y_3^\prime}} \hfill & { = {1 \over 2}{{{\beta _3}P{B_0}} \over {{\alpha _{B3}} - {\alpha _{\gamma 3}}}}\left[ {{e^{ - {\alpha _{\gamma 3}}{{2h} \over 3}}}} \right.\left. { - {e^{ - {\alpha _{B3}}{{2h} \over 3}}}} \right]{e^{ - {\alpha _{\gamma 3}}{{2h} \over 3}}}} \hfill \cr {} \hfill & { + {1 \over 2}{{{\beta _3}P{B_0}{e^{ - {\alpha _{B3}}{{2h} \over 3}}}} \over {{\alpha _{B3}} - {\alpha _{\gamma 3}}}}\left[ {{e^{ - {\alpha _{\gamma 3}}{{2h} \over 3}}} - {e^{ - {\alpha _{B3}}{{2h} \over 3}}}} \right]} \hfill \cr {} \hfill & { = {1 \over 2}{{{\beta _3}P{B_0}} \over {{\alpha _{B3}} - {\alpha _{\gamma 3}}}}\left[ {{e^{ - {\alpha _{\gamma 3}}{{4h} \over 3}}} - {e^{ - 2{\alpha _{B3}}{{2h} \over 3}}}} \right]} \hfill \cr } (5) PY3=PY3′.e−αγ32h3+PB0.e−2αB34h312β3αB3−αγ3[e−αγ32h3−e−αB32h3]=12β3PB0αB3−αγ3[e−αγ34h3−e−αB34h3]e−αγ32h3+12β3PB0e−αB34h3αB3−αγ3[e−αγ32h3−e−αB32h3]=12β3PB0αB3−αγ3[e−αγ3h−e−2αB3h] \matrix{ {P{Y_3}} \hfill & { = PY_3^\prime.{e^{ - {\alpha _{\gamma 3}}{{2h} \over 3}}}} \hfill \cr {} \hfill & { + P{B_0}.{e^{ - 2{\alpha _{B3}}{{4h} \over 3}}}{1 \over 2}{{{\beta _3}} \over {{\alpha _{B3}} - {\alpha _{\gamma 3}}}}\left[ {{e^{ - {\alpha _{\gamma 3}}{{2h} \over 3}}} - {e^{ - {\alpha _{B3}}{{2h} \over 3}}}} \right]} \hfill \cr {} \hfill & { = {1 \over 2}{{{\beta _3}P{B_0}} \over {{\alpha _{B3}} - {\alpha _{\gamma 3}}}}\left[ {{e^{ - {\alpha _{\gamma 3}}{{4h} \over 3}}} - {e^{ - {\alpha _{B3}}{{4h} \over 3}}}} \right]{e^{ - {\alpha _{\gamma 3}}{{2h} \over 3}}}} \hfill \cr {} \hfill & { + {1 \over 2}{{{\beta _3}P{B_0}{e^{ - {\alpha _{B3}}{{4h} \over 3}}}} \over {{\alpha _{B3}} - {\alpha _{\gamma 3}}}}\left[ {{e^{ - {\alpha _{\gamma 3}}{{2h} \over 3}}} - {e^{ - {\alpha _{B3}}{{2h} \over 3}}}} \right]} \hfill \cr {} \hfill & { = {1 \over 2}{{{\beta _3}P{B_0}} \over {{\alpha _{B3}} - {\alpha _{\gamma 3}}}}\left[ {{e^{ - {\alpha _{\gamma 3}}h}} - {e^{ - 2{\alpha _{B3}}h}}} \right]} \hfill \cr } where h shows the thickness of each phosphor layer in both DL and TL structures; the subscripted “2” and “3” indicate the DL and TL structures, respectively; β presents the conversion coefficient of blue light transformed to yellow light in the package; γ shows the yellow-light reflection coefficient; PB presents blue-light intensity and PY indicates yellow-light intensity; PB₀ is the LED-die emitted light intensity comprised of PB and PY; PY^′₃ is the intensity of the yellow light transmitted through both red- and green-phosphor layers; and α_B and α_Y indicate the fractions of blue-light and yellow-light energy loss, respectively.

The TL structure improves the lighting output of LEDs, and renders it significantly better than the results obtained from the dual-layer configuration, as can be observed in Eq. (6): (6) (PB3−PY3)−(PB2+PY2)(PB2+PY2)>e−2αB3h−e−2αB2he−2αγ3h−e−2αB2h> \matrix{ {{{(P{B_3} - P{Y_3}) - (P{B_2} + P{Y_2})} \over {(P{B_2} + P{Y_2})}}} \cr { > {{{e^{ - 2{\alpha _{B3}}h}} - {e^{ - 2{\alpha _{B2}}h}}} \over {{e^{ - 2{\alpha _{\gamma 3}}h}} - {e^{ - 2{\alpha _{B2}}h}}}} > } \cr }

As mentioned, Mie-theory [24, 25] is applied in the scattering analysis of phosphor materials and in determining the scattering cross-section C_sca. The power of transmitted light is calculated based on Beer's law: (7) I=I0exp(−μextL) I = {I_0}\exp \left( { - {\mu _{ext}}L} \right) Here, I₀ shows the power of incident light, and L and μ_ext demonstrate the thickness of each phosphor layer and the extinction coefficient, respectively. Besides, the calculation of the extinction coefficient is: μ_ext = N_r.C_ext. In this formula, N_r indicates the particles’ density distribution number (mm⁻³), and C_ext (mm²) presents the extinction cross-section phosphor spheres.

The effectiveness of using TL structure to improve the luminous flux can be inferred from Eq. (6). The lumen outputs of the DL and TL configurations are also expressed in Figure 6, from which we infer that the TL achieves about 2% enhancement in luminescence, compared to that achieved using the DL. Particularly, the luminous flux of the TL and DL increase when the CCT becomes higher. The superior lumen output of the TL in comparison with that of the DL can be observed clearly at high CCTs such as 7,700 K and 8,500 K. These impressive luminescence results in the TL model can be attributed to the higher green light emission when using the green phosphor Zn₂GeO₄:Mn²⁺, which is much greater than that from the DL, owing to the degradation of yellow-phosphor concentration for maintaining the stability of color temperature. This means the TL can reduce the backscattering and reabsorption of light, which enhances the efficiency of blue light transmission through the yellow-phosphor layers, facilitating the former to be converted effectively. It is also pointed out in Figure 3 that the TL results in higher red-light emission, thus resulting in greater enhancement in the color quality. Thus, the TL can yield high LO and enhance the chromaticity of LED at the same time.

Fig. 6

The angular chroma variation (D-CCT) of a LED package also plays an important role in light-chromaticity evaluation. The smaller the D-CCT presented, the more the chromatic homogeneity is improved. Accordingly, the results of D-CCT in the TL and the DL remote models are analyzed and presented in Figure 7. The color variations in the TL are much lower than in the DL. As can be seen, the D-CCT corresponding with the CCTs of TL is about 200–2,000 K smaller than that of the DL, which indicates the effectiveness of the TL in achieving higher chromatic homogeneity. This can result from the increase in the scattering properties of light when there are more phosphor layers in the remote structure. The improvement in light scattering probably leads to better light color mixing and distribution, and thus results in a reduction of the color variations. Yet, one disadvantage of scattering improvement is that it decreases the luminescence. However, the TL presents higher lumen output, better CQS, and significant reduction in backscattering events; therefore, a small decrease in its lumen intensity can be accepted.

Fig. 7

The effectiveness of the three-layered or TL remote phosphor configuration is demonstrated in this study. The green phosphor Zn₂GeO₄:Mn²⁺ and red phosphor LiLaO₂:Eu³⁺ are utilized for the model fabrication. Although the TL structure has a lower CRI, its CQS performs better since the color imbalance among red, yellow, and green is addressed. Also, the LO of the TL is higher than that of the DL due to the stronger green emission in the spectral region. The greater number of phosphor layers included in the TL structure enhances the scattering ability, which induces a reduction in backscattering, in turn leading to heightened blue-light transmission and conversion. The scattering improvement also increases the light color blending and results in a more uniform color distribution, thus lessening the color deviations. Though the higher scattering properties can cause a decrease in the lumen intensity, the benefits that the TL structure brings to the optical performance of LEDs via backscattering reduction have outweighed this drawback. Therefore, manufactures can apply the three-layered remote phosphor structure to accomplish higher-quality LED packages.