1. Introduction

With the cooperation of BMW and OSRAM, the M4 iconic laser headlight module was firstly introduced to the BMW i8 model in 2014 ^(1-2). Since then, Audi has also embarked on laser headlight development, raising expectations for future automotive front light systems ⁽³⁾. Laser headlight has twice the illumination distance and 10 times higher brightness than current LED headlight with high energy efficiency and small packaging size, which are attracting points as a future headlight’s light source for vehicles ^(4-6). However, in spite of its various advantages, the laser headlight is still technically challenging to be commercialized ^(7-8). Actually, it is required to improve the stability and economic advantage of phosphor material which convert the laser single beam to white light.

Phosphor type is usually classified into epoxy, ceramic and glass phosphor (or phosphor-in-glass) depending on the matrix materials which mix with phosphor particles, as shown in Table 1. Up to now, a ceramic phosphor with a high thermal durability has been used as a wavelength conversion material for the high power laser source ^(9-10). Usually, ceramic phosphor is applied with remote phosphor structure like OSRAM's Laser Activated Remote Phosphor (LARP) system to minimize thermal problem ^(11-12). However, since ceramic phosphor shows a relatively low quantum efficiency, many studies have been carried out to improve the optical efficiency of ceramic phosphor ^(13-15). Also, ceramic phosphor is expensive due to high sintering temperature and process conditions compared to other phosphor types.

In this study, we have considered the application of phosphor-in-glass (PiG) plate as a wavelength converter for the laser beam since it shows higher quantum efficiency and lower price than those of ceramic phosphor. There are some previous reports using PiG as a wavelength conversion material of laser ^(16-18). The internal quantum efficiency of PiG is over 80%, which is more than 10% higher than that of ceramic phosphor of 60~70%. In addition, even if it was exposed to relatively high temperature for long time, the luminous efficacy reduction and the color change was small ⁽¹⁹⁾. In addition, PiG is sintered at relatively low temperature with simple fabrication process, making it to be manufactured less expensively compared to ceramic phosphors ^(16-19). However, the PiG plate is not satisfied at whole requirements as a laser wavelength converter in the automobile headlight application since there are additional issues to be solved, including the deviations in the correlated color temperature (CCT) and intensity for white light beam.

A phosphor plate includes three inevitable limits when applied as a wavelength converter of laser, regardless of the type of phosphor materials due to the unique characteristics of laser itself, not the phosphor. The first one is the optical efficiency of color converting. As shown in Figure 1, when a laser is incident on the phosphor plate, only 40% of the light source is used for illumination due to the back emission, reflection, and scattering, which affects the brightness of the headlight. To improve the optical efficiency, the phosphor plate is fabricated as a reflective type instead of a transmissive type, which means the phosphor plate placed on the reflective mirror ⁽²⁰⁾. So in this study, a reflective type PiG plate is prepared to increase the optical efficiency.

Fig. 1. (a) Optical loss by back emission, scattering and reflection when a laser is incident on phosphor plate. (b) Spectrum of laser and illuminated light

The second problem is the color deviation of white light emitted from the phosphor plate. Figure 2 is the correlated color temperature (CCT) data of white light emitted from PiG plate as a function of angle when a laser has a normal incidence. In this case, the ΔCCT (maximum-minimum CCT) which indicates the deviation of angular CCT of the white light, is about 500K. This value shows that blue light is concentrated at the direction of the laser incidence. This problem also causes the intensity deviation as a function of emitted angle.

Fig. 2. CCT distribution of white light emitted by laser excited phosphor-in-glass (PiG)

The last one is the thermal problem due to the high power of the laser. A PiG plate shows thermal stability up to about 200°C with no thermal degradation. However, when the laser is incident continuously, there is locally temperature increase on the PiG plate, resulting in the thermal shock ^(21-22). Therefore, this study aims on solving above problems such as the optical efficiency, color temperature deviation, and thermal problem of the PiG color converter. By applying PiG plate with the scanned laser beam, a new optical system was proposed and the optical characteristics are evaluated experimentally to confirm the possibility of PiG application for the automotive laser headlight.

2. Design Concepts and Mechanisms

2.2 Additional scattering effect by TiO$_2$ film

For the additional improvement, we applied a scattering component of TiO$_2$ nanoparticles which is located under the PiG plate. Thus, a laser beam can be scattered at wider angle to increase the probability of meeting with the phosphor particle to improve the efficiency of color converting of PiG plate and more uniform color distribution of white light. A TiO$_2$ film was fabricated and inserted between the phosphor and reflective layer, allowing more scattering of blue laser and white light.

Fig. 3. Light spreading mechanisms of proposed structure applying MEMS scanned beam

Fig. 4. Additional scattering of unconverted laser light by TiO$_2$ layer embedded between PiG and reflector

3. Experiments and Results

3.1 Improvement of the optical performances

3.1.1 Reflective type of phosphor

Before examining the effect of scanned laser beam and TiO$_2$ film, we first experimentally evaluated the color conversion efficiency of the reflective type of PiG plate, which consists of YAG:Ce phosphor particles in the borosilicate glass. The thickness of 200um was optimized to show the best color conversion efficiency and desired CCT around 4000K. For the reflection type PiG, it is mounted on the reflective mirror. The external quantum efficiency of the reflective type PiG plate was measured as 0.78. Since the blue laser is re-entered to the PiG after reflecting the mirror, the color conversion efficiency increases in the reflective type. All further experiments were conducted based on this structure.

3.1.2 The improvement by the scanned beam

Experiments were conducted to confirm the optical improvement by the scanned laser beam. As shown in Figure 5, a MEMS scanner is located between the laser source and the reflective type PiG plate, where the wavelength of 450nm and optical power of 80mW blue laser was used. The laser beam from the MEMS scanner is incident on the reflective type PiG to convert to white light. The MEMS scanner oscillates in both x- and y-direction and scans the whole surface of PiG plate with the frequency at 28.8 kHz and at 60 Hz in the x- and the y-axis, respectively. Since it is scanned at such a high speed, it is not recognized with the naked eye, perceived as continuous light emission.

Fig. 5. Experimental setups for evaluating angular characteristics of white light emitted from PiG plate

Fig. 6. Experimental setup for measuring angular characteristics of white light depending on incident angle of 45° and 90° (normal direction)

The angular characteristics of white light is measured by a detector rotating around the reflective type PiG plate with quantum efficiency measuring equipment. The measured angular CCT and intensity data was evaluated to calculate the color deviation of the emitted white light.

The experiment was conducted with two different incidence angle 45° and 90° (or normal direction) of to the PiG as shown in Figure 6. In this experiment, we investigated how the scanning beam and its incidence angle effects on the characteristics and distribution of white light emitted from PiG plate.

As shown in Figure 7 and Figure 8, it was confirmed that the angular CCT and intensity distribution are improved in both incidence angles through scanning, which means that the color deviation of white light decreased. Especially, the ΔCCT which represents the degree of a color deviation decreased by more than 50%, as shown in Table 2. In case of 45° incidence angle, the intensity of beam highest at specular reflection angle, but the intensity distribution is asymmetric with higher value of CCT. In the 90° incident angle, the intensity distribution is symmetrical with the lower value of ΔCCT.

Fig. 7. Angular (a) intensity and (b) CCT measurement results at 45° incidence angle for single beam and scanned beam

Fig. 8. Angular (a) intensity and (b) CCT measurement results at 90° (normal direction) incidence angle for single beam and scanned beam

3.1.3 The application of TiO$_2$ film

The TiO$_2$ films were fabricated for the additional scattering effect, as shown in Figure 9. First, TiO$_2$ coating solution was prepared by mixing the TiO$_2$ nanoparticles in propylene glycol monomethyl ether acetate (PGMEA) solution with UV curable epoxy resin modifier (ERM) polymer. Then it was coated on the glass substrate and the film thickness was controlled by the gap gauge. The diameter of TiO$_2$ nanoparticles is 300nm, and the film thickness is fixed at 50um. Also, the film was prepared from the TiO$_2$ coating solution of 10wt% concentration since it resulted in highest scattering effect at our preliminary test. After inserting the film under the PiG plate, the CCT and intensity were measured again with the same methods. The schematic measurement system is shown in Figure 10.

Fig. 9. Fabrication process of TiO$_2$ film

Fig. 10. Additional scattering of blue laser and white light by TiO$_2$ nanoparticles between PiG plate and reflective mirror

As shown in Figure 11 and 12, the ΔCCT and the intensity deviation are decreased further when TiO$_2$ film was applied. In both cases, the overall distribution shape of CCT and intensity is similar to those of case without TiO$_2$ nanoparticles, but the deviations are decreased since blue and white light shows large angle scattering effect. Also, the CCT and intensity distribution were the best in the case of normal incidence angle.

Through these experiments, it was confirmed that the scanned laser beam spreads the blue light more widely and reduces the color temperature deviation, and it can be further improved by the additional scattering effect of TiO$_2$ nanoparticles. In addition, the distribution of the emitted light varies according to the incidence angle, resulting in better performance at normal incidence angle. Therefore, if the incidence angle is designed to be close to 90° in the automotive headlight, the best performance will be realized with the symmetric optical distribution.

Fig. 11. Angular (a) intensity and (b) CCT measurement results at 45° incidence angle with TiO$_2$ film

Fig. 12. Angular (a) intensity and (b) CCT measurement results at 90° (normal direction) incidence angle with TiO$_2$ film

Table 2. ΔCCT at the incidence angles of 45° and 90°

(Unit: K)

	45°	90°
PiG + Single beam	351.1	305.5
PiG + Scanned beam	315.5	246.9
PiG + TiO$_2$+Scanned beam	263	128.3

3.2 Thermal improvement

The thermal characteristics of PiG plate was evaluated with the similar setup, as shown in Figure 13. An IR camera was used to measure the temperature distribution on the PiG plate with the laser scanned beam. A laser beam of 80mW was entered on the PiG plate with and without the MEMS scanner for 15 minutes to confirm the temperature saturation.

Fig. 13. Measurement setup of temperature distribution after a laser beam focusing on PiG plate with and without MEMS scanner using IR camera

As a result, it was found that the temperature deviation on the PiG plate was changed from 10°C to 0.2°C without and with the laser scanning, respectively, as shown in Figure 14. Thus, it was confirmed that the scanned beam enables to prevent the thermal shock by reducing local temperature deviation. The quantitative temperature data is summarized in Table 3. The absolute temperature itself is also reduced due to the short illumination time in the case of laser scanning. Thus, the laser scanning can reduce the thermal problems of PiG plate which may occur in the case of continuous laser focusing case.

Fig. 14. IR images of temperature distribution during the incidence of (a) single beam and (b) scanned beam.

Table 3. Temperature comparison on PiG plate between continuous single beam and scanned beam cases

(Unit: °C)

	Single beam	Scanned beam
Max. Temperature	45.5	27.5
Min. Temperature	35.2	27.3
ΔT	10.3	0.2

3.3 Design of automotive laser headlight

Finally, we designed an automotive laser headlight system based on our concept, including a laser, MEMS scanner, PiG plate and TiO$_2$ film on the reflective mirror, as shown in Figure 15. Here, the incidence angle of laser to the PiG plate is kept as near-normal incidence to maintain the maximized symmetry and minimized color deviation of white light.

Fig. 15. The schematic configuration of the designed automotive laser headlight

The designed headlight consists of several unit modules like the LED headlamp, which enables to attach together as single system. In addition, PiG plate and laser diodes are located on the same heatsink, making it possible to manufacture more compact structure. The white light from these phosphor assemblies with an aid of MEMS scanner or actuator is reflected again on the surface of macro focal reflector together to realize circular white beam. In addition, each sectional module will be operated independently to realize various type of beam shape.

4. Conclusions

To realize commercial automotive laser headlight, the stability of color converting materials must be confirmed. Thus, new optical system concept of automotive laser headlight with the application of phosphor-in-glass (PiG) plate as a wavelength converter and the scanned light beam using MEMS scanner was proposed and its basic optical and thermal characteristics were evaluated experimentally. By applying laser scanning, it was improved on both deviations of correlated color temperature (CCT) and intensity in white light emitted through the PiG plate. A TiO$_2$ film was also applied under the PiG plate to further improve the CCT and intensity deviation by additional scattering effect. In addition, it was confirmed that there is little temperature deviation on the PiG plate of 0.2°C with the scanned laser beam to prevent the severe damage on the PiG plate due to thermal shock. Based on the results, it was possible to design a laser scanning module to improve both optical and thermal problems of automotive laser headlight. Finally, we have confirmed that PiG plate can be used as a good wavelength converter with an aid of laser scanning for various laser application fields.