A novel light-emitting diode (LED) packaging method, named the active packaging (AP) method, is presented in this paper. In this method, during the LED packaging process, the light emitted from a GaN LED chip itself is employed to package the LED encapsulant, thereby eliminating the need to utilize a mold. Current injection into a bare LED chip, triggers a photosensitive epoxy to polymerize, leading to the formation of mushroom lamp cap on the LED chip. The emission properties of LEDs fabricated by this method, including their emission beam profiles and light outputs, were characterized. The results showed that a self-focusing effect happened with the addition of an epoxy on the chip. The simulation demonstrated that the geometry the encapsulant controlled the beam pattern of emission. Further, the self-focusing effect was believed to be caused by the combination of the threshold energy of epoxy polymerization, the beam pattern and the power output of the LED chip.
© 2008 Optical Society of America
The interest for using light-emitting diodes (LEDs) for display and illumination applications has been growing steadily over the past few years. The potential for long life and reduced consumption of energy are the two key attributes of this rapidly evolving technology, which has generated tremendous interest for use in the abovementioned applications. Generally, an LED is fabricated by first bonding an LED chip to metal frame post and then capping it with a molded epoxy encapsulant. Often, epoxy is utilized to mold an LED in the shape of a small bullet: this structure is called a “bullet head” . In general, LED encapsulation is a potting process, frequently requiring the use of thermosetting or ultraviolet (UV) curable resins [2, 3].
A free radical polymerization reaction of the photosensitive polymer could be triggered when the photosensitive polymer is exposed to visible light or UV light, and this could result in rapid curing at room temperature. Such photon-induced curing has a number of unique advantages over conventional curing technologies, thereby attracting extensive attentions in various research disciplines and industrial sectors, such as packaging of LEDs  and MEMS devices , application of printing inks , and fabrication of photonic crystal structure  and microvascular networks .
However, irrespective of the type of packaging method used--- thermal curing or the UV curable packaging---the packaging procedures are nearly the same. The encapsulant, which is employed to protect and control LED emission properties, is formed with the involvement of a mold. One of the important aspects of this packaging method is that it does not consider individual LED chip emission properties: therefore, it is termed a passive packaging (PP) method in our study. Our method, as mentioned below, utilizes the light emitted from the LED chip itself to package the LED with a photosensitive epoxy: we name this technique the active packaging (AP) method. This photon-induced polymerization method, which leads to the formation of specific encapsulant shapes and characteristics, eliminates the need to utilize a mold. By controlling the packaging parameters, for example, the polymerization time (PT) and the polymerization current (PC), the emission properties of LEDs can be controlled. In this report, a photosensitive epoxy was employed to fabricate LED encapsulants by the AP method, and the relevant characteristics of the encapsulants were investigated also. In addition, we observed a self-focusing effect, which was generated during the experiment. The simulation of the LED emission beam profile confirmed that this kind of encapsulant could focus the light emitted from an LED chip.
The products utilized in this experiment----photoinitiator (Ciba Irgacure 819), difunctional monomer HDDA (1, 6-hexanediol diacrylate) and Ebecryl resin 80 (highly reactive oligomeric polyether tetraacrylate), were made by Ciba Specialty Chemical Inc (South Korea). The epoxy resin system was composed of photoinitiator/HDDA/Ebecryl resin 80 (0.2/25/74.5 wt.%). The sensitive wavelength of the photoinitiator ranged from 460 nm to shorter wavelengths, as shown in Fig. 1(a). The emission peak of the To-Can-type LED was observed at 455 nm with an FWHM of 15 nm under 1 mA current injection (Fig. 1(a)). The overlap of the spectrum between the electroluminescence (EL) of the To-can-type GaN LED and the absorbance of the photoinitiator was utilized for the experiment. The emission beam profiles of the LED chip under the corresponding work currents are given in Fig. 1(b). From this figure, it is clear that the emission beam profile of the LED chip is almost the same with respective light output intensities. The proposed threshold energy for triggering epoxy polymerization with 455 nm is indicated by the dotted line in Fig. 1(b).
2.2 Encapsulation process
As shown in Fig. 1(c), the bare LED chip was immersed in the photosensitive epoxy, and then the PC was injected into the LED chip. The photoinitiator was triggered by the illumination of the LED chip and consequently, the photosensitive epoxy was polymerized. This resulted in an encapsulant cap forming on the LED chip. The samples were after removed from the epoxy liquid and washed with acetone to remove any non-polymerized epoxy. The typical encapsulant was a mushroom-shaped lamp, as shown in Fig. 1(e) inset. The samples fabricated with the same current (1 mA) at different polymerization times (PT) (1, 2, 3, and 4 s) were named PT1, PT2, PT3, and PT4, respectively, as shown in Fig. 2(a). The samples that curved in 1 s with different PCs (1, 10, 20, and 30 mA) were named PC1, PC2, PC3, and PC4, respectively, and are presented in Fig. 2(b).
The images of samples were captured using a CCD camera. Their geometries were sketched from the side view of the samples. The beam profiles were measured under a 50 mA current. The light outputs and quantum efficiencies of the samples were characterized using an OL 770 Multi-channel Spectroradiometer (Optronic laboratories, Inc.) under work currents of 1, 10, 20, 30, 40, and 50 mA. The absorbance of the epoxy was characterized by an ultraviolet (UV)-visible absorption spectrometer (Thermo Spectronic, Genesys 6 spectrophotometer).
The simulation was performed by LightTools (Version 5.4). The parameters for the simulation were set similar to those of the samples. The GaN LED chip was set as a Lambertian source, with an area of 0.60 mm×0.60 mm and a height of 0.08 mm. The profiles of the encapsulants were sketched based on the samples. The reflective index of the chip and the encapsulant was 2.4 and 1.55, respectively. The total number of rays for the simulation was set to 1,000,000.
3. Results and discussion
The geometries of the LED encapsulants were characterized and presented in Fig 2(c), 2(d)and 2(e), 2(f) for the PT and PC cases, respectively. The figures show that the height of PT encapsulants increased from 2.0 mm to 2.2 mm to 2.9 mm to 3.4 mm as the diameter increased from 1.3 mm to 1.9 mm to 2.2 mm to 2.4 mm, respectively. This caused the height-to-diameter ratio fluctuate from 0.67 mm to 0.79 mm to 0.84 mm to 0.78 mm, respectively. On the other hand, the height of the PC encapsulants increased from 1.3 mm to 1.9 mm to 2.1 mm to 2.2 mm for PC1 to PC4 as the diameter increased from 2.0 mm to 3.4 mm to 3.5 mm to 3.9 mm, respectively. This resulted in the height-to-diameter ratio varied from 0.67 to 0.59 to 0.64 to 0.58, respectively. It could be concluded from the geometry and the size of the LED encapsulants that the encapsulant volume increases with an increase in both PT and PC. However, it is interesting to note that the encapsulant geometries are different in the PT and PC cases. In the PT case, the height-to-diameter ratio increases with the encapsulant volume increase while in the PC case, the height-to-diameter ratio decreases with the encapsulant volume increase. Such distinctive morphological development variations with respective packaging parameters prove that the packaging parameters, for example, PC or PT, could be employed to control the encapsulant morphology with good reproducibility.
In the PT case, the beam profile of the LED chip was defined; more polymerization occurred with the passage of time. This resulted in the encapsulant height and diameter increasing. Since the emitted light output of the LED chip at center was higher than that at edge, therefore, the polymerization rate is greater at the center than at the edge, resulting in the height increase being higher than the diameter increase; consequently, the height-to-diameter ratio decreased. The LED chip exhibits a curve shape similar to that of the beam profiles under corresponding work currents. However, for a certain light output, such as the proposed threshold energy for triggering epoxy polymerization in Fig. 1(c), the higher the work current, the wider was the emission angle of the chip at this energy. This caused the encapsulant diameter to increase, since epoxy polymerization occurs only when the light output of the chip is higher than the threshold energy triggering epoxy polymerization . On the other hand, the height increase would be less than the diameter increase, since the epoxy polymerization rate above the threshold energy did not increase significantly as compared to change in the polymerization rate from below to above the threshold energy. For this reason, the widening of the threshold energy emission angle caused the diameter to increase faster than the height with the PC increasing, resulting in the height-to-diameter ratio decrease.
Figure 3(a) shows the beam profiles of samples PT1 to PT4 under 20 mA current injection: here, the emission angles decrease from 132° for the bare chip to 130°, 97°, 91°, and 85° for PT1 to PT4, respectively. These decreases are accompanied with an increase in the peak intensities from 2.2 to 4.0 to 4.6 to 5.3 to 6.2, respectively. Both the narrowing of the emission profiles and the increase in the peak intensities demonstrate that the light emitted from the chip was focused. Figure 3(b) shows the simulation result obtained from the respective encapsulant profiles by using LightTools. The simulation shows that the emission angles decreased from 140 ° for a bare LED to 125°, 96°, 92°, and 80° for the four samples, respectively. The corresponding peak emission intensities varied as 2.2, 4.0, 4.6, 5.3, and 6.2, respectively. The simulation result was in good agreement with the experiment result, which indicates that the respective encapsulant profiles for the PT case resulted in the respective beam profiles for the samples.
The beam profiles of samples PC1 to PC4 (in Fig. 3(b)) illustrate that with an increase in PT, the emission angles of the samples decrease from 132° for the bare LED to 128°, 105°, 95°, and 82° for the four samples, respectively. In addition, the peak emission intensities correspondingly increased as follows: 2.1 (a. u.), 3.2, 4.2, 4.4, and to 5.2. The simulation results revealed that the emission angles decreased from 140° for the bare LED to 124°, 118°, 103° and 88° for the four samples, respectively. In addition, the peak emission intensities correspondingly increased as follows: 2.2, 3.2, 4.0, 4.4, and 5.4. The simulation result was in good agreement with the PC experiment result. The encapsulant has several functions. First, it serves as a transparent body that holds and protects the device; second, it behaves as a lens that focuses the light in the desired way and improves the light output of the LED die . Currently, encapsulants are controlled by molds that form required shapes. From our results, we concluded that encapsulants formed by the AP method, without utilizing a mold, can control the light emitted from chips. Irrespective of whether the PC case or PT case is considered, the beam patterns of the samples become narrow with an increase in PC or PT. The focusing effect is exceedingly apparent as the thickness of the continuous epoxy layer on the LED chip increases. In fact, as the encapsulant volume on bare chips increased, more light was focused (in Fig. 2(e) and 2(f)). This clearly proved that the self-focusing effect originated from the utilization of the AP process. The shape of the encausulant is close to a convex lens. The encapsulant magnified the LED chip 1.5 times (from 0.6 x 0.6 mm to 0.9 x 0.9 mm); the resulting shape appeared like a convex lens. These results demonstrate that the encapsulant fabricated by our method effectively focused the light emitted from an LED chip (Fig. 1(d) and 1(e)).
We reported a novel LED packaging method---the active packaging (AP) method, and discussed the characteristics of the proposed method. In the AP method, during the LED packaging, the light emitted from the GaN LED itself was employed to obtain the LED encapsulant. With current injection into the bare LED chip, mushroom encapsulant cap forms on the LED chip. It was observed that the emission angles of LEDs changed with the type of the encapsulant used, which means that the emission angles could be controlled by both PTs and PCs without the need to utilize a mold. Moreover, as the PC or PT increased, the emission profiles became narrower and could be reduced to 80° in our experiment. This proved that the self-focusing effect was generated. Lastly, the emitted peak intensities of the LEDs increased with the encapsulants volume increase
This work was supported by the Natural Fund of Guangdong Province, China (5300077), the Korea Research Foundation Grant funded by the Korean Goverment (MOEHRD)(KRF-2005- 005-J07501) and the Regional Technology Innovation Program (Grant No. RTI04-03-06) of MOCIE of the Korean Government and by the Korean Science and Engineering Foundation through the National Research Laboratory Program and through QSRC at Dongguk University.
References and links
1. D.A. Steigerwald, J.C. Bhat, D. Collins, R.M. Fletcher, M. O. Holcomb, M. J. Ludowise, P. S. Martin, and S. L. Rudaz, “Illumination with solid state lighting technology,” IEEE J. Selected Topics Quantum Electron. 8, 310–320 (2002). [CrossRef]
2. R.J.M. Zwiers, H.J.L. Bressers, B. Ouwehand, and D. Baumann, “Development of a new low-stress hyperred LED encapsulant,” Components, Hybrids, and Manufacturing Technology, IEEE Transactions on 12, 387–392 (1989). [CrossRef]
3. R. N. Kumar, L. Y. Keem, N. C. Mang, and A. Abubakar, “Ultraviolet Radiation Curable Epoxy Resin Encapsulant for Light Emitting Diodes,” J. Appl. Polym. Sci. 100, 1048–1056 (2006). [CrossRef]
4. J. N. Tey, A. M. Soutar, S. G. Mhaisalkar, H. Yu, and K. M. Hew, “Mechanical properties of UV-curable polyurethane acrylate used in packaging of MEMS devices,” Thin Solid Films 504, 384–390 (2006). [CrossRef]
5. A. Hancock and L. Lin, “Challenges of UV curable ink-jet printing inks - a formulator’s perspective,” Pigment & Resin Technology 33, 280–286 (2004). [CrossRef]
6. J. Serbin, A. Ovsianikov, and B. Chichkov, “Fabrication of woodpile structures by two-photon polymerization and investigation of their optical properties,” Opt. Express 12, 5221–5228 (2004). [CrossRef] [PubMed]
8. H. Wang, K. Lee, S. Li, L. Jin, S. Lee, Y. Wu, Y. Cho, and J. Cai, “Fabrication of CdSe-ZnS nanocrystal-based local fluorescent aperture probes by active polymerization of photosensitive epoxy,” accepted by Opt. Comm. (2008) http://dx.doi.org/10.1016/j.optcom.2007.11.086.
9. R. Bachelot, C. Ecoffet, D. Deloeil, P. Royer, and D. -J. Lougnot, Appl. Opt. 40, 5860 (2001). [CrossRef]