Effect of Thermal Stress and Short-wavelength Visible Radiation_2015

7/23/2019 Effect of Thermal Stress and Short-wavelength Visible Radiation_2015

1/6

Effect of thermal stress and short-wavelength visible radiation

on phosphor-embedded LED encapsulant degradation

Prathika Appaiah, Nadarajah Narendran , Indika U. Perera, Yiting Zhu, Yi-wei Liu

Lighting Research Center, Rensselaer Polytechnic Institute, Troy, NY, USA

a r t i c l e i n f o

Article history:

Received 25 November 2014

Received in revised form 9 March 2015

Accepted 10 March 2015

Available online 11 April 2015

Keywords:

Thermal degradation

Encapsulant

Short-wavelength radiation

White LED

Degradation

Reliability

a b s t r a c t

This study evaluated the effects of short-wavelength radiation and thermal stress leading to thermal

degradation of an LED encapsulant, which contributes to reduced light output. The effect of light trans-

mission was measured in encapsulant samples with different YAG:Ce phosphor concentrations mixed in

epoxy, which were subjected to short-wavelength optical radiation and thermal stress. Encapsulant sam-

ples with increasing phosphor concentrations showed lower degradation rates of the encapsulant under

thermal stress, while under short-wavelength optical radiation the samples with phosphor showed

higher degradation rates.

2015 Elsevier B.V. All rights reserved.

1. Introduction

Light-emitting diodes (LEDs) are finding success in numerous

applications including displays, signage, communication, and gen-

eral lighting. LEDs have equaled or surpassed traditional lighting

technologies such as incandescent, halogen, and fluorescent with

respect to efficacy, lifetime, and color rendering. Other value pro-

positions of LEDs are form factor, energy efficiency, and ease of

maintenance, which make them suitable for a variety of applica-

tions at the fixture or luminaire level. Even though LED general

lighting products are expected to be reliable based on reported life-

time, many factors affect their lifetime and therefore their reliabil-

ity, one being the encapsulant of the LED.

White phosphor-converted (pc) LEDs are commonly produced

with the combination of emitted optical radiation from a gallium

nitride (GaN) semiconductor LED and a yttrium garnet (YAG:Ce)phosphor. The phosphor is embedded in an encapsulant and pro-

duces a broad spectrum light by down-converting a portion of

the LEDs emitted optical radiation. The LED encapsulant, typically

made of epoxy or silicone, serves as an optical medium surround-

ing the LED die and helps to increase the extraction of generated

photons. The encapsulant also isolates the LED die from structural,

mechanical, and chemical stresses from the outside environment.

The degradation of the encapsulant plays a vital role in determin-

ing the useful life and reliability of the LED. The phosphor-embed-ded encapsulant is subjected to short-wavelength radiation and

thermal energy from the LED chip and the phosphor, due to the

losses in the down-conversion process, which cause degradation

of the encapsulant material. This degradation results in a reduction

in light output from the LED.

The degradation of encapsulants used in blue LEDs under the

effects of increasing drive current and increasing temperature

has been studied extensively [1]. In past studies, low-power LED

packages with epoxy encapsulant failed based on light output

degradation due to thermal stress[1,2]. This reduction in light out-

put was due to the yellowing of the encapsulant of the LED package

[2,3]. Similarly, in 5 mm pc-LEDs studies showed a reduction in

light output with time due to a change in transmittance of the

epoxy when operated at manufacturer-rated drive currents[3].In blue LEDs, the degradation at lower temperatures was due to

short-wavelength optical radiation but at higher temperatures, the

degradation due to heat was more pronounced[1]. In comparison,

pc-LEDs were found to degrade faster than blue LEDs [5]. This

increased rate in degradation was concluded to be caused by the

presence of phosphors in the epoxy [4,5]. Degradation due to

thermal stress and short-wavelength optical radiation leads to a

discoloration of the encapsulant, leading to a reduction in the relia-

bility of the encapsulant of the LED.

YAG:Ce is the commonly used phosphor for pc-LEDs. Studies

have shown YAG:Ce phosphors are stable at temperatures greater

http://dx.doi.org/10.1016/j.optmat.2015.03.030

0925-3467/ 2015 Elsevier B.V. All rights reserved.

Corresponding author at: Lighting Research Center, 21 Union St., Troy, NY

12180, USA. Tel.: +1 518 687 7100.

E-mail address:[email protected](N. Narendran).

Optical Materials 46 (2015) 611

Contents lists available at ScienceDirect

Optical Materials

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / o p t m a t
http://dx.doi.org/10.1016/j.optmat.2015.03.030mailto:[email protected]://dx.doi.org/10.1016/j.optmat.2015.03.030http://www.sciencedirect.com/science/journal/09253467http://www.elsevier.com/locate/optmathttp://www.elsevier.com/locate/optmathttp://www.sciencedirect.com/science/journal/09253467http://dx.doi.org/10.1016/j.optmat.2015.03.030mailto:[email protected]://dx.doi.org/10.1016/j.optmat.2015.03.030http://crossmark.crossref.org/dialog/?doi=10.1016/j.optmat.2015.03.030&domain=pdf


2/6

than 100 C [6]. Therefore, the main cause of degradation in the

encapsulant is most likely due to the epoxy degradation. In 2013,

Meneghini et al. showed an increase in the temperature of remote

phosphor samples when the irradiance level on the encapsulant

was increased [7]. This degradation of the encapsulant worsens

with an increase in duration and amount of exposure to short-

wavelength radiation and thermal stress.

From previous studies, it is clear that externally applied thermalstress and short-wavelength optical radiation are key factors that

degrade the encapsulant of the LED, but there is no clear under-

standing of the mechanism of degradation. This present study

evaluated the degradation of the LED encapsulant medium under

thermal stress and short-wavelength optical radiation with and

without phosphor in the medium.

2. Experiment

An experiment was designed to understand the effect of short-

wavelength optical radiation and externally applied thermal stress

on LED encapsulant degradation. The study utilized two experi-

mental setups, one to subject the encapsulant sample to applied

thermal stress at a specific temperature by heating the ambientaround the sample, and another to irradiate the samples with

short-wavelength optical radiation from blue LEDs. A separate

measurement setup was used to test the effect of the experimental

conditions by measuring the light transmission. Each encapsulant

sample was made on a glass substrate with a small depression that

contained the encapsulant material. The epoxy encapsulant mate-

rial was mixed with and without commercially available YAG:Ce

phosphor used in general illumination applications and cured in

an oven as per the curing profiles specified by the epoxy manufac-

turer. The epoxy encapsulant used in the study was a commercially

available, optically clear, two-part epoxy. The phosphor concentra-

tions chosen for the experiment were 10 mg/cm2 and 20 mg/cm2.

The weight of the phosphor added to each sample was consistent

with the values given inTable 1.The experimental setup was designed such that the encapsulant

samples were subjected to short-wavelength optical radiation and

external thermal stress separately. The short-wavelength optical

radiation was produced by 4 4 chip on board blue LED arrays

under a single primary lens with a peak wavelength of450 nm.

Each blue LED array was characterized in an integrating sphere

to determine the radiant power and peak wavelength. The LEDs

were placed on a thermoelectric cooler to ensure the operating

temperature was maintained. Temperature measurements were

recorded with thermocouples, which were attached to the refer-

ence temperature measurement locations of the LED array mod-

ules as per the LED manufacturers specifications, and were

maintained at 40 C.

The irradiance on the encapsulant samples was varied bychanging the distance from the LED array module and by changing

the drive currents to the LED array module. The external thermal

stress on the encapsulant samples was created using a heating ele-

ment controlled by a thermal controller. The experimental setup is

shown inFig. 1.

The measurement setup consisted of two integrating spheres

[8] to measure the light transmitted and reflected when the

phosphor samples were irradiated with the test LEDs (peak wave-

length 450 nm). Measurements were taken after the encapsulant

samples were subjected to either externally applied thermal stress

or short-wavelength optical radiation in the experimental setup.

The transmission level of each sample was normalized to that of

a clear epoxy sample that was not treated under either experimen-

tal condition. The normalized light transmission data collected

over time were again normalized to the data of each respectivesample at the initial stage before being subjected to either experi-

mental condition.

3. Results

The measurements from the two integrating sphere system

were normalized to epoxy sample data obtained at the time before

the samples were subjected to experimental conditions. The curve-

fit of the total light, which is the summation of both transmitted

and reflected light, was used for analysis while paired t-tests were

used to determine statistical significance at the 95% confidence

level.

3.1. Effect of external thermal stress as a function of phosphorconcentration

Epoxy with no phosphor and epoxy mixed with phosphor sam-

ples were subjected to a fixed, externally applied thermal stress

inside the experimental setup (Fig. 1), which maintained an ambi-

ent temperature around the encapsulant samples at 90 2 C. The

temperatures of the encapsulant samples and the ambient air were

measured using thermocouples.

The measured light transmission values of the encapsulant

samples are illustrated inFig. 2. The light transmission data were

curve-fit using an exponential function with the form:

y a expbx

where y is the normalized light transmission and x is the time inhours, whilea and b are the curve-fit parameters. The exponential

coefficient b is identified as the degradation rate of the material.

Evaluation of the exponential coefficients indicates that reac-

tion rates were similar for the three encapsulant samples under

the applied thermal stress condition. One-tailed paired t-tests

revealed there was statistical significance at 95% confidence inter-

val between the light transmission degradation of the encapsulant

samples.

To understand encapsulant degradation and the transmission of

the pump (blue) and the converted (yellow) light, the spectral

power distribution (SPD) of each sample was separated into two

parts at the local minima of the SPD between the pump and the

converted. The portion of the SPD with the wavelength segment

Table 1

Phosphor and epoxy weight added in creating encapsulant samples.

Epoxy

(no phosphor)

10 mg/cm2

of phosphor

20 mg/cm2

of phosphor

Epoxy weight (mg) 25.0 25.0 25.0

Phosphor weight (mg) NA 15.6 31.1

Phosphor weight (%) 0 38.4 55.4 Fig. 1. Experimental setup showing samples subjected to heat and short-wave-length radiation.

P. Appaiah et al./ Optical Materials 46 (2015) 611 7
http://-/?-http://-/?-


3/6

below the local minima was used to calculate the blue transmis-

sion component while the longer wavelength segment was used

to calculate the yellow transmission component.The normalized blue light transmission degradation rate

reduced as the phosphor concentration increased, as illustrated

in Fig. 3. The normalized blue light transmission for phosphor

mixed in epoxy samples showed an increase in the initial measure-

ments. This observation was more profound in the 20 mg/cm2

phosphor encapsulant sample, as evident from the initial data

points in Fig. 3. The degradation rate was highest for the epoxy

(no phosphor) encapsulant sample followed by the epoxy mixed

with phosphor at 10 mg/cm2 encapsulant sample, while the high-

est phosphor concentration encapsulant sample (20 mg/cm2) had

the lowest degradation rate. In addition, one-tailed paired t-tests

revealed that the degradation rates were distinct at 95% confidence

level when the three samples were compared with one another in a

paired comparison.The normalized yellow light transmission is illustrated in Fig. 4

based on the total transmitted yellow light for the phosphor mixed

in epoxy encapsulant samples. The degradation rate for the higher

phosphor concentration encapsulant sample was lower compared

with the lower phosphor concentration encapsulant sample. A

one-tailed paired t-test revealed that the yellow light total

transmission degradation rates were statistically significant at

95% confidence level between the two encapsulant samples.

Figs. 3 and 4 indicate that the light transmission degradation

decreased as the phosphor concentration in the epoxy increased.

3.2. Effect of short-wavelength optical radiation as a function of

phosphor concentration

To understand the effect of optical irradiation on LED encapsu-

lant material, encapsulant samples having phosphor mixed with

epoxy at 10 mg/cm2 and 20 mg/cm2 concentrations were subjected

to a constant level of short-wavelength optical radiation at 0.28 W/

cm2.Fig. 5shows the light transmission degradation as the phos-

phor concentration was increased. An epoxy encapsulant sample

without phosphor was irradiated at the same optical radiation

level and is also illustrated in Fig. 5. The equation of the trend

for the phosphor sample is given by:

y a1 expb1x a2 expb2x

The degradation rate of the epoxy (no phosphor) encapsulant

sample was lower than that of the phosphor mixed in epoxy

encapsulant samples.

The results showed that an increase in phosphor concentration

led to increased encapsulant degradation under short-wavelength

optical radiation.

Fig. 2. Normalized total light transmission with time for encapsulant samples

subjected to externally applied thermal stress. The parameters (a,b) specify fitted

curves of the formy a expbx.

Fig. 3. Normalized total blue light transmission with time for encapsulant samples

subjected to externally applied thermal stress. The parameters (a,b) specify fittedcurves of the formy a expbx.

Fig. 4. Normalized total yellow light transmission with time for encapsulant

samples subjected to externally applied thermal stress. The parameters (a, b)

specify fitted curves of the form y a expbx.

Fig. 5. Normalized total light transmission with time for encapsulant samples

subjected to short-wavelength optical radiation. The parameters (a1; b1 ,) and (a2; b2)specify fitted curves of the form y a1 expb1x a2 expb2x.

8 P. Appaiah et al./ Optical Materials 46 (2015) 611


4/6

3.3. Effect of varying externally applied thermal stress on a fixed

concentration of phosphor mixed in epoxy encapsulant

To investigate the effect of varying levels of externally applied

thermal stress on light transmission degradation, three encapsu-

lant samples were subjected to temperatures maintained at

73C, 76C and 89 C. The two encapsulant samples had phosphor

mixed in epoxy to a concentration of 20 mg/cm

2

. The normalizedlight transmission over time is illustrated inFig. 6.

The light transmission degradation of the encapsulant samples

increased as the externally applied thermal stress was increased.

3.4. Effect of varying short-wavelength optical radiation on a fixed

concentration of phosphor mixed in epoxy encapsulant

In order to understand the effect of short-wavelength optical

irradiance on encapsulant samples, two sets of samples with a

fixed concentration of phosphor mixed in epoxy at 10 mg/cm2

and 20 mg/cm2, respectively, were subjected to different short-

wavelength optical radiation levels. The irradiance values on the

encapsulant samples were 0.28, 0.50, and 0.57 W/cm2. The results

from the experiment are illustrated in Figs. 7 and 8.

The light transmission degradation increased as the short-

wavelength optical irradiance was increased. The 20 mg/cm2 phos-

phor concentration reiterated the light transmission degradation

trend that was observed with the 10 mg/cm2 phosphor concentra-

tion encapsulant samples.

4. Discussion

4.1. Effect of external thermal stress on phosphor embedded

encapsulant

The results for varying the phosphor concentration under exter-

nally applied thermal stress showed that the encapsulant samples

with phosphor had lower light transmission degradation compared

with the epoxy (no phosphor) encapsulant sample (Fig. 2). Since

YAG:Ce phosphor is thermally stable around 100 C, the cause of

degradation cannot be attributed to the quenching of the phosphor

[7]. Therefore, light transmission degradation is caused by epoxy

degradation under externally applied thermal stress. Past studies

have successfully shown that the addition of inorganic compounds

can reduce the degradation rate of epoxy [9,10]. These studies

found that the organic/inorganic hybrids were found to act as a

thermal insulator and a barrier to oxygen diffusion, therebymaking the composite more thermally stable.

An analysis of the blue and yellow portions of the total trans-

mitted light showed reduced light transmission degradation as

the phosphor concentration in the epoxy encapsulant was

increased (Figs. 3 and 4). The effect of encapsulant light transmis-

sion degradation due to short-wavelength optical radiation experi-

ments showed an increase in the rate of degradation as the

phosphor concentration was increased (Fig. 5). Past studies show

that pc-white LEDs (5 mm type) degrade faster than their blue

LED counterparts due to the presence of phosphor. One plausible

explanation is the increase in localized temperature of the epoxy

due to the heat generated in the phosphor particles during the

conversion of short-wavelength blue light into broadband white

light. This increase in heat is expected to increase with higherphosphor concentration in the epoxy encapsulant medium.

Fig. 6. Normalized total light transmission with time for encapsulant samples with

constant phosphor mixed with epoxy, subjected to varying levels of externally

applied thermal stress. The parameters (a, b) specify fitted curves of the formy a expbx.


phosphor mixed in epoxy at 10 mg/cm2, subjected to varying levels of short-

wavelength optical radiation. The parameters (a1; b1,) and (a2; b2) specify fitted

curves of the form y a1 expb1x a2 expb2x.


phosphor mixed in epoxy at 20 mg/cm2, subjected to varying levels of short-

wavelength optical radiation. The parameters (a1; b1,) and (a2; b2) specify fitted

curves of the form y a1 expb1x a2 expb2x.

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-


5/6

Encapsulant samples with higher phosphor concentrations will

result in a higher phosphor operating temperature compared with

encapsulant samples with a lower phosphor concentration,

increasing the degradation and reducing lifetime.

The effect of increased thermal stress on phosphor samples

(20 mg/cm2 of phosphor mixed in epoxy) revealed an increase in

encapsulant degradation (Fig. 6). Thermal stressing of the encapsu-

lant material leads to an alteration in the molecules of the polymer.The increase in temperature of a reaction increases the rate of the

reaction and deteriorates the polymer. The minimum energy

required for the initiation of a reaction is known as activation

energy. The sensitivity of a reaction rate to temperature can be

determined using the Arrhenius equation. The rate of a first-order

chemical reaction is given by:

kpd A exp EakT

where kpd; is the rate constant, A is the constant related to the

probability of the reaction occurring at any given temperature, Eais the activation energy measured in eV, k is the Boltzmann con-

stant, and T is the absolute temperature. The activation energy of

the encapsulant samples with phosphor mixed in epoxy subjectedto externally applied thermal stress was evaluated using their total

light transmission results. The calculation of activation energy was

carried out by manipulating the reaction rate equation stated above

to read:

lnkpd Ea1

kT

lnA

Fig. 9shows the variation of lnkpd versus 1=kT while the gra-

dient of this linear relation gives the activation energy of the

degradation.

Calculated activation energy for phosphor mixed in epoxy

encapsulant samples from the present study and epoxy resin

activation energy from past research [11,12] are reported in

Table 2.The activation energy of the encapsulant material increased as

the phosphor concentration increased (Fig. 9 and Table 2). This

suggests that the addition of phosphor reduces the thermal degra-

dation of the epoxy encapsulant. Therefore, presuming that the

reduction of light transmission is caused by epoxy degradation,

since the amount of epoxy in each encapsulant sample was the

same and the thickness of all encapsulant samples was within

0.180.26 mm, the activation energy of the encapsulant samples

should have been the same. The increase in normalized light

transmission and activation energy as the phosphor concentration

was increased therefore can be attributed to the degrading epoxy

(i.e., yellowing of the encapsulant) preferentially transmitting

phosphor-converted yellow light rather than absorbing more of

the blue light emitting from the LEDs.

Under thermal stress, higher activation energy is required for

samples with greater phosphor concentration than for clear epoxy

samples. This shows that the addition of phosphor in the samples

improves the thermal stability of the epoxy/phosphor composite.

4.2. Effect of short wavelength radiation on phosphor embedded

encapsulant

The light transmission in encapsulant samples with phosphor

mixed in epoxy reduced faster as the irradiance levels were

increased. In addition, the increase in the phosphor concentration

in the encapsulant further decreased the light transmission. Past

studies have also observed an increase in temperature of the

encapsulant samples with mixed in phosphor when irradiated with

short-wavelength optical radiation[7]. To assess the temperature

of encapsulant samples when subjected to short-wavelength opti-

cal radiation, two identical encapsulant samples were prepared

with 10 mg/cm2 phosphor mixed in epoxy. These samples were

subjected to the same 0.57 W/cm2 irradiance with one sample

being used for light transmission measurements while the other

encapsulant sample was attached with a thermocouple to measure

the in-situ operating temperature of the encapsulant. The

thermocouple was shielded from optical radiation in a similar

manner to the thermocouple shielding method proposed in past

research[13]. Fig. 10 illustrates the normalized total light trans-

mission and the encapsulant temperature as a function of time.

During the first 100 h of the experiment, a slight reduction in

light transmission from unity at the start of the test to 0.88 at

100 h was observed. This corresponded with a slight increase in

encapsulant temperature from 93C at initiation to 100 C at

100 h. This increase in temperature was caused by the initial epoxy

Fig. 9. Reaction rate variation with temperature of encapsulant samples. Here, mrepresents the slope of the line.

Table 2

Activation energy comparison of encapsulant samples with phosphor mixed in epoxy

and reported data for epoxy resin.

Epoxy

(no phosphor)

10 mg/cm2

phosphor

in epoxy

20 mg/cm2

phosphor

in epoxy

Activation energy,

Ea (eV)

0.180.50

[11,12]

0.53 0.75

Fig. 10. Total light transmission and temperature with time for an encapsulant

sample with 10 mg/cm2 phosphor mixed in epoxy. Inset images show the yellowingof the encapsulant sample at 0, 150, and 200 h.

10 P. Appaiah et al./ Optical Materials 46 (2015) 611


6/6

degradation resulting from the heat generated by the phosphor

during down-conversion of the short-wavelength radiation emit-

ted by the LED. The inset images inFig. 10show the yellowing of

the encapsulant sample. The initial reduction in light transmission

can be correlated to the slight yellowing of the encapsulant.

Approximately 150 h into the test, the temperature increased

rapidly along with a rapid reduction in light transmission. This

was caused by the epoxy absorbing nearly all the short-wavelengthoptical radiation emitted by the LED, contributing toward the heat

generation, which in turn increased the epoxy temperature. This

high temperature in epoxy caused further degradation, which

increased the absorption of optical radiation. The increase in

phosphor operating temperature reduced the phosphor down-

conversion efficiency, generating more heat and in turn degrading

the epoxy. These effects of heat generated in the down-conversion

process and yellowing epoxy led to rapid epoxy degradation, as the

inset images illustrate inFig. 10.

5. Conclusion

The present study was conducted to understand the LED encap-

sulant degradation caused by externally applied thermal stress and

short-wavelength optical radiation.

The thermal stress caused the encapsulant material to degrade

and at increased levels caused degradation rates to increase

further. The addition of phosphor as a whole did not change the

degradation rate of the encapsulant, as the main cause of degrada-

tion was due to epoxy yellowing and not due to phosphor degrada-

tion. The analysis conducted on the blue light emitted by the LEDs

and the phosphor-converted yellow light indicated that the

degraded epoxy acted as a blue light filter, which reduced the blue

light transmission compared with the yellow light. The activation

energy based on the degradation rates for the phosphor mixed in

epoxy samples revealed an increase in activation energy as the

phosphor concentration was increased in the encapsulant sample.

The reduction in degradation can be attributed to the blue light fil-

tering of the yellowing epoxy, as discussed.

The short-wavelength optical radiation degraded the encapsu-

lant samples and as the phosphor concentration was increased,

the degradation increased. Increases in the level of optical radia-

tion on the encapsulant samples further increased the degradation.

The degradation of encapsulant is related to the operating tem-

perature of the phosphor and as the phosphor temperature

increases, the conversion efficiency of the phosphor deteriorates

further, increasing the heat generated in the conversion process

leading to a temperature increase of the encapsulant, as discussed

earlier.

Curve-fitting seems to indicate there are at least two rate

reactions that govern the degradation of the encapsulant when

exposed to short-wavelength radiation, which is different from

the degradation rate when subjected to externally applied thermalstress at a similar encapsulant temperature. Further investigation

is necessary to obtain encapsulant temperature profiles/variations

as the encapsulant samples are exposed to short-wavelength radia-

tion to verify these observations.

Analysis of the data and the results revealed that the main cause

of the degradation is due to the degradation of the epoxy. The tem-

perature of the sample determines the degradation rate, indepen-

dent of how the temperature was increased, the ambient heat, or

short-wavelength radiation.

This experiment does have some limitations. The results from

this experiment are specific to the epoxy used in the study and

may not be generalized for all epoxies. The results might vary if

the epoxies have stabilizers that reduce the sensitivity of epoxy

to thermal stress or short-wavelength radiation. The epoxy

samples used for this study were bought from one manufacturer.

The results also are specific to the YAG:Ce phosphor used in this

study and might vary with different phosphors from differentmanufacturers. The results might vary when using phosphors that

are sensitive to heat. Additionally, it is common to mix a red phos-

phor into YAG:Ce to improve the color-rendering properties of the

white light [14]. In such cases, the encapsulant degradation rate

will be influenced more by the less efficient phosphor.

Role of the funding source

This research received no specific grant from any funding

agency in the public, commercial, or not-for-profit sectors.

Acknowledgments

We would like to thank Bridgelux and Cree for providing the

components for the experiments. We also thank the Lighting

Research Center staff, especially Martin Overington, Jean Paul

Freyssinier, Dr. John Bullough, and Andrew Bierman for their

valuable input. Jennifer Taylor is thanked for her invaluable

contribution towards preparing this manuscript.

References

[1] D.L. Barton, M. Osinski, P. Perlin, C.J. Helms, N.H. Berg, Life tests and failuremechanisms of GaN/AlGaN/InGaN light emitting diodes, in: 35th Annual Proc.IEEE Int., 1998, pp. 276281.

[2] A. Welsh Jr., L. Fu, W.W. So, H.B. Yuan, Die-attach epoxy reliability of InGaNLEDs, solid state lighting II, Proc. SPIE 4776 (2002) 6873.

[3] N. Narendran, N. Maliyagoda, A. Bierman, R. Pysar, M. Overington,Characterizing white LEDs for general illumination applications, Proc. SPIE3938 (2000) 240248.

[4] G. Meneghesso, S. Levada, E. Zanoni, G. Scamarico, G. Mura, S. Podda, M. Vanzi,S. Du, I. Eliashevich, Reliability of visible GaN LEDs in plastic package,Microelectron. Reliab. 43 (2003) 17371742.

[5] N. Narendran, Y. Gu, J.P. Freyssinier, H. Yu, L. Deng, Solid-state lighting: failureanalysis of white LEDs, J. Crystal Growth 268 (2004) 449456.

[6] T. Tamura, T. Setomoto, T. Taguchi, Illumination characteristics of lightingarray using 10 candela-class whiteLEDs under AC 100 V operation, J. Lumin.87(2000) 11801182.

[7] M. Meneghini, M. Dal Lago, N. Trivellin, G. Meneghesso, E. Zanoni, Thermallyactivated degradation of remote phosphors for application in LED lighting, IEEETrans. Dev. Mater. Rel. 13 (2013) 316318.

[8] Y. Zhu, Investigation of the Optical Properties of YAG:Ce Phosphor at DifferentWavelengths, M.S. thesis, Dept. Arch., Rensselaer Polytechnic Inst., Troy, NY,2006.

[9] C.L. Chiang, R.C. Chang, Y.C. Chiu, Thermal stability and degradation kinetics ofnovel organic/inorganic epoxy hybrid containing nitrogen /silicon/phosphorusby solgel method, Thermochim. Acta 453 (2007) 97104.

[10] C.L. Chiang, C.C.M. Ma, Synthesis, characterization and thermal properties of

novel epoxy containing silicon and phosphorus nanocomposites by solgelmethod, Eur. Poly. J. 38 (2002) 22192224.[11] M. Meneghini, L. Trevisanello, C. Sanna, G. Mura, M. Vanzi, G. Meneghesso, E.

Zanoni, High temperature electro-optical degradation of InGaN/GaN HBLEDs,Microelectron. Reliab. 47 (2007) 16251629.

[12] S. Ishizaki, H. Kimura, M. Sugimoto, Lifetime estimation of high power whiteLEDs, Illum. Eng. Inst. Jpn. 3 (2007) 1118.

[13] I.U. Perera, N. Narendran, Y. Liu, Accurate measurement of LED lens surfacetemperature, Proc. SPIE 8835 (2013) 883506.

[14] Y. Zhu, N. Narendran, Investigation of remote-phosphor white light-emittingdiodes with multi-phosphor layers, Jap. J. Appl. Phys. 49 (10R) (2010) 100203 .

http://refhub.elsevier.com/S0925-3467(15)00176-7/h0010http://refhub.elsevier.com/S0925-3467(15)00176-7/h0010http://refhub.elsevier.com/S0925-3467(15)00176-7/h0015http://refhub.elsevier.com/S0925-3467(15)00176-7/h0015http://refhub.elsevier.com/S0925-3467(15)00176-7/h0015http://refhub.elsevier.com/S0925-3467(15)00176-7/h0020http://refhub.elsevier.com/S0925-3467(15)00176-7/h0020http://refhub.elsevier.com/S0925-3467(15)00176-7/h0020http://refhub.elsevier.com/S0925-3467(15)00176-7/h0025http://refhub.elsevier.com/S0925-3467(15)00176-7/h0025http://refhub.elsevier.com/S0925-3467(15)00176-7/h0030http://refhub.elsevier.com/S0925-3467(15)00176-7/h0030http://refhub.elsevier.com/S0925-3467(15)00176-7/h0030http://refhub.elsevier.com/S0925-3467(15)00176-7/h0035http://refhub.elsevier.com/S0925-3467(15)00176-7/h0035http://refhub.elsevier.com/S0925-3467(15)00176-7/h0035http://refhub.elsevier.com/S0925-3467(15)00176-7/h0045http://refhub.elsevier.com/S0925-3467(15)00176-7/h0045http://refhub.elsevier.com/S0925-3467(15)00176-7/h0045http://refhub.elsevier.com/S0925-3467(15)00176-7/h0050http://refhub.elsevier.com/S0925-3467(15)00176-7/h0050http://refhub.elsevier.com/S0925-3467(15)00176-7/h0050http://refhub.elsevier.com/S0925-3467(15)00176-7/h0055http://refhub.elsevier.com/S0925-3467(15)00176-7/h0055http://refhub.elsevier.com/S0925-3467(15)00176-7/h0055http://refhub.elsevier.com/S0925-3467(15)00176-7/h0060http://refhub.elsevier.com/S0925-3467(15)00176-7/h0060http://refhub.elsevier.com/S0925-3467(15)00176-7/h0065http://refhub.elsevier.com/S0925-3467(15)00176-7/h0065http://refhub.elsevier.com/S0925-3467(15)00176-7/h0070http://refhub.elsevier.com/S0925-3467(15)00176-7/h0070http://refhub.elsevier.com/S0925-3467(15)00176-7/h0070http://refhub.elsevier.com/S0925-3467(15)00176-7/h0070http://refhub.elsevier.com/S0925-3467(15)00176-7/h0065http://refhub.elsevier.com/S0925-3467(15)00176-7/h0065http://refhub.elsevier.com/S0925-3467(15)00176-7/h0060http://refhub.elsevier.com/S0925-3467(15)00176-7/h0060http://refhub.elsevier.com/S0925-3467(15)00176-7/h0055http://refhub.elsevier.com/S0925-3467(15)00176-7/h0055http://refhub.elsevier.com/S0925-3467(15)00176-7/h0055http://refhub.elsevier.com/S0925-3467(15)00176-7/h0050http://refhub.elsevier.com/S0925-3467(15)00176-7/h0050http://refhub.elsevier.com/S0925-3467(15)00176-7/h0050http://refhub.elsevier.com/S0925-3467(15)00176-7/h0045http://refhub.elsevier.com/S0925-3467(15)00176-7/h0045http://refhub.elsevier.com/S0925-3467(15)00176-7/h0045http://refhub.elsevier.com/S0925-3467(15)00176-7/h0035http://refhub.elsevier.com/S0925-3467(15)00176-7/h0035http://refhub.elsevier.com/S0925-3467(15)00176-7/h0035http://refhub.elsevier.com/S0925-3467(15)00176-7/h0030http://refhub.elsevier.com/S0925-3467(15)00176-7/h0030http://refhub.elsevier.com/S0925-3467(15)00176-7/h0030http://refhub.elsevier.com/S0925-3467(15)00176-7/h0030http://refhub.elsevier.com/S0925-3467(15)00176-7/h0025http://refhub.elsevier.com/S0925-3467(15)00176-7/h0025http://refhub.elsevier.com/S0925-3467(15)00176-7/h0020http://refhub.elsevier.com/S0925-3467(15)00176-7/h0020http://refhub.elsevier.com/S0925-3467(15)00176-7/h0020http://refhub.elsevier.com/S0925-3467(15)00176-7/h0015http://refhub.elsevier.com/S0925-3467(15)00176-7/h0015http://refhub.elsevier.com/S0925-3467(15)00176-7/h0015http://refhub.elsevier.com/S0925-3467(15)00176-7/h0010http://refhub.elsevier.com/S0925-3467(15)00176-7/h0010http://-/?-

Documents

Effect of Thermal Stress and Short-wavelength Visible Radiation_2015