Service life of ultra-bright LEDs. Reasons for failure

Introduction to the concepts of "failures" and "service life" of LEDs. Many LED manufacturers claim a service life of up to 100,000 hours of continuous operation. However, this figure is misleading and largely depends on the quality of the product, the conditions of its use, and the criteria for assessing the reliability of the LEDs.

Even when using high-quality components, a decrease in luminous flux is inevitable - this is due to many factors, such as heat dissipation conditions, ambient temperature and ventilation, humidity and other parameters. Operating conditions, such as the magnitude and instability of the current, can also significantly reduce the service life. There are currently no standards defining the lifetime and reliability criteria for LEDs, although there are proposals from authoritative organizations to consider the lifetime as the time during which the luminous flux degrades to some value (30%) from the initial value.
Some companies have chosen to develop their own methods for predicting lifetime and reliability based on customer data, but the limited production volumes of most suppliers prevent the implementation of this approach.

While in most cases the performance of LEDs deteriorates gradually, sudden failures have also been observed due to dislocation growth from the periphery of the active region, destruction of the p-n-junction, dislocation growth from the oxidized end face or the intermediate region separating the end face and the dielectric coating, and catastrophic optical damage.

Secondly, customers working with LEDs have long realized that their reliability, especially in terms of degradation rate, often depends on the component supplier. Knowing these two hard limits requires that life tests be developed based on a fundamental understanding of failure mechanisms.

Users working with LEDs typically define the output power level at which the entire system will fail and then use physical models to predict the mean time between failures. A clear definition of failure is the most critical point, and most manufacturers and users have their own opinion on when an optoelectronic device can be considered to have failed. One method of determining failure is to clamp the current and monitor the output power of the device, considering the device to be inoperative when the output power drops below a certain level (usually 20% to 50%) of the initial value. Another method is based on monitoring the decline in the device's output power and compensating for it by increasing the drive current. When the drive current reaches a certain relative value (30%), the device is considered to have failed. Certain failure mechanisms and defects can also initiate LED failure.

 

Reliability engineers should not focus solely on the effects of temperature and current density, as this approach may lead to incorrect product selection. Active Region Degradation Light emission in an LED occurs as a result of recombination of injected carriers in the active region. Nucleation and growth of dislocations, as well as precipitation of nodal atoms, lead to degradation of the interior of this region.

These processes can only occur in the presence of a crystal structure defect; high injected current density, heating due to injected and leakage current, and emitted light accelerate the development of the defect. Material selection is also important, as the AlGaAs/GaAs system is much more sensitive to this failure mechanism than the InGaAs(P)/InP system. The InGaN/GaN system (for blue and green emission) is insensitive to defects such as those described above.

The active regions of these diodes may contain simple p-n junctions, embedded heterostructures, and multiple quantum wells. At the interfaces of such structures, changes in the chemical composition or even lattice parameters are inevitable. At high injection levels, chemical components can migrate by electromigration to other regions. Structural changes generate crystalline defects such as dislocations and point defects, which act as non-radiating centers that prevent natural radiative recombination and, as a result, generate additional heat inside the active layer. Electrode degradation Electrode degradation in LEDs mainly occurs at the p-region electrode (usually the device consists of an n-type substrate, and the p-region electrode is formed near the active region of the device).

The main reason for electrode degradation is the diffusion of metal into the inner region (the so-called peripheral diffusion) of the semiconductor. Diffusion increases with increasing injected current and temperature. Unfortunately, it is difficult to choose the rightohmic contact material to the p-region of InGaN/GaN systems is quite difficult due to the large band gap of p-type GaN. Because the electrode must have a smaller coefficient of mutual diffusion of the constituents, engineers sometimes use a barrier layer to suppress the effects of electromigration. Current saturation problems are more serious in high-power LEDs.

 

To solve these problems, engineers need to optimize the electrode design and the vertical component of the electric current. Electrodes made of some materials, such as transparent conductive indium tin oxide (ITO) or reflective metals (silver), are susceptible to problems such as electromigration and thermal instability.

Working edge degradation is a serious problem for visible-light-emitting AlGaAs/GaAs LEDs, but is not typical for InGaAsP LEDs. Oxidation by photochemical reactions leads to increased threshold current values ​​and, accordingly, to a decrease in the lifetime of the LED.

Another type of working edge failure is the so-called catastrophic optical defect (COD) - when the amount of light energy exceeds a certain level and the working edge begins to melt. Failure of optoelectronic devices, normally resistant to degradation of the working edge, can be initiated by damage during processing, foreign contamination and material defects. Thermal degradation The amount of heat generated by LEDs during operation requires their mounting on a radiator or heat-absorbing substrate, often using solder. If caverns in the solder create conditions for insufficient heat dissipation, the resulting hot spots lead to thermal degradation and failure. Thermal degradation due to solder cavities often dominates the first 10,000 hours of operation in LEDs.

Solder cavities can be caused by improper processing conditions or by metal diffusion at the interface (Kirkendall cavities). Cavity formation can also occur due to electromigration. When a sufficiently high current flows in a metal, vacancies and metal ions migrate to opposite poles, resulting in the formation of cavities (vacancies), crystals, tubercles, and whiskers. Whisker growth, which can be initiated by internal stress, temperature, humidity, and material properties, typically occurs at the solder-heatsink interface and can lead to short circuits. Semiconductors are susceptible to defects caused by electrostatic discharge (ESD). ESD failure modes can include sudden failure, parametric shifts, or internal damage leading to degradation during subsequent operation. According to current regulations, LEDs must have an ESD sensitivity greater than 100 V in a human body model test. Overload breakdown and ESD are significant problems for LEDs. Sometimes, designers use a Zener diode or a Schottky barrier to achieve a certain ESD class.

 

Most commercial InGaN/GaN LEDs are formed on sapphire substrates that have no electrical conductivity. This leads to the appearance of a residual electrical charge in the device, which makes it more susceptible to damage caused by ESD and overload. Thermal fatigue and short circuit
The difference in the coefficient of thermal expansion between the bonded parts and the solder leads to mechanical stresses during the manufacturing stage associated with thermal cycling, which can cause delamination in the bonded parts. When a power device is subjected to cyclic loading, the behavior of devices made, for example, with hard solder and soft solder may differ. Thermal fatigue is usually observed in devices made with soft solder, while devices made with hard solder are stable under thermal cycling.

Sometimes, inappropriate solder and process control can lead to a short circuit in the device. Due to its relatively high wettability, tin-based solder can overflow the edge of the contact pad and form a shorted trace. Failures associated with the assembly in the package can be caused by the encapsulant, electrode leads and phosphorus. Thermal stresses in the encapsulant are the most common cause of failure in LEDs. If - due to electrical overload or high ambient temperature - the case temperature reaches the transition temperature of the glass filler of the encapsulant (Tg), the resin begins to expand rapidly. The difference in the thermal expansion coefficient of the LED internal components can cause mechanical damage. At very low temperatures, cracking of the epoxy composition can occur, andfrom which the lenses are made.

High temperatures caused by internal heating and non-radiative recombination, reaching 150 C, cause the epoxy composition to turn yellow, which changes the output optical power or the color of the emitted light. If the refractive index of the sealant does not match the refractive index of the semiconductor material, the induced light remains in the semiconductor, resulting in an additional source of heat. Overheating of the epoxy composition may cause a break or separation of the electrode lead and a decrease in the strength of the connection between the crystal and the substrate. These problems, in turn, can lead to delamination of the crystal and the epoxy composition.

 

Mechanical stresses caused by lead conductors are another reason why a break may occur in the device. Failure to apply the correct pressure, position, and direction during soldering can result in mechanical stress at normal operating temperatures and bending of the leads dangerously close to the LED die.

Most white LEDs use yellow or red/green phosphors, which are susceptible to thermal degradation. When designers mix two or more different phosphors, the components must have comparable lifetimes and degradation patterns to ensure color saturation. The color temperature and color purity of the phosphor also degrade over time.