Theory of IQE Measurement Method

We first note that the traditional method to define the IQE could be realized by figure 4.1.

Generally, we used to perform a temperature dependent PL measurement and investigate the PL intensity variation curves with temperature. It is a non-destructive method to understand

the internal material characteristics and quality. It provides useful information about emission wavelength, light intensity and other optical properties. During the experiment period, the intensity gradually reduced with the rising temperature. We could reasonably assume there are no non-radiative recombination at 0 K and the IQE approaches 1 which means one photon generated in the MQW while one laser photon was injected into the MQWs. In the meantime, we assume the ExE would not change with the increasing temperature. Therefore, the IQE could be defined as equation 4-1 by the traditional method:

int

( ) ( )

(0) T I T

η = I (4-1)

where I(0) stands for the PL intensity at 0K and I(T) implies the PL intensity under specific temperature. We note that this method ignores the influence of injected laser power density which is not valid in blue GaN-based LED material system. Hence, we quote reference [7] to give a new definition by PL method. We performed the power-dependent PL measurement at low temperature and room temperature and define it by observing the tendency of the curves.

This could be expressed as the following equation 4-2:

int

where IPL is the collected PL intensity; Iinj is the injected laser intensity; EPL is the measured light photon energy; Einj is injected photon energy and C is a constant which is related to carrier injection efficiency, light extraction efficiency and correction efficiency of PL.

While the injected laser power increased, the collected light intensity also increased.

Generally, we could collect higher PL intensity at low temperature than that in room temperature. Therefore, the collected intensity versus the laser power at low temperature and room temperature could be depicted as figure 4.2 (a). We then can calculate the collected photons to be divided by the injected laser photons, by equation 4-2. A maximum number would be observed at some power of low temperature curve. After normalization to the peak, it could be shown as figure 4.2(b). Finally, we could define the IQE at room temperature at different laser power.

PL measurement setup

For excitation power and temperature dependent PL measurement, the laser source used in this research is femto-second-pulse Ti:sapphire. The output laser wavelength could be

tuned to be around 760-820 nm by adjusting the internal resonant length and gain profile. A frequency doubler and tripler crystal is setup at the laser output optical path. Here, we adopted the doubler crystal to have a wavelength at around 380-410 nm could be obtained. It is specifically important in avoiding the absorption of GaN material itself and ensured most of the injected laser photons could excited the photons inside the MQWs. We note that this assumption essential while we predicted the injected carrier density. The tuned the 380-410 nm laser was then incident into the sample surface vertically. The sample was mounted in a closed-cycle He cryostate with a temperature controller which could be precisely controlled in the temperature region around 15 to 300 K. By the temperature controller with a set of wire heater, we could perform a serious of temperature-dependent experiments from 15 to 300 K.

The luminescence signal dispersed through a 0.55-meter monochromator was detected by the photomultiplier tube (PMT). Not only optical properties like emission wavelength, intensity and FWHM were acquired, a time-resolved photoluminescence (TRPL) experiments could be performed simultaneously by setting a beam splitter to reflect part of light into a monochromator. It helps us to obtain the carrier lifetime from this lighting material. To sum up, we are able to realize the whole lighting mechanism by this PL measurement system. The setup detail of temperature dependent PL and TRPL are shown in figure 4.3.

Sample preparation

There are several types of the PSS arrangement like strip [8] or island-like [9] structures.

Not only the effect of geometric light reflection issue was considered, the crystal quality after epi-process is also an important issue here. So, the designed patterns on sapphire substrate in this research are two dimensional hole-arrays arranged as hexagonal lattice with lattice constant of 7 μm and hole diameter of 3 μm. After the patterns were defined, the PSS samples were then undergone the ICP dry etching process to make the PSS. The SEM images of the fabricated PSS samples were shown in figure 4.4. The etched hole is 0.5-μm in depth and has triangular-shaped C-plane in the center, surrounded by three {1-102} R-plane facets. The fabrication details of PSS can be found elsewhere [9, 10]. The LED structure was then grown on planar substrate and PSS by low-pressure MOCVD. The LED structure consisted of a 30-nm-thick AlN nucleation layer, a 2-μm-thick Si-doped n-type GaN, and an unintentionally doped active layer with InxGa1-xN/GaN MQWs, and a 0.2-μm-thick Mg-doped p-type GaN.

The doping concentration of n- and p-type GaN was nominally 5 x 10¹⁸ and 1 x 10¹⁹ cm^-3, respectively. The MQWs layers comprised 16 periods of an In0.15Ga0.85N well (~2 nm) and a GaN barrier (~16 nm). On the purpose of comparison, a LED grown on flat sapphire with

identical epi-structures was prepared.