Photoluminescence Spectroscopy (PL) and Cathodoluminescence Spectroscopy (CL)

When a solid is supplied with a certain type of energy it may emit photons, or undergo a process called luminescence. The luminescence process can be categorized according to the excitation source. Photoluminescence (PL) is due to photon excitation, chemiluminescence is due to energy supplied by a chemical reaction, thermoluminescence is due to energy supplied by heating, electroluminescence is due to excitation by application of an electric field, and cathodoluminescence (CL) is due to excitation by an electron source. The mechanisms leading to the emission of light in a solid are similar for the different types of excitation sources described above and can provide complimentary information [7, 8].

Photoluminescence (PL) depends on the nature of the optical excitation. The excitation energy selects the initial photoexcited state and governs the penetration depth of the incident light. The emission signal depends on the density of photoexcited electrons, and the intensity of the incident beam can be adjusted to control this parameter. When the type or quality of material under investigation varies spatially, the PL signal will change with excitation position. In addition, pulsed optical

excitation provides a powerful means for studying transient phenomena. Short laser pulses produce virtually instantaneous excited populations, after which the PL signal can be monitored to determine recombination rates.

PL is simple, versatile, and nondestructive. The instrumentation that is required for ordinary PL work is modest: an optical source and a spectrophotometer. A typical PL set-up is shown in Figure 2-11. Because the measurement does not rely on electrical excitation or detection, sample preparation is minimal. This feature makes PL particularly attractive for material systems having poor conductivity or undeveloped contact technology. Measuring the continuous wave PL intensity and spectrum is quick and straightforward. Instrumentation for time-resolved detection, such as single photon counting, can be expensive and complex. Even so, PL is one of the only techniques available for studying fast transient behavior in materials.

The advantages of PL analysis derive from the simplicity of optical measurements and the power to probe fundamental electronic properties. The chief drawback of PL analysis also follows from the reliance on optical techniques: the sample under investigation must emit light. Indirect-bandgap semiconductors, where the conduction band minimum is separated from the valence band maximum in momentum space, have inherently low PL efficiency. Nonradiative recombination tends to dominate the relaxation of excited populations in these materials. This problem can be augmented by poor surface quality, where rapid nonradiative events may occur. Nevertheless, once a PL signal is detected, it can be used to characterize both radiative and nonradiative mechanisms.

The choice of excitation is critical in PL measurement. The excitation energy and intensity will have profound effects on the PL signal. Although the excitation

conditions must be considered carefully, the strength of the PL technique relies heavily on the flexibility that these adjustable parameters provide. Because the absorption of most materials depends on energy, the penetration depth of the incident light will depend on the excitation wavelength. Hence, different excitation energies probe different regions of the sample. The excitation energy also selects the initial excited state in the experiment. Because lasers are monochromatic, intense, and readily focused, they are the instruments of choice for photoluminescence excitation.

A relatively inexpensive He-Cd or diode laser will often satisfy the basic requirement of light exceeding the bandgap energy. In more demanding experiments, the laser is chosen carefully to probe a particular depth or to excite a particular species.

Cathodoluminescence (CL) is a SEM-based technique that can be used for the characterization of semiconductor materials and devices. A typical CL set-up is shown in Figure 2-12. SEM-based and CL can provide information on the concentration and distribution of luminescent centers, distribution and density of electrically active defects, and electrical properties including minority carrier diffusion lengths and lifetimes. CL is the emission of light as the result of electron or “cathode-ray”

bombardment. The CL phenomenon was first reported in the middle of the 19th century during experiments on electrical discharges in evacuated glass tubes.

Luminescence was observed when cathode rays struck the glass tubes [7, 8]. The observation of luminescence due to cathode ray bombardment eventually led the J.J.

Thomson to the discovery of electron in 1897. Now, CL is widely used in cathode-ray tubebased instruments such as oscilloscopes, televisions, and electron microscope fluorescent screens.

The CL signal is generated by detecting photons, in the ultraviolet to infrared

range, that are emitted as a result of electronic transitions between the conduction band and valence band, between levels due to impurities and defects in the fundamental gap, or between impurity and defect levels and the valence band. The transition energies and probabilities can be affected by external perturbations, such as stress and electric fields. CL analysis performed in an electron microscope can be divided into spectroscopy and microscopy.

In CL microscopy, luminescence images or maps of areas of interest are obtained by a scanning or parallel beam. In panchromatic CL imaging the combined intensity of all CL wavelengths within the response of the detector are used to create the image.

In monochromatic CL imaging, the light is coupled into a monochromator and CL images can be created from a selected wavelength bandpass [7, 8].

In direct band gap semiconductor materials, the minimum of the conduction band and the maximum of the valence band occur at the same momentum value in an energy versus momentum plot. Momentum is conserved in direct band gap transitions and the transitions appear vertical on energy versus momentum plots. In these materials, the most likely radiative transitions are between the filled states of the conduction band minimum and the empty states of the valence band maximum. If the material is an indirect band-gap material, the maximum of the valence band and minimum of the conduction band do not occur at the same momentum value, and therefore phonon participation is required to conserve momentum. The recombination of electrons and holes results in the simultaneous emission of a photon and a phonon.

Since the probability of this process is lower than direct transitions, intrinsic emission is relatively weak compared to extrinsic luminescence in an indirect band gap material.

A simplified set of radiative transitions that lead to luminescence emission in

semiconductors containing impurities is shown in Figure 2-13 and a description of each process follows.

Process 1 produces intrinsic luminescence due to direct recombination between an electron in the conduction band and a hole in the valence band and results in the emission of a photon with energy close to that of the band gap. The recombination may occur from states close to the corresponding band edges, but the thermal distribution of carriers typically leads to a Gaussian shaped spectrum with the peak corresponding to the transition with the maximum electron and hole concentration [7, 8].

Process 2 is excition decay that is typically observable at lower temperatures. An excition is a bound electron-hole pair and excitionic states exist just below the conduction band. In most III-V compounds, recombination emission of the excitonic state produces photons of energies approximately equal to the band gap of the semiconductor; therefore this process can also be considered an intrinsic process. The intrinsic luminescence band is often referred to as the near-band-gap band because excitons and shallow recombination centers may contribute to the emission at room temperature [7, 8].

Process 3, 4, and 5 correspond to transitions that start or finish on localized states of impurities within the band gap, such as donors or acceptors. These transitions produce extrinsic luminescence. Transitions between deep donor and deep acceptor levels can lead to emission with photon energies significantly below the band gap.

Shallow donor or acceptor levels can be very close to the conduction band and valence bands. For example, silicon is a shallow donor in gallium nitride located approximately 20 meV below the conduction band [9]. Transitions that occur between

a shallow donor and acceptor states or between donor and acceptor states and the conduction band or valence bands may be difficult to distinguish from intrinsic luminescence. In these cases, CL measurements can be performed at cryogenic temperatures using liquid nitrogen or liquid helium as a cryogen. CL spectra can be sharpened into lines and series of lines corresponding to transitions between well-defined energy levels due to a reduction in the thermal excitation of carriers.

There is also an increase in the CL intensity as the temperature is lowered because radiative recombination becomes more favored as compared to the competing non-radiative recombination. An additional advantage of performing CL at cryogenic temperatures is the reduction of electron bombardment damage [7, 8].