Instrument to identify defects and impurities in wide band gap semiconductors via excited states

The researchers propose to develop a new instrument to measure the electrical properties of defects in wide band gap semiconductors. The most important wide gap materials at the present time are compounds made from metals from group III of the periodic table and nitrogen such as GaN and InGaN. These are referred to as III-N materials. They are used to make low energy lighting, LASERs and efficient RF and high power transistors. Today's generation of these devices does not function as well as would seem possible from the properties of the materials and at the present time functionality and performance is limited. LEDs and LASERs (other than blue) are less efficient at generating light than expected and in the case of the transistors several aspects of performance are less than desirable. This is due, at least in part, to the presence of defects in the component materials and devices. These defects are difficult to identify using existing techniques. They may be due to impurities or imperfections in the crystal lattice resulting from the crystal growth or introduced during the device manufacturing process.

The research group at Manchester have over thirty years of experience in solving defect problems in other materials such as GaAs and Si. Devices made from these materials have revolutionized society through mass produced electronics and communication technologies. The ability to measure, understand and control defects, particularly electrically active defects, has played a major role in this immense technological achievement and instruments devised, developed and licensed from Manchester have played a role in this. In the case of the III-N materials detecting defects and quantifying their properties is much more difficult and no technique exists at the moment which can look at all the band gap and quantify the recombination paths and trapping centres which degrade III-N devices.

The defining feature of the new instrument is that it uses sub-band gap light from tunable semiconductor LASERs to create excited states of the defects. Carriers are then thermally ionised to the semiconductor bands from the excited states. Because the optical excitation stimulates a bound to bound transition, a fine line spectrum can be obtained which is a fingerprint of the defect species and its location in the lattice. In the case of many defects being present, the emission rates will be separated using our existing Laplace DLTS processing. Recombination and trapping parameters can be obtained using the methodologies developed for variants of DLTS and LDLTS. One of our project partners (Santa Barbara University in California) will undertake theoretical studies aimed at associating the excited state spectra with chemical species and/or the structure of the defect with a view to generalized identification rather than using correlation with previously obtained spectra.

The instrument development is complementary to the EPSRC contracts currently in progress at many UK universities for the development of III-N materials and devices. In the initial phase of the instrument development, collaborations with consortia led by Cambridge and Glasgow for testing materials and power devices have been negotiated. This will be broadened to embrace other groups as the project progresses. Industrial interest in the project has resulted in strong support from five companies in the field of manufacture of III-N materials, LEDs, GaN power devices and instrumentation. Four of these are UK based. The potential benefits to society of a successful completion of this contract are enormous in facilitating greater improvements in domestic lighting and enabling new applications of the III-N materials to be developed for example in efficient short wavelength UV GaN LEDs. These could be used in cheap low maintenance drinking water sterilization, a pressing concern in the developing world.

The instrument proposed is to quantify and identify trapping and recombination centres in wide band gap semiconductors the most important of which at the present time is the GaN family of semiconductors. Consequently GaN is the material which will be used to develop the instrument and to undertake research which will demonstrate the instrument's capabilities. A direct impact of successful completion of the project will be the manufacture of the instrument under licence and its availability to the research and manufacturing community. How this might occur is detailed in the pathways to impact.

Although researchers and manufacturers of all wide gap semiconductors will benefit from the research, the major impact in the short term will be related to those researchers, materials producers and device manufacturers involved in the GaN family of materials. These fall into two main areas, light emitting diodes (LED) and transistors for power control and for communications. In the case of LEDs these already have a major market in high efficiency lighting. The instrument will impact in this sector by facilitating process control and further efficiency improvements. This is by providing information on electrically active defects in the semiconductor. At the moment techniques to do this are inadequate or in some cases non-existent. Green and UV LED are very much less efficient than the blue devices used with a phosphor to produce white light for general lighting. The efficiency limiting factor in the green and UV is non-radiative recombination via the deep states which this instrument will characterise and identify. The green LED will be used in more efficient lighting of improved spectral quality while efficient UV LEDs working at short wavelengths (~ 260nm) would provide a major impact to developing communities in enabling cheap and reliable water sterilisation on a domestic or neighbourhood scale with a low maintenance requirements. Being able to identify the non-radiative recombination paths would be a major step forward in enabling these devices, central to the application to be manufactured.

For the case of the power and RF transistors these have the potential to provide more efficient switching and amplification than is available today. However the devices suffer from deficiencies which limit their adoption into some major applications. These are degradation, current collapse and transient shift of characteristics which are due, at least in part, to deep states. The new instrument is expected to provide a way to measure the defect trapping parameter and identify their origins. This will impact strongly on researchers and manufacturers engaged in this field

The move towards efficiency improvement which this instrument will contribute to is of great importance to society. Fifteen percent of the world's electricity is used in lighting so major improvements in efficiency and widespread adoption of LED lighting will reduce carbon emissions. Similar arguments apply to the use of GaN power and RF devices which have the potential to provide even greater energy savings.