Jan 16

COS Reviews; Junction and carrier temperature measurements in deep-ultraviolet light-emitting diodes using three different methods

The Cranfield Optics Society (COS) was setup to enable its members to develop the skills and attributes required by researchers. To that end we regularly meet and discuss papers that are relevant to the work we do, and analyse them as a group. The papers are discussed on their clarity, their relevance to our work and their accuracy. Presenting research in the context of a paper is tricky; you strive to make the paper as accurate to the research you undertook while making it as brief as possible. You want it to be unambiguous and to allow others to replicate your work successfully, and therefore validate the usefulness of your research. To know how you should present your work, you need to understand how others have presented theirs and what you found beneficial in that format, and what needs to be improved.

The first paper that we are describing on the blog is; “Junction and carrier temperature measurements in deep-ultraviolet light-emitting diodes using three different methods” by Y.Xi, J.-Q. Xi, Th.Gessmann, J.M. Shah, J.K. Kim, E. F. Schubert, A.J. Fischer, M.H. Crawford, K.H.A Bogart and A.A. Allerman. This paper was published in Applied Physics Letters 86 in 2005. We choose this paper due to its relevance to the research that takes place in our department. LEDs are widely used in gas sensing, however their spectral properties and intensity can vary with temperature. The photonics department is broad in the range of research it does, but much of the equipment used is the same from experiment to experiment.

LED photo

The structure of the paper follows the traditional pattern of papers; Introduction, Theory, Method, Results and Conclusion. The sections aren’t delineated in the paper but flow as one continuous series of statements. I found this made it difficult at first to absorb the information, but this was a problem not shared by other members of COS. The Introduction in this case gave the background history of ultraviolet LEDs and why junction and carrier temperature are important. The introduction was concise, informative, gave quantified values of the power and wavelength of diodes produced by other groups and named methods used by other groups to measure junction and carrier temperature. This isn’t my field of study and I found the introduction useful. It provided enough information for me to read the paper and have a grasp of the context while having enough specific points that I could look them up for more detailed information on the area. It also explained the importance of junction temperature in short, not enough to fully understand the scope of its importance. It can be assumed that a potential reader would find it important, otherwise they wouldn’t be reading the paper.

The theory section covered the mathematical derivation for the temperature coefficient of the forward voltage. This is required to understand the methods employed to measure the temperatures of the junction and carriers. The derivation flows naturally and produces a valid result that is directly applicable to the work that is done. It is here that we come to the major problem of the paper, lack of description in regards to errors. This section states the temperature coefficient of the forward voltage for GaN diodes that is derived from the equation and from experimentation. What it doesn’t state is the error inherent in both and states that they are in good agreement. The theoretically derived value is -1.76mV/K and the experimentally derived value is -2.3mV/K. The errors for each value are clearly different as they have different numbers of significant figures, that is the only things I can ascertain about the uncertainty.

The method is broken into three parts, one for each of the different approaches used to gauge the temperatures. The description of the method is concise and quite descriptive. The controls in the experiment are well described; including how the diodes were constructed to give some minor information on how to repeat the work. The graphs in results are clear and have all the required information except error bars. The temperature coefficient of the forward voltage is stated for one of the diodes and what its theoretical value should be (-5.8mV/K and -2.04mV/K respectively) but since there is no error stated for either, it is hard to grasp how bad the disparity between them is. There is a good description of why they think there is a disparity but it isn’t quantified. It seems harsh to repeat the point but only figure 3 has error bars on it and for only one line. It is below figure 3 that there is a discussion on the errors in the results, where two methods are found to be in good agreement due to the large errors in one of the measurements.

In the final part of the results they discuss another of their controls, they tried the experiment with different heat sinks. This is a quite a useful comparison to use since it covers the effects of thermal transport in the experiment. The conclusion covers the key points of the paper; the most accurate method of measuring temperature and its error. Overall the paper is well written, it does contain all the necessary information that is summated in the conclusion. It should also be reliable due to the number of controls they state they use to arrive at their conclusion.