Friday, 10 August 2007

Summer Internship





Recently, I did a 6-weeks summer internship at Department of Chemical Engineering and Advanced Materials, University of Newcastle Upon Tyne. The internship started on 25th June until 3rd August. University of Newcastle is located at the heart of Newcastle city centre and about 5 mins walk to St. James Park. It is a nice university but I think it is a bit smaller than Sheffield Uni. 75 % of the university buildings are concentrated at one place and the others are scattered around the city centre.



There are 4 other students who also did the internship with me and all of them are studying at different universities. They come from Imperial College London (Chinese guy from Brunei), Cambridge University (English girl), University of Birmingham (Latin chick from Venezuela) and University of Queen Belfast, Ireland (Irish girl). All of us did different research.



I have conducted a research entitled, "Reflectance Spectroscopy for Monitoring Pharmaceutical Powder Mixtures". I was under the supervision of EPSRC researcher, Dr. Suresh Thennadil. This research has a balance aspects between experimental and theory. Basically, what I did is to figure out how to use reflectance spectroscopy to monitor pharmaceutical powder mixtures. Reflectance spectroscopy is a device which operate based on light reflection. If we illuminate a sample with light, a fraction of it with be reflected, absorbed, transmitted and scattered. Based on the reflected light, we can get chemical and physical information of the sample such as particle size and so on.

One of the main process in producing medicine is powder blending. Normally, manufacturer spend sufficiently long period of time just to make sure that all ingredients in the powder mixtures are evenly distributed. Significant time saving can be realised if the blending process can be stopped at an optimum time. So, if we can estimate the concentration of the components in the mixture at different regions of the blending unit, we can make a decision about the uniformity of the components in the mixture. Reflectance spectroscopy has a promising future to become an effective tool to monitor pharmaceutical powder mixtures because it of its ability to provide fast and non-destructive measurements without sample preparation.

Generally, there are 2 components in pharmaceutical powder mixture which is active ingredient and filler. Active ingredient is the one which has the ability to cure diseases and usually in small amount. The rest are just filler. In this research, I have used salicylic acid (usually found in aspirin) as the active ingredient and lactose monohydrate (usually found in milk) as the filler.


The first thing that I have been told to do is error check. I need to know what is the error due to measuring instrument, packaging of the powder mixture and blending process. I used 2 methods to analyse the reflectance spectra which is absorbance and Kubelka-Munk theory. There are several other methods (which gives better results) that can be used but due to time constraint, absorbance and Kubelka-Munk were chosen. I also built a calibration model to estimate the concentration of the active ingredient in the powder mixture. Then, I need to evaluate which method gives better prediction and at what wavelength the reflectance spectroscopy works best to monitor the powder mixture.

At the end of the internship, we were asked to make a report and a presentation. The audience of the presentation was lecturers and almost all PhD students at that department. I think I gave a good presentation but not excellent one because there are weaknesses here and there. The worst part of the presentation is during the Q & A session. I couldn't answer 2 questions but fortunately Raimundas (PhD student who also works in spectroscopy field and under Suresh supervision) helped me. One of the question was

"can we combine the absorbance and Kubelka-Munk techniques to give better result? rather than using them separately and then decide which gives better results".

The answer is yes, we can combine them but with added complexity. Because our aim is to make the error as lowest as possible. This means to make the difference between the estimated concentration and the real concentration as smallest as possible. One of the method that can be used to combine both techniques is PLS. The other question was,

"why do you picked only one wavelength? "

This is because it is cheaper to build a spectroscopy that works only at one wavelength. Spectroscopy that I used in the lab is very expensive because it works at wide range of wavelength, from UV-Vis-NIR region. That is why we need to pick the best wavelength which gives the lowest error.

I think all the questions are brilliant especially the first one. The person who asked that question must have a very critical thinking. This is one of the most important quality that every researcher need to have.


In my opinion, I have gained a lot of benefits from this internship even though I am not a chemical engineering student. In fact, spectroscopy is also used by physicist, semiconductor people etc. Now, I got a clear view of what research is all about. Besides, I also obtain the necessary skills to do research. My view towards supervisor also has changed. Before this, I think supervisor as the person that assess me. This is why sometimes don't ask questions to my supervisor because I am afraid that he might feel that I am to dependent on him and it might influence him when assessing my performance. Now, I think supervisor as a person who will guide me and a person who i can discuss about my project.

I also has improve my confidence especially when working with high tech devices. When I did my 3rd year project, I don't have enough confidence when working with the laser, cryostat etc because I am afraid that I will damage it. They are very expensive equipments. But now, everything has changed. I am full of confidence when working with the reflectance spectroscopy alone in the lab because I know what I am doing. Besides, I also learned how to use Matlab. Actually, it is not that hard.

Lastly, I would like to thank Suresh, Carlo, Elitsa, Maria and Raimundas for helping me during my 6 weeks internship. Thank you.


No comments: