6. Practical Approach to Automated FTIR Microscopy Problem Solving
Kodak European Research and Development
Telephone: +44 (0) 181 424 5346
Fax: +44 (0) 181 424 3750
Our 10 year old infrared microscope needed replacement. After years of heavy use it had poor stability and needed a lot of care and attention, so it was decided to allow it to retire. Our requirements for a replacement microscope were:
Good stability (robust)
Easy to use.
It was essential to be able to use our existing data; we have been adding spectra to our user generated libraries since 1983. The main uses of the microscope are problem solving and process understanding; it is a resource for all parts of the company.
We make full use of user generated and commercial libraries in various formats. These libraries contain over 45 000 spectra in total. We needed to be able to use all of these libraries on our new system, with minimum inconvenience. With the amount of data involved, going back many years, re-compiling spectral libraries was not an option. At this time we also started to think about the many dispersive reference spectra we had filed in the laboratory, which on paper were of limited use. We sorted through these reference spectra, selecting those that were of good quality and not duplicated in our electronic libraries. The spectra were sent to Mathshop(1) who did an excellent job of digitising them. A library was then compiled. The instrument supplier we selected for the new infrared microscope, supplied a program to convert our user generated libraries to their search format. Commercial libraries cannot be converted for copyright reasons, but can be searched using another supplied program.
Why Library Searching?
As soon as a member of staff has been trained to obtain spectra they can start to solve problems (IR knowledge not essential). A lengthy period of time spent training new members of staff before they can generate information is not a luxury we can afford. A certain amount of initial training is needed, so that the library user makes the correct inferences from the library search result. An experienced team member is still essential, of course, when more complex problems arise.
We have, perhaps, an unusual approach to building spectral libraries. Virtually all data is entered into our libraries, including mixtures and unknowns. Similar spectra can occur several years apart. On more than one occasion a recurring problem has been detected, even if a full identification of the chemical(s) present has not been possible. The background information of both can be used to help solve the current problem. Also, library searching can suggest the class of compound present, even if the exact identity is not obvious. An example of a library search result is below:
Figure 1. A typical library search report
The top spectrum is the sample spectrum and the other spectra are earlier spectra revealed by the library. Referring to the records of these earlier samples can enable identification of the major components present by even a fairly inexperienced person.
The image shows part of a filter used in a production area. These filters sometimes become blocked. Identifying the material causing the blockage is vital to improve the manufacturing process
Figure 2. A blocked filter from a production area
Problem Solving Process
Discussion with client… We need to know:
Background information to the problem
Possible causes of the problem
Contaminants that may be in the system
WHAT do they need to know in order to solve the problem?
WHEN do they need the information (usually yesterday!)
Talking is the most important part of the process.
Choice of Technique(s)
Other techniques may provide complementary information, such as:
Scanning Electron Microscopy
Atomic Absorption Spectroscopy
Infrared microscopy is a very useful problem solving tool, but other techniques may be appropriate instead or as well, depending on the situation. When deciding on the techniques to use, it is important to focus on the information needed to solve the problem. Unexpected results may result in further techniques being used.
It is important to have reference spectra of all substances that may be expected to be present. We ensure that all relevant reference spectra are in our libraries.
We then obtain spectra of the relevant samples concerned with the problem. We search the spectra against the libraries held on the instrument PC. If that does not reveal a good match we extend the search to other data held locally. If necessary, we then transmit spectra to our world-wide colleagues for their input. Consulting world-wide colleagues introduces a delay because of different time zones. We are looking at ways of sharing data more efficiently.
Further discussion with client
Is the information supplied to us enough to solve the problem?
Has the cause of the problem been identified?
Is any further analytical work needed?
We then issue a formal written report. Feedback is very important; additional information can then be included with the library entry. Should a similar situation arise in the future, then this information would be invaluable.
Examples of types of samples we look at:
Surface deposits on photographic film and paper
Deposits from solutions used in equipment at photo-finishers
Inclusions in polymer samples
The ordinary and the extraordinary!
Typically we are asked to identify unknown substances.
Techniques for Preparing Samples
For transmission samples, preparation is everything! Samples must be thin and flat. For preparation we use a good stereo zoom optical microscope – it is important to be able to see clearly what you are doing when you prepare a sample. It is not a good idea to try to use the infrared microscope for preparation, as there is the danger of causing damage to the optics. To manipulate samples, you need a fine pointed probe. We find butterfly mounting pins (!) are ideal, because they can be treated as disposable and have a very fine tip (2).
A roller is needed to flatten samples when using salt windows (we use 13x1mm, barium fluoride). Particularly useful for hard or darkly coloured samples are diamond windows (3). The use of hard diamond windows permits the sample in between to be squashed very thin. (Diamond windows are quite expensive, but never need replacing). Where possible one of the windows is removed before analysis to avoid fringing effects.
Samples such as polymers containing inclusions are prepared by obtaining a microtome cross section (5-10 microns thick).
The microscope system we chose is highly automated and allows us to select on-screen several areas to be analysed. We can then analyse automatically all areas marked (with different size apertures for each sample, if required). The image of the particle(s) can then be recorded (as below).
Figure 3. Several points have been selected for analysis
and been analysed automatically
Surface deposits (for example on photographic paper) can be analysed by Attenuated Total Reflectance (ATR) microscopy or ATR micro-mapping. The choice between the two depends on the appearance and the background knowledge of the sample (and whether your IR microscope system allows ATR mapping). Mapping is required if the surface deposit is believed to be inhomogenous.
ATR has the advantage of being non-destructive and so is particularly important for deposits on negatives when the consumers want them back afterwards.
The spectra below illustrate an amusing problem we were asked to look at:
A consumer sent some negatives into a photo–finisher for re-prints. When the photo-finisher opened the envelope the negatives had white bits all over them. Alarmed, the photo-finisher telephoned the customer. It emerged that the customer had spilt orange juice over the negatives!! To help the photo-finisher to try to clean the negatives, we were asked to identify the white debris and compare it with the envelope to see if it was the same material. The top spectrum shows the gelatin surface of the negative. The middle one shows the surface deposit and the bottom one shows the lining on the inside of the paper. The spectra are not of paper itself, but of inorganic coatings on the paper.
Figure 4. Example of problem solving using the ATR technique.
The deposit matched the lining of the negative pocket.
Our new FTIR microscope system (4) has proved to be:
Easy to use
A means of improving our productivity in analysing small samples
About the author:
I am 26 years old. I have been working at Kodak for nearly six years, of which I spent three years doing synthetic organic chemistry, completing an HNC in Chemistry. Since 1995, I have been working in the vibrational spectroscopy section. As well as infrared microscopy I do “routine” bulk infrared, near infrared and Raman spectroscopy. I am also Safety Officer for R&D (“part time”). I find problem solving work both interesting and enjoyable. If you (the reader) are a young person considering a career in this field, I wish you good luck and hope you enjoy it as much as I do.
I would like to thank Alan Strawn, my supervisor, for his encouragement and support.
Mathshop, 127 Middlesbridge Stret, Romsey, Hants, SO51 8HH.
Watkins and Doncaster, PO Box 5, Cranbrook, Kent, TN18 5EZ. They can send samples of probes so that you can choose the size best suited to your needs.
We use a Spectra-Tech Sample Plan fitted with diamond windows.
Perkin-Elmer AutoImage microscope coupled with a PE1000 spectrometer.