July 15: Final Update on UV-Vis

Last week Harbron finally decided to fully investigate the UV-Vis problem on her own. On Wednesday morning she wheeled in the spectrophotometer from PChem lab. She then ran some standard fluorescein samples in ethanol. She used the standard 1-cm-by-1-cm cuvette and found that both our machine and McNamara’s Cary 60 gave the same results. However, the two PerkinElmer spectrophotometer gave almost double the absorbance readings. That, Harbron said, was quite a separate issue from what we were trying to resolve. As long as our reading is the same as McNamara’s, which is a later model made by the same manufacturer, we should regard our machine as fixed.

But my sample still gives a lower reading. After some correspondence with the manufacturer, Harbron started to suspect that the special, narrow cuvettes I used for my experiments were causing the issue. A quick comparison between scans done with the standard square cuvette and my narrow cuvette confirmed her suspicion. Apparently part of the incoming detection beam was occluded by the quartz glass wall of the narrower cuvette that would otherwise be a clear passage in the standard one. The precise reason why this issue cropped up a few weeks ago was hard to know. It must have been that some internal misalignment occurred while we were moving the machine and replacing the sample holder. As suggested by the manufacturer, we fixed the issue by adjusting a long, metal screw that could tilt the entire sample holder by a minute angle, thus allowing the occluded light beam to avoid the glass and completely pass through the sample.

At this point the problem had been nominally fixed for now. And interestingly my samples now gave the same results on the Cary machines as they did on the PerkinElmer machines.