July 2: UV-Vis Update

After the UV-Vis was found out to be malfunctioning last Thursday, I had been anticipating the arrival of the new lamp that Professor Harbron hastily ordered. Meanwhile my experiment came to a halt so I was left pondering over past data and designing new experiments based on my findings. On Tuesday the lamp arrived early in the afternoon and Professor Molloy came in to install it later during the day. Following the instructions from a digital manual stored in the computer, we disconnected the spectrometer, lift it up on one side, and replaced the lamp, which was integrated into a module screwed onto the bottom surface of the spectrometer. We then connected it back and ran the Align application on the computer. A plastic lid tightly fitted to the front panel was then popped off, revealing two screwing positions that could control the alignment of internal optics. Professor Molloy carefully adjusted those two positions with his small screw driver until the signal registered on the Align application seemed to reach its maximum (about 10%T), which was not a fixed value but a reading that constantly fluctuated.

Then we ran my APESO sample. No, the problem was not gone, if not worse than before.

Harbron: This makes me cry a little bit.

Grace and Aaron started to evaluate my emotional state as to where it sat on the five-stage disappointment process – denial, anger, bargaining, depression, acceptance. I went to have dinner, and then came back, and then back again later that night to see if an operable workflow could be devised on the spectrometer sitting in the PChem lab. It turned out there were several problems with integrating that machine into my day-to-day experiment routine. First of all, it was not in our lab. So precursor solutions would need to sit for longer before they could be examined and moved into the next step. More travel and exposure to light would also introduce a lot of potential error. And then it was a slow machine coupled with a sketchy, old software. The software wouldn’t take file names longer than 8 characters and the old-school scheme of data retention and file saving bewildered me. Nonetheless, I familiarized myself with it and built up a tentative routine of using that machine. So on Wednesday morning I prepared a dye-doped nanoparticle suspension using that spectrometer to measure my dye concentration in free solution and determine the concentration thereof. But that sample didn’t go well. The thousand-ppm PFBT solution had gotten too concentrated so the resulting nano suspension was concentrated and under-doped. The sample also registered a bizarre spectrum on the pchem lab spectrometer, with big dips at around 560 nm and 470 nm. It seemed as if it couldn’t handle nanoparticle samples. I then diluted the sample and re-examined it on the Lambda 35 spectrometer in the spectroscopy room and the spectrum showed a 4% dye-to-CPN peak ratio. By this point in the experiment I knew that 4% was too low for effective quenching and that dye bleaching would too easily take over, so reversibility just couldn’t be achieved.

Presumably that pronounced total temporary death to my project. We were prepared to indefinitely postpone my UV-Vis kinetics experiments because our machine might need to be sent in for examination and repair while the pchem machine didn’t like nanoparticles. Professor Molloy came in and did several self-deployable performance tests on our instrument and sent the files to Agilent, which now owns Varian, the company that made our machine. Professor Harbron had run those same tests on Tuesday afternoon but they all returned PASS. This time it was no different. Despite the doldrums I encountered on the experimental end, on the theoretical front we were making big strides. After my discussion with Professor Harbron about progress I made on green light switching, she had been pumping out a lot of ideas and started “storyboarding” our – at this point, quite promising – paper.

Today (Thursday) in the afternoon Harbron suddenly came in and asked me about the sample holder. Yes indeed we swapped the old sample holder for another one that could be connected to the external peltier accessory. Harbron screwed that one off the base and put the old one in. She then ran a bunch of blank methanol and pink Rhodamine samples. The spectra now looked much better, with the peak absorbance almost matching what it was on the departmental one. But the UV region stilled looked off. Harbron was delighted to learn that the sample holder was a large part of the culprit. I then ran some APESO dye samples using the two sample holders and, voila, the difference between the two was very clear.

For a brief moment we thought Molloy could come back in, tweak the internal optics a little bit and all problems would then be gone. But Molloy decided not to rush surgery but instead to send more diagnostic data to the company. Meanwhile we re-considered the option of wheeling in the PChem spectrometer for temporary use. I ran the diluted nano sample from yesterday on that machine and miraculously it turned out fine this time.

The plan now is to wheel in the spectrometer and its computer from the pchem lab. We will wait for the company’s reply to decide what further action we should take to fix our own beloved machine.