Troubleshooting

From Course Wiki
Jump to: navigation, search

There were several issues that had to be addressed while performing this study. Among them, there were a few that will be mentioned here since they led to interesting conclusions or developments. In particular, we will refer to the issue presented by an off-spec laser, both in beam shape and wavelength, a laser power issue and a beam splitting phenomenon at the dichroic mirror level.

Misshapen beam

The laser issue arose when we realized that the 500 mW, 655 nm diode laser we had ordered had a very interesting beam shape, comprising three parallel lines forming a sort of sandwich, even when focused far away. The initial approach was to contact the company for the correct collimating optics, but since the latter is located in China, the sending would take more time than the one allotted for this class. Therefore, to partially solve the issue, we tried using a cylindrical lens to differentially focus the laser in one direction instead of the other. This proved to be fruitless since the strange shape was conserved at shorter distances (or the ones in our setup at least). The actual solution involved the use of a diaphragm that would constraint the used region of the beam to the most homogeneous one. This proved to be the best option for a more uniform sample illumination (though it still had some irregularity).

Off-wavelength light

After implementing this solution, we observed that the laser light was actually passing right through our dichroic and our emission filter. We noticed then that the laser specifications indicated that the wavelength of the laser was 655 ± 10 nm, and that our cutoff wavelength was around 665 nm. Therefore, the laser seemed to be right at the cutoff.

The initial inclination was to clean up the wavelength via an emission filter, but then diode lasers are usually narrow in wavelength, which meant the laser had a peak centered in the higher end of its possible interval. To tackle this issue we turned to a collaborator that provided a 680/10 emission filter, which is a little past our emission maximum but it still captures most of it while excluding the laser wavelength by far. This again proved fruitless, since we could still clearly see the laser through the filter even when we modulated it at low power.

The end result was that the laser was completely off specifications wavelength wise, so we should’ve obtained its spectrum before going through the hassles of getting more filters and such. The issue was solved by borrowing a 5 mW, 632.8 nm HeNe laser from another collaborator.

Emission filter instead of dichroic mirror

When aligning the HeNe laser for the setup, we noticed that there were still irregularities in the beam profile after it went through the objective. We saw 4 beams of different intensities instead of 1, which was rather odd. After checking most of the components, we discovered that the lens we thought was a dichroic mirror turned out to be an emission filter, but we took the filter set from a publication (van de Linde et al.) and we misinterpreted the methods section. The strange beam pattern originated from the fact that filters are designed for a particular light incidence angle, which is normally 90 degrees in excitation and emission filters. This is due to the way filters are made, since the thickness of the coatings used are tailored to block particular wavelengths, and if the angle changes then the effective thickness of each layer also changes. Dichroic mirrors (or beamsplitters) are designed for use at 45 degrees, and therefore that determines their essential character in microscopy setups. We solved the issue by ordering and actual dichroic mirror for the wavelengths we were going to use.

Low laser power

Finally, when STORM imaging was attempted, the laser power we were delivering wasn’t enough to photoswitch the dye molecules, which explains the reasoning behind using a high power laser. We tackled the issue by removing the beam expander from the path, which reduced the area being uniformly excited but at the same time concentrated the power of the laser. This proved to be very effective and photoswitching could clearly be seen in a small area within the image. Therefore we centered the laser on whatever feature interested us and took images there.