Plausibility of detecting viruses using optical spectra Ok, this is a huge long shot, and I also apologise if this is a poor question (I'm a newbie here...). This doesn't seem to me a strictly perfect fit for Chemistry.SE, Biology.SE, or Photography.SE either, so hopefully it's acceptable here, given that there is some level of Physics involved...(please move if necessary) Here goes...!
Research relating to taking optical spectral signatures of viruses seems somewhat hard to come by on the web (quite understandably so given their characteristics), but I am nonetheless curious as to whether the following (admittedly utterly crazy) apparatus might be at all theoretically plausible, or just an impossible fantasy:
Could it be possible, by fitting a spectrograph of sufficient resolution to a relatively low-cost optical camera (such that of a cellphone) under the following conditions:


*

*Adequate lighting conditions.

*Capturing large numbers of 'raw', uncompressed, unprocessed frames.

*Applying stacking techniques similar to those used in Astronomy to
increase effective exposure time and improve SNR when imageing very
faint objects.

*Performing the aforementioned stacking operation and saving the
result at extremely High Dynamic Range (due to the faintness of the
target signal relative to the noise and also the total amount of
light being collected).


Be able to detect some kind of super-faint characteristic spectral signature from the proteins or other molecules of, say, a virus (such as SARS-CoV-2) on a surface/clinical sample.
Given that these techniques work on extremely faint stars, galaxies, and other celestial objects, it seems the plausibility of this would depend heavily on how much of a (feeble) unique spectral signature (if any) molecules in a virus present on a surface might impart on light reflecting off that surface.
Alternatively might some kind of incredibly faint uniquely characteristic diffraction pattern caused by the virus particles be detectable using a similar method?
Thanks!
 A: I see two major problems with your approach: signal to noise ratio and specificity. 
Astronomy uses light sources on a dark background. The fact that the sources are faint and require all the techniques you mentioned is not that bad due to the dark background, but try to acquire the image of a star in broad daylight...The same goes with your virus. The number of viruses is relatively small, whatever technique you use. If you have other signals at the same wavelength, the signal will disappear.
This brings me to what I think is your best shot: fluorescence. For example, see https://www.ncbi.nlm.nih.gov/pubmed/15339207 which mentions that viruses contain molecules (mainly tryptophan in proteins) that can exhibit fluorescence. If you can control illumination (dark room with your own source at specific wavelength), you can filter the source wavelength and image the specific wavelengths created by fluorescence, you may be able to detect fluorescence created by viruses. Stacking images may even allow you to image distribution on a surface. 
The problem will probably be specificity. You will detect in this manner a lot of tryptophan containing object on your surface. Many viruses and proteins will be visible. I doubt you can get specific enough to detect a specific virus. This is why tests that use fluorescence tag viruses with a specific (fluorescent) molecule and then image the molecule, you then get the specificity you needed, but I may be wrong. If you can combine many fluorescence signals from the same item to obtain a signature, your detection scheme could just maybe be possible, but any device to do so would probably be very complex and expensive.
