I have varying levels of experience and skill with audio and rf frequency which have crossed over various times in my career and also in my hobbies. For example I noticed once when I cleaned a carbon rod with acetone that it sang. I went on intrigued by this to detect the frequency with a tuner and it was just above high E on a guitar, which led me to consider many things. If I were to play a high E near this rod would I be able to measure the vibration and see in a graph the effect? Well I could. And I did. I also played it with my violin bow. Anyway this is years and years ago and I have not stopped considering these things from then till now. I also have amateur radio licences and experiment there. I found that I could do exactly the same thing with oscillators and measured wavelength conductors. So all this and now ms-gc got me thinking.

Has anyone measured the resonance or the reflected energy from samples of alkaloids suspended in solutions that exhibit very little response them selves.

Allow me to elaborate an abstract theoretical idea that I have:

Let's say we took a sample of compound x and dissolved it in water in a glass tube.
We know the resonance of the tube and the volume of water plus the weight of the compound and the weight of the water and the weight of the tube of course.
Now let's say we try to resonate this sample in different ways. Perhaps we vibrate it by inducing frequencies on a mobile phone vibrating motor using pwm and we just measure audio. We could measure by direct contact using piezoelectric crystals. We could add some probes and send the signal through the solution and similarly we could measure the voltages in the solution. Jeez we could even induce fly back currents and measure some of the properties of the sample.

What I'm getting at is that I would love to go ahead and attempt to build something like this.
Seriously speaking, is there any merit in this idea? Is it possible?

I guess the basic process lets say for detecting some alkaloid in reed grass would be something like this:

You have a standard of the compound, referenced by all the methods available and the read outs.
You mash to a pulp some quantity of the suspect grass and then filter and clarify it as per a million teks.
Place some of the solution in a tube and submit it to the sequence of tests where it gets bombarded with audio waves via air then via direct contact then via probes then the same with rf waves sweeping through to UHF say 500Mhz.
Measurements are taken in so many ways almost in reverse but also maybe certain things happen with light during the vibrations!
So we shine a laser of who only knows what colour through the solution and measure that with a camera and ir channel.
Surely there is something amongst those conditions which are wide and variable that will create a unique signature.

Think of this like a virus database or a Chladni pattern where we are just looking for something unique that's only true when compound x is present.

I would love to hear anybody who has any ideas on this because I really feel like it has to work and further more I'm pretty certain that we could develop a system that could be useful in many other ways like checking a different compound has been eradicated from a solution.

Remember we are talking microphones, piezo electrics, vibrating motors, laser light diffraction and graphing.

That's all, and it's inside or embedded in most handheld devices and parts can be salvaged from printers/faxes/tvs/cd drives/ and all other manner of defunct electronics you and I all have sitting around gathering dust. The parts from a flatbed scanner are even almost a full package already. Like how the Rosetta project had various packages that ran their tests and spat out their telemetry.

A flatbed scanner could be used to visibly detect alkaloids in solution. I can't do it off the cuff but I can tell you how. Get the scanning arm to rotate round a beaker with alkaloids in solution directly opposite the rotating sensor.
Have it spew out pulsed light at all the frequencies available to the LEDs on board (this is where I'm lost) then have the sensor record the light through the solution and software graph it to a scale. Wavelengths seen? And amplitude? There you go. Isn't that spectroscopy?

So how can we get to a point where we can plug and play if you like with samples.
Please remember when you analyse something in a lab it's different to analysing a mimosa sample in this idea because although it may seem irresponsible I don't care to detect arsenic or strychnine or things outside the box of relevant I just want to confirm the presence of, and perhaps if I'm lucky, quantify specific compounds.

What can I do to some samples i have to create these references these standards or do some exist already?
What is the resonant frequency of harmine?

That's where I'm at with this idea so far but I continue to research and think on it. I read yesterday a guy modulated terahertz on to a laser to detect the compounds of explosi ves from a distance.

Food for thought.