# When does ratiometric measurement account for signal variations?

I've been reading about pH and ion sensors that rely on a ratiometric approach, which supposedly circumvents many problems of intensity-based methods such as dye bleaching, dye leaching, and fluctuations in source intensity. I can think of situations though when this might not be the case.

For example, fura-2 is excited at two independent wavelengths, and the two emissions are detected sequentially. If the excitation source intensity at either wavelength were to fluctuate, the ratio of the emissions would be off from the calibration. Another example would be a FRET sensor in which the acceptor dye leaches at a faster rate than the donor dye; the emission ratio of acceptor to donor would no longer match the calibration.

So when does ratiometric measurement account for signal variations like dye bleaching, dye leaching, and fluctuations in source intensity? Is it only when there is one dye and one excitation source? Am I missing any other requisites? It seems like if there's only a single dye that bleaches/leaches or a single excitation source that fluctuates, the change in dye fluorescence would remain proportional such that the emission ratio remains the same.

Thank you!

So when does ratiometric measurement account for signal variations like dye bleaching, dye leaching, and fluctuations in source intensity? Is it only when there is one dye and one excitation source? Am I missing any other requisites? It seems like if there's only a single dye that bleaches/leaches or a single excitation source that fluctuates, the change in dye fluorescence would remain proportional such that the emission ratio remains the same.

You are correct. Ratiometric readout of optical sensors works best if you use one excitation source and one probe molecule that gives you the two signals to be used.

Another example would be a FRET sensor in which the acceptor dye leaches at a faster rate than the donor dye; the emission ratio of acceptor to donor would no longer match the calibration.

This happend to me before, where donor and acepptor where actually tethered to the same molecule and the acceptor bleached faster then the donor, which also screwed up the measurment.

National Instruments$^{TM}$ gives a good explanation here, in a paper titled "Measuring Bridge-Based Sensors with the Ratiometric Approach, Publish Date: Feb 11, 2016". Here are a couple excerpts relevant to your question:

1) "The fact that the mV/V ratio stays constant for changes in the excitation voltage increases measurement stability. Small changes in the environment due to temperature differences typically affect excitation voltages. Because the excitation voltage is sensed and fed back to the ADC, any changes play a smaller role in the accuracy and stability of the measurement."

2) "By using the excitation as the source for the ADC reference, the output of the ADC provides measurements that are in units of V/V. This reduces the amount of scaling required to return the appropriate units."

So, the short of it is that noise, particularly that arising from power fluctuations, in either the source or detector in the measurement systems as you described is largely cancelled out when ratioed to a reference system.

This approach is most effective when you are rapidly switching, i.e. via a chopper mirror, between the sample measurement source and detector and that of your reference or blank measurement. When the signal from the source is then ratioed to that of the reference system, then any noise common to the two is effectively cancelled out, increasing your signal to noise ratio. This method addresses your examples of "sequential" measurements. In the system describe above, you are effectively taking your reference measurement simultaneously or in parallel rather than sequentially.