Quite frankly, the ST-L4620 looks like a dud to me. The output is fixed frequency (Fosc/32) and you have to make a judgment about the amplitude and wave shape to decide the liquid present/absent. That takes an ADC plus some arbitrary algorithm that you need to develop for yourself.

For less external parts than needed for the ST-L4620, you can use a 555 timer and monitor the pulse width of the output. The timer MUST be a CMOS 555 variant and not the bipolar type. The frequency varies enormously with moisture present/absent so the whole system is digital, needs no ADC, and can be set with good confidence that your pulse width decision points will be right first time.

Take a look at

http://www.emesystems.com/OL2mhos.htm

You can drop off the voltage regulator and drive the whole system from the processor supply voltage, I have used this system for moisture sensors and it is extremely sensitive to (water) moisture. Not sure how it goes with non-polar solvents but that is not covered for the ST-L4620 either.

HTH
Brian