Hi,
Some issues I see:
* no fast ceramics capacitors to stabilize power supplies.
* no ESD protection on the electrode connection
* no suitable low pass filtering in the input stage
* you talk about probe resistance, but I don´t see a resistance measurement at all. I just see a voltage masurement. (What exact pH probe are you using?)
* did you read the datasheet chapter "COMPENSATING FOR INPUT CAPACITANCE"? --> stability
* PCB layout
***
So I think it´s very likely that
* the amplifier gets killed by an ESD impulse when removing the probe,
* the OPAMPs start to oscillate.
Btw: measuring high impedance does not mean you have to avoid resistors.
Why are there different symbols for 10uF polarized capacitors?
Did you check supply voltages with DVM and scope?
Did you follow the signal path with a scope? (I mean under true application conditions)
Klaus