A ferrocene/ ferrocenium electrode is easy to make. Its redox potential is about +0.2V vs. Ag/AgCl.
Here a method I have used to make this in the past:
A 50ml solution of ferrocene (0.004moldm-3, 0.0037g), ferrocenium tetraflouroborate (0.004moldm-3, 0.0055g) and TBA BF4 (0.1moldm-3, 0.1646g) in DMSO was prepared for the reference electrode. A small amount of this solution was degassed with (dry) argon for 20 minutes, then placed in a glass tube with a semi-permeable frit at one end, a (flame cleaned) platinum wire was then placed in the tube and it was sealed with parafilm.
I usually use a silver wire pseudoreference electrode and add ferrocene as an internal standard (or cobaltocenium, decamethylferrocene, acetylferrocene). This also applies for DMSO.
Strictly speaking, the reference electrode must build on a redox couple of fast charge transfer kinetics (so that no or negligible polarisation occurs). In practice, this can be achieved using the Ag/Ag+ couple in which the Ag+ ion comes from a salt soluble in the internal electrolyte which is ionically connected to the external electrolyte via a salt bridge (ion conducting medium), ideally in the form of a membrane.
Simply using a metal wire (immersed in the electrolyte) as a pseudo-reference is commonplace for short time scale experiments, however, it is challenging to compare the potential data between experiments, unless a proper "internal reference" is present in the electrolyte. Thus, if you expect to do 2 or more experiments with electrolytes of different compositions, you should better prepare a proper reference electrode, which is fairly easy in most chemical laboratories. An example is shown by the attached drawing. Note that the solvent can be any organic solvent, as long as the Ag+ salt and supporting electrolyte can both dissolve in it.
Ag/AgCl electrode is very easy to build in lab. In case of measuring potentials in nonwater electrolytes you can use salt bridge system with internal electrolyte dioxanemixture with water in ratio up to 75:25 and electrolyte will behavior as water solution electrolyte (I am think that it can be interesting to try, but I am not guarantee result), But of course you can use pseudo reference electrodes Pt, Pg, Au and also Ag
Thank you, colleagues, for your helpful advice. I'm thinking on doing Ag/AgCl RE, but solubility of NaCl in DMSO is about 0,06M (KCl even less). Will it be enough?
Try the Ag/AgCl electrode with LiCl in DMSO. LiCl is soluble in a lot of organic solvents, usually rather slowly - it's reasonably soluble in DMSO, though I've no quantitative data.
Yes, 0.06 M concentration can be enough, it is necessary to ensure constant diffusion potential for all measurements. Reduced internal concentration are used in trace analysis to reduce contamination