A reference electrode is used as in equilibrium so that the net-current in circuit connected to the reference electrode is desired to be zero. Or rather it must be used under zero current. Connecting terminal for reference electrode in potentiostat has extremely high input impedance so that it is impossible for current to flow through the reference electrode. That is one of the important functions of potentiostat.

The most used reference electrode nowadays is silver-silver chloride electrode (SSCE). SSCE is environmentally friendly compared to calomel electrode using mercury and its salt. Its stability, reliability and reproducibility are also sufficient.

The reaction on SSCE is shown as below.

Ag + Cl- ⇔ AgCl + e- ​ ​(+0.20V)

The electrode potential of above reaction is given as follows.

E = E0Ag+ + (RT/F)ln aAg+ ​ ​eq.(1)

Substitution of the activity of silver ion with the solubility product constant of AgCl , KsAgCl results eq(1) into following equation (2).

E = E0Ag+ + (RT/F) ln KsAgCl - (RT/F) ln aCl- ​ ​eq.(2)

Using literature values of E0Ag+ and KsAgCl ( E0Ag+＝0.7991 V, and KsAgCl = [Ag+][Cl-] = 1.6×10-10), well-known equation (3) is obtained.

E = 0.222 - (RT/F) ln aCl- ​ ​eq.(3)

aCl- is the activity of chloride ion in internal solution of the electrode. So, the reference potential of SSCE depends on the activity of chloride ion in addition to temperature. An increase of the activity results in a cathodic shift of the potential. Hence, keeping internal chloride concentration to constant is important to obtain reproducible results.

Reference potential of saturated SSCE(in saturated KCl solution) is 0.197 V at 25°C versus standard hydrogen electrode(SHE). Electric contact of reference electrode with test solution is performed through the frit which permits mutual transfer between the internal solution and the test solution. Hence, dilution of the internal solution and contamination of the test solution provides a possible significant trouble sometimes.

Instead of KCl, sometimes NaCl is chosen as an internal electrolyte. When perchlorate anion is involved in test solution, NaCl is often employed as internal electrolyte because of the low solubility of potassium perchlorate compared with sodium salt. Deposition of potassium perchlorate within the frit may cause serious trouble like large increase of reference electrode impedance.

A main reason of using KCl is that potassium cation and chloride anion have almost identical mobility so that the liquid junction potential at the liquid interface composed of these ions is minimized. While, the liquid junction potential for the case using NaCl as internal electrolyte may become rather considerable. Which one is better choice might depend on a compromise among practical convenience.

Management of reference electrode is important, otherwise resulting in shift of reference potential. Reference electrode should be kept in solution of the same composition of internal solution, while it is not used. Contamination or deposition within the frit due to external substances should be excluded, otherwise resulting in big trouble such as oscillation of potentiostat caused by external noise etc.

Check method of Ag/AgCl reference electrode
For the check, if the reference electrode is properly, you can measure comparing with the new one of the same type. 
Since the reference electrode in the cell is the same, the ideal potential difference is 0 Volt, however in a practice a small potential difference will occur. It can be used if the potential difference, between the electrodes, is within 0 ± 20 mV. For the measurement, immerse the reference electrode in 3 M NaCl solution or saturated KCl solution, and then measure the potential difference between both them using a voltmeter.

Reference Potential for difference type of the reference electrode.

Equilibrium reaction shown below takes place in calomel electrode.

Hg + Cl- ⇔ 1/2Hg2Cl2 + e- ​ ​(1)

The electrode potential is given as following equation.

E = E0Hg22+/Hg + RT/2F ln Ks(Hg2Cl2) - RT/F ln aCl- ​ ​(2)

Early two terms in right side of equation (2) is replaced to E0Hg2Cl2/Hg which is calculated to be 0.273V from the literature values of redox potential
(E0Hg22+/Hg = 0.7960V) and the solubility product constant (Ks(Hg2Cl2)＝2×10-18). 

Replacing again E0Hg2Cl2/Hg to E0, the equation (3) is obtained, which is familiar formula involving the activity of chloride ion.

E = E0 - (RT/F) ln aCl- ​ ​(3)

Although a small discrepancy in above calculated value exists, E0 is known as 0.268 V for standard potential at 25°C. The electrode potential depends on the concentration of chloride ion as like silver-silver chloride electrode. It is 0.280 V at 1M KCl and 0.241 V at saturated KCl which is so often called as SCE as abbreviation of Saturated Calomel Electrode.

The reference potential of SCE is 0.241 V(at 25°C) , because of increased activity of chloride ion in internal solution of saturated KCl (ca. 4.8 M). In addition to the chloride concentration dependence of potential, the temperature dependence of potential is about -0.5 mV per degree. Except for SCE, the reference potentials are 0.280 V and 0.334 V for 1 M KCl and 0.1 M KCl, respectively. The higher the concentration of KCl is, the larger the temperature dependence of the potential is. 

We often receive the question of "Is SCE usable for experiment in organic solvent?" . Answer is "Yes". However, in the case when you dislike a contamination with water or chloride ion, you are recommended to use a salt bridge, in which both the same organic solvent and electrolyte used in your test solution is contained. Even so, you have to pay attention the fact that a rather large junction potential between SCE and the salt bridge still remains unknown. In addition to that, considering a growing risk of deposition in the frit, it might be better to use a reference electrode for non-aqueous solvent(such as Ag-Ag+ in organic solvent ) or pseudo-reference electrode (such as Pt wire), then refer to the internal standard such as ferrocene.

Usage of pseudo-reference just connecting to the reference terminal of potentiostat is recommendable from the stand point of low impedance.

You may not want to use the typical reference electrode (Ag-AgCl or calomel electrode. For example, you may be discouraged to use it in the cases that your test solution has to be kept away from a contamination with chloride ion or it is strong alkali or strong acid etc. In those cases, you have a choice of mercury-mercuric oxide (Hg-HgO) reference electrode for alkaline solution and mercury-mercurous sulfate reference electrode for neutral or acidic solution.

Mercury-mercuric oxide reference electrode (Hg-HgO electrode)

This is used for strong alkaline solution. It is utilized the equilibrium reaction between mercuric oxide layer formed on mercury and hydroxide anion in alkaline solution as internal solution. Consecutive reactions are shown below.

Hg ⇔ Hg2+ + 2e- ​ ​ (1)

HgO + H2O ⇔ Hg(OH)2 ⇔ Hg2+ + 2OH- ​ ​(2)

The electrode potential for the reaction (1) is given shown below as equation (3).

E= E0Hg/Hg2+ + (RT/2F) ln aHg2+ ​ ​(3)

Since the activity of mercuric ion is given by using the solubility product constant as aHg2+ = Ks(Hg(OH)2) /(aOH-)2, equation (3) is replaced to equation (4).

E = E0Hg/Hg2+ + (RT/2F ) ln Ks(Hg(OH)2) - (RT/F ) ln aOH- ​ ​(4)

As described in the section of silver-silver chloride, the standard potential of the reference electrode utilizing highly insoluble salt depends on significantly its solubility product constant.

Using E0Hg/Hg2+ = 0.8537 and Ks(Hg(OH)2) ≃ 10-25, equation (4) is finally given as below.

E= 0.116 - (RT/F ) ln aOH-  ​ ​(5)

As the internal electrolyte, 0.1 M, 1 M, 3 M and saturated solution of NaOH or KOH are employed. Saturated solution of Ca(OH)2 or Ba(OH)2 are sometimes used because of its low solubility compared with NaOH or KOH. Lower OH- activity of these alkali earth solution is favorable for use of glass container and raises the standard potential a little bit upward. The standard potentials of 1 M KOH, 1 M NaOH, saturated Ba(OH)2 and saturated Ca(OH)2 are 110 mV, 113 mV, 146 mV and 192 mV, respectively.

Mercury-mercurous sulfate electrode

The electrode reaction and the electrode potential are shown as follows.

Hg + 1/2SO42- ⇔ 1/2Hg2SO4 + e- ​ ​(6)

E = E0Hg/Hg22+ + (RT/2F) ln aHg22+ ​ ​(7)

The equation (7) is replaced by using the relation between the activity of mercurous ion and the solubility product constant Ks(Hg2SO4),

Ks(Hg2SO4) = [Hg22+][SO42-]) as follows.

E = E0Hg/Hg22+ + (RT/2F ) ln Ks(Hg2SO4) - (RT /2F ) ln aSO42- ​ ​(8)

Putting E0Hg/Hg22+ = 0.7960V(25°C) and Ks(Hg2SO4) ≃ 7×10-7, into equation (8), the final form is shown as below.

E= 0.6125 - (RT/2F ) ln aSO42- ​ ​(9)

The reference potential is considerably positive compared to that of calomel electrode reflecting rather large solubility of mercurous sulfate. Saturated K2SO4 or 1 M H2SO4 are used as the internal electrolyte. This reference electrode is an appropriate choice for the case to avoid a contamination of chloride ion.

Line-up of those kinds of electrode are available from ALS Japan as RE-61AP for Hg-HgO electrode and RE-2CP for Hg-Hg2SO4 electrode.