The effect of temperature on reference electrode potential
Dmitry Galyamin
Co-founder of Electroseek
·11 min read
Almost every reference electrode potential you will find is tabulated at 25 °C.
The problem is that not every experiment runs at 25 °C, and in many cases we do not even know precisely what temperature we are measuring at. What we call room temperature is around 20 °C, but it varies with region and season: a laboratory at 15 °C in winter is not the same as one at 30 or 35 °C in summer when the air conditioning cannot keep up. And when the experiment calls for specific conditions, we measure at 40, 60 or 80 °C.
The potential of a reference electrode is not an instrument constant. It is a thermodynamic property, and it moves with temperature.
What the specifications tell you, and what they do not
Reference electrode datasheets almost always state an operating temperature range. Ranges such as −10 to 100 °C or −5 to 80 °C are common, depending on the reference system and the body design, and you can compare them across the reference electrodes available.
That range tells you whether the electrode will survive. It does not tell you what potential it is reading within that range. These are two different things.
How much the potential shifts
This is not an open question: it has been measured and published for decades.
The expressions in use are empirical fits to electromotive force measurements, not theoretical derivations. Different groups measured real cells at different temperatures and fitted a polynomial to their results. The Handbook of Reference Electrodes (Inzelt, Lewenstam and Scholz, Springer) collects both the tables of measured values and the equations, with their original sources, and the Handbook of Electrochemistry (Zoski, Elsevier) gathers them into a quick-reference table in its chapter on reference electrodes.
The saturated KCl calomel electrode is one of the best characterised, so it serves well as an example. These are the values published in the Handbook of Reference Electrodes:
| Temperature | Potential vs SHE | Difference from 25 °C |
|---|---|---|
| 0 °C | 0.2568 V | +15.6 mV |
| 10 °C | 0.2507 V | +9.5 mV |
| 20 °C | 0.2444 V | +3.2 mV |
| 25 °C | 0.2412 V | — |
| 30 °C | 0.2378 V | −3.4 mV |
| 40 °C | 0.2307 V | −10.5 mV |
| 50 °C | 0.2233 V | −17.9 mV |
| 60 °C | 0.2154 V | −25.8 mV |
| 70 °C | 0.2071 V | −34.1 mV |
What about the other reference electrodes?
Calomel is an example, not a special case. Every reference system depends on temperature, each with its own curve. What changes is the magnitude and how far it has been characterised.
There is something worth pointing out: the most widely used electrode, silver/silver chloride, is the one with the fewest directly tabulated data. For calomel and for mercurous sulfate, the Handbook of Reference Electrodes publishes tables of measured values temperature by temperature. For Ag/AgCl what is published is the equation fitted by Bates and Bower, valid between 0 and 95 °C, and referring to the standard potential of the couple rather than to a commercial electrode filled with saturated KCl.
This comparison brings the three systems together. The calomel and mercurous sulfate columns are measured, tabulated values; the Ag/AgCl column was calculated by us from the published Bates and Bower equation:
| System | Shift at ~50 °C | Shift at ~70–75 °C | Source of the value |
|---|---|---|---|
| Calomel, saturated KCl | −17.9 mV (50 °C) | −34.1 mV (70 °C) | measured table |
| Ag/AgCl, standard potential E° | −17.9 mV (50 °C) | −34.6 mV (70 °C) | calculated from published equation |
| Hg/Hg₂SO₄, standard potential E° | −27.6 mV (55 °C) | −49.0 mV (75 °C) | measured table |
Three notes on these systems.
Silver/silver chloride. Keep in mind the limitations of using the standard potential. At 25 °C the E° of the couple is 0.2222 V, while a commercial electrode with saturated KCl reads around 0.197 V, because the chloride activity in saturated KCl is not unity. The slope with temperature is comparable, on the order of 1 mV per degree near room temperature. Other work has extended the characterisation above 100 °C.
Mercurous sulfate. Widely used in chloride-free media and in lead-acid work. It moves the most of the three, and is tabulated from 0 to 100 °C.
Mercury oxide. The usual choice in alkaline media. Polynomials have been published by the IUPAC group, with measurements extended between 283 and 363 K, and a coefficient on the order of 1 mV per degree. The Handbook itself warns that in some of those values the number of significant digits is probably exaggerated.
There is one important limitation common to all of them: do not take these figures as the exact value of your own electrode. They depend on the system, on the internal electrolyte and on the individual unit. To get a sense of the real uncertainty, the standard potentials published for mercurous sulfate at 25 °C by different authors range from 0.6125 to 0.6294 V. Seventeen millivolts of spread between careful studies, for the same type of electrode.
Around room temperature the shift is typically a few millivolts; at 40–80 °C it can readily reach tens of millivolts.
Tens of millivolts are not noise. If you are comparing the onset of two catalysts, or determining an overpotential at a given current density, that is where the difference between a real improvement and an artefact lies. The potentiostat will not warn you: the instrument measures correctly, what has shifted is the zero of the scale.
Why it shifts
Several factors contribute: the thermodynamics of the redox couple, the activity of the ions that set the potential, the solubility of the salts involved and, in some systems, the chemical equilibrium of the electrode itself.
One of them dominates in practice, and it shows clearly when you compare two calomel electrodes with different internal electrolytes. According to the published coefficients, the saturated KCl electrode moves on the order of 0.66 mV per degree near room temperature, and the 0.1 M KCl electrode on the order of 0.09 mV per degree. Same redox couple, same chemistry, almost a factor of ten.
The reason is that in a saturated electrode the chloride concentration is not set by you: it is set by the solubility of KCl, which rises on heating. In a fixed-concentration electrode, it does not change.
Hence a practical rule: for variable-temperature work, a fixed-concentration electrode is considerably more stable than a saturated one.
Is the standard hydrogen electrode exempt?
The standard potential of the SHE is assigned a value of 0.000 V at every temperature. This is a convention, not a statement that its absolute electrode potential is physically zero.
The practical consequence is what matters. When you say "potential versus SHE at 80 °C", the zero of your scale is the SHE at 80 °C, which is not the same physical state as the SHE at 25 °C. Two values, both quoted "vs SHE", may not be on the same scale if they were measured at different temperatures.
The SHE needs no correction because by definition it is the zero. But the zero moves with you.
When the reference and the solution are at different temperatures
One possible setup is that the cell operates at 80 °C while the reference electrode is kept in a side arm at room temperature to protect it.
With a temperature gradient along the bridge, a thermal liquid junction potential is established, which depends on the temperature difference and also on the bridge electrolyte. In low-viscosity solutions and moderate gradients it is estimated below 0.2 mV/K, and somewhat higher in strongly acidic media. In high-temperature, high-pressure systems it stops being a detail and becomes the main correction.
It is worth distinguishing two situations that give different results: the isothermal measurement, with working electrode and reference at the same temperature, and the non-isothermal one, with the working electrode hot and the reference cold. The published recommendation is to always correct isothermal data for the temperature coefficient of your reference electrode, determined with the same supporting electrolyte you use with your analyte.
And something basic that gets forgotten: the temperature shown on the hotplate or the bath is not the temperature of your solution. Measure with a thermometer inside the cell, close to the electrodes, and one in each compartment if you work with several.
Not every reference electrode tolerates the same range
Every system has its limit, and it does not depend only on the chemistry of the redox couple. It also depends on the internal electrolyte, the body material, the junction type and the seal. Two electrodes with the same chemistry can have different operating ranges because of how they are built.
As a general guide: calomel-based systems are the most temperature-limited, and above a certain point they stop being usable for chemical rather than mechanical reasons. Ag/AgCl systems generally tolerate wider ranges. And there are designs made for extreme conditions, including electrolytes formulated not to freeze below zero.
There is also an effect that is often overlooked: thermal hysteresis. Some electrodes do not return to exactly the same value after being heated and cooled. Correcting the number is not enough; it is worth checking the reference after a thermal cycle.
If you work in non-aqueous media, bear in mind that internal standards are affected too: ferrocene evaporates, and its evaporation rate increases markedly with temperature.
What temperature will you measure at, for how long, and in what medium? With those three details we can help you identify which reference systems suit your case and which manufacturers offer them. Get in touch and we will go through it with you, free of charge and with no obligation.
A warning about the equations in circulation
If you are going to apply a correction equation, check what it refers to. As we saw with Ag/AgCl, some expressions describe the standard potential of the couple rather than that of a commercial electrode with saturated KCl, and there are about 25 mV between them. Since they sometimes appear in the same table, it is easy to pick the wrong one.
Check where it comes from as well. These equations have been copied from one publication to another for decades, and there is at least one documented case of a transcription error. Contrast it against the original source.
What to do in practice
If you measure at a temperature other than 25 °C, correct for it. The equation for your reference is tabulated and it is a two-minute calculation.
Better than correcting, calibrate at your working temperature and with the same supporting electrolyte you will use.
If you can choose, avoid saturated electrodes for variable-temperature work.
If you are going to cycle temperature, check the reference after the cycle.
Measure the temperature inside the cell, not in the bath.
And report the conditions: which reference, which internal electrolyte, at what temperature and how you calibrated it. A good part of the spread in the literature would disappear with that alone.
The key takeaway
Temperature shifts the potential of your reference electrode, and under normal laboratory conditions that shift is on the order of tens of millivolts. It does not show up as a software error: it shows up as a plausible result.
To go further, our pH-based reference electrode guide covers the effect of the medium while explicitly leaving temperature aside, and the article on pseudo-reference electrodes and potential drift covers another common source of shift.
To stay updated on new content and insights, you can subscribe to our newsletter.
Do you need a reference electrode for work outside room temperature?
Bibliography:
- Handbook of Reference Electrodes, Springer, 2013, Ch. 5, Tables 5.3.1 and 5.3.2
- Handbook of Electrochemistry, Elsevier, 2007, Ch. 4, Table 4.14
- J. Electrochem. Soc. 1959, 106, 616–625
- ACS Energy Lett. 2025, 10, 1542–1549
- ACS Energy Lett. 2025, 10, 3269–3274
- Electrochim. Acta 2016, 212, 102–112
- J. Electrochem. Soc. 2009, 156, P56–P61
- Anal. Chem. 2026, 98, 2692–2700
- Pine Research, DRK10053, Overview of Reference Electrodes, 2016
Dmitry Galyamin
Co-founder of Electroseek
I am Dmitry Galyamin, PhD in Electrochemistry and co-founder of ElectroSeek. After more than ten years in academic research focused on electrocatalysis, electrochemical biosensors, and corrosion studies, I worked as a scientific consultant helping laboratories and companies solve practical challenges in electrochemistry. These experiences led me to create ElectroSeek, a platform designed to make it faster and easier for scientists to find the right electrochemical equipment and information for their work.