The Joule-Thomson Porous Plug Experiment Explained
Open a can of compressed air for cleaning a keyboard, hold the trigger for a few seconds, and the can turns cold in your hand. Let a carbon dioxide fire extinguisher discharge and frost forms at the nozzle. Gas escaping from high pressure to low pressure through a narrow opening tends to cool down. That everyday observation is one face of a phenomenon first measured carefully in the 1850s by two of the leading physicists of the Victorian era: James Prescott Joule and William Thomson, who later became Lord Kelvin.
Their experiment was simple in concept and demanding in practice. They pushed a steady stream of gas through a plug of tightly packed cotton or silk and measured its temperature just before and just after. The difference was small, often a fraction of a degree, but it was real, it depended on the gas, and it reversed sign for hydrogen. Within about half a century, engineers had used that small difference to liquefy air, then hydrogen, then helium, opening up the entire field of low-temperature physics.
This article explains why Joule and Thomson ran the experiment, how it worked, what they measured, why the effect happens, and how it shapes technology today.
The problem: does a gas cool when it expands?
By the mid-19th century, the idea that heat was a form of energy rather than a fluid was gaining ground. Joule had done much to establish it. His paddle-wheel experiments in the 1840s showed that a fixed amount of mechanical work always produced the same amount of heat. (See our article on the Joule paddle-wheel experiment for that story, and Rumford's cannon-boring experiment for an earlier challenge to the fluid theory of heat.)
One open question concerned gases. When a gas expands, does its temperature change even if no heat is added or removed and no outside work is done? The answer mattered for the emerging science of thermodynamics because it tested whether the energy of a gas depends only on its temperature or also on how spread out its molecules are.
Joseph Louis Gay-Lussac had studied this question earlier in the century, and Joule tried it himself in 1845. In Joule's version, now called free expansion or the Joule expansion, he connected a copper vessel of compressed air to an evacuated vessel, immersed both in water, opened a valve between them, and measured whether the water temperature changed as the air rushed into the empty vessel. He detected no change.
The trouble was sensitivity. The water and the metal vessels had a large heat capacity compared with the small mass of gas, so any temperature change in the gas was diluted to the point of being invisible. Joule's result was consistent with the gas having no change in internal energy, but it could not rule out a small effect.
Joule meets Thomson
William Thomson was a young professor at the University of Glasgow who had become one of the most important thinkers in the new science of heat. He had been skeptical of Joule's ideas at first, then became convinced and began working with him. Their collaboration became one of the most productive in 19th-century physics, with Joule doing most of the hands-on experimenting and Thomson analysing the results and suggesting further experiments.
Together they arrived at a different experimental design that could detect the small effect Joule's free-expansion setup had missed. Instead of a single sudden expansion into a vacuum, they could use a steady flow: push gas continuously through a restriction from a high-pressure side to a low-pressure side, and measure the temperature of the gas itself on each side once everything had settled into a steady state. Because the gas would flow continuously, the containers would reach thermal equilibrium with it, and the small temperature difference would no longer be swamped by the heat capacity of the apparatus.
The porous plug apparatus
The experiments began in 1852, and the results were published in a series of papers in the Philosophical Transactions of the Royal Society between 1853 and 1862 under the title "On the Thermal Effects of Fluids in Motion." The porous plug appears in the second part (1854). The apparatus was refined over the decade, but the core elements were:
- A pump driving a steady supply of gas at a constant high pressure.
- A coil of pipe immersed in a water bath, so the gas arrived at the plug at a known, steady temperature.
- A porous plug, made of tightly packed cotton wool or silk, held in a tube. The plug's job was to throttle the gas, letting it pass only slowly, so that it emerged at a lower pressure without picking up any significant speed.
- Thermometers placed in the gas stream just before and just after the plug.
- Insulation around the plug, so that no heat leaked in or out during the passage.
The design meant that the gas did no useful external work and exchanged essentially no heat with its surroundings. The only thing that happened was that it went from high pressure to low pressure through a narrow, tangled pathway.
Throttling: forcing a fluid through a restriction (a valve, a narrow tube, or a porous plug) so that its pressure drops without the fluid doing useful work or exchanging heat. A throttling process conserves a quantity called enthalpy.
What they found
Joule and Thomson tested several gases at a range of pressures and starting temperatures. Their main findings were:
- Air cooled after passing through the plug. The temperature drop was small, a fraction of a degree for each atmosphere of pressure drop at room temperature, but it was consistent and grew with the size of the pressure drop.
- Carbon dioxide cooled more strongly than air under similar conditions.
- Hydrogen warmed slightly after passing through the plug at room temperature. It behaved in the opposite way from the other gases.
- The cooling effect was smaller at higher starting temperatures. As the gas entering the plug was made hotter, the temperature drop shrank.
The fourth result hinted at something important. If cooling became weaker at higher temperatures, then above some temperature it might vanish entirely and reverse. Hydrogen, which already warmed at room temperature, would presumably cool if it started cold enough. That prediction turned out to be right.
| Gas | Result at room temperature | Inversion temperature (approx.) |
|---|---|---|
| Nitrogen | Cools on expansion | about 620 K (about 347 °C) |
| Oxygen | Cools on expansion | about 760 K (about 487 °C) |
| Carbon dioxide | Cools strongly on expansion | about 1,500 K |
| Hydrogen | Warms on expansion | about 200 K (about −73 °C) |
| Helium | Warms on expansion | about 40–45 K (about −233 to −228 °C) |
Gases cool on throttling only when they start below their inversion temperature. The air gases sit far below theirs at room temperature, so they cool. Hydrogen and helium sit above theirs, so they warm. Inversion temperatures also vary with pressure; the figures here are the commonly quoted maximum values, and sources differ by a few kelvin (for example, 621 to 623 K for nitrogen and 761 to 764 K for oxygen).
Why does the gas change temperature?
The cooling or warming happens because real gases are not ideal gases.
In an idealized gas, the molecules are treated as points that do not attract or repel each other. The energy of such a gas depends only on its temperature. If an ideal gas passed through the porous plug, its temperature would not change at all. (For background on the gas laws that describe that idealized behavior, see Boyle, Charles, and Avogadro's gas laws.)
Real molecules do interact. At moderate distances they attract each other weakly. At very short distances they repel strongly. Both forces affect what happens in throttling.
Attraction leads to cooling. As the gas expands through the plug, its molecules end up farther apart on average. Pulling apart molecules that attract each other takes energy. Since no heat is added from outside, that energy comes from the molecules' own motion. Slower molecules mean a lower temperature.
Repulsion and flow work can lead to warming. At higher temperatures, molecules collide harder and more often, and the short-range repulsion becomes more important. There is also a contribution from the "flow work" involved in pushing gas into and out of the plug. When these effects dominate, the net result is warming.
The inversion temperature is where these competing effects balance. Below it, attraction wins and the gas cools. Above it, the gas warms. Hydrogen and helium have very weak attractions between their molecules, so their inversion temperatures are very low. Nitrogen, oxygen, and carbon dioxide attract more strongly and have inversion temperatures far above room temperature.
The Joule-Thomson experiment thus provided some of the first quantitative evidence for intermolecular forces. Johannes Diderik van der Waals published his equation of state for real gases in 1873, which included terms for molecular attraction and molecular size. That equation predicts a Joule-Thomson effect and an inversion temperature of the general kind that Joule and Thomson observed.
The thermodynamics: why enthalpy is conserved
The porous plug experiment is a textbook example of an isenthalpic process, one in which the quantity called enthalpy stays the same.
Enthalpy combines the internal energy of a gas with the product of its pressure and volume. When gas is pushed into the plug from the high-pressure side, the gas behind it does work on it. When gas leaves on the low-pressure side, it does work pushing the gas ahead of it. Because no heat is exchanged and no other work is done, the internal energy change of the gas exactly balances the net difference in this pushing work. That is equivalent to saying enthalpy before equals enthalpy after.
For an ideal gas, enthalpy depends only on temperature, so constant enthalpy means constant temperature. For a real gas, enthalpy also depends on pressure, so the temperature has to shift when the pressure drops. The rate of that shift is called the Joule-Thomson coefficient: the change in temperature per unit change in pressure at constant enthalpy. A positive coefficient means cooling on expansion. A negative coefficient means warming.
This framing also explains the difference between Joule's 1845 free-expansion test and the 1852 porous-plug test. They measure different things. Free expansion keeps internal energy constant. Throttling keeps enthalpy constant. Both are zero-effect for an ideal gas, but for real gases the throttling effect is more practical to measure and more useful in engineering.
For more on how these ideas fit into thermodynamics as a whole, see how thermodynamics explains everything.
From a fraction of a degree to liquid air
The Joule-Thomson effect by itself is small. A single pass of air through a valve at room temperature drops its temperature by only a few degrees for a large pressure drop. That is not enough to liquefy air, which requires cooling to about −194 °C or below at ordinary pressure.
The breakthrough came in 1895, when Carl von Linde in Germany and William Hampson in England independently developed a regenerative cooling method. The key addition was a counter-current heat exchanger:
- Air is compressed to high pressure and cooled back to near room temperature.
- The compressed air flows through a heat exchanger and then through a throttling valve, where the Joule-Thomson effect cools it slightly.
- The cooled, low-pressure air flows back through the heat exchanger in the opposite direction, cooling the incoming high-pressure air before it reaches the valve.
- The incoming air now reaches the valve colder than before, so after throttling it emerges even colder.
- The cycle repeats, and the temperature at the valve ratchets downward with each pass until part of the air condenses into liquid.
The Linde-Hampson cycle turned a weak effect into a powerful one by letting it build on itself. Linde's company, which he had founded in 1879 as a refrigeration firm, grew into a major industrial gas company, and liquid air production became the basis for producing pure oxygen, nitrogen, and argon on an industrial scale by fractional distillation.
Hydrogen and helium: getting below inversion
Hydrogen and helium posed a problem. At room temperature, both warm on throttling, so the Linde-Hampson cycle alone would heat them rather than cool them.
The solution was to precool them below their inversion temperatures using another refrigerant first. In 1898, James Dewar at the Royal Institution in London liquefied hydrogen for the first time, precooling the gas with liquid air and then using Joule-Thomson expansion. The following year he produced solid hydrogen.
Helium, with the lowest inversion temperature of all, required precooling with liquid hydrogen. On July 10, 1908, Heike Kamerlingh Onnes at Leiden University in the Netherlands became the first to liquefy helium, reaching about 4.2 K, just above absolute zero. Three years later, in 1911, his group discovered superconductivity, the complete loss of electrical resistance in mercury cooled with liquid helium. Kamerlingh Onnes received the Nobel Prize in Physics in 1913.
The chain of discoveries runs straight back to the porous plug: without the Joule-Thomson effect and its inversion temperature, the path to liquid helium and superconductivity would have been far harder.
Where the Joule-Thomson effect matters today
Industrial gases. Air separation plants use refrigeration cycles that include Joule-Thomson expansion, often combined with turbine expanders that extract work for greater efficiency. Liquid nitrogen, liquid oxygen, and argon for medicine, steelmaking, electronics, and food processing come from these plants.
Liquefied natural gas. Natural gas is cooled to around −162 °C to make LNG for shipping. Joule-Thomson expansion is one of the cooling steps used in liquefaction processes.
Refrigeration and air conditioning. In a typical vapor-compression refrigerator, refrigerant passes through an expansion valve or capillary tube from high to low pressure. This is a throttling process, and the temperature drop that follows is the basis of the cold side of the cycle. In this case much of the cooling comes from the refrigerant partly boiling as its pressure falls, but the throttling step is the same isenthalpic process Joule and Thomson studied. For the full cycle, see how refrigerators work.
Miniature cryocoolers. Small Joule-Thomson coolers, fed by a bottle of high-pressure gas, are used to cool infrared detectors in some missile seekers and thermal cameras, and in cryosurgery probes that freeze diseased tissue.
Natural gas pipelines. When high-pressure natural gas passes through a pressure-reducing valve, it cools. If the gas contains water, the cooling can form ice or solid gas hydrates that block the valve, which is why such stations often preheat the gas.
Hydrogen refueling. Because hydrogen warms when it expands at room temperature, hydrogen fueling stations must manage heating as gas flows from high-pressure storage into a vehicle's tank. Stations typically precool the hydrogen to keep the tank within safe temperature limits.
Everyday examples. Spray cans of compressed gas and CO2 fire extinguishers cool during discharge. These involve the Joule-Thomson effect along with other processes, such as the liquid in the can evaporating.
Common misconceptions
"Gas always cools when it expands." Not always. Hydrogen and helium warm on throttling at room temperature, and any gas warms if it starts above its inversion temperature. Cooling is guaranteed only for gases below their inversion temperature.
"The Joule-Thomson effect and adiabatic expansion are the same thing." They are different. In adiabatic expansion that does work, such as gas pushing a piston or spinning a turbine, an ideal gas cools strongly because it gives energy away as work. In Joule-Thomson throttling, no useful work is extracted, and an ideal gas would not cool at all. The effect exists only because real gases have intermolecular forces.
"Joule did this experiment alone." It was a collaboration with William Thomson, in which Joule did most of the experimental work and Thomson analysed the results and suggested further experiments, and the steady-flow design worked where Joule's earlier free expansion had not. The effect is also called the Joule-Kelvin effect for this reason.
"The effect is too small to matter." It is small on a single pass. With a regenerative heat exchanger, as Linde and Hampson showed, repeated small drops add up to cryogenic temperatures.
Why the experiment mattered
The porous plug experiment shows how a careful choice of method can reveal an effect that a cruder method misses. Joule's 1845 free expansion was a reasonable test, and its null result was honest. But it was not sensitive enough. By shifting to a steady flow, Joule and Thomson let the apparatus come to equilibrium with the flowing gas and removed the dilution problem.
It also shows how basic research can lead to large industries without anyone intending it. Joule and Thomson were testing ideas about energy and the nature of gases. They were not trying to liquefy air. Within about four decades, their measurements had become the foundation of the industrial gas business, and in 1908, of the first liquid helium.
Finally, the experiment gave early, quantitative evidence that gas molecules attract each other. That insight fed into van der Waals's equation of state and into the broader picture of molecules as real objects with real forces, decades before their existence was universally accepted. For another historical study of gas behavior, see our article on Carnot and heat engines.
Key takeaways
- James Joule and William Thomson (Lord Kelvin) began the porous plug experiments in 1852 and published their results between 1853 and 1862.
- They pushed gas at steady flow through a plug of cotton or silk and measured its temperature before and after.
- Air and carbon dioxide cooled, while hydrogen warmed at room temperature.
- The effect arises because real gas molecules attract and repel each other, unlike in an ideal gas.
- Each gas has an inversion temperature, below which it cools on throttling and above which it warms.
- Carl von Linde and William Hampson used the effect with heat exchangers to liquefy air in 1895.
- James Dewar liquefied hydrogen in 1898, and Heike Kamerlingh Onnes liquefied helium in 1908, both relying on Joule-Thomson expansion after precooling.
Sources
- Joule, J. P., & Thomson, W. (1853–1862). "On the Thermal Effects of Fluids in Motion," Parts I–IV. Philosophical Transactions of the Royal Society of London, 143 (1853), 144 (1854), 150 (1860), 152 (1862).
- Wikipedia - Joule-Thomson effect
- schoolphysics - The Joule-Kelvin effect
- Cold Facts (Cryogenic Society of America) - Joule-Thomson effect
- JoVE Science Education - Joule-Thomson effect
- Wikipedia - James Dewar
- NobelPrize.org - Heike Kamerlingh Onnes biography
- Wikipedia - Hampson-Linde cycle
FAQ
Q: What is the Joule-Thomson effect in simple terms?
A: It is the change in temperature of a real gas when it is forced from high pressure to low pressure through a valve or porous plug without exchanging heat or doing useful work. Most gases cool at room temperature, while hydrogen and helium warm.
Q: Why did Joule and Thomson use a porous plug?
A: The plug throttled the gas, so it passed from high to low pressure slowly and steadily without gaining significant speed. This created a steady-state flow in which the small temperature change could be measured reliably, unlike Joule's earlier free-expansion experiment.
Q: What is the inversion temperature?
A: It is the temperature at which a gas switches from cooling to warming on throttling. Below it, the gas cools. Above it, the gas warms. Nitrogen's maximum inversion temperature is about 620 K, while hydrogen's is about 200 K and helium's only about 40 to 45 K.
Q: Why does hydrogen warm instead of cool?
A: Hydrogen molecules attract each other only very weakly. At room temperature, the effects that cause warming outweigh the small cooling from separating attracting molecules. Cooled below about 200 K, hydrogen does cool on throttling.
Q: How did the Joule-Thomson effect lead to liquid air?
A: In 1895, Carl von Linde and William Hampson added a counter-current heat exchanger, so gas cooled by throttling precooled the incoming compressed gas. Repeating this cycle drove the temperature down far enough for air to liquefy.
Q: Is the Joule-Thomson effect used in refrigerators?
A: The expansion valve in a vapor-compression refrigerator is a throttling device that works under the same constant-enthalpy conditions. Most of the cooling there comes from the refrigerant partly evaporating as its pressure drops.
Read next:
Follow new articles
Add SciHub101 to your feed reader to see new science and technology stories.
About the Author
SciHub101 is an independent science and technology publication founded and edited by Buğra Sözeri, a software engineer and app developer based in Türkiye. AI tools assist with drafting and research. Source documentation in the existing archive is uneven, and articles are being reviewed and corrected. The Editorial Team byline identifies editorial responsibility; it does not certify specialist or independent human review. Where we compare products, anything built by our own team is labelled as such — see the editorial guidelines and About page for how corrections and disclosures work.
Related Articles

The Joule Paddle Wheel Experiment: Proving Heat Is a Form of Energy
James Joule used falling weights and spinning paddles to prove that mechanical work and heat are the same thing — establishing the first law of thermodynamics and the SI unit that bears his name.

How Pressure Cookers Change Boiling Water
A pressure cooker traps steam to raise water's boiling point from 100°C to about 121°C, cutting cook times several-fold. Here is the physics and the math.
