Vapor pressure tells you how readily a liquid (or solid) sends molecules into the gas phase at a given temperature, and calculating it comes down to choosing the right equation for your situation. The simplest and most widely taught approach uses the Clausius-Clapeyron equation, which relates vapor pressure to temperature through a substance’s heat of vaporization. For everyday lab and engineering work, the Antoine equation is the go-to because published constants exist for thousands of chemicals. Beyond those two, several more specialized formulas handle extreme conditions, mixtures, tiny droplets, and even substances you have never measured before.
The Clausius-Clapeyron Equation
This is the equation most people encounter first, and it works well when you need a quick estimate over a moderate temperature range. In plain terms, it says that the natural logarithm of vapor pressure changes linearly with the inverse of absolute temperature. If you know the vapor pressure at one temperature and the enthalpy of vaporization, you can solve for the vapor pressure at any other temperature. The equation looks like this: ln(P₂/P₁) = −ΔH_vap/R × (1/T₂ − 1/T₁), where R is the gas constant (8.314 J/mol·K), T is temperature in kelvins, and ΔH_vap is the enthalpy of vaporization in joules per mole.
To use it in practice, suppose you know that water has a vapor pressure of about 3.17 kPa at 25 °C (298 K) and that its enthalpy of vaporization is roughly 40,700 J/mol. You want the vapor pressure at 50 °C (323 K). Plug in: ln(P₂/3.17) = −40700/8.314 × (1/323 − 1/298). That gives ln(P₂/3.17) ≈ 0.866, so P₂ ≈ 3.17 × e^0.866 ≈ 7.5 kPa. The actual measured value at 50 °C is about 12.3 kPa. The gap tells you something important about this equation’s limitations.
The Clausius-Clapeyron equation rests on two assumptions: that the vapor behaves like an ideal gas and that the enthalpy of vaporization stays constant across the temperature range. Neither holds perfectly. The vapor-phase assumption breaks down as you approach a substance’s critical point, and the enthalpy of vaporization actually changes with temperature. A preprint revisiting the derivation points out that while the linear behavior on a ln P versus 1/T plot is captured nicely, the underlying assumptions are not truly applicable over a wide temperature range.1ChemRxiv. Revisiting the Clausius/Clapeyron Equation and the Cause of Linearity For non-hydrogen-bonding organic chemicals, researchers have shown that the enthalpy of vaporization can be estimated from vapor pressure at 25 °C using a relationship grounded in Trouton’s rule, and this works acceptably across substances with vapor pressures spanning 15 orders of magnitude.2PubMed. Estimating enthalpy of vaporization from vapor pressure using Trouton’s rule In short, the Clausius-Clapeyron equation is a great starting point, but you should not lean on it for high-precision work or wide temperature spans.
The Antoine Equation
If Clausius-Clapeyron is the conceptual foundation, the Antoine equation is the practical workhorse. It takes the form: log₁₀(P) = A − B/(C + T), where A, B, and C are substance-specific constants and T is temperature (usually in degrees Celsius, though some tables use kelvins). These constants have been measured and tabulated for thousands of compounds, so you just look them up and plug in.
For water in the range of roughly 1 °C to 100 °C, one widely used set of Antoine constants (with pressure in mmHg and temperature in °C) is A = 8.07131, B = 1730.63, C = 233.426. At 25 °C, that gives log₁₀(P) = 8.07131 − 1730.63/(233.426 + 25) = 8.07131 − 6.6978 = 1.3735, so P ≈ 10^1.3735 ≈ 23.6 mmHg, which is close to the accepted value of about 23.8 mmHg. That is much better than what Clausius-Clapeyron delivers with a single enthalpy value, because the three fitted parameters absorb some of the nonlinearity that Clausius-Clapeyron ignores.
One thing to watch out for: different reference books publish slightly different values of A, B, and C. A comparative analysis of Antoine parameters from three independent data sources found that the temperatures predicted using any of the three were not statistically significantly different from a reference dataset, suggesting all three sources are useful.3PubMed Central. Comparing Antoine parameter sources for accurate vapor pressure prediction across a range of temperatures That is reassuring, but you should still make sure the temperature range listed alongside the constants covers the conditions you care about. Antoine parameters fitted to data between 20 °C and 100 °C will give poor results if you extrapolate them down to −40 °C.
When You Need Higher Accuracy
For engineering calculations where fractions of a percent matter, especially near a substance’s critical point, the Wagner equation is the standard. It uses four fitted constants and expresses the reduced vapor pressure (the ratio of the vapor pressure to the critical pressure) as a function of reduced temperature. The extra parameters let it capture the curvature that both Clausius-Clapeyron and Antoine miss at extreme conditions.
A study validating a simplified method for obtaining Wagner constants across 69 chemical compounds found an average absolute deviation in vapor pressure of just 0.039% when data spanned the full range from 1 kPa up to the critical pressure.4ScienceDirect (The Journal of Chemical Thermodynamics). Wagner liquid–vapour pressure equation constants from a simple methodology Even when predictions were extended up to the critical point using constants estimated from a more limited data range, average deviations stayed below about 0.3%. That is excellent performance for an equation you can implement in a spreadsheet. By comparison, the same study found deviations above 1% when using a simpler estimation scheme, making the Wagner equation the clear choice for process engineering, distillation column design, or any application where you cannot afford to be off by a few percent.
Calculating Vapor Pressure for Mixtures
Real-world systems rarely involve a single pure substance. If you are working with a liquid mixture, the most common starting point is Raoult’s Law, which says the partial vapor pressure of each component equals its mole fraction in the liquid multiplied by the pure-component vapor pressure at that temperature. For a two-component system, the total vapor pressure is just the sum of those partial pressures.
Consider a mixture of benzene and toluene at 25 °C. If benzene’s pure vapor pressure is about 95 mmHg and toluene’s is about 28 mmHg, and you have a 50/50 mole fraction mix, Raoult’s Law gives you a total vapor pressure of roughly 0.5 × 95 + 0.5 × 28 = 61.5 mmHg. This works well because benzene and toluene are chemically similar and form what chemists call a nearly ideal solution. For pairs of liquids that interact strongly with each other, like ethanol and water, Raoult’s Law gives only a rough starting point and you need activity coefficients to correct it.
For gases dissolved in liquids at low concentrations, the relevant relationship shifts to Henry’s Law, where the equilibrium ratio between a substance’s abundance in the gas phase and in the aqueous phase stays constant for a dilute solution.5Atmospheric Chemistry and Physics. Compilation of Henry’s law constants (version 4.0) for water as solvent Henry’s Law constants are widely published for common gases in water and are essential in environmental modeling, for instance when estimating how much of a volatile pollutant will partition out of a river into the atmosphere.
Water Vapor and Weather Applications
Water is by far the most commonly modeled substance when it comes to vapor pressure, and several simplified formulas exist specifically for it. Weather forecasting, irrigation scheduling, and HVAC system design all depend on accurately knowing the saturation vapor pressure of water at a given air temperature. The quantity that matters most in these fields is vapor pressure deficit (VPD), which is the difference between the saturation vapor pressure and the actual vapor pressure of the air. VPD drives evaporation, transpiration, and how “dry” the air feels.
Among the most popular approximations for saturation vapor pressure over water are the Teten formula, the Buck formula, and the Magnus formula, all of which give slightly different results. A study comparing these three against the more rigorous Goff-Gratch formula found that the Teten formula’s error increases linearly as temperature drops, and it stays acceptably accurate (below roughly 10% relative error in evapotranspiration calculations) only when the average daily temperature is above −10 °C. The Buck and Magnus formulas performed better in cold conditions and were recommended as feasible substitutes.6Procedia Engineering. Error of Saturation Vapor Pressure Calculated by Different Formulas and Its Effect on Calculation of Reference Evapotranspiration in High Latitude Cold Region
For practical use, the Buck equation is popular because it is simple enough to compute on a handheld calculator yet accurate across a wide temperature range. The general form for saturation vapor pressure over liquid water is: e_s = 6.1121 × exp((18.678 − T/234.5) × T/(257.14 + T)), with T in degrees Celsius and the result in hectopascals. At 20 °C, this gives about 23.4 hPa, which matches published values closely. If you are working in agriculture or greenhouse management, the difference between formulas rarely matters above freezing, but it can significantly affect irrigation models in cold climates or at high altitudes.
Vapor pressure deficit itself has a direct relationship to plant water use. Research on tall fescue found that transpiration was positively correlated with VPD, especially within a given temperature, though higher temperatures appeared to reduce the plant’s ability to regulate its own water loss.7Environmental and Experimental Botany. Assessing transpiration estimates in tall fescue: The relationship among transpiration, growth, and vapor pressure deficits In peanut, different genotypes responded differently to rising VPD: some showed a breakpoint around 2.2 kPa above which transpiration plateaued, while others increased transpiration linearly across the entire tested range.8Crop Science. Genotypic Variation in Peanut for Transpiration Response to Vapor Pressure Deficit Getting your saturation vapor pressure calculation right is the first step in any of these VPD-based models.
Vapor Pressure of Solids
Vapor pressure is not exclusive to liquids. Solids also exert a measurable vapor pressure through sublimation, and calculating it matters for applications from freeze-drying pharmaceuticals to understanding how ice crystals behave in the upper atmosphere. The Clausius-Clapeyron approach works here too, but you replace the enthalpy of vaporization with the enthalpy of sublimation, which is larger because you are breaking intermolecular bonds in a more ordered structure.
A complication that does not arise with simple atomic solids is that molecular solids have internal degrees of freedom, like vibrations and rotations within each molecule, that contribute significantly to the vapor pressure. Neglecting these internal modes can lead to underestimates of vapor pressure by many orders of magnitude. Researchers have developed working formulas that account for these contributions and validated them by computing the equilibrium vapor pressure of several molecular and atomic solids.9PubMed. Vapor pressure and sublimation rate of molecular crystals: role of internal degrees of freedom If you are trying to estimate how quickly a particular crystalline material will sublimate in a vacuum, the simple two-parameter Clausius-Clapeyron treatment is likely to be off by a wide margin unless you account for molecular complexity.
Small Droplets and the Kelvin Effect
All of the equations above assume a flat liquid surface. When the surface is curved, as in a tiny aerosol droplet or a nanoscale pore, the vapor pressure increases because surface tension effectively squeezes the liquid outward. The Kelvin equation quantifies this: the vapor pressure over a curved surface exceeds that over a flat surface by a factor that grows as the droplet shrinks. For droplets larger than about a micrometer, the correction is negligible. For nanometer-scale droplets, it can be substantial.
This effect is central to atmospheric science. Cloud formation depends on water vapor condensing onto tiny particles, and the Kelvin effect means that very small droplets need a higher-than-expected supersaturation to survive. Classic work on aerosol behavior used the Kelvin equation alongside adsorption theory to predict how aerosol particle size relates to relative humidity.10Journal of Colloid Science. Aerosol size and relative humidity If you are modeling cloud microphysics or the fate of airborne contaminant particles, ignoring the Kelvin correction gives you the wrong critical supersaturation and therefore the wrong nucleation rate.
High-Pressure Corrections
When the total system pressure is significantly higher than the substance’s own vapor pressure, you need one more piece of the puzzle: the Poynting correction. This factor accounts for the fact that an inert gas pressing down on a liquid increases its effective vapor pressure slightly. In most lab situations, the Poynting correction is close to 1 and you can ignore it. In high-pressure industrial processes or deep geologic settings, it becomes meaningful.
Rigorous calculation of the Poynting correction for water requires evaluating detailed thermodynamic models like the IAPWS-95 formulation, which is computationally expensive.11PubMed Central. Accuracy of Approximations to the Poynting Correction for Ice and Liquid Water Simpler approximations exist and work well in most practical cases, but knowing when you can and cannot get away with the shortcut depends on the total pressure relative to the saturation pressure. A good rule of thumb: if the total system pressure is less than a few times the substance’s vapor pressure, you can safely skip the Poynting correction. At hundreds of atmospheres, you cannot.
Fuel Blending and Industrial Use
One of the most commercially significant applications of vapor pressure calculation is in fuel formulation. Gasoline specifications are partly defined by Reid vapor pressure (RVP), which is measured at a specific temperature and reflects how volatile the fuel is. Too high, and you get excessive evaporative emissions and vapor lock in hot weather. Too low, and the engine has trouble starting in the cold.
When blending biofuels into gasoline, calculating the blend’s vapor pressure becomes a design problem. A study evaluating low-vapor-pressure bio-blendstocks found that Reid vapor pressure, distillation temperatures, and octane numbers were the key economic drivers of adding a bio-blendstock to a petroleum-derived base fuel.12Fuel. Potential economic values of low-vapor-pressure gasoline-range bio-blendstocks: Property estimation and blending optimization The economic value of different candidate molecules ranged from about $2.4 to $4.9 per gasoline gallon equivalent, with uncertainty in the property predictions leading to roughly 15% deviation in estimated value. Predicting the vapor pressure of novel blends accurately is therefore not just an academic exercise; it directly affects which biofuel candidates are worth scaling up.
Machine Learning Approaches
For substances with little or no experimental data, traditional equations are not an option because you have no measured constants to plug in. This is where computational predictions come in. Early approaches used quantitative structure-property relationship (QSPR) models, which relate a molecule’s vapor pressure to descriptors computed from its chemical structure. Neural-network-based QSPR models using descriptors derived from quantum mechanical calculations have been shown to predict vapor pressures of organic compounds with built-in error estimation.13PubMed. QM/NN QSPR models with error estimation: vapor pressure and logP
More recently, graph neural networks have taken this further. A model called GRAPPA, for instance, can predict the entire vapor pressure curve of essentially any organic molecule using nothing but its molecular structure as input.14Chemical Engineering and Processing: Advances. GRAPPA — A Hybrid Graph Neural Network for Predicting Pure Component Vapor Pressures These tools are especially valuable in drug development and environmental risk assessment, where you may need to evaluate thousands of candidate compounds and measuring each one experimentally is impractical. The predictions are not as precise as a well-fitted Antoine or Wagner equation based on lab data, but they fill a gap that no classical formula can address: estimating vapor pressure for a molecule nobody has ever synthesized.
Choosing the Right Equation
The right formula depends on what you need and what data you have. Here is a practical guide to picking among the options:
- Quick estimate, one known data point: Clausius-Clapeyron is fast and conceptually transparent. Fine for homework, rough process estimates, or sanity-checking a number.
- Standard lab or engineering work: Antoine equation with published constants. Covers most pure substances within the fitted temperature range with good accuracy.
- Near the critical point or high precision: Wagner equation. Deviations below 0.04% when well-parameterized, versus percent-level errors from simpler formulas.
- Liquid mixtures: Raoult’s Law for similar-chemistry components, modified with activity coefficients for non-ideal pairs.
- Water vapor in weather or agriculture: Buck or Magnus formula for saturation vapor pressure, then compute VPD from there.
- No experimental data at all: Machine learning or QSPR models that predict from molecular structure.
Every equation on this list is an approximation. Even the most accurate ones embed assumptions about how molecules interact with each other and with the gas phase. The art is in matching the precision of your equation to the precision you actually need, and in knowing the temperature and pressure range where your chosen formula breaks down. A well-fitted Antoine equation in its valid range beats a misapplied Wagner equation any day, and a simple Clausius-Clapeyron estimate scribbled on a napkin can save you from a catastrophic blunder in a chemical process before the detailed modeling is even set up.