Why is water a liquid?
Water and hydrogen sulfide are close chemical cousins: oxygen and sulfur sit one above the other in group 16, and both molecules are bent, with two hydrogens on the central atom. Yet hydrogen sulfide boils at −60 °C, while water boils at 100 °C. If water followed the pattern of its heavier relatives, it would boil somewhere around −90 °C, and Earth would have no oceans, no rain and no life as we know it.
The difference lies not in the bonds inside the molecules but in the much weaker forces between them. Whether a substance is a gas, a liquid or a solid at a given temperature is the outcome of a contest: thermal motion, which scatters molecules apart, against intermolecular attraction, which pulls them together. This lesson looks at both sides of the contest. We start with gases, where motion wins almost completely and simple laws hold; move to real gases and liquids, where attraction matters; and end with phase diagrams, which map who wins at every temperature and pressure.
Motion against attraction
Temperature measures the energy of random molecular motion. At 25 °C the typical kinetic energy available per mole is of order kJ/mol. Compare that with the attractions between molecules:
- In a gas the molecules are far apart (in air, about ten molecular diameters on average) and move freely at hundreds of metres per second, colliding often. Attractions are too weak, at that distance, to hold them.
- In a liquid the molecules touch their neighbours and attract them, but have enough energy to keep sliding past each other: a liquid flows and takes the shape of its container, but keeps its volume.
- In a solid attraction wins: each particle vibrates about a fixed position. In a crystalline solid those positions form a repeating lattice; in an amorphous solid such as glass they do not.
Heating a solid gives its particles more energy until they break free of fixed positions (melting); heating the liquid lets them escape the surface altogether (boiling). The stronger the attractions, the higher the temperatures needed, which is why the boiling point is a sensitive measure of intermolecular forces.
Definitions
Definition— pressure
Pressure is force per area, , in pascals (). Common units: Pa and the standard atmosphere, Pa exactly. A gas exerts pressure through its molecules hitting the walls.
Definition— ideal gas law
An ideal gas obeys , where is the thermodynamic temperature in kelvin and J mol⁻¹ K⁻¹ is the molar gas constant, exact since 2019 because both and the Boltzmann constant J K⁻¹ are. Real gases approach ideal behaviour at low pressure and high temperature.
Definition— partial pressure
In a mixture of ideal gases, each gas contributes the pressure it would exert alone in the same volume, its partial pressure , where is its mole fraction, and the partial pressures add up to the total (Dalton's law). Air at sea level is about 78% N₂ and 21% O₂ by amount, so the partial pressure of oxygen is about 21 kPa.
Definition— vapour pressure and boiling point
The vapour pressure of a liquid is the pressure of its vapour in equilibrium with it at a given temperature. A liquid boils when its vapour pressure equals the surrounding pressure; the normal boiling point is the temperature at which this happens at 1 atm.
Definition— phase diagram, triple point and critical point
A phase diagram shows which phase is stable at each temperature and pressure. Along the lines two phases coexist. At the triple point solid, liquid and gas coexist. At the critical point the liquid–gas line ends: above the critical temperature and pressure there is a single supercritical fluid, and no amount of compression produces a separate liquid.
The ideal gas and the motion behind it
The ideal gas law combines the experimental laws of Boyle (at fixed , ), Charles and Gay-Lussac (at fixed , ) and Avogadro (at fixed and , ).
Example— The air in a classroom
A classroom measures 10 m × 8 m × 3 m = 240 m³. At 20 °C (293.15 K) and 101.3 kPa it holds mol of air. With an average molar mass of 28.96 g/mol, that is about 289 kg: the air in a room weighs as much as four adults.
Kinetic theory explains the law. Model the gas as a huge number of tiny particles in random motion, with no forces between them except in brief elastic collisions. A molecule of mass bouncing off a wall transfers momentum; adding up all the impacts per second per area gives
where is the mean of the squared speeds. Comparing with shows that the average kinetic energy of a molecule is . Temperature is average molecular kinetic energy, for an ideal gas, and the root-mean-square speed is .
Molecules do not all share one speed. Collisions constantly redistribute energy, and the result, worked out by James Clerk Maxwell in 1860 and Ludwig Boltzmann soon after, is that each velocity component follows a normal (Gaussian) distribution with variance . The speed then follows the Maxwell–Boltzmann distribution
whose peak (most probable speed), mean and root-mean-square are , and . Rather than trust the formula, we can simulate it: draw random velocity components for 200 000 nitrogen molecules, compute their speeds, and compare.
import numpy as np
k_B = 1.380649e-23 # J/K, exact
N_A = 6.02214076e23 # 1/mol, exact
R = k_B * N_A # 8.314... J/(mol K)
T = 298.15 # K
M = 28.014e-3 # kg/mol: nitrogen, N2
m = M / N_A # mass of one molecule, kg
# Each velocity component of a gas molecule is normally distributed with variance kT/m.
rng = np.random.default_rng(seed=2026)
n = 200_000
v = rng.normal(0.0, np.sqrt(k_B * T / m), size=(n, 3))
speed = np.linalg.norm(v, axis=1)
theory = {"most probable": np.sqrt(2 * R * T / M), "mean": np.sqrt(8 * R * T / (np.pi * M)),
"root-mean-square": np.sqrt(3 * R * T / M)}
counts, edges = np.histogram(speed, bins=np.arange(0, 1601, 50))
centres = (edges[:-1] + edges[1:]) / 2
simulated = {"most probable": centres[np.argmax(counts)], "mean": speed.mean(),
"root-mean-square": np.sqrt((speed ** 2).mean())}
print("N2 at 25 °C simulated formula (m/s)")
for key in theory:
print(f"{key:17} {simulated[key]:9.0f} {theory[key]:9.0f}")
# The Maxwell-Boltzmann speed distribution, as a fraction per 50 m/s bin, against the simulation
def mb(u):
a = m / (2 * k_B * T)
return 4 * np.pi * (a / np.pi) ** 1.5 * u ** 2 * np.exp(-a * u ** 2)
print("\n speed/m/s simulated formula")
for c, k in zip(centres[:26:2], counts[:26:2]):
print(f"{c:9.0f} {k / n:9.4f} {mb(c) * 50:9.4f} " + "#" * round(400 * k / n))
# Kinetic theory: p = N m <v^2> / (3 V). One mole in the molar volume at 25 °C and 1 atm:
V = R * T / 101325
p = N_A * m * (speed ** 2).mean() / (3 * V)
print(f"\npressure from molecular speeds: {p / 1000:.1f} kPa (1 atm = 101.3 kPa)")The simulated mean (475 m/s) and root-mean-square (515 m/s) speeds match the formulas to the metre per second; the histogram's peak falls in the 400–450 m/s bin, around the formula's 421 m/s. And feeding the simulated speeds back into kinetic theory recovers a pressure of 101.2 kPa, within sampling error of one atmosphere.
Real gases and the van der Waals equation
Kinetic theory made two assumptions that real molecules violate: that molecules have no size, and that they do not attract each other. At atmospheric pressure both errors are small, because molecules are far apart. Squeeze a gas, or cool it, and they grow. The standard measure is the compressibility factor , exactly 1 for an ideal gas.
In 1873 Johannes van der Waals proposed a corrected equation:
The constant is the volume taken up by a mole of the molecules themselves, so the free volume is . The term accounts for attraction: a molecule about to hit the wall is pulled back by its neighbours, which lowers the pressure, and the effect grows with the square of the density because it needs pairs of molecules. Strongly attracting molecules have large .
import numpy as np
R = 8.314462618 # J/(mol K)
# van der Waals constants (SI): a in Pa m^6 mol^-2, b in m^3 mol^-1
GASES = {"N2": (0.1370, 3.87e-5), "CO2": (0.3640, 4.267e-5)}
# measured critical points (NIST Chemistry WebBook): T_c in K, p_c in MPa, molar volume in cm^3/mol
CRITICAL = {"N2": (126.19, 3.396, 89.4), "CO2": (304.13, 7.377, 94.1)}
def p_ideal(Vm, T):
return R * T / Vm
def p_vdw(Vm, T, a, b):
return R * T / (Vm - b) - a / Vm ** 2
T = 300.0
print("1 mol at 300 K ideal van der Waals Z = pV/RT")
for gas, (a, b) in GASES.items():
for litres in [24.0, 2.0, 0.5, 0.2]:
Vm = litres / 1000
pi, pv = p_ideal(Vm, T), p_vdw(Vm, T, a, b)
print(f"{gas:4} in {litres:5.1f} L {pi / 1e5:7.1f} bar {pv / 1e5:7.1f} bar {pv * Vm / (R * T):6.3f}")
# The van der Waals critical point: T_c = 8a/(27Rb), p_c = a/(27 b^2), V_c = 3b, so Z_c = 3/8 always
print("\ncritical point van der Waals measured")
for gas, (a, b) in GASES.items():
Tc, pc, Vc = 8 * a / (27 * R * b), a / (27 * b ** 2), 3 * b
Tm, pm, Vm = CRITICAL[gas]
print(f"{gas:4} {Tc:6.1f} K {pc / 1e6:5.2f} MPa {Vc * 1e6:4.0f} cm3 Zc={pc * Vc / (R * Tc):.3f}"
f" {Tm:6.1f} K {pm:5.2f} MPa {Vm:4.0f} cm3 Zc={pm * 1e6 * Vm * 1e-6 / (R * Tm):.3f}")
# Below T_c an isotherm has a loop: three volumes give the same pressure (gas, liquid, unstable)
a, b = GASES["CO2"]
p, T = 6.0e6, 290.0
roots = np.roots([p, -(p * b + R * T), a, -a * b]) # p Vm^3 - (pb + RT) Vm^2 + a Vm - ab = 0
real = np.sort(roots[np.abs(roots.imag) < 1e-12].real)
print("\nCO2 at 290 K and 6.0 MPa: molar volumes", ", ".join(f"{r * 1e6:.0f}" for r in real), "cm^3/mol")At 24 L per mole (room conditions) both gases are ideal to better than half a percent. Squeezed to 0.2 L per mole, nitrogen is still within 4% of ideal, but carbon dioxide's pressure is little more than half the ideal value (): its stronger attractions (larger ) pull the molecules together. At 300 K, just below CO₂'s critical temperature of 304 K, those attractions are on the verge of turning the gas into a liquid.
The equation's most remarkable feature is that it predicts condensation at all. Below the critical temperature, an isotherm of the van der Waals equation has a loop, and at 290 K and 6.0 MPa the cubic equation has three roots: 90 cm³/mol (a dense, liquid-like state), 229 cm³/mol (a gas-like state) and 126 cm³/mol between them, an unstable state that is never observed. A simple two-constant correction to the ideal gas law contains, in outline, the whole liquid–vapour transition.
Forces between molecules
The attractions that van der Waals' summarised come in three main kinds, all ultimately electrical:
- Dispersion (London) forces act between all molecules, polar or not. The electron cloud of a molecule fluctuates, creating a momentary dipole that induces a dipole in a neighbour, and the two attract. Larger molecules with more electrons, and more surface in contact, have larger dispersion forces. They are the only attraction between non-polar molecules such as methane, iodine or the noble gases.
- Dipole–dipole forces act between polar molecules, whose permanent dipoles (see Chemical bonding) tend to line up head to tail.
- Hydrogen bonds form when a hydrogen atom bonded to a small, very electronegative atom (N, O or F) is attracted to a lone pair on another N, O or F. The H atom, stripped of much of its electron density, is a concentrated positive charge that can come very close to the lone pair. At typically 10 to 40 kJ/mol, hydrogen bonds are the strongest of the three, but still at least ten times weaker than a covalent bond (the O–H bond in water: 463 kJ/mol).
Boiling points make the forces visible. The program below compares the hydrides of groups 14 to 17. Down each group the molecules get bigger and dispersion forces stronger, so the boiling point rises steadily. A straight line through periods 3 to 5 predicts what period 2 "should" do:
import numpy as np
# Normal boiling points in °C of the hydrides of groups 14 to 17, periods 2 to 5
# (typical handbook values, rounded; sources differ by a degree or so)
HYDRIDES = {
"group 14": [("CH4", -161.5), ("SiH4", -111.9), ("GeH4", -88.5), ("SnH4", -51.8)],
"group 15": [("NH3", -33.3), ("PH3", -87.7), ("AsH3", -62.5), ("SbH3", -17.0)],
"group 16": [("H2O", 100.0), ("H2S", -60.3), ("H2Se", -41.3), ("H2Te", -2.2)],
"group 17": [("HF", 19.5), ("HCl", -85.1), ("HBr", -66.8), ("HI", -35.4)],
}
period = np.array([2, 3, 4, 5])
print(" period 2 extrapolated from periods 3-5 difference")
for group, rows in HYDRIDES.items():
bp = np.array([b for _, b in rows])
slope, intercept = np.polyfit(period[1:], bp[1:], 1) # straight line through periods 3-5
guess = slope * 2 + intercept
print(f"{group}: {rows[0][0]:4} {bp[0]:7.1f} °C {guess:7.1f} °C"
f" ({slope:+.0f} °C per period) {bp[0] - guess:+6.0f} °C")
# Dispersion forces grow with size: the straight-chain alkanes
alkanes = [("methane", 16.04, -161.5), ("ethane", 30.07, -88.6), ("propane", 44.10, -42.1),
("butane", 58.12, -0.5), ("pentane", 72.15, 36.1), ("hexane", 86.18, 68.7),
("heptane", 100.20, 98.4), ("octane", 114.23, 125.7)]
print("\nalkane M/(g/mol) b.p./°C")
for name, M, bp in alkanes:
print(f"{name:10} {M:9.2f} {bp:8.1f} " + "#" * round((bp + 170) / 10))Methane falls slightly below the trend line (it has no hydrogen bonds, and is the smallest), but ammonia, hydrogen fluoride and above all water are far above it: ammonia by 93 °C, hydrogen fluoride by 132 °C, water by 193 °C. Water's extrapolated boiling point, about −93 °C, is the "no hydrogen bonds" water imagined at the start. Water wins by such a margin because each molecule has two hydrogens and two lone pairs and can join four hydrogen bonds, building a three-dimensional network, while HF and NH₃ are limited by having too few hydrogens or too few lone pairs.
The alkanes show dispersion forces alone: each extra CH₂ group raises the boiling point, by 73 °C from methane to ethane and by 27 °C from heptane to octane, which is why methane and propane are gases (fuel gas and bottled gas), octane a liquid (petrol), and long-chain alkanes waxy solids. Shape matters too: pentane, a chain, boils at 36 °C, but its compact isomer 2,2-dimethylpropane, with the same formula and less surface to touch its neighbours, boils at only about 10 °C.
Phase diagrams: water and carbon dioxide
Read a phase diagram by picking a point (a temperature and a pressure) and seeing which region it lies in. Moving horizontally means heating or cooling at constant pressure; crossing a line means a phase change. The two diagrams look alike but differ in two ways that matter.
Water's melting line leans backwards. For almost every substance, squeezing a liquid favours the denser solid, so the melting point rises with pressure and the line leans right. Ice is less dense than water, so pressure favours the liquid and the melting point falls, though only by about 0.0074 °C per atmosphere: the line is almost vertical.
Carbon dioxide's triple point is above 1 atm. At 518 kPa, about five atmospheres, it lies above everyday pressure, so at 1 atm solid CO₂ cannot melt: warmed, it turns directly into gas (sublimation) at −78.5 °C. That is why solid CO₂ is called dry ice. Liquid CO₂ exists only under pressure, for example inside a fire extinguisher, where it sits at about 60 bar at room temperature.
Above its critical point, 31 °C and 7.38 MPa, carbon dioxide becomes a supercritical fluid that diffuses like a gas but dissolves things like a liquid. Supercritical CO₂ is used industrially to remove caffeine from coffee beans and to extract flavours and fragrances, replacing organic solvents, and to transport captured CO₂ by pipeline in dense form.
Note
Dry ice is a hazard, not a toy. It can cause frostbite-like injuries if touched with bare skin, and it releases a large volume of gas: sealed in a closed container it can burst it, and in a small closed room the CO₂ can displace enough air to be dangerous. It belongs with adults who know how to handle it, with insulated gloves and good ventilation, and never in a sealed bottle or jar.
Vapour pressure and boiling at altitude
In a closed container a liquid evaporates until the vapour above it reaches a pressure at which molecules return to the liquid as fast as they leave: the vapour pressure. It rises steeply with temperature, because the fraction of molecules with enough energy to escape grows roughly like . Assuming the enthalpy of vaporisation is constant gives the Clausius–Clapeyron equation
For water, the international reference is the saturation-pressure equation of Wolfgang Wagner and Andreas Pruss, adopted by IAPWS, which agrees with the best measurements to a small fraction of a percent. The program compares it with Clausius–Clapeyron and then finds the boiling point of water at several altitudes, by bisection on the vapour-pressure curve.
import math
R = 8.314462618
def p_sat_iapws(T):
"""Vapour pressure of water in Pa: the IAPWS saturation-pressure equation (Wagner and Pruss)."""
Tc, pc = 647.096, 22.064e6
a = [-7.85951783, 1.84408259, -11.7866497, 22.6807411, -15.9618719, 1.80122502]
e = [1.0, 1.5, 3.0, 3.5, 4.0, 7.5]
tau = 1 - T / Tc
return pc * math.exp(Tc / T * sum(ai * tau ** ei for ai, ei in zip(a, e)))
def p_sat_cc(T, dH=40.66e3, T0=373.124, p0=101325.0):
"""Clausius-Clapeyron, assuming a constant enthalpy of vaporisation (its value at 100 °C)."""
return p0 * math.exp(-dH / R * (1 / T - 1 / T0))
print(" T/°C IAPWS/kPa Clausius-Clapeyron/kPa error")
for t in [0.01, 20, 40, 60, 80, 100, 150, 200]:
T = t + 273.15
a, b = p_sat_iapws(T), p_sat_cc(T)
print(f"{t:5.0f} {a / 1000:10.3f} {b / 1000:15.3f} {100 * (b - a) / a:+14.1f}%")
def boiling_point(p):
"""Temperature (°C) at which water's vapour pressure equals p, by bisection."""
lo, hi = 273.16, 647.0
for _ in range(60):
mid = (lo + hi) / 2
lo, hi = (mid, hi) if p_sat_iapws(mid) < p else (lo, mid)
return (lo + hi) / 2 - 273.15
def pressure_at(h):
"""Air pressure (Pa) at height h metres in the International Standard Atmosphere (h < 11 km)."""
return 101325 * (1 - 2.25577e-5 * h) ** 5.25588
print("\nplace (height) pressure/kPa water boils at")
for place, h in [("sea level", 0), ("Nairobi (1795 m)", 1795), ("Quito (2850 m)", 2850),
("La Paz (3640 m)", 3640), ("Everest summit (8849 m)", 8849)]:
p = pressure_at(h)
print(f"{place:30} {p / 1000:8.1f} {boiling_point(p):12.1f} °C")With a constant fixed at 100 °C, Clausius–Clapeyron is within 2% from 80 to 150 °C but overestimates the vapour pressure by 21% at 20 °C and 37% at 0 °C, because water's enthalpy of vaporisation is larger at low temperature (about 45 kJ/mol at 0 °C). It is a good local approximation, not a law over the whole range.
At altitude the air pressure is lower, so water boils sooner: at about 94 °C in Nairobi, 90.5 °C in Quito, 88 °C in La Paz and about 70 °C on the summit of Everest (where the real pressure, about 34 kPa, is a little above the standard-atmosphere model's 31 kPa because the tropical atmosphere is warm). Pasta and beans cook more slowly in the Andes, and mountaineers cannot make a properly hot cup of tea. A pressure cooker does the opposite: raising the pressure inside by about one atmosphere raises the boiling point to about 120 °C, and food cooks much faster. Its safety valve is what keeps that pressure from rising further, which is why it must never be blocked.
Solids, briefly
Solids are classified by what holds them together, and their properties follow:
| Type | Particles | Held by | Examples | Typical properties |
|---|---|---|---|---|
| Molecular | molecules | intermolecular forces | ice, sugar, iodine, dry ice | soft, low melting point, insulating |
| Ionic | ions | ionic bonds | NaCl, CaF₂, MgO | hard, brittle, high melting point; conduct when molten |
| Covalent network | atoms | covalent bonds | diamond, quartz (SiO₂), silicon carbide | very hard, very high melting point |
| Metallic | cations in an electron sea | metallic bonding | copper, iron, aluminium | malleable, conducting, lustrous |
The arrangement of the particles, the crystal structure, is the subject of crystal structures in Lumenary Materials, and phase diagrams there extend the ideas of this lesson to alloys.
The professional view
Equations of state in industry. Designing a gas pipeline, a refrigerator, a liquefied-natural-gas terminal or a carbon-capture plant needs accurate densities and phase boundaries at high pressure. Engineers use cubic equations of state descended from van der Waals', such as Peng–Robinson (1976), for mixtures, and highly accurate reference equations for pure fluids: IAPWS-95 for water, and those built into NIST REFPROP and the free NIST WebBook fluid tables, which reproduce the best measurements to a fraction of a percent. Mechanical engineers meet the same tables in the first law of thermodynamics and power cycles.
Simulating liquids. Liquids have no simple equation like , because their molecules are always in contact. Chemists study them with molecular dynamics: compute the forces between thousands to millions of molecules and integrate Newton's laws in femtosecond steps, much as our 200 000 nitrogen molecules were sampled, but with interactions. Water alone has dozens of competing models, each reproducing some properties better than others; Computational chemistry shows how such force fields are built.
Supercritical and dense CO₂. Captured carbon dioxide is compressed above its critical pressure for pipeline transport and injection into deep rock formations, where pressure keeps it dense. Getting the phase behaviour of impure CO₂ right is a safety issue for these projects, studied at research centres worldwide.
Common misconceptions
Misconception
"At a given temperature all the molecules move at the same speed." They have a broad distribution of speeds. Temperature fixes the average kinetic energy; individual molecules are constantly speeding up and slowing down in collisions.
Misconception
"The bubbles in boiling water are air." The first small bubbles, at the bottom of a pan before boiling, are dissolved air coming out of solution. The big bubbles of a rolling boil are water vapour: water molecules that have escaped into the gas phase inside the liquid.
Misconception
"Water boils at 100 °C." Only at 1 atm. Its boiling point depends on the surrounding pressure: 88 °C in La Paz, 120 °C in a pressure cooker.
Misconception
"Hydrogen bonds are the bonds to hydrogen inside a molecule." The O–H bonds inside a water molecule are covalent. Hydrogen bonds are the much weaker attractions between a hydrogen on one molecule and a lone pair on another.
Misconception
"Ice skates glide because their pressure melts the ice." The calculation does not support it: depending on the assumptions, a skater's weight lowers the melting point by a fraction of a degree to a few degrees, yet skating works perfectly well at −20 °C. The slipperiness of ice comes mainly from a thin, disordered, liquid-like layer at its surface and from frictional heating, and is still an active research topic.
Exercises
1. Gases, liquids and solids
Warm-upHumanWhat volume, in litres, does 2.0 mol of an ideal gas occupy at 300 K and 100 kPa? ( J mol⁻¹ K⁻¹.) Round to the nearest litre.
2. Gases, liquids and solids
CoreHumanAt the same temperature, how does the root-mean-square speed of hydrogen molecules (H₂, 2 g/mol) compare with that of oxygen molecules (O₂, 32 g/mol)?
3. Gases, liquids and solids
CoreHumanWhich of these substances form hydrogen bonds between their own molecules? Type the formulas separated by commas.
H2O, CH4, NH3, HF, CH3OCH3, CH3OH, H2S, HCl
4. Gases, liquids and solids
CoreHumanA block of solid carbon dioxide (dry ice) warms up in a room at 1 atm. What happens, and why?
5. Gases, liquids and solids
CoreHumanUse the Clausius–Clapeyron equation, with kJ/mol and a normal boiling point of 373.12 K at 101.325 kPa, to estimate the temperature (in °C, to the nearest degree) at which water boils inside a pressure cooker at 200 kPa.
6. Gases, liquids and solids
Warm-upHumanWhich hydrogen halide has the highest normal boiling point?
7. Gases, liquids and solids
CoreComputationalWrite boiling_point(p, dH, Tb, pb=101325.0), which uses the Clausius–Clapeyron equation to return the boiling point (in kelvin) at pressure p (Pa) of a liquid with enthalpy of vaporisation dH (J/mol) and boiling point Tb (K) at pressure pb. Use J mol⁻¹ K⁻¹. Solve the equation for directly; no iteration is needed.
8. Gases, liquids and solids
StretchComputationalWrite gas_molar_volume(p, T, a, b), which returns the molar volume (m³/mol) of a gas at pressure p (Pa) and temperature T (K) according to the van der Waals equation, with a in Pa m⁶ mol⁻² and b in m³ mol⁻¹. Multiply the equation out into the cubic , find its real roots with numpy.roots, and return the largest (the gas-like one). Use .
Further reading and resources
- International Association for the Properties of Water and Steam (IAPWS)The international reference formulations for the properties of water and steam, including the saturation-pressure equation used in this lesson.
- NIST Chemistry WebBook: thermophysical properties of fluid systemsAccurate densities, vapour pressures and other properties of common fluids, computed from reference equations of state.
- NIST REFPROPNIST's reference software for the thermodynamic properties of fluids and mixtures, used throughout industry.
- A new two-constant equation of state · Ding-Yu Peng and Donald B. Robinson · 1976The Peng–Robinson equation, still one of the most used cubic equations of state in chemical engineering.
- Over de continuïteit van den gas- en vloeistoftoestand (On the continuity of the gaseous and liquid states) · Johannes Diderik van der Waals · 1873The doctoral thesis that introduced the van der Waals equation; he received the 1910 Nobel Prize in Physics for it.