Entropy is a thermodynamic state function that quantifies the distribution of energy among accessible microscopic states and provides a rigorous measure of the direction and irreversibility of thermodynamic processes. In classical thermodynamics its differential for a reversible process is defined by the relation between reversible heat transfer and absolute temperature. Entropy is measured in energy per unit temperature, commonly joules per kelvin for an entire system or joules per kelvin per mole for molar entropy. The concept is central to the Second Law of Thermodynamics, which states that the total entropy of an isolated system cannot decrease during a spontaneous process.
In classical thermodynamics Entropy is a state function, meaning its change depends only on the initial and final equilibrium states rather than on the particular path connecting them. Heat itself is not a state function because the amount of heat transferred depends on the process. Entropy change can nevertheless be calculated from a reversible path connecting the same states even when the actual process is irreversible. For an isolated system the entropy change is nonnegative, being zero for an ideal reversible process and positive for an irreversible process. This property gives thermodynamics a directional character absent from the First Law alone.
Statistical Mechanics provides a microscopic interpretation of Entropy. The Boltzmann relation connects entropy with the logarithm of the number of microscopic configurations compatible with a macroscopic state. In this context a macrostate can correspond to many possible microstates, and macrostates with greater multiplicity have greater entropy. The relationship does not simply mean that entropy is a vague measure of disorder because the precise microscopic definition depends on the physical system, accessible states and statistical ensemble. Entropy is fundamentally connected to the number and probability of microscopic configurations consistent with macroscopic constraints.
Entropy is central to explaining irreversibility. Processes such as heat flow from a hot object to a colder object, free expansion of a gas and mixing of substances occur spontaneously under suitable conditions because the total entropy change is positive. The reverse processes are not forbidden by the First Law because energy conservation could permit them, but they would require conditions that produce compensating entropy changes elsewhere. Real processes involve friction, finite temperature differences, diffusion, viscosity, electrical resistance and other dissipative mechanisms that generate entropy. Reversible processes are ideal limiting cases with no net entropy generation.
Entropy changes occur in phase transitions, chemical reactions and changes of temperature or pressure. When a substance melts or vaporises at equilibrium, its entropy changes because the accessible microscopic states change substantially. Vaporisation generally produces a large entropy increase because molecules gain considerably greater translational freedom. For an ideal gas the entropy change can be expressed using changes in temperature and volume or pressure. These relations allow entropy to be calculated from measurable macroscopic variables without explicitly enumerating microscopic states.
Entropy is closely connected to information theory through mathematical relationships between probability distributions and information measures. The connection is particularly important in statistical mechanics, where thermodynamic entropy can be expressed using probabilities assigned to microscopic states. Information theoretic entropy measures uncertainty or information content within a probability distribution, while thermodynamic entropy has a physical interpretation tied to energy, temperature and accessible states. The mathematical similarity is deep, but the physical context and interpretation should not be assumed to be identical in every application.
A common misconception is that Entropy simply means disorder. Disorder can provide an intuitive illustration in some systems but does not constitute a complete thermodynamic definition. Entropy can increase because energy becomes distributed among more microscopic degrees of freedom even when a simple visual description of disorder is inadequate. Another misconception is that entropy of every subsystem must increase. The Second Law applies to the total entropy change of an isolated system, so one subsystem can decrease in entropy provided that another part increases by a sufficiently greater amount.
Entropy plays a major role in determining the maximum performance of heat engines, refrigerators and other thermodynamic devices. The Carnot limit follows from the requirement that entropy generation cannot be negative and establishes an upper bound on efficiency for engines operating between specified temperatures. Entropy generation is therefore also a practical measure of thermodynamic irreversibility. Engineering systems seek to reduce unnecessary entropy generation because it represents lost opportunities to convert available energy into useful work. Entropy consequently connects microscopic statistical behaviour with macroscopic limits on energy conversion and technological performance.
