Open Systems Semigroups

Open-system mathematics has two distinct tasks: describing a legitimate channel at one time and constructing a compatible continuous evolution over many times. This chapter develops the second task without hiding the infinite-dimensional hypotheses behind finite matrices. Its central result is the normal, norm-continuous GKSL theorem; its central warning is that a formal unbounded Lindblad expression does not by itself define a conservative evolution.

The physical models, approximation schemes, and applications live in Measurement and Open Quantum Systems. This chapter adds proofs and operator-theoretic qualifications to those canonical discussions, using the normal-state and measurement tools already developed in this volume.

Enter this chapter

Begin with Trace Class Operators, Normal States, and the mathematical Completely Positive Maps page. You should be able to pair a trace-class state with a bounded observable, distinguish normality from positivity, and interpret positivity after adjoining a finite ancilla.

The Mathematical Open-System Models page then separates four claims: an exact reduced trajectory, a channel on variable inputs, a continuous channel family, and a semigroup. Its composition-defect identity identifies the information discarded when an environment is reset between intervals. A swap model shows why tracing a joint unitary group does not automatically preserve the group or semigroup law.

Channels and dilations come before differential equations

• Quantum Channels in Infinite Dimensions constructs trace-preserving input cutoffs and distinguishes convergence on each state from worst-case convergence. An energy constraint supplies a uniform tail estimate when the energy subspaces are finite dimensional.
• Complete Positivity of Dynamics proves closure under compositions and appropriate limits. It explains why a nontrivial conservative generator is not a positive map and why a naive Euler step can have a negative eigenvalue.
• Stinespring Dilation constructs the dilation Hilbert space from complete positivity, proves minimal uniqueness, and identifies normality as the condition behind a spatial tensor-product environment.
• Kraus Representation Limits treats truncation errors, square-root jump amplitudes, and the absence of an infinite-dimensional normalized maximally entangled vector.

These pages complement the earlier Kraus representation theorem. A list of Kraus operators can describe each finite-time channel while varying nonsmoothly with time; the generator should therefore be analyzed as a derivative of the channel, not inferred from unjustified derivatives of that list.

Continuity determines the generator problem

Read Strong Continuity alongside Quantum Dynamical Semigroups. The first separates continuity of every trace-class orbit from uniform channel continuity. The second distinguishes differentiable master-equation solutions from continuous mild trajectories.

Generators of CP Semigroups proves that the derivative domain is dense and the generator is closed. Time averages regularize states into the domain; the Laplace resolvent produces a channel after normalization. Yosida approximations then give bounded channel semigroups converging to the original evolution on each input.

The GKSL–Lindblad Theorem characterizes normal uniformly continuous dynamics on a separable type-I algebra. A CP jump expansion proves sufficiency of bounded Hamiltonian and noise-intensity data. Countably many noise operators are allowed; their positive completeness sum must converge strongly to a bounded operator.

Bounded and Unbounded Generators marks the boundary of that theorem. Its explosive birth model has zero formal trace derivative on a dense test space but loses probability in the minimal evolution. The missing step is a domain and conservativity theorem, not another algebraic rearrangement.

Divisibility and equilibrium impose different structure

The Mathematical Language of Markovianity examines intermediate-channel existence when a reduced map is noninvertible. A lost input direction cannot later return under divisible evolution. The page also separates one-time divisibility from statements about memory under intermediate interventions.

Detailed Balance fixes a weighted operator inner product before discussing reversibility. GNS and KMS symmetry differ on noncommuting observables. A reversible reset generator has a nonnegative Dirichlet form, while a stationary directed cycle demonstrates that equilibrium alone does not imply detailed balance.

Check your command of the chapter

A useful exit test is to analyze one proposed master equation in four steps. Identify its state space and preparation assumptions; specify the continuity topology; determine the actual generator domain or the bounded theorem that supplies it; then prove complete positivity and conservation for the finite-time evolution. Only afterward classify divisibility, stationary states, and detailed balance.

For a bounded reset model, every step is explicit in this chapter. For an oscillator with unbounded jump operators, the same questions remain meaningful but the bounded GKSL theorem does not answer all of them. That distinction is the starting point for a sound mathematical model.

References

• Engel, Klaus-Jochen, and Rainer Nagel. One-Parameter Semigroups for Linear Evolution Equations. Graduate Texts in Mathematics 194, Springer, 2000. doi:10.1007/b97696. The Banach-space semigroup framework.
• Holevo, Alexander S. Quantum Systems, Channels, Information: A Mathematical Introduction, 2nd ed. De Gruyter, 2019. doi:10.1515/9783110642490. Normal quantum channels and their dynamical structure.
• Lindblad, Göran. “On the generators of quantum dynamical semigroups.” Communications in Mathematical Physics 48, 119–130 (1976). doi:10.1007/BF01608499. The bounded generator characterization.