ConceptIntermediateQuantum computing fundamentalsLesson 14

Decoherence

Decoherence is the physical process by which a quantum system becomes entangled with its environment and loses its relative phases; it is usually characterized by $T_1$ (energy relaxation) and $T_2$ (phase coherence).

Decoherence happens when a qubit interacts with degrees of freedom in its environment that we do not control: it becomes entangled with them, and the reduced state of the qubit alone becomes a mixture. In the language of the density matrix, the off-diagonal elements $\rho_{01}$, which carry the phase and superposition information, shrink, so the superposition gradually turns into a statistical mixture and the ability to interfere disappears. Zurek showed that this process explains how classical behaviour emerges from quantum mechanics, without by itself resolving every question about the interpretation of measurement.

Decoherence in qubits is described by two characteristic times. $T_1$ is the energy relaxation time, i.e. the rate at which $|1\rangle$ decays to $|0\rangle$ by releasing energy to the environment (amplitude damping); the probability of staying in $|1\rangle$ decreases as $e^{-t/T_1}$. $T_2$ is the phase decoherence time, i.e. the off-diagonal elements decay as $e^{-t/T_2}$. They are related by $$\frac{1}{T_2}=\frac{1}{2T_1}+\frac{1}{T_\varphi},$$ where $T_\varphi$ is the pure dephasing time, which gives the constraint $T_2\le2T_1$.

Decoherence should be distinguished from noise. Decoherence is a specific physical process resulting from unwanted coupling to the environment. "Noise" is a broader term covering every deviation from the ideal operation: decoherence, coherent control errors such as imprecise rotation angles, readout errors, crosstalk between qubits, and leakage of the state outside the two levels. So decoherence is one source of noise, not a synonym for it.

The practical consequence: a circuit must finish in a time much shorter than $T_1$ and $T_2$, which is why circuit depth is limited on NISQ devices. Quantum error correction protects information by encoding it in many qubits, provided that the physical error rates are below a certain threshold. The basic simulator in the lab is ideal and free of decoherence unless the noise model is explicitly enabled.

The Bloch sphere lab lets you enable a simplified model with the parameters $T_1$ and $T_2$ to watch the Bloch vector shrink inwards; it is a teaching model that does not describe any specific device.

Mathematical formula

$$\rho_{01}(t)=\rho_{01}(0)\,e^{-t/T_2},\qquad \frac{1}{T_2}=\frac{1}{2T_1}+\frac{1}{T_\varphi}$$

Worked example

A qubit in $|+\rangle$ with $T_1=100\,\mu s$ and $T_2=60\,\mu s$ (satisfying $T_2\le2T_1$). Ignoring frequency detuning, compute its state after $t=30\,\mu s$ using the standard relaxation model.

The transverse Bloch component: $r_x(t)=e^{-t/T_2}=e^{-0.5}\approx0.607$, i.e. $\rho_{01}\approx0.303$ instead of 0.5.

The longitudinal component approaches $|0\rangle$ (where $r_z=+1$): $r_z(t)=1-(1-r_z(0))e^{-t/T_1}=1-e^{-0.3}\approx0.259$.

The length of the Bloch vector is $\sqrt{0.607^2+0.259^2}\approx0.66<1$: the state has become mixed. The probability of getting $+$ in an $X$ measurement is $(1+r_x)/2\approx0.80$ instead of 1.

From the relation above: $1/T_\varphi=1/60-1/200$, i.e. $T_\varphi\approx85.7\,\mu s$.

Test your understanding

The quiz needs JavaScript.

Related topics

References

  1. Wojciech H. Zurek (2003). Decoherence, einselection, and the quantum origins of the classical. Rev. Mod. Phys. 75(3):715–775 doi:10.1103/RevModPhys.75.715
  2. Michael A. Nielsen; Isaac L. Chuang (2010). Quantum Computation and Quantum Information (10th Anniversary Edition). Cambridge University Press doi:10.1017/CBO9780511976667The standard reference for quantum computing; the Crossref record gives 2012 for the Cambridge Core online edition.
  3. John Preskill (2018). Quantum Computing in the NISQ era and beyond. Quantum 2:79 doi:10.22331/q-2018-08-06-79