3.2: Partial Measurements
( \newcommand{\kernel}{\mathrm{null}\,}\)
Let us recall how measurements work in single-particle quantum theory. Each observable Q is described by some Hermitian operator \hat{Q}, which has an eigenbasis \{|q_i\rangle\} such that
\hat{Q}|q_i\rangle = q_i |q_i\rangle.
For simplicity, let the eigenvalues \{q_i\} be non-degenerate. Suppose a particle initially has quantum state |\psi\rangle. This can always be expanded in terms of the eigenbasis of \hat{Q}:
|\psi\rangle = \sum_i \psi_i\, |q_i\rangle, \;\;\mathrm{where}\;\;\,\textrm{and}\;\, \psi_i = \langle q_i|\psi\rangle.
The measurement postulate of quantum mechanics states that if we measure Q, then (i) the probability of obtaining the measurement outcome q_i is P_i = |\psi_i|^2, the absolute square of the coefficient of |q_i\rangle in the basis expansion; and (ii) upon obtaining this outcome, the system instantly “collapses” into state |q_i\rangle.
Mathematically, these two rules can be summarized using the projection operator
\hat{\Pi}(q_i) = |q_i\rangle\langle q_i|.
Applying this operator to |\psi\rangle gives the non-normalized state vector
|\psi'\rangle = |q_i\rangle \langle q_i|\psi\rangle.
From this, we glean two pieces of information:
- The probability of obtaining this outcome is \langle\psi'|\psi'\rangle = |\langle q_i|\psi\rangle|^2.
- The post-collapse state is obtained by the re-normalization |\psi'\rangle \rightarrow |q_i\rangle.
For multi-particle systems, there is a new complication: what if a measurement is performed on just one particle?
Consider a system of two particles A and B, with two-particle Hilbert space \mathscr{H}_A \otimes \mathscr{H}_B. We perform a measurement on particle A, corresponding to a Hermitian operator \hat{Q}_A that acts upon \mathscr{H}_A and has eigenvectors \{|\mu\rangle\; |\; \mu = 1, 2, \dots\} (i.e., the eigenvectors are enumerated by some index \mu). We can write any state |\psi\rangle using the eigenbasis of \hat{Q}_A for the \mathscr{H}_A part, and an arbitrary basis \{|\nu\rangle\} for the \mathscr{H}_B part:
\begin{align} \begin{aligned} |\psi\rangle &= \sum_{\mu\nu} \psi_{\mu\nu}\, |\mu\rangle |\nu\rangle \\ &= \sum_\mu |\mu\rangle |\varphi_\mu \rangle, \;\;\;\mathrm{where}\;\;\; |\varphi_\mu\rangle\equiv \sum_\nu \psi_{\mu\nu}\,|\nu\rangle \;\in\; \mathscr{H}_B. \end{aligned} \label{psi_twop}\end{align}
Unlike the single-particle case, the “coefficient” of |\mu_i\rangle in this basis expansion is not a complex number, but a vector in \mathscr{H}_B.
Proceeding by analogy, the probability of obtaining the outcome labelled by \mu should be the “absolute square” of this “coefficient”, \langle\varphi_\mu|\varphi_\mu\rangle. Let us define the partial projector
\hat{\Pi}(\mu) \,=\, |\mu\rangle\langle \mu| \otimes \hat{I}.
The A slot of this operator contains a projector, |\mu\rangle\langle \mu|, while the B slot leaves the \mathscr{H}_B part of the two-particle space unchanged. Applying the partial projector to the state given in Equation \eqref{psi_twop} gives
|\psi'\rangle \,=\, \hat{\Pi}(\mu)\, |\psi\rangle \,=\, |\mu\rangle |\varphi_\mu\rangle.
Now we follow the same measurement rules as before. The outcome probability is
P_\mu = \langle\psi'|\psi'\rangle = \langle \mu|\mu\rangle\, \langle \varphi_\mu|\varphi_\mu\rangle = \sum_\nu |\psi_{\mu\nu}|^2.
The post-measurement collapsed state is obtained by the re-normalization
|\psi'\rangle \;\;\rightarrow\;\; \frac{1}{\sqrt{\sum_{\nu'} |\psi_{\mu\nu'}|^2}}\; \sum_{\nu} \psi_{\mu\nu} |\mu\rangle |\nu\rangle.
Example \PageIndex{1}
A system of two spin-1/2 particles is in the “singlet state”
|\psi\rangle = \frac{1}{\sqrt{2}} \Big(|\!+\!z\rangle|\!-\!z\rangle \,-\, |\!-\!z\rangle|\!+\!z\rangle\Big).
For each particle, |\!+\!z\,\rangle and |\!-\!z\,\rangle denote eigenstates of the operator \hat{S}_z, with eigenvalues +\hbar/2 and -\hbar/2 respectively. Suppose we measure S_z on particle A. What are the probabilities of the possible outcomes, and the associated post-collapse states?
- First outcome: +\hbar/2.
- The partial projector is |\!+\!z\rangle\langle+z| \otimes \hat{I}.
- Applying the projection to |\psi\rangle yields |\psi'\rangle = (1/\sqrt{2})\,|\!+\!z\rangle|\!-\!z\rangle.
- The outcome probability is \displaystyle P_+ = \langle \psi'|\psi'\rangle = \frac{1}{2}.
- The post-collapse state is \displaystyle \frac{1}{\sqrt{P_+}} |\psi'\rangle = |\!+\!z\rangle |\!-\!z\rangle
- Second outcome: -\hbar/2.
- The partial projector is |\!-\!z\rangle\langle-z| \otimes \hat{I}.
- Applying the projection to |\psi\rangle yields |\psi'\rangle = (1/\sqrt{2})\,|\!-\!z\rangle|\!+\!z\rangle.
- The outcome probability is \displaystyle P_- = \langle \psi'|\psi'\rangle = \frac{1}{2}.
- The post-collapse state is \displaystyle = \frac{1}{\sqrt{P_-}} |\psi'\rangle = |\!-\!z\rangle |\!+\!z\rangle.
The two possible outcomes, +\hbar/2 and -\hbar/2, occur with equal probability. In either case, the two-particle state collapses so that A is in the observed spin eigenstate, and B has the opposite spin. After the collapse, the two-particle state is no longer entangled.