These exercises practice finite-dimensional tensor products for distinguishable subsystems. They use the two-qubit basis order
∣ 00 ⟩ , ∣ 01 ⟩ , ∣ 10 ⟩ , ∣ 11 ⟩ , \lvert00\rangle,\quad
\lvert01\rangle,\quad
\lvert10\rangle,\quad
\lvert11\rangle, ∣ 00 ⟩ , ∣ 01 ⟩ , ∣ 10 ⟩ , ∣ 11 ⟩ ,
with subsystem A A A written first and subsystem B B B second. Identity factors are written explicitly when an operator acts only on one subsystem.
For identical particles, do not use these exercises as the final word: after forming tensor products one must impose exchange symmetry. See the identical-particle and Fock-space pages for that layer.
All state vectors below are normalized. Helpful background. Product bases explains the coordinate order, and operators on composite systems explains identity factors.
For qubits, use
I = ( 1 0 0 1 ) , X = ( 0 1 1 0 ) , I=
\begin{pmatrix}
1&0\\
0&1
\end{pmatrix},
\qquad
X=
\begin{pmatrix}
0&1\\
1&0
\end{pmatrix}, I = ( 1 0 0 1 ) , X = ( 0 1 1 0 ) ,
and
Y = ( 0 − i i 0 ) , Z = ( 1 0 0 − 1 ) . Y=
\begin{pmatrix}
0&-i\\
i&0
\end{pmatrix},
\qquad
Z=
\begin{pmatrix}
1&0\\
0&-1
\end{pmatrix}. Y = ( 0 i − i 0 ) , Z = ( 1 0 0 − 1 ) .
Thus
Z ∣ 0 ⟩ = ∣ 0 ⟩ , Z ∣ 1 ⟩ = − ∣ 1 ⟩ . Z\lvert0\rangle=\lvert0\rangle,
\qquad
Z\lvert1\rangle=-\lvert1\rangle. Z ∣ 0 ⟩ = ∣ 0 ⟩ , Z ∣ 1 ⟩ = − ∣ 1 ⟩ .
Let H A \mathcal H_A H A have basis { ∣ 0 ⟩ , ∣ 1 ⟩ , ∣ 2 ⟩ } \{\lvert0\rangle,\lvert1\rangle,\lvert2\rangle\} {∣ 0 ⟩ , ∣ 1 ⟩ , ∣ 2 ⟩} and let H B \mathcal H_B H B have basis { ∣ g ⟩ , ∣ e ⟩ } \{\lvert g\rangle,\lvert e\rangle\} {∣ g ⟩ , ∣ e ⟩} . Construct a product basis of H A ⊗ H B \mathcal H_A\otimes\mathcal H_B H A ⊗ H B and give the dimension.
Solution
Pair every basis vector of A A A with every basis vector of B B B . With A A A listed first, a natural order is
∣ 0 g ⟩ , ∣ 0 e ⟩ , ∣ 1 g ⟩ , ∣ 1 e ⟩ , ∣ 2 g ⟩ , ∣ 2 e ⟩ . \lvert0g\rangle,\quad
\lvert0e\rangle,\quad
\lvert1g\rangle,\quad
\lvert1e\rangle,\quad
\lvert2g\rangle,\quad
\lvert2e\rangle. ∣ 0 g ⟩ , ∣ 0 e ⟩ , ∣ 1 g ⟩ , ∣ 1 e ⟩ , ∣ 2 g ⟩ , ∣ 2 e ⟩ .
The dimension is
dim ( H A ⊗ H B ) = 3 ⋅ 2 = 6. \dim(\mathcal H_A\otimes\mathcal H_B)
=
3\cdot2
=
6. dim ( H A ⊗ H B ) = 3 ⋅ 2 = 6.
Expand the product state
( α ∣ 0 ⟩ + β ∣ 1 ⟩ ) ⊗ ( γ ∣ 0 ⟩ + δ ∣ 1 ⟩ ) \bigl(\alpha\lvert0\rangle+\beta\lvert1\rangle\bigr)
\otimes
\bigl(\gamma\lvert0\rangle+\delta\lvert1\rangle\bigr) ( α ∣ 0 ⟩ + β ∣ 1 ⟩ ) ⊗ ( γ ∣ 0 ⟩ + δ ∣ 1 ⟩ )
in the standard two-qubit basis order.
Solution
Distribute the tensor product:
∣ ψ ⟩ = α γ ∣ 00 ⟩ + α δ ∣ 01 ⟩ + β γ ∣ 10 ⟩ + β δ ∣ 11 ⟩ . \lvert\psi\rangle
=
\alpha\gamma\lvert00\rangle
+\alpha\delta\lvert01\rangle
+\beta\gamma\lvert10\rangle
+\beta\delta\lvert11\rangle. ∣ ψ ⟩ = α γ ∣ 00 ⟩ + α δ ∣ 01 ⟩ + β γ ∣ 10 ⟩ + β δ ∣ 11 ⟩ .
In the basis order ∣ 00 ⟩ , ∣ 01 ⟩ , ∣ 10 ⟩ , ∣ 11 ⟩ \lvert00\rangle,\lvert01\rangle,\lvert10\rangle,\lvert11\rangle ∣ 00 ⟩ , ∣ 01 ⟩ , ∣ 10 ⟩ , ∣ 11 ⟩ , the coordinate column is
( α γ α δ β γ β δ ) . \begin{pmatrix}
\alpha\gamma\\
\alpha\delta\\
\beta\gamma\\
\beta\delta
\end{pmatrix}. α γ α δ β γ β δ .
The coefficient array factors as c i j = a i b j c_{ij}=a_i b_j c ij = a i b j , so it has rank one when written as a 2 × 2 2\times2 2 × 2 coefficient matrix.
Decide whether
∣ ψ ⟩ = 1 10 ( ∣ 00 ⟩ + 2 ∣ 01 ⟩ + ∣ 10 ⟩ + 2 ∣ 11 ⟩ ) \lvert\psi\rangle
=
\frac{1}{\sqrt{10}}
\bigl(
\lvert00\rangle
+2\lvert01\rangle
+\lvert10\rangle
+2\lvert11\rangle
\bigr) ∣ ψ ⟩ = 10 1 ( ∣ 00 ⟩ + 2 ∣ 01 ⟩ + ∣ 10 ⟩ + 2 ∣ 11 ⟩ )
is a product state. If it is, factor it.
Solution
The coefficient matrix in the A , B A,B A , B order is
C = 1 10 ( 1 2 1 2 ) . C
=
\frac{1}{\sqrt{10}}
\begin{pmatrix}
1&2\\
1&2
\end{pmatrix}. C = 10 1 ( 1 1 2 2 ) .
The two rows are proportional, so the matrix has rank one. Therefore the state is a product state. One factorization is
∣ ψ ⟩ = ∣ 0 ⟩ + ∣ 1 ⟩ 2 ⊗ ∣ 0 ⟩ + 2 ∣ 1 ⟩ 5 . \lvert\psi\rangle
=
\frac{\lvert0\rangle+\lvert1\rangle}{\sqrt2}
\otimes
\frac{\lvert0\rangle+2\lvert1\rangle}{\sqrt5}. ∣ ψ ⟩ = 2 ∣ 0 ⟩ + ∣ 1 ⟩ ⊗ 5 ∣ 0 ⟩ + 2 ∣ 1 ⟩ .
Multiplying the two normalized factors gives the four amplitudes 1 / 10 1/\sqrt{10} 1/ 10 , 2 / 10 2/\sqrt{10} 2/ 10 , 1 / 10 1/\sqrt{10} 1/ 10 , and 2 / 10 2/\sqrt{10} 2/ 10 .
Compute the matrices of Z ⊗ I Z\otimes I Z ⊗ I , I ⊗ X I\otimes X I ⊗ X , and Z ⊗ Z Z\otimes Z Z ⊗ Z in the standard two-qubit basis order.
Solution
The first operator acts with Z Z Z on the first qubit:
Z ⊗ I = ( 1 0 0 0 0 1 0 0 0 0 − 1 0 0 0 0 − 1 ) . Z\otimes I
=
\begin{pmatrix}
1&0&0&0\\
0&1&0&0\\
0&0&-1&0\\
0&0&0&-1
\end{pmatrix}. Z ⊗ I = 1 0 0 0 0 1 0 0 0 0 − 1 0 0 0 0 − 1 .
The second operator flips the second qubit:
I ⊗ X = ( 0 1 0 0 1 0 0 0 0 0 0 1 0 0 1 0 ) . I\otimes X
=
\begin{pmatrix}
0&1&0&0\\
1&0&0&0\\
0&0&0&1\\
0&0&1&0
\end{pmatrix}. I ⊗ X = 0 1 0 0 1 0 0 0 0 0 0 1 0 0 1 0 .
The product operator multiplies the Z Z Z eigenvalues:
Z ⊗ Z = ( 1 0 0 0 0 − 1 0 0 0 0 − 1 0 0 0 0 1 ) . Z\otimes Z
=
\begin{pmatrix}
1&0&0&0\\
0&-1&0&0\\
0&0&-1&0\\
0&0&0&1
\end{pmatrix}. Z ⊗ Z = 1 0 0 0 0 − 1 0 0 0 0 − 1 0 0 0 0 1 .
Prove that local operators on different subsystems commute:
[ A ⊗ I B , I A ⊗ B ] = 0. [A\otimes I_B,\ I_A\otimes B]=0. [ A ⊗ I B , I A ⊗ B ] = 0.
Solution
Use the product rule for tensor-product operators:
( A ⊗ I B ) ( I A ⊗ B ) = A ⊗ B . (A\otimes I_B)(I_A\otimes B)
=
A\otimes B. ( A ⊗ I B ) ( I A ⊗ B ) = A ⊗ B .
In the opposite order,
( I A ⊗ B ) ( A ⊗ I B ) = A ⊗ B . (I_A\otimes B)(A\otimes I_B)
=
A\otimes B. ( I A ⊗ B ) ( A ⊗ I B ) = A ⊗ B .
The two products are the same, so their difference is zero:
[ A ⊗ I B , I A ⊗ B ] = 0. [A\otimes I_B,\ I_A\otimes B]=0. [ A ⊗ I B , I A ⊗ B ] = 0.
Compute the commutator
[ X ⊗ I , Z ⊗ Z ] . [X\otimes I,\ Z\otimes Z]. [ X ⊗ I , Z ⊗ Z ] .
Solution
Because the second factor is the identity in the first operator,
[ X ⊗ I , Z ⊗ Z ] = [ X , Z ] ⊗ Z . [X\otimes I,\ Z\otimes Z]
=
[X,Z]\otimes Z. [ X ⊗ I , Z ⊗ Z ] = [ X , Z ] ⊗ Z .
The Pauli commutator is
[ X , Z ] = − 2 i Y . [X,Z]
=
-2iY. [ X , Z ] = − 2 iY .
Therefore
[ X ⊗ I , Z ⊗ Z ] = − 2 i Y ⊗ Z . [X\otimes I,\ Z\otimes Z]
=
-2i\,Y\otimes Z. [ X ⊗ I , Z ⊗ Z ] = − 2 i Y ⊗ Z .
This shows a local operator need not commute with an interaction operator.
For real energy parameters ε A \varepsilon_A ε A , ε B \varepsilon_B ε B , and J J J , diagonalize the two-qubit Hamiltonian
H = ε A Z ⊗ I + ε B I ⊗ Z + J Z ⊗ Z . H
=
\varepsilon_A Z\otimes I
+\varepsilon_B I\otimes Z
+J Z\otimes Z. H = ε A Z ⊗ I + ε B I ⊗ Z + J Z ⊗ Z .
Solution
All three terms are diagonal in the standard product basis, so the product basis vectors are eigenvectors. Let z 0 = + 1 z_0=+1 z 0 = + 1 and z 1 = − 1 z_1=-1 z 1 = − 1 . For ∣ a b ⟩ \lvert ab\rangle ∣ ab ⟩ ,
E a b = ε A z a + ε B z b + J z a z b . E_{ab}
=
\varepsilon_A z_a
+\varepsilon_B z_b
+J z_a z_b. E ab = ε A z a + ε B z b + J z a z b .
Thus
E 00 = ε A + ε B + J , E 01 = ε A − ε B − J , E 10 = − ε A + ε B − J , E 11 = − ε A − ε B + J . \begin{aligned}
E_{00}&=\varepsilon_A+\varepsilon_B+J,\\
E_{01}&=\varepsilon_A-\varepsilon_B-J,\\
E_{10}&=-\varepsilon_A+\varepsilon_B-J,\\
E_{11}&=-\varepsilon_A-\varepsilon_B+J.
\end{aligned} E 00 E 01 E 10 E 11 = ε A + ε B + J , = ε A − ε B − J , = − ε A + ε B − J , = − ε A − ε B + J .
The coupling shifts energies according to whether the two Z Z Z eigenvalues are aligned or anti-aligned.
The controlled-phase operator has matrix
U C Z = ( 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 − 1 ) . U_{\mathrm{CZ}}
=
\begin{pmatrix}
1&0&0&0\\
0&1&0&0\\
0&0&1&0\\
0&0&0&-1
\end{pmatrix}. U CZ = 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 − 1 .
Show that it cannot be written as a single product A ⊗ B A\otimes B A ⊗ B of any two one-qubit matrices.
Solution
A product operator takes a product vector to another product vector, or to zero. But controlled phase takes ∣ + ⟩ ⊗ ∣ + ⟩ |+\rangle\otimes|+\rangle ∣ + ⟩ ⊗ ∣ + ⟩ , with ∣ + ⟩ = ( ∣ 0 ⟩ + ∣ 1 ⟩ ) / 2 |+\rangle=(|0\rangle+|1\rangle)/\sqrt2 ∣ + ⟩ = ( ∣0 ⟩ + ∣1 ⟩) / 2 , to a state with coefficient matrix
C = 1 2 ( 1 1 1 − 1 ) . C=\frac12\begin{pmatrix}1&1\\1&-1\end{pmatrix}. C = 2 1 ( 1 1 1 − 1 ) .
Its determinant is − 1 / 2 -1/2 − 1/2 , so its rank is two: the output is entangled. This rules out every product A ⊗ B A\otimes B A ⊗ B , including nondiagonal factors.
A Hamiltonian can be diagonal in a product basis and still generate entanglement. Evolve ∣ + ⟩ ⊗ ∣ + ⟩ |+\rangle\otimes|+\rangle ∣ + ⟩ ⊗ ∣ + ⟩ under H = J Z ⊗ Z H=JZ\otimes Z H = J Z ⊗ Z for J ≠ 0 J\ne0 J = 0 . Find its concurrence and the earliest positive time at which it is maximally entangled.
Solution
Put θ = J t / ℏ \theta=Jt/\hbar θ = J t /ℏ . The even-parity basis states have energy J J J and the odd-parity states have energy − J -J − J , so
∣ ψ ( t ) ⟩ = 1 2 [ e − i θ ( ∣ 00 ⟩ + ∣ 11 ⟩ ) + e i θ ( ∣ 01 ⟩ + ∣ 10 ⟩ ) ] . |\psi(t)\rangle=\frac12\left[
e^{-i\theta}(|00\rangle+|11\rangle)
+e^{i\theta}(|01\rangle+|10\rangle)\right]. ∣ ψ ( t )⟩ = 2 1 [ e − i θ ( ∣00 ⟩ + ∣11 ⟩) + e i θ ( ∣01 ⟩ + ∣10 ⟩) ] .
The coefficient determinant is − i sin ( 2 θ ) / 2 -i\sin(2\theta)/2 − i sin ( 2 θ ) /2 . For a normalized pure two-qubit state, concurrence is twice the absolute value of this determinant. Thus
C ( ψ ( t ) ) = ∣ sin ( 2 J t / ℏ ) ∣ , t f i r s t = π ℏ 4 ∣ J ∣ . C(\psi(t))=|\sin(2Jt/\hbar)|,
\qquad t_{\mathrm{first}}=\frac{\pi\hbar}{4|J|}. C ( ψ ( t )) = ∣ sin ( 2 J t /ℏ ) ∣ , t first = 4∣ J ∣ π ℏ .
The interaction changes relative phases between product-basis components; those phases generally prevent factorization. Having product eigenvectors does not imply that every superposition stays a product.
P. A. M. Dirac, The Principles of Quantum Mechanics , 4th ed., Oxford University Press, 1958.
M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information , 10th anniversary ed., Cambridge University Press, 2010.
J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics , 3rd ed., Cambridge University Press, 2020.
R. Shankar, Principles of Quantum Mechanics , 2nd ed., Springer, 1994.