Pauli and Dirac matrices

[61/56%]
Show
LaTeX

Find the 2×22 \times 2 matrices FF and F′F' from the two equations

(0010)=12(F+iF′),(0100)=12(F−iF′). \begin{pmatrix} 0 & 0 \\ 1 & 0 \end{pmatrix} = \frac{1}{2}(F+iF'), \quad \begin{pmatrix} 0 & 1 \\ 0 & 0 \end{pmatrix} = \frac{1}{2}(F-iF').

Find the anticommutator of FF and F′F', i.e. [F,F′]+≡FF′+F′F[F,F']_{+} \equiv FF'+F'F.

?

Let AA, BB be 2×22 \times 2 matrices over R\mathbb {R}. Find the conditions on AA and BB such that

(A⊗I2)12(1001)=(I2⊗B)12(1001). (A \otimes I_{2}) \frac{1}{\sqrt{2}} \begin{pmatrix} 1 \\ 0 \\ 0 \\ 1 \end{pmatrix} = (I_{2} \otimes B) \frac{1}{\sqrt{2}} \begin{pmatrix} 1 \\ 0 \\ 0 \\ 1 \end{pmatrix}.
?

Let MM be the Minkowski space endowed with the standard coordinates x0,x1,x2,x3x_{0}, x_{1}, x_{2}, x_{3} and the metric tensor field

g=−dx0⊗dx0+dx1⊗dx1+dx2⊗dx2+dx3⊗dx3 g = -dx_{0} \otimes dx_{0} + dx_{1} \otimes dx_{1} + dx_{2} \otimes dx_{2} + dx_{3} \otimes dx_{3}

and the quadratic form

q(x)=−(x0)2+(x1)2+(x2)2+(x3)2. q(x) = -(x_{0})^{2} + (x_{1})^{2} + (x_{2})^{2} + (x_{3})^{2}.

Let H→(2)\overrightarrow {H}(2) be the vector space of 2×22 \times 2 hermitian matrices. Consider the map φ:M→H→(2)\varphi : M \rightarrow \overrightarrow {H}(2)

φ(x)=X=(x0+x3x1−ix2x1+ix2x0−x3) \varphi (x) = X = \begin{pmatrix} x_{0}+x_{3} & x_{1}-ix_{2} \\ x_{1}+ix_{2} & x_{0}-x_{3} \end{pmatrix}

which is an isomorphism.

?
(i)

Find the determinant of XX.

(ii)

Find the Cayley transform of XX, i.e.

U=(X+iI2)−1(X−iI2) U = (X+iI_{2})^{-1}(X-iI_{2})

with the inverse

X=i(I2+U)(I2−U)−1. X = i(I_{2}+U)(I_{2}-U)^{-1}.

Let A,BA, B be two nonzero n×nn \times n matrices over C\mathbb {C}. Let Av→=λv→A\overrightarrow {v}=\lambda \overrightarrow {v} be the eigenvalue equation for AA. Assume that [A,B]=0n\left[A,B\right]=0_{n}. Then from [A,B]v→=0→\left[A,B\right]\overrightarrow {v}=\overrightarrow {0} it follows that

[A,B]v→=(AB−BA)v→=A(Bv→)−B(Av→)=0→. \left[A,B\right]\overrightarrow {v} = (AB-BA)\overrightarrow {v} = A(B\overrightarrow {v})-B(A\overrightarrow {v}) = \overrightarrow {0}.

Therefore A(Bv→)=λ(Bv→)A(B\overrightarrow {v})=\lambda (B\overrightarrow {v}). If Bv→≠0→B\overrightarrow {v} \neq \overrightarrow {0} we find that Bv→B\overrightarrow {v} is an eigenvector of AA with eigenvalue λ\lambda. Apply it to A=σ1⊗σ2A=\sigma_{1} \otimes \sigma_{2} and B=σ3⊗σ3B=\sigma_{3} \otimes \sigma_{3}.

?

Let A,BA, B be two nonzero n×nn \times n matrices over C\mathbb {C}. Let Av→=λv→A\overrightarrow {v}=\lambda \overrightarrow {v} be the eigenvalue equation for AA. Assume that [A,B]+=0n\left[A,B\right]_{+}=0_{n}. Then from [A,B]+v→=0→\left[A,B\right]_{+}\overrightarrow {v}=\overrightarrow {0} it follows that

[A,B]+v→=(AB+BA)v→=A(Bv→)+B(Av→)=0→. \left[A,B\right]_{+}\overrightarrow {v} = (AB+BA)\overrightarrow {v} = A(B\overrightarrow {v})+B(A\overrightarrow {v}) = \overrightarrow {0}.

Therefore A(Bv→)=−λ(Bv→)A(B\overrightarrow {v})=-\lambda (B\overrightarrow {v}). If Bv→≠0→B\overrightarrow {v} \neq \overrightarrow {0} we have an eigenvalue equation with eigenvalue −λ-\lambda. Apply it to A=σ1A=\sigma_{1} and B=σ2B=\sigma_{2}, where [σ1,σ2]+=02\left[\sigma_{1},\sigma_{2}\right]_{+}=0_{2}.

?

Consider the Pauli spin matrix σ2\sigma_{2} with the eigenvalues +1,−1+1, -1 and the corresponding normalized eigenvectors

12(1i),12(1−i). \frac{1}{\sqrt{2}}\begin{pmatrix} 1 \\ i \end{pmatrix}, \qquad \frac{1}{\sqrt{2}}\begin{pmatrix} 1 \\ -i \end{pmatrix}.

Then σ2⊗σ2\sigma_{2} \otimes \sigma_{2} admits the eigenvalues +1+1 (twice) and −1-1 (twice). Show that there are product states as eigenvectors and two sets of entangled eigenvectors for σ2⊗σ2\sigma_{2} \otimes \sigma_{2}.

?

Consider the Hilbert space Cn\mathbb {C}^{n}. Let A,B,CA, B, C be n×nn \times n matrices acting in Cn\mathbb {C}^{n}. We consider the nonlinear eigenvalue problem

Av→=λBv→+λ2Cv→ A\overrightarrow {v} = \lambda B\overrightarrow {v} + \lambda ^{2}C\overrightarrow {v}

where v→∈Cn\overrightarrow {v} \in \mathbb {C}^{n} and v→≠0→\overrightarrow {v} \neq \overrightarrow {0}. Let σ1,σ2,σ3\sigma_{1}, \sigma_{2}, \sigma_{3} be the Pauli spin matrices. Find the solutions of the nonlinear eigenvalue problem

σ1v→=λσ2v→+λ2σ3v→ \sigma _{1}\overrightarrow {v} = \lambda \sigma _{2}\overrightarrow {v} + \lambda ^{2}\sigma _{3}\overrightarrow {v}

where v→∈C2\overrightarrow {v} \in \mathbb {C}^{2} and v→≠0→\overrightarrow {v} \neq \overrightarrow {0}.

?

The four Dirac matrices which play a role in the Dirac equation are given by

γ1=(000−i00−i00i00i000),γ2=(000−100100100−1000), \gamma _1 = \begin{pmatrix} 0 & 0 & 0 & -i \\ 0 & 0 & -i & 0 \\ 0 & i & 0 & 0 \\ i & 0 & 0 & 0 \end{pmatrix}, \qquad \gamma _2 = \begin{pmatrix} 0 & 0 & 0 & -1 \\ 0 & 0 & 1 & 0 \\ 0 & 1 & 0 & 0 \\ -1 & 0 & 0 & 0 \end{pmatrix}, γ3=(00−i0000ii0000−i00),γ4=(1000010000−10000−1). \gamma _3 = \begin{pmatrix} 0 & 0 & -i & 0 \\ 0 & 0 & 0 & i \\ i & 0 & 0 & 0 \\ 0 & -i & 0 & 0 \end{pmatrix}, \qquad \gamma _4 = \begin{pmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & -1 & 0 \\ 0 & 0 & 0 & -1 \end{pmatrix}.
?
(i)

Show that the matrices can be written as Kronecker product of the Pauli spin matrices σ1\sigma_1, σ2\sigma_2, σ3\sigma_3 and the 2×22 \times 2 identity matrix I2I_2.

(ii)

Calculate the anticommutators [γ4,γ1]+[\gamma_4, \gamma_1]_+, [γ4,γ2]+[\gamma_4, \gamma_2]_+, [γ4,γ3]+[\gamma_4, \gamma_3]_+.

Let R∈Cm×mR \in \mathbb {C}^{m \times m} and S∈Cn×nS \in \mathbb {C}^{n \times n} be nontrivial involutions. This means that R=R−1≠±ImR = R^{-1} \neq \pm I_{m} and S=S−1≠InS = S^{-1} \neq I_{n}. A matrix A∈Cm×nA \in \mathbb {C}^{m \times n} is called (R,S)(R, S)-symmetric if RAS=ARAS = A. Consider the case m=n=2m = n = 2 and the Pauli spin matrices

R=(0110),S=(0−ii0). R = \begin{pmatrix} 0 & 1 \\ 1 & 0 \end{pmatrix}, \qquad S = \begin{pmatrix} 0 & -i \\ i & 0 \end{pmatrix}.

Find all 2×22 \times 2 matrices AA over C\mathbb {C} such that RAS=ARAS = A.

?

Let γ∈R\gamma \in \mathbb {R} and γ2<1\gamma^{2}<1. Consider the 2×22 \times 2 matrix

K=σ1−iγσ3. K = \sigma _{1}-i\gamma \sigma _{3}.

Show that the eigenvalues of KK and K∗K^{*} are given by ±1−γ2\pm \sqrt{1-\gamma^{2}}. Find the normalized eigenvectors.

?

Let σ1,σ2,σ3\sigma_{1}, \sigma_{2}, \sigma_{3} be the Pauli spin matrices. We define

τ1=iσ1,τ2=iσ2,τ3=iσ3 \tau _{1} = i\sigma _{1}, \qquad \tau _{2} = i\sigma _{2}, \qquad \tau _{3} = i\sigma _{3}

and τ0=I2\tau_{0} = I_{2}. Thus τj\tau_{j} (j=0,1,2,3j = 0, 1, 2, 3) are elements of the Lie group SU(2)SU(2).

?
(i)

We define the operator RR as (k≥2k \geq 2)

R(τj1⊗τj2⊗⋯⊗τjk):=12(τj1τj2∗⊗τj2τj3∗⊗⋯⊗τjkτj1∗). R(\tau _{j_{1}} \otimes \tau _{j_{2}} \otimes \cdots \otimes \tau _{j_{k}}) := \frac{1}{2}(\tau _{j_{1}}\tau _{j_{2}}^{*} \otimes \tau _{j_{2}}\tau _{j_{3}}^{*} \otimes \cdots \otimes \tau _{j_{k}}\tau _{j_{1}}^{*}).

Find R(τ1⊗τ2⊗τ3)R(\tau_{1} \otimes \tau_{2} \otimes \tau_{3}).

(ii)

Consider the unitary matrix (constructed from the four Bell states)

U=12(100−i0−i−100−i10100i)⇒U∗=12(10010ii00−110i00−i). U = \frac{1}{\sqrt{2}} \begin{pmatrix} 1 & 0 & 0 & -i \\ 0 & -i & -1 & 0 \\ 0 & -i & 1 & 0 \\ 1 & 0 & 0 & i \end{pmatrix} \Rightarrow U^{*} = \frac{1}{\sqrt{2}} \begin{pmatrix} 1 & 0 & 0 & 1 \\ 0 & i & i & 0 \\ 0 & -1 & 1 & 0 \\ i & 0 & 0 & -i \end{pmatrix}.

Show that U(iσ1⊗iσ1)UU(i\sigma_{1} \otimes i\sigma_{1})U, U(iσ2⊗iσ2)UU(i\sigma_{2} \otimes i\sigma_{2})U, U(iσ3⊗iσ3)UU(i\sigma_{3} \otimes i\sigma_{3})U are elements of the Lie group SO(4)SO(4).

References

[StHa17] Steeb W., Hardy Y. Problems and Solutions in Introductory and Advanced Matrix Calculus. 2nd ed., World Scientific Publishing, 2017.