Quantum Physics in Biquaternionic Form

Introduction

The companion article, Quantum Physics: Foundations and Structure, presented standard quantum physics in its conventional form: states are density matrices on a complex Hilbert space, observables are Hermitian operators, dynamics is $*$-unitary, and measurement is projective. It closed with a checklist of ten structural requirements that any comprehensive quantum theory must satisfy.

This article presents quantum physics in the biquaternion framework. The central object is the Hermitian subspace $\mathbb{B}_{\Psi}$ of the biquaternion algebra $\mathbb{B} = \mathbb{C}\otimes_\mathbb{R}\mathbb{H}$. Its elements are the states and the observables of the theory. This is a structural feature that standard quantum physics does not have. In its algebraic formulation the states and the observables are of different kinds — a normalized positive functional against an operator — and even in the density-matrix formulation, where both are Hermitian matrices, they are told apart by a normalisation and a positivity. In the biquaternion framework both are elements of the same four-dimensional real subspace, the state being the positive element normalised by $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12$ and the observable a general element of the subspace.

For the case of a single qubit, the framework reproduces the standard formalism: the Born rule, the Bloch sphere, and the projective measurement rule follow from the algebra, and the $*$-unitary dynamics has a natural algebraic formulation as rotor conjugation by a one-parameter group generated by a Hamiltonian. The purpose of this article is to develop this reformulation carefully, to state what is established and what is interpretation, and to flag the open questions.

The article is organized as follows. The Hermitian subspace $\mathbb{B}_{\Psi}$ and the two forms it carries are recalled. The pure states are identified with the idempotents and the mixed states with the Bloch ball; the observables are the general Hermitian elements and the Born rule follows from the Hermitian form. The dynamics follows, in the Schrödinger and Heisenberg pictures, then measurement as an algebraic operation, then purity, entropy, and fidelity, then the symmetries. The ten-point checklist is applied, the compatibility with relativity is stated, and the article closes with what is structural, what is new, and the open questions.

The Hermitian Subspace

An element of $\mathbb{B}_{\Psi}$ has the form

$$ \tilde H_\Psi = h_t\, e_t + i h'_x\, \vec e_x + i h'_y\, \vec e_y + i h'_z\, \vec e_z, \qquad h_t, h'_x, h'_y, h'_z \in \mathbb{R}. $$

The real scalar part is $h_t e_t$, and the imaginary vector part is $i(h'_x \vec e_x + h'_y \vec e_y + h'_z \vec e_z)$. In compact form,

$$ \tilde H_\Psi = h_t\, e_t + i\,\vec h_\Psi, \qquad \vec h_\Psi = (h'_x, h'_y, h'_z) \in \mathbb{R}^3. $$

The subspace $\mathbb{B}_{\Psi}$ is a real vector space of dimension 4. Its elements are Hermitian (fixed under the Hermitian conjugation ${}^{*}$).

The Forms on $\mathbb{B}_{\Psi}$

The algebra carries four forms, the scalar parts of the four general products. On $\mathbb{B}_{\Psi}$ they pair off and coincide, so two remain: the Hermitian form, which there equals the plain bilinear form, and the bilinear form $\langle\cdot,\cdot\rangle_{\natural}$, which equals the quaternionic sesquilinear form because complex conjugation and $\natural$ coincide on a Hermitian element.

The Hermitian form. For two Hermitian elements $\tilde{H}_{\Psi,1} = h_{t,1}\,e_t + i\vec h_{\Psi,1}$ and $\tilde{H}_{\Psi,2} = h_{t,2}\,e_t + i\vec h_{\Psi,2}$, it is the scalar part of the product with the conjugate,

$$ \langle\tilde{H}_{\Psi,1},\tilde{H}_{\Psi,2}\rangle_{*} = \mathrm{Sc}(\tilde{H}_{\Psi,1}\tilde{H}_{\Psi,2}^{*}) = h_{t,1} h_{t,2} + \vec h_{\Psi,1}\cdot\vec h_{\Psi,2}, $$

because $\tilde{H}_{\Psi,2}^{*}=\tilde{H}_{\Psi,2}$ on the sector. It is symmetric and positive definite, of inertia $(4,0)$: the Euclidean pairing on $\mathbb{R}^4$, the analogue of the Hilbert–Schmidt inner product, and the form the Born rule and the $*$-unitary evolution use. In the $2\times2$ image the same value appears multiplied by two. On the unit it gives $\langle e_t\rangle_{*}=1$, and the scalar part is the form against the unit, $\mathrm{Sc}(\tilde H_\Psi)=\langle\tilde H_\Psi,e_t\rangle_{*}$.

The biquaternion norm. The Lorentzian form on the sector has the biquaternion norm as its quadratic form,

$$ \langle\tilde H_\Psi\rangle_{\natural} = \tilde H_\Psi\tilde H_\Psi^{\natural} = h_t^2 - |\vec h_\Psi|^2, $$

of inertia $(1,3)$, the Lorentzian form whose future light cone defines the state space; the nonzero elements on the cone are the zero divisors of $\mathbb{B}$ that lie in $\mathbb{B}_{\Psi}$. Symmetrizing the product $\tilde{H}_{\Psi,1}\tilde{H}_{\Psi,2}^{\natural}$ polarizes it,

$$ \langle\tilde{H}_{\Psi,1},\tilde{H}_{\Psi,2}\rangle_{\natural} = h_{t,1} h_{t,2} - \vec h_{\Psi,1}\cdot\vec h_{\Psi,2}, $$

whose diagonal is the norm. The biquaternion norm is a quadratic form and this is its bilinear form; neither is a third object.

Form Expression on $\mathbb{B}_{\Psi}$ Inertia
Hermitian form, equal to the plain form $\langle\tilde{H}_{\Psi,1},\tilde{H}_{\Psi,2}\rangle_{*} = h_{t,1} h_{t,2} + \vec h_{\Psi,1}\cdot\vec h_{\Psi,2}$ $(4,0)$, Euclidean
Biquaternion norm, and its polarization $\langle\tilde H_\Psi\rangle_{\natural} = h_t^2 - \lvert\vec h_\Psi\rvert^2$ and $\langle\tilde{H}_{\Psi,1},\tilde{H}_{\Psi,2}\rangle_{\natural} = h_{t,1} h_{t,2} - \vec h_{\Psi,1}\cdot\vec h_{\Psi,2}$ $(1,3)$, Lorentzian

The Hermitian form is the inner product on the space of operators; the biquaternion norm is the Lorentzian form whose future light cone carries the states.

States

Pure States

A pure state of a qubit is a rank-one projection operator. In the biquaternion framework, the rank-one projections in $\mathbb{B}_{\Psi}$ are the idempotents:

$$ \tilde\Pi_{1,2}(\vec u) = \tfrac{1}{2}\left(e_t \pm i\,\vec u\right), $$

where $\vec u = u_x \vec e_x + u_y \vec e_y + u_z \vec e_z$ is a unit pure real quaternion, $|\vec u| = 1$.

Properties of the idempotents:

  1. Hermitian: $\tilde\Pi_{1,2}^{*} = \tilde\Pi_{1,2}$.
  2. Idempotent: $\tilde\Pi_{1,2}^2 = \tilde\Pi_{1,2}$.
  3. Unit scalar part: $\mathrm{Sc}(\tilde\Pi_{1,2}) = \tfrac12$.
  4. Complementary: $\tilde\Pi_1 + \tilde\Pi_2 = e_t$, and $\tilde\Pi_1 \tilde\Pi_2 = 0 = \tilde\Pi_2 \tilde\Pi_1$.
  5. Zero divisors: $\langle\tilde\Pi_{1,2}\rangle_{\natural} = 0$.

The parametrization is by the unit sphere $S^2 \subset \mathbb{R}^3$: the direction $\vec u$ determines the idempotent $\tilde\Pi_1(\vec u)$ uniquely, and the complementary idempotent $\tilde\Pi_2(\vec u) = \tilde\Pi_1(-\vec u)$ corresponds to the opposite direction.

Identification with standard states. Under the series isomorphism $\mathsf{M}_2:\mathbb{B}\to M_2(\mathbb{C})$ fixed by $\vec e_\eta\mapsto-i\sigma_\eta$ (Algebra and Element Conventions), the idempotent $\tilde\Pi_1(\vec u)$ maps to the standard pure state $|\vec u+\rangle\langle\vec u+|$, the spin-up state along direction $\vec u$. The map is computed in one line: $\mathsf{M}_2(\vec u)=-i\,\vec u\cdot\vec\sigma$, so $i\,\mathsf{M}_2(\vec u)=\vec u\cdot\vec\sigma$ and

$$ \mathsf{M}_2\bigl(\tilde\Pi_1(\vec u)\bigr)=\tfrac{1}{2}\left(I+\vec u\cdot\vec\sigma\right) =\tfrac{1}{2}\begin{pmatrix}1+u_z & u_x-i\,u_y\\ u_x+i\,u_y & 1-u_z\end{pmatrix}, $$

which is the Bloch density matrix of the spin-up state along $\vec u$, reducing to $\mathsf{M}_2(\tilde\Pi_1(\vec e_z))=\operatorname{diag}(1,0)$ on the axis. The set of pure states is the Bloch sphere $S^2$.

Mixed States

A mixed state of a qubit is a positive Hermitian matrix of unit trace. In $\mathbb{B}_{\Psi}$, the corresponding element is normalised by $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12\langle e_t\rangle_{*}$ and has the form

$$ \tilde{Ϛ}_{\Psi} = \tfrac{1}{2}\left(e_t + i\,\vec s_{\Psi}\right), \qquad \vec s_{\Psi} = (s_x, s_y, s_z) \in \mathbb{R}^3. $$

The positivity condition is the condition that the eigenvalues of the corresponding density matrix are non-negative:

$$ |\vec s_{\Psi}|^2 = s_x^2 + s_y^2 + s_z^2 \leq 1. $$

The eigenvalues of $\tilde{Ϛ}_{\Psi}$ are $\lambda_\pm = (1 \pm |\vec s_{\Psi}|)/2$, which are non-negative if and only if $|\vec s_{\Psi}| \leq 1$.

The state is pure if and only if $|\vec s_{\Psi}| = 1$, i.e., if and only if $\tilde{Ϛ}_{\Psi}$ is idempotent. It is mixed if $|\vec s_{\Psi}| < 1$. The maximally mixed state has $\vec s_{\Psi} = 0$, i.e., $\tilde{Ϛ}_{\Psi} = \tfrac{1}{2} e_t$.

The Bloch Ball

The set of all states (pure and mixed) is the Bloch ball of radius one in $\mathbb{R}^3$:

$$ B^3 = \{\vec s_{\Psi} \in \mathbb{R}^3 : |\vec s_{\Psi}| \leq 1\}. $$

The pure states form the boundary sphere $|\vec s_{\Psi}| = 1$; the maximally mixed state is at the centre.

A structural observation. The positivity condition $|\vec s_{\Psi}| \leq 1$ is equivalent to the condition $\langle\tilde{Ϛ}_{\Psi}\rangle_{\natural} \geq 0$ on the biquaternion norm of $\mathbb{B}$:

$$ \boxed{\;\langle\tilde{Ϛ}_{\Psi}\rangle_{\natural} = \tfrac{1}{4}\left(1 - |\vec s_{\Psi}|^2\right) e_t\;}. $$

So the Bloch ball is the intersection of the affine hyperplane $\{\mathrm{Sc} = \tfrac{1}{2}\}$ with the future light cone of the biquaternion norm. The pure states form the boundary of the cone, which is also the set of zero divisors of $\mathbb{B}_{\Psi}$ at the normalisation $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12$.

This is a clean geometric characterization: states are the positive elements of the future light cone of $\mathbb{B}_{\Psi}$ normalised by $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12$; pure states are the normalised elements on the cone itself.

Observables

General Observables

A general observable is a Hermitian element of $\mathbb{B}_{\Psi}$:

$$ \tilde H_\Psi = h_t\, e_t + i\,\vec h_\Psi, \qquad \vec h_\Psi = (h'_x, h'_y, h'_z) \in \mathbb{R}^3. $$

The scalar part $h_t e_t$ carries the number $\mathrm{Sc}(\tilde H_\Psi) = h_t$. The imaginary vector part $i\vec h_\Psi$ has zero scalar part.

The spectral decomposition of $\tilde H_\Psi$ is

$$ \tilde H_\Psi = \lambda_+ \tilde\Pi_1(\vec u_\Psi) + \lambda_- \tilde\Pi_2(\vec u_\Psi), $$

with

$$ \vec u_\Psi = \frac{\vec h_\Psi}{|\vec h_\Psi|}, \qquad \lambda_\pm = h_t \pm |\vec h_\Psi|. $$

The eigenvalues are $\lambda_\pm$, and the eigenvectors are the idempotents $\tilde\Pi_{1,2}(\vec u_\Psi)$. So every Hermitian element of $\mathbb{B}_{\Psi}$ has a spectral decomposition in terms of idempotents.

Observables with no scalar part. The Hermitian elements $i\vec h_\Psi$ (with $h_t = 0$) are the analogs of the scalar-free observables of the qubit (the Pauli matrices). They have eigenvalues $\pm|\vec h_\Psi|$ and eigenvectors $\tilde\Pi_{1,2}(\vec u_\Psi)$.

Compatibility

Two observables $\tilde{H}_{\Psi,1}, \tilde{H}_{\Psi,2} \in \mathbb{B}_{\Psi}$ are compatible (simultaneously measurable) if and only if their corresponding matrices commute. In terms of the vector parts, this holds if and only if $\vec h_{\Psi,1}$ and $\vec h_{\Psi,2}$ are parallel or one of them vanishes.

In the general case (non-parallel, nonzero $\vec h_{\Psi,1}$ and $\vec h_{\Psi,2}$), the observables are incompatible, and their commutator is

$$ [\tilde{H}_{\Psi,1}, \tilde{H}_{\Psi,2}] = -2\left(\vec h_{\Psi,1} \times \vec h_{\Psi,2}\right), $$

where $\vec h_{\Psi,1}\times\vec h_{\Psi,2}$ is the ordinary cross product.

Derivation. We compute

$$ \tilde{H}_{\Psi,1}\tilde{H}_{\Psi,2} = (h_{t,1}h_{t,2} + \vec h_{\Psi,1}\cdot\vec h_{\Psi,2})e_t + i(h_{t,1}\vec h_{\Psi,2} + h_{t,2}\vec h_{\Psi,1}) - \vec h_{\Psi,1}\times\vec h_{\Psi,2}, $$

$$ \tilde{H}_{\Psi,2}\tilde{H}_{\Psi,1} = (h_{t,1}h_{t,2} + \vec h_{\Psi,1}\cdot\vec h_{\Psi,2})e_t + i(h_{t,1}\vec h_{\Psi,2} + h_{t,2}\vec h_{\Psi,1}) + \vec h_{\Psi,1}\times\vec h_{\Psi,2}, $$

so

$$ [\tilde{H}_{\Psi,1}, \tilde{H}_{\Psi,2}] = -2(\vec h_{\Psi,1}\times\vec h_{\Psi,2}). $$

The result is a pure real quaternion (vector part only, with real coefficients): it lies in $\mathbb{B}_{\Phi} \cap \mathbb{B}_{H}$, where $\mathbb{B}_{\Phi}$ is the anti-Hermitian subspace and $\mathbb{B}_{H} = \operatorname{span}_{\mathbb{R}}\{e_t, \vec e_x, \vec e_y, \vec e_z\}$ is the real quaternion subspace, and this intersection is the real span of $\vec e_x, \vec e_y, \vec e_z$. This is the biquaternion form of the $\mathfrak{su}(2)$ angular-momentum commutation relations of the qubit.

The Born Rule

The Born rule gives the probability of an outcome of a measurement. In the biquaternion framework it is a consequence of the Hermitian form $\langle\tilde Q_1,\tilde Q_2\rangle_{*}=\mathrm{Sc}(\tilde Q_1\tilde Q_2^{*})$, positive definite on the whole algebra. Every probability formula below is written with the star: on $\mathbb{B}_{\Psi}$ the star acts as the identity on the arguments, and it is kept because it names the form.

Expectation Values

For an observable $\tilde H_\Psi = h_t e_t + i\vec h_\Psi$ and a state $\tilde{Ϛ}_{\Psi} = \tfrac{1}{2}(e_t + i\vec s_{\Psi})$, the state is normalised by the form against the unit,

$$ \langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12\langle e_t\rangle_{*},\qquad \langle e_t\rangle_{*}=1, $$

and the expectation value is the normalised Hermitian form,

$$ \langle \tilde H_\Psi\rangle_{\tilde{Ϛ}_{\Psi},*} = \frac{\langle\tilde{Ϛ}_{\Psi},\tilde H_\Psi\rangle_{*}}{\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}} =\frac{\mathrm{Sc}(\tilde{Ϛ}_{\Psi}\tilde H_\Psi^{*})}{\mathrm{Sc}(\tilde{Ϛ}_{\Psi})} = h_t + \vec s_{\Psi}\cdot\vec h_\Psi, $$ It is obtained by multiplying out $\tilde{Ϛ}_{\Psi}\tilde H_\Psi^{*}$ and taking the scalar part, with $(i\vec s_{\Psi})(i\vec h_\Psi) = \vec s_{\Psi}\cdot\vec h_\Psi - \vec s_{\Psi}\times\vec h_\Psi$ contributing only its scalar term. The rule is therefore a consequence of the form and of the placement of the states and the observables in $\mathbb{B}_{\Psi}$, not an independent postulate.

Probabilities

For a projective measurement of an observable $\tilde H_\Psi = h_t e_t + i\vec h_\Psi$ in a state $\tilde{Ϛ}_{\Psi} = \tfrac{1}{2}(e_t + i\vec s_{\Psi})$, the two outcomes are $\lambda_\pm = h_t \pm |\vec h_\Psi|$, with probabilities

$$ \boxed{\;p_\pm = \tfrac{1}{2}\left(1 \pm \vec u_\Psi\cdot\vec s_{\Psi}\right)\;}, $$

where $\vec u_\Psi = \vec h_\Psi/|\vec h_\Psi|$. The probabilities depend only on the angle between the measurement direction $\vec u_\Psi$ and the Bloch vector $\vec s_{\Psi}$.

The derivation is direct: $p_+ = \langle\tilde\Pi_1(\vec u_\Psi),\tilde{Ϛ}_{\Psi}\rangle_{*}/\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*} = \mathrm{Sc}(\tilde\Pi_1(\vec u_\Psi)\tilde{Ϛ}_{\Psi}^{*})/\mathrm{Sc}(\tilde{Ϛ}_{\Psi}) = 2\,\mathrm{Sc}(\tilde\Pi_1(\vec u_\Psi)\tilde{Ϛ}_{\Psi}^{*})$, and this is $\tfrac{1}{2}(1 + \vec u_\Psi\cdot\vec s_{\Psi})$.

Measurement Update

After a measurement of $\tilde H_\Psi$ with outcome $+$, the state is updated by the von Neumann–Lüders rule:

$$ \tilde{Ϛ}_{\Psi,\mathrm{out}} = \frac{\tilde\Pi_1(\vec u_\Psi)\,\tilde{Ϛ}_{\Psi}\,\tilde\Pi_1(\vec u_\Psi)}{2\,\mathrm{Sc}(\tilde\Pi_1(\vec u_\Psi)\tilde{Ϛ}_{\Psi}^{*})}. $$

Since $\tilde\Pi_1$ is idempotent and rank one, the two-sided operator is a real multiple of $\tilde\Pi_1$: in $M_2(\mathbb{C})$ a rank-one projection satisfies $P\,R\,P=(\operatorname{Tr}PR)P$, and $\operatorname{Tr}\mathsf M_2(\tilde Q)=2\,\mathrm{Sc}\tilde Q$ reads the multiple back as $p_+ = 2\,\mathrm{Sc}(\tilde\Pi_1\tilde{Ϛ}_{\Psi}^{*})$. Hence the post-measurement state is simply

$$ \boxed{\;\tilde{Ϛ}_{\Psi,\mathrm{out}} = \tilde\Pi_1(\vec u_\Psi)\;}. $$

So the state collapses to the idempotent associated with the outcome. This is the standard collapse, expressed as a composition of algebraic operations.

Normalisation and the Scalar Part

Why the unit weight is $\tfrac12$ and not $1$. The factor is forced by idempotency. Write a rank-one projection as $\tilde\Pi=\alpha(e_t+i\vec u)$ with $\vec u$ a unit pure real quaternion. Since $e_t$ is central and $\vec u^2=-e_t$,

$$ \tilde\Pi^2=\alpha^2\left(e_t^2+2i\vec u+(i\vec u)^2\right)=2\alpha^2\,\tilde\Pi, $$

so $\tilde\Pi^2=\tilde\Pi$ forces $\alpha=\tfrac12$. The alternative $\alpha=1$ gives $\tilde\Pi^2=2\tilde\Pi$, twice a projection and not a projection.

The normalisation is a statement about the scalar part, $\mathrm{Sc}(e_t)=1$. A normalised state has $\mathrm{Sc}(\tilde{Ϛ}_{\Psi})=\tfrac12$, and the Born weight is the ratio of two scalar parts,

$$ p=\frac{\mathrm{Sc}(\tilde{Ϛ}_{\Psi}\tilde H_\Psi^{*})}{\mathrm{Sc}(\tilde{Ϛ}_{\Psi})}, $$

the star being inert because both arguments are Hermitian. The matrix image carries a factor $2$ that the ratio does not see, because it multiplies both terms. The trace belongs to the image alone, $\operatorname{Tr}\mathsf M_2(\tilde Q)=2\,\mathrm{Sc}\tilde Q$, so the normalised state has unit trace, which is the density-matrix form of the statement; the element $\tilde{Ϛ}_{\Psi}$ itself carries no trace, only a scalar part. That the ratio is real, and that it lies in $[0,1]$, are the scalar-part and cone statements of States, Observables and the Born Weight.

Dynamics

The Schrödinger Picture

Time evolution of a pure state. A pure state is an idempotent $\tilde\Pi = \tfrac{1}{2}(e_t + i\vec u)$. It evolves in time by conjugation with a one-parameter group of $*$-unitary elements:

$$ \tilde\Pi(r'_t) = \tilde Q(r'_t)\,\tilde\Pi(0)\,\tilde Q(r'_t)^{*}, $$

where $r'_t=c\,t'$ is the coordinate time carried as a length (Physics Units and Common Elements Conventions) and $\tilde Q(r'_t)$ is a $*$-unitary biquaternion, $\tilde Q\tilde Q^{*} = e_t$, satisfying the Schrödinger equation

$$i\,\hbar\,c\,\frac{d}{dr'_{t}}\tilde Q(r'_t) = \tilde H_\Psi\,\tilde Q(r'_t), \qquad \tilde Q(0) = e_t, $$

with $\tilde H_\Psi = h_t e_t + i\vec h_\Psi \in \mathbb{B}_{\Psi}$ the Hamiltonian, a Hermitian element. The glyph carries the sector and needs no abbreviation: the evolution formulas write $\tilde H_\Psi$, not a bare $H$.

The solution. We have $-i\tilde H_\Psi = -i h_t e_t + \vec h_\Psi$. The scalar part and the vector part commute, so

$$ \boxed{\;\tilde Q(r'_t) = \exp(-i\theta_{\mathrm{evo}})\left(\cos(|\vec h_\Psi|\frac{r'_{t}}{\hbar\,c})\,e_t + \sin(|\vec h_\Psi|\frac{r'_{t}}{\hbar\,c})\,\vec u_\Psi\right)\;}, $$

where $\vec u_\Psi = \vec h_\Psi/|\vec h_\Psi|$ (for $\vec h_\Psi \neq 0$). The exponential of the vector part uses $\vec u_\Psi^2 = -e_t$, which gives the trigonometric form.

The angle is $\theta_{\mathrm{evo}}=h_t\,r'_t/(\hbar\,c)$, the scalar part of $\tilde H_\Psi$ times the interval over $\hbar\,c$, so that the central scalar in the display is $\exp(-i\theta_{\mathrm{evo}})=\exp(-ih_t\,r'_t/(\hbar\,c))$. The generator is anti-Hermitian, so its scalar part is purely imaginary and the angle is real.

The element $\tilde Q(r'_t)$ is $*$-unitary, $\tilde Q(r'_t)\tilde Q(r'_t)^{*} = e_t$, hence an element of the group $\mathrm{U}(2) \subset \mathbb{B}$. It is not Hermitian in general: with $\theta = |\vec h_\Psi|\frac{r'_{t}}{\hbar\,c}$ its conjugate is $\tilde Q(r'_t)^{*} = \exp(+i\theta_{\mathrm{evo}})\left(\cos\theta\,e_t - \sin\theta\,\vec u_\Psi\right)$, which differs from $\tilde Q(r'_t)$ except at isolated times, and the conjugation it performs is therefore not a projection.

The scalar part of the Hamiltonian does not affect the state. Since $\exp(-i\theta_{\mathrm{evo}})$ is a central complex scalar, and since the conjugation

$$ \tilde Q(r'_t)\,\tilde\Pi\,\tilde Q(r'_t)^{*} = \exp(-i\theta_{\mathrm{evo}})\tilde u\,\tilde\Pi\,\exp(+i\theta_{\mathrm{evo}})\tilde u^{*} = \tilde u\,\tilde\Pi\,\tilde u^{*} $$

(with $\tilde u = \cos(|\vec h_\Psi|\frac{r'_{t}}{\hbar\,c})e_t + \sin(|\vec h_\Psi|\frac{r'_{t}}{\hbar\,c})\vec u_\Psi$ a unit real quaternion), the phase $\exp(-i\theta_{\mathrm{evo}})$ cancels out. So only the vector part $\vec h_\Psi$ of the Hamiltonian affects the state evolution. This is the standard result: the scalar part of the Hamiltonian contributes only to a global, unobservable phase.

Time evolution of a mixed state. A general state $\tilde{Ϛ}_{\Psi} = \tfrac{1}{2}(e_t + i\vec s_{\Psi})$ evolves by the same conjugation:

$$ \tilde{Ϛ}_{\Psi}(r'_t) = \tilde Q(r'_t)\,\tilde{Ϛ}_{\Psi}(0)\,\tilde Q(r'_t)^{*}. $$

This is the von Neumann equation, expressed as a rotor conjugation. In differential form, it is

$$i\,\hbar\,c\,\frac{d}{dr'_{t}}\tilde{Ϛ}_{\Psi}(r'_t) = [\tilde H_\Psi, \tilde{Ϛ}_{\Psi}(r'_t)], $$

with the commutator defined in the biquaternion algebra.

Interpretation. The $*$-unitary $\tilde Q(r'_t)$ is an inner automorphism of the algebra: conjugation by a $*$-unitary biquaternion preserves multiplication. Its action on $\mathbb{B}_{\Psi}$ keeps the scalar part fixed and rotates the imaginary vector part, giving the standard precession of a spin-1/2 in a magnetic field, expressed as the biquaternion rotation.

The Heisenberg Picture

In the Heisenberg picture, the state is fixed and the observables evolve. For fixed Hermitian $\tilde O_\Psi\in\mathbb{B}_{\Psi}$, the evolved observable is

$$ \tilde O_\Psi(r'_t) = \tilde Q(r'_t)^{*}\,\tilde O_\Psi\,\tilde Q(r'_t). $$

The evolved observable takes the glyph $\tilde O_\Psi$ and the Hamiltonian keeps $\tilde H_\Psi$, so that the generator and the element it moves do not share a letter.

The evolution equation is the Heisenberg equation

$$i\,\hbar\,c\,\frac{d}{dr'_{t}}\tilde O_\Psi(r'_t) = [\tilde O_\Psi(r'_t), \tilde H_\Psi], $$

where $\tilde H_\Psi$ is the Hamiltonian.

Derivation. We compute the time derivative of $\tilde O_\Psi(r'_t) = \tilde Q^{*}\tilde O_\Psi\tilde Q$, using the Schrödinger equation $i\,\hbar\,c\,\frac{d\tilde Q}{dr'_{t}}= \tilde H_\Psi\tilde Q$ and its adjoint $-i\,\hbar\,c\left(d\tilde Q/dr'_t\right)^{*} = \tilde Q^{*}\tilde H_\Psi$:

$$ i\,\hbar\,c\,\frac{d}{dr'_{t}}\tilde O_\Psi =i\,\hbar\,c\left(\left(d\tilde Q/dr'_t\right)^{*}\tilde O_\Psi\tilde Q + \tilde Q^{*}\tilde O_\Psi\,d\tilde Q/dr'_t\right) = -\tilde Q^{*}\tilde H_\Psi\tilde O_\Psi\tilde Q + \tilde Q^{*}\tilde O_\Psi\tilde H_\Psi\tilde Q = \tilde Q^{*}[\tilde O_\Psi, \tilde H_\Psi]\tilde Q. $$

Since $\tilde Q(r'_t) = \exp(-i\tilde H_\Psi r'_t/(\hbar\,c))$ is a function of $\tilde H_\Psi$, it commutes with $\tilde H_\Psi$, and so does $\tilde Q^{*}$. The last expression is therefore $[\tilde Q^{*}\tilde O_\Psi\tilde Q, \tilde H_\Psi] = [\tilde O_\Psi(r'_t), \tilde H_\Psi]$, which is the Heisenberg equation. For the choice $\tilde O_\Psi=\tilde H_\Psi$ the right-hand side vanishes: the Hamiltonian is conserved under its own flow.

The two pictures are related by a $*$-unitary transformation and give the same physical predictions.

Reversibility

The evolution generated by a Hermitian Hamiltonian is $*$-unitary: the element $\tilde Q(r'_t)$ satisfies $\tilde Q(r'_t)\tilde Q(r'_t)^{*} = e_t$. The evolution is therefore reversible: the state at any later time can be recovered from an earlier time by applying $\tilde Q(r'_t)^{*}$.

Reversibility is one of the structural features of the biquaternion formulation. The $*$-unitary elements form the group $\mathrm{U}(2) \subset \mathbb{B}$, and their conjugation action on $\mathbb{B}_{\Psi}$ preserves the Hermitian form and the biquaternion norm.

Measurement as an Algebraic Operation

In the standard framework, the measurement postulate is stated separately from the $*$-unitary evolution. In the biquaternion framework, both are consequences of the same structure, the Hermitian form on $\mathbb{B}$, which restricts to the plain form on $\mathbb{B}_{\Psi}$.

Projective measurement. A projective measurement of $\tilde H_\Psi = h_t e_t + i\vec h_\Psi$ is specified by the spectral decomposition $\tilde H_\Psi = \lambda_+ \tilde\Pi_1 + \lambda_- \tilde\Pi_2$, with $\tilde\Pi_{1,2} = \tfrac{1}{2}(e_t \pm i\vec u_\Psi)$. The outcome $+$ has:

  • Probability $p_+=\langle\tilde\Pi_1,\tilde{Ϛ}_{\Psi}\rangle_{*}/\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\mathrm{Sc}(\tilde\Pi_1\tilde{Ϛ}_{\Psi}^{*})/\mathrm{Sc}(\tilde{Ϛ}_{\Psi})$, which at normalisation is $2\langle\tilde\Pi_1,\tilde{Ϛ}_{\Psi}\rangle_{*}$.
  • Post-measurement state $\tilde{Ϛ}_{\Psi,\mathrm{out}} = \tilde\Pi_1$, from the two-sided operation.

A structural point. The reversible–irreversible distinction is algebraic in the biquaternion framework. $*$-Unitary elements $\tilde Q$ preserve the biquaternion norm of $\mathbb{B}$ and generate reversible evolution. Idempotent elements $\tilde\Pi$ do not preserve the biquaternion norm (they satisfy $\tilde\Pi^2 = \tilde\Pi$ but $\langle\tilde\Pi\rangle_{\natural} = \tilde\Pi\tilde\Pi^{\natural} = 0$) and generate irreversible projection. The two kinds of process are distinguished by a property of the acting element, not by a separate postulate.

What the rewriting does not do. The update says which state follows a given outcome; it does not say why that outcome rather than another occurs. The selection problem is untouched by the algebra and is the same problem as in the standard formalism, stated in different letters (question 6 below).

Purity, Entropy, and Fidelity

Purity

The purity of a state $\tilde{Ϛ}_{\Psi} = \tfrac{1}{2}(e_t + i\vec s_{\Psi})$ is

$$ 2\,\mathrm{Sc}(\tilde{Ϛ}_{\Psi}^2) = \tfrac{1}{2}\left(1 + |\vec s_{\Psi}|^2\right), $$

which ranges from $\tfrac{1}{2}$ (maximally mixed, $\vec s_{\Psi} = 0$) to $1$ (pure, $|\vec s_{\Psi}| = 1$). The deviation of the element from being idempotent is the same number read the other way,

$$ \tilde{Ϛ}_{\Psi}^2 - \tilde{Ϛ}_{\Psi} = \tfrac{1}{4}\left(|\vec s_{\Psi}|^2 - 1\right)e_t, $$

so purity is the condition $\tilde{Ϛ}_{\Psi}^2 = \tilde{Ϛ}_{\Psi}$: the pure states are the fixed points of the squaring map.

Spectral Decomposition and Entropy

The spectral decomposition of $\tilde{Ϛ}_{\Psi}$ is

$$ \tilde{Ϛ}_{\Psi} = \lambda_+ \tilde\Pi_1(\vec u_s) + \lambda_- \tilde\Pi_2(\vec u_s), $$

with $\vec u_s = \vec s_{\Psi}/|\vec s_{\Psi}|$ and $\lambda_\pm = (1 \pm |\vec s_{\Psi}|)/2$. A function of the element is the function of its eigenvalues on its own idempotents, $\log\tilde{Ϛ}_{\Psi}=\log\lambda_+\,\tilde\Pi_1(\vec u_s)+\log\lambda_-\,\tilde\Pi_2(\vec u_s)$, with the extension by continuity at a vanishing eigenvalue. The von Neumann entropy is

$$ S(\tilde{Ϛ}_{\Psi}) = -2\,\mathrm{Sc}(\tilde{Ϛ}_{\Psi}\log\tilde{Ϛ}_{\Psi}) = -\lambda_+ \log \lambda_+ - \lambda_- \log \lambda_-, $$

which depends only on $|\vec s_{\Psi}|$ and vanishes for pure states.

Fidelity

For two pure states $\tilde\Pi_{\mathrm p} = \tfrac{1}{2}(e_t + i\vec u_{\mathrm p})$ and $\tilde\Pi_{\mathrm q} = \tfrac{1}{2}(e_t + i\vec u_{\mathrm q})$, the transition probability is

$$ 2\,\mathrm{Sc}(\tilde\Pi_{\mathrm p}\tilde\Pi_{\mathrm q}^{*}) = \tfrac{1}{2}\left(1 + \vec u_{\mathrm p}\cdot\vec u_{\mathrm q}\right), $$

which is the standard $\cos^2(\theta/2)$ in biquaternion letters, where $\theta$ is the angle between $\vec u_{\mathrm p}$ and $\vec u_{\mathrm q}$ on the Bloch sphere.

For mixed states the Uhlmann fidelity is the same scalar part with a square root, $2\,\mathrm{Sc}\sqrt{\tilde{Ϛ}_{\Psi,1}^{1/2}\tilde{Ϛ}_{\Psi,2}\tilde{Ϛ}_{\Psi,1}^{1/2}}$, and it closes in two dimensions because $\mathrm{Sc}\langle\tilde{Ϛ}_{\Psi}\rangle_{\natural}=\tfrac{1}{4}(1-|\vec s_{\Psi}|^2)$ is the determinant:

$$ F(\tilde{Ϛ}_{\Psi,1},\tilde{Ϛ}_{\Psi,2}) = 2\,\mathrm{Sc}(\tilde{Ϛ}_{\Psi,1}\tilde{Ϛ}_{\Psi,2}^{*}) + 2\sqrt{\mathrm{Sc}\bigl(\langle\tilde{Ϛ}_{\Psi,1}\rangle_{\natural}\langle\tilde{Ϛ}_{\Psi,2}\rangle_{\natural}\bigr)} = \tfrac{1}{2}\left(1+\vec s_{\Psi,1}\cdot\vec s_{\Psi,2}+\sqrt{(1-|\vec s_{\Psi,1}|^2)(1-|\vec s_{\Psi,2}|^2)}\right). $$

On pure states the square root vanishes and the fidelity is the transition probability above; on two maximally mixed states it gives $F=1$. The star stays explicit although it is the identity on $\mathbb{B}_{\Psi}$, so that the overlap is read as the Hermitian form.

Symmetries

The $*$-Unitary Group

The $*$-unitary group $\mathrm{U}(2)$ acts on $\mathbb{B}_{\Psi}$ by conjugation:

$$ \tilde{Ϛ}_{\Psi} \;\longmapsto\; \tilde Q\,\tilde{Ϛ}_{\Psi}\,\tilde Q^{*}, \qquad \tilde Q \in \mathrm{U}(2). $$

This action preserves the Hermitian property and the scalar part, so it maps states to states and observables to observables.

Unit Quaternions and Spin Rotations

The unit quaternions (elements of $\mathbb{B}_{H}$ with unit quaternion norm) form the subgroup $SU(2) \subset \mathrm{U}(2)$. Their conjugation action on $\mathbb{B}_{\Psi}$ preserves the scalar part and rotates the imaginary vector part, giving the rotation group $SO(3)$ acting on the Bloch sphere. This is the standard spin-1/2 rotation group, expressed in the biquaternion framework.

The Lie Algebra

The Lie algebra of the $*$-unitary group is the space of anti-Hermitian elements, $\mathbb{B}_{\Phi}$, with the commutator bracket:

$$ [\tilde{Q}_1, \tilde{Q}_2] = \tilde{Q}_1\tilde{Q}_2 - \tilde{Q}_2\tilde{Q}_1, \qquad \tilde{Q}_1, \tilde{Q}_2 \in \mathbb{B}_{\Phi}. $$

This is a Lie algebra structure on $\mathbb{B}_{\Phi}$ (which is not a subalgebra of $\mathbb{B}$, but is a Lie algebra under the commutator). The exponential map sends $\mathbb{B}_{\Phi}$ to the group of $*$-unitary biquaternions.

So the subspace $\mathbb{B}_{\Psi}$ carries the states and observables; the subspace $\mathbb{B}_{\Phi}$ carries the generators of the symmetry group. The two subspaces are complementary in the Hermitian decomposition of $\mathbb{B}$.

Application of the Ten-Point Checklist

The companion article listed ten structural requirements for a comprehensive quantum theory. Each is met here.

1. State space. The states are the positive elements of $\mathbb{B}_{\Psi}$ of unit weight, $2\,\mathrm{Sc}=1$: a canonical object of the algebra, not a postulate.

2. Observables. The observables are the Hermitian elements of $\mathbb{B}_{\Psi}$, with idempotents as eigenprojectors and the eigenvalues as the readings.

3. Born rule. The rule is the normalised Hermitian form $\langle\tilde{Ϛ}_{\Psi},\tilde H_\Psi\rangle_{*}/\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}$, a consequence of the form and not a separate postulate; the factor $2$ of the $2\times2$ image cancels in the ratio.

4. Dynamics. The dynamics is rotor conjugation by $*$-unitary biquaternions, with the Schrödinger equation for the rotor and the von Neumann equation for the state.

5. Measurement rule. Probability comes from the normalised Hermitian form and the update from the two-sided operation; the reversible–irreversible distinction is algebraic, $*$-unitary against idempotent.

6. Composition rule. Two qubits give $\mathbb{B}\otimes_\mathbb{C}\mathbb{B} \cong M_4(\mathbb{C})$, and the reduction over one factor is the doubled partial scalar part,

$$ \tilde{Ϛ}_{\Psi,A} = 2\,\mathrm{Sc}_2\bigl(\tilde{Ϛ}_{\Psi,AB}\bigr), $$

which is the partial trace on the image, $\operatorname{Tr}_2(a\otimes b)=2a\,\mathrm{Sc}(b)$. That the Hermitian elements of $\mathbb{B}\otimes\mathbb{B}$ exhaust those of $M_4(\mathbb{C})$ is question 1 below.

7. Symmetry group. The $*$-unitary group $\mathrm{U}(2)$ acts by conjugation, with $SU(2)$ generating the spin rotations and $SO(3)$ acting on the Bloch sphere.

8. Relativistic extension. The Lorentz group sits in the invertible elements through its double cover $SL(2,\mathbb{C})$, and $\mathbb{B}_{\Phi}$ carries the four-vectors. The formalism of this article is the non-relativistic restriction, its state space a fixed Bloch ball that a boost moves a state off; the relativistic development is in The Dirac Equation in Biquaternionic Form and in Canonical Quantization of the Biquaternion Dirac Field.

9. Classical limit. The limit of the qubit formalism is the classical spin vector, reached in a pure state under a precessional dynamics. The general limit needs decoherence and the correspondence principle, treated in Decoherence as Idempotent Projection and in The Lagrangian and the Non-Quantum Limit: The Action Principle.

10. Interpretation. The natural reading is algebraic: the states and the observables are elements of $\mathbb{B}_{\Psi}$ and the quantum structure follows from the algebra. It is compatible with the Copenhagen, relational, and informational readings without committing to any of them.

Compatibility with Relativity

Two halves, two physics. The algebra is one eight-dimensional real space, cut in two. One half, $\mathbb{B}_{\Psi}$, is the space of quantum physics: its elements are the states and the observables of this whole article, a real weight and an imaginary direction. The other half, $\mathbb{B}_{\Phi}$, is the space of relativity: its elements are the four-vectors, one real time and three real directions, and its norm is the interval.

The cut is the cut of the star. $\mathbb{B}_{\Psi}$ is the part the star fixes and $\mathbb{B}_{\Phi}$ the part it reverses, so $\mathbb{B}=\mathbb{B}_{\Psi}\oplus\mathbb{B}_{\Phi}$ is a direct sum, each half four real dimensional, the two meeting only at zero. Every element of the algebra is therefore one quantum part plus one relativistic part, and multiplication by the central $i$ carries one half onto the other. Because the two halves are complementary, the whole algebra is available to both readings at once, and neither reading has to be added from outside.

The coefficient slots show the difference. The element splits as $\tilde Q=\tilde Q_\Psi+\tilde Q_\Phi$ with

$$ \tilde Q_\Psi = q_t\,e_t + i\bigl(q'_x\vec e_x+q'_y\vec e_y+q'_z\vec e_z\bigr)\in\mathbb{B}_{\Psi}, \qquad \tilde Q_\Phi = i\,q'_t\,e_t + q_x\vec e_x + q_y\vec e_y + q_z\vec e_z\in\mathbb{B}_{\Phi}, $$

all coefficients real: the quantum half has the real scalar slot, the relativistic half the imaginary one. That single sign is what makes the two norms different, the same quaternionic norm being positive definite on $\mathbb{B}_{\Psi}$ and indefinite on $\mathbb{B}_{\Phi}$,

$$ \tilde Q_\Phi\tilde Q_\Phi^{\natural} = -q_t'^2 + q_x^2 + q_y^2 + q_z^2, $$

three directions positive and the temporal one negative, because the time coefficient carries the central imaginary and its square is negative. The zero set is the light cone, and the maps that preserve the interval are the Lorentz transformations. A state needs the positive definite form, and a four-vector needs the indefinite one; the two halves give one each.

One action, two groups. Both theories act by conjugation,

$$ \tilde Q_1 \;\longmapsto\; \tilde Q\,\tilde Q_1\,\tilde Q^{*}, \qquad \tilde Q \in \mathbb{B} \ \ \text{invertible}, $$

and they differ only in the subgroup over which the conjugating element ranges. The quantum operations are the $*$-unitary case, $\tilde Q\tilde Q^{*} = e_t$, forming $\mathrm{U}(2)$. The proper orthochronous Lorentz transformations of $\mathbb{B}_{\Phi}$ come from the elements of unit norm, the Lorentz rotors $\tilde U_{\mathrm{lor}}$ with $\tilde U_{\mathrm{lor}}\tilde U_{\mathrm{lor}}{}^{\natural} = e_t$, forming $SL(2,\mathbb{C})$, the double cover of that group. The two conditions are genuinely different and neither group contains the other, yet both act by the same formula on the same space, and their intersection is the rotation group:

$$ SU(2) \subset \mathrm{U}(2), \qquad SU(2) \subset SL(2,\mathbb{C}), \qquad \mathrm{U}(2) \cap SL(2,\mathbb{C}) = SU(2). $$

The intersection is the set of unit real quaternions. Its conjugation rotates the Bloch vector of a state in $\mathbb{B}_{\Psi}$ and the spatial three-vector of a four-vector in $\mathbb{B}_{\Phi}$ by the same $SO(3)$ rotation, with the same rotor, so the spin rotation of the qubit and the spatial rotation of the material sector are one algebraic operation. What separates the full groups is the character of the transformation: a $*$-unitary conjugation is an automorphism of $\mathbb{B}$, since it preserves multiplication, while a non-$*$-unitary one is only a congruence. The Euclidean Hermitian form is preserved by $\mathrm{U}(2)$ alone; the Lorentzian biquaternion norm is preserved by both, because $|\langle\tilde Q\rangle_{\natural}| = 1$ for a $*$-unitary element.

A boost is not a quantum operation. A pure boost has a Hermitian rotor, hence lies in $\mathbb{B}_{\Psi}$ itself, the complement of the sector it acts on: the rotation generators are material and the boost generators informational. With rapidity $\theta_{\mathrm{rpd}}$, $\tanh\theta_{\mathrm{rpd}}=v/c$,

$$ \tilde U_{\Psi,\mathrm b} = \cosh\frac{\theta_{\mathrm{rpd}}}{2}\,e_t + i\sinh\frac{\theta_{\mathrm{rpd}}}{2}\,\vec u_{\mathrm b}, \qquad \langle\tilde U_{\Psi,\mathrm b}\rangle_{\natural} = 1 . $$

It lies on the norm-one hyperboloid, whereas the states lie on the state slice of the future light cone, with $\langle\tilde{Ϛ}_{\Psi}\rangle_{\natural} = \tfrac{1}{4}(1 - |\vec s_{\Psi}|^2) \geq 0$ and the pure states on the cone. Boost conjugation preserves Hermiticity, positivity, and the biquaternion norm, but neither the normalisation $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12$ nor the plain form, so it moves the state off the slice: the image of the maximally mixed state has weight $\cosh\theta_{\mathrm{rpd}}$, and the physical state is recovered by renormalizing. This is the algebraic form of a familiar fact, that the finite-dimensional Lorentz transformation of a state is not $*$-unitary. The product of two boosts is not a boost: it carries a $*$-unitary factor, the Thomas–Wigner rotation, and there the relativistic group generates an element of the quantum group.

The shared imaginary unit. The $i$ of the Schrödinger equation, the $i$ of the phase $\exp(-i\theta_{\mathrm{evo}})$, and the $i$ of the material coordinate $i r'_t$ are one and the same element, the scalar imaginary. The generator of the quantum phase and the Lorentzian sign of the interval are therefore two consequences of one complex structure, and not independent conventions.

What the compatibility does and does not provide. The framework supplies a common space and one conjugation formula, with the rotation group as the intersection of the two groups. It does not supply, by itself, a relativistic quantum dynamics: a Lorentz-covariant state space, microcausality, and the many-particle structure lie outside the single-qubit formalism, whose state space is a fixed Bloch ball. Those belong to the biquaternion Dirac equation and to its canonical quantization, where the state space, and not only the transformation group, becomes relativistic.

What Is Structural and What Is New

What is standard QM, reformulated. The qubit state space (Bloch ball), the Born rule as the normalised Hermitian form $\langle\tilde{Ϛ}_{\Psi},\tilde H_\Psi\rangle_{*}/\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}$, the dynamics as rotor conjugation, projective measurement as the two-sided operation, the spectral decomposition in idempotents, and the standard forms of entropy and fidelity.

What is structurally new or newly visible.

  1. States and observables live in the same subspace, and positivity is a norm condition. In the standard formalism the two are of different kinds, or Hermitian matrices told apart by a normalisation and a positivity. In $\mathbb{B}_{\Psi}$ both are elements of one four-dimensional real subspace: the state is the positive element normalised by $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12$, the observable a general element, so the families are nested and a state is itself an observable. The condition $|\vec s_{\Psi}|\le1$ is the sector's norm condition $\langle\tilde{Ϛ}_{\Psi}\rangle_{\natural}\ge0$.

  2. The state space is canonical. The idempotents of $\mathbb{B}_{\Psi}$ are canonical objects of the algebra, and the Bloch sphere is their parametrization by $S^2$. The Bloch ball is the intersection of the normalisation hyperplane with the future light cone of the biquaternion norm.

  3. The Born rule and the state update are algebraic. Both follow from the Hermitian form and from the two-sided operation, not from independent postulates, and the reversible–irreversible distinction is a property of the acting element.

  4. The symmetry group and the observables share the same space. The $*$-unitary elements of $\mathbb{B}$ act on $\mathbb{B}_{\Psi}$ by conjugation, and their generators lie in $\mathbb{B}_{\Phi}$.

  5. The framework contains the algebraic structure for relativity. The Lorentz group is in the algebra as the group of biquaternions of unit norm, and $\mathbb{B}_{\Phi}$ carries the four-vectors, so the quantum formalism and the relativistic structure coexist in one space.

Open Questions

The idempotent formalism raises several questions.

1. Extension to many qubits. The formalism above describes a single qubit. The extension to $n$ qubits requires the tensor product $\mathbb{B}^{\otimes_\mathbb{C} n}$ (over $\mathbb{C}$), which is isomorphic to $M_{2^n}(\mathbb{C})$. How do the idempotents, states, and observables generalize? Is the tensor product natural in the biquaternion framework, or does it require additional structure? The tensor product, the reduction over one factor and the entangled two-qubit states are worked out in Entangled Subsystems in the Biquaternion Framework, and the reduced state is computed in Exercise: Entanglement Entropy and the Partial Trace.

2. Second quantization. The formalism is first-quantized. To describe creation and annihilation of particles, one needs a Fock space and operator-valued fields. How does the biquaternion framework extend to this setting? The biquaternion Fock space is built, with its creation and annihilation operators, in Fock Space and Creation/Annihilation Operators in Biquaternionic Form.

3. The full dynamics. The dynamics described here is for a qubit with a fixed Hamiltonian. The general dynamics of an open quantum system — with dissipation, decoherence, and feedback — is described by the Lindblad equation, which extends the von Neumann equation with additional terms. How does the Lindblad equation read in the biquaternion framework? It is written in the algebra in Quantum Channels and the Reversible/Irreversible Dichotomy, §The Lindblad Form.

4. Relativistic quantum theory. The biquaternion algebra contains the Lorentz group and the spinor representation, and the biquaternion Dirac equation describes a relativistic spin-1/2 field classically. The quantization of this field, with a Lorentz-covariant state space and a consistent many-particle structure, is carried out in Canonical Quantization of the Biquaternion Dirac Field, on the classical equation of The Dirac Equation in Biquaternionic Form. The relation between the single-qubit quantum formalism presented here and the relativistic spinor-field theory remains to be worked out in detail.

5. The interpretation of the informational sector. The companion article States, Observables and the Born Weight proposes an informational interpretation of the Hermitian subspace. Under this interpretation, the idempotents are "pure information states" and the Hermitian elements are "informational observables". The Born rule is the "probability of extracting information". Whether this leads to new structure is an open question.

6. The measurement problem. The idempotent formalism makes the Born rule and the state update algebraic, but it does not explain why one idempotent is selected in a given measurement. This is the same measurement problem as in standard quantum physics, in different notation. The deformation of an idempotent into a mixed state, and the channel that performs it, are the subject of Decoherence as Idempotent Projection; the selection itself is not explained there either.

7. Empirical content. The reformulation reproduces standard quantum physics for a single qubit. Does it predict anything new, or does it only reinterpret? If it predicts new physics, at what scale does the deviation from standard QM appear?

8. The path integral. The standard path integral formulation of quantum physics is not obviously expressible in the biquaternion framework. Is there a natural path integral in the biquaternion algebra, with a natural measure inherited from the algebra? A path integral is constructed in The Path Integral in Biquaternionic Form, with the Schrödinger, spin-half and Coulomb instances in its companion articles.

Summary

Quantum physics in the biquaternion framework is built on the Hermitian subspace $\mathbb{B}_{\Psi}$ of the biquaternion algebra. The states are the positive elements of $\mathbb{B}_{\Psi}$ normalised by $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12$ (unit weight), i.e., the Bloch ball, the pure states are the idempotents (the Bloch sphere), and the observables are the general Hermitian elements. The Born rule is the normalised Hermitian form $\langle\tilde{Ϛ}_{\Psi},\tilde H_\Psi\rangle_{*}/\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}$, the dynamics is rotor conjugation by $*$-unitary biquaternions, and measurement is the two-sided operation $\tilde{Ϛ}_{\Psi}\mapsto\tilde\Pi\tilde{Ϛ}_{\Psi}\tilde\Pi$.

The framework reproduces the standard qubit formalism for the states, observables, Born rule, dynamics, and measurement, with the structural advantages that states and observables live in the same space, the Born rule is a consequence of the Hermitian form, measurement is an algebraic operation, and the positivity condition on states is a biquaternion-norm condition.

The biquaternion space carries both structures at once — the Lorentz group in the invertible elements through the double cover $SL(2,\mathbb{C})$, and the material sector $\mathbb{B}_{\Phi}$ of four-vectors — and the quantum formalism presented here and the relativistic structure coexist within it. The full relativistic quantum theory of spinor fields is the natural continuation of the framework: the classical precursor is the biquaternion Dirac equation treated in The Dirac Equation in Biquaternionic Form, and its quantization is carried out in Canonical Quantization of the Biquaternion Dirac Field.

The extension to many qubits, the second-quantized version, the connection to quantum field theory, and the question of empirical content are open.

Summary of Notation

Symbol Meaning
$\mathbb{B}$ Biquaternion algebra
$\mathbb{B}_{\Psi}$ Hermitian subspace (states, observables)
$\mathbb{B}_{\Phi}$ Anti-Hermitian subspace (generators)
$\mathbb{B}_{H}$ Real quaternion subspace, $\operatorname{span}_{\mathbb{R}}\{e_t, \vec e_x, \vec e_y, \vec e_z\}$ (unit elements: $SU(2)$)
$\tilde H_\Psi = h_t e_t + i\vec h_\Psi$ Hermitian element (observable)
$\tilde{Ϛ}_{\Psi} = \tfrac{1}{2}(e_t + i\vec s_{\Psi})$ State
$\tilde\Pi_{1,2}(\vec u) = \tfrac{1}{2}(e_t \pm i\vec u)$ Idempotent (pure state)
$\langle\tilde{Ϛ}_{\Psi},\tilde H_\Psi\rangle_{*} = \tfrac12(h_t + \vec s_{\Psi}\cdot\vec h_\Psi)$, $\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*}=\tfrac12$ Hermitian form on a state
$\langle\tilde{Ϛ}_{\Psi},\tilde H_\Psi\rangle_{*}/\langle\tilde{Ϛ}_{\Psi},e_t\rangle_{*} = h_t + \vec s_{\Psi}\cdot\vec h_\Psi$ Born rule, the normalised Hermitian form
$\mathrm{Sc}(\tilde H_\Psi) = h_t$, $2\,\mathrm{Sc}(\tilde{Ϛ}_{\Psi}\tilde H_\Psi^{*}) = h_t + \vec s_{\Psi}\cdot\vec h_\Psi$ Scalar part, the value of the Hermitian form
$\langle\tilde H_\Psi\rangle_{\natural} = \langle\tilde H_\Psi,\tilde H_\Psi\rangle_{\natural} = \tilde H_\Psi\tilde H_\Psi^{\natural} = h_t^2 - |\vec h_\Psi|^2$ Biquaternion norm, inertia $(1,3)$
$\langle\tilde{H}_{\Psi,1},\tilde{H}_{\Psi,2}\rangle_{\natural} = h_{t,1} h_{t,2} - \vec h_{\Psi,1}\cdot\vec h_{\Psi,2}$ Polarization of the biquaternion norm
$\tilde{Ϛ}_{\Psi}\mapsto\tilde\Pi\tilde{Ϛ}_{\Psi}\tilde\Pi$ Projective measurement
$\tilde Q(r'_t) = \exp(-i\theta_{\mathrm{evo}})(\cos(|\vec h_\Psi|\frac{r'_{t}}{\hbar\,c})e_t + \sin(|\vec h_\Psi|\frac{r'_{t}}{\hbar\,c})\vec u_\Psi)$ $*$-Unitary evolution
$\tilde{Ϛ}_{\Psi}(r'_t) = \tilde Q(r'_t)\tilde{Ϛ}_{\Psi}\tilde Q(r'_t)^{*}$ Von Neumann evolution
$\tilde Q_1\mapsto\tilde Q\tilde Q_1\tilde Q^{*}$, $\tilde Q$ invertible Conjugation; quantum if $\tilde Q\tilde Q^{*} = e_t$ (group $\mathrm{U}(2)$), Lorentz if $\tilde U_{\mathrm{lor}}\tilde U_{\mathrm{lor}}{}^{\natural} = e_t$ (group $SL(2,\mathbb{C})$)

Further Reading

  • P. A. M. Dirac, The Principles of Quantum Mechanics (Oxford, 1930), for the foundational treatment.
  • John von Neumann, Mathematical Foundations of Quantum Mechanics (Princeton, 1932), for the density matrix and Hilbert space formalism.
  • Michael A. Nielsen and Isaac L. Chuang, Quantum Computation and Quantum Information (Cambridge, 2000), for qubits, the Bloch sphere, and quantum channels.
  • J. J. Sakurai and Jim Napolitano, Modern Quantum Mechanics (Pearson, 2017), for the standard textbook treatment of spin-1/2.
  • Asher Peres, Quantum Theory: Concepts and Methods (Kluwer, 1995), for the conceptual foundations.
  • Pertti Lounesto, Clifford Algebras and Spinors (Cambridge, 2001), for the Clifford algebra formulation of spin.
  • Chris Doran and Anthony Lasenby, Geometric Algebra for Physicists (Cambridge, 2003), for the geometric algebra approach to quantum physics.
  • S. L. Adler, Quaternionic Quantum Mechanics and Quantum Fields (Oxford, 1995), for the quaternionic formulation, its scalar-product obstruction and the comparison with the complex theory.
  • D. Finkelstein, J. M. Jauch, S. Schiminovich and D. Speiser, "Foundations of quaternion quantum mechanics", Journal of Mathematical Physics 3 (1962) 207, for the earliest systematic quaternionic quantum theory.