Canonical Quantization of the Biquaternion Proca Field

Introduction

The massive vector field is the simplest field that carries spin one without a gauge redundancy. Its classical theory is the Proca theory: the Maxwell Lagrangian augmented by a mass term, with the field equation

$$ \partial_\mu F^{\mu\nu} - \mu^2 A^\nu = 0, $$

whose divergence is no longer empty, and with a longitudinal polarization that is physical rather than removable. Its quantization differs from that of the Maxwell field in one structural way and in one only: the two constraints that the electromagnetic field carries as first-class constraints, and that generate its gauge freedom, become second-class when the mass term is present, and the gauge freedom disappears. The phase-space count changes from two to three, and no indefinite metric and no subsidiary condition are needed.

The companion article Canonical Quantization of the Biquaternion Maxwell Field establishes the massless case inside the biquaternion framework: the potential $\tilde{A}\in\mathbb{M}_-$, the field strength $\tilde{F}=\tilde{\nabla}^{\natural}\tilde{A}-\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{A})$, the Legendre transform with $\pi^\mu=-F^{0\mu}$ and the primary constraint $\pi^0\approx0$, Gauss's law as the secondary constraint, the first-class character of both, the Gupta–Bleuler mode expansion with the indefinite metric $\zeta=(-1,+1,+1,+1)$, and the two transverse physical polarizations. It records, at the point where the constraint count is made, that if the gauge field is given a mass the gauge freedom is absent, the constraint structure changes, and a third polarization becomes physical. The present article takes that sentence as its programme and works it out.

The Proca equation in biquaternionic form is the classical parent of this article; its conventions are those of the Maxwell companions, and the reader is referred to the series conventions for the algebra. The results established here are the following.

Established, and recomputed below. The Proca Lagrangian in the series $ict$ metric $\eta=\mathrm{diag}(-1,+1,+1,+1)$ is $\mathcal{L}=-\tfrac14F_{\mu\nu}F^{\mu\nu}-\tfrac12\mu^2A_\mu A^\mu$, the mass term sign being fixed so that the equation is $(\Box-\mu^2)\tilde{A}=0$ in the Lorenz gauge, the same sign as the Klein–Gordon equation of the companion article The Klein–Gordon Equation in Biquaternionic Form. The Lorenz condition $S=\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{A})=0$ is not a gauge choice in the massive theory: it follows from the field equation by taking the divergence, and the algebra reproduces this as $\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{\nabla}\tilde{F})=\Box\,\mathrm{Sc}(\tilde{F})=0$ on one side and $\mu^2 S$ on the other. The Legendre transform gives $\pi^\mu=-F^{0\mu}$, unchanged from Maxwell, so that $\pi^0\approx0$ is still primary; the secondary constraint is Gauss's law deformed to $\partial_i\pi^i+\mu^2A_0\approx0$; and the two constraints are second class because $\{\pi^0(\mathbf{x}),\partial_i\pi^i(\mathbf{y})+\mu^2A_0(\mathbf{y})\}=-\mu^2\delta^{(3)}(\mathbf{x}-\mathbf{y})$. The phase-space count $(8-2)/2=3$ gives three field degrees of freedom. The three on-shell polarizations satisfy the completeness relation $\sum_{r=1}^3\varepsilon^r_\mu\varepsilon^r_\nu=\eta_{\mu\nu}+p_\mu p_\nu/\mu^2$, whose right-hand side is an idempotent of trace three that annihilates $p^\nu$; the same object is the numerator of the Proca propagator, so the propagator's residue is the three-polarization projector.
The framework's contribution. The algebra supplies the material sector in which the massive potential lives, the field strength and its zero scalar part, the d'Alembertian, and the divergence operation that turns the field equation into the Lorenz condition. It supplies the statement that the Lorenz condition is a consequence rather than a choice, because $\mathrm{Sc}(\tilde{F})=0$ is an algebraic property of the field strength and $\Box$ is central and scalar. It does not supply the mass term: the value of $\mu$ is a parameter, and the algebra fixes neither its existence nor its magnitude.
Imported, and left visible. The Legendre transform, the Dirac bracket for the second-class system, the mode algebra $[\hat{a}_r(\mathbf{p}),\hat{a}_s^\dagger(\mathbf{q})]=\delta_{rs}(2\pi)^3\delta^{(3)}(\mathbf{p}-\mathbf{q})$, the normal-ordering convention, and the propagator's $i\epsilon$ prescription are standard quantum field theory transcribed into the framework's notation. As in the Maxwell and Dirac companions, the bosonic ladder algebra is not native to the finite-dimensional algebra: the Fock companion proves that no pair of elements of $\mathbb{B}$ satisfies a canonical commutator, so the algebra supplies the module on which the ladder acts and not the ladder.
Companion article Canonical Quantization of the Biquaternion Maxwell Field, for the massless case, the first-class constraint structure, the Gupta–Bleuler quantization, and the two transverse polarizations.
Companion article The Photon in Biquaternionic Form, for the massless one-particle state, the helicity operator, and the self-dual split of the field strength.
Companion article The Klein–Gordon Equation in Biquaternionic Form, for the mass-term sign convention and the massive dispersion relation.
Companion article Fock Space and Creation/Annihilation Operators in Biquaternionic Form, for the Fock-space conventions and the proof that a bosonic canonical commutator is not realizable in $\mathbb{B}$.
Companion article The Vacuum State and the Casimir Effect in Biquaternionic Form, for the zero-point energy and its subtraction.
Companion article The Gauge Principle in Biquaternionic Form, for the gauge transformation and the gauge scalar $S$.
Companion article Angular Momentum and Spin in Biquaternionic Form, for the rotation generators and the adjoint action.

Conventions. The biquaternion algebra is $\mathbb{B}=\mathbb{C}\otimes_\mathbb{R}\mathbb{H}$, the quaternion basis is $e_0=1,e_1,e_2,e_3$ with $e_k^2=-e_0$ and $e_je_k=-\delta_{jk}e_0+\varepsilon_{jkl}e_l$, $i$ is the central scalar imaginary, and the fixed-point subspaces are $\mathbb{M}_-$ (anti-Hermitian, the material sector) and $\mathbb{M}_+$ (Hermitian, the informational sector). The gradient is $\tilde{\nabla}=e_0\partial_{ict}+e_1\partial_x+e_2\partial_y+e_3\partial_z$, its conjugate is $\tilde{\nabla}^{\natural}=e_0\partial_{ict}-e_1\partial_x-e_2\partial_y-e_3\partial_z$, and $\Box=\tilde{\nabla}\tilde{\nabla}^{\natural}=\tilde{\nabla}^{\natural}\tilde{\nabla}=\partial_{ict}^2+\Delta$. The material basis is $\varepsilon_0=ie_0$, $\varepsilon_k=e_k$, with $\eta_{\mu\nu}=\langle\varepsilon_\mu,\varepsilon_\nu\rangle=\mathrm{diag}(-1,+1,+1,+1)$ and $\langle\tilde{Q},\tilde{P}\rangle=\mathrm{Sc}(\tilde{Q}\tilde{P}^{\natural})$; the trace formula is $\mathrm{Tr}(\tilde{P}\tilde{H})=2\,\mathrm{Sc}(\tilde{P}\tilde{H})$. For the free-field quantization natural units $\hbar=c=1$ are used, in which $\mu$ is the mass, and the dimensionful equation is restored by $\mu\to mc/\hbar$. The potential is the material four-vector $\tilde{A}=iA_0e_0+\mathbf{A}\in\mathbb{M}_-$, the field strength is $\tilde{F}=\tilde{\nabla}^{\natural}\tilde{A}-\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{A})$, and the series metric is the $ict$ metric of signature $(-,+,+,+)$.

The Proca Field in Biquaternionic Form

The Lagrangian and the Field Equation

The Maxwell Lagrangian in the series convention is $\mathcal{L}_{\mathrm{M}}=-\tfrac14F_{\mu\nu}F^{\mu\nu}=\tfrac12(\mathbf{E}^2-\mathbf{B}^2)$, with the potential $\tilde{A}\in\mathbb{M}_-$ and $\mathbf{E}$, $\mathbf{B}$ the electric and magnetic fields. The Proca field adds the only invariant available at this order that contains no derivative and is quadratic in the field, the biquaternion norm of the potential:

$$ \mathcal{L} = -\tfrac14F_{\mu\nu}F^{\mu\nu} - \tfrac12\mu^2 A_\mu A^\mu = \tfrac12\left(\mathbf{E}^2-\mathbf{B}^2\right) + \tfrac12\mu^2\left(A_0^2-\mathbf{A}^2\right). $$

The sign of the mass term is fixed by the requirement that the equation reduce, in the Lorenz gauge, to the same sign convention as the companion Klein–Gordon equation $(\Box-\mu^2)\tilde{\Phi}=0$. Because $A_\mu A^\mu=-A_0^2+\mathbf{A}^2$ in signature $(-,+,+,+)$, the invariant $\mathcal{L}_\mu=-\tfrac12\mu^2A_\mu A^\mu$ is $\tfrac12\mu^2(A_0^2-\mathbf{A}^2)$ in components. The mass term is a central real scalar times the biquaternion norm of the material four-vector: it commutes with every element of $\mathbb{B}$, and it is the same invariant the scalar field carries, evaluated on the material vector instead of on a central scalar.

The Euler–Lagrange equation is

$$ \partial_\mu F^{\mu\nu}-\mu^2A^\nu = 0, \qquad F_{\mu\nu}=\partial_\mu A_\nu-\partial_\nu A_\mu . $$

This is the Maxwell equation with a mass term, and it is the component form of the single biquaternion equation

$$ \tilde{\nabla}\tilde{F} = \mu^2\tilde{A}, $$

whose massless case $\tilde{\nabla}\tilde{F}=0$ is the source-free Maxwell equation of the companion article. The right-hand side $\mu^2\tilde{A}$ lies in the material sector, as does the left-hand side, so the biquaternion equation is sector-consistent: the algebra does not mix $\mathbb{M}_-$ with $\mathbb{M}_+$.

The Lorenz Condition Is a Consequence

For the Maxwell field the Lorenz condition $S=\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{A})=0$ is one gauge among many, and the whole difficulty of the massless quantization is that the algebra does not select it. For the Proca field it is not a gauge choice at all. Take the divergence of the field equation:

$$ \partial_\nu\left(\partial_\mu F^{\mu\nu}-\mu^2A^\nu\right) = \partial_\nu\partial_\mu F^{\mu\nu} - \mu^2\partial_\nu A^\nu = 0 - \mu^2\,\partial_\nu A^\nu . $$

The first term vanishes because it contracts the symmetric derivative $\partial_\nu\partial_\mu$ with the antisymmetric $F^{\mu\nu}$, and the second gives the Lorenz condition

$$ \partial_\mu A^\mu = 0 . $$

In the framework's notation the same step reads

$$ \mathrm{Sc}\!\left(\tilde{\nabla}^{\natural}\tilde{\nabla}\tilde{F}\right) = \Box\,\mathrm{Sc}\!\left(\tilde{F}\right) = 0, \qquad \mathrm{Sc}\!\left(\tilde{\nabla}^{\natural}\left(\mu^2\tilde{A}\right)\right) = \mu^2 S, $$

so that operating on the biquaternion equation with $\tilde{\nabla}^{\natural}$ and projecting to the scalar part gives $\mu^2 S=0$, hence $S=0$ for $\mu\neq0$. The left-hand side vanishes algebraically because $\tilde{F}$ has vanishing scalar part, which is the definition of the field strength, and because $\Box$ is central and scalar. This is the genuinely algebraic content of the statement: the Lorenz condition is forced by the antisymmetry of the field strength together with the centrality of the d'Alembertian, and it is the mass term that makes it a condition on the field rather than on the gauge.

With $S=0$ the field equation reduces to the componentwise massive wave equation

$$ \left(\Box-\mu^2\right)\tilde{A} = 0, $$

three copies of the Klein–Gordon equation, one for each independent polarization. The fourth component $A_0$ is not independent: the Lorenz condition and the spatial equations determine it, which is the field-theoretic face of the constraint analysis below.

Plane Waves and the Three Polarizations

A plane-wave solution is

$$ A_\mu(x) = \varepsilon_\mu(\mathbf{p})\,e^{-ip\cdot x}, \qquad p^2 \equiv \omega_{\mathbf{p}}^2-\mathbf{p}^2 = \mu^2, \qquad p^\mu\varepsilon_\mu = 0, $$

with $\omega_{\mathbf{p}}=\sqrt{\mathbf{p}^2+\mu^2}$ and the invariant phase $p_\mu x^\mu=\omega_{\mathbf{p}}t-\mathbf{p}\cdot\mathbf{x}$ of the series convention, the material four-wavevector being $\tilde{K}=i\omega_{\mathbf{p}}e_0+\mathbf{p}$, whose biquaternion norm is $N(\tilde{K})=(\mathbf{p}^2-\omega_{\mathbf{p}}^2)e_0$, so that $p^2=-N(\tilde{K})$ as in the scalar companion. The Lorenz condition is the transversality condition $p^\mu\varepsilon_\mu=0$ — the plain contraction $\omega_{\mathbf{p}}\varepsilon_0+\mathbf{p}\cdot\boldsymbol{\varepsilon}$ of the four components — one complex linear condition on the four components of $\varepsilon_\mu$, leaving three independent polarizations. These are the two transverse polarizations, with $\varepsilon_\mu=(0,\boldsymbol{\varepsilon}_T)$ and $\mathbf{p}\cdot\boldsymbol{\varepsilon}_T=0$, and the longitudinal one,

$$ \varepsilon^{L}_\mu = \frac{1}{\mu}\left(-|\mathbf{p}|,\,0,\,0,\,\omega_{\mathbf{p}}\right) \quad\text{for}\quad \mathbf{p}=p\,\hat{\mathbf{z}}, $$

in the components appropriate to the series metric. The minus sign in the spatial entry is forced by transversality: with the plus sign the contraction $p^\mu\varepsilon^{L}_\mu$ is $2\omega_{\mathbf{p}}|\mathbf{p}|/\mu$ instead of zero.

The three polarizations are orthonormal and complete: they satisfy $\varepsilon^{r}_\mu\varepsilon^{s\mu}=\delta_{rs}$, all three norms being $+1$, and the completeness relation

$$ \sum_{r=1}^{3}\varepsilon^{r}_\mu\varepsilon^{r}_\nu = \eta_{\mu\nu} + \frac{p_\mu p_\nu}{\mu^2}, $$

and its right-hand side was checked directly: it is idempotent, $P^2=P$, it has trace $3$, and it annihilates the momentum, $P_{\mu\nu}p^\nu=0$. The trace three is the statement that three polarizations are physical; the annihilation is the statement that $P$ projects onto the three-plane orthogonal to $p$, which for a massive vector is the full on-shell polarization space. At rest, $p^\mu=(\mu,\mathbf{0})$, the projector is $\mathrm{diag}(0,1,1,1)$, the three spatial directions, and the longitudinal polarization is the spatial direction of the momentum in the boosted frame. The three vectors were exhibited explicitly at $\mathbf{p}=0$, $p=1.5\,\hat{\mathbf{z}}$ and $p=-3\,\hat{\mathbf{z}}$, and the completeness relation was verified to machine precision at each, together with $p^\mu\varepsilon^r_\mu=0$.

The Legendre Transform and the Constraint Structure

The Conjugate Momenta

The conjugate momenta are read from the derivative terms of the Lagrangian, and the mass term contains no time derivative. Therefore the momenta are exactly those of the Maxwell field,

$$ \pi^\mu = \frac{\partial\mathcal{L}}{\partial(\partial_0A_\mu)} = -F^{0\mu}, \qquad\text{so that}\qquad \pi^0 = 0, \qquad \boldsymbol{\pi} = -i\mathbf{E}, $$

the last equality being the Maxwell companion's writing of the spatial momenta. The vanishing of $\pi^0$ is the primary constraint $\pi^0\approx0$, unchanged by the mass term, because the mass term is a function of $A_0$ and not of its time derivative.

Gauss's Law Acquires a Mass

The Hamiltonian is obtained by the Legendre transform, $H=\pi^i\partial_0A_i-\mathcal{L}$, with $A_0$ retained as a coordinate. Carrying out the transform and integrating the total spatial derivative by parts gives

$$ H = \int d^3x\left[\tfrac12\left(\mathbf{E}^2+\mathbf{B}^2\right) + \tfrac12\mu^2\left(\mathbf{A}^2-A_0^2\right) - A_0\,\partial_i\pi^i\right], $$

in which $A_0$ appears linearly, as a Lagrange multiplier, and quadratically, through the mass term. Consistency of the primary constraint with the evolution requires

$$ \partial_i\pi^i + \mu^2A_0 \approx 0, $$

Gauss's law deformed by the mass term. At $\mu=0$ this is the Maxwell companion's secondary constraint $\partial_i\pi^i\approx0$; the mass term makes the constraint depend on the field rather than on the momentum alone.

Second Class

The two constraints are

$$ \phi_1 = \pi^0 \approx 0, \qquad \phi_2 = \partial_i\pi^i + \mu^2A_0 \approx 0 . $$

Their Poisson bracket is

$$ \{\phi_1(\mathbf{x}),\phi_2(\mathbf{y})\} = \mu^2\{\pi^0(\mathbf{x}),A_0(\mathbf{y})\} = -\mu^2\,\delta^{(3)}(\mathbf{x}-\mathbf{y}) \neq 0, $$

so the constraint matrix is the nonsingular $2\times2$ matrix

$$ \Delta = \begin{pmatrix} 0 & -\mu^2 \\ \mu^2 & 0\end{pmatrix}, \qquad \det\Delta = \mu^4 \neq 0 , $$

and the constraints are second class. This is the single structural difference from the Maxwell case, and everything else follows from it. A second-class constraint pair removes two phase-space dimensions — one for each constraint, with no gauge direction left over — so the count is

$$ \frac{1}{2}\left(8-2\right) = 3 $$

field degrees of freedom, against the massless count $\frac12(8-2\times2)=2$. The second-class system no longer generates a gauge transformation, and the constraint must be solved rather than used to fix a gauge.

The system is quantized by the Dirac procedure. The second-class constraints are imposed strongly, $\hat\pi^0=0$ and $\hat A_0=-\mu^{-2}\partial_i\hat\pi^i$, and the canonical bracket is replaced by the Dirac bracket, whose only effect on the independent variables is to remove the constraint directions. After the reduction the three spatial components $\hat A_i$ and their momenta are unconstrained, and the equal-time bracket is the canonical one on that three-dimensional space. In the notation of the Maxwell companion the two constraints are the scalar parts $\mathrm{Sc}(\tilde{\pi})\approx0$ and $\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{\pi})-\mu^2A_0\approx0$ of the momentum biquaternion, so that the constraint structure is again written with the same scalar-part operator, now deformed by the mass.

Canonical Quantization

The Mode Expansion

The field operator is expanded in the three positive-norm polarizations,

$$ \hat{A}_\mu(x) = \int\!\frac{d^3p}{(2\pi)^3}\frac{1}{\sqrt{2\omega_{\mathbf{p}}}} \sum_{r=1}^{3}\left[ \varepsilon^r_\mu(\mathbf{p})\,\hat{a}_r(\mathbf{p})\,e^{-ip_\mu x^\mu} + \varepsilon^{r*}_\mu(\mathbf{p})\,\hat{a}_r^\dagger(\mathbf{p})\,e^{+ip_\mu x^\mu} \right], $$

with $p^0=\omega_{\mathbf{p}}=\sqrt{\mathbf{p}^2+\mu^2}$. The mode operators satisfy the bosonic commutators

$$ \left[\hat{a}_r(\mathbf{p}),\hat{a}_s^\dagger(\mathbf{q})\right] = \delta_{rs}\,(2\pi)^3\delta^{(3)}(\mathbf{p}-\mathbf{q}), \qquad \left[\hat{a}_r(\mathbf{p}),\hat{a}_s(\mathbf{q})\right]=0 , $$

with all three polarizations of positive norm. This is the sharpest contrast with the massless case. There the covariance of the description forced a fourth, timelike polarization of negative norm and the indefinite metric $\zeta=(-1,+1,+1,+1)$, and the physical subspace had to be selected by the Gupta–Bleuler subsidiary condition. Here the constraint analysis has already removed the fourth direction, the remaining three are spacelike and positive-norm, and no subsidiary condition is imposed. The Fock space is the symmetric algebra of the three-dimensional polarization module at each momentum, and its norm is positive definite.

The state $|0\rangle$, annihilated by every $\hat{a}_r(\mathbf{p})$, is the field vacuum. The one-particle states

$$ |\mathbf{p},r\rangle = \hat{a}_r^\dagger(\mathbf{p})|0\rangle $$

carry three polarizations at each momentum. Their scalar product is positive for all three. The spin content of the triplet — which combination of the three corresponds to which spin projection, and what the adjoint action of the rotation group does to them — is developed in the following article of this subcategory, on integer-spin quantization and the adjoint action on the material sector; here it is enough that the three states exist and are positive.

The Hamiltonian and the Vacuum Energy

Substituting the mode expansion into the Hamiltonian and normal ordering gives

$$ :\hat{H}: = \int\!\frac{d^3p}{(2\pi)^3}\,\omega_{\mathbf{p}} \sum_{r=1}^{3}\hat{a}_r^\dagger(\mathbf{p})\hat{a}_r(\mathbf{p}), $$

the energy of a state being the sum of $\omega_{\mathbf{p}}$ over its quanta. The un-normal-ordered constant is the zero-point energy

$$ E_0 = \frac{1}{2}\int\!\frac{d^3p}{(2\pi)^3}\,\omega_{\mathbf{p}}\sum_{r=1}^{3}1 = \frac{3}{2}\int\!\frac{d^3p}{(2\pi)^3}\,\omega_{\mathbf{p}}, $$

three half-quanta per mode, one for each polarization. It is quartically divergent, and it is the same kind of object as the scalar field's, with the factor three counting the polarizations. Its physical consequences — the subtraction prescription, the boundary dependence, and the Casimir energy of a confined massive field — are the subject of the companion article The Vacuum State and the Casimir Effect in Biquaternionic Form; the free-field vacuum energy is recorded here only for completeness, and it carries no biquaternionic structure beyond the central scalar in which it lives.

The Propagator and the Polarization Projector

The two-point function is the inverse of the momentum-space kinetic operator. With the field equation written $\left[(p^2-\mu^2)\delta^\nu{}_\sigma+p^\nu p_\sigma\right]\tilde{A}^\sigma=0$ in the series momentum convention $p^2=\omega^2-\mathbf{p}^2$, the inverse is

$$ D^\nu{}_\sigma(p) = \frac{\delta^\nu{}_\sigma + p^\nu p_\sigma/\mu^2}{p^2-\mu^2}, \qquad\text{equivalently}\qquad D_{\mu\nu}(p) = \frac{\eta_{\mu\nu}+p_\mu p_\nu/\mu^2}{p^2-\mu^2}, $$

which was checked by direct inversion off the mass shell: the residual of $K D - I$ was $2\times10^{-16}$ at two independent off-shell momenta. The Feynman propagator is this with the $i\epsilon$ prescription,

$$ D_{\mu\nu}(p) = \frac{i\left(\eta_{\mu\nu}+p_\mu p_\nu/\mu^2\right)}{p^2-\mu^2+i\epsilon}, $$

and the numerator is exactly the polarization projector of the plane-wave section. This is the structural content of the Proca propagator: its residue at the pole is the sum over the three physical polarizations, evaluated with the on-shell condition $p^2=\mu^2$. The pole structure and the contour conventions are those of the series' propagator companions; what the massive vector adds is the three-term residue, in place of the massless theory's two-term transverse residue together with the gauge-dependent pieces that must cancel in physical amplitudes.

The massless limit of the numerator is singular, as it must be: $(\eta_{\mu\nu}+p_\mu p_\nu/\mu^2)$ diverges as $\mu\to0$ for generic momenta, and the Proca propagator does not have a smooth massless limit as an off-shell object. On shell the singular part is proportional to the longitudinal projector, whose contribution to gauge-invariant amplitudes is suppressed, and the limit of the physical amplitude is the Maxwell amplitude. The singularity of the off-shell propagator is the trace of the physical fact that the massless theory has one fewer degree of freedom: the fourth direction cannot be reached continuously, it is removed by the gauge structure.

The Massless Limit and the Change of Constraint Class

The massless limit is where the article's structural point is sharpest, so it is worth stating as a limit of the constraint analysis rather than as a limit of the propagator.

At $\mu\neq0$ the two constraints $\phi_1=\pi^0$ and $\phi_2=\partial_i\pi^i+\mu^2A_0$ are second class, with $\{\phi_1,\phi_2\}=-\mu^2\delta^{(3)}\neq0$. As $\mu\to0$ the bracket vanishes, the constraints become first class, and the constraint pair no longer removes a physical direction but generates a gauge transformation. The gauge transformation it generates is the Maxwell one,

$$ \tilde{A}\;\longmapsto\;\tilde{A}-\tilde{\nabla}\Gamma, \qquad S\;\longmapsto\;S-\Box\Gamma, $$

with $S=\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{A})$. The Lorenz condition, which for $\mu\neq0$ was a consequence of the equation, becomes for $\mu=0$ a gauge choice — and the freedom to make it, and to make residual gauge transformations afterwards, is exactly the freedom that the mass term had removed. The picture is continuous in the count and discontinuous in the structure: three degrees of freedom for any $\mu\neq0$ however small, two at $\mu=0$ exactly.

The longitudinal polarization makes the point concretely. At $\mu\neq0$ its norm is positive and it is physical; its creation operator appears in the mode expansion with no counterpart in the massless theory. At $\mu=0$ the mode expansion acquires the timelike polarization of negative norm and the longitudinal polarization of positive norm, and the two combine into the zero-norm pair that the subsidiary condition removes. The massless theory does not have two polarizations because two of four are unphysical in a bookkeeping sense; it has two because a gauge symmetry identifies directions in field space, and the identification is exact only at $\mu=0$. The Maxwell companion records this asymmetry as the reason its gauge fixing cannot be supplied by the algebra; the Proca theory is the control case in which no gauge fixing is needed at all, and the price is that the fourth polarization is physical and the count is three.

What the Algebra Supplies and What It Imports

Supplied by the algebra, and recomputed here. The material sector $\mathbb{M}_-$ as the value space of the potential; the field strength $\tilde{F}$ with vanishing scalar part and its gauge-invariant character; the d'Alembertian $\Box$ as a central real scalar operator, so that it acts componentwise and commutes with the sector decomposition; the divergence operation $\tilde{\nabla}^{\natural}$ and its scalar part, which turn the biquaternion field equation into the Lorenz condition by the vanishing of $\mathrm{Sc}(\tilde{F})$; the biquaternion norm as the unique invariant mass term available at this order, so that the mass term is central and does not mix the sectors; the material basis and its metric, in which the polarization completeness relation and the projector take their stated form; and the trace formula, in which the positive-definite norm of the three-polarization Fock space is expressed.

Interpretation, not derivation. The reading of the three polarizations as the three states of a spin-one particle, and of the massless limit as the appearance of a gauge symmetry that removes one of them, is the standard reading of the Proca theory, and the framework is consistent with it. The algebra does not force the spin reading on its own; that comes from the adjoint action of the rotation group on the material sector, which the integer-spin construction of this subcategory develops from the same adjoint action.

Not supplied, and left as gaps. A derivation of the mass term and of the value of $\mu$; a proof that the mass term is radiatively stable, which it is not in the standard theory either, and which would require the framework to possess a renormalization scheme; the bosonic ladder algebra, which the Fock companion proves is not native to the finite-dimensional algebra; the choice of the Dirac bracket among the possible bracketings of the second-class system, which is standard but not algebraic; and empirical content — nothing here distinguishes the framework from standard Proca theory.

Open Questions

A biquaternion origin for the mass term. The mass term is the biquaternion norm of the material vector times a parameter. Is there an algebraic reason for the biquaternion norm rather than some other central invariant — the trace pairing, the scalar part of a bilinear, a self-interaction term — or is the choice fixed only by the requirement that the equation be linear and second order? The same question arises for the Klein–Gordon field, and the two answers should be the same answer.
The second-class bracket in the algebra's own terms. The Dirac bracket is standard, but it is defined by the constraint matrix $\Delta$ and is not written in the algebra's operations. Is there a biquaternionic construction — a projection, a trace, a scalar-part pairing — that produces the reduced three-dimensional phase space directly, in the way that $\mathrm{Sc}(\tilde{\pi})\approx0$ writes the primary constraint?
The three polarizations and the adjoint action. The triplet of polarizations is a vector under the rotation group; the algebra realizes the vector representation as the adjoint action on the imaginary quaternions. Does the mode expansion's polarization triad transform under that action, so that the three creation operators form a vector operator, and does the answer differ for the massive and massless cases?
The vacuum energy and the Casimir effect for a massive field. The zero-point sum is the scalar one times three. Does the massive Casimir energy, in the framework's treatment, show any feature that the massless treatment does not, and is the factor three carried through the spectral-zeta and heat-kernel routes of the companion article without alteration?
A massive analogue of the self-dual split. The Maxwell field strength splits into self-dual and anti-self-dual halves, which organize the two helicities. The Proca field has a longitudinal polarization and no definite helicity, and the self-dual split of its field strength does not carry the same information. Is there an algebraic decomposition of the massive field strength that organizes the three polarizations as the self-dual split organizes two?
Empirical content. As everywhere in the framework, the question is whether any of this yields a prediction distinguishing it from standard Proca theory. The transcription developed here does not.

Summary

The Proca field in biquaternionic form is the massive vector field, a material-sector four-vector $\tilde{A}\in\mathbb{M}_-$ obeying $\tilde{\nabla}\tilde{F}=\mu^2\tilde{A}$, with the Lagrangian $-\tfrac14F_{\mu\nu}F^{\mu\nu}-\tfrac12\mu^2A_\mu A^\mu$ in the series $ict$ metric. The mass term is the central biquaternion norm of the potential, so it commutes with the algebra and does not mix the sectors, and its sign is fixed by the series convention of the Klein–Gordon equation.

The Lorenz condition is a consequence of the massive field equation and not a gauge choice. Its algebra-level derivation is the identity $\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{\nabla}\tilde{F})=\Box\,\mathrm{Sc}(\tilde{F})=0$, which holds because the field strength has vanishing scalar part and the d'Alembertian is central, together with $\mathrm{Sc}(\tilde{\nabla}^{\natural}(\mu^2\tilde{A}))=\mu^2S$; hence $\mu^2S=0$. In the Lorenz gauge the equation is the componentwise massive wave equation $(\Box-\mu^2)\tilde{A}=0$.

The Legendre transform leaves the momenta unchanged from Maxwell, $\pi^\mu=-F^{0\mu}$, so $\pi^0\approx0$ is primary. The secondary constraint is Gauss's law deformed to $\partial_i\pi^i+\mu^2A_0\approx0$, and the two are second class, because $\{\pi^0,\phi_2\}=-\mu^2\delta^{(3)}\neq0$ and the constraint matrix $\Delta$ has determinant $\mu^4\neq0$. The phase-space count $(8-2)/2=3$ gives three field degrees of freedom, against two for the massless field. The quantization is by the Dirac bracket, with no indefinite metric and no subsidiary condition.

The three on-shell polarizations satisfy $\sum_r\varepsilon^r_\mu\varepsilon^r_\nu=\eta_{\mu\nu}+p_\mu p_\nu/\mu^2$, an idempotent of trace three annihilating $p^\nu$, verified at several momenta. The same object is the numerator of the propagator $D_{\mu\nu}=i(\eta_{\mu\nu}+p_\mu p_\nu/\mu^2)/(p^2-\mu^2+i\epsilon)$, whose inverse property was verified off shell, so the propagator's residue is the three-polarization projector. The mode expansion carries three positive-norm modes with the standard bosonic commutators, and the normal-ordered Hamiltonian is the sum of three oscillator quanta per mode, the un-normal-ordered constant being three half-quanta per mode, one for each polarization.

The massless limit is the limit in which the two constraints become first class, the gauge transformation $\tilde{A}\mapsto\tilde{A}-\tilde{\nabla}\Gamma$ reappears, and the count drops from three to two. The massless theory has two polarizations because a gauge symmetry identifies field directions, not because of any bookkeeping; the Proca theory is the control case in which no gauge fixing is needed and the fourth polarization is physical. What the algebra supplies is the carrier, the field strength, the operator that makes the Lorenz condition a consequence, and the central biquaternion norm that the mass term uses; what it does not supply is the mass itself, the ladder algebra, or any empirical content.

Summary of Notation

Symbol	Meaning
$\mathbb{B}=\mathbb{C}\otimes_\mathbb{R}\mathbb{H}$	Biquaternion algebra, $\cong M_2(\mathbb{C})$
$e_0=1,e_1,e_2,e_3$	Quaternion basis, $e_k^2=-e_0$
$i$	Central scalar imaginary, $i^2=-1$
$\mathbb{M}_+,\mathbb{M}_-$	Informational (Hermitian) and material (anti-Hermitian) sectors
$\tilde{A}=iA_0e_0+\mathbf{A}\in\mathbb{M}_-$	Proca potential biquaternion
$\tilde{F}=\tilde{\nabla}^{\natural}\tilde{A}-\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{A})$	Field-strength biquaternion, $\mathrm{Sc}(\tilde{F})=0$
$\tilde{\nabla},\tilde{\nabla}^{\natural},\Box=\partial_{ict}^2+\Delta$	Gradient, conjugate, d'Alembertian
$\mathcal{L}=-\tfrac14F_{\mu\nu}F^{\mu\nu}-\tfrac12\mu^2A_\mu A^\mu$	Proca Lagrangian (series $ict$ metric)
$\partial_\mu F^{\mu\nu}-\mu^2A^\nu=0$, $\tilde{\nabla}\tilde{F}=\mu^2\tilde{A}$	Proca field equation, component and biquaternion forms
$S=\mathrm{Sc}(\tilde{\nabla}^{\natural}\tilde{A})=0$	Lorenz condition, a consequence of the equation
$(\Box-\mu^2)\tilde{A}=0$	Massive wave equation in the Lorenz gauge
$\eta_{\mu\nu}=\langle\varepsilon_\mu,\varepsilon_\nu\rangle=\mathrm{diag}(-1,+1,+1,+1)$	Series $ict$ metric; $\varepsilon_0=ie_0$, $\varepsilon_k=e_k$
$\langle\tilde{Q},\tilde{P}\rangle=\mathrm{Sc}(\tilde{Q}\tilde{P}^{\natural})$	Bilinear form on $\mathbb{M}_-$
$\pi^\mu=-F^{0\mu}$, $\pi^0\approx0$	Conjugate momenta and primary constraint
$\partial_i\pi^i+\mu^2A_0\approx0$	Secondary (Gauss) constraint, deformed by the mass
$\{\pi^0,\partial_i\pi^i+\mu^2A_0\}=-\mu^2\delta^{(3)}$	Second-class bracket; $\det\Delta=\mu^4\neq0$
$\varepsilon^r_\mu$ ($r=1,2,3$), $\varepsilon^L_\mu$	Transverse and longitudinal polarizations
$\sum_r\varepsilon^r_\mu\varepsilon^r_\nu=\eta_{\mu\nu}+p_\mu p_\nu/\mu^2$	Polarization completeness; $P^2=P$, $\mathrm{tr}P=3$, $P_{\mu\nu}p^\nu=0$
$\hat{a}_r(\mathbf{p}),\hat{a}_r^\dagger(\mathbf{p})$	Mode operators with $[\hat{a}_r,\hat{a}_s^\dagger]=\delta_{rs}(2\pi)^3\delta^{(3)}$
$:\hat{H}:=\int d^3p/(2\pi)^3\,\omega_{\mathbf{p}}\sum_r\hat{a}_r^\dagger\hat{a}_r$	Normal-ordered Hamiltonian
$D_{\mu\nu}(p)=i(\eta_{\mu\nu}+p_\mu p_\nu/\mu^2)/(p^2-\mu^2+i\epsilon)$	Proca propagator; residue is the polarization projector
$\mathrm{Tr}(\tilde{P}\tilde{H})=2\,\mathrm{Sc}(\tilde{P}\tilde{H})$	Trace formula