The First-Order Root of the Biquaternion Klein–Gordon Operator

Introduction

The biquaternion Klein–Gordon operator has a first-order root, and it is not a single operator. The root is the gradient, and it factorises the operator only together with its conjugate. The square of the gradient is not scalar: the two orders of a mixed derivative add, because $e_t$ commutes with the three spatial units. A single first-order operator therefore has no scalar square in the algebra, and no single field carries a mass that way. The free equation stays first order in one field, and the mass is carried by the chiral pair.

Every derivative in this article is the corpus gradient (Operator and Calculus Conventions, §The Gradient). With the units $e_t,\vec e_x,\vec e_y,\vec e_z$ and the coordinates $Q_\mu$,

$$ \tilde\nabla_\Phi=\sum_\mu e_\mu\frac{\partial}{\partial R_\mu}, \qquad (\tilde\nabla_\Phi)^{\natural}=\sum_\mu\varepsilon_\mu e_\mu\frac{\partial}{\partial R_\mu}, \qquad \Box_\Phi=\tilde\nabla_\Phi(\tilde\nabla_\Phi)^{\natural}=(\tilde\nabla_\Phi)^{\natural} \tilde\nabla_\Phi=\frac{\partial^2}{\partial R_t^2}+\Delta_\Phi, $$

with $\varepsilon=(1,-1,-1,-1)$ and $\Delta_\Phi=\partial^2/\partial R_x^2+\partial^2/\partial R_y^2+\partial^2/\partial R_z^2$. The gradient is an element with derivative coefficients, and it multiplies its argument on the left. Three facts are used throughout: $\Box_\Phi$ is central and scalar, the product of the gradient with its conjugate is $\Box_\Phi$, and the square of the gradient alone is not scalar. No other letter names the gradient.

The equation is

$$ \left(\Box_\Phi-k_m^{2}\right)\tilde{Ξ}=0, \qquad k_m=\dfrac{mc}{\hbar}. $$

The companion The Klein–Gordon Equation in Biquaternionic Form records that the operator is central and scalar, that the equation decouples into four scalar equations, and that no scalar first-order operator in $\mathbb{B}$ has it as its square. The companion Klein–Gordon from the Dirac Square in Biquaternionic Form records the other side: the first-order equation that exists is the chiral pair, its two members act on the two parts of one element, and its mass couples the two. Together they leave Open Question 3 of the Klein–Gordon article: is the pair the only first-orderisation, and why four components?

There are two senses of "first-order root", and the article separates them. A single square is one first-order operator whose square is $\Box_\Phi-k_m^{2}$. A factorisation is a product of two first-order factors whose product is $\Box_\Phi-k_m^{2}$, with the physical equation one factor. The single square is the stronger notion, and the algebra does not admit it: the square of the gradient carries a vector remainder, and no constant zero-order term removes a term that the derivative part supplies. The factorisation is the weaker notion, and the algebra supplies it exactly, as $\Box_\Phi=\tilde\nabla_\Phi(\tilde\nabla_\Phi)^{\natural}$ and, on the two parts of the field, as the two lines composed, $\Box_\Phi-k_m^{2}$ on each part, the pair of §The Gradient.

Three results:

  • The root is a pair, not an operator. The product $\tilde\nabla_\Phi(\tilde\nabla_\Phi)^{\natural}$ is $\Box_\Phi$, and each factor alone is not a root: on the material sector the square of the gradient is $\Box_\Phi-2\Delta_\Phi-\frac{2i}{c}\sum_\eta\vec e_\eta\partial^2/\partial r'_t\,\partial q_\eta$. No first-order operator with constant coefficients in the algebra has a scalar square, because a scalar square needs four elements that anticommute pairwise with squares $+1,-1,-1,-1$, and the algebra contains no such quadruple.
  • The free equation is first order in one field. $\tilde\nabla_\Phi\tilde{Ξ}=0$ implies $\Box_\Phi\tilde{Ξ}=0$, and at a null momentum it selects a two-dimensional subspace of the four-dimensional solution space of the second-order equation.
  • The mass needs the two parts. A scalar mass commutes with the gradient, so it cannot be the zero-order term of a first-order factor. It is the coupling of the two members of the pair, and eliminating one member returns $\Box_\Phi\tilde{Ξ}=k_m^{2}\tilde{Ξ}$.

The four components are the four coefficients of a biquaternion. The count of sixteen maps is elementary linear algebra, and the identification of the field with an element of the algebra is the corpus's dictionary. Nothing in the article predicts a mass or a coupling.

The Equation

$\Box_\Phi$ is the biquaternion d'Alembertian of the series, one operator in both coordinate frames, and the article uses it without restating it (Operator and Calculus Conventions). It is read here in the Lorentz coordinates $x_0=r'_{t}$ and $r_\eta$, related to the algebra coordinates by $Q_t=i q'_t$ and $Q_\eta=q_\eta$. Derived from that relation, its Lorentz form is

$$ \Box_\Phi=-\frac{\partial^2}{\partial {r'_t}^2}+\Delta_\Phi , $$

because the substitution turns $\partial^2/\partial R_t^2$ into $-\partial^2/\partial x_0^2$ and leaves the three spatial derivatives alone. This is the form the rest of the article uses.

Throughout, $c$ is the speed of light in the medium and $\hbar$ the reduced Planck constant. The mass enters as the length $k_m=mc/\hbar$, and its reciprocal $k_m=mc/\hbar$ is the coefficient of the pair.

The Operator and Its Symbol

The gradient is the first-order operator

The first-order operator of the series is the gradient, and it is not a root. The reason is visible in its square. Because the partial derivatives commute,

$$ \tilde\nabla_\Phi^2=\frac{\partial^2}{\partial R_t^2}-\Delta_\Phi+2\sum_\eta\vec e_\eta\frac{\partial^2}{\partial R_t\,\partial R_\eta}, $$

and on the material sector, where $\partial/\partial R_t=-i\,\partial/\partial r'_t$, the same operator is

$$ \tilde\nabla_\Phi^2=\Box_\Phi-2\Delta_\Phi-2i\sum_\eta\vec e_\eta\frac{\partial^2}{\partial r'_t\,\partial q_\eta}. $$

The mixed term is the whole difference. The three spatial units anticommute in pairs and square to $-e_t$, so the spatial part of the square is scalar. The time unit does not join them: $e_t$ is the identity of the algebra, so it commutes with every spatial unit, and the two orders of $\partial^2/\partial R_t\partial R_\eta$ add instead of cancelling. The time coefficient of the gradient is the identity, and nothing anticommutes with the identity.

The product with the conjugate is scalar. Read on the units, the two operators differ only by the sign of the three spatial terms, and the product is central because $\varepsilon_\mu e_\mu^2=\varepsilon_\mu^2e_t=e_t$ for every $\mu$:

$$ (\tilde\nabla_\Phi)^{\natural}\tilde\nabla_\Phi=\tilde\nabla_\Phi(\tilde\nabla_\Phi)^{\natural}=\Box_\Phi . $$

The symbol and the mass shell

On a plane wave $e^{i\theta_{\mathrm{pwp}}}$ with the plane-wave angle $\theta_{\mathrm{pwp}}=k_{t}^{\prime}x_0-\vec k_\Phi\cdot\vec r$ real because every factor is real, every derivative is a multiple of the wave. The substitution is $\partial/\partial R_t\mapsto k_{t}^{\prime}$ and $\partial/\partial R_\eta\mapsto-i k_{\Phi\eta}$, so the gradient becomes one fixed element of the algebra,

$$ -i\tilde K_{\Phi}=k_{t}^{\prime}-i\vec k_\Phi=k_{t}^{\prime}-i\sum_\eta\vec e_\eta k_{\Phi\eta}, \qquad (-i\tilde K_{\Phi})^{\natural}(-i\tilde K_{\Phi})=-\tilde K_{\Phi}\tilde K_{\Phi}^{\natural}=k_{t}^{\prime\,2}-\lvert\vec k_\Phi\rvert^2 , $$

and the second identity is the symbol of $\Box_\Phi$. The mass shell is the level set

$$ k_{t}^{\prime\,2}-\lvert\vec k_\Phi\rvert^2=k_m^{2} \quad\Longleftrightarrow\quad E^2=\vec p^{\,2}c^2+m^2c^4 , $$

with $\vec k_\Phi=\vec p\,\frac{1}{\hbar}$. Here $k_{\Phi}=\frac{p}{\hbar}$ and $k_{t}^{\prime}=E/(\hbar\,c)$ are the momentum and energy wave numbers.

The symbol of $\Box_\Phi$ is the interval of the wave covector, and the mass shell is that symbol set to $k_m^{2}$. The symbol of the gradient is a zero divisor on exactly the same set: left multiplication by $-i\tilde K_{\Phi}$ is invertible off the cone and has rank two on it. The equation $\tilde\nabla_\Phi\tilde{Ξ}=0$ therefore has a nonzero plane-wave solution exactly where the massless Klein–Gordon equation does, and it halves the solution space there, as §What the Root Does Not Carry records.

The Factorisation

The two factors

The product of the gradient and its conjugate is the d'Alembertian, and the square of the gradient is not: $(\tilde\nabla_\Phi)^{\natural}\tilde\nabla_\Phi=\tilde\nabla_\Phi(\tilde\nabla_\Phi)^{\natural}=\Box_\Phi$, while $\tilde\nabla_\Phi^2$ keeps the vector remainder of §The gradient is the first-order operator.

The identity is exact, and it is the whole factorisation of $\Box_\Phi$ that the algebra supplies. It has two factors and neither is a root by itself. That is the sense in which the root exists: as a pair of first-order operators, not as one operator.

No single square

The objection is immediate. The two factors are not one operator, so perhaps a different first-order operator is one. The algebra answers no, and the answer is short.

A first-order operator with constant coefficients in the algebra and a constant zero-order term is

$$ \sum_\mu a_\mu\frac{\partial}{\partial R_\mu}+b, \qquad a_\mu,b\in\mathbb{B}. $$

Its square has second-order part $\frac{1}{2}\sum_{\mu,\nu}\left(a_\mu a_\nu+a_\nu a_\mu\right)\partial^2/\partial R_\mu\partial R_\nu$, because the derivatives commute and only the symmetric part of the coefficient survives. A scalar square therefore needs

$$ a_\mu a_\nu+a_\nu a_\mu=2\varepsilon_\mu\delta_{\mu\nu}e_t , $$

that is, four elements that anticommute pairwise, one of square $+1$ and three of square $-1$. No such quadruple exists in $\mathbb{B}$. The three of square $-1$ form a quaternionic triple; they generate the algebra, and an element that anticommutes with all three is zero (§No element anticommutes with the three spatial units), so no element of square $+1$ is left to be the fourth. The gradient is the instance: its three spatial coefficients are the units, and its fourth coefficient is the identity, which commutes with them.

The zero-order term does not repair this, because it enters the second-order part not at all. It contributes $\sum_\mu(a_\mu b+ba_\mu)\partial/\partial R_\mu+b^2$, first order and zeroth order, and the obstruction sits in the second order. So the single square fails for every operator of this form, and the factorisation is the strongest statement the algebra supports.

What the one-sided restriction costs, and what it does not. The no-go is a statement about coefficients that multiply the field on one side. An earlier form of this argument let them act on the right as well, and with that freedom the obstruction lifts: the fourth coefficient is a right multiplication, its anticommutation with the other three comes from the right factor, and the zero-order term is a multiple of the right multiplication by $\vec e_z$, the pseudoscalar $\omega=-\vec e_z$, which anticommutes with all four and squares to $-k_m^{2}$. That form has a single square, and the mass it carries is parity-odd, a pseudoscalar mass rather than a scalar.

The corpus's calculus has one side: every equation is written on the left, and a right multiplication is a left multiplication on the conjugate field (Operator and Calculus Conventions, §The Two Sides and the Renaming Theorem). The single square of that form therefore lies outside the calculus and is not a counterexample to the no-go above. What does not depend on the two-sided form, and what the rest of the article uses, is that the scalar mass has no single square and forces the pair.

The Field and the Four Components

The carrier is four-dimensional

The four terms of the gradient multiply by four different units, and a field that carries them is an element of the algebra. Every complex-linear map of the algebra into itself is a sum of the sixteen $\tilde{Ξ}\mapsto e_\mu\tilde{Ξ} e_\nu$, and those sixteen span $\operatorname{End}_{\mathbb{C}}(\mathbb{B})\cong M_4(\mathbb{C})$, so sixteen independent maps act on the four complex coefficients. A complex space of dimension $n$ carries $n^2$ independent complex-linear maps, so sixteen maps need $n\ge4$, and the algebra attains the bound with $n=4$.

The four components are therefore the four coefficients of a biquaternion, and they are not an added multiplicity. The root does not enlarge the field space relative to a biquaternion field. It changes the action, not the carrier.

No element anticommutes with the three spatial units

The no-go of §No single square uses one coefficient fact, proved here. Write a general element as $X=a_te_t+a_x\vec e_x+a_y\vec e_y+a_z\vec e_z$ with $a_\mu\in\mathbb{C}$. The anticommutator with $\vec e_x$ is

$$ \{X,\vec e_x\}=2a_t\vec e_x-2a_x e_t=0, $$

which forces $a_t=a_x=0$. Anticommuting with $\vec e_y$ then forces $a_y=0$, leaving $X=a_z\vec e_z$. But $a_z\vec e_z$ commutes with the third unit, so it does not anticommute with it. The only solution is $X=0$. A quaternionic triple other than the spatial units is a conjugate of them, and the same computation applies to it.

So the obstruction is not that the algebra lacks room for a fourth direction. It is that a fourth direction cannot be an element with the required anticommutation, and the identity is what the fourth coefficient of the gradient would have to be.

A single complex number has no first-order equation

A field whose value is one complex number satisfies the second-order equation: $\Box_\Phi\xi=0$ is one condition on one function. It satisfies no first-order equation of the series. The gradient acting on it returns an algebra element with four coefficients,

$$ \tilde\nabla_\Phi\xi=\sum_\mu e_\mu\frac{\partial\xi}{\partial R_\mu}, $$

and setting it to zero imposes four independent complex conditions, one per unit. The field is then constant. A root requires a field with room for the four units, and the algebra supplies exactly four complex dimensions.

| Field | First-order equation | Root? | |+-|+-|+-| | One complex number | $\tilde\nabla_\Phi\xi=0$, which forces $\partial\xi/\partial R_\mu=0$ | no, it has no solution that propagates | | Element of $\mathbb{B}$, massless | $\tilde\nabla_\Phi\tilde{Ξ}=0$ | yes, one factor of $\Box_\Phi$ | | Element of $\mathbb{B}$, massive | $\tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}$ and $(\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}$ | yes, the pair |

The Mass and the Chiral Pair

The mass is central

A physical mass is a scalar: it multiplies every element by the same constant, so it commutes with the gradient. It cannot therefore be the zero-order term of a first-order factor. No zero-order term anticommutes with all four left multiplications of the gradient: $\ell_{e_t}$ is the identity map, and a map $B$ with $\{\ell_{e_\mu},B\}=0$ for every $\mu$ is zero (§Verification).

The scalar mass therefore fails as the shift of a single operator, and the factorisation of $\Box_\Phi-k_m^{2}$ into two first-order factors on one field is not available in the algebra. What is available is the square of a first-order operator that carries the two parts.

The pair

Write the field as one element with two parts, $\tilde{Ξ}=\tilde{Ξ}_{L}+\tilde{Ξ}_{R}$. The first-order equation of the series is the chiral pair

$$ \tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}, \qquad (\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}, $$

with the same coefficient $k_m=m\,c/\hbar$ as above. Substituting the first into the second gives $(\tilde\nabla_\Phi)^{\natural}\tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m^{2}\tilde{Ξ}_{R}$, that is

$$ \Box_\Phi\tilde{Ξ}_{R}=k_m^{2}\tilde{Ξ}_{R} , $$

and the first member then defines $\tilde{Ξ}_{L}$. Each part therefore satisfies Klein–Gordon, and the pair is a bijection with the second-order solutions of one part, not a new restriction.

The factorisation is not available on one field, and the reason is the same vector remainder. With the mass as a shift,

$$ \left((\tilde\nabla_\Phi)^{\natural}-k_m\right)\left(\tilde\nabla_\Phi+k_m\right) =\Box_\Phi+k_m\left((\tilde\nabla_\Phi)^{\natural}-\tilde\nabla_\Phi\right)-k_m^{2} =\Box_\Phi-2k_m\sum_\eta\vec e_\eta\frac{\partial}{\partial R_\eta}-k_m^{2} , $$

which is not scalar. The two cross terms do not cancel, because they are different operators. Only the pair removes them, by substituting one member into the other.

The two lines

The two equations are two lines on two elements, and no single letter is written for them (Operator and Calculus Conventions, §The Two First-Order Equations and Their Symbol). Write $\tilde{Ξ}=(\tilde{Ξ}_{L},\tilde{Ξ}_{R})$ and

$$ \tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}, \qquad (\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}. $$

Composing each line with the conjugate of the other gives $\left(\Box_\Phi-k_m^{2}\right)\tilde{Ξ}_{L}=0$ and $\left(\Box_\Phi-k_m^{2}\right)\tilde{Ξ}_{R}=0$. This is the precise sense in which a mass needs a pair: the doubling is the carrier of the factorisation, not a doubling of particles, and the two parts are the two labelled halves of one field.

The mass term is a multiple of the identity on the two slots and cannot move between them; the gradient is the off-diagonal block, and it is the gradient that carries the mass across. A scalar mass cannot sit in a single factor, for the reason of §No single square; the pair supplies the second slot that the mass needs in order to appear as an equality.

The determinant

On a plane wave the two lines are the block matrix with $\ell_{i\tilde K_{\Phi}}$ and $\ell_{i\tilde K_{\Phi}^{\natural}}$ off the diagonal, where $\ell_M$ is left multiplication by $M$, and the determinant of that matrix with $-k_m$ on the diagonal is

$$ \det\left(\ell_{i\tilde K_{\Phi}^{\natural}}\ell_{i\tilde K_{\Phi}}-k_m^{2}\right) =\left(k_{t}^{\prime\,2}-\lvert\vec k_\Phi\rvert^2-k_m^{2}\right)^4 , $$

the fourth power of the symbol of $\Box_\Phi-k_m^{2}$: eliminating one block leaves the Schur complement $1/k_m\left(k_{t}^{\prime\,2}-\lvert\vec k_\Phi\rvert^2-k_m^{2}\right)$ times the four-dimensional identity.

The determinant is a product of eigenvalues, and it is not a square root. A determinant identity says that a power of the scalar symbol is the determinant of a first-order operator; a root needs one first-order operator whose square is the scalar operator, and the power four is the evidence that this operator lives on the doubled carrier. The vanishing is a consistency check on the factorisation, and not a construction of a single root.

What the Root Does Not Carry

The classification is a statement of form, and its boundaries are part of the result.

  1. It is stronger than Klein–Gordon. The massless first-order equation selects a subspace of the second-order solution space. At a null momentum the second-order equation admits every element, four complex dimensions, while $\tilde\nabla_\Phi\tilde{Ξ}=0$ admits two: at $k=(1,1,0,0)$ left multiplication by $-i\tilde K_{\Phi}$ has rank two and kernel two, and the kernel carries one state of each value of the chirality, the right action $\tilde{Ξ}\mapsto-i\tilde{Ξ}\vec e_z$. What the root halves is the room for the field at that momentum, not the spectrum.
  2. No mass and no coupling are predicted. The classification says which first-order structures have the Klein–Gordon second order; it does not say which one nature uses, and it does not determine $1/k_m$. The sign of the mass carries no content either: the pair at $k_m$ and the pair at $-k_m$ are the same equations with the sign of one part flipped.
  3. The coefficients are constant and the background flat. A position-dependent coefficient adds a term in which the derivative acts on the coefficient, and a gauged derivative adds a field strength; the constant-coefficient classification is then the starting point rather than the answer. On a curved background the derivatives become covariant, and the classification applies in the tangent space at each point. The $x$ coordinates are flat; the algebra coordinates give the same operator and not a different root.
  4. The real structure is open. The article works over $\mathbb{C}$. Whether the gradient can be given a real structure, and which sign that structure carries, is the charge-conjugation question, and the article does not settle it. The index question is a different one again, and the massless kernel of §The symbol and the mass shell splits by the chirality without the gradient separating the two values.

The Equation and the Root

Two statements that look like one. The Klein–Gordon equation is central: its operator multiplies every element by the same scalar, so a single complex scalar satisfies it. The Klein–Gordon root is not central: it is an element with derivative coefficients, and it multiplies the field on the left, so its action depends on the element it meets. The equation lives in the centre; the root acts with the four units.

The two statements are compatible because they are about different objects. The equation needs the scalar; the root needs the element, and the element is what supplies the four units. A product of two first-order factors is scalar, and no single factor is.

The doubling is not a doubling of particles. The two values of the label are the two parts of one element, and the second member of the pair is determined by the first through the gradient.

Verification

The identities were recomputed before they were written. The algebra multiplication and the left and right multiplications were built as four-dimensional complex matrices, and the symbol reductions were checked at random wave covectors. The residuals are the largest absolute difference over the stated samples, and $k_m$ is a fixed real mass coefficient.

| Check | Result | |+-|+-| | $(-i\tilde K_{\Phi})^{\natural}(-i\tilde K_{\Phi})-(k_{t}^{\prime\,2}-\lvert\vec k_\Phi\rvert^2)e_t$ on $100$ random wave covectors | $8.9\times10^{-16}$ | | the symbol of the square, $\tilde K_{\Phi}^2+\left(k_{t}^{\prime\,2}+\lvert\vec k_\Phi\rvert^2\right)e_t-2ik_{t}^{\prime}\vec k_\Phi$, on $100$ random wave covectors | $1.8\times10^{-15}$ | | the sixteen maps $\tilde{Ξ}\mapsto e_\mu\tilde{Ξ} e_\nu$ are independent | rank $16$ | | no element anticommutes with the three spatial units | kernel dimension $0$ | | no map anticommutes with the four left multiplications | dimension $0$ | | the same for a generic quaternionic triple, over $100$ conjugates | kernel dimension $0$ | | $\det\ell_{-i\tilde K_{\Phi}}-(k_{t}^{\prime\,2}-\lvert\vec k_\Phi\rvert^2)^2$ on $100$ random wave covectors | $5.7\times10^{-14}$ | | rank and kernel of $\ell_{-i\tilde K_{\Phi}}$ at five null wave covectors | rank $2$, kernel $2$ (exact) | | the same off the cone, at $k=(2,1,0,0)$ | rank $4$ | | the chirality on the null kernel, the eigenvalues of the right action by $\vec e_z$ | $\pm i$, one of each | | $\ell_{i\tilde K_{\Phi}^{\natural}}\ell_{i\tilde K_{\Phi}}-\Box_\Phi$ on the eight slots | $1.8\times10^{-15}$ | | $\det\left(\ell_{i\tilde K_{\Phi}^{\natural}}\ell_{i\tilde K_{\Phi}}-k_m^{2}\right)-(\Box_\Phi-k_m^{2})^4$ on $100$ random wave covectors | $2.5\times10^{-11}$ | | the elimination, $(\tilde\nabla_\Phi)^{\natural}\tilde\nabla_\Phi\tilde{Ξ}_{R}/k_m^{2}-\tilde{Ξ}_{R}$ on shell | $3.6\times10^{-15}$ |

The numerical checks are checks of polynomial identities at random points. The proof is the algebra of the preceding sections; the numbers guard against a sign or factor error in the transcription. Two exact values are quoted in the text: at $k=(2,1,0,0)$ the norm $(-i\tilde K_{\Phi})^{\natural}(-i\tilde K_{\Phi})$ is $3$, the determinant $\det\ell_{-i\tilde K_{\Phi}}$ is $9$, and at $k_m=1$ the determinant of the two lines, $\det(\ell_{i\tilde K_{\Phi}^{\natural}}\ell_{-i\tilde K_{\Phi}}-k_m^{2})$, is $16$.

Summary

The first-order root of the biquaternion Klein–Gordon operator exists, and it is a pair of first-order operators rather than one: the gradient and its conjugate, with $\tilde\nabla_\Phi(\tilde\nabla_\Phi)^{\natural}=\Box_\Phi$. The square of the gradient alone is not scalar, because $e_t$ is the identity and commutes with the three spatial units, so the mixed derivatives add. No first-order operator with constant element coefficients acting on the left has a scalar square either, because that needs four elements anticommuting pairwise with squares $+1,-1,-1,-1$, and the three of square $-1$ already generate the algebra.

The free equation is first order in one field, $\tilde\nabla_\Phi\tilde{Ξ}=0$, and it implies the massless Klein–Gordon equation. At a null momentum it selects two of the four complex dimensions of the second-order solution space, one state of each value of the chirality.

The mass is scalar and central. It commutes with the gradient, so it cannot be the zero-order term of a first-order factor; it is the coupling of the two parts of one element, and the two lines carry it, composing to $\Box_\Phi-k_m^{2}$ on each part. Eliminating one member returns $\Box_\Phi\tilde{Ξ}_{R}=k_m^{2}\tilde{Ξ}_{R}$, so the pair and the second-order equation on one part are equivalent.

A field whose value is one complex number has no first-order equation: the four terms of the gradient impose four independent conditions on one function. The results are a classification of form. No mass, no coupling and no dynamics are predicted. The open questions are the gauged, curved, position-dependent and real-structure extensions.

Summary of Notation

| Symbol | Meaning | |+-|+-| | $\mathbb{B}$ | The biquaternion algebra | | $e_t,\vec e_x,\vec e_y,\vec e_z$ | The four units | | $i$ | The central imaginary unit | | $Q_\mu$ | The algebra coordinates | | $x_\mu$ | The Lorentz coordinates | | $\tilde\nabla_\Phi$ | Biquaternionic gradient, an element with derivative coefficients, acting on the left | | $(\tilde\nabla_\Phi)^{\natural}$ | Its quaternion conjugate, the second factor of $\Box_\Phi$ | | $\Box_\Phi$ | d'Alembertian; its Lorentz form is derived in §The Equation | | $c$ | The speed of light in the medium | | $1/k_m$ | The mass length, $\hbar/(mc)$; its reciprocal $k_m=mc/\hbar$ is the inverse reduced Compton wavelength | | $k_{\Phi,\mu}$ | The wave covector of a plane wave | | $\omega=-\vec e_z$ | The pseudoscalar of the algebra, $\omega^2=-e_t$ | | $-i\tilde K_{\Phi}=k_{t}^{\prime}-i\vec k_\Phi$ | The symbol of the gradient, an element of $\mathbb{B}$ | | $\ell_M$ | Left multiplication by $M$ | | $\tilde{Ξ}$ | The field of the Klein–Gordon equation, an element of $\mathbb{B}$ | | $\tilde{Ξ}_{L},\tilde{Ξ}_{R}$ | The two parts of an element, the two values of the label | | $\tilde\nabla_\Phi\tilde{Ξ}_{R}=k_m\tilde{Ξ}_{L}$, $(\tilde\nabla_\Phi)^{\natural}\tilde{Ξ}_{L}=k_m\tilde{Ξ}_{R}$ | The pair as two first-order lines; each composed with the conjugate of the other is $\Box_\Phi$ per part | | the label | $L$ or $R$, two values, carried by the parts of an element |

Further Reading

  • P. A. M. Dirac, "The quantum theory of the electron," Proceedings of the Royal Society A 117 (1928) 610–624, for the first-order equation and its second-order consequence.
  • O. Klein, "Elektrodynamik und Wellenmechanik vom Standpunkt des Korrespondenzprinzips," Zeitschrift für Physik 41 (1927) 407–442, and W. Gordon, "Der Comptoneffekt nach der Schrödingerschen Theorie," Zeitschrift für Physik 40 (1926) 117–133, for the second-order relativistic wave equation.
  • J. D. Bjorken and S. D. Drell, Relativistic Quantum Mechanics (McGraw-Hill, 1964), for the factorisation of the wave operator into two first-order factors and the plane-wave solutions.
  • P. Lounesto, Clifford Algebras and Spinors (Cambridge University Press, 2001), for the classification of Clifford algebras, the minimal modules, and the square root of the wave operator.
  • H. B. Lawson and M.-L. Michelsohn, Spin Geometry (Princeton University Press, 1989), for the complex and real Clifford algebras and their irreducible modules.
  • I. R. Porteous, Clifford Algebras and the Classical Groups (Cambridge University Press, 1995), for the periodicity and the module dimensions.
  • V. V. Kravchenko, Quaternionic Analysis and Its Applications (Birkhäuser, 2003), for the quaternionic Dirac operator and its relation to the Euclidean Clifford algebra.
  • The corpus articles The Klein–Gordon Equation in Biquaternionic Form, Klein–Gordon from the Dirac Square in Biquaternionic Form, The Dirac Equation in Biquaternionic Form, The Biquaternion D'Alembertian and Its Green's Functions, and Algebra and Element Conventions for the conventions and the results this article builds on.
  • The corpus articles Operators Acting on Two Sides, Element and Operator Representations: The Three Actions, and The Polarized Module of the Biquaternion Algebra for the two-sided algebra and the nilpotent that exchanges the two parts of an element.