Introduction
The class of U ( 1 ) U(1) U ( 1 ) duality-invariant theories extends well beyond Maxwell and Born–Infeld theory [1 ] ; reviews of the bosonic and supersymmetric formulations are given in [2 , 3 ] . ModMax is a one-parameter nonlinear extension of four-dimensional Maxwell electrodynamics that preserves both conformal invariance and U ( 1 ) U(1) U ( 1 ) duality invariance [4 ] . With metric ( − , + , + , + ) (-,+,+,+) ( − , + , + , + ) , its Lagrangian is
x = − 1 4 F μ ν F μ ν , p = − 1 4 F μ ν F ~ μ ν , r = x 2 + p 2 , L M M = x cosh γ + r sinh γ . (1) x=-\frac{1}{4}F_{\mu\nu}F^{\mu\nu},\qquad p=-\frac{1}{4}F_{\mu\nu}\widetilde{F}^{\mu\nu},\qquad r=\sqrt{x^2+p^2},\qquad L_{\mathrm{MM}}=x\cosh\gamma+r\sinh\gamma. \tag*{(1)} x = − 4 1 F μν F μν , p = − 4 1 F μν F μν , r = x 2 + p 2 , L MM = x cosh γ + r sinh γ . (1)
The square root is nonanalytic at r = 0 r=0 r = 0 . Accordingly, our construction is a power series in the coupling on the domain r > 0 r>0 r > 0 , rather than a Taylor expansion in the field strength about the vacuum. The coupling of ModMax to charged particles was studied in [5 ] .
Supersymmetric Born–Infeld theory relates nonlinear electrodynamics to partial supersymmetry breaking [6 ] ; manifestly N = 2 \mathcal{N}=2 N = 2 actions and their iterative extensions were studied in [7 , 8 , 9 ] . Ivanov and Zupnik gave an auxiliary-superfield formulation of the N = 2 \mathcal{N}=2 N = 2 Born–Infeld theory with spontaneously broken N = 4 \mathcal{N}=4 N = 4 supersymmetry, verified its self-duality to tenth order, and proposed a recursive construction of the action [10 ] . Supersymmetric self-duality was formulated in [2 , 11 , 12 ] . Auxiliary-field formulations express self-duality as phase invariance of the auxiliary interaction. This approach was developed for bosonic electrodynamics in [13 , 14 , 15 ] and extended to superspace by Kuzenko [16 ] and, independently for N = 1 \mathcal{N}=1 N = 1 , by Ivanov, Lechtenfeld and Zupnik [17 ] . Bandos, Lechner, Sorokin and Townsend obtained the off-shell N = 1 \mathcal{N}=1 N = 1 ModMax action and its super-Weyl invariant coupling [18 ] . Further superconformal and nonlinear models appear in [19 , 20 , 21 , 22 ] , including higher-derivative deformations [23 ] .
The N = 2 \mathcal{N}=2 N = 2 vector multiplet decomposes into an N = 1 \mathcal{N}=1 N = 1 vector and a chiral multiplet Φ \Phi Φ . Setting Φ = 0 \Phi=0 Φ = 0 removes the complex scalar, second gaugino and two real auxiliary fields. Agreement with N = 1 \mathcal{N}=1 N = 1 ModMax therefore leaves their dynamics untested. Our equivalence uses a real vector-prepotential redefinition after this restriction; related nonlinear supersymmetric changes of variables were studied in [24 , 25 ] .
Kuzenko and Raptakis observed that an N = 2 \mathcal{N}=2 N = 2 superconformal extension of ModMax had not been constructed, identifying the dimensionless full N = 2 \mathcal{N}=2 N = 2 superspace measure as the main difficulty, and presented a model involving a compensating vector multiplet [20 section 4.3] . We compare the two constructions in § 8.4 . In this paper we construct a family of superconformal primary interactions built from the Abelian reduced chiral field strength alone, solve the self-duality equation to all orders with prescribed phase-invariant terms, and show that, with Φ \Phi Φ switched off, the resulting action is equivalent off shell to the N = 1 \mathcal{N}=1 N = 1 ModMax action. We work in ordinary flat N = 2 \mathcal{N}=2 N = 2 superspace and use the rigid superconformal Killing formalism of [26 ] . No compensator or other additional multiplet enters the action. The change of variables used to construct the completion is the N = 2 \mathcal{N}=2 N = 2 auxiliary-superfield representation of [16 section 4] , also used in [10 ] , which we rederive in our normalization. A recent account of nonlinear self-duality for different amounts of supersymmetry is given in [27 ] .
In this paper, we construct a compensator-free N = 2 \mathcal{N}=2 N = 2 superconformal and self-dual extension of ModMax electrodynamics. The action takes the form of a formal series:
S ~ t = S 0 + ∑ n ≥ 1 t n F ~ n , t = sinh γ , F ~ 1 = f m o d , Π 0 F ~ n = [ t n ] d ( t ) Δ ( n ≥ 2 ) , d ( t ) = 1 19 ( 2 t 1 + t 2 + 1 − t ) . (2) \begin{aligned}
\widetilde{S}_t&=S_0+\sum_{n\geq1}t^n\widetilde{F}_n,\qquad t=\sinh\gamma,\qquad\widetilde{F}_1=f^{\mathrm{mod}},\\
\Pi_0\widetilde{F}_n&=[t^n]d(t)\,\Delta\qquad(n\geq2),\\
d(t)&=\frac{1}{19}\left(\frac{2t}{\sqrt{1+t^2}+1}-t\right). \tag*{(2)}
\end{aligned} S t Π 0 F n d ( t ) = S 0 + n ≥ 1 ∑ t n F n , t = sinh γ , F 1 = f mod , = [ t n ] d ( t ) Δ ( n ≥ 2 ) , = 19 1 ( 1 + t 2 + 1 2 t − t ) . (2)
Here f m o d f_{\mathrm{mod}} f mod is the first-order interaction of § 5 , built from the minimal square-root interaction T 1 T_1 T 1 and one further primary density. The combination Δ = f m o d − T 1 \Delta= f_{\mathrm{mod}} - T_1 Δ = f mod − T 1 is given in (50) and vanishes under the N = 1 \mathcal{N}=1 N = 1 vector restriction Φ = 0 \Phi=0 Φ = 0 and therefore does not alter the reduction to the Bandos action. The phase average Π 0 \Pi_0 Π 0 is defined in (33) . The charged parts of the higher coefficients follow from the self-duality equation: the displayed phase-invariant series is not simply added to an already completed action. For comparison we also consider the canonical completion, in which all higher-order phase-invariant terms vanish. The choice of d ( t ) d(t) d ( t ) comes from the zero-vev scalar analysis. § 8.6 shows that, within the quadratic scalar form used here and with Maxwell initial data, the root-flow condition modulo equations of motion fixes this function. These data do not uniquely select the full superspace completion.
Two inductions establish the formal self-dual N = 2 \mathcal{N}=2 N = 2 action and its off-shell N = 1 \mathcal{N}=1 N = 1 equivalence. The auxiliary-superfield formalism is that of [10 , 16 ] ; its use with prescribed phase averages, and the all-order prepotential matching, are the steps needed here.
We study fluctuations about constant non-null electromagnetic backgrounds at vanishing scalar expectation value. The chiral-multiplet fluctuations are retained, so this calculation is distinct from setting Φ = 0 \Phi=0 Φ = 0 . The minimal interaction gives a superluminal scalar already at first order. One additional primary density reverses this coupling without changing the N = 1 \mathcal{N}=1 N = 1 reduction. For (2) , the zero-vev scalar quadratic action is determined to all orders and has an analytic continuation with v ⊥ 2 = e − γ v_{\perp}^{2}=e^{-\gamma} v ⊥ 2 = e − γ and v ∥ 2 = 1 v_{\parallel}^{2}=1 v ∥ 2 = 1 for finite γ ≥ 0 \gamma\ge0 γ ≥ 0 . On the same backgrounds, the N = 1 \mathcal{N}=1 N = 1 equivalence and S U ( 2 ) R SU(2)_R S U ( 2 ) R symmetry determine both gaugino operators and the auxiliary triplet. The gauginos remain luminal with positive kinetic coefficients, and the auxiliary fields remain algebraic.
The characteristic analysis of nonlinear theories distinguishes a second-order effective metric [28 , 29 , 30 ] from finite-order dispersive corrections and front velocities [31 , 32 ] . Photon causality in self-dual electrodynamics [33 , 34 , 35 ] does not settle the additional scalar sector. Here the scalar quadratic operator is exactly second order in derivatives on the specified backgrounds.
Stress-tensor flows offer another route to ModMax-type theories [36 , 37 , 38 ] ; marginal square-root deformations were also studied in two dimensions [39 ] . The relation between duality-preserving deformations and stress-tensor flows was established in [40 ] for Lagrangians depending on the field strength but not on its derivatives, and combined irrelevant and marginal flows were studied in [41 ] . The photon and zero-vev scalar quadratic sectors agree with the bosonic square-root flow, the latter modulo equations of motion and total derivatives. This agreement selects the scalar coefficient function; it does not independently establish a flow for the full action. Supersymmetric flows involve complete conserved supercurrent multiplets and their allowed improvements [38 , 42 , 43 ] . Constructing such a flow is a separate problem from the self-dual completion studied here.
Non-Abelian ModMax models with self-dual instantons, gravitating dyons [44 ] and merons [45 ] have also been studied. Instanton self-duality is distinct from the continuous electric–magnetic duality considered here for an Abelian N = 2 \mathcal{N}=2 N = 2 vector multiplet.
The all-order construction is formal. Convergence of the full nonlinear series and propagation on more general backgrounds remain open. The finite-coupling statements refer to the analytic continuation of the quadratic series at vanishing scalar expectation value.
Sections 2–5 give the formal construction; Sections 6–7 establish the N = 1 \mathcal{N}=1 N = 1 equivalence. Section 8 contains the propagation results and comparisons with earlier work. The appendices supply the finite field-redefinition identities, the scalar recursion and the superconformal variations. The first-order interaction is specified in Section 5.2; Appendix C gives the superconformal variations.
Conventions and constraints
Flat superspace and the reduced-chiral constraint
We work in ordinary four-dimensional N = 2 \mathcal{N}=2 N = 2 superspace with coordinates ( x μ , θ i α , θ ˉ α ˙ i ) (x^\mu,\theta^\alpha_i,\bar{\theta}^{\dot{\alpha}i}) ( x μ , θ i α , θ ˉ α ˙ i ) , where i = 1 , 2 i=1,2 i = 1 , 2 is an S U ( 2 ) SU(2) S U ( 2 ) internal isospin index. The flat spinor derivatives obey
{ D α i , D ˉ β ˙ j } = − 2 i δ j i ∂ α β ˙ , { D α i , D β j } = { D ˉ α ˙ i , D ˉ β ˙ j } = 0. (3) \{D^i_\alpha,\bar{D}_{\dot{\beta}j}\}=-2i\delta^i_j\partial_{\alpha\dot{\beta}},\qquad\{D^i_\alpha,D^j_\beta\}=\{\bar{D}_{\dot{\alpha}i},\bar{D}_{\dot{\beta}j}\}=0. \tag*{(3)} { D α i , D ˉ β ˙ j } = − 2 i δ j i ∂ α β ˙ , { D α i , D β j } = { D ˉ α ˙ i , D ˉ β ˙ j } = 0. (3)
Internal indices are raised and lowered with the antisymmetric S U ( 2 ) SU(2) S U ( 2 ) tensors, with ϵ 12 = 1 \epsilon_{12}=1 ϵ 12 = 1 and ϵ i j ϵ j k = δ k i \epsilon^{ij}\epsilon_{jk}=\delta^i_k ϵ ij ϵ j k = δ k i . All spinor squares are normalized with a factor 1 2 \frac{1}{2} 2 1 :
( D i ) 2 : = 1 2 D α i D α i , ( D ˉ i ) 2 : = 1 2 D ˉ α ˙ i D ˉ α ˙ i . (4) (D^i)^2:=\frac{1}{2}D^{\alpha i}D^i_\alpha,\qquad(\bar{D}_i)^2:=\frac{1}{2}\bar{D}_{\dot{\alpha}i}\bar{D}^{\dot{\alpha}i}. \tag*{(4)} ( D i ) 2 := 2 1 D α i D α i , ( D ˉ i ) 2 := 2 1 D ˉ α ˙ i D ˉ α ˙ i . (4)
Thus ( D 2 ) 2 (D^2)^2 ( D 2 ) 2 denotes the square of the derivative with internal index i = 2 i=2 i = 2 ; not to be confused with the N = 1 \mathcal{N}=1 N = 1 operator D 2 = 1 2 D α D α D^2=\frac{1}{2}D^\alpha D_\alpha D 2 = 2 1 D α D α introduced only after the truncation.
Define the symmetric triplet operators in the same normalization by
D i j : = 1 2 D α ( i D α j ) , D ˉ i j : = 1 2 D ˉ α ˙ ( i D ˉ j ) α ˙ . (5) D^{ij}:=\frac{1}{2}D^{\alpha(i}D^{j)}_\alpha,\qquad\bar{D}^{ij}:=\frac{1}{2}\bar{D}^{(i}_{\dot{\alpha}}\bar{D}^{j)\dot{\alpha}}. \tag*{(5)} D ij := 2 1 D α ( i D α j ) , D ˉ ij := 2 1 D ˉ α ˙ ( i D ˉ j ) α ˙ . (5)
Parentheses indicate symmetrization with unit weight. The Abelian vector-multiplet field strength W \mathcal{W} W is an S U ( 2 ) R SU(2)_{\mathrm{R}} S U ( 2 ) R singlet satisfying
D ˉ α ˙ i W = 0 , D i j W = D ˉ i j W ‾ . (6) \bar{D}_{\dot{\alpha}i}\mathcal{W}=0,\qquad D^{ij}\mathcal{W}=\bar{D}^{ij}\overline{\mathcal{W}}. \tag*{(6)} D ˉ α ˙ i W = 0 , D ij W = D ˉ ij W . (6)
The second relation is the Bianchi identity defining the reduced-chiral field strength and holds independently of the equations of motion. For the free Maxwell theory, the equation of motion additionally implies D i j W = 0 D^{ij}\mathcal{W}=0 D ij W = 0 . The reduced-chiral description is reviewed in [46 ] . The derivative with i = 1 i=1 i = 1 becomes the manifest N = 1 \mathcal{N}=1 N = 1 derivative only when the truncation to the N = 1 \mathcal{N}=1 N = 1 vector multiplet is made in § 6 .
Measures and functional differentiation
We write the spacetime and Grassmann integration measures explicitly. To abbreviate repeated functional pairings, define
I C [ A ] = ∫ d 4 x d 4 θ A , I F [ B ] = ∫ d 4 x d 8 θ B , \mathcal{I}_{\mathrm{C}}[A]=\int d^4x\,d^4\theta\,A,\qquad\mathcal{I}_{\mathrm{F}}[B]=\int d^4x\,d^8\theta\,B, I C [ A ] = ∫ d 4 x d 4 θ A , I F [ B ] = ∫ d 4 x d 8 θ B ,
I F [ B ] = I C [ P B ‾ ] , P : = ( D ˉ 1 ) 2 ( D ˉ 2 ) 2 . (7) \mathcal{I}_{\mathrm{F}}[B]=\mathcal{I}_{\mathrm{C}}[\mathcal{P}\overline{B}],\qquad\mathcal{P}:=(\bar{D}_1)^2(\bar{D}_2)^2. \tag*{(7)} I F [ B ] = I C [ P B ] , P := ( D ˉ 1 ) 2 ( D ˉ 2 ) 2 . (7)
Here d 4 θ d^4\theta d 4 θ is the chiral N = 2 \mathcal{N}=2 N = 2 measure and d 8 θ d^8\theta d 8 θ is the full N = 2 \mathcal{N}=2 N = 2 measure. The last equality fixes their relative normalization. The product of barred squares is invariant under constant S U ( 2 ) R SU(2)_{\mathrm{R}} S U ( 2 ) R changes of basis and maps an unconstrained scalar to a chiral one. The N = 1 \mathcal{N}=1 N = 1 measure normalization used after the truncation is in (53) .
The free action and the definition of a chiral functional derivative are
S 0 = 1 8 ( I C [ W 2 ] + c . c . ) , δ F = I C [ δ W δ F δ W ] + I ‾ C [ δ W ‾ δ F δ W ‾ ] . (8) S_0=\frac{1}{8}\left(\mathcal{I}_{\mathrm{C}}[\mathcal{W}^2]+\mathrm{c.c.}\right),\qquad\delta F=\mathcal{I}_{\mathrm{C}}\left[\delta\mathcal{W}\frac{\delta F}{\delta\mathcal{W}}\right]+\overline{\mathcal{I}}_{\mathrm{C}}\left[\delta\overline{\mathcal{W}}\frac{\delta F}{\delta\overline{\mathcal{W}}}\right]. \tag*{(8)} S 0 = 8 1 ( I C [ W 2 ] + c.c. ) , δ F = I C [ δ W δ W δ F ] + I C [ δ W δ W δ F ] . (8)
To define the functional derivatives, we treat W \mathcal{W} W and W ‾ \overline{\mathcal{W}} W as independent chiral and antichiral superfields, without imposing the Bianchi identity [11 ] . We differentiate the explicit superspace expressions given below. The Bianchi identity is imposed only afterwards, on physical configurations.
Notation
Several letters have different meanings in separate parts of the paper. Each is defined where it is introduced; the recurring cases are listed below.
Superconformal invariance of the minimal interaction
The bilinear ansatz
Let v = ( α α ˙ ) v=(\alpha\dot{\alpha}) v = ( α α ˙ ) and w = ( β β ˙ ) w=(\beta\dot{\beta}) w = ( β β ˙ ) . The operator Π \Pi Π symmetrizes undotted and dotted index pairs separately, each with unit weight. Define
X v w = Π [ ( ∂ v ∂ w W ) W ‾ + W ∂ v ∂ w W ‾ ] , Y v w = Π ( ∂ v W ) ( ∂ w W ‾ ) , Z v w = Π ∑ i [ ( ∂ v D β i W ) ( D ‾ β ˙ i W ‾ ) − ( D β i W ) ∂ v ( D ‾ β ˙ i W ‾ ) ] , T v w = Π ∑ i , j ( D α i D β j W ) ( D ‾ α ˙ i D ‾ β ˙ j W ‾ ) . (9) \begin{aligned}
X_{vw}&=\Pi\left[(\partial_{v}\partial_{w}\mathcal{W})\overline{\mathcal{W}}+\mathcal{W}\partial_{v}\partial_{w}\overline{\mathcal{W}}\right],\\
Y_{vw}&=\Pi(\partial_{v}\mathcal{W})(\partial_{w}\overline{\mathcal{W}}),\\
Z_{vw}&=\Pi\sum_{i}\left[(\partial_{v}D^{i}_{\beta}\mathcal{W})(\overline{D}_{\dot{\beta}i}\overline{\mathcal{W}})-(D^{i}_{\beta}\mathcal{W})\partial_{v}(\overline{D}_{\dot{\beta}i}\overline{\mathcal{W}})\right],\\
T_{vw}&=\Pi\sum_{i,j}(D^{i}_{\alpha}D^{j}_{\beta}\mathcal{W})(\overline{D}_{\dot{\alpha}i}\overline{D}_{\dot{\beta}j}\overline{\mathcal{W}}). \tag*{(9)}
\end{aligned} X v w Y v w Z v w T v w = Π [ ( ∂ v ∂ w W ) W + W ∂ v ∂ w W ] , = Π ( ∂ v W ) ( ∂ w W ) , = Π i ∑ [ ( ∂ v D β i W ) ( D β ˙ i W ) − ( D β i W ) ∂ v ( D β ˙ i W ) ] , = Π i , j ∑ ( D α i D β j W ) ( D α ˙ i D β ˙ j W ) . (9)
The symbol T v w T_{vw} T v w in this basis is not the physical stress tensor. All four structures have dimension four and zero U ( 1 ) U(1) U ( 1 ) charge; X , Y , T X,Y,T X , Y , T and i Z iZ i Z are real under graded conjugation with the displayed spinor-index ordering. The same bilinear sector occurs in the higher-spin currents of [47 ] ; see § 8.5 . Here the coefficients are fixed by superconformal invariance off shell, without using a conservation law or identifying the tensor with the stress tensor obtained by metric variation.
For the superconformal variation, contract the tensor indices with commuting spinors y α , y ˉ α ˙ y^{\alpha},\bar{y}^{\dot{\alpha}} y α , y ˉ α ˙ and use
p y = y α y ˉ α ˙ ∂ α α ˙ , D i ( y ) = y α D α i , D ‾ i ( y ˉ ) = y ˉ α ˙ D ‾ α ˙ i , F i = D i ( y ) W , G i = D ‾ i ( y ˉ ) W ‾ . (10) \begin{aligned}
p_{y}&=y^{\alpha}\bar{y}^{\dot{\alpha}}\partial_{\alpha\dot{\alpha}},&
D^{i}(y)&=y^{\alpha}D^{i}_{\alpha},&
\overline{D}_{i}(\bar{y})&=\bar{y}^{\dot{\alpha}}\overline{D}_{\dot{\alpha}i},\\
F^{i}&=D^{i}(y)\mathcal{W},&
G_{i}&=\overline{D}_{i}(\bar{y})\overline{\mathcal{W}}. \tag*{(10)}
\end{aligned} p y F i = y α y ˉ α ˙ ∂ α α ˙ , = D i ( y ) W , D i ( y ) G i = y α D α i , = D i ( y ˉ ) W . D i ( y ˉ ) = y ˉ α ˙ D α ˙ i , (10)
In the following polarized calculation we write D i ( y ) D^{i}(y) D i ( y ) and D ‾ i ( y ˉ ) \overline{D}_{i}(\bar{y}) D i ( y ˉ ) as D i D^{i} D i and D ‾ i \overline{D}_{i} D i . Products such as D i D j D^{i}D^{j} D i D j contain two polarized first-order derivatives, not the contracted square of (4). The polarization spinors are unrelated to the real variable y y y used later in the truncation to the N = 1 \mathcal{N}=1 N = 1 vector multiplet.
Constrained superconformal parameters
The real superconformal Killing vector ξ \xi ξ preserves chirality and satisfies [26 ]
[ ξ , D α i ] = ω α β D β i − σ ˉ D α i − Λ i j D α i , D ˉ α ˙ i σ = D ˉ α ˙ i ω γ β = 0 , [\xi,D^i_{\alpha}]=\omega_{\alpha}^{\ \beta}D^i_{\beta}-\bar{\sigma}D^i_{\alpha}-\Lambda^j_{\ i}D^i_{\alpha},\qquad\bar{D}_{\dot{\alpha}i}\sigma=\bar{D}_{\dot{\alpha}i}\omega^{\beta}_{\ \gamma}=0, [ ξ , D α i ] = ω α β D β i − σ ˉ D α i − Λ i j D α i , D ˉ α ˙ i σ = D ˉ α ˙ i ω γ β = 0 ,
D α k Λ i j = − 2 ( δ j k D α i − 1 2 δ i j D α k ) σ . (11) D^k_{\alpha}\Lambda^j_{\ i}=-2\left(\delta^k_jD^i_{\alpha}-\frac{1}{2}\delta^j_iD^k_{\alpha}\right)\sigma. \tag*{(11)} D α k Λ i j = − 2 ( δ j k D α i − 2 1 δ i j D α k ) σ . (11)
Reality, the conjugate identities and the traceless S U ( 2 ) SU(2) S U ( 2 ) condition complete the constraints. These parameter functions are determined by the finite set of rigid S U ( 2 , 2 ∣ 2 ) SU(2,2|2) S U ( 2 , 2∣2 ) parameters; they are not arbitrary local transformation parameters. The field strength transforms as
δ s c W = − ξ W − 2 σ W , δ s c W ‾ = − ξ W ‾ − 2 σ ˉ W ‾ . (12) \delta_{\mathrm{sc}}W=-\xi W-2\sigma W,\qquad\delta_{\mathrm{sc}}\overline{W}=-\xi\overline{W}-2\bar{\sigma}\overline{W}. \tag*{(12)} δ sc W = − ξ W − 2 σ W , δ sc W = − ξ W − 2 σ ˉ W . (12)
It preserves the reduced-chiral representation.
For the measure conventions of (7) , constant normalization factors do not alter the transformation laws of the measures:
L ξ ( d 4 x d 8 θ ) = 0 , L ξ ( d 4 x d 4 θ ) = 4 σ d 4 x d 4 θ . (13) \mathcal{L}_{\xi}(d^4x\,d^8\theta)=0,\qquad\mathcal{L}_{\xi}(d^4x\,d^4\theta)=4\sigma\,d^4x\,d^4\theta. \tag*{(13)} L ξ ( d 4 x d 8 θ ) = 0 , L ξ ( d 4 x d 4 θ ) = 4 σ d 4 x d 4 θ . (13)
The full superspace measure is invariant because the contributions of the four even and eight odd coordinates to the superdivergence of ξ \xi ξ cancel. The transformation of the chiral measure makes (8) invariant. Integration by parts is assumed to have no spacetime boundary term. The S U ( 2 ) R SU(2)_R S U ( 2 ) R transformations act on the internal derivative indices. Every internal index in the four bilinears is contracted, so the tensor is an S U ( 2 ) R SU(2)_R S U ( 2 ) R singlet. The remaining test is the cancellation of the inhomogeneous terms produced when derivatives act on the constrained superconformal parameters.
Special-supersymmetry calculation
We denote by A S A_S A S the inhomogeneous part of a special-supersymmetry variation: the remainder after subtracting the transport term, the Lorentz and R R R -symmetry rotations, and the homogeneous scaling terms. For pure special supersymmetry put
χ i = D i σ , χ ˉ i = D ˉ i σ ˉ . (14) \chi^i=D^i\sigma,\qquad\bar{\chi}_i=\bar{D}_i\bar{\sigma}. \tag*{(14)} χ i = D i σ , χ ˉ i = D ˉ i σ ˉ . (14)
They are constant odd parameters in the polarized calculation. The Killing identities give D i ω y β = − χ i y β D^i\omega_y^{\ \beta}=-\chi^i y^\beta D i ω y β = − χ i y β ; the derivatives of the S U ( 2 ) R SU(2)_R S U ( 2 ) R parameter are fixed by (11) . Differentiating (12) , including these parameter derivatives, gives
A S ( F i ) = − 2 χ i W , A S ( G i ) = − 2 χ ˉ i W ‾ , A S ( p y W ) = − i χ ˉ i F i , A S ( p y W ‾ ) = − i χ i G i , A S ( p y F i ) = − 4 χ i p y W − i χ ˉ j D j D i W , A S ( p y G i ) = − 4 χ ˉ i p y W ‾ − i χ j D ˉ j D ˉ i W ‾ , A S ( D i D j W ) = 4 ( χ j F i − χ i F j ) , A S ( D ˉ i D ˉ j W ‾ ) = 4 ( χ ˉ j G i − χ ˉ i G j ) . (15) \begin{aligned}
A_S(F^i)&=-2\chi^iW, & A_S(G_i)&=-2\bar{\chi}_i\overline{W},\\
A_S(p_yW)&=-i\bar{\chi}_iF^i, & A_S(p_y\overline{W})&=-i\chi^iG_i,\\
A_S(p_yF^i)&=-4\chi^ip_yW-i\bar{\chi}_jD^jD^iW, & A_S(p_yG_i)&=-4\bar{\chi}_ip_y\overline{W}-i\chi^j\bar{D}_j\bar{D}_i\overline{W},\\
A_S(D^iD^jW)&=4(\chi^jF^i-\chi^iF^j), & A_S(\bar{D}_i\bar{D}_j\overline{W})&=4(\bar{\chi}_jG_i-\bar{\chi}_iG_j). \tag*{(15)}
\end{aligned} A S ( F i ) A S ( p y W ) A S ( p y F i ) A S ( D i D j W ) = − 2 χ i W , = − i χ ˉ i F i , = − 4 χ i p y W − i χ ˉ j D j D i W , = 4 ( χ j F i − χ i F j ) , A S ( G i ) A S ( p y W ) A S ( p y G i ) A S ( D ˉ i D ˉ j W ) = − 2 χ ˉ i W , = − i χ i G i , = − 4 χ ˉ i p y W − i χ j D ˉ j D ˉ i W , = 4 ( χ ˉ j G i − χ ˉ i G j ) . (15)
The remaining variations organize into three structures:
U S = χ i W p y G i + χ ˉ i ( p y F i ) W ‾ , V S = χ i ( p y W ) G i + χ ˉ i F i p y W ‾ , M S = χ i F j ( D ˉ i D ˉ j W ‾ ) + χ ˉ i ( D i D j W ) G j . (16) \begin{aligned}
U_S&=\chi^iWp_yG_i+\bar{\chi}_i(p_yF^i)\overline{W},\\
V_S&=\chi^i(p_yW)G_i+\bar{\chi}_iF^ip_y\overline{W},\\
M_S&=\chi^iF^j(\bar{D}_i\bar{D}_j\overline{W})+\bar{\chi}_i(D^iD^jW)G_j. \tag*{(16)}
\end{aligned} U S V S M S = χ i W p y G i + χ ˉ i ( p y F i ) W , = χ i ( p y W ) G i + χ ˉ i F i p y W , = χ i F j ( D ˉ i D ˉ j W ) + χ ˉ i ( D i D j W ) G j . (16)
Applying the graded product rule gives
A S ( X ) = − 2 i U S , A S ( Y ) = − i V S , A S ( Z ) = 2 U S − 4 V S − i M S , A S ( T ) = − 8 M S . (17) A_S(X)=-2iU_S,\qquad A_S(Y)=-iV_S,\qquad A_S(Z)=2U_S-4V_S-iM_S,\qquad A_S(T)=-8M_S. \tag*{(17)} A S ( X ) = − 2 i U S , A S ( Y ) = − i V S , A S ( Z ) = 2 U S − 4 V S − i M S , A S ( T ) = − 8 M S . (17)
Appendix C derives the terms containing derivatives of the transformation parameters in (15) and gives the product expansions leading to every entry of (17) .
The primary tensor and the variation of the action
For K ^ = c X X + c Y Y + c Z Z + c T T \widehat{K}=c_X X+c_Y Y+c_Z Z+c_T T K = c X X + c Y Y + c Z Z + c T T , cancellation of (17) requires
− 2 i c X + 2 c Z = 0 , − i c Y − 4 c Z = 0 , − i c Z − 8 c T = 0. (18) -2ic_X+2c_Z=0,\qquad-ic_Y-4c_Z=0,\qquad-ic_Z-8c_T=0. \tag*{(18)} − 2 i c X + 2 c Z = 0 , − i c Y − 4 c Z = 0 , − i c Z − 8 c T = 0. (18)
Requiring the tensor to be primary fixes the solution up to scale within this four-term basis. Normalizing c T = − 4 c_T=-4 c T = − 4 gives
K ^ = − 32 X + 128 Y − 32 i Z − 4 T . (19) \widehat{K}=-32X+128Y-32iZ-4T. \tag*{(19)} K = − 32 X + 128 Y − 32 i Z − 4 T . (19)
The sum of its inhomogeneous terms is 64 i U S − 128 i V S − 64 i U S + 128 i V S − 32 M S + 32 M S = 0 64iU_S-128iV_S-64iU_S+128iV_S-32M_S+32M_S=0 64 i U S − 128 i V S − 64 i U S + 128 i V S − 32 M S + 32 M S = 0 . Translations, Q Q Q supersymmetry and constant Lorentz and R R R transformations are already respected by construction. Invariance under S S S and S ‾ \overline{S} S at the origin, together with the superconformal algebra, implies invariance under special conformal transformations. Thus the full rigid law is
δ s c K ^ v w = − ξ K ^ v w + L v u K ^ u w + L w u K ^ v u − 4 ( σ + σ ‾ ) K ^ v w , (20) \delta_{\mathrm{sc}}\widehat{K}_{vw}=-\xi\widehat{K}_{vw}+L_v{}^u\widehat{K}_{uw}+L_w{}^u\widehat{K}_{vu}-4(\sigma+\overline{\sigma})\widehat{K}_{vw}, \tag*{(20)} δ sc K v w = − ξ K v w + L v u K u w + L w u K v u − 4 ( σ + σ ) K v w , (20)
where L α α ˙ β β ˙ = ω α β δ α ˙ β ˙ + δ α β ω ‾ α ˙ β ˙ L_{\alpha\dot{\alpha}}{}^{\beta\dot{\beta}}=\omega_\alpha{}^\beta\delta_{\dot{\alpha}}{}^{\dot{\beta}}+\delta_\alpha{}^\beta\overline{\omega}_{\dot{\alpha}}{}^{\dot{\beta}} L α α ˙ β β ˙ = ω α β δ α ˙ β ˙ + δ α β ω α ˙ β ˙ .
On the domain where the body of K ^ ⋅ K ^ \widehat{K}\cdot\widehat{K} K ⋅ K is nonzero, and with a fixed branch of the square root, define
Q = K ^ ⋅ K ^ , T 1 = κ ∫ d 4 x d 8 θ W 2 W ‾ 2 Q . (21) Q=\sqrt{\widehat{K}\cdot\widehat{K}},\qquad T_1=\kappa\int d^4x\,d^8\theta\,\frac{\mathcal{W}^2\overline{\mathcal{W}}^2}{Q}. \tag*{(21)} Q = K ⋅ K , T 1 = κ ∫ d 4 x d 8 θ Q W 2 W 2 . (21)
Here K ^ ⋅ K ^ = K ^ v w K ^ v w \widehat{K}\cdot\widehat{K}=\widehat{K}_{vw}\widehat{K}^{vw} K ⋅ K = K v w K v w denotes contraction of the tensor indices. The coefficient κ \kappa κ is fixed by requiring the photon sector to reproduce the ModMax Lagrangian with the positive branch of the square root. On the constant electromagnetic backgrounds of Appendix § 11 , with field-strength parameters f , g f,g f , g defined in (121) , we obtain
Q 0 = 16 f g , [ W 2 W ‾ 2 ] θ 4 θ ‾ 4 = 4 f 2 g 2 , κ = 2. (22) Q_0=16fg,\qquad[\mathcal{W}^2\overline{\mathcal{W}}^2]_{\theta^4\overline{\theta}^4}=4f^2g^2,\qquad\kappa=2. \tag*{(22)} Q 0 = 16 f g , [ W 2 W 2 ] θ 4 θ 4 = 4 f 2 g 2 , κ = 2. (22)
Thus T 1 T_1 T 1 has constant-field density f g / 2 fg/2 f g /2 in these conventions; κ \kappa κ is not an independent coupling. Inverse powers of a superfield require its lowest commuting component to be nonzero; square roots additionally require a choice of branch. Accordingly, the interaction is defined where body ( K ^ ⋅ K ^ ) ≠ 0 \operatorname{body}(\widehat{K}\cdot\widehat{K})\ne0 body ( K ⋅ K ) = 0 . On the physical backgrounds considered here,
r = x 2 + p 2 > 0 , body ( K ^ ⋅ K ^ ) = ( 16 f g ) 2 > 0 , (23) r=\sqrt{x^2+p^2}>0,\qquad\operatorname{body}(\widehat{K}\cdot\widehat{K})=(16fg)^2>0, \tag*{(23)} r = x 2 + p 2 > 0 , body ( K ⋅ K ) = ( 16 f g ) 2 > 0 , (23)
where g = f ∗ g=f^* g = f ∗ . We choose the positive square root on these backgrounds. For varying physical configurations, we restrict attention to regions where this branch remains real and the denominators remain nonzero. For independent complex chiral and antichiral arguments, we use its local analytic continuation. Any additional denominators introduced below must likewise remain nonzero.
The scalar expectation value may vanish. However, constant configurations with F μ ν = 0 F_{\mu\nu}=0 F μν = 0 and vanishing fermions and auxiliaries have K ^ = 0 \widehat{K}=0 K = 0 , irrespective of the constant scalar value, and are excluded. The coupling expansion is therefore defined on the domain specified above, rather than as a weak-field expansion about the zero-field vacuum.
Numerator and denominator have the same homogeneous variation. Consequently
δ s c ( W 2 W ‾ 2 Q ) = − ξ ( W 2 W ‾ 2 Q ) , δ s c T 1 = 0. (24) \delta_{\mathrm{sc}}\left(\frac{\mathcal{W}^2\overline{\mathcal{W}}^2}{Q}\right)=-\xi\left(\frac{\mathcal{W}^2\overline{\mathcal{W}}^2}{Q}\right),\qquad\delta_{\mathrm{sc}}T_1=0. \tag*{(24)} δ sc ( Q W 2 W 2 ) = − ξ ( Q W 2 W 2 ) , δ sc T 1 = 0. (24)
More generally, let S K S_K S K denote the functional (21) with K ^ \widehat K K replaced by a denominator tensor K v w K_{vw} K v w of the same weights. If K v w K_{vw} K v w has inhomogeneous variation A v w A_{vw} A v w , then
δ S K = − κ J F [ W 2 W ‾ 2 ( K ⋅ K ) 3 / 2 K v w A v w ] . (25) \delta S_K=-\kappa\mathcal{J}_F\left[\frac{\mathcal{W}^2\overline{\mathcal{W}}^2}{(K\cdot K)^{3/2}}K^{vw}A_{vw}\right]. \tag*{(25)} δ S K = − κ J F [ ( K ⋅ K ) 3/2 W 2 W 2 K v w A v w ] . (25)
For (19), A = 0 A=0 A = 0 off shell.
Chiral functional derivative and phase invariance
Let L W χ = K ^ [ χ , W ‾ ] \mathcal{L}_{\mathcal{W}}\chi=\widehat K[\chi,\overline{\mathcal{W}}] L W χ = K [ χ , W ] be the linearization in the first argument. Define its graded formal adjoint by full superspace integration by parts. The chiral functional derivative of T 1 T_1 T 1 is the local expression
s : = δ T 1 δ W = κ ( D ‾ 1 ) 2 ( D ‾ 2 ) 2 [ 2 W W ‾ 2 Q − L W † ( W 2 W ‾ 2 K ^ Q 3 ) ] . (26) s:=\frac{\delta T_1}{\delta\mathcal{W}}=\kappa(\overline D_1)^2(\overline D_2)^2\left[\frac{2\mathcal{W}\overline{\mathcal{W}}^2}{Q}-\mathcal{L}_{\mathcal{W}}^\dagger\left(\frac{\mathcal{W}^2\overline{\mathcal{W}}^2\widehat K}{Q^3}\right)\right]. \tag*{(26)} s := δ W δ T 1 = κ ( D 1 ) 2 ( D 2 ) 2 [ Q 2 W W 2 − L W † ( Q 3 W 2 W 2 K ) ] . (26)
The adjoint is a finite differential operator determined by (9); it is not an inverse operator. Each recursive coefficient below is therefore a specified local functional on the same regular domain.
Regarded as a functional of unconstrained chiral superfields, T 1 T_1 T 1 is invariant under the rigid U ( 1 ) U(1) U ( 1 ) phase rotation:
T 1 [ e i α W , e − i α W ‾ ] = T 1 [ W , W ‾ ] , Im J C [ W s ] = 0. (27) T_1[e^{i\alpha}\mathcal{W},e^{-i\alpha}\overline{\mathcal{W}}]=T_1[\mathcal{W},\overline{\mathcal{W}}],\qquad\operatorname{Im}\mathcal{J}_C[\mathcal{W}s]=0. \tag*{(27)} T 1 [ e i α W , e − i α W ] = T 1 [ W , W ] , Im J C [ W s ] = 0. (27)
This is the first-order duality condition. The higher-order conditions determine the completion constructed in the next section.
Canonical self-dual completion to all orders
The functional equation and the normalization
The N = 2 \mathcal{N}=2 N = 2 self-duality equation [2 , 11 ] is
i M = 4 δ S δ W , D [ S ] : = Im J C [ W 2 + M 2 ] = 0. (28) i\mathcal{M}=4\frac{\delta S}{\delta\mathcal{W}},\qquad D[S]:=\operatorname{Im}\mathcal{J}_C[\mathcal{W}^2+\mathcal{M}^2]=0. \tag*{(28)} i M = 4 δ W δ S , D [ S ] := Im J C [ W 2 + M 2 ] = 0. (28)
It expresses covariance of the equations of motion and Bianchi identities under δ W = λ M \delta\mathcal{W}=\lambda\mathcal{M} δ W = λ M , δ M = − λ W \delta\mathcal{M}=-\lambda\mathcal{W} δ M = − λ W . It is not the off-shell invariance of S S S under a linear phase rotation of W \mathcal{W} W . The two-term expression S h y p = cosh γ S 0 + sinh γ T 1 S_{\mathrm{hyp}}=\cosh\gamma\,S_0+\sinh\gamma\,T_1 S hyp = cosh γ S 0 + sinh γ T 1 does not satisfy this equation. With S = 4 ( T 1 ) W \mathcal{S}=4(T_1)_{\mathcal{W}} S = 4 ( T 1 ) W , phase invariance of T 1 T_1 T 1 gives
D [ S h y p ] = − sinh 2 γ Im J C [ W 2 + S 2 ] . (29) D[S_{\mathrm{hyp}}]=-\sinh^2\gamma\,\operatorname{Im}\mathcal{J}_C[\mathcal{W}^2+\mathcal{S}^2]. \tag*{(29)} D [ S hyp ] = − sinh 2 γ Im J C [ W 2 + S 2 ] . (29)
To exhibit the obstruction, take constant chiral fields with only quadratic Grassmann components, W = W f + W y \mathcal{W}=\mathcal{W}_f+\mathcal{W}_y W = W f + W y , in the Lorentz and auxiliary-triplet representations. Their chiral pairing is orthogonal. Per unit spacetime volume put u 0 = ⟨ W f 2 ⟩ C u_0=\langle\mathcal{W}_f^2\rangle_C u 0 = ⟨ W f 2 ⟩ C and v 0 = ⟨ W y 2 ⟩ C v_0=\langle\mathcal{W}_y^2\rangle_C v 0 = ⟨ W y 2 ⟩ C . The denominator depends only on u 0 u ‾ 0 u_0\overline{u}_0 u 0 u 0 , and the photon normalization (22) fixes
L 1 = ∣ u 0 + v 0 ∣ 2 4 u 0 u ‾ 0 , S = 8 ( L 1 , u 0 W f + L 1 , v 0 W y ) . (30) L_1=\frac{|u_0+v_0|^2}{4\sqrt{u_0\overline{u}_0}},\qquad\mathcal{S}=8(L_{1,u_0}\mathcal{W}_f+L_{1,v_0}\mathcal{W}_y). \tag*{(30)} L 1 = 4 u 0 u 0 ∣ u 0 + v 0 ∣ 2 , S = 8 ( L 1 , u 0 W f + L 1 , v 0 W y ) . (30)
Imposing the auxiliary-field reality condition v 0 = v ‾ 0 v_0=\overline v_0 v 0 = v 0 gives Im ⟨ W 2 + S 2 ⟩ C = Im u 0 [ 1 − ∣ u 0 + v 0 ∣ 4 / ∣ u 0 ∣ 4 ] \operatorname{Im}\langle\mathcal{W}^2+\mathcal{S}^2\rangle_C=\operatorname{Im}u_0[1-|u_0+v_0|^4/|u_0|^4] Im ⟨ W 2 + S 2 ⟩ C = Im u 0 [ 1 − ∣ u 0 + v 0 ∣ 4 /∣ u 0 ∣ 4 ] . For u 0 = i ρ u_0=i\rho u 0 = i ρ , v 0 = ϵ ρ v_0=\epsilon\rho v 0 = ϵ ρ , with ρ > 0 \rho>0 ρ > 0 and real ϵ ≠ 0 \epsilon\ne0 ϵ = 0 with the sign required by that reality condition,
D [ S h y p ] / Vol = sinh 2 γ ρ ( 2 ϵ 2 + ϵ 4 ) ≠ 0. (31) D[S_{\mathrm{hyp}}]/\operatorname{Vol}=\sinh^2\gamma\,\rho(2\epsilon^2+\epsilon^4)\ne0. \tag*{(31)} D [ S hyp ] / Vol = sinh 2 γ ρ ( 2 ϵ 2 + ϵ 4 ) = 0. (31)
The failure already at γ 2 \gamma^2 γ 2 motivates the recursive completion. For S = S 0 + I S=S_0+I S = S 0 + I and g = I W g=I_W g = I W the equation reads
W 2 + M 2 = − 8 W g − 16 g 2 , δ 0 I = 4 Im J ~ C [ g 2 ] , δ 0 I = − 2 Im J ~ C [ W W ] . (32) \mathcal{W}^2+\mathcal{M}^2=-8\mathcal{W}g-16g^2,\qquad\delta_0 I=4\operatorname{Im}\widetilde{\mathcal{J}}_C[g^2],\qquad\delta_0 I=-2\operatorname{Im}\widetilde{\mathcal{J}}_C[\mathcal{W}\mathcal{W}]. \tag*{(32)} W 2 + M 2 = − 8 W g − 16 g 2 , δ 0 I = 4 Im J C [ g 2 ] , δ 0 I = − 2 Im J C [ WW ] . (32)
The phase rotation acts on unconstrained chiral superfields, and the same extension is used in all subsequent pairings. Different extensions give different results for functionals that depend on divergences of the field strength [21 ] .
Three U ( 1 ) U(1) U ( 1 ) actions must be distinguished. Electric–magnetic duality acts on the equations of motion and Bianchi identities through δ W = λ M \delta\mathcal{W}=\lambda\mathcal{M} δ W = λ M , δ M = − λ W \delta\mathcal{M}=-\lambda\mathcal{W} δ M = − λ W . The rigid phase rotation of unconstrained chiral superfields, W ↦ e i α W \mathcal{W}\mapsto e^{i\alpha}\mathcal{W} W ↦ e i α W , W ‾ ↦ e − i α W ‾ \overline{\mathcal{W}}\mapsto e^{-i\alpha}\overline{\mathcal{W}} W ↦ e − i α W , defines the phase-invariant functionals and the phase average Π 0 \Pi_0 Π 0 used below. For the free action i M = W i\mathcal{M}=\mathcal{W} i M = W , so this phase rotation agrees with a duality rotation with parameter λ = − α \lambda=-\alpha λ = − α ; they differ in the interacting theory. Finally, the R R R -symmetry U ( 1 ) R ⊂ S U ( 2 , 2 ∣ 2 ) U(1)_R\subset SU(2,2|2) U ( 1 ) R ⊂ S U ( 2 , 2∣2 ) also rotates the Grassmann coordinates and is used in Section 8.2 to exclude scalar–photon mixing. In what follows, phase invariance refers to the rigid phase rotation of the chiral arguments, and duality invariance refers to electric–magnetic duality.
Let δ 0 W = i W \delta_0\mathcal{W}=i\mathcal{W} δ 0 W = i W and δ 0 W ‾ = − i W ‾ \delta_0\overline{\mathcal{W}}=-i\overline{\mathcal{W}} δ 0 W = − i W . Define the phase average
Π 0 F : = 1 2 π ∫ 0 2 π d α F [ e i α W , e − i α W ‾ ] , (33) \Pi_0F:=\frac{1}{2\pi}\int_0^{2\pi}d\alpha\,F[e^{i\alpha}\mathcal{W},e^{-i\alpha}\overline{\mathcal{W}}], \tag*{(33)} Π 0 F := 2 π 1 ∫ 0 2 π d α F [ e i α W , e − i α W ] , (33)
which projects F F F onto its phase-invariant (charge-zero) part. For the canonical completion we impose the normalization
T 1 = f , Π 0 f = f , Π 0 T n = 0 ( n ≥ 2 ) . T_1=f,\qquad\Pi_0f=f,\qquad\Pi_0T_n=0\quad(n\ge2). T 1 = f , Π 0 f = f , Π 0 T n = 0 ( n ≥ 2 ) .
In this section f f f denotes an arbitrary first-order interaction, with T 1 = f T_1=f T 1 = f and I t = ∑ n ≥ 1 t n T n I_t=\sum_{n\ge1}t^nT_n I t = ∑ n ≥ 1 t n T n . The canonical completion of T 1 T_1 T 1 is obtained by choosing f f f to be (21); the modified family is specified in Section 5. Equalities of functionals are understood modulo superspace integration by parts. Each coefficient is local; the number of derivatives need not be uniformly bounded as n n n grows. The power series is formal in t t t , not an expansion about a vanishing field strength. The nonanalytic field dependence of the first-order interaction is permitted on its chosen regular domain.
Put g n = ( T n ) W g_n=(T_n)_W g n = ( T n ) W and P ( F , G ) = J ~ C [ F W G W ] \mathcal{P}(F,G)=\widetilde{\mathcal{J}}_C[F_WG_W] P ( F , G ) = J C [ F W G W ] . Substitution of the series in (32) gives
i M = 4 ( S t ) W = W + 4 ( I t ) W , D [ S t ] : = Im J ~ C [ W 2 + M 2 ] = 0 , i\mathcal{M}=4(S_t)_W=\mathcal{W}+4(I_t)_W,\qquad\mathcal{D}[S_t]:=\operatorname{Im}\widetilde{\mathcal{J}}_C[\mathcal{W}^2+\mathcal{M}^2]=0, i M = 4 ( S t ) W = W + 4 ( I t ) W , D [ S t ] := Im J C [ W 2 + M 2 ] = 0 ,
δ 0 T n = 4 Im R n , R n = ∑ k = 1 n − 1 P ( T k , T n − k ) , Im Π 0 R n = 0. (34) \delta_0T_n=4\operatorname{Im}R_n,\qquad R_n=\sum_{k=1}^{n-1}\mathcal{P}(T_k,T_{n-k}),\qquad\operatorname{Im}\Pi_0R_n=0. \tag*{(34)} δ 0 T n = 4 Im R n , R n = k = 1 ∑ n − 1 P ( T k , T n − k ) , Im Π 0 R n = 0. (34)
The last equality is an integrability condition: it must hold before δ 0 \delta_0 δ 0 can be inverted on the U ( 1 ) U(1) U ( 1 ) -charged part.
The main result of this section is that the self-duality equation can be solved order by order with this normalization. For every real, local, phase-invariant first-order interaction f f f on a regular domain preserved by the phase rotation, there is a unique formal interaction I t I_t I t satisfying this normalization and (34) . If f f f is invariant under rigid S U ( 2 , 2 ∣ 2 ) SU(2,2|2) S U ( 2 , 2∣2 ) transformations of unconstrained chiral superfields, so is every T n T_n T n . We establish existence and uniqueness in Sections 4.2 and 4.3, and superconformal invariance in Section 4.4.
Auxiliary-superfield form of the self-duality equation
For any real I = O ( t ) I=O(t) I = O ( t ) , introduce the auxiliary chiral superfield and the transformed functional
U = W + 2 g , g = I W , H [ U , U ‾ ] = I [ W , W ‾ ] + J ~ C [ g 2 ] + c.c. (35) \mathcal{U}=\mathcal{W}+2g,\qquad g=I_W,\qquad\mathcal{H}[\mathcal{U},\overline{\mathcal{U}}]=I[\mathcal{W},\overline{\mathcal{W}}]+\widetilde{\mathcal{J}}_C[g^2]+\text{c.c.} \tag*{(35)} U = W + 2 g , g = I W , H [ U , U ] = I [ W , W ] + J C [ g 2 ] + c.c. (35)
Here only the square term is conjugated in the last expression. The inverse W = W [ U , U ˉ ] \mathcal{W}=\mathcal{W}[\mathcal{U},\bar{\mathcal{U}}] W = W [ U , U ˉ ] is substituted on its right-hand side. It exists formally because U = W + O ( t ) \mathcal{U}=\mathcal{W}+O(t) U = W + O ( t ) . The composite U \mathcal{U} U is an auxiliary chiral variable, with no reduced-chiral constraint. The transformation acts on unconstrained chiral superfields and leaves the physical vector multiplet unchanged.
Vary (35), keeping all functional integrations and adjoints. Since these chiral functional derivatives are Grassmann even,
δ H = I ~ C [ g δ W + 2 g δ g ] + c . c . = I ~ C [ g δ U ] + c . c . , H U = g , \delta\mathcal{H}=\widetilde{\mathcal{I}}_{\mathrm{C}}[g\,\delta\mathcal{W}+2g\,\delta g]+\mathrm{c.c.}=\widetilde{\mathcal{I}}_{\mathrm{C}}[g\,\delta\mathcal{U}]+\mathrm{c.c.},\qquad\mathcal{H}_{\mathcal{U}}=g, δ H = I C [ g δ W + 2 g δ g ] + c.c. = I C [ g δ U ] + c.c. , H U = g ,
δ 0 U H = − 2 Im I ~ C [ U H U ] = − 2 Im I ~ C [ W g + 2 g 2 ] = 1 4 D [ S 0 + I ] . (36) \delta_{0}^{\mathcal{U}}\mathcal{H}=-2\operatorname{Im}\widetilde{\mathcal{I}}_{\mathrm{C}}[\mathcal{U}\mathcal{H}_{\mathcal{U}}]=-2\operatorname{Im}\widetilde{\mathcal{I}}_{\mathrm{C}}[\mathcal{W}g+2g^{2}]=\frac{1}{4}\mathcal{D}[S_{0}+I]. \tag*{(36)} δ 0 U H = − 2 Im I C [ U H U ] = − 2 Im I C [ W g + 2 g 2 ] = 4 1 D [ S 0 + I ] . (36)
The phase operator on the second line acts linearly on U \mathcal{U} U , not on W \mathcal{W} W at fixed I I I . The final equality follows directly from W 2 + M 2 = − 8 W g − 16 g 2 \mathcal{W}^{2}+\mathcal{M}^{2}=-8\mathcal{W}g-16g^{2} W 2 + M 2 = − 8 W g − 16 g 2 . Thus self-duality is equivalent to ordinary phase invariance of H \mathcal{H} H . This is the N = 2 N=2 N = 2 auxiliary-superfield formulation of [10 , 16 ] in our normalization. The phase-invariance condition is given explicitly in Eqs. (3.11)–(3.14) of [10 ] . The corresponding bosonic construction uses the auxiliary-bispinor formulation of [13 , 14 , 15 ] . Here U \mathcal{U} U is determined by (35); the physical action depends on W \mathcal{W} W and W ˉ \bar{\mathcal{W}} W ˉ alone.
Conversely, prescribe a real phase-invariant H = O ( t ) \mathcal{H}=O(t) H = O ( t ) and solve
U = W + 2 H U [ U , U ˉ ] , I [ W , W ˉ ] = H [ U , U ˉ ] − ( I ~ C [ ( H U ) 2 ] + c . c . ) . (37) \mathcal{U}=\mathcal{W}+2\mathcal{H}_{\mathcal{U}}[\mathcal{U},\bar{\mathcal{U}}],\qquad I[\mathcal{W},\bar{\mathcal{W}}]=\mathcal{H}[\mathcal{U},\bar{\mathcal{U}}]-\left(\widetilde{\mathcal{I}}_{\mathrm{C}}[(\mathcal{H}_{\mathcal{U}})^{2}]+\mathrm{c.c.}\right). \tag*{(37)} U = W + 2 H U [ U , U ˉ ] , I [ W , W ˉ ] = H [ U , U ˉ ] − ( I C [( H U ) 2 ] + c.c. ) . (37)
One determines the t n t^{n} t n coefficient of U \mathcal{U} U from the previous coefficients because H U = O ( t ) \mathcal{H}_{\mathcal{U}}=O(t) H U = O ( t ) . No inverse spacetime or spinor differential operator is used. Differentiating the second formula yields I W = H U I_{\mathcal{W}}=\mathcal{H}_{\mathcal{U}} I W = H U : its variation is I ~ C [ H U ( δ U − 2 δ H U ) ] + c . c . \widetilde{\mathcal{I}}_{\mathrm{C}}[\mathcal{H}_{\mathcal{U}}(\delta\mathcal{U}-2\delta\mathcal{H}_{\mathcal{U}})]+\mathrm{c.c.} I C [ H U ( δ U − 2 δ H U )] + c.c. and the bracket equals δ W \delta\mathcal{W} δ W . Therefore the two transformations are inverse to each other and the generated I I I is self-dual.
For the canonical normalization we need a further property. At fixed lower coefficients, adding a functional J J J at order n n n gives
H ⟼ H + t n J ⟹ I ⟼ I + t n J [ W , W ˉ ] + O ( t n + 1 ) . (38) \mathcal{H}\longmapsto\mathcal{H}+t^{n}J\quad\Longrightarrow\quad I\longmapsto I+t^{n}J[\mathcal{W},\bar{\mathcal{W}}]+O(t^{n+1}). \tag*{(38)} H ⟼ H + t n J ⟹ I ⟼ I + t n J [ W , W ˉ ] + O ( t n + 1 ) . (38)
Indeed, the change of U \mathcal{U} U starts at order n n n . Its contribution to H \mathcal{H} H is multiplied by H U = O ( t ) \mathcal{H}_{\mathcal{U}}=O(t) H U = O ( t ) , and every cross term in the square of the functional derivative also starts at order n + 1 n+1 n + 1 . The direct contribution is t n J ( U ) = t n J ( W ) + O ( t n + 1 ) t^{n}J(\mathcal{U})=t^{n}J(\mathcal{W})+O(t^{n+1}) t n J ( U ) = t n J ( W ) + O ( t n + 1 ) . Thus the normalization can be imposed order by order without changing the lower-order action.
Inductive construction and uniqueness
Start with H 1 = f H_{1}=f H 1 = f in H = ∑ j ≥ 1 t j H j \mathcal{H}=\sum_{j\geq1}t^{j}H_{j} H = ∑ j ≥ 1 t j H j . The action obtained from (37) has T 1 = f T_{1}=f T 1 = f . Suppose real phase-invariant H 1 , … , H n − 1 H_{1},\ldots,H_{n-1} H 1 , … , H n − 1 have been chosen so the action has the required canonical coefficients below order n n n . Form the finite polynomial H < n = ∑ j = 1 n − 1 t j H j \mathcal{H}^{<n}=\sum_{j=1}^{n-1}t^{j}H_{j} H < n = ∑ j = 1 n − 1 t j H j and let I t r i a l I_{\mathrm{trial}} I trial be its formal interaction obtained from (37). Define
K n = [ t n ] I t r i a l , H n = − Π 0 K n , T n = K n − Π 0 K n . (39) K_{n}=[t^{n}]I^{\mathrm{trial}},\qquad H_{n}=-\Pi_{0}K_{n},\qquad\boxed{T_{n}=K_{n}-\Pi_{0}K_{n}}. \tag*{(39)} K n = [ t n ] I trial , H n = − Π 0 K n , T n = K n − Π 0 K n . (39)
In H n H_{n} H n the arguments of the functional − Π 0 K n -\Pi_{0}K_{n} − Π 0 K n are renamed U , U ˉ \mathcal{U},\bar{\mathcal{U}} U , U ˉ ; the physical coefficient T n T_{n} T n retains W , W ˉ \mathcal{W},\bar{\mathcal{W}} W , W ˉ . The polynomial H < n \mathcal{H}^{<n} H < n is phase invariant, so the entire trial interaction is self-dual, including its order-n n n coefficient K n K_{n} K n . The H n H_{n} H n is real and phase-invariant. (38) leaves all lower coefficients unchanged and gives precisely the boxed T n T_{n} T n . It obeys Π 0 T n = 0 \Pi_{0}T_{n}=0 Π 0 T n = 0 , completing the induction.
This also establishes the integrability condition without a separate high-order Hessian expansion. At order n n n the exact trial action satisfies δ 0 K n = 4 Im R n \delta_{0}K_{n}=4\operatorname{Im}R_{n} δ 0 K n = 4 Im R n , where R n R_{n} R n is constructed only from the already fixed T 1 , … , T n − 1 T_{1},\ldots,T_{n-1} T 1 , … , T n − 1 . Therefore
0 = Π 0 δ 0 K n = 4 Im Π 0 R n for every n ≥ 2. (40) 0=\Pi_{0}\delta_{0}K_{n}=4\operatorname{Im}\Pi_{0}R_{n}\qquad\text{for every }n\geq2. \tag*{(40)} 0 = Π 0 δ 0 K n = 4 Im Π 0 R n for every n ≥ 2. (40)
The trial action exists by the explicit inversion (37) , independently of the phase recursion we are proving solvable.
Uniqueness follows in the same way. The lowest-order difference Δ T n \Delta T_n Δ T n of two canonical solutions obeys δ 0 Δ T n = 0 \delta_0\Delta T_n=0 δ 0 Δ T n = 0 because their lower coefficients, and hence their source R n R_n R n , agree. Consequently Δ T n = Π 0 Δ T n = 0 \Delta T_n=\Pi_0\Delta T_n=0 Δ T n = Π 0 Δ T n = 0 . Thus the induction constructs the same solution as phase decomposition, T n = ∑ m ≠ 0 ( − 2 R n , m / m + c . c . ) T_n=\sum_{m\ne0}(-2R_{n,m}/m+\mathrm{c.c.}) T n = ∑ m = 0 ( − 2 R n , m / m + c.c. ) with δ 0 R n , m = i m R n , m \delta_0R_{n,m}=imR_{n,m} δ 0 R n , m = im R n , m .
First coefficients and superconformal covariance
As a normalization check, let s = f W s=f_{\mathcal W} s = f W , B = P ( f , f ) = J C [ s 2 ] B=\mathcal P(f,f)=\mathcal J_C[s^2] B = P ( f , f ) = J C [ s 2 ] and C = P ( f , B ) C=\mathcal P(f,B) C = P ( f , B ) . The first coefficients are
T 1 = f , T 2 = B + B ˉ , T 3 = C + C ˉ . (41) T_1=f,\qquad T_2=B+\bar{B},\qquad T_3=C+\bar{C}. \tag*{(41)} T 1 = f , T 2 = B + B ˉ , T 3 = C + C ˉ . (41)
Choosing only H = t f \mathcal H=tf H = t f in the inverse transformation instead gives
I [ H = t f ] = t f + t 2 ( B + B ˉ ) + t 3 [ C + C ˉ + 2 P ( f , B ˉ ) ] + O ( t 4 ) . (42) \mathcal I[\mathcal H=tf]=tf+t^2(B+\bar{B})+t^3[C+\bar{C}+2\mathcal P(f,\bar{B})]+O(t^4). \tag*{(42)} I [ H = t f ] = t f + t 2 ( B + B ˉ ) + t 3 [ C + C ˉ + 2 P ( f , B ˉ )] + O ( t 4 ) . (42)
Symmetry of the mixed functional Hessian of real f f f makes P ( f , B ˉ ) \mathcal P(f,\bar{B}) P ( f , B ˉ ) real; its U ( 1 ) U(1) U ( 1 ) charge is zero. The induction therefore selects H 3 = − 2 P ( f , B ˉ ) H_3=-2\mathcal P(f,\bar{B}) H 3 = − 2 P ( f , B ˉ ) , with arguments renamed to U \mathcal U U , and removes exactly this term from the physical coefficient. A phase-invariant auxiliary interaction generates a self-dual action, while this additional subtraction selects the canonical completion.
If a functional F F F is superconformally invariant on unconstrained chiral superfields, its chiral functional derivative transforms in the same primary representation as W \mathcal W W . Its first variation and the chiral measure then imply
δ s c F W = − ξ F W − 2 σ F W , δ s c P ( F , G ) = 0 ( F , G invariant ) , \delta_{\mathrm{sc}}F_{\mathcal W}=-\xi F_{\mathcal W}-2\sigma F_{\mathcal W},\qquad\delta_{\mathrm{sc}}\mathcal P(F,G)=0\quad(F,G\ \text{invariant}), δ sc F W = − ξ F W − 2 σ F W , δ sc P ( F , G ) = 0 ( F , G invariant ) ,
δ s c T n = 0 ( n ≥ 1 ) , δ s c S t = 0 , D [ S t ] = 0 as formal series . (43) \delta_{\mathrm{sc}}T_n=0\ (n\ge1),\qquad\delta_{\mathrm{sc}}S_t=0,\qquad\mathcal D[S_t]=0\quad\text{as formal series}. \tag*{(43)} δ sc T n = 0 ( n ≥ 1 ) , δ sc S t = 0 , D [ S t ] = 0 as formal series . (43)
Phase decomposition commutes with the superconformal action, so inverting δ 0 \delta_0 δ 0 on terms with nonzero phase charge preserves invariance at each order. Equivalently, (37) is covariant and (39) preserves invariance. Finite differentiation and contraction also preserve locality at each order. The physical Bianchi identity is not imposed on a functional derivative or on U \mathcal U U to obtain these results.
Additional primary interactions and the phase-invariant terms
Primary tensors with unequal weights
The superconformal variation fixes the tensor within the four-term ansatz, but it does not exhaust the superconformal scalar interactions. Let A = W p A=\mathcal W^p A = W p and B ˉ = W ˉ q \bar{B}=\bar{\mathcal W}^{q} B ˉ = W ˉ q . In the polarized notation of Section 3, set
X L = ( p y 2 A ) B ˉ , X R = A p y 2 B ˉ , Y = ( p y A ) ( p y B ˉ ) , X_L=(p_y^2A)\bar{B},\qquad X_R=Ap_y^2\bar{B},\qquad Y=(p_yA)(p_y\bar{B}), X L = ( p y 2 A ) B ˉ , X R = A p y 2 B ˉ , Y = ( p y A ) ( p y B ˉ ) ,
Z L = ( p y D i A ) ( D ˉ i B ˉ ) , Z R = ( D i A ) ( p y D ˉ i B ˉ ) , T = ( D i D j A ) ( D ˉ i D ˉ j B ˉ ) . (44) Z_L=(p_yD^iA)(\bar{D}_i\bar{B}),\qquad Z_R=(D^iA)(p_y\bar{D}_i\bar{B}),\qquad T=(D^iD^jA)(\bar{D}_i\bar{D}_j\bar{B}). \tag*{(44)} Z L = ( p y D i A ) ( D ˉ i B ˉ ) , Z R = ( D i A ) ( p y D ˉ i B ˉ ) , T = ( D i D j A ) ( D ˉ i D ˉ j B ˉ ) . (44)
To restore the Lorentz indices, one symmetrizes the two undotted and the two dotted Lorentz indices separately. The resulting primary tensor is
K p q = 4 q ( q + 1 ) X L + 4 p ( p + 1 ) X R − 8 ( p + 1 ) ( q + 1 ) Y + 4 i ( q + 1 ) Z L − 4 i ( p + 1 ) Z R + T . (45) K_{pq}=4q(q+1)X_L+4p(p+1)X_R-8(p+1)(q+1)Y+4i(q+1)Z_L-4i(p+1)Z_R+T. \tag*{(45)} K pq = 4 q ( q + 1 ) X L + 4 p ( p + 1 ) X R − 8 ( p + 1 ) ( q + 1 ) Y + 4 i ( q + 1 ) Z L − 4 i ( p + 1 ) Z R + T . (45)
It has dimension p + q + 2 p+q+2 p + q + 2 and U ( 1 ) U(1) U ( 1 ) charge p − q p-q p − q , with K ^ = − 4 K 11 \widehat{K}=-4K_{11} K = − 4 K 11 . The six inhomogeneous special-supersymmetry coefficients, in the ordered basis of (44), are the rows of
( 0 − 2 i 0 0 − 2 p 0 − 2 i 0 0 2 q 0 0 0 0 − i − 2 ( p + 1 ) 0 0 0 0 − i 0 2 ( q + 1 ) 0 0 0 0 0 i − 4 ( p + 1 ) 0 0 0 − i 0 − 4 ( q + 1 ) ) (46) \begin{pmatrix}
0 & -2i & 0 & 0 & -2p & 0 \\
-2i & 0 & 0 & 2q & 0 & 0 \\
0 & 0 & -i & -2(p+1) & 0 & 0 \\
0 & 0 & -i & 0 & 2(q+1) & 0 \\
0 & 0 & 0 & 0 & i & -4(p+1) \\
0 & 0 & 0 & -i & 0 & -4(q+1) \tag*{(46)}
\end{pmatrix} 0 − 2 i 0 0 0 0 − 2 i 0 0 0 0 0 0 0 − i − i 0 0 0 2 q − 2 ( p + 1 ) 0 0 − i − 2 p 0 0 2 ( q + 1 ) i 0 0 0 0 0 − 4 ( p + 1 ) − 4 ( q + 1 ) (46)
They multiply the ordered structures
χ i A p y D ˉ i B ˉ , χ ˉ i ( p y D i A ) B ˉ , χ i ( p y A ) D ˉ i B ˉ , \chi^i A_{py}\bar D_i\bar B,\qquad\bar\chi_i(p_yD^iA)\bar B,\qquad\chi^i(p_yA)\bar D_i\bar B, χ i A p y D ˉ i B ˉ , χ ˉ i ( p y D i A ) B ˉ , χ i ( p y A ) D ˉ i B ˉ ,
χ ˉ i D i A p y B ˉ , χ i D j A D ˉ i D ˉ j B ˉ , χ ˉ i D i D j A D ˉ j B ˉ . \bar\chi_iD^iA p_y\bar B,\qquad\chi^iD^jA\bar D_i\bar D_j\bar B,\qquad\bar\chi_iD^iD^jA\bar D_j\bar B. χ ˉ i D i A p y B ˉ , χ i D j A D ˉ i D ˉ j B ˉ , χ ˉ i D i D j A D ˉ j B ˉ .
Multiplying (46) by the coefficient vector in (45) gives zero in every row. These are the weighted versions of the calculation of the parameter derivatives in Appendix C. They use only the chirality of W p W^p W p . Together with the homogeneous transformation law and the superconformal algebra, this establishes primarity.
The first-order interaction
For p = q = 3 p=q=3 p = q = 3 the tensor K 33 K_{33} K 33 has dimension eight and zero U ( 1 ) U(1) U ( 1 ) charge. Contracting it with K ^ \widehat K K gives a real, phase-invariant density whose superspace integrand has dimension zero,
D 33 = J F [ K ^ b a ( K 33 ) a b Q 3 ] , (47) D_{33}=\mathcal{J}_{F}\left[\frac{\widehat K^{a}_{\ b}(K_{33})^{b}_{\ a}}{Q^3}\right], \tag*{(47)} D 33 = J F [ Q 3 K b a ( K 33 ) a b ] , (47)
where the contracted indices are Lorentz vector indices. The interaction used in (2) is
f m o d = T 1 − 152 405 D 33 , (48) f_{\mathrm{mod}}=T_1-\frac{152}{405}D_{33}, \tag*{(48)} f mod = T 1 − 405 152 D 33 , (48)
with the normalization T 1 = 2 J F [ W 2 W ‾ 2 / Q ] T_1=2\mathcal{J}_{F}[W^2\overline{W}^2/Q] T 1 = 2 J F [ W 2 W 2 / Q ] . Both terms are real, phase invariant and superconformal. The interaction therefore satisfies the first-order self-duality equation for independent chiral arguments. The interaction D 33 D_{33} D 33 has zero N = 1 \mathcal{N}=1 N = 1 vector restriction, as shown in Section 6.3.
At vanishing scalar expectation value, the scalar Hessians of T 1 T_1 T 1 and D 33 D_{33} D 33 are − 18 Σ -18\Sigma − 18Σ and − 405 8 Σ -\frac{405}{8}\Sigma − 8 405 Σ in the conventions of Appendix B, and all their other independent-chiral scalar entries vanish. The coefficient − 152 / 405 -152/405 − 152/405 is fixed by requiring the total coefficient Σ \Sigma Σ : the additional term contributes 19 Σ 19\Sigma 19Σ . It changes the first-order scalar correction from superluminal to subluminal without changing the N = 1 \mathcal{N}=1 N = 1 reduction.
The results of Sections 7 and 8 use only the following properties of the first-order interaction f f f : it is real, phase invariant, superconformal and even under the reflection of Section 6.4; its vector restriction is Res 1 f = Res 1 T 1 \operatorname{Res}_1 f=\operatorname{Res}_1 T_1 Res 1 f = Res 1 T 1 ; and on the backgrounds of Section 8 its independent-chiral scalar Hessian is Σ ϕ ϕ ˉ \Sigma\phi\bar\phi Σ ϕ ϕ ˉ , with all other entries zero. Every interaction with these properties gives the same zero-vev results, with Δ = f − T 1 \Delta=f-T_1 Δ = f − T 1 in (50). The choice (48) uses only one additional primary density; the zero-vev conditions do not select it uniquely. Agreement of the quadratic actions on these backgrounds does not imply equivalence of the full interacting theories. The higher-order phase-invariant interactions are specified below.
Completion with prescribed phase-invariant terms
The construction of § 4 extends to any prescribed sequence of real, local, phase-invariant functionals C n C_n C n , provided they are defined on the same regular domain. In the induction of § 4.3 , replace (39) by
H n = C n − Π 0 K n , F n = K n − Π 0 K n + C n . (49) H_n = C_n - \Pi_0 K_n,\qquad F_n = K_n - \Pi_0 K_n + C_n. \tag*{(49)} H n = C n − Π 0 K n , F n = K n − Π 0 K n + C n . (49)
The lower-order coefficients remain fixed, and the phase-invariant auxiliary functional still generates a self-dual action. At the first order where two solutions with the same prescribed C n C_n C n differ, their difference is phase invariant and has zero phase average, so uniqueness holds with these data fixed. If every C n C_n C n is superconformal, the same covariance argument applies at each order. This establishes formal existence, uniqueness and superconformality without a convergence assumption.
Let F n F_n F n denote the canonical coefficients generated from f m o d f_{\mathrm{mod}} f mod and T n T_n T n those generated from the minimal interaction. The higher-order prescription uses the additional first-order term,
Δ = f m o d − T 1 = − 152 405 D 33 . (50) \Delta= f_{\mathrm{mod}} - T_1 = -\frac{152}{405}D_{33}. \tag*{(50)} Δ = f mod − T 1 = − 405 152 D 33 . (50)
It has zero vector restriction. At zero scalar expectation value its independent-chiral scalar Hessian is 19 Σ ϕ ϕ ˉ 19\Sigma_{\phi\bar{\phi}} 19 Σ ϕ ϕ ˉ in the component normalization of Appendix § 11 , and its other entries vanish, including those involving the highest scalar components. These properties are used in the recursive argument in Appendix § 11 .
The prescription in (2) is
C n = [ t n ] d ( t ) Δ , d ( t ) = − t 3 76 + t 5 152 − 5 t 7 1216 + O ( t 9 ) , C_n = [t^n]d(t)\Delta,\qquad d(t)=-\frac{t^3}{76}+\frac{t^5}{152}-\frac{5t^7}{1216}+O(t^9), C n = [ t n ] d ( t ) Δ , d ( t ) = − 76 t 3 + 152 t 5 − 1216 5 t 7 + O ( t 9 ) ,
F ~ 1 = f m o d , F ~ 2 = F 2 = P ( f m o d , f m o d ) + c . c . , F ~ 3 = F 3 − Δ 76 . (51) \widetilde{F}_1=f_{\mathrm{mod}},\qquad\widetilde{F}_2=F_2=\mathcal{P}(f_{\mathrm{mod}},f_{\mathrm{mod}})+\mathrm{c.c.},\qquad\widetilde{F}_3=F_3-\frac{\Delta}{76}. \tag*{(51)} F 1 = f mod , F 2 = F 2 = P ( f mod , f mod ) + c.c. , F 3 = F 3 − 76 Δ . (51)
The prescribed terms are real, phase invariant, superconformal and reflection even, with zero vector restriction. The closure argument in § 8.2 therefore fixes the quadratic scalar series for the chosen d ( t ) d(t) d ( t ) . All charged higher-order terms must be computed using (49) . Neither the zero-vev quadratic action nor its restricted flow selects a unique full superspace interaction.
Truncation to the N = 1 \mathcal{N}=1 N = 1 vector multiplet
Conventions and nilpotency identities
We use D 2 = 1 2 D α D α D^2=\frac{1}{2}D^\alpha D_\alpha D 2 = 2 1 D α D α and normalize the Maxwell action as
D 2 = 1 2 D α D α , W α = − 1 2 D ˉ 2 D α V , S M = 1 4 ∫ d 4 x d 2 θ W 2 + c . c . , V = V ˉ . (52) D^2=\frac{1}{2}D^\alpha D_\alpha,\qquad W_\alpha=-\frac{1}{2}\bar{D}^2D_\alpha V,\qquad S_{\mathrm{M}}=\frac{1}{4}\int d^4x\,d^2\theta\,W^2+\mathrm{c.c.},\qquad V=\bar{V}. \tag*{(52)} D 2 = 2 1 D α D α , W α = − 2 1 D ˉ 2 D α V , S M = 4 1 ∫ d 4 x d 2 θ W 2 + c.c. , V = V ˉ . (52)
The relative full/chiral measure convention used for the N = 1 \mathcal{N}=1 N = 1 calculation is
∫ d 4 x d 2 θ d 2 θ ˉ F = − 1 2 ∫ d 4 x d 2 θ D ˉ 2 F , δ S M = ∫ d 4 x d 2 θ d 2 θ ˉ E δ V , E = D α W α . (53) \int d^4x\,d^2\theta\,d^2\bar{\theta}\,F=-\frac{1}{2}\int d^4x\,d^2\theta\,\bar{D}^2F,\qquad\delta S_{\mathrm{M}}=\int d^4x\,d^2\theta\,d^2\bar{\theta}\,E\,\delta V,\qquad E=D_\alpha W^\alpha. \tag*{(53)} ∫ d 4 x d 2 θ d 2 θ ˉ F = − 2 1 ∫ d 4 x d 2 θ D ˉ 2 F , δ S M = ∫ d 4 x d 2 θ d 2 θ ˉ E δ V , E = D α W α . (53)
The conjugate chiral formula is understood. Define
A = D 2 W 2 , A ˉ = D ˉ 2 W ˉ 2 , N = W 2 W ˉ 2 , Ω = 4 N A A ˉ , A=D^2W^2,\qquad\bar{A}=\bar{D}^2\bar{W}^2,\qquad N=W^2\bar{W}^2,\qquad\Omega=\frac{4N}{A\bar{A}}, A = D 2 W 2 , A ˉ = D ˉ 2 W ˉ 2 , N = W 2 W ˉ 2 , Ω = A A ˉ 4 N ,
u = D α W α , v = D α W ˉ α , a = − 1 4 ( A + u 2 2 ) , b = 1 8 u 2 . (54) u=D_\alpha W^\alpha,\qquad v=D_\alpha\bar{W}^{\alpha},\qquad a=-\frac{1}{4}\left(A+\frac{u^2}{2}\right),\qquad b=\frac{1}{8}u^2. \tag*{(54)} u = D α W α , v = D α W ˉ α , a = − 4 1 ( A + 2 u 2 ) , b = 8 1 u 2 . (54)
and use the conjugate definitions of a ˉ , b ˉ \bar{a},\bar{b} a ˉ , b ˉ . The two chiral arguments are independent when differentiating functionals. Only after forming the recursive pairings do we impose the physical Bianchi identity u = v = E u=v=E u = v = E . This distinction is relevant to divergence-dependent actions, as discussed in [21 ] .
In these conventions, the N = 1 \mathcal{N}=1 N = 1 ModMax action of [18 eq. (4.19)] reads
S B ( γ ) = cosh γ S M + sinh γ R B , R B = ∫ d 4 x d 2 θ d 2 θ ˉ N A A ˉ . (55) S_{\mathrm{B}}(\gamma)=\cosh\gamma\,S_{\mathrm{M}}+\sinh\gamma\,R_{\mathrm{B}},\qquad R_{\mathrm{B}}=\int d^{4}x\,d^{2}\theta\,d^{2}\bar{\theta}\,\frac{\mathcal{N}}{\sqrt{A\bar{A}}}. \tag*{(55)} S B ( γ ) = cosh γ S M + sinh γ R B , R B = ∫ d 4 x d 2 θ d 2 θ ˉ A A ˉ N . (55)
This is an off-shell superspace action. We shall show that the truncated N = 2 \mathcal{N}=2 N = 2 action equals (55) after a local redefinition of the real prepotential, without eliminating the auxiliary field.
Let I \mathcal{I} I denote expressions containing at least one of the undifferentiated spinors. Their nilpotency gives
I = ( W 1 , W 2 , W ˉ 1 ˙ , W ˉ 2 ˙ ) , N I = 0 , W 2 W α = 0 , W ˉ 2 W ˉ α ˙ = 0. (56) \mathcal{I}=(W_{1},W_{2},\bar{W}_{\dot{1}},\bar{W}_{\dot{2}}),\qquad\mathcal{N}\mathcal{I}=0,\qquad W^{2}W_{\alpha}=0,\qquad\bar{W}^{2}\bar{W}_{\dot{\alpha}}=0. \tag*{(56)} I = ( W 1 , W 2 , W ˉ 1 ˙ , W ˉ 2 ˙ ) , N I = 0 , W 2 W α = 0 , W ˉ 2 W ˉ α ˙ = 0. (56)
Differentiation can remove the undifferentiated spinor factor. These vanishing products therefore cannot be used before the derivatives act. Products with N \mathcal{N} N are therefore always formed after the indicated differentiations, or after a chiral factor W 2 W^{2} W 2 has been taken through a barred derivative. In particular, ∂ μ W α \partial_{\mu}W_{\alpha} ∂ μ W α , D α W β D_{\alpha}W_{\beta} D α W β and D 2 W α D^{2}W_{\alpha} D 2 W α are not set to zero.
Reduction of the minimal interaction
Choose θ = θ 1 \theta=\theta_{1} θ = θ 1 and η = θ 2 \eta=\theta_{2} η = θ 2 . Switching off the N = 1 \mathcal{N}=1 N = 1 chiral multiplet sets Φ : = W ( x c h , θ , η = 0 ) \Phi:=\mathcal{W}(x_{\mathrm{ch}},\theta,\eta=0) Φ := W ( x ch , θ , η = 0 ) and all its components to zero. The remaining term in the η \eta η expansion gives
W = 2 i η α W α , Res 1 S 0 = S M . (57) \mathcal{W}=2i\eta^{\alpha}W_{\alpha},\qquad\operatorname{Res}_{1}S_{0}=S_{\mathrm{M}}. \tag*{(57)} W = 2 i η α W α , Res 1 S 0 = S M . (57)
The spacetime arguments in this formula are chiral coordinates. The full extra chiral multiplet is removed; the retained vector multiplet remains off shell. Let ρ η \rho_{\eta} ρ η denote evaluation at η = η ˉ = 0 \eta=\bar{\eta}=0 η = η ˉ = 0 after this embedding; it does not set the surviving θ , θ ˉ \theta,\bar{\theta} θ , θ ˉ to zero.
The minimal interaction is
T 1 = 2 ∫ d 4 x d 8 θ W 2 W ˉ 2 Q , Q = K ^ ⋅ K ^ , K ^ = − 32 X + 128 Y − 32 i Z − 4 T . (58) T_{1}=2\int d^{4}x\,d^{8}\theta\,\frac{\mathcal{W}^{2}\bar{\mathcal{W}}^{2}}{Q},\qquad Q=\sqrt{\widehat{K}\cdot\widehat{K}},\qquad\widehat{K}=-32X+128Y-32iZ-4T. \tag*{(58)} T 1 = 2 ∫ d 4 x d 8 θ Q W 2 W ˉ 2 , Q = K ⋅ K , K = − 32 X + 128 Y − 32 i Z − 4 T . (58)
The coefficient 2 is the photon normalization for the full measure in (7). The bilinears are those in (9), written directly with D α i D^{i}_{\alpha} D α i and D ˉ α ˙ i \bar{D}_{\dot{\alpha}i} D ˉ α ˙ i . The ordinary S U ( 2 ) R SU(2)_{R} S U ( 2 ) R invariance makes different choices of the retained supersymmetry equivalent.
On the regular domain, with the branch fixed by the photon normalization, the minimal interaction reduces to
Res 1 T 1 = R r = ∫ d 4 x d 2 θ d 2 θ ˉ N R 1 , R 1 = ( A + u 2 / 2 ) ( A ˉ + v 2 / 2 ) . (59) \operatorname{Res}_{1}T_{1}=R_{r}=\int d^{4}x\,d^{2}\theta\,d^{2}\bar{\theta}\,\frac{\mathcal{N}}{R_{1}},\qquad R_{1}=\sqrt{(A+u^{2}/2)(\bar{A}+v^{2}/2)}. \tag*{(59)} Res 1 T 1 = R r = ∫ d 4 x d 2 θ d 2 θ ˉ R 1 N , R 1 = ( A + u 2 /2 ) ( A ˉ + v 2 /2 ) . (59)
for arbitrary independent chiral spinors and before the Bianchi identity is imposed. To see this, note that the numerator is of top degree in η \eta η and η ˉ \bar{\eta} η ˉ :
W 2 W ˉ 2 = 16 η 1 η 2 η ˉ 1 ˙ η ˉ 2 ˙ N . (60) \mathcal{W}^{2}\bar{\mathcal{W}}^{2}=16\eta^{1}\eta^{2}\bar{\eta}^{\dot{1}}\bar{\eta}^{\dot{2}}\mathcal{N}. \tag*{(60)} W 2 W ˉ 2 = 16 η 1 η 2 η ˉ 1 ˙ η ˉ 2 ˙ N . (60)
The η \eta η -dependent shifts of the chiral coordinates contain an additional η \eta η or η ˉ \bar{\eta} η ˉ and therefore vanish in this product. Consequently the integration over η , η ˉ \eta,\bar{\eta} η , η ˉ involves only ρ η ( Q − 1 ) \rho_{\eta}(Q^{-1}) ρ η ( Q − 1 ) and no η \eta η -derivatives of the denominator. With the Maxwell normalization of the measures, the factor from the η , η ˉ \eta,\bar{\eta} η , η ˉ integration in (60) is 16. This is consistent with the chiral reduction factor 2 in (57), since our full N = 2 \mathcal{N}=2 N = 2 projector is ( D ˉ 1 ) 2 ( D ˉ 2 ) 2 (\bar{D}_{1})^{2}(\bar{D}_{2})^{2} ( D ˉ 1 ) 2 ( D ˉ 2 ) 2 , whereas (53) uses − 1 2 D ˉ 2 -\frac{1}{2}\bar{D}^{2} − 2 1 D ˉ 2 .
Write f α β = D α ( W β ) f_{\alpha\beta}=D_{\alpha}(\mathcal{W}_{\beta}) f α β = D α ( W β ) . The structures X , Y X,Y X , Y contain W \mathcal{W} W or a spacetime derivative of its η \eta η -independent component and hence vanish under ρ η \rho_{\eta} ρ η . The surviving terms of Z Z Z contain W α ∂ W ‾ α ˙ \mathcal{W}_{\alpha}\partial\overline{\mathcal{W}}_{\dot{\alpha}} W α ∂ W α ˙ or ( ∂ W α ) W ‾ α ˙ (\partial\mathcal{W}_{\alpha})\overline{\mathcal{W}}_{\dot{\alpha}} ( ∂ W α ) W α ˙ , and are annihilated by N N N . The mixed double-spinor derivatives in T T T give the outer product of f f f and f ˉ \bar{f} f ˉ . Thus, with the index ordering of (9),
N ρ η ( X ) = N ρ η ( Y ) = N ρ η ( Z ) = 0 , N ρ η ( K ^ α α ˙ β β ˙ ) = 32 N f α β f ˉ α ˙ β ˙ . (61) N\rho_{\eta}(X)=N\rho_{\eta}(Y)=N\rho_{\eta}(Z)=0,\qquad N\rho_{\eta}(\widehat{K}_{\alpha\dot{\alpha}\beta\dot{\beta}})=32Nf_{\alpha\beta}\bar{f}_{\dot{\alpha}\dot{\beta}}. \tag*{(61)} N ρ η ( X ) = N ρ η ( Y ) = N ρ η ( Z ) = 0 , N ρ η ( K α α ˙ β β ˙ ) = 32 N f α β f ˉ α ˙ β ˙ . (61)
An overall conjugation sign in the last relation would cancel in its square. The spinor product rule, followed by multiplication by N N N , gives
N f α β f α β = − N ( A + u 2 / 2 ) , N ρ η ( Q 2 ) = 32 2 N ( A + u 2 / 2 ) ( A ˉ + v 2 / 2 ) . (62) Nf^{\alpha\beta}f_{\alpha\beta}=-N(A+u^{2}/2),\qquad N\rho_{\eta}(Q^{2})=32^{2}N(A+u^{2}/2)(\bar{A}+v^{2}/2). \tag*{(62)} N f α β f α β = − N ( A + u 2 /2 ) , N ρ η ( Q 2 ) = 3 2 2 N ( A + u 2 /2 ) ( A ˉ + v 2 /2 ) . (62)
For example, take m α β = D α W β m_{\alpha\beta}=D_{\alpha}\mathcal{W}_{\beta} m α β = D α W β , with ϵ 12 = 1 \epsilon_{12}=1 ϵ 12 = 1 . Then u = m 21 − m 12 u=m_{21}-m_{12} u = m 21 − m 12 and N A = − 2 N det m NA=-2N\det m N A = − 2 N det m ; decomposing m m m into its symmetric and antisymmetric parts gives the first identity. The omitted terms in A A A contain W α D 2 W α \mathcal{W}^{\alpha}D^{2}\mathcal{W}_{\alpha} W α D 2 W α and are annihilated by N N N .
All quantities in the last identity have already been differentiated. Since their bodies are invertible, analytic functional calculus on the chosen branch gives
N ρ η ( Q − 1 ) = N 32 R 1 . (63) N\rho_{\eta}(Q^{-1})=\frac{N}{32R_{1}}. \tag*{(63)} N ρ η ( Q − 1 ) = 32 R 1 N . (63)
Equations (58) , (60) and (63) give the overall factor 2 ⋅ 16 / 32 = 1 2\cdot16/32=1 2 ⋅ 16/32 = 1 , and hence (59) .
Truncation of the additional interaction
For Φ = 0 \Phi=0 Φ = 0 , one has W = 2 i η α W α \mathcal{W}=2i\eta^{\alpha}\mathcal{W}_{\alpha} W = 2 i η α W α , so both W 3 \mathcal{W}^{3} W 3 and W ‾ 3 \overline{\mathcal{W}}^{3} W 3 , together with their derivatives, vanish identically. The tensor K 33 K_{33} K 33 in (47) is built from A = W 3 A=\mathcal{W}^{3} A = W 3 and B ˉ = W ‾ 3 \bar{B}=\overline{\mathcal{W}}^{3} B ˉ = W 3 through (44) , so D 33 D_{33} D 33 vanishes before integration. Therefore
Res 1 D 33 = 0 , Res 1 Δ = 0 , Res 1 f m o d = Res 1 T 1 . (64) \operatorname{Res}_{1}D_{33}=0,\qquad\operatorname{Res}_{1}\Delta=0,\qquad\operatorname{Res}_{1}f_{\mathrm{mod}}=\operatorname{Res}_{1}T_{1}. \tag*{(64)} Res 1 D 33 = 0 , Res 1 Δ = 0 , Res 1 f mod = Res 1 T 1 . (64)
These are off-shell identities for independent chiral spinor arguments, as required in the self-duality calculation. They allow arbitrary spacetime dependence and do not use the vector Bianchi identity.
Closure of the pairing of chiral functional derivatives
For a differentiable algebraic function L ( a , b , a ˉ , b ˉ ) L(a,b,\bar{a},\bar{b}) L ( a , b , a ˉ , b ˉ ) on the regular domain define
F [ L ] = ∫ d 4 x d 2 θ d 2 θ ˉ Ω L , f L = 4 L A A ˉ , r L α = δ F [ L ] δ W α . (65) \mathcal{F}[L]=\int d^{4}x\,d^{2}\theta\,d^{2}\bar{\theta}\,\Omega L,\qquad f_{L}=\frac{4L}{A\bar{A}},\qquad r_{L\alpha}=\frac{\delta\mathcal{F}[L]}{\delta\mathcal{W}^{\alpha}}. \tag*{(65)} F [ L ] = ∫ d 4 x d 2 θ d 2 θ ˉ Ω L , f L = A A ˉ 4 L , r Lα = δ W α δ F [ L ] . (65)
The arbitrary superfields W , W ‾ \mathcal{W},\overline{\mathcal{W}} W , W in these expressions can carry spacetime dependence and fermionic components. “Algebraic” only means that L L L is an ordinary function of the specified composites, not that their component fields are constant.
For any two such functions L , M L,M L , M , the pairing of their chiral functional derivatives closes on the same class:
1 2 ∫ d 4 x d 2 θ r L α r M α = F [ a L a M + b L b M ] . (66) \frac{1}{2}\int d^{4}x\,d^{2}\theta\,r_{L}^{\alpha}r_{M\alpha}=\mathcal{F}[a_{L}a_{M}+b_{L}b_{M}]. \tag*{(66)} 2 1 ∫ d 4 x d 2 θ r L α r M α = F [ a L a M + b L b M ] . (66)
The two chiral arguments remain independent in this identity. To derive (66) , we vary the full functional and integrate the spinor derivatives by parts. The exact chiral functional derivative of ∫ d 4 x d 2 θ d 2 θ ˉ N f \int d^{4}x\,d^{2}\theta\,d^{2}\bar{\theta}\,Nf ∫ d 4 x d 2 θ d 2 θ ˉ N f is
r α [ f ] = − 1 2 D ˉ 2 { 2 W α [ W ‾ 2 f + D 2 ( N f A ) ] + D α ( N f u ) } . (67) r_{\alpha}[f]=-\frac{1}{2}\bar{D}^{2}\left\{2\mathcal{W}_{\alpha}[\overline{\mathcal{W}}^{2}f+D^{2}(NfA)]+D_{\alpha}(Nfu)\right\}. \tag*{(67)} r α [ f ] = − 2 1 D ˉ 2 { 2 W α [ W 2 f + D 2 ( N f A )] + D α ( N f u ) } . (67)
Expanding the displayed D D D derivatives first, and using D α W ˉ α ˙ = 0 D_{\alpha}\bar{W}_{\dot{\alpha}}=0 D α W ˉ α ˙ = 0 , put it in the form
r α [ f ] = − D ˉ 2 [ W ˉ 2 ( W α P f + 1 2 ( D α W 2 ) Q f ) ] + W 2 X α [ f ] , P f = f + A f A , Q f = f u . (68) r_{\alpha}[f]=-\bar{D}^{2}\left[\bar{W}^{2}\left(W_{\alpha}P_{f}+\frac{1}{2}(D_{\alpha}W^{2})Q_{f}\right)\right]+W^{2}\mathcal{X}_{\alpha}[f],\qquad P_{f}=f+Af_{A},\quad Q_{f}=f_{u}. \tag*{(68)} r α [ f ] = − D ˉ 2 [ W ˉ 2 ( W α P f + 2 1 ( D α W 2 ) Q f ) ] + W 2 X α [ f ] , P f = f + A f A , Q f = f u . (68)
Here X α \mathcal{X}_{\alpha} X α collects the remaining local terms. They include derivatives of f A , f u f_{A}, f_{u} f A , f u . This is an exact factorization: for example W α D β W 2 W_{\alpha}D_{\beta}W^{2} W α D β W 2 is proportional to W 2 D β W α W^{2}D_{\beta}W_{\alpha} W 2 D β W α , and D ˉ \bar{D} D ˉ commutes through the chiral factor W 2 W^{2} W 2 . Thus all those terms retain an undifferentiated factor W 2 W^{2} W 2 even after the barred derivatives act. They do not contribute to the product of two functional derivatives, because the other derivative contains at least one undifferentiated factor W W W ; no term is discarded from an individual functional derivative.
The remaining expression carrying a spinor index is linear in W W W . In the component index basis used for (62) , its coefficient matrix is
C ( P , Q ) = ( P − m 12 Q m 11 Q − m 22 Q P + m 21 Q ) , W α P + 1 2 ( D α W 2 ) Q = C ( P , Q ) α β W β . (69) C(P,Q)=\begin{pmatrix} P-m_{12}Q & m_{11}Q \\ -m_{22}Q & P+m_{21}Q \end{pmatrix},\qquad W_{\alpha}P+\frac{1}{2}(D_{\alpha}W^{2})Q=C(P,Q)_{\alpha}{}^{\beta}W_{\beta}. \tag*{(69)} C ( P , Q ) = ( P − m 12 Q − m 22 Q m 11 Q P + m 21 Q ) , W α P + 2 1 ( D α W 2 ) Q = C ( P , Q ) α β W β . (69)
After multiplying the two functional derivatives, convert one chiral projector to a full integral using (53) . The remaining projector is multiplied by W ˉ 2 \bar{W}^{2} W ˉ 2 , so the identity
W ˉ 2 D ˉ 2 ( W ˉ 2 H ) = W ˉ 2 A ˉ H ˉ (70) \bar{W}^{2}\bar{D}^{2}(\bar{W}^{2}H)=\bar{W}^{2}\bar{A}\bar{H} \tag*{(70)} W ˉ 2 D ˉ 2 ( W ˉ 2 H ) = W ˉ 2 A ˉ H ˉ (70)
applies for any even H H H . All terms where a barred derivative acts on H H H still have an extra undifferentiated W ˉ \bar{W} W ˉ and vanish. The polarized determinant of (69) , together with N det m = − N A / 2 N\det m=-NA/2 N det m = − N A /2 , therefore gives
1 2 ∫ d 4 x d 2 θ r [ f ] α r [ g ] α = − ∫ d 4 x d 2 θ d 2 θ ˉ N ˉ A ˉ [ P f P g + u 2 ( P f Q g + Q f P g ) − A 2 Q f Q g ] . (71) \frac{1}{2}\int d^{4}x\,d^{2}\theta\,r[f]^{\alpha}r[g]_{\alpha}
=-\int d^{4}x\,d^{2}\theta\,d^{2}\bar{\theta}\,\bar{N}\bar{A}\left[P_{f}P_{g}+\frac{u}{2}(P_{f}Q_{g}+Q_{f}P_{g})-\frac{A}{2}Q_{f}Q_{g}\right]. \tag*{(71)} 2 1 ∫ d 4 x d 2 θ r [ f ] α r [ g ] α = − ∫ d 4 x d 2 θ d 2 θ ˉ N ˉ A ˉ [ P f P g + 2 u ( P f Q g + Q f P g ) − 2 A Q f Q g ] . (71)
This identity is applied only after all derivatives have acted, including the mixed-derivative terms in the barred projectors of (68) .
For f = f L f=f_{L} f = f L the ordinary chain rule gives
P f L = − L a A ˉ , Q f L = u ( L b − L a ) A A ˉ . (72) P_{f_{L}}=-\frac{L_{a}}{\bar{A}},\qquad Q_{f_{L}}=\frac{u(L_{b}-L_{a})}{A\bar{A}}. \tag*{(72)} P f L = − A ˉ L a , Q f L = A A ˉ u ( L b − L a ) . (72)
Substituting this and its M M M counterpart into (71) yields
− A ˉ [ P L P M + u 2 ( P L Q M + Q L P M ) − A 2 Q L Q M ] = 4 A A ˉ ( a L a M a + b L b M b ) , (73) -\bar{A}\left[P_{L}P_{M}+\frac{u}{2}(P_{L}Q_{M}+Q_{L}P_{M})-\frac{A}{2}Q_{L}Q_{M}\right]=\frac{4}{A\bar{A}}(aL_{a}M_{a}+bL_{b}M_{b}), \tag*{(73)} − A ˉ [ P L P M + 2 u ( P L Q M + Q L P M ) − 2 A Q L Q M ] = A A ˉ 4 ( a L a M a + b L b M b ) , (73)
which gives (66) . The cancellation holds for arbitrary u u u and v v v .
The full N = 2 \mathcal{N}=2 N = 2 pairing also reduces correctly. The reflection W ( η ) ↦ − W ( − η ) \mathcal{W}(\eta)\mapsto-\mathcal{W}(-\eta) W ( η ) ↦ − W ( − η ) leaves the term linear in η \eta η unchanged and reverses the scalar superfields at orders η 0 \eta^{0} η 0 and η 2 \eta^{2} η 2 . Both the minimal interaction T 1 T_{1} T 1 and f m o d f_{\mathrm{mod}} f mod are even under this reflection: K p q K_{pq} K pq acquires the factor ( − 1 ) p + q (-1)^{p+q} ( − 1 ) p + q together with the reflection of η , η ˉ \eta,\bar{\eta} η , η ˉ , so K ˉ = − 4 K 11 \bar{K}=-4K_{11} K ˉ = − 4 K 11 and K 33 K_{33} K 33 are even. The phase-invariant additions in (51) are even as well. The contractions of functional derivatives and the phase averages defining every coefficient preserve this reflection. At Φ = 0 \Phi=0 Φ = 0 , which is invariant under this reflection, the functional derivatives with respect to Φ \Phi Φ and Φ ˉ \bar{\Phi} Φ ˉ vanish. The chiral measure normalization then gives
Res 1 ( ∫ d 4 x d 4 θ F W G W ) = 1 2 ∫ d 4 x d 2 θ ( Res 1 F ) W α ( Res 1 G ) W α . (74) \operatorname{Res}_{1}\left(\int d^{4}x\,d^{4}\theta\,F_{\mathcal{W}}G_{\mathcal{W}}\right)=\frac{1}{2}\int d^{4}x\,d^{2}\theta\,(\operatorname{Res}_{1}F)_{\mathcal{W}}^{\alpha}(\operatorname{Res}_{1}G)_{\mathcal{W}\alpha}. \tag*{(74)} Res 1 ( ∫ d 4 x d 4 θ F W G W ) = 2 1 ∫ d 4 x d 2 θ ( Res 1 F ) W α ( Res 1 G ) W α . (74)
Consequently (59) and (66) apply inductively to the truncated canonical functionals.
All-order equivalence with the off-shell N = 1 \mathcal{N}=1 N = 1 ModMax action
Component field redefinitions have long been used to relate different forms of supersymmetric nonlinear electrodynamics and Goldstino actions [24 , 25 ] . Here the change of variables is made at the level of a real N = 1 \mathcal{N}=1 N = 1 prepotential, so that chirality and the vector Bianchi identity follow from its definition. The result concerns a formal off-shell equality on the regular domain and does not require a Goldstino interpretation of the retained photino.
Our main result for the N = 1 \mathcal{N}=1 N = 1 truncation is the following. On the regular domain specified below, the canonical completion of (21) admits a real, local, formal redefinition of the N = 1 \mathcal{N}=1 N = 1 prepotential, V o l d = V + O ( t ) V_{\mathrm{old}}=V+O(t) V old = V + O ( t ) , such that
Res 1 S sinh γ [ V o l d ( V ) ] = cosh γ S M [ V ] + sinh γ R B [ V ] (75) \operatorname{Res}_{1} S_{\sinh\gamma}[V_{\mathrm{old}}(V)] = \cosh\gamma\, S_{\mathrm{M}}[V] + \sinh\gamma\, R_{\mathrm{B}}[V] \tag*{(75)} Res 1 S s i n h γ [ V old ( V )] = cosh γ S M [ V ] + sinh γ R B [ V ] (75)
holds to every order in γ \gamma γ , modulo superspace integration by parts and with no equation of motion imposed on the retained N = 1 \mathcal{N}=1 N = 1 vector multiplet. The derivation has three steps. The exact truncation identities close the canonical recursion on algebraic coefficient functions. Their value at vanishing auxiliary invariant follows from uniqueness of the canonical solution. The resulting identity ensures that the prepotential redefinition remains regular at E = 0 E=0 E = 0 .
Truncated recursion at zero auxiliary invariant
With the composites and functional F \mathcal{F} F of § 6 , the truncated first-order interaction has density
Res 1 T 1 = F [ L 1 ] , L 1 = ( a + b ) ( a ˉ + b ˉ ) a a ˉ . (76) \operatorname{Res}_{1} T_{1} = \mathcal{F}[L_{1}], \qquad L_{1} = \frac{(a+b)(\bar{a}+\bar{b})}{\sqrt{a\bar{a}}}. \tag*{(76)} Res 1 T 1 = F [ L 1 ] , L 1 = a a ˉ ( a + b ) ( a ˉ + b ˉ ) . (76)
Let ρ L = L ( a , 0 , a ˉ , 0 ) \rho L=L(a,0,\bar{a},0) ρ L = L ( a , 0 , a ˉ , 0 ) denote evaluation of the coefficient functions at b = b ˉ = 0 b=\bar{b}=0 b = b ˉ = 0 . This is distinct from the evaluation ρ η \rho_{\eta} ρ η at η = η ˉ = 0 \eta=\bar{\eta}=0 η = η ˉ = 0 used in the reduction of T 1 T_{1} T 1 .
Res 1 P ( F , G ) = F [ p ( L , M ) ] , p ( L , M ) = a L a M a + b L b M b , \operatorname{Res}_{1}\mathcal{P}(F,G)=\mathcal{F}[p(L,M)], \qquad p(L,M)=aL_{a}M_{a}+bL_{b}M_{b}, Res 1 P ( F , G ) = F [ p ( L , M )] , p ( L , M ) = a L a M a + b L b M b ,
ρ p ( L , M ) = a ( ρ L ) a ( ρ M ) a , ρ Π 0 = Π 0 ρ , ρ L 1 = a a ˉ . (77) \rho p(L,M)=a(\rho L)_{a}(\rho M)_{a}, \qquad\rho\Pi_{0}=\Pi_{0}\rho,\qquad\rho L_{1}=\sqrt{a\bar{a}}. \tag*{(77)} ρp ( L , M ) = a ( ρ L ) a ( ρM ) a , ρ Π 0 = Π 0 ρ , ρ L 1 = a a ˉ . (77)
Here Res 1 F = F [ L ] \operatorname{Res}_{1}F=\mathcal{F}[L] Res 1 F = F [ L ] , Res 1 G = F [ M ] \operatorname{Res}_{1}G=\mathcal{F}[M] Res 1 G = F [ M ] in the reflection-even class of functionals generated by the first-order interaction. The reflection W ( η ) ↦ − W ( − η ) \mathcal{W}(\eta)\mapsto-\mathcal{W}(-\eta) W ( η ) ↦ − W ( − η ) , which was used to derive the truncated pairing (74), is preserved by every pairing and every phase average. Hence the functional derivatives with respect to the omitted chiral multiplet vanish at every order on Φ = 0 \Phi=0 Φ = 0 . The same holds for the conjugate pairing.
The factors b b b and b ˉ \bar{b} b ˉ ensure that the pairing at b = b ˉ = 0 b=\bar{b}=0 b = b ˉ = 0 depends only on the coefficient functions at those values. The functions and their derivatives are regular there, so no singular terms survive multiplication by b b b or b ˉ \bar{b} b ˉ . This evaluation is made on the coefficient functions; no condition is imposed on the superfields.
Let r a = a a ˉ r_{a}=\sqrt{a\bar{a}} r a = a a ˉ , C t = 1 + t 2 C_{t}=\sqrt{1+t^{2}} C t = 1 + t 2 and k = C t − 1 k=C_{t}-1 k = C t − 1 . At b = b ˉ = 0 b=\bar{b}=0 b = b ˉ = 0 , consider the interaction
J ∂ ( t ; a , a ˉ ) = k 2 ( a + a ˉ ) + t r a , J_{\partial}(t;a,\bar{a})=\frac{k}{2}(a+\bar{a})+tr_{a}, J ∂ ( t ; a , a ˉ ) = 2 k ( a + a ˉ ) + t r a ,
a ( J ∂ ) a 2 − a ˉ ( J ∂ ) a ˉ 2 = k 2 − t 2 4 ( a − a ˉ ) = − k 2 ( a − a ˉ ) . (78) a(J_{\partial})_{a}^{2}-\bar{a}(J_{\partial})_{\bar{a}}^{2}=\frac{k^{2}-t^{2}}{4}(a-\bar{a})=-\frac{k}{2}(a-\bar{a}). \tag*{(78)} a ( J ∂ ) a 2 − a ˉ ( J ∂ ) a ˉ 2 = 4 k 2 − t 2 ( a − a ˉ ) = − 2 k ( a − a ˉ ) . (78)
Since a a a and b b b have U ( 1 ) U(1) U ( 1 ) charge + 2 +2 + 2 , the phase variation of J ∂ J_{\partial} J ∂ is i k ( a − a ˉ ) ik(a-\bar{a}) ik ( a − a ˉ ) . The last line of (78) therefore gives δ 0 J ∂ = 4 Im ( J ∂ , J ∂ ) \delta_{0}J_{\partial}=4\operatorname{Im}(J_{\partial},J_{\partial}) δ 0 J ∂ = 4 Im ( J ∂ , J ∂ ) exactly. Its order-one coefficient is r a r_{a} r a ; all coefficients at higher orders are proportional to a + a ˉ a+\bar{a} a + a ˉ and have no phase-invariant part. The same uniqueness argument applies to these coefficient functions: at the first order where two solutions could differ, their difference has zero phase variation and zero phase average, and hence vanishes. Together with (77) , this identifies J ∂ J_{\partial} J ∂ with the canonical solution evaluated at b = b ˉ = 0 b=\bar{b}=0 b = b ˉ = 0 .
Once the Bianchi identity u = v = E u=v=E u = v = E is imposed, write the truncated density as r G ( t ; c , y ) r\,\mathcal{G}(t;c,y) r G ( t ; c , y ) , with G = c + y + ∑ n ≥ 1 t n ℓ n \mathcal{G}=c+y+\sum_{n\geq1}t^n\ell_n G = c + y + ∑ n ≥ 1 t n ℓ n . The real variables r , c , y r,c,y r , c , y are defined in (80) . The result is
G ( t ; c , 0 ) = c 1 + t 2 + t , (79) \mathcal{G}(t;c,0)=c\sqrt{1+t^2}+t, \tag*{(79)} G ( t ; c , 0 ) = c 1 + t 2 + t , (79)
ℓ 1 ( c , 0 ) = 1 , ℓ 2 j ( c , 0 ) = ( 1 / 2 j ) c , ℓ 2 j + 1 ( c , 0 ) = 0 ( j ≥ 1 ) . \ell_1(c,0)=1,\qquad\ell_{2j}(c,0)=\binom{1/2}{j}c,\qquad\ell_{2j+1}(c,0)=0\quad(j\geq1). ℓ 1 ( c , 0 ) = 1 , ℓ 2 j ( c , 0 ) = ( j 1/2 ) c , ℓ 2 j + 1 ( c , 0 ) = 0 ( j ≥ 1 ) .
This determines the truncated coefficients at y = 0 y=0 y = 0 to every order.
Construction of the prepotential redefinition
Now impose u = v = E u=v=E u = v = E and use a real prepotential. In our conventions
W α = − 1 2 D ˉ 2 D α V , E = D α W α , A = D 2 W 2 , N = W 2 W ˉ 2 , Ω = 4 N A A ˉ , W_\alpha=-\frac{1}{2}\bar{D}^2D_\alpha V,\qquad E=D_\alpha W^\alpha,\qquad A=D^2W^2,\qquad N=W^2\bar{W}^2,\qquad\Omega=\frac{4N}{A\bar{A}}, W α = − 2 1 D ˉ 2 D α V , E = D α W α , A = D 2 W 2 , N = W 2 W ˉ 2 , Ω = A A ˉ 4 N ,
x = − A + A ˉ + E 2 8 , p = i ( A − A ˉ ) 8 , r = x 2 + p 2 , c = x r , y = E 2 8 r . (80) x=-\frac{A+\bar{A}+E^2}{8},\qquad p=\frac{i(A-\bar{A})}{8},\qquad r=\sqrt{x^2+p^2},\qquad c=\frac{x}{r},\qquad y=\frac{E^2}{8r}. \tag*{(80)} x = − 8 A + A ˉ + E 2 , p = 8 i ( A − A ˉ ) , r = x 2 + p 2 , c = r x , y = 8 r E 2 . (80)
These are composites, not new fields. On the domain considered, r r r , A A ˉ A\bar{A} A A ˉ and the relevant square roots have invertible bodies. The branch near y = 0 y=0 y = 0 has ϖ : = 1 + 2 c y + y 2 = 1 \varpi:=\sqrt{1+2cy+y^2}=1 ϖ := 1 + 2 cy + y 2 = 1 there and extends wherever these denominators remain regular. The domain includes E = 0 E=0 E = 0 ; the construction uses no inverse power of E E E . Statements near y = 0 y=0 y = 0 are local analytic statements in the coefficient functions.
Homogeneity, reality, and the pairing closure imply that the truncated canonical coefficients have densities r ℓ n ( c , y ) r\ell_n(c,y) r ℓ n ( c , y ) . More concretely, the first-order interaction is polynomial in a , b , a ˉ , b ˉ a,b,\bar{a},\bar{b} a , b , a ˉ , b ˉ divided by a power of r a r_a r a ; the pairings and phase averages preserve this algebra. Once u = v = E u=v=E u = v = E is imposed, each coefficient is symmetric in a , a ˉ a,\bar{a} a , a ˉ , and the relations a a ˉ = r 2 a\bar{a}=r^2 a a ˉ = r 2 , a + a ˉ = 2 r c a+\bar{a}=2rc a + a ˉ = 2 r c reduce it to a polynomial in c , y c,y c , y at each order. In particular no coefficient develops an inverse power of y y y .
At first order ℓ 1 = 1 + 2 c y + y 2 = ϖ 2 \ell_1=1+2cy+y^2=\varpi^2 ℓ 1 = 1 + 2 cy + y 2 = ϖ 2 ; the recursion based on the chiral pairing gives ℓ 2 = 1 2 c ( 1 + 2 c y + y 2 ) \ell_2=\frac{1}{2}c(1+2cy+y^2) ℓ 2 = 2 1 c ( 1 + 2 cy + y 2 ) . These two coefficients fix the normalization of the matching calculation.
The Maxwell term and the square-root term of the target action can both be written as
S M = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r ( c + y ) , R B = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r ϖ . (81) S_{\mathrm{M}}=\int d^4x\,d^2\theta\,d^2\bar{\theta}\,\Omega r(c+y),\qquad R_{\mathrm{B}}=\int d^4x\,d^2\theta\,d^2\bar{\theta}\,\Omega r\varpi. \tag*{(81)} S M = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r ( c + y ) , R B = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r ϖ . (81)
Indeed r ( c + y ) = − ( A + A ˉ ) / 8 r(c+y)=-(A+\bar{A})/8 r ( c + y ) = − ( A + A ˉ ) /8 . Since A ˉ \bar{A} A ˉ is chiral, D ˉ 2 ( W ˉ 2 / A ˉ ) = 1 \bar{D}^2(\bar{W}^2/\bar{A})=1 D ˉ 2 ( W ˉ 2 / A ˉ ) = 1 ; the projection convention (53) and its conjugate then give the first identity. The second follows from A A ˉ = 16 r 2 ϖ 2 A\bar{A}=16r^2\varpi^2 A A ˉ = 16 r 2 ϖ 2 on the chosen branch.
With S M = 1 4 ∫ d 4 x d 2 θ W 2 + c . c . S_{\mathrm{M}}=\frac{1}{4}\int d^4x\,d^2\theta\,W^2+\mathrm{c.c.} S M = 4 1 ∫ d 4 x d 2 θ W 2 + c.c. and R B = ∫ d 4 x d 2 θ d 2 θ ˉ N / A A ˉ R_{\mathrm{B}}=\int d^4x\,d^2\theta\,d^2\bar{\theta}\,N/\sqrt{A\bar{A}} R B = ∫ d 4 x d 2 θ d 2 θ ˉ N / A A ˉ , the two sides to be matched are
Res 1 S t [ V ] = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r G ( t ; c , y ) , G ( t ; c , y ) = c + y + ∑ n ≥ 1 t n ℓ n ( c , y ) , \operatorname{Res}_1 S_t[V]=\int d^4x\,d^2\theta\,d^2\bar{\theta}\,\Omega r\,\mathcal{G}(t;c,y),\qquad\mathcal{G}(t;c,y)=c+y+\sum_{n\geq1}t^n\ell_n(c,y), Res 1 S t [ V ] = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r G ( t ; c , y ) , G ( t ; c , y ) = c + y + n ≥ 1 ∑ t n ℓ n ( c , y ) ,
S B ( t ) [ V ] = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r B ( t ; c , y ) , B ( t ; c , y ) = 1 + t 2 ( c + y ) + t ϖ . (82) S_{\mathrm{B}}(t)[V]=\int d^4x\,d^2\theta\,d^2\bar{\theta}\,\Omega r\,\mathcal{B}(t;c,y),\qquad\mathcal{B}(t;c,y)=\sqrt{1+t^2}(c+y)+t\varpi. \tag*{(82)} S B ( t ) [ V ] = ∫ d 4 x d 2 θ d 2 θ ˉ Ω r B ( t ; c , y ) , B ( t ; c , y ) = 1 + t 2 ( c + y ) + t ϖ . (82)
The second line is the complete off-shell N = 1 \mathcal{N}=1 N = 1 ModMax action of [18 ] in our conventions, including its fermionic and auxiliary-field terms.
Seek a formal change of variable Y = y + O ( t ) Y=y+O(t) Y = y + O ( t ) such that
G ( t ; c , Y ( t ; c , y ) ) = B ( t ; c , y ) , Y = y + ∑ n ≥ 1 t n q n ( c , y ) . (83) \mathcal{G}(t;c,Y(t;c,y))=\mathcal{B}(t;c,y),\qquad Y=y+\sum_{n\geq1}t^nq_n(c,y). \tag*{(83)} G ( t ; c , Y ( t ; c , y )) = B ( t ; c , y ) , Y = y + n ≥ 1 ∑ t n q n ( c , y ) . (83)
The coefficient of Y Y Y at t = 0 t=0 t = 0 is one, so each q n q_n q n is determined without dividing by a field-dependent Jacobian. To make the induction explicit, let Y < n = y + ∑ j = 1 n − 1 t j q j Y_{<n}=y+\sum_{j=1}^{n-1}t^j q_j Y < n = y + ∑ j = 1 n − 1 t j q j , and define β 1 = ϖ \beta_1=\varpi β 1 = ϖ , β 2 j = 1 2 ( c + y ) \beta_{2j}=\frac{1}{2}(c+y) β 2 j = 2 1 ( c + y ) and β 2 j + 1 = 0 \beta_{2j+1}=0 β 2 j + 1 = 0 for j ≥ 1 j\ge1 j ≥ 1 . Coefficient comparison gives
q n = β n − ∑ k = 1 n [ t n − k ] ℓ k ( c , Y < n ) . (84) q_n=\beta_n-\sum_{k=1}^{n}[t^{n-k}]\ell_k(c,Y_{<n}). \tag*{(84)} q n = β n − k = 1 ∑ n [ t n − k ] ℓ k ( c , Y < n ) . (84)
This determines each order from the lower ones and constructs the formal inverse around Y = y Y=y Y = y .
Suppose q j ( c , 0 ) = 0 q_j(c,0)=0 q j ( c , 0 ) = 0 below order n n n . At y = 0 y=0 y = 0 we have Y < n = 0 Y_{<n}=0 Y < n = 0 ; only the k = n k=n k = n term survives in the sum in (84) . The identity (79) gives
q n ( c , 0 ) = β n ( c , 0 ) − ℓ n ( c , 0 ) = 0 , q n = y d n with d n regular , q_n(c,0)=\beta_n(c,0)-\ell_n(c,0)=0,\qquad q_n=yd_n\ \text{ with }d_n\text{ regular}, q n ( c , 0 ) = β n ( c , 0 ) − ℓ n ( c , 0 ) = 0 , q n = y d n with d n regular ,
h = ∑ n ≥ 1 t n h n , h n = 1 2 ( d n − ∑ j = 1 n − 1 h j h n − j ) , Y = y ( 1 + h ) 2 . (85) h=\sum_{n\ge1}t^n h_n,\qquad h_n=\frac{1}{2}\left(d_n-\sum_{j=1}^{n-1}h_jh_{n-j}\right),\qquad Y=y(1+h)^2. \tag*{(85)} h = n ≥ 1 ∑ t n h n , h n = 2 1 ( d n − j = 1 ∑ n − 1 h j h n − j ) , Y = y ( 1 + h ) 2 . (85)
The base case is the same identity with n = 1 n=1 n = 1 . Since q n q_n q n is analytic and vanishes at y = 0 y=0 y = 0 , the ratio d n = q n / y d_n=q_n/y d n = q n / y has a regular extension to y = 0 y=0 y = 0 . The construction is therefore regular at a vanishing auxiliary field. The square-root recursion for h h h has leading coefficient 2, hence is uniquely solvable at every order.
Use the real, gauge-invariant prepotential shift
V o l d = V + N E A A ˉ h ( t ; c , y ) . (86) V_{\mathrm{old}}=V+\frac{NE}{A\bar A}h(t;c,y). \tag*{(86)} V old = V + A A ˉ N E h ( t ; c , y ) . (86)
Ω o l d = Ω , Ω r o l d = Ω r , Ω c o l d = Ω c , Ω y o l d = Ω Y . \Omega_{\mathrm{old}}=\Omega,\qquad\Omega_{r\mathrm{old}}=\Omega_r,\qquad\Omega_{c\mathrm{old}}=\Omega_c,\qquad\Omega_{y\mathrm{old}}=\Omega_Y. Ω old = Ω , Ω r old = Ω r , Ω c old = Ω c , Ω y old = Ω Y .
The last four identities follow from the finite superspace identities of Appendix A, valid for any local even h h h ; their use does not discard fermions or spacetime derivatives. Applied at each order, they remain valid for the formal h h h in (85) . Since h = O ( t ) h=O(t) h = O ( t ) , the map is locally invertible as a formal power series and preserves the prepotential gauge transformation and the field strength Bianchi identity. At first order the formula gives
q 1 = ϖ − ϖ 2 , d 1 = − ϖ ( 2 c + y ) 1 + ϖ , h 1 = − ϖ ( 2 c + y ) 2 ( 1 + ϖ ) . (87) q_1=\varpi-\varpi^2,\qquad d_1=-\frac{\varpi(2c+y)}{1+\varpi},\qquad h_1=-\frac{\varpi(2c+y)}{2(1+\varpi)}. \tag*{(87)} q 1 = ϖ − ϖ 2 , d 1 = − 1 + ϖ ϖ ( 2 c + y ) , h 1 = − 2 ( 1 + ϖ ) ϖ ( 2 c + y ) . (87)
which is manifestly regular at y = 0 y=0 y = 0 on the branch ϖ = 1 \varpi=1 ϖ = 1 . This supplies a direct normalization check on the general induction.
Combining (83) and (86) gives
Res 1 S sinh γ [ V o l d ( V ) ] = ( 1 4 cosh γ ∫ d 4 x d 2 θ W 2 + c . c . ) + sinh γ ∫ d 4 x d 2 θ d 2 θ ˉ W 2 W ˉ 2 ( D 2 W 2 ) ( D ˉ 2 W ˉ 2 ) . (88) \operatorname{Res}_1 S_{\sinh\gamma}[V_{\mathrm{old}}(V)]
=
\left(\frac{1}{4}\cosh\gamma\int d^4x\,d^2\theta\,W^2+\mathrm{c.c.}\right)
+\sinh\gamma\int d^4x\,d^2\theta\,d^2\bar\theta\,
\frac{W^2\bar W^2}{\sqrt{(D^2W^2)(\bar D^2\bar W^2)}}. \tag*{(88)} Res 1 S s i n h γ [ V old ( V )] = ( 4 1 cosh γ ∫ d 4 x d 2 θ W 2 + c.c. ) + sinh γ ∫ d 4 x d 2 θ d 2 θ ˉ ( D 2 W 2 ) ( D ˉ 2 W ˉ 2 ) W 2 W ˉ 2 . (88)
This equality holds to every order in γ \gamma γ , because t = sinh γ t=\sinh\gamma t = sinh γ is an invertible formal reparametrization at zero. The photon, photino and real auxiliary field of the vector multiplet remain off shell. The additional N = 1 \mathcal{N}=1 N = 1 chiral multiplet has been removed by Res 1 \operatorname{Res}_1 Res 1 , as specified in Section 6. This establishes (75) for the canonical completion of T 1 T_1 T 1 .
Equivalence for the modified completion
The equivalence (75) was established for the canonical completion of T 1 T_1 T 1 , namely S t = S 0 + ∑ t n T n S_t = S_0 + \sum t^n T_n S t = S 0 + ∑ t n T n . It extends to the action (2) by the following induction. Equation (64) gives the first-order equality. Suppose Res 1 F ~ k = Res 1 T k \operatorname{Res}_1 \widetilde{F}_k = \operatorname{Res}_1 T_k Res 1 F k = Res 1 T k for every k < n k < n k < n . Reflection invariance and (74) identify the truncated source of every chiral pairing in the two order-n n n self-duality equations. Hence
δ 0 Res 1 ( F ~ n − T n ) = 0 , Π 0 Res 1 ( F ~ n − T n ) = Res 1 C n = 0. (89) \delta_0 \operatorname{Res}_1(\widetilde{F}_n-T_n)=0,\qquad\Pi_0\operatorname{Res}_1(\widetilde{F}_n-T_n)=\operatorname{Res}_1 C_n=0. \tag*{(89)} δ 0 Res 1 ( F n − T n ) = 0 , Π 0 Res 1 ( F n − T n ) = Res 1 C n = 0. (89)
The first equality says the difference is phase-invariant; the second then makes it zero. The same argument with C n = 0 C_n=0 C n = 0 applies to the canonical modified family F n F_n F n . Therefore
Res 1 F ~ n = Res 1 F n = Res 1 T n ( n ≥ 1 ) , Res 1 S ~ sinh γ [ V ] = S B ( γ ; V B [ V ] ) . (90) \operatorname{Res}_1\widetilde{F}_n=\operatorname{Res}_1 F_n=\operatorname{Res}_1 T_n\quad(n\geq1),\qquad\operatorname{Res}_1\widetilde{S}_{\sinh\gamma}[V]=S_B(\gamma;V_B[V]). \tag*{(90)} Res 1 F n = Res 1 F n = Res 1 T n ( n ≥ 1 ) , Res 1 S s i n h γ [ V ] = S B ( γ ; V B [ V ]) . (90)
The same N = 1 \mathcal{N}=1 N = 1 prepotential redefinition therefore establishes the equivalence for the modified completion.
Physical interpretation and propagation
Scope of the propagation calculation
The following results concern quadratic fluctuations of (2) on constant non-null electromagnetic backgrounds with vanishing scalar expectation value. The quadratic action is determined to all orders in the coupling, and its coefficient functions are analytically continued. The additional chiral multiplet is retained: zero scalar vev is not the restriction Φ = 0 \Phi=0 Φ = 0 used for the N = 1 \mathcal{N}=1 N = 1 equivalence. No propagation claim on other backgrounds is made here. In particular, propagation at nonzero scalar expectation value lies outside the scope of this paper.
Scalar propagation at vanishing scalar expectation value
Take a constant non-null electromagnetic field and vanishing scalar, fermion and auxiliary backgrounds. A frame with parallel electric and magnetic fields covers the whole class:
E = E 0 e 2 , B = B 0 e 2 , r = E 0 2 + B 0 2 2 > 0 , z 0 = d 0 a = ψ 0 i = 0. (91) \mathbf{E}=E_0\mathbf{e}_2,\qquad\mathbf{B}=B_0\mathbf{e}_2,\qquad r=\frac{E_0^2+B_0^2}{2}>0,\qquad z_0=d_0^a=\psi_0^i=0. \tag*{(91)} E = E 0 e 2 , B = B 0 e 2 , r = 2 E 0 2 + B 0 2 > 0 , z 0 = d 0 a = ψ 0 i = 0. (91)
These configurations solve the equations at every formal order. The gauge equation is the divergence of a constant constitutive tensor. The background is fixed by the reflection of § 6.4 , so the action contains no terms linear in the scalar fluctuations. It is also an S U ( 2 ) R SU(2)_R S U ( 2 ) R singlet, excluding a term linear in the auxiliary triplet, while fermion parity excludes terms linear in the fermion fluctuations. These statements are preserved by the recursion and the prescribed phase-invariant interactions.
Scalar–photon mixing is absent in the quadratic action on this class. The discrete reflection of § 6.4 reverses the scalar components and leaves the vector components unchanged; the background is a fixed point. Equivalently, unbroken U ( 1 ) R U(1)_R U ( 1 ) R forbids a scalar–photon bilinear. The scalar is an S U ( 2 ) R SU(2)_R S U ( 2 ) R singlet whereas the auxiliary fields form a triplet, excluding their mixing. Fermion parity separates bosons and fermions. These selection rules are preserved by the full recursion and the prescribed phase-invariant coefficients. The scalar Hessian computed below therefore describes a decoupled sector.
Appendix § 11 computes the independent-chiral scalar Hessian at arbitrary momentum and establishes closure under the recursion. For the canonical modified completion, all higher scalar Hessians vanish: ( F n ) z ( 2 ) = 0 (F_n)^{(2)}_z=0 ( F n ) z ( 2 ) = 0 for n ≥ 2 n\geq2 n ≥ 2 . Its normalized quadratic density is
L z , c a n ( 2 ) = − ( η μ ν − t 2 r T F μ ν ) ∂ μ z ∂ ν z ˉ , v 1 , c a n 2 = 1 − t / 2 1 + t / 2 . (92) \mathcal{L}^{(2)}_{z,\mathrm{can}}=-\left(\eta^{\mu\nu}-\frac{t}{2r}T_F^{\mu\nu}\right)\partial_\mu z\partial_\nu\bar z,\qquad v^2_{1,\mathrm{can}}=\frac{1-t/2}{1+t/2}. \tag*{(92)} L z , can ( 2 ) = − ( η μν − 2 r t T F μν ) ∂ μ z ∂ ν z ˉ , v 1 , can 2 = 1 + t /2 1 − t /2 . (92)
Here T F T_F T F is the Maxwell stress tensor of the background. After analytic continuation in t t t , the coefficient of the transverse gradient term vanishes at t = 2 t=2 t = 2 and becomes negative for t > 2 t>2 t > 2 , giving a gradient instability. This is the reason for specifying higher phase-invariant terms rather than adopting the canonical prescription.
The scalar Hessian of Δ \Delta Δ is nineteen times that of f m o d f_{\mathrm{mod}} f mod . For the action (2), the same closure argument gives
( F ~ n ) z ( 2 ) = [ t n ] d ( t ) ( Δ ) z ( 2 ) , ( n ≥ 2 ) , (\widetilde{F}_n)_{z}^{(2)} = [t^n]d(t)(\Delta)_{z}^{(2)}, \qquad(n \ge2), ( F n ) z ( 2 ) = [ t n ] d ( t ) ( Δ ) z ( 2 ) , ( n ≥ 2 ) ,
β ( t ) = t + 19 d ( t ) 2 = t 1 + t 2 + 1 = tanh γ 2 , (93) \beta(t) = \frac{t+19d(t)}{2} = \frac{t}{\sqrt{1+t^2}+1} = \tanh\frac{\gamma}{2}, \tag*{(93)} β ( t ) = 2 t + 19 d ( t ) = 1 + t 2 + 1 t = tanh 2 γ , (93)
L z ( 2 ) = − ( η μ ν − β ( t ) r T F μ ν ) ∂ μ z ∂ ν z ˉ . L_z^{(2)}=-\left(\eta^{\mu\nu}-\frac{\beta(t)}{r}T_F^{\mu\nu}\right)\partial_\mu z\partial_\nu\bar{z}. L z ( 2 ) = − ( η μν − r β ( t ) T F μν ) ∂ μ z ∂ ν z ˉ .
Within this quadratic form, positive kinetic and spatial coefficients and subluminality hold if and only if 0 ≤ β < 1 0\le\beta<1 0 ≤ β < 1 ; these inequalities alone do not determine β ( t ) \beta(t) β ( t ) . The choice above is selected by the restricted root-flow condition of § 8.6 . It fixes every coefficient of the scalar quadratic series. The closed form of β ( t ) \beta(t) β ( t ) defines its analytic continuation along the positive real t t t axis.
In the frame (91), with contravariant indices,
T F μ ν r = diag ( 1 , 1 , − 1 , 1 ) , K z = ( 1 + β ) ( ω 2 − q ∥ 2 ) − ( 1 − β ) q ⊥ 2 . (94) \frac{T_F^{\mu\nu}}{r}=\operatorname{diag}(1,1,-1,1), \qquad K_z=(1+\beta)(\omega^2-q_{\parallel}^2)-(1-\beta)q_{\perp}^2. \tag*{(94)} r T F μν = diag ( 1 , 1 , − 1 , 1 ) , K z = ( 1 + β ) ( ω 2 − q ∥ 2 ) − ( 1 − β ) q ⊥ 2 . (94)
The negative spatial entry reflects the Maxwell tension along the field direction x 2 x^2 x 2 ; the time component is positive. Writing q ∥ = q 2 q_{\parallel}=q_2 q ∥ = q 2 and q ⊥ 2 = q 1 2 + q 3 2 q_{\perp}^2=q_1^2+q_3^2 q ⊥ 2 = q 1 2 + q 3 2 , one obtains
ω 2 = q ∥ 2 + e − γ q ⊥ 2 , v ∥ 2 = 1 , v ⊥ 2 = e − γ , 1 ± β > 0 ( 0 ≤ γ < ∞ ) . (95) \omega^2=q_{\parallel}^2+e^{-\gamma}q_{\perp}^2, \qquad v_{\parallel}^2=1, \qquad v_{\perp}^2=e^{-\gamma}, \qquad1\pm\beta>0 \quad(0\le\gamma<\infty). \tag*{(95)} ω 2 = q ∥ 2 + e − γ q ⊥ 2 , v ∥ 2 = 1 , v ⊥ 2 = e − γ , 1 ± β > 0 ( 0 ≤ γ < ∞ ) . (95)
Both real scalar components obey this dispersion relation. For a general function d ( t ) d(t) d ( t ) , (94) gives v ⊥ 2 = ( 1 − β ) / ( 1 + β ) v_{\perp}^2=(1-\beta)/(1+\beta) v ⊥ 2 = ( 1 − β ) / ( 1 + β ) with β = ( t + 19 d ( t ) ) / 2 \beta=(t+19d(t))/2 β = ( t + 19 d ( t )) /2 . The positivity, hyperbolicity and causality statements below therefore hold for every d ( t ) d(t) d ( t ) with 0 ≤ β ( t ) < 1 0\le\beta(t)<1 0 ≤ β ( t ) < 1 ; the root-flow choice of § 8.6 selects the value e − γ e^{-\gamma} e − γ . The photon, gaugino and auxiliary operators of § 8.3 follow from the N = 1 \mathcal{N}=1 N = 1 equivalence and S U ( 2 ) R SU(2)_R S U ( 2 ) R and do not depend on d ( t ) d(t) d ( t ) . The positive time and spatial coefficients give a positive quadratic Hamiltonian and a hyperbolic scalar equation. In the same frame the ray speed satisfies
∣ v g ∣ 2 = q ∥ 2 + e − 2 γ q ⊥ 2 q ∥ 2 + e − γ q ⊥ 2 ≤ 1. (96) \lvert v_g\rvert^2=\frac{q_{\parallel}^2+e^{-2\gamma}q_{\perp}^2}{q_{\parallel}^2+e^{-\gamma}q_{\perp}^2}\le1. \tag*{(96)} ∣ v g ∣ 2 = q ∥ 2 + e − γ q ⊥ 2 q ∥ 2 + e − 2 γ q ⊥ 2 ≤ 1. (96)
Hence the scalar characteristic cone lies inside or on the Minkowski light cone. Because the quadratic operator is exactly second order in derivatives, its characteristic cone is obtained directly, rather than from the low-frequency limit of a truncated dispersive operator. The interpretation in terms of an effective scalar metric is the usual one for quadratic fluctuations [28 ] , while the distinction between characteristic cones and well-posed evolution is discussed in [29 , 30 ] . The limit γ → ∞ \gamma\to\infty γ → ∞ is degenerate because the transverse coefficient tends to zero.
In the same parallel-field frame, the bosonic ModMax photon cones are [4 ]
ω + 2 = q ∥ 2 + q ⊥ 2 , ω − 2 = q ∥ 2 + e − 2 γ q ⊥ 2 . (97) \omega_+^2=q_{\parallel}^2+q_{\perp}^2, \qquad\omega_-^2=q_{\parallel}^2+e^{-2\gamma}q_{\perp}^2. \tag*{(97)} ω + 2 = q ∥ 2 + q ⊥ 2 , ω − 2 = q ∥ 2 + e − 2 γ q ⊥ 2 . (97)
The scalar cone lies between these two cones for positive coupling. The background need not preserve supersymmetry, so the scalar and photon cones need not coincide. The remaining quadratic sectors can be determined using the linearization of the vector-potential redefinition, as we now explain.
The remaining fields at vanishing scalar expectation value
The N = 1 \mathcal{N}=1 N = 1 equivalence also determines the photon and photino quadratic operators on (91) . The reflection used in § 6.4 excludes mixing with Φ \Phi Φ . Moreover, the linearization of the prepotential change (108) is particularly simple here. The background has E 0 = 0 E_0=0 E 0 = 0 , while N 0 = W 0 2 W ‾ 0 2 N_0=W_0^2\overline{W}_0^2 N 0 = W 0 2 W 0 2 already contains θ 2 θ ˉ 2 \theta^2\bar{\theta}^2 θ 2 θ ˉ 2 . Thus the linearized change is proportional to N 0 δ E N_0\delta E N 0 δ E : it changes only the auxiliary component of the real prepotential. It is the identity on the photon and photino. Their quadratic operators in the original variables therefore agree with those of the Bandos–Lechner–Sorokin–Townsend action.
On a constant electromagnetic background the term proportional to the derivative of the electromagnetic phase in Eq. (4.28) of [18 ] vanishes. Relative to the free Weyl operator, the photino coefficient is
Z λ = cosh γ + x r sinh γ ≥ e − γ > 0 ( 0 ≤ γ < ∞ ) . (98) Z_\lambda=\cosh\gamma+\frac{x}{r}\sinh\gamma\ge e^{-\gamma}>0 \qquad(0\le\gamma<\infty). \tag*{(98)} Z λ = cosh γ + r x sinh γ ≥ e − γ > 0 ( 0 ≤ γ < ∞ ) . (98)
The background preserves S U ( 2 ) R SU(2)_R S U ( 2 ) R , so both gauginos have this operator. There is no second independent mixed-chirality coefficient, and U ( 1 ) R U(1)_R U ( 1 ) R excludes same-chirality bilinears. Both gauginos consequently have the Minkowski light cone. This conclusion uses the quadratic linearization of the equivalence, as well as the background symmetries; equivalence of the full formal actions alone would not establish it.
The auxiliary coefficient can be obtained directly from the restricted recursion. In the notation of Section 7, expand its interaction density as
J = J ∂ + b u ( v , t ) + b ˉ u ( v ˉ , t ) + O ( b 2 , b b ˉ , b ˉ 2 ) , v = a a a ˉ = e i ϑ , cos ϑ = x r . (99) J=J_{\partial}+bu(v,t)+\bar{b}u(\bar{v},t)+O(b^2,b\bar{b},\bar{b}^2),\qquad v=\frac{a}{\sqrt{a\bar{a}}}=e^{i\vartheta},\qquad\cos\vartheta=\frac{x}{r}. \tag*{(99)} J = J ∂ + b u ( v , t ) + b ˉ u ( v ˉ , t ) + O ( b 2 , b b ˉ , b ˉ 2 ) , v = a a ˉ a = e i ϑ , cos ϑ = r x . (99)
Here u u u is a coefficient function, distinct from the superspace composite u u u of (54) . Equating the terms linear in b b b in δ 0 J = 4 Im p ( J , J ) \delta_0J=4\operatorname{Im}p(J,J) δ 0 J = 4 Im p ( J , J ) gives
− i ( 1 + t 2 + t cos ϑ ) ∂ ϑ u + u + u 2 = 0 , [ v − 1 ] u = t . (100) -i\left(\sqrt{1+t^2}+t\cos\vartheta\right)\partial_\vartheta u+u+u^2=0,\qquad[v^{-1}]u=t. \tag*{(100)} − i ( 1 + t 2 + t cos ϑ ) ∂ ϑ u + u + u 2 = 0 , [ v − 1 ] u = t . (100)
The last condition is the prescribed phase average: b b b has charge + 2 +2 + 2 , so the v − 1 v^{-1} v − 1 coefficient is fixed by the first-order interaction. Put τ = tanh ( γ / 2 ) \tau=\tanh(\gamma/2) τ = tanh ( γ /2 ) . The solution connected to u = 0 u=0 u = 0 at t = 0 t=0 t = 0 is
u = w 1 − w , w = ρ 1 + τ v v + τ , ρ ( 1 − τ 2 ) ( 1 − ρ τ ) 2 = t . (101) u=\frac{w}{1-w},\qquad w=\rho\frac{1+\tau v}{v+\tau},\qquad\frac{\rho(1-\tau^2)}{(1-\rho\tau)^2}=t. \tag*{(101)} u = 1 − w w , w = ρ v + τ 1 + τ v , ( 1 − ρ τ ) 2 ρ ( 1 − τ 2 ) = t . (101)
Substitution verifies the differential equation. The Laurent expansion of u u u about ∣ v ∣ = 1 |v|=1 ∣ v ∣ = 1 gives the last relation from [ v − 1 ] u = t [v^{-1}]u=t [ v − 1 ] u = t . For 0 < τ < 1 0<\tau<1 0 < τ < 1 , the function ρ ( 1 − τ 2 ) 2 − 2 τ ( 1 − ρ τ ) 2 \rho(1-\tau^2)^2-2\tau(1-\rho\tau)^2 ρ ( 1 − τ 2 ) 2 − 2 τ ( 1 − ρ τ ) 2 is strictly increasing on [ 0 , 1 ] [0,1] [ 0 , 1 ] , negative at zero and positive at one. Hence there is one solution with 0 < ρ < 1 0<\rho<1 0 < ρ < 1 . Since ∣ w ∣ = ρ |w|=\rho ∣ w ∣ = ρ on the unit circle, the auxiliary coefficient, relative to its free value, is
Z d = 1 + u + u ˉ = 1 − ρ 2 ∣ 1 − w ∣ 2 > 0 , Z d = 1 + 2 x r t + 2 x 2 r 2 t 2 + O ( t 3 ) . (102) Z_d=1+u+\bar{u}=\frac{1-\rho^2}{|1-w|^2}>0,\qquad Z_d=1+2\frac{x}{r}t+2\frac{x^2}{r^2}t^2+O(t^3). \tag*{(102)} Z d = 1 + u + u ˉ = ∣1 − w ∣ 2 1 − ρ 2 > 0 , Z d = 1 + 2 r x t + 2 r 2 x 2 t 2 + O ( t 3 ) . (102)
No spacetime derivative acts on the auxiliary fluctuation in this quadratic action. S U ( 2 ) R SU(2)_R S U ( 2 ) R gives the same coefficient for all three real auxiliary fields.
Together with (95) and (97) , these results determine the quadratic series of the whole multiplet on (91) . Its analytic continuation has causal scalar, photon and gaugino propagation and algebraic auxiliary fields at every finite γ ≥ 0 \gamma\ge0 γ ≥ 0 .
Comparison with compensator constructions
Kuzenko and Raptakis, Section 4.3 of [20 ] , note that no N = 2 \mathcal{N}=2 N = 2 superconformal ModMax theory had been constructed and identify the dimensionless full superspace measure as the obstacle. Equation numbers in this comparison refer to arXiv version 5. Their Eq. (4.28) supplies a compensator-dependent N = 2 \mathcal{N}=2 N = 2 model using an auxiliary chiral scalar and a compensating vector field strength. After switching off the additional physical N = 1 \mathcal{N}=1 N = 1 chiral multiplet and eliminating the auxiliary chiral components, their model reduces to N = 1 \mathcal{N}=1 N = 1 ModMax; see their Eqs. (4.29)–(4.32). An N = 2 \mathcal{N}=2 N = 2 model whose truncation is N = 1 \mathcal{N}=1 N = 1 ModMax therefore already exists in the literature. The present construction uses no additional compensator in the N = 2 \mathcal{N}=2 N = 2 action.
For the minimal interaction T 1 T_1 T 1 the relevant dimensions are
[ d 4 x d 8 θ ] = 0 , [ W 2 W ‾ 2 ] = [ Q ] = 4 , [ W 2 W ‾ 2 Q ] = 0. (103) [\mathrm{d}^{4}x\,\mathrm{d}^{8}\theta]=0,\qquad[\mathcal{W}^{2}\overline{\mathcal{W}}^{2}]=[Q]=4,\qquad\left[\frac{\mathcal{W}^{2}\overline{\mathcal{W}}^{2}}{Q}\right]=0. \tag*{(103)} [ d 4 x d 8 θ ] = 0 , [ W 2 W 2 ] = [ Q ] = 4 , [ Q W 2 W 2 ] = 0. (103)
The derivative composite Q Q Q supplies the required weight, and the additional density D 33 D_{33} D 33 in (48) also has a dimensionless superspace integrand. Both are constructed using only the physical field strength. Dimension counting alone is insufficient: § 3 establishes the cancellation of its inhomogeneous superconformal variation. This choice admits z 0 = 0 z_0=0 z 0 = 0 at nonzero electromagnetic invariants, at the cost of a nonanalytic denominator and a derivative-dependent interaction on the regular domain specified in subsection 3.4.
Section 4 of ref. [19 ] noted that the N = 1 \mathcal{N}=1 N = 1 superconformal ModMax model does not appear to have an N = 2 \mathcal{N}=2 N = 2 counterpart. Superconformal duality-invariant N = 2 \mathcal{N}=2 N = 2 models are constructed in [19 ] itself, and general methods are reviewed in [27 ] . Those models involve logarithms and inverse powers of W \mathcal{W} W and are defined on a domain where its lowest scalar component is nonzero. The present interactions instead admit z 0 = 0 z_0=0 z 0 = 0 on constant non-null electromagnetic backgrounds, subject to the regularity conditions specified in § 3.3 .
Higher-derivative ModMax deformations preserving duality and Weyl invariance are available in the bosonic and N = 1 \mathcal{N}=1 N = 1 supersymmetric settings [23 ] . They demonstrate that the symmetry requirements allow interactions beyond the original derivative-free electromagnetic Lagrangian. They do not determine the particular first-order interaction or choice of the phase-invariant terms used here. The additional primary density in (48) alters the first-order scalar Hessian while preserving Res 1 f m o d = Res 1 T 1 \operatorname{Res}_1 f_{\mathrm{mod}}=\operatorname{Res}_1 T_1 Res 1 f mod = Res 1 T 1 ; the higher phase-invariant coefficients then determine the scalar quadratic action at higher orders.
Comparison with higher-spin currents
Off-shell cubic couplings of the N = 2 \mathcal{N}=2 N = 2 hypermultiplet to higher-spin gauge superfields and the associated conserved supercurrents were constructed in harmonic superspace by Buchbinder, Ivanov and Zaigraev [48 , 49 ] . Kuzenko and Raptakis studied conformal and non-conformal N = 2 \mathcal{N}=2 N = 2 higher-spin supercurrent multiplets, including conformal currents built from the vector multiplet [47 ] . Higher-spin currents and symmetries of the vector multiplet were studied further in [50 ] .
The vector-multiplet currents J α ( s ) α ˙ ( s ) J_{\alpha(s)\dot{\alpha}(s)} J α ( s ) α ˙ ( s ) of [47 ] are conserved when the field strengths obey the free equations of motion. Their Eqs. (2.10)–(2.12) contain the relevant s = 2 s=2 s = 2 bilinear sector. Its first three structures have the ratios − X + 4 Y − i Z -X+4Y-iZ − X + 4 Y − i Z , also present in K ^ / 32 \widehat{K}/32 K /32 . The remaining structure uses contracted two-spinor derivatives. A complete identification with that current, including conjugate derivative ordering and possible equation-of-motion or improvement terms, has not been established. We do not claim novelty for the bilinear current itself.
Our construction uses its off-shell transformation law rather than a conservation equation valid on the free equations of motion. A primary superfield with a spin-two lowest component need not be the stress tensor defined by metric variation, and we do not identify K ^ \widehat{K} K with that tensor. The N = 2 \mathcal{N}=2 N = 2 supercurrent multiplet corresponds to a scalar superfield in [47 ] and couples to supergravity [46 ] . Primarity alone therefore implies neither causal propagation nor a stress-tensor flow equation.
Comparison with stress-tensor flows
Stress-tensor flows provide a separate way of constructing duality-preserving deformations [36 , 37 , 38 , 40 , 41 ] . We compare the component action with
∂ γ L = R [ T ] , R [ T ] = 1 2 T μ ν T μ ν − 1 4 ( T μ μ ) 2 . (104) \partial_{\gamma} L = R[T], \qquad R[T] = \frac{1}{2}\sqrt{T_{\mu\nu}T^{\mu\nu}-\frac{1}{4}(T_{\mu}^{\mu})^{2}}. \tag*{(104)} ∂ γ L = R [ T ] , R [ T ] = 2 1 T μν T μν − 4 1 ( T μ μ ) 2 . (104)
The four-dimensional square-root operator was introduced by Babaei-Aghbolagh et al. in Eq. (3.6) of [36 ] ; their Eqs. (3.5) and (4.1) give the normalization and ModMax flow used here. The choice of stress tensor matters. At zero scalar expectation value, we vary the photon and scalar quadratic densities with the metric, holding F μ ν F_{\mu\nu} F μν and the scalar fixed. A conformal scalar improvement changes the quadratic comparison only by a total derivative on a constant electromagnetic background. All scalar comparisons in this subsection are restricted to quadratic fluctuations at vanishing scalar expectation value.
For ModMax, T 0 μ ν = κ γ T F μ ν T_{0\mu\nu}=\kappa_{\gamma}T_{F\mu\nu} T 0 μν = κ γ T F μν , with κ γ = cosh γ + ( x / r ) sinh γ > 0 \kappa_{\gamma}=\cosh\gamma+(x/r)\sinh\gamma>0 κ γ = cosh γ + ( x / r ) sinh γ > 0 . Since T F μ ν T F μ ν = 4 r 2 T_{F\mu\nu}T_{F}^{\mu\nu}=4r^{2} T F μν T F μν = 4 r 2 , the positive square root gives R 0 = κ γ r = ∂ γ L M M R_{0}=\kappa_{\gamma}r=\partial_{\gamma}L_{\mathrm{MM}} R 0 = κ γ r = ∂ γ L MM . For the scalar quadratic density at z 0 = 0 z_{0}=0 z 0 = 0 , put
M μ ν = T F μ ν r , s z = ∂ μ z ∂ μ z ˉ , u z = M μ ν ∂ μ z ∂ ν z ˉ , L z ( 2 ) = − s z + β u z . (105) M^{\mu\nu}=\frac{T_{F}^{\mu\nu}}{r}, \qquad s_{z}=\partial_{\mu}z\partial^{\mu}\bar z, \qquad u_{z}=M^{\mu\nu}\partial_{\mu}z\partial_{\nu}\bar z, \qquad L_{z}^{(2)}=-s_{z}+\beta u_{z}. \tag*{(105)} M μν = r T F μν , s z = ∂ μ z ∂ μ z ˉ , u z = M μν ∂ μ z ∂ ν z ˉ , L z ( 2 ) = − s z + β u z . (105)
The tensor M M M must be varied along with the metric. The directional variation δ M g μ ν = M μ ν \delta M g^{\mu\nu}=M^{\mu\nu} δ M g μν = M μν gives δ M s z = u z \delta_{M}s_{z}=u_{z} δ M s z = u z and δ M u z = s z \delta_{M}u_{z}=s_{z} δ M u z = s z . Expanding the root yields
R ( 2 ) = 1 2 ( u z − β s z ) , R ( 2 ) − ∂ γ L z ( 2 ) − β 2 L z ( 2 ) = [ 1 2 ( 1 − β 2 ) − ∂ γ β ] u z . (106) R^{(2)}=\frac{1}{2}(u_{z}-\beta s_{z}), \qquad R^{(2)}-\partial_{\gamma}L_{z}^{(2)}-\frac{\beta}{2}L_{z}^{(2)}=\left[\frac{1}{2}(1-\beta^{2})-\partial_{\gamma}\beta\right]u_{z}. \tag*{(106)} R ( 2 ) = 2 1 ( u z − β s z ) , R ( 2 ) − ∂ γ L z ( 2 ) − 2 β L z ( 2 ) = [ 2 1 ( 1 − β 2 ) − ∂ γ β ] u z . (106)
With the Maxwell initial condition β ( 0 ) = 0 \beta(0)=0 β ( 0 ) = 0 , the vanishing of the bracket gives ∂ γ β = ( 1 − β 2 ) / 2 \partial_{\gamma}\beta=(1-\beta^{2})/2 ∂ γ β = ( 1 − β 2 ) /2 and hence β = tanh ( γ / 2 ) \beta=\tanh(\gamma/2) β = tanh ( γ /2 ) . Through β = ( t + 19 d ( t ) ) / 2 \beta=(t+19d(t))/2 β = ( t + 19 d ( t )) /2 , this fixes the function d ( t ) d(t) d ( t ) used in (2). The remaining term is proportional to the quadratic scalar action and hence to its equations of motion modulo a divergence. Equivalently, the change of normalization
z = cosh ( γ / 2 ) z ^ , L ^ z ( 2 ) = − cosh ( γ / 2 ) s ^ z + sinh ( γ / 2 ) u ^ z (107) z=\sqrt{\cosh(\gamma/2)}\,\hat z, \qquad\hat L_{z}^{(2)}=-\cosh(\gamma/2)\hat s_{z}+\sinh(\gamma/2)\hat u_{z} \tag*{(107)} z = cosh ( γ /2 ) z ^ , L ^ z ( 2 ) = − cosh ( γ /2 ) s ^ z + sinh ( γ /2 ) u ^ z (107)
gives ∂ γ L ^ z ( 2 ) = R ^ ( 2 ) \partial_{\gamma}\hat L_{z}^{(2)}=\hat R^{(2)} ∂ γ L ^ z ( 2 ) = R ^ ( 2 ) directly. The scalar characteristic cone is unchanged.
The photon and quadratic scalar sectors therefore satisfy the stated component root flow on the backgrounds considered here. This agreement fixes the chosen scalar coefficient function but does not establish a supersymmetric flow for the full action. The construction of an interacting conserved supercurrent and the associated flow lies outside the scope of this paper. The action (2) is defined by self-duality and the prescribed phase-invariant interactions.
Conclusions
The prescribed action (2) is a compensator-free, superconformal and self-dual formal series for an Abelian N = 2 \mathcal{N}=2 N = 2 vector multiplet. Its N = 1 \mathcal{N}=1 N = 1 vector restriction is equivalent off shell to the Bandos–Lechner–Sorokin–Townsend action through a local prepotential change. The construction is defined on a regular domain where the composite field-dependent denominator is nonzero and does not include constant zero-field vacua.
The scalar dynamics requires additional choices beyond symmetry and the vector restriction. On constant non-null electromagnetic backgrounds with vanishing scalar expectation value, the prescribed phase-invariant terms determine the quadratic action to all orders. Its analytic continuation has scalar transverse speed squared e − γ e^{-\gamma} e − γ , the ModMax photon cones, luminal gauginos and algebraic auxiliaries, with positive kinetic coefficients for finite γ ≥ 0 \gamma\geq0 γ ≥ 0 .
Supersymmetry of the action alone does not imply that all component fields have the same characteristic cone on a given electromagnetic background.
Within the quadratic scalar form studied here, the stress-tensor square-root flow modulo equations of motion and a total derivative fixes the function d ( t ) d(t) d ( t ) . This is a condition on that sector, not a derivation of the full action from a supersymmetric flow. Causal scalar propagation holds for every d ( t ) d(t) d ( t ) with 0 ≤ β < 1 0 \le\beta< 1 0 ≤ β < 1 ; the flow selects v ⊥ 2 = e − γ v_{\perp}^{2} = e^{-\gamma} v ⊥ 2 = e − γ . An interacting conserved supercurrent and a supersymmetric flow generating the full action have not been established.
The superconformal and self-duality requirements, the vector restriction, and the zero-vev propagation conditions leave freedom in the full superspace interactions. The zero-vev results hold for a class of first-order interactions, of which (48) is a simple representative. Propagation beyond the backgrounds considered here remain open. The analysis at nonzero scalar expectation value and the construction of a supersymmetric flow in terms of a conserved supercurrent will be addressed in a separate paper.
Finite identities for the prepotential redefinition
We now impose the Bianchi identity u = v = E u = v = E u = v = E and work with a real prepotential. For any real local even h h h , consider the finite shift
V ′ = V + N H , H = E A A ˉ h , W α ′ = W α − 1 2 D ˉ 2 D α ( N H ) . (108) V' = V + NH,\qquad H = \frac{E}{A\bar A}h,\qquad W'_{\alpha} = W_{\alpha} - \frac{1}{2}\bar D^{2}D_{\alpha}(NH). \tag*{(108)} V ′ = V + N H , H = A A ˉ E h , W α ′ = W α − 2 1 D ˉ 2 D α ( N H ) . (108)
In the application h = O ( t ) h = O(t) h = O ( t ) , so the transformation is formally invertible near the identity and preserves the gauge transformation of V V V . The corresponding W α ′ W'_{\alpha} W α ′ is a chiral field strength satisfying its off-shell Bianchi identity.
Define
δ = E 2 8 , x = − A + A ˉ + E 2 8 , p = i ( A − A ˉ ) 2 , r = x 2 + p 2 . (109) \delta= \frac{E^{2}}{8},\qquad x = -\frac{A+\bar A+E^{2}}{8},\qquad p = \frac{i(A-\bar A)}{2},\qquad r = \sqrt{x^{2}+p^{2}}. \tag*{(109)} δ = 8 E 2 , x = − 8 A + A ˉ + E 2 , p = 2 i ( A − A ˉ ) , r = x 2 + p 2 . (109)
On the regular domain, the finite shift (108) satisfies the superfield identities
Ω ′ = Ω , Ω x ′ = Ω x , Ω p ′ = Ω p , Ω E ′ = Ω E ( 1 + h ) , Ω δ ′ = Ω δ ( 1 + h ) 2 . (110) \Omega' = \Omega,\qquad\Omega x' = \Omega x,\qquad\Omega p' = \Omega p,\qquad\Omega E' = \Omega E(1+h),\qquad\Omega\delta' = \Omega\delta(1+h)^{2}. \tag*{(110)} Ω ′ = Ω , Ω x ′ = Ω x , Ω p ′ = Ω p , Ω E ′ = Ω E ( 1 + h ) , Ω δ ′ = Ω δ ( 1 + h ) 2 . (110)
The factors of Ω \Omega Ω are part of these identities and must be retained. To derive them, we expand the D D D derivative in (108) :
W α ′ = W α − 1 2 D ˉ 2 { W ˉ 2 ( D α W 2 ) H } + W 2 Y α = C α β W β + W 2 Y α . (111) W'_{\alpha} = W_{\alpha} - \frac{1}{2}\bar D^{2}\{\bar W^{2}(D_{\alpha}W^{2})H\} + W^{2}Y_{\alpha} = C_{\alpha}{}^{\beta}W_{\beta} + W^{2}Y_{\alpha}. \tag*{(111)} W α ′ = W α − 2 1 D ˉ 2 { W ˉ 2 ( D α W 2 ) H } + W 2 Y α = C α β W β + W 2 Y α . (111)
All derivatives of H H H in the omitted expression remain in the explicit W 2 Y α W^{2}Y_{\alpha} W 2 Y α remainder or in the chiral matrix C C C . The factorization is exact. The nilpotency of the Grassmann-odd spinor superfields implies
W ′ 2 = ( det C ) W 2 , N ′ = N det C det C ˉ . (112) W'^{2} = (\det C)W^{2},\qquad N' = N\det C\det\bar C. \tag*{(112)} W ′ 2 = ( det C ) W 2 , N ′ = N det C det C ˉ . (112)
Differentiate the exact expression for W ′ 2 W'^{2} W ′ 2 before multiplying by N N N . The product rule gives
N A ′ = N A det C , N A ˉ ′ = N A ˉ det C ˉ . (113) NA' = NA\det C,\qquad N\bar A' = N\bar A\det\bar C. \tag*{(113)} N A ′ = N A det C , N A ˉ ′ = N A ˉ det C ˉ . (113)
Every other term has an undifferentiated factor W W W or W ˉ \bar W W ˉ and vanishes when multiplied by N N N . Invertibility of the bodies then gives the finite identity
4 N ′ A ′ A ˉ ′ = 4 N det C det C ˉ A A ˉ det C det C ˉ = Ω . (114) \frac{4N'}{A'\bar A'} = \frac{4N\det C\det\bar C}{A\bar A\det C\det\bar C} = \Omega. \tag*{(114)} A ′ A ˉ ′ 4 N ′ = A A ˉ det C det C ˉ 4 N det C det C ˉ = Ω. (114)
For the remaining identities compute D α W β ′ D_{\alpha}W'_{\beta} D α W β ′ from the exact factorization (111) . Terms in which D D D acts on C C C , or the W 2 Y W^{2}\mathcal{Y} W 2 Y remainder, retain an undifferentiated factor W W W and vanish after multiplication by N N N . Evaluating the surviving matrix with (70) gives
N δ m α β = − 1 2 N A A ˉ H ε α β = − 1 2 N E h ε α β . (115) N\delta m_{\alpha\beta}=-\frac{1}{2}N A\bar{A}H\,\varepsilon_{\alpha\beta}=-\frac{1}{2}NEh\,\varepsilon_{\alpha\beta}. \tag*{(115)} N δ m α β = − 2 1 N A A ˉ H ε α β = − 2 1 N E h ε α β . (115)
This can be seen explicitly from
N ( C − 1 ) = N A A ˉ H ( m 12 − m 11 m 22 − m 21 ) , N m ′ = N m C T . (116) N(C-1)=N A\bar{A}H\begin{pmatrix}m_{12}&-m_{11}\\m_{22}&-m_{21}\end{pmatrix},\qquad Nm'=NmC^{\mathrm{T}}. \tag*{(116)} N ( C − 1 ) = N A A ˉ H ( m 12 m 22 − m 11 − m 21 ) , N m ′ = N m C T . (116)
The symmetric part of m m m is unchanged, whereas N E ′ = N E ( 1 + h ) NE'=NE(1+h) N E ′ = N E ( 1 + h ) . Applying the exact product rule for A ′ A' A ′ and the spinor determinant identity already used in (62) , one finds
N ( A ′ + 1 2 E ′ 2 ) = N ( A + 1 2 E 2 ) , N ( A ˉ ′ + 1 2 E ′ 2 ) = N ( A ˉ + 1 2 E 2 ) . (117) N\left(A'+\frac{1}{2}E'^{2}\right)=N\left(A+\frac{1}{2}E^{2}\right),\qquad N\left(\bar{A}'+\frac{1}{2}E'^{2}\right)=N\left(\bar{A}+\frac{1}{2}E^{2}\right). \tag*{(117)} N ( A ′ + 2 1 E ′ 2 ) = N ( A + 2 1 E 2 ) , N ( A ˉ ′ + 2 1 E ′ 2 ) = N ( A ˉ + 2 1 E 2 ) . (117)
Together with (114) , these equations give all the identities in (110) . The same identities hold for analytic functions of these invariants after multiplication by Ω \Omega Ω , on the chosen branch.
Scalar quadratic action and recursive closure
We calculate scalar fluctuations about (91) , retaining the additional chiral multiplet and arbitrary momentum. Independent chiral arguments must be varied before imposing reduced chirality, as required by the self-duality equation.
Conventions and background normalization
In chiral coordinates use
D α i = ∂ ∂ θ i α + 2 i θ ˉ β ˙ i ∂ α β ˙ , D ˉ α ˙ i = − ∂ ∂ θ ˉ α ˙ i . (118) D^{i}_{\alpha}=\frac{\partial}{\partial\theta^{\alpha}_{i}}+2i\bar{\theta}^{\dot{\beta}i}\partial_{\alpha\dot{\beta}},\qquad\bar{D}_{\dot{\alpha}i}=-\frac{\partial}{\partial\bar{\theta}^{\dot{\alpha}i}}. \tag*{(118)} D α i = ∂ θ i α ∂ + 2 i θ ˉ β ˙ i ∂ α β ˙ , D ˉ α ˙ i = − ∂ θ ˉ α ˙ i ∂ . (118)
The momentum matrix and ordered monomials are
p α α ˙ = ( ω + q 3 q 1 − i q 2 q 1 + i q 2 ω − q 3 ) , P = det p = ω 2 − q 1 2 − q 2 2 − q 3 2 , p_{\alpha\dot{\alpha}}=\begin{pmatrix}\omega+q_{3}&q_{1}-iq_{2}\\q_{1}+iq_{2}&\omega-q_{3}\end{pmatrix},\qquad P=\det p=\omega^{2}-q_{1}^{2}-q_{2}^{2}-q_{3}^{2}, p α α ˙ = ( ω + q 3 q 1 + i q 2 q 1 − i q 2 ω − q 3 ) , P = det p = ω 2 − q 1 2 − q 2 2 − q 3 2 ,
Θ = θ 1 1 θ 2 1 + θ 1 2 θ 2 2 , Θ ˉ = θ ˉ 1 ˙ 1 θ ˉ 1 ˙ 2 + θ ˉ 2 ˙ 1 θ ˉ 2 ˙ 2 , \Theta=\theta^{1}_{1}\theta^{1}_{2}+\theta^{2}_{1}\theta^{2}_{2},\qquad\bar{\Theta}=\bar{\theta}^{\dot{1}1}\bar{\theta}^{\dot{1}2}+\bar{\theta}^{\dot{2}1}\bar{\theta}^{\dot{2}2}, Θ = θ 1 1 θ 2 1 + θ 1 2 θ 2 2 , Θ ˉ = θ ˉ 1 ˙ 1 θ ˉ 1 ˙ 2 + θ ˉ 2 ˙ 1 θ ˉ 2 ˙ 2 ,
θ 4 = θ 1 1 θ 1 2 θ 2 1 θ 2 2 , θ ˉ 4 = θ ˉ 1 ˙ 1 θ ˉ 2 ˙ 1 θ ˉ 1 ˙ 2 θ ˉ 2 ˙ 2 . (119) \theta^{4}=\theta^{1}_{1}\theta^{2}_{1}\theta^{1}_{2}\theta^{2}_{2},\qquad\bar{\theta}^{4}=\bar{\theta}^{\dot{1}1}\bar{\theta}^{\dot{2}1}\bar{\theta}^{\dot{1}2}\bar{\theta}^{\dot{2}2}. \tag*{(119)} θ 4 = θ 1 1 θ 1 2 θ 2 1 θ 2 2 , θ ˉ 4 = θ ˉ 1 ˙ 1 θ ˉ 2 ˙ 1 θ ˉ 1 ˙ 2 θ ˉ 2 ˙ 2 . (119)
For parallel fields along x 2 x^{2} x 2 choose
W F = − f Θ , W ‾ F = g Θ ˉ , f = b F + i e F , g = b F − i e F , b F , e F ∈ R . (120) \mathcal{W}_{F}=-f\Theta,\qquad\overline{\mathcal{W}}_{F}=g\bar{\Theta},\qquad f=b_{F}+ie_{F},\qquad g=b_{F}-ie_{F},\qquad b_{F},e_{F}\in\mathbb{R}. \tag*{(120)} W F = − f Θ , W F = g Θ ˉ , f = b F + i e F , g = b F − i e F , b F , e F ∈ R . (120)
We may keep f f f , g g g algebraically independent during extraction and impose reality afterwards. We write the chiral Grassmann integral as C C > 0 CC>0 C C > 0 times the coefficient of the ordered monomial θ 4 \theta^{4} θ 4 . Component coefficients are converted to the normalization (7) before constants are compared. The relation to the physical fields is
B 0 = C C b F , E 0 = C C e F , B_{0}=\sqrt{CC}\,b_{F},\qquad E_{0}=\sqrt{CC}\,e_{F}, B 0 = C C b F , E 0 = C C e F ,
x = − C C 4 ( f 2 + g 2 ) , p = C C b F e F , r = C C 2 f g > 0 , ν : = f 2 + g 2 2 f g = − x r . (121) x=-\frac{CC}{4}(f^{2}+g^{2}),\qquad p=CCb_{F}e_{F},\qquad r=\frac{CC}{2}fg>0,\qquad\nu:=\frac{f^{2}+g^{2}}{2fg}=-\frac{x}{r}. \tag*{(121)} x = − 4 C C ( f 2 + g 2 ) , p = C C b F e F , r = 2 C C f g > 0 , ν := 2 f g f 2 + g 2 = − r x . (121)
The orientation is chosen so that p = E ⋅ B p=\mathbf{E}\cdot\mathbf{B} p = E ⋅ B . Reversing that convention leaves the quadratic results unchanged. The normalization of (22) gives
L 1 = 2 C C [ N Q ] θ 4 θ ˉ 4 , N = W 2 W ‾ 2 , Q 0 = 16 f g , H = K ^ 2 − Q 0 2 , L 0 , F = − C C 4 ( f 2 + g 2 ) = x , L 1 , F = C C 2 f g = r , L 0 , ϕ ( 2 ) = − 2 C C ∂ ϕ ⋅ ∂ ϕ ˉ , z = 2 C C ϕ . (122) \begin{aligned}
L_1&=2C_C\left[\frac{N}{Q}\right]_{\theta^4\bar{\theta}^4},\qquad N=\mathcal{W}^2\overline{\mathcal{W}}^2,\qquad Q_0=16fg,\qquad H=\widehat{K}^2-Q_0^2,\\
L_{0,F}&=-\frac{C_C}{4}(f^2+g^2)=x,\qquad L_{1,F}=\frac{C_C}{2}fg=r,\\
L_{0,\phi}^{(2)}&=-2C_C\partial\phi\cdot\partial\bar{\phi},\qquad z=\sqrt{2C_C}\,\phi. \tag*{(122)}
\end{aligned} L 1 L 0 , F L 0 , ϕ ( 2 ) = 2 C C [ Q N ] θ 4 θ ˉ 4 , N = W 2 W 2 , Q 0 = 16 f g , H = K 2 − Q 0 2 , = − 4 C C ( f 2 + g 2 ) = x , L 1 , F = 2 C C f g = r , = − 2 C C ∂ ϕ ⋅ ∂ ϕ ˉ , z = 2 C C ϕ . (122)
The positive branch of the square root in the photon sector and the standard sign of the free scalar kinetic term therefore follow from the same normalization. The purely magnetic specialization is f = g ≠ 0 f=g\ne0 f = g = 0 , with B 0 2 = C C f 2 B_0^2=C_Cf^2 B 0 2 = C C f 2 .
Scalar Hessian before imposing reduced chirality
Let T \mathcal{T} T denote the translation from antichiral to chiral coordinates. Introduce four independent scalar fluctuation amplitudes,
δ W = ϕ + θ 4 c , δ W ‾ = T ( ϕ ˉ + θ ˉ 4 c ˉ ) . (123) \delta\mathcal{W}=\phi+\theta^4c,\qquad\delta\overline{\mathcal{W}}=\mathcal{T}\left(\bar{\phi}+\bar{\theta}^4\bar{c}\right). \tag*{(123)} δ W = ϕ + θ 4 c , δ W = T ( ϕ ˉ + θ ˉ 4 c ˉ ) . (123)
Reduced chirality subsequently imposes c = 4 P ϕ ˉ c=4P\bar{\phi} c = 4 P ϕ ˉ and c ˉ = 4 P ϕ \bar{c}=4P\phi c ˉ = 4 P ϕ . The calculation uses eight odd coordinates and two commuting plane-wave amplitudes of opposite momenta. Terms quadratic in these amplitudes give the Hessian. For any numerator J J J multiplying Q − n Q^{-n} Q − n , put H = Q 2 − Q 0 2 H=Q^2-Q_0^2 H = Q 2 − Q 0 2 , Q 0 = 16 f g Q_0=16fg Q 0 = 16 f g . Since H H H vanishes on the background,
[ J Q n ] ( 2 ) = 1 Q 0 n [ J − n J H 2 Q 0 2 + n ( n + 2 ) J H 2 8 Q 0 4 ] ( 2 ) . (124) \left[\frac{J}{Q^n}\right]_{(2)}
=\frac{1}{Q_0^n}\left[J-\frac{nJH}{2Q_0^2}+\frac{n(n+2)JH^2}{8Q_0^4}\right]_{(2)}. \tag*{(124)} [ Q n J ] ( 2 ) = Q 0 n 1 [ J − 2 Q 0 2 n J H + 8 Q 0 4 n ( n + 2 ) J H 2 ] ( 2 ) . (124)
To evaluate the mixed lowest-scalar entry, take ϕ \phi ϕ and ϕ ˉ \bar{\phi} ϕ ˉ with momenta p α α ˙ p_{\alpha\dot{\alpha}} p α α ˙ and − p α α ˙ -p_{\alpha\dot{\alpha}} − p α α ˙ , respectively, and set c = c ˉ = 0 c=\bar{c}=0 c = c ˉ = 0 temporarily. Write ℓ = ∑ i θ i α p α α ˙ θ ˉ α ˙ i \ell=\sum_i\theta_i^\alpha p_{\alpha\dot{\alpha}}\bar{\theta}^{\dot{\alpha}i} ℓ = ∑ i θ i α p α α ˙ θ ˉ α ˙ i , so that T ϕ ˉ = e − 2 ℓ ϕ ˉ \mathcal{T}\bar{\phi}=e^{-2\ell}\bar{\phi} T ϕ ˉ = e − 2 ℓ ϕ ˉ . Let [ ⋅ ] ∗ [\cdot]_\ast [ ⋅ ] ∗ select the coefficient of θ 4 θ ˉ 4 ϕ ϕ ˉ \theta^4\bar{\theta}^4\phi\bar{\phi} θ 4 θ ˉ 4 ϕ ϕ ˉ , with the two plane-wave factors removed. Grassmann multiplication and the background tensor give
Θ 2 = − 2 θ 4 , Θ ˉ 2 = − 2 θ ˉ 4 , [ Θ Θ ˉ ℓ 2 ] θ 4 θ ˉ 4 = − 4 Σ , \Theta^2=-2\theta^4,\qquad\bar{\Theta}^2=-2\bar{\theta}^4,\qquad[\Theta\bar{\Theta}\ell^2]_{\theta^4\bar{\theta}^4}=-4\Sigma, Θ 2 = − 2 θ 4 , Θ ˉ 2 = − 2 θ ˉ 4 , [ Θ Θ ˉ ℓ 2 ] θ 4 θ ˉ 4 = − 4Σ ,
( K ^ 0 ) α β α ˙ β ˙ = 8 f g δ α β δ α ˙ β ˙ , K ^ 0 v w p v p w = 16 f g Σ , Σ = ω 2 + q 1 2 − q 2 2 + q 3 2 . (125) (\widehat{K}_0)_{\alpha\beta\dot{\alpha}\dot{\beta}}=8fg\,\delta_{\alpha\beta}\delta_{\dot{\alpha}\dot{\beta}},\qquad\widehat{K}_0^{vw}p_vp_w=16fg\Sigma,\qquad\Sigma=\omega^2+q_1^2-q_2^2+q_3^2. \tag*{(125)} ( K 0 ) α β α ˙ β ˙ = 8 f g δ α β δ α ˙ β ˙ , K 0 v w p v p w = 16 f g Σ , Σ = ω 2 + q 1 2 − q 2 2 + q 3 2 . (125)
The Kronecker symbols in this expression specify the components in the frame (120); they are not invariant spinor metrics. Tensor indices are raised with the antisymmetric spinor tensors as in the definition of K ^ ⋅ K ^ \widehat{K}\cdot\widehat{K} K ⋅ K .
For T 1 T_1 T 1 , put J = 2 W 2 W ‾ 2 J=2\mathcal{W}^2\overline{\mathcal{W}}^2 J = 2 W 2 W 2 and denote its terms of fluctuation order j j j by J j J_j J j . Similarly write H = H 1 + H 2 + ⋯ H=H_1+H_2+\cdots H = H 1 + H 2 + ⋯ . The terms needed in (124) are
J 0 = 8 f 2 g 2 θ 4 θ ˉ 4 , J 1 = 8 f g 2 Θ θ ˉ 4 ϕ − 8 f 2 g θ 4 Θ ˉ ϕ ˉ , ( J 2 ) ϕ ϕ ˉ = − 8 f g Θ Θ ˉ e − 2 ℓ ϕ ϕ ˉ , [ J 2 ] ∗ = 64 f g Σ . (126) \begin{aligned}
J_0&=8f^2g^2\theta^4\bar{\theta}^4,\\
J_1&=8fg^2\Theta\bar{\theta}^4\phi-8f^2g\theta^4\bar{\Theta}\bar{\phi},\\
(J_2)_{\phi\bar{\phi}}&=-8fg\Theta\bar{\Theta}e^{-2\ell}\phi\bar{\phi},\qquad[J_2]_\ast=64fg\Sigma. \tag*{(126)}
\end{aligned} J 0 J 1 ( J 2 ) ϕ ϕ ˉ = 8 f 2 g 2 θ 4 θ ˉ 4 , = 8 f g 2 Θ θ ˉ 4 ϕ − 8 f 2 g θ 4 Θ ˉ ϕ ˉ , = − 8 f g Θ Θ ˉ e − 2 ℓ ϕ ϕ ˉ , [ J 2 ] ∗ = 64 f g Σ. (126)
Only the 2 ℓ 2 2\ell^2 2 ℓ 2 term of the exponential contributes to the last coefficient. To obtain the denominator terms, expand K ^ = K ^ 0 + K ^ 1 + K ^ 2 \widehat{K}=\widehat{K}_0+\widehat{K}_1+\widehat{K}_2 K = K 0 + K 1 + K 2 . Then
H 1 = 2 K ^ 0 ⋅ K ^ 1 , H 2 = K ^ 1 ⋅ K ^ 1 + 2 K ^ 0 ⋅ K ^ 2 . (127) H_1=2\widehat{K}_0\cdot\widehat{K}_1,\qquad H_2=\widehat{K}_1\cdot\widehat{K}_1+2\widehat{K}_0\cdot\widehat{K}_2. \tag*{(127)} H 1 = 2 K 0 ⋅ K 1 , H 2 = K 1 ⋅ K 1 + 2 K 0 ⋅ K 2 . (127)
Let H ϕ H_{\phi} H ϕ and H ϕ ˉ H_{\bar{\phi}} H ϕ ˉ denote the coefficients of the two lowest-scalar waves in H 1 H_1 H 1 . For the wave of momentum p p p ,
D α i ϕ = − 2 θ ˉ β ˙ p α β ˙ ϕ , D α i D β j ϕ = 4 θ ˉ α ˙ θ ˉ β ˙ p α α ˙ p β β ˙ ϕ , D ˉ α ˙ i W ‾ F = − g ∂ θ ˉ ∂ θ ˉ α ˙ i . (128) D^{i}_{\alpha}\phi=-2\bar{\theta}^{\dot{\beta}}p_{\alpha\dot{\beta}}\phi,
\qquad
D^{i}_{\alpha}D^{j}_{\beta}\phi=4\bar{\theta}^{\dot{\alpha}}\bar{\theta}^{\dot{\beta}}p_{\alpha\dot{\alpha}}p_{\beta\dot{\beta}}\phi,
\qquad
\bar{D}_{\dot{\alpha}i}\overline{W}_{F}=-g\frac{\partial\bar{\theta}}{\partial\bar{\theta}^{\dot{\alpha}i}}. \tag*{(128)} D α i ϕ = − 2 θ ˉ β ˙ p α β ˙ ϕ , D α i D β j ϕ = 4 θ ˉ α ˙ θ ˉ β ˙ p α α ˙ p β β ˙ ϕ , D ˉ α ˙ i W F = − g ∂ θ ˉ α ˙ i ∂ θ ˉ . (128)
Inserting these expressions into (9) gives [ Θ ‾ K ^ 0 ⋅ ( X ϕ , Z ϕ , T ϕ ) ] θ ˉ 4 = 32 f g 2 Σ ( 1 , − 2 i , 8 ) [\overline{\Theta}\widehat{K}_0\cdot(X_{\phi},Z_{\phi},T_{\phi})]_{\bar{\theta}^{4}}=32fg^{2}\Sigma(1,-2i,8) [ Θ K 0 ⋅ ( X ϕ , Z ϕ , T ϕ ) ] θ ˉ 4 = 32 f g 2 Σ ( 1 , − 2 i , 8 ) at θ = 0 \theta=0 θ = 0 ; the subscript ϕ \phi ϕ denotes the coefficient of that wave. Multiplication by twice the coefficients ( − 32 , − 32 i , − 4 ) (-32,-32i,-4) ( − 32 , − 32 i , − 4 ) in K ^ \widehat{K} K gives the components selected by J 1 J_1 J 1 :
[ Θ ‾ H ϕ ] θ ˉ 4 ∣ θ = 0 = ( − 2048 − 4096 − 2048 ) f g 2 Σ = − 8192 f g 2 Σ , [\overline{\Theta}H_{\phi}]_{\bar{\theta}^{4}\mid_{\theta=0}}=(-2048-4096-2048)fg^{2}\Sigma=-8192fg^{2}\Sigma, [ Θ H ϕ ] θ ˉ 4 ∣ θ = 0 = ( − 2048 − 4096 − 2048 ) f g 2 Σ = − 8192 f g 2 Σ ,
[ Θ H ϕ ˉ ] θ 4 ∣ θ ˉ = 0 = 8192 f 2 g Σ . (129) [\Theta H_{\bar{\phi}}]_{\theta^{4}\mid_{\bar{\theta}=0}}=8192f^{2}g\Sigma. \tag*{(129)} [ Θ H ϕ ˉ ] θ 4 ∣ θ ˉ = 0 = 8192 f 2 g Σ. (129)
The three contributions in the first line come from the X X X , Z Z Z and T T T terms of (9), with their coefficients in K ^ \widehat{K} K ; the Y Y Y term vanishes because one field is constant. The second line follows from the barred calculation. The lowest component of ( K ^ 2 ) ϕ ϕ ˉ (\widehat{K}_2)_{\phi\bar{\phi}} ( K 2 ) ϕ ϕ ˉ is 192 p v p w ϕ ϕ ˉ 192p_{v}p_{w}\phi\bar{\phi} 192 p v p w ϕ ϕ ˉ : X X X contributes 64 p v p w 64p_{v}p_{w} 64 p v p w and Y Y Y contributes 128 p v p w 128p_{v}p_{w} 128 p v p w . The linear scalar variations of K ^ \widehat{K} K have positive Grassmann degree, so K ^ 1 ⋅ K ^ 1 \widehat{K}_1\cdot\widehat{K}_1 K 1 ⋅ K 1 has no lowest component. Consequently
( H 2 ) θ = θ ˉ = 0 = 6144 f g Σ ϕ ϕ ˉ , (H_2)_{\theta=\bar{\theta}=0}=6144fg\Sigma\phi\bar{\phi}, ( H 2 ) θ = θ ˉ = 0 = 6144 f g Σ ϕ ϕ ˉ ,
[ J 1 H 1 ] ∗ = 131072 f 3 g 3 Σ , [ J 0 H 2 ] ∗ = 49152 f 3 g 3 Σ , [ J 0 H 1 2 ] ∗ = 0. (130) [J_1H_1]_{*}=131072f^{3}g^{3}\Sigma,\qquad[J_0H_2]_{*}=49152f^{3}g^{3}\Sigma,\qquad[J_0H_1^{2}]_{*}=0. \tag*{(130)} [ J 1 H 1 ] ∗ = 131072 f 3 g 3 Σ , [ J 0 H 2 ] ∗ = 49152 f 3 g 3 Σ , [ J 0 H 1 2 ] ∗ = 0. (130)
The last equality holds because J 0 J_0 J 0 already contains all eight Grassmann coordinates and H 1 H_1 H 1 has no lowest component. Substitution in (124), with n = 1 n=1 n = 1 , gives
[ 2 W 2 W ‾ 2 Q ] ∗ = ( 4 − 16 − 6 ) Σ = − 18 Σ . (131) \left[\frac{2\mathcal{W}^{2}\overline{\mathcal{W}}^{2}}{Q}\right]_{*}=(4-16-6)\Sigma=-18\Sigma. \tag*{(131)} [ Q 2 W 2 W 2 ] ∗ = ( 4 − 16 − 6 ) Σ = − 18Σ. (131)
This calculation retains the two denominator contributions; expanding only the numerator would instead give 4 Σ 4\Sigma 4Σ .
For D 33 D_{33} D 33 , the nilpotency of Θ \Theta Θ gives W F 3 = W ‾ F 3 = 0 \mathcal{W}_{F}^{3}=\overline{\mathcal{W}}_{F}^{3}=0 W F 3 = W F 3 = 0 . The first variations are
A : = ( W 3 ) ( 1 ) = 3 f 2 Θ 2 ϕ = − 6 f 2 θ 4 ϕ , B ˉ : = ( W ‾ 3 ) ( 1 ) = − 6 g 2 θ ˉ 4 ϕ ˉ . (132) A:=(\mathcal{W}^{3})_{(1)}=3f^{2}\Theta^{2}\phi=-6f^{2}\theta^{4}\phi,
\qquad
\bar{B}:=(\overline{\mathcal{W}}^{3})_{(1)}=-6g^{2}\bar{\theta}^{4}\bar{\phi}. \tag*{(132)} A := ( W 3 ) ( 1 ) = 3 f 2 Θ 2 ϕ = − 6 f 2 θ 4 ϕ , B ˉ := ( W 3 ) ( 1 ) = − 6 g 2 θ ˉ 4 ϕ ˉ . (132)
The antichiral translation in B ˉ \bar{B} B ˉ adds no term because θ ˉ 4 \bar{\theta}^{4} θ ˉ 4 already contains all four barred coordinates. There is no c c c or c ˉ \bar{c} c ˉ contribution: multiplication by W F 2 \mathcal{W}_{F}^{2} W F 2 or W ‾ F 2 \overline{\mathcal{W}}_{F}^{2} W F 2 annihilates the corresponding highest scalar component. Thus K 33 K_{33} K 33 begins at second fluctuation order, and
( D 33 ) ( 2 ) = J F [ K ^ 0 ⋅ K 33 [ A , B ˉ ] ( 16 f g ) 3 ] . (133) (D_{33})_{(2)}=\mathcal{J}_{F}\left[\frac{\widehat{K}_0\cdot K_{33}[A,\bar{B}]}{(16fg)^{3}}\right]. \tag*{(133)} ( D 33 ) ( 2 ) = J F [ ( 16 f g ) 3 K 0 ⋅ K 33 [ A , B ˉ ] ] . (133)
There are no denominator variations at this order. Applying (118) to A , B ˉ A,\bar{B} A , B ˉ gives the following six contractions, with the tensor definitions in (44):
Table 134.
Here 576 = 36 × 16 576=36\times16 576 = 36 × 16 : the product A B ˉ A\bar{B} A B ˉ contributes 36 f 2 g 2 36f^{2}g^{2} 36 f 2 g 2 , and contraction of the two momenta with K ^ 0 \widehat{K}_0 K 0 gives 16 f g Σ 16fg\Sigma 16 f g Σ . For example, two derivatives on A A A give − p v p w A -p_{v}p_{w}A − p v p w A , proving the first entry. In Z L Z_L Z L , only the 2 i θ ˉ ∂ 2i\bar{\theta}\partial 2 i θ ˉ ∂ part of D A DA D A can supply four unbarred coordinates; the sum over the two internal indices gives 4 i p v p w 4ip_{v}p_{w} 4 i p v p w . The opposite momentum in Z R Z_{R} Z R reverses this sign. For T T T , both unbarred derivatives must use their 2 i θ ˉ ∂ 2i\bar{\theta}\partial 2 i θ ˉ ∂ parts. The Grassmann identity
θ ˉ A θ ˉ B ∂ C ∂ D θ ˉ 4 = ( δ D A δ C B − δ C A δ D B ) θ ˉ 4 (134) \bar{\theta}^{A}\bar{\theta}^{B}\partial_{C}\partial_{D}\bar{\theta}^{4}=(\delta^{A}_{D}\delta^{B}_{C}-\delta^{A}_{C}\delta^{B}_{D})\bar{\theta}^{4} \tag*{(134)} θ ˉ A θ ˉ B ∂ C ∂ D θ ˉ 4 = ( δ D A δ C B − δ C A δ D B ) θ ˉ 4 (134)
uses combined dotted and internal indices. After summing those internal indices and symmetrizing the tensor indices, its two terms give ( 8 − 16 ) p v p w = − 8 p v p w (8-16)p_{v}p_{w}=-8p_{v}p_{w} ( 8 − 16 ) p v p w = − 8 p v p w . This proves the last entry and hence
[ K ^ ⋅ K 33 Q 3 ] ∗ = 9 64 ( − 48 − 48 − 128 − 64 − 64 − 8 ) Σ = − 405 8 Σ . (135) \left[\frac{\widehat{K}\cdot K_{33}}{Q^{3}}\right]_{*}=\frac{9}{64}(-48-48-128-64-64-8)\,\Sigma=-\frac{405}{8}\Sigma. \tag*{(135)} [ Q 3 K ⋅ K 33 ] ∗ = 64 9 ( − 48 − 48 − 128 − 64 − 64 − 8 ) Σ = − 8 405 Σ. (135)
It remains to check the highest scalar components independently of reduced chirality. For the ϕ c \phi c ϕ c entry of T 1 T_{1} T 1 , the numerator contributes − 8 g 2 θ 4 θ ˉ 4 ϕ c -8g^{2}\theta^{4}\bar{\theta}^{4}\phi c − 8 g 2 θ 4 θ ˉ 4 ϕ c . The degree-two part of the c c c coefficient in H 1 H_{1} H 1 is 256 f g 2 Θ c 256fg^{2}\Theta c 256 f g 2 Θ c . Combining it with 8 f g 2 Θ θ ˉ 4 ϕ 8fg^{2}\Theta\bar{\theta}^{4}\phi 8 f g 2 Θ θ ˉ 4 ϕ in J 1 J_{1} J 1 gives [ J 1 H 1 ] ϕ c = − 4096 f 2 g 4 [J_{1}H_{1}]_{\phi c}=-4096f^{2}g^{4} [ J 1 H 1 ] ϕ c = − 4096 f 2 g 4 . Therefore the two contributions cancel:
( T 1 ) ϕ c / C C = − g 2 f + g 2 f = 0 , ( T 1 ) ϕ ˉ c ˉ / C C = − f 2 g + f 2 g = 0. (136) (T_{1})_{\phi c}/CC=-\frac{g}{2f}+\frac{g}{2f}=0,\qquad(T_{1})_{\bar{\phi}\bar{c}}/CC=-\frac{f}{2g}+\frac{f}{2g}=0. \tag*{(136)} ( T 1 ) ϕ c / C C = − 2 f g + 2 f g = 0 , ( T 1 ) ϕ ˉ c ˉ / C C = − 2 g f + 2 g f = 0. (136)
For ϕ c ˉ \phi\bar{c} ϕ c ˉ and c ϕ ˉ c\bar{\phi} c ϕ ˉ , the numerator vanishes because it contains more than four barred or unbarred coordinates; the same count annihilates J 1 H 1 J_{1}H_{1} J 1 H 1 . The pairs c c cc cc , c ˉ c ˉ \bar{c}\bar{c} c ˉ c ˉ , c ˉ c ˉ \bar{c}\bar{c} c ˉ c ˉ have no numerator contribution, since the square of a highest component and its product with the electromagnetic background vanish. Their H 2 H_{2} H 2 terms have positive Grassmann degree and are annihilated by J 0 J_{0} J 0 . For ϕ ϕ \phi\phi ϕϕ or ϕ ˉ ϕ ˉ \bar{\phi}\bar{\phi} ϕ ˉ ϕ ˉ , the numerator lacks one set of four coordinates, while every denominator contribution exceeds the available degree in the other set. This argument can be made in chiral or antichiral coordinates, respectively, since the total momentum is zero. All these entries of D 33 D_{33} D 33 vanish already by (132) , which contains only one lowest scalar from each chirality.
The complete ten symmetric entries, divided by C C C_{C} C C , are therefore
Table 138.
All background parameters and momentum components were retained in the contractions. The factors of f f f , g g g cancel separately, and the only momentum dependence is the quadratic polynomial Σ \Sigma Σ . The chosen coefficient in (48) follows from − 18 − ( 405 / 8 ) a = 1 -18-(405/8)a=1 − 18 − ( 405/8 ) a = 1 , giving a = − 152 / 405 a=-152/405 a = − 152/405 .
The Maxwell coefficient after imposing reduced chirality is 2 C C P ϕ ϕ ˉ 2C_{C}P\phi\bar{\phi} 2 C C P ϕ ϕ ˉ , fixing z = 2 C C ϕ z=\sqrt{2C_{C}}\phi z = 2 C C ϕ .
Hence
( f m o d ) ( 2 ) = ∫ d 4 x T F μ ν 2 r ∂ μ z ∂ ν z ˉ , ( Δ ) ( 2 ) = 19 ( f m o d ) ( 2 ) . (137) (f_{\mathrm{mod}})^{(2)}=\int d^{4}x\,\frac{T_{F}^{\mu\nu}}{2r}\partial_{\mu}z\partial_{\nu}\bar{z},\qquad(\Delta)^{(2)}=19(f_{\mathrm{mod}})^{(2)}. \tag*{(137)} ( f mod ) ( 2 ) = ∫ d 4 x 2 r T F μν ∂ μ z ∂ ν z ˉ , ( Δ ) ( 2 ) = 19 ( f mod ) ( 2 ) . (137)
The minimal interaction T 1 T_{1} T 1 instead has coefficient − 9 T F μ ν / r -9T_{F}^{\mu\nu}/r − 9 T F μν / r . The additional primary density thus changes the scalar Hessian while leaving the truncation to the N = 1 \mathcal{N}=1 N = 1 vector multiplet unchanged.
Closure under the recursion
The scalar first variations vanish on the background. Reflection excludes mixed scalar–photon variations, and S U ( 2 ) R SU(2)_{R} S U ( 2 ) R excludes scalar–auxiliary-triplet variations, including for the independent scalar components of the unconstrained chiral superfields. Thus (138) gives the complete scalar Hessian. For scalar component variations h h h , the second variation of a pairing is
d 2 P ( F , G ) [ h , h ] = ⟨ d 2 F W [ h , h ] , G W , 0 ⟩ C + 2 ⟨ d F W [ h ] , d G W [ h ] ⟩ C + ⟨ F W , 0 , d 2 G W [ h , h ] ⟩ C , ⟨ A , B ⟩ C = I C [ A B ] . (138) \mathrm{d}^{2}\mathcal{P}(F,G)[h,h]=\langle\mathrm{d}^{2}F_W[h,h],G_{W,0}\rangle_C+2\langle\mathrm{d}F_W[h],\mathrm{d}G_W[h]\rangle_C+\langle F_{W,0},\mathrm{d}^{2}G_W[h,h]\rangle_C,\qquad\langle A,B\rangle_C=\mathcal{I}_C[AB]. \tag*{(138)} d 2 P ( F , G ) [ h , h ] = ⟨ d 2 F W [ h , h ] , G W , 0 ⟩ C + 2 ⟨ d F W [ h ] , d G W [ h ] ⟩ C + ⟨ F W , 0 , d 2 G W [ h , h ] ⟩ C , ⟨ A , B ⟩ C = I C [ A B ] . (138)
Assume the scalar Hessians of F , G F,G F , G are constant multiples of Σ ϕ ϕ ˉ \Sigma\phi\bar{\phi} Σ ϕ ϕ ˉ , including zero. The linear variation of their chiral functional derivatives is then confined to the θ 4 \theta^4 θ 4 component: the lowest scalar variation pairs with θ 4 \theta^4 θ 4 under the chiral measure. The middle term vanishes by ( θ 4 ) 2 = 0 (\theta^4)^2=0 ( θ 4 ) 2 = 0 , as does its barred counterpart.
On this background, U ( 1 ) R U(1)_R U ( 1 ) R , S U ( 2 ) R SU(2)_R S U ( 2 ) R and fermion parity restrict the chiral functional derivatives evaluated on the background to the electromagnetic components:
F W , 0 = b F Θ , G W , 0 = b G Θ . (139) F_{W,0}=b_F\Theta,\qquad G_{W,0}=b_G\Theta. \tag*{(139)} F W , 0 = b F Θ , G W , 0 = b G Θ. (139)
Here b F , b G b_F,b_G b F , b G denote their coefficient functions, not the magnetic amplitude in (120). Let H F [ h , h ] = d 2 F [ h , h ] H_F[h,h]=\mathrm{d}^2F[h,h] H F [ h , h ] = d 2 F [ h , h ] be an independent-chiral scalar Hessian, with the wave directions and their momentum held fixed. Since W 0 = − f Θ W_0=-f\Theta W 0 = − f Θ , a fixed variation δ W = b G Θ \delta W=b_G\Theta δ W = b G Θ is δ f = − b G \delta f=-b_G δ f = − b G . Symmetry of the functional third variation then gives
⟨ d 2 F W [ h , h ] , G W , 0 ⟩ C = − b G ∂ f H F [ h , h ] , d 2 P ( F , G ) [ h , h ] = − b G ∂ f H F − b F ∂ f H G . (140) \left\langle\mathrm{d}^{2}F_W[h,h],G_{W,0}\right\rangle_C=-b_G\partial_f H_F[h,h],\qquad\mathrm{d}^{2}\mathcal{P}(F,G)[h,h]=-b_G\partial_fH_F-b_F\partial_fH_G. \tag*{(140)} ⟨ d 2 F W [ h , h ] , G W , 0 ⟩ C = − b G ∂ f H F [ h , h ] , d 2 P ( F , G ) [ h , h ] = − b G ∂ f H F − b F ∂ f H G . (140)
No derivative of b F b_F b F or b G b_G b G occurs: the third variation is evaluated in a fixed background direction. The antichiral pairing gives the corresponding ∂ g \partial_g ∂ g derivatives, with the sign fixed by W ‾ 0 = g Θ ‾ \overline{W}_0=g\overline{\Theta} W 0 = g Θ . Every derivative vanishes because the full independent-chiral Hessians in (138) are independent of f f f and g g g separately. Together with the vanishing middle term, this proves closure of the scalar Hessian under the pairing. The chiral and antichiral variations have remained independent throughout this calculation.
The independent-chiral scalar Hessian of Δ \Delta Δ has the same form, by (138). Thus each prescribed C n C_n C n contributes a Hessian of this form at the next order.
Every pairing that enters the recursion therefore has zero scalar Hessian if all lower scalar Hessians have the stated form. The phase average commutes with the expansion when the background and fluctuations are both transformed. The undetermined scalar coefficient at each order is phase-invariant and fixed by the prescribed phase-invariant part. Induction gives
( F n ) z ( 2 ) = 0 , ( F ~ n ) z ( 2 ) = C n ( 2 ) = [ t n ] d ( t ) ( Δ ) z ( 2 ) ( n ≥ 2 ) . (141) \boxed{(F_n)^{(2)}_z=0,\qquad(\widetilde{F}_n)^{(2)}_z=C_n^{(2)}=[t^n]\mathrm{d}(t)(\Delta)^{(2)}_z\qquad(n\ge2).} \tag*{(141)} ( F n ) z ( 2 ) = 0 , ( F n ) z ( 2 ) = C n ( 2 ) = [ t n ] d ( t ) ( Δ ) z ( 2 ) ( n ≥ 2 ) . (141)
There is consequently no higher-derivative scalar quadratic operator at any formal order on (91). This conclusion is specific to the stated backgrounds and prescribed coefficients.
Analytic continuation of the quadratic action
For the canonical completion the transverse spatial coefficient vanishes at
γ ∗ = arsinh 2 = log ( 2 + 5 ) (142) \gamma_*=\operatorname{arsinh}2=\log(2+\sqrt{5}) \tag*{(142)} γ ∗ = arsinh 2 = log ( 2 + 5 ) (142)
and becomes negative beyond it. For the action, summing the known quadratic coefficients gives β ( t ) = ( t + 19 d ( t ) ) / 2 = t / ( 1 + t 2 + 1 ) \beta(t)=(t+19\mathrm{d}(t))/2=t/(\sqrt{1+t^2}+1) β ( t ) = ( t + 19 d ( t )) /2 = t / ( 1 + t 2 + 1 ) , and hence (95). The real positive branch specifies an analytic continuation at every finite t ≥ 0 t\ge0 t ≥ 0 , with positive kinetic coefficients and second-order field equations. This second-order scalar operator determines the characteristic cone on these backgrounds. The characteristics of the coupled system on other backgrounds require a separate analysis.
Derivation of the special-supersymmetry variations
We derive the variations of the derivatives and bilinear products used in (15) and (17) . Use the polarized derivatives of (10) , with p y = p p_y=p p y = p in this appendix, and put F i j = D i D j W = − F j i F^{ij}=D^iD^j\mathcal{W}=-F^{ji} F ij = D i D j W = − F j i and G i j = D ˉ i D ˉ j W ‾ = − G j i G_{ij}=\bar{D}_i\bar{D}_j\overline{\mathcal{W}}=-G_{ji} G ij = D ˉ i D ˉ j W = − G j i . The derivative algebra is { D i , D ˉ j } = − 2 i δ j i p \{D^i,\bar{D}_j\}=-2i\delta^i_jp { D i , D ˉ j } = − 2 i δ j i p . The parameters χ i , χ ˉ i \chi^i,\bar{\chi}_i χ i , χ ˉ i and first descendants F i , G i F^i,G_i F i , G i are odd; the variation including its parameters is even.
For pure special supersymmetry, the flat Killing constraints [26 ] give
D i ω y β = − χ i γ β , D i Λ j k = − 2 δ j i χ k + δ j k χ i , D^i\omega_y{}^\beta=-\chi^i\gamma^\beta,\qquad D^i\Lambda_j{}^k=-2\delta^i_j\chi^k+\delta^k_j\chi^i, D i ω y β = − χ i γ β , D i Λ j k = − 2 δ j i χ k + δ j k χ i ,
D ˉ i ω ˉ y β ˙ = − χ ˉ i γ ˉ β ˙ , D ˉ i Λ j k = 2 δ i k χ ˉ j − δ j k χ ˉ i . (143) \bar{D}_i\bar{\omega}_y{}^{\dot{\beta}}=-\bar{\chi}_i\bar{\gamma}^{\dot{\beta}},\qquad\bar{D}_i\Lambda_j{}^k=2\delta_i^k\bar{\chi}_j-\delta_j^k\bar{\chi}_i. \tag*{(143)} D ˉ i ω ˉ y β ˙ = − χ ˉ i γ ˉ β ˙ , D ˉ i Λ j k = 2 δ i k χ ˉ j − δ j k χ ˉ i . (143)
Chirality sets the crossed derivatives of σ , σ ˉ \sigma,\bar{\sigma} σ , σ ˉ and of the Lorentz parameters to zero. The odd parameters are constant, and p p p annihilates the parameter functions. For either fixed internal index h h h (not summed), the Jacobi identity yields
[ ξ , p ] = i 2 ( { [ ξ , D h ] , D ˉ h } + { D h , [ ξ , D ˉ h ] } ) , [ ξ , p ] i n h = − i χ ˉ i D i − i χ i D ˉ i . (144) [\xi,p]=\frac{i}{2}\left(\{[\xi,D^h],\bar{D}_h\}+\{D^h,[\xi,\bar{D}_h]\}\right),\qquad[\xi,p]_{\mathrm{inh}}=-i\bar{\chi}_iD^i-i\chi^i\bar{D}_i. \tag*{(144)} [ ξ , p ] = 2 i ( {[ ξ , D h ] , D ˉ h } + { D h , [ ξ , D ˉ h ]} ) , [ ξ , p ] inh = − i χ ˉ i D i − i χ i D ˉ i . (144)
Indeed, the coefficients of D k , D ˉ k D^k,\bar{D}_k D k , D ˉ k inside the parentheses are − δ h k χ ˉ h − D ˉ h Λ h k = − 2 χ ˉ k -\delta^k_h\bar{\chi}_h-\bar{D}_h\Lambda^k_h=-2\bar{\chi}_k − δ h k χ ˉ h − D ˉ h Λ h k = − 2 χ ˉ k and − δ k h χ h + D h Λ h k = − 2 χ k -\delta^h_k\chi^h+D^h\Lambda^k_h=-2\chi^k − δ k h χ h + D h Λ h k = − 2 χ k .
Differentiating (12) once gives
δ F F j = − ξ F F j + ω y β D β j W − ( 2 σ + σ ˉ ) F F j − Λ j k F k − 2 χ j W . (145) \delta F^{Fj}=-\xi F^{Fj}+\omega_y{}^\beta D^j_\beta\mathcal{W}-(2\sigma+\bar{\sigma})F^{Fj}-\Lambda^k_jF^k-2\chi^j\mathcal{W}. \tag*{(145)} δ F F j = − ξ F F j + ω y β D β j W − ( 2 σ + σ ˉ ) F F j − Λ j k F k − 2 χ j W . (145)
A further D i D^i D i derivative has four inhomogeneous contributions:
( D i ω y β ) D β j W = − χ i F j , − ( D i Λ j k ) F k = − χ i F j + 2 χ j F i , (D^i\omega_y{}^\beta)D^j_\beta\mathcal{W}=-\chi^iF^j,\qquad-(D^i\Lambda^k_j)F^k=-\chi^iF^j+2\chi^jF^i, ( D i ω y β ) D β j W = − χ i F j , − ( D i Λ j k ) F k = − χ i F j + 2 χ j F i ,
− D i ( 2 σ + σ ˉ ) F j = − 2 χ i F j , − 2 D i ( χ j W ) = 2 χ j F i . (146) -D^i(2\sigma+\bar{\sigma})F^j=-2\chi^iF^j,\qquad-2D^i(\chi^j\mathcal{W})=2\chi^jF^i. \tag*{(146)} − D i ( 2 σ + σ ˉ ) F j = − 2 χ i F j , − 2 D i ( χ j W ) = 2 χ j F i . (146)
Their sum is A S ( F i j ) = 4 ( χ j F i − χ i F j ) A_S(F^{ij})=4(\chi^jF^i-\chi^iF^j) A S ( F ij ) = 4 ( χ j F i − χ i F j ) . For the mixed derivative, chirality gives D ˉ j F i = − 2 i δ j i p W \bar{D}_jF^i=-2i\delta^i_jp\mathcal{W} D ˉ j F i = − 2 i δ j i p W , so
A S ( p F i ) = − i χ ˉ j D j F i − i χ j D ˉ j F i − 2 χ i p W = − i χ ˉ j F j i − 4 χ i p W . (147) A_S(pF^i)=-i\bar{\chi}_jD^jF^i-i\chi^j\bar{D}_jF^i-2\chi^i p\mathcal{W}=-i\bar{\chi}_jF^{ji}-4\chi^i p\mathcal{W}. \tag*{(147)} A S ( p F i ) = − i χ ˉ j D j F i − i χ j D ˉ j F i − 2 χ i p W = − i χ ˉ j F j i − 4 χ i p W . (147)
The two contributions to − 4 χ i p W -4\chi^ip\mathcal{W} − 4 χ i p W come from the derivative anticommutator and the last term of (145) . The barred formulas follow by conjugation. One more spacetime derivative gives A S ( p 2 W ) = − 2 i χ ˉ i p F i A_S(p^2\mathcal{W})=-2i\bar{\chi}_ipF^i A S ( p 2 W ) = − 2 i χ ˉ i p F i and its conjugate.
For the products, X = ( p 2 W ) W ‾ + W p 2 W ‾ X=(p^2\mathcal{W})\overline{\mathcal{W}}+\mathcal{W}p^2\overline{\mathcal{W}} X = ( p 2 W ) W + W p 2 W and Y = ( p W ) ( p W ‾ ) Y=(p\mathcal{W})(p\overline{\mathcal{W}}) Y = ( p W ) ( p W ) immediately give the first two entries of (17) . For Z = ( p F i ) G i − F i ( p G i ) Z=(pF^i)G_i-F^i(pG_i) Z = ( p F i ) G i − F i ( p G i ) , moving all odd parameters to the left gives
A S ( Z ) = − 4 χ i ( p W ) G i − i χ ˉ j F j i G i + 2 χ ˉ i ( p F i ) W ‾ + 2 χ i W p G i − 4 χ ˉ i F i p W ‾ − i χ j F i G j i = 2 U S − 4 V S − i M S . (148) \begin{aligned}
A_S(Z)={}&-4\chi^i(p\mathcal{W})G_i-i\bar{\chi}_jF^{ji}G_i+2\bar{\chi}_i(pF^i)\overline{\mathcal{W}}\\
&+2\chi^i\mathcal{W}pG_i-4\bar{\chi}_iF^ip\overline{\mathcal{W}}-i\chi^jF^iG_{ji}
=2U_S-4V_S-iM_S. \tag*{(148)}
\end{aligned} A S ( Z ) = − 4 χ i ( p W ) G i − i χ ˉ j F j i G i + 2 χ ˉ i ( p F i ) W + 2 χ i W p G i − 4 χ ˉ i F i p W − i χ j F i G j i = 2 U S − 4 V S − i M S . (148)
In particular, ( p F i ) ( − 2 χ ˉ i W ‾ ) = + 2 χ ˉ i ( p F i ) W ‾ (pF^i)(-2\bar{\chi}_i\overline{\mathcal{W}})=+2\bar{\chi}_i(pF^i)\overline{\mathcal{W}} ( p F i ) ( − 2 χ ˉ i W ) = + 2 χ ˉ i ( p F i ) W . Antisymmetry gives
A S ( F i j G i j ) = − 8 ( χ i F j G i j + χ ˉ i F i j G j ) = − 8 M S . A_S(F^{ij}G_{ij})=-8(\chi^iF^jG_{ij}+\bar{\chi}_iF^{ij}G_j)=-8M_S. A S ( F ij G ij ) = − 8 ( χ i F j G ij + χ ˉ i F ij G j ) = − 8 M S .
These results establish the coefficient equations in § 3 without a Bianchi identity or field equation.
We now reserve p p p for the chiral weight and write the spacetime derivative as p y p_y p y . For A = W p p A=\mathcal{W}p^p A = W p p , differentiating the weight term replaces − 2 χ i W -2\chi^i\mathcal{W} − 2 χ i W by − 2 p χ i A -2p\chi^iA − 2 p χ i A . The parameter derivatives are unchanged, giving
A S ( D i A ) = − 2 p χ i A , A S ( p y A ) = − i χ ˉ i D i A , A_S(D^iA)=-2p\chi^iA,\qquad A_S(p_yA)=-i\bar{\chi}_iD^iA, A S ( D i A ) = − 2 p χ i A , A S ( p y A ) = − i χ ˉ i D i A ,
A S ( D i D j A ) = 2 ( p + 1 ) ( χ j D i A − χ i D j A ) , A_S(D^iD^jA)=2(p+1)(\chi^jD^iA-\chi^iD^jA), A S ( D i D j A ) = 2 ( p + 1 ) ( χ j D i A − χ i D j A ) ,
A S ( p y D i A ) = − 2 ( p + 1 ) χ i p y A − i χ ˉ j D j D i A , A_S(p_yD^iA)=-2(p+1)\chi^ip_yA-i\bar{\chi}_jD^jD^iA, A S ( p y D i A ) = − 2 ( p + 1 ) χ i p y A − i χ ˉ j D j D i A ,
A S ( p y 2 A ) = − 2 i χ ˉ i p y D i A . (149) A_S(p_y^2A)=-2i\bar{\chi}_ip_yD^iA. \tag*{(149)} A S ( p y 2 A ) = − 2 i χ ˉ i p y D i A . (149)
The conjugate formulas have weight q q q for W ‾ q \overline{W}^{q} W q . Applying these identities to the six products in (44) gives (46) .
Acknowledgments
Use of AI tools. OpenAI GPT-6 Astra and Anthropic Claude Opus 5.5 were used to assist with derivations, symbolic verification codes and editing of the texts. The tools were given the manuscript, calculation notes and specific derivation and verification tasks. Symbolic calculations were also used to check the component identities and quadratic propagation formulas. The derivations are given in the paper. The author takes full responsibility for the content of this paper.