betaFeedback

Reading v1View latest version

WorksPhysics and Astronomy

Can componentwise D7-tadpole cancellation yield a chiral Standard Model? A no-go for smooth O7−^{-} orientifolds

Abstract

We study a componentwise prescription for D7-tadpole cancellation that includes, as a special case, the local cancellation used in several moduli-stabilisation constructions. We work in smooth Type IIB Calabi–Yau orientifolds with O7−^{-} components of pairwise distinct divisor classes. For each component class [Ri][R_i], eight covering-space D7-branes are distributed among smooth irreducible representatives of that class in an orientifold-compatible way. Each component sector may therefore contain several orthogonal, unitary, or symplectic parent factors. We show that these configurations cannot supply the chiral Standard Model through their perturbative D7 spectrum. Without symplectic factors, the entire parent algebra of each component sector embeds into so(8)\mathfrak{so}(8) and cannot contain commuting color and weak-isospin subalgebras. Disjointness excludes massless fermions connecting different component sectors, so quark-doublet fermions are absent. Configurations containing symplectic factors require a separate argument: massless hypercharge, together with the same disjointness, forces the net quark-doublet chirality to vanish. Both arguments allow admissible D7 gauge backgrounds and diagonal gauge embeddings, with no restriction on h−1,1(X)h^{1,1}_{-}(X). We also give a geometric condition under which the tadpole forces purely local cancellation. For a specified finite set of allowed D7 support classes, a separate test identifies component sectors in which the tadpole forces componentwise cancellation. We discuss applications to toric constructions and axion potentials.

Motivation and relation to previous work

We study a componentwise prescription for D7-tadpole cancellation that includes, as a special case, the local cancellation used in moduli-stabilisation constructions and orientifold surveys. Each geometric O7 component RiR_i carries a charge proportional to its divisor class [Ri][R_i]. We assign the compensating charge +8[Ri]+8[R_i] to eight covering-space D7-branes, distributed among smooth representatives of that class together with their orientifold images. All stacks associated with the same class form one component sector. Thus componentwise refers to the charge allocation associated with each O7 component; cancellation need not hold pointwise on the O7 locus. This charge allocation is an additional restriction beyond the general D7-tadpole condition.

Purely local cancellation places all eight branes of each component sector on RiR_i, giving maximal parent algebra ⨁i=1nso(8)i\bigoplus_{i=1}^{n}\mathfrak{so}(8)_i. More general allocations may place some or all of these branes on other representatives of the same class. If the parent factors are all orthogonal or unitary, the complete algebra of each component sector still embeds into so(8)\mathfrak{so}(8), and one group-theoretic argument covers all such configurations. Symplectic factors require a separate argument. They can allow quark-doublet fermions, either within one stack or between stacks of the same component class, but massless hypercharge forces their net chirality to vanish.

The purely local case is the standard “local D7-tadpole cancellation” prescription. Its relation to a constant axio-dilaton is illustrated by the orientifold construction of Ref. [1]. It has appeared as an explicit ingredient in moduli-stabilisation constructions [2, 3, 4, 5]. It has also been used in studies of instanton corrections [6], flux vacua [7], and perturbative corrections [8]. Local cancellation is also adopted in constructions of orientifold geometries [9] and used as a reference configuration in geometric surveys [10, 11, 12]. Several studies compare the local prescription with more general non-local D7 configurations, including Whitney-brane constructions [10, 11, 13, 14]. General accounts of orientifold model building and its relation to F-theory can be found in Refs. [15, 16]. An explicit example with several disjoint O7 components, including both locally cancelled and non-invariant D7 configurations, is discussed in Ref. [17].

The interplay between chiral D7 sectors and nonperturbative moduli stabilisation was studied in Ref. [18]. Ref. [6] adopts purely local cancellation as a simplifying assumption and notes the need to relax it for concrete model building. We ask whether the broader componentwise prescription can supply the chiral Standard Model through its perturbative D7 spectrum after turning on gauge fluxes. Comparing the parent groups with the Standard Model gauge group does not settle this question. Fluxes can break a parent gauge group and generate chiral matter. As an illustration, Ref. [19] studies an O7/D7 stack with flux breaking SO(8)SO(8) to U(4)U(4) and producing chiral antisymmetric matter. The obstruction established here concerns the Standard Model spectrum, not chirality in general.

A quark doublet must carry both color and weak-isospin charges. Within the perturbative D7 spectrum, these charges must either arise within one component sector or from a string connecting two different sectors. For configurations without symplectic factors, we exclude both possibilities: each component sector’s parent algebra embeds into so(8)\mathfrak{so}(8) and cannot contain commuting color and weak-isospin algebras, while different component sectors have disjoint supports. The absence of massless matter connecting disjoint D7 sectors was used explicitly in the SO(8)12SO(8)^{12} construction of Ref. [2]; another construction involving disjoint smooth O7 components appears in Ref. [20]. The required representation-theoretic ingredients are standard; see, for example, Ref. [21].

We define the componentwise prescription and state the obstruction in Sec. II. After establishing the geometric and representation-theoretic ingredients, we present both arguments in Sec. IV. For configurations without symplectic factors, the obstruction follows from the parent gauge algebras and disjointness of different component sectors. For configurations containing symplectic factors, we show that nonzero net quark-doublet chirality would necessarily make hypercharge massive. The detailed derivation of this second argument is given in Appendix A. Both arguments include split allocations and diagonal gauge embeddings across component sectors. We also examine when the tadpole itself fixes the charge allocation. A geometric condition forces purely local cancellation, while a separate test for a finite set of allowed D7 support classes identifies component sectors in which the tadpole forces componentwise cancellation. We then discuss applications to toric hypersurfaces and the consequences of these allocations for axion potentials.

Setup and statement

Let XX be a smooth compact Calabi–Yau threefold with a holomorphic involution σ\sigma satisfying σ∗ΩX=−ΩX\sigma^{*}\Omega_X=-\Omega_X. We work in the perturbative geometric Type IIB orientifold description. Write the divisorial fixed locus and its class as

R=⋃i=1nRi,[R]=∑i=1n[Ri].(1)R=\bigcup_{i=1}^{n}R_i,\qquad[R]=\sum_{i=1}^{n}[R_i]. \tag*{(1)}

where RiR_i are the connected components of the divisorial fixed locus, each supporting an O7−O7^{-} plane. Throughout, “O7 component” refers to the geometric divisor RiR_i, while [Ri][R_i] denotes its cohomology class. These divisors are smooth and mutually disjoint, as explained later. Isolated O3 fixed points, if present, do not enter the argument. We assume that the O7-component classes are pairwise distinct,

[Ri]≠[Rj](i≠j).(2)[R_i]\ne[R_j]\qquad(i\ne j). \tag*{(2)}

We use “component class” for [Ri][R_i] and “D7 representative” for a divisor in that class.

Our componentwise D7-tadpole cancellation prescription is the following. For each ii, choose smooth irreducible divisors EiαE_{i\alpha} with [Eiα]=[Ri][E_{i\alpha}]=[R_i].1 Write Eiα′=σ(Eiα)E'_{i\alpha}=\sigma(E_{i\alpha}). Count each representative together with its image only once, combining coincident choices into one stack. Place NiαN_{i\alpha} D7-branes on EiαE_{i\alpha} and their images on Eiα′E'_{i\alpha}, where

∑αNiα=4,Niα∈Z>0,(3)\sum_{\alpha}N_{i\alpha}=4,\qquad N_{i\alpha}\in\mathbb{Z}_{>0}, \tag*{(3)}

with no additional D7-branes. Since [Ri][R_i] is orientifold-even,

[Eiα′]=[Eiα]=[Ri],[E'_{i\alpha}]=[E_{i\alpha}]=[R_i],
∑αNiα([Eiα]+[Eiα′])=8[Ri].(4)\sum_{\alpha}N_{i\alpha}\bigl([E_{i\alpha}]+[E'_{i\alpha}]\bigr)=8[R_i]. \tag*{(4)}

Summing over ii satisfies the D7-tadpole condition. All stacks and images assigned to [Ri][R_i] form its component sector. The allocation may occupy a single representative and its image, or several such choices within the same class.

The following list describes an individual stack together with its image. Different types may occur within the same component sector.

  1. O7-coincident: Eiα=RiE_{i\alpha}=R_i. The 2Niα2N_{i\alpha} covering-space branes lie on the O7− component and have maximal parent algebra so(2Niα)\mathfrak{so}(2N_{i\alpha}).

  1. Non-invariant: Eiα′≠EiαE'_{i\alpha}\ne E_{i\alpha}. There are NiαN_{i\alpha} branes on each of the two image divisors, giving maximal parent algebra u(Niα)\mathfrak{u}(N_{i\alpha}).

  1. Setwise invariant but distinct from the O7 component: Eiα′=Eiα≠RiE'_{i\alpha}=E_{i\alpha}\ne R_i. The 2Niα2N_{i\alpha} covering-space branes lie on this invariant divisor. With the symplectic Chan–Paton projection, their maximal parent algebra is usp(2Niα)\mathfrak{usp}(2N_{i\alpha}).

There is at most one O7-coincident parent stack in each component sector, since all branes on RiR_i are grouped together. Our notation usp(2N)\mathfrak{usp}(2N) uses the dimension 2N2N of the fundamental representation.

The maximal parent gauge algebra is

gmax⁡=⨁i=1ngi,gi=⨁αgiα,(5)\mathfrak{g}_{\max}=\bigoplus_{i=1}^{n}\mathfrak{g}_i,\qquad\mathfrak{g}_i=\bigoplus_{\alpha}\mathfrak{g}_{i\alpha}, \tag*{(5)}

where giα\mathfrak{g}_{i\alpha} is the algebra associated with the corresponding stack. Here maximal means before breaking by gauge backgrounds. Each component sector contains eight covering-space D7-branes, regardless of how its allocation is split. An unsplit allocation has Ni1=4N_{i1}=4 and gives so(8)\mathfrak{so}(8), u(4)\mathfrak{u}(4), or usp(8)\mathfrak{usp}(8).

Purely local cancellation places all eight branes of every component sector on its O7 component, giving ⨁iso(8)i\bigoplus_i\mathfrak{so}(8)_i. If some branes occupy other representatives, cancellation in that sector holds in cohomology rather than pointwise on RiR_i.

If RiR_i is the unique effective representative of its class, the eight covering-space D7-branes assigned to [Ri][R_i] must lie on RiR_i. If this holds for every ii, only purely local cancellation is possible, with maximal parent algebra ⨁i=1nso(8)i\bigoplus_{i=1}^{n}\mathfrak{so}(8)_i. Rigidity ensures this uniqueness within a linear-equivalence class.2

We allow orientifold-compatible holomorphic gauge bundles and worldvolume fluxes satisfying the relevant quantization and tadpole conditions. We also allow Wilson lines and Higgs backgrounds. Throughout, the D7 supports must remain those specified above, and the surviving gauge generators must lie in the parent algebra (5). We use the usual perturbative geometric open-string spectrum. Transitions to singular configurations with additional light sectors lie outside the setup.

Proposition: Obstruction to a D7 Standard Model: *In the setup above, suppose that color, weak isospin, and hypercharge are embedded in the unbroken D7 gauge algebra, with a massless hypercharge gauge boson before electroweak symmetry breaking. The net chiral index of the massless D7–D7 quark-doublet fermions is then zero: IQ=NQ−NQ‾=0I_Q=N_Q-N_{\overline{Q}}=0, where Q=(3,2)1/6Q=(3,2)_{1/6} and Q‾=(3‾,2)−1/6\overline{Q}=(\overline{3},2)_{-1/6}, and the multiplicities count left-handed Weyl multiplets in these representations.3 The chiral Standard Model, which requires IQ=3I_Q=3, is therefore excluded. If no symplectic factors are present, a stronger statement holds: no massless D7–D7 fermion transforms nontrivially under both color and weak isospin, for any hypercharge assignment.*

The proposition allows diagonal embeddings across component sectors and imposes no condition on h1,1(X)h^{1,1}(X). Finite central quotients of the gauge group do not change the argument.

Section IV presents the obstruction both without and with symplectic factors, including split allocations. Appendix A gives the detailed derivation of the symplectic argument.

Two ingredients

For configurations without symplectic factors, the obstruction follows from two facts. First, the distinct O7-component classes have mutually disjoint smooth D7 representatives in the stated setup. Second, the complete parent algebra of each component sector embeds into so(8)\mathfrak{so}(8), which cannot contain commuting color and weak-isospin algebras. We establish these statements before applying them to the D7 spectrum. The geometric result also enters the hypercharge argument in Section IV B, whose detailed derivation is given in Appendix A.

Disjoint O7 components and their D7 representatives

Let pp be a fixed point of the involution σ\sigma, so that σ(p)=p\sigma(p)=p. Starting with any Riemannian metric gg on XX, the averaged metric (g+σ∗g)/2(g+\sigma^*g)/2 is invariant under σ\sigma. Thus σ\sigma is an isometry of this metric and maps geodesics to geodesics.

Write

Lp:=dσp:TpX⟶TpX(6)L_p := d\sigma_p : T_pX \longrightarrow T_pX \tag*{(6)}

for the differential, or tangent map, of σ\sigma at pp. It describes how σ\sigma transforms an initial velocity: if γ(0)=p\gamma(0)=p and γ′(0)=v\gamma'(0)=v, then Lpv=dσ(γ(t))/dt∣t=0L_pv = \left.d\sigma(\gamma(t))/dt\right|_{t=0}. Because pp is fixed, both the original and transformed velocities belong to the same tangent space TpXT_pX. Moreover, σ2=id\sigma^2=\mathrm{id} implies Lp2=idL_p^2=\mathrm{id}.

The exponential map of the invariant metric therefore satisfies

σ(exp⁡pv)=exp⁡p(Lpv).(7)\sigma(\exp_p v)=\exp_p(L_pv). \tag*{(7)}

Here exp⁡pv\exp_p v is the point reached at parameter time one by the geodesic starting at pp with initial velocity vv.4 The identity says that following this geodesic and then applying σ\sigma gives the same endpoint as first transforming the initial velocity by LpL_p and then following the resulting geodesic.

Choose a sufficiently small ball BB around the origin in TpXT_pX, on which exp⁡p\exp_p is one-to-one. A nearby point exp⁡pv\exp_p v is fixed precisely when Lpv=vL_pv=v. Consequently,

Fix⁡(σ)∩exp⁡p(B)=exp⁡p(B∩ker⁡(Lp−id)).(8)\operatorname{Fix}(\sigma)\cap\exp_p(B)=\exp_p\bigl(B\cap\ker(L_p-\mathrm{id})\bigr). \tag*{(8)}

The space ker⁡(Lp−id)\ker(L_p-\mathrm{id}) is the +1+1 eigenspace of LpL_p. It is a linear subspace of TpXT_pX, so its image under these local coordinates is a smooth submanifold. Since the involution is holomorphic, its fixed components are also complex submanifolds.

Distinct divisorial branches cannot cross: even if each branch is smooth, their union would fail to be a smooth submanifold at a crossing. The divisorial fixed locus therefore consists of smooth, mutually disjoint components:

Ri∩Rj=∅(i≠j).(9)R_i\cap R_j=\varnothing\qquad(i\ne j). \tag*{(9)}

Let EiE_i and EjE_j denote arbitrary individual D7 supports from component sectors ii and jj, including any image supports. Then [Ei]=[Ri][E_i]=[R_i] and [Ej]=[Rj][E_j]=[R_j]. For i≠ji\ne j and a Kähler form JJ,

∫X[Ei]∧[Ej]∧J=∫X[Ri]∧[Rj]∧J=0.(10)\int_X [E_i]\wedge[E_j]\wedge J=\int_X [R_i]\wedge[R_j]\wedge J=0. \tag*{(10)}

By (2), EiE_i and EjE_j are distinct. If they intersected, their intersection would consist of holomorphic curves counted with positive multiplicities. The integral of the Kähler form JJ over each such curve is its positive area. Their intersection would therefore give a positive value in (10), contradicting its vanishing. Thus every support in one component sector is disjoint from every support in another. Representatives within the same component class may intersect, including brane–image pairs and separately chosen stacks. The absence of massless fermions connecting different component sectors follows from the holomorphic description below.

In the holomorphic large-volume description, Ext groups describe the relevant open-string fermion zero modes [24]. Let ιi:Ei→X\iota_i:E_i\to X be the inclusion of a D7 support, and let Vi\mathcal V_i be its gauge bundle. For supports in different component sectors, disjointness gives

Ext⁡Xk(ιi∗Vi,ιj∗Vj)=0for all k,i≠j.(11)\operatorname{Ext}_X^k(\iota_{i*}\mathcal V_i,\iota_{j*}\mathcal V_j)=0 \qquad\text{for all } k,\quad i\ne j. \tag*{(11)}

The same statement applies to the brane images.

Here ιi∗Vi\iota_{i*}\mathcal V_i means the bundle data viewed as a sheaf on XX, supported on EiE_i and zero away from it. Because the supports are disjoint, every point of XX has a neighborhood in which at least one of the two sheaves is zero. The local Ext sheaves therefore vanish, and so do the global Ext groups.

Vanishing in every degree excludes the entire massless fermion sector connecting the two supports, including vectorlike pairs. This holds for any holomorphic gauge bundles on the disjoint supports.5

Color and weak isospin without symplectic factors

Consider a component sector containing only orthogonal and unitary factors. A unitary transformation on NN complex components acts on 2N2N real components and preserves their Euclidean norm. Its real determinant is ∣det⁡U∣2=1|\det U|^2=1, so

u(N)⊂so(2N).(12)\mathfrak{u}(N)\subset\mathfrak{so}(2N). \tag*{(12)}

Applying this inclusion to each unitary factor gives

gi⊆⨁αso(2Niα)⊆so(8).(13)\mathfrak{g}_i\subseteq\bigoplus_\alpha\mathfrak{so}(2N_{i\alpha})\subseteq\mathfrak{so}(8). \tag*{(13)}

The middle algebra acts block diagonally on real vector spaces of dimensions 2Niα2N_{i\alpha}. Together these spaces have dimension ∑α2Niα=8\sum_\alpha2N_{i\alpha}=8, which gives the last inclusion. This argument applies even when the branes occupy different representatives of the same class. It remains to establish the following fact.

Lemma 1. The algebra so(8)\mathfrak{so}(8) has no subalgebra isomorphic to su(3)⊕su(2)\mathfrak{su}(3) \oplus\mathfrak{su}(2).

Suppose that su(3)\mathfrak{su}(3) is embedded in so(8)\mathfrak{so}(8). By definition, so(8)\mathfrak{so}(8) acts faithfully on R8\mathbb{R}^{8} by real antisymmetric matrices.6 Any embedded su(3)\mathfrak{su}(3) therefore acts on this same real eight-dimensional space. We classify the possible color actions and determine which infinitesimal rotations commute with them.

For this purpose, complexify the representation: allow the eight vector coordinates to be complex while retaining the same action matrices. The resulting space VCV_{\mathbb{C}} has complex dimension eight. The complex irreducible representations of su(3)\mathfrak{su}(3) of complex dimension at most eight are 11, 33, 3‾\overline{3}, 66, 6‾\overline{6}, 88. The representations 33 and 66 are inequivalent to their complex conjugates. Since the original action is real, each must appear in VCV_{\mathbb{C}} with the same multiplicity as its conjugate. In particular, a 66 would require an additional 6‾\overline{6}, exceeding the available dimension. The only possible nontrivial decompositions are therefore

VC∣su(3)=3⊕3‾⊕1⊕1,orVC∣su(3)=8.(14)V_{\mathbb{C}}\big|_{\mathfrak{su}(3)} = 3 \oplus\overline{3} \oplus1 \oplus1,\qquad\text{or}\qquad V_{\mathbb{C}}\big|_{\mathfrak{su}(3)} = 8. \tag*{(14)}

The all-trivial action is excluded because the defining representation of so(8)\mathfrak{so}(8) is faithful.

In the first case, the color triplet has three complex components, or equivalently six real components. The remaining two real directions are color singlets. On the triplet, the continuous orthogonal transformations that commute with color are the overall phase rotations, with Lie algebra u(1)\mathfrak{u}(1). The two singlet directions admit an independent ordinary rotation, with Lie algebra so(2)≃u(1)\mathfrak{so}(2) \simeq\mathfrak{u}(1). A transformation commuting with color cannot mix the triplet with the singlets, since they have different color transformation laws. The generators commuting with color therefore form the following algebra, called its centralizer:7

zso(8)(su(3))=u(1)⊕u(1).(15)\mathfrak{z}_{\mathfrak{so}(8)}\bigl(\mathfrak{su}(3)\bigr) = \mathfrak{u}(1) \oplus\mathfrak{u}(1). \tag*{(15)}

In the second case, the real eight-dimensional representation is the adjoint of compact su(3)\mathfrak{su}(3). Concretely, its vectors can be represented by traceless Hermitian 3×33 \times3 matrices HH, with color acting as H↦UHU†H \mapsto UHU^\dagger. There are eight real components, and the action preserves the positive definite inner product tr⁡(HK)\operatorname{tr}(HK).

The complexification of this representation is irreducible. Schur’s lemma states that an operator commuting with all generator matrices on an irreducible complex representation must be proportional to the identity. A commuting real operator is therefore a18a\mathbf{1}_8 with a∈Ra \in\mathbb{R}. To belong to so(8)\mathfrak{so}(8) it must also be antisymmetric, which requires a=0a=0. Thus no nonzero generator in so(8)\mathfrak{so}(8) commutes with this color action.

In the first case the centralizer is abelian, while in the second it vanishes. Neither can contain su(2)\mathfrak{su}(2). Since the two factors in su(3)⊕su(2)\mathfrak{su}(3) \oplus\mathfrak{su}(2) must commute, the claimed subalgebra cannot exist in so(8)\mathfrak{so}(8).

Together with (13), Lemma 1 excludes commuting color and weak-isospin algebras from every component sector without symplectic factors, including arbitrary splittings of its charge allocation.

Proof of the obstruction

We first use the two ingredients established above to exclude massless quark-doublet fermions in configurations without symplectic factors. We then treat configurations containing symplectic factors, where requiring massless hypercharge forces the net quark-doublet chirality to vanish.

Without symplectic factors

Consider any embedding of su(3)⊕su(2)L\mathfrak{su}(3) \oplus\mathfrak{su}(2)_L into gmax⁡=⨁igi\mathfrak{g}_{\max} = \bigoplus_i \mathfrak{g}_i. The embedding may be diagonal: the same color or weak gauge transformation can act simultaneously on several component sectors.

For each component sector, examine the projected color and weak generators in its complete parent algebra gi\mathfrak{g}_i. Because su(3)\mathfrak{su}(3) and su(2)\mathfrak{su}(2) are simple Lie algebras, each projected action is either zero or faithful: a nonzero projection represents the entire corresponding algebra without losing any independent generator.8 If both actions were nonzero, their commuting images would give su(3)⊕su(2)\mathfrak{su}(3) \oplus\mathfrak{su}(2) inside gi\mathfrak{g}_i and hence inside so(8)\mathfrak{so}(8) by (13), contradicting Lemma 1. Therefore at least one of color and weak isospin acts trivially on the entire component sector.

Equation (11) excludes massless fermions connecting different component sectors. Every massless D7–D7 fermion therefore receives its gauge charges from a single sector, although its endpoints may belong to different stacks within that sector. This includes strings between a representative and its image, as well as strings between separately chosen representatives of the same class. Since at least one of color and weak isospin acts trivially throughout the sector, no such fermion can transform nontrivially under both.

Admissible Abelian or non-Abelian gauge fluxes, Wilson lines, and Higgs backgrounds can change the unbroken gauge algebra and matter spectrum. They cannot enlarge the parent algebra or create a massless fermionic sector between disjoint supports. The proof therefore remains valid whenever the supports and parent algebras remain those specified in Sec. II; no classification of gauge backgrounds is needed.

Symplectic factors and massless hypercharge

When symplectic factors are present, color and weak isospin can act nontrivially within the same component sector. Quark doublets may then arise within one stack or between different stacks of that class. The obstruction instead comes from requiring massless hypercharge: the flux needed for nonzero net quark-doublet chirality necessarily gives hypercharge a Stückelberg mass. We explain this connection here and give its detailed derivation in Appendix A.

Consider a component sector ii capable of supplying an SM quark doublet. It has eight covering-space Chan–Paton labels, which are the gauge labels carried by the string endpoints. The color triplet must occur together with its conjugate, using six of these labels. The remaining two form the weak doublet, which must lie in a symplectic parent factor. Appendix A establishes that the endpoint representation has the form

Vi=(3,1)qi⊕(3‾,1)−qi⊕(1,2)0.(16)V_i = (\mathbf{3},\mathbf{1})_{q_i} \oplus(\overline{\mathbf{3}},\mathbf{1})_{-q_i} \oplus(\mathbf{1},\mathbf{2})_0. \tag*{(16)}

The color and image-color blocks have opposite hypercharges. On the weak doublet, an overall phase would fail to preserve the symplectic pairing, so the commuting hypercharge generator vanishes. All eight endpoint components have been used; there is no additional block in this sector that could contribute independently to the hypercharge-masslessness condition.

The charges in (16) belong to individual endpoints. A physical D7–D7 field carries two endpoint indices, and its hypercharge is the difference of their charges. A string joining the color and weak blocks can therefore carry (3,2)qi(\mathbf{3},\mathbf{2})_{q_i} even though the weak endpoint has zero hypercharge. For an SM quark doublet, qi=1/6q_i = 1/6.

Preserving color requires the gauge background to act identically on the three color components. The weak flux is orientifold odd, so its integral against the even restricted class [Ri][R_i] vanishes. Since the color and weak supports both have class [Ri][R_i], the orientifold-projected net quark-doublet index depends on a single color-flux integral:

Ac,i:=∫Ci[Ri]∧fc,i,(17)A_{c,i} := \int_{C_i} [R_i] \wedge f_{c,i}, \tag*{(17)}
IQi=ϵiAc,i.(18)I_{Q_i} = \epsilon_i A_{c,i}. \tag*{(18)}

Here CiC_i supports the color block, fc,if_{c,i} is its dimensionless gauge-invariant flux, including the pulled-back BB field, and ϵi=±1\epsilon_i = \pm1 sets the index convention. Appendix A derives this relation for both coincident and distinct supports, including the projection factor. The fluxes need not be restrictions of bulk flux classes.

The same integral determines the gauging of the C4C_4 axions under hypercharge. In the scalar description, the D7 Chern–Simons couplings make these axions shift under a hypercharge gauge transformation. For an even divisor class ω\omega, the corresponding axion charge is proportional to

KY(ω)=∑S∫Sω∣S∧tr⁡(YSFS).(19)K_Y(\omega) = \sum_S \int_S \omega|_S \wedge\operatorname{tr}(Y_S F_S). \tag*{(19)}

Here the sum runs over distinct occupied D7 support divisors SS in the covering space, including orientifold images. All branes sharing a support are included in its Chan–Paton trace. On each support, YSY_S is the action of the global hypercharge generator and FSF_S is the internal gauge-invariant field strength. The gauged axion kinetic terms give hypercharge a mass whenever any of these charges is nonzero, because the axion kinetic matrix is positive definite. Massless hypercharge therefore requires

KY(ω)=0for every ω∈H+1,1(X).(20)K_Y(\omega) = 0 \qquad\text{for every } \omega\in H^{1,1}_+(X). \tag*{(20)}

Let Ci′=σ(Ci)C'_i = \sigma(C_i) support the image-color block. Its flux satisfies fc′,i=−σ∗fc,if_{c',i} = -\sigma^*f_{c,i}. The color and image-color blocks thus have opposite hypercharges and opposite flux integrals against an even class. Their contributions to KYK_Y add rather than cancel. When tested against [Ri][R_i], each block contributes 3qiAc,i3q_i A_{c,i}, where the factor three counts its color components. The weak block contributes nothing because its hypercharge is zero. The contribution of this component sector is consequently KY,i([Ri])=6qiAc,iK_{Y,i}([R_i]) = 6q_i A_{c,i}.

Now apply (20) to ω=[Ri]\omega= [R_i]. Every D7 support SS in another component sector is disjoint from RiR_i, so [Ri]∣S=0[R_i]|_S = 0 in H2(S,R)H^2(S,\mathbb{R}). All other component sectors therefore drop out of this particular test. Using (17), we obtain

0=KY([Ri])=KY,i([Ri])=6qiAc,i=6qiϵiIQi.(21)\begin{aligned} 0 = K_Y([R_i]) = K_{Y,i}([R_i]) \\ &= 6q_i A_{c,i} = 6q_i \epsilon_i I_{Q_i}. \tag*{(21)} \end{aligned}

Since an SM quark doublet has qi≠0q_i \ne0, its sector must have IQi=0I_{Q_i} = 0. Nonzero net quark-doublet chirality would necessarily make hypercharge massive. Mixing hypercharge with generators in other component sectors cannot cancel this contribution, because their contributions to KY([Ri])K_Y([R_i]) vanish separately. The argument uses disjointness and does not require linear independence of the component classes.

By (11), every massless D7–D7 fermion belongs to a single component sector. Sectors without symplectic factors cannot supply quark doublets, while every sector that can supply SM quark doublets has zero net index by (21). Hence

IQ=∑iIQi=0.(22)I_Q = \sum_i I_{Q_i} = 0. \tag*{(22)}

This completes the proposition for split and unsplit allocations, mixed products, and diagonal embeddings. Massless vectorlike quark-doublet pairs are allowed, but the net chirality required by the Standard Model is excluded.

Scope and consequences

For configurations without symplectic factors, the obstruction is already the absence of even one massless quark-doublet fermion, independently of hypercharge. Configurations containing symplectic factors can allow such fermions, but massless hypercharge forces their net chirality to vanish. The Standard Model instead requires three net quark doublets. Thus both arguments obstruct the required spectrum before any analysis of realistic Yukawa couplings. A product of parent factors may contain the Standard Model gauge algebra, and fluxed stacks may support chiral matter, without supplying the chiral Standard Model spectrum.

Euclidean D3-brane (ED3) fluxes affect instanton actions and zero modes [26]. An ED3 intersecting several D7 sectors may carry charged zero modes for each, but these are integrated over in the instanton amplitude and are not propagating four-dimensional matter. Such instantons can generate operators involving existing fields; they do not supply new perturbative D7 fermions. Mass terms preserving the Standard Model gauge symmetry remove conjugate pairs and cannot change the net quark-doublet index. No assumption that ED3 contributions vanish is needed.

Neither argument requires any restriction on h−1,1(X)h^{1,1}_{-}(X). Smoothness gives disjoint O7 components regardless of the number of orientifold-odd classes. The hypercharge argument tests masslessness against the even class of an individual O7 component, so additional odd classes do not relax the resulting constraint. Equality of divisor classes does not imply equality of supports: distinct brane–image divisors can represent the same class even when h−1,1(X)=0h^{1,1}_{-}(X)=0.

The prescription assigns eight covering-space D7-branes to each class [Ri][R_i], including their orientifold images. Their supports must be smooth irreducible representatives of that class. All distributions satisfying (3) are included. Redistributing charge between component classes, using other divisor classes, adding further D7 branes, or allowing singular supports requires a separate analysis. Nonperturbative realizations of the visible sector also lie outside the proposition.

A geometric condition forcing local cancellation

A sufficient geometric condition forces purely local cancellation without imposing the componentwise prescription. Suppose there exists a Kähler form JJ such that

∫X[Ri]2∧J<0for every i.(23)\int_X [R_i]^2 \wedge J < 0 \qquad\text{for every } i. \tag*{(23)}

For any D7 configuration with effective holomorphic supports, let mi∈Z≥0m_i \in\mathbb{Z}_{\geq0} denote the covering-space D7 multiplicity on the actual O7 component RiR_i. Let SaS_a denote the remaining irreducible components, with covering-space multiplicities na∈Z>0n_a \in\mathbb{Z}_{>0}. Thus Sa≠RiS_a \ne R_i for every aa and ii. Define their total charge class by

[E]:=∑ana[Sa].[E] := \sum_a n_a [S_a].

The total covering-space D7 charge is therefore

[DD7]=∑imi[Ri]+[E].[D_{D7}] = \sum_i m_i [R_i] + [E].

Pairing the general tadpole condition [DD7]=8∑i[Ri][D_{D7}] = 8 \sum_i [R_i] with [Ri]∧J[R_i]\wedge J, and using disjointness of the O7 components, gives

(8−mi)∫X[Ri]2∧J=∫X[E]∧[Ri]∧J≥0.(8-m_i)\int_X [R_i]^2 \wedge J = \int_X [E]\wedge[R_i]\wedge J \geq0.

Indeed, each SaS_a and RiR_i have no common divisor component. Their intersection consists of holomorphic curves with positive multiplicities, or is empty. Every nonempty such curve has positive area with respect to the Kähler form JJ, so the integral on the right is nonnegative.9 The negativity in Eq. (23) therefore forces mi≥8m_i \geq8 for every ii.

Pairing the remaining tadpole ∑i(mi−8)[Ri]+[E]=0\sum_i(m_i-8)[R_i]+[E]=0 with J2J^2 then forces mi=8m_i=8 for every ii and the remaining sum over SaS_a to be empty: all coefficients are nonnegative, and every nonzero effective support has strictly positive Kähler volume. Thus the tadpole itself requires eight covering-space D7-branes on each RiR_i. The maximal parent algebra is ⨁iso(8)i\bigoplus_i \mathfrak{so}(8)_i, and the orthogonal-case obstruction applies.

Only the existence of one such Kähler form JJ is needed. Any proposed configuration at another Kähler point obeys the same cohomological tadpole and can be tested with JJ. The conclusion therefore holds throughout the Kähler cone of the fixed complex manifold XX.

We will apply this criterion to explicit orientifold geometries in a future work and identify the geometries it excludes from admitting a perturbative D7 Standard Model.

D7 allocations forced by the tadpole

The preceding geometric criterion forces purely local cancellation. We now ask what the tadpole forces when the allowed D7 support classes belong to a specified finite set. Let [Dα][D_\alpha], with α=1,…,m\alpha=1,\ldots,m, denote the distinct allowed classes, including the classes of all orientifold images. Each class must admit a smooth irreducible holomorphic representative, but need not be a prime toric class. Membership in the effective cone alone does not guarantee the required representatives [27]. All D7 configurations considered in this test use these classes. Also, for this allocation analysis, we allow different O7 components to have the same class.

Choose a real basis [HA][H_A] of divisor classes, with A=1,…,h1,1(X)A=1,\ldots,h^{1,1}(X), consisting of orientifold-even and orientifold-odd elements. Write

[Dα]=∑A=1h1,1(X)dαA[HA],[D_\alpha]=\sum_{A=1}^{h^{1,1}(X)}d_\alpha^A[H_A],
[Ri]=∑A=1h1,1(X)ρiA[HA].(24)[R_i]=\sum_{A=1}^{h^{1,1}(X)}\rho_i^A[H_A]. \tag*{(24)}

Here dαAd_\alpha^A and ρiA\rho_i^A are the coefficients of the classes in this basis.

Let xαx_\alpha count covering-space D7-branes in class [Dα][D_\alpha], including images and all occupied representatives. The tadpole gives one equation for each basis direction:

∑α=1mxαdαA=8∑i=1nρiA,A=1,…,h1,1(X),\sum_{\alpha=1}^{m}x_\alpha d_\alpha^A=8\sum_{i=1}^{n}\rho_i^A,\qquad A=1,\ldots,h^{1,1}(X),
xα≥0,α=1,…,m.(25)x_\alpha\geq0,\qquad\alpha=1,\ldots,m. \tag*{(25)}

The sum over ii includes every O7 component, so repeated O7 classes contribute with their component multiplicities. Since every [Ri][R_i] is orientifold even, ρiA=0\rho_i^A=0 in the odd basis directions. The corresponding equations therefore require zero total odd D7 charge.

To determine which allocations the tadpole forces, initially allow real nonnegative brane multiplicities xαx_\alpha. If (25) has no real solution, the allowed classes cannot cancel the tadpole. Otherwise, pairing the tadpole with J2J^2 fixes the total D7 volume. Since every allowed class has positive Kähler volume, each xαx_\alpha is bounded above. For each class, find

xαmin⁡=min⁡xα,xαmax⁡=max⁡xα,(26)x_\alpha^{\min}=\min x_\alpha,\qquad x_\alpha^{\max}=\max x_\alpha, \tag*{(26)}

over all real vectors xx satisfying (25).

A zero minimum means that a real allocation can omit that class. A positive minimum means that some charge in the class is compulsory. Equal minimum and maximum values mean that its multiplicity is fixed. Physical allocations must additionally obey integer multiplicities and brane–image counting. If the real bounds already fix a coefficient to eight, every allowed physical allocation also fixes it to eight.

Consider an O7 component whose class occurs among the allowed D7 classes and differs from every other O7-component class. Let α(i)\alpha(i) be the index for which [Dα(i)]=[Ri][D_{\alpha(i)}]=[R_i]. Let F\mathcal{F} contain all such indices ii for which

xα(i)min⁡=xα(i)max⁡=8.(27)x_{\alpha(i)}^{\min}=x_{\alpha(i)}^{\max}=8. \tag*{(27)}

Thus every allowed physical allocation contains exactly eight covering-space D7-branes in class [Ri][R_i] for each i∈Fi\in\mathcal{F}.

For any such configuration, let [Ta][T_a] label the remaining occupied D7 classes, with equal classes combined, and let PaP_a be their positive integer multiplicities. Subtracting 8[Ri]8[R_i] for each i∈Fi\in\mathcal{F} gives

∑aPa[Ta]=8∑j∉F[Rj],Pa∈Z>0.(28)\sum_a P_a[T_a]=8\sum_{j\notin\mathcal{F}}[R_j],\qquad P_a\in\mathbb{Z}_{>0}. \tag*{(28)}

For i∈Fi\in\mathcal{F}, pairing the right-hand side with [Ri]∧J[R_i]\wedge J gives zero because the O7 components are disjoint. Every remaining class [Ta][T_a] differs from [Ri][R_i], so its smooth irreducible support has proper intersections with RiR_i. The left-hand side is therefore a sum of nonnegative intersection integrals, and each integral must vanish. Every remaining D7 support is consequently disjoint from RiR_i:

Ta∩Ri=∅(i∈F).(29)T_a\cap R_i=\varnothing\qquad(i\in\mathcal{F}). \tag*{(29)}

Here TaT_a denotes any occupied representative of the class [Ta][T_a]. The same argument applies to every D7 representative of [Ri][R_i]: replacing RiR_i by such a representative leaves the intersection integrals unchanged. Thus the component sectors with i∈Fi\in\mathcal{F} are geometrically sequestered from the remaining D7 stacks: every support in these sectors is disjoint from every remaining D7 support. The argument of the proposition then shows that, in every component sector with i∈Fi\in\mathcal{F}, massless hypercharge is incompatible with nonzero net Standard Model quark-doublet chirality.

Any net chiral Standard Model quark doublets must therefore arise from the remaining stacks, whose charges must satisfy (28). If every O7 component belongs to F\mathcal{F}, the right-hand side vanishes. Pairing with J2J^2 forces the remaining sum to be empty, since every occupied support has positive Kähler volume. The entire configuration is then componentwise, and the no-go proposition applies.

The allocation test imposes no restriction on the number of smooth irreducible representatives of the allowed classes. If RiR_i is the unique effective representative of its class for some i∈Fi\in\mathcal{F}, its eight covering-space D7-branes must lie on the actual component RiR_i. That sector is purely local, with maximal parent algebra so(8)\mathfrak{so}(8).

A useful application is to favourable Calabi–Yau hypersurfaces obtained from four-dimensional reflexive polytopes in the Kreuzer–Skarke database. Their relevant coordinate divisors are DI=X∩{xI=0}D_I=X\cap\{x_I=0\}, with I=1,…,h1,1(X)+4I=1,\ldots,h^{1,1}(X)+4 [28, 29]. An inherited toric involution preserving XX can give an O7−^- orientifold with connected O7 components RiR_i. We restrict here. to smooth orientifold-compatible Calabi–Yau hypersurfaces, as required by our setup. An O7 component can arise from an ambient fixed divisor intersected with XX, or from an ambient fixed surface contained in XX, such as D^a∩D^b⊂X\widehat{D}_a \cap\widehat{D}_b \subset X [30]. Its class is orientifold even but need not coincide with an individual coordinate-divisor class [DI][D_I]. These constructions provide the divisor-class data needed for the allocation test. The allowed classes may include coordinate-divisor classes and additional classes admitting the required smooth irreducible representatives. Prime toric divisors are used as candidate QCD supports in statistical studies of these geometries [31, 32, 33].

These forced D7 allocations can also affect axion potentials. Statistical axiverse studies often model non-QCD axion potentials using ED3 contributions on prime toric divisors, with suppression e−2πτDe^{-2\pi\tau_D} [31, 32, 33]. When a component sector with i∈Fi \in\mathcal{F} is purely local and its low-energy theory is pure N=1\mathcal{N}=1 SO(8)SO(8) Yang–Mills, gaugino condensation generates Wnp,i∝e−2πTi/6W_{\mathrm{np},i} \propto e^{-2\pi T_i/6}, where Ti=τi+iθiT_i=\tau_i+i\theta_i [4]. The weaker exponential suppression can increase the associated axion masses relative to ED3-only estimates. Gauge backgrounds can modify the gauge kinetic function and the light matter content, changing the resulting nonperturbative contributions. We will examine these implications for axion spectra in a separate work [34].

Acknowledgments

This work is supported by the U.S. Department of Energy (DOE) under Grant No. DE-SC0010103. I am thankful to Federico Carta, Michele Cicoli, Jim Halverson, and Jakob Moritz for useful comments and questions.

Configurations with symplectic factors and massless hypercharge

This appendix gives the detailed derivation of the symplectic argument presented in Sec. IV B. We first determine the allowed Standard Model charges of the eight covering-space Chan–Paton endpoint components in each relevant sector. We then derive the orientifold-projected quark-doublet index and its relation to the flux-induced hypercharge-masslessness condition. The argument covers quark doublets arising within one stack or between different stacks of the same component class.

We assume that the Standard Model generators survive the D7 gauge backgrounds and regard their unbroken algebra as embedded in the parent algebra (5). We allow arbitrary diagonal embeddings:

φ:su(3)C⊕su(2)L⊕u(1)Y↪gmax⁡,φi:=πi∘φ,(30)\varphi:\mathfrak{su}(3)_{C}\oplus\mathfrak{su}(2)_{L}\oplus\mathfrak{u}(1)_{Y}\hookrightarrow\mathfrak{g}_{\max},\qquad \varphi_i:=\pi_i\circ\varphi, \tag*{(30)}

where πi\pi_i projects onto the complete algebra gi\mathfrak{g}_i of component sector ii, including all its parent factors. The projections preserve the commutation relations. In particular, the projected color, weak-isospin, and hypercharge subalgebras commute with one another.

Color, weak isospin, and hypercharge in one component sector

Consider a component sector capable of supplying a quark doublet or its conjugate. Both color and weak isospin must act nontrivially on it. Let VV be the direct sum of its covering-space Chan–Paton endpoint spaces, including images. It has complex dimension eight, whether the branes occupy one support or several.

The representation on VV is equivalent to its complex conjugate. Orthogonal and symplectic endpoint representations have this property individually, while each unitary endpoint representation occurs with its conjugate on the image stack. Consequently, a color representation inequivalent to its conjugate must occur together with that conjugate. A 66 would require an additional 6‾\overline{6}, exceeding eight dimensions. An irreducible color adjoint 88 would use all eight Chan–Paton endpoint components allowed by the componentwise allocation. Schur’s lemma would then leave only scalar matrices commuting with color, which cannot realize a nonzero weak su(2)\mathfrak{su}(2). The only possibility is therefore

V∣su(3)=3⊕3‾⊕1⊕1.(31)V\big|_{\mathfrak{su}(3)}=3\oplus\overline{3}\oplus1\oplus1. \tag*{(31)}

The color triplet and antitriplet each occur once. Transformations commuting with color can therefore act on each only by an overall phase. Weak isospin must act on the two remaining color-singlet components, which form one doublet.

This doublet must lie in a symplectic parent factor. An orthogonal action on the two remaining components provides only so(2)\mathfrak{so}(2), which cannot contain su(2)\mathfrak{su}(2). A doublet on a non-invariant unitary stack would require another doublet on its image, using four color-singlet components rather than the two available.

The possible parent algebras of a component sector supporting both color and weak isospin are therefore:

usp(8),so(6)⊕usp(2),u(3)⊕usp(2),usp(6)⊕usp(2).(32)\mathfrak{usp}(8),\qquad\mathfrak{so}(6)\oplus\mathfrak{usp}(2),\qquad \mathfrak{u}(3)\oplus\mathfrak{usp}(2),\qquad\mathfrak{usp}(6)\oplus\mathfrak{usp}(2). \tag*{(32)}

The first algebra uses all four brane–image pairs in one stack. The other three use a 3+13+1 split. Any multiplicity partition other than 44 or 3+13+1 has Niα≤2N_{i\alpha}\leq2 for every stack, and none of its parent factors can contain su(3)\mathfrak{su}(3). A diagonal embedding cannot evade this: its nonzero projection onto any parent factor would still have to embed su(3)\mathfrak{su}(3) into that factor. These decompositions agree with the standard subgroup tables [21].

Although the parent algebras in (A3) differ, their combined endpoint spaces have the same eight-dimensional decomposition under color and weak isospin. We now determine the allowed hypercharge action on these endpoint components. This specifies the charges carried by individual string endpoints; physical D7–D7 fields carry two such indices, and their charges will be obtained below.

A hypercharge generator commuting with color and weak isospin acts by a scalar on each irreducible endpoint block. The color and image-color blocks have opposite eigenvalues. On the weak doublet, an overall phase does not preserve the symplectic pairing, so the commuting hypercharge generator vanishes. The joint decomposition and hypercharge action are therefore

V∣φi(gSM)=(3,1)qi⊕(3‾,1)−qi⊕(1,2)0,(33)V|_{\varphi_i(\mathfrak{g}_{\mathrm{SM}})}=(\mathbf{3},\mathbf{1})_{q_i}\oplus(\overline{\mathbf{3}},\mathbf{1})_{-q_i}\oplus(\mathbf{1},\mathbf{2})_0, \tag*{(33)}

where gSM=su(3)C⊕su(2)L⊕u(1)Y\mathfrak{g}_{\mathrm{SM}}=\mathfrak{su}(3)_{\mathrm{C}}\oplus\mathfrak{su}(2)_{\mathrm{L}}\oplus\mathfrak{u}(1)_{\mathrm{Y}}. In this basis, the hypercharge generator is

Yi=qidiag⁡(13,−13,02).(34)Y_i=q_i\operatorname{diag}(\mathbf{1}_3,-\mathbf{1}_3,\mathbf{0}_2). \tag*{(34)}

All eight endpoint components have been used. There is no additional block in this sector that could contribute independently to the hypercharge-masslessness condition.

To obtain the physical matter representations, we must combine the two endpoint indices. A field represented by a map from one endpoint space to another transforms in the target representation and the dual of the source representation. Its hypercharge is therefore the target charge minus the source charge. In particular, a color endpoint of charge qiq_i and a weak endpoint of charge zero can give a quark doublet of hypercharge qiq_i.

Write the endpoint space as V=Vc⊕Vc′⊕VwV=V_c\oplus V_{c'}\oplus V_w, with dimensions 3, 3, and 2, respectively. The two ordered Chan–Paton blocks carrying quark-doublet charges are

Φ:Vw⟶Vc,\Phi:V_w\longrightarrow V_c,
Φ~:Vc′⟶Vw.(35)\widetilde{\Phi}:V_{c'}\longrightarrow V_w. \tag*{(35)}

Thus Φ\Phi is a 3×23\times2 matrix and Φ~\widetilde{\Phi} is a 2×32\times3 matrix. Under color and weak transformations U∈SU(3)CU\in SU(3)_{\mathrm{C}} and W∈SU(2)LW\in SU(2)_{\mathrm{L}},

Φ⟼UΦW−1,Φ~⟼WΦ~UT.(36)\Phi\longmapsto U\Phi W^{-1},\qquad\widetilde{\Phi}\longmapsto W\widetilde{\Phi}U^{T}. \tag*{(36)}

The source space Vc′V_{c'} transforms by U∗U^*, so the source index of a map transforms by the inverse matrix (U∗)−1=UT(U^*)^{-1}=U^T. This inverse action is what is meant by transforming in the dual of the source. For SU(2)SU(2), a doublet and its complex conjugate are equivalent representations.10 Both blocks therefore carry (3,2)(\mathbf{3},\mathbf{2}).

The corresponding hypercharge actions are

Yc,iΦ−ΦYw,i=qiΦ,Y_{c,i}\Phi-\Phi Y_{w,i}=q_i\Phi,
Yw,iΦ~−Φ~Yc′,i=qiΦ~.(37)Y_{w,i}\widetilde{\Phi}-\widetilde{\Phi}Y_{c',i}=q_i\widetilde{\Phi}. \tag*{(37)}

where Yc,i=qi13Y_{c,i}=q_i\mathbf{1}_3, Yc′,i=−qi13Y_{c',i}=-q_i\mathbf{1}_3, and Yw,i=02Y_{w,i}=\mathbf{0}_2. The zero hypercharge of the weak endpoint block therefore does not imply zero hypercharge for a physical weak doublet: the two endpoint charges enter with opposite signs. Both blocks in (A6) carry (3,2)qi(\mathbf{3},\mathbf{2})_{q_i}; the reversed maps carry the conjugate representation.

Here 3\mathbf{3} is defined relative to the global color algebra in (A1). A sector supplying Q=(3,2)1/6Q=(\mathbf{3},\mathbf{2})_{1/6} must consequently have qi=1/6q_i=1/6. The obstruction below needs only qi≠0q_i\ne0.

The quark-doublet index

Let CiC_i support the color triplet, let Ci′=σ(Ci)C'_i=\sigma(C_i) support its image, and let Wi=σ(Wi)W_i=\sigma(W_i) support the weak doublet. All three supports have class [Ri][R_i]. They may coincide or be distinct; the following argument covers both possibilities.

A gauge background preserving the projected color and weak groups must act identically on the components within each irreducible block of (A4). Since each block occurs once, its internal field strength has the form

Fc,i=fc,i13,Fc′,i=fc′,i13,\mathcal{F}_{c,i}=f_{c,i}\mathbf{1}_3,\qquad\mathcal{F}_{c',i}=f_{c',i}\mathbf{1}_3,
Fw,i=fw,i12.(38)\mathcal{F}_{w,i}=f_{w,i}\mathbf{1}_2. \tag*{(38)}

The ff’s denote dimensionless gauge-invariant flux classes, including the pulled-back BB field, on their respective supports. This form follows from preserving the gauge generators; it is not an additional choice of Abelian flux ansatz.

The orientifold exchanges the color and image-color blocks and preserves the weak block. Its action on the gauge field gives [35]

fc′,i=−σ∗fc,i,fw,i=−σ∗fw,i.(39)f_{c',i}=-\sigma^*f_{c,i},\qquad f_{w,i}=-\sigma^*f_{w,i}. \tag*{(39)}

The first pullback uses the map σ:Ci′→Ci\sigma:C'_i\to C_i, and the second uses σ:Wi→Wi\sigma:W_i\to W_i. Define

Ac,i:=∫[Ri]∣Ci∧fc,i,A_{c,i}:=\int_{[R_i]|_{C_i}}\wedge f_{c,i},
Ac′,i:=∫[Ri]∣Ci′∧fc′,i,A_{c',i}:=\int_{[R_i]|_{C'_i}}\wedge f_{c',i},
Aw,i:=∫[Ri]∣Wi∧fw,i.(40)A_{w,i}:=\int_{[R_i]|_{W_i}}\wedge f_{w,i}. \tag*{(40)}

The notation [Ri]∣S[R_i]|_S means the bulk divisor class restricted to the support SS. Since [Ri][R_i] is orientifold-even and σ\sigma preserves the complex orientations of these supports, (A10) implies

Ac′,i=−Ac,i,Aw,i=0.(41)A_{c',i}=-A_{c,i},\qquad A_{w,i}=0. \tag*{(41)}

In particular, the integral of the odd weak flux against the even restricted class vanishes.

The ordinary D7 bifundamental index applies both to intersecting supports and to coincident gauge blocks; in the latter case it contains the divisor self-intersection [25]. For distinct color and weak supports, the color-flux contribution is ∫Ci[Wi]∣Ci∧fc,i=Ac,i\int_{C_i}[W_i]|_{C_i}\wedge f_{c,i}=A_{c,i}, because [Wi]=[Ri][W_i]=[R_i]. The weak-flux contribution similarly equals Aw,iA_{w,i}. The same expressions apply when the supports coincide.

For an ordered block a←ba\leftarrow b, meaning a map Vb→VaV_b\to V_a, the source index transforms with the inverse gauge transformation. The two endpoints therefore couple with opposite signs to their gauge backgrounds. Before the orientifold projection, the chiral indices of the blocks in (A6) are

Icw=ϵi(Ac,i−Aw,i),(42)I_{cw}=\epsilon_i\left(A_{c,i}-A_{w,i}\right), \tag*{(42)}
Iwc′=ϵi(Aw,i−Ac′,i).I_{wc'}=\epsilon_i\left(A_{w,i}-A_{c',i}\right).

Here the indices count copies of the full (3,2)(\mathbf{3},\mathbf{2}) representation, and ϵi=±1\epsilon_i=\pm1 sets the common index convention.

The orientifold reverses the endpoint order and replaces each endpoint by its image: (a←b)↦(b′←a′)(a\leftarrow b)\mapsto(b'\leftarrow a'). Since w′=ww'=w and c,c′c,c' are exchanged, it maps c←wc\leftarrow w to w←c′w\leftarrow c' and vice versa. Both blocks carry the same quark-doublet charges.

The projection identifies the modes in these two blocks and keeps one copy of each paired mode. The physical index is therefore half the sum of the two unprojected indices [35]:

IQi=12(Icw+Iwc′).(43)I_{Q_i}=\frac{1}{2}\left(I_{cw}+I_{wc'}\right). \tag*{(43)}

The two blocks are exchanged rather than individually projected onto symmetric or antisymmetric representations. There is therefore no additional O7 correction for this pair of blocks.

Using (A12), both ordinary indices equal ϵiAc,i\epsilon_i A_{c,i}. Consequently,

IQi=ϵiAc,i.(44)I_{Q_i}=\epsilon_i A_{c,i}. \tag*{(44)}

The fluxes need not be restrictions of flux classes defined on the bulk space XX.

The massless-hypercharge condition

Equation (A15) expresses the net quark-doublet chirality in terms of a single flux integral, Ac,iA_{c,i}. The same integral determines this sector’s flux-induced hypercharge coupling. We will use disjointness to show that a nonzero value necessarily makes hypercharge massive when qi≠0q_i\ne0.

The four-dimensional axions θα\theta_\alpha arise from the scalar components of the RR four-form:

C4=∑αθαω~α+⋯ ,ω~α∈H+2,2(X).(45)C_4=\sum_\alpha\theta_\alpha\widetilde{\omega}^{\alpha}+\cdots,\qquad\widetilde{\omega}^{\alpha}\in H_+^{2,2}(X). \tag*{(45)}

Here the ω~α\widetilde{\omega}^{\alpha} form a basis of the even four-form classes. Let ωα∈H+1,1(X)\omega_\alpha\in H_+^{1,1}(X) denote the dual basis, normalized by

∫Xωα∧ω~β=δαβ.(46)\int_X\omega_\alpha\wedge\widetilde{\omega}^{\beta}=\delta_\alpha^{\beta}. \tag*{(46)}

Dimensional reduction of the D7 Chern–Simons couplings, together with the self-duality of the RR five-form field strength, gauges the shifts of these axions [36, 37]. The charges are determined by the flux integrals

KY(ω):=∑S∫SιS∗ω∧tr⁡(YSFS),(47)K_Y(\omega):=\sum_S\int_S\iota_S^*\omega\wedge\operatorname{tr}(Y_S\mathcal{F}_S), \tag*{(47)}

where SS denotes an occupied D7 support divisor and ιS:S↪X\iota_S:S\hookrightarrow X is its inclusion. The sum runs over the distinct supports in the covering space, including orientifold images. All branes sharing the same support are combined into one Chan–Paton bundle on SS. Here YSY_S is the action of the global hypercharge generator on this bundle, and FS\mathcal{F}_S is its internal gauge-invariant field strength. The trace includes all Chan–Paton components on SS.

Under a hypercharge gauge transformation with parameter λY\lambda_Y, the fields transform as

AYμ↦AYμ+∂μλY,A_{Y\mu}\mapsto A_{Y\mu}+\partial_\mu\lambda_Y,
θα↦θα+kαλY,kα∝KY(ωα).(48)\theta_\alpha\mapsto\theta_\alpha+k_\alpha\lambda_Y,\qquad k_\alpha\propto K_Y(\omega_\alpha). \tag*{(48)}

The common orientifold and field normalizations are absorbed into the proportionality. The gauge-invariant derivatives and their kinetic terms are therefore

Dμθα=∂μθα−kαAYμ,D_\mu\theta_\alpha=\partial_\mu\theta_\alpha-k_\alpha A_{Y\mu},
Laxion=−12GαβDμθαDμθβ,(49)\mathcal{L}_{\mathrm{axion}}=-\frac{1}{2}G^{\alpha\beta}D_\mu\theta_\alpha D^\mu\theta_\beta, \tag*{(49)}

where GαβG^{\alpha\beta} is the positive-definite axion kinetic matrix. Expanding these kinetic terms gives a hypercharge gauge-boson mass coefficient proportional to GαβkαkβG^{\alpha\beta}k_\alpha k_\beta. This is the Stückelberg mechanism. Positivity implies that this coefficient vanishes only when every kαk_\alpha vanishes. Since the ωα\omega_\alpha form a basis, a necessary condition for massless hypercharge is

KY(ω)=0for every ω∈H+1,1(X).(50)K_Y(\omega)=0\qquad\text{for every }\omega\in H_+^{1,1}(X). \tag*{(50)}

This condition applies also when hypercharge is a combination of generators from several sectors: YSY_S in (A18) is the projection of that same global generator onto each stack. Other U(1)U(1) gauge bosons may be Stückelberg massive; the final hypercharge combination must satisfy this condition.

For a component sector with endpoint decomposition (A4), let KY,i(ω)K_{Y,i}(\omega) denote its contribution to (A18). For an even class ω\omega,

KY,i(ω)=3qi∫Ciω∣Ci∧fc,i−3qi∫Ci′ω∣Ci′∧fc′,i=6qi∫Ciω∣Ci∧fc,i.(51)\begin{aligned} K_{Y,i}(\omega)&=3q_i\int_{C_i}\omega|_{C_i}\wedge f_{c,i}-3q_i\int_{C_i'}\omega|_{C_i'}\wedge f_{c',i}\\ &=6q_i\int_{C_i}\omega|_{C_i}\wedge f_{c,i}. \tag*{(51)} \end{aligned}

The last equality follows from the orientifold relation between the color fluxes. The weak block contributes nothing because its hypercharge eigenvalue in (A5) is zero. When the color and image-color blocks share a support, the two terms are contributions to the trace on that same support; when their supports differ, they are separate terms in (A18).

Disjointness isolates each sector

Apply (A21) to the even class ω=[Ri]\omega=[R_i]. Every D7 support SS belonging to a different component sector is disjoint from RiR_i, by the geometric argument in the main text. Consequently,

[Ri]∣S=0in H2(S,R).(52)\left.[R_i]\right|_S=0 \quad\text{in } H^2(S,\mathbb{R}). \tag*{(52)}

Indeed, for any closed two-cycle contained in SS, the integral of [Ri]∣S\left.[R_i]\right|_S is its intersection number with RiR_i. This vanishes because SS and RiR_i are disjoint.

All other component sectors therefore drop out of KY([Ri])K_Y([R_i]). Using (A22) and (A15), we obtain

0=KY([Ri])=KY,i([Ri])=6qiAc,i=6qiϵiIQi.(53)0=K_Y([R_i])=K_{Y,i}([R_i]) =6q_iA_{c,i}=6q_i\epsilon_i I_{Q_i}. \tag*{(53)}

Since a sector supplying an SM quark doublet has qi≠0q_i\ne0,

IQi=0.(54)I_{Q_i}=0. \tag*{(54)}

Mixing hypercharge with generators in other component sectors cannot evade this conclusion: their contributions to this test vanish separately. The argument uses disjointness, not linear independence of the component classes.

Finally, consider the arbitrary embedding (A1). Equation (11) implies that each massless D7–D7 fermion belongs to one component sector, with Standard Model transformation law determined by φi\varphi_i. Sectors without symplectic factors cannot supply quark doublets by the main-text proof. Every remaining sector that can supply an SM quark doublet has zero net index by (A25). Hence

IQ=∑iIQi=0,(55)I_Q=\sum_i I_{Q_i}=0, \tag*{(55)}

where sectors without the SM quark-doublet representation contribute zero. This proves the proposition for both split and unsplit allocations, mixed products, and diagonal embeddings.

The conclusion allows massless vectorlike quark-doublet pairs. It excludes the net chirality required by the Standard Model. The proof uses the fermion index and the D7 Chern–Simons couplings, and does not require the stacks to preserve a common four-dimensional supersymmetry.

References

  1. [1]A. Sen, F theory and orientifolds, Nucl. Phys. B 475, 562 (1996), arXiv:hep-th/9605150.DOI
  2. [2]F. Denef, M. R. Douglas, B. Florea, A. Grassi, and S. Kachru, Fixing all moduli in a simple f-theory compactification, Adv. Theor. Math. Phys. 9, 861 (2005), arXiv:hep-th/0503124.
  3. [3]D. Lust, S. Reffert, E. Scheidegger, W. Schulgin, and S. Stieberger, Moduli Stabilization in Type IIB Orientifolds (II), Nucl. Phys. B 766, 178 (2007), arXiv:hep-th/0609013.DOI
  4. [4]M. Demirtas, M. Kim, L. McAllister, J. Moritz, and A. Rios-Tascon, Small cosmological constants in string theory, JHEP 12, 136, arXiv:2107.09064 [hep-th].
  5. [5]L. McAllister, J. Moritz, R. Nally, and A. Schachner, Candidate de Sitter vacua, Phys. Rev. D 111, 086015 (2025), arXiv:2406.13751 [hep-th].arxiv.org/abs/2406.13751
  6. [6]R. Blumenhagen, X. Gao, T. Rahn, and P. Shukla, A Note on Poly-Instanton Effects in Type IIB Orientifolds on Calabi-Yau Threefolds, JHEP 06, 162, arXiv:1205.2485 [hep-th].DOI
  7. [7]T. Coudarchet, F. Marchesano, D. Prieto, and M. A. Urkiola, Symmetric fluxes and small tadpoles, JHEP 08, 016, arXiv:2304.0789 [hep-th].
  8. [8]X. Gao, A. Hebecker, S. Schreyer, and V. Venken, Loops, local corrections and warping in the LVS and other type IIB models, JHEP 09, 091, arXiv:2204.06009 [hep-th].DOI
  9. [9]D. Lust, S. Reffert, E. Scheidegger, and S. Stieberger, Resolved Toroidal Orbifolds and their Orientifolds, Adv. Theor. Math. Phys. 12, 67 (2008), arXiv:hep-th/0609014.arxiv.org/abs/hep-th/0609014
  10. [10]F. Carta, J. Moritz, and A. Westphal, A landscape of orientifold vacua, JHEP 05, 107, arXiv:2003.04902 [hep-th].
  11. [11]C. Criño, F. Quevedo, A. Schachner, and R. Valandro, A database of Calabi-Yau orientifolds and the size of D3-tadpoles, JHEP 08, 050, arXiv:2204.13115 [hep-th].DOI
  12. [12]R. Altman, J. Carifio, X. Gao, and B. D. Nelson, Orientifold Calabi-Yau threefolds with divisor involutions and string landscape, JHEP 03, 087, arXiv:2111.03078 [hep-th].DOI
  13. [13]P. Shukla, Classifying divisor topologies for string phenomenology, JHEP 12, 055, arXiv:2205.05215 [hep-th].DOI
  14. [14]X. Gao, A. Hebecker, S. Schreyer, and V. Venken, The LVS parametric tadpole constraint, JHEP 07, 056, arXiv:2204.04087 [hep-th].
  15. [15]R. Blumenhagen, B. Kors, D. Lust, and S. Stieberger, Four-dimensional String Compactifications with D-Branes, Orientifolds and Fluxes, Phys. Rept. 445, 1 (2007), arXiv:hep-th/0610327.DOI
  16. [16]T. Weigand, Lectures on F-theory compactifications and model building, Class. Quant. Grav. 27, 214004 (2010), arXiv:1009.3497 [hep-th].DOI
  17. [17]M. Cicoli, F. Muia, and P. Shukla, Global Embedding of Fibre Inflation Models, JHEP 11, 182, arXiv:1611.04612 [hep-th].DOI
  18. [18]R. Blumenhagen, S. Moster, and E. Plauschinn, Moduli Stabilisation versus Chirality for MSSM like Type IIB Orientifolds, JHEP 01, 058, arXiv:0711.3389 [hep-th].
  19. [19]M. Cicoli, I. García-Etxebarria, C. Mayrhofer, F. Quevedo, P. Shukla, and R. Valandro, Global Orientifold Quivers with Inflation, JHEP 11, 134, arXiv:1706.06128 [hep-th].DOI
  20. [20]R. Blumenhagen, T. W. Grimm, B. Jurke, and T. Weigand, F-theory uplifts and GUTs, JHEP 09, 053, arXiv:0906.0013 [hep-th].DOI
  21. [21]R. Slansky, Group Theory for Unified Model Building, Phys. Rept. 79, 1 (1981).DOI
  22. [22]R. Hartshorne, Algebraic Geometry, Graduate Texts in Mathematics, Vol. 52 (Springer, 1977).
  23. [23]A. Kanazawa and P. M. H. Wilson, Trilinear forms and chern classes of Calabi–Yau threefolds, Osaka Journal of Mathematics 51, 203 (2014), arXiv:1201.3266 [math.AG].arxiv.org/abs/1201.3266
  24. [24]S. H. Katz and E. Sharpe, D-branes, open string vertex operators, and Ext groups, Adv. Theor. Math. Phys. 6, 979 (2003), arXiv:hep-th/0208104.DOI
  25. [25]A. Hecker, S. C. Kraus, D. Lust, S. Steinfurt, and T. Weigand, Fluxbrane Inflation, Nucl. Phys. B 854, 509 (2012), arXiv:1104.5016 [hep-th].
  26. [26]R. Blumenhagen, M. Cvetic, S. Kachru, and T. Weigand, D-Brane Instantons in Type II Orientifolds, Ann. Rev. Nucl. Part. Sci. 59, 269 (2009), arXiv:0902.3251 [hep-th].DOI
  27. [27]N. Gendler, E. Sheridan, M. Stillman, and D. H. Wu, Holes in Calabi-Yau Effective Cones (2026), arXiv:2603.11173 [hep-th].arxiv.org/abs/2603.11173
  28. [28]V. V. Batyrev, Dual polyhedra and mirror symmetry for Calabi-Yau hypersurfaces in toric varieties, J. Alg. Geom. 3, 493 (1994), arXiv:alg-geom/9310003.arxiv.org/abs/alg-geom/9310003
  29. [29]M. Demirtas, A. Rios-Tascon, and L. McAllister, CY-Tools: A Software Package for Analyzing Calabi-Yau Manifolds (2022), arXiv:2211.03823 [hep-th].arxiv.org/abs/2211.03823
  30. [30]J. Moritz, Orientifolding Kreuzer-Skarke (2023), arXiv:2305.06363 [hep-th].arxiv.org/abs/2305.06363
  31. [31]N. Gendler, D. J. E. Marsh, L. McAllister, and J. Moritz, Glimmers from the axiverse, JCAP 09, 071, arXiv:2309.13145 [hep-th].arxiv.org/abs/2309.13145
  32. [32]E. Sheridan, F. Carta, N. Gendler, M. Jain, D. J. E. Marsh, L. McAllister, N. Righi, K. K. Rogers, and A. Schachner, Fuzzy axions and associated relics, JHEP 09, 016, arXiv:2412.12012 [hep-th].
  33. [33]M. Jain, D. E. Sheridan, D. J. E. Marsh, E. Heyes, K. K. Rogers, and A. Schachner, Bayesian inference on Calabi-Yau moduli spaces and the axiverse: Experimental data meets string theory, Phys. Rev. D 114, 026026 (2026), arXiv:2512.00144 [hep-th].arxiv.org/abs/2512.00144
  34. [34]M. Jain, F. Carta, and D. J. E. Marsh, The C4 axiverse in Kreuzer–Skarke orientifolds: rigid O(1) instantons, and D7 tadpole constraints (forthcoming).
  35. [35]A. Collinucci, F. Denef, and M. Esole, D-brane De-constructions in IIB Orientifolds, JHEP 02, 005, arXiv:0805.1573 [hep-th].
  36. [36]T. W. Grimm, M. Kerstan, E. Palti, and T. Weigand, Massive Abelian Gauge Symmetries and Fluxes in F-theory, JHEP 12, 004, arXiv:1107.3842 [hep-th].
  37. [37]C. Mayrhofer, E. Palti, and T. Weigand, Hypercharge Flux in F-theory: Anomalies and Gauge Coupling Unification, JHEP 09, 082, arXiv:1303.3589 [hep-th].

Paper details

vlnc:W_5b9p5ehs
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents