Lab/UFR-HEP0602/GNPHE/0602/VACBT/0602

On Local Calabi-Yau Supermanifolds and

Their Mirrors
arXiv:hep-th/0601215v1 29 Jan 2006

R. Ahl Laamara1,2, A. Belhaj3,∗, L.B. Drissi1,2 , E.H. Saidi1,2,†

1 Lab/ UFR de Physique des Hautes Energies

Faculté des Sciences, Rabat, Maroc

2 Virtual African Center for Basic Sciences and Technology (VACBT)

Groupement National de Physique des Hautes Energies (GNPHE)

Faculté des Sciences, Rabat, Maroc

3 Department of Mathematics and Statistics, University of Ottawa

585 King Edward Ave., Ottawa, ON, Canada, K1N 6N5

September 16, 2017

Abstract

We use local mirror symmetry to study a class of local Calabi-Yau super-

manifolds with bosonic sub-variety Vb having a vanishing first Chern class. Solv-
ing the usual super- CY condition, requiring the equality of the total U (1) gauge
P P P
charges of bosons Φb and the ghost like fields Ψf one b qb = f Qf , as b qb = 0
P
and f Qf = 0, several examples are studied and explicit results are given for
local Ar super-geometries. A comment on purely fermionic super-CY manifolds
P
corresponding to the special case where qb = 0, ∀b and f Qf = 0 is also made.

Key words: Mirror Symmetry, Local CY super-manifolds, Topological string

theory.

∗
[email protected]
†
[email protected]
Contents
1 Introduction 3

2 Mirror symmetry of local super CY manifolds 4

2.1 Bosonic CY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Mirror of local super-CY . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 Extended A-model . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.2 Extended B-model . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3 Mirror of Ar super-geometries 8
3.1 Local super A1 geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1.1 Extended model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.1.2 Mirror of extended model . . . . . . . . . . . . . . . . . . . . . . 9
3.2 Super Ap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4 More on mirror super-geometry 13

5 Conclusion 15
1 Introduction
Mirror symmetry has played a crucial role in superstring dualities. It provides a map
between Calabi-Yau (CY) manifolds used in the compactification of 10D superstring
models and topological string theory [1, 2]. In particular, it has been shown that the
topological A- and B-models are connected by mirror symmetry [2]; see also discussion
below. However, it has been realized, though, that rigid CY manifolds can have mir-
ror manifolds which are not themselves CY geometries. An intriguing remedy is the
introduction of CY super-manifolds in these considerations [3, 4]. It has thus been sug-
gested that mirror symmetry is between super-manifolds and manifolds alike, and not
just between bosonic manifolds.
On the other hand, it has been found that there is a correspondence between the
moduli space of holomorphic Chern-Simons theory on the CY super-manifold CP3|4 and,
self-dual, four-dimensional N = 4 Yang-Mills theory [5, 6]. This may also be related to
the B-model of open topological string theory having CP3|4 as target space. Partly based
on this work, CY super-manifolds and their mirrors have subsequently attracted a great
deal of attention [7]-[18]. It has been found, for instance, that an A-model defined on
the CY super-manifold CP3|4 is a mirror of a B-model on a quadric hypersurface in
CP3|3 × CP3|3 , provided the Kähler parameter of CP3|4 approaches infinity [7, 8].
Following this observation, an effort has been devoted to go beyond these particular
geometries. A special interest has been given to construct the mirror of Calabi-Yau super-
manifolds whose bosonic parts are compact toric varieties [19]. One of the objectives of
the present work is to extend the result of [19] by considering local Calabi-Yau manifolds
which have been used in type II superstring compactifications in presence of D-branes.
In particular, we discuss the mirror symmetry of the topological A-model on super-
manifolds whose the bosonic part is a local CY variety. The corresponding theory is a
supersymmetric U(1)p linear sigma model with (n + p) chiral superfields with charge qia
and 2p fermionic superfields with charge given by Qaα which is a p × 2p matrix. These
P P2p
charges satisfy the super local CY ( SLCY) condition p+n
i=1 qi
a
− a
α=1 Qα = 0 requiring
equality between the total charge of bosons and the ghost like fields.
In this paper we shall focus on the mirror super-geometry obtained by first choosing
a special form of the full spectrum of U(1)p gauge charges and integrating out some
fermionic fields in the topological B-model. In this way, the mirror B-models will still
have some fermionic directions. Our interest will be on the mirror of ADE super-
geometries and mainly on the fermionic extension of the ordinary Ar class. First, we
study the case of A1 super-geometry, which is found to be closely related to the equation
of the bosonic case in agreement with analysis using LG models. Ordinary A1 geometry is
recovered by canceling the fermionic directions. Then we work out the mirror of a class
of Ar local super CY manifold extending the A1 super-geometry. Finally, we discuss
the mirror symmetry of local higher dimensional super-CY geometries. In particular we
specialize on the mirror symmetry of the topological A-model using a fermionic extension
of line bundle over CPn .
The organization of this paper is as follows. In section 2, we review mirror symmetry
of local super CY manifolds. In section 3, we study mirror of ADE super-geometries
by exhibiting the method on the ordinary Ar series. In section 4, we consider mirror
super-geometries beyond ADE and in section 5, we give a conclusion.

2 Mirror symmetry of local super CY manifolds

In this section, we review mirror symmetry for local (bosonic) CY manifolds [20, 21];
then we give the extension to the super case.

2.1 Bosonic CY
To begin, let us consider a two-dimensional N = 2 supersymmetric linear sigma
model described in terms of n+p chiral superfields Φi with charge qia , i = 1, . . . , n+p; a =
1, . . . , p under U(1)⊗p gauge symmetry [22]. The geometry of the topological A-model
can be analyzed by solving the D-term potential (D a = 0) of the N = 2 linear sigma
model; that is
n+p
X
qia |φi |2 = r a , a = 1, . . . , p, (2.1)
i=1
a
where the r ’s are FI coupling parameters and where the φi ’s are the leading scalar fields
of the chiral superfield Φi . Dividing by U(1)⊗p gauge symmetry, one gets a n-dimensional
1
toric variety
Cn+p \ S
Vn =, (2.2)
C∗ p
where the p copies of C∗ actions indexed by a = 1, . . . , p are given by
a
C∗ p : φ i → λ q i φ i , i = 1, . . . , n + p, (2.3)

with λ a non zero complex number. The requirement for Vn to be a local CY manifold
is to impose the condition
n+p
X
qia = 0. (2.4)
i=1
1
Note that this geometry can be represented by a toric diagram ∆(Vn ) spanned by k = n+ p vertices
P a
n+p
vi in a Zn lattice satisfying qi vi = 0, a = 1, . . . , p.
i=1
On the supersymmetric field theoretic level, this relation implies that the underlying
linear sigma model flows in infrared to a conformal field theory.
Following [20]-[24], the mirror B-model is a Landau-Ginzburg (LG) model with peri-
odic fields {Yi } dual to {Φi } and connected as,

Re(Yi ) = |Φi |2 , i = 1, . . . , n + p, (2.5)

where Re(Yi ) denotes the real part of Yi . Under mirror transformation, eq(2.1) is mapped
to
X
qia Yi = ta , a = 1, ..., p, (2.6)
i

with r a = Re (ta ). Moreover the LG superpotential of the topological B-model reads as

n+p
X
W (Y1 , ..., Yn+p) = e−Yi . (2.7)
i=1

For convenience, it is useful to use the following field redefinitions

ŷi = e−Yi , i = 1, . . . , n + p, (2.8)

Then the superpotential W = W (ŷ1 , ..., ŷn+p) reads as

n+p
X
W = ŷi , (2.9)
i=1

and so eq(2.6) translates into the following projective hypersurface

n+p
Y qa
ŷi i = e−ta , a = 1, ..., p, (2.10)
i=1

with the manifest projective symmetry ŷi → λŷi following from the CY condition (2.4).
The solution of the constraint eqs(2.10) and projective symmetry defines a (n + p) − p −
1 − 1 = n − 2 dimensional toric manifold given by a holomorphic hypersurface in Cn−1

F (y1 , ..., yn−1) = 0. (2.11)

To recover the right dimension of the original manifold; that is a complex dimension
n local CY manifold, we generally use an ad-hoc trick which consist to add by hand
two extra holomorphic variables u and v combined in a quadratic form uv and modify
previous equation as
F (y1 , ..., yn−1) = uv. (2.12)

The main objective in what follows is to extend this analysis to a linear A-model with
fermionic (ghosts) fields and study the resulting mirror B-model. Besides the general-
ization of above results to local super-CY manifolds, one of the results following from
this fermionic extension is the re-derivation of eq(2.12) without need of adding by hand
of the term uv of right hand. As we will show later, the new manifold is given by a
hypersurface of type,
G(y1 , ..., yn−1) = χη, (2.13)
where, instead of u and v variables, we have now the variables χ and η which are
ghost like fields. As we will see, this relation defines an even complex n dimension
hypersurface of the complex superspace C(n−1)|2 . This geometry may be then viewed as
alternative elevation of (2.11). The standard elevation eq(2.12) is given by the purely
bosonic hypersurface in C(n+1)|0 .

2.2 Mirror of local super-CY

Here we want to study the mirror of the fermionic extension of the topological A-model
on local toric CY manifolds discussed in the previous subsection. Actually, this may be
viewed as an extension of paper [19] which has dealt with the case of compact bosonic
toric manifolds. Important examples of that work have been projective spaces and prod-
ucts thereof.

2.2.1 Extended A-model

Roughly, the extension corresponds to adding, to the usual bosonic superfield Φj , a set
of f -fermionic chiral superfields Ψα with Qaf charge under U(1)⊗p gauge symmetry. We
then have
P
ϑa qja
Φj → ei a Φj , j = 1, ..., n + p,
P
ϑa Q a
Ψα → ei a α
Ψα , α = 1, ..., f, (2.14)

with same transformations for the leading component fields φj and ψ α respectively and
where the ϑa ’s are the gauge group parameters. The full spectrum of U(1)⊗p charge
vectors q ′a = (q a |Qa ) thus takes the form


a
(q a |Qa ) = q1a , . . . , qp+n |Qa1 , . . . , Qaf , a = 1, . . . , p. (2.15)

The extended Da -term equations resulting from the above generalized A-model is given
by minimizing the Kahler potential of the 2D N = 2 generalized superfield action
Z Z 
2 4 2 2
SN =2 = d σd θK + d σd θW + cc , (2.16)

with respect to the gauge superfields Va . In above relation, K is the usual gauge invariant
Kahler term and W is a chiral superpotential with superfield dependence as,
 
K = K Φ1 , ..., Φ+ +
n+p ; Ψ1 , ..., Ψf ; V1 , ..., Vp ,

W = W [Φ1 , ..., Φn+p ; Ψ1 , ..., Ψf ] , (2.17)

as well as coupling constant moduli which have not been specified. Using the explicit
expression of K and putting back into
∂K
Da = |θ=0 = 0, (2.18)
∂Va
we get the following Da -term equations
p+n f
X X
qia |φi |2 + Qaα |ψ α |2 = Re(ta ), a = 1, . . . , p, (2.19)
i=1 α=1

where Re(ta ) stands for FI coupling constant. Strictly speaking, this is an even hy-
persurface embedded in the complex supermanifold Cp+n|f with dimension (p + n|f ).
Therefore, the space of vacua of above generalized supersymmetric action is a toric
super-manifold V n obtained by dividing Cp+n|f by U(1)⊗p gauge symmetry group in
same logic as in eq(2.2). Thus we have
Cp+n|f \ S
Vn = . (2.20)
C∗ p
With this relation at hand, one can go ahead and try to develop the toric super-geometry
of these local super-manifolds by mimicking the standard toric geometry analysis of toric
varieties. We shall not do this here; what we will do rather is to study some specific
examples with direct link to type II superstring theory compactifications. The first class
of these examples concerns specific fermionic extensions of ADE geometries. The local
super-CY (LSCY) condition reads as follows
p+n
X f
X
qia − Qaα = 0. (2.21)
i=1 α=1

This constraint equation is required by the invariance of holomorphic measure Ω(p+n|f )

of the complex superspace Cp+n|f ,
n+p
! f
!
Y Y
(p+n|f )
Ω = dφi dψ α , (2.22)
i=1 α=1

under C∗ p toric symmetry. In what follows, we shall fix our attention on those local CY
super-manifolds V n obeying the following special solution,
p+n f
X X
qia = 0, Qaα = 0, a = 1, . . . , p. (2.23)
i=1 α=1

In this particular class of solutions of eq(2.21), we have taken the bosonic subvariety as
a local CY manifold. This is the case for bosonic sub-varieties given by the fermionic
extensions of ADE geometries we are interested in here. It is also remarkable that
fermionic directions obey as well a CY condition for bosonic manifold. In conclusion
section, we will make a comment on this issue.
2.2.2 Extended B-model

Under T-duality, the bosonic superfield Φi of the linear super-toric sigma model is re-
placed by a dual superfield Yi as before, while the fermionic superfield Ψα is dualized by
the triplet (Xα , η α , χα ) [8]. The bosonic superfields Xα is related to Ψα as

Re(Xα ) = −|Ψα |2 , α = 1, ..., f, (2.24)

and the accompanying pair of chiral superfields {η α } and {χα } are fermionic superfields
required by the preservation of the super-dimension and hence the total central charge.
Under this dualization, the original complex superspace Cp+n|f gets mapped to

Cp+n+f |2f . (2.25)

The extended B-model, mirror to the above fermionic extended A-model with superfield
action SN =2 , is given in terms of the following path integral, see also [19],
Z " p # Z 
Y
Z = D̥ δ (̥a − ta ) exp W (Y, X, η, χ) , (2.26)
a=1
Q Q
where we have set D̥ = ( i dYi ) ( α dXα dη α dχα ). In this relation the ̥a ’s are the D-
terms of the extended A-model and W = W (Y, X, η,χ) is the extended LG superpotential
of the topological B-model. They are as follows,
n+p f
X X
̥a = qia Yi − Qaα Xα , a = 1, ..., p,
i=1 α=1
n+p f
!
X X
W = e−Yi + e−Xα (1 + η α χα ) . (2.27)
i=1 α=1

To extract informations on the local super-geometry of the B-model, we need to integrate

out the delta functions. Below, we shall focus our attention on the special case where
f = 2p and exemplify with models which have been used in type II superstring theory
compactifications.

3 Mirror of Ar super-geometries
Here we focus on the super-geometry extending the usual ordinary Ar geometries. A
quite similar analysis is a priori possible for the DE, affine and indefinite extensions.

3.1 Local super A1 geometry

To illustrate the construction, we initially consider the example of the model A1 . This is
a supersymmetric gauge theory with a U(1) gauge symmetry and three chiral superfields
Φi with charge (1, −2, 1) together with a real gauge superfield V . The D-term constraint
(equation of motion of V ) reads as

|Φ1 |2 − 2|Φ2 |2 + |Φ3 |2 = Re(t). (3.1)

This geometry describes the Kahler deformation of the A1 singularity of the ALE spaces

uv = z 2 , (3.2)

where u, v and z are the generators of gauge invariants. They are realized in terms of
the scalar fields as follows

u = Φ21 Φ2 , v = Φ23 Φ2 , z = Φ1 Φ2 Φ3 . (3.3)

For generalizations to rank r ≥ 2 ordinary ADE geometries as well as affine extensions

and beyond see [25].

3.1.1 Extended model

Basically, there is an abundance of possible fermionic extensions of above model. It may

be limited by imposing the LSCY condition (2.21). Since we are interested in the case
f = 2p = 2, the full spectrum of U(1) charge that one can have is given by the vector

q ′ = (q|Q) = (1, −2, 1|1, −1) . (3.4)

In this construction, A1 model appears as a subsystem while, as noted before and as

far as super-CY condition is concerned, there are several solution of eq(2.21). Using the
extension (3.4), the D-term for the A1 super-geometry becomes

|Φ1 |2 − 2|Φ2 |2 + |Φ3 |2 + |Ψ1 |2 − |Ψ2 |2 = Re(t), (3.5)

where Re(t) is the unique Kahler parameter of the model.

3.1.2 Mirror of extended model

Applying mirror transformation to above extended A-model with A1 super-geometry,

the associated mirror B-model is obtained in the same way as presented in previous sub-
section. The corresponding extended LG path integral eqs(2.26-2.27) takes the following
form
Z
Z = D̥δ [Y1 − 2Y2 + Y3 − X1 + X2 − t]
3 2
!
X X
× exp e−Yi + e−Xα (1 + η α χα ) . (3.6)
i=1 α=1
 Q3
Q2

with D̥ = i=1 dYi α=1 dXα dη α dχα . As usually, to extract information on the
mirror super-geometry of the B-model, we integrate out the fermionic fields η 1 , χ1 . Then
solving the delta function constraint by integrating out X1 yields,
Z 3
!

X  
Z = D̥ e e−Y1 +2Y2 −Y3 e−X2 exp e−Yi +e−X2 1 + η2 χ2 +et e−Y1 +2Y2 −Y3 , (3.7)
i=1
Q3

e =
where D ̥ i=1 dYi (dX2 dη 2 dχ2 ). Now, introducing new complex variables xi and
yi such that
x = e−X2 , yi = e−Yi , i = 1, 2, 3, (3.8)
the above partition function becomes
Z Y3   X3  !
dyi t y1 y3
Z = (dxdη 2 dχ2 ) exp yi + x 1 + η2 χ2 + e 2 . (3.9)
i=1
y23 i=1
y2

The rescaling x̃ = (x/y23) allows us to rewrite the above path integral as follows
Z X3  !
y y
1 3
Z = dy1 dy2 dy3dx̃dη 2 dχ2 exp yi + x̃y23 1 + η 2 χ2 + et 2 . (3.10)
i=1
y 2

In order to get the mirror of local super-geometry A1 , we can see x̃ as a Lagrange

multiplier; integrating it out one gets the following equation of motion
y1 y3 t
1 + η2 χ2 + e = 0. (3.11)
y22
The objective now is to interpret this equation as the mirror constraint equation of the
topological A-model on A1 super-geometry. In fact, we can solve (3.11) as
y1 y3
= −(1 + η2 χ2 )e−t . (3.12)
y22
Replacing now t by t′ = t + iπ, one absorbs the minus sign
y1 y3 ′ ′
2
= e−t + η 2 χ2 e−t . (3.13)
y2
Actually, this equation is quite similar to the bosonic one; except now we have the
presence of the additional contribution η 2 χ2 e−t induced by the fermionic fields. It is
′

easy to see that in the patch η 2 = χ2 = 0, we recover the bosonic mirror constraint
equation of ALE space with A1 singularity, namely
y1 y3 ′
2
= e−t . (3.14)
y2
Return to equation (3.13); a straightforward computation reveals that this equation can
be solved by taking the following parameterization
1
y1 = y, y3 = ,
y
1 t
y2 = (1 + η 2 χ2 )− 2 e 2 . (3.15)
where we have set t′ = t. We thus end with the following LG potential
1 1 t
y+ + (1 + η2 χ2 )− 2 e 2 = 0, (3.16)
y
which is mirror to sigma model on A1 super-geometry. This equation has the three
following remarkable features:
1. For η2 = χ2 , we recover the usual bosonic LG superpotential mirror to the bosonic
A1 geometry
1 t
y+ + e 2 = 0. (3.17)
y
2. In the case t → 0, one discovers the rule to define the super extension of the A1
singularity with U (1) charges as in eq(3.4). The mirror of super A1 singularity can be
then defined as follows
y1 y3 = y22 (1 + η2 χ2 ) (3.18)

in agreement with indication from conformal Landau Ginzburg field models where ad-
junction of quadratic terms do not modify the total central charge. Moreover, by using
fermionic statistics which forbids higher powers in η 2 and χ2 , one may define extensions
of above A1 singularity.
3. In the limit where the condensate modulus η 2 χ2 is small, eq(3.16) reduces to
1 t 1 t
y+ + e 2 − η 2 χ2 e 2 = 0. (3.19)
y 2
t
By making the identification 21 η2 χ2 e− 2 with the uv of the relation (2.11-2.12), one dis-
covers that the uv term added by hand in the bosonic case to recover the right dimension
of mirror manifold, is generated in a natural way in super-geometry. In the limit t → 0,
we have
1
y+ + 1 = η ′2 χ′2 , (3.20)
y
where we have set η′2 χ′2 = 21 η2 χ2 . This is a complex two dimensions even hypersurface
of C1|2 .

3.2 Super Ap
Now, we would like to push further the above results on super A1 to the class of Ap
super-geometry series having usual Ap geometry as local bosonic Calabi-Yau submani-
folds. To start, recall that Ap geometry has a description in terms of U(1)⊗p sigma model
involving (p + 2) chiral fields with the bosonic charge p × (p + 2) matrix
 
1 −2 1 0 0 ... 0 0 0 0 0
 
 0 1 −2 1 0 ... 0 0 0 0 0 
 
a  
qi =  0 0 1 −2 1 . . . 0 0 0 0 0 . (3.21)
 . .. 
 .. . 
 
0 0 0 0 0 . . . 0 0 1 −2 1

Basically, there are several fermionic extensions of the above A-model. However as we
mentioned before, we consider a model with 2p fermionic fields. In this way, the SLCY
condition may limit the choice of the charge matrix. For a reason to be specified later
on, we propose the following U(1)⊗p charge spectrum for ghost like superfields
 
1 −1 0 0 ... 0 0 0 0
 
 0 0 1 −1 . . . 0 0 0 0 
 
 
Qaα =  0 0 0 0 ... 0 0 0 0 . (3.22)
 . .. 
 .. . 
 
0 0 0 0 . . . 0 0 1 −1

This representation constitues a simple and natural extension of eq(3.4) recovering A1

super-geometry as a leading example; other representations are obviously possible. This
choice of U(1)⊗p charge matrix for ghost like fields allows us to handle each line as an
individual A1 super-geometry. In this way, we can easily repeat the same lines that we
have done for the super A1 case. Let us give some details below.
Roughly, LG mirror superpotential is given in terms of the following path integral
Z " p # !
Y X X
Z = D̥ δ (̥a − ta ) exp e−Yi + e−Xα (1 + η α χα ) , (3.23)
a=1 i α

Qp+2
Q2p

where now D̥ = i=1 dYi α=1 dXα dηα dχα and where we have set

̥a = Ya − 2Ya+1 + Ya+2 − X2a−1 + X2a . (3.24)

This partition function Z has p delta functions δ (̥a − ta ). To get the mirror super-
geometry, we first integrate out the fermionic field variables (η1a χ1a ) leaving only a
dependence on (η2a χ2a ), a = 1, 2, . . . , p and then we use delta functions to eliminate
the field variables X2a−1 . In doing so and following the same way as before, we get p
equations of motion,
Y qa
1 + η 2a χ2a = yi i , a = 1, . . . , p. (3.25)
i
To see how to obtain these equations, let us consider the case of A2 super-geometry. This
is a U(1)2 gauge theory with four chiral superfields (Φ1 , Φ2 , Φ3 , Φ4 ) and four ghost like
ones (Ψ1 , Ψ2 , Ψ3 , Ψ4 ). The full spectrum of U(1)2 gauge charges is given by

q ′1 = (1, −2, 1, 0|1, −1, 0, 0) ,

q ′2 = (0, 1, −2, 1|0, 0, 1, −1) . (3.26)

The above path integral reduces in present case to

Z 4 4
!
X X
Z = D̥δ (̥1 − t1 ) δ (̥2 − t2 ) exp e−Yi + e−Xα (1 + η α χα ) , (3.27)
i=1 α=1
Q4
Q4

with field measure D̥ = i=1 dYi α=1 dXα dηα dχα and D-terms as

̥1 = Y1 − 2Y2 + Y3 − X1 + X2 ,
̥2 = Y2 − 2Y3 + Y4 − X3 + X4 . (3.28)

Integrating in similar way as we have done for A1 super-geometry and making the same
variable changes, we get
Z " 4  #
X et1
y y
1 3
Z = D̥′ exp yi + x˜1 y22 1 + η 2 χ2 + 2
y 2
 i=1 t2

e y2 y4
× exp x˜2 y32 1 + η 4 χ4 + . (3.29)
y32
Q4

with D̥′ = i=1 dy i (dx˜1 dx˜2 dη2 dχ2 dη4 dχ4 ). In this case, we have two equations of
motion which are given by
y1 y3 ′
2
= (1 + η 2 χ2 ) e−t1 ,
y2
y2 y4
= (1 + η 4 χ4 ) e−t2 ,
′
2
(3.30)
y3
with t′a = ta + iπ. After solving these two equations, we come up with the following
mirror relation
 ′   
1 t1
2 −t′2
−t′1 1
+ e (1 + η 2 χ2 ) + y + y e e (1 + η 4 χ4 )(1 + η2 χ2 )
2 2 2 = 0, (3.31)
y
1
which should be compared with the usual mirror relation of ordinary A2 geometry y
+
2
1 + y + y = 0 associated to the limit t′a = 0 and η 2 = χ2 = 0.

4 More on mirror super-geometry

The method developed so far can be also used to build other local super CY mani-
folds. A simple extension of above A1 super-geometry analysis is given by a sigma model
with target space involving a fermionic extension of line bundle over CPp with p ≥ 2.
The case p = 1 corresponds exactly to the A1 super-geometry studied previously. Let us
analyze the case p = 2; that is the line bundle O(−3) over CP2 . It admits a U(1) sigma
model description in terms of four bosonic chiral fields with charge vector (1, 1, 1, −3)
and the corresponding D-term equation is given by

|Φ1 |2 + |Φ2 |2 + |Φ3 |2 − 3|Φ4 |2 = Re(t) (4.1)

Adding now 2 ghost like field variables Ψ1 and Ψ2 with vector charge (1, −1), as required
by SLCY condition, the D-term constraint equation of the extended A-model is given by

|Φ1 |2 + |Φ2 |2 + |Φ3 |2 − 3|Φ4 |2 + |Ψ1 |2 − |Ψ2 |2 = Re(t). (4.2)

The corresponding mirror super-geometry is given in terms of the following path integral

Z 4 2
!
X X
Z= D̥δ (̥ − t) exp e−Yi + e−Xα (1 + ηα χα ) . (4.3)
i=1 α=1
Q4
Q2

with D̥ = i=1 dYi α=1 dXα dηα dχα and

̥ = Y1 + Y2 + Y3 − 3Y4 − X1 + X2 . (4.4)

Now integrating out the fermionic fields η 1 , χ1 and solving the delta function constraint
by eliminating X1 , we get
Z


Z = D̥ e−Y1 −Y2 −Y3 +3Y4 e−X2
4
!
X
× exp e−Yi + e−X2 (1 + η2 χ2 + et e−Y1 −Y2 −Y3 +3Y4 ) , (4.5)
i=1
Q4

with D̥ = i=1 dYi (dX2 dη 2 dχ2 ). Using the following field re-definition

yi = e−Yi , x = e−X2 (4.6)

the above equation becomes

Z Y 4
! " 4  #
dyi X et y1 y2 y3
Z= (dxdη 2 dχ2 ) exp yi + x 1 + η2 χ2 + . (4.7)
y4
i=1 4 i=1
y43
x
By help of the following rescaling x̃ = y44
, the mirror geometry becomes
Z 4
" 4  #
Y X et
y 1 y 2 y 3
Z= dx̃dη2 dχ2 dyi exp yi + x̃y44 1 + η 2 χ2 + . (4.8)
i=1 i=1
y43

In this case, the equation of motion reads as

y1 y2 y3
= −(1 + η 2 χ2 )e−t . (4.9)
y43
Absorbing the minus sign by replacing t by t + iπ, the above equation becomes
y1 y2 y3 ′ ′
3
= e−t + η2 χ2 e−t . (4.10)
y4
This can be easily solved by the following parameterization
1
y1 = x, y2 = y, y3 =
xy
−1 t′
y4 = (1 + η2 χ2 ) 3 e3. (4.11)

The superpotential describing the mirror of the super-geometry reads as

1 −1 t′
x+y+ + (1 + η 2 χ2 ) 3 e 3 = 0. (4.12)
xy
For η2 χ2 = 0, we rediscover the usual bosonic relation.

5 Conclusion
In this paper, we have studied mirror symmetry of A-model on Calabi-Yau super-
manifolds constructed as fermionic extensions of local toric CY satisfying the SLCY con-
P P2p Pp+n a P2p
dition p+n a
i=1 qi =
a
α=1 Qα . By solving this condition as i=1 qi = 0 and
a
α=1 Qα = 0
separately; we have considered two classes of mirror super-geometries. The first class
deals with a special fermionic extension of ordinary geometries and the second class
concerns a set of sigma models with target space involving a fermionic extension of line
bundle over CPn with n ≥ 2. The representations studied here are not the general ones
since the bosonic subvariety of super-manifold considered here is taken as a Calabi-Yau
manifold. This condition is obviously not a necessary condition for building Calabi-Yau
super-manifolds. This work may be viewed as a extension of [19] which has dealt with
bosonic compact toric varieties. The mirror geometries studied in that paper have dealt
only with bosonic variables. However, here the mirror B-models involve fermionic direc-
tions captured by the ghost like fields. In dealing with the mirror of Ar super-geometries,
we have shown that these local CY super-manifolds are described by algebraic geome-
try equations quite similar to the bosonic case. The later can be obtained by canceling
fermionic directions. Moreover, we have found that in super-geometry, the right dimen-
sion of the bosonic CY subvariety is recovered in natural as shown on (3.20). Finally
we have shown that this approach applies as well to higher dimensional mirror super-
geometries; mirror of A-model on super line bundle over CPn studied in section 4 is an
example amongst others.
In the end of this study, we would like to add that along with ordinary CY man-
ifolds embedded in Cn|0 and super-CY manifolds embedded in complex space Cn|m , we
may also have super CY manifolds embedded in the purely fermionic space C0|m without
basic bosonic coordinates. These special super CY varieties are then hypersurface in
C0|m involving ghost fields like only.

Acknowledgement 1 :
AB would like to thank C. Vafa for kind hospitality at Harvard University, where a part
of this work is done. We thank J. Rasmussen and A. Sebbar for collaborations related to
this work. This research work is supported by the program Protars III D12/25, CNRST.

References
[1] S. Ferrara, J. A. Harvey, A. Strominger, C. Vafa, Second-Quantized Mirror Sym-
metry, Phys.Lett.B 361 (1995) 59-65, hep-th/9505162.

[2] M. Aganagic, A. Klemm, M. Marino, C. Vafa, Matrix Model as a Mirror of Chern-

Simons Theory, JHEP 0402 (2004) 010, hep-th/021109.

[3] S. Sethi, super-manifolds, rigid manifolds and mirror symmetry, Nucl. Phys. B 430
(1994) 31, hep-th/9404186.

[4] A. Schwarz, Sigma-models having super-manifolds as target spaces, Lett. Math.

Phys. 38 (1996) 91, hep-th/9506070.

[5] E. Witten, Perturbative gauge theory as a string theory in twistor space,

hep-th/0312171.

[6] A.D. Popov, C. Sämann, On supertwistors, the Penrose-Ward transform and N = 4

super Yang-Mills theory, hep-th/0405123.

[7] A. Neitzke, C. Vafa, N = 2 strings and the twistorial Calabi-Yau, hep-th/0402128.

[8] M. Aganagic, C. Vafa, Mirror symmetry and super-manifolds, hep-th/0403192.

[9] S. P. Kumar, G. Policastro, Strings in twistor superspace and mirror symmetry,

hep-th/0405236.

[10] A.D. Popov, M. Wolf, Topological B-model on weighted projective spaces and self-
dual models in four dimensions, hep-th/0406224.

[11] C. Ahn, Mirror symmetry of Calabi-Yau super-manifolds, hep-th/0407009.

[12] M. Rocek, N. Wadhwa, On Calabi-Yau super-manifolds, hep-th/0408188; On

Calabi-Yau super-manifolds II, hep-th/0410081.
[13] C. Zhou, On Ricci flat super-manifolds, hep-th/0410047.

[14] M. Wolf, On Hidden Symmetries of a Super Gauge Theory and Twistor String
Theory, JHEP 0502 (2005) 018, hep-th/0412163.

[15] S. Seki, K. Sugiyama, Gauged Linear Sigma Model on super-manifold,

hep-th/0503074.

[16] A. D. Popov, C. Saemann, M. Wolf, The Topological B-model on a Mini-Supertwistor

Space and Supersymmetric Bogomolny Monopole Equations, JHEP 0510:058,2005,
hep-th/0505161.

[17] T. Tokunaga, String Theories on Flat super-manifolds, hep-th/0509198.

[18] R. Ricci, Super Calabi-Yau’s and Special Lagrangians, hep-th/0511284.

[19] A. Belhaj, L.B. Drissi, J. Rasmussen, E.H. Saidi, A. Sebbar, Toric Calabi-Yau super-
manifolds and mirror symmetry, J.Phys. A38 (2005) 6405-6418, hep-th/0410291.

[20] K. Hori, C. Vafa, Mirror symmetry, hep-th/0002222.

[21] K. Hori, A. Iqbal, C. Vafa, D-Branes And Mirror Symmetry, hep-th/0005247.

[22] E. Witten, Phases of N = 2 theories in two dimensions, Nucl. Phys. B 403 (1993)
159, hep-th/9301042.

[23] M. Aganagic, C. Vafa, Mirror symmetry, D-branes and counting holomorphic discs,
hep-th/0012041.

[24] A. Belhaj, Mirror symmetry and Landau Ginzburg Calabi-Yau superpotentials in

F-theory compactifications, J. Phys. A 35 (2002) 965, hep-th/0112005.

[25] M. Ait Ben Haddou, A. Belhaj, E.H. Saidi, Geometric Engineering of N=2 CFT4s
based on Indefinite Singularities: Hyperbolic Case, Nucl.Phys. B674 (2003) 593-614,
hep-th/0307244
M. Ait Ben Haddou, A. Belhaj, E.H. Saidi, Classification of N=2 supersymmetric
CFT4s: Indefinite Series, J.Phys. A38 (2005) 1793-1806, hep-th/0308005.
A. Belhaj, E.H. Saidi, Non simply laced quiver gauge theories in superstring com-
pactifications, Afr.J.Math.Phys.1:29-51,2004.