Professional Documents
Culture Documents
Clifford-Finsler Algebroids and Nonholonomic Einstein-Dirac Structures
Clifford-Finsler Algebroids and Nonholonomic Einstein-Dirac Structures
I. INTRODUCTION
A class of spacetimes possessing noncommutative and/or Lie algebroid symmetries can be
defined as exact solutions in string and Einstein gravity.1,2 This leads to new developments in
formulation of classical and quantum field theories following the geometry of nonholonomic
manifolds3 possessing Lie algebroid symmetry.4 Lie algebroid structures in gravity are modeled by
generic off-diagonal metrics and nonholonomic frames 共vielbeins兲 with associated nontrivial non-
linear connection 共N-connection兲. The spacetimes provided with compatible metric, linear connec-
tion, and N-connection structures and possessing Lie algebroid symmetry are called Einstein–
Cartan algebroids, or 共in a more general context, for various extensions of the Riemann–Cartan
geometry兲 Lie N-algebroids. Usually, the Lie algebroids can be defined for a vector, or tangent,
bundle but, in general, they can be considered for any nonholonomic manifold provided with a
nonintegrable 共nonholonomic兲 distribution. 关In our works we use distributions defining
N-connection structures with the coefficients induced by the metric’s off-diagonal terms and cor-
responding vielbein’s coefficients. The geometric constructions are performed for nonholonomic
manifolds, i.e., spaces provided with nonintegrable distributions. In a particular case, when such
distributions are related to the exact sequences of subspaces defining an N–connection, the spaces
are called N anholonomic兴. In brief, such spaces are called Lie N-algebroids. Similar constructions
elaborated for the Einstein–Dirac spaces give rise to the geometry of Clifford algebroids. If the
curved spinor spaces are also enabled with Finsler, or Lagrange, structures, we deal with Clifford–
Finsler, or Clifford–Lagrange, algebroids.
We note that the methods of Finsler and Lagrange geometry5,6 were recently reconsidered in
a new way in order to solve physical problems related to standard theories of gravity and field
a兲
Electronic mail: svacaru@brocku.ca
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-2 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
The geometry of spinor spaces enabled with nonlinear connection 共N-connection兲 structure
was elaborated in a series of works11–13 共see also Refs. 19–21 for general references on Clifford
and spinor differential geometry and applications to physics兲. Here we note that the concept of
N-connection was originally proposed in the framework of Finsler geometry and geometric me-
chanics but such nonholonomic structures 关defined by exact sequences of subspaces of the tangent
space to the spacetime manifold and related nonintegrable distributions兴 may be also considered
on 共pseudo兲 Riemannian and Einstein–Cartan–Weyl spaces, see the discussion and historical re-
marks in Ref. 7. A class of nonholonomic spinor configurations can be defined by exact solutions
of the Einstein–Dirac equations parametrized by generic off-diagonal metric ansatz, nonholonomic
vielbeins associated to nontrivial N-connections, and arbitrary linear connections with nontrivial
torsion.
The aim of this section is to formulate the theory of Clifford algebroids provided with non-
linear connection 共N-connection兲 structures, i.e., the theory of Clifford N-algebroids. For holo-
nomic configurations, the Clifford algebroids can be defined as usual Lie algebroids4 but associ-
ated to a Clifford bundle instead of a vector or tangent bundle.
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-3 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
冋
e = ei =
xi
− Nai 共u兲 a ,ea = a .
y y
册 共3兲
Such vielbeins are called N-adapted frames. 关In order to preserve a relation with the previous
denotations,7,11–13 we note that e = 共ei , ea兲 and e = 共ei , ea兲 are, respectively, the former ␦
= ␦ / u = 共␦i , a兲 and ␦ = ␦u = 共di , ␦a兲 which emphasize that operators 共3兲 and 共2兲 define, corre-
spondingly, certain “N-elongated” partial derivatives and differentials which are more convenient
for calculations on such nonholonomic manifolds兴.
We denote by { : TV → TM the differential of a map : TV → hV, where hV is locally iso-
morphic to M, defined by fiber preserving morphisms of the tangent bundles TV and TM. The
kernel of { is just the vertical subspace vV , dim共vV兲 = m, with a related inclusion mapping
i : vV → TV and hV is a horizontal subspace. It should be emphasized that such maps and local
decompositions exist when V → M is a surjective submersion. A particular case is that of a fiber
bundle but we can obtain the results in the general case 关see the discussions and references in Refs.
3 and 7 related to almost sympletic manifolds, 共pseudo兲 Riemannian spaces, and vector bundles
and generalizations兴. A nonlinear connection 共N-connection兲 N on a manifold V is defined by the
splitting on the left of an exact sequence
0 → vV → TV → TV/vV → 0,
i.e., by a morphism of submanifolds N : TV → vV such that N ⴰ i is the unity in vV.
Equivalently, a N-connection is defined by a Whitney sum of horizontal 共h兲 subspace, hV
⯝ M 共we shall use the symbol “⯝” in order to emphasize some isomorphisms of spaces兲 and
vertical 共v兲 subspace, vV,
TV = hV 丣 vV. 共4兲
The spaces provided with N-connection structure are denoted by boldface symbols. For instance,
we write V for a manifold V provided with a distribution 共4兲 共being, in general, nonintegrable, i.e.,
nonholonomic—in the literature an equivalent term it is also used: anholonomic兲. Such manifolds
are called N-anholonomic with the nonholonomy defined by a N-connection structure. In a similar
manner, we can define nonholonomic manifolds enabled with certain more general nonintegrable
共nonholonomic兲 distributions of subspaces in TV, or in TTV, and so on, but in this paper we shall
restrict our considerations only to N-anholonomic manifolds with N-connection splitting on TV.
We shall use boldfaced indices for the geometric objects adapted to the N-connection.
Locally, a N-connection is defined by its coefficients Nai 共u兲,
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-4 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
N = Nai 共u兲dxi 丢 .
ya
The well-known class of linear connections consists on a particular subclass with the coefficients
being linear on y a, i.e., Nai 共u兲 = ⌫abj共x兲y b. Any N-connection is characterized by its N-connection
curvature
1
⍀ = ⍀aijdxi Ù dx j 丢 a ,
2 y
with N-connection curvature coefficients
B. Clifford N-algebroids
Let us state the notations for abstract 共coordinate兲 d-tensor indices of geometrical objects
defined with respect to an arbitrary 共coordinate兲 local basis, i.e., system of reference. For a local
basis on V, we write e␣ = 共ei , va兲. The small Greek indices ␣ ,  , ␥ , . . . are considered to be general
ones, running values 1 , 2 , . . . , n + m and i , j , k , . . . and a , b , c , . . ., respectively, label the geometrical
objects on the base and typical “fiber” and run, correspondingly, the values 1 , 2 , . . . , n and
1 , 2 , . . . , m. The dual base is denoted by e␣ = 共ei , va兲. The local coordinates of a point u 苸 V are
written u = 共x , y兲, or u␣ = 共xi , y a兲, where y a is the ath coordinate with respect to the basis 共va兲 and
共xi兲 are local coordinates on hV with respect to ei. We shall use “boldface” symbols in order to
emphasize that the objects are defined on spaces provided with N-connection structure.
We suppose that the N-anholonomic manifold V admits a d-spinor structure which allows us
to introduce spinor coordinates and parametrizations of geometrical objects. Let
´ ´ ˜k ´ ´ ˜k
e␣´ ␣´ = 共e´i1,e´i2, . . . ,e´i 共n兲
,eá1,eá2, . . . ,eá 共m兲
兲,
with boldfaced indices running coordinate values on dimensions of d-spinor spaces, k̃共n兲 and k̃共m兲,
being the coefficients of a d-spinor basis
Similar formulas hold for the associated d-spinor spaces provided with local bases e␣` ␣` and e␣` ␣` .
Such spinor bases are stated to be compatible to the N-connection splitting, i.e. to the vielbeins 共3兲
and 共2兲. For a given d-metric structure on V and its spinor decomposition, with associated spinor
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-5 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
bases e␣` = 共e`i , eà兲, which allows us to introduce the ␥ objects, we can define, for instance, a
N-adapted basis
´`
e␣ = 共␥␣兲␣´ ␣` e␣´ e␣` = 关ei = 共␥i兲iie´ie`i,ea = 共␥a兲áàeáeà兴.
As a result, we can alternatively consider spinor coordinates, for instance,
´`
u␣ = 共xi,y a兲 → u␣´ ␣` = 共xii,y áà兲.
For even dimensions of n, or m, further reductions are possible, when xii → xII⬘, or y áà → y AA⬘. This
´`
way, the d-tensor indices can be transformed into the d-spinor ones and inversely.
The standard definition of a Lie algebroid A ⯐ 共E , 关· , · 兴 , 兲 is associated to a vector bundle
E = 共E , , M兲, with a surjective map : E → M of the total spaces E to the base manifold M, of
respective dimensions dim E = n + m and dim M = n. The algebroid structure is stated by the anchor
map : E → TM 共TM is the tangent bundle to M兲 and a Lie bracket on the C⬁共M兲-module of
sections of E, denoted Sec共E兲, such that
关ea,eb兴 = Cab
c
共x兲ec 共8兲
and subjected to the structure equations
aj
ib
xj
− j a
b
i
xj
= cjCab
c
, 兺
cyclic共a,b,c兲
冉 aj
Cbc
d
xj
+ Cdaf Cbc
f
冊
= 0; 共9兲
for simplicity, we shall omit underlying coordinate indices if it will not result in ambiguities. Such
equations are standard ones for the Lie algebroids but defined on N-anholonomic manifolds. In
brief, we call them Lie N-algebroids.
Definition 2.1: A Clifford algebroid C共E兲 ⯐ 共Cl共E兲, s关· , · 兴, s兲 is associated to a Clifford bundle
Cl共E兲 ⬟ Cl共T*E兲 defined by the vector bundle E = 共E , , M兲 and provided with “spin” anchor s and
共Lie type兲 commutator structure s关· , · 兴 defined on the Clifford module Sec共Cl共M兲兲.
The Clifford algebroid structure on a manifold M is defined C共TM兲 ⯐ 共Cl共TM兲, s关· , · 兴, s兲.
In local form, the spinor structure functions are written
共eáà兲 = áà
i
共x兲ei = iáà共x兲i , 共10兲
ćc̀
关eáà,eb´b`兴 = Cáàb´b`共x兲ećc̀ , 共11兲
where we can consider a spinor decomposition on M with redefinition of indices like i → í , ì. Such
structure functions can be induced by pure spinor ones,
and
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-6 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
ć
关eá,eb´兴 = Cáb´共x兲eć ,
g f
f
Cag 共x,u兲 = e f f 共x,u兲eaa共x,u兲eg 共x,u兲Cag共x兲, 共14兲
respectively, into formulas 共7兲–共9兲. In general, the RC-algebroids are defined by the corresponding
sets of functions ˆ aj共x , y兲 and Cag
f
共x , y兲 with additional dependencies on v-variables y b for the
N-adapted structure functions. For such Lie N-algebroids, the structure relations became
关ed,eb兴 = Cdb
f
共x,y兲e f 共16兲
and the structure equations of the Lie N-algebroid are written
for Q fb ⬘e⬘ b⬘ e⬘ f b⬘
f ⬘bej = e b e e e f ⬘ e j共eb ee e f 兲 with the values e b and e f ⬘ defined by the N-connection. The Lie
b e f f
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-7 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
´ ´
e␣ = 共ei,ea兲 苸 ␥␣ = 共␥i, ␥a兲, ␣´ = 共i, á兲 苸 ␥␣´ = 共␥i, ␥á兲
for, respectively, elements of modules of d-vector and irreduced d-spinor fields 共see details in Ref.
11兲. D-tensors and d-spinor tensors 共irreduced or reduced兲 will be interpreted as elements of
corresponding ␥–modules, for instance,
␣´ ␥´ ␣´ ␥´ II⬘ ⬘
q␣. . . 苸 ␥␣, ´ . . .
苸 ␥ ´ . . .
, JK ⬘N⬘
苸 ␥IIJK ⬘N⬘
, ... .
We can establish a correspondence between the d-metric g␣ 共1兲 and d-spinor metric ⑀␣´ ´ for
both h- and v-subspaces of V by using the relation
1 ´`
g␣ = 共共␥共␣共u兲兲␣´ ␣` 共␥兲共u兲兲⑀␣´ ` ⑀´ ␣` , 共18兲
k̃共n兲 + k̃共m兲
where 共␣兲 denotes symmetrization on such indices and
⑀␣´ ` = 冉 冊
⑀´i j̀
0
0
⑀áb`
on the d-spinor space S = 共S共h兲 , S共v兲兲 may have symmetric or antisymmetric h共v兲 -components ⑀´i j̀
共⑀áb`兲. For simplicity, in this section 共in order to avoid cumbersome calculations connected with
eight fold periodicity on dimensions n and m on a N-anholonomic manifold兲 we shall develop a
general d-spinor formalism only by using irreduced spinor spaces S共h兲 and S共h兲 ⬘ .
1. D-covariant derivation
For a d-covariant operator
´`
D␣ = 共Di,Da兲 = 共␥␣兲␣´ ␣` D␣´ ␣` = 共共␥i兲iiD´i`i,共␥a兲áàDáà兲,
in brief, we shall write
D␣ = D␣´ ␣` = 共D´i`i,Dáà兲,
being constructed by using the coefficients of a d-connection, we define the action on a d-spinor
´
␥ as a map
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-8 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
´ ´ ´
D␣´ ␣` : ␥ → ␥␣ = ␥␣´ ␣`
satisfying conditions
´ ´ ´ ´ ´ ´ ´
D ␣共  + 兲 = D ␣  + D ␣ , D␣共f 兲 = fD␣ + D␣ f ,
´ ´ ´
for every  ,  苸 ␥ and f being a scalar field on V. It is also required that one holds the Leibnitz
rule
´ ´ ´
共D␣´ 兲 = D␣共´ 兲 − ´ D␣
and that D␣ is a real operator, i.e., it commuters with the operation of complex conjugation:
´ ´
共D␣共1兲 − D␣兲: ␥ → ␥␣´ ␣` . 共21兲
Because the action on a scalar f of both operators D␣共1兲 and D␣ must be identical, i.e.
D␣共1兲 f = D␣ f , 共22兲
´
the action 共21兲 on f = ´  must be written as
´
共D␣共1兲 − D␣兲共´ 兲 = 0.
We conclude that there is an element ⌰␣´ ␣` ´ ␥´ 苸 ␥␣´ ␣` ´ ␥´ for which
´
D␣共1兲 ␥´ ␥´ ␥´ 
´ ␣` = D␣´ ␣` + ⌰␣´ ␣` ´ 共23兲
and
D␣共1兲 ␥´
´ ␣` ´ = D␣´ ␣` ´ − ⌰␣´ ␣` ´ ␥´ .
´`
The action of the operator 共21兲 on a d-vector v = v can be written by using formula 共23兲 for both
indices ´ and ` :
´` ´ ` ` ´ ´ ` ` ´
共D␣共1兲 − D␣兲v = ⌰␣␥´ v␥´  + ⌰␣␥` v␥` = 共⌰␣␥´ e␥` + ⌰␣␥` e␥´ 兲v␥´ ␥` = Q␣␥
 ␥
v ,
where
´` ´ ` ` ´
Q␣␥ = Q␣´ ␣` ␥´ ␥` = ⌰␣␥´ e␥` + ⌰␣␥` e␥´ . 共24兲
The commutator D关␣兴D关兴 defines the d-torsion. Applying operators D关共1兲 共1兲
␣兴D关兴 and D关␣兴D关兴 on f
´
= ´ , we can write
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-9 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
`
e␣´ ␣´ e` D␣´ ` = D␣´ ` ,
we compute the components D␣´ ` ␥,
` ´ ␥´
e␣´ ␣´ e` e␥␥´ D␣´ ` ␥´ = e⑀´ ´ e´ ␥´ D␣´ ` ⑀´ + ⑀´ e⑀´ ␥´ D␣´ ` e⑀´ ⑀´ = D␣´ ` ␥´ + ⑀´ ␣´ ` ⑀´ , 共25兲
´ ´ ´ ´ ´
l␣␣e´ D␣ = D␣ + ␦ ␣␦´
共27兲
and
´ ␦´
l␣␣e´ D␣´ = D␣´ − ␦´ ␣´
共28兲
´
for d-covariant derivations D␣ and D␣´ .
We can consider expressions similar to 共27兲 and 共28兲 for values having both types of d-spinor
and d-tensor indices, for instance,
␦´ ␥ ␥ ⑀´
l␣␣l␥␥e␦´ D␣␦´ = D␣␦´ − ⑀␥´ ␣␦´
+ ␦´ ⌫␥␣
␦´
共we can prove this by a straightforward calculation of the derivation D␣共␦´ e␦´ l␥兲兲.
⑀´
Now we shall consider some possible relations between components of d-connections ␣␦´
␥ ␣´ `
and ⌫␣ and derivations of 共␥␣兲 . We can write
␣
= l␣␣Dl␣ = l␣␣D共␥兲⑀´ ` = l␣␣D共共␥兲⑀´ ` e⑀´ ⑀e` 兲 = l␣␣e␣´␣e⑀`⑀D共␥兲␣´ ⑀` + l␣␣共␥兲⑀´ ` 共e` De⑀´ ⑀ + e⑀´ ⑀De` 兲
´ ` ´ ` ` ´ ´ `
⌫
= l␣␣´ ⑀` D共␥兲␣´ ⑀` + l␣␣´ ⑀` e␣´␣e⑀` ⑀共␥兲⑀´ ` 共e` De⑀´ ⑀ + e⑀´ ⑀De` 兲,
´ ` ` ´ ´ `
` `
共␥␣兲´ ` 共␥兲␣´ ` ⌫␣␥ = 共␥兲␣´ ` D␥共␥兲´ ` + e` ␥␣´ + e␣´
´ ´
␥`
.
Contracting the last expression on ` and ` and using an orthonormalized d-spinor basis when
´
␥´
=0 共a consequence from 共26兲兲, we have
1 ´ ` `
␥␣´ = 共␥兲␣´ ` D␥共␥兲´ 兲,
´
共⌫ ␥ ␣´ `
− 共29兲
k̃共n兲 + k̃共m兲
where
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-10 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
´ ` `
⌫ ␥ ␣´ `
= 共␥␣兲´ 共␥兲␣´ ` ⌫␣␥ . 共30兲
The d-spinor connection 共30兲 can be defined by various type of d-connections, inclusively, by the
canonical one, see Ref. 1. Such formulas can be applied on Clifford algebroid
C共E兲 ⯐ 共Cl共E兲 , s关 · , · 兴 , s兲 or on a Clifford N-algebroid C共vV兲 ⯐ 共Cl共vV兲 , s关 · , · 兴 , s兲. We have to
change the v-derivatives into anchored ones, / y a → ia / xa, or in N-adapted form, ea → ˆ aje j,
and put the results in formulas 共29兲 and 共30兲. As a result, one defines a canonical covariant spinor
differential calculus, adapted to the N-connection structure, acting on the set of sections Sec共E兲 or
Sec共vV兲.
␥´
䊐␣´ ` ␥´ = X ´ ␣´ `
´ 共32兲
and
´
䊐␣´ ` ␥´ = X .
␥´ ␣´ ` ´
´
X␥´ ␦´ ␣´ ` = e␥´ ´ 䊐␣´ ` e␦´
and
␥´ ␥`
䊐␣´ ` f = T␣´ ` 䉮␥´ ␥` f .
´`
The d-spinor components R␦␦␥´ ␥` ␣´ ` of the curvature d-tensor R␦␥␣ can be computed by using
relations 共30兲, 共31兲, and 共33兲 and
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-11 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
´` ´`
共䊐␣´ ` − T␥´ ␥` ␣´ ` 䉮␥´ ␥` 兲V␦␦ = R␦␦␥´ ␥` ␣´ ` V␥´ ␥` , 共34兲
here d-vector V␥´ ␥` is considered as a product of d-spinors, i.e., V␥´ ␥` = ␥´ ␥` . We find
´` ´ ´ ` ␦` ` ´
R␦␦␥´ ␥` ␣´ ` = 共X␦␥´ ␣´ ` + T´ ` ␣´ ` ␦´ ` ␥´ 兲e␥`␦ + 共X ␥` ␣´ `
+ T´ ` ␣´ ` ␦´ ` ␥` 兲e␥´␦ .
It is convenient to use this d-spinor expression for the curvature d-tensor in order to get the
d-spinor components of the Ricci d-tensor,
´` ´ ´ ` `´` `
R␥´ ␥` ␣´ ` = R␦␦␥´ ␥` ␣´ ` ␦´ ␦` = X␦␥´ ␣´ ` ␦´ ␥` + T´ ` ␣´ ` ␦´ ␥` ␦´ ` ␥´ + X␦␥` ␣´ ` ␥´ ␦` + T´ ` ␣´ ` ␥´ ␦` ␦´ ` ␥␥` 共35兲
␦␣´ ``
´ ` ␣´
ឈ ´` ` ´ ` ` ´ ``
R = R␦␦␣´ ␣´ ` ␦´ ␦` = X␦␣´ ␣´ ` ␦´  + T´ ` ␣´ ` ␦´ ␦´ ` ␣´ + X␣´ ` ␦` + T␣´ ` ␦` ␦´ ` .
Finally, we write down the d-spinor components of the Einstein d-tensor G␥,
´ ´ ` `
G␥´ ␥` ´ ` = X␦␥´ ´ ` ␦´ ␥` + T´ ` ´ ` ␦´ ␥` ␦´ ` ␥´ + X␦␥` ´ ` ␥´ ␦` + T´ ` ´ ` ␥´ ␦` ␦´ ` ␥`
1 ´ ` ` ´ `` ` `
− ⑀␥´ ` ⑀´ ␥` 关X␦␣´ ␣´ ` ␦´  + T´ ` ␣´ ` ␦´ ␦´ ` ␣´ + X␦␣´ ` ␣´ ␦` + T´ ` ␣´ ` ␣´ ␦` ␦´ ` 兴. 共36兲
2
It should be noted that further reductions of 共35兲 and 共36兲 depend on dimensions n and m of
the, respectively, h- and v-subspaces, and that the symmetry properties are defined by the
⑀-objects. On Clifford N-algebroids, such formulas have to be considered for anchored
v-derivatives 共15兲 and 共7兲 共for d-spinor considerations, we have to apply spinor anchors 共10兲 and
共12兲兲, for instance, in the case of canonical d-connections and their spinor variants 共29兲.
ˆ ˆ
ei ⬟ eii共x,y兲ei, ei ⬟ eii共x,y兲ei , 共37兲
where
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-12 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
ˆ ˆ ˆˆ
gij共x,y兲eii共x,y兲ej 共x,y兲 = ␦i , gij共x,y兲eii共x,y兲e jj共x,y兲 = gij共x,y兲.
We define the algebra of Dirac’s gamma matrices 共in brief, h-gamma matrices defined by self-
adjoint matrices M k共C兲 where k = 2n/2 is the dimension of the irreducible representation of Cl共M兲
for even dimensions, or of Cl共M兲+ for odd dimensions兲 from the relation
ˆ ˆ ˆ ˆ ˆˆ
␥i␥ + ␥ ␥i = 2␦i I. 共38兲
We can consider the action of dxi 苸 Cl共M兲 on a spinor 苸 S via representations
ˆ ˆ i ˆ
−
c共dxi兲 ⬟ ␥i, −
c共dxi兲 ⬟ ␥i ⬅ e ˆi␥i . 共39兲
For any type of spaces TxM , TM or V possessing a local 共in any point兲 or global fibered
structure and enabled with a N-connection structure, we can introduce similar definitions of the
gamma matrices following algebraic relations and metric structures on fiber subspaces,
where
ˆ ˆ
gab共x,y兲eâa共x,y兲ebb共x,y兲 = ␦âb, gab共x,y兲eaa共x,y兲ebb共x,y兲 = hab共x,y兲.
Similarly, we define the algebra of Dirac’s matrices related to typical fibers 共in brief, v-gamma
matrices described by self-adjoint matrices M k⬘共C兲 where k⬘ = 2m/2 is the dimension of the irreduc-
ible representation of Cl共F兲 for even dimensions, or of Cl共F兲+ for odd dimensions, of the typical
fiber F兲 from the relation
ˆ ˆ ˆ
␥â␥b + ␥b␥â = 2␦âbI. 共41兲
The formulas 共38兲 and 共41兲 are respectively, the h- and v-components of relation 共18兲 共with
redefined coefficients which is more convenient for further constructions兲. The action of dy a
苸 Cl共F兲 on a spinor 쐓 苸 쐓S is considered via representations
쐓
c共dy â兲 ⬟ ␥â, 쐓
c共dy a兲 쐓 ⬟ ␥a 쐓 ⬅ eaâ␥â 쐓 . 共42兲
We note that in additionally to formulas 共39兲 and 共42兲 we may write, respectively,
ˆ
c共dy a兲 쐓 ⬟ ␥a 쐓 ⬅ eea â␥â 쐓
a
c共dxi兲 ⬟ ␥i ⬅ eiˆi␥i,
â
where
ˆ ˆ
g␣共u兲e␣ˆ␣共u兲e共u兲 = ␦␣ˆ , g␣共u兲e␣␣共u兲e共u兲 = g␣共u兲.
The second step is to consider gamma d-matrix relations 共unifying 共38兲 and 共41兲兲
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-13 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
ˆ ˆ ˆ
␥␣ˆ ␥ + ␥␥␣ˆ = 2␦␣ˆ I, 共44兲
with the action of du␣ 苸 Cl共V兲 on a d-spinor ˘ 苸 S resulting in distinguished irreducible represen-
tations 共unifying 共39兲 and 共42兲兲
In the canonical representation, we can write in irreducible form ␥˘ ⬟ ␥ 丣 쐓␥ and ˘ ⬟ 丣 쐓, for
instance, by using block type of h- and v-matrices, or, writing alternatively as couples of gamma
and/or h- and v-spinor objects written in N-adapted form,
1ⵜ i
ⵜS ⬟ d − ⌫ ␥i␥ jdxk . 共48兲
4 jk
On N-anholonomic spaces, it is possible to define spin connections which are N-adapted by
replacing the Levi-Civita connection by any d-connection.
The canonical spin d-connection is defined by the canonical d-connection,
1
ⵜˆ S ⬟ ␦ − ⌫ˆ ␣␥␣␥␦u , 共49兲
4
where the absolute differential ␦ acts in N-adapted form resulting in one-forms decomposed with
respect to N-elongated differentials ␦u = 共dxi , ␦y a兲 共2兲.
We note that the canonical spin d-connection ⵜ ˆ S is metric compatible and contains nontrivial
d-torsion coefficients induced by the N-anholonomy relations. It is possible to introduce more
general spin d-connections DS by using the same formula 共49兲 but for arbitrary metric compatible
␣
d-connection ⌫ . For the spaces provided with generic off-diagonal metric structure 共1兲 on a
N-anholonomic manifold, there is a canonical spin d-connection 共49兲 induced by the off-diagonal
metric coefficients with nontrivial Nai and associated nonholonomic frames in gravity theories.
In a particular case of N-anholonomic manifolds of even dimensions, we can define, for
instance, the canonical spin d-connections for a local modeling of a tangent bundle space with the
canonical d-connection ⌫ ˆ ␥ = 共L̂i , B̂i 兲. The N-connection structure N j states a global h- and
␣ jk jk i
v-splitting of the spin d-connection operators, for instance,
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-14 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
−ˆ ␦ k1 i
ⵜ ⬟ dxi i − dx L̂ jk␥i␥
j
共51兲
x 4
and
쐓ˆ c1 a
ⵜ ⬟ ␦ y a a − ␦y B̂ ␥a␥b
y 4 bc
being defined by the canonical d-connection, which 共in its turn兲 is completely defined by Nij共x , y兲
and gij共x , y兲.
The operators 共51兲 can be adapted to the Lie algebroid structure by anchoring the partial
v-derivatives. For instance,
␦ k b j
x i 共x ,y 共x 兲兲 =
x
a
i − Ni
y a =
x
冉 a k j
i − Ni a共x 兲
x
冊冉
k =
k
xi
冊
− Ni k 共xk,y b共x j兲兲
x
where the anchor ka 共7兲 induces a N-connection Nki ⯐ Nai ka. We can also perform a N-adapted
Clifford algebroid calculus by using the “boldface” algebroid ˆ aj 共15兲 with explicit dependence on
variables y b,
␦ k b
xi
共x ,y 兲 = i − Nai ea =
x
xi
冉
− Nai ˆ ka k =
x
冊冉
k
xi
− N̂i k
x
冊
for N̂ki = Nai ˆ ka. Such anchoring of partial/N-elongated derivatives has to be considered for the
canonical d-connection L̂ijk and B̂bc a
.
2. Dirac d-operators
We consider a vector bundle E on an N-anholonomic manifold M 共with two compatible
N-connections defined as h- and v-splittings of TE and TM兲兲. A d-connection
D共f 兲 = f共D兲 + ␦ f 丢
for any f 苸 C⬁共M兲, and 苸 Sec⬁共E兲 and ␦ defining an N-adapted exterior calculus by using
N-elongated operators 共3兲 and 共2兲 which emphasize d-forms instead of usual forms on M, with the
coefficients taking values in E.
The metricity and Leibniz conditions for D are written, respectively,
where fiber “inverse-caret” indices, in their turn, may split ␣ˇ ⬟ 共ǐ , ǎ兲 if any N-connection structure
is defined on TE. For some constructions of particular interest, we can take E = T*V , = T*V共g兲
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-15 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
and/or any Clifford d-algebra E = Cl共V兲 , Cl共V共g兲兲 , . . . with a corresponding treating of “acute” in-
dices to d-tensor and/or d-spinor type as well when the d-operator D transforms into respective
d-connection D and spin d-connections ⵜ ˆ S 共49兲, ⵜˆ 共g兲 . . . . All such, adapted to the N-connections,
computations are similar for both N-anholonomic 共co兲 vector and spinor bundles.
The respective actions of the Clifford d-algebra and the Clifford-Lagrange algebra can be
transformed into maps Sec⬁共S兲 丢 Sec共Cl共V兲兲 and Sec⬁共S共g兲兲 丢 Sec共Cl共V共g兲兲兲 to Sec⬁共S兲 and, re-
spectively, Sec⬁共S共g兲兲 by considering maps of type 共39兲 and 共45兲
D ⯐ − i共ĉ ⴰ ⵜS兲
쐓
=共共g兲 −D = − i共 −ĉ ⴰ −ⵜ共g兲兲, 共g兲 D = − i共 쐓ĉ ⴰ 쐓ⵜ共g兲兲兲. 共55兲
Such N-adapted Dirac d-operators are called canonical and denoted D̂ = 共 −D̂ , −쐓D̂兲 共共g兲 共g兲D̂
쐓
= 共共g兲 −D̂ , 共g兲 D̂兲兲 if they are defined for the canonical d-connection and respective spin d-connection
共49兲.
Now we can formulate the
Theorem 3.1: Let 共V , S , J兲 共共M 共g兲 , S共g兲 , J兲 be a spin N-anholonomic manifold. There is the
canonical Dirac d-operator 共Dirac N-anholonomic operator兲 defined by the almost Hermitian
spin d-operator
ˆ S˘ 兩
共ⵜ ˘ 兲 = ␦共˘ 兩
˘ 兲 + 共˘ 兩ⵜˆ S ˘兲 共56兲
and
for a 苸 Cl共M 共g兲兲 and ˘ 苸 Sec⬁共S共g兲兲 determined by the metricity 共52兲 and Leibnitz 共53兲
conditions.22,23
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-16 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
Proof: We sketch the main idea of such a proof being similar to that given in Ref. 24, Theorem
9.8, for the Levi-Civita connection, see also Ref. 25. In our case, we have to extend the construc-
tions for d-metrics and canonical d-connections by applying N-elongated operators for differen-
tials and partial derivatives and distinguishing the formulas into h- and v-irreducible components.
䊐
The canonical Dirac d-operator has very similar properties for spin N-anholonomic manifolds
and spin Lagrange, or Finsler spaces.11–13,7
冉 䊐+
n+m−2 ឈ
4共n + m − 1兲
冊
R + 2 共u兲 = 0. 共58兲
We have to elongate the covariant d-operator, D̆␣ = D␣ + ieA␣, and take into account the d-vector
current
冋
L共0兲共u兲 = 冑兩g兩 g␣e␣共u兲e共u兲 − 2 + 冉 n+m−2
4共n + m − 1兲
冊
共u兲共u兲 , 册 共59兲
where 兩g 兩 = det兩 g␣兩 and e␣ is defined by 共3兲, and must be anchored for Lie algebroid structures.
The N-adapted variations of the action with Lagrangian 共59兲 on variables 共u兲 and 共u兲 lead
to the energy-momentum d-tensor,
共0,c兲 1 共0兲
共u兲 = e␣共u兲e共u兲 + e共u兲e␣共u兲 −
E␣
冑兩g兩 g␣L 共u兲, 共60兲
and a similar variation on the components of a d-metric 共1兲 leads to a symmetric energy-
momentum d-tensor,
共0兲 n+m−2
E␣ 共u兲 = E共共0,c兲
␣兲共u兲 − 关R共␣兲 + D共␣D兲 − g␣ 䊐 兴共u兲共u兲. 共61兲
2共n + m − 1兲
We also conclude that the N-connection results in a nonequivalence of energy–momentum
d-tensors 共60兲 and 共61兲, nonsymmetry of the Ricci tensor, non-vanishing of the d-covariant deri-
ឈ ␣ ⫽ 0 and, in consequence, a corresponding modification of
vation of the Einstein d-tensor, D␣G
conservation laws on N-anholonomic manifolds.27
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-17 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
2. Proca equations
Let us consider a d-vector field ␣共u兲 with mass 2 共Proca field兲 interacting with exterior
gravitational field. From the Lagrangian
f ␣ = D␣ − D␣ ,
one follows the Proca equations on N-anholonomic manifolds
D␣␣共u兲 = 0. 共64兲
Putting 共64兲 into 共63兲 we obtain second-order field equations with respect to ␣:
共1兲 g␣
共u兲 = − g共f  f ␣ + f ␣ f 兲 + 2共␣ + ␣兲 − 共1兲
E␣
冑兩g兩 L 共u兲.
and
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-18 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
␣ 关
␣ 4 ␣␥ ␣ 兴
ជ = e + 1 C ˆ ˆ ˆ 共u兲l␣ˆ 共u兲␥ˆ ␥␥ˆ ,
䉮 共68兲
where coefficients C␣ˆ ˆ ␥ˆ = 共D␥l␣␣ˆ 兲lˆ ␣l␥␥ˆ generalize for N-anholonomic spaces the corresponding
Ricci coefficients on Riemannian spaces. Using ␥-objects ␥␣共u兲 共see 共19兲兲, we define the Dirac
equations on N-anholonomic manifolds:
ជ − 兲 = 0,
共i␥␣共u兲䉮 共69兲
␣
L共1/2兲共u兲 = 冑兩g兩兵关+共u兲␥␣共u兲䉮
ជ 共u兲 − 共䉮
␣
ជ +共u兲兲␥␣共u兲共u兲兴 − +共u兲共u兲其,
␣ 共70兲
where 共u兲 is the complex conjugation and transposition of the column 共u兲. We have to con-
+
J␣共1/2兲共u兲 = +共u兲␥␣共u兲共u兲.
We emphasize that interactions with exterior gauge fields can be introduced by changing the
locally anisotropic partial derivation from 共68兲 in this manner:
e␣ → e␣ + ie쐓B␣ , 共71兲
where e쐓 and B␣ are, respectively, the constant and the d-vector potential of gauge fields.
K␣⍀⌰⍀ = 共䉮 ␣䉮  − 䉮 䉮 ␣兲⌰ .
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-19 Clifford-Finsler algebroids and Einstein–Dirac structures J. Math. Phys. 47, 093504 共2006兲
For Yang–Mills fields, as a rule, one considers that BE is enabled with a unitary 共complex兲
structure 共complex conjugation changes mutually the upper and lower Greek indices兲. It is useful
to introduce instead of K␣⍀⌰ a Hermitian matrix F␣⍀⌰ = i K␣⍀⌰ connected with components of
the Yang–Mills d-vector potential B␣⌶⌽ according the formula:
1
F␣⌶⌽ = 䉮 关␣B⌶⌽ − iB关⌽␣兩⌳兩B⌶⌳ , 共72兲
2
where the spacetime indices commute with capital Greek indices. The gauge transforms are
written in the form:
⌰
B␣⌰⌽ 哫 B␣⌰ˆ ⌽̂ = B␣⌽⌰s⌽
⌽̂
q⌰ˆ + is⌽̂ ˆ⌰
⌰䉮 ␣q ⌰ ,
⌽̂
F␣⌶⌽ 哫 F␣⌶̂⌽̂ = F␣⌶⌽s⌽ q⌶̂⌶ ,
⌽̂ ⌶
where matrices s⌽ and q⌶̂ are mutually inverse 共Hermitian conjugated in the unitary case兲. The
Yang–Mills d-equations are written
䉮 关␣F␥⌰⌶ = 0. 共74兲
We must introduce deformations of connection of type, 䉮␣쐓 − → 䉮␣ + P␣, 共the deformation d-tensor
P␣ is induced by the torsion in the vector bundle BE兲 into the definition of the curvature of gauge
fields 共72兲 and motion equations 共73兲 and 共74兲 if the interactions are considered for nontrivial
torsions.
Among the subjects we will study in forthcoming papers, we note the following points:
• To construct exact solutions of the gravitational field equations in string gravity with
nontrivial limits to general relativity, parametrized by generic off-diagonal metrics and
nonholonomic frames and possessing Lie algebroid symmetries 共the first examples of
“gravitational” algebroids were analyzed in Ref. 2兲.
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp
093504-20 Sergiu I. Vacaru J. Math. Phys. 47, 093504 共2006兲
Finally, we note that the method of anholonomic frames with associated N-connection struc-
ture elaborated in Finsler geometry and further, in our works, applied to constructing exact solu-
tions in gravity was applied in this paper for a study of Dirac operators on nonholonomic mani-
folds possessing Lie algebroid symmetry. The constructions can be extended for spacetimes with
uncompactified extra dimensions and such investigations are regarded as interesting research in
modern physics and noncommutative geometry.
ACKNOWLEDGMENT
The author is grateful to the referee for hard work and constructive critics.
1
S. Vacaru, J. Math. Phys. 46, 042503 共2005兲.
2
S. Vacaru, gr–qc/ 0501057.
3
F. Etayo, R. Santamaría, and S. Vacaru, J. Math. Phys. 46, 032901 共2005兲.
4
A. Cannas da Silva and A. Weinstein, Geometric Models for Noncommutative Algebras 共American Mathematical Soci-
ety, Providence, RI, 1999兲.
5
R. Miron and M. Anastasiei, The Geometry of Lagrange Spaces: Theory and Applications 共Kluwer, Dordrecht, 1994兲.
6
S. Vacaru, E. Gaburov, and D. Gontsa, hep-th/0310133; in Clifford and Riemann- Finsler Structures in Geometric
Mechanics and Gravity, Selected Works, Differential Geometry – Dynamical Systems Monograph 7 共Geometry Balkan
Press, Bucharest, 2006兲, Chap 2; www.mathem.pub.ro/dgds/mono/va-t.pdfand gr-qc/0508023
7
S. Vacaru, math.DG/0408121; Clifford and Riemann- Finsler Structures in Geometric Mechanics and Gravity, Selected
Works, Differential Geometry – Dynamical Systems Monograph 7 共Geometry Balkan Press, Bucharest, 2006兲, Chap. 15;
www.mathem.pub.ro/dgds/mono/va-t.pdfand gr-qc/0508023
8
S. Vacaru and O. Tintareanu-Mircea, Nucl. Phys. B 626, 239 共2002兲.
9
S. Vacaru and F. C. Popa, Class. Quantum Grav. 18, 4921 共2001兲.
10
S. Vacaru and N. Vicol, Int. J. Math. Math. Sci. 23, 1189 共2004兲.
11
S. Vacaru, J. Math. Phys. 37, 508 共1996兲.
12
S. Vacaru, J. High Energy Phys. 9809, 011 共1998兲.
13
S. Vacaru and P. Stavrinos, Spinors and Space-Time Anisotropy 共Athens University Press, Athens, Greece, 2002兲,
gr-qc/0112028.
14
A. Weinstein, Curr. Appl. Phys. 7, 201 共1996兲.
15
P. Libermann, Arch. Math. 32, 147 共1996兲.
16
E. Martinez, Acta Appl. Math. 67, 295 共2001兲.
17
M. de Leon, J. C. Marrero, and E. Martinez, math.DG/0407528.
18
S. Vacaru, Phys. Lett. B 498, 74 共2001兲.
19
R. Penrose and W. Rindler, Two-Spinor Calculus and Relativistic Fields Spinors and Space-Time Vol. 1 共Cambridge
University Press, Cambridge, 1984兲.
20
R. Penrose and W. Rindler, Spinor and Twistor Methods in Space-Time Geometry Spinors and Space-Time Vol. 2
共Cambridge University Press, Cambridge, 1986兲.
21
C. P. Luehr and M. Rosenbaum, J. Math. Phys. 15, 1120 共1974兲.
22
A. Salam and J. Strathdee, Ann. Phys. 共N.Y.兲 141, 316 共1982兲.
23
J. M. Overduin and P. S. Wesson, Phys. Rep. 283, 303 共1997兲.
24
J. M. Gracia–Bondia, J. C. Varilli, and H. Figueroa, Elements of Noncommutative Geometry 共Birkhauser, Boston, 2001兲.
25
H. Schroeder, math. DG/ 0005239.
26
A. Ashtekar, Phys. Rev. D 36, 1587 共1987兲.
27
J. Kern, Arch. Math. 25, 438 共1974兲.
Downloaded 08 Nov 2006 to 128.100.216.128. Redistribution subject to AIP license or copyright, see http://jmp.aip.org/jmp/copyright.jsp