ebook img

Higher derivative corrections to Wess-Zumino action of Brane-Antibrane systems PDF

0.2 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Higher derivative corrections to Wess-Zumino action of Brane-Antibrane systems

Higher derivative corrections to Wess-Zumino action of Brane-Antibrane systems 8 0 0 2 Mohammad R. Garousi n a J 9 Department of Physics, Ferdowsi university, P.O. Box 1436, Mashhad, Iran 2 Institute for Studies in Theoretical Physics and Mathematics (IPM) ] P.O.Box 19395-5531, Tehran, Iran h t - p e h [ 3 Abstract v 4 5 By explicit calculation, we show that the expansion of the disk level S-matrix 9 1 element of one RR field, two open string tachyons and one gauge field that has . 2 beenrecently foundcorresponds tothederivative expansion oftheWess-Zumino 1 7 action of D-brane-anti-D-brane systems. 0 : v i X r a 1 Introduction Study of unstable objects in string theory might shed new light in understanding properties of string theory in time-dependent backgrounds [1, 2, 3, 4, 5, 6, 7]. Generally speaking, source of instability in these processes is appearance of some tachyonic modes in the spec- trum of these objects. It then makes sense to study them in a field theory which includes those modes. In this regard, it has been shown by A. Sen that an effective action of the Dirac- Born-Infeld type proposed in [8, 9, 10, 11] can capture many properties of the decay of non-BPS D -branes in string theory [2, 3]. This action has been found in [9] by studying p the S-matrix element of one graviton and two tachyons. Recently, unstable objects have been used to study spontaneous chiral symmetry break- ing in holographic model of QCD [12, 13, 14]. In these studies, flavor branes introduced by placing a set of parallel branes and antibranes on a background dual to a confining color theory [15]. Detailed study of brane-anti-brane system reveals when branes separation is smaller than the string length scale, the spectrum has two tachyonic modes [16]. The ef- fective action should then include these modes as they are the most important ones which rule the dynamics of the system. The effective action of a D D¯ -brane in Type IIA(B) theory should be given by some p p extension of the DBI action and the WZ terms which include the tachyon fields. The DBI part may be given by the projection of the effective action of two non-BPS D -branes in p Type IIB(A) theory with ( 1)FL projection [17]. We are interested in this paper in the − appearance of tachyon, gauge field and the RR field in these actions. These fields appear in the DBI part as the following [18]: S = T dp+1σTr V( ) det(η +2πα′F +2πα′D D ) , (1) DBI p ab ab a b − T − T T Z (cid:18) q (cid:19) where T is the p-brane tension. The trace in the above action should be completely p symmetric between all matrices of the form F ,D , and individual of the tachyon ab a T T potential. These matrices are (1) F 0 0 D T 0 T F = ab , D = a , = (2) ab 0 Fa(b2)! aT (cid:18)(DaT)∗ 0 (cid:19) T (cid:18)T∗ 0 (cid:19) where F(i) = ∂ A(i) ∂ A(i) and D T = ∂ T i(A(1) A(2))T. The tachyon potential which ab a b − b a a a − a − a is consistent with S-matrix element calculations has the following expansion: 1 V( ) = 1+πα′m2 2 + (πα′m2 2)2 + T T 2 T ··· 1 where m2 is the mass squared of tachyon, i.e., m2 = 1/(2α′). The above expansion is − consistent with the potential V( ) = eπα′m2T2 which is the tachyon potential of BSFT [19]. T This action has the following expansion: ′ πα = 2T T (2πα′) m2 T 2+DT (DT)∗ F(1) F(1) +F(2) F(2) + (3) DBI p p L − − | | · − 2 · · ! ··· (cid:16) (cid:17) where dots refers to the terms which have more than two fields. The WZ term describing the coupling of RR field to tachyon and gauge field of brane- anti-brane is given by [20, 21, 22] S = µ C STrei2πα′F (4) WZ p ∧ ZΣ(p+1) where the curvature of the superconnection is defined as: = d i (5) F A− A∧A the superconnection is iA(1) βT∗ i = , A βT iA(2) ! where β is a normalization constant. If one uses the tachyon DBI action (1) for describing the dynamics of the tachyon field then the normalization of tachyon in the WZ action (4) has to be [23] 1 2ln(2) β = (6) πs α′ The “supertrace” in (4) is defined by A B STr = TrA TrD . C D ! − Using the multiplication rule of the supermatrices [21] A B A′ B′ AA′ +( )c′BC′ AB′ +( )d′BD′ C D !· C′ D′ ! = DC′ +(−)a′CA′ DD′ +(−)b′CB′ ! (7) − − ′ where x is 0 if X is an even form or 1 if X is an odd form, one finds that the curvature is iF(1) β2 T 2 β(DT)∗ i = − | | , F βDT iF(2) β2 T 2 ! − | | 2 where F(i) = 1F(i)dxa dxb and DT = [∂ T i(A(1) A(2))T]dxa. Using the expansion 2 ab ∧ a − a − a for the exponential term in the WZ action (4), one finds many different terms. The terms which involve at most three open string fields are the following: µ (2πα′)C STri = µ (2πα′)C (F(1) F(2)) (8) p p p−1 ∧ F ∧ − µ µ p(2πα′)2C STri i = p(2πα′)2C F(1) F(1) F(2) F(2) p−3 2! ∧ F ∧ F 2! ∧ ∧ − ∧ n o +C 2β2 T 2(F(1) F(2))+2iβ2DT (DT)∗ p−1 ∧ − | | − ∧ µp(2πα′)3C STri i i = µp(2πα′)n3C 3iβ2(F(1) +F(2)) DT (DT)∗ o p−3 3! ∧ F ∧ F ∧ F 3! ∧ ∧ n o The coupling of one RR field C , two tachyons and one gauge field in the above terms p−1 can be combined into the following form: β2µ (2πα′)2 C d(A(1) A(2))TT∗ (A(1) A(2))d(TT∗) − p Σ(p+1) (p−1) ∧ − − − =R β2µ (2πα′)2n H (A(1) A(2))TT∗ o (9) − p Σp+1 (p) ∧ − R This combination actually appears naturally in the S-matrix element in the string theory side [23]. It has been shown in [23] that the effective actions (1) and (4) are consistent with an expansion of the S-matrix element of one RR, two tachyons and one gauge field. In the present paper, we would like to show that the expansion found in [23] is in fact the momentum expansion, i.e., the expansion which is consistent with the derivative expansion of the field theory of brane-anti-brane system. We will show this by explicitly calculating the higher derivative terms of the WZ field theory, i.e., (29), (35), and (44). Anoutlineoftherest ofpaper isasfollows. Inthenext section, westudy themomentum expansion of the S-matrix element of one RR and two tachyons, and the S-matrix element of one RR and two gauge fields. We shall find in this section the higher derivative extension of the coupling in the second line of (8) and the coupling in the last term in the third line of (8). In section 3, we study the momentum expansion of the S-matrix element of one RR, two tachyons and one gauge field. We shall find the higher derivatives extension of the coupling in the last line of (8) and the coupling in the first term in the third line of (8). In this section we will also find a class of higher derivative terms, i.e., (44) which has at least four derivatives, hence, they are not higher derivative extension of the two derivative couplings of the WZ terms. We discuss briefly our results in section 4, and give a general rule for finding the momentum expansion of any S-matrix element involving the tachyon fields. 3 2 Three-point function The three-point amplitude between one RR field and two tachyons in string theory side is given as [20, 23] iµ Γ[ 2u] AT,T,RR = 4p 2πΓ[1− u]2Tr(P−H/(n)Mpγa)ka . (10) (cid:18) (cid:19) 2 − where u = (k+k′)2 and k, k′ are the momenta of the tachyons. In the string theory side − ′ we have set α = 2. The trace is zero for p = n, and for n = p it is 6 32 Tr H/ M γa = H ǫa0...ap−1a . (n) p ±p! a0...ap−1 (cid:16) (cid:17) Wearegoingtocomparestring theoryS-matrixelements withfieldtheoryS-matrixelement including their coefficients, however, we are not interested in fixing the overall sign of the amplitudes. Hence, in above and in the rest of equations in this paper, we have payed no attention to the sign of equations. The trace in (10) containing the factor of γ11 ensures the following results also hold for p > 3 with H H for n 5. The tachyon vertex (n) (10−n) ≡ ∗ ≥ operator in string theory corresponds to the real components of the complex tachyon of field theory, i.e., 1 T = (T +iT ) (11) 1 2 √2 ′ Now if one replaces k in (10) with k p using the conservation of momentum, one will a − a− a find that the p term vanishes using the totally antisymmetric property of ǫa0...ap−1a. Hence a the amplitude (10) is antisymmetric under interchanging 1 2. This indicates that only ↔ the three-point amplitude between one RR, one T and one T is non-zero. 1 2 The momentum expansion of (10) is at u 0. Using the Maple, one can expand the → prefactor of (10) around this point, i.e., ∞ Γ[ 2u] 1 2π − = − + a uu . (12) Γ[1 u]2 u n 2 − nX=0 where some of the coefficients a are n a = 4ln(2) 0 π2 a = 8ln(2)2 1 6 − 2 a = π2ln(2) 3ζ(3) 16ln(2)3 (13) 2 −3 − − (cid:16) (cid:17) 4 1 a = 160π2ln(2)2 +3π4 960ζ(3)ln(2) 1280ln(2)4 3 120 − − 1(cid:16) (cid:17) a = 160π2ln(2)3 +9π4ln(2) 1440ζ(3)ln(2)2 +30ζ(3)π2 768ln(2)5 540ζ(5) 5 −90 − − − (cid:16) (cid:17) It is shown in [20] that the massless pole reproduced by the kinetic term of tachyon and the WZ coupling in the first line of (8). There is no higher power of momenta in the massless pole, hence, the kinetic term of tachyon and the WZ coupling have no higher derivative extension. Since the expansion (12) is in terms of the powers of p2, the other terms in (12) a correspond to the higher derivative corrections of the WZ action. It is easy to check that the following higher derivative terms reproduce the other terms in (12): ∞ ′ n α 2iα′µ a C (DaD )n(DT DT∗) (14) p n p−1 a 2 ! ∧ ∧ n=0 X The abovecouplings have on-shell ambiguity, since onecanreplaceT with∂a∂ T for theon- a shell external tachyon. However, we will show in the next section that the above couplings, with exactly the same coefficients a , appear in the contact terms aswell as inthe tachyonic n poleofthescattering amplitudeofoneRR,two tachyons andonegaugefield. This indicates that there is no on-shell ambiguity in the above couplings. We shall discus it more in the Discussion section. So, the above couplings are the higher derivative extension of the coupling in the second term in the third line of (8). The string theory S-matrix element of one RR and two gauge fields is given by [24, 25] Γ[ 2u] 2 − K (15) A ∼ Γ[1 u]2 − where K is the kinematic factor. Obviously the momentum expansion of this amplitude is around u 0. Expansion of the prefactor at this point is → ∞ Γ[ 2u] 2 − = b un . (16) n Γ[1 u]2 − − nX=−1 where some of the coefficients b are n b = 1 −1 b = 0 0 π2 b = 1 6 b = 2ζ(3) 2 5 19 b = π4 (17) 3 360 1 b = ζ(3)π2+18ζ(5) 4 3 1(cid:16) (cid:17) b = 55π6 +6048ζ(3)2 5 3024 (cid:16) (cid:17) In this case actually there is no massless pole at u = 0 as the kinematic factor provides a compensating factor of u. The amplitude has the following expansion: (4π)2µ ∞ A = i4(p 3p)!fa0a1f′a2a3εa4···apǫa0···ap  bnun+1δp,n+2 (18) − nX=−1   ′ ′ ′ ′ ′ where f = i(k ξ k ξ ), f = i(k ξ k ξ ) and ε is the polarizationof the RR potential. ab a b− b a ab a b− b a The above terms are reproduced by the following higher derivative couplings of field theory ∞ µ 2p!(2πα′)2Cp−3 ∧ bn(α′)n+1∂a1 ···∂an+1F(1) ∧∂a1 ···∂an+1F(1) − F(1) → F(2) (19) nX=−1 (cid:16) (cid:17)   We will see in the next section that these couplings, with exactly the same coefficients b , n appear in the massless pole of the scattering amplitude of one RR, two tachyons and one gauge field. The above couplings are the higher derivative extension of the coupling in the second line of (8). 3 Four-point function The S-matrix element of one RR field, two tachyons and one gauge field is given as [23] iµ ATTC = p Tr (P H/ M )(k .γ)(k .γ)(ξ.γ) Iδ +Tr (P H/ M )γa Jδ A 2√2π − (n) p 3 2 p,n+2 − (n) p p,n (cid:20) (cid:18) (cid:19) (cid:18) (cid:19) k (t+1/4)(2ξ.k )+k (s+1/4)(2ξ.k ) ξ (s+1/4)(t+1/4) (20) 2a 3 3a 2 a × − (cid:26) (cid:27)(cid:21) where I, J are : Γ( u)Γ( s+1/4)Γ( t+1/4)Γ( t s u) I = 21/2(2)−2(t+s+u)−1π − − − − − − Γ( u t+1/4)Γ( t s+1/2)Γ( s u+1/4) − − − − − − Γ( u+1/2)Γ( s 1/4)Γ( t 1/4)Γ( t s u 1/2) J = 21/2(2)−2(t+s+u+1)π − − − − − − − − − Γ( u t+1/4)Γ( t s+1/2)Γ( s u+1/4) − − − − − − where the Mandelstam variables are s = (k +k )2, t = (k +k )2, u = (k +k )2 1 3 1 2 2 3 − − − 6 k is momentum of the gauge field and k , k are the momenta of the tachyons. Note that 1 2 3 I, J are symmetric under s t. The traces in (20) are: ↔ 32 Tr H/(n)Mp(k3.γ)(k2.γ)(ξ.γ) δp,n+2 = ±n!ǫa0···apHa0···ap−3k3ap−2k2ap−1ξapδp,n+2 (cid:18) (cid:19) 32 Tr H/(n)Mpγa δp,n = ±n!ǫa0···ap−1aHa0···ap−1δp,n (21) (cid:18) (cid:19) Examining the poles of the Gamma functions, one realizes that for the case that p = n+2, the amplitude has massless pole and infinite tower of massive poles. Whereas for p = n case, there are tachyon, massless, and infinite tower of massive poles. The tachyon pole in particular indicates that the kinetic term of the tachyon has no higher derivative extension. Ithasbeenshown in[23]thattheleadingorder termoftheamplitude(20)expandedaround the following point : t 1/4, s 1/4, u 0 (22) → − → − → is consistent with the effective actions (1) and (4). We would like to find the field theory couplings which reproduce all terms of the expansion. Let us study each case separately. 3.1 p = n + 2 case For p = n+2, the amplitude is antisymmetric under interchanging 2 3, hence the four- ↔ point function between one RR, one gauge field and two T or two T is zero. The electric 1 2 part of the amplitude for one RR, one gauge field, one T and one T is given by 1 2 8iµ AAT1T2C = ±√2π(p p 2)! ǫa0···apHa0···ap−3k3ap−2k2ap−1ξap I (23) − h i Note that the amplitude satisfies the Ward identity, i.e., the amplitude vanishes under replacement ξa ka. → 1 Expansion of I around (22) is ∞ 1 I = π√2π b (s+t+1/2)n+1+ (24) n −u n=−1 X ∞ + c up((s+1/4)(t+1/4))n(s+t+1/2)m p,n,m  p,n,m=0 X  where the coefficients b are exactly those that appear in (17) and c = a are those that n p,0,0 p appear in (13). The constants c for some other cases are the following: p,n,m 2 c = π2ln(2), c = 14ζ(3),c = 8ζ(3)ln(2), (25) 0,0,2 0,1,0 0,0,3 3 − 7 1 c = 56ζ(3)ln(2) 1/2, c = (π4 48π2ln(2)2), c = 1/2 1,1,0 1,0,2 0,1,1 − 36 − − Inserting thefirst termof(24) into (23), onefinds amassless polewhich must bereproduced by field theory couplings. The couplings in (3) and (19), produces the following massless pole for p = n+2: = V (C ,A(1),A(1))G (A(1))V (A(1),T ,T ) (26) a p−3 ab b 1 2 A where iδ G (A(1)) = ab ab (2πα′)2T (u) p V (A(1),T ,T ) = T (2πα′)(k k ) (27) b 1 2 p 2 3 b − ∞ 1 Va(Cp−3,A(1),A(1)) = µp(2πα′)2(p 2)!ǫa0···ap−1aHa0···ap−3k1ap−2ξap−1 bn(α′k1 ·k)n+1 − nX=−1 where k is the momentum of the off-shell gauge field. Note that the vertex V (A(1),T ,T ) b 1 2 has no higher derivative correction as it arises from the kinetic term of the tachyon. The amplitude (26) becomes ∞ ′ n+1 2i α A = µp(2πα′)(p−2)!uǫa0···ap−1aHa0···ap−3k2ap−2k3ap−1ξanX=−1bn 2 ! (s+t+1/2)n+1 (28) this is exactly the massless pole of string theory amplitude. Note that there is no left over residual contact term in comparing above amplitude with the massless pole of the string theory amplitude. The contact terms of string theory amplitude (24) on the other hand are reproduced by the following couplings: ∞ ′ p α 2iα′(πα′)µp cp,n,m (α′)2n+mCp−3 ∂a1 ∂a2n∂b1 ∂bm(F(1) +F(2)) 2 ! ∧ ··· ··· p,n,m=0 X (DaD )pD D (D D DT D D DT∗) (29) ∧ a b1··· bm a1 ··· an ∧ an+1 ··· a2n For n = m = 0 case, the above couplings are the natural extension of the couplings (14) to C . Since there is no on-shell ambiguity for the couplings in (14), one expects there p−3 should be no on-shell ambiguity for the above couplings either. The above couplings are the higher derivative extension of the coupling in the fourth line of (8). 8 3.2 p = n case Now we consider n = p case. The string theory amplitude in this case is symmetric under interchanging 2 3. On the other hand, there is no Feynman amplitude in field theory ↔ corresponding to four-point function of one RR, one gauge field, one T and one T . Hence, 1 2 for p = n the string theory amplitude (20) is the S-matrix element of one RR, one gauge field and two T or two T . Its electric part is, 1 2 8iµ AAT1T1C = ±√2πpp! ǫa0···ap−1aHa0···ap−1 J (cid:20)(cid:18) (cid:19) k (t+1/4)(2ξ.k )+k (s+1/4)(2ξ.k ) ξ (s+1/4)(t+1/4) (30) 2a 3 3a 2 a × − (cid:26) (cid:27)(cid:21) Note that the amplitude satisfies the Ward identity, i.e., the amplitude vanishes under replacement ξa ka. → 1 The expansion of (s+1/4)(t+1/4)J around (22) is √2π 1 ∞ (s+1/4)(t+1/4)J = − + a (s+t+u+1/2)n n 2 (t+s+u+1/2) n=0 X ∞ d (s+t+1/2)n((t+1/4)(s+1/4))m+1 + n,m=0 n,m (31) (t+s+u+1/2) P ∞ + e (s+t+u+1/2)p(s+t+1/2)n((t+1/4)(s+1/4))m+1 p,n,m  p,n,m=0 X  where the coefficients a in the first line are exactly those appear in (13). Some of the n coefficients d and c are n,m p,n,m d = π2/3, d = 8ζ(3) (32) 0,0 1,0 − d = 7π4/45, d = π4/45, d = 32ζ(5), d = 32ζ(5)+8ζ(3)π2/3 2,0 0,1 3,0 1,1 − − 2 1 e = 2π2ln(2) 21ζ(3) , e = 4π4 504ζ(3)ln(2)+24π2ln(2)2 0,0,0 1,0,0 3 − 9 − (cid:16) (cid:17) (cid:16) (cid:17) Note that the contact terms in the last line of (31) do not have the structure of the contact terms in the first line of (31). They correspond to different couplings in field theory. The field theory, has the following massless poles for p = n: = V (C ,A)G (A)V (A,T ,T ,A(1)) (33) a p−1 ab b 1 1 A where A should be A(1) and A(2). The propagator and vertexes V (C ,A) are a p−1 iδ ab G (A) = ab (2πα′)2T (u+t+s+1/2) p 9

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.