Arxiv:Hep-Th/9708078V1 14 Aug 1997

Home , String (physics), Superstring theory, Type I string theory, Worldsheet

arXiv:hep-th/9708078v1 14 Aug 1997 fTertclPyis apc,Pln,1-2February 13-22 Poland, Karpacz, 1997 Physics, Theoretical of ∗ rneo,N,054 USA 08544, NJ, Princeton, a Lárus Thorlacius D-branes to Introduction uso fpbaesltnsltoso yeI su- II type of solutions soliton dis- p-brane a of with cussion begin II We type the theory. of We superstring context oriented the applications. in their D-branes attempt on all focus review no will to and made physics be will D-brane to troduction going still is that strong. D-branes involving explosion an activity to where of and led places before, solitons went in speculations string calculations only other explicit allow than often describe sim- considerably to are pler D-branes that along fact realization, the with This under transformations. states duality string fundamental to with attributes partners required be that the realized precisely was had it D-branes and solitons, including non-perturbative issues towards shifted du- string attention of heterotic ality, development the the With as theory. such promising string theories, less string much closed level, appears than perturbative theory the string at D- open and, associated of strings are open importance they with that the is overlooked that of was reason branes understanding One string present in the play- theory. duality in to and dynamics role theories non-perturbative key string a certain objects a ing in extended In existed ignored, being that from largely theory. gone and have string esoteric, they in time short stage relatively center to moved Introduction 1. the and D-branes to introduction basic strin a of give dynamics lectures non-perturbative the These in role important an etrspeetda h 3dKrazWne School Winter Karpacz 33rd the at presented Lectures eateto hsc,PictnUniversity, Princeton Physics, of Department -rnsaesrn hoysltn endb onaycond boundary by defined solitons theory string are D-branes hs etr oe r nedda re in- brief a as intended are notes lecture These have short, for D-branes or Dirichlet-branes, a ∗ hoyadhv on ierneo applications. of range wide a found have and theory g ntecs ftp I n I tigtheories string IIB is and theory IIA field type effective of the case the In ae otefnaetlsrn length string fundamental de- com- the large that to are that pared theory scales length field on strings effective scribes low-energy the is Supergravity II Type in Solitons 2. dy- their and D-branes explore of namics. we properties notes, lecture basic theory. the some field of effective remainder the the of In of p-branes realization the theory of string some the be to spacetime that charges in conserved show appropriate the we carry techniques, D-branes the- worldsheet string Using of motion required ory. of as equations symmetry, classical the conformal by theory is exact field strings has boundary The and open two-dimensional these them. simple on of a end theory that field strings worldsheet open of terms defined solitons in string introduce are then They We Dirichlet-branes. superstrings. field II effective type of low-energy theory the theory, pergravity n hi emoi ueprnr.Tebosonic The fields superpartners. bosonic fermionic massless theo- their of These number and a spectrum. contain string ries the the remove differ- from that between tachyon this projections choice GSO theory of consistent freedom string same two a the the to corresponds In or ence (A) su- chirality. opposite ten-dimensional (B) two the the have whether havepercharges to theories la- and refer these The A that B the and fact respectively. supersymmetries the independent IIB, to two refers and II IIA bel type of ity rdynamics. ir u trigpitfrdsusn tigsolitons string discussing for point starting Our toso pnsrnsta n nte.Te play They them. on end that strings open on itions D 0supergrav- 10 = l s = √ α ′ 1 . 2

fields are sufficient for our purposes here. The equations of motion involving h(5) are nonetheless type IIA and IIB supergravities have in common covariant and can be deduced from the D = 10 a symmetric metric tensor gµν , an anti-symmetric supersymmetry algebra [2]. rank two tensor potential bµν , and a scalar dila- These theories have classical solutions which ton φ. In string theory these fields are associ- describe extended objects called p-branes. A p- ated with worldsheet fermions with anti-periodic brane extends in p spatial directions and traces (Neveu-Schwarz) boundary conditions in both the out a p+1 dimensional worldvolume in space- left- and right-moving sectors of the closed string, time. The p-branes are topological solitons carry- and hence they are commonly referred to as NS- ing conserved charges that act as sources for the NS fields. The remaining bosonic fields in the various anti-symmetric gauge fields. supergravity theory are associated with periodic The NS-NS sector is the same in the IIA (Ramond) worldsheet fermions. In the type IIA and IIB theories, and includes a three-form field (1) theory the R-R fields are a vector potential aµ , strength H. Recall that a point charge in electro- and an anti-symmetric rank three potential a(3) , dynamics is a source for a two-form field strength. µνλ This generalizes to higher-dimensional objects; a while in the IIB theory they are a scalar a(0), and p-brane carries the charge of a (p+2)-form field anti-symmetric potentials of rank two and four, (2) (4) strength. The ‘electric’ charge that couples to H aµν and aµνλσ. is thus carried by a 1-brane. The corresponding The bosonic terms in the effective action for ‘magnetic’ field strength is a seven-form, type IIA supergravity are H˜ = e−2φ H , (3) 1 10 −2φ 2 ∗ SIIA = d x √ g e R + 4( φ) 2κ2 Z − h ∇ where H is the Hodge dual of H. The magnetic ∗ 1 1 2 1 2 charge that couples to H˜ is carried by a 5-brane. H2 h(2) + h′(4) There is a two-parameter family of black string −2 3! − 4 2 4! i 1· · solutions labeled by the electric H charge and the h(4) h(4) b , (1) −4κ2 Z ∧ ∧ mass per unit length [3]. The term ‘black’ refers where H = d b is the NS-NS antisymmetric ten- to the fact that for generic values of the mass sor field strength,∧ h(n) = d a(n−1) are R-R field the string is surrounded by an event horizon. At strengths, and h′(4) = h(4) +∧a(1) H. The NS-NS the extremal value, where the mass equals the ∧ −2φ −2 charge in appropriate units, the event horizon sector terms come with a prefactor of e gs reflecting that they arise at lowest order in∼ string shrinks away and we are left with an extended perturbation theory. The R-R sector terms, on fundamental string [4]. The supersymmetry al- the other hand, do not have such a prefactor. gebra dictates that the mass per unit length of a This can be traced to the structure of the R-R black string must always be larger than or equal vertex operators (see for example [1]) and ulti- to the charge per unit length, and the fundamen- mately to the fact that fundamental string states tal string saturates the inequality. do not carry R-R gauge charges. Similarly there is a two-parameter family of The bosonic part of the type IIB supergravity black 5-brane solutions [3], labeled by the mass action is and magnetic charge per unit five-volume. The supersymmetry algebra again implies a lower 1 S = d10x √ g e−2φ R + 4( φ)2 bound for the mass, M Q, which is satu- IIB 2 Z 2κ − h ∇ rated by the extremal NS-NS≥ 5-brane described 1 1 H2 ( a(0))2 in [5,6]. The mass bound is a generic feature of −2 3! − 2 ∇ p-brane solutions. The configurations that have ·1 (a(0)H + h(3))2 , (2) equal mass and charge per p-volume are referred −2 3! i · to as BPS states (short for Bogomolnyi-Prasad- plus some terms involving the self-dual h(5), Sommerfield). They are annihilated by some frac- which cannot be written in covariant form. The tion of the supercharges of the theory and as a 3 result they sit in short multiplets of the super- theory is very simple and the extended objects symmetry algebra [7]. If the theory has enough so defined indeed carry R-R charge and have the extended supersymmetry the mass and number of right mass density required by string duality. short multiplets in the spectrum is not changed by quantum corrections, even at strong coupling. 3. Dirichlet-Branes BPS states have therefore played a key role in the study of strong/weak coupling duality in gauge In order to describe R-R solitons in type II theory and string theory in recent years. string theory we introduce open strings into the The p-branes of most interest to us in these theory. We then have to prescribe boundary conditions for the worldsheet fields at the worldsheet lectures are associated with the R-R sector fields. µ The IIA theory has 0-branes that couple to the boundary. The bosonic worldsheet fields X , for two-form field strength h(2) and 2-branes that µ = 0,..., 9, are interpreted as the embedding couple to h(4). Then there are 4- and 6-branes coordinates of the string in spacetime. Let them that couple to the magnetic fields h(4) and satisfy ∗ h(2), respectively. In addition one can con- (N) na∂ Xm =0 m =0,...,p, ∗ a sider 8-branes. The corresponding ten-form field (D) Xi = const i = p +1,..., 9, (4) strength is not dynamical, but a constant field corresponds to a physical energy density [8,9]. where na is normal to the boundary. The N and D In type IIB theory there are p-branes with refer to Neumann and Dirichlet boundary condi- p=-1,1,3,5,7. The p=-1 case describes instan- tions. Conventional open strings in type I string tons, which are events in Euclidean spacetime, theory have Neumann boundary conditions for all and couple to the field strength of the R-R scalar the Xµ fields. With the above prescription the a(0). The R-R 1- and 5-branes couple to the open strings are constrained to end on a p+1 di- electric and magnetic field strengths h(3) and mensional hypersurface in spacetime but the ends h˜(7) = h(3), respectively. The 3-brane acts as a are free to move on this surface, see Figure 1. source for∗ the self-dual field h(5) and the 7-brane couples to the magnetic field h˜(9) = (da(0)). There are no 9-branes in type II supergravity.∗ The corresponding field strength would have to be of rank eleven and must therefore vanish in ten dimensions. In type I string theory, however, X P 9-branes play an important role in giving rise to 0 the SO(32) Chan-Paton gauge degrees of freedom X [1]. P+1 9 The black p-branes of type IIA and IIB su- X ,..., X pergravity are approximate classical solutions of FIGURE 1: Dirichlet p-brane with open string theory valid for sufficiently large mass per string attached. volume. Exact classical solutions of string theory are described by two-dimensional superconformal This defines the p+1 dimensional worldvolume field theories. The NS-NS sector 1- and 5-branes of an extended object, called a Dirichlet-brane. can be systematically studied using two dimen- At this stage the D-brane appears to be a rigid sional sigma models with the appropriate field sheet infinitely extended in p spatial directions. content and couplings [4,10], but this approach is Later on we shall see that a D-brane is in fact dy- not well suited for R-R sector backgrounds. For- namical object with certain massless open string tunately there is another way to proceed. It in- modes describing its undulations. A D-brane can volves a somewhat indirect construction in terms be given a finite volume in a spacetime with com- of auxiliary open strings but, as we shall see be- pact dimensions by wrapping around non-trivial low, the resulting two-dimensional conformal field cycles. Finite volume D-branes that can contract 4 to a point are presumably unstable and are not transverse to the p-brane. In string theory one described by simple classical solutions. could compute the one-point amplitude of a gravi- We also need to prescribe boundary conditions ton vertex operator inserted into a worldsheet of for the worldsheet fermions. The surface terms in disk topology with the D-brane boundary condi- the fermion equations of motion cancel if we let tions (4) and (5) applied. This amplitude directly the left-and right-moving fermions, ψµ and ψ˜µ, probes the gravitational field of the D-brane and reflect into each other up to a sign, at the world- allows the tension to be read off [12]. The disk sheet boundary. Since the fermion fields are only is an open string tree diagram and the amplitude defined up to an overall phase we can set ψµ = ψ˜µ therefore depends on the dilaton as e−φ. It fol- at one boundary, for example at σ = 0. The sign lows that the D-brane tension is proportional to −1 choice at the other end must be correlated with gs . Similarly, a one-point disk amplitude in- the N and D boundary conditions on the bosonic volving a vertex operator for the appropriate R- fields (4) in order to preserve worldsheet super- R sector gauge field yields the magnitude of any symmetry, R-R charge carried by the D-brane. These calculations are straightforward but re- (N) ψm = ψ˜m m =0,...,p, ± quire careful normalization, without which one (D) ψi = ψ˜i i = p +1,..., 9. (5) cannot read off the correct mass and charge. It ∓ turns out to be easier to instead work with a In flat spacetime the type II worldsheet field one-loop string amplitude which involves the ex- theory contains only free bosons and fermions and change of virtual closed strings between D-branes. is manifestly a conformal field theory. The lin- Following Polchinski we consider a pair of parallel ear boundary conditions (4) and (5) preserve the Dirichlet p-branes, separated by a distance Y in conformal symmetry and thus define exact classi- a transverse direction, and compute the annulus cal solutions of string theory. The construction is graph in Figure 2. indirect but it is very much simpler than the non- linear sigma model description of NS-NS solitons [11]. Dirichlet-branes are BPS states. This can be seen as a simple consequence of the boundary conditions on the worldsheet fields. The two supercharges of the type II theory are defined in terms of left- and right-moving worldsheet currents which get reflected into each other at the boundary. Only the linear combination Qα + cpQ˜α of the two supercharges is conserved, 9 9 were cp is a phase factor that comes from a par- X = 0 X = Y ity transformation in the transverse directions Xi FIGURE 2: Annulus amplitude for open strings with i = p +1,..., 9. In other words, a D-brane stretched between different branes. is invariant under half the supersymmetry of the original type II theory, which is precisely the BPS Worldsheet duality allows us to either view this condition. diagram as an open string vacuum loop with the string endpoints living on the two separate D- 4. D-Brane Tension and Charge branes or as a tree level closed string exchange The tension in a p-brane is by definition the of closed strings between the D-branes, see Fig- energy per unit p-dimensional volume. In super- ure 3. The diagram has a single real valued mod- gravity theory the tension can be determined by ulus which, in the closed string channel, can be an ADM procedure, considering the asymptotic taken to be the worldsheet time τ between the ends of the cylinder and runs from 0 to . In the falloff of the gravitational field in the directions ∞ 5

∞ limit of large τ the amplitude is dominated by F (e−πt) = eπt/3 (1 e−(2n−1)πt)8, (8) the exchange of massless closed sting states, the 3 − − nY=1 graviton and dilaton from the NS-NS sector and R-R gauge particles. By identifying the separate and they in fact cancel by Jacobi’s famous ab- contributions from these states one can read off struse identity. This is not surprising given that the tension and R-R charge of the D-branes. we are calculating a vacuum amplitude in a supersymmetric theory. The vanishing of the amplitude can also be given an interpretation in terms of closed string exchange. The exchange τ t =1/t of a graviton or dilaton gives rise to an attractive τ force between the D-branes, while the exchange t of a R-R gauge boson gives rise to the repulsive force between two D-branes with the same sign R- R charge. When all the contributions are added the net force turns out to be zero. This force bal- ance, which is a familiar feature of BPS objects, FIGURE 3: Open and closed string channels only holds when both D-branes are at rest in some for annulus amplitude. reference frame. Later on we will discuss the in- teraction between D-branes when one is moving To evaluate the diagram we first view it as an with respect to the other. open string loop and write the vacuum amplitude Even if the overall amplitude vanishes we can still extract useful information from the separate dp+1k ∞ dt terms. The amplitude (7) was expressed as an A = Vp+1 p+1 Z (2π) Z0 2t integral over t, the modulus of the annulus in the −2πt(k2+M 2 ) e {n} . (6) open string channel. The t 0 end of the inte- × → X{n} gration range is an ultraviolet limit from the open string point of view, but in the closed string chan- Vp+1 denotes the p-brane worldvolume, the inte- nel the modulus is τ =1/t and the corresponding gral is over the zero mode momentum in the p+1 limit is in the infrared. This is precisely where the directions along the worldvolume, and the sum is cylinder amplitude is dominated by massless ex- over the spectrum of physical modes of an open change and we expect to find information about superstring with ends on separate D-branes. The the tension and R-R charge of the D-brane. In or- mode sum and the momentum integral can be der to extract the t 0 behavior of the amplitude carried out explicitly, giving (7) we make use of the→ properties of the functions Fi defined in (8) under the modular transforma- dt 2 − (p+1) −Y 2t/2π A = V (8π t) 2 e tion t 1/t which relates the open and closed p+1 Z 2t string channels,→ F0 [F1 + F2 + F3]. (7) × −πt −1/2 −π/t F0(e ) = t F0(e ), The three terms in the brackets are −πt −π/t F1(e ) = F3(e ), ∞ F (e−πt) = F (e−π/t). (9) F (e−πt) = e2πt/3 (1 e−2nπt)−8, 2 2 0 − nY=1 Inserting this into (7) and expanding out the in- ∞ finite products to leading orders one finds F (e−πt) = 16e−2πt/3 (1 + e−2nπt)8, 1 − Y (p+1) n=1 dt 2 − A (1 1)V (8π t) 2 ∞ ≈ − p+1 Z t F (e−πt) = eπt/3 (1 + e−(2n−1)πt)8, 2 4 2 2 Y 16t e−Y t/ π. (10) n=1 × 6

Carrying out the t integral one arrives at [1,9] where n is some integer. The R-R charges obtained in (13) satisfy this relation with the mini- A (1 1)V 2π(4π2)3−pG (Y 2), (11) ≈ − p+1 q−p mum value of n = 1. 2 ± where Gd(Y ) is the radial Green function of a We conclude this section with some remarks massless scalar in d spatial dimensions. on the relation between D-branes and the cor- We now want to compare this result with a corresponding R-R charged solitons of supergravity responding field theory calculation. Consider a theory. They have the same tension and charge theory of a rank p + 2 field strength coupled to a density but apart from that they appear quite p-brane source, different. The D-branes are defined in terms of weakly coupled open strings that propagate in flat 1 2 S = d10x(h(p+2)) spacetime, whereas a supergravity p-brane solu- Z 2(p + 2)! tion describes a spacetime with non-trivial met- +iµ a(p+1). (12) ric and curvature. It depends on the value of p Z Vp+1 the string coupling which description is more appropriate. As g 0, the gravitational field of The leading order exchange amplitude is easily s → computed and has the same value as the R-R term the p-brane vanishes and the flat space D-brane in (11) provided we make the identification [9] picture is valid. I on the other hand, the coupling is not small, we are not justified in ignoring 2 2 3−p µp =2π(4π ) . (13) the gravitational response to the p-brane source and the supergravity solutions is a more accurate The analogous calculation for graviton and dila- description of the geometry. Supersymmetry and ton exchange in the effective field theory gives, the BPS property of extremal p-branes are crucial upon comparing with (11), the following D-brane ingredients in any argument that takes a result tension [9] obtained at weak coupling, such as the number √π(2π)3−p of multi D-brane states with some macroscopic τp = . (14) gs charges, and extrapolates it to the strong coupling regime in order to draw conclusions about Dirichlet-branes are thus seen to be BPS ob- a black hole of the same mass and charge. jects carrying R-R gauge charge and have tension inversely proportional to the string coupling. Their definition in terms of boundary con- 5. Probing D-branes With Strings ditions on auxiliary open strings leads to a two- The conformal field theory description of D- dimensional worldsheet theory with exact confor- branes is rather simple and allows explicit calcula- mal symmetry as required for classical solutions tion of various processes involving D-branes. For in string theory. This combination of properties instance one can compute amplitudes for closed makes a very convincing case for the identifica- string scattering off a D-brane [14–16]. This pro- tion of D-branes as the R-R sector p-branes of vides information about the structure of D-branes string theory. beyond the classical picture of infinitely thin hy- There is an important check on the above con- persurfaces. The leading contribution to the p- siderations. The p + 2 form field strength that brane form factor can be read off from the two- couples to a p-brane is Hodge dual to a 8-p form point function of closed-string vertex operators on field strength that couples to a (6-p)-brane. This a disk with the appropriate mixed Dirichlet and is analogous to the duality between electric and Neumann boundary conditions. This corresponds magnetic point charges in four-dimensional space- to a string colliding with a D-brane creating an time and accordingly the brane charge densities excitation that propagates along the worldvolume must satisfy a quantization condition analogous until the D-brane returns to its ground state by to the Dirac condition [13], emitting another string. For simplicity, we only µpµ6−p =2πn, (15) consider the scattering of NS-NS sector string 7 states. The generalization to R-R sector states string emitted from the D-brane interacting with is straightforward [15]. The amplitude takes the the external strings. The amplitude exhibits the form Regge behavior characteristic of string scattering,

t −1 1 1 (log s)t A = V1(z1, z¯1)V2(z2, z¯2) , (16) A s 2 e 2 , (21) h iDisk ∼ ≈ s where the vertex operator of a NS-NS sector indicating that the transverse thickness of a D- massless state (graviton or anti-symmetric ten- brane, as seen by strings, is of order the string sor) with polarization εIJ is scale at low energy and grows with increasing en-

I I ergy of the probe. This conclusion holds for all V (z, z¯) = εIJ (∂X + ikψψ ) Dirichlet p-branes with p 0. For scattering off J J ikX(z,z¯) (∂X¯ + ikψ˜ψ˜ )e . (17) a D-instanton the kinematics≥ is degenerate and × the form factor exhibits point like behavior. For evaluating the amplitude it is convenient to It is not surprising to find that stringy probes map the disk to the upper half of the complex cannot see structure on scales smaller than the plane and perform Wick contractions using the fundamental string length, but since D-branes in- following propagators for the worldsheet fields, teract by string exchange we can also draw some Xµ(z, z¯)Xν (w, w¯) = 2ηµν log z w conclusions about high energy collisions of D- h i − | − | branes. If the intermediate string states carry 2aµν log 1 zw¯ , − | − | high energy the Regge behavior will take over, so ηµν ψµ(z)ψν (w) = , in order to probe short distances by scattering D- h i z w branes of each other the collisions should be soft ia−µν ψµ(z)ψ˜ν (¯w) = , (18) and involve relatively slow moving D-branes. It h i 1 zw¯ − turns out that sub-stringy distances can indeed where aµν = diag 1, 1,..., 1, 1,..., 1 with be probed by slow D-branes, but we need to in- the -1 in the last 9-p{− entries reflecting− the− Dirich-} troduce a few more tools before we address that let boundary conditions. The resulting integrals question. can be carried out explicitly with the result One can also compute mixed closed and open string scattering amplitudes in a D-brane back- Γ(1 2s)Γ( t/2) ground using more or less standard techniques. A = f(εi, ki) − − , (19) Γ(1 2s t/2) The open string states correspond to excitations − − of the D-brane itself and the closed strings repre- where the functional form of the kinematic pref- sent a coupling to the rest of the world. Such am- actor f depends on the type of string states that plitudes play an important role in the application scatter off the D-brane [14]. The kinematic in- of D-branes to black hole physics, for instance in variants are the calculation of Hawking emission from a near- 2 2 s = kk = kk , extremal black hole modeled as a multi D-brane − 1 − 2 composite object. Such calculations are discussed k2 t = 2k 2k⊥ k⊥, (20) by Maldacena elsewhere in this volume. 1 − 1 · 2 k ⊥ where k and k are momenta parallel and per- 6. D-brane Dynamics pendicular to the D-brane worldvolume. The amplitude (19) is similar to standard string So far we have only discussed single D-branes scattering amplitudes. This is hardly surprising at rest. It is clearly important to develop a dy- since the D-brane by definition interacts with its namical theory for multiple D-brane configura- environment through the open strings that attach tions and D-brane motion. In full generality such to it. The poles in the s-channel correspond to a dynamical description remains a technical chal- an open string propagating along the D-brane, lenge. The formalism we have at our disposal de- while the t-channel poles correspond to a closed scribes D-branes in terms of a fixed background 8 conformal field theory and does not generalize in in moduli space where the worldsheet is a long any easy way to time-dependent backgrounds. In cylinder, i.e. the closed string channel, whereas order to proceed further we have to make approx- the low velocity amplitude is dominated by the imations. Following Bachas [17], we consider the open string channel where the annulus is long and forward scattering of two parallel p-branes in the thin. In the v 1 case, the light degrees of eikonal approximation, where one p-brane moves freedom are associated≪ with short open strings in a straight line past the other with impact pa- stretched between the D-branes. This suggests rameter b. The p-branes interact by string ex- an effective field theory description in terms of change but no back-reaction on their trajectories massless open string states, i.e. a gauge theory, or internal states is taken into account. The cal- which turns out to be very useful, as we shall see culation involves an amplitude which is a straight- later on. forward generalization of the one-loop open string The one-loop string amplitude that we have diagram we used previously to identify the D- found so useful can be generalized further [18,19]. brane tension and R-R charge, with the only dif- For instance, the two D-branes do not have to ference being that the separation between the p- have the same dimensionality. Consider the for- branes is no longer constant, ward scattering of a q-brane past a p-brane with p q. We can only have p-q be an even number 2 2 2 2 Y = b + v t . (22) since≥ we are either working in IIA theory where p and q are both even or IIB theory where they are The one-loop amplitude can still be evaluated ex- both odd. The new feature here is open strings plicitly and gives the phase shift for the forward that stretch between D-branes of different dimen- scattering. The resulting expression is somewhat sionality, with some fields that have ND bound- complicated and will not be repeated here. There ary conditions, i.e. Neumann at the p-brane but are two interesting limits. One is large s, where s Dirichlet at the q-brane. Once again the ampli- is the center-of-mass energy squared of the scat- tude can be obtained explicitly but is somewhat tering process. The imaginary part of the phase messy. It is useful to introduce an effective po- shift is then [17] tential as follows b2 ∞ 2 1 0 Im(δ) exp[ 2 ], (23) A(v,b )= dX V (v, r), (25) ∼ −log(s/M ) 2 √2π Z−∞ where M is the mass of the D-brane. The phase where r2 = b2 + (X0)2v2. This potential governs shift exhibits a characteristic length scale, b the motion of a non-relativistic q-brane interact- 2 1/2 ∼ [log(s/M )] , which grows with energy consis- ing with a p-brane with p q. The low velocity tent with the Regge behavior of the string scat- behavior depends on the value≥ of p-q. tering amplitudes considered above. For p=q we recover the case we discussed above The other limit is low velocity, v 1, where and the potential takes the form one finds [17] ≪ av4 v2 2 p = q : V (v, r)= 1+ O , (26) b 7−p 4 Im(δ) exp[ ]. (24) −r r ∼ − v where a is a positive constant. The velocity de- In this case the characteristic length scale is pendence means that identical D-branes at rest b v1/2 in string units, which is very small if feel no force, consistent with the BPS property. v ∼ 1. This is the first indication that slow- At non-zero velocity the system no longer has moving≪ D-branes can probe distances below the unbroken supersymmetry and there is a net at- string length. tractive force between moving D-branes. The The two limits involve different values of the expansion parameter at low velocities is v2/r4 real valued modulus of the annulus amplitude. which again suggests a characteristic length scale The large s behavior is dominated by the region r v1/2. 0 ∼ 9

For p = q + 4 there is still no force between the usual string theory procedure. The first step is to D-branes when they are at rest, but the veloc- identify the appropriate two-dimensional world- ity dependent potential is different from the p=q sheet theory with general couplings. Since D- case, branes are associated with open strings it is the possible boundary couplings that are of interest av˜ 2 v2 here, p = q + 4 : V (r, v)= 3−q 1+ O 4 , (27) −r r The leading term goes as v2, so it is natural to p absorb it into the kinetic term S = ds A (X0,...,Xp)∂ Xm b Z m t mX=0 1 ag˜ 2 q K 1 3−q v . (28) ∼ g − r 0 p i + νi(X ,...,X )∂nX , (30) The low velocity scattering can then be described i=Xp+1 in terms of geodesic motion on a moduli space with a non-trivial metric [19]. where ∂t and ∂n denote the tangential and normal Another case of interest is q =p ¯, the scattering derivatives at the worldsheet boundary. The pres- of a Dirichlet p-brane and anti-p-brane. These ence of the boundary action (30) leads to a gener- objects do not satisfy a BBS condition even at alization of the mixed Neumann/Dirichlet bound- rest and one expects a nontrivial potential already ary conditions for the open strings attached to the at v=0. There is in fact an instability if the p- D-brane. The above terms are the most general brane and antibrane get too close to each other boundary action of scaling dimension one (i.e. [20]. To see this, consider the one-loop amplitude renormalizable) for scalar fields Xµ, µ =0,..., 9. (7) for identical p-branes, with a small but criti- The function Am is interpreted as a U(1) gauge cal difference. In the present case the objects have potential that lives on the worldvolume and νi opposite sign R-R charge and the corresponding represents a shift in the D-brane position in the term, F3 in the square brackets in (7) enters with transverse directions. We have not written the the opposite sign. Now there is no longer a cancel- accompanying fermion terms required for world- lation but instead the terms add and the leading sheet supersymmetry. The form of the boundary behavior of the integrand for large t is action (30) is constrained by the Dirichlet bound- ∞ p+1 ary condition in the free theory in that A and ν dt 2 − m i A = Vp+1 (8π t) 2 are only functions of the worldvolume embedding Z0 t 2 coordinates. (π− Y )t −2πt e 2π 1+ O(e ) . (29) The next step towards the effective action is to × The integral diverges for Y 2 < 2π2. The system couple these boundary terms to the general non- is unstable against p p¯ annihilation at short dis- linear sigma model in the bulk of the string world- tance. The actual annihilation− process is presum- sheet and require the combined system to have ably complicated and results in the emission of a conformal symmetry in order to decouple nega- large number of closed strings. tive norm states from the string spectrum. This There is a similar instability in the amplitude places a set of restrictions on the coupling func- with p = q + 2. In this case the system contains tions of the theory which can be cast as dynam- two D-branes of different dimensionality which do ical field equations in the ten-dimensional target not annihilate each other but form a bound state space. The final step is to identify a target space with non-vanishing binding energy. action functional whose equations of motion are equivalent to the field equations that follow from 7. Effective Action for D-Branes two-dimensional conformal invariance. This procedure was carried out for D-branes in bosonic We can generate an effective action to describe string theory by Leigh [21] and the resulting D- low-energy processes involving D-branes by the brane effective action is a generalization of the 10

Born-Infeld action of non-linear electrodynamics, The supersymmetric generalization of the abelian Born-Infeld action (31) has been written down S = T dp+1σe−φ [24]. It is rather involved and will not be needed p pZ − here.

det(gµν + bµν +2πFµν ), (31) The U(1) worldvolume gauge field is associated ×q with massless open strings which end on the D- where the coefficient Tp is proportional to the D- brane. An interesting new feature appears when brane tension, gµν ,bµν and φ are spacetime fields we consider multiple D-brane configurations. If pulled back to the p-brane worldvolume, and the some of the D-branes coincide then there are ad- components of the gauge field strength are ditional massless open strings that go between different D-branes. When the D-branes are well F = ∂ A ∂ A , m,n =0,...,p mn m n − n m separated such strings have a large mass due to Fij = 0, i,j = p +1,..., 9, their tension and are suppressed at low energies. Fmi = ∂mνi. (32) When N D-branes are on top of each other there are N 2 massless open string modes that fill out The Fmn components are the worldvolume gauge the adjoint representation of U(N) and the world- field strength while the Fmi components reflect volume gauge symmetry is enhanced from U(1) to the D-brane velocity and tilt in the transverse U(N) [25,26]. directions. The full effective action also includes The D-brane position in the transverse direc- the usual ten-dimensional action for gµν , bµν , and tions, νi, is also promoted to a U(N) matrix. This φ. has played a crucial role in exploring D-brane dy- In type II superstring theory the NS-NS sector namics. It indicates that spacetime coordinates contributes a Born-Infeld term identical to (31) are replaced by non-commuting variables when and in addition the R-R sector gives rise to Chern- we get down to very short distances. Simons type terms [22,23] that can be neatly sum- This formalism also leads to a nice geomet- marized as ric picture for symmetry breaking when a gauge

′ (b+2πF ) theory can be realized as the worldvolume the- Sp = iµp a Stre , (33) ory of a collection of D-branes. The unbroken ZVp ∧ +1 phase corresponds to all the D-branes coinciding (n) where a = n a is the sum of all the R-R and then the gauge symmetry can be partially or gauge potentialsP in type II theory. The exponen- completely broken by moving some or all of the tial in (33) also involves a sum of terms but the D-branes apart. integral picks out forms of overall rank p+1 from The non-abelian generalization of the D-brane the product. The couplings between the world- effective action remains a subject of investiga- volume gauge field and the R-R sector potentials tion at the present time. In particular the gen- play an important role in the study of D-brane eralization of the non-linear Born-Infeld action bound states [1,22]. Consider for example a p- is troublesome [27]. In flat spacetime when the brane with a non-vanishing gauge field strength gauge field strength is weak and the D-branes are Fµν in its worldvolume. According to (33) such all moving slowly the non-linear terms in the D- a configuration carries (p-2)-brane charge in ad- brane action will be small and the effective ac- dition to its p-brane charge. It turns out to have tion is well approximated by the appropriate di- lower energy than the p- and (p-2)-branes sepa- mensional reduction of D=10, supersymmetric, rately and is in fact a realization of the (p,p-2) U(N) Yang-Mills theory [26]. Most studies of bound state mentioned at the end of the previous D-brane dynamics to date have been based on section. the Yang-Mills approximation rather than the full We have only kept track of the bosonic part of non-linear theory and we will describe an example the D-brane action but there are of course also in the following section. fermionic terms as required by supersymmetry. The dimensionally reduced Yang-Mills theory 11 will in particular include a potential term for the At any rate, this length scale is short compared to matrix valued D-brane positions which is propor- the fundamental string length ls when the string i j tional to T r([νi, νj ][ν , ν ]) and comes from the coupling gs is weak, and this is an indication that Fij components of the non-abelian field strength. slow moving D-branes can probe distances below The D-branes can only be separated along flat the string length. We will not present a detailed directions of this potential but these occur pre- study of the D-particle system here, but we note cisely when the νi all commute and the gauge that the dimensionally reduced Yang-Mills the- group is broken from U(N) to U(1)N , as expected ory gives rise to a velocity dependent potential from considering open strings attached to the D- V av4/r7 in agreement with our earlier consid- branes. erations≈ based on the one-loop open string amplitude. String duality requires there to be a bound 8. D-Particle Scattering state at threshold in this system. The existence of such a bound state has recently been proven As a simple application of the above ideas [30]. This required some fairly heavy mathemat- we briefly consider the scattering of two non- ical machinery but relatively simple arguments relativistic 0-branes [19,28,29]. This involves the can be given [19,28] that in such a bound state dimensional reduction of D=10 supersymmetric 1/3 the D-particles would be separated by r gs U(2) gauge theory to a 0+1 dimensional theory 2/3 ∼ on the D-particle worldline. The U(2) symme- and have velocities v gs , which is consistent with the relation r v∼1/2 obtained from the one- try factorizes into U(1) SU(2) in a natural way, ∼ where the U(1) factor× represents the center-of- loop string amplitude. mass degrees of freedom and the interesting dy- 9. T-Duality and D-Branes namics is in the SU(2) sector. The worldline action including the fermionic Our final topic is T-duality in type II string the- terms is ory and the transformation properties of D-brane configurations under this symmetry. D-branes 1 µν ¯ µ S = dt T r(Fµν F ) iT r(ψΓ Dµψ) , (34) were in fact originally discovered when investi- Z gs − gating how T-duality acts on open strings [12]. where A0, νi, and ψ take values in the adjoint Consider the closed string mode expansion representation of U(2) and Xµ(z, z¯)= xµ(z)+˜xµ(¯z), where z = eτ−iσ and F = ∂ ν + [A , ν ], 1 0i 0 i 0 i xµ(z)= xµ + i( αµlogz + αµ z−m) 0 − 0 m m Fij = [νi, νj ], mX6=0 D0ψ = ∂0ψ + [A0, ψ], 1 x˜µ(¯z)=˜xµ + i( α˜µlogz¯ + α˜µ z¯−m) (37) 0 − 0 m m Diψ = [νi, ψ]. (35) mX6=0 A simple scaling argument reveals the character- The spacetime momentum is the sum of the left- istic length scale of the dynamics [29]. In terms and right-moving zero modes, of the rescaled variables t˜= g1/3t, A˜ = g−1/3A , 0 0 µ µ µ −1/3 α +α ˜ =2p , (38) andν ˜i = g νi, the action takes the form 0 0 and their difference gives the change in Xµ(z, z¯) S = dt˜ T r(F˜ F˜µν ) iT r(ψ¯ΓµD˜ ψ) , (36) Z µν µ upon going around the closed string, − where the string coupling has been scaled out. µ µ 1 µ µ α0 α˜0 = X (σ +2π, τ) X (σ, τ) . (39) The characteristic length scale is therefore l − 2π − 1/3 ∼ µ g ls, which is in fact the Planck length of the For non-compact spacetime directions X must µ µ µ eleven-dimensional supergravity theory which is be single valued and it follows that α0 =α ˜0 = p . the strong coupling dual of type IIA string theory. If, on the other hand, one of the coordinates, say 12

X9, parametrizes a circle of radius R, then X9 In the open string sector T-duality acts on the only has to come back to itself up to an integer boundary conditions in non-trivial way. To see multiple of 2πR when σ σ +2π, this it is convenient to view the worldsheet bound- → ary in a closed string channel, as shown in Fig- α9 α˜9 = mR. (40) 0 − 0 ure 4. This can always be arranged by a con- The momentum in the compact direction takes formal transformation and choice of parametriza- 9 µ discrete values p = n/R so that tion. The Neumann boundary condition, ∂τ X = 2n 0, written in terms of complex coordinates z = α9 +α ˜9 = . (41) eτ−iσ, is 0 0 R µ The left- and right- moving zero modes in the 0 = (z∂z +¯z∂z¯)X (z, z¯) µ compact directions are then given by = z∂zx (z)+¯z∂z¯x˜(¯z). (46) n mR µ 9 Similarly, the Dirichlet condition, ∂σX = 0, be- α0 = ( + ), R 2 comes 9 n mR α˜0 = ( ). (42) 0= z∂ xµ(z) z∂¯ x˜µ(¯z). (47) R − 2 z − z¯ The mass spectrum and interactions of the When T-duality acts on a compact coordinate string theory is invariant under the T-duality the right-moving part of that coordinate changes transformation sign and thus Neumann and Drichlet boundary 2 conditions are exchanged. Now act with a dual- T9 : R ,m n. (43) → R ←→ ity transformation Tµ on a Dirichlet p-brane with worldvolume along (X0,...,Xp). If µ p the di- In other words, the theories at radius R and ≤ 2/R are exactly the same when the winding and mension of the D-brane is reduced by one, while momentum quantum numbers are interchanged. it is increased by one if µ>p. This is consistent The T-duality transformation can be written with the observation that T duality on an odd number of coordinates changes IIA theory into 9 ′9 9 9 T9 : X (z, z¯) X (z, z¯)= x (z) x˜ (¯z), (44) IIB and vice versa. → − T-duality can often be used to relate problems i.e. as a parity transformation acting on right that might at first sight seem quite different. For movers only. The corresponding transformation example, bound states of Dirichlet p- and (p-n)- on the spin fields is branes for a given value of n with different values T : Sα Sα, S˜α S˜α˙ . (45) of p can all be mapped into each other by T- 9 ←→ ←→ duality transformations. When T-duality is com- The chirality of the right-moving spin field is bined with the strong/weak coupling SL(2,Z) changed under T which means that the duality 9 duality of IIB theory the class of related prob- transformation relates IIA and IIB theory. This lems is further enlarged. remains true when T-duality acts on an odd number of compact dimensions while acting with T- 10. Parting Words duality on an even number of dimensions does not change the chiral character of the theory. D-branes are interesting and useful objects. They have enhanced our understanding of non- perturbative string theory and continue to find new applications. There is already a vast lit- τ erature on D-brane physics and it seems likely σ they will continue to play an important role in the future development of string theory. In these FIGURE 4: Worldsheet boundary in lectures we described how D-branes are intro- closed-string channel. duced into type II superstring theory and dis- 13 cussed some of their dynamical properties. We 15. S. Gubser, A. Hashimoto, I. Klebanov, and J. only scratched the surface of the subject and Maldacena, Nucl. Phys. 472 (1996) 231. many important topics were left out. For exam- 16. M. Garousi and R. Myers, Nucl. Phys. 475 ple, we did not cover unoriented string theory at (1996) 193. all, where a lot of interesting work involving D- 17. C. Bachas, Phys. Lett. B374 (1996) 37. branes has been done. We nevertheless hope that 18. G. Lifschytz, Phys. Lett. 388 (1996) 720. the reader comes away with some impression of 19. M.R. Douglas, D. Kabat, P. Pouliot and S.H. the power and beauty of D-brane technology. Shenker, Nucl. Phys. 485 (1997) 85. The author thanks the organizers of the 33rd 20. T. Banks and L. Susskind, Brane – Anti- Karpacz Winter School for the opportunity to lec- Brane Forces, hep-th/9511194. ture and for their kind hospitality. He also thanks 21. R. G. Leigh, Mod. Phys. Lett. A4 (1989) A. Peet for helpful comments on the manuscript. 2767. This work was supported in part by a US Depart- 22. M. Li, Nucl. Phys. B460 (1996) 351; M. ment of Energy Outstanding Junior Investigator R. Douglas, Branes within Branes, hep-th/- Award, DE-FG02-91ER40671. 9512077. 23. E. Bergshoeff, M. de Roo, M. B. Green, G. Papadopoulos, and P. K. Townsend, Nucl. B470 REFERENCES Phys. (1996) 113; E. Alvarez, J.L.F. Barbon, and J. Borlaf, Nucl. Phys. B479 1. J. Polchinski, TASI Lectures on D-Branes, (1996) 218; E. Bergshoeff and M. de Roo, hep-th/9611050 Phys. Lett. B380 (1996) 265; M.B. Green, 2. J.H. Schwarz, Nucl. Phys. B226 (1983) 269. C.M. Hull, and P.K. Townsend, Phys. Lett. 3. G. Horowitz and A. Strominger, Nucl. Phys. B382 (1996) 65. B360 (1991) 197. 24. M. Cederwall, A. von Gussich, B. Nilsson, P. 4. A. Dabholkar and J. Harvey, Phys. Rev. Lett. Sundell, and A. Westerberg, Nucl. Phys. 490 63 (1989) 719; A. Dabholkar, G. Gibbons, J. (1997) 179. Harvey and F.R. Ruiz, Nucl. Phys. 340 (1990) 25. J. Polchinski, Phys. Rev. D50 (1994) 6041. 33. 26. E. Witten, Nucl. Phys. 460 (1996) 335. 5. M.J. Duff and J.X. Lu, Nucl. Phys. B354 27. For an interesting recent proposal see A. (1991) 141. Tseytlin, On Nonabelian Generalization of 6. C. Callan, J. Harvey, and A. Strominger, Born-Infeld Action in String Theory, hep- Nucl. Phys. 367 (1991) 60. th/9701125. 7. D. Olive and E. Witten, Phys. Lett. 78B 28. U.H. Danielsson, G. Ferretti, and B. Sund- (1978) 97. borg, Int. J. Mod. Phys. A11 (1996) 5463. 8. L.J. Romans, Phys. Lett. B169 (1986) 374. 29. D. Kabat and P. Pouliot, Phys. Rev. Lett. 77 9. J. Polchinski, Phys. Rev. Lett. 75 (1995) 4724. (1996) 1004. 10. C. Callan, J. Harvey, and A. Strominger, 30. S. Sethi and M. Stern, D-Brane Bound State Nucl. Phys. 359 (1991) 611. Redux, hep-th/9705046. 11. See for example: C. Callan, J. Harvey, and A. Strominger, Supersymmetric String Solitons, hep-th/9112030. 12. J. Dai, R. Leigh and J. Polchinski, Mod. Phys. Lett. A 4 (1989) 2073. 13. R.I. Nepomechie, Phys. Rev. D31 (1985) 1921; C. Teitelboim, Phys. Lett. B167 (1986) 63, 69. 14. I. Klebanov and L. Thorlacius, Phys. Lett. B371 (1996) 51.