Arxiv:1512.02657V1 [Hep-Th] 8 Dec 2015 Rbe.I Is It Problem

Home , Andrew Strominger

arXiv:1512.02657v1 [hep-th] 8 Dec 2015 rbe.I is It problem. a hnb nesoda ttmn of statement t a generate as identities understood Ward be their then and may group BMS the of lations smttcKligsiosta smtt oagedepende angle to I asymptote that spinors Killing asymptotic recently of impact physical the standing The Supergravity 2 conclusions. our supports which topic, this on [ of gravitinos soft h oainrqie oeepaain ie metric a Given explanation. some requires notation The aebe niiae n[ in anticipated been have hnst h eakbesre fppr [ papers of series remarkable the to Thanks Introduction 1 hwta hr r eiulguesmere at symmetries gauge residual are there that transformati show supersymmetry local with associated density eto rvt,btas nYn–il hoy awl theo Maxwell theory, Yang–Mills in also but gravity, of ment n eaeteWr dniyt h otlmto gravitinos. of limit soft the to identity Ward the relate and R)fil ntesmls setting, simplest the in field (RS) eew rvd nudrtnigo eiullclsupersym local residual of understanding an provide we Here efis nrdc h etn of setting the introduce first We hl rprn hsatce ebcm wr fteindepen the of aware became we article, this preparing While r otie nlclsprymtytasomtos The transformations. supersymmetry local in contained are d = eiulLclSprymtyadteSf Gravitino Soft the and Supersymmetry Local Residual N 4, = N M ru,poiigasprymtiaino h M trans BMS the of supersymmetrization wit a associated providing charges group, into BMS commute charges ang these every that at show supersymmetry of statement the to equivalen and are gravitino symmetries these with associated identities d QD[ SQED 1 = S = [email protected] eso htteeeit nifiietwro emoi symme fermionic of tower infinite an exists there that show We = uegaiyato nte15odrfraimpoe most proves formalism order 1.5 the in action supergravity 1 4, 26 S N 2 ihteaypoi iln pnr on n[ in found spinors Killing asymptotic the with ] + = 25 S 3 uegaiyo naypoial a akrud h War The background. flat asymptotically an on supergravity 1 / ]. etrfrMteaia cec n Applications and Science Mathematical for Center 2 26 tvnG Avery G. Steven where .Hwvr euetemtositoue n[ in introduced methods the use we However, ]. S S 3 xodS,Cmrde A02138 MA Cambridge, St, Oxford 1 large / 8 oeS,Poiec,R 02912 RI Providence, St, Hope 182 2 2 d = = d = 2 2 or 1 κ κ eateto Physics of Department i = b 4, 2 2 residual a avr University Harvard rw University Brown g Z Z 4, N µν uesmer teeyangle every at supersymmetry d d a 4 4 N ukadU .Schwab W. U. Burkhard , 1 = Abstract xǫ e xe , = – 18 = uegaiy eakby ayo u results our of many Remarkably, supergravity. 1 µνρσ ag reo o nyi h eilsia treat- semiclassical the in only not freedom gauge e [email protected] hr a enteedu rgesi under- in progress tremendous been has there ] µ a µa uegaiyaddrv h ote charge Noether the derive and supergravity 1 e I ν b e ψ νb η aclt h soitdagbao charges, of algebra associated the calculate , µ ab R γ ( g 5 ω γ µν σ n rmteato.Sbeunl we Subsequently action. the from ons ) µνab D en h ilenfield vielbein the define , tsiosa smttcnl infinity null asymptotic at spinors nt esf ii ftegaiiowhich gravitino the of limit soft he ν y tigter [ theory string ry, uesmerz h supertrans- the supersymmetrize y etwr fVahsa yo [ Lysov Vyacheslav of work dent ψ otesf ii fthe of limit soft the to t ρ er fteRarita–Schwinger the of metry (2) . e diinly we Additionally, le. h (unextended) the h 28 b lations. .Tepeiul neglected previously The ]. ntelnug f[ of language the in re npure in tries Brown-HET-1689 ovnetfrour for convenient 27 19 ocnetthe connect to ] – 24 d ,admost and ], e µ a via 1 ]. (3) (1) 29 ] a with η = diag(−1, +1, +1, +1) and e = det eµ = √−g. The γµ-matrices are given by the contrac- a tion of the numerical matrices γa in a given Lorentz frame with the frame field eµ. The covariant derivative µ is given by ∇ κ ψ = D ψ + Γ ψ , (4) ∇µ ν µ ν µν κ ab where Dµψν = ∂µψν + ωµabγ ψν is the spin covariant derivative. We used the explicitly four dimensional form of S3/2 with γ5 = iγ0γ1γ2γ3, the product of the numerical matrices. Note that, in the 1.5 order formalism, the spin connection ωµab is an a priori independent variable. By solving its (algebraic) equation of motion, one finds

ωµab = ωµab(e)+ Kµab, (5) where the first part is defined by a function of the frame field and its derivatives. The contortion Kµab is quadratic in the gravitino field ψµ. Their exact forms can be found in, e.g., [30], and are inessential for the current discussion. The curvature two-form is defined via

1 ab [Dµ, Dν]= Rµνabγ . (6) 8 3 Noether two-form

For simplicity we assume a purely bosonic asymptotically flat background. Thus any terms pro- a portional to the torsion Tµν in the following will be dropped. The transformations that define local supersymmetry are given by the gauge transformation of the RS field and a corresponding transformation on the frame field (which we could drop in our chosen background)

a 1 a δe = ǫγ ψµ, δψµ = Dµǫ (7) µ 2 where ǫ is an anticommuting spinor. It was shown [27, 31] (see also [32]) that there is a Noether charge density two-form kµν that is associated with every local symmetry. To derive k for the transformations in (7), we use the formalism developed in [27]. Any variation of the action may be written as

δS = d4x(−EIδφ + ∂ θµ(φ , δφ )) (8) Z I µ I I

I a where E are the equations of motion for the set of ﬁelds φI = {eµ, ψµ}. For supersymmetry variations (7), θµ is given by

ν 1 µνρσ θ = iǫ Dρǫγ γσψµ . (9) 2κ2 5 Conversely, the action is only symmetric under (7) up to a total derivative term

1 4 µνρσ 4 µ δS = d x∂ν iǫ ψ γ γσDµǫ = d x∂µK , (10) 2κ2 Z ρ 5 Z where we dropped terms in θ and K proportional to the variation of the spin connection δωµ since they will not contribute to the result. These two total derivative contributions deﬁne the ordinary µ µ µ Noether current j = θ − K . In this combination the aforementioned terms proportional to δωµ drop out. Finally, we need to ﬁnd the weakly vanishing current Sµ from Noether’s identity

I I µ E δφI = ∆IE + ∂µS (11)

µ1...µi for local variations of the form δΦ = f (Φ)λ + f ∂ ∂ i λ. For (7) this is I I i I µ1 ··· µ P i Sµ = (ǫµνρσǫγ γ D ψ ) (12) κ2 5 σ ν ρ and so νµ ν ν e νµκ ∂µk = j − S = ∂µ (ǫγ ψκ) (13) hκ2 i where for the last form we made use of (42). Although we didn’t make this explicit in the calculation, it is possible to show that the torsion does not enter k.

2 4 Boundary conditions and Asymptotic Killing Spinors

Following the approach of [27], we must gauge fix the local, bounded supersymmetry transformations to define the path integral measure. We use Witten gauge /ψ = 0. In Witten gauge, the gravitino wave equation becomes Dψ/ ν = 0 with leading asymptotic behavior in Bondi coordinates (38), (0) (0) (0) ψu = ψu (θ) ψr = ψr (θ) ψA = rψA (θ). (14) The equation of motion fixes the subleading behavior. The two-form takes the form 1 kµν = ǫ¯ γνµ /ψ + 2γ[µψν] . (15) 2κ2 Demanding the charge is finite for the asymptotic Killing spinors in Witten gauge implies the fall-off condition 2γ[uψr] = O r−2 . (16) The initial data in (14) is constrained by the gauge conditio n /ψ = 0, and the fall-off condition (16). (0) To wit, we may take the spinors on the sphere ψA as the physical data. The gauge fixing condition leaves unfixed a discrete set of large residual supersymmetries, or asymptotic Killing spinors. These have already been discussed from a different perspective [28]: there is an infinite family, which are the “square root” of the BMS supertranslations. The residual supersymmetries are parametrized by spinors solving the Dirac equation, Dǫ/ = 0, for which the transformation (7) preserves the boundary conditions discussed above. Solutions are parametrized by spinors that asymptote to arbitrary angle-dependent spinors η(θ) on I ±, ǫ = η(θ)+ O r−1 . (17) The “small” subleading pieces are gauge-dependent and do not contribute to the charges.

5 Algebra

Now, define a charge Q[η] for large supersymmetry transformations η. Q[η] can be written as 1 Q[η]= ⋆k = dS ηγµνρψ 2 µν ρ (18) Zσ κ Zσ where σ I is an S2 at u =− on I +. Then [33–35] ⊂ 1 Q[η ] Q∞[η ] = δ Q[η ]= dS η γµνκD η 1 , 2 η1 2 2 µν 2 κ 1 (19) h i κ Zσ which can be reformulated as 1 δ Q[η ]= dS η γµνκD η − (D η )γµνκη η1 2 2 µν 2 κ 1 κ 2 1 (20) 2κ Zσ where we made use of the compactness of S2 to drop a total derivative term. The remainder takes the Nester–Witten form of the diffeomorphism charge [34, 35]. It is possible to rewrite this, see e.g., 2 [36], using the linear approximation Dµ = Dµ + Ωµ + O(h ) where h is the perturbation to the background metric gµν − gµν = hµν, D the background derivative, and Ω the linear part. The a a a a a frame field is then given by eµ = eµ + hµ such that hµν = eµhνa + eνhµa. We use g to raise and a αβ lower coordinate indices and eµ to transform to a local Lorentz frame. Since Ωµ = Ωµ γαβ and γσγαβ + γαβγσ = 2γσαβ along with the identity (42) the commutator may be written as 1 dS ǫµνκσǫ η γρη Ω αβ 2 µν σαβρ 2 1 κ . (21) κ Zσ µνκ We dropped a term proportional toη ¯ [2γ Dκη1] which may be though of as a potential central ρ ρ charge. Define now, as usual, ξ = η2γ η1 as the parameter of a coordinate transformation and contract the antisymmetric tensors to get the result 6 dS ξ[µΩ νκ] 2 µν κ . (22) κ Zσ

3 While it is almost a standard calculation, let us anyhow show that we can transform this result µν into the form of BMS charges. Note that the linearized spin connection Ωκ can be written as

µν σ[ν µ] Ωκ = g δΓσκ (23) with δΓ µ = 1 gµρ( h + h − h ). Then Ω can be written in two ways σκ 2 ∇σ κρ ∇κ σρ ∇ρ κσ µν 1 σ[ν µ]α Ωκ = gκα σ g h 2 ∇ 1 σαµν [µ ν]σ [µ ν]σ = gκα σH − σ δκ h − δκ g h (24) 2 ∇ ∇ µα σν να µσ where we defined the quantity Hσαµν = gσνh + gµαh − gσµh − gανh , the trace h = κ µν µν 1 µν hκ, and the trace reversed metric perturbation h = h − 2 g h. Inspecting the expression [µ νκ] ξ Ωκ and inserting the two expressions in (24) for Ω, we find that 1 Q[η ] Q[η ] = T[ξ]= dS ξ Hσαµν 1 , 2 2 µν α σ (25) h i κ Zσ ∇ which differs from the flux integral of the Barnich–Brandt Noether two-form [31] for diffeomor- phisms by a boundary term and is thus equivalent under the integral. Note that ξµ as defined above using the asymptotic Killing spinors (17) has a finite value at the boundary, i.e., T[ξ] consti- tutes a BMS translation [37, 38] in tune with [1, 26, 28]. Furthermore, the bracket [Q[η], T[ξ]] = 0 as expected.

6 Ward identity

The Ward identity associated with the two-form derived above is given by [27]

δη(Φ1 Φn) = i Φ1 Φn ⋆j[η]− ⋆j[η] (26) h ··· i h ··· ZI + ZI − i where Φ are the ﬁelds of d = 4, N = 1 supergravity. Here δΦ are local supersymmetry transformations (7) while the right hand side introduces the operator (18) into the path integral. Note that µ µν the Noether current j[η] is related to the two-form k in (13) simply via j = ∂νk up to equations of motion. µ + Using Witten gauge γ ψµ = 0 and Bondi coordinates (38), the integral over I on the right hand side of eq. (26) can be written as (analogously for I −) 1 Q[η]= dΣ η D γµνκψ + γµνκD ψ 2 µ ←−ν κ ν κ (27) κ ZI + h i similarly for I −. All derivatives in this section ff. are background derivatives D. We use the µνρ µ µ Witten gauge condition, the equations of motions γ Dνψρ = J with the supercurrent J , and the Majorana ﬂip to write

1 µ µ ν ν µ Q[η]= dΣµ ηJ − ψ γ Dνη + ψ γ Dνη . (28) κ2 ZI + h i µ The second term vanishes due to the residual gauge condition γ Dµη = 0. Thus with η|I = η(z,¯z) 1 2 r Q[η]= d zdu η J + ←−Dzγuψz¯ + ←−Dz¯ γuψz . (29) κ2 ZI + h i 7 Gravitino Soft Factor

From the Ward identities on correlation functions, we extend the result to S matrix elements by LSZ reduction following [39]. The boundary ﬁelds ψz,z ¯ |I in (29) can be found by a limiting pro- ± cess on the asymptotic mode expansions (see (43)) as in [5–11, 25]. The result is (cp is the gravitino annihilation operator)

i√2 + + −iEu − † iEu ψz = dEu (c e − (c ) e ) (30) 8π2(1 + zz¯) Z Exˆ Exˆ Exˆ

4 ± ∓ + − and ψz¯ the same with interchanged helicities. We used u = v . The spinor, u (u ), is right- µ handed (left-handed) in a helicity basis for the γ . Thus we may use the projectors PR (PL) to further reduce the charge to

1 2 r Q[η]= d zdu η J + ←−DzγuPLψz¯ + ←−Dz¯ γuPRψz . (31) κ2 ZI + h i

iE0u Only ψz,z ¯ are dependent on the coordinate u, so we deﬁne Ψz,z ¯ = limE0→0 du e ψz,z ¯ with a regulating factor of exp(iE0u) to extract the zero modes. η can be chosenR left-handed or right- handed to single out speciﬁc terms in Q[η]. The second and third term in the charge Q[η] can −1 −1 even be localized if η consists of a function (z − zi) (or (z¯ − z¯i) ) multiplied by a right-handed (left-handed) spinor that is constant with respect to Dz¯ (or Dz), i.e.,

1 1 ηα = χα, or ηα = χα. (32) ˙ z − w ˙ z¯ − w¯ More generally η = ǫ(z,¯z)χ with ǫ(z,¯z) an arbitrary function. The zero mode of the supercurrent has trivial action on the particle vacuum, duJr|0 = 0, but the rest of the charge Q[η] inserts i a zero momentum gravitino and acts like Rsoft charge Qs[η] in the terminology of [5–11, 25], i.e., Q[η]= Qh[η]+ Qs[η] generates a spontaneously broken symmetry. Finally, let us inspect the soft limit of scattering amplitudes Mn with n particles and one soft, + positive helicity gravitino denoted ψs with soft momentum ps. Fermions come in pairs; there is at least one more gravitino in the set {1,..., n}. All particles are considered outgoing. Then

n + ψ+ lim Mn+1(..., ψs )= Si Mn(..., DΦi,...) (33) ps→0 Xi=1

+ + ψ ǫs ·pi + + µ − th S = ǫ u γ v D 1/2 where i pi.ps i,µ ¯ s i and lowers the helicity of the i external leg by

Dh+ = ψ+, Dψ− = h−, Dψ+ = Dh− = 0. (34)

In the last two cases above, the amplitude on the right hand side vanishes. Using p2 = 0, let µ ¯ + ¯ pµσ = λpλp spinor helicity variables. Since gravitinos are Majorana we haveu ¯ s = (λs,0) and − + λxλ¯ i vi = (0, λi) in a helicity basis. Then the polarization can be written as ǫi .σ = hx,ii where λx is an arbitrary reference spinor and σµ the Pauli matrices. In the last equation we employ the ψ+ [s,i]hx,ii usual bracket notation, see, e.g., [40]. Then Si may be written as hs,iihx,si , compare [41]. Thus the positive helicity gravitino soft limit is the combination of a helicity lowering supersymmetry ψ+ transformation and the multiplication of an angle dependent factor Si [26]. The soft limit (33) can be related to the Ward identity (26) when employing an LSZ reduction [39]. Using asymptotic expansions for the graviton ﬁeld hµν and the gravitino ﬁeld ψµ (see (43)) with in-state annihilators a±, respectively c±, for particles with chirality , the action of a super- p p ± symmetry variation δs with large parameter η is given by

+ + + µ − + δηap = [Q[η], ap ]= ǫp,µηγ¯ vp cp , (35) − − + µ − − δηcp = [Q[η], cp ]= ǫp,µηγ¯ vp ap (36) where here, as above, Q[η] bosonic. The other two cases are similar and would be necessary for a discussion of the Ward identity for a negative helicity gravitino.We match this onto the right ψ+ hand side of the soft limit above by remarking that Si is a special case of the coefﬁcient of the ψ+ commutation relations above. This statement is clearer when writing Si with the help Bondi coordinates ψ+ 1 + zsz¯s + + µ − Si = u¯ s ǫi,µγ vi . (37) √2Es(zs − zi)

We may discard the divergence in Es by multiplying (33) with √Es since λ¯ s √Es as well as pull 1/2 ∝ ( + zz) √E η = 1 u− out the factor of 1 ¯ . Then, if we let s zs−zi s we see that it is exactly of the form given in (32). It follows that the S matrix statement of the Ward identity (26) is the leading soft gravitino limit. The analogous statement for a negative helicity gravitino can be derived in the same way.

5 Acknowledgments We are grateful to Vyacheslav Lysov for sharing a draft of his paper and co- ordinating the arXiv submission of our papers. BS was supported by the Center for Mathematical Sciences and Applications at Harvard University and the Cheng Yu-Tung fund, and NSF grant 1205550. BS is thankful to Andrew Strominger and Thomas Dumitrescu for discussions. SGA is supported by US DOE grant de-sc0010010.

A Conventions

We follow the conventions of [30]. Throughout the paper, the preferred background is asymptotically ﬂat space with metric gµν, strictly bosonic, and excluding black hole spacetimes. Retarded Bondi coordinates are given by the metric

ds2 =−du2 − 2dudr + r2dΩ2 (38)

2 4 ~2 ~ 2 with dΩ = ~ dθ . θ R covers the sphere once. In the text, we use z = θ + iθ . The frame (1+θ2)2 ∈ 1 2 is deﬁned by 0 1 A 2r A e = du + dr e = dr e = ~ dθ (39) 1 + θ2 with spin connection,

eA 1 ω =−ω = ω = (θ e − θ e ). (40) A1 1A r AB r A B B A The γ Clifford algebra is defined in the usual way by {γµ, γν} = 2gµν. All calculations use the Majorana flip extensively; given two anticommuting spinors ξ, χ and a set of γ-matrices, the flip is defined by ξγµ1 γµr χ = t χγµ1 γµr ξ (41) ··· r ··· where a convenient choice for tr is tr = −1 for r = 1,2mod4 and tr = 1 for r = 0, 3 mod 4 in d = 4. µνκ = 1 [µ ν κ] We also use the notation γ 3! γ γ γ and the identity

µνκ −1 µνκσ γ =−ie ǫ γ5γσ. (42)

Asymptotic expansions for the ﬁelds are given by

d3p h (x)= (ǫσ,⋆ aσeip.x + ǫσ (aσ)†e−ip.x) µν Z ( π)3 E p,µν p p,µν p σX=± 2 2 3 d p ⋆ ψ (x)= (ǫσ, uσcσeip.x + ǫσ vσ(cσ)†e−ip.x). (43) µ Z ( π)3 E p,µ p p p,µ p p σX=± 2 2

References

[1] Andrew Strominger. “On BMS Invariance of Gravitational Scattering”. In: JHEP 07 (2014), p. 152. DOI: 10.1007/JHEP07(2014)152. arXiv:1312.2229 [hep-th] (cit. on pp. 1, 4). [2] Andrew Strominger. “Asymptotic Symmetries of Yang-Mills Theory”. In: JHEP 07 (2014), p. 151. DOI: 10.1007/JHEP07(2014)151. arXiv:1308.0589 [hep-th] (cit. on p. 1). [3] Andrew Strominger and Alexander Zhiboedov. “Gravitational Memory, BMS Supertrans- lations and Soft Theorems”. In: (2014). arXiv:1411.5745 [hep-th] (cit. on p. 1). [4] Andrew Strominger. “Magnetic Corrections to the Soft Photon Theorem”. In: (2015). arXiv:1 509.00543 [hep-th] (cit. on p. 1). [5] Temple He et al. “New Symmetries of Massless QED”. In: JHEP 10 (2014), p. 112. DOI: 10.1 007/JHEP10(2014)112. arXiv:1407.3789 [hep-th] (cit. on pp. 1, 5). [6] Temple He et al. “BMS supertranslations and Weinberg’s soft graviton theorem”. In: JHEP 05 (2015), p. 151. DOI: 10.1007/JHEP05(2015)151. arXiv:1401.7026 [hep-th] (cit. on pp. 1, 5).

6 [7] Temple He et al. “2D Kac-Moody Symmetry of 4D Yang-Mills Theory”. In: (2015). arXiv:15 03.02663 [hep-th] (cit. on pp. 1, 5). [8] Daniel Kapec et al. “Asymptotic Symmetries of Massless QED in Even Dimensions”. In: (2014). arXiv:1412.2763 [hep-th] (cit. on pp. 1, 5). [9] Daniel Kapec et al. “New Symmetries of QED”. In: (2015). arXiv:1506.02906 [hep-th] (cit. on pp. 1, 5). [10] Daniel Kapec et al. “Higher-Dimensional Supertranslations and Weinberg’s Soft Graviton Theorem”. In: (2015). arXiv:1502.07644 [gr-qc] (cit. on pp. 1, 5). [11] Vyacheslav Lysov et al. “Low’s Subleading Soft Theorem as a Symmetry of QED”. In: Phys. Rev. Lett. 113.11 (2014), p. 111601. DOI: 10.1103/PhysRevLett.113.111601. arXiv:140 7.3814 [hep-th] (cit. on pp. 1, 5). [12] T. Banks. “The Super BMS Algebra, Scattering and Holography”. In: (2014). arXiv:1403.3 420 [hep-th] (cit. on p. 1). [13] Tom Banks. “Current Algebra on the Conformal Boundary and the Variables of Quantum Gravity”. In: (2015). arXiv:1511.01147 [hep-th] (cit. on p. 1). [14] Miguel Campiglia and Alok Laddha. “Asymptotic symmetries and subleading soft graviton theorem”. In: Phys. Rev. D90.12 (2014), p. 124028. DOI: 10.1103/PhysRevD.90.124028. arXiv:1408.2228 [hep-th] (cit. on p. 1). [15] Miguel Campiglia and Alok Laddha. “Asymptotic symmetries of gravity and soft theorems for massive particles”. In: (2015). arXiv:1509.01406 [hep-th] (cit. on p. 1). [16] Miguel Campiglia. “Null to time-like infinity Green’s functions for asymptotic symmetries in Minkowski spacetime”. In: JHEP 11 (2015), p. 160. DOI: 10.1007/JHEP11(2015)160. arXiv:1509.01408 [hep-th] (cit. on p. 1). [17] Miguel Campiglia and Alok Laddha. “Asymptotic symmetries of QED and Weinberg’s soft photon theorem”. In: JHEP 07 (2015), p. 115. DOI: 10.1007/JHEP07(2015)115. arXiv:15 05.05346 [hep-th] (cit. on p. 1). [18] Miguel Campiglia and Alok Laddha. “New symmetries for the Gravitational S-matrix”. In: JHEP 04 (2015), p. 076. DOI: 10.1007/JHEP04(2015)076. arXiv:1502.02318 [hep-th] (cit. on p. 1). [19] Steven G. Avery and Burkhard U. W. Schwab. “BMS, String Theory, and Soft Theorems”. In: (2015). arXiv:1506.05789 [hep-th] (cit. on p. 1). [20] Burkhard U. W. Schwab. “A Note on Soft Factors for Closed String Scattering”. In: JHEP 03 (2015), p. 140. DOI: 10.1007/JHEP03(2015)140. arXiv:1411.6661 [hep-th] (cit. on p. 1). [21] Burkhard U. W. Schwab. “Subleading Soft Factor for String Disk Amplitudes”. In: JHEP 08 (2014), p. 062. DOI: 10.1007/JHEP08(2014)062. arXiv:1406.4172 [hep-th] (cit. on p. 1). [22] Massimo Bianchi et al. “More on Soft Theorems: Trees, Loops and Strings”. In: Phys. Rev. D92.6 (2015), p. 065022. DOI: 10 .1103 /PhysRevD.92 . 065022. arXiv:1406 . 5155 [hep-th] (cit. on p. 1). [23] Anastasia Volovich et al. “Double Soft Theorems in Gauge and String Theories”. In: JHEP 07 (2015), p. 095. DOI: 10.1007/JHEP07(2015)095. arXiv:1504.05559 [hep-th] (cit. on p. 1). [24] Paolo Di Vecchia et al. “Soft theorem for the graviton, dilaton and the Kalb-Ramond field in the bosonic string”. In: JHEP 05 (2015), p. 137. DOI: 10.1007/JHEP05(2015)137. arXiv:1502.05258 [hep-th] (cit. on p. 1). [25] Thomas T. Dumitrescu et al. “Infinite-Dimensional Fermionic Symmetry in Supersymmetric Gauge Theories”. In: (2015). arXiv:1511.07429 [hep-th] (cit. on pp. 1, 5). [26] Marcus T. Grisaru and H. N. Pendleton. “Soft Spin 3/2 Fermions Require Gravity and Su- persymmetry”. In: Phys. Lett. B67 (1977), p. 323. DOI: 10.1016/0370-2693(77)90383-5 (cit. on pp. 1, 4, 5). [27] Steven G. Avery and Burkhard U. W. Schwab. “Noether’s Second Theorem and Ward Iden- tities for Gauge Symmetries”. In: (2015). arXiv:1510.07038 [hep-th] (cit. on pp. 1–4).

7 [28] M. A. Awada et al. “Conformal Supergravity, Twistors and the Super BMS Group”. In: An- nals Phys. 171 (1986), p. 52. DOI: 10.1016/S0003-4916(86)80023-9 (cit. on pp. 1, 3, 4). [29] Vyacheslav Lysov. “Asymptotical Supersymmetries from Gravitino Soft Theorem”. In: (2015). to appear (cit. on p. 1). [30] Daniel Z. Freedman and Antoine van Proeyen. Supergravity. Cambridge University Press, 2012. ISBN: 978-0-521-19401-3(cit. on pp. 2, 6). [31] Glenn Barnich and Friedemann Brandt. “Covariant theory of asymptotic symmetries, con- servation laws and central charges”. In: Nucl. Phys. B633 (2002), pp. 3–82. DOI: 10.1016/S 0550-3213(02)00251-1. arXiv:hep-th/0111246 [hep-th] (cit. on pp. 2, 4). [32] Vivek Iyer and Robert M. Wald. “Some properties of Noether charge and a proposal for dynamical black hole entropy”. In: Phys. Rev. D50 (1994), pp. 846–864. DOI: 10.1103/Phy sRevD.50.846. arXiv:gr-qc/9403028 [gr-qc] (cit. on p. 2). [33] C. M. Hull. “The Positivity of Gravitational Energy and Global Supersymmetry”. In: Com- mun. Math. Phys. 90 (1983), p. 545. DOI: 10.1007/BF01216185 (cit. on p. 3). [34] Edward Witten. “A Simple Proof of the Positive Energy Theorem”. In: Commun. Math. Phys. 80 (1981), p. 381. DOI: 10.1007/BF01208277 (cit. on p. 3). [35] James A. Nester. “A New gravitational energy expression with a simple positivity proof”. In: Phys. Lett. A83 (1981), p. 241. DOI: 10.1016/0375-9601(81)90972-5 (cit. on p. 3). [36] Sergio Cecotti. Supersymmetric Field Theories: Geometric Structures and Dualities. Cambridge University Press, 2015. ISBN: 9781107053816, 9781316213599. URL: http://www.cambrid ge.org/mw/academic/subjects/physics/theoretical-physics-and-mathe matical-physics/supersymmetric-field-theories-geometric-structures -and-dualities?format=HB (cit. on p. 3). [37] H. Bondi et al. “Gravitational waves in general relativity. 7. Waves from axisymmetric iso- lated systems”. In: Proc. Roy. Soc. Lond. A269 (1962), pp. 21–52. DOI: 10.1098/rspa.1962 .0161 (cit. on p. 4). [38] R. K. Sachs. “Gravitational waves in general relativity. 8. Waves in asymptotically ﬂat spacetimes”. In: Proc. Roy. Soc. Lond. A270 (1962), pp. 103–126. DOI: 10.1098/rspa.1962.0206 (cit. on p. 4). [39] Marcus T. Grisaru et al. “Supergravity and the S Matrix”. In: Phys. Rev. D15 (1977), p. 996. DOI: 10.1103/PhysRevD.15.996 (cit. on pp. 4, 5). [40] Henriette Elvang and Yu-tin Huang. “Scattering Amplitudes”. In: (2013). arXiv:1308.1697 [hep-th] (cit. on p. 5). [41] Zheng-Wen Liu. “Soft theorems in maximally supersymmetric theories”. In: Eur. Phys. J. C75.3 (2015), p. 105. DOI: 10.1140/epjc/s10052-015-3304-1. arXiv:1410.1616 [hep-th] (cit. on p. 5).