Progress in Parc and Nuclear Physics 95 (2017) 98-159 Aeron Contents lists available at ScienceDirect Se hl Progress in Particle and Nuclear Physics ELSEVIER journal homepage: www elsevier.con/loestelppnp Review Higgs boson production and decay at hadron colliders Dem Michael Spira Poul Scherer stu, C5222 Vien Ph, Swzerland ARTICLE INFO ABSTRACT ‘Brae hiory ‘Higgs physics at hadron colliders as the LAC e reviewed within the Standard Model (SM) ‘Available online 13 Apt 2017 and its minimal supersymmetric extension (MSSM) by summarizing the present state-of- ——_______________ the-art of theoretical predictions forthe production cross sections and decay rates, ewords ‘© 2017 Elsevier BV. All ight reserved. Higgs phys Standard moe Supersymmetty Higher-order corrections Contents 1. Introduction, 99 LL Organization ofthe review. 99 12, standard model 99 13. Minimal supersymmetric standard model 100 2. Hlggs-boson decays 103 2.1. Fermionic Higgs boson decays 103, 21, Standard model 103, 2.12. Minimal supersymmetric extension 105, 22. Higgs boson decays into intermediate gauge bosons. 107 22.1, Standard model. 107 222. Minimal supersymmetric extension. 109 2.3. Higgs boson decays into gluons 108 23:1, _ Standard model 108 2.32. Minimal supersymmetric extension a2 2.4, Higgs boson decays into photons. us 241, Standard model 5 2.42. Minimal supersymmetric extension ny 25. Higgs boson decays into photon and Z boson and Dalitz decays, ne 25.1, Standard model 19 252. Minimal supersymmetric extension. 120 26. Supersymmettc Higgs boson decays into Higgs particles cy 27. Supersymmetric Higgs boson decays 124 28. Branchingratis and total decay width, 128 28.1, _ Standard model 128 282. Minimal supersymmetric extension 19 3. Higgs-boson production. a 3.1. Gluon fusion: gg > #1 BI 3.11, Standard model BI 3.1.2. Minimal supersymmetric extension 1s ral address: Michael Sptapsch, pf doiorg 101036) ppap 2017-04001 (0145-5410}6 2017 Elsevier BY. Al ights reserved1M Spta/ Progress n Pace and Nuclear Physics 95 (2017) 98-159, 9 3.2. Vector-boson fusion: qq + qaV*V" > galt 136 32.1, Standard model 136 322. Minimal supersymmetric extension 139 33. Higgs Strablung: gj —> V* > VET 139 33.1, _ Standard model. 139 33.2. Minimal supersymmetric extension 140 34. Higgs bremsstrahlung off top and bottom quarks 141 34.1, Standard model 141 3.42. Minimal supersymmetric extension. us 35. Summary of single-Higgs boson production crass sections. 148 35.1, Standard model 18 352. Minimal supersymmetric extension. 149 35. Higgs boson pai production 149 36.1, Standard model 19 362. Minimal supersymmetric extension. 152 4. Summary 152 ‘Acknowledgements 153 References 153, 1 Introduction 1.1, Organization of the review In this work we will review all Higgs decay widths and branching ratios as well as all relevant Higgs boson production ‘cross sections at the LHC within the SM and MSSM, Previous reviews can be found in Refs. [1]. This work is a substantial update of Re, [2] The paper is organized as follows. This Section 1 will provide an introduction to the SM and MSSM Higgs sectors. In Section 2 we will review the decay rates of the SM and the MSSM Higgs particles providing explicit analytical results of the individual partial decay widths where possible. Section 3 will discuss the production cross sections for the Higgs bosons of the SM and MSSM and the impact of the numerical results on the profile of the Higgs bosons at the LHC. A summary will be given in Section 4, 1.2, Standard model The discovery ofa resonance atthe LHC [3] that is compatible with the Standard-Model (SM) Higgs boson [4] marked ‘a milestone in particle physics. The existence of the Higgs boson is inherently related to the mechanism of spontaneous symmetry breaking [5] while preserving the full gauge symmetry and the renormalizability ofthe SM [6]. Its mass, the last missing parameter of the SM, has been measured tobe (125,08-+0.24) GeV [4]. The existence of the Higgs boson allows the SM particles to be weakly interacting up to high-energy scales without violating the unitarity bounds of scattering amplitudes [7]. This, however, is only possible for particular Higgs-boson couplings to all ather particles so that with the knowledge of the Higgs-boson mass all its properties are uniquely fixed. The massive gauge bosons and fermions acquire mass through their interaction with the Higgs field that develops a finite vacuum expectation value in its ground state thus hiding the stil unbroken electroweak gauge symmetry. The minimal model requires the introduction of one weak isospin doublet of the Higgs field and leads ater spontaneous symmetry breaking to the existence of one scalar Higgs boson, while non-minimal Higgs sectors predict in general more than one Higgs boson, ‘Within the SM with one Higgs doublet the Higgs massis constrained by consistency conditions asthe absence of a Landau pole for the Higgs self-coupling up to high energy scales and the stability of the electroweak ground state [8] Ifthe SM is required to fulfil these conditions for energy scales upto the scale of Grand Unified theories (GUTS) of ~10” GeV the Higgs mass is constrained between 129 GeV and about 190 GeV [9]. The last condition of vacuum stability can be relaxed by demanding the lifetime ofthe ground state to be larger than the age of our universe | 10]-This reduces the lower bound on the Higgs mass to about 110 GeV [2]. The measured value of the Higgs mass indicates that our universe is unstable with a large lifetime far beyond the age ofthe universe If the SM is extended to the GUT scale, radiative corrections to the Higgs-boson mass tend to push the Higgs mass towards the GUT scale, if it couples to particles ofthat mass order. In order to obtain the Higgs mass at the electroweak scale the counter term has tobe fine-tuned to cancel these large corrections thus establishing a very unnatural situation that requites a solution. This is known as the hierarchy problem [1] Possible solutions are the introduction of supersymmetry (SUSY) 12,13], collective symmetry between SM particles and heavier partners asin Little Higgs theories [14] oraneffective ‘eduction ofthe Planck and GUT scales as in extra-dimension models at the TeV scale |15)100 1M Spta/ Progress Paice and Nuclear Physics 95 (2017) 9 ‘SM Higgs bosons are dominantly produced via the loop-induced gluon-fusion process gg — H with top quarks providing te leading loop contribution [15]. Other production processes as vector-boson fusion qq —» qgHt [17], Higgs-strahlung 4 > H+ W/2 [18] and Higgs radiation off top quarks gz, gj — «TH [19] are suppressed by more than one order of magnitude, While the dominant gluon-fusion process allows for the detection of the Higgs boson in the rare decay modes. into yy. four charged leptons and W-boson pairs, other decay modes are only detectable inthe subleading production modes 4 eg, the main Higgs-boson decay into bottom quarks in strongly boosted Higgs-strablung [20] or the Higgs-boson decay into ¢-leptons in the vector-boson-fusion process [21] 1.3, Minimal supersymmetric standard model Supersymmetric extensions of the SM are motivated by providing a possible solution to the hierarchy problem if the supersymmetric particles appear at the few-TeV scale [22]. Supersymmetry connects fermionic and bosonic degrees of freedom and thus links internal and external symmetries | 12]. It is the last possible symmetry-type of $-matrix theories, ice. the last possible extension of the Poincaré algebra, The minimal supersymmetric extension of the SM (MSSM) predicts the Weinberg angle in striking agreement with experimental measurements ifembedded in a GUT [23]. Moreover, it allows to generate electroweak symmetry breaking radiatively [24 and yields a candidate for Dark Matter if R-parity is conserved [25] which renders the lightest supersymmetric particle stable. Finally it increases the proton lifetime beyond experimental bounds in the context of supersymmetric GUTS [23,26] The MSSM requires the introduction of two isospin doublets of Higgs fields in order to maintain the analyticity of the superpotential and the anomaly-freedom with respect to the gauge symmetries [25,27], Moreover, two Higgs doublets are needed for the generation of the up- and down-type fermion masses. The Higgs sector is a TWwo-Higgs-Doublet model (2HDM) ‘of type Il The mass eigenstates consist of alight (h) and heavy (H) scalar, a pseudoscalar (A) and two charged (H) states. The self-interactions of the Higgs fields are entirely fixed in terms of the electroweak gauge couplings so that the self-couplings are constrained to small values. This leads to an upper bound on the light scalar Higgs mass that has to be smaller than. the Z-boson mass My at leading order (0). Ths is, however, broken by radiative corrections, which are dominated by top- ‘quark-induced contributions [28}. The parameter tg6, defined as the ratio of the two vacuum expectation values of the Scalar Higgs fields, will in general be assumed to be in the range 1 < tg < m,/ms, where m,(ms) denotes the top (bottom) ‘mass, consistent with the assumption that the MSSM is the low-energy limit of a supergravity model [29] and to avoid non-perturbative phenomena, If the soft SUSY-breaking parameters do not contain any complex phases the input parameters of the MSSM Higgs sector at LO are generally chosen to be the mass My of the pseudoscalar Higgs boson and tgp. All other masses and the mixing angle « between the scalar CP-even Higgs states can be derived from these basic parameters (and the top mass and SUSY parameters, which enter through radiative corrections). The radiative corrections can be approximated by the parameter ¢, ‘which grows with the fourth power of the top quark mass and logarithmically with the stop masses m,,,. supplemented by terms originating from soft SUSY-breaking parameters, i. the trilinear coupling A,, the higgsino mass j. and the third- generation squark mass Ms with X, = A, — 14/Ig6. 3G mio) [5 (mama) 2 (yt a Var? sint 6 |? \imzquny) me I Tame where Gy denotes the Fermi constant and 7x(Ms) the MS top mass athe scale Ms. These corrections ate positive and reach $ maximum (related to X,) for [ke = VOMg, They increase the squared mass ofthe ight neutral Higgs boson h to Mz mgeor pi] 6) z [a Me ‘The masses of the heavy neutral and charged Higgs bosons are determined by the sum rules (valid for this approximation), Mj = M3 +M} —M} +e Mi. = ME + Ma, 4 ‘The effective mixing parameter a between the CP-even scalar Higgs states canbe derived as MRM pa = tgp ——Mi Me __ Mea = 2b «jean 5)1M Spta Progress n Pace and Nuclear Physics 95 (2017) 98-159, 01 Tablet SSM Higgs couplings fo up- and Gown-sype fermions and gauge bosons w, ‘SM co a ae eo a v T i eosajsinp ——snajeosP snipe) Ho Sing/sing — casufeash cosa) A igs ie ° ‘The radiative corrections to the MSSM Higgs sector have been calculated up to the two-loop level in the effective potential approximation and in the diagrammatic approach [28]. The leading corrections are also known at the three-loop level ‘within the effective potential approach [30]. The corrections beyond next-to-leading order (NLO) are dominated by the QCD corrections to the top-quark-induced contributions. They decrease the upper bound on the light scalar Higgs mass Ma by about 10 GeV to ~135 GeV, while leaving a residual uncertainty of ~3 GeV on the light scalar Higgs mass In the context of inereasing lower mass bounds for the supersymmetric particles, a resummation of large logarithms related to the SUSY-particle masses has been performed forthe calculation of the MSSM Higgs-boson masses [31]. For large SUSY particle masses tis resummation may lead to effects ofthe order of 5 GeV (or lager for small values of tg) on the light scalar Higgs For on-shell external Higgs bosons there is an additional relevant effect emerging from finite external Higgs momenta leading to an additional mixing between different types of Higgs bosons. This can be treated by the introduction ofan external mixing matrix as discussed in Re. [32] that ensures that the external Higgs bosons are defined as proper on-shell states. The numerical effects are sizeable for certain MSSM scenarios and have to be taken into account in all processes with external MSSM Higgs bosons. ‘The calculation of the radiative corrections has also been extended to the effective trilinear and quartic self-interactions of the MSSM Higgs bosons that ate consistently defined at vanishing external momenta. The NLO corrections at (o,) and O(a) are available since along time [33], where ay = Gem? /(V2sin® Band ap = Gemi/(/2 cos? f).In the leading approximation in terms ofthe parameter ¢ the trilinear couplings of the neutral Higgs bosons are given up to O(a.) by [33] Daw = 3cos(2a)sin( p+ a) + aS ang = 2sin(2a)sin( +) cos(2a) cos(f +a) + fy imewte Aum = ~2sin(2e) cos.f + a) — cos(2a) sin( + a) + ae anny = 3 c05(2a) cOs(B + 2) 4 ie we Anan = c0s(26)sin( +a) + iz ere dana = ~ c05(28)c0s(6 + 0) + aS © Which have been normalized to V7G-M}. Similar but more involved expressions can be obtained forthe quartic Higgs self couplings, Some time ago the NLO corrections to the Higgs self-couplings have been extended by the next-to-next-to-leading ‘order (NNLO) corrections of O{axa,) [34] where a, denotes the strong coupling constant. The radiative corrections are large in general, while the NNLO partis of moderate size but reduces the theoretical uncertainties significantly tothe few per-cent level. ‘The couplings of the neutral Higgs bosons to fermions and gauge bosons depend on the angles « and f. Normalized to the SM Higgs couplings. they are listed in Table 1. The pseudoscalar particle A does not couple to gauge bosons at tree level, and its couplings to down (up)-type fermions are (inversely) proportional to tgB. For large values of tg6 the Yukawa couplings to (up) down-type quarks ae (suppressed) enhanced and vice versa apart from the regions where the light (heavy) scalar is close to its upper (lower) mass bound, where their coupling factors approach the SM-coupling strengths (up to sign differences for the heavy scalar). The couplings ofthe light scalar Higgs particle approach the SM values or large pseudoscalar masses, ie. in the decoupling regime. Thus it willbe dificult to distinguish the light scalar MSSM Higgs boson from the SM Higgs particle inthe region where all Higgs particles except the light scalar one are very heavy.102 IM Spta Progress in Pace and Nuclear Physics 95 (2017) 9 Fig. 1. Feynman diagrams contributing tothe bottom Yukawa couplings 3¢ (ele) and O00) (eight) In addition to these coupling factors the bottom Yukawa coupling in particular is strongly modified by radiative corrections due to sbottom-gluino exchange in the genuine SUSY-QCD part and by stop-chargino exchange in the leading SUSY-clectroweak part, see Fig, 1, The leading contribution behaves as O(s m CE /MBxy as") (with the gluino mass mz) ‘and can thus be very large fo lage values of gf [35]. In order to improve the perturbative results these contributions have to be resummed up to all orders. This can be done in the current-eigenstate basis involving the two Higgs doublets ¢ and 42 that couple to down- and up-type quarks, respectively. The leading correction to the bottom Yukawa coupling emerges {fom a coupling of the ‘wrong’ doublet gto the bottom quarks and can be discussed in terms of the effective Lagrangian az Ab oe ta = ~asie [ee + At ge |» the teh = mbt ie Z]o— afar (1— a oat (1+ a8) at (1- Ze) sa] o in the low-energy limit where the relations between current and mass eigenstates of the neutral Higgs components 9 = © [v, +Heosee = hsina + isin p — iG? cos] vi 1 f= AL [vp +Hsine + heose + iAcos B + iG? sin 8 aa 8 #) (8) have been used and G® denotes the neutral would-be Goldstone component. The indices L,R denote the chialtis ofthe bottom states y the Bottom Yukawa coupling ofthe MSSM Lagrangian and v ~ 246 GeV the SM vacuum expectation value The leading contributions to the correction Ay at NLO are given by [35] AB asl agit agin? Gr a (m2 =F Sms a tgp msm hy my ‘Aci 8 ond) hoe Vit? m2 M2 2 2 Mz y2) 3 2) m2 M2. u2) age My utes {008 mg, ma) ( +3) rom, m2) + +e) sn, MB } apt? = 2M, utes {en Bo ME) + STC ME, 2) + Steg Mn) + Bg wf ° where Cy = 4/3 and dy denotes the top Yukawa coupling and a = g/4, a = g?/4z the electroweak gauge couplings. The masses m, , and mt, , are the sbottom and stop masses, the higgsino mass parameter and M;,z the soft SUSY-breaking Dino and wino inass parameters, The variables s/c, = sin/coS@, are related to the stop/sbottom mixing angles 6, ». The function I is generically defined as. blog +belog? + cao Ka,b.0)= 10) )= ea Pxb = ca =) 9) ‘The effective Lagrangian of Eq. (7) can be parametrized in compact form as. fog =~ TB [ah n+ at HB is A] ay1M Spita Progress in Paice and Nustear Physics 95 (2017) 98-159. 02 f H--- wan dlagram describing H -> ff at lowest order. Soba ” Itshould be noted that in the decoupling limit of large pseudoscalar mass M, the mixing angle factor tga approaches —1/tg6 such that gf > gi! -> 1 so thatthe light scalar Higgs boson still becomes SMlike. By means of power counting it can be shown that the Lagrangian of Eq (7)1s valid up to all orders in j.tg@ and thus Ay [36-38] and potentially sizeable ‘Av-contibutions can be absotbed in a refined definition of A [37] The rewriting in terms of mass eigenstates describes the proper resummation ofthese terms. Two-loop QCD corrections to the dominant AE and 2" contributions have been calculated, They modify the size by a moderate amount of about 10% and reduce the scale dependence considerably to the level of a few per-cent [38,39] and thus yield a reliable prediction of the effective bottom Yukawa couplings. The couplings to all down-type fermions including the charged leptons can be dressed by the corresponding leading Ay terms Ssnalogotsly, They play a tole forthe strange and + Yukavra couplings, While the 4, contributions can be obtained ffom by straightforward replacements of the corresponding masses, the corrections for r leptons are given explicitly by [35] sit aso Mints {ion mi, M3) + ( 2) 4 (3-2) te, 23} Sr, Mea} (13) = - Ema wep flea 0) + Ea where s/c respectively. = sin/cos®, is related to the ¢ mixing angle 9, and ms, ;,m;, denote the stau and tau sneutrino masses, 2. Higgs-boson decays Higgs-boson couplings to SM particles grow with their corresponding masses so that for the SM Higgs boson its couplings to the intermediate gauge bosons W, Z and to the top and bottom quarks as well as z-leptons are the most relevant ones for decay and production processes, 2.1, Fermionic Higgs boson decays 21.1, Standard model ‘The decay widths of the SM Higgs boson into fermion pairs mediated by the fermion Yukawa couplings (see Fig. 2) is siven, for Mj, >> m}, by [40,41] My Win with Ne = 1(3) for leptons (quarks), the Higgs-boson mass My and the fermion mass my. Inthe leptonic case we use the lepton pole mass, while for quarks we use the MS quark mass ig. The QCD corrections for the decays into quark pairs can TlH > ffl = im} (1+ Sqco + 4 + Sms] (1 + Bete) (14) " Running-massefets with respect to QED corrections ae not taken nto account104 1M Spta Progress Pace and Nuclear Physics 95 (2017) 9 be expressed as [42-44] Seen = 5.672 (35,94 — 1 sen) ( 2 Pe sonzy( 2s(Mn)? + (164.14 —25.77N5 + 0.259np)( SY 2 y(n) (39.34 — 220.9Ny + 9.6851? — 0.0205N?)| (222) [1s — Jaa «Sr A] as a with the number N of contributing quark favours. The running quark mass and the QCD coupling are defined at the scale of the Higgs mass, absorbing inthis way large mass logarithms. The quark mass effects beyond the Yukawa coupling can be neglected in general, except for heavy quark decays in the threshold region.’ The QCD corrections in this case are given, in terms of the quark pole mass Mo, by [42] run > 0) = 2M ep? fy 4 16) WW > 06) = TE wg 5 (16) where f = (1 — 4M /mz)" denotes the velocity ofthe heavy quarks Q. The NLO QCD correction factor reads [42] py 2 2 13ptytog 28, 3 age SS SAB) + pgs 4 346? — 138 Ios AE + 7 - a7 with the function sou (“=2) vam, (22) ong Bg 2 sg =P fie 128) 204 ( 222) — tor! 4 a = 8 TSE 1+e ~ 205 vrs] = 3p lo (as) (Lis denotes the Spence function, Lix) = — /' dyy~" og(1 — y).) Electroweak corrections to heavy quark and lepton decays are well under control [45,46]. n the intermediate mass range they can be approximated by [47] } (as) My Ge 2 Me 3 2 /6up woth) + 22 faa [-s= 2 mesg] a with wy = 2ly — desi and ay = yy. ly denotes the third component of the electroweak isospin, e, the electric charge of the fermion f and sw = sin6w the sine of the Weinberg angle: « denotes the QED coupling, M; the top quark mass and ‘My.z the W, Z boson masses. The large logarithm log M3/M} can be absorbed in the running fermion mass analogous to the QCD corrections. The coefficient ky is equal to 7 for decays into leptons and light quarks; for b quarks itis reduced to 13, ,) adm) 4 > x {r-3 (Fon) e } forf Ab Bay Saw = 58) aay b> bx fi-aaeey (20) with g, = 72/6, The three-loop QCD corrections to the ky term can be found in [49]. Recently, the full mixed QcD- tlectroweak corrections Spiny have been determined {50}. These turn out tobe in the per-mile range if included in the factorized expression for QcD and electroweak corrections as shown in Eq, (14) The electroweak corrections in total are small, of (5%), in the intermediate mass range. The residual theoretical uncertainties in the intermediate mass range have been estimated a5 02x forthe QcD pat and 05% for electroweak corrections including mixed contributions inthe facorized approach [5 1]. 2 The inclusion ofthe full quark mas effects is mosivated by a proper sreatment of of-shell Msgs splittings into quark pais and by determining all subleadingquarkemas effects up to NID. Moreover, the heavy salt MSSM Higgs boson can aquire a mass lee to the location of the top-antcop Ahveshold1M Spta Progress in Paice and Nuclear Phys 95 (2017) 98-159, 10s, 2.1.2, Minimal supersymmetric extension ‘At lowest order the leptonic decay width of neutral MSSM Higgs boson® decays is given by [40,41] GeMe 42x where g? denotes the corresponding MSSM coupling factor, presented in Table 1, 6 = (1 — 4m?/M3)!* the velocity of the final-state leptons and p = 3(1) the exponent fr scalar (pseudoscalar) Higgs particle. The «pair decays play a significant role, with a branching ratio of up to about 10%, Muon decays can develop branching ratios of afew 10~*. All other leptonic decay modes are phenomenologically itrelevant. ‘The analogous leptonic decay width ofthe charged Higgs boson reads rw v= St a1 i) (22) The decay mode into xv, reaches branching ratios of more than 90% below the tb threshold and the muonic one ranges at afew 10-4. All other leptonic decay channels of the charged Higgs bosons are unimportant. For large Higgs masses [M3 > M3] the QCD-corrected decay widths ofthe MSSM Higgs particles into quarks can be ‘obtained from evaluating the analogous diagrams as presented in Fig, 2, where the SM Higgs particle H has tobe substituted by the corresponding MSSM Higgs boson & [42,44] ite retry (a? Pimp? ay Tle > @@l= = Ma(Mo Nes [1 + deco + 5°] (23) Neglecting regular quark mass effects, the QCD corrections Saco are presented in Eq. (15) and the top quark induced contributions read as 44] an AE, (etn) [ar Bo a tog Mata) n +e Me, if ( 256Mn) Mi Voge MalMa) E (2th) [nant ie Analogous to the SM case the large logarithmic contributions of the QCD corrections are absorbed in the running MS quark mass ig(Mg at the scale of the corresponding Higgs mass Mo. The quark decay width ofthe charged Higgs boson reads. inthe large Higgs mass regime My >> My + Mp, as 52.53) Grice avi (Eq, (24) is valid if either the first or the second term is dominant.) where U(D) denote heavy up(down)-type quarks. The Felative couplings g@ have been collected in Table | and the coefficient Vip denotes the CKM matrix element ofthe transition of D to U quarks. The QCD correction factor Saco is given in Eq, (15), where large logarithmic terms are again absorbed in the running MI masses My, o(Miy+) atthe scale of the charged Higgs mass My. Inthe threshold regions mass effects play a significant role. The partial decay widths of the neutral Higgs bosons @ = h, Hand A into heavy quark pairs, in terms ofthe ‘quark pole mass Ma, can be cast into the form [42] * rtH* > UB] == WMuol® (Mur Meee)? + Ap ( Mir eg) ] (1 + Baco) (2a) 0,2 yg? BP 4 Ga a 6 peg *] 25) where 6 = (1 — 4M3/M3)¥ denotes the velocity of the final-state quarks and p = 3(1) the exponent for scalar (pseudoscalar) Higgs bosons. The different powers in fare related to the property that the QQ pairs are produced in a ‘P-wave (S-wave) for (pseudo)scalar Higgs decays. To next-to-leading order, the QCD correction factor is given by Eq. (17) for the scalar Higgs particles h, H, while for the CP-odd Higgs boson A it reads as [42] I= EM) + eg lI9-+ 26° + 30% 108+ a z ‘with the function A(A) defined in Eg. (18). The QCD corrections in the threshold region are moderate, apart from a Coulomb singularity, which is regularized by taking into account the finite top quark decay width, 7B) (26) ¥ the following we denote the diferent types of neural Higes parties collectively by @ = h,H.A,105 1M Spta Progress Pace and Nuclear Physics 95 (2017) 9 The partial decay width of the charged Higgs particles into heavy quarks can be expressed as [53] runt > ub) = ee est a? {ea Fac = EE ea a0? {01 = ny = wo) | a = Mate ( +3 #%)| —4MyMp/ ao (0 + $t)} Qn where jz) = M?/M2,. and 4 = (1 — sty — us)? — 4uypip denotes the usual two-body phase-space function. The quark ‘masses Mp are the pole masses, The QCD factors S¥ (i,j = U, D)are given by [53] B= 2m +20) wy, B= mi —Hh+ Sy log + log xy + a aT) Mya EA = = Hy) oF 34 BEB tog Hy AEPO AH) fog + By 28 yy a34 ee ait Bx + (28) with the scaling variables x = 2y4,/[1 ~ ji — jj + 44] and the generic function [Ati(aay) —2Lig(—x) = 2Lig(—9) + 210g log(1 — x95) — logx;log{1 +) — log log( )=+[loet ay) + tors Tm SOT [lost = See Dog tad torn | + aye] 48 ‘ ‘The transition from the threshold region, involving mass effects, tothe renormalization-group-improved large Higgs mass regime is provided by a smooth linear interpolation analogous to the SM case in all heavy quark decay modes in the programme Hpecay [54]. ‘The full SUSY-electroweak and SUSY-QCD corrections to the fermionic decay modes have been computed [36,37,55]. ‘They turn out to be moderate for small values of tgp, of the order of about 10%. Only for large values of tgs > 10 do the luino corrections induce large corrections due to the A, terms of dovn-type fermions as discussed in the inttaduction. ‘These effects are in particular relevant for the MSSM Higgs boson decays into bottom quarks. Te final NLO decay width into bottom quarks can be written as 3GrMe — eo 1a? 4 eaten aig ele [1 + Seco + 8?) [2? + 22 5800] (29) ‘where the effective couplings 2 are given in Eq. (12) and thus resum the leading 4 terms. The contribution 82, denotes the remainder of the full SUSY-QCD corrections after absorbing the A, terms in the effective Yukawa couplings. This remainder turns out to be small fr all MSSM Higgs bosons even for large Higgs masses [37] Below the ¢ threshold, heavy neutral Higgs boson decays into offshell top quarks are sizeable, thus modifying the profile of these Higgs particles significantly inthis region. The dominant below-threshold contributions can be obtained from the ‘SM expression forthe scalar Higgs bosons [55] ro — BB) ar . me (30) with the reduced energies x,2 = 2E:»/My (@ = h, H. A), the scaling variables yy = 1 — x2, x; = M2/MZ and the reduced decay widths of the virtual particles y, = 17? /M3. The squared amplitudes can be written as [56] 18 = AC 91 = 92 bw ~ 5h) + ewe — Kw — keds + Ae) nya + (1 — Au )2y1 + Hw + He) 18 = VACA = ys — ya bevy — 4) + Qaw(vaya — Kae) = el¥aa = Dye — KW =) 6) ‘The differential decays width of Eq. (30) has to be integrated over the xy, x, region, bounded by ACT — xy — Xp + ke + ee — kw) + X1XD (32) Vt — 4ryad — dey In these formulae W- and charged-Higgs-boson exchange contributions are neglected, because they are suppressed with respect to the off-shell top quatk contribution to Web final states. However, forthe sake of completeness they are included in Hecay [54]. Their explicit expressions can be found in [56]. The transition from below to above the threshold is provided1M Spta Progress in Paice and Nuclear Phys 95 (2017) 98-159, 107 waren “a lamz Fig 3. Diagram contributing toH > W [V = W.2), by a smooth cubic interpolation. Below-threshold decays yield a t@ branching ratio at the per-cent level already for heavy scalar and pseudoscalar Higgs masses My, ~ 300 GeV. a Below the tb threshold off-shell charged-Higgs decays H* —> ¢"b -> bbW* are important. For Mye < Mi + My — where F; denotes the top width, their expression can be cast into the form [56] 3ciMs wi ee (kyr + 3k — kw) on + (8d — 4 — 3p + log AXES 2 (54 — aya — Daca + 442) + A (+-3)} (3) (+ ch wb) = swe) 1 + with the scaling variables xi = wan (i = £,W), The bottom mass has been neglected in Eq, (33), but it is taken into ‘account in HDECAY by performing a nuimerical integration of the corresponding Dalitz plot density, given in [56]. The off- shell branching ratio can reach the per-cent level for charged Higgs masses above about 100 GeV for small 8, which is significantly below the tb threshold My+ ~ 180 GeV. 2.2, Higgs boson decays into intermediate gauge bosons 22.1, Standard model ‘Above the WW and ZZ decay thresholds, the partial decay widths into on-shell paits of massive gauge bosons (V = W,Z) at lowest order (see Fig. 3) are given by [7|* GoM “evar with the abbreviations x = M3/M3, 6 = /T—A% and fy = 2(1) for V = W (2). ‘The NLO electroweak corrections are small and amount to less than about 5% in the intermediate mass range [45,57]. Furthermore the QCD corrections tothe leading top mass corrections of O{GeMz) have been calculated up to NLO. At NNLO they rescale the WW, ZZ decay widths by [58,59] [5-15 ~22)%]} G5) PU > ww) = Flt -» ww) {1 —x [5 ~(9 -233)]} (8) row) pO 4x4 122) Ga) TH > 22) = Figltt > ZZ) f1— with x: = Gri? (83:2). The three-loop corrections can be found in [50] Below threshold the decays into off-shell gauge particles are important [61,62]. For Higgs masses slightly larger than the cortesponding gauge boson mass the decay widths into paits of off-shell gauge bosons can be ast into the form [52] 1 2? ee dQzMy Ty -eP dQeMy Ty (= MyP + Mrye (G=uiy sur with Q2, Q2 being the squared invariant masses ofthe virtual gauge bosons, My and I their masses and total decay widths Tois given by rH vv) = G7) 1p = 6 Ge Mp [pa? imp +1 (38) = by EE VO. Fm [0003 0F: withthe phase-space factor do yiz) = (1- G9) “eventhough the SM Higgs mass isto small these expressions playa role for ofshell Higgs bosons sping into on-shell W- er 2-boson pis108 1M Spta/ Progress Paice and Nuclear Physics 95 (2017) 9 The branching ratios of double off-shell decays reach the per cent level for Higgs masses above about 100 (110) GeV for off-shell W(Z) boson pairs. Off-shell effects modify the partial decay widths by (10%) above the WW and ZZ thresholds. During the last years the complete electroweak corrections have been obtained for the full Higgs boson decays into four {fermions that include the resonant and non-resonant contributions to Higgs boson decays into intermediate gauge bosons W, Z [63] In Eq, (38) they can be approximated by the improved Born approximation [53] reo rixne[r4 Bie Bin 6 2 eGov 2h Ho ww > af [s+ (F)] Tyo Myx weft 4 3+ aww (TE | # GeME (Sx? eM (82 _avin 3) enon (Si) wa ( 6 eV Fat Been (Mi, OF, 2) + 99 on Oat com} (40) with the electric charges ey of the fermions and the fermionic couplings to the Z boson sin ___sindy by Fin cara Baty ae a {and the auxiliary functions Mi 2 2 2 tmz (4) = 20-+ 68? + 36((8? + I)logx + 3(1 — 6; log? x, Mi 3 ‘tow (4) = 84 1282 + 36362 — Dlogx + 3(1— Flog? x (22) where Bot 43) Bat (3) ‘and jx; denotes the complex top mass, given by p:? = m2 — im, J; with the top width F;. The factor bay describes the Coulomb singularity, 2 gry _ #0) Seva, QF. Q2) = am B (44) ‘where a(0) denotes the fine-structure constantiin the Thomson limit and the complex W mass s.2, = M3, —iMw Tw is given in terms of the W pole mass Miy and the W decay width Tiy. The function g() reads B= (1-PY (45) Finally the QCD correction factors to the gauge-boson decays are given by ye Hyak “1M Spita/ Progress in Pace and Nuclear Physics 95 (2017) 98-159, 108 n2228 9 ig 4 Diagrams contributing 0H —> gg atlowest order. and the continuum contributions can be fitted as ome = 3% cin = 4%. (a7) They constitute auxiliary corrections beyond the previous terms to compensate the remaining offset between the exact, cortections and the improved Born approximation [63]. In general there are additional non-resonant LO contributions beyond those of Eq, (37), However, they turn out to range at the per-cent level [63]. Recently the full electroweak NLO ‘expressions to the leptonic Higgs decays H > 4¢ have been matched to parton showers [54], indicating that there are a few exclusive observables where multiple photon effects may be relevant. The theoretical uncertainties due to missing corrections beyond NLO have been estimated as 0.5% in the intermediate mass range for the inclusive decay rate into four Termions [65,55 222, Minimal supersymmenic extension The dominant part ofthe partial widths of the scalar MSSM Higgs bosons into Wand Z boson pais can be obtained from the SM Higgs decay widths by escaling with the corresponding MSSM couplings gh" as listed in Table 1 Unf > VOvirly = (gy, (Hg > VOY, (48) Their branching ratios are strongly reduced by reduced MSSM couplings and additional open leading decay modes, and thus do not play a dominant role asin the SM case. Nevertheless the WW, ZZ branching ratios can reach values of (10%) for the heavy scalar Higgs boson H for small tgB.Off-shell WW, 22 decays can pick up several per-cent ofthe ight scalar Higgs decays at the upper end of its mass range. The pseudoscalar Higgs particle does not couple to Wand Z. bosons at tree level. The SUSY-electroweak corrections to scalar Higgs boson decays into off-shell WW and ZZ paits have been computed [57], While being similar to the SMHiggs case they can be enhanced to the 10%-level in particular MSSM 2.3, Higgs boson decays into gluons 23.1, Standard model The decay of the Higgs boson into gluons is mediated by heavy quark loops in the SM, with top quarks providing the dominant contribution, see Fig. 4. The partial decay width [40,68-70] is at lowest order given by - _ Geo Mi | P Tool > a8) = DA (49) with the form factor Ak) = Set Fann) fel=}, (50) [lee The parameter rq = 4M3/M} is defined by the pole mass Mg of the heavy loop quark @. For large quark masses the form factor approaches unity. The QCD corrections are known up to NLO including the full quark mass dependence [68] and up_no IM Spta Progress Pace and Nuclear Physics 95 (2017) 9 to N°LO in the limit of heavy top quarks [70-72], For a Higgs mass My ~ 125 GeV they can be expressed as (My) Pobeel= ne [r+0r7s 1360 +4 + (7020 738m 0s0ny +2375 +0653 ME) (5) 4 (153348 — 1062 82Np 4 52.62N} — 0.S378N? 4 (66.66 + 14.60N = 06887?) og Mi + (6.53 + 1.44Ns ~ 0.1118 on? Mt) (2) | % Tig (1-40.67 +0.20 + 0.02) 61 for Np = 5 light quark flavours, The term Ay *~ 0.024 denotes the NLO quark-mass effects from the top, bottom and charm quarks [68], The radiative corrections turn out to be very large: the decay width is increased by about 90% in the intermediate mass range. The approximation of the heavy top limit is valid for the partial gluonic decay width within about 510% for the whole Higgs mass range up to 1 TeV, while itis valid at the per-cent level in the intermediate mass range. The reason for the suppressed quark mass dependence of the relative QCD corrections is the dominance of soft and collinear ¢luon contributions, which do nat resolve the Higgs coupling to gluons and thus lead to a simple rescaling factor. The three- loop [71] and four-loop [72] QCD corrections to the gluonic decay width have been evaluated in the limit ofa heavy top quark, They contribute a further amount of O{20%) relative to the lowest order result and thus increase the full NLO expression by (10%), The reduced size of these corrections signals a significant stabilization of the perturbative result and thus a reliable theoretical prediction. The QCD corrections in the heavy quark limit can also be obtained by means ofa low-energy theorem [40.73]. The starting point is that, for vanishing Higgs momentum py > 0, the entire interaction of the Higgs particle with WZ bosons and fermions can be generated by the substitution M > M,x [+ (@=f.w,2) (652) where the Higgs field H acts as a constant complex number. At higher orders this substitution has to be expressed in terms of bare or MS. subtracted parameters [58,68]. In most ofthe practical cases the extemal Higgs particles are treated on-shell so that pi, = Mj; and the mathematical limit of vanishing Higgs momentum coincides with the limit of small Higgs masses ‘or large loop-particle masses. In order to calculate the effective Higgs couplings to gluons in the heavy-quark limit one starts from the heavy quark Q contribution tothe kinetic gluon operator [74] that is related to the decoupling relation of the strong coupling constant [75.76 1 May, an OB CH) (53) a= (ela) aug) ‘where Nr = 5 is the numberof light quark flavours. The gluonic field strength tensor Gis defined in the (Nr + 1)-flavour theory, while the superscript ofthe strong coupling a, defines the numberof active flavours taken into account. The inverse ofthe decoupling factor, develops a perturbative expansion that contains a logarithmic top mass dependence, i. itcanbe expressed in terms of the logarithm L; = log(s.3/m"(a3)), where M’(y.3) denotes the MS top mass at the renormalization scale jug. The substitution of Eq, (52) implies the shift = Blog (1 ' *) 4) ‘Keeping only Higgs-field dependent terms and transforming the gluonic field strength operator and the strong coupling ‘constant by their N;-flavour expressions one arrives at the effective Lagrangian in the heavy top-quark limit [74], eg = Sef ooe(14 2) toe (148) 4 Soe (144) - Set ( 65)IM Spta/ Progress in Pace and Nuclear Physics 95 (2017) 98-159, with the QCD corrections up to N*LO s=nSrn(®) +5 (2) +52)’ +00 n= m(Z) +n (2) +n) ere p= (2) +n(Z) +o) a= 0(%) +0103), (68) The perturbative coefficients are explicitly given by u 27 19 67 at = Ban (£-2 4 2e5 ' 16 3 96 5, — 897943, 2892659 , 209, 1733 * 9216 "41472 64° 288° gn, (4291 _ MOTT, 23, 55)) a (_H 77 6865 * \ 20736 Ts24 32°" Sat * 1728 * 31104, 12 1 11679301 84 = — eh nilog’2 + SP hiogs 2 + 85 nptogt2 + HOPPE loge 780 aan 1741824 95970579, tat 3751 sy 585 ya — TE gta 4 Et catog'2 — Ph eatog'2 — SS npestog'2 Broa 98 + qggNrtalos a s sora 08 97S3ON gg 9 92970579 gy hyesloe? — 290304 irtalor Tasis2 210! aes 8 45 885 yp, 4 T8301, 93970579 “D1, 3751 ine /™ 7576 sees DS a 270407 yp , 4081305 yp, S7BTST yay 11S y,, 480732 ~ rasaaea * Sanrre S'S ~ Ssz9g 4 S*~ 3945 — 7605"? 151369, 12171659669, 608462731, 313489, 735760 °° se707200 + Trerat60 PS * “gory NSS 76094378783,, 4692439 | 28121193841, | 4674213853, _ 807193, 522547200. | 145 19353600 °° 2903040” T78 * 854201072999 | ( 481, | 28257 ya, 21138), — + (tn nes - ue — ness szasa7z00* S184" * oxi 3as6 ~ a8 5160073, 9364157, 49087545) | | (__77_ yy _ 1267 yp, 41394, aia 12288 % 55296 eoi2' ~ 73824." * 2304 ° SOT) 2, (Ags 157.2, 275, | 2299) 2+ (one Sint 4 n+ SS) a 67) cy 05"? ~ 576°" + To" * 356 and 352 m=342N " m3 333» 58s = BB yy (8 BEL) on (Z 2 az | 4 car | asa ~ a! 481 9), ga (28297, _ 373637 429965 2985893 ne = Bhp np (B28 y, — 373887) 5 y, ( ZOR8E _ 2085883, 3592 08° ~ “31104 1728 ~ 13824 26296585, 143976701 | (77 yy M421yp , 90735, 45058), Tessa ~~ ap94g T728f ~ 3456" Ne 728 | 576) * (M4552 5 By, 4 649) 2 as a8 * 5 * oe 1697 | 175,22 m= Tag + 73" of = Bitaon 16259," g, (9899. 2077) ya (175 4 39). yy (le 7 me 1728 192°” \ 1728 * 144 384° 144° "V9 ~ 1728 51383 317,47 2 es By Aes Zp 864 36 24 27 8)m 1M Spta/ Progress Pace and Nuclear Physics 95 (2017) 9 by H An >= ~ tb 9 Fig. 6. Typical diagrams contributing to @ > gx at lowest onder. where Gi, denotes the gluon field strength tensor and a, the strong coupling constant with Nr = 5 active flavours. The constants used in these expressions are defined as a, = Li,(1/2)(n = 4, 5) and [77] rea Be. sthe, + Sog'2—32 te 2318 ao? 8088795462 (69) This effective Lagrangian describes the multi-Higgs couplings to gluons after integrating out the heavy top quark For the full calculation ofthe heavy quark limit given in Eq (51) the effective Hge() couplings have to be inserted into the blobs ofthe effective diagrams shown in Fig 5 for the NLO contribution. After evaluating these effective massless one-loop contributions the result coincides with the explicit calculation ofthe two-loop corrections in the heavy quark limit of Eq. (51) This calculation has been extended to N2LO analogously [71.72 Using the discussed low-energy theorem, the electroweak corrections of (GM?) to the gluonic decay width, which are mediated by virtual top quarks, can be obtained from the ©(GrM?) contribution tothe kinetic gluon operator. The final result leads to a simple rescaling of the lowest order decay width [49,78] ) ‘ese corrections enhance the gluonic decay width by about 0.3% and ate thus negligible. These O(G-M2) corrections are part ofthe fll electroweak corrections that have been determined fist analytically forthe partial light-fermion contributions [79] and finally numerically forthe ful electroweak corrections [80], The electroweak correctionsincrease the partial Higgs decay ‘width into gluons by about 5%. The residual theoretical uncertainties due to missing electroweak corrections beyond NLO hhave been estimated as 1% and due to missing higher-order QCD corrections as 3% 50 thatthe total uncertainty ranges at about 3% [51,65]. TH gx) = Tt [14 (60) 23.2, Minimal supersymmetric extension Since the h-quark couplings to the Higgs bosons may be strongly enhanced for large tg and the t-quark couplings suppressed in the MSSM, b loops can contribute significantly to the gg coupling so that the approximation MZ > MB can. in general no longer be applied. The leading order width for h, H — gg is generated by quark and squark loops. the latter contributing significantly for squark masses below about 400 GeV [81]. The contributing diagrams are depicted in Fig. 6. The partial decay widths are given by [68.81.82] Gre M |e te gh hgh Tolh/H > ge) = z ‘AN (xa) + etal ag 61 Hl -> ee) = To oe Mra) + Daag) (61) bite) 3 A Alin) = 50+ rE) AMM) = (62) with ¢; = 4M2/MBi4 (@-= @, @). The function f(r) is defined in Eq. (50) and the MSSM couplings gi! can be found in Table 1. The squark couplings a2 are summarized in Table 2. The amplitudes approach constant valves inthe limit oflarge1M Spta Progress in Pace and Nuclear Physics 95 (2017) 98-159, mm Table 2 MSSM Higgs couplings to charge Higgs bosons, chayginos related tothe mixing angles betweer the c to their fermionic superpareners, sfermonstelative to the corresponding SM Yekava couplings. O, andy (1 = 1,2) ate the sfermion sates define te weak charges according ° = @ iu 3H 7 3 ° Mss nia — oh Bes Bul — einen Saf x Bel — este reose+ 8) cog a) — eats] ° loop particle masses: Aa) > 1 FOr Min AMG AM) eb forty «403 The QCD corrections to the quark loops can be adopted from the SM Higgs case as given in Eq, (51) with the Higgs mass ‘My replaced by the corresponding MSSM Higgs mass My. The QCD corrections to the squark loops are known up to NLO including the Full mass dependence [83]. Up to NLO the partial decay width is given by PUn/H > gg) = Tio { (63) 95 , r-% for Min = AQ(r) ME I ay so that the quark contributions to the partial decay widths of the scalar and pseudoscalar Higgs bosons approach each ‘other, ie. the chiral symmetry is restored [the relative factor 3/2 is compensated by the different global coefficients of the partial decay widths of Eqs. (61), (55)]. The same is also true for the NLO QCD corrections (58). The finite parts of the virtual Cortections approach common logarithmic expressions, for M3 > M3 (72 with ¢, = 3 ifthe quark mass is defined to be the pole mass. The Abelian part of the large logarithms is related to the Sudakoy form factor [97] atthe virtual @b vertex and have been resummed down to the subleading logarithmic level [8]. The non-Abelian part of these logarithms, however, has not been resummed so far. The ral corrections determined by the coeflicient £2, are of subleading non-Abelian logarithmic order. The genuine SUSY-QCD corrections have been calculated in the limit of heavy SUSY-particle masses by means ofa heavy ‘mass expansion for the top-stop-sluino contributions [98-102] and a mixed heavy mass and large momentum expansion © This hasbeen confirmed explicit up to NLD in Ref [2]. the QCD corrections ate expanded in terms ofthe strong coupling with oly ive light active favours,1M Spta Progress n Pace and Nuclear Phys 95 (2017) 98-159, ns 1 teof Care @e>29 Preinsay 100F teB 0 oF ne cof = saponin * a My IGeV} Fig.7. The genuine SUSY-QCD corrections normalized to the LO bottom quark form factor. Real pat: red, imaginary part: ue. compareé to the Ay approximation (dashed lines). Ay has bees renormalized inthe MS scheme. (Fr interpretation ofthe references to colour in this igure legen, the reader {refered to che web version of thi rile) Source from Ret. 1105}, Fig. Typical diagrams contributing tof -» yy at lowest order. for the bottom-sbottom-gluino corrections [101-103]. A large part of the latter corrections in the bottom case can be absorbed by the effective bottom Yukawa couplings of Eq. (12). However, there is a sizeable remainder. This feature is confirmed by the full calculation of the SUSY-QCD corrections by means of a numerical integration of the corresponding two-loop diagrams [104,105]. The agreement with the approximate calculations is reasonable. The numerical results of the full calculation are presented in Fig. 7 for the modified small ey scenatio [106]. defined by the following choices of MSSM parameters [m, = 172.6 GeV}, Mg = 800GeV tmp = 30 M; = 1000GV 0p = 2 TW 73) M; = 500GeV 0 y= = —1.133 TeV Inthis scenario the squark masses amount to m, = 679GV —m, = 93560 rm, = 601GV —m;, = 961 GeV. 74) ‘ig. 7 display the genuine SUSY-QCD corrections normalized to the LO bottom quark form factor. ie ANC) > ANCE Chg $4), where the LO form factor has been defined in terms ofthe bottom Yukawa coupling g"" without ay-testmmation, The corrections canbe sizeable, but can be described reasonably well with the usual 4y approximation fy is renormalized in the MS scheme, however, with a sizeable remainder that determines the differences between the full and dashed curves of Fig. 7. 2.4, Higgs boson decays into photons 24.1. Standard model The decay of the Higgs boson into photons is mediated by W and heavy fermion loops in the SM, see Fig. 8; at lowest, order the partial decay width is given by [73] Goat MR iH yyl= So Ne eFAM (ay) + At Caw) (75) 128 V2?ns IM Spta Progress Pace and Nuclear Physics 95 (2017) 9 with the form factors ANC) = 2x [14 (1 = rfle)) Aye) = —[2+ 30 4312 — F(e)) and the function f(x) defined in Eq. (50). The parameters *, = 4M?/M} (i =, W) are defined in terms of the corresponding ‘of the heavy loop-particle masses, Fr large loop masses the form factors approach constant values Ho 4 2 (2 Nog for Mj, < 4M A> 7 for « 4M, (78) ‘The W loop is dominant inthe intermediate Higgs mass range, and the fermion loops interfere destructively. Only far above the thresholds for Higgs masses M, ~ 600 GeV, does the top quark loop become competitive, nearly cancelling the W loop contribution, This feature is still relevant for offshell Higgs contributions In the past the full two-loop QCD corrections to the quark loops have been calculated [8,107,108], The QCD corrections simply rescale the lowest order quark amplitude by a factor that only depends on the ratios of the Higgs and quark masses Ailsa) > Alea) x [1+ Calta) =] Gilt) > =1 for ME «ag om ‘According to the low-energy theorem discussed before. the QCD corrections in the heavy quark limit can be obtained from the effective Lagrangian [40,58,58,73] die (14 2) Bee (1+ 8) ] {o + dytoe (1 + “)- ‘with the QCD corrections up to NPLO n= m(2) +m(2)' +0008 p= a(S) +0108 79) ‘The perturbative coefficients are explicitly given by son od not b= 8 =-5-Sk+M(2-—)+ Ny 37 3 nis 5, = 2835,, 95339 961, 541 4693" 125) 312, 101 * = yep ~ 582 ~ Tag ~ Tos” + (aaa6 ~ 14a * 36 * 26" 54 324 ets) 2(H kb _ 49), dine ee om(-3+ + nw (Se +3)] _ 541101 NB (961 Tos ~ 216%" ~ 34 7 yh e (80) (where F,. denotes the photon field strength tensor and a, the strong coupling constant with Nr = 5 active flavours) that ‘can be derived from the top-quark contribution to the kinetic photon operator analogous to the QCD case. The results for 5, and 5; agree with previous determinations [75,109]. Partial results are also known up to N‘L0 [110] In order to improve the perturbative behaviour of the quark loop contributions they should be expressed preferably in terms of the running quark masses mg(Mj:/2), which are normalized to the pole masses Mg Via malHe Mg (1)1M Spta/ Progress in Paice and Nuclear Phys 95 (2017) 98-159. 7 AHA Fig 9. Typical diagrams contributing to 6 > yy atlowestorder with their scale identified with seq = M/2-within the photonic decay mode. These definitions imply a proper definition ofthe Q0 thresholds Mj, — 2Mq, without artificial displacements due to finite shifts between the pole and running quark ‘masses, as is the case for the running MIS masses, The residual QCD corrections are sma in the intermediate Higgs-mass range inereasing the photonic Higgs decay width by about 2X. Recently the full three-loop QCD corrections to the photonic Higgs decay have been calculated in the heavy top-quark limit [1 11]. They lead to a further contribution of a few per mille. The two-loop electroweak corrections have been evaluated in Ref. [112] for the W-boson and top-quark induced contributions. They decrease the partial photonic decay width by about 2% thus nearly cancelling the NLO QCD corrections. Both types of corrections only become large for Higgs masses above the tT-threshold, te or My > 2m, 2.42, Minimal supersymmetric extension ‘The decays of the scalar Higgs bosons into photons are mediated by Wand heavy fermion loops as in the SM and, in addition, by charged Higgs, sfermion and chargino loops, see Fig. 9. The partial decay widths [58,82] can be expressed as Gra M) 1eVviP Pth/H > yy) = Deepa) + Ay" ew) T vee AC) (82) + alaitoned + Dea D involving the form factors AMM (r) = Be [+ (1— rY Ce) ALG) = =t l= fC] AWM(r) = [2 +30 +3202 — OF(TI] and the function (x) as defined in Eq, (50) For large loop-particle masses the form factors approach constant values, Woy 4 4 > > ANC) > forty < aM, a i 2 2 Ae > 5 for Miva < AM ja 7 ANH) 27 for My < aM, ‘fermion loops start to be sizeable for sfermion masses My 300 GeV, while for larger sfermion masses they are negligible in general “The photonic decay mode of the pseudoscalar Higgs boson is generated by heavy charged-fermion and chargino loops, see Fig, 9, The partial decay width reads [68,82] Gros 2ehal A at PA yy) = SAIS Nachapatlen + DoateAs Ce) (83) with the amplitudes f(t) 34 For large loop-particle masses the pseudoscalar amplitudes approach unity. The parameters r, = 4M?/M3 (i = J,W,H®, 7, fate defined in terms of the corresponding heavy loop-particle masses and the MSSM couplings £°,, yj are summarized in Tables 1 and 2. ° The QCD corrections to the quark and squark loop contributions have been calculated. They are known for finite quark, squark and Higgs masses [68,83,107,108]. The QCD corrections rescale the lowest order quark amplitudes [68,83, 107,108,113], A$) + ARC0) [1 + Cole) ] (85)us 1M Sota Progress in Paice and Nuclear Physics 95 (2017) 9 Gyyulta) > - for Mijn < AMG Car) > 0 for M3 < AME Ng > AMQ[L + Blea] (65) Gynlrg) +3 for My «aM ‘The OCD corrections to the y'y decay width are defined in terms of the running quark masses in the same way as the SM photonic decay width. The QCD radiative corrections are moderate in the intermediate mass range [68,107,108], where this decay mode will be important. Owing to the narrow-width approximation of the virtual quarks, the QCD corrections to the pseudoscalar (scalar) decay width exhibit a Coulomb singularity at the ¢7 (QQ) threshold, which is regularized by taking into account the finite top quark and squarks decay widths [91]. The QCD corrections to the quark loops in the heavy quark limit can be obtained by means of the low-energy theorems for scalar as well as pseudoscalar Higgs particles, which have been discussed before. The result forthe scalar Higgs bosons ‘agrees with the SM result of Eq. (77), and the QCD corrections tothe pseudoscalar decay mode vanish inthis limit due to the ‘Adler-Bardeen theorem [95]. In complete analogy to the gluonic decay mode, the effective Lagrangian can be derived from the AB) anomaly [93] and is given to all orders of perturbation theory by [53] age 30 beg = RT Since there are no effective diagrams generated by light particle interactions that contribute to the photonic decay width at next-to-leading order, the QCD corrections to the pseudoscalar decay width vanish, in agreement with the explicit expansion ‘of the massive two-loop result. Starting at NNLO there are light quark-loop induced contributions involving the effective ‘Agg-vertexin the heavy-quark limit ‘Completely analogous the QCD corrections to the squark loops forthe scalar Higgs particles inthe heavy squark limi can be obtained by the extension of the scalar low-energy theorem to the scalar squarks. Their coupling to photons at NLO can be described by the effective Lagrangian [113]’ (87) He2 % pup ca bg = GEE [1432] (88) where only non-SUSY-QCD corrections are taken into account ie. diagrams with guino exchange or the squark four-point “couplings have been omitted. The genuine SUSY-QCD corrections can be deduced from the analogous gluon-gluon case as described and mentioned in Refs. [100,102,103]. Their numerical impact, however, has not been analysed so far. The electroweak corrections tothe pseudoscalar Higgs decays into photons have been calculated in the limit of heavy top quarks and W bosons some time ago within the 2HDM, ie. involving the non-SUSY part ofthe MSSM. The corrections ‘modify the partial decay wndth by [95] noxfine(s+ea)} ) with x; defined after Eq. (36). The correction ranges at the per-cent level For large values of tg6 the bottom-quark loops will take over the dominant role. In the limit of small quark masses the form factors develop logarithmic structures, TA yy) aos (90) so that the quark contributions to the partial decay widths ofthe scalar and pseudoscalar Higgs bosons approach each other, ice the chiral symmetry is restored {the relative factor 2 is compensated by the different global coefficients of the partial decay widths of Eqs. (82), (83)], The same is also true for the NLO QCD corrections, Cyltq) + sor : « -3[eif-«] dell for ane wo ‘where j4y denotes the renormalization scale of the running quark mass. The large logarithms are related to the Sudakov form factor [97] at the virtual bb vertex and have been resummed down to the subleading logarithmic level [93] 7 a to account the squark four-point couplings for degenerate squark masses, tc, the coefficient 3s cha1M Spta/ Progress in Pace and Nuclear Phys 95 (2017) 98-159, ne pw Z gi Z $ f He-- a H- Fig. 10, Types diagrams contributing to —+ Zy at lowest order, ws Z Nw ting te che QCD corrections to H — Zy. 25, Higgs boson decays into photon and Z boson and Dalitz decays 25.1, Standard model ‘On-shell z boson. The decay of the Higgs boson into a photon and aZ boson proceeds via W and heavy fermion loops, see Fig. 10. The partial decay width can be obtained at leading order as [82,114] TlH Zy) Ap) + Aivltw. dw) M2, M2 ee (92) Bax with the W and fermion form factors ag E28,” OH 6,3 1,23] Aiea) a) ‘cos Ow Aye.) = cosdy {13 — tan ayyts » +[(1+ 2) tan oy - (5+2) Jute of (93) The functions f. 12 are defined as vet é ae NOS AD = SG + ae eap UNF + lato) — 201 be) = 5 oF where the function g(r) is given by. Tarcsin: 1 r21 ve = a= Maas (94) a Eon] vt and f(r) is defined in Eq, (50). The parameters r, = 4M?/Mi and 2, = 4M2/M2 (i = f,W) are defined in terms of the corresponding masses of the heavy loop particles. Due to charge conjugation invariance, only the vectorial Z coupling ‘contributes to the fermion loop so that problems with the axial ys coupling do not arise, The W loop provides the dominant contribution in the intermediate Higgs mass range, and the heavy fermion loops interfere destructively The two-loop QCD corrections to the top quark loops have been calculated numerically [115] and analytically very recently [116]. They ate generated by virtual ghton exchange inside the quark triangle (see Fig. 11).Due to charge conjugation invariance and colour conservation, radiation of a single gluon is not possible. Hence the QCD corrections can simply be expressed as a rescaling of the lowest order amplitude by a factor that only depends on the ratios x, and a, (i = J, W)) defined above: A(x, 20) > Alita, da) x [1+ Datta ia] Dultqsdg) > —1 for < ME « aM (95)0 IM Spta/ Progress Paice and Nuclear Physics 95 (2017) 9 vt an Fig. 12, Generic diagrams contributing tothe Dalit decays H > yf In the limit My > 0 the quark amplitude approaches the corresponding form factor of the photonic decay mode (modulo couplings), which has been discussed before. Hence the QCD correction in the heavy quark limit for small Z masses has to coincide with the heavy quark limit of the photonic decay mode of Eq, (77). The QCD corrections for finite Higgs, Z and quark masses are presented in [115,116] as a function of the Higgs mass. They amount to less than 0.3% in the intermediate mass range, where this decay mode is relevant, and can thus be neglected, More important is the inclusion of off-shell effects of the final-state Z boson due to its finite width. This requires the analysis of Dalitz decays H —> yfJ. Dalitz decays. Dalit decays of the Higgs boson [117.118] are described by the diagrams of Fig. 12. The first diagram corresponds to the direct contributions, since these are mediated by the Yukawa coupling of the fermion f. The second line shows the triangle diagrams which include the resonant H —> yZ —> yff contribution as well as photon conversion H > yy" — yff-The leading contributions to the triangles emerge from W, top and bottorn loops. The last line exhibits box contributions and final-state photon radiation which, however, are numerically suppressed compared to the triangle diagrams, but are required for a gauge-invariant result Fig. 13 shows the invariant mass distributions for e*e~. jo y~ and rr pais including the individual contributions of the different types of diagrams normalized to the partial decay width into photons. A lower cut of E, > 1 GeV has been imposed on the photon energy to avoid the infrared singularity of photon radiation, For small invariant masses the photon conversion H —> yy* > y4*2~ provides the dominant contribution, while for invariant masses around tie Z-boson mass the Z-boson contribution H — yz" —> yff takes over the dominant role. At the end-point q? = Mj of the spectrum the yr~ (where it dominates in the whole q?-range). For a clean separation of the H -> yy,H —> yy* —» yt*e~,H —> Zy and H — €°€ contributions appropriate cuts have tobe implemented forthe Dalitz decays. The low-g? parthas to be attributed toll > yy. the q-part around M to H -» Zy and the end-point region close to M3 to the QED corrections toH > &*¢- 252. Minimal supersymmetric extension On-shelz boson. The decays of the salar Higgs bosons into aZ boson and photon are mediated by W and heavy fermion loops asin the Standard Model and in addition, by charged Higgs. sfermion and chargino loops as shown in Fig. 14. The partial decay widths are given by [68,119] GEM Mn (, MEY’) ur gny eee at) Eaaitovnn A ord) + Ay F[h/H + zy) (98)1M Spta Progress Particle and Nuclear Phys 95 (2017) 98-159, m 001 Ho yeten 108 107 109) Bea eae Ve (eV) Hote soo} S ‘ or iar T ag? cot Harty 108 107 “| ‘ on eG i a a a cl Fig. 13. The invariant mass dstibutonsin YF = Mes ex ofthe Dalit decays H > y 4 ¢°e° jp /r*2" normalized to FH —> yy withactt Ey > 1 {GeV onthe photon energy. The red ins show he contribution ofthe tee diagrams the tin Sli lines denote te contribution ofthe photon conversion Hy yy" ye! (aed the dashes ine the contrition fom the 2 exchange diagrams, hil the thik nes present the total centibutions. The dotted lines denote the contribution fem the box diagrams (int Hoot-Feyaman gauge) Source: From Ref [18 ww Z anf AHAzo> ie 1 Wat FAH anny Nw Fig. 14. Typical diegrams contributing to @ > Zy at lowest oder vith he form factors 4", A! of Eq, (93), and cos26y cosy where the function I;(r, 4) is defined after Eq. (93). Alle y= 1(e9) (7)m 1M Spta/ Progress Paice and Nuclear Physics 95 (2017) 9 Fig. 15. Typical diagrams contributing to Higgs decays with Higgs bosons in the final state The Zy decay mode of the pseudoscalar Higgs boson is generated by heavy charged-fermion and chargino loops, see ig. 14. The partial decay width reads [119] 4 GEM (,_ MB)" gan m rao 27) = TE (1) Eaten ane 3 abate se (98) with the fermion amplitudes Abe, A) = Ny SLE Ber Sit? 0g), (99) cos Bw ‘The contributions of charginos and sfermions involve mixing terms, Their analytical expressions can be found in {119}. For large loop-particle masses and small Z-boson mass, the form factors approach the photonic amplitudes modulo couplings ‘The parameters t, = 4M?/M3, A; = 4M2/MZ (i=J.W.H", Z#,f)are defined by the corresponding heavy loop-particle masses and the nop-mixing MSSM couplings, . Af summatized in Tables 1 and 2, while the mixing and Z boson couplings g? can be found in [82]. The branching ratios of the Zy decay modes range ata level of up to a few 10~* in the intermediate mass ranges of the Higgs bosons and are thus phenomenologically less important in the MSSM. Dalitz decays hhave not been studied within the MSSM framework. They involve additional contributions from charged Higgs bosons, charginos and sfermions inthe loop contributions including the box diagrams 26, Supersymmetric Higgs boson decays into Higgs particles The heavy scalar Higgs boson can decay into paits of light scalar as well as pseudoscalar Higgs bosons, see Fig, 15, At LO the partial decay widths are given by [82] PU hy = 23g, OM (100) 6 VE My eM TH AA) = ay — (101) Bay EME fy aS ‘The self-couplings Ayan and 24s can be derived from the effective Higgs potential [33,34] The decay mode into pseudoscalar particles is restricted to small regions of the MSSM parameter space, where the pseudoscalar mass My is small. The decay into light scalar bosons is dominant for small tg6 below the £7 threshold. The SUSY-electroweak corrections to these decays have been calculated [120], They modify the partial decay widths by a moderate amount in general depending on the MSSM scenario if the radiatively improved trilinear Higgs couplings and masses are used. absorbing in this way the bulk of the cortections in terms of effective parameters. Only in parameter regions where the LO decay widths are suppressed the radiative corrections can be larger, but there these decay modes are phenomenologically irrelevant. The contributions of final states containing off-shell scalar or pseudoscalar Higgs bosons may be significant and are thus included in Hoecay [54] and FevnHtces [121]. Their expressions read as [56] 4 +) ma 2 me BEM y(o-2 PU 06°) = Nigel Soa {= (fin) 1 (102) Viggot ‘where xs = MZ/Mj. They slightly extend the regions, where the ht, AA decay modes of the heavy scalar Higgs boson I! are sizeable. Analogous to the decays into W and Z boson pairs double off-shell decays play arole in some regions of the MSSM. arctanIM Spr rors Paticeand Nuclear Phys 95 (2017) 98-159 m parameter space below the corresponding thresholds. Neglecting the final-state fermion masses they can be cast into the form [122] i 2 04? 2 1 2 _dQe Myre =f 4Q3 Mole py, (103) @- (@- (My + agry =MjY + ML Mare with 02, Q2 being the squared invariant masses ofthe virtual Higgs bosons, My and I their masses and total decay widths; Toiszivenby [122] rH $6") , Sp? Nias ME fac03 OF: MBE * Team, OF OM ag with the phase-space factor (x, ¥: 2) of Ea, (38) “Moreover, Higgs bosons can decay into a gauge anda Higgs boson, see Fig 15. The various partial widths can be expressed tT (104) 2 GM PH 2) = Hyg un (10s: ( eet ) PH HW) = gy PCa (106) Ban PA 2) = ay ME 220g. MB: My (107) ‘Bin i PC BW) Hg eM 322, (108) 8var where the couplings Aye can be determined from the effective Higgs potential [33,34]. The branching ratios of these decay modes may be sizeable in specific regions of the MSSM parameter space. The electroweak corrections within the 2HDM framework have been calculated [123], They are of moderate size ifthe radiatively corrected Higgs couplings and masses are used! at tree-level. The genuine SUSY-electroweak corrections are only known for the effective parameters hy, and the Higgs masses. Below-threshold decays into a Higgs particle and an of-shell gauge boson turn out to be very important for the heavy Higgs bosons of the MSSM. The individual contributions read as [55,122] rl» 1") = is SEM, «09 9GzM4,My Pit WW) = Hy EM Gy (110) > SCM PA a) = hy EM any rue awh = Ser Sw a) ret aw) = Sei (113) with the generic functions Gy 1 = ‘sf 6) = 42-1 +6) adv] S + arctan ag +g ~2alogns + 41 -w [sa +) = Any 4 2] (115) ‘sing the parameters dy = <1 4 ei +2) — (ei GF nol (116) ‘ ™ yi with 8; = 7/12 ~ 10sin® &y /9 + 40 sin’ Gy /27. Off-shell h2* decays are important for the pseudoscalar Higgs boson for masses above about 130 GeV for small tgp [56]. The decay modes H= —» hW*, AW can reach branching ratios of several tens of per cent and lead toa significant reduction of the dominant branching ratio into rv final states to 2 level of 60-70% for small tgé [55]. This analysis can be extended to the double off-shell cases analogous to the previous decay modes into Higgs-boson pais. If final-state fermion masses are neglected the corresponding expressions are given by reo ovye 3 [Satine pee aah, am . (QP = MSP + Mgry (GMP amare” .ry 1M Spita/ Progress in Paice and Nuclear Physics 95 (2017) 9 with Q2, Q2 being the squared invariant masses of the virtual Higgs and gauge bosons, My, My and J's, their masses and total decay widths; 1 is given by Gem Ty = ay, SME 13002, 2: M2) cas: oy Some C0, OF) ) with the phase-space factor A(x. y;2) of Eq. (39) 27, Supersymmetric Higgs boson decays Sermion masses and couplings, - The scalar partners fof the left- and right-handed fermion components mix with each other, The mass eigenstates j,.2 ofthe sfermions fare related to the current eigenstates fx by mixing angles 0, fi = fucose + esiney fr = fisingy + fecos oy (119) ‘which are proportional to the masses of the related fermions. Thus mixing effects are only important for the third-generation sfermions , f, the mass matrix of which is given by [86] | MMP MA ar) T= | Meh, Fx (138) rath fs A] om) Some time ago the SUSY-QCD corrections to the stop and sbottom decays of the MSSM Higgs bosons have been the higher-order corrections in effective bottom and sbottom parameters. These corrections are included in Hpecay [54], FeynHiccs [121] and Hrotp [128]. The SUSY-electroweak corrections are known, too, and modify the partial decay widths by less than about 10% [129]. They are included in FeyNHices and HroLp. Chargino/neutralino masses and couplings, The charginojneutralino masses and couplings to the MSSM Higgs bosons are derived from the Higgs mass parameter . ‘and the SU(2) gaugino mass parameter M,. The mass mattix of the charginos reads as [86] Me swing ao) Vimycosp a Ie canbe diagonalized by two mixing matrices U,V, yielding the masses ofthe physical x stats: 1 fy 2 came =. [RAP CAR NEE nim} Ma = = {Mi 2 42M, = 0B a + ah cos 2p + AMR > atau snap)} aan TR1M Spta/ Progress in Pace and Nuclear Phys 95 (2017) 98-159, a Ifeither jx or My is large, one chargino corresponds to a pure gaugino state and the other to a pure higgsino state. The Higes ‘couplings to charginos [124,130] can be expressed as (k= 1.2, 3.4 corresponds to H. h, A, H*) ty 1 : Heo x Fue = FleavaUa — daVaUs) (142) where the coefficients ey and dy are determined as e = cosa = sina, & = —sing : 4, = -sina, d= cosa, ds = cosp (148) The mass matrix of the four neutralinos clepends in addition on the U(1) gaugino mass parameter M;, which is constrained bby SUGRA models to be My ~ § tanéyMy. Inthe bino-wino-higgsino basis it acquires the form [85] My ° i) (145) with the coefficients e, dy defined in Eq. (143) The charged Higgs couplings to chargino-neutralino pairs are fixed to be [124] 1 HE tx? Fug = cos 8 | Via + —2Va(No + tan OWN an af + latte + } i Fug = sing | UnNp — xix) = a | (Fe + Fhe) (1 -— SE ¢ HOD = Fig ea, | NT OT Hae uy where nya4—+1, ny — =1and dy ~ O unless the final state consists of two identical (Majorana) neutralinos. in which ase 5, — Tre) ~ 21 stands forthe sign of the ith eigenvalue ofthe neutralino mass matrix, which can be postive or negative For chatginos these parameters are always equal to unity. The symbols iy, ¢ phase-space functions of Eq, (39) If charginojneutralino decays are kinematically allowed, which may be the case for the heavy MSSM Higgs particles HA,H®, theit branching ratios can reach large values below the corresponding top quark thresholds, They can thus jeopardize the Higgs search at the LHC due to the invisibility ofa significant fraction of these decay modes [124]. Even above the corresponding top quark thresholds the chargino/neutralino branching ratios can be sizeable, For large Higgs masses they can reach common values of a couple of 10%. In the asymptotic regime My, >> M,, the total sum of decay widths into charginos and neutralinos acquires the simple form [124,130] . _ 3GrMiy Vian? t 1 (Ya) = Meas (14 fo) cy forall thre tiggs bosons H, A, H, which is independent of any MSSM parameter (gs, x9, Ma). Normalized to the total width, which is dominated by t, BB (C6) decay modes for the neutral (charged) Higgs particles the branching ratio of charginojneutralino decays wil exceed a level of about 20% even for small and large tg In some part of the MSSM parameter space invisible ight scalar Higgs boson decays into the lightest neutralino h -» will be possible with 3 relevant branching ratio [124,130]. The full SUSY-electroweak corrections tothe Higgs boson decays into gaugino pais have been calculated [131], Apart from parameter regions where the LO decay width s suppressed they ae of moderate size, ie. not larger than about 20%. It (M2, M2; M2) denote the usual two-bodyas IM Spta/ Progress n Paice and Nuclear Physics 95 (2017) 9 should be noted that a consistent renormalization of the gaugino sector is mandatory for reliable predictions of these partial decay widths. 28. Branching ratios and total decay width 281. Standard model ‘The determination of the branching ratios of Hliggs-boson decays requires the inclusion of the aforementioned higher ‘order corrections and a sophisticated estimate of the theoretical and parametric uncertainties, The first analyses of this type have been performed some time ago [132] and recently [55,133]. The parametric uncertainties are dominated by the uncertainties in the top, bottom and charm quark masses as well as the strong coupling a, We have used the MS masses for the bottom and charm quark [57] Tiy(My) = (4.18 £0.03) GeV, _HH,(3 GeV) = (0.986 + 0.026) GeV (14) and the top quark pole mass M, = (172.54 1)GeV (150) according to the conventions of the LHC Higgs Cross Section WGG (XSW) [51]. The MIS bottom and charm masses are ‘evolved from the input scale tothe scale of the decay process with 4-loop accuracy in QCD. If needed the pole mass of the bottom quark is obtained from the 3-Loop conversion between the MI and pole mass [154] 2 4 eh) pu My mala) los 16.110 + 17 lo feet 11 log* Mp My He 2 HRY 4 yyy (MY | (astueh)* eee “ty a(i)|( ud) + [9 297 + 128 sz0tog HE ++ 32.105! HE 4 3.68309" HE My My My > 415.371 log 2 it 3 05310g ME +.0.37010g° i) +NE (0 653 +.0.32010g Mt My Mi + 0.060209! HE +-0.00926I0g? #)) () ‘| F ota’) (sn) (29201 where Nr = 4 for the bottom quark and the strong coupling a, evolves with 5 active flavours, The mass-dependent term involving the light flavours can (for 0 < x = 1) be approximated by A(x) =F x- 0.579% 40.2308. (152) The charm pole mass Mis then determined by the renormalon-free relation [135] = My ~ 3.41 GeV (153) from the bottom pole mass Mj. The strong coupling a, is fixed by the input value at the Z-boson mass scale a,{Mp) = 0.118 + 0.0015, (say The lepton as well as W, Z-boson masses and widths are chosen according to their PDG values [135] im, = (0.510998928 + 0,000000011) MeV 1m, = (105.6583715 + 0.000035) MeV m, = (1776.82 + 0.16) MeV Myy = (80.385 + 0.015) GeV Tiy = (2.085 + 0.042) Gev My = (91.1876 + 0.0021) GeV Ty = (2.4952 + 0.0023) GeV. The complex pole masses of the W and Z boson can be obtained from the relations My — ity ps — arp = (v=w.2z) (156) Vlei1M Spta Progress in Pace and Nuclear Phys 95 (2017) 98-159, 29 ables tiated theoreti uncertainties rom missing higher ovdets nthe inte mediate Higs-mass range. Source omRel 1L, aria width wo Hiecroweaie i> bye ome Howe =e Hoy <1 Aowwesy “05x wh I ar q Db bi bbe ee foo tat te aa ak 125 126127128 128130 MyiGev) Fi. 16. Higes branching ratios and their uncertainties for the mass ange around 125 GeV. Sore! Fora Re [3 These are used for the predictions of the Higgs boson decay widths into four fermions according to Ref. [63]. Moreover, the Fermi constant has been chosen as [135] Gr = 1.1663787(6) 1075 GeV? (as) The total parametric uncertainty for each branching ratio has been derived from a quadratic sum of the individual impacts, of the input parameters on the decay modes. The theoretical uncertainties from missing higher orders in the perturbative expansion are summarized in Table 5 as discussed before for the individual partial decay processes. in order to be conservative the total parametric uncertainties are added linearly to the theoretical uncertainties. The final result for the branching ratios is shown in Fig, 16 for the leading Higgs decay modes with branching ratio larger than 10~* within the Higgs-mass range between 120 and 130 GeV. They have been obtained with ProrHecya® [53] for the decays H WW ZZ and Hoecay [54] for the other decay modes. The bands represent the total uncertainties of the individual branching ratios. For a Higgs mass My = 125 GeV the total uncercainty of the leading decay mode H —> bb amounts to less than 2%, since the bulk of it cancels out within the branching ratio. The uncertainty of P(H — bb), however, generates a significant increase of the uncertainties for the subleading decay ‘modes. The total uncertainties of BRH —> WW/ZZ) and BR(H —» ¢+r~/u*w~) amount to ~2%, while the uncertainties of BRC —> ge) and BR(H —> cé) range at ~6-7%, of BR(H —> yy) at ~3% and of BRH > Zy) at ~7%. The total decay width (of ~4.1 MeV can be predicted with ~2% total uncertainty. These results constitute a basic ingredient of the corresponding, LHC analyses. 28.2. Minimal supersymmetric extension ‘The final results for the MSSM Higgs branching ratios have been obtained with the public codes FevnHtccs [121] and Hoecay [54] using the SM input parameters of the previous section, Fig. 17 shows the neutral Higgs branching ratios for two30 1M Spta/ Progress Paice and Nuclear Phys 95 (2017) 9 ais Foye t ECA ate , 5 aa 7 a pateonate 'sa0" 200300" 400-600 600-700. €00 300-1000 '"sa0 200 300-400-800" 600-700 800800 1000 apts Faptggocchy aes Fomsgpetecay i i w E Evo" w w mira Simei wel sett fe totaal 15" Ba"SoAH" NS 80a S0 no 10) SAHOO ANS 20s a 20 no wes Fata Dechy , Fey NOECAY i i || wo ro q Pines enatio tanp-20 I i... 320nao| oth atp alata ara le late tata EE! 1 REND AW HLS ESE HSS no OS ESAS 20S HE aS no My [Ge] My, (Gevl Fig. 17 Neutral MSSM Migs branching ati as functions of he coresponding Higgs mass within the mf" scenario [37] fr two values og obtained ‘bya combination of Fevnices [121] and Hosen [5 Seurce: Hom Ret [3] values of tg within the mf! benchmark scenario [137] that is defined as 1000 Gev. 1000 GeV, = Mz = 200 GeV, Ay = A, = Ar, my = 1500 GeV, Mf, Meuse X05 = 1.5 Msusy (FD calculation), X¥5 = 1.6 Msusy (RG calculation), (158) The kinks visible in these plots are due to the opening of new decay modes according to the SM and SUSY-particle masses of the final-state particles, Fig. 18 displays the corresponding charged Higgs branching ratios within the mf" benchmark scenario [137] for two values of tg. The related uncertainties are not shown in the plots.1M Spita Progress in Pace and Nuclear Physics 95 (2017) 98-159, m BRK") a My [Ge] My Ge] Fig. 18. charged MSSM Higgs branching ratios ae functions ofthe charged Higgs mss within te mi" scen ‘ofpitaton of Fevlices[127]and Hen [5 Source: From Ret (6), [197] fortwo values of tg obtained bya 9 CeETT 1b 9 22200 Fig 18. Diagrams contributing to gg —+ Hat lowest order. 3. Higgs-boson production 3.1, Gluon fusion: gg —> H 3.1.1. Standard model ‘The gluon-fusion mechanism [15] pp -> gg +H dominates Higgs-boson production at the LHC in the entire relevant Higgs mass range. The gluon coupling to the Higgs boson in the SM is mediated by triangular top- and bottom-quark loops, see Fig. 19. Since the Yukawa coupling of the Higgs particle grows with the quark mass, the form factor reaches a constant value for large loop quark masses. Ifthe masses of heavier quarks beyond the third generation are fully generated by the Higgs mechanism, these particles would add the same amount to the form factor as the top quark in the asymptotic heavy quark limit. Thus gluon fusion can serve as a counter of the number of heavy quarks, the masses of which are generated by the conventional Higgs mechanism, On the other hand ‘within the three-generation SM gluon fusion will allow to measure the top quark Yukawa coupling. This, however, requires ‘a precise knowledge of the cross section within the SM with three generations of quarks The partonic cross section can be derived from the gluonic width of the Higgs boson at lowest order [16], 08(1—z) (159) = pact» gp) — 2H |S pH o ag" 88) 88 Jax > Ag(a)| with the scaling variables defined as z = Mj /3, cq = 4M3 /M}. The variable § denotes the partonic cm. energy squared and jug the renormalization scale, The amplitudes Al(rq) are given in Eq, (50), {In the narrow-width approximation the hadronic cross section can be cast into the form | 16] dct exo > H) = ooty — (160) (pp > H) in y with the gluon luminosity ace 1 dx Se = [Sate otic 08) is)a2 1M Spita/ Progress n Paice and Nuclear Phys 95 (2017) 9 9 woOySTTTTN g ocreayere o4 yo a Fig. 20. Typical ams contributing tothe virtual and rel QCD corrections to ge > H atNLO. atthe factorization scale gir, and the scaling variable is defined, in analogy to the Drell-Yan process, as 1 = M}/s, with s specifying the total hadronic cm. energy squared, The bottom-quark contributions interfere destructively with the top loop ‘and decrease the cross section by about 10% at LO, QCD corrections. In the past the (two-loop) NLO QCD corrections to the gluon-fusion cross section, Fig. 20, have been calculated including the full mass dependences {68,70, 108, 138-140]. They consist of virtual corrections to the basic gg —> H process and real corrections due to the associated production of the Higgs boson with massless partons, ge ig and gq—> Hg, 99> He. ‘These subprocesses contribute to the Higgs production at O(a3)-The virtual corrections rescale the lowest-order fusion cross section with a coefficient depending only on the ratios ofthe iiiggs and quark masses. Gluon radiation leads to two-parton final states with invariant energy > Mj in the gg, gq and qi channels at NLO. In general the hadronic cross section can be split into seven parts [68,70, 108,138,140} eu) (pp + HX) = 09 [14%] y+ ogg + Bog + Ang + Boy + Sey (162) ‘where the finite parts of vittual corrections C and the real corrections Ag, Agge and Aogs (same-flavour quark-antiquark. initial states) start to contribute at NLO, while Aggy (same-flavour quark-quark and antiquark-antiquark initial states) and Adag (different-flavour quark and antiquark initial states) appear for the first time at NNLO. The renormalization scale 1x of a, and the factorization scale jee of the parton densities are fixed properly, in general at zg = 4 = My /2. The quark-loop mass has been identified with the pole mass Mg, while the QCD coupling a, and the parton density functions are defined in the MS scheme with five active flavours. ‘We define the NLO K factor as the ratio Kyo = M2. (163) The NLO corrections are positive and large, increasing the gluon-fusion cross section at the LHC by about 60-90%, The QCD corrections to the bottom-quark contributions are significantly smaller ifthe bottom mass is used in terms ofthe pole mass ‘or the MS mass at the scale of the bottom mass. This choice is motivated by the numerical cancellation of squared and single logarithms of the relative QCD corrections at NLO [68], This feature modifies the destructive bottom-quark contribution to a reduction of the cross section by about 6% at NLO. Comparing the exact mass-dependent results with the expressions in the heavy-quark limit, it turns out that this asymptotic K factor provides an excellent approximation even for Higgs masses above the top-decay threshold,* We explicitly define the approximation by agp = Kiyol 20) X oF, fo) (184) Kjyo(0) = lim. Kuno where we neglect the b quark contribution in Kiug(20), while the leading order cross section avo includes the full tb, € {quark mass dependence. The comparison withthe fll massive NLO result i presented in Fig, 21. The sold line corresponds to the exact crass section andthe broken line tothe approximate one. For Higgs masses below ~1 TeV, the deviations of the asymptotic approximation from the full NLO result are less than 15%, whereas for My ~ 125 GeV they amount to ~5Y, ifthe {ull LO cross section i muplied by the approximate k-factor. This property ofthe NLO corrections suggest this tobe true also at higher orders since itis a consequence of the dominating sof and collinear gluon effects in the QCD corrections ‘Within the heavy top-quark limit the NNLO [ 142] and N*LO [143,144] QCD corrections have been calculated. The NNLO contributions increase the production cross section by about 20% beyond NLO, while the N’LO corrections range at the level of afew per-cent. These results indicate that the gluon-fasion cross section is under theoretical control despite the large size ofthe NLO corrections. Thisis further corroborate by the results obtained by a soft and collinear gluon resummation on top Large Higgs masses ae stil relevant for off-shell Higgs bosons.1M Spita Progress in Pace and Nuclear Phys 95 (2017) 98-159. 1m 0 tpp 9 H+ X06) 10 _ 1 10" ° BB 00 MylGev] Fig.21. Comparison ofthe exact and approximate NLO cross section o(pp —> H +X)at the LHC with cm. energy 3 = 13 TeV. The solid line shows the exact cess section including she ful, b,c quark mass dependence and the dashed lie eostespands to the approximation defined in Fa (158). The renormalization and facarization tales have been identified wit half ofthe Higgs mass stg — uy = My/2 andthe POFALHCIS NLO parton densities) ‘with NLO streng coupling [a Me) ~ 0.118) have been adopted. The top mass as been chosen as My = 172.5 Ge, he bottom mass as My = 4.84GeV and the charm macs as Me = 143 GeV, of the NO result. This approach resums the dominant factorizing contributions from soft and collinear gluon effects up to all perturbative orders. The soft corrections provide the leading ones ciose to the production threshold, while collinear gluon effect are of subleading order. Both can be treated systematically, The soft and collinear contributions provide a reasonable approximation of the full fixed-order results and thus a reliable estimate of missing higher-order effects beyond the fixed- order corrections. This resummation has been performed at the NNLL level [145] and the NLL level [146] in the heavy top-quark limit. The sizes ofthe different logarithmic orders follow roughly the pattern of the corresponding fixed-order corrections. Quite recently also finite top-mass effects have been included in the resummation framework [147] atthe NLL ‘order, where they are known exactly. Resummation effects beyond N’LO yield only a per-cent increase ofthe cross section for the central scale choice. However, they provide an approximation of effects beyond N°LO and contribute toa sophisticated estimate ofthe residual uncertainties by elaborating on the uncertainties due to the matching tothe fixed-order expression, Electroweak corrections. The electroweak corrections to the gluon-fusion cross section have been computed approximately first. The leading top mass corrections of (GrM?) coincide with the corrections to the gluonic decay mode of Eq. (60) and are thus small [49,78], This calculation has been refined by the determination of the NLO electroweak corrections {due to light-fermion loops [79] and finally by the full numerical integration of the exact NLO corrections to the top- and W. Z-induced electroweak corrections [80]. The electroweak corrections coincide with the ones to the H — gg decay. The NLO electroweak corrections have been extencled by a calculation of the mixed QCD-clectroweak corrections in the limit Mj. < Mi, [148] which can be attributed tothe corrections ofthe effective Hgg coupling, However, itis unclear how reliable this approximation isin practice. Due tothe dominance of soft and collinear gluon effects the bulk of the electroweak corrections will actorize from the pure QCD corrections. Inthe following the electroweak corrections will be combined with the QCD corrections in factorized form Total cross section. Theoretical uncertainties in the prediction ofthe Higgs cross section originate from three major sources, the dependence of the cross section on the parton densities, the unknown corrections beyond N'LO and the parametric ‘uncertainties originating from the input value of the strong coupling a, and to a lesser extent ofthe top and bottom quark ‘masses. The missing quark-mass effects beyond NLO have been estimated as less than 1% by an explicit large top-mass ‘expansion of the NNLO corrections beyond the heavy top-quark limit [149]. The total uncertainty of the prediction for the total gluon-fusion cross sections has been estimated as 4% and can dominantly be traced back to the renormalization- and factorization-scale dependence as well as the PDF, uncertainties [51]. The tenormalization- and factorization-scale dependence is depicted in Fig, 22 for the known perturbative orders in the heavy top-quark limit. A significant reduction from 50 — 100% to the few-per-cent level is visible from LO to N?LO. The results for the cross section are contained in the Public codes Hratu [150] up to NNLO as well as SusH [151], 1H1x [152] and cctlices [153] up to N?LO. The softfcollinear~ sluon-resummed results can be obtained with the code Trott for corrections beyond N?L0 [154] ‘Transverse-momentum distribution. Apart from the total cross section also distributions are relevant for analysing the Higgs properties. Of particular interest in this context is the transverse-momentum distribution of the Higgs boson that arises due to the additional radiation of gluons. The LO contributions are part ofthe real corrections to the gluon-fusion cross section