168
5 Direct Searches for New Physics
tive width, as done for example in [910, 913, 914]. The equivalence of these two
approaches can been shown [920], such that limits presented using one of these
approaches can be translated into the other. This also enables a future combination
of the variety of experimental results on single VLQ searches.
5.3.3 Production Through Resonance Decays
In ultraviolet-complete physical models, i.e. renormalisable theories with gauge
bosons resulting from symmetry requirements, VLQs are not the only new heavy
resonances in the particle content of the theory. The underlying symmetry groups
require the presence of spin-1 resonances, as for example in minimal composite
Higgs models [921, 922], where spin-1/2 VLQs appear naturally together with electrically neutral and charged spin-1 resonances ρ
0 and ρ
± [844, 845, 923]. In models
with extra dimensions [699], the lightest Kaluza-Klein excitation of the gluon, G
∗ ,
can couple to the lightest fermionic resonances, which are VLQs [632, 924, 925].
In general, a Higgs boson with a mass of 125 GeV requires fermions with masses
O(1 TeV) [926, 927], which are usually lighter than the spin-1 resonances. In such
models, decays of the heavy resonances, generically called Z
and W
, into VLQs
are allowed. If the mass of these resonances is larger than 2m VLQ , decays into pairs
of VLQs are possible with branching fractions of 60% and higher [925]. This usually results also in a large width of the spin-1 resonance, such that limits from the
non-resonant pair production of VLQs can be reinterpreted by rescaling the production cross section [928]. If the mass of these resonances is smaller than 2m VLQ ,
but larger than m VLQ + m q , then mixed decays into a SM quark with mass m q and
a VLQ can be dominant [929]. These heavy-light decays include W
→ T b and
W
→ t B, complementing the light-light decays W
→ tb and heavy-heavy decays
W
→ T B. While the light-light decays are covered by searches for tb resonances
and VLQ pair production, heavy-light decays need dedicated analyses at the LHC
in order to achieve the best sensitivity. For the neutral Z
the situation is analogous,
where the heavy-light decays Z
→ Bb and Z
→ T t are not covered by searches
for tt resonances and VLQ pair production, such that dedicated analyses for the
production of single VLQs in resonance decays have to be carried out.
The first search for the production of a single VLQ in the decay of a heavy
resonance has been performed by ATLAS using 19.5 fb
−1 of 8 TeV data [930]. It
targets the process pp → Z
→ Bb → Hbb → bbbb, as suggested in [931]. For
large m VLQ the H → bb decay is reconstructed with a large-R jet with p T > 300 GeV
and trimmed jet mass in the range 90–140 GeV. The large-R jet is required to be
matched to at least one b-tagged small-R jet. Two additional small-R jets have
to be present in the event. If no H jet is found, the event is reconstructed using
four small-R jets with p T >50 GeV, three of which have to be b-tagged. In this
case, the H boson candidate is reconstructed from the two jets with invariant mass
nearest to 126 GeV and p T of the dijet system larger than 200 GeV. The multijet
background is estimated from control regions, obtained from events with exactly
5 Direct Searches for New Physics
tive width, as done for example in [910, 913, 914]. The equivalence of these two
approaches can been shown [920], such that limits presented using one of these
approaches can be translated into the other. This also enables a future combination
of the variety of experimental results on single VLQ searches.
5.3.3 Production Through Resonance Decays
In ultraviolet-complete physical models, i.e. renormalisable theories with gauge
bosons resulting from symmetry requirements, VLQs are not the only new heavy
resonances in the particle content of the theory. The underlying symmetry groups
require the presence of spin-1 resonances, as for example in minimal composite
Higgs models [921, 922], where spin-1/2 VLQs appear naturally together with electrically neutral and charged spin-1 resonances ρ
0 and ρ
± [844, 845, 923]. In models
with extra dimensions [699], the lightest Kaluza-Klein excitation of the gluon, G
∗ ,
can couple to the lightest fermionic resonances, which are VLQs [632, 924, 925].
In general, a Higgs boson with a mass of 125 GeV requires fermions with masses
O(1 TeV) [926, 927], which are usually lighter than the spin-1 resonances. In such
models, decays of the heavy resonances, generically called Z
and W
, into VLQs
are allowed. If the mass of these resonances is larger than 2m VLQ , decays into pairs
of VLQs are possible with branching fractions of 60% and higher [925]. This usually results also in a large width of the spin-1 resonance, such that limits from the
non-resonant pair production of VLQs can be reinterpreted by rescaling the production cross section [928]. If the mass of these resonances is smaller than 2m VLQ ,
but larger than m VLQ + m q , then mixed decays into a SM quark with mass m q and
a VLQ can be dominant [929]. These heavy-light decays include W
→ T b and
W
→ t B, complementing the light-light decays W
→ tb and heavy-heavy decays
W
→ T B. While the light-light decays are covered by searches for tb resonances
and VLQ pair production, heavy-light decays need dedicated analyses at the LHC
in order to achieve the best sensitivity. For the neutral Z
the situation is analogous,
where the heavy-light decays Z
→ Bb and Z
→ T t are not covered by searches
for tt resonances and VLQ pair production, such that dedicated analyses for the
production of single VLQs in resonance decays have to be carried out.
The first search for the production of a single VLQ in the decay of a heavy
resonance has been performed by ATLAS using 19.5 fb
−1 of 8 TeV data [930]. It
targets the process pp → Z
→ Bb → Hbb → bbbb, as suggested in [931]. For
large m VLQ the H → bb decay is reconstructed with a large-R jet with p T > 300 GeV
and trimmed jet mass in the range 90–140 GeV. The large-R jet is required to be
matched to at least one b-tagged small-R jet. Two additional small-R jets have
to be present in the event. If no H jet is found, the event is reconstructed using
four small-R jets with p T >50 GeV, three of which have to be b-tagged. In this
case, the H boson candidate is reconstructed from the two jets with invariant mass
nearest to 126 GeV and p T of the dijet system larger than 200 GeV. The multijet
background is estimated from control regions, obtained from events with exactly