5.3 Vector-Like Quarks
169
two b-tagged jets and events where the H boson mass is reconstructed outside the
region 90–140 GeV. Five signal regions for the resolved and merged categories are
defined, depending on the reconstructed m VLQ and m Z . The observed numbers of
events in the ten signal regions are used in the statistical evaluation, performed as a
counting experiment. The analysis places upper limits on the product of cross section
σ ( pp → Bb) and branching fractions B(B → Hb)B(H → bb) for m VLQ < m Z /2
and m VLQ = m Z − 250 GeV in a plane of m VLQ versus m Z .
The case of a large coupling between the Z
and up-type quarks can be investigated in searches for pp → Z
→ T t. A promising decay channel is W bW b [904],
which differs from the kinematics of the Z
→ tt → W bW b resonance search in
two important aspects. The large mass of the T results in very different boosts of
the two W b systems. The top quark from the Z
decay can receive a large boost if
the mass difference m Z − m VLQ is large, while the boost of the T will be moderate
at most for m Z in the range 1.5–4 TeV and m VLQ between 0.7 and 3 TeV. When
considering the constraint m Z < 2m VLQ , Lorentz factors not larger than γ = 1.5 are
realised for the T , such that the W b system from its decay can not be reconstructed
in a single jet. However, the W and b will be approximately back-to-back, with large
p T in the laboratory rest frame. The second aspect is that the W b system from the T
decay will have a mass close to m VLQ m t , such that the usual selection employed
in tt resonance searches will result in a rejection of these events. Both aspects, the
different boosts and different masses of the two W b systems, result in an insensitivity of existing searches to this signal, despite the same final state. A dedicated
search has been carried out by CMS in the all-hadronic final state using 13 TeV data
with 2.6 fb
−1 [932]. The analysis selects events with a three-jet topology, with one
large-R t-tagged, one large-R W-tagged and one small-R b-tagged jet. The t and
W tagging relies on the soft drop mass, and τ 32 and τ 21 , respectively. It is checked
that the b-tagged small-R jet does not overlap with the two large-R jets. Two signal
regions are defined, depending on the presence of a b-tagged subjet in the identified
t jet. Both signal regions have approximately the same signal efficiency, with different background efficiencies and compositions. The subjet b tag reduces the multijet
background by a factor of about four, such that the corresponding signal region has
better sensitivity than the one without a subjet b tag. However, the latter still contributes to the overall sensitivity of this search and validates the multijet background
estimation, which is obtained from sideband regions with vetoes on b-tagged jets
and subjets. The uniquely identified decay particles of the signal decay chain allow
for a reconstruction of m Z and m VLQ , where both masses could be determined in
case of a potential signal in the data. The distributions in m Z and m VLQ are shown
in Fig. 5.17 for the signal region with a subjet b-tag. The relative mass resolution for
signal events is about 15%, such that pronounced peaks on the falling background
would be visible. Since both distributions are obtained from the same events in data,
only the distribution in m Z is used to extract upper cross section limits on a potential
signal.
While the all-hadronic search achieves high sensitivity for T → W b decays, the
channels T → Ht and T → Zt have been targeted by a dedicated search optimised
for pp → Z
→ T t → Htt and Ztt. The search has been carried out by CMS in the
169
two b-tagged jets and events where the H boson mass is reconstructed outside the
region 90–140 GeV. Five signal regions for the resolved and merged categories are
defined, depending on the reconstructed m VLQ and m Z . The observed numbers of
events in the ten signal regions are used in the statistical evaluation, performed as a
counting experiment. The analysis places upper limits on the product of cross section
σ ( pp → Bb) and branching fractions B(B → Hb)B(H → bb) for m VLQ < m Z /2
and m VLQ = m Z − 250 GeV in a plane of m VLQ versus m Z .
The case of a large coupling between the Z
and up-type quarks can be investigated in searches for pp → Z
→ T t. A promising decay channel is W bW b [904],
which differs from the kinematics of the Z
→ tt → W bW b resonance search in
two important aspects. The large mass of the T results in very different boosts of
the two W b systems. The top quark from the Z
decay can receive a large boost if
the mass difference m Z − m VLQ is large, while the boost of the T will be moderate
at most for m Z in the range 1.5–4 TeV and m VLQ between 0.7 and 3 TeV. When
considering the constraint m Z < 2m VLQ , Lorentz factors not larger than γ = 1.5 are
realised for the T , such that the W b system from its decay can not be reconstructed
in a single jet. However, the W and b will be approximately back-to-back, with large
p T in the laboratory rest frame. The second aspect is that the W b system from the T
decay will have a mass close to m VLQ m t , such that the usual selection employed
in tt resonance searches will result in a rejection of these events. Both aspects, the
different boosts and different masses of the two W b systems, result in an insensitivity of existing searches to this signal, despite the same final state. A dedicated
search has been carried out by CMS in the all-hadronic final state using 13 TeV data
with 2.6 fb
−1 [932]. The analysis selects events with a three-jet topology, with one
large-R t-tagged, one large-R W-tagged and one small-R b-tagged jet. The t and
W tagging relies on the soft drop mass, and τ 32 and τ 21 , respectively. It is checked
that the b-tagged small-R jet does not overlap with the two large-R jets. Two signal
regions are defined, depending on the presence of a b-tagged subjet in the identified
t jet. Both signal regions have approximately the same signal efficiency, with different background efficiencies and compositions. The subjet b tag reduces the multijet
background by a factor of about four, such that the corresponding signal region has
better sensitivity than the one without a subjet b tag. However, the latter still contributes to the overall sensitivity of this search and validates the multijet background
estimation, which is obtained from sideband regions with vetoes on b-tagged jets
and subjets. The uniquely identified decay particles of the signal decay chain allow
for a reconstruction of m Z and m VLQ , where both masses could be determined in
case of a potential signal in the data. The distributions in m Z and m VLQ are shown
in Fig. 5.17 for the signal region with a subjet b-tag. The relative mass resolution for
signal events is about 15%, such that pronounced peaks on the falling background
would be visible. Since both distributions are obtained from the same events in data,
only the distribution in m Z is used to extract upper cross section limits on a potential
signal.
While the all-hadronic search achieves high sensitivity for T → W b decays, the
channels T → Ht and T → Zt have been targeted by a dedicated search optimised
for pp → Z
→ T t → Htt and Ztt. The search has been carried out by CMS in the