5.3 Vector-Like Quarks
153
separation power between signal and background events. In the W t analysis, events
failing the selection criteria, e.g. not enough jets are found or no jet has been W
tagged, can not be reconstructed with this algorithm. Instead of discarding these
events, they are passed to a BDT in order to retain signal efficiency. The analyses
result in mass limits of 1.35 TeV for B(T /Y → W b) = 1 [877] and B(B/ X →
W t) = 1 [875]. The Z (νν)t analysis targets final states with one lepton and p
miss
T
>
350 GeV from the Z → νν decay. In addition, two large-R jets with trimmed jet
masses above 60 and 80 GeV are selected, representing the other VLQ decay in
the event. The backgrounds are already sufficiently suppressed in the signal region,
such that no specific assumptions on the second VLQ decay need to be made and
no explicit jet tagging is applied. Mass limits of 1.16 TeV for B(T → Zt) = 1 are
obtained [878]. The most involved of the four analyses is the analysis targeting
Ht + X final states [879], which also includes a p
miss
T +jets final state to improve the
sensitivity for Z (νν)t + X decays. Orthogonality to the Z (νν)t analysis in the +jets
final state is achieved by a lepton-veto. This analysis uses re-clustering of large-R
jets [431] and defines t jets by a trimmed jet mass larger than 140 GeV and at least
two subjets; H jets are required to have 105 < m jet < 140 GeV and either exactly
two subjets if p T < 500 GeV or one or two subjets if p T > 500 GeV. Selected events
are categorised into different regions depending on the small-R jet multiplicity, the
b-tagged jet multiplicity, and the H and t-tagged jet multiplicity. Overall, this results
in 34 signal regions and 26 validation regions, which are used to constrain the SM
background predictions. In each of these signal regions, the S T distribution (called
effective mass in the publication, m eff ) is used as sensitive variable to search for
a signal. The total event yields in the signal regions are summarised in Fig. 5.11,
where also the possible contribution from T T production is shown, where the T is
assumed to be in an electroweak doublet with B(T → Zt) = B(T → Ht) ≈ 0.5
and B(T → W b) = 0. The best sensitivity is observed in signal regions with one or
more H - or t-tagged large-R jets and (partially overlapping) b-tagged small-R jets.
The analysis reports the strongest mass limits up to date from a single analysis of
1.43 TeV for B(T → Ht) = 1.
There is one exception to the inclusive approach followed by CMS in the +jets
final state. The similarity of T T → W bW b with tt → W bW b, as well as the dominant decays T → W b and Y → W b, have prompted CMS to perform a dedicated
analysis in the +jets channel [880]. Because the νbqqb final state can be fully
reconstructed, it is possible to constrain the reconstructed neutrino and jet momenta
3
by a kinematic fit [881, 882], to achieve optimal resolution in the reconstructed
VLQ mass. In order to take decays of highly boosted W bosons into account, the
soft-drop mass of large-R jets is required to be between 60 and 100 GeV. Once a
W jet has been found, its two soft-drop subjets are used as input to the kinematic
fit instead of the corresponding one or two small-R jets. The kinematic fit results
in an excellent relative mass resolution of about 7%, which compensates for the
3 Also the lepton momentum is adjusted, but to a lesser degree, because the experimental resolution
is better than for jets.
153
separation power between signal and background events. In the W t analysis, events
failing the selection criteria, e.g. not enough jets are found or no jet has been W
tagged, can not be reconstructed with this algorithm. Instead of discarding these
events, they are passed to a BDT in order to retain signal efficiency. The analyses
result in mass limits of 1.35 TeV for B(T /Y → W b) = 1 [877] and B(B/ X →
W t) = 1 [875]. The Z (νν)t analysis targets final states with one lepton and p
miss
T
>
350 GeV from the Z → νν decay. In addition, two large-R jets with trimmed jet
masses above 60 and 80 GeV are selected, representing the other VLQ decay in
the event. The backgrounds are already sufficiently suppressed in the signal region,
such that no specific assumptions on the second VLQ decay need to be made and
no explicit jet tagging is applied. Mass limits of 1.16 TeV for B(T → Zt) = 1 are
obtained [878]. The most involved of the four analyses is the analysis targeting
Ht + X final states [879], which also includes a p
miss
T +jets final state to improve the
sensitivity for Z (νν)t + X decays. Orthogonality to the Z (νν)t analysis in the +jets
final state is achieved by a lepton-veto. This analysis uses re-clustering of large-R
jets [431] and defines t jets by a trimmed jet mass larger than 140 GeV and at least
two subjets; H jets are required to have 105 < m jet < 140 GeV and either exactly
two subjets if p T < 500 GeV or one or two subjets if p T > 500 GeV. Selected events
are categorised into different regions depending on the small-R jet multiplicity, the
b-tagged jet multiplicity, and the H and t-tagged jet multiplicity. Overall, this results
in 34 signal regions and 26 validation regions, which are used to constrain the SM
background predictions. In each of these signal regions, the S T distribution (called
effective mass in the publication, m eff ) is used as sensitive variable to search for
a signal. The total event yields in the signal regions are summarised in Fig. 5.11,
where also the possible contribution from T T production is shown, where the T is
assumed to be in an electroweak doublet with B(T → Zt) = B(T → Ht) ≈ 0.5
and B(T → W b) = 0. The best sensitivity is observed in signal regions with one or
more H - or t-tagged large-R jets and (partially overlapping) b-tagged small-R jets.
The analysis reports the strongest mass limits up to date from a single analysis of
1.43 TeV for B(T → Ht) = 1.
There is one exception to the inclusive approach followed by CMS in the +jets
final state. The similarity of T T → W bW b with tt → W bW b, as well as the dominant decays T → W b and Y → W b, have prompted CMS to perform a dedicated
analysis in the +jets channel [880]. Because the νbqqb final state can be fully
reconstructed, it is possible to constrain the reconstructed neutrino and jet momenta
3
by a kinematic fit [881, 882], to achieve optimal resolution in the reconstructed
VLQ mass. In order to take decays of highly boosted W bosons into account, the
soft-drop mass of large-R jets is required to be between 60 and 100 GeV. Once a
W jet has been found, its two soft-drop subjets are used as input to the kinematic
fit instead of the corresponding one or two small-R jets. The kinematic fit results
in an excellent relative mass resolution of about 7%, which compensates for the
3 Also the lepton momentum is adjusted, but to a lesser degree, because the experimental resolution
is better than for jets.