| : |
| : |
| : |
| : |
| : The Feynman diagrams of DM ($\chi$) production in association with the observed Higgs boson ($h$) arising from three theoretical models considered in this paper: (a) the $Z'_{B}$ model, (b) the $Z'$-2HDM model, and (c), (d), and (e) the 2HDM+a model. |
| : The Feynman diagrams of DM ($\chi$) production in association with the observed Higgs boson ($h$) arising from three theoretical models considered in this paper: (a) the $Z'_{B}$ model, (b) the $Z'$-2HDM model, and (c), (d), and (e) the 2HDM+a model. |
| : Caption not extracted |
| : Caption not extracted |
| Comparison of the \etmiss distribution in MC simulation and data, shown after the photon requirements but before the preselection requirements prior to BDT training. In particular, events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV. In addition to the signal used for training (blue), one representative signal is also shown for each of the three signal models. The $Z'$-2HDM, $Z'_B$, 2HDM+a ($\tan\beta=1, \sin\theta = 0.7, m_A = 300\ \GeV, m_a = 250\ \GeV$), and 2HDM+a ($\tan\beta=1, \sin\theta = 0.35, m_A = 600\ \GeV, m_a = 200\ \GeV$) signal models are normalized to their respective theoretical cross sections times branching ratios of 0.0815 fb, 0.411 fb, 0.269 fb, and 0.533 fb. The lower panel shows the ratio of data to MC. The uncertainty bands indicate the sum of statistical and experimental systematic uncertainties. |
| Comparison of the \etmiss distribution in MC simulation and data, shown after the photon requirements but before the preselection requirements prior to BDT training. In particular, events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV. In addition to the signal used for training (blue), one representative signal is also shown for each of the three signal models. The lower panel shows the ratio of data to MC. The uncertainty bands indicate the sum of statistical and experimental systematic uncertainties. |
| Data and MC distributions of the BDT input variables, \pTgg and \setmiss, after the photon requirements and the preselection requirements prior to BDT training. In particular, events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV. In addition to the signal used for training (blue), one representative signal is also shown for each of the three signal models. The $Z'$-2HDM, $Z'_B$, 2HDM+a ($\tan\beta=1, \sin\theta = 0.7, m_A = 300\ \GeV, m_a = 250\ \GeV$), and 2HDM+a ($\tan\beta=1, \sin\theta = 0.35, m_A = 600\ \GeV, m_a = 200\ \GeV$) signal models are normalized to their respective theoretical cross sections times branching ratios of 0.0815 fb, 0.411 fb, 0.269 fb, and 0.533 fb. The lower panel shows the ratio of data to MC. The narrow uncertainty bands (indicating the sum of statistical and experimental systematic uncertainties) at high \setmiss values are due to the absence of $\gamma\gamma$ and $\gamma$+jet contributions, which have large jet and \etmiss-related systematic uncertainties. |
| Data and MC distributions of the BDT input variables, \pTgg and \setmiss, after the photon requirements and the preselection requirements prior to BDT training. In particular, events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV. In addition to the signal used for training (blue), one representative signal is also shown for each of the three signal models. The lower panel shows the ratio of data to MC. The narrow uncertainty bands (indicating the sum of statistical and experimental systematic uncertainties) at high \setmiss values are due to the absence of $\gamma\gamma$ and $\gamma$+jet contributions, which have large systematic uncertainties. |
| Data and MC distributions of the BDT input variables, \pTgg and \setmiss, after the photon requirements and the preselection requirements prior to BDT training. In particular, events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV. In addition to the signal used for training (blue), one representative signal is also shown for each of the three signal models. The $Z'$-2HDM, $Z'_B$, 2HDM+a ($\tan\beta=1, \sin\theta = 0.7, m_A = 300\ \GeV, m_a = 250\ \GeV$), and 2HDM+a ($\tan\beta=1, \sin\theta = 0.35, m_A = 600\ \GeV, m_a = 200\ \GeV$) signal models are normalized to their respective theoretical cross sections times branching ratios of 0.0815 fb, 0.411 fb, 0.269 fb, and 0.533 fb. The lower panel shows the ratio of data to MC. The narrow uncertainty bands (indicating the sum of statistical and experimental systematic uncertainties) at high \setmiss values are due to the absence of $\gamma\gamma$ and $\gamma$+jet contributions, which have large jet and \etmiss-related systematic uncertainties. |
| Data and MC distributions of the BDT input variables, \pTgg and \setmiss, after the photon requirements and the preselection requirements prior to BDT training. In particular, events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV. In addition to the signal used for training (blue), one representative signal is also shown for each of the three signal models. The lower panel shows the ratio of data to MC. The narrow uncertainty bands (indicating the sum of statistical and experimental systematic uncertainties) at high \setmiss values are due to the absence of $\gamma\gamma$ and $\gamma$+jet contributions, which have large systematic uncertainties. |
| The BDT score for selected signals, the data control region, and the data sideband, with 1 being more signal-like and 0 being more background-like. The data sideband consists of events with two `tight' identified and isolated photon candidates, while the data control region consists of diphoton events where at least one photon candidate fails at least one of the identification or isolation requirements. All events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV, and events in the $120 < m_{\gamma\gamma} < 130$~\GeV\ region of the data sideband are vetoed. The error bars represent statistical uncertainties. |
| The BDT score for selected signals, the data control region, and the data sideband, with 1 being more signal-like and 0 being more background-like. The data sideband consists of events with two `tight' identified and isolated photon candidates, while the data control region consists of diphoton events where at least one photon candidate fails at least one of the identification or isolation requirements. All events are required to have $105 < m_{\gamma\gamma} < 160$~\GeV, and events in the $120 < m_{\gamma\gamma} < 130$~\GeV\ region of the data sideband are vetoed. The error bars represent statistical uncertainties. |
| The BDT score as a function of the input variables, \pTgg and \setmiss, with 1 being more signal-like and 0 being more background-like. The plot is produced by splitting the plane into equal cells of $\pTgg \times \setmiss$, and evaluating the corresponding BDT score. The black (red) lines indicate the minimum score boundaries for the BDT categories in the high \etmiss (low \etmiss) region, as defined in Table~\ref{tab:category_definition}. |
| The BDT score as a function of the input variables, \pTgg and \setmiss, with 1 being more signal-like and 0 being more background-like. The plot is produced by splitting the plane into equal cells of $\pTgg \times \setmiss$, and evaluating the corresponding BDT score. The black (red) lines indicate the minimum score boundaries for the BDT categories in the high \etmiss (low \etmiss) region, as defined in Table~\ref{tab:category_definition}. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The diphoton invariant mass spectra from the data and the corresponding fitted signal and background in each BDT category. The signal is the 2HDM+a model with $\tan\beta=1.0$, $\sin\theta=0.35$, $m_{A}=600$~\GeV, and $m_{a}=200$~\GeV. Different signal models have similar \mgg shapes and therefore will mainly differ in the relative signal contributions in the various categories. The non-resonant background and the predicted SM Higgs boson contribution are shown. The blue curve shows the sum of the signal, SM Higgs boson, and non-resonant background after the fit. The subplot shows the residual between the observed number of events and the fitted non-resonant and SM Higgs boson backgrounds. The error bars represent statistical uncertainties. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the $Z'_{B}$ model in the $m_{\chi}$--$m_{Z'}$ plane, for $\sin \theta = 0.3$, $g_q = 1/3$, and $g_{\chi} = 1$. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the $Z'_{B}$ model in the $m_{\chi}$--$m_{Z'}$ plane, for $\sin \theta = 0.3$, $g_q = 1/3$, and $g_{\chi} = 1$. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. |
| A comparison of the inferred limits with the constraints from direct detection experiments on the spin-independent DM--nucleon cross section in the context of the $Z'_B$ simplified model with vector couplings. Limits are shown at 90\% CL. The results from this analysis, in which the region on the side of the hatched band inside the contour is excluded, are compared with limits from the XENON~\cite{PhysRevLett.123.251801,PhysRevLett.123.241803,PhysRevLett.121.111302} and DarkSide-50~\cite{PhysRevLett.121.081307} experiments. The comparison is model-dependent and solely valid in the context of this model, assuming Dirac fermion DM, mixing angle $\sin\theta=0.3$, and the coupling values $g_q = 1/3$ and $g_{\chi}= 1$. The diagonal upper branch of the limit curve reflects the fact that there are no parameters of the model that predict a cross section larger than this limit. The impact of renormalization-group evolution effects~\cite{Crivellin:2014qxa,DEramo:2014nmf} when comparing collider and direct detection limits is not taken into consideration here. |
| A comparison of the inferred limits with the constraints from direct detection experiments on the spin-independent DM--nucleon cross section in the context of the $Z'_B$ simplified model with vector couplings. Limits are shown at 90\% CL. The results from this analysis, in which the region on the side of the hatched band inside the contour is excluded, are compared with limits from the XENON~\cite{PhysRevLett.123.251801,PhysRevLett.123.241803,PhysRevLett.121.111302} and DarkSide-50~\cite{PhysRevLett.121.081307} experiments. The comparison is model-dependent and solely valid in the context of this model, assuming Dirac fermion DM, mixing angle $\sin\theta=0.3$, and the coupling values $g_q = 1/3$ and $g_{\chi}= 1$. The diagonal upper branch of the limit curve reflects the fact that there are no parameters of the model that predict a cross section larger than this limit. The impact of renormalization-group evolution effects~\cite{Crivellin:2014qxa,DEramo:2014nmf} when comparing collider and direct detection limits is not taken into consideration here. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the $Z'$-2HDM model in the $m_A$--$m_{Z'}$ plane, for $\tan \beta = 1.0$, $g_{Z'} = 0.8$, $m_{\chi} = 100$~\GeV, and $m_{H^{0,\pm}}=m_{A}$. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. Above $m_A = 350$~\GeV, competing decays from $A \to t\bar{t}$ cause the $A\to\chi\chi$ branching ratio to decrease quickly with increasing $m_A$, resulting in the feature near $m_{Z'} = 1300$~\GeV. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the $Z'$-2HDM model in the $m_A$--$m_{Z'}$ plane, for $\tan \beta = 1.0$, $g_{Z'} = 0.8$, $m_{\chi} = 100$~\GeV, and $m_{H^{0,\pm}}=m_{A}$. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. Above $m_A = 350$~\GeV, competing decays from $A \to t\bar{t}$ cause the $A\to\chi\chi$ branching ratio to decrease quickly with increasing $m_A$, resulting in the feature near $m_{Z'} = 1300$~\GeV. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the 2HDM+a model in the $m_A$--$m_a$ plane, for $\tan \beta = 1.0$, $\sin \theta = 0.35$, and $m_{\chi} = 10$~\GeV. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. Around the threshold $m_A = 350$~\GeV, the competition between the resonant decay $A \to t\bar{t}$ and $A \to ah$ causes the $A\to a h$ branching ratio to decrease suddenly with a limited increase of $m_A$, resulting in the feature near $m_{a} = 200$~\GeV. Above $m_A = 350$~\GeV, the limit is mainly driven by the increased selection efficiency due to the harder signal \etmiss distribution. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the 2HDM+a model in the $m_A$--$m_a$ plane, for $\tan \beta = 1.0$, $\sin \theta = 0.35$, and $m_{\chi} = 10$~\GeV. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. Around the threshold $m_A = 350$~\GeV, the competition between the resonant decay $A \to t\bar{t}$ and $A \to ah$ causes the $A\to a h$ branching ratio to decrease suddenly with a limited increase of $m_A$, resulting in the feature near $m_{a} = 200$~\GeV. Above $m_A = 350$~\GeV, the limit is mainly driven by the increased selection efficiency due to the harder signal \etmiss distribution. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the 2HDM+a model in the $\tan\beta$--$m_a$ plane, for $m_{A,H^{\pm},H}= 600$~\GeV, $\sin \theta = 0.35$, and $m_{\chi} = 10$~\GeV. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. The region $\tan\beta<0.4$ is covered with a hatched band because there the decay width of the low-mass Higgs boson is greater than 20\% of its mass, which renders the cross-section calculation unreliable~\cite{Aaboud:2019yqu}. The shape of the limit curve closely follows the signal cross section, which is dominated by ggF for low $\tan\beta$ and $b\bar{b}$-fusion for high $\tan\beta$. |
| The observed (solid line) and expected (dashed lines) exclusion limit contours at 95\% CL for the 2HDM+a model in the $\tan\beta$--$m_a$ plane, for $m_{A,H^{\pm},H}= 600$~\GeV, $\sin \theta = 0.35$, and $m_{\chi} = 10$~\GeV. The dotted lines represent the $\pm1\sigma$ theoretical uncertainty for the observed limit. The $\pm1\sigma$ expected exclusion limit contour is shown as the yellow band. The region $\tan\beta<0.4$ is covered with a hatched band because there the decay width of the low-mass Higgs boson is greater than 20\% of its mass, which renders the cross-section calculation unreliable~\cite{Aaboud:2019yqu}. |
| The observed (solid line) and expected (dashed lines) exclusion limits at 95\% CL for the 2HDM+a model as a function of $\sin \theta$, for $m_{A,H^{\pm},H}= 600$~\GeV, $m_a = 200$~\GeV, and $\tan \beta = 1.0$. Since the predicted yield for this model vanishes at mixing angles of $\theta = 0$ and $\theta = \pi/2$, $\sin\theta$ is limited to the range $[0.1, 0.9]$. |
| The observed (solid line) and expected (dashed lines) exclusion limits at 95\% CL for the 2HDM+a model as a function of $\sin \theta$, for $m_{A,H^{\pm},H}= 600$~\GeV, $m_a = 200$~\GeV, and $\tan \beta = 1.0$. Since the predicted yield for this model vanishes at mixing angles of $\theta = 0$ and $\theta = \pi/2$, $\sin\theta$ is limited to the range $[0.1, 0.9]$. |