Search for the standard model Higgs boson in $\ell \nu \mathbf{+}\mathrm{\text{jets}}$ final states in $9.7{\mathrm{fb}}^{\mathbf{-}1}$ of $p\overline{p}$ collisions with the D0 detector

We present, in detail, a search for the standard model Higgs boson, $H$, in final states with a charged lepton (electron or muon), missing energy, and two or more jets in data corresponding to $9.7{\mathrm{fb}}^{-1}$ of integrated luminosity collected at a center-of-mass energy of $\sqrt{s}=1.96\mathrm{TeV}$ with the D0 detector at the Fermilab Tevatron $p\overline{p}$ Collider. The search uses $b$-jet identification to categorize events for improved signal versus background separation and is sensitive to associated production of the $H$ with a $W$ boson, $WH\to \ell \nu b\overline{b}$; gluon fusion with the Higgs decaying to $W$-boson pairs, $H\to WW\to \ell \nu jj$; and associated production with a vector boson where the Higgs decays to $W$-boson pairs, $VH\to VWW\to \ell \nu jjjj$ production (where $V=W$ or $Z$). We observe good agreement between data and expected background. We test our method by measuring $WZ$ and $ZZ$ production with $Z\to b\overline{b}$ and find production rates consistent with the standard model prediction. For a Higgs boson mass of 125 GeV, we set a 95% C.L. upper limit on the production of a standard model Higgs boson of $5.8\times {\sigma }_{\mathrm{SM}}$, where ${\sigma }_{\mathrm{SM}}$ is the standard model Higgs boson production cross section, while the expected limit is $4.7\times {\sigma }_{\mathrm{SM}}$. We also interpret the data considering models with fourth generation fermions, or a fermiophobic Higgs boson.

Figure 1

Data-derived muon trigger correction to account for the resulting efficiency gain in moving from single muon and $\mathrm{\text{muon}}+\mathrm{\text{jets}}$ triggers to inclusive triggers as a function of $H_T$ for $|\eta |<1.0$, shown (a) for events with ${\mathrm{E̸}}_{\mathrm{T}}<50\mathrm{GeV}$ and (b) for events with ${\mathrm{E̸}}_{\mathrm{T}}\ge 50\mathrm{GeV}$. The black circles show the correction when the muon is in the region of $\varphi$ ($-2<\varphi <-1.2$) where there is a gap in the muon coverage for detector supports, and the red triangles show the correction elsewhere in $\varphi$.

Figure 2

Average of the $b$-identification discriminant outputs of each jet in events with two jets. Discrepancies in data-MC agreement are within the systematic uncertainties described in Sec. 10.

Figure 3

Distributions for all selected events with two jets of (a) transverse mass of the lepton-${\mathrm{E̸}}_{\mathrm{T}}$ system, (b) charged lepton $p_{\mathrm{T}}$, and (c) ${\mathrm{E̸}}_{\mathrm{T}}$. The signal is multiplied by 1000. Overflow events are added to the last bin.

Figure 4

Distributions for all selected events with two jets of (a) leading jet $p_{\mathrm{T}}$, (b) second-leading jet $p_{\mathrm{T}}$, and (c) $\Delta \mathcal{R}$ between the leading and second-leading jets. The signal is multiplied by 1000. Overflow events are added to the last bin.

Figure 5

Invariant mass of the leading and second-leading jets in events with two jets and (a) zero $b$ tags, (b) one loose $b$ tag, and (c) one tight $b$ tag. The signal is multiplied by 1000, 500, and 200, respectively. Overflow events are added to the last bin.

Figure 6

Invariant mass of the leading and second-leading jets in events with two jets and (a) two loose $b$ tags, (b) two medium $b$ tags, and (c) two tight $b$ tags. The signal is multiplied by 200, 50, and 50, respectively. Overflow events are added to the last bin.

Figure 7

The multivariate discriminant output for (a) ${\mathrm{MVA}}_{\mathrm{MJ}}(VH)$ and (b) ${\mathrm{MVA}}_{\mathrm{MJ}}(H\to VV)$, for all events. The signal is multiplied by 1000.

Figure 8

Distributions of some of the most significant inputs to the final discriminant in events with exactly two jets and either one tight $b$ tag, two loose $b$ tags, two medium $b$ tags, or two tight $b$ tags: (a) $p_T^W/(p_T^{\ell }+{\mathrm{E̸}}_{\mathrm{T}})$, shown for events with one tight $b$ tag; (b) $\max |\Delta \eta (\ell ,\{j_1\mathrm{\text{or}}{j}_2\})|$ [85], shown for events with two loose $b$ tags; (c) $q^{\ell }\times {\eta }^{\ell }$ [86], shown for events with two medium $b$ tags; (d) $\sum _{}^{}(p_T{)}^{\mathrm{VIS}}$ [87], shown for events with two tight $b$ tags. The signal is multiplied by 200, 200, 50, and 50, respectively. Overflow events are added to the last bin.

Figure 9

Distributions of some of the most significant inputs to the final discriminant in events with exactly three jets and either one tight $b$ tag, two loose $b$ tags, two medium $b$ tags, or two tight $b$ tags: (a) $\max |\Delta \eta (\ell ,\{j_1\mathrm{\text{or}}{j}_2\})|$ [85], shown for events with one tight $b$ tag; (b) $q^{\ell }\times {\eta }^{\ell }$ [86], shown for events with two loose $b$ tags; (c) aplanarity [88], shown for events with two medium $b$ tags; (d) $m_{\ell \nu j_2}$ [89, 90], shown for events with two tight $b$ tags. The signal is multiplied by 200, 200, 50, and 50, respectively. Overflow events are added to the last bin.

Figure 10

Distributions of the final discriminant output, after the maximum likelihood fit (described in Sec. 12), in events with exactly two jets and (a) one tight $b$ tag, (b) two loose $b$ tags, (c) two medium $b$ tags, and (d) two tight $b$ tags. The signal is multiplied by 100, 100, 20, and 20, respectively.

Figure 11

Distributions of the final discriminant output, after the maximum likelihood fit (described in Sec. 12), in events with exactly three jets and (a) one tight $b$ tag, (b) two loose $b$ tags, (c) two medium $b$ tags, and (d) two tight $b$ tags. The signal is multiplied by 100, 100, 20, and 20, respectively.

Figure 12

Distributions of the most significant inputs to the final multivariate discriminants for the two-jets zero and one loose $b$-tag channels: (a) $\Delta p_T(j_1,j_2)$, shown for events with zero $b$ tags for $M_H=125\mathrm{GeV}$; (b) $\sum _{i=1}^2{p}_{\mathrm{T}}^{j_i}$, shown for events with one loose $b$ tag for $M_H=125\mathrm{GeV}$; (c) $(M_{\ell \nu }-M_{j_{12}})/(M_{\ell \nu }+M_{j_{12}})$, shown for all tags for $M_H=165\mathrm{GeV}$. The signal is multiplied by 1000, 500, and 100, respectively. Overflow events are added to the last bin.

Figure 13

Distributions of the most significant inputs to the final multivariate discriminants for the three-jets zero and one loose $b$-tag channels: (a) $|\Delta \eta (W,\ell )|$ [90, 91], shown for events with zero $b$ tags for $M_H=125\mathrm{GeV}$; (b) $M_T^W$, shown for events with one loose $b$ tag for $M_H=125\mathrm{GeV}$; (c) $\sum _{}^{}(p_T{)}^{\mathrm{VIS}}$ [87], shown for all tags for $M_H=165\mathrm{GeV}$. The signal is multiplied by 1000, 500, and 100, respectively. Overflow events are added to the last bin.

Figure 14

Distributions of the final discriminant output, after the maximum likelihood fit (described in Sec. 12), for events in the following channels: (a) two jets, zero $b$ tags for $M_H=125\mathrm{GeV}$, (b) two jets, one loose $b$ tag for $M_H=125\mathrm{GeV}$, and (c) two jets, all tags $M_H=165\mathrm{GeV}$. The signal is multiplied by 500, 500, and 200, respectively.

Figure 15

Distributions of the final discriminant output, after the maximum likelihood fit (described in Sec. 12), for events in the following channels: (a) three jets, zero $b$ tags for $M_H=125\mathrm{GeV}$, (b) three jets, one loose $b$ tag for $M_H=125\mathrm{GeV}$, and (c) three jets, all tags $M_H=165\mathrm{GeV}$. The signal is multiplied by 500, 500, and 200, respectively.

Figure 16

Distributions of the most significant inputs to background-specific multivariate discriminants for the $\ge \mathrm{\text{four}}$-jet subchannels: (a) $\cos \theta (\ell {)}_{\ell \nu \mathrm{CM}}$ [90, 92], input to discriminant against $V+\mathrm{\text{jets}}$ backgrounds, shown for events with zero $b$ tags; (b) ${\mathcal{S}\mathcal{I}\mathcal{G}}_{\mathrm{jets}}(\ell )$ [93], input to discriminant against diboson backgrounds, shown for events with zero $b$ tags; (c) $\sum _{}^{}{p}_T(\ell ,{\mathrm{E̸}}_{\mathrm{T}},j_1,j_2,j_3,j_4)$, input to discriminant against top-quark backgrounds, shown for events with one loose $b$ tag. The $M_H=125\mathrm{GeV}$ signal is multiplied by 250 in (c) and by 500 in (a) and (b). Overflow events are added to the last bin.

Figure 17

Distributions of the output of background-specific multivariate discriminants for the $\ge \mathrm{\text{four}}$-jet subchannels: (a) discriminant against $V+\mathrm{\text{jets}}$ backgrounds, shown for events with zero $b$ tags; (b) discriminant against diboson backgrounds, shown for events with zero $b$ tags; (c) discriminant against top-quark backgrounds, shown for events with one loose $b$ tag. The $M_H=125\mathrm{GeV}$ signal is multiplied by 250 in (c) and by 500 in (a) and (b).

Figure 18

Distributions of the most significant inputs, other than background-specific multivariate discriminants, to the final multivariate discriminants for the $\ge$ four-jet subchannels: (a) $m_{j_{12}\ell }$, shown for events with zero $b$ tags; (b) $\Delta \mathcal{R}(\ell ,j_3)$, shown for events with one loose $b$ tag. The $M_H=125\mathrm{GeV}$ signal is multiplied by 500 in (a) and 250 in (b). Overflow events are added to the last bin.

Figure 19

Distributions of the final discriminant output, after the maximum likelihood fit (described in Sec. 12), at $M_H=125\mathrm{GeV}$ for the four-or-more-jets channels with (a) zero $b$ tags and (b) one loose $b$ tag. The $M_H=125\mathrm{GeV}$ signal is multiplied by 500.

Figure 20

Final MVA discriminant output shown for the expected diboson signal and background-subtracted data rebinned as a function of $\log (S/B)$, after the maximum likelihood fit, summed over $b$-tag channels. The error bars on data points represent the statistical uncertainty only. The post-fit systematic uncertainties are represented by the solid lines. The signal expectation is shown scaled to the best fit value. The inset gives an expanded view of the high-$\log (S/B)$ region.

Figure 21

The expected and observed log-likelihood ratios as functions of the hypothesized Higgs boson mass $M_H$ for the (a) electron and muon, two- and three-jet, one tight and two-$b$-tag channels; (b) electron and muon, two- and three-jet, zero and one loose $b$-tag channels; (c) electron and muon, four-or-more-jets, zero and one loose $b$-tag channels; (d) combination of all channels. The dashed red and black lines correspond to the median LLR of the $\mathrm{\text{signal}}+\mathrm{\text{background}}$ and background-only hypotheses, respectively. The solid line corresponds to the LLR obtained from the data, and the shaded regions are the $\pm 1$ s.d. and $\pm 2$ s.d. values for the background-only hypothesis.

Figure 22

The MVA discriminant output distribution minus the total background expectation for $M_H=125\mathrm{GeV}$ rebinned as a function of $\log (S/B)$. The post-fit uncertainties are represented by the solid lines. The signal expectation is shown scaled to the best-fit value. The inset gives an expanded view of the high-$\log (S/B)$ region.

Figure 23

The expected and observed 95% C.L. upper limits on SM Higgs boson production for the (a) electron and muon, two- and three-jet, one tight and two-$b$-tag channels; (b) electron and muon, two- and three-jet, zero and one loose $b$-tag channels ($M_H\le 150\mathrm{GeV}$) and pretag channels ($M_H\ge 155\mathrm{GeV}$); (c) electron and muon, four-or-more-jets, zero and one loose $b$-tag channels; (d) combination of all channels. The limits are presented as ratios to the expected SM prediction. The dashed line corresponds to the expected limit, and the solid line corresponds to the limit observed in data. The shaded regions are the $\pm 1$ s.d. and $\pm 2$ s.d. values for the expected limit.

Figure 24

The expected and observed 95% C.L. upper limits on $\sigma (gg\to H)\times \mathcal{B}(H\to WW)$ compared to the prediction from the fourth generation fermion model.

Figure 25

The expected and observed 95% C.L. upper limits on fermiophobic Higgs boson production.