Measurement of the electron charge asymmetry in $p\overline{p}\to W+X\to e\nu +X$ decays in $p\overline{p}$ collisions at $\sqrt{s}=1.96\mathrm{TeV}$

We present a measurement of the electron charge asymmetry in $p\overline{p}\to W+X\to e\nu +X$ events at a center-of-mass energy of 1.96 TeV, using data corresponding to $9.7{\mathrm{fb}}^{-1}$ of integrated luminosity collected with the D0 detector at the Fermilab Tevatron Collider. The asymmetry is measured as a function of the electron pseudorapidity and is presented in five kinematic bins based on the electron transverse energy and the missing transverse energy in the event. The measured asymmetry is compared with next-to-leading-order predictions in perturbative quantum chromodynamics and provides accurate information for the determination of parton distribution functions of the proton. This is the most precise lepton charge asymmetry measurement to date.

Figure 1

(a) The $W$ boson rapidity ($y_W$) and electron pseudorapidity (${\eta }^e$) distributions in $p\overline{p}$ collisions. (b) The charge asymmetry for the $W$ boson and the decay electron. The electron asymmetry has a “turnover” due to the convolution of the $W$ boson asymmetry and the $V-A$ structure of the $W$ boson decay. These predictions were obtained using the MC event generator resbos [5] with the CTEQ6.6 [6] central PDF set, using the kinematic cuts $p_T^e>25\mathrm{GeV}$ and $p_T^{\nu }>25\mathrm{GeV}$.

Figure 2

Cross sectional view of the D0 central tracking detector in the $x\text{−}z$ plane.

Figure 3

Schematic view of a portion of the D0 calorimeters showing the transverse and longitudinal segmentation patterns. The rays indicate the pseudorapidity measured from the center of the detector (${\eta }_{\det }$).

Figure 4

Comparisons of the electron ${\eta }_{\det }^e$ distributions between data and the sum of signal and background predictions for selected $Z$ boson events, (a) event distribution and (c) value of ${\chi }_i$ for each bin between data and the MC predictions before applying the EMID correction, (b) event distribution and (d) value of ${\chi }_i$ for each bin after applying the EMID correction. ${\chi }_i=\mathrm{\Delta }{N}_i/{\sigma }_{N_i}$, $\mathrm{\Delta }{N}_i$ is the difference between the number of data and that of the MC prediction, and ${\sigma }_{N_i}$ is the statistical uncertainty in each $\eta$ bin.

Figure 5

Data and MC track matching efficiency comparison after $K_{\mathrm{eff}}^{q,p}$ correction, as a function of (a) electron $E_T$ and (b) electron $\eta$, for $e^{+}$ with positive solenoid polarity. There is an efficiency drop around 55 GeV, which comes from the tag-and-probe method, since in $Z\to ee$ events, when the tagged electron has high $E_T$, the other electron in the event is soft, resulting in inefficiency in the EMID.

Figure 6

The fitted mass value of CC-EC events (events with one electron in the CC and the other in the EC) in ${\eta }_{\det }^e$ bins for (a) data and (b) MC $Z\to ee$ events. The open blue points are the $Z$ boson mass peak values before applying the electron energy calibration, and the solid red points represent the peak values after applying the electron energy calibration.

Figure 7

Definitions of the $\eta$ and $\xi$ axes in $Z\to ee$ events, and the dielectron and the hadronic recoil system projections in these axes. The $\eta \text{−}\xi$ plane is transverse to the beam.

Figure 8

Charge misidentification probability as a function of ${\eta }^e$. The blue triangles represent the measured data charge misidentification probability, and the red dots represent that of the MC.

Figure 9

Comparisons of electron (a) $E_T^e$, (b) ${\cancel{E}}_T$, (c) ${\eta }_{\det }^e$, and (d) $W$ boson $p_T$ between data and the sum of signal and background predictions for selected $W$ boson events. Systematic uncertainties are not shown.

Figure 10

(a) Closure test I of the unfolding method for the kinematic bin $E_T^e>25\mathrm{GeV}$, ${\cancel{E}}_T>25\mathrm{GeV}$, using half of the MC sample as input for the unfolding procedure and the other half as pseudodata. The solid red points are the pythia generator-level electron asymmetries and the open blue points are the unfolded asymmetries. The asymmetries are shown after $CP$ folding. (b) $\chi$ distribution between predicted asymmetry and unfolded asymmetry, where ${\chi }_i=\mathrm{\Delta }{A}_i/{\sigma }_i$, $\mathrm{\Delta }{A}_i$ is the difference between the generator-level asymmetry and the unfolded asymmetry and ${\sigma }_i$ is the statistical uncertainty in bin $i$.

Figure 11

(a) shows different input asymmetries used in the second closure test for the kinematic bin $E_T^e>25\mathrm{GeV}$, ${\cancel{E}}_T>25\mathrm{GeV}$. The solid black points represent the default asymmetry distribution generated using the pythia generator with CTEQ6.1 L PDFs, the open red points represent the enhanced input asymmetry, and the open blue squares point represent the suppressed input asymmetry. (b),(c) show closure test II of the unfolding method using half of the MC sample as input for the unfolding procedure and the other half of the MC as pseudodata, where the generator-level asymmetries are enhanced or suppressed. A reweighting factor [$f=1-0.05\times {\eta }^e$ for (b), and $f=1+0.05\times {\eta }^e$ for (c)] has been applied to the number of electrons to ensure the generator-level asymmetries are far from the default values. The solid red points are generator-level electron asymmetries; the open blue points are the unfolded asymmetries. The asymmetries are shown after $CP$ folding.

Figure 12

Comparisons of the electron $E_T$ distribution between data and the sum of signal and background predictions for selected (a) $W^{+}$ and (b) $W^{-}$ events in the range $-2.7<{\eta }_{\det }^e<-2.4$. This bin is chosen as the one showing the poorest agreement. In different rapidity regions, the values of $x$ for $u$ and $d$ quarks are different; thus the $W^{+}$ and $W^{-}$ distributions are different from each other. The data uncertainty only represents the statistical uncertainty, and the bands represent the systematic uncertainty on the signal plus backgrounds, without any uncertainty from MC modelings.

Figure 13

The lepton charge asymmetry distribution after $CP$ folding with symmetric kinematic cuts $E_T^e>25\mathrm{GeV}$, ${\cancel{E}}_T>25\mathrm{GeV}$. (a) Comparison between the measured asymmetry and predictions and (b) the differences between the data and MC predictions and the predicted central value from mc@nlo using the NNPDF2.3 PDF set. The black dots show the measured electron charge asymmetry, with the horizontal bars showing the statistical uncertainty and the vertical lines showing the total uncertainty. The red triangles show the published D0 muon charge asymmetry [12]. The red dashed lines and cyan bands are the central value and uncertainty band from mc@nlo, respectively, using the NNPDF2.3 PDF sets. The blue dotted lines show the prediction from mc@nlo using the MSTW2008NLO central PDF set, and the green dot-dashed lines show the prediction from resbos using the CTEQ6.6 central PDF set.

Figure 14

The electron charge asymmetry distribution after $CP$ folding with symmetric kinematic cuts $25<E_T^e<35\mathrm{GeV}$, $25<{\cancel{E}}_T<35\mathrm{GeV}$. (a) Comparison between the measured asymmetry and predictions and (b) the differences between the data and MC predictions and the predicted central value from mc@nlo using the NNPDF2.3 PDF set. The black dots show the measured electron charge asymmetry, with the horizontal bars showing statistical uncertainty and the vertical lines showing the total uncertainty. The red dashed lines and cyan bands are the central value and uncertainty band from mc@nlo, respectively, using the NNPDF2.3 PDF sets. The blue dotted lines show the prediction from mc@nlo using the MSTW2008NLO central PDF set, and the green dot-dashed lines show the prediction from resbos using the CTEQ6.6 central PDF set.

Figure 15

The lepton charge asymmetry distribution after $CP$ folding with symmetric kinematic cuts $E_T^e>35\mathrm{GeV}$, ${\cancel{E}}_T>35\mathrm{GeV}$. (a) Comparison between the measured asymmetry and predictions and (b) the differences between the data and MC predictions and the predicted central value from mc@nlo using the NNPDF2.3 PDF set. The black dots show the measured electron charge asymmetry, with the horizontal bars showing statistical uncertainty and the vertical lines showing the total uncertainty. The red triangles show the published D0 muon charge asymmetry [12]. The red dashed lines and cyan bands are the central value and uncertainty band from mc@nlo, respectively, using the NNPDF2.3 PDF sets. The blue dotted lines show the prediction from mc@nlo using the MSTW2008NLO central PDF set, and the green dot-dashed lines show the prediction from resbos using the CTEQ6.6 central PDF set.

Figure 16

The electron charge asymmetry distribution after $CP$ folding with asymmetric kinematic cuts $25<E_T^e<35\mathrm{GeV}$, ${\cancel{E}}_T>25\mathrm{GeV}$. (a) Comparison between the measured asymmetry and predictions and (b) the differences between the data and MC predictions and the predicted central value from mc@nlo using the NNPDF2.3 PDF set. The black dots show the measured electron charge asymmetry, with the horizontal bars showing statistical uncertainty and the vertical lines showing the total uncertainty. The red dashed lines and cyan bands are the central value and uncertainty band from mc@nlo, respectively, using the NNPDF2.3 PDF sets. The blue dotted lines show the prediction from mc@nlo using the MSTW2008NLO central PDF set, and the green dot-dashed lines show the prediction from resbos using the CTEQ6.6 central PDF set.

Figure 17

The electron charge asymmetry distribution after $CP$ folding with asymmetric kinematic cuts $E_T^e>35\mathrm{GeV}$, ${\cancel{E}}_T>25\mathrm{GeV}$. (a) Comparison between the measured asymmetry and predictions and (b) the differences between the data and MC predictions and the predicted central value from mc@nlo using the NNPDF2.3 PDF set. The black dots show the measured electron charge asymmetry, with the horizontal bars showing statistical uncertainty and the vertical lines showing the total uncertainty. The red dashed lines and cyan bands are the central value and uncertainty band from mc@nlo, respectively, using the NNPDF2.3 PDF sets. The blue dotted lines show the prediction from mc@nlo using the MSTW2008NLO central PDF set, and the green dot-dashed lines show the prediction from resbos using the CTEQ6.6 central PDF set.