The ratios of the B meson production fractions have been measured for the first time in the CMS experiment.

B mesons are composite particles consisting of a bottom antiquark and another type of quark, such as an up, down, or strange quark, for B⁺, B⁰, and B_s meson, respectively. Since the era of the Large Electron–Positron Collider (LEP) at CERN, the probability of a bottom quark to hadronize into a specific type of B meson—a quantity known as the production fraction—has been studied in detail. The information is crucial for measurements of B meson decays, especially rare processes like B_s → μμ, where knowing the B_s production fraction is necessary to determine its branching fraction.

However, in practice, the B_s production fraction is not directly measurable in the experiment. One can think of this as a vending machine that randomly dispenses one of three drinks (representing B⁺, B⁰, and B_s mesons) based on some internal probability settings. Since these probabilities are inaccessible to us, we can only carefully measure the relative frequencies of the outcomes to infer the underlying probabilities. Similarly, B meson production studies rely on measuring ratios between different types of B mesons, known as production fraction ratios (PFRs), to indirectly determine their underlying production probabilities.

Figure 1: An interactive event display of B meson production at CMS. B and anti-B mesons are produced in a back-to-back orientation. One B meson decays into a final state that includes a muon, while the other decays into lighter hadrons.

In recent years, measurements of PFRs of B meson by the LHCb experiment have consistently reported a dependence of PFRs on the transverse momentum (p_T) of B mesons, a behavior that remains theoretically puzzling. This dependence implies that PFRs are not universal constants, and thus the PFR values measured by other experiments, such as LHCb, are not directly applicable to CMS measurements, as they cover a different p_T range. This challenge has motivated direct measurements of PFRs at CMS taking advantage of a new data-taking technique called data parking in 2018, which has enabled CMS to collect an unprecedented number of b hadrons.

Figure 2: The streams of collision events in CMS include the standard data stream, along with parking and scouting streams, which enable the collection of additional data with minimal impact on standard data-taking.

The CMS data-parking strategy is designed to collect a larger data set than the standard data stream. The idea behind data parking is to store the data intermediately and to perform the processing of the data only when sufficient computational resources are available. Figure 2 illustrates the different data-taking strategies at CMS, which includes data parking. This approach optimizes resource usage while enabling CMS to collect B physics events at a much higher rate than would otherwise be possible.

Figure 3: Examples of fits in the B_s candidate mass distribution (corrected by the D_s meson mass) used to extract the signal in the hadronic open-charm channels. The lower panels illustrate the deviation of data from the fit function, normalized by the statistical uncertainty in each bin.

The recent measurement of PFRs using the parking data marks the first determination of PFRs at CMS using open-charm hadronic decays (B meson decays through D mesons, which contain a charm quark). The large number of b hadron events in the parking data set allows CMS to achieve a statistically precise PFR measurement. A candidate for this decay is reconstructed using hadron tracks collected in the CMS detector, as shown in Fig. 1. The signal appears as a sharp peak in Fig. 3, corresponding to the B meson mass.

Additionally, this measurement tests the dependence of PFRs on p_T by using charmonium decays (B meson decays through the J/ψ meson). While charmonium decays offer cleaner signals at CMS, these measurements only provide relative PFR values since there is no reliable theoretical prediction. To address this issue, this CMS study also determines the conversion factor that translates them from relative to absolute PFRs.

Figure 4: The production fraction ratio between B_s and B⁰ (f_s/f_d) is presented as a function of p_T. The relative PFRs in charmonium decays are normalized to absolute PFRs using the conversion factors obtained in this study. PFR measurements from LEP are overlaid for comparison, along with the latest trend of f_s/f_d observed by LHCb at 13 TeV.

Figure 4 presents the measured PFR as a function of p_T for both hadronic open-charm and charmonium decays. The result indicates that the extrapolated trend from LHCb measurements at 13 TeV aligns with CMS results at low p_T, but a discrepancy emerges at high p_T. This flattening of PFR at high p_T has also been observed in a previous CMS study on relative PFRs

Further studies on measuring PFR at CMS continue to be of great interest to better understand the low p_T dependence of PFR—specifically, what causes them to depend on the B meson momentum and what their exact shapes are as a function of p_T. With the ongoing Run 3 data taking at the LHC, CMS has continued to collect an even larger amount of B physics data using the data-parking strategy. This additional data is anticipated to provide a more precise determination of the PFR and offers new insights into the observed low p_T dependence.

Read more about these results:

• CMS Physics Analysis Summary (BPH-21-007): "Coming soon!"

• @CMSExperiment on social media: LinkedIn - Facebook - Twitter - Instagram

• All CMS physics briefings

• All CMS publications

• All CMS preliminary results

• Search all CMS results