In the Standard Model all leptons are equal. The W and Z bosons as well as the photon couple equally to electrons (e), muons (μ) and taus (τ), a rule that is known as lepton universality. The only exception is the Higgs boson, that couples according to mass. Lepton universality was tested extensively, for instance at LEP or in semileptonic decays of hadrons containing b and c quarks.

However what applies to the Standard Model does not necessarily apply to potential New Physics contributions. Rare decays of beauty hadrons involving loops (also known as penguins) are particularly sensitive to such contributions as the Standard Model contribution is very suppressed. Typical examples are B⁺→K⁺μ⁺μ⁻ and B⁺→K⁺e⁺e⁻ decays, which have branching fractions of one in a million. Is the rate of these decays the same as it should? LHCb went on measuring this with data from LHC Run 1 and got 25% fewer muons than expected. The ratio (R_K) of the branching fraction of B⁺→K⁺μ⁺μ⁻ and B⁺→K⁺e⁺e⁻ decays was found to be 0.745, deviating by 2.5 standard deviations from the expectation of unity.

A 2.6σ deviation is not significant enough to call it an observation, not even an evidence, but the result raised a few eyebrows. It started becoming interesting when LHCb repeated the measurement with neutral B mesons and measured a ratio (R_K*)  of B⁰→K^*0μ⁺μ⁻ and B⁰→K^*0e⁺e⁻ of 0.7, again 2.5σ below the Standard Model expectation. Recently LHCb updated its R_K measurement, adding 2016 data, to find an increased value of 0.85, which thanks to smaller uncertainties is still in 2.5σ discrepancy with the Standard Model.

Enter Belle, the B factory experiment at KEK in Japan. A B factory is an electron-positron collider sitting at the Υ(4S) resonance, the lightest that produces a pair of B mesons. Such a setup has the advantage of producing very clean collisions, as only two B mesons are produced. The price to pay is a much reduced rate of B mesons compared with the enormous rates of the LHC. Recently Belle reported several measurements of these two ratios that are all consistent with, but slightly below one. These measurements are also all consistent with those of LHCb.

Figure: Measurements from LHCb, Belle, and BaBar (not mentioned in the text) of (left) RK and (right) RK*. The measurements are made in bins of dilepton mass squared.

The overall picture is still inconclusive, but that doesn’t prevent theorists from designing models that could explain these anomalies. Not only lepton universality in these decays needs to be accommodated: There are also hints of lepton non-universality in decays to tau leptons, and unexpected angular distributions in B⁰→K^*0μ⁺μ⁻. These anomalies hint toward a new vector coupling, or a vector-axial coupling (just like the weak interaction). Are there heavy Z’ and W’ bosons, or even leptoquarks waiting to be discovered by the LHC?

To answer these questions we need more data that could confirm if this is a fluke in statistics or a serious flaw in the Standard Model pointing to the existence of new physics. LHCb has yet to update the measurements with the full Run 2 data set. Moreover the ongoing upgrades of the various sub detectors will and the foreseen upgrade of the LHC luminosity will shed more light/increase the number/volume of available data to help us get answers. At the same time Belle II, the successor of Belle, is now operational in Japan and plans to record a dataset nearly 100 times larger than that of Belle.