
arXiv:2210.13930v1 [hep-ex] 25 Oct 2022
Revision of results on Υ(1S), Υ(2S), and Υ(3S) masses
A.G. Shamova, O.L. Rezanovaa,b
aBudker Institute of Nuclear Physics, 11, akademika Lavrentieva prospect, Novosibirsk, 630090, Russia
bNovosibirsk State University, 1, Pirogova street, Novosibirsk, 630090, Russia
Abstract
We have reconsidered the results on the masses of the narrow bottomonium states Υ(1S)–Υ(3S) obtained in 1982–1986 at CESR,
DORIS and VEPP-4 colliders in order to fix shortcomings of the mass determination procedures. For experiments at CESR and
DORIS this includes the incorrect accounting of the radiative corrections and usage of the electron mass value revised in 1986. In
analyses of all experiments the interference of the resonance production and the nonresonant process was ignored. The corrected
mass values for five experiments are suggested. The corrections vary from 0.1 to 0.4 MeV. The discrepancy between CESR and
VEPP-4 results on Υ(1S) mass has been reduced from 3.3 to 1.8 standard deviations.
Introduction
The new experiment on the high precision measurement
of the Υ–meson mass has been planned at the VEPP-4M col-
lider [1] with the KEDR detector [2]. The resonant depolariza-
tion method [3, 4] will be used for the beam energy determi-
nation. At the moment the laser polarimeter is under develop-
ment [5] and the test scan of the Υ(1S) has been performed [6].
In this context it is important to overcome known drawbacks
in the analyses of the preceding experiments and correct its re-
sults.
The mass of Υ(1S) was measured by the MD-1 detector at
VEPP-4 [7, 8] and CUSB detector at the collider CESR with
the accuracy of about 0.1 MeV [9]. With lower accuracy of
0.4-0.5 MeV the mass of Υ(2S) was measured by ARGUS and
Crystal Ball detectors at DORIS [10] and by MD-1 [11]. The
mass of Υ(3S) was determined with 0.5 MeV uncertainty by
MD-1 only [11]. In all these experiments the mass values were
obtained by fitting the inclusive hadronic cross section as func-
tion of the c.m. energy. The beam energy was determined using
resonant depolarization method.
In 2000 the results of the mass measurements at VEPP-4
were corrected [12] to the shift of the electron mass value oc-
curred in 1986 [13]. The results from CESR and DORIS stayed
intact.
Another problem to solve is accounting of the radiative cor-
rection in experiments [9] and [10] according to the work [14]
containing the mistake [15]. Despite to the existence of cor-
rect studies of the narrow resonance production since 1975, the
incorrect resonance shape from Ref. [14] was employed for de-
termination of leptonic widths and masses of ψ– and Υ–states
in many experiments. Concerning leptonic widths the problem
was solved in Ref. [16], the corrected values were included in
PDG tables. However, the masses of Υstates [9] and [10] were
not corrected neither in Ref. [16] nor in Ref. [12] where radia-
tive corrections were fixed up for J/ψ mass measurement by the
OLYA detector in 1980 [17]. The MD-1 experiments on masses
of upsilon states [7, 8, 11, 18] were performed with proper ra-
diative correction accounting.
Besides, there is a mistake in the calculation of the reso-
nance curve in Ref. [9], that will be discussed in details below.
The common drawback of all measurements of ψ– and Υ–
state masses mentioned above is ignoring of the interference
between resonant and nonresonant contributions to the hadron
production. First time it was accounted in the J/ψ– and ψ(2S)–
mass measurement in the experiment [19] and was discussed in
details in Ref. [20].
In all experiments under discussion except [9], the depen-
dence of the hadronic cross section on c.m. energy was not pub-
lished. In Refs. [12, 16], in order to correct the resonance lep-
tonic width [16] and mass [12], the equidistant data points were
simulated using the published values of the resonance curve pa-
rameters. Then two fits were performed with the correct fitting
function and that of the published paper. The variation of the
resonance parameter was added to its published value. This
method uses published values of parameters, biased by the in-
correct fit, and does not account for the specific layout of energy
points.
In contrast with that, in this work we obtained coordinates
of data points from the plots in electronic versions of publi-
cations using the graphical editor and converted them to the
physical quantities. Such data were not absolutely reliable thus
we shifted published values as described above. In Ref. [9] the
measured values of cross section and the energy were published
thus we could just refit the CUSB data.
In the next sections we describe the necessary corrections,
discuss data published by CUSB [9], and then obtain corrected
mass values for three resonances from five experiments.
1. Change of the electron mass value
As it was mentioned above, the experiments cited did em-
ploy the resonant depolarization method for the beam energy
determination. In this method the measured ratio of the spin
Preprint submitted to Elsevier October 26, 2022