---
title: 'X_b: Hidden-Bottom Partner Analysis'
url: https://www.emergentmind.com/topics/x_b
type: topic
---

# X_b: Hidden-Bottom Partner Analysis

In hadron spectroscopy, \(X_b\) most commonly denotes a hypothetical hidden-bottom counterpart of \(X(3872)\): an isoscalar \(J^{PC}=1^{++}\) structure tied to the \(B\bar B^*+B^*\bar B\) threshold and discussed variously as a \(B\bar B^*\) hadronic molecule, a mixed molecule–quarkonium state, or a threshold phenomenon shaped by coupled channels and form factors [2107.08451] [1810.03452]. The notation is not fully uniform: some works use \(X_b\) for the distinct open-flavor candidate \(X_b(5568)\), while in other subfields \(x_B\) denotes the Bjorken scaling variable or a scaled \(B\)-hadron energy [1603.00290] [2402.08199].

## 1. Definition, quantum numbers, and threshold setting

The dominant use of \(X_b\) in the hadron-spectroscopy literature is the bottomonium-sector partner of \(X(3872)\). In that usage, the relevant open-bottom thresholds are the \(B\bar B^*\) threshold at \(10604.44~\mathrm{MeV}\) and the \(B^*\bar B^*\) threshold at \(10650.20~\mathrm{MeV}\), and the expected quantum numbers are \(J^{PC}=1^{++}\) [2107.08451]. A closely related formulation describes \(X_b\) as an isoscalar \(J^{PC}=1^{++}\) structure tied to the \(B\bar B^*+B^*\bar B\) threshold, analogous to the way \(X(3872)\) is tied to \(D\bar D^*\) and \(\chi_{c1}(2P)\) [1810.03452].

Heavy Flavor Symmetry motivates such a state near \(M_B+M_{B^*}\sim 10.6~\mathrm{GeV}\), with smaller isospin breaking than in the charm sector because charged-neutral \(B^{(*)}\) mass splittings are smaller than for \(D^{(*)}\) mesons [2107.08451]. The small threshold splitting is central to the phenomenology. In one formulation, the neutral and charged thresholds are \(M(B^0\bar B^{*0})=(10604.8\pm0.4)\ \mathrm{MeV}\) and \(M(B^-B^{*+})=(10604.5\pm0.4)\ \mathrm{MeV}\), so a bound \(X_b\) is expected to remain almost purely isoscalar; this contrasts with \(X(3872)\), whose proximity to the neutral \(D\bar D^*\) threshold leads to unusually large isospin violation [1410.7729].

The hidden-bottom \(X_b\) literature does not converge on a single ontological category. Some papers treat it as a predominantly \(B\bar B^*\) molecular bound state; others emphasize strong mixing with nearby \(\chi_{b1}(3P)\) or \(\chi_{b1}(4P)\) quarkonium configurations; still others argue that the observable near-threshold enhancement need not coincide with a nearby pole and can be dominated by energy-dependent form factors [1512.07528] [1410.7729] [1810.03452]. A consistent theme is that the \(X_b\) problem is inseparable from threshold dynamics.

## 2. Symmetry arguments, quarkonium mixing, and representative spectra

Heavy Quark Spin Symmetry and Heavy Flavor Symmetry provide the initial organizing framework, but the more detailed calculations surveyed here repeatedly stress that symmetry alone is not predictive enough once nearby quarkonium levels and open-flavor channels are included [2107.08451]. In the coupled-channel analysis of “Symmetries, partners and thresholds: the case of the \(X_b\),” the relevant bare \(b\bar b\) levels are \(3^3P_1: 10512.76~\mathrm{MeV}\) and \(4^3P_1: 10737.27~\mathrm{MeV}\), positioned respectively below and above the \(B\bar B^*\) threshold; the sign of the induced interaction therefore depends on where the threshold region lies relative to the bare state through
\[
V_{\beta'\beta}^{\rm eff}(P',P) = \sum_\alpha \frac{h_{\beta'\alpha}(P')\,h_{\alpha\beta}(P)}{E-M_\alpha},
\]
so attraction or repulsion is not fixed by symmetry alone [2107.08451].

Karliner and Rosner emphasize a closely related but more qualitative point: a near-threshold \(B\bar B^*+\) c.c. molecule and the conventional \(\chi_{b1}(3P)\) are expected to be nearby in mass and should therefore mix strongly, just as \(X(3872)\) is plausibly a mixture of a \(\bar D D^*\) molecule and \(\chi_{c1}(2P)\) [1410.7729]. In that picture, observed \(\chi_{b1}(3P)\)-like signals may already contain an \(X_b\) component.

| Framework | Representative \(X_b\) result | Structural reading |
|---|---|---|
| Mixed molecule–quarkonium near \(\chi_{b1}(3P)\) | \(M(X_b)=10562~\mathrm{MeV}\), \(10585~\mathrm{MeV}/c^2\), or \(10604~\mathrm{MeV}/c^2\) quoted as expectations [1410.7729] | Near-threshold \(B\bar B^*+\) c.c. state mixed with \(\chi_{b1}(3P)\) |
| Coupled channels with HQSS/HFS and bare \(b\bar b\) states | \(M_{X_b}=10599.3^{+0.01}_{-0.02}\,\mathrm{MeV}\), \(\Gamma_{X_b}=0.51\pm0.01\,\mathrm{MeV}\) [2107.08451] | Predominantly \(B\bar B^*\) molecular \(1^{++}\) state |
| EFT line-shape analysis | Structure controlled by binding energy \(B\) and elementary probability \(Z\) [1512.07528] | Molecule, compact state, or mixture distinguished by near-threshold line shape |
| Extended Friedrichs scheme | Narrow peak around \(10615~\mathrm{MeV}\) with virtual state at \(10593~\mathrm{MeV}\) [1810.03452] | Observable enhancement dominated mainly by form factor rather than pole |

The most explicit molecular result among these is the coupled-channel calculation that finds a \(1^{++}\) state at \(10599.3^{+0.01}_{-0.02}\,\mathrm{MeV}\) with width \(0.51\pm0.01\,\mathrm{MeV}\), overwhelmingly molecular in composition:
\[
{\cal P}_{B\bar B^*}=93.65^{+0.08}_{-0.07}\%, \qquad
{\cal P}_{B^*\bar B^*}=5.84^{+0.01}_{-0.04}\%,
\]
with only tiny compact components,
\[
{\cal P}_{b\bar b(3^3P_1)}=0.33\pm0.02\%, \qquad
{\cal P}_{b\bar b(4^3P_1)}=0.14^{+0.05}_{-0.03}\% .
\]
That calculation attributes the binding primarily to coupled-channel dynamics, especially the nearby \(B^*\bar B^*\) threshold, rather than to a simple one-channel extrapolation from \(X(3872)\) [2107.08451].

## 3. Threshold dynamics, compositeness, and line shapes

A distinct line of work treats the \(X_b\) question as a near-threshold line-shape problem. In the EFT analysis “Structure of \(X_b\) from line shape analysis,” the physical state is described by a bare state coupled to a two-body continuum. The elastic amplitude after resumming bubble diagrams is
\[
\mathcal A = -\frac{g_0^2}{ E+B_0-g_0^2\frac{\mu}{2\pi}\sqrt{-2\mu E-i\epsilon} },
\]
with a physical bound-state pole at \(E=-B\) [1512.07528]. Weinberg’s compositeness parameter \(Z\) is then introduced through
\[
g^2 = \frac{2\pi\sqrt{2\mu B}}{\mu^2}(1-Z),
\]
where \(Z=1\) corresponds to a purely elementary state, \(Z=0\) to a purely molecular state, and \(0<Z<1\) to a mixed state [1512.07528].

In that framework, the near-threshold \(B^*\bar B\) line shape is sensitive to both the binding energy \(B\) and the compact-state probability \(Z\). For production through a compact short-distance source,
\[
\mathcal M \propto \sqrt{Z(1-Z)},
\]
so both a compact production component and a molecular coupling to \(B^*\bar B\) are required [1512.07528]. A pure molecular limit \(Z=0\) is not described by the same production mechanism; instead it requires direct \(B^*\bar B\) production plus rescattering through the molecular pole. This suggests that line-shape measurements can separate a predominantly molecular scenario from a mixed compact-plus-molecular scenario.

The most radical threshold interpretation among the surveyed works is the extended Friedrichs analysis of “Does the bottomonium counterpart of \(X(3872)\) exist?” That study predicts three nearby structures: a virtual-state pole
\[
z_v = 10593~\text{MeV},
\]
a narrow dressed \(\chi_{b1}(4P)\) resonance
\[
z_{R1}=10771\pm 3 i~\text{MeV},
\]
and a broad dynamically generated resonance
\[
z_{R2}=10672\pm 39 i~\text{MeV}.
\]
Yet the visible narrow enhancement in \(B\bar B^*\) scattering appears around \(10615~\mathrm{MeV}\), just above threshold, and is argued to be contributed mainly by the residue function \(f_i(E)f_f^*(E)\), not mainly by the nearby virtual-state pole in \(1/\eta\) [1810.03452]. The formal distinction is explicit in the \(S\)-matrix,
\[
S_{fi}(E,E')=\delta(E-E')\Big(\delta_{fi}-2\pi i \frac{ f_i(E){f_f}^*(E)}{\eta^+(E)}\Big),
\]
which separates pole information through \(1/\eta^+(E)\) from channel-dependent structure through the form factor [1810.03452].

That analysis gives a methodological warning with broader relevance: some threshold peaks may be generated by structures in form factors rather than by genuine nearby poles. A plausible implication is that the term \(X_b\) does not always identify a single spectroscopic object; in some models it identifies an experimentally visible threshold phenomenon whose line shape encodes both pole dynamics and nodal structure of high-radial-excitation wave functions.

## 4. Decay phenomenology and production mechanisms

A recurrent conclusion of the hidden-bottom \(X_b\) literature is that the discovery channel need not be the direct bottom analogue of \(X(3872)\to J/\psi\,\pi^+\pi^-\). Because \(X_b\) is expected to be almost purely isoscalar and because the charged-neutral \(B^{(*)}\) mass differences are small compared with the assumed binding energy, the isospin-violating channel \(X_b\to \Upsilon(nS)\pi^+\pi^-\) is expected to be greatly or highly suppressed [1502.02936] [1410.7729]. This is one reason null searches in \(\Upsilon(1S)\pi^+\pi^-\) have not been taken as decisive evidence against \(X_b\).

Several alternative decay modes have been proposed. In the effective-Lagrangian calculation of “Hunting for the \(X_b\) via hidden bottomonium decays,” the isospin-conserving rescattering decay
\[
X_b\to \Upsilon(1S)\omega
\]
has a partial width of about tens of keV, and if the total width is smaller than a few MeV like \(X(3872)\), the corresponding branching ratio may reach orders of \(10^{-2}\) [1502.02936]. Radiative decays
\[
X_b\to \gamma \Upsilon(nS),\qquad n=1,2,3,
\]
were predicted in a separate heavy-quark-symmetry loop analysis to have partial widths about \(1\) keV, with \(\gamma\Upsilon(2S)\) often the largest channel in the benchmark calculations [1402.6463].

More recently, the hidden-bottomonium transitions
\[
X_b\to \pi\pi\chi_{bJ}
\]
have been argued to be especially favorable. In the HH\(\chi\)PT study of \(X_b\to \pi\pi\chi_{bJ}\), the calculated partial width of \(X_b\to \pi\pi\chi_{b1}\) is about tens of keV and is \(1\sim2\) order(s) of magnitude larger than those of \(X_b\to \pi\pi\chi_{b2}\) and \(X_b\to \pi\pi\chi_{b0}\); if the total width is smaller than a few MeV, the branching ratio \(X_b\to \pi\pi\chi_{b1}\) may reach orders of \(10^{-2}\) [2311.15527]. By contrast, the isospin-breaking channels \(X_b\to \pi^0\chi_{bJ}\) are strongly suppressed once the charged and neutral loop cancellation appropriate to an isoscalar molecule is imposed [2311.15527].

The production literature is similarly channel-dependent. Radiative production from \(\Upsilon(5S,6S)\) through \(B^{(*)}\bar B^{(*)}\) and \(B_1^\prime\bar B^{(*)}\) loops yields branching ratios of order \(10^{-7}\sim10^{-6}\) [2301.07365]. A more favorable scenario is proposed for \(\Upsilon(10753)\), treated as an \(S\)-\(D\) mixed state: including \(P\)-wave \(B_1^{(\prime)}\) loops leads to a predicted radiative width \(\Gamma[\Upsilon(10753)\to\gamma X_b]\sim 13~\mathrm{keV}\) at the benchmark \(\epsilon_X=50\) MeV, corresponding to a branching fraction of \(10^{-4}\), and motivates searches in
\[
e^+e^-\to\gamma X_b,\qquad X_b\to\pi\pi\chi_{b1},
\]
near \(\sqrt s=10.754~\mathrm{GeV}\) [2403.01676]. Open-bottom \(B\bar B^*\) itself is also repeatedly emphasized as a key channel, and in the extended Friedrichs picture it is explicitly identified as the most promising place to observe the predicted threshold structure [1810.03452].

## 5. Experimental searches and current constraints

Dedicated searches have not produced a confirmed hidden-bottom \(X_b\) signal. CMS searched for a narrow state in
\[
X_b\to \Upsilon(1S)\pi^+\pi^-,\qquad \Upsilon(1S)\to \mu^+\mu^-,
\]
using \(20.7~\mathrm{fb}^{-1}\) of \(pp\) collisions at \(\sqrt s=8\) TeV. The search covered \(10.06\text{--}10.31~\mathrm{GeV}\) and \(10.40\text{--}10.99~\mathrm{GeV}\), found no evidence for \(X_b\), and set \(95\%\) CL upper limits
\[
R < 0.9\%\text{ to }5.4\%
\]
on
\[
R = \frac{\sigma(pp\to X_b)\,\mathcal B(X_b\to \Upsilon(1S)\pi^+\pi^-)} {\sigma(pp\to \Upsilon(2S))\,\mathcal B(\Upsilon(2S)\to \Upsilon(1S)\pi^+\pi^-)}.
\]
The smallest local \(p\)-value occurred at \(m_{X_b}=10.46~\mathrm{GeV}\) with local significance \(2.6\sigma\), reduced to \(0.8\sigma\) after the look-elsewhere effect [1309.0250].

ATLAS performed an analogous search in the same hidden-bottom dipion mode using \(16.2~\mathrm{fb}^{-1}\) of \(8\) TeV data. It scanned the mass ranges \(10.05\text{--}10.31~\mathrm{GeV}\) and \(10.40\text{--}11.00~\mathrm{GeV}\), found no evidence for a new narrow state, and set observed \(95\%\) CL upper limits on the relative production rate
\[
0.8\%\text{ to }4.0\%,
\]
excluding an \(X_b\) with relative production as large as the measured \(X(3872)\) benchmark \(R_{X(3872)}=6.56\%\) for all masses considered. For masses above about \(10.1\) GeV, the expected upper limits were more restrictive than those from CMS [1410.4409].

Belle searched near the \(\Upsilon(10860)\) region for radiative production
\[
e^+e^- \to \gamma X_b,\qquad X_b\to \omega\Upsilon(1S),
\]
using \(118~\mathrm{fb}^{-1}\) at \(\sqrt s=10.867\) GeV. No significant signal was observed for \(10.55\le m_{X_b}\le 10.65~\mathrm{GeV}/c^2\). At \(m_{X_b}=10.6~\mathrm{GeV}/c^2\), the fit gave \(N_{\rm sig}=-0.4\pm 2.0\) and the \(90\%\) CL upper limit
\[
\mathcal{B}(\Upsilon(10860)\to \gamma X_b)\,
\mathcal{B}(X_b\to \omega\Upsilon(1S))
< 2.9\times 10^{-5},
\]
with the limit varying from \(2.6\times10^{-5}\) to \(3.8\times10^{-5}\) across the scan range [1408.0504].

Belle II later searched for
\[
e^+e^-\to\gamma X_b,\qquad X_b\to\pi^+\pi^-\chi_{bJ},\qquad \chi_{bJ}\to\gamma\Upsilon(1S),
\]
in \(19.6~\mathrm{fb}^{-1}\) collected at \(\sqrt s=10.653,\ 10.701,\ 10.746,\ 10.805~\mathrm{GeV}\). Different hypotheses of the mass of \(X_b\) were evaluated, with the maximum probability found at \(m(X_b)=10.50~\mathrm{GeV}/c^2\), but no evident signal was found. Assuming \(M(X_b)=10.50~\mathrm{GeV}/c^2\), the \(90\%\) C.L. upper limits on
\[
\sigma(e^+e^- \to \gamma X_b)\times \mathcal{B}(X_b \to \pi^+ \pi^- \chi_{bJ})
\]
were \(0.14\), \(0.09\), \(0.17\), and \(0.32\) pb at \(\sqrt{s}=10.653\), \(10.701\), \(10.746\), and \(10.805\) GeV, respectively [2509.01917].

These null results constrain only specific channels and production mechanisms. Theoretical work repeatedly argues that the hidden-bottom dipion mode can be suppressed, while open-bottom or isospin-conserving channels may be more favorable. This suggests that current experimental non-observation is not a model-independent exclusion of \(X_b\), but rather a set of channel-dependent constraints.

## 6. Distinct objects and other uses of the notation

The notation \(X_b\) has also been used for the open-flavor candidate \(X_b(5568)\), which is unrelated to the hidden-bottom \(1^{++}\) state discussed above. In the light-cone sum-rule study “Width of the exotic \(X_b(5568)\) state through its strong decay to \(B_s^{0}\pi^{+}\),” \(X_b(5568)\) is assumed to be a scalar diquark–antidiquark tetraquark of type
\[
[su][\bar b \bar d],
\]
interpolated by
\[
J^{X_b}(x)=\varepsilon^{ijk}\varepsilon^{imn} \left[s^j(x) C\gamma_\mu u^k(x)\right] \left[\bar b^m(x)\gamma^\mu C\bar d^n(x)\right].
\]
The extracted strong coupling is
\[
g_{X_b B_s \pi}=(0.60\pm0.23)\,\mathrm{GeV}^{-1},
\]
leading to
\[
\Gamma(X_b\to B_s^0\pi^+)=(22.4\pm9.2)\,\mathrm{MeV},
\]
presented as compatible with the D0 result [1603.00290]. This is a different object from the hidden-bottom \(X_b\) near \(B\bar B^*\) threshold.

In other branches of high-energy and nuclear physics, the same characters usually refer not to a hadron but to the kinematic variable \(x_B\). In inclusive \((e,e')\) scattering, it is defined as
\[
x_B = \frac{Q^2}{2 m_N \omega},
\]
with \(Q^2 = 4 E_e E_{e'}\sin^2(\theta_{e'}/2)\), and values \(x_B>1\) select nuclear configurations requiring bound and moving nucleons rather than a free nucleon at rest [2402.08199]. In top-quark decay, \(x_B\) denotes the scaled energy of an observed bottom-flavored hadron,
\[
x_B=\frac{E_B}{E_b^{\max}},
\qquad
E_b^{\max}=\frac{m_t^2+m_b^2-m_W^2}{2m_t},
\]
and the central observable is \(d\Gamma/dx_B\) [1205.2528]. In small-\(x_B\) DVCS phenomenology, \(x_B\) is the DIS Bjorken variable
\[
x_B \equiv \frac{{\cal Q}^2}{2 P_1\!\cdot q_1},
\]
with the skewness approximation
\[
\xi = \frac{x_B}{2-x_B} + O(x_B^2), \qquad \xi \approx \frac{x_B}{2}
\]
in the HERA regime [0904.0458].

This notational ambiguity matters in bibliographic practice. In spectroscopy, \(X_b\) usually means the putative hidden-bottom partner of \(X(3872)\); in some hadron papers it means \(X_b(5568)\); and in several other subfields \(x_B\) is purely kinematic. A careful reading of context is therefore essential.

Source: https://www.emergentmind.com/topics/x_b