--- title: Generalized Linear Sigma Model url: https://www.emergentmind.com/topics/generalized-linear-sigma-model type: topic --- # Generalized Linear Sigma Model Searching arXiv for the specified paper and closely related generalized linear sigma model work. arxiv_search.query({"search_query":"all:\"generalized linear sigma model\" AND (id:1709.01834 OR title:\"Electromagnetic trace anomaly in a generalized linear sigma model\" OR title:\"Electromagnetic axial anomaly in a generalized linear sigma model\" OR title:\"Two-body decay widths of lowest lying and next-to-lowest lying scalar and pseudoscalar mesons in generalized linear sigma model\" OR title:\"Generalized linear sigma model with two glueballs\" OR title:\"Spinless glueballs in generalized linear sigma model\" OR title:\"Medium modifications of Heavy Quarkonia masses in a generalized Linear Sigma Model\" OR title:\"Heavy Quarkonium-nuclear bound states within a generalized linear sigma model\")","max_results":10,"sort_by":"relevance"}) Search completed. I’ll use the returned arXiv records corresponding to the specified GLSM papers as citations where needed. The generalized linear sigma model (GLSM) is an effective hadronic framework in which the field content of the linear sigma model is enlarged to accommodate additional QCD degrees of freedom and anomaly structures. In its standard low-energy \(SU(3)_L\times SU(3)_R\) realization, the model contains two chiral nonets, one with quark–antiquark structure and one with four-quark content, and is used to analyze masses, mixings, scatterings, and decay widths of light scalar and pseudoscalar mesons. In other realizations, the GLSM is supplemented by scalar and pseudoscalar glueball fields or by a scalar dilaton field \(\chi\) that simulates the gluon condensate, allowing the same general framework to address QCD axial and trace anomalies, in-medium heavy-quarkonium mass shifts, and quarkonium–nucleus bound states [1506.05575] [2212.12239] [2508.09474]. ## 1. Field content and symmetry assignments A central \(SU(3)\) formulation introduces two \(3\times 3\) matrix fields \[ M=S+i\Phi,\qquad M'=S'+i\Phi', \] where \(S,\Phi\) are the scalar and pseudoscalar nonets built of quark–antiquark degrees of freedom, while \(S',\Phi'\) are the corresponding four-quark nonets. Under \(SU(3)_L\times SU(3)_R\) they transform identically, \[ M\longrightarrow U_LMU_R^\dagger,\qquad M'\longrightarrow U_LM'U_R^\dagger, \] but under the anomalous \(U(1)_A\) they carry different charges, \(M\to e^{2i\alpha}M\) and \(M'\to e^{-4i\alpha}M'\). This distinction is one of the defining structural features of the model, because it permits separate encoding of conventional quarkonium and four-quark components while preserving the same chiral transformation law [1709.01834]. In the low-energy phenomenological applications summarized in the literature, the \(SU(3)\) version is typically truncated at leading order, corresponding to keeping effective terms that contain no more than eight underlying quarks and antiquarks. This truncation is used in studies of two-body strong decays, two-photon decays, and meson spectroscopy, and it provides the basis for the frequently used “scenario 3I” identifications in the scalar and pseudoscalar isosinglet sectors [1506.05575] [1701.00812]. A distinct \(SU(2)_L\times SU(2)_R\) realization is used in dense-matter applications. There the chiral multiplet \((\sigma,\pi^a)\) is combined with a scalar dilaton field \(\chi\), nucleon fields \(\psi_N=(\psi_p,\psi_n)^T\), and light-vector fields \(\omega_\mu,\rho_\mu^a\). Heavy quarkonium fields \(\Phi_Q\) do not enter at tree level; instead, their medium modifications are induced by the medium dependence of the dilaton field, which simulates the scalar gluon condensate of QCD [2410.15898]. ## 2. Effective Lagrangians and vacuum structure In the two-nonet \(SU(3)\) model, a representative leading-order effective Lagrangian has the form \[ \mathcal{L} =\mathcal{L}_{\rm kin}-V_0(M,M')-V_{SB}(S)+\mathcal{L}_{\rm EM\;trace}, \] with kinetic terms \[ \mathcal{L}_{\rm kin} =-\tfrac12\,\mathrm{Tr}\bigl[D_\mu M\,D^\mu M^\dagger\bigr] -\tfrac12\,\mathrm{Tr}\bigl[D_\mu M'\,D^\mu M'^\dagger\bigr], \] and covariant derivative \[ D_\mu M=\partial_\mu M-i\,e\,Q\,M\,A_\mu+i\,e\,M\,Q\,A_\mu,\qquad Q={\rm diag}\bigl(\tfrac23,-\tfrac13,-\tfrac13\bigr). \] The chiral-invariant potential is written, up to eight quark/antiquark lines, as \[ V_0 = -\,c_2\,\mathrm{Tr}[M M^\dagger] +c_4\,\mathrm{Tr}[M M^\dagger M M^\dagger] +d_2\,\mathrm{Tr}[M' M'^\dagger] +e_3\Bigl(\epsilon_{abc}\epsilon^{def}M^a_dM^b_e{M'}^c_f +{\rm h.c.}\Bigr) +c_3\Bigl[\gamma_1\,\ln\tfrac{\det M}{\det M^\dagger} +(1-\gamma_1)\,\ln\tfrac{\mathrm{Tr}(M M'^\dagger)}{\mathrm{Tr}(M' M^\dagger)}\Bigr]^2, \] while explicit chiral symmetry breaking is introduced through \[ V_{SB}=-\,2\,\mathrm{Tr}[A\,S],\qquad A={\rm diag}(A_1,A_1,A_3)\propto(m_u,m_u,m_s). \] In equivalent formulations the same structure appears with slightly different normalizations, but always with two chiral nonets, explicit symmetry breaking by a quark-mass spurion, and a logarithmic \(U(1)_A\)-breaking term [1709.01834] [1506.05575]. The vacuum structure is specified by diagonal condensates, \[ \langle S\rangle=\mathrm{diag}(\alpha_1,\alpha_2,\alpha_3),\qquad \langle S'\rangle=\mathrm{diag}(\beta_1,\beta_2,\beta_3), \] with the isospin limit \(\alpha_1=\alpha_2\), \(\beta_1=\beta_2\), \(A_1=A_2\). The extremum conditions \(\partial V/\partial S=0\) and \(\partial V/\partial S'=0\) determine the condensates in terms of the Lagrangian parameters. In the pseudoscalar-anomaly analysis there are initially 12 unknowns, namely six couplings \(\{c_2,c_4^a,d_2,e_3^a,c_3,\gamma_1\}\), two quark-mass parameters \(\{A_1,A_3\}\), and four condensates \(\{\alpha_1,\alpha_3,\beta_1,\beta_3\}\); the minimum conditions remove four of these, and the remaining eight are fixed by meson masses, \(F_\pi\), \(A_3/A_1\), and the trace and determinant conditions for the \(I=0\) pseudoscalar mass matrix [1701.00812]. In the \(SU(2)\) dense-matter version, the total Lagrangian is written as \[ \mathcal L=\mathcal L_0-V_G+\mathcal L', \] where \(\mathcal L_0\) contains the chiral- and scale-invariant kinetic and Yukawa terms, \(V_G\) is the scale-anomaly potential, and \(\mathcal L'\) contains the isovector–vector \(\rho\) terms. Here the vacuum expectation values are \(\sigma_0=f_\pi\approx93\,\mathrm{MeV}\) and \(\chi_0=\zeta\,\sigma_0\approx149\,\mathrm{MeV}\), with \(M_N=g_\sigma\sigma_0\) [2212.12239]. ## 3. Axial, electromagnetic, and trace anomalies One of the most distinctive aspects of the GLSM literature is the treatment of anomalies by nonderivative logarithmic terms. For the electromagnetic axial anomaly, the divergence of the \(U(1)_A\) current arising from \(\mathcal L_{\rm kin}+V_0+\mathcal L_{SB}\) vanishes identically in a linear realization, so an additional term is introduced to reproduce the Adler–Bell–Jackiw anomaly: \[ \partial_\mu J^\mu_{5(k)} = -\,\frac{e^2}{16\pi^2}\,\epsilon^{\mu\nu\rho\sigma}\,\mathrm{Tr}[\lambda_kQ^2]\,F_{\mu\nu}F_{\rho\sigma}. \] A convenient form is \[ X(M)= i\sum_{i=1}^3 a_i \Big\{\ln\mathrm{Tr}[x_iM+Mx_i]-\ln\mathrm{Tr}[x_iM^\dagger+M^\dagger x_i]\Big\} \,\epsilon^{\mu\nu\rho\sigma}F_{\mu\nu}F_{\rho\sigma}, \] with \(x_i=\mathrm{diag}(\delta_{i1},\delta_{i2},\delta_{i3})\), and matching gives \[ a_1=-\frac{e^2}{64\pi^2},\qquad a_2=-\frac{e^2}{64\pi^2},\qquad a_3=-\frac{2e^2}{64\pi^2}. \] An exactly analogous term \(X'(M')\) is constructed for the four-quark nonet, and the combined model may be written as \(\mathcal L_{\rm anom}=X(M)+X'(M')\) [1701.00812]. For the electromagnetic trace anomaly in the scalar sector, the minimal term is constructed by analogy with the ’t Hooft axial-anomaly term: \[ \mathcal{L}_{\rm EM\;trace} = b\,F^{\mu\nu}F_{\mu\nu} \Bigl\{ \tau_1\bigl[\ln(\tfrac{\det M}{\Lambda^3})+\ln(\tfrac{\det M^\dagger}{\Lambda^3})\bigr] +\tau_2\bigl[\ln(\tfrac{\mathrm{Tr}(M M'^\dagger)}{\Lambda^2}) +\ln(\tfrac{\mathrm{Tr}(M' M^\dagger)}{\Lambda^2})\bigr] \Bigr\}, \] with \[ 6\,\tau_1+4\,\tau_2=1,\qquad b=\frac{e^2}{12\pi^2},\qquad \beta(e)=\frac{e^3}{6\pi^2}. \] This term reproduces \(\theta^\mu_{\ \mu}=-\frac{\beta(e)}{2e}F_{\mu\nu}F^{\mu\nu}\) and yields effective \(f_i\gamma\gamma\) couplings after expansion around the vacuum expectation values [1709.01834]. In the dense-matter and glueball variants, the trace anomaly is instead encoded through a dilaton or scalar-glueball sector. In the \(SU(2)\) model the scale-anomaly potential is \[ V_G(\chi,\sigma,\pi) =B\,\chi^4\Big[\ln\big(\tfrac{\chi}{\chi_0}\big)-\tfrac14\Big] -\tfrac12\,B\,\delta\,\chi^4\ln\Big(\tfrac{\sigma^2+\pi^2}{\sigma_0^2}\Big) +\tfrac12\,B\,\delta\,\zeta^2\,\chi^2\Big[\sigma^2+\pi^2-\tfrac12\frac{\chi^2}{\zeta^2}\Big], \] with \(\delta=2N_f/(11N_c)=4/33\) for \(N_c=3\), \(N_f=2\), and \[ \Big\langle\frac{\alpha_s}{\pi}G_{\mu\nu}^aG^{a\mu\nu}\Big\rangle =\frac{24}{33-2N_f}\,B(1-\delta)\,\chi^4. \] In glueball-extended \(SU(3)\) versions, scalar and pseudoscalar glueballs are introduced through their connections with the trace and axial anomalies of QCD, respectively, with explicit anomaly terms \(f_A\) and \(f_S\) in the potential [2212.12239] [2508.09474]. ## 4. Mixing, mass eigenstates, and spectroscopy Because \(M\) and \(M'\) transform identically under \(SU(3)_L\times SU(3)_R\), physical scalars and pseudoscalars are mixtures of two-quark and four-quark fields. The model therefore employs orthogonal rotations from bare fields to mass eigenstates. In the \(I=1\) pseudoscalar sector, for example, \[ \begin{pmatrix}\pi(137)\\ \pi(1300)\end{pmatrix} =R_\pi^{-1} \begin{pmatrix}\phi^1_2\\ \phi'^1_2\end{pmatrix}, \] while in the isosinglet scalar sector one may define \[ f_a=\tfrac1{\sqrt2}(S^1_1+S^2_2),\quad f_b=S^3_3,\quad f_c=\tfrac1{\sqrt2}(S'^1_1+S'^2_2),\quad f_d=S'^3_3, \] and then \[ \begin{pmatrix}f_a\\f_b\\f_c\\f_d\end{pmatrix} =L_0 \begin{pmatrix}f_1\\f_2\\f_3\\f_4\end{pmatrix}. \] In the fit of Jora et al. designated “scenario 3I,” one identifies \(f_1\equiv f_0(500)\) and \(f_2\equiv f_0(980)\) [1506.05575] [1709.01834]. The same scenario is used in the pseudoscalar sector. Of six possible assignments for the heavier model-predicted \(\eta\) states, the “scenario 3I”—identifying \(\eta_3\equiv\eta(1295)\), \(\eta_4\equiv\eta(1760)\) and taking solution I of the resulting quadratic—gives the best overall \(\chi^2\) to the mass spectrum and to the decay \(\eta\to\eta\pi\pi\). For central inputs \(m[\pi(1300)]=1.30\,\mathrm{GeV}\) and \(A_3/A_1=30\), the model yields \(c_2 \simeq 1.85\times10^6\,\mathrm{MeV}^2\), \(c_4^a\simeq 9.9\), \(d_2\simeq1.6\times10^6\,\mathrm{MeV}^2\), \(e_3^a\simeq -4.1\times10^5\,\mathrm{MeV}\), \(c_3\simeq4.6\), \(\gamma_1\simeq0.50\), \(A_1\simeq5.3\,\mathrm{MeV}\), \(A_3\simeq160\,\mathrm{MeV}\), \(\alpha_1\simeq65\,\mathrm{MeV}\), \(\alpha_3\simeq78\,\mathrm{MeV}\), \(\beta_1\simeq108\,\mathrm{MeV}\), and \(\beta_3\simeq102\,\mathrm{MeV}\) [1701.00812]. Glueball-extended versions enlarge the isosinglet mass matrices further. In the 2025 spin-0 glueball analysis, the model contains two chiral nonets together with scalar and pseudoscalar glueballs. It is found that in order to satisfy the axial anomaly and at the same time accurately generate all seven eta masses, it is necessary to include at least two pseudoscalar glueballs, a physical one and an unphysical one that gets integrated out and yields an effective instanton-type term. In that analysis, the state that is dominantly made of glue in the pseudoscalar sector is a state with mass above \(2\,\mathrm{GeV}\), while in the scalar sector the three isosinglets in the \(1.5\text{–}2.0\,\mathrm{GeV}\) region can contain substantial glue, with the glue contents determined as functions of the scalar glueball condensate \(h_0\) [2508.09474]. ## 5. Strong and electromagnetic decays The GLSM is extensively used to compute tree-level strong two-body widths and, after unitarization, physical pole positions. For a generic scalar decay into two pseudoscalars, \[ \Gamma_{S\to P_1P_2} =\frac{|\vec q|}{8\pi\,m_S^2}\,|\gamma_{SP_1P_2}|^2\times(\text{isosym\ factor}), \] with \[ |\vec q| =\frac1{2m_S}\sqrt{[m_S^2-(m_{P_1}+m_{P_2})^2][m_S^2-(m_{P_1}-m_{P_2})^2]}. \] Loop and final-state-interaction effects are incorporated through K-matrix unitarization of the partial-wave amplitude, \[ T_l^I(s)=\frac{T_l^{I,B}(s)}{1-i\,T_l^{I,B}(s)}, \] and the pole positions give the physical masses \(\tilde m\) and widths \(\tilde\Gamma\). In the decay-width study of lowest lying and next-to-lowest lying scalar and pseudoscalar mesons, the two-body decay widths of lowest lying mesons are well predicted by this model while for the next-to-lowest lying mesons, only some of the decay widths agree with the experimental results [1506.05575]. The comparison with the ordinary single-nonet linear sigma model is a recurring theme. In the decoupling limit \(d_2,e_3^a\to0\), \(\gamma_1\to1\), or in the ordinary non-renormalizable single-nonet model, the predicted widths are much too large. The explicit comparison quoted for \(\sigma\to\pi\pi\) is \[ \Gamma_{\rm SNLSM}=830~\mathrm{MeV},\qquad \Gamma_{\rm GLSM\;(bare)}=531\pm99~\mathrm{MeV},\qquad \Gamma_{\rm GLSM,\;unitarized}=385\pm61~\mathrm{MeV}, \] with corresponding masses \(847\,\mathrm{MeV}\), \(645\pm42\,\mathrm{MeV}\), and \(476\pm4\,\mathrm{MeV}\). The same analysis states that including two- and four-quark mixing in GLSM brings all lowest-lying widths into the correct ballpark [1506.05575]. For pseudoscalar two-photon decays, the anomaly term yields \[ \Gamma(\pi_i\to\gamma\gamma)=f_i^2m_i^3/\pi. \] Using the fitted parameters, the model predicts \[ \Gamma(\pi^0(137)\to\gamma\gamma)=(7.67\pm0.01)\times10^{-3}\,\mathrm{keV}, \] versus \(7.64\times10^{-3}\,\mathrm{keV}\) experimentally, and also gives \[ \Gamma(\eta(547)\to\gamma\gamma)=0.39\pm0.04~\mathrm{keV},\qquad \Gamma(\eta'(958)\to\gamma\gamma)=7.0\pm0.7~\mathrm{keV}. \] Apart from the slight overestimate of \(\eta'(958)\to\gamma\gamma\), the agreement is at the \(20\text{–}30\,\%\) level, and the results suggest significant four-quark mixing in the heavier pseudoscalars [1701.00812]. For scalar two-photon decays, the trace-anomaly term gives \[ \mathcal L_{f_i\gamma\gamma}=-\tfrac12\,F_i\,f_i\,F^{\mu\nu}F_{\mu\nu},\qquad \Gamma(f_i\to\gamma\gamma)=\frac{F_i^2}{16\pi}\,m_{f_i}^3. \] Fitting the well-measured width \[ \Gamma(f_0(980)\to\gamma\gamma)=0.31^{+0.05}_{-0.04}\,\mathrm{keV} \] shows that agreement with experiment occurs only if \(\tau_2\gtrsim0.7\) or \(\tau_2\lesssim-0.8\), i.e. only when the direct mixing-sensitive term proportional to \(\tau_2\) is large. Once \(\tau_2\) is so constrained, the model predicts \[ \Gamma(f_0(500)\to\gamma\gamma)\gtrsim0.5~\mathrm{keV}. \] The same analysis states that “turning off” mixing, \(\tau_2\to0\), renders both decay widths far too small [1709.01834]. ## 6. Glueballs and anomaly-driven extensions The inclusion of explicit glueball fields extends the GLSM beyond the two-nonet mesonic sector. In the model with two glueballs, one introduces a scalar glueball \(G_s\equiv h(x)\) and a pseudoscalar glueball \(G_p\equiv g(x)\) in addition to the nonets \(M\) and \(M'\), and writes \[ \mathcal L_{\rm eff} =-\frac12\Tr(\partial_\mu M\,\partial^\mu M^\dagger) -\frac12\Tr(\partial_\mu M'\,\partial^\mu M'^\dagger) -\frac12\,\partial_\mu G_s\,\partial^\mu G_s -\frac12\,\partial_\mu G_p\,\partial^\mu G_p -V(M,M',G_s,G_p), \] with \[ V=V_0+V_A+V_S+V_{SB}. \] Here \(V_A\) mocks up the axial \(U(1)_A\) anomaly and \(V_S\) mocks up the QCD trace anomaly. In the decoupling limit, where glueball–quarkonia interactions are switched off, the analysis determines the properties of the pure scalar glueball and yields \(m_{G_s}\approx1.6\text{–}2.0\,\mathrm{GeV}\), stated to be in good agreement with lattice and QCD-sum-rule estimates [1807.10927]. The more detailed 2025 analysis with spinless glueballs shows that decay widths of isosinglet scalars as well as different self consistencies within this framework can be used to probe the glueball condensate. A histogram of successful decays versus \(h_0\) peaks sharply in the high-\(h_0\) band, and the combined mass-spectrum plus width analysis fixes the scalar-glue condensate to \(h_0\approx0.75\text{–}0.83\,\mathrm{GeV}\). In that treatment, \(f_0(1370)\), \(f_0(1500)\), and \(f_0(1710)\) all carry appreciable glue admixture, but the precise glue contents depend sensitively on \(h_0\) [2508.09474]. A plausible implication is that, within the GLSM family, the scalar sector above \(1\,\mathrm{GeV}\) is best viewed not as a single nonet problem but as a coupled system of quarkonium, four-quark, and glueball configurations whose relative weights depend on anomaly terms and vacuum condensates. ## 7. Dense matter, heavy quarkonia, and quarkonium–nucleus bound states In nuclear-matter applications the GLSM is reformulated as an \(SU(2)\times SU(2)\) chiral model with broken scale invariance incorporated through the dilaton field \(\chi\). The model relation \[ \Big\langle\frac{\alpha_s}{\pi}G^2\Big\rangle -\Big\langle\frac{\alpha_s}{\pi}G^2\Big\rangle_0 =\frac{24}{33-2N_f}\,B(1-\delta)(\chi^4-\chi_0^4) \] is inserted into the leading-order heavy-quarkonium mass-shift formula, \[ \Delta m_\psi =\frac{4}{87}\,B(1-\delta) \int dk^2\, \Bigl\langle|\partial_k\psi|^2\Bigr\rangle \frac{k}{k^2/m_Q+\epsilon} (\chi^4-\chi_0^4), \] where the momentum-space wave functions are Fourier transforms of harmonic-oscillator eigenfunctions with size parameter \(\beta\) fixed to reproduce the vacuum r.m.s. radius of each quarkonium state. The studies report an appreciable mass drop in the heavy-quarkonium states under consideration and note that the in-medium masses at finite densities should modify the in-medium partial decay widths of heavy quarkonia to open heavy flavor mesons [2212.12239]. The 2024 bound-state analysis converts the medium-modified dilaton field into quarkonium–nucleus potentials, \[ V_{Q\bar Q}(r)=\Delta m_{Q\bar Q}[\chi(r)], \] and then solves the Klein–Gordon equation \[ \Big\{ \nabla^2 + [E-V(r)]^2 - m^2 \Big\}\Phi(r)=0 \] for bound-state eigenvalues \(E