Probing Quark Electric Dipole Moment with Topological Anomalies Chao-Qiang Geng, Xiang-Nan Jin, Chia-Wei Liu, Bin Wu Abstract CP-odd observables originating from the strange-quark electric dipole moment ( d s d_{s} ) are identified in γ ∗ → K + ​ K − ​ π 0 \gamma^{\ast}\to K^{+}K^{-}\pi^{0} and evaluated in the chiral limit. A nonzero T-odd asymmetry A T A_{T} requires anomalous couplings descending from a five-dimensional Chern–Simons construction. Modeling the running of F A F_{A} and F V F_{V} with vector-meson dominance, we estimate sensitivities to d s d_{s} at the level of 𝒪 ⁡ ( 10 − 16 ) ​ e ⋅ cm \mathcal{O}(10^{-16})\,e\cdot\mathrm{cm} in e + ​ e − e^{+}e^{-} data at CMD–3 and 𝒪 ⁡ ( 10 − 18 ) ​ e ⋅ cm \mathcal{O}(10^{-18})\,e\cdot\mathrm{cm} using existing J / ψ J/\psi samples at BESIII. Future experiments at the Super Tau–Charm Facility and Belle II can further improve the reach by an order of magnitude. I Introduction Electric dipole moments (EDMs) are fundamental properties of fermions that quantify CP violation Smith:1957ht ; Bernreuther:1990jx . In particular, nucleon EDM searches provide by far the most stringent constraint on CP violation in QCD, forcing the vacuum angle \overline{\theta} , which naturally ranges from -\pi to \pi , to satisfy |\overline{\theta}|\lesssim 10^{-10} He:1989mbz ; Abel:2020pzs ; Chupp:2017rkp . Complementary constraints on hadronic CP violation are also obtained from atomic EDM searches Lamoreaux:1987zz . Accordingly, in the following we set \overline{\theta}=0 at the level of precision relevant for this work. On the other hand, the electron EDM currently provides the tightest constraint on physics beyond the Standard Model (SM) Roussy:2022cmp ; Barr:1990vd ; Ramsey-Musolf:2006evg ; Li:2010ax ; Jung:2013hka ; Inoue:2014nva ; Shu:2013uua ; Li:2021xmw ; He:1992dc ; Dorsner:2016wpm ; Fuyuto:2018scm ; Dekens:2018bci . Nevertheless, EDMs of quarks beyond the first generation remain subject to little direct experimental constraint Huang:2025ghw ; Belle:2021ybo ; Vecchi:2025jbb . In the lepton sector, the muon EDM provides a direct probe of CP violation beyond the first generation from storage-ring measurements Muong-2:2008ebm . Very recently, the BESIII collaboration reported the first direct bound on the \Lambda hyperon EDM from J/\psi\to\Lambda\overline{\Lambda} BESIII:2025vxm : \left|d_{\Lambda}(m_{J/\psi}^{2})\right|<6.5\times 10^{-19}\,e\cdot\mathrm{cm}\,. (1) Because QCD is nonperturbative in this kinematic regime, a direct relation between d_{\Lambda}(s) and the strange-quark EDM d_{s} can only be established at the zero-recoil point with an on-shell photon; d_{\Lambda}(0)=d_{s} He:1989mbz in the quark model. Once the momentum dependence of d_{\Lambda}(s) is taken into account, the inferred constraint on d_{s} is weakened relative to the bound on d_{\Lambda}(m_{J/\psi}^{2}) by three orders Chen:2025rab . In this work we propose a novel probe of d_{s} through angular distributions in e^{+}e^{-}\to K^{+}K^{-}\pi^{0} at e^{+}e^{-} colliders. In particular, J/\psi may appear as the intermediate state as shown in FIG. 1 . Compared with hyperon channels, this mode has a larger branching fraction and does not require subsequent decays, which would otherwise reduce the effective statistics by (\alpha_{\mathrm{sub}}{\cal B}_{\mathrm{sub}})^{2} Du:2024jfc ; Fu:2023ose , where {\cal B}_{\mathrm{sub}} and \alpha_{\mathrm{sub}} are the branching fraction and decay parameter of the subsequent decays. Within the SM, the amplitude for \gamma^{\ast}\to K^{+}K^{-}\pi^{0} at low energies is governed by the Wess–Zumino–Witten (WZW) term Wess:1971yu ; Witten:1983tw in chiral perturbation theory ( \chi PT) Weinberg:1978kz ; Gasser:1983yg ; Gasser:1984gg ; Leutwyler:1993iq , fixed by gauge anomalies Kaymakcalan:1983qq ; Chou:1983qy ; Wu:1986pr ; Fujiwara:1984mp , providing a controlled theoretical framework to isolate EDM effects. As been shown, the odd intrinsic processes are fixed at the chiral limit even in the presence of resonances Harada:2003jx ; Geng:2025fuc ; Geng:2026rbz . We construct a CP-odd observable from the angular correlations, match the d_{s} operator onto \chi PT, and estimate the achievable experimental sensitivity. Figure 1 : Schematic structure of the amplitudes for \gamma^{\ast}\!\to K^{+}K^{-}\pi^{0} . The left panel shows the SM contribution mediated by three-gluon annihilation. The right panel shows the EDM-induced contribution, where the boxed cross denotes the strange-quark EDM d_{s} . This paper is organized as follows. In Sec. II , we construct the CP-odd observable. In Sec. III , we match d_{s} onto \chi PT and present the WZW prediction. In Sec. IV , we estimate the attainable experimental precision. We conclude in Sec. V . II CP-odd observables Consider \gamma^{*}(q)\to K^{+}(p_{+})\,K^{-}(p_{-})\,\pi^{0}(p_{0}) . We employ the symmetric set of kinetic variables s=q^{2}, and s_{\pm}=(p_{\pm}+p_{0})^{2}. The total amplitude is parameterized as {\cal M}(\gamma^{*}\to K^{+}K^{-}\pi^{0})=\epsilon_{\mu}(q)\,J^{\mu}, (2) where \epsilon is the polarization vector of the virtual photon. Using a gauge-invariant decomposition, the hadronic current can be parameterized as J^{\mu}=F_{V}\Big(s,s_{+},s_{-}\Big)V^{\mu}\!+\!F_{A}\!\Big(s,s_{+},s_{-}\Big)A^{\mu}, (3) with the two transverse basis vectors \displaystyle V^{\mu} \displaystyle= \displaystyle(q\!\cdot\!p_{0})\,(p_{+}-p_{-})^{\mu}-\big(q\!\cdot\!(p_{+}-p_{-})\big)\,p_{0}^{\mu}, \displaystyle A^{\mu} \displaystyle= \displaystyle\,\epsilon^{\mu\nu\alpha\beta}\,p_{+\nu}\,p_{-\alpha}\,p_{0\beta}, (4) which satisfy q_{\mu}V^{\mu}=q_{\mu}A^{\mu}=0 . Invariance under C-parity demands F_{V} and F_{A} to be invariant under exchanging s_{+} and s_{-} ; see the figures in Ref. BESIII:2019apb . Since the final state has negative intrinsic parity, one sees that F_{V} and F_{A} are CP-even and CP-odd, respectively. In the \gamma^{*} centre-of-mass frame, q^{\mu}=(\sqrt{s},\vec{0}) and p_{i}^{\mu}=(E_{i},\vec{p}_{i}) . The amplitude can be written as {\cal M}=\vec{\epsilon}\cdot\Big(F_{V}\vec{V}+F_{A}\vec{A}\Big), (5) with \vec{V}=\sqrt{s}\,\big[E_{0}\,(\vec{p}_{+}-\vec{p}_{-})+(E_{+}-E_{-})\,\big(\vec{p}_{+}+\vec{p}_{-}\big)\big] and \vec{A}=\sqrt{s}\,(\vec{p}_{+}\times\vec{p}_{-}) . Taking the unpolarized e^{+}e^{-}\to\gamma^{*} case, we sum over the two transverse photon polarizations and obtain \overline{|{\cal M}|^{2}}=\frac{1}{2}\sum_{\lambda=\pm}|{\cal M}|^{2}=\frac{1}{2}\Big|\big(F_{V}\,\vec{V}+F_{A}\,\vec{A}\big)_{\perp}\Big|^{2}, (6) where \vec{X}_{\perp}\equiv\vec{X}-(\vec{X}\cdot\hat{z})\hat{z} denotes the component transverse to the beam axis. The interference term proportional to \Re(F_{V}F_{A}^{*})\,\vec{V}_{\perp}\cdot\vec{A}_{\perp} is C -even and CP-odd, and will be shown to be induced by d_{s} . The fully differential decay distribution for \gamma^{*}(q)\to K^{+}(p_{+})\,K^{-}(p_{-})\,\pi^{0}(p_{0}) in the \gamma^{*} rest frame is \frac{d\Gamma}{d\Phi}=\frac{1}{(2\pi)^{4}}\,\frac{|\vec{p}_{0}|\,|\vec{p}_{+}^{\,*}|}{16\,s}\;\overline{|{\cal M}|^{2}}, (7) where d\Phi\equiv dm_{KK}\,d\cos\theta_{0}\,d\Omega_{+}^{*}. Here, \theta_{0} is the inclined angle of the three-momentum of \pi^{0} and \gamma^{*} polarization in the \gamma^{\ast} rest frame, while \Omega_{+}^{\ast}=(\theta^{\ast},\phi) and \vec{p}_{+}^{\,\ast} are those of K^{+} in the K^{+}K^{-} rest frame; see FIG. 2 . Figure 2 : The definitions of the angles in Eq. ( 7 ). The angle \theta^{0} is the angle between the momenta of e^{+} and \pi^{0} in the center-of-mass frame. The angle \theta^{\ast} is the angle between \pi^{0} and K^{+} in the K^{+}K^{-} rest frame. The angle \phi is the dihedral angle between the two planes spanned by (\pi^{0},e^{+},e^{-}) and (\pi^{0},K^{+},K^{-}) . To isolate the contribution linear in F_{A} without assuming a specific kinematic dependence, we define the signed interference observable \vec{V}_{\perp}\cdot\vec{A}_{\perp} and construct the counting asymmetry A_{T}\equiv\frac{N\big(\vec{V}_{\perp}\!\cdot\!\vec{A}_{\perp}>0\big)-N\big(\vec{V}_{\perp}\!\cdot\!\vec{A}_{\perp}<0\big)}{N\big(\vec{V}_{\perp}\!\cdot\!\vec{A}_{\perp}>0\big)+N\big(\vec{V}_{\perp}\!\cdot\!\vec{A}_{\perp}<0\big)}\,. (8) Only the interference term \Re(F_{V}F_{A}^{*})\,(\vec{V}_{\perp}\cdot\vec{A}_{\perp}) contributes, providing a robust probe of \Re(F_{V}F_{A}^{*}) . Finally, we note that the framework here can be straightforwardly extended to the one with J/\psi as the initial state, as its J^{PC} is identical to \gamma^{*} . III Mapping d_{s} to \chi PT Starting from the flavor-conserving CP-violating strange-quark dipole operator, \mathcal{L}_{\text{eff}}=\frac{d_{s}}{2}\,\tilde{F}^{\mu\nu}\,\bar{s}\,\sigma_{\mu\nu}\,s, (9) where d_{s} parametrizes the strange-quark electric dipole operator induced by new physics. The QCD Lagrangian enjoys a global SU(3)_{L}\otimes SU(3)_{R} flavor symmetry, with q_{L}\to g_{L}\,q_{L} , q_{R}\to g_{R}\,q_{R} and q=(u,d,s)^{T} . To make chiral symmetry manifest in the presence of d_{s} , we introduce spurion fields t_{\pm} that transform as t_{+}\to g_{L}\,t_{+}\,g_{R}^{\dagger},\qquad t_{-}\to g_{R}\,t_{-}\,g_{L}^{\dagger}. (10) With these spurions, the effective Lagrangian can be rewritten as \mathcal{L}_{\text{eff}}=\frac{d_{s}}{2}\,\tilde{F}^{\mu\nu}\,\Big(\bar{q}_{L}\,\sigma_{\mu\nu}\,t_{+}\,q_{R}+\bar{q}_{R}\,\sigma_{\mu\nu}\,t_{-}\,q_{L}\Big). (11) Matching to Eq. ( 9 ), one sets t_{\pm}=\mathrm{diag}(0,0,1) in flavor space. With the spurion transformation rules above, \mathcal{L}_{\text{eff}} is formally invariant under SU(3)_{L}\otimes SU(3)_{R} . To match onto chiral fields, we note that U=\exp(2iP/f_{P}) with \displaystyle P\!=\!\!\!\ \left[\begin{array}[]{ccc}\frac{\eta_{0}}{\sqrt{3}}+\frac{\eta_{8}}{\sqrt{6}}+\frac{\pi^{0}}{\sqrt{2}}&\pi^{+}&K^{+}\\ \pi^{-}&\frac{\eta_{0}}{\sqrt{3}}+\frac{\eta_{8}}{\sqrt{6}}-\frac{\pi^{0}}{\sqrt{2}}&K^{0}\\ K^{-}&\bar{K}^{0}&\frac{\eta_{0}}{\sqrt{3}}-\frac{2\eta_{8}}{\sqrt{6}}\end{array}\right]\!, which transforms as U\to g_{L}\,U\,g_{R}^{\dagger}. The lowest-order chiral realization of the strange-quark tensor currents is \displaystyle{\cal L} \displaystyle= \displaystyle\frac{d_{s}}{2}\tilde{F}^{\mu\nu}\frac{\Lambda_{2}}{8}T_{\mu\nu}(t_{+},t_{-}) \displaystyle T_{\mu\nu}(t_{+},t_{-}) \displaystyle= \displaystyle i\mathrm{Tr}\!\Big[(t_{+}U^{\dagger}+Ut_{-})\,[\alpha_{\mu},\alpha_{\nu}]\Big] (16) \displaystyle\,-\,\epsilon_{\mu\nu\lambda\sigma}\,\mathrm{Tr}\!\Big[(t_{+}U^{\dagger}-Ut_{-})\,\alpha^{\lambda}\alpha^{\sigma}\Big]\,, where \alpha_{\mu}=-i\,\partial_{\mu}U\,U^{\dagger} and \Lambda_{2} is a low-energy constant. The tensor T_{\mu\nu}(t_{+},t_{-}) is constructed to satisfy \tfrac{1}{2}\,\epsilon^{\mu\nu\alpha\beta}T_{\alpha\beta}(t_{+},t_{-})=-\,i\,T^{\mu\nu}(t_{+},-t_{-}) , mirroring \tfrac{1}{2}\,\epsilon^{\mu\nu\alpha\beta}\sigma_{\alpha\beta}=-\,i\,\sigma^{\mu\nu}\gamma_{5} . The relevant piece reads \displaystyle{\cal L}^{(3)} \displaystyle=\frac{i2\sqrt{2}\,d_{s}\,\Lambda_{2}}{f_{K}^{2}f_{\pi}}F^{\mu\nu}\Big[K^{+}\,(\partial_{\mu}K^{-})\,(\partial_{\nu}\pi^{0})\Big], (17) where F^{\mu\nu}=\big(\partial^{\mu}A^{\nu}-\partial^{\nu}A^{\mu}\big). We focus on this channel because K^{\pm} is on shell and \pi^{0} can be reconstructed from two photons in relevant experiments. In the SM, the leading contribution is fixed by the WZW term. Since this term is the descendant of a five-dimensional Chern–Simons form, it is naturally written in the language of differential forms Wess:1971yu ; Witten:1983tw {\cal L}_{\chi{\rm PT}}^{\rm(WZW)}=\frac{\,N_{c}}{48\pi^{2}}\,{\rm Tr}\!\left[eQ\,A\wedge\alpha^{3}\right]. (18) Here A stands for the photon field. The overall normalization is fixed by the underlying chiral anomaly. Equivalently, it is topologically quantized by the five-dimensional Chern–Simons construction, so the coefficient is an integer, which is N_{c} in QCD. Expanding to {\cal O}(f_{P}^{-3}) yields the familiar local \gamma^{\ast}\to K^{+}K^{-}\pi^{0} vertex {\cal L}_{\chi{\rm PT}}^{\gamma KK\pi^{0}}=-\frac{N_{c}\,e}{12\pi^{2}\,f_{K}^{2}f_{\pi}}\,A\wedge d\pi^{0}\wedge dK^{+}\wedge dK^{-}. (19) The WZW-induced amplitude therefore provides the irreducible odd-parity baseline for \gamma^{\ast}\to K^{+}K^{-}\pi^{0} . The interference between the WZW term and the EDM naturally occurs because both induce odd-intrinsic-parity processes. In contrast, ordinary \chi PT conserves intrinsic parity and does not contribute. In the chiral limit, \chi PT yields \!\!\!\!F_{V}(0,0,0)=\frac{-N_{c}\,e}{12\pi^{2}\,f_{K}^{2}f_{\pi}},~~F_{A}(0,0,0)=\frac{\sqrt{2}\,d_{s}\,\Lambda_{2}}{f_{K}^{2}f_{\pi}}\,. (20) At \sqrt{s}>2m_{K}+m_{\pi} , resonances such as K^{*+}\to K\pi^{0} induce structures in F_{V} and F_{A} . In principle, the form factors F_{V} and F_{A} can be determined directly from the measured kinematic distributions; a dedicated Monte-Carlo study of this extraction will be reported elsewhere. To leading order, F_{V} and F_{A} can be modeled by the sequential decays of vector meson dominance (VMD) \gamma^{*}\to K^{+}K^{*-}(892)+c.c. as Sakurai:1960ju \displaystyle\!\!\!F_{V,A}\Big(s,s_{+},s_{-}\Big)=F_{V,A}(s,0,0) (21) \displaystyle\!\!\!\frac{1}{2}\left(\frac{m_{K^{*}}^{2}}{m_{K^{*}}^{2}-s_{+}-i\Gamma_{K^{*}}\sqrt{s_{+}}}+\frac{m_{K^{*}}^{2}}{m_{K^{*}}^{2}-s_{-}-i\Gamma_{K^{*}}\sqrt{s_{-}}}\right), where F_{V}(s,0,0) is a nonperturbative input to be fitted from the experiments. The determination of the ratio F_{A}(s,0,0)/F_{V}(s,0,0) needs to await lattice input, and here we take it as a constant in \gamma^{*}\to K^{+}K^{-}\pi^{0} . For J/\psi\to K^{+}K^{-}\pi^{0} , the leading form factor F_{V} is induced by three-gluon annihilation, and the ratio F_{A}/F_{V} is no longer given by Eq. ( 20 ). The leading form factor F_{V}(s,0,0) is extracted from the experimental branching fraction, while we adopt the resonant structure F_{A}(m_{J/\psi}^{2},0,0)=\frac{eQ_{c}f_{J/\psi}}{m_{J/\psi}}\frac{m_{A}^{2}}{m_{A}^{2}-m_{J/\psi}^{2}}F(0,0,0)\,, (22) for F_{A} , where Q_{c}=2/3 is the charge of the charm quark, f_{J/\psi}=(410.4\pm 1.7) MeV is the decay constant of J/\psi Hatton:2020qhk , and m_{A} is the nominal axial-vector meson mass and we vary it from 1 GeV to 2 GeV as an uncertainty of our framework. The dependencies on s_{+} and s_{-} can be well described by the VMD in Eq. ( 21 ). In particular, the BESIII collaboration found that more than 95\% of the {\cal B}(J/\psi\to K^{+}K^{-}\pi^{0}) is made of the sequential decay J/\psi\to K^{+}K^{*-}+c.c. BESIII:2019apb . IV Numerical result We propose to probe the EDM-induced interference in e^{+}e^{-}\to\gamma^{\ast}\to K^{+}K^{-}\pi^{0} . For a sample of N selected signal events, the purely statistical uncertainty is \delta A_{T}\simeq\sqrt{\frac{1-A_{T}^{2}}{N}}\;\approx\;\frac{1}{\sqrt{N}}\qquad(|A_{T}|\ll 1)\,. (23) The relation between A_{T} and d_{s} is denoted as d_{s}=\frac{-N_{c}e}{12\sqrt{2}\,\pi^{2}\Lambda_{2}}\,\frac{A_{T}}{R(s)}\,. (24) Here, R(s) is a phase factor that gauges our ability to extract d_{s} from A_{T} , and it can be computed numerically from Eq. ( 8 ). With the modeling of the VMD, R(s) is at order 1. In the following we take \Lambda_{2}=(18.20^{+1.17}_{-0.99})\times 10^{-3}~\mathrm{GeV} Jiang:2012ir . For CMD–3, the analysed sample ( L\simeq 34~\mathrm{pb}^{-1} ) Ryzhenenkov:2017lumi ; Erofeev:2017kkpi0 and the measured Born cross section \sigma(e^{+}e^{-}\!\to K^{+}K^{-}\pi^{0})\sim 0.05 – 1.1~\mathrm{nb} in \sqrt{s}=1.4 – 2.0~\mathrm{GeV} Erofeev:2017kkpi0 imply N_{\rm prod}\sim(1.7\times 10^{3}\!-\!2.7\times 10^{4})\,. The VMD behavior in Eq. ( 21 ) gives R(s)\approx 0.25 at 1.6 GeV. Taking a representative selected yield N\sim 10^{3} – 10^{4} , we obtain \delta d_{s}\sim(2.3\text{--}7.4)\times 10^{-16}\,e\cdot\mathrm{cm}\,. (25) A prospective VEPP–2000 data set of L\sim 1~\mathrm{fb}^{-1} Rastigeev:2024tek would improved the precision by an order. The overall normalization largely cancels in asymmetries. Similar measurements can be performed at BESIII at center-of-mass energies from 2.00 to 3.08 GeV, where the dominant resonance is expected to be the K_{2}^{\ast} BESIII:2022wxz , with an event yield of \mathcal{O}(10^{5}) . Here we focus on the J/\psi resonance, which is dominated by the K^{\ast}(892) contribution BESIII:2019apb . From the branching fraction of {\cal B}(J/\psi\to K^{+}K^{-}\pi^{0})=(2.88\pm 0.01\pm 0.12)\times 10^{-3} , we find F_{V}(m_{J/\psi}^{2},0,0)=-4.0\times 10^{-3} GeV -3 which gives R(s)=0.11\text{--}0.69 from Eq. ( 22 ). At BESIII, 182{,}972 candidate events were observed in 2019 BESIII:2019apb , and the J/\psi sample size has increased by about 45 times since 2009 BESIII:2021cxx . The expected statistical precision of A_{T} is around 3.5\times 10^{-4} , resulting in \displaystyle\delta d_{s}\;\simeq\;(0.3\text{--}1.8)\times 10^{-17}\;e\cdot\mathrm{cm}\,, (26) which can achieve better constraint than the one from d_{\Lambda} in Eq. ( 1 ) and ready to be tested at BESIII. The uncertainties of d_{s} quoted here mainly comes from the unknown of F_{A} . A solid determination can be achieved once F_{A}(s,0,0) is determined from a nonperturbative method. In general, all CP violating operators may contribute to F_{A} and may affect the constraint for d_{s} here. A detailed analysis with CP-odd parameters such as chromo-EDM will be presented elsewhere. The Super \tau Charm Facility will increase the production of J/\psi by two orders of magnitude STCF:2023CDR_PhysDet ; STCF:2025CDR_Acc and \delta d_{s} by an order of magnitude, comparing to BESIII. In addition, a border region of \sqrt{s} range from 4.2 GeV to 7 GeV may be explored. Complementary information may also come from Belle II, which aims at an integrated luminosity of \sim 50~\mathrm{ab}^{-1} Belle2LumiProj2024Dec , providing enormous statistics for exclusive hadronic final states with excellent charged–hadron identification. Although Belle II operates near the \Upsilon(4S) , the initial-state-radiation program enables contiguous measurements of e^{+}e^{-}\to K^{+}K^{-}\pi^{0} over a wide effective \sqrt{s} range BaBar:2007ceh . V Conclusion We evaluate the amplitude of \gamma^{\ast}\to K^{+}K^{-}\pi^{0} using a gauge-invariant basis of transverse vectors, introducing CP-even and CP-odd form factors F_{V} and F_{A} . Assuming C-parity invariance, we express the amplitude in the \gamma^{\ast} rest frame and identify the T-odd interference observable \vec{V}_{\perp}\cdot\vec{A}_{\perp} . Within \chi PT, the CP-even form factor arises from the WZW term, while a strange-quark electric dipole moment contributes via axial vector currents. We match the underlying operators onto chiral fields and obtain F_{V} and F_{A} in the chiral limit. The interference between the topological anomaly and d_{s} is a foreseeable result, since both break intrinsic parity. We define a signed interference asymmetry A_{T} to isolate d_{s} , and calculate the statistical sensitivity \delta d_{s} . Numerical estimates are performed for e^{+}e^{-}\to K^{+}K^{-}\pi^{0} at CMD-3, showing sensitivities at the 10^{-16}\,e\cdot\mathrm{cm} level with current data and better with higher luminosity. We extend the analysis to J/\psi\to K^{+}K^{-}\pi^{0} and note that existing BESIII samples could reach the order of \delta d_{s}\sim 10^{-18}\,e\cdot\mathrm{cm} . The asymmetry construction cancels overall normalization and many detector effects, providing a robust strategy to probe CP-violating strange-quark dipole moments.