Analog-antianalog isospin mixing in 47K $\beta$ decay The decay widths of isobaric analog resonances are modelled well by analog-antianalog isospin mixing via the Coulomb interaction. However, the effects of isospin mixing of the antianalog on nuclear beta decay angular correlations are generally measured to be much smaller than predicted. We have measured the isospin mixing of the I$^\pi$= 1/2$^+$ E$_x$=2.x MeV state in $^{47}$Ca with the isobaric analog 1/2$^+$ state of $^{47}$K. Using the TRIUMF atom trap for $\beta$ decay, we have measured a nonzero asymmetry of the emitted $^{47}$Ca with respect to the initial $^{47}$K spin polarization, which supported by the beta asymmetry implies a nonzero ratio of Fermi to Gamow-Teller matrix elements $M_F/M_GT$= -0.096$\pm$0.037. Interpreting as mixing between this state and the isobaric analog state implies a Coulomb matrix element of 72 $\pm$ 29 keV. The latter is an order of magnitude larger than observed on $\beta$ asymmetry measurements in similar decay systems, which we attribute to the simplicity of $^{47}$K and $^{47}$Ca states near doubly-closed shells, and thus a relatively unfragmented antianalog state. Intro The neutron beta decays to its isobaric analog state, the proton, as does tritium. Many other isotopes undergo beta minus decay to states of same spin I and parity pi, but because of the extra Coulomb energy at higher Z, decay to the isobaric analog state is energetically forbidden. So the Gamow-Teller operator dominates, while the Fermi operator linking isobaric analog states is only allowed if some low-lying final state of same Ipi is mixed by an isospin-breaking interaction with the excited isobaric analog. We see such isospin breaking in an I^pi=1/2^+ state in the 47Ca nucleus 80\% fed by the beta decay of 47K. Interference between Gamow-Teller and isospin-suppressed Fermi amplitudes produces a small asymmetry of the progeny recoil direction with respect to the initial nuclear spin, which we measure with TRIUMF's Neutral Atom Trap for beta decay (TRINAT). Our result below from a weighted average of recoil and beta asymmetries is Mf/Mgt=-0.096+-0.037, implying a Coulomb mixing matrix element 72+-26 keV, an order of magnitude larger than measured in other beta asymmetry measurements in nearby nuclei. Since 47Ca and 47K are near closed shells, that single known 47Ca 1/2+ state may contain much of the antianalog configuration and its predicted 190 keV mixing matrix element with the analog [Auerbach, Loc NPA 2022]. Sensitivity to time reversal-odd parity-even (TOPE) inherently isovector [Simonius PRL 1997] N-N interactions through a beta-nu-spin correlation is thought to be enhanced in these systems, as they are referenced to Coulomb rather than strong interactions [Barroso, Blin-Stoyle PLB 1973], while in this 47K decay case the nuclear matrix elements of such an isovector TOPE interaction may also escape fragmentation. Experiment: TRIUMF neutral atom trap Average of 1,000 atoms trapped over a 16 hr time period. Apparatus figure including trapped atoms image. Polarization by direct optical pumping, measured for the decaying atomns by comparing excited state population for circular polarized and linearlhy polarized light, measured by nonresonant photoionization. For spin-1/2 all the transitions have equal probabilities, making the deduced nuclear polarization simple to determine: Iz/I = 0.96 $\pm$ 0.04. Backgrounds from untrapped atoms. t1/2=19 sec and trap half-life 10 sec implies more then half the atoms decay after leaving the trap. There are finite backgrounds in the main eMCP-rMCP channel below, measured by deliberately releasing atoms from the trap. Contributions from other transitions. The 1/2+ to 1/2+ branch is 80$\pm$2 \% (J. Smith et al.) The 1/2+ to 3/2+ transitions total 20\%. They have no recoil asymmetry. We dilute our simulations below appropriately. The beta asymmetry is X for 1/2+ and Y for 3/2+ final states. The weighted average is Z +_ A. We need to fold this into the prediction of pseudoAbeta below. Results: Main channel is ion MCP wrt shakeoff electrons produced by charge states 2+ and higher. TOF spectrum shows +2 to +4. Their asymmetry wrt to polarization axis is shown to be nonzero, directly implying a nonzero Fermi contribution Modelled by a numerical integration including one 2 MeV gamma emitted The result is Mf/Mgt= -0.102 $\pm$ 0.041 Arecoil uncertainty budget: Z fit range 0.012 bkg 6$\pm$4\% 0.014 Polarization 0.004 Branching ratio negligible Fit Statistics 0.037 total 0.041 Beta's in coincidence with 47Ca recoils are also measured. If all 47Ca were measured, this would be a measurement of the beta asymmetry: however, some 47Ca escape the MCP, perturbing the asymmetry of beta's in coincidence by a well-defined combination of beta-neutrino correlation and its spin dependence. The uncertainty budget is in Table The result is consistent in sign with the recoil observable but with larger uncertainty. Subsection: critical to the $\beta$ asymmetry is discriminating $\beta$'s from $\gamma$'s, with expected prodution ratio 1:2, using conventional deltaE-E technique. Subsubsection delta E DSSSD Here backgrounds are negigibly small. Uncertainty budget for pseudoAbeta: Polarization 0.023 Backscatter correction of -0.012+-20\% -> 0.0024 Efield 0.025 Branching ratios ? Fit statistics 0.077 total 0.084 Recoil order corrections: We estimate the recoil order correction for weak magnetism for the 1/2+ -> 1/2+ transition assuming M_W/M_GT is given by a single-particle expression... For such a transition there is no contribution from orbital angular momentum, so the result is the nucleon value. Since the correction is small, we consider this adequate accuracy. Weighted average is -0.096$\pm$0.037 Comparison with theory Implications for future time-reversal measurement