betaFeedback

WorksPhysics and Astronomy

New high-sensitivity search for neutron to mirror-neutron oscillations at the PSI UCN source

Abstract

In a search for potential signals of neutron (n) to mirror-neutron (n') oscillations, a collaboration centered at the Paul Scherrer Institute (PSI) investigated the remaining parameter space claimed by anomalies with a dedicated high-sensitivity apparatus. An elaborate magnetic-field-mapping analysis and Monte Carlo simulation of the cumulative n-n' oscillation probabilities along the neutron trajectories inside the storage vessel complemented the neutron data analysis. Magnetic fields were scanned in the range 5 microT < B < 109 microT. No evidence of anomalous neutron losses was found. Consequently, new limits for the n-n' oscillation time constant were set. The parameter space, previously claimed for potential signals, has been excluded to 99.98 %.

A parallel mirror sector of particles as an exact duplicate of the ordinary particle sector was proposed a long time ago motivated by the issues of parity [1, 2, 3, 4]. In this picture all ordinary particles: the electron ee, proton pp, neutron nn etc., have mass degenerate mirror partners: e′e', p′p', n′n' etc. which are sterile to our Standard Model interactions. Besides gravity, mirror particles can interact with ordinary matter only via specific feeble interactions [5, 6]. Mirror matter could account for dark matter, provided that during the evolution of the Universe it had a lower temperature than the ordinary sector [7, 8, 9, 10].

Any neutral particle, elementary as e.g. neutrinos or composite as the neutron, can undergo a mixing with its mass-degenerate mirror partner. For example, ordinary-mirror (active-sterile) neutrino oscillations were discussed in [11, 12, 13]. Interestingly, interactions that induce such mixing violate B−LB-L in each sector. Thus, in the light of Sakharov’s hypothesis [14], they can co-generate both ordinary and mirror baryon asymmetries [15, 16] and thus naturally explain the concordance between dark and ordinary matter fractions in the Universe, ΩB′/ΩB≃5\Omega_{\mathrm{B}}'/\Omega_{\mathrm{B}}\simeq5 [17, 18].

The neutron and mirror-neutron mass mixing εnn′+h.c.\varepsilon_{nn'}+\mathrm{h.c.}, in contrast to the neutron-antineutron mixing [19], is not severely restricted by the experimental limits; they allow an n−n′n-n' oscillation time, τnn′=ℏε−1\tau_{nn'}=\hbar\varepsilon^{-1}, as small as few seconds [20], in agreement also with astrophysical bounds from cosmic rays [21, 22] and neutron stars [23, 24, 25, 26]. Fast n−n′n-n' transitions are not easily detectable as far as they are suppressed by environmental factors as the presence of ordinary and/or mirror matter or of ordinary and/or mirror magnetic fields [20, 27, 28]. Various proposals for experimental searches via neutron disappearance n→n′n\to n' and regeneration n→n′→n/nˉn\to n'\to n/\bar{n}, and neutron spin-precession distortion can be found in [27, 29, 30, 31, 32, 33]. The possible effects for neutron lifetime measurements and exotic nuclear transitions were discussed in [34, 35, 36].

The observation that the oscillation time τnn′\tau_{nn'} may be less than the neutron lifetime [20] triggered an intense experimental search for anomalous n→n′n\to n' losses using ultra-cold neutrons (UCNs). A series of experiments [37, 38, 39, 40, 41, 42, 43], performed at various magnetic fields testing energy degeneracy, established significant upper bounds on τnn′\tau_{nn'}. In a reanalysis of data from [37, 40] the authors of [28, 42] found significant deviations from the null-hypothesis, namely 3σ3\sigma and 5σ5\sigma effects. The data of [42] have shown a 2.5σ2.5\sigma deviation. They could explain the deviations via oscillations in the mass-degenerate n−n′n-n' system in the background of a mirror magnetic field, B′B', from mirror-photons [27, 28]. The next storage experiment performed by the nEDM collaboration at PSI [43, 44] has excluded a big part of the parameter space for the above deviations but some parameter regions were left.

In an effort to test the remaining parameter space, a collaboration at PSI assembled a new experiment with a dedicated apparatus described in [45, 46]. The measurement series was performed in 2021. In this Letter we report on the analysis and the obtained exclusion limits.

In the UCN storage experiments, n→n′n\to n' transitions would manifest as additional neutron losses unaccounted for by other loss channels such as neutron decay or during wall collisions. Since n′n' is sterile against conventional interactions, at each wall collision it would escape the storage chamber and thus elude detection. In the mass degenerate neutron – mirror-neutron system, the n→n′n\to n' transition probability as a function of free flight time tt between collisions is [27, 28]:

PBB′nn′(t)=sin⁡2[(ω−ω′)t]2τnn′2(ω−ω′)2+sin⁡2[(ω+ω′)t]2τnn′2(ω+ω′)2+(sin⁡2[(ω−ω′)t]2τnn′2(ω−ω′)2−sin⁡2[(ω+ω′)t]2τnn′2(ω+ω′)2)cos⁡β.(1)P_{BB'}^{nn'}(t)=\frac{\sin^2[(\omega-\omega')t]}{2\tau_{nn'}^2(\omega-\omega')^2}+\frac{\sin^2[(\omega+\omega')t]}{2\tau_{nn'}^2(\omega+\omega')^2} +\left(\frac{\sin^2[(\omega-\omega')t]}{2\tau_{nn'}^2(\omega-\omega')^2}-\frac{\sin^2[(\omega+\omega')t]}{2\tau_{nn'}^2(\omega+\omega')^2}\right)\cos\beta. \tag*{(1)}

Here ω=∣μnB∣/2ℏ\omega=|\mu_nB|/2\hbar and ω′=∣μnB′∣/2ℏ\omega'=|\mu_nB'|/2\hbar, and β\beta is the angle between the ordinary magnetic field BB and the mirror field B′B' vectors. This equation is valid as long as both fields are homogeneous and β\beta stays constant during the time tt.

In the limit B′→0B'\to0, the last term in Eq. (1) vanishes, and the n−n′n-n' oscillation probability is suppressed by the ordinary magnetic field, independently of its direction. In this case, the most stringent experimental limit τnn′>414 s\tau_{nn'} > 414\,\mathrm{s} (90% C.L.) was obtained in [38]. However, this limit becomes invalid if a large enough mirror field is present on Earth (assumed to be rotating with it [27]), B′≳0.1 μTB' \gtrsim0.1\,\mu\mathrm{T}. In this case, the transition probability, Eq. (1), is a function of both magnitude and direction of the applied magnetic field B\mathbf{B}. The oscillations will be suppressed at B=0B = 0 and resonantly amplified for B≈B′B \approx B'. Because the magnitude and direction of the mirror field are unknown, the magnetic field B\mathbf{B} needs to be scanned.

Our experiment employs tunable magnetic fields in the range 5 μT<B<109 μT5\,\mu\mathrm{T} < B < 109\,\mu\mathrm{T}. We work with the directional asymmetry ABA_B as the experimental observable [27]:

AB=n+B(ts)−n−B(ts)n+B(ts)+n−B(ts)=exp⁡(−msPBB′nn′)−exp⁡(−msP−BB′nn′)exp⁡(−msPBB′nn′)+exp⁡(−msP−BB′nn′).(2)A_B = \frac{n_{+B}^{(t_s)} - n_{-B}^{(t_s)}}{n_{+B}^{(t_s)} + n_{-B}^{(t_s)}} = \frac{\exp(-m_s P_{BB'}^{nn'}) - \exp(-m_s P_{-BB'}^{nn'})}{\exp(-m_s P_{BB'}^{nn'}) + \exp(-m_s P_{-BB'}^{nn'})}. \tag*{(2)}

Here, n±B(ts)n_{\pm B}^{(t_s)} represent neutron counts after a storage time tst_s, when applying magnetic fields of opposite polarities, ±Bez\pm B\mathbf{e}_z. With ez\mathbf{e}_z we denote the unit vector along the vertical axis of the storage vessel. PBB′nn′P_{BB'}^{nn'} is the mean probability, Eq. (1), for a time-lapse between wall collisions, and msm_s is the mean number of collisions of UCNs detected after storage time tst_s. All conventional losses cancel since they are independent of magnetic fields.

Until now, we described the case where B\mathbf{B} is homogeneous. However, this strong assumption for B\mathbf{B} cannot be made in general. Due to the remaining small magnetization of the storage vessel and external field disturbances, there can be an offset in magnitude between the fields set with opposite polarities under the conditions of the experiment described here. One can also have non-negligible transversal field components that do not exhibit rotational symmetry. Due to these components, the projection cos⁡β\cos\beta between the local B\mathbf{B} and B′\mathbf{B}' changes along the UCN trajectory. Assuming B′\mathbf{B}' from Earth is locally homogeneous and static, we parametrize B′\mathbf{B}' by its magnitude B′B', the polar angle bb, and the azimuth angle aa. Similarly, we take the ordinary field vector B\mathbf{B} as a function of a position vector x\mathbf{x}. Then

B′=B′(sin⁡bcos⁡asin⁡bsin⁡acos⁡b),B=B(x)(sin⁡θ(x)cos⁡ϕ(x)sin⁡θ(x)sin⁡ϕ(x)cos⁡θ(x))(3)\mathbf{B}' = B' \begin{pmatrix} \sin b \cos a \\ \sin b \sin a \\ \cos b \end{pmatrix}, \quad \mathbf{B} = B(\mathbf{x}) \begin{pmatrix} \sin\theta(\mathbf{x})\cos\phi(\mathbf{x}) \\ \sin\theta(\mathbf{x})\sin\phi(\mathbf{x}) \\ \cos\theta(\mathbf{x}) \end{pmatrix} \tag*{(3)}

The local cos⁡β\cos\beta projection in Eq. (1) thus depends on x\mathbf{x}:

cos⁡(β(x))=sin⁡bcos⁡asin⁡θcos⁡ϕ+sin⁡bsin⁡asin⁡θsin⁡ϕ+cos⁡bcos⁡θ.(4)\cos(\beta(\mathbf{x})) = \sin b \cos a \sin\theta\cos\phi + \sin b \sin a \sin\theta\sin\phi+ \cos b\cos\theta. \tag*{(4)}

In the case of an inhomogeneous magnetic field, we consider the local increment in the oscillation probability, that is the derivative of the probability function from Eq. (1) at a time tit_i, and an infinitesimally short duration. The oscillation probability for a single neutron during a single free flight path, PfP_f, after neglecting higher order terms, becomes the sum of all infinitesimal probabilities along the path. The oscillation probability of a single neutron for the entire duration of a storage measurement is obtained by summing the probabilities, PfP_f, for all free flight segments. This is equivalent to msPBB′nn′m_sP_{BB'}^{nn'} in Eq. (2), or, in Monte Carlo (MC) simulations, one can directly obtain the cumulated probability for the entire time spent by the UCNs in the magnetic field. In the general case of a rotationally asymmetric inhomogeneous field and an offset when changing the field polarity, Eq. (2) can be expressed as:

τnn′=12∣AB∣∣FA0(B,B′)+FA1(B,B′)cos⁡b+FA2(B,B′)sin⁡bcos⁡a+FA3(B,B′)sin⁡bsin⁡a∣12.(5)\tau_{nn'} = \frac{1}{\sqrt{2\lvert A_B\rvert}}\left|F_A^0(\mathbf{B},\mathbf{B}') + F_A^1(\mathbf{B},\mathbf{B}')\cos b + F_A^2(\mathbf{B},\mathbf{B}')\sin b\cos a + F_A^3(\mathbf{B},\mathbf{B}')\sin b\sin a\right|^{\frac{1}{2}}. \tag*{(5)}

The terms FAi(B,B′)F_A^i(\mathbf{B},\mathbf{B}'), referred to later as "resonance functions", represent asymmetries from B\mathbf{B} and −B-\mathbf{B} contributions to the cumulated oscillation probabilities (considered as ≲1\lesssim1) and are calculated in MUCUN [47, 48, 49] simulations. The term FA0F_A^0 arises due to the vertical field gradient and offset when changing field polarity. The terms FA1,FA2,FA3F_A^1,F_A^2,F_A^3 also include the rotational dependency of the B\mathbf{B} vector. The direction of the mirror field B′\mathbf{B}' is unknown, therefore τnn′\tau_{nn'} is given as an explicit function of the angles aa and bb.

A schematic view of the apparatus is shown in Figure 1. It was installed in Area West of the PSI UCN source [48, 49, 50, 51] and upgraded in 2021 [45]. UCNs entered through the West-1 beamport shutter (1) and traveled down the Ni/Mo-coated glass guides (2). By passing through the vacuum-tight shutter at the end of the horizontal guide (3) they entered a 1 m high, 1.5 m3^3 large evacuated storage vessel (4) made of stainless steel. During monitoring and after storage, the neutrons were released through a fast butterfly shutter [52] (5) and counted by a 20 cm ×\times 20 cm CASCADE [53] UCN detector (6). A set of eight rectangular coils (7), spanning across the entire vacuum tank of the storage vessel was used to generate the target magnetic field. The top lid of the storage vessel (8) could be opened to insert a device to perform spatially resolved measurements of the magnetic field in the storage volume. The setup was surrounded by a wall of concrete blocks for radiation protection (9).

Cut through the schematic (CAD) experimental setup

Figure 1. Cut through the schematic (CAD) experimental setup: (1) West-1 beamport shutter, (2) UCN guides, (3) horizontal guide shutter, (4) storage vessel, (5) butterfly shutter, (6) UCN detector, (7) magnetic field coils, (8) top lid of the storage vessel, (9) radiation shielding. Figure adapted from [46].

The magnetic field was generated by the eight coils wound on aluminium rails arranged around the storage vessel (compare also to [54]). Two independent, rectangular coils measuring each 3 m ×\times 4.6 m with distances of 1.8 m between them produced the fields in XX- and ZZ-direction. The field along the beamline axis (YY) was generated by two sets of two 3 m ×\times 3 m coils located at distances of ±0.9\pm0.9 m and ±2.2\pm2.2 m from the center of the storage vessel. The two neighbouring YY-coils were driven in series, resulting in a total of six independent currents I=(IX+,IX−,IY+,IY−,IZ+,IZ−)\mathbf{I} = (I_X^+, I_X^-, I_Y^+, I_Y^-, I_Z^+, I_Z^-). The surrounding field was constantly monitored by 16 three-axis fluxgates mounted on a wooden support structure around the storage vessel (see details on their sensitivities in [46]).

To characterize the magnetic field generated by the coils we performed spatially resolved measurements of the magnetic field response to the currents both inside the storage vessel, using a dedicated device, called a mapper, and the aforementioned monitoring fluxgates. The mapper (Figure 5 in Supplemental Material) held five fluxgates, which were mounted on a wood-polypropylene-profile spanning the diameter of the vessel. The profile was attached to a long, hollow aluminum rod that had to be inserted into the vessel from the top. The height and rotational angle of the mapper were tracked using potentiometer-based position sensors.

Picture of the mapper inside the electro-polished stainless steel UCN storage vessel

Figure 5. Picture of the mapper (see text) inside the electro-polished stainless steel UCN storage vessel.

A single map consisted of rotational scans at several heights. To characterize the magnetic field inside the storage vessel as a function of the applied coil currents, at least seven maps at linearly independent current combinations had to be taken.

Using the method proposed in [55] one can describe an arbitrary magnetic field as an infinite sum of harmonic polynomials as

B(r)=∑l,mGl,mΠl,m(r).(6)\mathbf{B}(\mathbf{r}) = \sum_{l,m} G_{l,m}\Pi_{l,m}(\mathbf{r}). \tag*{(6)}

The expansion coefficients Gl,mG_{l,m} are the generalized gradients for modes of degree ll. The harmonic polynomials Πl,m(r)\Pi_{l,m}(\mathbf{r}) are orthogonal and each basis state satisfies Maxwell’s equations.

By fitting the above field model to the data extracted from the magnetic field maps, we obtained an expression for the magnetic field inside the UCN storage vessel as a function of coil currents and position. Due to the design of the coils, it was sufficient to take into account gradients up to the 2nd order. These expansion coefficients served as magnetic field input for simulations with the MCUCN code, which were calibrated with measurements of the UCN counts after different storage periods and corresponding emptying time distributions [46].

A single UCN storage measurement was called a cycle. A detailed description of the order of events happening during one measurement cycle can be found in [46]. The detection

rate of the UCN counts as a function of time during a cycle is shown in Figure 2. At t=0t = 0 a proton beam pulse hit the lead spallation target of the UCN source. The beamport and horizontal guide shutters were open and neutrons entered the storage vessel. During the monitoring period, the butterfly shutter to the detector was briefly opened. After a storage time of 180 s, the butterfly shutter was opened again. The UCNs were released and counted for 95 s.

UCN counts during one storage measurement cycle

Figure 2. UCN counts during one storage measurement cycle (adapted from [46]).

To compensate for long-term drifts in UCN counts, the direction of the applied magnetic field was altered from one cycle to the next. This was done according to the pattern ABBA BAAB, where A ideally corresponds to a target field B=+∣B∣ez\mathbf{B} = +|B|\mathbf{e}_z and B to B=−∣B∣ez\mathbf{B} = -|B|\mathbf{e}_z. A complete set of eight cycles was grouped into a sequence. A sequence was the smallest unit of measurements considered in this analysis.

The UCN DAQ performed 5σ5\sigma-clipping of the normalized counts to reject sequences with outliers. These were caused by sequencing errors, proton beam interlocks, UCN source control problems, or synchronization problems with the optical link for the detector readout, leading to clear outliers in the monitoring or storage counts. Data were also rejected in case of problems with the magnetic field control system.

Strong superconducting magnets from neighbor experiments sometimes caused local field changes of up to several tens of microtesla. While the control system of the magnetic field was able to keep the field stable against small outside disturbances, the coil current supplies did not deliver enough current to fully compensate zero-order changes that were caused by these superconducting magnets. The coils were also not designed to counteract changes in gradients higher than second order. The influence of our coils on the field inside the vacuum storage tank was known; however, the effect of external magnetic field changes on the fields inside the vacuum vessel could not be fully characterized. Measurement cycles taken under these conditions were rejected.

We calculate the asymmetry between A- and B-cycles for an individual sequence jj, AjA_j from

Aj=∑inA,i−∑inB,i∑inA,i+∑inB,i.(7)A_j = \frac{\sum_i n_{A,i}-\sum_i n_{B,i}}{\sum_i n_{A,i}+\sum_i n_{B,i}}. \tag*{(7)}

The normalized counts n{A,B},in_{\{A,B\},i} per cycle (i=1,2,3,4i=1,2,3,4) are the ratio of counts after storage and the monitoring counts. The mean asymmetry ⟨AB⟩\langle A_B\rangle per target field and their standard errors δ⟨AB⟩\delta\langle A_B\rangle are plotted in Figure 3, showing that the values of ⟨AB⟩\langle A_B\rangle are consistent with the null-hypothesis in which the measured values of ⟨AB⟩\langle A_B\rangle are compatible with a random sampling of a Gaussian distribution centered at zero and with a standard deviation of the measured uncertainty. Consequently, no n−n′n-n' oscillations were observed, within 95% C.L. for almost all target fields with or without correcting for the look-elsewhere-effect. After correcting for the look-elsewhere-effect with the Bonferroni procedure [56], n−n′n-n' oscillations were excluded within 95% C.L. also at the remaining fields B=9 μTB=9\,\mu\mathrm{T} and B=19.5 μTB=19.5\,\mu\mathrm{T}.

Mean asymmetry as a function of target field

Figure 3. Mean asymmetry ⟨AB⟩\langle A_B\rangle as a function of target field BB. The light grey line shows the limits of the null-hypothesis H0H_0 at the 95% C.L., the dark grey outline shows the same limit after correcting for the look-elsewhere-effect. ⟨AB⟩\langle A_B\rangle is consistent with H0H_0 within a standard error δ⟨AB⟩\delta\langle A_B\rangle.

We can construct limits for τnn′\tau_{nn'} according to Eq. (5) by combining the resonance functions FA0F_A^0, FA1F_A^1, FA2F_A^2, FA3F_A^3 we obtained from Monte Carlo simulations of the n−n′n-n' oscillation probabilities with the distributions for the asymmetry obtained from the UCN counts. From sampling the asymmetry ABA_B in Eq. (5) as a normal distribution around the measured mean ⟨AB⟩\langle A_B\rangle with standard error δ⟨AB⟩\delta\langle A_B\rangle from Figure 3, we obtained the 95% C.L. limits for τnn′\tau_{nn'} as a function of the mirror magnetic field B′B'. For every target field we obtain one function for τnn′/∣cos⁡(b)∣\tau_{nn'}/\sqrt{|\cos(b)|}, the ratio utilized for the anomalies in [42]. Taking the upper envelope of all resonance functions (thus exploiting the best sensitivity from each target field) yields the limit for a given combination of aa and bb. When sampling these unknown angular coordinates of the mirror-magnetic-field vector, the majority of the (a,b)(a,b) tuples provides limits that completely exclude all three anomaly signals, where they were not yet excluded by PSI-2009 [41] (B′>5.4 μTB'>5.4\,\mu\mathrm{T}).

they were not yet excluded by PSI-2009 [41] (B′>5.4 μTB'>5.4\,\mu\mathrm{T}). The lower envelope of these limits is plotted as "PSI-2025" in Figure 4. Within 99.98% of the full 4π4\pi solid angle, all previously claimed anomalies are excluded at 95% C.L. There remains a small region ∣cos⁡(b)∣−1<10−4|\cos(b)|-1<10^{-4} in which these anomalies are not excluded. In this direction of the mirror magnetic field vector, the apparatus had the lowest sensitivity for the oscillation time τnn′/∣cos⁡(b)∣\tau_{nn'}/\sqrt{|\cos(b)|}. The parameter space is three-dimensional. In order to quantify the correlation of the limits τnn′/∣cos⁡(b)∣\tau_{nn'}/\sqrt{|\cos(b)|} as a function of B′B' and the angular coordinates, aa and bb, we calculated the limit curves able to exclude the anomaly signal if it were increased by a factor of 2 and 5 (Figure 4). The exclusion solid angles obtained were 99.62% and 89.84% of 4π4\pi, respectively. For a detailed mapping of these solid angle regions we refer to the Supplemental Material (Figure 6).

Anomalous signal regions and limits as a function of the mirror-magnetic-field

Figure 4. The anomalous signal regions of Berezhiani-2018 [42] from data in [37] (3σ\sigma, red), [40] (5.2σ\sigma, blue) and [42] (2.5σ\sigma, green) together with the 95% C.L. limits on τnn′/∣cos⁡(b)∣\tau_{nn^\prime}/\sqrt{|\cos(b)|} from PSI-2009 [41], PSI-2021 [43] and this work as a function of the magnitude of the mirror-magnetic-field, B′B^\prime. Along with the most conservative PSI-2025 limit (black line) excluding 99.98% of the full solid angle (4π4\pi), two higher limit curves but with less solid angle exclusion (99.62% and 89.84%) are plotted, corresponding to levels of factor 2 (dashed line) and 5 (dotted line) above the anomalous signals.

Exclusion regions of the directional parameters and zoom-in to the solid angle regions

Figure 6. Top − the exclusion regions of the directional parameters, aa and cos⁡(b)\cos(b), of the mirror-magnetic-field vector corresponding to the limit curves from this work shown in Figure 4, see text. Bottom − zoom-in to the solid angle regions at values ∣cos⁡(b)∣−1<10−4|\cos(b)| - 1 < 10^{-4}, where the anomalous signals were not excluded (orange on the scale).

Additional limits obtained by studying the frequency variations in the neutron spin-precession [57, 58] regard small values of B′B' (<1 μT<1\,\mu\mathrm{T}), outside the range of Figure 4. Other complementary efforts include the search of n−n′n-n' mixing in beam experiments [59, 60, 61, 62] but their results are applicable when nn and n′n' states have substantial mass splitting and do not interfere with the parameter space analyzed in our paper.

In conclusion, in a dedicated experiment we have demonstrated that previously claimed anomalies in the B′B' range of 5–109 μT\mu\mathrm{T} could be excluded at 95 % C.L. in 99.98 % of the 4π4\pi solid angle of the directional coordinates.

Acknowledgments

Acknowledgments — The experiment would not have been possible without the excellent technical support of Michael Meier and Luke Noorda. The authors acknowledge the valuable assistance of many support groups at PSI, especially the BSQ group operating the UCN source, the accelerator operating crews, the ‘Hallendienst’ and the mechanical workshops. We are also grateful for support from the ETH IPA and D-PHYS mechanical workshops and the vocational training division.

ETH and PSI appreciate the financial support from the Swiss National Science Foundation through projects 162574 (ETH), 169596 (PSI), 172626 (PSI), 172639 (ETH), 178951 (PSI), 188700 (PSI), 196416 (ETH), 200441 (ETH), 10003932 (ETH). We further acknowledge funding from the ETH Career Seed Grant SEED-13 20-2 and the SNF spark programme grant CRSK-2_196416. The collaborators from the Jagiellonian University Cracow wish to acknowledge support from the National Science Center, Poland, under grant No. 2016/23/D/ST2/00715, No. 2018/30/M/ST2/00319 and No. 2020/37/B/ST2/02349, and also by the Minister of Education and Science under the agreement No. 2022/WK/07. The work of Z.B. was supported in part by the research grant No. 2022J4RK “PANTHEON: Perspectives in Astroparticle and Neutrino Theory with Old and New messengers” under the program PRIN 2022 funded by the Italian Ministero dell’Università e della Ricerca (MUR) and by the European Union – Next Generation EU.

References

Supplemental Material

Magnetic field mapper for the n−n′n-n' experiment

A 3-D magnetic mapper, as shown in Figure 5, was built in order to characterize in detail the inhomogeneous field configuration. The five three-axis fluxgates of the type FLC3-70 from Stefan Mayer Instruments (see red marks), FGiFG_i are mounted at various positions along the black PEEK-profile. The central axis of the mapper can be lifted and rotated to ensure good coverage of the vessel’s inner volume. For further details, see the main text.

Calculations of the excluded solid angle ranges

To obtain a more general quantitative relation between the height of the exclusion limit in τnn′/∣cos⁡(b)∣\tau_{nn'}/\sqrt{|\cos(b)|} and the excluded solid angle associated to it, we performed additional calculations by repeating the same analysis on the signal exclusion but setting the signals ’artificially’ higher. The anomaly signals were all upscaled by a factor of 2, 5 and 10. The resulting solid angle ranges, in which the signal could still be excluded at 95% CL by our results, are shown in Figure 6 top (full-scale) and bottom (zoom-in).

In Figure 6-Top, the color of each (a,cos⁡(b))(a,\cos(b)) tile indicates the highest factor that the corresponding limit τnn′/∣cos⁡(b)∣\tau_{nn'}/\sqrt{|\cos(b)|} was able to fully exclude. The solid angle areas for factors 1, 2, 5 and 10 decrease accordingly as 99.98%, 99.62%, 89.84% and 50.85% of 4π4\pi, respectively. The lack of horizontal symmetry in the patterns is caused by the rotationally non-symmetric field maps that characterize the coil system.

In Figure 6-Bottom, we zoomed-in to the vertical axis values close to ∣cos⁡(b)∣−1<10−4|\cos(b)|-1<10^{-4}, to magnify the orange zones, where the anomalous signals were not excluded (compare color scale in Top). The uneven distribution of the orange tiles is caused by the statistical fluctuations in the simulated resonance functions and the discrete sampling of the (a,cos⁡(b))(a,\cos(b)) plane.

References

  1. [1]T. D. Lee and C.-N. Yang, Phys. Rev. 104, 254 (1956).
  2. [2]I. Y. Kobzarev, L. B. Okun, and I. Y. Pomeranchuk, Sov. J. Nucl. Phys. 3, 837 (1966).
  3. [3]S. I. Blinnikov and M. Khlopov, Sov. Astron. 27, 371 (1983).
  4. [4]R. Foot and R. R. Volkas, Phys. Rev. D 68, 023504 (2003).
  5. [5]Z. Berezhiani, Int. J. Mod. Phys. A 19, 3775 (2004), arXiv:hep-ph/0312335.DOI
  6. [6]R. Foot, Int. J. Mod. Phys. A 19, 3807 (2004), arXiv:astro-ph/0309330.DOI
  7. [7]Z. Berezhiani, D. Comelli, and F. L. Villante, Phys. Lett. B 503, 362 (2001), arXiv:hep-ph/0008105.DOI
  8. [8]A. Y. Ignatiev and R. R. Volkas, Phys. Rev. D 68, 023518 (2003), arXiv:hep-ph/0304260.DOI
  9. [9]Z. Berezhiani, P. Ciarcelluti, D. Comelli, and F. L. Villante, Int. J. Mod. Phys. D 14, 107 (2005), arXiv:astro-ph/0312605.DOI
  10. [10]Z. Berezhiani, S. Cassisi, P. Ciarcellutti, and A. Pietrinferni, Astropart. Phys. 24, 495 (2006), arXiv:astro-ph/0507153.DOI
  11. [11]R. Foot and R. R. Volkas, Phys. Rev. D 52, 6595 (1995), arXiv:hep-ph/9505359.DOI
  12. [12]Z. G. Berezhiani and R. N. Mohapatra, Phys. Rev. D 52, 6607 (1995), arXiv:hep-ph/9505358.
  13. [13]E. K. Akhmedov, Z. G. Berezhiani, and G. Senjanovic, Phys. Rev. Lett. 69, 3013 (1992), arXiv:hep-ph/9205230.DOI
  14. [14]A. D. Sakharov, Pisma Zh. Eksp. Teor. Fiz. 5, 32 (1967).
  15. [15]L. Bento and Z. Berezhiani, Phys. Rev. Lett. 87, 231304 (2001), arXiv:hep-ph/0107281.DOI
  16. [16]L. Bento and Z. Berezhiani, Fortsch. Phys. 50, 489 (2002).
  17. [17]Z. Berezhiani, Eur. Phys. J. ST 163, 271 (2008).
  18. [18]Z. Berezhiani, Int. J. Mod. Phys. A 33, 1843048 (2018).
  19. [19]D. G. Phillips, II et al., Phys. Rept. 612, 1 (2016), arXiv:1410.1100 [hep-ex].DOI
  20. [20]Z. Berezhiani and L. Bento, Phys. Rev. Lett. 96, 081801 (2006), arXiv:hep-ph/0507031.DOI
  21. [21]Z. Berezhiani and L. Bento, Phys. Lett. B 635, 253 (2006), arXiv:hep-ph/0602227.DOI
  22. [22]Z. Berezhiani and A. Gazizov, Eur. Phys. J. C 72, 2111 (2012), arXiv:1109.3725 [astro-ph.HE].DOI
  23. [23]Z. Berezhiani, R. Biondi, M. Mannarelli, and F. Tonelli, Eur. Phys. J. C 81, 1036 (2021), arXiv:2012.15233 [astro-ph.HE].DOI
  24. [24]D. McKeen, M. Pospelov, and N. Raj, Phys. Rev. Lett. 127, 061805 (2021), arXiv:2105.09951 [hep-ph].DOI
  25. [25]Z. Berezhiani, Universe 8, 313 (2022), arXiv:2106.11203 [astro-ph.HE].arxiv.org/abs/2106.11203
  26. [26]I. Goldman, R. N. Mohapatra, S. Nussinov, and Y. Zhang, Eur. Phys. J. C 82, 945 (2022), arXiv:2203.08473 [hep-ph].DOI
  27. [27]Z. Berezhiani, Eur. Phys. J. C 64, 421 (2009), arXiv:0804.2088 [hep-ph].DOI
  28. [28]Z. Berezhiani and F. Nesti, Eur. Phys. J. C 72, 1974 (2012), arXiv:1203.1035 [hep-ph].DOI
  29. [29]Y. N. Pokotilovski, Phys. Lett. B 639, 214 (2006), arXiv:nucl-ex/0601017.DOI
  30. [30]Z. Berezhiani, M. Frost, Y. Kamyshkov, B. Rybolt, and L. Varriano, Phys. Rev. D 96, 035039 (2017), arXiv:1703.06735 [hep-ex].DOI
  31. [31]Z. Berezhiani, Eur. Phys. J. C 81, 33 (2021), arXiv:2002.05609 [hep-ph].DOI
  32. [32]Z. Berezhiani, R. Biondi, Y. Kamyshkov, and L. Varriano, MDPI Physics 1, 271 (2019), arXiv:1812.11141 [nucl-th].DOI
  33. [33]Y. Kamyshkov, J. Ternullo, L. Varriano, and Z. Berezhiani, Symmetry 14, 230 (2022), arXiv:2111.01791 [hep-ph].arxiv.org/abs/2111.01791
  34. [34]Z. Berezhiani, Eur. Phys. J. C 79, 484 (2019), arXiv:1807.07906 [hep-ph].DOI
  35. [35]Z. Berezhiani, LHEP 2, 118 (2019), arXiv:1812.11089 [hep-ph].arxiv.org/abs/1812.11089
  36. [36]Z. Berezhiani, Eur. Phys. J. C 76, 705 (2016), arXiv:1507.05478 [hep-ph].DOI
  37. [37]G. Ban et al., Phys. Rev. Lett. 99, 161603 (2007), arXiv:0705.2336 [nucl-ex].DOI
  38. [38]A. P. Serebrov et al., Phys. Lett. B 663, 181 (2008), arXiv:0706.3600 [nucl-ex].DOI
  39. [39]K. Bodek et al., Nucl. Instrum. Meth. A 611, 141 (2009).
  40. [40]A. P. Serebrov et al., Nucl. Instrum. Meth. A 611, 137 (2009), arXiv:0809.4902 [nucl-ex].
  41. [41]I. Altarev et al., Phys. Rev. D 80, 032003 (2009), arXiv:0905.4208 [nucl-ex].DOI
  42. [42]Z. Berezhiani, R. Biondi, P. Geltenbort, I. A. Krasnoshchekova, V. E. Varlamov, A. V. Vassiljev, and O. M. Zherebtsov, Eur. Phys. J. C 78, 717 (2018), arXiv:1712.05761 [hep-ex].DOI
  43. [43]C. Abel et al. (nEDM), Phys. Lett. B 812, 135993 (2021), arXiv:2009.11046 [hep-ph].DOI
  44. [44]P. MohanMurthy, A Search for Neutron to Mirror-Neutron Oscillations, Diss. ETH No. 26525, ETH Zurich, Zurich (2020).
  45. [45]N. J. Ayres et al., Symmetry 14, 503 (2022), arXiv:2111.02794 [physics.ins-det].
  46. [46]I. Rienäcker, Improving ultracold neutron yields and the search for mirror neutrons at the PSI UCN source, Diss. ETH No. 28498, ETH Zurich, Zurich (2022).DOI
  47. [47]G. Zsigmond, Nucl. Instrum. Meth. A 881, 16 (2018), arXiv:1709.05974 [physics.ins-det].DOI
  48. [48]G. Bison, B. Blau, M. Daum, L. Göltl, R. Henneck, K. Kirch, B. Lauss, D. Ries, P. Schmidt-Wellenburg, and G. Zsigmond, Eur. Phys. J. A 56, 33 (2020), arXiv:1907.05730 [physics.ins-det].DOI
  49. [49]G. Bison, M. Daum, K. Kirch, B. Lauss, D. Ries, P. Schmidt-Wellenburg, and G. Zsigmond, Eur. Phys. J. A 58, 103 (2022), arXiv:2110.12988 [physics.ins-det].DOI
  50. [50]A. Anghel, F. Atchison, B. Blau, B. van den Brandt, M. Daum, R. Doelling, M. Dubs, P.-A. Duperrex, A. Fuchs, D. George, L. Gültl, P. Hautle, G. Heidenreich, F. Heinrich, R. Henneck, S. Heule, T. Hoffmann, S. Joray, M. Kasprzak, K. Kirch, A. Knecht, J. Konter, T. Korhonen, M. Kuzniak, B. Lauss, A. Mezger, A. Mtchedlishvili, G. Petzoldt, A. Pichlmaier, D. Reggiani, R. Reiser, U. Roh rer, M. Seidel, H. Spitzer, K. Thomsen, W. Wagner, M. Wohlmuther, G. Zsigmond, J. Zuellig, K. Bodek, S. Kistryn, J. Zejma, P. Geltenbort, C. Plonka, and S. Grigoriev, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 611, 272 (2009), particle Physics with Slow Neutrons.
  51. [51]B. Lauss and B. Blau, SciPost Phys. Proc. 5, 004 (2021), arXiv:2104.02457 [nucl-ex].arxiv.org/abs/2104.02457
  52. [52]G. Bison, F. Burri, M. Daum, K. Kirch, J. Krempel, B. Lauss, M. Meier, D. Ries, P. Schmidt-Wellenburg, and G. Zsigmond, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 830, 449 (2016).
  53. [53]CDT GmbH, (www.n-cdt.com, accessed April 2025).
  54. [54]T. Brys et al., Nucl. Instrum. Meth. A 554, 527 (2005).
  55. [55]C. Abel et al., Phys. Rev. A 99, 042112 (2019), arXiv:1811.06085 [physics.ins-det].DOI
  56. [56]O. J. Dunn, Journal of the American Statistical Association 56, 52 (1961), https://doi.org/10.1080/01621459.1961.10482090.
  57. [57]P. MohanMurthy, A. R. Young, J. A. Winger, and G. Zsigmond, Symmetry 14, 487 (2022), arXiv:2201.04191 [nucl-ex].DOI
  58. [58]P. MohanMurthy and J. Winger, PoS SPIN2023 112 (2024).
  59. [59]G. Ban et al., Phys. Rev. Lett. 131, 191801 (2023), arXiv:2303.10507 [hep-ph].DOI
  60. [60]L. Broussard et al., Phys. Rev. Lett. 128, 212503 (2022), arXiv:2111.05543 [nucl-ex].DOI
  61. [61]F. M. Gonzalez et al., Phys. Rev. D 110, 072022 (2024), arXiv:2402.15981 [hep-ex].arxiv.org/abs/2402.15981
  62. [62]H. Almazán et al., Phys. Rev. Lett. 128, 061801 (2022), arXiv:2111.01519 [hep-ex].DOI

Paper details

vlnc:W_b2j4f59a
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents