HOLMES

The experiment HOLMES, founded by the European Research Council, will perform a calorimetric measurement of the energy released in the electron capture of 163Ho to directly measure the neutrino mass with a sensitivity of ∼ 1 eV. This approach allows to eliminate the problematics connected to the use of external sources and the systematic uncertainties arising from decays on excited states. Such measurement will be performed with low temperature thermal detectors, where the decay energy is converted into a temperature signal measured by sensitive thermometers. HOLMES, besides of being an important step forward in the direct neutrino mass measurement with a calorimetric approach, will also establish the potential of this approach to extend the sensitivity down to 0.1 eV and lower. The best configuration has been defined with Monte Carlo simulations: HOLMES will collect about 3 × 1013 decays with 1000 detectors characterized by an instrumental energy resolution of the order of the eV and a time resolution of few microseconds. For a measuring time of 3 years, this translates in a total required 163Ho activity of about 300 kBq, equivalent to about 6.5 × 1016 163Ho nuclei, or 18 µg. The HOLMES detectors will have 163Ho implanted into Gold absorber coupled to Transition Edge Sensors, which will be read using microwave multiplexed rf-SQUIDs in combination with a ROACH2 based acquisition system. An extensive R&D activity is in progress in order to maximize the multiplexing factor while preserving the performances of the individual detectors. R&D activities aimed at optimizing the single detector performances, the 163Ho isotope production and embedding are in progress and will converge in a preliminary measurement of an array of 16 detectors planned by the end of 2016. We outline here the HOLMES project with its technical challenges, its status and perspectives.


Introduction
The end-point study of the spectrum of beta decay is currently the one and only experimental method which can provide a model independent measurement of the absolute scale of neutrino mass. Isotopes with low decay energy Q are preferable in order to maximize the statistics in the energy region of the end-point. 163 Ho decays via electron capture on 163 Dy with a very low Q of 2.8 keV [1]: the proximity of the M1 resonance to the end-point further enhances the statistics in the region of interest, making 163 Ho a very promising choice for the neutrino mass measurement. Within this framework, the European Research Council has funded HOLMES, an experiment to directly measure the neutrino mass [2]. HOLMES will perform a calorimetric measurement of the energy released in the electron capture decay of 163 Ho [3]. In a calorimetric measurement, the energy released in the decay process is entirely contained into the detector, except for the fraction taken away by the neutrino, eliminating both the problematics connected to the use of an external source and the systematic uncertainties arising from decays on excited final states. The most suitable detectors for this approach are low temperature thermal detectors, where all the energy released into an absorber is converted into a temperature signal that can be measured by a sensitive thermometer directly coupled to the absorber. In order to acquire the needed statistics while keeping the pile-up contribution as low as possible, HOLMES will deploy a large array of low temperature microcalorimeters with implanted 163 Ho nuclei. The resulting mass sensitivity will be as low as ∼1 eV/c 2 [2]. HOLMES will be an important step forward in the direct neutrino mass measurement with a calorimetric approach as an alternative to spectrometry.

Detectors
In order to reach a sensitivity of the order of 1 eV/c 2 it is necessary to collect statistics higher than 10 13 decays. To fulfill this requirement, the optimal experimental configuration has been defined with Monte Carlo simulations: in its optimal configuration, HOLMES will collect about 3×10 13 decays with an energy resolution of about 1 eV FWHM and a time resolution of about 1 μs. For a total measuring time of 3 years, this requires a total 163 Ho activity of about 300 kBq.
Deploying an array of 1000 detectors, each pixel must contain a 163 Ho activity of about 300 Bq. The total activity is given by about 6.5×10 16 163 Ho nuclei, equivalent to 18 μg. The detectors for HOLMES will have absorbers made of gold coupled to a Transition Edge Sensor (TES), a sensitive thermometer that exploit the steep variation of resistance of a superconductor across its transition. A series of TES prototypes suspended on a Si 2 N 3 membrane were produced at NIST and tested both at NIST and in Milano [4], achieving an energy resolution FWHM of 7 eV at the energy of the K α lines (5.9 keV) of a 55 Fe source [5].

Readout
The ideal multiplexing technique for such an experiment should preserve the native performances of the detectors both in terms of energy and time resolution and at the same time it should allow the readout of a large number of detectors with a conveniently small number of readout lines. For HOLMES the microwave frequency domain readout [6] was chosen. In this scheme, the variation of current flowing in the TES, biased at constant voltage, is sensed by a rf-SQUID amplifier which is inductively coupled to a λ/4-wave resonator. An increase of temperature seen by the TES is then translated into a variation of magnetic flux read by the SQUID, causing the SQUID inductance, and hence the resonant frequency, to vary. Each SQUID is then coupled to a resonant circuit which is designed to oscillate at a unique frequency. To linearize the periodic response of the SQUID, a linearly increasing DC signal (a ramp signal) is applied. The resonators will be probed by a comb of signals, each one at the resonant frequency of one resonator. In this picture, many resonators are capacitively coupled to a single feedline, allowing the readout of multiple channels with a rather simple cryogenic setup. Finally, the detector signal is reconstructed with the homodyne technique. The readout of the detectors will be done with the Reconfigurable Open Architecture Computing Hardware (ROACH2 [7]) system, whose bandwidth of 550 MHz sets the maximum number of multiplexed channels to about 30.

Isotope production
The 18  produce ≈ 270 kBq of 163 Ho in 1 week starting from 1 mg of 162 Er. In this scenario, considering the efficiencies of purification and implantation, 80 mg of 162 Er irradiated for 7 weeks should provide the amount of 163 Ho required by HOLMES. On the other hand, the burn-up process 163 Ho(n,γ) 164 Ho, which has a non negligible cross section around 200 barns, will affect the total yield reducing it. Other issues could also arise from the undesirable production of 166m Ho, which depends upon the abundance of 164 Er; this isotope is particularly critical because it decays beta with a half life of 1200 years causing background in the energy region below 5 keV, i.e. the region of interest for the 163 Ho electron capture.
In case other Er isotopes are present in the enriched Er 2 O 3 target, the neutron irradiation of the sample could create long-living isotopes such as 170 Tm and 171 Tm with very high activities. For this reason a chemical pre-purification and post-purification is performed at the Paul Scherrer Institute (PSI, Villigen, Switzerland) by means of ion-exchange chromatography [9], which ensures an overall efficiency of around 79%.

Implantation
The 163 Ho will be embedded in the gold absorber with a custom ion implanter, consisting of four main components: a sputter ion source, an acceleration section, a magnetic/electrostatic mass analyzer, and a magnetic vertical steering stage. This setup is being designed and optimized to separate the 163 Ho from contaminants not removed by chemical purification such as 166m Ho. This system will be integrated with a vacuum chamber that will allow the simultaneous evaporation of gold both to control the concentration of 163 Ho and to deposit the final layer of gold to prevent the oxidization of 163 Ho, which could cause a chemical shift of the end-point. The metallic cathode for the ion source will be made by a Gold matrix containing metallic 163 Ho which will be produced in the Holmium evaporation chamber. In this latter chamber a Tantalum heated Knudsen cell will be used for the reduction and distillation of Holmium from the irradiated Er 2 O 3 . The Ho 2 O 3 /Er 2 O 3 powder will be thermally reduced at about 2000 K using the reaction Ho+2Y(met)→2Ho(met)+ Y 2 O 3 .