- Email: [email protected]
The β-decay of 187Re studied with a cryogenic microcalorimeter
Nuclear Instruments
and Methods
in Physics
Research
A 370 (1996) 247-249
NUCLEAR INSTRUMENTS 8 METHODS IN PHYSICS RESEARCH Section A
ELSEWER
The P-decay of
’
87Re studied with a cryogenic microcalorimeter
F. Fontanelli,
F. Gatti*, A. Swift, S. Vitale
Abstract An experiment to study the P-decay of Ix7Re using cryogenic microcalorimeters is under course. The physical motivations of the experiment that can give a limit to the anti-neutrino mass are briefly discussed. Problems of the experiment and progresses obtained in the realisation of a detecting device are presented.
1. Introduction The kinematic limits to the electron anti-neutrino mass are presently obtained from experiments that, by means of high resolution magnetic or electrostatic spectrometers, measure the spectrum of the P-electrons emitted by a ‘H source in the end-point region. The effect of a finite anti-neutrino mass should manifest itself as a count deficit, in first approximation proportional to rnt, near the spectrum end-point. To set a meaningful limit to the neutrino mass, very small deviations must be excluded from the P-spectrum expected for m, = 0. The most likely value for ,rrt from all the recent results is negative, the particle data group average being -57+30 eV’ [I]. The negative rni value, that corresponds to an excess counting rate in the region approaching the end-point, has been speculated to arise from an incomplete knowledge of the final state effects or other systematic effects. To clarify the question of the squared negative mass, calorimetric experiments have been proposed in the past. The calorimetric method, where the full energy release following the P-decay is detected, removes the final state effect. Microcalorimeters are presently the only detectors with an energy resolution sufficient to set a sensitive limit to anti-neutrino mass and energy response independent of the kind of excitation. In particular. due to the limit in counting rate allowed by microcalorimeters. as explained in the following text, we have proposed the “‘Re as the p-emitter [3]. 2. Rhenium experiment In a calorimetric experiment, the full spectrum contributes to the detector counting rate, but only a very small fraction of decay events, proportional to (p”lQ)‘. (F” =
* Corresponding [email protected].
author.
Fax
+39
10
313358,
e-mail
0168.9002/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved SSDI 0168.9002(95)01 100-5
limit to be set to m,,), falls in the region of the spectrum sensitive to the anti-neutrino mass near the end-point. The source intensity must be limited to avoid pile-up problems, and the time needed to accumulate reasonable statistics in the end-point region may become exceedingly long unless Q is very small. With a tritium Q value of 18.6 keV and a microcalorimeter as total absorption spectrometer, the time needed to set an upper limit of 10 eV to my is not less than one year. The interest in the IX7Re P-decay originated since its Q value is the lowest known in nature, only 2.7 keV Because of the lower maximum decay energy of rhenium the allowed counting rate in the end-point region is much less stringent than for Tritium, reducing further the measure time to set a given limit to IPI,,. The theoretical analysis of the lX7Re P-spectrum shape is, at first glance, not too difficult, the decay being a first-forbidden unique transition. It is of note, however. that at such low energies great care has to be taken to check the validity of the theoretical approximations in evaluating the Fermi function F(Z. El and the spectral shape factor. Moreover, the decay of rhenium not as an isolated atom but in a crystal lattice is expected to produce an oscillatory modulation of the P-spectrum when the electron wavelength is not far from the lattice parameter [2]. The expected effect however is small and should be negligible above 1.5 keV Rhenium is a superconducting metal with critical temperature of 1.7 K. We have proved elsewhere [5,6] that a detector made with a superconducting rhenium single crystal coupled with a Ge NTD has energy thermalization that is reasonably fast and complete, provided that the operating temperature is kept above T=2X lo-J T, g 90 mK. It is then possible to build efficient microcalorimeters with a superconducting absorber, but, because of the lower limit on the operating temperature the overall heat capacity of the detector stays high limiting the achievable energy resolution.
VI. APPLICATIONS
248
F. Formrtelli
3. Progress
et ul. I Nucl. Instr. and Meth. in Phps. Res. A 370 (1996) 247-249
in detector development
In Figs. la and lb the P-spectrum and the Kurie-plot recently obtained with a test detector are shown. The two calibration X-ray peaks at 6490 and 5897 eV are from a “Fe source and are completely resolved. The test detector is made with a small natural metallic rhenium crystal and by an NTD germanium thermistor R17, 0.5 X 0.25 X 0.25 mm’ glued to the crystal with Ag epoxy. The rhenium mass is about 2 mg. imposed by the maximum allowed rate of “‘Re decays in the detector. Thermal and electrical connections to the refrigerator
100
60
60
40
zo
0
Energy (eV)
Fig. 2. View of the rhenium-Ge
NTD calorimetric
detector.
mixing chamber and electrical connections to a source follower, placed inside the refrigerator, are provided by two ultrasound bonded 20 pm aluminium wires. A schematic outline of the device is shown in Fig. 2. The amplified thermistor signal is recorded and analysed off-line: from each pulse we obtain, by adaptive filtering [7], an amplitude proportional to the energy release, and a shape factor [4] that measures the deviation of the pulse shape from a reference shape. The reference pulse shape is obtained by averaging several hundreds selected input pulses. The observed energy resolution at 5.9 keV, 55 eV RMS, is fully compatible with the signal to noise ratio. The noise level is scaling with the square root of the thermistor electrical resistance in the operating point (3.7 MO). as expected for the device Johnson noise: the contribution of the amplifier noise (0.8 nV/Hz”‘) is negligible. The signal amplitude is limited by the detector heat capacity, mainly due to the thermistor contribution, about IO pJ/K at 100 mK. Indeed the energy resolution obtained with a “Fe source and the complete detector, (rhenium crystal + thermistor) is not worse than that obtained by irradiating the thermistor alone.
4
2
0
1000
1500
2000
2500
3000
3500
4000
Energy (eV) Fig. 1. (a) Amplitude spectrum obtained with the rhenium-Ge NTD detector: the two small peaks are the 5897eV and the 6490 eV of the “Fe calibration source, the contmuous spectrum is the p spectrum of the ln7Re. The RMS resolution is 55eV (b) Kurie plot of the beta spectrum of ‘“‘Re.
Energy (keV) Fig. 3. LJ’Am spectrum. expanded.
In the inset the low energy
part is
F. Fontanelli
et al. I Nucl. Instr. and Meth. in Phys. Res. A 370 (1996) 247-249
0
0
0
Energy
(keV)
Fig. 4. Calibration plot of the rhenium-Ge NTD detector 60 keV. The non-linearity at the level of 1% is excluded 60 keV.
up to up to
The linearity of the Re-Ge microcalorimeter has been tested with a 24’Am gamma source up to 60keV The spectrum is shown in Fig. 3 and the calibration plot in Fig. 4. Deviation from the linearity at level of 1% can be excluded. In Fig. 5 is shown the plot of the energy resolution RMS vs. the energy of the spectral line. A clear increase of the line spread with the energy is found. With the present detector it should be possible. in about two months of data taking, to set an upper limit to the neutrino mass below 20 eV. To further improve the detector performance in the near future, we are testing rhenium detectors with a smaller thermistor (0.1 X 0.1 X 0.25 mm’). It was also observed [5,6] that by applying a thin coating of a convenient material, gold or bismuth, to the rhenium crystal a more efficient thermalization is obtained at the lowest temperatures. This effect that has to be investigated more as it could help us to operate a detector below T, /T, = 2 X 10mJ, with lower thermistor and parasitic heat capacity. We need also an efficient calibration procedure at low energy. The energy difference of the two lines of “Fe is rather small and the average energy is more than twice the end point energy of 18’Re. A third peak at a lower energy is necessary for a reliable calibration. An X-ray line of chlorine at 2.7 keV has been obtained by irradiating a sheet of PVC with a strong source of “Fe. For this purpose a system of thermal shields and collimators is under test.
~(,~‘,~,,‘~~~~‘~~,,‘~,,~‘~11~‘~~~~ 10
20
30
249
50
40
60
70
Energy (keV) Fig. 5. Energy resolution spectrum.
RMS vs. the energy
line of the ‘J’Am
4. Conclusion A cryogenic microcalorimeter with features adequate to set a meaningful limit to electron anti-neutrino mass studying the rhenium P-decay has been successfully tested, and preliminary data were obtained.
Acknowledgements The authors would like to thank Prof. E.E. Hailer of the Lawrence Berkeley Laboratory for kindly supplying the NTD Germanium sensors.
References [1] Review of Particle Properties, Phys. Rev. D 50 (1994). [2] SE. Koonin, Nature 354 (1991) 468. [3] E. Cosulich, G. Gallinaro, F. Gatti and S. Vitale, Phys. Lett. B 295 (1992) 143. [4] E. Cosulich and F. Gatti. Nucl. Instr. and Meth. A 321 (1992) 211. [5] S. Vitale, G. Gallinaro and F. Gatti, Proc. SPIE 1743 (1992) 368. [6] S. Vitale, G. Gallinaro and F. Gatti, J. Low Temp. Phys. 93(3) (1993) 262. [7] F. Gatti and A. Nostro, these Proceedings (Workshop on Low Temperature Detectors (LTD6). Beatenberg/Interlaken, Switzerland, 1995) Nucl. Instr. and Meth. A 370 (1996) 218.
VI. APPLICATIONS
and Methods
in Physics
Research
A 370 (1996) 247-249
NUCLEAR INSTRUMENTS 8 METHODS IN PHYSICS RESEARCH Section A
ELSEWER
The P-decay of
’
87Re studied with a cryogenic microcalorimeter
F. Fontanelli,
F. Gatti*, A. Swift, S. Vitale
Abstract An experiment to study the P-decay of Ix7Re using cryogenic microcalorimeters is under course. The physical motivations of the experiment that can give a limit to the anti-neutrino mass are briefly discussed. Problems of the experiment and progresses obtained in the realisation of a detecting device are presented.
1. Introduction The kinematic limits to the electron anti-neutrino mass are presently obtained from experiments that, by means of high resolution magnetic or electrostatic spectrometers, measure the spectrum of the P-electrons emitted by a ‘H source in the end-point region. The effect of a finite anti-neutrino mass should manifest itself as a count deficit, in first approximation proportional to rnt, near the spectrum end-point. To set a meaningful limit to the neutrino mass, very small deviations must be excluded from the P-spectrum expected for m, = 0. The most likely value for ,rrt from all the recent results is negative, the particle data group average being -57+30 eV’ [I]. The negative rni value, that corresponds to an excess counting rate in the region approaching the end-point, has been speculated to arise from an incomplete knowledge of the final state effects or other systematic effects. To clarify the question of the squared negative mass, calorimetric experiments have been proposed in the past. The calorimetric method, where the full energy release following the P-decay is detected, removes the final state effect. Microcalorimeters are presently the only detectors with an energy resolution sufficient to set a sensitive limit to anti-neutrino mass and energy response independent of the kind of excitation. In particular. due to the limit in counting rate allowed by microcalorimeters. as explained in the following text, we have proposed the “‘Re as the p-emitter [3]. 2. Rhenium experiment In a calorimetric experiment, the full spectrum contributes to the detector counting rate, but only a very small fraction of decay events, proportional to (p”lQ)‘. (F” =
* Corresponding [email protected].
author.
Fax
+39
10
313358,
0168.9002/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved SSDI 0168.9002(95)01 100-5
limit to be set to m,,), falls in the region of the spectrum sensitive to the anti-neutrino mass near the end-point. The source intensity must be limited to avoid pile-up problems, and the time needed to accumulate reasonable statistics in the end-point region may become exceedingly long unless Q is very small. With a tritium Q value of 18.6 keV and a microcalorimeter as total absorption spectrometer, the time needed to set an upper limit of 10 eV to my is not less than one year. The interest in the IX7Re P-decay originated since its Q value is the lowest known in nature, only 2.7 keV Because of the lower maximum decay energy of rhenium the allowed counting rate in the end-point region is much less stringent than for Tritium, reducing further the measure time to set a given limit to IPI,,. The theoretical analysis of the lX7Re P-spectrum shape is, at first glance, not too difficult, the decay being a first-forbidden unique transition. It is of note, however. that at such low energies great care has to be taken to check the validity of the theoretical approximations in evaluating the Fermi function F(Z. El and the spectral shape factor. Moreover, the decay of rhenium not as an isolated atom but in a crystal lattice is expected to produce an oscillatory modulation of the P-spectrum when the electron wavelength is not far from the lattice parameter [2]. The expected effect however is small and should be negligible above 1.5 keV Rhenium is a superconducting metal with critical temperature of 1.7 K. We have proved elsewhere [5,6] that a detector made with a superconducting rhenium single crystal coupled with a Ge NTD has energy thermalization that is reasonably fast and complete, provided that the operating temperature is kept above T=2X lo-J T, g 90 mK. It is then possible to build efficient microcalorimeters with a superconducting absorber, but, because of the lower limit on the operating temperature the overall heat capacity of the detector stays high limiting the achievable energy resolution.
VI. APPLICATIONS
248
F. Formrtelli
3. Progress
et ul. I Nucl. Instr. and Meth. in Phps. Res. A 370 (1996) 247-249
in detector development
In Figs. la and lb the P-spectrum and the Kurie-plot recently obtained with a test detector are shown. The two calibration X-ray peaks at 6490 and 5897 eV are from a “Fe source and are completely resolved. The test detector is made with a small natural metallic rhenium crystal and by an NTD germanium thermistor R17, 0.5 X 0.25 X 0.25 mm’ glued to the crystal with Ag epoxy. The rhenium mass is about 2 mg. imposed by the maximum allowed rate of “‘Re decays in the detector. Thermal and electrical connections to the refrigerator
100
60
60
40
zo
0
Energy (eV)
Fig. 2. View of the rhenium-Ge
NTD calorimetric
detector.
mixing chamber and electrical connections to a source follower, placed inside the refrigerator, are provided by two ultrasound bonded 20 pm aluminium wires. A schematic outline of the device is shown in Fig. 2. The amplified thermistor signal is recorded and analysed off-line: from each pulse we obtain, by adaptive filtering [7], an amplitude proportional to the energy release, and a shape factor [4] that measures the deviation of the pulse shape from a reference shape. The reference pulse shape is obtained by averaging several hundreds selected input pulses. The observed energy resolution at 5.9 keV, 55 eV RMS, is fully compatible with the signal to noise ratio. The noise level is scaling with the square root of the thermistor electrical resistance in the operating point (3.7 MO). as expected for the device Johnson noise: the contribution of the amplifier noise (0.8 nV/Hz”‘) is negligible. The signal amplitude is limited by the detector heat capacity, mainly due to the thermistor contribution, about IO pJ/K at 100 mK. Indeed the energy resolution obtained with a “Fe source and the complete detector, (rhenium crystal + thermistor) is not worse than that obtained by irradiating the thermistor alone.
4
2
0
1000
1500
2000
2500
3000
3500
4000
Energy (eV) Fig. 1. (a) Amplitude spectrum obtained with the rhenium-Ge NTD detector: the two small peaks are the 5897eV and the 6490 eV of the “Fe calibration source, the contmuous spectrum is the p spectrum of the ln7Re. The RMS resolution is 55eV (b) Kurie plot of the beta spectrum of ‘“‘Re.
Energy (keV) Fig. 3. LJ’Am spectrum. expanded.
In the inset the low energy
part is
F. Fontanelli
et al. I Nucl. Instr. and Meth. in Phys. Res. A 370 (1996) 247-249
0
0
0
Energy
(keV)
Fig. 4. Calibration plot of the rhenium-Ge NTD detector 60 keV. The non-linearity at the level of 1% is excluded 60 keV.
up to up to
The linearity of the Re-Ge microcalorimeter has been tested with a 24’Am gamma source up to 60keV The spectrum is shown in Fig. 3 and the calibration plot in Fig. 4. Deviation from the linearity at level of 1% can be excluded. In Fig. 5 is shown the plot of the energy resolution RMS vs. the energy of the spectral line. A clear increase of the line spread with the energy is found. With the present detector it should be possible. in about two months of data taking, to set an upper limit to the neutrino mass below 20 eV. To further improve the detector performance in the near future, we are testing rhenium detectors with a smaller thermistor (0.1 X 0.1 X 0.25 mm’). It was also observed [5,6] that by applying a thin coating of a convenient material, gold or bismuth, to the rhenium crystal a more efficient thermalization is obtained at the lowest temperatures. This effect that has to be investigated more as it could help us to operate a detector below T, /T, = 2 X 10mJ, with lower thermistor and parasitic heat capacity. We need also an efficient calibration procedure at low energy. The energy difference of the two lines of “Fe is rather small and the average energy is more than twice the end point energy of 18’Re. A third peak at a lower energy is necessary for a reliable calibration. An X-ray line of chlorine at 2.7 keV has been obtained by irradiating a sheet of PVC with a strong source of “Fe. For this purpose a system of thermal shields and collimators is under test.
~(,~‘,~,,‘~~~~‘~~,,‘~,,~‘~11~‘~~~~ 10
20
30
249
50
40
60
70
Energy (keV) Fig. 5. Energy resolution spectrum.
RMS vs. the energy
line of the ‘J’Am
4. Conclusion A cryogenic microcalorimeter with features adequate to set a meaningful limit to electron anti-neutrino mass studying the rhenium P-decay has been successfully tested, and preliminary data were obtained.
Acknowledgements The authors would like to thank Prof. E.E. Hailer of the Lawrence Berkeley Laboratory for kindly supplying the NTD Germanium sensors.
References [1] Review of Particle Properties, Phys. Rev. D 50 (1994). [2] SE. Koonin, Nature 354 (1991) 468. [3] E. Cosulich, G. Gallinaro, F. Gatti and S. Vitale, Phys. Lett. B 295 (1992) 143. [4] E. Cosulich and F. Gatti. Nucl. Instr. and Meth. A 321 (1992) 211. [5] S. Vitale, G. Gallinaro and F. Gatti, Proc. SPIE 1743 (1992) 368. [6] S. Vitale, G. Gallinaro and F. Gatti, J. Low Temp. Phys. 93(3) (1993) 262. [7] F. Gatti and A. Nostro, these Proceedings (Workshop on Low Temperature Detectors (LTD6). Beatenberg/Interlaken, Switzerland, 1995) Nucl. Instr. and Meth. A 370 (1996) 218.
VI. APPLICATIONS