Work package 1, task 1 and 5: M1 - Advanced materials for mirror substrates
M5 - Innovative materials for suspension in
advanced detectors
Task aims in the 4th year
 Characterization of the losses of Si fibres and cantilever blades at different
temperatures, reaching the Liquid Nitrogen temperature
 Investigation on the doping effect on Si (n- and p-type doping at cryogenic
temperatures)
 Study of dissipation mechanism at low (down to 10K) temperature in small sample
 Silicate bonding studies at low temperature
 Definition of last stage optimal parameters
Major achievements in the forth 12 months
Pairs of silicon disks were bonded in Glasgow using
different volumes of sodium silicate bonding solution (1
part commercial sodium silicate solution to 6 parts water,
volumes in the range 0.4l/cm2 to 0.1l/cm2). Extensive
studies are underway using a custom made facility in
Florence for investigating the thermal conductivity
through bonded silicon samples. The thermal conductivity
of the bonded sample was observed to be similar to pure
Figure 1: measured and literature values for
silicon at room temperature (see Fig. 1). These results, in
the thermal conductivity of silicon compared
to the measured values of bonded silicon.
combination with the observed robustness during repeated
temperature cycles in Glasgow, suggest the suitability of
silicate bonding for the quasi-monolithic silicon suspensions central to the construction of
third generation gravitational wave observatories from a both a reliability and thermal
loading point-of-view.
measured  of unbonded silicon
other literature values for silicon (MPDB)
measured  of bonded silicon
Figures 2: mechanical strength of silicate
bonded silicon samples - Perugia
Preliminary measurements of the strength of silicate
bonds have been carried out seperately by the
Univerisities of Perugia and Glasgow. New results from
the Perugia lab suggest the breaking strength of siliconsilicon bonds to be ~10 MPa, as shown in Figure 2,
with little evidence of a dependence on crystal
orientation and possible dependence on surface flatness
at /4. Further experiments are being planned to better
understand these dependencies in addition to other
possible effects, e.g. dependencies on the type (wet or
dry) of thermal oxide and the rate at which they are
grown.
Delays and possible alteration of the planning
None expected.
Work package 1, task 2: M2 - Advanced materials and techniques for resonant
detectors
Task aims in the 4th year
 Measurement of mechanical loss properties at low temperature in advanced materials
for resonant detectors.
 Investigation of the fabrication processes of silicate bonding for SiC, and electron
beam welding, explosion welding and cold welding for metals.
Major achievements in the forth 12 months
Materials:
We measured the quality factor of single crystal silicon wafers 4 inch in diameter and
thickness in the range 0.5-1 mm. In order to minimize the mechanical losses we decided
to minimize the contact area between the sample and the holder and to realize a nodal
contact between the two. This was done by sandwiching the wafer in the central part
between two sapphire balls.
In the temperature range 40-120 K and 130-300 K the dominant loss mechanism is the
thermoelastic loss (see fig1a). In the temperature range 120-130 K, where the predicted
thermoelastic losses are very low, we observed a strong correlation between the measured
quality factor of the first 12 modes and the wafer displacement averaged over a circle 0.5
mm in diameter (the estimated contact area between the sapphire ball and the disk)
calculated numerically using FEM. This suggests that in this temperature range the
dominating mechanism is the clamping loss. Below 40 K we achieved quality factors of
the order of 108. We have some evidence that in this temperature range surface losses are
very important.
A new nodal suspension has been tested. In this set-up a sapphire ball is glued at the
centre of the wafer, then the ball itself is glued on the samples holder. In this way no
pressure is applied at the wafer and one face keeps free and can be used to test coating
layers. The measured quality factor of the glued suspension is the same as the one of the
traditional nodal suspension.
Finally a bonded wafer composed by 3 wafer 4 inch in diameter and 0.3 mm in thickness
each has been bonded together using direct bonding (with no material added). The
bonded wafer has been tested at low temperature fig1b. In the explored temperature range
(4-300 K) the losses decrease monotonically with temperature. Preliminary analysis of
these results gives a low temperature loss angle value of the bonding layer of about 10-2.
Fig1: Left Measured loss angle for the silicon wafer modes at 380 Hz (circle) 1730 Hz
(plus) and 2515 Hz (cross). In grey the losses predicted by the thermoelastic model for
the mode at 1730Hz. Right Measured loss angle for several modes of the bonded silicon
disk.
As concerning the silicon carbide the quality factor of C/SiC disks 3 inch in diameter and
different thickness has been measured in the temperature range 2-300K. For one sample
the measurements has been extended down to 50 mK The disks were hung by a nodal
point suspension. The best measured quality factor at 50 m K is of the order of 500000
(fig2b), still too low for the detector Dual. The dominant dissipation mechanism is still
unknown but thermoelastic dissipation and surface losses can be excluded as the
measured loss angle (see fig2a.) does not exhibit the required mode frequency and disk
thickness dependence. In order to check if the SiC is intrinsically a dissipative material
single crystal pure SiC wafers of polytype 4H and 6H 2 inch in diameter and about 0.3
mm in thickness has been tested. In this case the best achieved quality factor is of the
order of 7•106 for the 6H crystal at 60 K.(fig2b), below this temperature the quality factor
drops. On the other hand, for the 4H polytype the Q-factor is monotonically decreasing
with temperature achieving the best value at low temperature (5 K) about 500000. These
results suggest that the quality factor of SiC depends on the considered politype but in
any case the mechanical dissipations are in general much higher than the in the case of
the silicon single crystal.
Fig2: Left: Cumulative plot of the loss angle measured for different SiC samples.
Right: Loss angle of C/SiC disks. The thickness of the disk 1 is 1mm while for the disk 2
is 2mm. For each disk 4 different modes has been characterized.
As regarding the dielectric losses, the measurements of the dielectric loss of a new
sapphire sample with higher purity have been performed at 10 kHz from 4 K down to 100
mK. The sample is a sapphire disk (SurfaceNet, purity 99.995%, diameter 40 mm,
thickness 0.270 mm), niobium sputtered on both sides, which is employed as capacitor
(C=440 pF) in a electrical resonator with a low loss superconducting coil (L=0.45 H).
From the free decay time constant of the oscillating resonator current (f=11283 Hz) an
upper limit to the dielectric loss of the sapphire capacitor is evaluated (see figure).
Fig3: Ultralow temperature upper limit of the dielectric loss of the sapphire estimated at
about 11.2 kHz
Delays and possible alteration of the planning
None expected
Work package 1, task 3: M3 - Investigation of superconducting materials
Task aim in the 4th year
The aim of this task is to fully characterize a prototype detector, based on a Parametric
Converter scheme, using High quality factor superconducting cavities as resonant
detector-transducer.
Major achievements in the forth 12 months
The prototype cavity installed in the cryostat was measured for the full characterization
of the Electromagnetic and mechanical properties of the system.
The preliminary results on the cavity Electromagnetic modes confirm the results of the
preliminary tests. Electromagnetic Quality factors in excess of 1010 @1.8 K, at a stored
energy of 5 joules, were measured.
After a careful mechanical tuning of the two cells of the RF structure the resulting mode
split (giving the maximum gain for the Parametric Converter detector) was measured, and
resulted in optimum detection frequency of 20 kHz according to the Design value sat
Room temperature.
After cooling down at the operating temperature of 1.5K the optimum frequency shifted
to 7 Khz due to the mechanical changes of the cavity dimension in after the cool down.
Preliminary tests of the parametric converter detector based on the showed a good
agreement with the parametric converter Projected sensitivity. We measured a strain
sensitivity of 10-19 [Hz-1/2]; the limitation being set by the wide band noise of the front
end RF amplifier used in the early stage of the tests to simplify the operation of the test
setup.
The activity will continue with the implementation of the Cryogenic RF amplifier and
electronic to fully exploit the cavity quality and the Parametric Converter detector
scheme.
The final aim of the experimentation is to measure the sensitivity (at 7 kHz) of the
detector and to find the noise contributions (Background Noise, Electronic Noise,
Residual Master RF oscillator Phase noise) affecting the ultimate sensitivity of the
Parametric converter detector.
Delays and possible alteration of the planning
None expected
Work package 1, task 4: M4 - Development of low loss dielectric coatings for
advanced detectors
Task aims in the 4th year
• Loss measurements on SiO2/Ta2O5 at low T
• Production of substrates and coatings
• Investigations of losses of substrates necessary to support diffractive coated optics
• Investigations of mechanical loss of diffractive coatings at low temperatures
• Prototype design and build
Major achievements in the forth 12 months
Investigation on mechanism of energy loss in coating still remains the most ambitious
and necessary advancement that has to be achieved for the third generation of
gravitational wave interferometers. The collaboration between the groups of Glasgow,
Lyon, Perugia, Jena and Firenze has been effective as ever and it had the support of
another group working in STREGA: the Legnaro group.
The major achievement, explained in full detail in the annual report, is the experimental
evidence of the loss mechanism nature: the relaxation process follows the
phenomenology of a double-well asymmetric potential system and the activation energy
has been measured. These fundamental results have been achieved in Glasgow.
The Legnaro group has its cryogenic facility able to reach sub Kelvin temperatures and in
collaboration with Glasgow and Jena they are taking measurements in that temperature
range to find out more about the relaxation mechanism.
In the mean time the Perugia group is comparing characteristics of silica in bulk and in
IBS layer in order to reveal similarities and differences between the two amorphous
materials.
Delays and possible alteration of the planning
None expected.
Work package 1, task 6: M6 - Study of thermoelastic effects caused by absorption of
\cosmic rays
Task aims in the 4th year
 Low temperature measurements on an Aluminium bar at the superconductive state
Major achievements the third 12 months
During last 12 mouths Task M6 activities have been focused on the preparation of the
cryogenic system for the measurements on the AL5056 at the BTF and the measurement
execution at T < 1 K. Data taking on the AL5056 bar was successfully performed at BTF
in an interval ranging from room temperature down to 0.5 K. Data collected in the
temperature range T < 1.3 K: for T > 0.9 K are in agreement with the expectations for the
normal state, the onset of the superconducting state effects appears at T < 0.9 K and at T
< 0.7 K the amplitude is larger than in the normal state. Further investigation are
scheduled for spring 2008.
Delays and possible alteration of the planning
None expected.