Transcript Gas measurements in the PVLAS experiment

Gas measurements in the PVLAS
experiment
Giuseppe RUOSO
INFN - Laboratori Nazionali di Legnaro
Summary
• Apparatus and test with gases
• Low pressure birefringence measurements
• Mixing of the photon with low mass particles
PVLAS Group
M. Bregant, G. Cantatore, F. Della Valle, M. Karuza, E. Milotti, E. Zavattini, G. Raiteri (Trieste)
S. Carusotto, E. Polacco (Pisa), U. Gastaldi, P. Temnikov (INFN - LNL)
G. di Domenico, G. Zavattini (Ferrara), R. Cimino (INFN - LNF)
Technical support
S. Marigo (LNL), A. Zanetti, G. Venier (TS)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
1
The PVLAS apparatus
Detect modifications of the polarisation state of
a linearly polarised light beam traversing a
dipole magnetic field in vacuum:
• ellipticity due to birefringence
• rotation of the polarisation plane
The two measurements are independent: by
inserting an optical element (Quarter Wave
Plate) one can switch from one measure to the
other OR using a Faraday Cell it is possible to
perform measurement simultaneously (Only in
recent data)
A Fabry Perot cavity (FP) increases the
effective optical path by a factor N ~ 5 104
Laser is green (532 nm) or infrared (1064 nm)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
2
Apparatus at LNL
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
3
Detection method
•
•
•
•
•
A pair of crossed polarisers (P, A) is used to sense polarization changes
The optical path length is increased by means of a Fabry-Perot resonator (finesse ~105) (mirrors M1 and
M2)
An intense magnetic field (~ 6 T) is generated by a superconducting dipole magnet
A removable quarter-wave plate (QWP) used to measure dichroisms
Heterodyne detection is employed to extract small signals
–
–
•
the interaction is time-modulated by rotating the magnet (this rotation also acts as a clock for all signals enabling
phases to be measured)
a carrier ellipticity is introduced by means of a modulator (SOM)
Light intensity transmitted through the last polarizer is detected and Fourier-analysed: the resulting
spectrum contains the physical information
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
4
Test with gases
Gases are ideal test for the apparatus due to the Cotton-Mouton effect:
Magnetic birefringence Dnu of a gas at pressure P in a dipole
magnetic field B
2


BT  P 
Gas
Dnu ( T ~ 293 K)
Dn  n||  n  Dnu
  
-13
1T
Nitrogen
- (2.47± 0.04) x 10

 Patm 
Oxygen
- (2.52± 0.04) x 10-12
Carbon Oxide
- (1.83± 0.05) x 10-13

Ellipticity  due to birefringence
  N
L

Dn sin 2 

a
 a  L n sin 2
bb

D


E'

b
a
Bext
E
L
With N ~ 50000 a few mbar of
nitrogen gives ellipticity ~ 10-4
L=1m
k
zona di campo
 = laser wavelength (532 nm, 1064 nm)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
5
Heterodyne detection - ellipticity
polariser
I0
magnetic field


M
ellipticity modulator
analyser
(SOM)
I

 
0
ITR  I 0  2       I0  2  cos  0t   cos  M t 
2

I
 0

2
  cos 0   M t  cos 0   M t 
M  2ROT
ITR()
2/2


2
2
cos 2 0t 



In the heterodyne detection, using a beat with
a calibrated effect, we have
• Signal linear in the birefringence
• Smaller 1/f noise

0 - M
Tr

0 + M
M0
PVLAS Day - Giuseppe Ruoso
0
High sensitivity
www.ts.infn.it/experiments/pvlas
6
Heterodyne detection - rotation
polariser
I0
magnetic field
M QWP 


ellipticity modulator
analyser
(SOM)
I
 
0
ITR  I 0  2       I0  2  cos  0t   cos  M t 
2

I
 0

2
  cos 0   M t  cos 0   M t 
M  2ROT
ITR()
2/2


2
2
cos 2 0t 



In the heterodyne detection, using a beat with
a calibrated effect, we have
• Signal linear in the birefringence
• Smaller 1/f noise

0 - M
Tr

0 + M
M0
PVLAS Day - Giuseppe Ruoso
0
High sensitivity
www.ts.infn.it/experiments/pvlas
7
Measurements
Heterodyne detection technique
(Rotating Magnet)
Measured effect given by Fourier
amplitude and phase at signal frequency
RUN 965, neon 15 mbar
B = 5.5 T, finesse = 61 000
-4
10
10-16
-5
10-17
-6
10
10-18
-7
10
Birefringence
Ellipticity
10
Vector in the polar plane
10-19
10-8
0
1
2
3
4
5
Frequency (units of magnet rotation frequency)
The amplitude measure the ellipticity/rotation
The phase is related to the triggers position
and magnetic field direction. True physical
signal must have a definite phase
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
8
Apparatus test with nitrogen
• Measure of Nitrogen CME
Dnu (N2) = -(2.4±0.1)10-13
Phase = 195 degree
• Fabry-Perot: finesse F amplification factor control
 = cavity decay time
d = 6.4 m cavity length
Run 573 FP,  ~ 510 ms, B = 5.0 T, P = 0.5 mbar
Run 580 NO FP, B = 5.3 T, P = 85.7 mbar
Expected amplification
N 2
F

2
c
PVLAS Day - Giuseppe Ruoso
d
 = 3.77 10-4
 = 1.52 106
Measured amplification
 47800
www.ts.infn.it/experiments/pvlas
N  48150
9
B Square check with Neon
During data taking the magnetic field diminishes and data must be
normalized to a standard field value before making comparison. In
order to do this we verified the B2 dependence of the effect
The fit to a quadratic
function optimizes
the chi square
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
10
Measurement of CME for Xe, Kr, He
Due to the extremely high sensitivity of the apparatus we were able to perform
precise measurement of very small CME in noble gases
Dnu ( T ~ 290 K,  =1064 nm)
Gas
Xenon
(2.44±0.22)x10-15
Kripton
(8.61±0.35)x10-15
Helium
(1.75±0.07)x10-16
Stability of the apparatus:
Helium CME for measurements
performed over a time > 1 year
Typical pressure plot: each point 100 s data record
2.2 10-16
8 10-18
HeG2 set
u
6 10-18
Unit birefringence Dn
Normalized birefringence Dn
1T
7 10-18
5 10-18
4 10-18
3 10-18
2 10-18
2 10-16
1.8 10-16
Infrared
Green
-16
1.6 10
1 10-18
1.4 10-16
HeG1 HeG2 HeG3 HeG4 HeG5
0
0
10
20
30
40
50
Pressure (mbar)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
HeIR2HeIR3 HeIR4
Data set
11
Gas system
High purity gas samples has to be
used in the measurements
(Helium is 99.9999% pure)
Gas bottles and An all metal gas insertion line
ensures the sample purity
insertion line
Lower vacuum
chamber with optics
We also use a cryopanel to prevent
contamination during gas filling
Chamber outgassing < 2 10-5 mbar/hour
Main components: H2, CO, H2O
Typical run lasts 3-4 hours
No contribution for measurements reported here
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
12
Gases at low pressure - ellipticity
Studying the amplitude of the gas ellipticity for pressures close to zero
it is possible to deduce the amplitude of the searched vacuum effect
Helium
1 10-6
Chamber filled with helium
Cavity amplification = 33 000
B=5 T
amp
8 10-7
360
300
Phase (degrees)
240
Ellipticity
-7
6 10
4 10-7
180
120
60
2 10-7
0
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
Helium pressure (mbar)
-60
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
Helium pressure (mbar)
Data indicates that vacuum is showing an effect which has
sign opposite to helium and thus there exists an helium
pressure at which the overall effect is zero!
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
13
Gases at low pressure - ellipticity - II
We performed the same
measurement with different gases
Log - Log scale
Helium, Neon, Nitrogen
10-5
November 2005 data
Neon
Helium
10-6
Ellipticity
Nitrogen has a CME with sign
opposite to neon and helium and
shows no zero crossing
Nitrogen
10-7
Data collected in two different
periods give similar results but
different vacuum amplitudes
PVLAS Day - Giuseppe Ruoso
10-8 -8
10
10-6
10-4
10-2
100
102
Pressure (mbar)
www.ts.infn.it/experiments/pvlas
14
Gases at low pressure - ellipticity - summary
• Zero pressure ellipticity effect of the order of 10-7 for 33000 passes in a 5 T
field for 532 nm light
• Similar results for infrared (lower statistics)
350
280
phase (degrees)
• The sign of the ‘vacuum’ signal is
opposite to noble gases birefringence
(CME) and same as nitrogen
Nitrogen phase
Helium phase
Neon phase
210
140
70
0
10-5
Nov 2005
10-4
10-3
10-2
10-1
100
101
Gas pressure (mbar)
Gas data in any case do not suggest the nature of the vacuum signal.
Explanation of this result is still unclear
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
15
Vacuum rotation
Possible interpretation
Rotation is actually a dichroism (selective absorption of a
polarization component) due to the mixing of the photon with a
low mass particle
Particle mass m ~ 1 meV
PVLAS Day - Giuseppe Ruoso
Inverse Coupling M ~ 4 105 GeV
www.ts.infn.it/experiments/pvlas
16
Mixing of the photon with low mass particle
If we suppose that the vacuum rotation signal is physical and due to a
particle we can use a gas to change the effect due to a change of the
effective mass of the photon (different index of refraction)
10-6
FB L sin x 
M,m, pgas

2 

8M
x 
2
L pgasn stp 1 m 2 

x  

2 
patm
2 


2
Curves for M = 4 10 5 GeV
10-7
Dichroism (rad)
2
ext
Vacuum dichroism
Dichroism with p > 0 mbar (~ 10 mbar Ne)
10-8
10-9
10-10
10-11
10-12
10-2
10-3
mass (eV)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
17
Mixing of the photon with low mass particle II
Increasing the pressure from vacuum the expected signal will decrease following a
[(sin x ) / x]2 function, with characteristic zeroes depending on the gas pressure P
(index of refraction)
Neon
(n-1) = 67.1 10-6 (P / Patm)
Helium (n-1) = 34.9 10-6 (P / Patm)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
18
Fabry -Perot cavities and ellipsometers
When an ellipticity is present in a Fabry-Perot cavity with birefringent mirrors, a
spurious dichroism is also generated due to a leakage between resonant modes of
the cavity that are almost degenerate
Gas in cavity with magnetic
field generates ellipticity
linearly proportional to
pressure through CME
PVLAS Day - Giuseppe Ruoso
A dichroism is also generated
linearly proportional to pressure
that amounts ~ 5 - 10 % of the
produced ellipticity
www.ts.infn.it/experiments/pvlas
19
Measurements - gas dichroism I
- gases do not generate rotation/dichroism
- small dichroism proportional to pressure due to Cotton-Mouton effect via cavity
birefringence (spurious effect)
- to reduce spurious effect we choose gases with largest ratio (n-1)/CME
Fitting function:
First measurement: neon
1.4 10 -12
yM,m, pgas b pgas  M,m, pgas
Normalized Rotation (rad)
The y axis has the measured
rotation/dichroism projected on
the physical axis and divided
by the number of passes in the
cavity
Measured data
Residual gas effect
1.2 10 -12
1 10 -12
8 10 -13
6 10 -13
y = m0*m2+1/4*(195*5e6/m1/2)...
Value
4 10 -13
2 10
-13
Error
M (eV)
2.86e+14
5.8e+13
m2
1.69e-13
3.68e-15
mass (eV)
0.00114
0.00011
Chisq
3.09
NA
R
0.998
NA
0
0
2
4
6
8
10
Neon pressure (mbar)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
20
Gas Dichroism II - still neon
Several measurements performed, some data show effect, some other no:
3 10-13
2.5 10-13
Fit compatible with straight line
Particle parameters compatible with 0
Errors values compatible with left side data
Difference (Measured data - residual gas effect)
2 10-13
-13
1.5 10
1 10-13
5 10-14
0
-5 10-14
Value
3 10 -12
Normalized rotation (rad)
y = m4+m0*m2+1/4*(195*5e6/m1...
y = m4+m0*m2+1/4*(195*5e6/m1...
2.5 10 -12
M (eV)
2.97e+14
m2
1.78e-13
2.53e-15
0.00117
0.000114
m4
-3.68e-13
3.66e-14
6.96
R
M (eV)
m2
NA
1
Value
3 10 -12
7.88e+13
mass (eV)
Chisq
2 10 -12
Error
Normalized rotation (rad)
3.5 10
AF05NMs runs 952 - 959
3.5 10 -12
AF02NMs runs 821-829
-12
NA
1.5 10 -12
1 10 -12
2.5 10
-12
2 10
-12
mass (eV)
4.27e+15
1.61e-13
1.14e-14
0.00027
m4
Error
1.49e+15
-1.19e-13
0.00716
1.87e-13
Chisq
10.5
NA
R
0.998
NA
1.5 10 -12
1 10 -12
Measured data
5 10 -13
5 10 -13
Residual gas effect
Measured data
Residual gas effect
0
0
0
5
10
15
20
0
Neon pressure (mbar)
5
10
15
20
25
Neon pressure (mbar)
If the non linearity is correct, is due to particle mixing or is there another
possible explanation?
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
21
Gas dichroism III - helium
To reduce linear effect due to Cotton Gas
Mouton we performed measurements
Neon
with helium
Helium
1.5 10 -12
M (eV)
Normalized rotation (rad)
mass (eV)
1 10-12
m4
67.1 10-6
(5.9 ± 0.1)x10-16
34.9 10-6
(1.75±0.07)x10-16
First data showed the non
linearity, but on following
runs this was not clear
Error
1.9e+14
m2
CME: Dnu
y = -4.32e-13 + 6.59e-14x R= 1
y = m4+m0*m2+1/4*(195*5e6/m1...
Value
(n -1) @ Patm
5.83e+13
6.59e-14
9.24e-15
0.00169
0.000239
-4.32e-13
1.33e-13
Chisq
14.3
NA
R
0.906
NA
5 10-13
0 100
Measured data
Residual gas effect
-5 10-13
0
5
10
15
20
25
Data analysis is still
underway, also with the
study of possible systematic
effects that could mimic the
non linear part
helium pressure (mbar)
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
22
Conclusions
Gas measurements are very important in the PVLAS experiment:
• Careful tests of the apparatus performances can be executed
• Vacuum magnetic birefringence/ellipticity measurements receive
a stronger validation from measurements with gas at low pressure
• The particle hypothesis can be tested measuring rotation /
dichroism in the presence of a gas. Regarding this point a clear result
needs more statistics and a careful control of systematics
PVLAS Day - Giuseppe Ruoso
www.ts.infn.it/experiments/pvlas
23