ESR Analysis

1
ESR Analysis

• Analysis of ESR Experiments
• H.G.Essel
• Experiment E049
• 205Hg, 140Pr Beta Decay of Highly Charged Ions

2
Single-Ion Spectroscopy of Two-Body b-Decay
Y.Litvinov
K. Beckert, F. Bosch, D. Boutin, C. Brandau, L.
Chen, H. Essel, M. Faber, T. Faestermann, H.
Geissel, M. Hausmann, A. Ivanov, P. Kienle, O.
Klepper, R. Knöbel, C. Kozhuharov, J. Kurcewicz,
S.A. Litvinov, Yu.A. Litvinov, L. Maier, M.
Mazzocco, A. Musumarra, C. Nociforo, F. Nolden,
W. Plaß, R. Reda, C. Scheidenberger, M. Steck, T.
Stöhlker, B. Sun, H. Weick, N. Winckler, M.
Winkler
3
E049 Experiment
Y.Litvinov
152Sm projectiles _at_ 500 MeV/uIntensity 108
particles/spill1032 g/cm2 9Be-targetBr-DE-Br
separation731 g/cm2 Al-degrader _at_ S2Several
particles (1-6) of 140Pr58 _at_ 400 MeV/u stored in
the ESRStochastic pre-cooling (4-7
seconds)Electron cooling
4
Schottky setup
5

Combined Stochastic and Electron Cooling
at the ESR
PPT animation
Mother Daughter Isomer
David Boutin, PhD JLU Giessen, 2005
10
6
Single-Particle Decay Spectroscopy
Y.Litvinov
Electron capture to ground state, emission of
neutrino
Sensitivity to single stored ions Recording the
correlated changes of peak intensities
corresponding to mother and daughter
ions Reliable determination of the number of a
few stored particles Investigation of a selected
decay branch, e.g. pure electron capture
decay Systematical effects such as late cooling
or feeding via atomic or nuclear decays can be
disentangled
7
Praseodym
8
Single shot time frame matrix (O(103))
Each time frame (row) is a Fourier spectrum
9
Total projection over time
10
Interactive approach

• Look at matrices and frames
• Determine interactively the decay times
• Control by several individuals
• Oszillations in decay time statistics would be
  exciting!
• Very time consuming!

11
Statistical approach

• Project decay track on time
• Decay time shows up as step
• Use signal filters to determine step ? time of
  decay

12
Track of decay
13
Projection over frequency window
Statistical approach
14
Signal filter quotient of integral windows
Statistical approach
15
Problems
Statistical approach

• Bad agreement with interactive results (10 - 50)
• Not adequate for nature of data
• Numerically instable
• Faked by additional decays
• Implementation slow (single files)

16
Analytical approach

• Validate each individual fourier spectrum (time
  frame) (106)
• Keep validated frames only
• A peak in decay frequency region is then valid
• But
• There are still peaks where no peaks should be
• There are missing peaks where peaks should be
• Therefore
• Apply confidence account (bonus) mechanism

17
Soft cooling to minimize decay frequency window
Analytical approach
Only possible with many mothers!
18
Time slices (get region of decay)
Analytical approach
19
Frequency projections over time slices
Analytical approach
Region of decay
20
Three time slices around decay
Analytical approach
Right part used for quality check
21
Fourier spectra (time frames)
OK
Not OK
22
Validity window of decay
Analytical approach
23
Valid frequency frames
Analytical approach
valid
Maximum in validity window
Maximum in validity window must be above 1.5
maximum outside
24
Bonus to determine decay time
Analytical approach
Confidence account
Bonus7, malus1
Bonus can be determined from distribution of
"holes"!
25
Comparison with interactive analysis
Analytical approach
Agreement 40 - 75 Statistical 10 - 50
26
Analysis parameters

• config file for hg0 run
• namepref hg0 // file name prefix
• maxfiles 5555 // maximum number of files
• maxitems 1000 // maximum items in container
• indoff 0 // offset of file numbers
• rawbinID 1 // header id
• normbinID 2 // header id
• anabinID 3 // header id
• decaybinID 4 // header id
• finalbinID 5 // header id
• histobinID 6 // header id
• IQaverage 1 // bins to average in IQ
• xrawbins 641 // bins x raw
• yrawbins 750 // bins y raw
• xunitfact 1.0 // units x (bin frequency)
• yunitfact 160.0 // units y (bin ms)
• pk1thresh 20.0 // minimum 1st peak

xnormbins 150 // bins x normalized ynormbins
750 // bins y normalyzed yoffset 30 //
offset for raw x projection yprobins 720 //
y bins raw x projection binsum 4 // peak
find noisemin 10 // peak find noisefact
1 // peak find pk1width 15 // half width
1st x peak pk2width 10 // half width 2nd x
peak pkdist 30 // distance x peak 1 and
2 xwinbins 40 // xbins sub matrix ywinbins
30 // ybins sub matrix xzoombins 100 //
xbins zoomed yzoombins 90 // ybins
zoomed winbonus 1 // bonus for window
slices bonus 7 // maximum hole
size pk2win 3 // half validity
window pk2signal 5 // signal to noise
distance backfact 1.5 // factor background
maximum
27
Must be determined interactively
2 pk2width
2 pk1width
xnormbins
xzoombins
xwinbins
pkdist
pk2win
28
Analysis design

• Statistical approach
• Implemented within ROOT (macros)
• Slow (bad design, single files)
• Analytical approach
• Implemented with C and CINT macros
• Optimised IO (container files)
• Very fast (20)

29
Analysis chain
Analytical approach
Raw
Zoom
Decay
Cool
Shot matrix nnnn.bin
Zoomed matrizes ana.bin
Shot matrizes raw_m.bin
Cooled matrizes cooled_m.bin
Results decay.bin
Histograms.his
Histograms.his
Histograms.his
Histograms.his
IQ raw nnnn.IQ
ROOT
Go4
LeA
Other
Analysis code, data storage, and visualisation
separate! Binary, self-describing container
files Binary, self-describing histogram
files ROOT (Go4) macros to read files
Matches exactly Go4 analysis steps!
30
Interactive analysis with Go4
31
Interactive analysis with Go4
32
Interactive analysis with Go4
Praseodym
Mother
Daughter
33
Next to do

• Beam time January
• Correlation mother doughter (proof of decay)
• Quality measure?
• Implement as analysis steps in Go4
• Take over by experimenters

Praseodym
Mother
Daughter