EPIC MOS long term stability & radiation damage effects · EPIC MOS long term stability & radiation damage effects ... No major solar flare since ... Asymmetry of radiation profile

Bruno Altieri -- SCI-SDX

XMM-Newton

EPIC MOS long term stability &radiation damage effects

EPIC consortium meeting

Palermo, 16 October 2003

B. Altieri

XMM-Newton

2


Talk outline

• Charge Transfer Inefficiency (CTI)

• Energy scale

• Energy resolution

• Bad pixels and CCD damages

• Background

• Radiation environment

• Conclusion

XMM-Newton

3


MOS2 CTI

• reduction of the parallelCTI by a factor 3, at the MOScooling in November 2003

•serial CTI: constant sincelaunch

• very slow degradation rateobserved since

CTI = 1.2 10-5 @ rev 700

-120C-100C

XMM-Newton

4


MOS1 CTI

• MOS1 CTI significantlyworse than MOS2• the worse of all 14 CCDs.

• Improved by a factor 2.2at cooling

No major solar flare sincerevolution 433

-100C -120C

XMM-Newton

5


MOS CTI modeling (CCFs)• 8 time periods defined

• CTI = (a + b δt) . Eα with α(t,CCD) =0.35-0.7

• Slow degradation since cooling now measurable, to be implemented

XMM-Newton

6


CTI degradation rates• CTI degradation rates comparable for all detectors

– Radiation management succeeded in limiting rate of CTI degradation to acceptable levels.

• PN back-illuminated CCDs degrading more shows that a relative low dose ofsoft-protons is reaching the CCDs.

– Soft-protons can only contribute an upper limit of ~20% of the total radiation damage (Smith& Holland)

– proton dose estimated to 106 /cm2 from pn discarded line counter (Altieri,2003), far below theradiation test with 109 /cm2, where no change of properties was seen (Kendziorra, 2000)

• CTI dominated by energetic penetrating radiation (for both BI & FI CCDs)

pn (BI)(150µ m)

beforecooling

aftercooling

BI FI

d(CTI)/dt /year 1.3E-05 1.3E-05 2.0E-06 4.4E-06 1.4E-06

d(CTI)/dt /year /100µ m 8.7E-06 3.3E-05 5.0E-06 1.8E-05 5.8E-06

EPIC ChandraMOS (FI)(40µ m)

ACIS-I(24µ m)

XMM-Newton

7


MOS1 energy scale

• some over-correction,by 5-10 eV

• MOS1 gain could betuned further (<2.10-3)

• No spatial dependenceof the energy scale

6 keV

1.5 keV

XMM-Newton

8


MOS2 energy scale

• slight over-correction inthe past

• temperature dependencegain of the gain notcorrected, but ~1.5% onlyof scientific observationaffected, during eclipseseasons

• MOS2 gain ~OK

1.5 keV

6 keV

XMM-Newton

9


MOS energy resolution• Ground software (SAS) manages to restore most of energy loss, but inevitably

detector energy FWHW widens due to imperfect correction and statisticalnoise of charge trapping.

– Energy resolution rather constant since cooling

140eV@6keV

[email protected] MOS2

6 keV

1.5keV

XMM-Newton

10


Decay of the internal calibration source

Half-life = 2.7 years

longer and longer exposuretimes will be required ...

XMM-Newton

11


MOS bad pixels

Number of hot pixels at >1% recurrence frequency per CCD

• Not an issue for TM bandwidth• Bad pixels have virtuallyvanished since cooling

•No or limited impact of radiation

XMM-Newton

12


Micro-meteoroid events• Two events in revs 108 & 325

• The figure shows a 3d surfaceof E1+E2 over all CCDs, forthe seconds impact.

• The intensity was notuniformly distributed butshows a smooth peak onCCD7 and large spikes onCCD6.

XMM-Newton

13


CCD impact damage• The figure shows the

distribution of new bad pixelsoverlaid with a contour ofE1+E2.

• In exposures following theevent 27 bad pixels appearedon CCD6 and CCD7.

• The new bad pixels coincidewith the location where theenergy was greatest.

XMM-Newton

14


MOS background (1)

XMM-Newton

15


MOS background (2)

XMM-Newton

16


XMM-Newton radiation historyM. Casale

R. Gonzalez

XMM-Newton

17


Earth magnetosphere

XMM-Newton

18


Asymmetry of radiation profile• Asymmetry results from a

combination :– asymmetry of earth magnetic field wrt

rotational axis/center of the earth– asymmetry of the earth’s magnetic field

wrt to solar wind pressure– synchronisation of the XMM orbit with

earth rotation

• The two legs see different regions oftrapped particle belts when at thesame radius.

XMM-Newton

19


Evolution of orbital parameters

• perigee altitude increased from 13.000 to16.000 km

• longitude of ascending node decreased bymore than 90 degrees since launch

• balanced by a 70 degrees increase of perigeeargument

•⇒ rotation of the orbit: perigee tends to aligntoward the equatorial plane.

XMM-Newton

20


Orbit configuration wrt magnetosphere

Winter: low radiation Summer: high radiation

XMM-Newton

21


Modeling the XMM radiation

• Goal:– minimise manual commanding– optimise science window duration– limit failed observation re-planning– avoid spoiled observations in the archive.

• Problem:– radiation trend linked only to the drift of orbital parameters ?

• then how will it evolve after the orbit change manoeuvre ?– or also related to solar cycle ?

• more particles trapped during the solar minimum ?– especially soft-protons in the quasi-trapped region

• Correlation of soft-proton radiation with the pointing direction ?

XMM-Newton

22


Are soft-proton chaotic flares predictable ?

Rev 687

Rev 688

XMM-Newton

23


A simple model to start with?• Advance start of EPIC science

observations :– Perigee Passage + 4 hours, from

revolution ~700 onwards– until Feb. 2004 ?– ==> initial CALCLOSED dropped.

• XMM-Newton: 20-40% affected by soft-proton flaring

• Chandra: <10% ?– Science observations only above

70000km

XMM-Newton

24


Conclusion

• Cooling of the MOS CCDs was highly beneficial– stable properties (energy resolution and scale)– should minimize number of CCFs

• CTI, gain, including redistribution CCFs

• CTI : very simple model implemented SAS (CAL/CCF)– no count rate dependence– over correction of the energy scale,

• Long-term stability prospect

• Need to understand radiation environment trends

EPIC MOS long term stability & radiation damage effects · EPIC MOS long term stability & radiation damage effects ... No major solar flare since ... Asymmetry of radiation profile

Documents