XIS Calibration on the Ground Status Report K. Hayashida (Osaka University) and the XIS-team

XIS Components XIS-Sensors AE/TCE DE CCID41-FI CCD EU (Engineering Unit) FM FI0,FI1,FI2,FI3 CCID41-BI CCD FM BI0,BI1 AE/TCE FM AE/TCE01,23 EM AE/TCE DE FM 4PPU+MPU FM Spare before Launch

Calibration Task Share Components Location X-ray Source QE reference Chip level CSR/MIT Fluorescent X-rays (C,O,F,Al,Si,P,Ti,Mn,Cu) ACIS chips calibrated at BESSY Camera without OBF +FM AE Osaka Grating Spectrometer 0.2-2.2keV Polypro-window Gas PC & XIS-EU Kyoto Fluorescent X-rays (Al,Cl,Ti,Mn,Fe,Zn,Se) Window-less SSD OBF Synchrotron Facility Synchrotron X-rays + monochrometer (Transmission measurement with PIN diode) Camera onboard the satellite ISAS/JAXA 55Fe

XIS Data Reduction Frame Data /8sec Event data XIS Response depends on the reduction procedure Frame Data /8sec Dark-level Subtraction Event Pickup (PH(E)>Event Threshold) 5x5 mode, 3x3 mode or 2x2 mode Event data Charge Trail Correction Grading / PHA-reproduction for PH(i)>Split Threshold PHA-dependent Split Threshold for BI Bad Columns Filter Spectrum / Image / Light Curve Onboard DE On the ground

Event Grades Grades 02346 are used as X-ray events. Event threshold 以上で最大の ピクセルレベルのピクセル Split threshold 以上でイベント に含むピクセル に含まないピクセル grade0 grade1 grade2 grade3 grade4 grade5 grade6 grade7 大きくイベントが広がる grade7 は、 エネルギー分解能を悪くするため データ処理には用いず、 grade02346 のみを使用する。 ※

BI1 Mn K incidence ＰＨ（２）= preceding pixel ，ＰＨ（７）=trailing pixel PH [ADU] Near readout node Far from readout node

Amount of Charge in the Trail Mn K ＢＩ１ CTI estimated from this trail PH(7)Center [ADU] CTI = (4.5±0.3)×10 [ /Transfer ] -6 *) temperature dependence was observed Number of V-Transfer in the Imaging Area

Incident X-ray Energy Dependence VCTI= (1.72・10 )×E －0.5 －4 HCTI= (6.06・10 )×E －0.5 －4 We can tell the amount of charge deposited in PH(7) and PH(5) => Charge Trail Correction

電荷漏れ補正後のPH(2)、ＰＨ（７）の波高分布 （Ｍｎ Ｋ） PH [ADU] PH [ADU]

Effects of Charge Trail Correction Correct Grade Branching Ratio and PH() Reduce Grade7 events due to Charge Trail. 10%-20% increase in Grade02346 ratio at high energies. Restore Non-uniformity in effective QE. (Partial) Restoration in the Energy Scale. Traps with Other time scales are not negligible.

電荷漏れ補正後のグレード分岐比　（ＦＩ２） バッドコラム除去

電荷漏れ補正後のグレード分岐比　（ＦＩ２）

電荷漏れ補正後のグレード分岐比　（ＢＩ１） バッドコラム除去

電荷漏れ補正後のグレード分岐比　（ＢＩ１）

Optimization of Split Threshold for BI1 G02346 event number FWHM (eV) Spth (ADU)

PHA-dependent SpTh

Bad (CTE) Columns Bad CTE X-ray image (number of events /pixel) Bad CTE Typically long trail in each event. Sometimes flickering pixel is observed. Rows near the readout node can be used. Identification logic without accumulating 10^7events was developed. EU= 21 bad columns/chip FI0=14, FI1=12, FI2=17,FI3=24 BI0=23, BI1=50 How should we do for adjacent columns ?

バッドコラム　その１ (1) BAD CTE / Trail Column PH22 PH7 PH0 PH2 PH11 ACTX ACTY V 25 55 381 271 1 / 668 325 V 27 61 1029 4 -1 / 668 336 V 30 64 1490 -2 -1 / 668 239 V 24 54 1471 1 1 / 668 729 V 22 62 1043 164 -2 / 668 239 V 22 44 1672 1 1 / 668 221 V 32 145 1316 3 0 / 668 570 <PH[7]-PH[2]>で判別可能(e.g. >8AU)

バッドコラム　その２ (2)BAD CTE Columns PH22 PH7 PH0 PH2 PH11 ACTX ACTY V 189 393 551 486 6 / 238 897 V 4 6 1565 7 0 / 238 97 V 179 353 549 346 3 / 238 798 V 0 5 1548 3 2 / 238 171 V 109 293 498 335 4 / 238 610 V 204 394 565 459 8 / 238 919 PH[7]>SplitTH(e.g.20) && PH[2]>SplitTH(e.g.20)のイベントの割合（e.g. >0.2)で判別可能

バッドコラムその３ (3) Flickerging Pixels PH22 PH7 PH0 PH2 PH11 ACTX ACTY V 26 74 81 42 17 / 51 879 V 1 5 58 1 -3 / 51 262 V -1 -2 54 2 3 / 51 262 V -1 3 75 1 1 / 51 262 V 0 671 945 2 0 / 51 272 V 1 2 1779 1 2 / 51 717 ピクセル毎のイベントの頻度で判別可能(e.g.>0.1c/frame)

Kyoto Cal System Fluorescent X-rays (Al,Cl,Ti,Mn,Fe,Zn,Se) Windowless Si-SSD is used as the reference counter, assuming 100% efficiency >1.5keV XIS FI-CCD QE=96%@4.5keV is assumed

Calibration Facility in the Osaka Clean Room Detector Chamber Manson Soft X-ray Generator Hetrick Spectrometer

Dispersion (Grating) Spectrum projection X-ray image X-ray energy FWHM～5eV O-Kα（0.53keV） こちらがCCDで得られたX線画像です。分散X線が上下方向に当たっています。低エネルギーほど下に分散されていまして、CCDのこの方向の位置とエネルギーは１対１に対応しています。このような明るい筋は、特性X線に対応しています。これを分散方向にプロジェクションしたのがこちらの図です。このようにして得たエネルギースペクトルは、CCDで得られるイベントのエネルギー情報よりもはるかに優れたエネルギー分解能のスペクトルが得られます。このスペクトルを我々は分散スペクトルと呼びます。以降、この分散スペクトルをもとに解析を行いました。 C-Kα (0.28keV) Dispersion direction Number of events/columns

Line profile against O-K line incidence Astro-E1 (FI) XIS Astro-E2 (FI) XIS 5 kV FI2 PHA(ADU) XIS1 (H.Katayama master thesis)

Line Profile model （１） Main Peak：Absorption in Depletion Layer （２）　Sub Peak：Lost charge below Split-threshold （３）　Triangle Comp.：Channel Stop origin （４）　Constant Comp.：Partial absorption in SiO2 F3 parameters： T1 (normalization), C1(center), S1(sigma) T2 (relative to T1), C2(spth/2, fixed), S2 (1.78×S1, fixed) T3 (relative to T1), F3(三角形の幅, 0.5×C1) T4 (T1で規格化した面積)　→　フリーパラメタ 6個 でフィット BI structure

Line profile for FI1 sensor Se-L line E=1.379keV O-K line E=0.525keV T2=0.033, T4=0.0052 T2=0.020, T4=0.0045 (Seg.B)

Line profile for BI1 sensor C-K line E=0.277keV O-K line E=0.525keV Al-K line E=1.487keV T2=0.19, T4=0.015 T2=0.071, T4=0.011 T2=0.078, T4=0.016

スペクトルの比較 FI2 g02346 カウント数 エネルギー分解能 [eV] 補正前 422615 134.3±0.2 補正後 422923 134.4±0.2 補正＋バッドコラム除去 415933 133.7±0.2

スペクトルの比較 FI2 g02346 カウント数 エネルギー分解能 [eV] 補正前 120393 171.4±0.5 補正後 120708 補正＋バッドコラム除去 118783 170.4±0.5

スペクトルの比較 BI1 g02346 カウント数 エネルギー分解能 [eV] 補正前 273272 133.1±0.2 補正後 302360 131.2±0.2 補正＋バッドコラム除去 288983 131.0±0.2

スペクトルの比較 BI1 g02346 カウント数 エネルギー分解能 [eV] 補正前 62403 165.7±0.6 補正後 72057 162 カウント数 エネルギー分解能 [eV] 補正前 62403 165.7±0.6 補正後 72057 163.0±0.6 補正＋バッドコラム除去 68894 162.6±0.6

Energy and Pulse-height Linearity FI1, Seg.C BI1, Seg.C

Energy Resolution (FWHM) FI-1, Seg.C BI-1, Seg.C

Quantumn Efficiency Measurement Relative Efficiencies of FM-FI0,FI1,FI2,FI3,BI0,BI1 and XIS-EU are measured by irradiating X-rays from the spectrometer to whole the CCD area. Generator beam current is always monitored and stabilized <1%. XIS-EU was cross-calibrated to a Gas PC on 2003Dec & 2004Jul. XIS-FM are not installed in the chamber with the Gas PC simultaneously. The gas PC was calibrated through the slant incident method in 2004 January.

Slant Incident Method: Application to Gas PC X-rays X-rays We determined to use the Gas PC as the reference counter

Gas PC Spectra for Different Incident Angle Counts PH (ch) 45° 30° 0°

Ratio of Counting rate of Gas PC 30°/0° 45°/0° Best fit estimate Poly propylene thickness 1.01±0.06mm H2O 0.281±0.048mm P10gas dead layer 79.6±9.7mm

2004年１月実験の再解析 比例計数管の窓のメッシュ 30°/0°と45°/0°を 同時フィット Ｘ線発生装置の長期変 動を補正 50mm厚、0.5mmピッチ 開口率0.763 (X線で実測） 断面○ではなく□を仮定 30°/0°と45°/0°を　　　　　　　　　　　　　　同時フィット Ｘ線発生装置の長期変　　　　　　　　　　　　　　　　　動を補正

XIS-EU and Gas PC cross calibration using a entrance slit

X-rays through Slit ( ~1mm) Dispersion Direction

PC Spectra and CCD Spectra PC (0.525keV) XIS-EU (0.525keV)

-1deg offset slant-PC is assumed Best Fit Estimates SiO2 0.471±0.039mm Si 0.205±0.029mm Si3N4 0.000±0.03mm Si depletion 65mm fixed Constant Factor 0.852±0.035 -1deg offset slant-PC is assumed

Relative QE XIS-EU FI0 FI1 Red=line component FI2 FI3

Relative QE of BI0,BI1 to XIS-EU ～80@0.28 keV ～10@0.6 keV

-1deg offset slant-PC is assumed Best Fit Estimates SiO2 0.443±0.039mm Si 0.181±0.029mm Si3N4 0.000±0.016mm Si depletion 68.9±1.7mm Constant Factor 0.857±0.003 -1deg offset slant-PC is assumed

XAFS near the O-Kedge FI-2 Eedge = 0.532 ±0.001 keV red. c2 = 1.3178 (d.o.f. = 418)

-1deg offset slant-PC is assumed Best Fit Estimates HfO2 0.005mm fixed Ag 0.001mm fixed SiO2 0.000±0.0005mm Si depletion 45.7±0.7mm Constant Factor 0.934±0.003 -1deg offset slant-PC is assumed

Upper limit of Surface dead layer in BI-CCD Dispersion Spectrum with BI 0.45 0.5 0.55 0.6 keV μm (H2O) H2O on BI1 <0.11 μm

PCＱＥ model

Detailed (FI) Gate Structure Area_ratio_gate=0.125(=3mm) Additional_gate_thickness / slab_thickness SiO2:0.7 Si:1.0 Si3N4 0.026mm ?

Channel Stop Parameters Area_channelstop=0.0833 (=2mm/24mm) Channelstop_thickness SiO2:0.35mm Si:0.45mm

Absolute QE issues Reconsider the assumptions 96% at 4.5keV for XIS-FI Check grade7 events ? Gas PC window model Mesh measurement ? ACIS BESSY calibration How was the effective area or normalization calibrated ? How about Channel Stop events ? Hidden dead space in FI ? Adopt the BI1 QE as a reference BI QE should not be 1 (at least a few % grade 7 events) How do we model F_data-reduction ? Application of the Slant Incidence method to BI0 after the Astro-E2 launch. Any other good way for the absolute QE cal ? Energy independent factor of 10% is not a problem. Edge structure of 10% might be a problem.

Summary We have completed the calibration experiments on the ground for XIS flight models. Data reduction procedures were updated for Astro-E2 XIS. Conversion to FTOOLS will be required. Profile has less tail component than Astro-E1 XIS. Relative QE between the XIS sensors were accurately measured <5%?. We need further work on absolute QE.