Background Rate Measurements at XRCF

Instrumental background measurements at XRCF were performed between flat-field tests, while the energy of the X-ray source was being changed. The data were acquired with the ACIS door in the open position. During each background measurement, the small gate valve where the X-ray beam line enters the thermal vacuum chamber was closed. Occasionally, however, the gate valve was opened by the TRW test director to facilitate tuning of the X-ray source in preparation for the next flat-field test. The background measurements were acquired in 11 science runs performed during the period 1997 May 10-18. For each run, ACIS was operated in either imaging or spectroscopy array mode, with all six detectors in the appropriate array read out. The ACIS telemetry event lists used in this analysis are identified by Science Run Number and TRW ID in Table 4.69.
 

Table 4.69: XRCF Phase I Science Runs Used for the Instrumental Background Analysis

Science Run	TRW ID	ACIS Array
76	I-BND-BU-2.027	Imaging
85	I-BND-BU-2.025	Spectroscopy
90	I-BND-BU-2.003	Imaging
99	I-BND-BU-2.029	Spectroscopy
105	I-BND-BU-2.013	Imaging
109	I-BND-BU-2.015	Spectroscopy
113	I-IAI-BG-3.001	Imaging
127	I-IAS-BG-3.002	Spectroscopy
150	I-IAS-BG-3.902	Spectroscopy
152	I-IAI-BG-3.901	Imaging
155	I-IAI-BG-3.901A	Imaging

Frames were processed for events by the ACIS electronics using the standard event thresholds for front-illuminated (38 ADU) and back-illuminated (20 ADU) ACIS detectors, along with the standard split threshold of 13 ADU. The exposure time per frame was 3.34 s. Light curves were then plotted showing the number of counts per frame vs. frame number. Periods of time when the gate valve was open were easily identified by the abrupt increase in the number of counts per frame. The range of frame numbers contaminated by X-rays were noted and all events in these frames were removed from the event lists. The events in each list were then sorted into spectra by detector quadrant and event grade.

For each detector, we combined the spectra from all the science runs to increase the signal-to-noise. The total exposure time thus obtained was $\sim63.9$ ks for each imaging array detector, and$\sim49.2$ ks for each spectroscopy array detector. Detectors S2 and S3 are in both arrays and so had total exposures of $\sim113.1$ ks each. To further improve the statistics, we combined the spectra from all eight front-illuminated (FI) flight CCDs. Unfortunately, we could not do the same for the back-illuminated detectors. The ACIS focal plane operating temperature at XRCF was $\rm -110\,^{\circ}C$; this is $\rm 10\,^{\circ}C$ warmer than the designed operating temperature. Consequently, the background data taken at XRCF with detector S1 were contaminated by an increase in the number of hot pixels and could not be used in the instrumental background analysis.

Rejection of particle-induced events in the ACIS detectors is accomplished using event amplitude and event grade selection criteria (i.e., the sum and distribution of charge in $3 \times 3$-pixel events). The passband of the ACIS detectors folded with the HRMA is$\sim 0.1$-10 keV, so we rejected all events with total pulse-height amplitudes greater than 10 keV. This eliminated 19.5% of the 781,015 events accumulated in the FI detectors and 80.0% of the 26,636 events accumulated in detector S3. We then applied the standard ASCA G02346 event grade selection criteria to the data to reject all events with pixel geometries that were inconsistent with the physical interaction of X-rays in the ACIS detectors (i.e., ASCA G157 events). The results are presented in Figures 4.103 & 4.104.
 

Figure 4.103: Background spectra measured at XRCF in the ACIS front-illuminated flight devices. The spectra have been summed over the four quadrants of all eight front-illuminated flight devices. The upper and lower panels show the spectra of the rejected (ASCA grades 157) and the unrejected (ASCA grades 02346) background events, respectively, in the passband 0.1--10~keV. The lower panel shows Mn K-alpha line emission due to scattered X-rays from the internal contamination monitor. The Au L-alpha line is caused by particle-induced fluorescence in the framestore covers.

Figure 4.104: Background spectra measured at XRCF in the ACIS back-illuminated flight device S3. The spectra have been summed over all four quadrants. The upper and lower panels show the rejected (ASCA grades 157) and the unrejected (ASCA grades 02346) background spectra, respectively, in the passband 0.1--10~keV.

The upper panel of Figure 4.103 shows the spectrum from 0.1-10 keV of the background events that were rejected by the ASCA G02346 selection criteria in the FI detectors. The rejected background rises abruptly from zero at low energies to a peak of about $\rm0.32\ ct\ cm^{-2}\ s^{-1}\ keV$ at around 0.65 keV, then decays towards higher energies as a power law with index around -2.2, and finally levels off above 6 keV into a long, flat tail with amplitude$\rm\sim1$-$\rm 2\times10^{-3}\ ct\ cm^{-2}\ s^{-1}\ keV^{-1}$. Spectrally-integrated background rates are listed in Table 4.71.

The lower panel of the figure shows the spectrum of the background events in the FI detectors that were not rejected by the event grading scheme. These unrejected background events masquerade as true X-ray events. The spectrum of the unrejected background appears essentially flat from 0.1-10 keV, with the exception of two emission lines. The Mn K$\alpha$ line at 5.894 keV is due to scattered X-rays from the internal contamination monitor (ICM), while the Au L$\alpha$ line at 9.671 keV is caused by particle-induced fluorescence in the detector framestore covers.

The spectrum of the rejected background events in detector S3 is shown in the upper panel of Figure 4.104. The spectrum suffers from low counting statistics, but a small peak can be seen around 0.4 keV; the spectrum then drops between 1 keV and 4 keV, before rising with energy in a broad hump that peaks around 13 keV.

The lower panel shows that the unrejected background in S3 is also relatively flat, with the exception of the Au L$\alpha$ line and a apparent excess in the number of counts below 2 keV. Note there is no sign of an Mn K$\alpha$ line in the spectrum.

To study the spatial distribution of the Mn K$\alpha$ line in the focal plane, we have measured the flux of the line in each quadrant. The Mn K$\alpha$ line fluxes in each detector are listed in Table 4.70, along with the fluxes of the Au L$\alpha$line and the unrejected continuum background. The last column contains the mean and RMS deviation of the count rates in the quadrants. The Mn K$\alpha$ rate varies systematically with position on the focal plane. All of the scattered Mn K$\alpha$ flux is observed in detectors I0 and I1, and the quadrants of I2 and I3 immediately adjacent to I0 and I1, respectively. The collimator and baffling of the internal contamination monitor was intended to require at least two reflections for a photon to travel from the source itself to the focal plane. The diffuse reflectance of the gold-coated surfaces near the detectors is$\le 2 \times 10^{-5}$ sr^-1at 6 keV. The (unbaffled) brightness of the ICM source itself was about $\rm 3\times 10^{3}\ ph\ sr^{-1}$ during the June 1997 XRCF measurements. Had the ICM baffling requirement been met, the observed flux at the focal plane (some 14 cm from the source) should have been less than $\rm 10^{-8}\ ph\ cm^{-2}\ s^{-1}$. The observed rate on I0 and I1 suggests that a subtantial fraction of the source flux is arriving at the detectors after a single reflection. The reason for this is not understood. For reference, a diagram showing the ICM, the focal plane, and the intended ICM beam patterns is presented in Figure 4.105.
 
 

Table 4.70: Instrumental Background Fluxes Measured at XRCF in the Passband 0.1-10 keV

Flight Device	Source	Quad A	Quad B	Quad C	Quad D	Mean (RMS)
		($\rm\times\ 10^{-4}\ ct\ cm^{-2}\ s^{-1}$)
I0	$\rm Mn~K_\alpha$	$2.59\pm 0.56$	$2.07\pm 0.51$	$2.07\pm 0.51$	$3.52\pm 0.63$	$2.56\pm 0.68$
	$\rm Au~L_\alpha$	$0.93\pm 0.40$	$1.76\pm 0.49$	$1.14\pm 0.42$	$1.24\pm 0.43$	$1.27\pm 0.35$
	$\rm Cont.$	$13.15\pm 1.48$	$11.50\pm 1.44$	$12.74\pm 1.44$	$10.87\pm 1.48$	$12.06\pm 1.06$
I1	$\rm Mn~K_\alpha$	$4.04\pm 0.68$	$3.21\pm 0.62$	$4.04\pm 0.68$	$3.62\pm 0.65$	$3.73\pm 0.40$
	$\rm Au~L_\alpha$	$1.14\pm 0.42$	$0.93\pm 0.40$	$0.93\pm 0.39$	$1.04\pm 0.40$	$1.01\pm 0.10$
	$\rm Cont.$	$11.60\pm 1.54$	$14.19\pm 1.57$	$12.74\pm 1.56$	$12.63\pm 1.54$	$12.79\pm 1.06$
I2	$\rm Mn~K_\alpha$	$3.31\pm 0.61$	$0.83\pm 0.36$	$0.00\pm 0.21$	$0.00\pm 0.21$	$1.04\pm 1.57$
	$\rm Au~L_\alpha$	$1.04\pm 0.39$	$1.86\pm 0.51$	$1.86\pm 0.50$	$1.14\pm 0.43$	$1.48\pm 0.45$
	$\rm Cont.$	$10.25\pm 1.43$	$13.25\pm 1.43$	$12.73\pm 1.34$	$13.36\pm 1.32$	$12.40\pm 1.46$
I3	$\rm Mn~K_\alpha$	$0.00\pm 0.15$	$0.10\pm 0.21$	$1.35\pm 0.43$	$3.73\pm 0.66$	$1.29\pm 1.73$
	$\rm Au~L_\alpha$	$0.73\pm 0.36$	$1.45\pm 0.44$	$1.24\pm 0.43$	$0.41\pm 0.33$	$0.96\pm 0.47$
	$\rm Cont.$	$11.70\pm 1.20$	$9.43\pm 1.17$	$11.19\pm 1.34$	$13.15\pm 1.53$	$11.37\pm 1.54$
S0	$\rm Mn~K_\alpha$	$0.00\pm 0.24$	$0.00\pm 0.19$	$0.00\pm 0.20$	$0.00\pm 0.14$	$0.00\pm 0.00$
	$\rm Au~L_\alpha$	$0.81\pm 0.43$	$1.88\pm 0.56$	$1.88\pm 0.57$	$2.02\pm 0.59$	$1.65\pm 0.56$
	$\rm Cont.$	$13.72\pm 1.48$	$9.55\pm 1.37$	$11.70\pm 1.48$	$11.97\pm 1.50$	$11.73\pm 1.71$
S2	$\rm Mn~K_\alpha$	$0.00\pm 0.16$	$0.06\pm 0.17$	$0.23\pm 0.20$	$0.23\pm 0.20$	$0.13\pm 0.12$
	$\rm Au~L_\alpha$	$0.64\pm 0.27$	$1.46\pm 0.34$	$0.99\pm 0.31$	$1.23\pm 0.33$	$1.08\pm 0.35$
	$\rm Cont.$	$13.92\pm 0.98$	$12.64\pm 0.99$	$12.87\pm 0.98$	$12.81\pm 0.99$	$13.06\pm 0.58$
S4	$\rm Mn~K_\alpha$	$0.00\pm 0.14$	$0.27\pm 0.30$	$0.00\pm 0.24$	$0.00\pm 0.27$	$0.07\pm 0.13$
	$\rm Au~L_\alpha$	$1.35\pm 0.51$	$1.21\pm 0.49$	$1.48\pm 0.54$	$1.08\pm 0.51$	$1.28\pm 0.17$
	$\rm Cont.$	$11.98\pm 1.44$	$12.92\pm 1.51$	$16.55\pm 1.67$	$17.09\pm 1.67$	$14.63\pm 2.56$
S5	$\rm Mn~K_\alpha$	$0.40\pm 0.33$	$0.54\pm 0.36$	$0.13\pm 0.27$	$0.00\pm 0.24$	$0.27\pm 0.25$
	$\rm Au~L_\alpha$	$1.61\pm 0.52$	$1.21\pm 0.49$	$1.08\pm 0.47$	$1.21\pm 0.49$	$1.28\pm 0.23$
	$\rm Cont.$	$10.36\pm 1.43$	$11.97\pm 1.49$	$11.03\pm 1.39$	$12.51\pm 1.46$	$11.47\pm 0.96$
S1$^{\ast}$	$\rm Mn~K_\alpha$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$
	$\rm Au~L_\alpha$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$
	$\rm Cont.$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$	$\cdots\hfil$
S3	$\rm Mn~K_\alpha$	$0.23\pm 0.24$	$0.06\pm 0.23$	$0.00\pm 0.21$	$0.00\pm 0.20$	$0.07\pm 0.11$
	$\rm Au~L_\alpha$	$0.53\pm 0.30$	$1.58\pm 0.40$	$0.82\pm 0.35$	$0.94\pm 0.34$	$0.97\pm 0.44$
	$\rm Cont.$	$22.47\pm 1.23$	$24.93\pm 1.33$	$26.80\pm 1.33$	$22.41\pm 1.23$	$24.15\pm 2.12$
$^{\ast}$ Background data taken at XRCF for detector S1 were contaminated by an increase in the number of hot pixels as a result of operating the focal plane at $-110^\circ$ C.

 

Figure 4.105: Cross-section through the ACIS detector housing showing the internal contamination monitor (ICM) X-ray source, its intended illumination pattern, and the ACIS focal plane. In the drawing the Observatory X-axis is up and the Z axis points to the right. Scattered X-ray flux from the ICM is detected on chips I0 and I1, as well as on high-Z quadrants of chips I2 and I3. The scattering path is unknown. The two dimensions shown are in inches.

The total Mn K$_{\alpha }$ flux in I0 and I1 is about $\rm 3.2\times10^{-4}\ ct\ cm^{-2}\ s^{-1}$, while the estimated continuum flux under the peak (unrejected) is about $\rm 4.1\times10^{-5}\ ct\ cm^{-2}\ s^{-1}$. Thus, the scattered Mn K$_{\alpha }$ photons increased the instrumental background in the $5.9\pm0.06$ keV spectral band by a factor of $\sim8.6$ at XRCF. Assuming AXAF is launched in January or February of 1999, about 20 months will have elapsed since the XRCF measurements were made. Given the 2.7 year half-life of the Fe 55 source in the ICM, the predicted on-orbit rate for the Mn K$_{\alpha }$ line drop by a factor of about 2/3. Moreover, since the expected amplitude of the on-orbit continuum exceeds that observed at XRCF by about a factor of about 50 (see Section 4.10.3 below), the scattered ICM is expected to increase the on-orbit background by about $10\%$ in the $5.9\pm0.06$ keV band.

Returning to Table 4.70, we see that the Au L$\alpha$fluxes and the continuum fluxes are roughly uniform across the focal plane, although a constant model is formally rejected in each case. The mean Au L$\alpha$ line flux was $\rm (1.25\pm0.069)\times10^{-4}\ct\ cm^{-2}\ s^{-1}$ in the FI detectors, and $\rm (0.97\pm0.22)\times10^{-4}\ ct\ cm^{-2}\ s^{-1}$ in S3. This ratio is consistent, within the rather large errors, with the expected relative quantum efficiency of the front- and back-illuminated detectors at 9.7 keV (see section 4.8.) Because the effective area of the ACIS/HRMA is so low at the energy of the Au L$\alpha$ line, researchers observing faint diffuse sources can easily remove this extra source of background by applying a high-energy cutoff.

The mean background continuum fluxes for ASCA grade G02346 and G157 events are listed in Table 4.71. The errors quoted in the table are the standard deviations of the means. These measurements indicate that the total particle-induced count rate in a frontside device is about 26 times higher than the rate in a backside device, but the rejection efficiency of the frontside device is about 52 times better, so the net result is that the unrejected background rate in a backside device is about twice that of a frontside device.

Table 4.71: Mean Background Continuum Fluxes Measured at XRCF in the Passband 0.1-10 keV

Flight	Unrejected Events	Rejected Events	Rejection
Device	(G02346)	(G157)	Efficiency
	($\rm ct\ cm^{-2}\ s^{-1}\ keV^{-1}$)	($\rm ct\ cm^{-2}\ s^{-1}\ keV^{-1}$)	(%)
FI	$1.24\pm0.29\times10^{-4}$	$2.02\pm0.05\times10^{-2}$	99.4
S3	$2.42\pm0.11\times10^{-4}$	$5.26\pm0.23\times10^{-4}$	68.5