Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0305-1-0.ps.gz Äàòà èçìåíåíèÿ: Wed Sep 4 16:35:02 2013 Äàòà èíäåêñèðîâàíèÿ: Mon Mar 3 13:13:02 2014 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: http astrokuban.info astrokuban

XMM­Newton CCF Release Note
XMM­CCF­REL­305
EPIC­MOS contamination parameters
S. Sembay & R. Saxton
September 4, 2013
1 CCF components
Name of CCF VALDATE List of Blocks changed Change in CAL HB
EMOS1 CONTAMINATION 0001.CCF 2000­01­01 CONTAM DEPTH YES
EMOS2 CONTAMINATION 0001.CCF 2000­01­01 CONTAM DEPTH YES
2 Change
It has become apparent that the response of the MOS cameras (primarily MOS2) has deteriorated at
low­energies (< 1 keV) over the course of the mission. It is suspected that this is due to contaminants
which have adhered to the surface of the cameras, absorbing a fraction of the incoming photons.
These new CCF elements tabulate the depths of contaminating elements as a function of time
(see Tab. 1). Depths have been tabulated from the beginning of the mission (2000­01­01) and
extrapolated until 2028­07­06. Carbon, Oxygen and Fluorine are common components of measured
contaminants on some other X­ray instruments (e.g. ACIS on Chandra). In this release the entries
in the CCF for Oxygen and Fluorine are placeholders and have zero depth. The MOS contaminant
is currently modelled as pure Carbon, similar to that observed on the RGS. There is no evidence
for a contaminant on the pn camera.
2.1 Determination of the contaminant
Figure 1 shows the observed count rate ratio in the 0.1­0.75 keV to 0.98­3.0 keV bands from ob­
servations of the SNR 1E0102.2­7219, which is known to have a highly stable spectrum. Di#erent
positions of the source on the detector are indicated by the color coding (see figure caption). Part
of the scatter in the ratio is due to position­dependent e#ects such as vignetting and charge re­
distribution, while the strong trend with revolution number, seen in MOS2, may only be plausibly
1

XMM­Newton CCF Release XMM­CCF­REL­305 Page: 2
Table 1: CCF column description
Column name Desciption
TIME observation date (MJD)
C DEPTH depth of carbon in microns
O DEPTH depth of oxygen in microns
F DEPTH depth of flourine in microns
explained by the existence of a thin but growing contaminant. The evidence for a temporal trend in
the MOS1 data is marginal. Similar trends are observed in the SNR N132D which also has a long
baseline of observations.
The continuous line on each plot shows the ratio which is predicted if the e#ective area (which
already accounts for position­dependent e#ects) is modified by an evolving Carbon absorber. The
depth of this material (in microns) is shown, as a function of time, in Figure 2. In Fig. 2 the points
show the thickness of Carbon required to produce a 0.1­0.75 keV to 0.98­3.0 keV count rate ratio
equal to that predicted by the standard IACHEC model of this source (Plucinsky et al. 2012). These
points have then been fitted with an exponential model to give a smooth deposition of contaminant
(although in the current epochs it appears to the eye to be fairly linear).
Note that directly observing a well­defined carbon edge in continuum sources is not possible due
to the complicated shape of the RMF at low energies.
All measurements were taken within 2 arcminutes of the camera optical­axis. It is assumed that
the contaminant has been deposited uniformly over the field of view and hence no spatial dependency
has been introduced into the calibration.
3 Scientific impact of this update
The e#ective area of the MOS cameras, calculated by the task arfgen, will be reduced at energies
< 1 keV. In practise, the major e#ect will be around the Carbon edge, between # 300 - 500 eV,
which will experience a reduction in the e#ective area, and a consequent rise in measured flux, in
this energy range, of # 10% for observations taken in 2013 with MOS­2. The e#ect will be less for
earlier observations and for those made with MOS­1.
4 Estimated scientific quality
This change should lead to a more consistent cross­calibration between the three EPIC cameras
than is presently available, especially at later epochs.
To test this we have fit the EPIC­pn and MOS­2 spectra of 3C 273 from revolution 2308 with

XMM­Newton CCF Release XMM­CCF­REL­305 Page: 3
[CR(0.1­0.75)/CR(0.98­3.0)]
0 500 1000 1500 2000 2500
Rev
1.50
1.55
1.60
1.65
1.70
Ratio
[CR(0.1­0.75)/CR(0.98­3.0)]
0 500 1000 1500 2000 2500
Rev
1.50
1.55
1.60
1.65
1.70
Ratio
Figure 1: The 0.1--0.75 keV / 0.98--3.0 keV count rate ratio from the SNR 1E0102.2­7219 as a function
of the revolution number; upper (MOS­1), lower (MOS­2). Points are colour­coded to reflect the
position of the source on the detector relative to the boresight: blue(above boresight), orange (to
the right), green (below), red (to the left). The solid blue line represents the expected ratio after
correcting for position­dependent e#ective area di#erences and a film of Carbon contaminant, whose
depth increases over the course of the mission.

XMM­Newton CCF Release XMM­CCF­REL­305 Page: 4
0 1 2 3 4 5
Time (s/1e8)
0.00
0.01
0.02
0.03
0.04
Depth
(microns)
MOS1
MOS2
500 1000 1500 2000 2500
Orbit
Figure 2: The depth of Carbon contaminant, in microns, inferred to lie on the MOS cameras as
a function of observation time (o#set from 2000­01­01T12:00:00). Solid curves represent a best­fit
exponential model.
a model of a double power­law, absorbed by the galactic column and multiplied by an instrument­
dependent constant (Fig. 3). The new e#ective area, including a correction for absorption by a layer
of Carbon, gives a better agreement between the EPIC­pn and MOS­2 data.
5 Test procedure and results
These CCFs will be used by the task arfgen to calculate the total e#ective area of the MOS cameras
including the e#ects of contamination. We have ran the task for the two cameras on data from
2000­01­01 and from 2013­01­01. The di#erence in the e#ective areas can be seen in figure 4.
6 Future changes
The depths have been extrapolated into the future using an exponential model. This should be
reviewed periodically to ensure that the deposition of the contaminant(s) continues at the expected
rate.
Should su#cient high­quality data become available for o#­axis sources it may be possible to
test the spatial dependency of the contamination over the field of view.

XMM­Newton CCF Release XMM­CCF­REL­305 Page: 5
0.1
1
10
normalized
counts
s
-1
keV
-1
Black (pn); Red (MOS2/Old Response); Blue (MOS2/New Response)
3C 273 - Orbit 2308
1
0.5 2
0.8
0.9
1
1.1
1.2
ratio
Energy (keV)
Figure 3: A fit to an observation of the QSO 3C273 from revolution 2308 with an absorbed dou­
ble power­law model. The model was first fit to the EPIC­pn points (black) and then applied
to the MOS­2 data, using the original e#ective area file (red) and an e#ective area corrected for
contamination (blue).
Figure 4: The ratio of e#ective areas between two observations taken in 2013­01­01 and 2000­01­01
for MOS­1 (left) and MOS­2 (right). The ratios are due to the contamination layers which have
formed on the MOS cameras since launch. This is much deeper in the MOS­2 camera.

XMM­Newton CCF Release XMM­CCF­REL­305 Page: 6
7 References
P.P.Plucinsky et al. ''Cross­calibration of the x­ray instruments onboard the Chandra, Suzaku, Swift,
and XMM­Newton Observatories using the SNR 1E 0102.2­7219'', 2012, SPIE, 8443, 12