6 Background line intensities: Comparisonsand implications

The instrumental line background of an instrument is determined by its material composition, the radiation environment, and the employed detection principle. In this respect it is interesting to compare the COMPTEL instrumental lines with those identified in the data from other low-energy $\gamma$-ray experiments on satellite platforms such as the NaI Gamma-Ray Spectrometer (GRS) onboard the Solar Maximum Mission (SMM) and the Ge Gamma-Ray Spectrometer (GRS) flown on HEAO 3. In doing so, one must keep in mind that an instrument's energy resolution is important for resolving lines. At 1 MeV, the energy resolution (FWHM) of COMPTEL is 9.8% and 8.8% in $E_{\rm tot}$and E₂, respectively (Schönfelder et al. 1993). At the same energy, the resolution of GRS-SMM was about 5.4% (Forrest et al. 1981), and that of GRS-HEAO 3 was $\stackrel{\textstyle _<}{_{\sim}}$0.3% (Mahoney et al. 1980). Therefore, the Ge spectrometer onboard HEAO 3 is far superior for $\gamma$-ray line studies compared to COMPTEL and GRS-SMM, which have similar resolution. The CGRO satellite is kept in a circular orbit of 28.5$^\circ$ inclination at altitudes between about 330 km and 515 km (see Fig. 9). The SMM and HEAO 3 missions were operated in circular orbits of 28.5$^\circ$ inclination at about 490-570 km altitude, and 43.6$^\circ$ inclination at about 500 km altitude, respectively. Similar to COMPTEL, the most abundant element in the material composition of both GRS-SMM (E. Chupp, priv. comm., 1999) and GRS-HEAO 3 (Wheaton et al. 1989) is Al in the instrument structures. Also, all instruments contain significant amounts of Fe, Ni, Cu, and Cr in, e.g., electronics components.

Five of the eight isotopes identified in the COMPTEL line background, namely ²²Na, ²⁴Na, ²⁸Al, ⁵²Mn, and ⁵⁷Ni, are due to activation of Al and Fe, Ni, Cu, and Cr nuclei in the instrument structure. As expected, each of these isotopes has also been identified in GRS-SMM (Share et al. 1989) and GRS-HEAO 3 (Wheaton et al. 1989) data. For GRS-SMM, however, an additional production channel for ²²Na and ²⁴Na were proton and neutron reactions, respectively, on ²³Na in the NaI detectors. Other isotopes identified in GRS-SMM and GRS-HEAO 3 that result from activation of the five elements and which have major lines below about 1 MeV should in principle also be present in the COMPTEL background, however, their events are suppressed by the D1 and D2 detector thresholds used in COMPTEL analyses. On the other hand, instrumental lines due to activation of Cs and I (GRS-SMM) or Ge (GRS-HEAO 3) are absent in COMPTEL since they exhibit a ToF value not used in the analysis (I) or since these elements are not present in the instrument (Cs, Ge).

The different detection principles employed in the three instruments result in significant differences in the relative importance of specific isotopes, their detectability, and rejectability. COMPTEL requires coincident interactions in the D1 and D2 detector and therefore is particularly susceptible to multiple photon events such as the $\beta^+$-decay of ²²Na or the $\beta^-$-decay of ²⁴Na (see Sect. 3). Consequently, multiple photon events are relatively more important than single photon events. On the other hand, individual photons from these decays produce photopeaks in E₂ that can be used for identifying individual isotopes and for determining their background contribution. In addition, the characteristic E₁-E₂ signature of multiple photon events provides a wide range of options for their rejection (or enhancement). In the SMM and HEAO 3 spectrometers events are triggered by individual photons. Unlike the COMPTEL veto system, which consists of plastic scintillator domes, the anti-coincidence systems of the SMM and HEAO 3 spectrometers were made of CsI crystals and therefore sensitive to photons at $\gamma$-ray line energies. For these two detectors therefore the probability for a radioactive decay to trigger the detectors is proportional to the number of emitted photons, as is the probability to trigger the anti-coincidence systems. To first order, the net probability of a radioactive decay for generating a background event is therefore independent of the photon multiplicity for GRS-SMM and GRS-HEAO 3. As long as the decays occur outside the detectors, individual photons will give rise to photopeaks in the latter two instruments, and these backgrounds are harder to suppress by event selections than in COMPTEL. As far as decays inside the detectors are concerned, $\beta$-decays are particularly hard to identify for GRS-SMM (e.g. activation of Na) and GRS-HEAO 3 (activation of Ge) since the added $\beta$-continuum will broaden $\gamma$-ray lines beyond recognition. In COMPTEL, activation in the D2 detector is effectively eliminated by event (ToF) selection.

Since COMPTEL is the first double-scattering Compton telescope operated in a near-Earth orbit, the instrumental background experienced during this mission may provide guidance for the design of future instruments. Below, some implications of the COMPTEL instrumental line background for conducting $\gamma$-ray line studies with this and future instruments are discussed.

The by far strongest astrophysical $\gamma$-ray line signal detected by COMPTEL is 1.8 MeV line emission from ²⁶Al in the interstellar medium (Diehl et al. 1995a). When observing along the galactic plane, the average event rate due to this extended line source is about 7 10^-4 s^-1 for imaging selections (see Appendix A.2). Typical event rates due to point sources observed by COMPTEL in the light of $\gamma$-ray lines, such as the supernova remnant Cas A (⁴⁴Ti at 1.12 MeV, Iyudin et al. 1997), or the type Ia supernova 1991T (⁵⁶Co at 0.85 MeV and 1.24 MeV, Morris et al. 1997b), or the Vela region (²⁶Al at 1.8 MeV, Diehl et al. 1995b) are about 3 10^-5 s^-1. Typical event rates arising from individual background isotopes are about 10^-1 s^-1 (see Fig. 8). Below 3 MeV, instrumental background lines account for 10-50% of the total background rate (see Fig. B.8). It follows that the signal-to-background ratio for astrophysical $\gamma$-ray lines in general is less than 1%, which can be enhanced to a few percent by fine-tuning the imaging event selections described in Appendix A.2 for analyzing a specific $\gamma$-ray line.

 

 

Table 2: Background isotope properties relevant for astrophysical $\gamma$-ray line studies. For each isotope the following quantities are listed: the efficiency for producing a background event under imaging event selections, the material(s) in which the isotope is produced, and the mission-averaged activity. Further details are given in the text

Isotope	Efficiency	Material	Activity
	(Imaging Sel.)		[g-1s-1]
2D	7 10-4	D1 scintillator	3.4 10-3
22Na	8 10-4	D1 Al structure	7 10-4
24Na	9 10-4	D1 Al structure	9 10-4
28Al	2 10-4	D1 Al structure	2 10-3
40K	2 10-4	D1 PMT glass	0.2
52Mn	2 10-3	Fe around D1	4 10-4
		Cr around D1	3 10-4
		Ni around D1	2 10-4
		Cu around D1	7 10-5
57Ni	5 10-4	Ni around D1	5 10-4
		Cu around D1	6 10-5
208Tl	2 10-3	D1 PMT glass	10-2

In this respect it is interesting to take a look at the efficiency, determined by Monte Carlo simulation, for various isotopes to produce a background event (see Table 2). For imaging selections, this efficiency is about 7 10^-4 and 2 10^-4 for ²D and ⁴⁰K, both of which give rise to type A (single photon) background events. A major source of type C multiple photon events is ²⁴Na with an efficiency of about 9 10^-4. The efficiency for ²⁰⁸Tl, a minor source of type C events which is assumed to have the same spatial distribution as ⁴⁰K, is about 2 10^-3. This is an order of magnitude larger than the efficiency for ⁴⁰K and again illustrates that COMPTEL is more susceptible to multiple photon decays than to single photon decays. The more stringent CDG event selections reduce these efficiencies by factors of 4-7 and 2-3 for type A and type C background events, respectively. For comparison, the detection efficiencies for celestial lines around 1-2 MeV are about 2 10^-4 using imaging event selections.

A goal for future Compton telescopes is to minimize background production by design. Passive material should be reduced, and manufactured from materials with low activation. Based on simulated efficiencies and the COMPTEL mass model, "mission-averaged'' activities (in decays per g of material per s) were derived for the eight identified isotopes using data representing the first 7 years of the mission (June 1991 to April 1998). These activities, together with the above mentioned simulated efficiencies and the activated materials, are summarized in Table 2. The relative yields for ⁵²Mn and ⁵⁷Ni were determined in hadron simulations (P. Jean, priv. comm., 1997). The numbers for the average activities are accurate within factors of 2-3. Their major sources of uncertainty are the simulated efficiencies and the mass normalization, which are currently based on the assumption that the activation is homogeneous in a particular material. This may not be accurate for isotopes produced by thermal neutrons, in particular ²⁸Al. As a cross-check, the average activities were compared to simulations of activation of COMPTEL materials, neglecting the contributions of secondary particles, for a solar-maximum radiation environment (P. Jean, priv. comm., 1997). The measured and the simulated activities are generally consistent within factors of a few for isotopes mainly produced by SAA protons. Hence these average activities may be used for estimating the line background rates of future instruments in a similar orbit, taking into account how the mass distribution differs from that of COMPTEL.