From naya@tgrs2.gsfc.nasa.gov Wed Nov 6 17:08:21 1996 Date: Wed, 6 Nov 1996 17:08:08 -0500 (EST) Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable To: GCNEWS@astro.umd.edu From: naya@tgrs2.gsfc.nasa.gov (Juan E. Naya) Subject: Galactic Al, Letter to Nature \title{Detection of high velocity 26Al towards the Galactic centre} \author{Juan E. Naya (1,2), Scott D. Barthelmy (1,2), Lyle M. Bartlett (3), Neil Gehrels (1), Marvin Leventhal (4), Ann Parsons (1), Bonnard J. Teegarden (1), \& Jack Tueller (1)} \institute{ NASA/Goddard Space Flight Center, Greenbelt, MD 20771, USA (2) Universities Space Research Association, 7501 Forbes Blvd. #206, Seabrook, MD 20706-2253, USA (3) NAS/NRC Resident Research Associate, Code 718, NASA/GSFC, Greenbelt, MD 20771, USA (4) Dept. of Astronomy, University of Maryland, College Park, MD 20742-2421, USA } \begin{abstract} Theory predicts that radioactive 26Al (which has a half-life of 0.72 Myr) is released into the interstellar medium by nova and supernova explosions, from winds of massive stars in the Wolf-Rayet phase, and from less-massive stars in very late stages of their evolution (in the Asymtotic Giant Branch phase). Observations of 1809 keV g-ray emission from 26Al can therefore be used as a tracer of Galactic Nucleosynthesis during the past million years. The irregularity of the emission in the plane of the Galaxy suggests that the dominant sources are likely to be massive stars and supernovae; the other predicted sources are older, and therefore expected to be distributed more uniformly. Here we report the detection of the 1809 keV emission line from the direction of the Galactic centre, and we show that the line width is approximately three times that expected from the effect of Doppler broadening due to Galactic rotation. The high velocities inferred from the line width favour an origin of 26Al in supernovae or Wolf-Rayet stars. Moreover, the fact that the 26Al has maintained such high velocities is difficult to reconcile with our current understanding of the propagation of material in the interstellar medium. \end{abstract} The Gamma Ray Imaging Spectrometer (GRIS)10 is a balloon-borne high-resolution g-ray spectrometer consisting of an array of seven germanium detectors surrounded by a thick (15 cm) active NaI anticoincidence shield. This instrument is one of the most sensitive high resolution gamma-ray spectrometers, as shown by results from previous flights such as the observations of the Galactic centre 511 keV positron annihilation line11 and the SN 1987a 56Co lines12. GRIS has been recently reconfigured with a wide field collimator (100šx75š FWHM field-of-view), and a 15 cm thick NaI blocking crystal to optimize its capability for observation of diffuse gamma-ray sources such as the cosmic diffuse background and Galactic 26Al emissions. The measurements reported here were made on a flight from Alice Springs, Australia, on 1995 October 24-26. The total germanium detector area and volume were 237 cm2 and 1647.6 cm3 respectively. The duration at float altitude was 32 hours at an average atmospheric depth of 3.8 g cm-2. The collimator was always pointed at the zenith. From Alice Springs, the Galactic centre, south Galactic pole, and Galactic plane (l=240š) transit nearly overhead (see top of Fig. 1). GRIS observed alternating 10 minute exposures with the blocking crystal open and closed. The variation of the 1809 keV line intensity during the flight is displayed in Fig. 1. These values include the instrumental background line due to interactions of cosmic-ray induced neutrons with Al in the instrument (i.e. 27Al(n,np)26Mg* and 27Al(n,2p)26Na(b-)26Mg*). Both the Galactic line and the background line are included in the model we have fit to the data. The variation of the rates with time clearly shows an excess at the Galactic centre and remains constant during the Galactic pole and Galactic plane transits. The excess is detected independently in each of the 6 operating Ge detectors, as expected from a real source detection. With a 100š field-of-view, the net flux is dependent on the distribution of 26Al emission. Using a model derived from the COS B measurement of the Galactic gamma-ray distribution (>70 MeV)13, which has been used to fit most previous results4,14-20, and taking into account the GRIS instrument response plus the effects of atmospheric absorption, we obtained a flux of 4.8 ± 0.7·10-4 photons s-1 cm-2 rad-1. This value is compatible with previous measurements and yields a detection at the 6.8s level. The measured astrophysical 1809 keV line spectrum is displayed in Fig. 2c. It was derived by subtracting the Galactic pole and Galactic plane accumulation (Fig. 2b) from the sum of both Galactic centre transits (Fig. 2a). Since the line rates observed at the Galactic pole and the Galactic plane transits are very similar (Fig. 1) and, we do not expect significant 1809 keV line emission from the Galactic pole region, the observed line in Fig. 2b should be mostly of instrumental origin. The spectrum was fit with a Gaussian profile model. The resulting astrophysical line centroid at 1809.0 ± 0.6 keV is consistent with the 26Al rest energy of 1808.65 ± 0.07 keV. The width of the 26Al line is the most remarkable result. The measured width (FWHM = 6.4 [+1.2,-1.1] keV) is the quadratic sum of the instrument resolution and the intrinsic width. The instrument energy resolution was precisely determined by fitting a line to the measured widths of many intense narrow lines in the background spectrum. The energy resolution at 1809 keV is 3.41 ± 0.1 keV, which implies that the intrinsic width of the astrophysical line is 5.4 [+1.4,-1.3] keV FWHM. We investigated the potential sources of systematic error that might generate such a broad feature in the GRIS spectrum. The instrumental background spectrum in this energy range contains a number of narrow lines, including the one at 1809 keV (see Fig. 2a and 2b). The background line at 1779 keV results from neutron capture, 27Al(n,g)28Al, followed by b-decay (mean life 3.23 min) to 28Si. It is interesting to point out that both the 1779 and 1809 keV background lines originate mainly from the interaction of neutrons in aluminum. Thus, systematic errors in the data analysis that could simulate a broad line or a net cosmic flux should give similar results for both lines. The 1779 and 1809 keV lines were both analyzed for the data corresponding to the four transits indicated in Fig. 1. A model consisting of a Gaussian plus a power law was fit to each line plus surrounding continuum . The resulting best fit line parameters are displayed in Fig. 3. The 1809 keV line intensity (see Fig. 3a) increases by a factor of three at the Galactic centre transits while the 1779 keV line intensity remains unchanged, within the statistics, during the whole flight. This indicates that the 1809 keV line excess observed at the Galactic centre transits is not instrumental in origin. The width of the lines shows a similar behavior (see Fig. 3b). The 1809 keV line gets broader in both Galactic centre transits and is consistent with a narrow line for both the Galactic pole and Galactic plane transits. The 1779 keV line width remains practically unchanged for the four targets, indicating that the broad nature of the 1809 keV line excess observed at the Galactic centre transits is not instrumental. Finally, the position of the centroid with respect to the line rest energy does not vary significantly with the target for either of the lines (see Fig. 3c), demonstrating that the quality of the gain correction does not change during the flight. We have estimated a conservative upper limit on the additional line broadening produced by systematic effects to be <0.5 keV. This limit was derived by assuming that the 1779 keV line parameter fluctuations shown in Fig. 3 were not of statistical origin and reflected real changes in the instrumental performance. The comparison of our present result with past observations is inconclusive, largely because of the low significance of previous line-width measurements. The width reported by the high resolution spectrometer HEAO-C14 (with a 43š field-of-view) was compatible with a narrow line, giving a 1s intrinsic width upper limit of 3 keV. Other high resolution instruments16,20 (with a ª20š field-of-view) have also reported a line-width compatible with a narrow line. An important exception is the previous GRIS measurement with a 20š field-of-view. The line was only detected at the 2.5s significance but the width observed was 5.1±2.2 keV FWHM. SMM15 and COMPTEL4 have made significant detections of this line, but the modest energy resolution of these instruments (based on scintillator detectors) does not constrain the width. A mode consisting of diffuse broad emission plus narrow emission concentrated at the Galactic plane could induce a line broadening dependent on the field-of-view of the instrument. However, the flux detected by GRIS does not reveal the existence of extra flux invisible to the narrow field instruments. Also, a single Gaussian model provides the best fit to the observed feature. The detection of such a wide line has very important implications for the origin of Galactic 26Al. The 1809 keV line was expected to have a width of <1.8 keV FWHM: because of its long lifetime, 26Al should have slowed down before decaying and it should therefore exhibit velocities typical of differential Galactic rotation (i.e. <150 km s-1) plus random motions of the interstellar medium8,9. The measured line broadening is a factor of three larger than predicted and implies that the 26Al decay occurs at velocities >450 km s-1. Only novae, supernovae and Wolf-Rayet stars are able to release this isotope at such high velocities. Thus without reheating or acceleration, AGB stars are not likely to be a major contributor to the production of the Galactic 26Al. Even with high initial velocities (>1000 km s-1), it is difficult to understand how 26Al could maintain these speeds for 106 years. High velocity material in the Galactic plane should be slowed down by interactions with the interstellar medium on timescales of less than 105 years3. New scenarios are required to account for such a broad line. One possibility is the broadening due to the contribution of local sources. Material ejected from local sources could contain freshly synthesized 26Al (less than 105 years old) that has not yet slowed down. In the Galactic centre direction, the closest source of this kind is Loop I. This radio structure, a residual of nearby supernova activity, is centered on the Sco-Cen association, has an angular diameter of 116š and, its centre is at a distance of 170 pc from the Sun21. While this source could emit the broad line, the predicted flux is about 20 times too low to account for our observation6. An alternate explanation is part of the ejecta could escape the high density disk before stopping. Such a phenomenon occurs in galactic 'chimneys,' which are cavities in the interstellar medium created by multiple supernova explosions that act as conduits for the efficient transport of hot gas from the galaxy's disk to its halo22. If these objects were very common in the Galaxy (i.e., typical features of clusters of massive stars) they could explain the observed line broadening. Such a scenario has been alluded to as a possible explanation for the origin of dark matter in the Galactic halo22,23. A third possibility is 26Al incorporated into high velocity dust grains (Pranztos private communication). From the observations of SN 1987A, we learned that high velocity grains form in the ejecta within a couple of years of the explosion24. These grains with a low charge/mass ratio could be only weakly coupled to the ejecta plasma and may not be stopped at the outer shock boundary of the plasma, continuing into the interstellar medium where they might survive in the hot phase for 106 years before being slowed down or destroyed. Unfortunately, the modest angular resolution for the GRIS instrument does not allow us to determine whether the broad line emission comes primarily from sources above the plane (and therefore attributable to chimneys), or mainly from the plane of the Galaxy. Progress on understanding the origin of 26Al requires improved observations performed with instruments combining good sensitivity with high angular and energy resolution. Measurements of the line profile and Doppler shifts due to Galactic rotation in different directions will provide valuable information about the location9 and properties of the 26Al sources. A more sensitive detector will also allow confirmation of the existence of 60Fe line emission, which has been predicted to be distributed in a similar way as the 26Al emission if supernovae were the major source of 26Al 25. The next generation of gamma-ray spectrometers (e.g., the INTEGRAL Spectrometer26) should be able to address these issues providing the most accurate information about the current nucleosynthetic activity in our Galaxy. References 1. Prantzos, N., Diehl, R. Physics Reports, 267, 1-69 (1996) 2. Ramaty, R., & Lingenfelter, R.E. Astrophys. J., 213, L5-L7 (1977) 3. Arnett, D.W. Ann. N. Y. Acad. Sci., 302, 90-92 (1977) 4. Diehl, R., Dupraz, C., Bennet, K., Bloemen, H., Hermsen, W., Knödlseder, J., Lichti, G., Morris, D., Ryan, J., Schönfelder, V., Steinle, H., Strong, A., Swanenburg, B., Varendorff, M., Winkler, C. Astron. Astrophys., 298, 445-460 (1995) 5. Diehl, R., Bennet, K., Bloemen, H., Dupraz, C., Hermsen, W., Knödlseder, J., Lichti, G., Morris, D., Oberlack, U., Ryan, J., Schönfelder, V., Steinle, H., Varendorff, M., Winkler, C. Astron. Astrophys., 298, L25-L28 (1995) 6 del Río, E., Diehl, R., Oberlack, U., Schönfelder, V. & von Ballmoos, P. in The Second Compton Symposium (eds Fichtel, C.E., Gehrels, N. & Norris, J.P.) 171-175 (Am. Inst. of Phys., College Park, 1994) 7. Chen, W., Gehrels, N., Diehl, R. Astrophys. J., 440, L57-L60 (1995) 8. Skibo, J., Ramaty, R. in Gamma-Ray Line Astrophysics, eds. Ph. Durouchoux and N. Prantzos, (New York, AIP) 168-170 (1991) 9. Gehrels, N., Chen, W. Astron. Astrophys. Suppl. Ser. in press 10. Tueller, J., Barthelmy, S.D., Gehrels, N., Teegarden, B.J., Leventhal, M., & MacCallum, C.J. in Nuclear Spectroscopy of Astrophysical Sources, ed. N. Gehrels & G.H. Share (New York:AIP), 439-443 (1988) 11. Leventhal, M. MacCallum, C.J., Barthelmy, S.D., Gehrels, N., Teegarden, B., Tueller, J. Nature, 339, 36-38 (1989) 12. Teegarden, B.J., Barthelmy, S.D., Gehrels, N., Tueller, J., Leventhal M., MacCallum, C.J. Nature, 339, 122-123 (1989) 13. Mayer-Hasselwander H.A., Bennet, K., Bignamy, G.F., Buccheri, Caraveo, P.A., Hermsen, W., G. Kanbach, G., Lebrun, F., Lichti, G.G., Masnou, J.L., Paul, J.A., Pinkau, K., Sacco, B., Scarsi, L., Swanenburg, B.N., Wills, R.D. Astron. Astrophys., 105, 164-175 (1982) 14. Mahoney, W.A., Ling, J.C., Wheaton, Wm. A., & Jacobson, A.S. Astrophys. J., 286, 578-585 (1984) 15. Share, G.H., Kinzer, R.L., Kurfess, J.D., Forrest, D.J., Chupp, E.L. and Rieger, E. Astrophys. J., 292, L61-L65 (1985) 16. MacCallum C.J., Huters A.F., Stang P.D., Leventhal M. Astrophys. J., 317, 877-880 (1987) 17. von Ballmoos, P., Diehl, R., Schönfelder, V. Astrophys. J., 318, 654-663 (1987) 18. Teegarden, B.J., Barthelmy, S.D., Gehrels, N., Tueller, J., Leventhal M., MacCallum, C.J. Astrophys. J., 375, L9-L12 (1991) 19. Varendorff, M., Schönfelder, V. Astrophys. J., 395, 158-165 (1992) 20. Durouchoux, Ph., Wallyn, P., Chapuis, C., Matteson, J., Bowman, B., Pelling, M., Peterson, L., Vedrenne, G., von Ballmoos, P., Malet, I., Niel, M., Lin, R., Feffer, P., Smith, D., Hurley, K. Astron. Astrophys. Supp. Ser., 97, 185-187 (1993) 21. Berkhuijsen, E.M., Hasam, G.C.T. & Salter, C.J. Astron. Astrophys. 14, 252-262 (1971) 22. Normandeau, M., Taylor, A.R., Dewdney, P.E. Nature, 380, 687-689 (1996) 23. Shull, M. Nature, 380, 668-669 (1996) 24. Colgan, S.W.J., Haas, M.R., Erickson, E.F., Lord, S.D., Hollenbach, D.J. Astropys. J., 427, 874-888 (1994) 25. Timmes, F.X., Woosley, S.E., Hartmann, D.H., Hoffman, R.D., Weaver, T.A., Matteucci, F. Astrophys. J., 449, 204-210 (1995) 26. von Ballmoos, P. Exp. Astron., 6, 85-96 (1995) ACKNOWLEDGMENTS Our thanks to the GRIS experiment development team, Stephen Deredyn, Stephen Snodgrass, Chris Miller and Kiran Patel. Launch and recovery support was provided by the crew of the National Scientific Ballooning Facility. We are also grateful to Wan Chen, Peter von Ballmoos, Jürgen Knödlseder and, Dieter Hartmann for discussions. Figure Captions Fig. 1 Evolution of the measured 1809 keV line rates during the October 25 1995 GRIS flight. The data include both astrophysical and background lines. The elevation angles of the different targets are shown at the top (GC 1: G. centre 1st transit; GP: south G. pole; GPl: G. plane at l=240 and; GC 2: G. centre 2nd transit). The count rates increase by a factor of three at the two Galactic centre transits, providing a clear detection of Galactic 26Al. Using a model derived from the Galactic gamma-ray distribution (>70 MeV) measured by COS B (Mayer-Hasselwander et al. 1982) (solid line), we obtained a flux of 4.8 ± 0.7·10-4 photons s-1 cm-2 rad-1 (6.8s level). FIG. 2 g-ray count rate spectrum near 1809 keV. (a) and (b) are the accumulations at both Galactic centre transits and at the Galactic pole and Galactic plane transits. Each channel has a width of 1 keV. These spectra show background line features at 1764 keV from the decay chain of 238U, 1779 keV from the deexcitation of 28Si, and 1809 keV from the deexcitation of 26Mg. The net Galactic gamma-ray spectrum, derived by subtracting spectrum (b) from spectrum (a), is displayed in (c). It shows line emission at only 1809 keV produced in the decay of interstellar 26Al. There is a small residual (~2 sigma) at the location of the 1779 keV background line that is probably due to a combination of small changes in the background line intensity and a small gain shifts. Because the signal-to-background ratio for the 1809 keV line is large (~ a factor of 2) the effect on the 1809 keV intensity will be small (<4 %). This effect is included in the systematic error estimate. The solid curve is the best fit of the data to a Gaussian line shape. Assuming an instrument energy resolution of 3.4 keV, the intrinsic width of the astrophysical line is 5.4 [+1.4,-1.3] keV FWHM, which is more than three times the values expected from previous theories. The significance of this measurement allows one to exclude a line-width merely reflecting Galactic rotation (i.e., 1.8 keV) within 97% probability. Fig. 3 Evolution of the intensity, width and centroid energy of the 1809 and 1779 keV line features from the spectra at the transits of the targets described in Fig. 1. Due to the similar origin of both background lines, systematic errors in the data analysis that could simulate a broad line or a net cosmic flux should give the same result for both lines. The intensity (a) and width (b) of the 1809 keV line increase dramatically at the Galactic centre transits while the 1779 keV background line remains unchanged. This shows that the 1809 keV line excess and broadening observed at the Galactic centre transits is of external origin. The energy of the lines (c) does not vary significantly, demonstrating that the gain correction was properly performed. A conservative upper limit <0.5 keV was estimated for the line broadening induced by possible systematic errors. This limit was derived by assuming that the intensity, width and centroid fluctuations of the 1779 keV line were not of statistical origin and reflected real changes in the instrument performance. ____________________________________________________________________ | Juan E. Naya, Visitor Scientist, Lab for High Energy Astrophysics \ | Code 661, Bldg 2, Rm 241, Email: naya@tgrs2.gsfc.nasa.gov \ | Greenbelt, MD 20771 _ _____ _______ __________________ \ \ / |/ / _ \/ __/ _ \ / ___/ __/ __/ ___/ \ \ Pho 1-301-286-6955 / / _ /\ \/ _ / / (_ /\ \/ _// /__ / \ Fax 1-301-286-1684 /_/|_/_//_/___/_//_/ \___/___/_/ \___/ / \_________________________________________________________________/ ----- End Included Message ----- ----- End Included Message -----