Discovery of recombining plasma associated with the candidate supernova remnant G189.6+3.3 with Suzaku
Shigeo Yamauchi, Moe Oya, Kumiko K. Nobukawa, Thomas G. Pannuti
aa r X i v : . [ a s t r o - ph . H E ] J u l Discovery of recombining plasma associatedwith the candidate supernova remnantG189.6 + Shigeo Y
AMAUCHI ∗ , Moe O YA , Kumiko K. N OBUKAWA , and Thomas G.P
ANNUTI Faculty of Science, Nara Women’s University, Kitauoyanishimachi, Nara 630-8506, Japan Department of Physics, Kindai University, 3-4-1 Kowakae, Higashi-Osaka, Osaka 577-8502,Japan Department of Physics, Earth Science and Space Systems Engineering, Morehead StateUniversity, 235 Martindale Drive, Morehead, KY 40351, USA ∗ E-mail: [email protected]
Received ; Accepted
Abstract
We present the results of an X-ray spectral analysis of the northeast region of the candidatesupernova remnant G189.6 + + + Key words:
ISM: individual objects (G189.6 + Introduction
Supernovae (SNe) release tremendous amounts of energy and are the main agents in the Universefor the synthesis of atoms of heavy elements. The supernova remnants (SNRs) help transfer the en-ergy of SNe into the surrounding interstellar medium (ISM) and are the leading candidates for theacceleration of cosmic-ray particles to approximately the ”knee” energy of the cosmic-ray spectrum.The expanding shock wave associated with SNRs sweeps up circumstellar matter and produces anX-ray-emitting plasma comprised of both ISM and stellar ejecta. One quantity that describes theplasma is the electron temperature kT e . This temperature, which corresponds to the temperature ofthe electrons of the plasma, gradually increases over time through Coulomb collisions between theelectrons and more energetic particles in the plasma. Furthermore, these energetic electrons ionizeneutral atoms within the plasma. The temperature of these ionized atoms, denoted as the ionizationtemperature kT i , is another quantity that describes the plasma and typically kT i follows kT e . In gen-eral, the X-ray emitting plasmas associated with SNRs are not in a collisional ionization equilibrium(CIE) state where kT e = kT i but instead in an ionizing plasma (IP) state where kT e > kT i . This isbecause the ionization dominant phase of SNRs typically lasts > yr. Therefore, the X-ray spectraof most young and middle-aged SNRs are well-fitted with an ionizing plasma (IP) model.In the last decade, strong radiative recombination continuua (RRCs) were discovered in theX-ray spectra of several Galactic SNRs, including IC 443 (Yamaguchi et al. 2009) and W49B (Ozawaet al. 2009). Since the RRC originates from radiative transitions where free electrons become boundto ions, the strong RRC is a sign of a recombining plasma (RP), which is characterized by kT e < kT i .Examples of Galactic SNRs that possess recombining plasmas – in addition to the SNRs mentionedabove – are G359.1 − − kT e drops below kT i byconductive cooling by a cold cloud (e.g., Kawasaki et al. 2002; Matsumura et al. 2017). Anotherscenario – known as the rarefaction scenario – proposes that adiabatic cooling occurs when the plasmabreaks out from a dense medium into a much less dense medium (e.g., Masai 1994; Yamaguchi et al.2018). Other proposed scenarios suggest that kT i increases by either photo-ionization by an externalX-ray source (e.g., Nakashima et al. 2013; Ono et al. 2019) or by ionization due to low-energy cosmicrays (LECRs; e.g. Hirayama et al. 2019). The precise origin of RPs associated with SNRs remains2ncertain and realizing a clear understanding of their origin within the evolution of SNR plasmas isan outstanding unresolved issue.The candidate SNR G189.6 + ∼ . ◦
5. The center of this SNR is offset from the center of the nearby prominent SNR IC 443by ∼ . ◦ + + years old and is therefore a well-evolved SNR. Asaoka andAschenbach (1994) argued that G189.6 + + + + The Suzaku observation of the northeast region of G189.6 + ig. 1. X-ray (ROSAT, color) and radio band (1420 MHz, processed by the Canadian Galactic Plane Survey Consortium, green contour) images ofG189.6 + data for each X-ray event were converted to Pulse Invariant (PI) channels using the xispi softwareand the calibration database version 2018-10-10. We screened the data using the standard criteria.During the observations, count rates of the non-X-ray background (NXB) of XIS 1 were systemati-cally higher than those of the NXB data generated by xisnxbgen (Tawa et al. 2008). Therefore, weutilized only the FI in the following analysis. The exposure times after applying the screening criteriaare 66.1 ks and 86.2 ks for XIS0 and XIS3, respectively. Figure 2 shows X-ray images in the ”soft” (0.6–3.0 keV) and ”hard” (3.0–8.0 keV) energy bandsof G189.6 + + ig. 2. XIS images of the northeast region of G189.6 + Based on the work of Masui et al. (2009), we modeled the sky background spectrum, which consistsof the Milky Way Halo (MWH), Local Hot Bubble (LHB), and Cosmic X-ray Background (CXB).The values of the parameters of the MWH and LHB were assumed to be the same as the values usedin Hirayama et al. (2019), while those of the CXB were fixed to the values in Kushino et al. (2002). Itis known that another diffuse source of X-ray emission known as the Galactic diffuse X-ray emission(GDXE) (Koyama 2018) exists along the Galactic plane. However, since G189.6 + b =3 . ◦
3, where the contamination of the GDXE issmall (Uchiyama et al. 2013; Yamauchi et al. 2016), we ignored this component of diffuse emissionin our spectral analysis.The source spectrum of G189.6 + xisnxbgen (Tawa et al. 2008) and was subtracted from the source spec-trum. Figure 3 shows the NXB-subtracted spectrum of G189.6 + + LHB + CXB) are indicated with a solid gray line.To fit the extracted source spectrum, we first applied a CIE model (corresponding to the vapec model in XSPEC) combined with a low-energy absorption model. The cross section of the photoelec-tric absorption and the abundance tables were taken from Balucinska-Church and McCammon (1992)and Anders and Grevesse (1989), respectively. In the spectral fitting, we corrected the energy scaleusing a linear function. Residuals were seen in the spectrum in the 0.6–1.0 keV band: to address5 − . . C oun t s s − k e V − (a) 1 102 5 − − ( da t a − m ode l ) / e rr o r Energy (keV) − . . C oun t s s − k e V − (b) 1 102 5 − − ( da t a − m ode l ) / e rr o r Energy (keV) − . . C oun t s s − k e V − (c) 1 102 5 − − ( da t a − m ode l ) / e rr o r Energy (keV) − . . C oun t s s − k e V − (d) 1 102 5 − − ( da t a − m ode l ) / e rr o r Energy (keV)
Fig. 3.
XIS spectrum of the east region (upper panel) and residuals from the best-fit model (lower panel), (a) model A, (b) model B, (c) model C, and (d) modelD. Errors of the data points are at the 1 σ level. The blue, red, and gray solid lines show emission from ISM, ejecta, and the sky background (MWH + LHB + CXB)components, respectively. these, we added another CIE model with solar abundances to represent emission from a shocked ISMcomponent. We denote this first combined model for fitting as Model A and note that Model A failedto provide a statistically-acceptable fit to the source spectrum ( χ /d.o.f.=224.6/151=1.49). We alsofound systematic residuals around 2.5 keV: the presence of these residuals – combined with positiveresiduals at the energy of the Si Ly α line – suggested the presence of an RRC from He-like Si (seefigure 3a).The implied presence of an RRC motivated us to apply an RP model (corresponding tothe vrnei model in XSPEC) for the ejecta component of the observed X-ray emission. Similarto our results with the CIE model, we again found residuals in the 0.6–1.0 keV energy range,and added the shocked ISM component. We denote this combined CIE + RP model as Model Band note that this combined model also failed to yield a statistically-acceptable fit to the spectrum6 χ /d.o.f.=235.4/150=1.57). Inspection of the fit obtained with this model revealed that the Mg He α line and the RRC structure remain in the fit residuals (see figure 3b). These suggest that kT e of thelighter elements differs from kT e of the heavier elements. As our next step, we applied a two-electrontemperature RP model (denoted as Model C). Specifically, the first RP component (denoted as RP1)consists of the elements H through Si, while the second component (denoted as RP2) consists of Sthrough Ni. While the electron temperatures of RP1 and RP2 were varied as free parameters indepen-dent of each other, other parameters, specifically the recombination timescale, the initial temperature kT , and the normalizations, were tied together. We found at last that Model C gave an acceptable fitto the source spectrum with χ /d.o.f.=151.0/148=1.02 (see table 1 and figure 3c).In the case of Model C, the plasma is assumed to have the same kT i for all the elements (whichis consistent with CIE) at the epoch where the plasma made a transition into an RP. However, we notethat attaining CIE is not an inevitable condition for plasmas associated with SNRs. For example, inanalyzing Suzaku data for W28, Sawada and Koyama (2012) examined the time evolution of RPswith the same kT for all elements (again consistent with CIE) and a different kT for each element,and showed that the spectrum could be fit adequately in both cases. Furthermore, Hirayama et al.(2019) applied a multi- kT RP model, where each element has its own value for kT , to the high-quality spectrum of IC 443, and showed the spectrum is represented by the model. We followedthe examples by these authors and attempted to fit the source spectrum of G189.6 + kT RP model, which we denote as Model D. We obtained a statistically-acceptable fit with thismodel as well ( χ /d.o.f.=146.1/144=1.01). The improvement from the Model C fit is not statisticallysignificant. The best-fitting parameters of this model are listed in table 1 and the model itself is plottedin figure 3d. Our work presented here confirmed that G189.6 + kT e ∼ + + able 1. The best-fit parameters of CIE (Model A), single-RP (Model B), two-temperature RP (Model C), and multi- kT RP (Model D).
Component Parameter ValuesModel A Model B Model C Model DAbsorption N ∗ H ± ± ± ± kT † e ± ± ± ± ‡ ± ± ± ± kT † e ± ± ± ± ‡ ± ± ± ± n e t § — 11 +7 − ± +4 . − . Ab k kT † Ab k kT † Ab k kT † Ab k H=He=C=N=O 1 (fixed) (link to Ne) 1 (fixed) (link to Ne) 1 (fixed) (link to Ne) 1 (fixed)Ne 6.0 ± ± +2 . − . ± +0 . − . +3 . − . Mg 1.2 ± ± ± > +1 . − . Si 0.8 ± ± ± > ± ± ± kT † e — — 0.73 ± ± ‡ — — (link to RP1) (link to RP1) n e t § — — (link to RP1) (link to RP1)H–Si — — — — 0 (fixed) — 0 (fixed)S=Ar=Ca — — — (link to Ne) 1.8 ± +2 . − . ± +3 . − . χ /d.o.f. 224.6/151=1.49 235.4/150=1.57 151.0/148=1.02 146.1/144=1.01 ∗ The unit is × cm − . † Units are keV. kT e is an electron temperature at the present time and kT is an initial ionization temperature at n e t =0. ‡ Defined as 10 − × R n H n e dV / (4 πD ) (cm − ), where D is the distance (cm), n H is the hydrogen density (cm − ), n e is theelectron density (cm − ), and V is the volume (cm ). § Recombination time scale, where n e is the electron density (cm − ) and t is the elapsed time (s). The unit is × cm − s. k Relative to the solar values in Anders and Grevesse (1989).
8e obtained statistically-acceptable fits to the extracted X-ray spectra of G189.6 + kT RP model). In the case of the fitwith Model C, the plasma is assumed to be in the CIE state at the epoch of transitioning into an RP.Note that the fitted initial temperature of the plasma with Model C is ∼ kT e , and fitted these spectra with a multi- kT i model which has a differentvalue for kT i for each element. With this approach, we found that an IP spectrum with kT e ∼ ∼ × cm − s is approximated by the multi- kT i model in which kT e is ∼ kT i of each element is consistent with kT of Model D. This result indicates that theinitial condition described by the fitted parameters of Model D can approximate the non-equilibriumionization state. Furthermore, the kT values of Model D appear to increase with increasing atomicnumber over the range of elements from Ne to Ca. This result parallels those of the different kT caseattained by an IP as shown by Sawada and Koyama (2012). Therefore, an acceptable fit to the spectrausing Model D implies that a scenario where the plasma transitions from an IP to an RP is possible.We note that Model D assumes the ion population of a CIE plasma with kT for each element, whichis actually different from that of a real IP. To investigate this scenario, a plasma model taking intoaccount the evolution from an IP at the initial epoch to an RP is needed.As described previously, different scenarios to produce an RP, such as the thermal conductionscenario, the adiabatic expansion scenario, the scenario of photo-ionization by an external source andthe scenario where ionization is accomplished by LECRs, have been presented in the literature. Asapplied to G189.6 + kT e dropped significantly and therefore the plasma transitioned to an RP from either an IP ora state of CIE. To investigate either of these scenarios further, it is crucial to determine if the ISMthat surrounds G189.6 + kT e We would like to express our thanks to all of the Suzaku team. The authors wish to thank the referee for constructive comments that improved themanuscript. This work was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Numbers JP16J00548 (KKN). References Anders, E., & Grevesse, N. 1989, Geochim. Cosmochim. Acta, 53, 197Asaoka, I., & Aschenbach, B. 1994, A&A, 274, 573Balucinska-Church, M., & McCammon, D. 1992, ApJ, 400, 699Braun, R., & Strom, R. G. 1986, A&A, 164, 193Ferrand, G., & Safi-Harb, S. 2012, Adv. Space Res., 49, 1313 irayama, A., Yamauchi, S., Nobukawa, K. K., Nobukawa, M., & Koyama, K. 2019, PASJ, 71, 37Kamitsukasa, F., Koyama, K., Nakajima, H., Hayashida, K., Mori, K., Katsuda, S., Uchida, H., & Tsunemi, H.2016, PASJ, 68, S7Katsuda, S., et al. 2016, ApJ, 832, 194Kawasaki, M. T., Ozaki, M., Nagase, F., Masai, K., Ishida, M., & Petre, R. 2002, ApJ, 572, 897Koyama, K. 2018, PASJ, 70, 1Koyama, K., et al. 2007, PASJ, 59, S23Kushino, A., Ishisaki, Y., Morita, U., Yamasaki, N. Y., Ishida, M., Ohashi, T., & Ueda, Y. 2002, PASJ, 54, 327Leahy, D.A. 2004, AJ, 127, 2277Masai, K. 1994, ApJ, 437, 770Masui, K., Mitsuda, K., Yamasaki, N. Y., Takei, Y., Kimura, S., Yoshino, T., & McCammon, D. 2009, PASJ,61, S115Matsumura, H., Tanaka, T., Uchida, H., Okon, H., & Tsuru, T. G. 2017, ApJ, 851, 73Mitsuda, K., et al. 2007, PASJ, 59, S1Nakajima, H., et al. 2008, PASJ, 60, S1Nakashima, S., Nobukawa, M., Uchida, H., Tanaka, T., Tsuru, T. G., Koyama, K., Murakami, H., & Uchiyama,H. 2013, ApJ, 773, 20Ohnishi, T., Koyama, K., Tsuru, T. G., Masai, K., Yamaguchi, H., & Ozawa, M. 2011, PASJ, 63, 527Ono, A., Uchiyama, H., Yamauchi, S., Nobukawa, M., Nobukawa, K. K., & Koyama, K. 2019, PASJ, 71, 52Ozawa, M., Koyama, K., Yamaguchi, H., Masai, K., & Tamagawa, T. 2009, ApJ, 706, L71Sawada, M., & Koyama, K., 2012, PASJ, 64, 81Tawa, N., et al. 2008, PASJ, 60, S11Uchida, H., et al. 2012, PASJ, 64, 141Uchiyama, H., et al. 2009, PASJ, 61, S9Uchiyama, H., Nobukawa, M., Tsuru, T. G., & Koyama, K. 2013, PASJ, 65, 19Yamaguchi, H., Ozawa, M., Koyama, K., Masai, K., Hiraga, J. S., Ozaki, M., & Yonetoku, D. 2009, ApJ, 705,L6Yamaguchi, H., et al. 2018, ApJ, 868, L35Yamauchi, S., Nobukawa, M., Koyama, K., & Yonemori, M. 2013, PASJ, 65, 6Yamauchi, S., Nobukawa, K. K., Nobukawa, M., Uchiyama, H., & Koyama, K. 2016, PASJ, 68, 59irayama, A., Yamauchi, S., Nobukawa, K. K., Nobukawa, M., & Koyama, K. 2019, PASJ, 71, 37Kamitsukasa, F., Koyama, K., Nakajima, H., Hayashida, K., Mori, K., Katsuda, S., Uchida, H., & Tsunemi, H.2016, PASJ, 68, S7Katsuda, S., et al. 2016, ApJ, 832, 194Kawasaki, M. T., Ozaki, M., Nagase, F., Masai, K., Ishida, M., & Petre, R. 2002, ApJ, 572, 897Koyama, K. 2018, PASJ, 70, 1Koyama, K., et al. 2007, PASJ, 59, S23Kushino, A., Ishisaki, Y., Morita, U., Yamasaki, N. Y., Ishida, M., Ohashi, T., & Ueda, Y. 2002, PASJ, 54, 327Leahy, D.A. 2004, AJ, 127, 2277Masai, K. 1994, ApJ, 437, 770Masui, K., Mitsuda, K., Yamasaki, N. Y., Takei, Y., Kimura, S., Yoshino, T., & McCammon, D. 2009, PASJ,61, S115Matsumura, H., Tanaka, T., Uchida, H., Okon, H., & Tsuru, T. G. 2017, ApJ, 851, 73Mitsuda, K., et al. 2007, PASJ, 59, S1Nakajima, H., et al. 2008, PASJ, 60, S1Nakashima, S., Nobukawa, M., Uchida, H., Tanaka, T., Tsuru, T. G., Koyama, K., Murakami, H., & Uchiyama,H. 2013, ApJ, 773, 20Ohnishi, T., Koyama, K., Tsuru, T. G., Masai, K., Yamaguchi, H., & Ozawa, M. 2011, PASJ, 63, 527Ono, A., Uchiyama, H., Yamauchi, S., Nobukawa, M., Nobukawa, K. K., & Koyama, K. 2019, PASJ, 71, 52Ozawa, M., Koyama, K., Yamaguchi, H., Masai, K., & Tamagawa, T. 2009, ApJ, 706, L71Sawada, M., & Koyama, K., 2012, PASJ, 64, 81Tawa, N., et al. 2008, PASJ, 60, S11Uchida, H., et al. 2012, PASJ, 64, 141Uchiyama, H., et al. 2009, PASJ, 61, S9Uchiyama, H., Nobukawa, M., Tsuru, T. G., & Koyama, K. 2013, PASJ, 65, 19Yamaguchi, H., Ozawa, M., Koyama, K., Masai, K., Hiraga, J. S., Ozaki, M., & Yonetoku, D. 2009, ApJ, 705,L6Yamaguchi, H., et al. 2018, ApJ, 868, L35Yamauchi, S., Nobukawa, M., Koyama, K., & Yonemori, M. 2013, PASJ, 65, 6Yamauchi, S., Nobukawa, K. K., Nobukawa, M., Uchiyama, H., & Koyama, K. 2016, PASJ, 68, 59