1 Introduction

Physical mechanisms of energy release and nucleosynthesis products ejection in supernova (SN) events are of fundamental importance for astrophysics. Optical and UV studies of the structure of SN remnants (SNRs) have revealed a complex metal ejecta structure with the presence of isolated high-velocity fragments of SN ejecta interacting with surrounding media. The most prominent manifestations of this phenomena are the fast moving knots (FMKs) observed in some young "oxygen-rich'' SNRs in the Galaxy: Cas A (e.g. Chevalier & Kirshner 1979; Fesen et al. 2002), Puppis A (Winkler & Kirshner 1985), G292.0+1.8, in the LMC N132D and 1E 0102.2-7219 in the SMC.

Optical FMKs in Cas A are showing very high abundances in O-burning and Si-group elements with high Doppler velocities. They have a broad velocity distribution around 6000  $\rm ~km~s^{-1}$, while bright radio knots are slower (e.g. Bell 1977). The optically emitting mass of the observed knots is only about 10 $^{-4} \mbox{$M_{\odot}$ }$ (Raymond 1984).

Ejecta-dominated X-ray emission from Cas A was observed with Chandra (Hughes et al. 2000; Hwang et al. 2001) and XMM-Newton (Bleeker et al. 2001; Willingale et al. 2002). Willingale et al. (2002) found an emission component that is probably due to heating of ejecta clumps sweeping up the ambient gas. They also note that the observed Fe-K line emission is confined to two large clumps. Recent Chandra images of G292.0+1.8 (Park et al. 2001) revealed a complex structure with multiple knots and filaments on angular scales down to the instrumental resolution.

X-ray knots were discovered with ROSAT in the Vela SNR by Aschenbach et al. (1995) as "shrapnel'', boomerang structures outside of the main shell. High-resolution Chandra observations of shrapnel A reported by Miyata et al. (2001) revealed a head-tail structure of the apparent size $8\farcm 4 \times 4\farcm 1$ ( $0.6 \times 0.3$ pc at 250 pc distance). They estimated the gas pressure in the head to be roughly ten times higher than that in the tail and $T_{\rm e} \sim$ 0.5 keV. The oxygen abundance was 0.34 ^+0.12_-0.08 of the solar value while that of Si was 3 ⁺²_-1 times of solar.

A localized region of 6.4 keV emission indicating the presence of Fe XVII or lower ionization states was found in the supernova remnant RCW 86 with ASCA instrument by Vink et al. (1997).

High resolution Chandra observation of the Galactic Central region reported by Wang et al. (2002) revealed that the observed He-like Fe K $_{\rm\alpha}$ emission is largely due to the discrete X-ray source population. Apart from the large-scale thin plasma emission in the Galactic Center region Bamba et al. (2002) have also observed with Chandra many small clumps of emission lines from neutral (6.4 keV) to helium-like (6.7 keV), with intermediate line energies between 6.5-6.7 keV. The problem of the origin of the observed large scale X-ray emission from the Galactic ridge requires a careful study of possible classes of abundant hard X-ray sources with $L_{\rm x} \sim 10^{29} \rm ~erg~s^{-1}$(e.g. Tanaka et al. 1999). In this respect modeling of an ensemble of X-ray point sources associated with fast moving SN ejecta fragments seems to be necessary.

In the paper we present a model of X-ray line production in a supersonically moving ejecta fragment. The line emission is due to K-shell ionization by nonthermal particles accelerated by the bow shock and then propagating through a cold metal-rich clump. The optical depth effects are important in the model and they are accounted for. Modern arcsec resolution instruments such as that aboard XMM-Newton and Chandra are sensitive to photons up to 10 keV with a good spectral resolution providing unique possibilities to study X-ray line emission of the ejecta fragments. These observations would allow to study stellar nucleosynthesis.