Near-infrared Spectroscopy of Dense Ejecta Knots in the Outer Eastern Area of the Cassiopeia A Supernova Remnant

Figure 3 shows the locations of the observed knots and line flux ratios of three bright emission lines (i.e., the [S ii] 1.03, [P ii] 1.189, and He i 1.083 μm lines) to the [Fe ii] 1.257 μm line, hereafter R([S ii]/[Fe ii]), R([P ii]/[Fe ii]), and R(He i/[Fe ii]). In the left panels, the line flux ratios are plotted as a function of the knot's radial distance from the explosion center, while in the right panels, the locations of the knots are shown in the (R.A., decl.) plane. In all panels, the color of the symbols represents the flux ratios, i.e., red (blue) means low (high) ratios. The squares and circles indicate the knots in the NE jet and Fe K plume areas, respectively. The red and green dashed lines in the right panels indicate the nominal location of the forward and reverse shocks, respectively. We can see that the NE jet knots are located outside of the forward shock, while most of the knots in the Fe K plume area are located between the forward and reverse shocks, except for one at P.A. = 142°, which is apparently located outside the forward shock.

Zoom In Zoom Out Reset image size

The ratio R([S ii]/[Fe ii]) of the ejecta knots varies over almost 2 orders of magnitude, i.e., 0.2–70 (see top row of Figure 3). A noticeable feature is a group of knots with low R([S ii]/[Fe ii]) at $r\sim 2\buildrel{\,\prime}\over{.} 7$. Most of them have a strong [Fe ii] 1.257 μm line without any detectable [S ii] 1.03 μm lines, so they are marked with upper limits in the figure. They are knots located around the forward shock front just outside the Fe K plume. Their R([S ii]/[Fe ii]) upper limits are more than 1 or 2 orders of magnitude lower than the R([S ii]/[Fe ii]) of the knots around the reverse shock at $r=1\buildrel{\,\prime}\over{.} 4$. There also appears to be a systematic variation in the R([S ii]/[Fe ii]) of the knots at r ≲ 2.3; i.e., the knots at larger radial distances have smaller ratios. The knots at $r=1\buildrel{\,\prime}\over{.} 4$ have ratios of a few tens, while those at $r\sim 2\buildrel{\,\prime}\over{.} 3$ have ratios of ∼10. On the other hand, the NE jet knots, except for one, have R([S ii]/[Fe ii]) > 4, which is comparable with the knots around the reverse shock. There is no systematic variation in the R([S ii]/[Fe ii]) of the NE jet knots.

The ratio R([P ii]/[Fe ii]) shows a similar pattern to the R([S ii]/[Fe ii]); i.e., the knots at $r=2\buildrel{\,\prime}\over{.} 7$ have a very low R([P ii]/[Fe ii]), and the knots at smaller radial distances have higher ratios (see middle row of Figure 3). Again, most of the knots around $r=2\buildrel{\,\prime}\over{.} 7$ have no detectable [P ii] 1.189 μm line. Most of the NE jet knots also have low R([P ii]/[Fe ii]), substantially lower than the knots around the reverse shock. In particular, most of the knots at $r\gtrsim 4^{\prime}$ do not have a detectable [P ii] 1.188 μm line. The 2σ upper limits of these knots are 0.05–0.3, which are a factor of 3–15 smaller than the ratios of the knots around the reverse shock.

The ratio R(He i/[Fe ii]) between the knots differs by a factor of 10³ (see bottom row of Figure 3). For most knots located at $r\lt 2\buildrel{\,\prime}\over{.} 8$, the He i 1.083 μm line is not detected, and the upper limit of R(He i/[Fe ii]) is 0.1–1.0. However, knots of high ratio abruptly appear at $r\sim 3\buildrel{\,\prime}\over{.} 0$, and the ratio reaches up to 100. Several knots in the NE jet area near $r\sim 3\buildrel{\,\prime}\over{.} 0$ also show exceptionally high R(He i/[Fe ii]). Seven knots were identified that exhibit only the He i 1.083 μm line with/without the very weak [C i] 0.985 μm line. These knots will be referred to as "He i knots" for the rest of the paper. These He i knots are detected mainly along the forward shock from the southern base of the NE jet to the Fe K plume area. It is worth mentioning that the radial velocities of most of the He i knots are very large, so they are not circumstellar material (see Section 5.2). In the NE jet, the knots near the tip of the jet have low (≲1.0) R(He i/[Fe ii]).

4.2.1. Knot Classification

In the previous section, we have found that the ejecta knots in different areas show distinct NIR spectroscopic properties. In this section, we classify the knots using three line ratios: R([S ii]/[Fe ii]), R([P ii]/[Fe ii]), and R(He i/[Fe ii]). We use the [S ii] 1.03 μm multiplet line instead of the [S iii] 0.953 μm line. The two lines have a good correlation, and the latter is usually brighter than the [S ii] 1.03 μm multiplet, i.e., F([S iii] 0.953)/F([S ii] 1.03) = 1.5 ± 0.6. But the [S iii] 0.953 μm line is very close to the blue limit of the spectroscope throughput, so the line becomes hardly detectable if it gets significantly blueshifted (e.g., if its radial velocity is less than −2000 km s⁻¹). There are two knots emitting only weak [S iii] 0.953 μm lines, but the expected [S ii] 1.03 μm fluxes of those two knots from F([S iii] 0.953)/F([S ii] 1.03) = 1.5 are smaller than their 3σ rms noise levels, indicating that the nondetection of the [S ii] line is probably due to the low sensitivity of our observations. For these two knots, we use the expected [S ii] 1.03 μm flux for their classification for convenience.

Figure 4 is a diagram of R([P ii]/[Fe ii]) versus R([S ii]/[Fe ii]). The NE jet and Fe K plume area knots are marked by squares and circles, respectively. We first examine the R([S ii]/[Fe ii]) of the knots. The R([S ii]/[Fe ii]) of the knots ranges from 0.2 to 70. The Fe K plume area knots are scattered over the entire range, while the NE jet knots have relatively high ratios, i.e., R([S ii]/[Fe ii]) ≳ 4. Note that most of the Fe K plume area knots with R([S ii]/[Fe ii]) ≲ 4 do not have detectable [S ii] emission. Such separation of the knots in R([S ii]/[Fe ii]) is similar to the properties of the knots in the main ejecta shell. For the knots in the main ejecta shell, Lee et al. (2017) performed principal component analysis of their NIR spectral properties and showed that they could be classified into two groups: (1) S-rich knots showing strong [S ii] and [P ii] lines and (2) Fe-rich knots showing strong [Fe ii] lines. They found that the two groups were well divided by the ratio F([S ii] 1.03)/F([Fe ii] 1.644) of ∼5. Following Lee et al. (2017), we adopt the threshold of R([S ii]/[Fe ii]) ∼ 3.7, assuming an intrinsic ratio of 1.36 for F([Fe ii] 1.257)/F([Fe ii] 1.644) (Nussbaumer & Storey 1988; Deb & Hibbert 2010). We therefore divide the knots into two groups: S-rich knots with R([S ii]/[Fe ii]) ≥ 3.7 and Fe-rich knots with R([S ii]/[Fe ii]) < 3.7. In Figure 4, S- and Fe-rich knots are represented by blue and red symbols, respectively. Note that all Fe-rich knots except one are the knots in the Fe K plume area and that almost of all of them have no detectable [S ii] and [P ii] emission. The Fe-rich knots are mostly located around the outer boundary of the Fe K plume (see Figure 3), and, as we will discuss in Section 5.2, this has an interesting implication for the SN explosion dynamics. Another thing to note is that all NE jet knots except one are S-rich knots and that they have R([P ii]/[Fe ii]) ratios that are generally lower than those of the S-rich knots in the Fe K plume area (see also Figure 3). We will explore the spectral properties of the NE jet knots in detail in Section 5.1.

Zoom In Zoom Out Reset image size

Figure 4 does not include all 67 ejecta knots because some knots do not exhibit [Fe ii] 1.257 μm (and [P ii] 1.189 μm) lines. This may appear strange because our target knots have been selected from the deep [Fe ii] image. But the observation was done with a 10'' slit in ABBA mode, so there could be non-[Fe ii] emission knots accidentally located within the slit. There are 12 knots with no detectable [Fe ii] 1.257 μm lines. Among these, four knots show only [S ii] 1.257 and/or [S iii] 0.953 μm lines, one knot shows only [S ii] 1.257 and He i 1.083 μm lines, and seven knots show only He i 1.083 μm lines. The former five knots belong to S-rich knots because the [Fe ii] 1.257 μm line has not been detected. (Their 2σ lower limits of R([S ii]/[Fe ii]) are ≥3.7.) But R([P ii]/[Fe ii]) is not available for them because both the [Fe ii] 1.257 and [P ii] 1.189 μm lines are not detected. Then there are seven He i knots that show only the He i 1.083 μm line with no [S ii] 1.03, [Fe ii] 1.257, or [P ii] 1.189 μm lines. Their He i 1.083 μm fluxes are comparable to those of the metal-rich ejecta knots, but their [S ii] 1.03, [Fe ii] 1.257, and [P ii] 1.189 μm fluxes are considerably lower than those of the other ejecta knots (Figure 5). We note that these He i knots are mostly an accidental detection and that the category of the He i knots is not very rigorously defined. Indeed, some spectra of S-/Fe-rich knots with high R(He i/[Fe ii]) if degraded could become indistinguishable from the He i knots. However, the number of such knots is less than a few and thus should not affect our discussion of He i knots. From Figure 5, we can also see that S- and Fe-rich knots have a comparable range of [Fe ii] 1.257 μm line strengths. This is not unexpected given that the knots that we observed are the bright ones in the deep [Fe ii] image. The two groups of knots, however, are clearly distinguished in their [S ii] line fluxes. In [P ii] emission, the Fe-rich knots generally have lower 1.189 μm fluxes than the S-rich knots, but not necessarily. A good fraction of the S-rich knots have [P ii] 1.189 μm fluxes as low as the Fe-rich knots. As we will see in Section 5.1, they are the knots near the tip of the NE jet.

Zoom In Zoom Out Reset image size

In summary, we have classified the 67 knots into three groups: S-rich (44), Fe-rich (16), and He i (7) knots. Figure 6 shows which group each individual observed knots belong to. We see that the majority (31/35) of the NE jet knots are S-rich knots. Only one knot is classified as Fe-rich, which has R([S II]/[Fe II])(=1.8) close to the threshold dividing S- and Fe-rich knots. For comparison, the majority (26/32) of the Fe K plume area knots are S- or Fe-rich with clear spatial separation between them; the S-rich knots are near the main ejecta ring within the SNR, while the Fe-rich knots are aligned in the north–south direction along the boundary of the SNR. Figure 6 also shows that those Fe-rich knots appear to be clustered in the position–velocity diagram, implying that they were ejected in a narrow cone. We will explore the kinematic properties of the Fe-rich knots in Section 5.2. The He i knots are detected along the forward shock front, from the southern base of the NE jet to the Fe K plume area. Most of the He i knots have large radial velocities, so they are different from the circumstellar knots that are bright in the He i 1.083 μm line (see Lee et al. 2017). We will explore their origin in Section 5.2. Table 5 summarizes the properties of the three groups, and Figure 7 shows the representative sample spectra of individual groups.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 5. Spectral Classification of the 67 NIR Ejecta Knots and Their Statistics

Knot Group	Spectral Properties	Number of Knots
		NE Jet Area	Fe K Plume Area
(1)	(2)	(3)	(4)
S-rich knots	Strong [S ii], [Fe ii] lines	31	13
	Strong [P ii] line (Fe K plume area); weak/no [P ii] line (NE jet area)
	R([S ii]/[Fe ii]) > 3.7
Fe-rich knots	Strong [Fe ii] line; weak or no [S ii], [P ii] lines	1	15
	R([S ii]/[Fe ii]) < 3.7
He i knots	Strong He i line; weak or no [C i] line	3	4
	R(He i/[Fe ii]) > 8
Total		35	32

Note. (1) Knot group, (2) line ratios, (3) and (4) number of knots in the NE jet and Fe K plume areas.

Download table as:  ASCIITypeset image

4.2.2. Spectral and Physical Properties of Three Knot Groups

In this section, we examine the basic spectral properties of three knot groups and compare the densities of their line-emitting regions, looking for differences other than the bright line ratios.

Figure 8 compares the distributions of the radial velocities and line widths of three knot groups. The radial velocities of the S-rich knots strongly peak near 0 km s⁻¹, which is mostly due to the knots in the NE jet. The radial velocities of the knots in the NE jet are mostly confined to a narrow range of v_r = 0 to +800 km s⁻¹, but there are also knots with high positive and negative velocities; e.g., the radial velocities of the knots in the southernmost stream are v_r = −2100 to −520 km s⁻¹. This is consistent with the previous result from optical studies that the bright knots in the jet streams lie close to the plane of the sky, although the knots in the NE jet region, including those in the jet base area in general, encompass a broad range of radial velocities (Fesen & Gunderson 1996; Fesen 2001; Milisavljevic & Fesen 2013). For comparison, the S- and Fe-rich knots in the Fe K plume area are mostly blueshifted. As already mentioned in Section 4.2.1, the Fe-rich knots are mostly located in the Fe K plume area, and their radial velocities are confined to a narrow range (−2000 to −200 km s⁻¹). On the other hand, the radial velocities of the He i knots are spread over a broad velocity range, from −3500 to +3500 km s⁻¹. In line width, the S-rich knots have a broad distribution centered at around +290 km s⁻¹, while the Fe- and He-rich knots have a relatively narrow distribution centered at approximately +250 km s⁻¹. The velocity widths may be considered as a characteristic shock speed for the knots. But as we will see in Section 4.3, the ejecta knots appear to have complex structures with subknots, so the shock speed could be substantially lower than this. Figure 8 also compares the distribution of [Fe ii] 1.257 μm line fluxes among the three knot groups. As we mentioned in Section 4.2.1, the S- and Fe-rich knots have a similar distribution of [Fe ii] 1.257 μm line fluxes, while the He i knots have the [Fe ii] 1.257 μm line undetected with an upper limit of mostly a few times 10⁻¹⁶ erg s⁻¹ cm² (see Figure 3). Note that the [Fe ii] 1.257 μm line flux in the table is the flux within the slit. The fluxes of the knots in the deep [Fe ii] image may be found in Koo et al. (2018).

Zoom In Zoom Out Reset image size

One of the physical parameters of the knots that can be easily obtained is electron density. In the J and H bands, there are several [Fe ii] lines originating from levels with similar excitation energies whose ratios depend mainly on electron densities and can be used as a density tracer (e.g., see Koo et al.2016; Lee et al. 2017). In Figure 9, we plot the F([Fe ii] 1.295)/F([Fe ii] 1.257) of the knots, which is sensitive to electron density between 10³ and 10⁵ cm⁻³. The x-axes of the panels in the figure are the [S ii] 1.03, [P ii] 1.189, and He i 1.083 μm fluxes normalized by the [Fe ii] 1.257 μm flux. For about two-thirds of the knots, the [Fe ii] 1.295 μm line has not been detected, so their upper limits are plotted. For the knots with the [Fe ii] 1.295 μm line detected, the electron densities inferred from their [Fe ii] line ratios are mostly in the range of (1–4) × 10⁴ cm⁻³, with a mean of 1.8 × 10⁴ cm⁻³. The S-rich knots have a slightly lower density than the Fe-rich knots, i.e., 1.5 × 10⁴ versus 2.5 × 10⁴ cm⁻³. Note that this is the density of the shock-compressed, line-emitting region of the knots. The initial density of a knot before being shocked should be much lower (e.g., ∼10² cm⁻³; see Figure S3 of Koo et al. 2013). There is no clear association between the line ratios of the bright lines and F([Fe ii] 1.295)/F([Fe ii] 1.257), except for a moderate positive correlation (ρ = +0.33) observed for R(He i/[Fe ii]), which could possibly be because the He i 1.083 μm line emissivity increases with density due to the contribution from the collisional excitation from the lower level (2³S; e.g., see Koo et al. 2023). For a given F([Fe ii] 1.295)/F([Fe ii] 1.257) ratio, the R([S ii]/[Fe ii]) of S- and Fe-rich knots differs by more than an order of magnitude. This suggests that the density of the shocked gas is not the primary factor causing the division between the two groups of ejecta knots.

Zoom In Zoom Out Reset image size

Numerous dense ejecta knots have been identified around the Cas A SNR in previous optical studies, and it will be interesting to check if our NIR ejecta knots have optical counterparts. We use the optical ejecta knot catalog of Hammell & Fesen (2008), who identified 1825 compact optical knots that lie beyond a radial distance of $\sim 2^{\prime}$ from the explosion center in Hubble ACS/WFC images taken with three different broadband filters (i.e., F652W, F775W, and F850LP) in 2004. They were able to classify the optical knots into three groups based on their flux ratios in the three filters: (1) [N ii] knots dominated by [N ii] λ λ6548, 6583 emissions, (2) [O ii] knots dominated by [O ii] λ λ7319, 7330 emissions, and (3) fast-moving knot (FMK)–like knots displaying filter flux ratios suggestive of [S ii], [O ii], and [Ar iii] λ7135 emission line strengths similar to the FMKs found in the remnant's main ejecta shell (see also Fesen et al. 2006). Of the 1825 knots identified by them, 444 were [N ii] knots, 192 were [O ii] knots, and 1189 were FMK-like knots. Their spatial distributions are distinct; [N ii] knots are arranged in a broad shell around the remnant, [O ii] knots are clustered around the base of jets, and FMK-like knots are mainly confined to NE and SW jet areas (see also Fesen & Milisavljevic 2016). We have compared the spatial positions of our 67 NIR ejecta knots with those of the optical outer knots shifted to the 2017 epoch assuming free expansion. Considering the seeing and slit width of our observations, we have searched their optical counterparts within a 05 radius. Note that 12 out of the 67 ejecta knots are located inside $r=2\buildrel{\,\prime}\over{.} 4$, where Hammell & Fesen (2008) did not search optical knots, so we can check the counterparts for only 55 NIR knots.

We found that 38 out of the 55 NIR knots have optical counterparts. All of them have FMK-like knots as a counterpart, except for one that appears to have a [N ii] knot as a counterpart. No NIR knots have [O ii] knots as a counterpart. It is not unusual that the NIR knots that are identified from the ground-based, deep [Fe ii] image are resolved into multiple knots in high-resolution HST images, implying that the knots can have a complicated structure consisting of subknots. Figure 10 shows the optical counterparts of the sample of the observed NIR knots. Some NIR knots appear as single, very compact optical knots, but they often have a more complex structure that is resolved into smaller subcomponents. In some cases, an NIR knot is found to be a part of a larger, more extended structure in the optical. We further note that optical knots tend to cluster together with other knots of the same type. However, there are rare cases where different types of optical knots are found in close proximity to each other, especially in the main shell region. While these can complicate the identification of optical counterparts, we do not think the overall trend will be affected. Figure 11 shows the pie charts for the optical counterparts of the three groups of NIR knots. About three-fourths of the S- and Fe-rich knots have optical counterparts, while about half of the He i knots have optical counterparts. It is worth noticing that one-fourth of the S-rich knots have no optical counterpart, although, since the F850LP filter is sensitive to the [S ii] 1.03 and [S iii] 0.907, 0.953 μm lines, the S-rich knots emitting strong [S ii] and [S iii] lines should at least be detected in the deep F850LP image. The S-rich knots not detected in the deep optical image may have appeared after 2004, when the HST images were taken. It has been found that the knots in the NE jet and Fe K plume areas reveal substantial emission variations ("flickering") in an interval as short as 9 months (Fesen et al. 2011; see also Figure 10). Indeed, we were able to confirm that several knots (e.g., knots 2, 7, and 15) were not present in the 2004 HST image but appeared in a subsequent 2010 HST image.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

We also found that 70% of the Fe-rich knots have FMK-like knots as their optical counterparts. This is inconsistent with our results that most Fe-rich knots do not have [S ii] 1.03 and [S iii] 0.953 μm lines. One possibility is that the flux detected in the F850LP image is not due to [S ii] or [S iii] lines but rather [Fe ii] lines. There are several bright [Fe ii] lines, e.g., 8619, 8894, and 9229 Å (Dennefeld 1986; Hurford & Fesen 1996; Koo et al. 2016), in the wavelength band (8000–10000 Å) of the F850LP filter, while there are few or no bright [Fe ii] lines in the wave bands of the F652W and F775W filters. For example, the expected [Fe ii] 8619/1.257 μm ratio ranges from 0.08 to 0.49 for electron densities of 10³–10⁵ cm⁻³ (Koo et al. 2016). Hence, the [Fe ii] lines can make a substantial contribution to the F850LP filter flux, and, together with the high interstellar extinction toward Cas A, the Fe-rich knots in our sample could have been classified as FMK-like knots in the optical band. Note that the majority of the optical knots in the catalog of Hammell & Fesen (2008) are very faint, with an F850LP flux of less than 5 × 10⁻¹⁷ erg cm⁻² s⁻¹, which is more than 2 orders of magnitude smaller than the [Fe ii] 1.257 μm fluxes of Fe-rich knots (Figure 5). Another possibility is that the knots have [S ii] and/or [S iii] lines but that the lines are "weak," i.e., R([S ii]/[Fe ii]) < 3.7, so they have been classified as Fe-rich knots in this work, but they were detected in the deep F850LP image and classified as FMK-like knots. Indeed, at least three of the Fe-rich knots with FMK-like knot counterparts do have [S ii] lines (see Table 6 in Section 5.2). Yet another possibility is a chance coincidence. The area where the Fe-rich knots have been detected is crowded with FMK-like knots (e.g., see Figure 1 of Hammell & Fesen 2008), so we cannot rule out the possibility that an FMK-like knot accidentally falls within the circle of 05 radius. But, since the distances between the matched knots are usually an order of magnitude smaller than the median distance between the optical knots around Fe-rich knots (i.e., a few arcseconds), most of the matches are probably genuine.

Table 6. Properties of Fe-rich Knots

Knot	K2019 No.	α (J2000)	δ (J2000)	Size	μ	F[Fe II]1.644 μm	vr	$\displaystyle \frac{F([{\rm{S}}\,{\rm{II}}])}{F([\mathrm{Fe}\,\mathrm{II}])}$	$\displaystyle \frac{F([{\rm{P}}\,{\rm{II}}])}{F([\mathrm{Fe}\,\mathrm{II}])}$	$\displaystyle \frac{F(\mathrm{He}\,{\rm{I}})}{F([\mathrm{Fe}\,\mathrm{II}])}$	Optical
No.				(arcsec)	(km s−1)	(erg cm−2 s−1)	(km s−1)				Counterpart
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
23	197	23:23:57.399	58:50:30.36	2.26 × 1.44	⋯	5.75E-16	−234	1.78	≤0.13	≤0.26	FMK
39	235	23:23:49.208	58:49:04.65	1.85 × 1.48	⋯	4.66E-16	−1286	≤1.72	≤0.10	0.71	FMK
40	234	23:23:49.044	58:49:03.51	1.69 × 1.44	⋯	5.17E-16	−1282	≤0.40	≤0.04	0.59	FMK
42	232	23:23:48.305	58:48:59.55	1.85 × 1.31	7291	4.81E-16	−1616	≤0.48	≤0.06	0.24	FMK
43	⋯	23:23:48.336	58:48:57.60	0.77 × 0.77	⋯	9.30E-17	−1987	≤1.13	≤0.20	≤0.60	⋯
45	⋯	23:23:51.235	58:49:51.88	1.47 × 1.39	⋯	3.82E-16	−1165	≤0.62	≤0.09	0.45	⋯
46	241	23:23:48.617	58:48:54.88	3.29 × 2.32	7221	2.98E-15	−1189	≤0.26	≤0.02	0.34	FMK
47	239	23:23:48.106	58:48:53.21	3.41 × 2.61	6989	2.89E-15	−1159	≤0.24	≤0.02	0.38	FMK
48	240	23:23:48.253	58:48:43.33	1.26 × 0.84	⋯	1.90E-16	−903	≤1.21	≤0.20	≤0.31	FMK
49	251	23:23:49.109	58:48:32.44	1.21 × 0.82	⋯	1.63E-16	−1554	3.13	0.35	0.86	FMK
51	250	23:23:48.909	58:48:28.59	2.26 × 1.76	⋯	9.13E-16	−207	≤0.39	≤0.13	8.02	⋯
53	259	23:23:47.850	58:48:17.75	5.01 × 2.89	7640	3.36E-15	−644	1.07	0.11	0.30	FMK
56	272	23:23:46.136	58:48:04.65	1.46 × 1.22	⋯	3.66E-16	−1413	≤0.85	≤0.09	≤0.16	⋯
63	290	23:23:44.342	58:47:32.27	1.22 × 0.89	⋯	1.87E-16	−2023	≤3.43	≤0.37	≤0.74	⋯
64	288	23:23:43.483	58:47:33.93	5.86 × 1.46	6580	1.10E-15	−1725	≤1.82	≤0.24	≤1.00	FMK
67	299	23:23:42.554	58:46:21.52	2.54 × 2.09	⋯	1.81E-15	1957	≤0.18	≤0.03	2.25	FMK

Note. (1) Knot serial number in this work, (2) knot serial number in Koo et al. (2018), (3) and (4) central coordinates, (5) major and minor axes, (6) velocity of proper motion, (7) observed [Fe ii] 1.644 μm flux, (8) radial velocity, (9)–(11) line ratios, (12) optical counterpart (see Section 4.3). Columns (3)–(7) are from Koo et al. (2018).

Download table as:  ASCIITypeset image

Lastly, it is interesting that about half of the He i knots have FMK-like knots as their optical counterparts. This may indicate that He i knots are just S-rich knots with [S ii], [S iii], and [Fe ii] lines below our detection limit. But, since most of the He i knots are also located in the area crowded with FMK-like knots, a chance coincidence is possible. We will explore the nature of He i knots in Section 5.3.

A most striking morphological feature in Cas A known from early optical studies is the NE jet structure (van den Bergh & Dodd 1970; Fesen & Gunderson 1996). It is composed of the streams of bright knots confined into a narrow cone well outside the northern SN shell that appears to have punched through the main ejecta shell. The jet structure extends to $\sim 3^{\prime}$ beyond the main shell, implying ejection velocities of up to 16,000 km s⁻¹, which is three times faster than the bulk of the O- and S-rich ejecta in the main ejecta shell (Kamper & van den Bergh 1976; Fesen & Gunderson 1996; Fesen 2001; Fesen & Milisavljevic 2016). Later, the SW "counterjet" composed of numerous optical knots with comparable maximum expansion velocities was detected on the opposite side (Fesen 2001; Milisavljevic & Fesen 2013). The origin of the bipolar NE jet–SW counterjet structure is not fully understood, although it must have been generated very near the explosion center during the first seconds of the explosion (see below). Critical information about the origin of the jet comes from its chemical composition. Previous optical studies showed that the NE jet knots exhibit strong S, Ca, and Ar emission lines with comparatively weak oxygen lines (van den Bergh 1971; Fesen & Gunderson 1996; Fesen 2001; Hammell & Fesen 2008; Milisavljevic & Fesen 2013; Fesen & Milisavljevic 2016). The knots lying farther out and possessing the highest expansion velocities show no detectable emission lines other than those of [S ii] (Fesen & Milisavljevic 2016). The knots near the jet base, on the other hand, show O and N emission too (Fesen & Gunderson 1996). The jet structure has been observed in the X-ray and IR too. In the X-ray, the jet material shows strong Si, S, Ar, and Ca lines but not a particularly strong Fe K line (Hughes et al. 2000; Hwang et al. 2004; Vink 2004). The Fe K emission has been found to be strong in other areas of Cas A, particularly the eastern area (Hwang & Laming 2003, 2012; Picquenot et al. 2021; see Section 5.2). These X-ray studies showed that the elemental composition of the jet, enriched in the intermediate-mass elements with a limited amount of Fe, is similar to that of oxygen or incomplete Si burning (see also Ikeda et al. 2022). In the IR, the ringlike jet base lifted from the main ejecta shell is pronounced in Ar emission (DeLaney et al. 2010). These findings suggest that the jet originated in the deep interior of the progenitor star where Si, S, Ar, and Ca are nucleosynthesized and penetrated through the outer stellar envelope.

Figure 12 gives a detailed view of the NE jet structure in [Fe ii] 1.644 μm emission. In the image, the NE jet appears as streams of bright knots confined to a fanlike structure with an opening angle of 10° viewed at the explosion center. The central stream at P.A. ≃ 64° is most prominent, extending to $4\buildrel{\,\prime}\over{.} 7$(=4.6 pc) from the explosion center. In the northern area above the central stream, there are many bright knots scattered over the area that appear to be aligned along several stream lines, while, in the southern area below the central stream, there is a prominent but fuzzy stream of ejecta material possessing a finite width at P.A. ≃ 70°, and not many bright knots are in between. This jet morphology is very similar to what we have already seen in [S ii] λ λ6716, 6731 emission (Fesen & Gunderson 1996).

Zoom In Zoom Out Reset image size

In Figure 12, we mark the ejecta knots observed in this work using different colors for S-rich, Fe-rich, and He i knots. We see that the NE jet knots are mostly S-rich (see also Figure 6). We have not detected an Fe-rich knot exhibiting only [Fe ii] lines. There is one knot classified as an Fe-rich knot located between the central and southern streams, but it exhibits [S ii] lines of moderate strength (R([S ii]/[Fe ii]) = 1.8). These results are consistent with the previous results from optical and X-ray studies that the jet material shows strong S lines but not particularly strong Fe lines. In the optical bands, the NE jet knots show strong O, S, and Ar lines like the other S-rich knots and are not clearly distinguished from the S-rich knots (FMKs) in the main ejecta shell, although, for example, a few outermost knots show stronger than usual [Ca ii] λ λ7291, 7324 lines that could be due to abundance differences (Fesen & Gunderson 1996). In the NIR, however, the NE jet knots are distinguished from the S-rich knots in the main ejecta shell in [P ii] emission. The majority of the dense knots in the NE jet show weak or no [P ii] lines, whereas the S-rich knots in the main ejecta shell show strong [P ii] lines. This is shown in Figure 13, where the line ratios of the NE jet knots are plotted as a function of distance from the explosion center (see also Figure 3). The figure shows that R([S ii]/[Fe ii]) remains roughly constant along the jet, whereas R([P ii]/[Fe ii]) and R([P ii]/[S ii]) appear to decrease systematically along the jet. Note that the [P ii] line has not been detected in the majority of the knots near the tip of the jet. For comparison, the three ratios for the S-rich knots in the main ejecta shell are (R([S ii]/[Fe ii]), R([P ii]/[Fe ii]), R([P ii]/[S ii])) = (14.2 ± 13.5, 1.46 ± 1.31, 0.117 ± 0.083) (see Figure 2 of Koo et al. 2013).

Zoom In Zoom Out Reset image size

The simplest interpretation of the low [P ii]/[Fe ii] and [P ii]/[S ii] ratios of the NE jet knots is that the amount of phosphorus present there is lower compared to the knots in the main shell. Although the knots in the NE jet and the main shell have different shock environments, the decrease in these ratios from the main shell to the tip of the NE jet while the [S ii]/[Fe ii] ratio remains constant suggests that the phosphorus abundance is probably the primary factor affecting the line ratios. Therefore, we consider the weakness of the [P ii] line in the NE jet knots as direct evidence that the NE jet originated from a layer with a high sulfur and low phosphorus content. And, since ¹⁵P is mainly produced in carbon- and neon-burning layers (Arnett 1996; Woosley et al. 2002), it implies that the NE jet was ejected from a region below the explosive Ne-burning layer, i.e., the O-burning or incomplete Si-burning layers. This is consistent with the result of previous X-ray studies, which showed that the elemental composition of diffuse ejecta material in the NE jet is similar to that of oxygen and incomplete Si burning (Vink 2004; Ikeda et al. 2022).

We now discuss the implication of our results on the physical origin of the NE jet. The origin of the NE jet and its role in the SN explosion have been subject to study since its discovery in the 1960s. Minkowski (1968) suggested that the jet structure (or the "flare" in his paper) could be a surviving part of a spherically symmetric shell due to nonspherical distribution of the ambient ISM (see also Fesen & Gunderson 1996). But it is now rather well established that the NE jet originated from a Si–S–Ar-rich layer deep inside the progenitor and penetrated through the outer N- and He-rich envelope (Fesen 2001; Fesen et al. 2006; Laming et al. 2006; Milisavljevic & Fesen 2013). The chemical composition of the jet material and the spatial correlation between the jet and the disrupted main ejecta shell are strong evidence for that. On its launching mechanism, there have been several theoretical propositions in relation to the core-collapse SN explosion models. In the classical "jet-induced" explosion model, where the magnetic field anchored to the collapsing core is amplified during the collapse and drives the SN explosion, a magnetocentrifugal jet can be ejected along the rotation axis (Khokhlov et al. 1999; Wheeler et al. 2002; Akiyama et al. 2003). Recent observational studies, however, suggest that the properties of the NE jet are not consistent with the jet-induced explosion model; there is no "pure" Fe ejecta in the jet (Ikeda et al. 2022), the opening angle is large, the kinetic energy is insufficient, and the kick direction of the neutron star kick is not along but perpendicular to the jet axis (Fesen 2001; Hwang et al. 2004; Laming et al. 2006; Milisavljevic & Fesen 2013; Fesen & Milisavljevic 2016; see Soker 2022). Our results are also in line with previous results and do not support the jet scenario. We also have not found dense Fe-rich knots around the tip of the NE jet in this study. The opening angle of the jet in Figure 12 is small (≲15°), but the total kinetic energy will be less than the previous estimate of 1 × 10⁵⁰ erg of the optical knots in the NE jet (Fesen & Milisavljevic 2016). Hence, our results also do not support the jet-induced explosion model.

In the neutrino-driven explosion model, which is the modern paradigm of the core-collapse SN explosion, narrow high-velocity ejecta can be generated either during or after the explosion by the newly formed neutron star. During the explosion, high-speed jetlike fingers can be produced from the core by neutrino convection bubbles (e.g., Burrows et al. 1995; Janka & Mueller 1996). In particular, the interface between Si and O composition is unstable, seeded by the flow structures resulting from neutrino-driven convection, leading to a fragmentation of this shell into metal-rich clumps (Kifonidis et al. 2003). Orlando et al. (2016) showed that Si-rich, jetlike features such as the NE jet in Cas A could be reproduced by introducing overdense knots moving faster than the surrounding ejecta just outside the Fe core. But a recent 3D simulation for modeling the Cas A SNR as a remnant of a neutrino-driven SN, going from the core collapse to the fully developed remnant, could not reproduce a structure similar to the NE jet (Orlando et al. 2021). Another possibility is a hydrodynamic jet from a newly formed neutron star. In a standard neutrino-driven explosion, a fraction of ejecta falls back to the newly formed neutron star to form a disk where an MHD jet can be launched (Burrows et al. 2005, 2007; Burrows & Vartanyan 2021; Janka et al. 2022). But as far as we are aware, there are few theoretical predictions that can be compared with observations for such an MHD jet. So, for the neutrino-driven explosion model, it is still an open question whether the NE jet can be explained by this model.

We discovered Fe-rich knots in the Fe K plume area (see Figure 14). This is the area where the Fe-rich plume with bright Fe K emission has been detected in X-rays (Willingale et al. 2002; Hwang & Laming 2003; Hwang et al. 2004; Hwang & Laming 2012; Tsuchioka et al. 2022). The ejecta material in this Fe K plume has no S or Ar, and its Fe/Si abundance ratio is up to an order of magnitude higher than expected in the incomplete Si-burning layer, indicating that it originated from a complete Si-burning process with α-rich freeze-out, where the burning products are almost exclusively ⁵⁶Fe because free α particles are unable to reassemble to heavier elements on a hydrodynamic timescale (Arnett 1996; Woosley et al. 2002). Recently, Sato et al. (2021) detected Ti and Cr in this area, the ratio of which to Fe is consistent with explosive complete Si burning with α-rich freeze-out. Therefore, the material in the Fe K plume has likely been produced in the innermost, high-entropy region during the SN explosion, where the temperature is high and the density is low (e.g., Arnett 1996; Thielemann et al. 1996). In 2D projected maps, however, the Fe K plume is located much further outside the main shell, which is composed of ejecta synthesized from incomplete Si burning, e.g., Si, S, and Ar, which could be interpreted as an overturning of the ejecta layers (Hughes et al. 2000). But a 3D reconstruction of the ejecta distribution using infrared and Chandra X-ray Doppler velocity measurements showed that, in the main shell, the Fe K plume occupies a "hole" surrounded by a ring structure composed of Si and Ar (DeLaney et al. 2010; see also Milisavljevic & Fesen 2013). As an explanation of the observed morphology, DeLaney et al. (2010) proposed a model where an ejecta "piston" faster than the average ejecta has pushed through the Si layers, rather than interpreting it as an overturning of the layers. In this model, the Si–Ar ring structure in the main shell represents the circumference of the piston intersecting with the reverse shock. Recently, Tsuchioka et al. (2022) showed that the velocity of the ejecta in the outermost Fe K plume is >4500 km s⁻¹, much higher than that of Si-/O-rich ejecta, suggesting that the ejecta piston was formed at the early stages of the remnant evolution or during the SN explosion, which is consistent with the results of numerical simulations. On the other hand, recent 3D numerical simulations demonstrated that Ni clumps, created during the initial stages of the SN explosion in the innermost region by the convective overturning in the neutrino-heating layer and the hydrodynamic instabilities, later decay to ⁵⁶Fe and can push out the less dense ejecta to produce structures such as the Fe K plume (Orlando et al. 2016, 2021). But Hwang & Laming (2003) pointed out that the morphology and the "ionization age" (n_e t, where n_e is the electron density, and t is the time since the plasma was shock-heated) of the Fe ejecta in the Fe K plume are inconsistent with a Ni bubble origin.

Zoom In Zoom Out Reset image size

Our detection of Fe-rich knots in the Fe K plume area has quite significant implications. Figure 14 shows the distribution of IR ejecta knots in the Fe K plume area. We see a group of Fe-rich knots aligned in the north–south direction along the forward shock front. They are spatially confined in the area immediately outside the boundary of the diffuse X-ray Fe ejecta. Their radial velocities are in a narrow range, from −2000 to −200 km s⁻¹ (see Figure 8 and Table 6), so, in a position–velocity diagram, they appear to be clustered (see Figure 6(b)). The ejection angle of this Fe-rich ejecta knot cluster is 14° ± 5° from the plane of the sky toward us. (In Figure 6(b), we see two Fe-rich knots located far off the cluster. One of them is knot 23 in the NE jet area, and the other is knot 67 near the southern edge of the field. The latter has a radial velocity of +1960 km s⁻¹, which is very different from that of the Fe-rich ejecta knot cluster. They are not considered to belong to the cluster.) The radial velocities of the X-ray-emitting diffuse ejecta in this area measured with the Chandra High Energy Transmission Grating are from −2600 to −500 km s⁻¹, comparable with those of the Fe-rich knots (Lazendic et al. 2006; Rutherford et al. 2013). On the other hand, the proper motion of the Fe-rich knots has a velocity of 6600–7600 km s⁻¹ (Table 6), which is considerably higher than that of the diffuse ejecta at the tip of the Fe K plume (i.e., 4500–6700 km s⁻¹; Tsuchioka et al. 2022). Hence, the Fe-rich knots move a little faster than the diffuse X-ray-emitting Fe-rich ejecta. The spatial and kinematic relations strongly support the physical association between the dense Fe-rich knots and the X-ray-diffuse Fe ejecta. And the spectral properties of the Fe-rich knots suggest that they are also produced by explosive Si complete burning with α freeze-out as the diffuse Fe ejecta. The He i 1.083 μm lines detected in the majority of Fe-rich knots might be from the helium remains of the α-rich freeze-out process (Table 6). In α-rich freeze-out, the burning products are almost exclusively ⁵⁶Fe, but a considerable amount of He could remain. The local He mass fraction can be as high as 0.2, depending on the model (Rauscher et al. 2002; Woosley et al. 2002; Thielemann et al. 2018). The detection of dense Fe ejecta knots implies that the Fe ejecta produced in the innermost region was very inhomogeneous.

Theoretically, dense metallic ejecta knots can be produced during the explosion. As mentioned above, numerical simulations have shown that when the innermost ejecta expanding aspherically encounters a composition shell interface, it is compressed to a dense shell, which fragments into small, dense metallic clumps by hydrodynamic instability (Kifonidis et al. 2000, 2003; Hammer et al. 2010; Wongwathanarat et al. 2017). It is uncertain whether the dense knots had initially higher velocities. However, it is possible that these dense knots were initially comoving with the surrounding diffuse ejecta, but once they encounter the reverse shock, the dense and diffuse ejecta are decoupled, i.e., the diffuse Fe ejecta are decelerated more, so the dense Fe clumps are located ahead of the diffuse Fe ejecta. This explanation has been proposed for dense Fe knots beyond the SNR boundaries detected in X-rays in young Type Ia SNRs (Wang & Chevalier 2001; Tsebrenko & Soker 2015). A caveat in this scenario is that, since the radioactive decay from ⁵⁶Ni to ⁵⁶Fe produces postexplosion energy, the Fe clumps are expected to expand unless they are optically thin to γ-ray emission and should be characterized by the diffuse morphology in the remnant phase (Blondin et al. 2001; Hwang & Laming 2003; Gabler et al. 2021). Indeed, in the interior of Cas A, the bubble-like structures in S-rich ejecta filled with ⁴⁴Ti are detected, supporting this scenario (Milisavljevic & Fesen 2013; Grefenstette et al. 2017). Therefore, the dense compact Fe-rich knots discovered in Cas A need an explanation. A possible explanation is that the dense Fe-rich knots are not ⁵⁶Fe decayed from ⁵⁶Ni but ⁵⁴Fe (Wang & Chevalier 2001). But ⁵⁴Fe is not a dominant element (see Hwang & Laming 2003, and below). In core-collapse SNe particularly, ⁵⁴Fe turns into ⁵⁸Ni (stable nuclei of nickel) in the α-rich freeze-out region of the complete Si-burning zone, so it is only detected in the incomplete Si-burning layer, where the S and Si abundances are comparable with or much larger than the ⁵⁴Fe and ⁵⁶Ni/⁵⁶Fe abundances (e.g., Thielemann et al. 1996, 2018). Therefore, the Fe in our Fe-rich knots is most likely ⁵⁶Fe, not ⁵⁴Fe. Another possibility, which was proposed by Hwang & Laming (2003) to explain the compact X-ray knots embedded in the Fe K plume, is that the optical depth of the knots is not large enough for the γ-rays from the radioactive decay from ⁵⁶Ni to ⁵⁶Fe to escape. Using a hydrodynamics model for ejecta evolution, they showed that, for a typical knot of diameter ∼3'' and electron density ∼10 cm⁻³, the optical depth of the γ-rays (∼1 MeV) becomes less than 1 about 5 days after explosion, which is much shorter than the ⁵⁶Ni and ⁵⁶Co lifetimes, i.e., 8.8 and 111.3 days, respectively. We consider that the Fe-rich knots discovered in our NIR study could be a denser and faster version of the X-ray knots. The NIR Fe-rich knots are dense (e.g., ≳10² cm⁻³; Section 4.2.2), and their apparent sizes measured with an ∼1'' resolution are comparable to the X-ray knots (Table 6). So, if we naively accept these numbers, the NIR Fe-rich knots are expected to be optically thick for γ-rays generated from the radioactive decay from ⁵⁶Ni to ⁵⁶Fe. However, the structure of the NIR Fe-rich knots is not resolved. It is possible that they consist of multiple compact subknots with relatively low column densities, enabling the γ-rays to escape primarily from the knot. The HST images in Figure 10 show that indeed, the intrinsic sizes of the NIR knots could be much smaller than their apparent sizes in the deep [Fe ii] image. Furthermore, the ejecta clumps are destroyed by the passage of the reverse shock (e.g., see Figure 4 of Slavin et al. 2020), so the initial sizes of the NIR and X-ray knots could have been considerably smaller. Hence, the Ni bubble effect may not be significant for either type of Fe-rich knot. We will conduct subsequent studies in a forthcoming paper.

Before leaving this section, it is worth noticing that our observation has been done toward bright knots in the deep [Fe ii] image obtained using a narrowband filter with a bandwidth of ±2800 km s⁻¹ (see Section 2), so there is a possibility that there are additional Fe-rich knots outside this radial velocity range. However, previous studies have shown that the diffuse Fe ejecta typically have radial velocities in the range of v_r = −3000 to −1000 km s⁻¹ (DeLaney et al. 2010; Tsuchioka et al. 2022). Therefore, it is likely that the majority of the bright Fe-rich ejecta knots are included in the [Fe ii] image of Figure 14, and our result provides insight into the distribution of the Fe-rich ejecta knots.

We detected seven knots that show only He i 1.083 μm lines. The He i 1.083 μm line intensities of these He i knots are not particularly strong in comparison with other metal-rich ejecta knots (Figure 5). But, since the [Fe ii] 1.257 μm line is not detected, they are noticeable by their very high R(He i/[Fe ii]), which is 1–2 orders of magnitude higher than the typical S- and Fe-rich ejecta knots (see the bottom panel of Figure 4). Table 7 lists the knots with R(He i/[Fe ii]) greater than 8, where we can see that the majority are He i knots.

Table 7. Ejecta Knots with High R(He i/[Fe ii])

	Knot	vr	FHe I 1.083 μm	$\displaystyle \frac{F([\mathrm{S\; II}])}{F([\mathrm{Fe}\,\mathrm{II}])}$	$\displaystyle \frac{F(\mathrm{He}\,{\rm{I}})}{F([\mathrm{Fe}\,\mathrm{II}])}$	$\displaystyle \frac{F([{\rm{S}}\,\mathrm{II}])}{F(\mathrm{He}\,{\rm{I}})}$	$\displaystyle \frac{F([{\rm{C}}\,{\rm{I}}])}{F(\mathrm{He}\,{\rm{I}})}$	Knot	Optical
	No.	(km s−1)	(erg cm−2 s−1)					Type	Counterpart
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
He I knots	26	3241	3.89E-15	⋯	≥8.21	≤0.55	≤0.25	He	FMK
	33	−3152	6.43E-15	⋯	≥27.40	≤0.55	0.31	He	FMK
	35	2314	2.22E-15	⋯	≥14.50	≤0.39	≤0.57	He	⋯
	38	−3312	1.34E-15	⋯	≥14.70	≤0.47	≤0.23	He	FMK
	41	−268	1.28E-15	⋯	≥22.09	≤0.35	≤0.24	He	⋯
	52	765	8.87E-15	⋯	≥83.32	≤0.08	0.11	He	⋯
	54	1186	1.05E-14	⋯	≥77.55	≤0.07	0.07	He	⋯
Other types of knots	1	2336	3.09E-14	17.21	73.23	0.24	0.07	S	FMK
	5	123	1.80E-13	71.82	33.10	2.17	0.08	S	FMK
	32	−3169	7.67E-15	≥46.20	≥54.43	0.85	≤0.09	S	⋯
	51	−206	1.49E-14	≤0.39	8.02	≤0.05	0.13	Fe	⋯

Note. (1) Knot serial number in this work, (2) radial velocity, (3) He i 1.083 μm flux, (4)–(7) line ratios where [C I] = [C I] 0.985 μm, (8) knot type, (9) optical counterpart (see Section 4.3).

Download table as:  ASCIITypeset image

He i knots are detected along the forward shock front, from the southern base of the NE jet to the Fe K plume area (see Figure 6). The radial velocities of the He i knots are spread over a very wide range, from −3300 to +3200 km s⁻¹ (Figure 8 and Table 7). Since most of them have very high radial velocities, they are probably not circumstellar material. Their spectral properties are also different from the low-velocity circumstellar clumps in Cas A, which are called quasi-stationary flocculi (QSFs). The QSFs are He- and N-enriched circumstellar material ejected by the progenitor before the SN explosion and have recently been shocked by the SN blast wave (Peimbert & van den Bergh 1971; McKee & Cowie 1975; see also Koo et al.2023 and references therein). In the NIR, QSFs show strong He i lines, together with moderately strong [Fe ii] and [S ii] lines and weak Paγ and [C i] 0.985 μm lines (Chevalier & Kirshner 1978; Gerardy & Fesen 2001; Lee et al. 2017). The ratio of Paγ to He i 1.083 μm is typically ∼0.03 (see Koo et al. 2023). For comparison, all He i knots show strong (>10⁻¹⁵ erg cm⁻² s⁻¹) He i lines without [Fe ii] or [S ii] lines, and about half of them also show weak [C i] 0.985 μm lines.

What is the origin of the He i knots? The first possibility that we can consider is that they are the debris of the He-rich envelope material of the progenitor star expelled by the SN explosion, like the high-velocity N-rich knots (NKs) detected in optical studies (Fesen et al. 1987, 1988; Fesen & Becker 1991; Fesen 2001; Hammell & Fesen 2008; Fesen & Milisavljevic 2016). The spectra of NKs are dominated by strong [N ii] λ λ6548, 6583 lines in the wavelength region 4500–7200 Å. For most NKs, the Hα line is either much weaker than the [N ii] lines or undetected, so NKs are considered important evidence that the Cas A progenitor had a thin N-rich and H-poor envelope at the time of the SN explosion. A deep, ground-based [N ii] imaging observation detected 50 NKs, whereas HST observations revealed numerous NKs distributed around the remnant. In particular, in the eastern area where He i knots have been detected, many NKs had been detected (Figure 15; see also Figure 5 of Fesen 2001 or Figure 9(E) of Fesen & Milisavljevic 2016). The radial velocities of the bright NK knots in the eastern area are in the range of −2000 to +2500 km s⁻¹ (Fesen 2001). So the association of He i knots with NKs seems plausible considering that He abundance might be enhanced in the CNO-processed, N-rich envelope of the progenitor star. Apparently, however, none of the He i knots has an optical NK counterpart (Table 7; see Section 4.3). Another difficulty in connecting He i knots to NKs is the nondetection of the He i λ5876 line in NKs (Fesen & Becker 1991; Fesen 2001). The recombination emissivity of the He i λ5876 line is a factor of 2–10 smaller than that of the He i 1.083 μm line, depending on the electron density (10²–10⁴ cm⁻³) in the Case B approximation (e.g., Draine 2003). The flux will be further reduced by interstellar extinction. Hence, the He i λ5876 line is expected to be much weaker than the He i 1.083 μm line, but it is not obvious if that can explain the nondetection of He i λ5876 lines. An interesting possibility is that the He i knots originated from the lower portion of the He-rich envelope of the progenitor star where N is depleted. If a large amount of carbon is dredged up by convection, nitrogen could be depleted because ¹⁴N mixed into the lower layer by convection participates the ${}^{14}{\rm{N}}{(\alpha ,\gamma )}^{18}{\rm{F}}{({\beta }^{+})}^{18}{\rm{O}}{(\alpha ,\gamma )}^{22}\mathrm{Ne}$ reaction. Such a ¹⁴N-depleted region formed by convective dredge-up is often found at the pre-SN stage in massive star models (Y. Jung et al 2023, in preparation). Observationally, our slits are positioned toward [Fe ii] line-emitting knots by design. So it is possible that we are picking up He-rich, N-depleted knots that are located closer to the explosion center than the optical NKs but coincidentally overlap with the same region of sky as the Fe-emitting knots (Figure 15). Future studies may further explore the possibility.

Zoom In Zoom Out Reset image size

Another possibility is that they are metal-enriched SN ejecta with a relatively high He abundance. We note that some S- and Fe-rich knots have very high R(He i/[Fe ii]), as high as the He i knots, and that the majority of He i knots have [S ii] 1.03 and [C i] 0.985 μm flux upper limits that are not particularly strong (Table 7). This explanation also appears to be consistent with the idea that half of the He knots seem to be FMK-like knots in optical classifications (Figure 11). The overabundance of low atomic elements could be attributed to significant mixing among the nucleosynthetic layers, as pointed out in previous X-ray and optical studies (Fesen & Becker 1991; Fesen & Gunderson 1996). From optical spectroscopy, Fesen & Becker (1991) reported the detection of a hybrid ejecta knot showing O and S emission lines along with [N ii] and Hα emission lines in the NE jet region, and they called it a "mixed emission knot" (MEK). Additional optical studies toward the outer region of the remnant have further detected several MEKs, which suggest chemical mixing among the nucleosynthetic layers between the explosive O-burning layer and the N-rich photospheric envelope of the progenitor (Fesen & Gunderson 1996; Fesen 2001). Yet another possibility is that the He i knots are the remains of the complete Si burning with α-rich freeze-out in the innermost region of the progenitor as helium in Fe-rich knots (see also Lee et al. 2017). But it seems to be difficult for this scenario to explain the difference in radial velocities between the He i and Fe-rich knots.

It is also worth noticing that Figure 6 does not represent the overall distribution of He i knots. Since our observation was done toward ejecta knots bright in [Fe ii] emission, the detection of He i knots was unexpected. The He i knots were detected because they fell on the slit by chance. In order to see the overall distribution of the He i knots, which is essential for understanding their origin, we need He i 1.083 μm narrowband observations of the Cas A remnant.

In addition to the 67 SN ejecta knots, we also detected three extended H emission features in the outer eastern region (Section 3), and their physical properties are listed in Table 8. They show only unresolved (FWHM ≲9 Å) H i lines, e.g., Paγ 1.094 and Paβ 1.282 μm lines, without any emission lines from metallic elements, and have radial velocities between −20 and −30 km s⁻¹, indicating that they are of CSM/ISM origin. We can estimate the line-of-sight extinction to the H-emitting gas using the flux ratio of the H i emission lines. All three features show the Paβ line, and the brightest one (H2) also shows the Paγ line. The observed H i line ratio, F(Paγ)/F(Paβ), is 0.41 ± 0.09. By comparing with the unreddened line ratio of 0.56 at T_e = 10⁴ K and n_e = 10⁴ cm⁻³ (Hummer & Storey 1987), we obtained an N_H of ${7.6}_{-5.1}^{+6.4}\times {10}^{21}$ cm⁻², which corresponds to an A_V of ${4.1}_{-2.7}^{+3.4}$ mag. The extinction-corrected F([Fe ii] 1.257)/F(Paβ) is <0.2 (2σ upper limit), which is at least an order of magnitude smaller than that in the typical shock-excited CSM/ISM detected in the several Galactic SNRs (2–20; Koo & Lee 2015, and references therein). This low F([Fe ii] 1.257)/F(Paβ) indicates that the H i lines are associated with a photoionized gas.

We noticed that the positions of the extended H emission features are coincident with the region where a bright, diffuse optical cloud is detected (Figure 16). The diffuse cloud in the eastern region (hereafter "East Cloud"), located 27 east of the center of the remnant, was first identified in an early optical photograph taken by R. Minkowski (Minkowski 1968; van den Bergh 1971), and it has a triangular shape with a size of $\sim 1^{\prime}$ (see also the right panel of Figure 16). In a deep Hα narrowband image, however, it turned out to be more extended and diffuse (see Figure 10 in Fesen 2001 or Figure 2 in Weil et al. 2020). The follow-up optical spectroscopy showed that the East Cloud emits a strong Hα emission line accompanied by relatively weak [O iii] λ λ4959, 5007, [N i] λ λ6548, 6583, and [S ii]λ λ 6716, 6725 lines (Fesen et al. 1987; Weil et al. 2020).

Zoom In Zoom Out Reset image size

The origin of the East Cloud, as well as its physical association with Cas A, has been controversial. The early optical studies suggested that it is a small, diffuse H ii region adjacent to the remnant (Minkowski 1968; van den Bergh 1971). However, the lack of OB stars around the eastern cloud suggested that it is a diffuse CSM/ISM near Cas A excited by the UV/X-ray emission from the SN outburst (Peimbert 1971; van den Bergh 1971; Weil et al. 2020). In this scenario, the East Cloud could be either the diffuse ISM surrounding the progenitor star (Minkowski 1968; Peimbert & van den Bergh 1971; Peimbert 1971; van den Bergh 1971) or the diffuse CSM blown out from the progenitor star during its red supergiant phase (Fesen et al. 1987; Chevalier & Oishi 2003; Weil et al. 2020). The peculiar structure of the East Cloud, its projected proximity to the NE jet, and its distinct spectroscopic properties relative to the surrounding diffuse emissions are indirect evidence suggesting the physical association of the East Cloud with the remnant (Fesen et al. 1987; Fesen & Gunderson 1996; Fesen et al. 2006; Weil et al. 2020), while the presence of brightened ejecta knots matching the East Cloud's emission structure provides strong evidence that the East Cloud is physically adjacent to Cas A (Weil et al. 2020).

According to our spectroscopy, the mean v_LSR of the Paβ lines weighted by 1/${\sigma }_{v}^{2}$ (where σ_v is the 1σ uncertainty of the central velocity) is −22 ± 5 km s⁻¹ (see Table 8). Considering the systematic uncertainty from the wavelength calibration (∼12 km s⁻¹; see Section 2), the v_LSR of the East Cloud is −22 ± 13 km s⁻¹. This is consistent with the previous result: v_LSR = −20 ± 75 km s⁻¹ from a few bright optical lines (Fesen et al. 1987) and v_LSR = −13 ± 10 km s⁻¹ from the Hα lines (Alarie et al. 2014). Previous radio observations, however, showed that Cas A is located at the far side of the Perseus spiral arm, implying v_LSR ∼ −50 km s⁻¹ for Cas A (Zhou et al. 2018, and references therein). Furthermore, recent NIR observations for the southern optical QSF (knot 24) reported the detection of narrow (FWHM ∼ 8 km s⁻¹) [Fe ii] lines from unshocked pristine CSM with v_LSR ≃ −50 km s⁻¹ (Koo et al. 2020), which is consistent with the systematic velocity center v_LSR ≃ −50 km s⁻¹ of Cas A. So the velocity of the East Cloud (v_LSR ∼ −22 km s⁻¹) is considerably different from the systemic velocity of the Cas A SNR. This might indicate that the East Cloud is not physically associated with the remnant but rather is located ∼2 kpc from us (more than 1 kpc closer than the remnant), assuming the flat Galactic rotation model with the IAU standard (R₀ = 8.5 kpc and v₀ = 220 km s⁻¹). Relatively low line-of-sight extinction of the East Cloud (A _V ∼ 4 mag) derived from the H i lines compared to that toward the remnant (A _V = 6–10 mag; see Figure 2) seems to support this possibility. On the other hand, if the East Cloud is the CSM material ejected from the progenitor star, as suggested by Chevalier & Oishi (2003) and Weil et al. (2020), the velocity difference (∼30 km s⁻¹) should represent the line-of-sight component of the ejection velocity of the CSM. And, considering that the East Cloud is located at the eastern outer edge of the remnant, the ejection velocity should have been much higher than 30 km s⁻¹ or the typical wind velocity of red supergiant stars, e.g., 10–20 km s⁻¹ (Smith 2014).