5 Abundance results

 

 

Table 6: Derived elemental abundances for our program stars. The abundances are given in the format ${\rm [X/H]} = \log{\rm (X/H)}_{\star} - \log{\rm (X/H)}_{\odot}$, where X denotes the element in question. The ionization stages are also indicated. The line-to-line scatter is given for each star and element when more than one line is used in the abundance analysis. The error in the derived abundance due to line-to-line scatter is thus = line-to-line scatter / $\sqrt n_{\rm lines}$, the number of lines used are given in parentheses. Oxygen abundances are given separately for the forbidden line at 630 nm and the triplet at 777 nm. No star has observations of both

	HD 10780	HD 32147	HD 99491	HD 104304	HD 121370	HD 145675	HD 182572	HD 196755
C I	0.28	-	-	-	-	-	-	-
O I630	-	-	0.27	0.37	-	-	-	-
O I777	-	-	-	-	-	0.48 0.15 (3)	0.62 0.10 (3)	0.11 0.05 (3)
Na I	-0.03	0.64	0.34	0.37	0.50	-	-	-
Al I	-0.01 0.03 (3)	0.48 0.05 (3)	0.41	0.25	-	0.54 0.02 (3)	0.55 0.04(3)	0.02 0.05 (3)
Si I	0.03 0.05 (5)	0.36 0.15 (5)	0.35 0.10 (7)	0.27 0.08 (8)	0.40 0.14 (6)	0.61 0.07 (3)	0.49 0.18(6)	0.09 0.05 (4)
						0.52 0.19 (4)
Ca I	0.13 0.10 (5)	-	0.18 0.08 (4)	0.15 0.07 (4)	0.11	0.21	0.28	-0.02 0.23 (2)
S I	-	-	0.56	0.49	0.72	-	-	0.38
Sc I	-	-	0.10	0.11	-	-	-	-
Sc II	-0.12	0.36	-	0.32 0.13 (3)	0.11	0.66	0.36	0.11
Ti I	0.10 0.07 (2)	0.66 0.18 (2)	0.17 0.10 (13)	0.11 0.10 (12)	0.22 0.22 (3)	0.62 0.03 (2)	0.32 0.03 (3)	0.12 0.09 (3)
Cr I	0.01 0.15 (3)	0.50 0.12 (3)	0.19 0.11 (6)	0.14 0.09 (5)	0.22 0.04 (2)	0.42 0.04 (3)	0.40 0.01 (2)	-0.03 0.16 (2)
Cr II	-	-	0.35	0.26	-	-	-	-
Fe I	-0.02 0.07 (39)	0.28 0.11 (39)	0.22 0.08 (42)	0.15 0.08 (44)	0.24 0.14 (32)	0.47 0.11 (30)	0.34 0.14 (29)	0.02 0.09 (28)
Fe II	-0.11 0.10 (4)	0.24 0.10 (4)	0.24 0.07 (5)	0.17 0.08 (5)	0.19 0.07 (4)	0.49 0.02 (3)	0.32 0.08 (3)	0.06 0.03 (4)
						0.56 0.14 (4)
Co I	-0.06 0.06 (4)	0.56 0.19 (4)	0.26 0.08 (5)	0.04 0.54 (6)	0.32	0.81 0.11 (4)	0.47 0.07 (3)	0.27
Ni I	-0.03 0.06 (17)	0.29 0.08 (15)	0.26 0.07 (20)	0.20 0.09 (19)	0.31 0.16 (13)	0.55 0.10 (11)	0.36 0.08 (18)	0.00 0.09 (13)

The stellar abundances derived in this study are summarized in Table 6. We will discuss the abundance determination for each element separately. For some elements only one or a couple of lines have been used and the results are therefore more tentative than firm. The number of lines used for each element are also indicated in the table.

Iron abundances

Iron abundances are derived from a large number of lines, 28 to 44 lines per star, which means that the errors in the mean are very small, typically less 0.02 dex. Thus, the error in Fe abundances is negligible in the error budget for the abundance ratios.

In Table 7 we compare the iron abundances in this study and those quoted by Taylor (1996). For HD 32147, HD 99491, HD 121370, HD 145675, and HD 182572 their SMR status is confirmed. HD 104304 is a marginal case and HD 10780 and HD 196755 are shown to not be SMR stars.

Oxygen

The forbidden O I line at 6300 Å was only measured in two stars, HD 99491 and HD 104304; they have [O/Fe] of 0.05 and -0.17dex, respectively. The errors in the O abundances are dominated by the measurement error for the [O I] line. Our spectra have S/N of $\sim$250, which translates into an uncertainty of $\sim$0.15 dex in [O/Fe]. Thus, our observed [O/Fe] are well within this scatter, and our data follow the trends found in Nissen & Edvardsson (1992) and Feltzing & Gustafsson (1998).

Three of our stars have useful observations of the triplet lines around 777 nm. For those three stars we get [O/Fe] = 0.01, 0.28, and 0.09 dex respectively. Line-to-line scatter is 0.1 dex or less for these stars which means that formal errors are less than 0.1 dex for all three stars. These oxygen abundances should be fairly reliable as we are dealing with stars that are similar to the Sun and our study is differential. Edvardsson et al. (1993) found a good correlation between oxygen abundances derived from the forbidden line and those derived from the triplet. Note, however, that Feltzing & Gustafsson (1998) found no such correlation for their very metal-rich sample. Thus, in conclusion, the [O/Fe] for HD 196755 derived from the triplet should be robust while the [O/Fe] for HD 145675 and HD 182572 are more uncertain in term of possible NLTE effects.

Sodium

Only one Na I line was available for analysis in our spectra, but it has been widely used in other abundance studies. Therefore, we are confident that it is giving us reliable Na abundances.

Aluminum

Aluminum shows a somewhat puzzling behaviour. Both Edvardsson et al. (1993) and Feltzing & Gustafsson (1998) found [Al/Fe] to be solar for all stars with [Fe/H] >0.0 dex. However, three of our stars show unexpectedly high [Al/Fe] abundances. The line-to-line scatter is small for all stars with all three lines measured.

Silicon

A flat distribution with some internal scatter is found. The line-to-line scatter is, as in Feltzing & Gustafsson (1998), on the larger side for the number of lines used.

Calcium

For three stars we observe several lines of Ca I. These stars all show [Ca/Fe] close to the solar value, as expected. For the other stars only one or two lines were available and the results are therefore uncertain.

 

 

Table 7: Comparison of [Fe/H] from Taylor (1996) and this work. We also give, in Cols. 3 and 4, the VSL and SMR status for the stars according to Taylor (1996)

ID	VSL	SMR	[Fe/H]	[Fe/H]
			(Taylor)	This work
HD 10780	no		0.396	-0.02
HD 32147	yes		>0.1	0.28
HD 99491	marg		0.115	0.20
HD 104304	marg		0.326	0.16
HD 121370		95%	0.305	0.25
HD 145675	marg	98%	0.38	0.47
HD 182572		98%	>0.341	0.35
HD 196755			0.500	0.02

Sulfur

Our linelist contains 2 S I lines, however, for those stars where we could determine S abundances only one line was available in each star. Our abundances are therefore uncertain. We note that [S/Fe] appears somewhat high.

Scandium

Sc abundances were derived from both Sc I and Sc II lines. The [Sc/Fe] values fall within the range expected from Feltzing & Gustafsson (1998) and the trend in our [Sc/Fe] data is flat at around 0.1-0.15 dex. We note though that our most metal rich stars are overabundant in Sc in contrast with Feltzing & Gustafsson (1998) which show a tendency for the most metal-rich stars to be underabundant.

Titanium

An overall flat trend is found for our stars in accordance with previous studies, i.e. Feltzing & Gustafsson (1998). HD 32147 stands out with an extremely high Ti abundance, perhaps indicative of an underestimated temperature (see Thorén & Feltzing 2000).

Chromium

The behaviour is the same as found by Feltzing & Gustafsson (1998), except for HD 32147. Essentially, [Cr/Fe] has a flat trend above solar metallicity.

Cobalt

We find a large scatter in the [Co/Fe] abundances. The origin of this large scatter is not entirely clear, but it is accompanied by a large line-to-line scatter as well, and thus the reason may be sought among the selection of stellar lines.

Nickel

A large number of Ni I lines were used, both to determine the effective temperature as well as the Ni abundances. All our stars have virtually solar [Ni/Fe] in agreement with Edvardsson et al. (1993) and Feltzing & Gustafsson (1998) for stars in this [Fe/H] range. Interestingly, we find that our most metal-rich star, HD 145675, shows a slightly enhanced [Ni/Fe]. Tentative enhancements of Ni in the most metal-rich stars can be found also in the studies by Feltzing & Gustafsson (1998) and Thorén & Feltzing (2000). All three studies adopted different techniques for derivation of stellar parameters, and thus the result seems to be significant. However, further investigations should be undertaken.