In Table 2 (click here) we summarize relevant information about our program stars.
The UBV, 
(de-reddened by using UVBYBETA code of Moon &
Dworetsky 1985) indices and 
values are from Renson
(1991). Geneva photometry is from Rufener (1980).
Orbital periods and eccentricities of
HD108651, HD116657 and HD 138213 are according to Conti & Barker
(1973), Gutmann (1965) and Lucy & Sweeney
(1971), respectively.
| Star | HD108561 | HD116657 | HD138213 |
| HR4751 | HR5055 | HD5752 | |
| U-B | 0.08 | 0.1 | 0.1 |
| B-V | 0.22 | 0.13 | 0.1 |
| V | 6.6 | 4.0 | 6.1 |
| E(b-y) | 0.006 | -0.01 | -0.014 |
| (b-y)0 | 0.114 | 0.063 | 0.046 |
| m0 | 0.233 | 0.239 | 0.191 |
| c0 | 0.832 | 0.911 | 1.141 |
 | 2.843 | 2.886 | 2.86 |
| U | 1.516 | - | 1.648 |
| V | 0.689 | - | 0.825 |
| B1 | 0.958 | - | 0.909 |
| B2 | 1.411 | - | 1.445 |
| V1 | 1.400 | - | 1.524 |
| G | 1.828 | - | 1.989 |
 [d] | 68.29 | 175.55 | 105.95 |
| e | 0.36 | 0.46 | 0.0 |
 [km  | 15 | 60 | 30 |
Table 3 (click here) summarizes the atmospheric parameters as derived from the
calibrations of different authors. Temperatures from UBV photometry
as well as those of Castelli & Kurucz (1994) from
index are all systematically lower compared with the other calibrations of
Strömgren or Geneva photometry. Nevertheless,
the most elaborated and two-dimensional calibrations of
Smalley & Dworetsky (1995) and Kobi & North
(1990) are both consistent. Thus, we accepted their rounded off means
as the best choice for model atmosphere parameters. All three stars are
SB1 binaries, hence the possible influence of their companions
was neglected.
| HD108561 | HD116657 | HD138213 | ||||
 |  | | | |
| Note |
| 69801 | - | 77701 | - | 80701 | - | U |
| 75502 | - | 80902 | - | 82902 | - | |
| 79001 | - | 84601 | - | 86601 | - | |
| 77702 | - | 82602 | - | 84702 | - | S |
| 81303 | 4.313 | 83903 | 4.43 | 84303 | 3.613 | |
| 78004,6 | - | 81504,7 | - | 79004,8 | - | |
| 78501 | - | - | - | 86101 | - | G |
| 79105 | 4.175 | - | - | 84105 | 3.685 | |
| 8000 | 4.2 | 8400 | 4.4 | 8400 | 3.6 | A |
,
7-for 
,
8-for 
.
A detailed spectrum synthesis of the spectral region around the
Li I 6708 line was accomplished using the SYNTH (Piskunov
1992) and SYNSPEC (Hubený 1987; Zboril 1989,
private communication) codes with the model atmospheres
interpolated from Kurucz (1992) ATLAS9 grid.
VALD atomic line database was used to create a line list
for the spectrum synthesis (Piskunov et al. 1995).
We first estimated the iron abundance. However, the normal
abundance of iron would result in a quite insufficient
depth of the synthetic absorption profile at 6705Å, while
at 6713Å it would be deeper by more than a factor of two.
Unspecified missing opacity at 6705Å was also
mentioned by Gerbaldi et al. (1995) and Hack et al.
(1997). Following one of the possibilities of how to cope with such
an inconsistency (van't Veer 1997, private communication), we adjusted
using the solar spectrum (Kurucz et al. 1984).
For the solar photosphere we used a model with 
, interpolated from Kurucz's (1992) grid.
The value of microturbulence for the theoretical spectrum computation was
taken 1 kms-1.
Comparing the theoretical and observed solar equivalent widths we estimated
the values of the lines as follows: Fe I 6705.101, -1.10 (VALD
-1.50);
Fe I 6712.676, -4.62 (-2.88) and Fe I 6713.046, -1.61 (-1.05).
Moreover, two lines,
Fe I 6713.745 and Fe I 6713.771 given in VALD seem to refer to the same
transition as they have identical values,
lower excitation potentials, as well as lower and upper J values.
The difference of 0.026Å in
wavelength is not unusual when referring to different sources.
Accepting both
the lines as different transitions, their values would have to be
less in order to reach a fit. In our line list only the latter one,
6713.771Å, was retained. This also satisfied to the solar spectrum.
Only after these adjustments could we fit the observed spectra of all
three stars.
With the known Fe abundance we determined the
microturbulent velocity value by fitting the profile of the blend at 6678Å
created mainly by a stronger Fe I line which showed to be sensitive
to 
. This allows the value to be set with a formal
accuracy 
for a given abundance value.
To derive the abundances of other elements, the computed spectra were
convolved with the instrumental profile (Gaussian of 0.22Å half-width)
and rotationally broadened to fit with the observed spectra. Though a good fit
was reached for a majority of the absorptions seen in the spectra
(see Fig. 4 (click here)),
there are still a few features like 6691Å in HD108651 or 6709Å
in the two hotter stars which are not explained sufficiently with the
derived abundances and known opacity sources.
The abundances obtained in this way are introduced
in terms of 
in Table 4 (click here).
Taking into account the accuracy of the atmospheric parameters, as
well as atomic data, the values for Li, Al, Si, Ti and Fe are determined within
dex, while the abundances of the other elements, occurring
mainly in weak blends, are only approximate.
| Sun | A | B | C | |||
 | 1 | 2 | 3 | 3 | 3 | |
| Li | 1.16 | +1.79 | +1.92 | +1.64 | +2.06 | |
| C | 8.60 | -0.27 | -0.4 | -0.4 | -0.2 | |
| N | 8.00 | +0.5 | 0.0 | +0.3 | ||
| Al | 6.47 | -0.66 | -0.81 | -0.49 | -0.02 | |
| Si | 7.55 | +0.34 | +0.53 | +0.45 | +0.67 | |
| Ca | 6.36 | -0.25 | -0.37 | -0.33 | +0.12 | |
| Ti | 4.99 | +0.09 | -0.04 | -0.04 | -0.04 | |
| Fe | 7.67 | +0.07 | +0.15 | +0.19 | +0.01 | +0.08 |
| Ni | 6.25 | +0.59 | +0.9 | |||
| Ce | 1.55 | +1.93 | +1.4 | +1.9 | +1.8 | |
| Sm | 1.00 | +0.82 | +1.2 | +0.8 | ||
| Gd | 1.12 | +1.27 | +1.8 | +1.7 | ||
| 5.6 | 1.8 | 2.7 | 0.5 | ||
|
| 20 | 20 | 48 | 32 | ||
- The entries are upper limits.
Table 5 (click here) lists the identified lines, their and approximate
theoretical equivalent widths in mÅ computed with the resulting abundances.
 & Ion | | A | B | C |
| 6675.260 N I | -1.98 | 0.6 | ||
| 6677.305 Fe I | -1.59 | 21.3 | 13.7 | 9.7 |
| 6677.759 Fe I | -2.17 | 1.6 | 0.6 | |
| 6677.955 Fe I | -3.63 | 1.7 | 0.6 | |
| 6677.987 Fe I | -1.42 | 64.5 | 33.3 | 25.5 |
| 6678.154 He I | 0.33 | 0.5 | ||
| 6678.803 Co I | -2.68 | 0.8 | ||
| 6678.837 Fe I | -0.45 | 1.5 | 1.2 | 1.9 |
| 6678.898 Fe I | -4.88 | 1.9 | 0.9 | 1.3 |
| 6679.222 Sm I | -0.83 | 1.2 | ||
| 6679.352 Ca I | -0.84 | 1.6 | 1.7 | 4.7 |
| 6679.566 Si I | -1.26 | 0.5 | ||
| 6679.569 Fe I | -3.12 | 3.7 | 2.1 | 3.0 |
| 6679.642 C I | -2.28 | 3.3 | 3.0 | 3.4 |
| 6679.748 Fe I | -0.47 | 1.5 | 1.2 | 1.8 |
| 6680.123 Ni I | -1.11 | 3.6 | ||
| 6680.133 Ti I | -1.85 | 6.0 | 4.1 | 4.8 |
| 6680.949 C I | -2.56 | 2.8 | 2.5 | 2.8 |
| 6681.199 Gd I | -1.48 | 0.9 | ||
| 6681.530 Sm I | -1.19 | 0.3 | 0.1 | |
| 6683.161 Si I | -2.16 | 4.7 | 2.7 | 2.7 |
| 6683.970 C I | -2.15 | 4.4 | 4.0 | 4.5 |
| 6684.179 Fe I | -0.65 | 1.0 | 0.8 | 1.2 |
| 6685.474 Fe I | -0.72 | 0.8 | 0.7 | 1.0 |
| 6685.622 N I | -1.78 | 1.0 | 1.0 | |
| 6685.891 C I | -3.14 | 0.5 | ||
| 6687.490 Fe I | -2.32 | 3.1 | 1.2 | 0.8 |
| 6687.797 Sm I | -0.97 | 0.4 | 0.1 | |
| 6688.794 C I | -2.09 | 5.0 | 4.6 | 5.2 |
| 6691.021 Ca I | -0.42 | 1.4 | 1.0 | 1.3 |
| 6691.325 Si I | -2.97 | 0.9 | 0.5 | 0.5 |
| 6692.265 Fe I | -2.95 | 0.8 | ||
| 6692.447 N I | -1.28 | 3.0 | 1.2 | 3.0 |
| 6693.169 Fe I | -0.73 | 0.5 | 0.8 | |
| 6693.210 Fe I | -2.67 | 2.3 | 1.6 | 2.4 |
| 6693.555 Sm I | -0.37 | 2.5 | ||
| 6694.721 Sm I | -1.41 | 0.2 | ||
| 6696.023 Al I | -1.35 | 2.3 | 3.1 | 5.2 |
| 6696.044 Si I | -1.83 | 9.9 | 6.0 | 5.8 |
| 6696.185 Al I | -1.58 | 0.5 | 0.9 | |
| 6696.320 Fe I | -2.04 | 2.1 | 0.9 | 0.6 |
| 6696.788 Al I | -1.42 | 0.8 | 1.3 | |
| 6698.673 Al I | -1.65 | 1.2 | 1.6 | 2.7 |
|
& Ion | | A | B | C |
| 6699.142 Fe I | -2.10 | 2.6 | 1.1 | 0.7 |
| 6699.164 Fe I | -4.04 | 0.6 | ||
| 6700.890 Ni I | -2.32 | 1.0 | ||
| 6702.862 N I | -1.81 | 0.9 | 0.9 | |
| 6703.567 Fe I | -3.16 | 3.5 | 1.1 | 0.9 |
| 6704.147 Gd I | -1.83 | 0.3 | 0.1 | |
| 6704.481 Fe I | -2.66 | 1.3 | 0.5 | |
| 6704.524 Ce I | -0.51 | 1.2 | 0.9 | |
| 6704.839 N I | -1.35 | 2.6 | 1.0 | 2.6 |
| 6705.101 Fe I | -1.10 | 16.4 | 7.3 | 5.0 |
| 6705.131 Fe I | -2.37 | 0.8 | ||
| 6706.051 Ce I | -1.25 | 0.5 | ||
| 6706.107 N I | -1.80 | 0.9 | 0.9 | |
| 6706.880 Fe I | -4.10 | 0.5 | ||
| 6706.980 Si I | -2.48 | 2.9 | 1.7 | 1.7 |
| 6707.473 Sm I | -1.48 | 0.3 | 0.1 | |
| 6707.761 Li I | -0.01 | 6.0 | 1.7 | 2.6 |
| 6707.912 Li I | -0.31 | 3.1 | 0.9 | 1.3 |
| 6708.759 N I | -1.79 | 1.0 | 1.0 | |
| 6708.885 Fe I | -0.52 | 1.4 | 1.1 | 1.6 |
| 6711.323 C I | -2.47 | 3.4 | 3.0 | 3.4 |
| 6711.575 Ni I | -3.81 | 1.0 | ||
| 6712.438 Fe I | -2.16 | 1.3 | 0.5 | |
| 6712.676 Fe I | -4.62 | 0.2 | 0.1 | 0.0 |
| 6713.046 Fe I | -1.61 | 6.8 | 2.9 | 2.0 |
| 6713.195 Fe I | -2.56 | 1.8 | 0.7 | 0.5 |
| 6713.586 C I | -2.17 | 6.6 | 5.9 | 6.6 |
| 6713.771 Fe I | -1.60 | 5.9 | 2.5 | 1.7 |
| 6715.383 Fe I | -1.64 | 7.0 | 2.9 | 2.0 |
| 6716.237 Fe I | -1.92 | 4.0 | 1.6 | 1.1 |
| 6716.973 Si I | -0.10 | 0.7 | ||
| 6717.298 Fe I | -1.96 | 1.8 | 0.7 | 0.5 |
| 6717.524 Fe I | -2.45 | 1.2 | 0.5 | |
| 6717.681 Ca I | -0.61 | 16.1 | 10.4 | 12.7 |
| 6717.794 Ti I | -1.80 | 6.4 | 4.4 | 5.2 |
| 6717.964 Fe I | -0.81 | 0.7 | ||
| 6718.130 Gd I | -1.00 | 0.7 | 0.3 | |
| 6718.883 Fe I | -0.73 | 0.7 | ||
| 6719.609 Si I | -2.50 | 2.0 | 1.1 | 1.1 |
| 6719.639 Fe I | -0.52 | 1.3 | 1.0 | 1.6 |
| 6720.280 Ce I | -1.35 | 0.5 | ||
| 6720.908 Si I | -2.41 | 3.9 | 2.2 | 2.2 |
| 6721.848 Si I | -1.49 | 26.9 | 16.1 | 15.3 |
| 6722.077 Fe I | -0.57 | 1.0 | 0.8 | 1.2 |
| 6722.610 N I | -0.98 | 5.8 | 2.4 | 5.9 |
| 6722.759 Co I | -0.81 | 1.3 | ||
| 6723.220 Co I | -0.99 | 0.8 |
HD108651 (HR4751, DM
, Sp.Am, 
)
is the well-known and studied Am binary. A number of authors
attempted a chemical composition determination and the most elaborated results
are compared with ours in Table 4 (click here).
Our microturbulent velocity, 
,
is remarkably lower than those given by other authors
(e.g. 5.6kms-1 by Savanov 1996, 7.0kms-1
Smith 1971). But although derived from a single line it is in
better accordance with the typical value of 2 kms-1 for B-A stars
(Lemke 1989; Adelman & Fuhr 1985).
As far as is concerned, besides the value introduced in Table
2 (click here),
various values can be found in previous papers. Bernacca et al.
(1971) gives 6 kms-1, while Uesugi & Fukuda
(1982) and Kraft
(1965) list 12kms-1.
Our value of 
is the same as that derived by Savanov
(1996).
In general, the abundance pattern with medium overabundances of rare earths
and a low Ca/Fe ratio (see Sect. 4.2) satisfies the pronounced Am
characteristics of this star.
HD116657 (HR5055, DM
, Sp.Am, 
). Even
if the projected rotational velocity value, 
,
derived by us is 
less than the one given in Table
2 (click here),
this star has one of the highest values among those so far studied
for Li abundance.
The microturbulent velocity, 
, is not
unusual for these stars. The abundance pattern, as well as the
Ca/Fe ratio, aligns the star with typical Am stars.
HD138213 (HR5752, DM
, Sp.Am, 
).
The value of the projected rotational velocity, 
, estimated by us is close to the one given in Table 2 (click here).
The microturbulence derived, following the procedure described, corresponds to
. Considering the lower surface
gravity,
, this is rather a low value.
The Am anomalies of this star are not as pronounced as in the two other
stars and the Ca/Fe ratio is larger.
If its true rotational velocity was smaller than that of HD 116657, which
is more peculiar, this could be an interesting paradox.
The mild anomalies could be a result of its lower
gravity, as evolved stars might loose their Ca deficiency
(Berthet 1992) and/or rotationally induced mixing might be more
effective in such stars (Michaud 1982).