Molybdenum Abundances in MINCE Stars

Abstract

Molybdenum (Mo, Z = 42) is a neutron-capture element with seven stable isotopes that can be produced by different processes. Previous studies have shown a large scatter in molybdenum abundances for metal-poor ([Fe/H] < $-1$) stars, indicating that multiple nucleosynthetic channels are responsible for molybdenum production even at very low metallicity. To understand which different nucleosynthesis processes are involved in the chemical enrichment of this element in the Galaxy, a large sample of precise molybdenum abundance is required. In this study, we present molybdenum abundances of 27 metal-poor stars from the Measuring at Intermediate Metallicity Neutron-Capture Elements project sample. We derived molybdenum abundances using three Mo i lines at 550.6 nm, 557.0 nm, and 603.0 nm, which proved to be reliable for measuring Mo abundances in giant stars with [Fe/H] $>-2$. Our derived [Mo/Fe] abundance ratios show on average slightly higher values (∼0.2 dex) compared to the literature samples. This may be due to an observational bias or to non-local thermodynamic equilibrium effects. We also found that Gaia-Sausage-Enceladus candidate stars have lower [Mo/Fe] than the sample average, while the only Sequoia candidate star has a higher [Mo/Fe] than most sample stars.

1. Introduction

Molybdenum (Mo) abundances in metal-poor stars are an excellent tool for understanding the different nucleosynthesis processes that contribute to the chemical enrichment of the Galaxy. This element, in fact, has seven stable isotopes that are produced by different processes: ${}^{92}\mathrm{Mo}$ and ${}^{94}\mathrm{Mo}$ are p-isotopes, and they can be produced by either proton captures (p-process) or photo-disintegration of heavier nuclei ($\gamma$-process) [1,2]; ${}^{96}\mathrm{Mo}$ is only made by the slow neutron capture process (s-process; see, e.g., [3] and references therein), while ${}^{100}\mathrm{Mo}$ is mostly produced by the rapid neutron capture process (r-process; see, e.g., [4] and references therein); and the other isotopes (${}^{95}\mathrm{Mo}$, ${}^{97}\mathrm{Mo}$, and ${}^{98}\mathrm{Mo}$) are produced by both the s-process and the r-process. Other possible nucleosynthesis sources of Mo are the intermediate neutron-capture process (i-process; e.g., [5,6]) and neutrino-driven ejecta from core-collapse supernovae (e.g., [7,8]).

According to [9], the most abundant Mo isotope in the Solar System is ${}^{98}\mathrm{Mo}$, which constitutes 24.4% of the total Mo abundance. This is followed by the isotopes ${}^{96}\mathrm{Mo}$, ${}^{95}\mathrm{Mo}$, and ${}^{92}\mathrm{Mo}$, which account for 16.7%, 15.8%, and 14.5% of the total Mo abundance, respectively. The remaining isotopes, on the other hand, collectively contribute less than 30% of the total Mo abundance (9.8% of ${}^{100}\mathrm{Mo}$, 9.6% of ${}^{97}\mathrm{Mo}$, and 9.2% of ${}^{94}\mathrm{Mo}$). Isotopic abundances of Mo can be measured in presolar silicon carbide (SiC) and graphite grains (see, e.g., [10]). In particular, high-precision measurements of Mo isotopes in presolar SiC grains can be used to quantify the relative s-, r- and p-process contributions to solar isotopic abundances, providing important new constraints on the nucleosynthesis of this element (see, e.g., [11]).

Stellar spectra allow us to determine the total abundance of Mo in the star, but not its isotopic abundances, because the isotopic shifts of the Mo i lines cannot be resolved. High-resolution spectroscopic studies of metal-poor stars ([Fe/H] < $-1$) have shown a substantial star-to-star scatter in Mo abundances among both dwarf and giant stars. This pronounced dispersion implies that multiple nucleosynthetic channels operate even at the lowest metallicities [12,13,14,15,16]. To shed light on the origin of Mo and its chemical evolution in the Galaxy, a large, homogeneously analysed sample of high-resolution, high-signal-to-noise-ratio spectra of metal-poor stars is required.

The present study is conducted within the framework of the Measuring at Intermediate metallicity Neutron-Capture Elements (MINCE) project, which seeks to collect heavy element abundances for several hundred stars at intermediate metallicity using multiple telescopes worldwide. The project focuses on tracing nucleosynthetic signatures preserved in old, metal-poor stars, with particular emphasis on elements with atomic number $Z>30$, namely the neutron-capture elements. This paper presents Mo abundances for a sample of 27 metal-poor stars with $-2.2<[\mathrm{Fe}/\mathrm{H}]<-1.0$ using the same set of stellar spectra analysed in [17].

2. Observational Data

The sample consists of 27 giant stars observed between September 2020 and August 2021 with the Ultraviolet and Visual Echelle Spectrograph (UVES, [18]) at the European Southern Observatory’s Very Large Telescope and second Unit Telescope of Paranal Observatory (Chile). The observations were obtained with a slit width of ${0.5}^{\prime \prime }$, providing a resolving power of 65,000 in the blue and 75,000 in the red. Two different instrument configurations were used, Dichroic#1 (346 + 580) and Dichroic#2 (437 + 760), giving a total spectral coverage from 304 nm to 945 nm for each star. The spectra have an average signal-to-noise ratio (S/N) of ∼150 at 580 nm.

3. Abundance Analysis

Previous studies of Mo abundances in metal-poor stars mostly relied on the Mo i line at 386.4 nm, since for stars with metallicity below $-2$ this is the only visible transition in the optical spectrum [14,15,16]. However, at higher metallicities, the determination of Mo abundance from this line becomes extremely dependent on the continuum placement and the correct reproduction of the blends, given its location in a region of the spectrum crowded with atomic and molecular features. Studies focusing on stars in the Galactic disc and bulge [19,20,21] have demonstrated that it is possible to derive Mo abundances from other lines located in the red part of the spectrum. These lines represent valid alternatives to the 386.4 nm line for determining Mo abundances in the stars in our sample, since they become visible in the spectra of giant stars with [Fe/H] $\gtrsim -2$.

In this study, Mo abundances were derived from three Mo i absorption lines at 550.6 nm, 557.0 nm, and 603.0 nm using the code MyGIsFOS [22]. MyGIsFOS is an automatic abundance analysis pipeline to derive stellar parameters and chemical abundances from stellar spectra. The code derives chemical abundances through a ${\chi }^2$ minimisation fit of the line profile using a grid of synthetic spectra. The grid has been computed using 1-dimensional (1D) local thermodynamic equilibrium (LTE) ATLAS12 model atmospheres and the spectral synthesis code SYNTHE [23,24]. The adopted atomic data for these lines are listed in

Table 1. Atomic data for the two Mo i lines adopted in this study. The columns gfflag and synflag are respectively the $gf$ relative quality and blending quality flags according to the Gaia-ESO survey line list [25].

Wavelength [Å]	Excitation Potential [eV]	log gf	Reference	gfflag	synflag
5506.493	1.335	0.060	[26]	Y	U
5570.444	1.335	−0.337	[26]	Y	U
6030.644	1.531	−0.523	[26]	Y	Y

. In the last two columns of Table 1 we also list the quality flags provided in the the Gaia-ESO line list version 6 [25]. A value of gfflag equal to Y means that the line has a good quality $log\mathrm{gf}$, while a value of synflag equal to Y means that the line is unblended. The two Mo lines at 550.6 nm and 557.0 nm have a synflag equal to U, which means that their blending quality is undecided.

For each star in the sample, we adopted the effective temperature (${\mathrm{T}}_{\mathrm{eff}}$), surface gravity ($log\mathrm{g}$), microturbulent velocity (${\mathrm{v}}_{\mathrm{turb}}$), and metallicity ([Fe/H]) derived in [17]. The adopted values are listed in

Table 2. Stellar parameters and Mo abundances for our sample of stars. The value $\sigma$ represents the line-to-line scatter of Mo abundance.

Star	Teff [K]	log g [dex]	vturb [km/s]	[Fe/H] [dex]	Nlin	[Mo/H] [dex]	$\mathit{\sigma }$ [dex]
BD-11 3235	4158	0.90	1.93	−2.11	1	−1.83
BD-13 934	4245	0.91	1.97	−1.93	2	−1.38	0.02
BD-13 3195	4280	1.15	1.81	−1.41	2	−0.65	0.02
BD-15 5449	4230	0.80	2.01	−2.19	1	−1.86
BD-16 2232	4367	0.98	1.96	−1.77	2	−1.33	0.03
BD-17 3143	4344	0.48	2.19	−1.68	2	−1.20	0.02
CD-23 1855	4551	1.23	1.96	−1.96	1	−1.56
CD-27 16505	4046	0.80	1.85	−1.33	3	−0.85	0.05
CD-28 10039	4100	0.71	1.91	−1.12	2	−0.75	0.01
CD-28 16762	4291	0.96	1.93	−1.63	2	−1.15	0.05
CD-28 17446	4063	0.75	1.93	−1.84	3	−1.45	0.02
CD-29 15930	4105	0.72	1.96	−1.79	2	−1.29	0.03
CD-33 2721	4343	0.78	2.06	−1.89	2	−1.43	0.01
CD-33 15063	4140	0.90	1.89	−1.80	3	−1.09	0.06
CD-36 518	4424	1.24	1.86	−1.74	2	−1.39	0.02
CD-39 6037	4082	0.92	1.79	−0.99	3	−0.74	0.01
CD-39 9313	4323	1.03	1.93	−1.68	3	−1.37	0.04
CD-43 7161	4167	0.82	1.93	−1.50	3	−1.15	0.01
CD-44 12644	4564	1.15	2.00	−1.44	2	−0.85	0.05
CD-44 15269	4265	1.00	1.87	−1.39	3	−0.98	0.02
CD-57 1959	4342	1.09	1.93	−2.00	1	−1.71
CD-62 15	4584	1.26	1.97	−2.11	1	−1.66
HD 6518	4285	1.15	1.81	−1.44	2	−1.03	0.07
HD 19367	4345	0.74	2.10	−2.12	1	−1.63
TYC 6199-351-1	4111	0.99	1.82	−1.66	2	−1.30	0.02
TYC 8161-753-1	4266	1.25	1.80	−1.77	2	−1.22	0.03
TYC 8633-2281-1	4118	0.58	2.06	−2.05	2	−1.74	0.02

. The stellar parameters were derived using the procedures described in [17,27].

4. Results and Discussion

The abundances of Mo derived for our sample of stars are listed in Table 2. The abundances are expressed in the form of the abundance ratio [Mo/H], where [Mo/H] = ${log}_{10}{(\mathrm{Mo}/\mathrm{H})}_{★}-{log}_{10}{(\mathrm{Mo}/\mathrm{H})}_{\odot }$. The adopted solar abundance of Mo is 1.92, as derived in [9]. The uncertainty $\sigma$ represents the line-to-line scatter when more than a single line of Mo was used to determine the total Mo abundance.

In our spectra, the strongest line is the one at 550.6 nm, which, however, is located on the blue wing of the Fe I line at 550.7 nm. Despite blending with this relatively strong iron line, the profile of the Mo line is distinctly visible in all of the analysed stars. The lines at 557.0 nm and 603.0 nm, on the other hand, do not appear to have any significant blends. In our sample, the line at 557.0 nm is visible in stars with metallicity [Fe/H] $>-2$, while the line at 603.0 nm is detectable for [Fe/H] $>-1.8$. In the cases where the three lines were visible in the spectrum of the analysed star, the abundances derived from the individual lines were in excellent agreement, with an average line-to-line scatter of 0.04 dex.

Figure 1 shows [Mo/Fe] (where [Mo/Fe] = ${log}_{10}{(\mathrm{Mo}/\mathrm{Fe})}_{★}-{log}_{10}{(\mathrm{Mo}/\mathrm{Fe})}_{\odot }$) as a function of [Fe/H] for our sample of stars compared to the Mo abundances derived for giant stars in [14,15,16].

In the metallicity range of our targets ($-2.2<[\mathrm{Fe}/\mathrm{H}]<-1.0$), our sample show a mean [Mo/Fe] (〈[Mo/Fe]〉) of 0.44 dex and a standard deviation ($\sigma$) of 0.12 dex, while the samples of [14,15] show 〈[Mo/Fe]〉 = 0.25 dex with $\sigma =0.15$ dex and 〈[Mo/Fe]$\rangle =0.21$ dex with $\sigma =0.16$ dex respectively. Our results are in general consistent with those of previous studies, although the values of [Mo/Fe] in our sample are on average higher than the ones in other samples at similar metallicities. Our sample also shows a slightly lower dispersion in [Mo/Fe] abundance ratios ($\sigma =0.12$ dex) compared to other samples in the same metallicity range ($\sigma =0.15$ dex and $\sigma =0.16$ dex). This offset in the mean [Mo/Fe] abundance ratios could simply be an observational bias due to the limited number of stars in our sample. However, since the other studies presented in Figure 1 derived Mo abundances exclusively from the 386.4 nm line, it is also possible that this difference could be due to non-LTE (NLTE) effects affecting each Mo line differently. In stellar atmospheres with temperatures and metallicities similar to those of the stars in this sample ($4000<$${\mathrm{T}}_{\mathrm{eff}}$$<4500$ K, [Fe/H]∼$-1.5$), neutral molybdenum is a minority species around $log\left({\tau }_{Ross}\right)=-3$, where the contribution function of the lines here examined peaks. It may therefore suffer from some deviation from LTE. Although no dedicated NLTE calculations for Mo i are available, we would expect Mo i to behave like other neutral minority species in FGK-type atmospheres. Previous studies showed that low-excitation lines of Fe i and Cr i are subject to stronger deviations from LTE than the higher-excitation lines (see, e.g., [28,29,30]). The Mo i 386.4 nm line is a low-excitation line, while the other three lines (550.6 nm, 557.0 nm and 603.0 nm) originate from higher-excitation levels; therefore we would expect the lines to respond differently to departures from LTE. This is in line with the results of [14], who suggested that the observed difference in [Mo/Fe] abundance ratios between dwarfs and giants in their sample could be due to NLTE effects.

Figure 2 shows [Mo/Fe] as a function of [Fe/H] for our sample of stars, colour-coded according to the kinematic classification derived in [17]. As shown in Figure 2, the four Gaia-Sausage-Enceladus (GSE) candidates have Mo abundances slightly below the sample average, with $0.25<[\mathrm{Mo}/\mathrm{Fe}]<0.39$ dex. On the other hand, the only Sequoia candidate is enriched in Mo compared to the other stars in the sample, with [Mo/Fe] = 0.71 dex. However, given the small number of GSE and Sequoia candidates, it is not possible at present to draw definitive conclusions regarding Mo abundances in these substructures.

5. Conclusions

In this paper we have derived Mo abundances for a sample of 27 giant stars with $-2.2<[\mathrm{Fe}/\mathrm{H}]<-1.0$ presented in [17]. No Mo abundances are available in the literature for this sample of stars, so they are presented for the first time in this study.

The main conclusions of this study are the following:

• The Mo i absorption lines at 550.6 nm, 557.0 nm, and 603.0 nm have proven to be reliable in measuring Mo abundance in giant stars with [Fe/H] $>-2$. In particular, the 603.0 nm line is detectable down to [Fe/H] = $-1.8$, while the other lines are detectable down to [Fe/H] = $-2$.
• The general trend observed for [Mo/Fe] as a function of [Fe/H] is in good agreement with the results of previous studies, even if the derived values of [Mo/Fe] are on average higher than the ones in other samples at similar metallicity. We believe that this may be due either to an observational bias or to NLTE effects.
• Currently, there are no NLTE corrections available for Mo i lines in the literature. Our results emphasise the need for these corrections in order to better understand the nucleosynthesis and chemical evolution of this element.
• In our sample, stars that are GSE candidates show [Mo/Fe] marginally below the sample average, while the only Sequoia candidate star has a higher [Mo/Fe] than most stars in the sample. However, the limited number of stars prevents us from drawing firm conclusions about [Mo/Fe] trends in these substructures.

As part of the MINCE project, we will extend the abundance measurements of Mo to the other stars in our MINCE sample, including the stars analysed in MINCE I and II, in order to obtain a large sample of Mo abundances derived in a homogeneous way in giant metal-poor stars to compare with Galactic chemical evolution models. We also plan to derive the abundances of Ru, Pd, and Ag in this sample of stars in order to obtain a complete pattern of light neutron-capture elements from Sr to Ag. This will allow us to study the abundance ratios of these elements and shed light on the nucleosynthesis processes that form them.