Warm Circumstellar Disk Gas Surrounding the 1.4 Gyr Old F2V Star HD 109085?

Unlike most stars with Kuiper Belt-like dusty debris disks which are gas-poor, some young (<100 Myr) A-type stars possess warm (T ∼ 1000–2500 K) circumstellar gas located within a few astronomical unit. The prototype example is the 20 Myr old A6V star β Pictoris, whose warm circumstellar disk gas was first detected using the Ca ii–K absorption line at 3934 Å. Subsequent observations revealed appreciable variability in the Ca ii–K absorption profile with weak features appearing on a nightly (and hourly) basis. This variable absorption behavior has now been well-modeled by invoking the evaporation of star-grazing planetesimals (exocomets) passing through our line-of-sight to the star (Beust et al. 1990). Recently over 20 other young (<150 Myr) A-type systems have been discovered that exhibit similar Ca ii absorption variability (Welsh & Montgomery 2015). However, no star of spectral type later than A6 or older than ∼400 Myr has yet been observed to exhibit similar Ca i–K absorption variability associated with circumstellar gas disk. We present Ca ii–K absorption spectra of HD 109085 (η Crv), which is a nearby (d = 18.3 pc), old (1.4 Gyr) F2V star surrounded by known dust belts, to determine if any circumstellar gas activity is present.

We observed HD 109085 on four occasions using the Sandiford spectrograph on the 2.1 m telescope at the McDonald Observatory, Texas on the nights of 2017 June 5th and 6th and 2018 April 27th and 29th at a resolution of R ∼ 65,000 (4.5 km s⁻¹). The data were extracted using standard techniques (Welsh & Montgomery 2015) with the resultant residual intensity profiles shown in Figure 1.

Zoom In Zoom Out Reset image size

Each profile shows weak absorption spanning the −15 to +15 km s⁻¹ velocity interval. This range encompasses both the expected velocity of any (variable) circumstellar absorption (i.e., V_rad = +0.7 km s⁻¹) and any sight-line contribution from passage through the Gemini Local Interstellar Cloud (projected velocity V ∼ −0.1 km s⁻¹) (Redfield & Linsky 2008). We assume that the absorption seen in each spectrum at V ∼ 0 km s⁻¹ is of interstellar origin and should be non-varying. A profile fit to this component in the 2017 June 5th spectrum results in a best-fit component at V = 0 km s⁻¹ with a column density of log N = 10.42 ± 0.08 cm⁻² and a Doppler parameter of b = 2.0 km s⁻¹, values similar to those found previously for the local ISM (Redfield & Linsky 2004). This interstellar component has been overplotted on the 4 spectra shown in Figure 1. By eye this component accounts for the majority of absorption spanning the −15 to +15 km s⁻¹ region in all but the 2018 April 29th spectrum.

We measure the equivalent width (EW_TOT) of each absorption profile over the −15 to +15 km s⁻¹ range, together with their measurement errors derived from the rms data point errors and the continuum placement error, adopted as the 1-σ error (Vallerga et al. 1993). We subtract the absorption contribution due to the local ISM which spans ∼−10 to +10 km s⁻¹ range. This contribution (as measured on 2017 June 5th) corresponds to an EW_ISM = 2.3 ± 0.4 mÅ, a value adopted for each of the other 3 observations. The residual equivalent width values, (EW_RES), for all 4 observations (i.e., EW_TOT–EW_ISM) are 0.6 ± 0.6 mÅ, 1.3 ± 0.6 mÅ, 0.4 ± 06 mÅ and 3.2 ± 1.1 mÅ respectively. The average residual for the first three nights observations is 0.8 ± 0.6 mÅ, a value which is essentially zero, indicating we have not observed absorption in excess of the interstellar contribution. However, the value of E_RES = 3.2 ± 1.1 mÅ derived for the 2018 April 29th observation (corresponding to a column density change of ∼log N = 10.52 ± 0.12 cm⁻² over a period of 48 hr across the entire line profile) is a change in EW at a statistical level of 2.9σ. This, (unfortunately), is only just below the widely accepted 3σ test for a result of confidence.

Our observations suggests we may have detected additional (circumstellar) absorption on the night of 2018 April 29th that was not present on the other three nights. This variable absorbing gas would be of a circumstellar origin and be located much closer to the central star than the cold dusty gas residing in the circumstellar debris belts. We note that no detectable Ca ii K-line absorption has been previously reported in observations of HD 109085 (Rebollido et al. 2018). Inspection of Figure 7 of that paper shows the HD 109085 data to be of a very low signal-to-noise ratio, thus accounting for their non-detection of the (expectedly) weak Ca ii absorption. Our present observations would be the first example of an F-type star, and of a star with an age >1 Gyr, that has an associated circumstellar disk of warm and ionized gas, that may be being replenished with evaporating exocomet material. If these swarms of exocomets contain ices and organic material then any terrestrial exoplanets in orbit around HD 109085 may still be being impacted and potentially seeded with life generating material.