Stephenson 2 DFK 1, also known as RSGC2-01 or St2-18, is a red supergiant (RSG) or possible extreme red hypergiant (RHG) star in the constellation of Scutum. It lies near the open cluster Stephenson 2, which is located about 5.8 kiloparsecs (19,000 light-years) away from Earth in the Scutum–Centaurus Arm of the Milky Way galaxy, and is assumed to be one of a group of stars at a similar distance, although some studies consider it to be an unrelated or foreground red supergiant.

The open cluster Stephenson 2 was discovered by American astronomer Charles Bruce Stephenson in 1990 in the data obtained by a deep infrared survey. The cluster is also known as RSGC2, one of several massive open clusters in Scutum, each containing multiple red supergiants.

The brightest star in the region of the cluster was given the identifier 1 in the first analysis of cluster member properties. However, it was not considered to be a member of Stephenson 2 due to its outlying position, abnormally high brightness, and slightly atypical proper motion, instead being categorized as an unrelated red supergiant.

In a later study, the same star was given the number 18 and assigned to an outlying group of stars called Stephenson 2 SW, assumed to be at a similar distance to the core cluster. The designation St2-18 (short for Stephenson 2-18) is often used for the star, following the numbering from Deguchi (2010). To avoid confusion from using the same number for different stars and different numbers for the same star, designations from Davies (2007) are often given a prefix of DFK or D, for example Stephenson 2 DFK 1 or simply D1 where the context is clear.

In 2012, Stephenson 2 DFK 1, along with 56 other red supergiants, was observed in a study regarding the maser emissions from red supergiants across the galaxy. The study derived the properties of those red supergiants using the Australia Telescope Compact Array (ATCA) and the DUSTY model. Stephenson 2 DFK 1 was among the red supergiants mentioned. That same year, it was observed again for a study regarding the types of masers on red supergiant stars in clusters. During 2013, in a study regarding the red supergiants in Stephenson 2, Stephenson 2 DFK 1 (referred to as D1) was observed. In several later studies, the star was described as being a "very late-type red supergiant."

It was also noted in Humphreys et al. (2020), albeit mistakenly referred to as RSGC1-01, another very large and luminous red supergiant in the constellation of Scutum.

When the cluster was originally discovered in 1990, Stephenson 2, and therefore Stephenson 2 DFK 1, was originally estimated to have a distance of around 30 kiloparsecs (98,000 light-years), much further than the cluster is thought to reside today.

A study in 2007 determined a kinematic distance of 5.83+1.91
−0.78 kiloparsecs (19000+6200
−2500 light-years) from comparison with the cluster's radial velocity, considerably closer than the original distance quoted by Stephenson (1990). However, because of Stephenson 2 DFK 1's doubtful membership, its distance was not directly estimated. This value was later adopted in a recent study of the cluster.

A similar kinematic distance of 5.5 kiloparsecs (18,000 light-years) was reported in a 2010 study, derived from the average radial velocity of four of the cluster's members (96 kilometers per second) and from an association with a clump of stars near Stephenson 2, Stephenson 2 SW, locating it near the Scutum–Centaurus Arm of the Milky Way. This value was later adopted in a 2012 study to calculate the star's luminosity. It is noted that the uncertainty in the distance was greater than 50%. Despite this, it is also stated that distances to massive star clusters will be improved in the future.

Verheyen et al. (2013) used the average radial velocity of the cluster (+109.3 ± 0.7 kilometers per second) to derive a kinematic distance of roughly six kiloparsecs (20,000 light-years) for the cluster. However, Stephenson 2 DFK 1's radial velocity is calculated to be only 89 kilometers per second and therefore leading to the study's statement that the star is a field red supergiant unassociated with the cluster.

Stephenson 2 DFK 1 is usually classified as a red supergiant, like the other stars in the cluster. However certain enigmatic properties, such as the star's significant infrared excess, have led the authors of Davies (2007) to state that it might be an extreme red hypergiant, much like VY Canis Majoris. It is also stated that Stephenson 2 DFK 1 is on the brink of ejecting its outer layers and evolving into a luminous blue variable (LBV) or Wolf–Rayet star (WR star).

The first calculation of the star's luminosity was published in 2010. It assumes membership of the Stephenson 2 cluster at 5.5 kpc and it is based on 12 and 25 μm flux densities, giving a relatively modest luminosity of 90,000 L_☉. Two years later, a new calculation for finding the bolometric luminosity by fitting the Spectral Energy Distribution (SED) using the DUSTY model gave the star a very high luminosity of nearly 440,000 L_☉.

The most recent calculation, based on SED integration (based on published fluxes) and assuming a distance of 5.8 kpc, gives a bolometric luminosity of 630,000 L_☉. It has been noted that Stephenson 2 DFK 1's SED is peculiar, and cannot be fitted to standard reddening laws. This would imply that the star has higher extinction and that it is actually more luminous, casting doubt on its membership. As stated in a 2012 study, the stellar association is spread over a large area, with Stephenson 2 blending into its immediate surroundings.

An effective temperature of 3,200 K was calculated in a 2012 study by SED integration using the DUSTY model, which would make it much cooler than the coolest red supergiants predicted by stellar evolutionary theory (typically around 3,500 K). However, this effective temperature is unlikely since one beyond the Hayashi line could indicate that it is not in hydrostatic equilibrium.

In 2007, Davies et al. estimated Stephenson 2 DFK 1's spectral type at M5 or M6, unusual and very late for even a red supergiant star, based on its CO-bandhead absorption. Negueruela et al. (2013) identified Stephenson 2 DFK 1's spectral type to be around M6, similar to the spectral type approximated by Davies et al. 2007, based on its spectrum and the characteristics of certain spectral features, such as titanium oxide (TiO) spectral lines.

A radius of 2,150 R_☉ (1.50×10 km; 10.0 au; 930,000,000 mi) was derived from a bolometric luminosity of nearly 440,000 L_☉ and an estimated effective temperature of 3,200 K, which is considerably larger than theoretical models of the largest red supergiants predicted by stellar evolutionary theory (around 1,500 R_☉). Due to the unreliable unlikely temperature and the luminosity used, the estimate is likely unreliable.

Stephenson 2 DFK 1 has been estimated to have a mass loss rate of roughly 1.35×10 M_☉ per year, which is among the highest known for any red supergiant star. It is possible that Stephenson 2 DFK 1 underwent an extreme mass loss episode recently due to its significant infrared excess. In 2013, an article describing the red supergiants in Stephenson 2 stated that Stephenson 2 DFK 1 (referred to as D1) and D2 (another member of Stephenson 2) have maser emissions indicating that they have the highest mass loss in the cluster. Only the stars with the highest bolometric luminosities in the cluster seem to present maser emissions. Stephenson 2 DFK 1 displays strong silicate emission especially at wavelengths of 10 μm and 18 μm. Water masers were detected in the star as well.

It has been debated for a while if this star is actually part of its supposed cluster. Due to its radial velocity being below the other cluster stars but with some signs of membership, some sources state that the star is unlikely to be a foreground giant; however, more recent papers considered the star an unlikely member due to its extreme and inconsistent properties.

Using radial velocities determined from silicon oxide (SiO) maser emission and IR CO absorption, a study of red supergiant masers in massive clusters considered Stephenson 2 DFK 1 as a field red supergiant, unrelated to Stephenson 2. This is due to its lower radial velocity that is significantly different compared to other stars from Stephenson 2.

Another possibility is that Stephenson 2 DFK 1 is actually a member, because its radial velocity is offset by an expanding optically thick envelope. The velocity difference between this star’s radial velocity and Stephenson 2 itself (20 kilometers per second) is a typical outflow speed for red supergiants. One study suggests that Stephenson 2 DFK 1 is part of a cluster related to Stephenson 2, Stephenson 2 SW, which is assumed to be at the same distance as the core cluster itself. This proposed cluster contains several other massive stars and red supergiants, including Stephenson 2 DFK 49.

The distance of Stephenson 2 DFK 1 has been stated to have a relative uncertainty greater than 50%, and the star's doubtful membership, uncertain distance and differing radial velocities compared to the rest of the stars in Stephenson 2 have led to some authors to consider the star as a red supergiant unrelated to Stephenson 2 or any of the red supergiant clusters at the base of the Scutum–Centaurus Arm. The radius of 2,150 R_☉ could possibly be an overestimation due to the method used and the fact that the limit for star size predicted by stellar evolutionary theory is estimated to only be roughly 1,500 R_☉. Luminosity estimates for the star are uncertain as well, as another estimate of the luminosity gave a value of 90,000 L_☉, however, it relies on a narrow range of fluxes, making it a likely underestimate.