Near-infrared observations of active asteroid (3200) Phaethon reveal no evidence for hydration

The study investigates whether the surface of the active near-Earth asteroid (3200) Phaethon shows evidence for hydration, addressing the broader question of what drives its observed activity (e.g., dust tail near perihelion). Phaethon, an Apollo-type NEA and parent of the Geminid meteor stream, develops a short-lived dust tail post-perihelion. Water-ice sublimation is considered unlikely due to the asteroid’s very small perihelion distance and high surface temperatures that render ice unstable. Phaethon is classified as a B-type asteroid (SMASS), similar in visible–NIR spectral behavior to other B-types such as (2) Pallas and (101955) Bennu, many of which show hydration signatures via strong 3-µm absorptions. Given Phaethon’s extreme thermal environment, hydrated materials could be destabilized. The primary objective is to acquire rotationally resolved spectra beyond 2.5 µm to search for the diagnostic 3-µm hydration band across most of Phaethon’s surface, thereby testing hypotheses for its current activity (volatile sublimation or phyllosilicate devolatilization) and assessing compositional links to the Pallas family.

Prior work established Phaethon’s rotational period (~3.604 h), size (~5.1 ± 0.2 km effective diameter), variable reported albedo (0.08–0.16), and B-type classification. Observations show Phaethon becomes weakly active near perihelion with a dust tail for a few days; ice sublimation is unlikely given thermal conditions. Dynamical studies indicate similarities to asteroid 2005 UD, suggesting both could be fragments of a primitive precursor. Other B-types, including Pallas and Bennu, commonly exhibit 3-µm absorptions linked to hydrated minerals. Bennu, though spectrally similar to Phaethon in 0.5–2.5 µm, has a lower albedo and clear hydration features, and is weakly active. A possible Pallas–Phaethon link has been proposed based on spectral slope, high inclination, and albedo; however, Pallas shows a deep 3-µm band unlike Phaethon’s results in this work. Lazzarin et al. reported a 0.43-µm absorption on Phaethon attributed to hydration, which this study evaluates in the context of 3-µm findings.

Observations were conducted with the SpeX spectrograph/imager in long-wavelength cross-dispersed (LXD) mode (1.9–4.2 µm) at the NASA Infrared Telescope Facility (IRTF) on 12 December 2017. Ten rotationally resolved LXD data sets (a–j) were obtained, sampling most of the surface visible during the apparition (primarily northern hemisphere and equatorial regions; aspect angle ~53°), effectively covering a full rotation. Data reduction followed established SpeX procedures: division by internal flat fields; A–B beam differencing to remove OH emission and sky thermal background; residual background subtraction; spectral extraction via summed flux within an aperture; division by a solar analog star (SAO 39985) observed at similar airmass to correct telluric absorptions; wavelength calibration using argon lines for λ < 2.5 µm and telluric lines for λ > 2.5 µm. Reduction employed Spextool (IDL, v4.0). Because spectra beyond 2.5 µm include significant thermal emission, thermal excess was modeled and removed using the Near-Earth Asteroid Thermal Model (NEATM), adopting visible geometric albedo pv = 0.122 ± 0.008 (Hanuš et al.) and slope parameter G = 0.06 (Ansdell et al.). The beaming parameter η was varied between 1.35 and 1.65 to best fit each set’s thermal excess while keeping pv fixed; emissivity (bolometric and spectral) was set to 0.9. A K/V reflectance scaling of 0.7 was applied to reconcile V-band albedo usage with K-band reflectance. The thermal model was subtracted from measured spectra to yield thermally corrected reflectance. The 3-µm band depth at 2.90 µm (D290) was computed as D290 = (Rc − R290)/Rc, where Rc is the continuum reflectance defined by setting reflectance at 2.45 µm to 1.0 with zero slope; uncertainties were propagated using δD290 = D290 * sqrt[(δRc/Rc)^2 + (δR290/(Rc − R290))^2], with δ terms derived from the propagated variances output by Spextool. Observational circumstances (airmass, magnitude, sub-Earth latitude, rotational phase) were recorded for each set; sets i and j had higher airmass and lower S/N.

- All thermally corrected LXD spectra (sets a–j) are featureless in the 3-µm region; no hydrated mineral absorption is detected within 2σ at 2.90 µm across the observed surface portions. - Band depths at 2.90 µm (D290) with 1σ uncertainties and beaming parameter used (from Table 1): • Set a: η=1.50, D290 = −1.90% ± 4.40% • Set b: η=1.50, D290 = −5.15% ± 9.97% • Set c: η=1.50, D290 = −5.00% ± 4.11% • Set d: η=1.45, D290 = −2.06% ± 1.27% • Set e: η=1.45, D290 = −5.43% ± 4.04% • Set f: η=1.35, D290 = +3.55% ± 4.48% • Set g: η=1.65, D290 = −2.68% ± 3.12% • Set h: η=1.45, D290 = +5.53% ± 7.74% • Set i: η=1.45, D290 = +0.72% ± 1.86% • Set j: η=1.45, D290 = +8.52% ± 9.60% - Observations covered virtually a full rotation, sampling most of the northern hemisphere and equatorial regions; only the area near the southern pole (~10% of the surface) was not observed. - Sets i and j were acquired at high airmass (1.849 and 2.130) and thus have lower S/N. - The lack of a 3-µm band contrasts with many B-types (e.g., Pallas, Bennu) that show strong hydration features; Phaethon’s spectrum remains blue-sloped in 0.5–2.5 µm but lacks the 3-µm absorption.

The absence of a detectable 3-µm absorption across the observed surface indicates that Phaethon’s surface (tens of microns depth) is anhydrous within measurement limits. This result contradicts earlier claims of hydration based on a 0.43 µm absorption feature, though hydration limited to the unobserved southern polar region (~10% of the surface) cannot be excluded. Surface uniformity in both NIR slope (from prior work) and lack of 3-µm features suggests substantial thermal processing and homogenization, potentially due to repeated intense heating near perihelion, resurfacing processes (e.g., ballistic redistribution), and non-uniform mass loss (~10^5 kg per orbit). Thermal history modeling indicates that different hemispheres have faced the Sun at perihelion over the past few thousand years, implying that the entire surface has experienced similar extreme conditions and could have been uniformly dehydrated. Comparisons to Pallas and Bennu show similar visible–NIR slopes and, for Phaethon, a higher albedo consistent with Pallas-family B-types; however, unlike Pallas’s deep 3-µm band (~20%), Phaethon lacks hydration features, consistent with surface dehydration at peak temperatures exceeding 1000 K—well above dehydration thresholds for serpentines and temperatures that cause volatile loss in carbonaceous materials (>500 K). Consequently, current activity is unlikely to be driven by volatile sublimation or phyllosilicate devolatilization; mechanisms such as thermal degradation or solar wind sweeping are more plausible. Whether Phaethon originated hydrated and subsequently dehydrated, or formed from intrinsically anhydrous material, remains unresolved; in situ observations (e.g., DESTINY+ flyby) may discriminate between these scenarios and clarify potential links to Bennu and the Pallas family.

Rotationally resolved 1.9–4.2 µm spectra of (3200) Phaethon show no evidence of a 3-µm hydration band across the observed northern hemisphere and equatorial regions. This indicates an anhydrous surface layer and argues against current activity being driven by volatile sublimation or phyllosilicate devolatilization. The findings are consistent with extensive surface dehydration due to extreme perihelion heating and support a compositional connection to the Pallas family despite the absence of hydration features seen on Pallas. Phaethon may have been originally hydrated and later dehydrated, or may have formed from anhydrous material; distinguishing these possibilities will likely require in situ investigation. Future work should include high-S/N coverage of the southern polar region and spacecraft observations (e.g., JAXA’s DESTINY+) to assess any localized hydration and better constrain Phaethon’s origin and activity mechanisms.

- Observational coverage excluded a region near the southern pole (~10% of the surface), so localized hydration there cannot be ruled out. - Later observation sets (i and j) were at higher airmass and lower signal-to-noise, potentially reducing sensitivity to weak features. - Thermal modeling involves degeneracy between beaming parameter and albedo, though parameter choices were constrained and do not materially affect the lack of detected absorption. - Results probe only the upper tens of microns of the surface; deeper subsurface hydration, if present, would not be detected.