Hydrogen Peroxide on the Surface of Europa

See allHide authors and affiliations

Science  26 Mar 1999:
Vol. 283, Issue 5410, pp. 2062-2064
DOI: 10.1126/science.283.5410.2062


Spatially resolved infrared and ultraviolet wavelength spectra of Europa's leading, anti-jovian quadrant observed from the Galileo spacecraft show absorption features resulting from hydrogen peroxide. Comparisons with laboratory measurements indicate surface hydrogen peroxide concentrations of about 0.13 percent, by number, relative to water ice. The inferred abundance is consistent with radiolytic production of hydrogen peroxide by intense energetic particle bombardment and demonstrates that Europa's surface chemistry is dominated by radiolysis.

The composition of a planetary surface is an important indicator of its evolution and subsequent chemical alteration. Europa's surface composition can be modified by extrusion of material from the interior (1), the infall of cometary and meteoritic material, photochemical processes, and by the deposition of material from the magnetosphere (2,3). Europa is subjected to intense bombardment by jovian magnetospheric particles—energetic electrons, protons, sulfur ions, and oxygen ions (4)—that could alter the composition through radiolysis (5, 6). The relative importance of these chemical alteration processes has not been established for Europa.

Spectra of Europa indicate a water-ice surface (7) with sulfur dioxide (SO2) (2, 8) and hydrated minerals (9). The SO2 could be produced by sulfate decomposition (10) or from implanted sulfur ions (2). The hydrated minerals may be evaporite salts (9), from brine extruded from a hypothetical subsurface ocean (11). Infrared (IR) spectra of Europa obtained by Galileo's near-infrared mapping spectrometer (NIMS) (12) showed absorption features (13) at wavelengths of 4.25, 4.03, and 3.50 μm. Carbon dioxide and SO2 were identified (13) as the 4.25- and 4.03-μm absorbers, similar to previous findings for Ganymede and Callisto (14). We show here that the 3.50-μm feature, corroborated with laboratory measurements and Galileo ultraviolet spectrometer (UVS) (15) data, indicates the presence of hydrogen peroxide (H2O2), formed in this environment by energetic plasma irradiation of Europa's surface (6).

NIMS reflectance spectra (Fig. 1A) of Europa's leading anti-jovian quadrant (16) show characteristic features of water frost and a feature at 3.50 μm that is not due to H2O. Hydrocarbons and ammonium-containing minerals were suggested to account for this feature (13), but such identifications are problematic because their absorption wavelengths do not match that observed, and both classes of compounds exhibit additional, stronger absorption features not evident in the NIMS spectra. For example, methanol (17) shows an absorption band at 3.53 μm and four other strong bands at 3.38, 3.35, 3.14, and 3.04 μm that are not apparent in Europa's spectrum (Fig. 1A). Similar arguments rule out other simple hydrocarbons and ammonium-bearing minerals (18). A more likely candidate for the 3.50-μm feature is H2O2, which produces 2ν2, ν2,6, and 2ν6 combination-bending-mode absorption (19, 20) at ∼3.5 μm and has been predicted to occur on icy satellite surfaces (6).

Figure 1

(A) Representative reflectance factor spectrum of Europa's leading, anti-jovian quadrant. The feature at 3.50 μm, identified as H2O2 absorption, is shown in the inset as the ratio to the interpolated continuum (the reflectance factor if H2O2 were absent). All other structure is due to water; the peak at 3.1 μm is a H2O restrahlen reflection feature. This spectrum is an average of nine individual spectra, covering 105°W to 125°W, 22°S to 2°N. The 3.50-μm band is found from <100°W to >140°W. (B) Laboratory diffuse reflectance spectrum (in percent) of a 0.5% frozen solution of H2O2 in water. Combination-bending transitions of H2O2 appear at 3.504 μm. No water absorption band occurs at this wavelength, as shown in the pure water ice curve (rescaled). Sample temperatures were 80 K.

For comparison to the NIMS data, we measured (21) the diffuse reflectance of H2O2 mixtures in water ice and found a band at 3.504 μm (Fig. 1B). Previous measurements (22) of UV-photolyzed, 10 K water ice showed the H2O2 feature at 3.509 μm, and it shifts to 3.505 μm at 70 K. These wavelengths are consistent with the Europa feature (3.50 ± 0.015 μm). The widths of the Europa and laboratory feature are also consistent, both being ∼0.06 μm wide (full width at half maximum) (Fig. 1). With the exception of this feature, frozen aqueous H2O2 solutions produce near-IR spectra that are indistinguishable from those of pure water ice.

An estimate of the surface concentration of H2O2 is obtained by laboratory IR-reflectance measurements of frozen H2O2-H2O solutions at various concentrations (21). Comparing the relative band depths of these spectra with the NIMS spectrum, we find a concentration on Europa of ∼0.13 ± 0.07% (by number of H2O2 molecules, relative to H2O; this convention is used throughout) (21). The estimate pertains to Europa's 3.5-μm-wavelength, optically sensed surface layer, which is limited by water spectral absorption to depths of about the ice-grain size (23), ignoring porosity. The grain size (∼60 μm) was determined with the observed IR reflectance factor (24) and radiative-transfer calculations for ice-grain surfaces (25).

Condensed H2O2 in Europa's surface may exist as a solid-state solution in ice or as crystals of pure H2O2 or H2O2·2H2O (20). However, the wavelength (3.52 μm) of the crystalline phases (at 4 and 80 K) is larger than the wavelength of Europa's feature and outside the range allowed by the estimated wavelength uncertainty. Pure amorphous H2O2 is precluded because its absorption band occurs at 3.56 μm (20).

Hydrogen peroxide absorbs UV radiation (26), so if our identification is correct we expect (6) a corresponding absorption signature to be present in Galileo UVS spectra (27). Disk-resolved UVS measurements (obtained concurrently with the NIMS observations) were compared with the diffuse reflectance of a frozen 0.16% solution of H2O2 in water (28). The spectra (Fig. 2) are consistent, both showing the onset of absorption at about 300 nm and exhibiting similar shapes at shorter wavelengths. The similarity of these spectra supports the IR identification of H2O2 on Europa's surface. However, numerous materials exhibit broad-band UV absorption spectra, so the UV measurements alone do not provide unique identification.

Figure 2

Europa UV reflectance factor compared with laboratory diffuse reflectance measurements and theoretical values. The Galileo UVS spectrum (•) is an average for longitudes 100°W to 140°W and was obtained simultaneously with the NIMS observations. The laboratory spectra (solid lines) are for a 0.16% H2O2 solution and a pure water-ice sample, both at 80 K. Each was measured relative to an aluminum mirror and collectively normalized to Europa data at long wavelengths. Theoretical reflectance factors (dashed lines) were computed for the IR-derived H2O2 concentration (0.13 ± 0.07%) and grain size (60 μm).

The concentration used for the laboratory measurements is about the same as the Europa value, but the grain size, which also affects the reflectivity, was not determined for this laboratory sample, so we cannot independently deduce H2O2 concentrations from these UV data. Instead, we used the IR-derived grain size and H2O2 UV absorption cross sections (26, 29) to compute theoretical reflectance factor spectra (25) for various concentrations of H2O2. Comparing these computed reflectance factors with the Europa UV spectrum (Fig. 2) shows that the UVS observations are consistent with the IR-derived abundance for wavelengths less than 250 nm. At longer wavelengths, the observed reflectance is higher than the theoretical values; this could be due to the use of 296 K, liquid-phase absorption coefficients in the calculations, which may be inaccurate for H2O2in low-temperature ice.

Production of H2O2 on Europa may be initiated through dissociation of surface water molecules into H + OH by the incident energetic plasma (6). Diffusion and chemical combination of two OH radicals yields H2O2. Reactive scattering of H and OH by water to form H2O2 also may occur (30). The energetic particle's energy flux (31) is Φ ≈ 5 × 1010 keV s−1 cm−2, and each 100 eV of energy will produce G ≈ 0.4 H2O2 molecules, on the basis of results for α-particles (32). Solar UV radiation may also dissociate H2O, but the average dissociating UV energy flux is <1% of that from the energetic particles (33). With the above parameters, the production rate of H2O2 isG Φ ≈ 2 × 1011 molecules s−1 cm−2. The incident particles can also destroy H2O2 molecules, either by direct dissociation or by production of H, which then reacts with H2O2 (34). We estimate the rate of destruction of H2O2 by the energetic particles using the flux (31) ϕ ≈ 1.2 × 108s−1 cm−2 and an assumed cross section of σ = 1 × 10−15 cm2, which is the molecular size of H2O2. This gives a loss rate of σ ϕ ≈ 1.2 × 10−7 s−1. Absorption of solar UV radiation, particularly in the 210- to 300-nm region, dissociates H2O2 molecules with a diurnally averaged rate of 2.5 × 10−6 s−1 (gas phase value) (33) and is effective to depths >250 μm (23). We ignore the recombination of the spatially contained dissociation products (the cage effect), which can reduce photon-induced loss by a factor of >3 (35). The net loss rate will therefore be (2.6 to 0.12) ×10−6 s−1 (lifetimes ∼ 4 days and ∼100 days, respectively), implying surface densities from (0.08 to 1.7) × 1018 cm−2. The vertical stopping distance for energetic electrons (at 60° incidence angle) is about 180 μm (31), and less for ions, so the average H2O2 concentration in this ∼180-μm-deep radiolysis layer will be (0.4 to 9) × 1019cm−3, or 0.013 to 0.3% relative to water ice. These estimated limits and our derived concentration of 0.13% are consistent.

The abundance of H2O2, and the existence of a Na and O2 atmosphere (3,36), thought to be produced by energetic-particle bombardment of the surface (3, 10,36), demonstrate that surface chemistry on Europa is dominated by radiolysis. At the energy influx rates used above, the optically sensed surface layer can be completely modified in a few tens of years. Temporal changes in the jovian magnetospheric energetic plasma can alter Europa's H2O2 abundance, thereby changing Europa's 210- to 300-nm reflectance; such changes have been observed in Europa's UV reflectance (37). Because CO2 and SO2 are present, themselves the possible products of radiolysis (10), related products such as H2SO4 and various carbon compounds should be investigated. Predictions, characterization, and identifications of surface species on Europa must consider radiolysis effects.

  • * To whom correspondence should be addressed at Mail Stop 183-601, Jet Propulsion Laboratory, 4800 Oak Grove Drive, Pasadena, CA 91109, USA. E-mail: rcarlson{at}


View Abstract

Navigate This Article