For this study we have constructed a high resolution 2000 band spectral file. The first 1000 bands are 1 nm wide (0.9 nm for the first band) and cover the wavelength range 0.1–1000 nm. The number of sub-bands over this range is 13 799, providing the resolution used for photolysis which can be seen in the spectral plots in Sect. . Bands 1001–2000 have a resolution of 10 cm⁻¹ and cover the range 1000 nm–0.01 m. Note that 1 nm resolution is equal to 10 cm⁻¹ resolution at 1000 nm. The switch from wavelength to wavenumber resolution is done so that the entire spectrum can be covered in a practical number of bands. This wide range allows for complete coverage of stellar and thermal radiative transfer. However, for the calculation of the photolysis rates, only wavelengths less than 1100 nm were considered. The absorption coefficients (k-terms) are calculated from the relevant input cross sections. These cross sections are taken from the sources described in Sect. , and listed in Table , alongside the photolysis reactions and branching ratios we adopted. The number of k-terms varies up to a maximum of 22 per band for the major gases.

In this study we include species that are important for Earth's stratosphere and were also part of PhotoComp. These species fall broadly into the categories: Ox, NOx and HOx, and organic species relevant to Earth. These selected species were chosen to compare with the output of the radiative transfer code Fast-JX used for the Regional Air Quality (RAQ) mechanism . For the purpose of the intercomparison and as there are no sources of opacity in the infrared in our calculations, any contribution to photolysis longward of 1100 nm is neglected. The species in addition to those within PhotoComp that we have added for exoplanets align with the high temperature network of which is designed for hot hydrogen-dominated exoplanets, such as hot Jupiters as this is intended for future studies of these objects. Species such as H2O and hydrocarbon species such as CH4 also have relevance for early-Earth-like exoplanets.

We incorporate Rayleigh scattering from O2 and N2 (air) down to 175 nm. This limit coincides with the threshold for O2→O(3P)+O(1D) photolysis. It is assumed that most of the absorbed flux from shorter wavelengths is used for dissociation and the atmospheric regions where the absorption occurs will have a significant atomic oxygen and nitrogen content. Rather than formulate a separate scheme that includes O and N scattering we assume absorption will dominate below 175 nm and any Rayleigh scattering can be neglected.

The Chemistry-Climate Model Validation, Stratosphere-troposphere Processes And their Role in Climate Evaluation (SPARC) CCMVal-2, was a climate model intercomparison initiative, which included an element on the benchmarking of photolysis models, PhotoComp 2008. The results were produced in the subsequent report . Initially conducted in 2008 using JPL 2006 data , PhotoComp was repeated in 2011 with predominantly JPL 2010 data . The primary goal of this photolysis intercomparison was to evaluate how different models calculated the photolysis rates in the stratosphere and troposphere. Part 1a of their experimental set-up was used for the comparisons in this paper. This consists of a clear sky with no aerosols, a Solar Zenith Angle (SZA) of 15° over the ocean, an albedo of 0.10 (Lambertian), an incoming Solar irradiance at top-of-atmosphere of 1365 W m⁻² and the inclusion of Rayleigh scattering. The pressure-temperature profile used by PhotoComp, and adopted here is shown in Fig. . The PhotoComp study also included tests of the accuracy of the actinic flux calculations for different atmospheric compositions. However, the accuracy of the Socrates radiative transfer calculations has been extensively validated previously for both Earth and exoplanets , therefore we restrict our work here to benchmarking the photolysis rates only.

The two reference models from PhotoComp we compare with

Data provided by Martyn Chipperfield and retrieved from https://homepages.see.leeds.ac.uk/~lecmc/sparcj (last access: 27 July 2025).

As photolysis is driven by short-wavelength flux, species and bands that absorb UV and visible light have a direct impact on the resulting photolysis rates as they dictate the actinic flux. The gases O2 and O3 are the main absorbers in this regime, and their absorption cross sections are shown in Fig. . The major bands for O3 are the Hartley bands (200–300 nm), which predominantly absorb in the stratosphere, with additional absorption longward of 300 nm through, for example, the Huggins (∼300–370 nm) and Chappuis bands (∼370–790 nm). For O2, absorption is mainly via the Schumann Runge bands (175 to 205 nm) and continuum (130 to 175 nm), as well as absorption of Solar Lyman-α emission (121.45 to 121.7 nm).

Figure  shows the actinic flux, as calculated by Socrates using Eq. () and detailed in Sect. , at the top-of-atmosphere (TOA, solid blue line), upper mid-atmosphere (at a pressure of 1 Pa) which represents the top model level specified by PhotoComp as used by the UCI-ref model (green line), and the lower mid-atmosphere (at a pressure of ∼2400 Pa) corresponding to the ozone layer (magenta line). The dominant absorption feature between 220–290 nm in the lower/mid-atmosphere is due to ozone absorption within the Hartley bands.

Temperature dependent ozone cross-sections used in this work have been compiled based on recommendations from the JPL 19-5 report augmented by more recent HITRAN 2020 data for the Hartley and Huggins bands between 244–346 nm. For wavelengths 110–244 and 346–830 nm the JPL recommended cross-sections have been used from the MPI-UV/Vis database . The cross-sections are extended to 1100 nm using data from . Temperature dependent quantum yields are from following the JPL 19-5 recommendation using data at 6 temperatures from the MPI-UV/Vis database.

Figure  shows the rates calculated for two possible dissociation reactions for ozone, namely O3→O(3P)+O2, and O3→O(1D)+O2 as the left and right columns, respectively, and as functions of pressure and wavelength as the top and bottom rows, respectively. Note that O(3P) refers to the ground state of the atom and O(1D) the first excited state. The top row of Fig.  shows the rates from UCI-ref (dashed orange line) and this work (solid blue line), demonstrating excellent agreement apart from a slight difference towards the top-of-atmosphere (TOA). At the TOA the photolysis rate is independent of the model radiative transfer and is governed by a simple convolution of the absorption cross section, quantum yield and stellar spectrum. The UCI-ref model is based on JPL recommended cross-sections from 2010 while Socrates is using an updated temperature dependent cross-section from HITRAN 2020 . The Socrates model will also have finer sub-band resolution. These differences are likely to be the main cause of the observed discrepancy although without access to the UCI-ref model and data it cannot be reliably determined. The bottom row of Fig.  shows the rates at the TOA and ∼100 Pa which approximately corresponds to the point where the models begin to disagree. The Socrates rates were also recalculated without contributions from wavelengths shorter than 177 nm. This had a negligible impact on the results, indicating that the inclusion of Lyman-α emission is not significant.

For oxygen, the JPL 19-5 recommended cross-sections are used for wavelengths > 205 nm. Over the Schumann Runge bands (179.2–202.6 nm) we use high-resolution cross-sections from . In the Schumann Runge continuum and the Lyman-α region (115–179 nm) we use data from . At shorter wavelengths down to 0.04 nm cross-sections are compiled from the data listed in Table . The cross-sections used are independent of temperature and pressure.

For the O2→O(3P)+O(3P) reaction we take the quantum yield to be 1 from the threshold at 242.3 nm down to 175 nm. Below 175 nm we take the quantum yield to be 1 for the O2→O(3P)+O(1D) reaction apart from the region around Lyman-α (121.35–122 nm) where the quantum yield is partitioned according to . In the EUV shortward of 102 nm the quantum yield falls below 1 using data from with photoelectron enhancement effectively increasing the quantum yield shortward of 65 nm using data from .

Figure  shows the total dissociation rate for O2→O+O for both the UCI-ref (orange dashed line) and our work (solid blue line) as a function of pressure (Pa) in the top left panel, as well as the separate Socrates rates for the dissociations O2→O(3P)+O(1D) and O2→O(3P)+O(3P) (solid green and blue lines respectively) in the top right panel. In the bottom panels the rates are shown as a function of wavelength for the separate reactions. The top left panel of Fig.  again shows excellent agreement between the rates calculated using Socrates and the UCI-ref values. The divergence in the rates near the surface occurs for values less than 1×10-19 s⁻¹ and is likely due to the use of different O2 cross-sections within the absorption window at wavelengths around 200 nm. There is also a slight departure at very low pressures (∼10 Pa, or above ∼64 km) where the UCI-ref model is no longer valid as it does not include EUV wavelengths (see discussion in Sect. ). The top right panel shows that the O2→O(3P)+O(3P) reaction is the main contributor to the total dissociation rate, while the O2→O(3P)+O(1D) reaction only contributes for wavelengths below the threshold at 175 nm (spectrum, bottom left panel). Repeating this comparison while omitting flux at wavelengths < 177 nm reduces the low-pressure disparity between the Socrates and UCI-ref results to negligible levels (not shown).

Absorption cross-sections for hydrogen peroxide (H2O2) follow the JPL 19-5 recommendations between 190–350 nm. Between 260–350 nm this includes the calculation of temperature dependent cross-sections at 7 temperatures between 200 and 320 K using the formulation from . FUV cross-sections between 106–190 nm are taken from . Cross-sections between 353–410 nm are taken from . Data from both of these sources was obtained from the MPI-UV/Vis database and extend the JPL recommended data without significant discontinuities.

Figure  shows the photolysis rate for the reaction H2O2→OH+OH as a function of pressure (Pa). Photolysis spectra for this reaction can be seen in Fig.  (bottom row). Photolysis occurs below the threshold wavelength of 557 nm with a quantum yield of 1 recommended by JPL 19-5 for wavelengths > 230 nm. The quantum yield at 193 nm is recommended to be 0.85. We use a simple step down to this value at the mid-point wavelength of 211.5 nm.

Photolysis is most significant at wavelengths coinciding with the Hartley bands of ozone (220–290 nm). This leads to a distinct drop in the photolysis rate across the ozone layer in the mid-atmosphere where the actinic flux is correspondingly reduced (see Fig. ). In the lower atmosphere the variation in photolysis rate with height is predominantly due to the temperature dependence of the H2O2 absorption cross-section. The photolysis rate profile calculated with Socrates matches the UCI-ref profile well with a slight departure towards lower pressures. This difference is significantly reduced if a quantum yield of 1 is used in the Socrates calculations below 211.5 nm.

Figure  shows the photolysis rates for the reactions NO2→NO+O(3P), NO3→NO2+O(3P) as a function of pressure. Photolysis spectra for these reactions can be seen in Fig. .

NO2 cross-sections are based on the recommendations of the JPL 19-5 report between 238 nm and 667 nm using data from at 220 and 298 K. We use the original high-resolution cross-sections obtained from the MPI-UV/Vis database rather than the band means reported in JPL 19-5. We also extend the cross-sections into the FUV and EUV using the data reported in Table .

For the reaction NO2→NO+O(3P) we use the temperature dependent quantum yields recommended in JPL 19-5 which are 1 up to the dissociation threshold wavelength of 398 nm and then rapidly decrease to zero for wavelengths > 422 nm. The photolysis rates from Socrates and UCI-ref in Fig.  (top) generally match to within a few percent. The match is particularly good at pressures less than ∼103 Pa where there has yet to be any significant absorption of the actinic flux over the wavelength region > 300 nm where significant photolysis occurs. Note, the match is improved further if quantum yields for 298 K are used (blue dashed line) without the temperature dependence.

The overall shape of the profile is predominantly governed by the temperature dependent cross-sections which introduce variations that mirror the temperature structure of the atmosphere shown in Fig. . The Socrates rates are further affected by absorption of the actinic flux at pressures higher than ∼103 Pa and begin to decrease, whilst the UCI-ref values only show evidence of absorption from ∼3×104 Pa. This difference is likely due to the use of updated ozone absorption cross-sections in Socrates from HITRAN 2020 (see Table ). We also use high-resolution cross-section data for NO2 which may contribute to the differences as there is fine structure in the near-UV region where photolysis occurs .

For NO3 the recommended cross-sections from JPL 19-5 for 298 K are used without a temperature dependence. The JPL report notes there is uncertainty in the absolute cross-section with reported values ranging by around a factor of 2. Quantum yields for the NO3 photolysis reactions have a strong dependence on both wavelength and temperature. We have used the recommended values from JPL 19-5 including the temperature dependence.

For the reaction NO3→NO+O2, the second row of Fig. , the reference and Socrates calculated rates match to within around 5 %. The reference models show very little variation in photolysis rate with pressure and we note the match with Socrates is improved if we use the quantum yields for 298 K without a temperature dependence (blue dashed line). For the reaction NO3→NO2+O(3P), the third row of Fig. , the opposite is true. The Socrates rates match the reference to within ∼5 % when temperature dependent quantum yields are used even though the shape of the profile with pressure differs significantly. When the quantum yields for 298 K are used the offset is more significant. It is important to note that between 400–640 nm, where photolysis for this reaction occurs, there are no significant sources of absorption of actinic flux. The difference in the rates as a function of pressure between our calculations and that of the reference, mimics the pressure-temperature profile, Fig. , and is almost solely due to the temperature dependent quantum yield.

Figure  shows the photolysis rates for the reactions N2O→N2+O(1D) and NO→N(4S)+O(3P), with N(4S) being the ground state of the nitrogen atom.

For N2O we use the recommended cross-sections from JPL 19-5 including the temperature dependence between 173–240 nm from . We extend the cross-sections into the FUV and EUV using the data reported in Table .

Socrates photolysis rates for the reaction N2O→N2+O(1D) match the UCI-ref values closely at pressures higher than ∼10 Pa. However Socrates rates continue to increase towards lower pressures while the UCI-ref rates do not. The quantum yield for this reaction is taken to be 1 for all wavelengths below the threshold at 336 nm. High TOA rates at FUV wavelengths, particularly around Lyman-α, are displayed in the photolysis spectrum shown in Fig. . If we only consider wavelengths > 177 nm in the Socrates calculations the rates near the TOA are reduced and more closely match the reference.

The photoabsorption cross section of NO gas features fine band structures. Fluorescence occurs, except within the “delta” bands δ(0-0) and δ(1-0), which correspond to the wavelengths 189.4–191.6 and 181.3–183.5 nm respectively . It is in these narrow regions that photolysis occurs with a quantum yield of unity. These regions coincide with the O2 Schumann Runge bands where there is also fine structure in the O2 absorption spectrum and therefore in the acintic flux. For these reasons an accurate rotational line list is needed for NO and this was sourced from the line list “XABC”, from Exomol . Absorption cross-sections at high-resolution were determined from the line parameters with a pressure and temperature dependence based on Voigt line profiles.

The photolysis spectrum calculated by Socrates for the NO dissociation shows the narrow photolysis regions have been clearly resolved (see Fig. ). The rates calculated by Socrates match the reference values reasonably well, as shown in Fig.  with slightly higher values at pressures below 50 Pa and slightly lower values for higher pressures. Note that an NO mass mixing ratio was included in the calculation of the photolysis rates as discussed in Sect.  which contributes to absorption of the actinic flux in these narrow bands and acts to reduce the photolysis rates at higher pressures.

For N2O5 we use the recommended cross-sections from JPL 19-5 between 200–420 nm including the temperature dependence from between 260–410 nm. We extend the cross-sections at short wavelengths down to 152 nm using data from .

For the reaction N2O5→NO3+NO2 we use the quantum yields from JPL 19-5 and IUPAC which recommend a value of 1 above 300 nm then stepping down to 0.85, 0.79, 0.62 and 0.08 in wavelength bins centred at 289, 287, 266 and 248 nm respectively. Figure  shows the photolysis rates yielded from this reaction which match the UCI-ref values reasonably well in the lower atmosphere. The differences in the upper atmosphere are likely attributable to the treatment of the quantum yield below 300 nm. The actinic flux at these wavelengths is attenuated by ozone in the stratosphere and so does not affect the photolysis rates in the lower atmosphere. If we repeat the Socrates calculations using a quantum yield of 1 at all wavelengths the photolysis rates closely match the UCI-ref values throughout the profile.

For HONO we use the recommended cross-section from JPL 19-5 between 184–396 nm extended to 400 nm using the 0.5 nm resolution data of . The JPL 19-5 cross-section contains a gap between 274–296 nm which we fill using the 0.5 nm resolution data between 292–296 nm and an interpolation in the logarithm of the cross-sections between 274–292 nm. The quantum yield for the reaction HONO→OH+NO is taken to be unity following the JPL 19-5 recommendation. The Socrates photolysis rates shown in Fig.  are found to be ∼15 % lower than UCI-ref at TOA indicating differences in the HONO cross-section used between the models. Towards higher pressures the Socrates photolysis rates decrease due to absorption of actinic flux in the ozone Huggins bands above 300 nm. We use updated HITRAN 2020 ozone cross-sections in this region compared to UCI-ref which displays less absorption.

For HNO3 we use the JPL 19-5 recommended temperature dependent absorption cross-sections. Between 240–315 K we use the data measured by . We extend the temperature dependence to 200 K using the recommended temperature coefficients from . Quantum yields for the reaction HNO3→OH+NO2 are reported by JPL 19-5 without a specific recommendation so we adopt the recommended values from IUPAC of 0.97 above 248 nm, 0.9 between 200–248 nm and 0.33 below 200 nm. The resultant photolysis rates from Socrates in Fig.  are significantly lower than those of UCI-ref. We note that a quantum yield of 1 for wavelengths > 200 nm is consistent with the range of reported values in JPL 19-5. If we use a quantum yield of 1 for all wavelengths (blue dashed line) the photolysis rates match the UCI-ref values very well.

For HO2NO2 we use the JPL 19-5 recommended cross-sections between 190–350 nm in the UV. For this comparison we ignore photodissociation in the overtone and combination bands in the infra-red. For the reaction HO2NO2→OH+NO3 we use the JPL 19-5 recommended quantum yields of 0.3 at wavelengths < 200 nm and 0.2>200 nm. The photolysis rates shown in Fig.  are significantly lower than UCI-ref. However, if we again use a quantum yield of 1 at all wavelengths (blue dashed line) the rates match the UCI-ref values reasonably well, indicating the reference results may actually be for the total HO2NO2 photolysis rate rather than the particular channel specified for PhotoComp.

Figure  shows the photolysis rates for the reactions H2CO→H+HCO, H2CO→H2+CO and OCS→CO+S(3P), where S(3P) is the ground state of the sulphur atom.

For formaldehyde, H2CO, the Socrates photolysis rates are of similar magnitude to the UCI-ref values but differ in the shape of the profile, with a particular increase towards lower pressures not seen in the reference models. As formaldehyde features fine structure within its cross section as a function of wavelength, we have adopted high resolution data as indicated in Table . We have implemented quantum yields for standard pressure (1 atmosphere) and 300 K from the JPL 19-5 report . However, for both the reactions H2CO→H+HCO and H2CO→H2+CO the quantum yield has a strong dependence on both pressure and temperature which we have not taken into account as the functionality to incorporate a pressure dependence has not yet been included in Socrates. This is likely to be the main cause of the discrepancy with the UCI-ref model values.

Figure  shows that for carbonyl sulphide, OCS, the Socrates rates and those of the reference agree very well. We use the temperature dependent cross-sections recommended by JPL 19-5 together with a quantum yield of 1 for all wavelengths.

Figure  displays the photolysis rates for the remaining organic species considered, namely CH3OOH, CH3CHO, PAN, CH3COCH3, CHOCHO, CH3ONO2, CH3COCHO and HOCH2CHO as a function of pressure (Pa). For the first reaction pathway of CH3C(O)OONO2 (polyacrylonitrile, or PAN), the Socrates rates match the reference particularly well, whereas for the second PAN reaction and CH3COCH3 there is a significant discrepancy. The quantum yields adopted for our calculations of the rates for PAN and CH3COCH3 (based on the JPL recommendations) could be the source of this discrepancy. The quantum yields are temperature and pressure dependent for CH3COCH3, however we currently only have the functionality to represent the temperature dependence. To account for the pressure dependence we used an appropriate tropospheric pressure in the formulation to calculate the quantum yields for the four temperatures in the look-up table (see Table ). Future work is needed to properly incorporate the pressure dependence and this could be a contributing source of the difference with the reference model for CH3COCH3 as well as some other organic species (e.g. formaldehyde).

There are even more significant differences between our rates and those of the reference for other organic species, namely, CHOCHO (glyoxal), CH3ONO2 (methyl nitrate) and CH3COCHO (methylglyoxal). For the first reaction of CHOCHO our Socrates rates are higher than those of the reference, while for the second they are lower. The quantum yields for CHOCHO again have a significant pressure dependence that we have not included which is likely to be the main cause of the discrepancy. Similarly, the quantum yields for CH3COCHO have a pressure dependence that we have not included leading to the marked difference in shape between the photolysis rate profiles of Socrates and the reference models.

The rates calculated by Socrates for CH3ONO2 match the UCI-ref values well in the lower atmosphere but are significantly higher than the references within and above the ozone layer. The JPL 19-5 report does not provide recommended values of the quantum yield but reports conflicting values measured at particular wavelengths. We use a quantum yield of 1 for wavelengths > 248 nm, 0.91 for wavelengths 241–248 nm and 0.7 for wavelengths < 241 nm. However the chosen limits are fairly arbitrary and likely to be the cause of the discrepancy between the Socrates and UCI-ref photolysis rates.

The higher levels of FUV and EUV flux for Proxima Centauri, compared to the Solar spectrum, shown in Fig. , result in a greater availability of actinic flux to drive photolysis below ∼175 nm. Figure  shows the actinic flux at three different atmospheric pressure levels, namely the TOA (blue), ∼20 Pa (upper mid-atmosphere, green) and ∼2400 Pa (lower mid-atmosphere, corresponding to the location of the ozone layer, magenta line) for both the Solar and Proxima Centauri spectra as the top and bottom panels, respectively. Figure  shows that below 175 nm, there is about an order of magnitude higher actinic flux for the Proxima Centauri spectrum compared to the Solar case, whilst at wavelengths greater than 175 nm there is significantly more actinic flux from the Solar spectrum.

Figure  shows the photolysis rates for the reactions O3→O2+O(3P), O3→O2+O(1D), O2→O(3P)+O(3P) and O2→O(3P)+O(1D), on a log scale as a function of pressure (left panel) and as a function of wavelength for the Solar spectrum (middle panel) and Proxima Centauri spectrum (right panel). For ozone, O3, the Proxima Centauri photolysis rates are significantly lower than the Solar case due to the lower stellar irradiance in the region 200 - 300 nm coinciding with the strong Hartley absorption bands of ozone (compare Figs.  and ). Proxima Centauri is a much cooler star than the Sun with a spectrum that peaks further towards the red, with significantly more power in the visible than the near-UV. For the O3→O2+O(3P) reaction this means there is a larger relative contribution from the weak Chappuis absorption bands beyond 400 nm than from the Hartley bands. These weak bands do not have a significant effect on the actinic flux in the visible region and as a result the photolysis rates do not experience the sharp decline across the ozone layer that is seen with the Solar case. In contrast, the O3→O2+O(1D) reaction has a threshold wavelength of 411 nm so there is no contribution from the Chappuis bands. For both reactions there is a more significant contribution from Lyman-α wavelengths for Proxima Centauri. This contribution falls off quickly in the upper atmosphere due to absorption of the actinic flux by oxygen.

The Proxima Centauri rates for O2→O(3P)+O(1D) are much higher than Solar (by about an order of magnitude) in agreement with . All contributions to the rates for this reaction originate from wavelengths below the threshold at 175 nm where the Proxima Centauri actinic flux is greater. For O2→O(3P)+O(3P) the quantum yield is zero below 175 nm except for a small region around Lyman-α where both dissociation reactions occur. For the Solar case, the major contribution is from the Schumann Runge absorption bands at wavelengths > 175 nm while for Proxima Centauri there is a much larger contribution from Lyman-α wavelengths. This leads to approximately equal total photolysis rates for O2→O(3P)+O(3P) at TOA. However the rates for Proxima Centauri decrease much more rapidly towards higher pressures due to stronger attenuation of Lyman-α wavelengths.

As an example of photolysis of HOx species, we focus on the dissociations of H2O, and specifically the reactions: H2O→OH(X2Π)+H where OH(X2Π) is the ground state, H2O→O(1D)+H2, and H2O→O(3P)+H+H. Figure  shows the rates for these dissociations as yielded by the Solar and Proxima Centauri spectra (blue and red lines respectively) against pressure on a log scale (left column), and as a function of wavelength for the Solar case (middle column) and Proxima Centauri case (right column) at the TOA (blue) and ∼20 Pa (green). The dissociation H2O2→OH+OH is also displayed in Fig.  for reference.

The absorption spectrum of H2O in the FUV consists of a broad continuum centred around 165 nm reducing to a minimum in absorption around 145 nm with increasing and more structured absorption towards shorter wavelengths. Based on values reported in the JPL 19-5 report, the quantum yield for the reaction H2O→OH(X2Π)+H is taken to be 1 for the broad continuum beyond 147 nm, while the quantum yield for the reaction H2O→O(1D)+H2 rises to 0.11 at wavelengths shorter than 147 nm. The reaction H2O→O(3P)+H+H is taken to have a quantum yield of 0.11 below its threshold wavelength of 129 nm.

Towards TOA the Proxima Centauri rates for all the H2O reactions are an order of magnitude higher than those of the Solar case due to the strong contribution from the Lyman-α region where the Proxima Centauri stellar irradiance is higher. For the reactions H2O→O(1D)+H2, and H2O→O(3P)+H+H where all the photolysis occurs <147 nm, the photolysis rates reduce to effectively zero by the mid-atmosphere due to attenuation of the actinic flux.

For the reaction H2O→OH(X2Π)+H there is a significant contribution from wavelengths > 175 nm where the Solar irradiance is higher than Proxima Centauri. The Schumann Runge absorption bands of O2 decrease in strength towards longer wavelengths between 175–200 nm with less attenuation of the actinic flux allowing photolysis to occur much lower in the atmosphere. This explains the difference in the photolysis rate profiles where the Proxima Centauri rates are higher near TOA due to the contribution around Lyman-α while the Solar rates are higher in the lower atmosphere where the Lyman-α wavelengths have attenuated and the dominant contribution is from wavelengths beyond 175 nm.

For photolysis of the NOx species, Fig.  shows the rates for the dissociation of NO2→NO+O(3P), NO3→NO+O2, NO3→NO2+O(3P), N2O→N2+O(1D) and NO→N(4S)+O(3P) as yielded by the Solar and Proxima Centauri spectra (blue and red lines respectively) against pressure (Pa) on a log scale (left column), and as a function of wavelength for the Solar case (middle column) and Proxima Centauri case (right column). Figure  shows the single plots of the dissociation rates of NO2→NO+O(3P) and NO3→NO+O2 as yielded by the Proxima Centauri spectrum only (red line) on a linear scale against pressure (Pa).

For the reaction NO2→NO+O(3P) the photolysis rates are much lower at all pressures for the Proxima Centauri case. The top right panels show the photolysis rates at the TOA (blue), upper-mid atmosphere ∼100 Pa (green) and lower-mid atmosphere ∼32000 Pa (magenta) indicating the dominant contribution between 300–400 nm is significantly reduced for the Proxima Centauri case with a greater contribution from shorter wavelengths around Lyman-α at the TOA compared to the Solar case. For the Proxima Centauri case, oxygen absorption at these shorter wavelengths impacts the photolysis higher up in the atmosphere resulting in the steep decline of photolysis rates with increasing pressures that is not seen in the Solar case. Figure , top panel, shows a zoom-in on the Proxima Centauri profile for this case. The small increase in rates at ∼102 Pa coincides with a peak in the temperature dependence of the cross section, which comes into effect at the longer UV wavelengths.

Interestingly for the reaction NO3→NO+O2, in the Proxima Centauri case, Fig.  (bottom panel) shows the rates changing as a function of pressure in a noticeably different way to that of the Solar case (see Fig. ). The spectra of the rates as a function of pressure for wavelengths between 580-640 nm are shown at the TOA (blue), ∼140 Pa (green) and ∼7600 Pa (magenta) in the second row right panels of Fig. . As there is very little absorption at these wavelengths, the spectral changes for different pressures are almost entirely due to the temperature dependence of the quantum yield. The TOA and 7600 Pa lines are essentially on top of each other because the temperature is about the same at these levels. The difference in the pressure dependence of the Solar and Proxima Centauri rates is due to the wavelength variation of this temperature dependence. Essentially, for shorter wavelength flux the quantum yield decreases as the temperature increases. Whereas, towards longer wavelengths the quantum yield increases as the temperature increases. As it is a lower temperature star, Proxima Centauri has a higher fraction of its flux at longer wavelengths and rates are therefore increased as the temperature increases. Whereas, for the Solar case, there is a larger fraction of the flux at shorter wavelengths leading to cancellation of the overall temperature dependence and a more muted effect on the shape of the photolysis rate profile.

Figure  shows the comparison of Solar and Proxima Centauri rates for the remaining NOx species. These rates are dominated by wavelengths > 200 nm and are therefore significantly higher in the Solar case.

Figure  shows rates for the dissociation of formaldehyde, H2CO→H+HCO (top row), H2CO→H2+CO (middle row) and OCS→CO+S(3P) (bottom row) as yielded by the Solar and Proxima Centauri spectra (blue and red lines respectively) against pressure on a log scale (left column), and as a function of wavelength for the Solar case (middle column) and Proxima Centauri case (right column) at the TOA (blue) and ∼320 Pa (green).

When examining H2CO, the photolysis rates are significantly lower for Proxima Centauri due to the lower contribution from near-UV (NUV, ∼200–400 nm) fluxes. The higher contribution in the FUV compared to the Solar case leads to the observed difference in the shape of the rate profiles. Oxygen absorption of FUV fluxes in the upper atmosphere leads to a sharp decrease in rates with pressure for Proxima Centauri, similar to behaviour displayed in the NO2 case in Fig. . The rates as a function of wavelength shown in the right panels of Fig.  display a similar structure to the corresponding data for NO2.

The photolysis rates for OCS as a function of wavelength, bottom row of Fig. , show contributions from wavelengths greater than 180 nm where the Solar spectrum is stronger than Proxima Centauri. Also evident in these spectra is the effect of the noisy structure of the Proxima Centauri irradiance spectrum in the NUV leading to the commensurate noise in the photolysis spectrum across this range, particularly around ∼200–210 nm.

Figure  shows the calculated rates as a function of pressure for the reaction CH3→CH2(1)+H (top row) and four dissociation rates of CH4. Note that CH2(1) is the methylene group where (1) refers to the excited singlet state. The methyl radical, CH3, has extremely limited data available (see Appendix , Table ), and only covered one band centred on 215.5 nm with zero rates elsewhere.

Figures  and  display the dissociation rates of the same organic species displayed in Fig.  for the Solar and Proxima Centauri spectra. The photolysis rates as produced by the Proxima Centauri spectrum for the species CH3CHO, CHOCHO, CH3COCH3, HOCH2CHO all display similar trends to that of OCS (see Fig. ).

Figure  shows the rates for photolysis reactions of C2H2, C2H3 and C2H4 as yielded by the Solar and Proxima Centauri spectra. For both C2H2 and C2H4, the photolysis rates for Proxima Centauri are higher than Solar at the top of atmosphere due to the greater contribution from Lyman-α wavelengths. At lower altitudes the contribution from the NUV dominates and Solar rates are higher than for Proxima Centauri.

Figure  shows the rates for photolysis reactions of C2H6 as yielded by the Solar and Proxima Centauri spectra. These reactions are dominated by wavelengths around Lyman-α and are correspondingly higher for Proxima Centauri.

In addition to the species already explored, some additional species are required for exoplanets where comparison rates under an Earth-like O3 profile are not available in PhotoComp, including H2O which is detailed in Sect. . Therefore, in this section we simply provide our calculated rates as a reference for future studies. Figure  shows the rates for the dissociation of CO→C+O(3P), CO2→CO+O(1D), CO2→CO+O(3P), HCN→H+CN and NH3→NH2+H as yielded by the Solar and Proxima Centauri spectra (blue and red lines respectively) against pressure on a log scale (left column), and as a function of wavelength for the Solar case (middle column) and Proxima Centauri case (right column) at the TOA (blue) and ∼20 Pa (green).

The cross sections of CO2 have a temperature dependence (see Table ) and the effect of this is evident for CO2→CO+O(3P) where we see a protrusion indicating an increase in rates around ∼100 Pa. The Proxima Centauri rates for CO2→CO+O(1D) are higher than the Solar rates due to the contribution around Lyman-α. The short wavelengths are attenuated before arriving at ∼100 Pa which is why we do not see a similar peak. The threshold for the production of O(1D) is 167 nm and the quantum yield is zero below 50 nm, therefore the flux supplied in the relevant wavelength range would be higher for Proxima Centauri than that provided by the Solar spectrum. Similar reasoning can be applied to CO while HCN and NH3 both have contributions from the NUV where the Solar flux is larger. For HCN the Lyman-α and NUV rates are balanced with Lyman-α dominating towards the top of the atmosphere and the NUV dominating below ∼100 Pa (similar to C2H2 and C2H4).

For ammonia (NH3) similar to the case of H2CO and NO2, as detailed in Sect.  and , the longer UV wavelength contribution is smaller for Proxima Centauri than when the Solar spectrum is used. For the rates calculated with the Proxima Centauri spectrum the shape of the rates as a function of pressure is mainly due to oxygen absorption of FUV flux at the top of the atmosphere, particularly around Lyman-α. The photolysis rates calculated with the Solar spectrum are dominated by the NUV where ozone absorption lower in the atmosphere is the most significant factor.