High Temperature Photochemistry in the Atmosphere of HD189733b
M. R. Line, M. C. Liang, Y. L. Yung
Recent infrared spectroscopy of hot exoplanets has begun to open up the door to their
atmospheric composition. Atmospheric composition is controlled by thermochemical
equilibrium deep in the planetary atmosphere. The impact of photochemistry occurs
higher up in the atmosphere at levels above ~1 bar.
These two chemistries compete
between ~1-10 bars in hot Jupiter-like atmospheres, depending on the strength of the
eddy mixing and temperature. HD189733b provides an excellent laboratory to study the
consequences of this hot atmospheric chemistry. The recent spectra of HD189733b and
HD209458b show signatures of CH4, CO2, CO and H2O. Here we identify the primary
chemical pathways that govern the abundances of CH4, CO2, CO and H2O in the context
of thermochemical equilibrium chemistry and photochemistry and how these chemistries
interact.
Our results suggest that these species can be photochemically enhanced above
or below the thermochemical equilibrium value, and that some caution must be taken
when assuming a strictly thermochemical equilibrium atmosphere.
1
Introduction
2
Of the several hundred exoplanets discovered thus far, only a small number are transiting hot-
3
Jupiters of which we can obtain limited spectral information. A variety of species have been
4
detected including atomic species like sodium (Na) (Charbonneau et al. 2002), and atomic
5
hydrogen (Vidal-Madjar et al. 2003) as well as molecular species including CO, CO2, H2O and
6
CH4 (Tinetti et al. 2007, Swain et al. 2009/2010). The detection of these species allows us to
7
begin to explore the chemistries that produce the observed abundances. The presence of these
1
8
detected species suggests hydrocarbon chemistry via CH4 photolysis as well as oxygen and water
9
chemistries.
10
The primary chemistries that determine chemical abundances in our own solar system are
11
thermoequilibrium chemistry and photochemistry. Current atmospheric modeling of hot Jupiter
12
atmospheres typically assume thermochemical equilibrium abundances (Sharp & Burrows 2006,
13
Showman et al. 2009, Showman et al. 2008, Burrows et al. 1997, Donovan et al. 2010, Rodgers
14
et al. 2009, Fortney et al. 2009). Photochemistry or other disequilibrium mechanisms, such as
15
quenching, have not received the same attention (with the exception of Cooper & Showman
16
2006, Zahnle et al. 2009, Liang et al. 2003/2004). Thermoequilibrium chemistry occurs in high
17
temperature and pressure regimes where chemical timescales are fast, typically deep down in the
18
atmospheres of giant planets (~1000 bars).
19
thermodynamic properties of compounds involved in the system via the minimization of the
20
Gibbs free energy (Yung & DeMore 1989). Photochemistry is a disequilibrium process driven
21
by UV radiation from the host star. Photochemistry should be especially important on these
22
close in highly irradiated giant planets (Liang et al. 2003).
Abundances are determined solely by the
23
Liang et al. (2003) was the first to explore the photochemistry that may occur on close in
24
giant planets through modeling the sources of atomic hydrogen in HD209458b. However, in that
25
investigation low temperature rate coefficients were used which are not suitable for these high
26
temperature regimes, and several key reactions governing the production and loss of H2O, the
27
most abundant observed species, were not included.
28
GCMs better estimates of temperature profiles and vertical mixing profiles can be obtained.
Additionally, with the use of modern
2
29
Zahnle et al. (2010) explored products of sulfur photochemistry and how they may be
30
responsible for the strong UV absorbers that cause thermal inversions as well as the formation of
31
hydrocarbon soot. There has been, however, no detection of sulfur species on these hot Jupiters.
32
The goal of this paper is to understand the chemistry that produces the observed
33
abundances of ~10-4, ~10-6, ~10-4, and ~10-7 for CO, CO2, H2O and CH4, respectively, as
34
detected in the atmosphere of HD189733b (Swain et al. 2009) through the combination of both
35
photochemical and thermochemical models and to compare the differences between the two.
36
Furthermore, It has been recently suggested by Madhusudhan & Seager (2009) that there may be
37
as much as 700 ppm of CO2 present, however, the discrepancy between this value and the value
38
from Swain et al. (2009) value has yet to be resolved. This discrepancy suggests that there is
39
much degeneracy in retrieving temperature and mixing ratio profile, and that the exact values of
40
the mixing ratios of the detected species are not well known.
41
important mechanisms that govern the abundance of these detected species using HD189733b as
42
an example.
In this study we identify the
43
44
45
46
47
48
Modeling
We use both a thermochemical and photochemical model to explain the observed
49
abundances of CO, CO2, H2O and CH4 in the atmosphere of HD189733b.
50
understand the effects that temperature and eddy mixing have on the photochemically derived
51
mixing ratios. We shall adopt a hot profile representative of dayside temperatures and cool
We want to
3
52
profile representative of night side temperatures for 30 N from Showman et al. (2009) (Figure 1).
53
We assumed isothermal profiles above the upper boundary of the Showman et al. (2009) GCM
54
for simplicity. These two profiles appear to have a thermal inversion near 1 mbar with a day-
55
night contrast of ~500 K. The use of two T-P profiles will give some indication of the day/night
56
contrast of the modeled species. Though HD189733b is not expected to have an inversion, we
57
still choose these T-P profiles because the bracket most other profiles in the literature (Fortney et
58
al. 2006, Tinetti et al. 2007, Burrows et al. 2008) and the existence of an inversion does not
59
significantly affect the major chemical pathways.
60
In order to determine the thermoequilibrium abundances we use the Chemical
61
Equilibrium with Applications model developed by Gordon & McBride (1994).
62
abundances at the appropriate lower boundary (explained later) will be used for our lower mixing
63
ratio boundary condition in the photochemical model. Thermochemical calculations require only
64
pressure and temperature along with the relative molar mixing ratios of the atomic species
65
involved in the compounds of interest, in this case C, O and H (no N or S because they have not
66
yet been detected).
67
abundance of these species ([C]/[H]~4.4×10-4, [O]/[H]~7.4×10-4, where [i] denotes the
68
concentration of species i (Yung & DeMore 1999 pg. 112)). The thermochemical model
69
computes the abundances of all possible compounds formed by those atomic species via a Gibbs
70
Free energy minimization routine (Gordon & McBride 1994.). We compute the equilibrium
71
abundances at each pressure-temperature level for our chosen temperature profiles. We would
72
expect to see thermochemical equilibrium abundances in an atmosphere that is not undergoing
73
any dynamical or photochemical processes, or where chemical timescales are much faster than
74
any disequilibrium timescales (Cooper & Showman 2006, Prinn & Barshay 1977, Smith 1998).
These
For simplicity, and lack of any other information, we assume solar
4
75
We use the Caltech/JPL-KINETICS 1D photochemical model to compute the
76
photochemical abundances of our species of interest (Allen et al. 1991, Moses et al. 2005,
77
Gladstone et al. 1996, Yung et al. 1984) for HD189733b. HD189733b is in a 2.2 day period
78
orbiting at 0.03 AU around a K2V star. We use the UV stellar spectrum from HD22049 which is
79
also a K2V star (Segura et al. 2003). The model computes the abundances for 32 species
80
involving H, C and O in 258 reactions including 41 photolysis reactions and includes both
81
molecular and eddy diffusion. The model uses the same hydrocarbon and oxygen chemistry as
82
in Liang et al. (2003, 2004) but with high temperature rate coefficients for the key reactions
83
involved in the production and loss of H, CH4, CO2, CO, OH and H2O. We have also added two
84
key reactions involved in the destruction of H2O and CO2 (R254 and R255 in table 1). We have
85
not however, added a complete suit of reactions as to achieve thermochemical equilibrium
86
kinetically (e.g., see Moses et al. 2009), however, we do not expect this to change our results for
87
we have included the key chemical pathways that govern the production and loss of our species
88
of interest. The model atmosphere for the photochemical model uses the two temperature profiles
89
described above. The lower boundary of the photochemical model is important in determining
90
the mixing ratios throughout the atmosphere. We will estimate this lower boundary using
91
quench level arguments rather than arbitrarily choosing some level. For more details on quench
92
level estimation we refer the reader to Cooper & Showman (2009) and Smith (1999).
93
Eddy and molecular diffusion are key parameters determining the distribution of the
94
abundances in the atmosphere. Eddy diffusion is the primary vertical transport mechanism in
95
our 1D model.
96
species become chemically quenched, and thus defines the lower boundary conditions for the
The strength of vertical mixing will determine where in the atmosphere the
5
97
photochemical model (Prinn & Barshay 1977, Smith 1997). Following Prinn & Barshay (1977),
98
The transport timescale is given by
99
$$$ all equation numbers should be at end and front of line $$$
100
 trans 
L2
Kz
101
(1)
102
where L is a vertical length scale typically chosen to be the scale height and Kz is the eddy
103
diffusion coefficient. The chemicalloss timescale of species i is given by
104
105
 chem,i 
(2)
[i]
Li
106
107

where [i] is the concentration of species
i and Li is the loss rate of species i, typically determined
108
by the bottleneck reaction. The quench level for species i is defined where τtrans = τchem,i . For
109
levels where τtrans < τchem,i the mixing ratio of species i is fixed at the quench level value. For
110
levels below the quench level, the compounds reach chemical equilibrium.
111
In order to determine the quench level in HD189733b’s atmosphere we must first
112
estimate the strength of eddy mixing and the timescale for the conversion of CO to CH 4 (Griffith
113
& Yelle, 1999 Prinn & Barshay, 1977). The adopted eddy diffusion profile in this model is
114
derived from a globally root mean square averaged vertical wind profile from a GCM (Showman
115
private communication) and is estimated by
116
117
K z ~ wL
(3)
118

6
119
where w is the root mean square of the vertical wind velocity. Smith (1997) suggests that the
120
appropriate length scale is some fraction of the scale height. Here we assume that it is the scale
121
height thus giving us the upper limit on eddy diffusion. The vertical winds range from 0 (at
122
~200 bars) to 7 m/s (~0.8 mbars). The vertical wind is assumed to be constant above this height.
123
Combining this with a typical scale height of ~200 km gives an eddy diffusion of ~1010 cm2 s-1
124
(Figure 1). Typical transport timescales from (1) are on the order of ~105 s.
125
126
The rate-limiting step in the conversion of CO to CH4, and thus the reaction determining
the chemical lifetime of CO, is
127
128
H  H2CO +M CH3O +M
(4)
129
130
(Yung et al. 1988, Griffith 
& Yelle 1999, Cooper & Showman 2006). The rate coefficient in
131
reaction 4 has not been measured in the lab, however it’s reverse reaction has been measured to
132
be
133
134
kr  1.4 106 T 1.2e7800/T
(5)
135
136
(Page et al. 1989) (in unitsof cm6s-1) where T temperature at which the reaction takes place.
137
The forward reaction rate, kf can be estimated via
138
(6)
139
where Keq is the equilibrium constant for the net thermochemical reaction (Yung et al. 1988)
kr
 K eq  e
(G f Gr )/ RT

140
141
kf
H  H2CO CH3O
(7)

7
142
143
and Gf and Gr are the Gibbs free energies of the reaction, given, respectively by
144
H[H]+H[H2CO]-T(S[H]+S[H2CO]) and H[CH3O]-TS[CH3O] with H[X] being the enthalpy of
145
formation of species X and S[X] being the entropy of species X. The enthalpies and entropies of
146
the given species can be found in Yung & DeMore (1989) pg 58. With the relevant
147
thermochemical data and equations (5) and (6) we can estimate the forward reaction rate of
148
reaction (4) to be
149
150
k f  5.77 1012 T 1.2e 3327/T
(8)
151
152
 then be determined with
The CO chemical lifetime can
153
154
 chem ~
(9)
[CO]
k f [H][H 2CO]
155

156
157
where the concentrations of CO, H and H2CO are determined via the thermochemical model.
158
Upon equating (9) with (1) using the dayside temperature profile we determine the quench level,
159
and thus the lower boundary to be ~3 bars (~1530 K) which is similar to the results of Cooper &
160
Showman (2006) for HD209458b.
161
Jupiter (~100 bars) (Prinn and Barshay 1979) and is similar to that of brown dwarfs (~6 bars)
162
(Griffith & Yelle 1999).
163
suggested by Smith et al (1999) can drop the quench level deeper. Using a length scale of H/10
164
instead of H drops the quench level to ~8 bars, but there is very little change in the
$$$ ??? This pressure level is much higher than that of
Choosing a different length scale lesser than the scale height as
8
165
thermochemical mixing ratios from ~3 bars (Figure 2). Additionally there is no significant
166
difference in quench level between the nightside and dayside because the two T-P profiles
167
converge near the quench level.
168
We assume a zero concentration gradient at the lower boundary so as to allow
169
photochemical products to sink down into the deeper atmosphere except for the observed species
170
of CO, H2O, CH4, CO2. For these species we fix the mixing ratios to be the thermochemially-
171
derived values at the ~3 bar quench level: 8.41×10-4, 6.36×10-4, 4.09×10-5, and 1.96×10-7,
172
respectively for the dayside and 8.39×10-4, 6.38×10-4, 4.25×10-5, and 1.98×10-7, respectively for
173
the nightside. We assume a zero flux boundary condition for the top of the atmosphere e.g, little
174
or no atmospheric escape, though this assumption may not be entirely true for atomic hydrogen
175
(Vidal-Madjar et al. 2003), however this does not effect the results significantly. $$$ should test
176
the escape boundary condition
177
178
Results
179
180
Thermochemical Results
The thermochemically derived mixing ratios (relative to H2) are shown in figure 2.
181
Again, these are the mixing ratios we would expect if there were no dynamical or photochemical
182
process occurring in the atmosphere, which we know not to be true. If we focus firstly, on the
183
dayside profiles we can see that CO is the dominant carbon bearing species and remains
184
relatively constant with altitude as do H2O and CO2. We also notice that CH4 falls off rapidly
185
with increasing altitude (decreasing pressure). We can understand this result by noting that CO,
186
CH4 and H2 abundances are related through the net thermochemcial reactions
187
9
188
CH4  H2O CO +3H2
(10)
189
190
(11)

CO  H2O CO2 +H2
191
192
and through Le Chateliersprinciple which states that as the total partial pressure of the
193
atmosphere decreases, the system will want to resist that decrease in order to maintain
194
equilibrium by producing more molecules (smaller molecules), which in this case results in the
195
production CO and H2. Upon comparing the daytime profiles to the nighttime cooler profile we
196
notice that CH4 becomes more abundant.
CH4 is more energetically favorable at lower
197
temperatures and is much more sensitive to the effects of temperature than CO and CO2. We
198
also note that atomic hydrogen is more abundant at warmer temperatures than at cooler
199
temperatures due to the entropy term in the Gibbs free energy.
From a thermochemical
200
perspective, we can expect ~10 mbar mixing ratios of the observable species, CO, H2O, CH4 and
201
CO2 to range from ,respectively, (2-9)×10-4, (6-13) ×10-4, (2.6-6758) ×10-7 , (4.7-16) ×10-7, due
202
to the day/night contrast. For comparison, the measured dayside emission values from Swain et
203
al (2009) for CO, H2O, CH4 and CO2 are respectively, ~10-4, ~10-4, ~10-7, and ~10-6
204
205
Photochemical Results
206
We run 4 cases of our photochemical model (figure 3) in order to compare the effects of
207
temperature and photolysis verse no photolysis on the mixing ratios (relative to H2) for H, CO,
208
H2O, CO2 and CH4 . In the following subsections we will discuss the important reactions
209
governing the production and loss of each of the relevant species.
210
10
211
H2O, OH, H
212
The primary reactions that govern the production and loss of H2O are
213
214
R71
H2O + hv  H + OH
215
R137
OH + H2  H2O + H
216
217
R254
H + H2O  OH + H2
218
R137 and R254 are fast enough that they readily recycle each other in such a way that the
219
abundance of H2O remains relatively constant with altitude at the quench level value of
220
~6.36×10-4 until the homopause at ~10 nbar. The photolysis of H2O does not significantly effect
221
its abundance in the observable atmosphere as can be seen in Figure 3 because the loss timescale
222
of H2O due to photolysis everywhere longer than the transport timescale thus allowing recently
223
photolyzed parcels to be readily replenished by upwelling. The photolysis of H2O, does, however
224
produce the important OH and H radicals that drive the remainder of the chemistry (Figure 4),
225
with the net result being the conversion of H2 to 2H.
226
H2O photodissociates into OH and H with photons shorter than 2398 Å. For HD189733b
227
228
229
below this wavelength there are ~8×1015 photons cm-2 s-1 available for H2O photolysis. For
comparison, the UV flux below this wavelength at Jupiter is ~3×1014 photons cm-2 s-1 and for
HD209458b, ~3×1018 photons cm-2 s-1.
OH and H increase with increasing altitude due to the
230
availability of more UV photons. The production of H at high altitudes via H2O photolysis may
231
be the driver of hydrodynamic escape on hot-Jupiters (Liang et al. 2003).
232
In short, the abundance of H2O is primarily set by the thermochemical equilibrium value
233
at the lower boundary condition, taken here to be the quench level, an the rapidly decreases
11
234
above the homopause. If the quench level changes, the observable value of H2O will change but
235
not significantly as can be seen in figure 2. The derived value here is slightly higher than the
236
237
Swain et al. 2009 observations of (0.1-1) ×10-4 but is more consistent with the value obtained by
Tinetti et al. 2007b of ~5×10-4. The day to night contrast is nearly unnoticeable in figure 3.
238
239
CO & CO2
240
Thermochemically, CO is the dominant carbon reservoir in hot atmospheres above ~10 bars
241
(Figure 2).
242
The abundance of CO is set by the quench level thermoequilibrium abundance of
8.4×10-4 until the homopause. The abundance of CO2 is determined via the interconversion of
243
oxygen from the large reservoirs of CO and H2O into CO2 via the OH radical. Deeper down in
244
the atmosphere, say, below the quench level, or in the presence of weak vertical transport (low
245
eddy diffusion), oxygen is moved into CO2 via the following reactions
246
247
R137
OH + H2  H2O + H
248
R152
OH + CO  CO2 + H
249
R254
H + H2O  OH + H2
250
251
R255
H + CO2  OH + CO
252
With R152 being the reaction that gives the oxygen from H2O via OH and the oxygen from CO
253
to CO2. There is no net production or loss of species from these reactions, meaning they will
254
assume thermochemical equilibrium. Assuming steady state, these 4 reactions can be combined
255
to give the thermochemical mixing ratio of CO2 in terms of the rate constants (k) and mixing
256
ratios (f) of the large reservoirs of CO and H2O
12
257
258
(12)
fCO2 ~
1.85 107 T1.5e 5542/T f H 2O fCO
259
260
261
k152k254
f H O fCO
k137k255 2

 the mixing ratio of CO2 in the absence of any disequilibrium
This relation would determine
262
mechanisms such as photochemistry or quenching. Using the thermochemical mixing ratios of
263
264
H2O (~6×10-4) and CO (~9×10-4) and evaluating the rate constants at the daytime temperature
(T~1200 K) we obtain a CO2 mixing ratio of ~4×10-7 which is consistent with figure 2.
265
In the photochemical limit (in the absence of eddy mixing), the photolysis reactions, R71
266
and R75 become more important and effectively replace R254 and R255 giving the important
267
chain of reactions
268
269
R137
OH + H2  H2O + H
270
R152
OH + CO  CO2 + H
271
R71
H2O + hv  H + OH
272
R75/76
CO2 + hv  CO + O/O(1D)
Net
OH+ H2  3H+O
273
274
275
276
Combining these reactions allows us to estimate the photochemical mixing ratio of CO2 with
277
278
(13)
fCO2 ~

k152J71
f H O fCO
k137J7576 2
13
 0.062T 0.1e1910/T
279
J71
J7576
f H 2O fCO
280
281
282
283
 of the indicated photolysis reaction. As an extreme case we can
where J is the photolysis rate
use the top of atmosphere photolysis rate of H2O to be ~10-5s-1 and the photolysis rate of CO2 to
be ~5×10-8 s-1 and a dayside temperature of ~1200 K giving fCO2 ~few × 10-5. Eq 13 suggests
284
that the abundance of CO2 becomes photochemically enhanced rather than destroyed. The
285
abundance of CO2 in the presence of only quenching (no photochemistry) will remain fairly
286
constant up until the homopause at ~1 nbar (Figure 3).
This is due to the lack of excess OH
287
produced in R71 used to drive R152 causing the chemical timescales involved to be much larger
288
than the eddy transport timescale. Again, for comparison, the observed mixing ratio of CO2 from
289
Swain et al. ( 2009) is ~10-7-10-6.
290
291
CH4 and Heavier Hydrocarbons
292
The primary fate of CH4 in the upper atmosphere is reaction with H to produce CH3, which
293
immediately reacts with H2 to restore CH4,
294
295
R28
CH4 + H  CH3 + H2
296
R53
CH3 + H2  CH4 + H2
297
298
The result is a do-nothing cycle. However, the above recycling is not perfect, and the following
299
sequence of reactions occur in the upper atmosphere
300
301
R28
2 ×[ CH4 + H CH3 + H2 ]
14
302
R4
2× [ CH3 + hv  CH2 + H ]
303
R48
CH2 + CH2  C2H2 + 2H
Net
2CH4  C2H2 + 2H2 + 2H
304
305
306
307
308
The net result is production of C2H2 in the upper atmosphere at the ~1ppm level. No other C2
309
hydrocarbons are produced in significant quantities. The primary fate of C2H2 from the upper
310
atmosphere is downward transport, followed by hydrogenation back to CH4. Additionally, CH4
311
can be destroyed by its reaction with the photochemically produced OH
312
313
OH + CH4  H2O + CH3
R141
314
315
The abundance of CH4 is ~4×10-5 which is several order of magnitudes larger then the ~10-7
316
detected by Swain et al (2009) and predicted by Liang et al. (2003).
317
318
319
Discussion
We have analyzed the important disequilibrium mechanisms, photochemistry and vertical
320
quenching, that govern the vertical distribution of species in hot Jupiter atmospheres.
321
derived abundances are consistent with the observations of Swain et al. (2009) with the
322
exception of methane. We obtained a value of ~4×10-5 where as the observations suggest two
323
orders of magnitude less. The observed value of ~10-7 corresponds to the thermochemical
324
equilibrium value at ~10 mbars. This would mean the quench level would have to be at this
325
level, suggesting an eddy diffusion on the order of ~103 cm2 s-1 from equations (9) and (1).
Our
15
326
However, Recent observations of HD209458b suggest that the methane mixing ratio may be as
327
high as ~10-5 which is more consistent with our model (Swain et al. 2009b).
328
may be possible that the value observed value may be probing the ~1 nbar level (Figure 3). Our
329
value of methane is also several orders of magnitude larger than the Liang et al. (2003) for
330
HD209458b. This is because the temperature at the lower boundary used in Liang et al. (2003)
331
for HD209458b is ~700 K hotter than our lower boundary temperature of ~1530 K and methane
332
is less stable at higher temperatures.
Alternatively, it
333
The metallicity of these hot jupiters is not well constrained, though we assumed solar
334
metallicity, we can still explore what might happen if this is not the case. Changes in metallicity
335
will affect the thermoequilibrium abundances. This will in turn change the lower boundary
336
mixing ratios. We varied the metallicity (taken here to be ([C]+[O])/[H]) from 0.1 solar up to 10
337
solar to see what effect it would have on our lower boundary mixing ratios (Figure 6). The
338
thermochemical 3 bar mixing ratio of CO, H2O and CO2 vary by several orders of magnitude
339
over the range of metallicities, where as CH4 changes very little. These orders of magnitude
340
change at the lower boundary due to metallicity will affect our photochemical results by that
341
same amount. With 10x solar metallicity we could expect mixing ratios of CO and H 2O to be as
342
high as ~0.1 and CO2 as high as 10-5. CO2 is more readily affected by metallicity than the other
343
species because it has two oxygen’s as opposed to CO’s, one oxygen.
344
The [C]/[O] ratio, rather than just the metallicity also can play a role in changing our
345
results. Here we vary the [C]/[O] ratio from 0.1 to 10 times the solar ratio of ~0.6 while keeping
346
the overall metallicity ([C]+[O])/[H]) constant at the solar value (Figure 6). The 3 bar mixing
347
ratio of CO does not vary significantly, but can get as high as ~10-3 given a slightly super solar
348
[C]/[O] ratio. CO2 rapidly decreases for ratios above solar and can get as low as 0.1 ppb for 10
16
349
times the solar ratio. As the [C]/[O] ratio increases past 1, H2O and CH4 swap roles in taking up
350
H and can change as much as 3 orders of magnitude.
351
There appears to be some compositional variability between the night and dayside,
352
though not much. Comparing the solid curves in the top of figure 3 to the dashed curves in the
353
bottom of figure 3 gives some sense of the magnitude of the day-night variability. There are no
354
photons on the night side so the quench level mixing solely determines the abundance throughout
355
the rest of the atmosphere.
356
factor of ~3 more CH4 on the night side over the dayside and up to a factor of 2 more CO2 on the
357
dayside.
358
photochemical reactions that only occur on the dayside (CH4 gets destroyed due to R141 and R4,
359
CO2 enhanced via equation 13). C2H2 would exhibit incredible variability since it is produced
360
strictly from photochemistry. We could expect to see up to 1 ppm on the dayside of these Hot
361
planets with very minute amounts on the nightside where it would be thermochemically recycled
362
back to methane.
There is a less than 1% maximum variability in CO and H2O, a
CO2 and CH4 experience more variability because they are most affected by
363
364
Conclusions
365
We have shown that both photochemistry and vertical quenching can significantly alter the
366
abundances of CO2, CH4 and C2H2 in hot Jupiter atmospheres. Vertical quenching determines
367
the lower boundary values and thus the mixing ratios of CO and H2O, which are not significantly
368
effected photochemically. CO2 can be significantly photochemically enhanced via equation (13),
369
and CH4 can be readily photochemically destroyed. These ideas can extend to other hot Jupiter
370
atmospheres, though we used HD189733b as our testbed. One can see from eq (13) that CO2’s
371
fate is determined by the temperature of the atmosphere, which depends on the stellar type and
17
372
the distance, and the ratio of the photolysis rate of H2O to that of CO2. Knowledge of these
373
terms will allow us to predict the abundance of CO2 in any hot Jupiter atmosphere. Finally, the
374
vertical distribution of species derived from thermochemical equilibrium can deviate
375
substantially from those derived via quenching, photochemistry and diffusion, and that the
376
simple assumption of thermochemical equilibrium is not valid in the observable regions of these
377
atmospheres.
378
$$$ need name/ Table 1/ need ref to rest of reactions …
40
CH + H2  3CH2 + H
5.51 × 10-16 T1.8 e-840/T
53
71
75
76
77
79
80
85
91
92
95
131
137
141
CH3 + H2  CH4 + H
H2O + hv  H + OH
CO2 + hv  CO + O
CO2 + hv  CO + O(1D)
HCO + hv  H + CO
H2CO + hv  H2 + CO
H2CO + hv  2H + CO
H2CCO + hv  1CH2 + CO
O + 3CH2 CO + 2H
O + 3CH2  CO + H2
O + C2H  CO + CH
O2 + C  O + CO
OH + H2  H2O + H
OH + CH4  H2O + CH3
1.14 ×1 0-20 T2.7 e-4739/T
9.95 × 10-6, 1.77 × 10-9
2.75 × 10-8, 7.18 × 10-11
1.64 × 10-8, 0
8.57 × 10-2, 8.57 × 10-2
3.57 × 10-3, 3.51 × 10-3
2.10 × 10-4, 1.56 × 10-4
1.44 × 10-1, 1.44 × 10-1
1.20 × 10-10
8.00 × 10-11
1.70 × 10-11
1.60 × 10-11
1.70 × 10-16 T1.6 e-1660/T
1.68 × 10-18 T2.2 e-1227/T
152
164
167
168
175
254
255
OH + CO  CO2 + H
CO + H + M  HCO + M
CO2 + 3CH2  CO + H2CO
HCO + H  CO + H2
2HCO  CO + H2CO
H + H2O  OH + H2
H + CO2  OH + CO
1.05 × 10-17 T1.5 e 250/T
5.29 × 10-34 e-370/T
3.90 × 10-14
1.50 × 10-10
3.01 × 10-11
7.50 × 10-16 T1.6e-9718/T
2.51 × 10-10 e-13350/T
Zabarnick et al.,
1986
Baulch et al., 1992
Frank, 1986
Frank, 1988, 1984
Baulch et al., 1992
Dorthe et al., 1991
Baulch et al., 1992
Srinivasan et al.,
2005
Baulch et al., 1992
Baulch et al., 1994
Tsang et al., 1986
Baulch et al., 1992
Tsang et al., 1986
Baulch et al., 1992
Tsang et al., 1986
a
For photolysis reactions, the photodissociation coefficient (s -1) is given at the top of the atmosphere and at
10 mbar. 2-body and 3-body rate coefficients are given in units of cm3s-1 and cm6s-1, respectively.
18
Figure 1—Temperature (solid) and eddy diffusion (dashed) profiles for the model
atmosphere. The cooler temperature profile is taken from 30 N from the night side of
the model by Showman et al., (2009). The hotter temperature profile is taken from the
dayside at the same latitude. The larger eddy diffusion is estimated as discussed in the
text (the larger values are for the dayside). The eddy diffusion is read along the top axis
where as the temperature is read along the bottom axis.
19
Figure 2—Thermochemical equilibrium mixing ratios derived from the temperature
profiles in figure 1. The top figure shows the mixing ratios derived for the dayside
(hotter) profile. The bottom figure shows the mixing ratios derived for the (nightside)
cooler profile.
20
Figure 3. Photochemical mixing ratios (solid) compared to the case with no
photochemistry (dashed) (but not thermochemical equilibrium) for the day (top) and
night (bottom) temperature profiles.
21
Figure 4. Important radical species involved in pathways governing the abundances of
CH4, H2O, CO and CO2. Solid is for the dayside temperature profile, dashed is for the
nightside temperature profile. The abundance of radicals increase with decreasing
pressure due to the availability of dissociating photons higher in the atmosphere.
22
Figure 5—Photochemical web illustrating the important chemical pathways that govern
the production and loss of the observable species. The boxes represent the observed
species and the circles represent species yet to be observed but are key in the production
and loss of the observed constituents.
23
Figure 6. The effects of changing metallicity (top) and C/O ratio (bottom) on the 3 bar
quench level mixing ratios for CO, H2O, CO2 and CH4.
24