Recent articles have invoked porous ice sands or other crustal grains as an additional potential reservoir for methane on Titan. (Selected examples: Sotin et al., 2009; Mitchell et al., 2009; Turtle et al, 2009,; Hayes et al 2008)

Last week, I set up a very simple laboratory experiment using an analogous system to try to answer the following questions:
1) How much methane could porous sands possibly hold per unit volume?
2) How will it affect the evaporation rate?
3) How could it affect the surface reflectivity?

As a laboratory analog for Titan’s hydrocarbon liquid mix (methane/ethane/nitrogen), I used solvent-grade heptane.
As an analog for Titan’s polar ice grains, I used either Flash-grade silica gel or analytical grade quartz beach sand. (The polar hydroxyl groups of the ice grains being analogous by the siloxy groups of silica)

The set-up
Three standard size 600 mL beakers were used:
Beaker A was charged with 400 mL silica gel
Beaker B was changed with 400 mL sand
Beaker C (control) was left empty.

Here are the initial images:
Click to view attachment

To initiate the experiment, a volume of Heptane was added to each beaker. Then, images and and weights were taken at key timepoints over a several day period to determine evaporation rate and monitor changes in visual appearance.

[Note: Although the temperature was held constant 298 K, the beakers were placed side-by-side in a fume hood with varying hoodflow (face velocity minimum was 100 cfm).]




Initial References:
Mitchell K.L., et al., LPSC 40 (2009) Abstract 1966. “A global sub-surface alkanifer system on Titan?”.

Hayes, A., et al. Geophysical Research Letters 35 (2008) L09204. “Hydrocarbon lakes on Titan: Distribution and interaction with a porous regolith”. doi: 10.1029/2008GL033409.

Sotin , C., et al. LPSC 40 (2009) Abstract 2088. “Ice-hydrocarbon interactions under Titan-like conditions: implications for the carbon cycle on Titan.”

Turtle, E. P., et al., Geophysical Research Letters 36 (2009) L02204. “Cassini Imaging of Titan’s High-Latitude Lakes, Clouds, and South-Polar Surface Changes.” doi: 10.1029/2008GL036186.

1) How much methane could porous sands hold?

Images after addition of heptane:
Click to view attachment
To 400 mL of silica gel in Beaker A was added 400 mL of heptanes. After stirring, the resulting slurry had a total volume of 425 mL. There was ca. 50 mL of clear solvent headspace.
Thus a volume of Flash grade silica can almost hold an equivalent volume of heptanes.
350 mL/400 mL = 87.5% volume equivalents

Click to view attachment
[I’ve been making flash silica slurries for most of my laboratory career, this result was a complete surprise!]

To the 400 mL of sand in Beaker B was added 200 mL heptanes. After stirring and settling, 25 mL of clear solvent head space remained.
225 mL/400 mL = 56% volume equivalents

Measured densities:
Flash grade silica: 0.488 g/mL (Lit.: 0.4 g/mL)
Sand: 1.607 g/mL (Lit.(dry sand): 1.6 g/mL)
Heptane 0.672 g/mL: (Lit.: 0.684 g/mL)


Based on this simulation, it is *possible* that crustal subsurface grains could hold a large (50-80% volume equivalent) amount of hydrocarbon solvent on Titan.

But for how long could it last???....

Wow! Neat experiment, Mike, thanks for sharing.

By any chance, did you try probing the slurry to see how firm it was? This sure seems to imply that Titan may have cryogenic quicksand in many spots; definite lander traps!

Thanks! I'm amazed at how much information can be extracted from a simple experiment. (Stay tuned!)

The sand was too heavy to slurry up. The silica was nice an fluid when agitated, but if you slowed down the stirring, it would set itself up.
(Think of stomping your feet in wet sand at the surf line. All OK when your feet are moving, but let the sand settle, and you are stuck!)

2) How will it affect the evaporation rate?

Beakers A (400 mL silica+400 mL heptane), B (400 mL sand+200 mL heptanes) and C (heptane only) were weighed and compared to the weight prior to heptane addition.

Here is a set of images showing how the headspace evaporates in the sand/heptane mixture, then the solvent level in the substrate(indicated by red arrow) can be seen to drop, although the rate slows down:
Click to view attachment

The chart of measured weights (and calculated volumes) below shows the amount of heptane remaining over time:
Click to view attachment
As can be seen from the graph, the heptane-only control evaporates at a roughly constant rate, while the sand/heptane and silica/heptane mixtures display a change in evaporation rate over time

The evaporation rate was calculated by calculating the amount of heptane volume change over the previous time period. A log-log graph shows the trend of the evaporation rate over time for the different substrate mixtures:
Click to view attachment
As the solvent goes deeper into the mixture, the evaporation rate decays exponentially. At all timepoints measured, the sand substrate had the lowest evaporation rate, although it contained less absolute volume.

In the present experiment, a 10-fold drop in evaporation rate was observed in both sand and silica by 2800 min (48 h) when the solvent level in the sand matrix had dropped to 10 cm below the surface.

Assuming a constant surface evaporation rate of the heptanes in the control beaker at 0.4 mL (=cm3) and a beaker diameter of 10 cm (surface area of 78 cm2), we derive an evaporation rate of 0.005 cm/min (=2,693 m/yr) at Earth STP (298 K).

According to Hayes et al., 2008: “The evaporation rate [on Titan] is taken to be 0.3 m/yr, consistent with an average windspeed of 0.1 m/s, methane mixing ratio of 0.35 in the lakes and methane relative humidity of 50% [T = 95 K)

[Putting this in perspective, an evaporation rate of 0.3 m/yr it is roughly equivalent to the aqueous evaporation rate in Needles, California or 3x the evaporation rate in the Piedmont region of North Carolina. Here is a pretty cool link for evaporation rates: http://www.grow.arizona.edu/Grow--GrowReso...?ResourceId=208]

From the simulation, it can be seen that even after 10x the expected lifetime of free heptane at the surface, 10% of the original heptane remains in the silica matrix. Similarly, after 7x the expected lifetime of a quantity of heptanes at the surface, 10% of the original heptanes was measured in the sand matrix.
Click to view attachmentClick to view attachment

If these results scale, (and I’d speculate that the lower absolute temperatures on Titan would accentuate the effect), it would be reasonable to expect that a few m of porous substrate would be sufficient to eventually slow the evaporation rate at least 10 fold and allow quantities to remain in the porous matrix longer than would be expected based on the free liquid evaporation rate.

As one implication, Turtle et al, 2009 provides evidence for South Polar lakes that were detected by ISS, but were not observed by SAR RADAR 2.5 years later. They postulate that “enough liquid could have evaporated or percolated into the subsurface during the intervening years to explain the lack of lakes observed by RADAR”.

Considering the change in evaporation rate evidenced by the laboratory simulation above, it is possible that significant amounts of hydrocarbon solvents could even still be trapped in the lake sediments and remain in buffered communication with the atmosphere.

Hmm. But presumably actual Titanian surface material grains are much lighter than silica; still sounds like a sticky situation to me!

Assuming the heaviest materials on Titan's surface are pure ice grains, they would have a density of 1 g/mL.
At Titan's temperature (95 K), liquid methane would have a density about 0.45 g/mL.
(Other non-halogenated organics would be roughly 0.6 g/mL)

On Earth, density ratio of rock/water is about 8/1.
On Titan the density ratio of ice/hydrocarbon solvent is about 2/1 or lower. And Titan has a lower gravity, so things would settle out even slower (however, pure methane's viscosity is much less than the viscosity of water).

Throw some emulsifying materials into the mix, and it could be a really gross mess. Titan's streams and lakes could be more of a slurry than a broth.

Quartz to water is around 2.6:1, not 8:1 - so it's on a par with your Titanian ice/solvent. Intriguing experiments though - any chance of doing similar at <100K and 1.5atm? ;-)

Andy

(I stand corrected. Quartz density is ca. 2.65 g/mL. Thanks!)

Great experiment, Mike. Nice and transparent. I'd have replied earlier
but was out in the field studying transient lakes in high-evaporation regions
(specifically Racetrack Playa in Death Valley) where there wasnt internet!

Evaporation rates from free liquid surfaces are controlled by the vapor pressure excess
(saturation v.p. minus ambient partial pressure), and windspeed. But in a regolith, as
the stuff starts drying out from the surface downwards, the labyrinth of pathways through
the pores becomes the limiting factor (pore size, tortuosity become factors) and the
evaporation rate falls off as inverse square root or log of time or something. Used to
be (maybe still is) a common problem examined for dessication of ice-saturated regolith
on Mars. Long ago Konrad Kossacki and I looked at the large-scale retention of porespace
on Titan regolith as a reservoir for liquid hydrocarbons (though the original idea was due
to Stevenson and Eluskiewicz)

http://www.lpl.arizona.edu/~rlorenz/kossacki.pdf

There's ample geomorphological evidence that Titan rivers can be vigorous, although that
doesnt mean some can be 'muddy'. Whether (or perhaps rather which and when) lakes
are 'tar' vs 'LNG' remains to be figured out, but I'd be shocked if at least some aren't LNG-like

Your basic point that the regolith can be a significant reservoir of even volatile hydrocarbons
is right on. It was basically your heptane-soaked silica (but methane/ethane-soaked ice perhaps)
that got rammed into the warm GCMS inlet at Huygens impact, and got sweated out.

Ok, shameless self-plugging, but I just spotted this at the end of that paper (written in 1996)...

"In this sense, a buried ocean behaves thermodynamically (in terms of ocean-atmosphere
equilibrium) just like a surface ocean would. Note, however, that we are not advocating
that all of Titan's hydrocarbons are concealed in a porous regolith: we only suggest that most
of the 'deep, global ocean' can be concealed. It is likely that exposed regions of nonporous
ice may exist, as well as surface reservoirs (perhaps crater lakes (Lorenz 1994) of hydrocarbons"

Well, I make no apologies for getting the lake morphology wrong - seemed a good idea at
the time - but I'd venture this overall picture holds up pretty well.

Thanks for sharing your thinking on that. (Plenty more lake science for Cassini to do!)

3) How could it affect surface reflectivity?

Here is an set of side-view images of the silica/heptane beaker over time:
Click to view attachment

(and the same image, contrast enhanced):
Click to view attachment

Something very interesting happens as the heptane evaporates: the silica very quickly changes appearance (to appear dry), even though the silica still contains a large amount of heptane. This occurs between 30 and 60 minutes, although only a relatively small amount of heptane evaporated during this period (8 mL or 2% of the total.) At the 126 min image, very complex banding can be seen (possibly from evaporation/recondensation?).

In this example, the appearance of the porous material can changes quickly, even though the amount of solvent inside the grains is almost the same.

Only a tiny amount of solvent change is necessary to change the visual aspect!

At 1384 minutes, the silica appears uniformly dry, although it still contains 32% heptanes.
So even though it looks "not damp" it is not dry!

Mike, would it be possible to place an optical microscope on its side, close enough to observe the grains through the glass, and allow it to move up and down on a platform (maybe a lab jack?) so that you could get some close up images of the grains at various levels? We have a couple of old and unused microscopes around our lab, I'm pretty sure no one would notice if I cannabalised one. I'm wondering if I can get chemistry to loan me some heptane so I can have a go myself!

Here are a couple close-up images at the "visual dry/visual wet" interface of the silica/heptane mix (taken at 86% heptane capacity),
Click to view attachment

and the sand/heptane mix (taken at 73% heptane capacity):
Click to view attachment

Both images are unaltered (i.e. no contrast adjustment).

The silica interface is fairly even (silica is also finer grained) but still shows a fractal "cracking" pattern at the interface. (Having your column crack when running a chromatographic separation is considered bad form). The wet blobs of silica tend to congregate in wetter blobs between the cracks.

The sand interface is fairly rougher (more uneven distribution horizontally), and shows a similar order of cracking, although the cracks are larger. Individual wet grains can be seen in the image.

Here are the images of the top surface of a second silica+heptane vessel as it dried (set up similar as before: T = 0, 35, 55, and 89 min):
Click to view attachment

And here is an animated GIF:
Click to view attachment
(click to animate)

As it dries, the silica/heptane matrix surface turns from transparant gray to a brigter opaque white. The blotchy white drying area starts from several initating points. Some of these grow quickly, while others remain the same size after initiation. (Kinda looks familiar, doesn't it?)

The surface appearance quickly becomes uniform white over a short period of time as the surface "dries" (= has slightly less heptane).

So silica/heptane changes appearance on wetting or drying. Would methane/water ice do the same thing on under Titan conditions?

Yup.

Check out: Sotin, C., et al. LPSC 40 (2009) Abstract 2088. "Ice-hydrocarbon interactions under Titan-like conditions: Implications for the carbon cycle on Titan". Available here: http://www.lpi.usra.edu/meetings/lpsc2009/pdf/2088.pdf
The authors set up an experiment at 90 K and at either 1.38 or 2.07 bar (Titan conditions = 95 K and 1.5 bar)) with liquid methane dripping into a depression in a pure ice substrate.

Key quotes clipped from the abstract (check out Figure 3):

"Figure 3 shows how the methane darkens the ice substrate as it presumably disperses into the ices' open pore space which may be cracks in the ice sample".

"...the elevated ice gets wetted by liquid methane before the pool starts filling up."

Presumably as the water ice dries from methane evaporation, the optical properties would go back to the initial lighter appearance of the ice. (Just like observed for silica).

Thinking about the optical properties of methane-wet silica sand and methane-wet ice sand prompted me to look up the relevant refractive indices. I found:

Methane/ethane 1.29
Water ice 1.31
Silica 1.45

The closeness of the first two implies that a methane/ice interface should reflect almost no light and refract it very little. Indeed if the liquid were able to fill all the pores and cracks in the ice (unlikely) the assemlage would be rendered virtually transparent. So maybe the changes in appearance you observed in your experiment with silica would be even more pronounced with ice.

Here is recently released Cassini ISS graphic PIA11147.
Click to view attachment

It shows the changes in the S Polar region from different observations. A 2007 RADAR observation noted that the “inundated” areas seen in 2005 appeared dry.
More information can be seen in this thread: http://www.unmannedspaceflight.com/index.p...c=5784&st=0
And the original article:
Turtle, E. P., et al., Geophysical Research Letters 36 (2009) L02204. “Cassini Imaging of Titan’s High-Latitude Lakes, Clouds, and South-Polar Surface Changes.” doi: 10.1029/2008GL036186.

As the authors noted, It is possible that transient the South Polar were detected as the surface matrix was saturated or inundated, and that the 2.5 year later observation by RADAR did not detect the “damp” material because the lakes had “evaporated or percolated into the surface” (Turtle et al, 2009).
[The hydrocarbon evaporation rate on the surface was estimated at 0.3 m/yr by Hayes et al., 2008]

By analogy to the silica/heptanes experiment above, it is possible that a porous material could appear dry, but still contain significant amounts of hydrocarbon material just under the surface. These wet sediments could still be in a buffered communication with the surface – still evaporating, but at a slowly decreasing rate. Here is a diagram showing this possible explanation related to the observations of the transient S Polar lakes:
Click to view attachment

Bottom line (!): Those S Polar Lake sediments could still be wet and “breathing” just below the surface.

Extending this to the Equatorial Sand Sea basins (including Mezzoramia), if the properties of the dune sea material with methane are analogous to silica/heptane, then damp porous material just under the surface dunes could also be a reservoir of methane.
Here is a possible side view of an Equatorial Sand Sea Basin:
Click to view attachment

At the center of the basin, a layer of porous sediments/dunes cover a damp subsurface that is slowly evaporating, even possibly after several years (decades? millennia?). Deeper still, this may blend into a much larger porous deep crustal reservoir that may be in global communications with other reservoirs.

The sand sea subsurface basin could act as an intermediate reservoir of methane. The evaporation rate (dependant on depth of covering dry sediments) is not as rapid as surface reservoirs. Yet the communication rate with the atmosphere may be faster than that of the much deeper crustal reservoirs described in Konsacki et al., 1996.

[The evaporation rate according to Hayes is 0.3 m/yr.
Taking the crustal diffusion rate (global?) in Konsacki et al of 2E5 kg methane s-1 after units conversion gives 6.3E12 kg methane y-1. Using a methane density of 0.5 g/mL (=cm3), we get 1.3E10 m3 methane y-1 diffusion rate. Assuming this is the global surface diffusion rate, and dividing by a surface area of 8E13 m2, we get 1.6E-4 m/yr. The diffusion rate is thus 2000 times slower than the evaporation rate.]

Here are volume capacities of heptane with other materials (measured by myself on a smaller scale):
Click to view attachment

Volume capacity of porous sediments in the sand sea basins?
Ballparking this conservatively:

Assume 10% of the sand sea basins (most likely the central depths) has a porous sediment depth of 100 m.
Assume the solvent capacity of the porous sediment is 50% (see above post).

Here are estimated areas/volumes of porous sediments in the sand sea basins (conservative, the areal extent is likely underestimated):
Click to view attachment

1.09E7 km2 total sand seas area (including Mezzoramia) x 10% (basin area with deep sediments) x 0.1 km (sediment depth) x 50% (solvent volume)
= 5.4E4 km3 hydrocarbon solvent could be held in porous sediments in the dune seas.

Compare with estimates of surface lake volume:
6.3E5 km (North polar lake area + 55 dark surface features southward of 69 S area from Turtle et al., 2009) x 0.2 km (200 m depth is 10x higher than Lorenz et al. 2007 estimate of 20 m average volume)
= 1.3E5 km3 hydrocarbon solvent held in surface features.

So even with a fairly conservative estimate of porous sediment capacity in the sand sea basins, it is within an order of magnitude of the amount of surface liquid estimated in the lakes.

Damp porous equatorial dune sediments could be a significant surface reserve for methane and other hydrocarbons on Titan.

Here is a diagram of Titan’s possible methane cycle showing the different possible points of hydrocarbon evaporation.
Click to view attachment

The surface reservoirs of hydrocarbon solvents could be in rapid atmospheric equilibrium, while porous basin sediments could act as a buffer due to the decreasing evaporation rates with depth. Even while the surface appears dry, damp sands hidden in the dune seas could “breathe” methane back into the atmosphere over timescale of years, decades, or even millennia. As mentioned above, this could explain the puff of methane when the Huygens probe touched down on the surface as well as the methane humidity profile (see Griffith, C.A., 2009).
Deeper still, subsurface crustal reservoirs (Hayes et al., 2008; Kossacki and Lorenz, 1996) could exist that effectively “caches” the hydrocarbons over much longer timescales (and with potentially larger capacity).

-Mike


(For a freely available recent discussion of Titan's methane cycle and it's relation to Huygens descent and surface data see: Griffith, C.A. Phil. Trans. R. Soc. A. 367 (2009) 713-728. "Storms, polar deposits, and the methane cycle in Titan's atmosphere." doi: 10.1098/rsta.2008.0245. html version (pdf also freely available) here: http://rsta.royalsocietypublishing.org/con...7/1889/713.full)

There are some striking turns of phrase in that paper. My favourite: "shallow unseen bogs".