Today is the first day of several weeks of conferences, so I thought a good idea would be to kick off the blog by summarising some of the interesting topics that come up during the talks. I can't possibly write down everything that gets said, so I'll just highlight some of the points that I found interesting.
This week and next week, I am attending a conference titled "WATSURF 2013. Water at interfaces: new developments in physics, chemistry and biology" at Les Houches, France (see below!). Today's session is on "From bulk to confined water."
Marie-Christine Maurel from UPMC in Paris presented a summary of theories on the origins of life, focusing on the RNA world hypothesis. One interesting idea in her talk, due to Attwater & co, is that ice acts as some sort of matrix to organize RNA activity. Not sure exactly what she meant, but may be worth exploring. An interesting experiment due to L. Orgel is that a single RNA strand in the presence of activated nucleotides "reproduces" spontaneously due to template directed synthesis followed by spontaneous formation of bonds between consecutive bases. The problem with this idea is that ribose is unstable in "life-like" solution conditions. Another interesting idea: Alternative Genetic Systems, where the backbone of DNA is replaced by something else: PNA, P-RNA, HNA, ANA, TNA. These systems can coexist and interact with DNA / RNA. One interesting candidate for fossils from the RNA world (Diener 1989) are viroids (30 known species): RNA strands with lots of complementarity that do not code for any protein, have no envelope and no capsid. Some show ribozyme activity. A final point that was also interesting is that high temperature and high pressure tend to have opposite effects on biological activity, so putting them together (e.g., at hydrothermal vents) yields similar biological activity as at STP.
José Teixeira (LLB, France) then talked about "Bulk and confined water". Confined water is very different from bulk water: thermodynamic properties (density, phase transitions, crystallisation), transport properties (viscosity, diffusion, ...), vibrational density of states, electronic properties. Where is confined water? Porous materials, earth rocks, hydrophilic surfaces, droplets, inverse microemulsions, foams, clays, membranes, biomolecules. [Look up simulations by Mounir Tarek for simulations]. H. E. Stanley and him have a curious super-simple model of hydrogen bond lifetimes that explains residence times of water in terms of H-bond lifetimes and populations of "mobile" (<2 h-bonds) and "immobile" particles (function of T). Very roughly and qualitatively, confined water (e.g. in pores or next to hydrophilic surfaces) has dynamics properties comparable to bulk water that is 20 C colder. Showed some experiments (Richard, Mercury, Massault, Michelot 2007) of enhancement of D vs H near silica-water interface in confined pores. A very curious isotopic effect is that cement built with D2O is hopelessly fragile compared to cement with H2O. Small change in diffusion constant causes big change in morphology of gel phases that form. Of course, the same thing happens upon changes in temperatures: real cement has lots of additives to compensate for the weaknesses of "raw" cement. He mentioned an interesting book on the subject: Tobias et al, "Water in confining geometries", Springer 2003.
Werner Kuhs (Göttingen, Germany) talked about nanoscopic atmospheric ice. Recommended a long, recent and comprehensive review on the subject, Bartels-Rausch et al (2012) Rev Mod Phys. According to Bernal-Fowler rules, there are six possible configurations of water at each site in ice, and their distribution is statistical. For ordered ices, QM calculations and experiment agree excellently. For disordered ice Ih, it's much more complicated to even do the QM calculation. So ice provides a good test-bed for QM calculations. "Bjerrum L-defects" (missing H between two O's) exist in ice at low concentrations (~10^-8). There's even disorder in the oxygens, which makes it hard to experimentally measure the H-bonding geometry in ice (he claims it hasn't yet been done right). This disorder makes the O-H distance in ice appear to be shorter than it actually is. An interesting detail: lattice constants can be got to 5-6 significant figures, so they provide very stringent test for theory that wants to be quantitative. Bartels-Rausch et al (2012) Atmos. Chem. Phys. Discuss. 12:30409-30541 has a discussion on the thickness of the disordered layer on ice Ih: the experiments & computations are all over the place. Some experiments on ice close to the melting point of ice, form droplets first, then as heating up close to melting point, a liquid-like layer forms below these droplets.
The second half of his talk was about "cubic ice". It turns out that nobody has experimentally formed hexagonal ice below ~ 190 K (you can form it at low temperatures and then cool it down). He claims that all "observations" of cubic ice can really be explained through stacking faults in ice Ih. Modeling these stacking faults very carefully and statistically, can fit the diffraction data for "cubic ice" very well. Furthermore, there is no interfacial water in "cubic ice" (unlike that seen in Moore & Molinero 2011 simulations). "Ice Ic" got from different other forms of ice has different (but reproducible) structures: some sort of "structural inheritance" in the long-range order of waters must play a role.
Thomas Loerting (Innsbruck, Austria) gave an excellent talk on about amorphous ices, i.e., LDA, HDA, VHDA, etc. My old office mate, David Limmer, has taken a deep look at this subject in the past few years coming from the skeptical camp (about the connection to LDL/HDL and the second critical point hypothesis), so I thought it interesting to see the experimental data presented coherently and clearly. They presented some very interesting results towards the end suggesting that they could see the beginning of an LDA->LDL transition upon heating which was, at the very least, suggestive. I'll have to get DL to clarify this (hint: comment on the blog!). First, a few preliminary trivia about equilibrium ice that I thought were worth jotting down. One obvious consequence of the Clausius-Clapeyron equation, phase boundaries that are parallel to the pressure axis are driven by entropy changes, those that are parallel to the temperature axis are driven by volume changes. Ice XI: proton-ordered hexagonal ice (in the presence of KOH); but W. Kuhs argued that this ordering is all due to the KOH, and not the water. Phase transformation beyond 100 MPa would require an ice colume more than 10km high, which is not found on the surface of Earth. But you can find such columns in space (e.g., Ganymede has ~900 km of ice layering, => up to 1GPa, layering of ice matches predictions from phase diagram well). Highest-density phase known is Ice X (breaks Bernard-Fowler rules), density of 2.5 g/cc at 100 GPa & melting point of 2,100 C. Claims of metallic shine in samples. Metastable ices (e.g. amorphous ice) are rare on Earth but common in space. To prepare amorphous ice, many options:
a) start from vapour & deposition to cold substrates (ASW => LDA)
b) extremely fast cooling (10^7 K/s, HGW => LDA)
c) pressurising ice Ih (HDA => VHDA)
HDA: take an ice cube, cool to 77K, then increase pressure. Around 1.2 GPa, nearly sudden reduction in density (Mishima et al, Nature 1984, 310, 392-395). When you release the pressure, the density doesn't rise back up (hysteresis). rho_HDA = ~1.15 g / cc. Almost no long-range order ("peaks" in XRD are very diffuse). If you heat HDA at 1.1 GPa, you get another transition to VHDA wuth hysterisis (rho_VHDA = ~1.26 g / cc). From Raman spectroscopy, of D2O & H2O mixtures, measure LDA O-D-O distance of 2.77 A, HDA: 2.82 A and VHDA: 2.85 A (must be due to higher coordination)
HDA <-> LDA transformation is apparently first order (at 130 K - 140 K), pressurising and depresurizing induces conversions. You can set up an experiment where LDA & HDA coexist, with a phase boundary between them => really first order.
Atomic Si, P, C, also compound SiO2 GeO2 and triphenylphosphate have polyamorphism and LDA & HDA state, as well as "anomalous" liquid properties.
Cooling rates of 10^5 K / s still produce crystalline ice from liquid water. 10^7 K / s achieved by spraying um-sized droplets onto plate at ~ 77 K. XRD of this splattered ice, "Amorphous Solid Water" (ASW), is equal to LDA and that of Hyperquenched Gaseous Water (?). However, ASW is a "microporous solid". You can anneal this at ~100 K / ~ 110 K to remove micropores. You can trap CO_2 in these pores, and then heating this thing up in a vacuum chamber produces clathrates.
He claims there is a glass transition at ~136 K in LDA if heating at 30 K / min (Nature 2005) if the samples are well-annealed (e.g. prepared by deposition on ~130 K [check numbers]). For HDA at 0.20 GPa, heating causes a continuous volume change with a kink at ~140 K and a step at ~ 150 K .. Can access kink reversibly (up to 144 K) by heating rate & cooling rate: 2 K / min. They then claim that HDA at ambient pressue has a (tiny) glass transition at 116 K, which can be accessed reversibly. Above the kink but below phase change, the observation says that there is a liquid. But perhaps its only that the hydrogens can start to move around and the oxygens are all trapped in particular sites => viscosity would be solid-like. But if they do an experiment with pushing a needle into the LDA sample, the needle is pushed out upon heating below T_g, then above T_g but below crystallisation, needle penetrates. They've filed a Guinness World Record for coldest liquid water ever observed (136 K)!
Importantly, they make no claim either way about the second liquid critical point hypothesis.
The last talk of the day was by Patrick Ayotte on photochemistry in ice and its link to HNO3 production. Since this is really outside my field and interests, I didn't take notes on this.
Alright, next blog entry tomorrow!
Patrick
No comments:
Post a Comment