Surfaces processes François Dulieu LERMA Paris Observatory and Cergy Pontoise University
Cold surfaces and astrophysics Image credit: J. Freitag and S. Messenger Smoky areas = dust grains Cold (10-15 K) ==> accretion of most of gas species ==> Catalytic centers ==> Increase with density of medium
Hydrogenation and surface chemistry The magic box? Chemistry on dust grains is supposed to work when gas phase do not! Gas phase Solid phase OH + H OH + H? One collision every day One collision every picosecond
Surface physics
Surface processes Temperature, Coverage, morphology
Surface processes Surface process I : Exp. Temperature, Coverage, morphology
Surface processes Surface process I : Exp. Surface process II : Chemical routes Temperature, Coverage, morphology
Surface processes Surface process I : Exp. Surface process II : Chemical routes Surface process III : Diffusions Temperature, Coverage, morphology
OVERVIEW Surface processes I Order of s Experimental methods (TPD, RAIRS, STM) Surface processes II Chemical routes Surface processes III Diffusion on surface Diffusion in bulk Energetic aspects
Surface science numbers : ML Magnitude Crystalline Ice (figure adapted from Guillot and Guissani, JCP, 2004) 1 cm3 ~ 1 g Cubic ice! Nvolume ~ 1/18 Navogadro ~ 3 1022 mol/cm3 Nsurface ~ Nvolume2/3 ~ 1 1015 mol/cm2 Lsites ~ Nvolume -1/3 ~ 3 10-10 m 1 ML ~ 1 1015 mol/cm2
Surface science numbers : accretion flux P bkg is UHV requiered? Cold surface Yes if you want to study any effect with a typical rate of 10^13 mol/cm2/s (show case of photodesorption) P bkg Flux* Time for 1 ML 1 10-10 mbar 0.0003 ML/s 8 hours 1 10-9 mbar 0.003 ML/s 50 minutes 1 10-8 mbar 0.03 ML/s 5 minutes 1 10-7 mbar 0.3 ML/s 3 seconds * For m = 18 g/mol T=300K
Interaction potential with the surface z x Eads x Ediff E0
Gas deposition methods Flux decreases as 1/d2 - good homogeneity - no side deposition (good T control) - low flux (10^12 10^13 mol/cm2/s) - big and expensive Flux measurement : surface saturation (see later) or calibrabration vs Bkg - high flux (10^13 10^15 mol/cm2/s) - can be inhomogeneous - possible side deposition Flux measurement : IR calibration vs Bkg deposition - good control of flux (see above ) - real stochastic deposition - not adapted for radicals - can induce parasitic signals (other cold parts) - can pollute the chamber Flux measurement : Direct pressure measurement
Mass spectroscopy Ionisation: - Local in a QMS head - Electron gun (or filament) - By laser induced ionisation Detection : + Time of flight + Multipole filtering - Measure only gas phase! (so require desorption...) - Can be quantitative, with high precision (%), every species - Can measure up to 0.001 ML - Except for laser : cheap and robust! Not only TPD : - King and Wells - Quantum state selected or Internal energy access - Kinetic energy measurements - Induced laser desorption
Infra-red spectroscopy - Transmittance - Reflexion, incident angle - Reflexion large incident angle - Measure solid phase! - Only active IR species - Solid spectroscopy (ambiguities, environment dependancy) - Can measure up to 0.1 ML of some species - Allow real time measurements - Is more applicable to astrophysics
Infra-red spectroscopy - Transmittance - Reflexion, incident angle - Reflexion large incident angle - Measure solid phase! - Only active IR species - Solid spectroscopy (ambiguities, environment dependancy) - Can measure up to 0.1 ML of some species - Allow real time measurements - Is more applicable to astrophysics
Infra-red spectroscopy - Transmittance - Reflexion, incident angle - Reflexion large incident angle - Measure solid phase! - Only active IR species - Solid spectroscopy (ambiguities, environment dependancy) - Can measure up to 0.1 ML of some species - Allow real time measurements - Is more applicable to astrophysics
STM - - Measure solid phase directly - Give insights at atomic scale - Only electronic density (Need a good quantum chemist!) - Can count directly species on substrate, homogeneities... - Mostly chemisorbed species and organised substrates
of molecules 0 UHV chamber 1 Ice sample (np) Ts = 10 K 2 D2 beam ( Tb 30 350 K) 3 Monitor (real time) partial pressure of D2 D2 ( or H2 )
of molecules ( King and Wells ' method) OUT IN IN = OUT Stick
of molecules ( King and Wells ' method ) OUT IN IN = OUT For H2 and D2, there is T dependence of the sticking And a rather large isotopic effect
OUT Some physics with hands IN IN = OUT IN = OUT H2 D2 (T/2)
of molecules ( King and Wells ' method ) OUT IN IN = OUT H2 D2 (T/2) If T for D2 is divided by a factor 2 the law is the same
Speed distribution of a beam gas OUT IN Temperature distribution, one dimension, effusive beam Tgas = 100 K Tgas = 200 K H2 D2 (T/2) Tgas = 300 K Speed in m/s
OUT Cut off model IN Temperature distribution IN = OUT Tgas = 100 K Tgas = 200 K Tgas = 300 K Stick Do not stick Speed in m/s
OUT Cut off model IN Temperature distribution IN = OUT Tgas = 100 K Tgas = 200 K Tgas = 300 K Stick Do not stick Speed in m/s
of light species OUT IN IN = OUT ß = 2.5 for isotropic thermal distribution Chaabouni et al A&A, 2012; Matar et al, JCP, 2010, Large isotopic effect H and D, never took into account (Tg > 50K) Amorphous silicate is harder than ASW ice
of heavier gases OUT IN IN = OUT Better mass matching Lower speed Higher adsorption energy Larger degrees of freedom For all the others molecules (H2O, CO, O2, CH4 ) No rejection mechanism with coverage S > 0.9 See Kimmel et al JCP, 2001, Acharyya et al, A&A 2007, Bisschop et al, A&A 2006
Interaction potential and thermal probability Before desorbing, most physisorbed species diffuse x Eads Ediff E0 For one adsorbate Pdesorption = ϖ exp (- Eads /kts ) Pdiffusion = ϖ exp (- Ediff /kts ) E0 = ½ h ϖ Usually Eads > Ediff ==> Pdiffusion >> Pdesorption So choosing ϖ = 10^13 /s corresponds to E0 = 21 mev and ϖ = 10^15 /s corresponds to E0 = 2068 mev! For D2 see Amiaud et al, JCP, 2006; Amiaud et al in prep;
Desorbing flux For N independent adsorbates Flux = N ϖ exp (- Eads /kts ) so called order 1 If the adsorbate is homogeneous and has a constant number of molecules on the surface No 1 ML (case of multilayer ) Flux = No ϖ exp (- Eads /kts ) so called order 0 HV accretion limit UHV accretion limit
Thermally programmed desorption Increase linearly the temperature Detection threshold, Diminution of the adsorbate number = reduction of the desorbing flux
Ideal Thermally Programmed vs desorption from amorphous substrate D2 desorption from porous water ice
Interaction potential
How to analyse?
model : Fermi-Dirac equilibrium
Isotopic segregation
Isotopic segregation
Isotopic segregation
Ortho and para segregation
Morphology of ice porous and compact
Morphology of ice porous and compact
Morphology of ice porous and compact
Morphology of ice porous and compact
Classical inversion method : easier! What is more important: substrate or adsorbate? Adsorbate Substrate Noble et al 2012, MNRAS
What is more important: substrate or adsorbate? Adsorbate Substrate Noble et al 2012, MNRAS
TPD from amorphous substrates (or heterogenous environments) A B C Where is the multilayer adsorption energy? (adsorbate-adsorbate interaction) A: intial separate peak B: Mixed peak C: 0th order behaviour immediately
Silicate substrate and other adsorbats Noble et al 2012, MNRAS
(Dis)Advantages of the IR is not very efficient for surface effects Dandling bounds Accolla thesis 2010 Oba et al, ApJ, 2009
STM is for coverage measurement better than TPD MULTILAYER REGIME Multilayer Adsorption energy Sub-Layer Regime Jorgensen thesis, 2011
for coverage measurement STM isbetter than TPD MULTILAYER REGIME Multilayer Adsorption energy Sub-Layer Regime
IR allows to have signature of some embedded intermediate radicals Ioppolo et al, MNRAS, 2011
Comparisons Theses tools are complementary, most teams use 2 of them. IR Mass A spectroscopy STM excellent - Identication of good stable compouds Possible isomeric discrimination Radicals Best - - 2 colors Sensitivity Low Best Surface quantities Very low Surface effects Poor good Best indirect Energetic aspects - Poor - Best B Lasers C