Diffusion Kinetics for Methanol in Polycrystalline Ice Patrick Marchand, Samuel Riou, and Patrick Ayotte* De ´ partement de Chimie, UniVersite ´ de Sherbrooke, 2500 BouleVard UniVersite ´ , Sherbrooke, Que ´ bec J1K 2R1, Canada ReceiVed: June 29, 2006; In Final Form: August 16, 2006 Quantitative analyses of the isothermal desorption kinetics from methanol-doped H 2 O films on Pt(111) reveal that transport kinetics for CH 3 OH in polycrystalline ice are much slower than previously reported. They also indicate that MeOH displays first-order desorption kinetics with respect to its instantaneous surface concentration below 0.1 mole fraction in ice. These observations allow isothermal desorption rate measurements to be interpreted in terms of a depth profiling analysis providing one-dimensional concentration depth profiles from methanol-doped polycrystalline ice films. Using a straightforward approach to inhibit ice sublimation, transport properties are extracted from the evolution of concentration depth profiles obtained after thermal annealing of binary ice films at high temperature. Heterodiffusion coefficients for methanol in polycrystalline (cubic) ice I c films are reported for temperatures between 145 and 195 K and for concentrations below 10 -3 mole fraction. Finally, diffusion kinetics for methanol in ice are shown to display a very strong concentration dependence that may contribute, in addition to variations in laboratory samples microstructure, to the disagreements reported in the literature regarding the transport properties of ice. I. Introduction Natural ice is a ubiquitous and continuously evolving molecular solid that presents heterogeneities on several length scales, ranging from molecular to kilometers. 1 These features represent a considerable challenge toward decoding the planetary atmospheric archives trapped in the polar ice caps. 2 Accordingly, the interpretation of climate proxies from ice cores has recently sparked renewed interest in the complex transport properties of ice. 3,4 Furthermore, as the latter are strongly coupled to the bulk uptake and interfacial reaction kinetics on ice particles and snow, 5,6 a better understanding of these phenomena is also crucial to help quantify the role played by ice in determining the chemical composition of the atmosphere 7 and the polar boundary layer. 8,9 It has thus been long recognized that the composition, structure, and morphology of ice particles and snowflakes encode chemical and physical clues of the environments in which they were formed and subsequently evolved. As they precipitate and accumulate on seasonally and permanently snow- covered areas, they form vertically stratified deposits within the snow pack. The complex transition from this initially highly porous, freshly fallen snow deposit, to the highly connected percolating pore structure of the fern, to dense polycrystalline ice still remains poorly understood. This formidably complex process controls the early-time evolution of the initially vertical concentration profiles. A quantitative understanding of aging processes in natural snow, such as sublimation and condensation, vapor and bulk diffusion, metamorphism, densification, creep, and flow, is thus required to properly date and interpret the concentration profiles retrieved from ice cores as well as improve our understanding of the role played by the snow pack in atmospheric chemistry phenomena at the polar boundary layer. The various impurity species used as climate proxies display complex interactions with the ice matrix that are controlled by their molecular properties, their chemical and physical state, but also the ice composition, morphology, and microstructure. Consequently, these features will also determine the transport and equilibrium properties of impurity molecules trapped within the ice matrix: from their initial spatial distribution within snow particles in the atmosphere to whether they will evolve to form microinclusions, 10 collect at grain boundaries, 11 or disperse more or less homogeneously within the ice crystallites that compose natural ices. Detailed knowledge of these parameters, as provided from analyses of natural samples, is thus required in order to guide laboratory investigations and provide environ- mentally meaningful kinetic parameters from model systems. Various experimental approaches have been proposed to probe the transport properties, and in particular the molecular diffusion kinetics, for various impurity molecules in artificial ice samples. 12-20 The acute and complex dependence of the phase, morphology, and microstructure of laboratory ices on preparation methods (vapor condensation or crystallization from the melt) 19-21 and growth conditions (flux/pressure, 22 temper- ature, 23 angle of incidence, 24 nature of the heterogeneous substrate, 25-28 etc.) and the great difficulties to experimentally quantify defect densities (dislocations, interstitials, vacancies, Bjerrum defects, etc.) 29 are all factors that severely limit meaningful comparisons between the results from these different studies. Compounded with our limited ability to characterize accurately and nondestructively the morphology and micro- structure of this delicate material, these considerations contribute to the large discrepancies between the results obtained from laboratory ice samples prepared by very different methods. Accordingly, uncontrolled and poorly characterized defects are often invoked as a possible source for the irreproducibility of bulk diffusion measurements in otherwise identically prepared macroscopic samples even within a single investigation 19,30 let alone comparing different studies. As a specific example, the * To whom correspondence should be addressed. Phone: 819-821-7889. Fax: 819-821-8017. E-mail: Patrick.Ayotte@USherbrooke.ca. 11654 J. Phys. Chem. A 2006, 110, 11654-11664 10.1021/jp0640878 CCC: $33.50 © 2006 American Chemical Society Published on Web 09/26/2006