January_AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | J A N U A R Y 2 0 1 8 2 9 uncoated ice by maintaining the ice around - 100°C [9] . The sublimation that occurs at this temperature also concen- trates the soluble impurities locally so that they can reach levels where they can be detected, although quantifica- tion is problematic (Fig. 3). X-ray fluorescence has been used to identify elemental impurities in 10-800 µm particles, in grain boundar- ies, and in 3-10 µm frozen brine inclu- sions in Vostok accretion ice (ice frozen onto the bottom of the core from the underlying Lake Vostok, Antarctica) held at - 140°C, while x-ray absorption near edge structure (XANES) spectros- copy was used to distinguish between sulfides, sulfites, and sulfates in the core [10,11] . In 2004, Iliescu, Baker, and Chang first demonstrated that both electron backscatter diffraction patterns (EBSPs) and channeling patterns could be ob- tained fromuncoated ice held at - 100°C (Fig. 4) [12] . Since then, several studies have used EBSPs to produce (both c - and a -axis) pole figures and grain boundary misorientation frequency data for ice, particularly from polar ice cores [13,14] . A combination of optical thin section micrographs coupled with EBSPs is a powerful way to study low-angle sub-boundaries in ice [13] . STUDYING DISLOCATIONS A significant body of work exists with regard to characterizing disloca- tions in ice, which have a Burgers Burgers vector. Glide can occur on a basal (0001) plane either as 60° or screw dislocations, or on a nonbasal (pris- matic or pyramidal) plane as an edge dislocation. However, the critical re- solved shear stress (CRSS) for non-bas- al slip is far higher than that for basal slip [1] . There are in fact two types of bas- al planes, which are referred to as the glide set, in which packing of oxygen at- oms is similar to that in an hcp metal, and the more widely-spaced shuffle set. The can, in principle, dis- sociate into partials on the glide set. Whether glide occurs on the glide set or shuffle set is not known, but observations of the dislocation of the dislocations dissociated in- to partials would definitely show that the glide set is the operating slip plane. Three techniques have been used to image dislocations in ice, trans- mission electron microscopy (TEM), etch-pitting replication, and x-ray to- pography. There have only been two published TEM studies of dislocations in ice [15,16] , and in both cases only grown-in dislocations were examined. Although, in principle, one could strain an ice specimen and then make a TEM foil, specimen preparation from the bulk is quite challenging, as the author can attest from personal experience. How- ever, even if an ice thin foil can be pro- duced and successfully transferred to the TEM, ionization damage in the elec- tron beam is a major problem. In their 1982 TEM study, as Falls et al. attempt- ed to analyze the grown-in dislocations, numerous voids formed in the thin foil and the dislocations rearranged them- selves substantially [16] . Etch-pitting has been used by sev- eral researchers to examine disloca- tions in ice [17,18] . Typically, the ice sur- face is coated with formvar and an etch pit forms where a dislocation intersects the surface. The etch pit itself can be ex- amined or a replica can be made of the etch pit. Etch pitting can be quite use- ful for obtaining an overview of the dis- location distribution over a large area. However, it has two disadvantages: 1) the typical etch pit of >3 µm diameter poses an upper limit to the observable dislocation density of ~1 x 10 11 m -2 , and 2) dislocations that don’t intersect the surface being examined are not observ- able. A further issue is that it is more difficult to generate etch pits from the basal dislocations, which emerge on a non-basal surface. There are several advantages for using x-ray topography for dislocation examination in ice, including the low x-ray absorption, which allows sam- ples 1-2 mm thick to be studied, and that examination can be performed in air at temperatures close to the melt- ing point. Dislocation densities are very low in ice compared to metals, hence the usefulness of this technique for ice whereas it is not very useful for metals. The downside of using conven- tional x-ray topography was revealed by the first use of synchrotron x-ray to- pography to examine ice in 1986 [19] . For synchrotron-based topography, a high- ly-collimated, polychromatic, area-fill- ing, high-intensity beam impinges on the ice specimen resulting in a trans- mission Laue pattern. Within each of the diffracted beams, which are record- ed on a photographic plate, there is an image of the dislocations for that par- ticular diffraction vector. Importantly, exposure times are only 1-2 seconds, enabling in-situ deformation studies to be performed. X-ray topography has been used to examine the dislocation in cores from a number of glaciers and polar ice sheets. While dislocation densities from 1 x 10 6 m 2 to 1 x 10 8 m 2 were observed in ice from the temperate Mendenhall Gla- cier in Alaska [20] , ice from cold polar ice sheets generally has dislocation densi- ties too high to be resolvable [21,22] . The distortion of diffraction spots in Laue x-ray diffraction patterns has also been used to examine dislocation densities in large single grains from the Vostok (Antarctica) ice core [23] . This is a useful averaging technique and candeal with higher dislocation densities than are observable using x-ray topography. However, it does not provide informa- tion on the spatial arrangements of the dislocations. EBSPs have alsobeenused to observe lattice distortion, rotations, and subgrain boundaries in ice in order to understand dislocation activity [24] . Fig. 4 — Electron backscattered diffraction pattern from ice, obtained using 15 keV elec- trons. The ice was held at around -100°C in a vacuum of 5 x 10-4 Pa. Note that the high quality pattern could be obtained as indicat- ed by the third order lines (arrowed) [12] .