Sun, H., Nelson, M., Chen, F. and Husch, J. 2009. Soil mineral structural water loss during loss on ignition analyses. Can. J. Soil Sci. 89: 603Á610. Water loss from soil minerals has been known to cause errors in the determination of soil organic matter when the loss on ignition (LOI) method is used. Unfortunately, no known published studies reliably quantify the range of structural water in the soil. To do this, 15 common reference minerals were analyzed by LOI to obtain their individual water loss. In addition, 14 upland, loamy soil samples and 3 wetland/hydric soil samples with varied mineral contents were analyzed to collect their X-ray powder diffraction spectra. Based upon X-ray spectra peak intensities, the modal abundance of minerals in each soil sample was determined using the RockJock computer program. The resultant modal weight percentages of all identified minerals in each soil sample were then multiplied by the LOI value for each mineral to obtain the mineral structural water loss (SWL) of that soil sample. For the 17 soil samples analyzed, the range of mineral water loss is 0.56 to 2.45%. Depending on the LOI values of the soil samples, the SWL:LOI ratios range from 0.04 to around 1.00. The SWL:LOI ratios are particularly low for top wetland soil when the LOI value is higher. The ratios are lower for surface soil samples than for subsurface soil samples because of the high LOI values in surface soil samples. Understanding soil mineral water loss and its relation to the LOI patterns from various environments is important for the accurate evaluation of soil organic matter when the LOI method is used.Key words: Mineral, structural water, loss on ignition Sun, H., Nelson, M., Chen, F. et Husch, J. 2009. Perte structurelle d'eau par les mine´raux du sol lors des analyses de perte par calcination. Can. J. Soil Sci. 89: 603Á610. On sait que l'eau perdue par les mine´raux du sol cause des erreurs lorsqu'on s'efforce de calculer la quantite´de matie`re organique dans le sol par la me´thode d'analyse de perte par calcination (LOI). Malheureusement, aucune des e´tudes publie´es ne quantifie de manie`re fiable la variation de l'eau structurelle dans le sol. Pour y reme´dier, les auteurs ont analyse´15 mine´raux de re´fe´rence communs par la LOI afin d'en mesurer la perte d'eau. Ils ont aussi analyse´14 e´chantillons de sol loameux des plateaux et 3 e´chantillons de sol hydrique des terres humides afin de recueillir le spectre de leur poudre par diffraction des rayons X. En partant de l'intensite´des pics obtenus aux rayons X, ils ont e´tabli l'abondance modale des mine´raux dans chaque e´chantillon graˆce au logiciel RockJock. Les pourcentages modaux ponde´re´s des mine´raux identifie´s dans chaque e´chantillon qui en re´sultent ont ensuite e´te´multiplie´s par la valeur LOI de chaque mine´ral, ce qui a donne´la perte structurelle d'eau (SWL) du mine´ral pour l'e´chantillon. Sur les 17 e´chantillons examine´s, la perte d'eau des mine´raux varie de 0,56 % a`2,45 %. D'apre`s la valeur LOI des e´chantillons, le ...

The optical and structural properties of amorphous sputtered films of Ge 2 Sb 2 Te 5 depend strongly on the preparation conditions. Films grown at higher growth rates exhibit greater local strains as indicated by the slope of the optical absorption in the exponential "band-tail" region, but these films also incorporate smaller densities of oxygen impurities. At slower growth rates the band-tail slopes are sharper (smaller local strains) but there is greater oxygen incorporation. We will discuss several experiments that suggest that the local strain relief in the films grown at slower growth rates is due to a greater ability of the atoms to rearrange on the growing surface and not to increased oxygen incorporation. Small angle x-ray scattering experiments show that the films exhibit small elliptical "voids" with long axes perpendicular to the growing surface. The approximate dimensions of these voids are 3 x 20 nm. These films can be switched optically with little change in surface topography as measured by atomic force microscopy. Electron spin resonance measurements indicate that paramagnetic defects exist in some films but are either absent or below the detection limit (~ 10 18 cm -3 ) in most films. The implications of these results for the switching mechanisms will be discussed.