Contrasting Size Distributions of Chondrules and Inclusions in Allende Cv3

Introduction: There are several leading theories on the processes that led to the formation of chondrites, e.g., sorting by mass [1,2], by X-winds [3], turbulent concentration [4], and by photophoresis [5]. The juxtaposition of refractory inclusions (CAIs) and less refractory chondrules is central to these theories and there is much to be learned from their relative size distributions. There have been a number of studies into size distributions of particles in chondrites but only on relatively small scales primarily for chondrules, and rarely for both CAIs and chondrules in the same sample (e.g. [6-17]). We have implemented macro-scale (25 cm diameter sample) and high-resolution microscale sampling of the Allende CV3 chondrite to create a complete data set of size frequencies for CAIs and chondrules. Methods: As part of an ongoing study to characterize nebular components in carbonaceous chondrites, CAIs and chondrules were characterized in X-ray phase maps obtained by scanning electron imaging (SEM) for seven Allende samples (including 0.50 cm, 0.68 cm, 0.70 cm, 0.77, cm, 0.72 cm, 0.79 cm, 0.80 cm sized pieces; 2 that were new and from a region of Allende from which a large (~ 25 cm) slab had previously been characterized [18] and 5 that were obtained previously by [19]). The SEM data were then combined with a representative section of the ~25 cm slab. The SEM data allow for accurate phase recognition and size characterization at the smallest sizes and the large size of the slab allows analysis at a much larger length scales. The latter should result in a better representation of larger CAIs, whose feature may not be fully apparenent from studies of smaller samples. Data Preparation. To create an accurate representation of the CAI and chondrule size distributions at smaller and larger scales, analyses of images of different scales were needed. Samples were imaged by SEM at a resolution of 3.34 μm/pixel to produce element maps for further CAI and chondrule identification. These SEM images were separately digitized by 4 researchers. The mosaic image (referred to as ‘macroscale’) was also digitized to gather data on the larger CAIs that may not be fully represented in any of the 7 micro-samples. Image Analysis. Digitized images were run through ImageJ and filtered at a confidence level of >80,000 um for CAIs. Digitized SEM images, including the mosaic image were run through ImageJ without filtering. All images were analyzed in ImageJ by the same process with the exception of the smallest, difficult to identify sizes in the macrophotography. The large slab sample was extensively imaged at a resolution of 13.88 μm/pixel. Both sides (one designated ‘LH’ and the other designated ‘RH’) were imaged by macrophotography and mosaic images were created using Adobe Illustrator. All image files were aligned so that each of the ~400 individual image files was in the same orientation. The mosaic images were divided into 9 equal size sections of ~ 1,000,000,000 um to facilitate easy comparison to other samples and other studies. Four researchers separately digitized CAIs and chondrules in each section. Compiling of Data. Frequency analysis was completed for the ImageJ results for each set of images (SEM and macro). The frequency data sets were merged to create an accurate size distribution of the Allende CAIs and chondrules. The SEM data and macro data were scaled up and down, respectively, to produce distributions on a ~ 1,000,000,000 um scale in order to compare with the micro data. After some iteration, it was determined that the best merging point for the SEM and macro data would be at the 225 μm (major axis) size limit. Particle size distributions below 225 μm were calculated from the SEM images while everything above was calculated using the macro images.