SOLUTION AND TIME OF FLIGHT LASER ABLATION ICP-MS MEASUREMENT OF MINOR AND TRACE ELEMENTS: AN EXAMPLE FROM DEATH VALLEY, CALIFORNIA

Chemically similar tephra beds may be difficult to distinguish with major elements measured by the electron microprobe. These difficulties present a growing need for minor and trace element analysis of individual glass shards, or small samples of shards. Solution and laser ablation inductively coupled mass spectrometry (ICP-MS) are possible methods for trace element measurement. To test the utility of ICP-MS we analyzed glass shards from a number of Death Valley tephra beds by both solution (S-ICP-MS) and time of flight laser ablation ICP-MS (TOF-LA-ICP-MS) and compared these data using the similarity coefficient commonly used in tephrochronological studies. Death Valley has multiple discontinuous tephra beds with glass shard compositions of Long Valley-Glass Mountain chemical affinities (families) ranging from the 3.35-Ma lower Mesquite Springs tuff to the 0.76-Ma Bishop ash bed. Major element similarity coefficients for this ash group are commonly >0.93, thus requiring minor and trace elements for correlation. In S-ICP-MS, dissolved, homogenized, bulk glass separates are used, producing “averaged” results, which may be spurious for tephra beds with a heterogeneous shard composition. TOF-LA-ICP-MS ablates along a designated pathway and rapidly collects data (~20 sec) before the shards are destroyed. Similarity coefficients using ten elements (Ce, Sm, Tb, Yb, Lu, Fe, Rb, Cs, Hf, Th) frequently used in tephrochronological studies measured by S-ICP-MS allow differentiation between tephra in the Upper Glass Mountain, Lower Glass Mountain and Mesquite Spring tephra families. Upper Glass Mountain and Mesquite Spring families are discernible using similarity coefficients measured by TOF-LA-ICP-MS. Individual tephra beds within each family may be correlated by rare-earth fractionation and variation diagrams. Thus, both S-ICP-MS and TOF-LA-ICP-MS produce data suitable for inclusion in correlation databases.