Michael C Mound (mike)
Senior Member Username: mike
Post Number: 33 Registered: 7-2007
| Posted on Thursday, September 27, 2007 - 1:35 am: | |
Folks, In my current work regarding silica spectral response in a limestone mineralogical context we have found as follows: SiO2 has no features in the 400-2500 nm region. That is due to the fundamental Si-O stretching and bending vibrations of the chemical bonds manifesting themselves in the 8-25 mm region (8000-25,000 nm). The quantum mechanical selection rules show no allowed Si-O transitions in the lower wavelength region (i.e., between 400-2500 nm). There are other minerals that are commonly mixed in with limestone (the key sampled material), for instance clay, which is a complex silicate and has Al-OH combination features at around 2200 nm and there is an Si-OH transition at 2225 nm that is visible in silica with water inclusions. We run an individual model for each oxide so there is no connection to any other oxide. However, if there is a strong correlation in the reference data (in this case, XRF) there will also be a strong correlation in the model data. That holds true for any of the included PCA's (reported as oxides). What do you think might be a solution for cluster associations (HCA) where there is no water (bound waters) in the lower NIR ranges? Thanks, Mike |