Solution 6.3 - Chemometrics: Data Analysis for the Laboratory and Chemical Plant

Skip to Navigation

Education Article

  • Published: Jan 1, 2000
  • Channels: Chemometrics & Informatics

1. The graph is given below.

2. The labelled scores plot is as follows.

 

The loadings plot is as follows.

 

The masses seem clustered. Those in the bottom right hand corner of the loadings plot appear to not to correspond to any compound and probably represent background or noise ions. Those in the bottom left hand corner are likely to correspond to one of the main compounds in the mixture.

3. Scaling the data and using all points in time gives the following graph.

There does not seem to be a particularly obvious pattern. However the first 3 points in time clearly dominate the plot. When scaling each row to a constant total, noise points have equal weight to those representing true compounds, and so can severely distort the analysis, thus removing them produces plots that are easier to interpret.

The scores plot is given below after the first three points have been removed. It is still slightly hard to interpret, but would suggest that between points 11 and 21 in time there are two components.

 

The loadings plot is given below.

 

 

The loadings appear mainly to lie on an approximate circle, which is an indication of two components. Furthermore there is now some more definitive groups.

4. The 10 masses are as follows.

96

78

95

136

154

155

41

172

97

68

 

5. The labelled scores plot is as follows.

The corresponding loadings plot is given below.

It is now quite clear that there are two components from the scores plot, with maximum purities at around times 5 and 20. The loadings plot now falls very clearly on a circle with m/z 68 and 154 corresponding closely to the fastest eluting component, and m/z 78 and 172 to the slowest eluting one. Because masses of low standard deviations probably represent noise, we now retain variables that arise from the compounds in the chromatogram. Because there are only two components in the data, the standardised loadings should fall on a circle, the reason for distortion in the loadings plot of question 3 is because there are a number of variables that really correspond to noise and will not fit into this pattern.

Social Links

Share This Links

Bookmark and Share

Microsites

Suppliers Selection
Societies Selection

Banner Ad

Click here to see
all job opportunities

Copyright Information

Interested in separation science? Visit our sister site separationsNOW.com

Copyright © 2017 John Wiley & Sons, Inc. All Rights Reserved