A Statistical Estimation Approach for Quantitative Concentrations of Compounds Lacking Authentic Standards/Surrogates Based on Linear Correlations between Directly Measured Detector Responses and Carbon Number of Different Functional Groups
Table 5
Comparison of percent difference (PD) between the actual and projected response factor (RF) values for all () or three chemical groups () in relation to three major variables of reference VOCs used for prediction: carbon number, molecular weight, and boiling point.
Order
Variables
Functional group
Number of chemicals
PDa
Predictive equationb
P value
Mean
SD
Slope
Intercept
1
Carbon number
All compoundsc
18
17.9
19.0
34,175
−82,472
0.9396
2
Aldehyde
4
9.83
6.55
24,836
−60,290
0.9725
3
Aromatic
6
2.03
0.76
29,456
−43,139
0.9713
4
Carboxylic
4
12.7
4.30
29,818
−58,701
0.9012
Mean
0.9462
0.0162
SD
0.0336
0.0239
5
Molecular weight
All compounds
18
40.6
27.4
2,784
−145,760
0.5445
6
Aldehyde
4
9.79
6.49
1,773
−88,781
0.9726
7
Aromatic
6
1.73
1.22
2,142
−33,879
0.9738
8
Carboxylic
4
12.7
4.28
2,126
−126,763
0.9013
Mean
0.8481
0.0163
SD
0.2052
0.0238
9
Boiling point
All compounds
18
81.5
112
609
31,947
0.1404
10
Aldehyde
4
16.2
10.0
961
−30,600
0.8870
11
Aromatic
6
2.08
0.89
949
58,007
0.9687
12
Carboxylic
4
17.7
7.19
1,358
−158,432
0.7813
Mean
0.6944
SD
0.3772
Percent difference (PD) = /RF(actual) * 100.
bThe predictive equations are derived from linear regression analysis between the number of carbon (x-axis) and actual RF values (y-axis).
cAA is not considered.