Report on the analysis of a mystery object/substance (Instrument Methods of Analysis)
TECHNIQUES AND INTERPRETATION IN INSTRUMENTAL METHODS OF ANALYSIS
and
INSTRUMENTAL METHODS OF ANALYSIS
AN UN-ASSESSED EXERCISE IN HANDLING REAL DATA
AIM: To provide students with real data to manipulate and investigate data significance.
Copies of the data sets can be found on Blackboard
Part 1) Limit of Detection (LOD) and Precision and Accuracy
Sample preparation and background.
The samples were crushed and 0.1 g was dissolved in a mixture of HF and HClO4, these acids were then evaporated and the
residue re-dissolved in HNO3 and made up to 100 ml with deionised water. The samples were then analysed by ICP-MS. The data
have already been quantified (you will gain more experience of this in the ICP-MS workshop).
As well as the samples there are two important analyses that you will look at today:
• Method blank: This is a sample containing nothing but deionised water and the same amount of HNO3 as the samples.
This was repeatedly analysed ten times. The data for these analyses are shown in table 1.
• Primary standard reference material: “1c argillaceous limestone”. This is a finely ground, homogeneous sample of
accurately known composition (the certified composition is shown in table 2). It was analysed as an unknown three times.
Limit of Detection (LOD)
The limit of detection tells the user the smallest possible amount that the instrument can determine from background noise.
One method for calculating the LOD is to run a blank sample ten times and the LOD is three times the standard deviation.
Element Na (ppb) Mg (ppb) Sr (ppb)
Method blank 52.39 3.93 0.50
Method blank 53.07 4.26 0.48
Method blank 52.39 4.21 0.50
Method blank 58.26 4.77 0.51
Method blank 53.61 4.24 0.50
Method blank 54.16 4.39 0.50
Method blank 58.39 4.55 0.49
Method blank 62.76 4.53 0.51
Method blank 66.45 4.64 0.50
Method blank 61.26 4.52 0.50
Mean 57.27 4.40 0.50
Standard deviation
LOD (ppb)
Limestone SRM 1804.32 20021.62 1102.70
Limestone SRM 1640.00 19027.03 1058.70
Limestone SRM 1618.38 19167.57 1058.05
Mean
Standard deviation
Table 1. Quantified data (ppb) for the method blank and the 1c Standard reference material, argillaceous limestone.
Q1) Fill in the standard deviation and LOD values in table 1. Comment on the results.
Na (ppb) Mg (ppb) Sr (ppb)
Certified value (ppb) 1483 25327 2536
uncertainty as % 50 9.5 0.067
Table 2. Certified values for selected elements in the 1c Standard reference material, argillaceous limestone. (Converted
from % weight oxide to ppb element).
Precision and Accuracy
Precision: Refers to the reproducibility of the measurement. A precise measurement is one which when re-determined gives the
same value.
Q2) Look at the data for the three analyses of the standard reference material (a graph or two might help) and comment on the
values.
Accuracy: How close a value is to the “true” value. An accurate reading is close to the true value.
Q3) Comment on the accuracy of the standard reference material data.
Q4) What happens if you consider errors on the measurement and on the certified value?
Part 2) Peak Data Exercise
Although we use mass spectrometers (to measure mass) or a spectroscope (to measure the spectrum) of a sample, what the
instrument obtains is not the full mass or spectrum range, but rather counts (which is the response of the detector) within
specific instrument channels. Each channel covers a portion of the mass or spectral range.
An analogy is a long row of polystyrene cups each one fitted with a different size mesh (each polystyrene cup being a
channel) into which is collected falling dust. The amount of dust in each cup is recorded as counts.
The data given in table 3 shows the growth of a peak, e.g. from XRF, over time from t= 20 to t= 300 seconds. The data
comprises 15 channels and the number of counts accumulated per channel over time, for t= 20 to t= 300 seconds. You need to do
four exercises:
Channel
Time (s) 1 2 3 4 5 6 7 8 9 10 11 12 13 14
15
20 10 11 9 8 9 12 12 9 11 11 9 12 10 10 8
40 15 11 9 10 10 13 15 14 10 11 10 14 12 11 10
60 25 14 12 19 15 20 25 22 14 13 12 16 18 15 19
80 40 31 35 32 37 38 42 39 32 35 31 26 28 37 38
100 45 46 40 41 47 58 60 57 41 40 46 48 41 43 44
120 75 75 79 78 79 78 80 79 78 79 75 75 71 70 69
140 100 99 101 99 107 115 120 116 99 101 99 100 101 99 98
160 120 121 118 121 140 160 170 159 121 118 121 120 119 119 118
180 130 132 132 132 150 195 210 197 132 132 132 130 130 132 131
200 160 161 158 159 180 265 350 258 159 158 161 160 161 162 162
220 170 171 172 169 220 465 550 450 169 172 171 170 168 167 165
240 190 192 194 195 320 647 700 654 195 194 192 190 185 184 182
260 210 212 211 212 450 712 850 699 212 211 212 210 211 211 209
280 230 231 229 230 568 880 1000 824 230 229 231 233 234 233 232
300 250 265 260 500 700 1000 1298 1000 690 500 270 250 260 265 255
Table 3. Number of counts accumulated per channel over 20 to 300 seconds
Q1) A peak can be confirmed, and a percentage determined for that element when the peak height achieves twice the background.
After what time interval is the peak confirmed?
For this exercise identify the peak centre, i.e. channel number, calculate an average background reading for each time
interval, e.g. by choosing 3 background readings on either side of the peak, identify at what time the peak height counts in
the central channel achieves double the background. Note the background changes at each time interval.
Q2) At t=300, define the width of the peak, in channels.
Q3) At t=300 calculate the full width of the peak at half-maximum height (the FWHM)
Q4) At t=300 calculate the net peak integral, i.e. peak are minus background for the width of the peak