limit of detection (lod)
TRANSCRIPT
Aldo Marchetto, Gabriele Tartari & Rosario Mosello
Limit of Detection (LOD)and
Limit of Quantitation (LOQ)estimation and use in the chemical lab
Life+ FutMon - Working Group on QA/QC in LaboratoriesMeeting of the Heads of the Laboratories
12-13 October 2009 in Warsaw
C.N.R. Institute of Ecosystem Study, Verbania (Italy)
e-mail: [email protected] http://www.iii.to.cnr.it
Internal and external QC
calibration blank control charts mean control charts LOD & LOQ result validation using ion balance andconductivity check
Internal quality cntrol
External quality control use of certified standard analyses of certified samples participation to WRTs
Why to estimate LOD and LOQ?
To evaluate the suitability of the used analytical technique and conditions to the aims of the monitoring.
To compare the quality of the determination with other published results in order to evaluate if there is necessity and possibility of improvement.
Aim of this presentation
To show simple approaches in order to encourage laboratories to start to estimate their LOD and LOQ using data they still have or they can collect at zero cost.
Not to present a state-of-the-art review of the literature about this subject.
IUPAC definition
The minimum single result which, with a stated probability, can be distinguished from a suitable blank value. The limit defines the point at
which the analysis becomes possible and this may be different from the lower limit of the determinable analytical range.
The Limit of Detection (LOD), expressed as the concentration, cL, or the quantity, qL, is derived from the smallest measure, xL, that can be
detected with reasonable certainty for a given analytical procedure.
The value of xL is given by the equation
xL = xbi + ksbi
where xbi is the mean of the blank measures, sbi is the standard deviation of the blank measures, and k is a numerical factor chosen
according to the confidence level desired
IUPAC definition1997
IUPAC definition
2002
ISO 13530 1997Water quality - Guidance on analytical quality control for
chemical and physicochemical water analysis
1997
Eurachem definition
1998
EPA
2004
Document 815-R-05-006
EPA Document 815-R-05-006 2004
EPA Document 815-R-05-006 2004
EPA Document 815-R-05-006
2004 LCMRLLowest Concentration Minimum Reporting Level
MRL Minimum Reporting Level UCMR Unregulated Contaminant Monitoring Regulation
Standard Methods 21° Ed. 2005: definitions
2005
Standard Methods 21° Ed. 2005: definitions
LOD LOQ
Detection Limit in analysisNORDTEST definition
2007
The Limit of Detection (LOD) expressed as the concentration is derived from the smallest measure, xL, that can be detected with reasonable certainty for a given
analytical procedure.
xL = xbi + ksbi
where xbi is the mean of the blank measures, sbi is the standard deviation of the blank measures, and k is a numerical factor chosen according to the
confidence level desired (2 < k < 3)
To obtain LOD, signals are transformed inconcentrations using the
calibration curve
SD of blank samples
signal
Calibration curve N-NH4 spectrophotometry indophenol blue
5 cm optical path
-0.1
0.1
0.3
0.5
0.7
0.9
0.000 0.050 0.100 0.150 0.200
concentration (mg N-NH4 /L)
sign
al (A
)
-0.1
0.1
0.3
0.5
0.7
0.9
0.000 0.050 0.100 0.150 0.200
concentration (mg N-NH4 /L)
sign
al (A
)
Calibration curve N-NH4 spectrophotometry indophenol blue
5 cm optical path
-0.01
0.00
0.01
0.02
0.03
0.04
0.05
0.000 0.002 0.004 0.006 0.008 0.010
concentration (mg N-NH4 /L)
sign
al (A
)
Calibration curve Chloride by Ion Chromatography determinationDionex AG19-AS19-AAES 100µL sample loop
-0.5
0.0
0.5
1.0
1.5
2.0
0 0.5 1 1.5 2
concentration (mg Cl /L)
sign
al (p
eak
area
)
Calibration curve Chloride by Ion Chromatography determinationDionex AG19-AS19-AAES 100µL sample loop
-0.5
0.0
0.5
1.0
1.5
2.0
0 0.5 1 1.5 2
concentration (mg Cl /L)
sign
al (p
eak
area
)
-0.05
0.00
0.05
0.10
0.15
0.20
0.00 0.05 0.10 0.15 0.20
concentration (mg Cl /L)
sign
al (p
eak
area
)
In practice…If you have mean control charts at different concentration,
use them to evaluate LOD and LOQ
else, if calibration slope is very stable (i.e. spectrophotometry)
use SD of the blanks (if possible)
or of a control sample of low concentration
or of the lowest standard
else,
apply Hubax-Vos method on as many calibrations possible,
or use the variability of the calibration standards
SD of blank samplesN-NH4 spectrophotometry indophenol blue
5 cm optical path
Calibrationmg N/L = 0.196 A + 0.0006
LOD 0.005 mg N/L = 0.196 x 0.021 + 0.0006
0.00
0.01
0.02
0.03
0.04
0.05
0.06
gen-05
giu-05ott-
05feb
-06ap
r-06
ago-06
mar-07
ago-07
ott-07
apr-0
8se
t-09
Abs
orba
nce
Mean
dataMean = 0.025 mA
SD = 0.007 mAxL = xbi + ksbi
xL = 0.025 + 3 x 0.007 = 0.046 mA
In the daily determinations and in the calibration, blank mean is subtracted from the values, then for estimating LOD
use 3SD (0.021 A), not mean + 3SD (0.046 A)
LOQ 0.014 mg N/L = 0.196 x 0.070 + 0.0006
Variability of the lowest standardChloride by Ion Chromatograpy determinationDionex AG19-AS19-AAES 100µL sample loop
Ratio Factor = peak area / ST concentrationRF = 0.057/0.05 = 1.14
LOD 0.018 mg Cl/L = 0.021/1.14
Lowest standard0.05 mg Cl/L
Peak area mean = 0.057Peak area SD = 0.007
xL = xbi + ksbi
for lowest standardxL = ksbi
xL = 3 x 0.007 = 0.0210.03
0.04
0.05
0.06
0.07
0.08
0.09
nov-05feb
-06ap
r-06
ago-06
nov-06feb
-07
mag-07
ago-07
nov-07feb
-08giu-08
ott-08
gen-09
mar-09
giu-09
Peak
are
a
Mean
data
LOQ 0.061 mg Cl/L = 0.070/1.14
Variability of calibration standards Chloride by Ion Chromatography determinationDionex AG19-AS19-AAES 100µL sample loop
LOD 0.0002 mg Cl/L
0
5
10
15
0 2 4 6 8 10 12Cl (mg L-1)
RSD
pea
k ar
ea RSD = 5.73 (mg/L)-0.2
LOD = RSD 33% = (5.73/33) (1/0.2)
LOQ = RSD 10% = (5.73/10) (1/0.2) LOQ 0.062 mg Cl/L
Standardmg/L
Peakarea
RSD
0.05 0.058 13.8
0.10 0.108 8.60.20 0.218 5.8
0.50 0.558 5.52.00 2.277 4.65.00 5.701 4.3
10.00 11.439 4.2
n >300 and 2-3 years of calibrations
Statistics mg L-1
Average 0.508Standard Deviation 0.013Standard Deviation recommended 0.025RSD % 2.7RSD % recommended 5.0Minimum 0.47Maximum 0.54N. data 115
0
5
10
15
20
25
0 1 2 3 4 5 6 7 8 9 10 11mg L-1
R.S
.D. C
ontr
ol C
hart
s
Using control chart
ChlorideIon Chromatography
C.N.R.- I.S.E.Verbania - Italy
0.40
0.42
0.44
0.46
0.48
0.50
0.52
0.54
0.56
0.58
0.60
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 101
106
111
mg
L-1
UWL
LWL
UCL
LCL
RSD = 3.37 (mg/L)-0.46
Single control chart in use from nov 2006 to aug 2007
All control charts used in the lab for IC determinations (in 20 years)
0
5
10
15
20
25
0 1 2 3 4 5 6 7 8 9 10 11mg L-1
R.S
.D. C
ontr
ol C
hart
sChloride by Ion Chromatography determination (100µL sample loop)
LOD = RSD 33% = (3.37/33) (1/0.46)
LOQ = RSD 10% = (3.37/10) (1/0.46)
RSD = 3.37 (mg/L)-0.46
LOD from control charts
LOD 0.007 mg Cl/L
LOQ 0.094 mg Cl/L
Hubaux-Vos
LOD from daily calibration (Hubaux-Vos)
Example from Dionex Chromeleon Software
Chloride by IC100µL sample loop
Range 0.05 – 10 mg/L7 standards in duplicate
daily calibration
LOD from daily calibration (Hubaux-Vos)
LOD 0.027 mg Cl/L
Example from Dionex Chromeleon Software
Chloride by IC100µL sample loop
Range 0.05 – 10 mg/L7 standards in duplicate
daily calibration
Range 0.05 – 10 mg/L7 standards
More than 300 calibrationsin 2-3 years
Standardmg/L
Peakarea
RSD
0.05 0.058 13.80.10 0.108 8.60.20 0.218 5.80.50 0.558 5.52.00 2.277 4.65.00 5.701 4.3
10.00 11.439 4.2
Chloride by Ion Chromatography determinationDionex AG19-AS19-AAES 100µL sample loop
LOD from average calibrations (Hubaux-Vos)
LOD 0.038 mg Cl/L
-0.05
-0.03
-0.01
0.01
0.03
0.05
0.07
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08Concentration
Sign
al
Data
Upper prediction limit
Upper confidence limit
Regression line
Lower confidence limit
Lower prediction limit
L.O.D.
yc
Chloride by Ion Chromatography determinationDionex AG19-AS19-AAES 100µL sample loop
Values obtained:
Lowest standard variability 0.018 0.061
Control charts RSD 33/10% 0.007 0.094
Standards variability RSD 33/10% 0.0002 0.062
Hubaux-Vos daily calibration 0.027 0.089
Hubaux-Vos average calibrations 0.038 0.073 *
Results obtained (mg Cl / L) LOD LOQ
* Gibbons’s AML
In practice…
If you have mean control charts at different concentration,
use them to evaluate LOD and LOQ
else, if calibration slope is very stable (i.e. spectrophotometry)
use SD of the blanks (if possible)
or of a control sample of low concentration
or of the lowest standard
else,
apply Hubax-Vos method,
or use the variability of the calibration standards
Conclusions Quality Control is an important tool to criticize analytical
activity, to improve it and to assure that it is suitable for
the purpose of monitoring.
The estimation of LOD and LOQ is an important part of
Quality Control.
LOD and LOQ can be simply evaluated using results of
control charts, blank samples or calibration curves. If you do not store these data yet, please store them
anywhere and in a few years you will have years of data...
Thank you for your attention