CHEM 4430

PRACTICES IN TESTING LABORATORY

Analytical methods
• Classification
• Sources of Methods
• Method Validation
Classification: - analytical principle
- analytical method
- analytical procedure
Ask to determine the heavy
metals as well as the
pesticides in Chinese Herbal
Medicine.

How to find the suitable methods?

AS EPA APHA
ASTM

ISO EN I h
In-house

AOAC BS Literature
Source of analytical methods
Standard methods
The test methods were subjected to intensive investigations and with proven
accuracy and precision before being awarded ‘standard’. Whenever possible,
laboratories should use standard methods.
methods

ASTM: American Society for Testing and Materials

AOAC: Association of Official Analytical Chemist (Food Analysis)
APHA
APHA: A
American
i P
Public
bli Health
H l h Association
A i i
(Standard Methods for the Examination of Water and Wastewater)
ISO: International Organization for Standardization
BS: British Standards
EN
EN: M h d published
Methods bli h d by
b CEN (European
(E Committee
C i for
f Standardization)
S d di i )
e.g. EN71 Toy Testing
AS: Australian Standards
DIN: German Standards

A list of organizations that provide standards:

http://www info gov hk/itc/eng/quality/psis/std body shtml#
http://www.info.gov.hk/itc/eng/quality/psis/std_body.shtml#
 http://www.astm.org
p g

American Society for Testing and Materials

Organized in 1898, ASTM is one of the largest voluntary standards development organizations in
the world.
ASTM is a not-for-profit organization that provides a forum for the development and publication
of voluntary consensus standards for materials, products, systems and services. More than
32,000 members representing producers, users, ultimate consumers, and representatives of
government and academia from over 100 countries develop documents that serve as a basis for
manufacturing, procurement, and regulatory activities.
ASTM develops standard test methods, specifications, practices, guides, classifications, and
terminology in 130 areas covering subjects such as metals, paints, plastics, textiles, petroleum,
construction, energy, the environment, consumer products, medical services and devices,
computerized systems, electronics, and many others. ASTM Headquarters has no technical
research or testing facilities; such work is done voluntarily by the ASTM members located
throughout the world.
More than 10,000 ASTM standards are published each year in the 73 volumes of the Annual
Book of ASTM Standards.
Standards These standards and related technical information are sold throughout
the world.
 Association of Official
Analytical Chemists (AOAC)

AOAC INTERNATIONAL is committed to the development and validation of

methods in the analytical sciences and to the improvement of quality assurance
procedures
d i laboratories.
in l b t i
During the past 100 years, AOAC evolved from a group of chemists in the U.S.
Department of Agriculture into an independent scientific Association of analytical
scientists
i ti t with ith members
b l t d throughout
located th h t theth world.
ld Today,
T d AOAC
INTERNATIONAL is the leader in providing validated methods, proficiency test
samples, accreditation criteria, and scientific information to industry, government
agencies and academic institutions.
agencies, institutions
AOAC is dedicated to methods validations, improved laboratory management, and
the professional development of analytical scientists.
Standard Methods for the Examination
of Water and Wastewater (20th Edition)
Edited by Lenore S. Clesceri, Arnold E.
G
Greenberg
b
and Andrew D. Eaton

Published by the American Public Health Association (APHA)

Approved by the U.S. Environmental Protection Agency (EPA)
Official methods
Methods mandated by government organization, e.g. methods from Environmental
Protection Agency (EPA),
(EPA) European Commission.
Commission
European Standard Titles
Standards reference
Bodies
CEN EN 71 Safety of toys:
CEN EN 71-1:1988
Part 1: Mechanical and physical properties
CEN EN 71-1:1998
Part 1: Mechanical and physical properties
CEN EN 71-2:1993
Part 2: Flammability
CEN EN 71-3:1994
71 3:1994
Part 3: Migration of certain element
CEN EN 71-4:1990
Part 4: Experimental sets for chemestry and related
activities
CEN EN 71-4:1990/A1 Part 4: Experimental sets for chemestry and related
activities
CEN EN 71-5:1993 Part 5: Chemical toys (sets) other than
experimental sets
CEN EN 71-6:1994 Part 6: Graphical symbol for age warning labelling
CENELEC EN 50088:1966 Safety of electric toys
CENELEC Amendment A1:1996 Safety of electric toys
at EN 50088:1996
CENELEC Amendment A2:1997 Safety of electric toys
at EN 50088:1996

CEN - European Committee for Standardization (http://www.cenorm.be/)

(http://www cenorm be/)
CENELEC - European Committee for Electrotechnical Standardisation
Literature methods
Methods published in the general-purpose analytical journals,
such as Analytical Chemistry, and the specialized journals such as
those dealing with food,
food pharmaceuticals,
pharmaceuticals plastics,
plastics metals,
metals paper,
paper
etc. The patent literature and also the methods developed by the
instrument manufacturers are also the source of literature
methods Validation is required in the laboratories.
methods. laboratories

In-house developed methods

Requires development and validation
M h d Validation
Method V lid i
- completion of an analytical development
- proof of proficiency when adopting an externally development method
((even standard methods have to be validated))
- business card of laboratory (formal requirement for accreditation and
show competence of a lab)

Analytical How qualified

Instrumentation a laboratory

Type, Extent and

d Adequate instrumental equipment and
Conduct of Correct a motivated management and
Validation competent employees
D l
Development
t off Analytical
A l ti l Procedures
P d
Standard
Methods Smaller modifications and Save
Scientific adaptation of equipment expenditure
P bli ti
Publications

Important especially for trace analysis

EN45001, Not to accept measurement commissions that do not lead

to meaningful results Insufficient sampling
Bad transport conditions
No stabilization of the sample or analyte
Under Statistical Control

- mean values (at low & high analyte conc.)of measurements constant for a long
period
- precision is sufficient and constant
- all other performance characteristics are constant
- valid
lid off calibration
lib i samplesl andd for
f analyte
l in i the
h corresponding
di matrix
i
Performance characteristics

- precision
- accuracy
y
- selectivity or specificity
- linearity & working range
- sensitivity
- limit of detection
- limit of quantitation
- ruggedness
d
Precision and Trueness
Acceptable Trueness
Good Overall
+ Accuracy
Good Precision
Precision
ii A study
d bby HHorwitz
i on
the cumulative results
s of collaborative
× 100%
x studies.
Repeatability
- precision under repeatability conditions
- the closeness of agreement
g between the results of independent
p
measurements of the same analyte carried out under the same method of
measurement, the same observer, the same measuring instrument, the
same location, the same conditions of use and repetition over a short
period of time.
- If possible, at least 8 measurements should be performed.

Reproducibility
- precision under reproducibility conditions
- the closeness of the agreement between the results of measurements of
the same analyte in distinct subsamples of a test material, where the
individual measurements are carried out changing conditions such as:
observer, measuring instrument, location, conditions of use, time, but
applying the same method.
Proof of Accuracy

(i) analyze
l Reference
R f Material
M t i l
or (ii) apply a totally different measurement principle
(even the sample preparation steps have to be different)

Calibration - (reliable)
- standards are free of errors
- the same precision over the whole working range
- the model is suitable, “straight line or curve”
- errors related to signals are randomly distributed (normal)
In practice, examine interdependence of precision and concentration
Limit of Detection
- the lowest concentration of analyte in a sample that can be detected,
b not necessarily
but il quantified
ifi d under
d theh statedd conditions
di i off the
h test.
(NATA Tech Note#13)

Any signal above the 3s line will be most

likely from the analyte in the sample.

+3σ 99.87% of the

mean sample
signals will be
blank value -3σ
3σ within ±3σ.

LOD = concentration of analyte gives a signal equal to 3 x

standard deviation
de iation of the blank
Li i off Quantification
Limit Q ifi i

- the lowest concentration of an analyte that can be determined with

acceptable precision (repeatability) and accuracy under the stated
conditions of the test.
(NATA Tech Note#13)

LOQ = concentration of analyte gives a signal equal to 10 x

standard deviation of the blank

How to obtain the LOD and LOQ for an

analytical method?
S l ti it or Specificity
Selectivity S ifi it

- the ability
y of a method to determine accurately y and
specifically the analyte of interest in the presence of other
components in a sample matrix under the stated
conditions of the test (NATA Tech Note#13)

Sensitivity

- the change of the signal from the instrument divided by

the corresponding change in the amount of the analyte

How to know?
Recovery
- the recovery is the amount of analyte that is determined by the method
divided by the amount of analyte that is present in a test sample.

fortified or spiked sample

Blank Analyte
Sample M '　
+ determined R= x 100%
y
Analyte M’ M
added
(M)

(certified) reference sample

Analyte
y Analyte
originally N'　
present determined R= x 100%
(N) N’ N
Ruggedness
- quality of data is independent of small operating variations of the
procedure
Two ways to determine ruggedness
(1) interlaboratory study (n≥8)
- follow one and the same procedure
high expenditure and not always possible

(2) one’s own lab: planned series of experiments by varying the most
important experimental parameters within the determined limited
tolerance proposed by AOAC
tolerance,

efficient and with minimum of assays

7 factors - 8 experiments
8-11 factors - 12 experiments
12 15 factors
12-15 f t - 16 experiments
i t
 Acceptable region

A large
l number
b off experiments
i t may be
b required.
i d Experimental
E i t l designs
d i
that enable one to extract as much information as possible from a
minimum number of experiments is required.
 A B,
A, B C,
C D,
D E,
E F andd G
7 Factors at two levels
a, b, c, d, e, f and g

Need 27 (128) experiments to study all the conditions

Factor Observed
1 2 3 4 5 6 7 Result
No. Expt.
1 A B C D E F G s
2 A B c D e f g t
3 A b C d E f g u
4 A b c d e F G v
5 a B C d e F g w
6 a B c d E f G x
7 a b C D e f G y
8 a b c D E F g z
Effect when changing from A to a:
Similar for other
= (s + t + u + v)/4 – (w + x + y + z)/4 effects!
Ruggedness study:

(1) ddefine
fi variable
i bl - most probably
b bl influence
i fl the
h result
l
(2) maximum tolerances for each variable in routine work
(3) draw up an adequate experimental plan
(4) perform and evaluate experiments in order to determine
i fl ti l factors
influential f t
(5) take measures
(i) keep
k exactly
tl within
ithi the
th target
t t values
l
or (ii) further optimization to reduce the influence of the
p
correspondingg variables
E ample:
Example:

A,B
C
D
E
F
G
H

resolution
The effect of factor A (initial concentration of acetonitrile):

(-4.6+9.5-4.7+7.5-5.6+9.7-1.1+8.5-5.3+9.5-2.1+7.1-5.4+7.7-1.3+6.7)/8
= -0.47

The effect of factor B (final concentration of acetonitrile):

(-4.6-9.5+4.7+7.5-5.6-9.7+1.1+8.5-5.3-9.5+2.1+7.1-5.4-7.7+1.3+6.7)/8
= -0.70
0 70

MeCNinitial
i iti l = -0
0.472091
472091 MeCNfinal
fi l = -0
0.69709
69709
molarity = 4.57791 pH = -2.22209
flow rate = 0.60291 temperature
p = -0.59709
injection vol. = 1.04655 wavelength = 0.40291
Relationship
R l i hi between
b purpose off the
h procedure
d andd procedure
d
characteristics to be determined
F
Frequency off V
Validation
lid ti
Comparison of Methods
Total new procedures - extensively validation
Less important
L i Influence the data
changes and
modification
accuracy and
precision ?
New
Perform Method in parallel
Old

First use
t-test for mean
F-test for precision

Then do the regression

Then,
The F test
The F test is based on the random distribution and serves
to determine whether two sets of data have significantly
different standard deviations. The two sets of data may y
be from different

• Analytical methods
• Analyses
• Samples
 2
s S1 standard deviation of data set 1
F= h1
where
2
s1 > s2
2 2

s 2
S2 standard deviation of data set 2

The calculated F value is then compared with Fc value from the F-test
table using the degree of freedom of the two data sets and the confidence
level.
If F > Fc, then the difference in standard deviation of the two data set is
significant.
Example of F-test
A reference analytical method, for which much experimental evidence is
at hand, shows a standard deviation sr=0.037 with nr=50. A few tests with
appurportedly
p y better method y
yield st=0.029 with nt=5.

(0 037/0 029)2 = 1.28

F = (0.037/0.029) 1 28
Fc = 5.7
On this evidence,
evidence no improvement is detectable.
detectable
For Fc < F, need nt>45 to bring down Fc, or bring down st to <0.0155.
Degreee of freedom oof test data set (denominator)) F-test Table for 95% confidence level

Degree of freedom of reference data set (numerator)

Tables of other confidence levels can be found in CRC

Handbook of Chemistry and Physics or many other
reference books
Example on Method Development and Validation
In house Method
In-house
Determintaion of quinoxaline-2-carboxylic acid (the major metabolite of
carbadox) in porcine liver by GC-ECNI-MS
Sin et al., Analytica Chimica Acta 508, 147 (2004).

Monooxy and Deoxy

metabolites

Carbadox
- monooxy andd ddeoxy metabolites
t b lit are suspected
t d carcinogens.
i
- banned in use as animal feeds in EU
- approved for use at 55μg/kg for swine up to 35kg bodyweight
- in HK, the maximum residue limit in porcine liver is set at 30 μg/kg
- quinoxaline-2-carboxylic acid as marker compound
Specificity
- analyzing 50 blank liver samples with known amount of metabolites of
carbadox spiked
No interference peaks on chromatograms

R
Recovery
- three sets, each of six, blank porcine liver samples fortified with 0.5, 1
and 1.5 times MRL of QCA (30μg/kg)

99%

Repeatability
- three sets, each of six, blank porcine liver samples fortified with 0.5, 1
and 1.5 times MRL of QCA (30μg/kg)

Intra-day precision 8.0%

Within-Lab
i hi b Reproducibility
d ibili
- fortified samples were analyzed by two independent operator on
different days

Linearity
- blank liver samples spiked with 0,
0 15,
15 30,
30 45,
45 60μg/kg

Show good linearity (r2 = 0.9998)

LOD and LOQ

- calculated concentration determined for the blank samples (n
(n=50)
50)

LOD = 3 x σ of blank = 0.2μg/kg

LOQ = 10 x σ of blank = 0.7μg/kg
Ruggedness
Results for Ruggedness Test