CH915: Elemental Analysis

 Module leader: Dr. Claudia Blindauer

 Lecturers:
 Dr. Claudia Blindauer
 Dr. John Fenlon (Statistics)
 Dr. Andrew Mead (Warwick HRI)
 Lab classes:
 Dr. Abraha Habtemariam
 Book recommendations, e.g.:
 D.C. Harris: Quantitative Chemical Analysis
 Vogel’s textbook of quantitative chemical analysis
 For the entire course: Skoog, Holler, Nieman: Principles of Instrumental
Analysis
 Aims of the module
 Introduce the Analytical Process
 Introduce concepts for quantitative analysis
 Including Statistics for Data Analysis
 Enable professional data analysis
 Introduce important methods for elemental
analysis of liquid and solid samples
 Enable selection of the best possible method
for a given analysis problem
 Enable to design experiments
 Module Overview
 5 sessions on chemical aspects of
quantitative and elemental analysis (C.
Blindauer, see handout)
 4 lab classes (A. Habtemariam)
 8 sessions on understanding data and
statistical aspects of quantitative analysis
(J. Fenlon, A. Mead, J. Lynn) – together
with MAOC and Systems Biology students
What is elemental analysis
and where is it applied ?
 What is Elemental Analysis ?
 Determine the elemental composition of
material
 Qualitative
 Quantitative
 CHNX: Combustion analysis for
verification of compound identity
 Other elements
 Elemental Analysis is applied in:
 Materials Sciences
 Metallurgy, glass, ceramics, cements, superconductors,
microelectronics…
 Geosciences
 geochemistry, mineralogy, geochronology…
 Environmental Sciences
 Biological Systems and Medicine

 In Industry:
 Quality control: Establish that produced material conforms in
terms of composition and purity
 Process control
 Food safety incl. packaging
 Forensics:
 Determine composition of soil, fibres, plastic, paint etc to
establish origin
 Trace analysis of Firearms Projectile Lead (FBI procedure)
 Elemental Analysis – Method
overview
 Classical methods:
 Qualitative Inorganic Analysis (Fresenius, Treadwell)
 Quantitative: Gravimetry,Titrimetry, Colorimetry…
 Instrumental trace analysis in solution
 Spectroscopic methods: AAS, ICP-AES/OES
 Mass spectrometry: ICP-MS
 Electrochemical methods ( CH914)
 Instrumental methods for solid materials
 X-ray methods (also spectroscopic)
 Mass spectrometry methods: SIMS and many other
 NB: Most instrumental methods are based on
physics, not chemistry of element
 Solid Analysis
state Select method in liquid
methods state
Acquire/define Acquire/define
sample sample

Process Process
sample sample

No
Chemical Soluble?
dissolution
Yes
The analytical
process Measurable
property?
General Yes
considerations Eliminate
No
interferences
and steps Change
chemical
Measure X
form

Calculate result

Determine error
 Method selection - considerations
 Destructive/non-destructive ?
 Non-destructive methods of analysis
 X-ray fluorescence, emission, etc.
 Destructive methods of analysis
 Combustion analyses
 Volumetric, gravimetric, electroanalytical analyses
 Atomic absorbance (AA) and inductively coupled plasma (ICP)
spectroscopy
 Mass spectrometry
 Expected analyte concentrations and performance
characteristics of method must match
 Sample must be compatible with required processing
and measurement
 Quantitative Analysis - Principles
1) Define sample amount (mass or volume)
2) Measure quantity proportional to analyte
concentration
 Measured property must vary in a defined way:
calibration with known standards necessary
 Analysis must be specific: Interferences must
be known and if possible be eliminated
 Accuracy: Proximity of measured value to accepted
(or "true") value: must be determined
 Precision: Closeness of measured values to one
another: must be defined and reported
 Performance characteristic of
quantitative analytical methods
 Accuracy
 Bias For definitions see:
 Recovery http://www.nmschembio.org.uk/
GenericArticle.aspx?m=98&amid=445
 Precision
 Reproducibility and Repeatability
 Detection capability
 Sensitivity
 Limit of Detection (LoD)
 Limit of Quantitation (LoQ)
 Selectivity and Specificity
 Linearity
 Working Range
 Robustness/Ruggedness
All these characteristics are intimately linked to the experimental error
 Experimental error

 Systematic error:
 Sources:
 Instrumental
 Method
 Personal
 Can be discovered and corrected
 Standard reference materials
 Blanks
 Controls, e.g. spiked samples
 Handle error by proper standardisation/calibration or
application of a correction factor

Systematic errors impact on Bias

 Experimental error
 Random error:
 Always present, can't be corrected
 Consequence of uncertainty of measurements
 electrical noise from instrument, causing fluctuations in
reading
 uncertainties in measurements of mass and volume
 Ultimate limitation in quantitation
 Must be aware of error and deal with it
 Repeated measurements

Random errors impact on Precision,

Reproducibility, Repeatability, LOD and LOQ

Both systematic and random errors affect accuracy

 Reporting quantitative data
 Errors can be defined via:
 Standard deviation (SD)
 Variance
 Relative std. deviation
N
 Coefficient of variation  i
( x  x ) 2

SD  i 1
 All quantitative data must N 1
be reported with error
– SD and RSD most common V  SD 2
 Propagation of errors
must be considered SD
RSD 
x
CV  RSD 100%
 Sampling errors: dealing with
heterogeneity
 “Real” samples are usually heterogeneous
 Examples: Foodstuffs, soils, water samples…
Lot
Sampling

Representative
bulk sample
Sample
preparation
 Random sampling: Homogeneous
 Sample fractions selected randomly lab sample
 Composite sampling:
 Samples taken at regular intervals and mixed

Aliquots
 Sampling error
 Overall error is composed of the errors introduced by the
analytical procedure (including sample preparation and
actual measurement(s)) and the sampling error:

SDo2 = SDa2 + SDs2

 SDo = overall standard deviation, SDa = sd of analytical procedure,
SDs = SD of sampling procedure

 If SDa << SDs or SDs << SDa, there is little point in trying to
reduce the smaller one
 Eg. If sa = 5% and ss = 10%, then so = 11%. Using a more expensive
and time consuming method whose sa = 1% will only reduce so to
10%
 Summary
 Elemental Analysis is important in a range of
sectors
 The analytical process consists of many steps
 Meaningful analysis must consider all steps
together
 Meaningful experimental design requires
understanding data
 Awareness of performance characteristics of
methods
 Awareness of statistics