Transcript ENV204 - RESEARCH METHODS IN ENVIRONMENTAL SCIENCE

EOE 2204
RESEARCH METHODS IN
ENVIRONMENTAL SCIENCE
Prof Paul Worsfold
Portland Square Development B520
Email [email protected]
(But not too often!)
Contents stored on sharepoint
EOE 2204
1
Introduction to Environmental Analytical Chemistry
This presentation is intended to provide support material for the
Wembury field and lab work. Any of the material in steps 1-5
could be included in an exam question.
The six steps in the “Analytical Approach” to any problem are:
1. Defining the problem
2. Sampling (Waters and Sediments)
3. Sample Treatment (Digestion of sediments, preservation of
waters)
4. Measurement
5. Data Treatment (Precision, Accuracy, Calibration, Limit of
Detection)
6. Report Writing
EOE 2204
2
Analytical Chemistry Definition
The science which deals with the detection,
identification and quantification of chemical species
in matrices of chemical, biological or environmental
origin. It is a multidisciplinary subject that has
chemistry at its core but also requires a knowledge
of other sciences, e.g. biology, biochemistry, physics,
mathematics, statistics and computing.
It is
concerned with the real world, with the primary aim
of determining WHAT constituents (qualitative
analysis) are present in a sample and HOW MUCH
of each constituent (quantitative analysis) is present.
EOE 2204
3
Environmental Analytical Chemistry
Common matrices/Application areas include;
• Hydrosphere
•e.g. sea, river, estuary, lake, porewater
• Atmosphere
•e.g. outdoor air, workplace air, air-sea interface
• Lithosphere
•e.g. soil, sediment. rocks
• Biosphere
•e.g. plants, animals, humans
EOE 2204
4
1. DEFINING THE PROBLEM
1.
2.
3.
4.
5.
Why am I doing this analysis?
What sites am I interested in?
What analytes am I interested in?
What concentrations do I expect?
What techniques should I use for
sampling and measurement?
6. What do my “customers” want
from this study/report?
EOE 2204
5
2. SAMPLING
Outline of the strategies which need to be
considered when designing and conducting a
sampling programme for environmental analyses
1.
2.
3.
4.
5.
6.
Sampling design
Economic and safety considerations
Sampling locations
Sampling time and frequency
Methods of sample collection
Sample storage and preservation
EOE 2204
6
Definition of objectives
Select determinands
and sampling positions
Design of sampling
programme for
environmental
analytical chemistry
Select number of samples
and time/frequency of sampling
Select appropriate analytical method
Apply this to e.g.
Wembury field trip
for stream water and
sediment analysis
Select methods for collecting samples
Select methods for sample preservation
Feedback from sampling,
analysis and interpretation
EOE 2204
7
Wembury Field and Laboratory Work
• Field work at Wembury (collection of water and
sediment samples and field measurements) at 3
locations (both morning and afternoon)
• Treatment of water and sediment samples
• Atomic absorption determination of Ca, Fe, Mg and
Zn in water samples and sediment extracts
• Determination of nitrate and phosphate in water and
sediment (P only) samples using a hand held
spectrophotometer
• Environmental assessment report
EOE 2204
8
Sampling strategy
Samples must be representative of;
•the bulk material (body) being sampled
•temporal variations
•spatial variations
Discrete (or Random/Grab) – does not account for variation in spatial
environment or time
Composite – pooled number of discrete samples
Systematic (or Grid/Transect) - a planned sampling strategy to accomplish
a specific objective e.g. a transect of an estuary to account for salinity
changes, identify different inputs, tidal cycle or time
(Pseudo)Continuous – ongoing monitoring to e.g. establish baseline
values for the environment and/or evaluate changes e.g. Atlantic Meridional
Transect (AMT) programme.
EOE 2204
9
Safety Considerations
• Appropriate health and safety precautions followed.
• Risk Assessment.
• Awareness of hazards: sample toxicity, site access,
site conditions, traffic, weather.
• Accessibility of sampling locations.
• Transportation.
Economic considerations
• Main constraints are time and resources.
• Set realistic/affordable objectives.
• Cost effective sampling design.
• Consider automated sampling (and analysis)?
EOE 2204
10
Sampling location
Consider the following issues;
• Programme objectives
• Heterogeneity of determinand distribution
• Accessibility (e.g. at bridge, tides)
• Exact position (e.g. centre of river, mid-depth)
• Point sources (e.g. STWs)
• Stratified water bodies (lakes, estuaries, seas)
• Mixing zones (merging streams, estuaries)
EOE 2204
11
Sampling Time and Frequency
Representative of temporal variations:
Random or cyclic (e.g. diurnal, seasonal, annual,
decadal).
Chemical, biological and physical processes.
Sampling options:
a)Occasional grab samples
b)At fixed times
c) At each part of a cycle
d)Continuously
Frequency of sampling constrained by economics
EOE 2204
12
Methods of Sample Collection
Sampling devices:
Water: Submersible pumps, Glass or plastic
(polyethylene) bottles, Depth sampling bottles.
Sediment and soils:, Scoops and trowels, Corers and
dredgers
Need to avoid
Contamination from sampling device and storage
container.
Glass – Use for organics. Ion exchange sites can
remove metal ions from solution.
Plasticware e.g. HDPE – Use for inorganics. Possible
leaching of organics from container.
EOE 2204
13
Sample bottles
Glass – is used for organic analysis. Easily cleaned but cant be
frozen.
Plasticware - e.g. high density polyethylene (HDPE) is used for
nutrients and trace metals. Can be frozen.
Cleaning protocols – should be defined in the sampling strategy
from the outset and can differ depending on the analyte under
investigation and containers used.
Cleaning protocols - trace level analysis
Pre-washing by soaking in hot detergent for 24 h
copious rinsing with pure water and then ultra high purity water
(UHP)
5 days in an acid bath (HCl)
copious (3x) rinsing with UHP
5 days in an acid bath (HNO3)
copious (3x) rinsing with UHP
EOE 2204
14
3. SAMPLE TREATMENT (WATERS)
Sample Preservation and Storage
The aim is to minimise potential changes to the
sample during storage. Processes that can affect
sample integrity include:
•Biodegradation (e.g. N and P compounds)
•Oxidation (e.g. Fe(II), organic compounds)
•Absorption of CO2 (e.g. pH, alkalinity)
•Precipitation (e.g. CaCO3, Al(OH)3)
•Volatilisation (e.g. O2, HCN)
•Adsorption (e.g. dissolved metals)
EOE 2204
15
Changes can be retarded (not prevented) by
Refrigeration (reduces biological activity)
Freezing (plastic bottles only)
Filtration (removes biotic and abiotic particulate
matter but not all bacteria and viruses)
Preserving agents: Concentrated; added during
sampling, Acidification (e.g. HCl) for trace metals,
Biocides (e.g. HgCl2) for biodegradable
determinands
EOE 2204
16
Filtration
• Separation of dissolved and particulate
determinands (0.45 m membrane filter) is
operationally defined
• Can use other sizes, e.g. 0.2 m
• Perform immediately
• Filters should be pre-washed
• Removes most biotic and abiotic particles but
not all bacteria, viruses and colloids
EOE 2204
17
3. SAMPLE TREATMENT (SEDIMENTS)
A sample size of ~ 0.1 – 5 g is usually enough for analysis.
This is obtained by dividing up the originally collected sample.
The less than 180 um size fraction of inorganic material is often
used and may require grinding. This helps provide a more
homogeneous sample and assists the dissolution (digestion)
process.
“Coning and quartering” - the reduction of the bulk sample to a
suitable sample size for analysis whereby the sample is dumped
to form a cone, which is then flattened. The circular layer is
divided into four equal segments and two opposite quarters are
discarded. The operation is repeated with the remainder until
the required amount is left.
EOE 2204
18
Methods of Sample Digestion
Wet digestion (ashing) in an open vessel (flask) is
the most common approach for elemental analysis.
Various acids can be used, e.g. a mixture of nitric
(oxidising) and hydrochloric (reducing) acids which is
known as aqua regia or nitric acid alone. To speed up
the process a closed pressurised vessel (e.g. PTFE
bomb) and/or a microwave can be used.
Dry ashing involves heating the sample in an open
platinum or glazed porcelain crucible at 450 - 550 oC
in a muffle furnace until the residue is white or nearly
so. It is then extracted with hot hydrochloric acid.
EOE 2204
19
Properties of commonly used acids for digestion of soils
and sediments
NITRIC acid (70%) is acidic and oxidising and forms soluble
salts (except oxides of Al, Nb, Ta, Ti, Sn, Sb, W). Used alone
or with bromine or hydrochloric acid.
HYDROCHLORIC acid (36%) is acidic and non-oxidising
(reducing for some higher oxidation states) and chlorides are
generally soluble (except Ag, Hg, Tl, Pb). Some chlorides
volatile (Hg, Ge, As, Sb, Se). Widely used for inorganic
matrices but not for organic matrices. Used alone or plus
either (1) a reducing agent such as tin(II) chloride or hyrazine
hydrochloride, or (2) an oxidising agent such as bromine,
nitric acid, perchloric acid.
EOE 2204
20
Sources of Contamination
•Reagents, e.g. acids, water
•Containers, material, cleaning protocols
•Environment, e.g. air quality, clean room, laminar flow
hood. Specific contamination sources include;
Facilities. Walls, floors and ceilings. Paint and coatings. Construction
material (sheet rock, saw dust). Air conditioning debris. Room air and
vapours. Spills and leaks.
People. Skin flakes and oil, Cosmetics and perfume. Perspiration. Clothing
debris (lint, fibres). Hair. A motionless person generates 100,000 particles of
<0.3 um per minute!
Fluids. Particulates floating in air (dust). Bacteria, organics and moisture.
Floor finishes or coatings. Cleaning chemicals. Plasticizers (outgasses).
Deionized water.
EOE 2204
21
Seastar (Canada) Baseline nitric acid specification
EOE 2204
22
Milli-Q water system (from Millipore Corp.)
•
•
•
•
•
•
Specifications
Resistivity (MΩ·cm at 25 °C) 18.2
TOC (ppb) 5–10
Pyrogens (EU/mL) NA
Bacteria (cfu/mL) <1
Flow Rate (L/min) 1.5
EOE 2204
23
4. MEASUREMENT
The major categories of techniques for the
chemical analysis of environmental matrices
Chemical Analysis
Instrumental methods
“Wet” methods
Classical
gravimetric
analysis
Classical
volumetric analysis
(titrations)
Optical
methods
uv/visible spectrometry
fluorescence
infrared
atomic absorption
(FAAS, ETAAS)
atomic emission (ICPAES, ICP-MS)
nmr spectroscopy
EOE 2204
Separation
methods
Other
methods
thin layer
chromatography
gas chromatography
liquid chromatography
electrophoresis
Electroanalytical
methods
voltammetry
potentiometry
amperometry
24
5. DATA TREATMENT
Precision – associated with random (indeterminate) error where
replicate measurements give different results. Individual results
distributed (Gaussian) around a mean value. Can be quantified
by the standard deviation.
Standard deviation (s.d.) - assessment of precision. Formula?
Relative s.d. (RSD) is s.d. expressed as a %
RSD (%) = (s/x) x 100
Accuracy – associated with systematic (determinate) error
(bias). How close is mean experimental value to the ‘true’ value?
Can be evaluated by intercomparison (round robin) exercises
and use of certified reference materials (CRMs).
EOE 2204
25
PRECISION AND ACCURACY
EOE 2204
26
Precision
Standard deviation (measure of random error)
Population standard deviation. The population standard
deviation (usually represented by the Greek letter sigma)
measures the variability of data in a population.
Sample standard deviation. The sample standard deviation
(usually represented by S) measures the variability of data in
a sample. It is easy to compute (compared to a population
standard deviation) because it is based on a small and
manageable number of measurements.
EOE 2204
27
The standard deviation assumes a Normal (Gaussian)
distribution of data from replicate determinations (usually 4 or
5). It is symmetrical about the mean and the greater the
value of s the greater the spread of the ‘bell shape’.
+/- 1s - includes 68 % of sample population
+/- 2s - includes 95 % of sample population
+/- 3s - includes 99 % of sample population
Mean (x) = Arithmetic estimate of true value (μ)
Spread or range = difference between largest and smallest
measured value
Confidence intervals can be used to test for systematic
errors – measure CRM, derive confidence interval and if
known value is not within confidence interval then systematic
error is probably present. EOE 2204
28
Accuracy
Seawater Reference Material for Trace Metals
NASS-5. The following table shows the twelve metals for which
certified values have been established. Certified values are
based on the results of determinations by at least two
independent methods of analysis. The uncertainties represent 95
percent confidence limits for an individual sub-sample. That is,
95 percent of samples from any bottle would be expected to
have concentrations within the specified range 95 percent of the
time.
Trace Metal Concentrations (micrograms/litre).
Arsenic 1.27 ± 0.12, Cadmium 0.023 ± 0.003, Chromium 0.110 ±
0.015, Cobalt 0.011 ± 0.003, Copper 0.297 ± 0.046, Iron 0.207 ±
0.035, Lead 0.008 ± 0.005, Manganese 0.919 ± 0.057,
Molybdenum 9.6 ± 1.0, Nickel 0.253 ± 0.028, Selenium(IV)
(0.018)*, Uranium (2.6)*, Vanadium (1.2)*, Zinc 0.102 ± 0.039.
*information value only
EOE 2204
29
CRM data sheets produced at Institute for National Measurement Standards, NRCC
CASS -4
Nearshore Seawater Reference Material for Trace Metals
NASS -5
Open Ocean Seawater Reference Material for Trace Metals
SLEW -3
Estuarine Water Reference Material for Trace Metals
SLRS-4
Riverine water Reference Material for Trace Metals
HISS -1
Marine Sediment Reference Material for Trace Elements and Other Constituents
MESS -3
Marine Sediment Reference Material for Trace Elements and Other Constituents
PACS -2
Marine Sediment Reference Material for Trace Elements and Other Constituents
CARP -2
Fish Reference Material for Dioxins, Furans and PCBs
DOLT -3
Dogfish Liver Reference Materials for Trace Metals
DORM -2
Dogfish Muscle Reference Materials for Trace Metals
LUTS-1
Non Defatted Lobster Hepatopancreas Reference Material for Trace Metals
TORT -2
Lobster Hepatopancreas Marine Reference Material for Trace Metals
ORMS -2
Elevated Mercury in River Water Reference Material
MOOS -1
Seawater Certified Reference Material for Nutrients
http://inms-ienm.nrc-cnrc.gc.ca/en/calserv/crm_e.php
EOE 2204
30
Calibration
Prepare series (at least 4) of standard solutions of known
concentration, analyse exactly as for sample.
Produce calibration graph – if linear, then derive the
concentration of the analyte by interpolation or use of
equation of straight line (y = mx + c).
y
r4
Response
signal
.
.
More sensitive
r3
r2
r1
.
.
.
.
.
e.g. y = 1.17x + 5.0833
r2 = 0.9988
. Unknown analyte
. standards
Less sensitive
c1
x
c2
Concentration
EOE 2204
31
Equation of a straight line y = mx + c (obtained from e.g. Excel)
e.g. y = 1.17x + 5.0833, r2 = 0.9988
y – signal, m – gradient, x - analyte conc, c – intercept of line
r2 value, known as ‘correlation coefficient’, (actually product-moment
correlation coefficient) represents line of best fit through the data
points. Indicates ‘quality’ of the calibration graph. Equation used to
interpolate unknowns, i.e. samples. The gradient of the line defines
the SENSITIVITY of the method.
Important considerations
•Always include blank value in calculation (never subtract blank
from other points).
•Always make rough manual plot of data in lab to check for errors.
•Is response linear? For analytical data need at least 0.99.
•What is the best straight line (slope and intercept) through the
data points?
EOE 2204
32
Limit of Detection (LOD)
This is a very important term that is used to determine the lowest
analyte concentration that can be confidently detected. It is
defined as the concentration giving a signal (e.g. absorbance
units) equal to the blank signal, YB, plus three standard
deviations of the blank, SB.
It is calculated by analysing the blank solution several (4 or 5)
times and calculating both the mean and standard deviation of
the blank data.
The value YB + (3 x SB) is then substituted into the equation of
the line (y = mx + c) and the equivalent x value (concentration)
determined. Assuming that the line goes through YB it follows
that x(LOD) = (3 x SB)/m.
EOE 2204
33
6. REPORT WRITING
See practical handbook for details of structure, length,
presentation and deadline
Presentation of data
The number of Significant figures given indicates the
precision of the experiment – in practice quote as sig. figs. the
digits which are certain e.g. 10.09, 10.10, 10.09, 10.11 - mean
is 10.102, s = 0.01304 – clearly uncertainty in 2nd decimal
place. Therefore best quoted as x +/- s = 10.10 =/- 0.01
Rounding up or down of figures – if consistent in one direction
e.g. up, introduces bias - therefore round to the nearest even
no e.g. 9.65 to 9.6, 4.75 to 4.8
EOE 2204
34
EOE 2204
35
7. Supplementary information and exercises/answers
t-test used to compare experimental mean with a known value
(thus checking for systematic errors) by
1. establishing a null hypothesis (H0) (e.g. there is no difference
between observed and known values other than that due to
random errors),
2. then testing whether it is true, i.e. the statistical probability
that difference between x and m arises from random chance.
The lower the probability that difference occurs by chance the
less likely the null hypothesis is true.
Usually null hypothesis rejected if probability is < 1 in 20 (i.e. p
= 0.05 or 5 %) i.e. the difference is significant at the 0.05 or 5 %
level. Increased certainty by using p = 0.01 or 0.001 (1 % or 0.1
%)
3. The value of the t is calculated, if this is less than a critical
value (taken from tables) then the null hypothesis is accepted.
EOE 2204
36
t-tests (cont’d) - also used to
Compare 2 experimental means e.g. those from a new
method with a reference method (Ho is that the 2 means are
equal and then test whether difference between the means
differs significantly from zero). Calculate t (nb diff calc than for
previous t) and use tables as before (n.b. assumes samples
are drawn from populations with equal s.d.). Can also use this
test to decide whether a change in experimental. conditions
effects the results. (n.b. if s.d. of the populations is different,
different equations are used).
Paired t test – comparison of differences in results obtained
by 2 analytical methods, when there are also differences in the
analyte concentrations in the samples analysed. Achieved at p
= 0.05 by testing the difference (d) between each pair of
results - Ho = d does not differ significantly from 0
EOE 2204
37
Outliers – when one or more results differs
substantially from the others – test data using
• Dixons Q test - Ho (null hypothesis) is that all
measurements come from the same population.
Usually use p = 0.05 (95 % confidence).
Q = Suspect value – nearest value/
largest value – smallest value
If calc Q value exceeds critical Q value (from
tables) then suspect value rejected.
• Grubbs test – compares the deviation of the
suspect value from the sample mean with the
standard deviation of the sample. It is now the
preferred outlier test.
EOE 2204
38
Other useful statistical tests
F-test - Significance test comparing standard deviations
and thus used to compare random errors between two data
sets.
1 sided F test - tests whether method A is more precise
than method B e.g. whether new analytical method is
more precise than old method.
2 sided F test – tests whether methods A and B differ in
their precision.
ANOVA – Analysis of variance. Comparison of more
than 2 means e.g. several different analytical methods
or different analysts using the same equipment. 2
different sources of variation – that always present due
to random errors and the variation due to controlled or
fixed-effect factors e.g. the solution storage conditions,
EOEor
2204
39
the analytical method used
the analyst.
UNITS
ppm = 1 part per million (106) = 1 mg L-1 = 1 ug mL-1
(for solutions) = 1 mg kg-1 = 1 ug g-1 (for
solids)
ppb = 1 part per billion (109) = 1 ug L-1 = 1 ng mL-1
(for solutions) = 1 ug kg-1 = 1 ng g-1 (for
solids)
ppt = 1 part per trillion (1012) = 1 ng L-1 etc.
EOE 2204
40
MORE UNITS
% = 1 part per hundred (percent)
‰ = 1 part per thousand (often used for seawater)
Mass (weight) percent (or m/m %) = (mass of
substance / mass of total solution) x 100
Volume percent (or v/v %) = (volume of
substance / volume of total solution) x 100
Mass/volume percent (or m/v %) = (mass of
substance in grams / volume of total solution in mL)
x 100
EOE 2204
41
Prefix
Symbol Factor
(10)
Prefix
Symbol Factor
(10)
tera
T
12
centi
c
-2
giga
G
9
milli
m
-3
mega
M
6
micro

-6
kilo
k
3
nano
n
-9
hecto
h
2
pico
p
-12
deca
da
1
femto
f
-15
deci
d
-1
atto
a
-18
EOE 2204
42
Molarity
The most common unit of concentration is molarity (moles per
litre), abbreviated as M (or mol L-1). A mole is defined as the
number of atoms of 12C in exactly 12 g of 12C. This number of
atoms is called Avogadros number and its value is 6.022045
x 1023. A mole is simply 6.022045 x 1023 of anything.
Note that M is the symbol for moles per litre whereas mol is
the symbol for moles.
The molecular weight (MW), now called relative molecular
mass (RMM), of a substance is the number of grams that
contain Avogadros number of molecules.
To convert from M to g L-1 multiply by the RMM. Then to
convert from g L-1 to mg L-1 multiply by 1,000.
To convert from g L-1 to M divide by the RMM.
EOE 2204
43
Exercise 1
Data for nitrate in water (mg L-1). The true value, as
determined by a CRM is 42.0 mg L-1.
A
42.5
41.6
42.1
41.9 41.1
42.2
B
39.8
43.6
42.1
40.1
43.9
41.9
C
43.5
42.8
43.8
43.1
42.7
43.3
D
35.0
43.0
37.1
40.5
36.8
42.2
E
42.2
41.6
42.0
41.8
42.6
39.0
Comment on the bias, precision and accuracy of each of
these sets of data.
EOE 2204
44
Exercise 1 Answers
1. Mean results for labs A-E are 41.9, 41.9, 43.2, 39.1, 41.5.
The true answer was 42.0.
A – precise, little bias, accurate.
B – poor precision, little bias, mean accurate but not very
reliable.
C – precise but biased to high values, not very accurate.
D – poor precision, biased to low values.
E – similar to A but the last result might be an outlier.
EOE 2204
45
Exercise 2
The following data sets were obtained for the determination of
copper, calcium and magnesium in river water (units mg L-1).
Calculate the mean, standard deviation and relative standard
deviation of each data set.
Cu
0.07
0.08
0.07
0.08
0.07
0.07
0.08
0.08
0.09
Ca
23.3
22.6
22.5
24.7
21.9
21.5
19.9
21.3
21.7
23.8
Mg
13.8
14.0
13.2
11.9
12.0
12.1
EOE 2204
46
Exercise 2 Answers
Means, standard deviations and relative standard deviations
as follows:
Cu
Mean = 0.077 mg L-1
s.d. = 0.007 mg L-1 RSD = 9 %
Ca
Mean = 22.3 mg L-1
s.d. = 1.4 mg L-1
RSD = 6.2 %
s.d. = 0.95 mg L-1
RSD = 7.4 %
Mg
Mean = 12.83 mg L-1
EOE 2204
47
Exercise 3
1. What dilutions would you use to prepare standards of 1, 2, 3,
4, 5 mg L-1 Ca from a stock solution of 1,000 mg L-1 Ca?
2. What dilutions would you use to prepare standards of 0.1, 0.5,
1.0 mg L-1 Mg from a stock solution of 1.000 mg L-1 Mg?
Exercise 3 Answers
1. Dilute stock 10X with water to give 100 mg L-1 Ca then pipette
1, 2, 3, 4 and 5 mL directly into 100 mL volumetric flasks and
make each up to 100 mL with water.
2. Dilute stock 10X with water to give 100 mg L-1 Mg then dilute
a further 10X with water to give 10 mg L-1 Mg then pipette 1, 5
and 10 mL directly into 100 mL volumetric flasks and make each
up to 100 mL with water.
Note: Always use ultrapure water e.g. from Milli-Q unit.
EOE 2204
48
Guided Reading
1. 1. ACOL Environmental Analysis by R.N. Reeve, 543.REE.
2. 2. Statistics for Analytical Chemistry, 4th Ed, by J.C.
Miller and J.N. Miller, 519.5024541.MIL.
3. 3. Any general Instrumental Analysis textbook, e.g.
Principles of Instrumental Analysis, 4th Ed, by D.A. Skoog
and J.L. Leary, 543.08.SKO
4. 4. An Introduction to Atomic Absorption Spectroscopy by L.
Ebdon, 543.0888.EBD.
5. 5. Encyclopedia of Analytical Science, Various articles,
543.003.ENC.
6. The Chemical Analysis of Water, 2nd Ed, by D.T.E. Hunt
and A.L. Wilson, 546.22.HUN.
7. http://ec.europa.eu/environment/guide/part2d.htm
EOE 2204
49