Method Development for Direct
Analysis of Mercury by Thermal
Decomposition
David L. Pfeil
dpfeil@teledyne.com
1
Webinar Topics
• How does thermal decomposition work?
• When does it make sense to use this technique?
• Applicable methods
• Instrumentation
• Concerns about samples
• Method development strategies
• Some examples
2
Hydra C - Principle of Operation
• Solid or liquid samples are weighed and introduced in
the Hydra C
• The sample is initially dried and then thermally
decomposed in an oxygen flow
• Combustion products are carried off and further
decomposed in a hot catalyst bed
• Mercury vapors are trapped on a gold amalgamator
and subsequently desorbed for quantitation
• The mercury content is determined using atomic
absorption
Hydra IIC – System Schematic
Absorption
Cells
High Sensitivity
Low Sensitivity
50-900°C
600°C
Decomposition Catalyst
Furnace
Furnace
O2 Supply
Drying
Tube
Amalgam
Furnace
Delay
Tube
When is Thermal Decomposition Most
Useful?
• The analysis of solids such as:
–
–
–
–
–
–
Soils, sediments & sludge
Coal & fly ash
Fish
Food & feeds
Plants & tissues
Ores
• Difficult to digest samples
• Samples with known chemical interferences
5
Benefits of Thermal Decomposition
• Fast turnaround time
• More universal calibration
• Interference reduction
• Better sensitivity for solids
• No hazardous chemicals used
6
Time Required for CVAA Sample Digestion
Time Required for Direct Analysis
Reference Materials as Standards
Sediment
Coal
Sediments
Dogfish
Oyster tissue
High Sensitivity Range
Correlation Coef. 0.9993
Dogfish
Low Sensitivity Range
Correlation Coef. 0.9996
Standard Reference Materials
SAMPLE
No.
Certificate
Measured
Recovery
(ppm)
(ppm)
(%)
1577
0.016
0.0178
111.7
Lypho 1
0.096
0.091
94.8
Lypho 2
0.039
0.036
90.9
Lypho 3
0.073
0.067
91.4
Dogfish
Dorm-2
4.64
4.34
93.5
Dogfish
Dolt-2
1.99
1.79
90
Oyster Tissue
1566
0.057
0.061
107
8406
0.06
0.061
101.7
2709
1.4
1.52
108.6
HC-35150
0.176
0.177
100.6
Tissue
Bovine Liver
Blood
Marine
Sediments &
Soils
Coal
Interference Reduction
Because the CVAAS technique requires a chemical reduction and the
thermal decomposition technique does not, some interferences can be
eliminated.
Interference with KI
120
100
Recovery (%)
Recovery (%)
120
Interference with Au
100
80
60
CVAAS
40
Thermal Decomp
20
0
0
1
2
KI (%)
3
80
60
CVAAS
40
Thermal Decomp
20
0
0
100
200
300
Au Conc. (ppm)
11
Detection Limits Comparison
Typical published detection limits
Classical CVAAS
CVAFS
CVAFS with amalgam
~1 ng/L
~0.1ng/L
<0.05 ng/L
Thermal Decomposition
0.005-0.001ng
Detection Limits in concentration for thermal decomposition
Divide the mass detection limit by the sample weight
For Example:
Assume the reported detection limit is 0.002ng and the sample weight is
0.2grams.
Detection limit = 0.002ng/0.2g = 0.01ng/g (ppb)
12
Apples to Apples
Typical published detection limits
Classical CVAAS
CVAFS
CVAFS with amalgam
~1 ng/L or ~0.001ng/g (ppb)
~0.1ng/L or ~0.0005ng/g
<0.05 ng/L
- For solid samples dilution with required digestion is not taken
into account.
- Often 50-100X dilution is required and the method detection
limit will be higher
For Example:
Assume 2.0g of sample is digested to a final volume of 100mls (or ~100g).
CVAAS Detection limit = (0.001ng/g)(100mls) /(2g) = 0.05ng/g (ppb)
Compare with 0.01ng/g by thermal decomposition for 0.2g sample
13
Chemicals Needed for CVAAS
• Acids
– Nitric
– Hydrochloric
– Sulfuric
• Oxidizers
– Potassium permanganate
– Potassium persulfate
Personnel safety
Gloves
Eye protection
Ventilation
Expense
Purchase price
Disposal costs
Contamination
Extra sample handling
Reagent purity
• Reductants
– Hydroxylamine
– Stannous chloride
– Sodium borohydride
14
Applicable Methods
• USEPA 7473
– Mercury In Solids and Solutions by Thermal
Decomposition, Amalgamation, and Atomic Absorption
Spectrophotometry
• ASTM D6722
– Standard Test Method for Total Mercury in Coal and
Coal Combustion Residues by Direct Combustion
Analysis
15
Instrumentation
Hydra IIC Modular Design
• Furnace
• Spectrometer
• Autosampler
16
Hydra IIC
- AA Spectrometer
Folding Mirrors
High Sensitivity Cell
Low Sensitivity Cell
Lamp Module
Reference Cell
Low Concentrations
• Measurement of real samples
– Contamination can be significant
• Boats
• Analyst
• Environment
– Sample homogeneity
18
Dryer / Amalgamator Module
Gold Amalgamation Trap
Nafion Dryer
Interface to Furnace
Interface to Control Board
High Concentrations
• Long (2”x5”) & Short Cell (1”) come standard
• High Concentration option is unique
1,500ng max or
25,000ng max
Catalyst
First section captures
catalyst poisons such as
sulfur and halogens
Sample boat enters
empty part of catalyst
tube
Third section traps
partially oxidized
residuals
Second section
provides oxygen to
assist combustion
21
Sample Concerns
• Homogeneity
– Is the aliquot measured representative of the sample?
– Can I get acceptable precision?
– How much sample is needed?
• Flash point, if combustible
– We are placing fuel & oxygen in a sealed vessel
• Volatility
– Will evaporation on the sampler affect accuracy?
• Concentration
– High or low concentrations may not fall in the analytical range
• Moisture
– Samples must be dried before the decomposition step
• Physical condition of sample
– Is the sample contained in the boat
22
Do we need to perform sample preparation?
• You might want to grind your
samples to get a more uniform
consistency
– For something like soils or
plant material a simple coffee
grinder might suffice
– Cost ~$20
23
Combustible materials
• What is combustible
– In an oxygen atmosphere materials are more
combustible than you might at first imagine
• Start small
– Control the amount of fuel added to the furnace
– Explosive conditions can exist when fuel and
oxidant are present in the right ratio
– If you know the sample is combustible, try to
keep the mass under 20ugs
– Use multiple deposits to increase the signal, if
necessary
24
Multiple Injections
• To deposit more volatile sample mass multiple injections
are typically offered in the software.
– Less mass of a sample is placed in each boat
– With each sample boat injection only the dry and combustion
phases are carried out
– Only after the last injection the mercury is eluted and measured
– Concentration is calculated based on the total mass injected in
all the boats
25
Example of Multiple Deposits
Total deposit = 1.077gm
26
What is volatile?
• If the mercury species is volatile (or reactive),
then results will be low if left on the sampler.
– Some acidic solutions can react with nickel boats,
reducing mercury.
• If a solvent is employed with the sample but the
mercury species is stable, loading the sampler
should be acceptable.
– If unsure, run a sample directly after loading and
repeat at various time intervals checking precision
27
Sample concentrations out of range
• Low concentrations
– Inject more sample
– Use multiple deposits to increase mass of mercury loaded
• High concentrations
– Inject less sample
• Analytical error during weighing
• Homogeneity concerns
– Dilution is difficult in solids
• Suitable blank material available?
– Digest the sample
• Liquid samples are easy to dilute
28
Moisture Control for Aqueous Samples
• As a rule of thumb allow 0.7 seconds of dry time
for every microliter of solution.
– With a sample capacity of 1.4ml the dry time
can be quite tedious
– 1400 x 0.7 = 980 seconds ( or >16 minutes)
• On the other hand solid samples may require little or no
dry time
29
Physical limitations
• Difficult to contain samples types can be wrapped
in a suitable material such as aluminum foil
– Hair
– Feathers
– Glass wool
• Sticky stuff needs to be transported carefully
– Previously frozen fish tissues
30
Method Development Strategies
• View from 30,000 feet
– Characterize your samples
– Determine sufficient dry time
– Run a small sample aliquot
– Estimate sample mass needed for mid-range
concentration
– Evaluate response characteristics
– Determine precision
– Adjust decomposition values, if necessary
– Determine accuracy
31
The Pyrolysis Phase
• After the sample is dry, furnace temperature increases
for sample pyrolysis or combustion.
• To satisfactorily complete this step sufficient oxygen
must be present.
–
–
–
–
Organic compounds require more oxygen
More mass deposited requires more oxygen
Multiple deposits may help reach complete oxidation
Determination of Hg in rice is a good example
32
Coal & Fly Ash
• Coal and fly ash are typically two of the easier matrices to analyze by thermal
decomposition; however, some coals high in sulfur may shorten catalyst
lifetimes.
Coal
Phase
Temperature (°C)
Time (s)
Drying
300
60
Decomposition
800
400
Catalyst
600
Wait Time
Amalgamator
60
600
30
• Notes:
– If moisture is high extend Drying Time
– Use nickel boats
– Calibrate on aqueous or matrix matched standards
33
Plant tissue
• Plant material may be best handled by pre-desiccation and grinding to a
consistent particle size; however, satisfactory results often may be
obtained without any pretreatment.
Plants
Phase
Temperature (°C)
Time (s)
Drying
300
0.7*mgs
Decomposition
800
160
Catalyst
600
Wait Time
Amalgamator
60
600
30
• Notes:
– Set dry time to 0.7 sec/mg sample weight if not desiccated.
– Use nickel boats
– Calibrate on aqueous or matrix matched standards
34
Soil
• Soil can contain plant roots and large rocks. It may make
sense to sieve and/or grind the soils to exclude these
materials.
Soil
Phase
Temperature (°C)
Time (s)
Drying
300
10
Decomposition
850
180
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
35
Soil Sample
Untreated
After Grinding
36
Fish
• Fish is recognized as the primary source of mercury
adsorption in humans.
• The FDA has a recommended concentration limit for Hg at
1ppm.
• The limit in much of Europe is set at 0.5ppm
Fish
Phase
Temperature (°C)
Time (s)
Drying
300
45
Decomposition
800
150
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
37
Rice
• Many grains, such as rice, are dietary staples and while
routinely low in Hg may deserve monitoring.
Rice
Phase
Temperature (°C)
Time (s)
Drying
300
60
Decomposition
800
400
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
• Note-limited mass required (≤50mg)
38
Oils
• We have looked at a variety of oils
– Used engine oils,
– Vegetable oil
– Marine oil
• The decomposition phase is hotter & longer than usual to
prevent incomplete decomposition.
Oils
Phase
Temperature (°C)
Time (s)
Drying
100
60
Decomposition
800
200
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
39
Hair
• Hair can be an excellent indicator of past exposure to mercury.
– Clean sample before clipping
– Wrap in tin foil to control positioning
Hair
Phase
Temperature (°C)
Time (s)
Drying
300
60
Decomposition
800
150
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
40
Blood & Urine
• Urine can be used to determine recent exposure to inorganic
mercury.
• Blood can be used to determine recent exposure to organic
mercury
Fluid
Phase
Temperature (°C)
Time (s)
Drying
300
60
Decomposition
800
150
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
41
Waters
• Waters can be determined by thermal decomposition but:
– Slow because of moisture removal
– Less sensitive than classical CVAAS
Water
Phase
Temperature (°C)
Time (s)
Drying
300
TBD
Decomposition
800
150
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
42
A Single Platform
AA
C
All run with
Envoy 2.0
Software
AFGold
AF
43
Sodium Hydroxide Solution
• Sodium hydroxide solutions unstable in the sample boat.
• Acidified solutions are stable.
– In our experiments we have used 6N HCl
NaOH
Phase
Temperature (°C)
Time (s)
Drying
300
TBD
Decomposition
800
150
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
44
Gypsum
• Synthetic gypsum used in the production of wallboard
– a byproduct of many coal-fired power plants using flue gas
desulfurization (FGD).
– As coal plant emissions become cleaner, concern is increasing in
byproducts, like gypsum.
Gypsum
Phase
Temperature (°C)
Time (s)
Drying
300
70
Decomposition
800
150
Catalyst Temperature
600
Wait Time
Amalgamator
60
600
30
45
Gypsum from one Power Plant
• A mean concentration of 170ng/g Hg
• 7.5 million tons of gypsum# is used in the production of
wallboard
• About 2,500lbs of mercury re-introduced into the environment.
Hg/yr = (0.170x10-6lbs/lbs)(7.5x106 ton/yr)(2000lbs/ton) =2,550lbs Hg/yr
# Based on ACAA 2006 Coal Combustion Product (CCP)
Production and Use Survey – 7,579,187 short tons used in
wallboard production
46
In Summary
• Very short turn-around time for results.
• Greener technique with no digestion chemicals or
waste products.
• Generic calibration for multiple matrices.
• Wide dynamic range but dilution is impractical.
• Multiple deposits for lower concentrations.
• Sensitivity better than classical CVAAS for solids,
not quite as good for liquids.
• Combo systems
47
Questions?????
48