Transcript 上課講義內容下載處

地質分析化學
Geo-analytical Chemistry
沈川洲 Chuan-Chou (River) SHEN; [email protected]
Course Description:
This course is intended for graduate students in the
Dept. of Geosciences, without an extensive background in
chemistry. It emphasizes both theoretical and practical aspects of
analytical techniques for geochemical research. Modern
instrumental techniques and experimental methods will be
included. Paper reading, discussion and presentation will also
play essential roles in class. After taking this class, students will
be able to design an overall experimental procedure, including
estimating sample size, sampling, choosing chemicals and
labwares and selecting measurement methods for their projects.
They will also learn how to independently accomplish chemistry in
a geochemical lab.
Literature:
Quantitative Chemical Analysis by Daniel C. Harris
Fundamentals of Analytical Chemistry by Skoog and West
Statistics for Analytical Chemistry
by Miller and Miller
Data Reduction and Error Analysis for the Physical Sciences
by Bevington and Robinson
Text materials: Selected chapters, class handouts, papers.
Credits: 2
Class hours: 7:00-9:00 pm, Thursday
Grading: Homework, 10%; Midterm exam, 30%;
Report and presentation, 20%; Final, 40%.
Syllabus:
Sept
Oct
Nov
Introduction, Uncertainty, Probability, distributions.
Error analysis.
Significance tests.
Regression correlation.
Methods for quantitative analysis.
Geochem lab: labwares, clean room, QA/QC.
Geochem lab: acids, standards, spikes, safety rule.
Samples and sampling.
Nov 27
Dec
Midterm exam.
Dec 18
Jan
2003 AGU Fall Meeting.
Jan 15
Final; oral exam.
Dissolution and separation techniques.
Ion exchange chromatography.
Analytical methods for geo-environmental samples.
Paper discussion. Geo-analytical techs.
Analytical methods for geo-environmental samples.
1. Introduction
1.1. Types of geochemical surveys
a. Rock surveys
b. Sediment and soil surveys
c. Stream, lake and ocean water surveys
d. Vegetation (biogeochemical) surveys
e. Gas surveys
a. Increased availability of toxic Al3+ to a pine tree in Germany near a coal-burning
power plant built in 1929. The increase is probably an effect of man-made acidity
in rainfall, which mobilizes Al3+ from minerals.
b. The growth of atmospheric CO2. CO2 comes from our burning of fossil fuel and
destruction of forests.
c. The growth in world population.
How long will our planet remain habitable if we do not control our population
and our impact on the environment?
2. Uncertainties in measurements
2.1. Analytical problems: qualitative
1. Does this distilled water sample contain any boron?
2. Could these two igneous rock samples have come from the
same site?
3. How much lead is there in this sample of tap-water?
2.2. Answers: quantitative
I can/cannot detect boron in this water sample.
A quantitative method capable of detecting boron at levels
of 1 mg/ml
This sample contains < 1 mg/ml boron
(detection limit).
This sample contains > 1 mg/ml boron (e.g. 1.4 mg/ml)
(1.4 ± 1.2 mg/ml; 1.4 ± 0.2 mg/ml; 1.40 ± 0.02 mg/ml).
2.3. Types of error
2.3.1. Gross error (gross mistake)
reversing a sign, using a wrong scale, arithmetic
mistakes…
2.3.2. Determinate error (systematic error)
a. Instrument errors
b. Method errors
c. Personal errors
Constant vs. Proportional
The accuracy of an experiment is dependent
on how well we can control or compensate
for systematic errors. Systematic errors affect
accuracy, i.e. proximity to the true value.
Errors make results different from the true
values with reproducible discrepancies.
How to reduce systematic errors?
2.3.3. Indeterminate error (random error)
Indeterminate error arises from uncertainties in a
measurement that are unknown and not controlled by the
scientist (operator). The precision of an experiment is
dependent on how well we can overcome random errors.
Random errors affect precision, or reproducibility, of
an experiment.
To reduce random errors:
a. Repeating the experiment
b. Improving the experimental method
c. Refining the techniques
Repeatability: within-run precision
Reproducibility: between-run precision
Repeatability: within-run precision
Reproducibility: between-run precision