Copyright © 2018 by the Water Environment Federation.

All Rights
Reserved. Permission to copy must be obtained from WEF.

Water Environment Research, WEF, and WEFTEC are registered

trademarks of the Water Environment Federation.

ISBN 978-1-57278-356-0

IMPORTANT NOTICE
The material presented in this publication has been prepared in accordance with generally
recognized engineering principles and practices and is for general information only. This
information should not be used without first securing competent advice with respect to its
suitability for any general or specific application.
The contents of this publication are not intended to be a standard of the Water Environment
Federation® (WEF) and are not intended for use as a reference in purchase specifications,
contracts, regulations, statutes, or any other legal document.
No reference made in this publication to any specific method, product, process, or service
constitutes or implies an endorsement, recommendation, or warranty thereof by WEF.
WEF makes no representation or warranty of any kind, whether expressed or implied, concerning
the accuracy, product, or process discussed in this publication and assumes no liability.
Anyone using this information assumes all liability arising from such use, including but not
limited to infringement of any patent or patents.
The Publisher works hard to ensure that the information in this publication is accurate and
complete. However, it is possible that the information may change after publication, and/or that
errors or omissions may occur. We welcome your suggestions on how to improve this publication
and correct errors. The Publisher disclaims all liability for, damages of any kind arising out of
use, reference to, or reliance on information included in this publication to the full extent
provided by state and federal law.

About WEF
The Water Environment Federation (WEF) is a not-for-profit technical and
educational organization of 35,000 individual members and 75 affiliated
Member Associations representing water quality professionals around the
world. Since 1928, WEF and its members have protected public health and
the environment. As a global water sector leader, our mission is to connect
water professionals; enrich the expertise of water professionals; increase the
awareness of the impact and value of water; and provide a platform for
water sector innovation. To learn more, visit www.wef.org.
 Foreword 2019 Edition
Over 20 years have passed since I wrote the original Foreword for this
publication. In looking back over the years, most things have not changed.
Indeed, if you are a new want-to-be analyst or an experienced analyst in
another testing position, there are still things you need to know about
environmental testing and the responsibilities (sometimes personal) that go
along with it.
The trend toward replacing wet chemical methods with multi-analyte
instrumental techniques continues unabated. Many of the wet methods in
the 18th edition of Standard Methods have been removed from the 23rd
edition in favor of instrumental methods. Although the U.S. Environmental
Protection Agency (U.S. EPA) still approves most of the older wet methods
for use, there is an increasing trend toward multi-analyte methods such as
ion chromatography, capillary ion electrophoresis, ICP/AES, ICP/MS, flow
injection, and others. Although computers may seem to run the instrument
and the laboratory, it is still humans that provide the chemical knowledge
and expertise.
The legal scrutiny never stops. In addition to state environmental
agency and U.S. EPA oversight, various organizations such as the American
Canoe Association have discovered that there is money to be made in
bringing civil lawsuits against water resource recovery facilities for non-
compliance with permits and regulations. In particular, they have found that
the laboratory is a prime target for non-compliance. Non-compliance may
take the form of an inability to meet legislated detection/reporting limits or
the use of non-approved methods.
You may think you are a simple laboratory analyst but, in actuality, you
are a prime target for a lawsuit. Moreover, in the event of a lawsuit, you
may be named personally as a defendant. Thus, you need to protect
yourself. Know the regulations and follow them to the letter. These include
40 CFR Part 136 for wastewater and 40 CFR Part 141 for water analysis.
Know the requirements of your facility’s permits. Make sure you document
everything, no matter how minor it seems.
 This is a wonderful industry to work in. Aside from the vital public
health function that we perform, it is also personally rewarding with
excellent growth potential. In June of 2000, I resigned from a commercial
analytical laboratory (Analytical Services, Inc.). Rather than stop working,
however, my situation has allowed more time to explore many new
challenges in the industry. In addition to continuing to write books and give
seminars to laboratory groups, I have been involved as an expert witness in
court, giving depositions and testimony about laboratory procedures,
analysis, quality control, and results. It seems that every day brings a new
opportunity.

ACKNOWLEDGMENTS
A work of this type, in a science as complicated as laboratory chemistry, is
impossible to produce in a vacuum. I owe an unpayable debt of gratitude to
the many persons who assisted in the donation of information, pictures, and
their time for technical creation of the manuscript. First and foremost, I
acknowledge the contributions and assistance of the company that paid for
my daily bread and allowed me to pursue such desires as the writing of the
first edition of this book, Analytical Services, Inc. Special thanks go to
Robert G. Owens, Jr., Denise S. Geier, Billy P. Dyer, G. Wyn Jones, Jeff
Newman, Lang Allen Reeves, and the rest of the technical staff at
Analytical Services for their help, data, and pictures. Also providing
enormous assistance in this work and (dare I forget it) restraint in putting up
with my oftentimes abrupt and abrasive personality were Paul McMinn,
Fisher Scientific, Norcross, Georgia; Susan Grable, Fisher Scientific,
Pittsburgh, Pennsylvania; James B. Carl, Painted Post, New York; David
Black and Jane Leisenring, Corning Inc., Corning New York; Lisa M.
Lazzara, Nalge Company, Rochester, New York; Paul Stinson, Ever Ready
Thermometer Co. Inc., West Patterson, New Jersey; Judy Poxon, Vee Gee
Scientific, Kirkland, Washington; Andy Sendelback and Chris Clark, Varian
Sample Preparation Products, Palo Alto, California; and Jennie Durant, Bill
Smutney, Maryanne Reves, and David Fenili, Kimble Kontes, Vineland,
New Jersey. Jon P. Henderson, Georgia Water and Wastewater Institute,
Georgia Water and Pollution Control Association, Carrollton, Georgia,
deserves mention for allowing me time with live classes of operators and
laboratory analysts to work out the presentation problems of many of the
concepts discussed in this book. The reviewers of the book—David
Kimbrough, California Environmental Protection Agency, Los Angeles,
California; Dr. Larry Keith, Radian Corp., Austin, Texas; Robert G. Owens,
Jr., ASI, Norcross, Georgia; Dr. Mark Bruce, Quanterra, North Canton,
Ohio; and Dr. David Carrick, Australian National Laboratories, Asquith
New South Wales, Australia, are thanked for their time and many useful
comments for improving the presentation and content. To steal a dedication
used by a classmate of mine from the graduating class of 1968, East
Greenwich High School, East Greenwich, Rhode Island, I dedicate this
book to the memory of my parents, Mary Eda Keller Smith and
Commander Roy Fowler Smith, USN, without whom I would not be here.
Finally, I owe a debt of gratitude to my editor, Lorna Ernst, and her staff
at the Water Environment Federation for the original idea for a second
edition of this work and the help they provided for its completion and
delivery.

ABOUT THE AUTHOR

Roy-Keith Smith began his laboratory career with the U.S. Army in 1969
and is a Vietnam veteran. After his discharge, he obtained his BS in
Chemistry from the Georgia Institute of Technology and graduated from
Colorado State University with a Ph.D. in Chemistry. He then completed a
one-year research faculty appointment at the California Institute of
Technology. After several years working with the Georgia Department of
Agriculture, he took a position as a laboratory manager in an environmental
laboratory. Taking time out for a stint as a college professor of
environmental chemistry, he returned to the environmental industry with,
first, a six-month sabbatical replacement position in the Hewlett-Packard
Analytical Education Center and then moved on to a position as Analytical
Methods Manager with Analytical Services, Inc. He left ASI in June 2000.
Dr. Smith served the industry as an educational consultant to the
Georgia Water and Wastewater Institute and was a member of the Georgia
Water Pollution Control Association Laboratory Committee, the Water
Environment Federation Laboratory Practices Committee, and as part
coordinator for Part 4000 and Part 5000 of Standard Methods. Through his
consulting firm Apichemical Consultants, he has published numerous books
on environmental analytical chemistry and served as an expert witness in
environmental analytical chemistry. He is the recipient of the 1994 WEF
Laboratory Analyst Excellence Award from the GWPCA.

Dr. Roy-Keith Smith

Apichemical Consultants
PO Box 1243
St. Augustine FL 32085
 Contents

Foreword
List of Figures
List of Tables

Chapter 1 Analytical Standards and Reagents

1.0 TYPES OF CHEMICALS
2.0 GRADES OF CHEMICALS
3.0 STORAGE OF CHEMICALS
4.0 MANUFACTURER-PREPARED STOCK SOLUTIONS AND
CERTIFICATION
5.0 REAGENT WATER

Chapter 2 Analytical and Toploading Balances

1.0 TYPES OF BALANCES
2.0 SITING AND ENVIRONMENTAL CONCERNS
3.0 CALIBRATION OF BALANCES AND CERTIFIED WEIGHTS
4.0 ACCESSORIES
5.0 OPERATION AND DAILY MAINTENANCE

Chapter 3 Laboratory Ware

1.0 TYPES AND CHARACTERISTICS OF GLASS
2.0 LABORATORY GLASSWARE DESCRIPTION
3.0 CERAMICWARE DESCRIPTION AND CHARACTERISTICS
4.0 PLASTICWARE DESCRIPTION AND CHARACTERISTICS
5.0 METALWARE DESCRIPTION AND CHARACTERISTICS

Chapter 4 Volumetric Devices and Their Use in Measurement

1.0 VOLUMETRIC MEASURING TOOLS
2.0 PROPER USE OF VOLUMETRIC GLASSWARE
3.0 CALIBRATION OF NONSTANDARD MEASURING TOOLS
4.0 REFERENCES

Chapter 5 Temperature Measurement

1.0 INTRODUCTION
2.0 DESCRIPTION OF TEMPERATURE MEASURING DEVICES
3.0 CALIBRATION CHECKING AND MAINTENANCE OF
LIQUID-IN-GLASS THERMOMETERS

Chapter 6 Preparation of Solutions and Dilutions

1.0 USE OF BALANCES AND VOLUMETRIC WARE
2.0 PREPARATION OF DILUTIONS
3.0 STORAGE OF PREPARED SOLUTIONS

Chapter 7 Calibrations
1.0 STANDARDIZATION OF SOLUTIONS
2.0 DIRECT STANDARDIZATION
3.0 INDIRECT STANDARDIZATION
4.0 SOLUTION LIFETIME
5.0 CALIBRATION
5.1 One-Point Calibrations
5.2 Multipoint Graphical Techniques
5.3 Multiple Standard Additions
6.0 FREQUENCY OF CALIBRATION AND CALIBRATION
CHECKS
7.0 REFERENCE

Chapter 8 Test Procedures

1.0 DISTILLATION
2.0 FILTRATION
3.0 GRAVIMETRIC DETERMINATIONS
4.0 COLORIMETRIC DETERMINATIONS
5.0 TURBIDIMETRIC DETERMINATIONS
6.0 TITRATION
7.0 pH AND ION-SELECTIVE ELECTRODES
8.0 REFERENCE

Chapter 9 Calculation and Reporting Results

1.0 SIGNIFICANT FIGURES
2.0 CALCULATIONS
3.0 DRY WEIGHT CORRECTIONS
4.0 REPORTING RESULTS

Chapter 10 Controlling the Test Procedure

1.0 THE ABILITY OF THE TEST TO ACTUALLY MEASURE THE
DESIRED PARAMETER IN THE SAMPLE
2.0 HOW CLOSE THE RESULT IS TO THE ACTUAL AMOUNT
OF ANALYTE IN THE SAMPLE
3.0 CAN THE SAME RESULT BE OBTAINED REPEATEDLY?
4.0 WHAT IS THE LOWEST LEVEL OF ANALYTE THAT CAN BE
DETECTED IN THE SAMPLE?
5.0 IS THE DETECTED PARAMETER ACTUALLY IN THE
SAMPLE?
6.0 BATCH ANALYSIS
7.0 REFERENCES
8.0 RECOMMENDED READING

APPENDICES

Appendix A Molecular and Formula Mass

Appendix B Molarity and Normality

Appendix C Types of Chemical Reactions

Appendix D Stoichiometry of Chemical Reactions

Appendix E Laboratory Analyst Training

1.0 INTRODUCTION
2.0 TRAINING GOALS
3.0 TRAINING PROGRAM
4.0 TRAINING DOCUMENTATION
5.0 CONCLUSION
6.0 REFERENCES

Index
 List of Figures
1.1 Acid storage cabinet
1.2 A deionizing unit for laboratory water purification
2.1 Triple beam balance
2.2 Top loading balance
2.3 Analytical balance
2.4 Aluminum weighing boats
2.5 Disposable plastic weighing boats
2.6 Reagent (weighing) funnel
2.7 Weighing scoop
3.1 (a) Tapered and (b) spherical joints
3.2 Diagrams of the size designation for a 24/40 male standard taper
joint (left) and a 35/25 male spherical joint (right). Units are
millimeters
3.3 O-ring joint
3.4 Stopcock plug
3.5 Buret with rotaflow valve
3.6 (a) Berzelius and (b) Griffin beakers
3.7 Erlenmeyer flasks with screw caps
3.8 Ribbed watch glass used with a beaker
3.9 Narrow mouth reagent bottle
3.10 Vacuum flask with side arm
3.11 Test tube
3.12 Kjeldahl flask and condenser
3.13 (a) West, (b) Graham, (c) Allihn, and (d) Liebig jacketed condensers
3.14 Jacketed condenser with ground glass fittings
3.15 Powder funnel
3.16 Plain long-stem filtering funnel
3.17 Perforated funnel with vacuum attachment
3.18 Porcelain Buchner funnel
3.19 Fritted Buchner funnel
3.20 Funnel and support assembly
3.21 Separatory funnel
3.22 Heavier-than-water continuous liquid-liquid extractor
3.23 Lighter-than-water continuous liquid-liquid extractor
3.24 Accelerated one-step continuous liquid-liquid extractor
3.25 (a) mortar and pestle, (b) spotplate, (c) crucibles, and (d) dishes
3.26 ASTM resin identification codes
3.27 Wash bottle without right-to-know label
3.28 Wash bottle with right-to-know labels
3.29 Variable-volume reagent dispenser
3.30 Petri dishes
3.31 Serological pipets
3.32 Disposable pipetter tips
3.33 Teflon beaker with metal bottom
3.34 Tubing connectors
3.35 Quick-disconnect tubing connector
3.36 Stainless steel adjustable hose clamp
3.37 Presterilized sample containers with dechlorination tablets for
coliform testing
3.38 Long-handle sampling spoons
3.39 Disposable filtration units
3.40 Reusable filtration units
3.41 Hand-operated vacuum pump
3.42 Single ferrule compression fitting
3.43 Variety of support clamps: (Top) Fixed position, medium 3 prong
dual adjust clamp. (Bottom) Fixed position, medium 2 prong single
adjust clamp
4.1 Class A volumetric flasks
4.2 (a) Class A volumetric pipet, (b) Class A measuring pipet (Mohr),
(c) a serological pipet
4.3 Class A graduated cylinder
4.4 Class AS micro buret with attached reservoir
4.5 Correct position of calibration mark and meniscus (left). Meniscus
positioned higher than calibration mark (right)
4.6 Finger (left) and thumb (right) pipetting techniques
4.7 Position of meniscus in a graduated cylinder estimated as 52.8 mL
4.8 Pipette bulb
5.1 Liquid-in-glass thermometer
5.2 Enclosed-chamber thermometers
5.3 Digi-Sense® Armored Liquid-In-Glass Thermometer
5.4 Digi-Sense® 7 Point Reversible Temp Label
6.1 Flask with ground glass stopper
6.2 Ultrasonic cleaning baths
6.3 Reagent dispenser bottle
7.1 Example of a three-cycle semilog graph
7.2 Example of a rectangular graph
7.3 Computer generated calibration plot of phosphate data
7.4 Line graph of phosphate data
7.5 Locating an absorbance value intersection with the calibration curve
7.6 Locating a concentration value from an absorbance result
7.7 Added phosphate calibration points demonstrating saturation
7.8 Regression line of all phosphate data
7.9 Regression line for phosphate data with top two points eliminated
7.10 Fluoride calibration data plotted on rectangular paper
7.11 Log of fluoride calibration data plotted on rectangular paper
7.12 Fluoride calibration data plotted on semilog paper
7.13 Multiple standard addition calibration for reactive phosphorous
8.1 Cyanide distillation system
8.2 Ammonia nitrogen distillation apparatus
8.3 Total Kjeldahl nitrogen (TKN) automated still
8.4 Kuderna-Danish concentrator
8.5 TurboVap II® Automated Solvent Evaporation System
8.6 Vacuum rotary evaporator
8.7 Distillation trap
8.8 Heating mantle and temperature controller
8.9 Water bath
8.10 (a) Cone, (b) Filter support with gasket, (c) Stopper
8.11 Vacuum trap
8.12 Dewar flasks
8.13 Solid-phase extraction apparatus for oil and grease
8.14 Solid-phase extraction disk
8.15 Glass desiccator
8.16 Plot of absorbance versus wavelength for a colorimeter test
9.1 Uncertainty of reading the position of a meniscus in a burette
9.2 Meniscus at 4.41-mL volume in a burette
9.3 Meniscus at 5.00-mL volume in a burette
10.1 Theoretical distribution of results controlled only by random error
around a mean
10.2 Frequency distribution for total suspended solids recovery
10.3 Accuracy and precision illustrated by target shooting
10.4 Benchsheet used for batch analysis
 List of Tables
1.1 Abbreviations frequently found in material data safety sheets
1.2 Comparison of price and quality of the common solvent acetone
from one supplier
1.3 Similar reagent-grade terms for some of the major suppliers
1.4 Primary standards and uses
1.5 Quality parameters for ASTM class waters
2.1 Tolerances (mg) of various ASTM and NIST classes of standard
weights
3.1 Rubber stopper sizes
3.2 Pores sizes of fritted glass filters
3.3 Chemical composition (percent) of elements other than iron in
stainless steel alloys
4.1 Comparison of Class A and Class B tolerances for volumetric flasks
4.2 Comparison of Class A and Class B tolerances for pipettes
4.3 Comparison of Class A and Class B tolerances for graduated
cylinders
4.4 Comparison of Class A and Class B tolerances for burettes
4.5 Comparison of relative percent accuracy of the calibration marks on
25-mL measuring devices for Class A and Class B tolerances
4.6 Comparison of percent accuracy for determination of dispensing
1.00 mL from a variety of Class A measuring devices
4.7 Flow times for TD transfer pipettes
4.8 Density variation of reagent water with temperature
5.l Comparison of Kelvin, Celsius, and Fahrenheit temperature scales
5.2 Thermocouple types, constructions, and temperature ranges
5.3 Infrared emissivity of various materials
7.1 Absorbance data for phosphate calibration
7.2 Fluoride calibration data
7.3 Multiple standard addition for a phosphorus test
8.1 Particle retention ratings for a variety of qualitative filter papers
8.2 Residue fractions determined in the laboratory
8.3 Characteristics of drying agents for potential use in a desiccator
8.4 Absorbance data obtained from varying the wavelength around a
method maximum
8.5 Some pH indicators for use in checking wavelength adjustment on
spectrophotometers
8.6 Methods that result in wavelength standards for checking
colorimeter
8.7 Exact wavelength setting variation depending on the direction of
approach
9.1 Implied uncertainties for different ways of writing a number on a
benchsheet
9.2 SI prefixes for metric units
10.1 Examples of relative percent difference calculated from matrix spike
and matrix spike duplicate recoveries
B.1 Serial dilutions performed on glucose solution
C.1 Common names of some laboratory chemicals
E.1 Introduction to quality assurance lesson plan
E.2 Analyst class schedule
 1
Analytical Standards and Reagents

1.0 TYPES OF CHEMICALS

2.0 GRADES OF CHEMICALS
3.0 STORAGE OF CHEMICALS
4.0 MANUFACTURER-PREPARED STOCK SOLUTIONS AND
CERTIFICATION
5.0 REAGENT WATER

Chemical testing can be divided into two types. The first type of testing
measures a bulk physical property of the sample, such as volume,
temperature, melting point, moisture content, or mass. These measurements
are characteristically nondestructive to the sample. Moreover, these
measurements are typically performed with an instrument, and one simply
has to calibrate the instrument to perform the test. Most analyses, however,
are of the second type of testing, in which a chemical property of the
sample is determined that generates information about how much of what is
in the sample. Measurements of this type are destructive to the sample.
The determination of chemical properties typically involves observation
of the reaction between the sample and selected chemicals, called reagents.
Without reagents, few chemical analyses could be performed.
An important class of reagents comprises analytical standards. These
are the chemicals used to calibrate a procedure so that the data generated
will correctly reflect the composition of the sample. Reagents allow the test
to be done, whereas analytical standards are necessary to obtain reliable
results from the test.
1.0 TYPES OF CHEMICALS
The three physical forms of matter encountered in the laboratory are solids,
liquids, and gases. Different chemicals exist in one of the three phases at
room temperature. When chemicals from commercial suppliers are
purchased, solids and liquids come in plastic or glass bottles, whereas gases
typically arrive compressed in steel cylinders. The chemical and physical
properties of the chemicals serve to classify them into appropriate groups
for handling and storage.
Many chemicals are not particularly corrosive or caustic, generating
solution pHs of approximately 7 (neutral). Some chemicals are highly
acidic, resulting in low pH solutions (1 to 4), whereas others are highly
basic and give solutions of high pH (10 to 14).
Two terms that are widely misused in chemistry are strength (also,
strong or weak) and concentration (also, concentrated or dilute). They are
completely different concepts.
Strength is a chemical property. It relates to the degree that individual
molecules of a substance remain intact in a water solution or break apart
into ions. A strong acid, such as hydrochloric acid, has no intact molecules
of HCl in the solution; instead, the molecules are all ionized (HCl → H+ +
Cl−). Examples of strong acids include hydrochloric acid, nitric acid,
sulfuric acid, hydrobromic acid, perchloric acid, and chromic acid. In a
similar vein, there are strong bases such as sodium hydroxide. There are no
intact molecules of NaOH in solution; instead, they are all ionized (NaOH
→ Na+ + OH−). Examples of strong bases include sodium hydroxide,
potassium hydroxide, barium hydroxide, and calcium oxide (lime). There
are also weak acids and bases. Weak means that the substance is largely
intact as molecules in the solution, with very little ionization. Acetic acid
(in vinegar) is a weak acid, and more than 95% of the molecules are intact
in solution with very little ionization (HOAc ↔ H+ + OAc−). Examples of
weak acids include acetic acid, phosphoric acid, hydrofluoric acid, lactic
acid, and ascorbic acid. Weak bases are also largely intact as molecules in
solution. An example of this is ammonia. A solution of ammonia represents
mostly intact ammonia molecules in solution with very little ionization
(NH3 + H2O ↔ NH4+ + OH−). Other weak bases include magnesium oxide,
nicotine, and many others.
Concentration is a physical property. It relates to how much of the
material is in the solution. If there is a lot of the substance in the solution, it
is termed concentrated, whereas, if there is only a small amount in solution,
it is called dilute. The strong acids are typically provided as concentrated
solutions such as hydrochloric acid (37%), sulfuric acid (96%), and nitric
acid (70%). The weak acid acetic acid is typically provided pure (100%,
glacial). Although these acids are diluted before use (i.e., 1.0-M
hydrochloric acid or 1.0-M sulfuric acid), they are still strong acids. A 1.0-
M solution of acetic acid is a dilute weak acid. The strong-base sodium
hydroxide is provided as a solid that must be diluted in water to provide a
working solution of, say, 1.0-M NaOH; however, it is still a strong base. A
concentrated or dilute solution of ammonia is always going to be weak
regardless of how powerfully it affects your eyes or nose.
A number of salts, such as ammonium chloride, ferric chloride,
aluminum sulfate (alum), and sodium acetate, form acidic or basic solutions
when dissolved in water. Although these solutions are not as strong as those
of the strong acids or bases, they can still be corrosive.
Acids and bases need to be stored separately in secure areas. Regardless
of whether they are strong or weak, acids and bases will generate a lot of
heat and can react violently when mixed.
Chemicals found in the laboratory are typically reactive; otherwise, they
would be of little use. Some reagents can donate (i.e., cause reduction) or
accept (i.e., cause oxidation) electrons from other materials. Some examples
of oxidizers include potassium persulfate, potassium permanganate,
chlorine gas, sodium dichromate, chromium trioxide, and hydrogen
peroxide. Nitric acid is an oxidizer and also a strong acid. Some common
reducers are manganese (II) sulfate, aluminum metal, zinc metal, sodium
borohydride, and sodium thiosulfate. Reducers and oxidizers should never
be stored together. Reducers and acids are not compatible and need to be
stored separately because the reaction between them often releases
hydrogen gas that can lead to an explosion or fire hazard.
Fires are always a concern in the laboratory. They can be caused by
mixing oxidizers and reducers or by electrical shorts in the equipment.
Some chemicals such as hexane, ether, alcohol, hydrogen gas, acetylene
gas, magnesium metal, and others are flammable. Once a fire starts with
flammable chemicals, it can be very difficult and hazardous to extinguish.
Many laboratory chemicals are toxic. Although a few such as water,
sodium chloride, and sugar are not poisonous, for the most part the
chemicals found and tested for in a water or wastewater laboratory are toxic
and should always be treated with respect. Chemicals should never be
tasted, and odor should be sampled with extreme caution. Inadvertent
contact with chemicals should be avoided by wearing gloves in the
laboratory. Eating, smoking, and drinking are forbidden in the laboratory
under Occupational Safety and Health Administration regulations because
of the possibility of ingesting toxic materials. Because samples can contain
hazardous chemicals (otherwise, they would not require testing), they, too,
should be regarded as dangerous. The best policy is to treat all samples and
laboratory chemicals as if “one touch could kill”.
Most pure chemicals are too concentrated for direct use as reagents and
must be diluted. The most common process of dilution involves dissolving
a known amount of the reagent in an inert liquid, called the solvent. The
most common solvent in a water or wastewater laboratory is water itself.
Water is used to dilute most inorganic reagents. Some reagents such as
concentrated hydrochloric acid (37%) and concentrated nitric acid (70%)
are provided in water solution. Sometimes, organic solvents are used to
dissolve reagents that are not soluble in water. One example is the acid-base
indicator phenolphthalein, which is dissolved in alcohol. Examples of
common organic solvents are alcohol, acetonitrile, hexane, methylene
chloride, methyl-t-butyl ether, and toluene. Many of the organic solvents are
flammable, and all are toxic.
The acidic, basic, oxidizing, reducing, toxic, and flammable properties
of the reagents will dictate how they are stored and handled. Each
laboratory chemical sold in the United States is required by law to come
with a safety data sheet (SDS, referred to in the past as a material safety
data sheet, or MSDS). The SDS summarizes all the physical and chemical
properties of the substance as well as the hazards. Directions for what to do
in case of fire or exposure are included as well as first aid measures. The
SDS must be in the laboratory and readily available to all employees. The
SDS can also be stored and accessed through a computer. Many chemical
suppliers have their entire collection of SDSs available online.
Abbreviations commonly used in an SDS are shown in Table 1.1
 Water is everywhere in the laboratory, including in the reagents. Some
reagents have well-defined with constant amounts of water in them and are
termed hydrates. Sodium thiosulfate pentahydrate (Na2S2O3 − 5H2O), for
example, has 5 molecules of water associated with each formula unit of the
thiosulfate. Reagents that are available as hydrates are also typically
available as anhydrous, meaning there is no water present. Both forms of
sodium thiosulfate are dry, well-defined crystalline materials; however, 5.00
g of the hydrate contains 1.82 g of water and 3.18 g of sodium thiosulfate.
Five grams of the anhydrous reagent contain 5.00 g of sodium thiosulfate.
In some cases (sodium thiosulfate is one example), the hydrate is much
more stable for storage purposes than the anhydrous reagent and, therefore,
is the preferred form.
Other reagents (e.g., sodium hydroxide, calcium carbonate, and zinc
nitrate) come from the supplier as dry-appearing, white, anhydrous solids.
However, when they sit on the shelf after being opened, they form a
clumpy, wet-looking solid and can eventually liquefy, which is a process
known as deliquescence. These materials are absorbing water in a
nonspecific fashion and present quite a problem to the analyst who wants to
know how much of 5.00 g is actually the desired reagent. Furthermore, in
addition to absorbing water, sodium hydroxide is reacting with atmospheric
carbon dioxide to form sodium carbonate.

TABLE 1.1 Abbreviations frequently found in material safety data sheets.

ACGIH American Conference of Governmental Industrial Hygienists

CAS Chemical Abstracts Service
CDC Centers for Disease Control and Prevention
CERCLA Comprehensive Environmental Response, Compensation, and
Liability Act
CL Ceiling Level
DOT Department of Transportation
EHS Extremely Hazardous Substances
EPA Environmental Protection Agency
IARC International Agency for Research on Cancer
LC50 Lethal Concentration for 50% of test population
LD50 Lethal Dose for 50% of the test population
LEL Lower Explosive Limit
MSDS Material Safety Data Sheet
N/A Not Available
NFPA National Fire Protection Association
NTP National Toxicology Program
OSHA Occupational Safety and Health Administration
PEL Permissible Exposure Limit
REL Recommended Exposure Limit
RQ Reportable Quantity
RTECS Registry of Toxic Effects of Chemical Substances
SARA Superfund Amendments and Reauthorization Act
STEL Short-Term Exposure Limit (< 15 minutes)
TLV Threshold Limit Value
TPQ Threshold Planning Quantity
TSCA Toxic Substances Control Act
TSD Treatment, Storage, or Disposal
TWA Time-Weighted Average
UEL Upper Explosive Limit

2.0 GRADES OF CHEMICALS

The success of the analysis and the utility of the data are completely
dependent on the quality of the reagents used in the procedure. A trace
analysis for zinc is doomed to fail if a technical-grade acid containing 0.1%
zinc as an impurity is used for the sample digestion. Chemicals are hard to
purify and the purity of the reagent is directly related to its cost.
Practical- and technical-grade chemicals are designed for industrial
process use, not for laboratory use. These grades may be as high as 95% to
97% pure, but a 95% pure material still contains 50,000 parts per million
(ppm) of impurities. Most analyses in the water and wastewater laboratory
are looking for contaminants in the 1- to 10-ppm range or even lower.
Trying to find 10 ppm of a target analyte in a mixture with up to 5,000
times as much miscellaneous garbage is almost impossible, particularly
when some of the garbage is comprised of the same material as the target
analyte of the test procedure.
There are several nationally recognized chemical-purity grading
standards. The most important standards for water and wastewater
laboratories are those published by the American Chemical Society (ACS)
in Reagent Chemicals, American Chemical Society Specifications, which is
in its 10th edition. Chemicals that meet the listed specifications are sold as
“ACS Reagent Grade”. This standard is the only one designated specifically
for chemical laboratory purposes. The other recognized standards are the
United States Pharmacopeia, National Formulary, and Food Chemicals
Codex. These are designed for the food and drug industries that have
different requirements than those of a chemical laboratory. All reagents
used in chemical analysis should be at least ACS Reagent Grade.
Unfortunately, not all laboratory chemicals are listed in Reagent
Chemicals. Furthermore, the listed specifications may not meet the
requirements for all possible laboratory uses. This is particularly true for
trace analysis of metals and organics. For example, mercury analysis
requires use of potassium permanganate. The ACS Reagent Grade is 99+%
pure, with the low mercury version having ≤0.05 ppm mercury. That
represents 50 parts per billion (ppb) mercury for use in an analytical
procedure that has a detection limit of 0.2 to 0.5 ppb and a regulatory limit
of 2.0 ppb mercury.
Because it is not practical, and in many cases impossible, to purify
reagents and solvents to meet all laboratory uses, purified materials are
available for specific purposes. For example, a solvent sold as “Pesticide
Residue” grade is suitable for use in the sample preparation and analysis of
pesticide residues by gas chromatography, but is not suitable for high-
performance liquid chromatography (HPLC) methods. There is a separate
grade of solvents for HPLC use. For the most part, these highly purified,
special-use reagents and solvents are of a higher grade for a given chemical
than the ACS Reagent Grade, and the cost is significantly higher. A cost
comparison for different grades of one supplier’s acetone, based on 2016
prices, is shown in Table 1.2.
Each batch of manufactured reagents has a lot number and certificate of
analysis, which often are incorporated in the label on the bottle. Reliable
suppliers of reagents always provide a copy of the certificate of analysis
upon request if the certificate does not come with the bottle of reagent.
Willingness to provide this information indicates that the company stands
behind the stated quality of their chemicals and is willing to provide
technical assistance when questions arise. Refusal by the supplier to provide
a certificate of analysis for a purchased reagent is a sure sign of a less-than-
reliable or incompetent company.
All reagents have a usable shelf life when stored under proper
conditions. The shelf life may range from weeks to months to years. Often,
an expiration date is listed on the label. If none is listed, a year from date of
receipt is commonly assumed, although prepared solutions may have a
much shorter shelf life. Reagents should not be used past their expiration
date.
There is a plethora of confusing terms used by the various chemical
supply houses for the different grades of available reagents. Table 1.3
attempts to draw some comparisons.

TABLE 1.2 Comparison of price and quality of the common solvent

acetone from one supplier (March 2016).

Grade Quantity Price ($)

Pesticide residue GC 4L 237.90
Pesticide residue 4L 206.87
Spectrometry 2.5 L 101.99
ACS 4L 206.33
HPLC 4L 139.95
Purified 1 gal 120.82
Histology 4L 147.72
TABLE 1.3 Similar reagent-grade terms for some of the major suppliers.

A special group of reagents, called primary stands, are important in

laboratory work. They have a common set of characteristics, as follows:

• Solid and easily purified to a reliable level;

• Stable to oven-drying temperatures (100 to 150°C);
 • Stable, with a long shelf life;
• Do not quickly absorb water on standing;
• React with a well-defined and constant stoichiometry; and
• Readily available from a number of commercial sources.

Primary standards are dried in an oven at 103 to 105 °C before use to

remove any absorbed water. Some examples of primary standards and their
use are presented in Table 1.4.
Upon receipt of a reagent or standard, the analyst should write the date
on the bottle and the certificate of analysis. The chemical is then placed in
its proper storage area, the certificate of analysis filed in a binder, and the
SDS checked to see whether it has been updated since the last one on file.
An SDS should be updated every 2 to 3 years and the most current
information kept on file.
When purchasing chemicals, the amount projected to be required for use
is the primary consideration. Even though smaller amounts of chemicals,
for instance 10 g, cost more per unit than, say, 1 kg, there are no savings if
only 10 g is used and the other 990 g must be thrown away after a year.
There are also added costs if the chemical must be processed as a hazardous
waste. Most laboratory chemicals are classified as hazardous. Laboratories
tend to build up substantial stocks of outdated reagents, which simply
contribute to the hazardous environment of the facility and increase the
chances of a leak, spill, or breakage. Furthermore, it is always a challenge
to find an unlabeled bottle of a chemical on a shelf and an illegible remnant
of a label adjacent to the bottle. The laboratory needs to keep a list of on-
hand chemicals arranged by purge and expiration date so that stocks can be
kept at a minimum.

TABLE 1.4 Primary standards and uses.

Primary standard Use

Benzoic acid Bomb calorimetry
Potassium biiodate Standard source of iodine
Potassium dichromate Standardization of redox reactions
Oxalic acid dihydrate Standardization of potassium permanganate
solutions
Potassium hydrogen Standardization of strong base solutions
phthalate
Sodium carbonate Standardization of strong acid solutions
Sodium chloride Total dissolved solids, chlorides

3.0 STORAGE OF CHEMICALS

As mentioned in the previous section, chemical reagents can be classified as
acids, bases, oxidizers, reducers, flammable solvents, gas cylinders, and
“miscellaneous”. Reagents should always be segregated according to class
and stored in separate areas. There are specialized chemical storage cabinets
for each of the hazard classes. An acid storage cabinet is illustrated in
Figure 1.1. Flammable solvent storage cabinets should also be vented to a
safe area outside the laboratory. Readily available chemically resistant
plastic trays can be placed under the bottles on each shelf for added
protection. The cabinet itself can be placed on a large tray to provide a
holding dam in case of catastrophic container breakage.
Some reagents and standards must be stored at lower than room
temperature. Explosion-proof and chemically resistant refrigerators should
be used rather than those purchased from the local appliance store. Samples,
reagents, and standards must never be stored in the same refrigerator
because reagents and standards can contaminate the samples and vice versa.
In certain areas of the country, such as the West Coast, earthquake bars
are provided on the front of each shelf to prevent bottles from “walking” to
the edge and falling off. In general, it is a good idea to have all chemicals
stored in cabinets with doors that have a positive latch, even in areas not
prone to earthquakes.
FIGURE 1.1 Acid storage cabinet. Reprinted with permission from Metal
Arc, Inc.

4.0 MANUFACTURER-PREPARED STOCK

SOLUTIONS AND CERTIFICATION
Many chemical suppliers provide reagents as certified solutions. In most
cases, these prepared solutions are more expensive than the solutions each
analyst can prepare from the stock chemical—even including the time and
reagents used in the standardization of the solution. Potassium
permanganate is one of those few prepared solutions that is less expensive
purchased from chemical suppliers because of the involved filtration and
standardization of the solution. It is an absolute necessity that the analyst
obtain and retain on file a copy of the manufacturer’s certificate of analysis
for any purchased solutions. It is also imperative that the analyst checks the
titer of the reagent solution when it is opened and thereafter at intervals
depending on the particular chemical.
Discrepancies in results must be resolved, not swept under the rug.
Verification of a manufacturer’s solution titer is an excellent test of the
analyst’s ability to perform a standardization. Even purchased check
samples and standards should be assayed against the laboratory’s existing
standards. Repeated inability to verify a titer is a cause for concern.
Although the most common cause is a systematic (procedural or
calculation) error by the analyst, sometimes it is an indication of a
manufacturer error. Everybody makes mistakes, including the big chemical
supply houses, and all reagents change titer with time. When the laboratory
quality assurance manager has an exhaustive set of data proving beyond a
shadow of a doubt that there is an error in a purchased solution, he/she
should contact the supplier about the problem. A reputable supplier will
resolve the problem. There is nothing more satisfying than having a
supplier admit an error about a purchased solution, but nothing more
embarrassing than to be shown where your error is.

5.0 REAGENT WATER

Water is the most common solvent found in the laboratory. In general,
drinking water, although safe for human consumption, is not suitable for
use in the laboratory without further purification. The American Society for
Testing Materials (ASTM) has developed a purity classification scheme for
the quality of laboratory water. The quality parameters for these classes are
outlined in Table 1.5.
TABLE 1.5 Quality parameters for ASTM class waters.

Water is a difficult substance to purify. The significant contaminants

present in water are dissolved ionized salts, gases, non-ionized solids
(organics), suspended solids, microorganisms, and organic degradation
byproducts. Removing all the contaminants from water, especially in
multigallon quantities, is expensive and time-consuming. It is more cost-
effective to examine the intended use of the water and prepare it to be free
of the target analytes and interferences for that specific purpose.
The main methods for purification of laboratory water are deionization,
distillation, filtration, reverse osmosis, adsorption, ultrafiltration,
sterilization, and ultraviolet irradiation; of these, the oldest method is
distillation. For the most part, some of these methods are the same
procedures used on a large scale at water purification or treatment plants.
Distillation, by itself, will remove most inorganic salts, but little else. The
older stills made of copper actually increase the copper, lead, and zinc ion
load of the distilled water. Distillation is typically used as one of the last
steps in a water purification train, and then it is typically performed in an
all-glass still.
 Deionizing cartridges filled with either anion-specific or cation-specific
exchange resins, or cartridges filled with a mixture of the ion-exchange
resins, are used to remove charged ions and particles from water. Self-
contained, stand-alone units are also available (Figure 1.2).
Cartridges filled with activated carbon are used to remove organic
contaminants from water. The order of the cartridges in the purification
train is important and depends on the final use of the water. The resin
cartridges can contribute organic material to the water, whereas the
activated carbon cartridges leach metal ions and anions. Water for metals
analysis should pass through the ion-exchange unit last, whereas water
intended for determination of organic parameters needs to be carbon-
filtered last. Most laboratories start with drinking water, pass it through
deionizing cartridges and/or an activated carbon adsorption cartridge, and
then distill the reagent water.
FIGURE 1.2 A deionizing unit for laboratory water purification.
Reprinted with permission from Hach Company.

The analyst may find that, for parameter testing at very low levels, even
the aforementioned procedures are inadequate for production of analyte-
free water. It is possible to prepare water for a specific test that will exhibit
no demand on the reagents used in the test by distillation of the water from
the determinative reagent. For example, if potassium dichromate is the
determinative reagent, the analyst can add some potassium dichromate to
the intended water, and then distill the water in an all-glass still to prepare
oxidant-demand-free water for use as dilution water or for preparation of
reagent solutions. Water suitable for preparation of potassium permanganate
solutions can be distilled from potassium permanganate. Reductant-
demand-free water can be prepared by distillation from sodium thiosulfate.
Chloride-free water can be prepared by distillation from silver nitrate in the
still pot. Sulfate-free water can be distilled from barium chloride.
In practice, the first 50 mL or so of water that distills is used as a rinse
of the system and discarded; then, the water can be collected for use. No
more than 90% of the original contents of the distillation flask should be
collected; the approximately 10% remaining in the still should be discarded.
The still pot should never be allowed to distill to dryness.
Once the water is distilled, it is typically stored in a large glass bottle,
then filtered through a 0.45- to 0.2-μm filter as it is dispensed. Reagent
water is extremely difficult to keep pure. It will absorb carbon dioxide,
ammonia, and other contaminants from the air, and is prone to develop
microorganism and fungus populations. The storage bottle needs to be
protected from contact with the atmosphere to prevent contamination. As a
rule, the laboratory should never prepare more purified water than it can use
in one day. Most modern systems are designed to provide purified water on
demand. These systems should always be flushed daily through the last
stage and dispenser outlet before any collection of reagent water takes
place.
There are many commercial suppliers of water purification units
designed for direct attachment to drinking water sources. These units
typically have replaceable cartridges that are returned to the supplier when
exhausted for recharge/reuse. Indeed, no cartridges are designed to last
forever. The laboratory needs to develop a protocol for checking the quality
of the reagent water produced on a daily basis from the purification system.
This may be as simple as logging the conductivity of the water daily for
metals and general chemistry use. Or, it may be as involved as daily
evaluation of a portion of the product water as if it is a regular sample. As
soon as the quality of the purified water falls below the standards needed to
produce an analyte-free blank, the cartridges need to be replaced. By
keeping a record of the volume of water successfully purified by a set of
cartridges, the laboratory can determine the expected lifetime for the system
and arrange to have the cartridges replaced before analysis has to stop
because of substandard water.
 2
Analytical and Toploading Balances

1.0 TYPES OF BALANCES

2.0 SITING AND ENVIRONMENTAL CONCERNS
3.0 CALIBRATION OF BALANCES AND CERTIFIED
WEIGHTS
4.0 ACCESSORIES
5.0 OPERATION AND DAILY MAINTENANCE

The first analytical instrument encountered by the laboratory technician is

the balance. The balance is a device used to determine mass. Specifically
designed to work on the surface of the earth, it measures the acceleration
force on the mass caused by the earth’s gravity and translates this
measurement into mass units. Balances used for mass determinations on
earth do not work accurately in space. Because balances measure the effect
of gravity on an object, they will give varied readings as the gravitational
force changes—a problem most commonly encountered at differing
altitudes.

1.0 TYPES OF BALANCES

The statue of justice at the entrance to the U.S. Department of Justice in
Washington, D.C., holds an archaic balance dating to ancient Egypt and
Greece. It consists of a balance arm held on a knife’s edge (pivot point),
with a pan attached to the end of each arm.
 The unknown sample is placed in one pan and known weights are added
to the other pan until the balance arm is level.
With a few technological upgrades, the modern laboratory balance
works much the same way. The single and triple (Figure 2.1) open-beam
balances have movable weights attached to a swing arm and can be used to
determine weights of a gram or more. The position of the weight along the
beam indicates the weight of the unknown. Single and triple open-beam
balances are frequently encountered in the microbiology laboratory, where
they are used in the preparation of media.
For more accurate weight determinations, the toploading and analytical
balances are used. These have limited weight ranges. In general, a balance
with a 4000-g capacity will only weigh to an accuracy of 0.1 g, whereas a
200-g capacity balance can measure with a 0.001-g sensitivity. Toploading
balances (Figure 2.2) are used in the laboratory to weigh out reagents for
preparation of solutions and for sample dispensing with a sensitivity down
to 10 mg, depending on the maximum capacity. Most balances have
incorporated electronics to some degree to speed up the determinations. The
electronics allow automatic zeroing, calibration checking, and simplified
tare determination. Some models will send the determined weights directly
to a computer system for automated data-logging.
The analytical balance (Figure 2.3) can be distinguished from the
toploading balance by the draft shield over the weighing pan. Analytical
balances are much more sensitive than toploading balances and have a
decreased weight range. The 200-g maximum capacity analytical balance
generally reads to 0.1 mg, whereas the 40-g capacity models will read to
0.01 mg. Microbalances are available that read to 0.0001 mg, with a
maximum capacity of 2 g.
The operation of the analytical balance is based on the leveling of a
balance arm on a knife-edge. In older mechanical models, weights are
added to the standard end of the arm as the dials are turned. In electronic
models, the standard end of the arm is coupled to an electromagnetic coil
with a light beam, and a photodetector cell is used to determine when the
balance arm is level. The amount of current required by the coil to bring the
balance arm into level is related to the mass on the determination pan by
comparison with standards.
FIGURE 2.1 Triple beam balance. Reprinted with permission from
OHAUS.

FIGURE 2.2 Top loading balance. Reprinted with permission from

OHAUS.
FIGURE 2.3 Analytical balance. Reprinted with permission from
OHAUS.

2.0 SITING AND ENVIRONMENTAL

CONCERNS
Balances are sensitive to gravity, humidity, temperature, static electricity,
dust and dirt, vibrations, and air currents. These mechanical devices
translate the effects of gravity into mass determinations; so, for maximum
performance, each balance must be calibrated in the location where it will
be used. The more sensitive the balance, the more important this becomes.
It is impossible to calibrate a balance in one location, move it, and expect it
to remain in calibration. Another facet of the interaction of the balance with
gravity is that the balance must be absolutely level with respect to gravity.
When the balance is level, the forces on the balance arm are straight up and
down at the knife-edge pivot point. If a balance is not level, part of the force
goes to a side-to-side motion along the knife edge, leading to an inaccurate
reading and a shortened life expectancy of the mechanism because of wear
at the pivot. Most balances have a bull’s-eye spirit level, and the feet adjust
up and down so that level can be achieved.
As the ambient temperature changes, the metal components of the
balance mechanism expand or contract, altering the length of the balance
arm and affecting the readings of the instrument. The best location for
analytical balances is in an isolated room by themselves with temperature
and humidity control. This may be a luxury that most laboratories cannot
provide. However, if the utmost accuracy is required, then a room must be
made available.
Humidity control is required to prevent the condensation of moisture
within the mechanical parts of the balance, which can affect the readings
and possibly lead to metal corrosion. The samples themselves are affected
by the humidity because of water adsorption during the weighing process.
This can be dramatically demonstrated by placing five or six fresh pellets of
sodium hydroxide in a weighing boat on the balance and then watching the
weight increase over 15 to 20 minutes. Laboratories in areas with high
humidity find it extremely difficult to obtain stable readings of oven-dried
materials.
Dust and dirt are the banes of balances because they limit the ease of
movement of the mechanical parts. For instance, the pan on a toploading
balance moves up and down through a guide underneath the pan. Spilled
chemical or sample on the top of the balance becomes lodged between the
guide and the travel rod, adding friction to the measurement. If the chemical
is corrosive, the metal parts of the balance can become chemically etched,
rendering the whole instrument useless. Most problems encountered with
toploading balances are directly attributable to dirt and spills. Chemicals
and samples should never be weighed directly on the pan of the balance.
Instead, a weighing boat or paper should be used.
 Vibrations prevent the balance arm from coming to rest at level and lead
to an oscillating reading. Both low- and high-frequency vibrations affect
balances. The higher-frequency vibrations can be eliminated by placing the
balance on a heavy solid platform such as a marble balance table or slab.
The slab should be supported by vibration dampers. This is not particularly
effective if the table or bench upon which the slab is resting does not
provide solid contact with the floor. Nor is it very effective when the analyst
is leaning an arm on the slab during the weighing process. Low-frequency
vibrations are more difficult to eliminate and may require locating
instruments on the lowest level of the building, which has a concrete floor
in direct contact with solid ground. Again, vibration dampers are generally
useful. In locations adjacent to railroads or highways, it may be necessary
to suspend using the balance when trains or heavy trucks are passing.
All analytical balances are equipped with draft shields, which prevent
the pan from swaying because of air currents. The draft shields should
always be closed when weighing. Balances should not be situated where
they are subjected to air flows from heating, ventilation, and air
conditioning vents. Balances should not be placed in halls or rooms with
doors on both ends unless the doors are kept closed to prevent air flow
through the room. Air flow through a room is strong enough to shake the
balance and can defeat the accuracy gained through the use of slabs,
vibration dampers, and marble balance tables.

3.0 CALIBRATION OF BALANCES AND

CERTIFIED WEIGHTS
The only person who can calibrate a balance is a balance service technician.
This is typically done on an annual or semiannual basis and involves a
general cleaning and lubrication of the internal mechanism of the balance.
The balance is calibrated by testing several Class l weights and adjusting
the mechanical and electronic response of the balance to give the correct
weight. Class 1 weights are weights that conform to construction materials
and mass tolerances established by ASTM (standard E617-91). There are a
variety of classes of calibration weights and their tolerances are shown in
Table 2.1.
 The National Institute of Standards and Technology (NIST) calibration
weights have been superseded by the ASTM standards and are becoming
very difficult to find. That is not to say that the existing NIST or the even
older National Bureau of Standards (NBS) calibration weight sets need to
be thrown away. A set of the Class S, Class S-1, or Class 2 weights should
be present in the laboratory to perform a daily calibration check on the
analytical and toploading balances. This is merely a check to ensure that the
balance is not in need of a service call. The indicated weight of the standard
from the daily calibration check needs to be recorded; it should be within
the certified tolerance of the weight and the sensitivity of the balance. For a
balance that reads to 0.1 mg, this would suggest that a 100-mg to 5-g Class
S or Class 2 weight should be determined daily and that ±0.1- to ±0.2-mg
readings would be acceptable. A range of weights should be read on a
weekly basis. For an analytical balance reading to 0.1 mg, this might consist
of the whole range from 10 mg to the maximum capacity of the balance.
This allows continuous monitoring of balance performance. When the
balance begins to consistently read higher or lower than the standards or
gives erratic day-to-day readings, it is time to call the service technician.
TABLE 2.1 Tolerances (mg) of various ASTM and NIST classes of
standard weights.

4.0 ACCESSORIES
The modem electronic balance with its keypad and digital display is
substantially faster to operate than the older chain-drive-operated double
pan balance, which has been relegated to the museum or display case.
However, the electronic balance is no more accurate than its predecessor,
and the same analytical considerations of appropriate measurements still
apply.
The maximum accuracy of the balance is obtained in the midrange of
the scale. Unfortunately, the maximum accuracy required in weighings is
frequently on the low end of the range. The maximum accuracy in weight
determination is obtained when the weight difference between the sample
and the sample container is large. These realities often compete against each
other. For example, a solids analysis is being performed, and the analyst has
a choice of a value based on the difference between the following weights:

Tare Final Weight difference (%)

35.0231 35.0256 0.0071
0.1231 0.1256 2.03

As far as the balance is concerned, the first set of weighings has an

uncertainty of 3 parts in 1,000,000, while the second set of weighings has
an uncertainty of 1 part in 1000.
However, the second case represents about 500 times greater confidence
in the analytical result. This is a specific example of the analytical concept
that measurement of a small difference on a large background is a much
poorer analytical situation than measurement of a large difference on a
small background.
The term, tare, mentioned in the preceding example is an entirely
different concept than zero. Zero is setting the balance to read exactly zero
(0.0000 on an analytical balance) when the pan is empty. Tare is the weight
of the empty sample container. The tare weight of the container should be
recorded in the balance logbook or on the sample analysis benchsheet. This
weight is subtracted from the final weight of the sample in the container to
arrive at the weight of the sample alone. Many electronic balances can
perform the subtraction automatically by setting the digital readout to zero
when the tare function is pressed; then when the sample plus container is
read, the indicated value is that of the sample alone. This is poor analytical
practice for several reasons. First, it gives the illusion that for a balance
with a maximum range of 200 g, the whole range is available for sample
determination when the tare is set to zero. In reality, the full 200-g range is
available only for the sample plus container with the empty pan set at zero.
Second, there is almost always an elapse of time between determination of
the weight of the sample container and determination of the final weight.
During this elapsed time another technician could have used the balance
and changed the tare value, making the final weight meaningless. Third,
there is no written record of the chain of calculations, which begins with the
initial sample amount tested and ends with the final parameter
concentration in the sample. Fourth, the ability of the balance to read
exactly zero with an empty pan is in itself a calibration check of the
balance. An inability of the balance to read zero on the empty pan is an
indication that the service technician needs to be called.
Samples and reagents should never be weighed directly onto the pan of
the balance. Instead a sample container is required. There are a large variety
of containers available from the supply houses that are appropriate for use
with balances. These containers can be broken down into weighing boats,
dishes, funnels, papers, and scoops. The disposable boats are made of
aluminum (Figure 2.4) or a variety of plastics (Figure 2.5). Ceramic boats
are commonly reused. Some may be specially designed sample introduction
containers for a particular piece of testing instrument. Dishes may be either
disposable or reusable. Reagent funnels (Figure 2.6) are made of
borosilicate glass and are designed to be upended into the neck of a
volumetric flask and to have the contents washed down into the flask with
solvent. Weighing papers are made of glassine or polyethylene-coated paper
to minimize sample sticking to the paper during transfer. The scoops
(Figure 2.7) are constructed of stainless steel and may be coated with
Teflon® to further reduce reactivity.
FIGURE 2.4 Aluminum weighing boats. Reprinted with permission from
Eagle Thermoplastics, Inc.

FIGURE 2.5 Disposable plastic weighing boats. Reprinted with

permission from Eagle Thermoplastics, Inc.
 Several accessories make working with analytical balances easier. The
first is a brush for removing small spills of reagents and samples in the
balance area. These brushes are constructed of camel hair or synthetic fibers
and frequently include an antistatic device, which eliminates static charges.
The antistatic device is a polonium cartridge that emits electrons, canceling
the static. When weighing light particles, such as polystyrene flakes or dusts
on air filters, static repulsion can make it next to impossible to get the
particles in the sample container. There are also antistatic guns available
that shoot out an electron stream when the trigger is depressed. Resting the
balance on a grounded antistatic pad, which is available from computer
supply companies, can also help with the static electricity problem. Another
useful tool is a set of forceps so that direct contact with the sample
container or calibration weight can be avoided. Fingerprints have mass and
can be detected with a sensitive balance. Forceps also make it easier to
manipulate small items.

FIGURE 2.6 Reagent (weighing) funnel. Reprinted with permission from

TWD Scientific, LLC.
FIGURE 2.7 Weighing scoop. Reprinted with permission from OHAUS.

Spatulas and scoops are essential for dispensing solid samples and
reagents into the weighing boat. A vibrating spatula is available for
dispensing small amounts of material to hit the exact desired weight of
reagent. Solvent and water squeeze bottles make quantitative transfer of the
weighed material into the flask or beaker easier by simply washing the
reagent off the boat directly into the final container.
Balances should be covered when not in use. This prevents accidental
spilling of reagents, solutions, and samples into the balance. Covering the
balance also minimizes dust accumulation in the mechanism.

5.0 OPERATION AND DAILY

MAINTENANCE
The balance is easy to use, as this example of total suspended solids (TSS)
determination demonstrates. First turn the balance on and let it warm up. Be
sure that the doors on the draft shield are closed every time a weight is
determined. Check to make sure that the bubble in the bull’s-eye level
indicator is exactly centered in the center circle. The second step is to
depress the tare button with the pan empty. It should stabilize at zero.
Record the reading and then select one of the calibration weights from the
set and carefully place it on the balance using the forceps. A weight about
the same mass as the intended weighings is the most representative. After
the readings stabilize, record the exact weight on the benchsheet or the
balance daily calibration check log. If the recorded weight is within the
tolerances of the calibration weight, weigh and record each of the sample
filters. Pick up the filter with the forceps, place it on the balance pan, and
close the draft shield door.
When the readings stabilize, record the weight on the benchsheet.
Remove the filter from the pan, and check that the balance returns to
exactly zero after each use. After all the filters are weighed, close the doors
on the balance, and clean up the area. Some analysts will reweigh the
calibration weight after the last sample to ensure that the balance is still in
perfect working order.
After the TSS test is completed and the filters are ready to be
reweighed, perform the final weight determinations by repeating the steps
in the above paragraph. With care and regular calibration by a service
technician, an analytical balance should give years of trouble-free
operation.
 3
Laboratory Ware

1.0 TYPES AND CHARACTERISTICS OF GLASS

2.0 LABORATORY GLASSWARE DESCRIPTION
3.0 CERAMICWARE DESCRIPTION AND
CHARACTERISTICS
4.0 PLASTICWARE DESCRIPTION AND CHARACTERISTICS
5.0 METALWARE DESCRIPTION AND CHARACTERISTICS

To be a successful carpenter, one needs to know about chisels, screwdrivers,

saws, hammers, drills, and so on. In a similar vein, a successful laboratory
analyst needs to know about the many items used in the laboratory. The
most common items in a laboratory are made of glass, plastics, ceramics,
and metal.

1.0 TYPES AND CHARACTERISTICS OF

GLASS
Containers made of glass are the single most common item found in a
laboratory. According to V.O. Altemose’s 1962 presentation, “Gas
Permeation Through Glass”, at the Seventh Symposium on the Art of
Glassblowing, glass is transparent, rigid, essentially nonpermeable to gases
(compared to plastic or metal containers), resistant to many chemicals, and
can withstand moderate heating, cooling, and moderately increased or
decreased pressures. Unfortunately, it is also fragile and subject to thermal
shock. In the old days more than 50 years ago, laboratory glassware used to
be made of soda lime glass or soft glass, so called because the glass would
soften and become workable in the flame of a Bunsen burner.
Laboratory technicians were expected to be able to make repairs to
chipped or broken glassware and to fabricate simple laboratory apparatuses.
Soft glass is still found in stock glass tubing and melting-point tubes. The
borosilicate, aluminosilicate, and quartz glasses used now require high
temperatures and specialized tools for fabrication. Although the borosilicate
glasses can be worked in an oxyacetylene torch, this is becoming a rare skill
in the laboratory, and most glassware is purchased from commercial
suppliers.
Glass is a supercooled liquid, not a crystalline solid, that becomes
strong when it is placed under compression. Tempering is the process of
heating glass to just under the softening point, then rapidly cooling the
surface with cold air. As the inside of the glass cools and shrinks, the
already hard surface is compressed and becomes strong. The piece
maintains the strength until a scratch relieves the compression force and the
piece disintegrates. Tempering can also be accomplished chemically by
substituting large ions such as potassium for small ions such as sodium.
Annealing is the process of allowing the fabricated piece to cool slowly to
dissipate all internal stresses. Annealed glass is nowhere near as strong as
tempered glass, when subjected to physical shock, but it has a higher
extreme working temperature. In general, as long as the glass surface is not
marred by a scratches, chips, or cracks, it is resistant to breakage. However,
a surface defect can lead to breakage at the slightest bump. All glassware
that is chipped or cracked should be discarded. Although the borosilicate
and quartz glasses are more resistant to breakage and not as sensitive to
thermal shock, the quartz is much more difficult to work with and will still
break when dropped.
Another aspect of glass as a super-cooled liquid is that it will soften and
deform at elevated temperatures, rather than sharply melting. Annealed
borosilicate glass containers, the most common laboratory glass, have a
normal working temperature range of up to 230 °C and should never be
used at temperatures above 490 °C. Ignoring this temperature restriction
will result in “slagging” of the container, which is particularly easy in a
muffle furnace. Quartz (Vycor®) and sintered powdered quartz containers
are appropriate for use in muffle furnaces, but ceramics are a less expensive
solution.
New glassware is slightly alkaline and should be soaked in 1%
hydrochloric or nitric acid overnight, then cleaned before use. Glass is
commonly cleaned with a soft brush, hot water, and any nonabrasive
detergent. Once the surface particulates and organics are removed, the piece
is rinsed with lots of hot water and then reagent-grade water. Stubborn dirt
can be removed by soaking in chromic acid solution (made from 10 g
sodium dichromate and 200 mL hot concentrated sulfuric acid) or one of
the commercially available chromic acid replacements. Chromium is a toxic
environmental pollutant and analysts should minimize its use in the
laboratory. Although potassium hydroxide-isopropanol baths have been
widely used for glassware cleaning, the actual process involves a chemical
attack and removal of the surface of the glass by the strong alkali. This
weakens the glass and makes it more prone to breakage. Ground-glass
surfaces are particularly sensitive to attack by alkali.
Glass is resistant to many chemicals, but not to hot phosphoric acid,
strong alkali, and substances containing ionic fluoride, such as hydrofluoric
acid and potassium fluoride. Metallic impurities in the glass are capable of
exchanging with metal cations in solution, particularly under acidic
conditions. Percentage-level materials found in glass are silicon, aluminum,
sodium, and boron, with subpercentage amounts of lithium, zirconium,
potassium, fluorine, chlorine, sulfur, and antimony.
In the early days of laboratory glassware and up until the last 40 years,
glass apparatuses, such as condensers, flasks, and tubes, were joined with
corks. Although true cork has been largely replaced by natural and synthetic
rubbers such as neoprene and polyethylene, the more common method of
joining glassware is with a ground-glass joint (Figure 3.1).
FIGURE 3.1 (a) Tapered and (b) spherical joints. Reprinted with
permission from Technical Glass.

The two common joint configurations are a tapered joint and a spherical
joint. Separate pieces of these types of joints are largely interchangeable
from one manufacturer to another because of American Society for Testing
and Materials (ASTM) standard E676 for standard taper joints (symbolized
as ) and ASTM E677 for the spherical (ball and socket) joints
(symbolized as ). The standard taper joints are identified by the largest
ground-surface diameter and taper height. For example, “24/40” means the
joint has a 24-mm diameter at the widest part of the ground surface and a
joint surface that is 40-mm long with a 1:10 taper. Spherical joints are size-
identified by the largest diameter of the ground-glass joint and the inside
diameter. For example, “35/25” means the joint has a 35-mm diameter
spherically ground surface and a 25-mm tube opening through the joint
(Figure 3.2).
The ball and socket joints have to be held together by clamps. These
range from a plastic ball and socket clip to a stainless steel screw pinch
clamp. The plastic clips do not hold the pieces of glass together firmly,
which results in leaks. They are also subject to breaking, which suggests
they are usable for only the most trivial joints, a rarity in the laboratory. The
metal pinch clamp is the only suitable way to ensure that a ball and socket
joint will stay together.
The O-ring joint is a variation of the ball and socket joint (Figure 3.3).
An ASTM standard for the joint does not exist yet, and each manufacturer
has a particular way of specifying the size. One manufacturer specifies the
size by the internal diameter of the tube running into the joint, whereas
another manufacturer specifies the pinch clamp size for the related ball and
socket joint.
FIGURE 3.2 Diagrams of the size designation for a 24/40 male standard
taper joint (left) and a 35/25 male spherical joint (right). Units are
millimeters.
FIGURE 3.3 O-ring joint. Reprinted with permission from Adams and
Chittenden Scientific Glass.

Stopcocks are an integral part of a number of glassware pieces, such as

burets and separatory funnels. The 1:10 taper of the stopcocks is the
same as that of the stoppers and standard taper joints. The size indication of,
for instance, a 2 “12/30” single straight bore glass stopcock means the
borehole in the stopcock is 2 mm in diameter, the diameter of the stopcock
at the centerpoint of the borehole is 12 mm, and the length of the ground-
glass surface is 30 mm. The symbol indicates a stopcock that is
manufactured to the Product Standard, ASTM E911for stopcock plugs
(currently made of polytetrafluoroethylene only). Sized similar to the
stopcocks, they have a 1:5 taper rather than a 1:10 taper. There are Teflon®
stopcock plugs commercially available that are designed to replace
glass plugs (Figure 3.4).
The machined surfaces of ground-glass joints have tight tolerances on
the sizing and curvature, which result in interchangeable parts to form a
leakproof joint, regardless of the manufacturer. This is strictly true only for
U.S. manufacturers because other countries have different standards. A
leak-free joint, of course, depends on the mating ground surfaces being
clean and particle free. It is possible to fracture a joint by having a hard
particle of grit between the ground surfaces of a spherical joint and then
attempting to form a leak-free seal by overtightening a pinch clamp. The
ground surfaces should always be wiped and inspected for grit before they
are lubricated and fitted together.

FIGURE 3.4 Stopcock plug. Courtesy of Corning Incorporated.

Glass-to-glass joints should always be lubricated. This prevents the

glass surfaces from freezing together and aids in the formation of a leak-
free seal. Silicone and petroleum-based lubricants are widely used when
organic solvents are not used. Water-based lubricants, such as glycerine, are
sometimes suitable when organic solvents are being used. Teflon® sleeves
are commercially available for use inside ground joints to prevent direct
glass-to-glass contact. As long as care is taken to keep the Teflon® sleeve
clean and free of cuts, this is the preferable solution. Teflon® bushings are
also available for use inside ground-glass joints. These have the advantage
of being thicker and thus sturdier than the sleeves, but at a substantial
increase in cost. The main advantage of Teflon® stoppers, sleeves, and
stopcock plugs is that they are chemically inert to most reagents and do not
require lubrication. However, they are subject to changes in size when the
temperature changes. A handy technique for freeing a stuck Teflon® stopper
or plug is to place the item in a freezer and wait for the Teflon® to shrink
and become loose. Teflon® stoppers and plugs are softer than glass, and
care must be taken to avoid cuts and deformations in the surface.
Particularly for stopcock plugs, abrasive dirt and small particles that get
into the joint between the ground glass and Teflon® surfaces can scratch or
cut the Teflon® when the plug is turned, which can lead to a leaky joint.
Stopcock plugs should never be stored tightened because the Teflon® can
deform. Teflon® forced into the glass borehole can then be shaved off when
the plug is turned, again leading to a leaky joint.
Stopcock metering valves, a variation of the standard Teflon® or glass
stopcock, allow more precise control of flow from burets. The flow of
liquid through a standard stopcock is regulated by the partial alignment of
the bore in the buret and the hole in the plug. With practice, the flow
regulation becomes easy for the experienced technician; however, each
individual buret has slightly different flow characteristics that can change
with how tight the plug is in the joint. The stopcock metering valve has a
screw-threaded stem that moves in and out along the axis of the plug to
restrict the size of the borehole, allowing much finer control over the flow.
The plug is turned in the stopcock metering valve only as an on-off control.
This type of flow control is similar to that used in the Rotaflo® buret valves
(Figure 3.5).

2.0 LABORATORY GLASSWARE

DESCRIPTION
Beakers come in two general types, the Griffin beaker and the Berzelius
beaker, both of which have a spout molded on the rim for easy pouring
(Figure 3.6). Spoutless beakers are also available. The Griffin design is
shorter and wider than the Berzelius for the same volume. Beakers come in
standard and heavy-wall construction. The heavy-duty rim of the heavy
wall beaker results in a more rugged container, which is better able to
tolerate mechanical washers and other shocks, but less able to withstand
rapid temperature changes. Beakers are also available with Teflon® coating
on the rim for better pouring characteristics and greater resistance to
chipping and cracking. Beakers are used for mixing solutions, holding
samples, performing reactions and digestions, and other general work. The
volume graduations on the sides are only approximate. Watch glasses are
commonly used to cover beakers.

FIGURE 3.5 Buret with rotaflow valve. Courtesy of Corning

Incorporated.
FIGURE 3.6 (a) Berzelius and (b) Griffin beakers. Courtesy of Corning
Incorporated.

Erlenmeyer flasks (Figure 3.7) are general-purpose containers and

reaction vessels. They are available in standard and thick-wall construction
with wide and narrow mouths. The wide mouth is better for mixing and
performing titrations. A low-actinic Erlenmeyer has a UV absorbing stain
fused on the surface, giving the glass a red color for working with solutions
that are sensitive to light. Erlenmeyer flasks are made with plain rim
openings for closure with stoppers of cork, soft rubber, hard rubber,
neoprene, silicone, polyethylene, and other materials. Rubber stoppers are
sized according to the chart in Table 3.1 (cork stoppers have different size
designations). The 000 stopper is 21-mm long, sizes 00 to 13 are 25-mm
long, and sizes 14 and 15 are 39-mm long. Size 13½ is provided variously
as 25- or 35-mm long. The natural and synthetic material stoppers have a
wide range of chemical sensitivity and resistance, so the stopper material
must be matched to the chemicals used. This would require having a large
variety of different stoppers available in the laboratory. Most laboratories
avoid this problem by not storing materials in plain-top Erlenmeyer flasks;
rather, they use the more specialized ground-glass or screw-thread capped
flasks. If a top needs to be placed on a plain Erlenmeyer flask to prevent
contamination, an inverted beaker or small watch glass placed over the
opening works well. Another technique is to invert a much smaller
Erlenmeyer flask to the neck of the flask. A glass cap that functions like the
inverted smaller Erlenmeyer flask

FIGURE 3.7 Erlenmeyer flasks with screw caps. Courtesy of Corning

Incorporated.
TABLE 3.1 Rubber stopper sizes.

is also available. For sterilization of culture media in Erlenmeyer flasks, a

wad of cotton gauze is often used as a stopper. Use of the inverted beaker or
flask as a cover for flasks in the autoclave is another alternative.
For storage of solutions and materials, Erlenmeyer flasks are available
with groundglass, polyethylene, or Teflon® stoppers, or with molded
threads for screw caps. The ground-glass tops are to the standard.
Watch glasses (Figure 3.8) are shallow, bowl-shaped circles of glass
widely used as tops for beakers so that the technician can observe what is
happening in the beaker without opening it. The main function of the watch
glass is to prevent contaminants from falling into the open beaker. Watch
glasses are made with either a plain, smooth surface or with ribs molded on
the outside of the curve. The ribs are present so that the beaker is not sealed
while the contents are heated to boiling and the hot vapors will condense on
the cooler watch glass, dripping back into the solution.
Glass bottles are used to hold samples, reagents, and solutions for
storage and to perform bacteriological tests on samples of 100 mL and
larger (Figure 3.9). Some bottles, such as acid storage and other reagent
bottles, come with ground-glass necks to accommodate stoppers; however,
most bottles have a threaded neck to allow closure with a screw cap. A
special type of thick-wall bottle is used for centrifuging large amounts of
material. Sampling bottles vary in size, from the 40-mL vial for samples for
volatile organic analysis to the 120-mL wide-mouth jar for solid samples to
the 1000-mL narrow-mouth amber glass bottle for general semivolatile
organics analysis, and the 1000-mL wide-mouth clear glass bottles for
solids and sludges.
Many solvents and concentrated acids are packaged in 4-L screw-cap
glass jugs by suppliers. These fragile containers, particularly those used for
concentrated acids, for the last several years have been coated with plastic
to contain the potentially catastrophic spill of about a gallon of concentrated
acid when the jug is accidentally dropped. Jugs are particularly prone to
having the bottom break off when they are cracked by tapping against
another jug or hard object. Bottles should never be used for the preparation
of reagents or heated on hotplates; they are only storage containers.
FIGURE 3.8 Ribbed watch glass used with a beaker. Courtesy of Corning
Incorporated.

FIGURE 3.9 Narrow-mouth reagent bottle. Reprinted with permission

from Corning Incorporated.

Vacuum flasks (Figure 3.10) are also referred to as filtering flasks.

These are heavy-wall Erlenmeyer-style flasks with a side arm in the neck
for attachment to a vacuum source and a flattened rim to facilitate
placement of one of a variety of funnels on top. A rubber, neoprene, or
other soft-material gasket should be used between the funnel and the flask;
direct glass-to-glass contact is to be avoided at all times. Vacuum flasks are
not designed to be heated; however, placement in a cold bath may be
possible with care. The flasks are available with either molded side arms or
replaceable side arms that are attached by a grommet.
Test tubes and culture tubes are used for many purposes in the
laboratory (Figure 3.11). They are available with a plain top (rimless), a
beaded rim for greater strength, a ground-glass closure top, or a molded
screw-thread top. They are also available in a standard wall and a thick wall
for more rugged use. Common test tubes have a constant internal diameter
for the length of the tube and a rounded bottom. They are sized in
millimeters by the outer diameter and the length of the tube. For example, a
16 × 125 tube has a 16-mm outer diameter and is 125-mm long, with an
approximate capacity of 15 mL. Centrifuge tubes are heavy-wall versions of
the test tube, designed to withstand the pressures generated in a centrifuge.
They are frequently tapered in a long or shallow “V” at the bottom, and are
marked with volume graduations for rough assessment of sediment
amounts.
FIGURE 3.10 Vacuum flask with side arm. Courtesy of Corning
Incorporated.
FIGURE 3.11 Test tube. Courtesy of Corning Incorporated.

Kjeldahl flasks (Figure 3.12) are thick-wall, round-bottom flasks with a

long neck for use in the Kjeldahl digestion procedure for total organic
nitrogen content. They are designed for the severe thermal and chemical
abuse encountered in the procedure. Care must be taken to avoid direct
contact between the bottom of the flask and the heating element used in the
digestion. Kjeldahl flasks are available in a variety of sizes, from 30 to 800
mL.
FIGURE 3.12 Kjeldahl flask and condenser. Courtesy of Corning
Incorporated.

Condensers are used to convert solvent vapors from a heated flask back
into a liquid by cooling. One of the two basic designs is a jacketed tube
with coolant passing through the jacket. The vapors are cooled on the walls
of the inner tube and return to the liquid state. Condensers of this design
include the West, Graham, Allihn, and Liebig styles (Figures 3.13 and
3.14). The Liebig and West designs have a straight, constant-bore inner
tube, the Allihn has a bulbed inner tube, and the Graham has a coiled inner
tube. In the other basic design, vapors pass over and around a central cooled
shape. Condensers of this design include the Friedrichs, Hopkins, cold
finger, and dry-ice condenser.
The coolant source attachment nipples are designed for direct
attachment of a rubber hose. A metal band screw-type hose clamp can be
added over the hose on the nipple to form a secure attachment. A less
desirable temporary solution is to wrap a turn or two of wire around the
hose on the nipple and tighten it with a pair of pliers. If the wire is tightened
too much it can cut the hose on the nipple and make the attachment less
secure.
FIGURE 3.13 (a) West, (b) Graham, (c) Allihn, and (d) Liebig jacketed
condensers. Reprinted with permission from Corning Incorporated.

FIGURE 3.14 Jacketed condenser with ground-glass fittings. Reprinted

with permission from DWK Life Sciences.
 Some coolant nipples have molded screw threads for attachment of the
coolant hoses with a threaded connector. These are particularly useful when
equipment is constantly being set up and taken down. Another alternative in
a setup is to attach a short piece of hose to the condenser using a hose
clamp and attach a rapid-disconnect adapter to the other end. The hose
should still be secured to the threaded connector or adapter with a hose
clamp.
The source of coolant for condensers has, for many years, been the cold
water tap of a nearby sink. This is an amazing waste of water that has been
purified for human consumption. A more sensible source of coolant is a
recirculating cooler. Added advantages of the recirculating cooler are that it
can be set up anywhere there is an electrical outlet and the coolant, typically
1:1 ethylene glycol:water, can be made as cold as 215 °C, which is much
colder than any tap water. The coolant is pumped with sufficient force that
up to eight or 10 condensers can be serviced in either a series or parallel
configuration by a single cooler. This increased pumping pressure
absolutely requires that all hose connections be made using hose clamps. If
a hose comes loose from a condenser, the ethylene glycol-water mixture
will spray up to approximately 15 m (50 ft) from the end of the hose,
soaking every item in the laboratory. After the cooler has emptied itself
(about 15 to 30 seconds), the cooling chamber will have no load on it,
which could result in cooler burnout. Most recirculating coolers are also
capable of heating the circulating fluid so that it can be used as a heat
source.
Coolant should be introduced to the condenser to completely fill the
cooling space. In the jacketed condensers, coolant enters at the bottom of
the jacket and exits at the top. In the internal condensers, a tube typically
extends from one of the nipples to the very bottom of the cooling space.
This is the entry nipple. There should be no air spaces in the coolant area
when the condenser is used properly. Condensers can be linked together in
series by running a piece of hose from the exit nipple of one condenser to
the entry nipple of the next. The first and last condensers in the series are
attached to the coolant source. Extra coolant must be added to the
recirculator to keep all the condensers filled as well as the cooling chamber.
FIGURE 3.15 Powder funnel. Courtesy of Corning Incorporated.

Funnels come in a large variety of shapes and sizes. Their primary use is
to assist in the transfer of liquids or solids from one container to another.
The use of funnels with large-diameter stems (called powder funnels; Figure
3.15) reduces the time for the transfer. An air escape should always be
provided when transferring liquids to avoid having air bubbles travel
backward up the stem, slowing the liquid flow and often resulting in liquid
being slopped out of the funnel when the air bubbles break. A simple air
escape can be provided by placing an opened paper clip over the rim of the
container before placing the funnel in position. The second main use for
funnels is to separate liquids from solids. How this is performed and the
apparatus used depend, to a great extent, on whether the liquid or solid is of
greater interest. When the liquid is the phase of interest, a simple gravity
filtration will often suffice. A suitable apparatus is prepared by placing a
folded filter paper into the funnel and then pouring the sample through it.
Funnels and filter paper circles come in a range of sizes (Figure 3.16).
Funnels are sized by the internal diameter of the top of the funnel, and filter
paper circles are sized by the diameter of the circle. Because most funnels
are of a 60-deg. angle, doubling the diameter of the top of the funnel will
give the approximate paper size. For example, a 65-mm funnel would
require a 12.5-cm filter paper. A filter paper should never extend above the
rim of the funnel; therefore, a slightly smaller diameter than calculated is
the appropriate size.

FIGURE 3.16 Plain long-stem filtering funnel. Courtesy of Corning

Incorporated.
 Papers are meant to be folded before being placed in the funnel. Folding
the paper in half and then in half again, then opening one side up, will give
a shape that closely conforms to the sides of the funnel. This looks very
neat; however, when the sample is poured into the paper in the funnel, the
liquid seals the paper to the sides of the funnel and the only area of rapid
filtering is at the tip of the folded paper. Some funnels have molded flutes
on the inside surface of the funnel to allow a larger portion of the filter
paper surface to contribute to the filtering. If the solid to be filtered is not
heavy, all of the surface of the paper can be involved in the rapid filtering
by making the second fold of the paper a little shy of exactly half and by
tearing off approximately 6 mm (0.25 in.) of the corner. The paper will then
sit in the funnel supported only by the top edge. In many cases, however,
the weight of the solid and liquid forces the wetted paper to conform to the
sides of the funnel with a subsequent reduction in flow. A pleated paper,
also called fluted paper, is the common solution. Filter papers can be folded
into pleats, which is somewhat time-consuming, or they can be purchased
already pleated.
FIGURE 3.17 Perforated funnel with vacuum attachment. Reprinted with
permission from DWK Life Sciences.

The problem with using the whole surface of the paper for filtering is
solved by changing the design of the funnel. A Buchner funnel has straight
sides and a perforated support disk for the paper (Figure 3.17). The older
Buchner funnels are ceramic (Figure 3.18), but the more recent models are
glass, with perforated or fritted (blown) glass bottoms (Figure 3.19). A filter
paper of the same diameter as the bowl of the funnel is placed in the
bottom, wetted with liquid; then, the sample is added. The solid and paper
are removed from the funnel by simply inverting it over a container. In
some applications, the fritted glass disk itself is used as the filter rather than
as a support for filter paper, particularly when the solution, such as a strong
acid, will dissolve the paper. This works for coarse solids, which are easily
removed from the funnel by dumping and rinsing. However, once fine
particles have become lodged in the pores of the disk, it is almost
impossible to remove them. The fritted disk itself is fragile and can be
abraded by scraping with glass stirring rods or metal spatulas. The fritted
disks are available in a range of pore sizes, as indicated in Table 3.2.

FIGURE 3.18 Porcelain Buchner funnel. Courtesy of VEE GEE Scientific,

Inc.
FIGURE 3.19 Fritted Buchner funnel. Courtesy of Corning Incorporated.

TABLE 3.2 Pore sizes of fritted glass filters.

Class Pore size (μm)

Coarse 40–60
Medium 10–15
Fine 4–5.5
Very fine 2–2.5
Ultrafine 0.9–1.4

Fritted glass Buchner funnels should be cleaned by rinsing solvent or

water through the filter in the reverse direction. This can be done by simply
inverting the funnel in the sink and filling the stem with rinse liquid.
Gravity serves to pull the liquid down through the fritted disk. A faster
washing can be achieved by inverting the funnel into a large rubber gasket
in the mouth of a vacuum flask. A vacuum source is used to suck the liquid
rapidly through the disk. This method has the added advantage of sucking
the disk dry once the washing and rinsing are complete. Another technique
is to run a length of hose from the faucet on the sink to the end of the stem
of the funnel. Turning on the faucet forces water through the disk with some
pressure. Caution should be exercised when turning on the faucet, as heavy
water flow can loosen the hose and spray water all over the technician and
the laboratory. Additionally, if the pressure causes the technician’s grip on
the funnel to slip, the funnel typically breaks when it hits the side of the
sink. The advantage of this technique is that it often will dislodge particles
caught in the pores of the filter disk.
Membrane filters are synthetic materials with controlled pore size
ranges. In general, they are fragile, and special glass apparatuses called
vacuum filter holders are used to support them during use (Figure 3.20).
The bowl and membrane support base are two separate parts held together
by either a clamp or a magnetic catch. The stem to the membrane support
base may have a ground-glass joint for attachment to a specially designed
receiving flask, or it may have a rubber stopper for use with a vacuum flask.
In general, the latter design is more versatile and less expensive.
Separatory funnels (Figure 3.21) are designed to facilitate liquid–liquid
extractions. The idea behind the liquid-liquid extraction is to remove
organic analytes from the water by partitioning them into an organic
solvent. The basic design of the separatory funnel has not changed for many
years, although the materials used for the stopcock plug and the stopper
have changed from the original glass to Teflon® and other plastics. Liquid–
liquid extractions involve mixing the sample, which is typically water
based, with an organic solvent such as methylene chloride or hexane,
shaking for approximately 2 minutes, and then allowing the layers to
separate. The lower layer can be drawn off through the stopcock into a
beaker or flask used as a receiver. If a heavier-than-water solvent like
methylene chloride is used, the bottom layer is the organic layer. If a
lighter-than-water solvent, such as hexane, is used, it will end up on the top.
FIGURE 3.20 Funnel and support assembly. Reprinted with permission
from DWK Life Sciences.
FIGURE 3.21 Separatory funnel. Courtesy of Vee Gee Scientific, Inc.

A significant problem encountered when using the separatory funnel is

the formation of emulsions that interfere with the clean separation of the
liquids into two layers. A number of standard procedures can be used to
break the emulsion-centrifugation. These include physically stirring the
emulsion with a glass or Teflon® rod or heating with a hot air dryer;
however, these are all tedious remedies. The continuous liquid-liquid
extractor is designed to avoid formation of an emulsion. It heats the solvent
in the receiver to boiling, then condenses the vapors to a liquid above the
sample, passes them through the sample as drops, and returns them to the
receiver. There are different designs based on the density of the extracting
solvent. In heavier-than-water extractors (Figure 3.22), solvents condensing
in the cooling tower simply drip down through the sample, collect in a pool
at the bottom of the sample reservoir, then are returned by siphon to the
receiver. In lighter-than-water extractors (Figure 3.23), solvents condense in
the cooling tower, fall into a tube that opens at the bottom of the sample
reservoir, float upward through the sample, collect in a layer over the
sample, and then simply overflow back into the receiver.

FIGURE 3.22 Heavier-than-water continuous liquid-liquid extractor.

Used with permission of Sigma-Aldrich Co. LLC.
FIGURE 3.23 Lighter-than-water continuous liquid-liquid extractor. Used
with permission of Sigma-Aldrich Co. LLC.

The continuous liquid-liquid extractor is a more efficient extractor than

the separatory funnel when one considers hands-on technician time required
for operation and the degree of recovery of target analytes from the water
matrix. Still, it takes 16 to 24 hours of operation for reasonable transfer of
the desired analytes to the receiver. This is mostly because the small
droplets of solvent containing the extracted materials are diluted into the
much larger solvent layer in the sample reservoir and then slowly
transferred back to the solvent receiver. Complete removal of highly water-
insoluble materials such as diesel fuel with methylene chloride as the
solvent is completed within the first hour of operation, whereas more water-
soluble analytes, such as the nitrophenols, can require 4 to 6 hours of
operation. Replacement of the siphon transfer tube and elimination of the
solvent pool at the bottom of the sample reservoir by placing a solvent-
permeable membrane at the bottom of the reservoir cuts the required
operation time for the continuous liquid-liquid extractor to the 4 to 6 hours
necessary for only the analyte extraction process (Figure 3.24). As soon as
the solvent droplets reach the membrane at the bottom of the sample
reservoir, they are passed through to the solvent receiver for recycling
through the apparatus. An added advantage of the device is reduced solvent
use (less than 100 mL compared to 300 to 500 mL for the traditional
design). Further, the membrane serves to remove water from the solvent,
thus eliminating a separate drying step. This apparatus is only available for
use with heavier-than-water solvent.

3.0 CERAMICWARE DESCRIPTION AND

CHARACTERISTICS
Ceramic or porcelain materials are used whenever high temperatures are
required in the procedure and a chemical resistance similar to that of glass
is needed. Ceramics are made from clays that are molded when wet into the
desired shape, allowed to dry, glazed, and then fired at high temperatures
inside kilns. Unglazed porcelain is a porous, rough material with a low
degree of resistance to abrasion. The glaze seals the surface of the ceramic
and protects the underlying bulk of the piece. Ceramics are not as fragile as
glass; however, they will chip. Once a piece is chipped and the underlying
porcelain exposed, it will absorb water and other liquids easily, rendering
the piece useless for gravimetric work.
FIGURE 3.24 Accelerated one-step continuous liquid-liquid extractor.
Courtesy of Corning Incorporated.

The most common uses of ceramic materials in the laboratory are for
spot plates, crucibles, dishes, mortars and pestles, and spatulas (Figure
3.25).
Crucibles are available with solid, nonporous bottoms or with
perforated, porous bottoms and are designed for ignition of the contents at
temperatures up to 1150 °C. The latter design is called a Gooch, or filtering
crucible. Gooch crucibles were once prepared for use by addition of a slurry
of asbestos fibers in water to form a mat on the bottom. A small, perforated
ceramic disk called a Witte plate was then placed over the asbestos mat,
then a second layer of asbestos was laid down over the Witte plate. After
drying and firing, the Gooch crucible was ready for use. Concern over
asbestos as a health hazard has led to the use of glass fiber filters instead of
the asbestos mats and, as a result, the Witte plates have disappeared from
commercial supply house catalogs. Solid-bottom crucibles and dishes are
still commonly used for evaporating and drying water and sludge samples
for determining the various classes of solids.

FIGURE 3.25 (a) mortar and pestle, (b) spotplate, (c) crucibles, and (d)
dishes. Courtesy of Vee Gee Scientific, Inc.
 A mortar and pestle are used for grinding and mixing solids. The inside
surface is unglazed because the roughness adds to the efficiency of the
grinding. This also results in the inner pores becoming clogged with the
ground sample, rendering them virtually impossible to decontaminate. Once
a ceramic mortar is used for one type of sample or chemical, it should be
reserved for that application in the future. Other materials used for mortar
and pestle construction, such as agate, glass, aluminum oxide, synthetic
sapphire, and iron, are more expensive, but easier to clean, and may be
more appropriate than ceramic in certain applications.

4.0 PLASTICWARE DESCRIPTION AND

CHARACTERISTICS
Synthetic materials are rapidly becoming dominant in the laboratory. For
some applications, they are unrivaled as fabrication materials, whereas in
others they offer significant challenges to the use of more traditional
materials. Plastic laboratory ware is reusable, leakproof, unbreakable, and,
in general, inexpensive. The generic term, plastic, covers a large number of
different materials with a wide range of physical, chemical, and mechanical
properties. The mechanical properties of the plastic place limitations on
what shapes and laboratory devices can be molded from the material.
Tubing molded from a hard, brittle resin will not be as usable as a softer,
tougher formulation, whereas a beaker or flask must be made from a rigid
material. The physical properties will define the temperature range within
which the plastic can be used. Except for the fluorocarbon resins, most
plastics cannot be used above 100 °C and significant softening may occur
slightly above room temperature. Maximum operating temperatures for
some resins usable at higher than 100 °C are polycarbonate—135 °C,
polypropylene—135 °C, polysulfone—165 °C, polymethylpentane—175
°C, and polyphenyleneoxide—121 °C. Many plastics become brittle and are
subject to shattering below 0 °C, whereas others are designed for use to
−100 °C and lower. In general, manufacturers do not mold products from
inappropriate materials, and, as long as the item is used as it was intended,
it will serve well. For instance, polyethylene squeeze bottles are ideal for
storing acid or base solutions; however, preparing the solution by diluting
the concentrated acid or solid base directly into the squeeze bottle will
generate a lot of heat and result in the bottle slagging.
Of greater importance to the laboratory worker is the chemical
resistance of the material, or perhaps chemical compatibility is a better way
to look at the concept. The technician who is used to glass containers and
other highly chemically resistant laboratory tools may be somewhat
surprised when a polystyrene disposable pipet clouds and dissolves on
contact with a methylene chloride solution. Manufacturers of plasticware
provide a chemical resistance chart in their catalogs, and it is up to the
technician to know what chemicals may or may not be used with which
tools. Following the same line of thought, it is also necessary to know what
resin is used to make each product. To this end, material or recycling codes
are often molded into the bottom of the piece, particularly bottles, beakers,
and flasks. The recycling codes and symbols were established by the
Society of the Plastics Industry (SPI). The ASTM codes are also found
molded into plastics and are similar to those used by SPI, but without the
arrows in the triangle (Figure 3.26).
FIGURE 3.26 ASTM resin identification codes. Reprinted with
permission from ASTM.

Items constructed of one of the fluorocarbon resins exhibit a wide range

of chemical resistance, along with extended use to high temperatures (up to
200 to 250 °C), depending on the exact material. They are also the most
expensive. The fluorocarbons include the Teflon® (E.I. du Pont de
Nemours) materials and others. Teflon® is a trademarked class term for
three related resins, polytetrafluoroethylene (TFE or PTFE),
tetrafluoroethylene-hexafluoropropylene copolymer (FEP), and
®
perfluoroalkyoxy polymers (PFA). Teflon tetrafluoroethylene is white,
opaque, dense, and one of the slickest materials known, with a usable
temperature range from −200 to 250 ° C. Tetrafluoroethylene-
hexafluoropropylene copolymer is translucent, with a lower operating range
(−270 to 205 °C) and similar chemical resistance.
Perfluoroalkyoxy polymers are translucent, resistant, and have a usable
temperature range from −270 to 250 °C, with better strength and creep
resistance than polytetrafluoroethylene. Other fluorocarbon materials are
Tefzel® (E.I. du Pont de Nemours) (ethylene-tetrafluoroethylene
copolymer; ETFE or PETFE); Halar® (Ausimont) (ethylene-
chlorotrifluoroethylene copolymer; ECTFE or PECTFE); Kel-F® (3M)
(polychlorotrifluoroethylene; PCTFE, also called Halon and Dallon);
Tedlar® (E.I. du Pont de Nemours) (polyvinyl fluoride); and Kynar®
(Pennwalt Corporation) (polyvinylidene fluoride; PVDF).
The next most chemically resistant group of resins is that of the
polyolefin resins. These include polyethylene, polypropylene,
polymethylpentene (PMP or TPX), polystyrene, and polyvinyl chloride
(PVC). Polyethylene is available in many forms, including low-density
polyethylene (LDPE), high-density polyethylene (HDPE), linear low-
density polyethylene (LLDPE), and cross-linked high-density polyethylene
(XLPE). High-density polyethylene containers are preferred for sampling
bottles because of their low permeability, high chemical resistance, and
ability to hold the molded shape while still being somewhat flexible at 0 °C.
Low-density polyethylene molded items are flexible, but permeable to
gases and solvents. The cross-linked HPDE forms huge three-dimensional
molecules, resulting in high chemical resistance with superior structural
integrity and is particularly suitable for large storage tanks. Fluorinated
polyethylene or polypropylene molded items have been exposed to fluorine
gas, which results in substitution of fluorine for hydrogen on the surface of
the polymer. This creation of a chemically bonded layer of fluorocarbon
improves the chemical resistance of the item without the substantial
increase in cost of a solid fluorocarbon structure. The layer is thin, and care
must be taken to avoid scratching or scraping the surface, which exposes
the underlying material. Polyethylene terephthalate (PETE or PETG) and
polybutylene terephthalate (PBT) are common resins used to mold soft
drink containers, but they are not suitable for laboratory ware. Items made
from polystyrene or polymethylpentene are transparent, rigid, and brittle
materials, and the polystyrene items are subject to organic solvent attack.
Pure PVC molded items are translucent and offer good chemical resistance,
being somewhat flexible when the walls are thin. Blending PVC with
phthalate plasticizers gives a soft, flexible tubing marketed as Tygon®
(Norton Company) that is crystal clear. Other available tubings are made of
PVC, LDPE, polytetrafluoroethylene, and polyurethane (PUR) resins,
without any plasticizer additives. These, in general, are stiffer than the
Tygon® tubings, but they reduce laboratory contamination problems. With
the exception of Tygon® tubing, items molded from the polyolefin resins
are generally made without plasticizers.
Other resins used to make laboratory items include acetal (ACL,
Delrin® [E.I. du Pont de Nemours]), polyetheretherketone (PEEK), nylon
(NYL, polyamide), polycarbonate (PC), polysulfone (PSF), polyethylene
terephthalate (PETE or PETG), polyurethane (PUR), polyphenylene oxide
(PPO), nitrile, and neoprene. The acetal and PEEK resins are used to make
test tube racks and other items with machined surfaces that have to hold
their rigid shape under repeated use-for example, the screw threads on
tubing connectors. The fluorocarbon and olefin-based resins generally are
too soft for these purposes. Acetal is made from polymerized formaldehyde.
Subjecting racks made of acetal to sterilizing conditions in an autoclave will
depolymerize the resin, resulting in release of large amounts of toxic,
choking, formaldehyde fumes. Polyetheretherketone will swell slightly
upon exposure to tetrahydrofuran, methylene chloride, or dimethyl
sulfoxide, which can result in thread-joined items freezing together.
Allowing the item to sit in the air for several hours will reverse the swelling
and the joined parts can be screwed apart. For ferrules made of PEEK,
exposure to these chemicals actually makes the connection more secure.
Polycarbonates are rigid, tough, and crystal clear, making them ideal
materials for safety glasses and blast shields. Polycarbonate safety glasses
are opaque to UV light, offering further protection for the user.
 Probably the first plastic item encountered by the analyst in the
laboratory is a wash bottle (Figure 3.27). The bottle is most commonly
made of LDPE with the stem and closure constructed of polypropylene. A
right-to-know safety label may be permanently printed on the bottle (Figure
3.28). Fluorinated versions of the right-to-know wash bottles for organic
solvents are available. Teflon®-constructed wash bottles are used in cases
when contamination is of utmost concern. Preprinted right-to-know labels
made of self-adhesive PolyPaper® (Nalge Company) will stick to the
Teflon® wash bottles, which otherwise are difficult to label.
A variation of the wash bottle is the variable-volume dispenser (Figure
3.29). These are particularly useful when an approximate volume of
solution is required to be repetitively dispensed. The plunger is raised and
then depressed to dispense an approximate volume of solution.

FIGURE 3.27 Wash bottle without right-to-know label. Reprinted with

permission from Thermo Fisher Scientific.
 The membrane filtration procedure in microbiology testing calls for use
of 50-mm-diameter Petri dishes for the incubation of the filter membrane
(Figure 3.30). These dishes are most frequently purchased as presterilized
polystyrene, with a sterile pad already inserted in the dish, ready for the
addition of the broth and membrane filter. Other sizes of culture dishes are
widely used in microbiological testing procedures. They come presterilized,
ready for the addition of the agar solution and sample. Disposal consists of
autoclaving the used dishes to sterilize them, followed by incineration or
other means.

FIGURE 3.28 Wash bottle with right-to-know labels. Reprinted with

permission from Thermo Fisher Scientific.
FIGURE 3.29 Variable-volume reagent dispenser. Reprinted with
permission Thermo Fisher Scientific.
FIGURE 3.30 Petri dishes. Reprinted with permission from Hach
Company.

Gloves are used in the laboratory to protect the technicians’ hands from
infectious materials, toxic or irritating solvents, caustic chemicals, broken
glassware, and hot or cold surfaces. In the laboratory, no one type of glove
will perform all these functions. When analysts must protect their hands
from chemicals or biological agents and a “rubber” glove is to be worn,
they are faced with a surprising array of choices. A latex glove may be
suitable for working in the general chemistry laboratory, but a nitrile glove
may be more appropriate if organic-solvent extractions are being
performed. For most purposes, the main considerations are resistance to the
particular class of chemicals, required dexterity when wearing the glove,
and potential contamination of the analytical samples. Secondary
considerations are comfort and cost. Gloves are available in natural latex
rubber, neoprene, PVC, nitrile, butyl, polyethylene, Teflon®, and Viton (E.I.
du Pont de Nemours) (fluoroelastomer); each has a specific range of
chemical resistance, which should be checked by the analyst. When
working with hot items in a muffle furnace or cold materials around dry ice
(solid carbon dioxide) temperatures, temperature-resistant gloves should be
used. A heat-resistant glove woven with a silica fiber outside and cotton
lining inside is very different from a cryogenic temperature-resistant glove
with polyolefin insulation between layers of nylon. Cut-resistant gloves
made of Kevlar™ (E.I. du Pont de Nemours) fiber and cotton lining with
nitrile dots for gripping are useful when cleaning up broken glassware.
Heavy-duty gloves made of butyl, Viton, or neoprene are typically used
when handling bottles of concentrated acid. A glove made of two edge-
sealed Teflon® sheets cut in the shape of a mitten offers the ultimate in
chemical resistance, but is about as flexible as a baseball glove. Dexterity is
related to the stiffness and thickness of the glove. As more dexterity (i.e.,
thinner glove) is required, the comfort, durability, and chemical resistance
suffer.
In general, a medical examination glove will not be useful in a
laboratory. There may be situations that call for a sterile glove; however,
other factors will drive glove selection. Gloves that are powdered or treated
with hand lotion present a large risk of sample contamination. Moisturizing
hand lotion applied by the analyst may also pass through the glove and
reach the sample. Analysts can check their gloves for contamination
potential by processing a pair of the gloves as if they were a sample. They
can be submerged in a portion of reagent water, then the reagents used in
the test added sequentially. The results can frequently lead one to a change
in gloves, sacrificing comfort for chemically compatible handwear. The
analyst may want to consider wearing two pair of gloves, the inner for
comfort and the outer for protection.
Disposable pipets are frequently used in the microbiology laboratory.
They come in either polystyrene (polystyrene sereological pipet is shown in
Figure 3.31) or polypropylene. The polystyrene pipets are truly single use
and disposable, and are used only for water solutions. The polypropylene
pipets are resistant to many chemicals, unbreakable, reusable, and
autoclavable; these are features not shared with the polystyrene pipets. The
polypropylene pipets are available in transfer, Mohr, and serological forms.
FIGURE 3.31 Serological pipets. Reprinted with permission of Thermo
Fisher Scientific.

Disposable pipetter tips are most commonly made of polypropylene

(Figure 3.32). They are available presterilized for microbiological use. A
special metal-free tip is preferred for use in trace metals analysis because
plastics are frequently prepared with a variety of metals used as
polymerization catalysts, which end up in the plastic resin as trace-level
contaminants. The disposable pipetter tips are excellent for use with water-
based samples, reagents, and solutions; the possibility of contamination or
analyte adsorption precludes their use with organic solvents or in organic
analysis.
FIGURE 3.32 Disposable pipetter tips. Reprinted with permission from
Bio Plas, Inc.

Chemically resistant and nonbreakable, Teflon® beakers (Figure 3.33),

evaporating dishes, flasks, separatory funnels, and other containers are
durable and ideal for trace analyte analysis, but expensive. The drawback of
using Teflon® laboratory ware as a replacement for glassware is that it is
subject to surface scratches and cuts, offering places for dirt to accumulate
and leading to sample contamination through carryover. Teflon® beakers
and flasks are excellent for use with hot acid digestions on hot plates as
long as the analyst takes care to not let them boil dry and exceed the upper
temperature limit of the fluorocarbon. Microwave digestion vessel liners are
made of Teflon® because they are inert to the hot acid matrix and
transparent to the microwaves. This includes use of hydrofluoric acid as the
digestion medium, an impossibility with glass digestion vessels.
 Tubing is used for many purposes in the laboratory. Latex and natural
rubber tubing have been in use for more than 100 years. Although they are
subject to chemical attack and grow brittle, they are still used in some
applications, particularly in the connection of vacuum sources to laboratory
apparatuses. The most commonly encountered synthetic tubing is Tygon®,
composed of PVC and bis (2-ethylhexyl) phthalate. It is soft, flexible, and
transparent, but becomes stiff when cold and is prone to leach the phthalate
plasticizer into samples. Tygon® tubing is serviceable for attachment of
coolant water lines to cooling towers and condensers. By soaking the end of
the tube in acetone for about a minute, the Tygon® will easily expand and
slip over the glass fitting. Allowing it to dry for 10 to 20 minutes, then
clamping it in place with a steel band screwed tubing clamp, affords a firm,
leak-free, positive attachment. Some care must be taken to avoid prolonged
soaking in acetone because it does extract the plasticizer from the Tygon®.
Tubing made of silicone polymer offers similar physical properties;
however, it is manufactured without plasticizers or additives, is odorless,
and offers good chemical resistance.
FIGURE 3.33 Teflon beaker with metal bottom. Photo copyright © Cole-
Parmer.

Other synthetic tubings that offer greater chemical resistance and fewer
sample contamination problems are constructed of polyurethane,
polyethylene, polypropylene, and fluorocarbon polymers. These tubings
offer a range of flexibility, chemical resistance, gas permeability, and
transparency.
FIGURE 3.34 Tubing connectors. Reprinted with permission from Bel-Art
—SP Scienceware.
 In general, the higher the chemical resistance, the less flexible the
tubing. Soft, flexible tubing can be fit together through use of glass or
plastic T and Y connections (Figure 3.34). For an apparatus that is
constantly being assembled and broken down again, there are a variety of
quick connections available constructed of either metal or plastic (Figure
3.35). These may be of a simple force-fit design, or they may have a
positive screwlock. Some also incorporate an on/off flow valve.
The stiffer tubings are attached together by screwed fittings with
ferrules or flanges for leak-free connections. The fittings can be of metal,
such as Swagelok® (Crawford Fitting Company) design, or they may be
made of one of a variety of fluorocarbon polymers (Figure 3.36).

FIGURE 3.35 Quick-disconnect tubing connector. Reprinted with

permission from Bel-Art—SP Scienceware.

Sample containers are a largely overlooked part of laboratory analysis.

Technicians commonly grab any available container and go get a sample.
However, more than 50% of all analytical problems actually arise during
the sampling process, and a good many relate to the container used for the
sample. The ideal container is inert to the sample, easily cleaned,
unbreakable, and inexpensive. Unfortunately, the ideal container does not
exist. The best general-purpose sample container is made of Teflon® and it
is also the most expensive. A good alternative is the fluorinated HDPE
container with a Teflon® liner for the cap. From there, the choices are
HDPE and borosilicate glass, with Teflon® lid liners. There are a large
variety of shapes and sizes of sample containers available, such that each
container is designed for a particular set of test parameters on a particular
matrix. For example, a plastic presterilized sample container for coliform
analysis of drinking water is available. It contains a sodium thiosulfate
tablet for dechlorination and a locking mechanism for ensuring sample
security (Figure 3.37). It is premarked for the appropriate size sample such
that the analyst merely has to open the container, add the appropriate test
media (Colilert™), then reclose the container and place it in the incubator.

FIGURE 3.36 Stainless steel adjustable hose clamp.

 Items used in sampling need to be inert to the sample. Monitoring well
bailers, shovels, dippers, spatulas, mixing spoons, trays, and bowls are more
and more frequently being made of solid fluorocarbon, HDPE, or
polypropylene construction. Sampling tools made of synthetics are easy to
clean and decontaminate and are also inert to most environmental
conditions and samples (Figure 3.38).

FIGURE 3.37 Presterilized sample containers with dechlorination tablets

for coliform testing. Reprinted with permission from Corning Incorporated.
FIGURE 3.38 Long-handle sampling spoons. Reprinted with permission
from Bel-Art—SP Scienceware.
FIGURE 3.39 Disposable filtration units. Reprinted with permission from
Thermo Fisher Scientific.

Plastic filtration units that are either presterilized and disposable (Figure
3.39) or autoclavable and reusable (Figure 3.40) are becoming more
common in laboratories and in the field. The disposable units are typically
made of polystyrene, whereas the reusable items are made of polysulfone.
They can be configured for either pressure or vacuum filtration. Designed
for filtration of aqueous samples, chemical-resistance charts should be
consulted before attempted filtration of organic liquids. Hand-operated
vacuum pumps made of PVC, die-cast zinc, and aluminum allow field
filtration without bulky extra equipment and power supplies (Figure 3.41).
FIGURE 3.40 Reusable filtration units. Used with permission of Sigma-
Aldrich Co. LLC.

5.0 METALWARE DESCRIPTION AND

CHARACTERISTICS
Metallic elements are frequently found in the laboratory in the form of
standards and reagents. They are also used as fabrication materials for many
items. The most common metals are iron (and steel), aluminum, copper
(and brass), titanium, and platinum. The most common items are mixing
and reaction containers, utensils, tubing, fittings, clamps, support rods, and
foils.
Although items of solid iron are rare, steels are used frequently,
particularly stainless steels. A variety of stainless steel formulations are
available (Table 3.3). Stainless steel 316 is the most acid-corrosion
resistant, and stainless steel 304 is the second choice for acid resistance.

FIGURE 3.41 Hand-operated vacuum pump. Reprinted with permission

from Lincoln Industrial.
TABLE 3.3 Chemical composition (percent) of elements other than iron in
stainless steel alloys.

Items constructed of 316 stainless steel are corrosion resistant; however,

they will still react with many chemical reagents, particularly concentrated
acids, found in the laboratory. Thus, the use of stainless steel is limited to
countertops (generally only in microbiology areas), tongs for transfer of hot
items from the oven and furnace, tweezers for manipulation of small items,
and spoons and scoops for portioning out solid reagents and samples for
weighing. A strongpoint of stainless steel is its inertness to organic solvents
and reagents. Stainless steel components are frequently found in
instruments, such as the swing pan on the analytical balance or the
weighing pan on the toploader. Membrane-filter holders, used with syringes
for filtering organic solvents, are made of stainless steel. Stainless steel
bowls, scoops, spoons, core samplers, and even shovels are found in sludge
and soil sampling operations. In most cases, the solids being sampled are
not highly corrosive and the ruggedness of the stainless steel, coupled with
corrosion resistance under normal conditions, makes stainless steel an ideal
construction material. Stainless steel dippers are highly recommended for
obtaining water and wastewater samples from outfalls and other hard-to-
reach areas.
While, in most cases, metals are not the preferred materials for
laboratory containers, there is one application where a metal serves an ideal
function as a reaction vessel. The metal is platinum and the application is
muffle furnace ashings and fusions. Platinum, one of the three coinage
metals (gold, silver, and platinum), is malleable (i.e., soft), chemically
resistant, has a high melting temperature (1773 °C), is a good electrical
conductor, and is very expensive. It is attacked by hot aqua regia
(concentrated nitric and hydrochloric acids), but is resistant to single
mineral acids. Platinum crucibles are expensive: a 25-mL crucible lists for
$600, and the cover adds on another $171. Platinum crucibles are used in
applications where a ceramic crucible is not suitable, for instance, the
muffle furnace fusion for silica determinations.
The tubing that is used to conduct compressed gases, such as hydrogen,
nitrogen, and oxygen, from the cylinder to the instrument using the gas is
most frequently made of copper. Copper tubing is, to a certain extent,
flexible, easy to work, and available in refrigeration grades (ASTM B68 or
B75), which means it is seamless and can conduct gases without leaking.
Copper tubing is not subject to corrosion in most applications where the
laboratory is humidity controlled and the ventilation system prevents a
buildup of acid or organic solvent vapors. As it is supplied from the
manufacturer, copper tubing is contaminated with oil and grease on the
inside, and this must be removed before installation. A suitable procedure is
to rinse the inside of the tubing with toluene to remove the oils. After the
toluene has evaporated, the copper tubing will not contribute contamination
to the gas flowing through it as other tubings such as Tygon® will. Further,
copper tubing, unlike plastic tubings, does not become brittle with age and
crack.
The other metal tubings found in the laboratory are constructed of
stainless steel and titanium. These are much more difficult to work and less
flexible than copper, but offer improved resistance to corrosion. They are
used when the tubing conducts liquids, such as chromatography solvent
mixtures. The only practical way to attach metal tubing to regulators on gas
cylinders and the other end of the tubing to instruments so that the
connection is leak-free is with compression-design tube fittings (Figure
3.42). Both single ferrule and two-piece ferrule compression fittings are
available.
The fitting forms a leak-free connection when the back ferrule
compresses the front ferrule into the body and clamps down onto the tubing
walls. The compression and sealing action takes only one and a quarter turn
past the finger-tight point. Tightening the fitting until no further movement
between the nut and body is possible only crushes the tubing in the fitting.
The fittings are engineered so that a gauge can be used on the gap between
the body and nut to check for a proper seal. Once the seal is made, the nut
and ferrules are permanently attached to the tubing end. The connection can
be disassembled and reassembled without difficulty or leaks.
The fittings will provide a leak-free fitting for a considerable length of
time. However, in situations where the fittings are subjected to changes in
temperature or vibration, they can work themselves loose. It is prudent to
check each fitting with a leak detector at least every 6 months. Liquid soap
solutions (available as SNOOP®), methanol, and electronic gas detectors
are commonly used to check for leaks. The electronic detectors are
somewhat expensive; however, they do not leave a residue that the liquids
do. Also, if a leak exists, it is not a one-way situation. Gas is exiting the
fitting and liquid can enter it, resulting in contamination to the inside of the
tubing.
FIGURE 3.42 Single ferrule compression fitting. Reprinted with
permission from Parker Hannifin.

Because of the way the fitting forms a seal with the tubing, it is
necessary that the fitting be made of a harder metal than the tubing. Brass
fittings are suitable for connecting copper tubing; however, stainless steel
fittings are needed for stainless steel or titanium tubing.
Although the Swagelok® compression-design pipe fitting is the most
commonly used, there are no industrywide standards. This means that
pieces of fittings cannot be interchanged between manufacturers.
Grease or other liquid/solid compounds, such as graphite lubricants,
should never be used on metal fittings that are part of gas systems.
Invariably, the lubricant will get inside the tubing and contaminate the gas
flowing through the pipes. In some cases, this can contribute to a fire
hazard, particularly if oxygen is the gas. Further, these types of lubricants
can cause gas cylinder regulators and flow controllers to malfunction. If a
lubricant is needed, Teflon® tape can be used. A single layer is wrapped
around only the threads of the fitting; tape on the mating surfaces of the
fitting will cause a leak. When the fitting is taken apart, all traces of the
residual tape must be removed from both parts of the fitting and a new layer
applied.
Pipe fittings that are found to leak after the manufacturer’s tightening
directions have been followed can never be made satisfactorily leak-free by
continued tightening of the joint. Brute strength will only result in stripping
the threads of the fittings and permanently damaging the joint mating
surfaces. This situation is caused by either dirt or other debris on the mating
surfaces of the joint or by a mismatch of parts from two different
manufacturers. The solution is to take the fitting apart, inspect it for debris,
and verify that the male and female parts are from the same company.
Support rods and clamps are used to hold the laboratory apparatus
upright and in place. The support may be a single upright rod attached to a
heavy metal base, called a ring stand, or it may be part of a larger collection
of horizontal and vertical rods called a lattice. The rods are frequently
constructed of aluminum; however, some older rods are steel. In some
situations, such as inside fume hoods used for acid digestions, the rods can
become quite corroded. Solid rods of synthetic materials may be a solution
for these applications. Solid aluminum stock of appropriate diameter (13
mm, or 0.5 in.) can be purchased and the lattices assembled with lattice
connectors for custom installations. Hollow 1.2-m × 1.2-m (4-ft × 3 4-ft)
aluminum posts are used to form the lattice frame and for attachment to
laboratory benches.
There are a wide variety of clamps used to attach the apparatus to the
support lattice (Figure 3.43). The ring is used to support separatory funnels.
Two-prong and three-finger pinch clamps constructed of either corrosion-
resistant metal or reinforced nylon are available in a variety of sizes.
Double buret pinch clamps are widely used in titration stations. Chain
clamps are useful for large items such as beakers and condensers.
FIGURE 3.43 Variety of support clamps. (Top) Fixed position, medium 3
prong dual adjust clamp. (Bottom) Fixed position, medium 2 prong single
adjust clamp. Reprinted with permission from Troemner LLC.
 Many clamps come with a clamp holder as an integral part. Other
clamps just have a rod attached to the jaws, and a clamp holder must be
used to attach the clamp to the lattice. A wide variety of clamp holders are
available. These range from the standard right-angle double-screw clamp
and the hooked single-screw holder, to the multiple-screw swivel and
contort holders. Lattice connectors, which require a screwdriver, box
wrench, or Allen wrench for adjustment, can also be used as clamp holders,
particularly when the clamp is part of a permanent setup.
Clamps and clamp holders require maintenance. They should be kept
clean and free of chemical residues to prevent corrosion. The screw
adjustment requires lubrication with either a light oil or silicone grease—
not a large amount, just enough so that the screw threads do not freeze.
Grease is the better solution because it does not evaporate and protects
against corrosion. Lubrication also makes it easier to adjust the position of
the clamp when you are holding an expensive glass apparatus in one hand
and trying to attach it to the lattice with the other.
Another concern is how much to tighten the clamp. The answer is just
enough to keep the apparatus in position. Pliers should not be used to
tighten clamps and clamp holders because the screw may be overtightened
to the point where it can be wrung apart or the clamp jaw can bend and
break. Again, lubricating the screw threads will avoid this problem.
 4
Volumetric Devices and Their Use in
Measurement

1.0 VOLUMETRIC MEASURING TOOLS

2.0 PROPER USE OF VOLUMETRIC GLASSWARE
3.0 CALIBRATION OF NONSTANDARD MEASURING TOOLS
4.0 REFERENCES

Most quantitative assays performed in water and wastewater laboratories

depend on precise measurement of volumes of solutions and samples for
accurate and repeatable results. The quoted precision and accuracy for a test
assumes that an analyst is experienced and is making use of the best
analytical tools available. Use of an inappropriate measuring device will
introduce additional sources of error into the procedure and directly
translates to a totally misleading analysis. This chapter will present the
various types and standards of calibration of volumetric measuring tools, the
proper use of volumetric glassware, and how to perform and document the
calibration of nonstandard measuring devices.

1.0 VOLUMETRIC MEASURING TOOLS

Erlenmeyer flasks and beakers, the “pots and pans” of laboratory glassware,
are not volumetric measuring tools. Indeed, although in many cases they
have markings on the side of the vessel, these are only approximate volume
indicators. Even in the most expensive Erlenmeyer flasks and beakers, there
is a ±5% uncertainty in where the line actually needs to be. This is the key to
understanding the use of calibrated volumetric measuring devices. That is, it
is not how careful the technician is in filling the device to the volume line,
rather, it is what is the tolerance with which the device is marked at the
correct volume.
There are only five volumetric measuring devices recognized as suitable
for accurate and precise analytical work: burettes, volumetric flasks,
volumetric (transfer) pipettes, measuring (Mohr) pipettes, and graduated
cylinders. Produced and calibrated in accordance with American Society for
Testing and Materials (ASTM) standards, they are available in three grades:
Class A (Figure 4.1) and Class B that meet the ASTM standards and student
or economy grade that has no standard. Applicable standards are ASTM
E694-99 (2010) Standard Specification for Laboratory Glass Volumetric
Apparatus, and ASTM E542-01 (2012) Standard Practice for Calibration of
Laboratory Volumetric Apparatus. Specific standards are ASTM E287-02
(2012) Standard Specification for Laboratory Glass Graduated Burets,
ASTM E288-10 (2007) Standard Specification for Laboratory Glass
Volumetric Flasks, ASTM E969-02 (2012) Standard Specification for Glass
Volumetric (Transfer) Pipets, ASTM E1272-02 (2012) Standard
Specification for Laboratory Glass Graduated Cylinders, and ASTM 1293-
02 (2012) Standard Specification for Glass Measuring Pipets. The ASTM
standards define the construction and design of the tools, but, more
importantly, define the tolerances within which the markings are placed on
the device, with Class A glassware having the smallest tolerances. The Class
B tolerances are in general twice the acceptance range of the Class A
specifications (see Tables 4.1 to 4.4). All plastic (Teflon®, Nalgene®,
polypropylene, polymethylpentane, polyethylene, etc.) volumetric devices
are, at best, Class B, although no standard exists for their calibration. Class
A volumetric glassware will always have a large “A” prominent near the
label on the piece. Glassware without an obvious “A” is of Class B or lesser
quality.
FIGURE 4.1 Class A volumetric flasks. Reprinted with permission from
Thermo Fisher Scientific.

Some glassware manufacturers have been offering two types of Class A

volumetric ware. The first is a Class A line that meets the ASTM standards.
The second type is a certified and serialized Class A line that comes
complete with documentation and traceability to the National Institute of
Standards and Technology, which is required for certification to International
Standards Organization (ISO) standards, such as ISO 9000. The substantial
added expense of the certified/serialized glassware is only justified if the
laboratory performs a lot of work for overseas and international clients.
However, the average water or wastewater laboratory has only the local
municipality or county as its main client and thus gains no benefit by
purchasing this specialized line other than traceability. The tolerances on the
placement of the volume markings are the same for the registered and
unregistered Class A lines, and, therefore, the extra expenditure is not
warranted.

TABLE 4.1 Comparison of Class A and Class B tolerances for volumetric

flasks.
TABLE 4.2 Comparison of Class A and Class B tolerances for pipettes.

TABLE 4.3 Comparison of Class A and Class B tolerances for graduated

cylinders.
TABLE 4.4 Comparison of Class A and Class B tolerances for burettes.

Volumetric flasks are supplied in a variety of shapes and sizes, in

addition to the different calibrated classes. Some have wide bases formed on
the bottom of the flask, while others have the bottom flattened so they can
stand unsupported. The most common shape for the bottom of the flask is
round, although a square shape is also available. In general, the flasks with
formed bottoms are more prone to breakage, but less likely to be tipped over.
The 1- and 2-mL sizes are available as calibrated tubes without a base,
similar in appearance to test tubes.
Closures for the flasks include plastic caps that snap over the top of the
flask, screw-threaded tops for use with screw caps and standard-taper
ground-glass joints for stoppers of polyethylene, Teflon®, or glass. Some
flasks have no closure and are designed to be used with corks or rubber
stoppers. Corks are rapidly becoming an archaic form of closure because of
leaks and cross-contamination. Stoppers of rubber or one of the other
common plastic materials (Teflon®) can be used, but are not preferred
because of becoming loose at refrigerator temperatures. This can be used to
advantage when a plastic stopper becomes stuck in a ground-glass joint.
Simply placing the item in the freezer will shrink the stopper enough to
allow easy extraction. The plastic caps are designed to replace the corks, but
they leak. The screw caps and ground-glass joint closures are optimal for
effective sealing of the flask. Glass stoppers give a better seal at different
temperatures; however, they are prone to freezing (sticking) in the flask
because of either chemical reaction between the material in the flask and the
glass surfaces of the joint and the stopper or simple friction. Teflon® stopper
sleeves are available that create a chemically inert surface between the glass
stopper and the ground-glass joint, allowing a tight seal and no tendency to
become stuck. Screw-top caps can be overtightened, resulting in breaking of
either the cap or the glass threads. The cap can also break when the flask is
placed in the freezer. This is not to suggest that volumetric flasks are
designed to be storage containers; they are not. Volumetric flasks are
designed to be used to prepare known dilutions of samples, reagents, and
standards, then the solution should be transferred into a separate plastic or
glass storage bottle.
All volumetric flasks are calibrated “to contain” (“TC”). This means that
when the flask is filled with liquid to the mark, it contains the specified
amount of liquid, such as 100.00 mL. To prepare a solution in a volumetric
flask, a known amount of the solute is placed in the flask, some solvent is
added, and the flask is agitated until the solute is dissolved in the solvent.
After the solution returns to room temperature (some dissolutions generate
heat whereas others reduce the temperature of the solution), more solvent is
added until the flask holds 1 to 3% less than the final volume. The solution is
mixed then allowed to settle. The volume is finally adjusted to the mark by
careful addition of more solvent until the bottom edge of the meniscus sits
just on top of the calibration mark. Final mixing of the contents of the flask
is performed by stoppering the flask, then inverting it and shaking. This is
repeated two or three times. Solids that dissolve slowly can be mixed with
some solvent in an Erlenmeyer flask using a magnetic stir bar to agitate the
mixture. The solution can then be quantitatively transferred to a volumetric
flask for final volume adjustment.
Measuring pipettes can be differentiated from volumetric pipettes by the
presence of a graduated volume scale on the side of the measuring pipette
(Figure 4.2). There are two types of graduated pipettes commonly found in
the laboratory. The measuring (Mohr) pipette is graduated from a zero mark
near the top of the pipette to a baseline near the tip of the pipette. The pipette
is labeled with the total volume and the volume divisions; for example, “5
mL in 0.1” means the total capacity is 5.00 mL and there are volume
markings every 0.1 mL. Mohr pipettes are intended to indicate the delivered
volume by a difference technique between the initial and final liquid
positions, with delivery of the maximum calibrated volume leaving the tip
full of liquid. Thus, the Mohr pipette is a “to deliver” (“TD”) device.
Measuring pipettes are calibrated with either Class A or Class B pipette
tolerances; however, each volume marking on the Mohr pipette has the same
tolerance as the maximum volume marking. Dispensing 1.00 mL of liquid
from a 10-mL Class A measuring pipette has the tolerance of the 10-mL
pipette (±0.02 mL), not the tolerance of a 1.00-mL pipette (±0.006 mL).
Mohr pipettes are designed for repetitive delivery of reagent volumes when
the reagents are not the determinative reagent in an analysis.
A serological pipette is graduated from a zero mark near the top of the
pipette to the tip of the pipette. Serological pipettes are labeled with the total
volume and the volume divisions similar to the Mohr pipettes (e.g., “1 mL in
0.01”). Serological pipettes are marked to indicate the difference between
the initial and final liquid levels; to deliver the entire marked amount, the
pipette is blown out so that no liquid remains in the tip. The best available
serological pipettes are calibrated to only Class B tolerances, and all the
volume markings on the pipette have the same tolerance. Dispensing 1.00
mL from a high-quality 10-mL glass serological pipette is done with a
±0.04-mL uncertainty. A special type of serological pipette is available with
a wide opening at the tip for measuring viscous liquids and samples with
particulate content that would clog a narrow-tip pipette. The wide-tip
pipettes are useful for dispensing samples for biochemical oxygen demand
(BOD) testing and other procedures.
FIGURE 4.2 (a) Class A volumetric pipet, (b) Class A measuring pipet
(Mohr), (c) a serological pipet. Courtesy of Corning Incorporated.

Volumetric (transfer) pipettes have a bubble in the shaft and only a single
volume mark. Some volumetric pipettes are calibrated to contain (TC), but
they are rarely encountered. The most common volumetric pipettes are
calibrated to deliver (TD). There is a special volumetric pipette, termed a
dual-purpose pipette, that has calibration marks with Class A tolerances for
both to deliver and to contain use. Volumetric pipettes are designed for
delivery of a single set volume of aqueous solution with maximum
repeatable accuracy. Volumetric pipettes are available in all unit volumes
from 1.00 to 10.00 mL and in larger volumes of 15.00, 20.00, and 200.00
mL, in addition to those listed in Table 4.2.
A final type of pipette is the bacteriological pipette, which is designed
for use by the dairy industry. These are available in 1.1-, 2.2-, 10-, and 11-
mL sizes in both narrow and wide-tip forms. The volume tolerances on these
pipettes are wider than the Class B standards for other pipettes. They are
intended to make serial dilutions into standard glassware easier. These may
find possible use in the microbiology section of the water and wastewater
laboratory, but they are not suitable for analytical chemistry procedures.
Graduated cylinders are tubes of glass (or plastic) that have volume
markings on the side (Figure 4.3). Glass graduated cylinders are available
calibrated to deliver or to contain to the Class A or Class B ASTM
standards, and are most commonly used for dispensing large portions (.50
mL) of sample for testing purposes. They also find use in determining, after
the fact, how much sample was analyzed. This is accomplished by marking
the original liquid level of the sample in the sample container with a marker,
dispensing all the sample for analysis, then refilling the container to the
mark with water. The water is transferred to a graduated cylinder for volume
measurement. Graduated cylinders are not designed for accurate preparation
of solutions or dilutions, but are invaluable for preparing solvent mixtures.
For example, if a 30% water-isopropanol mixture is needed, a graduated
cylinder is the most appropriate volumetric measuring tool for the
preparation. Graduated cylinders should never be used as a mixing container
for reagents that generate heat (i.e., exothermic dissolution or reaction) in the
mixing process because thermal stresses can easily crack the cylinder.
 Burettes (Figure 4.4) are indispensable tools in titrations and other
laboratory uses such as repetitive dispensing of reagents. A burette is a
measuring pipette with a valve (stopcock) at the bottom and allows
dispensing of tightly controlled volumes of reagent with known accuracy.
Graduated with the zero mark at the top, they indicate dispensed volume (to
deliver). Burettes are available in Class A or Class B tolerances (Table 4.4).
The variety of burette sizes allows a matching between the anticipated
volume for the titration and the total capacity of the burette. The most
accurate use of the burette is obtained when the dispensed volume is greater
than 80% of the total volume capacity of the burette. Burettes are available
with two-position stopcocks that allow filling the burette from a gravity-
powered reservoir of reagent with a simple twist of the wrist. This allows
fast titrations in situations where the analyst has many titrations to perform
in a limited amount of time. The two-position stopcocks are particularly
necessary when using 5-mL or smaller capacity burettes. They are almost
impossible to fill from the top by pouring a reagent solution into the burette
because of surface tension of the liquid around the air bubbles and the
narrow diameter of the tube. Some burettes come with a flared top for easy
filling without spilling. Small funnels are also available that fit into the top
of the burette for easy filling.
FIGURE 4.3 Class A graduated cylinder. Courtesy of Corning
Incorporated.
FIGURE 4.4 Class AS microburet with attached reservoir. Courtesy of
BRAND.

With a selection to be made between several different volumetric

measuring tools, the relative accuracy needed in the measurement will
govern the choice of tool. The relative accuracy (because of allowed
tolerances) for 25-mL devices is presented in Table 4.5. Some error
consequences from the decision to dispense 1.00 mL from a variety of Class
A measuring tools is illustrated in Table 4.6. Remember that this considers
only the allowed tolerances for the placement of the marks on the device;
there is added error associated with how consistently the device is used in
terms of technique and this error increases with the total volume of the
measuring device. The point is that it is essential to select the proper
measuring tool of the proper size to obtain the best results.

TABLE 4.5 Comparison of relative percent accuracy of the calibration

marks on 25-mL measuring devices for Class A and Class B tolerances.

2.0 PROPER USE OF VOLUMETRIC

GLASSWARE
The first consideration in the use of volumetric glassware is the temperature.
All glass expands and contracts with the rise and fall of the temperature.
Different types of glass have different rates of expansion, so there is no
universally applicable correction factor for temperature variation. Instead,
the temperature at the time of calibration is typically etched on the device. If
none is listed, then 20 °C is assumed because it is the standard.

TABLE 4.6 Comparison of percent accuracy for determination of

dispensing 1.00 mL from a variety of Class A measuring devices.

Measuring device Percent accuracy

1-mL volumetric pipet ±0.6
1-mL volumetric flask ±1.0
5-mL measuring pipet ±1.0
5-mL microburet ±1.0
5-mL serological pipet ±2.0
10-mL graduated cylinder ±8.0
10-mL measuring pipet ±2.0
10-mL buret ±2.0
25-mL measuring pipet ±3.0
25-mL buret ±3.0

Further, the solutions themselves change volume with temperature, with

most expanding as the temperature rises. Water is unusual in its behavior,
shrinking in volume from 0 to 4 °C, then expanding as the temperature is
increased. This volume variation affects the concentration of reagents in the
solution. As the solution shrinks, the concentration of the dissolved reagents
increases. Reagent and calibration standards should always be dispensed at
the same temperature at which they are prepared. Preparation of some
reagents can cause a temperature change in the solution: dissolution of
calcium chloride, as well as dilution of most acids and bases, result in
heating the solution, whereas dissolving ammonium nitrate or ammonium
chloride in water causes the solution to cool. Final volume adjustment
should be delayed until the solution is at room temperature.
“Dilute to the mark” and “read the meniscus” are common instructions
for preparing solutions or performing titrations. There are right ways and
wrong ways to do this. Most calibration marks on volumetric glassware
extend all the way around the device. This is to aid in orienting the eye to the
exact plane of the mark. No trace of a circle or ellipse should be seen when
the eye is in the correct position relative to the mark, rather, only a single
straight line.
The width of a calibration mark is immaterial and not mentioned in the
calibration specifications. Instead, the upper edge of the calibration mark is
the point of calibration. Thus, the correct position for the meniscus is with
the upper edge of the calibration mark just lining up with the bottom edge of
the meniscus. Figure 4.5 illustrates the correct positioning of the meniscus
and calibration mark. Figure 4.6 illustrates the finger and thumb techniques
for holding the pipette.
Reading graduated cylinders, measuring pipettes, and burettes frequently
requires estimating the position of the meniscus between calibration marks.
The space between the calibration marks is imagined to have 10
subdivisions, then the position of the bottom edge of the meniscus is
estimated (Figure 4.7).
If the container is a “to contain” device, the entire contents are the
correct volume. When transferring the contents to another container, simply
pouring the solution from the to contain device to the new container
maintains the concentration of the solution; however, all the liquid will not
be transferred. This is acceptable if the concentration of the reagent or
standard is the important point. If all the solution must be transferred, then a
quantitative transfer must be made. This entails blowing the pipette out with
a pipette bulb (Figure 4.8) and rinsing the inside of the pipette with
additional solvent from a squeeze bottle or rinsing the flask or cylinder with
additional solvent. For a “to deliver” device, a flow time is specified (Table
4.7). The tip internal diameter is designed so that the desired flow time is
established as the pipette empties by gravity. Pushing the liquid out with a
pipette bulb defeats the design. A pipette with a chipped or dirty tip will not
deliver the liquid at the proper rate. Pipette tips that are dirty can be cleaned
by suspending the pipette in a beaker of soapy water in an ultrasonic cleaner.
Copious quantities of reagent water should be used to rinse any traces of
soap residue from inside the tip. Particles lodged in a pipette tip can be
removed by attaching a piece of tubing to the tip and flushing the blockage
out with water from a faucet. A wire or pipe cleaner is not a preferred
method because the tip can be cracked or chipped. The same considerations
apply to burette tips.

FIGURE 4.5 Correct position of calibration mark and meniscus (left).

Meniscus positioned higher than calibration mark (right). Courtesy of
Apichemical Consultants.
FIGURE 4.6 Finger (left) and thumb (right) pipetting techniques. Courtesy
of Apichemical Consultants.

FIGURE 4.7 Position of meniscus in a graduated cylinder estimated as

52.8 mL. Courtesy of Apichemical Consultants.

The volumetric (or transfer) pipette is a carefully designed piece of

glassware. First and foremost, it operates by gravity. Any disruption of the
effect of gravity on the pipette, such as the use of a pipette pump, is going to
result in either too much or too little liquid being dispensed. Using your
mouth to draw liquid into the pipette is a good way to get poisoned and very
poor laboratory technique. The proper tool for filling the pipette is a pipette
bulb (Figure 4.8).

FIGURE 4.8 Pipette bulb. Courtesy Apichemical Consultants.

TABLE 4.7 Flow times for TD transfer pipettes.

Pipet volume (mL) Class A flow time (sec) Class B flow time (sec)
1 8–60 3–60
5 8–60 8–60
10 15–70 8–70
25 25–70 15–70
50 25–70 15–70
100 30–70 20–70

The bulb is placed over the wide end of the pipette, and the bulb is
squeezed. Then, the tip of the pipette is placed under the surface of the liquid
and the blub is relaxed to gently pull liquid into the pipette until the liquid
level is above the calibration mark on the pipette. The bulb is taken off the
pipette and either a finger or thumb placed over the opening (Figure 4.6).
Touching the tip of the pipette to the bottom of the container to restrict flow
while doing this makes the process easier. The pipette tip is then taken out of
the liquid and the meniscus carefully lowered to the calibration mark (Figure
4.5). The pipette is now ready for the transfer.
Proper dispensing of solution from a to deliver transfer pipette requires
touching the tip to the inside wall of the receiving container, then allowing
the pipette to drain by gravity, then maintaining contact between the tip and
container for two seconds after flow ceases. Pipettes have a built-in flow
time and disrupting the flow time by pushing liquid out of the pipette defeats
the purpose and design of the pipette. There will be liquid left in the tip of
the pipette. This must not be blown out as then excess liquid will have been
delivered. For solutions that have significantly different density or viscosity
than pure water, the flow times and retained liquid in the tip are not correct.
In these cases, the analyst may want to consider using a to contain pipette or
a measuring pipette because the error associated with the excess retained
solution in the tip of the transfer pipette generally exceeds the manufacturing
tolerance of the to contain pipette.
 There is also the option of basing the delivery method on the weight of
the liquid delivered. Densities are easy to measure, and volumes easily
converted into mass (D = mass/volume). Dispensing the correct mass is easy
with a balance, and then washing the material into the reaction vessel with
either water or an appropriate solvent is quantitative.
The following is an excellent exercise for developing skill with
volumetric pipettes. Take a Class A 10-mL volumetric pipette and use it to
deliver 10.00 mL to a 100-mL Class A volumetric flask 10 times. The
resulting meniscus should be exactly at the 100-mL calibration mark of the
flask.
Volumetric flasks are not storage containers. They should be used to
prepare the solution, then the solution transferred to a plastic or glass storage
bottle. Volumetric flasks are expensive; storage bottles are considerably less
expensive.
For analysts who are trying to achieve the highest level of reproducibility
in their work, how they use volumetric ware will have an effect on the
results. Suppose a calibration is being prepared and eight standards must be
made. If eight different volumetric flasks and eight different pipettes are
used, then the full range of the flask and pipette volume marking tolerances
will be transferred to the calibration. If instead a single flask and pipette are
used, then the random error associated with the tolerances will be changed to
a systemic error, and as long as the analyst continues to use the same pipette
and flask in sample preparation, the error is self-correcting and vanishes. For
example, if a 1.00-mL volumetric pipette is required in the method for
adding the determinative reagent, and the pipette actually dispenses 1.05
mL, but is the only pipette used, then the calibration corrects for this
systemic error of 0.05 mL.

3.0 CALIBRATION OF NONSTANDARD

MEASURING TOOLS
There are numerous tools used in the laboratory for fluid measuring and
transfer that one would like to calibrate to the maximum extent possible.
Examples include BOD bottles, colorimetric tubes, syringes, diluters, bottle-
top dispensers, and disposable-tip adjustable- and fixed-volume pipetters.
The placement of an additional 201.0-mL or 99.00-mL mark on a volumetric
flask for accurate determination of the proper volume of sample for titration
in the Winkler dissolved oxygen determination is another example. These
can all be calibrated to better than Class A tolerances. All one needs is an
analytical balance, reagent water, a standardized thermometer, a calibration
notebook, and a table of water densities (or specific gravity) such as that
found in the Handbook of Chemistry and Physics or in the Handbook of
Environmental Analysis. This is presented in Table 4.8.
First, let everything equilibrate to room temperature and place the
thermometer into the container of reagent water. Dispense a volume of water
from the device to be calibrated onto the analytical balance (use a tared
container). Record the weight of the water in grams. Repeat this at least two
or three times, then calculate the average result. Record the temperature.
Look up the density of water in the table and divide the average weight of
dispensed water by the density. The result is the actual volume dispensed in
milliliters, with a tolerance tighter than the Class A standards. Tag the device
and keep a record of the calibration in the notebook.

TABLE 4.8 Density variation of reagent water with temperature. Reprinted

with permission from R.-K. Smith Handbook of Environmental Analysis, 4th
Ed., Genium Publishing Schenectady, New York, 1999.
 Calibration of bottles and other containers is performed by calculating
the weight of water that corresponds to the desired volume at the ambient
room temperature. Place the device on a balance and add reagent water until
the required weight is obtained. Use a permanent laboratory marker to
indicate the proper meniscus position, then place an etched line using one of
the following techniques.
An old fashioned approach to placing a permanent mark on the glass
starts with using a laboratory marker to mark the correct place. A layer of
melted paraffin wax is brushed over the mark and allowed to cool. Using a
razor blade, remove a thin groove of wax over the mark down to the glass.
Fill the groove in the wax with concentrated hydrofluoric acid and wait an
hour or so. Hydrofluoric acid has a low surface tension when in contact with
glass and will flow freely over the glass surface unless confined by the wax.
The acid will etch a permanent mark in the glass at the desired point. Rinse
the excess acid off the glass, then clean the wax from the glass with
methylene chloride or hexane.
A more modern approach to achieve a high-quality permanent mark on
the glass uses a sand blaster, which is available at many rental companies or
plate glass companies. After the desired volume is marked on the glass, the
mark is covered with a piece of heavy, transparent tape. Then, the tape is cut
away from the mark with a razor blade. The sand blasting only takes a few
seconds to make a permanent etch on the glass. A number of pieces can be
prepared for sand blasting and the etch marks made in a very short period of
time. Tape is also easier to work with than hot wax.
Another marking technique is to use a file or diamond or silicon carbide
marking pen to make a scratch at the correct position. Be aware that the
scratch creates a weak spot in the glass and makes the device more prone to
fracture at that point. Etched marks, either chemically or mechanically, tend
to be more resistant to fracture. Permanent ink markings can be made as a
last resort for placing volume markings on the glass; however, they rub off
after a short period of time and can be washed off entirely with certain
solvents.
Glass syringes require calibration verification only once as the contents
and gaskets are visible. When the plunger is discovered to leak, it is time to
discard the syringe and calibrate a new one. Analysts should be aware that
the heat from hands will change the volume of the syringe and the density of
the contents. Therefore, it is important to handle the syringe by the raised
ring at the end of the syringe.
For devices that require preventive maintenance, such as diluters,
dispensers, and disposable-tip pipettes, the calibration verification should be
repeated at regular intervals (e.g., weekly). This serves as a check on the
seals and lubricant inside the device. For example, the metals preparation
laboratory at Analytical Services, Inc., checks the calibration of disposable-
tip pipettes weekly and logs the relative variation of the mean of three
determinations from the labeled volume and the relative standard deviation.
To test how often the checks need to be made, the following experiment was
performed on a 1.00-mL adjustable volume disposable-tip pipette. The first
week’s calibration verification gave 1.0141 mL (+1.59% variation and
0.15% relative standard deviation for three determinations). The second
week after no maintenance gave 1.0170-mL (+1.86%) and 0.65% relative
standard deviation. After 3 weeks, the results were 1.0454-mL (+4.8%
deviation) and 5.1% relative standard deviation, indicating that the seals or
lubricant inside the pipette body were failing. After cleaning, repair, and
relubrication, the calibration verification gave 1.0066-mL (+0.9%) and
0.28% relative standard deviation. Other than calibration verification checks,
there was no indication that the device was failing. For adjustable volume
pipettes, at least two volumes, one high and one low, should be checked. The
overall experience is that the disposable-tip pipettes fail irregularly, but
require significant cleaning or replacement after 1 to 3 months.
Needless to say, ASTM Class A volumetric and measuring pipettes,
volumetric flasks, graduated cylinders, and burettes are the preferred
measuring tools for solutions in the laboratory. Other devices can be
calibrated using the aforementioned procedures to expand your ability to
measure solution volumes reliably and improve the repeatability of your
laboratory work.

4.0 REFERENCES
American Society for Testing and Materials (2012a) Standard Practice for
Calibration of Laboratory Volumetric Apparatus; ASTM E542-01;
American Society for Testing and Materials: West Conshohohocken,
Pennslvania.
American Society for Testing and Materials (2012b) Standard Specification
for Laboratory Glass Graduates Burets; ASTM E287-02; American
Society for Testing and Materials: West Conshohohocken, Pennslvania.
American Society for Testing and Materials (2012c) Standard Specification
for Laboratory Glass Graduated Cylinders; ASTM E1272-02; American
Society for Testing and Materials: West Conshohohocken, Pennslvania.
American Society for Testing and Materials (2012d) Standard Specification
for Glass Measuring Pipets; ASTM 1293-02; American Society for
Testing and Materials: West Conshohohocken, Pennsylvania.
American Society for Testing and Materials (2012e) Standard Specification
for Glass Volumetric (Transfer) Pipets; ASTM E969-02; American
Society for Testing and Materials: West Conshohohocken, Pennslvania.
American Society for Testing and Materials (2010) Standard Specification
for Laboratory Glass Volumetric Apparatus; ASTM E694-99; American
Society for Testing and Materials: West Conshohohocken, Pennsylvania.
American Society for Testing and Materials (2007) Standard Specification
for Laboratory Glass Volumetric Flasks; ASTM E288-10; American
Society for Testing and Materials: West Conshohohocken, Pennslvania.
 5
Temperature Measurement

1.0 INTRODUCTION
2.0 DESCRIPTION OF TEMPERATURE MEASURING
DEVICES
3.0 CALIBRATION CHECKING AND MAINTENANCE OF
LIQUID-IN-GLASS THERMOMETERS

1.0 INTRODUCTION
Temperature is a measure of the motion of molecules. Molecules that have a
lot of motion are at a higher temperature than molecules that are moving
more leisurely. Heat is the transfer of temperature from one item to another.
There are places where high temperatures exist, but there is little heat; the
fringes of earth’s atmosphere represent one of these places. The molecules
are zipping around with high temperatures, but there are not enough of them
to effect any large degree of temperature transfer from one item to another.
Temperature is most commonly determined by quantitating a heat
transfer from the object of interest to some measuring instrument. Three
temperature scales are in use in the United States: Fahrenheit, Celsius, and
Kelvin. The Kelvin scale is the official scientific temperature scale. Its zero
is the absolute zero, where all atomic and molecular motion ceases. A single
unit of temperature is called a kelvin (K). The Celsius (centigrade)
temperature scale is based on the degree Celsius (°C), with 1 °C representing
the exact same increase in temperature as 1 kelvin. The zero on the
centigrade scale (actually +0.01°) is the ice point (triple point) of water, and
100 °C is established as the boiling point of water at 1 atmosphere (760 mm
of mercury [mm Hg]) of pressure). Absolute zero on the Celsius scale is
−273.15 °C. The Fahrenheit scale is based on the degree Fahrenheit (°F),
with the ice point of water at 32 °F and the boiling point at 212 °F.
Conversions between the scales are easily accomplished as follows:

There are no common points between the Kelvin scale and the Celsius
scale. However, −40.0° is the same temperature on the Fahrenheit and
Celsius scales, and 574.6 is the same on the Kelvin and Fahrenheit scales.
The internationally recognized fixed-temperature reference standards are
the boiling point of oxygen at 1 atmosphere of pressure (−182.962 °C), the
triple point of water (+0.01 °C), the boiling point of water (100.00 °C), and
the freezing point of zinc (419.58 °C). The official temperature measuring
standard is the platinum resistance thermometer. A comparison of the
temperature scales is illustrated in Table 5.1.
The effects of increased temperature are typically an increase in volume
of a substance because of the increased motion of the constituent particles,
an increase in chemical reaction rates, an increase in the flowrate for solids
and liquids, and possible phase changes from solid to liquid to gas. Water is
an odd material that actually shrinks in volume in going from a solid at 0 °C
to a liquid at +4 °C. From that point on, water expands in volume. This is
why ice floats on liquid water.

2.0 DESCRIPTION OF TEMPERATURE

MEASURING DEVICES

TABLE 5.1 Comparison of Kelvin, Celsius, and Fahrenheit temperature

scales.
Temperature can be variously measured with a liquid-in-glass thermometer,
an electrical thermocouple, an electrical resistance (platinum or solid-state
thermistor) thermometer, a mechanical bimetal strip, an infrared-sensitive
meter, temperature-sensitive liquid crystal strips, thermochromic crayons
and paints, or the decorative Galileo buoyancy thermometer. For precise
measurement, the liquid-in-glass thermometer is the most common device.
The main parts of the liquid-in-glass thermometer are the bulb and the
stem (Figure 5.1). The stem may include an auxiliary scale for calibration of
the ice point, a contraction chamber, an immersion line, the main scale, and
an expansion chamber.
The bulb is the reservoir of the expanding liquid, whereas the stem is
made from a thick-wall capillary tube of strictly controlled internal diameter.
All liquid-in-glass thermometers rely on expansion of the liquid column
with increased temperature. The liquids encountered in glass thermometers
are mercury, mercury-thallium alloy, and a variety of organic liquids.
Mercury thermometers are regarded as the most precise and have been
constructed for measurements over the temperature range from above the
freezing point of mercury (−38.9 °C) to well over 300 °C, the upper limit
being set by the properties of the glass. The mercury-thallium alloy extends
the lower end of the operating range to −56 °C. For even lower temperatures,
an organic liquid such as pentane, toluene, or alcohol, alone or in mixtures
with other compounds such as chloroform, methylene chloride, ethyl
bromide, trans-dichloroethylene, or trichloroethylene, is used.
Glass thermometers come in three designs: total immersion, partial
immersion, and complete immersion. The total immersion thermometer is
designed to have the bulb and the entire column of liquid up to the indicated
temperature mark on the scale immersed in the medium being measured. The
rest of the stem above the column of liquid sticks up out of the medium. The
partial immersion thermometer has a line scribed around the stem below the
main scale. The bulb and stem are to be immersed up to the line in the
medium to be measured. The rest of the stem is to be at ambient temperature.
The complete immersion thermometer is designed to be entirely under the
surface of the measured medium or inside the oven. The most commonly
encountered example is the refrigerator thermometer that comes in a small
bottle that acts as an attached liquid heat sink (Figure 5.2).

FIGURE 5.1 Liquid-in-glass thermometer. Courtesy of AASHTO.

Stem corrections need to be made to readings from partial immersion

thermometers in the most accurate work. These consist of measuring the
temperature of the stem at the high point of the indicating liquid column
with a second thermometer and applying a correction to the observed
reading from the main thermometer. The magnitude of the correction is
based on the temperature difference between the stem and the observed
temperature. For large differences between the measured medium and the
ambient stem, two or more readings along the length of the stem may be
necessary. Stem corrections are explained in detail in NIST Special
Publication 1088, Maintenance, Validation, and Recalibration of Liquid-in-
Glass Thermometers, available from the National Institute of Standards and
Technology (NIST). (Cross, C.D., W.W. Miller, D.C. Ripple, and G.F.
Strouse, Maintenance, Validation, and Recalibration of Liquid-in-Glass
Thermometers. NIST Special Publication 1088, January 2009, U.S.
Department of Commerce, Washington, D.C.) Other corrections explained in
the monograph include use of a total immersion thermometer in a partial
immersion style and use of a partial immersion thermometer in a total
immersion fashion.
FIGURE 5.2 Enclosed-chamber thermometers. Reprinted with permission
from Thermco Products, Inc.

Within the environmental laboratory, thermometers calibrated by NIST

or traceable to a calibrated NIST thermometer are regarded as the standard.
These thermometers have serial numbers engraved on the stem and come
with certificates that attest to their lineage to a NIST calibration. The
certification paperwork will list the temperatures at which calibration was
performed and the necessary corrections to be added or subtracted from the
indicated temperature to obtain an accurate temperature.
 Unless the thermometer can be viewed and read through a glass door or
port in the side of a temperature-controlled enclosure such as a freezer,
refrigerator, incubator, or oven (Figure 5.2), the bulb of the thermometer
needs to be immersed in a medium to serve as a heat sink. This allows the
thermometer to be removed from the enclosure and read without the
indicated temperature changing rapidly as the bulb comes in contact with the
room-temperature atmosphere. For cold or moderately warm enclosures, the
bulb is immersed in propylene or ethylene glycol, whereas oven
thermometers are typically placed in beakers of sand. Certified thermometers
are available with attached heat sink reservoirs.
Glass thermometers are fragile instruments and, especially for the NIST-
certified thermometers, may be quite expensive. The thermometer may be
protected by enclosing it in an armored case; although this will not protect it
from breakage when it is dropped, it will allow it to survive occasional
bumps against hard objects (Figure 5.3). There are also Teflon®-coated
thermometers, which will confine the glass shards and liquid contents upon
breaking.
Electronic-indicating thermocouples are widely used for field
measurement of temperature. These are based on the generation of DC
potential on the opposite sides of a junction formed from two different types
of metal. The common thermocouple constructions are listed in Table 5.2.
Exposed-wire thermocouples have the fastest response time and longer
operating range; however, they are subject to corrosion and other chemical
reactions. Enclosed-tip thermocouples provide isolation from the
environment and are commercially available.

FIGURE 5.3 Digi-Sense® Armored Liquid-In-Glass Thermometer. Photo

copyright © Cole-Parmer.

Moving molecules generate electromagnetic waves in the infrared part of

the spectrum. The higher the temperature of the object, the stronger the
infrared radiation. This can be measured with an infrared detector. Not all
materials emit infrared radiation with the same efficiency (emissivity),
making a universal infrared temperature monitor an impossibility. Emissivity
is a scale ranging from 0 (no infrared emission) to 1 (perfect emission)
(Table 5.3). However, the infrared thermometer can be calibrated for a
particular object and becomes useful for noninvasive fast-repetitive
temperature measurements of this item. An example of this is the checking
of the temperature of samples received at the laboratory by the sample-
receiving personnel. The infrared thermometer is calibrated for each type of
sample container and used to check the arrival temperature without having to
open the container. These infrared thermometers are available either as
complete units or as probes that plug into appropriate (J- or K-type)
thermocouple meters.

TABLE 5.2 Thermocouple types, constructions, and temperature ranges.

Thermocouple type (range °F) +pole −pole

J (−310–1,832) Iron Constantan (NiCu)
K (−418–2,507) Chromel Alumel (NiAl)
T (−418–752) Copper Constantan
E (32–1,650) Chromel Constantan
R (32–3,210) Pt–13%Rh Pt
S (32–3,210) Pt–10%Rh Pt

Temperature indicator strips are based on color changes of liquid crystal

spots with changing temperatures (Figure 5.4). They are available as either
single-temperature indicating spots or multiple spots on a strip of plastic.
Irreversible and reversible indicating strips are manufactured. They can be
attached to the sides of sample containers to give an immediate indication of
the temperature. The reversible, multipoint, self-adhesive design can be left
on the container as it makes its way through the laboratory, allowing
temperature checking at any time. As the temperature rises, the color of the
liquid crystal changes from brown to black (brown, yellow, green, blue,
violet, and black). A green color indicates that the temperature is at the rated
temperature written on the indicator.

TABLE 5.3 Infrared emissivity of various materials.

Material Emissivity
Ceramic 0.80–0.95
Glass 0.7–0.85
Metal 0.02–0.21
Plastic 0.35–0.85

FIGURE 5.4 Digi-Sense® 7 Point Reversible Temp Label. Photo copyright

© Cole-Parmer.

3.0 CALIBRATION CHECKING AND

MAINTENANCE OF LIQUID-IN-GLASS
THERMOMETERS
Liquid-in-glass thermometers are fragile precision instruments and should be
treated gingerly. Aside from breaking the thermometer, the most common
problems with their use involve liquid column separations and changes in
the bulb. Liquid column separations occur infrequently if the thermometer is
stored in an upright (vertical) position and not subjected to a lot of vibration
or jarring. The thermometers filled with organic fluids are much more prone
to column separation than the mercury-filled ones. This is especially true if
the thermometers are stored horizontally in a drawer. The easiest way to re-
establish a continuous column is to slowly and carefully cool only the tip of
the bulb in a dry ice (solid carbon dioxide) slurry of acetone, toluene, or
alcohol. This contracts the liquid into the bulb and, if it is mercury-filled,
solidifies it. Upon allowing the bulb to gradually warm to room temperature,
the liquid refills the capillary with a continuous column. Low-temperature
organic fluid-filled thermometers may require a liquid nitrogen bath to
completely contract the fluid into the bulb. Another technique applicable to
the organic liquid-filled thermometers is to gently heat (with your fingers or
hand) the glass capillary at the spot of the separation and carefully tap the
stem. This will allow the fluid to flow around the separation and move the
separation up to the top of the column.
The bulb is the most delicate part of the thermometer, and changes in its
volume will affect the calibration. All bulbs age and will change volume.
They are made of glass, and the glass is a supercooled liquid, subject to
creep and flow. The aging process is hastened by prolonged heating of the
bulb, especially to temperatures above 250 °C. The calibration of the
precision thermometer, regardless of type, should be checked at least
annually. Calibration checking at the ice point of water is the typical
procedure.
A suitable ice bath is formed by shaving ice made from the purest water
available (distilled) into a Dewar flask, then adding pure water to form a
slurry. The slurry is compressed and the excess water decanted. The bath
temperature stabilizes after 15 to 30 minutes, and the thermometer is
immersed to the appropriate point and the temperature determined. The
auxiliary scale is provided for calibration checking on thermometers that do
not include the ice point on the main scale. The difference between the 0.01
°C mark on the thermometer and the actual reading in the ice bath is the
calibration correction, which is applied to all subsequent temperature
readings. This procedure works because the change in calibration is
attributable to size changes in the bulb, not changes in the stem.
Many laboratory certification programs require an annual calibration
check of each thermometer used in the laboratory against a NIST-traceable
thermometer. For most purposes, this can be accomplished by checking a
single-point cross-reading of the thermometer against the traceable
thermometer. For example, the water bath for mercury digestions is required
to be at 95 °C. Placing the NIST-traceable thermometer and the water bath
thermometer in a water bath at 95 °C and verifying that both thermometers
read 95 °C, then writing the procedure and observations down in a logbook,
serves as adequate calibration checking of the water bath thermometer. The
water bath thermometer should be tagged with a cross reference to the
logbook entry.
In the microbiology laboratory, a more extensive annual thermometer
calibration check against the NIST-traceable thermometer is required to be
documented. The procedure involves suspending the NIST-traceable
thermometer and the thermometers to be checked in a stirred water bath of at
least 2000-mL capacity, raising the temperature to around 50 °C, then
allowing the water bath to cool. As the water bath cools, the temperatures on
the NIST-traceable thermometer and the checked thermometers are recorded
every 1 minute over the temperature range of either 50 to 40 °C or 40 to 30
°C, depending on which incubator temperature (44.5 or 35.0 °C) the
thermometer is intended to monitor.
 6
Preparation of Solutions
and Dilutions

1.0 USE OF BALANCES AND VOLUMETRIC WARE

2.0 PREPARATION OF DILUTIONS
3.0 STORAGE OF PREPARED SOLUTIONS

Most analytical chemistry is carried out in solution. When proper attention

is paid to the preparation, storage, and manipulation of solutions, analytical
procedures can be performed with a high degree of accuracy and
reproducibility. Many solutions are available from chemical supplyhouses
with a certificate of analysis attesting to the accuracy of the stated
concentration. With only a few exceptions, it is possible for the laboratory
technician to prepare and standardize the needed solutions to the same or
greater accuracy and at a substantial savings in reagent costs. The major
reason a laboratory would purchase a prepared solution is that the time
spent by the analyst in preparing the solution is more valuable than the
added expense of the commercially available solution. The time required by
the analyst to standardize a solution has to be spent regardless of whether
the solution is purchased or prepared. Other considerations that will
influence the decision to prepare or purchase a solution are the shelf life of
the needed reagents and a severe lack of appropriate storage area. When a
reagent with a short shelf life is used to prepare a solution and the
laboratory ends up disposing of the excess, then there is little savings in not
using the commercial preparation. Waste disposal costs need to be factored
into most laboratory decisions and will become more important in the
future.
1.0 USE OF BALANCES AND VOLUMETRIC
WARE
A common solution preparation starts with one or more solid compounds
that need to be dissolved in water. The analyst needs to know what purpose
the solution will serve in the analytical procedure. The solution may be used
to add excess reagent to the reaction, and the amount of excess is not
considered critical. Or, it may be that the amount of reagent added has a
critical effect on the outcome of the analysis. If the concentration of the
reagent is included in the calculation of the answer, this is probably the
case. Another clue is in how the amounts of materials are specified in the
method. For example, in a method, solution A is prepared from 5 g of
magnesium chloride dissolved in 1 L of water, whereas the preparation of
solution B calls for 1.378 g of potassium hydrogen phthalate dissolved in
water and diluted to 1000 mL. It should be obvious to the analyst that the
exact concentration of solution A is not going to have a great effect on the
calculated answer, whereas use of 1.391 g instead of 1.378 g of the
phthalate for the preparation of solution B is going to give a significantly
different answer. Solution B is variously called a determinative or
stoichiometric reagent. It is perfectly acceptable for the analyst to prepare
solution A on a toploading balance; however, solution B requires an
analytical balance. If a decision cannot be made as to what type of reagent
is being prepared, it is good practice to treat the solution as if it is a
determinative reagent.
In some methods, preparation of a determinative reagent is as
previously described for solution A, followed by a detailed procedure for
the standardization of the reagent solution. The standardization serves to
establish the exact concentration of the prepared solution. Solution
preparation instructions like this are encountered when the purity of the
reagents is not guaranteed because they are unstable or contain varying
amounts of water. Preparing sulfuric acid solutions for titrations is an
example. Concentrated sulfuric acid contains anywhere between 2 and 8%
water. The analyst prepares a solution by diluting a specified number of
milliliters of concentrated acid to a liter with water, then establishes the
exact concentration by standardization with either sodium carbonate or
sodium hydroxide solution.
 After the analyst has decided what needs to be done, the next step is to
go to the storage area and obtain the required reagents. If 2.314 g of
potassium phosphate dibasic is needed for the solution, the proper solution
is not going to result if 2.314 g of potassium phosphate tribasic, potassium
phosphate dibasic trihydrate, or potassium phosphate monobasic is used.
Although the trihydrate can be used to prepare the desired concentration of
solution, the analyst needs to calculate the proper amount to be weighed
(see Appendix A). The names of chemicals are specific, and it is generally
not sufficient to match three of the four words of the name of the chemical
with what is found on the shelf. Likewise, the spelling of the chemical name
is very specific. Sodium sulfide cannot be substituted for sodium sulfite, nor
potassium ferricyanide for potassium ferrocyanide.
Unfortunately, many compounds have more than a single name, with
phosphate salts being the worst offenders. Sodium phosphate dibasic,
sodium biphosphate dibasic, sodium hydrogen phosphate dibasic, and
disodium hydrogen phosphate all refer to the same exact substance (i.e.,
Na2HPO4). If at all possible, the analyst should try to match the chemical
formula on the bottle label with the chemical formula specified for the
reagent in the method.
Once the proper reagent has been found, the substance needs to be
examined to decide whether it needs to be dried. There is no general rule for
this decision. Some substances, such as sodium sulfide or phenol, will
adsorb water readily once the bottle has been opened; however, attempted
drying in the oven will create a toxic hazard in the laboratory. Some
hydrated compounds will lose part of their water if dried, whereas others
are stable. If there is any doubt, the analyst should consult the material
safety data sheet for the chemical (a good idea in any case) and ask the
advice of a more knowledgeable chemist. If the method specifies that the
reagent is to be dried, the analyst should place slightly more than is going to
be needed in a clean beaker or on a watch glass, then place the material in
an oven at the correct temperature for the specified length of time. Although
104 °C is often the suggested drying temperature, some materials, such as
sodium carbonate, require hotter ovens. Potassium hydrogen phthalate
(KHP) is a commonly used primary standard, and it should be dried before
weighing. Drying at 104 °C for an hour is the proper procedure; however, if
the oven is 125 °C or hotter, the KHP can react with itself to form water,
phthalic anhydride (which is somewhat volatile), and potassium phthalate.
After the substance is dry, it should be transferred to a desiccator while it is
still hot to cool in a dry atmosphere.
In general, it is not a good practice to attempt to weigh a solid directly
into a volumetric flask. The mouth of the flask is too small to get all of the
reagent in without spilling, and the tare weight of the flask will frequently
surpass the capacity of the balance. Reagent funnels are designed to solve
these problems. They are available in a variety of sizes and, because they
are made of glass, can be cleaned and reused. The desired amount of
material is weighed into the reagent funnel, which is then upended into the
mouth of the volumetric flask. Liquid reagents can also be weighed directly
into the reagent funnel using a disposable glass pipet. The reagent is
quantitatively washed into the volumetric flask with water from a squeeze
bottle. Disposable weighing boats can be used for larger amounts of
reagents. A powder funnel should be placed in the mouth of the volumetric
flask to avoid spilling the reagent during the transfer from the weighing
boat. Again, wash the weighing boat out with reagent water from a squeeze
bottle to obtain a quantitative transfer of the reagent into the flask.
After the reagent is in the flask, add water until the flask is about 90%
filled. Stopper the flask and, holding the stopper in place with the thumb,
invert the flask and shake. It is poor practice to close the flask by covering
its mouth with a thumb. Right the flask so that the liquid drains out of the
neck, then invert and shake it at least two more times. Examine the flask to
ensure that the reagent has dissolved completely. This can be deceiving
because some solids exhibit various stages in the dissolution process.
Sugars appear to dissolve readily; however, stabilization of the solution
may take hours. Once the solution has stabilized to room temperature (some
dissolutions generate heat, whereas others cause the solution to become
colder), carefully remove the stopper, add water to almost the volume mark,
and repeat the stoppering and inversion with shaking. The flask should not
be diluted to volume before forming the solution because some materials
will cause the total volume to shrink, whereas others will cause an
expansion upon dissolving.
Some solutions can be successfully prepared directly in a volumetric
flask. However, if the compound needs prolonged stirring or heating to go
into solution, or if the dissolution process is exothermic or a pH adjustment
is required, a secondary container is probably indicated, followed by
transfer and dilution to volume in an appropriate-size volumetric flask.
About 90% of the final volume of water is added to a beaker or Erlenmeyer
flask (Figure 6.1), followed by the weighed reagent, then a magnetic stir bar
is added. Some materials, such as concentrated acids, can react violently
when water is added; as such, they should be cautiously stirred into water.
A cover is typically placed over the opening of the container to prevent dust
and other materials from contaminating the solution and to prevent droplets
of the solution from splashing out. A watch glass over a beaker or inverted
beaker over the mouth of an Erlenmeyer flask are functional covers. A
smaller Erlenmeyer flask inverted in the neck of the larger flask is suitable.
The solution is stirred, with heating if appropriate, until everything is in
solution and it has cooled to room temperature. An alternative is to place
the covered beaker or flask in an ultrasonic cleaning bath and use
ultrasound to dissolve the solid (Figure 6.2).
FIGURE 6.1 Flask with ground glass stopper. Reprinted with permission
from DWK Life Sciences.
FIGURE 6.2 Ultrasonic cleaning baths. Reprinted with permission from
Grant Instruments.

A record of the solution preparation needs to be made in either a

laboratory notebook maintained for this purpose or on the back of the
benchsheet that uses the solution. Information to be recorded includes the
date and time of the preparation, the person performing the preparation, the
manufacturer’s lot number of the reagents used (along with a cross-
reference to the location of the manufacturer’s certificate of analysis), the
weights and volumes of the preparation, and other comments or
observations such as a pH check on the reagent, if required.

2.0 PREPARATION OF DILUTIONS

It is important for the technician to recognize the limitations of laboratory
equipment and to know how to get the best results from the available tools.
For most analyses in the water and wastewater laboratory, the target
analytes are being analyzed at the part-per-million level or lower.
Consider the preparation of an ammonia standard at the 5.0-mg/L level.
Ammonia standards are typically prepared from ammonium chloride, which
has a molecular mass of 53.49 g/mol and an equivalent mass of 3.141
g/1.000 g of ammonia. Although it is possible to weigh out 15.7 mg of
ammonium chloride on the analytical balance and dilute it to 1000 mL, the
balance has limitations; thus, the known concentration of this standard has
an uncertainty between 1 and 6%, ignoring any contribution from the
adsorption of water during the weighing. There is also the problem of
weighing anything to a desired value of ± 0.1 mg, a virtual impossibility
because of the size of the individual crystals of most materials. A much
better procedure is to prepare a concentrated stock solution of ammonia
standard, say at the 500.0-mg/L level, by weighing 1.570 g of ammonium
chloride and diluting to 1000 mL. In this case, the desired weight is ±1.0
mg, which is possible to obtain with care. A 10.00-mL aliquot of the stock
solution is removed with a 10.00-mL volumetric pipet and diluted to 1000
mL in a volumetric flask to form a working standard of 5.000 mg/L of
ammonia known to ± 0.001 mg/L and an uncertainty much less than 0.1%.
An added benefit of this procedure is the ability to dilute 1.00-, 2.00-, 5.00-,
and 20.00-mL aliquots of the stock solution to form a whole series of
working standards from a single, highly reliable weighing, thus removing a
large potential source of random variation in multilevel calibrations.
The following simple equation is used to predict the final concentration
of dilution procedures

Ci and Cr are the initial and final concentrations, respectively, while Vi

and Vr are the initial and final volumes of the reagent solution used in the
dilution. Any concentration units can be used as long as they are the same
units for the initial and final states. Suppose the analyst wishes to prepare
50.0 mL of a 6.00 mg/L standard from a 100-mg/L stock solution. Ci is 100
mg/L, Cr is 6.00 mg/L, and Vf is 50.0 mL; thus, the only unknown is Vi (i.e.,
the volume of the stock solution to be diluted). The values are plugged into
the equation, and the value of Vi is calculated as 3.00 mL. The analyst
knows that the desired solution will be prepared by adding 3.00 mL of the
stock solution to the 50.0-mL volumetric flask and diluting to volume.
The preparation of stock solutions allows for use of the analytical
balance in a highly accurate part of its range and minimizes the number of
required weighings. Another advantage is that the working standards can be
quickly prepared as needed—even daily—from the single bottle of stored
stock solution, rather than having to store a separate bottle of each of a
number of standards.
Other techniques commonly used to indicate a dilution are the 1 + 4 and
1:5 notations. The two notations result in the same solution and are
frequently used interchangeably, except in cases where the total volume
contracts when the dilution is prepared, such as the preparation of sodium
hydroxide solutions. The first (1 + 4) tells the analyst to take 1 volume of
the stock solution and add to it 4 volumes of dilution solvent. The second
(1:5) tells the analyst to take 1 volume of the stock solution and add solvent
until the total volume is 5 times the original volume. The exact size of the
volume is immaterial. It could be 1.0 mL or 7.0 L. In the 7.0-L case, the
analyst would take 7.0 L of the stock solution and add to it 28.0 L of the
solvent to satisfy the 1 + 4 requirement; 7.0 L of the stock solution would
be diluted out to 35.0-L total volume to satisfy a 1:5 requirement.
Decade dilutions or serial dilutions are required in many microbiology
procedures. These consist of taking the sample and preparing solutions of
1:1, 1:10, 1:100, 1:1000, and so on. Suppose the analyst is going to need to
prepare a 1:10 and a 1:100 dilution of the sample on a regular basis. One
way to prepare for this is to sterilize some dilution bottles with 90 mL of
dilution water in them and other bottles with 99 mL of dilution water. The
addition of 10 mL of the sample to the 90-mL bottle gives the 1:10 dilution,
and 1 mL of the sample added to the bottle with 99 mL of dilution water
gives the needed 1:100. An alternative is to prepare all the dilution bottles
with 99 mL of dilution water, then add 1.0 mL to one bottle to give the
1:100 dilution and 11.0 mL to the next bottle to give the 1:10 dilution
(11:110 = 1:10). Many nominally 10-mL serological pipets actually have
markings to 12 mL to make this easier.
3.0 STORAGE OF PREPARED SOLUTIONS
Volumetric flasks are not storage vessels for prepared solutions. Once a
solution has been prepared, it should be transferred to a clean glass or
plastic bottle, as appropriate. The storage bottle should have a screw-cap
closure with a Teflon® cap liner. Ground-glass stoppered reagent bottles
should be used only for stable acidic solutions. Basic solutions will cause
the stopper to freeze in a glass bottle. Basic solutions should be stored in
high-density polyethylene or Teflon® bottles. Reagents should be labeled
with the contents, concentration, date of preparation, date of last
standardization, expiration date, proper storage location, and preparer’s
initials.
Some reagent solutions can be stored on the laboratory bench; however,
storage in the dark in an appropriate cabinet is generally a better idea.
Solutions should be segregated as to acids, bases, oxidizers, reducers, and
flammables, the same as for the stock reagents. Self-adhesive, right-to-
know labels and hazard placards on individual bottles of reagent solutions
are not a bad idea. Reagent solutions should never be stored with samples
because of cross-contamination problems.
Many solutions should be refrigerated. The volume and concentration of
the reagent solution changes with temperature; thus, the bottle should be
removed from the refrigerator and allowed to warm to room temperature
before dispensing.
FIGURE 6.3 Reagent dispenser bottle. Reprinted with permission from
Bel-Art— SP Scienceware.

Water and other solvents will evaporate from bottles, even when the cap
is tight. This is particularly true when reagent solutions are stored in
dispenser bottles because they are prone to solvent evaporation (Figure 6.3).
Evaporation can be significant over a period of time as little as a week at
room temperature. Therefore, each time solution is removed from the bottle,
a permanent ink line indicating the liquid level should be drawn on the
outside of the bottle and dated.
Reagents should never be dispensed directly from the container by
pipet. Instead, a portion of the reagent should be poured into a beaker and
the pipet filled from the beaker. The excess reagent is disposed of, rather
than returned, to the container. This avoids contamination of the reagent by
a dirty pipet.
 Reagents are the primary means for performing analysis in the
laboratory. The care that is taken with the preparation and storage of the
reagents will be directly related to the quality of the analytical results
generated by the laboratory.
 7
Calibrations

1.0 STANDARDIZATION OF SOLUTIONS

2.0 DIRECT STANDARDIZATION
3.0 INDIRECT STANDARDIZATION
4.0 SOLUTION LIFETIME
5.0 CALIBRATION
5.1 One-Point Calibrations
5.2 Multipoint Graphical Techniques
5.3 Multiple Standard Additions
6.0 FREQUENCY OF CALIBRATION AND CALIBRATION
CHECKS
7.0 REFERENCE

Instruments must be calibrated to obtain reliable analytical results. Every

procedure in the U.S. Environmental Protection Agency (U.S. EPA) method
manuals and Standard Methods for the Examination of Water and
Wastewater (APHA et al., 2012) lists steps for preparing standard solutions
and calibrating the test.

1.0 STANDARDIZATION OF SOLUTIONS

Quantitative analysis requires the use of standard and reagent solutions of
known concentration. In general, it is poor analytical practice to prepare a
solution and use it without verifying the concentration. First, no chemical is
100% pure, and many chemicals are available in a range of purities, which
may or may not be correctly stated on the label.
Second, chemicals often react with the solvent or with trace impurities in
the solvent such as oxygen, carbon dioxide, ammonia, or heavy metals.
Although primary analytical standards can be weighed out and diluted to
volume to form solutions with concentrations known to a high degree of
accuracy, the limited number of primary standards means that some other
measures are required to establish the strength of most solutions. The idea is
to chemically relate the concentration of the reagent solution back to a
primary analytical standard. In some cases, this can be done directly;
however, the more common situation is for two or more steps to be involved.

2.0 DIRECT STANDARDIZATION

An example of direct standardization is the ferrous ammonium sulfate (FAS)
titration of total residual chlorine using N,N-diethyl-p-phenylenediamine as
the indicator (SM18 4500-CI F). Solutions of chlorine with known
concentrations are difficult to prepare and are not stable. The procedure is
calibrated by standardization of the FAS titrant against the primary standard,
potassium dichromate. Given the directions in Section 2.c of the Standard
Methods procedure, when the FAS is exactly the correct titer, 2.82 mL of the
0.100-normal (N) dichromate is required for neutralization of 100 mL of the
FAS solution (see Appendix D, where this result is worked out in detail).
The FAS solution is prepared as in the method directions. An aliquot is
dispensed into a beaker or Erlenmeyer flask, and a stir bar is added along
with the required acid and indicator. Then the solution is titrated with the
standard potassium dichromate solution. Averaging at least two repeats of
the determination gives a reliable concentration known to at least three
significant figures and a solution of FAS ready for residual chlorine
titrations.
Another example is the standardization of solutions of bases such as
sodium or potassium hydroxide with the primary standard potassium
hydrogen phthalate (KHP). The KHP is dried in an oven at 104 °C for an
hour, then a portion is weighed and dissolved in 100 mL of reagent water.
The KHP has a high equivalent weight (204.23 g/mol), allowing easy and
accurate weighing. For standardization of a 0.1-N base solution, typically
approximately 400 mg of KHP is required for titration with approximately
20 mL of base. Several drops of visual indicator solution, such as
phenolphthalein, is added, then the mixture is stirred and then titrated until a
permanent pink color is formed. The calculation of the base concentration is
as follows:

In practice, two or three standardizations with KHP are performed, then

the concentrations are averaged to arrive at an accurate titer. Titer means the
same thing as concentration and is commonly used when discussing the
concentration of titration reagents. Base solutions are particularly sensitive
to degradation in concentration because of carbon dioxide absorption and
should be standardized each day they are used.
Acid standard solutions can be directly determined by using sodium
carbonate. The carbonate has a tendency to absorb water from the air rapidly,
and an error may be introduced when aliquots are weighed in humid
environments. A preferred procedure is to standardize a base solution with
KHP and then use the base to standardize the acid solution.

3.0 INDIRECT STANDARDIZATION

Many reagents do not react directly with an existing primary standard. To
establish an acceptable chemical relation to a primary standard, a number of
reactions may be required. An example is the standardization of phenol
solutions presented in SM225530 C 3.a.1. The phenol solution is reacted with
a bromate-bromide solution to form tribromophenol. The excess bromate-
bromide remaining in solution after the reaction with phenol is determined
by titration with sodium thiosulfate solution. The titer of the sodium
thiosulfate solution is established by titration of iodine prepared from the
primary standard potassium biodate.
 Standard Methods has many other indirect standardizations that are used
to establish the exact concentration of reagents. It is sometimes quite
challenging as an exercise of your chemical knowledge to relate the
standardization procedure to the actual analysis.

4.0 SOLUTION LIFETIME

Reagents and standards for quantitative analysis have a limited shelf life. For
many reagents, the lifetime is given in the method; however, for others, a
maximum life of 1 year is assumed. This is not to imply that the analyst
needs to standardize a solution only once, immediately after preparation, and
then use the solution until the bottle is empty or a year has elapsed. Solution
titers change through processes as passive as liquid evaporating from the
bulk solution and condensing as droplets of pure solvent on the walls of the
bottle at the top (a sub-boiling phenomenon). Reagents should be agitated to
mix the contents before use. A loose top can allow the solvent to evaporate,
even in the refrigerator. This results in the reagent titer increasing slightly.
To keep track of evaporation problems, the liquid level and the date should
be marked on the outside of the bottle with a permanent marker each time
the reagent is used.
Other reagents react either directly or indirectly with gases in the
atmosphere. This is the case with sodium thiosulfate and sodium sulfite
solutions. Every time the analyst shakes the solution and opens the bottle,
the titer is going to drop over the next week or so. Sodium hydroxide and
other base solutions are constantly absorbing carbon dioxide from the air,
which is what causes the crust of hard, white material to form around the
opening of the sodium hydroxide bottle. It is not sodium hydroxide left after
evaporation of the water; it is sodium carbonate. This is also one of the two
reasons sodium hydroxide solutions are not to be stored in bottles with
ground-glass closures: the sodium carbonate formed upon reaction with
atmospheric carbon dioxide helps seal the ground surfaces together. The
other reason is that sodium hydroxide reacts chemically with the ground
surfaces and bonds them together. Acidic solutions will absorb atmospheric
ammonia and neutralize the solution slowly over time.
 All these reasons suggest that reagent titers should be checked more
frequently than once a year. General guidelines applicable to all reagents
cannot be given. Some reagents need to be standardized only once a week,
whereas others require daily checking. Analysts who want to perform only
the best-quality work will standardize the solution each time they use
reagents.

5.0 CALIBRATION
Calibration consists of establishing a relationship between the analytical
response of a test and the concentration of target analyte in the sample. This
is done by testing samples of known concentration of the analyte and
recording the analytical response. The analytical response, for example, may
consist of a voltage, resistance, or current difference between an electrode
and a reference, a decrease in the amount of light of a specific wavelength
passed through a solution (absorbance), or the number of milliliters of titrant
required to reach an endpoint.
The most common procedure for calibrating an instrument is to perform
the test on standards (at least three if another number is not specified in the
method) and then create a graph of the response against concentration.
Graph paper is available in any office supply store or from scientific supply
houses. Papers with 10 × 10 or 20 × 20 squares to the inch are the most
versatile sizes for laboratory work, although others can be used. Some
methods need to be calibrated with semilog paper (Figure 7.1); however, for
most methods, rectangular graph paper (Figure 7.2) is appropriate.
FIGURE 7.1 Example of a three-cycle semilog graph.
FIGURE 7.2 Example of a rectangular graph.

5.1 One-Point Calibrations

The only one-point calibrations permissible in the realm of environmental
analysis, which includes most of the operations in a water or wastewater
laboratory, are titrations. The reason for this is that 1 mL of a titration
solution will only react with a certain amount of target analyte and no more.
Titrant is added until all the analyte is consumed, and the number of
milliliters of titrant used is directly related to the amount of analyte in
solution. Most other analytical procedures are measuring some physical
property of the sample, which is often established through a chemical
reaction. Physical properties are concentration dependent, and large amounts
of the property can result in additive or subtractive interference with the
measurement. Some of the older pH meters will only accept a single-point
(pH 7.00) calibration, but the laboratory needs a new pH/ion-selective
electrode meter anyway for fluoride and ammonia analyses. One-point
calibrations work in the drunk driver program of the local law enforcement
establishment because they have a dedicated instrument that reads either
above or below the legal breath alcohol limit. The actual value above or
below the regulatory limit is of passing interest only to the news media. It is
a pass/fail test, and thus a calibration of the breath analyzer at the legal
regulatory limit is acceptable. However, most of the instruments in water and
wastewater laboratories are used for numerous tests, and a more detailed
calibration is required for each analysis.

5.2 Multipoint Graphical Techniques

The example presented herein will demonstrate the variety and
characteristics of multipoint calibrations. An analyst has made phosphate
standards and analyzed them by method SM224500-P E with a colorimeter to
give the data in Table 7.1.
To plot the data on 8.5 in. × 11-in. graph paper with 10 × 10 squares to
the inch, the analyst should examine the vertical and horizontal axes of the
paper, decide which will be the response axis and which the concentration
axis, and establish the scale. The best procedure is to try to use the whole
length of an axis by examining the calibration data. In the aforementioned
case, the concentration values run from 0 to 1.500 mg/L and the absorbance
values from 0 to 0.793. The paper has 20 major blocks up the vertical axis
and 14 along the horizontal axis. If every dark line on the horizontal axis is
labeled with a whole number for concentration, only about one-tenth of the
graph paper is used. A more usable graph results from labeling every dark
line with a decimal number (0.1, 0.2, 0.3, 0.4, and 0.5). This allows the full
length of the axis to be used; however, the last point falls off the axis. A
good technique is to count the number of dark lines available (14) and divide
by the needed range (1.5 units). The result should be rounded down to the
next integer; in this case, the result is 10. This means that, beginning with
the leftmost line on the graph labeled 0, every tenth dark line should be
labeled with a unit number. For the vertical axis, there are 20 dark lines and
the needed range is 0 to 0.793. Round the range up to 0.800. Eight major
divisions (0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, and 0.8) will cover the needed
range, so divide the number of dark lines (20) by the number of divisions
(8), which gives 2.5. Round the result down to the nearest integer, 2.
Beginning with the bottom line on the graph labeled 0, every second dark
line should be labeled with an absorbance major division. With the axes
labeled and the graph ready to go, the points are plotted. The points are
placed on the graph at the intersection of the horizontal line representing the
absorbance value and the vertical line representing the concentration value.
An example of this plot, called a scattergram, is shown in Figure 7.3; a
statistics program for a personal computer was used to generate the graph.

TABLE 7.1 Absorbance data for phosphate calibration.

Phosphate (mg/L) Absorbance

0.010 0.011
0.020 0.018
0.050 0.036
0.100 0.070
0.200 0.141
0.300 0.210
0.500 0.359
0.750 0.520
1.000 0.660
1.250 0.727
1.500 0.793

The next task is to use the plotted data to generate a calibration curve.
The easiest way to construct the calibration curve is to simply connect the
dots. This is shown in Figure 7.4, for the data in Table 7.1. In this step, an
assumption is being made that if the absorbances are known for two
concentrations, then there is a regular change in the value of the absorbance
as the concentration of other samples is tested, as long as the unknown
concentration lies between the original two standards. If the standards are
FIGURE 7.3 Computer generated calibration plot of phosphate data.
FIGURE 7.4 Line graph of phosphate data.

fairly close together in concentration, say 0.1 mg/L and 0.5 mg/L, then this
may be a valid assumption. However, the farther apart the concentrations of
the standards are, the weaker the rationale upon which the calibration is
constructed. Suppose that only the first and last standards in Table 7.1 had
been tested and the calibration curve constructed by connecting only the
lowest and highest points with a straight line. It is obvious that incorrect
values could be obtained for sample results determined with such a graph.
The calibration plot is used by running the test procedure on an unknown
sample, then reading the concentration value from where the absorbance
value intersects the curve. For example, an absorbance of 0.450 is located on
the absorbance axis, then a straight line is followed over to the intersection
with the curve (see Figure 7.5).
Next, a vertical line is dropped from the point of intersection of the
absorbance with the calibration curve down to the concentration axis. This
process, illustrated in Figure 7.6, gives a value of about 0.65.
FIGURE 7.5 Locating an absorbance value intersection with the
calibration curve.

There are several important characteristics of the calibration curve in

Figure 7.4. First, the set of data points actually curves. This is not a problem
as long as the response changes continuously as the concentration increases.
However, when the response flattens out, the increased concentrations of
analyte do not give differing results. This phenomenon is called saturation
and is illustrated in Figure 7.7. In the example, saturation of the colorimeter
response has occurred at phosphate concentrations of 2.0 mg/L and higher.
FIGURE 7.6 Locating a concentration value from an absorbance result.

The analyst knows that there is a usable calibration up to 1.5 mg/L, but
no higher.
In most Standard Methods and U.S. EPA procedures that specify a set of
concentrations for the calibration standards, the range of the calibration that
is linear has already been determined and the calibration standards are
chosen to encompass only the linear range. In many instances, such as this
phosphate data example, the usable range of calibration is, in fact, much
larger than that specified in the method. It is in the best interest of the analyst
to use the maximum range of calibration possible for the test. However, if
the analyst is going to use nonlinear portions of the calibration, it is also
imperative to determine where the curve begins to saturate.
The phenomenon of saturation is why an analyst must never report a
value for a sample that is extrapolated by extending the calibrated range
beyond the highest calibrated point. This is a misapplication of the
assumption under which the calibration curve was constructed. The proper
procedure is to either repeat the test with a smaller amount of sample or
dilute the sample with reagent water and repeat the reading, then correct the
determined concentration by the dilution factor. For example, using the
aforementioned calibration curve for phosphorus, a sample reads 0.912
absorbance, which is above the calibrated range. The sample is diluted with
4 parts of reagent water (represented by 1:5 or 1+4), which is a dilution
factor of 5. The sample is re-read to give a 0.693 absorbance and a diluted
concentration of 1.14 mg/L. The diluted concentration is multiplied by the
dilution factor (5) to give the true concentration of 5.70 mg/L.

FIGURE 7.7 Added phosphate calibration points demonstrating saturation.

Linearity is a quality of a calibration in which a single straight line

connects all the calibration points. An equation of the calibration can be
determined of the following form:
where a and b are constants. This equation can be programmed into a
calculator or personal computer to generate the concentration of the sample
given the response value. A distinct advantage to constructing a linear
calibration is that it tends to minimize random analytical errors that occurred
during the testing of the calibration standards. These errors may result from
variation in the exact volumes of solutions used to make the standards,
variation in the exact calibration of the glassware used to make the
standards, differences in the time and temperature conditions under which
the reaction occurred, the presence of random contamination of the
glassware, and other causes. Another advantage is that if a standard has been
prepared incorrectly, it sticks out because its point lies far off the line
connecting the rest of the points and it can be eliminated from the
calibration. This is one reason to use more points than the minimum required
to calibrate a test. Eliminating a point from consideration in the calibration is
termed censuring.
Although the use of a calibration equation can be quite a timesaver and
lead to automation of a test procedure, there are a few hidden pitfalls in its
use. The first pitfall is that a negative response will generate a negative
concentration, which is analytically impossible. A human analyst would
never report a negative result from a negative response, but a computer
generally is not programmed to make such a decision. The second common
pitfall is that a high response will be converted into a concentration that may
be well beyond the saturation level of the test.
FIGURE 7.8 Regression line of all phosphate data.

If a series of calibration points appears to be linear, the analyst can use a

straight edge to construct a straight line through the points. Frequently, a
statistics program or graphing program on a computer is used to
automatically generate a linear equation for the set of points. A regression
equation can be generated for almost any calibration, even those that are
obviously inappropriate. An example is shown in Figure 7.8, where a linear
fit equation is determined for all the phosphate calibration data in Table 7.1.
The regression coefficient (R2) is generated at the same time as the
regression equation by the computer program and gives an indication as to
how well the derived equation describes the set of points. Regression
coefficients range from 0 to 1, with 1 being a perfect fit of the points to the
equation. In this case, a regression of R2 = 0.978 is obtained, which may be
fine for other sciences, but not for analytical chemistry. If the top two points
are eliminated from the set of calibration data, then the linear equation has a
much better fit to the points, generating the plot in Figure 7.9. A regression
coefficient of R2 = 0.998 is obtained, but the calibration range is now limited
to 1.0 mg/L.
The regression equation generated by most of the statistics programs is
obtained from a table of results, or written on the bottom of the plot, as
illustrated in Figures 7.8 and 7.9.
The equation is written with a computer version of scientific notation for
numbers with lots of zeros. The “E-3” means 10−3. A “*” is used to
represent multiplication and “^” represents an exponent. The regression
results in Figure 7.9 thus are interpreted as phosphate (mg/L) = −0.009895 +
1.491(absorbance), R2 = 0.998.

FIGURE 7.9 Regression line for phosphate data with top two points
eliminated.

Tests such as hydrogen ion, fluoride, and ammonia, which use a millivolt
reading from an electrode as the analytical response, give calibrations that
are curved. Refer to Table 7.2 for fluoride calibration data. These curves all
fit an exponential equation of the following form:

Where
mV = the reading in millivolts and
a and b = constants determined for each test.

The hydrogen ion determination is reported in terms of pH, which

substitutes into the equation as

The analyst can plot the millivolt readings against the concentration of
the standards on rectangular graph paper (Figure 7.10). However, these
electrode-based procedures have long working ranges. For ammonia, the
range is typically 0.1 to 1000 mg/L, whereas the hydrogen ion range is from
pH 1 to 13, a 1012 range of concentration. Most of the good data would be
unreadable from a rectangular graph. Further, the degree of curvature
requires testing a lot of standards to obtain an accurate calibration. Two
alternatives present themselves. The first is to use rectangular graph paper
and plot the log of the concentration on the horizontal axis rather than the
concentration itself (Figure 7.11).
This requires use of a calculator for each sample. The other alternative is
to use the semilog paper illustrated in Figure 7.1 and simply read the
concentration off the graph (Figure 7.12). The number of cycles is related to
the exponential range of the paper: two-cycle paper has two powers of 10
(from 1 to 100), three-cycle has three powers of 10 (from 1 to 1000), and
four-cycle has four powers of 10 (from 1 to 10 000).

TABLE 7.2 Fluoride calibration data.

 Fluoride (mg/L) log (F) rnV
0.10 −1.00 160.8
1.00 0.00 107.1
10.0 1.00 48.6
100 2.00 −11.0

FIGURE 7.10 Fluoride calibration data plotted on rectangular paper.

FIGURE 7.11 Log of fluoride calibration data plotted on rectangular paper.
FIGURE 7.12 Fluoride calibration data plotted on semilog paper.

The question arises as to how many points are required to establish a

calibration.
This depends on the test and the method. Cyanide requires at least six
points, whereas the methylene blue active substances test for anionic
surfactants calls for 10. In general, if the calibration is shown by the analyst
to be linear, then at least a low calibration point, a high calibration point, and
a midrange standard should be performed. This is because any two points
can be connected with a straight line, whereas three points in a single line
demonstrate linearity with a higher degree of confidence. For calibrations
that are curved, a minimum of four points should be used to demonstrate that
the calibration equation actually fits the curve, and many more points may
be necessary to demonstrate the structure of the curve with a suitable degree
of confidence. U.S. EPA recently proposed that 10 points are necessary to
demonstrate a calibration curve with an order higher than linear.
 A related question deals with blanks and whether they are part of the
calibration. The blank is defined as a test performed on a sample containing
none of the target analyte. Its purpose is to set an absorbance zero for the
instrument and, more importantly, to detect the presence of analyte
contamination in the test reagents. Zero target analyte in the sample requires
that not a single molecule be present; no tests are sufficiently sensitive to
make such a determination. A reported concentration of zero is relative only
to the sensitivity of the test, which is more properly referred to as the
detection limit. Just considering the tests commonly performed in the water
and wastewater laboratory, in almost every case of a nondetect on a sample,
there is a more sensitive instrument available that will result in detection at a
lower level. Although there are established protocols for determination of
the method detection limit (Title 40, Code of Federal Regulations, Part 136,
Appendix B), the proper procedure in calibration is to select the lowest
calibration standard for preparation of the curve. Tested samples that give a
response lower than the lowest calibration standard are reported as “less
than” plus whatever the value of the lowest calibration standard is or as a
“nondetect”.
A separate calibration should be performed on each instrument used by
each analyst who performs the test. Calibration data and graphs should be
labeled with the test performed, the date of calibration, the name of the
analyst, and the instrument, if applicable. These data become part of the
legal documentation required for verification of a test result.

5.3 Multiple Standard Additions

Multiple standard additions is a technique for establishing a calibration in
the specific sample for the desired analyte. It eliminates matrix effects,
which interfere with the determination of the target parameter. A blank is
prepared for the test in order to set the instrument zero. If the sample is
known or suspected to contain analyte at a level above the normal calibrated
range of the test, the sample is diluted to bring the analyte level down into
the lower to middle part of the calibration. At least three aliquots of the
sample are spiked with increasing levels of the calibration standard stock
solution, with the concentration levels chosen to bracket the suspected level
of the parameter in the original sample. The test procedure is then performed
on the blank, sample, and spiked aliquots of the sample. The analytical
responses are recorded, and a plot is prepared of the known spike level of
analyte against the response. The best-fit line is constructed through the
points and extended down to intercept the concentration axis at the zero
response level. The intercept value is the negative of the amount of target
analyte in the sample. An example of this procedure is drawn from reactive
phosphorus analysis using the ascorbic acid combined reagent (SM224500-P
E.).
Three spiked samples are prepared by diluting the stock phosphorus
standard 1:10 to give a 5.00-μg/mL working standard solution and then
diluting 1.00, 2.00, and 4.00 mL of this working standard to 100.0 mL with
the sample. The test procedure is performed using the normal reagent blank
to set the absorbance zero. Refer to Table 7.3 for multiple standard addition
values for a phosphorus test.
The three spike solutions are plotted and a linear regression is
determined. The regression coefficient must indicate a high degree of
linearity of the spiked sample data points (typically, a minimum of 0.995 is
required). This example generated an R2 of 0.999. The intercept can be
determined graphically, or the value from the regression equation can be
used. The sample result in this example is 72 μg/L (see Figure 7.13).
The results have to be examined closely to verify that they are reliable.
The absorbance of the highest spike must be below the saturation level on
the regular calibration curve for the method. The spike results should bracket
the true value of the sample to ensure that an extrapolation of no more than a
factor of 2 is being attempted. The absorbance of the unspiked sample
should lie very close to where the extrapolated regression line crosses zero
concentration. Failure to meet all these points should suggest to the analyst
that the multiple standard addition procedure for this sample may need to be
repeated or that the matrix interference is so severe that the method for
determination of the parameter is unusable.

6.0 FREQUENCY OF CALIBRATION AND

CALIBRATION CHECKS
For many procedures, once the test is calibrated, it will remain in calibration
for a long period of time. There are other procedures that require
recalibration every time the test is performed because of sensitivity to
environmental variations or an inherent lack of instrument stability.
Procedures that require daily calibration are generally specified as such in
the method.

TABLE 7.3 Multiple standard addition for a phosphorus test.

Assuming we are dealing with a procedure that remains stable over

several days or weeks, there are guidelines as to when to recalibrate. If there
is any change in the instrument, such as a preventive maintenance or
cleaning, replacement of a light source or sample cell, or adjustment of the
photodetector, then recalibration is needed. If there is any change in the
reagents used in the procedure, such as preparation of fresh reagent or
standard solutions or purchase and use of a new manufacturer’s lot number
of reagent, then recalibration is needed. New analysts performing the test
should do their own calibrations because no two analysts perform a
procedure in exactly the same way. If the instrument is moved to a new
location, it must be recalibrated. An environmental change, such as changes
in indoor humidity, lighting, or temperature, may require recalibration. If
none of the listed changes have occurred (and, in most laboratories, one or
more will have), then the procedure still needs to be completely recalibrated
at least once a year.
FIGURE 7.13 Multiple standard addition calibration for reactive
phosphorus.

The best way to monitor a calibration is to perform a calibration check

every time the method is performed. This consists of choosing one of the
calibration standards and running it as a sample. The analytical response of
the calibration check should be close to that of the original calibration. Any
other result indicates that the procedure may be out of calibration and a new
standardization needs to be performed. How close the calibration check
should be to the original response depends on, first, the specified
requirements in the method and, second, the analyst’s desires to produce
good data. Many methods will specify that the calibration check must be
within a certain percentage of the response of the original calibration (e.g.,
95 to 105%). The size of the allowed variation is a function of the ability of
the test procedure to produce accurate data when it is newly calibrated, the
stability of the calibration, the amount of time required for recalibration, and,
finally, the quality of the data required on a daily basis.
7.0 REFERENCE
American Public Health Association; American Water Works Association;
Water Environment Federation (2012) Standard Methods for the
Examination of Water and Wastewater, 22nd ed.; American Public Health
Association: Washington, D.C.
 8
Test Procedures

1.0 DISTILLATION
2.0 FILTRATION
3.0 GRAVIMETRIC DETERMINATIONS
4.0 COLORIMETRIC DETERMINATIONS
5.0 TURBIDIMETRIC DETERMINATIONS
6.0 TITRATION
7.0 pH AND ION-SELECTIVE ELECTRODES
8.0 REFERENCE

In water and wastewater laboratories, many test procedures use techniques

that have long analytical traditions, but have been modified to meet the
information needs of the treatment plant. Distillation, filtration, and
gravimetric analysis are time-honored names with modern interpretations.
While fractionation columns, asbestos fiber mats, and ashless paper were
essential in the traditional forms of these methods, they are now useless in
most water and wastewater laboratories. This is not to say that the older
cautions and considerations about the techniques are to be discarded; there
is simply a new list.

1.0 DISTILLATION
Distillation is the heating of a solution until the liquid is boiled to a vapor.
The vapor is condensed back into a liquid on a cold surface and allowed to
drip into a receiver. If the vapors are passed through a tortuous path with a
gentle temperature gradient, separations of liquids with different boiling
points can occur. Although some larger laboratories use this latter aspect of
distillation for the recovery and purification of expensive or hazardous
solvents, for the most part this function is not performed in the treatment
plant laboratory.
Distillation has three uses in the environmental laboratory. The first is
final purification of reagent water (discussed in Chapter 1). The second is
gross separation of a target analyte from interferences in the sample, as in
cyanide, sulfide, ammonia, fluoride, and total phenols procedures. The final
use of distillation is for the concentration of the target analytes from a large
bulk of solvent to a smaller volume, as in oil and grease, trace metals
digestion, and semivolatile organic extraction.
When distillation is used to isolate the target analyte from the sample
matrix, the object is to boil the analyte out of the sample and then trap it.
Cyanide, sulfide, and ammonia are all converted to a gas form by adjusting
the sample pH. The problem is that the gas form may have measurable
solubility in the water. Thus, the sample is heated to boiling and the gas,
along with water vapor, is separated from the bulk liquid. In the case of
cyanide and sulfide, the water vapor content needs to be reduced from the
desired gas stream. This is done by passing the hot vapors over a cold
surface, which allows the water to condense and drop back into the bulk
sample, while a constant air flow causes the desired analyte to be swept into
the receiver. A strong base solution in the receiver traps the cyanide or
sulfide as a nonvolatile salt.
Acid is added to the sample through the tube in the side of the
distillation flask. The distillation flask is heated with an electrical heating
mantle to a rapid boil. The cold finger condenser is cooled from an external
source of cold water; a recirculating chiller provides better results than a
cold water tap, particularly in the warmer seasons. The gaseous analyte is
swept from the distillation flask to the receiver with a vacuum pump. The
receiver is equipped with a gas dispersion tip on the entry tube to separate
the entering vapors into fine bubbles and make the absorption of the
gaseous analyte into the sodium hydroxide solution more efficient. The gas
dispersion tip is porous blown glass, which can be destroyed when stored in
the sodium hydroxide solution for prolonged periods of time. The keys to
successful use of the apparatus for good recovery of sulfide or cyanide are
filling the flask less than half full, a rapid boiling of the sample, sufficient
acid to reduce the pH to less than 2, efficient cooling of the condenser, and,
finally, a rapid stream of gas through the system, but not so fast that the
trapping solution is blown out of the receiver.
A recent development in cyanide stills is to miniaturize the equipment
and reduce the distilled sample size. These multiple distillation units, which
can handle up to 12 distillations at once, come complete with heat sources
in the base, vacuum available through a manifold, and timers for automated
venting and cooling (Figure 8.1).
The cyanide still and the ammonia still are essentially the same
apparatus, with the difference being that the vent pipe in the sample flask is
blocked off and is not used in the ammonia still. The ammonia still (Figure
8.2) is designed so that the boiled water is condensed and collected separate
from the bulk sample. The ammonia is collected along with the condensed
water in a pH-adjusted trap. Cold water is run through the condenser to
assist in the condensation of the distillate. The tip of the condenser is placed
underneath the liquid in the receiver to ensure trapping all the ammonia in
the buffer solution. Caution must be exercised when the distillation is
complete. When the still pot is no longer heated, the system begins to cool
and the hot gases contract. This results in the contents of the receiver being
sucked back up the condenser and into the still pot, ruining the sample. To
prevent this, the receiver should be removed from the tip of the condenser
as soon as the heat is shut off.
FIGURE 8.1 Cyanide distillation system. Reprinted with permission from
DWK Life Sciences.

An automated version of the ammonia still, which includes a gas purge

with steam, is available. The automated still also can be used for the final
step in the total Kjeldahl nitrogen digestion and distillation. It adds a
neutralizing agent to a digested sample and follows the addition with a
steam and gas purge (Figure 8.3).
FIGURE 8.2 Ammonia nitrogen distillation apparatus. Reprinted with
permission from DWK Life Sciences.

The fluoride distillation apparatus looks similar to the ammonia still, but
the fluoride is isolated through a different series of steps. The still pot is
initially filled with concentrated sulfuric acid and contains a few soft soda
glass beads as boiling stones. The sample is added to the sulfuric acid. The
fluoride in the sample is turned by the acid into hydrofluoric acid, which
reacts with the soft glass to form hydrofluorosilicic acid. This is distilled
out of the sulfuric acid solution along with the water from the sample. The
procedure in Standard Methods for the Examination of Water and
Wastewater (APHA et al., 2012) (SM22 4500-F-B) recommends a
condenser that is not available commercially. Our laboratory has used
Graham condensers on fluoride stills for many years with no difficulty,
despite the specific prohibition in the method.
The total phenols distillation apparatus can look like the ammonia still;
however, a third principle of operation, azeotropic or steam distillation, is at
work. An azeotrope is a combination of two or more components that boil
at a lower temperature than the pure components alone. Steam distillation is
a particular form of azeotropic distillation in which water is the significant
component. It can be supplied to the sample as steam from a separate
generator, commonly introduced through a side arm. Or, the steam may be
generated by simply boiling the sample itself.
FIGURE 8.3 Total Kjeldahl nitrogen (TKN) automated still. Courtesy of
Labconco.

A specialized trap called a Dean-Starke trap is used for azeotropic

removal of water from a sample for water determination. The sample is
mixed with toluene, then rapidly heated. The water and toluene azeotrope
liquefies in the condenser and falls into the trap. The water, being heavier
than the toluene, settles to the bottom of the trap, and the excess toluene
flows back to the still pot. Some trap designs have a stopcock at the bottom
for removal and measurement of the collected water.
Distillation is also used to remove excess solvent from a target analyte.
Examples of this are oil and grease and semivolatile organic analyses. A
volume of an organic solvent such as hexane or methylene chloride is used
to isolate the target analyte from the water matrix, then the solvent volume
must be reduced or entirely removed before analysis. In these distillations,
it is the residue that is of interest, not the distillate. The distillate may be
collected for recycling or proper disposal, but quantitative recovery is not
the main objective. The various apparatus configurations used are a
stillhead/condenser combination attached to a flask, a Kuderna-Danish
concentrator (Figure 8.4), a TurboYap® solvent concentrator (Zymark)
(Figure 8.5), and a vacuum rotary evaporator.
The conventional stillhead/condenser combination is easily attachable to
the boiling flask and allows temperature monitoring of the distillation
process. It is ideally suited to recovery of the distillate; however, it is
somewhat space-consuming if a number of units are used. A vacuum source
can be attached to the end of the condenser for faster distillations. The
maximum rate of distillation is determined by the condensed solvent
flooding the stillhead.
FIGURE 8.4 Kuderna-Danish concentrator. Image courtesy of Corning
Incorporated, Life Sciences.

All stills should be operated with boiling stones, or ebulators, in the

boiling flask. These may consist of a small stick of wood (like the handle of
a cotton swab or applicator stick), glass beads or tubes, ceramic pieces, or
Teflon® chips. A magnetic stir bar can be used if the heat source can be
adapted to contain a magnetic stirrer. The purpose of the ebulator is to
provide irregularities in the boiling flask to initiate boiling (bubble
formation) and prevent superheating. The wood sticks contain trapped air,
which expands and is released upon heating; however, the wood can also be
a significant source of contamination. Superheating of the boiling flask
results in spontaneous formation of large bubbles of solvent vapor that drive
hot liquid up and out of the boiling flask, flooding the head of the still and
the condenser. This is called bumping, and is to be avoided.

FIGURE 8.5 TurboVap II® Automated Solvent Evaporation System.

Courtesy of Biotage.

Almost every technician has watched in dismay as a favorite magnetic

stir bar falls out of a flask and disappears down the drain in the sink.
Although a magnet on a stick can sometimes be used to fish the bar out of
the drain, more often than not the stir bar is gone forever. This problem can
be avoided by purchasing a polyethylene colander from the grocery store
and always decanting flask residues through the colander. The colander is
self-supporting on its molded feet in the sink. A large ceramic Buchner
funnel can also be used for this purpose; however, these funnels are
breakable and can shatter a flask if it slips out of wet hands into the funnel.
The Buchner funnel can be supported in a ring clamped to the neck of the
faucet and pivoted out of the way for washing glassware. This technique
also works well for recovery of Teflon® chips used as boiling stones.
Soaking them in concentrated sulfuric acid followed by a thorough rinsing
with reagent water cleans them sufficiently for reuse.
The Kuderna-Danish concentrator (Figure 8.4) is designed to remove
low boiling solvents from large volumes of sample, while at the same time
retaining target analytes that are somewhat volatile. It consists of a three-
ball fractionation column, called a Snyder column, a boiling chamber, and a
receiver. The unit is assembled with sample filling up to half the volume of
the boiling chamber; then, boiling chips are added, and the apparatus is
placed in a water bath so that the receiver is immersed in the water bath and
the bottom portion of the boiling chamber is exposed to steam from the
water bath. Proper operation of the unit calls for the balls in the Snyder
column to be rapidly bouncing up and down (“chattering”) in a pool of
condensed solvent from vapor. The liquid in the column traps the volatile
components from the sample and returns them to the boiling chamber.
Most losses of the volatile analytes arise from starting the concentration
without any liquid surrounding the balls in the Snyder column. It is
important to control the temperature of the water bath that heats the unit. If
it is too hot, bumping can be quite a problem, even to the extent of
separating the receiver from the boiling chamber, which dumps the whole
sample into the water bath. The analyst should never attempt to reduce the
solvent volume to the exact desired final volume in a Kuderna-Danish unit.
A preferable procedure is to halt the concentration when the boiling
chamber becomes empty, remove the receiver, and further concentrate the
sample under a gentle stream of nitrogen gas.
An alternative to the Kuderna-Danish concentrator uses a rapid flow of
gas to remove solvent from the sample. The technique works best for highly
volatile solvents, such as hexane and methylene chloride. The sample is
immersed in a warm water bath at or slightly above the boiling point of the
solvent; the water bath should never be as hot as that required for operation
of a Kuderna-Danish concentrator. The main purpose of the water bath is to
replace the heat lost by the system from the rapid evaporation of the solvent
by the gas stream. Rapid volume reduction is possible with minimal loss of
analyte from the sample. The solvent vapors from the unit can be ducted to
a condenser for solvent recovery for recycling or proper disposal. Boiling
stones are not required for use with the TurboVap® system (Figure 8.5)
because the solvent never achieves a boiling temperature.
Another technique for concentrating samples uses a rapidly spinning
boiling flask immersed in a warm water bath with a mild vacuum to rapidly
remove solvent (Figure 8.6). The spinning receiver spreads the solution in a
thin layer over the internal surface of the flask, which increases the
evaporation rate and prevents bumping. Addition of boiling stones to the
flask can actually cause bumping by interfering with the smooth, thin layer
of solvent spread over the flask. A distillation trap (Figure 8.7) can be
attached between the sample flask and the rest of the unit to prevent any
inadvertent bumps from reaching the cooling coils of the evaporator.
Keeping the cooling coils of the rotary evaporator at a low temperature
through use of a recirculating chiller while at the same time cooling the
solvent receiver can result in efficient recovery of the solvent and provide
for rapid volume reduction of the sample.
FIGURE 8.6 Vacuum rotary evaporator. Courtesy of Buchi Corp.
FIGURE 8.7 Distillation trap. Courtesy of Laboyglass.

The boiling flask or sample container is heated by either a heating

mantle (Figure 8.8) or by a water bath (Figure 8.9). The water bath allows
more control over the maximum temperature of the heat source because of
the boiling point of water, which is an important consideration if the residue
is to be taken to dryness. Heating mantles without a thermocouple
temperature-monitoring probe can result in empty flasks reaching
temperatures of more than 300 °C. On the other hand, heating mantles are
dry sources of heat and their use avoids the problem of splashes and high
humidity that the water baths present. Heating mantles are oriented toward
application of individually controlled heat to a single boiling flask for
applications where high boiling rates are desired or the solvent, such as
water, has a high boiling point. The water bath is capable of supplying the
same even amount of heat to a number of units at the same time.
Cooling the condenser of the distillation unit can present a number of
problems. Water from the tap is a waste of good drinking water (on par with

FIGURE 8.8 Heating mantle and temperature controller. Copyright©

Cole-Parmer.
FIGURE 8.9 Water bath. Reprinted with permission from Poly Science.

using tap water to wash your driveway or water the lawn) and is not cool
enough to work effectively, especially for low-boiling solvents or chained
cooling towers. A recirculating chiller is the correct device for laboratory
cooling needs. Use of a 1:1 mixture of ethylene glycol and water allows the
temperature range of the cooling liquid to go as low as −25 °C, which gives
very effective cooling. The high heat capacity of this mixture presents the
opportunity to attach up to eight condensers in a series for cooling multiple
units. The outlet (top port) of one condenser is attached to the inlet (lower
port) of the next condenser in the chain. All connections of tubing to
glassware must be firmly attached with metal wormscrew tubing clamps.
The pumping force of the recirculating pump will spray the cooling mixture
up to 50 ft if a tube comes off, making quite a mess throughout the
laboratory. Further, the unit will empty itself in about 15 seconds, which can
cause a rapid burnout of the chiller.

2.0 FILTRATION
The most common tests involving filtration in the treatment plant laboratory
are membrane filtration coliform counts, solids determinations, and sample
preparations that require digestions, such as total phosphorus or trace metals
procedures. The general idea behind filtration is to separate solids from
liquids. A recently introduced technique, solid-phase extraction, also has
many of the same considerations as filtration, although the purpose is
different. How the filtration is performed is, to a large extent, dependent on
the size of the desired and undesired materials. Size in filtration is measured
in micrometers (μm, one millionth of a meter), and generally refers to the
diameter or largest dimension of the particle if it is not round; for instance,
0.001 in. is equivalent to 25.4 μm. The smallest visible particles are about
40μm in diameter. Light microscopes can see down to about 0.2 μm.
Bacterial cells range in size from 0.3 to 10 μm. Yeast and fungi cells range
from 0.6 to 4.0 μm. Virus particles range from 0.004 to 0.08 μm. Dissolved
metal ions are smaller than 0.002 μm. General filtration will handle sizes
down to 10 μm, microfiltration covers from 0.1- to l0-μm particle sizes,
ultrafiltration concerns the range from 0.001 to 0.1 μm, and less than 0.001
μm is the realm of reverse osmosis.
Modern filtration media have proceeded quite a bit beyond the older
methods of formation of asbestos mats in the bottom of Gooch crucibles.
The main classes of disposable filter media used are membrane, glass fiber,
and paper.
Membrane filters are thin sheets of filter media that appear to have a
solid surface. They can be subdivided into screen and depth filters. A screen
filter is a smooth-surfaced thin sheet of impervious material (commonly
polyester or polycarbonate) with uniform-size pores randomly distributed
over the surface. Screen filters are manufactured by exposing the sheets of
the screen to ionizing radiation from a nuclear reactor. The duration of the
exposure determines the number of pores per unit area. The sheet is then
treated with a strong alkaline solution, and the damage tracks from the
ionizing radiation are preferentially dissolved. The length of time in the
alkaline solution determines the diameter of the pores.
Depth filters are characterized by a rough surface with a random size
distribution of openings. The openings extend into the body of the filter
with a large number of twists and turns, which result in an absolute pore
size for materials passing through the filter. A large number of materials are
used for fabrication of depth filters, such as silver, polypropylene,
polysulfone, polyethersulfone, polyvinyl chloride (PVC), the various
fluorocarbon resins, nylon, cellulose acetate, and nitrocellulose. These
materials (except for the fluorocarbon filters) are created by evaporating the
solvent from a solution of the filter material under controlled conditions of
rate and temperature. The fluorocarbon filters are formed by stretching the
initial cast sheet of membrane to create the desired pore sizes.
The main differences between the two types of membrane filters are as
follows:

• The screen filter will efficiently pass most particles that have a
smaller size than the pore openings, whereas the depth filter will
retain a higher proportion of particles that are smaller than the rated
pore size;
• The depth filter will lose fibers or particles of filter material and
possibly deform under pressure or vacuum filtration;
• The screen filter has the lower particle capacity before plugging;
and
• The depth filter has the higher effective surface area.

Another important property of the membrane filters is their behavior

toward water and other solvents. Filters that become wet and pass water
with ease are termed hydrophilic. Hydrophobic filters reject being wetted
with water and, in some cases, can be used to remove traces of water from
the filtered solution. The fluorocarbons are examples of hydrophobic
materials. Attempting to filter a water solution through a hydrophobic
membrane that has not been properly prepared can be a difficult, if not
impossible, task.
Few membrane filters have a tolerance for heat, especially at oven
temperatures. Their normal reaction is to shrivel into a ball, and some
materials simply melt. For procedures that require the membrane to be
sterile, autoclavable membrane filters are available.
However, the more common solution is to purchase presterilized filters.
Sterilization is done by the manufacturer by exposing the filter to either
ethylene oxide or gamma radiation. Ethylene oxide can react with
impurities in the filter or with the membrane itself to form toxic
compounds. It is prudent for the analyst to check each manufacturer’s lot
number of presterilized membranes for sterility (by culturing a blank of
filtered dilution water) and for inhibitory characteristics, which means that
growth of cultures is slowed or prevented. Inhibitory checks are made by
filtering a sample known to contain the target organism, such as
Escherichia coli, and then seeing if the microorganisms grow on the
membrane.
Another property of the membranes that needs to be considered is their
compatibility with different solvents. Manufacturers of membrane filters
provide solvent compatibility charts in the back of their catalogs, and these
should be checked. The membrane filter holder must also be compatible
with different solvents. Many filters come prepackaged in a holder that
attaches to a syringe, hose, air pump, or other device. The membrane itself
may be tolerant of the solution; however, the holder may interact with it in
unexpected ways. These concerns also work in the other direction, an
example being the sampling of airborne particulates for trace metals. These
are typically collected on a membrane filter, then the filter is dissolved in
hot nitric acid to prepare a solution suitable for analysis. The mixed
cellulose ester membranes readily dissolve in acid and are designed for this
application. They come prepackaged in a plastic air-sampling cassette that
attaches to the air-sampling pump. However, PVC membrane filters also
exist that are available in the same style of air-sampling cassette and are
designed for sampling total particulates in air. Attempted digestion of the
PVC membrane in hot nitric acid for trace metals analysis leaves a gooey
mess of melted plastic mixed with acid in the beaker.
Glass fiber filters are made of strands of borosilicate glass compacted
into a flat disk. These work as depth filters and are rated by an approximate
minimum size particle retention, such as 5 μm. The glass fiber filters have
all the properties of glass in that they are autoclavable, resistant to many
chemicals (except solutions of alkali or fluorides), inert, and temperature-
stable to about 550 °C for brief periods of time. They are fragile and tend to
fall apart upon attempted folding or other rough handling. To improve their
mechanical strength, some glass fiber filters have organic polymer binders
mixed into the filter.
The glass fiber filters have rapid flowrates and do not plug easily. Not
only are they used by themselves for many filtration analyses, such as total
suspended solids (TSS), they also find use as prefilters for membrane
filtration applications. In this capacity, they serve to keep the large
particulates in the sample away from the surface of the membrane and help
prevent surface plugging of the fine filter. In practice, the membrane filter is
placed directly on the support of the filter holder, then a glass fiber filter of
the same diameter is placed over it. The reservoir of the filter holder is then
added, and the whole clamped together. For difficult samples, a stack of
glass fiber filters might be added over the membrane disk with the coarsest
filter on top, followed by layers of increasingly fine filters down to the
membrane. Another alternative for difficult-to-filter samples is to increase
the diameter of the filters. A doubling of the diameter of the filter increases
the available area for filtration by a factor of 4, with a similar increase in
filtrate throughput. Commonly available sizes are 25, 37, 47, 90, and 142
mm, each with its own appropriately sized filter holders.
Paper filters have been the traditional filter media used in laboratories
for decades. They are provided in a variety of grades, sizes, and ratings. The
main classes are qualitative and quantitative. The qualitative papers are the
most useful in the treatment plant laboratory because most of the functions
of the quantitative papers have been taken over by the glass filter and
membrane filters. Qualitative filters are designed for rapid filtration of
samples to remove particulates. They are available in a number of particle
retention ratings, as indicated in Table 8.1.
TABLE 8.1 Particle retention ratings for a variety of qualitative filter
papers.

Paper filters are used as the disposable filter media in glass funnels and
in ceramic Buchner funnels. The proper size of filter paper for the Buchner
funnel is the internal diameter of the bowl. The correct size paper for a glass
funnel can be determined by measuring the distance from the rim to the
attachment of the stem of the funnel, then doubling the number and
rounding to the next lower whole number. Another technique is to measure
the diameter of the bowl opening of the funnel and to double it.
If the paper is folded in half and in half again (quarters), then opened
and placed into the funnel so that it is in contact with the sides of the funnel
in all spots, the only active site of filtering is at the very tip of the folded
paper. The rate of filtration can be increased by using a funnel with vertical
raised ribs or cutout slots on the inside of the bowl. These funnels serve to
break the contact of the paper with the sides of the bowl and allow more
surface area to be used for filtration. The alternative is to fold the paper so
that it sits away from the sides of the filter bowl. This is called fluted paper
and can be purchased already folded.
All filters, regardless of type, have the potential to contaminate a
sample. Filter paper is often overlooked as a significant source of sample
contamination, particularly in the organics laboratory. The term, leachable,
refers to materials that can be dislodged from the filter and end up in the
filtrate. The leachable materials may be solid fibers or particles from the
filter, or they may be dissolved from the filter into the liquid of the sample.
All filters must be pretreated before use. This may consist of simply
running a portion of reagent water or solvent through the filter and
discarding the filtrate before the sample is filtered, or it may be more
involved. An example is TSS analysis, in which the filter is taken through
the entire analytical process filtration—drying in the oven, cooling in the
desiccator, and weighing before the filter is used for a sample. If a volatile
solids determination is to be made, then the filter must also be ashed in the
muffle furnace as part of the pretreatment. This is especially important if
there are organic binders in the filter. Not taking them into account in the
filter tare weight can result in lower final weights of filter plus residue than
the initial filter weight alone. In addition, organic binders are burned out of
the filter at 550 °C, with the filter itself becoming fragile and difficult to
handle. The analyst should avoid using the glass fiber filters with organic
binders if ashing is part of the regular analytical procedure.
Blown glass filters, called fritted glass, are available permanently
attached to the bottom of the funnel or at the ends of gas dispersion tubes.
These filters come in a variety of porosities, termed coarse (40 μm),
medium (10 to 15 μm), fine (4 to 5.5 μm), very fine (2 to 2.5 μm), and ultra
fine (0.9 to 1.4 μm). The direct filtration of material through the fritted
funnels is ideal for certain applications. If they plug or become dirty, they
can be reverse-direction washed under either mild vacuum or pressure (up
to 15 psi) with soap or strong acid solutions, rinsed with lots of reagent
water, dried, and reused. Because caustic solutions will dissolve the fritted
filters, they should not be exposed to or stored in caustics for any length of
time. For the Buchner-style funnel, the coarse, fritted filter can be used as a
support for a disposable paper or glass fiber filter. Fritted filters should not
be cleaned by scrubbing because the glass is sensitive to abrasion. Use of
metal scrapers for collection of filtered particulates is to be avoided unless
extreme caution is taken. While teaching a freshman quantitative laboratory
class, I had the experience of watching a student get a 500% yield of a
barium sulfate precipitate while scraping it out of a fritted filter funnel with
a stainless steel scoop. He complained, “It was really stuck in there”.
 Filtration can be performed using gravity to pull the liquid through the
filter; however, in most cases, a more rapid filtration is desired. This extra
force is applied either as pressure above the sample (pressure filtration) or
as a vacuum below the filter. Vacuum filtration is the more common
approach for liquids that are not volatile; however, for liquids such as
hexane, ether, or methylene chloride, a strong vacuum can result in
complete evaporation of any filtrate.
Vacuum filter flasks are thick-wall flasks with a side arm for attachment
of a hose to the vacuum source. Direct glass-to-glass contact must be
avoided by placing rubber or other cushioning material between the filter
funnel and the flask to avoid chipping the rim of the flask. The cushions can
take the form of a one-hole stopper on the stem of the funnel or a hollow
cone or flat gasket placed on the rim of the flask. The cushions also allow
an airtight seal between the funnel and the flask (Figure 8.10).
A laboratory vacuum source that has been used for many years is a
Venturi-type water aspirator that attaches to a water faucet. Good vacuums
(10- to 30-mm Hg, depending on the water temperature) result from this
device, but it is a tremendous waste of potable water. Although hand-
operated vacuum pumps are available and useful for filtrations in the field,
the more common source of vacuum is an electric vacuum pump. These
electric pumps use oil as the transfer medium for creating the vacuum. The
pumps are rated first by the maximum vacuum that can be produced, second
by the capacity, and third by use. The maximum vacuum achievable by the
pump is limited by the vapor pressure of the pumping fluid, typically an oil,
and is measured in millimeters of mercury, also called torr. The capacity (or
displacement) of a vacuum pump, which is the amount of air it can pump, is
the important consideration in the laboratory. Capacity is measured in liters
per minute (L/min) or in cubic feet per minute (cfm). Operation of a four-,
six-, or eight-station vacuum filtration manifold is going to take a lot more
capacity than operation of a single vacuum filter flask.
The intended use of a vacuum pump will dictate the purchase of a
corrosion-resistant, solvent-resistant, or explosion-proof model, as well as
the choice of accessories. No pump should be directly connected to a
vacuum flask because no pump is designed to have liquid filtrate sucked
into the vacuum intake. A trap must always be placed between the pump
and the vacuum flask. The trap may be a second vacuum flask or it may be
a specially designed vacuum trap. For large-capacity vacuum manifolds, a
19-L (5-gal) flask for the filtrate would still be followed by a smaller trap.
When volatile organic solvents are being filtered, they will evaporate and
end up in the pumping oil, thereby diluting it. Two traps may be needed, the
first cooled to catch the vapors and the second empty to catch any overflow.
For solvents such as methylene chloride or hexane, a trap cooled with dry
ice and acetone, dry ice and toluene, or liquid nitrogen will effectively trap
the vapors and protect the pump. The vacuum traps (Figure 8.11) are
designed to be placed in long, narrow Dewar flasks (Figure 8.12) for
holding the cooling mixture. When Dewar flasks are used, the vacuum trap
must be securely clamped in place to avoid touching the insides of the
Dewar, which often results in an implosion of the Dewar. When vacuum
traps are used, they should be checked and emptied daily before any use of
the vacuum pump.
FIGURE 8.10 (a) Cone. Reprinted with permission from Thermo Fisher
Scientific. (b) Filter support with gasket. Reprinted with permission from
DWK Life Sciences. (c) Stopper. Reprinted with permission from Thermo
Fisher Scientific.

Another consideration in using the vacuum trap is that, if the pump

stops while a vacuum exists in the system, the oil from the pump will be
sucked out of the pump and up into whatever is under vacuum. Vacuum
pump oil is one of the nastier materials to try to remove from the inside of
glass apparatuses and from samples. The vacuum should always be released
before the pump being stopped. A three-way valve between the vacuum
trap and the pump, with one leg open to the atmosphere, is a useful
accessory for venting the system before pump shutoff. Another advantage
of the valve, particularly when a manifold is in use, is that it allows the
manifold to return to atmospheric pressure while still connected to the
filtrate catch trap, at the same time maintaining the load on the vacuum
pump.
FIGURE 8.11 Vacuum trap. Reprinted with permission from Menglab,
Inc.
 Other useful accessories include a vacuum gauge for monitoring the
vacuum of the pump and a smoke eliminator or an oil trap for the outlet side
of the pump to reduce air contamination from oil droplets being sprayed out
of the pump. Vacuum tubing is essential for connecting the various parts of
the vacuum train. This is comprised of thick-wall tubing of rubber or one of
a variety of synthetic materials. The use of the vacuum system will dictate
which tubing is appropriate: solvent-resistant if solvents are being filtered
or corrosion-resistant if acids or caustics are filtered. Vacuum tubing should
be checked every day for leaks and replaced on a regular preventive
maintenance schedule, such as monthly.

FIGURE 8.12 Dewar flasks. Reprinted with permission from Pope

Scientific, Inc.
 The idea behind solid-phase extraction is to isolate target analytes from
a liquid sample by adsorption on a solid material while at the same time
removing the bulk of the liquid, then selectively removing the target
analytes from the solid into a different solvent for analysis. Operationally,
the technique is similar to a filtration, with the filter media being replaced
with an adsorbent. Two designs for solid-phase extraction are commercially
available: column and disk. The column design looks like a plastic or glass
disposable syringe barrel filled with the solid adsorbent (Figure 8.13). The
disk design looks like a thick membrane filtration disk (Figure 8.14).
The preparation of the disk or column is crucial to the success of the
technique. Improper preparation can result in no recovery of the desired
analytes and in massive contamination of the sample because of leachable
materials from the adsorbent. Manufacturer’s directions for the preparation
and use of the media should be closely followed.
FIGURE 8.13 Solid-phase extraction apparatus for oil and grease.
Reprinted with permission from Xenosep Technologies.

Apparatuses for the solid-phase extractions typically have limited

sample capacity in the reservoir above the adsorbent. The analyst can
continuously add more sample to the reservoir as the liquid is pulled
through the adsorbent; however, an automated sample introduction is more
efficient. This is achieved by placing the sample in a large separatory
funnel, tightly putting the stopper in the top of the funnel to achieve an
airtight seal, and then hanging the funnel over the reservoir so that the tip of
the funnel is below the top of the reservoir. The stopcock to the funnel is
opened, and the sample flows into the reservoir. When the sample level in
the reservoir reaches the tip of the funnel, the liquid prevents air from
entering the tip of the funnel and the weight of the liquid still in the
separatory funnel creates a vacuum in the top of the funnel, which prevents
more sample from flowing into the reservoir. When the liquid level in the
reservoir drops below the tip of the separatory funnel because of sample
being pulled through the adsorbent media, air enters the funnel, allowing
more liquid to flow into the reservoir until the liquid level is again above
the tip of the funnel. This continues in a controlled fashion until all the
sample has been added to the reservoir.
FIGURE 8.14 Solid-phase extraction disk. Reprinted with permission
from Xenosep Technologies.

Once the sample has passed through the solid-phase extraction media,
the next step in the process is to elute the analytes from the adsorbent. With
all the time spent separating the analytes from the aqueous matrix, it is not
desirable to empty filtrate from the vacuum flask, then elute the sample into
the dirty flask. One solution is to use a second clean vacuum flask. Another
is to empty the original vacuum flask, then add a clean large test tube to the
flask to catch the elution solvent.
3.0 GRAVIMETRIC DETERMINATIONS
Gravimetric determinations were once the mainstay of analytical chemistry.
Combined with melting-point determinations and chemical derivatizations,
they formed the primary means of compound identification and a
significant form of quantitative analysis. Now, we measure the residue
content of samples by gravimetry. Some of the technique concerns remain
the same; however, others have changed. The steps in a solids or residue
analysis consist of first establishing the tare weight of the container,
followed by filtering, drying, cooling, and weighing the residue plus
container. Subtracting the container tare weight from the final weight gives
the residue amount.
The exact equipment used will be determined by which residue fraction
is of interest. The choices and requirements are shown in Table 8.2.
TABLE 8.2 Residue fractions determined in the laboratory.

The general choice of weighing containers is limited to glass fiber filters

for the nonfilterable fractions (TSS and total volatile suspended solids) and
ceramic or Vycor® (quartz) dishes for the other fractions, assuming that the
volatile fraction is a common determination, as it is in most laboratories. If
volatile fractions are not determined, more choices are available because the
laboratory ware does not have to survive a 550 °C muffle furnace.
Drying and cooling are conducted in ovens and desiccators,
respectively. A kitchen oven generally does not have the temperature
stability that is required for a laboratory oven. The drying of glassware and
drying of samples for residue determinations should be performed in
separate ovens. Totally sealed ovens are not desirable for residue
determinations because of the large amount of water vapor that is driven off
the samples. Two common types of ovens have the capability to remove the
water vapor: gravity flow (gravity convection) and forced air (mechanical
convection). The forced-air oven has a fan that circulates air around the
samples and allows faster drying. Samples that produce light, fluffy
residues may present difficulties from the residues blowing away. Vacuum
ovens that reduce the pressure inside the oven while heating the sample are
also available. Although fast drying can occur in these ovens, the residue
results may be different from those obtained from atmospheric pressure
ovens. The oven interior should be constructed of stainless steel to prevent
corrosion. If liquids other than water are being removed from the samples,
an explosion-proof oven may be required for flammable solvents.
Temperature control in ovens can be performed with either a liquid-in-
glass thermometer or a thermocouple built into the oven’s temperature
control system. The tip of the liquid-in-glass thermometer should be buried
in a heat sink, such as a beaker of sand or nonvolatile liquid (silicone oil), to
avoid temperature fluctuations caused by opening the door of the oven. The
thermocouple reading should be checked against a liquid-in-glass
thermometer at regular intervals (a minimum of once a year). The
temperature of the oven should be recorded on the benchsheet at the
beginning and end of each drying period.
A muffle furnace is required for the ashing part of solids analysis.
Ashing depends on the presence of oxygen, and a forced air furnace can
provide a fast ashing, on the order of 10 minutes or less. The furnace only
has to heat to 550 °C, and added temperature range (such as 1700 °C) is a
waste of money. The temperature on the muffle furnace needs to be checked
by an independent thermocouple at regular intervals not to exceed a year.
Drying occurs in the oven, not in the desiccator. The purpose of the
desiccator is to provide a dry atmosphere in which the residue can cool to
room temperature before weighing (Figure 8.15). The aim is to avoid
having the residue absorb moisture during cooling. Taking an oven-dried
residue and placing it in a desiccator for long periods of time can actually
accomplish the opposite purpose, particularly when the residue has a
stronger affinity for water than the drying agent in the desiccator. The
drying agents reduce the humidity in the desiccator by either reacting
irreversibly with the water vapor, as is the case with phosphorus pentoxide,
which forms phosphoric acid with water, or by absorbing water to form a
hydrate, which is the action of calcium sulfate (Drierite®) or magnesium
perchlorate. Drierite® is a commonly used drying material because it comes
with an indicator, which is blue when the desiccant is good and red-pink
when its drying capability is exhausted. It is not efficient, however, and its
capacity is low. When the indicator has turned pink, the atmosphere inside
the desiccator is really wet. Table 8.3 presents some of the characteristics of
the drying agents; however, it does not present data on the rate of drying. In
general, irreversible reagents
FIGURE 8.15 Glass desiccator. Reprinted with permission from DWK
Life Sciences.

dry considerably faster than reversible agents. When determining solids in

the treatment plant laboratory, this is a key consideration because of the
samples’ short residence time in the desiccator.
TABLE 8.3 Characteristics of drying agents for potential use in a
desiccator.

Analysts commonly make the mistake of removing dried samples from

the oven, placing them on a tray, and walking into the next room to put
them in the desiccator. Instead, the desiccator must be immediately adjacent
to the oven or furnace for quick transfer of the item while it is still hot.
Once a hot item is placed in the desiccator and it is closed, a competition
for the water vapor that entered the desiccator while it was open begins
between the hot item and the desiccant. As long as the item is warm, the
desiccant has the advantage. As soon as the item is cooled, it should be
weighed, then placed back in the oven for another cycle of heating, cooling,
and weighing. If the desiccators cannot be placed immediately adjacent to
the ovens and the analytical balance, a reasonable solution is to place the
desiccators on a rolling cart for transfer of the filters or crucibles from the
ovens to the balance. This works well when the ovens and balances are in
different rooms.
A further area of concern in determination of residues is laboratory
contamination, which results in falsely high or low values. Contamination
that results in low values has been attributed to the organic binders used in
some glass fiber filters in the volatile solids procedures and can be avoided
by properly preparing the sample container before analysis. Contamination
that gives high results can come from touching the weighing containers
with fingers, gloves, or dirty tongs or forceps. Fingerprints have weight,
which can be detected with the analytical balance. More massive
contamination can arise from glove prints, which can leave a residue of
powder and lotion. A hot crucible can even melt off some of the glove
material onto the weighing container. Tongs or forceps made of metal
should be the only tools used for direct contact with the weighed item. Even
then, the analyst must be careful not to contaminate the tongs with the
residue from one sample and transfer it to the next touched sample. Placing
a hot crucible on a plastic tray is also a good way to inadvertently pick up
extra mass. The hot items should be immediately transferred from the oven
to the porcelain or stainless steel support in the desiccator and from there to
the balance pan, without any stops in between.

4.0 COLORIMETRIC DETERMINATIONS

Colorimeters are common instruments used for many different analytical
determinations. The colorimeter is used to measure the intensity of a
colored derivative formed in an analytical procedure. If the derivative
appears red, it absorbs most strongly in the blue (low) end of the visible
spectrum, whereas a blue solution absorbs most strongly in the red (high)
end of the spectrum. In general, the complementary color of the solution
color is the measured wavelength. Colorimeters are limited to the visible
light spectrum wavelengths (400 to 900 nm) and are restricted versions of a
general class of instruments called spectrophotometers. A UV-visible
spectrophotometer has the same capability as a colorimeter and is often
used as a more expensive substitute. The added expense arises from the
addition of extra lamps that produce light in the UV part of the spectrum,
special mirrors and lenses that can perform their functions on UV light, and
a photodetector that is sensitive to both UV and visible light.
Colorimeters consist of a power transformer, light source,
monochromator, sample cell, photodetector, and some electronics to convert
the output of the photodetector to absorbance or transmittance units on a
dial or digital display. Each part of the instrument requires attention to
obtain the best, most consistent results.
Although the power transformer is simply either working or not, the
service line to which it is attached should have ample current for all the
instruments connected to it. Everyone is familiar with the situation in the
bathroom in the morning when the hair curlers, clothes iron, and electric
razor are all running, and the lights dim. The same situation occurs in the
laboratory. Indeed, when a number of instruments such as incubators,
refrigerators, water baths, and microscopes are attached to the same circuit
as the colorimeter, there will be fluctuations in the power supply and the
output of the power transformer, and the lamp will dim. A steady power
supply must be available for consistent operation.
The lamp in the colorimeter is the analytical probe of the instrument. As
lamps age, the intensity of the output light decreases slowly until failure
occurs. The decreasing output affects the results, particularly if the same
calibration curve is used for long periods of time. The actual effect seen is a
slight loss of sensitivity on the low end of the calibration and a lowering of
the onset of saturation in the upper end of the curve. However, other than
these effects, the aging of the lamp can be masked by setting the absorbance
of the reagent blank solution to zero. Some of the more expensive
colorimeters include a direct probe of the lamp condition as part of the
startup electronics of the software; however, in most instruments, this is
simply not taken into consideration. The analyst can perform a check on the
lamp intensity by zeroing the absorbance on the empty sample cell, then
recording the absorbance of the reagent blank before resetting the
colorimeter to zero for the calibration curve. Keeping day-to-day records of
the lamp output will provide early warning of lamp failure.
Filter photometers are a class of colorimeters that use filters to block
transmission of light through the sample above and below a desired range
of wavelengths. Changing the wavelength range of operation requires
changing the set of filters. Most colorimeters are continuously variable, but
the wavelength in use should not be thought of as monochromatic (single
wavelength). Rather, a narrow range of wavelengths is passed through the
sample, set by the dispersion of the light into a spectrum and the
narrowness of the entrance slit that is used to select a portion of the
spectrum for use. Typically, these variables are preset by the manufacturer.
In general, the more precise the wavelength range of light available for
analysis, the less the intensity. Most variable-wavelength colorimeters have
wavelength bands of less than 5-nm wide; filter photometers are 30- to 60-
nm wide.
In the colorimeter, the monochromator is either a prism or a diffraction
grating placed in the path of the light from the lamp to separate it into a
spectrum. The desired part of the spectrum is then isolated by focusing it
onto a slit in the sample holder. Various means are used to direct the light of
proper wavelength to the slit. The prism or diffraction grating may move on
a mechanical pivot, or a mirror may be moved. The movement may be
activated by a hand-turned wheel or an electric motor. Most colorimeters
use a hand-turned wheel to dial to a desired wavelength, say, for instance,
880 nm. Few analysts check to see if the selection mechanism is
functioning as desired so that 880 nm is actually the center of the
wavelengths of light passing through the sample cell. The absorbance of
light by solutions is typically a broad band rather than a sharp spike at a
specific wavelength. The specified wavelength in the procedure is typically
the maximum absorbance within the band. This can be checked with the
colorimeter by placing a standard in the sample cell, then recording the
absorbance as the wavelength is changed. The instrument should be zeroed
on the reference at each reading. The maximum absorbance should be
obtained at the method-specified wavelength, as illustrated in Figure 8.16,
for a check on the absorbance of a standard from the hexavalent chromium
procedure (SM22 3500-Cr B). The maximum absorbance for this instrument
is found to be about 5 nm off the correct value of 540, as illustrated in
Figure 8.16, which was prepared from the data in Table 8.4.
FIGURE 8.16 Plot of absorbance versus wavelength for a colorimeter
test.

TABLE 8.4 Absorbance data obtained from varying the wavelength

around a method maximum.

Wavelength (nm) Absorbance

500 0.221
510 0.280
520 0.335
530 0.377
535 0.385
540 0.393
 545 0.398
550 0.393
560 0.368
570 0.325
580 0.274
590 0.205
600 0.144

An alternative is to check a known standard against the literature

maximum absorbance. A readily available set of standards are the pH
indicators used in the laboratory (Table 8.5). They have the added
advantage of providing two wavelength checks for each compound, chosen
by adjustment of the pH.

TABLE 8.5 Some pH indicators for use in checking wavelength

adjustment on spectrophotometers.
 Repeating the wavelength check on a regular basis can also serve as a
check on the stability and condition of the mechanical linkages between the
dial and the pivots. A selection of possible solutions to use that will cover
the range of the colorimeter is found in Table 8.6.
Another check on the performance of the colorimeter is to record the
absorbance for a sample at a particular wavelength, move the wavelength
selector about 100 nm away, then return it to the original setting. The
absorbance should return to the original value. For the colorimeter tested in
Table 8.4, it was found that if the desired 540-nm wavelength was
approached from the low side (500 nm), it was about 5 nm off; however, if
it was approached from the high side (600 nm), it was exactly on the mark
(Table 8.7). This was found to be repeatable, and the preferred direction of
adjustment was determined to be from a higher wavelength.
Some procedures, such as the Nesslerization method for ammonia
determinations, use different parts of the broad-band absorbance to extend
the range of the test. The wavelength of maximum absorbance is used for
the most sensitive test, whereas high levels of ammonia are determined by
use of a wavelength of lesser absorbance.
The sample is the next area of concern. Anything that can interfere with
the transmission of light from the lamp to the photodetector will affect the
analysis. The sample must be entirely free of particulates; turbid samples
give nonspecific absorbances and scatterings that result in reduced light
transmission and falsely high values for the test. Air bubbles in the sample
prevent light from reaching the photodetector and, in some cases, may
actually increase the transmission because of refraction effects. Air bubbles
can be removed from the sample by brief exposure in an ultrasonic bath.
Samples with intrinsic color can exhibit specific absorption at the
wavelength of interest. This is a difficult problem to correct. One procedure
is to measure the absorption of the sample without any of the analytical
reagents in it and subtract this absorption from the analytical sample.
Problems crop up when part of the color is caused by an unanticipated
reaction between the sample and the reagents. In some instances, a sample
blank can be prepared when a key reagent is deleted from the normal blank,
such as occurs in the sodium 2-(parasulfophenyl-azo)-1,8-dihydro-3,6-
naphthalene disulfonate (SPADNS) procedure for fluoride analysis. A
powerful and not very time-consuming technique is to prepare a calibration
curve in the sample through the method of multiple standard additions,
which was discussed in Chapter 7.
Dirt on the inside or outside of the sample cuvette can result in
nonspecific absorbance. Fingerprints and condensation of moisture (from a
cold sample) on the outside of the cell should be removed immediately
before insertion of the cell in the sample holder by wiping the cell with a
clean, lint-free tissue or cloth. A disposable tissue is preferred because
reusing the cloth can transfer dirt from one wipe to the next.

TABLE 8.6 Methods that result in wavelength standards for checking

colorimeter.

Substance Method Wavelength (nm)

SPADNS-zirconium mixed SM18 4500-F- D 570
reagent
Indophenol SM18 4500-NH3 630
D
Molybdenum blue SM18 4500-P E 880
DPD SM18 4500-Cl G 515
Zincon zinc SM18 3500-Zn F, 620
SM22 3500-Zn B
Dithizone SM18 3500-Zn D 620
Dithiazone-zinc SM18 3500-Zn D 535
Dithiazone-silver SM18 3500-Ag D 462
Dithiazone-mercury SM18 3500-Hg C 492
Dithiazone-cadmium SM18 3500-Cd D 518
Dithiazone-lead SM18 3500-Pb D 510
Potassium permanganate SM18 3500-Mn D 525
Phenanthroline-iron SM18 3500-Fe D, 510
SM22 3500-Fe B
Bathocuproine-copper SM18 3500-Cu E, 484
SM22 3500-Cu C
Neocuproine-copper SM18 3500-Cu D, 457
SM22 3500-Cu B
Diphenylcarbazide-chromium SM18 3500-Cr D, 540
SM22 3500-Cr B
*Most of these methods were deleted from Standard Methods beginning with the 19th edition.

TABLE 8.7 Exact wavelength setting variation depending on the direction

of approach.

Beginning (nm) Final (nm) Absorbance

500 540 0.393
600 540 0.399
500 540 0.396
600 540 0.399

No two cuvettes have exactly the same transmission properties. The

manufacturers of the cuvettes recognize this by having an orientation line
marked on the outside of the cuvette. All colorimetric determinations
should be made in the same cuvette, oriented in the light beam in exactly
the same manner from one determination to the next. Using two different
cuvettes just introduces more variation in the analysis. If the cuvette
becomes scratched or hazed, it should be discarded. The U.S.
Environmental Protection Agency (U.S. EPA) recommends the use of a thin
layer of silicone oil on the outside of the cuvette to eliminate the effect of
scratches and scrapes, particularly for turbidity measurements. The silicone
oil has the same index of refraction as the glass and serves to fill the
scratches, making them invisible. However, this is not a good analytical
technique. The oil will coat the inside of the sample holder and trap more
dust and dirt inside the instrument. If the cuvette is unsuitable without the
oil, then it’s important to get a new one.
Some colorimeters are equipped with flowthrough sample cells that can
be either automatically or manually activated. These sample cells provide
fast analysis times for sets of samples, but are prone to becoming dirty on
the inside. The washing and purging cycles for the flowthrough cells can
actually take up more time than manually filling and reusing a cuvette.
When the cells become fouled, a decision must be made as to whether a
new cell should be purchased or the old cell cleaned manually. Some of the
flowthrough cells are impossible to clean manually, so the only option is
replacement. There are also the added preventive maintenance concerns of
the pump and tubing for the flowthrough cell as well as periodic emptying
of the waste jug.
The photodetector and other electronics in the colorimeter can decrease
in sensitivity and fail. In most cases, they are simply replaced by the service
technician. The lenses and mirrors associated with them can become dusty
and dirty and need periodic cleaning. The instrument itself should always
be spotlessly clean and covered with a dustcover when it is not in use. This
provides protection from accidental spills. The lifetime of the colorimeter as
a whole can be greatly extended by locating it in a clean area with stable
humidity and temperature, not moving it physically from one location to
another, and providing a stable electrical supply. When colorimeters are
subjected to environmental vibrations, mirrors and other parts of the optics
get out of alignment and a service call is required for adjustment.
Whether the instrument should be turned off when not in use is up to the
technician. Leaving it on eliminates having to wait for it to warm up before
use, extends the life of the lamp, and reduces wear and tear on the switches.
However, if the colorimeter is left on in an area with poor ventilation,
excessive heat buildup can lead to premature failure of the electronics. Our
laboratory turns the instruments off overnight, but leaves them on during
the day.
5.0 TURBIDIMETRIC DETERMINATIONS
Many of the considerations that apply to the use of the colorimeter also
apply to the turbidimeter or nephelometer. The objective is to pass a beam
of white light through the sample, then measure the light scattered at right
angles to the analytical beam. An assumption made in this analytical
method is that the measured amount of scattered light is proportional to the
particulate content of the sample. Although it is possible to establish a
relationship between the turbidity reading and the concentration of
particulates for a constant sample source, in general, the relationship will be
different for each sample. As discussed in the method, highly refractive
particles, such as sand or quartz, will give high refractions and high
apparent turbidities, whereas light-absorbing particles, such as charcoal and
organic degradation products, will give relatively low readings. True color
in the sample will also absorb light and reduce the amount of light reaching
the photodetector.
Turbidity in a sample will change over time and with conditions. In
most samples, reducing the temperature will cause material to precipitate
from solution and increase the turbidity, but increasing the temperature will,
in general, not decrease the turbidity. Many samples will exhibit an increase
in turbidity just upon sitting at room temperature over several days. This is
attributable to the vast majority of sample solutions not being at a stable
point of existence, termed equilibrium. Another cause for an increase in
turbidity over time is the multiplication of bacteria and microscopic plants
in suspension.
Changing the pH of the sample also changes the turbidity. Lowering the
pH by adding acid will typically reduce the turbidity caused by inorganic
materials. Turbidity caused by organic material suspended in the sample
will not be affected. Raising the pH by adding base will typically increase
the turbidity because of the precipitation of inorganic salts and colloids. For
these reasons, turbidity is most frequently measured immediately after
sample collection without any preservation.
There are many types of turbidity standards in existence, the most
commonly encountered being the freshly prepared formazin suspension,
synthetic polystyrene-divinyl benzene beads, and a variety of solid
standards constructed of hard plastic or soft gels. The beads and the
formazin are the only standards recognized as suitable as primary standards
in either Standard Methods or by U.S. EPA.
Turbidity readings are affected by anything that either increases the
scattering of light in the sample or standard or decreases the amount of light
hitting the photodetector.
Scratches anywhere on the sample cell, not just those in the direct path
of the light, will increase the scatter level. A thin layer of silicone oil on the
outside of the sample cell or solid standard is one way to reduce the effect
of scratches and scrapes. This is also a way to increase the amount of dirt
on the outside of the sample cell. The analyst should not neglect the top of
solid standards as a source of error because light can be reflected back
down through the sample from a dusty top surface and give a falsely high
scatter reading.
The inside of the sample compartment of the instrument must be kept
absolutely clean and free of dust. A freshly wiped sample cell typically has
a small static charge on it that will attract dust to the surface from within the
sample compartment.

6.0 TITRATION
Titration is the addition of a reagent solution of known concentration from a
buret to a known volume of sample until some change indicates complete
reaction. Based on the volume and concentration of the reagent added and
knowledge of the chemistry of the process, a concentration of a target
analyte or parameter in the sample can be calculated. A simple operation on
the surface, this time-honored procedure serves as a true indicator of the
ability of the analyst to perform bench-level chemical manipulations. A
number of factors affect the outcome of the titration analysis, including the
equipment, the chemistry of the titration, the indicator, the technique of the
operator, and the recording of the data.
The equipment for most titration procedures consists of a buret, a buret
support stand, a bottle of reagent, one or two empty beakers or flasks, a
graduated cylinder, a pipet or volumetric flask for volumetric measurement
of the sample, the container of sample, a receiving flask, a magnetic stir bar
and magnetic stirrer, a bottle of indicator, a squeeze bottle of reagent water
or other appropriate solvent, and a notebook and pen for recording data.
The buret and other volumetric ware should be Class A glassware and
appropriately cleaned for the test procedure, as are the rest of the glassware
and the stir bar. All the equipment should be conveniently arranged on a
clean surface before beginning the titration.
The first order of business is the titration reagent and the buret. The
buret chosen for the titration should present the maximum accuracy
possible for the expected amount of reagent to be dispensed. The maximum
accuracy for the buret occurs at the 80 to 100% capacity level. It is poor
practice to dispense 3.0 mL of titrant out of a 50-mL buret. If 3.0 mL is
expected to be used, a buret with a total capacity of 5.00 mL should be
chosen. Another alternative is to dilute the reagent so that a greater volume
is required. In the aforementioned example, a 1:7 dilution would call for
21.0 mL of titrant, and a 25-mL buret could be used to achieve maximum
accuracy.
The reagent should be at room temperature before filling the buret.
Most solutions expand as they warm, and the changing volume affects the
amount of reagent delivered during the titration. A temperature-indicating
strip on the side of the reagent bottle can be used as a quick check to ensure
that the reagent is at room temperature. A portion of reagent is dispensed to
a small beaker with a pour spout for filling the buret. Some analysts also
use a small glass funnel that fits into the top of the buret for filling. The
object is to avoid spilling any reagent on the outside of the buret. There are
three reasons for this.
First, most analysts try to avoid direct hand contact with reagents, and
having the outside of the buret and buret stand covered with reagent
increases the possibility of contact. Second, a drop of reagent may fall from
the outside of the buret into the sample during the titration and introduce an
error to the volume of reagent dispensed. Third, it is messy and contributes
to a sloppy work area, increasing the chance of sample contamination. Once
reagent is dispensed, it should never be returned to the reagent bottle. A
small beaker aids in dispensing just the right amount of solution.
After the buret is filled to above the zero marker (most analysts
remember to close the stopcock while filling the buret after one experience
of leaving it open), another beaker or flask is placed under the buret tip, and
the stopcock is opened to allow flow until the air bubble in the tip of the
buret is flushed out. The air bubble must be removed or it will disappear
during the titration, resulting in a false reading of the volume dispensed.
Many analysts perform the air bubble flushing as a part of the process of
rinsing the buret with the titration reagent. More reagent is added to the
buret until it is filled above the zero mark, then the stopcock is opened to
lower the reagent to the zero mark. The reagent dispensed into the beaker or
flask during the rinsing and flushing is discarded, not returned to the buret
or reagent bottle. Finally, the volume level on the buret is recorded on the
benchsheet or in the notebook, even if it is 0.00 mL.
If two burets are being used on the same buret support and they contain
different reagents—or even if they contain the same reagent at different
concentrations—they should be labeled. All too often an analyst will place
a buret with standard hydrochloric acid and one of sodium hydroxide on the
same stand and then express puzzlement about the extremely high alkalinity
of that day’s effluent sample, caused by attempted titration with the sodium
hydroxide solution.
The receiving flask should be an Erlenmeyer flask or other container
with a narrow mouth (e.g., a biochemical oxygen demand bottle). A beaker
is too prone to having drops splash out of the container. The receiving flask
should also have a flat bottom to facilitate the stirring action of the stir bar.
Some analysts will clamp the receiving flask to the support stand to avoid
inadvertent spills.
The sample to be titrated is measured out with the graduated cylinder,
pipet, or volumetric flask into the receiving flask. If a “to contain” (TC)
flask or graduated cylinder has been used, it must be rinsed with several
small portions of reagent water from the squeeze bottle and the washings
added to the receiving flask to effect a quantitative transfer of the sample.
The minimum amount of water should be used for the transfer because a
large dilution of the sample can affect the results, particularly when an
electrode is used to establish an endpoint to the titration. If other reagents
need to be added to the receiving flask, it is done at this time. The amount
of sample taken for titration is written down on the benchsheet.
The buret and flask should be oriented so that the drops of titrant fall a
minimum distance to the surface of the solution in the flask without hitting
the sides of the flask. Thus, the tip of the buret should be lowered inside the
mouth of the flask until there is just enough room to turn the plug of the
stopcock.
The solution should be stirred at a moderate rate to provide rapid
dispersion of the drops of reagent; however, splashing the solution is to be
avoided. The walls of the container can be washed with reagent water from
the squeeze bottle so that any material adhering to the walls can be titrated.
Most titrations should be performed rapidly during the first part of the
reagent addition, then slowed to a dropwise addition as the endpoint is
approached. This process can be hastened by performing a rapid rough
titration at a fast rate of addition and noting at about what volume the
indicator changed. The titration is repeated on a second aliquot of sample
with a fast rate of addition up to a slightly smaller volume than is required
for completion, then making a slow dropwise finish to the endpoint.
Another technique is to withdraw 1 to 5% of the sample from the receiver,
titrate it to an approximate endpoint, then add it back into the receiver.
Dividing the amount of titrant required for the small titration by the decimal
fraction of the small portion withdrawn gives an indication of the total
titrant volume needed.
The endpoint of the titration is that volume of titrant necessary to just
achieve the desired indicator change. One drop of sample added to the
solution should be able to reverse the indicator change. Another technique
to verify the indicator change is to note the reading on the buret and then
add another drop or two of titrant to check the completion of the titration.
The indicators, for the most part, are not neutral observers to the
chemical reaction, but are actual participants. For acid-base indicators, such
as bromocresol green or phenolphthalein, the indicator consumes acid or
base to perform the color change. Starch used in iodometric titrations binds
with the iodine to form the blue-black color; however, if the starch-iodine
complex is allowed to sit for any length of time, the iodine irreversibly
reacts with the starch. This is why the starch is added only when there is a
bare trace of iodine left in the solution; once the starch is added, the titration
is quickly finished. Indicators should be added to the solution in the
minimum amount necessary to visualize the color change. Excess amounts
of indicator can actually obscure the endpoint, particularly when the
indicator changes from one color to another. For example, a large amount of
the wine-red Eriochrome Black T indicator will mask the initial onset of the
blue color of the endpoint.
 Each sample should be titrated to the same intensity of indicator color.
Titrating one phenolphthalein-containing sample to a slight pink tinge and
another to a deep red-pink will just introduce an avoidable variation in the
results. It is generally a good idea to prepare an endpoint comparison
sample of the desired color and use it to check the endpoint color of tested
samples. This can occasionally present problems when intrinsic color in the
sample generates a different endpoint hue than the standard. Most titrations
are ended when the indicator develops a permanent color, not a color that
appears for a second or two then fades back to the original state. Normally,
this situation is a sign that the endpoint to the reaction is very close, but that
the titration is not quite complete. A full drop of titrant may be more than
the amount required to complete the reaction. Partial drops can be allowed
to form at the tip of the buret, then touched off to the side of the flask and
washed into solution with a squirt of reagent water from the squeeze bottle.
Directly washing the partial drop from the buret tip with reagent water can
leach more reagent from the borehole in the buret tip and reduce the
accuracy of the titration. Stopcock metering valves can be used to produce
very small partial drops of reagent.
Reagent water is difficult to prepare and has a great tendency to absorb
gases and other materials from the air. This changes the purity of the water
even though it still appears clean. Contaminated reagent water can affect the
titration by either increasing the amount of titrant required or, in some
cases, reducing the amount of reagent necessary because of a previous
hidden reaction with the target analyte. When using reagent water, it is
essential to determine its possible contribution to the titration. This can be
done by treating a portion of the reagent water as a sample and performing
the titration upon it. For most titrations, only one or two drops of reagent
should be required to reach the endpoint. The reverse condition should also
be checked, in which the reagent water may contain material that reacts in a
similar fashion chemically to the titrant and the addition of indicator results
in an endpoint signal; for example, addition of starch indicator to the
reagent water gives a colorless solution. A drop of weak iodine solution
should be sufficient to produce the blue-black iodine-starch complex. If
more than a single drop is required, the water is reductive and either this
effect must be quantitated and accounted for during the use of the water or
the contamination must be removed.
 When the titration is completed, the buret should be emptied of excess
reagent and cleaned. Often, simply rinsing the buret with reagent water
suffices. The buret should be stored upside down with the Teflon® stopcock
plug loosened. Burets are not storage containers for titration reagents. It is a
good practice to dedicate a buret to a single reagent and purpose. Although
burets are accurate volumetric devices, always using the same buret for the
same titration cuts out another possible random error variable.
Documentation of the titration consists of the last date of
standardization and the value of the titrant solution; results of reagent water
analysis; the volume of sample titrated; the beginning and ending readings
off the buret for each titration; the way in which the final result was
calculated; and the date, time, and name of the analyst. Titrations should
always be performed in duplicate and the average result reported.

7.0 pH AND ION-SELECTIVE ELECTRODES

The attachment of a voltage or amperage meter to a chemical sensor to form
an analyte selective electrode is an area of intense activity, with new
products being introduced all the time. Most analysts are familiar with pH
electrodes and meters, the most common method of pH measurement in the
United States. Other electrodes that are selective for ammonia, cyanide,
fluoride, nitrate, potassium, or dissolved oxygen are described in the most
recent editions of Standard Methods. Many other ion-selective electrodes
(ISE), which are selective for parameters, such as residual chlorine,
chloride, bromide, calcium, carbon dioxide/carbonate, lead, lithium,
sodium, and ionic surfactants, to name a few, are available but not widely
used in the water and wastewater community. However, this will change
over time. A fascinating development is the coupling of a fiber-optic line to
a photodetector and a minute chemical sensor for measuring color changes
in the sensor resulting from exposure to target analytes.
The advantages of the electrode are that it does not require a lot of
chemical reagents for use, the analysis is generally fast, and the
concentration range over which it is reliable often spans a factor of 1000 or
more. The drawbacks are that the individual electrode is often expensive
($200 to $600 is not uncommon) and, although preventive maintenance can
be performed to extend its useful life, it is often viewed as expendable.
Electrodes come in two basic types: single electrode and combination
electrode. The single electrode requires use of an additional reference
electrode to make a measurement, whereas the combination electrode has a
reference electrode built into it. The two most common reference electrodes
are the silver-silver chloride reference and the calomel (mercury chloride)
reference. The combination electrodes frequently incorporate a silver-silver
chloride reference.
Most electrodes are temperature sensitive, and all samples and
calibration solutions need to be at the same temperature for accurate
measurement. Some electrodes, such as those used for pH measurement,
have temperature compensation devices either built into the single electrode
body or available as another probe to stick into the sample. These are not a
substitute for having the samples and standards at the same temperature.
The pH/ISE workstation contains the meter and a supporting holder for
the electrodes, a magnetic stirrer, a squeeze bottle of reagent water, and a
waste container. The squeeze bottle is for rinsing the electrodes between
each determination. Other items may include a cup of storage solution,
electrode-filling reagent, and tissues for gently wiping the sensing tip of the
electrode should it become fouled. A sample conditioning reagent is
required for many ISE procedures to control pH, suppress significant
interferences, or adjust the overall ionic level within the sample.
Most electrodes are designed to consume a small amount of analyte
from solution. This requires that the solution be stirred during the
measurement. This is particularly true for the electrodes that measure free
gases in solution, such as the ammonia, dissolved oxygen, and carbon
dioxide sensors. The solution must be stirred so that a constant level of the
analyte is available to the electrode surface, but the solution should not be
stirred so fast that the gas is aerated from the sample. This also suggests that
gas sensing electrodes can be used only once on each aliquot of sample. It is
not possible to set out cups of calibration solutions, measure them, then try
to measure them again and expect the same results.
For most direct measurements, a portion of the sample is poured into a
disposable plastic cup of about 50-mL capacity. Some analyses require a
larger volume and a larger cup should be used. Glass beakers can be used in
some cases, but this only increases the number of items to be washed and
cleaned. Each sample should be measured in a separate, clean container.
Fluoride analysis must be performed in plastic containers because fluoride
reacts with glass.
The analyst must read and be familiar with all parts of the
manufacturer’s directions for use of the electrode. Preventive maintenance
steps that must be complied with for successful use and long life of the
electrode include the following:

• Conditioning steps to be performed to get the electrode ready for

initial use,
• Electrode body filling solution and frequency,
• Sensing membrane replacement and frequency,
• Storage conditions during periods of non-use,
• Required sample pretreatment, and
• Reference electrode requirements.

All electrode systems need to be calibrated each day. The response of

the electrode is sensitive to environmental conditions to the point that
yesterday’s calibration is generally useless today. Fortunately, the
calibration procedure only takes about 5 to 10 minutes.
Many pH/ISE meters can hold a calibration and will directly read in
concentration units. Although this is a generally accepted procedure for pH
determinations, it is poor analytical practice and is completely lacking in
documentation that a calibration has been performed. The analyst should
record the response of the electrode to the calibration standards and prepare
a calibration graph on a daily basis. This daily graph is a good tool for
determination of the length of time required for a stable reading on samples
of low concentration. The gas-sensing electrodes, in particular, are known
for their slow response; for example, the ammonia electrode can take up to
10 minutes to accurately read a 0.1-mg/L standard. The analyst can prepare
a calibration graph for the higher standards (1.0, 10.0, and 100.0 mg/L for
the ammonia example) and extend the line down to the lowest level (0.1
mg/L), then use a stopwatch to determine how long is required for the
electrode to reach the proper response.
Electrode-based determinations are typically not affected by color or
turbidity in the sample; however, they can become fouled with oil, grease,
or precipitating materials from the solution. It is also possible to poison the
electrode. Therefore, the analyst needs to be aware of common
interferences listed in the method and in the manufacturer’s directions.

8.0 REFERENCE
American Public Health Association; American Water Works Association;
Water Environment Federation (2012) Standard Methods for the
Examination of Water and Wastewater, 22nd ed.; American Public
Health Association: Washington, D.C.
 9
Calculation and
Reporting Results

1.0 SIGNIFICANT FIGURES

2.0 CALCULATIONS
3.0 DRY WEIGHT CORRECTIONS
4.0 REPORTING RESULTS

After data have been collected from the analysis, they typically must be
transformed (using a formula) into concentration units for reporting. Up
until the early 1970s , slide rules were used to ease the labor of hand
calculation, particularly multiplication and division. The slide rule is an
approximation device and can be read to only about one part in a thousand
(0.001). This also happens to be the level of uncertainty in most results
obtained in the laboratory.
Results calculated from the slide rule for the most part expressed the
same uncertainty as the values generated by the analytical tests, and
analysts did not have to pay much attention to how they expressed the final
number. For example, suppose alkalinity is calculated from a titration that
uses 8.18 mL of 0.106 N standard acid on 200.0 mL of sample using the
following formula:

If a slide rule is used for the calculation, the result is 217 mg/L. If an
electronic calculator is used, the answer generated is 216.770 mg/L.
Mathematically, the answer using the calculator is more specific; however,
a real question that must be asked is, “Does the answer misrepresent the
analysis?” The answer, of course, is yes.
Before we explore why the misrepresentation exists, a few conventions
must be established. The first is the concept of significant figures as they
apply to laboratory work.

1.0 SIGNIFICANT FIGURES

The idea of significant figures is grounded in the uncertainty of
measurements. In laboratory work, there is always an uncertainty in the
measurement attached to each procedure. In the aforementioned alkalinity
example, if a 25-mL Class A burette is used for the titration, the position of
the meniscus can be readily determined as being between 8.1 and 8.2 mL
because there are calibration marks at 8.1 and 8.2 mL (Figure 9.1). The
meniscus is below the 8.1 mark and above the 8.2 mark. A more specific
result is obtained by estimating where the meniscus lies between the two
marks. Based on the analyst’s judgment, the meniscus is within the last
third of the distance between the two volume marks, and 8.17 is a good
estimate of the position. Therefore, the analyst writes down “8.17 mL” on
the benchsheet, with the understanding that there is uncertainty in the last
decimal place. It might be a 6 or it might be an 8; however, 7 is the best
estimate within that range. The measurement has three significant figures.
In general, analysts always interpret numbers on benchsheets and other
laboratory records to have an uncertainty of ±1 in the last digit unless a
definite uncertainty is written down with a value, such as 44.5 ±0.2 °C. If a
2.00-mL volume is dispensed with a Class A volumetric pipette and the
analyst writes down “2 mL” on the benchsheet, there is going to be a
definite problem in understanding when someone else reviews the data, as
illustrated in Table 9.1. The analyst who created the record and knows what
it means probably will not be around when the record is examined months
or years later. As illustrated in Figure 9.2, the correct value to record for this
titration is 4.41 mL, not 4.4 mL. In Figure 9.3, the correct value to record is
5.00 mL, not 5 mL.
 For most volumetric measuring devices, the maximum accuracy in the
volume determination and thus the number of significant figures that can be
reported is limited first by the tolerance of the placement of the volume
markings on the burette, flask, pipette, or cylinder, and second by the ability
of the analyst to reproducibly dispense the same volume from the device.
As a general rule, three significant figures represent the maximum
reproducibility of these measurements, although, occasionally, four can be
justified.
FIGURE 9.1 Uncertainty of reading the position of a meniscus in a
burette. Courtesy of Apichemical Consultants.

TABLE 9.1 Implied uncertainties for different ways of writing a number

on a benchsheet.
 Number Implied uncertainty
2 1 to 3
2.0 1.9 to 2.1
2.00 1.99 to 2.01

As an example, a weight determination on an analytical balance can

generate a measurement with up to five and sometimes six significant
figures. If the balance is the only measuring tool used in the entire analysis,
then results with five or six significant figures may be possible. However,
the significant figures in the result must reflect the uncertainty of the least
accurate device used in the analytical procedure. In a total suspended solids
determination, the balance may have given a residue difference of 1.2345 g;
however, the analyst used a 100-mL Class A graduated cylinder for sample
dispensing. The result is limited by the uncertainty of the graduated
cylinder (±0.35 mL, or 99.6 to 100.4 mL), not the balance. The volume
uncertainty range of the cylinder is greater than three significant figures, but
less than four, so three significant figures are reported. The reported result
for the test should be 12 300 mg/L, reflecting the uncertainty of the
graduated cylinder, instead of 12 345 mg/L with the five significant figures
derived from the balance reading.
FIGURE 9.2 Meniscus at 4.41-mL volume in a burette. Courtesy of
Apichemical Consultants.

Another example is encountered in the preparation of reagent solutions.

The reagent is weighed out on the balance to give, for instance, 0.1234 g of
standard, which is then dissolved to volume in a 100.0-mL Class A
volumetric flask. The concentration can be reported as 0.01234 g/L. The
standardization procedure then calls for taking a portion of the solution and
titrating it against a primary standard. The burette used in the titration may
give only a three-figure volume (8.12 mL), and thus the concentration of
the standardized solution will have to be reduced to three significant figures
(0.0123 g/L) to reflect the uncertainty in the titration. If the analyst arranges
the titration to use more than 10.00 mL, then retaining the four significant
figures (0.01234 g/L) is warranted. In general, the smaller the volume or
weight measured, the fewer significant figures that can be reported and the
greater the overall uncertainty in the measurement.
FIGURE 9.3 Meniscus at 5.00-mL volume in a burette. Courtesy of
Apichemical Consultants.

When calculations are performed, the reported result must reflect the
overall uncertainty of the testing process, or, more correctly, the
significance of the most uncertain measurement in the process. For
instance, suppose a result is calculated from the following measurements:

The question is what implied uncertainty should be reported. The

correct answer is suggested by examining what the numbers are interpreted
to mean. The implied errors are as follows:
 The answer needs to reflect the most uncertain measurement, which is
the one associated with the ±1. Therefore, the analyst is correct in choosing
to report only an integer value. Statistics texts on that discuss scientific
measurements that give involved procedures for generation of error
estimates, but, in a simplistic view, the entire measurement error in the
process can be roughly approximated by adding the individual
uncertainties, which in this case gives ±1.2101. Simply choosing to present
the final result for the sample as the measurement with the greatest error, ±1
from the 105 value, is a fair approximation of the appropriate implied error.
It is now easy to understand why the calculator-derived number may
misrepresent the analytical result. If the analyst reports all the digits that the
calculator provides (e.g., 12.54693), without any regard for significant
figures and analytical reality, the result appears more significant than it can
possibly be.

2.0 CALCULATIONS
With the exception of dissolved oxygen meter and pH meter readings, the
final result of most procedures that are performed in the water and
wastewater laboratory is not directly determined from the actual
measurement. Some conversion of the measurement value must take place.
For example, alkalinity is measured in the laboratory from a volumetric
titration, but the final reported result is not the number of milliliters of
standard acid. Similarly, the absorbance readings from the colorimeter in
phosphorus analysis must be converted to the final result in milligrams per
liter.
For the most part, the conversion formula is included as part of the
method. The alkalinity formula is listed as follows:

Where

A = the volume of acid used in the titration, mL;

 NB = the normality of the acid used; and
C = the volume of the sample tested, mL.

The constant in the formula (50 000 in this case) is derived from a
combination of conversion factors including milliliters to liters, normality
to molarity, and the stoichiometric factors of reaction equivalents and
molecular mass of calcium carbonate (discussed further in Appendix D). It
is important to observe that the correct units of measurement are used in the
formula. If the calculation requires milligrams for a weight, the use of
grams will give an incorrect result.
The analyst should always include a copy of the final result calculation
used in the method on the bottom or back of the benchsheet that documents
the analysis. At least one of the results should be worked out in some detail
so that there is no confusion as to what formula was used to generate the
values. It is poor analytical practice to read the number of milliliters used in
a titration off the burette, punch the values into a calculator, and record only
the final result on the worksheet. If a calibration graph is used to generate a
value, its location must be referenced on the benchsheet, or better yet, a
copy can be attached to the benchsheet. Sufficient detail must be provided
to allow another person to check the analyst’s work from the raw analytical
response obtained during the measurement to the final reported result.
Everyone makes calculation and data entry mistakes. It is not shameful to
have work checked by another analyst to make sure the correct results have
been generated; rather, it is only prudent.

3.0 DRY WEIGHT CORRECTIONS

Dry weight corrections are used for conversion of results obtained on solid
samples that contain water to a parameter result that would be obtained if
there were no water in the sample. This is frequently done so that
comparisons between different samples can be made on a common basis.
For instance, realistic lead contamination comparisons between batches of
sludge are impossible if one sludge represents 6% water and another 15%
water.
 Performing the dry weight conversion requires knowledge of the water
content of the sample, which means a total solids determination on the solid
must be made in addition to the needed parameter tests. Once the residue
value is obtained from the total solids test, the percentage of solids can be
determined as follows:

If a dry residue of 0.1452 g is obtained from a sample weight of 0.9672 g,

then the percentage of solids is as follows:

The reported value has been rounded to three significant figures to reflect
the general laboratory convention of reporting only three significant figures.
In the actual practice of determining the percentage of solids in the
laboratory, three significant figures may in fact be an understatement of the
uncertainty of the test. Moisture content can vary widely over short ranges
within a solid, and it is not uncommon to see variations in the percentage of
solid determinations over a 3 to 4% (and larger) range for repeated tests of
the same sample.
The percentage moisture of the sample is often desired as a reported
characterization of the sample. It is easily obtained by subtracting the
percentage of solids from 100. For the aforementioned example, this gives
the following:

The conversion of a tested parameter to a dry-weight basis is performed by

dividing the parameter result for the wet sample by the decimal fraction of
the solids content of the sample, as follows:
For example, a chromium test has been made on a sludge to give 4.8 mg/kg
(wet weight), and the percentage of solids determined as 16%. The dry
weight result for chromium becomes as follows:

Dry weight correction always gives a higher value for the parameter, which
serves as a check on the calculation. If a lower value is obtained, the result
needs to be recalculated.

4.0 REPORTING RESULTS

Most laboratory results, with the exceptions of pH and a few other
parameters, are reported in parts per million (ppm) units, either milligrams
per liter (mg/L) or milligrams per kilogram (mg/kg). Because 1000 mg/L is
equivalent to 0.1%, the custom is to avoid using parts per thousand as a
reporting unit; instead, percentages are used. When the reported parameter
amounts are very low, a common procedure is to use parts per billion (ppb)
units, which are equivalent to micrograms per liter (μg/L) or micrograms
per kilogram (μg/kg). One thousand parts per billion is equal to 1 ppm. The
correct symbol for micro is the small greek m, “μ”. Parts per trillion (ppt)
are seen as nanograms per liter (ng/L) and nanograms per kilogram (ng/kg).
The other occasionally seen prefixes are presented in Table 9.2.
Another reporting unit is the equivalent (eq) or milliequivalent (meq)
per unit volume or mass. These units are based on the atomic mass of the
parameter, such as sodium, potassium, or chloride. Moreover, these units
are commonly encountered in medical laboratories, but only a few
environmental tests use them. One example is cation exchange capacity,
which is reported in milliequivalents per kilogram (meq/kg). The
conversion from milligram to milliequivalent is performed by dividing the
number of milligrams by the atomic mass of the element. For example, the
atomic mass of sodium is 23, and 23 mg of sodium is, therefore, equal to
1.0 meq of sodium.
TABLE 9.2 SI prefixes for metric units.

Prefix Abbreviation Multiplier

zepto- z 10−21
atto- a 10−18
femto- f 10−15
pico- p 10−12
nano- n 10−9
micro- μ (u) 10−6
milli- m 10−3
centi- c 10−2
deci- d 10−1
deca- da 10
hecta- h 102
kilo- k 103
mega- M 106
giga- G 109
tera- T 1012
peta- P 1015
exa- E 1018

Calculation of anion/cation balances requires further adjustment of the

milliequivalent value to account for the charge on the ion. The convention
is to divide the milliequivalent value by the charge. For example, sodium,
potassium, nitrate, chloride, and bromide have a charge of 1, thus no
conversion is needed. Calcium, sulfate, and magnesium have a charge of 2,
thus the milliequivalent value is divided by 2. Phosphorus and aluminum
have a charge of 3, thus the divisor is 3.
Most process control decisions are based on values measured in gallons
per million gallons or pounds per million gallons. Because a gallon per
million gallons expresses a part per million, a milligram-per-liter result only
requires a direct substitution of units. For example, 1.45 mg/L is equivalent
to 1.45 ppm and 1.45 gal/mil. gal. To convert to weight measures, the
common convention is to multiply by 8.34, which is the number of pounds
of water in 1 gal. Thus, the aforementioned 1.45 mg/L is multiplied by 8.34
to give 12.1 lb/mil. gal.
Many parameters have reporting units that are not immediately obvious.
Nitrogen tests are reported as nitrogen (N), regardless of form, be it total
Kjeldahl nitrogen, ammonia, nitrate, or nitrite. Acidity, alkalinity, and
hardness are reported as calcium carbonate (CaCO3). Phosphorus results are
reported as phosphate and potassium as potassium (K) in the water and
wastewater laboratory. However, in an agricultural laboratory, phosphorus
is reported as phosphorus pentoxide (P2O5) and potassium as potash (K2O).
These reporting conventions may be established, as in the case of the
nitrogen parameters, in the calibration standards. Directions for the
preparation of the 100-mg/L stock solution for nitrate analysis is actually a
443 mg/L solution if the concentration is expressed as nitrate instead of
nitrogen, and 722 mg/L if the concentration is expressed as potassium
nitrate (SM22 4500−NO3−B). In other cases, the calculation equation given
in the method contains the conversion, as in the hardness and alkalinity
report form.
One of the signs of a quality laboratory that produces reliable data is
how much care the analysts apply to the calculation and reporting of results.
Analysts should never trust their memory or mental ability to convert
results; rather, they should write down what has been done and have
someone else check the results.
 10
Controlling the Test Procedure

1.0 THE ABILITY OF THE TEST TO ACTUALLY MEASURE

THE DESIRED PARAMETER IN THE SAMPLE
2.0 HOW CLOSE THE RESULT IS TO THE ACTUAL
AMOUNT OF ANALYTE IN THE SAMPLE
3.0 CAN THE SAME RESULT BE OBTAINED REPEATEDLY?
4.0 WHAT IS THE LOWEST LEVEL OF ANALYTE THAT CAN
BE DETECTED IN THE SAMPLE?
5.0 IS THE DETECTED PARAMETER ACTUALLY IN THE
SAMPLE?
6.0 BATCH ANALYSIS
7.0 REFERENCES
8.0 RECOMMENDED READING

In this chapter, we introduce the concept that, when absent, drives

laboratory owners and managers up the wall and, when present, makes
everything work smoothly. The concept is, of course, analytical sense.
Analytical sense is the recognition that something did not work quite right
in the analysis and it is going to have to be repeated. It is the difference
between the analyst who seldom misses a check sample and the technician
who can never get the right answer. Analytical sense is typically a concern
about what the analyst is doing and all the factors that may come into play
and affect the result of the analytical process. This book represents an
attempt to present some of the arcane knowledge necessary to promote
analytical sense, but no analyst can even hope to be able to take into
account all of the factors that can influence a result of a tested parameter.
Control over many of the identifiable factors related here will result in a
drastic reduction in the number of “matrix interference” excuses reported in
the case narratives. Admittedly, a skilled analyst will still end up with
sample and batch results that demonstrate poor precision or matrix spike
recoveries; however, these results will be few and far between when
analytical sense is being used by the chemist.
In spite of the presence of analytical sense among laboratory workers,
quantitative measures of the success of the analytical procedure, such as
percent recovery (%R) for accuracy and relative percent difference (RPD)
for precision, are still necessary. This is because of some types of problems
making their presence known slowly over time. These quantitative
measures can indicate a looming failure in the procedure. It is like a leak.
And no leaks go away on their own; they only get worse. The best
procedure is to detect the leak early on and fix it. The same philosophy
applies to laboratory work: find the problem early on and fix it.
For each test that is performed, there is a common set of concerns about
the analysis that should be answered to the satisfaction of both the
technician and the persons who use the results. These concerns are as
follows:

• The ability of the test to actually measure the desired parameter in the
sample,
• How close the result is to the actual amount of the analyte in the
sample,
• Whether the same result can be obtained repeatedly on the sample,
• The lowest level of analyte that can be detected in the sample, and
• Whether the detected parameter is actually in the sample.

A reality of performing chemical analysis is that the same answer is not

obtained every time the analysis is performed. This is because each step in
the procedure is not repeated exactly the same way every time. Two types
of errors are recognized as accompanying each analysis: random and
systematic. A systematic error is a consistent mistake or bias in a single
direction from the true value, such as an error in calibration or an incorrect
volume measurement. An example of this is an analyst who prepares a
stock calibration solution of 100 mg/L, but mislabels it as “50 mg/L”. Every
result generated from the calibration curve with the mislabel will always
give a value that is one-half of the correct value.
Random errors result from the inability to perform a procedure exactly
the same way twice in a row. Random error results are always distributed
equally around a central value. Figure 10.1 illustrates the theoretical
distribution of results from a test controlled solely by random error. Figure
10.2 illustrates a real example in which the recovery from total suspended
solids tests is plotted as a frequency distribution. The average value for the
distribution is 93.9. The true value is l00, thus there is a systematic error of
6.1 to the low side. The width of the distribution is a measure of the random
errors that contribute to the analysis. The distribution is even around the
central point of 93.9, thus random errors are the main contributing factor to
the width of the distribution.
FIGURE 10.1 Theoretical distribution of results controlled only by
random error around a mean.

Over time, protocols have been developed to provide the answers to

these questions. These protocols are termed quality controls. A quality
control is a technique that measures a single aspect of a test’s success.
Examples of quality controls are matrix spikes, continuing calibration
verifications, blanks, and duplicate analysis. The combination of all the
quality controls performed in the laboratory along with the training,
facilities, instrumentation, supplies management, sampling techniques, and
containers, and anything else that affects the ability of the laboratory to
produce reliable data, is termed the quality assurance program. It is
typically formalized in a written document called the quality assurance
manual.
FIGURE 10.2 Frequency distribution for total suspended solids recovery.
1.0 THE ABILITY OF THE TEST TO
ACTUALLY MEASURE THE DESIRED
PARAMETER IN THE SAMPLE
This aspect of the testing process is more properly consigned to the research
and development department rather than the water and wastewater
laboratory. For the most part, laboratories answer this concern by using
methods of analysis that are mandated under one or more regulatory
programs. All regulatory programs that specify monitoring requirements
will also provide lists of approved methods for providing appropriate data.
Most commonly, approved methods are published in documents from the
regulatory agency, such as the U.S. Environmental Protection Agency (U.S.
EPA), or they will cite laboratory industry standard compilations of
methods (e.g., Standard Methods for the Examination of Water and
Wastewater [APHA et al., 2012]). When questions arise about whether a
method is approved or not, the regulatory agency is the only source of
reliable information. Claims of approval from a manufacturer’s literature or
sales representatives exist for the purpose of selling the product, period.
The monitoring reports required by the regulatory authority all must be
signed by the plant or laboratory manager. This signature acknowledges
legal responsibility that the data included in the report were obtained using
approved methods. The water and wastewater laboratory personnel are
solely responsible for ensuring that only approved methods are used for
data collection.

2.0 HOW CLOSE THE RESULT IS TO THE

ACTUAL AMOUNT OF ANALYTE IN THE
SAMPLE
The technical term for the closeness of the obtained result to the true value
of the target analyte in the sample is accuracy. The most appropriate way to
measure accuracy is to add a known amount of the target analyte to the
sample, then reanalyze the sample and see how close to the added amount
the result is. This is called a matrix spike. Accuracy is an average concept,
although it is typically determined from a single matrix spike sample.
Accuracy is quantitatively assessed by calculation of percent recovery
(%R). This can be done one of two ways, as illustrated in Equation 10.1.

The first equation is used when there is no detected target parameter in the
sample. The second equation is the correction necessary when there is target
analyte in the sample before the addition of the matrix spike. The true value
is the amount of the matrix spike added to the sample.
Perfect accuracy gives a value of 100 for the percent recovery. For most
general chemistry procedures, acceptable accuracy results are from 80 to
120, although the method should be consulted for the proper range.
The idea behind the matrix spike is to obtain data on how capable the
test is of measuring the target analyte level in the sample. A low matrix
spike recovery may indicate that the sample is interfering with the
measurement of the parameter by sequestering the parameter in some
manner that makes it unavailable for reaction with the testing reagents.
Low recoveries may also indicate interference with the measurement of
the reaction product of the reagent with the target analyte. High recoveries
may be an indication that the sample is reacting with the reagent to form a
product that may have different proportions of the analyte in it than the
normal product. Regardless of the exact reason behind the high or low
recoveries, the analyst has definite information that the test procedure is not
generating an accurate result.
As far as the actual analysis of the matrix spike is concerned, the analyst
must be careful that the value for the matrix spike lies in the calibrated
range for the test procedure. A common error is to ignore the existing
(background) amount of the target analyte in the sample before the addition
of the matrix spike. Although the matrix spike amount and the background
amount individually may lie within the calibrated range, the sum of the
values may be above the range, requiring a dilution for analysis. For
difficult samples, the matrix spike procedure is invaluable in providing an
indication of how well the test is generating a reliable answer. The ultimate
expression of the matrix spike procedure is the multiple standard addition
calibration, discussed in Chapter 7.
Unfortunately, there are some target analytes that cannot be matrix-
spiked. This is because there is no known solution that can be added to the
sample to produce a predictable increase in the target analyte level.
Examples of analytes in this category are biochemical oxygen demand
(BOD), pH, alkalinity, and conductivity. There are many reasons for the
failure in the matrix spike for the example parameters, but they boil down
to a lack of a reasonable degree of predictability of the interaction between
the spike solution and the sample. This does not mean that analysts should
abandon their attempt to assess the accuracy of the procedure; they merely
have to use other techniques. For instance, they can either spike a sample of
reagent water with the target analyte (used in the BOD) or test a prepared
sample from a commercial supplier that contains a known level of target
analyte. Both of these are termed laboratory control samples, although the
purchased solution is sometimes also referred to as a quality control check
sample. The results from the laboratory control samples are treated in the
same way as those obtained from the matrix spike, but they do not give the
analyst the exact same information about how the test worked on the
individual sample.

3.0 CAN THE SAME RESULT BE OBTAINED

REPEATEDLY?
Precision is the term given for the ability to repeatedly generate the same
result from the same sample. It is a measure of the width of the distribution
of results around a central value. Precision is measured in the laboratory on
a daily basis by analyzing sample duplicates. Sample duplicates are
performed by pouring out two identical aliquots of the sample and treating
them as separate samples throughout the analytical procedure. Then, the
relative percent difference is calculated with the final results (Equation
10.2). The values A and B are the individual results obtained from the
sample and sample duplicate analysis. A perfect result for relative percent
difference, obtained by getting exactly the same value for the sample and
sample duplicate, is zero.

Relative percent difference from sample duplicates is a good measure of the

precision of the analysis as long as two criteria are met. The first is that the
target analyte is found in every sample analyzed. This is always true for pH
and conductivity, and it is frequently true for many general chemistry
parameters such as BOD, chemical oxygen demand, nitrate, chloride,
fluoride in drinking water, sodium, total dissolved solids, and other tests.
Different samples will have different levels of the target analyte. One day a
high sample will be tested, the next a low sample, and, over time, a good
picture of the precision capability of the laboratory will emerge at all levels
of the calibration range. For other parameters (e.g., cyanide), heavy metals
(e.g., mercury, arsenic, and lead), and most organic target analytes (e.g.,
benzene, acenaphthene, and bis[2-ethylhexyl]phthalate), the analyte is
infrequently found and the most common result is below detection limits
(BDLs) or not detected. Although a BDL or not detected is a valid result,
the relative percent difference calculated from many days of not detected is
not a reliable measure of a laboratory’s ability to measure the parameter;
rather, it is only a measure of how consistently it cannot be detected. For
this reason, relative percent difference is most frequently determined in
these cases from the matrix spike and a matrix spike duplicate. Using this
approach, the analyst is able to calculate a relative percent difference on a
real amount of analyte every day. It does have the limitation that only a
single level of analyte is tested for precision because most laboratories only
use a single matrix spike stock solution.
The second criterion for successful use of the relative percent difference
metric, as related to precision, is that the recovery of the target analytes is
around 100%. This is easily seen from an example in which the relative
percent difference is calculated from the matrix spike and matrix spike
duplicate results for a test, illustrated in Table 10.1.
In cases 1 and 2, the values of relative percent are centered on good
accuracy (100) and the calculated relative percent differences of 2 and 100
indicate good and poor precision, respectively. However, in case 3, the poor
recovery translates into a poor relative percent difference, even though
common sense suggests that the ability to reproduce the poor recovery is
actually an indication of good precision. This further points out the often-
quoted observation that precision always gets worse the lower the analyte
level.
This same phenomenon is seen when relative percent difference is
calculated from sample duplicates and low levels of target analyte are
encountered.
Another measure of the variation of results is the relative standard
deviation (RSD). Relative standard deviation is a statistical concept that
attempts to generalize the results of a few observations to the behavior of a
population of results. Relative standard deviation is calculated by dividing
the standard deviation of a set of results by the mean of the data. Relative
standard deviation can be applied to two results to obtain Equation 10.3.
Comparison of this equation with that for relative percent difference in
Equation 10.2 indicates that the multiplier of 2 has been replaced with the
square root of 2, about 1.41. There is nothing incorrect about expressing
precision as a statistical generalization (RSD); however, real precision
determined with the relative percent difference metric will not be
comparable to that calculated as RSD.

TABLE 10.1 Examples of relative percent difference calculated from

matrix spike and matrix spike duplicate recoveries.
Precision and accuracy are distinctly different concepts as measures of the
success of an analysis. Their relationship is most commonly illustrated with
an example from target shooting (Figure 10.3).
The top two targets illustrate high accuracy with associated high and
low precision. Remember, accuracy is an average concept. A good test of
the comprehension of the concepts of accuracy and precision, as applied to
laboratory analysis, is to assume that the best situation of high accuracy and
high precision (upper left) is not possible, then select the second most
desired picture. The analyst should select the lower left picture, that is, low
accuracy and high precision. The reason behind this choice is that there is
limited time for analysis and reanalysis of samples. An analyst cannot
afford to test the sample five to 10 times to obtain a correct average result,
as depicted in the upper right target, and a sample and sample duplicate
result in this situation give no indication where they lie within the target.
FIGURE 10.3 Accuracy and precision illustrated by target shooting.
Reprinted with permission from Handbook of Environmental Analysis, 2nd
Ed., Genium Publishing, Schenectady, N. Y. (1995).

The results may be adjacent to each other, to one side of the center, or they
may bracket the center. Rather, the analyst must know that every result on
the sample is going to lie within a small area (lower left target), then an
adjustment (percent recovery) can be applied to the result by the end user of
the data to arrive at the correct value. This is called bias correction. The
laboratory never applies a bias correction to the data it generates. Instead,
the laboratory should supply sufficient accuracy and precision data so that
the end user can perform a bias correction.

4.0 WHAT IS THE LOWEST LEVEL OF

ANALYTE THAT CAN BE DETECTED IN
THE SAMPLE?
This is a difficult question to answer and one that is the subject of intense
debate within the analytical and regulatory community. The answer to this
question also has some subtle, but important ramifications. U.S. EPA has a
procedure (Title 40, Code of Federal Regulations, Part 136, Appendix B)
that is mandated under many regulatory programs for determination of the
method detection limit (MDL). The MDL procedure attempts to define the
minimum quantity of analyte that can be determined as different from zero
with 99% confidence. In short, this procedure requires spiking reagent
water with the target analyte at a level two to five times the expected
detection limit. Seven aliquots of the spiked water are analyzed by the
complete test procedure. The mean and standard deviation of the seven
determinations are computed, then the standard deviation is multiplied by
the one-tailed t statistic corresponding to one less than the number of
replicates analyzed (3.143 for seven repetitions). If the final result fits the
two- to five-times criterion of the original spike level, then the method
detection limit is considered to be valid. An example and calculation of a
method detection level is worked out in detail in the 4th edition of
Handbook of Environmental Analysis (Smith, 1999).
One of the ramifications that arises from the use of this definition and
procedure is that most of the data supplied with the published methods
concerning detection limits are derived from studies conducted on spiked
samples of reagent water, not real samples. The method detection level is
determined in a “perfect” matrix, and a simple correction consisting of
multiplying the MDL by a factor resulting from the use of a smaller sample
in an attempt to obtain usable data for nasty samples is not really justified.
An additional ramification is that the MDL is a measure of the precision of
the method and has little to do with any accuracy considerations. Another
important ramification concerns the calibration procedure, in which it is
necessary to include an MDL concentration standard as a calibrated point.
The subtle part of the definition stems from how the end user of the data
views a BDL (below detection limit) result. Some users give the sample a
value equal to the MDL, whereas others assign the sample one-half the
MDL. Still others decide that zero is the proper interpretation. Zero is
acceptable until a better method or instrument is used for analysis of the
sample, in which case, what once was zero now has a confidently assigned
value.
As far as the water and wastewater laboratory is concerned, it should
follow the procedure that is given in 40 CFR 136 Appendix B and attempt
to duplicate the detection limits presented in the published method.

5.0 IS THE DETECTED PARAMETER

ACTUALLY IN THE SAMPLE?
This question actually has two parts. The first is a question of identification
and should perhaps be rephrased as, “Does the analytical signal that I have
obtained arise from my target analyte, or is it a result of analytical
interference?” Most methods have an introductory section that discusses the
presence of interferences with the analytical procedure.
These interferences can either contribute to or detract from the
analytical response. Sulfamic acid is added to the still pot of the manual
distillation procedure for cyanide to prevent formation of cyanide from
nitrite and organic materials present in the sample (SM224500-CN−C).
Another example is the presence of a signal in a target analyte retention
time window on a chromatogram resulting from use of a gas chromatograph
with an electron capture detector. The signal can arise from a large number
of compounds other than the target analyte and, thus, reanalysis on a
column with different retention characteristics is required for reliable
identification.
When I was in school, the following saying was going around: “Six
months in the laboratory will save you a week in the library”. This is a left-
handed way of expressing the time-consuming nature of laboratory work
and pointing out that other workers have probably had the same
experiences. Analysts can save time by taking advantage of others’
experiences and knowledge. The section on interferences in each method is
a short recitation of problems with the procedure and solutions that other
technicians have developed over the years. It is much easier to read and
follow the advice given there than to attempt to rediscover the solution.
The second part of the question about the actual presence of the target
analyte in the sample assumes that the target analyte was correctly
identified and a result has been generated from the analysis. Now, the
concern is whether the analyte was originally in the sample or was added to
the sample during the sample collection and analysis processes. Laboratory
contamination is the inadvertent addition of target analytes to the sample. It
is a rampant problem throughout the environmental industry and, unless
active measures are taken to prevent it, all the results generated from the
laboratory must be suspected of representing the level of contamination in
the laboratory rather than an indication of what is in the sample.
Laboratory contamination is determined by the analysis of a reagent
blank every time a sample analysis is performed. A reagent blank is a
sample of analyte-free water that is taken through the entire analytical
process and analyzed as a sample. The results from this quality control
should be a nondetect of the target analyte. If the target analyte is found in
the reagent blank, it is poor analytical practice to assume that the same level
of contamination exists in all samples and that the correct result can be
obtained by subtracting the blank value from the initial sample result. A
much preferred decision is to stop the analysis, find the source of the
contamination, correct the source of contamination, and then resume
analysis.
A reagent blank is not the zero point on the calibration curve. An
absorbance blank is sometimes used to set the background absorbance on
the spectrophotometer to zero, but it is not a substitute for a reagent blank.
These are separate quality controls.
Sources of laboratory contamination vary. It can become an exercise in
ingenuity to find a source, particularly if the contamination is intermittent.
Sources of laboratory contamination that have been documented include air
fresheners, perfumes, technicians’ dirty hands, rubber gloves worn by
technicians, dirty glassware, the soap used to clean the glassware, improper
grade of reagents, exhausted reagent water systems, improper order of the
cartridges in the reagent water system, mislabeled reagent bottles, plastic
bags used as containers of bulk chemicals, fingernail polish, carpet glue,
and many others. This is not to suggest that contamination can arise from
only the laboratory. It can also occur in the sampling process through use of
improperly cleaned sample containers or sampling tools, contaminated or
inappropriate preservatives, or carryover contamination between
consecutive samples. Because of the wide variety of contamination sources,
it is imperative to maintain a program of detection through the use of
reagent blanks with every analysis.

6.0 BATCH ANALYSIS

Application of the procedures and techniques relayed in the preceding
section will serve to control the analysis and generate reliable results on
each sample. However, there is an economic reality that must be addressed
in most situations. This has already been touched upon in relation to the
importance of precision over accuracy. To obtain the best analytical results,
a duplicate and matrix spike should be performed on each and every
sample. However, there is not enough time available to do this, and the cost
of the extra chemicals and personnel used in the multiple analytical
repetitions is prohibitive. In a compromise to maintain quality control over
analysis, but decrease overall expense and time, the concept of batch
analysis has evolved.
A batch is defined by U.S. EPA and most other federal and state
regulatory agencies as a set of 10 to 20 samples that are of a similar matrix,
reportable under a similar regulation, and processed as a group within a
single period of time. The minimum quality controls that are performed
with each batch consist of a reagent blank, a calibration verification, sample
duplicates (at a rate of two per 20 samples), and a matrix spike (see
benchsheet in Figure 10.4). Other quality controls can be added to the
minimum, such as laboratory control samples, duplicates, matrix spike
duplicates, quality control check samples, postdigestion spikes, surrogate
standards, system performance check standards, and many others
encountered and required as part of the different methods. The results for
these quality controls are generalized to represent the expected results of
similar controls applied to each and every sample in the batch.
An example of a suitable set of samples for a batch analysis is a group
of samples of water drawn from all the drinking fountains in a building for
lead and copper analysis. Another example is a group of groundwater
leachate samples from the monitoring wells surrounding a landfill. Still
another is a set of biosolids samples being evaluated for potential land
application. An inappropriate use of the batch concept would be to assign
all these samples to the same batch and make the assumption that the matrix
spike performed on the drinking water fountain sample is applicable to the
biosolids sample.
A precaution that should be observed in batch analysis concerns the
comparability of data. This question arises when unfiltered (also called
total) and filtered (also called dissolved) samples are analyzed and data
differences between the two types of samples are important. The filtered
sample should not exhibit higher levels of analytes than the unfiltered
sample; however, if the samples are processed in different batches or, even
worse, processed by different methods, there is a strong possibility that the
results for the dissolved sample may be slightly higher than the total
sample. Although these differences may well arise from contamination of
the filtered sample because of the filtering apparatus, analyzing the total
and filtered samples in the same batch by exactly the same method will
result in elimination of many of the variables in the analysis and give more
comparable data.
Most laboratories use the batch concept for quality assurance purposes
and, to a large degree, it is successful as long as the aforementioned
guidelines are followed for assigning samples to a batch. Other information
on quality control and quality assurance programs can be obtained from
Part 1000 of Standard Methods as well as variety of books and U.S. EPA
publications on the subject (see Section 8.0, “Recommended Reading”).
FIGURE 10.4 Benchsheet used for batch analysis.

It is important to realize that quality assurance and quality controls are

not something management dreamed up one night for the express purpose
of making analysts’ jobs more difficult. Laboratory work is important for
the continued protection of public health. No measure taken to ensure that
reliable data are available for decisions on public health can ever be
construed as unnecessary.

7.0 REFERENCES
American Public Health Association; American Water Works Association;
Water Environment Federation (2012) Standard Methods for the
Examination of Water and Wastewater, 22nd ed.; American Public
Health Association: Washington, D.C.
Smith, R.-K. (1999) Handbook of Environmental Analysis, 4th ed.; Genium
Publishing: Schenectady, N.Y.

8.0 RECOMMENDED READING

An enormous quantity of books are available concerning environmental
analysis, laboratory procedures, and quality control. They have varying
relevance to the actual performance of tests in the municipal treatment plant
laboratory. For further information about the various approved test
procedures used in the water and wastewater laboratory, the following
references are recommended.

• O’Dell, J. (2002) Methods for the Determination of Inorganic

Substances in Environmental Samples; EPA/600/R-93/100; U.S.
Environmental Protection Agency: Washington, D.C. NTIS
PB94121811.
• U.S. Environmental Protection Agency (2002) Methods for
Chemical Analysis of Water and Wastes; EPA/600/4-79/020; U.S.
Environmental Protection Agency: Washington, D.C. NTIS
PB84128677.

The lists of approved methods for water and wastewater monitoring are
found in Title 40, Code of Federal Regulations, Parts 136, 141, and 142.
The most recent edition, available from the U.S. Government Printing
Office (202-783-3238), is the one to be consulted.
Other useful references concerning quality control and laboratory
techniques are as follows:

• U.S. Environmental Protection Agency (1979) Handbook for

Analytical Quality Control in Water and Wastewater; EPA-600/4-79-
019, NTIS PB-297451 (available from NTIS at 1-800-553-NTIS).
• Keith, L. H., Ed. (1996) Principles of Environmental Sampling, 2nd
ed.; American Chemical Society: Washington, D.C.
• Keith, L. H. (1991) Environmental Sampling and Analysis; Lewis
Publishers: Chelsea, Michigan.
• U.S. Environmental Protection Agency (2005) Manual for the
Certification of Laboratories Analyzing Drinking Water. Criteria
and Procedures Quality Assurance, 5th ed.; EPA 815-R-05-004;
U.S. Environmental Protection Agency: Washington, D.C.
• Taylor, J. K. (1987) Quality Assurance of Chemical Measurements;
Lewis Publishers: Chelsea, Michigan.

Finally, for further information about the correct methods to use when
attempting to follow a test procedure, e-mail the Standard Methods
Manager at NEdman@awwa.org or call 1-303.347.6241. E-mail is best
because a question can then be easily distributed to a number of Standard
Methods volunteers to obtain an answer. The Standard Methods manager
can answer questions about Standard Methods policy or such questions as
which methods are U.S. EPA-approved. For more technical questions, the
manager will distribute the e-mail message to volunteers.
 Appendix A
Molecular and Formula Mass
Compounds are collections of elements in fixed numerical and weight
(mass) proportions. An example of a compound is common table salt,
sodium chloride. For each sodium atom in the compound, there is one
chlorine atom. All samples of sodium chloride will have equal numbers of
sodium and chlorine atoms. Looking on a periodic chart, one will find that
the mass of a sodium atom is 22.990 atomic mass units (amu) and a
chlorine atom is 35.453 amu. If the atomic mass units are replaced with an
equal number of grams, then one has defined the mass of a mole of sodium
chloride, in this case 58.443 g. In general, to find the molecular (or formula)
mass of a compound, the exact number and types of elements in the
compound must be known. Multiplying the mass of each element by the
number of atoms of that element in the compound and summing the results
for all the elements will give the molecular (or formula) mass. The
following are examples:

Sodium chloride (NaCl)

Calcium chloride (CaCl2)

Calcium carbonate (CaCO3)

Hydrates are compounds that have a defined number of water molecules

associated with each molecular unit of the compound. Examples of
frequently encountered hydrates are bluestone (copper sulfate pentahydrate)
and potassium hydrogen phosphate trihydrate. The first compound has five
molecules of water for every unit of copper sulfate, the second has three
molecules of water for every unit of potassium hydrogen phosphate.

Water (H2O)

Copper sulfate (CuSO4)

Copper sulfate pentahydrate (CuSO4·5H2O)

Potassium hydrogen phosphate (K2HPO4)

Potassium hydrogen phosphate trihydrate (K2HPO4·3H2O)

A key point is that 228.219 g of potassium hydrogen phosphate

trihydrate contains exactly 174.174 g of potassium hydrogen phosphate.
The only difference is that the first compound contains water and the
second does not. For example, if a procedure calls for dissolving 5.00 g
potassium hydrogen phosphate in 1000 mL of water, and only the trihydrate
is available, the appropriate amount of the trihydrate to use is determined by
taking the ratio of the mass of the trihydrate to that of the anhydrous
compound and multiplying by the needed 5.00 g.

Needed mass of trihydrate = 5.00 × (228.219/174.174) = 6.55 gi 228

There is just one compound that contains only the elements sodium and
chlorine, and that is sodium chloride, in which the ratio of atoms is 1:1.
Calcium chloride has two chlorine atoms for every calcium, and the ratio is
1:2 of calcium to chlorine. Other elements will combine in a variety of
different ratios. For example, iron and chlorine will combine to form
ferrous chloride (FeCl2) and ferric chloride (FeCl3).The first compound has
a ratio of 1:2, while the second has 1:3. These are different compounds with
different masses, and different chemical and physical properties. If a
procedure calls for use of ferric chloride and ferrous chloride is substituted,
the procedure probably will not work.
All the above examples used inorganic compounds. Organic compounds
are similar in that the chemical formula must be known to calculate the
molecular weight. Organic compounds for the most part contain carbon and
hydrogen with assorted amounts of other elements, although a few organic
compounds without hydrogen are known, such as carbon tetrachloride
(CCl4).

Glucose (C6Hl2O6)

Carbon tetrachloride (CCl4)

 Regardless of whether an inorganic or an organic compound is under
consideration, the formula weight, formula mass, molecular mass,
molecular weight, gram atomic weight, equivalent mass, or other terms for
the same calculation represent the same number of molecules or formula
units, which is called a mole. For example, 153.823 g of carbon
tetrachloride, 180.156 g of glucose, 228.219 g of potassium hydrogen
phosphate trihydrate, and 58.443 g of sodium chloride all contain the same
number of compound units. Further discussion of the concept of the mole
and examples of its use are in Appendices B and D.
A frequently used concept in the environmental laboratory industry is to
report differing forms of an element in terms of that element alone. An
example is reporting ammonia, organic nitrogen, nitrite and nitrate, all in
terms of nitrogen. The formula weight of each of these is illustrated below.

Ammonia (NH3)

Nitrite (NO2)
Nitrate (NO3)

The percent nitrogen is different for each compound form of nitrogen,

and the percent nitrogen for organic nitrogen is unknown because many
different organic compounds can contain nitrogen. A possible guide to
organic material content based on an organic nitrogen determination can be
derived from protein analysis using nitrogen contents of 6.25% for meats,
6.38% for dairy products, and 5.7% for grain products; however, the portion
of organic nitrogen in environmental samples that can be attributed to
protein is probably low. These percent nitrogen values are calculated as
follows:

Nitrogen results can easily be converted to the mass of the ammonia,

nitrite, or nitrate by dividing by the amount of nitrogen, expressed as a
decimal, in the compound. The following example converts 1.00 g of
nitrogen to the corresponding mass of ammonia, nitrite, and nitrate.
 The reason this convention is followed is to allow easy determination of
the total mass of the nutrient nitrogen in the system and calculation of the
percent distribution of the total nutrient load among the different forms. In
the above example, there is a total of 4 g of nitrogen in the system with 25%
being in each of the chemical forms although the exact form of the organic
nitrogen is unknown.
 Appendix B
Molarity and Normality
Manipulating and weighing solid reagents, standards, and samples is
tedious and messy. A further limitation is placed upon weighing by the
physical range of the scales and balances. Many analytical determinations
are in the low part per billion range, which is micrograms per liter, and few
analytical balances will weigh below 100 μg. For these reasons, most
chemistry in the water and wastewater laboratory is performed in solution.
The ability to manipulate solutions with a high degree of accuracy and
obtain known concentrations of chemicals at low levels is the strongest
argument for the widespread use of solution chemistry.
Quantitative solution chemistry is based upon the mole concept.
Molarity is the number of moles of substance per liter of solution. The
molarity of the solution is symbolized by a number followed by an
italicized capital M.

M = moles/liters
Although other quantitative conventions exist, molarity is the concept
most widely used in the analytical laboratory A 1.000 M solution of sodium
chloride contains 1 mole (58.443 g) of sodium chloride dissolved in water
to a final volume of 1000 mL (1.000 L); 1.00 mL of this solution contains
0.00100 of a mole (0.058443 g) of sodium chloride. The analytical balance
can accurately weigh 58.443 g of sodium chloride, but it cannot accurately
weigh 0.058443 g. The 1000-mL volumetric flask and the 1.00-mL
volumetric pipet are used to achieve the desired accuracy from the initial
weighing of 58.443 g.
A 0.100 M solution is prepared by dissolving 0.100 mol of substance in
water and diluting to a final volume of 1000 mL (1.000 L).

0.100 M = 0.100 mol/1.000 L

 An alternative preparation is to dissolve 0.010 mol in 100 mL (0.100 L).
This allows preparation of a known concentration solution in an amount
appropriate for laboratory needs without wasting reagents and taking up
valuable storage space. Note, however, that the accuracy of the knowledge
of the solution concentration has decreased by a decimal place, because of
the limited accuracy of the analytical balance.

0.10 M = 0.010 mol/0.100 L

By knowing the chemical formula of the compound, the desired
molarity, and the desired volume, any solution of known concentration can
be prepared. For example, 75.0 mL of a 0.247 M ferrous ammonium sulfate
solution is needed and the only material on hand is ferrous ammonium
sulfate hexahydrate (Fe[NH4]2[SO4]2·6H2O). The first step is to calculate
the number of moles needed:

Moles = molarity × liters = 0.247 M × 0.075 L = 0.0185 mol

The second step is to calculate the molecular weight of the ferrous
ammonium sulfate hexahydrate.
Fe[NH4]2[SO4]2·6H2O
 Multiplying the number of moles needed by the molecular weight of the
substance gives the amount of material needed to be weighed out on the
analytical balance:

Needed weight = moles × molecular weight = 0.0185 mol × 392.13g/mol =

7.254 g
All that remains to be done is to dissolve this weight of material in
water and dilute to a final volume of 75 mL.
The real power of solutions comes from the ability to make a
concentrated solution from a highly accurate determination on an analytical
balance and then prepare more dilute solutions of known concentration with
a high degree of accuracy. If 1.00 mL of a 1.000 M sodium chloride
solution is diluted to 1000 mL in a volumetric flask, the final concentration
can be calculated using a dilution equation,

where V; is the initial volume, C; is the initial concentration, Vi is the final

volume, and Cf is the final concentration. For this particular example,

A 0.0010 M sodium chloride solution could be made directly by

dissolving 0.058 g sodium chloride to 1000 mL in water, but by performing
a dilution the accuracy of the known concentration is increased by two
decimal places (a factor of 100).
A series of dilutions can be used to develop an intriguing paradox. A
mole contains 6.02 × 1023 molecules, and a molecule is the smallest unit of
a compound. Suppose 180.156 g (1 mol) of glucose is dissolved in water
and diluted to 1000 mL. The resulting solution is 1.000 M and contains 6.02
× 1023 molecules of glucose. Diluting 1.00 mL of this solution to 1000 mL
gives a 0.001 000-M solution, which contains 0.180 156 g and 6.02 × 1020
molecules of glucose. Diluting 1.00 mL of this solution to 1000 mL gives a
0.000 001 000-M solution containing 0.000 180 156 g and 6.02 × 1017
molecules of glucose. Several more serial dilutions are performed as
illustrated in Table B.1.
The question, of course, is how many molecules of glucose are in flask
9, even though 1.801 56 × 10−22 g are there with a high degree of accuracy.
Normality is an archaic but frequently encountered measure of the
concentration of a solution. Normality is based upon the reactions for which
a solution is intended. When acid-base reactions are considered, the concept
of normality is straightforward. A 1.00 normal solution has one equivalent
of acid or base per liter. A 1 M solution of a monoacidic material is also 1
N. (N is the symbol for normality.) However a 1 M solution of a diacidic
material such as sulfuric acid contains 2 mol of acidity per liter and is thus 2
N. The idea is that 1.0 mL of a 1 N solution always has the same amount of
acidity regardless of whether a monoacidic or a diacidic acid is used. Base
solutions can be dealt with in the same fashion.

TABLE B.1 Serial dilutions performed on glucose solution.

 The problem with normality arises when reactions other than acid-base
reactions are considered. A common application is to redox reactions (see
Appendix C). Two examples are illustrated below:

In the first example, there is a 5:1 relationship between iron and

permanganate. A 1-N iron solution would be 1 M, and a 1-Npermanganate
solution would be 0.2 M. In the second reaction, which is conducted under
alkaline conditions, there is a 3:1 relationship, and a 1-N permanganate
solution would be 0.33 M. Needless to say, labeling a 1-N solution of
permanganate and sitting it on the shelf can lead to a great degree of
confusion, particularly when there are a large number of procedures being
performed in the laboratory. Standard Methods is currently moving away
from the normality notation; however, its use will probably linger like a bad
odor for many years.
 Appendix C
Types of Chemical Reactions
There are many, many different types of chemical reactions, and a good
textbook on quantitative inorganic analysis should be consulted for an
exhaustive discussion. However, two—acid-base and redox reactions—are
of particular importance in analysis and will briefly be described here.
Acids are chemicals that donate a hydrogen ion in reaction.
Representative common strong mineral acids include nitric (HNO3),
hydrochloric (HCl), and sulfuric (H2SO4). These ionize completely in water
to form hydrogen ions and a negatively charged anion. There are also
organic acids, such as acetic acid and phenol. Bases are substances that can
either donate a hydroxyl ion (OH−) in a reaction or can directly react with a
hydrogen ion. Ammonia (NH3) is an example of the latter type of base. It
reacts with a hydrogen ion to form the ammonium ion (NH4+).
Representative common strong bases include sodium hydroxide (NaOH),
potassium hydroxide (KOH), calcium oxide (CaO), and calcium hydroxide
(Ca[OH]2). Other common bases are sodium carbonate (Na2CO3) and
sodium bicarbonate (NaHCO3). These materials have been known for
hundreds of years and have common names as indicated in Table C.1.
Acids and bases react in a neutralization reaction. The most common
products are a salt and water. A shorthand notation for the reaction between
hydrochloric acid and sodium hydroxide to produce salt and water is

The first reaction encountered by a child is often the volcano

demonstration in grade school. This is an acid-base reaction between
vinegar (CH3CO2H) and baking soda (NaHCO3) to produce water, a salt
(sodium acetate; CH3COONa) and carbon dioxide (CO2). The release of the
carbon dioxide gas makes the volcano froth:

TABLE C.1 Common names of some laboratory chemicals.

Chemical name Common name

Hydrochloric acid Muriatic acid
Sulfuric acid Oil of vitriol
Acetic acid Vinegar
Phenol Carbolic acid
Sodium bicarbonate Baking soda
Sodium carbonate Soda ash
Sodium hydroxide Caustic soda, soda lye, lye
Potassium hydroxide Caustic potash
Calcium hydroxide Slaked lime
Calcium oxide Lime, quicklime

Redox reactions involve the exchange of electrons between atoms. The

atom that accepts the electrons is being reduced; the atom that donates the
electrons is being oxidized.
Reduction always takes place with an associated oxidation. That is why
the reactions are called redox. Metal atoms are frequently participants in
these electron exchange reactions. An example is the reaction between
ferrous ion and permanganate under acidic conditions mentioned in
Appendix B.
 In this reaction, the iron is donating an electron and is being oxidized.
The manganese ion is accepting the electrons and is being reduced:

It is impossible for electrons to exist floating around loose in solution,

so every electron produced in an oxidation reaction must be immediately
consumed in a reduction reaction.
Chlorine is a powerful oxidant, and its reactions are all of the redox
type. The determination of chlorine by iodometric titration involves two
redox reactions. The first is the reduction of chlorine to chloride ion with
the simultaneous oxidation of iodide ion to iodine. The second redox
process is titration of the iodine with thiosulfate to regenerate the iodide:
 Appendix D
Stoichiometry of Chemical Reactions
Stoichiometry is the quantitative relationship between substances involved in
chemical reactions. It is a reflection of the law of conservation of matter; that
is, matter can be neither created nor destroyed in a chemical change. It is the
fundamental concept and limitation underlying all quantitative chemical
analysis. In the most elemental acid-base reaction between hydrochloric acid
and sodium hydroxide, the stoichiometry of the reaction is that 1 unit of
hydrogen chloride (HCl) will react with 1 unit of sodium hydroxide (NaOH)
to form 1 unit of water and 1 unit of salt. All the elements on one side of the
reaction must be present in the same numbers on the other side of the
reaction arrow. This is what is called a balanced reaction . Without a
balanced chemical equation in hand, no stoichiometry calculations will
successfully predict the experimental results.

This information can be used to predict how much of what will be

required to produce desired amounts of products. Suppose the reaction is
conducted on the 1-mol scale, then, by knowing the molecular mass or
weight, appropriate gram amounts of hydrogen chloride (36.46 g) and
sodium hydroxide (40.00 g) can be combined to give predictable amounts of
salt (58.44 g) and water (18.02 g).
 If only 5.00 g of sodium hydroxide is available, it is possible to calculate
exactly how much hydrochloric acid will be required for complete reaction
and how much salt and water will be produced.

With solutions, these calculations and determinations become simple;

they are based on the solution concentration. For example, how much of a
0.15-M hydrogen chloride solution will be required for complete reaction
with 250 mL of a 0.23-M sodium hydroxide solution? The answer is 383
mL.
 The titration for alkalinity is based on an acid-base reaction, which is
illustrated for hydrochloric acid. The results of this determination are always
expressed as milligrams of calcium carbonate per liter. This reaction requires
2 mol of hydrogen chloride for every 1 mol of calcium carbonate. The
procedure calls for titration of a 200-mL sample with 0.10 M of hydrochloric
acid. If 18.8 mL of the acid is required for the titration, what is the
concentration of calcium carbonate alkalinity? The first step is the
calculation of the number of moles of hydrochloric acid used. This is 0.001
88 mol. The next step is to consider the reaction, which requires two
equivalents of hydrogen chloride to react with each equivalent of calcium
carbonate. If there are 0.001 88 mol of hydrogen chloride used, then only
half this amount, or 0.000 94 mol, of calcium carbonate was present. The
molarity of the calcium carbonate solution is next calculated: 0.004 70 M .
The results are typically reported in terms of milligrams per liter. If the
molecular weight of calcium carbonate is 100.09 g/mol and there are 1000
mg in a gram, then the molecular weight can also be expressed as 100 090
mg/mol. The final result is 470 mg/L, as follows:
 This result should be compared to that obtained from the calculation
given in Standard Methods , where A is the number of milliliters of titrant.
All the same considerations are in the calculation, with the constants of
molecular weight, stoichiometry, and gram-to-milligram conversion
compressed into the multiplier of 50 000, as follows:

Alkalinity, mg/L CaCO3 = (A × N × 50 000)/mL sample = 470

(Standardization of ferrous ammonium sulfate with dichromate from SM18

4500-CI F, Section 2.c.)
The reaction of the standardization is as follows. The preparation and
concentration of the solutions used in the standardization follows:
 Thus, the standardization of the ferrous ammonium sulfate titrant at the
stated concentrations and amounts will require exactly 2.82 mL of the
dichromate solution.
 Appendix E
Laboratory Analyst Training

1.0 INTRODUCTION
2.0 TRAINING GOALS
3.0 TRAINING PROGRAM
4.0 TRAINING DOCUMENTATION
5.0 CONCLUSION
6.0 REFERENCES

1.0 INTRODUCTION
There has been a proliferation of instant chemistry test kits within our
industry. I refer to them as “pseudo-chemistry” because it takes no skill to
generate numbers using these kits. I once taught my 8-year-old son to use
several of them. He was quite proud of his success, but I would never
describe him as an analyst. He completely lacked any understanding of
what he was doing or why. It takes a lot of time and effort to learn all the
skills necessary to be an expert laboratory analyst. The immense popularity
of the test kits is a symptom of the shortage of trained analysts in our
industry. Many people have an expectation of instant gratification, and the
test kits provide both “instant” and “gratification” without any great
expenditure of effort or time. Thus, there is a very low level of
professionalism in our industry.
Aside from the desire to raise the overall level of professionalism
among laboratory analysts, there is also a legal necessity to have trained
analysts perform tests in treatment plants and commercial laboratories. As
described in Environmental Laboratory Data Evaluation (Berger et al.,
1996) and Handbook of Environmental Analysis (Smith, 1999),
demonstration and documentation of the level of training of the analyst is
an important aspect of the use of scientific evidence in court cases. Much
scientific evidence has been refused admission or has been severely tainted
because of a lack of documented training of the “expert”. A noteworthy
example is the photographic analyst who testified for the defendant in the
O.J. Simpson civil trial that the pictures of the shoes presented by the
plaintiffs were faked. It was subsequently brought out during cross-
examination that the “expert” had absolutely no training in photographic
analysis and was probably a fraud himself. The same can and has happened
in regulatory hearings and court cases in which laboratory results are
submitted as evidence. Imwinkelried’s (2014) standard reference lists and
discusses six known weaknesses in analyst training as tempting targets for
legal challenge. They are as follows:

1. The witness is unqualified to vouch for the theory’s validity

○ Lack of understanding of the theory,
○ Lack of theoretical background, and
○ Insufficient theoretical background;
2. The witness is unqualified to vouch for the instrument’s reliability
○ Unfamiliarity with the instrument or technique;
3. The witness was unqualified to maintain the equipment;
4. The witness was unqualified to operate the equipment and conduct
the test
○ Whether a credential is required and
○ Whether the witness possesses the credential;
5. The witness did not use proper test procedures in conducting the
test; and
6. The witness is unqualified to interpret the test result.

It is important to remember that any result generated from a municipal

or commercial laboratory in support of National Pollutant Discharge
Elimination System (NPDES) compliance monitoring requirements,
hazardous waste characterization, industrial pretreatment monitoring
verification, or any of the other myriad regulatory programs has the
potential of ending up under regulatory review or in court—sometimes in
criminal court where the evidentiary requirements are much more stringent.
This implies that all analysts need to be trained if legal defensibility of the
data is to be assured.
This article is broken up into three parts. The first part is a discussion of
the skill areas that an analyst needs to know to be successful. The second
part presents a training program that addresses and meets these goals, and
the third part describes documentation of analyst training and skill.

2.0 TRAINING GOALS

Before we can address the issue of how to train an analyst, we need to
examine what an analyst needs to know to function in a safe and
responsible fashion. First and foremost is knowledge of chemistry
laboratory technique. Back when I was in college, there were fewer than 30
undergraduates in my class in the chemistry degree program, but, in the
Biology Department three buildings down, there were more than 100
students in the biology degree program. I often wondered what biology
majors did for a living if they didn’t go on to graduate school. After I began
working in a commercial analytical chemistry laboratory I found out the
answer: they go to work in analytical chemistry laboratories. This is strange
because none of the biology majors I knew ever took analytical chemistry
as a course; most stopped taking chemistry after taking general, organic,
and maybe biochemistry. This is not to imply that chemistry majors learn
any great amount of laboratory technique in analytical chemistry classes.
Most individuals graduating with college science majors have at best a
smattering of proper laboratory technique that they picked up by accident
during their studies. It’s definitely not because of any systematic training
program in proper laboratory technique.
Second, a detailed knowledge of chemistry laboratory safety is an
absolute necessity. High schools, colleges, and universities are notorious for
having a distressing lack of awareness of safety in their laboratories. Sure,
they comply with the fire regulations and provide extinguishers, blankets,
showers, and eye-wash stations, but that is about it. Most teaching
laboratories represent accidents waiting to happen. The cavalier attitude
toward chemical toxicity and other health hazards, and especially toward
responsible disposal (“pour it down the sink”), is prevalent. These attitudes
and practices have to be changed to bring analysts into compliance with
Occupational Safety and Health Administration (OSHA) and other
regulations dealing with workplace safety, chemical exposure, and proper
waste disposal practices. There is also a distinct need to instill common
sense practices in the analyst regarding chemicals and laboratory
equipment.
Third, a detailed awareness of environmental regulations is needed. The
analyst operates within a compliance monitoring framework, and certain
test methods are approved for use while others are not. The mere possession
of the most recent edition of Standard Methods for the Examination of
Water and Wastewater (APHA et al., 2012) will not satisfy regulatory
needs. First, the most recent edition is probably not the approved edition for
compliance monitoring methods, and, second, not all the methods found in
Standard Methods are approved for use. The analyst must be familiar with
the Code of Federal Regulations, plus any state environmental regulations
that govern the analyst’s sphere of responsibility. The fastest way to have
analytical results rejected as evidence in court cases is to use unapproved
methods to generate the data.
Fourth, an in-depth knowledge of the chemistry of the test is necessary.
You can frequently recognize the untrained analyst when you hear the
question, “Is this step really necessary?” Persons writing test methods do
not waste time and space by adding unnecessary steps to the procedure.
Even though the chemist following the procedure may not know exactly
why each step is performed, he or she has sufficient trust and experience to
recognize that each step is important to the successful generation of the
result. On the other hand, the analyst may find that the sample needing
analysis is not completely amenable to the written test procedure and that
some modifications are necessary. Further, it is advantageous to know the
chemistry of the test procedure so that the analyst has an awareness of the
limitations of the test. No test procedure works equally well for all samples
and the analyst must know the symptoms that indicate when the test is not
working. Learning to avoid or correct for test interferences is the hallmark
of the expert analyst (Baird and Smith, 2002; Smith, 1999, 2003).
 Fifth, a knowledge of the use and interpretation of quality control
procedures is necessary (Smith, 2000, 2001, 2005). Without quality control,
there is no confidence in results. Quality controls are always part and parcel
of every approved method. Knowledge of how to interpret the quality
control results is needed to assist the analyst in making the determination of
whether the test is working or not working for any particular sample. The
knowledge of specific quality controls must also be accompanied by
knowledge of where they fit within the overall quality assurance program.
We can summarize these job knowledge goals as follows:

• General laboratory technique,

• Safety and chemical hygiene,
• Regulatory requirements,
• Chemistry of specific test procedures, and
• Quality control.

3.0 TRAINING PROGRAM

When training is considered, the trainee typically is a new employee who
we would like to make productive as soon as possible. It is not in our best
interest to teach laboratory technique, then, when that subject is finished,
move on to the next item on the list. It is also not productive to sit the
employee down in a class for 1 or 2 weeks, drill them with everything they
possibly need to know, and then move them into the laboratory and expect
them to remember or understand everything that was covered. Learning to
be an analyst takes a long time and the most rewarding experience occurs
when the skills learned in the laboratory are immediately reinforced with
material discussed in the classroom.
Successful laboratory training is never passive. The simple presentation
of a block of information in a class is rarely sufficient by itself. It must be
followed up with supervised practical application on the job. Further, the
person receiving the training must be evaluated and evaluation must
continue over the lifetime of the employee. This is especially true in the
case of laboratory work, where, over time, an analyst will introduce
shortcuts in his or her work. Sometimes, these shortcuts will introduce
valuable savings in time and materials; however, the more common
occurrence is that the quality of the work suffers. Periodic re-evaluation
serves to identify when the quality of work is deteriorating.
I have found that the most successful approach is that of tiered training.
The tiers are characterized as “introduction”, “development”, and
“maintenance”. The introduction consists of 8 hours of formal classroom
lecture/demonstration that the student receives over the first 3 weeks of
employment, while simultaneously receiving on-the-job training by the
immediate supervisor. The 8 hours of formal class contain 2 hours of
chemical hygiene and safety training, 1 hour of radiation safety training, 3
hours of quality assurance training, 1 hour of Laboratory Information
Management Systems (LIMS) orientation, and 1 hour of administration
orientation. The quality assurance training is broken up into three segments,
with one segment presented each of the first 3 weeks of employment. This
allows the employees to digest the information and integrate it to the on-
the-job instruction they are receiving from their supervisor. A lesson plan
for the quality assurance training is presented in Table E.1.
All employees are required to complete the introductory training,
regardless of job area. At the next level of training, the topics are more
directed toward job function. Field service personnel take an OSHA and
hazardous waste field sampling course, while all laboratory analysts
participate in a 40-hour analyst course. Many states require laboratory
analyst licensing for persons performing drinking water or wastewater
analysis. The licenses are obtained through written examination. The
analyst course is oriented toward helping employees pass the certification
exams. The topics of the analyst class are chosen to present a wide variety
of general laboratory knowledge (Table E.2) that encompasses the subject
range of the certification exams. The class is presented in 1-hour segments,
with two classes a week during the lunch hour. The participants eat lunch
and listen to the class. The entire course takes 20 weeks and, at the end of
the schedule, the presentations cycle back to hour number 1 and are
repeated. The presentations are largely independent and analysts can join at
any point in the schedule. Problem sets (homework) are frequently handed
out and the answers discussed at the next meeting of the class.
TABLE E.1 Introduction to quality assurance lesson plan.

Definitions of Quality assurance and quality control are defined.

quality
assurance &
quality control
Analytical Definitions and how the two requirements of
validity & environmental regulatory analysis are described. The legal
legal accountability and personal responsibility of each analyst
defensibility for their work are described.
Quality The purpose, use, and frequency of updates for the quality
assurance assurance manual are described.
manual
Standard How SOPs are written and updated is described. The
Operating format and authority of the SOP is described.
Procedures
(SOPs)
Approved The types of regulatory analysis performed and the
methods location of approved methods are described, along with
the role of 40 CFR and other regulatory documents.
Holding times Definition of holding times, how to calculate, and location
of holding time lists in quality assurance manual
Preservatives Definition of preservatives and location of lists in quality
assurance manual and regulatory sources
Containers Location of container lists in quality assurance manual and
other regulatory sources
Accuracy & Terms are defined, target analogy, and methods of
precision calculation
Batch QC Batch quality control is described and how it is
implemented.
Control charts How control charts are generated and maintained, where
control limits come from
MDL Mean and standard deviation are defined and illustrated;
MDL is defined and U.S. EPA method of determination
presented.
Significant Source of concept and use in laboratory for data reporting
figures
Error Demonstration of correct procedure for correcting
correction benchsheets
Unit ppt, ppb, ppm, and % defined and related to ug/L, mg/L,
conversions ug/kg, and mg/kg
Volumetric Definition of Class A calibrated glassware and use
glassware
Volumetric Definition and demonstration of correct use
pipettes
Volumetric Use of water density and temperature to calibrate non-
calibration Class A devices
Representative Discussion and demonstration of obtaining representative
samples samples from a container
Dilution Discussion of dilution equation
factors C1 × V1 = C2 × V2
Percentage Percentage moisture determination described and dry
moisture weight reporting demonstrated
Performance Performance evaluation samples and required
Evaluation documentation are described
samples
Analyst Reasons, requirements, and forms are discussed
certification
Organizational Laboratory management structure described
chart
Client Role of project managers
contracts
Flow of work Receipt, log-in, and location of samples. Use of LIMS for
 sample management described
Record Retention and storage of records. Legal requirement for
keeping records
Bar-code Use of bar codes for tracking samples in LIMS
system
Building Custody and security of samples and building, wearing
security name badges, escorting visitors

Obviously, a 1-hour class on semivolatile organics is not going to make

a technician proficient in the analysis of organic target analytes using a gas
chromatograph or any other instrument or involved technique. For these job
positions, not only will the analyst complete the 40-hour analyst training,
they will also receive specialized training from either the instrument
manufacturer’s representative or other person certified to present the
training. Although it is sometimes advantageous to hold the specialized
training in-house, for the most part these courses require the analyst to
travel to the manufacturer’s facility, particularly when extensive instrument
hands-on instruction is involved. Examples of these course titles include
basic and advanced Gas Chromatograph Operation and Maintenance, Gas
Chromatograph-Mass Spectrometer (GC-MS) Operation and Maintenance,
Mass Spectral Interpretation, Inductively Coupled Plasma (ICP)-Atomic
Emission Spectrometry Operation and Maintenance, High Performance
Liquid Chromatograph Operation and Maintenance, ICP-Mass
Spectrometer Operation and Maintenance, Ion Chromatograph Operation
and Maintenance, Atomic Absorption Spectrometer Operation and
Maintenance, Capillary Ion Electrophoresis Operation and Maintenance,
Microbiological Species Identification, and LIMS.

TABLE E.2 Analyst class schedule.

Hour
1 Molecular formulas and names of chemicals, Part 1
2 Molecular formulas and names of chemicals, Part 2
3 Molarity, solutions, and dilutions
4 Stoichiometry and calculations, Part 1
5 Stoichiometry and calculations, Part 2
6 Chlorine chemistry and analysis, Part 1
7 Chlorine chemistry and analysis, Part 2
8 Chloride and fluoride analysis
9 Nitrate, nitrite, total Kjeldahl nitrogen (TKN), and ammonia
analysis, Part 1
10 Nitrate, nitrite, TKN, and ammonia analysis, Part 2
11 Dissolved oxygen, BOD, and COD, Part 1
12 Dissolved oxygen, BOD, and COD, Part 2
13 Dissolved oxygen, BOD, and COD, Part 3
14 Fecal and total coliforms, Part 1
15 Fecal and total coliforms, Part 2
16 Regulatory programs and regulations
17 Sample receipt, chain-of-custody, and LIMS
18 Sampling, holding time, container and preservatives
19 Accuracy, precision, and MDL (data quality objectives)
20 Regulatory reporting levels (NPDES, Safe Drinking Water Act,
Corps of Engineers, etc.)
21 Reagents, standards, and laboratory water (specific gravity and
conductance), silver nitrate test
22 Glassware and volumetric ware
23 Calibrations
24 Temperature measurement and conductivity
25 Solids, Part 1
 26 Solids, Part 2
27 Phosphorus
28 Sulfate, sulfite and sulfide
29 Metals
30 Volatile organics
31 Semivolatile organics
32 Turbidity and color analysis
33 Odor and taste
34 Color analysis
35 Oil & grease, and total petroleum hydrocarbons
36 Surfactants
37 Cyanide
38 pH, alkalinity, and hardness, Part 1
39 pH, alkalinity, and hardness, Part 2
40 Jar test

At the end of development training, and continuing on throughout his or

her career, formal evaluations of the analyst are performed. Many states
require licensing of water and wastewater analysts. A common requirement
for licensing is obtaining a passing grade on a certification exam often
sponsored by the state. The certification exams are prepared by the
Association of Boards of Certification (ABC; Ankeny, Iowa). For states that
do not sponsor a certification exam, ABC will administer certification
exams to individuals through a designated proctor. Both drinking water and
wastewater laboratory exams are available. The water laboratory
examination has only one level, whereas the wastewater laboratory exams
have four levels of proficiency as follows:

• Class I—Plant operators who perform process control tests: alkalinity,

biochemical oxygen demand (BOD), carbonaceous biochemical
 oxygen demand, residual chlorine, coliforms, color, dissolved
oxygen, odor, oxygen uptake, pH, sludge density index, solids,
turbidity, and so on;
• Class II—Intermediate-level treatment plant laboratory analysts
who perform regulatory monitoring and process control: all topics
in Class I plus chemical oxygen demand (COD), conductivity,
nitrogen, oil and grease, phosphorus, and so on;
• Class III—Advanced laboratory analysts in larger municipal or
commercial laboratories: all topics in Class II plus bioassay,
cyanide, inorganics, metals, organics, phenols, and so on; and
• Class IV—Expert laboratory analysts/laboratory managers: all
topics in Class III plus detailed instrumental analysis (AA [Atomic
Absorption], ICP, gas chromatograph, GC-MS, ion chromatograph,
etc.) and management skills.

Most state licensing programs for laboratory analysts are based on the
Class II examination. I have found this to be an excellent evaluation of the
basic skills required of a laboratory analyst. The higher-level exams are
useful in evaluating an analyst’s progress as it develops during his or her
career. Water Environment Federation, in association with ABC, published
a study guide to help analysts prepare for these exams (WEF, 2000).
There are no written examinations available that measure analyst skills
outside of the water or wastewater arenas (ABC used to offer an
Environmental Laboratory Analyst Exam, but it is no longer available). Our
experience has been, however, that the wastewater examinations are an
accurate measure of the analyst’s general success in other areas of
environmental analysis, with the single exception of knowledge of specific
regulations. Part of the reason, I believe, lies in the requirements of the
compliance monitoring programs, which are much more stringent than
those of other U.S. Environmental Protection Agency (U.S. EPA) and state
regulatory areas. It is easier to move from a very strict regimen of testing to
a less stringent protocol, rather than vice versa.
Evaluations of the analyst’s hands-on technical capacity are separate
from evaluations of the analyst’s general knowledge, but are no less
important. Initial demonstration of ability (IDA) represents method-specific
evaluations of the ability of the analyst to perform a particular test before
working on real-world samples. Many U.S. EPA test methods contain a
detailed description of a required IDA. Typically, it consists of four to seven
repetitions of the test on a spiked reagent water sample, with the obtained
accuracy and precision of the replicate results compared to listed
performance standards. Successful completion of a Method Detection Limit
(MDL) study, as described in 40 CFR 136, Appendix B, is an evaluation
that is performed at least once a year for each test analyte for which the
analyst is responsible.
Other important evaluations include performance audits of analyst skill
in following test procedures. A performance evaluation sample is a blind
test of the analyst’s ability to obtain an acceptable result on samples
containing unknown concentrations of target analytes. U.S. EPA used to
provide performance evaluation samples twice a year for both drinking
water (water supply) and wastewater (water pollution); however, the
program was terminated in the 1990s. Currently, performance evaluation
samples that cover a wide range of target analytes in a variety of matrices
are available from commercial suppliers. Performance evaluation sample
results serve as an excellent measure of analyst capability and are
frequently the first indicator that there are egregious problems in the way
the analyst is following a test protocol.
Another form of performance audit is obtained in conjunction with
preparing and updating performance expectations (data quality objectives)
for detection/reporting limits, accuracy, and precision. This evaluation
reviews quality control results over a period of time and compares the
current results with either historical laboratory data or method-specified
performance.
Periodic system audits are valuable evaluations. System audits take
many forms and may be conducted by either in-house quality assurance
personnel or by outside visitors to the laboratory. Most state certification
programs and many federal programs require an on-site visit and audit as
part of the certification or validation process. The visit may be conducted
by a state or federal official, or the audit may be contracted out to a third-
party accreditation organization. These audits from persons outside the
laboratory are extremely valuable as an independent source of evaluation.
Often, we who work in the laboratory get so involved with the day-to-day
operations that we can’t see the forest for the trees. Our objectiveness is
further clouded by the personal relationships that exist in the laboratory,
frequently leading to the decision that, “It’s not really that big of a deal and
I don’t want to hurt their feelings”. Regardless of the feelings of the analyst,
failure to follow prescribed procedures hurts the laboratory. Outside
auditors are free from these personal relationships and give a more
objective evaluation.
System audits compare, in detail, what the analyst is actually doing on
the bench with what is prescribed in the official approved method, the
laboratory’s quality assurance manual, and the laboratory’s standard
operating procedure (SOP). An annual review of the SOP by the analyst and
the laboratory’s most knowledgeable analyst is an excellent procedure. A
system audit should always be triggered by an unacceptable result on a
sample.

4.0 TRAINING DOCUMENTATION

As was discussed in the introduction, the training of an analyst is a
necessary part of the foundation evidence to support scientific data in court
cases and regulatory reviews. Proof of the training is best supported through
documentation to give credence to any testimony claiming proper training.
This suggests that individual training files need to be maintained for each
employee.
The file should contain the analyst’s resum©, which summarizes any
formal technical training or experience that he or she had before
employment at the laboratory. A one-page resum© is often more than
adequate. If the analyst has attended or graduated from a college or
university, a copy of the transcript is often useful.
It is necessary to document each class or course that the analyst attends
while he or she is at the laboratory. If the course consists of more than one
session, having a sign-in sheet or spreadsheet is useful for keeping up with
who has attended which session. Once a course of study is completed, the
responsible person for the training issues a signed certificate. Necessary
information for the certificate are the name, date(s), and reference source
for the course, the name of the student, and the name and dated signature of
the instructor. For in-house courses, it is frequently beneficial to attach a
copy of the lesson plan to the certificate in the file, especially if the record
is to be admitted in a court case. For courses taken outside the laboratory, a
copy of the training completion certificate should be included in the file.
Evaluations also need to be documented. Passage of a certification exam
is accompanied by issuance of a certificate by the sponsoring organization.
A copy should be in the file. If any license is granted by the state, a copy of
that license and any renewals should also be in the training file.
Copies of completed initial demonstration of ability (IDA) and method
detection limit (MDL) studies should be available in the training records.
Acceptable results on performance evaluation samples should be
documented. It may be beneficial that a copy of the official report from the
organization responsible for the performance evaluation sample be attached
to the certificate.
Written reports of audits, regardless of whether they are internal or
external, should be included in the training file. They should indicate by
whom and when the audit was conducted, the findings of the audit, and any
corrective actions recommended to correct deficiencies. Most audits require
a written response, and a copy of the response should be attached to the
audit report. Any documentation that is generated as a corrective action to
the deficiencies should be copied and included.
There are a number of ways to keep these records. In some laboratories,
the personnel, technical training, and safety training records may be kept
together in a single folder. In other laboratories, where the
personnel/finance, training, and chemical hygiene/safety functions are
managed by different departments at different locations, the three sets of
records may be separated. Regardless of where the records are stored, it is
important to give the employee a copy of the record and to keep at least one
copy for filing, either in paper form or electronically. If the training records
are required to be produced in court, it is often 3 to 6 years after the event
of the analysis, although I was recently involved in a civil law case and
records up to 25 years earlier were desired, with several of the analysts
retired or, in one instance, deceased. Still, the necessity existed to prove the
individuals were trained and capable of doing the test procedure at the time
the test was performed.
It is beneficial to the training manager to maintain training summaries.
These allow one to tell at a glance who has had what training. Spreadsheets
can be generated in-house or by one of the available commercial computer
programs. Hardcopy printouts of these training summaries are useful in
marketing efforts by the laboratory. Project proposals can be enhanced
through inclusion of lists of employees who hold particular
certifications/licenses, such as OSHA field sampling or drinking water
analysts.

5.0 CONCLUSION
No one is born an expert analyst. Training is absolutely necessary to
produce a knowledgeable, competent laboratory worker. In this appendix, I
have described a program that I have used for years with some degree of
success. Hopefully, with some situational modifications, this will work in
your facility.

6.0 REFERENCES
American Public Health Association; American Water Works Association;
Water Environment Federation (2012) Standard Methods for the
Examination of Water and Wastewater, 22nd ed.; American Public
Health Association: Washington, D.C.
Baird, R. B.; Smith, R.-K. (2002) Third Century of Biochemical Oxygen
Demand; Water Environment Federation: Alexandria, Virginia.
Berger, W. H.; McCarty, H.; Smith, R.-K. (1996) Environmental Laboratory
Data Evaluation; Genium Publishing: Schenectady, N.Y.
Imwinkelried, E. J. (2014) The Methods of Attacking Scientific Evidence,
5th ed.; LEXISNEXIS: Dayton, Ohio.
Smith, R.-K. (1999) Handbook of Environmental Analysis, 4th ed.; Genium
Publishing: Schenectady, N.Y.
Smith, R.-K. (2001) Interpretation of Inorganic Data; Genium Publishing:
Schenectady, N.Y.
Smith, R.-K. (2000) Interpretation of Organic Data; Genium Publishing:
Schenectady, N.Y.
Smith, R.-K. (2003) Guide to Environmental Analytical Methods, 5th ed.;
Genium Publishing: Schenectady, N.Y.
Smith, R.-K. (2005) Interpretation of Biological Data; Genium Publishing:
Schenectady, N.Y.
Water Environment Federation (2000) Wastewater Laboratory Analysts’
Guide to Preparing for the Certification Examination; Water
Environment Federation: Alexandria, Virginia.
 Index

A
Accuracy, 187
Acetal resins, subjecting racks, 55
Acetic acid (in vinegar), 2
Acid-base reaction, 209-210
Acids, 209
ACS Reagent Grade, 6-7
Alkalinity, 173
Alkalinity formula, 178
American Chemical Society (ACS), 6
Analytical balance, 16
equipped with draft shields, 19
location for, 18
operation of, 16
weight determination on, 175
Analytical chemistry, 101
Analytical determinations, 205
Analytical sense, 183-184
Anhydrous reagents, 4
Anion/cation balances, calculation of, 181-182
Annealed glass, 28
ASTM Class water, quality parameters for, 11, 12t
Azeotrope, Dean-Starke trap, 135

B
Balances, 15
calibration of, and certified weights, 19-20
dust and dirt, 18
high-frequency vibrations, 19
 humidity control, 18
low-frequency vibrations, 19
maximum accuracy of, 21
operation and daily maintenance, 24-25
samples and reagents, 22
tare, 21
types of, 15-17
vibration dampers, 19
weight determination, maximum accuracy in, 21
with gravity, 18
Balances and volumetric ware
determinative or stoichiometric reagent, 102
use of, 102-106
Bases, 209
Batch, 195
Batch analysis, 194-196
comparability of data, 195
minimum quality controls, 195
quality assurance and quality controls, 196
quality assurance purposes, batch concept for, 195-196
Beakers, 33
Below detection limits (BDLs), 183, 189
Berzelius beaker, 33
Biochemical oxygen demand (BOD), 188
Blending PVC with phthalate plasticizers, 55
Blown glass filters/fritted glass, 146-147
Borosilicate glasses, 28
Burettes, 80-82

C
Calibration graph, 179
Calibrations, 111-130
blank, 126
calibration curve, characteristics of, 119-120
calibration equation, use of, 122
calibration plot, 119
censuring, 122
 detection limit, 127
direct standardization, 112-113
frequency of, and calibration checks, 128-130
environmental change, 129
monitoring a calibration, 129
indirect standardization, 113
linear calibration, 121
linearity, 121
multiple standard additions, spike solutions, 127-128
multipoint graphical techniques, 116-127
procedure for calibrating an instrument, 114
regression coefficients, 123
regression equation, 123
saturation, 119-120, 121f
scattergram, 117
solutions, standardization of, 111-112
Centrifuge tubes, 38
Ceramic boats, 22
Ceramic mortar, 52
Ceramic or porcelain materials, 50-52
Ceramic or porcelain materials, uses in laboratory, 51
Chemical suppliers, 4, 11
discrepancies in results, 11
potassium permanganate, 11
Chemical testing, chemical properties, determination of, 1
Chemical-purity grading standards, 6
Chemicals
acid storage cabinet, 10f
chemical and physical properties of, 2
chemically resistant plastic trays, 10
concentrated, 2
concentration (also, concentrated or dilute), 2
dilute, 2
grades of, 6-9
hazard classes, 10
storage of, 10-11
strength (also, strong or weak), 2
 types of, 2
Chemistry laboratory technique, 219
training documentation, 228-229
training goals, 219-220
training program, 220-228
Chloride-free water, 13
Class 2 weights, 19
Class A volumetric pipette, 174
Class l weights, 19
Class S or Class 2 weight, 20
Class S weights, 19
Class S-1 weights, 19
Colorimeters, 157-158
absorbance of light by solutions, 159
ammonia determinations, Nesslerization method, 161
lamps, aging of, 158
maximum absorbance, 159-161
monochromator, 158-159
sodium 2-(parasulfophenyl-azo)-1,8-dihydro-3,6-naphthalene
disulfonate (SPADNS) procedure for fluoride analysis, 161
Colorimetric determinations, 157-164
Compounds, 199
Condensers, 39
Condensers, coolant as a source 41-42
Conversion formula, 178
Coolant nipples, 41
Copper tubing, 67
Cross-linked high-density polyethylene (XLPE), 54
Cross-linked HPDE, 54
Crucibles, 51-52
Culture tubes, 37-38

D
Deliquescence, 4
Deliquescence, determinative reagent, preparation of, 102
Dilutions, preparation of, 106-107
1 + 4 and 1:5 notations, 107
 decade dilutions or serial dilutions, 107
stock solutions, preparation of, 107
Direct standardization, 112-113
acid standard solutions, 113
ferrous ammonium sulfate (FAS) titration of total residual chlorine
using N,N-diethyl-p-phenylenediamine, 112
sodium or potassium hydroxide with the primary standard potassium
hydrogen phthalate (KHP), 112-113
standardization of solutions of bases, 112
Distillation, 131-142
ammonia still, 133-134
apparatus configurations, 136
boiling flask, superheating of, 137
bumping, 137
ceramic Buchner funnel, 138
concentration of the target analytes, 132
conventional stillhead/condenser combination, 136
cyanide stills, 132-133
ebulator, purpose of, 137
fluoride distillation apparatus, 134
gas dispersion tip, 132
gross separation of a target analyte, 132
heating mantles without a thermocouple temperature-monitoring
probe, 140
Kuderna-Danish concentrator, 138
magnetic stir bar, 136-137
reagent water, final purification of, 132
removing excess solvent from a target analyte, 135-136
total phenols distillation apparatus, 134-135
uses in environmental laboratory, 132
Dry weight corrections, 179-180
Dual-purpose pipette, 80

E
Electron capture detector, 193
Electronic balances, 21
Electronic balances, accessories, 21
 polonium cartridge, 23
set of forceps, 24
spatulas and scoops, 24
weighing papers, 23
Electronic-indicating thermocouples, 97
Equivalent (eq) or milliequivalent (meq) per unit volume or mass, 181
Erlenmeyer flasks, 34
low-actinic Erlenmeyer, 34
rubber stoppers, 34
sterilization of culture media in, 36
storage of solutions and materials, 36
with screw caps, 35f
Explosion-proof and chemically resistant refrigerators, 10
Exposed-wire thermocouples, 97

F
Filter photometers, 158
Fires, 3
Fluorinated HDPE container with a Teflon® liner, 62
Fluorinated polyethylene or polypropylene molded items, 54
Fluorocarbon resins, 55
Fluorocarbons, 54
Fritted filters, 147
Fritted glass Buchner funnels, 46
Funnels, 42-43
air escape, 42-43
Buchner funnel, 44-45
fritted disks, 45
paper folding, 43-44
pleated paper/fluted paper, 44
uses of, 42-43

G
Glass
annealing, 28
as a super-cooled liquid, 28
ASTM E676, 30
 ASTM E677, 30
ball and socket joints, 30
joint configurations, 29-30
leak-free joint, 31
metallic impurities in, 29
O-ring joint, 30
percentage-level materials found in, 29
spherical joint, 30
stopcock plugs, 32
stopcocks, 31
surface defect, 28
tapered joint, 30
Teflon® bushings, 32
Teflon® stoppers, sleeves, and stopcock plugs, 32
tempering, 28
types and characteristics of, 27-50
water-based lubricants, 32
Glass bottles, 36
Glass fiber filters, 144-145
Glass fiber filters, total suspended solids (TSS), 144
Glass graduated cylinders, 80
Class A or Class B ASTM standards, 80
Glass thermometers, 95, 97
Glass-to-glass joints, 32
Glassware, 28
Glassware cleaning, 28-29
Glassware cleaning, potassium hydroxide-isopropanol baths, 29
Gooch, or filtering crucible, 51
Graduated cylinders, 80
Gravimetric determinations, 152-157
contamination, 157
desiccators, purpose of, 154-156
Drierite®, 155
drying and cooling, 154
irreversible reagents, 155-156
ovens, temperature control in, 154
vacuum ovens, 154
 weighing containers, 154
Griffin beaker, 33
Ground-glass joints, machined surfaces of, 31
Ground-glass or screw-thread capped flasks, 33
Ground-glass surfaces, 29

H
Hand-operated vacuum pumps, 65
Hazardous laboratory chemicals, 9
HDPE and borosilicate glass, with Teflon® lid liners, 62
High recoveries, 188
High-density polyethylene (HDPE), 54
High-density polyethylene containers, 54
High-performance liquid chromatography (HPLC) methods, 7
Hydrates, 4, 200
Hydrophilic filters, 143
Hydrophobic filters, 143

I
Infrared thermometer, 97-98

K
Kjeldahl flasks, 38-39
Kuderna-Danish concentrator, 138
rapid volume reduction, 138
rapidly spinning boiling flask, 139
Snyder column, 138

L
Laboratory contamination, 193-194
Laboratory control samples, 189
Laboratory glassware, 27, 29
Latex and natural rubber tubing, 60
Linear low-density polyethylene (LLDPE), 54
Liquid-in-glass thermometer, 95, 99
expansion of the liquid column with increased temperature, 95
 liquid column separations, 99
Low recoveries, 188
Low-density polyethylene (LDPE), 54
Low-density polyethylene molded items, 54

M
Manufactured reagents, 7
Manufacturer-prepared stock solutions and certification, 11
Matrix spike, 187-188
MDL concentration standard, 192
Membrane filters, 46, 142-143
behavior toward water and other solvents, 143
depth filters, 143
manufacturers of, 144
screen filters, 142-143
Membrane filtration, 142-152
Dewar flasks, 148
disk or column, preparation of, 150-151
membrane filters, 142-143
modern filtration media, 142
purchasing presterilized filters, 144
solid-phase extraction, 150-152
vacuum filtration, 147
vacuum pump, use of, 147-148
vacuum trap, use of, 148-149
vacuum tubing, 150
Meniscus, 174
Mercury analysis, 6
Metal band screw-type hose clamp, 39
Metal tubings, fitting
brass fittings, 69
chain clamps, 69
clamps and clamp holders, maintenance of, 70-71
compression-design tube fittings, 68
double buret pinch clamps, 69
grease or other liquid/solid compounds, use of, 69
lattice connectors, 70
 leakfree connection, 68
solid rods of synthetic materials, 69
stainless steel and titanium used in construction, 68
standard right-angle double-screw clamp, 70
support rods and clamps, 69
Swagelok® compression-design pipe fitting, 69
Metallic elements, 65-71
316 stainless steel items, 67
solid iron items, 65
stainless steel items, 65
Method detection limit (MDL), 192
Molarity, 205-207
Monitoring reports, 187
Mortar and pestle
used for grinding and mixing solids, 52

N
National Bureau of Standards (NBS) calibration weights, 19
National Institute of Standards and Technology (NIST) calibration weights,
19
Neutralization reaction, 209
NIST or traceable to a calibrated NIST thermometer, 96-97
NIST-traceable thermometer, 100
Nonstandard measuring tools
calibration of, 88-90
glass syringes, calibration verification of, 90
permanent ink markings, 90
sand blasting, 89-90
Winkler dissolved oxygen determination, 88
Normality, 207-208

O
Olefin-based resins, 55
One-point calibrations, 116
Organic liquid-filled thermometers, 99
Organic solvents, 4
Oxidizers, 3
P
Paper filters, 145-146
fluted paper, 146
leachable, 146
TSS analysis, 146
Partial immersion thermometer, 95-96
Perfect accuracy, 188
Perfluoroalkyoxy polymers (PFA), 54
Pesticide residue grade, 6-7
pH and ion-selective electrodes (ISE), 169-171
advantages of, 170
combination electrodes, 170
electrode-based determinations, 171
pH/ISE meters, 171
pH/ISE workstation, 170
preventive maintenance, 171
single electrode, 170
Phenol solutions, standardization of, 113
Plastic, 52
ASTM codes, 53
ASTM resin identification codes, 53f
chemical resistance/chemical compatibility, 53
disposable units, 64
fluorocarbon resins, 54
mechanical properties of, 52
physical properties, 52
plastic filtration units, 64
recycling codes and symbols, 53
sampling tools made of synthetics, 63
Society of the Plastics Industry (SPI), 53
Plastic presterilized sample container, 62
Plasticware
description and characteristics, 52-65
manufacturers of, 53
Platinum, 67
Platinum crucibles, 67
Polybutylene terephthalate (PBT), 54
Polycarbonate safety glasses, 55
Polycarbonates, 55
Polyetheretherketone resins, 55
ferrules, 55
Polyethylene, 52, 54
Polyethylene terephthalate (PETE or PETG), 54
Polyolefin resins, 54
Polytetrafluoroethylene (TFE or PTFE), 54
Potassium dichromate, 13
Powder funnels, 42
Practical-grade chemicals, 6
Precision, 189
Prepared solutions, storage of, 108-109
Primary standards and uses, 9
Primary stands, 8
Purchased reagent, 7
Pure PVC molded items, 55
PVC membrane filters, 144

Q
Qualitative filters, 145
Quality assurance manual, 185
Quality assurance program, 185
Quality control, 185
Quality control check sample, 189
Quantitative solution chemistry, 205

R
Random errors, 184-185
Reagent blank, 194
Reagent chemicals, 6
Reagent funnels, 22
Reagent solutions, preparation of, 176
Reagent titer, 113-114
Reagent water, 11-14
American Society for Testing Materials (ASTM), 11
cartridges filled with activated carbon, 12-13
 deionizing cartridges, 12-13
distillation, 12
laboratory water, purification of, 12
resin cartridges, 12-13
water purification units, commercial suppliers of, 14
Reagents, 1
as certified solutions, 11
lifetime, 113-114
reliable suppliers of, 7
Redox reactions, 210
Reducers, 3
Reductant-demand-free water, 13
Reduction, 210
Relative percent difference, 189-190
Relative standard deviation (RSD), 190
Reporting results, 180-182

S
Safety data sheet (SDS), 4
Sample duplicates, 189
Sampling bottles, 36
Separatory funnels, 46-48
continuous liquid-liquid extractor, 48-50
emulsion-centrifugation, 48
heavier-than-water extractors, 48-49
liquid-liquid extractions, 46, 48
Single matrix spike stock solution, 190
Single open-beam balances, 16
Slide rule, 173-174
Sodium thiosulfate pentahydrate (Na2S2O3 - 5H2O), 4
Soft glass, 28
Soft, flexible tubing, 61-62
Solid-bottom crucibles and dishes, evaporating and drying water and sludge
samples, 52
Solution titers, lifetime, 113
Solutions and dilutions, preparation of, 101-109
chemical supplyhouses, 101
 reagent with a short shelf life, 101
waste disposal costs, 101-102
Solutions, preparation of, 102-106
reagent funnels, 103-104
solution preparation instructions, 102
Solutions, standardization of, 111-112
primary analytical standards, 112
quantitative analysis, 111
Solvent, 4
Spectrophotometers, 157
Stainless steel
components, 67
dippers, 67
inertness to organic solvents and reagents, 67
use of, 67
Stiffer tubings, 62
Stoichiometry, 213-216
balanced reaction, 213
titration for alkalinity, 215
Stopcock metering valves, 33
Strong acids, 2-3
Strong bases, 2
Strong-base sodium hydroxide, 3
Sulfate-free water, 13
Supplier, refusal by, 7
Synthetic materials, 52
Synthetic tubings, 61
Systematic error, 184

T
Target analyte, 188
Technical-grade chemicals, 6
Teflon®
beakers and flasks, 60
constructed wash bottles, 55
E.I. du Pont de Nemours materials, 54
laboratory ware, 60
 microwave digestion vessel liners, 60
tetrafluoroethylene, 54
Temperature measurement, 93-100
bulbs, changes in volume affect calibration, 99
Celsius (centigrade) temperature scale, 93
emissivity, 97
Fahrenheit scale (Fahrenheit (°F)), 94
fixed-temperature reference standards, 94
increased temperature, effects of, 94
kelvin (K), 93
suitable ice bath, 100
temperature indicator strips, 98
Test tubes, 37-38
tetrafluoroethylene-hexafluoropropylene copolymer (FEP), 54
Titration, 165-169
acid-base indicators, 168
contaminated reagent water, 168-169
documentation of, 169
equipment for procedures, 165
reagent and the buret, 166
reagent water, 168
titration procedures, equipment for, 165
Toploading balances, 16
Total immersion thermometer, 95
Trihydrate, 201
Triple open-beam balances, 16
Tubing, 55
purposes in laboratory, 60
Swagelok® (Crawford Fitting Company) design, 62
Turbidimetric determinations, 164-165
Turbidity, 164
increase over time, 164
readings, 165
standards, 164-165
Tygon® (Norton Company), 55, 60
tubing, 60-61
tubings, 55
U
Unglazed porcelain, 50-51
UV-visible spectrophotometer, 157

V
Vacuum filter flasks, 147
Vacuum filter holders, 46
Vacuum flasks /filtering flasks, 37
Venturi-type water aspirator, 147
Volumetric (transfer) pipettes, 80
calibrated “to contain” (“TC”), 77-78
calibrated “to deliver” (“TD”), 80
Volumetric glassware, proper use of, 83-88
calibration specifications, 84
“dilute to the mark” and “read the meniscus”, 84
reagent and calibration standards, 84
temperature, 83-84
“to contain” device, 84-85
“to deliver” device, 85, 87
volumetric (or transfer) pipette, 87
Volumetric measuring devices, 73-90
1- and 2-mL sizes, 77
ASTM E287-02 (2012) Standard Specification for Laboratory Glass
Graduates Burets, 74
ASTM E288-10 (2007) Standard Specification for Laboratory Glass
Volumetric Flasks, 74-75
ASTM E542-01 (2012) Standard Practice for Calibration of
Laboratory Volumetric Apparatus, 74
ASTM E694-99 (2010) Standard Specification for Laboratory Glass
Volumetric Apparatus, 74
Class A line, 75, 77
Class A or Class B pipette tolerances, 78
Class A volumetric glassware, 75
Class B tolerances, 75
corks, 77
flasks, closures for, 77
glass stoppers, 77
 graduated cylinders, 80
graduated pipettes, 78
maximum accuracy in the volume determination, 174-175
Mohr pipette, “to deliver” (“TD”) device, 78
screw caps and ground-glass joint closures, 77
serological pipette, 78
stoppers of rubber, 77
Teflon® stopper sleeves, 77
volumetric pipettes, 78, 80

W
Wash bottle, 55, 56f
cut-resistant gloves made of Kevlar™, 58
disposable pipets, 58
disposable pipetter tips, 59-60
disposal, 56-57
gloves, 57-58
gloves, dexterity of, 58
heat-resistant glove, 58
heavy-duty gloves, 58
membrane filtration procedure in microbiology testing, 56
polypropylene pipets, 58
polystyrene pipets, 58
preprinted right-to-know labels, 55
right-to-know labels for organic solvents, fluorinated versions of, 55,
56f
trace-level contaminants, 59-60
variable-volume dispenser, 55-56, 58f
Watch glasses, 36
Water, in reagents, 4
Weak acids, 2
Weak bases, 2
Witte plate, 51-52
 This le was downloaded from Z-Library project

Your gateway to knowledge and culture. Accessible for everyone.

z-library.se singlelogin.re go-to-zlibrary.se single-login.ru

O cial Telegram channel

Z-Access

https://wikipedia.org/wiki/Z-Library
ffi
fi