United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-91-011
April 1991
I/
Air
&EPA The Impact of Declaring
Soybean Oil Exempt from
VOC Regulations on the
Coatings Program
7
-------
EPA-450/3-91-011
The Impact of Declaring Soybean Oil
Exempt from VOC regulations
on the Coatings Program
Final Report
EPA Contract No. 68-02-4378
Work Assignment 138
ESD Project No. 89/12A
Radian Project No. 239-017-38
April 1991
Edited by
Madeleine Strum, Ph.D
Office of Air Quality Planning and Standards
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
DISCLAIMER ,
The work in this report was an effort of the Emissions Standards Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection Agency. Mention of
trade names or commercial products is not intended to constitute endorsement or
recommendation for use. Copies of this report are available - as supplies permit - from the
Library Services Office (MD-35), U.S. Environmental Protection Agency, Research Triangle
Park, NC, 27711, or may be obtained, for a fee, from National Technical Information
Services, 5285 Port Royal Road, Springfield, Virginia 22161.
Publication No. EPA-450/3-91-011
-------
ACKNOWLEDGEMENTS
We acknowledge Candace Blackley and Joan Bursey of the Radian Corporation who
carried out a great deal of the research and initial preparation of the report in fulfillment of
assignments under EPA Contract 68-02-4378. Thanks are due to Ms. Rima Dishakjian of the
Emission Measurement Branch, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency for conducting the Method 24 and 24A analyses'. We
appreciate the efforts of Mr. Dave Ailor of the National Oil Processor's Association for
coordinating most of the oil samples, and Dr. Stoil Dirlikov for the vernonia oil sample and
technical information. Further thanks go to Dr. Len O'Neil of the Paint Research Association
for sending the papers and bulletins on the numerous studies undertaken by him in the 1940's
and 50's.
-------
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION ' 1
2.0 CONCLUSIONS 3
3.0' CHEMICAL & PHYSICAL PROPERTIES ' 5
3.1 FATTY ACID COMPOSITION 5
3.2 DRYING CLASSIFICATION 9
3.3 VOLATILITY/PHOTOCHEMICAL REACTIVITY 10
3.4 THERMAL STABILITY 12
3.5 INDIVIDUAL OILS 12
3.5.1 Canola Oil 12
3.5.2 Castor Oil 13
3.5.3 Coconut Oil 13
3.5.4 Cottonseed Oil 14
3.5.5 Linseed Oil 14
3.5.6 Safflower Oil 15
3.5.7 Soybean Oil 15
3.5.8 Sunflower Oil , 18
3.5.9 Tung Oil '". 18
3.5.10 Vernonia Oil 19
4.0 PROCESSING AND REFINING VEGETABLE SEED OILS 21
5.0 DRYING/CURING PROCESSES 23
5.1 OXIDATION 23
5.2 CHEMICAL REACTIONS OF THE DRYING/CURING PROCESS 24
5.3 PRODUCTS OF COATING OXIDATION 25
6.0 VOC TEST PROCEDURES AND FINDINGS 27
6.1 TESTS CONDUCTED BY THE PAINT RESEARCH ASSOCIATION .... 28
6.2 APPLICATION OF GAS CHROMATOGRAPHY/MASS
SPECTROMETRY TECHNIQUES 29
6.3 VOC TEST RESULTS 30
7.0 IMPACT OF TEST RESULTS ON TEST METHODS AND REGULATIONS .... 33
8.0 REFERENCES 35
IV
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1
2.0 CONCLUSIONS 3
3.0 CHEMICAL & PHYSICAL PROPERTIES '. ' , 5
3.1 FATTY AGED COMPOSITION 5
3.2 DRYING CLASSIFICATION 9
3.3 VOLATILITY/PHOTOCHEMICAL REACTIVITY 10
3.4 THERMAL STABILITY 12
3.5 INDIVIDUAL OILS 12
3.5.1 Cariola Oil 12
3.5.2 Castor Oil 13
3.5.3 Coconut Oil 13
3.5.4 Cottonseed Oil 14
3.5.5 Linseed Oil 14
3.5.6 Safflower Oil 15
3.5.7 Soybean Oil 15
3.5.8 Sunflower Oil 18
3.5.9 Tung Oil 18
3.5.10 Vernonia Oil ? 19
4.0 PROCESSING AND REFINING VEGETABLE SEED OILS 20
5.0 DRYING/CURING PROCESSES 21
5.1 OXIDATION 21
5.2 CHEMICAL REACTIONS OF THE DRYING/CURING PROCESS 22
5.3 PRODUCTS OF COATING OXIDATION 23
6.0 VOC TEST PROCEDURES AND FINDINGS 24
6.1 TESTS CONDUCTED BY THE PAINT RESEARCH ASSOCIATION .... 25
6.2 APPLICATION OF GAS CHROMATOGRAPHY/MASS
SPECTROMETRY TECHNIQUES 26
6.3 VOC TEST RESULTS 27
7.0 IMPACT OF TEST RESULTS ON TEST METHODS AND REGULATIONS 30
8.0 REFERENCES 32
-------
LIST OF TABLES
Page
Table 3-1 CHARACTERISTICS OF VEGETABLE SEED OILS 8
Table 3-2 FATTY ACID CONTENT OF COMMON OILS 9
Table 3-3 COMPOSITION OF NORMAL FATTY ACIDS 10
Table 6-1 TEST RESULTS 29
VI
-------
-------
1.0 INTRODUCTION
A food processing facility alleged that soybean oil used in frying its products should
not be classified as a volatile organic compound (VOC) because its very low vapor pressure
would preclude it becoming airborne, hence it could not contribute to photochemically
produced ozone in the troposphere.
In addition to widespread use in the food industry, soybean and other vegetable oils
are constituents in some inks, paints, and perhaps, other coatings. Thus the ramifications of
any decision to declare it non-photochemically reactive on the longstanding regulatory
program for coatings were important. This project was initiated to resolve the following
questions:
Does soybean oil contribute VOC's to the atmosphere? In other words, is soybean oil
an ozone precursor?
Is soybean oil unique among the vegetable oils? Are its physical characteristics
significantly different from other vegetable seed oils in terms that would affect its
potential to be a precursor to ozone formation?
What impact would declaring soybean oil exempt from VOC regulations have on
current testing procedures and regulatory compliance programs for coatings'1
This report presents the results of an investigation which addressed these questions.
The information was obtained from a literature search, contacts in industry and academia, and
limited experimental analyses. The literature search included the following computerized
databases: Chemical Abstracts, Predicast Terminal System Predicasts Overview of Marketing
and Technology (PTS PROMPT), Food Science and Technology Abstracts (FSTA), Agricola,
CAB Abstracts, and Biosis Previews. Tables were prepared to compare various
characteristics of ten vegetable seed oils (some of which are frequently found in coatings):
canola. castor, coconut, cottonseed, linseed', safflower. soybean, sunflower, rung, and vernonia
oils. The common uses and availability of each oil were identified, and the oil bean refining
and drying/curing processes were examined. Studies of the VOC associated with vegetable
-------
oil oxidation were reviewed, and test results applying EPA's reference methods 24 (RM-24)
and 24A (RM-24A) for measuring VOC's in coatings to various vegetable oils were analyzed.
Section 2.0 contains the summary and conclusions. Section 3.0 discusses chemical
and physical properties and general characteristics of the oils. Refining processes are
described in Section 4.0 and drying/curing processes in Section 5.0. Section 6.0 discusses
VOC test procedures and findings. Regulatory impacts are included in Section 7.0. Section
8.0 contains the references and Section 9.0 the Appendices.
-------
2.0 CONCLUSIONS
1) A decision to exempt soybean oils from VOC regulation will have no effect on the
longstanding VOC reduction program associated with coatings.
2) Soybean oil and other vegetable seed oils have no measurable vapor pressure at room
temperature and atmospheric pressure.
3) None of the vegetable seed oils analyzed with EPA Method 24 (RM-24) or EPA Method
24A (RM-24A) contained a measurable amount of VOC.
4) Some samples analyzed by RM-24 or RM-24A actually gained weight (presence of VOC
would normally cause a weight loss.)
5) Enforcement of air pollution regulations using RM-24 and RM-24A would not falsely
jeopardize a source using coatings with a vegetable oil constituent.
6) Soybean oil, like the other vegetable seed oils listed in Table 3-1. is a complex fatty
acid triglyceride.
7) The literature provides no obvious evidence that soybean oil can be distinguished from
other vegetable seed oils in any significant way from the standpoint of volatility or
photochemical reactivity.
8) Vegetable oils vary somewhat in physical properties and from batch to batch because of
weather, soil and species.
9) All vegetable seed oils autoxidize when exposed to air. The resulting oxidation reaction
cleaves the molecule with formation of some amount of many VOC's, chiefly hydrocarbons
and aldehydes.
10) Oxidation is the mechanism by which coatings containing vegetable seed oils dry or
cure.
11) Soybean oil will not boil at atmospheric pressure. Progressively increasing application
of heat ultimately chars the oil, thermally cracking the fatty acid triglycerides to release
hydrocarbons and aldehydes, precursors of ozone.
-------
12) Soybean oil can contribute to ozone formation", as a precursor of VOC, although the
contribution has not been quantified.
13) Since autoxidation of oils is a surface phenomenon, any process which maximizes the
surface area of the oil will accelerate the autoxidation process. Thus, processes which employ
oil as a mist or spray accelerate the oxidation of the oil and increase the release of ozone
precursors.
14) This program, with a primary focus on the use of oils in the coatings arena, has posed
new questions regarding volatile organic compounds which may be released in the food
industry from processes involving cooking of foods in heated oils and spraying of oils.
Appendix iv provides a list of references on this topic obtained from a computerized literature
search.
Note that any attempt to sample a hot stack above a soybean oil cooking process poses
challenges. A typical VOC sampling arrangement distinguishes between paniculate and
gaseous phases. One would normally conclude that the ozone precursors are present in the
gaseous phase. Clearly, that would not be the case here. Inasmuch as the oxidation of
soybean oils is not a rapid process, little of the ozone precursors would be formed in the
stack. As a result, the total ozone formation potential of the exhaust gas includes the VOC in
the gas phase and some unknown portion of the aerosol in the paniculate phase.
'.Although not a product of this study, clearly the addition of moisture-laden foodstuffs such
as potatoes, apples, and other materials to hot cooking oils causes splattering of the oil. The
airborne particles have dramatically increased surface to mass ratios which would accelerate the
oxidation reaction. Further, many of the foodstuffs are complex carbohydrates which, during
cooking, can and do thermally crack to form lower molecular weight more volatile organic
compounds which become airborne and available to participate in photochemical chemistry.
Odorous compounds during the cooking operation are evidence of these volatile organic
materials.
-------
3.0 CHEMICAL & PHYSICAL PROPERTIES
3.1 FATTY ACID COMPOSITION
Vegetable seed oils are water-insoluble substances of vegetable origin. Chemically
they are triglycerides, i.e., compounds made up of glycerol-esters of long chain fatty acids.1
H H
I
H-C-OH HOOCR H-C- OOCR
i I
H-C-OH -f HOOCR' 1^ H-C - OOCR'
; I
H-C-OH HOOCR" H-C-OOCR" 4- 3H 0
; i .
H H
Glyceroi Fatty Acids Triglyceride
Since more than one kind of fatty acid is usually present, most vegetable oils are mixed
triglycerides. The type of fatty acid found in the oil is the most important factor influencing
the properties of the vegetable oils.2 Most vegetable oils are composed of fatty acids with 18
carbon chains with varying degrees of unsaturation (containing double bonds). Examples of
common saturated fatty acids are caprylic (C8), capric (CIO), lauric (C12), myristic (C14),
palmitic (C16), and stearic (CIS).3 Common unsaturated fatty acids include oleic (CIS),
linoleic (CIS), linolenic (CIS), alpha-eleostearic (CIS), beta-eleostearic (CIS), and ricinoleic
(CIS).4 Tables 3-1 and 3-2 show the predominant fatty acids found in the ten selected oils.
Table 3-3 shows the composition of these fatty acids. The type and percentage of fatty acids
vary from oil to oil. For example,
-------
V5
"yj
UJ 't?
Q Z W
^> ^r (g
CO Q. _
g
o uj
U. 3
— _t
^-
co
UJ LU
Z ^
0 $
U)
Q 3
0 d
>
|s|
2 < £?
§5l"
Q. U. O
>—
L1J U-l
K LU
^ Z W -?
O O u. s^
o O
' Q
H*
C3 O
— tb
CC c/5
0
Q
LU O
2 O
D U,
>lf
§8*
en u.
I
H- g-
! o ^
0<£
O
LL
O
LU
Q-
^«.
I—
o
CM
CO
CM
.c
O
en
^
CM
"a
-C
CM
"|
0
o
in
ra
0
c
ra
O
^3 JO
m •* o
' T CM CO
CM CM CJ
o o o o o o o
co m en en o en en
s 2 g 0^ | | a
m CM ^— -T ^r co -T
"o* '"^
o A "u" = "cr "c*
-005355
3i s« =» °i pi °» °n
^ r^» — *T f^ o — c\j _ un _ *™ — c\j
.,
O O O) C O C S
m m o in o O) O)
? i" £ i, | |
-0 T3 -0 g> _ _ _
c c c 1, E E E
O O O w 05 Q) O
Z Z 2 Q CO CO CO
JD J
^ fi
^—i CM
IT) -: r: ^
I CM , CO CM
O O O O
^ = 1 , | s i
o § § 1 o • o
ui o « yi z; >. c
ra o o ~ ra o =
O O O Jj CO CO CO
o
en
CD 7-.
i£ 8
CO —
~ ™ ~
5 « ~o
D ^
O OJ
m T
c
s-
Q
r- _
CO 0
in to
0 O
2
§ 5
i- >
en 3
— o
a"-
•a —
§.0
o >
Ua
S S
3
Ol
<
^8s
5 en 2-
o — .
c
.c
c
o
1
.
— O) O C
£ £ n" «
ra in Q- ri
Q. ra 09 TJ
= "5 ra -
3 = o «
c Q- j= ra
< O O CD
"1-1
l-'S-t
= O) < nj"
O ^ •£"§
cc3° ™
o ° -o O
ffl "= § 'O
o a>
CD ra
CM 5
W 1 LU §
c ra « o
o Q. -g o
fH C C^ C^
Q) Q) O O
T3 ~O C3 C
O
a
-f >• 2
^ O Q)
^ j^ tr
gsS
2 Q O
u — ra
I 2 S
i" s
0 = -D
o S §
S §r
^ _o o
o o c
O T3 O —
------- An error occurred while trying to OCR this image.
-------
u
D
s
a:
g
CJ
T
U
u
CJ^
y
u
JL
cj^
u
w —
~ ^
5 o
o
o
CJ CJ CJ Cj
u u u u
o o o c
U U
_!!_
CJ
~ c
~ 2 !
59
R
oc
U
>,
50
o
c.
-3
U
-------
tung oil consists of 80 percent eleosteanc acid, whereas soybean oil is made up of 51 percent
hnoleic acid.
The acid value in Table 3-1 indicates the amount of free fatty acids in an oil,
expressed as the number of milligrams of potassium hydroxide required to neutralize the acids
m 1 gram of oil,5
3.2 DRYING CLASSIFICATION
The double bonds in unsaturated fatty acids are chemically reactive sites and are the
points at which oxygen reacts with an oil to produce drying.*1 Therefore, the degree of
unsaturauon of the fatty acid determines the drying properties of the oil. Vegetable oils are
often classified as nondrying, semi-drying, or drying oils, with the first group containing a
higher content of saturated fatty acids and the last group a higher content of unsaturated fatty
acids. The choice of oils for use in coatings is influenced by the drying properties, and this
classification can be helpful in that decision.
Nondrying oils have low unsaturation or contain less than 20 percent of linoleic acid.
Coconut, cottonseed, and castor oil are members of this group. Large quantities of these oils
arc consumed in edible products, and others are used in soaps or surface-active products.
Drying oils have greater unsaturation or greater quantities of fatty acids with higher
unsaturation, such as linoleic acid or linolenic acid.8 Oils in this group include linseed, tung,
and dehydrated castor oil which are found in paints, varnishes, and other coatings.
Semi-drying oils are intermediate between the non-drying and drying oils in properties
and contain about 40 to 60 percent linoleic acid.9 Safflower. soybean, and sunflower oils are
classified in this group, because they can be used in products of both classes after
modifications. For example, soybean oil can be deodorized and inhibited with anrioxidants to
make an edible salad oil. or it may be modified to make a long oil alkyd performing much
9
-------
like linseed oil.10
Semi-drying and drying oils were long used as binders in the protective-coating
industry but have been largely replaced by synthetic binders in many of the coatings used
now. The binder is the non-volatile resinous or resin-forming constituent of coating,
responsible for forming the film.11 The binder plus a solvent is referred to as the vehicle.
The term iodine value in Table 3-1 indicates the degree of unsaturation of an oil, and
is expressed as the centigrams of iodine absorbed by one gram of oil under controlled
conditions.12 Usually oils with high iodine values have better drying properties, although this
statement is not applicable when comparing conjugated with non-conjugated oils.
Conjugated double bonds are two or more double bonds which alternate with single bonds in
an unsaturated compound (e.g., -CH=CH-CH=CH-). Non-conjugated double bonds occur
when the unsaturated sites are separated by one or more additional methylene groups
(e.g., -CH=CH-CH2-CH=CH-).
3.3 VOLATILITY/PHOTOCHEMICAL REACTIVITY
Volatility is defined as the tendency of a solid or liquid material to pass into the vapor
state at a given temperature. Experimentally, the volatility is determined by dividing the
vapor pressure of a component by its mole fraction in the liquid or solid.14 The vapor
pressure of a substance (often expressed in millimeters of mercury, mm Hg) is the pressure
characteristic at any given temperature of a vapor in equilibrium with its liquid or solid
form.15 Very little information relating to the determination of volatility-(direct determination
of vapor pressure for oils) was retrieved during the literature search and data gathering phase
of this program. Vapor pressure information was obtained only for soybean oil. At a
temperature of 254°C, soybean oil has a vapor pressure of 0.001 mm Hg. At a temperature
of 308°C, soybean oil has a vapor pressure of 0.05 mm Hg.16 The extrapolation of these
vapor pressures to standard conditions (27°C) would yield no measurable vapor pressure at
room temperature. Boiling points for the oils cannot be measured at atmospheric pressure,
10
-------
since decomposition of the oil (as indicated by discoloration of the oil, charring, and
ultimately evolution of smoke)"* occurs before the oil can actually boil. Therefore, the
most direct indicators of volatility, vapor pressure and normal boiling point, although not
directly available for the vegetable seed oils, obviously are such that one would not expect
the oils to vaporize to any significant extent. Boiling points could be obtained tor many of
the fatty acids which constitute the major portion of the oil molecule. These fatty acid
boiling points were determined under reduced pressure (typically 1 - 20 mm Hg rather than at
760 mm Hg which is atmospheric pressure) because decomposition is observed when the fatty
acids are heated at atmospheric pressure.
The Glossary for Air Pollution Control of Industrial Coating Operations defines VOC
as:
any organic compound which participates in atmospheric photochemical
reactions; that is. any organic compound other than those which the
Administrator designates as having negligible photochemical reactivity. VOC
may be measured by a reference method, an equivalent method, an alternative
method or by procedures specified under any subpart. A reference method, or
an alternative method, however, may also measure nonreactive organic
compounds. In such cases, an owner or operator may exclude the nonreactive
organic compounds when determining compliance with a standard.
This definition gives precedence to the analytical method specified for any specific standards
under the New Source Performance Standards (NSPS) program. The definition presented in
40 CFR, pans 51 and 52, specifically identifies those organic compounds (all of which,
except for methane and ethane, are halogenated) which the Agency has deemed of negligible
reactivity. The Agency permits any emissions of these materials to be excluded from a
measurement such as RM-24/24A of total volatiles. New materials are being added to the list
of exempt compounds.
Vegetable seed oils are a VOC bv this definition since the definition contains no
Decomposition, of course, causes some portion of the oils to become airborne and available
for photochemical reactions.
-------
specific criteria pertaining directly to volatility, vapor pressure, number of carbon atoms, or
boiling point.
3.4 THERMAL STABILITY
No oil decomposition temperatures (the point at which oils start to decompose upon
heating) were found from the literature search. However, the smoke points of four of the
vegetable oils have been determined and are found in Table 3-1. Bailey's provides the
followins definitions regarding thermal stabilitv.
The smoke, fire, and flash points of a fatty material are measures of its thermal
.nubility when heated in contact with air. The smoke point is the temperature
at which smoking is first detected in a laboratory apparatus protected from
drafts and provided with special illumination. The temperature at which the
material smokes freely is usually somewhat higher. The flash point is the
temperature at which the volatile products are evolved at such a rate that they
are capable of being ignited but not of supporting combustion. The fire point
is the temperature at which the volatile products will support continued
combustion.18
Impurities such as free fatty acids lower the smoke point of the oil drastically.1 Canadian
government specifications require that a good frying oil have_a smoke point above 200°C.
Clearly, any event (e.g. heating) which causes the oil to smoke results in airborne
decomposition products that are available for photochemical reactions.
3.5 INDIVIDUAL OILS
3.5.1 Canola Oil
Canola seed is bred from rapeseed, a member of the mustard family. Most of the
world's supply of canola oil comes from Canada and Europe. Research is being performed to
solve production problems in order to increase yield since demand surpasses supply.21 The
12
-------
oil is composed of 58 percent oleic acid which makes it lower in saturated fat than coconut,
soybean or any of the other oils." Therefore, there is a large demand for canola oil from
health-conscious consumers. Canola consumption has risen from 100,000 Ibs in 1986 to 500
million Ibs in 1989, and it is expected to grow 7 percent each year to 700 million Ibs by
1994."3 Even with the higher price in the grocery store, $3.99 for 16 oz. canola oil versus
SO.89 for 16 oz. of vegetable (unspecified) oil, sales are increasing.24 Canola oil is, however,
used more in food products than in coatings.
3.5.2 Castor Oil
Castor oil is extracted from the seed of the castor plant which is grown commercially
in Brazil, India, and in some of the warmer sections of the United States. The seeds
generally contain about 45 percent oil.25 Not only is the oil itself useful in a variety of
compositions, but it is capable of undergoing several kinds of chemical transformations (such
as dehydration) by which it is convened into useful derivatives. Castor oil differs from other
vegetable oils by having higher viscosity, greater solubility in alcohol, and lesser solubility in
petroleum solvents.~& These properties are important for its uses in paint, varnish.
plasticizers, hydraulic fluids, and cosmetics.
Dehydrated castor oil is among the most valuable drying oils for coatings. It dries
faster, heat-bodies (thickens when heated) faster, has equal color retention and imparts
superior film toughness to alkyds than soybean oil.27 However, usage of castor oil is
98
restricted by its somewhat higher cost (S().4()5/lb as compared to S0.25/lb for soybean oil).
3.5.3 Coconut Oil
Coconut oil is obtained from copra, the dried, broken kernels of coconuts; copra
contains about 65 percent oil.29 With an average iodine value of 10 or less and only 9
percent unsaturated acids, coconut oil is more stable to oxygerrand heat than any other
natural oil.J° Because of this stability, alkyd coating resins that contain coconut oil are
13
-------
31
permanently flexible and are highly color retentive, even when baked at high temperatures.
Coconut oil is relatively resistant to the type of rancidity caused by oxidation at
ordinary temperatures, but it is susceptible to a type of rancidity caused by microorganisms,
especially various molds.32 Therefore, food products and other materials containing coconut
oil and moisture are subject to this type of spoilage. Even so, this oil is used for food
products, soaps, and coatings. However, due to its high saturated fat levels, its use in food
products is declining.33
3.5.4 Cottonseed Oil
Cottonseed oil is extracted from the seed of the cotton plant, which is grown primarily
in the United States, India. U.S.S.R, Brazil, China, and Egypt.34 Practically all of the crude
cottonseed oil produced in this country is refined for use in edible products such as
shortening, margarine, and cooking oils."5 Cottonseed oil demand is stimulated by its
amenability to many frying applications.'36 However, due to its low unsaturation it could also
be used as a lower cost substitute for coconut oil in color retentive baking alkyds. Alkyds
made with cottonseed oil have better color retention than those made with soybean and
safflower oils but are inferior in gloss and toughness.
3.5.5 Linseed Oil
Linseed oil is a low-melting oil, is slightly less viscous than most vegetable oils, and
has a high iodine value which reflects the high degree of unsaturation of its fatty acids.
"JO
Linseed oil is obtained from flaxseed, which averages about 35 percent oil content. Linseed
oil is not well suited for use as an edible fat, but it is used in the manufacture of paints
(particularly artists' paints), varnishes, and printing inks. It is the sole binder in oil house
paint in which the binder is usually a blend of thin linseed oil, raw or refined, with
heat-bodied linseed oil.39 A thin, alkali refined linseed oil is also used for alkyd paints where
fast drying is important and good color retention is not required.
14
-------
One property of linseed oil that must be considered when choosing a linseed oil paint
is "after-yellowing." Linolenic acid yellows more than other fatty acids and largely governs
the yellowing of oils. Since linseed oil contains the highest proportion (52 percent) of this
acid, it yellows more than other oils, thereby causing a yellow tint to linseed oil-based
paints.40
A disadvantage of the use of linseed oil-based paints is its high suceptibility to mildew.
Mildew, a fungus which flourishes in damp environments, is a common cause of the
discoloration and degeneration of house paint and unfinished wood. Many paints contain
mildewcides which serve to prevent and cure mildew problems.41
The demand for fats and oils for all uses will grow 1.8 percent per year to about 21
billion pounds in 1994, and growth in the use of linseed oil in coatings and inks will
continue.42
3.5.6 Safflower Oil
The safflower plant, from which safflower oil is obtained, is grown in Africa, and the
Middle East, as well as in the western portion of the United States. The seeds contain about
30 percent oil. Safflower oil is used in some food products but is used more frequently as a
drying oil in coatings. The oil falls oetween soybean oil and linseed oil in total unsaturation
and in price. As for its drying properties, safflower oil is closer to linseed than to soybean.
Unless modified, safflower oil does not dry fast enough for use in house paints.
Approximately 75 percent of the fatty acids of safflower oil are linoleic and only one percent
are linolenic, which compares with 9 percent linolenic in soybean oil. Because of this unique
composition, safflower oil dries faster and yellows less than soybean oil coatings.
3.5..7 Soybean Oil
Soybean oil. also called soya oil. is extracted from the seeds of soybean plants which
15
-------
are grown in many areas of the United States and South America. In fact, the United States
is the world's leading producer of soybeans, which is the American farmer's second-largest
cash crop.4 Crude soybean oil usually contains no more than 1.5 percent free fatty acids,
often less than 0.5 percent, and from 1.5 to 4 percent other impurities.46 These constituents
are removed from the oil during refining and degumming processes. Refined soybean- oil
consists primarily of the triglycerides of oleic, linoleic, linolenic, saturated acids, and about
0.8 percent impurities (referred to as unsaponifiable material).47 Most varieties of oil seeds
contain from 18 to 22 percent oil.48
The soybean is an annual crop that is well suited to mechanized planting, cultivation,
and harvesting. Varieties have been developed to suit various climates and soils, and new
varieties continue to be developed. The price of soybeans remains one of the lowest of
\egetable oils (about SO.25 per pound of crude).49 Because of the large number of varieties
of soybeans grown and the diversity of soils and climate in which they can be successfully
cultivated, the physical and chemical properties of soybean oils will vary. For example, the
iodine value of soybean oil varies with the plant variety and also varies within the same
variety grown in different locations and in different years.50 In one study in which samples
of cultivated soybeans were extracted under uniform conditions in the laboratory, oils having
iodine values ranging from 99.6 to 143.2 were obtained. These samples represented ten
varieties grown in several locations in different crop years.'1 Oils produced commercially
tend to vary less because of the blending of different lots of seeds that takes place in ordinary
commercial handling and milling. One comparison of 87 soybean oil samples produced
commercially showed a range in iodine values from 129.3 to 136.7.52 These variations may
be partially due to the variability in weather conditions such as temperature and humidity
which affects the metabolic processes involved in the biosynthesis of lipids (fats) in the
growing seed.53'54 The moisture content of soybeans at the time of harvesting depends a
great deal upon the weather and the maturity of the beans, and it is an important factor in
determination of the grade of soybeans?
The principal uses of soybean oil are in food products and as an ingredient in paints,
16
-------
inks, and other drying oil products. Soybean oil's low saturated fat content and lack of
cholesterol have made it a popular choice for food manufacturers and fast food chains.=6-57
Soybean oil is the most widely consumed of the vegetable oils with a projected growth rate of
1.9 percent per year through 1994.58
Because of the semi-drying nature of soybean oil, no paints are made from straight
soybean oil. However, alkyds (oil based paints) made from soybean oil have good drying
properties, and it is used more than any other oil for making alkyd resins found in
architectural and industrial coatings."9 Soya alkyds are also widely employed as binders for
interior and exterior air-drying enamels, marine paints, and alkyd type house paints.60
Soya alkyds are classified as color retentive, and satisfy the color requirements of the
majority of white and light colored top coats, both air dried and baked.61 There is also
evidence that colored inks made with soybean oil have greater color rendition (brightness)
than inks made without it.6" News inks made from soybean oil also resist rub-off better than
conventional inks.6 For these reasons, there has been an increase in the use of soybean oil in
the printing industry.
• In general, paints and inks contain solvents and/or petroleum distillates which emit
VOC\ to the atmosphere. Based on Method 24 (see Section 6.0), Flint Ink Corporation
found that the substitution of a mixture of soybean oil and corn oil for the middle petroleum
distillates in a sheetfed offset ink greatly reduced the VOC content.64 We expect the use of
soybean and perhaps other vegetable oils in place of the more volatile petroleum distillates
to reduce ozone formation since their slower evaporation rate greatly offsets the oxidation
reaction that releases some unquantified amount of light hydrocarbons (Section 5.0). The
Hazardous Waste Research and Information Center in Illinois is carrying out a research
project entitled "Toxic Substance Reduction for Sheetfed Offset Printers" to determine the
waste reduction advantages of substituting soybean oil for petroleum based inks and organic
solvent based cleaners. ~
17
-------
A benefit of using soybean oil in coatings is that it is a renewable, domestically
produced and processed product. The fact that soybeans can be planted and harvested year
after year is an advantage. Since soybeans are commonly used in food products, toxicity does
not appear to be a problem. Disposal of soybean-based products could also be considered
less problematic than petroleum-based products since the oils appear to be more readily
biodegradable.65
3.5.8 Sunflower Oil
Sunflower oil. a light yellow oil well suited for use as a salad and cooking oil, is
obtained from the sunflower seed plant that is grown primarily in Europe and the U.S.S.R.66
The oil content of the seed may average about 29 percent.57 Total unsaturation of the oil is
comparable with that of soybean oil. but the linolenic acid found in soybean oil is lacking in
sunflower oil. Sunflower oil is classified as a semi-drying oil and is used in the
manufacturing of oil-modified alkyd resins and similar products.08 Like soybean oil, the
iodine value and composition of sunflower oil are influenced by the plant variety,
temperature, soil fertility and moisture supply. Oils differing widely from each other in
iodine value usually differ in the ratio of linoleic to oleic acid rather than in any large
difference in total content of unsaturated acids.69
3.5.9 Tung Oil
Tung oil, also known as China wood oil, was originally obtained from China but is
now grown in the southern United States. Tung oil is a highly unsaturated oil with a
viscosity considerably higher than that of linseed oil. Since tung oil has laxative if not
poisonous properties, it is not considered edible.70 However, since tung oil is classified as a
drying oil it is utilized in varnishes, enamels, and other products in which the drying quality
of the oil is important. Tung oil works well in combination with relatively inexpensive resins
and. compared with nonconjugated oils, tends to confer superior water resistance, alkali
resistance, hardness, and durability on the products in which it is used.'1
18
-------
Eleostearic acid is the principal fatty acid in tung oil, and it is the chemical nature of
this acid that differentiates tung oil from other drying oils. Eleosteanc acid is an isomer of
linolenic acid, and it has its three double bonds in the conjugated relationship with each
7"1
other. " This conjugated unsaturation is responsible for the fact that tung oil thickens and
gels more readily than other oils when heated. The very rapid bodying of tung oil precludes
its use as a large part of the oil in alkyd resins. •
3.5.10 VernnniaOil
Vernonia oil is extracted from the vernonia plant, a new potential oil seed crop grown
in Africa. AMU. Central America, and in the southwest United States. Vernonia seeds contain
about 42 percent oil in contrast to soybean seeds, which contain 17 to 18 percent oil.'
Vernonia oil contains predominately (about 80 percent) tnvernolin, a triglyceride of vernolic
acid. A unique property of this oil is its homogeneous molecular structure consisting
primarily.of identical tnglyceride molecules with three equal vernolic acid residues. In
contrast, all other vegetable oils consist of a heterogeneous mixture of triglycerides with
different fatty acid residues.75
Vernonia oil is a transparent homogeneous liquid at room temperature with low
viscosity and excellent solubility in numerous organic solvents, diluents, and paints. Due to
these properties, vernonia oil is being considered as a reactive diluent (reduces viscosity like a
conventional >olvent but also reacts with the coating resin in the drying process) for
high-solids alkyd, epoxy, and epoxy-ester coating formulations by-replacing conventional
T/- -j-j
solvents which ultimately'are released as VOC emissions.' '
19
-------
20
-------
4.0 PROCESSING AND REFINING VEGETABLE SEED OILS
The following description of the processing and refining procedures may vary for
different vegetable seeds, but in general, the steps are similar. " In order to extract oil from
vegetable seeds, the seeds are first cleaned, dried, cooled, and transferred to tempering bins to
allow the moisture within the seeds to redistribute itself.78 The seeds are cracked, dehulled,
and rolled into flakes until the oil cells are exposed. These flakes are then extracted with the
solvent hexane to remove fats. The solvent is removed, and the flakes are toasted with steam,
dried, cooled, and ground into meal.79 The remaining solvent-oil mixture is distilled to
separate the two into solvent and crude vegetable oil.
In us crude state, vegetable seed oil contains, besides the triglycendes, a number of
impurities such as gums, phosphatides. free fatty acids, pigments, and traces of oxidation
products from oil degradation.80 The purpose of refining oils is to remove these unwanted
materials to attain quality standards for end-use products, whether they be food or non-food
products. Degumming is an optional first step in purification in which water is added to the
oil to dissolve the hydratable phosphatides, and the mixture is then separated by
centrifugation.81'82 The next step is to neutralize the free fatty acids with alkali and to
separate the resultant soap. This process is referred to as saponification: the term
suponification number found in Table 3-1 is defined as the number of milligrams of
potassium hydroxide required to saponify 1 gram of fat. The next step is to react the oil
with an activated bleaching clay to remove pigments and oxidation products: the last is to
deodorize by steam distillation at reduced pressure. This treatment removes most of the
remaining free fatty acids, some color bodies,-and some unwanted flavor. 4 Vegetable oils
used in coatings generally do not require the deodorization step, only refining and
bleaching.85 After deodorization, the oil can be made more viscous by injecting air into the
QS
oil at elevated temperatures to partially oxidize the oil to create bodied or blown oil.
21
-------
-------
5.0 DRYING/CURING PROCESSES
5.1 OXIDATION
Oxidation is the reaction of oxygen with another substance in which electrons are
transferred. Autoxidation, the oxidation of vegetable oils with the oxygen as present in the
atmosphere, affects the oils in various ways. Fats and oils intended for food can become
rancid due to oxidation if not handled and stored properly. However, the drying of paint
films is dependent on the oxidation process. Most coatings, in spite of their apparent
diversity, dry by much the same oxidizing mechanism, whether they are alkyds, epoxy esters,
V "7
urethane alkyds, or oil-based coatings. All are based on drying oils, the naturally occurring
o o
tnglyceride esters of certain unsaturated fatty acids bearing multiple double bonds.
There are a number of factors that affect the oxidation process:89
The reaction is accelerated by light.
Film thickness influences oxidation rate, because oil oxidizes more rapidly in
shallow layers than in deep layers.
Temperature influences reaction rate. For example, heat usually accelerates
oxidation.
The rate of oxidation is greatly dependent upon unsaturation and the structure
of the fatty acids present in the oil. Unsaturated fats are more susceptible to
oxidation than saturated fats.
The presence or absence of antioxidants or prooxidants will affect oxidation.
Antioxidants, which sometimes occur naturally in vegetable oils, delay or
reduce the oxidation process. Prooxidants accelerate the reaction, and are often
added in the form of driers. Driers are oil-soluble metallic compounds such as
cobalt or manganese deliberately added to oils for the purpose of accelerating
the oxidation and drying of oils used in coatings.
The fewer impurities in the oil, the faster the oxjdation and drying processes.
23
-------
5.2 CHEMICAL REACTIONS OF THE DRYING/CURING PROCESS
Since the drying/curing processes of oils are generally described in terms of coatings,
the following process description reflects oils in coatings, not oils alone.
Autoxidation of oils is the basis for the drying/curing of coatings. Drying is defined
as the polymerization of the glycerides of unsaturated vegetable oils induced by exposure to
air or oxygen.91 Curing is the conversion of a raw product to a finished and useful condition,
usually by application of heat and/or chemicals which induce physicochemical changes.
The mechanism, while not fully understood, involves the reaction of oxygen with the double
bounds or active methylene groups in the oil or resin medium. The reaction usually entails a
free-radical mechanism and is initiated by heat, light, metallic catalysts, or free-radical
generators.93 Autoxidation may be thought of as being analogous to addition
polymerization.94
In autoxidation, oxygen attacks the fatty acid chains at or near the double bonds.
Non-conjugated fatty acids are thought to first undergo a rearrangement of their bonding to
give conjugated forms. This oxygen attack involves the formation of hydroperoxiues (ROOH,
where R is a fatty acid chain) that subsequently break down to form free radicals (R0«)-
These free radicals readily attack other fatty acid chains at the methylene group between the
two double bonds, producing other free radicals (R*), and allowing bond rearrangement.95
With the conjugated double bonds available, drying proceeds as it does with
conjugated acids. Oxygen attacks the double bonds to form peroxy radicals (R-O-O*), which
interact to form polyperoxides (long chains of fatty acid molecules linked together by -O-O- •
groups), or which may react with other fatty acids to form large, complex, crosslinked
materials. The radicals can also decompose, leading to the evolution of volatile organics
such as aldehvdes, ketones and alcohols.97
24
-------
5.3 PRODUCTS OF COATING OXIDATION
Emissions from the formation of a coating film result from:
(1) evaporation of the solvent portion of the coating,98 and
(2) reaction by-products emitted during the chemical reaction that takes place in
coating films at cure temperatures.
The latter are denoted cure volatiles, and include a variety of compounds such as aldehydes,
ketones. acids, water, carbon dioxide, hydrogen, polymerized fats, epoxides, and peroxides.
If heat is not applied, as with many inks, some of these compounds such as aldehydes remain
in the film instead of volatizing.101 A more detailed discussion of the oxidation products of
vegetable oils is found in Section 6.0.
25
-------
26
-------
6.0 VOC TEST PROCEDURES AND FINDINGS
The methodology chosen to evaluate the impact of a policy decision declaring soybean
oil exempt from regulation on the Agency's program for coatings was to determine the VOC
content of the oil with the Agency's analytical methods for measuring the VOC content of
coatings and inks.
EPA's Reference Method 24 is applicable to the determination of volatile matter
content, water content, density, volume solids, and weight solids of paint, varnish, lacquer, or
rekued surface coatings (solvent-borne and waterborne). The quantity of VOC in surface
coatings is defined by the results of Method 24. Using the test protocol for Method 24,
samples are weighed and then heated at 110 ± 5°C for one hour. The samples are weighed
again after heating. Loss in weight represents evolution of volatile material which would
include water. VOC and non-reactive (exempt) volatile compounds. Procedures are included
within Method 24 to determine water content either by direct injection into a gas
chromatograph or by Karl Fischer titration either to a colorimetric end point or to a
potentiometric end point. Coating density and solids content may also be determined by
procedures included in Method 24. In commercial and regulatory applications, the pounds of
VOC per gallon of coating less water and exempt compounds is calculated from the measured
loss of weight and the exempt compound and water content of the coating.
EPA's Reference Method 24A is applicable to the determination of VOC content and
density of solvent-borne (solvent-reducible) printing inks and related coatings. The VOC in .
printing inks and related coatings is defined by the results of Method 24A. For Method 24A,
a sample is heated at 120 ± 2°C for 24 hours in a forced draft oven, or the sample may be
heated for a shorter period of time in a vacuum oven. Method 24A also incorporates
procedures for the determination of coating density and solvent density, but procedures for the
determination of water content are not required in Method 24A"since the method applies to
solvent-borne printing inks and related coatings.
27
-------
The EPA reference methods provide a single number which is characteristic of the
weight percent VOC. Other methods found in the literature have been applied to characterize,
qualitatively, the VOC's which are emitted from vegetable oils under certain conditions. The
next sections will discuss the results of this work, and results from the EPA test methods.
6.1 TESTS CONDUCTED BY THE PAINT RESEARCH ASSOCIATION
Long ago, the Research Association of British Paint, Colour, and Varnish
Manufacturers measured water, carbon dioxide, volatile acids, and aldehydes evolved from the
drying process (autoxidation) of oil films at ambient temperatures.1 ~ To determine the nature
of the volatile compounds which were emitted in the course of the drying process, the British
chemists coated the interiors of bottles with a thin film of oil. A stream of air was then
passed through six bottles in series for seven days so that acidic volatile products which were
formed could be collected in a solution of sodium bicarbonate and characterized. The British
chemists found that the main acidic compound which was formed from linseed oil was formic
acid. The amount of volatile products evolved from oil films after seven days of aging in the
laboratory could account for a significant degradation of the acid portion of the triglyceride
molecule. Reaction in the molecule appeared to begin by cleavage at or adjacent to a double
bond, followed by progressive degradation along the chain with release of volatile acids and
carbon dioxide. The total amount of volatile materials released varied with the type of oil
tested. These experiments were performed approximately forty years ago, prior to the
availability of now common chromatographic techniques. Not all hydrocarbons were
characterized. Aldehydes and acids formed solid derivatives which could be characterized by
melting point and other physical properties.
Since the oxidation is a surface phenomenon, the oxidation process is accelerated
when more surface area is created for the oils. The British studies, using oils in a thin film at
room temperature, accelerated the autoxidation process. Industrial applications which require
spraying of oil create an aerosol of the oil which dramatically increases the surface area over
that of bulk oil. Autoxidation processes are accelerated in these spraying applications,
28
-------
thereby enhancing the production of volatile oxidation products.
6.2 APPLICATION OF GAS CHROMATOGRAPHY/MASS SPECTROMETRY
TECHNIQUES
Autoxidation of vegetable seed oils with the resulting formation of odors and flavors
has been studied extensively for many years. Oil manufacturers, oil chemists, and chemists in
the areas of flavors and fragrances have extensively studied the evolution of volatile material
from oils under various conditions. Autoxidation of oils proceeds slowly under room
temperature conditions, but the reactions still occur: oil becomes yellow and darkens under
prolonged storage. If the period of storage is long enough or if conditions are favorable for
accelerated oxidation, the oil becomes rancid. In the last ten to twenty years, the powerful
combined techniques of gas chromatography and mass spectrometry have been available to
characterize the products obtained from the oxidation of oils under various conditions. For
the most part, investigators have been concerned with the oxidation of oils used in cooking
applications, since heat accelerates the oxidation process and volatile decomposition products
formed in oils used in cooking can affect the taste and smell of foods cooked in the oil. No
formal reference test methodology has been available for these applications: investigators
have developed their own methodology to meet their needs in characterizing the volatile
materials emitted from the oils. Quantitative data have not been obtained because the
decomposition process in heated oils is an ongoing process. Oxidation reactions proceed as
long as the oil is heated and sites of reaction (double bonds) remain in the triglyceride
molecule. Thus, volatile decomposition products will continue to form until no more reaction
sites are available and the oil is decomposed to a tar.
While the primary concern has been with the cooking process, studies- have been
performed at room temperature. In order to identify the volatiles formed during the storage
of soybean oil, a time versus volatile decomposition study was performed on oil aged under
normal laboratory conditions.103 The experiment involved vacuum stripping of volatile
materials from the oils over a period of five to six hours. The volatile materials stripped
29
-------
from the oils were collected cryogenically, the cryotrap was heated, and the ultimate analysis
was performed by gas chromatography/mass spectrometry. The investigators found that
hydrocarbons and aldehydes were the principal volatile compounds which resulted from
soybean autoxidation by comparing fresh oil with oils which had been allowed to sit under
laboratory conditions for a period of six weeks. The oil which had been allowed to sit
exhibited larger quantities of the hydrocarbons and aldehydes, but new compounds were not
formed as the oil aged over the longer time period.
Most other studies available in the published literature have been performed on the
headspace generated above heated oils, with qualitative analysis of the organic compounds
emitted from the oil under these conditions.104'105 The compounds which are observed are
characteristic of the oxidation/decomposition of a fatty acid triglyceride and include
hydrocarbons of various lengths and conditions of saturation, as well as aldehydes and
ketones. In most cases, volatiles were generated in the headspace above oils heated to 180°C,
since 180°C is the temperature used in cooking applications. In general, fresh oil samples
produce few chromatographic peaks of low intensity, whereas the oils which had been aged
for varying periods of time produced relatively more volatile materials.1 In addition, the
volatile compounds identified were those expected from the autoxidation of the unsaturated
fatty acid components of each of the vegetable seed oils.
6.3 VOC TEST RESULTS
The Emissions Measurement Branch (EMB) conducted RM-24 and RM-24A testing
(Appendix i) on various vegetable seed oils for the purpose of this study at the Research
Triangle Park laboratories. For the Method 24 testing, the sample was weighed, heated in a
forced draft oven at 11()°C for 1 hour, then re-weighed. For Method 24A, the heating time
was 24 hours at 120°C in a forced draft oven. Alternatively, for "Method 24A, the heating
may take place for 0.5 hour in a vacuum oven, but the atmospheric pressure technique was
used for these samples. Test procedures shown in Appendix i were followed, with the
exceptions that no solvent was added to the sample and no moisture determinations were
30
-------
performed.
Method 24 testing was performed on two samples of soybean oil purchased from the
grocery store (Food Club vegetable oil, Pittsburgh, PA; and Wesson oil, Durham, NC). The
results of the Method 24 "determinations, shown in Appendix ii, showed that these oils did not
exhibit weight loss, indicating that these samples contained no VOC.
Method 24A testing was performed for eleven vegetable oil samples representing
.seven different oils, shown in Table 6-1. The vernonia oil was supplied by Stoil Dirlikov of
the Coatings Research Institute at Eastern Michigan University. Other oil samples were
supplied by David Ailor. Director of Regulatory Affairs, of the National Oilseed Processors
Association (NOPA). The results of the R-M-24A testing are shown in Appendix iii, and all
of the test results are summarized in Table 6-1. Table 6-1 shows that all of the samples
except for vernonia oil exhibited a gain in weight ranging from 0.01 percent to 3.56 percent.
In addition, under the testing conditions, several of the oils solidified or crystallized,
indicative of a change in physical condition and/or chemical composition. A solid structure is
indicative of cross-linking between molecules or components of molecules, or oxidation with
a concomitant gain in weight. However, there are multiple oxidation reactions which could
occur, ranging from the formation of peroxides to formation of alcohols, aldehydes, ketones,
and acids. Where the oils have changed structure, it is also possible that volatile materials are
entrained in the solid oils, trapped beneath the surface film. These volatile materials will
eventually permeate the surface film and be emitted from the oil, but the timetable for this
permeation process may be hours or days. The oils were not homogeneous in composition
prior to the heating process, so the number of potential reactions is so large that it is not
possible to assess accurately how much oxidation has occurred in order to account
quantitatively for the gain in weight observed under the test conditions of Method 24A.
Further, this inability to determine the weight gain attributable to oxidation precludes our
ability to judge the amount of VOC evolved by the oil during the fractionation - oxidation
reaction. What did seem clear is that enforcement of air pollution regulations using RM-24
and 24A would not falsely jeopardize a source using coatings with a vegetable oil constituent.
31
-------
Table 6-1 Test Results
EPA RM-24A RESULTS
(% VOC or Average
weight loss)
EPA RM-24 RESULTS
(% VOC or Average
weight loss)
Canola
Castor
Coconut
Cottonseed
Linseed"1
Saffiower
Soybean
Sunflower
Vemonia0
-0.87
-0.01, -0.63
-2.14. -2.30
-0.51, -0.66. -1.25
-0.62
-3.56
+3.54
0, +0.1, -0.1, -0.3
solidified : crvstallized ; cdried -
The results include the organics and moisture. Different samples of each vegetable oil
were used for each test.
-------
7.0 IMPACT OF TEST RESULTS ON TEST METHODS AND REGULATIONS
The results obtained by the application of Method 24 to soybean oil demonstrate that
this oil and, by extrapolation, other vegetable seed oils do not contain VOC. Since the .
literature provides no obvious evidence that soybean oil can be distinguished from other
vegetable seed oils in any significant way from the standpoint of volatility or photoreactivity,
the obsen/ations have been extended to the other vegetable seed oils. Thus, the presence of
vegetable seed oils in coatings should neither mask nor interfere significantly with Method 24
testing of coatings. Measurement of VOC content of a coating by RM-24A may yield an
erroneous value due to possible weight gain by the oil in the coating under the conditions of
the test. However, the degree'of the error is both small and will provide a margin of safety
to the source since it would indicate less VOC than what is actually emitted. Therefore, there
should be no significant impact on any of the coating regulatory compliance programs.
It should be noted, however, that this conclusion is not the same as concluding that
vegetable oils do not contribute to ozone formation in the troposphere. RM-24 and RM-24A
test conditions do not subject the oils to the temperature used in frying applications: a
temperature of 180°C and atmospheric pressure is used in most frying applications, although
some commercial frying is performed under pressure. Analytical results reported in the
literature show that under the temperature conditions used for frying with soybean oil, VOC's
.Mich as hydrocarbons and oxygenated compounds are emitted to the atmosphere.107 Since "
those VOC's are formed as a result of the temperature-catalyzed oxidation/decomposition of
the fatty acid triglycerides which make^ up the oils, the decomposition process occurs at an
accelerated rate as long as the oils are subjected to the elevated temperature, and sites of
unsaturation remain in the fatty acid carbon chains. Also, when food (mostly complex
carbohydrates) is cooked at temperatures of 180°C in heated oil, the composition of the food
is changed and volatile compounds are emitted from the food. This evolution of volatile
compounds is evidenced by the odors when foods are cooked. Thus, cooking food also
contributes to VOC emissions, although the magnitude of the contribution is unknown.
33
-------
Therefore, although the vegetable seed oils themselves contain no VOC's when analyzed
by Method 24 or Method 24A. the oils are precursors of VOC which are formed very slowly
under conditions of standard temperature and at an accelerated rate under elevated
temperature. The oxidation/decomposition reactions are also accelerated when the oil is
present as an aerosol or mist,, since a .far greater surface area is created.
The only way to determine the volatile organic compound emissions from a deep fat
frying operation would be to perform a stack test in which the emissions are speciated and
quantified. The results, however, would be controversial. Aerosol oils, for example, would
be measured as paniculate matter. In as much as the oxidation, would the entire mass be
considered VOC? Obviously it should not be, but what portion should be appropriately
tagged for future cleavage? Clearly, such testing would yield a conservative estimate of VOC
since the contribution of airborne oxidation of soybean oil to the formation of ozone
precursors is not instantaneous.
34
-------
8.0 REFERENCES
1. Vegetable Oils. Chemical and Process Technology Encyclopedia, 1974. p. 1129.
2. Ibid.
3. Ibid., p. 1130.
4. Ibid., p. 1131.
5. Fox. F. Oils for Organic Coatings. Federation Series and Coatings Technology.
Philadelphia. Pennsylvania. August 1974. 47 pp.
h. Ibid., p. 20.
Vegetable Oils. p. 1132.
S. Ibid. p. 1133.
l>. Ibid.
10. Vegetable Oils. p. 1135.
11. Riegel's Handbook of Industrial Chemistry. Van Nostrand Reinhold. New York, 1983.
12. Fox. p. 20.
13. Ibid.
14. Hawley's Condensed Chemical Dictionary, llth Edition. Van Nostrand Reinhold. New
York. 1987.
15. Ibid.
16. Bailey's Industrial Oil & Fat Products. Vol. I. 4th Edition, 1979.
17. Glossary for Air Pollution Control of Industrial Coating Operations, Second Edition.
U.S. Environmental Protection Agency. Research Triangle Park, North Carolina.
Publication No. EPA-450/3-83-of3R.' December 1983."p. 22.
18. Bailey's Industrial Oil & Fat Products.
-------
19. Riegel's Handbook.
20. Canola Council of Canada. Product Information. 1990.
21. Blaney, C. State Sees a Future in. Canola Plant. The News and Observer. December
10, 1990.
22. Canola Council of Canada.
23. CPI Purchasing. October 1990. pp. 53, 57.
24. Supermarket News. January 29, 1990. p. 26.
25. Fox. p. 35.
26. Eckey, E. Vegetable Fats and Oiis. Remhold Publishing Corporation. New York.
1954' p. 596.~
27. Fox. p. 36.
28. Cost Information. American Paint and Coatings Journal. November 5, 1990. p. 24.
29. Fox. p. 34.
30. Ibid., p. 35.
31. Ibid.
32. Eckey, p. 320.
33. Research Studies. September 10, 1990. pp. 1-102.
34. Ibid., p. 640.
35. Ibid., p. 659.
36. Research Studies.
37. Fox, p. 35.
38. Ibid., p. 31.
39. Ibid., p. 32.
-------
40. Ibid., p. 31.
41. Cassens. Daniel L. and Feist. William C. Finishing Wood Exteriors. US Department
of Agriculture Agriculture Handbood No.. 647, May 1986
42. CPI Purchasing.
43. Fox, p. 32.
44. Ibid.
45. Sinclair and Valentine. Product Information. October 1990.
46. Eckey. p. 513.
47. Ibid., p. 514.
48. Moeller, R. Mothers of Tomorrow's Printers May Gaze Fondly on Their Offsprings
and Sigh, "The Child Has Soybean Oil in his Veins." American Ink Maker. January
1989. ~
49. Cost Information.
50. Eckey. p. 511.
51. Ibid.
52. Ibid., p. 512.
53. Sakla. A., et al. The Effect of Environmental Conditions on the Chemical
Composition of Soybean Seeds: Relationship Between the Protein, Oil, Carbohydrate
and Trypsin Inhibitor Content. Food Chemistry. November 1987.
54. Al-Kahtani, H. Quality of Soybeans and Their Crude Oils in Saudi Arabia. JAOCS,
Vol. 66, no. 1, January 1989.
55. Eckey. p. 510.
56. Chemical Marketing Reporter. May 7. 1990. p. 27.
57. Journal of Commerce. July 26. 1990. p. 7A.
58. Research Studies.
37
-------
59. Fox. p. 33.
60. Ibid
61: Ibid.
62. Moeller, p. 22.
63. Ibid.
64. Telecon. Madeleine Strum. EPA, with Stanley Field, Flint Inks, April 11, 1991.
65. Telecon. Candace" Blackley, Radian Corporation, with Ruth Felland, American
Newspaper Association. February 19, 1991.
66. Eckey. p. 772.
67. Ibid., p. 775.
68. Ibid.
69. Ibid., p. 776.
70. Ibid., p. 570.
71. Ibid., p. 574.
72. Ibid., p. 570.
73. Fox. p. 37.
74. Telecon. Candace Blackley, Radian Corporation, with Stoil Dirlikov, Coatings
Research Institute. March 8, 1991.
75. Dirlikov, S., and M. Islam. Vernonia Oil: A New Reactive Diluent. Modem Paint
and Coatings. August 1990.
76. Ibid.
77. Ibid.
78. Moeller.
79. Ibid.
38
-------
80. McCabe, A. Continuous Alkali Refining of Vegetable Oils. Food Processing
Industry. February/March 1974.
81. Moeller.
82. Wiedermann, L. Degumming, Refining, and Bleaching Soybean Oil. JAOCS. March
1981.
33. Bailey's Industrial Oil and Fat Products, Vol. I, 4th Edition, 1979.
84. Moeller.
85. Telecon. Candace Blackley, Radian Corporation, with Cecil Wilcoxon, Townsends,
Incorporated. January 28, 1991.
86. Moeller.
S7. Hare, Clive H. Anatomy of Paint. Journal of Protective Coatings and Linings.
October 1989.
88. Ibid.
89. Eckey.
90. Hutchinson. G.H. Some Aspects of Drying Oils Technology, JOCCA. 1973.
91. Hawley's Condensed Chemical Dictionary.
92. Ibid.
93. Ibid.
94. Eckey.
95. Hare.
96. Ibid.
97. Hancock. R. A., and N. J. Leeves. Studies in Autoxidation. Pan I. The Volatile
By-Products Resulting from the Autoxidation of Unsaturated Fatty Acid Methyl Esters.
Progress in Organic Coatings. Vol. 17, p. 321-336, 1989.
39
-------
98. Joseph, Ron, and Associates. Getting Into Compliance With Environmental
Regulations for Paints. Coatings, and Printing Facilities. 1990.
99. EPA Glossary, Industrial Finishing Magazine. August 1990.
100. Eckey.
101. Telecon. Madeleine Strum, U.S. Environmental Protection Agency, with Chris
Serauskas, Lauter Chemical. November 2, 1990.
102. The Autoxidation of Drying Oils. IV. The Volatile Products of Oxidation. Technical
Paper No. 159, Vol. 8, No. 4. The Research Association of British Paint, Colour, and
Varnish Manufacturers. January, 1949.
103. Selke. E., H. A. Moser, and W. K. Rohwedder. Tandem Gas Chromatography-Mass
Spectrometry Analysis of Volatiles from Soybean Oil. JAOCS, Vol. 47, p. 393-397,
1970.
104. Hancock, R. A., and N. J. Leeves. Studies in Autoxidation. Part I. The Volatile
By-Products Resulting from the Autoxidation of Unsaturated Fatty Acid Methyl Esters.
Progress in Organic Coatings. Vol. 17, p. 321-336, 1989.
105. Snyder, J. M., E. N. Frankel, and E. Selke. Capillary Gas Chromatographic Analyses I
of Headspace Volatiles from Vegetable Oils. JAOCS, Vol. 62, No. 12. p. 1675-1679, "
1985.
106. Ibid.
107. Ibid.
40
-------
APPENDICES
Appendix i
VOC Test Methods
EPA Reference Method 24
EPA Reference Method 24a
(40 CFR Ch. 1, Pt. 60. App. A, 7-1-90 Edition)
ASTM D2369
ASTM D4017
ASTM D3792
(Annual Book of ASTM Standards, Vol 06.01, 1987)
Appendix ii
EPA RM-24 Test Results
Appendix iii
EPA RM-24A Test Results
Appendix iv
Frying Operations References
41
-------
-------
APPENDIX i
42
-------
-------
190
Environmental Protvcrton Agency
Pt. 60, App. A, Merh. 24
iaxno* 24— DtxnucniATioK or VOUHTLI
MATTO CONTEXT. WATXB Coirmrr. Dm-
smr, VOLUME Sous*. AJTO WCZOHT
Souo* or StntrACt COATXXG*
1. .Applicability and Prtnctpte
1.1 Applicability. This method applies to
the determination of volatile nutter con-
tent. water content, density, volume solids.
and weight solid* of paint, varnish. lacquer.
or related lurt ace coating*.
1.2 Principle. Standard methods are uied
to determine the volatile matter content.
water content, density, volume solids, and
weight solids of the paint, varnish, lacquer.
or related surface coattna.
2. jtppiicaote Standard JfetfuxU
Use the apparatus, reacents, and proce-
dure* specified in the standard methods
below:
2.1 ASTM D 1475-40 (Reapproved 1980).
Standard Test Method for Density of Paint.
Varnish. Lacquer, and Related Products (in-
corporated by reference— see I 90.17).
2.2 ASTM D3369-41. Standard Test
Method for Volatile Content of Coatinn
(Incorporated by reference— «ee t 90.17).
2.3 ASTM D3792-79. Standard Test
Method for Water Content of Water-Reduc-
ible Paints by Direct Injection Into a Oas
Chromatocraph (incorporated by refer-
ence—Me t 40.17).
2.4 ASTM D4017-81, Standard Test
Method for Water In Paint* and Paint Ma-
terial* by the Karl Fischer Tltration
Method (Incorporated by reference— «ee
i 40.17).
3. Procedure
3.1 Volatile Matter Content. Use the pro-
cedure in ASTM D2369-81 (Incorporated by
reference — se« 1 80.17) to determine the
volatile matter content (may Include water)
of the ""*Mng Record the following infor-
mation:
W,- Weight of dish and sample before heat-
ing, g.
W,- Weight of dish and sample after heat-
ing. f.
W.- Sample weight, f.
Run analyse* In pairs (duplicate set*) for
each coating until the criterion In Section
4.3 U met. Calculate the weight traction of
the volatile matter (W.) for each analyst* a*
follows:
W, .
W,-W.
W.
Eq. 24-1
3.2 Water Content. Por waterbome
(water reducible) costings only, determine
the welfht fraction of water (W.) usinc
either -Standard Content Method Test for
Water of Water-Reducible Paint* by Direct
- Injection Into a Oas Chromauxrraph" or
"Standard Test Method for Water in Paint
and Paint Materials by Karl Fischer
Method." (These two method* are Incorpo-
rated by reference— see IM.17.) A water-
borne coating ls any coating which contains
more than 5 percent water by weight in It*
volatile fraction. Run duplicate set* of de-
termination* until the criterion In Section
4.3 Is met. Record the arithmetic average
(W.).
3.3 Coating Density. Determine the den-
sity
-------
ft. 60, App. A, M»th. 24A
statement. If liter several attempts It 15
concluded that the ASTM procedures
cannot be used (or the specific coating with
the established within-laboratory precision.
the Administrator will assume responsibility
(or providing the necessary procedures for
revising the method or precision statements
upon written request to: Director. Emission
Standards and Engineering Division. (MD-
13) Office of Air Quality Pluming and
Standards. U.S. Environmental Protection
Agency. Research Triangle Park. NC 27711.
4.4 Confidence Limit Calculations (or
Waterborne Coatings. Based on the be-
tween-laboratory precision statements, cal-
culate the confidence limits (or waterbome
coatings as follows:
To calculate-the lower confidence limit.
subtract the appropriate between-laborato-
ry precision value from the measured mean
value for that parameter. To calculate the
upper confidence limit, add the appropriate
between-laboratory precision value to the
measured mean value for that parameter.
For W. and Dn use the lower confidence
limits, and for W.. use the upper confidence
limit. Because V, Is calculated, there is no
adjustment for the parameter.
5. Calculation*
S.I Nonaqueous Volatile Matter.
5.1.1 Solvent-borne Coatings.
W.-W. Eq. 24-2
Where:
W.-Weight fraction nonaqueous volatile
matter, g/g.
5.1.2 Waterbome Coatings.
W.-W.-W. Eq. 24-3
5.2 Weight Fraction Solids.
W.-l-W, Eq. 24-4
Where:
W,-Weight solids, g/g.
MSTHOO 34A—DRVIMXHATIOK or VOLATXLI
MATTTB Coirrorr AMD Dmsrrr or Panrr-
[NO IlOU AJTD RJELATO COATDMM
1. XppUcoMmv and PrineivU
1.1 Applicability. This method applies to
the determination of the volatile organic
compound iVOC) content and density of
solvent-borne (solvent reducible) printing
Inks or related """irtgr
1.2 Principle. Separate procedures are
used to determine the VOC weight fraction
and density of the coating and the density
of the solvent in the mating. The VOC
weight fraction Is determined by measuring
the weight IOM of a known sample quantity
which has been heated for a specified
length of time at a specified temperature.
The density of both the coating and solvent
are measured by a standard procedure.
Prom this Information, the VOC volume
fraction Is calculated.
40 CFI Ch. I (7-1-90 Editi*,,
2. Procedure
2.1 Weight Fraction VOC.
2.1.1 Apparatus.
2.1.1.1 Weighing Dishes. Aluminum rou
58 mm in diameter by 18 mm high, with *
flat bottom. There must be at least thr»»
weighing dishes per sample.
2.1.1.2 Disposable Syringe. S ml.
2.1.1.3 Analytical Balance. To measure to
within 0.1 mg.
2.1.1.4 Oven. Vacuum oven capable or
maintaining a temperature of 120-2'C u^
an absolute pressure of 510 ±51 mm Hg far
4 hours. Alternatively, a forced draft oven
capable of maintaining a temperature of 120
rZ'C for 24 hours.
2.1.2 Analysis. Shake or mix the sample
thoroughly to assure that all the solids art
completely suspended. Label and weigh ta
the nearest 0.1 mg a weighing dish and
record this weight (M^).
Using a 5-tnl syringe without a needle
remove a sample of the coating. Weigh the
syringe and sample to the nearest 0.1 mj
and record this weight . Transfer 1 to
3 g of the sample to the tared weighlnt
dish. Rewelgh the syringe and sample to the
nearest 0.1 mg and record this weight (Mm)
Heat the weighing dish and sample in i
vacuum oven at an absolute pressure of 510
=51 mm Hg sad a temperature of 130 t2*C
(or 4 hours. Alternatively, heat the welch-
ing dish and sample in a forced draft oven
"at a temperature of 120 ±2'C for 24 hours.
After the welching dish has cooled, reweigh
It to the nearest 0.1 mg and record the
weight (M*>. Repeat this procedure (or i
total of three determinations for each
sample.
2.2 Coating Density. Determine the den-
sity of the ink or related coating accordini
to the procedure outlined In ASTM D 147J-
90 (Reapproved 1980), (Incorporated by ref-
erence—see I 40.17).
2.3 Solvent Density. Determine the den-
sity of the solvent according to the proce-
dure outlined In ASTM D 1475-60 (reap-
proved 1980). Make a total of three determi-
nations (or each coating. Report the density
D. as the arithmetic average of the three
determinations.
3. Calculation*
3.1 Weight Fraction VOC. Calculate the
weight fraction volatile organic content W.
using the following equation:
M,
W. -
aeport the weight
average
Volume Fra
volume fraction v<
the followuu
V. >
4.
4.1 Standard T<
p^int. Varnish. Li
uct*. ASTM Desli
proved 1980).
4.3 Teleconvera
Corporal
Qorporati
Ink Anal:
4.3 Teleeonvers
Robert. Oravure
Burt. Rick. Radii
S, 1079. Oravure I
25 — Dm
OUS NOKMTTH,
GAJLSON
1.1
the measurement
pound* (VOC) as
organic* (TOKM
emissions. Organ
Interfere with th<
paniculate filter
detectable for U
carbon.
When carbon
vapor are preaei
they can produi
sample. The mag
on the concentr
vapor. As a guide
centratlon, expr
time* the water >
product does not
considered Insigi
bias Is. not signif 1
percent CO, and
it would be slum
detection limit t
20 percent water
This method li
applies) to the
Costs, logistics,
source testing mi
980
-------
I (7-1-90 Edition)
n VOC
>UheS. Aluminum (oil.
y 18 mm high, with a
nuat be at least three
unpie.
Synnge. 5 ml.
Salance. To measure to
lum oven capable of
rature of 120 = 2'C and
of 510 =51 mm Hg for
y. a forced draft oven
ig a temperature of 120
ne or mix the sample
•.hat ail the solids are
|d. Label and weigh to
a weighing dish and
Inge without a needle
the coating. Weigh the
|u> the nearest 0.1 mg
fit (M.Y,;. Transfer 1 to
the tired weighing
inze and sample to the
cord this weight (M.n).
dish and sample in a
bsolute pressure of S10
mperature of 120 ±2'C
lively, heat the weigh-
in a forced draft oven
±2'C for 24 hours.
. has cooled, reweigh
'mg and record the
"this procedure for a
erminatlons for each
ty Determine the den-
iated-coating according
lined In ASTM D 1475-
). (incorporated by ref-
ty. Determine the den-
Lccordlng to the proce-
JTM D 147S-80 (reap-
, total of three detenni-
ling. Report the density
§: average of the three
Ion VOC. Calculate the
Kile organic content W.
|q oat ion.
Environmental Protection Agoncy
w.,
Eq. 24A-1
Report the weight fraction VOC W. as the
Arithmetic average of the three determina-
tions.
3.2 Volume Fraction VOC. Calculate the
volume fraction volatile organic content V.
using the following equation:
V.-(W.D./D.
Eq. 24A-2
4. SioiioempAv
4.1 Standard Test Method for Density of
paint. Varnish. Lacquer, and Related Prod-
ucts. A3TM Designation D 1475-40 (Reap-
proved 1980).
4.2 Teleconvenatlon. Wright. Chuck.
lomont Corporation with Reich. R. A..
Radian Corporation. September 29. 1979.
Oravure Ink Analysts.
4.3 Teleconversatlon. Oppenheimer.
Robert. Orarure Research Institute with
Bun. Rick. Radian Corporation. November
j. 1979. Oravure Ink Analysis.
&LKTHOD 26—DBTOMXHATIOK or TOTAL OAO>
o0s NomorrsAjri OEOAJTIC Emssiows AS
CAUON
1. AmUcaMHtv and Principle
1.1 Applicability. This method applies to
the measurement of volatile organic com-
pounds (VOC) as total gaseous nonmethane
organic* (TONMO) as carbon in source
emissions. Organic paniculate matter wul
interfere with the analysis and. therefore, a
paniculate filter is required. The minimum
detectable for the method is 50 ppm as
csrbon.
When carbon dioxide (CO,) and water
vapor are present toother in the stack.
tney can produce a positive bias in the
ample. The magnitude of the bias depends
on the concentrations of COt and water
vapor. As a guideline, multiply the CO, con-
centration, expressed as volume percent.
tunes the water vapor concentration. IT this
product does not exceed 100. the bias can be
considered iTnignifb^mt Por example, the
bias Is not significant for a source ha vine 10
percent COi and 10 percent water vapor, but
tt would be significant for a source near the
detection limit having 10 percent CO, and
20 percent water vapor.
This method Is not the only method that
applies to the measurement of TONMO.
Costs, logistics, and other practical)ties of
wurce testing may make other test methods
Ft. 60, App. A, Moth. 25
more desirable for measuring VOC contents
of certain effluent streams. Proper Judg-
ment is required In determining the most
applicable VOC test method. Por example.
depending upon the molecular weight of the
ornnlcs In the effluent stream, a totally
automated semicontlauous nonmethane or-
games (NMO) analyser Interfaced directly
to the source may yield accurate results.
This approach has the advantage of provid-
ing emission data semicontlnuously over an
extended time period.
Direct measurement of an effluent with a
flame ionizatlon detector (FID) analyzer
may be appropriate with prior characteriza-
tion of the gas stream and knowledge that
the detector responds predictably to the or-
ganic compounds In the stream. If present.
methane (CHJ wul. of course, also be meas-
ured. The FID can be applied to the deter-
mination of the mass concentration of the
total molecular structure of the organic
emissions under any of the following limited
conditions: (1) Where only one compound is
known to exist: (2) when the organic com-
pounds consist of only hydrogen and
carbon: (3) where the relative percentages
of the compounds are known or can be de-
termined, and the FID responses to the
compounds are known: (4) where a consist-
ent mixture of the compounds exists before
and after emission control and only the rela-
tive concentrations are to be assessed: or (5)
where the FID can be calibrated against
mas* standard* of the compounds emitted
(solvent emission*, for example).
Another example of the use of a direct
FID Is as a screening method. If there is
enough Information available to provide a
rough estimate of the analyzer accuracy.
the FID analyser can be used to determine
the VOC content of an uncnaracterued gas
stream, With a sufficient buffer to account
for possible inaccuracies, the direct FID can
be a useful tool to obtain the desired results
without costly exact determination.
In situations where a qualitative/quanti-
tative analysis of an effluent stream Is de-
sired or required, a gas chromatocraphlc
FID system may apply. However, for
sources emitting numerous organics, the
time and expense of this approach will be
formidable.
1.2 Principle. An emission sample Is with-
drawn from the stack at a constant rate
through a heated filter and a chilled con-
densate trap by means of an evacuated
sample •«»»» After sampling Is completed.
the TONMO are determined by independ-
ently analyzing the condensate trap and
sample tank fractions and combining the
analytical results. The organic content of
the condenwte trap fraction Is determined
by oxidizing the KUO to CO, and quantita-
tively collecting the effluent in an evacuat-
ed vessel: then a ponton of the CO, is re-
981
-------
-------
Designation: 0 2369 - 87
>t2
Standard Test Method for
Volatile Content of Coatings1
Ti;s standard is issued under the fixed designation D 236"* '.he numner immediaielv following the designation indicates the vear of
^ncinai adoption or. in the case of revision ine veai ot last reMsion A number in parentheses indicates me vear ol last reapprovaj \
superscript epsilon u> indicates an editorial cnange since tne iast revision or rcapproval
r-!i < ,'f 5: Teinod ha.', revn approvea .'or we r\ agencies in me Department be
suostituted with mutual agreement of the producer and user.
See Note 3.
1.5 Tnis stanaard ma\ involve hazardous materials, oper-
ations, and equipment. This standard does not purport to
aadress ail of the satetv problems associated with its use It is
tne responsibility of the user of this standard to establish
Appropriate saletv and health practices and determine the
appitcac:itt\ ot reguiaton, limitations prior to use \ specific
hazard statement is given in 7 3.1.
-. Referenced Documents
2 1 AST.U S;andards
D343 Specification for 2-Ethoxyeth>i Acetate (95^
Grader
D 362 Specification for Industrial Grade Toluene-'
D 1193 Specification for Reagent Water"
E 145 Specification for Gravity Convection and Forced-
Ventilation Ovens5
E 180 Practice for Determining the Precision Data of
ASTM Methods for Analysis and Testing of Industrial
Chemicals0
3. Summary of Test Method
3.1 A designated quantity of coating specimen is weighed
into an aluminum foil dish containing 3 mL of an appro-
priate solvent, dispersed, and heated in an oven at 110 ± 5*C
for 60 mm. The percent volatile is calculated from the loss in
weight.
4. Significance and Use
4 1 This test method is the procedure of choice for
determining volatiles in coatings for the purpose of calcu-
lating the volatile organic content in coatings under specified
test conditions. The inverse value, nonvolatile, is used to
determine the weight percent solids content. This informa-
tion is useful to the paint producer and user and to
environmental interests for determining tne \oiatiles emitted
by coatings.
5. Apparatus
5.1 Aluminum Foil Dish. 58 mm in diameter b> 18 mm
high with a smooth I planar i bottom surface. Precondition
the dishes for 30 mm in an oven at 110 r 5°C and store in a
desiccator pnor to use.
5.2 Forcea Dratt Oven, T>pe 11A or T>pe IIB as specified
in Specification E 145
5.3 Svnnim. 5-mL. capable of properK dispensing the
coating under test at sufficient rate that tne specimen can be
dissolved in the solvent (see " 2)
5.4 Test Tube, with new cork stopper
5.5 ll'etghim? or Dropping Bottle
This '.esi rnetrtoe •! under the mnsdiction ot ^STM Committee D-l on Paint
'id Re:a'.fC Coatings and Materials and is the direct responsiDimv 01 Suocom-
'nine* L>" ; -n Comical Anaivsis of Paints ana Paint viateiais
'-'JIT?"* eCiucn .irDroveu June .0 i-^7 Published AuauM -" ~ ' "iir.^ls
JuBiisneij _' J I ""--'>- " ._jsi rresiou^ edition D !;^*> - •>"
Discontinued, see /WC \nnuul BOOK <>i 4STV S'^naaras Pan >
1 ^r>nual tto'ai m 4ST\I ^lanaura:, v ol u*ui
nuardi VQI • 1 Oi
-------
45ft) D 2369
o. Reagents
K 1 P\r :\ ,>t Reagents—Reagent grade chemicals shall be
-sod in ar tests. Unless otherwise indicated, it is intended
:r-._: all r;agents shall conform to the specifications of the
L , rrmittee on Analytical Reagents of the American Chem-
-.u. Societv where such specifications are available. Other
grades rr,a> oe used, provided it is first ascertained that the
'eaesr.t is of sufficiently high punty to permit us use without
.-ssening ne accuracy of the determination.
- 2 Pitr;-\ ot \\'aier—Unless otherwise indicated, refer-
;~;es :o water shall be understood to mean T>pe II of
Specification D 1 193.
- 3 Ti'.'.tene. technical grade. Specification D 362.
- 4 .?-£.•'i.jvm/u1/ Acetate, technical grade. Specification
". Procedure
1 Mix the sample, preferably on a mecnamcal shaker or
roiler. until homogeneous. If air bubbles become entrapped.
;•.:: t?> hand until the air has been removed.
" 2 Using an appropnate weighing container (5 3. 5 4. or
f 5 with the svnnge preferred for highest precisioni. weigh to
: mg. b\ difference, a specimen of 0.30 = 0.10 g for
:3atmgs believed to have a volatile content less than 40
weight '"c or a specimen of 0.50 ± 0.10 g for coatings believed
to nave a volatile content greater than 40 weight °c. into a
:ared aluminum foil dish (5.5* into which has been added 3
r ! mL of suitable solvent (6.2. 6.3. or 6 4). Add the
specimen dropwise. shaking (swirling) the dish to disperse
tr.e specimen completely in the solvent. If the material forms
a lump that cannot be dispersed, discard the specimen and
prepare a new one. Similarly prepare a duplicate
NOTE 2—If the specimen cannot be dispersed in the solvents listed
*• -. 6 3. or o 4) a compatible solvent may be substituted provided it is
-.0 less volatile than I-ethoxyethvi acetate (6 •»>
" 3 Heat the aluminum foil dishes containing the dis-
persed specimens in the forced draft oven (5 2) for t>0 mm at
iO r 5'C
~ 3 ! Warning—Provide adequate ventilation, consistent
•Mth accepted laborator> practice, to prevent solvent vapors
"Reagent Chemicals. American Chemical Societv Specifications." Am Chem-
:ai Soc Washington DC For suggestions on the testing ot reagents not listed bv
".e American Chemical Societv. see "Reagent Chemicals and Standards." bv
.' jsepn Rosin D Van Nostrand Co . Inc.. Sew \ork. VY ana the United Slates
Pharmacopeia '
from accumulating to a dangerous level.
7.4 Remove the dishes from the oven, place immediateK
in a desiccator, cool to amoient temperature and weigh to 0. i
mg.
NOTE 3—If unusual decomposition or degradation of the specimen
occurs during heating, the actual time and temperature used to cure the
coating in practice mav be substituted for the time and temperature
specified in this test method, subject to mutual agreement of the
producer and user.
8. Calculations
8.1 Calculate the percent volatile matter. I', in the liquid
coating as follows:
r % - 100- [«»: - »',)/S) x ioo]
w-here:
H', = weight of dish.
ir, = weight of dish plus specimen after heating, and
S = specimen weight.
8.2 The percent of nonvolatile matter. .V. m the coating
may be calculated by difference as follows:
V, "I = 100 - volatile matter
9. Precision and Bias
9.1 The precision estimated for tests at 60 mm at 110 ±
5°C are based on an interlaboratory study3 in which 1
operator in each of 15 laboratories analyzed in duplicate on 2
different days 7 samples of water-based paints and 8 samples
of solvent-based paints containing between 35 and 72 °c
volatile material. The paints were commercially supplied.
The results were analyzed statistically m accordance with
Practice E 180. The within-laboratory coefficient of variation
was found to be 0.5 % relative at 213 degrees of freedom and
the between-laboratones coefficient of variation was 1.7 ^
relative at 198 degrees of freedom. Based on these coeffi-
cients, the following criteria should be used for judging the
acceptability of results at the 95 ^'confidence level.
9.1.1 Repeatability—Two results, each the mean of dupli-
cate determinations, obtained by the same operator on
different days should be considered suspect if they differ b>
more than 1.5 'c relative.
9 1.2 Reproduability—T'HO results, each the mean of
duplicate determinations, obtained by operators in different
laboratories should be considered suspect if they differ by
more than 4.7 °l relative.
9 2 Bias—Bias has not been determined.
1 Supporting data are availaBle from ASTM Headquarters Request RR
001-1026
-------
D2369
APPENDIX
(Nonmandatory Information*
'. ! Oven residence time of 10 mm for the paint test
imen at 110 ± 5*C was the ongmal procedure for this
method. For information purposes, the precision state-
is for 20-mm residence time are:
', 1 ! The precision estimates are based on an mterlabo-'
n study m which 1 operator in each of 15 laboratories
vzed in duplicate on 2 different days 7 samples of
r-based paints and 8 samples of solvent-based paints
.aming from 35 to ^2 ^ volatile material. The paints
1 commercially supplied. The results were anahzed
sticaily in accordance with Practice E 180. The within-
raion coefficient of variation was found to be 1 1 ^c
relative at 193 degrees of freedom and the between-labora-
tones coefficient of variation was 2.5 ro relative at 178
degrees of freedom. Based on these coefficients me following
criteria should be used for judging the acceptability of results
at the 95 ^ confidence level.
u) Repeatability—Two results each the mean of duplicate
determinations, obtained by the same operator on different
days, should be considered suspect if they differ by more
than 2.9 ^ relative.
is) Reproducibilitv—Two results, each the mean of dupli-
cate determinations, obtained by operators in different
laboratones should be considered suspect if they differ by
more than
relative.
"Ka American Society 'or Testing ana Materials :axiu no position rtsoecung trie vaiiarty of any often rights assertea in connection
**" any uem mentioned :n ri/s sra/xJaro Users ot 'His sttnovo art txortss'y aovista tnat aettrminwion ot ro« validity ot any sucr
:-ateni ngms ana int "s* of mtnngtmtm at sucr) rtqnu are entirety ineir own rasoonsioiirty
~":s stanaara is suoiec! to revision at any time oy 'ne resoonsiOte technical committee ana must D» ravieww ev«ry five years ana
' -or revisea Mner reaoorovea or wrtnartwn Tour comments are invitea eitrttr tor revision ot tna snnawa or tor aaOiticnti sltnatras
ara snouia oe aaaressea to ASTM Hetaautrters Your comments will receive cvetui consiaerttion at a mtming of th« resoonsioia
•ecnmcai commmett wnicn you may ertena II you ten tnat your comments nave riot receiv*} a lair netting you snouio mate your
< ews Known to :ne ASTM Committee on Stanaaras. '9'6 Race St, Pniiaaeionia. PA 19103
-------
-------
Designation: D 4017 - 88
Standard Test Method for
Water in Paints and Paint Materials by Karl Fischer Method1
Tuis Mjndjrd is 'ssucU jnder me n\ej acM^naiion D -10! ~ 'he numrwr immediately follow IRK the designation indicates trie <.JT .M
Tit-'nji j^option or m me ca»< ol reMSion me ^eJ^ ui last revision ^ numner in parentheses maicaies ;he '.ear oi .asi reappro1* ji , \
superscript fpstton i* mdicjics jn cdiion..! v.hanae since t-ne usi reMSion or reappro^ai
1. Scope
1.1 This test method is applicable to all paints and paint
materials, including resins, monomers, and solvents, with the
exception of aldehydes and certain acti\e metals, metal
oxides, and metal hydroxides. While the evaluation was
limited to pigmemed products containing amounts of water
in the 50 to "0 ^c range, there is reason to believe that higher
and loNver concentrations can be determined by this test
method.
1.2 This standard may involve Hazardous materials oper-
ations, ana equipment This standard does not purport to
address ail or the safety problems associated *ith its use It is
the responsibilit\ ot the user of this standard to establish
appropriate sa/en and heanh practices and determine the
applicant!ir. ot revutaton- iimitations prior to use Specific
hazard statements are given in Section 7
2. Referenced Documents
11 ASTM Standards
D 1193 Specification for Reagent Water
E 180 Practice for Determining the Precision of ASTM
Methods for Analysis and Testing of Industrial
Chemicals'"
E 203 Test Method for Water Using Karl Fischer Reagenr
3. Summary of Test Method
3 1 The material is dissolved in pyndme. or another
appropriate solvent, and titrated directly with standardized
Karl Fischer Reagent, to an electrometnc end point The
sluggish reaction with water in pyndme is accelerated with a
chemical catalyst. 1-ethylpipendine.
3.2 Pyndme is used as a solvent to minimize interference
problems caused by ketones. It is also used because the more
commen solvent, methanol. will not dissolve many common
resins and because methanol reacts with some resins to
produce water
4. Significance and Use
4 1 Control of water content is often important in control-
ling the performance of paint and paint ingredients, and it is
cntical m controlling volatile organic compound (VOC)
content.
4.2 Paint matenals are often insoluble in common Karl
Fischer solvents such as methanol. Pyndme has been found
to be a nearly universal solvent for these matenals: however.
the Kari Fischer reaction is too slow m that solvent at room
temperature. To speed it up. 1-ethyl-pipendme is added at
5 °c as a buffer, or "catalyst"
5. Apparatus
5.1 Karl Fischer Apparatus, manual or automatic, encom-
passed by the descnpuon in Test Method E 203. Apparatus
should be equipped with a 25-mL buret. Class A. or
equivalent.
5,.2 Syringe. 100-uL capacity, with needle.
5 3 Svrmqes. 1-mL and 10-mL capacity, without needle.
but equipped with caps.
6. Reagents
6.1 Purity oj Reagents—Reagent grade chemicals shall be
used in all tests. Unless otherwise indicated, it is intended
that all reagents shall conform to the specifications of the
Committee on Analytical Reagents of the Amencan Chem-
ical Society, where such specifications are available."1 Other
grades may be 'used, provided it is ascertained that the
reagent is of sufficiently high punty to permit its use without
lessening the accuracy of the determination.
6.2 Purity of Water—Unless otherwise indicated, refer-
ences to water shall be understood to mean reagent grade
water conforming to Type II of Specification D 1193
6.3 Karl Fischer Reagents
6.4 Pyndme
6.5 1-Eihylpipendine.
6.6 Hydrochloric Acid (HC1) Concentrated
7. Hazards
" 1 Karl Fischer reagent contains four toxic compounds.
namely iodine, sulfur dioxide, pyndme. and methanol or
glycol ether. The reagent should be prepared and dispensed
in a hood. Care must be exercised to avoid inhalation or skin
contact. Following accidental contact or spillage, wash with
large quantities of water.
".2 Pyndme and methanol solvents should be treated with
the same care as Karl Fischer reagent.
is under the jurisdiction o! \sTM Committee D-, on Pjint
and Materials and is "ic direu responsibility ,,i ^uncum-
cmitji VnuivMs -i Pjinis jnd Pjint Materials
prn.td ' V • '•"" PufMi-ned Sei.emrx."' '-"<>! I'-n-irjn.
4 "Heazcni Chemicals Amencan C nemical SiKie:* Specineanons \T,
i^ai X)c Washington CX For suyjesnons on me 'esting ui r-acen:1- not iis
ne Xmencan Lriemaai Viuets sei. Reauent r,icmicais -ind i'jnr.jrc
j iM.rr Kosm D V-n V>s(randt ) 'r^ Ne1.1- VJTK \> .ind :nc ' -liuJ
-------
04017
TABLE 1 Specimen Guidelines
Specimen
Aeigm
•0-20
•:-zo
::-25
•5-25
-c'.h\ ipipendine is of unknown toxicity and. there-
•;re snoald be handled with the same care as the above
materials
8. Procedure
- : ^'^r.aard;:j;;on of Karl Fischer Reagent:
? . ! Aad enough fresh pyndine to cover the electrode up.
plus i rr.L of 1-ethylpipendme catalyst per 20 mL of
P'.ndme Catahst performs best at a concentration of about
5 ~- of me solume present.
* 2 Fill the 100-uL synnge to about half full with
-nailed water and weigh to the nearest 0.1 mg.
^ 1 3 Pretitrate the pyndme to the end point indicated by
me equipment manufacturer, by adding just enough Karl
F-scher Reagent 11KFR) to cause the end point to hold for at
'east 50 s
5 i 5 I The use of the catalyst greatly increases the reac-
tion rate between water and Karl Fischer reagent. To obtain
reliaole results, increase the electrode sensitivity and reduce
•.itration rate to a minimum. Most instruments have controls
for these functions. Consult the instructional manual for
.nformauon on these controls.
S : 4 Empty the contents of the svnnee into the titrator
'•essei Immediately replace the stopper of the sample port
ana titrate with KFR to the end point 35 described in 8.1.3.
S I 5 Repeat standardization until replicate values of F
agree wuhin 1 °~c. Determine the mean of at least two such
Determinations Carp, out'calculations retaining at least one
extra decimal figure bevond that of the acquired data. Round
"T figures alter final calculations.
•> ! £ C. *;,cutaiion:
^ : o i Calculate the KFR titre F as follows:
¥ = J r1
wnere.
-' = water added, g. and
P = KFR used. mL.
The ^alue for F should be recorded to the four significant
J;g;ts and should be the mean of at least two determinations.
T-picai ^lues are in the range of 0.004000 to 0.006000
i rr;L
? 1 i"jr, s/s ui Samples ll'ith More Than n : °J Hater
^ 2 ' The utration vessel should already contain
p-euiratcd pvndine and catalvst. as descnbed in steps 8.1 1
-nd ? ! 3 in the standardization procedure. Best results are
• rtamed with fresh solvent, that is. contain no previously
'.I'.rated specimen m the vessel.
^22 With a 1-mL or 10-mL svnnge. draw the amount of
material indicated m Table 1
r 2 I Remove the svnnge trorn the specimen, pull the
r'^n.i'.r u u l:itie tunhcr -Aipe the -j^ess matenui oil'me
svnnge. and place a cap on the synnge tip VVeign me filled
synnge to the nearest 0 1 mg.
8.2.3 Remove the cap. and emptv the s\nnge contents
into the pretitrated pyndme vessel. Pull the plunger out am:
replace the cap. Titrate the specimen with KFR :o the erv
point descnbed in 8 1.5.
8.2.4 Reweigh the emptied synnge. and calculate tr
specimen weight by difference.
8.2.5 Calculation:
8.2.5.1 Calculate the percent water L as follows-
L = (P x F'x lOO)/5
8.3 Analysis of Materials With Less Than u : ~ Hater
8.3.1 For 0.1 to 0.5°^. foilow procedure in 82 (1-g
specimen), except substitute a 1-mL microburet for the
25-mL buret m the Karl Fischer apparatus.
8.3.2 For less than 0.1 ^o. use a 1-mL microburet ani
increase specimen size as much as needed, up to 10 g. 1
should be possible to measure moisture levels down to 1 pprr,
(0.0001 °l) by this approach.
NOTE—Specimens witti less than 0 1 "i *ater rna\ require special
handling tecnmques to prevent pickup of aimosphenc moisture The
precision of tnis test metnoa was determined wun specimens containing
nigher *ater levess.
9. Maintenance
9 1 Cleanup—Clean the utration vessel by nnsing with
fresh pyndme. Do not use methanol or other solvents.
9 2 Dn-ness—Check frequently to be sure that all drying
tubes, are in good condition and tightly connected. Replace
dessicant wnen indicator color changes througn naif of in*.'
moe
9.3 Electrode Performance—If electrode response is slug-
gish or otherwise off standard, take the following steps, in
turn, to correct the problem. Test the eiectrode with a
utration after each step, to determine if the next step is
required.
951 Wipe the electrode tip with a clean paper towel.
^32 Wash the electrode by dipping in concentrated
hydrochlonc acid for at least 1 mm. Rinse first with distilled
water, then with methanol.
933 Follow manufacturer's instructions on resetting end
point meter.
9 3.4 Replace power source. See manual for replacement
procedure.
935 Replace the electrode.
10. Precision and Bias
10.1 The precision estimates are based on an inter-
laboratory study 'in which one operator in each of sever.
different laboratones analyzed in duplicate, on two differen:
days, seven samples of water-based paints of \anous tvpes
containing between 25 to 75 cc water. The results were
analyzed statistically in accordance with Practice E lisO. The
wuhin-laboratorv coefficient of vanation was found to be
1 " ^ relative at 98 df. and the between-laborator. coeffi-
cient of vanation was 5 3 °c relative, at 42 df. Based on these
coefficients, the following cntena should be used for judging
the acceptability of results at the 95 ^ confidence levei
M) i 1 Repfiiiahiiu\—Two results, each tne mean of du-
plicate determinations, obtained bv the «ame .>oerator ^r
-------
D4017
d'.fferent aa\s sneuid be considered suspect i: tries citfer b\
nore tr.an - " r- relative.
KJ '. 1 •?t""v>i;;/i'//v — Two results, each the mean of
cuplicate determinations, obtained b> operators in different
laroratenes snomu be considered suspect if thev ditTer ps
more than i 5 n ~- relative.
!0.2 Bias — Bios has not been determined for :ms test
metncxl.
"• Inde* Terms
11.1 This test method is indexed under the following
terms: K.ari Fischer reagent method: moisture content, vvater
content (paints, related coatings).
"e American Society s>ceratton at a meeting c' me 'esoons'O'e
u' comments nave 101 'ece'vao a tair nearmg you snouia mane your
-c me *4S~M Corrrr'rree jn Stsrcarcs '9'c ^ace S'
-------
-------
Designation: D 3792 - 86
Standard Test Method for
Water Content of Water-Reducible Paints by Direct Injection
Into a Gas Chromatograph1
TVs uandard is issued under the ti\ed designation D '"••>; the numrrr immediately following the desisnauon indicates the '•car o:
onejrui adoption or. in the --ose ji vision ;ne ^ear 01 last revision A lumber in parentnews indicates tne vear of last rejppro'.u A
superscript epsilon i>i indicates an euuonai ..tun^e since '.he last re^'sion or reapproval
. Scope
1 1 This test method is for tne determination of the total
• ater content of water-reducible paints. It has been evalu-
.ted for latex s\ stems (styrene-butadiene. poKiMnvlacetatei-
.cnlic. acr%lici. It has not >et been evaluated for -other
vater-reducible paints but is beneved to be applicable The
•staDlished Corking range of this method is from 40 to 55 ^
vaier There is no reason to bene\e that it wni not work
outside of tms range.
1 2 i '"s siarujra ma\ :nvoi\e hazardous >naier:ais iTtr-
2i:ons ~>;a eautrment This s^nuara J.vi >vr purport to
uflfl'r .../ :V tr.e 'atetv pronerris ^ssoc'.atea ^nn :i\ :i\c !: :s
tne re*?onsiru r. m the ;w •" this ^.anaura '<> establish
appropriate sa'e'. Jnd neattn prac:ices and determine me
jppi!LjT:;::\ >•' re?uunor\- ::rnuauons prior ,v use For
specific nazara statements, see Section S
Referenced Documents
Z 1 -LSTU S^rMarat
D 1 19} Spec::"ication for Reagent v\ater
Dl36-i Test Method for \\ater :n Volatile Solvents
•F'.scner Reagent Titration Method)'
E ISO Practice for Determining Precision Data of ASTM
Methods for Analysis and Testing of industnai Chem-
3. Summarv of Test Method
3 '. A suitable aliauot of v^noie paint is internalK stan-
Jardizec uun an.T.drous 2-oropanoi. diluted witn dimeth\l-
iormamic;e. ar.c '.Men miected into a gas chromatograpnic
column ;ontamipg a porous polymer packing tnat separates
^ater from otner volatile components The water content is
determined from area calculations of the materials pro-
ducing peaks on tne chromatogram
4. Significance and Lse
-i ' V\;th tne need to calculate volatile organic content
iVCCi of uater-reducibie paints, it is necessan. to Knov. the
water content This gas cnromaiograpmc test method pro-
vides a relative!1, simple and direct \>.av to determine
content
- ^ 'en — e'noc s jnder :ne unsaictuT .M \^TM r.immittee D- >;r. P:int
ic Rfiatec ' 'Ji.-zs jnd Matenais ana •<, 'ne direct resoonsiPnr,'. 01 ^ucx.om-
'Ht;; f>'i . n C"tTicai Analysis M Pami^ jnu Pjmi Matenj i
^r-^.ni r^',,>r _-^'ri\c:0 Nov I1* ^^^ ^JTSPCd Jjnu.tP. -«" ' ""'^ £. r. Jl I >.
TABLE 1 Instrument Condition*
Dstectof
Temo«r»tur»s. 'C
Samow mwt
Detector
Column*
Initial
F-n*
Program rate
Gamer Gas
=iow rate mL/mm
Detector current
Soecimen size
triermai conauctivitv
200
240
80
170
30;rmn
neitum or nitrogen
50
150 mA
* cor isotnermai ooeration set tne column temoerature at 'iC°C After ;ne
Z-orooancH nas oeareo tne column adiust the temoerature to i'C°C -mn OMF
: ears Tie coiumn Reset tne temoerature to 140°C tor suoseaue"! ains
5. Apparatus
5 1 Oas Chromaiosraph—An> gas-iiauid chromato-
graphic instrument having a detector mav 'DC used. Temper-
ature programming capability is preferaole. but isothermal
operations may be adequate. See Table 1
5 2 Coiumn—The column should be •» ft (1 12 m i of • s-m.
13 I-mmi outside diameter tubing of stamiess steel, or other
iuitaoie material, packed with 60'80 mesn i ISO to 250 u.m>
oorous polymer packing material.* A reoiaceable glass sleeve.
glass wool plug, or other suitable material mav be placed on
r.ne entrance end of the column to retain anv nonvolatile
-natenais. This will minimize sludge buildup in tne column.
5 3 Recorder—A recording potentiometer with a full-scale
reflection of 1 to 10 mV. full-scale response f.rne of 2 s or
less and sufficient sensitivuv and staouitv to meet the
requirements of 5 I.
5 4 Liquid CHarziru* Devices—Micro svnnges of 10 or
2 5 -uL rapacity.
6. Column Conditioning
6 1 Procedure—The packed column is installed in the gas
chromatographic unit leaving the exit end disconnected from
me detector. This will prevent any contamination of the
detector with the coiumn bleed. Set the nelium or nitrogen
flow rate at 20 to 30 mL/min if a '?-m i3 2-mmi outside
diameter column is used Purge the column 5 or 10 mm
be to re heating. Heat the column from room temperature to
200°C at 5°C'min and hold this temperature tor at least 12 h
At the end of this time, heat ;r>e coiumn at
• PjrjpaK 0* 3 trademarn 01 Waters A5soc . Inc Mntora MA ^as oeen tound
:i-ia*' >r\ An^ other porous ponmer pacmng or otner . j'urr- i,-
JDC- -r cxMlurmancc md> "e used Trirse proouc:s -:re
- -"-i i^r-pn suppliers jnd d:stnnut
-------
D3792
:=C mm to 250°C (the maximum temperature for this
rackine) and hold for several hours. Cool the column to
room temperature and connect the column detector. Reheat
::ie column to 130'C at 5°C mm to observe it" there is
:oiumn oleed. Optimum conditioning of this column ma>
•.axe several cscles of the heating program before a good
recorder baseline is achieved.
*.2 Before each calibration and senes of determinations
•or daily) condition the column at 200'C for 1 h wuh earner
gas flow
". Reagents and Materials
I P;ir;rv of R^agcnts—Reagent.grade chemicals shall be
used in ail tests. Unless otherwise indicated, it is intended
that all reagents shall conform to the specifications of the
Committee on Analytical Reagents of the American Chem-
ical Societv where such specifications are available.5 Other
grades may be used, provided it is first ascertained that the
reagent is of sufficiently high purity to permit its use without
lessening the accuracy of the determination.
~2 P:int\- or ll'aier—Unless otherwise indicated, refer-
ence to water shall be understood to mean reagent water
conforming to T\pe II of Specification D 1193.
~ 5 Corner Gas—Helium of 99.995 ^o or higher punty
H;gh-punt\ nitrogen may also be used.
" 4 D:mein\titvrnamide iDWF) \Ann\drous} gas chroma-
tography. spectroohotometnc quality 'Note 11
5 2-Pr'^an<>i i An''.\artnni Mn D V an No^ir^na C^ " Nr» > jrk N> ind the "Lnitcd Suits
response factor to water relative to the standard is deter-
mined b> rneans of the following procedure. See Fig. 1 for a
typical chromatogram. It is good practice to determine the
relative re:ention time daily or with each senes of determi-
nations.
10.2.1 Weigh about O.I g of water and 0.2 g of 2-propano!
to 0.1 me into a septum sample vial. If it has beer
determined that a correction for the water content i<
necessar-. weigh 2 mL of dimethylformamide (DMF) into
the vial. If the DMF is anhydrous, simply add 2 mL of it as
weighing is not necessary.
10.2.2 Inject a 1-uL aliquot of the above solution into the
column and record the chromatogram. The retention order
and approximate retention times after the air peak are (h
water, about 0.7 mm: O 2-propanol. about 2.8 mm; and (J)
DMF. about 7 mm.
10.2.3 The preferred procedure to obtain the water con-
tent of the DMF is the Karl Fischer mration (Note 1». If this
has been determined, calculate the response factor for water
by means of the following equation:
1H,O
where:
R
II
H.O
' "H:0 T rr» j"V
response factor.
weight of 2-propanol,
weight of water added,
weight of dimethylformamide,
area of water peak.
area of 2-propanol peak, and
weight ^c water in DMF
10.2.
lowing
H56
4 If Karl Fischer utrauon
procedure may be used
is not available, the fol-
to obtain a reasonable
FIG. 1 Typical Chromatogram
-------
D3792
timate of the response factor:
10.2 4 I Inject the same size aliquot of DMF and 2-
jropanoi mixture, but \vitnout added water, as a blank Note
he area ot the \vater peak in the blank.
102-1 The response factor for water is calculated bv
neans of the following equation:
R =
wnere.
R
H.O
A,
B
= -esponse factor.
= weight of 2-propanol.
= weight of the water.
= area of 2-propanol peak.
= area of the water peak, and
= area of the water peak in the blank.
11. Procedure
1! i ^ eign to 0 1 me. 0 6 g of water-reducible paint (see
Note 2- ana 0 2 g of 2-propanol into a septum vial. Add
2-mL of DMF into the vial. Seal the vial. Prepare a blank
containing the 2-propanol and DMF but no paint.
N'OTE ^—Check eacn paint svstem to be analvzed for interfering
pea« Coalescing agents do not interfere with tnis determination
11 2 Shake the vials on a wnst action shaker or other
suitable cevice for 15 mm. To facilitate settling of solids
"How the vials to stand for 5 mm just pnor to injection into
ic cnrorr.atograpn. Low-speed centnfugation may also be
used.
! i 3 Inject a 1-uL sample of the supernatant from the
prepared solutions into the chromatographic column.
Record tne chromatograms using the conditions described in
Table 1
12. Calculations
12 ; Measure the area of the water peak and the 2-
propanoi internal standard peak and multiply each area by
'.he apprccnate attenuation factor to express the peak areas.
on a common basis. L'se of an electronic integrator is
recommenced to obtain tne best accuracy and precision.
However tnangulanon. plammeter. paper cut out. or ball
and disk integrator may be used.
122 Calculate the water concentration in the paint by
means of tne following equation:
H,0 - - X '00
-IH 0 x
where
-1 x It'
= aria of water peak.
= area of 2-propanoi peak.
II' = weignt of 2-propanol added.
IT, = weight of paint, and
R = response factor determined in 10 2
12.3 Correction lor H ater Content (» Sunc'U
12.3 1 If the blank indicates the presence of a detectable
peak for water in the dimethylformamide used as solvent.
make a correction in the calculation.
12.3.2 The water content of the dimethvlformarmde de-
termined by either chromatography (1024) or. preferably.
Karl Fischer titranon (10.2.3) is used to make the correction
Calculate the water content due to the solvent by using the
following equation:
H_,o,S^c-'-™^
w here:
Ji"5 = weight of dimethylformamide.
»', = weight of paint, and
P = weight ~c water in DMF
100
12.3 3 The water content of the oamt m this case is the
difference between the total percent determined in 12.2 and
the correction for the solvent water content as determined in
13. Precision and Bias6
13.1 The precision estimates are based on an inter-
laboratory study in which nine different laboratories ana-
lyzed in duplicate on two days four samples of water-
reducible paints containing from 40 to 55 °c H-O
(theoretical). The results obtained were analyzed statistically
in accordance with Practice E 180. The withm-laboratorv
coefficient of variation was found to be i.O ^ relative at 34
degrees of freedom and the between-laboratones coefficient
of variation 2.6 % relative at 30 degrees of freedom. Based on
these coefficients, the following criteria should be used for
judging the acceptability of results at the 95 ^c confidence
level.
13 1.1 Repeatabilin-—Two results, each the mean of du-
plicate detemmations. obtained bv the same operator on
different days should be considered suspect if they differ by
more than 2.9 ^ relative.
13 1.2 Reproducibilin—Two results, each the mean of
duplicate determinations, obtained by operators in different
laboratories should be considered suspect if-thev differ by
more than 75^ relative.
13.2 Bias—Bias has not been determined.
' Supporting data are
'"rom ASTV1 Headauaners Request RR D"i -
*ie American Society lor Testing ana Materials taxes no rxsition resoecting tne validity ot any patent 'igrtts asserted in connection
Mitn any item mentioned in tnis stannara users ot mis stanoara are exoressiy aavisea mat determination ot me validity ot any sucn
patent ngnts ana tne 'is* ot infringement ot sucn ngnts are entirely tneir own resoonsiomty
'"is standard 'S suDiect to revision at any rims oy me resoonsioie tecnmcal commmet ana must oe reviewed every live years ana
' iot -tvtsea eitner reaoorovea or #irnarawn four comments are mvitea eitner lor revision ot tnis stanaara or tor adartionai stanaaras
in? rnouia oe aaaressea 'O /»STW neaoooarrers 'our comments will receive careful consiaeration at a meeting ot me resoonsioie
•eci"cai commires *mci sou n?ay attena it /ou 'eei mat your comments nave not 'eceivea a tair Bearing fou snouid mate yo^r
. ews mown :o me *STM Committee on S:anaaras '9'6fface5f Pi/iaae/onia P* '9'C3
-------
-------
APPENDLX ii
43
-------
-------
Soybean Oil Volatility Test Results
Test Date: December 5, 1990
Analysis: Gravimetric portion of Method 24
Equipment: Mettler PM 100 Balance, Fisher Forced Draft Oven
Temperature: 234 °F, 112.2 °C
Analyst: Rima Dishakjian
Prepared for: Madeleine Strum, CPB, MD-13
Food Club Soybean Oil
Sample
Number
1
2
3
Percent
Volatile
0%
0.1%
0.1%
Wesson Soybean Oil
Sample
Number
1
2
3
Percent
Volatile
0%
-0.1%
-0.3%
It would be safe to assume that the soybean oils tested are not volatile
according to the Method 24 definition. Any positive or negative percent
volatile value is probably due to variability in the balance used to weigh the
samples and is not real.
-------
-------
APPENDIX iii
44
-------
Ul
-4
.•^
0
0)
1— 1
A
0
*J
4)
o>
4)
>
*4
0
U)
—H
tn
>^
>-4
«
C
i4{
«^
*4>
PM
TJ
0
AJ
4)
X
C
4)
>
O
jj
'-u
fl
'M
Q
TJ
41
O
U
0
b.
a
s
4)
4J
O
U)
M
U
4>
£.
OJ
— t
Cfa
«
41
U
C
iH IQ
en ^^
X <8
<*i a
^ CJ
\o
MOO
r-t (N
O ^
4J X
cu •
^ Ch
en CD
\^4 CO
.. .. u
« *> 3
-U C 4J
« o n
o a u
a 41
•a •* a
n 3 a
4) O* 4)
6- Cd fr«
n
rH
1
a
J£
•
a
04
u
«
a
3
u
4J
cn
c
a , ig
,r^ a
IQ 4)
C U
< a>
,
•^ 3
fi
&« Cd
~ {J
CJ Z
M <
2 U
£.4 a;
s M
W f.-j
O <-*
£.. >_
5 M
Cd O
2
~
O. 1-1
O
__^
>» 5>
N4 ^^
Z 2£
Cd <
0.
C
0
0
J
o
o
Ou
£4
in o
(N VO
r* O
•* CM O CN n ^
o n o o o O
(N £4 <*i CJ
o
en Q
i
a a
i i
IX &
CN
4
O
00 ^O *tf
o o o
tn eo tn
T> tn eo
m m ^
•«r ts r-
O CN m
^ »H O
4 M
in ^ r-
m m •
• • n
n n
<~> (N n
•«r a
-»T* nr»n M-*^
* /^ i^ /»< « »^
«„(>, n^^ co«reo
^4^f^ t^^O O <*"! O
• ^H * • ^4 rH • • •
,-( ^
o»^fN n^m vo^eo
•O ' J T3
J 0> M 4)
M J -^ O "*
O^J " ' !^^
P^ •** -^
O -H Q -*
rf T3 Cd TJ
J O-^ Cd-t
O cd ^ cn -i
Z Cd O Z 0
< w u M ia
CJ Z iJ
M
O ^4 •• Cd ••
1 4) Q 4)
da a jj 3 -u
1 .10 ES 0
« EX Z CJ Z
•a
4)
c
M-I
I
-------
o
(U
A
.u
4)
0
O
-~*
ta
I
*J
0)
c
Q]
>
O
(0
u
Q
H)
O
CU
4)
4J
O
a
U
V
ta
•H
b.
y
H IQ
cn ^
I
a
z
03
a.
o
H
4J
en
Pt 03
^ O
O O O .0)
-4 P«~* C
0 i-t 1-1-^
>-4 Ch JS O
0\ 4) CQ T3
\F^ 00 -^ (0
\ 9) (0
ess a ••
.. ~* U
O OS • O
• • •• ^4 *W
O 4J 3
-U C 4J •• 'O
ifl OJ 10 -U OJ
Q a u tn w
a oj >«
jj -H o,^ a
a 3 a to oj
® tr v c u
&< Cd &« < 0,
ww
II
go
6- Ed
U Z
W <4S
cd r
x u
= *
y ]~
M >£
cd j
Cb
o ~
6* -2
s
U J
Ed O
-•"•N
4* CT*
^
5 _i
a. M
o
^^
> 2
ft.
•*•
Ziv
%
CU <<
04
§
0
J
HI
o
h.
O
Cd
a,
£
if) m
n o
i
vo vo o rr o -» 00 vO 00 00
•» p» »H if) n •*
in in if) P"> T •»
r» m r» 1-1 oo if)
PI n o ^ <*i (N
^ * o o •* *
ON O ••* Pvl rt ^*
r4 Cd
O-" 03
Ui >
•< -o o
M Cfl
Z
o ** w
Z 0 Q
A •U 3
Cd O CS
> 2 CJ
c-l
0
O
o o n
O i-4 >H
O O O
1
r^P^ vo
p* ^ If)
pi • r-
• n •
rt rt
C*" ^P C4
VO O M
o • r>
• P4 •
M P>l
^ H
i-^
(0
Jj
U)
u
J U
M
O
U O
Z I*
3 0
£1 Z
MO
o
3SS
O 0 0
00 P» »
•v m oo
o • .
n n
00 If) VO
fM O PI
O ^-t O
• • •
(M
* If) •
eft co *r
n ^ PI
^ M n
<-> pn n
J
H
O
z
i4i
Cd
03
^4
O
03
a
i
03
1
a
TJ N
TJ OJ -H
coo
•H fl TJ
-------
-------
APPENDIX iv
45
-------
-------
FRYING OPERATIONS REFERENCES
Kawada, T., R.G. Krishnamurthy, B.D. Mookherjee, and S.S. Chang.
Chemical Reactions Involved in the Deep Fat Frying of Foods II.
dentification of Acidic Volatile Decomposition Products of Corn
oil. JAOCS. Vol. 44. February, 1967.
Krishnamurthy, R.G., and S.S. Chang. Chemical Reactions Involved
in the Deep Fat Frying of Foods III. Identification of Nonacidic
Volatile Decomposition Products of Corn Oil. JAOCS. Vol. 44.
February, 1967.
Keijbets, M.J., G. Ebbenhorst-Seller, and J. Ruisch. Deep-Fat
Finish-Frying of French Fries in Unhydrogenated Refined Soybean
Oil. Fette. Seifen. Anstrichmettel. 1986.
Eskin, N.A., et al. Stability of Low Linolenic Acid Canola Oil
to Frying Temperatures. JAOCS. Vol. 66. August, 1989.
Zeddelmann, H. The Spoilage of High - Quality Baked Products by
Inappropriate Choice of Fats and Processing Methods. CCB Review
for Chocolate, Confectionery and Bakery 2(1): 23-26, 1977.
Miller, L.A., and P.J. White. High-Temperature Stabilities of
Low-Linolenate, High-Stearate and-Common Soybean Oils. JOACS
65(8): 1324-1327, August, 1988.
Taha, F.S., H.E. Helmy, and A.S. El-Nockrashy. Changes in
Cottonseed Oil When Used for Frying Vegetable Products Containing
Chlorophyll. JOACS 65(2): 267-271, February, 1988.
Khattab, A.M., et al. Stability of Peroxidised Oils and Fat to
High Temperature Heating. Journal of the Science of Food and
Agriculture 25(6): 689-696, 1974.
White, P.J., and Y.C. Wang. A High-Performance Size-Exclusion
Chromatographic Method for Evaluating Heated Oils. JOACS 63(7):
914-920, 1986.
cao.068
fry_opt.ref
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on the revene before completing/
REPORT NO.
EPA-450/3-91-011
3. RECIPIENT'S ACCESSION NO
TITLE AND SUBTITLE
The Impact of Declaring Soybean Oil Exempt -from VOC
Regulations on the Coatings Program
5 REPORT DATE
April 1991
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
Madeleine L. Strum, EPA
Candace Blackley, Radian
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND AOORESS
Emission Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency (MD-L3)
Research Triangle Park, NC 27711
10. PROGRAM ELtMENT NO.
11 CONTRACT/GRANT NO
68-02-4378
2. SPONSORING AGENCY NAME AND AOORESS
Emission Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14 SPONSORING AGENCY CODE
5. SUPPLEMENTARY NOTES
6. ABSTRACT
This document presents the findings of a study to evaluate the impact of declaring
soybean and other vegetable seed oils exempt from VOC regulation on the coatings
program. The physical and chemical characteristics of 10 vegetable seed oils are
tabulated and their uses are discussed. Tests conducted with EPA reference Methods
24 and 24a showed no weight loss, indicating that the oils contain no VOC. However,
the study discloses that VOC's are emitted during the autoxidation reaction which
occurs when these oils are in contact with atmospheric oxygen.
-------
------- |