Journal of the Air Pollution Control Association

ISSN: 0002-2470 (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/uawm16

Odor Threshold Determinations of 53 Odorant

Chemicals

Gregory Leonardos , David Kendall & Nancy Barnard

To cite this article: Gregory Leonardos , David Kendall & Nancy Barnard (1969) Odor Threshold
Determinations of 53 Odorant Chemicals, Journal of the Air Pollution Control Association, 19:2,
91-95, DOI: 10.1080/00022470.1969.10466465

To link to this article: https://doi.org/10.1080/00022470.1969.10466465

Published online: 16 Mar 2012.

Submit your article to this journal

Article views: 43324

View related articles

Citing articles: 30 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uawm20
 Gregory Leonardos, David Odor Threshold Determinations
Kendall and Nancy Barnard
Arthur D. Little, Inc. Of 53 Odorant Chemicals
In order to assist in assessing potential odor problems arising from chemical manufacturing operations, the
odor thresholds of 53 commercially important odorant chemicals have been determined using a standardized
and defined procedure. The odor threshold data previously available have shown wide variation reflect-
ing the diversity of procedures and techniques used. Factors that may affect the odor threshold measure-
ment include the mode of presentation of the stimulus to the observer, the influence of extraneous odorants
in the presentation system, the type of observer used, the definition of the odor response, the treatment of
the data obtained, and the chemical purity of the odorant. The experimental approach used has mini-
mized these variations. The odoranfs were presented to a trained odor panel in a static air system utiliz-
ing a low odor background air as the dilution medium. The odor threshold is defined as the first concen-
tration at which all panel members can recognize the odor. The effect of chemical purity has been deter-
mined by measuring the odor threshold of materials representing different modes of manufacture or after
purification by gas chromatographic procedures. The threshold concentrations range over six orders of
magnitude. Trimethylamine exhibited the lowest threshold (0.00021 ppm volume); methylene chloride
was not recognizable below 214 ppm. Of the 53 chemicals, sulfur bearing compounds exhibit low thres-
hold values on the order of parts per billion. Aside from the sulfides, it is not possible to anticipate the
odor threshold of a material based on its chemical structure or functionality.

Odor thresholds for 53 commercially
important odorant chemicals have been
determined under controlled laboratory
conditions to permit comparison of their
absolute odor strengths. The odor
threshold data previously available
have shown wide variation reflecting
the diversity of procedures and tech-
niques used. Factors that may affect
the odor threshold measurement include
the mode of presentation of the stimulus
to the observer, the influence of extrane-
ous odorants in the presentation system,
the type of observer used, the definition
of the odor response, the treatment of
the data obtained, and the chemical
purity of the odorant. The experi-
mental approach used has minimized
these variations. Laboratory deter-
mination of recognition threshold con-
centrations is only one type of data re-
quired to establish meaningful air
quality criteria.
In 1967 the Manufacturing Chemists'
Association asked Arthur D. Little,
Inc., to determine the odor thresholds of
53 odorant chemicals using a standard-
ized and defined procedure to permit a
comparison of their relative odor im-
portance.
Odor thresholds in the past have been
determined by a variety of methods.
This may in part account for the varia-
tion in the reported data.1 Factors that
Mr. Leonardos, Mr. Kendall and
Miss Barnard are Staff Chemists
with Arthur D. Little, Incorporated,
Acorn Park, Cambridge, Massachu-
setts 02140.
affect the odor threshold measurement
include the mode of presentation of the
stimulus to the observer, the influence
of extraneous odorants in the presenta-
tion system, the type of observer used,
the definition of the odor response, the
treatment of the data obtained, and the
chemical purity of the odorant.
The experimental procedures used in
determining the odor thresholds include
an odor test room offixedvolume (static
system) and low background odor for
presenting the stimulus to the observer.
Trained odor analysts, who are capable
of describing the character and intensity
of odor, were used throughout the study.
The odor response was defined rigor-
ously by requiring each panel member
to describe the odor quality. By
minimizing the variation in procedure,
sample presentation, and panelist, it is
possible to use a smaller number of re-
sponses. The odor threshold has been
defined as the first concentration at
which all the panel members have been
•able to recognize the odor sensation
that is characteristic of the chemical
and have been consistent in their re-
sponse at all higher test concentrations,
even when these are presented in a ran-
dom fashion to minimize sample order
bias.
The experimental approach, then,
was to determine the odor thresholds of
a wide variety of chemicals of com-
mercial importance using the standard-
ized and defined procedure. The odor
thresholds determined in this study are
internally consistent and on this basis
should find wide use in assessing poten-
tial odor problems.
The experimental procedure used
minimizes variation in the factors affect-
ing the odor threshold by utilizing the
best available technology and proce-
dures for measuring odor. We recog-
nize the effect of chemical purity on the
odor threshold and have studied this, on
a limited basis, by evaluating the odor
threshold of the same chemical produced
from different starting materials,
method of production, or the addition of
stabilizers. Selected chemicals were
purified by gas liquid chromatographic
procedures to note the effect of chemical
purity of the material on the odor
threshold.
Odor and the Odor Threshold
Before describing the techniques used
and the results obtained, it is important
to have an understanding of the odor
sense, what it is, how and what is mea-
sured, and the significance of the odor
threshold. At the present time or in
the foreseeable future, there are no
instruments available for measuring
odor other than the human and his nose.
The human nose has a well-deserved
reputation as a detector with high sensi-
tivity to certain chemicals at extremely
low concentrations. At the present
time, instruments are available with
sufficient sensitivity to detect concen-
trations of some odorant chemicals (such
as methanol) that are below the odor
threshold. An odor measuring instru-
ment must not only parallel the sensi-
tivity of the human nose but, more
important its response must parallel the
response of the biological system.
Odor can be defined as sensations re-

February 1969 Volume 19, No. 2 91

suiting from the interaction of volatile
chemical species inhaled through the
nose, making contact with the olfactory
area and registering in the brain. Not
all chemical compounds impinging on
the olfactorium produce the odor sensa-
tion. Odor may be considered as hav-
ing at least two parameters—quality
and intensity. Both of these pa-
rameters are of importance in the odor
threshold measurement. The qualita-
tive aspect of the odor sensation may be
described in associative terms (odor
character notes) or broadly as hedonics,
i.e., "like-dislike." The lay observer
has a tendency to give a "like-dislike"
response. Through training, experi-
ence, and attitude, the odor analyst will
describe the odor sensation in terms of
familiar qualitative experiences like
fatty, sour, sweet, which are termed
odor character notes.
Food flavors which are largely aromas
are comprised of complex mixtures of
hundreds of chemicals which act as
stimuli to the human sensor}^ organs.
Some of the chemicals are directly
identifiable, but most are either buried
in a complex blend or are read by the
human system in combination as a char-
acter note. These principles have
evolved into a method of measuring
flavor and odor—The Flavor Profile
Method2—that has been successfully
used for solving flavor and odor prob-
lems for some 25 years. Odor char-
acter notes may be attributed to the
odor of a single chemical or an inte-
grated response to a mixture of chem-
icals. Reference to these chemicals or
mixtures of chemicals allows for the de-
velopment of a common terminology in
describing odor. "Single" chemical
compounds may be described with one
or more character notes, depending on
the nature of the chemical and the pres-
ence of impurities. The use of trained
analysts who describe the odor response
allows for increased confidence in assess-
ing data obtained for determining odor
thresholds. In most threshold studies,
emphasis has been placed on the in-
tensity of the odor response and not
enough on the quality of the odor re-
sponse.
The second parameter of the odor re-
sponse is the intensity (strength of re-
sponse). Although there is consider-
able argument, intensity is generally
accepted to be a logarithmic function of
the concentration of the chemical species
(as expressed by the Weber-Fechner
Law). The Weber-Fechner Law has
found wide application in the senses of
sound and sight and appears to be
applicable to the sense of smell. As
with all biological systems, the response
to odorants is not elicited at all concen-
trations. Threshold is defined as the
degree of stimulation of a nerve or nerve
92
center that just produces a response.
Concentrations of odorant chemical be-
low this point will be judged not odor-
ous. The concept of threshold, to be
effective, requires a rigorously defined
response. In the odor literature, there
are essentially two thresholds. Some
investigators have defined threshold as
the point where there is a detectable
difference from background. Another
definition of threshold is the minimum
identifiable odor (MIO) or recognition
threshold. Both may be properly con-
sidered odor thresholds. However, it
has been our experience that the detec-
tion threshold (first change from back-
ground) is less reliable and difficult to
reproduce as it relies on a poorly defined
judgment of the observer. Often there
is confusion as to the definition of the
first change as well as changes that can
occur in the background odor. The
recognition threshold, on the other
hand, utilizing trained observers, al-
lows for a rigorous definition of the odor
response. The recognition odor thresh-
old can be defined as the first concentra-
tion at which an observer can positively
identify the odor quality of the odorant
chemical and is consistent in his re-
sponse at all higher concentration levels.
Some experimental work3 carried out
previously at ADL suggests, that for a
limited number of odorants over a
range of concentrations, it is possible to
distinguish by triangulation a difference
in intensity produced by an increase of
concentration of approximately 60 per-
cent. Thus, there are 25-30 just-notice-
able differences (JND). The JND is de-
fined as the minimum change in con-
centration which can be recognized
statistically as a difference in odor in-
tensity. In practice, most organoleptic
methods of flavor and odor analysis rely
on a seven to nine point intensity scale.
The threshold intensity is regarded as a
fixed concentration with other intensity
ratings representing a range or spread of
concentrations. Based on our observa-
tions in this work, it is probable that the
odor threshold for an individual observer
for a chemical may show some variation.
However, it is not likely that such
variation will be significantly greater
than an order of magnitude. In this
study, concentrations evaluated occur
in equal logarithmic steps somewhat
greater than two JND's.
The detection threshold has been re-
lied upon to a great extent in the past.
This has been defined as that concentra-
tion of odorant which produces the first
change in odor over background. It is
possible that much of the literature data
although reported as detection thresh-
olds are in reality recognition thresholds
due to misinterpretation of the odor
response. It has been our experience
that the detection threshold is not re-
Journal of the Air Pollution Control Association
liable nor reproducible. It may be
affected by subtle changes in back-
ground odor, and by its very nature, is a
poorly defined odor response. Interest
on the part of odor research workers has
switched to the recognition threshold
concept which affords a better definition
of the odor change. The statement,
"We feel comfortable with the known,"
could be extended to "and the recogniz-
able" when applied to measuring odor
thresholds.
The odor thresholds determined in
this study are derived under near-ideal
conditions of low background odor. It
is not known what the effect on the odor
threshold is as a result of the interaction
of other odor contributing materials in
the environment. In the real situation,
a high background odor partly through
the mechanism of adaptation becomes
the reference point against which
changes are perceived.
It is likely that the odor thresholds
determined in this study may be lower
than those determined under environ-
mental background conditions in view
of their being determined under near-
ideal conditions. However, we would
like to point out the possible influence
on odor threshold of varying back-
ground odors to which the lay popula-
tion is subjected. The human reacts to
change, and odor problems occur as a
discernible change. At the present
time, the degree of this discernible
change is not known. This study
should give some indication as to the
extent of potential odor problems as a
result of chemical manufacturing opera-
tions.
Experimental
The chemicals obtained represent the
highest purity commercially available
from large scale production. Addi-
tional samples of carbon tetrachloride
and toluene representing different modes
of production or starting material were
evaluated. Samples of styrene with
and without inhibitor were evaluated to
note any effect on the odor threshold of
the inhibitor. In addition, two samples
of hydrogen sulfide were evaluated.
Test Room
The odor test room offers the conve-
nience of presenting the odor stimulus
simultaneously to a group of analysts
in a low odor background. The test
room consists of two chambers — an
antechamber and the test room — with
supporting equipment such as fans,
ducts, and activated carbon filters for
removing odorized air and introducing a
controllable "odor-free background"
that serves as a dilution system.4 The
test room and antechamber are lined
with polished aluminum that has been
 shown to be relatively nonabsorptive to
odorant materials. Influent air that is
used to purge the room is passed through
a bank of activated carbon filters, cir-
culated through the antechamber and
test room by five electric fans, and ex-
pelled through an exhaust port in the
ceiling. The test room has a volume of
approximately 500 cubic feet (13,200
liters).
The antechamber serves as a buffer
between the test room and the external
environment and serves to acclimate
the panel members to the relatively
odor-free conditions prior to each test
exposure.
The chemical to be studied is intro-
duced quantitatively at the appropriate
concentration level into the test room by
a variety of methods depending on its
physical properties. Three electric fans
circulate the odorant for five minutes,
after which time the fans are shut down
and the air in the test room is allowed to
come to rest. The panel members enter
the test room from the antechamber and
make their individual odor observations
describing the odor character and the
intensity of each odor character note
that may be present. Upon completion
of the analysis, the test room and ante-
chamber are purged with the carbon-
treated air for twenty minutes. In
only a few instances was the twenty-
minute purge period inadequate for
odor removal. In these cases, purging
was continued until the room was odor
free or a complete wash-down of the
room was required.
Odorization of the Test Room
The chemicals evaluated included
liquids, gases, and solids. The air in
the test room (volume of 13,200 liters)
served as the dilution medium. Liq-
uids were injected into the test room
using Hamilton (Gas Tight) micro-
syringes. The syringes can accurately
deliver as little as 0.1 ul of material and
are used extensively in gas chromatog-
raphy. To obtain the desired concen-
tration of odorant chemical in the test
room, the "Conversion Tables for Gases
and Vapors" (Handbook of Physics and
Chemistry, 44th edition, 1962, p. 2288)
were used to convert from ppm volume
to milligrams of chemical per liter of air.
The total milligrams of material re-
quired for the test room were calculated
by multiplying the mg/liter by test
room volume. The specific gravity as
supplied by the manufacturer was then
used to calculate the volume of liquid
odorant chemical to be injected into the
test room. As little as 0.1 jul of liquid
odorant could be directly injected into
the test room. Liquids whose threshold
concentration was such that an injection
of less than 0.1 n\ was required were
diluted in an appropriate solvent. The
February 1969 Volume 19, No. 2
solvents used were other liquid mate-
rials whose odor properties had been
studied within the program. Usually
ethanol or mcthanol were used. The
dilutions were made in such a manner
that the amount of solvent present in
the test room was considerably below
threshold.
Solid materials used to odorize the
test room were introduced in two ways —•
injection of a solution containing the
dissolved solid or by volatilizing a
weighed amount with a flow of hot air.
As an example, phenol was dissolved in
odor-free water, and the appropriate
volume of the phenol solution was in-
jected into the test room. Other solids
were weighed into a suitable receptacle
which was placed in the test room, and
volatilized by a flow of warm air pro-
vided by a hot air fan. Usually five
minutes was sufficient to volatilize the
solid material.
Gases were diluted with odor-free air
in a dilution apparatus outside the test
room. The apparatus consisted of
three glass dilution bulbs which could
be evacuated to as low as 0.01 mm Hg
by a mercury diffusion pump coupled
to a liquid nitrogen trap. The three
glass bulbs were used to prepare inter-
mediate dilutions of the gas. A 1.3
liter glass bulb was attached to the
apparatus, evacuated, and then charged
with a precisely calculated pressure
from one of the three dilution bulbs.
The charged 1.3 liter bomb was then
brought to atmospheric pressure with
odor-free air. The contents of the 1.3
liter bomb were then swept into the test
room with compressed air.
Odor Panel
A trained panel of four staff members
of the Food and Flavor Section of
Arthur D. Little, Inc. was used for
determining the odor threshold of each
chemical. Panel members were selected
from a pool of approximately fifteen
observers with more than one year of
analytical odor work. Only one chem-
ical was observed per day. Prior to the
test room observations, the panel
examined the odor over water at various
dilutions to become acquainted with the
odor type and to develop a common
terminology for describing the odor.
Each chemical was examined at a
minimum offivedifferent concentrations
in the odor test room. The first odor
observation of the day for the odor
panel was a determination of the back-
ground level of the test room. These
sessions were included to ensure that the
room was not contaminated and also to
acquaint the panel members with any
background odors that are inherent to
the room. The background odor was
usually described as very slightly musty.
Each panel member was required to
be present for all the odor examinations
scheduled for the day, as a different con-
centration was evaluated at each ses-
sion. The panel members were not in-
formed of the concentration of the
chemical in the test room. A concen-
tration range was selected that was
thought to include the threshold con-
centration. The order of presentation
of concentrations was on a random
basis. The panel members did not
make continuous observations; each
observation was separated by a mini-
mum of 25 minutes which was the time
necessary to purge the room and intro-
duce a new concentration of material
into the test room. Olfactory fatigue
is thus no factor in the observations.
The concentration levels examined
for each chemical are increased in equal
logarithmic steps. It is presumed that
odor is a logarithmic function of con-
centration. The concentrations ex-
amined were multiples by ten of 1, 2.1,
and 4.6 ppm (by volume). The large
incremental steps in concentration were
selected so as to make the odor changes
more evident and minimize the effects of
individual variation.
The results obtained from each pan-
elist were examined separately. A posi-
tive response is indicated for each con-
centration at which the panelist de-
scribed the odor of the chemical. As
the concentration level was varied at
random, the threshold concentration
for each subject was determined by
noting the positive responses as a func-
tion of the concentration. The thresh-
old was taken as the lowest concentra-
tion at which the panelist could define
the odor and that which could be con-
sistently recognized at higher concen-
trations.
The threshold concentration for each
chemical is compiled in this manner for
each panel member. The odor thresh-
old reported is the concentration at
which all four panel members could
positively recognize the odor of the
chemical.
Three chemicals were selected for
study to note the effects of the purity of
the chemical on the odor threshold.
Toluene, styrene, and tolylene diiso-
cyanate were purified on a gas chro-
matographic column (20' x 3/8" 50%
SE-30 on 80/100 Chromport XXX,
preparative) under isothermal condi-
tions. The chromatographic column
effects a high plate efficiency distilla-
tion. Nitrogen was used as a carrier
gas at a flow rate of 110 ml per minute.
A number of repeated injections and
collections were made to amass sufficient
quantities of purified material for the
odor test room study.
Results
The odor threshold reported for each
chemical is the concentration at which
93
Table I. Odor thresholds in air (ppm volume). all panelists recognize the odor. All
11
concentrations are calculated as parts
. Chemical Odor Threshold Odor Description per million by volume. The minimum
0.21 Green sweet,
threshold observed was with trimethyl
Acetaldehyde
Acetic acid 1.0 Sour amine at 0.00021 ppm (0.21 ppb).
Acetone 100.0 Chemical sweet, pungent Methylene chloride was not described
Acrolein 0.21 Burnt sweet, pungent below a concentration of 214 ppm. The
Acrylonitrile 21.4 Onion-garlic-pungency 53 compounds studied represent a range
Allyl chloride 0.47 Garlic-onion pungency.'green of six orders of magnitude in threshold
Amine, dimethyl 0.047 Fishy concentrations (Table I). Odor de-
Amine, monomethyl 0.021 Fishy, pungent scriptions (other than chemical name)
Amine, trimethyl 0.00021 Fishy, pungent used to describe the odor quality of the
Ammonia 46.8 Pungent
odorant chemical are included in Table
Aniline 1.0 Pungent
4.68 Solvent
I.
Benzene
Benzyl chloride 0.047 Solvent Sulfur Containing Compounds
Benzyl sulfide 0.0021 Sulfidy
As a group, compounds with the sulfur
Bromine 0.047 Bleach, pungent
0.001 Sour
atom in their structure have the lowest
Butyric acid
Carbon disulfide 0.21 Vegetable sulfide
thresholds of the compounds evaluated
Carbon tetrachloride (chlorination of CS2) 21.4 Sweet, pungent (Table II). All the sulfides with the
Carbon tetrachloride (chlorination of CH4) 100.0 exception of carbon disulfide and sulfur
Chloral 0.047 Sweet dioxide have threshold concentrations at
Chlorine 0.314 Bleach, pungent the parts per billion level.
Dimethylacetamide 46.8 Amine, burnt, oily Hydrogen sulfide gas from a cylinder
Dimethylformamide 100.0 Fishy, pungent exhibited a lower threshold than hy-
Dimethyl sulfide 0.001 Vegetable sulfide drogen sulfide gas produced by acidify-
Diphenyl ether (perfume grade) 0.1
Burnt rubbery
ing sodium sulfide. The discrepancy in
Diphenyl sulfide 0.0047
Ethanol (synthetic) 10.0 Sweet
the determined odor threshold for the
Ethyl acrylate 0.00047 Hot plastic, earthy two hydrogen sulfide samples may be
Ethyl mercaptan 0.001 Earthy, sulfidy due to impurities that are present in the
Formaldehyde 1.0 Hay/straw-like, pungent samples used; further studies are being
Hydrochloric acid gas 10.0 Pungent pursued to determine the factors ac-
Hydrogen sulfide (from Na2S) 0.0047 Eggy sulfide counting for this discrepancy between
Hydrogen sulfide gas 0.00047 samples. It is of interest that carbon
Methanol 100.0 Sweet disulfide, sulfur dichloride, and sulfur
Methyl chloride (above 10 ppm) dioxide affect the other chemical senses
Methylene chloride 214.0 (taste and feel) as well as having odorous
Methyl ethyl ketone 10.0 Sweet
Methyl isobutyl ketone 0.47 Sweet
effects.
Methyl mercaptan 0.0021 Sulfidy, pungent Nitrogen-Containing Compounds
Methyl methacrylate 0.21 Pungent, sulfidy Of the 53 odorant chemicals eval-
Monochlorobenzene 0.21 Chlorinated, moth'balls
Nitrobenzene
uated, trimethyl amine exhibited the
0.0047 Shoe polish, pungent
Paracresol 0.001 Tar-like, pungent
lowest threshold (0.00021 ppm). Di-
Paraxylene 0.47 Sweet methyl formamide is recognizable at
Perchloroethylene 4.68 Chlorinated solvent 100 ppm. The nitrogen-containing
Phenol 0.047 Medicinal compounds indicate the wide range of
Phosgene 1.0 Hay-like threshold concentrations that can occur
Phosphine 0.021 Oniony, mustard (Table III).
Pyridine 0.021 Burnt, pungent, diamine
Styrene (inhibited) 0.1 Solventy, rubbery Oxygenated Compounds
Styrene (uninhibited) 0.047 Solventy, rubbery, plasticy Table IV lists the odor thresholds of
Sulfur dichloride 0.001 Sulfidy oxygenated compounds according to
Sulfur dioxide 0.47 chemical class. Extent of oxidation of
Toluene (from coke) 4.68 Floral, pungent, solventy the ethanol series (ethanol, acetalde-
Toluene (from petroleum) 2.14 Moth balls, rubbery hyde, acetic acid) does not appear to
Tolylene diisocyanate 2.14 Medicated bandage, pungent
Trichloroethylene
have an affect on the odor threshold
21.4 Solventy
trend. One might expect a lower
a threshold as the oxidation state is in-
Other than chemical name
creased. Chloral (trichlorinated analog
of acetaldehyde) does have a substan-
tially lower threshold (0.047 ppm) than
Table II. Odor thresholds — sulfur Table III. Odor thresholds —nitrogenous
acetaldehyde. Considering other chlo-
pounds (ppm volume). compounds (ppm volume). rine-containing compounds studied, it is
not possible to make a generalization as
Hydrogen sulfide (cylinder) 0.00047 Trimethyl amine 0.00021 to the effect of chlorination on the odor
Dimethyl sulfide 0.0010 Nitrobenzene 0.0047 threshold.
Ethyl mercaptan 0.0010 Monomethyl amine 0.021
Sulfur dichloride 0.0010 Pyridine 0.021 Unsaturated Compounds
Benzyl sulfide 0.0021 Dimethyl amine 0.047 The presence of unsaturation in an
Methyl mercaptan 0.0021 Aniline 1.0 odorant chemical is not associated with
Hydrogen sulfide (Na2S) 0.0047 Acrylonitrile 21.4
Diphenyl sulfide 0.0047 Ammonia 46.8
low threshold concentrations (Table V).
Carbon disulfide 0.21 Dimethyl acetamide 46.8
Ethyl acrylate and methyl methacrylate
Sulfur dioxide 0.47 Dimethyl formamide 100.0 are isomeric, however, the threshold
concentrations are quite different. The

94 Journal of the Air Pollution Control Association

Table IV. Odor thresholds — oxygenated odor descriptions do differ (see Table I).
compounds (ppm volume). odor response provide consistent and
This disparity points up the difficulty of internally comparable data for one
Carbonyls making extended generalizations per- parameter of odor. These recognition
Chloral 0.047 taining to the odor threshold based on threshold concentrations demonstrate
Acetaldehyde 0.21 similar chemical structures. the broad range in sensitivity of human
Acrolein 0.21 olfaction to various chemical structures
Methyl isobutyl 1ketone 0.47 Benzenoid Compounds
under ideal test conditions.
Phosgene 1.0 The effect of substitution on the It should be noted that, in addition
Formaldehyde 1.0 benzene ring also produces a wide varia-
Methyl ethyl ketone 10.0 to the threshold concentration for a
tion in odor thresholds Table VI. given chemical compound, there are at
Acetone 100.0
A single methylation of the benzene least three sensory attributes and two
Esters ring (toluene) lowers the threshold one
Ethyl acrylate 0.00047 psychological parameters to be con-
concentration step. Further methyla- sidered in evaluating its contribution to
Methyl methacrylate 0.21
tion (p-xylene) succeeds in lowering the odor problems. The three sensory
Carboxylic Acids
odor threshold by an order of magnitude. attributes are: (1) the change in intensity
Butyric acid 0.001
Acetic acid 1.0
The —CH2CI group (benzyl chloride) (odor strength) with concentration;
produces an odor threshold that is 100 (2) the qualitative character of the
Alcohols
p-Cresol
times lower than benzene. The only odor; and (3) the type and degree of
0.001
Phenol 0.047
generalization possible with benzenoid- interaction with other odorants. Psy-
Ethanol 10.0 type materials is that substitution on chologically, people react either posi-
Methanol 100.0 the ring reduces the odor threshold by tively or negatively to odor types —
as much as a thousandfold depending on they like or dislike them. A problem is
the nature of the group added. usually associated with those that
Purity of the Chemical people dislike. Secondly, people react
Table V. Odor thresholds —variously Odor thresholds of samples purified more strongly to things that are differ-
substituted ethylenic compounds ( p p m
by gas liquid chromatography show ent and, therefore, what is normal or
volume).
slight differences, if any, from the start- expected is frequently accepted be it
Odor ing material. With tolylene diiso- good or bad.
Thresh- cyanate, the prime effect of the purifica- The work discussed in this paper is a
Chemical R Group(s) olds
tion appears to eliminate or reduce sub- first step in providing basic informa-
Ethyl acrylate -C:OOC2H6 0.00047 sidiary odor character notes and leaves tion on a number of chemicals which
Styrene —CeHs 0.047 a sharper odor impression of the prime are industrially important. The recog-
Acrolein —HC:O 0.21 odor character note. In this case, nition odor thresholds reported in this
Methyl meth- —CHs, 0.21 purification did not affect the recogni- study were developed under ideal lab-
acrylate —C:OOCH 3 oratory conditions and are not recom-
Allyl chloride —CH2CI
tion threshold. Table VII lists the
0.47
Tetrachloroethylene —CI(4X) 4.68
odor thresholds of chemicals before and mended for air quality criteria and
Trichloroethylene —CI(3X) 21.4 after purification. standards since no effort was made to
Acrylonitrile —C=N 21.4 Toluene (from coke) purified chroma- define the degree of objectionability of
tographically has a slighly lower odor the odorant chemicals. Future phases
threshold; styrene, however, on purifica- of this program will attempt to provide
tion showed a higher odor threshold. data on the other aspects of the problem,
Samples of carbon tetrachloride and such as the method of sample presenta-
Table VI. Odor thresholds'—variously
substituted benzenoid compounds (ppm toluene representing different modes of tion, the use of untrained observers, the
volume). manufacture do show variation in the change in intensity (and quality) with
odor threshold concentration (Table I). concentration, and the odor interactions
Odor with typical impurities.
Thresh- Carbon tetrachloride manufactured by
Chemical R Group(s) old the chlorination of carbon disulfide has a Studies are presently being conducted
lower threshold (21.4 ppm) than carbon to determine and compare the odor
Paracresol —OH, —CH 3 0.0010 tetrachloride produced by the chlorina- thresholds of selected chemicals ob-
Nitrobenzene —NO2 0.0047
Phenol —OH 0.0470
tion of methane (100 ppm). Toluene tained from various methods of odor
Benzyl Chloride —CH2CI 0.0470
produced from coke has a slightly higher presentation. These include the syringe
Styrene —CH:CH 2 0.0470 threshold concentration (4.68 ppm) dilution method and a dynamic (odor
Monochlorobenzene —Cl 0.21 than toluene produced from petroleum hood) method of odor presentation.
p-Xylene —CHs, —CH 3 0.47 (2.14 ppm). The odor character notes In addition, the effect of varying types
Aniline —NH 2 1.0 (Table I) used to describe the two sam- and levels of background odor on the
Toluene —CH 3 2.14 ples reflect the difference in mode of recognition odor threshold are being
Benzene 4.68 manufacture. The purity of the chem- evaluated.
ical can affect the odor quality as well as
the odor threshold. Thus it is difficult References
to predict with any certainty the role 1. Adams, E. M., "Air Pollution Abate-
Table VII. Odor thresholds of chemicals that impurities present in the compound ment Manual—Physiological Effects,"
purified by GLC. may play in affecting the odor threshold. Gosline, ed., 22-26, Ch. 5, Manual
Sheet p. 6, Manufacturing Chemists'
t o l y l e n e diisocyanate 2.14 ppm Conclusions
Assoc, Washington (1951)
Tolylene diisocyanate* 2.14 ppm
2. Sjostrom, L. B., Cairncross, S. E., and
Recognition odor thresholds have Caul, J. F., "Methodology of the flavor
Toluene (coke) 4.68 ppm
been determined in the laboratory by an profile," Food Tech. XI, 20, (1957)
Toluene (coke) a 2.14 ppm 3. Hainer, R. M., Emslie, A. G., and
Toluene (petroleum) 2.14 ppm
expert panel for 53 commercially im- Jacobson, A., "An information theory
portant chemicals. By standardizing of olfaction," Annals N.Y. Acad. of
Styrene (inhibited) 0.10 ppm
the method of sample presentation and Sci. 58, 158-173 (1954)
Styrene (inhibited) 3 0.21 ppm 4. Deininger, N., and McKinley, R. W.,
Styrene (uninhibited) 0.047 ppm minimizing extraneous sensory inter-
"The design, construction, and use of an
ference, the concentrations determined odor test room," ASTM, Sp. Tech.
«• Purified by GLC proceedures as producing the minimum identifiable Publ. No. 164 (1954)';

February 1969 Volume 19, No. 2 95