Citation
Development of a technique for the measurement of methanol in the ambient atmosphere

Material Information

Title:
Development of a technique for the measurement of methanol in the ambient atmosphere
Creator:
Palomares, Arturo
Publication Date:
Language:
English
Physical Description:
xii, 75 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Methanol -- Measurement ( lcsh )
Atmospheric chemistry ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Chemistry
Statement of Responsibility:
by Arturo Palomares.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
25394757 ( OCLC )
ocm25394757
Classification:
LD1190.L46 1991m .P34 ( lcc )

Full Text
DEVELOPMENT OF A TECHNIQUE FOR THE
MEASUREMENT OF METHANOL IN THE AMBIENT
ATMOSPHERE
by
Arturo Palomares
B.A., University of Colorado, 1984
A,,thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Chemistry
1 991


This thesis for the Master of Science
degree by
Arturo Palomares
has been approved for the
Department of Chemistry
by

Date


Palomares, Arturo (M.S., Chemistry)
Development of a Technique for the Measurement of
Methanol in the Ambient Atmosphere
Thesis directed by Associate Professors Larry G.
Anderson and John A. Lanning
. In the presence of nitrogen dioxide, which is found
in exhaust gases and in the ambient air, alcohols give
rise to the formation of alkyl nitrite. The alkyl
nitrite plays a very important role in the formation of
photochemical smog as it accelerates ozone formation by
producing the OH radical. The OH radical may in turn
react with methanol, which leads to the formation of
formaldehyde.
The role,and behavior of alcohols in photochemical
smog formation have not been studied, to date, due to
the unavailability of methods for their determination in
the ambient air.
Methanol vapors were found to be readily derivatized
on a silica gel-cartridge with 3,5-dinitrobenzoy1
chloride.
The benzoate derivative formed is measured by
reverse phase high performance liquid chromatography
with ultra-violet detection at 224 nm. It was


IV
determined that the derivatization took place during the
first fifteen minutes of reaction time, at a sampling
rate of 0.5 1/min.
Moisture effects were studied to determine if
water would interfere with the derivatization
process. It was found that water did not appear to
play a significant role. Derivatization of the alcohol
took place even in the presence of large amounts of
water (sampling vapors over a 99.9% water/methanol
solution).
Breakthrough was studied and found to be occurring
during the early experiments. This was attributed to
the large concentrations of methanol that were being
sampled.
The author reports here an acetonitrile based
solvent system for the derivatization of methanol
designed to accommodate the low levels of methanol
expected to be found in the ambient air.
Presently there are no reliable, simple,
sensitive, and selective techniques for the routine
measurements of alcohols in the ambient air. A technique
that is able to determine the levels of alcohols in the
ambient air in the presence of large amounts of other
substances is needed.


V
The form and content of this abstract are
approved. We recommend its publication.


DEDICATION
I dedicate this thesis to my wife Teresa for all her
help and support.
I also dedicate this thesis to Mr. Danny Martinez
for giving me a chance when everybody else said no.
r


CONTENTS
CHAPTER
I. INTRODUCTION ........................... 1
Research Objectives .................. 4
II. BACKGROUND AND THEORY .................. 7
Approach to the Problem .............. 11
III. EXPERIMENTAL SECTION .................... 13
Instrumentation ....................... 13
Experimental Apparatus .............. 14
Preparation of Phenylisocyanate
Standard ............................ 16
Preparation of Methanol Standard ...... 16
Preparation of Mixed Methanol-
Isocyanate Standard ................... 17
Re-crystallization of Salicylic
Acid .................................. 17
Preparation of Methyl Salicylate ...... 18
Preparation of 3,5-dinitrobenzoyl-
chloride Standard ..................... 18
Preparation of Methyl-3,5-
dinitrobenzoate ....................... 19
Preparation of 2% Sodium Carbonate ... 20
IV. RESULTS AND DISCUSSION .................. 21
Salicylic Acid Experiments
27


VIII
3,5-Dinitrobenzoy1 Chloride
Experiments ............................ 32
Laboratory Sampling Studies ........... 42
Ambient Air Studies ................... 52
V. CONCLUSIONS ................... 67
Recommendations for Future Studies ... 69
REFERENCES ..................................... 7 4


FIGURES
Figures
3.1 Silica gel cartridge ..................... 15
4.1 U.V., absorption spectra of
phenylisocyanate ......................... 22
4.2 U.V., absorption spectra of
phenylisocyanate-methanol derivative .... 23
4.3 U.V., absorption spectra of
phenylisocyanate-water mixture .......... 24
4.4 Total ion chromatogram and
mass spectra of phenylisocyanate/
methanol mixture ........................ 26
4.5 HPLC chromatograms of in-house synthesized
methylsalicylate and commercially
prepared methylsalicylate ............... 29
4.6 HPLC chromatograms of methylsalicylate
standard, sampled vapors over pure
methanol ................................ 30
4.7 HPLC chromatograms of 3,5-dinitrobenzoy1
chloride and methyl-3,5-dinitrobenzoate
derivative .............................. 36
4.8 Total ion chromatogram and mass spectra
of in-house synthesized methyl-3,5-
dinitrobenzoate ......................... 38
4.9 Total ion chromatogram and mass spectra
of commercially prepared methyl-3,5-dinitro-
benzoate ............................... 39
4.10 HPLC chromatograms of coated blank
cartridge, benzoate standard, derivatizing
agent, methanol vapors .................. 41


X
4.11 HPLC chromatograms of benzoate standard,
benzoyl chloride standard, vapors over
pure methanol ........................... 44
4.12 HPLC chromatograms of 21 jag/mL 3,5-
dinitrobenzoyl chloride standard and same
standard with 1.1 mL water added ........ 4 9
4.13 HPLC chromatograms of 18.2 jjg/mL methyl-3,5-
dinitrobenzoate standard and ambient air
sample at 50X dilution .................. 53
4.14 Gas chromatograms of benzoate standard and
ambient air sample ...................... 5 5
4.15 Gas chromatograms of laboratory blank
and field blank cartridges .............. 56
4.16 Gas chromatograms of ambient air sample
at 10X dilution and benzoate standard
at 100 ng/_pl ........................... 58
4.17 Gas chromatograms of laboratory blank
and field blank cartridges .............. 59
4.18 HPLC chromatograms of ambient air
sample and methyl-3,5-dinitrobenzoate
standard ................................ 62
4.19 HPLC chromatograms of laboratory blank
cartridge and field blank cartridge
63


XI
TABLES
Table
4.1. Concentrations of Vapor's over Pure
Methanol ................................. 43
4.2 . Results of Breakthrough Experiments ..... 45
4.3. Results of Breakthrough. Experiments .... 47
4.4. Moisture Effect Results ................. 50
4.5. Vapors over Pure Methanol ................ 5 1


XII
ACKNOWLEDGMENTS
I would like to express my appreciation to Professor
Larry G. Anderson for his support and guidance throughout
the course of this work. I would also like to thank Dr.
John A. Lanning for his advice and helpful discussions.
Special thanks to Dr. John N. Gillis for the use of
the mass spectrometer. Thanks also to Mr. Dean F. Hill
for encouraging me to pursue a Master of Science degree.
Finally, I would especially like to express my
appreciation to Teresa, Sergio and Emilio for their love
and support during my studies.


CHAPTER I
INTRODUCTION
Alternative fuels will play an increasingly
important role as more and more cities look for ways to
combat their air pollution problems'. Presently, 101
U.S., cities have failed to meet the National Ambient
Air Quality Standard for ozone (1). Replacement of
gasoline and diesel powered vehicles with alternative
fuel powered vehicles is one approach that has been
suggested as a means of reducing atmospheric ozone
concentrations. One of the alternative fuels which is
expected to play a major role is methanol. Presently,
methanol is being formulated as M85, which is composed
of 85% methanol and 15% gasoline, and as M100, which is
composed of 100% methanol.
As the number of vehicles using this type of fuel
increases, it is reasonable to suggest that the
concentrations of alcohols in the atmosphere will
likewise increase. The increase in the alcohol
concentrations could be attributed to evaporative
losses, to unburned fuel in the exhaust, and to losses


2
during the fueling process.
, Methods which have been used for the determination
of formaldehyde have been modified to determine methanol,
in solution (2-7). Most of these methods determine the
alcohol concentration colorimetrically. Typically, the
alcohol is oxidized to formaldehyde and the resulting
colored dye is determined by ultra-violet
spectrophotometry. Bhatt and Gupta (2) oxidized methanol
with acidic potassium permanganate. The excess oxidant
was removed with hydrogen peroxide. The alcohol was
identified by the color reaction with oxalydihydrazide
and copper(II) in an acetate medium. Jephcott (3) also
oxidized methanol in the same manner, however, he used
oxalic acid to remove the excess oxidant before finally
developing a color with Schiff's reagent (4). Edvin (5)
used chromotropic acid for the determination of oxidized
methanol. Barns and Speicher (6) used phenylhydrazine
and Schryver's (7) method for the determination of
methanol.
Verma and Gupta (8) developed a procedure for
determining methanol in air. The procedure involved
oxidizing the alcohol to formaldehyde with acidic
potassium permanganate. The formaldehyde was determined
in an acidic medium with an p-aminoazobenzene and sulphur
dioxide reagent system by spectrophotometry. Maeda et
al. (9) determined trace amounts of alcohols in the


3
ambient air by using a heterogeneous reaction of the
alcohol with nitrogen dioxide. The alkyl nitrite formed
was detected on the surface of Pyrex glass by electron
capture gas chromatography.
The major disadvantage of each of these methods is
that they are cumbersome and somewhat complex. Another
disadvantage is that other compounds can be similarly
oxidized to formaldehyde. For example, Glycerols and
glycols both form formaldehyde on oxidation.
A technique which has been used to measure alcohol
emissions is the bubbler method (10). This technique
takes advantage of the high solubility of methanol and
low solubility of hydrocarbons in water. The technique
simply involves bubbling the air sample through water,
which traps the alcohol. The aqueous solution is
injected onto a gas chromatograph equipped with a flame
ionization detector.
The disadvantages of this approach are: (A)
hydrocarbons, which may be absorbed in the aqueous
solution during the bubbling process, could be counted as
the alcohol (methanol) by the method; and (B) the
sensitivity of the flame ionization detector is adversely
affected by the presence of large amounts of water.
Long path fourier transform infra-red spectroscopy
has been used to measure the concentrations of methanol
in the ambient air (11). The measurements were made with


4
an instrument that was equipped with a cell having a 23
meter long base path and a 1260 meter optical path.
Obviously, this technique uses very expensive and
delicate components and thus cannot be used for routine
measurements, since it would be very difficult to
transport from sampling site to sampling site.
Research Objectives
The primary goal of this research was to develop a
technique for measuring the concentrations of alcohols
(methanol) in the ambient air by reverse phase high
performance liquid chromatography utilizing an
ultraviolet detector (HPLC-UV). The technique would need
to be simple, selective, sensitive, and reproducible.
. This approach did not appear to be easily
accomplished because methanol absorbs at wavelengths
below 190 nm.
The following provides a description of the
constraints that had to be observed during development of
the technique.
In order to measure and detect methanol and/or other
primary alcohols by HPLC-UV, at wavelengths above 190 nm,
the alcohol would need to be derivatized and the
derivatization would need to take place on a coated
silica gel cartridge. This would require that enough


5
derivatizing agent be adsorbed on the silica gel
cartridge to be available for derivatization of the
alcohol(s).
The derivatizing agent would need to be selective
toward alcohols. Therefore, the ideal derivatizing agent
should not derivatize any of the'other species which are
expected to be found in the atmosphere, such as aldehydes
and ketones.
Water should not have a significant impact on the
derivatizing agent nor should it significantly affect the
ability of the derivatizing agent to derivatize the
alcohol.
While many organic reactions require the application
of thermal energy or the use of a catalyst to obtain the
desired products, the proposed technique would require
that the reaction take place at ambient environmental
temperatures and in the absence of a catalyst.
No clean-up agents could be used since these could
have a deleterious effect on the derivatizing agent or on
the derivative itself.
To make this an attractive and portable technique,
sample preparation would need to be kept down to a
minimum.
The technique for derivatizing the alcohols would
need to be compatible with ultra-violet reverse phase
high performance liquid chromatography.


6
The instrumental analysis time would need to be kept
down to a minimum to allow the greatest number of samples
to be analyzed in the shortest amount of time.
High yields of the derivatized alcohol would be a
requirement, however, yields of less than 100% would be
acceptable as long as the technique is reproducible.
The restrictions imposed on the design of the
technique by the constraints mentioned above are
mandatory in a system to be used for routine ambient
measurements of alcohols. Any scheme for the measurement
of alcohols (methanol and/or ethanol) in the atmosphere
carried out under the conditions described above should
be attractive for routine air monitoring of alcohols.


CHAPTER II
BACKGROUND AND THEORY
Initial consideration of the objectives of this
research focused on the question of which derivatizing
agent or agents could be suitably used for the
derivatization of methanol. By derivatizing the alcohol
greater sensitivity and detection above 190 nm could be
obtained.
Classical derivatization techniques have been used
for fluorescence and ultra-violet absorption analysis in
solution. Derivatization techniques have also been used
in nuclear magnetic resonance (NMR) studies and in gas
chromatographic analysis.
Typically, a large molar excess of the derivatizing
agent is used to ensure high derivative yields. There
are three types of derivatization techniques: pre-column
derivatization, on-column derivatization and post-column
derivatization. In pre-column derivatization the
derivative is formed before it is introduced onto the
chromatographic column. On-column derivatization, mostly
used in gas chromatography, requires that the


8
derivatization take place on the chromatographic column
interest is derivatized after it has eluted from the
column but before passing into the detector.
Our studies focused on pre-column derivatization.
Pre-column derivatization was chosen over the other two
derivatization processes due to the previously mentioned
constraints placed on the technique.
Several derivatizing agents have been used to
derivatize alcohols. All of these studies, however, haye
been carried out in solution.
No investigations, prior to this one, have been
carried out to determine if alcohol vapors could be
adequately derivatized on a silica gel cartridge.
Phenylisocyanate has been used to form
carbanilates when reacted with aliphatic alcohols and
carbohydrates in a dry pyridine solvent system (12).
Phenylisocyanate has been used to derivatize 2-
ethoxyhexanol and other aliphatic alcohols (13). The
substituted urethane is formed when the isocyanate is
i
treated with the alcohol. The reaction scheme is as
follows (13):
This approach to derivatizing the alcohol appeared
to be a good candidate since neither thermal energy nor
itseif. In post-column derivatization the substance of


Sf
the presence of a catalyst were required. However,
phenylisocyanate reacts vigorously with water in a ratio
of 2:1. The isocyanate readily undergoes hydrolysis to
form as a product an N-substituted carbamic acid which is
quite unstable and breaks down to form an amine and
carbon dioxide. The presence of large amounts of water
in the atmosphere, therefore, make this reaction scheme
undesirable. Also, the proposed technique requires that
one of the components of the mobile phase be water; which
could react with phenylisocyanate during the
chromatographic process.
Alcohols have been derivatized to form benzoates
(14), p-nitrobenzoates (15), p-methoxybenzoates (16), and
3,5,-dinitrobenzoates derivatives (17, 18). Salicylic -
acid and methanol have been combined to form methyl
salicylate (19). The methyl salicylate is prepared by
the esterification of the carboxylic acid group of
salicylic acid with methanol. The reaction is as
+ n,o
The esterification of acids with alcohols, can only
be accomplished if there is a means available to drive
the equilibrium to the right. There are several ways in
which this can be accomplished. These include: having an


10
excess of one of the reactants, removal of the products
(water or the ester) by distillation, removal of the
water by azeotropic distillation and removal of the water
by using a molecular sieve or by adding a dehydrating
agent. Also, typically, this type of reaction requires
the use of a catalyst, such as sulfuric acid.
The disadvantages of this approach are : A) water
would always be present in large amounts in the
atmosphere and, therefore, it would be improbable that
one could always maintain an excess of one of the
reactants; B) removal of the water or ester by
distillation would not be possible for the proposed
technique; C) neither molecular sieves nor dehydrating
agents could conveniently remove the' large amounts of
water present in the ambient air and D) water is one of
the products of the reaction and therefore the
equilibrium, according to Le Chatlier's principle, would
be shifted to the reactants side of the reaction leading
to little or no ester formation. Therefore, it would
appear that this scheme would not be suitable for
derivatizing alcohols present in the atmosphere.
3,5-dinitrobenzoyl chloride appears to be an
excellent candidate for derivatizing alcohols. The
reaction is as follows:
+ ROH
2n
o i
02n
0
t + HC!
'OR


11
The reaction between 3,5-dinitrobenzoy1 chloride and
methanol leads to the formation of methyl-3,5-
dinitrobenzoate, which can be readily monitored by
reverse phase high performance liquid chromatography with
U.V., detection at 224 mn.
A possible disadvantage of this approach is that
3,5-dinitrobenzoy1 chloride readily reacts with water to
form 3,5-dinitrobenzoic acid, by the following process:
Therefore, both, water and methanol compete for 3,5-
dinitrobenzoyl chloride. Thus, the benzoyl chloride
would need to be present in excess &t all times.
The success of this approach is dependent on the
ability of methanol to consume the derivatizing agent at
a significantly greater rate than water.
Based on the previous information on the
derivati.zation of alcohols and the' constraints imposed by
this specific problem, a method using 3,5-dinitrobenzoyl
chloride adsorbed on a silica gel carttidge was devised.
During the course of this research several
different derivatizing agents were investigated and
Approach to the Froblem


12
several different concentrations of the coating solution
adsorbed on the silica gel cartridge were also
investigated.
The resulting derivatives were measured by high-
performance liquid chromatography employing an ultra-
violet detector.


CHAPTER III
EXPERIMENTAL SECTION
LJ.V., spectra were obtained with a Perkin-Elmer
Hitachi 200 Spectrophotometer equipped with a
Perkin-Elmer Hitachi 200 recorder. And with a Perkin-
Elmer 552A UV/VIS Spectrophotometer equipped with a
Perkin Elmer 561 recorder.
The Hitachi 200 Spectrophotometer was set to scan
from 370 to 190 nm at a rate of 30 nm/min., and set with
a slit width of 3.0 nm.
The 552A Spectrophotometer was set to scan from 450
nm to 190 nm at rate of 120 nm/min., and set with a slit
width of 2.0 nm.
High-performance liquid chromatography was
performed using a Varian 5000 Liquid Chromatograph
equipped with a Hewlett-Packard 3388A integrator and a
Varian UV-50 detector. The mobile phase composition is
listed with each chromatogram. The flow rate was set at
1.0 mL/min for all of the studies. The analytical columns
used were a C18 jaBondapak, 25 cm long and a Water's RP18,
25 cm long.


14
Gas chromatography/ mass spectrometry was performed
using a Hewlett-Packard 5890 Gas Chromatograph equipped
with a 5970 series Mass Selective Detector. Mass spectra
for the phenylisocyanate methyl ester were obtained by
scanning mass to charge ratios from 50-400. The electron
impact ionizing voltage was 70 eV.
The gas-chromatography/mass spectrometry conditions
are described in each figure or in the text.
Experimental Apparatus
A sequential air sampler, built by Dr. Larry G.
Anderson of the University of Colorado at Denver, was
used to obtain all of the air samples.
The air flow rate was measured using a Precision
Scientific Wet Test Meter.
The air pump was a II5 volt, 4.2 amp pump made by
GAST MFG Corporation.
The silica gel cartridges, which were polyethylene
syringe barrels, Figure 3.1, were prepared by placing a
polypropylene porous disk at the base of the cartridge
and adding 0.75 grams of 100-200 mesh silica gel. A
second porous disk was placed on top of the silica gel.
The silica gel cartridge was rinsed with 10 mL of
acetonitrile. The cartridges were then coated with 10 mL
of the derivatizing agent. The coating solution was


15
Figure 3.
^ Silica gel
described
cartridge preparati
m the text.
on
as


16
drained through the cartridges by gravity. Once coated
the cartridges were dried overnight with helium in an
oven at 56 C (Blue M Electric Company of Blue Island,
Illinois).
Preparation of Phenylisocyanate Standard
All reagents were used as received unless otherwise
specified. The phenylisocyanate 98%+ (Aldrich) standard
was prepared by taking l mL of phenylisocyanate (density
= 1.096 g/mL) and pipetting it into a 100 mL volumetric
flask and diluting to mark with acetonitrile. This
solution served as the stock standard solution. Various
concentrations were prepared from this stock standard by
performing the necessary dilutions.
Preparation of Methanol Standard
The methyl alcohol used was anhydrous grade received
from Mallinckrodt. The stock methanol standard was
prepared by pipetting 0.13 mL of methanol, density 0.791
g/tnL, into a 100 mL volumetric flask and diluting to
the mark with acetonitrile. This standard served as the
stock solution. Various concentrations of methanol
standards were prepared from this stock standard
solution.


17
Preparation of Mixed Methanol/Isocyanate Standard
t
A mixed standard was prepared by taking 1 mL of a
500 ^ag/mL phenylisocyanate standard and mixing it with 1
mL of a. 500 ug/mL methanol standard solution.
Re-crystallization of Salicylic Acid
The salicylic acid was obtained from Fisher
Scientific Company. 20.6 g of salicylic acid were
weighed into a 250 mL Erlenmeyer flask and refluxed.
Acetone was added slowly maintaining a gentle reflux of
the solvent in the flask. The solvent was added until no
more of the salicylic acid dissolved. The solution was
brought to boiling while stirring rapidly, to avoid
bumping. The hot solution was allowed to cool on the
laboratory benchtop. Once cooled the solution was placed
in an ice bath. The crystals were collected by suction
filtration using a Buchner funnel and were washed with
generous volumes of cold solvent. The melting point of
the recrystallized salicylic acid was determined to be
158 C to 159.1 C. The literature melting point of
salicylic acid is listed as 157 C to 159 C (20). The
melting point was obtained by using a Thomas Hoover


18
Unimelt, Capillary Melting Point Apparatus, Arthur H.
Thomas Company, Philadelphia Pa.
Preparation of Methyl Salicylate
1.00747 grams of the recrystallized salicylic acid
were weighed into a test tube and 10 mL of methanol were
added which was acidified by adding five drops of
concentrated sulfuric acid. The test tube was placed in
a hot water bath maintained at a temperature of 70 C for
15 minutes. All of the salicylic acid dissolved. A
strong minty, odor was sensed.
Preparation of 3,5-dinitrobenzoyl chloride
Standard
3,5-dinitrobenzoyl chloride (Kodak) was
recrystallized by taking 3.12 grams and weighing into a
500 mL Erlenmeyer flask. To the flask were added 300 mL
petroleum ether. A funnel was placed at the mouth of the
flask and the solution was refluxed for several minutes.
The 3,5-dinitrobenzoyl chloride did not completely
dissolve. Methylene chloride was added to the refluxing
solution one milliliter at a time until all of the 3,5-
dinitrobenzoy1 chloride dissolved. The total volume of
methylene chloride added was 65 mL. This solution was


19
boiled for approximately 38 minutes. The solution was
allowed to cool on the laboratory bench top. After a few
minutes, the solution was placed in an ice bath. The
solution immediately became very cloudy and small
crystals were observed to be forming. Approximately five
minutes after placing the solution in the ice bath large
amounts of crystals had been formed. The crystals were
collected by suction filtration using a Buchner funnel.
The crystals were still a bit yellowish in color even
after generous washings with cold solvent. The melting
point was obtained and determined to have a range of 64
to 65 C. The literature melting point of 3,5-
dinitrobenzoyl chloride is listed as 74 C (21). The
approximately 10 C difference between the literature
value and the value obtained by this researcher was
attributed to a faulty thermometer. The thermometer was
not calibrated.
Preparation of Methyl-3,5-dlnitrobenzoate
The methyl-3,5-dinitrobenzoate was prepared as
described by Shriner et al. (22). The melting point was
obtained and determined to have a range of 101 C to
102 C. The literature melting point for methyl-3,5-
dinitrobenzoate is listed as 112 C (21). The
approximately 10 C difference between the literature


value and the value obtained by this researcher was
attributed to a faulty thermometer.
Preparation of 2% Sodium Carbonate
About 0.20 grams of sodium carbonate were weighed
into a graduated cylinder and dissolved in 10 mL of
water.


CHAPTER IV
RESULTS AND DISCUSSION
The first compound investigated for derivatizing
methanol was phenylisocyanate. Figure 4.1 shows the U.V.,
spectra of a 1 ,pg/mL phenylisocyanate standard solution.
The figure clearly illustrates that phenylisocyante has
two absorption bands one at approximately 226 nm and one
at approximately 200 nm. Figure 4.2 is the U.V., spectra
obtained after mixing equal volumes of a 1 jjg/mL solution
of phenylisocyanate and a 1 ^pg/mL solution of methanol.
Clearly the two spectra are very different indicating
that an interaction between the isocyanate and the
alcohol took place. The product formed'appears to have
its maximum absorption at about 214 nm. Figure 4.3 is
the U.V., spectra obtained after mixing equal volumes of
a 1 jjg/rnL phenylisocyanate solution and a 1 jug/mL water
solution. This spectra is very different from the
spectra in Figures 4.1 and 4.2. The spectra clearly
shows that the product of the interaction between the
isocyanate and water has a maximum absorption at about
230 nm and another less intense absorption band at


iFigure 4.1. U.V., absorption spectra of phenylisocyanate.
The absorption spectra was obtained by using
the following conditions: Scan from 370 nm j
to 185 nm, slit: 2.3 nm., scan speed was set
at 30 nm/min, and the source was a deuterium
lamp.


23
37'
J 33c) 29c) 25cT
WAVELENGTH (nm)
210
170
Figure 4.2. U.V., absorption spectra of phenylisocyanate
-methanol mixture.
The spectra was obtained by using the
instrumental conditions described in figure
4.1.


24
WAVELENGTH Cnm)
Figure 4.3. U.V., absorption spectra of phenylisocyanate
-water mixture.
The spectra was obtained by using the
instrumental conditions described in figure
t


25
approximately 260 nm. Therefore, it would appear that
phenylisocyanate interacts with both methanol and water,
as evidenced by the very different U.V., spectra obtained
for each of the different species.
A mass spectrum (A) of a mixed solution containing
500 jag/mL of phenyisocyanate and 500 jjg/mL of methanol
was obtained and is illustrated in Figure 4.4. The mass
spectra and total ion chromatogram are shown. A mass
spectral library search was performed to determine the
identification of the peak with a retention time of 6.22
minutes. The mass spectra library match of 96.37%
indicated that the peak was that of the phenyl methyl
ester. Indicating that derivatization of the alcohol had
occurred. The figure also shows the standard mass
spectrum (B).
The gas chromatography-mass spectrometry conditions
used to obtain the mass spectrum in Figure 4.4 were as
follows: the oven temperature was set and held at 80 C
for 1 minute and increased at a rate of 10 C/minute up
to a temperature of 260 C. A 1 jjL splitless injection
was used. The mass spectrometer was scanned from m/z 50
to 400. The column used was an XL dimethylsilicone with a
film thickness of 0.33 jam., and a length of 12.5 meters
with an internal diameter of 0.2 mm.
This researcher was unable to obtain an HPLC
chromatogram of the derivatized methanol. A possible


Detector Response Abundance
26
M/Z
A. Time ( min )
B
M/Z
Figure 4.4. Total Ion chromatogram and mass spectra of
phenylisocyanate/methanol mixture.
The instrumental conditions are described
in the text.


27
explanation is that the methyl ester was hydrolyzed to
the acid once the methyl ester derivative was introduced
into the HPLC mobile phase, which was composed of 50%
water/50% acetonitrile. Several attempts were made to
chromatograph the methanol derivative with no success.
This compound was not investigated further as a possible
derivatizing agent. It is important to note that
Bjorkqvist et al. (13) were able to derivatize several
aliphatic alcohols in solution by utilizing a DMF solvent
system and a 50% acetonitrile and 50% water mobile phase
system. However, Bjorkqvist et al. used an excess of the
phenylisocyanate to derivatize the alcohols. The unused
excess reagent was destroyed by using an amine or an
alcohol. The excess isocyanate needs to be destroyed
because it reacts vigorously with water. This researcher
did not conduct any studies to determine if methanol
vapors could be derivatized on a silica gel cartridge
with phenylisocyanate.
Salicylic Acid Experiments
The second compound investigated as a possible
derivatizing agent was salicylic acid. Initial
investigations consisted of taking 1 gram of salicylic
acid and adding 5 mL of methanol and three or four drops
of concentrated sulfuric acid in a test tube. The test


28
tube was placed in a water bath maintained at 70 C for
15 minutes. The methylsalicylate was formed almost
immediately as evidenced by a strong minty odor. In
order to determine if the salicylate had been formed by
the process described above, the synthesized
methylsalicylate was compared, by HPLC, against the
commercially prepared methylsalicylate. The
chromatograms are shown in Figure 4.5. The retention
time of the commercially prepared methylsalicylate (B)
was determined to be 1.40 minutes. The retention time of
the methylsalicylate prepared in our lab (A) had a
retention time of 1.39 minutes. Strongly indicating that
methylsalicylate had been formed by the above noted
process.
Salicylic acid appeared, therefore, to be an
excellent candidate for derivatizing methanol. Silica
gel cartridges were prepared and coated with a 1000 jjg/mL
solution of salicylic acid. The cartridges were dried
overnight with helium at a temperature of 56 C. After
sampling, the cartridges were removed from the sampling
apparatus and eluted with 5 mL of acetonitrile. Figure
4.6 shows the chromatogram obtained by sampling the
vapors over pure methanol for two hours (B) at a rate of
1 L/min., and the chromatogram of the methylsalicylate
standard (A). A very small peak at 1.39 minutes matches
the retention time of the methylsalicylate standard.


29
iu
tn
:z
o
CL
0C
TIME (min)
Figure 4.5. HPLC chromatogram.*; of in-house synthesized
methylsalicylate (A) and commercially
prepared methylsalicylate.
The instrumental conditions were: the
mobile phase was. acetonitrile/water (70:30),
wavelength 306 nm, flow 1 mL/min.,
absorbance range 1 AUFS, and the column was
a Cl 8 ^lBondapak.


DETECTOR RESPONSE
30
TIME (min)
Figure 4.6. HPLC chromatograms of methylsalicylate
standard (A), sampled vapors over pure
methanol (B) and (C).
The instrumental conditions were: mobile
phase was acetonitrile/water (60:40),
wavelength 306 nm, flow 1 mL/min.,
absorbance range 1 AUFS, and the column was
a CIS jaBondapak.


31
Thus, it appeared that the methanol was being derivatized
but not to a quantitative extent. It was decided that
perhaps a longer sampling time and a faster sampling rate
would increase the derivative yields. Figure 4.6 also
shows the chromatogram obtained after sampling the vapors
over pure methanol for 20 hours (C) at a sampling rate of
3 liters/minute. No peak matching the retention time of
fnethylsalicylate was observed.
Thus, it would appear that a longer sampling period
and a faster sampling rate did not increase the
derivative yield. It appears that the kinetics of the
reaction are slow at room temperature. Other
explanations for the lack of quantitative derivative
formation on the silica gel cartridge are: no heat was
applied to the sampling system and the methyl salicylate
was hydrolyzed during the longer sampling times. Since,
the objective of this research was to derivatize alcohols
at ambient environmental temperatures, no external heat
was applied to the sampling apparatus. No experiments
were conducted to determine if heat was needed to form
quantitative yields of the derivative on a silica gel
cartridge. This derivatizing agent was not investigated
further.


32
3/5-Dinitrobenzoyl Chloride Experiments
The third derivatizing agent investigated was 3,5-
dinitrobenzoyl chloride. Valdez and Reier (23) described
a method for the formation of dinitrobenzoate esters of
alcohols at low levels in aqueous solutions. The Valdez-
Reier derivatization process involved derivatizing the
alcohol with 3,5 dinitrobenzoyl chloride in an
acetonitrile solvent system. Their process consisted of
drying the acetonitrile and the 3,5-dinitrobenzoyl
chloride (DNBC) with molecular sieves. The alcohol
standards ranging in concentration from 20 to 2000 ppm
were prepared in a water solvent system. 100 jjL of the
dried acetonitrile and 10 jil, of the standard alcohol were
mixed together in a 2-mL vial, which contained several
molecular sieves. After allowing the sample to dry for
15 to 20 minutes, 1000 jiL of a 12 mg/mL solution of DNBC
was injected into the vial. This solution was then
heated at 75 C for 90 to 120 minutes. The resultant
product was then directly injected into a liquid
chromatograph.
Our study differs considerably from the Valdez-Reier
investigation in that no molecular sieves were used in
any of our studies. Also, our studies did not use heat
to prepare the alcohol derivative. Our derivatization


33
process occurred at ambient temperatures and on a silica
gel cartridge.
DNBC has been reported as amenable to alcohols and
amines in aqueous solutions. Most of these studies,
however, have been carried out in nonaqueous solutions
and under anhydrous conditions (24, 25). DNBC has been
used to derivatize glycols and other similar compounds
(26). Sample preparation prior to instrumental analyses,
however, proved to be rather tedious. Often involving
the washing of the reaction product with an acid or a
base. Lehrfield (12) was successful in derivatizing
carbohydrates with 3,5-dinitrobenzoy1 chloride. However,
sample preparation prior to instrumental analyses was a
long time-consuming process. Jupille (27) reported a
procedure for derivatizing alcohols in an aqueous
environment. The process involved ;the use of a two-phase
system of aqueous ethanol and benzene. The top organic
layer was directly injected into the chromatograph. It
was assumed, therefore, that all of the alcohol
*
derivative was extracted into the organic layer,
neglecting the possibility that some of the derivative
might be retained in the aqueous layer. The major
disadvantage of the techniques described above is that
they all suffer from extensive sample preparation prior
to injection into the chromatographic system. And in the
Valdes-Reier study, great pains were taken to remove


34
water.
This research report describes a process for
deirivatizing methanol at ambient temperatures on a silica
gel cartridge without taking precautions to remove
moisture. No washings or extensive extractions are
necessary. A single solvent phase system was used to
avoid problems associated with partitioning of the
alcohol derivative between the different solvents in a
two-phase solvent system. Although, it appeared that the
3,5-dinitrobenzoic acid and the associated anhydride were
formed during the derivatization process they did not
appear to affect the overall derivatization process.
Confirmation that the acid and associated anhydride
formed during the derivatization process needs to be
investigated. Methyl benzoate, benzoic acid and
anhydride formation may occur as shown below:


35
The mechanism chosen to ensure quantitative yields
of the methyl ester relied on the ability to have the
derivatizing agent, DNBC, in excess. Although, neither
the kinetics of the reaction nor the mechanism for methyl
ester formation were studied, it is reasonable to deduce
that the rate of formation of the benzoate is dependent
on the DNBC concentration. Therefore, the quantity of
derivative obtained is dependent on the original
concentration of the alcohol of interest. The
possibility of formation of the acid by reaction between
the derivatizing agent and water, and the subsequent
anhydride formation could reduce the efficiency of the
reaction between the 3,5-dinitrobenzoyl chloride and
methanol. The concentration of the DNBC could be greatly
decreased, which in turn would decrease the yields of the
methyl ester. It is believed that the reaction with
water must be the limiting step for the reaction of the
alcohol with DNBC when both are present.
Our first task was to determine the optimal
instrumental parameters and to determine the retention
time of both the derivatizing agent, DNBC, and that of
the benzoate derivative, methyl-3,5-dinitrobenzoate.
Figure 4.7 shows the chromatogram (A) of a 21 jjg/mL
standard solution of 3,5-dinitrobenzoyl chloride. The
chromatogram was obtained with a mobile phase composed of
70% acetonitrile and 30% water at a flow rate of 1


DETECTOR RESPONSE
36
Figure 4.7. HPLC chromatograms of 3,5-dinitrobenzoyl
chloride (A) and methyl-3,5-dinltrobenzoate
derivative (B).
The instrumental conditions are described
in the text.


37
mL/min., and at a wavelength of 224 nm. Figure 4.7 also
shows the chromatogram (B) obtained after adding 1 mL of
pure methanol to the above noted standard solution. The
sample chromatogram was obtained using the same
instrumental conditions as those used to obtain the
standard chromatogram. Chromatogram A clearly shows that
the retention time of approximately 3.20 minutes
corresponds to the derivatizing agent. Chromatogram B
dramatically illustrates that the peak at 3.20 minutes is
completely absent while another peak appears, which has a
retention time of approximately 4.61 minutes. The peak
at 4.61 minutes must, therefore, correspond to the
derivative, methyl-3,5-dinitrobenzoate. This experiment
was repeated several times and the same results were
obtained each time.
In order to confirm that the derivative was being
formed and that it had a retention time of 4.61 minutes,
a methyl-3., 5-dinitrobenzoate standard was synthesized in
our laboratory by following the procedure as described by
Shriner et al. (22). Figure 4.8 is the total ion
chromatogram and the mass spectra of the methyl-3,5-
dinitrobenzoate standard prepared in our laboratory. To
further characterize our standard it was compared to a
commercially prepared standard. Figure 4.9 is the total
ion chromatogram and the mass spectra of the commercially
prepared methyl-3,5-dinitrobenzoate. Clearly the mass


ABUNDANCE DETECTOR RESPONSE
38
TIC a* RP.rPnniR.Q
3. OE4:
1.UE4
1.BF.-1
1.4Et:
PE.1
innnn
noon
soon
4000
soon
o
TIME (min)
Figure 4.8. Total ion chromatogram and mass spectra
of in-house synthesised methyl-3,5
dinitrobenaoate.
Instrumental conditions are described
in the text.


39
TIME (min)
z
<
p
z
D
m
<
Scan 73S C1G.913 min'! of RTRFE2GR.D
IB no
1 J00:
1300'
(000;
BOO-
BOO;
400;
300 ;
D .ji. j llvlm-^il)J
GO BO 100 130
M/Z
Figure 4.9. Total ion chromatogram and mass spectra
of commercially prepared methyl-3,5-
dinitrobenzoate.
Instrumental conditions are described
in the text.


40
spectra match is quite convincing that indeed the
laboratory prepared standard is the methyl benzoate and
therefore, the HPLC peak of 4.61 minutes is that of
methyl-3,5-dinitrobenzoate. All mass spectra and gas
chromatograms were obtained by using the following
instrumental conditions: initial oven temperature 60 C
held for 1 minute and increased at a rate of 7 C/minute
to a final temperature of 260 C, total run time was
39.57 minutes, the injection port temperature was set at
225 C, the transfer line temperature was set at 270 C,
linear flow velocity was set at 30 cm/second @ 80 C, the
carrier was helium, the column was an XL Dimethyl
Silicone 12.5 meters by 0.2 millimeters internal
diameter, the film thickness was 0.33 jimeters, solvent
delay time was 3.00 minutes, set to scan from 50 to 500
m/z at 0.95 scans per second.
Silica gel cartridges were prepared as described
previously. The coating solution was prepared by weighing
0.1 grams of the derivatizing agent into a 100 mL
volumetric flask and diluting to the mark with
acetonitrile. Figure 4.10 illustrates the HPLC
chromatograms of the eluent blank coated cartridge (A),
which shows only one peak with a retention time of 3.18
minutes corresponding to the retention time of the
derivatizing agent which is shown in (C). The methyl-
3, 5-dintrobenzoate standard at a concentration of 1.04


DETECTOR RESPONSE
41
Figure 4.10 HPLU chromatograms of coated blank
cartridge (A), benzoate standard (B),
derivatizing agent (C), methanol vapors
(D). Instrumental conditions are
described in the text.


42
jjg/mL (B), shows a peak with a retention time of 4.60
minutes. The 3,5 dinitrobenzoyl chloride at a
concentration of 1.03 jjg/mL (C) has a retention time of
3.18 minutes. The chromatogram obtained from the eluent
after sampling the vapors over pure methanol for 15
minutes at a rate of 0.5 L/min., are shown in (D). The
sample chromatogram (D) shows three major peaks, one at
3.18 minutes, one at 3.48 minutes and one at 4.60
minutes. The 3.18 minute and 4.60 minute peaks
correspond to the derivatizing agent and the derivative,
respectively. The peak at 3.48 minutes was not
identified. The instrumental conditions were: the mobile
phase was set at 70% acetonitrile and 30% water, the
wavelength was set at 224 nm, the mobile phase was set at
1 mL/min., the integrator attenuation was set at 2^8, the
integrator chart speed was set at 0.50 cm/min., and the
column was a C18 jjBondapak.
Laboratory Sampling Studies
Our early studies of this compound focused on
derivatizing methanol vapors (28). An experiment was
devised in which vapors over pure methanol were sampled
into a coated silica gel cartridge. The vapors of an
open bottle of methanol were sampled.
The silica gel cartridges were prepared by taking


43
0.75 grams of 100-200 mesh silica gel and adding it to a
cartridge, shown in Figure 3.1. The silica gel
cartridges were rinsed with 10 mL of acetonitrile. The
cartridges were then coated with 10 mL of a 1032 jag/mL
solution of 3,5-dinitrpbenzoyl chloride. The coating
solution was drained through the cartridges by gravity.
Once coated the cartridges were dried overnight with
helium in an oven at 56 C. The sampling rate was set at
0.5 L/min., and the sampling times were 15 minutes, 30
minutes, and 45 minutes. Table 1 shows the results. The
results are based on an assumed purity of 100 % of the
methyl benzoate standard. The high performance liquid
chromatograms are shown in Figure 4.11. Tables 4.1-4.3
clearly demonstrate that DNBC is the limiting reagent.
Table 4.1. Concentrations of Vapors over Pure Methanol.
jjg MeOH rig Expected
Samplinq Time on Cartridge on Cartridge
15 minutes 4.76 4.76
30 minutes 6.14 9.52
45 minutes 11.9 14.3
As shown in Table 4.1 very small amounts of methanol
were derivatized.
It became a concern that not all of the methanol
vapors were being derivatized to the methyl ester. Two


44


45
explanations are offered. Not enough derivatizing agent
was present on the cartridge to derivatize all the
available methanol vapors or breakthrough was occurring.
An experiment was set-up in which two cartridges were
placed in series. Once again the sampling rate was set
at 0.5 liter/minute and the sampling times were 15
minutes, 30 minutes, 45 minutes, and 60 minutes. Table
4.2 shows the results obtained from this experiment. The
concentration of the coating solution (derivatizing
agent) used was 1 000 jug/mL.
Table 4.2. Results
Sampling Time
15 minutes
30 minutes
45 minutes
60 minutes
of
First Cartridge
jag MeOH
9.95 (9.95)
11.8 (19.9)
8.35 (29.8)
14.5 (39.8)
Experiments.
Second Cartridge
jag MEOH
29.6 (29.6)
11.8 (59.2 )
5.29 (88.8)
8.36 (118)
Breakthrough
The numbers in parentheses refer to the expected ug
of methanol on the silica gel cartridge. Thus, if enough
derivatizing agent was on the cartridge to derivatize all
of the available methanol then the concentration of
methanol on the silica gel cartridge should increase.
Clearly methanol vapors were breaking through the first
cartridge. This can be easily explained when it is


46
realized that large volumes of methanol vapors were being
sampled. In essence the first cartridge was overloaded
with methanol vapors. The concentration of derivatized
methanol decreased as sampling times were increased.
This was attributed to the larger volume of air which
were being sampled along with the methanol vapors. Thus,
the longer the sampling time, the larger the air volume
and therefore, the lower concentrations of methanol.
To prove that more derivatizing agent needed to be
placed on the cartridge, in order to collect these large
concentrations of methanol, another experiment was
conducted whereby the concentration of the coating
solution was increased by a factor of ten. The
cartridges were coated as previously described and once
again the cartridges were placed in series. The sampling
rate, and sampling times were as described above. The
results of this experiment are shown in Table 4.3.
Clearly more methanol was derivatized on the silica
cartridge.


47
Table 4.3. Results of Breakthrough Experiments.
First Cartridge
Sampling Time PPMV MeOH
Second Cartridge
PPMV MeOH
15 minutes
9.72
3.45
30 minutes
8.59
1 .98
45 minutes
4.46
1 .86
60 minutes
4.40
1.41
As Table 4.3 notes breakthrough was still evident
when using a coating solution with a concentration of
1 0,000 jjg/mL. Table 4.3 also shows that still more
methanol vapors were derivatized on the silica gel
cartridge than in previous experiments.
As noted previously a potential problem with this
approach to derivatizing methanol vapors was water.
Therefore, several experiments were conducted whereby
solutions containing varying amounts of water and
methanol were prepared and their vapors sampled.
Obviously, air was also sampled along with the methanol
and water vapors. To determine the retention time of the
acid peak, 1.1 mL of water were added to the 21 jug/mL
standard solution of 3,5-dinitrobenzoy1 chloride. Note
that this is the same standard solution to which 1 mL of
methanol was added ( HPLC chromatogram is illustrated in
Figure 4.7). The HPLC chromatogram (B) obtained after
adding 1.1 mL of water to the benzoyl chloride standard


48
is illustrated in Figure 4.12. Note that the peak at
1.74 minutes increased in size, when compared to the peak
at 1.72 minutes (A) for the standard. Also, the peak
height of the benzoyl chloride decreased by more than
half. This strong indication that the chloride does
indeed react with the water, but the acid peak elutes at
a much earlier time than does the benzoate derivative.
The chromatogram illustrated in Figure 4.12 was obtained
with the following HPLC instrumental conditions: mobile
phase 70% acetonitrile-30% water, flow at 1 mL/min.,
absorbance range set at 1 AUFS, integrator attenuation at
2^13, wavelength set at 224 nm, the detector bandwidth
was set at 16 nm and the analytical column was a
Lichrosorb RP18, 25 cm in length. Having established the
retention time of the derivatizing agent, the benzoate
derivative and the benzoic acid, it was decided to
investigate the varying solutions of methanol-water.
Silica gel cartridges were prepared and coated as
previouslydescribed. A 10% methanol / 90% water
Solution was prepared by pipetting 10 mL of methanol into
a 100 mL volumetric flask and diluting to the mark with
water. The vapors above this solution were sampled at a
rate of 0.5 L/min., for 15 minutes, 30 minutes, 45
minute, and 60 minutes. The coating solution was
prepared by weighing 10 grams of 3,5-dinitrobenzoy1
chloride into a 100 mL volumetric flask and diluting to


DETECTOR RESPONSE
49
A
TIME (min)
TIME (min)
Figure 4.12. HPLC chromatograms of 21 jjg/mL
3,5-dinitrobenzoyl chloride standard (A)
and same standard with 1.1 mL water
added (B).
Instrumental conditions are described
in the text.


50
the mark with acetonitrile. After sampling, the
cartridges were removed from the sampler and eluted with
acetonitrile. The eluent was collected in a 5. mL
volumetric flask. An HPLC chromatogram (obtained as
described above), for the 15 minute time period sampling
event is shown in Figure 4.13. Table 4.4 illustrates the
results.
Table 4.4
Sampling Time
15 minutes
30 minutes
45 minutes
60 minutes
Moisture Effect Results.
PPMV MeOEl
56.0
32.8
8.94
3.94
PPMV Expected
56.0
28.0
18.7
14.0
Table 4.4 clearly illustrates that methanol vapors
are being derivatized to the methyl ester in the presence
of large amounts of water. Also, illustrated by Table
4.4 is the rapid decrease of derivatized methanol after
15 minutes. As noted before as the sampling time
increases the concentration of methyl ester decreases.
For comparison purposes the experiment described above
was repeated except that this time vapors over pure
methanol were sampled. Table 4.5 shows the results
obtained with this experiment.


51
Table 4.5. Vapors over Pure Methanol.
Samplinq Time jjg MeOH jiq Expected
15 minutes 4070 4070
30 minutes 3065 8140
45 minutes 4445 12210
60 minutes 284 1 1 6280
BLANK (60 minutes) 651
In comparing Tables 4.4 and 4.5 it is realized that
lower concentrations of the methyl ester were obtained in
the presence of water. Indicating that water was
affecting the ability of the derivatizing agent to
derivatize all of the available methanol vapors. The 30
minute sampling period data showed a higher concentration
of methyl ester in the 10% methanol/ 90% water solution
than in the 100% methanol solution. The trend of lower
derivative concentration as sampling time increased was
once again observed.
Thus far we have proven that methanol and methanol
vapors could be derivatized with 3,5-dinitrobenzoy1
chloride both in solution and on a silica gel cartridge,
in the absence of catalysts or thermal energy. Also,
proven, thus far, is that the methanol derivative could
be adequately determined by reverse phase high
performance liquid chromatography by utilizing an ultra-
violet detector.


52
Having established that methanol vapors could be
derivatized on a silica gel cartridge with 3,5-
dinitrobenzoyl chloride and that the derivative could be
adequately determined by high performance liquid
chromatography it was decided that ambient air samples
should be studied.
Our first experiment was very surprising in that
methanol appeared to be present, at high levels, in the
Denver, Colorado metropolitan area.
Ambient Air Studies
In the first ambient air sampling event, performed
at the Auraria ambient air sampling station on December
1, 1990, the silica gel cartridges were prepared as
described previously except that 81.86 grams of 3,5-
dinitrobenzoyl chloride were weighed into a 100 mL
volumetric flask and diluted to volume with acetonitrile.
The sampling rate was set at 2.5 Liters/minute and the
sampling period was 4 hours. The chromatogram obtained
is shown in Figure 4.13 (B) along with the benzoate
standard (A). It is quite clear from the dilution
factor, 50X, that needed to be used that large
concentrations of methanol were present. The calculated
concentration of methanol present in ambient air,
uncorrected for the blank, was 0.117 ppmv.


DETECTOR RESPONSE
Figure 4.13. HPLC chromatograms of 18.2 jug/mL
methyl-3,5-dinitrobenzoate standard (A)
and ambient air sample (B) at 5OX
dilution.
Instrumental conditions are described
in the text.


54
Another sampling event was performed, at the same
sampling site on January 11, 1991, as described above.
The gas chromatogram of the methyl-3,5-dinitrobenzoate
standard (A) and the sample chromatogram (B) are
illustrated in Figure 4.14. Clearly the retention time
of 16.829 for the sample (B) matches the retention time
of 16.897 for the benzoate standard (A). The number of
moles of methanol present was 4.95E-05, compared to
5.18E-07 moles of methanol present in the field blank and
9.22E-07 moles of methanol present in the laboratory
blank.. The blank gas chromatograms are shown in Figure
4.15. The top chromatogram is the laboratory blank
cartridge. The laboratory blank cartridge was prepared
the same way as the sample cartridge except that it was
never taken out of the laboratory. The bottom
chromatogram is that of the field blank. The field blank
cartridge was prepared in the same manner as the sample
cartridge and was taken to the sampling site and placed
in the sampling apparatus the same way as the sample
cartridge except that no air was flowed through the
cartridge. The difference between the laboratory blank
and the field blank is that the field blank was exposed
to the atmosphere, whereas the laboratory blank cartridge
was not exposed to the ambient air. Clearly, then, as
shown by the results presented above it appears that
methanol is present in Denver's atmosphere. However,


DETECTOR RESPONSE. DETECTOR RESPONSE
55
TIME (rain)
TIME (min)
Figure 4.14 Gas chromatograms of benzoate standard (A)
and ambient air sample (B).
Instrumental conditions are described in
the text.


DETECTOR RESPONSE
Figure 4.15 Gas chromatograms of laboratory blank (A)
and field blank (B) cartridges.
Instrumental conditions are described in
the text.


57
concentrations of 0.117 ppmv and 2.47 ppmv are a bit too
high to be true. This researcher did not expect to find
methanol present in Denver's atmosphere at the observed
concentrations. However, it must be noted that this
researcher was unable to find historical data on the
concentrations of methanol present in Denver's air.
Therefore, no comparable data exists to compare to this
researchers results.
Based on the thought that perhaps this researcher had
contaminated the laboratory glassware, all the laboratory
glass ware was scrupulously washed with soap and water.
The glassware was dried in an oven at 56 C for 12 hours.
Another sampling event was performed at the Auraria
ambient air sampling station on January 15, 1991. All
silica gel cartridges were prepared as previously
described. The air sampling was performed as in previous
experiments. Great care was taken to minimize laboratory
cross contamination. The results from this sampling
event are illustrated in Figures 4.16 and Figure 4.17.
Once again the environmental sample appears to indicate
the presence of methanol at a concentration of 2.16 ppmv
( 4.33E-05 moles of methanol) in Denver's atmosphere.
The laboratory blank contained 6.90E-08 moles of methanol
and the field blank contained 2.20E-07 moles of methanol.
Clearly, no significant contamination of either the
silica gel cartridges, the solvents or.the glassware


DETECTOR RESPONSE
58
TIME (min)
Figure 4.16 Gas chromatograms of ambient air sample
(A) at 10X dilution and benzoate
standard (B) at 100 ng/^il.
Instrumental conditions are described in
the text.


59
TIME (min)
Figure 4.17 Gas chromatograms of laboratory blank (A)
and field blank cartridges (B).
Instrumental conditions are described in
the text.


60
appears to have occurred. The levels of methanol in the
blanks is approximately the same for all the studies
reported above. Thus, it appears that the blanks are
only showing background levels of methanol. The low
levels of methanol, in the blanks, indicate that cross
contamination did not occur.
Having ruled out the possibility of cross contamination
as the source for the high values of methanol present in
the sample cartridges, the following explanations are
offered: 1). this researcher inadvertently contaminated
the sample cartridges but not the blank cartridges; 2).
this researcher purposely added methanol to the sample
cartridges and not to the blank cartridges; 3). the
methanol source must be in the sampling apparatus, which
affects only the sample cartridges and not the blank
cartridges; and 4). the measured levels of methanol are.
an indication of the amount of methanol present in
Denver's ambient air.
In all of the ambient air sampling experiments, the
derivatizing coating solution was prepared by taking 80
grams of 3,5 dinitrobenzoyl chloride (obtained from
KODAK) and weighing into a 100 mL volumetric flask and
diluting to the mark with acetonitrile.
At this point it was decided that the amount of
3,5-dinitrobenzoyl chloride used to make up the silica
gel cartridge derivatizing solution should be decreased;


61
to see if the amount of derivatizing agent present had a
profound effect on the derivatization process.
Therefore, 10 grams of 3,5-dinitrobenzoy1 chloride were
weighed into a 100 mL volumetric flask and diluted to the
mark with acetonitrile. The cartridges were prepared as
previously described.
Denver's atmosphere was again sampled at the Auraria
sampling station on April 20, 1991. The sampling rate
and sampling duration were the same as in all previous
ambient air studies reported in this report. This time
the sampling was conducted by placing the cartridges in
series (one on top of the other) to determine if
breakthrough was occurring. As can be seen by the HPLC
chromatograms shown in Figures 4.18 and 4.19 the sample
chromatograms, at 1X, look similar to the laboratory and
field blank chromatograms. The chromatograms were
obtained by using a 50% acetonitrile and 50% water mobile
phase mixture, the wavelength was set at 224nm, the
integrator attenuation was set at 2~12, the column was a
Lichrosorb RP18, 25 cm long.
The top silica gel cartridge wa's found to contain
4.07E-07 moles of methanol, which corresponds to 0.020
ppmv methanol. The second top cartridge was found to
contain 3.76E-07 moles of methanol, which corresponds to
0.019 ppmv methanol. The third top cartridge was found
to contain 4.10E-07 moles of methanol, which corresponds


DETECTOR RESPONSE
62
Figure 4.18. HPLC chromatograms of (A) ambient air sample
and (B) methyl-3,5-dinitrobenzoate standard.
Instrumental conditions are described in the
text.


DETECTOR RESPONSE
63
TIME (min)
Figure 4.19. HPLC chromatograms of (A) laboratory blank
cartridge and (B) field blank cartridge.
Instrumental conditions as in Figure 4.18.


64
to 0.021 ppmv methanol. The laboratory and field blank
were found to contain 3.70E-07 and 4.60E-07 moles of
methanol. Clearly, the sample cartridges have the same
amount of methanol as the blanks. Thus, it is concluded
that no ambient methanol was derivatized on any of the
cartridges. Based on these results the other cartridges
were not analyzed. It would seem, therefore, that the
amount of (derivatizing agent used to coat the cartridges
plays a role in determining if ambient methanol will be
derivatized.
It must be noted that all previous experiments
were conducted by using the derivatizing agent (3,5-
dinitrobenzoyl chloride) obtained from Kodak. This
derivatizing agent was very light yellow in color and
consisted of solid granules. When the cartridges were
coated with the 80 gram/100 mL derivatizing solution the
silica gel cartridges took on a yellowish appearance as
opposed to their original white color. Thereby indicting
that the derivatizing agent had been deposited on the
silica gel.
In the experiments just discussed the derivatizing
agent used was obtained from Aldrich with a claimed
purity of 98+%. This derivatizing agent was very brown
in color and was of a very fine nature. When the silica
gel cartridges were coated they took on a very brown
color. This derivatizing agent was very different in


65
color and texture than the Kodak derivatizing agent.
/
After the cartridges were oven dried it was noticed that
a very thin brown solid layer of material was on top of
the silica gel cartridges (probably undissolved DNBC).
The laboratory blank and the field blank exhibited the
same effect. No similar event was noticed in any of the
cartridges that were coated with the Kodak prepared
derivatizing agent solution.
It is not known whether there is a significant
difference between the 3,5-dinitrobenzoyl chloride
obtained from Kodak and that obtained from Aldrich.
However, the Kodak chloride looked to be of a purer
nature.
Another sampling event was performed on May 3,
1991, using the Aldrich benzoyl chloride. The results
were essentially the same as in the previous experiment
in which the Aldrich derivatizing agent was used.
This researcher has shown that in three different
instances high concentrations of methanol were found in
Denver's atmosphere. This researcher has also shown that
on two other separate occasions no quantifiable methanol
was found to be present in Denver's ambient air.
However, there are several factors that could explain
these results. First, the amount of derivatizing agent
used to coat the cartridges was reduced by approximately
10 times; second the source of the derivatizing agent was


66
changed from Kodak to Aldrich, third and perhaps more
significant is that the earlier experiments were
conducted in the winter months of December and January.
The later studies were done in April and May. It could
be that Denver's air was cleaner in April and May than it
was in December and January. However, since historical
data on methanol concentrations in Denver's air is
nonexistent it is difficult to prove that the weather
plays a significant role.
Another experiment was attempted in which 3 grams of
3.5- dintrobenzoyl chloride (obtained from Kodak) were
weighed into a 50 mL volumetric flask and 30 mL of
acetonitrile were added with a volumetric pipet. The
cartridges were then prepared as previously described.
The sampling duration and rate was the same as in all the
previous ambient air sampling experiments. It must be
noted that these three grams were the last quantities of
3.5- dintrobenzoyl chloride available at this researchers'
laboratory. The experimental results once again showed
that the quantities of methanol present in the sample
cartridge were the same as those present in the blank
cartridges.


CHAPTER V
CONCLUSIONS
A sampling and analysis that can be used for the
derivatization of methanol in the ambient atmosphere
without the aid of catalysts or thermal energy were
developed during the course of this study. The 3,5-
dintrobenzoy1 chloride reacts with methanol to form the
methyl-3,5-dinitrobenzoate. The benzoate is then
monitored by reverse phase high performance liquid
chromatography with ultra violet detection.
The ambient methanol is derivatized on a silica gel
cartridge; which is coated with the appropriate
quantities of 3,5-dintrobenzoyl chloride.
This report presented the results of both laboratory
controlled and ambient air experiments. It is concluded
based on the laboratory studies that 3,5-dinitrobenzoyl
chloride is a good derivatizing agent for converting an
alcohol (methanol) to a benzoate.
The ambient air Studies provided mixed results.
Some studies clearly demonstrated that 3,5-dinitrobenzoyl
chloride was indeed derivatizing methanol on a coated


68
silica gel cartridge. However, the concentrations of
methanol found in Denver's atmosphere appear to be too
high to be accepted with a high degree of confidence.
Other studies did not find methanol above background
quantities exhibited by the blanks. However, as
explained earlier it is suggested that Denver's
atmosphere was cleaner in the months of April and May
than in December and January. Clearly, the weather must
play a role on the quantities of methanol that can be
found in Denver's atmosphere at any one time. Denver's
winters are typically very wet and cold. This leads to
dirty air being trapped in the lower layers of the
atmosphere.
The studies presented in this report can be used
as a foundation for conducting other studies. This
researcher was interested in finding a derivatizing agent
or agents capable of derivatizing ambient quantities of
methanol on a silica gel cartridge. Not a whole lot of
time was spent on focusing on a single approach. Several
different derivatizing agents and several different
derivatizing solutions concentrations were studied.
The results presented here indicate that 3,5-
dinitrobenzoyl chloride needs to be investigated further
to determine if it is an adequate agent for derivatizing
ambient concentrations of methanol in the ambient air.
The experimental results tentatively indicate that 3,5-


69
dinitrobenzoyl chloride does possess the ability to
derivatize methanol present in the ambient air.
Several more studies need to be performed in order
to gain a great degree of certainty in the ability of the
chloride to derivatize ambient quantities of methanol.
Recommendations for Future Studies
It became apparent after conducting the ambient air
sample analyses that isocratic instrumental conditions
were not optimal for separating all the possible products
formed by the reaction between 3, 5-dintrobenzoy1 chloride
and water and 3,5-dintrobenzoy1 chloride and methanol and
of course the left over 3,5-dinitrobenzoyl chloride. As
was shown earlier the 3,5-dinitrobenzoic acid elutes form
the column at an earlier time than the benzoyl chloride
which in turn elutes before the benzoate. The ambient
air studies chromatograms clearly illustrate that a big
broad peak is present from approximately 1 minute to
approximately 7 minutes when isdcratic conditions are
used. A goal was to separate the benzoic acid peak from
the benzoyl chloride peak. Gradient elution was
attempted to try to separate the peaks. Several
different gradient mobile phase compositions were used
but none provided better separation or resolution of the
peaks of interest than the isocratic conditions. The


70
benzoate peak elutes at approximately 9 minutes when a
50%/50% acetonitrile/water mobile phase composition is
used. This provides enough separation from the benzoic
acid and benzoyl chloride peaks that quantitation of the
benzoate is possible. By observing the HPLC
chromatograms obtained from the ambient air studies
(Figure 4.13 ) it was noticed that 'the sample
chromatograms did not differ significantly from the
laboratory or field blanks which would strongly indicate
that the derivatizing agent was not significantly
depleted by the water concentrations present in the
ambient air. If water had consumed all of the
derivatizing agent it would be expected that the sample
HPLC chromatograms would differ in appearance from the
laboratory blank. The laboratory blank cartridges were
not exposed to the ambient air as were the sample
cartridges.
However, it was difficult to prove with a great
degree of confidence that the benzoyl chloride was not
depleted by the ambient water; since the benzoic acid
peak, if present, elutes from the column before the
benzoyl chloride peak. Also, since large quantities of
derivatizing agent were used, the derivatizing agent peak
if not completely consumed by the ambient water or
ambient methanol would give rise to a very large and
broad peak. This researcher believes that water did not


71
significantly deplete the concentration of derivatizing
agent from the silica gel cartridge and therefore the
large broad peak observed at the beginning of each
chromatogram is primarily due 3,5-dinitrobenzoy1
chloride.
It is highly recommended that normal phase high
performance liquid chromatography be attempted to support
the above noted conclusions. In normal phase liquid
chromatography the elution order would be reversed to
that obtained by reverse phase high performance liquid
chromatography. This would mean that methyl-3,5-
dinitrobenzoate would elute from the column first
followed by the 3,5-dinitrobenzoy1 chloride and the last
component to elute would be the 3,5-dinitrobenzoic acid.
If this approach does not separate the chloride from the
acid peak than gradient conditions must be explored.
It is suggested that perhaps the purity of the 3,5-
dinitrobenzoyl chloride plays a role in the ability of
the derivatizing agent to derivatize methanol. A
reasonable experiment that should be attempted to verify
that the purity of the derivatizing agent plays a role is.
to repeat the laboratory controlled experiments by
coating several silica gel cartridges with a coating
solution that was prepared from the Aldrich benzoyl
chloride. If the results are significantly different
than the results obtained with the studies presented here


72
then clearly the purity of the derivatizing agent is very
important.
The use of two analytical columns should be explored
as this would provide better resolution of the target
analyte peaks.
Sampling time and sampling duration (volume) must be
evaluated to determine if sampling speed or sampling time
play a significant role in the ability of the
derivatizing agent to derivatize the methanol. Clearly,
the volumes of air samples would affect the rate at which
the chloride can convert the alcohol to the benzoate.
This must be explored further.
The kinetics of the reaction between water and 3,5-
dinitrobenzoyl chloride and the kinetics between 3,5-
dinitrobenzoyl chloride and methanol must be studied and
compared to determine if water or methanol compete on an
equal or greater basis for the derivatizing agent.
Gas chromatography should also be explored as the
instrumental technique. Gas chromatography with a flame
ionization detector or with a mass selective detector
offers several advantages. The most important advantage
is that the 3,5-dinitrobenzoyl chloride does not gas
chromatograph and neither does the 3,5-dinitrobenzoic
acid. However, the methyl-3,5-dinitrobenzoate does
chromatograph nicely. This would eliminate the benzoic
acid-benzoyl chloride elution problem that was observed


73
by reverse phase liquid chromatography. Also, since such
low concentrations of methanol wer^ quantitated by
reverse phase HPLC it is reasonable to suggest that the
flame ionization detector could reach much lower
detection limits than the ultra violet detector used in
the' HPLC experiments.


REFERENCES
(l) U.S. Environmental Protection Agency. Reported in
"EPA Adds 37 Additional Cities to Ozone Non-
Attainment List". Air/Water Pollution Report,
1989, 245.
(2) Bhatt, A. 1980, 2_2, ; Gupta, V. K. Ind. J. Environ. Health 203 .
( 3 ) Jephcott, C. M. Analyst 1935, 60, 588.
(4) Schiff, H . Ann. 1866, 140, 92.
( 5 ) Edvin, F. Z. Z. Anal. Chem. 1937, 110, 22.
(6) Barns, E. 1942, 23, C.; Speicher, H. W. J. Ind. Hyq. Toxical 1 0.
( 7 ) Ruch, W. E. "Quantitative Analysis of Air
Pollutants", Ann-Arbor. Science Publishers, 1975,
1 46.
(8) Verma, P.; Gupta, V. K. Talanta 1984, 3J_ ( 5 ), 394.
(9) Maeda, Y.; Fujio, Y.? Suetaka, T.; Munemori, M.
Analyst 1988, 113, 189.
(10) Gabele, P.; Ray, W.; Duncan, J.; Burton, C.
"Characterization of Exhaust Emissions From
Methanol and Gasoline Fueled Automobiles",
EPA 460/3-82-004, August 1982.
(11) Hanst, P. L.; Wong, N. W.; Bragin, J. Atmos.
Environ. 1982, J_6, ( 5 ), 969.
(12) Lehrfield, J. L. J. Chromatoqr. 1976, 120, 141.
(13) Bjorkqvist, B.; Toivonen, H. J. Chromatoqr 1978,
153, 265.
(14) Higgins, J. W. J. Chromatoqr. 1976, 121, 329.
(15) Fitzpatrick, F. A.; Siggia, S. Anal. Chem. 1973, 45,
2310.


75
(16) Porcaro, P. J.; Shubiak, P. Anal. Chem. 1972, 44,
1865.
(17) Galetto, W. G.; Kepner, R. E.; Webb, A. D. Anal.
Chem. 1966, 38., 34.
(18) Carey, M. A.; Persinger, H. E. J. Chromatoqr.
Sci. 1972, 10, 537.
(19) Pavia, D. L.; Lampman, G. M.; Kriz, G. S.
"Introduction to Organic Laboratory Techniques",
W. B. Saunders Company, 1976, 97.
(20) "The Merck Index" Martha Windholz, ed, Merck & Co.,
Inc. 1976.
(21) "Handbook of Chemistry and Physics" Robert C. Weast,
ed, Chemical Rubber Co. Press, Inc.:Boca Raton,
Florida, 1979.
(22) Shriner, R. L.; Fuson, R. C.; Curtin, D. Y. "The
Systematic Identification of Organic Compounds"
John Wiley & Sons, Inc. 1965, 247.
(23) Valdez, D.; Reier, J. C. J. Chromatoqr. Sci. 1986,
24, 356.
(24) Ross, M. S. F. J. Chromatoqr. 1977, 141, 107.
(25) Nozawa, A.; Ohnuma, T. J. Chromatogr. 1980, 187,
261 .
(26) Salvage, T. Analyst 1970, 95., 363.
(27) Jupille, T. J. Chromatoqr. Sci. 1979, J_7, 160.
(28) Palomares, A.; Anderson, L. G.; Lanning, J. A.
Presented at the 32nd Rocky Mountain Conference,
July 29-August 3, 1990, Denver, Colorado.