Citation
Carbonyl compounds in the Denver-Metro Area, with emphasis on the C3 and larger compounds

Material Information

Title:
Carbonyl compounds in the Denver-Metro Area, with emphasis on the C3 and larger compounds
Creator:
Molefe, Ofentse
Publication Date:
Language:
English
Physical Description:
xii, 106 leaves : illustrations ; 28 cm

Thesis/Dissertation Information

Degree:
Master's ( Master of Science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Chemistry, CU Denver
Degree Disciplines:
Chemistry

Subjects

Subjects / Keywords:
Carbonyl compounds -- Analysis ( lcsh )
Air -- Pollution -- Colorado -- Denver Metropolitan Area ( lcsh )
Air -- Pollution ( fast )
Carbonyl compounds -- Analysis ( fast )
Colorado -- Denver Metropolitan Area ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 104-106).
Thesis:
Department of Chemistry
General Note:
Department of Chemistry
Statement of Responsibility:
by Ofentse Molefe.

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:
42611973 ( OCLC )
ocm42611973
Classification:
LD1190.L46 1999m .M65 ( lcc )

Full Text
CARBONYL COMPOUNDS IN THE
DENVER-METRO AREA, WITH
EMPHASIS ON THE C3
AND LARGER COMPOUNDS
by
Ofentse Molefe
B.S., University of Botswana, 1993
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Chemistry
1999


This thesis for the Master of Science
degree by
Ofentse Molefe
has been approved
by
4*-//2 / 99
Date


Molefe, Ofentse (M.S., Chemistry)
Carbonyl Compounds in the Denver-Metro Area,
with emphasis on the C3 and larger Compounds.
Thesis directed by Dr. Larry G. Anderson
ABSTRACT
Formaldehyde, acetaldehyde and acetone have attracted attention due to-
their health hazardous nature at elevated concentrations. These lower
molecular weight carbonyls can have both primary and secondary
sources. Previous downtown Denver studies have shown a strong
correlation of these carbonyls with carbon monoxide, suggesting a net
motor vehicle source. In addition, previous results show that
acetaldehyde and acetone concentrations are higher in Boulder than in
Denver. This observation may be due to a net photochemical production
of these carbonyls in Boulder. In this work, the possibility of biogenic
carbonyl contribution to Denver, Boulder and other outlying sites is
explored. Methacrolein and methyl vinyl ketone, which are C4 carbonyls
produced by the oxidation of isoprene (a biogenic hydrocarbon), are used
as indicators in this work. The 2,4-dinitrophenylhydrazine (DNPH)
derivatives of methacrolein and methyl vinyl ketone are synthesized and
used to prepare calibration standards which are subsequently used in the
determination of methacrolein and methyl vinyl ketone at the mentioned


sites, alongside with formaldehyde, acetaldehyde and acetone. The
presence of other higher molecular carbonyl compounds is also
determined. The instrumentation employed in this study includes a Hewlett
Packard High Performance Liquid Chromatograph, with a built-in UV-
Visible diode array detector. A Zorbax (5pm, 4.6 x 250 mm) reversed-
phase chromatographic column is used. A water/acetonitrile solvent
system was used and detection was done at 360 nm, 376 nm and 380 nm.
The carbonyl measurement results are presented for Boulder, West
Boulder, Golden, Barr-Lake and Denver. These results include the
presence (or the absence thereon) of methyl vinyl ketone and
methacrolein at the above sites; and the quantitative comparison of
formaldehyde, acetaldehyde and acetone for the various sites studied.
These samples were collected and analyzed during the summer of 1998.
The results shows in general, the carbonyl average concentrations are
higher in the outlying areas than in Denver. In addition, the daytime
concentrations are higher than in the nighttime. Propionaldehyde and
methyl ethyl ketone were identified at the studied sites. In addition there
are traces of other unidentified higher molecular carbonyl compounds.
This abstract accurately represents the content of the candidates
thesis. I recommend its publication.
Signed
Larry G. Anderson


DEDICATION
To
Lele


ACKNOWLEDGMENT
I am indebted to my supervisor and advisor, Dr. Larry G. Anderson, for his
guidance and patience with me during the course of this work, and also for
his advice and support during the entire course of my academic pursuits. I
would also like to thank Dr. John Lanning for his advice and suggestions. I
am thankful to Dr. D. Zapien for his suggestions and also to Mike Pribil,
Dr. Gabriela Weaver, Jeff Boon, Julie Hull and Gordon Pierce for their
assistance. I would also like to thank Mai Pham, Rene Rabouin and Gene
Bouche' for their moral support.


CONTENTS
Figures...........................................................x
Tables...........................................................xi
Chapter
1. Introduction ................................................ 1
1.1 Purpose of Study..............................................1
1.2 Aldehydes and Ketones ........................................2
1.2.1 Health Effects of Aldehydes and Ketones ....................3
1.2.2 Sources of Aldehydes and Ketones............................4
1.3 Sources of OH, 03 and N03....................................12
1.4 Carbonyl Measurement Techniques..............................13
1.4.1 The 2,4-Dinitrophenylhydrazine Coated Silica-Gel Technique.13
1.4.2 The DNPH-Carbonyl Reaction ................................14
1.4.3 UV-Visible Spectroscopy ...................................15
1.4.4 Liquid Chromatography .....................................17
2. Experimental Methods ........................................20
2.1 Sampling Protocols...........................................20
2.1.1 Sampling Sites.............................................20
2.2 Analytical Methods...........................................24
2.2.1 Preparation of Calibration Standards.......................24
2.2.2 Liquid Chromatographic Analysis............................26
2.2.3 Qualitative Analysis....................................... 27
2.2.4 Quantitative Analysis......................................28
vii


2.2.5 Method Performance........................................ 28
3. Results and Discussions.......................................33
3.1 Synthesis of Methyl Vinyl Ketone and Methacrolein derivatives.33
3.2. HPLC Analysis of the Field Samples.........................41
3.2.1 The Boulder-Marine Site....................................41
3.2.2 The West -Boulder Site.....................................43
3.2.3 The Barr-Lake Site.........................................47
3.2.4 The Auraria Site......................................... 47
3.2.5 The Golden (NREL) Site.....................................52
3.2.6 Methacrolein and Methyl Vinyl Ketone at the Sites..........56
3.2.7 Comparison of the Sites Studied............................58
3.2.8 Detection Limit Measurements...............................59
3.2.9 Method Quality Assurance Results...........................63
3.2.10 2,4-DNPH Coated Cartridge Carbonyl Collection..............65
3.2.11 Cartridge Extraction Efficiency...........................66
3.2.12 The HPLC Precision Measurements ..........................69
3.2.13 Retention Times Reproducibility Results...................72
3.2.14 Mass Spectrometry Results for the MTA-DNP and MVK-DNP
Hydrazone ................................................73
4. Conclusions................................................87
4.1 Some Recommendations for Further Studies...................89
viii


Appendix
A. Preparation of Calibration Standards (Calculations)..............91
B. Detection Limits Calculations....................................95
C. The Permeation Tube System.......................................97
D. Retention Times Reproducibility Data.............................99
Works Cited.........................................................104
IX


FIGURES
Figure
1.1 Carbonyl DNPH Reaction..................................15
2.1 The Permeation-Tube System................................ 23
3.1 UV-Spectrum for MVK-DNP Hydrazone.........................35
3.2 UV-Spectrum for MTA-DNP Hydrazone..........................36
3.3 Calibration Curves for MVK-DNP Hydrazone..................38
3.4 Calibration Curve for MTA-DNP Hydrazone...................40
3.5 Sample Chromatogram for Boulder-Marine Site................42
3.6 Sample Chromatogram for West-Boulder Site..................45
3.7 Sample Chromatogram for Barr-Lake Site.....................48
3.8 Sample Chromatograms for Auraria Site......................50
3.9 Sample Chromatograms for Golden Site.......................53
3.10 Mass Spectra for MTA-DNP Hydrazone........................74
3.11 Mass Spectra for MVK-DNP Hydrazone........................75 -
3.12 Mass Spectrum for 2-butenal-DNP Hydrazone.................77
3.13 Mass Spectrum for the DNPH Reagent........................78
3.14 Mass Spectrum for Formaldehyde-DNP Hydrazone..............79
3.15 Mass Spectrum for Acetaldehyde-DNP Hydrazone..............80
3.16 Methyl Vinyl Ketone (2,4dinitrophenylhydrazone).........83
3.17 Methacrolein (2,4dinitrophenylhydrazone).................84
3.18 1HNMR Spectrum of MVK-DNPH..............................85
3.19 1HNMR Spectrum of MTA-DNPH..............................86
x


TABLES
Table
1.1 Biogenic Volatile Organic Compounds Emissions...............7
1.2 Some Biogenic Hydrocarbons and their Sources ...............9
3.1 Melting Points for the MTA-DNP and MVK-DNP Hydrazones.....33
3.2 Standards Calibration Data for MTA-DNP Hydrazone...........37
3.3 Standards Calibration Data for MVK-DNP Hydrazone...........39
3.4 The identified Carbonyl Chromatographic Peaks at the
Boulder-Marine Site.......................................41
3.5 The Unidentified Carbonyl Chromatographic Peaks at
the Boulder-Marine Site...................................44
3.6 Chromatographic Peaks at the West-Boulder Site.............46
3.7 Carbonyl Chromatographic Peaks at the Barr-Lake Site.......49
3.8 Carbonyl Chromatographic Peaks at the Auraria Site.........51
3.9 Carbonyl Chromatographic Peaks at the Golden Site..........52
3.10 Comparison of the Daytime and the Nighttime Carbonyl
Concentrations at all Sites Studied.......................55
3.11 Comparison of the Sites Studied............................60
3.12 Results of the Replicate Injections of the 1 ppbv MTA-DNP
Hydrazone Standard........................................61
XI


3.13 Results of the Replicate Injections of a 2 ppbv
MVK-DNP Hydrazone Standard.................................62
3.14 Results of the MTA-DNPH Permeation-
Tube Sampling Analysis.....................................63
3.15 Results of the MVK-DNP Hydrazone Permeation-Tube
Sampling Analysis..........................................64
3.16 Cartridge-Collection Efficiency Study Results for MTA DNP
Hydrazone..................................................65
3.17 Cartridge-Collection Efficiency Study Results for MVK-DNP
Hydrazone................................................. 66
3.18 Cartridge-Extraction Efficiency Study Results for MTA-DNP
Hydrazone..................................................67
3.19 Cartridge-Extraction Efficiency Study Results for
MVK-DNP Hydrazone......................................... 69
3.20 HPLC Method Precision Results for Formaldehyde-DNP
Hydrazone..................................................70
3.21 HPLC Method Precision for Acetaldehyde-DNP Hydrazone.......70
3.22 HPLC Method Precision for Acetone-DNP Hydrazone............71
3.23 Sample Retention-Time Reproducibility Study Results........72
3.24 Summary of Mass Spectra Results for MTA-DNP
and MVK-DNP Hydrazones.....................................76
3.25 1H-NMR Results of MTA-DNPA and MVK-DNPH....................82
XII


1. Introduction
1.1 Purpose of the Study
Previous carbonyl studies have indicated that acetaldehyde (ethanal) and
acetone (propanone) concentrations are generally higher in the outlying
areas than in Denver. For example, Susan D. Riggs (1) has conducted a
carbonyl study in the Denver-Metro area and her findings indicates that the
acetaldehyde and acetone concentrations tend to be higher in Barr-Lake,
Niwot Ridge and Boulder, than in Denver. In addition, Anderson et al. have
analyzed carbonyl data they collected from 1991 to 1993 (2) and their
findings point to the same conclusions. Susan Riggss findings indicates
that in the Summer of 1996 the acetaldehyde and acetone concentrations
in Denver ranged from 0.41 ppbv to 5.75 ppbv respectively; while in
Boulder, acetaldehyde ranged from 0.66 ppbv to 21.74 ppbv and acetone
from 0.08 ppbv to 39.20 ppbv (1). This observation seems to contradict the
conclusion that the major source of carbonyl compounds in the Denver-
Metro area is motor vehicle exhausts. Anderson et al. have correlated the
winter source of carbon monoxide (CO) to that of formaldehyde and
acetaldehyde, suggesting that these compounds have a common source,
namely motor vehicle exhaust pipes. The much higher motor vehicle
density in Denver suggests that the concentration of the aforementioned
compounds should be much higher in Denver than in the outlying areas,
but this is not the case. A few explanations for this observation can be
brought forward. Firstly, the higher concentrations of these compounds in
the outlying areas may suggest a net photochemical production in those
1


areas. A photochemical possibility may be further strengthened by the
observation that in these outlying areas, and also in Denver, the summer
carbonyl compounds concentrations are substantially larger than those in
winter. In addition, in either season, the daytime carbonyl concentrations
are notably higher than the nighttime concentrations. The hypothesis of
this study is that the higher carbonyl concentrations in the outlying areas is
due to a net photochemical production in those areas. The possibility of
biogenic contribution to these sites (using methacrolein and methyl vinyl
ketone as indicators) is explored. In addition the presence and
identification of some higher molecular weight carbonyl compounds are
looked into.
1.2 Aldehydes and Ketones
Both aldehydes and ketones are carbonyl compounds. Carbonyl
compounds contain the central structural feature C = O, the carbonyl
group. In ketones, the carbonyl group is bonded to two carbon atoms so
that the simplest ketone, acetone (propanone) bears the following
structure:
H3C
= 0
H3CX
(Acetone)
Aldehydes on the other hand have only one or no carbon atom and one
hydrogen atom bonded to the carbonyl group. The simplest aldehyde will
thus be formaldehyde (methanal) with the following structure:
2


H
(Formaldehyde)
1.2.1 Health Effects of Aldehydes and Ketones
Carbonyl compounds in general have been reported as mutagens and
possible carcinogens(3). For example formaldehyde and acetaldehyde are
irritants to the eyes and the upper respiratory system.
Formaldehyde, acetaldehyde and acetone have been classified as toxic
and hazardous substances by the Environmental Protection Agency (EPA)
fact sheets (4). Exposure to substancial acetaldehyde concentrations can
cause dizziness, lightheadedness, or even cause one to pass out (4).
Repeated exposure can even cause skin allergies, which may lead to skin
irritation from exposure to very small concentrations of the chemical. As a
result of the toxicities of these carbonyl compounds, the National Institute
for Occupational Safety and Health (NIOSH) recommends formaldehyde
exposure limit of 16 ppbv for an eight hour average (5). The current
formaldehyde exposure limit however, set by the Occupational Safety and
Health Administration (OSHA US Department of Labor) is 750 ppbv for
an eight hour average (6). This is the Permissible Exposure limit (PEL) for
formaldehyde. The PEL is described as the maximum exposure over an
eight hour period.The Short Term Exposure Limit (STEL) is the exposure
over a fifteen minute period. NIOSH has formaldehyde STEL
recommendation of 100 ppbv (5). Other OSHA occupational PELS are 200
ppm for acetaldehyde, 1000 ppm for acetone and 200 ppm for methyl ethyl
ketone (2-Butanone) (5). Thus, employers whose workplaces are covered


by these standards are bound by law to monitor their employees exposure
to these compounds.
1.2.2 Sources of Aldehydes and Ketones
The sources of the aldehydes and ketones can be divided into
anthropogenic and biogenic. The biogenic sources are natural while
anthropogenic sources are a result of man's activities. In addition, both
biogenic and anthropogenic sources of aldehydes and ketones can be
further subdivided into primary and secondary sources. Primary sources
emit the compound in its original state, while secondary sources emit a
precursor compound which is subsequently converted to the compound of
interest by chemical reactions in the atmosphere.
1.2.2.1 Anthropogenic Sources
Carbonyl compounds can be emitted by both industrial sources and motor
vehicle exhausts. For example, the more important primary sources of
formaldehyde and acetaldehyde in the atmosphere are believed to be
direct emissions, from motor vehicles and other combustion sources (7). In
addition formaldehyde is used in the production of synthetic resins, and is
also employed as a chemical intermediate in some industrial processes.
Acetaldehyde is used in the manufacture of acetic acid and acetone is
employed by autobody shops, printing companies, chemical and metal
industries. Motor vehicle emissions of formaldehyde and for some fuels,
acetaldehyde, have been shown to increase with the use of oxygenated
fuels, including methanol, ethanol, and methyl tertiary butyl ether (MTBE)
(7). The motor vehicle and industrial processes emissions of carbonyls are
especially important in winter, when photochemical carbonyl production is
4


minimal due to the relatively low sunlight intensity. Traces of methacrolein
and methyl vinyl ketone are also emitted by automobile exhausts; emitted
as part of tobacco smoke and biomass combustion (8). In addition, methyl
vinyl ketone is also emitted by whiskey manufacturing plants (8). As shall
be explained later however, most of the ambient methacrolein and methyl
vinyl ketone are a result of biogenic sources. In summer, the
photochemical sources of carbonyls become more significant, due to the
increased sunlight intensity, which plays an important role in
photochemical activities. These include the products of the gas-phase
reactions of ozone and the hydroxyl radical (OH) with alkenes (9), and the
volatile organic compounds (VOCs) released by industries.
1.2.2.2 Biogenic Sources
(Isoprene and the Monoterpenes)
Vegetation naturally releases organic compounds to the atmosphere (10).
There are many organics emitted by different types of vegetation and
these include isoprene and the monoterpenes. In general decidous trees
(hardwoods) like oaks, aspens, willows and poplars emit mainly isoprene
while monoterpenes are emitted mainly by coniferous forests (softwoods)
like pine trees, firs, juniper, spruce, cedars, and redwoods (11). Isoprene
(2-methyl-1,3-butadiene) has the following structure:
CH2=CHC(CH3) = CH2
Concentrations of isoprene and the monoterpenes in ambient air are
typically ~ 1 10 ppbC outside forest canopies and ~ 10 ppbC within
them, and usually comprise < 10 % of the total hydrocarbons in the rural
sites (11). Roberts et al. conducted measurements of monoterpene
hydrocarbons at Niwot Ridge, Colorado (12). Their GC/MS analyses
5


provided positive identification of a-pinene, d-limonene, mycerne, A-3-
carene, camphene and (3-pinene; plus a tentative identification of (3-
phellandrene and a-thujene. (3-pinene and a-pinene were found to be the
dominant monoterpenes. In addition, the summertime average
concentration for the sum of the five major monoterpenes (camphene, d-
limonene, A-3-carene, (3-pinene and a-pinene) was 0.30 ppbv (12).
Although the authors report having detected isoprene in this study, they did
not quantify it, but expect its concentration to be atleast 20% of that of the
monoterpenes. Another Colorado study was conducted on the
measurements of hydrocarbons, oxygenated hydrocarbons, carbon
monoxide and nitrogen oxides, at a site on the western perimeter of
Boulder (13). At least twenty five hydrocarbons are reported to have been
identified and quantified in this study. The authors report concentrations of
isoprene lower than 100 pptv in this study. The authors also report
considerable quantities of methyl tertiarybutyl ether (MTBE), which is a
motor fuel additive (for CO control).The Isoprene and the monoterpenes
emissions have been shown to increase with increases in temperature and
light intensity (11). One would expect therefore, that the largest emissions
occur in summer, when the solar influx (sunlight reaching the earth)
reaches its maximum. Table 1.1 shows the amount of biogenic volatile
organic compounds emitted in Colorado in 1988, 1990, 1991 and 1995.
The other states, California, Arizona and Texas are included for
comparison. The information was obtained from an Environmental
Protection Agency (EPA) web page (14). In the below table, at least 60 %
of the emissions (in general) occurred in summer. In general the biogenic
emission rates have been shown to be affected by climatology, land use
and temperature fluctuations. Monoterpenes have the chemical formula
6


C-ioH-io and maybe acyclic or bicyclic. Table 1.2 (11) gives names,
structures and sources of some monoterpenes as well as isoprene. Both
isoprene and the monoterpenes are olefins (they possess at least one
double bond in their structures). This renders these compounds
susceptible to oxidation in the atmosphere. These and other biogenic
compounds have received
Table 1.1. Biogenic Volatile Organic Compound
Emissions (x 1000,000 lbs)
State 1988 1990 1991 1995
Colorado 1778 1496 1634 1652
Arizona 1070 1084 1034 1096
California 3630 3556 3422 3588
Texas 5422 5728 4488 5298
attention for their contribution to the production of ozone plus other
compounds and the formation of aerosols in urban and rural areas (11).
They react with ozone (03), the nitrate radical (N03) and the hydroxyl
radical (OH) to produce carbonyl compounds and radicals which undergo
further reactions. The rate constant for the reaction of isoprene with the
OH radical at room temperature, in cm3moleculeV1 is 1.0 x 10'1,
1.4 x 10-17 with respect to 03 and 5.8 x 1013 with respect to N03 (10). Sara
M. Aschmann and Roger Atkinson (15) conducted a study to determine the
formation yields of methyl vinyl ketone and methacrolein from the gas-
phase reaction of 03 with isoprene. The mechanism shows the formation
of significant amounts of methyl vinyl ketone and methacrolein as well as
formaldehyde (15). The percentages are the yields from the overall
7


reaction. The mechanism proceeds by initial addition of 03 to the C = C
double bonds to form two primary ozonides:
o o
I _
CHj---CHC(CHj)=CHj xl
CHe=CHCpay-CHg
The first ozonide shown above then decomposes to give methacrolein
(major product), formaldehyde and a biradical, as shown below:
l I ________
CHg---CHCfCHjJ^CH*
(i) CH2=C(CH3)CHO (methacrolein) + CH200
(ii) HCHO + [CH2=C(CH3)CHOO
The other ozonide breaks down into methyl vinyl ketone, more
formaldehyde, a biradical and a very small amount of an
epoxide:
1 I -----------------
CH2=CH6(CHb)-CHj
B
(i) CH2=CHC(0)CH3 (methyl vinyl ketone) + CH200
(i) HCHO + CH2=CHC(CH3)00
8


Table 1.2 Some biogenic hydrocarbons
and their sources.
Hydrocarbon Structure Source
Isoprene \ Emitted by
/\ deciduous trees
a -Pinene I Numerous
<8 Conifers
P-Pinene & California black sage and general conifers
d-Limonene $ Emitted by loblolly pine, California -black sage and
disturbed eucalyptus.
9


The OH radicals also form in this reaction. The biradicals formed in this
reaction undergo further reactions to ultimately form formaldehyde:
[CH2=CHC(CH3)00]_^ [CH2=CHC(OOH)=CH2] _
CH2=CHC(0)CH2 + OH
CH2=CHC(0)CH2 + 02 CH2=CHC(0)CH200
CH2=CHC(0)CH20
CH2=CHC(0)CH20 CH2=CHCO + HCHO (formaldehyde)
CH2=CHCO + 02 CH2=CHC(0)00 _ CH2=CHC(0)0 _
ch2=ch + co2
CH2=CH + 02 HCHO (formaldehyde) + HCO
Methacrolein and methyl vinyl ketone have also been found to be the
major products of the OH-isoprene reaction (10). The reaction mechanism
starts by the addition of the OH radical to the C = C double bonds The
OH can add at different parts of the isoprene double bonds, and in each
case the resulting intermediate species goes on to react with oxygen, then
with nitric oxide and finally with more oxygen to form either methacrolein or
methyl vinyl ketone, formaldehyde and a peroxy radical. The following
reaction sequence results when the OH adds on the more substituted
isoprene double bond. Methyl vinyl ketone is formed, together with
formaldehyde:
O-

<->2


NO
OH.

0-.
o
+ HOj- + [HCHO
When the OH adds to the less substituted double bond, the reaction
10


sequence leads to the formation of methacrolein, together with the
formaldehyde and the peroxy radical:
o2
OH
+ Hey +
HCHO
In addition, the isoprene OH reaction can assume the following reaction
paths, which also yields more methyl vinyl ketone and methacrolein:
OH
NO
00*
OH
O-
+
ho2- +
HCHO
00-
OH
OH
NO
OH
0
+ H02- +
HCHO
11


It must be mentioned that although the chemistry of isoprene has been
extensively studied in environmental chambers in laboratories, the gas-
phase oxidation of terpenes under atmospheric conditions is much less
understood (16). The atmospheric oxidation of these compounds yield
products mostly different from those produced by the atmospheric
oxidation of isoprene. For example, the major carbonyl products of d-
limonene, a-pinene and p-pinene (some of the most abundant
monoterpenes in the atmosphere) are 4-acetyl-1-methylcyclohexene, 6,6-
dimethylbicyclo [3.1.1] heptan-2-one and 4-acetyl-1-methylcycohexene,
respectively. These carbonyls are further oxidized to yield more
compounds, including formaldehyde as a minor product.
1.3 Sources of OH, O3 and NO3
These species which are responsible for the oxidizing nature of the lower
atmosphere, are reproduced as follows: Ozone may be transported
downwards from the stratosphere. It is also produced by the photolysis of
nitrogen dioxide gas (11):
N02 - NO + O (X < 420 nm)
0 + 02 O3
The photolytic destruction of ozone serves as the primary source of the
OH-radicals (11):
03 ^02 + O (1D); (X< 320 nm
0(1D) + H20-^ 2 OH
Photolysis of oxygenated compounds like carbonyls also produce the OH
radicals, for example, formaldehyde photolysis yield OH radicals as follows
(11):
12


HCHO - H + CHO ; X < 335 nm
H + O2 + M HO2 + M
CHO + O2 -HO2 + CO
The H02 lead to the formation of hydrogen peroxide, the photodissociation
of which yield more OH radicals (11):
2 H02 H202 + 02
H2O2 - 2 OH; X < 350 nm
In addition, OH radicals result from oxidation of nitric oxide (11):
NO + HO2 -NO2 + OH
1.4. Carbonyl Measurement Techniques
1.4.1 The 2,4-Dinitrophenylhydrazine
Coated-Silica Technique
Several techniques are presently used to measure gas-phase carbonyl
compounds. The 2, 4-dinitrophenylhydrazine (DNPH) coated cartridge -
liquid chromatography (LC) method, which was first employed to measure
ambient carbonyls some fifteen years ago, is the most widely used (17).
This technique involves the conversion of carbonyl compounds to their
hydrazone derivatives and then analyzing them with high performance
liquid chromatography (HPLC), employing ultraviolet detection (UV) (or
the DNPH-LC-UV-technique). The technique involves packing of solid
sorbents into cartridges and then coating the sorbents with acidified
DNPH. The most commonly used sorbents are silica gel and C18 silica gel
(18). In this work, silica gel, which is a polymer of silicon and oxygen with
surface hydroxyl groups, is employed as the packing material. The DNPH-
LC-UV is the method currently recommended by the Intersociety
13


Committee and by air pollution control agencies in several countries (18).
In addition, the Environmental Protection Agency (EPA) has specified the
use of the method for measuring airborne carbonyls as part of a new
nationwide program, which began in 1994 (18). When using this method,
retention times, UV-visible spectra and response factors are employed for
diagnostic purposes. The continous emission of the deuterium lamp can be
utilized in conjunction with a UV- diode array detector and this helps in the
improvement of analytical selectivity (17). That is, by monitoring more than
one wavelength at a time, changes in relative responses of
chromatographic peaks can be better followed, leading to improved
qualitative identification of compounds in a mixture. In addition, the diode
array detector serves as a useful way of gaining increased sensitivity
especially in the analysis of complex mixtures of compounds. This is so
because the detector allows the operator to monitor different compounds
at their wavelength of maximum absorption.
1.4.2 DNPH-Carbonyl Reaction
In HPLC-UV analyses, derivatization can be utilized to enhance the
detectability of certain compounds, which would otherwise not be easily
detectable. The DNPH part of the DNPH-carbonyl derivatives serves as
the major chromophore. The hydrazine group, (-HN-NH2) reacts with
carbonyls to form hydrazones. This acid-catalyzed reaction occurs when
the lone pair of electrons on an -NH2 group attached to another hetero-
atom attacks the carbonyl carbon of an aldehyde or ketone (19). In the
same way, the reaction of DNPH with aldehydes and ketones is employed
to synthesize the corresponding DNP-hydrazones. The reaction of a
carbonyl compound with 2,4-DNPH is shown in figure 1.1 below.
14


(Carbonyl)
nhnh2
(DNPH)
NH-N = CRR
(DNP Hydrazune)
Figure 1.1 Carbonyl-DNPH Reaction
1.4.3 UV-Visible Spectroscopy
DNPH exhibit absorption maxima at 357 nm (17). Simple aldehydes and
ketones shows only weak absorption in the ultraviolet region of the
spectrum due to the n to n* electronic transitions of the carbonyl group
(20). However, when carbonyls are derivatized with DNPH the resulting
hydrazones have their absorption shifted more toward the visible region of
the spectrum, due to the DNPH chromophore. The aromatic (conjugation)
in the 2,4-Dinitrophenylhydrazones (DNP) lead to intense absorption due
to the ft to ft* transitions, with a much higher extinction coefficient of the
absorption maxima. As implied above, the DNPH part of the hydrazones is
the main chromophore. However the following observations can be made
when different types of carbonyls are derivatized with DNPH (17).


1.4.3.1 DNPH Derivatives of Aliphatic
Monofunctional Carbonyls
The carbonyl substituents Ri and R2 (Ri and R2 are aliphatic, and R2 = H
for aldehydes) have only a small influence compared to that of the main
chromophore, (N02)2C6H3NH-N=CRiR2t on the overall absorption of the
DNPH derivative, so that the absorption maxima of such derivative is only
about 360 nm. For example the absorption maximum for the DNPH
derivative of formaldehyde is 352 nm, and that for the acetaldehyde is
362 nm (17).
1.4.3.2 DNPH Derivatives of Aromatic Carbonyls
Contribution toward the absorption maxima of the derivative increases
when the carbonyl substituents are aromatic. There is a bathochromic
shift from 360 nm to about 385 nm. This is attributable to the conjugation in
the aromatic ring of the carbonyl. For example, benzaldehyde-DNPH has
its absorption maxima at 385 nm (17).
1.4.3.3 DNPH Derivatives of Dicarbonyls
These types of carbonyls are exemplified by glyoxal, methylglyoxal and
glutaraldehyde.The di-DNPH chromophore of these carbonyls results in a
large bathochromic shift of the absorption maxima, from about 360 nm to
about 410 430 nm. For example, glyoxal-DNPH has its maximum
wavelength of absorption at 437 nm (17).
16


1.4.3.4 DNPH Derivatives of Unsaturated
Carbonyls
Methyl vinyl ketone and methacrolein are encompassed in this group.
The derivatives exhibit a bathochromic shift from 360 nm to about 376 nm.
Methyl vinyl ketone-DNPH absorbs maximally at 376 nm. The shift is
probably due to the unsaturation in the carbonyls structures.
1.4.4 Liquid Chromatography
Column chromatography is a physical method of separation in which the
molecules to be separated are distributed between a stationary phase and
a mobile phase. When molecules pass through the column, they spend
some of the time in the mobile phase and some in the stationary phase .
The time spent in the stationary phase depends on the affinity of the
molecules to the coating material of the phase. The primary components
of a high liquid chromatograph are as follows:
High pressure pump: It delivers a constant flow of solvent (eluent) to the
system.
Injector System: The HP1090 system employed in this study involved an
auto-injector. The sample injection valve, a Rheodyne model 7010, was a
six port rotary valve. It operates by rotation of a flat rotor against a flat
stator.
Column: The separation column is the heart of chromatography, providing
versatility in the types of analyses that can be performed, due to the wide
choice of materials for the stationary and mobile phases (21). It is thus
possible to separate molecules that differ only slightly in their physical and
17


chemical properties. The column/injector system is a critical area where
efficiency can be easily lost. In this study a Zorbax analytical column,
which is a reversed-phase C-is column, was employed. The stationary
phase is a hydrocarbon-bonded surface formed by bonding octadecylsilyl
groups (Ci8H37Si-) to silica. During a chromatographic elution, the most
polar molecules elute first since they have little affinity for the non-polar
stationary phase. The advantages of this system is that the highly polar
molecules will have relatively shorter retention times and less peak tailing
(22).
Detector: By far the most widely used detector is the UV detector, which
measures the change in the UV absorption as the solute passes through a
flow cell in a UV transparent solvent (22). UV detectors have very low
noise characteristics and have very high sensitivity (17). Compounds with
high molar absorption coefficients (e = 10,000 to 20,000) can be detected
even at only a few nanograms concentrations, due to the stability and high
sensitivity of the current UV detectors (22). UV detectors are concentration
sensitive and thus, in addition to being used for qualitative measurements,
they are also employed for quantitative analyses. In this work a UV-diode
array detector, which is an example of a scanning wavelength detector,
was used. The diode array works in parallel, simultaneously monitoring all
wavelengths (21). This permits the attainment of a real-time spectrum for
each solute as it elutes. In addition to the above HPLC components, others
like recorders, integrators, in-line filters and more may be used. The
retention time is the piece of qualitative information acquired from a liquid
chromatographic analysis. Quantitative column chromatography is based
on comparison of either peak height or peak area of the analyte to that of
the standard. In this work, peak area was the quantitative tool of choice.
18


The instrument employed has an electronic integrator, which permits
precise estimation of peak areas. Quantitative analysis can involve either
the use of external standards calibration method or the internal standards
one. The internal standards method involves the introduction of an
accurate volume of an internal standard and sample. The ratio of analyte
to internal standard then serves as the analytical tool. This method
minimizes the uncertainties associated with the sample injection. The
external standard calibration method, which was used in this work,
involves the preparation of a series of standard solutions. Subsequently
the peak areas are plotted against the concentrations and the linear
regression analysis results of the data serves as the analytical tool.
19


2. Experimental Methods
2.1 Sampling Protocols
2.1.1 Sampling Sites
Data was collected at the following sites:
Auraria: This site is located on the Auraria campus along the west edge of
downtown Denver, and is about a quarter mile from the same. It is
surrounded by the following major highways/corridors: Speer Boulevard,
Auraria Parkway, Colfax Avenue and the Interstate-25. The samples were
collected continuously for a period of 12 hours each, from 17th to 22nd
September, 1998.
Boulder Marine: This site is located at 2320 Marine street, and is near
downtown Boulder in a residential area, near a well travelled two-lane
street. This is an outlying site. Boulder itself is about 35 km northwest of
downtown Denver, and it represents an urban environment. Samples were
collected from 22nd to 27th September, 1998. Each sample ran for a period
of 12 hours.
Boulder West: This site is in a well wooded setting and represents a rural
type site. Since it is in a wooded area, its data might give better insight
pertaining to the presence of biogenically derived carbonyl compounds.
The samples were collected from 3rd to 8th September, 1998.
Barr Lake: Another outlying area, approximately 30 kilometers north-east
of Denver. Samples at this location were collected from 10th to 15th
September, 1998. Like at the other sites, each sample was collected for a
period of 12 hours.
20


Golden at the National Renewable Energy Lab (NREL):
This is another outlying site, about 20 30 kilometers west of Denver.
Samples at this site were collected from Wednesday, 26th to 31st August,
1998. Each sample was collected for a period of 12 hours. The above
sampling sites were chosen for their variability as described above, and
also for their convenience. In addition, some previous carbonyl data are
available for some of these sites, which could be utilized for comparison
purposes. The Boulder Marine and the Golden sites are well prepared
sampling sites operated by the Colorado Department of Health. The
Auraria site is also a regular sampling site operated by the University of
Colorado at Denver. An automated, portable, sequential sampler unit was
employed in this study. The sampler housed the following: a voltmeter
calibrated for flow, a vacuum pump and a vacuum gauge, a series of
solenoid valves through which air can be sucked when the pump is
operated; an internal computer which switched the valves on and off
according to its program; and the associated circuitry and tubing. During
the sampling periods the sampler was placed inside the sampling station
and connected to the sampling line by a polyethylene tube. The sampling
line in turn, was connected to the outside environment. The internal
computer could be programed to switch on the valves sequentially or
according to the desired order and time. In this work, the timer was
programmed to switch on the valves after every 12 hour sampling period.
The daytime samples ran from 8:00 am to 8:00 pm of the same day, while
the nighttime samples ran from 8:00 pm to 8:00 am (of the following day).
Carbonyl air samples were collected by drawing air directly through the
DNPH coated sampling cartridges, from the sampling line and via the
polyethylene tubing which linked the sampling line and the sampler. During
21


the field sampling, one field blank was collected with each weeks samples.
The field blank was prepared and handled exactly as the other cartridges,
but not sampled (had no air drawn through it) during the sampling
procedure. The blank cartridge carbonyl concentration was subtracted
from the sample cartridge concentration, during quantitative analyses. In
addition, two lab blanks were stored for each batch of 70 cartridges. These
were also prepared and handled as the other cartridges but were not
sampled, but kept in the lab for later evaluation of storage and handling
effects. In the experiments involving the method performance (calibration
of the instrument and the technique), that is, the experiments involving the
HPLC detection limit, the coated cartridge carbonyl collection and
extraction efficiencies, a permeation tube system was employed. Figure
2.1 shows the outline of the system. It consists of a glass casing (where
the carbonyl permeation device are placed). The glass casing itself is
placed in a temperature regulated waterbath. A thermostat heats and
mixes the bath so that it can be kept at a consistent temperature. Air flows
through the glass the glass casing to the exhaust. There is also the dilution
air, which can be regulated to control the concentration of the carbonyl
compound available for sampling. The permeation devices (tubes) were
prepared in the laboratory, and were made from approximately 31/2 inches
of Teflon tubing, which was capped on both ends with tight-fitting glass
rods after the desired carbonyl compound has been introduced inside. The
permeable nature of the Teflon tubes allows a steady release of the
vaporized carbonyl compound into the glass casing, for subsequent
sampling. It is vital that the whole system be allowed several days to
reach a state of equilibrium, in terms of temperature and the carbonyl-
release (vaporization) rate.
22


pump
-TTTT-
Flow meter
DNPH cartridge
Back-up
cartridge
air
mass flow
controller
Front
flartriilon
exhaust
mass flow
controller
permeation tube
air
Figure 2.1. Schematic Diagram of the
Permeation-Tube System.
23


2.2 Analytical Methods
2.2.1 Standards Preparation
2.2.1.1 DNPH Purification
DNPH (Aldrich Chemical Company Inc) was recrystallized twice from
acetonitrile (methyl cyanide) (HPLC grade) and rinsed twice with cold
acetonitrile and dried in a dessicator. An acetonitrile solution of the purified
DNPH was then run on the HPLC to see if it was pure and free from any
DNPH oxidation products or hydrazone impurities.
2.2.1.2 Preparation and Coating of the Cartridges
12 mm i.d. polyethylene syringe bodies (the cartridges) are each packed
with 0.6 0.7 grams of 40 microns silica gel. Two porous but firm
polyethylene filters are used to hold the silica in place, and the silica is
pressed firmly in place with a small light rod. After packing with the silica
gel, the cartridges are coated with the acidified DNPH solution. The
coating solution itself is prepared by mixing the following solutions in a 500
mL volumetric flask: 50 mL of saturated 2,4-DNPH solution; 1 mL of
concentrated hydrochloric acid. The flask is filled to the mark with
acetonitrile (HPLC grade). The saturated DNPH solution was itself
prepared by equilibrating purified 2,4-DNPH crystals over HPLC grade
acetonitrile solvent. Immediately after its preparation, the coating solution
was coated onto the silica -packed cartridges to prevent any possible
contamination of the solution, especially by the ambient carbonyls. During
24


the coating procedure, 5.0 mL of the solution was transferred into each
packed cartridge and the excess solution being allowed to drain to the
waste! Once packed, the cartridges are immediately dried in a 40 C oven.
In addition, once prepared, the coating solution must be used entirely, and
on the same single batch of cartridges, to maintain consistency insofar as
possible contamination and random errors are concerned.
2.2.1.3 Synthesis of the Carbonyl -
DNPH Derivatives
The carbonyls employed in this study were methacrolein (95%, Aldrich)
and methyl vinyl ketone (99%, Aldrich). For each carbonyl, 0.5 g were
dissolved in acetonitrile to make a 10 mL solution (solution I). Then 0.5 g of
the purified DNPH crystals was dissolved in acetonitrile, and 2 drops of
concentrated hydrochloric acid added to make a 10 mL solution
(solution II). Then solution I and II were mixed and the mixture allowed to
stand for 20 30 minutes to allow the carbonyl DNPH crystals to form.
The product was filtered on Whatman paper. The crystals were then
recrystallized once from hot ethyl acetate and washed twice with the cold
ethyl acetate. The crystals were then stored in a dessicator and allowed to
dry.
2.2.1.4 Preparation of the Calibration Standards
A detailed account of the mathematics involved in the preparation of the
standards is outlined in appendix A. This procedure employed the use of
the carbonyl derivatives which were previously synthesized and purified in
the laboratory. A specified amount of the dry carbonyl-hydrazone was
accurately weighed and dissolved in enough acetonitrile to make a solution
25


of an exactly known volume. The solution served as a stock solution of an
accurately known concentration. Then using volumetric pipettes, accurate
volumes of the stock solution were transferred to pyrex volumetric flasks
and diluted to accurate volumes, to make a series of standard solutions.
The series of standards were later used to construct an external standard
curve. Methyl vinyl ketone-dinitrophenylhydrazone (MVK- DNP
hydrazone) standards with concentrations ranging from 0.49 ppbv to
9.73ppbv and those ranging from 0.50 ppbv to 10.01 ppbv for
Methacrolein-dinitrophenylhydrazone (MTA DNP hydrazone) were
prepared.
2.2.2 Liquid Chromatographic Analysis
The analytical protocols of this method have been discussed implicitly in
the introduction, and are very similar to the EPA Method TO-11 (23).
Following sample collection, each cartridge was eluted with 5.0 mL of the
HPLC grade acetonitrile. Then aliquots of the extract were analyzed by
liquid chromatography, with a UV- visible detector. The system used was a
Hewlett Packard HP1090M HPLC with a built-in UV-visible diode array
detector, and an auto-injector. The system control was the HPLC
Chemstation (Pascal series). The analytical column used was a Zorbax
(5 pm, 4.6 x 250 mm). The eluent consisted of an acetonitrile-water
(purified) mixture which was continually degassed with helium gas, at
pressures not exceeding 45 psi. The acetonitrile solutions of the carbonyl-
DNP hydrazone standards and the air samples were analyzed by the
above described system (HP1090M HPLC, UV-visible diode array
detector). The detection was at the following wavelengths, which were
monitored at the same time: 360 nm, 376 nm and 380 nm. Thus for every
26


compound/standard, a set of three liquid chromatograms were recorded. In
preliminary and initial trial LC runs, both gradient and isocratic elutions, at
various acetonitrile: water ratios were employed, to try and find the best
separation system (peak separation and resolution). In the final analysis,
both an isocratic and a gradient elution, with eluent composition from 60:40
acetonitrile:water ratio to one of 75:25, running over a 25 minutes period
were employed. The flow rate was 1.000 mL/minute, the column pressure
was about 113 bars, the column temperature of 20 C and the injection
volume was 20.0 pL. The UV-visible absorption spectra of 200-600 nm of
the pure DNPH, carbonyl-DNP hydrazones standards and the carbonyl-
DNP hydrazones (from samples) were recorded every 640 ms and in 2 nm
steps.
2.2.3 Qualitative Analysis
Qualitative analysis involved the use of retention times (of peaks) and UV-
spectra as the diagnostic tools. The retention times of the carbonyl-DNP
hydrazones in the air samples were compared to those of the carbonyl-
DNP hydrazones standards synthesized in the lab (methyl vinyl ketone-
DNPH and methacrolein-DNPH). In addition, the UV-visible spectra (200-
600 nm) of the chromatogram peaks in the sample were compared to that
of methyl vinyl ketone-DNPH and methacrolein-DNPH standards.
Moreover, since the absorption maxima of methyl vinyl ketone-DNPH and
methacrolein-DNPH are at longer wavelengths than those of the more
abundant carbonyls like formaldehyde, acetaldehyde and acetone, three
wavelengths (360 nm, 376 nm and 380 nm) were monitored at the same
time and the response of the relevant peaks monitored. Formaldehyde,
acetaldehyde and acetone were also analyzed in each sample. The use of
27


the retention times and response factors as diagnostic tools require that
these parameters be reproducible over time. The reproducibility procedure
is described elsewhere in this chapter.
2.2.4 Quantitative Analysis
The quantitative analysis of the carbonyl-DNPH derivatives in the air
samples involved the use of external standards. The series of standards
prepared earlier were used to construct an external standard calibration
curve-absorbance (peak area vs concentration). The calibration factors
obtained from the subsequent least square linear regression of the data
were used to calculate the parts-per-billion by volume (ppbv)
concentrations of the carbonyls that were conclusively identified in the air
samples. The DNP hydrazones standards for formaldehyde,
acetaldehyde, acetone, propionaldehyde, acrolein and methyl ethyl ketone
were also employed in this procedure.
2.2.5 Method Performance
2.2.5.1 Cartridge Collection Efficiency Test
The collection efficiency experiments were performed with respect to
methyl vinyl ketone (MVK) and methacrolein (MTA). The experiments
results furnished the following two pieces of information:
Whether or not the coated cartridges actually collect the carbonyls.
Efficiency with which the cartridge collects the carbonyls. The experiments
were done with the same sampler used in the field. Two cartridges were
placed in series and the air from the permeation tube system (containing
the two carbonyls) pulled through the cartridges, at 1.6 L/minute for 45
28


minute periods for methacrolein and 1 hour periods for methyl vinyl ketone.
A volt meter (The air sampler was calibrated using the volt meter)
monitored the air sampling rate, while a dry test meter connected to the
exhaust end of the sampler measured the total air volume sampled per
cartridge.The collection efficiency would be 100% if all of the MVK and the
MTA molecules in the sampled air reacted with the 2,4-DNPH coating on
the silica gel of the front cartridge. After efficiency test sampling, both the
front and the back-up cartridges were eluted and analyzed by LC. The
peak response of the back-up cartridge was expressed as a percentage of
the front cartridge and the average of such values represented the
approximate cartridge collection efficiency.
2.2.5.2 Cartridge Extraction Efficiency Test
In each test, a carbonyl permeation tube was dropped into the permeation
system and the whole system allowed to equilibriate at a temperature of 40
degrees Celsius. Then the air from the permeation system was sampled
through two coated cartridges connected in series, for a 1 hour period in
the case of methyl vinyl ketone and 45 minutes for methacrolein. Six
replicate samples were done per carbonyl derivative. The sampled
cartridges were then each eluted twice in a row. Both extracts were then
analyzed by LC and the peak response of the second extractant expressed
as the percentage of the first. The efficiency would be 100% if all the
carbonyl content of the cartridge was removed by the first extraction and
all mass of compounds accounted for. Six such tests were performed, and
an average value established for both methacrolein and methyl vinyl
ketone.
29


2.2.5.3 HPLC Precision
The precision of the HPLC system used was determined by a series of
replicate injections of five formaldehyde/acetaldehyde/acetone 2,4-
DNPH standards. The standards concentrations in ppbv were 0.93, 2.31,
4.62, 8.34, 13.86 for formaldehyde-DNPH; 0.96, 2.40, 4.8, 8.66 and 14.4
for acetaldehyde-DNPH and 0.78, 1.92, 3.85, 6.95 and 11.55 for the
acetone-DNPH derivative. For each standard, three replicate injections
were performed and run on the HPLC. The analysis was isocratic and the
solvent system was 60% acetonitrile: 40% water; the injection volume was
20pl and although the detection was done at the three wavelengths of 360
nm, 376 nm and 380 nm, only the 360 nm data were utilized as it is the
approximate maximum absorption wavelength for the three compounds.
2.2.5.4 HPLC Detection Limit
Detection limit can be defined as the quantity of analyte that can be
detected with reasonable certainty for a given analytical procedure (25). It
can also be described as the concentration of analyte that gives a signal,
X, significantly different from the background signal, Xb (21). What might
be "significantly different" thus depends on the analysts technique and the
precision and accuracy of the method and the instrumentation employed.
Poole and Poole (25) states that the detection limit can be determined as
the signal equal to three times the standard deviation of the gross blank
signal. Using Willard et al. (21) definition above, the detection limit of a
technique can thus be quantified as: X Xb = 3Sb where,
X is the signal with minimum detectable analyte concentration, Xb is the
30


signal of the blank and Sb is the standard deviation of replicate blank
signals. In my analyses, there was no integratable peak at the area of
interest in the blanks, which leaves the final expression for the detection
limit as:
X 0.00 ppbv = 3 Sb, or
X = 0.00 ppbv + 3 Sb
Where X becomes the minimum identifiable signal, or the detection limit.
Replicate injections of two standards (one of MVK-DNPH and the other of
MTA-DNPH) were prepared and 10 replicate injections in the case of
methacrolein, and 12 for methyl vinyl ketone were performed by the HPLC.
The peak responses were recorded. The absolute standard deviation in
the peak responses and the regression output of the calibration curve were
then used to compute the approximate detection limit of the technique, that
is Sb in the above equation is the standard deviation in the responses
(converted to concentration units) of the replicate injections of the
standards described above.
2.2.5.5 Quality Assurance of the Method
This procedure was developed to determine the reproducibility of the
permeation tube system and the analysis technique employed, as regard
the sampling, the extraction and the analysis method. The permeation
system requires several days to equilibrate. The method involved sampling
MVK and MTA from the permeation tube system, at 1 liter/minute for the
following period of times: 45 minutes for methacrolein and 60 minutes for
methyl vinyl ketone. The cartridges were eluted and analyzed; the
standard deviation in the peak responses was used as a measure of the
precision of the method.
31


2.2.5.6 Mass Spectrometry and Nuclear Magnetic
Resonance of the Synthesized Hydrazone.
The MTA and MVK hydrazones were run on a mass spectrometer,
interfaced to an HPLC chromatogram, the results of which were used to
confirm the structures of the compounds. After several hours of
optimization of the mass spectrometer, the compounds were analyzed at a
pressure of 80 psi and a temperature of 55 C. The HPLC conditions were
27.8 C, 92 bars, and the solvent composition was 60 % acetonitrile and
30 % water, at a flow rate of 0.6 ml/min. In addition, the two compounds
were analyzed on a Nuclear Magnetic Resonance (NMR) spectrometer,
set to the Proton Nuclear Magnetic Resonance (1H-NMR) mode. NMR is a
molecular spectroscopy process whereby the radio frequencies absorbed
and emitted by a compound (subjected to a magnetic field) are measured,
and their patterns correlated to the molecular structure of the compound.
The 1H-NMR thus gives information about the number and types of
hydrogens in the compound under consideration. The fundamental
components of a NMR spectrometer are a powerful magnet, a radio
frequency generator, a radio frequency detector and a sample tube.
32


3. Results and Discussions
3.1 Synthesis of the Methyl Vinyl Ketone and
Methacrolein Derivatives
Both derivatives were bright-orange in color and were about one minute
apart in terms of HPLC retention times. They also have very similar
ultraviolet visible spectra. These similarities are not surprising because
the two compounds are structural isomers. Figure 3.1 shows the UV
spectrum and the HPLC chromatogram for methyl vinyl ketone DNP
hydrazone; and figure 3.2 shows the UV spectrum and the HPLC
chromatogram for the methacrolein DNP- hydrazone derivative. Table
3.1 shows the melting points of the derivatives synthesized in the lab,
together with the literature values (21). The melting range for the
synthesized MTA-DNP hydrazone (205-206 C) compares very well with
the literature value of 206 C; which suggests that the compound is pure.
Table 3.1. The melting points of the DNP hydrazones
Derivative Melting Point (Range) (Degrees Celsius) Literature Value (Degrees Celsius)
MVK-DNP hydrazone 145-147
MTA-DNP Hydrazone 205 206 206
The MVK DNP hydrazone has a melting range of 2 degrees Celsius.
The methyl vinyl ketone used in the lab had a lot of impurities. All but one
33


impurity was removed by the purification method employed
(recrystallization). The impurity peak had a retention time about 0.6
minutes longer than the methyl vinyl ketone DNPH peak. From
experience of working with the two derivatives, MVK DNP hydrazone
seems to be more soluble in acetonitrile than the MTA derivative. This
made the recrystallization of MVK -DNP hydrazone from acetonitrile a
relatively difficult task. Recrystallization of this compound from ethyl
acetate proved to be more productive. Table 3.2 shows the standard
calibration data for MTA DNP hydrazone. The r2 value of the data
(regression output) is 0.995, indicating a good correlation of the standards
concentration values and the peak response values. The curve is shown in
figure 3.3 and as can be seen, it is very linear. The regression equation for
the data is:
Y = 0.00488*C + 0.0645
34


Figure 3.1. UV-Visible Spectrum and HPLC
Chromatogram for MVK-DNP Hydrazone.
35


Figure 3.2. UV-Visible Spectrum and HPLC
Chromatogram for MTA-DNP Hydrazone
36


Table 3.2. External Standard Data for MTA DNP
hydrazone.
Standard Concentration (ppbv) Peak Response Regression Equation Concentrations (ppbv)
0.50 98.37 0.545
1.01 186.57 0.975
2.01 413.07 2.081
5.03 988.5 4.891
10.05 2056 10.103
37


Peak Response (au)
Figure 3.3 Standard Calibration Curve for
MTA-DNPH hydrazone
38


Table 3.3 shows the standard calibration curve for the MVK- DNP -
hydrazones. The regression equation is Y = 0.007464*C + 0.995 and the r2
value was 0.9995.
Table 3.3. Standard Calibration Data for MVK DNP
hydrazone.
Standard- Concentration (ppbv) Peak Response Regression Concentrations (ppbv)
0.49 66.22 0.59
0.97 121.72 1.01
1.95 236.52 1.87
4.86 660.51 5.02
9.73 1327 10.00
The graph (figure 3.4) is very linear, showing the proportionality of the
relationship between the standards concentration and the peak responses.
It also shows the good accuracy with which the standards were prepared.
The x-coefficient of the MTA-DNPH is 0.00488 while that of MVK-DNPH is
0.007464, suggesting an apparent stronger absorption by the MVK
derivative.
39


Standard Concentration (ppbv)
Peak Response
Figure 3.4 Standard Calibration Curve for
MVK-DNPH
40


3.2 HPLC Analysis of the Field Samples
3.2.1 The Boulder-Marine Site
Figure 3.5 shows a sample of chromatograms for the Boulder site
samples. Table 3.4 shows the peaks which were positively identified by
virtue of their HPLC retention times. It should be noted that the retention
times for the same peaks may differ from one injection to another, due to
slight variations in the chromatographic responses, and also in the
integration variability of the computer employed. The chromatograms are
arranged in an alternating daytime -nighttime order (the first being a
daytime sample).
Table 3.4. The Peaks and their Retention Times,
identified at the Boulder (Marine) Site.
Peak Retention- Time Identification Daytime Average Conc/ppbv Nighttime Average Conc/ppbv
5.750 Formaldehyde 6.61 2.20
7.385 Acetaldehyde 15.51 2.96
9.862 Acetone 28.89 8.26
10.840 Propionaldehyde - -
15.272 Methyl Ethyl Ketone
41


Figure 3.5 Sample of Chromatograms for the
Boulder-Marine Site
42


The propionaldehyde and the methyl ethyl ketone peaks were not
quantified due to the unavailability of the quantified standards. Judging
from the relative peak responses of these compounds, it can be deduced
that their ambient concentrations are much smaller than those of
formaldehyde, acetaldehyde and acetone. In addition, the peak responses
indicate that the daytime concentrations of these compounds are larger
than those of the nighttime. In addition, there were a number of other
peaks that although were not positively identified, were consistently
detected at this Boulder site. Table 3.5 shows the unidentified peaks which
were consistently detected at this site. It should be noted that in all
chromatographic runs, the retention times for the same peaks may differ
from one run to another, for the same reasons as mentioned above. These
unidentified peaks shows that there are some other carbonyls (other than
those positively identified) in the above site. Their relatively small peak
responses however, suggest that they exist at minute concentrations. An
important general observation at this site (and indeed at all the other sites)
is that the daytime responses for the carbonyls detected is larger than the
nighttime ones. This may suggest that the source of these compounds is
more efficient during the day; or that there is an additional source (most
probably of photochemical nature) during the daytime periods.
3.2.2 The West Boulder Site
Figure 3.6 shows a sample of the HPLC chromatograms of the samples
collected at the Boulder -West sampling site. In addition, table 3.6 shows
the peaks and their retention times (both the identified and the
unidentified). This West-Boulder site had fewer unidentified peaks than the
Boulder-Marine site. Although it had occasional peaks beyond the
43


18.000 minutes region, this region was generally clean, in most samples.
Propionaldehyde was identified at this site; however, methyl ethyl ketone
was not.
Table 3.5 The Unidentified Peaks at the
Boulder Site.
Peak Retention Time Identification Status
11.552 Unidentified
14.685 Unidentified
15.743 Unidentified
18.521 Unidentified
22.447 Unidentified
27.515 Unidentified
31.871 Unidentified
32.366 Unidentified
44


Figure 3.6 Sample of Chromatograms for the
West-Boulder Site
45


Table 3.6 The Peak Retention Times and their
Identification Status, at the West-Boulder Site.
Peak Retention Time Peak Identity Daytime Average Conc/ppbv Nighttime Average Conc/ppbv
6.141 Formaldehyde 4.06 1.76
7.890 Acetaldehyde 1.97 1.10
10.925 Acetone 3.30 1.54
11.975 Propionaldehyde - -
15.396 Unidentified - -
17.057 Unidentified - -
17.665 Unidentified - -
46


3.2.3 Barr-Lake Sampling Site
Figure 3.7 shows a sample of the HPLC chromatograms for the Barr-Lake
site. Table 3.7 shows the analysis of these chromatograms. The peaks are
arranged in the order of increasing retention times and their identity (where
applicable) written down. In addition to the identified carbonyl compounds,
there were traces of an occassional peak at about 33.300 to 34.000
minutes. As in the other sites, the formaldehyde, acetaldehyde and
acetone responses are larger during the daytime than during the nighttime,
suggesting increased production of these compounds during the day. This
site also shows a variety of unidentified carbonyls between 16.00 and 30
minutes, indicating the presence of high molecular weight compounds in
that area, though in very small concentrations. Of significance notice is the
peak at 29.804 minutes, due to its consistently relatively large size at the
Barr-lake site. It suggests the presence (and in significant concentrations)
of a high molecular weight carbonyl compound at this site.
3.2.4. Auraria Site
Figure 3.8 shows samples of chromatograms for alternating
daytime/nighttime samples for the Auraria site.Table 3.8 stipulates the
peak retention times and their identity where applicable. The identities
were based on standard retention times and/or UV-visible spectra
matches. At this site, the daytime responses for formaldehyde,
acetaldehyde, acetone and propionaldehyde are clearly larger than those
for the nighttime, implying increased production of these compounds
during the day.
47


Figure 3.7 Sample of Chromatograms for the
Barr-Lake Site
48


Table 3.7 Peak Retention Times, and their
identity Barr-Lake Site.
Peak Retention Time Peak Identity Daytime Average Conc/ppbv Nighttime Average Conc/ppbv
5.874 Formaldehyde 9.80 3.35
7.580 Acetaldehyde 6.38 2.43
10.168 Acetone 6.64 2.89
11.212 Propionaldehyde Daytime response larger
14.082 Unidentified No significant difference
15.863 Methyl Ethyl ketone No significant difference
16.398 Unidentified No significant difference
21.500 Unidentified No significant difference
23.428 Unidentified No significant difference
24.855 Unidentified No significant difference
29.804 Unidentified Daytime response larger
49


Figure 3.8 Sample of Chromatograms for the
Auraria Site
50


Table 3.8 Peak Retention Times/ldentities
- Auraria Sampling Site.
Peak Retention Time (Minutes) Peak Identity Daytime Average Conc/ppbv Nighttime Average Conc/ppbv
5.601 Formaldehyde 7.13 4.86
7.150 Acetaldehyde 3.89 2.78
9.485 Acetone 2.19 1.88
10.390 Propionaldehyde - -
14.550 Unidentified - -
14.964 Methyl Ethyl Ketone
22.351 Unidentified - -
26.620 Unidentified
29.542 Unidentified - -
51


3.2.5 Golden (NREL) Site
Figure 3.9 shows a sample of the chromatograms obtained from the
analysis of the carbonyl samples collected at the National Renewable
Energy Lab in Golden. Table 3.9 shows the retention times and the identity
status of the peaks which were consistently detected during the numerous
HPLC runs of the samples obtained at this site.
Table 3.9. Peak Retention times and Peak Identities
- Golden Site.
Retention Time (Minutes) Peak Identity Daytime Average Conc/ppbv Nighttime Average Cone (PPbv)
5.746 Formaldehyde 5.15 3.40
6.530 Formaldehyde Pre-peak (unidentified)
7.374 Acetaldehyde 9.84 3.65
9.830 Acetone 15.03 8.24
10.811 Propionaldehyde - -
15.198 Methyl Ethyl Ketone
15.690 Unidentified - -
29.804 Unidentified - -
52


Figure 3.9 Sample of Chromatograms for the
Golden Site
53
122.212


The results for this Golden site indicate that the daytime response for
formaldehyde, acetaldehyde, acetone and propionaldehyde is larger than
the nighttime one, which shows that the daytime production of the above
compounds is more. As previously suggested, this could be due to the
effect of photochemical production of these compounds. Apart from
formaldehyde, acetaldehyde and acetone, the other carbonyls which were
positively identified using retention times match were propionaldehyde and
methyl ethyl ketone. Table 3.10 shows the summary of the daytime and
the nighttime concentrations of the quantified carbonyls, at all the 5 sites.
The values are average (where number of observations is 5) ppbv
concentrations. The observations which can be made from the table are
that the daytime average concentrations for these compounds are
predominantly larger than the nighttime ones. This could be attributed to
the net photochemical production of these compounds, as has previously
been mentioned. The acetone concentrations at the outlying areas is much
higher than that at the Auraria site, which is in agreement with previous
studies (1). In particular, for this study, the Boulder-Marine has a daytime
acetone concentration of over 28 ppbv and NREL has over 15 ppbv,
which are very large compared to 2.187 ppbv for Auraria site. Moreover, all
the outlying sites but West-Boulder shows higher acetaldehyde
concentrations compared to the Auraria site.
54


Table 3.10 Comparison of the Daytime and the
Nighttime Carbonyl Concentrations at the Various
Sites Studied.
Site Formaldehyde (Average Cone.) Acetaldehyde (Average Cone.) Acetone (Average Cone.)
Night (ppbv) Day (ppbv) Night (ppbv) Day (ppbv) Night (ppbv) Day (ppbv)
Auraria (current) 4.86 7.13 2.78 3.89 . 1.88 2.19
Auraria previous 3.28 3.46 1.48 1.96 1.01 1.55
Boulder- Marine (current) 2.20 6.61 2.96 15.51 8.26 28.89
Boulder- Marine previous 2.02 4.16 2.07 7.92 3.74 10.83
Barr- Lake (current) 3.35 9.80 2.43 6.38 2.89 6.65
Barr- Lake Previous 1.24 1.24 1.98
West- Boulder /Current 1.76 4.06 1.10 1.97 2.73 3.30
West- Boulder Previous
NREL (Current) 3.40 5.15 3.65 9.84 8.24 15.03
NREL Previous - - -
55


The median carbonyl concentrations from the previous studies (included in
table 10 above) were obtained from a study by Susan Riggs (1), in the
summer of 1996, and are included for comparison purposes. Although in
general the overall carbonyl concentrations from the current study are
higher, both studies point out that the nighttime carbonyl concentrations
are smaller than the daytime ones. In addition, both studies indicates that
the outlying areas generally have higher acetaldehyde and acetone
concentrations. This is in agreement with other studies done in the past, as
outlined in the purpose of the study section.
3.2.6 Methacrolein and Methyl Vinyl Ketone at
the Boulder-Marine, West Boulder, Golden,
Barr-Lake and the Auraria Sites
Methacrolein and methyl vinyl ketone derivatives of 2,4-DNPH were
synthesized in the lab and used in the analysis of the samples from all the
sites to look for these biogenic-derived carbonyl compounds. Based on the
results from matching the retention times and the spectra of the sample
peaks to those of the derivatives, these compounds were not conclusively
identified at any of the sites. There were occasional small peaks around
the area of interest, but these could not be conclusively attributed to the
existence of either methyl vinyl ketone or methacrolein. It should be
mentioned that if the compounds actually exist, but were not detected due
to any chromatographic or detector limitations (for example detection limit),
then those compounds probably exist in very minute quantities,
a, p-unsaturated aldehydes and ketones are expected to react with OH,
NO3 and O3 to form other compounds. It is thus a possibility that MVK and
MTA (which fall in the above category), oxidize at a rate which
substantially reduce their ambient concentrations.
56


For a second order reaction between an organic compound G and an
oxidizing species O, the reaction can be represented as:
G + O k0 ^ Products, where k0 is the rate constant of the reaction.
If the tropospheric concentration of the oxidizing compound, [O], and k0are
known, then the tropospheric lifetime (x) of the hydrocarbon can be
calculated as:
x= 1/(k0 [O])
The rate constants for the reaction of MVK and MTA with OH, in
cm3 molecule'1 s'1 are 1.9 x 10'11 and 3.3 x 10"11, respectively (11). Using
the typical daylight OH concentration of 1 x 106 molecules cm'3 (13), the
tropospheric lifetimes of MVK and MTA with respect to reaction with OH
are 14.8 hours (0.6 days) and 8.4 hours (0.4 days), respectively. For
comparison purposes, other values are 1.3 days for formaldehyde, 0.7
days for acetaldehyde, 50 days for acetone, 0.6 days for acrolein
(CH2=CHCHO) and 0.4 days for 1-butene. Although the tropospheric
lifetimes of MTA and MVK due to the reaction of these compounds with
NO3 and O3 are not presented here, those of formaldehyde, acetaldehyde
and 1-butene are 2.7 months, 0.6 months, and 4.8 days, respectively for
the NO3 reaction; and 21 hours and 13.8 days for the reaction of 1-butene
and acrolein with O3 (O3 is not expected to react with formaldehyde,
acetaldehyde or acetone, because they are saturated hydrocarbons). The
typical concentration of 03 used in these calculations was 50 ppb
(1.2 x 1012 cm3 molecules'1 s'1), and that of the N03 was ~ 10 ppt
(2.5 x 10 8 cm3 molecule1 s'1) (11). From the above lifetime values, it can
be seen that the hydrocarbon OH reaction is the most important (gives
shortest lifetimes) compared to that of the reactions involving O3 or NO3.
The lifetimes of 1-butene (due to its oxidation) are probably an
57


underestimate with respect to those of MTA and MVK; and those of
formaldehyde and acetaldehyde, an overestimate. The lifetimes of acrolein
(with respect to its oxidation by ozone) should compare well with those of
MVK and MTA (the three compounds are all a,p-unsaturated carbonyl
compounds). MVK and MTA being branched hydrocarbons, are however
expected to react faster than acrolein, with concomitant shorter
tropospheric lifetimes. From the preceding discussion, the lifetimes of MTA
and MVK (due to their reaction with NOs), are expected to be longer than 4
days (1-butene) and shorter than 2.7 months (formaldehyde). Moreover,
the tropospheric lifetimes of MVK and MTA with respect to their reaction
with 03, may be expected to be longer than 21 days (1-butene) but shorter
than 13.8 days (acrolein). The lifetimes of MVK and MTA due to the
reaction of these compounds with OH, though short, are comparable to
those of formaldehyde, acetaldehyde and acrolein, an a,p-unsaturated
carbonyl compound. A reasonable conclusion to explain why MTA and
MVK were not detected in this study is thus that the source of these
compounds (isoprene) is probably available in small concentrations,
leading to these compounds being formed in minute quantities in the
troposphere.
3.2.7 Comparison of the Boulder-Marine,
West Boulder, Barr-Lake, Golden and
the Auraria Sites
Table 3.11 shows the comparison of the above studied sites, in terms of
which peaks were detected.
58


3.2.8 Detection Limit Measurements
Table 3.12 shows the peak responses of the 10 replicate injections of the
1 ppbv standard of the MTA-DNP hydrazone. The detection limit for MTA-
DNP hydrazone, calculated from these data was found to be 0.36 ppbv.
A detailed account of the calculations involved is outlined in the appendix
(Appendix B). Table 3.13 shows the peak responses of the 11 replicate
injections of the 2 ppbv standard of the MVK-DNP hydrazone. The
standard deviation in the peak responses was 5.49 and the detection limit
was found to be 0.42 ppbv. The details of the calculations are found in the
appendix section (Appendix B).
59


Table 3.11. The Peaks Identified at the
Sites Studied
Carbonyl/Peak Site
Auraria Boulder- West- Barr-Lake Golden
Site Marine Boulder Site Site
Formaldehyde X X X X X
Acetaldehyde X X X X X
Acetone X X X X X
Propion- aldehyde X X X X X
Methyl ethyl- Ketone X X X X
Methyl Vinyl- Ketone
Methacrolein - - - - -
14.647 minutes X X X X
16.398 minutes X X X X X
24.867 minutes X X X X
60


Table 3.12. Peak areas of the 10 replicate injections
of the MTA DNP hydrazone standard.
Injection Number Peak Response
1 199.86
2 191.83
3 197.05
4 201.16
5 201.68
6 198.13
7 199.77
8 191.77
9 199.58
10 200.02
61


Table 3,13. Peak responses for 12 replicate
injections of a 2 ppbv MVK-DNP standard.
Injection Number Peak Response
1 239.57
2 243.94
3 249.55
4 242.08
5 256.06
6 245.32
7 239.93
8 245.52
9 254.97
10 249.62
11 252.42
62


3.2.9 Quality Assurance Results
Table 3.14. The results from the MTA-DNPH
permeation system.
Sample Number Volume of Air Sampled Peak Response Peak Response Corrected for Air Volume
1 83.57 838.43 8981.36
2 64.73 652.54 9024.59
3 76.88 772.13 8990.90
4 69.89 709.38 9086.36
5 60.92 613.18 9010.61
6 64.09 647.77 9048.08
The relative standard deviation in the peak responses for the MTA-DNPH
permeation-tube experiment was found to be 35.5 and the relative
standard deviation was 0.393 %. The relevant calculations can be found in
the appendix section (Appendix C). The standard deviation in the peak
responses for the MVK-DNPH permeation-tube experiment was found to
be 290.38 and the relative standard deviation was 1.24 %. The details of
the calculations can be found in the appendix section (Appendix C).
63


Table 3.15 Peak Responses from the MVK-DNPH
Permeation Tube System
Sample Number Air Volume Sampled Peak Response Volume- Corrected Peak Response
1 115.33 2968.98 23045.75
2 89.23 2298.98 23064.83
3 57.96 1542.98 23831.85
4 92.19 2423.98 23538.08
5 56.1 1472.98 23504.98
6 52.51 1386.98 23645.80
64


3.2.10 2,4-DNPH Coated Cartridge Carbonyl
Collection Efficiency
3.2.10.1 MTA-DNP hydrazone Collection
Efficiency
Table 3.16 shows the peak responses of both the front cartridge and the
backup cartridge.
Table 3.16. Cartridge Collection Efficiency Results,
for MTA-DNP hydrazone.
Sample Front Cartridge (Peak Response) Back-Up Cartridge (Peak Response) Efficiency (%)
1 838.43 0.00 100
2 652.54 0.00 100
3 772.13 0.00 100
4 709.38 0.00 100
5 613.18 0.00 100
6 647.77 0.00 100
3.2.10.2. Cartridge Collection Efficiency for
MVK-DNP Hydrazone
Table 3.17 shows the results for the coated cartridge collection efficiency
for the MVK derivative.
65


3.2.11 Cartridge Extraction Efficiency
3.2.11.1 Extraction Efficiency for MTA-DNP
Hydrazone
The results of how well the sampled MTA-DNP hydrazone cartridge was
extracted are summarized in table 3.18 in the following page.
Table 3.17. Cartridge Collection Efficiency for
MVK-DNP hydrazone
Sample Front Cartridge (Peak Response) Back-Up Cartridge (Peak Response) Collection Efficiency
1 2968.98 0.00 100
2 2298.98 0.00 100
3 1542.98 0.00 100
4 2423.98 0.00 100
5 1472.98 0.00 100
6 1386.98 0.00 100
66


Table 3.18 MTA Cartridge Extraction Efficiency
Results
Sample Front Cartridge (Peak Response) Back-UP Cartridge (Peak Response) Extraction- Efficiency (%) [Front/(Front + Back-up)] 100%
1 838.43 25.83 97.01
2 652.54 0.00 100
3 772.13 3.03 99.61
4 709.38 5.27 99.26
5 613.18 5.23 99.15
6 647.77 5.90 99.10
67


From the table 3.18, the mean extraction efficiency of the MTA-DNP
hydrazone coated cartridge is 99.02 %. The standard deviation in the
individual sample efficiencies is 0.95 %.
Thus, the 95 % confidence limits (CL) in the extraction efficiencies is
computed as follows:
CL = Mean (t a)/VN
Where mean refers to the mean extraction efficiency
a = standard deviation
N = sample size (Number of observations)
t = statistical parameter, defined as (x p)/s (27).
where x = an individual observed value, s the sample standard deviation
and p is the population mean. This statistical value is 2.57 for 5 degrees of
freedom (27).
CL = Mean (2.57 0.95)/2.45
CL = 99.02 0.99 %
3.2.11.2 Cartridge Extraction efficiency for
MVK-DNP Hydrazone
The results are outlined in table 3.19. From the table, the mean extraction
efficiency for the MVK-DNP hydrazone sampled cartridge is 99.57 %. In
addition, the standard deviation is 0.37 %. Thus, the 95 % confidence
limits (CL) in the extraction efficiency is:
CL = Mean (2.57 0.37)/ 2.4495
CL = 99.57 0.39 %
68


Table 3.19. Cartridge Extraction Efficiency for
MVK-DNP hydrazone.
Sample Front Cartridge (Peak Response) Back-Up Cartridge (Peak Response) Extraction- Efficiency (%)
1 2968.98 9.20 99.69
2 2298.98 28.21 98.79
3 1542.98 2.05 99.87
4 2423.98 3.99 99.84
5 1472.98 8.24 99.44
6 1386.98 3.19 99.77
3.2.12 The HPLC Precision Measurements
The following table depict the instrument precision calculation results. The
data in table 3.20 show that the relative standard deviation in the peak
areas for the replicate injections of the formaldehyde-DNPH standards
ranged from 0.24 % 2.9 %. The relative standard deviations in the
acetaldehyde-DNPH peak responses are shown in table 3.21. These
ranges from 0.28% 2.0 %.
Table 3.22 shows the relative standard deviations in the peak responses of
the acetone-DNPH standards. They range from 0.58 % 3.6 %.
69


able 3.20. Results of the HPLC Precision
:alculations. (Formaldehyde-DNPH)
Std Cone (ppbv) Peak Area (au) Peak Area (au) Peak Area (au) Mean Peak- area Std Devi ation Relative Std Deviation (%)
0.93 63.13 62.22 59.8 61.72 1.41 2.3
2.31 153.65 155 145.04 151.23 4.41 2.9
4.62 302.97 302.96 296.29 300.74 3.15 1.0
8.34 552.49 554.9 551.73 553.04 1.35 0.24
13.86 900.41 909.86 890.57 900.28 7.88 0.88
[able 3.21. HPLC Precision Results
[Acetaldehyde-DNPH)
Cone of Std (ppbv) Peak- Area (au) Peak- Area (au) Peak Area (au) Mean Peak- Area Std Dev. Relative Std dev. (%)
0.96 69.25 70.69 68.74 69.56 0.83 1.2
2.40 170.3 171 173.66 171.65 1.45 0.84
4.80 331.67 345.63 346.58 341.29 6.82 2.0
8.66 610.3 622.65 617.49 616.81 5.06 0.82
14.4 1011 1015 1018 1014.67 2.87 0.28
70


Table 3.22. Results for HPLC Precision
Calculation (Acetone-DNPH)
Cone, of Std (ppbv) Peak- Area (au) Peak- Area (au) Peak- Area (au) Mean Peak- Area Standard- deviation Relative Std Dev. (%)
0.78 54.79 52.29 57.14 54.74 1.98 3.6
1.92 136.31 135.66 134.16 135.38 0.90 0.66
3.85 264.31 261.86 268.12 264.76 2.58 0.97
6.95 484.93 489.29 478.37 484.20 4.49 0.93
11.55 803.24 792.32 795.31 796.96 4.61 0.58
71


3.2.13 Retention Times Reproducibility
The following table shows the relative standard deviations of the retention
times of the identified compounds. For each injection of the sample, three
chromatograms were obtained (360, 376 and 380 nm). The relative
standard deviations in the averages of these chromatograms were
calculated.
Table 3.23. Retention Times Reproducibility.
RELATIVE STANDARD DEVIATIONS (%)
Site Form. Aceta. Aceto. Propio. MEK
Auraria 0.44 0.47 0.54 0.54 0.63
Boulder 0.10 0.13 0.12 0.15 0.11
W. Boulder 0.33 0.43 0.35 0.36 0.42
NREL 0.25 0.26 0.27 0.22 0.24
Barr-Lake 0.12 0.16 0.24 0.35 0.22
As displayed in the above table, the relative standard deviations in the
retention times were below 1.00 %, at all sites. The relatively low relative
standard deviations reflect a good reproducibility of the retention times.
The reproducibility (precision) of retention times is important since the
parameters are used for diagnostic purposes.The data used to compute
the above relative standard deviations can be found in the appendix
section (appendix D).
72


3.2.14 Mass Spectrometry and 1H-NMR
Spectrometry of the Hydrazones
Figure 3.10 and 3.11 shows the mass spectra of the MTA and MVK DNP
hydrazones, respectively. These compounds are CioH10N404 structural
isomers, with a molecular weight of 250 amu. Table 3.24 shows the
summary of the abundant peaks observable in the spectra. Both isomers
shows a molecular ion peak at 250 amu, which is consistent with the
molecular weight of the compounds. The 203 peak can be attributed to the
loss of a proton and a -NO2 group from the aromatic ring, and the 202
peak to the loss of an N02 group and a two protons. The 54 peak, which
shows as the base peak for the MTA isomer is probably due to the
CH2=CHC=N + a proton. The 53 peak may be attributable to the
CH2=CHC=N fragment in the case of the MTA derivative and CH2=CCH=N
for the MVK derivative. The mass spectrometer library search matched the
two compounds to 2-butenal, (2,4-dinitrophenylhydrazone) which is a
structural isomer of the two compounds. The mass spectrum of this
compound is shown in figure 3.12. The spectrum shows a fragmentation
pattern similar to the MTA and MVK derivatives. For example, it shows the
250 molecular ion peak, the 202/203 peaks, the 156 peak and the 53/54
peaks. Figures 3.13, 3.14 and 3.15 shows the literature mass spectra of
2.4- Dinitrophenylhydrazine (the derivatizing reagent used), formaldehyde-
2.4- dinitrophenylhydrazone (the simplest carbonyl DNPH derivative), and
that of acetaldehyde-2,4-dinitrophenylhydrazone, respectively (28).
73


Figure 3.10 Mass Spectrum of the
MTA-DNPH Standard


Figure 3.11 Mass Spectrum of
MVK-DNPH Standard


Table 3.24. Summary of the Mass Spectra of
MVK and MTA DNP Hydrazones
MTA- DNPH (Peak) Fragment MVK- DNPH (Peak) Fragment
251 MH
250 Molecular Ion (M) 250 (Base) Molecular Ion
203 (M-H) N02 203 (M-H) N02
202 (M-2H) -N02 202 (M-2H)-N02
79 Ring fragmentations & Rearrangements 120 Ring fragmentations & Rearrangements
63 Ring fragmentations & Rearrangements 63 Ring fragmentations & Rearrangements
56 54 CH2=CHC=N + H
55 CH2=C-CH=N + 2H 53 ch2=chc=n
54 (Base) CH2=CCH=N + H
53 ch2=cch=n
76


Figure 3.12 Mass Spectrum of
2-Butenal-DNPH


#33469: Hydrazine. (2.4-dlnltrophenyl)- SCALED
Abundance
Figure 3.13 Mass Spectrum
of 2,4-Dinitrophenylhydrazine


#39127: Formaldehyde, (2.4-dlnltrophenyl)hydrazone SCALED
Abundance
Figure 3.14 Mass Spectrum of
Formaldehyde-DNPH


Figure 3.15 Mass Spectrum of
Acetaldehyde-DNPH


The spectra are included for comparison purposes. In particular, all the
above compounds (including the MTA and the MVK derivatives) yield
peaks at 63 and 79 amu, indicating that the peaks are probably due to the
ring fragmentations and/or rearrangements. The structures of the MVK and
the MTA DNPH derivatives are shown in figures 3.16 and 3.17,
respectively. Figures 3.18 and 3.19 shows the 1H-NMR spectra of the
MVK-DNPH and the MTA-DNPH, respectively. Table 3.25 shows the
interpretation of the two spectra. Both derivatives shows singlets at
5 = 11.2 and 9.1, and two doublets at 5 = 8.3 and 8.0. Both observations
are consistent with the protons in the DNPH part of the derivatives as
mentioned in table 3.25. The peak at 5 = 7.3 is the solvent peak. Both
derivatives also shows a 8 = 2 prominent peak, attributable to the methyl
group observed in both compounds. In addition to the aforementioned
peaks, the MTA derivative shows a 5 = 9.7 and 8.0 doublets, while the
MVK derivative shows two quartets at chemical shift values of 5.6 and 6.6.
These observations are consistent with the difference in the proton
distribution of the two compounds, as indicated in table 3.25.
81


Table 3.25.1H-NMR Results of MTA-DNPH and
MVK -DNPH
Peak (Chemical Shift) MTA-DNPH MVK-DNPH Comments
11.2 (Singlet) N-H N-H -
9.1 (Singlet) Ar-H Ar-H The ring proton between the two N02 groups.
8.3 (Doublet) Ar-H Ar-H Splitting by the two non-identical neighboring protons on the ring, (a "doublet of doublets")
8.0 (Doublet) Ar-H Ar-H Splitting by the two non-identical neighboring protons on the ring, (a "doublet of doublets")
7.8 (Singlet) -CH- - -
7.3 (Singlet) Solvent Solvent Deuterochlorofor m (CDCI3)
6.6 (Quartet) - ch2=ch- Splitting (Non-identical)
5.6 (Quartet) - ch2-ch- Splitting (Non-identical)
2.2 (Singlet) -ch3 -ch3 3-identical protons
82


Figure 3.16. Methyl Vinyl Ketone (2,4 diphenylhydrazone)
83


Figure 3.17 Methacrolein (2,4 diphenylhydrazone)
CH2
84


II
Figure 3.18 H-NMR Spectrum of MVK-DNPH


00
O)
Figure 3.19 1H-NMR Spectrum of MTA-DNPH


4. Conclusions
The following conclusions have been drawn from the findings of the study:
Methyl vinyl ketone and methacrolein were not conclusively identified at
any of the sites studied (Denver at Auraria, Boulder at Boulder-Marine,
West-Boulder, and Golden at NREL).
Formaldehyde, acetaldehyde and acetone are the dominant (in terms
of concentrations) carbonyls detected at these sites.
Propionaldehyde [CH3CH3CHO] and methyl ethyl ketone
[CH3C(0)CH2CH3] have been identified consistently at four of the five
sites.
There are other unidentified (but detected) higher molecular carbonyl
compounds at the above sites.
Boulder, NREL and Barr Lake sites have more acetaldehyde and
acetone than the Denver site. Moreover, from the peak response
measurements, these sites appears to have more propionaldehyde and
methyl ethyl ketone than the Denver site. The preceding conclusions
leave a few points to consider further. If methyl vinyl ketone and
methacrolein were not conclusively identified at these sites, does that
mean there is no biogenic contribution at these sites? If there is no
biogenic carbonyl contribution to the carbonyl compounds in these
areas, then why are ...
acetaldehyde and acetone concentrations higher at the outlying sites than
in Denver? From these study, and also based on information from some
previous studies as cited in the introduction, the following points can be
87


asserted: The apparent higher carbonyl concentrations at the outlying sites
is:
Not due to local anthropogenic emissions (the car density in these
areas is much smaller than the one in Denver).
Not due to isoprene emissions (methyl vinyl ketone and methacrolein,
which are major products of isoprene oxidation, at least in theory, were
not identified or were too small to be identified at these areas).
Thus, in the midst of the uncertainties, errors and instrumental limitations
talked about in the results and discussions part, my final conclusions are
that the higher carbonyl concentrations in the outlying areas:
Might be due to the chemistry of terpenes.
Might be due to the chemistry of other biogenic compounds.
Might be due to the chemistry of anthropogenic hydrocarbons.
Might be due to the transportation of these compounds to the outlying
areas, via the air mass drift.
Might be due to primary biogenic emissions; for example, microbial
activities.
Results analysis without a description of possible errors and uncertainties
maybe incomplete. This is so because total control of experimental
variables is very difficult, if not impossible. Every step involved in analysis
is a potential source of error: In this work the potential sources of error are
the sampling methods, the analyses techniques, the instrumental response
and indeed the data handling. There are generally two types of errors: The
systematic (determinate) and the random (indeterminate) errors (21).
The effect of systematic errors is to deviate the results from the true value
in a consistent manner. The following are therefore, sources of systematic
errors : Insufficient purity of reagent, improper operation of the
88