 

a... 3..-
. ‘ .. 2:}.
, ‘ .

I. .1
:J

.r. . A
main? is? .

L1.
3 l L
n. n :v .. I
. a; tunin... ‘

 

 

(uh. :Ew
L: H”; K.

 

 

i 9:“;

u .. yo. I

driftsuuunﬂu.

2- .33.

r: . 1

.513. ..
.1

1+...
.53. 2..
.. 6.4.9... :u

31R:

3 , . . . . . . . .
. . I!!! .03 vat/lung”.
‘ «nib. . .Hhu...
ii . .. 2.5L
3;... . 1 . I. i.

if:
a

'17

v 11 ._

g '12;
r.

as .77..) x4 I)»; .
ain‘tvv. vpluvllvn

2

‘2 .l- 751:. . .. . . . L . \ K
at...“ 3!. s. . . . . . .- Ll £36 . .ﬁm

. .. , . , .2... . . _ .1535... L..§=,_~.wgéu\._=a.:. .§L_?_!é.a a; um... ﬂagshiﬁgﬁné .. . . 1.5M...
. .I. In). . 2:“. I. Ind-r; ._ . : 1’ ' 2%? . .. . A: .1? u: ﬁx .i ...v.. . . . .. . . . . S v.\ 5..

”It

 

MW?”

LIBRARIES
MICHIGAN STATE UNIVERSITY
EAST LANSING, MICI-I 48824-1048

This is to certify that the
thesis entitled

METHOD DEVELOPMENT OF AN ADAPTIVE AIR
SAMPLING DEVICE FOR USE WITH PORTABLE GAS
CHROMATOGRAPHY IN FIELD FORENSIC ANALYSES

presented by

Scott Alan Ramsey

has been accepted towards fulfillment
of the requirements for the

degree In Forensic Chemistry

 

 

wﬁuj

(/Major Pr6fe or’ s Signature
// No o‘/

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity Institution

-—..—_-—.-_-<-.. -

|
1'

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c1/CIRC/DateDuep65-p. 15

 

 

METHOD DEVELOPMENT OF AN ADAPTIVE AIR SAMPLING DEVICE
FOR USE WITH PORTABLE GAS CHROMATOGRAPHY IN FIELD
FORENSIC ANALYSES
By

Scott Alan Ramsey

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE
School of Criminal Justice

2004

ABSTRACT
METHOD DEVELOPMENT OF AN ADAPTIVE AIR SAMPLING DEVICE FOR
USE WITH PORTABLE GAS CHROMATOGRAPHY IN FIELD
FORENSIC ANALYSES
By

Scott Alan Ramsey

The RVM Adaptive Air Sampler injection system, developed by RVM
Scientiﬁc, Inc, has been evaluated as a potential improvement over conventional
air sampling systems. This prototypical system utilizes both solid sorbent
microtrap and solid phase microextraction (SPME) technology in a unique
design. The primary air concentrator is a self-heating, SPME-coated nickel alloy
wire contained in the lumen of two concentric tubes, creating a restrictive
airspace. The secondary concentration stage incorporates a microtrap
containing 10-mg of solid sorbent which provides rapid injection of the ensuing
vapor plug into a 60 for analyte separation and quantitation. The two-stage
sorbent design enables dynamic, high-ﬂow (4-5 L/min) air sampling which has
not previously been investigated with SPME technology. Laboratory analysis of

a 39-component, US. EPA compendium TO-14A gas mixture at concentrations

of 0.5-150 ppbv (where ppbv = 1 part in 109 by volume) by the RVM injection

system illustrated linearity with benchmark data in terms of sensitivity at low
ppt(v) levels, detection limits, and desorption efﬁciency. The RVM Adaptive Air
Sampler provided several advantages over a conventional air concentration
system in terms of sampling ﬂow rate, preconcentration time, and power

consumption.

To Tessa, my beautiful and loving wife

ACKNOWLEDGMENTS

I would like to thank my mentor, Brian A. Eckenrode, Ph.D. for support
and guidance he so graciously offered. Over the past year, Brian has provided
me with invaluable Ieaming experiences and helped to develop the
chromatographer that I am today. Without his help, this thesis would not have
been possible.

Many thanks are extended to the associates at the FBI Counterterrorism
Forensic Science Research Unit, and Oak Ridge Institute for Science and
Education (ORISE) for funding my internship and making this research
opportunity possible.

Much recognition is given to Robert V. Mustacich, Ph.D., of RVM
Scientiﬁc, Inc., for instrumentation development. Andrew Neushul and Jim
Everson, also of RVM Scientiﬁc, provided many hours of technical support, which
is deeply appreciated.

My appreciation is extended to Jay Siegel, Ph.D. and the Michigan State
University School of Criminal Justice for allowing me to be apart of a research
opportunity at the FBI Academy.

I would like to also thank David Foran, Ph.D. for providing a great referral
for the FBI internship, from which the research of this thesis arose.

I am eternally grateful for my wife, Tessa Ramsey, whose incredible

patience over the past three years has ﬁnally come to fruition. Thank you.

Finally, I would like to thank my Lord and Savior, Jesus Christ, for building
my faith in ways unimaginable during the journey along this path.

Psalms 37:4

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................... ix
LIST OF FIGURES ....................................................................................... x
CHAPTER 1 - INTRODUCTION .................................................................. 1
1.1 Purpose of Research ............................................................... 1
1.2 Solvent-free Sample Preparation ............................................ 5
1.3 Gas Phase Extraction .............................................................. 6
1.3.1 Headspace Extraction ................................................... 6
1.3.2 Supercritical Fluid Extraction ......................................... 8
1.4 Membrane Extraction .............................................................. 9
1.5 Sorbent Extraction ................................................................... 11
1.5.1 Solid Phase Extraction .................................................. 11
1.5.2 Solid Phase Microextraction .......................................... 17
1.6 Low Thermal Mass Gas Chromatography ............................... 21
1.7 References .............................................................................. 26
CHAPTER 2 - EXPERIMENTAL .................................................................. 29
2.1 Adaptive Air Sampler: HSA-SPME Element Design ............... 29
2.2 Adaptive Air Sampler: Focusing Preconcentrator Design ....... 34
2.3 Instrumentation ........................................................................ 39
2.3.1 Benchmark Air Concentrator ..................................... 39
2.3.2 Gas Chromatograph ................................................... 41
2.3.3 Mass Selective Detector ............................................ 42

vi

2.3.4 RVM HSA-SPME Element/Focusing

Preconcentrator Microtrap Assembly ........................ 42

2.3.5 GC Capillary Column ................................................ 45

2.3.6 Dynamic Gas Generator ............................................ 46

2.3.7 Sampling Pump ......................................................... 46

2.3.8 Digital Multimeter ....................................................... 47

2.4 Reagents ................................................................................. 48

2.5 Experimental Procedure .......................................................... 52

2.5.1 EPA Toxic Organics (TO-14A) Gas Mix .................... 52

2.6 References .............................................................................. 55

CHAPTER 3 — RESULTS AND DISCUSSION ............................................. 56

3.1 US. EPA Toxic Organics (TO-14A) Gas Mix .......................... 56

3.1.1 Baseline Gas Chromatograph .................................... 56

3.1.2 Experimental Desorption Time Proﬁles and

Gas Chromatograms ................................................. 58

3.1.3 Extraction Time Proﬁles ............................................. 70

3.1.4 Calibration Curves ..................................................... 72

3.1.5 Limits of Detection ..................................................... 81

3.1.6 Method Comparison .................................................. 85

3.1.7 Power Consumption .................................................. 88

3.2 References .............................................................................. 92

CHAPTER 4 - CONCLUSIONS ................................................................... 93

APPENDIX A ................................................................................................ 97
APPENDIX B ................................................................................................ 100

vii

BIBLIOGRAPHY ........................................................................................... 1 15

viii

Table 2.1

Table 2.2

Table 2.3

Table 2.4

Table 2.5

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Table 3.5

Table 3.6

Table 3.7

Table 3.8

LIST OF TABLES

Technical data for the Stablohm 675 nickel alloy wire
used in the HSA—SPME element ............................................. 30

Temperature parameters for the three trapping modules
in the Entech 7100 Air Concentrator ...................................... 40

Flow and volume parameters for trapping in the
Entech 7100 Air Concentrator ................................................. 40

EPA TO-14A, 39-Component Gas Mixture
(1 ppmv nominal each) ............................................................ 49

LTM-GC/MS method parameters for the TO-14A
experimental gas mix ............................................................... 54

Analytes exhibiting co-elution in the Entech baseline
Chromatograph ........................................................................ 58

Analytes exhibiting co-elution in the RVM experimental
Chromatograms ........................................................................ 67

Desorption efﬁciencies for baseline and experimental air
concentrators. 100 ppbv (nominal) TO-14A gas sampling
performed @ 0.2 L/min (Entech), 0.3 Umin (microtrap),

and 4 Umin (HSA-SPME element) .......................................... 68

Calibration curves for the Entech 7100 benchmark and
the HSA-SPME experimental air concentrators with

their respective correlation coefﬁcients (R2) ............................. 74

Calibration curves for the Focusing Preconcentrator 2
Microtrap air concentrator with correlation coefﬁcients (R )....77

Limits of Detection (LCD) for the benchmark and
experimental air concentrators of the TO-14A components... 83

Regression line parameters for the HSA—SPME
element versus benchmark cryotrap experiment ..................... 87

Power consumption proﬁles for the Entech 7100
benchmark and coupled RVM Adaptive Air Sampler/
LTM-GC systems ..................................................................... 90

Figure 1.1

Figure 1.2

Figure 1.3

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 2.7

Figure 2.8

Figure 2.9

Figure 2.10

LIST OF FIGURES

Images in this thesis are presented in color

An RVM LTM A68 retroﬁt GC oven door and LTM
GC column installed on an Agilent 6890 gas
Chromatograph ........................................................................ 22

RVM Low thermal mass GC design. R.V. Mustacich,
US Patent 6,209,386 (2001) .................................................... 23

Chromatographic comparison of a 15-component
alkane mix separated by conventional and low

thermal mass 60 capillary columns ........................................ 25
Cross—section of HSA-SPME element design ........................ 31
A prototype HSA-SPME Element ............................................ 32
Scanning electron microscopic image of a Carboxen/

PDMS HSA-SPME element at 1300X magniﬁcation ............... 33
A ﬁrst-generation focusing preconcentrator which

integrates the HSA—SPME element with the microtrap ............ 35
Flow path of an embedded microtrap GC system ................... 36

Flow path of a GC system with simple integration of
microtrap (labeled “adsorbent”) and air inlet ............................ 36

Close-up view of the O-ring clutch drive belt system
for the RVM focusing preconcentrator microtrap carriage ....... 37

Installation of the RVM Adaptive Air Sampler on the
LTM A68 retroﬁt oven door on the Agilent 6890 GC ............... 41

Installation of 1/16” 3-way tee; helium ﬂows to the
RVM Adaptive Air Sampler during focusing
preconcentrator injection ......................................................... 43

The Gilair-5® used in sampling Tedlar bags at 4 Umin
into the HSA-SPME element ................................................... 47

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

Figure 3.10

Figure 3.11

US. EPA Toxic Organics (T O-14A) volatiles gas mixture
baseline Chromatograph. See Table 2.4 for peak
identiﬁcation ............................................................................ 57

Diagram Showing a 10-second temperature ramp
typically observed during microtrap thermal desorption .......... 60

Overlayed Chromatograms of gas sampling by the
focusing preconcentrator microtrap and subsequent
canyover blank ........................................................................ 61

Diagram Showing both resistive and non-resistive
power ramps and respective temperature proﬁles .................. 63

Contrasting Chromatograms illustrating increased

retention times, peak tailing, and convolution in

non-optimal desorption conditions, vs. well-deﬁned

peaks under optimal conditions ............................................... 64

Overlayed Chromatograms of a US. EPA TO-14A

volatiles mixture separated by the RVM

Adaptive Air Sampler experimental air concentration

system .................................................................................... 66

Representative selection of analyte extraction time
proﬁles from a 100 ppbv (nom) TO-14A gas mix ..................... 71

Comparison of 16 ppbv (nominal conc., moisture added)
sample loading by benchmark and experimental air
concentration systems ............................................................. 80

Calibration curves for the BTEX series in the TO-14A

gas mixture. The HSA—SPME element has comparable
sensitivities to that of the benchmark over a 1-minute

sampling time .......................................................................... 82

Dynamic ﬂow HSA—SPME response versus benchmark
cryotrap air concentration response for benzene in air ........... 87

Total power consumption for Entech benchmark

and experimental RVM ooncentratorlLTM-GC
temperature ramp .................................................................... 91

xi

CHAPTER 1 - INTRODUCTION

1.1. Purpose of Research

With the advent of global and domestic terrorism within the last decade,
there has risen a critical need to perform on-site and real-time air analyses for
the detection of various volatile organic compounds (VOCS) which may be used
in terrorist attacks. Although more expensive to produce and acquire, chemical
warfare (CW) agents are a serious threat, but more readily available “toxic
industrial materials” (TIMS) such as hydrogen cyanide and phosgene, mass-
produced by many companies in the United States and internationally, are easier
to obtain [1]. Released into the environment, TIMS have the potential to cause
injury and irreparable, long-term health problems to large numbers of people.
This situation necessitates the development of a portable, rapid air sampling
system that will accurately assess volatile TIMS components which will provide
crucial information to help enable law enforcement agents to execute a threat-
Speciﬁc response.

The detection of human remains, drugs, and explosives by police canines
is also an important component of investigative law enforcement. Cases
involving missing persons have been resolved by the ability of trained dogs to
store a victim’s distinct chemical signature, and track this speciﬁc odor to crime
scene locations or buried remains. Canines have also been trained speciﬁcally
to discern human scent on living persons and have been successful in matching

criminal suspects to proffered scent evidence. Because there are no national

qualiﬁcation standards for human scent-discriminating canines, scent detection is
not currently accepted as evidence, although positive responses are generally
accepted as probable cause by court systems [2]. It would therefore be
beneﬁcial for law enforcement to mechanically reproduce the canine’s ability to
extract and detect VOCS from air to determine target compounds and their
respective concentrations. Rapid, in Situ detection and analysis at this level
requires the ability to extract volatile components from large volumes of air at
high ﬂow rates (lein), assuring representative sampling in a minimal amount of
time. This detection system could be integrated with a portable detection device
(i.e. gas chromatography coupled with mass spectrometry, GC/MS) to be used
with scent-discrimination canine teams in the ﬁeld. During a search, the handler
can utilize the detection device for providing conﬁrmatory information of positive
responses from canines at the speciﬁed area of interest (i.e. scent evidence or
buried remains).

Improvements have been made in the miniaturization of the laboratory
versions of the few commercial GC/MS systems available for ﬁeld use, but they
remain limited by substantial power consumption, and lengthy sampling times
using headspace and/or low ﬂow dynamics (mL/min). The extra time needed for
sampling reduces to total number of samples that may be obtained over a given
period, and this may be crucial in emergency situations. Field-portable GC/MS
systems currently utilize in-house solid sorbent microtraps and/or valve loops that
require long extraction times due to their low sample capacity. There are also no

ﬁeld-portable systems which currently have the ability to apply both preliminary

and conﬁrmatory testing to an air sample, which could provide valuable court-
defensible data. These shortcomings can be addressed by implementing an air
sampling device that incorporates solvent-free extraction that minimizes sample
preparation times and power consumption for efﬁcient operation [3].

Solid-phase microextraction (SPME) coupled with laboratory GC/MS has
been demonstrated to perform solvent-free extraction of volatiles from the
headspace of forensic specimens including drugs and explosives, and shows
potential in developing an understanding of the complicated process of canine
odor detection [4]. SPME has proven to be useful in a wide variety of analytical
applications, mainly due to its excellent sensitivity, cost effectiveness, and
reproducible results. In its current syringe-based design, however, it remains
limited by lengthy extraction times and requires bulky, power-thirsty thermal
injection ports to passively heat and desorb analytes from the SPME ﬁber.
Therefore most ﬁeld sampling is performed via grab sampling using such
materials as Tedlar bags, vacuated stainless steel canisters, charcoal strips, or
desorption tubes, and analyzed in a laboratory setting.

This research project is part of a continuing endeavor to provide portable
instrumentation for the Federal Bureau of Investigation (FBI) in the effort to
perform rapid and reliable volatile chemical analyses in the ﬁeld. The ﬁrst phase
of this project was the method development of the RVM low thermal mass (LTM)
GC (RVM Scientiﬁc, Inc.) which proved to offer rapid (>100° C/min temperature
ramps) yet highly controllable and reproducible chromatographic separations

unmatched by conventional GC systems, operational at ﬁeld-deployable size and

power [5, 6]. The subsequent phase of this project involves the development of a
method which will characterize the RVM Adaptive Air Sampler air injection
system as an effective alternative to headspace SPME and other forms of grab
sampling used today. This system utilizes both SPME and solid phase extraction
(SPE) technology in a design speciﬁc for sampling at high ﬂow rates (L/min).
Trace-level volatiles may be extracted from large volumes of air using only a
fraction of the time required for traditional SPME headspace sampling, while
retaining laboratory-level sensitivity. This extraction speed is representative of
canine olfactory capabilities which is desired for rapid detection [7]. The air
sampler requires no cryogen or extraneous solvents, making it amenable to ﬁeld
portability. The RVM Adaptive Air Sampler, coupled with proven LTM-GC
technology, has the potential to provide rapid, cost-effective, and reproducible air
sampling and analysis in the ﬁeld while retaining laboratory-level performance.

To determine the system’s capabilities and limitations, the RVM Adaptive
Air Sampler was evaluated on a conventional Agilent 6890/5973 GC/MS
laboratory system generally employed by many forensic and environmental
laboratories. Comparative benchmark data was obtained using the Entech 7100
Air Concentrator coupled with a resistively heated LTM-GC column (installed on
a retroﬁt oven door on the Agilent 60). Upon completion of baseline evaluation,
the RVM Adaptive Air Sampler injection system was evaluated under similar
conditions. A 39-component, US. EPA compendium method toxic organics (TO-
14A) hydrocarbon gas mixture was analyzed under optimal extraction and

desorption conditions using the laboratory Entech 7100 Air Concentrator system

for comparison against varying RVM experimental conditions. Performance of
both systems was evaluated in terms of linear dynamic range, sensitivity, limits of

detection, desorption efﬁciency, and power consumption.

1.2. Solvent-free Sample Preparation

The operating principle behind any sample preparation method is to
partition analytes between the sample matrix and an extracting phase [8]. Ideal
analytical separations, especially those performed in ﬁeld situations, should be
rapid with minimal sample loss, simple to peIfonn, and cost effective. For
sampling trace-level organic species in the ﬁeld, a number of procedures can be
applied. Presently, many analytical laboratories perform liquid-liquid extractions
(LLE) such as the Soxlet method, requiring large volumes of solvent and 2-24
hours of sample preparation. Acceptable recoveries are obtainable from the
procedure, but at the expense of hazardous air pollution from evaporated
solvents and additional concentration procedures. This cumbersome and time-
consuming process was the rate-deterrnining step in the overall analytical
process, and impractical for ﬁeld implementation.

Solvent-free sample preparation methods have been introduced over the
last three decades and have gained wide support from the scientiﬁc community
for enabling more efﬁcient, laboratory throughput by reduction of sample
preparation times while retaining analyte sensitivity and selectivity. These

methods are also the driving force behind ﬁeld sample preparation. Solvent-free

extraction techniques are Classiﬁed into three categories: gas phase, membrane,

and solid sorbent extractions.

1.3 Gas Phase Extraction
1.3.1 Headspace Extraction

Static headspace (SHS) extraction is a widely used method for the
recovery of volatiles. This method requires no solvents and little sample
preparation, but may necessitate bulky instrumentation, depending on sample
matrix type. For liquid matrices, the volatile components in a sample are allowed
to equilibrate with the air above the sample (the headspace). The analytes must
have a vapor pressure greater than or equivalent to the matrix for partitioning into
the headspace. Equilibration of analytes between the matrix and headspace
occurs after a period of time, by which a portion of the headspace can be
sampled directly for on-column GC focusing or GC/MS analysis.

The SHS method can be applied to ﬁeld samples, whereby a sampling
instrument can “sniff” directly above the area in question. Sampling in this
manner is rather low in sensitivity Since trace-levels of analyte (ppb or sub-ppb)
in a dynamic environment cannot be detected without being concentrated to
remove interferences such as air and humidity. Lizzani-Cuvelier et al [9]
illustrated that compound recoveries from olive oil by SHS were poor, even at
various temperatures (40-110°C) and extraction exposure times (30-120 min.).
Bicchi et al [10] tested the relative efﬁciencies of SHS and seven other solvent-

free extraction techniques for 16 components found in Arabica coffee. It was

discovered that the SHS method extracted 1-2 orders of magnitude less than the
average recovery for all other techniques.

Dynamic headspace methods are an improvement over static headspace
methods and involve the addition of a gas purge of volatiles into a second
trapping stage which concentrates the analyte for effective GC separation. In
this “purge & trap” technique, air, nitrogen, or helium iS forced through an
aqueous sample matrix at ambient temperature, and the resulting effervescence
removes volatile components from the sample which are collected in a cold
(cryogenic) or sorbent trap. Unlike the SHS method, dynamic headspace
extraction is an exhaustive approach which requires complete removal of analyte
for quantitation, and thus realizes higher recoveries. US. Environmental
Protection Agency (EPA) methods 601 and 624 utilize this proven method for
quantitation of volatile petroleum contaminants in groundwater, and chlorinated
pesticides in potable water, respectively. Le Pape etal [11] successfully isolated
16 volatiles of the red algae Palman'a palmate by crushing and diluting the
seaweed for purge and trap dynamic headspace extraction, at recoveries
comparable to simultaneous distillation-extraction (SDE), a popular yet
cumbersome and artifact-producing method. Drawbacks to dynamic headspace
extraction include long purge times (10-30 minutes per sample), contamination of
sample transfer lines due to foaming from purging undiluted sample, and

canyover signal from previous analyses.

1.3.2 Supercritical Fluid Extraction

The air pollution problems associated with the use of chlorinated organic
solvents in liquid-liquid extraction (LLE) was addressed in the mid-1980s with the
advent of supercritical ﬂuid extraction (SFE). It was discovered that many semi-
volatile organic compounds (SVOC) of analytical interest were soluble in certain
solvents once the temperature and pressure exceeded the critical point of the
solvent. The critical point is where the solvent is neither liquid nor gas, but a
“dense gas” containing physical properties characteristic of both phases that
favor higher extraction efﬁciencies. Supercritical ﬂuid exhibits viscosity, surface
tension, and diffusion properties comparable to a gas, enabling ﬂow through
pores of a sample and penetrates crevices of a material at a faster, more efﬁcient
rate than a liquid. It also favors the higher solubilizing power found in liquid
matrices, adding solvating Characteristics normally not found in a gaseous phase.
Therefore contaminated soils, sludges, ash, and other material can be easily
analyzed by SF E for the presence of SVOCS.

Many supercritical ﬂuids are gases at ambient temperature and pressure
(ATP). Recovery therefore becomes straightfonivard when compared to organic
liquids, which must ﬁrst be evaporated (ergo the pollution-related problems) to
concentrate the analyte and thus contribute to lower recoveries. The
supercritical ﬂuid can be separated from the analyte by simply releasing the
pressure. For volatiles analysis, the analyte stream can be bubbled through a
suitable solvent for the analyte, which will dissolve in the small volume of solvent

[12]. Using supercritical carbon dioxide as the solvent, Seitz et al [13] proved this

type of SF E/Purge 8. Trap procedure to be more effective in extracting 80 volatile

compounds (Cg-Ce) from whole and ground grain samples than by purge & trap

methods alone. Wong et al [14] showed that air toxicant extracts could be
recovered from air sampling adsorbents by SF E without additional sample

concentration steps. However, lower molecular weight volatiles, such as

bromochloromethane (<02) were lost due to volatilization in the expanding C02

stream during decompression of the supercritical solvent. The implementation of
SFE methods is preferred when conducting extractions of analytes of medium-to-

low volatility, but do require extraneous concentration to ensure recovery of high-

volatile components, such as C1-Cz halocarbons. For air analysis, SFE is best

utilized in tandem with grab sampling by means of a portable adsorbent bed and
a ﬁnal purge and trap concentration method.

The only known portable supercritical ﬂuid extractor has been developed
[15] for the Federal Bureau of lnvestigation’s Hazardous Materials Response
Unit, for the extraction of hazardous materials from investigation sites. This

device remains in the beta-testing stages, however, and requires lengthy

extraction procedures and heavy 002 tanks which increase instrument weight.

1.4 Membrane Extraction
A faster, more portable alternative to gas phase extraction is the removal
of volatiles through use of membrane material which can be directly coupled with

GC or GC/MS for rapid, real-time monitoring. In membrane extraction, VOCS

selectively diffuse out of an air or water sample matrix through a hydrophobic
polymer membrane into a nitrogen gas stripping phase. The polymer most
associated with the removal of VOCS and SVOCS from air is a hollow-ﬁber
membrane composed of silicone rubber [16]. It is more widely used than the
sheet membrane predecessor because it offers a higher ratio of surface area to
volume, which allows more efﬁcient extraction [8]. Semi-volatile compounds may
also be extracted with the use of liquid solvent streams, however at the expense
of longer elution times and loss in sensitivity.

The stripping phase containing the extracted organics is directly interfaced
with a GO or GC/MS, or is passed through an intermediate concentration stage
such as a cryotrap or an adsorbent-ﬁlled trap. These traps are designed to
further enrich volatiles and quickly desorb them as a vapor plug into the 60 for a
more efﬁcient separation. Pawlizsyn et al. [17] Showed that this membrane
extraction with sorbent interface (MESI) approach provides a useful tool for real-
time air monitoring of biogenic volatile organic compound (BVOC) emissions
released by freshly cut branches from a Eucalyptus dunnii tree. Here, an
automated sampling system was developed using a polydimethylsiloxane
(PDMS) membrane coupled to a dual-sorbent trap containing Tenax solid
sorbent and PDMS for trapping semi-volatiles and BVOC, respectively. It was
demonstrated that BVOC concentrations diminished over a 24-hr sampling
period, but semi-volatile emissions increased over the same period. It was

observed that applying a higher stripping gas ﬂow increases the extraction

1O

efﬁciency of the membrane by enhancing analyte mass transfer between
membrane and sorbent.

Hauser and Popp [16] optimized a method incorporating
thermodesorption/GC with membrane extraction of VOCS from a water matrix.
Detection limits for nine compounds had a range of 002-01 pglL, and
quantitative results corresponded well with those achieved with conventional
headspace-GC/FID. The rate limiting factors for overall analysis however were a
30-minute enrichment time required for optimized system performance, and the
decreasing extraction rates for volatiles whose boiling points approached 220°C.
The method could not be used for the analysis of semi-volatile compounds,

owing to slow diffusion across membrane boundaries.

1.5 Sorbent Extraction
1.5.1 Solid Phase Extraction

As opposed to liquid and gas phase extraction techniques, trace
concentration from both air and water matrices can be more easily accomplished
by trapping on a solid sorbent support to preconcentrate the analytes of interest
to a concentration sufﬁcient for analysis. Adsorbents commonly used for this
method include charcoal, macroreticular porous polymers, polyurethane foams,
bonded-phase materials, and ion-exchange resins [18]. Different sorbent
materials are designed for the collection of speciﬁc organic chemicals based on
the predominant functional group, polarity and/or molecular weight of the analyte.

Some common characteristics of sorbent material include large sample capacity,

11

large breakthrough volume, and high afﬁnity for non-polar organics coupled with
its hydrophobicity toward water and humidity. Unlike traditional solvent extraction
(LLE), solid phase extraction is simple to perform and requires less overall time
than other methods, uses little or no solvent, and is relatively inexpensive.
Solid Phase Extraction from Water Matrices

The use of bonded-phase sorbents for the solvent extraction of semi-
volatiles from water samples has gained wide acceptance as a valid

environmental laboratory technique in the past two decades. Here, non-polar

C5 - C13 carbon Chains are dispersed in a polymer coated on ﬂat disks or

contained in tube cartridges. In this method, an aqueous sample containing
suspected SVOCS is passed through a SPE disk or cartridge in which the
compounds of interest are retained on the polymer coating. Retention is driven
by hydrophobic interactions from the net repulsion between solute and solvent
(water), afﬁnity for solute to the non-polar hydrocarbon phase on the SPE
support, and selective polar interactions [18]. Using a small quantity of organic
solvent, the analytes are ﬂushed from the SPE support into a collection tube, and

further concentrated by evaporating the solvent with a nitrogen gas stream to a

quantity amenable for instrument detection. Enrichment factors of 103 - 107 can
be obtained by this method [18]. The procedure works well for exhaustive
extraction (complete analyte removal) for quantifying SVOC in drinking water,

which normally contains minimal amounts of interfering compounds. Recoveries

are usually between 85-100%. However, waste water and aqueous biological

12

matrices may contain particulate matter (i.e. humic material) which can clog the
sorbent material and diminish overall analyte recovery.

Although the use of solid phase extraction for liquid matrices has lessened
the need for large amounts of solvent used in LLE, and actual extraction times
are relatively short (< 1 hr), extensive preparation procedures increase the total
extraction time and can require as much as 4 hours to perform up to eight
simultaneous extractions using automated instrumentation. The SPE sorbent
must be conditioned with solvent prior to use to remove any manufacturing
contaminants. Conditioning may not remove all contaminants, such as bis-
(ethylhexyI)-phthalate, a plasticizer found in manufacturing materials, which is
also a target analyte for US. EPA method 625.2 (pesticides in drinking water).
Extracts require drying with sodium sulfate to remove residual water. Depending
on extract volume, the solvent may need to be evaporated further to achieve
proper analyte concentration. The technique is limited to semi-volatile
compounds with boiling temperatures substantially above that of the desorption
solvent temperature [8].

Solid Phase Extraction from Air Matrices

The isolation and concentration of volatile organic compounds in air can
be performed efﬁciently with solvent-free solid phase extraction. Because these
high vapor pressure-organic species are normally encountered at the ppb and
sub-ppb levels in the environment, preconcentration from a large sample size is
required to achieve detectable levels. Traps made with solid sorbent materials

mentioned are an effective means to accomplish this task. When an air stream

13

containing organics passes through a trap composed of a solid sorbent material
at ambient temperature, the Chemical vapors become adsorbed onto the support
while air and humidity pass through, effectively concentrating VOCS over a given
period of time. The VOCS are released from the sorbent by ballistic resistive
heating of the trap as it is back-ﬂushed with a carrier gas. The carrier gas injects
this concentrated “plug” of chemical vapor onto the head of a GC column for
chemical separation. The range of compounds to be studied usually includes
molecules of varying volatility and polarity. Therefore the ideal sorbent would
permit large breakthrough volumes for light gases while providing rapid
quantitative desorption of larger molecules [19].

There is currently a wide variety of commercially available sorbent
materials, each with its own afﬁnity for certain chemical compounds. Traps can
be constructed with single or multiple sorbencies, depending on the molecular
weight and polarity of the compounds of interest. In the instance of multiple
sorbents, beds are arranged in the trap in order of increasing sorption activity
and decreasing mesh size from the sampling inlet. During the sampling phase of
a purge and trap method, compounds of higher volatility and lower molecular
weight pass through the larger mesh solid sorbents of the sorbent trap, but are
trapped by succeeding smaller mesh beds. During desorption, the trap is
backﬂushed by helium gas so higher molecular weight compounds do not come
into contact with the strongest adsorbents [20]. Optimum desorption conditions
will theoretically produce high analyte recoveries, sharp, chromatographed

peaks, and accurate quantiﬁcation with little to no carryover between runs.

14

The Microtrap: Extraction of Volatiles in Air

The efﬁciency of VOC trapping by solid sorbents has expanded their use
in gas chromatography as miniaturized versions in portable instmmentation.
These “microtraps” are produced by packing capillary or small-bore tubing with a
solid adsorbent [21]. An on-line microtrap (OLMT) can effectively bypass the
conventional GC injection port to make direct injections onto the GC capillary
column. Trapped and concentrated organics are released by ballistic heating of
the microtrap and swept onto the GC column by carrier gas. Due to small
dimension and careful choice of adsorbents, microtraps have enabled the
concentration and subsequent detection of analytes at the ppb and sub-ppb level
from voluminous amounts of air without solvents or the need for cryogenic
refocusing [22], a frequently incorporated procedure in laboratory air
concentrators to aid in trapping lighter molecular weight VOCS and for removing
moisture.

The implementation of the microtrap with gas chromatography has
increased its use in on-site and near real-time monitoring of VOCS without the
need of transporting gas samples to the laboratory for analysis [23]. Mitra et al
[23] showed that a solid sorbent microtrap (L = 15 cm) coupled with a membrane
extractor could effectively trap VOCS from catalytic incinerator exhaust over a
wide linear dynamic range (sub-ppb to ppm levels), and maintain system
response stability after 140 injections of a 1 ppm toluene standard (R80 4%) at
a 10 mL/min sampling rate. The microtrap was able to sample and desorb in 3-

minute intervals, providing near-real-time and continuous chromatographic

15

analysis. Bassford et al [22] used a microtrap packed with three solid sorbents of
3-5 mg each to sample atmospheric halocarbons (oceanic and pollutant) in one-
hour intervals continuously over a one month period. The analytical system was
able to detect and quantitatively measure a wide range of halocarbons in 0.2 L
air samples at concentrations as low as 0.1 ppt(v). The continual monitoring of
the halocarbons in the ﬁeld provided information on short-term ﬂuctuations in
concentration which could not otherwise be assessed by grab sampling, which
requires transportation of samples to a laboratory for analysis and can cause
halocarbon degradation if storage is required.

The use of solid sorcent materials for the extraction of VOCS has
increased the versatility of air sampling protocol to the point where analysis can
be performed in the ﬁeld with use of portable gas chromatography or GC/MS.
However, for complete desorption of captured vapor, the microtrap is limited to a
small internal volume which can be rapidly heated to sufﬁcient desorption
temperatures in an amount of time short enough for tight vapor injections. This
small volume severely limits the sampling ﬂow rate to the mL/min range by which
extraction can take place. Sampling with the microtrap therefore requires a long
amount of time (10 min — 2 hrs per sample) to obtain representative analyte
mass loading from a relatively large volume of air. Thus, sampling by microtrap

alone may still be the limiting factor in performing air analysis in the ﬁeld.

16

1.5.2 Solid Phase Microextraction

In 1987, Pawliszyn and Liu [24] discovered that several micrograms of
chemicals of varying volatility could be easily desorbed from the chemically
modiﬁed surface of a fused Silica optical ﬁber by laser desorption for rapid GC
injection. In a follow-up study in 1990, Pawliszyn and Arthur [25] created the
technique of solid phase microextraction (SPME) by illustrating how a fused silica
ﬁber, coated with stationary phase, could adsorb analytes from a liquid matrix
and rapidly desorb them on-column via heated GC injection port . In the
analysis, a 1-cm length fused Silica ﬁber was coated with a polyimide stationary
phase (171 t 5 pm thickness), and placed in an aqueous sample inside a closed
container. Analytes within the sample partitioned into the stationary phase during
a two-minute exposure. The ﬁber was removed from the sample, and injected
into a GC injection port via syringe which enabled rapid thermal desorption of
analyte from the ﬁber, resulting in efﬁcient, reproducible, on-column separations.
Pawliszyn argued that SPME could retain the advantages of solid phase
extraction (SPE); namely, the elimination of solvent and reduced analysis time
over LLE methods, as well as surpass SPE extraction and desorption efﬁciencies
by eliminating the need to perform exhaustive extractions to acquire quantitative
analytical results. Solid phase microextraction, unlike SPE, only requires that the
analyte establish equilibrium between sample matrix and stationary phase. The
amount of time required to establish equilibrium depends on the diffusion

properties and distribution constant of the analyte.

17

The successful implementation of SPME in water analysis broadened the
scope of its use for other applications, including air analysis. In a preliminary
study, a polydimethylsiloxane (PDMS) coating was used as the stationary phase
on fused silica ﬁber to effectively capture atmospheric VOCS [26]. Pawliszyn in
1997 introduced the concept of extracting analyte from a dynamic airﬂow using
SPME ﬁbers [27]. He proved that a constant ﬂow rate of gas mixture containing
a stable concentration of analyte passing over a SPME ﬁber exhibited Similar
amounts of analyte loading for light volatiles (K $5000) to that of static
headspace extraction over the same time period. Compounds in the headspace
analysis with K 26000 required a doubling of the extraction time to reach
equivalent loading observed with the dynamic ﬂow analysis, since species with

increasing molecular weight have decreasing rates of diffusion, where:
Cair = Cﬁber /K (1)
K = analyte partition coefﬁcient

C3,): concentration of analyte in air (in mg/m3)

Cﬁber = concentration of analyte on the SPME ﬁber (in mg/m3)

Pawliszyn, using a known concentration of analyte on the SPME ﬁber, the known
temperature, and rearranging equation (1) with the ideal gas law and Clausius-

Clapeyron equations, derived the expression that determined the unknown

concentration of an airborne analyte (ppmvair):

R T
MWPVf low/7“”) (2) [271

 

ppm(v)air : (Cﬁber)

18

where:

R = ideal gas law constant
MW = analyte molecular weight

P = analyte vapor pressure at a known temperature T (in Kelvin)

Vf= volume of SPME ﬁber

 

 

Using equation (2), a battery of tests were performed to characterize

calibration procedures for a VOC/SVOC gas mix using PDMS-coated SPME

ﬁbers of varyi

ng length and thickness. The results are summarized as follows:
Upon normalization of SPME ﬁber lengths, interﬁber analyte
recovery precision was better than 9% RSD for 10 different ﬁbers
lntraﬁber day-to-day reproducibility (n = 8) was < 9% RSD for all
analytes tested

Humidity in amounts < 95% relative humidity did not result in
reduction of analyte mass loading

Analytes with larger K values have lower limits of detection (LOD)
Similar results were observed between SPME ﬁber sample loading

at both laminar and turbulent ﬂow rates

19

The last point indicates that the SPME method has the potential of incorporating
ﬂow rates up to the L/min maximum without compromising loading accuracy and
reproducibility.

In 2003, Ciucanu et al [28] developed a microtrap design that incorporated
SPME technology. Using a straight piece of wire (d = 0.1 mm) as support, a
helical aluminum wire (d = 0.07 mm) coated with PDMS stationary phase was
wrapped down the length of the straight wire at a constant pitch. The wires were
then placed lengthwise in the lumen of an inert, stainless steel tube (80 mm x
0.75 mm id. x 0.95 mm o.d.). Desorption of the wire in the microtrap consisted
of traditional resistive heating of the outer surface of the stainless steel tube, in
which passive convection would internally heat the wires. Ciucanu claimed that
this microtrap design increased the amount of time required to observe analyte
breakthrough to ten times over that of a straight, thick-ﬁlm SPME capillary trap
with a Similar amount and thickness of PDMS phase. It was theorized that
analyte adsorption was aided by the helical design creating turbulence at low
ﬂow, increasing analyte mass transport from the air matrix to the surface of the
stationary phase [28]. The experimental RVM Adaptive Air Sampler design
implements a similar helical shape for the adsorbent to take advantage of
turbulent ﬂow, albeit higher ﬂows (Umin) than the optimal 5 mLJmin ﬂow rate

incorporated in the Ciucanu device.

20

1.6 Low Thermal Mass Gas Chromatography

A capillary gas Chromatograph functions primarily to separate vaporized
analyte mixtures by means of forcing them through a capillary tube whose inner
walls have been coated with a polymer. This polymer is the stationary phase.
The analyte vapor, moving through the column by a gaseous mobile phase,
interacts with the stationary phase . The interactions are governed by analyte
boiling point, vapor pressure, and molecular weight against the physical
properties of the stationary phase. Since most chemical Species each exhibit
variations in these properties, some analytes will interact with the stationary
phase more strongly than others. The analytes in the vaporized mixture can thus
become separated from each other through the course of traversing the column.
The analytes will emerge from the column at distinct times (retention times), and
are then detected by means of various commercially available detector devices
(FID, MS, etc.). By utilizing controlled temperature ramps to increase the
temperature of the column during the chromatography, less volatile analyte
becomes more volatile, eluting from the column earlier than if temperature was
isothermal. Therefore the application of heat to the column decreases the overall
analysis time, and is implemented in many 60 methods used today.

Most conventional commercial benchtop GCS utilize large convection
ovens to achieve stable temperatures for uniform column heating. However,
their large thermal cavity and signiﬁcant power consumption (> 1kW for a

0.5°C/min temperature ramp to 200°C) prohibits their use in ﬁeld situations [5].

21

RVM Scientiﬁc, Inc. has created a GC design with reduced size and power
requirements using proprietary technology [29]. This new lightweight, compact

design (Figure 1.1) utilizes low thermal mass (LTM) technology which

LTM-GC
Column

 

Figure 1.1. An RVM LTM A68 retroﬁt GC oven door and LTM GC column
installed on an Agilent 6890 gas Chromatograph

effectively replicates the function of a conventional GC oven, but at a fraction of
power and space. Because the LTM-GC is highly efficient and small, it can
incorporate fast temperature ramp rates and portability unachievable by
conventional GC systems.

The concept design and ability of the LTM-GC has been previously
discussed [5, 6]. The RVM device combines a commercially available GC

capillary column with a ceramic ﬁlament-wrapped heating element [29]. These

22

ceramic filaments provide thin insulation and uniform heating up to 1200°C.
Temperature sensors are also bundled with the GC column, and the assembly is
tightly wound into a torus, which is then covered in tin foil (See Figure 1.2). The
exchange of heat from the insulated heating element to the column is tightly
controlled by the temperature sensors, and heat loss is minimized by the
insulation of the tin foil, thus realizing the low thermal mass effect. The low
thermal mass of the tin foil also enables rapid column cool- down (3 minutes)
upon completion of analysis by means of a heat sink fan in Close proximity to the
column.

Column’SensorVWI'e

‘ ‘ ~ ‘ ‘ ASSembly

hm aIe'I

. \ .
, Sensor CC (,D-Iimn
. ./

\

I i
WI

Ir‘;;Il-I

. .
‘ Insulator? ﬂamingwjre

 

Figure 1.2. RVM Low thermal mass GC design. R. V. Mustacich, US. Patent
6,209,386, (2001 ).

The result of low thermal mass GC column technology is improved speed

of analysis and sensitivity. An illustration contrasting conventional vs. low

23

thermal mass chromatography is shown in Figure 1.3. 1 [IL injections of a 15-
component aqueous alkane mixture were used as the test mix to observe
differences in total run time and analyte peak height for an LTM-GC (90°C/min)
vs. conventional GC (20°C/min) using capillary columns of Similar dimension and
stationary phase. It is observed that the LTM-GC offers peak resolution >1.5,
one order of magnitude greater in peak height (higher analytical sensitivity), in
1/3 the total run time required in the conventional GC system operated at a
maximum controllable temperature ramp rate.

LTM-GC technology is a proven method with much potential in the
emerging ﬁeld of near-real-time portable chemical detection devices. Combining
LTM-GC with an adaptable front-end air preconcentration system with rapid
sample extraction and injection capabilities could offer analytical advantages with
many forensic applications, including canine scent tracking, clandestine lab
sampling, and volatile chemicals monitoring for the Federal Bureau of

Investigation and other law enforcement entities.

24

mcano >538 00 mmmE
REP: 26. van .m:o_Em>coo .3 3:238 5:. mega 28888-9 m Co comtmano oEaEmoﬁEoEo .né 2:9".

 

 

 

.58. as:
a a L. o m e n a F o
8+mod
SE:
83:: w
. :. 8&3 w
3.8 is; 3.8%. us.
Sign ...._., m Beam Sam...
. ,. . . 8+mo.~
. ,... , . nu w. Jar—.250?—
353163; .52.. occx?..o~0.n0.w..2=§_2 ~03.
8&3

ASP—n: news. .2522. 30.. .n> 3.53.5250
"newton—Eco >cnucuoquoEo

 

 

25

1.7

10.

11.

References

Bennett, M., TICS, TIMS, and Terrorists, in Today's Chemist at Work.
2003. p. 21-25.

Stockham, R., D. Slavin, and W. Kift, Specialized use of Human Scent in
Criminal Investigations. Forensic Science Communications, 2004. 6(3).

Pawliszyn, J., Sample Preparation: Quo Vadis? Anal. Chem., 2003.
75(1 1): p. 2543-2558.

Furton, K., et al., Laboratory and field experiments used to identify Canis
lupus var. familian's active odor signature chemicals from drugs,
explosives, and humans. Analytical and Bioanalytical Chemistry, 2003.
376(8): p. 1212-1224.

Sloan, K., Method Development and Implementation of a Novel
Temperature-Programmable Gas Chromatograph for Rapid Forensic
Analysis in the Field, in Columbian School of Arts and Sciences. 2001,
George Washington University: Washington, 0.0. p. 47.

Eckenrode, B., K. Sloan, and R. Mustacich, Development and Evaluation
of a Low Thermal Mass Chromatograph for Rapid Forensic GC-MS
Analysis. Field Analytical Chemistry and Technology, 2001. 5(6): p. 288-
301.

Settles, G. and D. Kester, Aerodynamic Sampling for Landmine Trace
Detection. SPIE Aerosense, 2001. 4394(paper 108).

Pawliszyn, J., Solid Phase Microextraction: Theory and Practice. 1st ed.
1997: Wiley-VCH, Inc. 247.

Lizzani-Cuvelier, L., et al., Comparison of Static Headspace, Headspace
Solid Phase Microextraction, Headspace Sorptive Extraction, and Direct
Thermal Desorption Techniques on Chemical Composition of French Olive
Oils. Journal of Agricultural and Food Chemistry, 2003. 51(26): p. 7709-
771 6.

Bicchi, C., et al., Headspace Sorptive Extraction (HSSE), Stir Bar Sorptive
Extraction (SBSE), and Solid Phase Microextraction (SPME) Applied to
the Analysis of Roasted Arabica Coffee and Coffee Brew. Journal of
Agricultural and Food Chemistry, 2002. 50(3): p. 449-459.

Le Pape, M., et al., Optimization of Dynamic Headspace Extraction of
Edible Red Algae Palman'a palmata and Identiﬁcation of the Volatile
Components. Journal of Agricultural and Food Chemistry, 2004. 52(3): p.
550-556.

26

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Skoog, D., J. Holler, and T. Nieman, Principles of Instrumental Analysis.
1998, Philadelphia: Harcourt Brace & Company.

Seitz, L., M.S. Ram, and R. Rengarajan, Volatiles Obtained from Whole
and Ground Grain Samples by Supercritical Carbon Dioxide and Direct
Helium Purge Methods: Observations on 2, 3-Butanediols and
Halogenated Anisoles. Journal of Agricultural and Food Chemistry, 1999.
47(3): p. 1051-1061.

Wong, J.M., et al., Determination of Volatile and Semivolatile Mutagens in
Air using Solid Adsorbents and Supercritical Fluid Extraction. Anal. Chem..
1991. 63(15): p. 1644-1650.

Wright, 8., T. Zemanian, and S. Anthony. Prototype Performance
Evaluation for the Federal Bureau of Investigation Field-portable
Supercritical Fliud Extractor. in 223rd ACS National Meeting. 2002.
Orlando, FL: American Chemical Society.

Hauser, B. and P. Popp, Membrane Extraction Combined with
Thennodesorption/Gas Chromatography and Mass Selective Detection for
the Analysis of Volatile Organic Compounds in Water. Journal of High
Resolution Chromatography, 1999. 22(4): p. 205-212.

Pawliszyn, J., et al., Sampling and Monitoring of Biogenic Emissions by
Eucalyptus Leaves using Membrane Extraction with Sorbent Interface
(MESI). Journal of Agricultural and Food Chemistry, 2002. 50(22): p.
6281 -6286.

Poole, CF. and SA Schuette, Isolation and Concentration Techniques for
Capillary Column Chromatographic Analysis. Journal of High Resolution
Chromatography, 1983. 6: p. 526-549.

Mitra, S. and C. Feng, Breakthrough and Desorption Characteristics of a
Microtrap. J. Microcolumn Separations, 2000. 12(4): p. 267-275.

Supelco, l., Chromatography Products for Analysis and Puriﬁcation.
2003/2004: Sigma-Aldrich Co.

Mitra, S., et al., A Microtrap Interface for Continuous Monitoring of VOCS
in Air Emissions. Journal of Mass Spectrometry, 1999. 34(478).

Bassford, M., P. Simmonds, and G. Nickless, An Automated System for
Near-Real-Time Monitoring of Trace Atmospheric Halocarbons. Anal.
Chem., 1998. 70(5): p. 958-965.

Mitra, S., N. Zhu, and Z. Li, Application of On-Iine Membrane Extraction
Microtrap Gas Chromatography (OLMEM-GC) for Continuous Monitoring
of VOC Emission. J. Microcolumn Separations, 1998. 10(5): p. 393-399.

27

24.

25.

26.

27.

28.

29.

Pawliszyn, J. and S. Liu, Sample Introduction for Capillary Gas
Chromatography with Laser Desorption and Optical Fibers. Anal. Chem.,
1987. 59(10): p. 1475-1478.

Pawliszyn, J. and C. Arthur, Solid Phase Microextraction with Thermal
Desorption Using Fused Silica Optical Fibers. Anal. Chem, 1990. 62(19):
p. 2145-2148.

Pawliszyn, J. and M. Chai, J. Environmental Science and Technology,
1995. 29: p. 693-701.

Pawliszyn, J. and P. Martos, Calibration of Solid Phase Microextraction for
Air Analyses Based on Physical Chemical Properties of the Coating. Anal.
Chem., 1997. 69(2): p. 206-215.

Ciucanu, l., et al., Helical Sorbent Microtrap for Continuous Sampling by a
Membrane and Trap Interface for On-Iine Gas Chromatographic
Monitoring of Volatile Organic Compounds. Anal. Chem., 2003. 75(4): p.
736-741.

Mustacich, R. and A. Neushul, Adaptive Sampling Technology Final
Report. 2003, RVM Scientiﬁc, Inc.: Santa Barbara.

28

CHAPTER 2 - EXPERIMENTAL

2.1 Adaptive Air Sampler: HSA-SPME Element Design

In 2003, BA Eckenrode of the Federal Bureau of Investigation and RV.
Mustacich of RVM Scientiﬁc, Inc. developed a novel design utilizing solid phase
microextraction (SPME) for trapping trace-levels of volatile and semi-volatile
compounds from large volumes of air at high ﬂow rates (L/min). Unlike the 10-
mm length SPME-coated ﬁbers traditionally employed in headspace syringes
which use passive heating (via GC inlet) for desorption, the solid sorbent SPME
phase is coated on a 100 mm-length (d = 0.127 mm) nickel alloy wire which can
be self-heated by application of a small voltage. The alloy wire (Stablohm 675,
California Fine Wire 00., Grover Beach, CA) is speciﬁcally manufactured to resist
oxidation at higher temperatures to ensure thermal stability over the course of
repeated ﬂash heating. To ensure uniform desorption, the wire was calibrated for
resistance by controlled heating in a laboratory oven. Using the temperature
dependence of the resistance, both power and current were calculated as a
function of temperature. From these measurements it was found that
temperature dependence on current was independent of the length of wire with
uniform diameter [1]. Technical data for the Stablohm 675 alloy wire is shown in
Table 2.1. The high-surface area, SPME-coated wire, or “HSA—SPME element,”
is therefore capable of rapid, controllable desorption of adsorbed VOCS by
means of a programmable current source, regardless of minor variations in the
length of wire used. The nickel alloy can rapidly achieve proper desorption

temperatures (ZOO-400°C) at a rate of 4000°Clmin from ambient temperature [1],

29

consuming considerably less power than most conventional air concentration

systems.

Table 2.1. Technical data for the Stablohm 675 nickel alloy wire used in the

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HSA-SPME element
Electrical Properties
Speciﬁc Resistance (Ohms/CMF) 675
Speciﬁc Resistivity (micro-ohms/cm) 112.22
Commercial Resistance Tolerance 5 00%
(<0.020) '
Temperature Coefﬁcient of Resistance 0 00015
(ohms/ohm/°C (0 — 100°C} '
Thermal EMF vs. Copper +0.002
Physical Properties
Density (gm/cm?) 8.247
Density (lbs/in3) 0.298
Thermal Conductivity (watts/cm/°C) 0.132
Coefﬁcient of Linear Expansion (x 10' 14
6 Win °C)°C
Melting Point °C 1350
Melting Point °F 2462
Composition
Nickel 60%
Chromium 16%
Iron Balance
Miscellaneous
Magnetic Faint
Operatim Temperature (°C) 900
Operating Temperature (°F) 1652

 

 

 

 

To provide optimum extraction efﬁciency under high ﬂow rates, the air
sampling device incorporating the HSA-SPME element requires a design (Figure

2.1) which will maximize contact between air ﬂow and the HSA-SPME element

30

 

 

  
 
   

Outer glass tube
/ (boundary layer]

4.18 Amm

 

Ni alloy wire '
odsorbonf coating "

 

‘L,

 

inner glass tube"

 

Figure 2.1. Cross-section of HSA-SPME element design

adsorbent coating. In this particular concept, the element is wound in helical
fashion at a constant pitch down the outer surface of a 50 mm x 1.2 mm o.d. x
1.0 mm id. borosilicate glass tube which provides the element with inert physical
support. The element termini are brazened to heating wire leads, with one lead
running inside the length of the glass tube, so that both SPME contact leads
terminate at the same end of the tube. The ends of the tube are then tightly
plugged with silane-treated glass wool to minimize air ﬂow through the lumen of
this tube. The tube is then inserted into another borosilicate glass tube of larger
diameter (78.5 mm x 4.8 mm o.d. x 3.0 mm id), whereby the helical HSA-SPME

element is contained in the annular space between the two tubes. A prototype of

31

this design is shown in Figure 2.2. During sampling, as air is forced into the 3-
mm id. tube, the air stream comes into contact with the inner tube and is further
restricted to the annular space where the HSA-SPME element is located. The air
is forced into a helical rotation and the analytes come into continuous contact
with the HSA—SPME element down the length of the inner tube. Thus the self-
heated, helical SPME coated wire design in a restrictive air space increases
mass transfer of volatile analyte from a turbulent, high ﬂow of air to the solid

sorbent.

 

 

 

 

 

[llllllllllllllllIIIIIIIIIIIII'IIIIIInflllll'lllIjiHillIII'! llli[‘l'll‘illlllllllIIIlIII
0 mm 10 20 30 40 50 60 70 80

—-—— - _..-_I-. “- _ .‘._— .1

 

 

 

, _..._.._. _
rm-ﬂ-H-I- .. ——-—-...-———r—-—.-—---—-—v-o o-d— --‘-—.— ---——...-—— -—--- .-

Figure 2.2. A prototype HSA-SPME Element

 

 

The rate of mass transfer is dictated by the size of the boundary layer of the
SPME coating, which is essentially the area created by friction between the air
ﬂow and sorbent surface where analyte diffusion takes place (See Figures 2.1 &
2.3) [2]. Because sample ﬂow through the HSA-SPME element is very high (4
Umin in this study), the boundary layer is reduced to an inconsequential size, so

that mass transfer is essentially unlimited. Therefore, in a short, 1-minute

32

sampling period at a ﬂow of 4L/min, a large, representative volume of air can be
sampled to concentrate a qualitative amount of organic compound, and quickly
then desorbed without the need of a bulky, high thermal mass injection port

usually required for passive desorption of 1 O-mm SPME ﬁbers .

Polymer- "
coated wire ' , Inner
glass tube

 

Figure 2.3. Scanning electron microscope (SEM) image of a
Carboxen/PDMS HSA-SPME element at 1300X
magniﬁcation (outer glass tube not shown)

The calculated HSA-SPME element surface area/ vapor volume ratio is
1:2.8. In comparison, the theoretical SPME surface areal volume ratio of a
typical syringe ﬁber (L = 10 mm, equivalent phase thickness of 65 pm) in the
annular space of this tube is 1:41, and 1:376 for a headspace extraction in a
40mL glass vial. The relatively high surface area to vapor volume ratio for the
HSA-SPME element should result in higher extraction efﬁciency than a syringe

ﬁber by minimizing sample loss even at higher sampling ﬂow rates (4-5 L/min).

33

Flow rates of this magnitude are possible in this particular design due to a low

pressure drop across the relatively large-diameter, Short-length tube.

2.2 Adaptive Air Sampler: Focusing Preconcentrator Design

The desorbed vapor volume from the HSA-SPME element is too large to
be directly injected into a gas Chromatograph for efﬁcient chromatography; it
must be further concentrated and re-focused by an intermediate sampling stage.
RVM Scientiﬁc, Inc. and associates have obtained the patent [Mustacich and
Richards, US. Patent # 6,223,584 (2001)] to develop this "focusing
preconcentrator,” (See Figure 2.4) which is an adsorbent-based microtrap
integrated with a small valve [1]. The innovative design integrates the air
sampling inlet directly with the microtrap opening, eliminating the requirement for
extra valving and heated intermediate sample lines normally encountered in
imbedded microtrap 60 systems (See Figures 2.5 & 2.6). Therefore less wattage
is required for valve and micro-pump operation, as well as for heating the
microtrap and transfer line during injection.

The microtrap used in the focusing preconcentrator is composed of a 3.81
cm x 0.165 cm o.d. x 0.142 cm i.d stainless steel hypodermic needle tubing
which contains 10 mg of solid sorbent material. The tubing is wrapped with
thermally insulated heating wire which provides high-efﬁciency heating upon
desorption/injection, controlled by two K-type thermocouples located at the
heating wire termini. The microtrap is encased in stainless steel housing to

retain low thermal mass during heating.

The microtrap carriage is composed of block aluminum which moves in
one degree of freedom (fonrvards and backwards) by an actuator. The default

mode of the actuator is in the sampling position, whereby the microtrap is fully

 

HSA-SPME
Element

 

A!"

”1 GCIMicrotrap
Interface , ,
Microtrap , Microtrap
in 8.8. Electronic —. ‘ Carriage
Valves

Housing ,. , ,1
L." T fir
Figure 2.4. A ﬁrst-generation focusing preconcentrator which integrates the
HSA-SPME element with the microtrap

 

exposed to the air inlet. During sampling at ambient temperature, the micro
pump creates a vacuum on the backside of the microtrap, which pulls air from the
air inlet through the microtrap to effectively trap and concentrate organic vapor
via solid sorbent. To begin an injection, the actuator turns a screw-ﬁtted drive
shaft by means of a simple o-ring, which moves the microtrap carriage into the

Teﬂon-lined GC interface to effectively seat the opening of the microtrap onto

35

 

AIMS“

 

 

 

 

true Ion
Injector 5
Carrier Aﬁotboﬂ
Gas "9

Isothermal. 200-220 'C

 

Ramp Heaed to 220 'C Capillty

 

 

 

 

 

var
imi ‘
Pump
i
Wli”
Valve
Mouton ‘ INLET cm
Valve 1 H Flow G“ rum
"as l .3... i—“TEJT—ee a..."
A, m" 77.
W.200-220'C
Gaul-Ir
wmmbm m...

 

 

 

Figure 2.6. Flow path of a GC system with simple integration of microtrap
(labeled “adsorbent") and air inlet

36

the GC transfer line opening. This movement also seals the microtrap opening
from the air inlet. Placement of the actuator in-line with the microtrap allows
movement in only one degree of freedom, reducing the possibility of incorrect
seating of the microtrap onto the transfer line if the carriage succumbs to physical
shock during transport. The clutch mechanism allows slippage of the o-ring to
prevent the carriage from becoming stuck in the injection position if the actuator
overdrives the carriage with excessive force. A close-up view of this mechanism
is shown in Figure 2.7. Once the microtrap is ﬁrmly seated on the GC transfer

line interface, helium carrier gas back-ﬂushes the microtrap as it is ﬂash-heated.

 

. Ii"

Figure 2.7. Close-up view of the O-ring clutch drive belt system for the
RVM focusing preconcentrator microtrap carriage

The analytes are subsequently desorbed from the adsorbent as a vapor “plug,”

and injected into the heated, ﬂow-restricted transfer line located inside the GC

37

interface plate and onto the head of the GC column for efﬁcient separation. After
injection, the actuator reverses to reposition the microtrap carriage for
subsequent sampling, and terminates movement when an electronic optical
switch detects an aluminum pin connected to the carriage.

The addition of the HSA-SPME element seated on the air inlet of the
focusing preconcentrator comprises the dual-adsorbent, RVM Adaptive Air
Sampler injection system (See Figure 2.4). For sampling, the HSA-SPME
element can be detached from the focusing preconcentrator and attached to the
outlet of a sampling pump, by which air can be pumped through the SPME tube
at ﬂow rates of up to 5 L/min using a portable air pump (See Figure 2.10).
Analytes are effectively extracted from the air sample matrix at ambient
temperature and adsorbed into the Carboxen/PDMS phase. The HSA-SPME
element is then disconnected from the sampling pump and reattached to the air
inlet of the focusing preconcentrator. During focusing preconcentrator sampling,
the micropump ﬂow of ~0.3 L/min is directed to the HSA-SPME element as it is
ﬂash-heated to a temperature necessary to effectively desorb analyte from the
element. VOC analytes are carried via air inlet into the microtrap for a ﬁnal
concentration of analyte by adsorption into the microtrap solid sorbent material,
followed by desorption onto the GC column. Used in tandem, the HSA-SPME
element and focusing preconcentrator microtrap enables rapid, large-volume
sampling at high ﬂow rates necessary for ﬁeld analysis. The dimensions (2.5” x
2.5” x 5”) and weight (approx. 4 lbs) of the device is conducive for use with ﬁeld-

portable instrumentation.

38

2.3 Instrumentation
2.3.1 Benchmark Air Concentrator

The Entech 7100 Air Concentrator (Entech Instruments, Simi Valley, CA)
was used as a benchmark for comparison with experimental results from the
RVM Adaptive Air Sampler injection system. The baseline Entech 7100 houses
three in-Iine cryogenic trapping modules designed speciﬁcally for the
preconcentration of VOCS, including very light gases, from whole air samples in a

laboratory setting. The ﬁrst module consists of a thermally protected, 1/8” i.d.
glass bead cryotrap with a temperature range of '180°C to 230°C and a
360°C/min ramp rate. The second module incorporates a 1/8” i.d. Tenax TA
adsorbent cryotrap. Used in conjunction, these traps offer advanced H20 and

00;; management by effectively resisting adsorption of these interfering

molecules [3]. The ﬁnal trapping module comprises an internal megabore

cryofocusing trap with a temperature range of '190°C to 100°C. During
concentration, sample was heated via heated transfer line and extracted by
modular adsorbent at the speciﬁed cryogenic temperature. Rapid heating of the

module desorbed the vapor plug into the succeeding module held at cryogenic

temperature. For on-column GC injection, the ﬁnal module was back-ﬂushed

with helium while ballistically heated at a rate of 10,000°C/min from '190°C to a

ﬁnal temperature of 100°C, without overshoot. The Entech 7100

preconcentration system parameters are listed in Tables 2.2 and 2.3.

39

 

 

Table 2.2. Temperature parameters for the three trapping modules in the Entech

7100 Air Concentrator
A ' ’ ‘ T ’ “T'éﬁrjoéiéiur‘és” 7 -_ " ‘ 7"
(deg c) . . _ .
Module 1: Module 2: Module 3:
Event Glass Beads/ Tenax TAI Megabore with
Cryogen Cryogen Cryofocusing
Concentration -150 -50 -150
Preheat 20 --- --
Desorption 20 180 100

Bakeout 130 for 5 min 190 --

Table 2.3. Flow and volume parameters for trapping in the Entech 7100 Air

 

Concentrator
7 7 7 7‘ ,, WFIows grid Volumes , , a -,, 7 7 i
. Preﬂush Flow Rate Volume
Medium (sec) _ (myL/min) (mL)
Calibration 2 200 200
Standard 2 200 200
Sample 2 100 75
Purge Flush

 

 

The instrument’s mass ﬂow controller was programmed to trap sampled
air at its maximum rate of 200 mL/min, with continual ﬂow adjustment capability.
The Entech 7100 was controlled using a personal computer with Entech

Instruments 7100/7000 Concentrator version 2.49-6 software.

40

2.3.2 Gas Chromatograph

An LTM A68 gas Chromatograph (RVM Scientiﬁc, Inc.) was used to
perform analyte separation. The LTM A68 housed a high-efﬁciency, low thermal
mass GC capillary column capable of rapid, controllable temperature ramp rates
(4000°C/min to 400°C) at low power which affords faster analytical run times over
conventional benchtop GC ovens [4]. To enable mass spectrometric detection,
the LTM A68 assembly was installed as a retroﬁt to the GC oven door of an
Agilent 6890 series, model 1530A (serial #USOOO110032) gas Chromatograph

(See Figure 2.8). The Agilent GC oven was used primarily to control helium ﬂow

 

Adaptive Air j \ LTM A68
Sampler Controller
Figure 2.8. Installation of the RVM Adaptive Air Sampler on the LTM A68
retroﬁt oven door on the Agilent 6890 GC

 

41

by means of an electronic pressure control, and to heat transfer lines (210°C) of
the respective preconcentrator systems to the LTM-GC. and from the LTM-GC to
the mass selective detector situated adjacent to the Agilent GC (See Figure 2.8).
The Agilent 6890 GC was controlled using a personal computer with

Chemstation version B.01.00 software.

2.3.3 Mass Selective Detector
An Agilent 5973 mass selective detector (MSD), model G1099A (serial #
US72821157) was used for identiﬁcation and quantization of chromatographed

analytes. The MSD system was equipped with a turbomolecular pump able to

provide a vacuum-housing pressure of at least 3 x 10.5 Torr. A two-stage rotary-

vane diffusion pump was also used as a secondary pumping system to reduce
pressure ahead of the turbopump and provide an exhaust to the atmosphere.
The instrument was autotuned daily using US. EPA Compendium TO-14A ion
abundance criteria for 4-Bromoﬂuorobenzene as the tuning standard. The
Agilent 5973 MSD was controlled by the Chemstation software program also

used to control the Agilent 6890 GC.

2.3.4 RVM HSA-SPME Element/Focusing Preconcentrator Microtrap
Assembly
The RVM Adaptive Air Sampler injection system was originally designed
for use in portable GC instrumentation; for experimental purposes, it was

installed on the LTM A68 retroﬁt door (RVM Scientiﬁc, Inc.) which sits adjacent to

42

the LTM-GC column (Figure 2.8). The focusing preconcentrator transfer line was
ﬁtted through an insulated opening in the door, and connected to a three-way low
dead volume tee (Valco, Inc.). The heated transfer outlet line from the Entech
7100 Preconcentrator was fastened to a second opening of this tee, enabling
simultaneous connection of both preconcentrators to the LTM-GC column ﬁtted
on the LTM A68. To divert helium ﬂow to the RVM Adaptive Air Sampler from
the Agilent GC injection port during experimental injections, a 3-way tee was
installed in the 1/16” stainless steel helium line leading to the 6890 GC injection
port (See Figure 2.9). Helium carrier gas is diverted to the RVM focusing
preconcentrator during injection by means of an electronic valve located on the

microtrap carriage (See Figure 2.4).

 

Figure 2.9. Installation of 1/16” 3-way tee; helium ﬂows to the RVM
Adaptive Air Sampler during preconcentrator injection.

43

The sorbents selected for this study were previously tested [5, 6] for
effectiveness in trapping VOCS in air. The HSA-SPME element was coated with
Carboxen dispersed in polydimethylsiloxane (PDMS) to a thickness of 65 um,
and the focusing preconcentrator microtrap was packed with ~5 mg each of
60/80 mesh Carbopack B and 60/80 mesh Carboxen 1000 (Supelco lnc.,
Bellefonte, PA). Sorbents were retained and separated by silane-treated glass
wool (Supelco Inc.).

The RVM Adaptive Air Sampler consisted of the following major
components and associated characteristics:

1. HSA-SPME Element 100 mm (d = 0.127 mm) Stablohm 675
Nickel alloy wire (60% Ni, 16% Cr, bal Fe; California Fine Wire
Company, Grover Beach, CA), coated with Carboxen dispersed in
PDMS with 65pm thickness (Supelco, Inc., Bellefonte, PA),
contained inside annulus of two glass tubes (1 & 3 mm id.
Borosilicate, Fisher Scientiﬁc, Fair Lawn, NJ); separate 24 V, 4.2 A
power supply (RVM Scientiﬁc, Inc.)

2. Focusing Preconcentrator. Aluminum base and carriage assembly
(RVM Scientiﬁc, Inc.); a 1.5” x 1/8” i.d. microtrap containing ~5 mg
each of Carbotrap B and Carboxen 1000 (Supelco, Inc.); a 264:1
planetary gearhead actuator motor (Micromo, lnc., Switzerland)
which provides high holding torque; three solenoid valves
(www.SMCUSA.C0m, model #SY114) for helium and air ﬂow

control; and a Knewberger micropump (www.KNF.com, model

44

#UNMP50KVD6) used for air sampling. Ceramic-insulated heating
wires (RVM Scientiﬁc, Inc.) were used to heat the microtrap and
transfer line; type K thermocouples were used for thermal control.

3. Focusing Preconcentrator Temperature Controller. mT-TC3
custom controller card (RVM Scientiﬁc, Inc.)

4. Power Supply. RVM Scientiﬁc, Inc. custom 24 V, 4.2 A, providing
an average peak current (varying with heating rate) of 0.66 A.

5. ND Converter. Model 203, 20-bit data acquisition with RS-232

connection (Lawson Labs, Inc., Malvem, PA)[7]

During the installation of the preconcentrators, the 6890 GC injector port
and Agilent 5973 MS detector functions remained unchanged. Injection was
controlled by the respective preconcentrator, while Chemstation software
continued to control the gas ﬂow, GC oven temperature, and detector functions.
The RVM preconcentrator components were powered and controlled
independently. The HSA-SPME element was heated by a separate, 24 V power
supply. The focusing preconcentrator functions were controlled by the RVM A68
using a separate personal computer and DOS-based software written in C++

version 1.52 (See Appendix B).

2.3.5 GC Capillary Column
Laboratory chromatographic separations for the Entech baseline and RVM
experimental studies were accomplished using a commercially available 30-m

DB-5MS low thermal mass (LTM) capillary column (RVM Scientiﬁc, Inc.) The

45

leading column end was installed in the third port of the three-way low dead
volume tee containing the terminal ends of both preconcentrator transfer lines
previously described. The DB-5MS column had a nominal inner diameter of 250

pm and a stationary phase thickness of 0.25 pm.

2.3.6 Dynamic Gas Generator

Gas dilutions were performed using Kin-Tek Models 491 M-B Precision
Gas Standards Generator and DGB 491M Direct Gas Blending Module (La
Marque, TX). The two modules used in tandem performed controlled, dynamic
mixing of ﬂowing gas streams of diluent air and component gas (T O-14A). Gas
mix humidiﬁcation was provided by the Kin-Tek model 491 M HG Humidiﬁcation
Module. This module added moisture to the ﬁnal mixture by controllable
saturation of a portion of the dilution gas before addition to the mixture [8]. The
resulting diluted, humidiﬁed outlet gas ﬂow rate was monitored and veriﬁed by an
ADM 3000 Flowmeter from J 8. W Scientiﬁc, Inc. (Folsom, CA). Relative humidity
and temperature of dilution gas was measured using a Humidity Stick (SKC lnc.,
Eighty Four, PA). Final diluted gas samples were captured in 5 L Tedlar bags
furnished by SKC, Inc.

2.3.7 Sampling Pump
To achieve sampling at high ﬂow rates comparable to canine capabilities,
a GILAlR-5 Portable Air Sampling Pump (Sensidyne/Gilian lnc., Cleanrvater, FL)

was employed to sample Tedlar bags containing the gas dilutions. For sampling,

46

one end of the HSA-SPME element was attached to the outlet of the GILAlR-5
sample pump to capture VOCs from a Tedlar bag attached to the pump inlet

(See Figure2.10).

 

Figure 2.10. The Gilair-5® used in sampling Tedlar bags
at 4 Umin into the HSA-SPME element
The pneumatic system of the GlLAlR-5 contained a motor driven, dual
piston pump, powered by a 6.0 Volt, 1.8 Ampere-hour Nickel Cadmium battery,
capable of ﬂows up to 5,000 scc/min. Unlike stainless steel air sampling
canisters which must be evacuated before sampling, the use of Tedlar bags for
gas dilutions permitted replication of actual ﬁeld sampling protocol at standard

atmospheric pressure.

2.3.8 Digital Multimeter

A Metex digital multimeter, model #M3860-M, was used to monitor power

47

consumption of the Entech and RVM Adaptive Sampler components during
preconcentration events. Temperature measurements were performed with a
thermocouple (Type K, Omega Engineering, Stamford, CT) connected to the
multimeter. The multimeter was equipped with an RS-232 serial link and
MultiView version 2.60 software to enable data collection with a personal
computer [7]. Power readings were collected directly in terms of Watts (W) using

an adapter rated at 13 amps and 120 AC volts.

2.4 Reagents

A US. EPA compendium Toxic Organics (EPA TO-14A) 39-component volatiles
mix, each certiﬁed at 1 ppmv nominal concentration, was obtained from Restek
Corp. (Bellefonte, PA). Table 2.4 lists the components of the gas mix, which
ranges from light gases (e.g. Freons) to substituted benzenes. Table 2.4 also
lists each component’s respective molecular formula, expected retention time
and molecular/qualiﬁer ions to be quantitated during analysis. (Note: these
retention times were obtained from US. EPA method TO-14A and represent
analysis by conventional GC columns; they do not reﬂect shortened retention
times from LTM-GC columns incorporated in this study). Ultra-pure air carrier
gas furnished by Airgas (Radnor, PA) was used as the sample gas diluent. The
Entech 7100 preconcentrator required liquid nitrogen and ultra high purity
(99.999%) nitrogen gas obtained from Air Products (Allentown, PA) for
cryofocusing purposes. All analytical separations were accomplished using GC

grade helium (99.9995% purity) with built-in puriﬁer (Air Products, Inc.).

48

Table 2.4. EPA TO-14A, 39-Component Gas Mixture (1 ppmv nominal each)

 

Target Analyte

Molecular
Formula

Retention
Time
(min)

Ion/Abundance
(amul % base
Peak)

 

—l

. Dichlorodiﬂuoromethane*

CCI2F2

5.00

85/100
87/31

 

2. Chloromethane

CH3C|

5.34

50/100
52/34

 

*

w

. Dichlorotetraﬂuoroethane

CzCI2F4

5.46

85/100
135/56
87/33

 

4. Chloroethene

CzH5Cl

5.71

62/100
27/125
64/32

 

UI

. Bromomethane

CH3Br

6.40

94/1 00
96/85

 

OI

. Chloroethane

CzH3C|

6.67

64/100
29/ 140
27/ 140

 

N

. Trichloroﬂuoromethane*

CCI3F

7.59

101/100
103/67

 

. 1 ,1-Dichloroethene

CzH2C|2

8.41

61/100
96/55
63/31

 

<0

. Methylene Chloride

CHzClz

8.64

49/1 00
84/65
86/45

 

10. Trichlorotriﬂuoroethane

CzCI3F3

8.71

151/100
101/140
103/90

 

11. 1,1-Dichloroethane

C2H4CI2

9.94

63/100
101/140
103/90

 

12. (Z)-1,2-Dichloroethane

C2H4CI2

10.73

61 /1 00
96/60
98/44

 

13. Chlorofonn*

 

 

CH3C|

 

11.00

 

83/100
85/65
47/35

 

49

 

Table 2.4. (Cont’d.)

 

Target Analyte

Molecular
Formula

Retention

Time
(min)

lonlAbundance
(amul % base

POGKI

 

14. 1,2-Dichloroethane

CzH4CI2

11.90

62/100
27/70
64/31

 

15. 1 ,1 ,1 -Trichloroethane*

02H3Cl3

12.03

97/100
99/64
61/61

 

16. Carbon Tetrachloride“

CC|4

12.51

1 1 7/100
199/97

 

17. Benzene"

Cch

12.58

78/ 1 00
77/25
50/35

 

18. 1 ,2-Dichloropropane

C3H50l2

13.39

63/100
41/90
62/70

 

19. Trichloroethylene*

C2HCI3

13.44

1 30/1 00
1 32/92
95/87

 

20. (Z)-1,3-Dichloropropene

C3H4Cl2

14.38

75“ 00
39/70
77/30

 

21 . (E)-1 ,3-Dichloropropene

C3H4Cl2

14.99

75/1 00
39/70
77/30

 

22. 1 ,1 ,2-Trichloroethane

CzH3Cl3

15.21

97/100
83/90
61/82

 

23. Toluene*

CrHa

15.46

91/100
92/57

 

24. 1,2-Dibromoethane

C2H4Br2

16.27

107/100
109/96
27/1 15

 

 

25. Tetrachloroethylene*

 

CzCl4

 

16.48

 

166/100
164/74
131/60

 

* Denotes most signiﬁcant volatile Chemicals detected at the surface of 1-yr

old human burials

50

 

Table 2.4. (Cont’d.)

 

Target Analyte

Molecular
Formula

Retention
Time (min)

Ion/Abundance
(amul % base

Peak)

 

26. Chlorobenzene

CeH5C|

17.47

1 12/100
77/62
1 14/32

 

27. Ethylbenzene*

C8H10

17.75

91/100
106/28

 

28I29. m,p-Xylenes*

C8H10

18.02

91/100
106/40

 

30. Styrene*

C8H8

18.51

1 04/100
78/60
1 03/49

 

31. 1 ,1 ,2,2-Tetrachloroethane

CszCl4

18.60

83/100
85/64

 

32. o-Xylene*

C8H10

18.61

91/ 100
106/40

 

33. 1,3,5—Trimethylbenzene

CeH12

20.19

1 05/1 00
1 20/42

 

34. 1,2,4-Tn'methylbenzene

09H12

20.70

105/1 00
1 20/42

 

35. 1,3-Dichlorobenzene

06H4C|2

20.93

146/100
148/65
1 1 1/40

 

36. 1,4-Dichlorobenzene

CeH4C|2

21.01

146/100
148/65
1 1 1/40

 

37. 1,2-Dichlorobenzene

CeH4CI2

21.44

146/1 00
148/65
1 1 1/40

 

38. 1,2,4-Trichlorobenzene

09H12

23.80

1 80/1 00
1 82/98
1 84/30

 

39. Hexachlorobutadiene

 

 

C4C|6

 

24.23

 

225/1 00
227/66
223/60

 

* Denotes most signiﬁcant volatile chemicals detected at the surface of 1-yr

old human burials

51

 

2.5 Experimental Procedure
2.5.1 EPA TO-14A Gas Mix

A set of experiments were performed to develop a method to illustrate the
functionality of the new tandem HSA-SPME element/microtrap design. A US
EPA method TO-14A gas was used as the test mix because although the this
mix is normally used for environmental purposes, Vass et al. has demonstrated
that approximately 14 of the 39 target compounds included in the TO-14A gas
mix have also been identiﬁed as major components in decomposition of 1 year-
old buried human remains [9]. These components are noted in Table 2.4. Thus
the continued study of these speciﬁc compounds is relevant in law enforcement
and forensic science initiatives.

The Entech 7100 Air Concentrator was used as a benchmark to provide a
performance measurement against the experimental RVM Adaptive Air Sampler
preconcentration system. Initial calibration of the Entech 7100 Preconcentrator
was performed using a series of dilutions (0.5-32 ppbv, 60 t 10% rel. humidity) of
the TO-14A gas mixture to establish comparative sensitivity data to that of the
performance of the RVM Adaptive Air Sampler. Each calibration point was
performed in triplicate with corresponding responses averaged. A 13-minute
preconcentration time was necessary for proper operation of the Entech 7100,
which involved cryotrapping at ‘150°C, a standard laboratory practice for the
concentration of polar and non-polar light gases. The RVM Adaptive Air Sampler
required considerably less concentration time than the Entech Since the dual

adsorbents trap at ambient temperature and does not use cryogen.

52

Although both systems are capable of trapping at ﬂow rates of 200
mL/min, the additional HSA—SPME element on the front-end of the RVM permits
trapping at higher ﬂow rates of up to 5 Umin as well. Testing of the RVM
Adaptive Air Sampler was accomplished using gas dilutions approximately
equivalent to benchmark dilutions, as well as utilizing two contrasting ﬂow rates:
one calibration was performed with sampling at 0.3 Umin using only the focusing
preconcentrator microtrap, and another at 4 L/min utilizing both the HSA—SPME
element and focusing preconcentrator in tandem. The two ﬂow rates represent
laboratory and ﬁeld sampling situations, respectively.

Evaluation of the RVM Adaptive Air Sampler involved a comparison study
between the benchmark and HSA—SPME element/microtrap in tandem, as well
as with the microtrap sans HSA-SPME element. Studies included desorption
and extraction time proﬁles, analytical sensitivity, limits of detection and overall
power consumption. Analyte recoveries from the three sampling modes were
compared over 1-minute sampling times for each mode. Power consumption
data was collected with the Metex multimeter during multiple analyses for both
the benchmark and experimental concentrators. A quantitative method
comparison between HSA-SPME element and benchmark systems was
performed using uniform parameters, as well as a sampling ﬂow rate of 0.2 Umin

for both systems. These method parameters are listed in Table 2.5.

53

Table 2.5. LTM-GC/MS method parameters for the TO-14A
experimental gas mix

 

 

 

 

 

 

 

 

 

 

 

 

AGILENT GC oven
Temperature 210° C
LTM-GC TEMPERATURE PROGRAM
Initial Temperature 30° C
Initial Hold Time 2.00 min
Temperature Rate 1 20° C/min
Intermediate Temperature 1 175° C
Hold Time 0.00 min
Temperature Rate 2 60° C/min
Final Temperature 210° C
Final Hold Time 0.00 min
INLET

Mode Splitless
Temperature 250°C
Carrier Gas Helium

COLUMN PROGRAM
Mode Programmed Pressure
Initial Pressure 28.00 psi
Initial Hold Time 2.00 min
Pressure Rate 1 2.02 psi/min
Intermediate Pressure 1 38.11 psi
Hold Time 0.00 min
Pressure Rate 2 6.01 psi
Final Pressure 2 46.10 psi
Final Hold Time 0.00 min
Average Linear Velocity 53 cm/sec

MSD SETTINGS
Acquisition Mode Scan
Scan Range 29-180 m/z, 0-1.5 min
34-280 m/z, 1.5-11 min
Scan Rate 14.64 Hz, 0-1.5 min
10.13 Hz, 1.6-9.5 min

Threshold 150
Sampling (2”) n = 1
Solvent Delay 0.00 min
MS Quad Temperature 150°C
MS Source Temperature 230°C
MS Transfer Line
Temperature 200°C

 

 

 

2.6

References

Mustacich, R. and A. Neushul, Adaptive Sampling Technology Final
Report. 2003, RVM Scientiﬁc, Inc.: Santa Barbara.

Ciucanu, I., et al., Helical Sorbent Microtrap for Continuous Sampling by a
Membrane and Trap Interface for On-Iine Gas Chromatographic
Monitoring of Volatile Organic Compounds. Anal. Chem., 2003. 75(4): p.
736-741.

Entech, Entech Instruments 7100 Operators Manual. 1997: Simi Valley,
CA. p. 1-4.

Sloan, K., Method Development and Implementation of a Novel
Temperature-Programmable Gas Chromatograph for Rapid Forensic
Analysis in the Field, in Columbian School of Arts and Sciences. 2001,
George Washington University: Washington, DC p.47.

Tuduri, L., V. Desauziers, and J. Fanlo, Potential of Solid-Phase
Microextraction Fibers for the Analysis of Volatile Organic Compounds in
Air. Journal of Chromatographic Science, 2001. 39: p. 521-529.

Shirey, R., SPME/GO Analyses of Sulfur Gas and VOCS, Using a New
Carboxen/PDMS Fiber. Vol. 16. 1997, Bellefonte: Supelco, Inc. 7-8.

Eckenrode, B., K. Sloan, and R. Mustacich, Development and Evaluation
of a Low Thermal Mass Chromatograph for Rapid Forensic GC/MS
Analysis. Field Analytical Chemistry and Technology, 2001. 5(6): p. 288-
301.

Kin-Tek Laboratories, l., Model 491M-B Operation Manual. 2001, Kin-Tek
Labs: La Marque, TX. p. 1.

Vass, A., et al., Decompositional Odor Analysis Database. Journal of
Forensic Sciences, 2004. 49(4): p. 1-10.

55

CHAPTER 3 - RESULTS AND DISCUSSION

3.1 US EPA Toxic Organics (T O-14A) Gas Mix
3.1.1 Baseline Gas Chromatograph

The TO-14A gas mixture was analyzed by the Entech 7100 benchmark air
preconcentration system using the parameters listed in Table 2.5 to determine
optimal air concentrator conditions for analyte quantitation. The baseline total
ion Chromatograph of the TO-14A components separated by the LTM-GC
installed on the Agilent GC oven door is shown in Figure 3.1. Peak assignments
were conﬁrmed with the mass selective detector (mass spectra not shown).
Total baseline analysis time for the 39-component mixture was 22 minutes,
including 13 minutes for sampling and preconcentration, and 8 minutes for
chromatographic separation. The average analyte peak width at half height
(PWHH) was 0.83 seconds, with a peak tailing ratio of 1.0 (ratio of area right of
peak center to area left of peak center, Gaussian = 1).

Analytes experiencing co-elution are shown in Table 3.1. Although MSD
data acquisition was performed in scan mode, co-elution was of minimal
consequence since chromatographic analysis was performed by quantitating the
area of each analyte molecular and qualiﬁer ion peak (listed in Table 2.4) rather
than total ion Chromatograms. Co-elution of m-Xylene and p-Xylene components
required quantitation as a single peak, since these structural isomers exhibited
identical molecular and qualiﬁer ions, and could not be resolved based upon

differences in the ion scan.

56

ESBEEE a8
.2 4N «3...; 8m £93325 8:32 933:. was 8.32, 23.2 3590 2x8 Eu .3. .3 2.5...

 

 

.383 93 anmm . 3.5 E:

a h o m v n N r
oo+mod

$9..ch

02mg?

owepunqv

¢°+wcd

OO+NOd

 

heme. —.

23...: 30 (v TOP (mm .m.: . Louuhuceocoeoi .3 {25.055

 

 

57

Table 3.1. Analytes exhibiting co-elution in the Entech baseline

 

Chromatograph
mlz Retention Time
Co-eluted Analytes Quantitated (min)

1. Dichlorodiﬂuoromethane 85,87

2. Chloromethane 50,52 1 66-1 68
3. Dichlorotetraﬂuoroethane 85,135.87 ' '
4. Chloroethane 62,27,64

16. Carbon Tetrachloride 117,119 2 94
17. Benzene 78,77,50 '

21. Toluene 91,92

22. (E)-1,2-Dichloropropene 75,39,77 “1'4”
28. m-Xylene 91,106 5 21
29. p-Xylene 91,106 '
#30. Styrene 1O4,T3,1O3

31. o-Xylene ‘ 91,106 5°43'5'44

The resulting Chromatograms of the TO-14A analytes sampled and separated by
the benchmark air concentrator/LTM-GC system exhibited results acceptable to

initiate experimental air concentrator analysis.

3.1.2 Experimental Desorption Tlme Profiles and Gas Chromatograms
To better characterize the adsorption capabilities of the HSA-SPME
element design, two separate analytical studies were performed on the
experimental RVM Adaptive Air Sampler. The ﬁrst study consisted of 4 Umin
gas sampling by the HSA-SPME element, followed by desorption onto the
focusing preconcentrator microtrap. This contrasted with a second study

performing 0.3 L/min gas sampling by the microtrap alone. Both studies

58

maintained 1-minute sampling times to replicate ideal, rapid sampling executed
in ﬁeld situations.

Desorption time tests for the HSA-SPME element and microtrap were
required to optimize a number of variables, including quality of chromatographed
analyte peaks, sorbent bleed, and total run time. Optimal desorption conditions
would require a short analysis time (8-10 minutes, including sampling), with a
minimal amount of sorbent bleed, followed by low analyte canyover on the SPME
element and microtrap sorbents. Initial tests were performed on the lone
microtrap to determine proper desorption time and temperature necessary to
remove from the trap any analyte carryover. Once these parameters were
established, they were used in the microtrap for desorption time tests for the
HSA—SPME element to ensure that a majority of observed carryover and sorbent
bleed came from the lone HSA—SPME element.

Microtrap Desorption Time Profile

In order to minimize sorbent bleed, it was found that the microtrap required
relatively short desorption times (a few seconds) at maximum temperatures for
the particular sorbent phase (330-385°C). Longer desorption times at lower
temperatures failed to effectively reduce carryover, and the resulting peak tailing
prevented accurate analyte quantitation. Using longer desorption times while
increasing desorption temperature reduced carryover, however at the expense of
increased sorbent bleed. A desorption temperature of 385°C sustained for 1.5
minutes caused complete sorbent breakdown in the microtrap, requiring

replacement. It was determined that ﬂash-heating the microtrap up to its highest

59

allowable temperature (330-385°C) for 10 seconds would offer comprehensive
analyte desorption without sustaining a large amount of sorbent breakdown.
Since the microtrap contained thermocouples within the heating wire coil, rapid
and controllable temperature ramping was possible without overshoot (Figure

3.2).

 

Programmed and Actual Voltages During Microtrap

 

 

 

 

 

 

 

500 Desorption

400 ‘ - —Actual Tenperature
MK — Target Temperature

300

 

 

 

Terrperatune
'S’
/

 

 

 

 

 

 

 

 

 

O

I I I I

01:00 01:30 02:00 02:30 03:00 03:30
Time (nin)

 

 

 

Figure 3.2. Diagram showing a 10-second temperature ramp typically
observed during microtrap thermal desorption. The actual starting
temperature is higher than the target temperature due to heat
emanating from proximal heated GC components.

Chromatograms illustrating microtrap sample loading and the subsequent
carryover blank at optimum conditions are shown in Figure 3.3. The top

Chromatograph displays 38 of the 39 components trapped, along with 2 bleed

peaks and one sample bag remnant peak. Although large sorbent bleed peaks

60

.0023 CO 939382 E:E_me m E $803 :9 m_ 5:90on .xcﬂn Lc>oEmo
E63333 ucm agoﬁE Loumbccocooca 9:258 9.: E mEEEmm new be mEEmoumEezo Cc>mtc>o .n.n 2:9".

 

 

 

 

 

 

 

 

 

 

 

 

 

AEEV DEE.
N we Ndr Nd Nd «N No N6 «6 «6
i4 :1 . III: . . «<1 e 1 It . 808a
mo+uoo.—. n «02.5
I Iii III? mo+m°o.—.
1. :
mo+woo.u
- mo+mo9n
OPEL
E33: w 9. 53> JEEm Lo>o>tao i
Ens—=01 a
a a a 5n. . mo+mooé
an o. Euw .3 55:6: 0: 65.30.. QmLuoLosi
too—m «coeom a
. ti. .1 t.
L l 808...”

 

 

 

82:8 F .8 55: no O 55.958
Ea... 3 5: «.3 STOP - 359.25.

J

 

 

61

are Observed upon loading, they are signiﬁcantly reduced in the carryover blank
by the addition of a 10-second helium purge after microtrap sample loading but
before thermal desorption. The helium purge decreases oxidation and
subsequent bleed of sorbent by effectively removing oxygen from the trap before
desorption.
HSA-SPME Element Desorption Time Profile

At the time of initial HSA-SPME element desorption tests, thermocouples
were not installed on the heating wire; heating was controlled by an 18-ohm
resistor placed in-Iine with the heating wire. A maximum sustained temperature
in this conﬁguration was 294 i 2°C after 4 minutes; however, the minimum
temperature of 250°C required for achieving proper analyte desorption from the
Carboxen/PDMS phase was not reached until 0.5 minutes after thermal initiation.
Even after a 4-minute desorption time, carryover was still present, and sorbent
bleed was very high. Based on the favorable desorption characteristics observed
with ﬂash-heating the microtrap, the resistor was removed from the HSA-SPME
heating wire, and replaced with a 24V battery with manual power switch. A
thermocouple was added to the inner lumen of the SPME tube and connected to
a multimeter to observe HSA-SPME element temperature ramps (°C). With the
resistor removed, the SPME wire could be ﬂash-heated to 350-400°C in 3-5
seconds and immediately switched off. In this conﬁguration, both carryover and
sorbent bleed remained minimal over successive trial runs. Figure 3.4 illustrates
the desorption temperature proﬁles for the resistive and non-resistive power

ramps. By driving off analyte from the solid sorbent materials by quick bursts of

62

heat, the desired run times were realized, and carryover was minimized.
Chromatograms contrasting optimal and non-optimal desorption times using the

HSA-SPME element and microtrap in tandem are illustrated in Figure 3.5.

 

Programmed Power and Actual Temperature Ramps - With and

 

    
 

 

      
   

Without Resistance
450 — - , - - - 60
—Temperature - 18 Ohm Resistance

400 —Temperature - No Reslstance 50
3* 350 —Power - 18 Ohm Resistance 3
:5
— .1 - . . - , . — — . . E
4: 300 . .. _ . -. . W” . 4o 9.
9' ';, .2,“ ”fits” r, < “ ' O
2 250 g
3 30 g
g 200 3
3. '9.
E 150 20 g.
a
" 100 E

10
50
0 0
00:00 00:30 01:00 01:30 02:00 02:30 03:00 03:30 04:00

 

Time (min)

 

 

Figure 3.4. Diagram showing both resistive and non-resistive power ramps and
their reﬂective temperature profile. Higher temperature is achieved
in only a few seconds without resistance, thereby reducing overall
analysis time by 2 minutes.

63

mcoEocoo .mEsao 595 E 9.me 8&3463 .w> .mcozﬁcoo 5:90on .mEzao
.8: 5 8520250 ucm .mc___S xmoa .mmEz 8:532 @3385 9:962: mEmumoﬁEoEo mcsmmbcoo .m.n 2:9”.

 

 

E25 2:: A
t m. 2 3 m h m n F

: :1 i _ 4d a < _ < _ _ _ 8+m8.o
—1 EEDv @ >25 2: BEE-ham

o omNN @ 55. N "oEF Eamon 983.25.
0 gnaw @ 55. v ”25... £9.39 NSF—$.43: I oo+mcoé
”.mEzaqcoz

 

 

 

 

 

27mm. r u «onto

‘ a“ €37] gas.“

 

 

   

m.
T 8+m8d w
m

 

EEDV @ >93 9 65.956
0 comm @ n or BEE. £030 catches.

 

 

 

 

i + n
0 use @ m m "05.... 9.080 winwu<mr we woo v
".2530
, oc+moc.m
wo+wood

 

mbwuwsﬂbﬂn—
5:983 $553.52 ER .5530 3.25 8.25 saws 3:92 - <39 Sum .mj

 

 

Optimized Experimental Gas Chromatograms
Under optimized desorption conditions for the HSA-SPME element and

microtrap, total experimental analysis time, including a 1-minute gas sampling
time followed by chromatographic separation and detection, was 9.5 minutes,
with an average PWHH of 0.95 seconds and peak tailing ratio of 1.2 (See Figure
3.6). Of the 39 components in the TO-14A gas mixture, analytes Chloromethane
and Chloroethene were not detected by the Agilent 5973 Mass Spectrometer
when gas sampling was performed by the HSA—SPME element, and only
Chloromethane went undetected under lone microtrap sampling. These
components are known to be poorly retained by the solid sorbents used in the
microtrap [1]. Longer retention times (~1.5 minutes) observed by the
experimental system relative to the benchmark system are due to the fact that
mass spectra capture was initiated at the onset of experimental HSA-SPME
desorption/microtrap sampling, prolonging retention by 1.16 minutes over the
benchmark system, which remotely initiated MS capture only after a 13-minute
sampling and preconcentration period.

Analytes from the experimental system experiencing co-elution are shown
in Table 3.2. Despite the increased retention times, both experimental studies
appeared to give acceptable peak widths and tailing comparable to that of the

baseline system.

65

.cozmoEEoE xmma .8 vN mam... com .8296 cozmbcoocoo :w .958thon 5.2.8 .2

m>=amu< _>_>m 05 3 8:28am 933E 8:59 8:590 2x8 <n_m_ .mj m v6 mEEQOEoEo 83:90 .od 959".

 

 

 

 

 

 

 

 

 

2:52.:
a a h a
—h ‘r __ —“ p
j X
en
8
8 3 a
8
m
a mu 8
5 N; 8
8 mm
8.3

 

 

om

 

3.3

5.0.:

 

 

 

55: e © mimic: 5? 3855'

5:5 «6 © nahuocoEI

 

 

€25; $8 © 2:52 F .2

5:. mam «3-9 E55 5% or so 3358

 

 

9.3.2.2 mam <36... (mm .m.: . boas—am .__< ¢>=nau< E>¢

N
7 oo+mod

, oc+wo._.

oo+wo.N

, oc+mo.n

co+moé

 

ii i mc+mod

esuodseu

 

 

66

Table 3.2. Analytes exhibiting co-elution in the RVM experimental

 

Chromatograms
mlz Retention Time
c°'°'“t°d Ana'ytes Quantitated (min)
16. Carbon Tetrachloride 1 17,1 19
4.44
17. Benzene 78,77,50
21. Toluene 91 ,92
4.75
22. (E)—1,2-Dichloropropene 75.39.77
28. m—Xylene 91,106
6.60
29. p-Xylene 91 ,106
31. o-Xylene 91,106
6.84
32. 1,1,2,2-tetrachloroethane 83,85

Desorption Efficiencies

Upon establishing proper desorption parameters for the RVM
experimental system, desorption efﬁciencies were determined for the HSA—SPME
element, microtrap, and baseline systems (See Table 3.3). Here, a humidiﬁed,
100 ppbv (nominal) TO-14A gas mix was sampled (n = 5) and desorbed by the
respective concentrator, and resulting analyte peak areas quantitated. Blank
samples were then run to determine the amount of sample carryover from the
previous desorption. The benchmark system performed optimally, with minimal
carryover for most analytes (average desorption = 99.7 :t 0.7%). The RVM
focusing preconcentrator microtrap had equivalent results (average desorption =

97.5 :I: 2%). The average desorption for the HSA—SPME element

67

 

 

 

 

 

 

 

 

 

 

 

EN 8 3 3 o.o 02 8 18 2.3288229“;
3 a 2 mo. to 2: 8 <8 2328
an 8 so 2: od 8* 8 2: ”28202.3 8&8
no 8 mo 8 od 2: 8 5: 22.8.2654:
3 5 mo 8 o.o 8F 2: 98 «answegzgos;
5 3 so 3 o.o 2: 8 :0 5.8925
3 8 od 8 ad 2: 8 «.8 «558.2;2onfﬁ
P. 3 mo 8 od 2: 8 2m ocasoeozgoss
m...” 8 3 um od 2: 8 Es. 22.883.222.35
an 8 so 8 od 8? 8 Q8 8:25 2.238:
no 8 so 8 9o 2: 8 En 22.8.2554;
9m mm m; mm 0.0 cow om wa o:~£oEo.o=:o.oEu..F

.. oz to 2: o.o 2: S was «$588.8
3 5 3 mm od 2: em 3 acaeososem
«N 8 to on: o.o 2: mm v.2- 22.8.25
3 8 mm 8 od 2: 08 E. .. aunts”, : L

.. oz -- oz to 8s 2: New- 8285225
5. 3. No 2: 9o 2: 3 now. 828528538205

lane ﬁns ﬁns 2...: 8.;
E E g .

own minim: om”. assess. own c.3588 _aw.__%_wo have? £35893

 

 

 

nausea 82 n oz

285% ozomadzv 5:: .3 Ba $9655 55: no .Axaeﬁcgv as: No © 855th @5353
mom <36» cmEEOE Boo 09 992588 E... .mEoE:ooxm com xomEcocoo .8 3.82qu 5:90on dd sin...

68

 

 

 

 

 

 

 

 

 

 

 

 

a.» on no 8 od 3 8 8. a . - 1 .u
e: 2 mo 5 .od, 5 8 $3 gasseozozhtus
3 B to 8 no 8 8 v: . 23:303_ﬁ_o.~;
m.» 8 No 8 to 8 8 38 83:322554;
3 8 no 8 we 8 8 at 235852293
N; 3 No 8 to 8 8 $2 a5~=3_>._$s_ﬁ.v.~.s
3 3 No 2: No 2: 2: 5.2 23592355....5;
3 «a mo 8 3 2: NS «.3: 9.233
3 8 to 2: No 2: 8 «.9: a:2.oo.o_._ue.£.~.~€.
E 8 no 8 No 2: 8 no: 225m
3 «a. No 8 No 8. NS 39 853x95
w... mm md mm to cor Nor N62 season—Em
3 8 so 8 to 2: «.9 NS 8358.25
ms 8 mo 3 to 2: Es :9 222822585
5 8 so 2: od 2: mm 32 25585225-?
N; 8 no 8 no mo 9: so: 23.3
mo 8 ad 8 9o 2: 2: as: gasseeﬁEaﬁs
ca 5 3 8 to 02 8 m: 28288654.?sz
to 8 to 2: o.o 2: mm 32 23052255-”. 15
no 8 o. s S to 2: m2 3 Swansea‘s:
ﬁns ans ﬁns 2%: 6...
3.. 3.. g .
new”. m2nm.<m: oMm 3.39.3.2 QM”. xouﬁsocom .mwwmwo mun—“Wm 8222 $93

 

 

3:58 89%? 8:988 a...” 033

69

however was 93.3 :I: 4%, which was expected to be less than the average
microtrap desorption efﬁciency since analyte must be further concentrated on the
microtrap before ﬁnal GC injection. Nevertheless, the experimental desorption
efﬁciencies are found to be at acceptable levels based upon data from the

Iiteratu re [4].

3.1.3 Extraction Time Proﬁles

Having established proper desorption parameters for the RVM
experimental system, it was necessary to determine the time required for the
extraction process to reach equilibrium during high-ﬂow sampling with the HSA-
SPME element. Normally, 1-cm SPME ﬁbers require long periods of time (5
minutes to 2 hours) for headspace extractions, and shorter times for agitated, or
low-ﬂow samples (5-10 minutes @ 4-5 ml/min). With a 10-cm length, helical
HSA-SPME-coated wire under high ﬂow (4 L/min), it is possible to load a much
larger amount of analyte mass onto the ﬁber in a shorter amount of time. For this
analysis, a humidiﬁed, 100 ppbv (nominal concentration) TO-14A gas mixture
was sampled at 4 L/min by the HSA-SPME element for predetermined times, up
to the one-minute sampling time threshold. Individual analyte responses were
plotted against time to determine extraction time proﬁles (See Figure 3.7).

Since this particular study concentrates on a more qualitative approach
towards fast ﬁeld sampling, the focus was to determine the approximate amount
of sample which could be loaded on to the SPME phase within a one-minute time

interval. As can be observed in Figure 3.7, most of the early eluting light gases

7O

x_E mam (3-9. E05 >93 oop a E0: moan 9:: cozomﬁo 929:» *o c2828 025E823”. Hm 2:9".

 

ASE. 05:. Eauaxm
r 0.0 66 1.0 «.0 o

cc+mcod

 

oca£§o=ﬁo.5tp n5!
ocgoeaét§25=71u
8320 messaging...
assassins. V II:
ocuSoSoEoEmv. _. III
oconconogoEothtu. F Ill
E32220 IT
ocacu§o_;o_o-~. TE 10..

mo+moc.m

Demo? —.

 

o:a£8.o_:o_o.~.7.l ._ 8&8...
oconconssmeﬁédé LT v
oco_>x-o
ocoNcomnll 8&8..."
ocoEOhlbl

 

 

 

  

0%. on EEDv © msz.<m: .msasum.
3.2.2:; .2... no.
dam (SHOP 29:—:05 >93 2:. SEE—«mm 8E8.“

   

 

 

 

71

5.: wow <3-O.r E55 Ban 8.. m Eat onoa 0E: cozogxo 929...... do 5:028 o>=8comoaom Hm 2:2".

 

AEEV 05.... 0.500an
r ad ad v.0 «6 9
8.52.6

<\I(II.II.| (

fllfllllllllflllul‘ 1r

\I-

mo+mo9m

8+mco. r

co+mowz r

 

 

 

 

 

 

co+maod

 

”a”, an.“ m ,. Uowdn .EEDc msEm. w: ”mix—Ea”M
, w... 35.53: {undo
,. dam <3.-O._. ABE—:05 >nnn oer "oEEum

 

 

 

 

8.723...“

1

 

 

 

71

reach equilibrium (noted by reaching a plateau) within 0.25-0.50 minutes.
Medium-molecular weight gases reached equilibrium in 0.65-0.85 minutes.
However, many of the late.eluting compounds, including the substituted
benzenes (toluene, o-Xylene) continue their upward slope at the one-minute
mark. On a quantitative basis, these compounds would require longer extraction
times to reach equilibrium between the air matrix and SPME phase. Since rapid
ﬁeld sampling is sought, long extraction times are not necessary or desired. It is
therefore possible to choose shorter extraction times for those analytes whose
equilibration times are longer than one minute [2]. Sample ﬂow rate,
temperature, extraction time, and humidity should be controlled in order to obtain
reproducible data. It is of course not possible to control slight changes in
temperature and humidity in a dynamic environmental setting. For this reason,
sample-to-sample precision in quantitation may not be achieved in HSA-SPME
ﬁeld applications. To obtain better precision, simultaneous sampling with
multiple HSA—SPME elements with the same phase and phase thickness may be
implemented. For purposes of obtaining qualitative data for a compound in rapid
ﬁeld sampling, a representative signal can be achieved within one minute for 100

ppbv concentrations for all compounds listed in the TO-14A mix.

3.1.4 Calibration Curves
Based on SPME theory, the amount of analyte extracted from the sample
under set conditions should be directly proportional to the concentration of the

analyte in the sample [2]. Therefore, plotting the analytical signal of the analyte

72

vs. the known concentration should produce a linear calibration curve. In this
study, a 7-point calibration curve (0-32 ppbv nom, moisture added) was
performed with both of the Entech benchmark and RVM experimental air
concentrators. Again, sampling ﬂow rates respective to the particular
concentrator (0.2 L/min for the Entech, 0.3 lein for the microtrap, and 4 Umin
for the HSA-SPME element) were utilized, with extraction times remaining
constant at one minute. Each calibration point was performed in triplicate with
the ensuing responses averaged. The resulting analytical signals (y) at the
respective concentration (x) were plotted by the least squares technique to obtain
a line of linear regression using the equation y = ax + b, where a is the
regression slope, and b is the y-lntercept (See Tables 3.4 and 3.5). Using the y
residuals from each plot, the random errors were calculated for a and b at 95%

conﬁdence intervals [3].

Linear ﬁt of the data was quantified in terms of the R2 value, which
represents the variance in y attributable to the variance in x [4]. The R2 value is

a measure of linear ﬁt ranging from 0 to 1, with a perfect linear ﬁt having an R2

value of 1.000. The linear ﬁt of the benchmark data and RVM preconcentrator
data are also listed in tables 3.4 and 3.5. With exceptions of the undetected

analytes Chloromethane, Chloroethene, and weakly retained bromomethane,

comparison between the benchmark and experimental HSA—SPME element R2

values indicate that the linear ﬁt for a given analyte calibration curve was about

the same (Table 3.4). The experimental focusing preconcentrator microtrap R2

73

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Emmd v H N md H Nd vmdmd m H m- md H Nd o:u53..0=.o.o-N...
38.0 m H N Nd H v.N Sodd o H m- vd H vd 5.82020
wvmdd 9 H .V dd H dd Nmmmd m H m- Nd H dd o:a£o€0...o.n-N. 1N.
ommdd N H No rd H md nmmmd d H o dd H Em 3553.25.94;
wmmmd md H md o H wd mwdmd m H T Nd H m6 o:a5o£o:=§o..o...ot...
38d m H m vd H mé Smad w H V. dd H md cateco c.8352...
mmmmd N H m Nd H mN Emmd m H m- vd H dd 323225.04...
mammd md H Nd mod H md oumad m H .V md H dd o:n£oES0=:22zoth
oz 02 02 28.0 N H ..- Nd H we 2853.5...0
Ngd md H wd vod H vod $8.0 N H F- rd H 0.. 2.2355505
mmmdd md H o Nod H md nomad 9 H m dd H B 2353.320
38d .d H rd mod H rd ommmd CON H F 3 H : 3289838633205
oz 02 oz Fded m H m od H md o:a£oE9.0...U
moomd Nd H wd mod H mod ommmd 3 H N- Ne H NF o:u£oE0..o==.uo.o.:oE
Nm 33.3....» 2.06 Nu. 333:...» 2.05 22.23 893
EoEo.o msdmém: eta—5.28m

85 5E 299.888 8 .888 mzaw-<wx 2.. c8 8588.. 8: 82cm 9.. 8. 8:8 888:8 .3” 28»

v

8.8.8 82 u oz

. or x momcoamo. 8.8209 ..< Ava mEQoEooo 5.6.9.8 @2883.

74

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Nam... N H m m... H m...- Nvmm... : H .w m... H 2. o823§.a£§.m.3
28... m H m no H m 88... N. H m- m... H m... 88.3.6
:8... N... H . to H N. 88... o H N- .8 H .8 a8...8.o...§.£.N.N.H.H
28... .V H .V H... H N8 NNmm... o. H m- N... H 8.... 225m
88... m: H t 3 H N. 88... N H o- 5 H t 88......8.
58... m H m N... H E Nam... m. H N- o. H N. o8~=3......m
88.8 N. H 2 Ho H 3 9.8... w H H- o... H m... 88:322....
88... m H m N... H N... 88... o H .V. .8 H m... a8..£.88=.§.£
38... o... H m... to H N... Nomad N H .- N... H oN 82.88553
88... : H : 8... H N... 3.8... S H m- m... H a... 2.8.8
38... N H to to H o... 5.8... m H N- N... H N a8...8.o...o..H-H.H.H
88... N... H to S H to 88... to H . to H to o8..o...o.o_..o.a.H.H-.m.
88... m... H o... to H N... .88... v H N- N... H I ¢:&0.No.o.;o.o.H.H-.N.
38... m H N N... H .8... N88... m H m- to H m...” 8838228.:
9.8... N H . N... H m. Nam... H H N- H... H wN 22.28.28.9N;
88... o. H t N... H E Ram... o. H N- N... H 3 28:8
N88... N... H o o H N... 88... m H m- to H o8 8.8.88.3 8&8
58... N H P H... H o8 RS... 8 H m- to H H... c8...oo._o...o.ﬁ.H.H.H
Nu. 2.3.2....» 2.2m Nu. ado-.82...» 2.05 03.22 Ema...
8.8m mama-<2. 882.8

8.80. .3" 2.3

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mammd Nd H No. Hd H 0; ONmmd N H o Nd H N.N «5.8839623wa
38... 06 H ed .8 H mé aommd w H .- md H m.N w:o~:aao.o...o..hé.N.w
Nmmmd m H m md H N.m ovmmd N H P- md H Qm acoucwao..o...o.o.N._-
586 N H or md H N.m FNmmd m H m- md H ob 8533032391..
ommmd N H m md H Qm 58... m H v. 0.0 H m... 339.320.50.06...
NNmmd m H o v.0 H Om EVA-wad :- H v. Nd H 0w o:m~=mn.>£m£.....-1N..-
Na 38.2....» 2.05 N”. 33.2...-» 2.2m 3322 «cm-5...
EoEwﬁ Mina-(w... mecocam

.888. .3. 22¢

76

Table 3.5. Calibration curves for the focusing preconcentrator microtrap with

correlation coefﬁcients (R2). All regression responses x 104

ND = Not Detected

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Target Analyte Slope y-Intercept R2
Dichlorodiﬂuoromethane 0.08 t 0.01 O :t 0.1 0.9933
Chloromethane ND ND ND
Dichlorotetraﬂuoroethane 0.6 t 0.08 -0.8 t 1 0.9875
Chloroethane 0.2 i 0.01 0 i 0.1 0.9994
Bromomethane 0.04 t 0.01 0.3 :t 0.2 0.9442
Chloroethene 0.2 i 0.03 0.1 t 0.4 0.9797
Trichloroﬂuoromethane 1.2 :t 0.1 -0.3 :I: 0.8 0.9984
1,1-Dichloroethene 1.1 :1: 0.02 -0.1 t 0.3 0.9997
Methylene Chloride 0.8 i 0.02 -0.01 :l: 0.3 0.9994
Trichlorotriﬂuoroethane 0.8 t 0.03 0.5 :t 0.5 0.9985
1,1-Dichloroethane 1.1 :I: 0.05 0.2 t 0.6 0.9987
(Z)-1,2-Dichloroethane 0.9 t 0.04 0.2 :I: 0.6 0.9983
Chloroforrn 1.2 :l: 0.07 0.3 :1: 0.9 0.9977
1,2-Dichloroethane 0.9 :t 0.03 0.1 :I: 0.4 0.9994
1,1,1-Trichloroethane 0.8 :I: 0.03 0.1 :l: 0.4 0.9989
Carbon Tetrachloride 0.2 :L- 0.01 -0.1 :t 0.1 0.9988
Benzene 1.8 :t 0.1 2 :I: 1 0.9983
1,2-Dichloropropane 0.6 t 0.05 0.5 t 0.6 0.9957
Trichloroethylene 1.2 :1: 0.1 0.9 1: 1 0.9959
(Z)-1,3-Dichloropropene 0.1 t 0.02 0.2 :l: 0.3 0.9698
(E)-1,3-Dichloropropene 0.02 :l: 0 0.07 :l: 0.1 0.9215
1,1,1 -Trichloroethane 0.6 t 0.05 0.5 :l: 0.7 0.9949
Toluene 1.3 1: 1.0 1 :l: 1 0.9975
1,2-Dibromoethane 0.4 t 0.04 0.2 :t 0.6 0.9896
Tetrachloroethylene 1.1 :1: 0.1 0.8 :1: 1 0.9952

 

77

 

Table 3.5. (Cont’d.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Target Analyte Slope y-lntercept R2
Chlorobenzene 1.4 1 0.08 0.8 1 1 0.9977
Ethylbenzene 1.9 1 0.1 1 1 1 0.9979
m/p-Xylenes 3.1 1 0.2 2 1 3 0.9963
Styrene 1.1 1 0.06 0.5 1 0.8 0.9983
1,1,2,2-Tetrachloroethane 0.4 1 0.04 0.4 1 0.5 0.9947
o-Xylene 1.5 1 0.1 1.0 1 2 0.9959
1,3,5-Trimethylbenzene 1.6 1 0.2 2 1 2 0.9916
1,2,4-Trimethylbenzene 1.5 1 0.2 2 1 2 0.9914
1,3-Dichlorobenzene 1.2 1 0.1 1 1 2 0.9905
1,4-Dichlorobenzene 1.3 1 0.2 1 1 2 0.9910
1,2-Dichlorobenzene 1.2 1 0.1 1 1 2 0.9899
1,2,4-Trichlorobenzene 0.8 1 0.1 0.5 1 1 0.9943
Hexachlorobutadiene 0.9 1 0.1 0.9 1 2 0.9930

 

 

values also had similar results when compared to the baseline system (Table
3.5).

The a term, or the slope in the linear regression equation, is also deﬁned
as the calibration sensitivity of the particular analyte, and can be used to
compare relative sensitivities between methods [5]. It can be observed in Table
3.4 that the HSA—SPME element was less sensitive to the light gases (up to 1,1-
Dichloroethene) than the benchmark system by 1-4 orders of magnitude. This
may be attributed to the relatively high initial desorption temperature of 57°C
detected in the microtrap (Figure 3.2), causing either a lack of initial uptake or

premature desorption. This assumption seems valid since HSA-SPME slope

78

values for the ﬁrst 11 analytes are more agreeable with microtrap values (Table
3.5), even though the theoretical analyte load on the HSA—SPME element was
13.3X higher than microtrap loading in the same 1-minute sampling time. For the
mid- and late-eluting analytes, the HSA—SPME element adsorbed relatively
similar amounts as the benchmark at given concentrations, although theoretical
loading was 20X higher in the HSA-SPME system. This is to be expected since
analyte adsorption by the HSA-SPME element is non-exhaustive and analyte
breakthrough is large at higher sampling ﬂow rates (4 L/min). Nevertheless, the
HSA-SPME element samples a more representative volume of air in the same
amount of time while retaining sensitivities observed with the benchmark, which
uses near-exhaustive cryogenic sorption at lower sampling ﬂow rates (0.2 Umin).
The histogram in Figure 3.8 illustrates this point. The chart shows individual
analyte load proﬁles (averaged triplicate runs) of a humidiﬁed, 16 ppbv (nominal
conc.) TO-14A gas mixture performed by the benchmark and experimental
systems. For each system, total sampling time was exactly one minute at
respective system ﬂow rates. Analyte loading of the light gases (compounds 1-
11) by the benchmark system was more efﬁcient due to cryogenic trapping than
with either experimental system trapping at ambient temperature. However, in
compounds 12-39, with the exception of carbon tetrachloride, the HSA-SPME
element trapped relatively similar amounts of analyte as the benchmark system,
but from ~20X the volume of air, without using cryogen. At this rate, the HSA-

SPME element could realize qualitative signal from low and sub-ppb levels

79

 

 

 

 

 

I Microtrap @ 0.3 Umin (6 ng nom.)

 

I HSA—SPME/Microtrap @ 4 L/min (80 ng nom.)

l'_'l Baseline @ 0.2 Umin (3.2 ng nom.)

 

 

—. 1% 85%

 

 

 

3.0E+06 "—"4

2.5E+06 .7—

2.0E+06 ”4

1 .5E+06

esuodseu saw

1 .0E+06 1

5.0E+05 -
o 0E+00 L
8‘ '6‘
6596596”
\0
00

 

 

80

Figure 3.8. Comparison of 16 ppbv (nominal conc., moisture added) sample loading by benchmark and experimental

air concentration systems

of analyte by concentrating from a large volume of air in the targeted time frame.
By plotting the linear regression of a particular analyte from each air
concentrator on the same graph, a visual comparison can be made regarding
sensitivity towards the analyte. Steep slopes are equivalent to higher sensitivity
to a speciﬁc concentration of analyte versus ﬂat slopes. Sensitivity comparison
of the BTEX components (Benzene, Toluene, Ethylbenzene, and o-xylene) of the
TO—14A gas mix is illustrated in Figure 3.9. Again, the HSA-SPME element
illustrated sample uptake equivalent to the benchmark study at the lower (low
and sub-ppb) concentrations, even though the HSA—SPME sampling ﬂow rate
was 20X faster than benchmark ﬂow rates. As expected, the microtrap
sensitivities were far less than both benchmark and HSA-SPME studies, since
the non-cryogenic, exhaustive approach traps 0.33X and 13.3X less over a 1-

minute period than these studies, respectively.

3.1.5 Limits of Detection

By deﬁnition, the limit of detection of an analyte is the minimum
concentration required to give an instrument response which can be
distinguished from the response given in a blank (no analyte present) at the

same retention time [3]. The minimum concentration to give such a signal is

equal to the blank signal, ya, plus three standard deviations of the blank, $3 [3]:

y—yB :35}?

81

 

Sensitivity Comparison in BTEX Components
0-32 ppbv, 1 Minute Sampling Times

 

HSA-SPME Element/Microtrap

5.00E+05

 

0 5 1o 15 20 25 30 35
Concentration (ngIL)

 

 

 

Figure 3.9. Calibration curves for the BTEX series in the TO-14A gas mixture.
The HSA—SPME element has comparable sensitivities to that of the
benchmark in identical 1-minute sampling time.

82

Based upon the published method [3] for determining the limit of detection (LOD)
for each air concentration system, ﬁve pure air blanks were sampled, and upon

LTM-GC analysis, corresponding background signals were obtained and
standard deviations (sB) determined for each component in the TO-14A gas

mixture. This value is multiplied by 3 to give a 95% conﬁdence level of detection,
as noted by Kaiser [6] and Long and Winefordner [7]. Using the respective

analyte linear regression equation (Tables 3.4 & 3.5), the value of y at x = 0.0
concentration (ya) was acquired. Rearranging the above equation and solving

for y, the concentration x at the limit of detection could be determined by solving

for x in the linear regression equation [5]. These results are listed in Table 3.6.
As can be observed, both the focusing preconcentrator microtrap and

HSA-SPME element contained detection limits from the low ppb to low ppt range.

This range is consistent with benchmark preconcentrator limits, indicating that

Table 3.6. Limits of Detection (LOD) of the benchmark and RVM
experimental air concentrators for the TO-14A components

 

83

Table 3.6. (Cont'd.)

 

 

 

 

 

Benchmark Microtrap

Target Analyte t) t
Methylene Chloride 19 35
Trichlorotriﬂuoroethane 2 1281
1,1-Dichloroethane 8 16
(Z)-1 ,2-Dichloroethane 17 13
Chloroforrn 5 15
1,2-Dichloroethane 5 21
1,1,1-Trichloroethane 3 37
Carbon Tetrachloride 17 21
Benzene 69 263
1,2-Dichloropropane 27 132
Trichloroethylene 65 16
(Z)-1,3-Dichloropropene 11 192
(E)-1,3-Dichloropropene 162 1633
1,1,1-Trichloroethane 85 47
Toluene 466 234
1,2-Dibromoethane 39 43
Tetrachloroethylene 89 30
Chlorobenzene 119 17
Ethylbenzene 134 23
m,p-Xylene 182 20
Styrene 153 1 1
1,1,2,2-Tetrachloroethane 132 37
o-Xylene 123 6
1,3,5-Trimethylbenzene 171 19
1,2,4-Trimethylbenzene 167 19
1,3-Dichlorobenzene 188 1 1
1,4-Dichlorobenzene 144 8
1,2-Dichlorobenzene 247 14
1,2,4-Trichlorobenzene 1193 68
Hexachlorobutadiene 973 8

 

HSASPME
t

64
1222

15

203 "5'
459.
337
723

21 ,
1405
631

129‘

 

 

the experimental concentrators were able to achieve relatively low background
noise for the particular analyte ions scanned. A comparison of these results to
literature data [8] indicates that the benchmark and experimental air
preconcentrators are both effective in detecting at low levels often encountered
in large-volume sampling. The low-ppt values are also a consequence of the
relatively small peak width values obtained by LTM column chromatography (See
Figure 1.3). These values contribute to taller chromatographed peaks and
subsequent increased analytical sensitivity (detector response) toward analyte.
In this investigation, calibration curves from the low ppt to low ppb range,
coupled with the extraction proﬁles obtained in Figure 3.7, indicate a linear
dynamic range of four orders of magnitude for TO-14A components #11-39 in

Table 2.4.

3.1.6 Method Comparison

To determine whether the HSA-SPME method could be compared
quantitatively to the EPA TO-14A method, sampling parameters were normalized
to 0.2 Umin over a one-minute period for both HSA-SPME and benchmark
systems. A Carboxen/PDMS HSA-SPME element with a 25 ,um phase thickness
was used to ensure desorption of the BTEX components at lower desorption
temperatures of 250-285°C.

Ten TO-14A gas samples were prepared in Tedlar bags such that nominal
sample concentrations were 8, 16, 32, 40, 60, 80, 100, 120, and 140 ppb(v)

except for m/p-xylenes, which were approximately twice these concentrations.

85

These gas samples were individually sampled by the benchmark concentration
system at 0.2 L/min to produce nominal mass loading of 1.6, 3.2, 6.4, 8, 12, 16,
20, 24, and 28 ng, respectively. The sampling was repeated by the HSA-SPME
element using the same concentrations and sampling ﬂow rate to produce
quantitatively similar mass loading. Benchmark and experimental calibration
standards were composed at concentrations of 10, 20, 50, 100, and 150 ppb(v);
the curves for all the BTEX components produced linear correlation coefﬁcients
greater than 0.99 Equally weighted regression lines were used to statistically
compare the analytical techniques.

The data derived from the benchmark air concentration system were plotted on
the x—axis (reference axis) of a regression graph, and responses from the HSA-
SPME element were plotted on the y-axis. Each point on the graph thus
represents one sample analyzed by the two methods. A regression plot for
benzene is given in Figure 3.10. Perfect agreement between the methods for all
samples would theoretically yield a slope and correlation coefﬁcient of unity (1.0)
and a y-intercept of zero [9]. The regression line parameters for each BTEX
component with 95% conﬁdence intervals [3] are listed in Table 3.7. From Table
3.7, it can be observed that there is strong correlation between the benchmark

cryotrap method and the experimental HSA—SPME method.

86

 

concentrations: 1.6-28 ng nom.
TO-14A, hunidlfled
rate: 0.2 Unin

method

1 .018x +
R2 = 0.9926

HSA-SPME Reponse (nom ng)

 

0 5 10 15 20 25 30
Benchmark Reponse (nom ng)

 

 

 

Figure 3.10. Dynamic ﬂow HSA—SPME response versus benchmark cryotrap
air concentration response for benzene in air.

Table 3.7. Regression line parameters for the BTEX components of the TO-14A
gas mix. Method comparison of HSA-SPME element versus the
benchmark cryotrap

 

 

 

 

 

 

 

 

 

 

Regression Parameters: Benchmark System on x-axis

Analyte 2

Slope y-Intercept Correlation (R )
Benzene 1.02 1 0.078 0.016 1 1.2 0.9926
Toluene 1.05 1 0.104 -0.189 1 1.6 0.9900
Ethylbenzene 1.05 1 0.076 0.154 1 1.2 0.9934
m/p-xylene 1.05 1 0.076 0.257 1 1.2 0.9935
o-xylene 1.04 1 0.089 0.261 1 1.3 0.9932

 

 

 

3.1.7 Power Consumption
Baseline Air Concentrator

Because the Entech 7100 Air Concentrator requires the use of lengthy
heated transfer lines, multiple pneumatic valving, and electronic temperature
control for modular heating and cooling, it consumes more power than the
prototype RVM Adaptive Air Sampler. The RVM uses no dedicated heated lines,
replaces pneumatic valving with two 1” x ‘/4 “ electronic valves, and consumes a
minimal amount of power during operation for micropump sampling and ballistic
heating of the HSA-SPME element and microtrap.

Power consumption was measured for the benchmark Entech system for
the length of the preconcentration (13 minutes), including transfer line heating
and sampling events. Prior to initiating the preconcentration program, the
instrument registered a standby power reading of 248 W with the transfer lines
stabilized at 150°C. During preconcentration, power consumption was sporadic,
with no deﬁned power inclines or declines with speciﬁc preconcentration events,
as shown in Figure 3.10. The benchmark system gave a maximum power
reading of 712 W during the preconcentration program.

Experimental Air Concentrator

Power consumption for the RVM Adaptive Air Sampler prototype
concentration system was recorded during preconcentration (micropump
sampling/HSA-SPME desorption and microtrap desorption), for a total time of
1.16 minutes. The standby power reading before preconcentrator initiation was

10 W, which is primarily due to supplying power to the internal controller cards.

88

Maximum power observed during HSA-SPME element heating was 62 W, and 40
W for microtrap heating.

To give a more deﬁnitive estimate of total power consumption for in situ
sampling and analysis, power readings for the RVM Adaptive Air Sampler and
low thermal mass-GC (LTM-GC) were taken as a complete, stand-alone system
to illustrate total power required for ﬁeld portability. Other components required
for a complete system are a laptop computer, an on- board detector device, and
carrier gas peripherals (electronic pressure controls, carrier gas getters, etc.), all
requiring a power source as well. Figure 3.10 shows the power consumption as
a function of time for the RVM Adaptive Air Sampler followed by a ZOC/min
temperature ramp on the LTM-GC. The standby power observed was 52 W
before preconcentration, which includes a 10 W standby amount from the RVM
Adaptive Sampler. The difference conﬁrms the LTM-GC standby reading of 42W
observed by Sloan [4]. Upon preconcentrator initiation, power rose to a
maximum amount of 101 W and 88 W for HSA-SPME element and microtrap
ballistic heating, respectively. The average watts/min for the combined RVM
prototype air concentrator/LTM-GC was 64, which represents 83% less power
required to operate the lone baseline air concentrator at 375 W/min. The LTM-
GC showed a maximum power reading of 101 W during the column temperature
ramp. Power consumption proﬁles for the baseline and RVM Adaptive
Sampler/LTM-GC systems are shown in Table 3.8. The total power consumption
for the RVM Adaptive Air Sampler and LTM-GC devices during operation is

equivalent to supplying power to a 75 W light bulb. At this rate, the experimental

89

system could theoretically perform a large number of analyses in the ﬁeld utilizing

a portable power source. At an average of 70 W with a 2.91 ampere load, the

tandem RVM devices could theoretically operate for 3 hours [10] on an 18 amp-

hour, 115 VAC portable power supply, providing up to 18 separate analyses

using experimental GC parameters.

Table 3.8. Power consumption proﬁles for the benchmark Entech 7100 versus
the coupled RVM Adaptive Air Sampler/LTM-GC systems

Min Watts

Max Watts
”Avg. Watts

Median Watts
”Percent RSD

Total Power (W)

Entech 7100
90' i .
712
346
343
- 35 -
4875 in 13 min

90

RVM / LTM-GC
101
71
72
18
759 in 11.3 min

 

13:59.2. .2 .8... m. :o_EE:mcoo .958 00.5.5 new .0888 ._< @2583. _>.>m dEm.
6.39882 00-5. 5.5.8.5618 s.>m .mEoEtoqxo 98 65.58.. .895. 8.8.8.1818 .958 .90... .36 8.5m...

1...... 88$
003—. Oouﬂr 005—. Donna 6°56 0°36 OOHNO ocuoc
QEo... a
Panda 9238.5. 2:.
Eamon.
OON
con
88 W
com
COD
25%.! 5.2 I: 2:
cozahceocooei 8352-9 Eoeucm
.0301
con

 

0042b:— £~.>> “5.1300 .33an .__< o>3nau< .2>~.
.a> .35.»:oocooek. coeucm No catnﬁamcoo .eioa No coarse—:00

 

91

3.2

10.

References

Brown, J. and R. Shirey, A Tool for Selecting an Adsorbent for Thermal
Desorption Applications. 2001 , Supelco, Inc.: Bellefonte, PA. p. 1-36.

Pawliszyn, J., Solid Phase Microextraction: Theory and Practice. 1st ed.
1997: Wiley-VCH, Inc. 247.

Miller, J.C. and J.N. Miller, Statistics for Analytical Chemistry. 1984,
Chichester, West Sussex: Ellis Horwood Limited. 202.

Sloan, K., Method Development and Implementation of a Novel
Temperature-Programmable Gas Chromatograph for Rapid Forensic
Analysis in the Field, in Columbian School of Arts and Sciences. 2001,
George Washington University: Washington, DC p.47.

Skoog, D., J. Holler, and T. Nieman, Principles of Instnrmental Analysis.
1998, Philadelphia: Harcourt Brace & Company.

Kaiser, H., Anal. Chem, 1987. 42(53A).
Long, G.L. and JD. Winefordner, Anal. Chem, 1983. 55(712A).

Bruns, M. and F. Li, Field Concentration of Toxic Organics in Air. Vol. 14.
1995, Fremont: Supelco, Inc. 10-11.

Pawliszyn, J ., et al., Headspace Solid-phase Microextraction versus Purge
and Trap for the Determination of Substituted Benzene Compounds in
Water. Journal of Chromatographic Science, 1994. 32: p. 317-322.

Altronix, Tech Tips: Application Note 102. 2004,
http/l:www.altronix.com/html/an102.htm.

 

92

CHAPTER 4 — CONCLUSIONS

Results from the analysis of the US. EPA TO-14A gas mixture indicate
that the experimental high-surface area SPME element, combined with the
focusing preconcentrator microtrap, provides performance similar to the Entech
7100 benchmark air concentration system for the BTEX components under low
and high-ﬂow sampling conditions at ambient temperature. For the majority of
volatile organic compounds tested, linear dynamic range and desorption
efﬁciencies were comparable to the benchmark system. While the benchmark
system achieved lower limits of detection for the lighter molecular weight
compounds, the lower limits found by the experimental system for the late-eluting
compounds illustrates superior trapping efﬁciency for these compounds, even at
20 times the benchmark sampling ﬂow rate. Therefore, sampling at a high ﬂow
rate while retaining high sensitivity enables a decrease in detection limits for
these compounds.

The experimental air concentration system outperformed the benchmark
system in preconcentration time and power consumption. Total sample and
preconcentration time for the Entech 7100 required thirteen minutes to perform
cryogenic trapping in three successive sorbent modules before ﬁnal desorption
and subsequent GC analysis. HSA-SPME element and microtrap
preconcentration, including initial sampling, required 89% less time than the
benchmark preconcentrator, theoretically enabling approximately 10X sample
throughput over a 1-hr period using experimental parameters. Power

consumption for the RVM Adaptive Air Sampler, coupled with the RVM LTM-GC.

93

required a maximum of 101 W during preconcentration and not more than 80 W
for an entire chromatographic operation. The experimental system consumed
seven times the power on average over the experimental system. Clearly, the
RVM experimental system can provide representative ﬁeld sampling in a
relatively short amount of time, without large power requirements, or loss in
sensitivity.

Future Work

Since HSA-SPME elements can be individually heated, the
implementation of multiple elements in an array along the same ﬂow path affords
an adaptable front-end with sampling variation. By varying the SPME adsorbent
coatings (e.g. polyacrylate and PDMS), detection of a broader range of
compounds (SVOCS and VOCs) may be achieved, rather than using one type of
adsorbent alone. In this arrangement, both preliminary and conﬁrmatory
analyses may be performed from the same air sample. Future extension of this
project include testing SPME arrays by sampling compounds with a broader
range of volatility in both controlled and open (ﬁeld) environments.

Human scent detection and tracking by police canines has been
successful in the ﬁeld of forensic investigation, but canine training methods
imposed in the United States have recently come under judicial scrutiny.
Currently, there are standard procedures for training and certiﬁcation, however
there are no scientiﬁc studies that identify the chemical content of human scent
“signatures,” or respective concentrations upon which canines are responding.

With the RVM Adaptive Air Sampler, it may be possible to sample at ﬂow rates

94

and achieve detection levels comparable to canine olfactory systems to aid in
forensic investigations in determining human scent chemical composition.

At the time of this writing, the RVM Adaptive Air Sampler and LTM-GC
systems have been installed in a prototypical, portable GC. complete with on-
board micro pulsed-discharge helium ionization detection (uPDHID). The next
phase of this ongoing project is to develop methods that will optimize this
detector, along with sampling and analytical methods for this novel, stand-alone

system.

95

APPENDICES

96

APPENDIX A
GLOSSARY OF TERMS AND ABBREVIATIONS
analyte - a single component of mixture separated by gas chromatography
ballistic - rapid heating action
cmlsec - centimeters per second

carrier gas — a gas, usually diatomic, which carries sample analytes through the
gas chromatograph

capillary column — a long marrow tube composed of glass silica that contains a
stationary phase through which the mobile phase (carrier gas) is forced under
pressure

chromatogram - a graphical record of chromatography containing a series of
peaks which represent the detector response as a function of time

convolution - analyte peaks which are overlapped on a chromatograph

cryogen - liquid nitrogen; used to rapidly cool sorbent traps to effectively trap
lighter volatile gases

deconvolution - the process of distinguishing individual peaks from multiple
overlapped peaks

efﬁciency - relative ease of performance in terms of analyte separation, power
consumption, and/or time

gas chromatograph - an instrument that uses a capillary column to separate
mixtures on the basis of chemical interactions with a stationary phase and a
mobile phase

60 - gas chromatograph or gas chromatography

HSA-SPME - high surface area-solid phase microextraction

K - partition coefﬁcient for an analyte

L - liter

Umin — liters per minute

LOD - limit of detection [ppt(v)]

97

LTM-GO - low thermal mass gas chromatography

m - meter
mg — milligram (10'3 gram)

um - micrometer (10'6 m)

mL - milliliter (10'3 L)

MS — mass spectrometry or mass spectrometer
MSD - mass selective detector

min - minute

Ohm — a unit of electrical resistance (0)

ppbv — parts per billion by volume

ppmv — parts per million by volume

pptv - parts per trillion by volume

precision - the agreement of results between repeated experiments

psi - pounds per square inch

quadrupole mass spectrometer - a scanning mass spectrometer that ﬁlters out
ions based on mass as they are accelerated through the quadrupole area

RH - relative humidity

resolution - a quantitative measure of a column’s ability to separate two

analytes

retention time - the time required from the point of injection for the analyte to

exit the column
RSD - relative standard deviation
sec - second

SPME - solid phase microextraction

98

SVOC — semi-volatile organic compound

temperature program — a series of time-based and controlled temperature
changes applied to a GC column during separation

thermal mass — a measure of a material's ability to store energy in the form of
heat

toroid - a closed-loop circular coil rotated about a second larger circular axis
V - volts
VOC - volatile organic compound

W - watts

99

APPENDIX 8

Original software provided by RVM Scientiﬁc, Inc., to control the Adaptive
Air Sampler system

l/Door Preconcentrator.CPP 6/09/03

#include <stdio.h>
#include <math.h>
#include <dos.h>
#include <string.h>
#include <stdlib.h>
#include <time.h>
#include <conio.h>
#include <graph.h>
#include "pcl4c.h"
#include <ctype.h>
#undef tolower

#deﬁne DAC_enable 8
#deﬁne ON 1
#deﬁne OFF 0

unsigned char digital_out, PORT, gain, bipolar, bits_24, average, ﬁlter,
polled, channel, char_count, dos _yscale, analog_out_a, analog_out_b,
video = 1, term = 1, b[4], mode_reg_hi, mode_reg_mid, mode_reg_lo,
c_mode, ch, cl, gain_prev;
int maxcnt=16383, digital_in (void), Seg, rate, num_ave, time_base_count,
counts_per_volt, time_base,
wait_time, count_port, counter, wait_time_prev,
ambient,iso_1_temp=60,iso_1_time=0,iso_2_temp=60,
iso_2_time=90,sampling_time,trap_init,trap_desorb,trap_bake,trap_heatin
g_time,
trap_update_freq,injection,bake_time,
t_last=0, ramp_time,inj_delay=0, i=0, j, kbhit(void), v1, v2, v3, v4, inj_time,
samp_time;

ﬂoat DAC_init,baud,f_time,data_rate, ramp_rate=60.0, time _per_update=0.1,
t=0, T_target,DAC_A, DAC_A_lnit,

DAC_A_Fin,datum,chromatogram_time, pressure, cvoltage, feedback;
static char LRxBuffer[128+16], LTxBuffer[128+16];
char far *LPtr,fname[13];
clock_t start;
char s1 [15];

100

void lnit_variables(void);
void Ready_com_port(void);
long Time_in_sec(void);
void Sign_on(void);
void delay_sec(int sec);
void delay(int ms);
void send _packet(unsigned char b1, unsigned char b2);
void wait_for_char(int nchar, int wait_in_sec);
void Set_DAC_voltage(char DAC, ﬂoat voltage);
ﬂoat RTD(ﬂoat T);
void bit_8 (unsigned char action);
void bit_7 (unsigned char action);
void bit_6 (unsigned char action);
void bit_5 (unsigned char action);
void bit_4 (unsigned char action);
void bit_3 (unsigned char action);
void bit_2 (unsigned char action);
void bit_1 (unsigned char action);
void abort(void);
void DAC_Init(void);
void reposition(void);
void digital_in (unsigned char action);
void lnit_AZD_mode(void);
void Calc_AZD_regs(void);
void DAC_Init(void);
void Set_AZD_mode(void);
ll
// Main Program
//
void main(void)
{
char s[20], c='y':
int i, v1, v2, v3, v4;
Init_variables();
v1=0,v2=0,v3=0,v4=0;
lllnitialize R8232 port for communication with Lawson multi l/O board
Ready_com_portO;
IlEstablish RS232 communication with Lawson board
Sign_on();
lllnitialize analog outputs
DAC_Init();
DAC_init = RTD(20);
bit_1(OFF);
bit_2(OFF);
bit_3(OFF);
bit_4(OFF);

101

bit_5(OF F);
bit_6(OFF);
bit_7(OFF);
bit_8(OFF);
//
reposition();
while (c != 'x')
{
_c|earscreen(_GCLEARSCREEN);
printf("Press (I) to move motor ln\n");
printf("Press (O) to move motor Out\n");
printf("Press (P) to turn pump on and ofﬂn");
printf("Press (1) to open or close Valve 1\n");
printf(”Press (2) to open or close Valve 2\n");
printf(”Press (3) to open or close Valve 3\n");
printf("Press (D) to desorb\n");
printf("Press (S) to check status of optical switch\n");
printf("Adjust motor in IN position and leave in OUT position before exiting
(X)\n");
do c = tolower(getch());
while (c!='1' && c!='2' && c!='3' && c!='p' && c!='d' &&
c!='i' && c!='o' && c!=‘s' && c!='x');
switch (c)

l .
case 'l':
bit_6(ON);
delay_sec(1 );
bit_6(OFF);
break;
case '0':
bit_5(ON);
delay_sec(1 );
bit_5(OFF);
break;
case '1':
v1 = (v1 - 1);
v1 = abs(v1);
if (v1) bit_4(ON); else bit_4(OFF);
break;
case '2':
v2 = (v2 - 1);
v2 = abs(v2);
if (v2) bit_2(ON); else bit_2(OFF);
break;
case '3':
v3 = (v3 - 1);

102

v3 = abs(v3);

if (v3) bit_3(ON); else bit_3(OFF);

break;

case 'p':

v4 = (v4 - 1);
v4 = abs(v4);

if (v4) bit_7(ON); else bit_7(OFF);

break;
case 'd':
Set_DAC_voltage('A'. 1);

else printf("Valve is Open”);

}
c='y’;

while (c != 'x')

{

// turn on valves 2 and 3

bit_2(ON);

bit_3(ON);

delay_sec(3);

Set_DAC_voltage('A',0);

bit_2(OFF);

bit_3(OFF);

break;

case 's':

if (digital_in()) printf("Valve is Closed");

delay_sec(1 );
break;

_c|earscreen(_GCLEARSCREEN);

printf("lnjection Time: %d sec\n", inj_time);
printf("Sample Time: %d sec\n",samp_time);
printf("Change (T)ime to lnject\n");

printf(" (S)ample time\n");

printf(" Accept values, e(X)it menu, and start run\n");

do 0 = tolower(getch());

while (c!= 'x' && c!='t' && c!='s');

switch (c)

{

case 't':

printf("lnjection Time = ");
scanf("%d",&inj_time);

break;
case '8':

printf("Sample Time = "X

103

scanf("%d",&samp_time);
break;

}

// Starting new method for preconcentrator sampling
II
[I here is where motor is repositioned
printf(”Repositioning motor\n");
reposition();
//
// turn on valve 1 (sample valve)
// simultaneously turn on pump - can use same digital channel in future
_c|earscreen(_GCLEARSCREEN);
printf(”\nNew Method Starting\n");
printf("Tuming on valve 1 and pump to sample\n");
bit_4(ON); // valve 1
bit_7(ON); // pump
delay_sec(samp_time);
[I turn off valve 1 and pump
bit_4(OFF);/l valve 1
bit_7(OFF);/l pump
//
// Injection phase
//
// close motor, heat transfer line, heat preconcentrator for 2 seconds
bit_6(ON);
delay_sec(3);
bit_6(OF F);
Set_DAC_voltage('A', 1);
delay_sec(3);
ll turn on valves 2 and 3
bit_3(ON);
bit_2(ON);
printf(”lnjecting\n");

// insert appropriate injection delay here
delay_sec(inj_time);

// turn off valves 2 and 3 to sample

// here is where heating of transfer line and preconcentrator is turned off
Set_DAC_voltage('A', O);

bit_3(OF F);

bit_2(OFF);

ll Start LTM A68

bit_8(ON);

delay(100);

bit_8(OFF);

104

}

ll Reposition motor

ll

printf(”Repositioning motor\n”);

reposition();

//

delay_sec(1 );

SioDone(PORT); //Reset R8232 port

// end of main program

//

void lnit_variables()

{

l/

}

PORT = 0; llcom1 = 0, com2 = 1, etc.

gain = 1; llAlD gain in powers of 2

bipolar = 1; I/polarity mode ﬂag

bits_24 = 1; //AlD word length ﬂag

average = 0; //AlD averaging of points as a power of 2

ﬁlter = 2; ”ND cutoff frequency as 4*10E(ﬂlter)

polled = 1; //polled mode ﬂag

rate = 10; ”ND data rate in Hz

num_ave = 1; llparameter for wait_time calc

baud = 0; llparameter for wait_time calc

channel = 0; llanalog input channel selection

wait_time = num_ave/rate + baud/rate + 2; llwait time variable
char_count = 3; //number of characters to transmit in Lawson l/O
dos _yscale = 1; l/AlD calibration ﬂag

f_time = 0.03; l/timing value

time_base_count = 10000; llDAC counter

counts_per_volt = (int)ceil(time_base_countl5); llDAC initialization
time_base = 0; l/DAC initialization

analog_out_a = 1; llDAC port 'A' value

analog_out_b = 2; //DAC port '8' value

digital_out = 0; llunsigned char for digital output

inj_time = 5; // injection time

samp_time = 4; // sample time

void Ready_com_port(void)

{

char string[80];
char buffer[1];

l/setup 128 byte receive buffer

LPtr = (char far *)LRxBuffer;

Seg = FP_SEG(LPtr) + ((FP_OFF(LPtr)+15)>>4);
SioRxBuf(PORT,Seg,Size128);

105

Ilsetup 128 byte transmit buffer
LPtr = (char far *)LTxBuffer;
Seg = FP_SEG(LPtr) + ((FP_OFF(LPtr)+15)>>4);
SioTxBuf(PORT,Seg,Size128);
Ilset port parameters
SioParms(PORT,NoParity,OneStopBit,WordLength8);
//reset port for initial operation at 300 baud
SioReset(PORT,Baud300);
SioDTR(PORT,'S');
SioRTS(PORT,'C');
if (video) printf("\nRS232 initialized using COM%d @ 300
Baud\n",1+PORT);

if (term)

{
strcpy(string,"R8232 initialized using COM");
_itoa(1+PORT,buffer,10);
strcat(string,buffer);
strcat(string,” @ 300 Baud");

}
}
long Time_in_sec(void)
{
time_t ltime;
time(&ltime);
return ltime;
}
void Sign_on(void)
{
long time;
int i,j;

time = Time_in_sec();

INVake up Lawson multi l/O board

while ((i != 3) && (T ime_in_sec()—time < 5))
{

i = SioPutc(PORT,'\O');
delay(200); ”increased from 200 for laptop
i = SioGetc(PORT,5);

}

if (video) printf(”Lawson board output = %d\n",i);
delay(200);

//Send sign-on token

SioPutc(PORT,0x88);

106

delay(200);
SioPutc(PORT,'\O');
delay(200);
SioGetc(PORT,0);
l/Lawson board should be ready to communicate at 9600 baud
SioBaud(PORT,Baud9600);
if (video)
{
printf("Echo test:\n");
delay(200);
for (i=1 ; i<11; i++)
{
SioPutc(PORT,i);
delay(200);
j = SioGetc(PORT,0);
printf(”/ad %d\n",i,j);
}

}

delay(200);
SioPutc(PORT,'\0');
lllnitialize ND converter
lnit_A2D_mode();
send_packet(average,ﬁlter);
if (polled) send_packet(0,1);
else send_packet(0,0);
delay(200);

”Check to see if Lawson board correctly returns initialization values
for (i=0; i<3; i++)

b[i] = SioGetc(PORT,1);
delay(200);

}
if ((b[O] == mode_reg_hi) && (b[1] == mode_reg_mid) 8.8.
(b[2] == mode_reg_lo))
if (video) printf(”LOGGED ON AT 9600 BAUD IN POLLED
MODE\n");
else if (video)
{
printf("PROBLEM LOGGING ON");
exit(0);

}

void send _packet(unsigned char b1 ,unsigned char b2)
{

unsigned char check_sum;

107

}

SioPutc(PORT,b1);
SioPutc(PORT,bZ);
check_sum = (b1+b2) & Oxff;
SioPutc(PORT,0heck_sum);

void DAC_Init(void)

{

send_packet(5,64); llset resolution to 14 bit

Set_DAC_voltage('A',0);
Set_DAC_voltage('B'.O):

}
void wait_for_char(int nchar, int wait_in_sec)
{
long start;
start = Time_in_sec();
while ((SioRxQue(PORT) != nchar) | (Time_in_sec()start > wait_in_sec));
if ((SioRxQue(PORT) != nchar) && video)
printf("lO Error: not enough received characters”);
}
void Set_DAC_voltage(char DAC, ﬂoat voltage)
{
ﬂoat analog_duty_cycle;
unsigned char anahigh,analow;
if (voltage < 0) voltage = 0;
if (voltage > 5) voltage = 5;
analog_duty_cycle = (ﬂoat)maxcnt*voltage/5;
anahigh = (unsigned char)(analog_duty_cycle/256);
analow = (unsigned char)(analog_duty_cycle-anahigh*256);
send_packet(8,analow); llset analog output low byte
switch (DAC) llset high byte
{
case 'A': send_packet(9,anahigh & 0x7f); break;
case 'B': send_packet(9,anahigh | 0x80);
}
}
ﬂoat RTD(ﬂoat T)
{
ﬂoat x,Vref,Rf,R2,alpha,delta;
Vref = 2.501; /N

108

Rf = 13687; llgain resistor, ohms

R2 = 9946; /lohms
alpha = 0.00385; /lPt temperature coefﬁcient
delta = 1.5065; llPt temperature coefﬁcient

x = Vref‘Rf/R2*alpha*(T—delta*(T/100-1)*(T/100)); ”Calendar-Van Dusen
// Divide x by 2 to run TC3 board

x=ﬂz

return x;

}

void bit_8 (unsigned char action)
switch (action)

{
case OFF: digital_out &= 0x7f ; break;

case ON: digital_out |= 0x80 ;
}
send_packet(2,digital_out);

void bit_7 (unsigned char action)

{ switch (action)
{ case OFF: digital_out &= Oxbf; break;
case ON: digital_out |= 0x40;
send_packet(adigital_out);
iroid bit_6 (unsigned char action)

switch (action)

{
case OFF: digital_out &= 0xdf; break;
case ON: digital_out |= 0x20;

}
send_packet(2,digital_out);

void bit_5 (unsigned char action)

{
switch (action)
{
case OFF: digital_out &= Oxef; break;
case ON: digital_out |= 0x10 ;
}
send_packet(2,digital_out);
}

109

void bit_4 (unsigned char action)

{

switch (action)

{
case OFF: digital_out &= 0xf7 ; break;
case ON: digital_out |= 0x08 ;

}
send_packet(2,digital_out);

void bit_3 (unsigned char action)

{

switch (action)

{
case OFF: digital_out &= Oxfd ; break;

case ON: digital_out |= 0x02 ;
}
send_packet(2,digital_out);

void bit_2 (unsigned char action)

{

switch (action)

{
case OFF: digital_out &= Oxfb ; break;

case ON: digital_out |= 0x04 ;
send_packet(2,digital_out);

void bit_1 (unsigned char action)

{

switch (action)

{
case OFF: digital_out &= Oxfe ; break;
case ON: digital_out |= 0x01 ;

}
send_packet(2,digital_out);

int digital_in(void)
//
// Optek optical switch on digital input RS 1
l/
{
send_packet(0x80,0x4c);
wait_for_char(2,1 );
for (j=0; i<2; I“)
{

b[i] = SioGetc(PORT,1);

110

delay(100);
}
if (b[0] != 0x80)
{

printf("digital input command token not retumed\n");

exit(0);
}
return (b[1] & 0x01);
}
//
void reposition (void)
{
bit_5(ON);
while(digital_in()==1 );
bit_5(OFF);
bit_6(ON);
while(digital_in()==0);
bit_6(OFF);
void delay_sec(int sec)
{
int i;
for (i=0; i<sec; i++) delay (1000);
}

void delay(int ms)

int i,j,k;
for (i=0; i<ms; i++)
for (j=0; j<10; j++)
for (k=0; k<5; k++);

}
void lnit_A2D_mode(void)

Calc_A2D_regs();
send_packet(mode_reg_himode_reg_mid);
send_packet(mode_reg_lo,mode_reg_m id);

}
void lnit_A2D(unsigned char channel)

{
char string[80]. 1':

int i, channel_prev;
sprintf(string,"# Initializing A/D channel %d",channel);

send_packet(0x01,channel*16 + DAC_enable);
if (dos_yscale)

111

gain_prev = gain;
channel_prev = channel;

if (gain != 1)
{
gain = 1;
Set_A2D_mode();
}
channel = 7;

wait_time_prev = wait_time;

wait_time = wait_time + f_time + 2;

”Do offset calc

send_packet(0x82,16*channel + DAC_enable);
wait_for_char(char_cou nt+1 ,wait_time);

for (i=1; i < char_count+2; i++) j = SioGetc(PORT,1);
wait_time = wait_time_prev;

channel = 6;

wait_time_prev = wait_time;

wait_time = wait_time + f_time + 2;

”Do full scale cal

send_packet(0x83,16*channel + DAC_enable);
wait_for_char(char_cou nt+1 ,wait_time);

for (i=1; i < char_count+2; i++)j = SioGetc(PORT,1);
wait_time = wait_time_prev;

if (gain != gain _prev)

{

gain = gain_prev;

Set_A20_mode();

channel = 7;

wait_time_prev = wait_time;

wait_time = wait_time + f_time + 2;

if (video) printf("Repeating offset cal”);

llDo offset calc

send_packet(0x82,16*channel + DAC_enable);
wait_for_char(char_cou nt+1 ,wait_time);

for (i=0; i < char_count+1; i++)] = SioGetc(PORT,1);
wait_time = wait_time_prev;

channel = channel_prev;
send_packet(0x01,channel*16 + DAC_enable);
delay(ﬂoor(f_time*6500)+1 800);

dos _yscale = 0;

}
void Set_AZD_mode(void)
{

112

int j;

send_packet(0x84,0);

wait_for_char(1 ,wait_time);

Calc_A2D_regs();
send_packet(mode_reg_himode_reg_mid);
send_packet(mode_reg_lo,mode_reg_mid);
delay(6500);

//Check for proper token return after sending mode
j = SioGetc(PORT,0);

if ((i != 0x84) && video) printf("Wrong start token");
for (j=0; j<3; j++)

{

b[j] = SioGetc(PORT,1);
delay(3300);

}
if (((b[0] != mode_reg_hi) | (b[1] != mode_reg_mid) |
(b[2] l= mode_reg_lo)) 88 video) printf("PROBLEM SENDING A

MODE");

}
void Calc_A2D_regs(void)

{

int mode_code, 9;
ﬂoat AD_clock = 19531.25;
int temp, standby = 0;

rate = 10;

l/calc mode register data

mode_code = (int)ﬂoor(AD_clocklrate);

mode_reg_lo = mode_code & Oxff;

mode_reg_mid = (mode_code & 0x700)/256;

rate = ﬂoor(AD_clock/(((mode_reg_mid & 0x7)*256)+mode_reg_lo));
if (video) printf(”Calc rate to send: %d\n",rate);

9 = 0:

temp = gain;

while (temp != 1)

if (temp > 1) temp /= 2;

g++;
}
llset gain and power down bits
mode_reg_hi = g*4 | standby;
if (lbipolar) mode_reg_mid += 16; l/unipolar
if (bits_24) mode_reg_mid += 128;

113

BIBLIOGRAPHY

114

BIBLIOGRAPHY

Altronix, Tech Tips: Application Note 102. 2004,
httpl/zwww.altronix.com/html/an102.htm

Bassford, M., P. Simmonds, and G. Nickless, An Automated System for Near-
Real-Time Monitoring of Trace Atmospheric Halocarbons. Anal. Chem., 1998.
70(5): p. 958-965.

Bennett, M., TICs, TIMs, and Terrorists, in Today's Chemist at Work. 2003. p. 21-
25.

Bicchi, C., et al., Headspace Sorptive Extraction (HSSE), Stir Bar Sorptive
Extraction (SBSE), and Solid Phase Microextraction (SPME) Applied to the
Analysis of Roasted Arabica Coffee and Coffee Brew. Journal of Agricultural and
Food Chemistry, 2002. 50(3): p. 449-459.

Brown, J. and R. Shirey, A Tool for Selecting an Adsorbent for Thermal
Desorption Applications. 2001, Supelco, lnc.: Bellefonte, PA. p. 1-36.

Bruns, M. and F. Li, Field Concentration of Toxic Organics in Air. Vol. 14. 1995,
Fremont: Supelco, Inc. 10-11.

Ciucanu, I., et al., Helical Sorbent Microtrap for Continuous Sampling by a
Membrane and Trap Interface for On-line Gas Chromatographic Monitoring of
Volatile Organic Compounds. Anal. Chem., 2003. 75(4): p. 736-741.

Eckenrode, B., K. Sloan, and R. Mustacich, Development and Evaluation of a
Low Thermal Mass Chromatograph for Rapid Forensic GC-MS Analysis. Field
Analytical Chemistry and Technology. 2001. 5(6): p. 288-301.

Entech, Entech Instruments 7100 Operators Manual. 1997: Simi Valley, CA. p.
14.

Furton, K., et al., Laboratory and ﬁeld experiments used to identify Canis lupus
var. familiaris active odor signature chemicals from drugs, explosives, and
humans. Analytical and Bioanalytical Chemistry, 2003. 376(8): p. 1212-1224.

Hauser, B. and P. Popp, Membrane Extraction Combined with
Thennodesorption/Gas Chromatography and Mass Selective Detection for the
Analysis of Volatile Organic Compounds in Water. Journal of High Resolution
Chromatography, 1999. 22(4): p. 205-212.

Kaiser, H., Anal. Chem., 1987. 42(53A).

115

Kin-Tek Laboratories, l., Model 491M-B Operation Manual. 2001, Kin-Tek Labs:
La Marque, TX. p. 1.

Le Pape, M., et al., Optimization of Dynamic Headspace Extraction of Edible Red
Algae Palmaria palmata and Identiﬁcation of the Volatile Components. Journal of
Agricultural and Food Chemistry, 2004. 52(3): p. 550-556.

Lizzani-Cuvelier, L., et al., Comparison of Static Headspace, Headspace Solid
Phase Microextraction, Headspace Sorptive Extraction, and Direct Thermal
Desorption Techniques on Chemical Composition of French Olive Oils. Journal of
Agricultural and Food Chemistry, 2003. 51 (26): p. 7709-7716.

Long, G.L. and JD. Winefordner, Anal. Chem, 1983. 55:(712A).

Miller, J.C. and J.N. Miller, Statistics forAnalytical Chemistry. 1984, Chichester,
West Sussex: Ellis Horwood Limited. 202.

Mitra, S. and C. Feng, Breakthrough and Desorption Characteristics of a
Microtrap. J. Microcolumn Separations, 2000. 12(4): p. 267—275.

Mitra, S., et al., A Microtrap Interface for Continuous Monitoring of VOCs in Air
Emissions. Journal of Mass Spectrometry, 1999. 34(478).

Mitra, S., N. Zhu, and Z. Li, Application of On-line Membrane Extraction
Microtrap Gas Chromatography (OLMEM-GC) for Continuous Monitoring of VOC
Emission. J. Microcolumn Separations, 1998. 10(5): p. 393-399.

Mustacich, R. and A. Neushul, Adaptive Sampling Technology Final Report.
2003, RVM Scientiﬁc, Inc.: Santa Barbara.

Pawliszyn, J., Sample Preparation: Quo Vadis? Anal. Chem., 2003. 75(11): p.
2543-2558.

Pawliszyn, J., Solid Phase Microextraction: Theory and Practice. 1st ed. 1997:
Wiley-VCH, Inc. 247.

Pawliszyn, J., et al., Sampling and Monitoring of Biogenic Emissions by
Eucalyptus Leaves using Membrane Extraction with Sorbent Interface (MESI).
Journal of Agricultural and Food Chemistry, 2002. 50(22): p. 6281-6286.

Pawliszyn, J. and C. Arthur, Solid Phase Microextraction with Thermal
Desorption Using Fused Silica Optical Fibers. Anal. Chem., 1990. 62(19): p.
2145-2148.

Pawliszyn, J. and M. Chai, J. Environmental Science and Technology, 1995. 29:
p. 693-701.

116

Pawliszyn, J. and S. Liu, Sample Introduction for Capillary Gas Chromatography
with Laser Desorption and Optical Fibers. Anal. Chem., 1987. 59(10): p. 1475-
1478.

Pawliszyn, J. and P. Martos, Calibration of Solid Phase Microextraction for Air
Analyses Based on Physical Chemical Properties of the Coating. Anal. Chem.,
1997. 69(2): p. 206-215.

Poole, CF. and SA. Schuette, Isolation and Concentration Techniques for
Capillary Column Chromatographic Analysis. Journal of High Resolution
Chromatography, 1983. 6: p. 526-549.

Seitz, L., M.S. Ram, and R. Rengarajan, Volatiles Obtained from Whole and
Ground Grain Samples by Supercritical Carbon Dioxide and Direct Helium Purge
Methods: Observations on 2, 3-Butanediols and Halogenated Anisoles. Journal of
Agricultural and Food Chemistry, 1999. 47(3): p. 1051-1061

Settles, G. and D. Kester, Aerodynamic Sampling for Landmine Trace Detection.
SPIE Aerosense, 2001. 4394(paper 108).

Shirey, R., SPME/GC Analyses of Sulfur Gas and VOCs, Using a New
Carboxen/PDMS Fiber. Vol. 16. 1997, Bellefonte: Supelco, Inc. 7-8.

Skoog, D., J. Holler, and T. Nieman, Principles of Instrumental Analysis. 1998,
Philadelphia: Harcourt Brace & Company.

Sloan, K., Method Development and Implementation of a Novel Temperature-
Programmable Gas Chromatograph for Rapid Forensic Analysis in the Field, in
Columbian School of Arts and Sciences. 2001, George Washington University:
Washington, DC. p. 47.

Supelco, lnc., Chromatography Products for Analysis and Puriﬁcation.
2003/2004: Sigma-Aldrich Co.

Tudor, L., V. Desauziers, and J. Fanlo, Potential of Solid-Phase Microextraction
Fibers for the Analysis of Volatile Organic Compounds in Air. Journal of
Chromatographic Science, 2001. 39: p. 521-529.

Vass A, et al., Decompositional Odor Analysis Database. Unpublished, 2004.

Wong, J.M., et al., Determination of Volatile and Semivolatile Mutagens in Air
using Solid Adsorbents and Supercritical Fluid Extraction. Anal. Chem., 1991 .
63(15): p. 1644-1650.

Wright, 8., T. Zemanian, and S. Anthony. Prototype Performance Evaluation for

the Federal Bureau of In vestigation Field-portable Supercritical Fluid Extractor. in
223rd ACS National Meeting. 2002. Orlando, FL: American Chemical Society.

117

   

lljlllljlllllllllllllllllllllllllll