Citation
Identification and quantification of process water contaminants from paper mills using old corrugated containers

Material Information

Title:
Identification and quantification of process water contaminants from paper mills using old corrugated containers
Creator:
Kortmeyer, Jordan Cassidy
Publication Date:
Language:
English
Physical Description:
69 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Corrugated paperboard -- United States ( lcsh )
Containers -- Recycling -- United States ( lcsh )
Pressure-sensitive adhesives ( lcsh )
Factory and trade waste -- Pollution ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 68-69).
Thesis:
Department of Chemistry
Statement of Responsibility:
by Jordan Cassidy Kortmeyer.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
54697136 ( OCLC )
ocm54697136
Classification:
LD1190.L46 2003m K67 ( lcc )

Downloads

This item has the following downloads:


Full Text
IDENTIFICATION AND QUANTIFICATION OF PROCESS WATER
CONTAMINANTS FROM PAPER MILLS USING OLD CORRUGATED
CONTAINERS
by
Jordan Cassidy Kortmeyer
B.S., California Polytechnic State University, 1995
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Chemistry
2003


This thesis for the Master of Science
degree by
Jordan Cassidy Kortmeyer
has been approved
by
Doijglas Dyckes
Jorge Yordan

Date


Kortmeyer, Jordan Cassidy (M.S., Chemistry)
Identification and Quantification of Process Water Contaminants from Paper Mills
Using Old Corrugated Containers
Thesis directed by Professor Larry G. Anderson
1 ABSTRACT
The use of recycled corrugated containers (OCC) has been steadily growing in the
United States over the last two decades. More of these products are being recycled,
decreasing the quantities disposed in landfills. OCC is used in the production of
new paperboard products such as new corrugated containers and the outer sheets of
gypsum board.
Using OCC for the production of new paper products can introduce contaminants
such as pressure sensitive adhesives (PSAs) and hot melts in the manufacturing
process. OCC contaminants can affect appearance and strength of new product.
Conventional water treatment techniques can be used to remove contaminants down
to about 1 micron in diameter. Contaminants that are at or below 1 micron can pass
through screens or hydrocyclones (cleaners) and remain in the manufacturing
process water. Eventually, the micro-contaminants agglomerate to cause
manufacturing problems.
This new method was developed because currently there is no method for
identifying and quantifying PSAs and hot melts in process water. The new method
uses solvent extraction to collect the PSAs and hot melts from the process water.
Standard addition method was used to increase the overall accuracy of the test. The
extracted samples were analyzed using GC/MS. Results showed the new method to
be reproducible enough to warrant its use.
m


This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Signed
IV


ACKNOWLEDGEMENT
My thanks to my advisor, Larry Anderson, for his guidance and help in completing
this project. I would also like to thank my other advisors: Douglas Dyckes, Susan
Schelble, Jorge Yordan and co-workers at Luzenac America for their support and
understanding. In addition, I would like to thank the Colorado Commission on
Higher Education (CCHE) for funding this project.


TABLE OF CONTENTS
Figures.....................................................................viii
Tables........................................................................x
Chapter
1. Introduction.........................................................1
1.1 Purpose of the Study.................................................1
1.2 History of Pressure Sensitive Adhesives and Hot Melts................2
1.2.1 Pressure Sensitive Adhesives (PSAs).................................2
1.2.2 Hot Melts............................................................3
1.3 Literature Search....................................................4
2. Experimental Methods................................................12
2.1 Standards and Samples...............................................12
2.1.1 Preparation of Standards............................................12
2.1.2 Preparation of Samples..............................................13
2.2 Instrumentation.....................................................13
2.2.1 Gas Chromatography (GC).............................................13
2.2.2 Mass Spectrometry (MS)..............................................15
2.3 Extraction Techniques for PSAs and Hot Melts.......................22
2.3.1 Solvent Extraction..................................................22
2.3.2 Solid Phase Extraction (SPE)........................................24
2.4 Standard Addition Method............................................24
2.5 Identification T echniques..........................................25
vi


2.5.1 Analysis of Standards...............................................25
2.5.2 Analysis of Samples.................................................26
2.6 SPE of Adhesive Standards...........................................27
2.7 SPE of Mill Process Water...........................................27
2.8 Soxhlet Solvent Extraction of Adhesive Standards....................28
2.9 Solvent Extraction of Mill Process Water............................28
2.10 Precision...........................................................28
2.11 Fixed Detection Limit...............................................29
3. Results and Discussion..............................................30
3.1 Process Water Characterization......................................30
3.2 PSA Standard Characterization.......................................30
3.3 Hot Melt Standard Characterization..................................36
3.4 Precision...........................................................41
3.5 Mill A Characterization.............................................49
3.6 Mill B Characterization.............................................56
4. Conclusions.........................................................62
5. Future Work.........................................................64
Appendix
A. General Concentration Calculation...................................65
References...................................................................68
vii


FIGURES
Figure
1.2.1.1 Vinyl Acrylate Monomer...........................................3
1.2.2.1 Ethylene Vinyl Acetate Monomer...................................4
1.3.1 Pulp Sample Extraction Scheme....................................6
1.3.2 Water Sample Extraction Scheme...................................7
1.3.3 Mill Deposit General Extraction Scheme..........................10
2.2.1.1 Gas Chromatograph Parameters....................................14
2.2.2.1 Transfer Line and Ion Trap Assembly.............................15
2.2.2.2 Ion Trap Assembly...............................................16
2.2.2.3 Mass Spectrometer Parameters....................................16
2.2.2.4 Electron Impact Source..........................................18
2.2.2.5 Ion Trap........................................................19
2.2.2.6 Electron Multiplier.............................................20
2.2.2.7 Electron Multiplier Continuous Dynode Setup.....................21
3.2.1 PSA Standard Chromatograph Run 1................................31
3.2.2 PSA Mass Spectrum Peak 2589 ....................................32
3.2.3 PSA Mass Spectrum Peak 4305 ....................................32
3.2.4 PSA Mass Spectrum Peak 4456 ....................................33
3.2.5 PSA Mass Spectrum Peak 5011 ....................................33
3.3.1 Hot Melt Standard Chromatograph Run 1...........................37
3.3.2 Hot Melt Spectrum Peak 3092 ....................................37
3.3.3 Hot Melt Spectrum Peak 3359 ....................................38
3.3.4 Hot Melt Spectrum Peak 3827 ....................................38
viii


3.3.5 Hot Melt Spectrum Peak 4254 .................................39
3.4.1 Baseline Peak Integration of Peak 4456 for PSA Standard......42
3.4.2 PSA Standard Chromatograph Run 2.............................45
3.4.3 PSA Standard Chromatograph Run 3.............................45
3.4.4 Hot Melt Standard Chromatograph Run 2........................46
3.4.5 Hot Melt Standard Chromatograph Run 3........................47
3.5.1 Mill A Sample Baseline Chromatograph.........................50
3.5.2 Mass Spectrum for Peak 3 827 From Mill A Baseline Chromatograph. ..51
3.5.3 Mass Spectrum for Peak 4456 From Mill A Baseline Chromatograph...51
3.5.4 Mill A 50/50 Sample and PSA Standard Chromatograph...........53
3.5.5 Mill A 50/50 Sample and Hot Melt Standard Chromatograph......55
3.6.1 Mill B Sample Baseline Chromatograph.........................57
3.6.2 Mass Spectrum for Peak 3827 From Mill B Baseline Chromatograph...57
3.6.3 Mass Spectrum for Peak 4456 From Mill B Baseline Chromatograph...58
3.6.4 Mill B 50/50 Sample and PSA Standard Chromatograph...........59
3.6.5 Mill B 50/50 Sample and Hot Melt Standard Chromatograph......60
IX


.23
23
30
.34
39
43
43
44
44
47
48
48
49
.52
.54
,55
.58
.60
61
TABLES
Solubility of Various Polymers used in Adhesive Manufacturing
Solvency Testing..........................................
Initial Characterization of Process Water.................
PSA Ion Fragments......;. *...............................
Hot Melt Ion Fragments....................................
PSA Peak 2589.............................................
PSA Peak 4307.............................................
PSA Peak 4456.............................................
PSA Peak 5011.............................................
Hot Melt Peak 3092........................................
Hot Melt Peak 3359........................................
Hot Melt Peak 3827........................................
Hot Melt Peak 4254........................................
Mill A Peak Areas for Baseline Chromatograph..............
Peak Areas for PSA Standard and 50/50 Sample..............
Standard Solution for Mill A
Peak Areas for Hot Melt Standard and 50/50 Sample.........
Standard Solution for Mill A
Mill B Peak Areas for Baseline Chromatograph..............
Peak Areas for PSA Standard and 50/50 Sample...............
Standard Solution for Mill B
Peak Areas for Hot Melt Standard and 50/50 Sample.........
Standard Solution for Mill B
x


1. Introduction
1.1 Purpose of the Study
In North America the recycling of Old Corrugated Containers (OCC) has grown by
more than 50 percent since 1990 to 19 million tons per year (1). This is partly due to
the fact that landfill space is decreasing nation-wide. Technological advances in
paperboard manufacturing have made it possible to use more OCC in the production
of new paperboard products such as the liner and medium for corrugated containers.
OCC contains contaminants such as pressure sensitive adhesives (PSAs) and hot
melts (see Section 1.2) which can affect the appearance and physical properties of the
finished paperboard product. During the manufacturing of paperboard, these
contaminants can agglomerate when subjected to shear, pH or temperature shock.
The agglomerated contaminants can form large sticky particles that deposit on
process equipment and finished product. Some of the larger stickies can be
removed from the process using mechanical, mineral or chemical methods such as
screening, cleaners, mineral pacification or flocculant removal.
Over the years, contaminants in OCC pulp have been studied using a variety of tests.
The most widely used method is solvent extraction followed by Fourier Transform
Infrared spectroscopy (FTIR). This method is only qualitative and only allows
contaminant identification but does not quantify the contaminants. In addition, the
above method is only used to identify contaminants that are mainly with the pulp
fibers and not in the aqueous phase. As the paperboard industry moves into the
future, more mills are recycling their process water to reduce manufacturing cost and
meet EPA regulations. The recycling of process water allows the small PSAs and
hot-melt particles to concentrate in the process water since they are too small to be
1


removed by mechanical equipment. Depending on the concentration, surfactants can
be used to keep the micro-contaminants suspended in water, but as the concentration
increases over time, the contaminants are more likely to deposit on process equipment
and finished product. Flocculants can be used to increase the particle size of the
contaminants so they can be removed using mechanical means but this method is
rather costly to mills.
There is a need for a test method that could both identify and quantify OCC
contaminants in process water. This would help mills monitor their process and to
find solutions to remove contaminants and run efficiently. In addition, a test method
would allow mills to determine the efficiency of their current methods for removing
and/or pacifying the micro-contaminants. In the end, mills could avoid deposit
products associated with contaminant concentration building up in their process
water.
1.2 History of Pressure Sensitive Adhesives and Hot
Melts
1.2.1 Pressure Sensitive Adhesives (PSAs)
The history of adhesives goes back many thousands of years. Archaeologists have
discovered a burial site from 4000 B.C. where broken clay pots had been put back
together with tree sap. The first adhesives were made out of natural substances such
as tree sap or from the protein of animal bones, hide, hoofs and horns. Today, a lot of
synthetic adhesives exist for a variety of applications. One of the largest applications
for adhesives is in the packaging industry (2).
Pressure sensitive adhesives (PSAs) are used for labels and packaging tapes. The
fact that they are tacky at room temperature makes them sensitive to pressure. PSAs
can be made out of natural and synthetic elastomers or acrylic polymers. Today, most
of the formulations for PSAs use acrylic polymers such as vinyl acrylate (See Figure
2


1.2.1.1) and a tackifier. As the name suggests, a tackifier is used to increase the
tackiness of adhesive so it will adhere to many substrates. PSAs are supplied in an
emulsified suspension to meet EPA regulations regarding volatile organic
compounds.
Figure 1.2.1.1 Vinyl Acrylate Monomer
Molecular Weight = 98 amu
H H
I 1 s
-{-G-C)-
1 i n
O H
0=0
t
H-C
H
1.2.2 Hot Melts
The use of hot melts first started with people using waxes to seal documents and
letters. It was not until the 1960s when the art of hot melts was advanced with new
formulations (2). Now, their primary use is for sealing boxes and cartons. They are
made with thermoplastic polymers and are solid at room temperature. At elevated
temperature hot melts become a viscous fluid and are easy to applied. Three
components are used to formulate hot melts: high molecular weight copolymers, a
resin tackifier and wax. The most common copolymer system used to formulate a hot
melt is ethylene vinyl acrylate (EVA) (See Figure 1.2.2.1). The tackifier resin is
usually a low molecular weight polymer that gives the hot melt specific viscosity and
adhesion properties. Wax is added in the form of paraffin or synthetic hydrocarbons
and helps control the surface wetting characteristics, viscosity, melting point and
flexibility of the hot melt.
3


Figure 1.2.2.1 Ethylene Vinyl Acetate Monomer
Molecular Weight = 114 amu
H H H H
I I 1 I .
H H O H
C=0
H-C-H
H
1.3 Literature Search
A significant amount of research has recently been aimed at quantifying the resin and
fatty acid content from wood fibers in relation to organic deposit (pitch) problems in
pulp or paper mills. The most commonly used analytical methods use a solvent
extraction step with the analysis being performed using FTIR or gas chromatography
/ mass spectroscopy (GC/MS) (3). These methods are variations of two methods
written in the 1970s by the National Council of the Paper Industry for Air Stream
Improvement (NCASI) and British Columbia Research (BCR) (3). Both original
methods used packed column GC for analysis but varied in analyte isolation. For
example, the NCASI method used acidified (pH 2-3) diethyl ether solvent extraction
while the BCR method used adsorption of the analyte onto a polymeric resin under
alkaline conditions (pH 9-10). Generally, these methods are used to extract the resin
and fatty acids from the wood fibers itself. Now, there is higher interest in analyzing
a mills process water since more mills are discharging process water. Monitoring
and compliance of discharge process water is a major undertaking under the penalty
of heavy fines if violations are committed.
4


During the middle eighties a new method was reported for determining resin and fatty
acids in mill process water that included solvent extraction using methyl-t- butyl ether
under alkaline conditions (3). The extract was methylated using diazomethane and
analyzed with a GC with a flame ionization detector (FID). The results showed this
method to be relatively fast and accurate and did not produce foam, which sometimes
happens when a solvent is mixed with an aqueous sample.
While solvent extraction was being studied, research for the paper industry was being
done in solid phase extraction. A paper on analyzing organics from mill deposits,
water process streams and black liquor clean up samples using solid phase extraction
was published in the late eighties (4). IR and GC/MS spectra were obtained of mill
process chemicals such as oil based defoamer and rosin size used in pulp process for
a particular mill. Samples were initially dissolved in a solvent and solid phase
extraction was used to collect organic compounds within the sample such as
defoamer, resin acids and rosin size. The extracted resin acids were methylated with
diazomethane before being analyzed by GC/MS. For the mill deposits, the organic
components were removed using methanol Soxhlet extraction. The extracted sample
was put through an octadecysilane (Cis) solid phase extraction cartridge and hexane
and chloroform were used to elute the organic compounds from the cartridge. The
solvents eluted different defoamer hydrocarbons and wood resin components from the
cartridge. Additionally, solid phase extraction was used to separate the organics from
a mills paper machine white water (6). The white water was directly passed through
the cartridge without performing prior solvent extraction. The organics identified in
the paper machine white water were hydrocarbon oils, resin, and fatty acids. The
solid phase extraction cartridges were able to remove about 80% of the organic
compounds from the white water. Directly passing black liquor samples through two
solid phase extraction cartridges was able to remove 70% of the organics. The author
explained the decrease in collection efficiency in terms of the large amounts of
5


organics present in the black liquor. In general, simple solid phase extraction was
found to be a convenient and efficient method for separating out organic compounds
found in deposits and water process streams.
As the 80s came to an end, a paper was published that reported some results for wood
extractives from pulp and water samples from mechanical pulping process of Spruce
(5). This research was aimed at helping the mill discharge less contaminants (wood
extractives) that can be damaging to aquatic life. Compared with earlier papers, the
authors developed a simple scheme (Figure 1.3.1 and Figure 1.3.2) breaking down the
extractives into lipophilic (pitch) extractives and polar extractives.
Figure 1.3.1 Pulp Sample Extraction Scheme
6


Figure 1.3.2 Water Sample Extraction Scheme
In the new scheme, dichloromethane (DCM) was used first to extract the lipophilic
compounds followed by an acetone / water solution to extract the polar compounds.
The DCM extractables were methylated with diazomethane and then silylated. The
acetone extractables were only silylated. The extractives were quantified with
GC/MS. The mass spectra from previous spruce research were used to identify the
lipophilic and polar compounds. The water samples used were obtained by filtering
the pulp fibers from the process water. For the filtered samples, the lipophilic
extractives were extracted using diethyl ether while the polar extractives were
concentrated by evaporating the water by vacuum or freeze drying and then ethanol
was added to the dried solids. As before, the lipophilic extractives were methylated
and then silylated whereas the polar compounds were only silylated. Both types of
compounds were analyzed using GC/MS. The GC was able to resolve a majority of
the compounds found in the mixture using a capillary column. Mass spectra of
compounds from previous work were used to identify almost 50 compounds.
7


Additional research has been performed in the area of extractives analysis from pulp
and mill effluent over the last decade. In the early 90s, a paper was published
outlining a new method for determining wood extractives in papermaking process
water and effluent (6). In this work, water samples were first collected from inside
the mill and from outside the mill (effluent). Following centrifugation to remove
wood fines and non-dissolved solids, the water phase was extracted with methyl-t-
butyl ether to extract the organics. Extraction was performed at three pH values 3.5,
6.5 and 8.0. The most acidic extraction had the highest quantity of extractables.
Each sample was silylated and analyzed using GC with a flame ionization detector
(FID). Reportedly, some compounds could not be resolved with their GC using a 5-
meter column. The authors mentioned that a longer column could possibly resolve all
the compounds in the sample. UV spectrometry was performed at 280nm to
determine if all organic compounds can be extracted using methyl tert-butyl ether.
The absorbance results showed that 80% of the pitch was extracted. The 20%
remaining non-extractables were attributed to lignin and other polar colloidal
substances that were not extracted. The method proved to be convenient for
analyzing multiple samples at one time.
During the later part of the nineties, the major source of pulp fiber for the paper
industry continued to be trees. Over that time, more research was done to determine
the wood extractables in paper mill effluent due to the rising concern over the toxicity
of paper mill discharge water. A paper was written that outlined the analysis of wood
extractives for a variety of species used by European paper mills (7). This study
included the analysis of the effluent from the mills. The lipophilic components of the
wood species were extracted using acetone and analyzed using a GC with an FID
detector. A bioassay was performed on the mill effluent to determine the toxicity of
the water. A correlation was drawn between the lipophilic components and bioassay
8


for the different wood species. This study was the first to show a relationship
between the toxicity of process water and the lipophilic components from wood.
Over the same period, a research study was reported on the analysis of resins from
wood and pulp from two wood species that were processed with two different pulping
methods (8). This study looked at the effect on resins from sulfite and Kraft pulping
on Radiata pine and Eucalyptus. During the different pulping processes resin acids
can be liberated from the wood chips, which can cause deposit problems or increase
the toxicity of the effluent. Organics were extracted using a variety of solvents in a
Soxhlet apparatus. Acetone was the most polar solvent used in this study. GC/MS
showed the liberated resins found in Eucalyptus were largely unaffected by the
different types of pulping methods. The liberated resins found in Radiata pine were
more affected by the different types of pulping. As demonstrated by this study, a
variety of solvents are necessary to extract the total organics from the pulp samples.
As more people become concerned about the environment, paper mills will be forced
to use more recycled fibers such as old newspapers (ONP), old magazines (OMG),
mixed office waste (MOW) and old corrugated containers (OCC). These new fiber
sources contain contaminants, which can deposit onto process equipment and cause
production problems. To better understand these contaminants some research studies
have been undertaken over the years.
In the mid 90s, a paper was published on the analysis of deposits from newsprint
mills using recycled fibers (9). A variety of analysis methods were used to analyze
the deposit samples. Solvent extraction was used to liberate the organic from the
samples using ethanol, hexane, chloroform and acidic acetone. A solid phase
extraction cartridge was used to separate hydrocarbon oils from wood resins. The
calcium and aluminum soaps in the samples were determined by inductively coupled
9


plasma (ICP) spectroscopy. Fourier Transform Infrared spectroscopy (FTIR) was
performed on the deposit samples as well as the solvent extractables. Nuclear
Magnetic Resonance Spectroscopy (NMR) was performed on the chloroform
extractables. Thermogravimetric Analysis (TGA) was performed on the non-
extractables such as wood fiber, organic polymers, and calcium carbonate. GC was
performed on the ethanol extractables such as resin and fatty acids. Analysis took
place after methylation with diazomethane. Energy Dispersive Spectroscopy (EDS)
was used to quantify the inorganics in the samples. A general scheme (Figure 1.3.3)
was devised to analyze the deposit samples with emphasis on identification of major
classes of compounds found in the deposits.
Figure 1.3.3 Mill Deposit General Extraction Scheme
ORIGINAL MILL DEPOSIT
It was determined from this scheme that the major compounds found in the deposits
for this newsprint mill came from deinking chemicals and ink residue.
A paper about the analysis and characterization of contaminants from OCC furnish
was published in 2000 (10). In this paper, samples of stickies contaminants were
10


collected from three mills using 100 % OCC furnish. First, the stickies were
subjected to a solvent immersion cycle with ethanol using a Soxhlet apparatus for one
hour at 80C. Then the solvent immersion cycle was performed with acetone for two
hours. The solid stickies and extraction samples were characterized qualitatively by
FTIR. In addition, differential Scanning Calorimetry (DSC), X-ray fluorescence,
Scanning Electron Microscope (SEM) and particle size analyzer were used to
characterize the samples qualitatively. X-ray fluorescence and SEM were used to
determine the inorganic content in the samples. The Differential Scanning
Calorimetry was used to determine the glass transition temperature (Tg) of the
polymers in the samples. From the literature, this study showed that one could
identify the different organic polymer components in the samples by comparing Tg
literature with Tg analyzed. The samples that were characterized for particle size
were also quantified Using FTIR. A correlation was found between the polymer type
responsible for a given particle size found in the sample. This study used a novel
approach of characterizing the polymers that make up the contaminants found in OCC
furnish.
The test method development in this thesis is based on the above-mentioned studies.
It is evident that a significant amount of work has been done previously for the pulp
and paper industry on using solvent and solid phase extraction as well as Gas
Chromatography / Mass Spectrometry. However, the quantification of polymeric
contaminants in board mills using OCC deserves special attention both from the
process improvement and environmental standpoints.
11


2. Experimental Methods
2.1 Standards and Samples
A commercial PSA and hot melt were utilized as standards to be investigated by
GC/MS. Samples from two paperboard mills were analyzed during the development
of this test method.
2.1.1 Preparation of Standards
The PSA (Techcryl 2007) was obtained from Dyna-Tech Adhesives Corporation and
the hot melt (Final Bond) was obtained from Industrial Adhesives. Techcryl 2007 is a
polyacrylic pressure sensitive adhesive that is used as the adhesive for packaging tape
and package labels. This PSA was utilized in this study because it has similar
chemistry to the other PSAs used for packaging adhesives. Plus, the US Postal
Service certified Techcryl 2007 as meeting their packaging adhesive test
requirements. Final Bond is a hot melt copolymer of ethylene and vinyl acetate
(EVA) used widely for sealing the bottom of cardboard boxes. This hot melt has
similar chemistry to the other hot melt adhesives used for sealing boxes. A series of
dilutions ranging from 0.10 to 1.00 percent were prepared for the standards.
Chloroform was used as the solvent. A 25-mL volumetric flask was used to prepare
the standard solution. The density of chloroform was used to calculate the initial
weight of the 25 mL of solution. Initially, the weight of standard was based on the
weight of 25 mL of solvent. The weight of standard was corrected for the final
weight of the solution. The standards were initially heated to 50C to reduce the
preparation time. The 0.10 percent standard for both Techcryl 2007 and Final Bond
was the only stable standard at room temperature (approximately 25C). At and
above 0.25 percent, the standards formed large polymer particles after 24 hours.
12


2.1.2 Preparation of Samples
Samples were collected from two OCC Mills. One of the mills is located in the east
while the other is located in the northwest. Neither of the mills process water systems
were 100 percent closed. They clarify their process water and some of it is
discharged to a local sewage waste plant for treatment prior to being discharged into a
river. The samples were first vacuum filtered using number 40-Whatman filter paper
to remove fibers from the process water. Initial testing was performed to characterize
the process water. Once the initial characterization was completed, 300ml of process
water was placed into a 1L beaker. The 1L beaker was placed into an oven for 48
hours at 110C. After drying, the 1L beaker was placed in a desiccator for one hour
to cool. The dried sample was scraped from the bottom of the beaker using a metal
spatula. The sample solutions were prepared using the same protocol as the standard
solutions. Sample solutions were prepared using 1.00-percent dried sample based on
total solution weight. Solutions were heated at 50C for three hours and allowed to sit
for twenty-four hours before analyzing.
2.2 Instrumentation
2.2.1 Gas Chromatography (GC)
A widely used instrument for separating organic compounds from a mixture is a gas
chromatograph (GC). The GC used for this research was a Varian 3400. The Varian
3400 needed some repairs before it could be used for this study. The injection port
where the sample is first vaporized needed to be replumbed because the pressure was
not sufficiently high. In addition, the heating block for the injection port had to be
rewired to ensure a constant vaporization temperature. To vaporize the sample, the
injection port is kept at 260C. The injector is being used in split mode to ensure that
the detector (MS) does not get overloaded with sample. Once vaporized a carrier gas
or mobile phase pushes the vaporized sample into a column where separation of the
13


organic compounds takes place. New tubing and filters were installed for the carrier
assembly. For the purposes of this research helium was the inert carrier gas, and a
Varian capillary column (CP-SIL 8CB-MS) was used. The column was 30 meters
with and inner diameter of 0.25 millimeters and a film thickness of 0.25 micrometers.
Once the sample is in the column, separation takes places through absorption and
desorption of the organic compounds in the sample mixture from the stationary phase
on the capillary column. For this research, the stationary phase was: 5%
phenylpolysiloxane and 95% dimethylpolysiloxane column coating. As the column is
heated, the organic polymers that make up the PSAs and hot melt contaminants
desorb from the column coating. Separation takes place because different polymers
will have different affinities for the column coating. As the temperature of the
column is increased these affinities decrease. The temperature ramp that was used for
this research is shown in Figure 2.2.1.1. The temperature range was between 60C
and 285C. The recommended maximum temperature for the column that was used
was 300C.
Figure 2.2.1.1 Gas Chromatograph Parameters
Seg Tenp Rate Tine Total
1 60 0.0 1.00 1.00
2 285 7.9 20.40 29.40
3 205 0.0 45.52 75.00
Start | M|*C | Event 11
End | fefl|C [Event^ I
Tine f 1.001 nin [Event 3 ]
Rate | B.B]C/nin | Event 4~|
1077
Colunn Injector Xfer Line
Sot 60 *C 1 260|C | 2601C
Actual 60 'C 260 "C 260 *C
14


The total time to perform one run was 75 minutes. The molecular weight of the
polymers will dictate the rate the polymers elute the column as the temperature is
increased. Once the polymer elutes from the column, the helium carrier gas pushes it
through the transfer line and into the ion trap.
2.2.2 Mass Spectrometry (MS)
The mass spectrometer that was used for this research was a Varian Saturn 3. The
MS unit had not been used in over six years. Luckily, the turbo pump was
operational and in good working condition. Only the o-ring between the transfer line
and ion trap had to be replaced (Figure 2.2.2.1) (11). The trap was taken apart and
cleaned (Figures 2.2.2.2) (11).
Figure 2.2.2.1 Transfer Line and Ion Trap Assembly
The lens plunger, lens, gate plunger, compression springs and ceramic insulators were
replaced. To ensure efficient ionization the filament assembly was replaced as well,
which is located on top of the ion trap. The parameters that were used in this research
are shown in Figure 2.2.2.3.
15


Figure 2.2.2.2 Ion Trap Assembly
Figure 2.2.2.3 Mass Spectrometer Parameters
|_5B| to |156| m/z
Hass Range
Scan Tine
( 4 uScans)
Segment Length
Fil/flul Delay
Peak Threshold
Hass Defect
Background Hass
Tune File
Description
. 5001 seconds
75.001 minutes
5.00 minutes
l| counts
01 mu/100u
4S.[ ct/z
Ion Mode
Ion Preparation | MS/flS |
Ion Control
Cal Gas
mci |
Auto

G n 1

| Fixed |
C:\SflIUKHSHEIHODS\TUN_DFLI | | Use Last Actiue File |
DEFAULT TUNE PABATIETEBS FILE
Edit/Load
The mass range was 50 to 150 M/Z to reduce background noise. A five-minute delay
was used to ensure that the chloroform had already passed through the filament
16


before it is energized. The relatively large amount of solvent in the testing solutions
would overload the filament and inevitably bum it out in a short period of time. For
this particular test, the ion trap was being used in the following manner. Extractives
were ionized using electron impact (Figure 2.2.2.4) (12). Ions were generated when
the extractive collide with the tungsten filament. A 70-volt potential is used to attract
the extractives toward the tungsten filament. The molecules and electrons in the
electron impact ion source travel at right angles and intersect in the middle where
collisions and ionization take place. Primarily, the ions coming out of the El source
are singly charged positive ions (M + e - M*+ + 2e') produced when energetic
electrons force extractives to lose electrons through electrostatic repulsion. Only a
small number of extractives undergo the primary reaction due to the low efficiency of
the El source. The ions accelerate through a series of slits toward the grid in the
upper end cap of the ion trap due to a 5-volt potential difference that is applied
between the grid and the repeller. Once the ions are in the ion trap, the sides of the
trap contain a pair of doughtnut-shaped ring electrodes, which a variable radio-
frequency voltage is applied to (Figure 2.2.2.5) (13). Ions with a certain M/Z value
stay in the ion trap cavity in a stable orbit surrounded by the electrodes.
17


Figure 2.2.2A Electron Impact Source
To mass
analyzer
The lighter ions destabilize as the radio-frequency voltage is increased, while the
heavier ions are stabilized. The lighter ions collide with the walls of the ion trap as
they become destabilized. The ion trap works by scanning the radio frequency of the
ring electrodes so that the trapped ions become destabilized and are emitted through
the lower end cap of the ion trap into the electron multiplier.
18


Figure 2.2.2.S Ion Trap
Electron
multiplier
transducer
Ion signal
The positively charged ion fragments travel into the electron multiplier (EM) after
traveling through the slits of the bottom end cap electrode shown in Figure 2.2.2.5.
where their signal intensity is increased. The particular EM used in this study is
shown in Figure 2.2.2.6. A variety of EMs can be used in GC/MS instruments. This
particular EM is the continuous dynode type and consists of a lead oxide/glass,
funnel-like resistor (the cathode), and a cup
19


Figure 2.2.2.6 Electron Multiplier
(the anode) at the exit end of the cathode (see figure 2.2.2.7) (12). The top of the
cathode has a negative potential between 800 and 3000 volts applied to it and a
ground potential applied to the exit end of it. The positive ion fragments are attracted
to the top of the cathode by the large negative voltage.
20


Figure 2.2.2.7 Electron Multiplier Continuous Dynode Setup
The positive ion fragments hit the cathode walls with enough energy to dislodge
electrons from the inner curve of the cathode. The dislodged electrons travel further
into the EM because of the decreasing negative potential gradient that exists toward
the bottom of the EM. Due to the curve of the EM, electrons do not travel far before
hitting the cathode walls again. More electrons are ejected from the cathode and
travel further into the EM. This process happens many times before the electrons
reach the bottom of the EM. The cascade of electrons will increase the signal
intensity of the initial polymer fragment approximately 105 times. The anode at the
end of the EM collects all the electrons, which generates an ion current signal that is
sensed by the electrometer circuit of the mass spectrometer than transferred to the
computer and graphed out as a chromatogram.
21


2.3 Extraction Techniques for PSAs and Hot Melts
As previously mentioned, PSAs and hot melts have been analyzed using various
techniques. For this research, various methods were explored for solvent extraction
of PSAs and hot melts from an aqueous phase. Once the contaminants were
extracted from the aqueous phase they were analyzed using GC/MS.
2.3.1 Solvent Extraction
A variety of techniques exist for extracting PSAs and hot melt from an aqueous
phase. Due to its selectivity, solvent extraction is the preferred method for PSAs and
hot melts. Simple solvent extraction can be performed in using a separatory funnel or
a more complicated solvent extraction can be performed using a Soxhlet apparatus
where the solvent is heated at reflux. The challenge with solvent extraction is finding
a solvent that is selective to the compounds of interest. Table 2.3.1.1 shows a variety
of solvents that can be used to extract the different polymers used in PSA and Hot
melt formulations (13). To determine the best solvent for this particular research, a
series of solvency tests were performed at 0.25 percent solids to determine the best
solvent with 100 percent dissolution for Techcryl 2007 and Final Bond (Table
2.3.1.2). The solvency tests showed chloroform and methyl sulfoxide to have good
dissolution after 24 hours at room temperature for the standard adhesives.
Chloroform was used for this study because it is more readily available in the lab
where this study was performed.
22


Table 2.3.1.1 Solubility of Various Polymers used in Adhesive Manufacturing
Extraction Sequence Water Methanol Acetone Pet. Ether Toluene Chloroform n-Heptane
Polymers
Styrene- isoprene N N N S S S N
Styrene- butadiene N N N N S S N
Natural rubber N N N - N S N
Polystyrene N N PS SS S S N
Polyiosbutene N N N SS S N N
Butyl rubber N N N s s N N
Polyvinyl acetate N S S s S N
Polyvinyl alcohol N N N N N N N
EVA N N N s s S N
Polyacrylates N N S - s S N
Polyethylene N N N N N N N
Polypropylene N N N N N N N
Polyvinyl ether N - - - S N
N: non-soluble; S: soluble; PS: Partially soluble; SS: Slightly soluble
Table 2.3.1.2 Solvency Testing
Solvents Techcryl 2007 Final Bond
1. Acetone Soluble Non-soluble
2. Methyl Chloride Soluble Non-soluble
3. Chloroform Soluble Soluble
4. Diethyl Ether Soluble Non-soluble
5. Formamide Soluble Non-soluble
23


Table 2.3.1.2 Solvency Testing (Continued)
Solvents Techcryl 2007 Final Bond
6. Methyl Sulfoxide Soluble Soluble
7. Acetonitrile Soluble Non-soluble
8. Hexane Soluble Non-soluble
9. Toluene Soluble Non-soluble
2.3.2 Solid Phase Extraction (SPE)
In solid phase extraction technique, an adsorbing disk or cartridge is used to extract
the compound of interest from an aqueous sample. A vacuum manifold was used
with the adsorbing disks. About 20ml of process water was pulled through an
adsorbing disk under vacuum instead of using gravity feed. The cartridges that were
used can be attached to a syringe using an adaptor. Using a 10-ml syringe, the
process water is pulled through the cartridge using the syringe plunger. The adsorbing
disk and cartridges contain a solid phase sorbent material, such as alumina or florosil,
or silica coated with compounds such as octdecylsilyl compounds (C-18), which
adsorb the contaminant from the aqueous phase as the sample passes through the
cartridge. Oxygen containing substituents can be added to the Cis to control the
hydrophobic/hydrophilic ratio of the adsorbing disk or cartridge. A variety of SPE
disks and cartridges were tried for this research. Once the compound in question is
collected, a solvent is used to remove it from the collection disk or cartridge. A
concern associated with SPE is that the solvent being used does not dissolve the
adsorbed compound.
2.4 Standard Addition Method
Standard addition method was used to improve the reliability of the test method. The
purpose of standard addition was to decrease the effect that baseline noise had on
24


analyte peak integration. For example, an analyte peak within a sample
chromatograph would be small without the addition of a standard and peak
integration would be more susceptible to being influenced by baseline noise. The
addition of a standard increases the size of the analyte peaks and decreases the overall
influence the baseline noise has on the integration of the analyte peak. Using
standard addition increased the overall precision of the test.
2.5 Identification Techniques
2.5.1 Analysis of Standards
The parameters for the GC/MS were optimized using the PSA and hot melt standards.
Initially, standard solutions ranging in concentration from 1.0 to 0.1 percent were
tested to determine solution stability and chromatograph response. Before the
analysis, each standard was heated to 50C. Standards were injected into the GC
using a 10pl syringe. The standard solution was homogenized before a 5pL aliquot
was drawn into the syringe for analysis. A lpL aliquot of air was drawn into the
syringe after the sample aliquot. The air helped to ensure that no sample was lost
during the process of puncturing the GC septum. The syringe was inserted into the
GC injection port and the GC analysis program was started. To equilibrate the
syringe to the injection port temperature, the syringe was in the injection port for 15
seconds before injection. After injection, the syringe was left in the injection port for
an additional 15 seconds to ensure that the syringe did not interfere with the sample
traveling toward the column. Once the analysis was over, a chromatograph of
standard sample was displayed. The different peaks within the chromatograph
represent the different compounds that make up the PSA or hot melt standard. Three
large peaks within each standard chromatograph were chosen to be their fingerprint
peaks. For each adhesive, the mass spectrum for the three fingerprint peaks was
25


obtained to characterize the standard. The mass spectrum of the fingerprint peaks
was used to identify PSAs, and/or hot melts in samples
2.5.2 Analysis of Samples
GC injections for the samples were performed using the same protocol as the standard
solutions. To obtain baseline chromatographs, samples were analyzed with standards.
The baseline chromatographs were used to determine if the samples contained
fingerprint peaks for PSA, hot melt or both. Peak 4456 is a fingerprint peak for
PSAs and peak 3827 is a fingerprint peak for hot melts. Standard addition technique
was utilized to determine the concentration of adhesive in the sample solutions. First,
4-ml of fifty percent standard at 0.1 percent and fifty percent sample solution was
prepared for analysis. Each adhesive standard was analyzed separately to ensure that
they did not interfere with one another. The fingerprint peaks for the standard/sample
solutions were integrated to determine the change in peak size due to the 0.1 percent
standard (PSA Peaks 2589,4456) (hot melt Peaks 3092, 3827). The concentration of
the separate adhesives can be calculated in the following manner. In the case of the
PSA, the dilution factor is calculated by dividing the peak area of peak 2589 for the
50/50 solution by the peak area of peak 2589 for the PSA standard. The peak area of
peak 4456 is calculated by subtracting the dilution factor multiplied by the peak area
of peak 4456 for the standard. This value is the calculated peak area for peak 4456.
The PSA concentration is calculated by dividing the calculated peak area of peak
4456 by the peak area of peak 4456 for the standard and then multiplying by the
standard concentration, which in this case is 0.1 percent. A similar calculation is used
for hot melts using peaks 3092 and 3827.
26


2.6 SPE of Adhesive Standards
Initial research was performed to determine the best extraction method for removing
PSAs and hot melts from mill process water. Solid phase extraction was performed
following the procedure described in Sections 2.3.2. Experimentation was done using
the PSAs and hot melt standards to determine if SPE could be a suitable collection
technique. The PSA and hot melt standards were dissolved in water using heat. The
temperature ranged between 50 and 60 C, which is the normal processing
temperature for OCC mills. Standards were tested at a variety of concentrations. To
reduce the plugging of the extraction disk and cartridge, a concentration of 0.10 %
was used. After the standards were dissolved, they were passed through the
extraction cartridge and disk. Visually, the adhesives could be seen sticking to the
medium in the extraction disk and cartridge. Adhesive could be seen sticking to the
medium for both extraction disk and cartridge after numerous solvent washes. The
chloroform was heated to 50C and passed through the disk and cartridge. The hot
chloroform seemed to dissolve more of the adhesive standards but only after allowing
the disk and cartridge to soak in the chloroform for two hours. The molecular weight
of the polymers could be too high for the extraction medium to handle. It is believed
that the long chain hydrocarbon polymer is being held tightly by the non-polar
medium of the disk and cartridge. A more polar medium was used with the same
result. The fact that both adhesives would not pass through the extraction disk or
cartridge could be due to the fact that their molecular weights were too large to allow
complete elution of the adhesive from the material.
2.7 SPE of Mill Process Water
The process water of Mills A and B were passed through the extraction disks and
cartridges. In both cases, the amount of process water varied between 5 and 20 mL.
Initial testing was done at 20 mL but the disk clogged after 10 mL and cartridge
clogged after 5 mL. The extraction disks developed a brown spot on top after passing
27


10 mL of the process water through them. The brown spot did not decrease in color
after passing 100 mL of chloroform through the disks. The brown spot did not
decrease in color after being soaked in 50C chloroform for two hours. The
cartridges developed the same brown color after 5 mL of process water, which could
not be eliminated. Chromatographs of the chloroform collected after SPE of the
process water were low in signal, which made differentiating contaminant peaks from
noise impossible. The low signal made the analysis impossible.
2.8 Soxhlet Solvent Extraction of Adhesive Standards
The adhesive standards could be dissolved in chloroform after following the
procedure described in Section 2.3.2. A Soxhlet extraction apparatus was tried at
80C cycling temperature. The adhesive standards clogged the cellulose thimble so
some of the PSA and/or hot melt could be seen at the bottom of the thimble after 4
hours of chloroform cycles.
2.9 Solvent Extraction of Mill Process Water
The mill process water was prepared using the procedure described in Section 2.5.2.
Soxhlet extraction was performed using the same cycling temperature as above. The
chromatographs for the Soxhlet collected chloroform samples were similar to the
separator funnel chloroform extraction samples. Straight chloroform extraction is
preferred because it worked for the adhesives standards and mill process water
samples.
2.10 Precision
The reliability of GC/MS and test method is based on the PSA and hot melt
standards that were made and analyzed. Standards were produced at various
concentrations ranging from 0.1 to 0.5 percent. The 0.1 percent PSA and hot melt
standard had better reproducibility in results than the higher concentration
28


standards. At the higher concentrations, the standards were harder to keep in
solution and in turn harder to keep homogenized, which affected their overall
reproducibility. The 0.1 percent standards were analyzed three times to determine
the reproducibility of the GC/MS used in this study. The reproducibility is based
on the error associated with the heights and areas of the fingerprint peaks for PSA
and hot melt. In addition, the analysis was performed using the same
instrumentation and parameters to help reduce the overall instrument error and
increase the precision.
2.11 Fixed Detection Limit
The fixed detection limit for this test method has been averaged to be a peak area
greater than 1950. A peak area of 1950 is three times the noise peak area around
the analyte peaks. The quantitative limit for this test method is an analyte peak
area greater than 6500. A peak area of 6500 is ten times the noise peak area around
the analyte peaks.
29


3. Results and Discussion
3.1 Process Water Characterization
The process water was tested to determine if organics were present before the
extraction step. The results are shown in Table 3.1.1.
Table 3.1.1 Initial Characterization of Process Water
Mill pH Charge (pequal/L) Conductivity Chemical Oxygen Demand (mg/L)
A 6.49 -130.3 2.50 mS 1534
B 4.84 -106.7 1494 ^iS 1980
Both process water samples contained negatively charged organics, which is
indicative of the water containing compounds such as PSAs or hot melts. The
negative charge is from the fact that PSAs and hot melts contain carbonyl groups,
which are strong electron withdrawing groups. The carbon next to the carbonyl
group will have one extra electron, which gives the polymers a negative charge.
The process water from Mill A contained more negative charge than Mill B as
shown in table 3.1.1. Mill A had a higher conductivity than Mill B, which would
suggest that Mill A contains more organic than Mill B, but the chemical oxygen
demand (COD) data does not support this theory. The COD result for Mill B was
higher, which indicates that Mill B is higher in organics than Mill A.
3.2 PSA Standard Characterization
The standards were analyzed by GC/MS to determine the fingerprint peaks that
make up the adhesives. Throughout this paper, scan numbers will be used to
distinguishing chromatograph peaks and not retention times.
30


Figure 3.2.1 PSA Standard Chromatograph Run 1
Figure 3.2.1 shows the chromatograph for the PSA standard. Four peaks were
chosen for mass spectral identification (Peaks 2589, 4305, 4456, and 5011).
Figures 3.2.2- 3.2.5 show the mass spectra for the four peaks. Table 3.2.1 shows
the possible ions that make up the fragments shown in the mass spectra. The ion
fragments were identified for the polyvinyl acrylate molecule. The monomer was
identified with the M+l peak, which had an atomic mass unit of 99. The mass
spectrum for peak 4305 looks different than the mass spectra for the other
fingerprint peaks. Peak 4305 could be from an additive that was used in the PSA
formulation, which could be a reason why there is a difference in its mass
31


spectrum. The area of peak 4305 will not be used to calculate the PSA
concentration because it may not be part of the polyvinyl acrylate polymer.
Figure 3.2.2 PSA Mass Spectrum Peak 2589
lee* S7
Figure 3.2.3 PSA Mass Spectrum Peak 4305
W/ 70
32


129
Figure 3.2.4 PSA Mass Spectrum Peak 4456
Figure 3.2.5 PSA Mass Spectrum Peak 5011
i88X 57
33


Table 3.2.1 PSA Ion Fragments
34


Table 3.2.1 PSA Ion Fragments (Continued)
Ion Fragments Molecular Weight (amu)
H H 1 i -C-C-H 1 1 0 H c=o ** H-C H 99
H H H 1 1 1 o-o-o- 1 1 1 OHH c=o h-6 H 112
HHHH 1 l 1 1 -c-c-c-c- 1 1 1 ] OHH C=0 H-C H 125
35


Table 3.2.1 PSA Ion Fragments (Continued)
Ion Fragments Molecular Weight (amu)
HHHH lii
ill' c-c-c-c- III
o H H? CO 141
H-C H
3.3 Hot Melt Standard Characterization
Figure 3.3.1 shows the chromatograph for the hot melt standard. Four peaks were
chosen for mass spec identification (Peaks 3092, 3359, 3827, and 4254). Figures
3.3.2- 3.3.5 show the mass spectra for the four peaks. Table 3.3.1 show the
possible ions that make up the fragments shown in the mass spectra. The
fragments shown in the table are more realistic for the hot melt standard than the
fragments shown for the PSA standard because fewer carbons are unprotected.
36


Figure 3.3.1 Hot Melt Standard Chromatograph Run 1
Figure 3.3.2 Hot Melt Mass Spectrum Peak 3092
37


Figure 3.3.3 Hot Melt Mass Spectrum Peak 3359
Figure 3.3.4 Hot Melt Mass Spectrum Peak 3827
s?
38


Figure 3.3.5 Hot Melt Mass Spectrum Peak 4254
Table 3.3.1 Hot Melt Ion Fragments
Ion Fragments Molecular Weight (amu)
H H H H i ill
1 1 1 55
H H H
39


Table 3.3.1 Hot Melt Ion Fragments (Continued)
Ion Fragments Molecular Weight (amu)
H 1
1 -0- 1
0 57
00 1

H H H H i i i
1 i i i -cc-c-c- 1 1 1 1 71
1 1 1 1 H H 0 H
H H H H H i iii
1 i i i i -cc-c-c-c- 1 1 1 1 1 85
H H 0 H H
H H H H H i iii
1 i i i i -cc-c-c-c- 1 1 1 1 1 97
H H 0 H H
-C- i
40


Table 3.3.1 Hot Melt Ion Fragments (Continued)
Ion Fragments Molecular Weight (amu)
H H H H H | 1 1 1 1 -cc-c-c-c- 1 1 1 1 1 H H 0 H H 113
6=o i
H H H H H i iii
1 1 i 1 i -cc-c-c-c- 1 1 1 1 1
1 1 1 1 1 H H OHH 0=0 i H-C-H i 127
3.4 Precision
Tables 3.4.1 through 3.4.4 show the height and area for the PSA peaks from
Section 3.2. The percent error for the height for the peaks ranges between 9.2 and
27.7. The percent error for the area of the peaks ranges between 7.9 and 30.1. The
error is reasonable considering that an older instrument was be used to perform this
study. Good laboratory procedures were used to minimize the error associated for
this test method. Figures 3.4.2 and 3.4.3 show the chromatograph for the last two
runs of the PSA standard. Visually, the chromatographs look identical to the first
run. The baseline error associated with the integration of the peaks should be
consistent throughout the analysis because the automated integration was used.
The automated integration drew the baseline for the peaks as shown in figure 3.4.1.
41


This baseline picture was taken from automated peak integration of peak 4456 for
the PSA standard.
Figure 3.4.1 Baseline Peak Integration of Peak 4456 for PSA Standard
42


Table 3.4.1 PSA Peak 2589
Run Height (Ion Counts) Area (Ion Counts)
1 20334 129857
2 23777 153778
3 20406 135659
Average 21506 139765
Standard Deviation 1967 12478
% Error 9.2 8.9
Table 3.4.2 PSA Peak 4307 Scans
Run Height (Ion Counts) Area (Ion Counts)
1 26602 330316
2 16177 220808
3 28317 413473
Average 23699 321532
Standard Deviation 6570 96632
% Error 27.7 30.1
43


Table 3.4.3 PSA Peak 4456
Run Height (Ion Counts) Area (Ion Counts)
1 10844 53695
2 6866 45923
3 9466 51053
Average 9059 50227
Standard Deviation 2020 3947
% Error 22.3 7.9
Table 3.4.4 PSA Peak 5011
Run Height (Ion Counts) Area (Ion Counts)
1 10036 87082
2 11811 65210
3 15191 77322
Average 12346 76538
Standard Deviation 2619 10957
% Error 21.2 14.3
44


Figure 3.4.2 PSA Standard Chromatograph Run 2
Figure 3.4.3 PSA Standard Chromatograph Run 3
45


Figures 3.4.4 and 3.4.5 show the chromatograph for the last two runs of the hot
melt standard. Visually, the chromatographs look identical to the first run. Tables
3.4.5 through 3.4.8 show the height and area for the hot melt peaks from the
Section 3.4. The percent error for the height for the peaks ranges between 2.9 and
17.1. The percent error for the area of the peaks ranges between 1.6 and 14.6. The
error is reasonable considering that an older instrument was be used to perform this
study.
Figure 3.4.4 Hot Melt Standard Chromatograph Run 2
m:y r
46


Figure 3.4.5 Hot Melt Standard Chromatograph Run 3
lBte
Table 3.4.5 Hot Melt Peak 3092
Run Height (Ion Counts) Area (Ion Counts)
1 70541 574651
2 69347 567567
3 66729 585878
Average 68872 576032
Standard Deviation 1950 9233
% Error 2.8 1.6
47


Table 3.4.6 Hot Melt Peak 3359
Run Height (Ion Counts) Area (Ion Counts)
1 16979 116864
2 21872 149310
3 17450 117299
Average 18767 127824
Standard Deviation 2699 18608
% Error 14.4 14.6
Table 3.4.7 Hot Melt Peak 3827
Run Height (Ion Counts) Area (Ion Counts)
1 25026 233521
2 33974 286403
3 26183 222686
Average 28394 247537
Standard Deviation 4867 34092
% Error 17.1 13.8
48


Table 3.4.8 Hot Melt Peak 4254
Run Height (Ion Counts) Area (Ion Counts)
1 16858 192067
2 21770 241260
3 17114 218903
Average 18581 217410
Standard Deviation 2765 24630
% Error 14.9 11.3
3.5 Mill A Characterization
Figure 3.5.1 shows the baseline chromatograph for Mill A process water. Mill A
does in fact contain both PSAs and hot melts as shown by the presence of peaks
3827 and 4456. Peak 3827 is representative of hot melts while peak 4456
represents PSAs. For PSA, the chromatograph does not contain peaks 2589 and
4305, which are also fingerprint peaks. The chromatograph contains peak 5011,
but this peak will not be used for this study. In the case of hot melt, the
chromatograph does not contain peaks 3092 and 3359. There is a peak at 3827,
which is not proportional in height to peak 4254 as shown in Figure 3.5.1. For the
standard hot melt chromatograph, peak 3827 is larger in height than peak 4254. In
addition, peak 4254 did not have the same fragmentation as peak 4254 for the hot
melt standard. The two large peaks at longer retention times are present on the
standard chromatographs for PSA and hot melt but are small relative to the
fingerprint peaks. Some of the fingerprint peaks could be missing from the sample
chromatograph because of the manufacturing process and chemicals added to the
process water.
49


Figure 3.5.1 Mill A Sample Baseline Chromatograph
180 /.
J01-
Peak 4456
Peak 3827


-i-i--------1>----1
2888 4800
18.66 33.34
(Mb
49.99
66.66
In addition, OCC mills use flocculants to agglomerate contaminants together to form
large particles, which could be why peaks only show up at the higher retention times.
The small particles would come out sooner with smaller retention times. It is
assumed that the peaks that are being investigated in this study are particles because
the retention times for the peaks that do show up are similar in retention times to the
peaks on the standard chromatograph.
50


Figure 3.5.2 Mass Spectrum of Peak 3827 From Mill A Baseline Chromatograph
lBBz 57
Figure 3.5.3 Mass Spectrum of Peak 4456 From Mill A Baseline Chromatograph
100Z 55
51


Figures 3.5.2 and 3.5.3 show the mass spectrums for peaks 3827 and 4456
respectively, which were used to help confirm the above assumption. The spectra
are similar to the standard mass spectra for peaks 3827 and 4456. Some
differences do exist. The base peak for the mass spectrum of peak 4456 has shifted
from 129 amu to 55 amu. The shift is due to fragmentation that was able to take
place when the sample was analyzed and was not possible when the standard was
analyzed. From the mass spectra of baseline chromatograph peaks 3827 and 4456,
one can assume that the peaks in fact represent hot melt and PSA respectively
because they contain the same mass peaks as the mass spectra for the standards.
Table 3.5.1 Mill A Peak Areas for Baseline Chromatograph
Peak Peak Areas Measured from Baseline Chromatograph (Ion Counts) Calculated Peak Areas Step 2 of Cone. Calculation (Ion Counts)
3827 4960 53850
4456 78945 72349
The measured and calculated peak areas for peaks 3827 and 4456 are shown in
Table 3.5.1. Measured peak areas are the integrated peak areas from the baseline
chromatograph. The measured peak areas are used to calculate the peak area in
Appendix A taking into account the dilution factor, which takes into account any
error in sample preparation and instrument injection. Calculated peak areas are
determined in step 2 of Appendix A and are used to calculate the concentration of
adhesives within the sample. From Table 3.5.1, there is a difference between the
measured and calculated areas for the peaks 3827 and 4456. As mentioned
previously, the difference between calculated and measured peak areas are
probably attributed to background noise and/or sample preparation error. The peak
areas are calculated to ensure that all possible experimental errors are accounted for
52


before the contaminant concentrations are calculated. A greater difference exists
between the measured and calculated hot melt peak (3827), which suggests that
there was more error associated with the hot melt analysis for Mill A.
Figure 3.5.4 Mill A 50/50 Sample and PSA Standard Chromatograph
The chromatograph in Figure 3.5.4 shows what happens to the chromatograph
when Mill A sample is mixed with the PSA standard. A noticeable peak at 2589 is
formed with the addition of PSA standard. In addition, a noticeable increase in
height is seen in peak 4456. The chromatograph contains a peak at 4305 but it is
not proportional to peak 4456 as seen in the PSA standard chromatograph. The
height of peak 4305 is larger than peak 4456 in the standard PSA chromatograph.
The difference between the standard and standard/sample peak heights for peaks
4305 and 4456 could be due to the process chemicals that are still present in the
sample. The formation of the peak 2589 is directly related to the addition of PSA
standard as well as the increase in area to peak 4456.
53


The data shown in Table 3.5.2 represent the total areas for peaks 2589 and 4456 for
both the PSA standard and 50/50 sample standard solution for Mill A. As one can
see, the area of peak 2589 for the PSA standard has decreased by approximately 50
percent when it was mixed with sample. This observation helps confirm that the
changes to the baseline chromatograph for sample A are from the addition of PSA
standard and not associated to the GC/MS instrument. The PSA areas in Table
3.5.1 were used in the calculations in Appendix A for determining the PSA
concentration for Mill A. The concentration of PSA for Mill A was calculated to
be 0.14 percent.
Table 3.5.2 Peak Areas Tor PSA Standard and 50/50 Sample Standard Solution for Mill A
Peak PSA Standard Peak Area (Ion Counts) PSA Standard and Sample Area (Ion Counts)
2589 139765 59863
4456 50227 93445
The chromatograph in Figure 3.5.5 shows what happens to the chromatograph
when Mill A sample is mixed with the hot melt standard. The development of a
peak at 3092 is from the addition of PSA standard to the sample. The height of
peak 3827 is increased. The area of peak 3092 will be used to determine the
relative contribution of the standard to be 3827, which is the fingerprint peak for
hot melts.
54


Figure 3.5.5 Mill A 50/50 Sample and Hot Melt Standard Chromatograph
Table 3.5.3 Peak Areas for Hot Melt Standard and 50/50 Sample Standard Solution for
Mill A
Peak Hot Melt Standard Peak Areas (Ion Counts) Hot Melt Standard and Sample Areas (Ion Counts)
3092 576032 192029
3827 247537 135538
The data shown in Table 3.5.3 represent the total areas for peaks 3092 and 3827 for
both the hot melt standard and 50/50 sample standard solution for Mill A. The area
of peak 3092 for hot melt Standard and sample is approximately 65 percent less
than the area for peak 3092 in the hot melt standard. A reduction in peak area is
expected since the hot melt standard was reduced in concentration by fifty percent
when it was next with the sample. This helps confirms that changes in the
chromatograph are from the addition of hot melt standard. The areas in Table 3.5.2
55


were used in the calculations in Appendix A for determining the hot melt
concentration for Mill A. The concentration of hot melt for Mill A was calculated
to be 0.022 percent.
3.6 Mill B Characterization
Figure 3.6.1 shows the baseline chromatograph for Mill B process water. As seen
before in Mill A sample, Mill B does contain both PSAs and hot melts as shown
by the presents of peaks 3827 and 4456. The chromatograph for Mill B is very
similar to Mill A and the same observation as described before for Mill A also
apply to Mill B. This could indicate that the manufacturing process and fiber
supply for Mill B is similar to Mill A. Plus, the addition of PSA and hot melt
standard had a similar effect on the chromatograph Mill B as seen in Mill A.
The mass spectra for peaks 3827 and 4456 of the baseline chromatograph for Mill
B are shown in Figures 3.6.2 and 3.6.3. As with Mill A, the mass spectrums for
Mill B baseline chromatograph are similar to the mass spectrums for the hot melt
and PSA standards. As seen previously for Mill A, the base peak for the mass
spectrum of baseline chromatograph peak 4456 has shifted from 129 amu to 55
amu. The mass spectra for Mill B baseline chromatograph peaks 3827 and 4456
are similar to the mass spectra for Mill A. Mill B mass spectra for the baseline
chromatograph contain similar mass peaks as the mass spectra for the standard, so
one could assume that the peaks in question represent the standards in the baseline
chromatograph.
56


Figure 3.6.1 Mill B Sample Baseline Chromatograph
Figure 3.6.2 Mass Spectrum of Peak 3827 From Mill B Baseline Chromatograph
laa* 57
57


Figure 3.6.3 Mass Spectrum of Peak 4456 From Mill B Baseline Chromatograph
100* 55
Table 3.6.1 Mill B Peak Areas for Baseline Chromatograph
Peak Peak Areas Measured from Baseline Chromatograph (Ion Counts) Calculated Peak Areas Step 2 of Cone. Calculation (Ion Counts)
3827 2329 10801
4456 343252 301502
Table 3.6.1 shows the peak areas for the fingerprint peaks and the peak areas as
calculated in step 2 in Appendix A for Mill B. The same differences and trends
exist as mentioned previously for Mill A. The difference in measured and
calculated peak areas for peak 3827 is greater than it was for Mill A. For peak
4456, the difference in measured and calculated peak area is also greater than it
was for Mill A. It would seem that the standard addition method is more beneficial
58


for peak 3827 (hot melt peak) because its peak area is very small for the baseline
chromatograph.
Figure 3.6.4 Mill B 50/50 Sample and PSA Standard Chromatograph
Figure 3.6.4 shows the formation of peak 2589 when PSA standard is added to Mill
B sample. One can see that the peak area of peak 2589 for the PSA standard and
sample is once again decreased by approximately 50 percent as compared to the
PSA standard. This helps confirm that the addition of PSA standard is responsible
for the changes in the chromatograph. Comparing Mill A to Mill B, the area of
peak 2589 for the PSA standard and sample does not change a great deal. A larger
change is noticed and measured for the area of peak 4456 between the two mills.
59


Table 3.6.2 Peak Areas for PSA Standard and 50/50 Sample Standard Solution for Mill B
Peak PSA Standard Peak Area (Ion Counts) PSA Standard and Sample Area (Ion Counts)
2589 139765 59655
4456 50227 323100
The data in Table 3.6.2 were used to calculate the PSA concentration for Mill B.
Calculation can be seen in Appendix A. The calculated PSA concentration for Mill
B is 0.60 percent. Mill B has a higher concentration of suspended PSAs than Mill
A, which could be due to the fact that Mill B is using a different type of dispersant
than Mill A. The high concentration of PSA is dispersed in the process water
because of the fact that the water went through the filter paper at the sample
preparation step.
Figure 3.6.5 Mill B 50/50 Sample and Hot Melt Standard Chromatograph
Figure 3.6.5 represents the standard addition of hot melt standard to Mill B sample.
Peak 3092 appears when hot melt is added to the sample. One can see that the area
60


of peak 3092 for the hot melt standard is once again reduced by approximately 65
percent when mixed with sample. The peak area for peak 3092 for Mill B is
smaller than it is for Mill A. The difference in peak area represents the difference
in hot melt concentration between the two mills.
Table 3.6.3 Peak Areas for Hot Melt Standard and 50/50 Sample Standard Solution for Mill B
Peak Hot Melt Standard Peak Areas (Ion Counts) Hot Melt Standard and Sample Areas (Ion Counts)
3092 576032 197585
3827 247537 94964
The data in Table 3.6.3 was used to calculate the hot melt concentration for Mill B.
Calculation can be seen in Appendix A. The calculated hot melt concentration for
Mill B is 0.004 percent.
61


4. Conclusions
PSAs and hot melts cause process and production problems for paper mills using
old corrugated containers. The problems usually occur when the small particles
agglomerate or flocculate to form large deposits, which are visible to the human
eye. A test method has been developed that quantifies the small particles of PSAs
and hot melts in the process water from these mills using OCC to make new
paperboard products.
In practice, mill personnel would use this method to monitor the concentration of
PSAs and hot melts in the mill process water. This will enable them to determine
the concentration at which PSAs and hot melt start to cause problem for the mill.
By determining the problematic concentration level for PSAs and hot melts, the
mill will be able to stay below this level to decrease down time for cleaning the
production equipment and in doing so increase productivity. Plus, by using this
test method the can determine if their process water cleaning equipment is running
efficiently or they can optimize the amount of cleaning chemicals they use.
This new method uses gas chromatography/mass spectrometry to measure the PSA
and hot melt concentrations in mill process water. Standard addition is used to
increase the reproducible and decrease experimental error to warrant further study.
For calculating the concentrations of PSA and hot melt, the dilution factor is taking
into account to determine the peak area that would be expected if the experimental
error associated with sample preparation was negligible.
Finally, calculations showed Mill B to have a higher PSA concentration than Mill
A, but Mill A had a higher hot melt concentration than Mill B. Mill B had an
overall higher adhesive concentration than Mill A, which is supported by the fact
that the COD was higher for Mill B than A. The COD values are very useful for a
62


mill to know the total organic concentration of their process water. But, COD
values do not indicate the types of organic compound in the process water.
63


5. Future Work
To further the usefulness of this test method more work should be done using the
equipment in a mill environment to ensure that it will meet their needs. In
addition, one could do more work determining the actual compounds that could be
influencing the MS fragmentation for PSA. This will ensure that the analysis is
actually looking at vinyl acrylate and not another compound within the PSA.
Analysis should be performed to determine the reproducibility between multiple
instruments. To simplify the test method, work could be done using Pyrolysis-
GC/MS, which can eliminate the use of solvent extraction to collect the adhesive
contaminants.
64


Appendix A
General Concentration Calculation
PSA Concentration Calculation
Step 1: Calculate dilution factor for 50/50 Solution PSA and Sample
Peak 2589 Area for 50/50 Solution
Peak 2589 Area for Std.
= Dilution Factor
Step 2: Calculate sample peak area for 50/50 Solution PSA and Sample.
Peak 4456 Area for 50/50 Solution- (Dilution Factor) (Peak 4456 Area for Std.)
= Calculated Sample Peak 4456 Area
Step 3: Calculate PSA concentration
(
Calculated Sample Peak 4456 Area
Peak 4456 Std. Area
(Std. Cone.) = Sample Cone.
Hot Melt Concentration Calculation
Step 1: Calculate dilution factor for 50/50 Solution hot melt and Sample
Peak 3092 Area for 50/50 Solution
Peak 3092 Std. Area
= Dilution Factor
Step 2: Calculate sample peak area for 50/50 Solution hot melt and Sample.
Peak 3827 Area for 50/50 Solution (Dilution Factor) (Peak 3827 Area for Std.)
= Calculated Sample Peak 3827 Area
Step 3: Calculate hot melt Concentration
(
Calculated Sample Peak 3827 Area
Peak 3827 Std. Area
(Sta. Cone.) = Sample Cone.
65


Mill A Concentration Calculations
PSA Concentration
59863 = 42
139765
93445 (.42)(50227) = 72349
72349
50227
* 0.1% =0.14%
Hot Melt Concentration
-1920-29 = .33
576032
135538 (.33)(247537)= 53850
53850
247537
* 0.1% =0.022%
66


Mill B Concentration Calculations
PSA Concentration
59655 _
139765
323100 -(.43)(50227) = 301502
301502
50227 0.1% 0.60%
Hot Melt Concentration
197585
576032
0.34
94964 (.34) (247537) = 10801
10801
247537 -10/o = 0.004%
67


References
(1) J. Routson and J. Kincaid, Eds., 2000 Pulp and Paper Fact Book. San
Francisco: Miller Freeman, 2000.
(2) C. Nicholson, ESC Report: History of Adhesives, [Online Document]
July 1991, [2001 Nov. 20], Available at HTTP: bsahome.org /escreports/
historyoradhesives.org
(3) R. H. Voss, and A Rapsomatiotis. An Improved Solvent-Extraction Based
Procedure For the Gas Chromatographic Analysis of Resin and Fatty Acids
in Pulp Mill Effluents. Journal of Chromatography, vol. 346, no. 1, pp.
205-214.
(4) K. M. Sweeny, Solid-Phase Extraction Techniques in the Pulp and Paper
Industry. Tappi Journal, vol.71, no. 1, pp. 137-140.
(5) R Ekman, and B. Holmbom, Analysis by Gas Chromatography of the
Wood Extractives in Pulp and Water Samples from Machanical Pulping of
Spruce. Nordic Pulp and Paper Research Journal, vol. 1, no. 1, pp. 16-24.
(6) F. Orsa, and B. Holmbom, A Convenient Method for the Determination of
Wood Extractives in Papermaking Process Waters and Effluents. Journal
of Pulp and Paper Science, vol. 20, no. 1, pp. J361-J366.
(7) P. Charlet, G. Lenon, B. Joseleau, and P. Chareyre. Analysis of
Extractives From Different Wood Species. Int. Svmp. Wood Pulping
Chem.. 1997, pp. 15-1 -15-4.
(8) A. F. Wallis, and R. H. Weame, Gas Chromatographic Analysis of Resin
in Woods and Pulps. Int. Svmp. Wood Pulping Chem.. 1997, pp.118-1-
118-4.
(9) X. Y. Guo, and M. Douek. Analysis of Deposits/Stickies from Newsprint
Mills Using Recycled Fibre. Journal of Pulp and Paper Science, vol. 22,
no. 1, pp. J431-J439.
68


(10) J. D. Holbery, D. L. Wood, and R. M. Fisher, Analysis and
Characterization of Contaminants in OCC Recycle Furnishes, Tappi
Journal, vol. 83, no. 1, pp. 1-9.
(11) Varian Associates Inc, Saturn GC/MS Manual, Walnut Creek: Instrument
Group, 1995.
(12) D. A. Skoog, F. J. Holler, and T. A. Nieman, Principles of Instrumental
Analysis, Englewood Cliffs: Prentice Hall, 1998.
(13) C. Negro, and M. C. Monte, Stickies Problems in Recycling, Paper
Recycling: An Introduction to Problems and Their Solutions. Madison:
TAPPI Press, 1997, pp. 48-85.
(14) M. R. Doshi, and J. M. Dyer, Methods to Quantify Stickies A Mill
Survey, Paper Recycling Challenge: Stickies. Appleton: Doshi and
Associates, 1997, pp. 171-193.
(15) D. D. Blevins, M. F. Burke, T. J. Good, P. A. Harris, K. C. Van Horne, N.
Simpson, and L.S. Yago, Sorbent Extraction Technology. Harbor City:
Varian Sample Preparation Products, 1993.
(16) R. R. Rosenberger, and C. J. Houtman, Quantification of Pressure
Sensitive Adhesives, Residual Ink, and Other Colored Process
Contaminants Using Dye and Color Image Analysis, Recent Advances in
Paper Recvcling-Stickies. Madison: TAPPI Press, 2002, pp.69-77.
(17) M. Douek, Overview of Research on Stickies at the Pulp and Paper
research Institute of Canada (PAFRICAN), Paper Recycling Challenge:
Stickies. Appleton: Doshi and Associates, 1997, pp. 15-21.
69


Full Text

PAGE 1

IDENTIFICATION AND QUANTIFICATION OF PROCESS WATER CONTAMINANTS FROM PAPER MILLS USING OLD CORRUGATED CONTAINERS by Jordan Cassidy Kortmeyer B.S., California Polytechnic State University, 1995 A thesis submitted to the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Science Chemistry 2003

PAGE 2

This thesis for the Master of Science degree by Jordan Cassidy Kortrneyer has been approved by Susan Scheible

PAGE 3

Kortmeyer, Jordan Cassidy (M:S., Chemistry) Identification and Quantification of Process Water Contaminants from Paper Mills Using Old Corrugated Containers Thesis directed by Professor Larry G. Anderson I ABSTRACT The use of recycled corrugated containers (OCC) has been steadily growing in the United States over the last two decades. More of these products are being recycled, decreasing the quantities disposed in landfills. OCC is used in the production of new paperboard products such as new corrugated containers and the outer sheets of gypsum board. Using OCC for the production of new paper products can introduce contaminants such as pressure sensitive adhesives (PSA's) and hot melts in the manufacturing process. OCC contaminants can affect appearance and strength of new product. Conventional water treatment techniques can be used to remove contaminants down to about 1 micron in diameter. Contaminants that are at or below 1 micron can pass through screens or hydrocyclones (cleaners) and remain in the manufacturing process water. Eventually, the micro-contaminants agglomerate to cause manufaoturing problems. This new method was developed because currently there is no method for identifying and quantifying PSA's and hot melts in process water. The new method uses solvent extraction to collect the PSA's and hot melts from the process water. Standard addition method was used to increase the overall accuracy of the test. The extracted samples were analyzed using GC/MS. Results showed the new method to be reproducible enough to warrant its use. lll

PAGE 4

This abstract accurately represents the content of the candidate's thesis. I recommend its publication. IV

PAGE 5

ACKNOWLEDGEMENT My thanks to my advisor, Larry Anderson, for his guidance and help in completing this project. I would also like to thank my other advisors: Douglas Dyckes, Susan Scheible, Jorge Yordan and co-workers at Luzenac America for their support and understanding. In addition, I would like to thank the Colorado Commission on Higher Education (CCHE) for funding this project.

PAGE 6

TABLE OF CONTENTS Figures .......................................................................................... viii Tables .............................................................................................. x Chapter 1. Introduction ......................................................................... 1 1 .1 Purpose of the Study ............................................................... .1 1.2 History of Pressure Sensitive Adhesives and Hot Melts ...................... 2 1.2.1 Pressure Sensitive Adhesives (PSA's) ........................................... 2 1.2.2 Hot Melts ............................................................................. 3 1.3 Literature Search .................................................................. .4 2. Experimental Methods ............................................................ 12 2.1 Standards and Samples .......................................................... .l2 2 .1.1 Preparation of Standards ......................................................... 12 2.1.2 Preparation of Samples ........................................................... 13 2.2 Instrumentation ................................................................... 13 2.2.1 Gas Chromatography (GC) ...................................................... 13 2.2.2 Mass Spectrometry (MS) ......................................................... 15 2.3 Extraction Techniques for PSA's and Hot Melts ............................. 22 2.3.1 Solvent Extraction ................................................................ 22 2.3.2 Solid Phase Extraction (SPE) ................................................... 24 2.4 Standard Addition Method ....................................................... 24 2.5 Identification Techniques ........................................................ 25 vi

PAGE 7

2.5.1 Analysis ofStandards ............................................................ 25 2.5.2 Analysis of Samples .............................................................. 26 2.6 SPE of Adhesive Standards ..................................................... 27 2.7 SPE of Mill Process Water. ..................................................... 27 2.8 Soxhlet Solvent Extraction of Adhesive Standards ........................... 28 2.9 Solvent Extraction of Mill Process Water ..................................... .28 2.10 Precision ............................................................................. 28 2.11 Fixed Detection Limit ............................................................. 29 3. Results and Discussion .......................................................... 30 3.1 Process Water Characterization ................................................ .30 3.2 PSA Standard Characterization .................................................. 30 3.3 Hot Melt Standard Characterization ........................................... .36 3.4 Precision .............................................................................. 41 3.5 Mill A Characterization ........................................................ .49 3.6 Mill B Characterization .......................................................... 56 4. Conclusions ........................................................................ 62 5. Future Work ........................................................................ 64 Appendix A. General Concentration Calculation ............................................. 65 References ....................................................................................... 68 Vll

PAGE 8

Figure 1.2.1.1 1.2.2.1 1.3.1 1.3.2 1.3.3 2.2.1.1 2.2.2.1 2.2.2.2 2.2.2.3 2.2.2.4 2.2.2.5 2.2.2.6 2.2.2.7 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.3.1 3.3.2 3.3.3 3.3.4 FIGURES Vinyl Acrylate Monomer .......................................................... 3 Ethylene Vinyl Acetate Monomer ............................................... 4 Pulp Sample Extraction Scheme .................................................. 6 Water Sample Extraction Scheme ............................................... 7 Mill Deposit General Extraction Scheme .................................... 1 0 Gas Chromatograph Parameters ............................................... 14 Transfer Line and Ion Trap Assembly .......................................... 15 Ion Trap Assembly ................................................................ 16 Mass Spectrometer Parameters .................................................. 16 Electron Impact Source ........................................................... 18 Ion Trap ............................................................................ 19 Electron Multiplier ................................................................ 20 Electron Multiplier Continuous Dynode Setup ............................... 21 PSA Standard Chromatograph Run 1 .......................................... 31 PSA Mass Spectrum Peak 2589 ................................................ .32 PSA Mass Spectrum Peak 4305 ................................................. 32 PSA Mass Spectrum Peak 4456 ................................................. 33 PSA Mass Spectrum Peak 5011 ................................................ .33 Hot Melt Standard Chromatograph Run 1 .................................... 3 7 Hot Melt Spectrum Peak 3092 .................................................. 37 Hot Melt Spectrum Peak 3359 .................................................. 38 Hot Melt Spectrum Peak 3827 .................................................. 38 Vlll

PAGE 9

3.3.5 Hot Melt Spectrum Peak 4254 .................................................. 39 3.4.1 Baseline Peak Integration of Peak 4456 for PSA Standard ................ .42 3.4.2 PSA Standard Chromatograph Run 2 ......................................... .45 3.4.3 PSA Standard Chromatograph Run 3 ......................................... .45 3.4.4 Hot Melt Standard Chromatograph Run 2 .................................... .46 3.4.5 Hot Melt Standard Chromatograph Run 3 .................................... .47 3.5.1 Mill A Sample Baseline Chromatograph ....................................... 50 3.5.2 Mass Spectrum for Peak 3827 From Mill A Baseline Chromatograph ... 51 3.5.3 Mass Spectrum for Peak 4456 From Mill A Baseline Chromatograph ... 51 3.5.4 Mill A 50150 Sample and PSA Standard Chromatograph .................. .53 3.5.5 Mill A 50/50 Sample and Hot Melt Standard Chromatograph ............. 55 3. 6.1 Mill B Sample Baseline Chromatograph ....................................... 57 3.6.2 Mass Spectrum for Peak 3827 From Mill B Baseline Chromatograph ... 57 3.6.3 Mass Spectrum for PeakA456 From Mill B Baseline Chromatograph ... 58 3.6.4 Mill B 50/50 Sample and PSA Standard Chromatograph ................... 59 3.6.5 Mill B 50/50 Sample and Hot Melt Standard Chromatograph ............. 60 IX

PAGE 10

Tables 2.3.1.1 2.3.1.2 3.1.1 3.2.1 3.3.1 3.4.1 3.4.2 3.4.3 3.4.4 3.4.5 3.4.6 3.4.7 3.4.8 3.5.1 3.5.2 3.5.3 3.6.1 3.6.2 3.6.3 TABLES Solubility of Various Polymers used in Adhesive Manufacturing ......... 23 Solvency Testing ................................................................... 23 Initial Characterization of Process Water .................................... 30 PSA Ion Fragments ........ ..................................................... 34 Hot Melt Ion Fragments ......................................................... 39 PSA Peak 2589 ................................................................... 43 PSA Peak 4307 ................................................................... 43 PSA Peak 4456 ................................................................... 44 PSA Peak 5011 ................................................................... 44 Hot Melt Peak 3092 ............................................................. .47 Hot Melt Peak 3359 ............................................................. .48 Hot Melt Peak 3827 .............................................................. 48 Hot Melt Peak 4254 .............................................................. 49 Mill A Peak Areas for Baseline Chromatograph ............................. 52 Peak Areas for PSA Standard and 50/50 Sample ...... : ..................... 54 Standard Solution for Mill A Peak Areas for Hot Melt Standard and 50/50 Sample ....................... 55 Standard Solution for Mill A Mill B Peak Areas for Baseline Chromatograph ............................. 58 Peak Areas for PSA Standard and 50/50 Sample ............................ 60 Standard Solution for Mill B Peak Areas for Hot Melt Standard and 50/50 Sample ....................... 61 Standard Solution for Mill B X

PAGE 11

1. Introduction 1.1 Purpose of the Study In North America the recycling of Old Corrugated Containers (OCC) has grown by more than 50 percent since 1990 to 19 million tons per year (1). This is partly due to the fact that landfill space is decreasing nation-wide. Technological advances in paperboard manufacturing have made it possible to use more OCC in the production of new paperboard products such as the liner and medium for corrugated containers. OCC contains contaminants such as pressure sensitive adhesives (PSA's) and hot melts (see Section 1.2) which can affect the appearance and physical properties ofthe finished paperboard product. During the manufacturing of paperboard, these contaminants can agglomerate when subjected to shear, pH or temperature shock. The agglomerated contaminants can form large sticky particles that deposit on process equipment and finished product. Some of the larger "stickies" can be removed from the process using mechanical, mineral or chemical methods such as screening, cleaners, mineral pacification or flocculant removal. Over the years, contaminants in OCC pulp have been studied using a variety of tests. The most widely used method is solvent extraction followed by Fourier Transform Infrared spectroscopy (FTIR). This method is only qualitative and only allows contaminant identification but does not quantify the contaminants. In addition, the above method is only used to identify contaminants that are mainly with the pulp fibers and not in the aqueous phase. As the paperboard industry moves into the future, more mills are recycling their process water to reduce manufacturing cost and meet EPA regulations. The recycling of process water allows the small PSA's and hot-melt particles to concentrate in the process water since they are too small to be 1

PAGE 12

removed by mechanical equipment. Depending on the concentration, surfactants can be used to keep the micro-contaminants suspended in water, but as the concentration increases over time, the contaminants are more likely to deposit on process equipment and fmished product. Flocculants can be used to increase the particle size of the contaminants so they can be removed using mechanical means but this method is rather costly to mills. There is a need for a test method that could both identify and quantify OCC contaminants in process water. This would help mills monitor their process and to find solutions to remove contaminants and run efficiently. In addition, a test method would allow mills to determine the efficiency of their current methods for removing and/or pacifying the micro-contaminants. In the end, mills could avoid deposit products associated with contaminant concentration building up in their process water. 1.2 History of Pressure Sensitive Adhesives and Hot Melts 1.2.1 Pressure Sensitive Adhesives (PSA's) The history of adhesives goes back many thousands of years. Archaeologists have discovered a burial site from 4000 B.C. where broken clay pots had been put back together with tree sap. The first adhesives were made out of natural substances such as tree sap or from the protein of animal bones, hide, hoofs and horns. Today, a lot of synthetic adhesives exist for a variety of applications. One of the largest applications for adhesives is in the packaging industry (2). Pressure sensitive adhesives (PSA's) are used for labels and packaging tapes. The fact that they are tacky at room temperature makes them sensitive to pressure. PSA' s can be made out of natural and synthetic elastomers or acrylic polymers. Today, most of the formulations for PSA's use acrylic polymers such as vinyl acrylate (See Figure 2

PAGE 13

1.2.1.1) and a tackifier. As the name suggests, a tackifier is used to increase the tackiness of adhesive so it will adhere to many substrates. PSA's are supplied in an emulsified suspension to meet EPA regulations regarding volatile organic compounds. Figure 1.2.1.1 Vinyl Acrylate Monomer Molecular Weight= 98 amu HH I 1 -+yy-+n 0 H I C=O I fi-H H-C I H 1.2.2 Hot Melts The use of hot melts first started with people using waxes to seal documents and letters. It was not tmtil the 1960's when the art of hot melts was advanced with new formulations (2). Now, their primary use is for sealing boxes and cartons. They are made with thermoplastic polymers and are solid at room temperature. At elevated temperature hot melts become a viscous fluid and are easy to applied. Three components are used to formulate hot melts: high molecular weight copolymers, a resin tackifier and wax. The most common copolymer system used to formulate a hot melt is ethylene vinyl acrylate (EVA) (See Figure 1.2.2.1 ). The tackifier resin is usually a low molecular weight polymer that gives the hot melt specific viscosity and adhesion properties. Wax is added in the form of paraffin or synthetic hydrocarbons and helps control the surface wetting characteristics, viscosity, melting point and flexibility of the hot melt. 3

PAGE 14

Figure 1.2.2.1 Ethylene Vinyl Acetate Monomer Molecular Weight= 114 amu H H H H I I I 1 -t-c-c-c-c-t I I I I n H H 0 H I C=O I H-C--H H 1.3 Literature Search A significant amount of research has recently been aimed at quantifying the resin and fatty acid content from wood fibers in relation to organic deposit (pitch) problems in pulp or paper mills. The most commonly used analytical methods use a solvent extraction step with the analysis being performed using FTIR or gas chromatography I mass spectroscopy (GC/MS) (3). These methods are variations of two methods written in the 1970s by the National Council of the Paper Industry for Air Stream Improvement (NCASI) and British Columbia Research (BCR) (3). Both original methods used packed column GC for analysis but varied in analyte isolation. For example, the NCASI method used acidified (pH 2-3) diethyl ether solvent extraction while the BCR method used adsorption of the analyte onto a polymeric resin under alkaline conditions (pH 9-10). Generally, these methods are used to extract the resin and fatty acids from the wood fibers itself. Now, there is higher interest in analyzing a mill's process water since more mills are discharging process water. Monitoring and compliance of discharge process water is a major undertaking under the penalty ofheavy fines if violations are committed. 4

PAGE 15

During the middle eighties a new method was reported for determining resin and fatty acids in mill process water that included solvent extraction using methyl-tbutyl ether under alkaline conditions (3). The extract was methylated using diazomethane and analyzed with a GC with a flame ionization detector (FID). The results showed this method to be relatively fast and accurate and did not produce foam, which sometimes happens when a solvent is mixed with an aqueous sample. While solvent extraction was being studied, research for the paper industry was being done in solid phase extraction. A paper on analyzing organics from mill deposits, water process streams and black clean up samples using solid phase extraction was published in the late eighties (4). IR and GC/MS spectra were obtained of mill process chemicals such as oil based defoamer and rosin size used in pulp process for a particular mill. Samples were initially dissolved in a solvent and solid phase extraction was used to collect organic compounds within the sample such as defoamer, resin acids and rosin size. The extracted resin acids were methylated with diazomethane before being analyzed by GC/MS. For the mill deposits, the organic components were removed using methanol Soxhlet extraction. The extracted sample was put through an octadecysilane (Cis) solid phase extraction cartridge and hexane and chloroform were used to elute the organic compounds from the cartridge. The solvents eluted different defoamer hydrocarbons and wood resin components from the cartridge. Additionally, solid phase extraction was used to separate the organics from a mill's paper machine white water (6). The white water was directly passed through the cartridge without performing prior solvent extraction. The organics identified in the paper machine white water were hydrocarbon oils, resin, and fatty acids. The solid phase extraction cartridges were able to remove about 80% of the organic compounds from the white water. Directly passing black liquor samples through two solid phase extraction cartridges was able to remove 70% of the organics. The author explained the decrease in collection efficiency in terms of the large amounts of 5

PAGE 16

organics present in the black liquor. In general, simple solid phase extraction was found to be a convenient and efficient method for separating out organic compounds found in deposits and water process streams. As the 80s came to an end, a paper was published that reported some results for wood extractives from pulp and water samples from mechanical pulping process of Spruce (5). This research was aimed at helping the mill discharge less contaminants (wood extractives) that can be damaging to aquatic life. Compared with earlier papers, the authors developed a simple scheme (Figure 1.3.1 and Figure 1.3.2) breaking down the extractives into lipophilic (pitch) extractives and polar extractives. Figure 1.3.1 Pulp Sample Extraction Scheme PULP SAMPLE I I Freeze-drying and extraction with dlchloromethane followed by acetonetwater (9: 1 V/V) DICHLOROtJIETHANE SOLUBLES ( Lipophilic eKiractlva& ) methylation allylatlon AQUEOUS-ACETONE SOLUBLES ( Polar extractlvee ) hydrolysis methylation &llylatlon sllylatlon GAS CHROMATOGRAPHY Free tatty aclde Free eterol& Free tatty alcohol& Realn acid& Dltarpana alcohols Total latty acids Total aterola Total tatty alcohol& Other totals ( tree + esterified cpda.) 6 Uanana Other phenolic& Sugars Other polar cpda.

PAGE 17

Figure 1.3.2 Water Sample Extraction Scbeme WATER SAMPLE Extraction with Vacwm evaporation dlethyl ather or freeze-drying ( UpophiOc extractives ) C Polar extractives ) methylation sllylallon hydrolysis methylation allylatlon GAS CHROMATOGRAPHY Free latty acids Free sterole Free fatty alcohols Resin acids Ditarpane alcohols Total latty acids Total aterola Total latty alcohols Other totals ( free .. esterified cpds.) sllylatlon Llgnans Other phenolics Sugars Other polar cpds. In the new scheme, dichloromethane (DCM) was used first to extract the lipophilic compounds followed by an acetone I water solution to extract the polar compounds. The DCM extractables were methylated with diazomethane and then silylated. The acetone extractables were only silylated. The extractives were quantified with GC/MS. The mass spectra from previous spruce research were used to identify the lipophilic and polar compounds. The water samples used were obtained by filtering the pulp fibers from the process water. For the filtered samples, the lipophilic extractives were extracted using diethyl ether while the polar extractives were concentrated by evaporating the water by vacuum or freeze drying and then ethanol was added to the dried solids. As before, the lipophilic extractives were methylated and then silylated whereas the polar compounds were only silylated. Both types of compounds were analyzed using GC/MS. The GC was able to resolve a majority of the compounds found in the mixture using a capillary column. Mass spectra of compounds from previous work were used to identify almost 50 compounds. 7

PAGE 18

Additional research has been performed in the area of extractives analysis from pulp and mill effluent over the last decade. In the early 90s, a paper was published outlining a new method for determining wood extractives in papermaking process water and effluent (6). In this work, water samples were first collected from inside the mill and from outside the mill (effluent). Following centrifugation to remove wood fines and non-dissolved solids, the water phase was extracted with methyl+ butyl ether to extract the organics. Extraction was performed at three pH values 3.5, 6.5 and 8.0. The most acidic extraction had the highest quantity of extractables. Each sample was silylated and analyzed using GC with a flame ionization detector (FID). Reportedly, some compounds could not be resolved with their GC using a 5meter column. The authors mentioned that a longer column could possibly resolve all the compounds in the sample. UV spectrometry was performed at 280nm to determine if all organic compounds can be extracted using methyl tert-butyl ether. The absorbance results showed that 80% of the pitch was extracted. The 20% remaining non-extractables were attributed to lignin and other polar colloidal substances that were not extracted. The method proved to be convenient for analyzing multiple samples at one time. During the later part of the nineties, the major source of pulp fiber for the paper industry continued to be trees. Over that time, more research was done to determine the wood extractables in paper mill effluent due to the rising concern over the toxicity of paper mill discharge water. A paper was written that outlined the analysis of wood extractives for a variety of species used by European paper mills (7). This study included the analysis of the effluent from the mills. The lipophilic components of the wood species were extracted using acetone and analyzed using a GC with an FID detector. A bioassay was performed on the mill effluent to determine the toxicity of the water. A correlation was drawn between the lipophilic components and bioassay 8

PAGE 19

for the different wood species. This study was the first to show a relationship between the toxicity of process water and the lipophilic components from wood. Over the same period, a research study was reported on the analysis of resins from wood and pulp from two wood species that were processed with two different pulping methods (8). This study looked at the effect on resins from sulfite and Kraft pulping on Radiata pine and Eucalyptus. During the different pulping processes resin acids can be liberated from the wood chips, which can cause deposit problems or increase the toxicity of the effluent. Organics were extracted using a variety of solvents in a Soxhlet apparatus. Acetone was the most polar solvent used in this study. GC/MS showed the liberated resins found in Eucalyptus were largely unaffected by the different types of pulping methods. The liberated resins found in Radiata pine were more affected by the different types of pulping. As demonstrated by this study, a variety of solvents are necessary to extract the total organics from the pulp samples. As more people become concerned about the environment, paper mills will be forced to use more recycled fibers such as old newspapers (ONP), old magazines (OMG), mixed office waste (MOW) and old corrugated containers (OCC). These new fiber sources contain contaminants, which can deposit onto process equipment and cause production problems. To better understand these contaminants some research studies have been undertaken over the years. In the mid 90s, a paper was published on the analysis of deposits from newsprint mills using recycled fibers (9). A variety of analysis methods were used to analyze the deposit samples. Solvent extraction was used to liberate the organic from the samples using ethanol, hexane, chloroform and acidic acetone. A solid phase extraction cartridge was used to separate hydrocarbon oils from wood resins. The calcium and aluminum soaps in the samples were determined by inductively coupled 9

PAGE 20

plasma (ICP) spectroscopy. Fourier Transform Infrared spectroscopy (FTIR) was performed on the deposit samples as well as the solvent extractables. Nuclear Magnetic Resonance Spectroscopy (NMR) was performed on the chloroform extractables. Thermogravimetric Analysis (TGA) was performed on the non extractables such as wood fiber, organic polymers, and calcium carbonate. GC was performed on the ethanol extractables such as resin and fatty acids. Analysis took place after methylation with diazomethane. Energy Dispersive Spectroscopy (EDS) was used to quantify the inorganics in the samples. A general scheme (Figure 1.3.3) was devised to analyze the deposit samples with emphasis on identification of major classes of compounds found in the deppsits. Figure 1.3.3 Mill Deposit General Extraction Scheme ORIGINAL MILL DEPOSrT + It was determined from this scheme that the major compounds found in the deposits for this newsprint mill.came from deinking chemicals and ink residue. A paper about the analysis and characterization of contaminants from OCC furnish was published in 2000 (10). In this paper, samples of stickies contaminants were 10

PAGE 21

collected from three mills usmg I 00 % OCC furnish. First, the stickies were subjected to a solvent immersion cycle with ethanol using a Soxhlet apparatus for one hour at 80C. Then the solvent immersion cycle was performed with acetone for two hours. The solid stickies and extraction samples were characterized qualitatively by FTIR. In addition, differential Scanning Calorimetry (DSC), X-ray fluorescence, Scanning Electron Microscope (SEM) and particle size analyzer were used to characterize the samples qualitatively. X-ray fluorescence and SEM were used to determine the inorganic content in the samples. The Differential Scanning Calorimetry was used to determine the glass transition temperature (T g) of the polymers in the samples. From the .literature, this study showed that one could identify the different organic polymer components in the samples by comparing Tg literature with T g analyzed. The samples that were characterized for particle size were also quantified using FTIR. A correlation was found between the polymer type responsible for a given particle size found in the sample. This study used a novel approach of characterizing the polymers that make up the contaminants found in OCC furnish. The test method development in this thesis is based on the above-mentioned studies. It is evident that a significant amount of work has been done previously for the pulp and paper industry on using solvent and solid phase extraction as well as Gas Chromatography I Mass Spectrometry. However, the quantification of polymeric contaminants in board mills using OCC deserves special attention both from the process improvement and environmental standpoints. 11

PAGE 22

2. Experimental Methods 2.1 Standards and Samples A commercial PSA and hot melt were utilized as standards to be investigated by GC/MS. Samples from two paperboard mills were analyzed during the development of this test method. 2.1.1 Preparation of Standards The PSA (Techcryl 2007) was obtained from DynaTech Adhesives Corporation and the hot melt (Final Bond) was obtained from Industrial Adhesives. Techcryl2007 is a polyacrylic pressure sensitive adhesive that is used as the adhesive for packaging tape and package labels. This PSA was utilized in this study because it has similar chemistry to the other PSA's used for packaging adhesives. Plus, the US Postal Service certified Techcryl 2007 as meeting their packaging adhesive test requirements. Final Bond is a hot melt copolymer of ethylene and vinyl acetate (EVA) used widely for sealing the bottom of cardboard boxes. This hot melt has similar chemistry to the other hot melt adhesives used for sealing boxes. A series of dilutions ranging from 0.10 to 1.00 percent were prepared for the standards. Chloroform was used as the solvent. A 25-mL volumetric flask was used to prepare the standard solution. The density of chloroform was used to calculate the initial weight of the 25 mL of solution. Initially, the weight of standard was based on the weight of 25 mL of solvent. The weight of standard was corrected for the final weight of the solution. The standards were initially heated to 50C to reduce the preparation time. The 0.10 percent standard for both Techcryl 2007 and Final Bond was the only stable standard at room temperature (approximately 25C). At and above 0.25 percent, the standards formed large polymer particles after 24 hours. 12

PAGE 23

2.1.2 Preparation of Samples Samples were collected from two OCC Mills. One of the mills is located in the east while the other is located in the northwest. Neither of the mills process water systems were 1 00 percent closed. They clarify their process water and some of it is discharged to a local sewage waste plant for treatment prior to being discharged into a river. The samples were first vacuum filtered using number 40-Whatman filter paper to remove fibers from the process water. Initial testing was performed to characterize the process water. Once the initial characterization was completed, 300ml of process water was placed into a 1L beaker. The 1L beaker was placed into an oven for 48 hours at 11 0C. After drying, the 1 L beaker was placed in a desiccator for one hour to cool. The dried sample was scraped from the bottom of the beaker using a metal spatula. The sample solutions were prepared using the same protocol as the standard solutions. Sample solutions were prepared using 1.00-percent dried sample based on total solution weight. Solutions were heated at 50C for three hours and allowed to sit for twenty-four hours before analyzing. 2.2 Instrumentation 2.2.1 Gas Chromatography (GC) A widely used instrument for separating organic compounds from a mixture is a gas chromatograph (GC). The GC used for this research was a Varian 3400. The Varian 3400 needed some repairs before it could be used for this study. The injection port where the sample is first vaporized needed to be replumbed because the pressure was not sufficiently high. In addition, the heating block for the injection port had to be rewired to ensure a constant vaporization temperature. To vaporize the sample, the injection port is kept at 260C. The injector is being used in split mode to ensure that the detector (MS) does not get overloaded with sample. Once vaporized a carrier gas or mobile phase pushes the vaporized sample into a column where separation of the 13

PAGE 24

organic compounds takes place. New tubing and filters were installed for the carrier assembly. For the purposes of this research helium was the inert carrier gas, and a Varian capillary column (CP-SIL 8CB-MS) was used. The column was 30 meters with and inner diameter of 0.25 millimeters and a film thickness of 0.25 micrometers. Once the sample is in the column, separation takes places through absorption and desorption of the organic compounds in the sample mixture from the stationary phase on the capillary column. For this research, the stationary phase was: 5% phenylpolysiloxane and 95% dimethylpolysiloxane column coating. As the column is heated, the organic polymers that make up the PSA's and hot melt contaminants desorb from the column coating. Separation takes place because different polymers will have different affinities for the column coating. As the temperature of the column is increased these affinities decrease. The temperature ramp that was used for this research is shown in Figure 2.2.1.1. The temperature range was between 60C and 285C. The recommended maximum temperature for the column that was used was 300C. Figure 2.2.1.1 Gas Chromatograph Parameters 285 c j/ 68 II .1 ("inutes) Colunn Start I Euent :11 End IEuent zl Set Seg Tenp Rate Tine 1 611 11.11 1.110 z 285 7.9 28.49 3 295 0.0 45.52 75.88 11177 Co lunn I njectcr lifer Ll ne 68 c [ill] c c=::!]0C/11tin Tine Actual 60 c 260 c 260 c Rate 14 Total 1.1111 29.49 75.00

PAGE 25

The total time to perform one run was 75 minutes. The molecular weight of the polymers will dictate the rate the polymers elute the column as the temperature is increased. Once the polymer elutes from the column, the helium carrier gas pushes it through the transfer line and into the ion trap. 2.2.2 Mass Spectrometry (MS) The mass spectrometer that was used for this research was a Varian Saturn 3. The MS unit had not been used in over six years. Luckily, the turbo pump was operational and in good working condition. Only the o-ring between the transfer line and ion trap had to be replaced (Figure 2.2.2.1) (11 ). The trap was taken apart and cleaned (Figures 2.2.2.2) (11). Figure 2.2.2.1 Transfer Line and Ion Trap Assembly GC The lens plunger, lens, gate plunger, compression springs and ceramic insulators were replaced. To ensure efficient ionization the filament assembly was replaced as well, which is located on top of the ion trap. The parameters that were used in this research are shown in Figure 2.2.2.3. 15

PAGE 26

Figure 2.2.2.2 Ion Trap Assembly --------Figure 2.2.2.3 Mass Spectrometer Parameters lon llode l_sal to rralz Scan Ti111e .secilnds c 4 uScansl Ion Preparation I L Seg111ent Length l 75 BEll ;, i nu tes rillllui Dela!;l Peak Threshold llass [31 l!lui1B8u Backgroundlla8s [3 n/z Ion Control Fix,:,d I CalGas . Tune Flle. ._I .c_: ..... \S_ii_T_UR_K_W_T_H_ODS'\.,..._T_U_H;:.D_FL_. r __.l I u,;e Last. Active Fiie I Description TUHE PAIIAIIl:rEl!s FILE EditiLmid The mass range was 50 to 150 M/Z to reduce background noise. A five-minute delay was used to ensure that the chloroform had already passed through the filament 16

PAGE 27

before it is energized. The relatively large amount of solvent in the testing solutions would overload the filament and inevitably burn it out in a short period of time. For this particular test, the ion trap was being used in the following manner. Extractives were ionized using electron impact (Figure 2.2.2.4) (12). Ions were generated when the extractive collide with the tungsten filament. A 70-volt potential is used to attract the extractives toward the tungsten filament. The molecules and electrons in the electron impact ion source travel at right angles and intersect in the middle where collisions and ionization take place. Primarily, the ions coming out of the EI source are singly charged positive ions (M + e ---+ M+ + 2e") produced when energetic electrons force extractives to lose electrons through electrostatic repulsion. Only a small number of extractives undergo the primary reaction due to the low efficiency of the EI source. The ions accelerate through a series of slits toward the grid in the upper end cap of the ion trap due to a 5-volt potential difference that is applied between the grid and the repeller. Once the ions are in the ion trap, the sides of the trap contain a pair of doughtnut-shaped ring electrodes, which a variable radio frequency voltage is applied to (Figure 2.2.2.5) (13). Ions with a certain M/Z value stay in the ion trap cavity in a stable orbit surrounded by the electrodes. 17

PAGE 28

Figure 2.2.2.4 Electron Impact Source The lighter ions destabilize as the radio-frequency voltage is increased, while the heavier ions are stabilized. The lighter ions collide with the walls of the ion trap as they become destabilized. The ion trap works by scanning the radio frequency of the ring electrodes so that the trapped ions become destabilized and are emitted through the lower end cap of the ion trap into the electron multiplier. 18

PAGE 29

Figure 2.2.2.5 Ion Trap End caps lJ Filament ----___-/ e-eCOE9 E9EBO 00 0 Electron multiplier transducer Ion signal Ring electrode The positively charged ion fragments travel into the electron multiplier (EM) after traveling through the slits of the bottom end cap electrode shown in Figure 2.2.2.5. where their signal intensity is increased. The particular EM used in this study is shown in Figure 2.2.2.6. A variety of EMs can be used in GC/MS instruments. This particular EM is the continuous dynode type and consists of a lead oxide/glass, funnel-like resistor (the cathode), and a cup 19

PAGE 30

Figure 2.2.2.6 Electron Multiplier (the anode) at the exit end of the cathode (see figure 2.2.2.7) (12). The top of the cathode has a negative potential between 800 and 3000 volts applied to it and a ground potential applied to the exit end of it. The positive ion fragments are attracted to the top of the cathode by the large negative voltage. 20

PAGE 31

Figure 2.2.2.7 Electron Multiplier Continuous Dynode Setup Multiplier Voltage -,Botto1m Endcap Electrode Amplified lon Current ---to Electrometer Circuit Anode The positive ion fragments hit the cathode walls with enough energy to dislodge electrons from the inner curve of the cathode. The dislodged electrons travel further into the EM because of the decreasing negative potential gradient that exists toward the bottom of the EM. Due to the curve of the EM, electrons do not travel far before hitting the cathode walls again. More electrons are ejected from the cathode and travel further into the EM. This process happens many times before the electrons reach the bottom of the EM. The cascade of electrons will increase the signal intensity of the initial polymer fragment approximately 1 05 times. The anode at the end of the EM collects all the electrons, which generates an ion current signal that is sensed by the electrometer circuit of the mass spectrometer than transferred to the computer and graphed out as a chromatogram. 21

PAGE 32

2.3 Extraction Techniques for PSA's and Hot Melts As previously mentioned, PSA's and hot melts have been analyzed using various techniques. For this research, various methods were explored for solvent extraction of PSA's and hot melts from an aqueous phase. Once the contaminants were extracted from the aqueous phase they were analyzed using GC/MS. 2.3.1 Solvent Extraction A variety of techniques exist for extracting PSA' s and hot melt from an aqueous phase. Due to its selectivity, solvent extraction is the preferred method for PSA's and hot melts. Simple solvent extraction can be performed in using a separatory funnel or a more complicated solvent extraction can be performed using a Soxhlet apparatus where the solvent is heated at reflux. The challenge with solvent extraction is finding a solvent that is selective to the compounds of interest. Table 2.3 .1.1 shows a variety of solvents that can be used to extract the different polymers used in PSA and Hot melt formulations (13). To determine the best solvent for this particular research, a series of solvency tests were performed at 0.25 percent solids to determine the best solvent with 100 percent dissolution for Techcryl 2007 and Final Bond (Table 2.3.1.2). The solvency tests showed chloroform and methyl sulfoxide to have good dissolution after 24 hours at room temperature for the standard adhesives. Chloroform was used for this study because it is more readily available in the lab where this study was performed. 22

PAGE 33

Table 2.3.1.1 Solubility of Various Polymers used in Adhesive Manufacturing Extraction Pet. Sequence Water Methanol Acetone Toluene Chloroform n-Heptane Ether Polymers Styrene-N N N s s s N isoprene StyreneN N N N s s N butadiene Natural N N N -N s N rubber Polystyrene N N PS ss s s N Polyiosbutene N N N ss s N N Butyl rubber N N N s s N N Polyvinyl N s s s s N acetate Polyvinyl N N N N N N N alcohol EVA N N N s s s N Polyacrylates N N s -s s N Polyethylene N N N N N N N Polypropylene N N N N N N N Polyvinyl N s N --ether N: non-soluble; S: soluble; PS: Partially soluble; SS: Slightly soluble Table 2.3.1.2 Solvency Testing Solvents Techcryl2007 Final Bond 1. Acetone Soluble Non-soluble 2. Methyl Chloride Soluble Non-soluble 3. Chloroform Soluble Soluble 4. Diethyl Ether Soluble Non-soluble 5. Formamide Soluble Non-soluble 23

PAGE 34

Table 2.3.1.2 Solvency Testing (Continued) Solvents Techcry12007 Final Bond 6. Methyl Sulfoxide Soluble Soluble 7. Acetonitrile Soluble Non-soluble 8. Hexane Soluble Non-soluble 9. Toluene Soluble Non-soluble 2.3.2 Solid Phase Extraction (SPE) In solid phase extraction technique, an adsorbing disk or cartridge is used to extract the compound of interest from an aqueous sample. A vacuum manifold was used with the adsorbing disks. About 20ml of process water was pulled through an adsorbing disk under vacuum instead of using gravity feed. The cartridges that were used can be attached to a syringe using an adaptor. Using a 10-ml syringe, the process water is pulled through the cartridge using the syringe plunger. The adsorbing disk and cartridges contain a solid phase sorbent material, such as alumina or florosil, or silica coated with compounds such as octdecylsilyl compounds (C-18), which adsorb the contaminant from the aqueous phase as the sample passes through the cartridge. Oxygen containing substituents can be added to the C18 to control the hydrophobic/hydrophilic ratio of the adsorbing disk or cartridge. A variety of SPE disks and cartridges were tried for this research. Once the compound in question is collected, a solvent is used to remove it from the collection disk or cartridge. A concern associated with SPE is that the solvent being used does not dissolve the adsorbed compound. 2.4 Standard Addition Method Standard addition method was used to improve the reliability of the test method. The purpose of standard addition was to decrease the effect that baseline noise had on 24

PAGE 35

analyte peak integration. For example, an analyte peak within a sample chromatograph would be small without the addition of a standard and peak integration would be more susceptible to being influenced by baseline noise. The addition of a standard increases the size of the analyte peaks and decreases the overall influence the baseline noise has on the integration of the analyte peak. Using standard addition increased the overall precision of the test. 2.5 Identification Techniques 2.5.1 Analysis of Standards The parameters for the GC/MS were optimized using the PSA and hot melt standards. Initially, standard solutions ranging in concentration from 1.0 to 0.1 percent were tested to determine solution stability and chromatograph response. Before the analysis, each standard was heated to 50C. Standards were injected into the GC using a 10f.ll syringe. The standard solution was homogenized before a Sf.I.L aliquot was drawn into the syringe for analysis. A 1 f.LL aliquot of air was drawn into the syringe after the sample aliquot. The air helped to ensure that no sample was lost during the process of puncturing the GC septum. The syringe was inserted into the GC injection port and the GC analysis program was started. To equilibrate the syringe to the injection port temperature, the syringe was in the injection port for 15 seconds before injection. After injection, the syringe was left in the injection port for an additional 15 seconds to ensure that the syringe did not interfere with the sample traveling toward the column. Once the analysis was over, a chromatograph of standard sample was displayed. The different peaks within the chromatograph represent the different compounds that make up the PSA or hot melt standard. Three large peaks within each standard chromatograph were chosen to be their fingerprint peaks. For each adhesive, the mass spectrum for the three fingerprint peaks was 25

PAGE 36

obtained to characterize the standard. The mass spectrum of the fingerprint peaks was used to identify PSA's, and/or hot melts in samples 2.5.2 Analysis of Samples GC injections for the samples were performed using the same protocol as the standard solutions. To obtain baseline chromatographs, samples were analyzed with standards. The baseline chromatographs were used to determine if the samples contained fingerprint peaks for PSA, hot melt or both. Peak 4456 is a fingerprint peak for PSA's and peak 3827 is a fingerprint peak for hot melts. Standard addition technique was utilized to determine the concentration of adhesive in the sample solutions. First, 4-ml of fifty percent standard at 0.1 percent and fifty percent sample solution was prepared for analysis. Each adhesive standard was analyzed separately to ensure that they did not interfere with one another. The fingerprint peaks for the standard/sample solutions were integrated to determine the change in peak size due to the 0.1 percent standard (PSA Peaks 2589,4456) (hot melt Peaks 3092, 3827). The concentration of the separate adhesives can be calculated in the following manner. In the case of the PSA, the dilution factor is calculated by dividing the peak area of peak :2589 for the 50150 solution by the peak area of peak 2589 for the PSA standard. The peak area of peak 4456 is calculated by subtracting the dilution factor multiplied by the peak area of peak 4456 for the standard. This value is the calculated peak area for peak 4456. The PSA concentration is calculated by dividing the calculated peak area of peak 4456 by the peak area of peak 4456 for the standard and then multiplying by the standard concentration, which in this case is 0.1 percent. A similar calculation is used for hot melts using peaks 3092 and 3827. 26

PAGE 37

2.6 SPE of Adhesive Standards Initial research was performed to determine the best extraction method for removing PSA's and hot melts from mill process water. Solid phase extraction was performed following the procedure described in Sections 2.3.2. Experimentation was done using the PSA's and hot melt standards to determine if SPE could be a suitable collection technique The PSA and hot melt standards were dissolved in water using heat. The temperature ranged between 50 and 60 oc, which is the normal processing temperature for OCC mills. Standards were tested at a variety of concentrations. To reduce the plugging of the extraction disk and cartridge, a concentration of 0.10 % was used. After the standards were dissolved, they were passed through the extraction cartridge and disk. Visually, the adhesives could be seen sticking to the medium in the extraction disk and cartridge. Adhesive could be seen sticking to the medium for both extraction disk and cartridge after numerous solvent washes. The chloroform was heated to sooc and passed through the disk and cartridge. The hot chloroform seemed to dissolve more of the adhesive standards but only after allowing the disk and cartridge to soak in the chloroform for two hours. The molecular weight of the polymers could be too high for the extraction medium to handle. It is believed that the long chain hydrocarbon polymer is being held tightly by the non-polar medium of the disk and cartridge A more polar medium was used with the same result. The fact that both adhesives would not pass through the extraction disk or cartridge could be due to the fact that their molecular weights were too large to allow complete elution of the adhesive from the material. 2.7 SPE of Mill Process Water The process water of Mills A and B were passed through the extraction disks and cartridges. In both cases, the amount of process water varied between 5 and 20 mL. Initial testing was done at 20 mL but the disk clogged after 10 mL and cartridge clogged after 5 mL. The extraction disks developed a brown spot on top after passing 27

PAGE 38

10 mL of the process water through them. The brown spot did not decrease in color after passing 1 00 mL of chloroform through the disks. The brown spot did not decrease in color after being soaked in 50C chloroform for two hours. The cartridges developed the same brown color after 5 mL of process water, which could not be eliminated. Chromatographs of the chloroform collected after SPE of the process water were low in signal, which made differentiating contaminant peaks from noise impossible. The low signal made the analysis impossible. 2.8 Soxhlet Solvent Extraction of Adhesive Standards The adhesive standards could be dissolved in chloroform after following the procedure described in Section 2.3.2. A Soxhlet extraction apparatus was tried at 80C cycling temperature. The adhesive standards clogged the cellulose thimble so some of the PSA and/or hot melt could be seen at the bottom of the thimble after 4 hours of chloroform cycles. 2.9 Solvent Extraction of Mill Process Water The mill process water was prepared using the procedure described in Section 2.5.2. Soxhlet extraction was performed using the same cycling temperature as above. The chromatographs for the Soxhlet collected chloroform samples were similar to the separator funnel chloroform extraction samples. Straight chloroform extraction is preferred because it worked for the adhesives standards and mill process water samples. 2.10 Precision The reliability of GC/MS and test method is based on the PSA and hot melt standards that were made and analyzed. Standards were produced at various concentrations ranging from 0.1 to 0.5 percent. The 0.1 percent PSA and hot melt standard had better reproducibility in results than the higher concentration 28

PAGE 39

standards. At the higher concentrations, the standards were harder to keep in solution and in turn harder to keep homogenized, which affected their overall reproducibility. The 0.1 percent standards were analyzed three times to determine the reproducibility of the GC/MS used in this study. The reproducibility is based on the error associated with the heights and areas of the fingerprint peaks for PSA and hot melt. In addition, the analysis was performed using the same instrumentation and parameters to help reduce the overall instrument error and increase the precision. 2.11 Fixed Detection Limit The fixed detection limit for this test method has been averaged to be a peak area greater than 1950. A peak area of 1950 is three times the noise peak area around the analyte peaks. The quantitative limit for this test method is an analyte peak area greater than 6500. A peak area of 6500 is ten times the noise peak area around the analyte peaks. 29

PAGE 40

3. Results and Discussion 3.1 Process Water Characterization The process water was tested to determine if organics were present before the extraction step. The results are shown in Table 3 .1.1. Table 3.1.1 Initial Characterization of Process Water Chemical Mill pH Charge (J..Lequai/L) Conductivity Oxygen Demand (mg/L) A 6.49 -130.3 2.50mS 1534 B 4.84 -106.7 1494 J..LS 1980 Both process water samples contained negatively charged orgamcs, which is indicative of the water containing compounds such as PSA's or hot melts. The negative charge is from the fact that PSA's and hot melts contain carbonyl groups, which are strong electron withdrawing groups. The carbon next to the carbonyl group will have one extra electron, which gives the polymers a negative charge. The process water from Mill A contained more negative charge than Mill B as shown in table 3 .1.1. Mill A had a higher conductivity than Mill B, which would suggest that Mill A contains more organic than Mill B, but the chemical oxygen demand (COD) data does not support this theory. The COD result for Mill B was higher, which indicates that Mill B is higher in organics than Mill A. 3.2 PSA Standard Characterization The standards were analyzed by GC/MS to determine the fingerprint peaks that make up the adhesives. Throughout this paper, scan numbers will be used to distinguishing chromatograph peaks and not retention times. 30

PAGE 41

Figure 3.2.1 PSA Standard Chromatograph Run 1 101 zaaa th.(>? 1880 33.3;j Peak 5011 6000 49.99 l sooe (,f) .(.f] Figure 3.2.1 shows the chromatograph for the PSA standard. Four peaks were chosen for mass spectral identification (Peaks 2589, 4305, 4456, and 5011) .. Figures 3.2.23.2.5 show the mass spectra for the four peaks. Table 3.2.1 shows the possible ions that make up the fragments shown in the mass spectra. The ion fragments were identified for the polyvinyl acrylate molecule. The monomer was identified with the M+ 1 peak, which had an atomic mass unit of 99. The mass spectrum for peak 4305 looks different than the mass spectra for the other fingerprint peaks. Peak 4305 could be from an additive that was used in the PSA formulation, which could be a reason why there is a difference in its mass 31

PAGE 42

spectrum. The area of peak 4305 will not be used to calculate the PSA concentration because it may not be part of the polyvinyl acrylate polymer Figure 3.2.2 PSA Mass Spectrum Peak 2589 BIIC I. 69 91 98 Figure 3.2.3 PSA Mass Spectrum Peak 4305 "I 18 I I l ?1'1 .5? 83 32 lOS I It? I 133 12J 11\49 188 118 128 tail jg:, c r'l 1(-a 188 11Z

PAGE 43

Figure 3.2.4 PSA Mass Spectrum Peak 4456 196z ss I I 91 01 i I nj t . Figure 3.2.5 PSA Mass Spectrum Peak 5011 i 33 117 141 r 141 I I 1,17

PAGE 44

Table 3.2.1 PSA Ion Fragments Ion Fragments Molecular Weight (amu) H H H I I I -0-C-C-I I I 57 0 H H I H H I I -0-C-I I 71 0 H I C=O I H H I l -0-C-I I 0 H 83 I C=O I -cI 34

PAGE 45

Table 3.2.1 PSA Ion Fragments (Continued) Ion Fragments HH I 1 -o-o-H I I 0 H I C=O I YH H-C I H HHH I I I -c-c-c1 I I 0 H H I 0=0 C-H H-8 I H HHHH I I I I -c-c-c-c1 I I I OHH I c>O I Y.-H H-C I H Molecular Weight (amu) 99 112 125 35

PAGE 46

Table 3.2.1 PSA Ion Fragments (Continued) Ion Fragments Molecular Weight (amu) HHHH I 1 1 I -e-c-c-eI I I I 0 H HO I I 141 C=O J tr-H H-C H 3.3 Hot Melt Standard Characterization Figure 3.3.1 shows the chromatograph for the hot melt standard. Four peaks were chosen for mass spec identification (Peaks 3092, 3359, 3827, and 4254). Figures 3.3.23.3.5 show the mass spectra for the four peaks. Table 3.3.1 show the possible ions that make up the fragments shown in the mass spectra. The fragments shown in the table are more realistic for the hot melt standard than the fragments shown for the PSA standard because fewer carbons are unprotected. 36

PAGE 47

Figure 3.3.1 Hot Melt Standard Chromatograph Run 1 r Peak 3092 \ Peak 3359 Peak 3827 I Peak4252 I / l lOT _.. .... 11l60 r.ooo 9688 2eri!8 1(..(i(o n.a3 sum M.r.? Figure 3.3.2 Hot Melt Mass Spectrum Peak 3092 18\b: sS l I 83 37 r

PAGE 48

Figure 3.3.3 Hot Melt Mass Spectrum Peak 3359 JJJB;C. $? r 71 I Figure 3.3.4 Hot Melt Mass Spectrum Peak 3827 57 r 71 05 38

PAGE 49

Figure 3.3.5 Hot Melt Mass Spectrum Peak 4254 [ l 71 J i l II tltlftU '. I a!l ... I "I ' trlj""T'IlfT' E.8 78 1118 118 lZ8 l36 1.4fl 158 168 Table 3.3.1 Hot Melt Ion Fragments Ion Fragments Molecular Weight (amu) H H H H I I I I -e-c-c-e-55 I I I I H H H 39

PAGE 50

Table 3.3.1 Hot Melt Ion Fragments (Continued) Ion Fragments H I -c1 0 I C=O I -C--H I H H H H I I I 1 -c-c-c-c1 I I I H H 0 H I H H H H H I I I 1 I -C-C -C-C-C-I I I I I H H 0 H H I H H H H H I I I 1 I -c-c-c-c-c1 I I I I H H 0 H H I -c1 Molecular Weight (amu) 57 71 85 97 40

PAGE 51

Table 3.3.1 Hot Melt Ion Fragments (Continued) Ion Fragments Molecular Weight (amu) H H H H H I I I 1 I -C-C -C-C-CI I I I I 113 H H 0 H H I C=O I H H H H H I I I 1 I -C-C -C-C-CI I I I I H H 0 H H 127 I C=O I H-C-H I 3.4 Precision Tables 3.4.1 through 3.4.4 show the height and area for the PSA peaks from Section 3.2. The percent error for the height for the peaks ranges between 9.2 and 27.7. The percent error for the area of the peaks ranges between 7.9 and 30.1. The error is reasonable considering that an older instrwnent was be used to perform this study. Good laboratory procedures were used to minimize the error associated for this test method. Figures 3.4.2 and 3.4.3 show the chromatograph for the last two runs of the PSA standard. Visually, the chromatographs look identical to the first run. The baseline error associated with the integration of the peaks should be consistent throughout the analysis because the automated integration was used. The automated integration drew the baseline for the peaks as shown in figure 3 .4.1. 41

PAGE 52

This baseline picture was taken from automated peak integration of peak 4456 for the PSA standard. Figure 3.4.1 Baseline Peak Integration of Peak 4456 for PSA Standard Baseline for Automated Peak Integration 42

PAGE 53

Table 3.4.1 PSA Peak 2589 Run Height (Ion Counts) Area (Ion Counts) 1 20334 129857 2 23777 153778 3 20406 135659 Average 21506 139765 Standard 1967 12478 Deviation %Error 9.2 8.9 Table 3.4.2 PSA Peak 4307 Scans Run Height (Ion Counts) Area (Ion Counts) 1 26602 330316 2 16177 220808 3 28317 413473 Average 23699 321532 Standard 6570 96632 Deviation %Error 27.7 30.1 43

PAGE 54

Table 3.4.3 PSA Peak 4456 Run Height (Ion Counts) Area (Ion Counts) 1 10844 53695 2 6866 45923 3 9466 51053 Average 9059 50227 Standard 2020 3947 Deviation %Error 22.3 7.9 Table 3.4.4 PSA Peak 5011 Run Height (Ion Counts) Area (Ion Counts) 1 10036 87082 2 11811 65210 3 15191 77322 Average 12346 76538 Standard 2619 10957 Deviation %Error 21.2 14.3 44

PAGE 55

Figure 3.4.2 PSA Standard Chromatograph Run 2 18ft.-. 'i'Or Peak 4305 Peak4456 Peak 2589 '\ Peak 5011 \ 1/ /_ ..... t I 2006 16.6? 4Q88 33.JJ Figure 3.4.3 PSA Standard Chromatograph Run 3 11111?[ t. ,-b968 56.81 Peak 4456 TOt. I 11889 66.(,(> r I Peak 4305 Peak 5011 Peak2589 ':a. / \ I l-. 2fle8. 16.68 45 _I. 6008 se.et

PAGE 56

Figures 3.4.4 and 3.4.5 show the chromatograph for the last two runs of the hot melt standard. Visually, the chromatographs look identical to the first run. Tables 3.4.5 through 3.4.8 show the height and area for the hot melt peaks from the Section 3.4. The percent error for the height for the peaks ranges between 2.9 and 17 .1. The percent error for the area of the peaks ranges between 1.6 and 14.6. The error is reasonable considering that an older instrument was be used to perform this study. Figure 3.4.4 Hot Melt Standard Chromatograph Run 2 tee?Peak 3092 \ Peak 3359 Peak 3827 I Peak4252 1/ 46 r f

PAGE 57

Figure 3.4.5 Hot Melt Standard Chromatograph Run 3 18ft& Peak 3092 Peak 3359 lOT Peak 3827 u ....... 1 ..... .. -' zeee 16.67 Table 3.4.5 Hot Melt Peak 3092 t 4888 33.31 6009 50.61 8006 ()(l.(,'t r l Run Height (Ion Counts) Area (Ion Counts) 1 70541 574651 2 69347 567567 3 66729 585878 Average 68872 576032 Standard 1950 9233 Deviation %Error 2.8 1.6 47

PAGE 58

Table 3.4.6 Hot Melt Peak 3359 Run Height (Ion Counts) Area (Ion Counts) 1 16979 116864 2 21872 149310 3 17450 117299 Average 18767 127824 Standard 2699 18608 Deviation %Error 14.4 14.6 Table 3.4.7 Hot Melt Peak 3827 Run Height (Ion Counts) Area (Ion Counts) 1 25026 233521 2 33974 286403 3 26183 222686 Average 28394 247537 Standard 4867 34092 Deviation %Error 17.1 13.8 48

PAGE 59

Table 3.4.8 Hot Melt Peak 4254 Run Height (Ion Counts) Area (Ion Counts) 1 16858 192067 2 21770 241260 3 17114 218903 Average 18581 217410 Standard 2765 24630 Deviation %Error 14.9 11.3 3.5 Mill A Characterization Figure 3.5.1 shows the baseline chromatograph for Mill A process water. Mill A does in fact contain both PSA's and hot melts as shown by the presence of peaks 3827 and 4456. Peak 3827 is representative of hot melts while peak 4456 represents PSA's. For PSA, the chromatograph does not contain peaks 2589 and 4305, which are also fingerprint peaks. The chromatograph contains peak 5011, but this peak will not be used for this study. In the case of hot melt, the chromatograph does not contain peaks 3092 and 3359. There is a peak at 3827, which is not proportional in height to peak 4254 as shown in Figure 3.5.1. For the standard hot melt chromatograph, peak 3827 is larger in height than peak 4254. In addition, peak 4254 did not have the same fragmentation as peak 4254 for the hot melt standard. The two large peaks at longer retention times are present on the standard chromatographs for PSA and hot melt but are small relative to the fingerprint peaks. Some of the fmgerprint peaks could be missing from the sample chromatograph because of the manufacturing process and chemicals added to the process water. 49

PAGE 60

Figure 3.5.1 Mill A Sample Baseline Chromatograph taex --.--.---rzeoo 1..6(, 6000 4g';gg In addition, OCC mills use flocculants to agglomerate contaminants together to form large particles, which could be why peaks only show up at the higher retention times. The small particles would come out sooner with smaller retention times. It is assumed that the peaks that are being investigated in this study are particles because the retention times for the peaks that do show up are similar in retention times to the peaks on the standard chromatograph. 50

PAGE 61

Figure 3.5.2 Mass Spectrum of Peak 3827 From Mill A Baseline Chromatograph 1BBY5? 71 as SI1P BHS 99 184 96 1Z7 118 149 76 89 111 lr.lill 143 518 Y.11 I I 111111 ill 'I 'I 'I 'I" 'I 'I 'I "I "I "I 4B 58 68 78 BB 9B 1BB UB 1ZB 138 148 158 168 Figure 3.5.3 Mass Spectrum of Peak 4456 From Mill A Baseline Chromatograph 181l;r. ss SI1P BHG "'I" II j, "I 81 71 91 T1 IIIII II II "I 117 129 141 99 111 Ll J111 till! ll ,,, IL158 'I "I 'I'' 'I '"I 'I 'I' 'I 'I' 51

PAGE 62

Figures 3.5.2 and 3.5.3 show the mass spectrums for peaks 3827 and 4456 respectively, which were used to help confirm the above assumption. The spectra are similar to the standard mass spectra for peaks 3827 and 4456. Some differences do exist. The base peak for the mass spectrum of peak 4456 has shifted from 129 amu to 55 amu. The shift is due to fragmentation that was able to take place when the sample was analyzed and was not possible when the standard was analyzed. From the mass spectra of baseline chromatograph peaks 3827 and 4456, one can assume that the peaks in fact represent hot melt and PSA respectively because they contain the same mass peaks as the mass spectra for the standards. Table 3.5.1 Mill A Peak Areas for Baseline Chromatograph Peak Areas Measured Calculated Peak Areas from Baseline Peak Step 2 of Cone. Calculation Chromatograph (Ion Counts) (Ion Counts) 3827 4960 53850 4456 78945 72349 The measured and calculated peak areas for peaks 3827 and 4456 are shown in Table 3.5.1. Measured peak areas are the integrated peak areas from the baseline chromatograph. The measured peak areas are used to calculate the peak area in Appendix A taking into account the dilution factor, which takes into account any error in sample preparation and instrument injection. Calculated peak areas are determined in step 2 of Appendix A and are use.d to calculate the concentration of adhesives within the sample. From Table 3.5.1, there is a difference between the measured and calculated areas for the peaks 3827 and 4456. As mentioned previously, the difference between calculated and measured peak areas are probably attributed to background noise and/or sample preparation error. The peak areas are calculated to ensure that all possible experimental errors are accounted for 52

PAGE 63

before the contaminant concentrations are calculated. A greater difference exists between the measured and calculated hot melt peak (3827), which suggests that there was more error associated with the hot melt analysis for Mill A. Figure 3.5.4 Mill A 50/SO Sample and PSA Standard Chromatograph 168:1. Peak 4456 1 rot l Peak 2589 \ . \ w-, t. f N.,.,,.,, .. 2888 4889 6800 1&.(i7 3:1 .3 SIUU The chromatograph in Figure 3.5.4 shows what happens to the chromatograph when Mill A sample is mixed with the PSA standard. A noticeable peak at 2589 is formed with the addition of PSA standard. In addition, a noticeable increase in height is seen in peak 4456. The chromatograph contains a peak at 4305 but it is not proportional to peak 4456 as seen in the PSA standard chromatograph. The height of peak 4305 is larger than peak 4456 in the standard PSA chromatograph. The difference between the standard and standard/sample peak heights for peaks 4305 and 4456 could be due to the process chemicals that are still present in the sample. The formation of the peak 2589 is directly related to the addition of PSA standard as well as the increase in area to peak 4456. 53

PAGE 64

The data shown in Table 3.5.2 represent the total areas for peaks 2589 and 4456 for both the PSA standard and 50/50 sample standard solution for Mill A. As one can see, the area of peak 2589 for the PSA standard has decreased by approximately 50 percent when it was mixed with sample. This observation helps confirm that the changes to the baseline chromatograph for sample A are from the addition of PSA standard and not associated to the GC/MS instrument. The PSA areas in Table 3.5.1 were used in the calculations in Appendix A for determining the PSA concentration for Mill A. The concentration of PSA for Mill A was calculated to be 0.14 percent. Table 3.5.2 Peak Areas for PSA Standard and 50/50 Sample Standard Solution for Mill A PSA Standard Peak Area PSA Standard and Sample Area Peak (Ion Counts) (Ion Counts) 2589 139765 59863 4456 50227 93445 The chromatograph in Figure 3.5.5 shows what happens to the chromatograph when Mill A sample is mixed with the hot melt standard. The development of a peak at 3092 is. from the addition of PSA standard to the sample. The height of peak 3827 is increased. The area of peak 3092 will be used to determine the relative contribution of the standard to be 3827, which is the fingerprint peak for hot melts. 54

PAGE 65

Figure 3.5.5 Mill A 50/50 Sample and Hot Melt Standard Chromatograph tor Peak 3827 Peak 3092 I zoe a I 1801) 33.33 68811 se:s1 Table 3.5.3 Peak Areas for Hot Melt Standard and 50/50 Sample Standard Solution for MiliA Hot Melt Standard Peak Hot Melt Standard and Peak Areas (Ion Counts) Sample Areas (Ion Counts) 3092 576032 192029 3827 247537 135538 The data shown in Table 3.5.3 represent the total areas for peaks 3092 and 3827 for both the hot melt standard and 50/50 sample standard solution for Mill A. The area of peak 3092 for hot melt Standard and sample is approximately 65 percent less than the area for peak 3092 in the hot melt standard. A reduction in peak area is expected since the hot melt standard was reduced in concentration by fifty percent when it was next with the sample. This helps confirms that changes in the chromatograph are from the addition of hot melt standard. The areas in Table 3.5.2 55

PAGE 66

were used in the calculations in Appendix A for determining the hot melt concentration for Mill A. The concentration of hot melt for Mill A was calculated to be 0. 022 percent. 3.6 Mill B Characterization Figure 3 .6.1 shows the baseline chromatograph for Mill B process water. As seen before in Mill A sample, Mill B does contain both PSA's and hot melts as shown by the presents of peaks 3 827 and 4456. The chromatograph for Mill B is very similar to Mill A and the same observation as described before for Mill A also apply to Mill B. This could indicate that the manufacturing process and fiber supply for Mill B is similar to Mill A. Plus, the addition of PSA and hot melt standard had a similar effect on the chromatograph Mill B as seen in Mill A. The mass spectra for peaks 3827 and 4456 of the baseline chromatograph for Mill B are shown in Figures 3.6.2 and 3.6.3. As with Mill A, the mass spectrums for Mill B baseline chromatograph are similar to the mass spectrums for the hot melt and PSA standards. As seen previously. for Mill A, the base peak for the mass spectrum of baseline chromatograph peak 4456 has shifted from 129 amu to 55 amu. The mass spectra for Mill B baseline chromatograph peaks 3827 and 4456 are similar to the mass spectra for Mill A. Mill B mass spectra for the baseline chromatograph contain similar mass peaks as the mass spectra for the standard, so one could assume that the peaks in question represent the standards in the baseline chromatograph. 56

PAGE 67

Figure 3.6.1 Mill B Sample Baseline Chromatograph 188'1. Tl)l Peak 4456 Peak3827 \ 1 ,iJ 41188 33.3'1 I 60011 5fL81 ABell li(d)? Figure 3.6.2 Mass Spectrum of Peak 3827 From Mill B Baseline Chromatograph 57 SlfP BKG 71 85 97 49 58 69 78 88 98 198 118 128 138 149 159 169 57

PAGE 68

Figure 3.6.3 Mass Spectrum of Peak 4456 From Mill B Baseline Chromatograph 188:< 55 129 117 141 SI'IP 81 BKG 91 &7 99 7377 95 1ll5 Jl 11,,11 Ill, 111 Ill Ill ,,I . llr II,IJ .1.1111 'I' 'I''" .... .. '1""1' 'I' I""! 48 SB 68 78 BB 98 188 UB 128 138 148 158 lf>B Table 3.6.1 Mill B Peak Areas for Baseline Chromatograph Peak Areas Measured Calculated Peak Areas from Baseline Peak Step 2 of Cone. Calculation Chromatograph (Ion Counts) (Ion Counts) 3827 2329 10801 4456 343252 301502 Table 3.6.1 shows the peak areas for the fingerprint peaks and the peak areas as calculated in step 2 in Appendix A for Mill B. The same differences and trends exist as mentioned previously for Mill A. The difference in measured and calculated peak areas for peak 3827 is greater than it was for Mill A. For peak 4456, the difference in measured and calculated peak area is also greater than it was for Mill A. It would seem that the standard addition method is more beneficial 58

PAGE 69

for peak 3827 (hot melt peak) because its peak area is very small for the baseline chromatograph. Figure 3.6.4 Mill B SO/SO Sample and PSA Standard Chromatograph 168% ror Peak 4456 Peak 2589 \ \ \ zeea u . &? 4008 33.33 Figure 3.6.4 shows the formation of peak 2589 when PSA standard is added to Mill B sample. One can see that the peak area of peak 2589 for the PSA standard and sample is once again decreased by approximately 50 percent as compared to the PSA standard. This helps confirm that the addition of PSA standard is responsible for the changes in the chromatograph. Comparing Mill A to Mill B, the area of peak 2589 for the PSA standard and sample does not change a great deal. A larger change is noticed and measured for the area of peak 4456 between the two mills. 59

PAGE 70

Table 3.6.2 Peak Areas for PSA Standard and 50/50 Sample Standard Solution for Mill B PSA Standard Peak Area PSA Standard and Sample Peak (Ion Counts) Area (Ion Counts) 2589 139765 59655 4456 50227 323100 The data in Table 3.6.2 were used to calculate the PSA concentration for Mill B. Calculation can be seen in Appendix A. The calculated PSA concentration for Mill B is 0.60 percent. Mill B has a higher concentration of suspended PSA's than Mill A, which could be due to the fact that Mill B is using a different type of dispersant than Mill A. The high concentration of PSA is dispersed in the process water because of the fact that the water went through the filter paper at the sample preparation step. Figure 3.6.5 Mill B 50/50 Sample and Hot Melt Standard Chromatograph 166. TOT 2006 16.6? I 681!8 5U.aa 8889 (1(,.(,6 Figure 3.6.5 represents the standard addition of hot melt standard to Mill B sample. Peak 3092 appears when hot melt is added to the sample. One can see that the area 60

PAGE 71

of peak 3092 for the hot melt standard is once again reduced by approximately 65 percent when mixed with sample. The peak area for peak 3092 for Mill B is smaller than it is for Mill A. The difference in peak area represents the difference in hot melt concentration between the two mills. T bl 3 63 P kA a e .. ea reas or ot e t tan ar an ample tan ar o utlon or I ti H M I S d d d 50/50 S I S d d S I ti M"ll B Hot Melt Standard Peak Hot Melt Standard and Peak Areas (Ion Counts) Sample Areas (Ion Counts) 3092 576032 197585 3827 247537 94964 The data in Table 3.6.3 was used to calculate the hot melt concentration for Mill B. Calculation can be seen in Appendix A. The calculated hot melt concentration for Mill B is 0.004 percent. 61

PAGE 72

4. Conclusions PSA' s and hot melts cause process and production problems for paper mills using old corrugated containers. The problems usually occur when the small particles agglomerate or flocculate to form large deposits, which are visible to the human eye. A test method has been developed that quantifies the small particles ofPSA's and hot melts in the process water from these mills using OCC to make new paperboard products. In practice, mill personnel would use this method to monitor the concentration of PSA's and hot melts in the mill process water. This will enable them to determine the concentration at which PSA's and hot melt start to cause problem for the mill. By determining the problematic concentration level for PSA's and hot melts, the mill will be able to stay below this level to decrease down time for cleaning the production equipment and in doing so increase productivity. Plus, by using this test method the can determine if their process water cleaning equipment is running efficiently or they can optimize the amount of cleaning chemicals they use. This new method uses gas chromatography/mass spectrometry to measure the PSA and hot melt concentrations in mill process water. Standard addition is used to increase the reproducible and decrease experimental error to warrant further study. For calculating the concentrations ofPSA and hot melt, the dilution factor is taking into account to determine the peak area that would be expected if the experimental error associated with sample preparation was negligible. Finally, calculations showed Mill B to have a higher PSA concentration than Mill A, but Mill A had a higher hot melt concentration than Mill B. Mill B had an overall higher adhesive concentration than Mill A, which is supported by the fact that the COD was higher for Mill B than A. The COD values are very useful for a 62

PAGE 73

mill to know the total organic concentration of their process water. But, COD values do not indicate the types of organic compound in the process water. 63

PAGE 74

5. Future Work To further the usefulness of this test method more work should be done using the equipment in a mill environment to ensure that it will meet their needs. In addition, one could do more work determining the actual compounds that could be influencing the MS fragmentation for PSA. This will ensure that the analysis is actually looking at vinyl acrylate and not another compound within the PSA. Analysis should be performed to determine the reproducibility between multiple instruments. To simplify the test method, work could be done using Pyrolysis GC/MS, which can eliminate the use of solvent extraction to collect the adhesive contaminants. 64

PAGE 75

Appendix A General Concentration Calculation PSA Concentration Calculation Step 1: Calculate dilution factor for 50/50 Solution PSA and Sample Peak 2589 ... l\rea for 50/50 Solution = Dilution Factor Peak 2589 Area for Std. Step 2: Calculate sample peak area for 50/50 Solution PSA and Sample. Peak 4456 Area for 50/50 Solution(Dilution Fac1or) (Peak 4456 Area for Std.) = Calculated Sample Peak 4456 Area Step 3: Calculate PSA concentration ( Calculated Sample Peak 4456 Area ) ( Std C ) -s 1 c one. -amp e one. Peak 4456 Std. Area Hot Melt Concentration Calculation Step 1: Calculate dilution factor for 50/50 Solution hot melt and Sample Peak 3092 Area for 50/50 Solution D'l t' F t Peak 3092 Std. Area -I 0 Ion ac or Step 2: Calculate sample peak area for 50/50 Solution hot melt and Sample. Peak 3827 Area for 50/50 Solution(Dilution Factor) (Peak 3827 Area for Std.) = Calculated Sample Peak 3827 Area Step 3: Calculate hot melt Concentration ( Calculated Sample Peak 3827 Area) (St C ) -s 1 c o. one. -amp e one. Peak 3827 Std. Area 65

PAGE 76

Mill A Concentration Calculations PSA Concentration 59863 139765 = .42 93445(.42)(50227) = 72349 72349 50227 Hot Melt Concentration 0.1 /o = 0.14 o/o 192029 ---=.33 576032 135538 (.33)(247537) = 53850 53850 247537 0.1 /o =0.022/o 66

PAGE 77

Mill B Concentration Calculations PSA Concentration 59655 139765 .43 323100 -(.43)(50227) = 301502 301502 50227 0.1 /o = 0.60/o Hot Melt Concentration 197585 576032 = 0 3 4 94964-(.34)(247537)= 10801 10801 247537 0.1 /o = 0.004/o 67

PAGE 78

References (1) J. Routson and J. Kincaid, Eds., 2000 Pulp and Paper Fact Book, San Francisco: Miller Freeman, 2000. (2) C. Nicholson, ESC Report: History of Adhesives," [Online Document] July 1991, [2001 Nov. 20], Available at HTTP: bsahome.org /escreports/ historyoradhesives.org (3) R. H. Voss, and A Rapsomatiotis. "An Improved Solvent-Extraction Based Procedure For the Gas Chromatographic Analysis of Resin and Fatty Acids in Pulp Mill Effluents." Journal of Chromatography, vol. 346, no. 1, pp. 205-214. (4) K. M. Sweeny, "Solid-Phase Extraction Techniques in the Pulp and Paper Industry." Tappi Journal, vol.71, no. 1, pp. 137-140. (5) R Ekman, and B. Holmbom, "Analysis by Gas Chromatography of the Wood Extractives in Pulp and Water Samples from Machanical Pulping of Spruce." Nordic Pulp and Paper Research Journal, vol. 1, no. 1, pp. 16-24. (6) F. Orsa, and B. Holmbom, "A Convenient Method for the Determination of Wood Extractives in Papermaking Process Waters and Effluents." Journal of Pulp and Paper Science, vol. 20, no. 1, pp. 1361-1366. (7) P. Charlet, G. Lenon, B. Joseleau, and P. Chareyre. "Analysis of Extractives From Different Wood Species." Int. Symp. Wood Pulping Chern., 1997, pp. 15-1 -15-4. (8) A. F. Wallis, and R. H. Weame, "Gas Chromatographic Analysis of Resin in Woods and Pulps." Int. Symp. Wood Pulping Chern., 1997, pp.ll8-1-118-4. (9) X. Y. Guo, and M. Douek. "Analysis of Deposits/Stickies from Newsprint Mills Using Recycled Fibre." Journal of Pulp and Paper Science, vol. 22, no. 1, pp. J431-J439. 68

PAGE 79

(10) J. D. Holbery, D. L. Wood, and R. M. Fisher, "Analysis and Characterization of Contaminants in OCC Recycle Furnishes," Tappi Journal, vol. 83, no. 1, pp. 1-9. ( 11) Varian Associates Inc, Saturn GC/MS Manual, Walnut Creek: Instrument Group, 1995. (12) D. A. Skoog, F. J. Holler, and T. A. Nieman, Principles of Instrumental Analysis, Englewood Cliffs: Prentice Hall, 1998. (13) C. Negro, and M. C. Monte, "Stickies Problems in Recycling," Paper Recycling: An Introduction to Problems and Their Solutions, Madison: TAPPI Press, 1997, pp. 48-85. (14) M. R. Doshi, and J. M. Dyer, "Methods to Quantify Stickies -A Mill Survey," Paper Recycling Challenge: Stickies, Appleton: Doshi and Associates, 1997, pp. 171-193. (15) D. D. Blevins, M. F. Burke, T. J. Good, P. A. Harris, K. C. Van Horne, N. Simpson, and L.S. Yago, Sorbent Extraction Technology, Harbor City: Varian Sample Preparation Products, 1993. (16) R. R. Rosenberger, and C. J. Houtman, "Quantification of Pressure Sensitive Adhesives, Residual Ink, and Other Colored Process Contaminants Using Dye and Color Image Analysis," Recent Advances in Paper Recycling-Stickies, Madison: TAPPI Press, 2002, pp.69-77. (17) M. Douek, "Overview of Research on Stickies at the Pulp and Paper research Institute of Canada (PAPRICAN)," Paper Recycling Challenge: Stickies, Appleton: Doshi and Associates, 1997, pp. 15-21. 69