CD A U.S. Environmental Protection Agency Industrial Environmental Research      EPA-600/7-78-017
til  f^ Office of Research and Development  Laboratory                        •m*7O
                        Research Triangle Park, North Carolina 27711 February 197o
             EFFECT OF HANDLING
             PROCEDURES ON SAMPLE
             QUALITY
            Interagency
            Energy-Environment
            Research and Development
            Program Report

-------
                       RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology.  Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields.  The seven series
are:

     1.  Environmental Health Effects Research
     2.  Environmental Protection Technology
     3.  Ecological Research
     4.  Environmental Monitoring
     5.  Socioeconomic Environmental Studies
     6.  Scientific and Technical Assessment Reports (STAR)
     7.  Interagency Energy—Environment Research-and Development

This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series.  Reports in this series result from
the effort funded under the 17-agency Federal Energy/Environment
Research and Development Program.  These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems.  The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology.  Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
                                                                      I

This document is available to the public through the National Technical
Information Service, Springfield, Virginia  22161.

-------
                                              EPA-600/7-78-017
                                                 February 1978
EFFECT  OF HANDLING PROCEDURES
             ON SAMPLE QUALITY
                             by
                J.W. Adams, T.E. Doerfler, and C.H. Summers

                        Arthur D. Little, Inc.
                          Acorn Park
                    Cambridge, Massachusetts 02140
                   Contract No. 68-02-2150, T.D. 10501
                     Program Element NO.EHB529
                  EPA Project Officer: Larry D. Johnson

                Industrial Environmental Research Laboratory
                 Office of Energy, Minerals, and Industry
                   Research Triangle Park, N.C. 27711
                          Prepared for

               U.S. ENVIRONMENTAL PROTECTION AGENCY
                  Office of Research and Development
                      Washington, D.C. 20460

-------
                      TABLE OF CONTENTS

                                                           Page
      TABLE OF CONTENTS	   iii

      LIST OF TABLES	   iv

      LIST OF FIGURES	   vi

      ACKNOWLEDGEMENT   	  vii

      SUMMARY   	  viii

  I.  INTRODUCTION  	    1

 II.  EXPERIMENTAL CONSIDERATIONS   	    3

      A.  General Considerations  	    3
      B.  Experimental Design 	    3
      C.  Sample Identification   	    £
      D.  Sample Formulation  	    6
      E.  Handling Cycle  	    8
      F.  Analytical Procedures 	    8
      G.  Data Analysis and Interpretation  	    9

III.  EXPERIMENTAL RESULTS	   11

 IV.  DISCUSSION OF RESULTS   	   22

      A.  Statistical Analysis of Experimental Data ...   22
      B.  Interpretation of Results   	   25

  V.  CONCLUSIONS AND RECOMMENDATIONS   	   37

      REFERENCES	   38

      APPENDICES	   40
                      I
          APPENDIX A  - Formulation Procedures for
                        Samples	   41

          APPENDIX B -  Analytical Results of Metallic
                        Content Contained Within Green
                        Liquor, White Liquor and Green
                        Liquor Extract  	   45

          APPENDIX C -  Gas Chromatography Conditions
                        Employed During Analysis  ....   48
                             iii

-------
                     LIST OF TABLES


Table No.                                             Page

    1      Admissible Experimental Conditions  ...    5

    2      Model Compound Chemical and Physical
              Properties  	   7
    3      Analytical Results from Stored Probe Wash
              Samples (Set I) (Percent Deviation
              from Expected)	, „

    4      Analytical Results from Stored Sorbent
              Trap Condensate Extract Samples (Set II)
              (Percent Deviation from Expected)  ...  14

    5      Analytical Results from Stored Sorbent
              Resin Samples (Set III) (Percent
              Deviation from Expected)  	  15

    6      Statistical Summary of Standard Calibra-
              tion Samples	  16

    7      Results of Extraction Step Experiment
              (Percentage Deviations from Unspiked
              Standard)    	17

    8      Analysis of Variance - Probe Wash Experi-
              ment (Phenol)	23

    9      Probe Wash Experiment (Set I) (Average
              Percentage Deviation from Expected) . .  24

   10      Probe Wash Experiment (Set I) Signifi-
              cant Two-factor Interactions  	  26

   11      Analysis of Variance - Condensate Extract
              Experiment (Phenol) 	  28

   12      Condensate Extract Experiment (Set II)
              (Average Percentage Deviation from
              Expected)    	29

   13      Condensate Extract Experiment (Set II)
              Significant Two-factor Interactions . .  30


                                                 continued.
                            iv

-------
                      LIST OF TABLES  (continued)
      Table No.                                            Page

         14     Analysis of Variance, Sorbent Resin
                   Experiment  (Phenol)  .  .  . .	    32

         15     Sorbent Resin Experiment  (Set III)
                   (Average Percentage  Deviation  from
                   Expected)	    33
APPENDIX A Table No.

         Al     Formulation  Procedure  for Probe Wash
                   Samples (Set I)    	    42

         A2     Formulation Procedure for Sorbent Trap
                   Condensate Extract Samples  (Set II)  .    43

         A3     Formulation Procedure for Sorbent Resin
                   Samples (Set III)    	    44
APPENDIX B Table No.

         Bl     Metallic Content of Green and White
                   Liquors by ICPOES    	    46

         B2     Metallic Content of White Liquor, Green
                   Liquor and Green Liquor Extract by
                   Emission Spectrometry  	    47
APPENDIX C Table No.      (

         Cl      Gas Chromatographic  Conditions   ....    49

-------
                       LIST OF FIGURES
Figure No.                                               Page


    1       GC/MS Chromatogram of Set II Sample
               P-NN-H-RL-S	    19

    2       GC/MS Chromatogram of Set II Sample
               P-NN-R-RD-S	    20

    3       GC/MS Chromatogram of Set III Sample
               AG-AA-C-CD	    21

    4       Probe Wash Experiment (Set I) Significant
               Two-Factor Interactions 	    27

    5       Condensate Extract Equipment (Set II),
               Significant Two-Factor Interactions .  .    31

    6       Sorbent Resin Experiment (Set III)
               Significant Two-Factor Interaction  .  .    34
                               vi

-------
                     ACKNOWLEDGEMENTS









     The authors wish to express their sincerest gratitude




to Drs. Judith C. Harris and Philip L. Levins, as well as to




Messrs. Stephen Spellenberg and James Stauffer of ADL for




their contributions during the design, performance and analy-




sis of the experimental scheme.  Appreciation is also extended




to Drs. Larry Johnson, Robert Statnick and Raymond Merrill of




the Process Measurements Branch, Industrial Environmental




Research Laboratory of the Environmental Protection Agency at




Research Triangle Park for their suggestions and comments ,on




experimental planning and review.
                             vii

-------
                            SUMMARY
     The results of a study designed to evaluate the effects of
typical shipping and storage handling procedures on organic mate-
rials collected in Level 1 studies are presented.  Specific parame-
ters reviewed included sample container composition (amber glass
and high density linear polyethylene), head space composition (air
or nitrogen), temperature (38°C,  21°C,  5°C),  lighting condition
(dark and diffuse sunlight) and catalytic species content.  Three
unique sample sets, representative of fractions obtained during a
Level 1 environmental assessment, were formulated containing six
model organic compounds.  A three-week simulated shipping and
storage cycle was used as representative of elapsed time between
sample collection and analysis.  All three experiments were de-
signed in accordance with statistical principles appropriate for
conducting factorial experiments.  Experimental results were ana-
lyzed by the Analysis of Variance technique in order to assess the
relative effect of each shipping/storage condition studied.

     The major findings of these experiments are as follows:

  •  Generally, any reasonable combination of shipment temperature,
     head space composition, and storage condition can be utilized
     for the short term storage of Level 1 survey analysis samples.

  •  These same conditions can be employed for the short term
     storage of samples undergoing quantitative chemical analysis
     where the accepted range of variation is i 10 percent.

  •  Amber glass containers are preferred for sample storage over
     polyethylene bottles, as it has been found that spurious con-
     taminants may be extracted from the container walls by contact
     with organic solvents.

  •  Diffuse sunlight can penetrate polyethylene containers and
     cause degradation in species that are photoreactive.  Airiber
     glass on the other hand will absorb the harmful wavelengths
     prohibiting photodegradation.
                              viii

-------
                       I.  INTRODUCTION
      Sampling is conducted as part of a Level 1 environmental assess-
 ment study which provides data on the overall composition of process
 and effluent streams.  The collected samples constitute a small por-
 tion of the stream and are assumed to be representative of the entire
 stream.  Therefore, the immediate goal of any process involving
 sample collection is to insure that the results obtained from that
 portion are truly representative of the whole.  As such, all sam-
 pling and analysis schemes are developed using best state of the
 art procedures in an attempt to minimize the occurrence of un-
 characteristic changes.  If interferences are encountered, ap-
 propriate steps are implemented to either quantify their contri-
 bution or to eliminate them altogether.  Failing this, the methods
 are rejected and alternative procedures explored.
     Despite  the rigorous method development practices,  there still
are areas where changes  can  occur and not be detected.   One such
instance occurs between  the  time a sample is first placed within a
container,  and when  it is removed for analysis.  During  this time,
the sample  may be exposed to a number of adverse conditions, such
as heat or  light, that can result in change or degradation.

     Normally, it is preferable  to minimize the length of time that
any sample  is retained prior to  analysis, as this reduces the avail-
able time for changes to occur.   However, in many instances, this
cannot be done, and  alternative  precautionary preservation proce-
dures must  be devised to control sample quality.

     Three  factors are of concern in the development of  any preser-
vation technique:  contamination,  loss and chemical reactivity.
Any one or  combination of these  factors can happen during the ship-
ment and storage of  samples,  compromising the value of their analysis.
It is important, therefore,  that practices be designed to alleviate,
if not eliminate, their occurrence.
                        i
     Many handling procedures are documented within the  literature,
all of which are designed to fulfill a specific requirement.  For
example, plastics of any sort (excluding Teflon) have been shown to
adsorb sample constituents or introduce extraneous contaminants (1, 2).
Specific classes of contaminants  introduced include plasticizers,
antioxidants, colorants and  stabilizers (3).  Thus, many researchers
indicate a  preference for glass  containers, equipped with Teflon or
aluminum foil cap inserts (1, 2,  4).  However, even this technique
is not foolproof as both Teflon  and aluminum may be contaminated
with up to  400 parts per billion  of di-2-ethylhexyl phthalate (5).

-------
     Elaborate container and apparatus cleaning procedures are out-
lined in a number of references (6, 7, 8).  Coupled with extensive
pre-utilization container storage specifications, these procedures
minimize spurious contamination but do not completely eliminate
their possible occurrence.

     During actual shipment and storage operations, many researchers
(1,7,9,10) have noted preferences for either sub-zero temperature
conditions or freeze drying of sample media, but even these pre-
cautions do not eliminate possible compositional variations over
extended periods of time.

     To further complicate matters, many desirable handling practices
cannot be implemented and rigorously followed under adverse field
collection conditions.  Assuming that the analyst is able to collect
samples and package them according to defined practices, their desti-
nations and environments are not easily regulated unless they are
always under the immediate control of the investigator during ship-
ment and storage.

     Recently, the Process Measurements Branch of the Industrial
Environmental Research Laboratory (IERL) published a manual detail-
ing specific procedures to be used for the collection and analysis
of materials during a Level 1 environmental assessment.  However,
since little quantitative data was available describing the impact
of handling procedures, few recommendations were presented for
proper shipment and storage practices.  To supplement the manual,
the Process Measurements Branch of the IERL commissioned this study
to review normally encountered procedures and provide data describ-
ing the impact of chosen conditions upon the sample quality of
organic species in the samples.

     Although a complete review of methods is impractical, it is
hoped that sufficient data will be produced to highlight practices
that may be practically implemented and result in minimum sample
degradation and contamination.

-------
                  II.  EXPERIMENTAL CONSIDERATIONS

A.  General Considerations

    Three main factors contribute to the degradation of sample
quality:  contamination, sample loss, and chemical reactivity or
instability.  Given appropriate conditions, any one or combination
of these factors may occur and become a source of interference
within the retained sample, detracting from its representative
nature.  A number of alternative shipment and storage procedures
have been proposed and implemented, many of which include provi-
sions for the regulation of at least one of the following parameters.

   •  shipping and storage time

   •  sample container composition and quality

   •  shipment and storage temperature

   •  container head space composition

   •  storage lighting condition

   •  presence of catalytic species

    Little quantitative information is available detailing the poten-
tial impact of these three factors upon mixed samples.  Within this
study, various levels of each of these parameters were experimentally
controlled to a assess their contribution to degradation  within
samples.

    It was decided at the outset that test samples would be developed
which were representative of the type of sample collected during a
Level 1 environmental assessment and which contained a variety of
classes of compounds.  These representative samples would be pack-
aged, stored under simulated shipping conditions and returned to the
laboratory where they would await analysis.  The simulated shipment
time would span one week and the laboratory storage time an addi-
tional two weeks.  Analyses would be performed within one week of
the termination of the storage period.

B.  Experimental Design

    Using criteria presented in the Level 1 Procedures Manual,
three experiments were designed to study the effects of the follow-
ing parameters, or factors, on sample integrity:

   •  sample container composition (amber glass or polyethylene),

   •  head space (air, nitrogen, or a combination),

   •  shipment and storage temperatures (38°C, 21°C, 5°C),

-------
    •  storage lighting conditions  (dark or diffuse sun), and

    •  catalytic species content  (acid extractable stainless steel
       components).

    Sets of samples were devised  to simulate representative probe
wash (Set I), sorbent trap condensate extract  (Set II),  and sorbent
resin samples (Set III) which are obtained during a Level 1 source
assessment.  Once  formulated and  packaged, each sample was handled
according to practices normally encountered during transport and
storage cycles.

    Some combinations were eliminated as non-applicable, such as
the conversion back to an air head space over  samples after prelimi-
nary packaging under nitrogen.  Furthermore, a variation in cataly-
tic species content was considered only in the Set II (condensate
extract) experiment.  The sorbent resin experiment (Set  III) was
conducted with additional limitations in the factor/level combina-
tions tested.  However, some of the factor/level combinations were
repeated in order  to achieve greater precision in estimating experi-
mental error.  The final array of factors and  levels chosen for the
three experimental sets is given  in Table 1.

    As the experimental program progressed it  was established that
certain economies  in experimentation were possible by employing
statistical principles of experimental design.  Therefore, only
twenty-seven unique combinations  were actually tested in the con-
densate extract experiment (Set II).  However, the set of twenty-
seven input conditions were selected according to the theory of
fractional factorial experimentation (11, 12), so that meaningful
inferences could be made about each factor independently, as well
as certain predetermined two-factor interactions.  These concepts
are described in greater detail in Section IV-A.

C.  Sample Identification

    A coding system was developed that would allow for the immediate
identification of any sample without reference to a master log.  The
code was derived from the first letter of each important shipping
or storage condition; an example  of which is given below:

                                                           Catalytic
            Head Space       Shipment       Storage         Species
Container   Composition    Temperature     Condition        Content

    AG-         NN-             H-             CD-             S
     +          t               t              .f               .+
Packaged    Shipped in Na  Shipped under  Stored in      Contains acid
in an       Stored in N£   elevated       the  cold       extractable
amber                      temperature    and  in the     stainless
glass                      (hot)          dark           steel consti-
bottle                                                   tuents

A complete list of sample coding  abbreviations is presented in Table 1.

-------
FACTORS
             TABLE 1
Admissible Experimental Conditions
                           LEVELS

(1)
Container
(2)
Shipment/Storage
Head Space
Composition
(3)
Shipment
Temperature
(4)
Storage Condition
Temperature/Lighting
(5)
Catalytic Species
Code
AG
P
AA
AN
NN
H
R
C
RD
KL
CD
N
S
Possible Combination^
Probe Wash
Experiment
(Set I)
Amber Glass
Polyethylene
Air/Air
Air /Nitrogen
Nitrogen/Nitrogen
Hot (37°C)
Room (21 °C)
Cold (5°C)
Room/ Dark
Room/Light
Co Id /Dark
Not analyzed
33 x 2 = 54
Condensate Extract
Experiment
(Set II)
Amber Glass
Polyethylene
Air/Air
Air/Nitrogen
Nitrogen/Nitrogen
Hot
Room
Cold
Room/Dark
Room/Light
Co Id/ Dark
None
Stainless Steel Com-
ponents
33 x 22 = 108
Sorbent Resin
Experiment
(Set III)
Amber Glass
(Not analyzed)
Air/Air
(Not analyzed)
Nitrogen/Nitrogen
Hot
Room
Cold
Room/ Dark
(Not analyzed)
Cold/Dark
Not analyzed
3 x 22 = 12

-------
D.  Sample Formulation

    Model compounds selected for review within this study were
chosen on the basis of the following criteria:

    •  Representative of a variety of different compound classes.

    •  Representative of a boiling point range where losses due to
       volatility may be possible, but do not pose a major
       influence.

    •  Included in concurrent Level 1 analysis studies.

    •  Applicability to ready analysis by gas chromatography.

    Based on these selection criteria, the following six compounds
were subsequently used:

                      Phenol
                      N-Methylaniline
                      4-Chlorobenzaldehyde
                      Acenaphthene
                      n-Hexadecane
                      4,4'-Dichlorobiphenyl

    Specific structural and physical parameters of the chosen species
are detailed in Table 2.

    Using information detailed within the "Level 1 Procedures Manual"
(6), three sets of simulated field samples were formulated, namely
probe wash (Set I), sorbent trap condensate extract (Set II), and
sorbent resin (Set III).  Each set was compounded such that typically
expected field concentrations of selected model species were contained
within the individual sample.  To minimize the variability of compound
concentration within any given sample, master batches of each sample
type were premixed and aliquots removed for packaging.

    As compounded, the probe wash and condensate extract series con-
tained between 30 and 50 milligrams of each model compound, diluted
to a final volume of 300 milliliters with carrier solvent.  As out-
lined in the Procedures Manual, the probe wash (Set I)  solvent was
1:1 methylene chloride:  methanol, and the condensate extract (Set II)
solvent was methylene chloride.  The sorbent resin samples contained
approximately 3 to 5 milligrams of each model compound on 15 +1 grans
of XAD-2.  In this instance a master batch of spiking solution was
made up in pentane and 5 mis pipetted•onto the resin.

    During formulation of the condensate extract set, an attempt was
made to simulate conditions encountered during field utilization of
a.  SASS train.  It was reported that sorbent trap condensate samples
obtained from sources having hydrochloric acid mist and sulfur dioxide
within the effluent stream, contained quantities of hydrochloric and

-------
                                TABLE 2
             Model Compound Chemical and Physical Properties
Species

Phenol
  Class

Oxygenated
  compound
  (phenol)
N-Methylaniline  Nitrogen
                   compound
              Structure
Molecular
  Weight


    94
                          -NHCH3   107
                                              Boiling
                                              Point °C
                                                                182
                                               196
4-Chloroberiz-
  aldehyde
Acenaphthene
n-Hexadecane
Oxygenated
   ygenated        r	^
   compound   Cl -('   Y>-CHO
   (Aldehyde)     V^=L/
                                 141
Polynuclear
  Aromatic
  Hydrocarbon
Straight Chain
  Hydrocarbon   CH3(CH2)ii+CH3
                                                  154
                                                  226
                                               214
                                                277
                                                287
4,4'Dichloro-
  biphenyl
 Halogenated
   Compound    Cl
( Polychlorinated
   biphenyl)
                       -//  \\-Cl
                                                   223
                                               315

-------
sulfuric acid and exhibited a characteristic pale green color  (13).
As  the sorbent trap is fabricated of stainless steel, the color
possibly results from the etching of the metal by the acid and is
indicative of trace metallic content within the condensate liquor.
To  approximate this condition, and assess its potential impact upon
storage stability parameters, simulated green and white liquor solu-
tions were formulated and extracted with methylene chloride for
Set II samples.  Specific details of this process and all other
sample formulation operations are provided within the appendices of
this report.  (Tables A1-A3)

E.  Handling Cycle

    After formulation, samples were placed within the desired  con-
tainer type, with either a nitrogen or air head space, and sealed.
They then were subjected to a three-week handling cycle.  During
the first week, or simulated shipment phase, selected samples were
stored within closed boxes and placed in controlled temperature
zones to simulate conditions expected to arise during truck trans-
port from a remote collection point to a central laboratory.   Three
such zones were used:  one of 38 °.C (100°F), one of 21°C (70°F), and
one of 5°C (40°F).  All samples were then unpacked and rerouted for
two additional weeks to a storage location.  Samples requiring head
space composition change were opened, purged with nitrogen and then
resealed before being stored.  During the storage phase, three
basic combinations existed:

               21°C (70°F) and    diffuse sunlight (northern window)

               21°C (70°F) and    dark (closed cabinet)

                5°C (40°F) and    dark (closed refrigerator)

F.  Analytical Procedures

    Gas chromatography (GC) was utilized as the major quantitative
analysis tool during this evaluation.  Specific experimental details
are presented in the appendices of this report (Table C-l).  As the
probe wash and condensate extract samples were already dilute solu-
tions, no sample preparation was required prior to analysis.  For
sorbent resin samples, a continuous 24-hour Soxhlet extraction was
required to remove the spiked compounds from the XAD-2 resin.  Since
the sample recovery of this step was unknown, several non-stored
spikes were similarly prepared and analyzed along with the stored
samples.

    This was accomplished by placing 15 il gm  of XAD-2 resin into
an  amber glass bottle, and spiking it with 5 mis of a freshly pre-
pared standard solution.  The contents of the bottle were then imme-
diately transferred into a Soxhlet thimble and extracted for 24-hours
with roughly 200 milliliters of methylene chloride.  The resulting
solution was brought to a final volume of 250 milliliters with
methylene chloride,  from which a 5  microliter aliquot was removed

-------
and directly analyzed by GC.  Additional details of this procedure
are discussed in Section III of this report.

     In  all  cases,  a  fresh  standard  solution was  prepared  immediately
before  the  analysis  of  individual sample sets.   This  standard was
required  for  several reasons.   First,  this  sample was useful for
monitoring  the  instrumental parameters.  If integrated peak area
responses were  found to change  between repetitive injections of the
standard  solution, steps could  be taken to  retune the GC  to original
conditions.

      Similarly,  as it was  impractical  to analyze all  stored samples-
on  the  same day,  the fresh standard provided a mechanism  through
which the data  from  non-contiguous  analysis could be  compared.  At
the beginning of each day, the  standard solution was  injected
several times to insure that the instrumental response was equiva-
lent  to prior repetitions.   If  not,  all parameters were rechecked
and adjusted, prior  to  proceeding with the  analysis of the stored
samples.

      Finally, this sample  served as  a  calibtation standard.  Formu-
lated to  duplicate the  composition  of  the stored samples, when they
were  originally  packaged and placed in the simulated shipping se-
quence, this  standard solution  represented  the stored samples in an
unaltered state.   By directly comparing the observed  peak area re-
sponses of  the stored sample to the  fresh standard, and correcting
for minor initial  compounding variations, it becomes  possible to
compute a percentage deviation  attributable to the imposed handling
procedures.

      The  metallic content  of the  green and  white liquors  and  the
methylene chloride extract of the  green liquor was  determined by
spark emission  spectroscopy (ES).   Attempts to  obtain more accurate
quantitative  data for the  methylene chloride extract  by  the induc-
tively  coupled  plasma optical emission spectroscopy  (ICPOES)  tech-
nique was frustrated by the lack of compatibility between this  tech-
nique and methylene  chloride.  However, quantitative  results were
obtained  for  both of the aqueous phase solutions (green  and white
liquor) by  ICPOES.   All results of  these  analyses are given in
Appendix  B, Tables B-l  and B-2.

G.  Data  Analysis and Interpretation

    The extent of sample degradation is computed as the ratio of the
stored  samples'  area response to that  of the freshly  prepared standards,
For example,  in  the  hypothetical situation  where a sample initially
containing  100 pg/ml of phenol  yielded an integrated  area of 500
microvolts-seconds and  a standard containing 105 ug/ml phenol exhibi-
an  area response of  515 microvolts-seconds, the  equation would be

-------
                 microvolts-seconds
                      sample
                 microvolts-seconds
                      standard
                                         100 yg/ml
     A value of 1.00, representing the expected concentration, is
subtracted from this value and the resulting decimal expressed as
a percentage, i.e.,

                       1.019 - 1.00  =  0.019 - 1.9%

This figure is the percentage deviation of the stored samples' final,
post handling, concentration from the initially compounded amount as
represented by the fresh standard solution.   As computed, positive
values imply that an increase in concentration has occurred within
the returned sample, while negative values indicate sample loss.
Values approximating zero infer no change.

     As day to day fluctuations are known to influence chromato-
graphic responses for duplicate sample injections, data analysis was
limited to experiments performed on the same day under similar con-
ditions.  The average response factors of bracketing standard solu-
tion injections was determined, and individual samples compared
directly with this value.  This procedure was useful for normalizing
the effects of instrumentation variability.

     Experimental results were analyzed according to a standard sta-
tistical technique known as the Analysis of Variance (ANOVA).  The
basic assumption implicit in utilizing this technique is that the
experimental observations (i.e., the deviations of the stored sample
from the standard) are random variables, with well-defined distribu-
tional properties, that can be expressed as an additive linear func-
tion of the parameters that have been controlled in the experiment.
If the appropriate assumptions are satisfied, it is then possible to
test hypotheses regarding the effect of the individual parameters,
or factors, that have been varied over two or more levels.  If there
is reason to believe that the necessary assumptions do not apply for
the data, alternative analytical methods could be utilized; e.g.,
transforming the data, using non-parametric test methods, and/or
performing an Analysis of Co-variance in which the standard calibra-
tion value is the co-variate and the observed peak area response the
dependent variate.  These techniques are described in detail in most
statistical methods texts (11).
                                10

-------
                    III.  EXPERIMENTAL RESULTS

     Results obtained  from this  three-set stability review are pre-
sented in Tables 3 through 5.  The tabulated values reflect percent-
age variation of the stored samples from the expected sample content,
based upon the initial  formulation criteria.  As such, values near
zero imply best handling practice implementation, while negative
values infer species loss and positive values imply an increase in
concentration or, more  likely, contamination.  Deviation on either
side of ideal is viewed as equally bad, as corrective practices
would need to be implemented in  either case.

     Generally, the variability  attributed to the analysis itself
has been found to be small.  Table 6 reflects the average peak area
response, standard deviation and coefficient of variation for each
compound within the standard preparation, as determined by repeti-
tive injections throughout the individual analysis sequences.  In
addition, the 95 percent confidence levels surrounding the calcula-
ted means are presented.

     As mentioned, the  samples of Set III (sorbent resin set) con-
sisted of the model compounds supported on XAD-2 solid sorbent resin;
a combination which is  unsuitable for direct GC analysis.  In order
to analyze these samples, it was necessary to extract them within a
continuous Soxhlet extraction apparatus.  The resulting solutions
were then adjusted to  the appropriate final volume and directly
analyzed by GC.

     Since the extraction process could possibly have interfered with
the fina.l results obtained, six  non-stored samples were prepared and
analyzed simultaneously with the stored samples to estimate the effect
of this step.  The results obtained are given in Table 7.

     Observed average sample recoveries from the extraction process
are found to range from 91.4 to  98.4 p'ercent.  However, as several
of the listed data groups exhibit large variability around the com-
puted average, no attempt has been made to adjust the results of
Set III samples to account for the extraction process effects.
Instead, tabulated values include all factors involved in sample
storage, handling, and  extraction.

     Generally the data show that most compounds are recovered within
±10% of their original  concentration for all three sets of conditions.
The results are reviewed and analyzed in detail in Section IV.

     Interpretation of  the results obtained during Sets II and III
analyses of n-Hexadecane concentration was difficult, due to the in-
ordinately high recoveries which can not be assigned to any single
parameter.   Within Set  II analyses, these variances, ranging from 10
to 65 percent above the formulated concentrations, were initially
felt to arise from contamination.  Observations made during analysis
                                   11

-------
10
                                                  TABLE 3'
                             Analytical Results from Stored Probe Wash Samples  (Set I)
                                     (Percentage Deviation from Expected)
Run
No.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Sample
Identification
© © (D © *
AG-AA-H-RD
-AA-R-RD
-AA-C-RD
AG-AA-H-RL'
-AA-R-RL
-AA-C-RL
AG-AA-H-CD
-AA-R-CD
-AA-C-CD
AG-AN-H-RD
-AN-R-RD
-AN-C-RD
AG-NN-H-RD
-NN-R-RD
-NN-C-RD
AG-AN-H-RL
-AN-R-RL
-AN-C-RL
AG-NN-H-RL
-NN-R-RL
-NN-C-RL
AG-AN-H-CD
-AN-R-CD
-AN-C-CD
AG-NN-H-CD
-NN-R-CD
-NN-C-GD

Phenol

14.3
4.9
0.0
-8.2
-0.5
-1.6
-6.0
-1.9
-2.3
9.6
-15.5
-3.2
-0.4
-2.1
-4.2
-8.2
-7.6
-13.7
0.0
3.2
2.6
-4.1
-17.6
-16.1
2.3
2.0
-26.3
                                               N-Methyl.
                                                .aniline
 24.1
  0.5
  6.4
 -0.3
  5.7
  3.6
  5.5
 13.7
  2.4
  4.0
 -0.7
  1.9
  5.2
  1.0
  0.8
 -0.8
 -0.5
 -3.8
  3.1
  7.6
  8.3
  0.7
 -8.8
-10.0
 -0.3
  4.9
  7.7
Acenaphthene

     3.6
    -1.6
     0-0
    -2.3
     3.8
     1.2
    -1.4
    -0.4
    -2.1
     2.9
    -6.0
    -2.7
     1.8
    -1.4
    -1.2
    -2.3
                                                                   -1
                                                                   -3
       9
     5.4
     2.7
     5.8
     5.8
      .2
     1.2
     0.8
     3.1
     2.8
     3.6
a-Hexadecane

    10.2
     5.3
     5.0
     0.4
     6.3
     3.7
     3.8
     5.5
     2.7
     8.0'
    -2.2
     1.4
     6.0
     3.4
     1.7
    -0.4
    -2
  9
7.7
5.3
8.6
9.8
1.8
5.6
5.6
7.7
8.1
8.3
           4,4'-Dichloro-
              biphenyl

                 3.6
                -6.3
                 1.5
                -8.5
                 7.2
                 3.6
                -2.7
                 1.0
                 0.6
                 2.8
                -6.7
                11.0
                 0»0
                -0.5
                -1.5
                -8.5
-7
                     -1
  6
8.7
6.7
8.2
9.2
 ,4
2.1
2.1
2.0
1.3
3.2
       * Denotes  factor  combination  observed during experiment (see Table 1)
                                                                                          continued....

-------
                                       Table 3   (continued)
                    Analytical Results from Stored Probe Wash Samples (Set I)
                               (Percent Deviation from Expected)
Run
No.

28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Sample
Identification
© (D(D ©*
P-AA-H-RD
-AA-R-KD
-AA-C-RD
P-AA-H-RL
-AA-R-RL
-AA-C-RL
P-AA-H-CD
-AA-R-CD
-AA-C-CD
P-AN-H-RD
-AN-R-RD
-AN-C-RD
P-NN-H-RD
-NN-R-RD
-NN-C-RD
P-AN-H-RL
-AN-R-RL
-AN-C-RL
P-NN-H-RL
-NN-R-RL
-NN-C-RL
P-AN-H-CD
-AN-R-CD
-AN-C-CD
P-NN-H-CD
-NN-R-CD
-NN-C-CD

Phenol

2.2
1.8
-17.6
4.7
9.8
-16.2
-14.7
31.6
2.9
7.6 .
17.2
0.4
1.7
15.5
-7.9
12.7
6.4
8.7
12.7
11.7
6. '8
-2.0
1.2
-8.3
-16 . 8
-15.5
-9.1
                                      N-Methyl-
                                       aniline

                                          6.6
                                          4.0
                                          6.1
                                          3.4
                                         -4.5
                                        -10.7
                                         11.4
                                         18.9
                                          7.4
                                         10.5
                                         27.8
                                          6.8
                                          6.4
                                         24.4
                                          5.1
                                          1.1
                                          2.5
                                         -4.5
                                         -2.6
                                         -4.2
                                         -5.6
                                          1.7
                                         16.2
                                          8.0
                                         -8.3
                                          2.0
                                          7.0
Acenaphthene


      4.7
      2.6
     -2.3
      5.7
      5.1
      4.3
      0.3
      3.2
      4.4
      6.7
      9.5
      2.5
      4.6
      7.7
      0.3
      4.0
      8.3
      4.0
      6.0
      4.0
      3.6
     -1.3
     -4.3
      3.6
      4.7
    -10.2
     -0.7
n-Hexadecane

    -4.2
    -5.0
    -7.2
    -2.7
     0.0
    -0.2
    -4.3
    -1.4
     2.5
    -1.2
     3.5
    -2.2
    -3.6
    -0.7
    -2.4
    -0.7
     3.7
     0.7
    -2.2
    -0.7
     1.5
    -6.3
    -9.2
    -0.5
    -2.5
   -12.5
    -2.5
4,4'-Dichloro-
   biphenyl

      6.1
      0.0
     -4.6
      1.7
     -1.6
      1.1
     -1.0
     -0.6
      5.6
      7.8
     11.4
      3.9
      1.6
      3.4
     -2.9
     -2.0
    -15.6
      7.5
      4.4
     -1.6
     10.2
     -2.0
     -5.6
      1.1
    -1.
   -15.
    -7.0
Denotes factor combination observed during experiment (see Table  1)

-------
                                          Table 4
     Analytical Results from Stored  Sorbent  Trap Condensate  Extract Samples  (Set II)
                          (Percent Deviation from Expected)


                       Phenol

                         0.8
                         7.6
                         8.4

                         2.9
                         3.0
                        -6.2

                        -0.8
                        -3.4
                        -2.3

                        -0.3
                         0.0
                         5.0

                         6.0
                        -6.8
                         0.0

                        -0.3
                         0.0
                        -6.3

                         1.9
                        -3.3
                         O.8

                         3.6
                         2.6
                        -1.3

                        -1.0
                         2.9
                        -3.1

+ Equivalent sample identification codes indicate separate preparations of duplicate samples.
  Presented in randomized order as performed in experiment.

* Denotes  factor combination  observed during experiment (see Table 1).
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Sample
Identifies ti
9-N?Sc?S
P-AN-C-RL-S
P-AN-C-RL-S
AG-AN-H-RD-N
P-AA-H-CD-S
AG-AA-R-RL-N
AG-NN-R.-RD-S
AG-AA-H-CD-S
AG-NN-H-RL-S
P-AA-C-RD-S
P-AN-R-CD-S
P-NN-H-RL-S
P-NN-H-RL-S
AG-AA-C-RD-S
AG-NN-C-CD-N
P-NN-H-RD- S
P-AN-H-RU-N
AG-AN-R-CD-S
P-NN-R-RD-S
P-NN-C-CD-N
P-AN-R-CD-S
P-AA-C-RD-S
P-AA-H-CD-S
P-AA-R-RL-N
P-AA-R-RL-N
P-AN-H-RD-N
AG-AN-C-RL-S
H-Methyl-
aniline
2.6
-6.0
-16.1
21.6
12.8
7.1
15.3
12.5 '
15.2
8.6
10.2
-10.9
-12.2
6.8'
4.5
12.9
15.2
6.9
13.5
10.6
5.4
8.5
12.5
-2.3
3.1
17.5
9.1
4-Chloro-
benzaldehvde
3.0
7.1
8.9
0.7
9.0
-0.7
.10.1
2,1
3.1
5.2
5.2
13.0
« 2.5
0.0
0.7
6.0
«.8
1.5
6.0
2.3
5.4
1.4
7.6
3.3
-1.5
W.3
5.3
Acenaph-
thene
-0.7
2.3
7.4
2.5
0.7
-4.4
3.6
-1.6
0.9
0.0
0.3
3.4
0.7
-8.2
-8.0 *
0.3
0.3
-6.7
1.3
-4.4
-2.0
-0.7
0.7
-3.3
-4.7
3.8
-4.0
n-Hexa-
decane
0.9
1.9
35.6
0.9
53.7
19.6
25.4
22.8
23.3
42.0
43.9
*5.2
64.2
14.9
.14.2
18.6
18.9
16.6
24.2
14.1
42.8
46.3
20.1
15.2
14.4
21.5
20.3
4,4'-Dichloro-
biphenvl
2.9
3.9
11.6
-3.4
•2.4
4.2
4.2
12.0
-5.0
-5.8
7.4
10.5
-0.5
-1.0
0.0
8.3
13.2
8.7
2.9
4.9
4.8
5.8
11.1
7.3
-0.5
•6.8
-1.9

-------
                                               TABLE 5
                  Analytical Results  from Stored  Sorbent Resin Samples (Set III)
                                  (Percent deviation from expected)
Run
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Sample
Identification"*"
 © vD ©*
AG-NN-R-RD
AG-AA-C-CD
AG-AA-H-CD
AG-NN-H-RD
AG-NN-C-CD
AG-AA-R-CD
AG-NN-R-CD
AG-NN-H-CD
AG-AA-R-RD
AG-NN-C-RD
AG--AA-H-RD
AG-AA-C-RD
AG-AA-H-CD
AG-NN-R-CD
AG-NN-C-RD
AG-AA-C-CD
AG-AA-R-RD
AG-NN-H-RD
Phenol
0.0
3.1
-0.5
1.1
4.3
-9.9
1.8
1.8
0.0
6.0
-0.6
1.2
0.6
3.5
-4.1
-2.9
3.9
4.9
N-Methyl-
aniline
-3.4
1.3
2.9
-9.0
2.9
-2.1
8.1
1.3
1.3
2.7
5.0
6.8
3.6
-6.0
1.6
6.3
2.8
6.3
4-Chloro
benzaldehyde
-4.0
-2.1
-1.2
-6.1
-4.0
-6.1
-0.7
-3.0
-1.5
-3.0
-0.8
-0.8
-3.1
-0.8
-4.9
-4.2
-8.8
-11.2
Acenaph-
thene
-2.4
-1.0
0.0
-6.9
-3.7
-5.6
-1.2
0.8
0.8
-1.3
0.0
-0.5
-3.2
-1.1
-2.2
-0.9
-0.7
-1.4
n-Hexa-
decane
5.2
24.2
19.7
11.7
15.1
15.3
15.5
19.0
18.6
17.1
18.4
18,8
15.9
18.4
17.0
17.9
19.7
17.5
4,4'-Dichloro'
biphenyl
-1.4
-0.3
1.2
-8.2
-5,1
1.0
0,3
-1.4
-2,1
3.1
-1.1
-2.2
-4.9
-2.6
-3,7
-1.3
-1.0
-2.0
+ Equivalent sample identification codes indicate separate preparation of duplicate samples.
* Denotes factor combination observed during experiment (see Table 1).

-------
                                           Table 6




                    Statistical  Susmary of Standard  Calibration  Samples

Set I
Probe Wash
Series

50 Observations
Set II
Condensate
Extract
Series
-
12 Observations
Set III
Sorbent Resin
Series

16 Observations

Mean
st'd dev.
coef. of var.
95*
confidence limits
Mean
st'd dev.
coef. of var.

95*
confidence limits
Mean
st'd dev.
coef. of var.
95*
confidence Units
Phenol
274
36.9
13. 5Z

263.5 - 284.5
376
9.4
• * 2.5*


370.1 - 381.9
15.6
1.1
7.1*

15.0 - 16.2
N-Methyl-
anillne
356
27.3
7.7*

348.2 - 363.8
366
8.3
2.3*
*

360.8 - 371.2
18.5
1.4
7.6*

17.8 - 19.2
4-Chloro -
benzeldehyde

not
analyzed


261
9.8
3.8*
1


254.8 - 267.2
16.0
0.9
5.6*

15.5 - 16.5
Acenaphthene
265
9.7
3.7*

262.2 - 267.8
295
10.9
3.7*


288.2 - 301.8
42.0
2.6
6.2*

40.6 - 43.4
n-Hexadecane
504
20.1
4.6*

498.3 -509.7
456
15.1
.3.3*


446.5 - 465.5
21.9
1.7
7.82

21.0 - 22.8
4,4'-Dlcbloro-
biphenyl
200
21.8
10.9*

193.8 - 206.2
236
19.4
8.2*
-

223.8 - 248.2
29.6
1.8
6.1*

28.6 - 30.6
This species-was not present in Set I samples.

-------
                  TABLE 7
    Results of Extraction Step Experiment
(Percentage Deviations from Unspiked Standard)
Standard
Spikes
Run No. 1
2
: 3
4
5
6
Average
95% Confi-
dence Interya]
Limits
Phenol
-5.6
-1.9
-8.6
-14.1 '-
-9.0
-3.0
-7.03
-2.3
to
-11.7
N-Methyl
Aniline
-0.5
+3.4
+3.9
-12.5
-3.7
-5.3
-2.45
-8.9
to
+4.0
4-Chloro-
Benzaldehyde
-8.5
-3.6
-4.2
-7.8
-13.7
-13.7
-8.58
-4.0
to
-13.6
Acen aph thene
-1.4
-2.0
-2.3
-2.1
-6.4
-0.7
-2.48
-0.4
to
-4.6
n-Hexa-
decane
-1.4
-1.3
-1.3
-2.3
-2.3
-0.9
-1.58
-1.0
to
-2.2
4,4'-Dichloro-
biphenyl
-1.7
-4.1
-4.5
-2.0
-1.3
-0.7
-2.38
-0.7
to
-4.0

-------
 tended to support this theory, as a shoulder was noted on the chro-
 ma togram part way through the hexadecane peak.   In addition,  where
 a relatively smooth baseline had once existed,  several smaller but
 reproducible peaks were in evidence.   To determine the contaminants'
 composition, two representative samples of the  condensate extract
 set were studied by capillary gas chromatography-mass spectrometry
 (GC-MS).   The chromatograms obtained from these analyses are  presented
 in Figures 1 and 2.

      Referring to Figure 1, it is clear that a  homologous series of
 olefins is present within the retained sample (P-NN-H-RL-S),  and
 that the species identified as a CIG  olefin is  the probable cause of
 much of the observed extra hexadecane content (65%).   As olefins are
 major constituents of both mold release compounds and polyethylene
 bottles,  it is likely that the contaminant series results from the
 extraction of container walls by the  methylene  chloride.


      However, the excessive n-hexadecane content observed in most
of the samples of Set II and Set III cannot be singly attributed to
the extraction of contaminants from the polyethylene sample bottles.
With reference to Figure 2, a chromatogram of a second sample
(P-NN-R-RD-S) from Set II, where a 25 percent excess of n-hexadecane
had been reported, indicates that even though the same homologous
series of contaminants is present, their apparent contribution to the
reported excess is not large (even on a proportional basis).   Further-
more, reference to the results of Set I analyses, indicates that not
all samples stored in polyethylene bottles of the same lot and under
similar conditions will become contaminated.

      Comparably, review of a typical sample from the serbent  resin
set (see Figure 3) shows that no olefin species  are contained  within
a sample exhibiting a net gain of 24 percent n-hexadecane.  As all
the tested resin combinations in this set, and many samples in Set II
as well, were stored in amber glass bottles, the extraction of olefins
from the container walls is not considered likely.

      Thus, the origin and nature of all factors influencing the
n-hexadecane concentration within Set II and Set III samples is not
completely clear.  While it is obvious that some of the samples have
become contaminated with components extracted from the walls of the
polyethylene sample bottles, this alone cannot account for all of the
encounted hexadecane excesses.  Other factors, including possibly in-
accurate sample formulation, must be considered as having influenced
some of the reported results, making the interpretation of Set II and
III ji-hexadecane results impossible.
                              18

-------
VO
300
                                                                    350
    •"""I
400      450
                             Figure 1.  GC/MS Chromatogram of Set II Sample P-NN-H-RL-S

-------
100
                                       ii

                                       i
                           •s
                           *
                       I
                M-l
                a)
                H
                O

                to
                r-4
                O
                                                              M
                                     s
                                      el
           50
100
200
250
300
350
400
450
                          Figure 2.  GC/MS Chromatogram of Set II Sample P-NN-R-RD-S

-------
00 |
r-
•r
9
iH
£
1






o
0
01
£
50 100
I
f

4) V
X <
1
s
.0
O
l-c
O
6
^
1'


150 200 250 300
i
a
n
a
»1
B










i T"j ij i i i i . i i
350

-------
                        IV.  DISCUSSION OF RESULTS
A.   Statistical Analysis of Experimental Data
      The changes in composition (percent deviation) of each compound,
 given in Tables 3-5, are observed to vary over the set of experi-
 mental input conditions considered.   The statistical technique ANOVA
 provided a mechanism to allocate this variability in the data to
 identifiable sources; namely, those  conditions which were controlled
 in the design of the experiment (i.e., type of container, head space
 composition, shipment temperature, and storage condition).  Further-
 more, the ANOVA technique provides an estimate of the normal or
 chance variation in the data that has been an inherent part of
 performing the experiment.   By comparing the magnitude of the dif-
 ference attributable to varying each factor over the levels tested
 to the normal variation in the data, those factors having a measur-
 able (i.e.,  statistically significant) effect can be identified.


     To illustrate this  concept, an ANOVA table is  given in Table 8
 for  the set  of 54 phenol observations  reported in  Table 3.  Although
 the  basic experiment has not been replicated, it is possible  to
 estimate chance variation (i.e., experimental error) by assuming
 all  three and four  factor interactions are negligible.  For the
 phenol data, the experimental error variance is estimated  to be
 82.81: thus  the standard error is V82.81 = i 9.1  percent  on  a per
 observation  basis.

     The ANOVA table indicates those effects and two-factor inter-
 actions that appear to  contribute significantly to the variability
 of the data.  To illustrate this concept, the average values of
 each level of each  factor are given in Table 9.  For example, the
•significant  contribution attributable  to varying types of  container
 on the phenol data  can be explained by the fact that aniber glass, on
 the  average, yielded a  loss of 3.7 percent, while  the 27 experiments
 using polyethylene yielded a positive  deviation of 1.8 percent phenol.
 This  difference of 5.5 percent is statistically significant, since
 the  ANOVA table provides the data to calculate a 95 percent confi-
 dence interval about this difference:
    Estimated "container effect" = (-3.7 - 1.8) ± (2.048)\

              -  -5.5% ± 5.1%

Therefore, this experiment indicates the likelihood that the use of
polyethylene results in phenol measurements of 0.4 percent to 10.6
percent higher than would be observed using amber glass, a positive
(non-zero) contribution.
                                 22

-------
                               TABLE 8..
                            Analysis of Variance
                          Probe Wash Experiment (Phenol)
Source of Degrees of
Variation Freedom
Main Effects
Container 1
Head Space Composition 2
Shipment Temperature 2
Storage Condition 2
Interactions
Container x Head Space
Composition 2
Container x Shipment
Temperature 2
Container x Storage
Condition 2
Head Space Composition x
Shipment Temperature 4
Head Space Composition x
Storage Condition 4
Shipment Temperature x
Storage Condition 4 i
Error 28
Total 53
Sum of
Squares
406.3
38.6
675.2
574.3
426.4
363.5
159.6
327.9
689.1
175.7
2318.8
6155.5
Mean
Square
406.3
19.3
337.6
287.1
213.2
181.7
79.8
82.0
172.3
43.9
82.8
F
4.91*
< 1
4.08*
*
3.47
2.57
2.19
< 1
< 1
2.08
< 1

Standard Error per sample =  \82.81   = '  +  9.

* Indicates significance at 5% test level
                                   23

-------
                                                    TABLE 9
                                         Probe Wash Experiment  (Set I)
                               (Average Percentage Deviation from Expected)
Factor
Container
Head Space
Composition
Shipment
Temperature
Storage Con-
dition (Temper-t.
ature/Light)
Level
Amber Glass
Polyethylene
Air /Air
Air/Nitrogen
Nitrogen/Nitro-
gen
Hot
Boom
Cold
Room/ Dark
Room/light
Cold/ Dark
Average
Standard Error Per Sample
No. of Ob-
servations
27
27
18
18
18
18
18
18
18
18
18
54
Significant Interactions:
Phenol
-3.7 j
f *
1.8 \
0.2
-1.8
-1.3
0.4 x
- 2.5>*
-5.8^
1.3,
1.3 >*
-5.6'
-1.0
9.1
None
N-Methyl
aniline
3.0
5.1
5.8
2.9
3.5
4.0
6.1
2.0
7.8i
-O.lj>**
4.4'
4.1
6.3
Cont.x Hd.)
Sp. Comp. -7
Cont. x K
5to.r.Cond.J
Acenaph-
thene
°-7},
3.0 )
1.6
1.5
2.4
2.2
1.6
1.7
1.8x
3.5 > *
0.2^
1.8
3.4
None
n-41exa-
decane
4-7L*
-2.2\
1.1
0.7
1.8
0.8
0.9
2.0
0.9
2.1
0.7
1.2
3.5
None
4,4'-
Dichloro-
biphenyl
1.2
0.1
0.4
0.5
1.1
0.5\
-1.5 i*
3.0'
1.7
1.3
-1.0
0.6
4.9
Hd. Sp.
Comp.x :i
Stor.
Cond.
to
        * Indicates significance at 5% test level.
       ** Indicates significance at 1% test level.

-------
    Similarly, all other significant differences are identified in
Table 9.  The Set I experiment was structured in such a way that
it was possible to investigate not only the influence of each factor
independently, but also the interactive or combined effect of all
factor pairs.  Therefore, significant two-factor interactions are
also given in Table 9.  The nature and interpretation of the two-
factor interactions observed to be significant are demonstrated
by the two-way tables and graphs given in Table 10 and Figure 4.
For example, it is observed that n-methylaniline readings were ap-
proximately 2 percentage points higher with glass than with poly-
ethylene when tested with air/air or nitrogen/nitrogen head space
compositions.  However, this situation was very different when air/
nitrogen was used, in that glass yielded nearly 10 percent lower
readings than polyethylene.

    The condensate extract (Set II) experimental design was dif-
ferent from Set I in that all possible factor/level combinations
were not tested.  It had been possible as a result of analzying the
data from Set I to eliminate some of the small effect combinations,
The plan used, known as a Fractional Factorial design, required only
twenty-seven unique test conditions in order to estimate each factor
independently.  Furthermore, only the two-factor interactions involv-
ing the container effect were estimable.  This analysis required the
assumption that all other two-factor and higher order interactions
had a negligible effect on the data.

    An illustrative example of the components of variation are given
for phenol data in Table 11.  It is observed that only the container
effect is statistically significant; i.e., an increase between 1.4
and 7.4 percent can be expected in phenol measurements when poly-
ethylene is used instead of amber glass.

    Table 12 summarizes the results of all ANOVA's computed for Set
II.  Significant interactions are illustrated in tabular (Table 13)
and graphic form (Figure 5) as described for Set I data.  Similar
analytical results are given for the Sorbent Resin (Set III) experi-
ment in Tables 14 and 15.

B.  Interpretation of Results

    The overall purpose of this set of experiments has been to detect
measureable differences in the recoveries of select model compounds
that are directly attributable to the variation of specific shipment
and storage procedures.  Subsequent to detection, the extent of dif-
ference found would be directly compared with the accuracy require-
ments of a Level 1 survey analysis to determine if, in fact, the
measured variability could interfere with the desired goals of such
a survey.

    As such, the laboratory experiments were designed and conducted
in such a way that differences were likely to be detected if, indeed,

                                  25

-------
                           TABLE  10
                Probe Wash Experiment (Set I)
            Significant Two-factor Interactions
COMPOUND
  N-Methylaniline:
      Container
                 Glass
                 Poly
                         Head Space Composition
                            AA
               AN
NN
6.8
4.7
-2.0
7.8
4.3
2.7
      Container
                 Glass
                 Poly
                         Storage Condition
                                    RL
                       CD
4.8
10.9
2.5
-2.8
1.8
7.1
  4,4*-Dichlorobiphenyl:
                         Storage Condition
       Head Space
       Composition
AA

AN

NN
                            RD
               RL
CD
0-0
5.0
0 0
0.6
-2.9
6.2
0.5
-0.6
-3.0
  * Indicates significance at 5% test level.
                              26

-------
            Compound:   N-Methylaniline
            Container X Head Space Composition
          12
       .1   8

-------
                                TABLE  11
                          Analysis of Variance
                 Condensate Extract Experiment  (Phenol)
Source of Degrees of
Variation Freedom
Main Effects
Container
Head Space,
Composition
Shipment
Temperature
Storage
Conditions
Catalytic
Species
Interactions
Container x Head
Space Composition
Container x Shipment
Temperature
Container x Storage
Condition
Container x Cata-
lytic Species
Error
Total
1
2
2
2
1

2
2
2

1
11
26
Sum of
Squares
113.8
31.0
29.6
5.9
4.7

34.0
12.9
12.6

38.3
122.1
405.0
Mean
Square
113.8
15.5
14.8
2.9
4.7

17,0
6.4
6.3

38.3
11.1
F
**
10.16
1.40
1,33
<1
<1

1.53
<1
<1

3.45

   Standard Error per sample = -y/11.10  =  3.33%

** Indicates significance at the 1% test level.
                                     28

-------
                                                      TABLE   12
                                    Condensate Extract  Experiment  (Set II)
                                 (Average Percentage Deviation from Expected)


Factor
Container

Head Space
Composition

Shipment
Temperature
Storage Condi-
tion (Tempera-
ture/Light)
Catalytic
Species


Level
Amber Glass
Polyethylene
Air/Air
Air/Nitrogen
Nitrogen/Nitro-
gen
Hot
Room
Cold
Room/Dark
Room/ Light
Cold /Dark
Stainless Steel
None
Average
Standard Error Per Sample
No. of
Observa-
vations
9
18
9
9
9
9
9
9
9
9
9
18
9

27
Significant Interaction




Phenol
-2.9)
1.5J
-1.1
1.5
-0.3
0.7
-1.5
0.8
0.3
0.3
-0.6
0.3
-0.6
0.01
3.33
None



N-Methyl-
aniline
11.0)
4.8 j *
7.7
7.1
5.7
9.4 )
8.0} *
3.2)
13.3 j
- -1.4 /**
8.7)
5.8
8.9
6.8
3.78
Cont. x|
Stor. ?**
Cond. /
4-Chloro-
benzalde-
hyde
2.5\ A
5.9 j
2.9
6.1
5.2
6.6
3.9
3.8
5.6
4.6
4.1
5.5
3.2
4.8
3.62
None



Acenaph-
thene
-2.9)
0.3)
-2.4
0.4
-0.3
1.3)
-1.7 >*
-1.8J
0.3
-0.2
-2.4
-0.1
-2.1
-0.8
2.49
Cont. x)
Ship. /*
Temp . /

n.-Hexa-
decane
17.6
30.2
27.7
22.5
27.8
32.3
24.5
21.1
23.6
28.9
25.5
32. 3^
13. 3/
26.0
16.9
None


4,4'-
Dichloro-
biphenyl
2.0
5.4
3.9
5.7
3.1
5.2
5.3
2.3
3.4
3.3
6.0
4.4
3.9
4.3
5.4
None


VO
           * Indicates significance at 5% test level.
          ** Indicates significance at 1% test level.

-------
                              TABLE 13


               Condensate Extract Experiment (Set II)
                Significant Two-factor Interactions"*"
  COMPOUND
    N-Methylaniline:
                             Storage Condition
      Container
Glass

Poly
                              RD
                   RL
CD
14 = 6
12.7
10.5
-7.4
8.0
9.0
                                                  **
    Acenaphthene:
                             Shipment Temperature
      Container
                   Glass
                   Poly
                              Hot
                   Room   Cold
0.6
1.6
-2.5
-1.3
-6.7
0.6
  Only those interactions involving the factor "Container" are
  estimable from this design.
^Indicates significance at 5% test level.
  Indicates significance at 1% test level.
                                  30

-------
o
CD

Q
       Compound:   N—Methylaniline


                  Container X Storage Condition
    15
    10
   -10
             RD         RL


             O - Amber Glass


             A— Polyethylene
CD
       Compound:   Acenaphthene

                   Container X Shipment Temperature
 c
 tg
 '
 -
 <0    9

 '>    *•
 tt)

 O
    -6
             Hot        Room    Cold


             O— Amber Glass


             ^— Polyethylene
FIGURE  5.  CONDENSATE EXTRACT EQUIPMENT (SET II)


            SIGNIFICANT TWO-FACTOR INTERACTIONS
                       31

-------
                            TABLE  14

                       Analysis of Variance

                Sorbent Resin Experiment  (Phenol)

     Source of        Degrees of   Sum of     Mean
     Variation          Freedom   Squares    Square      F

Main Effects

    Head Space              1        33.1      33.1       1.64  (non-sig.)

    Shipment Temperature    2        7.4       3.7     <1

    Storage Condition       1        6.2       6.2     <1

Interactions

    Head Space Composition x
      Shipment Temperature  2        3.5       1.8     <1

    Head Space Composition x
      Storage Condition     1        17.9      17.9     <1

    Shipment Temperature
      Storage Condition     2        8.2       4.1     <1

    Error                   8      161.4      20.2
    Total                  17      237.7
                               i^^^^^^^^
Standard Error per sample  =  \20.2  =  +4.5%
                                  32

-------
                                                    TABLE  15.
                                         Sorbent Resin Experiment  (Set III)
                                  (Average Percentage Deviation from Expected)

Factor

Bead Space
Composition
Shipment
Temperature

Storage Condi-
tion
(Temp/Light)

Level
Air /Air
Nitrogen/Nitro-
gen
Hot
Room
Cold
Room/Dark
Cold /Dark
Average
Standard Error Per Sample
No. of
Observa-
tions
9
9
6
6
6
9
9
18



Significant Interactions

Phenol
-0.6
2.1
1.2
-0.1
1.3
1.4
0.2
0.8
4.5


None

N-Methyl-
aniline
3.1
0.5
1.7
0.1
3.6
1.6
2.0
1.8
5.6


None
4-Ctjloro-
benzalde-
hyde
-3.2
-4.2
-4.2
-3.6
-3.2
-4.6
-2.8
-3.7
2.1
Hd. Sp.
Comp . x ^
Ship.
, Temp. ,

Acenaph-
thene
-1.2
-2.2
-1.8
-1.7
-1.6
-1.6
-1.8
-1.7
2.2


None

n-Hexa-
decane
18.7
15.2
17.0
15.4
18.4
16.0
17.9
16.9
4.3


None
4,4'-
Dichloro-
biphenyl
-1.2
-2.3
-2.7
-1.0
-1.6
-2.1
-1.5
-1.8
3.3


None
oo
           *  Indicates  significance at the 5% test level.

-------
                  Compound: 4—Chlorobenzaldehyde
                                    Shipment Temperature
                                   Hot      Room      Cold
Head Space
Composition
Air
Nitrogen
-1.7
-6.8
-5.5
-1.8
-2.4
-4.0
                  Head Space Composition X Shipment Temperature
             o
             +3   o
             .s

             Q
             35
                 ^



                -6


                -8
                         Hot
Room
Cold
                    O -Air/Air
                    A -Nitrogen/Nitrogen
Indicates significance  at the 5%  test level.
              FIGURE  6.  SORBENT RESIN EXPERIMENT (SET III)
                          SIGNIFICANT TWO-FACTOR INTERACTION
                                34

-------
they did exist  (that is, the design was "powerful" in the statistical
sense).  Furthermore, it should be recognized that these experiments
were intended to yield estimates of the magnitude of the true, but
unknown, average effect caused by varying a parameter over two or
more levels, and that these estimates were to be measured with
reasonable analytical precision.  Thus, as an example, it was found
in Set I experiments that a sample stored in polyethylene can be
expected to yield a larger phenol concentration than comparable
samples stored within amber glass containers.  Furthermore, the con-
centration of phenol in polyethylene can be expected to be from 0.4
to 10.6 percent higher on the average.

    Although this suggests that such a difference is "real" in the
sense that repetitions of this experiment would yield similar find-
ings, the added contribution of even 10 percent is considered to be
well within normally accepted analytical limits for Level 1 analyses.

    With this in mind, it becomes clear that none of the tested com-
binations produce differences large enough to impact upon the results
of a Level 1 survey analysis.  Thus, any reasonable combination of
shipment temperature, storage condition, container, head space com-
position and catalytic species content can be used for a short
period of time.  Furthermore, inasmuch as a ±10 percent range is
normally acceptable for quantitative chemical analysis of samples,
it appears that most of these same conditions can be directly ap-
plied to quantitative analysis routines.  However, the tabulated
data also suggest that some caution should be exercised before these
conditions are unilaterally applied.

    One occasion implying caution, occurs during the analysis of the
n-Hexadecane results in Set II experiments.  As was previously men-
tioned, a homologous series of olefins apparently was extracted
from several, but not all, of the polyethylene containers used
during this experimental sequence.  The Cig member of this series
contributed directly to the observed excessive n-hexadecane concen-
trations, as it was integrated as part of the model compound GC
response, causing an average increase of 26 percent to be found
within the stored samples.  However, the distribution of the olefin
series was not uniform throughout all similarly handled  samples
implying that some of this observed variability was produced by other
factors which can not be identified by the experimental conditions
used.  As the origin of all factors contributing to the excessive
hexadecane content are unclear, judgment of the handling condition
effect upon species similar to n-hexadecane must be reserved.

    What is important from this experiment, however, is the fact
that it emphasizes that polyethylene containers should not be used
for the storage of samples in organic solvents.  While other results
of this study imply that plastic containers do not affect the model
compounds recoveries (due to loss or concentration), this sequence
of experiments shows that undesirable contaminants can arise during
confinement within polyethylene and interfere with eventual analysis.


                                  35

-------
As there is no way to insure that spurious contaminants will not be
extracted from the polyethylene, only those materials unlikely to
react with the container should be retained in them.

    A review of N-Methylamiline data from both Set I and Set II also
is of interest.  As listed, only those samples stored in diffuse sun-
light and at room temperature exhibit a net decrease in the N-Methyl-
aniline content.  From Tables 9 and 12 it is noted that decreases of
0.1 and 1.4 percent are encountered while other tested storage condi-
tions (room or cold temperature and in the dark) produce increases
ranging from 4.0 to 14.0 percent of the intially formulated concen-
trations.  The noted decreases are even more pronounced when a review
of the storage condition x container composition interaction is made.
As seen from Tables 10 and 13, those samples stored in polyethylene
and exposed to diffuse sunlight lose between 2.5 and 7.5 percent
of the initially compounded N-Methylaniline content, while all other
tested conditions (including both polyethylene and amber glass con-
tainers stored in the dark, and amber glass stored in diffuse sun-
light) show increased levels ranging from approximately 1.5 to 15.0
percent.

    As N-Mathylaniline is known to be photoreactive, it is presumed
that the noted species loss is attributable to some amount of photo-
oxidation.  While the amber glass bottle appears to have adequately
screened out the reactive ultraviolet light, the polyethylene did
not.   This is expected because polyethylene is transparent in the
ultraviolet region and will allow some light to penetrate and degrade
the aniline.  The amber glass, on the other hand, absorbs the ultra-
violet light prohibiting photodegradation of the amine.  Once again
this implies that the utilization of polyethylene should be restric-
ted to those instances where it is known that stored samples will
not degrade solid samples.
                              36

-------
                  V.  CONCLUSIONS AND RECOMMENDATIONS

    Excluding the storage of certain materials within polyethylene
containers or in the presence of diffuse sunlight, any reasonable
combination of handling procedures appears to be suitable for the
short term storage of Level 1 survey analysis samples.  As Level 1
procedures only require the identification of sample constituents
within a range of ±2 or 3 times of what actually exists, the rela-
tively small perturbations encountered during this experimental
review will not detract from the value of the accumulated data.
In fact, based on these results, it appears that several of the
tested conditions, including shipment temperatures (below 38°C) ,
storage conditions (storage at either room or reduced temperature
and in the dark) head space compositions (air or nitrogen) and
amber glass containers, are appropriate for the storage of quanti-
tative samples where variation must be held to ilO percent.
    The utilization of polyethylene sample bottles for storage of
organic materials is not appropriate, as it has been demonstrated
during Set II experiments that spurious contaminants can be extrac-
ted even from rigorously precleaned containers.  Furthermore, poly-
ethylene is not suitable for the storage of photo-reactive sub-
stances.  As noted from Sets I and II, N -Methylaniline results,
samples stored for short periods of time in sunlight (even in-
direct) can be lost through photo-degradation.  "While amber glass
bottles absorb the harmful ultraviolet wavelengths, polyethylene,
which is transparent, allows the light to penetrate.  Therefore, it
is recommended that these containers only be used for the storage
of samples that are known not to degrade.

    Furthermore, it is recommended that attempts be made to minimize
the contact of organics species with catalytic agents.  Although the
presence of acid extractable stainless steel constituents has not
been proven to be significant during this analysis , the effect of
these and other catalytic species has not been fully explored.
                               37

-------
                             REFERENCES
 1.  Bruce, H. E. and Cram, S. P. "Sampling of Marine Organisms and Sedi-
     ments for High Precision Gas Chromatographic Analysis of Aromatic
     Hydrocarbons" in Marine Pollution Monitoring, Procedings of Workshop
     held at NBS, Gaithersburg, MD, NBS Special Publication 409, pp. 181-
     182, 1974.

 2.  Federal Working Group on Pest Management," Guidelines on Sampling and
     Statistical Methodologies for Ambient Pest Monitoring" Washington,
     D.C. (1974).

 3.  Lawrence, W. H., Mitchell, J. L., Guess, W. L. and Avtian, J.,
    "Toxicity of Plastics Used in Medical Practice" Journal of Pharma-
     ceutical Sciences, 52_, 958-963, 1963.

 4.  Goerlitz, D. F. and Brewn, E. "Methods for the Analysis of Organic
     Substances in Water," Techniques of Water Resources Investigations,
     USGS Cat. #19:15/5, BK 5, Chapter A3 (1972).

 5.  Giam, C. S. amd Chan, H. S., "Control of Blanks in the Analysis of
     Phthalates in Air and Ocean Biota Samples," Accuracy in Trace Anal-
     ysis:  Sampling, Sample Handling, Analysis, NBS Special Publication
     No. 422, Vol. 2, pp. 701-708, 1976,

 6.  Hamersma, J. W., Reynolds, S. L., Maddalone, R. F., IERL-RTP Proced-
     ures Manual:  Level 1 Environmental Assessment, EPA Publication,
     EPA-600/2-76-160a, 1976.

 7.  Hertz, H. S., et. al., "Methods for Trace Organic Analysis in Sedi-
     ments and Marine Organisms," Marine Pollution Monitoring, NBS Special
     Publication 409, pp. 197-199, 1974.

 8.  Hamilton, P. B. and Myoda, T. T., "Contamination of Distilled Water,
     HCl, and NH.OH with Amino Acids, Proteins and Bacteria," Clinical
     Chemistry 20, pp. 687-691, (1974).

 9.  Straughan, D.,"Field Sampling Methods and Techniques for Marine
     Organisms and Sediments." Marine Pollution Monitoring, NBS Special
     Publication 409, pp. 183-187, 1974.

10.  Bristol, D. ''Effects of Storage Conditions on Residues of 2,4-D and
     2,4-DCP on Potatoes," Accnracy in Trace Analysis:  Sampling, Sample
     Handling, Analysis, NBS Special Publication, Vol. 2, pp. 737-746, 1976.

11.  Snedecon,  G.W.,  and Cochran,  W.G.,  Statistical Methods,
     Iowa State University Press,  Ames,  Iowa,  1967.
                                     38

-------
References (continued)
12.  Addelman, S., and Kempthorne, 0., Orthogenal Main Effect Plan,
     Aeronautical Research Laboratory Report No. 79, Nov. 1961.

13.  Private Communication with Dr. Robert Statnick of IERL
     Laboratory, Research Triangle Park,  North Carolina.
                                39

-------
APPENDICES
      40

-------
APPENDIX A - Formulation Procedures for Samples
TABLE Al   -  Formulation Procedure for Probe Wash
              Samples (Set I)

TABLE A2   -  Formulation Procedure for Sorbent Trap
              Condensate Extract Samples  (Set II)

TABLE A3   -  Formulation Procedure for Sorbent Resin
              Samples (Set III)
                              41

-------
                           TABLE Al

      Formulation Procedure for Probe Wash Samples (Set I)

Amber Glass Samples - 27 samples of 300 mis each

   9 liters total volume of a 1:1 mix of methylene chloride:methanol

                                       cone Mg/ml
                 Phenol                    168
                 N-Methylaniline           167
                 Acenaph th ene               9 5
                 n-Hexadecane              166
                 4-4'-Dichlorobiphenyl     103


Polyethylene Sample - 27 samples of 300 mis each.

    9 liters total volume of a 1:1 mix of methylene chloride:methanol

                                       cone yg/ml

                 Phenol                    155
                 N-Methylaniline           162
                 Ac enaph then e              10 3
                 n-Hexadecane              164
                 4,4'-Mchlorobiphenyl      94
                                   42

-------
                             TABLE A2

              Formulation Procedure for Sorbent Trap
                Condensate Extract Samples (Set II)

White liquor extract - unspiked - 9 samples of 300 mis each.

     White liquot - 3 liters of an 0.05 M HC1 and 0.005 M H2SOit
     in distilled water prepared (metallic content analyzed by
     ICPOES) and then extracted with an equal volume of tnethylene
     chloride.  Methylene chloride extract retained as white
     liquor extract.  To this, the following concentrations of
     model compounds was added.
               Phenol
               N-Methylaniline
               4-Chlorobenzaldehyde
               Acenaphthene
               n-Hexadecane
               4,4'-Dichlorobiphenyl
Green liquor extract - spiked - 18 samples of 300 mis each.

     Green liquor - A piece of 316 stainless steel placed  in 100 mis
     of 1.0 M HC1 and stirred for 3 hours, producing green acidic
     solution.

     A 6-liter volume of 0.05 M HC1, 0.005 M H2S04 and distilled water
     was prepared using the concentrate as a base (metallic content
     analyzed by ICPOES).  The total volume was then extracted with
     an equal amount of methylene chloride.  Methylene chloride (green
     liquor extract) retained and metallic content analyzed.  (See Ap-
     pendix B).  Aqueous phase discarded.

                                             ug/ml

               Phenol                         165
               N-Methylaniline.                169
               4-Chlorobenzaldehyde           167
               Acenaphthene                   101
               n-Hexadecane                   168
               4,4'-Dichlorobiphenyl          101
                                  43

-------
                              TABLE A3

        Formulation Procedure for Sorbent Resin Samples (Set III)


Amber Glass bottles - 24 samples

     Stock solution - seven model compounds added to 200 mis pentane
     to produce following concentrations:
                Phenol
                N-Methylaniline
                4,-Chlorobenzaldehyde
                Acenaph thene
                n-Hexadecane
                4,4'-Dichlorobiphenyl
      Five mis of stock solution pipetted onto 15^ 1 gm of XAD-2
      resin contained in amber glass bottles.   After handling period,
      samples were removed and extracted (Soxhlet) with methylene
      chloride for 24 hours.   Extract volume brought to 250 mis and
      analyzed.
                               44

-------
APPENDIX B  -  Analytical Results of Metallic Content
               Contained Within Green Liquor, White Liquor
               and Green Liquor Extract
TABLE Bl    -  Metallic Content of Green and White Liquors
               As Determined by Inductively Coupled Plasma
               Optical Emission Spectrometry (ICPOES)

TABLE B2    -  Metallic Content of White Liquor, Green
               Liquor and Green Liquor Extract by Emission
               Spectrometry As Determined by Spark Emission
               Spectrometry
                          45

-------
                          TABLE  Bl
             METALLIC  CONTENT  OF GREEN AND WHITE
                       LIQUORS  BY ICPOES
                       (Values  in yg/ml)
"^^vQompound
Element — ..
Li
Be
B
Na
Mg
Al
Si
K
Ca
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Sr
Mo
Ag
Cd
Sn
Ba
Pb
White
Liquor
ND
ND
ND
ND
ND
ND
0.01
ND
ND
ND
ND
ND
0.004
ND
0.13
ND
0.08
ND
ND
0.02
ND
ND
ND
ND
ND
Green
Liquor
ND
ND
ND
0.9
0.01
0.11
0.5
ND
0.15
0.003
0.009
2.3
0.22
8.3
0.005
1.6
0.14
0.13
0.008
0.31
ND
ND
ND
0.003
ND
Detection
Limits
(in H20)
0.3
0.001
0.05
0.3
0.006
0.05
0.01
1
0.001
0.002
0.003
0.005
0.001
0.003
0.005
0.01
0.01
0.008
0.002
0.01
0.01
0.001
0.01
0.001
0.01
ND = not detected
                               46

-------
                              TABLE  B2
          METALLIC CONTENT OF WHITE LIQUOR, GREEN LIQUOR
                    AND GREEN LIQUOR EXTRACT BY
                       EMISSION SPECTROMETRY
                         (Values in yg/ml)
\Cpmpound
Element ^\.
Li
Be
B
Na
Mg
Al
Si
K
Ca
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Sr
Mo
Ag
Cd
Sn
Ba
Pb
White
Liquor
ND
ND
ND
0.5
0.005
ND
0.01
ND
0.01
ND
ND
ND
ND
0.01
ND
ND
0.05
ND
ND
ND
0.001
ND
ND
ND
ND
Green
Liquor
ND
ND
ND
3
0.01
0.05
0.5
ND
0.01
ND
ND
4
0.8
10
ND
3
0.10
ND
0.1
0.1
0.005
ND
ND
ND
ND
Green
Liquor
Extract
ND
ND
ND
0.05
0.001
ND
0.005
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.001
ND
ND
ND
ND
Detection
Limits
0.1
0.001
0.005
0.05
0.001
0.01
0.01
10
0.005
0.01
0.01
0.01
0.005
0.01
0.05
0.05
0.001
0.05
0.01
0.01
0.001
0.01
0.05
0.1
0.01
ND - Not detected
                                    47

-------
APPENDIX C  -  Gas Chromatography Conditions Employed
               During Analysis
                             48

-------
                              TABLE Cl
Gas chromatographic conditions:



Instrument:  Varian Assoc. Model 2700 with FID




Column:      6' x 1/8" SS containing 10% 0V - 101 on 100/120 Supelcoport



Conditions:




             He  30 ml/min




             H_ ^30 ml/min    flows adjusted slightly to optimize performance




            air ^300 ml/min




             injection port 260°-270°C




             detector 300°-310°C
Program:     50° (5 min) 10°/min 215° (4 min)




Calculations:  Spectra-Physics System I Computing Integrator
                                  49

-------
                                TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
 1. REPORT NO.
  EPA-600/7-78-017
                           2.
                                                      3. RECIPIENT'S ACCESSION- NO.
 4. TITLE AND SUBTITLE
 Effect of Handling Procedures on Sample Quality
                              5. REPORT DATE
                               February 1978
                                                      6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.W. Adams, T.E. Doerfler, andC.H.  Summers
                                                      8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts  02140
                              10. PROGRAM ELEMENT NO.

                              EHB529
                              11. CONTRACT/GRANT NO.

                              68-02-2150, T.D. 10501
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC 27711
                              13. TYPE OF REPORT AND PERIOD COVERED
                              Task Final:  12/76-11/77
                              14. SPONSORING AGENCY CODE
                                EPA/600/13
is.SUPPLEMENTARY NOTES IERL-RTP project officer is Larry D. Johnson, Mail Drop 62,
919/541-2557.
16. ABSTRACT
          JThe report gives results of an evaluation of the effects of typical shipping
and storageTiandling procedures on organic materials  collected in Level 1 environ-
mental assessment (EA) studiesJ Parameters reviewed included: sample  container
composition (amber glass and high-density linear polyethylene), head space  compo-
sition (air or nitrogen), temperature  (38 C, 21 C,  and  5 C), lighting (dark and diffuse
sunlight), and catalytic species content.  Three sample sets,  representing fractions
obtained during a Level 1 EA and containing six model  organic compounds, were used.
A simulated 3-week shipping and storage cycle represented elapsed time between
sample collection and analysis.  All three experiments were in  accordance with
statistical principles appropriate for conducting factorial experiments.  Experimental
results were analyzed using analysis of variance,  to assess the relative effect of each
shipping/storage condition studied.
 7.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                  b.IDENTIFIERS/OPEN ENDED TERMS  C. COS AT I Field/Group
Pollution
Organic Compounds
Sampling
Quality
Materials Handling
  argo Transportation
Storage
Shipping Contai-
 ners
Containers
Analysis of
 Variance
Pollution Control
Stationary Sources
Environmental Assess-
 ment
13B
07C
14B      13D
                         15E,13H
                                  12A
 8. DISTRIBUTION STATEMENT

 Unlimited
                  19. SECURITY CLASS (ThisReport)
                  Unclassified
                        21. NO. OF PAGES
                              58
                  20. SECURITY CLASS (Thispage)
                  Unclassified
                                           22. PRICE
EPA Form 2220-1 (9-73)
                                          50

-------