'.1.

A.;.::..u..r.
.

:33... .

. 9-K}? a1

_. 2.":

..x.‘.....mn.a.....a....

1...: 3.11.135.
: 3.3. x91...

5.. c a
.. .. x .11..
.1 1 . ,
In}. {it
a l 1 33:3}. és.l)\.§:i
{.55. stvk\l\1xvr
51...?! r

z 1\ q . . . «1...... .
vuyIVIXu v.11 . , ‘ ‘ ‘ .1... 11.2.11?“ _5! i .31...
r L\. ; usnhvx... :3.

l .1 171...... 131:!
5:11.25: ‘ . $ .1... 1.)\
1.... 12.1.. x 5;

{1‘ 3

: .7hv 1x11
.31... 22:1... 11....)
:‘Ir

.5. 15 .n:1§h»11.§|t|\ . z
...!;>..\1‘ a). i \i?! :3)?
9413:4138)
. ktifi
{LI

119....)
,1113x17...r.|\\2 t I

.1515;
p7:)qa\tlv. >1

,1): :1. :7)...
y..\\x)..7)1x..:. III:
:9")... :71»;
I...\ v.
.35.!
it; ‘(.Ivua
i:::...91

xxx y‘all)
b ..:.......1fin{..v~l.
.zwsleyly... lave-12x rial; v

1? ‘ 5: .

x:..).~::.yl.1.{..
\ ‘xl 1: i3.

 

1‘3...
. .5311 3.3.1;
1.? (5:11.,

p 5.!!! 3:31.,1?
hi): \nX...PL..! 5.!

 

teases
’W

J

llillflllil|U|||H||||HN||||lllllHlllllllllllllHlHllHllil

31293 01563 5745

This is to certify that the
dissertation entitled
Reai-Time Anaiysis of Light Aikenes at Eievated
Temperatures and Pressures by Fiber Optic
Near Infrared Spectroscopy

presented by

Engin Deniz Yaivac

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Chemistry

 

 

 

Major professor

Date /2//,/.7(

MSU is an Affirmative Action/Equal Opportunity Institution 042771

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

fig
DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

ll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution
abimmpms-m

REAL-TIME ANALYSIS OF LIGHT ALKENES AT ELEVATED TEMPERATURES
AND PRESSURES BY FIBER OPTIC NEAR INFRARED SPECTROSCOPY
By

Engin Deniz Yalvac

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirement
for the degree of
DOCTOR OF PHILOSOPHY
Department of Chemistry

1996

Copyright by
Engin Deniz Yalvac

1997

ABSTRACT

REAL-TIME ANALYSIS OF LIGHT ALKENES AT ELEVATED TEMPERATURES
AND PRESSURES BY FIBER OPTIC NEAR INFRARED SPECTROSCOPY
By

Engin Deniz Yalvac

This study had as its goal, the simultaneous determination of ethylene and
monoalkylated light alkenes at elevated temperatures and pressures via fiber Optic Fourier
Transform Near Infrared (FT -NIR) Spectroscopy to enable the in situ, real-time
monitoring of these compounds. This goal was accomplished in two stages: First the
simultaneous determination of ethylene and monoalkylated light alkenes at room
temperature and pressure was studied. Then, the effects of elevated temperature and
pressure were investigated.

Ethylene and monoalkylated light alkenes namely, propylene, l-butene, l-pentene, 1—
hexene, l-Octene, l-decene, were selected for the determination. Various other alkylated
alkenes, i.e., 1,1 dialkylated, cis, trans dialkylated, and trialkylated alkenes were
investigated to determine the effects of the molecular structure for this analysis. The first

overtone of the asymmetric =CH2 stretch of the monoalkylated alkenes was found to be
unique for the light alkenes in the NIR region. This region of the Spectrum was used to

built a model based on Classical Least Squares (CLS) regression to demonstrate the

feasibility of the determination at room temperature and pressure. Ethylene and l-octene

solutions in Isopar B were studied as example. Concentrations of the alkenes were
determined with an error less than 1 wt %.

In the second stage of the study, the effects of elevated temperature and pressure on
the spectra were investigated via fiber Optic FT-NIR spectroscopy. The effects of
elevated pressures in the range studied (> 500 psi) on spectra are found to be insignificant
for these compounds. However, the temperature effects are critical, especially for the
lighter molecules such as ethylene and propylene. Therefore, the determination of these
compounds at elevated temperatures (25-140 °C) requires the incorporation of the
temperature effects into the calibration model. An apparatus to collect the spectra of
these compounds at elevated temperatures and pressures by fiber optic FT -NIR
spectroscopy was constructed. The determination of ethylene and l-octene was
evaluated. Calibration models were built using partial least-squares (PLS) regressions.
Predictive ability of the model, expressed as standard error of prediction (SEP), was less
than 0.5 wt % for both ethylene and l-Octene and the absolute error of determination for

alkenes is expected to be less than i 0.75 wt %.

Dedicated to my parents;
my husband, Selim;

my son, Arda. ..

ACKNOWLEDGMENTS

The author would like to express her gratitude to the following people and
institutions who have generously contributed to the completion of this thesis.

To Professor Stanley R. Crouch for his invaluable assistance and advice throughout
the course of this work. To the other members of the doctoral committee, Professor Kris
Berglund, Professor George E. Leroi, Professor John L. McCracken for their interest and
assistance. To Dr. L. David Rothman of the Dow Chemical Co. for his help and support
as on-site advisor.

My co-workers are also an intricate part of this research, without their contributions
this publication couldn't be possible. Among them I would like to thank Dr. Mary Beth
Seasholtz and Dr. Randy J. Pell for their assistance with chemometric tools, Dr. M. Anne
Leugers for her help with my endless questions in vibrational spectroscopy, Dr. Mark
Beach for building the high temperature, pressure apparatus. All the others: Dr. J. Paul
Chauvel, Dr. Del Lafevor, Dr. Mark Dittenhaffer, Dr. Bill Heeschen, Dr. Dave Albers,
Dr. Mark LaPack, Mr. Robert A. Bredeweg, Mr. Jeff Larson, Dr. Richard Hamer, thanks
for being available for many fruitful discussions.

Finally, I would like to thank the Dow Chemical Co. as the sponsor of this research

and allowing me fulfill a personal dream. "IT'S NEVER TOO LATE!"

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................................................ ix
LIST OF FIGURES ............................................................................................................................... x
CHAPTER 1 .......................................................................................................................................... 1
INTRODUCTION ................................................................................................................................. 1
In situ Real-Time Analysis ............................................................................................................. l
Near-Infrared vs Mid-Infrared Analysis .......................................................................................... 2
Why Fourier Transform Near-IR .................................................................................................... 5
Historical Background .................................................................................................................... 6

In situ, Real-Time Determination of Alkenes by Vibrational Spectroscopy ............................. 6
Determination of Alkenes in Mid IR Range Using Modified Reactors as IR Cells ................... 6
Determination of Alkenes in Mid IR Range Using Internal Reflectance Crystals ..................... 7
Determination of Alkenes in Mid IR Range Using Attenuated Total Reflectance Element ....... 8
Determination of Alkenes in Mid IR Range Using Flow-Through IR Cells .............................. 9
Determination of Alkenes by Near IR Spectroscopy .............................................................. 13
Chemometrics for Near-IR Spectroscopy ...................................................................................... 16
Goal of This Research ................................................................................................................... 23
CHAPTER 2 ....................................................................... 25
THEORETICAL BACKGROUND ...................................................................................................... 25
Effect of Temperature and Pressure on Vibrational Spectroscopy ................................................. 25
Pressure or Temperature Induced Phase Transition ................................................................ 26
Frequency Shifts Due to Pressure and Temperature ............................................................... 28

Band Shape and Intensity Changes ........................................................................................ 30
Splitting of Degenerate Vibrations ......................................................................................... 34
Doubling of Absorption Vibrations ........................................................................................ 34
Sensitivity of Vibrations to Expansion of Molecular Volume ................................................. 35
Multivariate Calibration Techniques ............................................................................................. 36
Classical Least Square Regression ......................................................................................... 36

Partial Least Square Regression ............................................................................................. 38

vii

CHAPTER 3 ........................................................................................................................................ 42
QUANTITATIVE ANALYSIS OF LIGHT ALKENES BY FT -NIR SPECTROSCOPY: ROOM

TEMPERATURE PRESSURE FEASIBILITY STUDIES ................................................................... 42

Experimental Procedures .............................................................................................................. 43

Materials ............................................................................................................................... 43

FT -NIR System ..................................................................................................................... 45

Fiber Optic Cell Assembly .................................................................................................... 47

Calibration Standards ............................................................................................................ 48

Data Analysis and Results ............................................................................................................ 49

Interpretation of Spectra ........................................................................................................ 49

Quantitative Determination ................................................................................................... 64

Summary ............................................................................................................................... 68

CHAPTER 4 ........................................................................................................................................ 70
IN SITU REAL-TIME DETERMINATION OF ALKENES BY FIBER OPTIC FT—NIR SPECTROSCOPY

AT ELEVATED TEMPERATURES AND PRESSURES ...................................... 70

Experimental Procedures .............................................................................................................. 71

High Pressure Temperature Apparatus ................................................................................... 72

Pressure Vessel ........................................................................................................ 72

Expansion Containment Chamber ............................................................................ 73

Head Space Assembly ............................................................................................. 73

Fiber Optic Probe-Holder Assembly ........................................................................ 74

Single Sided Transmission Probe ............................................................................. 74

Pressure Measurement and Control .......................................................................... 75

Forced Air Oven ...................................................................................................... 76

Temperature Measurement and Control ................................................................... 76

FT-NIR Spectrometer and the Detector .................................................................... 77

Determination of Vessel and Head Space Volume ................................................................. 79

Experimental Design ............................................................................................................. 79

Experiments for the Determination of Pressure Effects on Spectra ........................................ 82

FT-NIR Calibration Procedure ............................................................................................... 82

Data Analysis and Results ............................................................................................................ 83

Effects of Pressure on Spectra ................................................................................................ 83

Effects of Temperature on Spectra ......................................................................................... 85

Calculation of Liquid Phase Concentration at Each Temperature Setting ............................... 86

Calibration Models ................................................................................................................ 99

viii

Determination of PLS Model Rank ........................................................................ 104

Examination of Outliers ......................................................................................... 107

Examination of Spectral Residuals ......................................................................... 110

Model Prediction Errors ......................................................................................... 113

Summary ............................................................................................................................. 116
CHAPTER 5 ..................................................................................................................................... 1 17
CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH ................................. 117
Conclusions ................................................................................................................................ 1 17
Recommendations for Future Research ....................................................................................... 120
APPENDIX A: FT-NIR Spectra of Light Alkenes Used in the Study ................................................. 122
APPENDIX B: Excel Spread Sheets Showing the Calculation of Liquid Phase for Each Run ........... 142
APPENDIX C: Data File Names and Their Correlating Concentrations ............................................ 182
LIST OF REFERENCES ................................................................................................................... 185

LIST OF TABLES

Table l: Ethylene and Monoalkylated Alkenes .......................................................... 44
Table 2: Cis and Trans Dialkylated Alkenes .............................................................. 44
Table 3: 1,1 Dialkylated Alkenes and Trialkylated Alkenes ...................................... 45
Table 4: Characteristic IR Frequencies (cm-1) of Alkylated Ethylenes ...................... 59
Table 5: Shift of location of first overtone of asymmetric =CH2 stretch of

monoalkylated alkenes ................................................................................................ 61
Table 6: Predicted Concentrations of l-Octene and Ethylene ..................................... 67
Table 7: Experimental Design Versus Actual Experiments ........................................ 81
Table 8: Antoine Constants for l-octene and Isopar E ............................................... 88

Table 9: Literature Values for Constants in Temperature Dependent Density Equation
................................................................................................................................... 90

Table 10: Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for
Run # 26 ..................................................................................................................... 93

Table 11: Concentrations of Liquid Phase for Standards Containing Ethylene in 1-
Octene, or Isopar E ..................................................................................................... 95

Table 12: Concentrations of Liquid Phase for Standards Containing 1-Octene in Isopar
E ................................................................................................................................ 96

Table 13: Concentrations of Liquid Phase for Standards Containing Ethylene, l-Octene,
and Isopar E ............................................................................................................... 97

Figure 1:

Figure 2

Figure 3

Figure 4:

Figure 5:

Figure 6:

Figure 7:

Figure 8:

Figure 9:

LIST OF FIGURES

Schematic diagram of the FT-NIR Set-up .................................................... 46
Schematic Diagram of the Fiber Optic Cell .................................................. 47

Spectra of 98 wt %. l-Pentene, l-Hexene, l-Octene, l-Decene and Ethylene,
Propylene, l-Butene Solutions in Isopar E (monoalkylated alkenes) in 6,080 to
6,150 cm-l Region in Overlaid Format at Room Temperature and
Pressure .......................................................................................................... 50

Spectra of 98 wt %. Pure Octene Isomers (cis 2-octene, cis 3-octene, cis 4-
octene and trans 2-octene, trans 3-octene, trans 4-octene) (cis and trans
dialkylated alkenes) in 6,080 to 6,150 cm-l Region in Overlaid Format ..... 51

Spectra of 98 wt %. Pure Trialkylated Alkenes (cis 3-methyl-3-heptene, trans
3-methyl-3-heptene) in 6,080 to 6,150 cm—l Region in Overlaid Format 52

Spectrum of 98 wt %. Pure 2-Ethyl-1-Hexene (1,1 dialkylated alkene) and
Octane in 6,080 to 6,150 cm-l Region in Overlaid Format ........................... 53

Spectra of 98 wt %. l-Pentene, l-Hexene, 1-Octene, 1-Decene and Ethylene,
Propylene, l-Butene Solutions in Isopar E (monoalkylated alkenes) in 4,800 to
4,850 cm-l Region in Overlaid Format ......................................................... 54

Spectra of 98 wt %. Pure Octene Isomers (cis 2-octene, cis 3-octene, cis 4-
octene and trans 2-octene, trans 3-octene, trans 4—octene) (cis and trans
dialkylated alkenes) in 4,800 to 4,850 cm-l Region in Overlaid Format ...... 55

Spectra of 98 wt %. Pure Trialkylated Alkenes (cis 3-methyl-3-heptene, trans
3-methyl-3-heptene) in 4,800 to 4,850 cm-l Region in Overlaid Format. 56

Figure 10: Spectrum of 98 wt %. Pure 2-Ethyl-1-Hexene (1,1 Dialkylated alkene) and

Octane in 4,800 to 4,850 cm-l Region in Overlaid Format .......................... 57

Figure 11: Spectrum of 10 wt % l-Octene, 1-2wt % Ethylene (20 times enlarged) and

Pure Isopar E in 6,080 to 6,150 cm-l Region in Overlaid Format at Room
Temperature and Pressure .............................................................................. 65

Figure 12: Measured and Estimated Pure Spectra for l-Octene and Isopar E in 6,080 to

6,150 cm-l Region in Overlaid Format at Room Temperature and
Pressure .......................................................................................................... 65

xi

Figure 13: Schematic Diagram of the High Temperature, Pressure Apparatus .......... 72

Figure 14: Schematic Diagram of the High Temperature, Pressure Vessel ................ 73
Figure 15: Guided Wave Single Sided Transmission (SST) Probe ............................ 75
Figure 16: Schematic Diagram of the FT-NIR Set-up; Spectrometer, Detector, Fiber
Optic Probe, and Probe Holder Assembly .................................................. 78
Figure 17: Effect of Pressure on Spectra of Ethylene in Isopar E ............................. 85

Figure 18: Estimated Pure Ethylene, Propylene, and l-Octene Spectra as a Function of
Temperature ............................................................................................... 86

Figure 19: Corrected Liquid Phase Concentrations at Each Temperature for All Runs.

....................................................................................................................................... 98
Figure 20: Full Data Used in PLS Modeling of Ethylene and 1-Octene ................... 102
Figure 21: Second Derivative of the Data used in PLS Model ................................. 103

Figure 22: PLS Ethylene Model PRESS and SEP vs Number of Principal Component
Factors ...................................................................................................... 106

Figure 23: PLS l-Octene Model PRESS and SEP vs Number of Principal Component

Factors ...................................................................................................... 106
Figure 24: Ethylene PLS Model Outliers ................................................................. 109
Figure 25: l-Octene PLS Model Outliers ................................................................ 109
Figure 26: Ethylene PLS model spectral Residuals ................................................. 111
Figure 27: l-Octene PLS Model Spectral Residuals ................................................ 112
Figure 28: PLS Fit Errors for Ethylene Over All Temperatures .............................. 114
Figure 29: PLS Fit Errors for l-Octene Over All Temperatures .............................. 115

xii

Chapter 1

INTRODUCTION

In situ, Real-Time Analysis

In situ,, real-time monitoring of the chemical and physical changes that occur during
processing of chemicals enables the processes to be optimized and controlled."2‘3’4’5
Information collected in situ by sensors is transmitted in, real-time to the process control
computers which compare the state of the process to the desired state and take corrective
action. Chemical information obtained by chemical analyzers, is invaluable in process
control along with the information obtained from physical sensors.6’7 Benefits of closed
loop process control where chemical analyzers are used are not limited to better process
control or optimum operating conditions. Closed loop process control with feedback
from chemical analyzers eliminates delays and the hazards of manual sampling and
analysis.8’9‘10 An additional benefit for the environment is waste reduction or control in

many instances. ' 1.12.13

Optimization studies in research and development areas also benefit from, real-time in
situ monitoring via chemical analyzersm’IS These tasks often require investigation of
numerous conditions such as different process parameters or raw materials, solvents etc.

In situ monitoring via chemical analyzers can lead to faster and fewer numbers of

experiments and thus to more rapid reaction or process optimization.””17 Another area
that in situ, real-time monitoring excels is where the reaction or the process conditions
prevent the use of an alternative analysis via conventional samplinglg'19 Unstable
intermediates or products that cannot be determined under any conditions other than the
reaction conditions are good examples of these cases.20 Often, the, real-time analysis

required for fundamental research falls into the category of in situ monitoring.”

Various chemical analysis techniques are available for in situ, real-time
monitoring.22’23’24’25 Among a host of spectroscopic, separation based, dielectric and
acoustic monitoring techniques, vibrational spectroscopy represents a particularly
attractive choice owing to the unparalleled breadth and wealth of the molecular level

information that can be obtained.26’27'28'29

Near-Infrared vs Mid-Infrared Analysis

The NIR region contains weaker overtones of the fundamental absorptions and
encompasses bands that result from the harmonic overtones of fundamental and
combination bands associated with hydrogen atoms. Therefore, the compounds
containing O—H, N—H, and/or C-H bonds lend themselves favorably to analysis by NIR

spectroscopy.

The first overtone of a fundamental stretching vibration generally appears at a location

approximately equal to twice the fundamental frequency plus a relatively small shift to a

, l
h30 3 32

longer wavelengt which depends on structural and/or ambient conditions.
However, this prediction rule does not always hold.33’34’35 For instance, the systems
characterized by the presence of more complex combination bands, where a variety of
fundamental and overtone transitions occur simultaneously, are among these
unpredictable cases.36’37'38 The vibrational spectra of liquid solutions appear as broad
bands rather than the narrow lines associated with the spectra of gaseous compounds, and
these bands get broader with increasing energy, leading to substantial spectral
overlapping problems in the NIR region. In these instances, it is difficult to distinguish

overtones of fundamental stretching and/or overtones of combination bands for a

meaningful study of molecular structure or for quantitative analysis.

Due to the severe overlapping of broad bands, historically, the NIR region of the infrared
spectrum was considered of little utility in molecular structure determination, and little
attention has been paid to it until the last couple of decades. Only after the development
of fiber optic technology and the implementation of some of the statistical tools for
quantitative analysis, did N IR spectroscopy start to find more applications especially in in

situ,, real-time analysis.

One of the most important instrumental advantages of the N IR range over the MIR range
is the fact that the inexpensive and readily available silica-type optical fibers can be used.
The fibers, available in the MIR range are not as convenient and inexpensive as the silica-
type optical fibers. They are quite brittle requiring stainless steel meshing as cladding to

prevent breakage. Alternatives to optical fibers in the MIR, such as ATR (Attenuated

Total Reflectance) probes via light pipes, suffer from similar problems especially for in
situ,, real-time applications of process monitoring and control in manufacturing
environments. Therefore, while MIR offers advantages in spectra] interpretations, N IR is

exponentially more robust.

The NIR range is characterized by lower absorptivities than the MIR range and is, hence,
conducive to measurements of thicker samples or to the use of longer pathlengths. This is
considered another advantage of N IR over MIR for in situ,, real-time analysis
applications since the use of small pathlengths in process analysis is troublesome. Many
processes contain solid particles which cause plugging of the small pathlength (> 1 mm)

IR cells.

Straightforwardness of the sample interfaces, such as fiber optic light transmission probes
in the NIR region, is the reason for these devices to be commercially available at
affordable prices; thus their popularity in use for in situ,, real-time analysis.

Transmission and reflectance probes are commonly used sample interfaces in the N IR
region. Despite numerous designs, the principles of transmission and reflectance probes
are all based on the absorption of the sample according to Beer’s law. The use of these
types of probes in systems where the analysis needs to be done under process conditions
opened invaluable opportunities for the investigations of high temperature and pressure
systems and/or systems involving reactivity or flammability hazards when sampled

conventionally.

Why Fourier Transform Near-IR?

Although the superiority of FT-IR over conventional dispersive IR analyzers has been
well accepted for some time, FT-NIR has a more recent history.39’4o’41 The grating or
dispersive NIR analyzers suffer less compared to their MIR counter parts due to having
more sensitive detectors and more energetic light sources in the NIR region. Recently,
NIR instruments have begun to be used in process analysis applications where stability
and the reproducibility are critical. Vibration and temperature fluctuations in these areas
are a few of the causes of instability. In order to make more stable NIR instruments

manufacturers have started to move towards FI‘ technology.

FT-NIR also benefits from Fellgett's advantage (multiplex advantage) and J acquinot's
advantage (throughput advantage) as is the case in the MIR region. Additionally, the
Connes advantage (accurate frequency scale) leads to a stable wavelength axis in FT-IR
or FT-NIR compared to scanning or grating instruments. All modern FI‘ instruments
employ a high accuracy laser in the interferometer where all frequencies can be
mathematically related to the laser frequency leading to a very accurate frequency scale.
Hence, FT-NIR instruments are advantageous over dispersive NIR instruments for in

situ,, real-time monitoring applications.

Historical Background

In situ, Real-Time Determination of Alkenes by Vibrational Spectroscopy

Historically, the use of conventional (off-line) vibrational spectroscopy to study chemical
systems has been popular; recently, the use of vibrational spectroscopy for in situ
monitoring of processes is making inroads in, real-time analysis. In the MIR range in situ
determinations are mostly restricted to modified cells where the chemical reaction takes
place and the reaction mixture is monitored in, real-time. Another alternative is to have a
flow-through cell where the reaction mixture or the sample can flow through enabling,
real-time monitoring. This option, while, it is pretty popular with the analysis of gaseous
samples, it can be quite troublesome with liquid samples due to the possibility of having
different phases (solid particles, gas bubbles) and the coating of the cell windows.
Moreover, none of these cells lend themselves to use at high temperatures and pressures
without extensive modifications. Regardless, they have been used in the laboratory

environment, with extensive modifications, quite successfully.

Determination of Alkenes in Mid IR Range Us_in_g Modified Reactors as IR Cells
Among the applications performed in the laboratory environment, the ones pertaining to
the determination of alkenes can be considered precedent to this study. For example,
Karge and Bolding42 studied the reactions of olefins by in situ IR using a zeolite catalyst.
An IR cell was modified which simultaneously served as the fixed-bed flow reactor.
Similarly, hydroformylation of olefins were studied by a number of different groups

where the IR cell functioned as the reaction chamber as well. Yin et a1. 43 used Rh-based

supported liquid phase catalysts in heterogeneous hydroformylation of olefins at 50-150
°C by using a stainless steel tube (6 mm diameter 350 mm long) as the reactor /IR cell
and monitored the reaction mixture, real-time. Chuang and Pien44 studied
hydroformylation of ethylene on rhodium/silica catalysts and monitored the reaction
mixture by in situ IR spectroscopy. Xia et al.45 demonstrated that the hydroformylation of
olefins is faster with a rhodium-phosphine complex catalyst in 2-ethylhexanol using in
situ IR spectroscopy. Garland and Bor46 in another study of hydroformylation of olefins
with rhodium containing catalyst used 3,3-dimethyl-l-butene, and monitored the reaction
intermediates via in situ,, real—time IR analysis. Hydroformylation of propylene was
studied similarly under high temperature and high pressure (up to 17 atm) at reaction
conditions by Pan et al.47 using in situ IR. Cobalt catalysts were used by Mirbach et a1. 48
for the hydroformylation of 1-octene in hydrocarbon solvents and the in situ IR

spectroscopy revealed that excess l-octene is isomerized to internal olefins under these

conditions, but not hydroformylated.

Determination of Alkenes in Mid IR Range Using Internal Reflectance Crystals

A similar hydroformylation study was carried out by Moser et a1.49 However, a different
cell design was used in this study to monitor 1-hexene and its conversion products. The
reactor was embedded with a cylindrical internal reflectance crystal (CIR) which was
coupled with the FT-IR system (CIR-FTIR). These type of reactors, embedded with the
internal reflectance crystal have been popular in in situ monitoring of reactions where a
sample can not be removed from the system. Several examples can be found in literature

. . . . . . . o, 1. 2
about their implementauons in areas concerning other than olefinic compounds.5 5 5

An alternative way of using an internal reflectance crystal is to install it into a probe tip or
in a flow-through cell. The connection to the spectrometer and the detector is in this case
accomplished by either the light pipes or the fibers that can transmit in IR region of the
electromagnetic spectrum. One of the best example applications of this type of
instrument design is given by Moser et a1.53 describing the CIR-FT IR system using
chalcogenide glass optical fibers between the cylindrical internal reflectance element and
the FT-IR. Remote analysis in this fashion provides more flexibility between the FT-IR
and the reaction chamber. In this paper cobalt catalyzed hydroformylation of olefins
described where CIR containing reaction chamber was operated under high temperature

(90°C) and high pressure (750 psi) quite successfully.

Determination of Alkenes in Mid IR Using Attenuated Total Reflectance Elements
Comparable to CIR, Attenuated Total Reflectance (ATR) elements can be used embedded
in either reactors, in flow-through cells or at the tip of the probes for in situ monitoring of
various reactions. Altman and Jalkian54 describe in a patent where an ATR element was
installed in a flow-through cell to follow alkylation of olefins. A stream of olefins and a
stream of isobutanes are contacted in the reactor in the presence of an HF acid catalyst.
Further, the information obtained from the analyzer was used in controlling the HF,
water, and sulfolane feed streams to the reactor. Rekoske and Barteau55 used the ATR
element in a probe to monitor the carbonyl coupling in liquid-solid and gas-solid
reactions with reduced titanium reagents. They coupled the ATR probe with a FF-IR

system and monitored the olefins formed as intermediate compounds. The signal strength

and background response are some of the concerns that both the ATR and the CIR
element coupled systems encounter leading to lower signal to noise and lower sensitivity

when compared with conventional IR or FT-IR instruments.

Determination of Alkenes in Mid IR Range Using Flow-Through IR Cells

Among the in situ IR techniques the most popular ones are where flow-through cells are
used. This is especially practical with gas phase reactions. Several applications
involving determination of alkenes via in situ IR or FT-IR coupled with flow—through

1.56 monitored the

cells at various pathlengths can be found in the literature. Atkinson et a
gas phase reactions of alkenes with ozone by using in situ FI-IR. 1-pentene, l-hexene, 1-
heptene, l-octene, 2,3—dimethyl—1-butene, cyclopentene, and 1-methylcyclohexene were
investigated at room temperature and under slight vacuum (740 Torr). Intermediate
compounds of the reaction was determined as well which led to defining the reaction
mechanisms for these reactions. An earlier study by the same research group57 was
conducted for the gas phase reactions of a series of l-alkenes and l-methylcyclohexene
with the OH radical in the presence of N O. 1-pentene, l—hexene, l-heptene, 1-octene,
2,3-dimethyl-1-butene, cyclopentene, and 1-methylcyclohexene were investigated at

similar conditions using in situ FT-IR where this time a gas chromatograph is used to

correlate and confirm the IR results.

The mechanism of the Wacker Oxidation of alkenes over Cu-Pd-Exchanged Y Zeolites

158

was investigated by Espeel et a where an in situ IR analyzer is used for the

determination of the effects of reactant partial pressure in the rate of oxidation. Jiang et

10

al.59 used in situ FT -IR similarly to study the surface oxidation of light hydrocarbons on
prseodymium where the reaction temperatures reached to 573 °K. They demonstrated
that having a flow-through cell allows the cooling of the reaction products and that the

determination of the reaction products under reaction conditions is not always necessary.

Products and the mechanism of the gas-phase reactions of alkenes and N 03 was
investigated by Cariati and Rindone60 by using an in situ IR in a 480 L reactor. Products
were also investigated using gas chromatography by sampling the gas mixture
periodically. These authors demonstrated the value of in situ,, real-time monitoring with
almost instantaneous results over the sampling techniques with some response time.
Martens and Crouset61 used an in situ IR as an on-line analyzer for the monitoring and
control of the manufacturing of olefins and diolefins by steam cracking of hydrocarbons.
The reaction mixture of the steam cracking contains paraffinic hydrocarbons, naphthenic
hydrocarbons, and aromatic hydrocarbons, and the cracking temperature was controlled
by the response obtained from the on-line IR analyzer based on the concentrations of

naphthenic compounds which were monitored in the range of 0.8-2.6 pm.

The mechanism of selective oxidation reactions of 1-butene, 1,3-butadiene to maleic
anhydride was studied by Wenig62 using in situ IR spectroscopy. In this extensive study
other analytical techniques such as x-ray diffraction, laser Raman spectroscopy, SEM, X-
ray energy dispersive spectroscopy were also used for the determination of the catalyst
and the reactions that take place on the surface of the catalyst. Similarly, Moser63 studied

the selective oxidation of olefins and paraffins to maleic anhydride using in situ IR and in

4 .
k, 6 ’65'66 studied the same

situ Raman techniques. Wenig and Schrader, in earlier wor
reaction at elevated temperatures (up to 300 °C) where reaction products were determined

in a flow-through FT-IR cell.

Another implementation of IR or FT-IR flow-through cells has been in the area of
coupling with other analytical tools such as Gas, Liquid or Super Fluid Chromatography
(GC, LC, SFC). Infrared spectroscopy can be used as the detection system for these
techniques; the column effluent is interfaced with the IR via a flow-through cell. Infrared
spectroscopy or the FT—IR enhances the versatility of these separation techniques.
Several examples of the determination of alkenes can be found in the literature as

described below.

For instance Koizumi et a1.67 separated gasoline over a silica gel column (Devosil-60, 5
pm particle size, 25 cm long, 4.6 mm ID) and determined groups of hydrocarbons
(alkanes, alkenes, and aromatic hydrocarbons) by using FT-IR as the detector. Column
effluent was pumped through a flow cell (24 uL with 0.3 mm path length) where the FT-
IR determination was accomplished. First the alkanes then the alkenes eluted from the
column and the aromatics were back-flushed in order the reduce the analysis time.
Another LC-IR determination of alkenes was accomplished by Dobrov et al.68 for the
monitoring of the phenol alkylation process by olefins. A silica gel column (particle size
0.074-0.088 mm) was used and the mobile phase was optimized for the separation of

paraffinic and olefinic hydrocarbons.

12

In the area of GC-IR and GC/FT-IR coupling, several applications can be found
concerning alkenes. Sojak and co-workers69 combined a capillary gas chromatography
with a FT-IR and a Mass Spectrometric Detector (MSD) for the identification of isomers
of n-nonadecenes. A mesogenic stationary phase was found to be better than a
nonmesogenic stationary phase for the separation of 17 possible isomers of n—nonadecene.
Identification of compounds was confirmed both by the IR and the MSD where IR
provided distinctly different spectra for the geometrical isomers. Two separate groups
Listemann et al.70 and Gurka et a].71 similarly studied number of compounds linked with
environmental extracts including numerous alkenes via GC/FT-IR/MSD. In one of these
studies 106 compounds were identified many of them jointly based on functional groups.
FT-IR/MSD detector combination was found to be invaluable in enhancing the detection

and reinforcing the identifications.

Implementation of FT -IR via a flow-through cell as the detection system in SFC has also
been quite popular since the determination of high molecular weight compounds or
thermally labile compounds where SFC excels has always been in need of better detection
techniques. Unfortunately, many of the modifiers used in SFC interfere with the IR
determination. However, there are still several IR vibrations such as the C=O or the C=C
stretching that are interference free. Moreover, the use of deuterated solvents as
modifiers allows the detection of OH stretching vibrations while the use of chlorinated

solvents (e.g., CCl4) preserves the fingerprint region.72 However, determination of light

l3

alkenes is not commonly investigated by SFC since less complicated alternative

techniques are available.

The biggest disadvantage of flow-through cells is the fact that solids cannot be easily
analyzed. Diffuse reflectance IR spectroscopy can be used for in situ monitoring of solid
surfaces. The best examples of these type of applications pertaining to determination of
alkenes are in the area of catalytic reactions of alkenes on surface of solid catalysts. For
instance, using in situ diffuse reflectance FT-IR, Takeuchi et a1.73 studied the noble metal
promoted (Co/Si02), selective vapor phase hydroformylation of ethylene and obtained
strong absorption bands of linear and bridged CO species under the reaction conditions.
In the course of the reaction, changes observed in these bands suggested that the linear
CO species plays a major role in the CO insertion process to ethylene. Johnson et al. used
diffused reflectance IR in determination of hydrocarbons for the catalytic studies of
molecular sieve. Along with a gas chromatograph which was used in injections of
samples collected during the course of the reaction, diffuse reflectance IR monitored the
C-H stretching of the hydrocarbons that formed in the channels and pores of the

molecular sieve.74

Determination of Alkenes by Near IR Spectroscopy
Relatively fewer number of vibrational spectroscopic determinations of alkenes have been
reported using the Near-IR (NIR) region and as in situ measurements.75’76’77’78’79’80’81

Maggard,82 in a patent describes the determination of PIANO compounds (paraffin,

isoparaffin, aromatics naphthenes, and olefins) in fuel streams using NIR spectral region.

14

The technique uses 1672-1698 and/or 1700-1726 nm for naphthenes, 1622-1650 and/or
2064-2234 nm for olefins, 1092-1156 and/or 822-884 nm for aromatics, and 880-974 nm
and/or 1152-1230, 1152-1230, 1320-1380, 1470-1578, 1614—1644, 1746-1810, 1946-
1810, 1940-2000, and/or 2058-2130 nm for paraffins and isoparaffins. The statistical
treatment of the data such as multiple regression has been used to help the determination
of these families of hydrocarbons as well. This technique is especially designed to aid in
blending of fuels such as diesel fuels. A similar fuel analysis, determination of octane

number of gasoline is described by Parisi et al.83

This analyzer was used on-line via fiber
optic cables and the NIR cell was installed in a side stream of the process. The method

checked well with the values measured by conventional techniques.

Ethylene was determined by NIR spectroscopy to monitor its polymerization reaction
induced by an exciplex laser at high pressure (3200 bar), and high temperature (190-230
°C) in gas phase in a study by Breckemann et. al.84 A kinetic scheme was presented
which adequately describes the polymerization kinetics as a function of temperature,
overall density, laser pulse energy, and conversion. Buback and Tups85 in an earlier work
studied the high pressure co-polymerization of ethylene and carbon monoxide via IR and
NIR spectroscopy. The reaction was monitored in a stainless steel cell. Laser light was

used for the photochemical initiation of the polymerization likewise.

Lysaght86 describes a field portable fiber optic near infrared spectrometer for fuel analysis
applications. This is a short wavelength (700-1100 nm) NIR instrument designed to be

light weight (slightly larger than a briefcase) where the instrument is controlled by a

15

laptop computer. All of the components of the analyzer are chosen with this criterion in
mind. A tungsten lamp was used as the light source. A thermal electric cooled 1024
element silicon photodiode array detector was used along with a holographic grating for
wavelength dispersion. Volume percent ratios of aromatic and saturated compounds in
fuel were determined as an example application. Chemometric data treatment was used

for this determination.

All of the NIR applications discussed above are used for the general determinations of
olefins as a group of compounds and that they do not distinguish among its members. In
this research, an attempt to determine ethylene and the other monoalkylated light alkenes
simultaneously at elevated temperatures and pressure has been attempted for the first time

in the NIR spectral range.

Ethylene and monoalkylated light alkenes are industrially important compounds used in
manufacturing of polyethylene and similar polymeric materials. The use of on-line, at-
line and off-line vibrational spectroscopy for the control and monitoring of polyethylene
solution processes has been investigated earlier. Lange et al.87 described an on-line FT-
IR analyzer for the simultaneous determination of ethylene and l-octene in the MIR
spectral range where ethylene was determined at 1909 cm‘1 and l—octene at 1829 cm".
The information obtained from the analyzer is used in control of the feed addition to the
polyethylene solution process. Other on-line measurements such as density, viscosity,
melt index, and gas chromatography have also been studied for monitor and control of

P01yethylene manufacturing processes.”89

l6

Chemometrics for N ear-IR Spectroscopy

The application of multivariate statistical calibration and prediction methods to
quantitative FT-NIR spectroscopy has been one of the most important reasons for its
popularity today. Chemometrics can be broadly defined as the application of statistical
and mathematical methods for the design or optimization of chemical experiments and
for the efficient extraction of information from chemical data.90 Multivariate data
analysis has been shown to improve precision, accuracy, reliability and applicability of
infrared spectral analysis relative to conventional univariate methods of data analysis.
Multivariate methods drive their power from the simultaneous use of multiple intensities
(i.e. multiple variables) in each spectrum. Thus, the problem of spectral interferences can
be eliminated with the use of any one of a number of multivariate calibration and
prediction methods.9l’92 These methods include classical least squares (CLS), inverse
least squares (ILS), q—matrix, partial least squares (PLS), and principal component
regression (PCR). Correlation and Kalman filter methods have received less attention in
the infrared spectrosc0py therefore, and will not be described here. PLS and PCR exhibit
the greatest range of applicability and are attracting the most attention in the NIR

applications.

In spectroscopy, CLS is a multivariate least-squares procedure based directly on Beer's
law. Infrared spectroscopists have sometimes referred to this method as the K—matrix
method.”94 The CLS model accounts for errors in the spectral measurements. CLS can

accommodate spectral intensities at all frequencies for all calibration samples. In general,

17

all overlapping spectral components should be known for optimal performance of CLS.
By being a full-spectrum method, CLS has the ability to achieve improved precision since
there is a signal averaging effect when many or all the spectral intensities are included in

the analysis.95

The ILS method is a least-squares method [sometimes called P-matrix, or multiple-linear
regression (MLR) when applied to near-infrared]. It uses the inverse of the Beer's law as
its model. That is, concentration is modeled as a linear combination of absorbances. The
ILS model accounts for errors in the reference concentrations. The ILS technique is a
frequency-limited method and, therefore, is not capable of the precision improvements of
CLS from signal averaging of multiple intensities. However, ILS can often be a useful

method even if only one component is known for the calibration samples.96

The factor analysis methods of PLS and PCR are capable of being full-spectrum methods.
Like ILS, PLS and PCR can be employed even when only one component is known in the
calibration samples. Both PLS and PCR methods factor the spectral data calibration
matrix into the product of two smaller matrices. This amounts to a data compression step
where the intensities at all frequencies used in the analysis are compressed to a small
number of intensities in a new full-spectrum coordinate system. This new coordinate
system is composed of loading vectors that can be used to represent the original spectral
data. The intensities in the new full—spectrum coordinate system (called scores) are then
used in a model where concentration is presumed to be a linear function of these

intensities. Thus, PLS and PCR are methods that are concerned with modeling both

spectra and concentrations during calibration. PCR performs the factoring of the spectral
data matrix without using information about the concentrations. Therefore, there is no
guarantee that the full-spectrum basis vectors that are associated with PCR are relevant
for concentration prediction. PLS, on the other hand, performs the spectral factoring
trying to account for the spectral variation while assuring that the new basis vectors relate
to the calibration concentrations. Thus, PLS sacrifices some fit of the spectral data
relative to PCR in order to achieve better correlation to concentrations during

predictions.97

The diversity of multivariate methods available for application to quantitative
spectroscopy can create problems in terms of which one to use. The choice of
multivariate method for NIR spectroscopy often depends on the particular data set.
Therefore, the spectroscopist usually is faced with comparing the predictive ability of a
couple of different algorithms. However, lately more significant comparisons of these
methods from a statistical point of view have been made and methods were compared.
One of the best examples of this kind of comparison was made by Thomas et al.98 The
quantitative prediction ability of the four most heavily used multivariate calibration
methods in infrared spectroscopy (CLS, ILS, PLS, and PCR) have been compared by
using extensive Monte Carlo simulations. The simulations were performed assuming that
Beer's law holds and that spectral measurement errors and concentration errors associated
with the reference method are normally distributed. Eight different factors that could

affect the relative performance of the calibration methods were varied in a two-level,

eight—factor experimental design in order to evaluate their effect on the prediction abilities

of the four methods.

It was found that each of the three full spectrum methods has its range of superior
performance. The frequency-limited ILS method was never the best method, although in
the presence of relatively large concentration errors it sometime yields comparable
analysis precision to the full-spectrum methods for the major spectral component.
Among the full-spectrum methods, PCR and PLS were found to be very similar. The
major difference between these two methods was that PLS seems to predict better than
PCR in the cases when there are random linear baselines or independently varying major
spectral components which overlap with the spectral features of the analyte. This result
was not surprising when one considers the fact that the PCR decomposition is based
entirely on spectral variation without regard to component concentrations while the PLS
decomposition is dependent on the component concentrations. Because the spectral
variations caused by the presence of a random linear base line or major spectral
components can be reasonably large, the PCR decomposition is significantly influenced
by variations which have no relevance to the analyte concentrations. Therefore, PCR is

not able to predict as well as PLS in these situations.

It was also found that the differences between CLS and PLS are more numerous and
complex, but are also affected by the algorithms used in the analysis. The PLS method
was the optimal performer or close to optimal over the wide range of conditions

COHSidered in this study. Unlike CLS, all overlapping spectral components do not have to

20

be known, nor does the spectral base line have to be explicitly modeled. It seems that the
only inherent dangers of using PLS result from over- or underfitting by an inappropriate

number of factors.

It is important to realize that factors other than prediction ability often need to be
considered when choosing a calibration method. For example, the estimated pure-
component spectra generated by CLS contain significant qualitative information which is
very useful for the determination. The PLS method can also yield qualitative information
of better quality than is generally possible with PCR, but of poorer quality than CLS.
However, it should be noted that this study assumed that Beer's law holds. The biggest
problem comes in real-life applications where the concentrations of the components are
high enough to be outside the linear range of Beer's law. In these instances the choice is

usually PLS.

It seems that there are certain applications that CLS will be preferable not because its
predictive ability is superior, but it can provide quick, qualitative information. One
example of such a case can be a feasibility study in the laboratory environment where
several different chemical systems are being screened. The advantage of CLS in this case
would be the number of standards needed to build the model, since in CLS, one can built

a reasonable model by only using the pure spectra of the components.

However, robust models to be implemented in process analysis applications would

always have to be built with a sufficient number of standards drawn from careful

21

experimental design schemes representing the concentration ranges used in the
application. These types of models also need to be cross validated before they can be
implemented in process analysis applications which will require preparation of an
adequate number of standards. Therefore, for such process analysis application CLS

would lose its advantage in terms of a lower number of standards.

Numerous applications have appeared in the literature in last five years that compared
CLS and PLS. The PLS method has been the choice, overwhelmingly, for applications in
process analysis due to its robustness and ability to deal with possible sample matrix
changes from chemical interferences or temperature effects along with non-linearities
from Beer's law. For instance: Marjoniemi and Mantysalo99 compared PLS, PCR, ILS
and CLS to measure dye components in the visible region and PLS gave the best results.

Lin and Brown100

measured the salinity of sea water in NIR (680-1230 nm) again PLS
calibration produced a better predictive model. Ortho and metha cresol in water was
determined for an array of four piezoelectric crystals by Wei.101 et a1. PLS performed
better than CLS because there was some co-linearity problem with the data. In another
study, PLS gave somewhat better results than CLS in the determination of ascorbic acid
in pharmaceutical preparations, but only in the more complex samples.102 Slightly better

3 a visible

results for PLS were also obtained in a NIR tobacco analysis investigation,10
absorption measurement of myoglobin oxygen saturation,'04 and in on-line analysis of
sugars in a fermentation process.105 The PLS method was said to offer a more workable

approach than CLS in the analysis of mixtures, due to the mixture constraint.106 It was

also claimed that PLS could take nonlinear effects into account better than CLS.107

22

N i,108 has studied the reaction of the trace metals Co, Ni, Cu, Zn, and Fe with a
chromogenic reagent in the visible region with multivariate calibration methods. He
found that PLS performed better than CLS. Van de Voort and co-workers109 used FT-IR
for the analysis of milk and found a viable technique for this purpose when it was used
with PLS. Miller in a NIR study combined the first overtone and the second overtone
regions with the multivariate methods of PLS and CLS for the analysis of EPDM
(ethylene-propylene-dien monomers) polymers. The PLS coefficient spectra and the CLS
reconstructed spectra obtained from the calibrations were used to determine the sources
of the unknown spectral effects. Results indicated that the combination, first overtone,
and second overtone regions of the spectrum can be used to determine ethylene and
propylene concentrations in the terpolymers, and the combination region can be used to
determine diene concentrations. Because the unknown interaction effects are present in

the spectra of these materials, the PLS calibration method gives more accurate

calibrations than CLS.l ‘0

On the other hand, some studies indicated that CLS and PLS results were equivalent for

their sample sets, which included the determination of cadmium by inductively coupled

[11

plasma mass spectrometry, the determination of aromatics and saturates in aviation fuel

by NIR,112 and the estimation of crude lipid content in trout by NIR.113

23

Goal of This Research

The overall goal of this study was to find distinguishable features for ethylene and
monoalkylated light alkenes in the NIR region, so that fiber optic probes could be used
for the simultaneous, in situ,, real—time determinations of these compounds at elevated
temperatures and pressures. The first part of this research is involved with measurements
made at room temperature and pressure to establish the feasibility of the analysis. The in
situ measurements for the determination of the effects of the temperature and pressure on
this analysis was done in a separate system where a high pressure, temperature apparatus

was coupled with the fiber optic FT-NIR spectrometer.

An FT-NIR instrument incorporating a sample cell via fiber optic cables was used for the
feasibility study. Ethylene and monoalkylated light alkenes namely propylene, l-butene,
l-pentene, 1-hexene, l—octene, and l-decene were used in this determination, Various
other alkylated alkenes (e.g., 1,1 dialkylated, cis, trans dialkylated, and trialkylated

alkenes) were investigated to determine the effects of molecular structure on this

determination.

The first overtone of the asymmetric =CH2 stretch of the monoalkylated alkenes (6,080-

6 150 (:rn~ 1) was found to be unique for alkenes, and was used to build a model based on

ClaSsical Least Squares (CLS) regression. A mixture of ethylene and l—octene was taken

th example mixture to prove feasibility at room temperature and pressure. The
as e

24

reliability of the results was tested using the residual spectra which were obtained by

subtracting the estimated spectra from the measured spectra.

In the second part of this research the effects of elevated temperature and pressure for the
determination of light alkenes were investigated, and calibration models were developed

using high pressure, temperature apparatus. The ethylene and l-octene system, evaluated
during the feasibility study was investigated at elevated temperatures and pressures to

provide continuity from the feasibility study.

Considerable difficulties are involved in preparing well-known mixtures of the
components to build calibration models for accurate quantitative determination at
elevated temperatures and pressures. Consequently, for this particular work, in situ fiber
optic probe technology was essential. A special apparatus was built to obtain the spectra
of known concentration mixtures at varying temperatures and pressures. A high pressure
vessel was modified to incorporate a fiber optic probe which was coupled to the FT-NIR
spectrometer. The high pressure vessel was placed in an oven. Calibration standards
were prepared gravimetrically and the concentration of the liquid phase was corrected for
each temperature setting. Calibration models were built using partial least-squares (PLS)
regression to predict the concentrations of these compounds at varying temperatures and

pressures.

 

Chapter 2

THEORETICAL BACKGROUND

Effects Of Temperature And Pressure On Vibrational Spectroscopy

The effects of temperature and pressure on electronic and vibrational spectroscopy have
been studied in solids more than other phases since these effects can be investigated over
a wide temperature and pressure range without observing too many phase

changes.m'”5" 16

Effects of either high pressure or temperature on vibrational modes are mostly due to
changes relative to spin states, oxidation states, and changes in geometry.”7’118 However,

all the general changes in vibrational spectroscopy can be listed as follows:

1. Pressure or temperature induced phase transitions,
2. Frequency shifts due to pressure and temperature,
3. Band shape and intensity changes,
4. Splitting of degenerate vibrations,
5. Doubling of absorption vibrations,

6. Sensitivity of vibrations to expansion of molecular volume.

25

26

Other effects of pressure, such as the effects on hydrogen-bonded systems and mode
softening, will not be discussed here. Additionally, the effects of temperature and
pressure on the vibrational spectroscopy of liquids and dissolved gases in liquids are the
focus rather than gas phase (only) or solid phase (only) systems. However, where there
are similarities no differentiation will be made. For instance, some of the effects are
either the same or very similar for solids and liquids. Similarly, effects can be the same

for gaseous compounds and their solutions in liquids.

Pressure or Temperature Induced Phase Transitions

If the applied pressure or the temperature causes a phase change, substantial differences
can be observed in the vibrational spectra. This is a consequence of the differences
between the selection rules for the two different phases.l 19"” A change in the molecular
symmetry of the compound would be the most important reason for substantial changes in
spectra. However, spectral changes occur during phase changes where there are no
known changes in the molecular symmetry of the molecule. These changes are due to a
change in the distance between molecules in going from gas to liquid and to solid phase.
When the molecules are closer together more molecular interactions occur which may

effect the vibrational energy levels of the molecule.

The molecular interactions may be in the form of collisions. Two types of collisions are
described in literature: adiabatic and diabatic. After an adiabatic collision the molecule is

left in the same energy level that it had before the collision. With diabatic collisions,

27

there is a change in the energy level of the molecule. This type of energy changes which
may be caused by temperature or pressure changes will influence the vibrational energy
levels of the molecule since AE = hcfi and may cause frequency shifts, intensity changes,

and band broadening. ”"122

Mixtures behaves similar to the pure compounds when there is a phase transfer from
vapor to liquid phase. Additional intermolecular interactions are expected between the
compound and the solvent molecules which will affect the vibrational spectra. A good
example was reported by Morin and co-workers123 using supercritical carbon dioxide and
liquid carbon dioxide as solvent. Spectra of numerous compounds were collected in the
gas phase, in supercritical carbon dioxide, and in the liquid carbon dioxide. Substantial
frequency shifts were observed for the (C-H), (C=C), (C-H out of plain), (C-H), and
(C=O) vibrations studied. The frequency shifts from gas phase to solutions of the
compounds of interest in the supercritical phase observed were 1-3 cm’1 less than the
shifts observed from supercritical fluid to liquid phase (3-23 cm"). This illustrates the
temperature, pressure induced phase changes on spectral shifts due to the increased

molecular interactions.

Another example of intensity and band shape changes and frequency shifts in vibrational

SpeCtrosc0py because of phase changes is given by Taha and Tosson.124

The intensity and
frequency of N-O stretching mode of crystalline Ba(NO3)2 was monitored. When the
phase changed from ordered (I) to disordered (11) between 25-400°C, changes in band

shape, intensity and frequency shifts were observed. These spectral changes due to phase

 

 

 

28

change were explained similarly by density changes of the compound which affect

intermolecular interactions.

The light alkenes investigated in this study mostly remain in one phase under the
experimental conditions used. Thus, no phase shift transitions are expected to be

observed for these compounds.

Frequency Shifts Due to Pressure and Temperature

Pressure and temperature can induce other changes in the geometric configuration of the

126J27

compounds, (e. g., conformational changes, isomerizations,‘25 deformation, and

128

reorientations). When this happens substantial changes can occur in vibrational

frequencies. There are numerous studies that indicate conformational changes and related

129J30

IR or Raman frequency shifts in liquids, gases,I31 and lately with biochemical

compounds such as proteins.132 One of these applications given by Kato et al. '33
investigates the temperature, pressure and solvent effects on the 1,3 dichloropropane
conformational equilibrium between the trans-gauche and gauche-gauche conformations

and provides a correlation between enthalpy differences due to pressure and temperature

changes and the frequency shifts of the C-Cl stretching mode.

Other frequency shifts due to temperature occur with the thermochromic compounds.134

The cause of this type of effect is usually associated with the intramolecular proton or

135
l.

electron transfer. One case is described by Hoshino et. a where they studied the

thermally induced proton transfer in N,N'-bis(salicylidene)—p-phenylenediamine crystals.

29

The O—H vibrational stretch was monitored in the IR region showing considerable

broadening and a frequency shift toward lower energy.

For molecules that are not undergoing a phase transition or a conformational change, the

normal coordinates of the vibration remain unchanged. Since the masses also remain

unchanged, any frequency shifts are due to changes in force constants.136‘137

Sherman has developed a model to determine the effect of pressure on force constants

and thus, frequencies.138 According to this model, a molecule can be described by a

potential

V=A-Br'n+Cr'In [1]

where A, B, C, n, and m are constants and r is the interatomic distance. The rate of

change of the force constant k with r is given by

dk/dr = -(n + m + 3) kO/ro [2]

where kg and r0 are the ambient pressure values of k and r. A Lennard-J ones potential

shows a force constant change of

dk/ko= -21 (If/1'0 [3]

which amounts to a 21% increase in k for 1% decrease in r.

30

Pressure effects for solids cause changes for both inter- and intramolecular interactions.

This is illustrated by the examination of the equation

kE/drE = kl/drl [4]

where kE is the external force constant and k1 is the internal force constant. The values rE
and rI represent the respective compression. The order of frequency shifts with pressure

is 1-3 cm'1 kbar'l for external modes. For internal bond-stretching modes, the range is
0.3-1 cm'l kbar’l. Bond-bending modes are less affected and are on the order of 0.1-0.3
cm'1 kbar’l. These estimates are for average behavior and there are exceptions which do

not follow the general trends.

Since the pressures required for these types of effects (~14,000 psi) are much higher than
these employed in this study (max experimental pressure < 750 psi), no frequency shifts
are expected for the first overtone of the asymmetric =CH2 stretch of ethylene and the

monoalkylated alkenes.

Band Shape And Intensity Changes

It is known that band shape and intensity changes occur as well as frequency shifts due to
temperature changes in vibrational spectroscopy. These changes can be described by the
Boltzmann distribution functionm’140 The relative populations of the various energy

levels are calculated by:

31

n.
_1 : e—(Ei—EoflkT [5]

no
Where 11, and no are the numbers of molecules in the ith and zero energy levels, (Ei-Eo) is
the energy difference between the levels, k is Boltzmann’s constant and T is the absolute
temperature. For vibrational spectroscopy, the lowest possible energy level is where v=0.

If we substitute the energy term AB = th)’ into equation [5], we obtain

_i _ e—hCB/kT [6]

As can be seen from this equation at higher temperatures the number of molecules at zero
energy level will decrease since a fraction of the molecules will be at v: 1, and a smaller
fraction in v=2, etc. This will result in a decrease in the absorbance since the absorbance
is proportional to the population of molecules in the v=0 level according to the Beer's

law.

In the liquid phase, a general broadening in the shape of the band occurs along with a
decreased intensity of the band when temperature is increased. Several different studies
relate the shape of the band to experimentally measurable physical parameters such as
temperature, area or height of the absorbance band. One of these techniques is called the
correlation-function approachm’142 According to the correlation-function calculations

the shape of a spectral band can, in general be defined by band moments M(n).

32

M(n) = j (to — m0)“ I(u))d(o [7]

—@

where n = 0,1,2. . ., (1) = 21w , I((i)) is the height of the band at the angular frequency (1) of

the vibration (for instance I = ((00) is the height of the band at the center). These band

moments can be expended in a Taylor-series. However, only the first few moments can
be evaluated experimentally and related to temperature changes.143 For example M(2) =
2kT/I where T is the absolute temperature, and I is the moment of inertia of the linear

solute molecule for two atomic molecules such as HCl or DC].

Other methods correlating the band shape to the experimentally measurable parameters
use mostly least-squares fitting via computer programs.144 The cause of band shape
changes can be explained by population changes in the rotation-vibrational energy levels.
The temperature changes are similar to those observed with vibrational energy levels and

can be explained by Boltzmann statistics too.

The light source in vibrational spectroscopy causes an excitation of the rotational energy
levels along with the vibrational excitation levels. The fine structure of the rotational
vibrations are more visible in the vibrational spectra of gas phase compounds. In the
liquid phase or with solutions of gases in liquids, the rotational energy levels between the

vibrational energy levels are closer together, and often the rotational fine structure is lost.

33

In the liquid phase, a general broadening in the shape of the band occurs along with a
decreased intensity of the band maximum. Several studies relate the shape of the band to
physical parameters such as temperature.'45‘l46 Band shape changes can be explained by a
population change in the rotation-vibrational energy levels which are influenced by
temperature similarly to the vibrational energy levels. The radiation source in vibrational
spectroscopy causes excitation of the rotation—vibrational energy levels as well. This

rotation-vibrational fine structure is more visible in the spectra of gas phase compounds.

In the liquid phase or with solutions of gases in liquids, the rotation-vibrational levels are
effected more than gases due to collisions between the molecules. In the liquid phase, the
distance between the molecules get smaller leading to increased number of collisions.
Therefore, in condensed phase, the rotational structure is blurred by frequent molecular
collisions and IR vibrational bands are broad with no rotational structure.147 Solutions of
small molecules carries the characteristics of the gas phase more than the liquid phase.
Thus, For solutions of small gas molecules such as ethylene and propylene, fewer
collisions occur compared to the condensed phase. Therefore, more pronounced
population change in the rotation-vibrational energy levels is observed compared to the
condensed phase which will lead to more significant broadening in the band due to the
temperature change. This is why a more pronounced temperature effect can be seen in
the vibrational spectra of solutions of ethylene and the other gas members of
monoalkylated light alkenes (propylene, l-butene) compare to the liquid members.
Overall, the increased temperature tends to decrease the band intensity and increase the

width of the band.

34

Splitting of Degenerate Vibrations

A loss in the degenerecy of E- or F- type vibrations with pressure is also possible. In
some cases, this may be due to a lowering of the symmetry of the molecule where a
compound may have two or more molecules per unit cell. It is possible that the vibrations
in the unit cell might couple, causing a factor-group splitting, or Davydov splitting. '48’149
Dovydov splitting (also known as crystal field splitting) is observed in solids with a
crystalline structure. Wu C.K. et a1.150 studied the infrared active rocking modes of
crystalline polymethylenes in orthorhombic and monoclinic structures. Also, instances
are possible whereby unallowed vibrations may become allowed by application of
pressure.15 ' No such observations are expected for ethylene and the monoalkylated

alkenes since no symmetry changes are predicted to be induced under the experimental

conditions used and these compounds are not crystalline.

Doubling of Absorption Vibrations

In the course of various studies it is observed that a doubling of bands can occur with
pressure or temperature.15 2"5 3 This may be due to a lowered site symmetry.
Alternatively, two accidentally overlapping vibrations may occur at the same frequency.
These may be induced to separate because of a difference in the pressure dependencies
manifested by the two vibrations, or separate due to factor—group splitting.‘54 Fermi

. . . 1 5
resonance can also cause some band doubling as the pressure 1S increased. 5

35

It is also possible that at high pressure and temperature, normally forbidden modes might

become allowed (in a lower site symmetry) due to structural interconversions.156

No such observations are predicted for ethylene and monoalkylated alkenes due to
changes in pressure and temperature, since no site symmetry changes are expected to be

induced under the experimental conditions employed.

Sensitivity of Vibrations to Expansion of Molecular Volume
Generally, for complexes, molecular expansion is expected to affect the molecular
symmetry which can lead to different vibrations and the disappearance of some and

'57 Additional changes are observed due to molecular volume

appearance of others.
changes caused by bond distance changes. The bond distance changes that do not affect
the molecular symmetry will shift the vibrational bands to higher energy.158 A correlation
between the vibrational energy levels and the molecular volume changes due to

temperature or pressure can be given in the following equations!”160

AV = RTnO / P [8]

Substituting no from the Boltzmann equation

AV = RTn, /1>e-"°“"‘T [9]

36

No frequency changes for the first overtone of the asymmetric =CH2 stretch caused by
structural interconversions due to temperature and pressure are predicted for ethylene and

the monoalkylated alkenes under the experimental conditions employed.

Multivariate Calibration Techniques

Classical Least Square Regression

Classical Least Squares (CLS) regression in general describes the procedure where the
calibration model represents the physical law which relates the variation in the spectra
with composition, i.e., Beer’s law where spectral absorbance is represented as a linear
function of component concentration. '6' Since CLS explicitly uses a Beer’s law model, it
is referred as a hard model or explicit model method.162 Although the CLS method is less
flexible than soft modeling methods such as PLS and PCR, it provides greater qualitative
information in the range that it is defined.163 However, lack of fit can occur due to
deviations from Beer’s law such as curvature in the absorbance versus concentration plot,

or the existence of interactions between components in a mixture.164

CLS assumes that the mixture spectrum is a linear combination of the pure component

spectra. Therefore, a mixture spectrum A, based on Beer’s law can be written as a

combination of all pure-component spectra, K.165

A: CK+E [10]

37

where A is the m by n matrix consisting of the absorbances of each of the m samples at n
frequencies. C is the m by 1 matrix of the 1 component concentrations in the m samples,
and K represents the [by n matrix of pure-component spectra. E is the m by n matrix of

spectral noise or model error present in the spectra.

During calibration, the least-squares solution for the pure-component spectra K is:

K = (C'C)"C'A [11]

Where K is used to represent the least-squares estimate of the matrix K that minimizes

the sum of squared spectral errors in equation [10]. C' is the transpose of matrix C.

During prediction, the unknown sample spectrum vector a is equal to:

a=K'c+e [12]

where c is the unknown concentration vector and e is the error vector and K' is the
transpose of matrix K.. Note the order of matrix multiplication is reversed from that of

equation [10] since vectors are being represented as column vectors.

The least-squares solution for the component concentrations of the unknown is obtained

by using K from equation [11].

38

c=(f<f<')"l‘<a [13]

where K ' is the transpose of matrix K.

Similarly by using least—squares regression an estimate of the error in the predicted
concentrations can be calculated. This error is equal to noise when the set of calibration
spectra chosen for the analysis fit exactly with the sample spectrum. The fit will degrade
when nonlinearities are experienced or when the chemical impurities exist but are not

accounted for in the modelingwé’lm‘I68

In CLS, the absorbance of the spectrum at a given wavelength is assumed to be an
additive function of the concentration of the constituents. Hence, the CLS technique can
only be applied to systems where the make up of the sample has to have exactly the same
components as the standards. In other words, if there is an additional compound
(impurity) in the sample CLS will still assume that the sample is the mixture of the
components used in the training set. CLS also assumes that no other spectral variation

exist and effects the data.

Partial Least Square Regression
The theoretical background of partial least squares lies in factor—based regression which is
outside of the scope of this research. Only a brief description of the mechanics will be

provided here to be able to interpret the results of the PLS algorithm.

 

39

Partial least squares is a decomposition technique which consolidates the variation in
spectra. The spectral variations are due to mostly, the concentration variations of the
constituents, the inter-constituent interactions i.e. solvent effects, instrument variations
such as detector noise, temperature effects, density changes etc. PLS uses those
variations as partial spectra which are often called eigenvectors, (or spectral loadings, or
loading factors or principal components or simply factors) to construct the unknown
spectra. Presumably, these eigenvectors can be used to reconstruct the spectrum of a
sample by multiplying each one by a different constant scaling factor and adding the
results together until the new spectrum closely matches the unknown spectrum. These

scaling constants used to reconstruct the spectra are generally known as scores.169

In PLS both the spectral data and the concentration data are decomposed into their
respective most common variations. This generates two sets of eigenvectors and two sets
of corresponding scores: one set for the spectral data, and the other for the constituent
concentrations. These two sets of scores are related to each other through some type of
regression which constructs the calibration model. PLS performs the decomposition on
both the spectral and concentration data simultaneously. As each new factor is calculated
for the model, the scores are "swapped" before the contribution of the factor is removed
from the data. The newly reduced data matrices are then used to calculate the next factor,

and the process is repeated until the desired number of factors is calculated."59

40

The eigenvector that represents the common variations in the spectral data is called the
spectral "loading" and the eigenvector that represents the changes in the spectra
corresponding to the regression constituents is called the spectral weight.
Correspondingly, there are two sets of scores: one for the spectral data and another for the
concentration data. The mechanics of the calculation require, a weighting score to be
taken as the starting value, and through the regression the convergence of the calculation
is expected. After setting an initial value for the weighting scores, a spectral weighting
vector is calculated, which in the next step is normalized to unit length. In the following

step, the spectral scores are obtained.'69

Calculations of concentration loading vectors is done next and normalized to the unit
length. After this step, new weighting scores are calculated and convergence is checked
by new scores to the previous pass. If the scores are effectively the same, the PLS cross
product is calculated which then followed by the obtaining the spectral loading vector.
This spectral loading vector is normalized by the spectral scores and its contribution is
removed from the spectral data and also from the concentration data. After this point the
vector counter is increased by one and the steps are repeated until all desired factors are

calculated. '69

One of the difficult tasks in PLS is determining the correct number of loading factors to
use to model the data. The earlier factors in the model are most likely to be the ones
related to the concentration of the constituents, while later factors generally have less

information that is useful for predicting concentration. In fact these factors will

41

eventually begin to model the system noise (which usually provides the smallest
contribution to the data). However, if too few vectors are used to construct the model,
the prediction accuracy for unknown samples will suffer since not enough terms are being
used to model all the spectral variations that compose the constituents of interest.
Therefore, it is very important to define a model that contains enough vectors to properly
model the components of interest without adding too much contribution from the noise.
Models that include noise vectors or more vectors than are actually necessary to predict
the concentrations are called overfit. Models that do not have enough factors in them are

known as underfit.169

PLS can account for the compound of interest accurately if there are impurities in the

sample that were not accounted for during the calibration.

Chapter 3

QUANTITATIVE ANALYSIS OF LIGHT ALKENES BY F T-NIR
SPECTROSCOPY: ROOM TEMPERATURE PRESSURE FEASIBILITY

STUDIES

To find distinguishable features for ethylene and monoalkylated light alkenes in the NIR
region, and to achieve their simultaneous, in situ,, real-time determination at elevated
temperatures and pressures, a feasibility study was performed at room temperature and
pressure. The in situ measurements for the determination of the effects of the temperature
and pressure on this analysis were done in a separate system where a high pressure/
temperature apparatus was coupled with the fiber optic FT-NIR spectrometer (see

Chapter 4).

For the feasibility study, an FT -N IR instrument incorporating a sample cell via fiber optic
cables was used. Ethylene and monoalkylated light alkenes, namely propylene, 1—butene,
l-pentene, l-hexene, l-octene, and l-decene, were used in this determination. Various
other alkylated alkenes (e. g., 1,1 dialkylated, cis, trans dialkylated, and trialkylated

alkenes) were investigated to determine the effects of molecular structure.

The first overtone of the asymmetric =CH2 stretch of the monoalkylated alkenes (6,080-

5,150 cmll) was found to be unique for alkenes, and was used to build a model based on

Classical least squares (CLS) regression (multiple linear regression). A mixture of

42

43

ethylene and l-octene was taken as the mixture to prove feasibility at room temperature
and pressure. The reliability of the results was tested by means of residual spectra which

were obtained by subtracting the estimated spectra from the measured spectra.

Experimental Procedures

Materials

The light alkenes investigated during the feasibility study were ethylene and
monoalkylated ethylenes namely propylene, l-butene, 1—pentene, l—hexene, l-heptene, 1-
octene and l-decene (Table 1). Additionally, cis and trans-dialkylated alkenes, 1,1
dialkylated alkenes and trialkylated alkenes were investigated to determine the effects of
the molecular structure on the quantitative analysis of these compounds. Most of the cis
and trans dialkylated alkenes were isomers of octene. These were cis 2-octene, cis 3-
octene, cis 4-octene and trans 2-octene, trans 3-octene, and trans 4-octene (Table 2). Cis-
and trans-3-methyl-3-heptenes were the trialkylated alkenes used in the study. The
compound 2-ethyl—l-hexene was used as an example of a 1,1 dialkylated alkene. Table 3

shows the formula for 1,1 dialkylated alkenes, trialkylated alkenes and n-octane.

Isopar E, a liquid cut of saturated hydrocarbon distillation (Exxon Company, USA), was
used as a solvent during these experiments. Isopar E contains about 60 volume % iso-

octane and a variety of other saturated hydrocarbons.

44

Table l: Ethylene and Monoalkylated Alkenes

 

 

 

 

Compounds Formulae
1. Ethylene CH2=CH2
2. Monoalkylated Alkenes
2. l. Propylene CH2=CH=CH3

 

2.2. 1-Butene

CH2=CH=CH2=CH3

 

2.3. l-Pentene

CH2=CH=CH2=CH2=CH3

 

 

 

 

2.4. l-Hexene CH2=CH=CH2=CH2=CH2=CH3
2.5. l-Octene CH2=CH=CH2=CH2=CH2=CH2=CH2=CH3
2.6. l-Decene CH2=CH=CH2=CH2=CH2=CH2=CH2=CH2=CH2=CH

 

3

 

Table 2: Cis and Trans Dialkylated Alkenes.

 

Compounds

Formulae

 

Cis-2-Octene

CH(CH3)=CH(=CH2=CH2=CH2=CH2=CH3)

 

Trans-2-Octene

CH(CH3)=C(CH2=CH2=CH2=CH2=CH3)H

 

Cis-3-Octene

CH(CH2=CH3)=CH(=CH2=CH2=CH2=CH3)

 

Trans-3-Octene

CH(CH2=CH3)=C(CH2=CH2=CH2=CH3)H

 

Cis-4-Octene

CH(CH2=CH2=CH3)=CH(=CH2=CH2=CH3)

 

 

Trans-4-Octene

 

CH(CH2=CH2=CH3)=C(CH2=CH2=CH3)H

 

 

 

45

Table 3: 1,1 Dialkylated Alkenes and Trialkylated Alkenes

 

 

 

 

 

 

 

 

 

 

Compounds Formulae
1. Trialkylated Alkenes
(1.1) Cis-3-Methyl-3-Heptene CH3(CH3=CH2)C=C(CH2=CH2=CH3)H
( 1.2) Trans-3-Methyl-3-Heptene (CH3=CH2)CH3C=C(CH2=CH2=CH3)H
2. 1,1 Dialkylated Alkenes
2-Ethyl- l-Hexene CH2=C(CH2=CH3)(CH2=CH2=CH2=CH3)
Octane CH3=CH2=CH2=CH2=CH2=CH2=CH2=CH3
FT-NIR System

A Diamond 20 model FT -N IR spectrometer manufactured by KVB/Analect (Pasadena,
CA) was used in these experiments. The spectrometer was coupled with a cell assembly
via optical fibers as illustrated in Figure l. The Diamond 20 FT-NIR system has a 1.5
cm’1 transept IV hermetically sealed interferometer. Calcium fluoride optics were used
for the 12,000 to 1,200 cm”l region. The spectrometer includes an internal source and all
optical control electronics. It is controlled by a 66 MHz 486 based data system operated
under MS DOS/Windows 3.1. The spectrometer and the data station were interfaced with
a LAN interface card provided by the manufacturer. The system includes the
manufacturer’s FX-70, DOS based software program, and FX-80, Windows 3.1 based

software for the control and operation of the analyzer.

 

46

The spectrometer was coupled with a three stage, Indium—Arsenide (InAS), SP-2426
model detector repackaged and sold by KVB/Analect as well. The detector was cooled
with a thermoelectric cooler. The light from the spectrometer to the sample cell and back
from the sample cell to the detector were carried with fiber optic cables. A fiber optic

receiving module was used at the interferometer outlet and at the detector inlet.

Figure 1: Schematic diagram of the FT-NIR Set-up.

5mm Pathlength Cell

 

 

   

 

Launcher I]
FTNIR —| g
SPECTROMETER _l T

 

OPTICAL FIBERS Cell Assembly

Receiver i
--

Detector

 

, . ___ .-___—_._ __.. .

47

Fiber Optic Cell Assembly

A cell holder was designed as illustrated in Figure 2 to interface the optical fibers with an
Optical cell. Plano-convex lenses (5 mm in diameter) were used to collimate and focus
the light as shown. A removable cell was used in the middle of the holder. A set screw

was used to keep the cell from moving once in its place.

Figure 2: Schematic Diagram of the Fiber Optic Cell.

<—5rrm—->

light

(wear Fiber < E > Opt:

/

Collirmting Lens

 

 

 

 

 

 

 

Focusing Lens

 

 

A 5 mm path length cell was used in this study to obtain about 0.3 absorbance units (a.u.)
for the first overtone (6,120 cm") of the asymmetric =CH2 stretch of l-octene at a
concentration of 10% by weight. Background spectra were obtained with empty clean

optical cells for all of the absorbance measurements.

 

»a— - .— _—.\.__ ..

 

48

The fiber optic cables were 550 um, low OH content fibers supplied by SpecTran
Specialty Optics (Avon, CT). The cable was purchased as bulk fiber, and the fiber optic
connectors were installed by using a termination kit supplied by SpecTran, at the end of
the fibers, after cutting to the required length. About 20 feet of fiber optic cable was used

to connect the cell assembly and the spectrometer/detector.

Calibration Standards

The spectra of the most of the monoalkylated ethylenes and di and tri alkylated alkenes
were collected in pure form (98% or more purity) and in solutions using Isopar E as
solvent. All solutions were prepared at room temperature and pressure except for the
ethylene and propylene solutions in Isopar E which were prepared at 4 °C. For this
purpose, the gas (ethylene or propylene) was bubbled into Isopar E in a small vial placed
in an ice bath. This solution was quickly weighed and transferred to the optical cell
where the spectrum was obtained immediately. The l-butene solution in Isopar E was

prepared at room temperature due to the higher solubility of l-butene in Isopar E.

Additionally, mixtures of 2%, 4%, 6%, and 10% by weight of l-octene and the other
alkenes were prepared. The spectra obtained from these mixtures were used in checking
the validity of the calibration model. The spectra of all of these compounds were

obtained at 4 cm‘1 resolution and as an average of 64 scans from 4,000 to 12,000 cm".

49

Data Analysis and Results

Interpretation of the Spectra
Despite the immense amount of information in the literature for assigning the

‘ ' ° . . . . 4,
fundamentals of alkenes 1n the rmd-IR reglon,170171 '72 ‘73 ‘7 ‘75

there is comparatively
little information for the assignments of the first overtones in the NIR region. Therefore,

to be able to find distinct bands that can be related to alkenes in the NIR range, the

overtones of the fundamentals had to be estimated.

A coarse inspection of the NIR spectra obtained from mono, di and trialkylated alkenes
reveals that there are two main regions where differences in spectra occur. These are the
6,100 cm‘l region and the 4,750 cm'l region. Figures 3 through 6 show the 6,100 cm'l
region of the spectra, for ethylene and monoalkylated alkenes (Figure 3), cis and trans
dialkylated alkenes (Figure 4), trialkylated alkenes (Figure 5), and 1,1 dialkylated alkenes
(Figure 6) in overlaid mode. The 4,750 cm'l region of the spectra for these groups of

compounds, similarly, are shown in Figure 7, Figure 8, Figure 9, and Figure 10.

Additionally the full NIR spectra of all compounds used in the study are given in the

Appendix A Figures A1 through A17.

50

9E8 59635225

 

   
   

onwoo omow onwrw om—w onwmo ammo

0
25.305 lo 0
l c
a
In. D
..x. .—
ocoooa; . \ ..c. o
\ ..c. L m

885:;
..\ e
0:200; .. ...4 m; <

M) --\f

. . . 225m-—

 

 

 

.2335 use 2383th Eoom 3 358m @8130 E =o_wom 7:8 o2 .o 8 owed E Amuse—E cofigfioaofiv
m Emofl E masts—ow 25:5; Jae—30E dam—bum can 258D; .2580-— .ocoxoZL .ocoEom; so :5 ma mo 88on um 9.53m

51

?Eov 89.5562,

omen 88 38 85 £5 8% ammo

III“. .ll . .

 

OCOuooonem—hflhh

 

M o
W ocoaooéocahh / - no u
o:200-~.mc2._. K a
r . a
.

W .l N. D
_ .—
. 2.28.456 o
m
I. V. D
9.300.?«6 <

I. Q.

 

 

 

.8385 28 2380359 :53. 8 “£50m @8130 E :2me .-Eo c2 .c 9 owed 5 32—8.? noun—5.36 max: 28 £8 3:28
-v was: .ocouoom we“: .o:88-~ mag 98 25803 m6 6:384” m8 .o:28-~ $8 $088— 2880 85 do 3» mm mo .88on "v 9.5»?—

52

S-E$ 338.5335

 

 

 

 

one- as .38 gm .. one . ..eme . one..- . _ 38. o.
T.
. it-
o
ocouuoréifioséécufi o
i i l I l: illio
:
a
.—
oeoao:.n._>5o2.n.o_o 1N. e
m
no. a
<
TV.
'0.

 

 

 

.2335 can 05:23:89 800% “a 588m @8130 E 530%
_-Eo 02.0 2 08.0 E Aocoaozmaxfiofié was: .ocofionAAzfiufié mmov mosey—2 coax—33th. 8?— do :5 ma Lo «.5on “m 95$..—

53

...E3 coop—5:263

 

 

  

£8 8.8 85 8.8 8-3 omme £8 £8
lN.u
2.300.:
\ I I ... ... l t l llllllllllllllllllllllll o u
I .... o
..N. =
a
1v. a
.—
...o. o
f m
eeeso=.~._§m-~ .
im a
m <
l
_

 

 

 

 

.8885 use 2280585. 8005 3 88:5 2256
E commom ..Eo cad 2 256 E 2880 28 856:“ 3812sz :v 0:283-— -.»5m-~ 25 .Qe :5 mo 50 828on 6 9.53m

54

$-63 5953833

 

 

 

 

ommv owov omrov oats omhv 2me owwv ommv
... 883.5; 8.2585 ..... . ....... no
......... s.-- .. .
s o
w c
. l a
.. n
.. .—
\.fl .. . l m.w O
. 8580-— 2. ...... m
... 888:-— a
.1L\\ rm <
0880; ...
. lop—8:5-—
“ lmd

 

 

 

.8985 new 83838.5. 885 3 2285 228.50 E 5885 .-Eo omw.v 9 08% 5 $88.? eBflEEocoEv
m 8385 E 80:28 88.5; 88.505 8835.".— eca 8880; .8880; 8.88:; .8885-— .8 “3 mm «o 888m K 9:.me

55

:83 8983583

 

 

   

 

  

 

ommv 9.5V owhv omuwv ommv
m #414...) .... l... i .. . . . 5.x»! ._ .. 1......... 1.. .. .. . ... ... ....l ...Ib...h...r......q.:.ao «r... .. ill-.... 1.. ......tuiuthd) . N-
3800.938...
Will I I “Ni O
N 3809985.... I u ... ...... o
M e
M 4/ .
m 2.200.888... a
'0. .—
W \ O
y o o o
. \ m
\ c o
\...~. \..‘ ...” to. n
s
.. 2.200.820 <
. \ 8.800.920 . \
. . 8234.20 ...
. s
. ~ I w

 

 

.8385 use 83885th 885 um 3:85 68:85 5 =885. -Eo omwd 2 83. S A8855 882.336 8a: 25 23 3880
-v 8...: .288 m 8.8 .888- N 88 can 888. v 8 .0880. m 90 .888- N 88 88:83 8.800 85 68 “3 ma .8 8885.015 2&5

56

$.83 eons—.582,

owov of? ombu ommv ommv oomv ommv

 

 

 

ocoao:.m._>£o2.n.o_o -m.

<amonna=oo

IN.

 

 

 

.8385 was 83885th 8005 “a 2255 28:30 E 8.85
.-Eo omwxo 8 come 5 €=ofio¥mifio§$ 8m: .o:o&on-m-§:88-m 38 828.? 8833—35. 85 68 S, mm 50 888m 5 9—sz

557

$53 83832325

 

 

 

 

 

 

came 39. 2.5 89. £9.
. , .. . : . ... .- .. ro
......
8:30Oé
o
10. =
a
-o. e
.—
Iw O
m
8854.384 -..... e
I <
-3

 

 

 

.8385 93 83805th 8005 “a 3885 2.2830
s 8&8 _.:.o one... 9 o8... s 838 Ba 88:“ 83.5.35 1: 2.88.7384 28 as E 8 no 5.58% .2 2&8

58

A closer look at these regions indicates that the differences in the 6,100 cm'1 are more
significant comparatively. Therefore, this region was taken during the study to build the
CLS calibration model for the determination of ethylene and the monoalkylated alkenes.
To be able to interpret the differences in spectra in this region the fundamentals of the C-

H stretches in the MID-IR region have been inspected.

Table 4 shows the characteristic frequencies of alkylated ethylenes.176 Three important
vibrational stretching bands (fundamental bands in mid-IR) for the alkenes have the
potential for an overtone band in the NIR region.l77"78’l79' 180"“ These are: the =CH2
asymmetric stretch (3,090-3,075 cm'l), the =CH stretch (3,020-2,995 cm’l) and the in

phase (symmetric) stretch of =CH2 (3,000—2,980 cm'l). Other fundamentals, for instance
the C=C stretch, the CH2 wagging, the CH2 deformation and the CH2 rocking are at

longer wavelength; thus, their first overtones should not be in the NIR region. The
second or the third overtones of these bands in the NIR region would be too weak to be

significant.

Table 4: Characteristic IR Frequencies (cm!) of Alkylated Ethylenes.

 

 

Group cm" IR R Assignment
H \ / H 3090-3075 m m CH2 asym. str.
C=C 3020-2995 w s, p CH str.
R/ \ H 3000-2980 m w CH2 sym. str.
1840-1805 w — 2 x 910cm'
M2223?” 1650-1638 ms vs, p C=C str.
1420-1412 mw s, p CH, scis. def.
1309-1288 w s, p CH rock
995-985 vs w trans CH in—phase wag
910—905 vs w CH1 wag
688-611 w w cis CH in-phasc wag
R\ / H 3090-3075 m w CHz asym. str.
C=C 3000-3075 m s, p CHz sym. str.

9/ \H 1810-1770 w — 2 x 890cm"
1.1-Dialky1 1660-1640 ms vs, p C=C str.
(vinylidinc) 1420-1400 w m. p CH, scis. def.

900-885 vs w CH2 wag

H \ / H 3020-2995 m m CH str.

C=C 1662-1631 mw vs. p C=C str.
R / \ R 1429-1397 m —— cis CH asym. rock
cis-Dialkyl 1270-1250 w s 'cis CH sym. rock
730-650 ms vw cis CH wag
H \ / R 3010-2995 m m CH str.
C=C 1676-1665 vw vs, p C=C 511'.
R / \ H 1325-1300 vs s trans CH sym. rock
trans-Dialkyl ca. 1295 mw — trans CH asym. rock
980-965 vs vw trans CH wag
R \ / H 3040-3020 w m CH str.
c=c 1680-1664 w vs, p C=C str.
R / \ R 1360-1322 — w CH rock
Trialkyl 840-790 m vw CH_ wag
R \ / R 1680-1665 — vs. p C=C str.
C=C
/ \
R R

Tctraalkyl

60

A thorough investigation of this table and the NIR spectra of the alkenes used in these
experiments shows that the band observed around 6,120 cm'1 is unique to monoalkylated
alkenes and 1,1 dialkylated alkenes both of which contain a fundamental =CH2 stretch in
the mid—IR. Cis and trans octenes are cis and trans dialkylated alkenes and do not contain
a =CH2 group; thus, no overtone of either the symmetric or asymmetric stretch of this
group is expected. Experimentally, no overtones are observed experimentally in the
6,100 cm'1 region which confirms the prediction. None of the trialkylated alkenes (cis
and trans 3-methyl—3—heptenes) containing a single =CH stretch have any bands in this
region. Therefore, the band observed at 6,120 cm“l cannot be due to the first overtone of
the =CH stretch. Similarly, the symmetric =CH2 stretch can be eliminated since the
location of the fundamental is at lower energy to meet the energy level of 6,120 cm'1
region. Therefore, the band that is observed at 6,120 cm‘l region is more likely to be the

first overtone of the asymmetric =CH2 stretch of alkylated ethylenes.

The anharmonic oscillator approximation can be used to estimate the location of the first
overtone of the asymmetric =CH2 stretch (3,090—3,075 cm'l).182"83"84

The ener y levels (cm") of the anharmonic oscillator are iven by
g g

Ev,b=§,(v+1/2) - x.,§,(v+1/2)2 [14]

where 5, is the fundamental (cm"), v is the vibrational quantum number and Xe is the

anharmonicity constant.

61

For the first overtone v = 2 to v = 0

A5,, = 2v. — 65.x, [15]
For a typical value of Xe = 0.0025,185"86 the calculated wavenumber of the first overtone
using a fundamental of 3090 cm’1 is AEvib = 6180 - 46 = 6134 em" while the observed

value is 6129.8 cm". Hence, it was concluded that the observed band is the first overtone

of the asymmetric =CH2 stretch.

The spectra of ethylene, propylene, l-butene, l-pentene, l-hexene, l-octene and 1—decene

are shown in Figure 3 in overlaid mode, in the first overtone of the asymmetric =CH2

stretch region. Careful examination of this region of the spectra indicates that the first

-1
overtone of the asymmetric =CH2 stretch in 6,080 to 6,160 cm region shifts towards

longer wavelengths or towards lower energy with increasing chain length. Table 5 lists
these shifts among the members of the monoalkylated alkenes. The shift decreases with
an increasing number of carbon atoms in the molecule. Among l-pentene, l-hexene, 1-

octene and l-decene this shift is about only a couple of wavenumbers.

62

Table 5: Shift of location of first overtone of asymmetric =CH2 stretch of
monoalkylated alkenes.

 

 

 

 

 

 

 

 

 

Alkene Number of C Max of the band ACi A(C2-Ci)
atoms location (cm'l)

Ethylene C2 6,129.8

Propylene C3 6,123.1 6.7 6.7

l-Butene C4 6,119.5 3.6 10.3

l-Pentene C5 6,118.2 1.3 11.6

l-Hexene C6 6,116.1 2.1 13.7

l-Octene C7 6,114.3 1.8 15.5

l-Decene C8 6,113.6 0.7 16.2

 

 

 

 

 

 

In addition to the decrease in the frequency of the first overtone of the asymmetric =CH2

stretch, a decrease in the intensity of the band with the increasing chain length was also
observed. The decrease in the dipole moment derivative of the molecule due to the
increased chain length is known to be the cause of a decrease in the integrated area of the

bands in vibrational spectroscopy.187

Furthermore, the spectrum of 2-ethyl-l-hexene (a 1,1-dialkylated alkene) shown in Figure
6, clearly presents two overlapped bands that can be interpreted similarly. The band

observed at 6,126 cm"1 is due to the asymmetric stretch of =CH2 and the adjacent band is

due to the symmetric stretch of =CH2 (6,109 cm'l).

63

This interpretation can be used in developing quantitative determinations of these
compounds by NIR spectroscopy. For instance, simultaneous determination of mixtures
of monoalkylated alkenes starting from l-pentene to l-decene would be difficult because

the first overtones of the asymmetric =CH2 stretching vibrations overlap severely.

However, since a more significant shift is observed between the first overtone of the

asymmetric =CH2 stretching vibration from ethylene to l-butene, the simultaneous

determination of these light members of the monoalkylated alkenes is possible.

Additionally, the determination of pairs consisting of small and larger molecules should
also be more feasible than the determination of adjacent pairs. For instance, the
determination of the ethylene and 1-octene pair separation should be more feasible than
the determination of the l-octene and l-decene (0.7 cm'l) pair or the 1-butene and 1-

pentene pair (1.3 cm").

Moreover, the determination of mixtures containing more than 2 compounds, may be
challenging as well. Therefore, for demonstrating the feasibility of the quantitative
analysis of these compounds, the readily distinguishable, ethylene and 1-octene mixture

was taken as an example.

64

Quantitative Determinations:
For the ethylene/l-octene example, the 6,080—6,160 cm"l region (the asymmetric stretch

of =CH2) was used instead of whole spectrum since this region was found to be sufficient

to allow quantitative determination. Thorough examination of the FT-NIR spectra of 1-
octene, Isopar E and ethylene in this region revealed a small interference from Isopar B

(see Figure 11).

Additionally, the compounds that absorb in this region such as l- pentene through 1-
decene and 2-ethy1-1-hexene will interfere with the determination and bias both the
ethylene and 1-octene estimates for the reasons described earlier. Therefore, this
determination is limited to mixtures where there is no l-pentene, l-hexene, 1-heptene, 1-
octene, and longer chain members. However, 2-Ethy1-l—hexene has a unique band in
4,800-4,600 cm‘l region as shown on Figure 10. Therefore, if both regions are used 2-

ethyl-l-hexene will not bias the ethylene or l-octene results.

It was found that using the spectrum of l-octene measured from pure l-octene gave
biased results, attributed to deviations from Beer's law. Therefore, a pure spectrum of 1-
octene in the presence of Isopar E was estimated from spectra of samples containing
Isopar E and l-octene at concentrations of 10 to 20 wt % levels. The measured and
estimated pure spectra for l-octene and Isopar E are shown in Figure 12, in overlaid
mode. The l-octene spectrum in the presence of Isopar E appears to be shifted to higher

wave numbers.

65

Figure 11: Spectrum of 10 wt % l-lOctene,1-2wt% Ethylene (20 times enlarged) and
Pure Isopar E in 6, 080 to 6,150 cm Region 1n Overlaid Format at Room Temperature
and Pressure.

 

0.25- .
0.2- .
0
‘c’
50.15” QCIBDC -
a , x
i I \
.D
\
63 01.. I, \\ ’
...... /-—------- .
n.-- / V\ “-45.02%
00 - 1’ \ “
.s ..-’ ___~ 2‘ .
----. a” ~ ‘~\~
----------- a." ethYIene ~‘~:==

 

 

 

O 1 1 g 1
6080 6090 6100 6110 6120 6130 6140 6150 6160
wavenumber

Figure 12. Measured and Estimated Pure Spectra for 1- Octene and Isopar E in 6, 080 to
6,150 cm Region 1n Overlaid Format at Room Temperature and Pressure.

 

1.5 . . . . .

absorbance

 

 

 

 

0 1 1 1 l 1 1 , 7
6080 6090 6100 6110 6120 6130 6140 6150 6160
wavenumber

66

Since an appropriate instrumental arrangement to obtain the pure spectra of ethylene and
propylene was not available during the first part of the study, the ethylene spectrum was
estimated by subtracting the Isopar E spectrum from the spectrum of the mixture. The

error in estimation of this concentration could be as large as 1 wt % absolute. However,

obtaining high accuracy was not one of the goals of the study at this stage.

Concentrations of standard mixtures were estimated using the CLS model. The CLS was
chosen among the available linear regression techniques due to its simplicity (see
discussion on comparison of multivariate calibration techniques in Chapter 2).
Numerical results for the analysis of multiple samples are given in Table 6 . The table
provides the sample number, the true and predicted values of octene and ethylene, as well
as the uncertainties associated with these predictions. The uncertainties are from the
regression analysis, and increase as the noise in the data increases, or as the fit degrades.
The samples indicated by an asterisk are the samples that were used to estimate the l-
octene and Isopar E spectra. For samples 12-19, the absolute errors in the 1-octene
prediction ranged from -0.2 to 0.1%, with l-octene levels ranging from 4.8 to 20%. The
errors in the ethylene prediction were impossible to ascertain accurately, as there were

large errors in the true amount of ethylene in the samples.

The uncertainties shown in Table 6 are a result of the least squares calculation and
represent the confidence in the predicted concentrations. The uncertainties increase
dramatically for the prediction of the 100% l-octene samples, as these are not well

described by the estimated spectra, due to deviations from Beer's law.

67

Table 6: Predicted Concentrations of l-Octene and Ethylene.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

# True Predicted Delta True Predicted Delta Uncertainty Uncertainty in
l-Octene l-Octene (Difference) Ethylene Ethylene (Difference) in l-Octene Ethylene
1 0 -0.3 0.3 0 0.0 0.0 0.1 0.0
2 -0.2 0.2 0 0.0 0.0 0.1 0.0
3 0 -0.3 0.3 O 0.1 -0.1 0.1 0.0
4 100 98.5 1.5 0 -2.3 2.3 1.5 0.7
5 100 98.0 2.0 0 -2.6 2.6 1.3 0.6
6 100 101.6 -1.6 0 -4.0 4.0 1.6 0.7
7 100 101.2 -1.2 0 —4.3 4.3 1.4 0.6
8:1: 0 0.0 0.0 x 1.0 ? 0.0 0.0
9 0 -0.2 0.2 x 1.0 ? 0.1 0.0
10 0 -0.2 0.2 x 0.7 ? 0.1 0.0
11 0 -0.3 0.3 x 1.3 ? 0.1 0.0
12* 10 9.9 0.1 0 0.0 0.0 0.1 0.0
13* 20 20.1 -0.1 0 0.] 0.1 0.0 0.0
14* 4.8 5.0 -0.2 0 0.0 0.0 0.0 0.0
15* 9.9 9.8 0.1 0 0.0 0.0 0.0 0.0
16* 15.4 15.4 0.0 0 -0.1 0.1 0.1 0.0
17 4.8 5.0 -0.2 x 1.1 ? 0.0 0.0
18 9.9 9.9 0.0 x 1.2 ? 0.0 0.0
19 15.4 15.5 -0.1 x 1.1 ? 0.1 0.0

 

 

 

 

 

 

 

 

 

x True ethylene concentration around 1.5i 1 % wt

:1: Ethylene concentration defined to be 1.0 % wt. This was the pure spectrum used for ethylene. The true

concentration is around 1.5 i 1 % wt. All ethylene predictions are relative to this.

? = Unknown value

* Samples used to estimate l-octene and Isopar E spectra

Data shown in this table was generated by Mary Beth Seasholtz using MATLAB (Trademark of the

Mathworks Inc.)

68

The calibration model is expected to change from room temperature and pressure
conditions, to elevated temperature and pressure conditions, since the spectra of all of
these compounds will be slightly different at elevated temperature and pressure.

Therefore, this model is limited to room temperature and pressure conditions.

Summary

Distinguishable features for light alkenes in NIR region were established at room
temperature and pressure. Light alkenes ranging from ethylene to l-decene were selected
for the determination and to study the effect of molecular structure. It was found that the

first overtone of the asymmetric stretch of =CH2 for light monoalkylated alkenes i.e.

ethylene, propylene, l—butene, shifts toward the lower energy which may be enough to

distinguish and detemiine these compounds simultaneously in a mixture.

Additionally, the determination of mixtures containing small and larger molecules such as
ethylene and l-octene, ethylene and l-hexene, or ethylene and l-decene should be
possible as was demonstrated with ethylene and l-octene mixture in this study. Among
the heavier members of the monoalkylated alkenes, the addition of a carbon atom to the

molecule causes a slight shift in location of the first overtone of the asymmetric =CH2

stretch towards lower energy. Therefore, this method may not be suitable to determine 1-
pentene through 1-decene in a mixture simultaneously. Other alkylated light alkenes such
as 1,1 dialkylated, cis and trans—dialkylated and trialkylated alkenes were evaluated to
make sure that the technique is unique for the determination of the monoalkylated

alkenes. 1,1 dialkylated alkenes were found to be an interference in the region of the

69

asymmetric =CH2 stretch; however, they can be resolved if desired, by the use of the

-1
4,750 cm region in the model. Cis, trans dialkylated and trialkylated alkenes do not

interfere with the determination.

The first overtone of the asymmetric =CH2 stretch of the monoalkyl group was used to

build a model based on CLS regression The reliability of the results was checked from
the residual spectrum for the ethylene/ 1-octene example, obtained by subtracting the
estimated spectrum from the measured spectrum. Determination error was expressed as

predicted minus measured and, was less than 1% for the ethylene/ 1-octene analysis.

Chapter 4

IN SI T U REAL-TINIE DETERMINATION OF ALKENES BY FIBER OPTIC F T-

NIR SPECTROSCOPY AT ELEVATED TEMPERATURES AND PRESSURES

In the second part of this research the effects of elevated temperature and pressure for the
determination of light alkenes were investigated and calibration models were developed
using high pressure, temperature apparatus. The ethylene and 1-octene system, evaluated
during the feasibility study was investigated at elevated temperatures and pressures to

provide continuity.

Considerable difficulties are involved in preparing well-known mixtures of the
components to build calibration models for accurate quantitative determination at
elevated temperatures and pressures. Consequently, for this particular work, in situ fiber
optic probe technology was essential. A special apparatus was built to obtain the spectra
of known concentration mixtures at varying temperatures and pressures. A high pressure
vessel was modified to incorporate a fiber optic probe which was coupled to the FT-NIR
spectrometer. The high pressure vessel was placed in an oven. Calibration standards
were prepared gravimetrically and the concentration of the liquid phase was corrected for
each temperature setting. Calibration models were built using partial least-squares (PLS)
regression to predict the concentrations of these compounds at varying temperatures and

pressures.

70

71

Experimental Procedures

A high pressure, temperature apparatus was constructed to obtain spectra of light alkenes
in Isopar E at elevated temperatures and pressures. A block diagram of the set up
illustrating the components of the system is given in Figure 13. The apparatus consists of
a pressure vessel, and a circulation loop which incorporates the fiber optic probe holder
assembly and a variety of pressure transducers. The circulation loop employs a magnetic
circulation pump to recycle the contents of the vessel through the loop. Thus, the liquid
phase is circulated through the fiber optic probe holder where the sample and
spectrometer interface is provided. A magnetic stirrer is used for agitation within the
vessel. The vessel and the circulation 100p assembly, including the fiber optic probe

holder, are placed inside a forced air oven.

The apparatus is controlled and data acquired by a Camile (Segion, Midland, MI) brand
name data acquisition and control system. Camile is used in controlling the temperature
set points of the oven, the operation of the circulation pump, and the speed of the agitator.

Additionally, it is used to record all the experimental conditions.

72

Figure 13: Schematic Diagram of the High Temperature, Pressure Apparatus.

 

N2 C2114 Vacuum

Cam ile
‘ Controller

 

 

HIGH TEMPERATURE OVEN

 

 

 

Pressure Transducer fi9 Pressure Transducer

1:? 1??

 

1
I
I
I
|

1

 

   
 

    
 

 

 

 

 

 

 

 

 

 

HIGH ’:

PRESSURE gum

pnc

VESSEL Probe Cab],

Holder
SAFETY I
CONTAINMEN
l-‘TNIR

CHAMBER D SPECTROMETER

M AGNETIC STlRRER

 

High Pressure, Temperature Apparatus

Pressure Vessel:

The apparatus incorporates an Autoclave Engineers (Erie, PA) Eze-Seal model 100 mL
volume high pressure vessel (PLimit = 3000 psig, TLirnit = 315°C). The vessel consists
of two parts: a top (cover) and the body. These two parts of the vessel are sealed together
with a metal gasket (double delta seal) as shown in Figure 14. The top portion of the
vessel provides several inlets and exits via dip pipes and other connections incorporating

either high temperature and high pressure valves. All of these high temperature, pressure

73
connectors were welded to the line and a metal gasket was used for a seal between them.
Each time the vessel is removed from the system, these metal gaskets were replaced.

Figure 14: Schematic Diagram of the High Temperature, Pressure Vessel.

Cover

 

 
 

4— Optional Adapter

     
  

Double Delta Seal

 

 

Loose Flange

Vessel Body ——>

Expansion Containment Chamber

The high pressure vessel contains a safety relief valve rated at 2000 psi which is
connected to a six liter, high pressure, stainless steel vessel (rated at 3000 psi). In case of
an unexpected temperature increase or an unexpected reaction which would increase the

pressure of the vessel above 2000 psi, the expansion would be contained in this chamber.

Head Space Assembly

The high pressure vessel is connected to a circulation loop consisting of a probe/holder

assembly, circulation pump, and a variety of pressure transducers. The circulation loop

74

employs a Ruska brand name (Model 23 30-802) magnetic circulation pump to recycle the
contents of the vessel through the loop where the sample and spectrometer interface is
located. The head space is also connected to a vacuum pump to evacuate the system
when needed, and to a distributor assembly to discharge nitrogen and ethylene into the

system.

Fiber Optic Probe-Holder Assembly

The fiber optic probe assembly is a modified 3A" stainless steel tee. High temperature,
pressure connections are welded on both sides of the Tee to provide inlet and outlet ports;
a 3%" Swagelok fitting is welded on the third side to seal the probe into the holder. The
probe attaches to this assembly with 3%" Swagelok nut and ferrules. The probe is inserted

into the holder such that the sample gap is in the middle of the circulating stream.

Single Sided Transmission Probe:

The Single Sided Transmission (SST) probe is manufactured by UOP/Guided Wave (El
Dorado Hills, CA). It is a Visible/NIR transmission probe with the transmission Slit on
the side as shown in Figure 15. The probe internally contains a pair of 500 pm in
diameter fibers to provide the light interface at the windows. All of the optics are
installed in a 34 inch O.D., 15 inch long 316 stainless steel body. Two sapphire windows,
sealed by gold brazing, provide the sample interface. Two fiber optic connectors at the
end of the probe connect the probe to the spectrometer and the detector via fiber optic

cables.

75

Figure 15: Guided Wave Single Sided Transmission (SST) Probe.

Probe BOdY Protective Cover for Fiber Connections
Water-Tight Seal/Strain Relief

Fiber Optic Cables

—1 0
I
II
0.69" I4 12.0" >

... ,4 19.7" >
Pathlength 10mm

 

 

 

 

 

 

 

 

 

The Guided Wave SST probe is available from the manufacturer in 2 mm, 5 mm, 10 mm
and 20 mm fixed pathlengths. The optimum probe pathlength for the experimental
concentrations of ethylene and l-octene is chosen as 10 mm, deduced from the feasibility
study. The fiber Optic cables used to connect the SST probe to the interferometer and to
the detector are 500 um, low hydroxyl fibers supplied by SpecTran Specialty Optics.
About 20 feet of fiber optic cable is used to connect the SST fiber optic probe and the
spectrometer/detector set up. The fiber optic cable carries the light from the spectrometer

to the probe and from the probe to the detector.

Pressure Measurement and Control
Pressure measurements are done with one of the following pressure transducers

depending on the pressure of the system (Maximum 750 psi): MKS Model 315BA-

76

25,000 (0.1% accuracy), and GP250 Model 33 l-RM-GP/GJ/GX (0.25% accuracy), or
Setra C206 (0.13 % accuracy). Each transducer is calibrated against a digital pressure
gauge with an accuracy of 0.01%. Pressure calibrations are performed at several
pressures at each temperature set point. Before performing a calibration the pressure

transducers are allowed to equilibrate for at least two hours at each temperature setting.

Forced Air Oven

A Hotpack (Philadelphia, PA) model 212034 forced-air oven is used with a 9.5 cubic feet
internal volume. Oven walls are modified to establish several external connections such
as vacuum lines, pressure transducer signal lines, magnetic stirring driver. This floor
stand type oven has a 25"x22"x30" door which enables the removal of the pressure vessel
conveniently and easy operations and modifications of the system. The solid-state
electronics with multi-mode control option provides a remote access set point with an
accuracy of i 2°C from the Camile data acquisition and control system. The internal
oven temperature gradient below 200 °C is less than 0.9 °C. The Hotpack model 212034

oven contains a 4.4 watt 230V heating element with a 3 oC per min heating rate.

Temperature Measurement and Control

Temperature measurements are made with a Burns 12092 Platinum Resistance
Thermometer (PRT) connected to a Hewlett-Packard HP34401A digital multimeter. This
PRT is calibrated against another calibrated Fluke Model 2189A platinum resistance
thermometer. Several PRT’S are installed in the system to obtain the temperature of the

head space, liquid phase and different locations of the oven.

77

FT-NIR Spectrometer and the detector:

The Diamond 20 model FF-NIR spectrometer manufactured by KVB/Analect (Pasadena,
CA) is used in these experiments. The spectrometer is coupled with a Single Sided
Transmission (SST) fiber optic probe manufactured by UOP Guided Wave via optical
fibers as illustrated in Figure 16. The Diamond 20 FT-NIR system has a 1.5 cm”1 transept

IV hermetically sealed interferometer. NIR CaF2 optics are used for the 12,000 to 1,200

cm’l region. The spectrometer includes an internal source and all optical control
electronics. It is controlled by a 66 MHz 486 based data system operated under MS
DOS/Windows 3.1. The spectrometer and the data station are interfaced with an interface
card provided by the manufacturer. The system includes the manufacturer’s FX-70, DOS
based software program, and FX-80, Windows 3.1 based software for the control and

Operation of the analyzer.

78

Figure 16: Schematic Diagram of the FT-NIR Set-up; Spectrometer, Detector, Fiber
Optic Probe, and Probe Holder Assembly.

   
     
 

 

Launcher

FTNIR
SPECTROMETER

 

 

1

OPTICAL FIBERS

Receiver i

 
   
    

 

  
 

Detector

Fiber Optic Probe
Holder

The spectrometer is coupled with a three stage, Indium—Arsenide (InAS), SP-2426 model
detector repackaged and sold by KVB/Analect. The detector is cooled with a

thermoelectric cooler. Radiation is carried from the spectrometer to the fiber optic probe
and back to the detector with fiber optic cables. A fiber optic receiving module is used at

the interferometer outlet and at the detector inlet.

The spectrometer is coupled to the Camile process control computer by an input/output
(I/O) hardware. Via this connection, the data collection by the FT-NIR spectrometer is
triggered whenever the system reaches equilibrium under the desired conditions. An
analog signal (4—20 ma) from Camile is sent to the FT—NIR PC to initiate the data

collection.

79

Determination of Vessel and Head Space Volume

In order to determine the total volume of the apparatus (vessel plus the head space) a
calibration vessel with a volume of 144.37 ml is used. This calibration vessel is
connected to the system and to the pressure transducers via a specially designed valve
assembly. The calibration vessel is pressurized and its pressure is measured at constant
temperature. By opening the valve connecting the calibration vessel and the system, the
entire system is pressurized and the new pressure is measured. Since the initial volume
and the pressure are known and the second pressure is measured, the second volume is
calculated from: V2=P1V1/P2 at constant temperature. During the course of the
experiments several modifications were made to the head space assembly which required
the determination of the system volume each time. The system volume used in the

experiments ranged from 249 ml to 264 ml.

Experimental Design

The number of standards needed to develop a representative calibration model for
multivariate calibration techniques is an important factor in experimental design. The
concentration range of the standards and the similarity of the standards to the unknown
sample matrix are also factors that contribute to the accuracy of the model. The
concentration range of ethylene was chosen as 0.5 to 7.5 percent weight in solution, and
the range of 1-octene was selected as 2 to 30 percent weight in Isopar E for these
experiments. PLS models require the number of standards to be higher than the number

of variables in the system. However, the more data , the higher the confidence in the

80

analysis and in the statistics. Since the concentration range of l-octene was at least twice
the concentration range of ethylene, a rectangular two factorial experimental design was
used to represent the concentration range of the each component. For l-octene a 4 level,
and for ethylene a 3 level concentration design was configured. Thus, the experiments
consisted of three concentration levels of ethylene and four concentration levels of 1-
octene. Each run was performed at five different temperature (25, 60, 85, 110, 140 °C)
since the temperature constitutes another factor in the experimental design. Table 7
shows the experimental design layout and the target concentrations for each run. The
same table also lists the actual concentrations (liquid plus gas phase) and temperatures.
The experimental difficulties in addition of ethylene into the system resulted in slightly
different concentrations for the components than the targeted concentrations. However,
achieving actual concentrations around the targeted concentrations was sufficient for the
calibration purposes. Several runs were aborted due to experimental or equipment
problems. Ten to twelve spectra were collected for every run at each temperature once

equilibrium is established

81

 

m— ..aacm— fl 3:25: ...

 

 

 

 

 

03 .0: .00 .00 .3 $5: $3.00 mm
03 .0: .3 .00 .mm $3.0 $00.0 R 03 .0: .3 .00 .3 $000— $00.0m 0m
03 .0: .3 .00 .3 $36 $3.3 0m 03 .0: .00 .00 .3 $00.m $000.6. 0—
03 .0: .3 .00 .mm $wm._ $3.0m nN 03 .0: .3 .00 .mN $000 $000— w—
03 .0: .3 .00 .mm $2.0. $8.0m em 03 .0: .3 .00 .mN $00.0_ $000— 5
03 .0: .3 .00 .3 $86 $3.: mm 03 .0: .3 .00 .mm $00.m $00.0— 2
03 .0: .3 .00 .8 $0_._ $05.: mm 03 .0: .3 .00 .mm $0m.0 $00.2 m—
03 .0: .nw .00 .3 $20 $50.2 a 03 .0: .3 .00 .mm $00.0_ $00.0 3
03 .0: .00 .00 .mm $98 $0m.0 M: 03 .0: .00 .00 .3 $006 $00.0 m—
0: .0: .3 .00 .mm $30.— $0w0 S 03 .0: .3 .00 .3 $000 $00.0 N—
03 .0: .3 .00 .mm $.04v $Nh.0 2 03 .0: .3 .00 .mm $00.0_ $00.N 2
0v. .0: .2 .00 .mm $3.5 $3: 2 03 .0: .3 .00 .nm $00.m $00.N 0—
03 .0: .00 .00 .nm $00.m $8; 3 03 .0: .3 .00 .mm $0m.0 $00.N 0
03 .0: .3 .00 $mm._ $00; 2 03 .0: .2 .00 .mm $00.0 $00.0m w
03 .0: .3 .00 $00.0 $00.0m S 03 .0: .3 .00 .mN $00.0 $00.9 h
03 .0: .3 .00 $00.0 $00.9 Z 03 .0: .3 .00 .mm $00.0 $00.0 0
0300 1mm $00.0 $00.5 0— 03 .0: .3 .00 .mm $00.0 $00.N m
0.2.0— fimw $00.0 $00.N 0 03 .0: .3 .00 .mm $00.0_ $00.0 v
03.0— 1mm $000— $00.0 mwo 03 .0: .00 .00 .mm $00.m $00.0 m
03.0.12 $wm._ $00.0 <00 03 .0: .3 .00 .3 $0m.0 $00.0 N
0.1.02.3 $00.0 $00.0 h 03 .0: .3 .00 .mm $00.0 $00.0 _
Away—afl—RHJHV $ .—.>> $ .5.» Amh.v.—..m.—..N.H.~HV $ E?» $ P3
De Jar—Ema. Hzmawlkm HZWHUOA #23”— Ue mflfiDfiéflm—th films—VIBE film—H.004 #7::—
mZOEkJ—PZMUZOU J<DBU< ...mZO—HAEPZEUZOO Gmhmwm<h
mEZm—E—zmmxm A<DHU< ZU—mm—G A<HZmznflmmxm

 

 

mucoEEonm 038.33 m520> :wmmon— Rucofitomxm uh 035,—.

 

82

Experiments for the Determination of Pressure Effects on Spectra

Since the experiments are run at as high as 140 °C, the system pressure is expected to
reach several hundred psi due to the vapor pressure of ethylene, with the exact pressure
dependent on the ethylene concentration in the solution. Preliminary experiments were
conducted to determine the effects of pressure on the spectroscopic bands of interest.
Separate runs were prepared with 1-octene in Isobar E and ethylene in Isobar E at
constant temperature where the system was pressurized with nitrogen at several different

pressures (300 psi, 400 psi, 485 psi). Spectra were collected under each set of conditions.

FT-NIR Calibration Procedure

Several standards were prepared for the ethylene, l-octene system targeting the
experimental design. The actual ethylene concentrations ranged from 0.5 wt % to 13 wt
% and l-octene ranging from 0.5 wt % to 30 wt %. These standards were prepared
gravimetrically in the high pressure vessel at room temperature. The amount of each
component added was determined by mass difference. The vessel was first evacuated via
the vacuum pump, valves were closed and disconnected from the apparatus. After
weighing the empty vessel, the l-octene/Isopar E mixture, prepared in advance, was
loaded into the vessel using a 100 ml glass syringe. A 15 gauge syringe needle was
modified to fit to the high temperature, pressure connector of the vessel inlet. The vessel
was weighed again after degassing it under vacuum for several minutes. Next ethylene
was charged into the vessel, from a small (5 lb) cylinder via a distribution system
allowing the evacuation of the lines from cylinder to the vessel. After the ethylene

addition, the vessel was weighed and reconnected to the apparatus.

83

The apparatus was heated and allowed to equilibrate at the required temperature set point.
When temperature equilibrium was reached, the vessel pressure remained constant. At
each temperature set point, 10-12 spectra were obtained of the liquid phase. The system
pressure and temperature at equilibrium were recorded in order to calculate the vapor
phase composition, which was used to correct the liquid phase concentration of each

component.

Data Analysis and Results

Effects of Pressure on Spectra

According to theory, the effects of pressure in vibrational spectrosc0py are significant
when there is a phase change and/or a change in the symmetry group of the compound
due to the applied pressure. Either situation requires application of significant amounts of
pressure. The experimental conditions used during this study did not cause phase
changes or lead to a change in the symmetry group. The system pressure originates from
the vapor pressure of ethylene at the set-point temperature. At the highest concentrations
of ethylene ( 13 wt % at room temperature) and the highest temperature set-points (140
°C), the vapor pressure of the system is less than 750 psi. The supercritical temperature
(Tc) and pressure (PC) of ethylene are 10 °C and 52.2 atm (762 psi), respectively. Since
all experimental temperatures are above the supercritical temperature of ethylene,
ethylene is in the supercritical phase in this matrix. Furthermore, monoalkylated alkenes,

under the experimental conditions used, are in the liquid phase and, thus, no phase

84

transitions are expected. Also, according to theory, high pressures (~14,000 psi) are
necessary before starting to change the symmetry group of molecules. Hence, no change
in the first overtone of the asymmetric =CH2 stretch of the monoalkylated alkenes is

predicted.

Results of the experiments correlate well with predictions. Figure 17 shows the spectra
of 6.31 wt % ethylene in Isobar E (at room temperature) in the 6100 cm'1 region at several
different pressures. If any effects of pressure on band shape or frequency shifts do occur
in this system it would be most observable for ethylene. As seen in Figure 17, spectra of
ethylene solutions collected at different pressures showed no difference from each other
in the region where the first overtone of the asymmetric stretch of =CH2 is monitored.
Pressures above the vapor pressure of ethylene were obtained by pressuring the system
with nitrogen at constant temperature. After these experiments, the remainder of the

calibration was carried out at system pressure.

85

Figure 17: Effect of Pressure on Spectra of Ethylene in Isopar E.

0.6 1—

 

—— 485 si
0.5 —- p

. vapor pressure
— 400 psi
.. 300 psi

 

 

 

Absorbance

 

 

 

-o.1 1 1 1 1 + + 1 :
6400 6350 6300 6250 6200 61 50 6100 6050 6000
Wavenumber

Effects of Temperature on Spectra

The determination of the effect of temperature on the C—H stretch in NIR range for
ethylene and monoalkylated ethylenes was accomplished by varying the temperature from
room temperature to 140°C (25°C, 60°C, 85°C, 110°C, 140°C) for ethylene, propylene
and l-octene. A known concentration of ethylene in Isobar E was prepared and spectra
were collected at these 5 different temperatures. The spectra were corrected for
concentration changes due to the temperature/density changes. When these bands of
interest were overlaid, significant differences were observed on band shape. Mainly,
bands become broader and showed decreased peak intensity at elevated temperatures.

Similar experiments were done with propylene and 1-octene in Isobar E. The effect

86

observed on the band of interest by temperature gets smaller and becomes less significant
with increasing chain length. The effect on band shape is much more significant for
ethylene than is for 1-octene. These effects are shown in Figure 18 for ethylene,

propylene and l-octene for the first overtone of the asymmetric stretch of =CH2.

Figure 18: Estimated Pure Ethylene, Propylene, and 1-Octene Spectra as a Function of
Temperature.

 

Ethylene, 25 °C

.15 Ethylene, 140 °C

  

A \\
b
S l.
o
O
r 1 Propylene, 65 C /
b /
a I, \
n Propylene, 125 C ’ . \
7 '. k
C 05 I -
' Ar
e

1 -Octene, /'

25-140°C / , '
/’ , \~\-€\-

62 - o 6 20 6160 6130 61'20 61'00

 

 

Wavenumber cm'1

Calculation of Liquid Phase Concentration at Each Temperature Setting

Due to thermal expansion, a volume for expansion of the liquid needs to be left in the
vessel where the liquid is heated. If the head space volume is significant relative to the

total volume, the partitioning of the compounds between the liquid and the gas phase will

87

be important affecting the concentration of the liquid phase. When the vessel is heated to
higher temperatures, there will be more number of molecules in the gas phase than there
are at lower temperatures. Consequently, the concentration of the liquid phase will be
changed at each temperature setting. This is particularly significant with volatile

compounds such as ethylene.

In order to account for this change, and to be able to calculate the liquid phase
concentrations accurately, the concentration of the gas phase was estimated using
experimental parameters such as temperature and the pressure of the head space and the
vapor pressures of the compounds from physical property tables. Then, the calculated

vapor phase amount was subtracted from the initial concentration of the liquid.

The accuracy of the calibration model will directly be related to the accuracy of the
standards and the accuracy of the experiments. In order to calculate the head space
concentration, head space volume had to be calculated. This requires an estimation of the

liquid volume and the solution density at each temperature set point since

vvapor phase = Vtotal ' vliquid

Vapor phase concentration is calculated according to the Ideal Gas Law by using each

component’s partial pressure. The number of moles of each component in the vapor

phase is (niv):

88

 

nY= ‘ mm

The partial pressure of the components in the vapor phase (Piv ) for l-octene and Isobar E

are determined from the pure component vapor pressure of the compound (Pf) at that

temperature, by using the Raoult's Law. Since the total vapor phase pressure is measured
and known, ethylene’s partial pressure can be determined from the difference according

to the following equation.

PC2 = PTm — PC8 — P

Meas lsoparE

UN

Pure component vapor pressures of l-octene and Isobar E at room temperature were

obtained from physical property tables and corrected for temperature by using the Antoine

equation. ”38"”

B
1 I>=A————- m
0a ) T+C 1]

Antoine constants are given in the following Table 8 for l-octene and Isobar E.188

Table 8: Antoine Constants for l-octene and Isopar E

 

A B C

 

Isopar E 3.40926 1 185.31 222.048

 

 

l-octene 4.05935 1358.21 213.311

 

 

 

 

 

89

Using the pure component vapor pressure, the partial pressures of l-octene and Isopar E

are given by Raoult’s Law:

Pi = xiPi" [19]

where X, = the liquid mole fraction of component i.

In addition, the vapor or head space volume is required (VV) in equation [16]. This is

determined by subtracting the liquid volume which is sum of the volume of each liquid

L T
component, Vi , from the total vessel volume, V .

vv = vT — Enfvf [20]

Where V, = the molar volume of component i in cc/mol (or partial molal volume for

ethylene),

In order to find the liquid volume at each temperature set point, the liquid phase solution
density must be calculated. An overall solution density is obtained by dividing the total
molar mass of the material by the overall molar volume. The overall molar volume of the

solution is the sum of the molar volumes of each component, the latter are calculated by

90

dividing the masses of each component (number of moles of each component times their
molecular weight) by their pure component densities. The pure component density is

calculated at each set point temperature by equation [21]:

pi =A1+BiT [21]

where
pi = the density of component i in g/cc,
T = the temperature in degrees °C.

Ai= Experimental constants (literature values) for component i.

Bi: Experimental constants (literature values) for component i.

Literature values of A, B, and M, (molecular weight for component i) are given in Table

9 for l-octene, Isobar E and ethylenezlgg’189

Table 9: Literature Values for Constants in Temperature Dependent Density Equation

Mi Ai Bi
123.00 0.73468 -8.245E-4

1 12.215 0.73502 -9.167E-4
28.054 0.46757 0

 

After obtaining the overall molar volume, an overall density can be calculated by dividing
the overall mass by the overall molar volume. This density can be used for calculation of
the liquid volume by dividing the total mass originally loaded in the vessel by the overall
density (assuming that all the components are in the liquid phase). After calculating this

approximate liquid phase volume, the volume of the vapor phase and the number of

91

moles in the vapor phase can be calculated according to equations [16] and [21]. The

number of the moles in liquid phase is equal to the difference between the total number of

moles (n?) and the number of moles in the vapor phase (niV ).

n.L = n.T°‘ — n." [22]

 

xL = ' [23]

This calculation process summarized in Eq. [16]-[23] is repeated until the calculated

system pressure (PiTOt ) which is equal to:

P,” = PC2 + PC8 + PI [24]

soparE

converges with the measured system pressure (PTOI ). It was found that the iteration

Meas

converged after one pass. An example of this calculation for one of the runs of the
experiment is shown in Table 10. For this calculation process, the system is assumed to

exhibit ideal behavior (fugacity is equal to the partial pressure).

92

 

 

_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

50500.0 N0 053.00: 0000; 0005.0 00.05 $00.00— 0K5 O5 .EN».V0« w2340> 4mmmm>
V0020. «0005.0 00 «050.00 0000.0 v5«¢.0 50.«0 $00.00 5:02:00:
vo-wm«.0. 00:05.0 m 5:02:09 v05«.0« 0«0«.0 000«.0 00.«« $««.0« 00 00.0 05.: NO
m < V0000. 0:«.0 000—0 056 $00.0 NO 00.00 p.050<low:\00
5m+<nQ 00.02.00 5.600: 2: :0: 202.50%: .002 MNAOE d 600:2 85>; $53 a .0062:
_ r
0 _

 

 

 

0:500:00 \ «2:29. 000:: :0: .«0 2: :0 2:002: 2220:» :0:0::0 00: 2: :0 220 :50: A: 2.52:0 002:: :9. S «u :0 022: :0 0 2: 2 .202 «0 520:3.

 

«205.030.: \ 00> 02.2% :2. . 20:02 2: :0 .02: .020 00:00:: 00: 2: :0 220 :50: :0 005200 000:: :9. S 20:02 :0 022: :0 .0 2: 2 .0050 5:332 :O:<>

 

9195003250529. 000:: .02. . 00 2: :0 .02: 20:: 5:00:00 00: 2: :0 220 2:0: 5.: 005200 000:: :2. 5 2.200.. :0 022: :0 a 2: 2 .52 00 55:3.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _ 000:: 2:0: 2: :0 2:29.- .0002. 2: :0 2:29. :22». :23: 000:: 2259. 2: :0 020.5:

_ 0 0:28: .3 02.32. 22.: 350:0 2: :0 82: .20." 022.: 20:: 032.. 2: :0 0:39.

«503‘ n: :23? 0 2: 5: 2022:2000 2:29. :22: :22 2: A: 00030: 002: :22: 2: :0 002200 222.: 2: :0 5552::

1:50:00 2: 5.: 02:00: 22: :22: 2:20:20: :000 2: :0 E0» 50: :02200 «20:02 60.30 2:232:00 2: :0 2:29. :22: 2: :0 2:0

30:00. :0 200.252.220-500 :0 2:32: 20:22:22 2: 2 20:00 2 : .«o :0 2:32: 2220: 2: 2 25 < .0: «0

20:02 :0 00:02:05.~m:0:02 :0 :00 n 20:02 :0 2:32: 2220: 2: 2 25 1.0: 23:02

:3 a 8.32 803.: 00 _ 00 :0 00:02:08.0...» :0 :50... 8 32:32: 3:2. 2: a. :5 < .0: 00

0v0.««« 8.00: 0«000. 0 m :<:02 00 0. 22:20 0220200 20:02 :0 2:32: 20:2. 2: 2 =5 2:: 23:02

0 m L1 .20 2.200.: 2:. 0.2.02 .2 0 2.5. 05230200 02552.; 0 0 _ 0

050322.000.— 2 022000: 2:20:00 01:22:: :20: :0 0220200 20:20:52 :50: :00 2: .0 00 :0 2032: 20:2. 2: 2 =5< :> 00
_ _ _ _ _ _ _ ...:q :0 2:0: :2 2:32: 00:02:00 2 25.x :0

25:00:20 30:23«5:210:000005: 2:002: 2: :0 :0:2:l:00 2: 0. 0220000 2:32: 2:202:00 2 Sm: : 52:00

_ _ 20:20:82 :50: :0» 09.0 2: 2 50:5

23:02.0: 2:32: 0:300 2: :0 000002 2: 2 25:00::

30:00:02 2:32.32: 2: :0 30:02 2: 2 .5005:

3000.0 530.0 3 80.0 5«0 . .0 5 E ..0 300.0 «.00: 50.5: 0000.0 3000.0 0«0«.. 0 — «0.. 00.0. 0.05« 5.50. «0.05« 00.03
00000.0 05«00.0 50500.0 5«nr.0 03.0 0000.0 «. 00— 00. 5. 0000.0 0«000 505«.— 00—0.: 05.0— 0.V5« 00.00.. 00.00« mm.50«
«0500.0 00 50.0 05000.0 0505.0 055.0 :0«0.0 v. 00: 00. 0. 5000.0 005 —.0 0000.0 0000.0 00.0.. 0.02 00.005 v«.5m« «0.00«
00500.0 00 50.0 05000.0 059.0 053.0 :0«0.0 0.03 50.0— 0000.0 505 —.0 0000.0 0500.0 00.0: 0. on« 0000— 00.58 «0.08
3500.0 0900.0 05000.0 059.0 03.0 50«0.0 «.00: 00.0: 0000.0 0.5—.0 0000.0 0000.0 00....— m. 08 00.00: 0«.58 «0.08
00000.0 00000.0 00000.0 000 F .0 009.0 0000.0 0.00: 0«.? 000 p .0 0050.0 v9.0.0 0« 5.0 05...? «.mm— 0«.00 3.. .8 00.08
50000.0 00000.0 00000.0 003.0 002.0 0000.0 0. v0. 0«.? 503.0 5050.0 «03.0 0«50 00.0. «.00: 0«.-:0 :0. 50« 00.00«
00000.0 00000.0 00000.0 00»~ 5.0 vn«p.0 0500.0 0 00— 0«.? 0005.0 0050.0 030.0 005.0 00.9 5.00— 00.3 «0. :8 «0.00«
0300.0 00000.0 0—0000 3.3.0 00«5.0 0000.0 0. on. 0«. 3 500.0 0000.0 mow—.0 0««—.0 «0. C 0.00— 5.00 00.50— 5060—
«3000 3000.0 0500.0 3.3.0 820 0000.0 0. on: 0.. 2 0000.0 0000.0 «02.0 m««..0 5.: «.00: 00.00 0«.50: 5060—
000000 3000.0 0800.0 3.3.0 00: ..0 0000.0 m. on: 3.: 5000.0 5000.0 003.0 0««.. 0 00.: 2.00. 00.3 0:05 5060.
5000.0 v .0000 3000.0 003.0 mm. — ..0 0000.0 0. m: 8.0 «0«00 «000.0 030.0 5:.«0 0 v0.0 0.««— 50.0« :0.««— 5.0:
0030.0 «500.0 00000.0 003.0 cm: ..0 2.00.0 v.3; 00.5 3.00 0000.0 600.0 50—0. 0 50.5 p.53 5«.«« 00.5: «0.0:

.002 s .002 0 :02 9. 0 $5 .0 .3 42mm 5 =5: .0: 25... .0: =5: .0: 25¢ .:> 25.: W: 25< .: Sm: : 5%: 5:02: 50005:

«0 53:02 00 02: 40> 02:09. 55:02:: #00 «o :<:OM: 00 :1:02 00 5:00

53:3. MOSS :lo:<> 0«0200

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0m 2.. 00m :00 20020030 0002008000 000:: 2:05 0.: 053000 002.0 00200 Boxm 0m 030,—.

93

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2:002: :20002: 2: 2 0022.26 22...: 20:02 :00 00 .00 2: .0 020002: .2t2w 2: .0 5:0 2: 2 3.51 ~: 0..<0

 

_ _ «0 .0 2.82.3528 .0 :...u «o .8582: 5:2. 2.. 0. ...... < .0: «0

 

20:02 .0 5:02. 3:220:02 .0 :>.n20:02 :0 2:002: .220: 2: 2 25< .0: :<:o.m...

 

00 .0 00.32.35.200 .0 2.... 00 .8522: 2:2. 2.. a. :5... .0... 00

 

0.0.... ... 20......» .o 5:8: .2: 2.. a. u. ..0... 00 0.00...

 

 

2...... ... 20:8. .o 00:8: .2: 2.. 2 ... ..02 03.02 0.00...

 

:30: ... 2.200.: :0 20:02: .2: 2: 2 u. do: 00 930...

 

0.0.... ... 20.2.0 .0 E82. 2.0.9: 2.. 0. .02: «o 0.00...

 

2...... ... .223. .0 38.»: 2.0.2. 2.. u. 0:... 5.000. 0.00...

 

 

0.0.... ... 228.. .o .522. £22. 2.. u. .2... 00 0.00:

 

 

 

 

 

 

 

 

 

 

 

 

0:20.50 .0 30.0.: 2:022: 2: .....s 002.: 2:0: 2: :. 0:0. ...0 :0 00.2: .0 202:0: 0.: 0:: .::.:E .3 :02200 02...: 20.... :. 0:03.50 .0 2.0.2:. 2: 2 0 «0 0.00...

 

20:02 .0 30.0.... 3300.2: 2: 2...... 000:: 2:0: 2: :. 20:02 .0 022: .0 .025... 2: 0:13.22: .3 :02200 002.: 2:0: ... 20:02 .0 2.0.0:. 2: 2 0 5::00. 0.00...

 

 

4 0:200 .0 20.2.. 22:00.2: 2: .....s 002.: 20.... 2: :. 0:200: .0 00.2: .0 32:0: 2: 0:32:20 .3 :02200 002.: 2:2. :. 0:200: .0 50.0... 2: 2 0 00 0.00...

 

 

— .00 .0 00.2: .0 0 .22 2: E2: 000:: 02.2. 2: :. 2.22:0 .0 00.2: .0 0 2: 05.02300 :0 2.2200 002.: 2:0: :. 2.22:0 .0 00.2: .0 0 2: 2 ..02 «.0 0.00...
20:02 .0 00.0.: .0 ... .22 2: E2. 000020202. 2: :. 20:02 .0 00.2: .0 0 2: 02.02300 N: :02200 002.: 2:2. :. 20:02 .0 00.2: .0 0 2: 2 ..02 :560. 0.00....

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.0.00 .0 022: :0 .0020: .20. 2: 52. 002.5302... 0.: 0. 0:200: .0 00.2: .0 ..0050: 2: 02.02300 ...: 2.22:0 002.: 200.. :. 0:209: ..0 00.2: :0 ..02::: 2: 2 20:. 00 0.00...
00.0: 00.0: 00.5: who 00.0 050 5 .0 5500.0 0050.0 $00.0 $00.50 $00.00 205.0 00.00 00.00 0000.0 5000.0 000: .0
05.0: 05.0: 00.5: E 00.0 E .0 0500.0 0050.0 $00.0 $ 50.50 $00.00 05.0 v0.00 00.00 5000.0 5000.0 0005.0
00.0: 00.0: :00— 00.0 0 5 .0 000 5 .0 0000.0 :0500 $00.0 $00.50 $00.00 00.0 00.00 3.00 0 :0 : .0 0000.0 0000.0
00.0: 00.0: 00.0: 00.0 0:0 000 p .0 5000.0 :0500 $00.0 $00.50 $00.00 00.0 00.00 00.00 0 :0 : .0 000:0 0000.0
00.0: 00.0: 00.0: 00.0 0 _. .0 :00 5 .0 5000.0 :0500 $00.0 $00.50 $00.00 00.0 00.00 00.00 0:0 _. .0 0000.0 0000.0
00.0: 00.0: 50.0 : 00.0 00.0 00: : .0 0000.0 0050.0 $55.0 $00.50 $50.00 v0.0 00.00 00.00 500 : .0 0000.0 0000.0
00.0: 00.0: 50.0: 00.0 00.0 «03.0 0000.0 0050.0 $55.0 $00.50 $50.00 00.0 00.00 0:00 500:.0 000:0 0000.0
00.0: 00.0: 00.0: 00.0 00.0 :0: : .0 0000.0 0050.0 $55.0 $00.50 $50.00 00.0 00.00 00.00 500 : .0 0000.0 0000.0
00. 5: 00. : p 0 5 .: — 00.0 00.0 00: : .0 0050.0 0050.0 $00.0 $00.50 $00.00 50.0 :0.00 0.00 000 _. .0 0000.0 0000.0
:0. 3 :0.:: 0:. S. 00.0 00.0 00:. : .0 0050.0 «050.0 $00.0 $00.50 $00.00 50.0 :000 :000 000 : .0 0000.0 0000.0
00. :5 00. 5: 55.: 00.0 00.0 003.0 0050.0 0050.0 $00.0 $00.50 $00.00 50.0 :000 :000 000—0 0000.0 0000.0

00.0 00.0 50.0 00.0 :00 000—0 0050.0 0000.0 $00.: $00.50 $05.00 00.0 00.00 00.00 00: :.0 0500.0 5000.0
50.5 50.5 00.5 [00.0 50.0 :00 : .0 0 :50.0 0000.0 $00.: $00.50 $05.00 00.0 00.00 00.00 50: :.0 0500.0 500%]
35¢. a: =5< ... 35¢ 0: =5< 20.: 3.51 .0: ... ..OE .... ..OE ... 403 $52 05.5.: 8.5.: u a u .2 u .. do: . ...OE . 403 .
aim: 0.:‘0 NO 5.3.00. 00 NO I (:00. 00 00 25:00. 00 NO 11:00. 00 NO 15:00. 00
05:0... @003 £00... 0500... 05:0... £0... 05:0... £00... 05:0... 0.00... 0.00... 0.00...

 

 

 

00:00.: .0: 02.23.00 00.02.5050 80:: 2:05 0:: 00.3000 .0000 002:0 68:0 “3:000 0: 050,—.

 

94

The results of the calculations for each run are tabulated and the corrected liquid phase
concentrations (in mole fraction) for each component are given in Table 11 (standards
containing ethylene and Isopar E), in Table 12 (standards containing l-octene and Isopar
E), and in Table 13 (standards containing all three components). Additionally, calculated
liquid phase concentrations are plotted at each temperature and shown in

Figure 19.

The Excel spreadsheets showing the calculations of liquid phase concentrations of each

run are given in Appendix B Tables B1 through B20.

95

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00
0000.0 080.0 530.0 $00.00 $00.0 $.00 $00.00 $8.0 $00.9 0.509 00.50.. (835$ 00
030.0 080.0 .0900 $3.00 $00.0 $8.0 $8.00 $8.0 $00.— 0. 909 00. 00.. <8u75m 90
0000.0 080.0 50 5.0 $3.00 $00.0 $0v.0 $8.00 $00.0 $00.9 04.9 00.3 (00345 00
09
0005.0 080.0 80.0 $00.3 $00.0 $00.0 $8.90 $8.0 $3.0 N 0 S. 8.50.. 50.95.“. 9
035.0 080.0 .0000 $030 $8.0 $50.0 $8.5 $00.0 $v0.0 0.50 00.89 50075”. 5..
085.0 0000.0 0800 $8.3 $00.0 $05.0 $8.90 $8.0 $.00 5.58 3.3 5025 9
5005.0 080.0 0900 $8.3 $00.0 $00.0 $8.90 $00.0 $v0.0 0.000 00.00 502i 0..
0055.0 080.0 5000 $00.00 $00.0 $00.0 $8.5 $8.0 $3.0 0.509 00.00 5025 S.
0..
9005.0 080.0 050.0 $8.8 $00.0 $05.0 $00.00 $00.0 $8.9 0.8V 00.509 08.345 0..
0005.0 080.0 3.00 $9.00 $8.0 $50.0 $00.00 $8.0 $8.9 0.03. 00.89 0895a 2.
0005.0 080.0 05.0.0 $3.00 $00.0 $00.0 $00.00 $00.0 $8.9 0.08 004.0 08075”. 0..
0
0.00.0 8.0 .0500 $50.00 $00.0 $00...N $ 90.00 $8.0 $00.0 00.80 8.89 5007-5 0
00.0.0 080.0 59.0 $00.00 $00.0 $000 $ 90.8 $00.0 $00.0 0.080 05.00 502.3 5
.... 49... k g l 492 $5) $5) $.5— $§ $5. $3. <2 E (El .l 58>... 0
$18. 8 00 S18. 8 8 $18. 8 00 8 g 0
050: 0302 Q30... Q8: Q8... 950: as: as: Q30: v
055.4% Egnam .3238 0
0
mag. «5 W658 Z. 8 “6% ..

O z s. v. 0 . G u. m _ D O m _ <

 

 

 

 

 

 

 

 

 

 

 

 

 

m $08. 8 .2530... 5 28.5.3. 0558on $50530 80 SEE 2:05 0o 82.8.5800 u: v.05.

 

96

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

09 900.0 500 90.0 $50.05 $00.00 00.05 00.00 50.00 05.509 0:.me
0300.0 500 90.0 $90.05 $00.00 00.05 00.00 00.9 00.09 02.23.“.
50000.0 00900 $00.05 $00.00 00.05 00.00 00.0 00.3 02.23.”.
05000.0 000 90.0 $00.05 $50.00 00.05 00.00 00... 00.00 02.22.
9000.0 00009.0 $5.00 $00.3 00.00 00.09 50.00 50.509 92.2.3
300.0 00.90 $00.00 $00.3 00.00 00.09 0.0:. 00.09 92.23.".
00500.0 90009.0 $90.00 $00.3 00.00 00.09 3.0 0.03 302.5
00500.0 09009.0 $00.3 $5.0— 00.00 00.0.. 09.0 00.00 3075.".
00000.0 00v50.0 $09.00 $50.0 00.00 00.5 00.00 00.509 02.23.“.
00000.0 00050.0 $00.00 $00.0 00.00 00.5 00.: 09.09 902......
50 900.0 0.0500 $5.00 $00.0 00.00 00.5 3.0 _.0.3 9..me
90050.0 00000.0 $0.200 $3.9 00.00 00.0 00.00 00.509 09.23.“.
00050.0 09000 $00.00 $00.9 00.00 00.0 ..0. C 09.09 00.23.".
00050.0 05900.0 $90.00 $00.. 00.00 00.0 3.0 90.3 00%23fl
... ..03. n. ..03 u. .63. $5.: $.5— $5§ $.50 $53 $5.: 3”. m. 5mg... 3.“.
53.00. 8 N0 5.3.00. 8 00 5.5.00. 8 8 5.28
Q59... DEG... 930... 9:0... 938... Q50... Q50... 5.50... 950...
EDEQZSOW EDENUSGN 43530.0
wmaaw w £65.00. 2. 8 m..O 2052.230

 

 

m. 308. a. 0:200... 055850 830530 .0. 85.0 2:03 0o 2008:8280 ”N. 203,—.

 

97

Table 13: Concentrations of Liquid Phase for Standards Containing Ethylene, l-Octene, and Isopar E.

 

Figure 19: Corrected Liquid Phase Concentrations at Each Temperature for All Runs.

1-Octene Concentratlons

98

 

35.0%

30.0% H m

25.0% -~

20.0% ..

15.0% fl

10.0% ‘-

5.0% -~

X X“

0.0% -—X
0.0%

 

IO

*4 Xi‘.

X‘O

X/.I

2.0%

 

025C I60C BSC
X110C X140C

 

 

 

XX I 9

X:/ I

 

 

% fine We ; %
4.0% 6.0% 8.0% 10.0%
Ethylene Concentrations

12.0%

99

Calibration Models

The most popular Chemometrics method for quantitative analysis, partial (PLS) least
squares regression has been investigated for the determination of light alkenes at elevated
temperatures and pressures. The classical least squares has been the choice earlier in
proving the feasibility of the work due to its simplicity. However, partial least squares is
needed for modeling the data obtained at elevated temperatures (see discussions on
multivariate calibration techniques in Chapter 2). The goal in this regression technique is
to find a mathematical relation between two data sets. In spectroscopy one of these data
sets represents concentrations and the other the spectra. The spectra and the known
composition concentrations of the standards form the calibration set (also known as the
training set) from which the calibration equations (calibration model) are built. Once the
calibration equations or the model is created, predictions of the concentration values from

measuring the spectrum of the unknown can be accomplished.

During the initial study, CLS was chosen to investigate the feasibility of the analysis, and
the goal was not to develop highly accurate models but provide qualitative analysis.
However, for the second part of the study the goal was to be able to do the analysis
accurately at elevated temperatures and pressures. All of the chemometric techniques
were screened for this purpose and PLS was chosen as the most promising technique.

Models were developed using this technique for ethylene and l-octene separately.

Extensive investigation of the spectra and the spectral features of the light alkenes in the

NIR region was reported in Chapter 3. Unique bands (first overtone of asymmetric

100
stretch of =CH2) were classified for the identification of these compounds. Using this

unique region (6268 to 6071cm-l) of the spectrum with the regression techniques
produces accurate models. For many of the chemometric techniques (e. g., CLS and PLS),
the whole spectrum can be and sometimes should be utilized in the model in order to
obtain information about the impuritieslgo'191 However, in other cases, as is the case with
light alkenes, utilizing the whole spectrum does not produce the best model.192 For
instance, in the NIR spectrum of these compounds highly absorbing bands in the C-H

stretching region are observed due to the high concentration of paraffinic hydrocarbons.

Therefore, in the multivariate data modeling in this study, only the 6,268 to 6,071 cm-l

spectral region is included. The spectra are baseline corrected at 7,650 cm'1 during these

studies by subtracting the absorbance at 7,650 cm-l.

One way to validate the predictive ability of the calibration model is to perform a cross
validation. One of the popular methods of cross validation is done by leaving one-sample
or a selected small group of samples out of the model, building a new model, and
predicting the concentration of the left out sample or samples. For the PLS cross
validation, the models are estimated by leave-one sample out cross validation technique

as well.

PLS models for the ethylene and l-octene system was accomplished by using Pirouette
(Trademark of Infometrix, Inc.) software. Two separate PLS models were built for
prediction of ethylene and l-octene. Additionally models that excluded some outlier data

were also generated and the best predictive model was chosen.

10]

A mean centering preprocessing technique was used in modeling via Pirouette software.
Mean centering involves subtraction of the mean of the column in a data matrix from
each entry in the column. This technique helps take care of the intercept term in the
regression model and helps stabilize the regression. It is recommended that mean

centering should always be performed as the preprocessing technique for PLS.I89

Figure 20 shows the full data set used in building this model. Also in Appendix C the
data file numbers and correlating concentrations used in construction of the PLS model

for both ethylene and l-octene are given in Table C1

The second derivative of the data is used as the smoothing/transformation technique in
building the PLS model. Figure 21 shows the second derivative of the full data used in

the model.

102

9.5.0 wmmm232m><>>

-—: LLZ'9809
3— 98681709
-- 689'L909
-— 898'6909
—— 2901909
~— LBA'VLOQ
I 96172809
-- 608'0609
-— 8261609

:— LSQ'SOLQ
-- lSS'SLLQ
—— 990mg
—— GLA'BZLQ
:— caveats
.Z 20217149
—~ lZG'LSLQ
« 989'69L9

" 617819 L9

: 8909219
—- 6L6'9089

-‘* SSQ'SLZQ

   

 

y -- L617'06L9
:- 1-— 902'96L9

 

 

 

 

mN

0580; can ecu—Eam mo 3:082 me E comb San. 25m 5N v.5»:—

EONVQHOSGV

€28 mummzaz m>§>

mod.

 

i lLZ'9809
? 6891909

I
l
l
‘ i
l
1
l

L 6966909

1
l

r {:9 1.901909

103

 

 

 

      

  

 

m m m m m. w w. w
9 I. 8 6
.w. m .0 .1. z .9 z .0
I. .7 Z 6 8 0 .7
8 6 0 Z, v 9 6
l 9 6 8p 6 8 l
little-1.101313. . a u. i in E- n
. .x‘ ........ fl“ I'll-[Ill
[VII 1 “““
\\\
I
fl‘.\
1 \\\
/ \\
/t x
/l\

903'86 L9
6 l6'9089
8898 189

-1 v0.0-

+ mod.

: No.0.

 

... 5.?

o

u. l.” n It“: )D.l.lun..nil

.r 3.0

1. No.0

 

no.0

 

(.53 4.5“.— “.0 m>_._.<>_m_mo 0200mm

3.002 mam 5 now: «an 2: no 038330 vacuum 3N 0.53.”—

BONVBUOSEV

104

Determination of PLS Model Rank

As it was discussed in Chapter 2, one of the difficult tasks in PLS is determining the
correct number of loading factors to model the data. One of the most effective ways of
determining these factors and the predictive ability of the model is to calculate the
Prediction Residual Error Sum of Squares (PRESS) for every possible factor.193’194 The
PRESS is calculated by building a calibration model with a number of factors, then
predicting some samples of known concentration (usually the "leave one out cross
validation" method for PLS) against the model. 195 "96'197 The sum of the squared

difference between the predicted and known concentrations give the PRESS value for that

rnodel.l69,l98.l99

PRESS = gimp” — C,,)2 [25]

i=1 j=l

Where n is the number of samples in the training set, and m is the number of constituents.
Cp is the matrix of predicted sample concentrations from the model, and C is the matrix
of known concentrations of the samples. The smaller the PRESS value, the better the

model is able to predict the concentrations of the calibrated constituents.

Another useful tool in determining the prediction ability of the model and the number of

factors in PLS is the standard error of prediction SEP.

105

NI—

 

SEP = ‘=‘ [26]

 

 

where 6, is the predicted concentration of analyte in sample 1' with the use of a calibration

model that excludes sample 1', c, is the known concentration, and N is the number of

calibration sampleszoo’zm 202

Both SEP and PRESS values were calculated using the “leave one sample out” cross
validation technique for ethylene and l-octene separately. Figure 22 and Figure 23 show
the error prediction of the model (both SEP and PRESS) versus the number of loading
factors (principle components) for ethylene and l-octene models. As can be seen for the
ethylene model, the error prediction after the 3rd principle component of the variance
remains practically unchanged. Hence, this number was taking as the rank of the ethylene

model. In the same way, rank of 6 was chosen for the l—octene model

106

Figure 22: PLS Ethylene Model PRESS and SEP vs Number of Principal Component
Factors.

C2 MODEL ERRORS

 

 

 

 

 

 

 

 

 

 

 

0.025
. SEP
3 0.02 ----- Press Val
in
E 0.015~
O
z
< 0.01-
e.
6
0.005—
0 i """F""""'¥"":""%""f““
v- N (0 V In (D l\ CO 0) O
O O O 0 O O 0 O O *—
(L CL 0. a n. 11 CL 0. a. 8

NUMBER OF PRINCIPAL COMPONENTS

Figure 23: PLS l-Octene Model PRESS and SEP vs Number of Principal Component
Factors.

 

 

 

 

 

 

 

 

 

 

 

 

C8 MODEL ERRORS
0.1 .
0.09 SEP
-. 0'08 .. ------- Press Val
3 0.07 -» .
E 0.06 -.
O
3 0.05 -~
l 0.04 -
3 0.03-
0.02 ~-
0.01 .-
0 ‘1 r . 5 1'. _[ .— r _%
'- N CO V I!) (D N Q 0') O
O O O O O 0 O O O ‘-
0. 0. 0. 0. CL 0. CL 0. a. 8
W85? OF PRINCIPAL COM PONWS

 

 

 

107

Examination of Outliers

A number of different aspects in PLS can also shed light into the selection of the proper
rank for the model. Pirouette offers outlier plots. Monitoring the spread of outliers
within the concentration space is a good way of providing confidence in the model. If an
unusual spread of data is observed, it can indicate an improper rank choice or there might
be problems with the data which may require pretreatment. However, the latter is not
expected in this study since these are synthetic samples which were prepared based on an

experimental design.

Pirouette provides outlier plots showing the leverage of the sample against the residual
distances. Influential or high leverage samples are of particular interest in model
development. If a sample's profile differs greatly from the average training set profile, it
will have a great influence on the model, drawing the model closer to its location in factor
space. A sample's influence is quantified by its leverage, which represents a sample's
distance from the centroid of the training set. As the model size grows, the leverage

increases until all samples have equal influence on the model.

However, influential samples are not always outliers. If a sample lies a large distance
from the center of the training set because it has an extreme value of the dependent
variable, it contributes important information to the model. In examining outliers the
leverage is usually plotted against studentized residuals instead of Y-residuals since
studentized residuals takes leverage into account. The studentized residuals weigh the

value of the residuals by taking into account the greater precision near the middle of the

108

data set versus the extremes. The studentized residuals tend to be distributed according to
the student t distribution. Therefore, when eliminating outliers a good reason associated
with the outlier needs to be present. Only three samples were eliminated from this model.
These were the samples with very low ethylene concentration in which the errors
involved with their preparation were significant due to the lower limit of the balance
used. Measurement of a small amount of ethylene (less than 0.5 g) was not accurate

enough.

Figure 24 and Figure 25 show the outliers for ethylene and the l-octene models with no
apparent outliers. Two samples in the ethylene model and one sample in the l-octene
model appearing far out on the leverage axis are not outliers. They are the samples with

extreme concentrations.

STU RESIDUAL

STU RESIDUAL

109

Figure 24: Ethylene PLS Model Outliers

C2 MODEL OUTLIERS

 

:0
‘M 1 . i i

 

. ' t l
0... was 9 0.1 : .015 0.2 0.25 0.3

6.
-‘” O

900

O

 

0.35

0

 

 

LEVERAGE

Figure 25: l-Octene PLS Model Outliers

CB MODEL OUTLIERS

 

 

9
¢
J
"I'

0.2 0.3 0.4 0.5

0.6

 

 

LEVERAGE

 

110

Examination of Spectral Residuals

Examining the spectral residuals also provides information about the model. For
instance, if the spectral residuals present a spectrum looking structure it indicates
problems with the model, since they should look like random noise if the model is
adequate.

Figure 26 and Figure 27 show the enlarged spectral residuals with no particular spectra

structure.

F>CU~VJM7~5

111

Figure 26: Ethylene PLS model spectral Residuals

C2 MODEL RESIDUALS

 

10 20 30 40 50

Variable Number

mF>CIUHmFJW

112

Figure 27: l—Octene PLS Model Spectral Residuals

C8 MODEL RESIDUALS

 

10 20 30 40 50
Variable Number

113

Model Prediction Errors

Obtaining the correlation between the calculated versus measured concentrations is the
ultimate indication of the model's predictive ability. Usually measured values (actual
values) are plotted against the difference between the predicted and the measured
(predicted-measured) values. These plots are given for ethylene and l-octene in Figure

28 and Figure 29.

As it can be clearly seen from these plots, the absolute errors in the PLS model for both
ethylene and l-octene were less than i 0.5 wt %., and :I: 0.75 wt % (absolute)

respectively.

As a result the predictions of both (ethylene and 1-octene) models were found satisfactory
when tested using the training set (cross validation). PLS will be a robust model for the
real-life samples as well since it can tolerate the existence of impurities and/or changes in
the Isobar E composition. Additionally, since PLS includes the temperature and density

effects in the model, it will not necessitate the measurement of the sample temperature.

114

Figure 28: PLS Fit Errors for Ethylene Over All Temperatures.

 

 

MEASURED- PREDICTE

ETHYLENE MODEL PREDICTION ERRORS

 

 

 

 

 

0.80%
O
0.60%
o
, o
0.40%
o
3 O O
o
0.200/0 .
e o
,9 o 0
9’ o .
0.00% 0 . fi' 6 t .9 t i I
0.00%.. 2.00% ’ 04.00% 6.00% 8.00% 100090 12.00%
8 O . . o 9
-o.20% ~- .0 09
0 ° .
o o 9’
-0.40% 4- ..
o
-0.60% ..
o
-0.80%

M EASLBEJ CONCENTRATIONS WT %

 

 

115

Figure 29: PLS Fit Errors for l—Octene Over All Temperatures.

 

 

MEASURED-PREDICTE

1-OCTENE MODEL PREDICTION ERRORS

0.60% -
0.5070 0
0.40%
o
’ o
0.30% . .
0.20% 9 . o
o
0.10% , o ’
3 O :0
0.0070 T i v 9 —IL I i
0. 76‘ 5.00%: 10.00% 15.00% 3 20.00% 25.00% 30.00% 35.00%
-O.10°/o Q
: 9 O 9
'0.20°/o O O ‘
-0.30% , .

 

 

 

 

M EASLIE) CONCBTI'RATIOINS WT %

 

 

116

Summary

Determination of light alkenes via fiber optic FT-NIR spectroscopy at elevated
temperature (140 °C) and pressure (500 psi) has been demonstrated. The effects of
elevated pressures on spectra were found to be insignificant for these compounds.
However, the temperature effects are critical, especially for the lighter molecules such as
ethylene and propylene. Temperature effects on spectra decrease with increasing chain
length among the light members of monoalkylated alkenes. Calibration models were
constructed incorporating the temperature effects. Calibration standards were prepared
and spectra obtained at elevated temperatures and pressures. An apparatus for this
purpose was built and used for the collection of spectra. The determination of ethylene
and 1-octene was used as the example system. Partial least-squares regression was
investigated for both ethylene and 1-octene. The calibration covered the range of about 1-
13 wt % ethylene, and 2-30 wt % l-octene over the temperature range of 25 to 140 °C.
Absolute prediction errors are expected to be less than i- 0.75 wt % for both ethylene and

1-octene.

CHAPTER 5

CONCLUSIONS AND RECOMNIENDATIONS FOR FUTURE RESEARCH

A viable method has been developed to achieve the simultaneous determination of
ethylene and monoalkylated light alkenes at elevated temperatures and pressures via fiber
optic Fourier Transform Near Infrared (FT -NIR) spectroscopy for in situ, real-time

monitoring of these compounds.

An additional goal of the study, to evaluate the effects of molecular structure of the
monoalkylated light alkenes on this determination, has also been accomplished. In
addition, the effects of elevated temperature and pressure on the vibrational spectroscopy

of the light alkenes in the NIR region were investigated.

Conclusions

By far the most important result of this investigation is the development of a method for
the simultaneous determination of ethylene and monoalkylated light alkenes at elevated
temperatures and pressures via FT-NIR spectroscopy for in situ,, real-time monitoring of
these compounds. Also, the effects of molecular structure on this determination have
been shown along with the effects of elevated temperatures and pressures. Other

conclusions drawn from this study are summarized below:

117

118

. NIR spectroscopy is a viable tool for in situ,, real-time monitoring of systems both in
the laboratory environment and in process analysis.

. Coupling NIR spectroscopy with fiber optic probes enables the in situ, real-time
monitoring of systems especially for process analysis applications.

. FT-NIR spectroscopy is more stable and a more reliable tool than conventional NIR
spectroscopy.

. NIR has many advantages over MIR for in situ,, real-time monitoring.

. The first overtone of the asymmetric stretch of =CH2 is unique to ethylene and

monoalkylated light alkenes. Thus, the simultaneous determination of several light
alkenes in presence of di and tri alkylated alkenes is possible in the NIR region.

. The first overtone of the asymmetric stretch of =CH2 for ethylene and the
monoalkylated light alkenes shifts towards lower energy with an increasing number of
carbon atoms in the molecule. This shift is significant enough to allow for the

simultaneous determination of these compounds in the NIR region.

. The intensity of the first overtone of the asymmetric stretch of =CH2 for

monoalkylated light alkenes decreases with increasing molecular weight.

. Simultaneous determinations of ethylene and monoalkylated light alkenes are more
feasible for those compounds with the biggest shift difference between them. For
instance, the determination of ethylene and any other monoalkylated light alkene in a
mixture is possible. However, determination of 1-pentene through l-decene in a

mixture may be difficult with this technique due to the spectral overlap.

9.

10.

11.

12.

13.

14.

15.

16.

119

Additional features other than the first overtone of the asymmetric stretch of =CH2
(use of different regions of NIR spectrum such as 4,800 cm'l region) can be an aid for
determination of these compounds in presence of interferences such as that from 2-
ethyl- l-hexene in this determination.

Multivariate data analysis is necessary in NIR studies.

Classical least square regression is an adequate tool to use as a multivariate data
analysis tool in NIR spectroscopy. Its simplicity makes it the preferred tool for
qualitative laboratory feasibility studies.

Pressure in the range used in this study (less than 750 psi) has no effect on vibrational
spectroscopy of the light alkenes in terms of absorption intensity, band shape changes,
or the vibrational frequency.

Temperature effects are significant for the gas phase molecules or their solutions in
liquids. The smaller members of the light alkenes (ethylene, propylene, and l-butene)
are more influenced than the heavier members of the light alkenes. Within the
temperature range (25-140°C) implemented in this study the effects of temperature
are enough to influence the determination of light alkenes at elevated temperatures.
Temperature effects on the first overtone of the asymmetric stretch of the =CH2
decrease with increasing molecular weight among the light alkenes.

Temperature effects in the range studied (25-140°C) for the light alkenes are in the
form of decreased intensity and broadened absorption bands for the first overtone of
the asymmetric stretch of the =CH2.

The effects of temperature on the intensity of the absorption band for the first

overtone of the asymmetric stretch of =CH2 can be explained by the population

120

changes in the energy levels. With increased temperature the population of the
molecules at ground vibrational energy level decreases since some of the molecules
are elevated to the first vibrational energy level.

17. The effects of temperature on the shape of the absorption band for the first overtone
of the asymmetric stretch of =CH2 can be explained by a change in the energy
distribution among the rotational energy levels which are located between the
vibrational energy levels. Since rotation vibrational energy levels are also excited by
the radiation. A similar redistribution of the population of the rotational energy levels
will influences the shape of the band.

18. Partial least squares regression can be used for modeling the training set for the
simultaneous determination of light alkenes in NIR region. The absolute errors

predicted by this model are less than i 0.75 wt % for ethylene and l-octene

Recommendations for Future Research

The feasibility of simultaneous determination of light alkenes in situ,, real-time at
elevated temperatures and pressures opens several areas of investigation. First of all via
this technique even higher temperatures and pressures can be studied to determine their
effects on the vibrational spectroscopy of these molecules. Secondly, ethylene as a
supercritical fluid at the elevated temperatures and pressure can be investigated in terms

of its supercritical properties such as its solvation power or its intermolecular interactions.

121

The development of this type of in situ, real-time analysis technique via fiber optic probes
can be implemented in other systems with distinguishable features in NIR. For instance
the similar hydrocarbons (paraffinic, olefinic, aromatic) can be investigated. Expansion
of this determination to simultaneous determination of similar compounds along with
light alkenes, such as the determination of di or trialkylated light alkenes should also be

feasible.

The possibilities of using this tool in process analysis in similar applications for process
optimization and control will benefit the control of manufacturing chemicals in industry.
Thus, further research should seek to implement NIR spectroscopy for process analysis of

many other compounds.

 

APPENDICES

APPENDIX A

APPENDIX A

This appendix contains the FF-NIR spectra (2,500-12,000 cm'l) of ethylene, mono
alkylated alkynes, cis and trans dialkylated alkenes, trialkylated alkenes, 1,1 dialkylated
alkenes, n-octane and the Isopar E used in this study. All of the spectra were obtained
with 4 cm'1 resolution and with a 5 mm long pathlength at room temperature and

pressure.

122

123

LIST OF SPECTRA

Figure A-30: FT-NIR Spectrum (4,000 12,000 cm-l)of Ethylene in Isopar E Solution
Figure A-31: FF-NIR Spectrum (4,000 — 12,000 cm-1)of Propylene in Isopar E Solution
Figure A-32: FT-NIR Spectrum ( 4,000 - 12,000 cm—1)of l-Butene in Isopar E Solution
Figure A-33: FT-NIR Spectrum (4,000 - 12,000 cm—l) of l-Pentene 99.9 % pure

Figure A-34: FI-NIR Spectrum (2,500— 12,000 cm-l) of 1-Hexene 99.9 % pure

Figure A-35: FT-NIR Spectrum (2,500- 12,000 cm— 1) of l-Octene 99.9 % pure

Figure A-36: FT—NIR Spectrum (2,500- 12,000 cm-l) of 1-Decene 99.5 % pure

Figure A-37: FT-NIR Spectrum (2,500— 12,000 cm-l) of Trans—2—Octene 98 % pure
Figure A-38: FT-NIR Spectrum (2,500- 12,000 cm-l) of Trans-3-Octene 99 % pure
Figure A-39: FT-NIR Spectrum ( 4,000 - 12,000 cm-l) of Trans-4-Octene 99 % pure
Figure A-40: FT-NIR Spectrum ( 2,500 — 12,000 cm-l) of Cis-2-Octene 96 % pure
Figure A-4l: Ff—NIR Spectrum ( 2,500 - 12,000 cm-l) of Cis-3-Octene 96 % pure
Figure A-42: FT-NIR Spectrum ( 2,500 - 12,000 cm-l) of Ciss-4-Octene 98 % pure
Figure A-43: FT-NIR Spectrum ( 2,500 - 12,000 cm-l) of Cis—3-Methyl 3-Heptene 98 %
pure

Figure A-44: FT-NIR Spectrum ( 2,500 - 12,000 cm-l) of Trans-3-Methyl 3-Heptene 98
% pure

Figure A-45: FT-NIR Spectrum ( 2,500 - 12,000 cm-l) of Isopar E

Figure A-46: FT -NIR Spectrum ( 2,500 - 12,000 cm-l) of n-Octane 99 % pure

Figure A-47: FT-NIR Spectrum ( 2,500 - 12,000 cm-l) of 2-Ethy1—1-Hexene 99 % pure

 

 

124

 

 

8:28 m 982 a 225m its 25.2 Sci 9.58% 524.2 .84. 6.5»:—

 

125

 

8:38 a 238— ... 26.22.. «2.5 98.2 . 25.3 .5586 «2.: :3. 26mm

 

 

:.

L
2
w

m

1 26
1:35“

..w

 

Egan—em H 538— E 2.8.5; “21:5 22.3“ . 256 v Baboon—m %.,—h “NM < 0.3mm

 

 

127

 

 

2.... as 5.5 225: 5 :56 25.5 . 25.5 5.58% 22.2.2 63. 2.55

 

 

FT-NIR Spectrum (2,500- 12,000 cm‘l) of l-Hexene 99.9 % pure

0
-

Figure A-34

 

 

 

129

 

 

6.5.. as 35 2.35; 5 26...... 25.2 25.8 52.8% 22.2..— 53. 2.52

130

 

 

 

2. mg 6582.2 .5 p.86 25.2 25.5 52.8% 22 .E 53. 252

 

131

 

 

2.... 2. .5 65.5.2222. 5 26.5 25.2 25.5 5.58% 22.: 23. 252

 

 

 

132

 

2.... 66 a 65.55.22... 5 29.5 25.2 25.8 9.58.5 22.2.... 53. 2.52

 

 

52:. .
. ..........<...

 

133

 

9:5 6.5 a 02899....522B we 9.25 aqua . :25; 2.2.5925 Hula—rm an)» 23E

 

 

134

 

 

2.5 as 5 2650.38 5 26.5 25.2 . 25.u 5 5.5855 Ez E 53. 2.5.5

FT NIR Spectrum ( 2,500 - 12,000 cm") of Cls-3-0ctene 96 % pure

Figure A-4l

135

 

.«y
.1:

.»x t
.«

 

 

 

 

136

 

2.... as 5 2.38-926 5 26:6 25.2 . 25." V 55855 Ez .E u2..< 2.52

137

 

 

222 .2. no oEfi—oflé E:~o§.m.£0 no 9.59 25.2 .. 8m.n V 25.5925 Mid—h ”91" 85mm"—

138

(2:

.. .. .4. ‘ .V
_. g 9“.»

 

 

2.... 95 a Suaié .282 was; .... A15 25.2 . 83 V 52.8% Ez ,E «I» 0.5mm

 

 

A: E 2, 31.3.5. 2,73,3“V

 

n

m
w
t

w

x

139

» '4!

 

m 5.8— % $5 25.2 . 83 V 5.58% «Eur.— nmv< 25mm

 

140

 

 

9:5 § as 2380.: ac A159 Exams .. Sm." v 5.58.5 mnzéh “31‘ 253m

 

141

9::—

§ 3 2.86:.

 

vixen .... $5 25.2 . Sm.” V 5.58% m

 

EE— §1< 2:me

 

 

APPENDIX B

APPENDIX B

This section contains the excel spread sheets showing the calculations of the liquid phase

concentration for each run.

142

Table B- 1:

Table B-2:

Table B-3:

Table B—4:

Table B-5:

Table B-6:

Table B-7:

Table B-8‘:

143

LIST OF TABLES

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # O7

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 08A

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 08B

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # O9

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 10

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 11

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 12

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 13

Table B-9: Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 14

Table B-lO:

Table B-1 1:

Table B-12:

Table B-13:

Table B-l4:

Table B-lS:

Table B-l6:

Table B-17:

Table B-18:

Table B- 19:

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 15

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 17

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 18

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 2]

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 22

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 23

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 24

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 25

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 27

Excel Spread Sheet Showing the Liquid Phase Concentration Calculations for Run # 28

144

 

 

 

 

 

 

 

22.8. F 2883 2.8 F .25» as 8.8“ M23? dam;
«838 2283 8.0 3.2 .283 m «View.

o o o o o 8 o :3 8
28.3 2226 $820 5m 0582 8 o 3.2 $8.30
5.32 402 2:05. a .32. «Es 25 m .352

28

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. 05.8% .0 \ 2:29. 32.: :2. .No 2: :0 2:32: 3:22 c2328 an: 2: 3 22m :15»: a: 322:: 82.: :2. 2 No :o «22: :o a 2: 2 .5: «o :0:§

 

05.3%.: 1.2. 82:12:. . 22:02 2: :0 d2: .32 5:230 23. 2: :o 22» us»: an 322:0 023: :2 S .2302 :0 .22: :0 : 2: 2 40.: :<:o2 :o:<>

 

9:50:2222223 32:: :2. . 00 2: :0 62: 5:2 .223? 23. 2: 3 22a :2»: A: 322:: 323‘ :2. 2 2.23.: :0 922: :o a 2: 2 .53 :0 :O:§

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.6:: 22V: 2: :0 2:22.- 382. 2: B 2:29. 22 "$23: 82:: 3:2. 2: :0 $5.5:

22% a: 32%: use: Earle 2: 3 82: 32 ... :2: 82: 2a: 2: :o 2:39.

Dbl: u: 5:2. o 2: a: :22332 2:29. 222: 22 2: A: 32:2: 32: 222: 2: N: 522:0 2:32 2: :o tfizmn

5:28: 2: A: :223: 32: 222: 2:26:88 :23 2: :0 =5» 59: 322:0 «2302 60.92 2:30:53 2: :o 2:29. :22: 2: :0 Sam

33:3: :0 2:32: 3:32-80 :0 2:32: 222.3122: 2 23:0 2 2 No :0 2:32: Eta: 2: 2 32.1 am: «0

23:2 :9 20:92:02.~m2::o2 :0 :i n 98:02 :0 2:82: 22:: 2: 2 2.36: 23:02

:38 5.82 823 8 _ 8 3 328235.90 3 2V... 8 28582: 2:2: as 2 =5. .0: 8

2.0.08 5.3: 803.0 m :<:02 :0 2 22:5» b22223 2:32 :o 2:32: 3:2 2: 2 2.52:: 53:02

0 m < "a. 2.23.. Ea 22.8. .2 .o .22 9:55.28 mz_o»z<_ _ _ _ _

«02.292353 2 9.2288 2:22.00 222.3 :2»: A: :22:2uo 2:33:52 :29: .8 2: .u :0 :o 2:32: 3:2. 2: 2 3:1 C: :0

_ _ _ _ _ _ _ 2:: :0 2:5 2 2:32: :20230 2 2:: M:

2:93 :o 32.22. 2223:: 32:23: 2:32: 2: so 5:223: 2: 2 2:230: 2:32: :2823 2 Sm: .: :moo

_ _ 222322 .50: :8 :25 2: 2 5.2:.

282.32. 2:32: 2808 2: :0 32:2 2: 2 :03»:

32522. 2:82: :2: 2: :0 3282 2: 2 9:5.“
00:00.0 00 50.0 km. F0 m9q — .0 080.0 mdmp and: mm 5.0 0000.0 095.0 5.3 0.55 v0.00— mw. EN
00:00.0 00 50.0 mm. to 53.0 030.0 mdm— mmdp 35.0 0000.0 02.0.0 5.0— QRN 8.00— mm. 5N
00000.0 00 50.0 09 2.0 53.0 080.0 909 5.9 005.0 0000.0 «v5.0 00.0? v.08 00.00.. 0508
530.0 00 50.0 09 —.0 000 —.0 55.0 —.mm. 0.5.. 008.0 0000.0 080.0 Qum— fi 5w 5.5 50.va
530.0 00 50.0 00. 2.0 000 F .0 85.0 fimm— 0«.? «08.0 0000.0 830 Sam. m. 5w 5.5 0v.vmm
0300.0 00 50.0 09 2.0 000 «.0 35.0 «.09 L99 38.0 0000.0 93.0 may. «.08 0.2.5: 5.00m

.20.: s 40.: > .52 > 41$ a .3 «EH > 2:: 6: =2: .0: 2.: .o: 2...: .:> 2.; .:> 2.2.-‘1: Sm: .: 5MB. :62: 9:12.:

:0 $202 8 $5.5: MEI: 40> \CEZIMQ Sam «0 5:02 00 :<:02 :0 ::00
L053 20:3. 50:3.— 52::—

 

 

 

 

 

 

 

 

5 :2 :3: :0: 2.0025250 55:50:00 023: 2:03 20 22328 Beam 085m 30%: ”Tm 2an

145

 

3:0: 5 2:209: :0 50:02: :2: 2:: 2 k 40:: 00 922.:

 

 

030: s 08:50 :0 88:2: 30:0: 2:: 0: 5.: 8 0:00:

 

030:: 5 20:8: :0 3080: 80:0: 2:: 0: .2: $800: 080:

 

 

 

 

 

 

 

 

 

 

 

030:: s 20:00.: :0 8020: 30:0: 2:: 0: 2:: 8 0:0:

 

0:03.50 :0 30:02 88020:: 2: 5:: 002:: 2:0: 2: :: 0:0: 50 :0 020:: :0 82:5: 05 01::

 

:52: :0 005030 002:: 030:: 5 32:50 :0 30:2. 2:: 0: 0 8 050:

 

20:02 :0 50:03 88022: 05 5:: 002:: 2:0: 2:: 5 20:02 :0 022: :0 8.5:: 05 0:00:52: :0 85200 002:: 2:0: :: 20:02 :0 50:2: 0:: 0: 0 $500: :50:

 

0:200 :0 30:2: 85020:: 05 5:: 002:: 2:0: 05 :: 08:00.: :0 00:2: :0 805:: 2: 0:00:52: :0 85200 002:: 2:0: 5 08:00.: :0 50:2: 05 2 0 00 9:0:

 

 

NO :0 00:2: :0 : :20: 2: 50:: 002:: 8:00 2: :: 0:03.50 :0 020:: :0 0. 05 05:02:80 N: 85030 002:: 2:0: 5 0:00:50 :0 022: :0 ... 2: 2 40:: No :30:

 

20:02 :0 020:: :0 ... :20: 05 50:: 82:30:00 2: :: 20:02 :0 020:: :0 a 05 05:02:58 :3 :05200 002:: 2:0: :: 98:02 :0 020:: :0 ... 2: 2 :03: $502 :30:

 

0:0:00 :0 022: :0 828:

:20: 05 50:: 000:: 8:00 2: :: 08:00.: :0 00:2: :0 805:: 05 05:02:80 :0 005200 002:: 2:0: 5 08:00.: :0 022: :0 805:: 2: 2 .53: :0 050::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmdw mmd 5: p .0 mowwd 20mm: {ammuma 5.0 a —.N0 mm: 9.0 500.0
00.2 00.0 :28 020.0 0.00... .8000 S... 2.00 23.0 68.0
:02 00.0 023 22.0 0.8.: .280 8.0 2.00 83.0 88.0
3.2 2.0 82.0 a .8 .08.: 28.8 0:. 2.00 23.0 800.0
3.2 2.0 82.0 28 28... 28.8 3... 800 5.8 800.0
00.9. 0N6 :0 w .0 mm 5.0 2000.0 o\oow.m0 m ....w wNNo 0.3. v.0 mlwmwd
:2 .0: :8. .0: :2 0: ..: 00: ... .8: n: 00: 02: E .2: 0 .: 0 .: 0.: 00:: 00:: 00::
8 03:00: 8 8 5.000: 00 8 2800: 8 8 03:00: 8 8 2800: 8
030: 2:03 0.80: R00: 2:0: 050: 0:00: 2:0: 2:0: 0:00: 0I::0: :50:

 

00 t :3: :0: 000020200 00000000000 08:: 2:05 2: 9:325 805m 082$ 38:: ”3:00: Tm 030.0

 

146

 

_

_

 

 

 

 

 

«aw—N: _. N8 35d Nfim — 442.0... 4.: oudmu w2340> Amwwm>

080...“: mgmvmd 5980.0 cm amwmomd m m<:0w_

o o o o o 00 0 N4 N0

vans—0.. :monmod mkwvod N; web—dd NO d on :<:Ow_:m0
20:05:: .52 wMJOE d .mwSz .x....>> 29:5 9 .wm<2,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. $5.003. .0 \ «2:22. 32:: .:s. .NO 2: :0 2:32: :2::::: :0::::0 3m 2: :0 0:20 :50: A: 32:30 32:: .:s. 5 N0 :0 020:: :0 ... 2: 2 20:: «0 05:3.

 

0:530:20 1:0: 32:: :2. . 2:02 2: :0 .02: :2: :0::::0 3m 2: :0 22019:»: A: :05030 32:: .:s. .: 20:02 :0 3:2: :0 ... 2: 2 40:: :<:02 :Q:<>

 

05.83.2222? 32:: .:s. . no 2: :0 3:: 3:: :0:0::0 30. 2: :0 22.350: :0 352:0 32:: .:s. S 2200.: :0 3:2: :0 :: 2:2 .53: :0 :O:§

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

32:: :32: 2: :0 2:22.. :32. 2: :0 2:29. :20: u «:23: 32:: 3:2. 2: :0 2:320:
22% a: 323: :32. :ato 2:: :o 8.5 :50: u :23: 82:: :30: 2:: :o 2:32.
D93: u: :0:0: 0 2: :0 3:22:00: 2:22. :22: :20: 2: A: 3:30: 32: :22: 2: :0 :02200 222.: 2: :0 Ptmzmn
23:0: 2: :3 3:30: 32: :22: 25:02:00 :03 2: :0 2:» 20:: 352:0 @6202 60.8: 2:262:00 2: :0 2:29. :22: 2: :0 :3»
«2.302 :0 2:32: 2:02.30 :0 2:32: 2:02:22: 0: 23:0 2 :: No :0 2:32: Etc: 2: 2 =5: .0: N0
.2302 :0 :0:02::0E.~m:::02 :0 :>: n 20:02 :0 2:32fl2ta: 2: 2 2.220: m:<:o2
:02: 8.82 880.: 00 :0 :0 8202:0530 :0 2:" :0 32:82: it»: 2:: 2 =5: .0: :0
2.0.3“ 5.3: 800.3 :<:o2 00 0: 3:23 3:23:00 96:02 :0 2:32: 3:2. 2: 2 3:. «1:: m:<:o2
0 m < a... 228-. :5 mags. 3: .0 .25 32.25200 .265: d _ _
~0+:.:xm.<n~ m0: 0: :22002: 2:22:00 2:25: :20: .3 3:030:00 22202.6: :20: :3 2: :u 00 :0 2:32: 3:2 2: 2 2.; .:> :0
_ _ _ _ _ 2:0 :0 2E: :: 2:32: :20230flH—v4“
E:mv:: :0 30:22: 2.22.2: :00::0:2: 2:32: 2: :0 :0:20::0 2: 0: 3:20:00 2:32: 3:02:00 2 Sm: .: 5:00
_ _ 05:20:20: :50: :3 :26 2: 2 Eu»:
333:2: 2:32: ::003 2: :0 50:02 2: 2 :63
33:22: 2:32: :2: 2: :0 30:3: 2: 2 9:22.:
mvmmod 330.0 B: :d Ev :d 55d «.00: and 068.— 38.: 003.: end— mfim: 3.59 02mm—
vamod 830.0 :m: :d F S. :d om Ed «.00: 0nd new”: 0000.: momma: cud: msm: 00.8— 02mm:
mvmmod 830.0 E: :d F 3: :d 5 5d Ndm: omd mmmmé among 83.: Ed: m.:.m: 3.an 02mm—
mo 50d mvmood wow : d can :d N 36.0 08.. 5d mmmod mused nvond mad m. 5: 8.8: Edo—
mwpmod memood Nowpd 000 :d N 30.0 0mm: 5.0 800.0 92.0.0 medhd mad m. 8: dude: undo—
00500 3.80.0 «02.0 000 ..0 0 30.0 0.00. 8.0 2.8.0 350.0 0860 3.0 m. 8: 3.00: 55.00:
0580.0 0:: Edd mvw F d m :w :d 0 Food as: end nomad Bmmd as Ed m :5 mid: 8.5 3.3
033.0 0: Sod mvw : d 0 5 :d 0 Emd 0N: nod 008.0 «08.0 3 5d a Pd 0. :0 no.5 2.:
530.0 mm Sod mvm : .0 m Pm : .0 P 50.0 QR: and nomad mmvmd mm 5.0 00.: odd: 92.: mod:
:03: > .53: > .53: > .: :S ..: :3 45mm > 3.5: .0: 3:.< :0: E<:0F: 2.5: .:> 3:: W: E< .: 3%: Nn: :MMFE :03: 3:33.:
«0 5:02 :0 mi: :0: m2b20> >226: 3.8% No :<:o2 8 :<:o2 :0 5:00
:0: S. :O:§ 50:5. «60:23:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

<wo .... 00m :0: 32.20030 0032.08.80 03:: 2:21: 2:: 2:325 Beam 302$ 38:: ”Wm 030:.

 

 

 

 

147

 

 

 

 

 

 

 

 

 

 

 

20000:: 020002: 2: 0: 00:00:00 :03... 00:00: 0:... 0o .00 0.: :0 02000203202: 2: :0 E00 0.: 0: Sb: .0: 0.30

 

_

_

00 :0 00.02.38.000 :0 .....n «0 .8532: :25: 2:. 0: s: 0 .0: «0

 

0.00%»: :0 00:00:: 3.0.0.008. :0 00:20:00: :0 20002: 0...... 2:. 0: E0 .0: 00:00:

 

00 :0 0008:0800 :0 min 00 .0283... 0:00.: 2:. 0: 3:0 .0: 00

 

03¢: 0: 2:00:50 :0 02:00:: :02: 2:: 0: n: :03 N0 050::

 

030:: C: 9.2000: :0 20:30:: :2: 2:: 0: ...: 40:: 01109030:

 

0:02: :: 2.800.: :0 :0:00:: :2: 2:: 0: ..: :03: :0 050::

 

0:00.: ... 80.0.... :0 .08.... 30.2.. 2:. 0: 0.... 00 0:00...

 

0.00:: c: 0.008: :0 .5223000: 2:. 0: 00.: 00:00: 0:00...

 

 

 

 

 

 

 

 

 

 

 

0:00.: 0: 000.00.. :0 .0022: 30:9: 0... 0: 0.... 00 0:00...

 

0:03.:0 :0 3E0... 0:000:08 0.: 5:0. 000:: 0:00: 2:: :: 2:0: .:0 :0 00:2: :0 32:0: 2: 0::

 

00.3... .... 0050.00 002:: 0.0:... ... 000.050 :0 .00.... 2:. 0: 0. 00 0:00...

 

0:00:00: :0 330:: 3:000:08 2: $0: 002:: 0:00: 2: ...: 0:00:00: :0 020:: :0 :00E0: 0.: 0:32:08 :0 005030 000...: 0:00: :: 0.2000: :0 30:2: 2:: 0: m 03:00.: 050:

 

0:200 :0 30:00: 0:000:08 2: 5:0. 000...: 0:06.: 2: :: 0:209: :0 00:2: :0 :0050: 2: 0:30:38 0:: 00::030 000:1: 0:00.: :: 0:209: :0 30:00. 0.: 0: m 00 0:09.:

 

 

No :0 00:2: :0 ... :20: 2: ES: 002:: 3:00 2: :: 0:00:50 :0 00:2: :0 ... 2: 05.02300 :0 005030 000:: 0.00:: :: 0:00:50 :0 00:2: :0 0 2: 0: :03: «o 0:02.:

 

0:00:00: :0 00:2: :0 0 :03. 2: E0: 003%.?) 2: :: 0:00:00: :0 00:2: :0 ... 2: 05.02300 :3 005030 000...: 0:00: :: 0:030: :0 00:2: :0 ... 2: 0: :03: 03:00: 0:02.:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000:00 :0 00:05 :0 .0050: :0:0: 0::: 50:: 002%: :0Q0) 0C: 0: 0:0:004 :0 00:02.: :0 :00800 0:: “0200:2200 NO 000:0:00 80km 0:33: 0: 00300.: :0 00:02.: :0 :00500 00: 0: 40: GO 0:301:
000 00.. .300 00000 0.000 000.00 00.0 00.00 00.00 0000.0
00.0 00.. ... .00 00000 0.00 0.00.00 00.0 00.00 00.00 0000.0
00.0 00.. 3.00 0000.0 00.00 0000.00 000 00.00 00.0.0 0000.0
00.0 :00 00.00 00000 0.000 0000.00 .00 0000 00.00 00000
00.0 :00 00 .00 0.000 0.000 0.00.00 .00 0.00 00.0.0 .0000
00.0 .00 00.00 0.000 0000 0.00.00 .00 0.00 00.0.0 .0000
00.0 00.0 .0 .00 00000 0.000 000000 00.0 00.00 00.00 0.000
00.0 00.0 .0000 00.00 0000 0.00.00 00.0 00.00 00.00 0000.0
2.0 00.0 I 0000.0 00000 L! 0000 09.0.00 00.0 00.00 00.00 0.000

:5: 00.: 0.00%.... 0...... .0: u. 00:. ... .0... u: ..0... 0E. .20. 0...... d .. 0 .. 0: 00:: 00:: 00:.
00 00:00: 00 «0 00:00: 00 «0 03:00: 00 00 00:00: 00 «0 00:00: 00
0.000... P00... 0.00... 0.000... 0i.00:.. 0l.00:.. 0:00... 0.00... #00... S0F0... E00... 0.00...

 

<8 :0 =3: :0: 0.8003200 02.00.5050 000.5 2005 0.: 9:326 .025 002% _00xm ”3.50: NA: 030:.

 

 

 

148

 

 

_

 

 

 

 

 

 

 

 

.280 00 2.0.0 88.. 0.8.0 8.0. .088. 00.0. ._e_ 8.80 0539 000000
8.02.... 08m? 00 {8.00 85.0 «80.0 8.2 0.8.8 00000. _
3.08.... 88.0 05.00. 080.0 080.0 080.0 8.0 .v08... 00 8.0. 8.0. 00
0 0 030.2 88.0 8.0.0 8.2.08.0. 00 8.2 0.00 0000.00
.0+<n0 8.808 0080 2.. .2 20:00.0... .0... 0.300. 0 .00.... at: .00: 0 .0000.

 

_

_

 

_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.0.0000... \ 300.00 00000 .000 .00 0... .0 0.00020 .0...00. 02.0000 000 0... .0 0.80 00.00 0.. 000.800 00000 .000 0. No .0 00.00. .0 0 0... 0. .0... No 0003.

 

«......00006130 00000 .000 . .0000. 0... .0 .020 ..00. 0000000 000 0... .0 0.0.0 00.00 00 000.800 00000 .000 0. b.0000. .0 00.0.0 .0 0 0... 0. ..03. 03.00. 0000.0

 

......0000.0\~0E0.00 00000 .000 . 0o 0... .0 .020 ..0

. 0000000 00.. 0... .0 0.80 00.00 0.0 000.800 00000 .000 0. 000.00.. .0 00.0.0 .0 0 0... 0. ..os. 00 0003.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ a 00000 0.0.... 0... .0 00.0.00- .00000 0.... .0 0.00.00 .80. n «.8... 00000 .0000 0... .0 2.5.00
_ _ 0.0000 .3 000.000 0.0.... .0593 0... .0 000.0 .80. u 08.... 0000.0 0.00... 0... .0 $0.30..
90+... no 08.00 0 0... 0... 00.0.0800. 00.0.00 .800. .80. 0... 0... 000.000 0000. .820 0... .3 000.800 20.0.0. 0... .0 EME
5.0000 0.. 00 000.000 0000. .800. 800000.000 0000 0... .0 0.00 0.0.. 000.800 3.0000. 60.00. 800000.000 0... .0 0.00.00 .800. 0... .0 :30
3.0000. .0 0.00020 0.000200 .0 0.00020 .0.t00...0.0.0 0. .0000 0... No .0 0.00020 .8000 00. 0. 8.... .00 «0
0.0000. .0 00..00...00.. 3.0000. .0 00. u m.0000. .0 200020 .0...00 0... 0. 3.0.1.00 m0<0om..
20.0.0 .080. 0800.0 00 . 00 .0 00.00.005.00 .0 00.u 00 30.00020 .0000 2.. 0. 0.... .00 00
03.0.00 5.00: 0080.0 03000. 00 0. .0..E.0 08800.00 0.00000. .0 0.00020 .0000 0... 0. ...... (W0 £03.00.

0 0 L". "o... 2.200.. 0:0 0.2.8. .o. .0 .25 00230200 02.0.2.0 _ _ _ .
~0+L.\m-<u~0.00.. 0. 0.0.0.0000 0.00.0000 00.0.0... 00.00 .3 08800.00 20.0.0050. .0.00 .00 0... .0 00 .0 0.00020 .0000 0... 0. 3...... .00 00
A ~ _ H _ _ _ ..0.0 .0 8.00 0. 0.00020 00.02.00 0. 0...... .0
2.000.... .0 3000.8. «0300"... .0000000.. 0.00020 0... .0 00028.00 0... 0. 06.0.0000 20000.0 00.02.00 0. 500 .0 0000
A J: _ . 0.0.0.0050. .0.00 .00 0000 0... 0. ENE.
.0000000.. 200020 000000 0... .0 .00000. 0... 0. 000.0
.0000002. 200020 .0... 0... .0 .00000. 0... 0. 00.2.0.0!
0080.0. 0080.0 008.0 0000.0 0000.0 0.80 00.0. 2.5.00 030.00 030. .0. 9.8 0.000 00.00. mm. 50 00.00..
0080.0. 880.0 38.0 ovmwd amend ...th med. 0.. 5.8 00 5.0m mmo... 5. 9.8 0.80 8...? No. Km 00.09.
0008.0. 830.0 008.0 9.3.0 5.6.0 0.000 00.0. 0000.00 0000.00 RS. .0. ~30 0.00.. 00.00. mm. am 00.00..
momood 505... 0.00.0 momwd nomad 0. SN «N... mmm~.vw «@960 RS. Z. mvdw m6... 8.8. omdmv 006:.
«88.0 9.086 83.0 vommd ommmd 0. 8m 2... mommiw BNNdm menu. 5 mvdw 0.03. 8.8. 3.80 6.90
«0806 03.5.0 moved momwd oommd oémm mp... mvmwéw oommdm mam. 5 ovdw 0...; 5.8. 3.80 mud...
0008.0 0008.0 0 Bod 88.0 000.8 0.90 00.0 0000.0. 030.3 006.00 00.00 0.000 00.00 3.000 00.000
0080.0 805.0 800.0 ooomd 000.3 0.03 3.0 800.0. 008.3 mwmmdm 5.3 0.000 00.00. 00.2.0 00.80
0008.0 0008.0 300.0 080.0 0000.0 0.05 mmd «000.0. 83.00. 008.3 8.3 0.000 00.00 00.2.0 00.000
..03. 0 ..03. 0 ..03. 0 .. .00 .. .... 4.0.0 0 3.; .00 SE. .00 3.; .00 8.... M.— .E.< ~0.. Eu. .0 300 .0 ENE. 0005M! 00.3.2.0

No 5.000. 00 $5.00 ME. ..00 00.02MB :30 «0 03.00. 00 5.000. 00 0.0.0 ,

L00 <0 .000 <0 m00<0 00000.00

 

 

mwo ... =3. .0. 20:38.5 02.8.0880 820 2.65 05 3.05.5 .026 002% .085 5m 03....

 

149

 

.2883 8382: 2: 0. 892300 :22: $302 23 :0 we 2: 3 83803 3:23 2: 3 go 2: 2 5.; .3 0.3.0

 

F 8 B 53235.8 3 can 8 .8332: 2:3 2: a. :3 .8 8

 

9302 3 3:25 3:29.302 3 33.63302 3 2:803 .223 2: 2 2.; .03 $302

 

8 .0 83338.8 :0 at... 8 32:32: 3:3 2: 2 :3 .0: 8

 

3:9: 5 233.50 3 23:09: 3.: :5 2 k :03 NO 930:

 

Bag 5 288. B 8:8: BE 2: 2 m .3: 5.32 030:

 

3:3: 5 0:800.“ 3 20202: BE :5 2 k :03 Q0 0503

 

22: s .832. B 2.22: 222. 2: 2 23 «0 :50:

 

33:: 5 N330»: 3 3003.: 392: 05 2 «at: @8309 930:

 

 

 

 

 

 

 

 

 

 

 

use: 5 328.. :o .323 3:2. 2: 2 2:. 8 :50:

 

2.22:0 3 :33: 330205 2: 5:: 823 33:: 2: 5 0:2 :3 3 835 3 3:25: 2: :5:

 

3:35 .3 2.58:0 823 22: s 82.33 3 322. 2: 2 m 8 050:

 

N332 3 320:. 330205 2: 5:: 823 53:: 2: 5 9302 3 835 3 3:53: 2: 2:33:35 :: 858:0 823 23:: 5 $302 3 520:. 2: 2 m 5302 9:0:

 

0:200 3 522: 330205 2: 52. 823 23:: 2: 5 2.200-. 3 835 3 3:53: 2: 323:3... A: 852:0 823 2:0: 5 0:200-» 3 332: 2: 2 m :0 :50:

 

 

No 3 0205 3 ... 33: 2: 50: 823 3:2. 2: 5 0:22.50 3 835 3 2. 2: M53238 2.: 852:0 823 2:6: 5 2.22:0 3 835 3 x 2: 2 :9: No 9303

 

M3302 3 835 3 ... 33: 2: 50.: 823 332. 2: 5 $302 3 «205 3 ... 2: 9:25:30 2.: 852:0 823 23:: 5 9302 3 835 3 ... 2: 2 :9: ”213020503

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

°20~0° ‘0 QO~OE ‘0 nOfiEDE inbu on: 50.: OMQCN Lon“) QC» 2‘ thuoo.» ‘0 QQ~OE ‘0 kficsat at» QCBOquQDQ Na BOC~Q~OO OMQCQ b~3fi= p: ot0~0°oh :0 O¢~OE \O kQOEs-C 0n: Q~ 40: Q0 Qfiuo:
28- 8.8 223 88.0 .22.2 .298 .2. 8.2 :28 2.85
8d. 2.8 28: 826 .2.2.2 2.2.8 .2. .82 .28 98.0
23. 2.2 285 88.0 n2.2.2 22.2 s.» 8.2 .23 286
B... 8.2 285 28.0 22.2 22.2 2.: 2.2 85.0 8qu
a: 2.2 28.0 28.0 $2.2 22.8 2.» 2.2 23 823
a... 2.2 88.0 38° 22.2 22.8 2.» 2.2 22. 885
2.» 3.2 :23 85.0 28.2 28.2 3.» 2.2 .32 26.0
2: 2.2 22.0 836 ..2.8.2 22.8 :3 2.2 :53 86.0
2: 2.2 82b 83W 282 23.8 8.» 2.2 38.0 86.0

:3 6m :3 .8 :3 .8 ... ..0: m ..0: m ..0: 2t: 2:: EV u. .. u ._ u . 40:: 40:: 40:.
8 5:02 8 8 $.32 8 8 $202 8 8 5:32 8 8 532 8
82.22.. 28: 0l50: 02:0: 0.2.30: 2:2,. 030: J23: 0.50: 0.3: L0l50: 2:0: 050:

mwo # :sm .8 macaw—5030 2232:0030 823 2:3..— 2: wagonm 325 303m 382: ”3:08 m-m 033.

150

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Euro}: «0 ovolndw: 88.: mm~od 8.0:. 8.8: A .:<.:.0.:.
vo.m:.:.0. «800.0 00 38.8: 53.0 005.0 01mm coda—w 5x009
vommwd. 00000.0 05E mmmvd 0:000 58.0 00.: 80 00 00.0. 8.0 N0
0 Lw: 88.0 88.0 88.0 8.0 8.0 N0 00.« 0.0:. 3:02:00
:m+<uQ 00::0000 000000 00: :00 200.0500 :03: mmgos: 0. .0050 020$ at: 0 $00.3:
0000.0 \ «:00 00000 .000 . :0000: 2: :0 .020 :02 00:0000 00 2: :0 0:20 00:00 .3 005200 00000 .000 0: 20000: :0 00:20 :0 0 2: 0: 40:: 01000: 0005:
~¥:.:.0000.0\«00:0:0> 002:0 .000 . 00 2: :0 .020 :00: 00:0000 000 2: :0 220 00:00 .3 002200 002:0 .000 0: 00200.: :0 00:20 :0 0 2: 0: 40:0 00 0003.
l _ 002:0 0:02: 2:: :0 0:00:00. :0000> 2: :0 200:0; :20: u «:23: 00000 :000> 2: :0 3.3.5:.
7 0:22. :0 00000:. 030:: 0.50:8 2:: :o 02.8 :02 n 0000:8004 0:00: 2:: :o 000 00:.
an 00:00 0 2: :0 0220200: 200:0: :200: :20: 2: .3 0000.00 0020 :0:20 2: :0 002200 20020 2: :0 Emzma
0.00000 0:: .3 0000.00 0020 :200: 200000200 0000 2: :0 E00 20:: 002200 «20000: 60.00: 200002000 0.: :0 200:0; :0:20 2: :0 30m
_ _ _
20002—3 00:00::—:00:.~m:02_00: :0 2.: M 20000: :m: 200020 :0:::00 2: 0: 3:0. (.00 mm<0OMh
: :00 :0 5.80: 088.0 00 _ 00 :0 0008038000 3 0:0“ 00 :85820 0:02: 2:: 0: 0:0 6.: 00
00920 3.00 :: 880.0 5:000: 00 0: :0:::0:0 0220200 20000: :0 200020 :000> 2: 0: 20:00:: 23.00%:
0 0 <_ a... 2.28.: 05 0.2.8. B: .0 .25: 0:25.028 026:2: _ _ _ J
«01.203 0.: 2 0220000 2020000 00:20< 00:00 :3 0220200 2220202 :0:00 :00 2: :0 00 :0 200020 :0000 2: 0: :20. .00 00
_ _ _ _ _ _ _ .000 :0 0::00 0: 200020 0202:0flqm.
2.05:0: :0 :00025: «02.0.3.0: :0000000: 200020 2: :0 00:22:00 2: 0: 0220000 200020 00:02:00 0: (:00 .0 0000
: _ 2220200: :0:00 :00 0000 2: 0: 5%
:0000002: 200020 000000 2: :0 :00002 2: 0: 06050
:0000002: 200020 :00: 2: :0 :00000: 2: 0: 9:00:31:
«88.0 mm :86 @000 ovu :0 mowed :3: 0 Km: 88.0 88.: 83.: mm.— dem 00.5: 00.3 00.8
«88.0 3:80 08:0 ovm :.o mowed 0.3: 003.: 88.0 8.8.: 0000.: 00.: «0.9.: 00.8: made 00.8
«008.0 8:80 «B :.o :vu :d Bumd 0.3: 003.: 9.8.0 «08.: 030.: am.— Nm.m~ 00.8: mmdm 3.8
«080.0 088.0 0:30 mo. : :6 03.0.0 mdm: 800.0 00 5.0 008.0 08:6 Mb :0. :: o :00: 00.0: mu. ::
«88.0 038.0 m :0 : .0 mm : :0 03.0.0 0.8: vowed :0 5.0 «05.0 :. :o:..o mmd :o. : : 8.8: 3.0: mm. ::
«008.0 «38.0 0:30 00 : :6 030.0 0.00: 0000.0 «05.0 83.0 0000.0 W6 :0. :: 2.00: ~00: 002::
2.50.0 $08.0 W030 mm : :.o 2.8.0 :00: amend 88.0 0000.0 00:00 00.0 00.0 5.00 3.0 00.0
:0 :80 m 88.0 00:. : .0 mm : :0 008.0 000: a 3.0.0 008.0 800.0 mm :00 5.0 00.0 8.00 :.~.m 00.0
mm :86 508.0 mmv : .0 mm : :.o «08.0 0.3: 0000.0 008.0 09.0.0 0000.0 Wad 00.0 5.00 W0 8...:
00:: 0 .53: 0 .: .00 a .:0 000m 0 ::_:.:.< .00 2.; .00 8:.< N0 32% 0.0 E< .0 (:00 .0 5&0.» 062.0 02.0%
03000: 00 E0400 E400 Emzumn Elam 0(000: 00 00.002 00 E8
I000 <0 #0003: 00020::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mo 0 00M :0: 000020200 00000000000 00000 2005 2:: 90.52% :000m 002% Boxm ”Ym 050:.

 

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

200020 0200020 2: 0: 0020:. 000 00.0.: 20000: 000 00 .«0 2: :0 0200020 :2200 2: :0 E00 2: 0: ELQ .0 0.:10 «00:.
. _ _ .
0:000? :0 00:00:: 20.36000: :0 0>wnm:0000: :0 200020 .2000 2: 0: 3:8. .00 5.000:
00 :0 =0..00.:.o=..:00 :0 0...... 00 .8508... 0.000.... 0. 0:0 .00 00
_ . . . _ 00...
0:00.: 0: 20000: :0 00:00:: :00. 2: 0: 0 :03: 01002032.:
0:00“: 0: 000:00_.: :0 0000—2: :20 2:— 0: 0 :00: WEDGE
0.0.... 0: 2008. :0 .0022. 0.0.0... 2.. 0. .0... 00000. 0.00...
00...... 0. 0028... :o .0940 :30... fl... 0. 0:... .00 0.00...
00.00: :0 2:20.... :0:000:20 2: 5:0. 000020.02: 2: 0: 0:080: :0 00:20 :0 :02000 2: 00:22:00. ...0 002200 00000 0:00.: 0: 0:280: :0 0:20.: 2: 0: m 0<000: 0:02.:
00200 :0 30:00. :0:000:20 2: 5:... 002:0 0:02: 2: 0: 00200.: :0 00:20 :0 :00050 2: 00:22:20 :0 005200 00000 0:00.: 0: 00200.: :0 0:20.: 2: 0: m 00 050::
_ . _ _ .
0 00: :0 0_0:20 :0 0 :20: 2:: 0.0:: 00000 300.. 2: 0: 20000: ::_: 00:20 :0 :0 2: 000000000 .3 000.200 002:0 0:000: 0: 0:0000w :0 00:20 :0 0 2: 0: :03 5:000: 0:000:
0 :02000 .20: 2: :02: 00000 :0000 2: 0: 00200.: :0 00:20 :0 :02000 2: 00002300 2: 002200 00005202: 0. 00200.: :0 00:20 :0 :02000 2: 0: :0: :0 n.:|:.:0:.:
8.0:. 00.: «88.0 0«86 0000.0 080.0 0«.: 00.0 00.: 8.: 003.0 88.0 $2.00 $00.: 00...: 00.: «80.0 0«80
08.: 00.: «80.0 m«8.o 088.0 080.0 0«.: 00.0 08.: 8.: 090.06 8«o.o $080 $8.: 00.00 00.: «80.0 0«80
00.: 0...: «80.0 0«86 0800.0 080.0 0«.: 8.0 00.: 8.: 008.0 :80... $880 $8.: 08...: 00.: «88.0 0«80
8.0:. 00.: «08.0 008.0 008.0 88.0 80.0 «0.0 000 00.0 008.0 0890 $00.00 $00.: 000:. 00.: «08.0 008.0
mod: on: «08.0 88.8 wooed 88.0 88.0 8.8 who 8.0 888.0 08.08 $80.88 $80.: 860 00.: «08.0 008.0
00.0: 0...: «08.0 008.0 0000.0 88.0 08.0 8.0 000 8.0 88.0 0800 $80.00 $80.: 00.8: 00.: «08.0 008.0
8.0:. 00.: n: 8.0 008.0 080.0 800.0 8.0 8.0 80.0 00.0 «08.0 0:«o.o $8.00 $00.: 0«.0: 08.: u: 8.0 008.0
..«.0:. «0.: m: 8.0 008.0 080.0 «80.0 00.0 8.0 :0... 00.0 008.0 n:«o.o $8.00 $00.: 0«.»: «0.: s: 8.0 808.0
0«.0: 8.: 0: 8.0 88.0 :. 806. 080.0 00.0 8.0 00.0 8.0 88.0 0:«o.o $00.00 $00.« 8.0: 8.: 0: 8.0 8 8%]
I0 .: u : :03: : 003:: 003:0 00:... 2.2: .00 3.2: .00 3.0.0. .0 25.0 .0 0 00:... 0 40:: $.05 $.5— 0 .: 0.: 000:: :03:
01000: 00 0.0000: 00 5:000: 00 Weiom: 00 «(0‘3 0.:10 5.000: 00 00.000: 00 0<000: 00 01000: 00
[P503 £02.: GEO: @000: m00§ 000$. P303 0:02.: MSG: MEG: 2.00:: 0%: go: 050::

 

 

mo 0 =3. .0: 0:00.233“: 02.2.0850 00...... 0.0.0... 0... 8.325 .028 2.2% .086 .3000. ..-m 0.2....

 

_

 

30306 N0 635 :.«N: 668. : mane. o 60.: 6666: 40.0.0: ...... 63. :33 wining Awwmw>

 

 

 

 

 

 

 

 

 

3.22.- 88:... 00 88...: 88... 88. o 8.: 8.8 00000. _
8.000... 88:... 000000 0.....de 8528...... 0...... 8... 00 8... 8.... 00

0 0W 88.2 88.288. o 8... 8... 00 8.: ... .... 0000.00
2.0.0 8.08.. 0.08.. ......om. 200000.30... 00.0... 0 .000... 0.0: 0...... 0 ..000...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0000.0 \ :00 00000 .00.. . .0000: 0.: :0 .020 000: 02.0000 00 0.: :0 20.0 00:00 :0 002200 00000 .2... 0: 0.0000: :0 00:0... :0 0 2: 0: ..02 00000: 0003.
05.008.000.029. 00000 .00.. . 00 0.: :0 .020 000. 02.0000 003 0.: :0 20.0 30:00 00 002200 00000 .20.. 0: 00200. : :0 00:00. :0 0 0.: 0: :02 00 00000
_ 00000 0.0:... 0.: :0 200.0... .0002. 0.: :0 00.0.9. :22 u 00...: 00000 0000.. 0.: :0 N20020:.
_ 0.800 .... 000.30 0.0:... 00.0.... 2.. .o 82.. .32 ... 02.... 8...... 00...... 2.. .o 03.0 ..0..

 

 

 

 

152

«0 00:09 0 0.: .3 020.0200: 00.0.0: .200. :22 0.: :0 0000.00 0000. 00.0.0 0.: .40 000.200 20.0.0. 0.: :0 E0200

 

2.0000 0:: 00 000:..00 0000. 00:00. 200000.000 0000 0... :0 0.00 0.0.: 000.200 3.0000: .00.«0.10:00000.000 0:: :0 00.0.00 .20... 0.: :0 2B

_ _ _ : _ _ _ _

0.0000: :0 00:02:00.. 9.0000: :0 00: a 0.0000: :0 200020 .2000 2: 0: 200.00 00.0000:

 

 

 

 

 

 

 

 

 

 

 

 

 

:0... .0 .082 088... 00 _ 00 .0 8080.08.00 .0 0...u 00 .8508... .00.... 2.. 0. E0 .0... 00

88.««« 8.8:: 836...“ 00000: 00 0. .0:.E:0 020.0200 0.0000: :0 200020 .009. 0.: 0. 2.: 0.00 000000:
0 ...... 2.28.: 0.... 0.2.8. 3. o .....0 0:20.828 025:2... _ _ _ _

~0+.:..\m.<nN:30.: 2 350.0000 0.020000 00.200. 30:00 :0 02200.00 20.20022 :0:00 .00 0.: .0 00 :0 200020 0000.. 0.: 0. 2.810.. 00

_ _ _ _ h L _ .000 :0 2:00 0. 200020 0202.00 0: 2 .:.< .0

 

2.09:2 :0 .0002"... «0.50"... 0000000.: 200020 0.: :0 00:20:00 0.: 0. 30.0.0000 200020 0202.00 0: 0.00 .0 0000
_ _ 22200.02 :0:00 :00 0000 0.: 0: EMF...
0.000002. 200029000000 0.: :0 .00002 00. 0: 000I..:|.0
00000000... 200020 .2: 00. :0 .00002 00. 0: 92:00

 

 

 

 

 

 

 

 

 

3386 mm :86 man :6 3:.N :6 «owed 6.03: o :o«.: 30: :6 :83: omvm. : 33.: «3.33. 30.03: 03.33 03.33
3386 mm :86 mom :6 30m :6 «omwd 6N3: mmow. : «a: :6 omen: momm. : 3m. : dew 35.03: 3333 03.33.
03.0666 03 :86 men :6 ovw :6 536 0....3: $8.: 03: :6 «303.: 333.: 33.: «3.03 03.0.3: 3333 03.3w
63366. 6 336066 : :0: 6 Now : 6 33036 n. 33: «mmod 3386 33536 332.. o 36 m«.: : o :30: N03: 30. : :
63366. 6 030606 : :0: 6 «6N:6 63.06 n. 33: 3336 mowed «3036 n :on. o 2.6 mm. : : 8.36: 06.0: 30. ::
53666 @3866 N :v :6 :ON :6 30036 3.33: «3 :06 ommod 36506 mmmod 0.56 mm. : : 0.0.60: WW3: 30. : :
no :86 omoocd 300 :6 mo: :6 3.336 3.33: 3836 3.3.6 33036 mm :36 336 03.3 5.03 v3.N 02m
no :606 omoood 33. :6 mm: :6 00036 3.33: mummd 6036 @3036 mm :36 $36 03.3 63.03 03. N 006
33 :86 86666 300:. o 03: :6 30336 3.33: 0:. :36 3386 33036 33636 036 03...“ 5.03 0«. 3 006

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

..0... .. ..0... .. .. .... .. .... ......u .. 0:0 .9. 4...... .o... :00 .2. 0:0 .... =00 ... 0.0.. ... film... 00...... 00.0....
00000. 00 2.0.0.. 00000.. 2.0200 ...00 00000. 00 00000. 00 0000
0000.. 0000.. 0.0200

 

o— ... :3: .0: 2.2.0.0200 02.8.0880 80:0 203... 0:. 3.32.3 .0023 000.03 .0800 “Wm 0300.

 

153

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

20000:: :0:00qu 2: 0: 00:30:00 :25: 92:00: 0:0 00 000 2: :0 080002: 2:0: 2: :0 E00 2: 0: SE: .: o:<u OE.
_ L _ _ _ _
90:00: :0 00:00:: 3:230:00: :0 :>:nm:0:00: :0 2000023220: 2:: 0: Sb: .0: 0:00:00:
00 :0 308035.30 :0 min 00 32:32.: 00:2: 20 a: 0.: < .0: 00
_ _ _ _ _ 00+:
0:00: :: 002:00: :0 :0:02: :2: 2: 0: u: :03: meiom: O59:
0:02: :: 0:23.: :0 :0:02: :2: 2:: 0: n: :9: 00 @009:
_ : _ r _
0:02: :: 20:00: :0 300:0: 2:203 2: 0: 2:: 0:360: O59:
0:000 0: 0:28.: :0 .520: 30:3 2:: 0: 2:. 00 0:00:
_ _ : _ :
:00: :0 30:00. 20022: 2: 5:0. 002:: 200:: 2: :: 20:02 :0 00:2: :0 02:0: 2:: 0:32:08 0.: 00::2:0 002:: 200:: :: 90:00: :0 30:0: 2: 0: 0 03:00.: O39:
0:200 :0 30:00. 20022: 2: 2:0. 002:: 0:00: 2: :: 0:202: :0 00:2: :0 :02:0: 2: 0:32:02: 0: 002030 002:: 0:00: :: 0:209: :0 2:200. 2: 0: 0. 0o O59:
_ _ _ _ _ _ _ _ _ _ _ _ _
:0:00: :0 022: :0 0 :20: 2: 80:: 002:: :0:00 2: :: 22:00: :0 00:2: :0 0 2: 02:02:00 0.: 002200 002:: 0:02: :: 90:00: :0 020:: :0 0 2: 0: :03: m<100:OSO::
0 02:0: :20: 2: 50:: 002:: :0:00 2: :: 0:209: :0 022: :0 02:0: 2: 3:002:00 0.: 005200 002:: 0:00: :: 0:209: :0 020:: :0 02:0: 2: 0: :03: :0 @003
00. K 00.0 000.0 0000.0 0000.0 0 :00.0 00.: 0:.0 00.: 00.: 0000.0 03.0.0 A00:00 A000.0 00. K 00.0 00:00 K0000
00. K 00.0 000.0 0000.0 0000.0 0 50.0 :0: 0:.0 00.: 00.: 0000.0 0000.0 A00:00 6000.0 00. K 00.0 00:00 K0000
00. K 00.0 005.0 K0000 0000.0 080.0 :0.: 0:.0 00.: 00.: 0000.0 003.0 $3.00 00.00.0 00. K 000 020.0 000.0
00. K 00.0 0000.0 Kvod 0000.0 0000.0 00.0 00.0 K0 00.0 0000.0 K0000 n000.00 0000.0 00. K 00.0 0000.0 K30
00. K 00.0 0000.0 030.0 0000.0 0000.0 00.0 00.0 Kd 00.0 0000.0 K0000 $00.00 n000.0 00. K 00.0 0000.0 030.0
00. K :00 0000.0 050.0 0000.0 K0000 00.0 00.0 K0 :00 300.0 0080 A$00.00 $00.0 0.0. K 0.0 0000.0 02.0.0
00. K 0:..0 0000.0 :0000 200.0 0000.0 00.0 00.0 00.0 00.0 0000.0 :0K0.0 oxo :000 0000.0 0K. K 0:..0 0000.0 :0:00
0:.. K 9.0 0000.0 :0000 200.0 0000.0 00.0 00.0 00.0 00.0 0000.0 :0K0.0 $5.00 .0000 0:.. K 00.0 0000.0 :0000
0K. K 00.0 0000.0 :0000 200.0 0000b 00.0 I000 00.0 W00 0000.0 :0000 n0:000 o000.0 0K. K 00.0 0000.0 :030
0. : 0.: :92: :02: :Ost. :O:.:> E; .0: 3:1 .0: Eu: .: 3.; .: u: :02 ..: :03: 02.3 00.5. 0. : u: :03: :02:
.0330: 00 0:00:00: 00 0:530: 00 0:530: 00 0<W= 0:30 5:100: 00 50:00: 00 m<:00: 00 52:00: 00
O59: OISE: O52: R09: mo:§ 05:3. O52: O52: OFDO: @003 go: O52: P09: O39:

 

0: :0: 00m :0: 000020200 000000000000 80:: 0:03: 2: 053000 000:0 002:0 Boxm ”3:00: Wm 030:.

 

 

154

 

 

 

 

 

 

 

0%"de N0 0000. 5F 008.: 0000.0 8.55 8.80 .:<.:.0.:.
09w: 9.0. 0800.0 8 0000.00 0 0000.0 «000.0 00.00 8.00 IviOm:
0°-w00.0- 0000:..0W teipM: 0:0 0 .00 :00 ...o 000 F .o 00. C 8.0.. m0 006 8.0 NO
m 0: 88.0 88.0 800.0 86 8.0 NO 8.0— Qt. 3:09:50
~m+<nQ 50:25: 5:26: 2: :0": 20:05:: :03: mqus: 0 .mm<3: RE 295:. u .mmcs

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0: #00000 \ «2:50: 002:: i0: .00 2: :0 2:002: 320:: :0:0::0 00: 2: :0 0:20 0.50: A: :05030 002:: :0: 5 No :0 00:2: :0 0 2: 0: 40:: No :O:§

 

0000.0 \ :0: 002:: :0: . :0:00: 2: :0 .02: 00:: :0

:0:00 00: 2: :0 20:0 :50: :3 :05030 002:: $0: 5 20:00: :0 00:2: :0 0 2:0: 40:: 0:330: :0:00—

 

¥::.0000.0\~0E::0> 002:: :0: . 00 2: :0 .02: .:0

:0::0:::0 00: 2: :0 0:0:0 050: :3 005200 002:: :0: 5 00200.: :0 00:2: :0 0 2: 0: .53: 0o :0:<>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

002:: 5:0: 2: :0 2:52.. :0000: 2: :0 2:52. :20: n «:23: 002:: :0:0> 2: :0 $0340:
0:000: 0: 0000.00 030:: 350:0 2:: :0 000:: :80: ... :23: 80:: 0:00: 2: :0 0230:
n: :0:0: 2: A: :0:0::0:00: 2:52. :22: :0:0: 2: .3 :0::>0: 000.: :22: 2: .3 :05030 2:35 2: :0 E022:
50:0: 0:: N0 :0::>0: 000:: :22: 0::0:0:Eo0 :000 2: :0 5:0 50:: :05200 020:8: 60.00: 0::0:0N::00 2: :0 0550:. :0:0:: 2:: :0 :30
«20:00: :0 2:002: 35:5. «00 :0 2:002: :0:::0::.:0:0:: 0: :0:00 0: :: .00 :0 0::002: 3:0: 2: 0: :E. < .0: 00
20:00: :0 :0:00:::0:: «20:00: :0 :>: n 20:00: :0 2:002: 320: 2: 0: 5.510: 33:00]:
$0.20 3.82 088.0 00 _ 00 :0 02:03:35.:00 :0 0:0" 00 32:80:: 300: 2:: 0: :20 .0: 00
000.000 .0030: 00000.0 :0:00: 00 0: :0::E:0 :0:0::0:00 20:00: :0 0::000:m:0m0> 2: 0: :_:.:.< :> 2:00:00:
0 .2. 2.300.: 0... 22:8. 3: 0 2:: 00250200 026021 _ _
:0+§B.<u~::00.: 0: 052000: 2:20:00 052:: 0.50: :3 :0:0::0:00 2::2NNE0: :50: :00 2: :0 0o :0 2:002: :0:0:. 2:: 0: =5: :0 0o
_ _ _ _ _ _ _ .50 :0 0:::: 5 2:002: :0:00::00 0: 00.2: In:-
::.:m::: :0 50:22: «5.4590: :00::0:0:: 2:002: 2: :0 3:25:00 2: 0: 0520000 2:002: :0:00::00 0: Sm: .: 5:00
_ H 0:220:20: :50: :00 :26 2: 0: :mmlli
:00::0:0:: 2:002: ::0000 2: :0 30:02 2: 0: :00II.:.:
:00::0:0:: 2:002: :2: 2: :0 30:00: 2: 0: 00:00::
00000.0 00 50.0 0:0: .0 000:0 00 :00 0.00: 0000.: 0000.0 0000.: 0000.: 00.: 00.00 00.00: 00.00 00.00
00000.0 00 50.0 000: .0 000:0 00 :00 0.00: 0000.: 0000.0 0000.: 0000.: 00.: 00.00 00.:.0: 00.00 00.00
00000.0 00 80.0 000: .0 000:0 00 :00 0.00: 0000.: 800.0 0:00.: 0000.: 00.— 00.00 00.00: 00.00 00.00
00000.0 00000.0 0:000 :.0 p :0 0000.0 0.00: 0000.0 00: :0 0000.0 0000.0 0:0 00. 2. 00.00: 00.0: 00. 3
00000.0 00000.0 0:000 00: :0 0000.0 0.00: :0000 :0: :0 00:00 R000 :50 00. 2 00.00: 00.0: :0. I
000000 80000 0:000 :0— :0 0000.0 0.00: 0000.0 00: :0 00:0. 0 0000.0 Rd 00.: 00.00: 00.0: 00. 2
00:00.0 00000.0 000:0 00: :0 0000.0 :00: 0000.0 0.00.0 0000. 0 0: 5.0 00.0 00.0 00.00 00.0 00.0
00:80 EEEEFE IEBHLIBRbllmmEbI |1000 000 00.00 00.0 00.0
00:00.0 00000.0 000:.0 00: :0 0000.0 0.00: 0:000 0:00. 0 0000.0 00:00 00.0 00.0 00.00 00.0 00.0
00000.0 00000.0 000.0 00 p :0 0000.0 0.2.: 000:0 00 :0. 0 B0— .0 :.00 :0 0:0 0:0 00.00 00.0- 0:.0
00000.0 00000.0 80:0 00— :0 0000.0 02.: 000:0 00 5.0 000: .0 000:.0 3.0 00.0 00.00 00.0- 0:.0
00000.0 00000.0 000.0 00:. 0 0000.0 3.: 000.0 M0 8.0 000 :0 0:0 :0 |0F_..0 0:.0 0%0 430.0- 0.0
.53: > 40:: a .: :2. .: 3 08mm : 3.2: .0: 3.3: .0: 35‘ .m: :E. <E Ea: : 50m .: rm: :0..th 00:002.:
:330: 00 E340: MRI: .5: >5:ng Elam :<:om.: 00 5:00: 00 E000
06:5. :93: 202::

 

 

Z :0 00m .5: 00000—0200 00000000000 000:: 20:03 2:: 05398 Beam wwoam BBQ 60: 030.0

 

155

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6.580.... 02:82.. 2.. 0. 80.0.. :8 8...... 92.8. 2.0 00 2.. .0 8.082.. .202. 2.. .0 5:0 2.. a. 2.; .0. 0.30 0....
_ _ _ _ _ .
N880. .0 00000... .05. @8000. .0 q>vum.2.02 .0 9.080... .202. 2.. 0. 0...... 6.. SRO“.
8 .o 22.02.3528 .o ...0" 8 .8508... .202. 2.. 2 =5. .2. 8
_ _ _ . «.03.
2:2. ... 22.8. .o 02.8.. .2: 2.. 2 h. ..00. 5&8. 0.00...
0.00”. ... 2.0.2“. .o 00.8%.. .0... 0.... 0. ... ..Os. 00 0.00...
2:2. ... 238. .o .522. 20.2: 2.. 2 2E 010020.00...
2:0... 2. 2.23... .0 28.12.20: M... 2 2.5 8 0.00...
%2 .o 20.2: .2322: 2.. 5.... 82... 2:2. 2.. c. 22.02 .0 3.2.. .o .380: 2.. 0:32.38 .3 .3228 82... 2:2. 0. 0.88. 3 £0.03 2.. 2 0 05.00. 0.00...
2.200 .0 2.0.2: .2820... 2.. 0...: 82... 0.00.. 2.. ... 2.2.8.. .0 8.2.. .0 .0080: 2.. 00.03.35 .3 000.200 82... 0.00.. ... 2.200.. .0 0.0.9: 2.. n. 0 00 0.00...
_ 0 _ _
2.00. .0 ...—0.0... .0 .. .20. 2.. Em... 002.308.. 0... 0“ .0880. .0 8.0:. .0fiocyz..0§08 .3 000.200 82.0 0.00..— ... 9.2.00. .0 8.0:. .0 0. 2.. n. .5... 50.00. 0.00...
o .0050: .20. 2.. E0... 82... .08.. 2.. ... 2.200.. .0 00.2.. .0 .0050: 2.. 05.02.08 .3 80.200 82.... 0.00.. ... 2.200.. .0 8.2.. .0 80:5: 2.. a. do... 00 0%..
00.00 00... 0000.0 0.0..0 0000.0 0 .000 00.. 00.0 00.. 00.. .0000 000.0 $. .00 $00.... 0...... 00... 0000.0 0.0..0
8.8 00. .. 0000.0 0.0..0 00000 0 .000 00.. 000 00.. «0.. .080 08.0 $. .00 $00.... 8.3 v0. .. 8000 0.0.0
8.00 00. .. 00000 0.0..0 00000 0 .000 00.. 00.0 00.. 00.. .080 000.0 $. ..00 $00.... 8.00 we. .. 8000 0.0..0
00.00 0... .. .0000 0 5.0 0000.0 0 .000 00.0 ...0 0.0 00.0 0800 0.0.0 $00.00 $00.... 00.00 3.... .0000 0.0..0
00.00 3.. .. .0000 000 ..0 0000.0 0000.0 00.0 ...0 ...0 00.0 800.0 0 .0.0 $00.00 $00.... 00.00 3.... .0000 000.0
00.00 0... .. 0000.0 000.0 0000.0 0000.0 00.0 ...0 ...0 00.0 8000 0.0 .0 $00.00 $00.... 00.00 0... .. 0000.0 000.0
00.00 .0. . . 00000 000 . .0 0 .000 .0000 00.0 00.0 000 000 8000 000 ..0 $ .000 $00.0. 00.00 .0. . . 00000 000 ..0
0000 .0. .. 0000.0 |mmb.0 0 .000 00000 mm... 000 mmb ..00 0.00.0.0! :wa3 $800 $00.0. 00.00 .0... 00000 000.0
00.00 .0... 0000.0 000.0 0.000 00000 00.0 00.0 00.0 ..00 0000.0 000 ..0 $ .000 $00.... 00.00 .0... 00000 000.0
00.00 ..0. .. 0.000 000 ..0 00000 .0000 0.0 000 v.0 0.0 0080 000 ..0 $00.00 $.00. 00.00 v0. .. 300.0 000.0
00.00 00.. . 0.000 000.0 0000.0 .0000 0.0 00.0 ...0 0.0 008.0 000 .0 $00.8 $.00. 0.0.00 .0... 300.0 000 .0
00.00 00. .. 0.000 000.0 0000.0 .0000 0.0 00.0 N . .0 M .0 0.000 000.0 $00.8 $.00. 00.00 00. .. 0.000 000.0
0. .. 0. . .50.. .33.. 00:... 00:... =h<4mm E1 6.. Ev. ... 2.5. ... ... .5: u. .52 E 3...... 0. .. 0. .53.. .52.
5.100. 00 03.00. 00 $0.00. 00 0.3.00. 00 «<0... 0.08 0.3.00. 00 50.00. 00 23.00. 00 0.3.00. 00
[9:003 E030... @002 0.00... tons. E3. E0... p00... E... 0.00... 0.00... 0.00... 0.00... 0.00...

 

 

 

__ ~ 0... 05M .5.“ mags—00:00 GomawbcoocoU mesm @394 05 WEBOSW “005% Cucumw 30me ”AJCOUV Gum 03.0%

156

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8806 No 888:: 88.: 8:66 8.:.:. 8.8: :50:
com: :8. «68:6 00 88.8 386 886 06.8 86:. m Siom:
com—8.0. 8086 mEPm: 88.8 856 886 0:8 8.8 8 86 86 mu
m < 886 88.6 886 86 86 NO 8.8 «.5: 3:02:00
$2.0": 00:00:00 8000 05 :0": 20:00:“: .62 800:: 0 80:: 2:: 25. m .802
0806 x :2: 0023 i0: . :0:000: 2: :0 .02.: :02 :0:0000 00 2: :0 20:0 :52. N: 005030 002:: 2:00 5 82:00: :0 00:05 :0 :: 0.: 0: 40:: 5:000: mom—S
9:.:.8N0.0\«0E=:0> 002:: 2:2. . 00 2: :0 .02.: :02 00:00:00 00: 2:: :0 20:0 :50: .3 005030 002:: .23 5 00200.: :0 00:2: :0 : 2: 0: 40:: 00 g: S:
002:: 0:00: 2: :0 2:52.- 3000:. 2: :0 2:50) :20: u «:23: 000cm 3&2. 2: :0 NE: 40:.
50:00 .3 000300 0:00: 050.00 2:: :0 80:0 :80: n 00::.:: 82:2 0.50: 2:: :0 .25 40:.
an: 00:00 2: 2: 00:53:00: 2:50: :52: :20: 2: 2: 000:>00 002: :0:05 2: .3 005030 225:: 2: :0 :CBZNQ
2:0:00 0:: .3 000300 000:: :52: 2:238:00 :000 2: :0 5:0 50:: 005030 3:80: .033: 2:288:00 2: :0 2:59. :52: 2: :0 Sam
0:0:00:_:0 00:05:08.:200r: :0 :0: n 20:00: :0 203.2: 5:00 new: 0:: 0.0.: 000000:
: :8 :0 5.08: 0.8.8.: 00 _ 00 :0 5:02:3580 :0 .8" 00 32082: 3:00 2:: a: 2:0. .00 00
mvo.««« 5.8:: 88:8 8:109 00 0: :0:::::0 00:58:00 82:00: :0 20000:: :23. 2: 0: 3:: «12: 8:362
o m < :20 0:209: 0:0 H230. :0: .0 .25 2.225200 mlz.o.:z<_ :
~0+.::\n.<u~&u0.: 0: 2503000 2:20:00 00:25:30: :3 0203200 222022.: :52: :00. 2: :0 00 :0 0,5000% :23. 2: 0: 2.; .2. 0o
_ _ ~ :50 :0 2:2: 5 0.50002: 00:02:00 0: :3; n:
that: :0 30023: 232:9: 30:00:02 20002:: 2: :0 8:25:00 2:: 0: 0150:0000 05000:: 0200200 0: Sum m: 5:00
H 0:222:50: :52: :00 :26 2: 0.: GNP:
800082: 2:002: 0:008 2: :0 :00002 2: 0: @093.
5:00:00: 20320 5:: 2:: :0 30002 0:: 0: 003::
:886 0:.«86 «8:6 :8:6 8:06 N8: 0086 886 88.: 88.: 8.: 8.8 8.8: 8.8 006.
8:.86 6886 «8:6 :8:6 8:06 :8: $86 886 88.: «:8: 8.: 8.8 :58: 8.8 006.
8:86 6:.«86 «8:6 :88 0:. :m6 :8: 886 886 08a: m :8: 8.: 8.8 :08: 8.8 006.
0886 8:86 936 8«:6 886 «.8: $86 :«86 « :86 $86 86 86: «0.8: «N0: 8.2
0886 8:86 :::.:6 «8:6 886 «.8: 6806 886 886 886 86 86: R86: «No: 0 :6:
2886 06:86 :3: 6 «8 :6 886 «.8: :806 ::««6 E86 286 86 86: 8.8: 8.0: 86:
8 :86 8866 80:6 8: :6 886 060: 886 6:96 336 8 :md 86 «Nm 8.00 3.« «:8
E886 030:6 E86 0.8: 886 «3:6 E896 86 8.? 8.3 Ed .
8 :86 8686 80:6 8: :6 886 0.8: K86 86:6 8:66 0:. 56 86 ««.m 8.00 3...: «:8
8:86 6886 08:6 8: :6 386 34.: 86:6 68.: 08:6 0«« : 6 86 8.0 8.8 «:8. ::.:.
8:86 :886 08:6 8: :6 886 8:: 68:6 :866 «8:6 8«:6 86 8.0 8.8 8.0. ::.0
. 8 :86 :886 30: 6 8% 886 0.3.: :88 886 68:6 lw :« :6 me6 8.0 moan 8.0. 8::
.5: > 40:: > .: c5 443 .:::N > 2.340.: :E.< .0: :E.< .2. 3:0. .2. 35¢ m: (:9: .1 5m?— 163Q 85::
523%: 00 m8 40> 85 40> :8sz 23m m<10m: :0 5:10»: 8 5:00
IMOKS: EOQS. «32::

 

 

«: :: :3: :0: 80303030 8:228:00 08:: 03:05 0:: 83:25 :88 002% 700%,“: ”Wm 030:.

157

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

630002: 0030005 2: 0: 000:2. :00 00:5: 92:00: 2:0 00 2: :0 0030002: 3:22: 2: :0 500 2: 0: E0< .1 0:06 9:0
_ _ _ : A :
92:00: :0 50:00:: :05. «22:00: :0 §:nm:2:00: :0 030002: 3:02: 2:0: 20:0 .0: 0:10:00:
00 :0 5080:8200 :0 00:“ 00 323020 30.3: 2:: 0: E0 6.: 00
_ _ _ 0:03.
030:: 0: 022:00: :0 00:00:: :05 2: 0: h: :03: 50:00: 930::
032—: 0: 00200.: :0 00:00:: :05 2:— 0: ..: :03: :00 900::
030:: c: 2008: :0 50200 20:9: 2:: 0: E0000: 0000::
035—: 0: 00200.: :0 500:2: 30:00. “a: 0: 00:2 _00 930::
00:00: :0 E :02 2:300:05 2: 5:: 002:: 030:: 2:: c: 92:00: :0 00:05 :0 :0052: 2: 00:33:35 :0 005030 002:: 030:: 0: 9:00:00: :0 2:030. 2:: 0: 0 5:100: 930::
00200 :0 30:00. 2:300:05 2:: 2:3. 002:: 030:: 2: c: 28:00.: :0 00:05 :0 :0050: 2:: 00:03:35 :3 005200 002:“: 030:: 0: 28:00.: :0 30:00. 2: 0: 01 0o :59:
5&0: :0 0_0:05 :0 0 “0:0: 2: 52—2: 002:: 00000 2: 0“ 22:00: :0 00:05 :0 m: 2: 0:30—2:30 0.0 00:00:00 002:: 030:“ :: u:0000w:0 00:05 :0 0 2: 0: we: EEO“: 900::
0 :0052: :20: 2:: 50:: 002:: 0000:. 2: 0: 00200.: :0 00:05 :0 .8052: 2: 00:02:30 0.: 00:00:00 002:: 030:: 0: 28:00.: :0 00:05 :0 00050: 2: 0: :0! :0 OM30:
5.00 00.00 0000.0 0000.0 0000.0 0000.0 00.0 00.0 00.0 00.0 200.0 00.0.0 0000.00 60.00.00 5.00 00.00 0000.0 0000.0
00.00 00.00 0000.0 0000.0 0000.0 0000.0 00.0 00.0 00.0 00,— 300.0 00:0.0 e000.00 e0.00.00 0.00 00.00 0000.0 0000.0
00.00 00.00 500.0 . 0000.0 0000.0 0000.0 00.0 00.0 00.: 00,— 200.0 00:0.0 60.00.00 A0900.00 .000 00.00 :0000 0000.0
00.00 00.00 0000.0 0000.0 000.0 080.0 00.0 00.0 00.0 00.0 0000.0 00 :00 00.5.00 6000.00 00.00 00.00 0000.0 0000.0
00.00 00.00 0000.0 0000.0 :0000 080.0 00.0 00.0 00.0 00.0 0000.0 00:0.0 e0.6.00 9000.00 00.00 00.00 0000.0 0000.0
00.00 00.00 0000.0 0000.0 :0000 080.0 00.0 00.0 00.0 00.0 0000.0 00 :00 90.5.00 e0.00.00 00.00 00.00 0000.0 0000.0
00.00 00.00 300.0 0000.0 0 50.0 0000.0 00.0 00.0 00.0 00.0 0000.0 00 :00 60.00.00 o0.00.00 00.00 00.00 600.0 0000.0
TEm 0000 EmFlbmpmd 0000.0 #9000 00.0 2.0 bub 00.0 00.0.0.0 0900 $00.00 nx.00.0.le.0[0.00 0:00 FEE
00.00 00.00 600.0 0000.0 0 80.0 0000.0 00.0 00.0 00.0 00.0 0000.0 00 :00 200.00 n$00.00 00.00 00.00 0000.0 00000
00.00 00.00 500.0 0000.0 0 50.0 0000.0 2.0 00.0 00.0 0:.0 0000.0 00 :00 A0000.00 0000.00 00.00 00.00 800.0 00000
00.00 00.00 800.0 0000.0 0.00.0 0000.0 3.0 00.0 00.0 00.0 0000.0 00 :00 60.00.00 Av0.00.00 00.00 00.00 500.0 0000.0
00.00 00.00 2000.0 0000.0 0 Bob 0000.0 2.0 00.0 00.0 M0 080.0% 00 :00 60.00.00 o\w0.F0.00 00.00 00.00 .000 P 0000.E
0. : 0.: :03: :90: :03 > :03: > 000010..“ 0:01.01 301(40 :0: m: n: :03: ..: :90 000:: 00.02 0 : 0: :03: :03:
0:330: 00 23:60: 0.0 0:330: 00 0:10:00: 00 «(mi 0:3: 0:00:00: 00 5030: 00 13:00: 00 0:00:00: 0.:
EDGE 950:: MOSS: 9030:: :19: 08 mom—0:: £39: 550:: £39: £39: 530:: 930:: 0.859: 330::

 

0: 0 :3: :0: 00000320“: 0000050000 002:: 030:: 05 053000 .0000 000000 :00xm: ”3:00: Wm 200:.

 

 

 

 

 

 

 

 

 

 

 

 

 

:
31:85 00 0.80:: 88.: 88.0 8.5 0.8.8: .50: .2. 8:8 L205? 0008;
ENE. 883 00 82...: 28.0 ~82. 8.0.0. 0.8.8 00200:
3.0.8.0. 082.0 00000: 88.0. 88.0 3.56 5.: $8.: 00 8.0 .00.: 00
m L0 0.3.: 28.0 88.0 8.: 0.8.: 00 8:: 6.8 $60000
£20 04:88 002—2 055: 2000000 000 @300 0 .0000 2:: 0:0: 0 .0050
_ _

 

 

 

 

 

 

 

 

 

 

 

 

 

0:032:00 .0 1200?: 002:0 .00: . 0 2: :0 20002d :0:::00: 00:00:00 000 2: :0 0:0:0 00:00 :0 00:00:00 002:0 .00: 0: 00 :0 00:00: :0 0 2:: 0: :03: 00 0003:

 

 

9:50:00 .0 \ «:0: 002:0 .00: 00000: 2: :0 .020 .000: 00:00:00 000 2: :0 0:20 00:00 :0 00:00:00 002:0 .Nm: 0: 000000: :0 00:00: :0 0 2: 0: :02 0(000: 000$:

 

900.0000 002020: 002.:0 i0: . 00 2:: :0 .020 .000: 00:00:00 000 2:: :0 20:0 00:00 :0 000:0:00 002:0 .00: 0: 00200.: :0 00:20 :0 0 2:: 0: 00:: 00 0003:

 

00000 0:00: 00: :0 20:00:. :0000: 00: :0 200:0: :20: n 00:3: 00000 0000: 00: :0 03300:

 

 

002.00 :0 22:00 00.:: 05000 .5 :o 82.. :02 u 08:: 82:0 032.: 20 :0 000.00:

 

«0.0.20 n0 00:00 2:: 00 0220200: 200:0: 00:20 :20: 2:: :0 0000.00 0020 00:20 2:: N0 00:00:00 20000: 2:: :0 26200

 

 

 

00000:: 0:: :0 000::00 000:0 00:20 200000200 0000 2: :0 0:00 :02: 020030 «20000: 60.00: 200000.000 2:: :0 200:0: :0:00: 2:: :0 :50

 

«000000: :0 200020 0:00:00 :0 200020 :0:::00:.:0:0:0 0: :0000 0: :: .00 :0 200020 :0:00 2:: 0: s: < 60$.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000000: :0 00:02:20. «000000: :0 5:: u N00000: :0 200020 :0:000 2: 0: E<.00 $3000:

30.0 :N 5P9 00000.0 00 _ 00 :0 00:02:05.3“: :0 0::u 00 :0200020 :0:00 2:: 0: 32¢: N00 00

000.wa 5.00:. 0080.0 0:000: 00 0: 00000 3:030:00 000000: :0 200020 :000: 00: 0: 2.50.0: 050000:
0 0 .20 2.28.: 0... 0.3.8. 3: .0 .55: 2.25.0200 02.0.51 _ _ : _

«0+ :0. u~< 0.: 20520000 8020000 0:00:00 00:00 :0 02:00:00 22200.00: :0:00 :00 2:: :0 00 :0 200020 :000: 2:: 0: :20. 0: 00

_ _ _ _ _ _ _ .000 :0 0:00 0: 200020 00:02:00§

E0002: :0 30022: «03:00:: 00000002: 200020 2:: :0 00:20:00 2:: 0: 0520000 200020 00:02:00 0: (:00 .0 0000

_ _ 20200.00: :0:00 :00 00:0 00: 0: :MMIIE.

 

158

00.3‘0200: eagohq 3.00000 0:: :0 :DOUQOL on: 0‘ 10:94th

 

00000002: 200020 :2: 2:: :0 :00002 2: 0: kmmmfi

 

 

 

Foowoo @008. o Npooo. o 8:... fio owmwo 0056 wow: nod mmoué pmmoo woos.— Lmlhvm... 5.0 «Kb 00.:.me 5.09 8.0.

 

vmowod 808. o N 58. o 8N fio own 9 .o 00 ..oo coop mod «mom; «mood mmmmé 80m... and 0.8 «0.09. whoo— 00.0.

 

cmowoo «008.0 N 58.0 vow —.o a 5 p .0 mo 56 $62 was.” vuow... owmoo Saw; movmé 3.0 0.8 8.2: whoop ovo.

 

vww —o.o hemooo oooood 3.0 «d anpo n 30.0 Emo— mwé «ammo m0 8. o 35.0 mnmoo 8.0 o.~n made No.2 med.

 

v 5 —o.o 0886 @8006 :5 fio «mm P .o :. 39o Ema: vwé mowmo mv 5. o 50.0. o 030. o 8.0 h. K undo: 00.00 36.

 

o 5 poo 088. o oooood va P .o NNN F .o a 3:06 0.69 0N0 —~No.o 3.. so mono. o wmmo. o 8.0 h. 7:. undo: ohms N :6

 

5m 56 mm —oo. o mooood mum _..o vowfio 0 506 ad: and Sand ooooo vomm. 0 on 5. o med 0mm moém o flow 00mm

 

Fom—oo $506 0886 mum—.o 08—6 93:: «db— Nod vvmmd 886 300.0 «056 mod Nam $.00 o—dm «Nam

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 5.0 8 :8 o 88.0 22.0 :83 m :86 «.2: 8.0 8.8.0 88.0 28.0 8 .2. 8.0 08 8.8 2.8 2.8
805.0 R80. 0 .886 236 8 : ..o N :80 0.00.: 8.: 83¢ 88.0 82.0 :28 8.: :0 8.8 0.. z. 2.8
8: 8.0 :88. o .880 m z. 3 8 : ..o 0. :80 Q2: 8.: :88 88.0 82.0 82.0 8.: 3.0 8. 8 8. z. 2.8
I005: #080 .880 23.0 S P ..o m :80 0.2: :0: 23.0 88.0 :88 «:2. 0 4mm.“ :0 8. 8 :0. S 5.8
000 : .000 : 00:: : .0 ...: .: 0: 03:0} : 0:0 .00 0:0 .00 0:0 .00 :5. .m: :5. .m: :5. .0 S00 .0 60.2 002ml. 00020
00 00000: 00 E000: E000: 200200 F00 00 00000: 00 05:00: 00 0000
I002: 00%.: 0000: 0:0:I0M

 

m: 0 :3: :0: 000003200 00000000000 00000 20:05 05 2:325 325 002.5 3020 ”3:00: w-m 0300.

 

159

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2000200200005 2: 0: 000.0500 00:5: 05000: 000 00 .00 05:0 0050020 :0:00 2:: :0 500 2: 0: 3:: < .0 0:30

 

: 00 :0 0282:3508 :0 00:» 00 3200020 00:00 2:: 0: 0E: .00 00

 

0:250: :0 00:00:: :05. @0000: :0 0>:nm:0000: :0 2000232000 2:: 0: 3:0 .00 03:00:

 

00 :0 02:02:08.5 :0 0:? 00 :8:=00o:0 00:2: 2:: 0: :23: 3.: 00

 

0:00: 0: 0:00:50 :0 00:00:: :05 0n: 0: n: 40:: N0 050::

 

0:00: 0: 0:00:00: :0 00:00:: :2: 05 0: n: :0: 5:000: 0:003

 

0:00: 0: 00800.: :0 00:00:: :05 0:: 0: u: 40:: 9.0 0:003

 

0:00: 0: 0:00:50 :0 2:00:00 2:002 05 0: 02:. N0 050::

 

0:00: :0: N508: :0 :0020WES02 0:.:: 0: 20:). 10.000: 050:.—

 

 

 

 

 

 

 

 

 

 

0:::0:: 0: 00800.: :0 2:00:00 2:90): 2:: 0: «at: 8 0:00:.:

 

 

2:00:50 :0 50:00. 0:000:05 05 5:3 002:0 0:00: 2:: 0: 0:0: 50 :0 00:05 :0 :0050: 2: m5:

 

:0:00... :0 08:22. 32:0 030:: s 80:50 :0 H000 2:: 0: 0 00 0:00:

 

$0000: :0 50:00. 5:000:05 2: 5:2 002:0 0:00: 2: 0: 0.0000: :0 00:05 :0 :00500 2: 05:00:05 :0 005200 002:0 0:00: 0: E0000: :0 50:03 2: 0: 0 5:000: 030::

 

fl 5200 :0 50:00. 0:000:05 2: 5:0. 002:0 0:00: 2: 0: 20:00.: :0 00:05 :0 :00500 2:: m0::.:0:::05 :0 005200 002:0 0:00: 5 20:00.: :0 30:2: 05 0: 0 00 050::

 

_

Ill

.00 :0 00:05 :0 0 :20: 2: 50:: 002:0 500:. 2:: 0: 2:00:50 :0 00:05 :0 0 2: 05005000 :3 005200 002:0 0:00: 0: 0:00:50 :0 00:05 :0 0 05 0: :03: No 050::

 

90000: :0 00:05 :0 0 :20: 05 50:: 002:0 :0000 05 0: .0008: :0 00:05 :0 0 05 05005000 :0 005200 002:0 0:00: 0: 90000: :0 00:05 :0 0 2:: 0: 00:: 03:00.: 050::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00300 :0 00:02: :0 000500 :50: 2:: 50:: 000.:Q :0Q0) 2:: 0: 00200.: :0 00:02: :0 000500 2:: %fiknau NO 000:0:00 0000:: 0:00: 0: 0:00:09: :0 00:02.: :0 .8050: 2:: 0: do: 8 Q::.:°:-:
$3 63 $3 .0: 83 $033 .833 :83 .:.$3 0.9.3 .33.. $3 $2 8.: $83 283 $63
$3 $3 33 a: $3 383 $83 3333 0.303 .303 .33.: $3 $2 8.: $83 283 $83
$3 $3 $.0 .0: $3 3233 $33 383 .:.$3 .303 .33.: $3 $2 8.: $53 283 $63
$3 83 $3 $3 53 $83 $23 233 0.333 0.3.3 .33.: $3 $2 8.: 283 $83 $83
$3 $3 $3 $3 83 .233 $333 283 0.33 0.33.3 0.33.: $3 $2 6.: 283 $333 $53
$3 $3 2.. $3 :33 $33 $83 3:83 0.33 0.3.3 .33.: $3 $2 6.: 283 $83 3:33
mod mmd No.0 and 5.0 nmwod 500.0 N 50.0 2090.0 oxamnfim o\omo. _. mmd 8.00 5. w oomod moved 9. p 0.0
$3 $3 $3 $3 :33 383 3.33 283 .:.$3 .:.$.3 .33.: $3 $2 6.: 833 333 $53
$3 $3 $3 $3 :33 $83 333 283 .333 .:.$3 .33.: $3 $2 5.: 3833 3:03 $63
$3 $3 33 23 83 $83 $33 :83 0.33 0.33 0.33.: $3 $2 6.: .233 3.33 3.63
$3 $3 $3 23 333 $333 $33 :83 .383 0.3.3 0.8.: $3 $2 3.: $83 3:33 .333
$3 2 3 3.0 23 83 $333 $33 :83 0.33 0.3.3 0.33.: $3 $2 3.. $033 353 .333

:2“... :5 .0 :20 0.: 0.5 3.: 05 3.: 0 .0: 0 :00 ..: :00 .5: :20 .20 0 .: 0 .: 0 : 00:: :00: :00:
03.0.: 0.: <0 00 05:00: 00 :0 05:00: 00 00 0.0.00: 00 00 50:00: 00 00 05:00: 00
2:200. @003 E00: E00: E00: P020 2003 @003 K00: 2003 R00: @003 0:00:

 

m: 0 00m :0: 000020230 00030000000 32:0 2005 05 mEBQE 82% 00025 Boxm ”3:00: w-m 030:.

 

 

160

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _ _
:22... 00 22.8. 88.: sum? 8.0: 0.8.8: .30: E 8.50 @239 ..mwmml>
3.0%? 802.0 00 22.8. 88.0 B80 5.: 22.2.. 02:00:
3.02.0. 022.0 22090: 38.: 85.0 88.0 8.: 28.. 00 8.0 2.0 00
0 0%: ~82. 82.0 mo. 8 2.0 28.0 00 8.: 02 5:00:00
£2": 8:28 00%: 2% 20:02:: 000 00000 0 .000: 0:: 0.95. 0 .002:
_

 

A

 

$5.88.: \ «050:0; 002:: dds . o 2: :0 20000:: 520:: :0:0000 000. 2: :0 0:20 0500 :0 005200 .

 

52:: :00 :: N0 :0 00:2: :0 :: 2:: 0: 40:: N0 t0:<>

 

«5.5.88.0 \ :00 002:: :2. . :0:00: 2: :0 00:88:: :0:0000 00: 2: :0 0:20 0500 :3 005200 002:: :00 5 m:0:00: :0 00:2: :0 0. 2: 0: 40:: :0:00: mo:<>

 

~5b00~00x~0550> 002:: :00 . 00 2: :0 00:: 00:: :0:0000 00: 2: :0 0:0:0 0500 :3 005030 002:: :00 5 0:209: :0 00:05 :0 a 2: 0: .55 00 :O:§

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ 002:: 0:00: 2: :0 2:59.. :0000> 2: :0 050:9. :20: u «:23: 002:: :0:00 2: :0 $5300

_ 3:0:0: :3 0000.00 0:00: 359:0 2: :0 002: :20: n «:23: 002:: 0:05.: 2: :0 m5: .5:

Dm+< u: :0:00 0 2: A: 00:202m: 200:0:— :0:0:: :20: 2: :0 00:30: 002: :0:0E 2: :0 005200 0:002: 2: :0 xtmzm:

50:00 0:: :0 0000.00 002: :0:05 25:02:00 :000 2: :0 500 50:: 005200 3:0:00: 60.00: 2:000:53 2: :0 2:500 :22: 2: :0 =50

3:0:00: :0 200008333100 :0 20000:: :0::0::.:0:0:: 0: :0000 0: :: No :0 20000:: 3:20 2:: 0: :21 :0: No

m:0:00: :0 :0:00:::0E. 3:0:00: :0 :0: n m:0:00: :0 20000:: :53: 2: 0: 5.210: N50200:
:02: 5.002 088.0 00 _ 00 :0 02:02:05.8 :0 2:" 00 30332: :0:00: 2:: 0: :3

08.80 5.00: 0080.0 :0:00: 00 0: :0::E:0 00:23:00 m:0:00: :0 200023020. 2: 0: s: <.:> 2:00:00:

0 0 “W0 3... 2.200.: Emacs 3: .0 .25 00250200 02.05: : A _ _

~0+t\m.<n~ 0.: 0:d.::0:0000 2:20:00 050::< 0.500 :0 0220200 0:0:0:.0|::|:0: :50: :00 2: :0 00 :0 20000:: :0:00 2: 0: 5.3: .:> 00

_ _ 4 _ H _ _ :50 :0 0:::0 5 20000:: 00:00::00flhlclm

2:95: :0 3000:»: «03:52:: :0000002: 20000:: 2: :0 :0::0:0::00 2: 0: 0503000 20000:: 00:00:00 0: $0: .: 5:00

Jr : _ 0:220:50: :50500 :000 2: 0: 59:.

:0000002: 20000:: 000000 2: :0 :00000: 2:: 0: “GE

:000:0:0:: 20000:: :0::: 2: :0 :00000: 2: 0: .500»:

528.0 808.0 28.0 :08 088 858 0.05 3.0 Rub. : 008.0 88.: 88.— 6.0 0.09 8.8: :83 vmd
088.0 0 38.0 208.0 m 68 088 00:08 0.0: 00.0 008.: 008.0 0:8.— mo~0.: 00.0 0.09 0 £9 :0. SF «No
088.0 «88.0 088.0 808 508 008.0 .85 8.0 «000.0 858 38.0 008.0 00.:. 0.8— 8.8: 3.0: 8.0
808.0 2&88 088.0 «00 8 608 008.0 :82 8.0 800.0 85.0 500.0 008.0 00.:. 0.8: 00.03 8.0: vwd
088.0 038.0 880.0 «8 8 88 008.0 No: 8.0 800.0 mm 5.0 0088 0808 00.:. 0.8: 3.8.. was: cud
8088 8 :88 0888 .30 8 o 58 ovmwd 0.00: 00.0 888 0088 800.0 8 5.0 3.0 2.0 00.00 8.8 «8
808.0 00 :88 088.0 .30 8 o E 8 008.0 0.00: 8.0 88.0 008.0 0000.0 85.0 8.0 0.8 5.00 «:00: «8
0008.0 00 :88 88.0 000 8 m E 8 300.0 0.00: 00.0 0808 008.0 R008 35.0 00.0 0.5 8.00 8.02 N8
«0088 28.0 808.0 ~98 8: 8 85.0 N. we: 8.: 0008 88.0 088 80.8 5.0 0.8 «0.00 8.2 «8
Emma: 888 68.08 «00 8 8: 8 88.0 m. 8: 00.0 «on 8 .388 4.808 «mm 8 I000 0.8 8.00 J 008:. end
.50: > 00:: :. .53: 0 .: .00 .: .S .05.: > 5.; .0: 5:: .0: 5.20 q0: 32.1 NS 5.5: .:> E< .: Sm: .: ENE. :00»: 025::

«o 50:90.: 00 E400 E049: :Crmkm: fin: no :<:o.m:: 00 530%: 00 E00

L033 :0:§ :0:<> 20234:.

 

E :: =3: :0: 80:28:00 8300:0080 002:: 0:85 05 852$ 02% 002% :00xm 5A: 033..

 

161

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

20000:: E05002: 2: 0: 00900000 02:... 22:00: 0:0 00 .«0 2: :0 00.0.0002: .222: 2: :0 :50 2: 2 SE: .1 0.3.0

 

_

_

«0 :0 02.02.352.00 :0 0...». «0 .8308... .2000 2.. 0. E: .00 «0

 

90:00: :0 00:00:: 3.0.3.300: :0 manning: :0 03000:... .030: 2: 2 3.; .00 59:00:

 

00 :0 02.000.05.80 :0 0..." 00 .8582: .200: 2.. .... E: ...0 00

 

330:: S 0.3350 :0 52:00:: :05 05 a: l 40:: N0 050:

 

03.... ... 2008. :0 02.02: .2: 2:. 0. ..: ..0: 0:000. 0.00...

 

3%: E 23:09:. :0 52:00:: :02: 05 a: u: do: no 9303

 

03.... 5 000......» :o .0802. 20.2. 2.. 0. ...0 «0 0.00...

 

0.2... 0. 2008. :o .0080: 30.00 2.. 0. 2:. 0:000. 030...

 

 

 

 

 

 

 

 

 

 

 

030.. ... 0:23-. :0 .0800: 0.0.0.: 0:. 0. a... 00 050...

 

000350 :0 30:2: 22:020.: 2: 5:... 00003.30: 2: 0: 0:2 0:0 :0 020.: :0 .0050: 2: m0:

 

.003... .0 000.030 000:: 030.. 0. 000.050 .6 .039: 2.. 0. 0 «0 0:00...

 

9&9: :0 2:20.. :0:000:20 2: 5:: 002:: 202: 2: :: H.230: :0 020.: :0 32:0: 2: 9.32:2: :0 300.030 002:: 20:: 0: 92:00: :0 230.: 2: 0: m 5:002 930::

 

0:200 :0 220.: :0:000:20 0.: 5:: 0000.: 2:2: 2: :: 00200.: :0 020.: :0 32:0: 2: 0:32:30: ...0 00:20:00 002:: 2:2: 0: 00200.: :0 0:20... 2: 0: u 00 2:0:

 

 

.NO :0 020.: :0 ... :20: 2: 50:: 000.3%00 2: 0: 000150 :0 020.: :0 ... 2: 3:00:00... .3 000.200 002:: 2:2: 0: 2.25:0 :0 020.: :0 ... 2: 0: .53: N0 930::

 

9.0000: :0 020.: :0 ... :20: 2: 80:: 002:: :0:00 2: 0: 22:00: :0 020.: :0 0 2: 05.02300 :0 000.200 0000.: 2:2: 0: 92:00: :0 00:2: :0 ... 2: 0: 40:: 5:102 052.:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Otbuuo ‘0 Q20:— \0 kOQEDC inc» 0:: 59¢ OhQCQ LOQQ) 0C: a: 0:200... \O MO~OE so :OQESC Ont MC-OQ§Q=Q Mn przfiuho OMQCQ B‘s-U: a: OCG~OOoh .3 OO~OE ‘0 :OQEDC on: Q: 40: $0 E~d
.0... .0... 8... m... 8... 0...... 8...... as... $2.“ «.50.. .03.. ....m 8.: «0.. 2...... 0.8... 00......
...... 2.... a... N... 8... 0...... 88... 3...... 0.20 0.0.0.0.. 2...... ....N 8.... «0.. R8... 28... 00......
0... 0... 8... ...... ...... ...... .... .08... .... ...... ......00 ...... ...... $3.. 0«.... .0... «0.. 08.... 88... 00......
.... c... 8... 00... ...... 9...... .000... .... ...... 0.2.0 ......8 0.3.. 8.0 00.: «0.. 08.... 88... 8......
0:. 0:. ...... 00... ...... 8. .... 00...... a ...... 0.8.... ......00 o...... 000 00.... «0.. 88... 88... 8......
.....0 3.0 8... 8... ...... 0.. .... 88... 0.. ...... $80 ..00.... 0.3.. 00.0 2.... «0.. .08... 28... .... ......
8.0 8.0 00.0 8... ...... 0...... .80... 8...... ...0...“ $8.00 00...... .0... 2.: am... 008... 28... 8......
:96 wad mod and 5.0 how pd omood mm 8.0 .xme $uo.mm o\ovm. _. m0.~ 053.. um... omwod whom... on 5.0
5.: 5.: 50.: v — .o 8.6 mam — .o ohmwd Na 5.0 o\oN w .m o\ovm.vm o\ovm.w ovd when mm... onwod omowd cm 5.0
...... ...... 00... ...... 8... 08.... 20.... N05... ...0... 00.0.3 .....0. S: 2.... 00.. 5...... 88... 8 .mwl

E: .0 E: .0 3...: ...0 E: ...0 E: ...0 0 ..0... ... ..0: ... ..0... 0...... ...0. E 0 .. ... .. 0. ..00. 00:: 00....
0:0... 00:0 0.0 0:000. 00 «0 0:000. 00 «0 0:000. 00 8 0:000. 00 «0 0:000. 00
0.000... 0.....0... 0.00... £00... £9.00... R00... 2.00... E00... 0.50... 0.00... 2.00... 0.00...

 

 

E :0 :3. :0: 002.0320“: 02.2.08on 00...... 2:3..— 0... mcmaonm .028 008.5 30%. ”3:00. m-m 030,—.

 

162

 

 

_ _ _
:m:r0...0 00 82.8 88.. 008... E0 088. ..0:0: 0.58.08 M239 .8008

 

 

 

 

 

 

 

00.2.0. 88:... 00 .088 82.0 8.0.0 8.0: .0.0.8 00000. _

8.88.... 88:... 0.000. .22.. 8.0.0 8.0.0 ...... .08.. 00 8.0 00.0 00
0 LMI 88.: 8.8.0 28.0 8... 6:8.: 00 8.: 0:: 0000000

.028. 00:88 0080 050... 2000000 .00 00.00 0 .0000 0:0: 020. 0 .000...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

65.0000: \ «200:9. 0023 is. .00 2: :0 200003 :0:03: 00:00:00 000. 2: :0 0:0:0 00:00 3 000:030 0023 i§ 0: «0 :0 00:00: :0 ... 0:: 0: 40:: NO 1900.:

«05.00006 \ «:00 0023 is. . :300: 2:: :0 .020 03: 00:00:00 00:: 2: :0 0:0:0 0500 3 000:030 0023 is. 0: 2300: :0 00:20 :0 0 2: 0: .6: 5300: 0003:
9530006305030 0000i is. . 00 2: :0 .003 .:0 : 00:00:... 000 2: :0 0:0:0 :0:00 3 000:030 0023 is. 0: 00200.: :0 00:20 :0 0 0.: 0: .53: 00 09:3:

 

 

 

 

 

 

 

 

_ _ 0000:0030: 0.: :0 0:00:00. :00000 2:: :0 200:2. :20: u «:0::.:: 0000i :03: 0:: :0 90:30:.

4 _ 0.0000 00 0000.00 0.00:: .05000 0.0 :0 80.0 :00: u 00.:: 002.0 0.00... 0.0 :0 0:00:00
DQ+< :6 00:00 2:: 3 0220200: 200:0: :0:00: :0:0: 2: 3 0000.00 0000: :0:00: 00: 3 005030 20:50: 2: :0 3.52%:
50000 0:: :3 0000.00 0000: :0:00: 20003500 0000 2: :0 E00 50:: 005030 «22:00: 60.00: 200002000 2:: :0 200:2. :0:00: 2: :0 35m

 

 

«0:300: :0 200003 :0:00i:.«00 :0 200003 :0:00i:.:0:0:i 0: :0000 0: :: No :0 20002:: :0:03 2: 0: 3:: 1 «oi «0
2300: :0 00:00:::0.0.«m:0i00: :0 i0: .0. m:0i00: :0 200023 :0:00i 2: 0: :E. (6:: NERO“:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.888 8.08. 880.0 00 _ 00 :0 00008000080 :0 0:0... 00 :0:000:20 0.000 2:: 0. .20 .00 00

03.08... 5.00: 083.0 00000.: 00 0: :0::E:0 0220200 20000: :0 200003 :25. 2:: 0: 2.3.0:: 53000:
0 m < .20 2:800.— 000 W000: :0: .0 it: 3.23.0200 mz_o.pz<

«o...:.::m.<u« 0.: 0: 0050:0000 0:00:0000 050:0... 0500 3 0220200 20:20 .00: :50 :00 2:: :0 0o :0 20002 :0 s. 2:: 0: 3:8. ...0. 00

fl _ _ _ _ _ _ .53 :0 0:00 0: 0000003 00:02:00fl

3:33: :0 30022: «0:.in: :0000002: 200003 2:: :0 00:03:00 2: 0: 0503000 200003 00:02:00 0: 30...: .i 5:00

fl _ 20:22:80: :53 :00 0000 2:: 0: EMF

 

00000000: 200003 000000 2: :0 :00000: 2: 0: i000
3000002: 200003 :05: 2: :0 :00000: 2: 0: :mmm:

 

 

...:
.Q

 

 

 

880.0 088.0 088.0 08.0 8.8 82.0 :8. 8.8 88.0 88.0 28.. 8.0.. 8.8 0.08 88. 8.8.. 8.08
880.0 888.0 088.0 08.0 8.... 82.0 :8. 8.8 88.0 0.8.0 88.. 800.. 8.8 0.08 88. 0.8.. 8.8..
880.0 088.0 088.0 08.0 8.8 82.0 :8. 8.8 8.00 0.8.0 88.. 28.. 8.8 :88 ...8. 8.8.. 8.8..
088.0 8.8.0 088.0 08.0 8:0 82.0 :8. 8.8 020.0 ...0... 88.0 82.0 8.8 0.08 8.8. 8.08 2.0.8
088.0 8.8.0 .88... 08.... 8.8 82.0 :8. 8.8 02.8 ...0... 020.0 08:... 8.8 0.08 8.8. 8.08 2.08
088.0 8.8.0 A0.88.0 08.0 8.8 82.0 :8. 8.8 .200 28... 88.0 38.0 8.8 .08 .:.8. 8.08 2.08
:88... 2.8... 888.0 08.0 8.8 82.0 :8. 0.8. 88.0 88.0 :08... 2.8.0 8.8. 0.08 8.00 00.08 828
280.0 2.8.0 «88.0 08.... 8.8 82.0 :8. 8.0. 88.0 88.0 88.0 8.0.0 8.0. 0.08 00.8 00.08 8.8
280.0 2.8... 088.0 08.0 8.8 82.0 :8. 8.0. 28.0 88.0 88.0 38.0 8.0. 0.08 .08 00.08 8.8
220.0 880.0 .880 08.0 8.8 82.0 :8. 2.0. 8.8 0.8.0 80.... 82.0 8.0. :88 8.8 8.8 8.08
820.0 .88... .880 08.0 8.8 82.0 :8. ...0. 8:... 0.8.0 80.... 82.0 8.0. 008 8.8 8.08 8.08
0020.0 0800.0 .8800 08.0 8.8 82.0 :8. 0.0. 8:... 8.8.0 82.0 k2... i8|2 0.08 8.8 P000: 00:8
00:... ..00 s. ..00 > .: .3. .. .0. g 0 0:0 .00 0:0 .00 0:0 .00 :00 .s. :00 .0: 000.0 .000 .0 .08.: 000:0 8.00:0

00 00000. 00 E000 E0000. 2.10.30 #00 00 00000. 00 00000. 00 0000
..lmo0§ 000.0. 0003. 0.0200

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m— t. 63% :0.“ mflOUfl—SO—QU GOUNbGDOGOU Owasm 6:37—1— 05 wCMBOa—m Hoop—m Gavan” ~00Km ”O: ..m 3an

 

163

 

 

 

 

 

 

 

 

 

 

 

 

630020 0200020 2: 0: 00030000 003:: $0000: 000 00 .00 2: :0 00.50020 5000 2: :0 0:00 2: 0: SE: .0 0.30

 

00 :0 02:80:05.3 :0 0:0 00 30:082.: 5000 2:: 0: =5: .00 00

 

E0060: :0 00:00:: 3003:0000: :0 0003:0000: :0 200020 2000 2: 0: 3:50: .00 5:000:

 

00 :0 00:08:05.5 :0 00:... 00 3000023000: 2:: a: 0:0 .00 00

 

0:00: 0: 0:25.50 :0 00:00:: :00: 2:: 0: ..: 40:: N0 930::

 

0:00: 0: 03000: :0 00:00:: :00: 05 0: k 40:: thOm: 630:.—

 

USE: 0: 23:00.: :0 £02000: :05 2:: 0.: n: 40:: 00 930:

 

030:: 0: 30:00. :0 08:00 30:2. 2:: 0: 00:. 00 0:00:

 

0:00: 0: 0:008: :0 002.0 30:9... 2:: 0: :00 00000: 0:00:

 

 

 

 

 

 

 

 

 

 

 

:0:00: 0: 00030.: :0 3003:3032: 2:: 0: $3. 00 030::

 

002050 :0 30:03 :0:000:20 2: 5:0. 002:0 0:00: 2: 0: 000: 50 :0 00:00: :0 :00500 2: 00::

 

:0:00: :0 005030 32:0 030:: 0: 22.000 :o 30.00 2:: 0: 0 00 0:00:

 

E0300: :0 36:00. :0:000:20 2: 5:0. 002:0 0:00: 2: 0: M00000: :0 02203 :02000 2: 00:30:30: 50 005030 002:0 0:00: 0: M00000: :0 30:00. 2: 0: 0 5:000: 0:000:

 

00200 :0 30:03 :0:000:20 2: 5:: 002:0 0:00: 05 0: 00200.: :0 00:00: :0 :00500 2: 00:30:33 5:: 000:030 002:0 0:00: 0: 00200.: :0 30:00. 2: 0:0 00 050::

 

 

.00 :0 00:00: :0 0 :20: 2: 30:: 002:0 :0000 2:: 0: 002050 :0 00:00: :0 0 2: 0032300 50 005030 002:0 0:00: 0: 00233 :0 020.0 :0 0 2: 0: 40:: 00 050:.—

 

m:0000: :0 00:20 :0 0 :03: 2: 50:: 002:0 :0000 05 0: 92:00: :0 00:00: :0 :__ 2: 0502300 :0 000:030 002:0 0:00: 0: M3000: :0 00:00: :0 0 2:: 0: 40:: 03:00.: 030::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00200 :0 00:00: :0 :00500 :20: 05 50:: 002:0 :0000 2: 0: 00200.: :0 00:20 :0 :0:0000 05 0502300 :0 000:030 002:0 0:00: 0 _ 00200.: :0 00:00: :0 :02000 2: 0: 40:: 00 030::
modN 666N :66N 06.: 666 mom :6 6 :666 65:66 $65.: $ :666 $66. : on.” N :65 mm... 656 : .6 56 :66 66 :66
666N 66.6N :6.mN 06.: 666 666:6 6:666 65:6.6 $65.: $ 5.66 $66.: 66.6 N:.m5 8.: .0596 56:66 66:66
666N 06.6N oo.mN 06.: 66.6 666:6 6 :666 65:66 $65.0 $ :666 $66.: 666 N:.m5 6m: .0596 66:66 on :66
N66N N66N ovNN :66 :66 N56 :6 :6656 65 :66 $66.6 $6 :66 $66. : 06.0 :665 vm. : :vv : .6 MN :66 56 :66
N6.6N N66N vaN 06.6 :66 :56 :6 5656 65 :66 $66.6 $6 : .66 $66.: 06.0 5.65 06.: :3: : .6 6N :66 56 :66
66.6N 66.6N ovNN om6 :66 :56 :6 565.6 65 :66 $66.6 $6 : .66 $66.: 06.0 N665 vm: :3: : .6 MN :66 56 :66
66.6: m6.6: 56.6: 6N6 :66 6:6 :6 6565.6 65 :66 $VN6 $56.N6 $66. : mN.v N065 06.: o :m : .6 N6 :66 56 :66
.066: 66.6: 56.6: 6N6 :66 96:6 @5656 65 :66 $0N.m $56.N6 $66.: mN.:. N065 0m... 6:6 :6 N6 :66 56 :66
66.6: 66.6: 66.6: 5N6 :66 606 : .6 6565.6 65 :66 $0N.m $56.N6 $66.: 6N.:. N165 06.: 5:6 :6 No :66 56 :66
6N6: 6N6: m :6: N:6 66.6 666N6 66556 05:66 $666 $wm.N6 $66.: :90 60.65 06.: 666:6 on :66 56 :66
6N6: 6N6: o :6: N : .6 66.6 666N6 5055.6 05 :66 $66.6 $wm.N6 $66.: :60 60.65 06.: 566 :6 on :66 56 :66
0N6: 0N6: :6: N:.6 F666 666N6 @655b 05:66 $666 $wm.N6 $66.: 5.0 60.65 3.: 666 :6 66:66 56:66

EQ C 3.51 :0 EQ _0n: EQ 00:: =5< ~00 n: 40: ...: do: u: 40:: E E E u .: 0 .: q: 403: 40:: 403:
m<m§ 0410 NO t<009 00 NO 5:969 00 NO ”300%: 00 N0 ”0.000%: 00 N0 RS002 00
@003 P00: @003 P00: 0|:00: £5.03 0|:00: $.00: 5:500: @003 .0500: 0:00:

 

 

.2 0 =3: :0: 000003200 00000000000 002:0 0:005 05 wEBonm 0026 00800 302.0: ”3008 6: ..m 030:.

 

164

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ : : :
$.30... 00 8808. 88.: 22.... 80h: 0.8.8: 00:0: .2800: mza..n> .000?
amt... ~82... 00 92.8 :53 3:3 8. K .0008 00:00:
vomlmwo. 85:... 0.000: :8.: $8... 23... :3 0.2.0 00 8... 8.0 00
0 <_ 8.3 .08.... :33 8.0 0.8.0 00 8.: 00:. 0:00:00
5+0": ::.—0.38 0.4% 2.. :0... 20003”: .00 00.00: 0 .0000: .00: 0.0:. 0400.0:
_ _

 

 

 

 

 

 

95.0000 6 \ «0:00:00 002:: :00 . 0 2:: :0 20000::

:0:::0:: 00:00::0 00: 2: :0 0:0:0 00:00 06 005030 002:: :00 0: N0 :0 00:00: :0 0 2:: 0: :03 «0 20:3:

 

O:l:«:.00~0 6 1:00 002:: .30.. :0:00: 2:: :0 .02: 00:: 00:00:00 000 2:: :0 20:0 60:00 ::.: 00:00:00 002:: :00 0: 0:0:00: :0 00:00: :0 0 2:: 0: :00: m<:om: :0:00:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0:052:00. 2:200:00 002.:: :00 00 2:: :0 .02: 00:: 00:00:00 000 2: :0 0:0:0 :0:00 :0 005200 002:: .50. 0: 00200.: :0 00:20 :0 0 2: 0: :00: 00 @0500
: _ 002:: 0:00: 2: :0 0:00:00. :00000 2: :0 00:0:00 :0:0:I I 00:3: 002:: :0:00 2: :0 $00400
H _ 0::0000 0.0 0000.00 0:00: :00:0::0 2:: :0 0000: :20: u :23: 002:: 0:02.: 00: :0 00:: :00
D03: 00 00::00 2:: 0.: 0203200: 0:00:00 :0:00: :20: 2:: :0 0000.00 0000: :20.0 2:: 0.0 005030 20000. 2: :0 0.20200
:::0000 0:: :0 000:000 0000: :0:00: 200002000 0000 2:: :0 0:00 0.0:: 005030 «0:0:00: 60.00: 200003000 2:: :0 0:00:00 :0:00: 2:: :0 2:0
38:00: :0 20000:: 00:00:18 :0 20000:: :0:00::.:0:0:: 0: :0000 0: :: .00 :0 0:0002: :0:00: 00: 0: 0:... .0: N0
00:3: :0 00:00:00.0. «0:0:00: :0 :0:n 0:0:00: :0 20002: :0:00: 2:: 0: ::.: <6: 05:00:
::n.m:N 8.86: 8080.0 00 _ 00 :0 00002000080 :0 :0:n 00 :020002: :0:00: 2:: 0: ::.: .0: 00
avoNNN 860:: 86006 :0:00: 00 0: :0::E:0 00:0:00:00 0:0:00: :0 20002: :0:00 2:: 0: ::.: < :0 05:00:
0 0 .2. 2.28.: .2... 02.8. a. 0 ...0: 0550200 W265: _ _ : :
0+::\N.<n:m::uo.: 0: 0520000 0:020:00; 00:20... 00:00 00 00.20200 20:0:0I:.00: :0:0: :00 2:: :0 00 :0 20002: :0:00 2:: 0: 0:: < :0 00
_ _ _ H _ _ .000 :0 0::00 0: 20002: 00:02:00 i
::.:020: :0 0.002%: «03:09:: :0000002: 20000:: 2:: :0 00:22:00 2:: 0: 6520000 0:0002: 00:02:00 0: :00: .: ::00
: H20:0:0:E0: :0:0: :00 0000 2:: 0: :0E
:0000002: 20000:: 000000 2:: :0 :00002 2:: 0: :03:
:0000002: 20000:: :2: 2:: :0 :00000: 2:: 0: :000::
66806 : ::.86 :386 6:0:6 N: : :6 no ::6 6.8: 8N: NNNo.: 30:6 0:8,: «:8.: 066: 0.08 8.8: mmNNN N66 :N
6NNmod m :08. o 2.806 0:6:6 N: : :6 8::6 0.8: 6:.N: :6 8.: :6606 N66N: 88.: :66: 0.08 No.8: moNNN 66.0 :N
68006 880. o 8086 R0 :6 N: : :6 mm ::6 0.8: 60. : : :686 836 0::06 63:6 8.: : 0. :: : 860: 8.0:: 8.8:
0:636 N886 oNooo6 0:0:6 N: : :6 no ::6 0.8: 60. :: 586 8:66 8:06 80:6 8.: : v. ::: m :60: 86:: 00.8:
80.56 8806 8086 0:6 :6 N: : :6 mm ::6 6.8: :0 :: ::de N636 8:06 8 5:6 00.: : :. ::: 6060: :6.::: :98:
::0006 3:86 0:806 ::.:006 N: : :6 8 ::6 0.8: 8. 6 mml:N6 0086 0806 06 :06 86 93.: :30 66.3.: 00.3:
Imgod 8:86 0:86 0:96 N: : :6 8 ::6 0.8: 86 8:N6 0806 N866 06 :06 86 8.00: 8.00 66.3.: 01:0:
80:66 8:86 o 586 6:96 N:::6 no ::. o 0.8: :06 00:N6 38. o 0:86 8:66 :06 N00: ::.00 560: N:.:o:
66686 8806 0886 0:6 :6 N: : :6 8 ::. o 0.8: N6: 8N : 6 6:8. 0 68:6 6NN:6 8.0 06:: 66.8 56:: 6060:
06686 60086 0886 0:96 N: : :6 09:6 0.8: N6: 8N: 6 6:86 :8:6 oNN:6 86 0.0:: 8.8 066:: 6060:
66686 8806 0886 0|:m:6 N: : :6 60:6 0.8: N6: 8N: 6 6:86 68:6 VNN:6 86 06:: 00.8 56:: 6060:
0:806 0886 :886 0:6 :6 N: : :6 no ::6 0.8: «:6 6N3. 0 :N86 6806 6N86 m :6 066 0:. :0 N06: :06:
8686 :N806 5806 0:6 :6 N: : :6 8 ::6 0.8: 666 88. o 0 :86 8:66 886 86 060 N :6N 606: 00.::
00806 6 586 :886 m.:o :6 N: : :6 8::6 0.8: U006 88. o m :86 vaod 88. o :06 0.010 g mPNI: 6:6:
00:: 0 00:0: 0 :00: 0 .: .00 .: ~:0 40.x: 0 E:<00: 2:1 .0: 2:1 :0: :5: M0 ::.: Ii:0 3:! .: Sm: :: 50>: :00:...I: :mmflF
N0 :0:00: 00 MR: :00 W00: :00 00.3%,: :00 «0 :<:00: 00 :<:00: 00! 5:00
:0:<0 m0:<0 :0:<0 0:023:—

 

 

 

 

 

0: 0 00m :0: 0000300000 00000000000 002:: 0:03:— 0:. $0505 .025 002:5 :0oxm .: Tm 030:.

 

 

 

 

 

 

 

 

 

 

 

 

 

20002: 020002: 2: 0: 0092500 :0:00. 22:00: 0:0 00 .00 2: :0 0200020 :0:00: 2: :0 E00 2: 0: 5.: .1 0.30

 

_ _ 00 :0 02:02:05.:«0 :0 0:0" 00 303820 :0:00 2:. 2 =5. .00 00

 

92:00: :0 00:02: :0E. «000000: :0 m>:um:0000: :0 20002.: :0:00“: 2: 0: 0.2—x .0: 0:360:

 

00 :0 02:80:35.:00 :0 ::.:u 00 323020 2:00 as a: 3:0 .00 00

 

0:00: 5 0:03.50 :0 50:00:: :05 2:: 0: u: .003 NO 9303

 

0:00: 5 0:00:00: :0 20:09:: :2: 2.: a: k 40:: I<lOm< 030:4

 

030:: F: 00800-: :0 =0:005 :05 2.: 0: n: 40:: 00 9:03

 

03¢: :: 23:50 :0 2:00:00 2:902 2:: 0.: 2.0: N0 030:

 

030:: s .2008: :0 38:00 30:0: 05 a: 0...: :0:00: 050:

 

 

 

 

 

 

 

 

 

 

0:00: 5 02200.: :0 300:2: 330:. 2:: m: 00:: 00 950:

 

 

 

0:01.50 :0 30:2: 003022: 2: 5:: 002:: 0:00: 2: :: 0:0: ::0 :0 00:2: :0 :00E:: 2: 0:: :0:0E :0 005200 002:: 030: :: 0:20.50 :0 30:2: 2: 0: 0 N0 0:002

 

22:00: :0 30:2: 020023: 2: 5:: 002:: 0:00: 2: :: 22:00: :0 00:05 :0 :00E:: 2: 0:32:0E :0 00::030 002:: 0:00: :: 22:00: :0 30:2: 2: 0: m 0:350: 050::

 

0:200 :0 30:2: :0:000:0E 2: 5:: 002:: 0:00: 2: E 0:200: :0 00:0E :0 :00E0: 2: 0:000:3E :0 005200 002:: 0:00: :: 05:00.: :0 30:2: 2: 0: 0 00 0:00:

 

 

N0 :0 00:0E :0 0 :20: 2: E2: 002:: 200.2. 2: :: 0:01.50 :0 00:0E :0 ... 0.:[05002300 :0 005200 002:: 0:00: :: 0:00:50 :0 00:0E :0 0 2: 0: 40:: N0 0303

 

22:00: :0 00:2: :0 0 :20: 2: E2: 002:: .602. 2: :: 22:00: :0 00:0E :0 :: 2: 00:00:03 :3 005200 000:: 0:00: :: 20.000: :0 00:2: :0 0 2: 0: 00:: 0:00:00: 030::

 

 

 

165

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0:200 :0 00:0E :0 00:5: :20: 2: E2: 002:: :0:00 2: :: 0:209: :0 00:0E :0 00:5: 2: 05:02:26 N: 005200 0023030: 5 0:209: :0 00:0E :0 02:0: 2: 0: 40:: 00 0303
00.9 00.0F 0R.N F 0F.F F F.0 F0000 R0000 0000.0 0R00.0. 900000 0\..00.0 0R.F «0.0:. «0.0 R000 00:.00 0R00.0
F0.0F F00 F FRNF 00. F F F.0 0000.0 0000.0 0000.0 .0000 e000.00 o000.0 0R. F ~0.0R «0.0 000.0 00R0.0 0R00.0
00. F F 00. F F 00. F F R00 00.0 R0000 0000.0 0000.0 $00.0 $00.00 e000.0 00. F 00. FR 00.0 0000.0 0R0.0 0R00.0
00. F F 00. F F 00. F F R00 00.0 R0000 0000.0 0000.0 $00.0 n0900.00 AU000.0 00. F 00. FR 00.0 0000.0 0R0.0 0R00.0
00. F F 00. F F 00. F F 00.0 00.0 R0000 0000.0 0000.0 0000.0 A0900.00 0000.0 00. F 00. FR 00.0 0000.0 0R0.0 0R00.0
00.0 00.0 00.0 00.0 «0.0 R0000 R F000 0000.0 $00.0 e000.00 °000.0 00. F RF. FR 00.0 0000.0 00R0.0 R000
00.0 00.0 «0.0 00.0 00.0 R0000 RF00.0 0000.0 e000.0 e000.00 $00.0 00.F RF.FR 00.0 0000.0 00R0.0 RR00.0

F00 8.0 00.0 00.0 «0.0 0000.0 050.0 0000.0 .0000 9000.00 e000.0 00F RF.FR 00.0 0000.0 00R0.0 RR00.0
00.0 00.0 00R 0F.0 50 R0000 F0000 0000.0 $00.0 $00.00 6000.0 00.0 0~.FR 00.0 FOR0.0 00R0.0 0R00.0
00.0 00.0 N0.R 0F.0 F0.0 RO0F.0 F0000 0000.0 $00.0 e000.00 n00.0.0 00.0 0~.FR 00.0 33.0 00R0.0 0R00.0
00.0 00.0 00R 0F.0 50 R0000 F0000 0000.0 $00.0 o000.00 0.000.0 00.0 0«.FR 00.0 FOR0.0 00R0.0 0R00.0
0 F0 0 F0 0 F0 00.0 00.0 00F F0 00 5.0 0R00.0 $00.0 e000.00 200.0 00.0 00. FR R00 0000.0 ROR0.0 0R00.0
00.0 00.0 00.0 00.0 00.0 FOF F0 00 5.0 0R00.0 903.0 $00.00 o000.0 Rd 00. FR R00 0000.0 ROR0.0 0R00.0
F00 F00 0R.0 00.0 00.0 FOF F.0 0R F0.0 0RO0W $00.0 $00.00 Av000.0 00.0 F0. FR R0.0 0 F000 ROR0.0 0R00.0
2:0. .0 E0 .0 =5. .00 5:: a 3:10: 0 .00: 0 00: 0 .00: x0: *0: at: 0 0 0 0 0 : 00:: 00:: 00::
0.33: 0.30 «0 5:100: 00 «0 03:00: 00 «0 03:00: 00 «0 :0:00! 00 «0 50:00: 00
0.503 £32.: 0B: 0503 @003 @003 E03 @003 @003 0.50: 0:50: 0:503

 

0: 0 00m 00: 000003200 0000008000 00000 2003 05 0005020 00000 000000 Boxm ”3000: ::-m 030:.

 

166

 

 

 

 

 

 

 

 

 

 

_
hmhovd N0 50 5.0 : : 0000.: 0 55.0 mm. 5 0095.50: 25.0: _Eun.3n leDJO> qwmwm>
vo.w.:.:.m- «0000.5 :0 505.00: 0:095 $50.0 5.0:. $00. 5 E<l09
cownwd. 00:00.0 I<E vimfi ::556 095.0 00.0 2..de 8 cod. 8% N0
m Ln“: ammo... 0505.5 :56 8.: $00.: N0 8.x. Qah (Q0950
hm+<nQ 20:03:00 ~€23: 2: :0: 20:05:: 203: WMQOS: m 69‘: 00.::— RE a .mm<:<

 

:

 

_

_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6:200:06 12:20: 002:: :0: .«0 2: :0 2:002: :0:.:0:: :0:0:00 00: 2: :0 220 :20: :0 302200 002:: :0: :: «0 :0 00:2: :0 0 2: 2 40:: No :O:<>

 

0530000.: 1:0: 000:: :0: . :0:00: 2: :0 .02: 00:: :0:0:00 00: 2: :0 220 :20: :0 302200 000:: :0: :: 90:02 :0 020:: :0 0 2: 2 20:: 5200: 05:3:

 

~¥::.0000.0\~0E::0: 002:: :0: . 00 2: :0 .02: 00:: :0:0:00 00m 2: :0 20:0 :20: :0 302200 002:: :0: :: 2:209: :0 00:2: :0 0 2: 2 .53: 00 :0:00—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _ 0023212: 2: :0 2:29.. :0000: 2: :0 2:20: :22 u :23: 002:: :0:00 2: :0 MR: .5:

_ L 0.22. .3 02.3% 030:: 350:0 2: :0 8:: is u :03: 82:0 0:00: 2: :0 ms: .6:

:03. u: :0:00 o 2: :0 08:00:00: 2:29. 50:. :20: 2:. :0 000300 002: 50.: 2: :0 02.500 2:00.: 2: :0 2:02.010

20:0: 2: A: 33:00: 002: :22: 25:02:00 :000 2: :0 E00 E02 302200 «20:02 00 00: 25:02:00 2: :0 2:22. :22: 2: :0 ::H0

3:02:00: :0 2:002: 3:202. .«00 :0 2:002: :0::0::.:22: 2 :0:00 2 :: .00 :0 2:002: 350.: 2: 0: ::.: .0:|l 0.0

90:02 :0 :0::00:::0E.«m:0:02 :0 :NT. m:0:02 :0 2:002: :0:.:0: 2:22.310: 33:0me

:00 .0 0.82 088.0 00 2 00 :0 3002:3828 :0 3.:- 00 32:32: 5:2: 2: 0: =5: .0: 00

03.000 5.00: 0009.. 0 :0:00: 00 o: :0:00» 0:03:00 .2008: :0 2:002: 30:. 2: 0: :23: 00.200:

0 L: .2. 2.200.: 0... 0.2.8. 3: 0 .2: 0550200 02.92.; _ _ T

~0+t\m.<u~:::0.: 2 0233000 2:20:00 022:1 u::0: :3 3203200 222%.:2 :50: :00 2: :0 0o :0 2:002: :0:00 2: 0: 2.5: ::. 00

T _ — _ 4 _ _ .20 :0 02:: :: 2:002: 30:00:00 2 =2: _:

::022 :0 3032.2: «::.—ES: 30:30:02 2:002: 2: :0 :0:0.:2I:00 2: 2 9:33.000 2:002: 30:02:00 2 $0: : 5:00

_ _ 05:20:22 :20: :00 :28 2: 2 :mmi.

30:30:02 2:002: ::0000 2: :0 ::0302 2: 2 rink!

30:30:02 2:002: :02: 2: :0 50:02 2: 0: 500::

E05
Ema: 00000.0 00000.0 0:: :.0 0:05 0:.:0.0 ::.0: Ex: 05:. : 0005 0000.: 0000. : Ema: Ema: 00.:.0: 0:.00: :0.:.0
:0:00 00000.0 0500.0 :00:.0 000:0 :0:00 0.00: 00.:. ::00. 0 0000.0 0000.0 05:. 0 3.0 0.0:. 00.00: 05.0: 00.00
00: 5.0 00000.0 0 500.0 :00 : .o 000: .0 0000.0 0.00: 00.:. 550.0 55.0 0000.0 0000. o m : .0 0.0: 00.00: 0:00 00.00
0055.5 00000.0 0500.0 :00: .o 000:0 0000.0 0.00: 00.: :0:00 005.0 0000.0 0000.0 ::.0 :.0:. 00.00: 05.00 00.8
0005.0 00 :000 0500.0 0005 80:0 0000.0 0.0:: 00.0 0 50.0 0000.0 0000.0 :0:00 ::.0 0.00 00.00 030 00.::
0005.0 00:00.0 0500.0 0005 50:0 0000.0 0.0:: 0:0 0 :000 0000.0 500.0 005.0 ::.0 0.00 00.02 00:0 :05.
0005.0 00 50.0 0500.0 000:.0 50:.0 0000.0 0.0:.— 0:..m : 50.0 0000.0 0000.0 :0:00 ::.0 0.00 00.00. 0:.00 00.:.0
0005.0 00000.0 00000.0 000:.0 00: :.0 0000.0 0.9:: 0:.0 000:0 500.0 000:.0 000:.0 00.0 :00 00.00 5.00 00.00
0005.0 00000.0 00000. 0 000:.0 00: :.0 0000.0 0.2.: 0:.0 000:0 500.0 000:0 000 :.o 50.0 :60 00.00 00.00 00.00
50 5.0 0080.0 5000. o 000:0 00: :.0 0000.0 0.00: 0:.0 :0: .0 0000.0 000:.0 0:0:0 00.0 :00 05.00 :0:00 00.00
000 :00 00000.0 :0000. 0 30:0 00: :0 000:0 0.00: 00.0 0000.0 :. .000 0000.0 0000.0 00.0 0.00 00.00 00.:: 00.0:
:00 :00 00000.0 8000.0 305 00: :.0 000:0 0.00: 00.0 0:000 050.0 0000.0 0000.0 00.0 0.00 00.00 00.0: 00.0:
000 5.0 0080.0 5000.0 000 : .o 00: : .0 0005.0 0.00: I 90.0 0000.0 0 50.0 005.0 0500.0 00.0 0.00 00.00 50.0: 00.0:
.53: > 00:: > 20:: 0 q s: .: .S .:E\: > 2.2:: .0: ::.: :0: 2.: .0: 2.: .m: ESQ: ::.:«J: Sm: .: WM: 2.00:0: 500::
00 03:00: 00 $2.30: M12340: 502%: :30 0. SEOJM: 0o :1:OM: 00 5:00

:0:§ 50:3. :0:§ 0:025

 

E 0 00m :0: 000320200 00300000000 80:: 2:05 2: 023000 000:0 00800 Boxm ”NTm 030:.

167

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

23823 32382: 2: 2 808388 :22: 08332 3.8 :0 .00 2: 8 823823 8:83 2: 8 Ea» 2: 2 25.x .3 0.30

 

_

_ 8 8 8:28:80 8 :32... 8 82:82: 583 2: 2 :8. .o: 8

 

08332 8 3283: 89:. «08302 8 330nm83o2 8 23823 8:83 2: 2 25.x .03 233003

 

8 8 88288.8 8 :8" 8 823823 283 2: 2 s: < .o: 8

 

030: S 28:50 3o 30:00:: BE 05 2 3 403 No 0303

 

283 s 288: 8 :28: BE 2: 2 m .8: 5.82 232L

 

030: E 2880.5 30 30:00.: BE 2: 2 .3 do: 00 9:0:

 

28: s 253:. 8 .528 £23 2: 2 at. «o 233:

 

22.: s 288. 8 .828 £22. 2: 2 8: 5.82 233:

 

 

 

 

 

 

 

 

 

 

 

28: s 228. to .528 .839: 2: 2 .8: 8 232..

 

8:23.28 8 83.83. 88888 2: ::.: 823 233: 2: 5 8:2 :8 8 88.: 8 82:3: 2: 0::

 

 

8:35 8 3.528 82:3 232: 5 82:8 8 222. 2: 2 m 8 23l.3:

 

08332 8 30.83. 8888.: 2: 2:: 823 233: 2: :3 083:2 8 888 8 82:3: 2: 0%:3338 .3 8:83: 823 233: 2 0882 8 :38: 2: 2 0 0(302 QSIIOZ

 

_ 2888 8 :38: 82388.: 2: 2:: 88:3 233: 2: :2 8:28.. 8 88.: 8 82:3: 2: 0:33:88 :3 32:28 823 233: :3 3:28.. 8 238.: 2: 2 0 30 25F:

 

_ .00 8 888 8 : 88. 2: :8: 823 .832. 2: E 823:8 8 88.: 8 ... 2: 058288 A: 8228 823 233: S 823:8 8 888 8 ... 2: 2 .302 00 Sis—'10:

 

08332 8 88.: 8 : 88. 2: .8: 823 832. 2: :: N838. 8 88.: 8 3 2: 028288 :3 32:28 823 233: :2 08302 8 888 8 ... 2: 2 .5: 5302 0503

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3:28 8 88.: 8 82:3: 88. 2: :8.: 823832. 2: 5 2888.. 8 88.: 8 82:3: 2: 028283: .8 8:828 8225233: 2889. .8 88.: 8 82:3: 2: 2 .202 :0 050:
00.0 00.0 00.0
ENG: .300: Ema: Ema: Ema: Ema: Emma Emm: Emma Ema: Emma .Jmllw"; 00.05 000 El: 0000.0 003.0
0 ...0 0 p .0 03.». 00.0 00.0 003.0 0000.0 0050.0 .33. — $00.00 e5.00.0 Jvl0l. 0 00.0.5 50.0 00000 0 30.0 003.0
9.0 0.0 03.0 00.0 00.0 003.0 0000. 0 0050.0 n$3; a2900.00 95.00.0 30 00.05 50.0 0000.0 0300 0030
3.0 :0 536 00.0 00.0 003.0 0000. 0 0050.0 33.. 65.00.00 02.00.0 30 00.05 50.0 0000.0 030 0 003.0
3.3 3.0 0|5.0 3.0 00.0 003.0 5.00. 0 0050.0 93.00; 9:00. .0 e3.00.0 50.0 0— 65 0M0 02000 0000. 0 503.0
3.». :6 RM —00 00.0 0030 5300 0050.0 2.00; °5.00. .0 e\o000 500 0: .v5 00.0 0300 0000. 0 503.0
:6 — fie 05.0 3.0 00.0 003.0 5300 0050. 0 ..5.00.. o\o00. —0 n2900.0 50.0 0 :35 00.0 — F00. 0 0000.0 503. 0
50.0 50.0 0—0 30 3.0 053.0 0050.0 0050. 0 $9.— 280. —0 $00.0 ..00 00.3.5 00.0 0000. 0 00000 003. 0
50.0 50.0 0'0 3 .0 30 053.0 0050.0 0050.0 2.0 —. . 293. ,0 $00.0 3.0 0005 00.0 0000. 0 0000.0 003.0
50.0 50.0 0 —.0 3 .0 30 #0530 0050.0 0050.0 2.0 F .p $30. 3 Ax00 0 20.0 00. V5 00.0 0000.0 0000.0 0030
00.0 00.0 00.0 3.0 00.0 $000 00500 0050. 0 2900; .x00. —0 95.00. 0 00.0 00. 05 00.0 0000.0 0000.0 0030
00.0 00.0 00.0 3.0 00.0 900.0 5050.0 0050. 0 ..3.00.— 63.50. '0 93.00. 0 00.0 00. 35 00.0 000.0 0000 0 0030
3.0 3.0 00.0 3.0 00.0 0000.0 0050.0 v.0lm0. 0 210.— 2ol50. —0 0000.0 00.0 00.05 00.0 0000.0 0000. 0 0030
E5... .3 ::< .3 35.383 E53 «:3 35¢ ‘33 3 do: 3 302 3 30: $53 *5:- xts 0«.: u .3 0 4 40:: 40:: 30::
0<m|= 0.30 00 m<302 30 00 5302 :0 00 33302 00 00 530M 00 00 013003 00
05003 E0: E3 0503 0303 @5003 05003 900: 0303 Jnsg JOSOS 0503

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

: t :3— 8: 22.2325 2.2322880 82.3 2335 2: 0532.0 82.0 8230 .8320 ”2.500 2-0 22,—.

 

168

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _
$500.0 «0 508.8 82: F060 88 0.8.8F .3505 .5808 01.239 .0085
3.053- «805.0 00 58.02 080.0 88.0 8.55 0.8.8 00000:
00.08.... 8005.0 3000. 82.0 :80 88.0 50.0 ..0.8.0 00 8.0 90 m0
0 L0 580.5 080.0 88.0 00.0 0.9.5 «0 8.5 0.8 000030
50+<u0 J38... 3048 as .3 20:00.00. .3: 0MB: 0 .000: at: $0: a. .000:
F _

 

 

.

 

005.0800 12.53.. 823 .0: .00 2: ..

 

0 2:002: 332: 00:00.00 000 2: 3 033 003: 0.3 000330 0023 .5: 5 No 3 0038 3 ... 2: 00 do: «o 5013.

 

«205.030.: \ :0: 0023 is. . 3mg. 2: 3 .02: .35: 3:250 000 2: 3 0.20 0:30 5.0 02.330 0023 .12. S N308. 3 0035 3 0 2: 0: .53 5:30. tom—S

 

«k~b00~0.0x«2§3> 0023 i2. . 0o 2: 3 .020 02: 3:250 000 2: 3 033 0:3: 3 02.330 0023 .00.. .0 2.200.: 3 003.: 3 0 2: 2 do: 00 mom—S

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _ 002R :30: 2: 3 08:30.. 3002. 2: 3 2:03.. 33: u «30.: 00ch 3&2. 2: 3 3.3.5:
_ _ :0:00 .3 0.2.50 030: =59... 2: 3 002.. is a 03:0 8:3 030: 2: 3 0339
559; an :0:00 05%.: 00330300 0:53.. .33.: 33. 2: 3 00.0300 002: 33E 2: 3 000330 03:05 2: 3 0500,30
.0300: 0: .3 00030: 0020 .33.: 03232000 0000 2: 3 E00 E0: 000330 @3000: 60.0.00 03232000 2: 3 25.3: 33E 2: 3 SB
$3000: 3 0.50003 3:30:00 3 030003 30:00:33.0 0. 30$ 0: : .00 3 030020 3:30 2: 00 32% .00 No
.0332 3 00:00.:3E. @3000: 3 SC I m3m00. 3 030003 3.03: 2: 0.. E5010“. mm<lomL
F F00 Fm F082 088.0 00 _ 00 3 8.0023530 3 0.0.. 00 3950020 3:00 2: 0. :50. .00 ml0
000.000 8.09 F 00000.0 $3000: 00 0. 3:53 0033030 03:00. 3 0.50003 33.. 2: 0: 251* .1: $3.30.

0 0 0 .2. 2.28; 0... 0.3.8. .2 0 .25 052050200 02.05: _ *0
«0:03.34 0.. 0. “033000 0320:00H 003:3 0:3: .3 0033030 0.533200. 33m .00 0.: 3 00 3 030003 3005 2: 0: =50. 1> 00
_ _ _ H W _ _ .83 3 0:0: 0.. 050003 3.023010%
5.5.09.3 3 300030 3:550 300023: 050003 2: 3 00:03:00 2: 0. 0.0.3.3000 030003 00:00:00 0: 301 .0 $60
_ _ 0333202 33m .00 0000 2: 0: Fame.
300023: 050003 000000 2: 3 30002 2: 0: F5095.
30:00.3: 0.50003 30: 2: 3 30002 2: 0: 5.000%”:

Emu:

Emma 00000.0 00000.0 009.0 003.0 0000.0 0.00.. Em: 0000.0 0500.0 000NF 5500.F Ema.» Em": 00.59 00.000 00.000
FNF50.0 00 F000 0 5000 «00:0 000 F0 0500.0 0.50F 00. F0 0000.0 0000.0 0050.0 0500. 0 05. F0 0.90 00.09 05.000 00.000
0F F500 0050.0 0 F0000 009.0 000 F0 0500.0 0.59 F0. F0 F0000 5000. 0 0050.0 0000. 0 F5. F0 F0 F0 5509 00.000 00.000
F~F50.0 0080.0 0500.0 009. 0 000 3 0500.0 F.50F 0F. Fm 0000.0 0000. 0 «000.0 0000.0 00. F0 0.05 00.03 3.000 0 F000
00500.0 00000.0 50000.0 59. 0 009.0 0000.0 0.09 0F.0F F0000 «08.0 F F000 056.0 00.9 0.050 05.00 00.050 0350
00500.0 00000.0 50000.0 53. 0 000 3 0000.0 0.09 0F.0F 0000.0 008.0 0 F000 055.0 00.3 0.050 05.00 00.050 00. F50
00500. 0 00000. 0 50000.0 F00 F0 000 F0 0000.0 0.09 NF.0F 5000.0 55. 0 0000.0 F550 00.? 0.050 00.00 00. 050 00. F50
0900. 0 00000. 0 00000.0 00F F. 0 F00 3 0000.0 5.03 00...: 09 F0 0500. 0 ~00 F0 000 3 F F.0F F000 50.00 00. 000 00.000
0900. 0 00000 0 00000.0 00F F. 0 FO0F. 0 0000.0 5.03 00.0F 09 F0 0500.0 000 F0 meFd F F.mF Fdww «0.00 00. 000 00.000
F9000 00000.0 00000.0 00F F .0 FO0F. 0 F0000 5.03 50.3 09 3 0000.0 000 F0 mFmFd 00.9 0. F00 05.00 05.0 Fm 00.03
Fm F000 0 500.0 5000.0 05F F.0 NFnF. 0 0500.0 0.03 00.9 0000.0 0 80.0 0000.0 0000.0 05.9 «.52 F500 0003 05.00F
5060.0 0 500.0 5000.0 55F F0 05 F0 F5000 0.03 05 3 0000.0 080.0 0000.0 5000. 0 05. 0F 0.00F 00.00 00.03 8de
00000.0 0 F0000 5000.0 05F F.0 0FOF. 0 0000.0 5.03 00. F F [M0000 .mF00.0 I0000.0 0000. 0 00. FF 0.00F I550 0F. FmF 05.09
.3: ._ .3: > .3: > 44...... 0 .3 «EEI > :50 .00 :5. ...0 :50 .8 35.4% :50 ..5> .050 .0 S00 0 .00.: 00050 302%

No 5.000. 00 m2: 40> NH: 40> Emzmo EM no 0308.. 00 21000: 00 E8
L003. No.33 mom—S 032::—

 

0_ 0. :3— 000 00000—0030 00000000000 000:0 03E..— 05 053000 30:0 000000 Boxm “07m 030,—.

 

 

 

 

 

 

_ _ _ _ 0 _ _ _ _ _ _

 

65002: 00.50005 2: 2 000.2. :00 :22: $2202 0:0 00 .00 2: B 020002: 22:00 2: B :50 2: 2 2.; .0 0.30

 

_ _ 00 3 808205.200 3 Sc" 00 32:82: 22:00 2: 2 =5. .0: 00

 

022202 B :0:00: BE. «0:032 B thum02202 B 20002: 22:0: 2:2 20¢. .0: 5x002

 

00 B usageskgo 0.0a 00 .8582: 20.0.5 2 Eq .8 00

 

3:0: :2 0:20.50 B :0:00: BE 05 2 m 203 N0 930:

 

030: S 0:232 B :0:00: BE 05 2 m .63 13.1002 900:

 

2:0: 5 0:200.— B :0:00: BE 05 2 l 20: 00 930:

 

2:2. 5 8225.. B .580: 322.. 2: 2 .23 00 050:

 

use: 5 20002 B .522. £22. 2: 2 2.... 2202 050:

 

 

 

 

 

 

 

 

 

 

 

2:0: 5 2.28.. :o 3830 322. 2: 2 2:. 00 030:

 

 

0:20:00 B 2:022: 220022: 2: 5:: 000:: 200: 2: :2 0:2 50 B 022: B 202:0: 2: 0:: 22:2: 0: 002200 0023 200: :2 0:02:50 B 3020.: 2: 2 0 No 0303

 

002202 B 30.2: 20022: 2: 5:: 000:: 2:0: 2: :2 022202 B 022: B .0050: 2: 0:50:22: 00 002200 000:: 2:0: 2 20:02 B 322: 2: 2 0 25.02 0303

 

0:200 B 320.: 220022: 2: 2:3. 000:: 2:0: 2: :2 0:200; B 022: B .8080: 2: 0:30:22: 00 002200 000:: 2:0: :2 0:200.» B 0:20: 2: 2 0 00 0503

 

 

.No B 022: B ... BB. 2: :20: 00225622 2: :2 0:20.:0 B 022: B 0. 2: 02.002000 0: 002200 000:: 2:0: :2 0:20:20 B 0205 B 0 2: 2 20:. No 0503

 

9.2202 B 022: B ... .22 2: $0: 000:: :0:00 2: :2 0.0002 B 022: B 0 2: 05.02300 00 002200 002B 200: :2 9.2202 B 022.: B 0 2: 2 .5: $2302 050:

 

0:200 B 022: B 000:5: 222 2: 20: 000:: Bums 2: :2 0:200.“ B 022: B 002:0: 2: 02.002000 :0 005200 000:: 200: S 0:200; B 022: B :00—:0: 2: 2 do: 00 050:

 

 

169

00.0 00.0 00.0

 

 

 

00.0N 00.0N 00.0N 00.0 00.0 0N0...0 0000.0 NN00.0 $00.0 $00.00 $00.0 0N6 00.00 00.0 0N0—.0 0000.0 0N00.0
00. N 00. ..N 0 F. N 00.0 00.0 0000.0 0000.0 0—000 2900.0 o\oN0.00 e0.00.0 00.0 00.00 00.0 0000.0 0N00.0 —N00.0

 

0. —N ..0. N 0 0. N 00.0 00.0 000 p .0 0000.0 0 000.0 0000.0 o\oN0.00 o\o00.0 00.0 *0 00 00.0 000 _. .0 0N00.0 0N00.0

 

00. N 00. N 0 F. N 00.0 00.0 000 P .0 0000.0 0 50.0 $00.0 o\oN0.00 A$00.0 02v V0.0 00.0 000 0.0 0N00.0 0N00.0

 

00.00 00.00 0 F .0 p 0N.0 No.0 N 00 0.0 N0v0d 0 50.0 $0 —.0 o\o FN.00 $00.0 «0.0 N0.00 00.0 0 —0 0.0 0N00.0 NN00.0

 

00.09 00.0.. 0 0.0? 0N.0 No.0 N —0 —.0 N0v00 0 000.0 00.0 0.0 o\o FN.00 $00.0 v0.0 N000 00.0 0 F0 «.0 0N00.0 NN00.0

 

 

00.00 00.0— 0 —.0w 0N.0 No.0 N ..0 F .0 N000.0 0 000.0 2.0 p .0 o0 —N.00 $00.0 00.0 N000 00.0 0 00 F .0 0N00.0 NN00.0

 

 

F 0.00 3.00 00.3. N ...0 00.0 N00 0.0 0N00.0 0 000.0 .x. ..0.0 90.00.00 Av0.00.0 000 00.00 00.0 000 w .0 N000 0 NN00.0
P «.9 3.00 00.3 N —.0 00.0 900 0.0 0N00.0 0 000.0 o\o 00.0 G0.00.00 $00.0 00.0 00.00 00.0 N00 —.0 N000.0 NN00.0
00.0— 00.0w 00.: N 0.0 00.0 N00 _. .0 0N00.0 0 F000 o\o 00.0 e0.00.00 $00.0 00.0 00.00 00.0 000 0.0 N000.0 NN00.0

 

00.00 00.00 00.0— 00.0 00.0 PO0N.0 0000.0 0000.0 o0N0.0 e0.00.00 o\o—0.0 00.0 N000 00.0 0000.0 0000.0 0N00.0

 

00.00 00.9 00.00 00.0 00.0 000N .0 0000.0 0000.0 o0..N0.0 90.00. 00 00 00.0 00.0 N000 00.0 000 v.0 0000.0 0N00.0

 

 

 

 

 

00... F 00... — 00. p F 00.0 00.0 000N.0 0v00.0 0000b. °0.00.0 °0.00.00 $00.0 00.0 N000 00.0 0000.0 0000.0 0N00.0
E01 .1 E00. «l 0.00.: cm 20< .01 ER .01 k 203 u 402 k 40: E E 00.03 m 4 u .— u 4 403. 4032 20‘:
01m: 0.36 0‘ 003002 00 NO 211602 00 N0 0.23.2002 00 8 thOQ 00 N0 @3009 00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

030: 10000: 2:03 200: .0000: R00: 030: 030: 030: 2.0000... E00: 050:

I’ll

 

 

M: 0 ::.m .80 28020200 200830250 0225 2:05 2: 053020 000:0 00200 385 ”3:000 070 030,—.

 

170

 

 

_

0

_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0003.0 00 550$ 88F .83... 2.3 2.8.8. 259 .5330 #02085 000%
3.02.0. «SRO 8 080.8 800.0 800.0 2.3.0.00k 0E8.
3W3? 853 $.62 300.2 :08 002.0 5.: $8.: 8 S... 3.. 00
m <_ 982 ~83 .83 8... 0.2.0 00 8.2 0.8 E0200
F0+<n0 c433 0088 2. an. 20:30”. 00: 0&0: 0 .0050 at: E 0 .0032
_
F _

 

 

 

05.8000 \ F2522. 82:. is. .00 as 3 anaei 22:20 $0.50. 2. as 3 22. 052. 3 .8528 8:3 is. s 00 Fe .22: B u 2: 2 .6: 00 :83.

 

05.8000 \ 09. 8:3 i2. . =32 as 3 «ti 220 c2030. =0 .5 3 22.052. ‘3 .8528 8:3 i2. 5 minis. .o .22: 0o 0 2: 2 .6: 0362 :85.

 

~0300009~052§ 802m .02. . 00 2: F0 .02: <20 :00-000 30 as Fe 28. 052. an 005030 323 is. E 2500; 3 .22: 0o 0 05 0F .5: 00 .0903.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

— _ 823 3:0: 05 in 2239.. F809. 2: 00 2222. 32 u @230 0023 3mg 05 00 0:: 09..

_ _ 03:2. .3 083% 2:0: 32%.. 2: 3 82: .22 a 02:0 82R 3:0: 20 B 0:39.

Dh+< '0 =23: 2: 00 08-32000 25:9. 52: 52 as .3 00230 32: EBB 2: am 022800 2:38 as F0 >h0200

#0230 a: .3 000300 3:: =22: 85:09:00 :23 05 F0 E00 E00 005800 «Home». .0000. 2:233:00 05 00 2:29.. 52: 20 0° :00

«03%: 00 2332: 3:20:00 F0 2032: 322028.50 2 F300 3 FF .00 3 9532.0 5:3 05 0F 3: 6m 00

0:332 Fo coaoEFFoEKmiqunF 00 at u minnow. so 2:303 Etna 0523101 00339

F F00 F0 F0002 0800... 8 _ no 3 coaoéaséo .0 90- 8 32:82.“. Sui as 2 E1 .8. 00

03.pr F000: 0008.0 0562 8 2 5:5» 0223:... 0:32 3 eaaei Bis. 2: 2 s: «ck. 05.62

0 m .2. 2.28; 2.. 0.2.8. .2 o .25 0530200 M265: _ _ _ J

40+t\0.<n~<&moa 8 050303 3:20:00 052:! 050: >0 02033.... 8322058 .500 «on 2: «a 00 F0 2038.0 333 2: 2 E: %— 00

_ F _ F _ _ _ ..50 3 8:5 5 2082.0 08023... 2 Es < .1

Bank: 00 3002-0 «59.590 30:02.2. 2:895 05 so coaqiahao 05 8 053080 20393 080230 2 3010.0 5000

_ _ 2222052 .520 F3 :98 05 3 F00?—
0800232 2:32: 0:38 2: io 30an 05 & F.meflufl
08:02.2. 2:320 F2: 05 0o 30qu 05%
00000.0 F0080 00000.0 009.0 003.0 F0000 0.09 00.9 09.0. 0 09.00 0000. F F000.F F 30 0.000 9.09 0.0000 00.000
00000.0 00 F000 00000.0 0F 2.0 009.0 900.0 0.09 00.9 000v. 0 009.0 0000.0 0000.0 .39 0.000 00.09 00.30 00.000
00000.0 F0 F00. 0 00000.0 9 F 3 0090 F F000 0.09 00.9 09.0 009.0 0000.0 900.0 0F .9 0.000 00.09 F0000 00.000
00000.0 00 F00. 0 00000.0 9 F 3 000 F0 0 F000 0.09 00.9 009.0 99.0 0000.0 0000.0 0F .9 0.000 00.09 00.30 00.000
00000.0 00000.0 0 F0000 03 F0 3.9.0 «050 0.00F FF..0F 0000.0 003. 0 0000.0 020. 0 00.9 F. F00 00.3— 0.0000 0009
00000.0 00000.0 0 F0000 00F F0 390 0000.0 0.9.9 3.9 0000.0 003. 0 0000.0 090. 0 00.9 F. F00 F0.F.0_ 9.000 0009
00000.0 00000.0 0 F0000 03 3 F00 3 090.0 0.09 3.9 0000.0 0000. 0 0900 00 F0. 0 00.9 F. 80 00.00 0v.000 00.09
00000. 0 00000.0 00000.0 005. 0 009.0 0000.0 0.09 9. FF 009.0 090. 0 009.0 000F. 0 001 FF 0.09 00.00 F009 0009
0.0000. 0 00000.0 00000.0 00 FF. 0 009.0 0000.0 0.F.0F 0F. FF 009.0 090. 0 009.0 009. 0 0NF F 0.09 00.00 00.09 00.0I0IF.
00000. 0 00000.0 00000.0 00F F. 0 009.0 0000.0 0.09 0 F. F F 009.0 00 F0. 0 009.0 000F.0 00. F F v.00F 00.00 00.09 00.00F
89.0.0 0 F0000 00000.0 0.0000 000F.0 0000.0 0.03 F0.» .0000 0000.0 v9.0.0 0000.0 00.x. 0.9 F 05.0 00. F9 00.09
F0900 0 F0000 00000.0 00000 009.0 0000.0 0.0! F00 .0000 0000.0 003.0 0000.0 v0.0 0.0: :60 00. F9 00.0.9
F0030 9000.0 00000.0 F0000 009.0 0000.0 0.03 F0.» 0000.0 0000.0 33.0 0000. 0 v0.0. 09 F M060 b0. FOF 00.09

40: > do: > do: > 4L8. .F .3 .FlExue > 35.0 on 00.00. .01 0:1 .20 ER .03! Ski at. fibflk 50.0 «1 “Wk—... Quit human

No .0360. 00 mi: 40> m3: 40> >h02m0 3:0 0 5:30. 00 5062 00 S000
533» 0053 F6130 .0023:
0 0 FF :3. 80 mcozfisofiv 0000008000 000:."— 2005 05 053000 00020 00200 3810 ”E -m 0300.

171

 

 

_ _ _ _ _

_

_

 

20000:: 020002: 2: 2 30:00:00 00:0... 90:00: 0:0 00 .00 2: :0 020002: 090: 2: :0 E00 2: 0: ::< .: 0.:(0

 

_ _

00 :0 02:02:35.0“: :0 .00" 00 32082: :0:00: 20 a: 3.3 ...: 00

 

90:00: :0 00:00:: 35:90:00: :0 ::.: 90:00: :0 20000:: :0:00: 2: 0: ::.Q .0: 52:00:

 

00 :0 5002:3500 :0 00:" 00 3.582.: 00:2: .5 .: =5: ...: 00

 

0:00: 5 0000050 :0 00:00:: :2: 2:: 0: n: 400: N0 050::

 

0:00: 0: 0:0:00: :0 00:00:: :2: 2.: 0: n: 400.: $5200: 050::

 

030:: E 00200.: :0 00:00:: :2: 2.: 0: u: :0: 00 0:30:

 

2000 s 80:5. :0 2.2.: 20:00 .5 2 .2... 00 0:00:

 

0:000 0: 2.08: :0 0.8.2: 30...... 20 2 0:: :0:00: 0:00:

 

 

 

 

 

 

 

 

 

 

 

030:: t: 2:200.“ :0 300:2: Egg; 2:: 0: 02:: 00 930::

 

 

000350 :0 30:2: :0:000:20 2: 0:0: 002:: 0:00: 2: 5 0:0: 0:0 :0 00:2: :0 :0050: 2: 05:

:0:00... :0 020.30 ...0: 0300 s 2200:. :. 20:... 20 .0 0 00 0:00:

 

90:00: :0 50:2: 0:000:05 2: 5:: 002:: 0:00: 2: 5 90:00: :0 00:2: :0 :0050: 2: 05:05:05 :0 005200 002:: 0:00: 5 90:00: :0 E0020 2: 0: 0 50:00: 030::

 

0:200 :0 50:2: 0:000:05 2: ::.: 002:: 0:00: 2: 5 00200.: :0 00:2: :0 :0050: 2: 0:005:05 :0 005200 0000: 0:00: 5 2200.: :0 2:002: 2: 2d 00 030::

 

 

.00 :0 00:2: :0 0 :20: 2: 50:: 0000: :0:00 2: 5 0:20.50 :0 00:2: :0 0 2: 0535000 :0 005200 002:: 0:00: 5 0:20.50 :0 00:2: :0 .0 .2: 0: 00:: 00 050::

 

90:00: :0 00:2: :0 0 :20: 2: 50:: 002:: :0:00 2: 5 90:00: :0 00:2: :0 0 2: 05300000 :0 005200 0000: 0:00: 5 90:00: :0 00:2: :0 .0 2: 0: 00:: 02:00: 030::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0:200 :0 00:2: :0 :0050: :20: 2: 50:: 002:: :0:00 2: 5 0:209: :0 00:2: :0 :0050: 2: 05:00:00» N0 005200 002:: 0:00: 5 00200.: :0 00:2: :0 .3050: 2: 0: 00:: 00 @003
00.0 00.0 00.0
2.00 2.00 00.0: 00. 0 00. 0 000:. 0 000:.. 0 00000 2.00.0 2.00.0: 2.00.: 00.0 3.00 0:..0: 009.0 0000.0 0.0—.0
2.0: 2.0: 00.0: :.0. 0 0:. 0 000: .0 0000. 0 000.0 2.00.0 2.2.2 2.002.: 00.0 00.:.0 00.0: 000:0 300.0 00000
.200 2.0: 00.0: 00.0 0.0 000:. 0 0000. 0 00000 2.00.0 2.2.? 2.00.: 00.0 00.00 00.0: 003.0 0000.0 .0000
2.0: 5.0: 00.0: 00.0 0.0 0000.0 0000.0 000.0 2.00.0 2.: 00:. 2.00.2 00.0 00.00 00.0: 003.0 0000.0 .0000
00.0: 00.0: 00.0: 00. 0 00.0 :00 P .0 0000.0 0:000 2.00.0 2.00.2 2.00.5 00.0 00.00 00.0: 009.0 0000.0 000 :0
00.00 00.00 00.0: 00. 0 00.0 0000.0 0000.0 0:.0 :.0 2.00.0 2.00.00 2.00.: 00.0 00.:.0 00.0— 0000.0 0000.0 00.0—.0
00.0: 00.0: 00.0: 00.0 00.0 00000 0000.0 0:0 :0 2.00.0 2.00.00 2.00.: 00.0 00.00 00.0: 0.80.0 0000.0 00000
00.: 00.: 0 0: 2.0 00.0 003.0 0000.0 000.0 2.00.0 2.00.2 2.00.: 3.0 00.00 00.02 00: :.0 0000.0 000:.0
00. S 00. F p 0:. 2 3.0 00.0 0:0 : .0 0000.0 000 :0 2.00.0 2.00.0» 2.00:.— 0:.0 00.00 00.0— :0: 0.0 0000.0 000 F0
00. 2 00.: p 0 0.: 2.0 00. 0 0:0 — .0 0000.0 000 :0 2.00.0 2.00.00 2.00.2 3.0 00.:.0 00.0: :0: :0 0000.0 000.0
00.:. 00.:. «0.0 00.0 00. 0 000: .0 0000 0 000:.0 2.00.0 2.0000 2.00.2 00 0 0:..:.0 00.0: 00000 0000.0 000.0
00.0 00.:. 00.:. 00.0 00.0 000:. 0 0000. o 000 «.0 2.00.0 2.00.00 2.00:.— 00. 0 00.00 00.3 000 —.0 0000.0 000 —.0
00K 00.:. 3:. 00. 0 00.0 000 p .0 0000. 0 000 :0 2.00.0 2.0:..0:. 2.00:.— 00.0 0:..:.0 00.0: 000 0.0 0000.0] 000 «.0
=22 ...: E< _: 05.010: ::.—2 .0:l|.1 =C< 0: ..: 00:: h: 00:: ..: 00:: E Rt: E 0 .: Um .: 0 .: 0000: .002: 003::
00:90 0.30 00 52:00: 00 00 52:00: 00 00 £3.50: 00 00 21:00: 00 00 50:00: 00
@000 @502 H.003 P00: @002 #00:: P303 E03 @003 P500: P30: @000:

:0 0 =3: :0: 000020200 0000008000 002:: 2003 2: 053000 00000 00200 305 ”30000 07m 050:.

 

172

 

 

 

 

 

 

 

 

 

 

 

 

—
ban—Mote N0 hvmnd: : 080.: :thd oo.:o $8.8: #90:. .Euoéou W238 > ..mmmw>
VOMfid. momohd 8 new“; 085d «:5. c coda o\oNO.:d m ”$309
vomlmmd. ooXKd ::.:(lom «moodw 3.9.0 :52. 3.3 33.5: 8 86 oo.: N0
m LN 38.: vmvod mmood oo.: o\od:.: NO 8.0: 0.8 (10950
hm+<uQ 20:95: :33: 0::: :0“: 20:03:": :03 mm:0::<, a .mm<3: «PCS E m .mm§

 

:

_

 

_

:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2:589: : 3:53.. 8.5: is. .«0 2:: :o 2:88: 5:3: 8:38 20 2:: :o a...» new... .3 0.538 ...-c: d: s «0 :o 8:2: :0 ... 2:: a: 40:: «0 :03:

 

Ottodwao \ :9: 000:: 30:. .. :300: 0:: :0 .003 :3: :0:030 03 0:: :0 0:80 N::00 ::.: 005330 000:: 30:. :: $300: :0 00:2: :0 ‘ 0:: 0: :9: .3300: comes

 

~¥:.:..0owao\~0::0:0> 000:3 4:3. . no 0:: :0 .003 :3: :0:00—:0 03 0:: :0 0:80 9:0: :3 005200 0003 30:. :: 30:00.: :0 00:2: :0 a 0:: 0: :9: 00 moms.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _ 82:: 2:9: 2:: :o 2:29.. :88: 2:: :o 2:29. :22 1:23:82: :85 2:: :o 0230:.

_ _ 50:00 A: 3:30: 20:: :0::m.:0 0:: :0 000:: :0:0: u «:83: 0003 £03.: 0:: :0 “2:30:—

Em}: A: :0:00 0:: A: 3:030:00: 050:0: :0:0E :80: 0:: a: 3:30: 000:: :0:0E 0:: :3 305030 0:03.: 0:: :0 Emzme

53:0: 0:: A: 30:30:: 000:: :0:0E 8:33.50 :30 0:: :0 :50 80:: 005030 0:300: 60.8: 0::0:3E00 0:: :0 050:9. :0:0E 0:: :0 Sam—w:

3:300: :0 20002303318 :0 050003 :0:t3:.:0:0:n: 0: 30:0 0: :: No :0 050003 :0:::3 0:: 0: ::.: :01 NO

m:0%00:| :0 :0::00:::0E.«m:300: :0 3:: .- m:300: :0 050003 :0:::3 0:: 0: 2:10“: 3:369

:0mm 3.89 088... 8 _ 00 :o coco-::oEHFB ::.:u 00 32:32: Era: 2:: 0: 3%

ovodmm 5.8:: 88:8 555%: no 0: :0::E:0 00:03:00 m:300: :0 200003 :38. 0:: 0: 323%: $3000:

0 0 Id“ .8. 2.23.: ::.. mica. 3: o .25: 0:250:00 mz_o.:z<_ _

«03:33.; 0.: 0: m:::::0000 2:80:00 0::0::< 9:03 A: :0:-:00:00§E0: :50: :00 0:: :0 no :0 050003 330:. 0:: 0: 3:1 1:: 8

_ _ H _ _ _ : .E:0 :0 0::: :: 20000.3 00:02:00§

2.19:3: :0 30:25: «55:3: 30330:: 050003 0:: :0 :0::0::I::~0 0:: 0: w:::::0000 200003 00:00:00 0: Sm: .n: 550

H 0:::E3E0: 3:03 :00 :03 0:: 0: :mmP:

300005: 050003 ::0000 0:: :0 30:3: 0:: 0: Eng

30330:: 050003 :0:: 0:: :0 30:00: 0:: 0: mam—n:

:8 5.0 508d mooood mm: :d «on :d we :md odd: and toad «womd 83.: 3mm: and odd VNNM: «odd: oodo
:mo:o.o ommoo. o :88. o mm: :.o :83 o:. :oo 0.8: v0.0 823 386 moon: m :3: ood Q8 5.5: 3.3 moda—
moflod no :8. o 9.08. o 2.: :d N :o: .o vovo. o o. :o: oo.: ow :md RN : .o :08. o ammo. 0 $6 «do Veda: undo Edm—
oode 3 :oo. o 0:68. o o: :d ::o:.o move. 0 o. :o: oo.: no :md .RN : .0 mode 0 wood o co... «.8 3.8: 3.3 :odfl
«856 2:86 9586 on: :.o :96 oovo. o m. 5: mod onomo «8:6 0:8. 0 Com. o no.0 0.5 3.69 no.8 omdm
ooo:o.o o: :86 emooo. o o :m :.o «8:6 :68. o :.o:.: 3.0 :oowo osmo. o 9:3 0 on :3 mod mom 8.3: 8.3 code
coo:o.o o: :oo.o vwooo. o o 53 ~33 mooo. o :.o:.: and 0086 2.8. 0 nova. o um :06 no.0 0.00 5.3 oo.? 8.9.
com :3 0: :23 «~80. o m :N : .0 Km : .o noood :d:.: and woowd :30 0 did 00 :md mod odm No.3 mod: nwdv
03 5.0 880.0 o 58.0 omw :d vow : .o wooed ads: ::..N own :d owmo. o 89. 0 3w :d :o.~ Qwv 8.3 mon mwdw
cum :od omoood o 58.0 com :d me:d ooood odn: ::..m omw :d mwmo. o «8:. o mum : .o :o.~ adv :bdm mmdm modm
:o~:o.o omoooo o:ooo.o own 3 vowto ooooo o.o:.: Ed owm :.o ammo. o «8:6 mm“: o mod tmv moon omdo oudm
o: : :o.o :.~ooo.o voodoo mom :6 mowfio 886 :08: :~.~ momoo «mood ooood 03o. o RN v.8 omsm omo: v0.0:
ow: :od Roood 308.0 oowfio :om:d «mood vdo: Ed owned ooood mowed moved HEN v.8 :.~.:.o I and: cod:

:0: .. 40:: > .6: > .: 8. .: .3 3:“ > :3 .9: :5. .o: :5. a: :5: ...:5 :5: ms :3... EM: .1 RM: 10:: $003

No 5300: 00 $5.39: min 40> 3%: Shaw «0 530m: 00 135%.: 00 thou

53$. tomes momS: «$23:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

an t :3: 8: 22:33:00 3:05.880“: 82: 20:03 0:: wagosm :oonm 329m Boxm 67m 033.

173

 

 

_

_

_ q

_

_

 

20002: 0200005 2: 0: 00900000 00:00. 00:00: 0:0 00 .8 2: :0 020002: :0:t0: 2: :0 500 0.: 0: ::.: .: 0.30

 

: 00 :0 0282:3800 :0 0:0... 00 3030020 :0:00: 05 0: ::.: .00 00

 

00:00: :0 00:00:: :05. 30:00: :0 :>:um:0:00: :0 20000:: :0:::2~2: 0: ::.; .0: 03:00:

 

00 :0 8080:0500 :0 ::.fia 320002: 0:02:05 0: 3E: .00 00

 

:0:00: 5 2:00:50 :0 :0:00: :05 0::: 0: n: 40:: N0 050::

 

030:: 0: 0:2:00: :0 00:09:: :2: 2:: 0: n: :030:S:00: Q00:

 

0::—0:: 2: 0:209: :0 20:00:: :05 0::: 0: n: 40:: 00 030::

 

030:: 5 05:50 :0 38:00 30:9: 2:: 0: 0.0: 00 030:

 

030:: 0: 0:008: :0 .0022: 30:9: 2:: 0: 0:3 5.000: 050:

 

 

 

 

 

 

 

 

 

 

030:: 0: 0:88.: :o 38:00 201:9: 2: 0: 0.3. 00 0:00:

 

000::50 :0 30:03 :0:000:20 2: 5:0. 002:: 0:00: 2: 5 0:0: ::0 :0 00:05 :0 0052: 2: 05:

 

 

:0:00... :0 08:53 00000 030:: 0: 32:50 :0 30:0: 2:: 0: 0 00 0:00:

 

00:00: :0 50:03 :0:000:20 2: 5:3 002:: 0:00: 2: 5 00:00: :0 00:05 :0 .0052: 2: 02:05:05 :0 005200 002:: 0:00: 5 00:00: :0 530:: 2: 0: 0 :0:00: 050::

 

02200 :0 50:00. :0:000:20 2: 5:2 002:: 50%: 2: 5 00200-: :0 00:05 :0 0052: 2: 02:05:35 :0 005200 002:: 0:00: 5 00200.: :0 30:00. 2: 0: 0 00 030::

 

 

.00 :0 00:05 :0 ... :20: 2: 50:: 002:: :0:00 2: 5 000::50 :0 00:05 :0 0 2: 05:00:03 :0 005030 002:: 030:: 5 22:50 :0 00:05 :0 0 2: 0: :03: «0 0:09.:

 

00:00: :0 00:05 :0 0 :20: 2: 50:: 002:: :0:00 2: 5 00:00: :0 00:05 :0 0 2: 05:02:30 :0 005030 002:: :0:00: 5 00:00: :0 00:05 :0 0 2:: 0: 40:: 03:00: 030::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000:00 :0 00:02: :0 :0053: 3:0: 0::: 50:: 000:: :0Q0) 0::: F:: 00200.: :0 00:02.: :0 :00—00:: 0::: m:::00.::030 NO 000:0:00 000th: U:::U:: 0: 00300.: :0 00:02.: :0 :0083: 0::: 0: :0: EU Q:=°:n:
0n.o 0n.0 00.0 F0. F 0N0 0.00.0 000n.0 000 F0 $00.0 $00. F0 $00.nF 00.0 00.no 00.3 Fn 5.0 0000.0 NNO F.0
00.0 00.0 00.0 F0. F 0N0 300.0 moond 000F.0 $00.0 $00. F0 $00.nF 00.0 00.n0 00.3 Fn F00 0000.0 NNO F .0
00.0 00.0 00.0 N00 0 F0 NON0.0 0N0n.0 000 F .0 $00.0 $00. F0 $n0.n F 00.0 N0.n0 00.0 F 00N0.0 0 F000 nN0 F .0
00.:: 00.0 00.0 N00 0 F0 NON0.0 0N0n.0 000 F .0 $00.0 $00. F0 $n0.nF 00.0 N0.n0 00.3 00N0.0 0 F000 nNOF.0
No.0 No.0 n0.0 N00 0F.0 00N0.0 0N0n.0 000 F .0 $00.0 $00. F0 $n0.nF 00.0 00.n0 00.0F 00N0.0 :0 F000 nN0 F.0
no.0 no.0 00.0 nN.0 00.0 0000.0 N End 000 F .0 $0n.0 $00. F0 $n0.nF F0.0 N0.n0 F0.: 0 FN0.0 0N00.0 0N0 F.0
no.0 no.0 v0. 0 nN.0 00.0 0000.0 N F0n.0 000 F .0 $0n.0 $00. F0 $n0.nF F0.0 N0.n0 F0.: n FN0.0 NN00.0 0N0 F.0
00.0 00.0 N0. 0 nN.0 00.0 0000.0 N F0n.0 000 F .0 $0n.0 $00. F0 $n0.n F 6.0 00.no F0.F0 F 0 FN0.0 NN00.0 0N0 F.0
F0.N Ed on. N 0 F0 No.0 0N00.0 00nn.0 nn0 F .0 $nn.0 $00. F0 $00.n F 00.0 00.n0 00.:. F 0NN0.0 0N00.0 000 F0
F0.N F0.N on.N N F.0 No.0 0N00.0 00nn.0 nn0F.0 $nn.0 $00. F0 $00.n F 00.0 00.n0 00.3 0NN0.0 0N00.0 000F.0
00.N 00.N vnN N F.0 No.0 0N00.0 00nn.0 nn0F .0 $nn.0 $00. F0 $00.nF 00.0 00.n0 00. 0 F 00N0.0 0N00.0 000F.0
nN.N nN.N FN.N 00.0 5.0 0000.0 00nn.0 0n0F.0 $N0.0 $N0. F0 $00.: 00.0 00.00 00. 0F 00N0.0 F000. 0 FO0F.0
nN.N [nN.N l FN.N 00.0 5.0 0000.0 00nn.0 0n0 F .0 $N0.0 $00. F0 $00.n F 00.0 00.00 v0. 0 F 30th 0 F000. 0 F00 F.0

En< a: E< .: En< 6: ::.n< 0: E< .00: n: 403 n: 40:: n: :03 $.03 $.05 $.03 0 d u .: 01.: 40‘: : 40:0: : :03 :
00:03: 0.:(0 N0 03:00: 00 N0 03:00: 00 N0 ::(QOQ 00 «0 15:00: 00 N0 03:00: 00
030:: 0.50:: @003 0B: 0|:00:.: £00: 0'30: @503 0h00: @003 0B0: 0:003 ,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm :: =3: :0: 28003200 5:238:00 002:: 2:05 on: 052650 32:0 00200 .081: ”3:00: 2-0 030:.

 

 

174

 

 

_

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.0.3... 00 «03.8: 88.: 83.0 3.00 0.8.8: .50: .0003: M239 .0009
30:3. 883 00 28.8. :05... 900.0 8.0: 0.5.2 M00000:
8.03.... 8:03 00:00: :80... 82.0 ~83 5.2 0.8.: 00 8.0 8.0 00
0 < 3.3. 82.0 82.0 8.0 0.00.: 00 8.2 00: 3:00:00
Elfin 8:838 004% 2:: 0.: 2003:“: .62 00:03:, 0 00.5. 0::: 00s 0 .0002
:

 

 

 

. 55.8.3:

 

: \ «20:00:. 002:: :0: .«o 2:: ::

 

: 05000:: :0:::0M:: 00:00:00 00: 0::: :0 0:20 :0:00 :0 005030 002:: .:0:. 0: No :0 00:00: :0 0 2:: 0: :02 no :O:<>

 

Ottawa: 1:0: 002:: 05:. . :0:00: 0:: :0 00:: 00:: 00:00:00 00::q00: :0 0:20 :0:00 :0 000:0:00 002:: .:0:. 0: m:0:00: :0 00:00: :0 0 0::: 0: :00: :<:Om: 20:3:

 

«05.000025520; 002:: .:0:. . 00 2: :0 00:: 00:: 00:00:00 00: 2: :0 0:0:0 :0:00 :0 000:0:00 002::: :0; 0: 00200.: :0 00:00: :0 0 2: 0: :00: 00 :O:§

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ _ 002:: 0.50:: 2: :0 0055:? 5000:. 2: :0 005:9. :20: n «:23: 002:: :0:00 2: :0 $5.30:—

_ _ 0:82. 5 0.032. 030:: 050.00 2:: :o 00:: :30: u 0.5:: 82:0 0.50: 2:: :o 03:30:.

99;: In 00:00 2: :5 0205200: 005:9. 5:20 :20: 2: :0 0000.00 0020 :20:0 2: :0 000.330 0500.0 2: :0 ::.—:0:00

$0000 0:: :0 0000.00 0020 5:20 200002000 0000 2:: :0 050 80:: 005200 35:00: 60.00: 2:000:88 2:: :0 200:9. :0:00: 0::: :0 0:00

W5a0: :0 05000:: 5:05:00 :0 05000:: :0:00::.:0:0:: 0: :0:00 0.: :: .00 :0 05000:: :0:00: 0:: 0.: ::.: .0: «0
W550: :0 00:02:20. @550: :0 ::.: n 90:00: :0 05000:: :0:00: 0::: 0: 3:16: 33.3%

:0... a .080: 088... 00 : 00 :0 00:80:05.5 :0 ::.:u 00 303002: :0:00: 2:: 0: ::.: < .00 00

03.3: 8.00:: 038.0 05:00.: 00 o: 5:80 :8.-30:00 00.08:: 2:002: :00: 2:: 0: 2:12. 000000:

0 0 < "2. 2.200.: a... 0.2.8. :o: .0 .50: 05.20200 02.0023 _ _ J _

«0+txmiN 0.: 0: 0:0:0000 0:00:0000 00:0:0< :0:00 :3 0205200 05:20:50: :50: :00 2: :0 00 :0 05000:: 5:2. 2: 0: ::.: .:> 00

W _ _ — _ _ _ _ ..0:0 :0 0:00 0: 05000:: 00:00:50 0.: 3:51 ::

::.:003 :0 50005:: «03009:: 5000000: 05000:: 2: :0 00:05:00 0:: 0: 0505000 05000:: 00:02:00 0: $0: .: ::00

I _ _ 05:05:80: :03: :00 0000 0::: 0.: ENE.

5000000: 05000:: 000000 0.: :0 50000: 0:: 0.: ::005:
5000000: 95000:: :0:.:: 0:: :0 50000: 2: 0: hwwflm

Smuod 0008.0 8806 on: :0 at. to 8:06 0.02 00.0 3.5.0 836 38.: 0000.— ovs :00: 00.8: B. : 2 5.0::
vowwoo monooo 3086 o: :.o 2.36 oopoo mot go m :36 vao ommwé noon: 00.:. hoop vusm: 3. p S 5.0— —
880.0 R 80.0 380.0 «mm; :mio 080.0 0.00: 00.... 800.0 mm: ..o 83.0 0000.0 8.0 0.00 8.8: 8.02 0%
moowoo stood vvoooo «mm to 30 to 5006 v.5— vod ommvo on: :.o good 386 omd moo o:..oo: 8.8— ow. 5—
ooomoo mtooo evoooo mmmfio 536 owned n.5— mnd @036 on: :6 «wood oomoo and mom ovoo: 05.3. 3.8—
oom So «986 mwoooo mom fio own :6 owmoo mdo— mm... 836 0 Foo 300.6 «and mod 0. E 00.3 nods oodo-
oom So «986 «Noooo nomfio 896 owned mdo: no.0 336 n Boo 3006 «end mod 0. ::. 00.02 nods omdo
mom —o.o Nowooo mmoooo momfio com :6 owmoo Duo: 5.0 $36 0 :mod 356 $56 mod 0. K 3.3— nodh oodo
moo—co mmoooo 3086 «8.6 So :.o 336 Qum— mod owed oomoo : :96 5m :6 mod Yon No.8. 006m 3.:
moo —o.o «moooo o woooo «on p .o 3.0 F .o o Koo 0.0m: and moo : .o oowoo woo : .o mum : .o 8.0 0.3 awn 00.8 are?
3856 $806 o—oooo mom—o :00 :.o ::.oo 0.59 mod woo—.o oomoo mooto ommfio mod 0.8 dem 3.0:... 3.00
ooopoo toooo Nooooo omm—o vamfio mmooo tum: mad 886 986 0366 mnwoo mod 0.0.: «now no.9. No.8
«an :o.o 0586 «886 omnfio own :6 £006 o. E: 3N. oomoo moooo mmvoo ImWNoo mad adv 0:..mw oodv No.8

:03: :. :03: :. :02 :. .: 9:. .: .5 .00 > E< .0: SE: .0: ::.: .0: 0::: _.:l:. 231:: ::.: :: 30: .m 5% ::.:wks: bumps:

00 5:00: 00 ME: 40:. W01: 40:. ::.»:umzrm: :30 «0 5:00: 00 :0:00: 00 PER:

Imo:§ Lco:§ :O:§ 0002::—

 

 

mm 0 =3: :0: 20:33:00 02.8.0850 0:00: 20:05 0::. 33.2% 30.5 002% Boxm ”o: ..m 030:.

175

 

 

6.30002: 00.0000... 2: 0: 000.0260 00:00. $0000: 000 00 $0 2: :0 00:032.: 3:30: 2: :0 .000 2: 0: 32%.... 0.3.0

 

_

_

00 .0 20.80.000.200 .0 .....n 00 .0288... 0.085.. 0. =5. .00 «0

 

$0000: :0 00:00:: :05. 3.2000: :0 Q>:um.0m00: :0 03000:: :0:00: 2: 0: ::.—Q .00: ...—Row:

 

00 .0 00.00.008.00 .0 0...» 00 .00.:0020 0.000 2.. 0. 2:0 .00 00

 

0:03: 0: 000250 :0 :0:00: :02: 2:: 0: u: 403 N0 930:

 

0.00.. 0. 0.2.8. .0 00.80.. .0... 2.. 0. ... 00.0 03.00. 0.00...

 

:0:00: ...: 0:200.» :0 00:00:: :08 0::: 0: u: 40:: no 9303

 

0:03: ...: 2.0350 :0 300:2: 2:90.: 0:: 0: 0&0: N0 930:

 

0.00.. ... 0.008. .0 .0022: .000... 2.. 0. 0:... 05.00. 0.00...

 

 

 

 

 

 

 

 

 

 

 

0:09: ...: 2.800.: :0 ::00:0@ 3%.: 0:: m: 02.: no 930:.—

 

0:00:50 :0 .0900. :0:000:20 2: 0::... 002::: 0.00:: 2: 52.0: 50 :0 00:05 :0 .0080: 2: m5.

 

.003... :0 0050.00 000.... 0.0.0.. ... 05.3.0 .0 2.0.0... 2.. 0. 0. 00 0.00...

 

9.0000: :0 2.0.0.: :0:000:20 2:: ::.... 002::: 0.00:: 2: 5 9.0000: :0 00:05 :0 .0052. 2: 332.3... ...0 005200 002::: 0.00:: 5 92:00: :0 .0500. 0.: 0: a 50.002 0.00::

 

2.0.00 :0 .090... :0:000:20 0.: 5:... 002::: 0.00:: 0.: 5 000.00.: :0 00:0... :0 :00500 2:: 0520::05 N0 005200 002::: 0.00:: 5 2.200-: :0 .020... 2: 0: m 00 0:00...

 

 

No :0 00:0... :0 ... :20: 2:: .00.: 002::: .000: 0.: 5 2.03.50 :0 00:05 :0 ... 2:: 05:00:03 ...0 005200 002::: 0.0%.: 5 000:2:0 :0 00:0... :0 .0 2: 0: 005 No 900...

 

92:00: :0 00:05 :0 ... :20: 2: 50.: 002::: :0:00 0.: 5 9.0000: :0 00:0... :0 0 2: $4202.03 ...0 005200 002::: 0.00:: 5 mimom: :0 00:0... :0 ... 2: 0: .50: 25:02 050:0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

QQOKVO ‘0 @022: \O MOO-=3: i=8 Otu EOL‘ OQQCEOQQ> Gt. 2: OCOuoonh ‘0 «6:0—5: ‘0 :QQEDE 0.2 Qttouhnam :AO UOC‘Q~OO OOQCQ U~=U= C: 020:9?“ \O QO~OE ‘0 :QOEDC Q9: 9‘ 40: Q0 Q~D°~d
8.: 8.: .00 8.0 8.0 .020 800.0 000.0 0.00 0.82 0.00:. 8.0 8.0: .02 8. ..0 0.:00 0:0..0
8.: 8.: .00 000 8.0 .020 .800 000 ..0 0.00 0.8.2 0.00:. 8.0 8.0: .02 8. ..0 0.:00 0:0..0
8.0 8.0 00.0 :...0 0.0 .030 .0000 000 ..0 0.000 0. .02 0.00:. 8.0 8.2 8.2 :0. ..0 8:00 ::0..0
8.0 8.0 00.0 8.0 0.0 .88 0000.0 000.0 0.00.0 0.00.2 0.00:. 8.0 8.2 8.0. :2.0 020.0 ::0..0
00.0 00.0 .00 8.0 ...0 .020 .0000 000.0 0.00.0 0.62 0.00:. 8.0 8.2 8.2 :2 ..0 8:00 ::0..0
00... 00... :00 8.0 00.0 ......0 080.0 000.0 0.:00 0.00.0: 0.00:. 8.0 .00: 8.2 02.0 00:00 000.0
00... 00... :0... 8.0 8.0 ......0 800.0 000.0 0.00 0.00.0: 0.00:. 8.0 .00: 8.0. 02.0 020.0 000.0
00... 00... :0... 8.0 8.0 ......0 080.0 000.0 0.00 0.00.0: 0.00:. 8.0 .00: 8.0. 02.0 00:00 80.0
mod wad end :6 No.0 vmv P .o N 36.0 vow —.o o\o Cum okmadh 2°va— de finds. motm— mm: .6 unhmd 5m v.0
8.0 8. 00.0 ...0 00.0 80.0 0.000 000.0 0..:.0 0.00.0: 0.00:. 00.0 :00: 8.2 «2.0 :0:00 .020
00.0 8.0 00.0 ...0 000 80.0 0.000 000.0 0..:.0 0.00.0: 0.00:. 00.0 :00: 8.2 «2.0 :0:00 .020
8.0 8.0 8.0 00.0 00.0 08.0 .0000 000.0 020 08.0: 0.00:. 00.0 .00: 00.0. 08.0 ...:00 .020
NQN mod and .I no.0 I ood 03. —.o omwwd mmo _. .o o\omn.o 20805 2&0: and wwdh om.m— oomho Shmd Nam Pd

.20. ... 0:0 ... .20 00.. 0:0 0.. 0:0. .00 h. 00.0 h. 00.. u. ..0: 0::... 0::... 005 0 .. 0 0 0 .. 00.... ..0... 00....
2.0:: 0000 «0 03.00. 00 «0 03.00. 00 00 03.00. 00 «0 03.00. 00 00 00000. 00
200... 0:00... P00... P00... 200... $.50... $00... L060... E00... L200... 200... 900...

mm ... =3: 00: 0:28—00:00 00:00:000000 002::: 03:05 2:. 9.355 002% 002% :00xm “3:00: 07m 030:.

 

 

 

 

176

 

 

 

 

 

 

 

 

_ _
0:02.000 8 030.8 88.. 2.8.: 8.00 2.8.8. .300 .6363— 0239 .002;
3.02.0- «82.0 8 82.8 3:6 89.0 6.8 2.2.8 25.00:
8.03.0- 832.0 52% 82.8 82.0 3.80 8.9.... 2.8.8 8 8.0 8.: 8
0 Lm 83. : 080.0 8:3 00. : 2.8.2 8 8.8 ...0: 50028
23.8 5:28 002% as 3“: 295.2%: 00.20: 0 .002: at: at: a .09.:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0: 5.0000 .0 1022202. 002::: do; .00 2: :0 2:820 322:: 20:20:00 80 2: :0 220 02:»: N: 022200 .

 

.220 .02. s 8 :o 3:2: :o 0 2:. 2 40: «0 20.3.

 

00.0.0000 .0 \ 00> 82:: .02. . 22:02 2: :0 .00:: 22:: 20:22.0 80 2: :0 220 02:00 :0 002:200 82:: .22. 2: $0000: :0 8:02: :0 ... 2: 2 40:: 0:350: @013.

 

«¥:h.0000.0\~0\25:0> 82:: is. . 00 2: :0 .02.: 220: 20:38 80 2: :0 220 022. Am 005200 002:: .02. 2: 02200.: :0 020.2 :0 a 2: a: 40:: 00 comes

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

82:02:: 2: :0 2220). :88: 2: :0 2220:. :20: I 02:4: 82:9 :23. 2: :0 N25402:
002% .3 082% 033: 0502.0 .5 :o 8:: 28 u 02:: 82:2 2.2.: 22 :o 2.5 00>
QD+< an 0020: 05 Na 2:030:00: 025:0:— .:0:0:: :30: 0::: NO 000300 002:: :0:05 05 000 005800 2005:: 0::: :0 EWZWQ
03200 2: .3 0002.00 322 2222 220200.200 2.80 2: :0 250 .20: 022200 «02:02 .0000E20200500 2: :0 2229. 2222 2: :0 :50
@8202 :0 2082.: 3220.80 :0 20823002022221 0: :0:00 2 2 .00 :0 2082.: 220% 2: 2 ::.Q 0%
I 92:02 :0 20:02:02.. @8000: :0 «::.: n M8002 :0 2:32: :222: 2: 2 32.1.0: $3100:
:83 5.002 088... 8 _ 8 2. 50823528 2. $0.. 8 32:82.: .228 as 2 a: 6.00%
000.000 :000 2 000010 50:00: 00 0: 8:200 022028 M8002 :0 20820 2002. 2: 22:1.Emcflom:
0 0 <40 "2. 2.28.: a... 0.2.8. 3: .o ...E: mpzfimzoo M2651 _ r

«0+C\0.<N&004 0: 0420:0000 2220200 02:0:2< 02:00 :0 022028 02220222 ERE :00 2: :0 00 :0 0:088: 2002. 2: 0.: 2.2x .10 00
I _ _ _ fl _ 4 _ .220 :0 2:20 2: 20320 0202:00§
tank: :0 30022: «232250 20000020.: 0.2.08.2: 2: :0 20:20:00 2: 0: 0502080 02000280200200 2 30.: .0 5:00
.. I H _ 2220222 :20: :8 22.0 2: a: 50):.
2800022: 2032.: 020000 2: :0 30002 2: 2 Emma
2800022: 200020 :2: 2: :0 30002 2: 2 5.00%..
:00020 00:00.0 00000.0 003.0 003.0 0000.0 0.03 00.0.. 0000.0 3.00.0 0:00.: 00.0.: 3.3. 0.000 00.00: 00.000. 00.000
300 :0 00000.0 00000.0 000:0 003.0 0: :00 0.00: 03.0 0050.0 0:0 :0 0000.0 0:000 00.00 0.000 00.00: 0 :00 00.000
0000 :0 00000.0 00000.0 00000 003.0 00.0.0 0.00: 0.2.0 3.000 000 :0 0000.0 0000.0 00.:.0 0.000 0100: 00600 3.000
0000 :0 00000.0 00000.0 000 00 00v :0 00:00 0.00.. 00.:.0 0000.0 :00 :0 0000.0 0000.0 00.:.0 0.000 5.00: 00.000 00000
000: :0 0000.0 00000.0 000 :0 00.. 20 0000.0 0.00: 0 :00 :3 20 0 :000 00:00 00 0.0 00.00 0.0:. 0062 3.00:. 00.00:.
000: 00 00000.0 00000.0 0000 003.0 0000.0 0.00: 0:00 0300 0 :000 00.0.0 :0:00 00.00 0.0% v0.v0_ 3.00:. 0.00:.
000: :0 00000.0 00000.0 000:0 002 :0 :0000 0.00: 0:00 00:. :0 0 00.0 030.0 0:. :00 00.00 0.000 00.00. 00.00:: .0002
0000 :0 000000 0 500.0 :.0— :0 000 :0 03.00 0.00 — 02.00 0000.0 00000 «.00 :0 000 :0 00.00 2.000 00.00 00.000 00.002
0000.0 0000.0 0 500.0 :0: :0 :.00 20 030.0 0.00: 00.00 000.0 0000.0 000:0 v00 20 00.00 0.000 00.00 00.000 00.03
00000 000000 0800.0 :0: :0 000.0 «3.0.0 0.00 P 00.00 0000.0 0000.0 000 20 :00 :0 00.00 0.000 00.00 00.000 00.03
00 :000 0500.0 00000.0 00: :0 :00 :0 0000.0 0.00: :00— 0000.0 0000.0 0000.0 0000.0 00.0: :00 00.00 00.:.00 00.000
3.0000 00000.0 00000.0 02 ..0 000 ..0 0000.0 v.00 : 00.0: 005.0 030.0 030.0 0000.0 00.0: 3.00 0:..00 00.000 0.000
00000.0 00000.0 00000.0 0:; 20 0 :0 20 0000.0 0.00: 00.:; 00 80 0.00.0 R000 800.0 00.0: 0600 £0.00 00.000 I00.0u0

40:: > 40:: > 40:: > 4 05 4 .S 422.0 > 32:: ~02 2.: 00a: :21 .01 32.0. W> ::.... W: 2.3. .0 50.: m: ENE. 002.00 .50th

0.0 23:00: 00 ME: 40> m2: 40> 220% E00 00 23:0M 00 03:00: 00 5:00

:003. M003. moms. 00020....

 

 

 

 

 

 

 

 

 

 

 

 

 

00 0 00m :0: 000020200 2000028230 3020 0:003 2: 053020 00020 00800 Boxm ”07m 200,—.

 

177

 

 

 

.2885 8.382: 2: 2 893:8 8:3. N882 28 :0 .No 2: 8 8882: Etna 2: 8 Eu» 2: 2 E»? .m 0.36

 

8 8 88285.8 .0 3:" 8 .8382: Eta: 2.. 8. ::.. 6.. 8

 

H.282 8 8:8: 8.53882 8 3.89.282 8 2582: 8:8: 2: 2 2.; .01 580.2

 

8 .6 8.8.85.8 8 min 8 .8582... 3:8 2.. 8. E... 6.. 8

 

Ban: 5 28:38 8 5285: BE 05 2 k 403 N0 0594

 

28.. s 288. .6 :88: BE 2.. a. u .82 2‘82 98:

 

and: s 838.. .o :88: .6... 8. .... u .8: 8 98:

 

28.. s 08:5. .o .822... 38.2. 2.. a. $3. 8 Q8:

 

.28.. s 288. .o .888 £92. 2.. a. $3 SE8. Q8:

 

 

 

 

 

 

 

 

 

 

 

28.. s 2.28-. B .828 £92. 2.. 2 $5 8 2:0:

 

2.22:8 8 332: 88822: 2: 53. 828 8:5: 2: S 22 So 8 88:. 8 82.5: 2: he:

 

3:38 .8 8.8.8 823 use: 5 2.2.38 3 299: 2: 2 a 8 28:

 

$282 8 :88: 88885 2: 5:: 828 2g 2: S 9282 8 888 8 .285: 2: 9.32:2: a: 82:28 82K 23:: S 9282 8 2842: 2: 2 m 5802 9:0:

 

2.226 8 822: 88885 2: 5:: 828 ::.o: 2: S 2.28.. 8 885 8 82.5: 2: 832:2: €528 82:82.5: 5 2.200.. 8 892: 2: 2 n no 0303

 

 

N0 8 88:. 8 a .22 2: E9: 828 8:2. 2: E 2.22:8 8 885 8 ... 2: 0158288 .3 82:28 828 8:5: 5 2.22:0 8 88.: 8 .8 2: 2 .53. No 050:

 

9.302 8 885 8 ... .22 2: :8: 828 8:2. 2: 5 M882 8 888 8 u 2: 8:09.88 .3 8228 828 8:5: 5 widow: 8 88.: 8 8. 2: 2 .5: 5:32 9:03

 

2:28 8 88:. 8 .285: .22 2: So: 828 822. 2: S 2.28.. 8 88:. 8 82:2. 2: #158288 .3 8528 828 83:: 5 2:23.. .8 885 8 82.5: 2: 2 do: 8 9:03

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

86 86 66.6
8.3 3...... 8.2. 86 86 886 886 886 $8.8 $8.8 $2.8 8.8 8.8 8.8 886 886 886
8.8 8.8 8.8 86 26 286 82.6 886 $8.8 $8.8 $5.8 :8 8.8 8.8 886 82.6 886
8.8 8.8 8.8 86 26 286 886 886 $8.8 $8.8 $2.8 :8 8.8 8.8 886 886 886
8.8 8.8 8.8 86 26 286 886 886 $8.8 $8.8 $2.8 2.8 8.8 8.8 886 886 886
8.8 8.8 2.8 26 86 886 886 886 $86 $2.8 $8.8 88 3.8 8.8 886 2.86 886
8.8 8.8 2.8 26 86 886 886 886 $8.8 $3.8 $86.8 8.8 8.8 8.8 886 2.86 886
8.8 8.8 2.8 26 86 886 886 886 $8.8 $2.8 $86.8 8.8 3.8 8.8 886 2.2.6 886
8.8 8.8 8.8 86 86 886 82.6 886 $2.2 $8.8 $8.88 8.8 8.8 8.8 886 886 886
8.8 8.8 8.8 86 86 886 886 886 $2.2 $8.8 $8.8 8.8 8.8 8.8 886 886 886
8.8 8.8 8.8 86 86 886 82.6 886 $2.2 $8.8 $8.88 8.8 8.8 8.8 886 886 886
8.2 8.2 8.2 86 86 886 22.6 .886 $8.2 $8.8 $8.88 8.8 8.8 8.8 886 82.6 886
8.2 8.2 8.2 86 86 286 82.6 286 $8.2 $8.8 $8.8 2.8 8.8 8.8 886 2.86 886
8.2 8.2 8.: 86 86 886 83.6 286 $2.2 $8.8 $2.88 .8 8.8 8.8 886 2.86 886
2.3 ... ::.. ... =5: .8 =5. .3 =5. .8 m .6: ... .8: m .8: $8. $5 $2. a .. u .. u .. .83: 8:: 8::

88:. 38 8 582 8 8 528. 8 8 582 8 8 5.82 8 8 8.82 8
8:52.. #8: go: £32.. @603, 2.8: 8:82.. 2.83 28: 2W8: :88: use:

 

 

VN a. cam uOm mCOflw—So—wnv comm—NHHCOOGOU Owwnm U—SU—i— 05 mcmgonm “OOH—m UNOHQW EQXW “AJZOUV mhnm 296R.

 

178

 

mm # 5.x H8 macaw—5030 cogs—80:00 82m 2:34 2: wagonm Beam 38mm Boxm ume 2an

 

179

 

 

_

_

A

r

_

 

68000:: 008820: 2: 0: 000.8280 22:0. W882 28 no .8 2: :0 0080020: 8:80: 2: :0 E8 2: 2 :21 .0: 0.30

 

_

8 :0 0000220580 :0 min 8 8200020 02:23.: 2 :00—0 .00 8

 

M8102 :0 :0:00: 800N838: :0 0:388:02 :0 088000 8:80: 2: 2 ::.; .00 003002

 

8 :0 0000200508 :0 2:» 8 82000000 02000 05 2 :02 .00 8

 

200:: 0: 002050 :0 00:02: :20 05 2 0 020 8 052.:

 

3:5: 0:: 080:0»: :0 03:09:: :05 2:: a: k 402 t110QQSG:

 

00:00:: 0:: 28:00-0 :0 20:09:: :20: 05 a: k 40: Q0 9303

 

202: s 002050 :0 000200 30:00. 05 2 00:. 8 032.:

 

202: 0: 20002 :0 000200 80:00. 05 2 00:. 05002 032.:

 

 

 

 

 

 

 

 

 

 

 

200:: s 00200.: :0 000200 30:00. 05 2 00:. 8 052.:

 

2:20.50 :0 2:22: 88022: 2: 0::: 002::: :0:00: 2: 0:: 2:2 0:0 :0 022: :0 80:8: 2: We:

 

0020.0 00 002200 00000 202: s 002050 :0 E20... 05 2 0. 8 032.:

 

N80200: :0 592: 88022: 2: 0::: 82:0: 20:0: 2: 0: N882 :0 0220: :0 82:2: 2: 9:32:00: :3 022200 002:: 0:02: 0: W882 :0 2:92: 2: 2 a 0:539 930::

 

28:00 :0 2:92: 88022: 2: 0::: 002:: 0:00: 2: 0: 28:00.0 :0 0200: :0 82:2: 2: 8:52:08 03 020830 002:: 082: c: 28:00.: :0 28:2: 2: 2 m 8 039.:

 

 

do :0 02000 :0 0. :20: 2: 80:: 82:: 88> 2: 0: 2:20.50 :0 0.22: :0 0. 2: 00:00:08 0.: 022200 82:0 032: 0: 202.38 :0 022: :0 :0 2: 2 .00:: «o 039.:

 

.2882 :0 02000 :0 0 8:0: 2: 0:0: 82:: 88> 2: 0: M8002 :0 02000 :0 0 2: 20:00:08 0.0 002200 28:: :0:00: 0:: N882 :0 022: :0 0 2: 2 .00:: 15002 0:39.:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Oflbuoo so mfisoe \O \OQEDC ‘fiuOu OE» 50.: QMfiCQ k0lfi> 00: Cu 0p3~00nh s0 OSOE ‘0 .EQEszL 0C: Mttufihhbm kn UO—éfiuflo OQNCQ b~30§ a: 020~00uh a0 Q0~OE \O kOQESC G2” 0‘ do: .wQ Q~3°~d
00.0 00.0 :0 00.0 00.0 0000.0 0000.0 020.0 2.000 2.00.00 2.00.00 00.0 0020 00.00 00.00 0000.0 000.0
00.0 00.0 :.0 00.0 00.0 0000.0 0000.0 000.0 2.00.0 2.00.00 2.00.00 00.0 00. .0 00.00 00.00 0000.0 000.0
00.0 00.0 00.0 00.0 00.0 0000.0 0000.0 00.00 2.000 2.00.00 2.00.00 00.0 00. .0 00.00 0000.0 0000.0 0000.0
00.0 00.0 00.0 00.0 00.0 0000.0 0000.0 00:00 2.000 2.00.00 2.00.00 00.0 00. .0 00.00 00.00 0000.0 0000.0
00.0 00.0 :00 00.0 0.0 0000.0 0000.0 020.0 2.000 2.00.00 2.00.00 00.0 00.00 00.00 00.00 0000.0 0000.0
00.0 00.0 00.0 00.0 0.0 0000.0 0000.0 0000.0 2.000 2.00.00 2.0000 00.0 00.00 5.00 0000.0 0000.0 000.0
00.0 00.0 :00 00.0 0.0 0000.0 0000.0 0000.0 2.00.0 2.00.00 2.00.00 00.0 00.00 00.00 00 5.0 .0000 000:0
00.0 00.0 00.0 00.0 00.0 0000.0 0000.0 0000.0 2.00.0 2.00.00 2.00.00 00.0 00.00 0.00 00 00.0 00000 000 :0
00.0 00.0 :00 :0 00.0 0000.0 0000.0 0000.0 2.00.0 2.00.00 °0.00.00 00.0 :00 00.00 200.0 0000.0 .0000
00.0 00.0 00.0 :0 00.0 0000.0 0000.0 0000.0 2.00.0 2.00.00 2.0000 00.0 :00 00.00 000.0 0000.0 000.0
00.0 00.0 00.0 :0 00.0 0000.0 0000.0 0000.0 2.00.0 2.00.00 2.0000 00.0 :00 00.00 000.0 0000.0 .0000
00.0 00.0 00.0 00.0 6.0 0000.0 0000.0 0000.0 2.00 2.00.00 2.00.00 00.0 00.00 00.00 0000.0 0000.0 00000
00.0 I000 00.0 00.0 00.0 0000.0 0000.0 0000.0 2.00 2.00.00 2.00.00 00.0 00.00 00.00 0000.0 0000.0 000 F

20.2 .0 200. .0 a: .00 :00. .00 000.0 .00 0 .0000 0 .8: 0 .8: 000.: 000:: 000.: 0 0 0 .2 0 .2 00:: .00:: 02.:
200:: 0.80 8 02002 8 8 02002 8 8 00.002 8 8 00.002 8 8 8002 8
E8: 0.0.2.020 E00: @520 PS2: 00:20 082.: £02.: Elia: 2:03 2.02.: 052.:

 

mm 0: Si 00: 002003200 002000000000 00020 2:05 05 9:326 000$ 082$ Boxm ”3:00: me 030E

 

 

180

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.0.0006 N0 508.06: 88.: 0086 SINK 00.8.8: A 20.00:. 88.0mm ml55._0> dwwm>
vomit? 0800.6 00 6866 866 886 86 00.86 W 0:000:02
vomlmwd. 802.6 N $0.102 306.06: 8006 886 86:. $8.00 00 86 6: No
Q < 800.: 0866 0086 6: 00.00.— No 8.8: 0.6:. 300280
8.08 0000000 000000 28 2200200 .8: 00:0: 0. .002: 000:: 2.00. 0 .0002

«05.88.: \ «0822. 002:0: . 00 .NO 2: :0 0800000: 8:80: 8:080 00: 2: :0 0:08 2:00 8 8228 002:0: .80 c: No :0 0208 :0 ... 2: 2 .53: No 0.50:3.

«0::. r886 \ «:0: 002:: .80 . 882 2: :0 .00:: .80: 8:080 00: 2: :0 0:08 88: N0 02:28 000cm .22. 0:: N882 :0 00:20: :0 :: 2:: 2 20:: 52:62 0:00:30
«2:882:6088200 002::: .80 . 00 2: :0 .0000: 80:: 8:080 00: 2: :0 0:08 8:2. 8 00528 002::: .80 0:: 28:00.: :0 020:: :0 t 2: 2 40:: 00 No.03.

002:: :28: 2.: :0 28:00. 8000:. 2: :0 28:00 :20: a «8:3: 002::: 88:. 2: :0 N3529—

0::0000 00 002000 200:: 022:0 05 :0 000.0 :20: u 02:: 00000 200: 05 :0 28.80

«5+1 nu 00:0: 0 2: 8 88:00:00: 2022. 8:08 :20: 2: 8 000800 82.: 8:08 2: 8 80:28 08:20: 2: :0 EmZNn

2:28: 0:: 8 00:80: 0020 8:08 2.88880 :08 2: :0 0080 80:: 82200 «N882 80.8: 2:888:00 2: :0 208:2. 8:08 2: :0 33m

«N882 :0 080008 8:82.80 :0 088000: 8:82:22: 0: 880 2 :: .000 :0 0800000: 8:80: 2: 2 SE: .0m N0

N882 :0 8:02:08. «N882 :0 0::: n N882 :0 0800000: 8:80: 2: 2 3:100: 8:362

05.0 Fm 8609 088.0 00 _ 0o :0 8:000:08. «00 :0 W00:" 00 808808.822: 2:: 2 ::.;0 .00 00

86.000. 8.8: 00806 N $0002 00 0: 8:88 00:28:00 N882 :0 08000:: 008:. 2: 2 ::.: 00m: NISBN:

0 0 0. no... 00300.. 000 E00. .0: .0 .25 00250200 M25021 _ _

.u+.:.:\N.<u $8.: 0: £90000 2:22:00 2:02? 8:20 8 00:28:00 088888: :0:00: :00 2: :0 00 :0 0800000: 88:. 2: 2 :5. (40:: 00

_ « _ _ _ _ _ .050 :0 2:0: 0:: 0800000: :0:00200§

0019:: :0 80:08:: «038.2: 80828: 08000:: 2: :0 8:22:00 2: 0: 8800000 080002: 00:00:80 2 Sm: .0: :0:00

I I _ _ 088888: :0:00: :00 .800 2: 2 8E

82828.: 0800000: 28000 2:: :0 82000: 2: 2 :00:me

80828.: 0.88000: :2: 2: :0 80:80: 2:: 2 .5800:

0:086 886 6:36 8006 0806 «.00.: 00.0 6000.: 88.: EN 6.8: 00.09 8.8: 963
03.86 0886 936 80:6 886 «.000 006 03.0.: 00.8.: 00.:. 6.8: 00.09 8.8— 08:
0 .086 8866 :36 80:6 8606 ::.: 00.0 080.— 56.: 8.0 0.80 00.80 20.8: 86:
$086 0886 836 8:6 «086 660: 8.0 986 886 :06 6.8 8.8: 00.8 30:.
8086 8086 80 F 6 8: 8 0086 660: 8.0 8006 886 Ed 0.8 00.8: 00.8 86:.
80066 06086 536 mm: :6 0086 0.8: 8.0 00006 6086 ohm 0.8 580 00.8 00.00
00086 cm :86 80:6 03 :6 0086 2.09 00.0 8006 w: 86 ~04. 0.8 00.02 0.060 0.0.8.
8086 60:86 80.06 m: 8 6086 6.03 8.0 0806 886 «0.0 0.8 3.3— 560. $60:
8086 0386 80:6 00:6 «0006 0.00: 00.0 3006 8606 «0.0 0.8 862 560 3.8.
0 F :86 0886 000 8 0.6: :6 00506 :60: ”00.0 :m 0 8 600 :6 86 6.0.0 5.8 8.0.0 8.0.:
m P :86 0886 0mm :6 06— 8 0.0006 0.00: 006 00: 06 0 00 F 6 86 6.00 0060 8.00 8.00
N w :86 00866 08 8 0o: :6 81006 6.00: med 8 0 :6 o :0 :6 066 0.8 860 00.00 8.0.0.0
0886 m 0866 0096 00606 0806 0.8: RN 0086 6086 6.0 v. S. 8.00 8.3. 8.00

.60: 0 20:: 0 .33: 0 .: 00> 20:. 48:: > 2.: .00: 3.: .00: 2.: .00: 3.; ...“: 05% 0m: E< .0: S2: .0: ENE. r588: 0Nm00m
No 0:000:02 00 3.5.5:. N3549— Emzmn 33m «0 0:000:02 00 N352 00 5:00

0:903. no.3: 0:00:30 682:”:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mN : =3: :0: 08:28:00 0200:8080 0025 E83 2: 836% 0028 08:05 885 ”m: -m 030,—.

181

 

 

200020 00:00.88 2:: 0: 0008080 00:00: N882 08 00 .00 2:: :0 0200020 8080 2:: :0 800 2:: 2 E; .0 0.:<0

 

_

_

00 :0 0000200580 :0 0:0:" 00 80500230800 2:: 0: 0::... .00 00

 

N808: :0 02:00:: :08. «N808: :0 0>:nN882 :0 0:00.020 80882:: 2 8.0! .00 08.000:

 

00 :0 0000200580 :0 0:0:" 00 80.00008 .0800 0s 0: ::.... .00 00

 

0:03: 0: 0020.50 :0 00:00:: :08 2:: 0: n: .003 NO 930:

 

030:: 0: 08%02 :0 02:00:: :oE 2:: 0: u: 00: 0,0300%: Q50:

 

0:03: 0:: 28:00-: :0 :0200:: :05 0::: 0: .0 00:0 00 9:00:

 

0:000 0: 0000000 :0 800.00 .0000. 0s 0: 00:0. 00 0:00...

 

0:000 0: 0.01000: :0 800.000.0000. 0.0 0: 00.0. .8000: 0:02.:

 

 

 

 

 

 

 

 

 

 

 

0:000 0: 000.00.: :0 800.00 30:00. 0::: 0: .80. 00 0:00:

 

2:00:08 :0 80:9: 8:000:08 2:: 0::... 002:0 0:02: 2:: 0: 0:0: 0:0 :0 00:08 :0 8080: 2:: 0:::

 

00:00.0 00 000.030 000...: 0:00.: 0: 000.050 :0 0:200. 0::: 0: 0 00 0:00:

 

N882 :0 084000 8:000:08 2:: 5:0. 002:0 0:02: 2:: 0: N882 :0 00:08 :0 8080:: 2:: 8:200:08 00 008200 002:0 0:02: 0: N882 :0 80:2: 2:: 2 0 0<002 030:0

 

28:00 :0 30.800 8:000:08 2:: 0:0: 002:0 0:02: 2:: 0: 28:00.: :0 00:08 :0 8080: 2::d0.50:0:::08 8 022200 002:0 0:02: 0: 28:00.: :0 30:00.. 2:: 2 0 00 030:

 

 

.00 :0 00:08 :0 0: :20: 2:: 80:: 002:0 :080 2:: 0: 2:20.50 :0 00:08 :0 ... 2:: 05:02:00» 0.0 020200 002:0 0:02: 0: 2:20.50 :0 00:08 :0 ... 2:: 0: 00:: N0 030:

 

N882 :0 00:08 :0 ... :20: 2:: 80:: 002:0 :0000 2:: 0: N882 :0 00:08 :0 ::. 2:: 08:02:08 00 008200 002:0 0:02: 0: N882 :0 00:08 :0 .0 2:: 2 00:: 00.000: 030:0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0:00:00 :0 OO:OE 00 00:20:30: 3:0: 002 :0:00: OWN—«Q kbnwfi) 0:: 0:: 00:800.: ‘0 Q3020 :0 0002230: 0:: Qantas; 0A0 tatsfiuao OWQCQ @205: 0:: 0000:0000 :0 @302: :0 00032:: 05 Q« do: 6.0 3:308:
.00 .00 00..., 8.0 00.. 0000.0 88.0 0000.0 0.00.0 .0200 00.0 8.0 8.00 020.0 0000.0
.00 .00 00.... 8.0 0.... 0000.0 88.0 0000.0 .0000 0.00.00 00.0 8.0 8.0: 020.0 0000.0
0.0.0 00.: 00.0 8.0 0...: 0000.0 88.0 0000.0 .0000 0.00.00 «...0 8.0 6.2 020.0 0000.0
.00 ::.... 00.... 8.0 00.0 0000.0 88.0 0000.0 .0000 .0300 00.0 8.0 00.0: 006.0 00.0.0.0
:00 .00 00.0 8.0 00.0 800.0 88.0 0000.0 .000... 0.0.8 3.0 8.0 00.0: 008.0 0000.0
00.0 00.0 8.... 8.0 00.0 0000.0 88.0 0000.0 0000.0 .0300 00.0 8.0 00.00 020.0 0000.0
0:... 00... 00... 8.0 8.0 0000.0 88.0 0000.0 .0000 $00.00 0... 8.0 5.2 008.0 500.0
0:... 0:... 00... 8.0 8.0 0000.0 88.0 0000.0 0000.0 .0800 00.0 8.0 .00:. 00 6.0 .0000
00... 00.0 00... 8.0 8.0 0000.0 080.0 0000.0 .0000 0000.00 00.0 8.0 :02 005.0 .000...
00.0 00.0 00.0 8.0 2.0 0000.0 88.0 020.0 0.00.0 0000.00 5.0 8.0 5.00 66.0 0000.0
00.0 00.0 2.0 8.0 2.0 0000.0 88.0 0000.0 .0000 0.00.00 .00 8.0 8.2. .050 0000.0
00.0 00.0 00.0 8.0 2.0 0000.0 88.0 020.0 0000.0 0.00.00 :00 8.0 :02 .200 800.0
.00 :00 00.0 8.0 00.0 28.0 88.0 300.0 .0000 0.00.00 00.0 8.0 00mm 0.00.0 0000.0

0E: .0 ::.: .0 0.0.: .00 0:0 .00 0800M 0 .80: ... .80: 0 .80. 020. 0.000 020. 00: 0 .. 0 .: .00: .80: .82:
28:: 0.80 00 0.800: 00 00 .8000: 00 00 0.800: 00 00 50000: 00 00 8.80: 00
P30: E00: 0000: 0000: 0:00: 0000: 0:00: @003 :0:00: .0000: R00: 050:

 

00 0: :3: :0: 2500—00—00 0002800000 0800 00.03 2:: 053000 8000 000.00 Boxm ”3:00: 07m 030.0.

 

 

APPENDIX C

APPENDIX C

This section contains the data used in PLS modeling concentration for each file number correlating to each
run

 

182

183

Table C-l: Data Files and the Correlating Concentrations Used in PLS Model

 

184

Table C-l (Cont.): Data Files and the Correlating Concentrations Used in PLS Model

 

LIST OF REFERENCES

LIST OF REFERENCES

Beebe, K.R., Blaser, W.W., Bredeweg, R.A., Chauvel, J.P., Harner, R.S., LaPack,
M., Leugers, A., Martin, D.P. Wright, L.G. Yalvac, E.D. Anal.Chem., 1993, 65,
199R-216R.

Yalvac, E.D. Paper No. 536, 16th Annual Meeting of the Federation of Analytical
Chemistry and Spectroscopy Societies, 1989, Chicago, Illinois.

Yalvac, E.D., Paper No. 209, 17th Annual Meeting of the Federation of Analytical
chemistry and Spectroscopy Societies, 1990, Cleveland, Ohio.

Yalvac, E.D., Bredeweg, R.A., Albers, D.R. Paper No. 66, 18th Annual Meeting of
the Federation of Analytical Chemistry and Spectroscopy Societies, 1991, Anaheim,
California, 1991.

Yalvac, E.D., Bredeweg, R.A. Process Quality and Control, 1992, 6(2-3), 91-6.

DesJardin, M.A., Doherty, S.J., Gilbert, J .R. LaPack, Shao, J ., Process Control and
Quality, 1995, 6(4), 219-29.

Aaljoki, K., Hukkanen, H., J okinen, P., Process Control and Quality, 1994, 6(2-3),
125-33.

Mayes, D.M., Koza, D.S., Cavinato, A.G., Process Control and Quality, 1993, 5(1),
1-9.

Babji, B.S., Saraf, S.N., Process Quality and Control, 1993, 4(3), 181-197.

Heeschen, W.A., Greminger, D.C., Yalvac, E.D., Process Control and Quality,
1992, 2(3), 215-25.

Cardis, T.M., Process Control and Quality, 1994, 6(2—3), 109—15.
Nielsen, J ., Process Control and Quality, 1992, 2(4), 371-85.
Morabito, P.L., Melcher, R.G., Process Control and Quality, 1992, 3(1-4), 35-43.

Garrison, A.A., Moore, C.F., Roberts, M.J., Hall, P.D., Process Control and Quality
1992, 3(1-4), 57—65.

185

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30

31

32

33

34

35

186

Moessner, R.C., Process Control and Quality, 1992, 2(3), 237—49.
O'Rourke, P.E., JNMM, 1989, 18(1), 85.

Harner, R.S., Proc. SPIE-Int Soc. Opt. Eng., 1992, 1681 (Opt. Based Meth. Process
Analysis), 256.

Burgess, L.W., Afromowitz, M.A., Proc. SPIE-Int. Soc. Opt. Eng. 1992, 1637
(Environ. Process. Monit. Technol.), 118.

Doyle, W.M., Proc. Control Qual.,l992, 2, 11.

LaPack, M.A., Tou, J.C., Enke, C.G., Anal. Chem, 1991, 63(15), 1631.
Srinivas, G., Chuang, S.C., J. Phys. Chem, 1994, 98(11), 3024-31.

Callis, J .B., Illman, D.L., Kowalski, B.R., Anal. Chem., 1987, 59(9), 624A.
Blaser, W.W., Ruhl, H.D., Bredeweg, R.A., Am. Lab., 1989, 21(1), 69.
Clevett, K.J., Bioprocess Technol., 1990, 47.

Guillemin, C.L., J. Chromatogr. 1988, 441(1), 1.

Farquharson, S., Chauvel, J.P., J. proc. SPIE-Int. Soc. Opt. Eng., 1988, 918, 80.
McPeters, H., Anal Chim. Acta., 1990, 238, 83.

Doyle, W.M., Jennings, N., Spectroscopy, 1990, 5, 34.

Saltzman, R.S., Mathews, M., Dell, C.D., ISA Calgary 89 Symp., 1989, 217.
Herzberg, G., Infrared and Raman Spectra, D. Van Norstrand Co.: New York, 1945.
Gordy, W., Martin, P. C., J. Chem. Phys. 1939, 7, 99-102.

Gordy, W., Phys. Rev. 1946, 69, 604—607.

Plyler, E. K., Benedict, W. 8., Ibid, 1952, 49, 1-6.

Rank, D. H., Rix, H. D., Wiggins, T. A., J. Opt. Soc. Am. 1953, 43, 157-162.

Wiggins, T. A., Shull, E. R., Rank, D. H., Ibid, 1953, 21, 1368—1373.

36

37

38

39

40

41

42.

43.

45.

46.

47.

48.

49.

50.

51.

52.

187

Goddu, F. R., Delker, D. A., Anal Cheml960, 32, 1, 140-141.
Huges, R. H., Martin, R. J., Coggeshall, N. D., J. Chem. Phys., 1956, 24,489.
Coggeshall, N. D., Anal. Chem, 1950, 22, 30, 381-395.

Griffiths, P. R., deHaseth, J. A., Fourier Transform Infrared Spectrometry, John
Wiley & SonszNew York 1986.

Esposito. L. D., Koenig, J. L., Fourier Transform Infrared Spectroscopy, Academic
Press:New York, 1978.

Spiro, I. J. Schlessinger, M., Infrred Technology Fundamentals, Marcel Dekker:
New York, 1989

Karge, H.G., Boldingh, E., Catal. Today, 1988, 3(5), 379-386.

Yin, Y., Te, M.,Jiao, F., Su, G., Proc. Int. Conf. Pet. Refin. Petrochem.
Process.,1991, 2, 614—20.

Chuand, S.C., Pien, 8.1., J. Mol. Catal., 1989, 55(1-3), 12-22.
Xia, C.,Sun, Y., Li, D., Fenzi Cuihua, 1989, 3(1), 62—71.

Garland, M., Bor, G., Inorg. Chem, 1989, 28(3), 410-413.

Pan, W., Liu, J .; Li, S.; Xia, Y.; Liu, D., (Dep. Chem. Chem. Eng, Qinghua Univ.,
Beijing, Peop. Rep. China). Gaodeng Xuexiao Huaxue Xuebao (Zhuankan), 1982,
107-13.

Mirbach, M.J., Mirbach, M.F., Saus, A., Topalsavoglou, N., Tuyet, N.P., J. Am.
Chem. Soc., 1981, 103(25) 7590-4

Moser, W.R.; Papile, C.J.; Brannon, D.A.; Duwell, R.A.; Weininger, S.J., J. Mol. Catal.,
1987, 41(3), 271-92.

Sheridan, R., Rein, A., Research and Development, 1991, 33(10).
Landau, R.N., Rein, A.J., Research and Development, 1993, 35(13), 39-40.

Giusto, R.A., Sevent RC User forum USA, Proceedings, Mettler-Toleddo, Inc,
Hightstown, NJ, 1993.

 

53.

54.

55.

56.

57

58

59

60.

61.

62.

63.

65.

66.

67.

68.

69.

70.

71.

188

Moser, W.R., Berard, J .R., Peter, J .M., Burger, R.J., Applied Spec., 1992, 46(7),
1105-12.

Altman, L.J., Jalkian, R., US Patent, 1993, Patent No., 5,407,830.
Rekoske, J .E. Barteau, M.A., Ind. Eng. Chem. Res., 1995, 34(9), 2931-9.

Atkinson, R., Tuazon, E.C., Aschmann, S.M., Environ. Sci. Technol., 1995, 29(9),
1860-6.

Atkinson, R., Tuazon, E.C., Aschmann, S.M., Environ. Sci. Technol., 1995, 29(6),
1674-80.

Espeel, P.H., De Peuter, G., Tielen, M.C., Jacops, P.A., J. Phys. Chem., 1994,
98(44), 11588-96.

Jiang, S., Li, C., Xin, Q., Cuihua Xuebao, 1993, 14(3), 229-33.

Cariati, F., Rindone, B., Phys.=CHem. Behav. Atmos. Polut., 1990, 400-7.
Martens, A., Crouzet, P., Eur. Pat. Appl., 1989, EP-345182.

Wenig, T.W.,Ames Lab Report, IS-T-1313; 1989, Order No. DE870109337, 220pp.
Moser, T.P. Ames Lab. Report, IS-T—1307, 1987, Order No. DE87010935, 143pp.
Wenig, T.W. Schrader, G.L., J. Phys. Chem.,1987, 91(22), 5674-80.

Wenig, T.W. Schrader, G.L., J. Phys. Chem.,1987, 91(7), 1911—18.

Wenig, T.W. Schrader, G.L., J. Phys. Chem.,1986, 90(24), 6480—8.

Koizumi, H. Suzuki, Y., Bunseki Kagaku, 1983, 32(8), 489-93.

Dobrov, V.S., Klimentko, P.L., Sopkina, A.K., Gordash, Y.T., Malova, L.G.,
Neftepererab. Neftekhim, 1980, 19, 41-5

Sojak, L., Ostrovsky, I., Kubinec, R., Kraus, A., J. Chromatogr., 1992, 609(1—2),
283-8.

Listemann, M.L., Waller, F.J., Herman, F.L., Spectroscopy, 1993, 8(5), 40-5.

Gurka, D.F., Titus, R., Anal Chem., 1986, 58(11), 2189—94.

72.

73.

74.

75.

76

77

78

79

80

81

82.

83.

84.

85.

86.

87

88

89

90.

189

Morin, P. Caude, M., Rosset, R., J. Chromatogr., 1987, 407, 87—108.

Takeuchi, K., Hanaoka, T., Matsuzaki, T., Sugi, Y., Ogasawara, S., Abe,
Y.,Misono, T., Stud. Surf. Sci. Catal., 1995, 92, 269-74.

Johnson, S.A., Rinkus, R.M., Diebold, T.C., Maroni, V.A., Appl. Spectrosc., 1988,
42(8), 1369-75.

Lachenal, G., Vibrational Spectroscopy, 1995, 9, 93—100.

Khettry, A., Hansen, M. G., Antec, 1995, 2824-2831.

DeThomas, F. A., Hall, J. W., Monfre, S. L., Talanta, 1994, 41, 3, 425—431.
Mijovic, J ., Andjelic, S., Kenny, J ., Polymers for Advanced Technologies, 1995

Osborn, B. G., Feam, T., Near Infrred Spectroscopy in food Analysis, Wiley, New
York, 1980.

Watson, C. A., Anal. Chem., 1977, 49, 9, 835A-840A.

Archibald, D., Miller, C. E., Lin, L. T., Honigs, D.E., Apl. Spectrosc., 1988, 42,
1549.

Maggard, S.M., US Patent, 1990, Pat. No., 0,506,391.
Parisi, A.F., Nogueiras, L., Prieto, H., Anal. Chim. Acta., 1990, 238(1), 95—100.

Brackemann, H., Buback, M., Voegele, H.P., Makromol. Chem., 1986, 187(8),
1977—92.

Buback, M., Tups, H., Physica B+C, 1986, 139—140(1-3), 626-8.

Lysaght, M.J., Air Force Inst. of Tech., Report No., AIFT/CI/CIA—91—013D,l991,
258 pp.

United States Patent, 1992, Pat. Num. 5,151,474.
United States Patent, 1992, Pat. Num. 5,155,184.
United States Patent, 1984, Pat. Num. 4,469,853.

Sharaf, M.A., Illman, D.L., Kowalski, B.R., Chemometrics, 1986, John Wiley &
Sons, New York.

190

91. Haaland, D.M., Thomas, E.V., Anal. Chem. 1988, 60, 1193.
92. Haaland, D.M., Thomas, E.V., Anal. Chem. 1988, 60, 1202.
93. Haaland, D.M., Anal. Chem. 1988, 60, 1208.

94. Brown, C.W., Spectroscopy, 1986, 1(4), 32.

95. Jackson, J E A User Guide to Principle Components, 1991, John Wiley, New
York.

96. Massart, D.L., Brereton, R.G., Dessy, R.E., Hopke, P.K., Spiegelman, C.H.,
Wegscheider, W., Eds. Chemometrics Tutorials, 1990, Elsevier, Amsterdam.

97. Kowalski, B.R., Chemometrics, Mathematics and Statistics in Chemistry., 1983, D.
Reidel Publishing Co., Dordrecht, Holland.

 

98. Thomas, E.V., Haaland, D.M., Anal. Chem., 1990, 62, 1091-99.

99. Marjoriemi, M., Mantysalo, E., J. Am Leather Chem. Assoc., 1992, 87, 249-58.
100. Lin, J ., Brown, C.W., Appl. Spectrosc. 1993,47, 239.

101. Wei, W., Mo, 2., Yao, S., Anal. Chim. Acta, 1991, 25, 143-8.

102. Blanco, M, Coello, J ., Iturriaga, H., Maspoch, S., de la Pezuela, C., Talanta, 1993,
40, 1671-6

103. Varadi, M., Hruschka, W., Norris, K.H., Acta. Aliment, 1992, 21, 95-106.
104. Arakaki, L.S.L., Burns, D.H., Appl. Spectrosc.,l992, 46(12), 1919-28
105. Bellon, V., Sens. Actuators, B 1993, 12, 57-64.

106. Kettaneh-Wold, N., Chemom. Intel]. Lab. Syst., 1992, 14, 57-69.

107. Gauglitz, G., Kraus, G., GIT Fachz. Lab. 1992, 36, 232-5, Chem. Abstr. 1992, 268,
115-22.

108. Ni. Y, Anal. Chim. Acta., 1993, 284(1), 199-205.

109. Van de Voort, F., Sedman, J ., Emo, G., Ismail, A.A., J. AOAC Int.,1992, 75(5),
780-5.

110.

11

p—-

112.

113.

114.

115.

116.

117.

118.

119.

120.

12.

p—a

122.

123.

124.

125

126.

127.

128.

191

Miller, C.E., Appl. Spectrosc.,1989, 43(8), 1435-43.

. Ketterer, M.E., Biddle, D.A., Anal. Chem, 1992, 64, 1819-23.

Lysaght, M.J., Kelly, J.J., Callis, J.B., Fuel, 1993, 72, 623-31.

Lee, M.H., Cavinato, A.G., Mayes, D.M., Rasco, B.A., J. Agric. Food Chem, 1992,
40, 2176-81.

F. Seitz, Modern Theory of Solids, 1940 Me Graw-Hill, New York.
Bridgman, P.W., Physics of High Pressure, 1949, G. Bell and Sons, Ltd., London.

Frank, C.W., Electronic Transition and the High Pressure Chemistry and Physics of
Solids, 1973, Chapman and Hall, London.

Ferraro, J .R., Vibrational Spectroscopy at High External Pressures, 1984, Academic
Press, Inc., New York.

Griffith, J .S., The Theory of Transition Metal Ions, 1964, Cambridge University
Press, London.

Fisher, D.C., Drickamer, H.G., J. Chem Phys., 1971, 54, 4825.

Bargeron, C.B., Drickamer, H.G., J. Chem Phys., 1971, 55, 3471.

Drier, T., Schiff, G., J. Chem. Phys., 1994, 100(9), 6275-89.

Ulivi, L., Lu. 2., Tabisz, G.C., Phys. Rev., A: Gen. Phys., 1989, 40(2), 642—51.

Morin, P.H., Beccard, B., Caude, M., Rosset, R., J. of High Res. Chromatog. &
Choromatog. Communications, 1988, 11, 697-705.

Taha, S., Tosson, M., Thermochim. Acta, 1994, 236 (1-2), 217—226.

Lin, S.B., Koenig, J.L., J. Polym. Sci., Polym. Phys. Ed., 1983, 21(10), 2067-83.
Fernandez Bertran, J., La Serna, B., J. Mol. Struct., 1979, 56(2)283-8.
Fukushima, K., Kawai, T.,J. Mol. Struct., 1984, 112 (1—2), 51-7.

Campbell, J.H., Fisher, J.F., Jonas, J., J. Chem Phs., 1974, 61(1),346-60.

129.

130.

13.

p—

132

133

134.

135.

136.

137.

138.

139.

140.

141

142

143

144

145

146

147.

148.

192

Karlsson, A., Klaeboe, P., Nielsen, C.J., J. Raman Spectrosc., 1992, 23(3), 167-80.
Tanabe, K., Hiraishi, J., Tamura, T., J. Mol. Struct., 1976, 33(1), 19-37.

Van der Veken, B.J.,Coppens, P., Sanders, R.S., Daeyaert, F.F., Durig, J .R., J. Mol.
Struct. 1992, 272, 305-46.

Femandez-Ballester, G., Castresana, J ., Arrondo, J .L.R., Ferragut, J .A., Gonzalez-
Ros, J.M., Biochem. J., 1992, 288(2), 421-6.

Kato, M., Taniguchi, Y., J. Phys. Chem., 1994, 98 (10), 2688-93.
Carles, M., Eloy, D., Pujol, L., Bodot, H., J. Mol. Struct., 1987, 156(1-2), 43-58.

Hoshino, N., Inabe, T., Mitani, T., Maruyama, Y., Bull. Chem. Soc. Jpn., 1988,
61(12), 4207-14.

Minomura, S., Drickamer, H.G., J. Chem Phys., 1961, 35 903.
Perez—Albueme, E.A., Drickamer, H.G, J. Chem Phys., 1965, 43, 1381.

Sherman, W.F., Wilkinson, G.R., Advances in Infrared and Raman Spectroscopy,
1980, Heydon, London.

Snyder, R.W., Sheen, C.W., Painter, P.C., Applied Spectroscopy, 1988, 42(3), 503—
8

Efimov, Y.Y., Naberukhin, Y.I., Opt. Spektrosk. 1982, 52(6), 999-1003.
Gordon, I.G., Adv. Magnetic Resonance, 1968 3, 1.

Gordon, R.G., J. Chem. Phys. 1963, 39, 2788.

Arroume, M., Idrissi, A., Turrell, G., J. Molecular Struct., 1992, 266, 153-158.

Sugitani, A, Ikawa, S., Konaka, S., Chem. Phys., 1990, 142, 423-30.
Gordon, I.G., Adv. Magnetic Resonance, 1968 3, 1.

Gordon, R.G., J. Chem. Phys. 1963, 39, 2788.

Ingle, J.D., Crouch, S.R., Spectrochemical Analysis, 1988, Prentice Hall, New Jersey.

McOmber, J .I., Shriver, D.F., Ratner, M.A., Ferraro, J .R., J. Phys. Chem. Solids,
1982, 43, 903.

149.

150

15.

p..-

15.

N

15.

L»)

154.

155.

156.

157.

15

00

159.

160.

16.

p—g

162.

16.

b.)

164.

165.

166.

167.

193

Adams, D.M., Tan, T.K., J. Phys. Chem. Solids, 1981, 42, 559.

Wu, C.K., J .Polymer Science, 1974, 12, 2493-506.

Adams, D.M., Ekejiuba, I.O.C., J. Chem. Soc. Faraday Trans, 1981, 77, 851.
Ferraro, J.R., Coord. Chem. Rev., 1979, 29, 1.

Adams, D.M., Sharma, S.K., J. Chem. Soc. Faraday Trans, 1978, 2 (74), 1355.
Adams, D.M., Appleby, R., Inorg. Chim. Acta., 1978, 26, 643.

Adams, D.M., Sharma, S.K., J. Mol. Struct., 1981,71, 121.

Breitinger, D., Kirsten, U., J. Solid State Chem., 1977, 22, 233.

Dultz, W., Krause, H., Winchester, L.W., J. Chem. Phys., 1977, 67, 2560.

. Greig, D., Shriver, D.F., Ferraro, J.R., J. Chem. Phys., 1977, 66, 5248.

Gardiner, D.J., Walker, N.A., Dare-Edwards, M.P., Spectrochimica Acta.,l987,
43A(1), 21-34.

Gardiner, D.J., Walker, N.A., Dare-Edwards, M.P., Spectrochimica Acta.,l987,
43A(10), 1241-47.

Haaland, D. M., Easterling, R. G., App1. Spec., 1980, 34, 539.
Haaland, D. M., Easterling, R. G., App1. Spec., 1982, 36, 665.
Haaland, D. M., Easterling, R. G., Vopicka, D. A., Appl. Spec., 1985, 39, 73.

Brown, C. W., Lynch, P. F., Obremski, R. F., Lavery, D. S., Anal. Chem., 1982, 54,
1472.

Haaland, D. M., Practical Fourier Transform Infrared Spectroscopy, Academic
Press, Inc.: New York, 1990.

Martens, H., Naes, T., Multivariate Calibration, Wiley, New York, 1989.

Thomas, E. V., Anal Chem., 1994, 66, 795A

168.

169.

170.

17.

p—d

172.

173.

174.

175.

176.

177.

178.

179.

180.

18

H

182.

183.

184.

185.

194

Brown, S.D., Blank, T.B., Sum, S.T., Weyer, L.G., Anal. Chem., 1994, 66 (12),
315R.

User Guide, PLSplus/IQ, Galactic Industries Co. 1996, Salem, NH.

Bellamy, L. J .; The Infrared Spectra of Complex Molecules, John Wiley & Sons,
New York, 1975.

Dollish, F. R., Fateley, W. G., Bentley, F. F. Characteristic Frequencies of Organic
Compounds, John Wiley & SonszNew York, 1974.

Nyquist, R. A. The Interpretation of Vapor-Phase Spectra, Group Frequency Data,
Sadtler Research Laboratories, 1984.

Pouchert, C. J ., The Aldrich Library of FT-IR spectra, Aldrich Chemical Company,
Inc., 1985.

Rea, D. G., Anal. Chem. 1960, 32, 1638.

Crawford, Jr., B. L., Lancaster, J. E., Inskeep, R. G., J. Chem. Phys. 1953, 21, 678.
Lin-Vien, D., Colthup, N. B., Fataley, W. G., Grasselli, J. G., The handbook of
Infrared and Raman Characteristic Frequencies of Organic Molecules, Academic
Press, Inc., San Diego, CA, 1991.

Thompson, H. W., Torkington, P., Trans. Faraday Soc. 1945, 41, 246.

Rasmussen, R. S., Brattain, R. R., J. Chem. Phys. 1947, 15, 120.

Rasmussen, R. S., Brattain, R. R., J. Chem. Phys. 1947, 15, 131.

Rasmussen, R. S., Brattain, R. R., J. Chem. Phys. 1947, 15, 135.

. McMurray, H. L., Thornton, V., Anal. Chem., 1952, 24, 318.

Colthup, N. B., Daly, L. H., Wiberley, S. E.; Introduction to Infrared and Raman
Spectroscopy; Academic Press:San Diego, 1990.

Struve, W. S.; Fundamentals of Molecular Spectroscopy, John Wiley & Sons: New
York, 1989.

Barrow, G. M.; Introduction to Molecular Spectroscopy, McGraw-Hill: York,l962.

Kaye, W., Spectrochemica Acta, 1954, 6, 257-287.

186.

187.

188.

189.

190.

19.

p—n

192.

193.

194.

195.

196.

197.

198.

199.

200

20

H

195

Gaytguer, G.; J .Phys. Radium, 1953, 14, 19-24.
Mecke, R., Discuss. Faraday Soc., 1950, 9, 161-171.

Vaan Ness, H.C., Abbott, M.M., Classical Thermodynamics of Nonelectrolyte
Solutions, 1982, McGraw-Hill, New York.

Reid, R.C., Prausnitz, J .M., Poling, B.E., The Properties of Gases and Liquids,
1987, 4th Edition, McGraw-Hill, New York.

Navarro-Villoslada, F., Perez—Arribas, L.V., Leon—Gonzalez, M.E., Polo-Diez,
L.M., Anal. Chim. Acta., 313, 1995, 93-101.

Rohrback, B.G., Trends in Anal. Chem. 1991, 10(9), 269-261.

Kelly, J.J., Barlow, C.H., Jinguji, T.M., Callis, J.B., Anal. Chem. 1989, 61(4), 313-
320.

Malinowski, E.R., Factor Analysis in Chemistry, 1991, 2nd edition, Wiley, New
York.

Geladi, P., Anal. Chim. Acta., 1986, 185, 1—17.

Haaland, D.M., Computer—Enhanced Analytical Spectroscopy, 1992, V3, 1-30,
Plenum Press, New York.

Harris, J.M., Frans, S.D., Poston, P.E., Wong, A.L., IN Computer-Enhanced
Analytical Spectroscopy; Meuzelaar, H.L.C. Ed.,Plenun Press, New York, 1990,
2,3-36.

Miller, J.N., Analyst, 1991, 116,3—14.

Brown, P.J., Anal. Proc., 1990, 27,303-6.

Mark, H., Anal. Chim. Acta, 1989, 223, 75-93.

Adams, J .M., Chemometrics in Analytical Spectroscopy, The Royal Society of
Chemistry, 1995, Cambridge, England.

. Meloun, M., Militky, J ., Forina, M., Chemometrics for Analytical Chemistry, 1992,

Ellis Horwood, LTD., West Sussex, England.

196

202. Brereton, R.G., Chemometrics Application of Mathematics and Statistics to
Laboratory Systems, 1990, Ellis Horwood LTD., West Sussex, England.

 

   

"111111111111111111111115