PARTIAL OXIDATION
OF SOLID ORGANIC WASTES
This final report (SW-7rg) on work performed under
Research Grant No. EC-00263
(formerly successive Grant Nos. were SW-00022 and UI-00552)
to the Rensselaer Polytechnic Institute
was written by WILLIAM W. SHUSTER
and has been reproduced as received from the grantee with
the exception of a minor change in the introduction.
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
Environmental Health Service
Bureau of Solid Waste Management
1970
-------
Public Health Service Publication No. 2133
LIBRARY OF CONGRESS CATALOG CARD NO. 74-609260
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, B.C., 20402 - Price $1
-------
FOREWORD
Recent applications of scientific and engineering knowledge to the
development of new products and processes have had significant effects
upon the health, comfort, and well-being of the Nation's people. While
such technical advances are, of course, most desirable, they have at the
same time, ever increasing stresses upon man's environment. One particularly
pressing problem is that of solid waste disposal. Not only have changing
patterns in social structure affected the quantity of solid wastes, but
also the quality and composition have changed. The introduction of
increased amounts of materials resistant to biological and chemical
degradation has made the disposal of solid wastes increasingly difficult.
The approach used in the present work has been to consider the
possibilities of utilizing the large organic portions of domestic solid
wastes as building blocks for useful compounds through a process of
pyrolysis, and partial combustion. Such an approach is entirely consistent
with the objectives of the Solid Waste Disposal Act of 1965,to study ways
for reducing the amount of solid waste and to consider the recovery and
utilization of potential resources in solid wastes.
This report has been prepared to describe the approach that has been
used to produce useful products from the Nation's huge store of solid waste
materials. The methods and techniques described and the results obtained
to date will be useful to other workers in the field in pointing the
direction to a potentially useful new method of solid waste disposal.
RICHARD D. VAUGHAN, Director
Bureau of Solid Waste Management
±±±
-------
-------
Table of Contents
III. Theoretical Considerations
I. Introduction 1
II. Background 3
A. General 3
B. Incineration 3
C. Pyrolysis ..... 5
D. Gas Chromatography 6
E. Infrared Spectroscopy 7
A. Partial Combustion 9
B. Fluidization 11
C. Gas Chromatography 12
D. Infrared Spectroscopy 14
IV. Materials and Apparatus 17
A. Materials 17
1. Analytical Standards 17
2. Carrier Gases 17
3. Indicating Papers 18
4. Reagents for the Analysis of Formaldehyde
and Methanol 18
5. Reagents for Syringe Reactions 18
B. Apparatus 18
1. Reactor System Equipment 20
2. Sample Collection Equipment 24
3. Analytical Equipment 24
V. Experimental Procedure 27
A. Paper 27
1. Reactor Operation 27
2. Gas Chromatograph Operation 29
-------
VI.
VII.
VIII,
IX.
Page
B. Dried Sewage Sludge (Orgro) 32
1. Reactor Operation 32
2. Analytical Run 33
C. Dried Leaves 37
1. Reactor Operation 37
2. Analytical Runs 38
Results 43
A. Reaction Runs Using Paper 43
B. Reaction Runs Using Orgro 53
C. Reaction Runs Using Leaves 67
Discussion 85
A. Studies with Paper 85
B. Studies with Orgro 87
C. Studies with Leaves 89
Summary and Conclusions 95
Literature Cited 97
List of Tables
Table 1 Table of Infrared Absorption Frequencies
for Liquid Samples
Table 2 Reagents and Conditions for Syringe Reactions
Table 3 Feed Materials Composition
Table 4- Summary of Paper Reaction Runs
Table 5 Summary of Retention Times
Table 6 Summary of Leaf Reaction Runs
Table 7 Identification of Chromatogram Peaks for
Pyrolysis of Leaves at 250 C-300°C
Table 8 Summary of Identified Reaction Products
Page
16
40
43
51
52
69
84
96
VI
-------
List of Figures
Page
Figure 1 Cellulose Molecule 10
Figure 2 Reactor System Schematic 19
Figure 3 Brass Reactor 21
Figure 4 Black Iron Reactor 22
Figure 5 Ceramic Reactor 23
Figure 6 Gas Sampling Apparatus 31
Figure 7 Infrared Gas Cell Loading Apparatus 34
Figure 8 Typical Preparative Gas Chromatograph
Collection Trap 36
Figure 9 Chromatograph Run - Sample #8 45
Figure 10 Chromatograph Run - Sample #9 46
Figure 11 Chromatograph Run - Sample #11 47
Figure 12 Chromatograph Run - Sample #8 + Acetic Acid 48
Figure 13 Chromatograph Run - Sample #11 + Formaldehyde 49
Figure 14 Chromatograph Run - Sample #11 + Formic Acid 50
Figure 15 Chromatogram of Organic Liquid Effluents from
a Partial Oxidation of Dried Sewage Sludge 54
Figure 16 Chromatogram of Organic Liquid Effluents from
a Pyrolysis of Dried Sewage Sludge 56
Figure 17 Infrared Absorption Spectrum of Low-Boiling
Effluents from Partial Oxidation Held at -78°C
and Evacuated to 1/2 Atmosphere 57
Figure 18 Infrared Absorption Spectrum of Low-Boiling
Effluents from Partial Oxidation Held at 0 C
and Evacuated to 1/3 Atmosphere 58
Figure 19 Infrared Absorption Spectrum of Low-Boiling
Effluents from Partial Oxidation Held at 25°C
and Evacuated to 1/3 Atmosphere: First
Extraction 59
vii
-------
Figure 20 Infrared Absorption Spectrum of Low-Boiling
Effluents from Pyrolysis Held at -78°C and
Evacuated to 1/2 Atmosphere 60
Figure 21 Infrared Absorption Spectrum of Low-Boiling
Effluents from Pyrolysis Held at 0 C and
Evacuated to 1/3 Atmosphere 61
Figure 22 Infrared Absorption Spectrum of Low-Boiling
Effluents from Pyrolysis Held at 25°C and
Evacuated to 1/3 Atmosphere: First Extraction 62
Figure 23 Infrared Absorption Spectrum Using a Sodium
Chloride Liquid Cell of an Unknown Liquid
(approximately 95% Pure) 63
Figure 24- Infrared Absorption Spectrum Using a Sodium
Chloride Liquid Cell of 99.9% Pure Toluene 64
Figure 25 Potassium Bromide Pellet of Unknown Salt 65
Figure 26 Potassium Bromide Pellet of Reagent-Grade
Ammonium Carbonate 66
Figure 27 Infrared Absorption Spectrum of an Unknown
Liquid Fraction (95% Pure) 68
Figure 28 Chromatogram of Vapor Above Liquid Sample
From Pyrolysis of Leaves Before Venting 70
Figure 29 Chromatogram of Vapor Above Liquid Sample
From Pyrolysis of Leaves After Venting 71
Figure 30 Chromatogram of Vapor Above Liquid Sample
and Acetic Anhydride 72
Figure 31 Chromatogram for Pure Methanol and Acetic
Anhydride 73
Figure 32 Chromatogram of Vapor Above Liquid Sample
and Sodium Hydroxide 74
Figure 33 Chromatogram of Vapor Above Liquid Sample
and Potassium Permanganate 75
Figure 34- Chromatogram of Vapor Above Liquid Sample and
Sodium Borohydride 76
Figure 35 Chromatogram of Vapor Above Liquid Sample
and Sodium Borohydride Plus Ethyl Alcohol 77
viii
-------
Page
Figure 36 Chromatogram of Vapor Above Liquid Sample and
Isopropyl Alcohol 78
Figure 37 Chromatogram of Vapor Above Liquid Sample and
Acetone 79
Figure 38 Chromatogram of Vapor Above Liquid Sample and
Hydroxylamine Hydrochloride 80
Figure 39 Chromatogram of Vapor Above Liquid Sample and
Concentrated Sulfuric Acid 81
Figure 40 Chromatogram of Vapor Above Liquid Sample and
Hydrogen Iodide and Sodium Bicarbonate 82
Figure 41 Chromatogram of Vapor Above Liquid Sample and
7:3 Sulfuric Acid 83
-------
Acknowledgement
The author wishes to recognize the assistance of Edward F.
Maziarz, Jr., Joel S, Gilbert, and Laurence K. Burnell in the technical
phases of this study. In addition, thanks are given to various members
of the staff of the Chemistry Department at Rensselaer Polytechnic Insti-
tute for their advice and assistance in the analytical phases of this
work, and for the use of their infrared spectrometer.
This investigation was supported by Public Health Service Grants
Nos. 1R01-SW-00022-01, 5R01-SW-00022-02 and 5R01-UI-00552-03.
-------
Abstract
A study has been made of the possibility of utilizing the organic
content of solid municipal waste by reforming high molecular weight organic
compounds into simpler compounds of economic interest. Major components of
waste such as paper and leaves have been investigated. In addition, a dried
sewage sludge used as a soil conditioner and containing a large amount of or-
ganic matter was also studied.
Reforming of organic molecules in the solid waste was accomplished by
a process of partial combustion. Finely divided waste was supported as a
fluidized bed in an air-nitrogen stream containing less oxygen than that re-
quired for complete combustion. Heat was supplied electrically through the walls
of the reactor to supplement the exothermic heat of reaction.
Gaseous products of reaction evolved from the reactor were condensed
and collected in a series of traps held at progressively lower temperatures.
The fractions obtained included a tar fraction, an aqueous solution, liquid
organics, and a fraction of uncondensed gases. Analytical methods which were
applied and utilized in the examination and identification of major components
in the complex mixtures included wet chemical methods, gas chromatography used
in conjunction with peak attenuation and a syringe reaction technique, infrared
spectrescopy and mass spectroscopy.
Runs were made in a temperature range of 250°C to 1000°C and with air-
nitrogen mixtures that ranged from 0% air to 100% air. Products which were
obtained and identified included water, acetic acid, formic acid, formaldehyde,
methanol, acetone, toluene, acetaldehyde, methyl acetate, ethyl vinyl ether,
methane, carbon dioxide, carbon monoxide, propylene, ethylene, ammonia, ammonium
xi
-------
carbonate and hydrogen. Many of the same products were found in all runs made,
In general, runs with high air-nitrogen ratios favored the formation of more
highly oxygenated compounds , while low air-nitrogen ratio promoted the forma-
tion of hydrocarbons.
xii
-------
Introduction
One of the most pressing problems facing our nation at present is the
problem of disposing of the huge quantities of solid wastes which accumulate as
a by-product of our modern society. Until fairly recently, people of this country
have not been particularly concerned with solid waste disposal. In the past, it
has been common practice to dispose of refuse by the most expedient method at
hand. Such a method might be open burning or the use of an open dump.
This situation has now changed. With the rapid increases in population
during the past few decades, people are now crowding in on one another to the
point where it has become increasingly difficult to hide our refuse. In addition
to the increase in total quantity of waste produced through population increase,
the quantity of waste produced per capita has virtually doubled since 1920 be-
cause of changes in our living standards. Besides these changes in total quantity,
the chemical and physical composition of refuse has changed, which in general, has
complicated the problem of satisfactory disposal. The introduction of new plastic
materials which are resistant to or completely unaffected by bacterial or chemical
action, has been on a scale to completely alter the approach to refuse disposal.
Changes of packaging techniques and the concept of the throw-away container have
introduced large quantities of materials which are difficult to handle in the
traditional disposal processes. Thus, the need for considering totally new ap-
proaches to the problem is apparent.
In the present study, the problem has been approached by considering
refuse not so much as material to be gotten rid of, but rather as a resource.
As we look at the composition of refuse we find that well over half of the ma-
terial is organic in nature, and as such, is made up largely of carbon, hydrogen
-------
and oxygen in about the proportions found in cellulose. Thus, the possibility
exists of reforming the organic portion of solid wastes into useful and economi-
cally attractive materials in much the same way that natural gas or petroleum
may be reformed.
Consequently, studies have been made of the types of products which can
be obtained by reacting certain organic materials, commonly discarded at present
as solid wastes, with limited quantities of oxygen at highly elevated temperatures.
The purpose of such treatment is to reform the organic components of waste into
simpler organic compounds of economic interest. The materials studied in this
work included paper, sawdust, leaves, and "Orgro", a dried sewage treatment plant
sludge.
The principal efforts in this study have been directed towards developing
suitable laboratory equipment for reacting the materials, to collecting, separating
and identifying the products formed and to establishing relations between operating
conditions and the types of products obtained.
-------
Background
A. General
A review of the history of waste treatment reveals that various heat
processes have been applied with varying degrees of success to the decomposition
of the organic components of solid waste materials. In the majority of processes,
the prime objective has been to reduce the bulk of the material and to reform the
organics into relatively simple, stable compounds. In such cases , it has been
found expedient to use large quantities of air in processes usually categorized
as burning or incineration.
More recently, however, attempts have been made to heat solid wastes to
high temperatures in the absence of air. Such a process, usually called pyrolysis,
has as its prime objective, the decomposition of the organic material to relatively
simple compounds which are economically attractive.
A brief review of these processes, particularly as they apply to the
problem of recovery and utilization of components of solid waste, will be useful.
A short discussion of appropriate analytical techniques is also included.
B. Incineration
The use of equipment specifically designed to handle and burn refuse is
probably somewhat less than one hundred years old. It is reported that attempts
were made to burn municipal refuse in England in 1874, but that combustion was
unsatisfactory because of large quantities of wet garbage incorporated in the
refuse. In general, temperatures in the early incinerators were low, and the
heavy production and accompanying odors was a problem. The use of auxiliary fuel
was found to be necessary for satisfactory combustion.
Apparently, the first practical units in the United States were made at
various government installations. These were followed by municipal incinerators
in a few of the larger cities at about the turn of the century and fairly wide
acceptance followed soon thereafter.
-------
Early attempts were made to separate rubbish, garbage and ashes before
burning, and to utilize rubbish as a fuel for waste heat recovery. In general,
these early attempts did not prove practicable because of the lack of operating
4
dependability.
Many types of designs have evolved over the years. Recently, such im-
provements as continuous feed, mechanical stoking, facilities for fly ash removal,
air pollution control devices, and other automated features have greatly increased
2 15
the efficiency and lowered the operating costs of incineration. '
The disposal of sewage sludge by incineration has been attempted for quite
a few years. This method has met with limited success , due to problems encountered
with respect to smoke, odor, and weather conditions. Initial attempts to burn
these filter cakes with conventional stoker furnaces have been generally unsuccessful
because of clinker formation and high ash percentage by dry weight of solids in
the filter cakes. This material, at best, can be considered only a very low-grade
fuel. Special equipment has been designed to provide for improved incineration of
sewage sludge. Today, sewage plant incinerators generally use either the flash
2
drying incineration process or the multi-hearth process for the burning of sludge.
Attempts to utilize the heat of combustion of organic matter in solid
wastes for the generation of steam has been one way of seeking to reduce the cost
of disposal. Such methods have been employed in a number of instances in several
European countries. In some cases attempts have been made to recover salvageable
materials from incinerator residue, or to utilize the residue as a building
4.
material. The gaseous products evolved from incinerators are largely oxides of
carbon and water vapor. The sulfur content of refuse is low so that sulfur dioxide
is usually not a serious air pollutant. Similarly, nitrogen oxides production is
usually not a problem, because the temperatures are usually lower than that required.
-------
It has been reported that when the oxygen supply is deficient, additional products
are formed which are often considered to be nuisances. These include: carbon
monoxide, ammonia, benzene, aldehydes, ketones , phenols, organic acids and esters.
These compounds are valuable by themselves, and efforts have been made to separate
, ^, 15
and recover them.
C. Pyrolysis
Some work has been reported where organic material has been burned, or
pyrolyzed, in the absence of oxygen. Research performed by Dimitri, Jongedyk,
and Lewis who used destructive distillation along with fluidization for the break-
down of powdered hardwood has been described. The sawdust was heated while sus-
pended in a stream of hot non-oxidizing gas. A recovery system was used to separate
tar, acetic acid, methanol and other products from the outlet gas. The described
system operated between 250 C and 380 C. It was proposed that this system be com-
bined with a second stage operating at a higher temperature which would strip off
all residual volatile matter and leave a charcoal.
Morgan, Armstrong and Lewis followed up this work in 1953, and they at-
tempted to add several refinements to the system along with making a more complete
analysis of the products. As in the earlier work, an isothermal fluidized bed was
used. The sampling train consisted of a water-cooled condenser, a mist trap, and
two containers filled with activated charcoal, for the recovery of solvent vapors.
These vapors were recovered by desorption under the influence of heat and vacuum,
followed by condensation in a trap cooled with methyl-ethyl ketone and dry-ice with
a clean-up trap cooled by liquid nitrogen. The condensates were then tested by wet
Ofi
chemical analysis for the presence of acetic acid and methanol.
-------
Of more recent origin is the work performed by Hoffman on the pyrolysis
of solid municipal wastes. The material studied consisted of so-called "typical
San Diego solid wastes". As pyrolysis progressed, the tars and heavier fractions
were condensed immediately outside the reactor in a tar trap. The lighter frac-
tions were collected in receiving bottles submerged in dry-ice and acetone and
the gases trapped in a gas receiving balloon. The gases evolved were analyzed by
gas chromatography and were found to consist of carbon dioxide, carbon monoxide,
methane, ethylene and ethane. The volume proportions of these gases proved to be
dependent upon the temperature at which the reaction was run. The inert residue
20
proved to be sterile and appeared to have potential as a suitable fill material.
D. Gas Chromatography
A variety of gas chromatography techniques have proven valuable for analysis
of products from combustion processes. The gas chromatograph has been used for
both qualitative and quantitative analyses of mixtures of known and unknown compounds,
In 1960 Feldstein made a chromatographic analysis of incinerator effluents.
His work was confined to the analysis of low molecular weight hydrocarbons present
in the effluents. He employed a multiple column system and a method of trapping
which allowed the collection and identification of a range of products. The instru-
12 39
ment was calibrated by injection of known quantities of standard hydrocarbons. '
Gas chromatography work on a polluted atmosphere by Bellar and others in
1962 substantiated Feldstein's analysis. The trapping system described by
Feldstein had been used in earlier work by Brenner and Ettre with similar success
Q
using silica gel as a trapping column adsorbent.
Cvejanovich developed a chromatographic technique in.which three columns
in series were employed to separate C, through C_ paraffins and olefins as well as
-------
to give the quantitative determination of hydrogen, oxygen, carbon monoxide and
carbon dioxide.
A chromatography study of the Los Angeles atmosphere was conducted by
Altshuller and Bellar in 1963 using silica gel to separate ethane, ethylene ,
acetylene, propane, propylene, n-butane, isobutane, n-pentane, and isopentane,
and a carbowax column for benzene, toluene, ethylbenzene, and m, p, and o-xylene.
A flame ionization detector was used during this analysis instead of the thermal
conductivity detector because of the greater sensitivity that was needed due to
3
the low concentrations of the impurities in the sample.
The gas chromatograph has been used with some success for quantitative
work. Maher employed a technique of measuring relative peak areas before and
after treatment of samples to remove olefins , aromatics and normal paraffins. The
amounts by which the areas of peaks changed compared to the increase in the largest
component was taken as a measure of the component removed by this treatment. This
proved somewhat satisfactory for the analysis of several hydrocarbon mixtures,
yielding information in terms of hydrocarbon types , but suffered limitations asso-
23
ciated with the efficiency of the reagents used to remove the hydrocarbon groups.
Hoff and Feit have described a technique for carrying out syringe reactions
using various reagents to remove or change functional groups. This method provided
19
information on the types of compounds present in organic mixtures.
E. Infrared Spectroscopy
Infrared spectroscopy has proven to be extremely useful in the qualitative
42
and quantitative analysis of unknown compounds. Yokum, Hein, and Nelson were
successful in identifying methanol, acetaldehyde, and carbonyl sulfide in the
gaseous effluents from back yard incinerators. They employed evacuated bulbs to
-------
collect these gases and loaded their contents into an infrared gas cell. These
trapped gases were analyzed using an infrared spectrophotometer. They found that
by drawing the vapors off the liquid effluents they were able to identify acetone,
methanol, benzene, and an ester.
31
Sawicki and Hauser were successful in identifying carboxylic acids and
aldehydes in air-borne particulates through the use of high resolution infrared
spectroscopy. As a result of a detailed study of the behavior of known carboxylic
acids and aldehydes, it was possible to identify unknown mixtures of these compounds.
Similar studies were conducted on primary and secondary aliphatic amines by
38 V
Stewart. Quantitative analyses were performed by Brattain, et al. , but were
limited to mixtures of known compounds due to the necessity of pure component trans-
mission measurements.
A promising application to pollution studies has been the combined use
of gas chromatography (to separate and collect pure components of an unknown mixture)
18
and infrared spectroscopy (to analyze the collected components). Haslam, et al.
concerned themselves with the development of micro sampling apparatus and sample
o c oc
handling techniques for these instruments. Similar attempts by others " have
been made to solve these problems.
A strong advantage to the use of infrared spectroscopy for the identifica-
17
tion of complex organic materials has been demonstrated by Harms. He suggests
that the infrared spectra of the pyrolysis products of complex organic molecules
could be used as a means of identification. He demonstrates that the infrared
spectrum of the pyrolyzate can be used simply as a unique characterzing pattern,
without special regard for the actual chemical composition reflected in the spectrum.
-------
Theoretical Considerations
A. Partial Combustion
As is well known, most organic materials may be burned under conditions
of total combustion to yield such products as carbon dioxide, sulfur dioxide, oxides
of nitrogen and water vapor. However, if the reaction occurs in a deficiency of
oxygen, the combustion is incomplete and a wide variety of products are formed.
Some typical combustion reactions under both total and partial combus-
tion conditions are as follows :
Total combustion
Partial combustion
+ 0 - > 2CO
CO + H - ^ HCHO (formaldehyde)
CO + 2H0 - > CH0OH (methanol)
Z o
Thus , even the simplest of hydrocarbons may yield a variety of products
under conditions of partial combustion. As the complexity of the fuel increases,
so also does the variety of possible products increase. In the case of a material
such as paper, a major component of domestic refuse, the list of possible products
is considerable. In Fig. 1 is shown a representation of the cellulose molecule,
which is the chief constituent of paper. As can be seen, this molecule is al-
ready partially oxygenated and it would be expected that a great variety of degrada
tion products should be possible .
Because of the increased complexity in the composition of leaves and saw-
dust, it would be expected that the variety of compounds resulting from the partial
combustion of these materials would be correspondingly greater. Likewise, dried
-------
'CM
O
w
O
w
J
O
s
P-i
O
I
O
J
w
O
10
-------
sewage sludge contains a myriad of organic compounds, and heat treatment might
be expected to yield a product mixture of tremendous complexity. Through
partial combustion somewhat simpler organic compounds are potentially possible.
For instance, hydrocarbons that are present may oxidize to mixtures of alcohols
which in turn might oxidize to aldehydes and ketones. Further oxidation would
25 2 8 34-
result in the corresponding organic acids. ' '
B. Fluidization
Chemical reactions between solid and gaseous materials may often be carried
out most conveniently in a fluidized bed. The process of fluidization involves
bringing a fluid into intimate contact with a bed of finely divided solids at a
velocity sufficient to keep the solids in suspension. At fluidization, the rate
of fluid flow through the bed will be such that the pressure across the bed, due
to frictional drag on the particles, will equal the effective weight of the bed.
When a fluidized bed is examined at this point, it is found to be in a turbulent
state. It bubbles and boils and the solids move about in an erratic manner.
A fluidized bed has three characteristics that make it desirable for use
in a reactor system. The first is that the violent agitation and rapid redistribu-
tion of solids in a fluidized bed results in the elimination of hot spots in the
reaction zone. This allows for close control over reaction temperatures. Secondly,
the large gas-solid contact area makes a fluidized bed reactor an efficient system
for catalytic and non-catalytic reactions. Lastly, the solids can be handled con-
veniently by pneumatic systems when they are in a fluidized state. If the solids
have to be introduced and removed from the reaction bed, this can often be done
conveniently since the solids can be pumped into and out of the system like a fluid.
There are several limitations of a fluidized bed reactor, primarily as-
sociated with non-ideal flow. The ramification of non-ideal flow is that reactor
11
-------
size increases much more rapidly than for plug flow reactions for high conversions.
In addition, the extent of deviation from plug flow must be determined by experi-
22
ment to adequately design these units.
C. Gas Chromatography
The analysis of organic mixtures can be performed in a gas chromatograph.
The sample, either liquid or vapor, is injected into a vaporizer which generates a
"plug" of vapor which is then swept into a packed column by an inert carrier gas.
Separation of the components in the column results from the differences in the
multiple forces by which the column materials tend to retain each of the components.
The nature of the retention may be absorption, solubility, chemical bonding, polar-
ity, or molecular filtration, but in any event, the column retains some compounds
longer than others. Upon emerging from the column, the gases immediately enter a
detector. In the thermal conductivity detector, the difference in thermal conduc-
tivity of the product stream and a reference stream is used to generate a signal
which is amplified and used to cause a pen deflection in a strip chart recorder.
The selection of a suitable packing material which will effectively
separate the unknown sample components is critical to the successful use of gas
chromatography. Besides being dependent on the material used as a column packing,
the separation achieved is strongly dependent in the manner in which a column is
packed.
Gas chromatography, while being an excellent analytical tool, suffers
from two limitations in its application to pollution studies. The first limitation
is that if components of a wide boiling range or different characteristics are
present in the sample, the analysis often cannot be carried out in one run because
it is very hard to find "general" chromatographic columns. The second limitation
is that although gas chromatography shows an extremely high sensitivity, it is not
12
-------
high enough for ranges of compounds in concentrations such as are commonly of
Q T 1 0*7 QQ
interest in air pollution studies. ' ' ' While these two shortcomings may
be formidable in some cases, they can often be overcome. The problem of a wide
variety of compounds can often be overcome by temperature programing. In this
technique, the column is kept at a low temperature until all low boilers have been
eluted. Then the column temperature is linearly programed which accelerates the
elution of the higher boiling components. The sensitivity problem can be solved
by collecting the samples using a concentration technique.
The gas chromatograph can often be used for quantitative as well as quali-
tative work. The length of time that a component is in a column, called its re-
tention time, is specific for any particular compound and solid phase under the
21
same instrument conditions. This time may, therefore, be used for qualitative
identification of the component. A homologous series of organic compounds will
yield a straight line or smooth curve when some property of the compounds is
plotted against retention time on semi-logarithmic paper. The retention time on
the log scale may be plotted against carbon number, molecular weight or boiling
point. With an unknown sample, this plot can be used for tentative identification.
Confirmation of a tentative identification from retention times with a particular
column packing may be accomplished by duplicating the separation on another column
using a different packing. Confirmation may also be made by adding a small amount
of the suspected component to the mixture and rechromatographing. If the peak of
the component is increased, then more positive identification has been made. This
is known as the peak attentuation technique. Complications may arise when compo-
21 37
nents having similar properties are present. Then peak interference may occur. '
Quantitatively, the height of the elution peak or the area under the peak
is approximately proportional to the amount of the component that produced it.
The peak height is influenced by the temperature of the column or any other factor
13
-------
that affects the activity of the adsorbent. Since the area under an elution peak
is practically independent of the sample size, it is generally a more satisfactory
quantitative measurement than peak height. The area can be computed by any of a
number of ways.
D. Infrared Spectroscopy
Infrared spectra originate primarily from the vibrational stretching and
bending modes within molecules. Most organic and inorganic compounds have charac-
teristic absorption wavelengths in the infrared region. A scan of the infrared
region presented as wavelength versus absorbance (or transmittance) is called an
infrared spectrum. This spectrum is a powerful tool for the study of molecular
structure. Both qualitative and quantitative measurements are possible with an
infrared spectrophotometer. In addition, a match between the infrared spectrum
of an unknown material and the spectrum of a reference sample is almost without
M-l
equal as an empirical proof of identity. It is quite unlikely that any two com-
pounds have the same response in the infrared region.
Although an infrared spectrum is a characteristic of a specific compound,
certain functional groups of atoms give rise to absorption bands at or near the
same frequency regardless of the structure of the remainder of the molecule. It
is the presence of these characteristic group frequencies that enables one to
examine an unknown compound and pick out its key functional groups. A table of the
more commonly used functional group absorption frequencies appears in Table 1.
These frequencies were taken from reference (35) but appear in almost any text
covering infrared spectroscopy.
To realize the full potential of infrared spectroscopy in qualitative
analysis, a large collection of properly indexed reference spectra of known com-
pounds must be readily available. Even though there are enormous amounts of
14
-------
spectra available, the problem is that present indexing and retrieval systems
suffer from shortcomings, and the spectra that are available vary widely in
quality and the manner in which they are presented.
15
-------
TABLE 1
TABLE OF INFRARED ABSORPTION FREQUENCIES
FOR LIQUID SAMPLES
Type of Compound
General
Characteristic Group
Vibrations
C-H Stretching
C-H Bending
C=C Stretching
C=C Stretching
Frequency (Microns)
3.3
6.8
6.0
4.4
3.5
7.3
6.25
4.8
Alcohols and
Phenols
0-H Stretching
0-H Bending
C-0 Stretching
2.7 - 3.1
7.0 - 7.5
7.9 -10.0
Ketones
C=0 Stretching
5.35- 6.5
Aldehydes
C-0 Stretching
C-H Stretching
5.7
3.5
6.0
3.7
Carboxylic
Acids
0-H Stretching
C-0 Stretching
2.8
5.6
2.9
5.9
Nitriles
C=N Stretching
4.4 - 4.5
*These frequencies have been taken from Reference (35).
16
-------
Materials and Apparatus
A. Materials
1. Analytical Standards
The analytical standards used in the peak attenuation technique were
all analytical reagent grade compounds. They were used in conjunction with the
research gas chromatograph in an attempt to analyse samples for their presence.
These standards included:
Alcohols - methanol, ethanol, propanol, isopropanol,
butanol, isobutanol, pentanol, and hexanol
Acids - formic, acetic, propionic, butyric, and
valeric acids
Aldehydes - formaldehyde, acetaldehyde, and
propionaldehyde
Ketones - acetone and butanone
Paraffins - methane, ethane, propane, and butane
Miscellaneous - carbon dioxide, carbon monoxide,
nitrogen, benzene, toluene, xylene , carbon
tetrachloride, and carbon disulfide
The gases listed in these last two categories, with the exception of
nitrogen, were obtained from the Matheson Company in lecture bottles. The nitrogen
gas was purchased from the Linde Division of Union Carbide. The liquid standards
were obtained from the Fisher Scientific Company.
2. Carrier Gases
Helium, nitrogen, and air were used as carrier gases and were obtained
from the Linde Division of Union Carbide. The nitrogen was used as a carrier gas
in the preparative gas chromatograph and in the reactor during the pyrolysis
17
-------
reactions. The helium was used in the research gas chromatograph, and the air
was used in the reactor during the partial oxidation reactions.
3. Indicating Papers
Accutint was used to determine the pH of several of the liquid samples
that were collected. The entire range of this indicating paper proved to be use-
ful. This paper is a product of Anachemia Chemicals, Ltd. At a later stage of
the research Alkacid test paper by Fisher Scientific was employed.
4. Reagents for the Analysis of Formaldehyde and Methanpl
The two reagents required in the analysis of formaldehyde were chromo-
tropic acid and concentrated sulfuric acid. The former was purchased from
Matheson Coleman and Bell, and the latter acquired from the Fisher Scientific
Company.
The analysis of methanol required the use of potassium permanganate,
sodium bisulfite, chromotropic acid, sulfuric acid, and pyrogallol. These were
all obtained from Matheson Coleman and Bell with the exception of the sulfuric
acid.
5. Reagents for Syringe Reactions
The reagents used for the syringe reactions included the following:
concentrated sulfuric acid, hydrogen iodide, sodium borohydride, potassium
permanganate, hydroxylamine-hydrochloride, acetic anhydride, sodium hydroxide and
hydrochloric acid. The reagents were obtained from Matheson Coleman and Bell
with the exception of the acids and sodium hydroxide which were obtained from
Fisher.
B. Apparatus
A schematic diagram of the apparatus used is shown in Fig. 2.
18
-------
o
I-H
H
O
CO
s
w
H
CO
>-"
CO
o
H
O
19
-------
1. Reactor System Equipment
Three separate reactors were used at various times for this work. They
were generally similar in overall dimensions but differed in materials (brass ,
iron and ceramic) and in some details. Diagrams of the three reactors are
shown in Figs.3,4 and 5.
Each reactor was approximately two inches in diameter and 24 inches in
length. A metal flange at the bottom was used to join the reactors to a 12 inch
entrance section. A 100 mesh brass screen was supported between the flanges
which in turn supported the bed of material in the reactor. A 1/4 inch side
arm was connected to the reactor near the top to allow removal of product gases.
A similar arm was provided on the iron reactor for removal of solid residue.
The metal reactors had black iron reducers at top and bottom. The bottom reducer
connected to the inlet gas line. In the brass reactor, the top reducer allowed
the introduction of a thermocouple well the length of the reactor. In the iron
and the ceramic reactor, the top of the reactor was connected to an injection
feed system. This system consisted of two 1 1/2 inch brass gate valves separated
by a 10 inch section of 2 inch iron pipe. A 1 1/2 inch black iron nipple con-
nected the lower valve to the reducer at the top of the reactor, and an identical
nipple was connected to the upper valve. A rubber stopper fitted tightly with
an aluminum plunger was forced to fit tightly into the upper nipple. This enabled
feed to be injected into the reaction zone while the reactor was at operating
temperature.
Heat was supplied to the reactors by heating tape. A model BIH-62 1/2
heating tape manufactured by the Briskeat Company was found to be satisfactory and
capable of heating the reactor to 2000 F. Current was controlled by a Powerstat.
The reactors were insulated by a high temperature pipe lagging to prevent
undue heat losses.
20
-------
Exit Port
24
10"
III
I''
FIGURE 3
BRASS REACTOR
2 x 7/8" Reducer
Insulation
Combustion Chamber
Thermocouple Well
Bolted Flanges
2 x 7/8" Reducer
21
-------
2 x iJg" Reducer
Exit Port
Gate Valve
Feed Reservoir
1%" Gate Valve
Combustion Chamber
Thermocouple Location
Insulation
Exit Port
Bolted Flanges
2 x 7/8" Reducer
FIGURE i*
BLACK STEEL REACTOR
22
-------
23$ x 7/8" Reducer
Exit Port
FIGURE 5
CERAMIC REACTOR
23
IV Gate Valve
Feed Reservoir
IV Gate Valve
Insulation
Combustion Chamber
Thermocouple Location
Exit Port
Sand
Glass Wool
2% x 7/8" Reducer
-------
Reaction temperatures were measured by chromel-alumel thermocouples
made and calibrated in the laboratory. These were used in conjunction with an
L and N Potentiometer, Model 8691.
Air and nitrogen streams were measured by two model 2-1355 Sho-Rate
rotameters manufactured by Brooks Instrument Division of Emerson Electric Company.
They were calibrated by a Precision Scientific wet test gas meter.
As shown in Fig. 2, a Lucite pilot plant was connected in parallel with
the reactor. This was periodically used as an indicator of the type of fluidiza-
tion obtained under a particular set of operating conditions.
The solids fed to the reactor were fluidized by a mixture of nitrogen
and air. Nitrogen was fed from a cylinder and air was supplied from an oil-less
air compressor. In later runs cylinder air was used. The gases were blended to
provide the desired oxygen concentration and feed to the bottom of the reactor.
2. Sample Collection Equipment
The product gas stream leaving the top of the reactor was passed to a
water-cooled Liebig condenser. Here the high boilers were condensed and collected
in a flask connected to the condenser. From here the product stream passed
through a series of glass traps maintained successively at lower temperatures.
Each trap was immersed in a cold bath held in a Dewar thermos flask. Icewater,
dry ice, and dry ice-acetone mixtures were used in the cooled baths. In some runs,
the gases from the traps were passed through a column packed with silica gel made
of 2 feet of 1/4 inch copper tubing. Gases were adsorbed in this unit for subse-
quent analysis.
3. Analytical Equipment
Much of the analytical work was done on a Model 5751A research gas
chromatograph manufactured by Hewlett Packard. This instrument was equipped with
24
-------
both dual flame ionization and thermal conductivity detectors. A Model 7128A
strip chart recorder was used in conjunction with this unit. This recorder had
two channels and one disc integrator.
On some tests a Hewlett Packard, Model 775, Prepmaster gas chromatograph
was used to fractionate the liquid samples. This unit had a built-in strip chart
recorder, and the capability of completely automatic operation.
Several column packings were used for both the analytical and preparative
instruments. Most of the work was done with Polypack-1, manufactured by Hewlett
Packard, and Porapak Q or Porapak N, both manufactured by Waters Associates.
A soapfilm flowmeter having a 10 ml scale was used in conjunction with
a stop watch to determine the volumetric flow rate of carrier gas through the
analytical columns in the research gas chromatograph. Carrier gas flow in the
preparative unit was measured using the built-in rotameters.
A Perkin Elmer, Model 137, sodium chloride spectrophotometer was used
for all of the infrared analyses, including gas, liquid and solid analyses. The
sample holders included a 10 cm gas cell, several sodium chloride liquid cells,
and a potassium bromide pellet holder. The pellets were formed by use of an
evacuable potassium bromide die.
A collection system was fashioned for use with the preparative gas chro-
matograph in order to eliminate excess sample transferrals between collection in
gas chromatograph and final analysis in the infrared spectrophotometer. This
system consisted of special collection cells called Extrocells , Model 195918;
plugs for these cells to prevent evaporation, Model 195919; polystyrene packs to
hold the cells while centrifuging the sample into the small slit in the bottom of
these cells, Model 195924; a cell mount to fit the sodium chloride spectrophotom-
eter, Model 1954-79; and a cell holder that slides the Extrocell into the cell
25
-------
mount, Model 195929. This equipment was manufactured by the Scientific Instru-
ments Division of Beckman Instruments.
26
-------
Experimental Procedure
During the course of this study three classes of materials were examined
as potential sources of useful products. In chronological order in which they
were studied, they were 1) finely divided paper, 2) Orgro, dried sewage sludge
from the Schenectady N. Y. sewage treatment plant, and 3) finely divided dried
leaves. Since these materials differed somewhat in their properties, somewhat
different techniques of handling and of analysis were required. Consequently, the
procedures used for each material are considered in order.
A. Paper
1. Reactor Operation
Most of the runs on paper were made using printed newspaper. A few runs
were made with ink-free unused newspaper but no significant differences were noted
and no distinction between runs appeared to be justified.
A considerable effort was devoted to finding a suitable way of preparing
the paper so that it could be fluidik,ed. It was found that because of its low
density, rather large particle sizes were required for good fluidizations. Particles
in the size range of 5 to 10 millimeters were generally satisfactory. Smaller sizes
tended to clump together and fluidized erratically.
Unheated runs were made in the transparent pilot column to determine opti-
mum particle size, and gas velocity as a function of reactor batch size.
Several methods were used to prepare the paper sample. A Waring blender
was found to reduce the paper to a convenient size and was used in some runs. A
conventional kitchen meat grinder was also found to produce satisfactory samples.
Other methods of preparation were found to be less effective. The sample size
varied between 15 and M-0 grams. In most runs the sample was taken as 15 grams.
During later runs larger samples were taken to see if sample size had any effect on
results.
27
-------
Once the sample size was decided upon and the sample prepared, it was
loaded into the cold reactor by removing the reactor cap and dropping the sample
in by hand. After the reactor was loaded, it was connected into the line and
sealed against leaks. Silver Goop, a sealant distributed by the Crawford Fitting
Company, was found to be a reasonably good sealant under the conditions encountered.
The Dewar flasks, in which the liquid and gas traps were immersed, were
filled with the appropriate cooling medium and allowed to come to equilibrium.
The flows of air and nitrogen were then established to give both the desired total
flow for good fluidization and the desired air to nitrogen ratio. The ratios
studied varied from total air or a complete combustion run, to total nitrogen or a
pyrolysis run.
Once the bed was fluidizing properly and base measurements were recorded,
the electrical heating coils in the reactor were turned on. The temperature in
the reactor was measured at intervals with the chromel-alumel located in a well at
the centerline of the reactor.
A reaction run usually lasted from 2.5 to 3.0 hours. As the run progressed
the temperature rose fairly steadily to about 400 F. At this point an evolution
of smoke was usually noted and the temperature rose rapidly. The maximum tempera-
ture attained depended upon how much heat was supplied to the heating coils , the
size of sample and the gas flow rate, but usually was in the range of 860 F to
1100°F.
After a run was completed, the heat to the reactor was discontinued, and
after a brief cooling period the gas flows were stopped. The cooling baths were
removed from the liquid traps and the traps were allowed to come to room tempera-
ture. While remaining connected to the system, nitrogen was passed through the
trap to purge any low boilers to the gas trap which remained at the low temperature.
The remaining liquid sample was then transferred to a labeled test tube, the pH
determined, and the tube stored in a dry-ice chest for subsequent analysis. The
28
-------
gas trap was sealed and removed from the system and likewise stored in the dry-
ice chest. After cooling, the reaction residue was removed and weighed.
2. Gas Chromatograph Operation
Basic to the successful operation of the chromatograph was the selection
of suitable columns for the constituents. After considerable investigation,
columns with dimensions of 4 ft. by 1/4 in. diameter, packed with Porapak Q, were
selected.
For all runs , helium was used as a carrier gas since it gave fewer peak
inversions. As a general procedure, the gas flow was established, once the columns
were in place, and the system checked for leaks. The flows through the reference
column and the operating column were balanced by means of a soap film flowmeter and
maintained at 50 cc/min.
A number of standardization runs were made by sending various liquids and
gaseous samples of known materials through the chromatograph at different tempera-
tures. This provided a relationship between retention time and molecular weight
for these materials. All of the liquid samples were of analytical reagent grade.
The liquid samples were injected into the chromatograph with a 10 microliter
syringe and the retention times were measured at 150 C, 175 C and 200 C. Results
are shown in a later section.
Gaseous standards were also processed in a similar manner, but injected
by a special technique developed for gases, and described later.. An oven tempera-
ture of 100 C was used during these runs, since the gases tended to move through
the column at a faster rate than the liquid samples. Results are given in a later
section.
The analysis of liquid samples from reaction runs were made as soon after
being obtained as possible. In preliminary analyses, oven temperature was maintained
at 200 C and several runs made until suitably reproducible chromatograms were ob-
29
-------
tained. It became apparent that many of the same components were being obtained
in virtually all of the reaction runs, although in different proportions. Conse-
quently , it was decided to make programmed runs in subsequent work to achieve
better separation of components. In these runs the temperature was raised at a
constant rate of 10 C/min from a temperature of 150°C to 200°C during the course
of the analysis. Based upon the results from previously determined standards, it
was possible to make postulations regarding the identity of several of the compo-
nents. To substantiate these postulations, a peak attenuation technique was tried,
in which a small amount of the suspected component was added to the original sample
and its effect upon the resulting chromatogram observed. This method was tried
with numerous pure materials as indicated in the results.
Attempts were made to apply infrared spectroscopy to the identification
of unknowns in the liquid samples. Results, however, were inconclusive with the
samples tested in this part of the study.
A limited number of analyses were made on the gases which were not con-
densed in the liquid traps. The sampling apparatus shown in Fig. 6 was used to
collect the off-gases and to allow their introduction to the gas chromatograph.
The gases were adsorbed in a silica gel column cooled in a dry ice-acetone bath
as shown. When an analysis was desired, the carrier gas used in the chromatograph
was diverted to flow through the column which was heated in boiling water. Thus,
the adsorbed gases were carried into the instrument and analysed. Most runs were
carried out by running the chromatograph at 100 C for 20 minutes, followed by a
program in which the temperature was raised to 200 C at a rate of 10 C/min. The
helium flow was directed through the trap for 1 minute and then by-passed the trap
for the remainder of the run. Standard gas samples were introduced into the gas
stream through a septum just ahead of the silica-gel trap.
30
-------
s
H ft
fc ?
k a) &
m H M
o -i- ft o o
6 -p -p
e m m
3 co 6
-H O
o
0)
c
w
w
IT)
O<
0) w
rH
(1) rH
0) rcJ
iz; >
a
§
Pk
PH
a
<
CO
w
nS
0) 4i
H o.
(0 6 W)
O O O
fc -H
H
H
0)
O
fn
^2
O
31
-------
B. Dried Sewage Sludge (Orgro)
1. Reactor Operation
In early attempts to fluidize Orgro, it was found that considerable
channeling took place because of the heterogeneous nature of the dried sewage
sludge. Good fluidization was obtained by screening out fines smaller than 60-
mesh, and mixing the remainder with about an equal volume of glass beads or sand
of approximately the same mesh size. To prevent the loss of a small amount of
fines present in the solids mixture, a glass wool plug was inserted in the exit
gas port.
It was found that a 2 inch layer of coarse sand supported on a 100 mesh
screen at the bottom of the bed acted as an excellent gas distributor to the bed,
and also acted as a gas preheater because of its high heat capacity. This sand
did not fluidize at gas velocities required for the runs.
In a reaction run, the solid mixture was introduced into the reactor to
give a static reaction mixture bed depth of approximately 1 foot. Upon adjust-
ment of the gas flows to the desired values, the bed usually expanded by an addi-
tional 2 inches. This height, combined with the sand height above the brass
screen brought the total height of the bed to within a few inches of the exit port.
The desired carrier gas flow, and the desired ratio of air to nitrogen,
were made by making appropriate adjustments to the gas rotameter settings. Air
was supplied from a compressor in the early runs, but the supply was later replaced
by cylinder air because of a smoother flow. Temperatures in the range of 900 F
to 1900°F were studied.
In the early runs, using the brass reactor, the procedure followed was
exactly as described for the batch runs using paper. Later, when the black iron
reactor and the ceramic reactor were used, attempts were made to develop a procedure
-------
for making continuous runs. A continuous feed of solid mixture was pumped to
the reactor, and solid residue was continuously removed. This worked well with
a cold reactor, but with a hot reactor, the inlet and exit lines quickly plugged.
Consequently, a procedure for semi-continuous runs was evolved. Here, a batch
of feed was introduced into the reactor from a feed reservoir on top of the
reactor by opening the valves on top of the reservoir. A plunger aided in in-
troducing the feed quickly to the gas stream flowing in the hot reactor. This
procedure was adopted for all runs with the iron reactor and the ceramic reactor.
Gases evolved from the reactor were sent through a series of traps held
at successively lower temperatures and liquid fractions were collected in each
trap. After a run was completed, the gas flows were discontinued and the reactor
cooled. The traps were sealed and the contents held at a low temperature for
subsequent analysis.
2. Analytical Run
The samples collected in the traps held in the dry ice-acetone baths
were found to contain absorbed gases. It was found convenient to analyse these
gases by use of an infrared spectrophotometer. To load the gas cell with the
absorbed gas, an apparatus shown in Fig. 7 was used. The infrared cell and
connecting system was evacuated by a vacuum pump. A collection trap immersed in
its cold bath and containing the desired sample was connected to the system, and
the pump then isolated from the system. The valve at the top of the trap was
opened slowly to allow some of the gases to enter the system. When the pressure,
as indicated by the gauge, reached 1/2 atmosphere absolute, the valves on the gas
cell and the trap were closed. The system was then vented to atmospheric pres-
sure, the infrared cell disconnected, and an infrared spectrum of the gas was
determined. The cell was then reattached to the system, the vent closed and the
system pumped down again so that the cell could be reloaded. The trap for this
33
-------An error occurred while trying to OCR this image.
-------
portion of the run was placed in ice water. The above procedure was followed
except that the run was discontinued when the pressure reached 1/3 of an atmos-
phere. The cell was again removed for another spectrophotometer determination.
This procedure was repeated once more with the trap at room temperature.
Liquid samples were collected in each of the traps for all runs. The
samples in the dry ice-acetone cooled traps usually separated into an aqueous
layer and an organic layer. Many analytical methods were attempted with each
sample, but most proved unsatisfactory because of the complexity of the mixtures.
Mass spectroscopy was attempted but merely resulted in a spectrum of almost every
conceivable mass number. Wet chemical methods were attempted but indicated that
there were interferences present. The most satisfactory technique was the use of
gas chromatography in conjunction with infrared spectroscopy.
The aqueous samples were analyzed qualitatively using a research gas
chromatograph. The resultant chromatogram showed one large peak for water, fol-
lowed by a large number of other compounds of relatively low concentration. When
the organic layer was run on this chromatograph under the same conditions, it was
indicated that the same compounds present in the aqueous layer were also present
in the organic layer. Since this organic sample was too complex to analyse using
infrared spectroscopy without further refinement, fractions were collected using
a preparative gas chromatograph.
The collection traps furnished with the preparative unit were too large
to be used with the collected samples. A special trap was devised, as shown in
Fig. 8, which proved quite satisfactory and allowed collection efficiencies of
about 90%. The technique used to separate and collect components was basically
the following: first, approximately 3 microliters of the unknown mixture were
injected into the analytical side injection port to see what type of separation
35
-------
Vapor Entrance
Connection to
Manifold
Pyrex Glass Tube
Glass Wool
UJ
Vapor Exit
Silicone Elastomer
Septum Material
1/16" Outside Diameter
Stainless Steel Tubing
Electrocell or Test Tube
FIGURE 8
TYPICAL PREPARATIVE GAS CHROMATOGRAPH COLLECTION TRAP
36
-------
could be achieved. Then, the oven temperature was programmed at several rates
until a suitable separation was found. When the right conditions were decided
upon, the preparative side of the unit was adjusted to duplicate these conditions.
It was found that the most efficient operating conditions were an injection port
temperature of 210 C, a column oven programmed from 110 C to 240 C at 2.5 C/min.
and a detector temperature of about 250 C. The column oven programming rate was
delayed 4- minutes after injection of the sample to further increase separation.
The collection traps were attached to the manifold of the chromatograph
and placed in dry ice-acetone baths. Four of these traps were employed to collect
pure components, and a fifth was used to collect the remaining mixture. The col-
lection valve was operated manually in such a way as to collect the same components
in each trap on successive sample injections. Only the four components which ap-
peared to be in the largest concentrations were collected in these first four
traps. Two injections of one ml. each were thought necessary, and by subsequent
inspection of the collection traps, this was verified. The collected fractions
were then rechromatographed to check for purity. This was accomplished by ex-
tracting approximately one microliter of a collected fraction and injecting it
into the analytical side of the chromatograph. If the resulting chromatogram
proved to be free of impurities , the sample was then put in an infrared liquid
cell and run on the infrared spectrophotometer.
In addition to the technique described above for component identification,
the peak attenuation method previously described was also used.
C. Dried Leaves
1. Reactor Operation
All of the work on this material was carried out in the black iron
reactor under essentially isothermal conditions. The operational procedure was
37
-------
similar to that employed during the Orgro runs. The feed mixture was prepared
by grinding leaves in a Waring blender and blending the finely divided solids
with an equal volume of glass beads. This mix was introduced into a hot reactor
from the feed reservoir, and the gas flow adjusted to give the desired fluidiza-
tion. Runs were made at 500°F and 1500°F. Evolved gases were collected in
series of cold traps, described for the Orgro runs, and held at trapping tem-
peratures for analysis.
2. Analytical Runs
The research chromatograph was used extensively in this phase of the
work. Helium was used as the carrier gas at a rate of 60 cc/min. The column
packing was Porapak Q. Liquid samples were programmed at 2 C/min and gas
samples were programmed at 4 C/min.
19
A technique of syringe reactions, described by Hoff and Feit was used
as an aid in product identification. This technique involved conducting reac-
tions between products and certain reagents in a syringe, and then observing
the chromatogram of the resulting reaction product mix. By observing changes
in the chromatograms of the original samples one may draw conclusions as to com-
pounds or functional groups that are present. In making a run, the sample in the
trap was allowed to warm at room temperature for 30 minutes. This allowed a
saturated vapor above the liquid and insured uniformity in the samples. An eye-
dropper bulb acted as a septum on the trap. Vapor samples were withdrawn by
inserting the needle of a syringe through the septum. The reagents for the re-
action had previously been distributed on the walls of the syringe. After drawing
a sample into a prepared syringe, it was set aside for 3 minutes to allow reaction
to occur. The sample was then injected into the chromatograph for analysis.
38
-------
The reagents used and the conditions required for the reaction are given in
Table 2. To supplement the data obtained by the syringe reaction technique,
peak attenuation was also used to aid in component identification.
39
-------An error occurred while trying to OCR this image.
-------
m
o
cu
w
o
ft
ft
cu
id p
g C
O cu
H 6
13 CU
id ft
ra
r-H
O
e
£
-p
H
ft
P
H
fi
CO
0)
O
O
eu
cu
o
2:
co
rH
o
a
o
g
LK
co
ft
0)
ra
cu
w
0)
o
o
ft
p-1
o
o
bO
G
H
O
nd
o
ft
ft
«
ra
ft
cu
-p
ra
cu
CO
cu
N ra
>,H
H o
O X!
ft 0
id o
>,H
Sc id
CO
T3
G
3
0
It*
O
O
13
CU
P
rd
ft
3
p
rd
ra
G
co
cu
O
e
cu
«
CO
Td
C
O
ft
B
O
O
r-H
!>)
fi
O
r\
p
rd
o
bO
G
H
O
rj
13
O
ft
ft
CO
cu
G
H
B
id
ra
cu
O
B
«
CN
Cl)
o
ra
cu
CO
13
O
ft
E
O
O
CU
H
""^J
H
O
CO
ft
cu
Id
^
w
CU
w
rd
cu
ft
o
cu
o
1
cu
ft
cu
B
o
IS]
o
ra
ra
cu
0
X
cu
ra
-p
rd
B
rH
e
H
H
H
CO
CU
rH
B
O
N
O
ra
0)
0
3
13
ra
fi
0
ft
B
O
0
0
H
13
H
O
ra
cu
-P
rd
fi
H
B
H
H
W
0)
B
O
CU
G
O
CU
I
fi
o
o
0) -rH
ft B
w
O cu
ft B
X -rH
H H
CN
CU
cu B
bO
B '
rH CU
ft N
00 CO
P p
G G
3 cu
O M
0 id
P4
0
B ra
O -p
H G
id bO
ft id
id cu
ft PH
cu
ft
(X,
LO
cu
ft
cu
ft
ft
>)
H
X!
CO
cu
ft
0)
is1
-p
G
cu
bO
rd
CD
oi
ft
-p
H
C
B
H
13
O
CO
o
cu
ft
p
rH
0
13
H
o
O
O id
G fi
co -H rd
G CNft
3 O cu
O S, -P
c3 2; £ O
CO
H bfl H CM
id e S
3 LO
0* 0 Zj
,
cu
H
ft
13
rj
G
rd
O
CO
CM
x --^
id
CO O
ft G
o o
ft O
T3 ^
o
H
G
Z£
O
id
2;
M
LO
CN
B
rH
id
o
CO
ft
cu
p
rd
[5
H
6
0
LO
CU
id
rH
X
0
ft
id
Xi
G
cu
bO
x"1
O
fi
H
CU
a
0
SI
o
0)
B
o
0
LO
H
O
nc
rH
0
LO
CN
C
CU
W)
o
ft
id
x
o
LO
C
ft
/-^ cu
13 -P
o id
B 15
O
O H
>-' B
cu
13
H
ft
O
rH
XI
0
ft
cu
p
rd
B
rH
ft
CNCU
0 ii
co rd
< Z
cu
p
rs1
H CU
13 CO
O ft
W <
CO
O ft
o cu
r3 id
Si Z
Si H bOrH
bO LO
o o
LO LO CN LO
CU
-p
rd
o
$t
H rd
13 o
O -rH
41
-------
-------
Results
The composition of paper and leaves used as feed materials to the
reactor in these studies is shown in Table 3. No analyses for the dried sewage
sludge were given because of extreme variability of samples.
Table 3
Feed Materials Composition
Percentages
PaperLeaves
51.21
5.85
41.14
0.04
0.12
1.57
55.72
5.37
28.72
4.20
0.11
5.88
Moisture 4.07
(as received)
Ultimate Analysis (dry)
C
H2
°2
N2
S
Ash
Material balances on the system invariably gave low product yields be-
cause of the small size of the initial charge and leaks which were inevitable to
reactor operation.
A. Reaction Runs Using Paper
The products obtained from the partial combustion of paper were classified
into three groups in accordance with the trap in which the products were found.
High boiling compounds having a viscous, tarry appearance were obtained in the
first two traps, cooled respectively by water at room temperature, and by ice-
water. These traps did an excellent job of removing the tars and prevented them
43
-------
from contaminating the other traps in the system. In the third trap a liquid
fraction was obtained having a color varying from a pale straw color to a rather
intense yellow. This trap was immersed in a dry ice-acetone bath and was de-
signed to accomplish the removal of any water vapor and other high boiling prod-
ucts formed from the reaction. Low boiling products were collected in the silica
gel trap at the end of the collection train.
The small amount of high boilers which was obtained in most runs , and
the high viscosity of this fraction made analysis most difficult. Attempts to
determine the components in this group were made but no definitive results were
obtained.
Principal efforts were directed towards identifying products present in
the liquid fraction. Gas chromatography proved to be the most effective method
available. Porapak Q was found to be a good column packing for separation of
components. While the combustion reaction was carried out under varying propor-
tions of air and nitrogen, the same compounds appeared to be formed in each case
but in differing proportions. A summary of the operating conditions investigated
for various runs, together with the sample size and amount of residue obtained,
is given in Table 4.
As an aid in identifying the product materials , the retention times for
a number of pure materials, thought possibly to be present, were determined.
These times are given in Table 5. By matching retention times of components with
those of known samples, a number of products have been identified with a reasonable
degree of certainty. These products included water, formaldehyde, acetic acid and
formic acid. These tentative identifications were substantiated by adding small
amounts of the suspected pure components to the original samples and observing
the changes produced in their chromatograms. These results are shown quite graph-
ically in Figures 9-14. The presence of acidic compounds was verified by the low
44
-------
20
16
12
o
Chromatograph Run
Flow = 50 cc/min
Temp.
Packing - Porapak Q
= 150-200°C @ 10°C/min
Sample #8
a
rH
0)
8
0
4 6
Retention Time (min)
8
10
45
FIGURE
-------
20
16
12
Chromatograph Run
Flow =50 cc/min
Temp. = 150-200°C @ 10°C/min
Packing - Porapak Q
Sample #9
g
a) 8
cu
0
4 6
Retention Time (min)
FIGURE 10
46
8
10
-------
20
18
Chromatograph Run
Flow = 50 cc/min
Temp. = 150-200°C @ 10°C/min
Packing - Porapak Q
Sample
12
o
bO
H
01
JS
0)
o.
8
4 6
Retention Time (min)
8
10
FIGURE 11
47
-------
20 r
16
Chromatograph Run
Flow =50
Temp.
Packing - Porapak
cc/min
= 150-200°C @
10 C/min
12
Sample #8 + Acetic Acid
o
bfl
H
a) 8
P..
Retention Time
6
(min)
8
10
FIGURE 12
48
-------
20
16
12
Chromatograph Run
Flow - 50 cc/min
Temp. = 150-200°C @ 10°C/min
Packing - Porapak Q
Sample #11 + Formaldehyde
u
bfl
H
0)
O-i
8
I
I
0
4 6
Retention Time (min)
8
10
FIGURE 13
49
-------
20 r
16
12
Chromatograph Run
Flow = 50 cc/min
Temp. = 150-200°C @ 10°C/min
Packing - Porapak Q
Sample #11 t Formic Acid
-p
a
*<-\
(U
SJ
s
8
4 6
Retention Time (min)
8
10
FIGURE
50
-------
Table 4
Summary of Paper Reaction Runs
Sample
No.
1
2
3
4
5
6
7
8
9
10
11
12
Sample Size
Grams
15
15
15
20
20
20
25
30
35
40
40
40
Air:N
Ratio
2:1
2:1
1:1
00
0
2:1
0
3:1
00
00
2:1
3:1
Residue
Grams
3.5
0.7
1.4
0.1
6.7
6.0
7.3
6.2
7.0
10.0
10.4
10.0
Temperature
°C
458
480
532
534
534
523
546
529
580
532
558
548
Liquid
pH
2.5
2.5
3.5
3.5
3.0
2.5
3.0
2.5
3.0
3.0
3.0
3.0
51
-------
Table 5
Summary of Retention Times
Porapak Q
Compound 150 C
Air 0.5 min
Water 0.90
Alcohols
Methanol 1.70
Ethanol 3.50
Iso-
propanol 8.20
n-
propanol
Butanol
Acids
Formic
Acetic
Propionic
Butyric
Aldehydes
Formal-
dehyde
Ketones
Acetone 7.40
But an one
175°C
0 . 5 min
0.80
1.25
2.40
3.70
4.70
10.40
2.15
4.30
9.30
14.80
8.20
3.50
7.00
200°C
0.5 m:
0.60
0.90
1.45
2.10
2.50
5.10
1.45
2.55
5.00
9.60
4.20
2.05
3.75
Porapak N
3.10
6.70
9.30
12.70
9.60
1.90
3.75
6.05
8.45
8.50
5.45
10.60
1.30
2.20
3.35
4.20
8.70
4.35
6.55
12.70
2.90
5.60
Note: - indicates that the compound wasn't run at this temperature either
because of poor separation or lengthy retention time.
Flow = 50 cc/min
52
-------
pH of these liquid samples which ranged from 2.5 to 3.5. Specific values are
given in Table 4. The presence of the above products was tentatively confirmed
in several samples by wet chemical methods. These analyses also suggested the
presence of methanol.
A limited amount of work was done on the gas adsorbed in the last trap.
Chromatography showed the presence of at least nine components in this gas.
Methane was found in all cases, together with carbon dioxide, carbon monoxide
and formaldehyde.
B. Reaction Runs Using Qrgro
The products of reaction obtained in the first trap, which was water
cooled, were found to be high viscosity, high-boiling tars. Because of the small
quantity and the difficulty of handling,little progress was made in analysis of
this fraction.
The second trap which was cooled by ice-water contained a liquid which
was over 95% water. The small amount of organic matter present was roughly sim-
ilar in composition to the material in the third trap. The third trap which was
cooled by dry ice-acetone contained the bulk of the organic products. This
material was liquid at the trap temperature but also contained dissolved gases
which were analysed.
A chromatogram of the separation obtained using the preparative gas
chromatograph is shown in Fig. 15. The sample treated was 1.0 cc of the organic
layer collected in the dry ice-acetone trap from the partial oxidation of Orgro.
Not all of the eluted peaks have been shown in this figure. Those that occurred
after 46 minutes were quite small and were not collected due to their low con-
centration. There were a total of 34 peaks that could be discerned. The compo-
nents that were collected in preparative collection traps were those at retention
-------An error occurred while trying to OCR this image.
-------
times of 14-, 16, 33 and 37 minutes. The other components were collected in the
by-pass trap. These four fractions were run on an infrared unit to be discussed
later.
Fig. 16 is a chromatogram of the organic layer collected in the dry ice-
acetone trap from a pyrolysis run using Orgro. Nitrogen was used as the fluidizing
medium and no air was added. This chromatogram shows remarkable similarities
to the chromatogram in Fig. 15. These two chromatograms represent the planned
extremes in operating conditions, i.e., those of partial oxidation and pyrolysis.
Figs. 17, 18 and 19 represent the infrared work done with the absorbed
gases resulting from the partial oxidation of Orgro. As described previously,
these gas samples were taken at sequentially higher bath temperatures and loaded
into a 10 cm path length infrared gas cell. The spectra obtained were interpreted
29
in terms of published data. The presence of carbon dioxide, propane, methanol,
and acetone is indicated.
Figs. 20, 21 and 22 represent the results comparable to the above for the
absorbed gases resulting from the pyrolysis of Orgro. These spectra indicate the
presence of carbon dioxide, carbon monoxide, methane, ethylene, propane and ammonia.
Fig. 23 is an infrared absorption spectrum of the fraction collected in
one of the preparative collection traps at a retention time of 14 minutes , after
being rechromatographed. An infrared spectrum of pure toluene was determined, as
shown in Fig. 24. The striking similarity confirms the presence of toluene.
Considerable quantities of a white salt was recovered by addition of
acetone to the aqueous layers. This salt appeared to sublime at a temperature
of 60 C with the evolution of a strong odor of ammonia. This information suggested
that the salt was ammonium carbonate. An infrared spectrum of ammonium carbonate,
shown in Fig. 26, confirms this suspicion.
55
-------An error occurred while trying to OCR this image.
-------
O
O
o
o
CN
O
CO
o o oo
NT 10 ON.
57
8
o
H
h
£ -p
-------
CO
c
to i
S3 H o
H 0} -P
W -H
!D -P tJ
E td -P
3 &< oj
-p -P O
t) d Cj
O -H >
P. o w
en o<
I 'Tj
c to n
o a a)
H T-l
p H
o
co
a;
M
a
-------
8
3DNVQ3OSSV 59
-------
o
CN
O
O
o
00
13
-P rH
cd a;
to
S -p
O S3
0>
O 3
ho w
C tn o
H -H B
in co .p
_ H ,
S O -<,C\J
£H >^ O
-PP.4-
PI n
Pf -H W
O -H
in o
n rn
-------
o
s
8
cd
O
P
flj
x: H
-P 0)
tn
SS
a
o
H
0)
!-t
0
cd P3 ft
t0 05 O
C -H E
H W -P
W '
P O
O p p
0) C
PI -H 'a
CO O tt>
" p
C
Q
H
bO 3
C O
U3>
S315
w W
rtJ t-1
eU O
|