Pilot Plant for Production of Single-Cell
Protein from Cellulosic Wastes
-------
CONSTRUCTION OF A CHEMICAL-MICROBIAL
PILOT PLANT FOR PRODUCTION OF
SINGLE-CELL PROTEIN FROM CELLULOSIC WASTES
This report (SW-24a) was prepared for
the Federal solid waste management program
by C. D. CALLIHAN, and C. E. DUNLAP
Department of Chemical Engineering
Louisiana State University
Baton Rouge, Louisiana, under
Contract No. PH 86-68-152
U.S. ENVIRONMENTAL PROTECTION AGENCY
1971
-------
An environmental protection publication
in the solid waste management series (SW-24c)
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.26
Stock Number 6502-0027
-------
FOREWORD
The 1965 Solid Waste Disposal Act (Public Law 89-272) as amended
by the Resource Recovery Act of 1970 (Public Law 91-512), placed great
emphasis on the recovery and reuse of resources now being wasted. Cellu-
lose represents more than 50 percent of municipal refuse and a much
larger fraction of agricultural wastes. As the use of paper and paper
products grows, the amount of cellulose waste also increases.
Louisiana State University, under contract to the Federal solid
waste management program of the U.S. Environmental Protection Agency, has
constructed and operated a pilot plant for the fermentative production
of bacterial single-cell protein from cellulosic wastes. This plant was
constructed at the Mississippi Test Facility (MTF) of the National
Aeronautics and Space Administration (NASA), whose participation and
aid greatly enhanced the value of the project. This report on the results
of that contract (No. PH 86-68-152) was prepared by the Department of
Chemical Engineering of the University. The Federal solid waste manage-
ment program was represented by Thomas C. Purcell and Clarence A. demons
during the implementation of the contract and the preparation of the
report.
—RICHARD D. VAUGHAN
Deputy Assistant Administrator
for Solid Waste Management
iii
-------
CONTENTS
Page
SUMMARY 1
CONCLUSIONS AND RECOMMENDATIONS 2
INTRODUCTION 5
AVAILABILITY OF CELLULOSIC WASTES 7
Urban Wastes 8
Agricultural Wastes 8
Economics of Cellulose Wastes 9
THE PROBLEM OF PROTEIN 11
INITIAL DEVELOPMENT OF THE PROCESS 13
Cellulose Properties and Treatment 13
Cellulose Fermentation 19
THE PILOT PLANT 26
Cellulose Handling Section 29
Cellulose Treatment Section 33
Sterilization Section 43
Fermentation Section 50
Harvesting Section 57
PILOT-UNIT OPERATION 61
Size Reduction of Solids 61
Solids' Dry Handling 62
Alkali-oxidation Treatment 63
Media Composition 69
-------
Page
Fermenter and Feed Stream Sterilization 74
OC
Inoculation and Fermentation °J
Harvesting ^2
PRODUCT QUALITY AND BY-PRODUCT USAGE 104
ECONOMIC POTENTIAL 109
Productivity
Cell Harvesting 115
Product Purification H7
Product Value
Summary of Product Cost
REFERENCES 123
INDEX OF FIGURES
1. Effect of pH on Growth of the Organism and Activity of the
Cellulase Enzyme 25
2. Pilot-plant Flow Sheet 27
3. Pilot-plant Floor Plan and Equipment List 28
4. Knife Grinder —Initial Size Reduction 30
5 . Solids Blower —Air-conveying System 31
6. Solids Cyclone, Hopper, and Metering Feeder 32
7 - Alkali-solids Slurry Tank 34
8. Cellulose Treatment Section 36
9. Solid-liquid Separator 37
10. Solid-liquid Separator and Oxidation Oven 38
11. Infrared Oyen—Heating Elements 40
12. Infrared Oven—Control Panel 41
13. Reslurry Tank 42*
14. Steam Inj ector 44
vi
-------
Page
15. Steam Injection Heater 46
16. Steam Injector, Holding Section, and Evaporative Cooler 48
17. Chilled-water Heat Exchanger 49
18. Fermenter 51
19. Pilot-plant Fermenter 53
20. Control Panel—Fermenter 55
21. Control Panel—Fermenter 56
22. Cellulose and Cell Concentration 58
23. Mixer/Settlers and Flocculent Tank 59
24. Growth of Cellulomonas on Bagasse Treated for 5 min in
Different Alkali Concentrations 66
25. Growth of Cellulomonas on Bagasse Treated for 2 hr in
Different Alkali Concentrations 67
26. Effect of Different Carbon Sources on the Growth of
Cellulomonas 72
27- Effect of Different Nitrogen Sources on the Growth of
Cellulomonas 73
28. Effect of Phosphate Level on the Growth of Cellulomonas 76
29. Effect of Sodium Chloride Level on the Growth of
Cellulomonas 78
30, Effect of Trace Mineral Level on the Growth of
Cellulomonas 80
31. Initial Sterilization Profile 83
32. Equilibrium Continuous-sterilization Temperatures 84
33, Effect of Changing Agitation in a Batch Fermentation 87
34. Cell Density and Soluble Carbohydrate Concentration Versus
Time for a Batch Fermentation 89
35. Cell Density Versus Time for a Continuous Fermentation 90
36. Fermentation of Unwashed, Treated Bagasse 92
37, Fermentation of Washed, Treated Bagasse or Purified Wood Pulp.. 92
vii
-------
Page
38 . Calculation of dX/dt Values ................................... 93
39 . Calculation k. Values .......................................... 94
40. Determination of Maximum Cell Production and Optimum
Feed Rate .................................................. 94
41. Single-cell Protein, Freeze-Dried and Drum-Dried .............. 107
INDEX OF TABLES
1. Proximate Analysis of Agricultural Cellulosics ................ 15
2, In Vitro Rumen Fluid Digestibilities of Cellulosic Wastes
Before and After Treatment ................................. 20
3. Descriptive Chart of Cellulose-utilizing Organism ............. 23
4, Amount of Carbohydrate Solubilized by Alkali Treatment ........ 68
5. Pilot-plant Media Composition for Runs Three, Four,
and Five [[[ 70
6. Effect of Different Carbon Sources on the Growth of
Cellulomonas ............................................... 71
7- Effect of Nitrogen Level on the Growth of Cellulomonas ........ 74
8. Effect of Phosphate Level on the Growth of Cellulomonas ....... 75
9. Effect of Sodium Chloride Level on the Growth of Cellulomonas.. 77
10, Effect of Trace Mineral Level on the Growth of Cellulomonas ... 79
11 , Replacement Inorganic Nutrients ............................... 81
12 . Pilot Plant Nutrient Media for Runs 6 Through 15 .............. 82
13. Fermentation Batch and Continuous Run Data for 141-gal
Pilot-plant Fermenter ...................................... 96
14 . Product Analysis .............................................. 103
15. Growth Yields of Cellulomonas on Carboxymethyl Cellulose ...... 105
16. Essential Amino Acid Content of the Cell Protein (g of
Amino Acid per 100 g Protein) ........... , .................. 106
17- Comparable Volumetric Production Efficiencies for
Continuous Fermentations . .................................. 113
#
18 . Symbiotic Growth .............................................. Hg
-------
SUMMARY
A pilot plant was designed, constructed, and operated that produced
microbial single-cell protein from waste sugarcane bagasse. Bagasse was
ground and given a mild alkaline-oxidation treatment before fermentation
Bagasse was slurried at as much as 10 g per liter dry weight in water with
a simple nutrient salts mixture to form the fermenter feed stream. The
process was operated in both batch and continuous-flow patterns.
Cellulomonas, sp. bacteria were used in a pure culture for most runs,
but a mixed culture run of Cellulomonas and the symbiotic organism
Aloal'lgenes fa&cal'ls showed much higher production capabilities. Maximum
cell density obtained with pure Cellulomonas was 1.7 g dry weight per liter
and 6,24 g per liter for the mixed culture. Culture mass-doubling times
during log-phase growth were usually from 3.2 to 3.7 hr.
The maximum experimental volumetric production efficiency (VPE) of a
continuous run using pure Cellulomonas was about 0.10 g of dry cell mass
per liter of fermenter capacity per hr. The mixed culture run had a
theoretical VPE of 0.512.
Single-cell protein was produced as a light brown to yellow-brown
powder and had a crude protein content of 50 to 55 percent.
-------
CONCLUSIONS AND RECOMMENDATIONS
The following conclusions were drawn from data and experience obtained
during the fulfillment of contract PH#86-68-152.
1. Alkali-treated sugarcane bagasse can be fermented on a
continuous-flow basis by Cellulomonas, sp. bacteria for the
production of single-cell protein.
2. Cellulomonas will preferentially metabolize soluble carbohydrate
rather than insoluble cellulose if both are present in the media.
3. Bagasse jnust be subjected to alkali treatment before appreciable
bacterial attack can occur.
4. As much as 90 percent of treated bagasse can be solubilized in
batch fermentations, but not all of this is metabolizable
carbohydrate.
5. About 26 percent of whole, bone-dry bagasse is converted to cell
mass at a continuous fermenter efficiency of 75 percent.
6. Usual log-phase culture mass-doubling time for Cellulomonas is
3.2 to 3.7 hr.
7- Increased agitation increases final cell density and growth rate
probably by improving oxygen and substrate mass transfer.
8. Fertilizer-grade and industrial-grade chemicals may be used in
most cases to replace laboratory-grade or reagent-grade salts
used in the nutrient jnedia.
9. The presence of alkali in the feed stream improves the efficiency
of continuous sterilization.
-------
10. Mixed-culture fermentation with Cellulomonas and A.lcal'lgenes
faecalis gives much higher cell density than a pure Cellulomonas
culture.
11. The maximum cell density obtained with a pure Cellulomonas
culture is 1.66 g per liter dry weight. The mixed-culture run
had a cell density of 6.24 g per liter.
12. Maximum experimental volumetric production efficiency obtained
by a pure Cellulomonas culture in a continuous run was 0.098 g
of dry cell mass per liter of fermenter volume per hour.
Calculated VPE for the mixed culture was 0.512 g per liter per
hour.
13. Cellulomonas contains about 50 to 55 percent crude protein
(Kjeldahl method) and has a good amino acid balance.
14. Passage of the alkali-treated solids through the oven was found
not to be necessary to assure degradation by the organisms.
The following recommendations have been made in light of experience
gained during the performance of the contract.
1. Further work needs to be done in defining the limits and
capabilities of mixed-culture fermentation of cellulose.
2. The pilot plant should be modified to permit less severe alkali
treatment of the cellulosic before fermentation.
3. A second fermentation vessel should be added to the pilot plant
to permit two^stage fermentation.
4. A harvesting method needs to be perfected that will give a clean
cellulose-free cell product that is low in salt.
5. further work needs to be done on recycle and use of the solid
and liquid by-product streams.
-------
6. An automatic antifoam addition system needs to be added to
the fermenter, and concurrent analysis of cell density and
substrate concentration should be perfected.
7. The infrared oven should be removed as a pre-fermentation
processing step in the treatment of bagasse.
-------
CONSTRUCTION OF A CHEMICAL-MICROBIAL
PILOT PLANT FOR THE PRODUCTION OF
SINGLE-CELL PROTEIN FROM CELLULOSIC WASTES
The 1970's have been introduced as a decade of conservation and
environmental reclamation. It is fervently hoped that this will prove to
be the case. The life-supporting resources of the earth and its atmosphere
are being depleted and polluted at an alarming rate while the population
increases ever more quickly. The use of the earth's ecosystem by man has
resulted in the absolutely predictable problem of too much waste and too
few resources. The historical role of industry has been one of processing
raw jnaterials into usable consumer goods. The inevitable fate of these
goods has been to become wastes. Nature is then left with the task of
reconverting these wastes into resources.
Increasing population and industrialization have increased waste output
to a point where natural reclamation pathways cannot keep up. The synthesis
pathways of nature are overloaded. It is clear that new methods of waste
reclamation and reuse must be developed. The usable lives of raw materials
must be lengthened, and more efficient ways of returning wastes to resources
must be found.
In September 1968, the Department of Chemical Engineering of Louisiana
State University, under contract to the Bureau of Solid Waste Management of
the Department of Health, Education, and Welfare, began the design and
construction of a pilot-plant unit for the conversion of waste cellulose
materials into bacterical single-cell protein (SCP) by fermentation.
-------
Prior laboratory work had shown that the proposed process was
technically feasible, and the preliminary economic data were encouraging.
The scope of the contract was to design, construct, and operate a pilot
unit; produce bacterial single-cell protein for analysis and testing; and
evaluate process economics more completely. The initial operation of the
pilot unit has been limited to a single waste cellulosic substrate,
sugarcane bagasse. This material represented a rather typical heterogeneous
cellulosic agricultural waste, was easily accessible, and was a good example
of an under-utilized raw material. Purified ground wood pulp was also used
as a control substrate in several runs.
The process was designed for a dual purpose. The first was the use of
present wastes and potential pollutants; the second was the production of
inexpensive, high-quality protein for food. The success of the idea may be
judged from the tremendous interest and response this project has received
from industry; the success of the process will be proved by its future
industrial exploitation.
-------
AVAILABILITY OF CELLULOSIC WASTES
Cellulose is by far the most widespread and readily available of all
solid organic materials. It comprises almost one-third of the weight of all
trees, vines, grasses, and straws. Unlike other resources such as oil and
minerals, cellulose is constantly replenishing itself by photosynthesis and
growth. Vast quantities of this material accumulate as waste products from
activities such as food processing, lumbering, paper making, and cereal
grain harvesting. Additionally, municipal and industrial wastes of paper,
rags, boxes, wood, excelsior, grass, and leaves raise the available amount
of cellulose-bearing material to astronomical proportions.
The ready availability of such a quantity of cellulose has been the
impetus for voluminous research into novel and diverse methods of utilizing
this material. Wallboard, door cores, and mulch have been made from cereal
grain straws and sugarcane bagasse; chemical-grade cellulose from cotton
linters, wood, and bast fibers; animal feed pellets from pea vines and other
fibrous vegetable roughages; animal bedding from oat, rice, and wheat straw;
furfural and brake shoes from sugarcane bagasse; various chemicals from corn
cohsj and limitless other examples. Still, most waste cellulose is dumped,
buried, burned, or used as a fuel supplement; and its chemical energy or
physical utility is thereby wasted or only fractionally retained.
Cellulose is a major constituent of all woody plants, grasses, and
vegetables. The few plants that have become industrially important for
theij cellulose—trees, cotton, flax, etc—have been chosen, in most cases,
for their physical rather than chemical natures.
7
-------
The availability of waste cellulose may be brought into better
perspective by considering urban or municipal wastes and agricultural
wastes separately. Cellulose found in urban wastes; for example, paper,
rags, cardboard, has usually been subjected to some type of processing,
while cellulose agricultural wastes are almost always found in the native,
heterogeneous state.
Urban Wastes
The United States alone produces more than 250 million tons of urban
2 3
wastes per year. Cellulose comprises from 40 to 50 percent of this total.
Most of the cellulose of urban waste is from paper, while leaves, grass, and
A
wood supply most of the rest. Only 18 to 20 percent of the waste paper is
reclaimed and reused for paper stock, and all of the rest is either
incinerated or used as landfill.
Of the cellulose found in urban refuse, much more than 20 percent could
be reused in paper making. Japan currently reuses more than 40 percent of
her urban waste cellulose. Huch of the cellulose in urban wastes is,
however, either of such poor fiber quality or so intimately mixed with
noncellulosics that it is not suitable for recovery and reuse. Yet, if this
cellulose could be used for its chemical rather than physical properties, a
far greater portion could be removed from incinerators and landfills.
The yearly appearance of about 70 million tons of cellulose in urban
wastes, less about 10 million tons reclaimed for reuse, leaves 60 million
tons per year net waste cellulose in the United States alone.
Agricultural Wastes
*
The IJ. S. Department of Agriculture has estimated that more than 200
million tons of cellulosic agricultural wastes such as plant stems, straw
-------
leaves, grasses, bagasse, and husks are produced every year in the United
States. More wastes come from canning and food packaging and preparation
plants. Less than 1 percent of these agricultural wastes are -used; most are
left in the fields to rot. Some of these wastes accumulate at some central
point during processing. Sugarcane bagasse, sugar beet vinasse, rice and
wheat husks, corn cobs and husks, and several others are generally brought
to a central point in their usual processing cycle. When these materials
accumulate, they present a disposal problem and rank as pollutants simply by
their volume and lack of profitable use. Much of this material is burned as
fuel to fire plant boilers. Some, however, is reused and made into various
construction materials and agricultural products.
Total world production of sugarcane bagasse is about 36 million tons
yearly. The United States contributes about 13 million tons. Most of this
is burned as boiler fuel in the sugar mill, but some is used for paper stock,
hardboard, furfural production, and charcoal.
Economics of Cellulose Wastes
Since waste materials do not usually enter the economic or industrial
cycle, it is often difficult to quote a price or value for them. Because
they are of rather low economic value, the costs of handling, storage,
transportation, and preparation become large factors. Some cellulosic waste
like mixed urban refuse can be attributed a negative cost, usually equal to
the disposal costs, and the value of some is determined as fuel replacement
value,
The cost of sugarcane bagasse in the United States ranges from about $6
to $13 per ton baled. The bagasse contains about 50 percent moisture.
Cellulose makes up from 50 to 60 percent of bone-dry bagasse, and
-------
10
hemicelluloses add from 10 to 20 percent. The remainder is largely lignin
and ash. The mill cost of bone-dry bagasse carbohydrate then is from $15 to
$40; an average cost would be about $20 per ton of bone-dry carbohydrate
from bagasse. Added to this cost are the costs of baling, handling,
storage, and short-haul transportation. This usually adds about $10 per
ton. Total cost would therefore be about $30 per ton, or 1.5 cents per Ib
of fermentable carbohydrate from bagasse. The cost of Number One Mixed
Grade waste paper is about the same.
In every case dealing with waste cellulose reuse, the cost of
transportation makes it necessary to limit hauling distances to a minimum.
If costs are calculated on materials such as cereal grain straws, then
gathering costs must also be added. Yet, with a maximum cost of 1.0 to 1.5
cents per Ih, low-grade waste cellulose remains a relatively inexpensive
raw material.
-------
THE PROBLEM OF PROTEIN
Coupled with the problem of increasing waste loads is the problem of
dwindling resources. A major resource in short supply now appears to be
food, and more particularly, protein.
The scope of the world food problem up to the present has been limited
mainly to the underdeveloped continents of Asia, Africa, and Latin and
South America. These nations face a staggering food deficit before 1975.
Predictions show, however, that the more developed nations will eventually
face the same problems. Food and Agricultural Organization (FAO) estimates
for nations such as Pakistan and India run well over 100 percent beyond
Q
current demands.
In particular, the problem of protein deficiency has received much
interest since high-quality protein is in such shortage in countries like
9 10
India, Pakistan, and Brazil. ' Protein demand for the future largely
follows the predictions for calory requirements. The protein problem is
further complicated by the protein quality factor. To be of usable
metabolic quality, protein must contain all the amino acids necessary for
growth., maintenance, and reproduction in a balance suitable for efficient
use. This high-quality protein has been supplied almost entirely by meat,
jnilk? fish, eggs, and poultry. Present trends show, however, that
population growth and increasing food demands will severely exceed protein
supply from these sources. Better animal husbandry is certainly possible,
especially in the underdeveloped countries, but that alone cannot cope with
11
-------
12
increasing requirements. Technology is faced then with the problem not
only of increasing the world protein supply by 50 percent in the next
20 years, but also of producing protein of low cost, high quality, easy
distribution, and high social acceptability.
The process developed at Louisiana State University (LSU) is the
fermentation of insoluble cellulose by a cellulolytic bacterium. The
bacteria are then harvested from the media for use as a food protein. The
process was developed for the utilization of excess sugarcane bagasse, and
bagasse has been retained as the sole carbon source for most of the pilot
unit runs. The SCP produced is a light brown-to-yellow powder having a
crude protein content of from 50 to 60 percent. The SCP has a good amino
acid pattern and has served as a protein source in successful rat-feeding
studies. More of the economic aspects, quality considerations, and market
properties of this SCP product will be discussed later in this report.
The term fermentation, as used in this report, means the respiration
and subsequent growth of a bacterial culture under aerobic conditions.
-------
INITIAL DEVELOPMENT OF THE PROCESS
Studies on the sources of protein show that protein production rates
and efficiencies differ widely. Differences in protein production rates
were demonstrated by Thaysen with the example that a 1,000-lb bullock can
synthesize 0.9 lb of protein every 24 hr, whereas 1,000 Ib of soybeans
synthesize 82 lb of protein in the same length of time, and 1,000 lb of
yeast could produce more than 50 tons of protein in 24 hours. The
difference in the efficiency of protein production between, for example,
the bullock and yeast growth is equally marked. If carbohydrate were fed
to both, the amount ultimately formed into protein would be no greater than
5 percent in the bullock but more than 25 percent in the yeast.
Such drastic difference in production rates and efficiencies have
claimed the interest of many investigators whose proposals have, in some
way, been concerned with either shortening the protein production chain or
increasing the efficiency of one or more of the steps.
A comparison of the land area necessary for the support of each of
these protein-producing activities would show again the savings and
efficiency enjoyed by the yeasts. In the LSU process, a bacterium was used,
and the carbohydrate source was cellulose.
Cellulose Properties and Treatment
The natural carbohydrates that occur in plants are of a rather motley
character. Many different sugar molecules may be found; and these, in the
native cellulosic, are bound together by several different types of
13
-------
14
linkages. The classical definition of cellulose is a linear polymer of
anhydroglucose units linked at the one and four carbon atoms by a
beta-glucoside bond. The number of repeating units may range from about
30 to 5,000 or more. The highest degree of polymerization (DP) recorded
is about 15,000 units. In native celluloses the polymer chains are of
widely varying lengths; and so, polymer weights and degrees of polymerization
differ throughout a sample.
In addition to the carbohydrates, these native agricultural wastes
also contain lignin, protein, fat, and ash. A proximate analysis gave the
relative fractions of these constituents in several different types of
cellulosic agricultural wastes (Table 1).
The macromolecular physical structure, or fine structure, of a
cellulosic material in the native solid state is a complex function of
intermolecular and intramolecular forces between and within individual
cellulose polymer chains; between cellulose fibrils in fibers; and between
fiber units and the hemicelluloses, lignin, gums, resins, and minerals.
Since formation of the physical structure of cellulose occurs as a growth
process, the interrelationship of the various components is in a dynamic
and eyer-changing state.
HL. .Mark states that all properties of cellulose (both chemical and
physical) are, in the last analysis, determined by chemical structure but
that forces between the cellulose polymer chains produce a super-molecular
12
texture that profoundly influences most properties of the material.
General agreement has been reached on the more important points of
gross physical structure of cellulose as follows:
1. The polymer chains of natural cellulose exist in differing degrees
of order with respect to each other.
-------
TABLE 1
PROXIMATE ANALYSIS OF AGRICULTURAL CELLULOSICS1
Substrate
Cotton linters - raw
Cotton linters - purified 3.27
Bagasse - whole
Bagasse - pith
Bagasse - fiber
Rice straw
Sawdust - pine
Johnson grass
Prairie grass
Alfalfa meal
Cottonseed hulls
Corn cobs
Oat straw
Wheat straw
Sorghum bagasse
*NFE is Nitrogen-Free Extract.
rotein
2.45
3.27
2.68
2.97
1.74
5.05
0,42
6,71
5.32
22.11
5.53
2,61
7,74
2.71
2,73
Fat
0.96
1.69
1.0
5.23
0.76
2.41
2.89
2.87
2.06
2.38
1.51
0.87
2.0
1.25
2.73
NFE*
4.48
3.69
31.1
30.83
32.78
36.70
13.16
39.44
41,84
39.46
15.83
46.89
35.50
35.58
41.60
Fiber
85.7
90.0
52.74
47.14
52.39
34.39
74.83
37.00
36.73
23.88
50.54
37.10
36.94
46.08
40.54
Lignin
5.0
0.84
10.63
9.28
10.23
4.06
8.45
5.01
7.39
5.39
23.31
10.44
7.41
8.02
7.17
Ash
1.27
0.42
1.79
4.52
2.06
17.36
0.21
8.95
6.63
6.74
3.25
2.06
10.39
6.36
4.21
-------
16
2. The most highly ordered fraction of cellulose gives a definite
and unique crystalline diffraction pattern in x-ray diffraction
photographs. The exact size and lattice angles may be computed
for a characteristic single crystal.
3. The least ordered fraction is entirely amorphous and shows no
regularity whatsoever.
4. The crystalline fraction, whether composed of discrete crystallites
or of a continuous nature, is difficultly penetrable by solvents,
enzymes, or reagents.
5. The amorphous region, whether interstitial or sequential, or both,
is easily penetrable by solvents, enzymes, or reagents.
6, Cellulose in either region has the same chemical structure.
The relative crystallinites of several cellulosics, determined by the
13 1
.x-ray diffraction technique of Segal are listed here.
Cotton linters, raw 72.8
Cotton linters, purified 81.0
Bagasse - whole 42.5
Bagasse - pith 32.3
Bagasse - fiber 42.3
Rice straw 43.3
Sawdust - pine 38.0
Johnson grass 33.3
Prairie grass 39.2
Alfalfa meal 33.3
Cottonseed hulls 42.0
Corn cobs 28.6
Oat straw 38.7
Wheat straw 46.6
Sorghum bagasse 42.2
It is evident that relative crystallinities differ widely among the
materials, whole sugarcane bagasse being in an intermediate position.
Seyeral investigators have noted that the reactivity or enzyme degradation
*
14
of cellulose varies inversely with its crystallinity.
-------
17
An Important property of cellulose is that it can be penetrated and
swollen by some strongly electrolytic solvents. This phenomenon has long
been used in treatment of cellulose to increase reactivity, to change the
solid physical state, and to improve dyeability and surface properties. A
partial list of chemical agents that will penetrate and swell cellulose to
some degrees follows.
2% sodium hydroxide + carbon disulfide
Calcium thiocyanate
Cuprammpnium hydroxide
Sodium hydroxide
Sulfuric acid
15% sodium hydroxide + carbon disulfide
Ruthenium red
Copper ethylendeiamine
Phosphoric acid
85% formic acid + zinc chloride (80:20) AmZn
Trimethylbenzylammonium hydroxide
Iron tartrate complex
Hethacrylate embedding
Sodium zincate
Cadoxen
Of the reagents listed, cuprammonium hydroxide, sodium hydroxide,
sulfuric acid, and copper ethylenediamine have received most interest and
industrial application. Other chemical agents such as nitrogen dioxide in
dimethylsulfoxide, and zinc chloride have recently gained industrial
interest. An excellent review of the literature on action of
15
celluloses-swelling agents has been published.
The importance of the swelling action of some reagents upon cellulose
has been realized for a long while. Almost all processes that involve
reactions of cellulose utilize either a pretreatment to swell the cellulose
or a reaction solvent that also acts as a swelling agent. The formation of
alkali-
-------
18
The degradation of cellulose is generally understood to mean a decrease
in the average degree of polymerization of the polymer. McBurney has
16
characterized the four major types of cellulolytic degradation:
1. Hydrolytic - cellulose is reduced in DP and shows an increase in
reducing power.
2. Oxidative - cellulose is reduced in DP, and the product shows a
development of carbonyl and carboxyl groups.
3, Microbiological - cellulose may be reduced in DP, but loss of
strength is the most pronounced effect.
4. Mechanical - cellulose is reduced in DP if the fiber is subjected
to severe physical treatment.
Cellulose can be degraded microbially by microorganisms or enzyme
systems of these organisms. Ruminant animals, termites, some snails, and
many bacteria and fungi depend upon their ability to degrade cellulose and
metabolize its degradation products for their supply of carbon and energy.
In each case the instrument of degradative action is an enzyme or enzyme
system produced by the organism.
The ideas used in the design and development of the pre-fermentative
treatment of the cellulosic wastes for the LSU process were directed towards
making the cellulose more bio-degradable. The criteria of the treatment
process were;
1. To decrease the relative crystallinity of cellulose
2. To disrupt the. physical structure of the lignin in the material
3. To decrease the average DP of the cellulose
4. To obtain these changes at a competitive cost.
A mild alkaline swelling of the cellulose followed by an air oxidation
seemed to meet all these criteria. Sodium hydroxide is available at low
-------
19
cost from by-product streams of chlorine manufacture. It swells the
cellulose and solubilizes part of the lignin. The swelling causes the
lignin sheathing to be disrupted, and the cellulose is made more accessible
to enzymes. If this treatment is followed by an air oxidation, the
cellulose is degraded to a lower DP and the relative crystallinity is
reduced.
An average component analysis of bagasse before and after the treatment
process showed that the carbohydrate fraction remaining in the bagasse after
a water extraction was increased from about 57 percent to almost 75 percent.
The water-soluble fraction of carbohydrate rose from about 2 percent to
almost 18 percent. This rise in relative carbohydrate content was caused by
preferential removal of noncarbohydrate components such as resins, gums,
lignin, fat, protein, and dirt. The increase in water-soluble carbohydrate
was caused by the oxidative de-polymerization of a fraction of the cellulose
present.
Bagasse showed that a threefold decrease in DP during the treatment,
averaging a DP of more than 800 before treatment and less than 300 after.
The overall degree of relative cellulose crystallinity also was lowered from
about 50 percent to 10 percent by the treatment.
Ten different cellulosics were treated by alkaline-oxidation, and the
biodegradability of their cellulose fractions was determined by the in vitro
rumen fluid method of Baumgardt. The treatment increased cellulose
digestibilities on an average of 85 percent (Table 2).
Cellulose Fermentation
All industrial use of carbohydrates for microbial substrates has been
limited to those that are water soluble. There has been considerable
-------
20
TABLE 2
IN VITRO RUMEN FLUID DIGESTIBILITIES OF CELLULOSIC
18
WASTES BEFORE AND AFTER TREATMENT'
Bagasse - whole
Bagasse - pith
Bagasse - fiber
Rice straw
Johnson grass
Prairie grass
Corn cobs
Oat straw
Wheat straw
Sorghum bagasse
(% Dry cellulose
Untreated
15.1
26.5
30.1
7.3
66.5
16.2
19.3
35.5
25.4
30.0
digested)
Treated
57.0
55.0
50.3
54.1
88.0
45.7
44.0
66.0
44.0
61.5
interest, however, in the utilization of soluble sugar fractions produced
from insoluble carbohydrate materials. Since Delbruck in Germany found that
food-grade yeast could be produced from solubilized wood sugars, sporadic
interest has recurred such as in Germany during World War II.19'20 The
German yeast process produced Candida utilis and several strains of lesser
importance from wood sugars generated by the sulfuric acid hydrolysis of
softwoods. Other yeast plants there and current U. S. plants use waste
21
sulfite liquor from paper pulp processing as a substrate. Processes have
also been developed for production of bacterial single-cell protein from
22
spent sulfite liquor.
-------
21
These processes have, in a sense, used insoluble material as substrate,
but in all cases the material has been chemically solubilized before
introduction into the actual fermentation process.
Viewed as a fermentable carbohydrate, cellulose differs rather
radically from carbohydrate substrates in general use. It is insoluble, it
is polymerized by a one-to-four beta-glucosidic linkage, it generally has a
highly-ordered crystalline fraction, and it is invariably found closely
associated with hemicelluloses and lignin. The use of cellulose as a
microbial substrate actually adds but one step to the overall mechanism of
carbohydrate fermentation. The cellulose must be solubilized before
entering cell metabolic pathways. This solubilization is an enzymatic step
catalyzed by the cellulase enzyme system of some bacteria and fungi.
For a long while researchers thought native cellulose was either
23
unfermentable or, at least, very resistant to biodegradation. Langwell,
O / O ^ O £i
Olson, Petterson, and Sherrard, Acharya, and Fontaine reported native
cellulose unfermentable or inhibited to a prohibitive extent by lignin.
Virtanin and others found that cellulose was biodegradable but that periods
27
of 3 to 4 weeks were necessary for cellulose breakdown. Hajny, Gardner,
and Ritter have reported cellulose biodegradation by thermophiles, but
28
residence times were 2 to 6 days. Grey reported conversion of cellulose
to cell tissue in fungi, but again rate is measured in days, and maximum
29
carbohydrate-to-protein conversion efficiencies are 17 percent.
Recently the isolation of several mesophilic and thermophilic bacteria
with faster growth rates and of fungi with highly active cellulases has
30 31 32
increased the prospects of profitable cellulose fermentation. ' '
The bacterium used in the LSU pilot fermentation process was isolated
32
and identified by Srinivasan and Han. The following is a summary of the
methods used in isolation and characterization of the organism.
-------
22
Isolation medium consisted of 6.0 g/liter sodium chloride, 1.0 g/liter
ammonium sulfate, 0.5 g/liter potassium phosphate dibasic, 0.5 g/liter
potassium phosphate monobasic, 0.1 g/liter magnesium sulfate, 0.1 g/liter
calcium chloride, with 0.1 percent yeast extract and a strip of filter
paper. About 1 g of a rotting sugarcane and soil mixture was inoculated
into the isolation media. After 3 to 7 days incubation at 30 C on a
reciprocal shaker, a portion of filter paper was transferred into fresh
media. This process was repeated several times to enrich the aerobic and
mesophilic cellulose-utilizing organisms. The filter paper from the
enriched culture was removed, macerated in a small amount of sterile water,
and streaked onto plates containing each of the following media: nutrient
agar, carboxymethyl cellulose agar, and filter paper agar (a plate of
nutrient agar covered with filter paper). Representatives of the various
colonies that developed on each of these media were transferred to test
tubes containing nutrient salts and cellulose. Tubes showing visible
degradation of filter paper were selected and alternately transferred into
liquid and solid media in order to enrich and isolate the cellulolytic
organism. Isolated colonies were further purifed by the terminal dilution
method. The purity of the isolated culture was confirmed by the microscopic
examination and colony morphology of several agar plate colonies. The
isolated culture was subjected to diagnostic tests for identification of the
strains. A characterization comparison was made with the isolated
cellulolytic organism and two organisms of the genus Cellulomonas (Table
3). The organism was able to grow well between the temperatures of 25 C and
35 C. Below 20 C and above 40 C, there was a marked decrease in the growtR
rate. The effect of pH on the growth rate and activity of the cellulase
enzyme was determined and a pH optimum was observed between 6.0 and 7.0
(Figure 1).
-------
23
TABLE 3
DESCRIPTIVE CHART OF CELLULOSE-UTILIZING ORGANISM
32
C.
C. uda
Isolate
Morphological
characteristics
Form
Size
Motility
Gram stain
Rods, curved
0.4-0.6y
X
0.7-1.8y
Nonmotile
Variable
Rods
0.5y
X
1.0-1.5y
Nonmotile
Negative
Rods, short
0.3-0.5y
X
0.7-1.2y
Nonmotile
Negative
Cultural
characteristics
Agar slant
Broth
Gelatin stab
Filter paper
in peptone
broth
Optimum
temperature
Agar colonies
Smooth, glistening, Moderate, flat,
opaque, yellow grayish white
Uniformly turbid
Slow liquefaction
Fibers separate on
slight agitation
28-33 C
Uniformly turbid
Slow liquefaction
Fibers separate on
slight agitation
28-33 C
Smooth,
glistening
opaque, yellow
Uniformly turbid
Slow liquefaction
Fibers separate on
slight agitation
25-35 C
Bluish,
transparent,
smooth, flat,
circular; grow
feebly on nutrient
agar
-------
TABLE 3 (Continued)
24
Biochemical
char ac t er is t ic s
Starch
Nitrate
MR test
VP test
Indole
production
Glucose
Lactose
Sucrose
Maltose
Dextrin
Starch
C. flavigena
Hydrolyzed
Reduce to NO-
—
Negative
—
Acid
Acid
Acid
Acid
—
Acid
C. uda
Hydrolyzed
Reduce to NO-
—
Negative
—
Acid
Acid
Acid
Acid
—
Acid
Isolate
Hydrolyzed
Reduce to NO-
Negative
Negative
Negative
Acid
Acid
Acid
Acid
Acid
__
-------
25
300
280
260
240
220-
200
180
!60-
O)
>-•§ 120
|| 100
80
a) 60
Ul wi
Z • 20
1.0
0.1
If
m o
Q o
>o
O °
c
0)
0.01
45678
pH
Figure 1. Effect of pH on growth of the organism
and activity of the cellulase enzyme.
-------
THE PILOT PLANT
The purpose of the contract was to construct a pilot fermentation unit
based on the findings of previous laboratory research. Criteria dictated
by both engineering and microbiological practice were used in the design.
Many of the larger pieces of process equipment were designed by LSU and
constructed and furnished by NASA. Most instrumentation and specialized
equipment was obtained with contract funds.
The plant was designed so that fermentations could be carried out in
both batch and continuous-flow operation. The general flow sheet (Figure 2)
and floor plan (Figure 3) of the pilot plant are about the same as proposed
but may be changed to some extent depending on optional use of some
equipment. The plant is housed in room 142 of building 8100 at the National
Aeronautics and Space Administration's (NASA) Mississippi Test Facility,
Bay Saint Louis, Mississippi. This room consists of a 30-ft by 60-ft ceramic
tile floor, tile walls, a 16-ft ceiling, a deionized water treatment and
storage system (1,500-gal capacity), a 50-gal chilled water system (50 F
water at a 90,000 BTU per hour heat load), a more than adequate ventilation
system, and a steam generator capable of 600 Ib of 100 psi steam per hour.
Available utilities consist of 115, 220, and 440 60-cycle AC electrical
power; potable water; 90 psi compressed air; and 90 psi compressed nitrogen.
The pilot plant's equipment can be grouped into five distinct process
sections: cellulose-handling, treatment, sterilization, fermentation, and
*
cell harvesting.
26
-------
Cellulose
waste
Water or Recycled
Liquor
Nutrients
Solid/Liquid
Separator
[Oxidation Oven
Steam
I
Blower
Mixer/Settler
FUxculent
Mixer/Settler
Liquor to -*
use or
Recycle
Dryer
Dry
-------
28
KNIFE GRINDER
BLOWER
HOPPER AND FEEDER
ALKALI SLURRY TANK
ALKALI STORAGE
OXIDIZING OVEN
RESLURRY TANK
STEAM INJECTOR
HOLDING SECTION
EVAPORATIVE COOLER
ACID STORAGE TANK
WORKING STAND
ROTARY DRUM DRYER
CHILLED-WATER HEAT
EXCHANGER
NUTRIENT STORAGE TANK
FERMENTER
FERMENTER CONTROL
PANEL
CENTRIFUGE
MIXER/SETTLER NO.l
FLOCCULENT ADDITION
TANK
MIXER/SETTLER NO.2
BULK FEED TANK
Figure 3. Pilot-plant floor plan and equipment list.
-------
29
Cellulose Handling Section
The first step in processing the cellulose is particle size reduction.
A five-bladed knife grinder fitted with a 1/8-in. sizing screen is used for
this purpose (Figure 4). The five fixed blades and five rotating blades
are 18 in. long. The power is supplied by a 7 1/2-hp electric motor that
provides for a 300 to 400 Ib per hr capacity. The grinder is fitted with a
collection hopper and a pneumatic solids-handling pickup attachment. The
cellulose is fed in manually and chopped to 1/8-in. particle lengths in
this first piece of equipment.
The chopped cellulose is transferred from the grinder to a
solids-separating cyclone by the solids blower (Figure 5). The blower is
powered by a 7 1/2-hp electric motor that turns a 24-in. fan. The inlet and
outlet are approximately 30 sq in., which is more than adequate to handle
the maximum grinder through-put.
The solids-separating cyclone collects the ground cellulose and
discharges the conveying air out the top port (Figure 6). The cellulose
drops into the hopper. The cyclone seems to handle all blower and solids
flows and is moderately dust free. No power is consumed in the cyclone
operation.
The solids are then collected in a vibratory live-bin hopper (Figure 6).
The hopper has a capacity of 15 cu ft, which provides inventory for several
hours of cellulose treating. Vibratory action is supplied by a 1-hp
electric motor operating an eccentric cam device. The hopper as purchased
had no level monitoring system. To remedy this, a vertical slit 1-in. wide
was cut in the side of the hopper and a clear plastic plate was used to seal
the slit. This provided a convenient, visible means of monitoring hopper
level.
-------
30
Figure 4. Knife grinder—initial size reduction,
-------
31
Figure 5. Solids blower—air conveying system.
-------
SOLIDS FROM
GRINKER
VIBRATORY
LIVE-BIN HOPPER
METERING FEEDEB
Figure 6. Solids cyclone, hopper, and metering feeder.
-------
33
The vibratory hopper supplies cellulose to the vibratory metering screw
feeder (Figure 6). The feeder has a 2-in.-diameter screw powered through a
variable-ratio belt and pulley assembly by a 3/4-hp electric motor. The
feeding capacity ranges from 0.33 to 2 Ib per min of chopped bagasse of bulk
density of 5.5 to 6.0 Ib per cu ft. This feeder was intended to meter the
amount of cellulose being fed into the feed makeup for the feed stream to
the fermenter. The required feed rate was found to be much smaller than the
range capability of the feeder. The feeder was, therefore, operated on a
manual time-cycling basis.
With the exception of the oversized feeder, all equipment in the
cellulose handling section worked with no problems. The mounting and support
frames in this equipment section, as in the rest of the pilot plant, were
designed by LSU personnel and constructed by Mississippi Test Facility (MTF)
support personnel. Virtually all equipment and associated frames in this
section were constructed from carbon steel.
Cellulose Treatment Section
The ground cellulose output of the handling section was fed into the
slurry tank (Figure 7) by the vibratory feeder where it underwent alkali
contacting. The slurry tank is a 60-gal, 304 stainless steel, cylindrical
vessel equipped with a propeller-type agitator, an automatic temperature
control system (ambient to boiling), and a liquid-level control that
activates a solenoid valve to admit makeup alkali solution. The heat for
the temperature control system is generated by three 3,300-watt electric
heating bands around the outside of the vessel under a 2-in. layer of
insulation. The tank is slightly elevated by three legs to make room for
the outlet plumbing. The outlet is 1-in. pipe located at bottom center of
the tank.
-------
VIBRATORY METERING
FEEDER
TANK TEMPERATURE
CONTROL
ALKALI-SOLIDS
SLURRY TANK
AUTOMATIC LEVEL
CONTROL
Figure 7. Alkali-solids slurry tank.
-------
35
The makeup alkali solution is supplied by a 500-gal elevated storage
tank. The alkali solution flows through an outlet on the tank to the
liquid-level-control solenoid valve and into the slurry tank by gravity.
The contents in the slurry tank are continuously pumped out during
cellulosic treatment by a 1/2-hp stainless steel centrifugal pump. From this
pump the slurry stream may take one of several routes (indicated as Routes
1, 2, and 3) through the rest of the cellulose treatment section (Figure 8).
Route Number 1. From the slurry tank pump, the slurry flows to the
solid-liquid separator (Figure 9) where the cellulose is partially
de-watered. It consists of a pair of rubber-coated squeeze rollers, between
which a 10-in.-wide monel screen belt passes; an idler roller that holds
tension on the belt; and a catch pan that catches the alkali solution
squeezed from the cellulose by the squeeze rollers (Figure 9). The tension
on the belt is adjustable by increasing or decreasing the spring pressure on
the idler roller shaft. The pressure between the squeeze rollers is also
adjustable by varying the spring pressure on the shaft of the top roller.
The alkali solution is recycled to the slurry tank by way of the catch pan
and gravity flow. The separator is also equipped with a spring-loaded
scraper that removes the de-^watered cellulose from the belt. The lower
squeeze roller is driven by the drive system associated with the oxidation
oven through a chain and sprocket system. All parts of the separator are
stainless steel with the exception of the monel screen belt.
The de-watered cellulose drops from the separator scraper onto the
oxidation oven belt (Figure 10). The cellulose is carried through the oven
by means of a 12-in.^wide monel screen belt. The belt rides on a drive
roller and an idler roller. The drive roller is connected to a 1/6-hp DC
variable-speed, electric motor through a gear reduction box and a chain and
-------
Ground
Cellulose
Alkali
Solution
Solid/Liquid
Separator
•••.
Sterilization
Figure 8. Cellulose treatment section (numbered arrows indicate alternate routes).
UJ
-------
SOLID-LIQUI1
SEPARATOR
LIQUID RECYCLI
LINE
Figure 9. Solid-liquid separator,
-------
Cellulose Slurry
Feed
Variable
Spring Pressure
Variable
Speed Drive
Radiant Heater
Banks
Alkali Solution
Recycle
Treated Cellulose
Figure 10. Solid-liquid separator and oxidation oven.
OJ
00
-------
39
sprocket. Heat for the oven is provided by three banks of three infrared
strip heaters (Figures 10 and 11), which are temperature controlled by time
cycling each bank of heaters independently. The heaters in bank one are
rated at 5,400 watts. Heaters in banks two and three are rated at 3,300 and
2,400 watts, respectively. The oven is also equipped with an air-sparging
system. Air is sparged onto the cellulose from under the screen belt by
perforated tubing, and* flow is regulated by a rotameter. The oven is
equipped with a control panel from which the heater temperatures, the
residence time (2 to 12 min) of the cellulose in the oven (belt speed), and
the aeration rate are controlled (Figure 12)- A scraper is used to remove
the cellulose from the belt as was done in the separator. All parts of the
oven contacting the cellulose are stainless steel.
As the cellulose continues along Route 1 of the treatment section, it
falls from the oven belt scraper into the reslurry tank where it is mixed
with de-ionized water and appropriate nutrient salts. This is a 35-gal,
cylindrical, steam-jacketed stainless steel vessel (Figure 13). It has an
automatic liquid-level control device that adds de-ionized water to a
constant level and is agitated by a 1/8-hp, single-propeller agitator. It
is temperature controlled by regulating the steam pressure in the steam
jacket and is insulated to minimize heat losses. This reslurry operation is
the final step in the treatment section. The slurry in this vessel is pumped
out a 1-in. bottom-centered outlet by the main feed pump and enters the
sterilization section.
Route Number 2. This treatment method uses the same equipment that
Route 1 uses except for the reslurry tank. Route 1 requires a very slow
feed from the slurry tank to the solid-liquid separator. In Route 2, the
alkali treatment in the slurry tank and the oxidation treatment in the oven
-------
OVEN HEATER ELEMENTS
RE-SLURRY TANK
Figure 11. Infrared oven-heating elements
-------
INFRARED HEATER
BANK CONTROLS
BELT SPEED
CONTROL
Figure 12. Infrared oven-control panel.
-------
42
s'UTRIENT SALTS
?EED LINE
Figure 13. Reslurry tank.
-------
43
are done at a high feed rate, and the treated cellulose is manually
reslurried in the large feed storage tank. This tank is a 500-gal,
cylindrical, stainless steel tank with a conical shaped bottom. A large
quantity of makeup feed slurry could be prepared in this tank which
permitted continuous feeding for many hours without continuous operation of
the cellulose-handling and treatment sections. The main feed pump can be
supplied directly from this tank instead of the small reslurry tank, as was
given in Route 1. A 1/2-hp, single-propeller agitator is used to agitate
the feed storage tank.
Route Number 3. The final optional cellulose treatment route is
indicated as Route 3 (Figure 8)- This treatment method is used when it is
not necessary to treat the cellulose in the oxidation oven. The slurry from
the slurry tank is fed through a 40-mesh screen filter to remove the
cellulose from the alkali solution. The cellulose is manually reslurried
in the feed storage tank as in Route 2. The alkali solution is collected
in a 2-ft-wide by 3-ft-long by 2-ft-high stainless steel tank. The screen
filter fits over the tank and can easily be removed for cleaning. The tank
and filter system is mounted, on caster wheels and can easily be moved about
the pilot plant for other filtering uses. The excess alkali solution is
recycled to the slurry tank.
Sterilization Section
The makeup feed slurry in either the reslurry tank or the feed storage
tank is pumped continuously into the steam injector system by the main feed
pump. This a Lapp Pulsafeeder diaphragm metering pump (Figure 14) and can
be set for flow rates ranging from 0 to 2.2 gpm. The fermenter residence
time is determined by the feeding rate chosen. The pump can handle slurry
-------
TEMPERATURE
CONTROL VALVE
LIQUID-STEAM H
MIXER .
I FEED STREAM
1 FLOW SHOWN
BY ARROWS
^RECYCLE
PUMP
Figure 14. Steam injector.
-------
45
densities of up to about 5 percent cellulosic solids by weight without
clogging. Care must be taken to prevent clogging of the pump upon starting
and stopping the slurry flow through it. Power for the pump is supplied
by a 1-hp electric motor.
The main feed pump feeds the slurry into the steam injection system
where the feed stream is brought up to sterilization temperature and
pressure. The steam injection system (schematic in Figure 15) consists of
a temperature-controlled steam injector, a recirculating liquid-steam mixer,
and a support stand.
The steam injector injects steam into the liquid-steam mixer loop where
it is mixed homogeneously with the feed stream to produce a uniform stream
temperature. Steam is metered into the system through an automatic valve
that is controlled by a temperature probe located downstream of the steam
injector. This valve can be set to produce a feed stream temperature
ranging from 260 F to 320 F. There is also a bypass valve that permits
manual injection of steam. The inlet steam line is equipped with the
necessary filter, pressure relief valve, and trap; a check valve is located
between the automatic valve and the steam injector.
The liquid-steam mixer is a recirculation loop constructed of a 1/2-in.
stainless steel pipe. The feed stream is pumped through this loop by a
1/2-hp centrifugal pump. The outlet of the liquid-steam mixer is equipped
with a 0 to 100 psi bourdon pressure gauge, a copper-constantan thermocouple
connected to the 24-point multipoint recorder on the fermenter control
panel, a pressure relief valve set at 100 psi, and a 0 to 400 F dial
indicating thermometer.
The support stand for the injection system is of carbon steel, unistrut
construction. All parts that contact the feed stream are stainless steel.
-------
Main
feed
pump
Slurry
feed
/I
Ns.
.^Jemperoture
control
valve
Filter
Water
Pressure
T.R.-
7T-JUJ
ionl
Recycle
pump
o
Temp.
Feed
stream
Figure 15. Steam injection heater.
-P-
CTi
-------
47
The hot feed stream flows into the sterilizing holding section where
the stream is routed in turbulent plug flow through a variable number of
insulated tubes (Figure 16)- There are four banks of tubes through which
the hot feed stream may be routed. These banks provide for four variations
in the residence time in the holding section so that a definite
temperature-time sterilization sequence can be effected. This variation
in residence time is accomplished by the manual switching of four 1-in.
three-way plug valves located at the beginning and end of each tube bank.
The first tube bank is composed of 22 stainless steel tubes of 1/2-in.
diameter that are 10-ft long. Banks two, three, and four each have ten
1-in.-diameter, 10-ft-long stainless steel tubes. All tubing is surrounded
by insulation. The outlet of the holding section is equipped with a
diaphragm-protected, 100 psi bourdon pressure gauge and a copper-constantan
thermocouple attached to the 24-point multipoint temperature recorder.
The final step in the sterilization process is the cooling of: the feed
stream to a temperature compatible with the contents of the fermenter. This
is accomplished in two steps by an evaporative cooler and by a counterflow,
double-pipe, chilled-water cooler, respectively (Figures 16 and 17).
The evaporative cooler consists of a coil of 1/2-in. stainless steel
tubing positioned in a standard 55-gal drum with a 1/2-in. tube spray ring
over the coil (Figure 16)- The coil is 18 in. in diameter and contains 10
turns of tubing. The feed stream flows through the coil, and water is
trickled over the coils by the spray ring. The drum collects the excess
water that flows by gravity to the drain. At low feeding rates, this cooler
drops the temperature of the feed stream to approximately 100 F, and the
chilled-water cooler is not needed.
The counterflow, double-pipe, chilled-water cooler (Figure 17) is
composed of seven 6-ft sections of jacketed tubes attached in series. The
-------
FEED STREAM FLOW
SHOWN BY ARROWS
PLUG-FLOW
ADJUSTABLE LENGTH
HOLDING SECTION
COOLING WATER
SPRAY RING
EVAPORATIVI
COOLER
Figure 16. Steam injector, holding section, and evaporative cooler.
-------
49
i
FEED STREAM FLOW
SHOWN BY ARROWS
CHILLED WATER IS IN
COUNTERFLOW PATTERN
IN JACKETS
BACK PRESSURE
CONTROL VALVE
Figure 17. Chilled^water heat exchanger,
-------
50
inner tubes are 1/2-in. stainless steel and the outer tubes 1 1/4-in. carbon
steel pipe. The feed stream flows through the inner tubes while the 50 F
chilled water flows countercurrently through the outer jacket. The flow
rate of the chilled water is regulated so that the outlet temperature of the
feed stream is equal to that of the fermenter. The seven jacketed tubes are
supported on a unistrut frame. Maximum design heat load is limited by the
maximum heat load on the available chilled-water system and is 90,000 BTU
per hour. The inlet and outlet temperatures are monitored by thermocouples
connected to the multipoint recorder.
Fermentation Section
The feed stream flows from the chilled-water cooler and the
sterilization section through a back-pressure control valve into the
fermenter. The feed stream can be recycled to either the reslurry tank or
the feed storage tank by a manual valving operation. This permits starting
and stopping of feed to the fermenter without stopping and starting any of
the feed stream preparation equipment.
The fermenter (Figure 18) is a 150-gal, jacketed and insulated vessel.
The inner wall of the fermenter is stainless steel sheet rolled into a
23 1/2-in. inside diameter (ID) by 7-ft-long cylinder. Slip-on flanges are
used to seal the ends. The outer jacket is stainless steel sheet rolled
into a 26 1/4-in. ID cylinder. A 1-in. layer of insulation surrounds the
outer jacket. The top blind flange is fitted with an 8-in. weld neck flange
that serves as the mounting flange for the agitator. It is also equipped
with a 4- by 6-in. hand hole and cover, a 1-in. pipe feed stream inlet and
*
inoculation port, and a 1-in. pipe air vent outlet. The hand hole is used
in the initial filling of the fermenter and for cleaning purposes. The
-------
Figure 18. Fermenter,
-------
52
bottom blind flange is fitted with a 2-in. weld neck flange and a 1/2-in.
pipe air inlet port. The weld neck flange is the outlet for fermented
media. There is a 3/4-in. pipe inlet and an outlet at the top and bottom of
the outer jacket. These are used for steam heating of the fermenter
contents for initial autoclaving of the fermenter and for temperature
control of the fermenter during a fermentation run. There are two
5-in.-diameter view ports near the top of the fermenter internals. At the
same level as the center of the view ports, there is a 1-in. pipe inlet
through the inner wall of the fermenter that is used for the liquid-level
control probe. Opposite this probe is a 2-in. flanged port that is not used
at present but was designed into the fermenter for possible future operation
of the fermenter in series or parallel with a second fermenter. At
mid-height on the fermenter is a 3-in. flanged outlet that is used as a
sampling port and a temperature probe inlet. Four small tabs are welded on
the inside of the fermenter for attaching internal baffles or draft tubes
(Figure 19). Design pressure of the fermenter is 150 psi.
Mixing of the fermenter media is accomplished in one of two ways.
Initially, a draft tube was used. This was simply a long cylindrical tube
suspended vertically in the fermenter by the attachment tabs. A draft flow
is caused by sparging air around the bottom and outside of the draft tube.
The air sparger is a ring of 1-in.-diameter tubing attached to the air inlet
on the bottom blind flange and is concentric with the draft tube and the
walls of the fermenter. Sixteen air injection nozzles with 0.063-in.
orifices are positioned on the tube ring.
The agitator is a 3-hp, variable-speed drive with a double mechanical
*
seal, two flat blade turbines, and one pumping turbine at the bottom of the
shaft. The rpm ranges from 117 to 300. A steady bearing was attached to
the bottom blind flange to support the bottom end of the agitator shaft.
-------
53
Jacket
Water "
Out
Jacket
Water-
In
Feed
In
Air
Out
Vi V
5
t
^a
«
0»
fiS B
O
0
O
0
R SSA-
3
0
•o
•0
\\ ^"\V"\V \V"\VH»^\v"AV \\'"^
\\
! 1
Cul ture
Out
Air
In
.Sample Port
. Temperature
Probe
'Dissolved
Oxygen
Probe
Figure 19. Pilot-plant fermenter.
-------
54
The fermenter is equipped with various controlling and monitoring
instrumentation (Figures 20 and 21). The temperature of the media is
recorded and controlled by a Honeywell temperature recorder and controller
that activates a three-way proportioning valve, which mixes hot and cold
water flowing through the outer jacket of the fermenter. The fermenter
temperature is also recorded on the multipoint recorder.
The pH of the media is recorded by a strip chart recorder and
controlled by a pH meter and controller. The pH is adjusted up when the
controller opens a solenoid valve that allows anhydrous ammonia to flow into
the inlet air line and is adjusted down when the controller activates a pump
that feeds hydrochloric acid into the feed inlet. The flow of anhydrous
ammonia is indicated by a rotameter.
The amount of air being sparged into the fermenter media is indicated
and controlled by an automatic rotameter. The rotameter regulates a
reverse-acting pneumatic control valve. The flow rate range is from 0 to
115 scfm at 0 psig and 60 F. The automatic rotameter is on the outlet air
line. The inlet air is passed through one of two air filters in parallel.
Only one filter is used at a time. Each filter is a jacketed and insulated
tube. Steam is passed through the outer jacket to keep the filter at
sterilization temperature. The inner jacket is packed with Dow-Corning
extra-fine-tempered glass wool. Air flows into the inner jacket through the
top and out the bottom.
The degree of agitation is indicated by a tachometer and is manually
regulated by a hand crank on the agitator.
The fermenter is equipped with a sampling system that, through the use
of a positive-displacement sampling pump, provides a sterile seal during
sampling. A sampling probe extends into the fermenter approximately 6 in.
-------
FERMENTER TEMPERATUH
RECORDER-CONTROLLER
FERMENTER pH RECORDER-
CONTROLLER
ULTRAVIOLET-VISIBLE
SPECTROPHOTOMETER
ANHYDROUS AMMONIA
CYLINDERS
INLET AD
TERS
Figure 20. Control panel—fermenter.
l/i
Ln
-------
FERMENTER AUTOMATIC
LEVEL CONTROL
INLET AIR LINE PRESSURE
mmmmmmmm
FILTERED AIR PRESSURE
STEAM PRESSURE
F ER ME NTERPRESSUR E
ROTAMETER
AIR FLOW RATE
AUTOMATIC
ROTAMETER
UR SAMPLING ASSEMBLY
Figure 21. Control panel—fennenter,
~
-------
57
through the sampling port flange. The probe is wrapped with a monel screen
shield that helps to prevent clogging of the probe by cellulosic material.
The sample of medium is pumped from the fermenter by a nutating disk
peristaltic pump. The sample medium is then passed over the pH sensing
probe and finally collected for laboratory investigation.
An automatic liquid-level controller on the fermenter maintains a
working volume of 140 gal during continuous runs. As the liquid level
reaches a predetermined level, the controller opens a pneumatic canister
valve on the fermenter outlet line and allows the fermenter media to flow
into the harvesting section of the plant.
Power to pH controlling pump, solenoid valves, recorders, and control
equipment totals about 1 hp. All material contacting the media in the
fermentation section is stainless steel.
Harvesting Section
Several methods are currently used to harvest the cells from the
fermented medium. Figure 22 shows the first of these schematically. The
medium is dumped into the first 500-gal mixer/settler tank (Figure 23),
where the unused cellulose is settled out and drawn off as an underflow
stream. The mixer/settler has four outlets at various levels on the tank
wall. The residence time of the medium in the settling tank is varied by
choosing a particular overflow outlet.
The overflow from the first tank then goes into the pH adjustment tank
(Figure 23) where hydrochloric acid is pumped in for pH adjustment. The pH
is recorded and controlled by a strip chart recorder/controller. The tank
is agitated by a propeller-type agitator.
The overflow from the pH adjustment tank goes into the second
mixer/settler tank. This tank is essentially like the first mixer/settler
-------
Culture
Media
Floccu lent
Mixer/Settler No.l
Mixer/Settler No. 2
Undigested
Cellulose
Flocculated
Cells
Figure. 22. Cellulose and cell concentration.
00
-------
MIXER/SETTLER NO. 1
MIXER/SETTLER NO. 2
pH ADJUSTMENT TANK
Figure 23. Mixer/settlers and flocculent tank.
-------
60
except that the overflow outlets are on both sides of the tank and the legs
are a few inches shorter. The acid causes the cells to flocculate and
precipitate to the bottom of the tank. The overflow from this tank goes to
the drain or to reuse. The cells are taken off the bottom and are either
centrifuged in the Sharpies Super centrifuge or drum dried on a steam-heated
double-drum drier.
-------
PILOT-UNIT OPERATION
The pilot unit could be run as either a batch process or a
continuous-flow process. Batch operation required operation of the
cellulose-treating equipment until enough treated solids were accumulated
to charge the fermenter. The fermenter was then loaded with medium,
sterilized, inoculated, allowed to operate for a period of time, and dumped.
Cells were harvested by a batch precipitation or by centrifugation.
For continuous operation of the plant, a culture medium was prepared
and grown as already described to a point within the logarithmic phase of
culture growth. The addition of a continuous-feed stream was started, and
the volume of culture medium was regulated by periodic culture withdrawals.
The volume of the culture medium in the fermenter was never allowed to
vary more than 1 percent above or below the standard working volume. Culture
withdrawal cycles were rapid enough and withdrawal volumes were small enough
to permit a true continuous-flow scheme of operation. Cells were harvested
from the exit stream either by direct centrifugation or by continuous
precipitation.
Size Reduction of Solids
The primary purpose of the size reduction step was to make the rough
agricultural wastes more homogeneous in size. The mill run bagasse was
extremely hard to meter, flow, pump, or mechanically handle in any way
without some size reduction. When cut through a 1/8-in. nominal screen,
the material could be hoppered and fed without undue problems.
61
-------
62
It has been reported that there is a correlation between the size of
Q O
cellulose particles and rate of enzymatic attack. The effects of particle
size do not, however, become apparent until the cellulose is reduced to
100-mesh size or smaller. It is prohibitively expensive to reduce particle
size to this level on an industrial scale. A 1/8-in. screen was, therefore,
chosen because it would produce a material that could easily be handled and
permitted an economical grinding operation.
A rotary knife grinder was chosen as the size reducer primarily because
of power costs. A hammer mill has a lower initial cost but uses two to
three times as much horsepower per unit of capacity. The knife cutter
produces a clean and homogeneous output and is more dust free than an
attrition mill such as a hammer mill or shredder.
Solid.sj_ JDry_.. Han_d_lJ.n.g.
Cellulose particles are fibrous and not free flowing. Their usual bulk
density is from 5 to 8 Ib per cu ft (at 10 percent moisture). They have a
high and irregular angle of repose and tend to bridge and arch badly,
especially when their moisture content is more than 15 percent. The solids
are noncompressible, and wh,en dry, are rather abrasive to high-speed grinders,
The pilot plant uses a pneumatic conveying system to transfer solids
from the grinder to the hopper. Solids are collected in a cyclone and
deposited into a vibratory hopper. The vibratory action of the hopper
ensures a constant feed to the vibratory metering screw feeder. The
vibratory 2-in. screw feeds a continuous stream of relatively constant-mass
flow rate to the slurry tank.
When the cellulosic solids become moist or damp, it is extremely
difficult to avoid bridging in hoppers and almost impossible to prevent
-------
63
jamming in screw flights or Moyno pumps. The pilot plant's splids-handling
system was designed to avoid metering streams that were not either dry
solids or light water slurry.
Alkali-oxidation Treatment
Cellulosic solids were contacted with alkali, de-watered, and either
contacted with air in an oven or used directly after de-watering.
Solids flowing from the feeder were fed into the alkali slurry tank
where they were slurried with sodium hydroxide of from 2 to 4 percent
concentration. Trace amounts of an oxidation catalyst (cobaltous chloride)
were sometimes added to the slurry. Solids density in the slurry tank was
usually held between 4 and 8 percent solids by weight. The tank was
agitated, and the temperature was varied from ambient to 160 F. Residence
times of the solids in the slurry tank were varied from 30 min to more than
1 hr.
The slurry was pumped continuously from the slurry tank to a
de-watering step. If the solids were to receive further oxidative
treatment, they were passed through the de-watering squeeze rollers on the
oxidation oven. The liquid was returned to the slurry tank. Solids exiting
the squeeze rolls retained about 60 percent moisture. These solids were
passed through the oven where they were heated and contacted with air.
Solids exiting the oven contained from 20 to 40 percent moisture, depending
on the severity of the oxidation treatment. Temperature, airflow rate, and
residence times in the oven were variable. Surface temperature of the
radiant heating elements in the oven was varied between 600 F and 900 F, and
residence time of solids in the oven set at from 2 to 6 min. The extent pf
the oxidation reaction could be controlled in this way, which determined the
fraction of cellulose and other carbohydrates degraded to water-soluble
-------
64
products. The cellulosic solids coming out of the oven were dropped in the
reslurry or feed stream make up tank. Necessary inorganic salts, antifoam
agents, and special nutrients were mixed with the cellulose and make up
water in the reslurry tank to prepare the complete fermenter feed stream.
Solids not oxidized in the oven were pumped from the slurry tank into
a batch de-watering screen filter. Alkaline liquid was removed and the
solids were washed with water to obtain a feed material x^ith a predetermined
level of water-soluble carbohydrates. This material was prepared for
fermenter feed by batch mixing with inorganic salts, antifoam agents, special
nutrients, and water in the 500-gal feed storage tank.
Alkali treatment of sugarcane bagasse is believed to cause
depolymerization, decrystallization, delignification, and swelling of the
cellulose fiber and thus increases the digestibility of the cellulose by
microorganisms. The digestibility of alkali-treated cellulose was reported
to be markedly different between different kinds of cellulosics. The effect
was said to be considerably more pronounced with hardwoods than with
34 35
softwoods. ' Thus, the optimum conditions of alkali treatment should be
established for each substrate used. Since the direct concern of the
contract was the maximum production of SCP, the effect of the severity of
alkali treatment on the growth of Cellulomonas has been evaluated.
Finely ground bagasse obtained from the vibratory live-bin hopper in
the pilot plant was mixed into various concentrations of alkaline solution
(50 g of bagasse per liter of solution) and kept at room temperature with
frequent agitation. A portion of the slurry was withdrawn at frequent time
intervals and quickly neutralized with hydrochloric acid. The excess liquid
*
was then removed from the cellulose by squeezing the slurry in cheese cloth.
One percent wet weight of the treated bagasse was used as a substrate with
the basal media already described.
-------
65
The various media were then inoculated with an equal amount of actively
growing Cellulomonas culture and incubated at 30 C on a rotary shaker. The
growth rate was determined by measuring the turbidity with a Klett
colorimeter. The kinetics of cell growth on the bagasse treated with
different concentrations of alkali and different lengths of time showed that
alkali treatment of bagasse significantly increased the growth rate and the
maximum cell density of the organism. When all bagasse samples were treated
for 5 min, the growth rate and maximum cell density were increased
proportionally with the increase of alkali concentration (Figure 24).
When the bagasse samples were treated for 2 hr with the same
concentrations of alkali, no significant difference was observed among the
alkali-treated samples within the range of 0.5 to 10 percent sodium
hydroxide concentrations; however, a highly significant difference was noted
between the treated samples and an untreated bagasse control (Figure 25).
Since alkali treatment depolymerizes the cellulose polymer, it was
suspected that the growth-promoting effect of the alkali treatment was to
provide more soluble carbohydrate, which is more readily utilizable by the
organism than the insoluble form of cellulose. It was shown that more
carbohydrate was indeed solubilized by the alkali treatment (Table 4);
however, the level of soluble carbohydrate in the final media was negligible
because all the soluble carbohydrate was washed from the bagasse before the
media were made. The increase in growth rate of the organism on
alkali-treated cellulose was not, therefore, due to providing soluble
carbohydrate but to some changes in the insoluble form of cellulose fiber.
This inference was confirmed by the result that alkali-treated and
thoroughly washed bagasse also supported the growth of the organism as well
as alkali-treated, unwashed bagasse.
-------
66
200
010%
x5%
x
in
"c
3
£ 100
in
C
0)
a
O
00.5%
20-
40 60
Time ( hours)
80
100
Figure 24. Growth of Cellulomonas on bagasse treated
for 5 min. in different alkali concentrations.
-------
67
200
X
I/I
'c
3
100
w>
C
-------
68
TABLE 4
AMOUNT OF CARBOHYDRATE SOLUBILIZED BY ALKALI TREATMENT
Treatment
(time, alkali cone, %)
5 min 0
0.5
1.0
5.0
10.0
30 min 0
0.5
1.0
5.0
10.0
2 hr 0
0.5
1.0
5.0
10.0
Soluble carbohydrate (mg/ml)
In the alkaline In the growth media
slurry (washed substrate)
0.105
0.340
0.360
0.550
0.480
0.140
0.370
0.460
0.690
0.630
0.105
0.480
0.500
0.615
0.580
0.150
0.008
0.009
0
0.008
0.160
0.008
0.008
0.008
0.008
0.120
0.008
0
0
0.008
-------
69
Media Composition
The pilot plant culture media were composed of the cellulose source,
water, inorganic nutrient salts, trace minerals, special nutrients, and
ant if cam agents. The original media used the same chemicals as were used
for the laboratory shake-flask cultures. The media were changed somewhat
for the pilot-plant runs three through five (Table 5).
In order to determine the substrate utilization characteristics of the
Cellulomonas organism, shake-flask cultures were prepared with nine
different carbon sources (Table 6). All the substrates, except the
substituted hydroxyethyl and methyl celluloses, supported growth, lactose
and glycerol being the best substrates (Figure 26).
Laboratory tests were run to seek optimum levels of nutrient
inorganics. Nine different nitrogen sources were used to find which best
supported cell growth. Ammonium bisulfate and ammonium bicarbonate
performed better than the rest, and sodium nitrate and ammonium acetate
supported no growth at all (Figure 27).
Inorganic nitrogen, in the form of ammonium sulfate, was used at
different levels in shake flasks to determine optimum nitrogen level (Table
7). The test showed that the nitrogen level had little effect on initial
growth. Ultimately, of course, nitrogen level would cause growth limitation
and would have to be maintained at a certain level.
The effects of the level of inorganic phosphate was also checked (Table
8). Optimum levels were seen to be from 0.04 to 0.08 percent phosphorus.
In addition to effects on cell density levels, cell growth rate was also
affected (Figure 28).
In the original nutrient media used in the isolation of the organism,
sodium chloride was included at a level of 6.0 g per liter. Tests have
-------
70
TABLE 5
PILOT-PLANT MEDIA COMPOSITION FOR RUNS THREE, FOUR AND FIVE
Component
Amount for 1 liter (g)
Substrate: Treated bagasse (dry weight)
Nutrients: Ammonium sulfate
Potassium phosphate (dibasic)
Potassium phosphate (monobasic)
Magnesium sulfate
Calcium chloride
Sodium chloride
Other: Yeast extract
Trace minerals*
Polyglycol P-2000
Water
*Trace minerals composition (g/liter)
Calcium chloride
Ferric chloride, 6HLO
Zinc sulfate, 7H20
Copper sulfate, 5H_0
Cobaltous chloride, 6H20
Ethylenedinitrilotetraacetic acid
6.0
3.0
0.5
0.5
0.1
0.1
3.0
0.3
1.0 ml
0.1 ml
To 1.0 liter
0.5
16.7
0.18
0.16
0.18
20.1
-------
71
TABLE 6
EFFECT OF DIFFERENT CARBON SOURCES ON THE
GROWTH OF CELLULOMONAS
Carbon source Optical density
(10 g/liter) (Klett unit)
Glycerol
Glucose
Galactose
Cellobiose
Maltose
Lactose
CMC
Methyl cellulose
Hydroxyethyl cellulose
335
250
270
195
142
340
227
40
65
Cell yield3
(dry wt)
(g/liter)
0.436
0.325
0.351
0.254
0.185
0.442
0.290
0.052
0.085
*Cellulomonas grown in basal media for 96 hr in shake tubes at 30 C.
shown that less sodium chloride will give much the same growth response
(Table 9 and Figure 29). Total exclusion of sodium chloride in pilot-unit
ferraenter runs produced, however, low cell growth and slow growth rates.
The level of trace mineral solution in the culture media was tested,
and optimum level was found to be from 0.5 to 1.0 ml per liter of media
(Table 10 and Figure 30).
Since it would not prove industrially feasible to use the
laboratory-grade chemicals that had been used in the pilot-plant and
laboratory-fermentations, various industrial-grade and fertilizer-grade
chemicals were tested to find less expensive replacements for the nitrogen
-------
72
340
300
260
20
40 60
Time (hours)
80
Lactose
Glycerol
Galactose
Glucose
CMC
-Cellobiose
Maltose
Hydroxyethyl
eellulose
Methyl
cellulose
100
Figure 26. Effect of different carbon sources on the
growth of Cellulomonas.
-------
73
220
180
c
3
JS> 140
100
o
V
^
Q.
O
60
20
20 40 60
Time (hours)
80
100
1. AMMONIUM BISULFATE 6.
2. AMMONIUM BICARBONATE 7-
3. AMMONIUM CHLORIDE 8.
4. AMMONIUM NITRATE 9.
5. AMMONIUM SULFITE
AMMONIUM SULFATE
UREA
SODIUM NITRATE
AMMONIUM ACETATE
Figure 27. Effect of different nitrogen sources on the
growth of Cellulomonas.
-------
74
TABLE 7
EFFECT OF NITROGEN LEVEL ON THE GROWTH OF CELLULOMONAS
Nitrogen level
[%(NH4)2S04]
0
0.1
0.3
0.6
1.0
N/P
o-°-125
0.3 „,
0.8 * '
fct-'-"
1*0 _ i oc
0.8 1<25
Optical density
(Klett units)
0
260
290
292
272
Cell yield3
(dry wt)
(g/liter)
0
0.339
0.378
0.380
0.354
aCellulomonas grown in basal media plus various levels of nitrogen
for 92 hr in shake tube at 30 C. Basal media contained CMC 10 g/liter;
K2HP04 2.0 g/liter; KH2P04 2.0 g/liter; mgS04 0.1 g/liter; Ca C12 0.1
g/liter; yeast extract 0.5 g/liter.
and phosphorus sources. Several fertilizer-grade chemicals were tried as
nutrient replacements with varying degrees of success (Table 11).
From these tests media were developed that used some of the
fertilizer-grade and some of the laboratory-grade chemicals. A
nitrogen-to-phosphorus-to-potassium (N-P-K) ratio of about 5 to 1 to 0.1 was
maintained. These media were used in pilot-plant runs 6 through 15 (Table
12).
Fermenter and Feed Stream Sterilization
*
Before the start of each run the fermenter was filled with media and
sterilized. All air and feed inlet and outlet lines, sampling lines and
valves, inoculating ports, and acid and base inlet ports were sterilized at
-------
75
TABLE 8
EFFECT OF PHOSPHATE LEVEL ON THE GROWTH OF CELLULOMONAS
Phosphate level
(% phosphate)
0
0.004
0.01
0.02
0.04
0.08
0.2
N/P Optical density
(Klett units)
40
15.7 100
6.3 200
3.1 220
1.6 265
0.8 265
0.31 200
Cell yield3
(dry wt)
(g/liter)
0.052
0.13
0.26
0.286
0.345
0.345
0.26
Cellulomonas grown in basal media plus various levels of phosphate
for 48 hr in shake tube at 30 C.
the same time. Steam at 40 psig was used for sterilization of all lines,
valves, and filters and was used in the fermenter jacket to heat the culture
media. Initial sterilization temperature curves for the fermenter and all
lines, filters, and ports show the severity of initial sterilization
techniques (Figure 31)•
When the cell density in the fermenter had reached the value desired,
the continuous-feed sterilization section was sterilized and continuous-feed
sterilization started.
The main feed pump and the stream injector recycle pump were started,
and steam was fed into the injector. During this time, the pumped liquid
stream was flowing through the sterilizing holding section, the evaporative
-------
76
.08%
.04%
-.004%
0%
10 20 30
Time (hours)
40
50
Figure 28. Effect of phosphate level on the growth of
Cellulomonas.
-------
77
TABLE 9
EFFECT OF SODIUM CHLORIDE LEVEL ON THE GROWTH OF CELLULOMONAS
NaCl level
(% NaCl)
0
0.1
0.3
0.6
1.0
Klett unit
125
135
145
100
60
Cell yield3
(dry wt)
(g/liter)
0.104
0.175
0.190
0.130
0.078
aCellulomonas grown in basal media for 24 hr on shake tube at 30 C,
cooler, and the chilled-water heat exchanger and was recycled to the
reslurry tank or bulk feed tank.
The steam injector temperature control valve was set to maintain a
pre-chosen temperature (usually 300 F) for liquid exiting the injector.
This hot stream was flowed through the rest of the sterilizer system and
recycled to the reslurry or feed tank. The sterilizing holding section,
evaporative cooler, and chilled-water heat exchanger were heated to about
250 F for from 30 min to 1 hr before the coolers were started. After the
heat exchangers and feed piping were sterilized, the cooling water was
turned on in the evaporative cooler and the chilled-water heat exchanger.
The temperature of the feed stream was then adjusted and allowed to reach
equilibrium in all parts of the sterilizing-cooling section. Equilibrium
temperatures were maintained in all parts of the continuous-feed
sterilization section during a continuous fermentation (Figure 32).
-------
78
260
220 -
(0.3%)
(0.1%)
( 0% )
(0.6%)
(1.0%)
20
40 60
Time (hours)
80
100
Figure 29. Effect of sodium chloride level on the growth
of Cellulomonas.
-------
79
TABLE 10
EFFECT OF TRACE MINERAL8 LEVEL ON THE GROWTH OF CELLULOMONAS
Trace mineral
solution
(ml/liter)
0
0.1
0.5
1.0
5.0
10.0
'StLneral
a Klett
unit
216
219
228
230
210
26
solution contains (g/liter) :
CaCl2 0.5
FeCl3 . 6H20 0.167
Cell yieldb
(dry wt)
(g/liter)
0.280
0.285
0.297
0.300
0.274
0.034
. 7H20 0.18
. 5H00 0.16
-------
80
260
220
180
140
Q>
100
c
Q)
TJ
S 60
20
1 ml/liter
0.5 ml/liter
0 ml/liter
0.1 ml/liter
5 ml/liter
10 ml/liter
20 40 60
Time (hours)
80
Figure 30. Effect of trace mineral level on the
growth, of Cellulomonas.
-------
TABLE 11
REPLACEMENT INORGANIC NUTRIENTS
81
Nitrogen sources
Cell density
Heavy Medium Light
Ammonium sulfate (NH,)~SO,-as used
in laboratory medium
Urea, industrial grade (Company A)
Urea, fertilizer grade (Company A)
Ammonium polyphosphate, TVA
fertilizer grade
Ammonium chloride, industrial
grade
Ammonium nitrate-urea,
fertilizer mix (Company B)
Ammonium polyphosphate, fertilizer
grade (Company B)
X
X
Work by Han and others has shown that the germicidal effect of sodium
hydroxide considerably enhances the kill rate of bacteria and spores in a
nf
heat sterilization.
For this experiment, a spore-forming bacterium was isolated from
sugarcane bagasse. The spore suspension was then subjected to various
combinations of time, temperature, and alkali concentration, and the rates
of destruction were determined for each set of combinations. A series of
survival curves and thermal-death time curves revealed a different mode of
death between death by heat and by alkali. When alkali was incorporated
with heat, the death rates of bacterial spores were increased and the slope
of the thermal-death time curve was changed.
-------
82
TABLE 12
PILOT PLANT NUTRIENT MEDIA FOR RUNS 6 THROUGH 15
Component Amount (g per liter)
Substrate: treated bagasse or purified 5.0
ground wood pulp
Ammonium polyphosphate (15 to 27.1 to 0) 0.73
Ammonium chloride, industrial grade 3.4
Sodium chloride, industrial grade 3.0
Calcium sulfate (or calcium chloride) 0.1
Magnesium sulfate 0.1
Potassium phosphate, dibasic 0.75
a
Yeast extract or yeast lysate 0.2
Trace minerals solution, as in Table 5 1.0 ml
Polyglycol, P-2000 0.1 ml
Water to 1.0 liter
xeast lysate was prepared by the concentration and lysing of brewers
yeast obtained fresh from the Jackson Brewing Company, New Orleans, La.
From a series of thermal-destruction curves and alkaline-destruction
curves, an empirical equation expressing the relationship between the death
rate, temperature, and alkali concentration was established. The equation
expresses that the death rate of the bacterial spore is affected
exponentially by temperature and directly by alkali concentration. With the
equation, sterilization time for various combinations of heat and alkali was
determined, and the overall correlation index between the experimental data
and the computed value was 0.877.
-------
83
300
Fermenter
media
200
Q>
u
3
0)
a
E
a>
100
Air inlet and
outlet lines and
filters
1 2
Time ( hours)
Figure 31. Initial sterilization profile.
-------
80 F 300 F 260 F 120-180 F
90 F 93 F
Steam
Water
T
-e
Water
Reslurry Steam Holding Evaporative Chilled water
tank injector section cooler exchanger
Fermenter
Figure 32. Equilibrium continuous-sterilization temperatures.
oo
-------
85
Inoculation and Fermen^atjLpn
Cellulomonas, sp. bacteria were kept in pure slant cultures on nutrient
agar. From the test tube containing a slant culture of Cellulomonas, a
loopful of culture was transferred to a test tube containing sterile media
(basal media plus a strip of filter paper) and incubated for 2 days until
visual turbidity was observed. The actively growing culture was then
propagated up to 15 liters by using 5 to 10 percent inoculum volume for each
transfer step. Filter paper was used as a sole carbon source in early
stages of propagation, and alakali-treated bagasse or ground wood pulp
replaced filter paper in the 15-liter culture. All the flask cultures were
incubated on a rotary shaker at 30 C, while the 15-liter carboy was
incubated at room temperature and aerated with filtered air. The culture
was allowed to reach a cell density of from 300 to 500 Klett units in the
15-liter carboy.*
When the temperature of the fermenter had cooled to the proper value
(shown in Figure 31), the 15-liter inoculum was pumped in through the
previously sterilized inoculation port. All fermentation variables were set
at their proper control points, and batch growth of the organism was started.
The culture was agitated either by use of a draft tube or by a turbine
agitator and baffle system. The draft tube left about 30 percent dead
volume in the fermenter and was not as efficient in promoting oxygen
37
transfer and mixing as the mechanical agitator was. The mechanical
*A Klett unit is a reading of optical density in a Klett-Summerson
colorimeter; with Cellulomonas organisms, 1,000 Klett units equal a
bacterial-cell density of about 1.3 g per liter dry weight.
-------
86
agitator had essentially zero percent dead volume, promoted better oxygen
transfer, and kept the contents of the vessel more homogeneously mixed.
The agitator was usually run at 117 rpm, but subsequent runs have shown that
higher agitation speeds promote faster growth rates and better oxygen
transfer, and agitation speeds up to about 300 rpm have been used (Figure
33). The faster agitation rates, however, promote formation of a stable
foam that is hard to break by addition of antifoam agent.
Air used in fermenter aeration was taken directly from the 90 psi
missile-grade air system of the building. The air was filtered before
entering the fermenter in one of two parallel, steam-jacketed, glass fiber
depth filters. Aeration rate was set and automatically controlled by the
fermenter pressure-aeration rate control system. Aeration rate was usually
set at about 2 volumes of air at standard temperature and pressure per
volume of culture media per minute. This high rate was chosen to ensure a
sufficient oxygen supply to the culture.
The concentration of dissolved oxygen in a growing culture was
monitored in both draft tube-agitated and turbine-agitated cultures, and in
pure Cellulomonas and mixed cultures. The transfer of oxygen from gas to
liquid phase was directly effected by degree of agitation, and an increase
in agitator speed was usually followed by an increase in growth rate (Figure
33). At high cell densities and high growth rates the culture would be
limited by oxygen transfer if enough substrate was present. As substrate
was consumed, growth would become substrate limited, and dissolved-oxygen
concentration would gradually build up to saturation level.
The pH of the culture media was maintained between 6.5 and 6.7 during
a run by addition of either acid or base. During active fermentation fche pH
tended to drop owing to the fermentative production of by-product and
nucleic acids. During the growth phase only base, anhydrous ammonia, was
-------
bo
Optical density (Klett units)
O
O
LO
M
Mi
Hi
(D
O
rt
s-
TO
H-
TO
H-
§
3
Pi
rt
O
hh
n>
S
rt
O
0
O
c
0 "^
00 £,
10
O
Agitation at 140 rpm
Agitation ,
increased to
210 rpm
Agitation
increased to
275 rpm
oo
•vl
-------
88
added if the feed material was not too highly alkaline. If the process was
left untended for any length of time during active growth, the pH would fall
to 6.0 or lower, and fermentation would slow or cease.
If the pH rose to a value of 8.5 or 9.0 by addition of alkali, the acid
produced by the fermentation would bring it back down to the proper range,
but cellulase activity at such a high pH was not optimum.
The temperature of the culture media was automatically controlled and
recorded. The usual setting was between 91 F and 94 F. A temperature rise
to 104 F or 105 F did not hurt the growing culture, and a drop to 80 F or
85 F had only a slight slowing effect on the growth rate.
Fermenter pressure was controlled by the fermenter's air outlet valve.
The vessel was usually run with an internal pressure of 20 psig. This was
done to permit smooth, positive operation of the aeration control system and
to minimize chances of contamination through leakage around valves or shaft
seals. Higher pressures were used to increase dissolved oxygen in runs that
were oxygen limited.
Note here that no effort has been made to define the quantitative
effects of pH, aeration, agitation, pressure, or temperature. The settings
of these variables were chosen either by extrapolation from laboratory data
or by experience. Their quantitative effects should and will be defined,
but that work has not been completed.
The concentrations of cells and soluble carbohydrates were determined
by periodic sampling of the culture media. Cell concentration was reported
in Klett optical density units, and soluble carbohydrate was determined as
mg per liter. Cell and soluble-carbohydrate concentrations for a batch
fermentation where unwashed alkali-oxidation-treated sugarcane bagasse was
used as substrate are presented (Figure 34). The cell density curve for a
continuous run is also presented (Figure 35). Typical curves are presented
-------
89
1000
c
D
x 100
c
0)
TJ
0)
U
10
X X
*
/
1000
100
O)
E
0)
o
o
-0
o
u
_0
_Q
"5
to
10
10
20 30
Time (hours)
40
50
Figure 34. Cell density and soluble carbohydrate concentration
versus time for a batch fermentation.
-------
500- •
400- -
VI
'c
= 300
c 200
0
u
100-•
10
15 20
Time (hours)
35
Figure 35. Cell density versus time for a continuous fermentation.
-------
91
for the same variables in batch runs with unwashed, treated bagasse (Figure
36) and washed, treated bagasse or wood pulp (Figure 37).
Values for cell density versus time have been collected for all
pilot-plant batch and continuous runs. If the data from the batch runs are
analyzed by the method of Adams and Hungate, constants may be evaluated for
38
use in the continuous-flow process.
For evaluating the cell production potential for a particular batch
culture in a continuous-flow fermenter, the cell concentration (X) is
plotted versus the time (t) (Figure 38). Then the slope of the growth curve
dX/dt is taken at several points and plotted versus the cell concentration
(Figure 39). For finding the growth rate constant (k), the slope of a line
from the origin to any point on the dX/dt-versus-X curve is taken. In a
single stage, backmix fermenter k equals D (the dilution rate, hours ) if
the cell growth is substrate limited. The predicted equilibrium cell
concentration may be taken from the dX/dt-versus-X curve by dropping
perpendicularly to the ordinate X and reading the concentration. A curve of
all cell production (P) may then be plotted, which is the feed rate times
equilibrium cell concentration versus the feed rate, to find the optimum
operating range of the continuous fermenter (Figure 40). The maximum
volumetric production efficiency (VPE) of the fermenter may be calculated
as grams dry cells per liter per hour.
In continuous runs, the feed stream was started during the log phase
of growth (Figure 35). An arbitrary feed rate was chosen initially, and the
cell density usually rose or fell depending on the rate of feed. The feed
rate was adjusted to the desired value to permit the cell density to reach
equilibrium. Volumetric production efficiency was then determined and some
variable changed. The cell density was again allowed to reach equilibrium
-------
92
c
0)
u
C
o
u
O
\
0)
o
o
w
-O
_3
Q
Time,_t, (hours)
Figure 36. Fermentation of unwashed, treated bagasse.
t>0
0)
O
.
41
o c
O
10
Time,_L, (hours)
Figure 37. Fermentation of washed, treated bagasse
or purified wood pulp.
-------
93
c
o
V
u
C
o
4)
X+dX
t+dt
Time (hours)
Figure 38. Calculation of
dX
dt
values,
-------
94
dX
dt
Cell concentration
Figure 39. Calculation of k values.
o
in
E
o
i_
O)
c
o
TJ
O
O
U
Feed rate (liters/hour)
Figure 40. Determination of maximum cell
production and optimum feed rate.
-------
95
and the VPE again calculated. This procedure was continued to the end of
the run.
Data from all successful batch and continuous runs of the pilot plant
were collected and used for calculation of fermentation rates and
efficiencies (Table 13). Batch runs of from 24-hr to 182-hr duration were
made, and continuous flow was maintained for from 30 to 74 hr. Bagasse or
wood pulp was initially charged at from 5 to 10 g per liter of media, and
the continuous-feed streams held 5 g per liter. From 53 to 91 percent of
the bagasse fed was solubilized during the batch runs, the longer runs
generally consuming more of the bagasse. This was not strictly true in all
cases, and it is felt that the initial alkaline-oxidation treatment may
control the initial degree of bagasse solubilization. The more severely
treated bagasse dissolves (and is thus metabolized) more thoroughly than
bagasse with more mild treatment. Continuous-flow cultures could probably
be maintained at a utilization efficiency of from 50 to 70 percent,
depending on residence time.
Log-phase cell mass-doubling times ranged from 1.8 to 4.5 hr for
bagasse, and from 2.5 to 5 hr for purified wood pulp. Average cell
mass-doubling time for bagasse-grown cultures was about 3.6 hr. VPE values
were calculated from all batch-run data, and the values ranged from 0.02 to
0.162 for pure cultures and 0.512 for the symbiotic culture. Experimental
continuous VPE values were 0.033 and 0.098. This showed that about 1 g of
dry SCP could be produced for each 6 liters of pure Cellulomonas culture
volume every hour.
When unwashed, alkali-oxidation-treated bagasse was used as the
substrate in a run, the initial level of soluble carbohydrate was from 500
to 800 rag/liter. The level depended on the severity of the pre-fermentation
-------
TABLE 13 96
FERMENTATION BATCH AND CONTINUOUS RUN DATA
FOR 141-gal PILOT-PLANT FERMENTER
Run Number
4Ba
Length of run (hr) 30
Weight of initial bagasse
substrate (g dry weight) 3120
4Ca 5 6
74 27 24
5.0 3200 2650
Weight of bagasse recovered
after fermentation (g dry
weight)
Percent bagasse solubilized —
Weight of cells at end of
batch (g dry weight) 244
Log phase mass doubling
time, (t,-hr) 4.4
mo
Log-phase growth-rate
constant* (hr"1) 0.145
Maximum cell density (g
dry weight/liter) 0.47
Calculated volumetric^
production efficiency 0.0525
Actual volumetric production
efficiency (maximum)0 —
1500
53.0
340
1.8
0.24
0.17 0.64
0.070
0.033
633
76.0
259
3.7
0.189
0.49
0.0675
—
»s
B and C designations represent the respective batch and continuous
flow parts of the same run. Values given are for the culture in equilibrium.
Volumetric production efficiency is given as g of dry cell mass
produced per liter of culture media per hr. These values are calculated
from batch data.
Volumetric production efficiency values experimentally determined
from continuous run data.
-------
97
TABLE 13 (Continued)
FERMENTATION BATCH AND CONTINUOUS RUN DATA
FOR 141-gal PILOT-PLANT FERMENTER
Run Number
8 9
Actual volumetric production
efficiency (maximum)0
10
Length of run (hr)
Weight of initial bagasse
substrate (g dry weight)
Weight of bagasse recovered
after fermentation (g dry
weight)
Percent bagasse solubilized
Weight of cells at end of
batch (g dry weight)
Log-phase mass-doubling
time, (t,-hr)
ma
Log-phase growth-rate
constant^ (hr )
Maximum cell density (g
dry weight/liter)
Calculated volumetric^
production efficiency
117
2650
299
89.0
434
3.2
0.128
0.82
0.033
90 182
2650 2630
248
90.8
890 417
3.3 3.9
0.140
1.66 0.78
0.056
o
w
H
<
l-l
S3
<
H
*Z
U
Cellulosic weights are given as g initially charged in batch runs
and as grams per liter of feed in the continuous runs.
SPurified ground wood pulp.
Log-phase growth rate at maximum theoretical volumetric production
efficiency.
-------
98
TABLE 13 (Continued)
FERMENTATION BATCH AND CONTINUOUS RUN DATA
FOR 141-gal PILOT-PLANT FERMENTER
Run Number
11 12 13 14Ba
Length of run (hr) 48 114 46 21.5
Weight of initial bagasse
substrate (g dry weight)d 2650 2650 2650 2650
Weight of bagasse recovered
after fermentation (g dry
weight) — — — —
Percent bagasse solubilized
Weight of cells at end of
batch (g dry weight)
Log-phase mass-doubling
time, (t ,-hr)
ma
Log-phase growth-rate
constant (hr"1)
Maximum cell density (g
dry weight/liter)
Calculated volumetric
production efficiency
Actual volumetric production
efficiency (maximum)
410
3.0
0.114
0.77
0.055
—
446 710
5.0 4.0
0.075 0.0769
0.96 1.33
0.0197 0.056
—
539
3.6
0.1768
1.005
0.069
—
ff
BSymbiotic run with Alaaligenes faecalis and Cellulomonas.
-------
TABLE 13 (Continued)
FERMENTATION BATCH AND CONTINUOUS RUN DATA
FOR 141-gal PILOT-PLANT FERMENTER
99
14C'
Weight of bagasse recovered
after fermentation (g dry
weight)
Percent bagasse solubilized
Weight of cells at end of
batch (g dry weight)
Log-phase mass-doubling
time, (t ,-hr)
Log-phase growth-rate
constant (hr~^)
Maximum cell density (g
dry weight/liter)
Calculated volumetric,
production efficiency
Actual volumetric production
efficiency (maximum)c
0.52
Run Number
15 16
p
H
Z
<
H
&
O
0.098
17*
Length of run (hr)
Weight of initial bagasse
substrate (g dry weight)
28.5
5.0
32
5.0e
69.5
10.0
465 3340
3.9 4.5
0.20 0.0532
0.871 6.24
0.162 0.512
-------
100
treatment and on the concentration of bagasse fed. When washed bagasse
and wood pulp were used, the soluble-carbohydrate level was from 10 to
100 mg/liter.
Cellulomonas grew on all three substrates; however, substrate
utilization mechanisms seemed to be quite different. The
soluble-carbohydrate concentration of cultures with high inital values
fell quite rapidly during lag and initial logarithmic phases (Figure 34).
Both the cell mass increase and the soluble-carbohydrate decrease may be
determined quantitatively during this time. When both these values were
calculated over a discrete time period, as long as it was in the logarithmic
phase of cell growth, it was seen that the yield coefficient—grams of cell
mass increase per gram of soluble-carbohydrate decrease—was 0.5, or 50
percent. This is the accepted value for aerobic cell growth on a
carbohydrate substrate. This was interpreted to mean that if metabolizable
soluble carbohydrate is present in the media in sizable amounts, it will be
metabolized before the insoluble cellulose. This has, of course, been well
recognized previously and has been attributed to repression or inhibition of
cellulase enzymes by the soluble carbohydrate, particularly cellobiose.
When the soluble carbohydrate reached a low level in these batches, the
growth rate slowed considerably.
In batches where washed, treated bagasse was used, the initial
concentration of soluble carbohydrate ranged from less than 10 to 100
mg/liter. The cells grew and exhibited almost the same kinetics as when
grown with larger soluble-carbohydrate concentrations, and the concentration
of soluble carbohydrate did decrease somewhat. The soluble carbohydrate
*
no longer supplied, however, a sufficient quantity of substrate to
explain cell mass increase. The insoluble cellulose was, therefore,
being degraded and metabolized to some extent whenever initial soluble
-------
101
carbohydrates were low, and the transfer of insoluble cellulose to soluble
carbohydrate by enzyme action was appreciable.
In batches with pure wood pulp as the substrate, the initial
concentration of soluble carbohydrate was usually less than 50 mg/liter.
During the run, the concentration increased until the end of the log-growth
phase and then decreased (Figure 37). It never rose above about 150
mg/liter. In these cultures all the soluble substrate was being produced
by enzymatic breakdown of cellulose.
Y. W. Han has found that the severity of the alkaline-oxidation
treatment of bagasse directly affects the amount of carbohydrate that is
solubilized but, after a certain point, does not seem to increase the rate
o/r
or level of organism growth. These data indicate that the usefulness of
the thermal-oxidation step is minimal, and it is distinctly harmful if an
active cellulase enzyme system is desired. In other words, it may be best
to minimize oxidative breakdown of cellulose. A certain mild alkali
swelling treatment is, however, necessary since untreated bagasse is
metabolized at an extremely slow rate.
The rate of mechanical agitation was shown to have a definite effect
on the limit and rate of growth of the organism (Figure 33). In a batch
run the agitation rate was changed twice after cell growth started. Each
time the growth rate increased after an increase in agitation rate. After a
period of growth, the culture would begin to enter the stationary phase.
Another increase in agitation would initiate another cycle. It is presumed
that the increased agitation led to higher dissolved-oxygen transfer rates
and relieved the oxygen limitations of the culture.
-------
102
Harvesting
The SCP product and the by-product undigested cellulose were harvested
from both batch and continuous-flow cultures by one of three methods,
depending on the type product desired.
The cellulose waste that passed through the fermenter without being
solubilized had to be cleared from the product stream. This was done either
by direct filtration of the effluent stream by a 40-mesh screen filter or
by settling. All but the fine particles of fiber could be removed from the
stream with the filter. An even cleaner stream could, however, be obtained
by batch or continuous settling of the product stream. This was
accomplished in a variable volume, and thus, a variable residence time
mixer/settler. A residence time of from 2 to 6 hr was provided in this
vessel, and a cleared overflow stream was obtained. The undigested
cellulose could then be taken off as an underflow.
The overflow stream containing the cells was flowed either to the
centrifuge or to the flocculent addition tank and the second mixer/settler
unit. Cells were precipitated in the second mixer/settler and removed as an
underflow in a heavy cream. Organisms spun out of the media by
centrifugation were removed from the centrifuge as a heavy sludge.
The cell cream from the second mixer/settler could be drum dried or
freeze dried directly from the settler. The underflow stream contained from
4 to 5 percent solids by weight. When this stream was dried without further
cell concentration, however, a large concentration of salts was obtained in
the product. The salts came from the nutrient inorganics that passed
through the fermenter unchanged and from those that were generated in the
fermenter. Salt concentrations in the harvested dry cell product have run
as large as 25 percent.
-------
103
The cell cream from the precipitation step could be centrifuged prior
to drying to remove most of the slats. Centrifugation of the precipitated
cream was a much easier and more economical step than that of the total
cellulose-free media.
The dried SCP product contained from 50 to 60 percent crude protein
as determined by Kjeldahl analysis, and was quite low in fiber and lignin
(Table 14).
TABLE 14
39
PRODUCT ANALYSIS
Sample Selected component (% on a dry basis)
Protein Fat Fiber Ash Lignin ADFa
Untreated bagasse 2.92 1.87 40.6 4.68 7.6 53.5
Treated bagasse 1.66 1.59 39.2 23.4 3.07 46.1
Freeze-dried, centrifuged
cells
Cellulose remaining after
fermentation
57.8
7.7
2.53
2.67
2.53
68.0
9.0
2.88
1.37
7.6
3.7
74.0
3ADF is acid detergent fiber.
The cellulose remaining undigested or insoluble after the fermentation
contained a much larger relative fiber content than the unfermented bagasse.
It is probable that the alkali-oxidation treatment and the enzyme action in
the fermenter degraded and solubilized protein, fat, lignin, and
hemicelluloses, leaving a relatively larger fraction of insoluble fiber in
the recoverable effluent solids. This fiber was essentially de-pithed and
clean and had a higher relative crystallinity than the unfermented samples.
-------
PRODUCT QUALITY AND BY-PRODUCT USAGE
Cellulomonas sp bacteria were grown under controlled conditions in
the laboratory on a carboxymethyl cellulose substrate. An analysis of the
harvested cells showed a protein content of 46.2 percent and a nonproteinaceous
40
nitrogen level of 7.7 percent on a dry-weight basis (Table 15).
The essential amino acid content of the cell protein of the organism
was determined, and the values obtained were compared with the FAO
reference-protein values and with those of proteins from other plant and
animal sources (Table 16). Also included are the analyses of single-cell
proteins from petrochemicals. The essential amino acid pattern of the
Cellulomonas compares favorably with that of FAO reference protein. The
lysine content, which is deficient in a number of foods, particularly cereal
grains, was larger than that of the reference protein. The content of other
essential amino acids, such as leucine and valine, were extremely large when
compared with the proteins of other sources and the FAO reference protein.
The methionine content was comparable with that of wheat flour or
single-cell protein produced from hydrocarbon.
Feeding studies were conducted on male weanling rats of the
41
Sprague-Daw ley strain. It was found the Cellulomonas cells were superior
to Pseudomonas cells produced on hydrocarbons but were inferior to casein.
The rats held their weight on a diet with 20 percent protein supplied by
Cellulomonas and showed gains on a diet containing 40 percent Cellulomonas
protein. The cells were not toxic even when fed at the 80 percent level (40
104
-------
TABLE 15
GROWTH YIELDS OF CELLULOMONAS ON CARBOXYMETHYL CELLULOSE
105
Yields
(mg/mg) in 100 ml (%)
a b
Cell mass /CH^O consumed
Protein /cell mass
Nonprotein-N /cell mass
Protein/CH^O consumed
13.0/26.0
6.0/13.0
1.0/13.0
6.0/26.0
50.0
46.2
7.7
23.0
Cells grown 2 days on a basal medium containing 0.1% of CM-cellulose
were harvested by centrifugation. Cell crops were dried at 110 C to obtain
a constant weight.
Difference of CH^O concentrations in initial and final medium. CtLO
concentrations were measured by phenol sulfuric acid method.
CBy micro-Kjeldahl method after extracting nucleic acids with 5 percent
TCA at 90 C for 30 min.
dDifference of N content in whole cells and hot-TCA-treated cells.
percent crude protein). The addition of L-methionine improved the quality
of the protein considerably- Thus, it was believed that methionine is the
first limiting amino acid of Cellulomonas protein. Large fecal-nitrogen
content of the rats fed intact Cellulomonas cells indicated the resistance
of the cell wall of Cellulomonas to digestion. It was felt that cell
homogenization or lysis before feeding would improve the efficiency of
protein utilization.
The SCP product when dried and ground is a free-flowing powder with a
dark brown-to-yellow color depending on amount of lignin inclusion (Figure
41). At low moisture levels the storage properties and shelf life are very
-------
TABLE 16
106
ESSENTIAL AMINO ACID CONTENT OF THE CELL PROTEIN
(g OF AMINO ACID PER 100 g PROTEIN)
32
a
Cellulomonas
Amino acid cell protein
Arginine
Hist i dine
Isoleucine
Leu cine
Lysine
Methionine
Phenylalanine
Tyrosine
Threonine
Valine
9.21
2.30
4.74
11.20
6.84
1.86
4.36
2.67
5.37
10.71
FAOb Wheat0
reference protein flour
—
—
4.2
4.8
4.2
2.2
2.8
2.8
2.8
4.2
4.2
2.2
4.2
7.0
1.9
1.5
5.5
—
2.7
4.1
Beef°
7.7
3.3
6.0
8.0
10.0
3.2
5.0
—
5.0
5.5
BPC
protein
5.1
5.1
4.6
3.1
6.0
1.1
8.1
—
11.0
7.0
sample was hydrolyzed with 6 N HC1 at 100 G for 22 hr and analyzed
with a Beckman model 116 amino acid analyzer in the laboratories of Dr. S. P.
Yang, School of Home Economics, Louisiana State University.
National Academy of Science - National Research Council.
Q
lyengar, M. S., 1967. Protein from petroleum. Paper presented at
Single-cell Protein Conference at MIT, Cambridge, Massachusetts.
BP protein designates the single-cell protein obtained from
hydrocarbons by British Petroleum Company.
-------
Figure 41. Single-cell protein, freeze-dried and drum-dried.
-------
108
good. The product is easy to store, handle, and ship, and can easily
be mixed with food or feed materials for nutritional uses.
Two by-product streams are currently produced by the pilot plant. The
slurry underflow from the first mixer/settler, or the filter cake both
contain the insoluble fiber fraction. In addition to the unused fiber, this
solid stream also contains lignin, ash, salts, and some absorbed enzymes and
organisms.
This unused fiber could be recycled to the cellulose treatment section
for additional treatment and more nearly complete utilization, or it could
be recovered and used in paper stock or the manufacture of chemicals.
The liquid stream that exits either the second mixer/settler or the
centrifuge may either be recycled to the feed reslurry tank to replace
makeup water or may be used as a nutrient solution for hydroponic gardening.
A plant growth test has been made with this stream as a plant nutrient
42
source, and it was found to be satisfactory.
Owing to the relatively large volume of water used in the process, it
is apparent that most of the effluent liquid should be recycled. In
addition to decreasing water volume usage, this procedure would permit more
efficient utilization of the inorganic salt nutrients.
-------
ECONOMIC POTENTIAL
The importance or utility of any process depends upon its economics.
It must be profitable to be commercially developed and industrially
widespread. The future of SCP as a protein source for food or feed is
involved not only in current supply and demand but also in the great
protein needs of the future.
Many factors enter into the economics of microbial protein production.
Some are typical chemical engineering economics that are familiar while
others are not so familiar and require special consideration. Several
factors can be defined that are probably the most important to the economics
43
of microbial protein production, as follows.
1. Raw material
a. Ease of collection to a central area
b. Availability to a given site
c. Bulk-handling properties
d. Seasonal fluctuations in availability
2. Sterility requirements
a. Microbial encroachment
3. Fermentation
a. Residence time in reactor (doubling time)
b. Cell concentrations attainable
c. Operating temperature (cooling water versus refrigeration)
d. Total oxygen requirements
e. Power requirements for mass transfer
f. Heats of reaction
o. Cell yields per Ib of substrate consumed
h. Foaming tendency
4. Cell-harvesting techniques
a. High-speed centrifuges versus thickeners
109
-------
110
5. Washing and purification techniques for removal of:
a. Substrate residues
b. Raw-material impurities
c. Nucleic acids
d. Metabolic by-products
6. Product value
a. Percentage protein
b. Limiting amino acid
c. Digestibility
Factors such as drying costs and bagging and handling costs are omitted
since these are generally typical and familiar to all chemical engineers.
Certainly they will need to be included in the final analysis, but they must
be about the same regardless of the type of product.
Many companies, both in the United States and abroad, are producing
microbial proteins from hydrocarbons either in pilot plants or on a large
44
scale. A good example is British Petroleum with their pilot plant in
Scotland, their 1-ton-a-day plant in India, or their new
30-million-ton-per-year plant in France. The Institut Francais des Petroles
is already operating a 1-ton-per-day plant, while Esso Nestle's, Chinese
Petroleum Corporation, and others are operating pilot units with obvious
intentions of large plants in the future.
The cost of the carbon substrate used for SCP production varies not
only with the substrate selected but also with geographical locations and
the required purity of the starting material. For example, the cost of
hydrocarbon substrate is doubled in going from gas oil at 1 cent per pound
to purified n-paraffins at 2 cents per pound. The additional product
purification costs of the proteins produced on gas oil may more than offset
this initial 1-cent-per-pound difference in starting materials. As for
geographical location, methane in the Gulf Coast area is about 0.25 cent per
pound while the same material on the eastern seaboard would be about twice
that.
-------
Ill
The use of waste cellulose as a carbon substrate for microbial protein
production does not present the same problems of purity and geographical
location as those presented by hydrocarbon substrates. The production of
SCP from high-purity cellulose is actually a deterrent to microbial growth.
This same general conclusion was reached by Dr. N. J. King working with
brown rot fungi grown on alkali-treated wood. The waste cellulose
requires, however, an alkali pretreatment for cellulose swelling and
disruption of the lignin sheathing that protects the cellulose fibrils.
The geographical location of the plant would appear to have little
bearing on the cost of the cellulose substrate since waste cellulose seems
to be readily available everywhere. In rural areas the cellulose is
available as corn cobs, rice hulls, wheat straw, sawdust, bagasse, and so
on. In urban areas the solid wastes contain large portions of excelsior,
books, newspapers, rags, towels, wood, and so forth. Not only are these
wastes available, but also a credit rather than a debit may be given to the
cellulose consumer for getting rid of unwanted solid wastes. One of the
advantages of using bagasse as the source of cellulose is that it is
collected at one central point, the sugar mill, and can be obtained for its
fuel replacement value.
Bagasse costs have been calculated previously, and the raw material
cost per pound of fermentable carbohydrate has been established at 1.5
cents. The cost per ton of air-dry baled bagasse is about $18.90. This
material contains about 10 percent moisture. These are conservative costs;
they would probably be lower.
Productivity
Insufficient experimentation has been carried out to determine the
maximum cell density that can be obtained with cellulose feed. The
-------
112
difficulty arises from the fact that two or more consecutive reactions must
be carried on at about the same rate. If the enzymatic reaction goes too
fast, then allosteric inhibition of feedback repression from the
disaccharide causes a decrease in the cellulase activity. On the other
hand, if rapid bacterial growth is reached too quickly (in a batch process),
the log-phase growth rate becomes essentially irrecoverable. Doubling times
of 2 to 4 hr have been observed with Cellulomonas, corresponding favorably
to rates obtained on hydrocarbons. This means that holding time or dilution
time would be about the same in both cases.
It has been shown that volumetric production efficiencies of about 0.16
g dry cell mass per liter of fermenter capacity per hour have been obtained
with pure cultures, and as much as 0.51 has been obtained with symbiotic
cultures (Table 13). Data have been collected and computed to compare these
rates of SCP production with those of other comparable processes (Table 17).
It is seen that these values range from less than 1/2 to 7 times the rate
obtained in our plant. Limiting these values to those that have been
experimentally verified, the range goes up to about five times our current
rate. This was obtained with yeast growth and sulfite liquor. Comparable
values for yeast grown on hydrocarbons should be in this same range.
The culture mass-doubling time of pure Cellulomonas cultures at about
3.5 to 3.7 hr is well within the range of values currently considered to be
industrially feasible. The equilibrium cell density, however, is lower than
in the yeast processes. That the equilibrium cell density can be improved
from 0.5 g per liter is certain since the pilot plant has been operating at
low substrate loadings; to what extent, however, is not known at this time.
*
A rather interesting discovery was made during previous research that
could lead to improved fermentation economics although it is still too early
-------
TABLE 17
COMPARABLE VOLUMETRIC PRODUCTION EFFICIENCIES FOR CONTINUOUS FERMENTATIONS
Organism
Substrate
VPE
Reference
Yeast
Yeast
Yeast
Fungi
Mixed culture
Bacteria (Cellulomonas)
Bacteria (Cellulomonas
and Alcaligenes)
Sulfite liquor
Acid-solubilized
wood sugars
Alkanes
Cellulose
Sulfite liquor
Cellulose
Cellulose
2.5 (Experimental)
3.66 (Theoretical)
3.46 (Theoretical)
0.042 (Calculated
from batch)
0.57 (Experimental)
0.162 (Experimental)
0.512 (Calculated
from batch)
21 (Inskeep)
4 (Meller)
44 (Wang)
46 (Updegraf)
47 (Araberg)
Table 13
Table 13
aVPE « g of dry cell mass produced per liter of fermenter capacity per hr,
-------
114
to be certain of all the aspects. During one fermentation of treated
sugarcane bagasse, the rate and extent of visible cellulose breakdown
increased markedly over the values from previous runs. The culture was
found to be contaminated by an organism other than Cellulomonas. The second
organism was isolated and identified by Drs. V. R. Srinivasan and Y. W. Han
as of the genus Alcaligenes. A mixed culture of Cellulomonas and Alcaligenes
was prepared and grown by them, and the culture was found to be composed
primarily of Cellulomonas, and it exhibited better growth characteristics
than either of the pure bacteria in the same media.41 Concurrent experiments
in the Chemical Engineering Department on measurement of the amount of
solubilized carbohydrate present in the menstruum at any time showed that
the amount solubilized remained essentially constant at 200 to 300 mg per
liter in a pure culture of Cellulomonas. This knowledge, when combined with
previous observations, indicates that what could be happening in the symbiotic
culture was that the Alcaligenes was consuming the hydrolysis product that
had previously been inhibiting enzyme activity. The inhibiting product is
thought at the present time to be cellobiose. In other words, the enzyme
activity that had previously been inhibited by the presence of the reaction
product disaccharide in the menstruum was now free to proceed without
inhibition.
The last fermentation run finished before the issuance of this report
used Cellulomonas and Alcaligenes bacteria in a mixed-culture fermentation
of treated bagasse. Cell densities increased five-fold over all previous
pure culture runs, and growth rate was comparable. The calculated
theoretical VPE value was almost five times higher than comparable values
for pure cultures (Table 13). The only reason growth stopped at 6.24 g»of
cells per liter was lack of substrate. Higher cell densities should be
obtainable.
-------
115
In laboratory tests to seek other organisms that might be symbiotic
with Cellulomonas, several types of cellobiose-metabolizing yeasts were
grown with Cellulomonas in shake tubes. Most of the yeasts tested showed
higher cell densities than either Cellulomonas or themselves grown
separately in the same type media. None, however, showed as good growth
as Cellulomonas and Alcaligenes (Table 18).
Amino acid patterns of Cellulomonas and Alcaligenes were determined for
42
comparative purposes. Alcaligenes showed a larger methionine content than
Cellulomonas. Since methionine is the limiting amino acid in Cellulomonas
protein metabolism, the addition of Alcaligenes should enhance the
digestibility of the total SCP.
Work is continuing on the possibilities of symbiotic fermentation that
could be of great benefit to both cell density and culture growth rates.
Cell Harves ting
The concentration of the cells in the effluent from the fermenter is
very important when the cell concentration is to be increased in a
desludging centrifuge. This is due to the fact that the cost of such
equipment is normally based on the volumetric through-put rates.
Furthermore, a great deal more power is required to handle the additional
volume since the liquid must be subjected to forces 5,000 to 15,000 times
gravity. This is much more noticeable in the harvesting of bacteria because
their cell size is usually about one-fifth that of yeast, and power
requirements are inversely proportional to the square of the particle size.
Initial tests on the separation of Cellulomonas have shown that these
cells can be settled from suspension by adjusting the pH down to 5.2 or
below or by adding a polyionic flocculent. A continuous thickener is used
in the pilot plant in place of centrifuges. The power requirements and
-------
TABLE 18
SYMBIOTIC GROWTH
116
fl
Organism
C
A
C + A
C + Ycl
C + Yc2
C + Yc3
C + Yc4
C + Yc5
C + Yc6
C + Yc7
C + Yc8
C + Yc9
C + YclO
C + Yell
C + Ycl2
C + Ycl3
0 hr
12
14
18
22
21
16
17
17
24
17
15
20
22
17
15
15
Growthb (Klett
68 hr
40
14
270
60
60
60
70
62
67
52
60
80
65
100
70
100
uni ts )
92 hr
47
30
280
78
78
70
77
55
80
65
70
85
80
110
90
110
C = Cellulomonas, A = Alcaligenes, Yc = cellobiose-utilizing
yeast,
Growth is the average of duplicate shake-tube culture.
-------
117
maintenance on this equipment are extremely low compared with those for
high-speed centrifuges. The cell cream from this vessel may then be
concentrated much more economically by centrifugation.
Product Purification
The extent of purification required will depend upon the end use
expected for the product. Probably the ultimate objective of essentially
everyone's work in this area is to produce human food since this is a market
that could afford to pay a good price for the product. The big stumbling
block is nontoxicity and FDA approval. If this is the ultimate market for
SCP grown on gas-oil with its attendant conglomeration of products that
cannot be consumed by bacteria, then FDA would be quick to recognize that
such products would not be very nourishing to human beings. This means,
then, that for these products a large cost will be involved in solvent
extraction or whatever to remove these components from the final product.
Single-cell protein grown on hydrocarbon then will have some rather
costly purification steps unless an organism can be found that will easily
disengage itself from the residual hydrocarbons. The FDA has gone on record
in the past as ruling that the addition of as little as 200 ppm of mineral
oil to the human diet is objectionable.
Microbial proteins grown on cellulose may not encounter such severe
restrictions as products grown on other substrates, because cellulose is not
considered objectionable in the human diet. In fact, it is often added to
the diet as a bulking agent in the form of a water-soluble derivative. Some
of the undigested lignin remaining in the effluent stream is certainly going
to go out with the product, but lignin also is not particularly
objectionable in the human diet, since one already gets fair quantities of
-------
118
it in certain garden vegetables and fruits. The lignin passes through the
digestive tract without being assimilated.
Perhaps the largest worry that all SCP producers have is the presence
of very large amounts of nucleic acid in the protein product. Nucleic acids
have a deleterious effect on rats, causing gastric disturbances, skin rash,
and other effects, and very little is known about the tolerance limits of
such products in the human diet.
Product Value,
The market value of the product will determine how high the production
costs can go. It would appear at the moment that very few single-cell
proteins will be able to compete with soybean protein pep se. Still, there
are many factors that contribute to the value of protein other than adding
up the total nitrogen content and multiplying by 6.25. It turns out that in
the human diet the value of a protein is based on the digestibility of the
product, and this is usually limited by either the difficulty of cell wall
rupture or the amount of the limiting amino acid, or both.
The quality of a given protein is based on how much weight gain results
per gram of protein consumed. This is variously determined as PER (protein
efficiency ratio), or the amount of weight gain per gram of protein intake;
as BV (biological value); as NPU (net protein utilization); and so forth.
In the final analysis, the data that would be most desirable would be the
protein quality divided by the cost or the weight gain per unit of cost for
each of the various proteins. Unfortunately, these data appear to be rather
limited for unconventional proteins.
-------
119
Summary of Product Cost
It has been impossible to obtain a detailed cost analysis on the
production of SCP on hydrocarbon substrate from one of the large producers
because of the competitive nature of this endeavor. About the only cost
information available appears to be the final selling price, and even that
has to be looked at with some skepticism because it could change with supply
and demand. Cost data available show quite a wide range (Table 19). Esso
Research and Engineering has indicated an approximate selling price of 17
cents per pound of 50 percent protein single cells that have been grown on
n-paraffin. British Petroleum has indicated a price in the range of 10 to
20 cents per pound for 50 percent protein grown on gas-oil. Soybean
proteins are quoted at a price of 6 to 7 cents a pound on the same basis
while fish flour is in the range of 15 to 20 cents per pound. It should be
obvious from these figures that if the market for SCP is to be animal feed
supplement, then the price to beat is the 6 to 7 cents per pound for soybean
flour.
The current operating data for the pilot plant do not permit an
extensive economic analysis. A key point, however, in the evaluation of
any SCP process is the VPE or fermenter productivity value. Most economic
analyses in the literature give a VPE of from 1.5 to 3.5 as an economic
break-even point. Our current VPE of from 0.1 to 0.5 is considerably below
this point. The key to enhancement of the processing economics is an
»
increase in the VPE value. Our current VPE would make it necessary to
operate large fermenters and cell recovery equipment, and this hurts the
present economic picture.
Since, however, the little optimization done so far has yielded
favorable results in production rates and efficiencies, it is felt that VPE
-------
120
values for this process can be raised to a competitive range. Moreover,
symbiotic-culture growth has shown great potential, and other economic
savings will be realized with further work.
-------
TABLE 19
COST OF CONVENTIONAL AND UNCONVENTIONAL PROTEIN*
Food product
Conventional protein foods :
Wheat flour
Skim milk powder
Fish, dried
Cheese
Chicken
Beef
E8gs
l^ess conventional protein foods :
Cottonseed flour
Soy protein flour
Fish protein concentrate
Peanut Cake
Price of product
(C/lb)
5
15
14
32
26
21
24
7
7
12.5
7
Percent protein
12
36
37
24
15
12
11
55
52
85
42
Price of prot'
(C/lb)
42
43
39
133
173
175
211
12
13
14
17
-------
TABLE 19 (Continued)
COST OF CONVENTIONAL AND UNCONVENTIONAL PROTEIN*
Food product
Unconventional protein foods :
Algae
Yeast (petroleum)
Leaf protein
Spirulina
Yeast (vegetable)
C
Bacteria (cellulose)
Price of product Percent protein
3 50
6-8 50
37-47 50
18 65
10-14 50
10-15 50-55
Price of prote:
(Cl/b)
6
12-16
74-94
28
20-28
20-30
All prices, except where noted, are from Abbott, J. C., 1966. Unconventional protein. Presented
at engineering research conference at Santa Barbara, California.
Wang, D. I. C. Proteins from petroleum. Chemical Engineering, p. 99, Aug. 26, 1968.
Data from this report, based on current operating data.
Ni
NO
-------
REFERENCES
1. Dunlap, c. E. Proteins from waste cellulose by chemical-microbial
processing. Ph.D. Thesis, Department of Chemical Engineering,
Louisiana State University, Baton Rouge, 1969.
2. Vaughan, R. D. Reuse of solid wastes: a major solution to a major
national problem. Waste Age3 1(1):10, 14-15, Apr. 1970.
3. Personal communication. C. E. Dunlap, Louisiana State University,
to T. C. Purcell, Bureau of Solid Waste Management, Mar. 25, 1969.
4. Meller, F. H. Conversion of organic solid wastes into yeast; an
economic evaluation. Public Health Service Publication No. 1909.
Washington, U.S. Government Printing Office, 1969. p. 15.
5. Personal communication. C. E. Dunlap, Louisiana State University,
to W. F. Carr, U.S. Forest Products Laboratory, Mar. 24, 1969.
6. Atchison, J. E. Sugar cane bagasse as a raw material for pulp, paper,
and board. Presented at United Nations. Educational, Scientific,
and Cultural Organization and Food and Agricultural Organization
Regional Symposium on Pulp and Paper Research and Technology for
the Middle'East and North Africa, Beirut, Lebanon, Dec. 10-14, 1962.
7. Audubon Sugar Factory, Louisiana State University. Unpublished
records [1967].
8. The world food problem. A report of the President's Science Advisory
Committee, v. 1. Report of the Panel on the World Food Supply.
Washington, The White House, 1967. 127 p.
9. Altschul, A. M., ed. World Protein Resources. Washington, American
Chemical Society, 1966. 285 p. (Advances in Chemistry Series,
No. 57).
10. United Nations Economic and Social Council Advisory Committee on the
Application of Science and Technology to Development. Increasing
the production and use of edible protein. Feeding the expanding
world population: recommendations for international action to avert
the impending protein crisis. United Nations Economic and Social
Council Agenda Item E/4343. [n.p.], 1967. [110 p.]
11. Thaysen, A. D. Food and fodder yeast. In Roman, W., ed. Yeasts.
New York, Academic Press Inc., 1957- p. 155-246.
12. Mark, H. Interaction and arrangement of cellulose chains. In
Ott, E., H. M. Spurlin, and M. W. Grafflin, eds. High polymers.
v. 5. pt- !• Cellulose and cellulose derivatives. New York,
Interscience Publishers, 1954. p. 217-230.
123
-------
124
13. Segal, L., J. J. Creely, A. E. Martin, Jr., and C. M. Conrad.
An empirical method for estimating the degree of crystallinity
of native cellulose using the x-ray diffractometer. Text-lie
Research Journal, 29(10):786-794, Oct. 1959.
14. Baker, T. I., G. V- Quicke, 0. G. Bentley, R. R. Johnson, and
A. L. Moxon. The influence of certain physical properties of
purified celluloses and forage celluloses on their digestibility
by rumen microorganisms in vitro. Journal of Animal Science^
18(2):655-662, May 1959.
15. Warwicker, J. 0., R. Jeffries, R. L. Colbran, and R. N. Robinson.
A review of the literature on the effect of caustic soda and
other swelling agents on the fine structure of cotton. Shirley
Institute Pamphlet No. 93. Didsbury, Manchester, England, The
Cotton, Silk and Man-made Fibres Research Association, 1966.
247 p.
16. McBurney, L. F. Kinetics of degradation reactions. In Ott, E.,
H. M. Spurlin, and M. W. Grafflin, eds. High polymers, v. 5,
pt. 1. Cellulose and cellulose derivatives. New York,
Interscience Publishers, 1954. p.99-129.
17. Baumgardt, B. R., M. W. Taylor, and J. L. Cason. Evaluation of
forages in the laboratory. II. Simplified artificial rumen
procedure for obtaining repeatable estimates of forage nutritive
value. Journal of Dairy Soienee, 45(l):62-68, Jan. 1962.
18. Dunlap, C. E., and C. D. Callihan. Fermentative utilization of
sugar cane bagasse. Unpublished data, Louisiana State University,
Baton Rouge. [Presented at Meeting, American Society of Sugar
Cane Technologists, June 5, 1969.]
19. Delbrueck, M. Hefe ein edelpilz. [Yeast, a peculiar fungus.]
Wochensohrift fuer Brauerei, 27(31):373-376, July 30, 1910.
20. Locke, E. G., J. F. Saeman, and G. R. Dickerman. The production
of wood sugar in Germany and its conversion to yeast and alcohol.
FIAT [Field Information Agency, Technical] Final Report No. 499.
Joint Intelligence Objective Agency, Office of the Military
Government for Germany (U.S.), 1945.
21. Inskeep, G. C., A. J. Wiley, J. M. Holderby, and L. P. Hughes.
Food yeast from sulfite liquor. Industrial and Engineering
Chemistry., 43(8) :1702-1711, Aug. 1951.
22. Amberg, H. R. Aerobic fermentation of spent sulfite liquor for the
production of protein concentrate animal feed supplement. National
Council for Stream Improvement Technical Bulletin No. 82. New Yor£,
1956. 31 p.
-------
125
23. Langwell, H. Cellulose fermentation. Journal of the Society of
Chemical Industry, Chemistry and Industry, 51(49):988-994,
Dec. 2, 1932.
24. Olson, F. R., W. H. Peterson, and E. C. Sherrard. Effect of
lignin on fermentation of cellulosic materials. Industrial and
Engineering Chemistry., 29(9) : 1026-1029, Sept. 1937.
25. Acharya, C. N. Studies on the anaerobic decomposition of plant
materials. IV. The decomposition of plant substances of varying
composition. Biochemical Journal., 29(6) : 1459-1*67, June 1935.
26. Fontaine, F. E. The fermentation of cellulose and glucose by
thermophilic bacteria. Ph.D. Thesis, University of Wisconsin,
Madison, 1941. 81 p.
27. Virtanen, A. I., and J. Hukki. Thermophilic fermentation of wood.
Soumen Kemistilehti3 19B:4-13, 1946.
28. Hajny, G. J., C. H. Gardner, and G. J. Ritter. Thermophilic fermen-
tation of cellulosic and lignocellulosic materials. Industrial
and Engineering Chemistry, 43(6):1384-1389, June 1951.
29. Gray, W. D. Fungi and world protein supply. In Altschul, A. M., ed.
World protein resources. Washington, American Chemical Society,
1966. (Advances in Chemistry Series, No. 57). p.261-268.
30. Ghose, T. K. Continuous enzymatic saccharification of cellulose with
culture filtrates of trichoderma viride QM 6a. Biotechnology and
Bioengineering, 11(2):239-261, Mar. 1969.
31. Bellamy, W. D. Cellulose as a source of single-cell proteins—a
preliminary evaluation. General Electric Research and Development
Center Report No. 69-C-335. Schenectady, General Electric Company,
1969. 5 p.
32. Han, Y. W., and V. R. Srinivasan. Isolation and characterization
of a cellulose utilizing bacterium. Applied Microbiology, 16(8):
1140-1145, Aug. 1968.
33. Dehority, B. A., K. El-Shazly, and R. R. Johnson. Studies with the
cellulolytic fraction of rumen bacteria obtained by differential
centrifugation. Journal of Animal Science, 19(4):1098-1109,
Nov. 1960.
34. Pew, J. C., and P. Weyna. Fine grinding, enzyme digestion, and the
lignin-cellulose bond in wood. Tappi, 45(3):247-256, Mar. 1962.
35. Stranks, D. W. Fermenting wood sub-strates with a rumen cellulo-
lytic bacterium. Forest Products Journal, 9(7):228-233, 1959.
-------
126
36. Han, Y. W., H. A. Shuyten, Jr., and C. D. Callihan. The combined
effect of heat and alkali for sterilizing bacterial spore in
sugar cane bagasse. Presented at South Central Branch Meeting,
American Society for Microbiology, New Orleans, Nov. 21-22, 1969.
24 p.
37. Campuzano, A. Mathematical models for fermentation processes.
M.S. Thesis, Louisiana State University, Baton Rouge, 1969.
38. Adams, S. L., and R. E. Hungate. Continuous fermentation cycle
times. Prediction from growth curve analysis. Industrial and
Engineering Chemistry, 42(9):1815-1818, Sept. 1950.
39. Yang, S. P., and W. F. McKnight. Nutritional studies of single
cell protein. In Single-cell protein from cellulosic wastes.
Final report, v. 2. Baton Rouge, Louisiana State University,
1970. p.527-549.
40. Srinivasan, V. R., and Y. W. Han. Utilization of bagasse. In
Hajny, G- J., and E. T. Reese, eds. Cellulases and their appli-
cations. Washington, American Chemical Society, 1969. (Advances
in Chemistry Series, No. 95). p.447-460.
41. Yang, S. P. Nutritional value of single cell protein. Presented
at 31st Annual Meeting, Chemurgic Council, New Orleans, Oct. 23,
1969.
42. Michaud, M. Unpublished data, 1969.
43. Callihan, C. D., and C. E. Dunlap. The economics of microbial
proteins produced from cellulosic wastes. Compost Soienoe,
10(1-2):6-12, Spring-Summer 1969.
44. Wang, D. I. C. Proteins from petroleum. Chemical Engineering,
75(18):99-108, Aug. 26, 1968.
45. Private communication. Dr. N. J. King, Forest Products Research
Laboratory, Aylesbury, Bucks, England, to C. E. Dunlap, Louisiana
State University, 1968.
46. Updegraff, D. M. Degradation of waste paper to protein. Research
in microbial fermentations. Denver Research Institute Report
No. 2534. University of Denver, 1969. 100 p.
47. Amberg, H. R. Aerobic fermentation of spent sulfite liquor for the
production of protein concentrate animal feed supplement. National
Council for Stream Improvement Technical Bulletin No. 82. New York,
1956. 31 p.
Via462
------- |