The Pennsylvania State University

The Graduate School

College of Engineering

MOVING TO SUSTAINABILITY: IMPROVING MATERIAL

FLOWS IN THE IRON CASTING INDUSTRY

A Dissertation in

Environmental Engineering

by

He Huang

© 2010 He Huang

Submitted in Partial Fulfillment of the Requirements

for the Degree of

Doctor of Philosophy

August 2010

The dissertation of He Huang was reviewed and approved* by the following:

Fred S. Cannon Professor of Environmental Engineering Dissertation Advisor Chair of Committee Brian A. Dempsey Professor of Environmental Engineering Sridhar Komarneni Distinguished Professor of Clay Mineralogy Adjunct Professor of Civil and Environmental Engineering Robert C. Voigt Professor of Industrial Engineering Angela Lueking Associate Professor of Energy and Mineral Engineering Peggy A. Johnson Professor of Civil Engineering Professor's Name Head of the Department of Civil and Environmental Engineering

*Signatures are on file in the Graduate School

iii

Abstract

Sustainable engineering solutions were developed to improve the sustainability of

the iron casting process. These engineering solutions aimed to modify the material and

energy flow in the iron casting industry in a manner that previously wasted energy and

resources can be utilized to control the pollution and reduce energy and material cost for

the iron foundries.

The first approach was to reclaim the thermal energy from cupola furnace exhaust

gas to produce and regenerate porous carbons that could be employed to adsorb the

volatile organic compounds from the iron casting process. Saturated porous carbons

could further be reused in the green sand mold as the carbon additive. Therefore no extra

cost will be posed to the iron casting to remove its VOC emissions. The pore structure

developments of different coals under simulated thermal conditions were investigated.

The adsorption of typical volatile organic compounds from the iron casting process on

the in-situ porous carbons was also studied and compared with a commercial activated

carbon from similar precursors.

The second approach was to replace expensive foundry coke by waste anthracite

fines. The iron melting process was carefully investigated for the design of the alternative

fuel. The thermal energy in the preheat zone of the cupola furnace was employed as free

energy to create silicon carbide binding in-situ. The traditional material flow into the

cupola furnace was also rearranged to assist the in-situ ceramic binding, minimize the

change of chemistry in the cupola furnace, and eliminate the additional cost with silicon

additives. Different Si-containing materials tested in this study showed different binding

mechanisms at high temperature. Bindered anthracite with silicon powders had the

iv

highest post-pyrolysis strength provided by the nanowires generated in-situ at high

temperature. The binding strength from the nanowires was further enhanced by

decreasing the anthracite grain size which allowed more direct connections of the

anthracite particles by individual nanowires. The post-pyrolysis unconfined compressive

strength of anthracite pellets (2.86 cm in diameter and 1.875” in length) made from

pulverized anthracite fines with 9% silicon powders reached as high as 3.6 Mpa (535 psi).

Anthracite pellets made from 50% pulverized anthracite fines and 50% original anthracite

fines were only slightly weaker than the anthracite pellets made from 100% pulverized

anthracite fines. The nanowires generated between silicon powders and anthracite fines at

high temperature are silicon carbide (3C-SiC or β-SiC) nanowires with highly

crystallized face-center cubic zinc-blender structures. These nanowires, which were

grown though the vapor-solid mechanism by stacking the (111) lattice plane along the

[111] direction, were typically 30-60 nm in the diameter and could grow tens of

micrometers in length. In the lab-scale pyrolysis system used in this study, the silicon

carbide nanowires started to form at temperature as low as 1100 °C. At 1400 °C the

formation of silicon carbide was very fast and finished within 10 minutes. The

replacement of foundry coke by waste anthracite fines could save significant amount of

energy, greatly reduce carbon dioxide emission, and avoid other pollutions from the

coking process.

v

Table of Contents

List of Figures ................................................................................................................... vii List of Tables ..................................................................................................................... ix Acknowledgements............................................................................................................. x Chapter 1 Introduction ........................................................................................................ 1

1.1 Industrial pollution and sustainable environmental engineering solutions ............... 1 1.2 Material and energy flow in the iron casting industry .............................................. 4 1.3 Problem statement, objectives, and hypothesis of this study .................................... 9

Chapter 2 Pore structure development of in-situ pyrolyzed coals for pollution prevention in iron foundries.......................................................................................................... 13

Abstract ......................................................................................................................... 13 2.1 Background and objectives ..................................................................................... 14 2.2 Experimental ........................................................................................................... 16

2.2.1 Raw material..................................................................................................... 16 2.2 Pyrolysis .................................................................................................................. 18 2.3 TGA-Mass spectrophotometry (TGA-MS)............................................................. 19 2.4 Characterization of in-situ pyrolyzed carbon .......................................................... 19 2.3 Results and Discussions .......................................................................................... 21

2.3.1 Effect of pyrolysis temperature on pore structure development in raw coals of several ranks .............................................................................................................. 21 2.3.2 Effect of pyrolysis time and tube-furnace pyrolysis on pore structure development .............................................................................................................. 23 2.3.3 The potential of using moisture content and presoaked water as the activation agent .......................................................................................................................... 25 2.3.4 Effect of raw coal grain size on pore structure development ........................... 27 2.3.5 Surface acidity change of lignite after pyrolysis .............................................. 29 2.3.6 Compare the pore volume distribution and benzene sorption of in-situ pyrolyzed carbon with commercial lignite-based AC............................................... 29

2.4 Conclusion............................................................................................................... 31 Chapter 3 Binding waste anthracite fines by low emission renewable binders and ceramic

materials for foundry coke replacement ..................................................................... 48 Abstract ......................................................................................................................... 48 3.1 Background and objectives ..................................................................................... 49 3.2 Experimental ........................................................................................................... 55

3.2.1 Raw materials ................................................................................................... 55 3.2.2 Anthracite pellet preparation ............................................................................ 56 3.2.3 Mechanical strength tests ................................................................................. 56 3.2.4 Pyrolysis ........................................................................................................... 57 3.2.5 Characterizations .............................................................................................. 57

3.3 Results and discussion............................................................................................. 58 3.3.1 The effect of pellet composition on mechanical strength................................. 58 3.3.2 Crystal structure changes within anthracite pellets with difference Si-containing materials................................................................................................... 61 3.3.3 After pyrolysis: the morphologies of anthracite pellets that contained various Si-containing materials.............................................................................................. 62

vi

3.3.4 Improving the silicon carbide nanowire binding strength by reducing the anthracite grain size................................................................................................... 64 3.3.5 Coke replacement's opportunity on sustainability and its limitations .............. 65 3.3.6 Burning rates of bindered anthracite pellets..................................................... 67 3.3.7 Full-scale test of briquetted anthracite fine bricks in an operating foundry cupola ........................................................................................................................ 68 3.3.8 Effect of preheating and organic crosslinking on the binding strength of the collagen binder .......................................................................................................... 70

Chapter 4 Formation of silicon carbide nanowires on the surface of anthracite fines...... 93 Abstract ......................................................................................................................... 93 4.1 Introduction ............................................................................................................. 93 4.2 Experimental Section .............................................................................................. 95

4.2.1 Raw materials ................................................................................................... 95 4.2.2 Preparation of bindered anthracite pellets ........................................................ 96 4.2.3 The pyrolysis process ....................................................................................... 96 4.2.4 SEM, EDS, and TEM ....................................................................................... 97 4.2.5 Ambient-temperature XRD and real-time high-temperature XRD .................. 97 4.2.6 Unconfined compressive strength .................................................................... 98

4.3 Results and Discussion............................................................................................ 98 4.3.1 Morphology and crystal structure of the silicon carbide nanowires formed at 1400 °C...................................................................................................................... 98 4.3.2 The effect of pyrolysis temperature on the development of the SCNWs....... 101 4.3.3 The effect of silicon content on the growth of SCNWs at 1400 °C ............... 103 4.3.4 Discussion on the growth mechanism ............................................................ 103 4.3.6 Summary......................................................................................................... 105

Chapter 5 Conclusions .................................................................................................... 115 5.1 Conclusions ........................................................................................................... 115

5.1.1 Original contributions to science and engineering ......................................... 116 5.1.2 The impacts on sustainability beyond the iron casting industry..................... 119

5.2 Recommendations for future studies..................................................................... 119 Appendix A: The crystal structure of 3C-SiC and the stacking sequence of the SiC

nanowires .................................................................................................................. 125 Appendix B: Appling raw rice husks as an addition renewable silicon sources for binding

the anthracite fines .................................................................................................... 131 References:...................................................................................................................... 136

vii

List of Figures

Figure 2-1 TGA recorded mass loss of different coals during pyrolysis (Temperature started increasing at about 2 minutes, with the rate of 50 °C/min to the prescribed temperatures)............................................................................................................... 36

Figure 2-2 Pore volume distributions (listed in descending order) of several ranks of coals pyrolyzed in TGA at 600-900 °C for 1 hour. (Heating rate 50 °C/min)..................... 37

Figure 2-3 Evolution of Methane, BTX, H2O, and COx from coals during pyrolysis (L-left y axis; R-right y axis) ........................................................................................... 38

Figure 2-4 Pore volume distributions of Sabine lignite powders pyrolyzed in TGA at 800 °C and 900 °C for different pyrolysis time ................................................................. 39

Figure 2-5 Pore volume distribution of Sabine lignite pyrolyzed by tube-furnace pyrolysis (800 °C)....................................................................................................................... 40

Figure 2-6 Effect of different water resources on the development of pore structures in powdered Sabine lignite during pyrolysis (All pyrolysis was conducted at 800 °C for 15 minutes).................................................................................................................. 41

Figure 2-7 Pore volume distribution of in-situ pyrolyzed carbon from powder Sabine lignite and granular lignite. (TGA pyrolysis at 800 °C) ............................................. 42

Figure 2-8 Mass loss of powder and granular Sabine lignite during pyrolysis at 800 °C, with a heating up rate at 50 °C/min ............................................................................ 43

Figure 2-9 Volatile releases from granular (········) and powdered (——) Sabine lignite during pyrolysis detected by TGA-MS....................................................................... 44

Figure 2-10 Pore volume distribution of in-situ pyrolyzed carbon from Sabine lignite (TGA pyrolysis, 800 °C for 15 min) and commercial lignite-based activated carbon45

Figure 2-11Comparing the adsorption of benzene onto the in-situ pyrolyzed porous carbon and a commercial activated carbon................................................................. 46

Figure 2-12 Adsorption of typical volatile organic compounds from green sand molds onto the in-situ pyrolyzed porous carbon.................................................................... 47

Figure 3-1 The operation schematic for a cupola furnace that includes bindered anthracite bricks........................................................................................................................... 80

Figure 3-2 The setup of the drop-shatter tester used in this study .................................... 81 Figure 3-3 SEM images of anthracite pellets with different Si-containing materials after

the pyrolysis at 1400 °C. Pellets with 9% kaolinite powders (a&b); Pellets with 9% solid sodium silicate(c&d); Pellets with 9% silicon powders (e&f)........................... 82

Figure 3-4 Crystal structure changes of anthracite pellets with different ceramic additives (9%) after the pyrolysis at 1400 °C ............................................................................ 83

Figure 3-5 EDS spectrum of (a) a sphere on the surface of an anthracite particle in pyrolyzed anthracite pellets that contained kaolinite powders; (b) a ceramic bridge in a pyrolyzed anthracite pellet that contained sodium silicate ...................................... 84

Figure 3-6 High resolution TEM of the silicon carbide nanowires formed at 1400 °C within the anthracite pellets with 9% silicon powder ................................................. 85

Figure 3-7 Different amount of powdered anthracite and silicon: effect on the unconfined compressive strength after pyrolysis at 1400 °C ........................................................ 86

viii

Figure 3-8 The gradual change in morphology within the pyrolyzed anthracite pellets with 9% silicon content as the amount of powdered anthracite increased (a-made from 100% raw anthracite fines; b-made from 75% raw anthracite fines and 25% powdered anthracite; c- made from 50% raw anthracite fines and 50% powdered anthracite; d- made from 100% powdered anthracite ....................................................................... 87

Figure 3-9 Changes of upper stack temperature and cupola back pressure during coke replacement by anthracite fine bricks ......................................................................... 88

Figure 3-10 SEM images of anthracite particles bound by 1% collagen-based binders at room temperature (a), at 70 °C (b), and 1% collagen+1% Fructose at 70 °C (c, d) ... 89

Figure 3-11 The effect of adding fructose on the unconfined compressive strength and anti-breakage strength of the anthracite pellets at ambient conditions ....................... 90

Figure 3-12 Intensive drop shatter test of pellets with different binder and sugar contents..................................................................................................................................... 91

Figure 3-13 The crosslinking between a peptide chain and a monosaccharide molecule 92 Figure 4-1 EDS spectrum of the ash content from the anthracite used in this study...... 106 Figure 4-2 The pyrolysis setup ....................................................................................... 107 Figure 4-3 SEM images of silicon carbide nanowires grown from anthracite fines (#10 ×

80, with 9% silicon and 1% collagen) at 1400 °C. Inserts in d are EDS spectrums from different locations on a nanowire..................................................................... 108

Figure 4-4 TEM image of the SCNWs grown from anthracite fines (#10 × 80, with 9% silicon and 1% collagen) at 1400 °C. The up corner insert in a is the selected area electron diffraction pattern of the SCNWs ............................................................... 109

Figure 4-5 XRD patterns of raw anthracite fines, and anthracite fines with 9% silicon powders and 1% collagen before and after thermal treatment at 1400 °C................ 110

Figure 4-6 Real-time change of the XRD pattern at 1400 °C within mixed anthracite and silicon powders (9% silicon and 91% anthracite)..................................................... 111

Figure 4-7 Crystal structure change as temperature increased within anthracite pellets made from #10 × 80 anthracite fines with 9% silicon content ................................. 112

Figure 4-8 Change of morphology as temperature increases within anthracite pellets made from #10 × 80 anthracite fines with 9% silicon content (a-25 °C; c -1000 °C; e,f-1100 °C; g, h-1200 °C, b and d show the EDS spectrums on silicon particles treated at the 25 °C and 1000 °C respectively)......................................................... 113

Figure 4-9 SEM images of SCNWs grown from anthracite fines (#10 × 80) with different silicon contents (a,b-4.72%, c,d-1.74%) at 1400 °C................................................. 114

Figure 5-1 The change of material and energy flow in the iron casting process by introducing the in-situ porous carbons...................................................................... 123

Figure 5-2 The modified material flow into the cupola furnace by replacing a part of the coke with bindered anthracite fines .......................................................................... 124

ix

List of Tables

Table 2-1 Elemental analysis of the raw coals.................................................................. 33 Table 2-2 Sieve analysis of the received anthracite and bituminous coal ........................ 34 Table 2-3 Change of slurry pH of powdered lignite after pyrolysis in TGA for 1 hour... 35 Table 3-1 Sieve analysis of the original anthracite fines .................................................. 74 Table 3-2 Compare the elemental analysis and heat content of the anthracite tested in this

study to a typical foundry coke................................................................................... 75 Table 3-3 The strengths of anthracite pellets bindered with an array of binders and

additives recipes.......................................................................................................... 76 Table 3-4 Comparison on the burning rates of different anthracite pellets and a foundry

coke in a tube furnace ................................................................................................. 77 Table 3-5 The effect of preheating on the binding strength of the collagen based binder 78 Table 3-6 Mass loss from the evaporation of the water generated from the crosslinking

between collagen and fructose .................................................................................... 79

x

Acknowledgements

I am heartily grateful to my advisor, Dr. Fred Cannon, whose guidance, patience,

and support from the beginning to the final stage of this study allow me to develop a

better understanding on this subject. It is my great pleasure and honor to work with him

for my doctoral study. I am indebted to many of my colleagues, who kindly offer their

help in a number of ways at times when I desperately need it.

I want to dedicate this thesis to my wife, Jinghua, for her understanding, support

and great faith in me.

1

Chapter 1 Introduction

1.1 Industrial pollution and sustainable environmental engineering solutions

Industrial facilities (including manufacturing, mining, oil and gas extraction, and

service industries) is an important component of the modern society. As in the United

States, industrial facilities consume the largest amount of energy and generate the most

pollution among all human activities. Recent studies have found that human industrial

activities started affecting the earth environment at very early ages (Hong et al. 1996). By

measuring the copper concentration in Greenland ice dated for several million years,

scientists discovered higher than normal copper concentration in the atmosphere started

2500 years ago. This early large-scale pollution of the atmosphere of the northern

hemisphere is attributed to emissions from highly polluting smelting technologies used

for copper production during Roman and medieval times, especially in Europe and China.

The industrial revolution that began in the 18th century influenced almost every aspect of

daily life in some way, and it also dramatically increased the pollution from industrial

activities to a new level and finally intrigued the public awareness of environmental

pollution after the World War II. In order to protect the environment and the health of the

residents, different environmental legislations were enacted to restrict industrial facilities.

Nowadays, most of the nations worldwide have enacted legislation to regulate pollution

from various industrial activities. To achieve the emission limits set in these legislations,

industrial facilities have to employ pollution control as a step in their industrial activities.

The traditional pollution control refers to “end of pipe” treatment technologies.

These pollution control processes were primarily installed for appropriate disposal of the

2

contaminations that generated from the industrial facilities. Traditional pollution control

processes did cut down the emission from industrial activities. As a result the

environmental qualities in many developed countries were greatly improved during the

last several decades. However, pollution control at the end of pipe also has many

inherited shortcomings. Normally, the engineers that designed the pollution control

process only focused on what came out at the end of the manufacturing process; and then

select appropriate unit operations to remove the contaminates in the effluent to achieve

the emission standards. Therefore, traditional pollution control poses extra effort to the

industrial activities. As a result the cost of the industrial production increased. It is

actually appropriate in some aspect, since the environmental impact, which has been

neglected for a long time, should also be counted as manufacturing cost. However, from

the sustainable development point of view, pollution control should also be an integral

part of the industrial manufacturing instead of a separated system. In the last two decades,

the concept of sustainable development has been emerging in the environmental and

social studies (World-Commission-on-Environment-and-Development 1987). It pursues

the balance among the environment, the economy, and the human society. In the

sustainable development for industrial manufacturing, cleaner production and life cycle

analysis are receiving increasing popularities. The basic idea is that sustainability should

be taken into consideration from the first step in the industrial production from raw

material extraction to the final disposal of the product.

Thus, in the new era of environmental engineering, the challenge for

environmental engineers is that the pollution control is no longer a simple step at the end

of the manufacturing. A sustainable engineering solution requires the environmental

3

engineers to have a comprehensive understanding in the specific industrial manufacturing

process from which the pollution was generated. It is meaningful to notice that pollution

usually was generated from the consumption of energy and raw materials. A sustainable

engineering solution should integrate the pollution control unit operations with the

manufacturing process to improve the material and energy flow in the whole system. By

improving the material and energy flow, the consumption of energy and raw materials

can be further reduced, and less pollution will be generated from the whole process. Also

it is straightforward that with less energy and raw material consumption, the overall cost

for the industrial manufacturing and the pollution control can also be decreased.

Therefore, a sustainable pollution control solution should benefit the industrial

manufacturing both environmentally and economically. Furthermore, sustainable

solutions can also be achieved across different industries. That is an optimized material

and energy flow can be designed for multiple industries. In other words, the waste

materials from one industry may be usable resources for another manufacturing process.

Multi-industry approach is the one of the ultimate goals of sustainable environmental

engineering on industrial pollution control since it can significantly reduce the

environmental impact on the human society from the industrial practices.

In this research, sustainable engineering solutions were investigated in order to

control the pollution and reduce the cost from the iron casting industry. At the same time,

extra resources from upstream and downstream industries were also utilized in the iron

casting process to extend the sustainable impacts of this study.

4

1.2 Material and energy flow in the iron casting industry

The history of metal casting can be dated back to around 3000 B.C. in the Middle

East. However metal casting at large scale started 2500 years ago in Europe and China

for iron casting. It had immediate impact on the earth atmosphere as discovered from the

ice core record (Hong et al. 1996). Although many improvements have been introduced

into iron casting process during the last 2000 years, the basic concepts for iron casting are

still the same. First the iron has to be melt at very high temperature in a cupola furnace.

Tremendous amount of energy is consumed in this iron melting process. Then the molten

iron is poured into a prepared solid mold. As the molten iron cools down and solidifies in

the mold, it takes the shape of the mold. Therefore a casting process can be roughly

divided into mold making, iron melting and pouring, solidification and shake out of the

product. Major material flows in the iron casting process are the iron flow, the molding

material (sand, clay, and coal) flow, and the fuel flow (foundry coke).

The iron flow is very straightforward. The raw iron (ie. pig iron) is added into the

cupola furnace to be melted and then poured into the mold. After the shake out, the iron

becomes the product. In order to melt the iron, foundry coke has to be continuously

supplied into the cupola furnace to maintain a very high temperature. Air also has to be

continuously supplied to provide oxygen for the combustion. After the depletion of the

oxygen, this air flow leaves the cupola furnace with a temperature of 900 to 1100 °C. In

order to be released to the atmosphere this super hot off-gas has to be cooled in a drop-

out box, in which water is used as the coolant. Thus huge amount of heat energy is

wasted in the cooling process.

5

In the molding materials flow, the raw materials are mixed with water first. This

mixture was then made into mold with a desired shape. After the molten iron was cooled

down and solidified, the sand mold was then destroyed in the shake out process. The used

molding materials were then separated from the iron product. Usually only a portion of

the used molding materials was recycled. The others were disposed in landfills.

Most of the organic air pollution in foundries is generated when these two major

materials flows meet. The super heated molten iron drives out the VOCs in the high

volatile bituminous coal used in the green sand mold. Also when sand cores are included

in the green sand mold for more complicated casting shapes, volatile organic compounds

would also be released from the organic binders used in the sand cores. These organic

pollutants emit from the sand during the shake out process. The regulations on these

VOC emissions are becoming more and more stringent.

Meanwhile, the foundry industry is also facing a crisis in their fuel supply.

Foundry coke is currently the universal fuel for the cupola furnace, as it has many

desirable properties for iron casting. Produced by pyrolyzing special coking bituminous

coal at 1000 °C for 14-36 hours, foundry coke has high fixed carbon, low volatile content,

low sulfur content, and low ash (Avallone et al. 2006). More important is that the

pyrolysis fused the carbon contents in the coking coals into a porous structure. In this

way the coke has both the mechanical strength to support a fuel bed in the cupola furnace,

and a fast combustion rate. After the discovery of coke, it replaced anthracite chunks and

charcoals and became the only fuel for cupola furnace (Kirk 1903; Moldenke 1917).

However, some drawbacks on using coke as the cupola furnace emerged in the last

decades.

6

The first problem with foundry coke is that foundry coke can only be produced

from special coking bituminous coals. The bituminous coal must have a free swelling

number that is larger than 5 to be qualified as coke precursor. However, if the free

swelling number is too high, the produced coke may have a low mechanical strength. In

addition to the free swelling number, the bituminous coal should also have low ash

content, low sulfur content, and low phosphorus content. Because of these strict

requirements, only a very small portion of the bituminous coal is suitable for coke

production. These coking bituminous coal sources depletes at a fast rate. And as the

world iron and steel production increased dramatically, the rate of coke production could

not catch up with the demand.

The second problem is that the production of foundry coke is a process with high

energy demand and emits tremendous amount of pollutions. Foundry coke is produced by

pyrolyzing the coking bituminous coals at 1000 °C for about 36 hours. In a well managed

coke plant, where the tars and volatile matter released from coals during pyrolysis were

fed back as fuels, 5900 MJ of energy is consumed for per tone of coke produced.

Considering the energy contents of coke (which is typically about 30000 MJ/tone), 20%

of energy was consumed in the coking process. Other than the volatile matter, the coke

process also generates pollutions such as, CO2, SO2, particulate matter, and wastewater.

These two issues drove up quickly the price of the foundry coke which generated

huge interests in searching for a suitable replacement for the foundry coke and reduce the

cost for the iron casting industry. In the last century, many research activities were

carried out to look for possible coke replacement. These research focused mainly on the

utilization of other coal resources. One of the most notable applications is the production

7

of “form coke”(Easler et al. 1982; 1985; 1990; 1991; Gill and Chaklader 1984; Paul et al.

2002; Plancher et al. 2002; Schinzel 1981; Smoot et al. 2007). During the 1950s and

1960s, FMC Corporation and US Steel sponsored a research program in order to explore

the feasibility to convert non-coking coals into fuels with the quality of coke (Schinzel

1981). As a result of this decade long research, a new product “form coke” was invented.

The form coke process consists of several steps described briefly as following:

First, finely ground dried coal is pyrolyzed at about 500 °C; the volatile species

are condensed to form tar. The remaining char fraction is then carbonized at a higher

temperature (800 °C) to further remove any residual volatile matter; Then this char is

now ready for mixing and will become the main source of fixed-carbon in the form coke

briquettes (Schinzel 1981); Following its separation from the coal, the tar fraction is

subjected to an air-blowing stage to produce a heavy tar fraction by incorporating oxygen

into the tar structure. This accelerates internal condensation of products formed from the

decomposition of tar and improves the quality of semi-coke grain structure. The

carbonized char is recombined with the oxygenated binder and mixed at approximately

100 °C. The mixture is then pressed into the so-called green briquettes under pressure

(Schinzel 1981) . These green briquettes are subjected to a low-temperature (250 °C)

oxidation that promotes co-polymerization between the char and binder phases (Taylor

and Coban 1987a; b). During the process, briquettes also lose volatile matter. Cured

briquettes are finally subjected to a high temperature (900 °C) carbonization that reduces

briquettes to virtually pure carbon.

Through this form coke process many non-coking bituminous coal could be made

into fuel with similar quality as coke. Furthermore, some lignite could also be utilized as

8

form coke precursors. However, as described above, this form coke process is a relative

complicated process and involves many steps of pyrolysis and thermal treatment, which

consume huge amount of energy. Till now the application of form coke in iron casting is

very limited.

Because of its lower cost, anthracite is being considered again as cupola fuel.

However, there were many unsuccessful experiments on using anthracite in recent

decades. Several foundries had tried to add anthracite chunks into their cupola furnaces.

However because the combustion rate of anthracite chunk is much lower than that of

coke, the temperature in the cupola furnace dropped dramatically. More anthracite is

required to melt the same amount of iron. Therefore it may cost more to burn anthracite

in the cupola furnace than to burn coke.

The combustion rate of anthracite fines is much higher than that of coke, therefore

anthracite fines were also considered as the replacement of coke. During the last decade,

several foundries in Pennsylvania had tried to add anthracite fines and coke breeze

directly into their cupola furnace. However, the high speed upward air flow blew away

the fine particles before they reached the cupola burning zone. Most of these anthracite

fines and coke breeze were not combusted and finally were collected in the drop-out box

and bag house.

In the late 1970s and the early 1980s Reclasource Corporation, a subsidiary of

Berwind Corporation, developed a synthetic coke that was made from anthracite fines

and coke breeze (Rehder 1981). Anthracite fines and coke breeze were mixed first, and

then an organic binder was applied to bind the anthracite fines and coke breeze together.

This finally mixture was then made into briquettes, and supposed to be used as foundry

9

coke. This process accomplished the first step which is to bind the anthracite fines into

bigger structure. However, it did not address the issue of maintain the structure at iron

melting temperature inside the cupola furnace. After these briquettes were added into a

cupola furnace, the organic binders would lose their strength and the briquetted structure

will crush.

Therefore, the real challenge of coke replacement by anthracite fines is how to

maintain the binding strength at such high temperature in the cupola furnace. Also, the

binding materials that bind the anthracite fines at high temperature should not

significantly decrease the burning rate of the bindered structure.

In summary many waste energy and materials were generated from iron casting in

the traditional processes. With thoughtful engineering design, many of these materials

and energy can be reclaimed. By accomplishing this, the foundry industry will be

benefited both environmentally and economically.

1.3 Problem statement, objectives, and hypothesis of this study

With the stringent environmental regulations and increasing cost in energy and

raw materials, the foundry industry is facing tough challenges. Improvements for the

metal casting process are necessary to keep the foundry operations sustainable, and there

are many sustainable opportunities within foundry operations (Lee and Benson 2004). At

Penn State, research has being conducted to apply the advanced oxidation unit operations

in-situ in the metal casting process (Huang et al. 2009; Wang et al. 2007a; Wang et al.

2005; Wang et al. 2009; Wang et al. 2006; Wang et al. 2007b). This approach reduced the

emission and can reclaim waste raw materials from iron foundries at the same time.

10

However, there are still many sustainable opportunities in foundries that can be addressed.

First, the VOC emission from the metal casting is still a major environmental issue for

the metal casting process. Volatile organic compounds pose health threats to the foundry

personal and the local residents. Also emission permits on VOCs prevent some foundries

from expanding their production capacities. Second, the skyrocketing price of foundry

coke put tremendous financial burdens on the foundry industry. At the same time, there

are many valuable resources in foundries that can be reclaimed. The main proposal of this

study is to design sustainable engineering protocols to utilize the waste resources to

remove the volatile organic compounds emission and provide alternative fuel for the

foundries.

Improvements within the iron casting were designed first to address the volatile

organic compounds emission. Porous carbons that can adsorb volatile organic

compounds were produced in-situ by reclaiming the waste heat energy from the metal

casting process. The idea is to install simple and barely controlled heat exchange

instruments for the porous carbon production. The challenge is to generated robust pore

structures over a wide range of thermal conditions. Scientific research was conducted to

obtain further details on the pore formation and development within the carbon structures,

including the effects of pyrolysis temperature and duration, micro- and macro-structures

of the raw materials, and presoaked activation agents.

Sustainable engineering approaches were also designed to apply the byproducts

from food processing industry and anthracite mining industry to develop an alternative

fuel for the cupola furnaces in iron casting industry. Collagen binder, which is a

byproduct from food processing, was used to bind the waste anthracite fines from

11

anthracite mining. The bindered materials can be used as a coke replacement. Several

improvements are necessary for collagen bindered anthracite fines to become a successful

alternative fuel for cupola furnaces. Physical and chemical treatments were investigated

to improve the room temperature strength of the collagen binder. To increase the

mechanical strength at extremely high temperature, inorganic Si-containing materials

were added to the bindered anthracite fines. The preheat zone in the cupola furnace

provided the free heat energy for the silicon carbide binding to form. Several different

types of Si-containing materials were selected based on different redox level of the

silicon.

For this research, the following hypotheses were proposed regarding the in-situ

porous carbon production and foundry coke replacement:

1. Micro-structures, especially the graphitization stage of the raw carbon materials

will have significant effect on the robustness of the pore structure developed

under variable thermal conditions.

2. Marco-structures of the raw materials (mainly grain size) may have limited effect

on the final pore structure

3. Thermal treatment temperature may have significant effects on the pore structures,

as lower porosity will be the result of either a temperature too high or too low.

Thermal treatment duration has less effect on the pore structures after the pores

were full developed.

4. Presoaked water in the carbon pore structures may have limited activation

functions.

12

5. Crosslinking of collagen with organics containing aldehyde groups can increase

the strength of the collagen binder.

6. Si-containing materials added into the bindered anthracite fines can form silicon

carbide under the high temperature and reducing condition in the cupola furnace

preheat zone.

7. Silicon carbide is expected to provide strong mechanical strength for the bindered

anthracite fines at extremely high temperature, especially if silicon carbide

nanowires could be formed to connect neighboring anthracite particles.

8. The reaction between silicon (0) and anthracite fines to form silicon carbide is

very quick at high temperature.

13

Chapter 2 Pore structure development of in-situ pyrolyzed coals for pollution prevention in iron foundries

Abstract

A protocol was devised for preparing pyrolyzed coals that could be made in-situ

at foundries to capture volatile organic compound (VOC) emission. This pyrolysis

created extensive micropore volume in lignite over a broad range of temperature and time;

and could use waste heat from cupola exhaust gases by a heat-exchange tube. For

foundry application, moderate porous carbon with relatively uniform pores over wide

ranges of temperature and time would be more practical than highly porous activated

carbon (AC) that requires narrowly-controlled operations. This pyrolysis protocol was

developed in a thermogravimetric analyzer (TGA) and in a small tube furnace, while

using lignite, bituminous coal, and anthracite. The lignite yielded the most pore volume;

and this was relatively uniform (0.1-0.13 mL/g of pores) while temperatures were 600-

900 °C, and times were 0-60 minutes. Smaller grain sizes yielded improved porosity; and

this corresponded to more release of phenols and naphthalenes from smaller grains, as

discerned by TGA-mass spectroscopy (MS). TGA-MS also revealed that improved pore

development between 600-800 °C corresponded to the release of CO2 and H2O; and

concurrently higher slurry pH linked to less oxygenated functionality. Adsorption of

benzene was compared between the in-situ porous carbon and a commercial AC.

14

2.1 Background and objectives

Among the 189 hazardous air pollution (HAP) compounds that are listed in 1990

Clean Air Act Amendment (CAAA), at least forty have been identified in foundry

emissions. (Allen et al. 1991; Fox et al. 2002; Glowacki et al. 2003) These compounds

include such suspected carcinogens as benzene, toluene, xylene (BTX), and other toxic

volatilized organic species that might pose a threat to the health of foundry workers and

local residents. As a result, foundries receive pressure from local communities and the

EPA to control their HAP emission. Conventional end-of-pipe type emission control

devices would markedly increase costs of operating foundries in countries that practice

environmental restraint; and this would adversely impact US foundry competitiveness.

The intent of on-going research at Penn State, with collaborators, has been to devise

means that concurrently cut foundry operating costs while also decrease pollution and

material use (Fox et al. 2007; Fox et al. 2008a; Fox et al. 2008b; Wang et al. 2007a;

Wang et al. 2005; Wang et al. 2009; Wang et al. 2006; Wang et al. 2007b). In order to

achieve such innovative solutions one must thoughtfully and creatively analyze the flow

of materials and energy in an iron foundry.

Green sand mold casting is the most common casting process used in foundries.

During the pouring of the melted iron, the volatile matter of the bituminous coal in the

green sand mold is thermally released in a manner that provides a gaseous carbonaceous

blanket between the sand mold and iron surface. This carbonaceous blanket facilitates a

lustrous finish for the iron surface; and it also facilitates the separation of the iron from

the green sand mold in subsequent shake out (Dempsey et al. 1997; Green and Olender

15

1980; Wang et al. 1973). In this case the VOCs are also useful resources to the casting

process.

Roughly half of the foundries use cupolas, within which iron is melted as it mixes

with burning coke. The off-gas leaves the copula at temperatures of 1150 °C, and then

conventionally cools down in a drop-out box before being discharged into the atmosphere.

With such high temperature, a typical mechanized foundry with a cupola furnace

consumes about 5500 million joules energy to melt one ton of iron (Energetics-

Incorporated 1999). In contrast, the stoichiometric heat required is only 900 million

joules based on the iron’s heat capacity and change of phase energy. The temperature of

the exhaust air is similar to the temperature that commercial manufacturers used to

produce activated carbon (AC) (800-1000 °C) (Bansal et al. 1988). Moreover, VOC

emission could be usefully reclaimed as a resource for the gaseous carbonaceous blanket,

etc. if they could be cost-effectively captured on porous carbons. Commercially available

AC can capture such VOCs, but it is expensive relative to foundry operations. For this

reason, only a few foundries are using AC to control the VOC emission.

In light of the above, the objective of the research herein was to test the

hypothesis that by harnessing the waste heat from cupola exhaust gases, a coal source

could be pyrolyzed in-situ into a moderately porous carbon. This pyrolysis could occur

within a heat-exchange tube that could be introduced into the drop-out box. Meanwhile,

the volatile generated during the pyrolysis could be easily directed from the heat-

exchange tube to the cupola to avoid secondary pollution. Then the pyrolyzed porous

carbon could be used in the baghouse to adsorb the VOC emission; and finally the VOC

loaded porous carbon can be reused into the green sand mold. In this way the cost of

16

VOC pollution control could be substantially reduced for an iron foundry. During iron

pouring, only a small portion of the bituminous coals in green sand mold was heated and

released VOCs. Therefore it is not necessary for the in-situ porous carbon to contain very

high adsorption capacity. Rather it is important to have a product contain certain amount

of pore volume through board range of pyrolysis conditions, especially the pyrolysis

temperature which is the hardest to control in the in-situ pyrolysis situation. Thus, a

simple and reliable produce for production of the porous carbon in foundries will further

cut down the investment and operating cost. Several important factors for pore structure

development were tested in this study, such as the coal rank, the grain size of the raw coal,

the pyrolysis temperature, time, and the possibility of mild activation by pre-loaded water

contents in the raw coal. The in-situ pyrolyzed carbon was compared with a commercial

AC from similar raw materials in case of pore structure and benzene adsorption.

2.2 Experimental

2.2.1 Raw material

Pyrolysis tests appraised anthracite, bituminous, and lignite coals as raw materials.

The anthracite originated from Jeddo Coal Company (Wilkes-Barre, PA). The bituminous

coal was provided by Neenah Foundry (Neenah, WI), who obtained it from Massey

Energy Company (Richmond, VA). Experiments appraised two lignites from Mississippi

Lignite Mining Company: one from the Red Hills Mine in Ackerman MS, and the other

from Sabine Mine in Hallsville TX. The elemental analyses of these raw coals appear in

Table 1, along with that for pyrolyzed Red Hills lignite and Sabine lignite. Elemental

analysis for C, H, N and S was performed by a carbon-hydrogen-nitrogen determinator

17

LECO CHN600 and a total sulfur determination LECO SC132 (LECO Corporation, St

Joseph, MI), while O was calculated by difference.

The anthracite was received in small granular form and the bituminous coal was

received mainly in powder form (see Table 2 for detailed sieve analysis). While the

lignite sources arrived as chicken egg-size chunks. Before pyrolysis, the lignites were

pulverized by a SPEX 8000 mixer/mill from A.O. Smith Corp. (Milwaukee, WI) with 8

steel balls for 10 minutes. The product was sieved, and the pulverized coals were

separated for subsequent use; and then the unpulverized particles were refilled into the

SPEX mill to repeat the pulverization process. Several cycles of this pulverization-

sieving process yielded more than 95% of the lignite ending up in the US mesh

#200X600 (0.074 to 0.024mm) range.

Granular lignite was produced by passing lignite chunks through a coffee grinder,

and then sieving. The particles between US mesh #20 (0.841mm) and #50 (0.297mm)

were kept as granular lignite in this study. Powder lignite was used in all pyrolysis tests

herein unless otherwise identified. As compare to Table 2 US mesh #20X50 (0.841 to

0.297mm) is on the high end and US mesh #200X600 (0.074 to 0.024mm) is on the low

end of the range that foundries typically specify for green sand molds.

These original lignite coals contained about 30% moisture on a weight base. The

moisture content in lignite will change as it is exposed to the atmosphere, and the

moisture content offered the potential of mildly activating (i.e. gasifying) the in-situ

pyrolyzed coal. Therefore the raw coals used in this study were not pre-dried, unless

otherwise specified.

18

2.2 Pyrolysis

At bench scale, the raw coals were pyrolyzed by a TG-131 thermogravimetric

analyzer (TGA) from Thermo Cahn (Newington, NH). A basket made from mesh #400

(0.037mm) stainless steel wire cloth held about 1 gram of raw coal, and the TGA

pyrolyzed this while nitrogen gas flowed through the TGA at 100 mL/min. During this

protocol, the temperature in the TGA was kept at room temperature for about 2 minutes;

and then it increased at 50 °C/min to prescribed temperatures that ranged from 600 to 900

°C, for a specified pyrolysis duration that ranged from 0 to 60 minutes. That duration at

the plateau temperature has been defined herein as the “pyrolysis time”. For lab-scale

experiments, TGA pyrolysis has the advantage in control of various conditions. And the

biggest advantage of using TGA is that the real time mass loss can be recorded, which is

a critical message from pyrolysis. However the TGA rate was lower than what we would

expect in the foundry’s full-scale heat-exchange tube. Therefore, as a further factor, in

order to simulate the situation in the heat exchange tube and confirm the effect of faster

heating rate on pore structure development, we also employed a 5 cm (2”) diameter

vertical tube-furnace system for some of the more favorable thermal treatment conditions.

The vertical tube-furnace very rapidly pyrolyzed the raw coals. Specifically, this furnace

became preheated to the prescribed temperature with laminar flow of a nitrogen gas

stream up through the furnace’s tube. Then a 5 gram sample of coal was held in a basket

that was descended into the heated tube-furnace. After this, the thermocouple-recorded

temperature climbed from 30-40 °C to 800 °C within about 2-3 minutes, where the

temperature remained for the prescribed time. This protocol herein has been denoted as

tube-furnace pyrolysis. Representative tube-furnace pyrolysis trails were performed and

19

compared with TGA pyrolysis trails under the same conditions to confirm whether the

difference in heating rate between these two systems affected the pore structure of

pyrolyzed coals significantly.

2.3 TGA-Mass spectrophotometry (TGA-MS)

TGA-MS analysis was employed to monitor the emissions from coals during

pyrolysis. About 150 mg of coal samples were placed in a TGA 2050 (TA Instruments,

Newcastle, DE) and pyrolyzed from room temperature to 1000 °C with a heating rate of

20 °C/min under an argon atmosphere. The gaseous effluent from the TGA was carried

by argon through a Thermostar GSD 301T mass spectroscopy (Pfeiffer Vacuum Inc.,

Nashua, NH) for emission analysis. Additional details of this protocol are described by

Wang et al.(Wang et al. 2007b).

2.4 Characterization of in-situ pyrolyzed carbon

The pore volume distribution and specific surface area of in-situ pyrolyzed

carbons were determined by adsorption of argon vapor onto GAC samples by means of

46-100 point isotherms with progressively increasing relative pressures from 10-6 to

0.993 atm/atm by an accelerated surface area and porosimetry system (ASAP) 2010 units

from Micromeritics Instrument Corporation, (Norcross, GA). These analyses employed

the protocol of Moore et al.(Moore et al. 2001); Replication of pore volume distribution

analysis of a pyrolyzed carbon was conducted by pyrolyzing a new raw carbon sample

under the same conditions and then performing pore volume analysis on the ASAP again.

20

The difference between samples pyrolyzed under same condition was relatively small

(less than 5% difference in micro-pore volume). Duplications were conducted for most

pyrolysis test. Representative duplication results have been presented in the figures to

exhibit the reproducibility.

The slurry pH of a coal (pyrolyzed or raw) sample was measured by mixing 0.4

gram coal powder with 4 mL of water in a glass vial. For the pyrolyzed coals the protocol

was as follow: after 1 hour of pyrolysis at prescribed temperature, the pyrolyzed coals

were mixed with deionized water immediately after being cooled to room temperature

under nitrogen protection. The glass vial was sealed and shaken for 5 minutes. Then the

sealed vial was settled for more than 24 hours. After that the pH value of the water was

measured by an Accumet Model 10 pH meter from Fisher Scientific (Pittsburgh, PA).

Adsorptions of benzene on in-situ pyrolyzed carbon and lignite-based commercial

AC HD4000 (from Norit Americas, Marshall, TX) were performed in the TGA. Powder

carbon samples held by a stainless steel basket was loaded in the TGA. The temperature

was held at 60 °C all the time and only nitrogen was introduced at the beginning to

remove the pre-sorbed species until no more weigh change was recorded. Then benzene

was introduced at about 5000 ppmv in nitrogen gas. The weight change of carbon

samples was recorded by the TGA.

Adsorptions of VOCs released from sea-coal used in green sand mold were also

performed on two-section column tests. Each adsorption column contains a in-situ

pyrolyzed coal section (1 g pyrolyzed coals were packed) followed by a commercial

activated carbon (2 g commercial activated carbon were packed). The emission from 1 g

sea coal was passed through the column. Then carbons from each section of the column

21

were washed by CS2 and measured by GC-FID separately to determine which emissions

were captured by the in-situ pyrolyzed coal and which emission broke through.

2.3 Results and Discussions

2.3.1 Effect of pyrolysis temperature on pore structure development in raw coals of

several ranks

Anthracite, bituminous coal, and lignite were pyrolyzed at a range of temperatures

from 600 °C to 900 °C. The weight loss profiles of these coals during pyrolysis are

shown as derivative thermo-gravimetric (DTG) and weight loss curves in Figure 1. On

the DTG curves, the peak below 200 °C represents mainly the release of physically

adsorbed water (moisture content) in the raw coal; the other peak at about 500 °C

represents the release of volatile matter from the raw coal. For pyrolysis of lignite and

bituminous coals, Figure 1 reveals that a considerable amount of mass released between

600 °C and 800 °C; however beyond 800 °C the weight changed only slightly. After

being pyrolyzed at the temperatures listed above, particles of this particular bituminous

coal (seacoal), agglomerated into a single composite, whereas the lignite and anthracite

coals did not agglomerate.

The pore volume distributions of these pyrolyzed coals are shown in Figure 2.

There are several major results indicated by the figure as listed below:

1. The rank of coal yielded a significant effect on the pore volume distribution of the

in-situ pyrolyzed carbon.

2. For both the bituminous coal and lignite coal, the pore volume increased as the

pyrolysis temperature increased from 600 °C to 800 °C, and then for the

22

bituminous coal, the pore volume decreased significantly as temperature increased

yet further to 900 °C. For the lignite coal, pore volume decreased slightly as

temperature rose from 800 °C to 900 °C.

3. The pore structure in pyrolyzed lignite was much more stable over the broad

temperature range of 600 °C to 900 °C.

4. The pyrolysis process produced mainly micropores in the coals.

Pyrolysis of coal has been studied extensively with a relatively small amount of

research pertains to the change of pore structures during pyrolysis (Alonso et al. 2001;

Centeno et al. 1995; Chattopadhyaya et al. 2006; Chiche et al. 1965; Cypres et al. 1985;

Feng et al. 2006; Kok 2003; Maloney and Jenkins 1985; Nandi et al. 1964; Oda et al.

1981; Puente et al. 2000; Radovic et al. 1985; Radovic et al. 1983; Sakintuna et al. 2004;

Simons 1979; 1983; Simons and Finson 1979; Singla et al. 1983; Solomon et al. 1992a;

Solomon et al. 1992b; Toda 1973a; b; Toda et al. 1970; Tomeczek and Gil 2003; Xu et al.

1994; Yu et al. 2007). In the literature, during pyrolysis, releasing of volatile matter was

accepted as the reason of increasing pore volume in coals while graphitization at higher

temperature was considered as the cause of pore closure (Franklin 1951; Marsh and

Menendez 1988; Nsakala et al. 1978; Sakintuna et al. 2004; Strugala 2002; Toda 1973b;

Tomeczek and Gil 2003). However it is known that most of the volatile matter released

when the temperature reached 600 °C (Berkowitz 1985; Jenkins et al. 1973; Nsakala et al.

1978; Sathe et al. 2002). Results in Figure 2 showed substantial amount of increase in

pore volume between 600 °C and 800 °C, although the mass loss between these two

temperatures was very small as shown in Figure 1. In Figure 3, TGA-MS revealed that

majority of the volatile matter released between 400 °C and 600 °C. On the other hand,

23

for CO2 and H2O additional peaks released at 800 °C; and increase of CO release at 800

°C. Release of COx and H2O reflects breaking of oxygen functional groups. Above 600

°C the primary oxygen substituents that would remain in the graphene structure would be

the oxygen in ether (which is the crosslink between graphene layers), and the oxygen in

quinoid groups (Berkowitz 1985; Figueiredo et al. 1999; Jiang et al. 1988; Leon y Leon

D and Radovic 1994; Nsakala et al. 1977). Result from Figure 3 confirmed that breaking

of crosslinks between graphene layers at higher temperature created more pores in the

pyrolyzed coals between 600 °C and 800 °C.

Therefore, there are three major mechanisms that direct the development of pore

structure at different temperatures: release of volatile matter around 550 °C generates

major portion of the pores; break of crosslinks around 800 °C generates additional pores;

graphitization at temperature above 800 °C closes up some pores. Results from Figure 2

suggest that lignite is the most suitable precursor for the in-situ porous carbon because

pyrolyzed lignite not only provides the highest pore volume, but also has the most stable

pore structure within the tested pyrolysis temperature range.

2.3.2 Effect of pyrolysis time and tube-furnace pyrolysis on pore structure

development

Sabine lignite was pyrolyzed in the TGA by ramping the temperature up to a

plateau of 800 °C, and then leaving the temperature at 800 °C for the times of for

different plateau time durations from 0 minutes to 60 minutes. The pore volume

distributions that resulted from this pyrolysis exposure appear in Figure 4. As shown, the

micropore volume was highest with 0 minutes of plateau temperature time; then was 13%

less than that with 7.5 minutes of plateau time, then 8% less than at 0 minutes when the

24

plateau time was 15-60 minutes. The lower pore volume at 7.5 minutes may have

reflected that the graphene planes were continuing to rearrange during the initial stages of

an 800 °C plus temperature regime. This trend could also have been reflected in the

Figure 3 TGA-MS responses, which showed that as a lignite experienced temperatures

proceeding above 800 °C, the CO gas continued to be released; and this was an indicator

that oxygen bonds were continuing to be broken-perhaps as the graphene planes

continued to rearrange. Several other authors have posed the notion that pore closing can

occur via structural transformations that lead to graphitization; and these transformations

can occur at temperature as low as 700-800 °C (Chiche et al. 1965; Green et al. 1982;

Toda 1973a; b). After 7.5 minutes, the rearrangement of the graphene layer groups was

still proceeding. Also it may be possible that 7.5 minutes are not long enough for some

volatile compounds to diffuse out of the coal bulk. Then the pore volume could have

stabilized by 15 minutes of pyrolysis and thereafter. This interpretation is concurrent with

the weight loss profile of lignite pyrolysis at 800 °C as shown in Figure 1, where the

weight of pyrolyzed lignite continued to decline during the first 5-10 minutes of pyrolysis

at 800 C, but then remained stable with no further thermal decompositions after about 10

to 15 minutes of pyrolysis.

However when pyrolyzed at 900 °C, distinct difference lays in pore volume

distribution between the coals pyrolyzed for 15 minutes and 60 minute. This result

indicates that after 15 minute pyrolysis at 900 °C the lignite coal was continuing to

graphitize. Therefore for proposed full scale application shorter pyrolysis time between

10 and 15 minutes is recommended to minimize the effect of graphitization.

25

Since significant difference in heating rate can affect the properties of pyrolysis

product (Cai et al. 1996; Chitsora et al. 1987; Cui et al. 2007; Hayashi et al. 2002;

Hayashi et al. 2000), Sabine lignite was also pyrolyzed by the tube-furnace pyrolysis

system, where the heating rate was similar to that anticipated in a foundry’s full scale

heat-exchange system. The pore volume distributions of the tube-furnace pyrolyzed

carbons (Figure 5) exhibited the same results following either 15 or 30 minutes pyrolysis.

Moreover, by comparing the Figures 4 and 5 data for 15-30 minutes pyrolysis, one

observes that the difference in heating rates between the tube-furnace pyrolysis and TGA

pyrolysis yielded no discernible differences in pore structure. Also, this comparison

verified that the lab scale TGA pyrolysis at a 50 °C/min heating rate can closely represent

a situation that more comprehensively mimics full-scale conditions.

2.3.3 The potential of using moisture content and presoaked water as the activation

agent

Under certain pyrolysis conditions, the native moisture contents in raw lignite

may have the potential to mildly activate the pyrolyzed coal (Hayashi et al. 1999; Yip et

al. 2007). TGA pyrolysis of pre-dried (105 °C overnight) and un-dried Sabine lignite was

compared; and the resulting pore volume distributions were identical, as shown in Figure

6.

Thus, although steam is a common activation agent in producing activated carbon,

the moisture content of this raw lignite did not influence pore development in the TGA.

This lack of influence may have occurred because: (1) with gradual temperature increase,

all this water evaporated by the time the TGA reached activation temperatures; (2) the

amount of steam generated from natural moisture content was not adequate, especially at

26

high temperature where the activation reaction occurs; or (3) there were not sufficient

opened active sites in the raw coal to react with steam.

In order to increase the amount of in-situ water, raw lignite was mixed with de-

ionized water for 24 hours and then the mixture (containing 38% raw lignite and 62%

water), was held within a #400 mesh (0.037mm) basket so as to pyrolyze in the TGA at

800 °C for 15 minutes. Furthermore, this same procedure was also performed on a coal

sample which had already been pyrolyzed in TGA at 800 °C for 15 minutes (the mixture

also contained about 38% coal and 62% water). By mixing pyrolyzed coal with water and

then pyrolyzing the mixture again, some features of the traditional steam activation

process were simulated; only here, the water was pre-soaked into the pyrolyzed lignite in

a manner that would eliminate the need for foundry operators to control steam flow rates.

Results presented in Figure 6 show 10% increase in pore volume by pre-loading water in

raw coal. Yet another 10% increase in pore volume was achieved by pre-loading water

into pyrolyzed coal. Yet further, when employing a tube-furnace protocol on pre-

pyrolyzed coal that was water-soaked, the resultant pore volume distribution was nearly

the same as when coal with this preparation protocol was heated gradually in the TGA.

This meant that some water remained available for activation in both these cases. Thus,

the soaked water influence pore development during both the tube-furnace and TGA runs.

The presence of remaining water in the TGA runs could be attributed to two factors: First,

the nitrogen flowrate in the TGA reaction-tube chamber was low relative to the amount

of water retained within the carbon grains within the TGA. Also, this soaked-up water

had penetrated into the pores and needed longer time and higher energy to fully diffuse

out of the pores. These results confirmed the second and third hypothesized limitations of

27

utilizing native moisture content for mild activation in the pyrolysis systems tested in this

study.

2.3.4 Effect of raw coal grain size on pore structure development

In order to be reused in green sand molds, the grain size the in-situ pyrolyzed

carbon is required to within the range of that of the seacoal used in foundries. When we

compare the extremes of size, #20X50 (0.841 to 0.297mm) vs #200X600(0.074 to

0.024mm), we want to see how much the grain size will affect the pore structure

development during pyrolysis. There is a few data available in literature regarding the

effect that coal grain size has on pore structure development during pyrolysis, as

acknowledged by several authors. (Centeno et al. 1995; Duz et al. 2005; Gavalas 1982;

Hanson et al. 2002; Howard 1981; Kok et al. 1998; Nsakala et al. 1977; 1978; Yu et al.

2007). Several have appraised the effect of particle size on surface area (Centeno et al.

1995; Nsakala et al. 1977; Yu et al. 2007). Still none of them discussed the possible

mechanism of the particle size effect. In this study, Sabine lignite with both #20X50

(granular) and #200X600 (powder) grain sizes experienced pyrolysis in the TGA at 800

°C. The pore volume distribution results appear in Figure 7.

As shown in Figure 7, the pore volume decreased greatly with increasing grain

size. However the granular lignite with pre-soaked water developed similar amount of

pore volume during pyrolysis as did powder lignite with pre-soaked water. This result

indicates that the effect of grain size on pore structure development could be easily offset

by mild activation from the pre-soaked water. In literature, one hypothesis is that the

transport of volatile matter from within coal to its exterior is limited in larger particles

(Gavalas 1982). This effect would keep some volatile matter trapped within the coal

28

which could render less weight loss and pore development for the granular coal than for

the powdered coal.

In order to appraise this, we monitored the weight loss profiles of both grain sizes

during pyrolysis (following overnight drying at 105 °C). The results appear in Figure 8.

In contrast to literature-derived expectations, the weigh loss profiles of granular and

powder Sabine lignite were virtually identical. However, in contrast, the TGA-MS

profiles of granular and powder lignite did indeed reveal telling distinctions regarding the

pyrolysis volatile releases from these two different coal sizes (Figure 9).

The species that yielded the higher ion currents included low molecular weight

compounds, such as H2O, CO2, CH4, and BTX the amount released from granular lignite

was very close to those from powder lignite. Release of H2, H2O, CO, CO2, and CH4

corresponds to more fundamental changes of pore structure development during pyrolysis,

such as the losing of functional groups and graphitization. In contrast, the phenols, and

naphthalenes were released far less from granular lignite than from powder lignite. These

compounds are general larger in size and probably will be restricted in diffusion in the

narrow pores. Therefore the difference in pore volume between pyrolyzed powder and

granular lignite is most likely caused by the blockage of pore structure by slow diffusive

compounds. This kind of blockage could be broken by steam activation. And this may be

the reason why the pre-soaked granular lignite yielded similar amount of pore structure as

the pre-soaked powder lignite.

On a similar note, Suzuki et al. observed that when phenol and naphthol are

loaded onto activated carbons, only limited portions of these compounds become released

during a thermal regeneration (Suzuki et al. 1978).

29

Therefore in summary, smaller raw lignite grain size will favor the development

of pore structure by promoting quick release of the species that have higher molecular

weights, lower diffusion rates, and lower propensity to become thermally released.

2.3.5 Surface acidity change of lignite after pyrolysis

Pyrolysis not only decreases the amount of H and O in coals (as shown in Table

1), it also will reduce the coal’s acidity. These effects were inter-related: many of the

oxygen-containing functional groups cause acidity; and these include carboxyls and

lactones (Menendez et al. 1996). Specifically, as shown in Table 3, slurry pH increased as

the pyrolysis temperature increased. In particular, a comparable increase in pH was

detected from the raw coal to the 600 °C product, as from 600 °C and 800 °C, although

the mass loss from 600 °C to 800 °C was much smaller than from ambient to 600 °C.

This increased slurry pH is important in foundry applications for two reasons: first, green

sand molds perform better at higher pH levels (Neill et al. 2001); and a significant reason

that lignites become employed in green sand systems so infrequently is because raw

lignite adversely suppresses the green sand mold pH. Second, activated carbons generally

adsorb VOCs more favorably when the carbons host a higher pH (Nowack et al. 2004).

2.3.6 Compare the pore volume distribution and benzene sorption of in-situ

pyrolyzed carbon with commercial lignite-based AC

Figure 10 presents the pore volume distribution of pyrolyzed Sabine lignite (TGA,

800 °C, 15 minutes), as compare to a commercial lignite-based AC (HD4000) that we

obtained from Norit Americas (Marshall, TX). By comparing these two carbons, one

realizes that the larger pores (20-500Å wide) in lignite-based AC were created during the

30

steam and CO2 activation process. The in-situ pyrolyzed carbon contained as much

micropore volume, in the <13Å range as did the commercial lignite AC. The micropore

volume generally dictates adsorption capacity of small VOCs, while mesopore volume is

an important factor related to diffusion rate. Normally, an adsorbent with sufficient

mesopore volume needs less time to reach adsorption equilibrium.

Adsorption of benzene in nitrogen gas by in-situ pyrolyzed carbon and a lignite

based commercial activated carbon was monitored by the TGA (shown in Figure 11). The

recorded weight change showed a faster adsorption of benzene by the in-situ pyrolyzed

carbon (Sabine lignite powder 800 °C in TGA for 15 minutes) in the initial 2 minutes,

such after 2 minutes the in-situ pyrolyzed carbon adsorbed 1 mg benzene/g carbon; and

HD4000 had adsorbed 50% as much. However after a 10 hour pseudo-equilibrium time

(adsorption equilibrium reached on the commercial AC 4 hours after the adsorption

began, and never reached on the in-situ pyrolyzed carbon), the adsorption amount on the

commercial AC was about 2.5 times of that on in-situ pyrolyzed carbon.

Figure 12 shows the adsorption performance of the in-situ pyrolyzed porous

carbon on different VOC species generated from the sea coal used in green sand molds.

The in-situ porous carbon had very good performance in removing benzene, toluene,

xylene and other VOC species that contain more than 6 carbon atoms in the molecular

structure. The in-situ porous carbon had adsorbed plenty of these larger species, and none

of them was detected in the breakthrough check portion. The removal of smaller species

such as methane was not completed. As shown in Figure 12, although some of the

methane was detected in the in-situ porous carbon, it can also be detected in the

commercial activated carbon section.

31

2.4 Conclusion

Pyrolyzing lignite in-situ at a foundry is a potential way to produce a cost-

effective adsorbent for controlling VOC emission in iron foundries. The similarity of the

pore volume distributions of tube-furnace pyrolyzed lignite and TGA pyrolyzed indicates

that TGA pyrolysis at 50 °C/min can closely represent the pyrolysis condition in a bench-

scale tube-furnace. Further anticipation is that both these protocols can usefully simulate

the thermal treatment condition of the proposed heat-exchanger tube. The lab scale

experimental results show that lignite-based in-situ pyrolyzed carbon is more stable

product than bituminous coal or anthracite. Indeed the pyrolyzed lignite product remained

relatively uniform in pore structure over a relatively board range of thermal temperature,

thermal time, and native coal moisture content. For a temperature range from 600 °C to

900 °C, and thermal durations of 7.5 to 60 minutes, the in-situ pyrolyzed carbon

produced from lignite reliably contained a considerable amount of pore volume.

The development of pore structure in coals during pyrolysis is mainly caused by

the releasing of volatile matter, COx and H2O within the coal structure. This pyrolysis

process produced mainly micropores; and the amount of pore volume with width <13Å

was comparable to steam activated lignite-based commercial AC. The native moisture

contents in raw lignite did not facilitate more pore volume development in the pyrolysis

systems appraised in this study, because of inadequate amount of steam remained at high

temperature and there were not enough open active sites in raw coals as compare to

carbonized coals. The grain size of the raw lignite did affect pore development; and

larger grain sized corresponded to lesser pore development, because the larger and more

sterically hindered VOCs such as naphthalenes and phenolics could not escape from the

32

larger grains during pyrolysis. However mild activation from pre-soaked water in

granular lignite can eliminate this effect and open the blocked pore volumes during

pyrolysis.

33

Table 2-1 Elemental analysis of the raw coals

Proximate analysis % Ultimate analysis % (dry ash free) Coal Moisture V.M.

(dry) F.C. (dry)

C H N S O Ash (dry)

Anthracite1 3.0 4.99 82.3 94.3 2.25 0.89 0.38 2.59 12.8Bituminous1 Coal 5.1 33.4 56.5 86.4 5.74 1.68 0.79 6.22 10.9Red Hills Lignite 25.8 26.5 48 68.42 5.79 1.26 0.90 23.62 25.5Sabine Lignite 33.2 32 49.6 70.51 6.03 1.29 0.96 21.20 18.4Pyrolyzed Red Hills Lignite2 N/D N/D N/D 96.48 1.05 1.76 0.85 0 36.1Pyrolyzed Sabine Lignite2 N/D N/D N/D 94.86 1.08 1.65 0.84 1.56 22.41 As determined by suppliers 2 Pyrolyzed in TGA, HTT 800 °C, heating rate 50 °C, pyrolysis time 1 hour. V.M.-volatile matter; F.C.-Fixed Carbon; N/D-not measured

34

Table 2-2 Sieve analysis of the received anthracite and bituminous coal

Anthracite Bituminous Coal (Seacoal)1

Mesh number (size) Retained % Passing % Retained % Passing % 28 (0.60 mm) 60.0 40.0 5.1 94.9 35 (0.42 mm) 18.5 21.5 7.1 87.8 60 (0.25 mm) 14.5 7.0 15.8 71.9 100 (0.15 mm) 4.5 2.5 16.3 55.6 200 (0.075 mm) 1.5 1.0 21.9 33.7 300 (0.045 mm) 0.5 0.5 10.7 23.0

1. Provide by Neenah foundry, with sieve sizes as conventionally specified and used

35

Table 2-3 Change of slurry pH of powdered lignite after pyrolysis in TGA for 1 hour Pyrolysis temperature Coal Source

Raw 600 °C 800 °C

Sabine 6.0 8.6 11.1 Red Hills 6.0 9.3 11.7

36

Figure 2-1 TGA recorded mass loss of different coals during pyrolysis (Temperature

started increasing at about 2 minutes, with the rate of 50 °C/min to the prescribed temperatures)

37

Figure 2-2 Pore volume distributions (listed in descending order) of several ranks of coals

pyrolyzed in TGA at 600-900 °C for 1 hour. (Heating rate 50 °C/min)

38

Figure 2-3 Evolution of Methane, BTX, H2O, and COx from coals during pyrolysis (L-

left y axis; R-right y axis)

39

Figure 2-4 Pore volume distributions of Sabine lignite powders pyrolyzed in TGA at 800

°C and 900 °C for different pyrolysis time

40

Figure 2-5 Pore volume distribution of Sabine lignite pyrolyzed by tube-furnace pyrolysis (800 °C)

41

Figure 2-6 Effect of different water resources on the development of pore structures in

powdered Sabine lignite during pyrolysis (All pyrolysis was conducted at 800 °C for 15 minutes)

42

Figure 2-7 Pore volume distribution of in-situ pyrolyzed carbon from powder Sabine

lignite and granular lignite. (TGA pyrolysis at 800 °C)

43

Figure 2-8 Mass loss of powder and granular Sabine lignite during pyrolysis at 800 °C, with a heating up rate at 50 °C/min

44

Figure 2-9 Volatile releases from granular (········) and powdered (——) Sabine lignite during pyrolysis detected by TGA-MS

45

Figure 2-10 Pore volume distribution of in-situ pyrolyzed carbon from Sabine lignite (TGA pyrolysis, 800 °C for 15 min) and commercial lignite-based activated carbon

46

Figure 2-11Comparing the adsorption of benzene onto the in-situ pyrolyzed porous

carbon and a commercial activated carbon

47

Figure 2-12 Adsorption of typical volatile organic compounds from green sand molds

onto the in-situ pyrolyzed porous carbon

48

Chapter 3 Binding waste anthracite fines by low emission renewable binders and ceramic materials for foundry coke replacement

Abstract

Bindered waste anthracite fines were proposed as an alternative fuel to replace the

expensive and depleting foundry coke. In order to increase the mechanical strength at the

iron melting temperature, several Si-containing materials were introduced with the

bindered anthracite with the aim of inducing binding strength that would hold at high

temperatures (1400 °C). The binding mechanism of each Si-containing materials,

especially the formation of desirable silicon carbide, was dictated by the redox level of

the silicon in the Si-containing materials. Zero valent silicon in the silicon powder

induced the greatest strength after pyrolysis at 1400 °C, by generating silicon carbide

nanowires (SCNWs) that spanned between the anthracite grains. Silicon carbide that did

not provide binding strength was also formed from amorphous Si (+4). No silicon carbide

was formed from crystallized Si (+4). The post-pyrolysis strength of SCNW bindered

anthracite pellets was also affected greatly by the anthracite grain size and the silicon

content. The exothermic reaction between elemental silicon and carbon is favored for the

cupola furnace operation. Furthermore, the physical and chemical conditions within a

cupola furnace allowed a sustainable engineering solution to be designed to rearrange the

materials flow into the cupola furnace which improves the sustainability of the iron

casting process. Replacing coke by waste anthracite fines allows industry to avoid the

energy consumption and CO2 emission that inherently occur when bituminous coals are

49

coked at elevated temperatures before they are used in cupolas. This also offers the

opportunity to reduce costs and pollution.

3.1 Background and objectives

The price of energy sources and raw material has increased dramatically in

recently years. For iron foundries using cupola furnace, foundry coke is the universal fuel

and foundry coke’s price increased 450% from the year 2002 to 2008. The market price

for a short ton of typical foundry coke reached as high as $570 in 2008, and remained at

$500 till 2010. The foundry industry in the United States is carrying a tremendous burden

in securing raw material and controlling energy cost; and foundries are actively seeking

alternative options.

Coke has become the universal fuel for iron casting cupola furnaces since the

middle of the 19th century. Coke is produced by pyrolyzing special coking bituminous

coals that will fuse into a highly porous rigid structure when they have experienced about

1000 °C for 14 to 36 hours (Avallone et al. 2006; Kirk 1903; World-Bank 1998). After

this coking procedure, the product coke displays desirable properties for the cupola

furnace operation. These properties include high fixed carbon, low volatile content, low

sulfur, developed porosity and enlarged surface area from a strong fused-carbon structure

which allows a fast burning fuel bed to be built up above the molten iron(Avallone et al.

2006). Moreover, the coke’s rigid fused structure allows the coke to maintain its

structural integrity while it is roughly handled and heated in the cupola’s heat zone and

burned throughout in the cupola’s melt zone. With its high combustion rate and carbon

contents, a ton of foundry coke will melt 10 tons of iron in cupola furnace (Kirk 1903).

50

However, only a few coal sources will form (fused-porous) coke when they are

pyrolyzed. As an inherent limitation, the very production of coke consumes 15-25% of

the energy of the raw coal. It is worth noting that as an alternate to conventional coking,

the formcoke technology emerged in the 1970s (Easler et al. 1982; 1985; 1990; 1991;

Gill and Chaklader 1984; Paul et al. 2002; Plancher et al. 2002; Smoot et al. 2007). This

process involves briquetting of coke and other raw coals with a binder material generated

by pyrolyzing these coal sources. For the formecoke technology, the raw coals can be

non-coking bituminous coal or lignite. However, this process inherently does not

circumvent the pre-coking energy requirements and thus saves no energy relative to the

conventional coking of bituminous coal.

There are three significant zones in a cupola: the drop zone, preheat zone and melt

zone (Figure 3-1). The solid carbon fuel should host high strength and resistance to

shattering when roughly handled and dropped 5-15 feet in the midst of scrap irons; then

in the preheat zone it should hold its structural integrity as newly added scrap iron and

fuel was piled on top of it; yet further it should hold its structural integrity in the meltn

zone to build a fuel bed above the molten iron. Coke's fused porous carbon structure

processes enough strength for the cupola operation. Also this porous carbon structure

renders a very fast burning rate that keeps the temperature high in the cupola.

The ultimate challenge for a successful alternative fuel is that the new fuel should

exhibit a fast combustion rate and enough mechanical strength to hold its structural

integrity under the cupola operational conditions. With regard to carbon content,

anthracite has similar carbon content and energy as coke. Indeed, anthracite once was a

widely used fuel source in cupola furnaces before the emergence of coke (Kirk 1903).

51

However, the combustion rate of solid anthracite chunks is slow due to its condensed

structure and low exposed surface area. The ratio of anthracite coal-to-melted iron has

been 1:5 when using 8-18” diameter anthracite pieces; and 1:8 when using 4-8” diameter

pieces(Kirk 1903). But this is lower than the 1:10 ration with coke (see above). With yet

smaller size anthracite pieces, the iron melting ratio can be further improved, but when

the grains become too small, the intense air flow can blow the fine anthracite grains out

of the cupola and into the bag-house before they burn. Because of this confounding

limitation, iron cupola furnaces and other industrial operations pose minimum size

specifications for their solid carbon fuels. Therefore, anthracite fines often represent an

unusable residue from the anthracite mining process; and they are often discarded in the

valleys of eastern Pennsylvania’s coal region. The opportunity explored herein has been

to devise a mean of binding these anthracite fines together in a manner that will allow the

bindered chunks to be suitable for use in cupolas. Others have attempted binding

anthracite fines with organic binders, but these have not offered an adopted coke

replacement (Rehder 1981). The authors herein are not aware of any other viable binding

of anthracite fines that inherently alleviates the energy requirement for pre-coking while

also providing fast burning and structural integrity.

Organic binders along would be inadequate because the organics would be

decomposed as the bricks experienced elevated temperatures in the cupola; and then the

bricks would lose their strength and disintegrated into fines that would incur a pressure

buildup through the tuyeres, and inadequate oxygen transfer to the carbon fuel. Without

adequate oxygen transfer, the fuel could not burn hot enough to melt the iron; and this

may damage the whole melting process. Thus, in over view, the biggest challenge for

52

every fuel to be successful in the cupola furnace is to maintain a structural integrity in the

melt zone. The paradox is that many of the binders used for ambient condition binding

can not maintain their strength at the iron melting temperature.

The aim herein has been to utilize, as much as possible, the materials components

that are introduced into a cupola already, and reconfigure these into a bindered composite

that effectively replace (at least a portion of) the coke, while also offering the favorable

features of coke. In typical iron cupola operations, for every 1000 pounds of scrap iron

charged into the furnace, there are also approximately 100 pounds of coke, 1 pounds of

silicon, and 20 pounds of limestone, along with other additives. The overall dwell time of

these materials in a cupola is typically 40-80 minutes. This materials flow avails typically

about 20-40 minutes in the preheat zone where the materials experience 1200-1400 °C in

a starved-air (pyrolysis) environment.

The elemental metallic silicon is added into cupolas to form carbide and improve

the cast iron quality. One premise has been that since silicon must be charged into the

furnace anyway, it would be useful to bind together the silicon and anthracite fines with a

collagen-based binder, and feed these all together into the cupola as bricks. By this

strategy, the organic binders could provide the ambient-temperature strength during

storage and handling into the drop zone. Then the silicon metal (or other Si source) could

react with anthracite in the 1200-1400 °C preheat zone to form SiC that could hold the

anthracite fines together while the bricks subsequently burn in the melt zone.

If bindered bricks of anthracite fines with Si could be used, the modified materials

flow would be as shown in Figure 3-1. Thus the focus of this study was to employ an

array of materials that contain silicon at different redox levels (0 and +4); and bind these

53

to anthracite fines with collagen. Then discern which Si source facilitated the formation

of a briquetted product that would exhibit high unconfined compressive strength after it

experienced high temperature pyrolysis.

The organic binder used to bind all materials at room temperature has been a

collagen-based binder produced from food byproduct by Hormel Food Company (Austin,

MN). Physical and chemical procedures were also tested in this study to enhance the

binding strength of this low emission renewable binder at ambient conditions. Preheating

of the collagen binder and water mixture helped to denaturalize the collagen structure and

form a gelatin solution. Also the amino acids in the protein peptide chain structure could

react with chemicals that contain aldehyde groups such as sugars (Duan and Sheardown

2005; Eyre and Wu 2005; Fujimoto and Horiuchi 1986; Tanzer 1973). Via this reaction

the aldehydes can crosslink with the peptide chains in the collagen structure to provide

more binding strength (Chan and So 2005; Pieper et al. 2000; Sajithlal et al. 1998a; b;

1999; Sell et al. 2005; Usha and Ramasami 2000; Verzijl et al. 2002).

Silicon is the second most abundant element in Earth’s crust after oxygen, making

up 25.7% of the crust by mass. Silicon typically exists as various forms of silicon dioxide

(silica) or silicates. Silicon in these species is the +4 valence. Pure silicon crystals (zero

valent) are very rarely found in nature. They usually exist in volcanic exhalations which

had experienced extremely high temperature and pressure conditions. Zero-valent silicon

can be produced produced commercially by the reaction of high-purity silica with wood,

charcoal, and coal, in an electric arc furnace using carbon electrodes. The silicon

produced via this process is called metallurgical grade silicon and is at least 98% pure.

Ultra high temperature (~1900 °C) and huge energy input are also required for this

54

reaction. As a result the price of metallurgical grade silicon is also very high ($1.4 per

pound).

Theoretically silicon carbide can be formed more easily by reacting carbon with

zero valent silicon than Si (+4) species (equations [1] and [2]). However, the high price

and energy demand of zero valent silicon make it economically and environmentally

unfavorable to introduce additional zero valent silicon into the cupola furnace. The

typical amount of silicon added in cupola operation is about 1% of the coke added. For

example, if 5% zero-valent silicon is required to provide enough strength in the bindered

bricks, the bindered anthracite bricks can only replace 20% of the foundry coke without

introducing additional zero valent silicon.

Si (+4) species are much cheaper than the zero valent silicon. If silicon carbide

can be formed between anthracite and Si (+4) species under the conditions in the cupola

preheat zone to provide binding strength to the anthracite bricks, it may eliminate the

addition of zero valent silicon into the cupola. This improvement can further decrease the

operation cost and energy demand of the whole metal casting process. However, the

reaction between carbon and Si (+4) to form silicon carbide requires energy input (as

shown in equation [2]). It is unknown whether silicon carbide could be formed under the

conditions in the cupola pre-heat zone. Furthermore, the energy input for the formation of

silicon carbide may lower the temperature in the cupola furnace, which is unfavorable to

the cupola operation. The Si (+4) species can be further divided into two groups. The first

group is crystallized Si (+4) species and the second group is amorphous Si (+4) species.

Amorphous Si (+4) species can react slightly favorably with carbon due to their non-

crystallized structures.

55

3.2 Experimental

3.2.1 Raw materials

Anthracite fines used in this research were obtained from Jeddo Coal Company

(Wilkes-Barre, PA). These were Jeddo’s “#5” anthracite fines (identified as “anthracite

fines” herein). The sieve analysis of Table 1 shows that most of the anthracite fines are

with in the range of US mesh # 10 to 80 (2000 to 177 μm). For some experiments, these

coal fines were also crushed into powders that can pass through a U.S. mesh #100 (150

μm) sieve (identified as “anthracite powder” herein). The approximate analysis and the

heat content of the anthracite fines were compared with a typical foundry coke in Table 2;

and the dry heat content of the anthracite fines was nearly the same as for the coke. The

foundry coke was that which has been used by Ward Manufacture (Blossburg, PA). The

collagen-based binder was provided by Entelechy representing Hormel Foods Company

(Austin, MN), and it was received as dry small granular form. Fructose (98%-102%) was

obtained from Mallinckrodt Baker, Inc. (Phillipsburg, NJ).

Materials that contain silicon at three different redox levels were tested. These

levels included metallic Si (0), amorphous Si (+4), and crystallized Si (+4). Elemental

silicon lumps (10cm&down, 98.4% purity) were purchased from Alfa Aesar (Ward Hill,

MA). The lumps were crushed into powders (less than #100 mesh) before being added

into the anthracite. The Amorphous Si (+4) containing materials was sodium silicate

(Na2SiO3) solution; and it was provided by J.B. DeVenne Inc. (Berea, OH). Sodium

silicate is also a frequently used inorganic core binder that can resist high temperature

conditions. The crystallized Si (+4) was kaolinite powder (KGa-1, Al2Si2O5(OH)4) that

was obtained from Washington County, GA.

56

3.2.2 Anthracite pellet preparation

Anthracite fines were dried at 105 °C overnight to remove the moisture content

first. The collagen binder (1g, typically) was dissolved in 12 g water at 70 °C (unless

otherwise listed) so as to form a gelatin solution. Fructose, when added, was dissolved

into the water together along with the collagen binder. When the Si-containing material

was in powder form, it was mixed with the anthracite fines first. If instead the Si-

containing material was a solution (i.e. sodium silicate), it was added in the gelatin

solution before being mixed with the anthracite grains. The final mixture was packed into

a cylindrical mold (2.86 cm in diameter, 4.76 cm in length) with 275 Kpa (40 psi)

pressure applied on both ends. Finally the pellet was extruded from the mold and cured

under ambient conditions. During the curing, evaporation released 10 g of the 12 initial

grams of the water. At least three anthracite pellets were produced from the same batch

for replications. Only the binding material (binder, additives, and water) was preheated.

The anthracite fines were added at room temperature.

3.2.3 Mechanical strength tests

Mechanical strengths of bindered anthracite fines at room temperature were

monitored (Richards 1990), by the unconfined compressive strength test and the drop

shatter test. The drop shatter test is a standard for foundry coke to test its strength against

breakage (anti-breakage strength). In the drop shatter test in this study the anthracite

pellets were dropped onto a steel plate through a 1.83 meter long PVC tube that had an

inner diameter of 3.81 cm. First one flat end was down, next the other, until a single

pellet was dropped for a total of 10 times. Then the weight of the major remaining piece

57

was obtained. Also if no a single piece larger than 50% of the original weight of the pellet

remained after a drop, the test was stopped and 0% was recorded for the test.

The unconfined compressive strength of anthracite pellets was determined by a

Simpson-Gerosa electronic universal sand strength machine. A horizontally moving arm

applied pressure on a pellet until failure. Final compressive force was calculated based on

the diameter of the original pellet sample, assuming negligible shape change during the

test.

3.2.4 Pyrolysis

The pyrolysis of the anthracite pellets was conducted in a Lindberg Blue M tube

furnace from Thermo Scientific (Newington, NH). A horizontal alumina tube was

employed. A slow nitrogen gas flow (~ 2 standard cubic centimeters per minute) was

used to prevent the anthracite from burning. A three-step pyrolysis procedure was

employed. First the furnace was ramped up to the prescribed temperature at 3 °C/min.

Then this prescribed temperature was maintained for 2 hours. Finally the furnace was

cooled down to room temperature at 3 °C/min. The pyrolyzed anthracite pellets were then

removed from the tube-furnace for further tests. Unless otherwise identified, the

prescribed maximum temperature was 1400 °C, which was close to the temperature that

the anthracite bricks would encounter in the preheating zone of a cupola furnace.

3.2.5 Characterizations

Scanning electron microscopy (SEM) analysis was performed on an FEI Quanta

200 Environmental SEM. The instrument was operated under low-vacuum conditions

(10-103 Pa) using a Gaseous SE detector. The high voltage was set at 20 kV, and the spot

58

size of the electron beam was set at 4 nm. A transmission electron microscope (TEM,

Model 2010, JEOL, Tokyo, Japan) was used to determine the morphology and particle

size, for electron diffraction.

Ambient temperature XRD patterns were obtained via a PANalytical X’Pert Pro

MPD diffractometer. The diffraction patterns were collected for two theta between 5-70

degrees. The pellets were crushed into powders by the ball mill, and the powders were

placed into a special aluminum sample holder. So the powder XRD pattern offered a bulk

measurement, and represented the average situation in the pyrolyzed pellets.

3.3 Results and discussion

3.3.1 The effect of pellet composition on mechanical strength

Anthracite pellets were prepared with an array of recipes. These included 100 g

anthracite fines, 1 g collagen, 0-1 g fructose, and 0-10 g of silicon, kaolinite, or sodium

silicate. The unconfined compressive strength and drop shatter retentions have been

presented in Table 3-3. The most favorable unconfined compressive strengths were

achieved when employing 10 g silicon metal powder (1300 kPa before pyrolysis and 720

kPa after pyrolysis) (Table 3-3). The 10g sodium silicate addition offered considerable

unconfined compressive strength (2344 kPa) before pyrolysis, but only 28 kPa after

pyrolysis. The pellets that include kaolinite offered scant strength after pyrolysis. For all

three Si sources, less Si corresponded to lower strengths.

The differences in morphology and crystal structure confirmed that each of the

three Si-containing materials tested in this study exhibited distinct binding mechanisms,

relative to their ability to hold the anthracite fines together at high temperature. Moreover,

59

the redox level of the silicon in those materials significantly affected the development of

the binding systems.

The crystallized Si (+4) in the kaolinite powder could not react with the anthracite

to form silicon carbide under the conditions tested in this study. Instead, the kaolinite

powders became sintered into larger ceramic features,via mechanisms that are similar to

the process of making porcelain from kaolinite. Also, numerous spheres formed on the

anthracite surface (Figure 3-3b) indicate a weak affiliation between the kaolinite

alteration products and the anthracite particles. Therefore the strength of the pyrolzyed

anthracite pellets was very weak too.

The amorphous Si (+4) in the sodium silicate solution was able to form silicon

carbide with the anthracite at 1400 °C. However, the amount of SiC formed was smaller

(much smaller diffraction peaks in the XRD pattern), as compare to the silicon powder

system. The silicon was already in the +4 valence, it is very difficult for the silicon to

enter the vapor phase and grow SiC nanowires via the classic vapor-liquid-solid or vapor-

solid growth mechanism. This situation is similar to the commercial SiC production, in

which silica has been reduced by carbon to SiC at 1600 to 2500 °C. This procedure

usually produces SiC powders and bulk materials, but does not produce SiC nanowires.

Furthermore, there is no evidence that the silicon carbide generated from the amorphous

Si (+4) provided binding strength for the pyrolyzed anthracite pellets. The ceramic

bridges that held the anthracite particles together were most likely the remnants of the

solid sodium silicate.

The zero valent silicon powders reacted with the anthracite at 1400 °C and grew

silicon carbide nanowires from the anthracite surfaces. Silicon carbide itself is a material

60

with very high mechanical strength that is resistant at high temperatures. Even better is

that the silicon carbide nanowires were grown from the anthracite surface and attached to

the anthracite firmly through the formation of silicon carbide between carbon and silicon.

From the basic thermodynamics of the chemical reactions, it is noticed that the

reaction from the element silicon to silicon carbide is energetically favorable, as the

overall reaction has a negative enthalpy change as shown in equation [1]. The overall

reaction from Si (+4), for example quartz, to silicon carbide has a positive enthalpy

change as shown in equation [2] (other SiO2 forms are only slightly different in the

formation enthalpy value). Thus, energy input is required to form silicon carbide from Si

(+4) and carbon. This endothermic reaction will decrease the temperature in the cupola

preheating zone. Alternatively, exothermic reactions such as reaction [1], are favorable in

the cupola furnace operation.

molkJ=ΔHmolkJ=ΔGSiC(s)C(s)+Si(s) θθ /11.73/139.69 −−→ [1]

molkJ=ΔHmolkJ=ΔG(g)CO+SiC(s)(s)+(s)SiO θθ22 /241.444/261.690 2C → [2]

molkJH SiCf /11.73, −=θ

molkJ=GθSiCf, /139.69−

molkJH SiOf /86.9102, −=θ

molkJG SiOf /66.8532, −=θ

molkJ=θG2COf, /26.94−

molkJ=θH2COf, /509.393−

61

But in order to derive the metallic Si that can be loaded into the cupola, a manufacturer

would need to expend energy to form the Si metal from SiO2 or other sources. So from an

energy perspective, there are tread-offs. As a first step, none-the-less, foundries have the

opportunity to use the silicon metal that is added to the melt anyway, and bind it with the

anthracite fines in a manner that offers this alternative carbon energy source.

3.3.2 Crystal structure changes within anthracite pellets with difference Si-

containing materials

XRD patterns in Figure 3-4 show the crystal structure change within the

anthracite pellets induced by the pyrolysis. The raw anthracite fines tested in this study

contain muscovite (KAl2(AlSi3O10)(F,OH)2), kaolinite (Al2Si2O5(OH)4 ), and quartz

(SiO2) as detected by the XRD (Figure 3-4). These materials also contain crystallized Si

(+4). After the pyrolysis at 1400 °C, muscovite and kaolinite were converted into mullite

(Al6Si2O13), and the carbon became more crystallized. Pyrolysis of the anthracite fines

alone yields no silicon carbide, as detected by the XRD.

When 10 g kaolinite powder was bindered to 100 g anthracite fines with 1 g

collagen, the XRD exhibited that the kaolinite diffraction peaks that appeared before

pyrolysis were converted mostly to mullite after 1400 °C pyrolysis. This combination did

not yield silicon carbide.

In contrast, when they collagen bindered anthracite fines included sodium silicate

as the Si source, 1400 °C pyrolysis did yield beta silicon carbide (3C-SiC), per XRD

(Figure 3-4). This means that this pyrolysis condition incurred a net-reduction of the

silica from Si (+4) to Si (0) valence. The carbon in anthracite reacted with the Si (+4)

62

within the sodium silicate, and the amorphous structure of the sodium silicate apparently

enabled this redox reaction.

The formation of SiC became manifest yet greater when 10 g silicon metal

powder was bindered with 100 g anthracite fines and pyrolyzed at 1400 C. After being

pyrolyzed at 1400 °C for 2 hours, the silicon metal XRD response was depleted, while

very strong diffraction peaks from the cubic crystal structure of 3C-SiC were observed.

Under the pyrolysis condition, zero valent silicon and carbon reacted with each other to

form silicon carbide. The real reaction pathway may include several intermediate steps as

suggested in the literature (Chiu et al. 2007; Park et al. 2004; Yang et al. 2009), but the

final product is silicon carbide.

3.3.3 After pyrolysis: the morphologies of anthracite pellets that contained various

Si-containing materials

SEM images in Figure 3-3 show the morphology of the anthracite pellets with

various Si-containing additives after the pyrolysis at 1400 °C. For the kaolinite system

after the pyrolysis (Figure 3-3a) the SEM images showed that kaolinite powders were

sintered into yet larger ceramic features that spanned across neighboring anthracite

particles. A close look on the surface of anthracite pellets (Figure 3-3c) revealed many

spheres (presumably silica as indicated by the EDS spectrum in Figure 3-5a) generated

from the alteration of kaolinite to mullite and silica; and these shapes indicated that the

spheres were experiencing a high surface tension.

When sodium silicate was included, the SEM images of the pyrolyzed products

exhibited sponge-like ceramic structures that were reminiscent of their sodium silicate

63

precursor structures (Figure 3-3c&d). The ceramic blocks were not composed by silicon

carbide, as indicated by the EDS spectrum in Figure 3-5b.

In contrast, when 10 g silicon was included with 100 g anthracite fines, the 1400

°C pyrolysis created a labyrinth of nanowires. These nanowires grew to 20 μm generally;

and some times to 100 μm. The nanowires altered the color of the pellets from the black

of anthracite to the light green of the nanowires. These nanowires were often 30 nm in

diameter, and they were coated by an amorphous thin layer of silicon oxide (about 2 nm

thick) as shown in the HRTEM image in Figure 3-6. Figure 3-6 also indicates the

nanowires were highly crystallized (except for this thin amorphous exterior layer). As per

the XRD patterns (above) the silicon carbide was 3C-SiC, which has a face centered

cubic structure. The space between two planes in Figure 3-6 was about 0.25 nm which is

concurrent with the calculated value in 3C-SiC crystal structure (2.51A). The stacking

pattern in Figure 3-6 confirmed that the nanowires were grown by stacking the (111)

lattice plane of 3C-SiC in the [111] direction (Attolini et al. 2008; Attolini et al. 2009;

Baek et al. 2006; Chen et al. 2008; Feng et al. 2003; Hyun et al. 2009; Ju et al. 2007;

Kang et al. 2004; 2005; Khongwong et al. 2009; Lai et al. 2000; Li and Wen 2007; Li et

al. 2008; Li et al. 2001a; Li et al. 2001b; Li et al. 2009; Li et al. 2003; Liang et al. 2000;

Meng et al. 1999a; Meng et al. 1999b; Meng et al. 1999c; Meng et al. 1998; Meng et al.

1999d; Miranda et al. 2009; Shen et al. 2006; Wen et al. 2008; Yang et al. 2009; Yang et

al. 2006; Yao et al. 2007; Zhang et al. 2009a; Zhang et al. 2006; Zhang et al. 2008; Zhang

et al. 2007; Zhou et al. 2006; Zhou et al. 2000a; Zhou et al. 2000b; Zhou et al. 1999; Zhu

et al. 2005).

64

3.3.4 Improving the silicon carbide nanowire binding strength by reducing the

anthracite grain size

The anthracite pellets could be packed more tightly by decreasing the grain size.

In this way the gaps between the anthracite particles will be small enough for the SiC

nanowires to directly connect two neighboring anthracite particles. Figure 3-7 shows the

post-pyrolysis strength of anthracite pellets made from different contents of powdered

anthracite and silicon. With the same silicon content, anthracite pellets made totally from

the powdered anthracite were 5 times stronger than the anthracite pellets made totally

from the original anthracite fines. With partial powdered anthracite, the post-pyrolysis

strength of the anthracite pellets could also be enhanced. As shown in Figure 3-7, the

anthracite pellets made from 50% powdered anthracite and 50% original anthracite fines

were only slightly weaker than that made from 100% powdered anthracite.

SEM images in Figure 3-8 exhibited a gradual change within the pyrolyzed

anthracite pellets as the percentage of powdered anthracite increases. When the

anthracite pellets were made from the original anthracite fines only (Figure 3-12a), there

were many relatively big void spaces among the original anthracite particles. The

anthracite pellet in Figure 3-8b was made from 75% original anthracite fines and 25%

powdered anthracite. In this pellet some smaller anthracite particles filled in the void

spaces between the bigger anthracite particles. This filling reduced the gaps between two

anthracite particles which made it easier for the nanowires to directly connect them.

Figure 3-8c shows a pyrolyzed pellet made from 50% original anthracite and 50%

powdered anthracite. In this pellet the big anthracite particles were surrounded by the

smaller anthracite particles. It is clear that the average gap size in Figure 3-8c is much

65

smaller than that in Figure 3-8b, and similar to that in Figure 3-8d which was made solely

from powdered anthracite. Thus, anthracite powders can fill in the gaps between the

bigger original anthracite fines, and it is an efficient way to reduce the gap size within the

anthracite pellets.

3.3.5 Coke replacement's opportunity on sustainability and its limitations

The engineering approach for partial coke replacement presented in this study

rearranged the materials that flow into the cupola furnace as shown in Figure 3-1. By first

binding the anthracite with the silicon metal powder, and then add them together into the

cupola furnace, a portion of the foundry coke could be replaced by the cheaper anthracite

fines and the silicon metal that is added into cupolas anyway.

Also replacing expensive foundry cokes by waste anthracite fines also can

increase the sustainability of the iron casting process by reducing the energy consumption

and the carbon dioxide emission from the coke-making process. As discussed in the

introduction section, coke is produced by pyrolyzing special bituminous coals at above

900 C for 24 hours. The pyrolysis step is necessary to fuse the carbon and drive out the

volatile species in the bituminous coals. Anthracite has already obtained these properties

through millions of years’ natural carbonization process and does not need the pyrolysis

process to meet the requirement of an efficiently compacted carbon source for the cupola

furnace operation.

In a well managed coke plant about 5054 to 5940 mJ of energy is consumed in-

situ to produce one metric ton of coke (Energetics-Incorporated 1999; Ertem and

Ozdabak 2005; Stubbles 2000). The energy for coke production is obtained by burning

coals, and then the heat releasing reaction is:

66

molkJ=ΔH(g)CO(g)O+C(s) θf2 /393.52 −→

Using the following calculation:

tonkgmolkg

molmJ

tonmJ

/664044.03935.0

5940=×

Therefore for a metric ton of coke produced, 664 kg of CO2 would be released into the

atmosphere by burning carbon. In comparison to this, a tone of coke offers 30.74 billion

joules of energy, and a metric ton of coke release 3667 kg of carbon dioxide when burned.

Thus, 20% of all energy and CO2 represented in a ton of product coke is consumed in

merely making the coke. It is this energy that can be circumvented by replacing a part of

the coke with binderd anthracite fines.

Coking processes also generates other pollution (World-Bank 1998), such as

particulate matter, SO2, NOx, volatile organic compounds, and wastewater, that would not

be generated by using anthracite fines. Furthermore, anthracite fines are often a waste by-

product from anthracite mining, and the approach herein would facilitate their use as a

valuable fuel source for the iron casting and steel industries; and this provides additional

revenue for the anthracite mining industry. This research exhibited a practical example on

improving the sustainability by adapting sustainable engineering solutions in the

traditional industrial manufacturing.

However, there are some limitations of utilizing the silicon bindered anthracite

pellets. The major limitation comes from the requirement of elemental silicon. Although

silicon is also demanded in the cupola furnace for iron casting, the amount of silicon is

relatively small. In normal operations, the amount of silicon is about 1% of the fuel on

67

weight basis. As from the result in this study, the silicon content can be reduced to as low

as 4% in the anthracite pellets and the pellets still can have more than 690 kPa (100 psi)

of post-pyrolysis unconfined compressive strength. In this way, the anthracite fines can

replace up to 25% of the coke without introducing additional elemental silicon.

Additional silicon usage will introduce additional cost and decrease the environmental

impact of coke replacement, because elemental silicon is also produced through energy

intensive procedures.

3.3.6 Burning rates of bindered anthracite pellets

The burning rate of the bindered anthracite fines with silicon were tested and

compared with the foundry coke. Pieces (about 5 gram in mass) from the anthracite

pellets were dropped into a tube furnace which was preheated to 1050 °C. After the piece

was heated in the tube for about 2 minutes, air was passed through the tube to provide

oxygen for burning. The final time was recorded when the piece was burned out, and no

more fleshing could be observed.

As shown in Table 3-4, the non-bindered anthracite fines had a much faster

burning rate than the foundry coke. When silicon was added to hold the anthracite

together at high temperature, the anthracite pellets made totally from anthracite powder

burned slightly faster than did the foundry coke; while the anthracite pellets made from

25% anthracite powders and 75% anthracite fines burned slightly slower than did the

foundry coke. One phenomenon was noticed was that anthracite pellet pieces burned

faster than the coke at the beginning of the test. Then the silicon carbide held the ash of

the burned outer anthracite particles on the piece which slowed the burning of inner

anthracite particles. The air flow inside the tube can peel of the ash from the foundry

68

coke, but could not peel down the ash generated from silicon added anthracite pellets.

However, inside the cupola furnace, the air flow is much more turbulent, and the molten

iron drips should be able to peel down the surface ash easily.

3.3.7 Full-scale test of briquetted anthracite fine bricks in an operating foundry

cupola

Briquetted anthracite fine bricks made from anthracite fines, collagen binder,

fructose additive, and silicon powders were fed into an operation foundry cupola to

replace a part of the foundry cokes. All together, 500 pounds of the anthracite bricks had

been prepared. The bricks included anthracite fines (70 parts), anthracite powder (30

parts), silicon (8 parts), collagen (1 part), fructose (1 part), some moisture, and other

additives. Each brick was 5.75inches (14.6 cm) diameter by 2.25 inches (5.7 cm) high,

and weighed 1.8 pounds (0.82 kg).

For about one hour, these bricks were fed into the cupola at a rate that they

replaced 10% of the coke that was otherwise fed into the cupola. Together with other

materials, the anthracite bricks were roughly dropped 10 feet into the hopper and bucket,

along with the scrap metal, coke, silicon bricks, limestone, etc. During the time of the

test, the feeding rate of the silicon bricks was diminished by about 5-8%. No breakage of

the anthracite bricks was observed as they dropped into the bucket (as per visual

inspection), nor as they were conveyed into the top of the cupola (as per remote camera).

While these anthracite bricks descended through the cupola melt zone past the tuyere

windows, 10-12 shapes that exhibited the tell-tale smooth curvature of a 5-6” circle were

observed; and it appeared that a number of these were the anthracite bricks. In some

cases, just an edge of the bricks could be observed (as the windows were somewhat

69

narrow), and in a couple cases, the full round brick could be observed. Through the

tuyeres, none of the bricks that we observed had been broken in pieces. In the melt zone,

these bricks appeared to be burning throughout their structure, just as were the coke

chunks. Moreover, the bricks were resisting abrasive degradation as molten iron

splattered onto the bricks, just as for coke.

The parameters appraised during this demonstration included visual inspection of

the bricks’ structural integrity (as above), product metal quality and chemical

composition, cupola temperatures and pressures, etc. The metal product quality and

chemistry parameters were within the experimental error of one another. The back

pressure and the temperature (represented by the upper-stacker temperature) of the cupola

were increased during the addition of the anthracite bricks as shown in Figure 3-9.

Cupola operation is complicated and affected by many factors. It is hard to draw a

conclusion that the changes in temperature and back pressure were caused solely by the

addition of the anthracite bricks. However these increases were within the operation

range of the cupola furnace.

In summary, by replacing 10% of the foundry cokes, the cupola furnace could still

operate in normal condition. There was no significant change in the operation parameters

and there is no damage to the cast iron product either. For the small scale foundry where

the full-scale test was carried out, 600 ton of coke could be replaced per year at 10%

replacement rate. It could save the foundry at least $120,000 a year in fuel cost. Also, per

the calculation in section 3.3.6, more than 360 tones of carbon dioxide emission could be

reduced per year from the coking process. For the entire U.S. foundry industry, which

70

consumes about 1.4 million tonnes of coke each year, the saving in fuel cost and the

reduction in carbon dioxide emission could be even more significant.

3.3.8 Effect of preheating and organic crosslinking on the binding strength of the

collagen binder

Table 3-5 compares the unconfined compressive strength and drop shatter test

results of pellets with 1% collagen-based binder. Only the binding material (binder,

additives, and water) was preheated. The anthracite fines were added at room temperature.

It is clear that when mixed with water at room temperature the collagen-based

binder didn’t have strong binding strength. The anthracite pellets made at 25 °C had an

unconfined compressive strength of only 138 kPa and was completely destroyed in the

drop shatter test. The mechanical strength of anthracite pellets kept increasing with

increasing temperature. The biggest difference was between 25 °C and 50 °C. The anti-

breakage strength of the pellets increased from 0 to 67% and the compressive strength

increased about 500%. In order to pursue the balance between better binder performance

and easy-control procedure, most of the pellets tested in following experiments were

made with collagen-based binder and additives solution preheated at 70 °C. The addition

of 1% fructose in the binder mixing process had little effect on the compressive strength.

However, it dramatically increased the anti-breakage strength of the anthracite pellets.

SEM photos (Figure 3-10) revealed the effect of preheating on the collagen binder.

If the binder and water were mixed at room temperature, there was no obvious sign

indicating the existence of binders. The binders bind the anthracite particles only at the

places where the particles contacting each other. Preheated binders covered the surface of

many anthracite particles and formed strings that stretched to link with the binding

71

materials attached the neighboring anthracite particles. At room temperature the collagen

binder adsorbs water and becomes sticky but does not dissolve in water. As temperature

increases, the mixture of water and collagen binders becomes a viscose solution. When it

is cooled again, the solution becomes a whole piece of gel. This sol-gel process improved

the performance of the binders. Because the binders were in the same phase as the water,

they could be much easily and evenly distributed through the anthracite pellet. Thus the

sol-gel process provided much higher integrity for the collagen binders in the anthracite

pellets. Also as shown in Figure 3-10, when 1% fructose was added into the binder

mixture, more binder strings were formed on the surface of the anthracite particles. It

confirmed that adding monosaccharide at least increased the amount of organic materials.

This effect is more noticeable when very low amount of collagen binder (for example

0.2%) was used, as shown in Figure 3-11.

With only 0.2% collagen the existence of 1 gram monosaccharide dramatically

increased the compressive strength of the anthracite pellets. While with 0.5% and 1%

binder, adding up to 1 gram monosaccharide didn’t increase the compressive strength of

the anthracite pellets. In contrast with increasing sugar amount the compressive strength

of the anthracite pellets decreased. When 5 gram monosaccharide was added, the pellets

became very soft and could not support heavy loads. Without any sugar additives the

anthracite pellets with 0.5% or 1% collagen binders had lower anti-breakage strength

than the foundry coke did. With increasing monosaccharide amount, the anti-breakage

strength of the pellets increased. With monosaccharide the anthracite pellets also

increased in elasticity. That means within certain deformation range, the pellets with

monosaccharide can return to their original shape when the force is removed. However,

72

as mentioned above, too much sugar additives greatly decreased the compressive strength

of the pellets. Results in Figure 3-11 confirmed that with 1% collagen and 1% fructose

the anthracite pellets had better anti-breakage strength than the foundry coke tested in this

study, and they also had acceptable compressive strength.

Intensive drop shatter tests were performed on anthracite pellets made from

different recipes as shown in Figure 3-12. It can be seen from Figure 3-12 that the

anthracite pellets made with 1% collagen binder and 1% fructose always had higher anti-

breakage than the foundry coke did within the 30 drops of the intensive drop shatter tests.

The pellets with 2% collagen binders had similar anti-breakage strength as the pellets

with 1% collagen and 1% fructose did after 10 drops. However, the results from 20 and

30 drops showed that the 1% additional collagen binder did not increase the anti-

breakage strength as did the 1% fructose. The pellets with 2% collagen binders showed

much less strength than the pellets with 1% collagen and 1% fructose did after 20 and 30

drops, and they were even weaker than the foundry cokes. But still collagen binder is also

an important component of the whole binding system. As shown in Figure 3-12, although

some pellets with 0.5% collagen-based binder had better anti-breakage strength than the

coke in the normal drop shatter test, they were weaker in the intensive drop test.

It is known that monosaccharide can crosslink with proteins such as collagen

through glycation reaction, and it is initiated between the aldehyde group of the sugar and

an additional amino group on the peptide chain (Bailey et al. 1993; Eyre and Wu 2005;

Fujimoto and Horiuchi 1986; Sajithlal et al. 1998a; Tanzer 1973). Figure 3-13 shows the

reaction between glucose and a peptide chain with lysine. Because free water was

generated during this reaction, it can be verified by monitoring the weight change of the

73

system in a dried environment. Samples contained 5 grams collagen-based binder and 5

grams sugar were mixed in water solution. And then the mixture was dried at 70 °C for

more than one week. Samples with only sugar or only collagen-based binder were also

treated through the same procedure as blanks. As water generated during the crosslink

reaction evaporated from the system, the weight of the system decreased as shown in

Table 3-6. Strength of collagen materials can be enhanced by crosslinking with sugar

(Charulatha and Rajaram 2003; Ohan et al. 2002; Usha and Ramasami 2000; Verzijl et al.

2002). Thus crosslinks from monosaccharide increased the strength of the collagen binder.

Adding appropriate amount of fructose will increase the anti-breakage and anti-abrasive

strength and these properties are important for the anthracite pellets both in the

transportation process and in the drop-zone of the cupola furnace.

74

Table 3-1 Sieve analysis of the original anthracite fines

Sieve number (opening size µm)

10 (2000)

20 (841)

40 (400)

80 (177)

100 (149)

Pan

Retained (%) 0.26 33.87 39.26 23.81 1.42 1.38

75

Table 3-2 Compare the elemental analysis and heat content of the anthracite tested in this study to a typical foundry coke

Proximate analysis % Ultimate analysis % (d.a.f.) Name V.M. (dry)

F.C. (dry)

Ash (dry)

C H N S O Dry heat content (mJ/kg)

Anthracite 4.99 82.3 12.8 94.3 2.25 0.89 0.38 2.59 29.88 Coke 1.30 90.81 7.89 96.1 0.75 1.61 0.68 0.95 30.74

V.M.-volatile matter; F.C.-fixed carbon

76

Table 3-3 The strengths of anthracite pellets bindered with an array of binders and additives recipes

Binder contents (100 gram anthracite fines size #10-80, l gram collagen binder, other additives as listed)

Unconfined compressive strength before pyrolysis (kPa)

Drop shatter remain (%)

Unconfined compressive strength after pyrolysis (kPa)

No other additives 1234 76 0 1g fructose 1095 99 0 3g sodium silicate (0.7g as silicon) N/D 90 0 6g sodium silicate (1.4 g as silicon) 2455 99 0 10g sodium silicate (2.3 g as silicon) 2342 98 ~28 1g kaolinite (0.22 g as silicon) N/D 91 0 10g kaolinite (2.2 g as silicon) 1340 93 ~7 2g silicon 1230 77 69 5g silicon N/D 80 345 10g silicon 1250 91 690 1g fructose, 10g silicon 1300 99 720 Foundry Coke1 2758 93 2758 10g silicon2 N/D N/D 3600

N/D- not determined; 1-foundry coke itself; 2-using 100 gram anthracite powders (<#100) instead of anthracite fines

77

Table 3-4 Comparison on the burning rates of different anthracite pellets and a foundry coke in a tube furnace

Sample Set temp °C

Accrual temp

with air °C

Preheat time (min)

Final time (min)

Burning time (min)

Coke piece 1050 850-900 2 20.75 18.75 Coke piece 1050 850-900 2 21.5 19.5

Anthracite pellet piece (100 powdered

anthracite, 4 silicon, 1 collagen)

1050 850-900 2.5 21 18.5

Anthracite pellet piece (100g powdered

anthracite, 4g silicon, 1 g collagen)

1050 850-900 2.5 20 17.5

Anthracite pellet piece 5 g

(25g powder anthracite, 75g anthracite fines, 10g silicon, 1g

collagen)

1050 850-900 2.75 24 22.25

Anthracite fines (#10x80)

1050 850-900 2.75 8 5.25

Anthracite fines (#10x80)

1050 850-900 2.5 6 4.5

Anthracite pellet piece (100 g

anthracite fines, 1g collagen)

1050 850-900 2.5 - - * the piece fall into

single fines after the preheat

78

Table 3-5 The effect of preheating on the binding strength of the collagen based binder 1% Collagen 1% Collagen+1% Fructose Temperature Unconfined compressive strength (kPa)

Drop shatter retained (%)

Unconfined compressive strength (kPa)

Drop shatter retained (%)

90 °C 1323 85 1364 96 70 °C 1233 76 1096 96 50 °C 792 67 916 96 25 °C 139 0 124 0

79

Table 3-6 Mass loss from the evaporation of the water generated from the crosslinking between collagen and fructose

O.M. (g) F.M. (g) M. L. (g) H2O(g)a C 5 4.635 0.37 - F 5 5 0 - F+C 5+5 8.275 1.725 1.355 a: water generated from the crosslink O.M.-original mass, F.M.-final mass after dried at 70 °C, M.L.-mass loss, C-collagen binder, G-glucose, F-fructose

80

Figure 3-1 The operation schematic for a cupola furnace that includes bindered anthracite bricks

81

Figure 3-2 The setup of the drop-shatter tester used in this study

82

Figure 3-3 SEM images of anthracite pellets with different Si-containing materials after the pyrolysis at 1400 °C. Pellets with 9% kaolinite powders (a&b); Pellets with 9% solid

sodium silicate(c&d); Pellets with 9% silicon powders (e&f)

83

Figure 3-4 Crystal structure changes of anthracite pellets with different ceramic additives

(9%) after the pyrolysis at 1400 °C

84

Figure 3-5 EDS spectrum of (a) a sphere on the surface of an anthracite particle in

pyrolyzed anthracite pellets that contained kaolinite powders; (b) a ceramic bridge in a pyrolyzed anthracite pellet that contained sodium silicate

85

Figure 3-6 High resolution TEM of the silicon carbide nanowires formed at 1400 °C

within the anthracite pellets with 9% silicon powder

86

Figure 3-7 Different amount of powdered anthracite and silicon: effect on the unconfined compressive strength after pyrolysis at 1400 °C

87

Figure 3-8 The gradual change in morphology within the pyrolyzed anthracite pellets with 9% silicon content as the amount of powdered anthracite increased (a-made from 100% raw anthracite fines; b-made from 75% raw anthracite fines and 25% powdered anthracite; c- made from 50% raw anthracite fines and 50% powdered anthracite; d-

made from 100% powdered anthracite

88

15

20

25

30

35

06:00:00AM

07:12:00AM

08:24:00AM

09:36:00AM

10:48:00AM

12:00:00PM

Time

Cup

ola

back

pre

ssur

e (o

si)

800

1100

1400

1700

2000

Upp

er s

tack

tem

pera

ture

(F)

Cupola Back Pressure

Upper StackTemperature

Anthracitefeedingperiod

Figure 3-9 Changes of upper stack temperature and cupola back pressure during coke

replacement by anthracite fine bricks

89

Figure 3-10 SEM images of anthracite particles bound by 1% collagen-based binders at

room temperature (a), at 70 °C (b), and 1% collagen+1% Fructose at 70 °C (c, d)

90

0

200

400

600

800

1000

1200

1400

-1 0 1 2 3 4 5 6Fructose amount (g per 100g anthracite)

Unc

onfin

ed c

ompr

essi

ve s

tren

gth

(Kpa

)

Fructose+0.2% collagen

Fructose+0.5% collagen

Fructose+1% collagen

0

10

20

30

40

50

60

70

80

90

100

-1 0 1 2 3 4 5 6Sugar amount (g per 100g anthracite)

Ant

i-bre

akag

e st

reng

th

(wei

ght r

etai

ned

afte

r dro

p sh

atte

r %)

Fructose+1% collagen

Fructose+0.5%collagen

Foundry Coke

Figure 3-11 The effect of adding fructose on the unconfined compressive strength and anti-breakage strength of the anthracite pellets at ambient conditions

91

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15 20 25 30 35Drop Times

Wei

ght R

etai

ned

% 1C+2.5F0.5C+2.5F1C+2F1C+1.5F1C+1FCoke2C1C+0.5F0.5C+2F0.5C+1.5F0.5C+1F0.5C+0.5F

Figure 3-12 Intensive drop shatter test of pellets with different binder and sugar contents

92

C

NH

H (CH2)4

C

NH2

O

Amino Acid

Amino Acid

C OH

(CHOH)4

CH2OH

C

NH

H (CH2)4

C

NH

O

Amino Acid

Amino Acid

C OCH2

(CHOH)3

CH2OH

H2O+ +

C

NH

H (CH2)4

C

NH

O

Amino Acid

Amino Acid

C OCH2

(CHOH)3

CH2OH

C

NH

H (CH2)4

C

NH

O

Amino Acid

Amino Acid

C

O

CH2

HCOH

HCOH

OH

HCOH

HCHRearrangement

Figure 3-13 The crosslinking between a peptide chain and a monosaccharide molecule

93

Chapter 4 Formation of silicon carbide nanowires on the surface of anthracite fines

Abstract

Silicon carbide nanowires (SCNWs) were grown from anthracite surfaces through

a simple one-step carbothermal process with silicon powder as the Si precursor.

Progressive thermal tests exhibited that the formation of the SCNWs started from 1100

°C; and was favored at 1400 °C. No extra metal catalyst was needed for the growths of

the SCNWs. Characterizations were performed by XRD, SEM, EDS, TEM, and SAED.

The SCNWs were 30-60 nm in diameter and were typically grown by stacking the (111)

lattice plan of 3C-SiC along the [111] direction. Many non-epitaxial branches of the

nanowires were also formed through this one-step process as observed by TEM. The

results suggest the SCNWs were most likely grown through the Vapor-Solid mechanism.

This inexpensive and fast formation of SCNWs made it possible to bind the waste

anthracite fines as alternative fuel in foundry cupola furnaces, as the SCNWs provided

the crucial hot crushing strength.

4.1 Introduction

Silicon carbide is a ceramic material with extraordinary mechanical properties,

and it has been employed to reinforce other ceramic materials (Takahashi et al. 2003;

Zhang et al. 2009b). It is well know that silicon carbide nanowires (SCNWs) could be

grown from many different substrates (Deng et al. 2006; Yang et al. 2004; Yang et al.

94

2005; Zhu et al. 2005), including carbon-based materials (Gao et al. 2002; Munoz et al.

2002). This in-situ growth of silicon carbide nanowires provides an opportunity to bind

fine carbon particles into bigger structures. Unlike the structures held by organic binders,

these silicon carbide nanowires bindered structures will retain their strengths under high

temperature conditions. One emerging application is to bind waste anthracite fines into

bigger bricks, if SCNWs could be grown in-situ from anthracite surfaces through a

relatively simple protocol. In this way, the waste anthracite fines could be utilized as a

high quality fuel in the iron casting process in which silicon carbide is also a desirable

additive.

SCNWs have received much attention partially due to their electronic properties

and their high mechanical strength even at extremely high temperatures (Wong et al.

1997; Zhang et al. 2008; Zhou et al. 1999). Although various precursors and processes

have been employed (Fu et al. 2009; Khongwong et al. 2009; Li et al. 2010b; Shi et al.

2000; Zhang et al. 2009a), most of them are chemical vapor depositions via either Vapor-

Liquid-Solid (VLS) (McMahon et al. 1991; Wagner and Ellis 1964) or Vapor-Solid (VS)

mechanism(Yang et al. 2009). Usually these processes have required high temperature, a

continuous supply of precursors, and metallic catalysts (for VLS). Typically, the

precursors that contain silicon and carbon have been introduced continuously during the

process, to provide raw materials for the growth of the nanowires. The metallic catalyst

has been essential in forming the liquid phase in VLS growths that has led to the growth

of the nanowires, but it is not necessary for VS growths.

In contrast to the above complexity, a simplified process has been devised for

growing SCNWs on the surface of anthracite fines. Silicon powders were bindered with

95

anthracite fines and exposed to simulated thermal conditions. Yet further, the SCNWs

generated herein were then tested so as to discern their mechanical strength for binding

the anthracite fines together at extremely high temperatures. For full-scale foundry fuel

applications there are several interesting points associated with the environment that the

SCNWs were proposed to be applied in. First, high temperature conditions exist in the

cupola furnace;(Huang et al. 2009) and the typical temperature above the iron melting

zone is from 1000 °C to 1400 °C. So the first question was whether the SCNWs could

grow well within this temperature range. Secondly, the anthracite fines contain

considerable traces of metals; and the question was whether these could effectively

catalyze the growth of the SCNWs. The third question was whether there would be

enough raw materials for the SCNWs to grow without a continuous inflow of precursors.

In this study, analytical techniques such as scanning electron microscopy (SEM),

energy dispersive X-ray spectroscopy (EDS), transmission electron microscopy (TEM),

and X-ray diffraction (XRD) were employed to monitor the morphology and

crystallography changes in the silicon and anthracite system. The effects of anthracite

grain size, pyrolysis temperature, and silicon content were evaluated.

4.2 Experimental Section

4.2.1 Raw materials

The anthracite fines used in this study was obtained from Jeddo Coal Company

(Hazelton, Pennsylvania). The elemental analysis of this coal source has been listed in

Table 1, and the EDS spectrum on the ash of this coal is shown in Figure 4-1. Most of

these anthracite grains (about 97%) were within the range of U.S. mesh #10 × 80 (2000

96

μm to 179 μm). This grain size has been indentified as “anthracite fines” throughout the

text herein. For some experiments, these anthracite fines were also crushed into powders

that passed through a U.S. mesh #100 (150 μm) sieve. This grain size has been identified

as “anthracite powder” throughout the text herein.

The silicon metal was purchased from Alfa Aesar (Ward Mill, MA) as 98.4%

silicon lumps. The lumps were crushed into silicon powders by a ball mill; and then

screened to less than U.S. mesh #100 (150 μm) sieve. The typical grain size of this

silicon powder was 10 to 20 μm as observed by SEM.

4.2.2 Preparation of bindered anthracite pellets

Anthracite fines were dried at 105 °C overnight to remove the moisture content.

Then 100 grams of anthracite fines were mixed with 10 grams of silicon powder (i.e. 9%

Si) unless otherwise stated. In some cases, 2 or 5 grams of silicon was used. Meanwhile,

the collagen binder (1 gram by dry weight) was dissolved in water at 70 °C to form a

gelatin sol. The anthracite and silicon mixture were added into this sol with constant

mixing. The final mixture was packed into a cylindrical mold (2.9 cm dia. x 4.8 cm long)

with about 280 kPa pressure applied on both ends. Finally, the pellet was extruded from

the mold and cured under ambient conditions. At least three anthracite pellets were

produced for each recipe and protocol, so as to achieve statistical replication.

4.2.3 The pyrolysis process

The pyrolysis of the anthracite pellets was conducted in a horizontal alumina tube

furnace as shown in Figure 4-2. A slow nitrogen gas flow (~ 2 standard cubic centimeters

per minute) was used to prevent the anthracite from burning. A three-step pyrolysis

97

procedure was employed. First the furnace was ramped up to the prescribed temperature

at 3 °C/min. Then this prescripted temperature was maintained for 2 hours. Finally the

furnace was cooled down to room temperature at 3 °C/min again. The pyrolyzed

anthracite pellets were then removed from the tube-furnace for further tests. Unless

otherwise identified, the prescripted maximum temperature was 1400 °C, which is close

to the temperature that the anthracite bricks would encounter in the preheating zone of a

cupola furnace. As identified below, other prescribed temperatures of 1000 °C-1300 °C

were also tested so as to investigate the crystal structure change in the system as the

temperature increased.

4.2.4 SEM, EDS, and TEM

SEM and EDS were performed on an FEI Quanta 200 Environmental SEM. The

instrument was operated under low-vacuum conditions (10-103 Pa) using a Gaseous SE

detector. The high voltage was set at 20 kV and the spot size of the electron beam was

set at 4 nm. The EDS spectrums were collected from 0-10 keV within 60-seconds.

A transmission electron microscope (TEM, Model 2010, JEOL, Tokyo, Japan)

was used for electron diffraction; and to determine the morphology and particle size.

4.2.5 Ambient-temperature XRD and real-time high-temperature XRD

Ambient temperature XRD patterns were obtained via a PANalytical X’Pert Pro

MPD diffractometer. The diffraction pattern within two theta of 5-70 degree was

collected. For these analyses, the pellets were crushed into powders by the ball mill and

placed into a special aluminum powder sample holder. So the powder XRD pattern was a

bulk measurement and represented the average situation in the pyrolyzed pellets.

98

Real-time high-temperature XRD tests were also performed on the PANalytical

X’Pert Pro MPD diffractometer. The anthracite pellets were pulverized before the

pyrolysis. Next, the powders that contained both silicon and anthracite were placed on an

alumina strip. A vacuum condition was provided around the alumina and the strip was

heated from room temperature to 1400 °C at 100 °C per minute. Once the temperature

reached 1400 °C, the diffractometer started to collect diffraction signals from 10° to 70°

repeatedly. Each scan from 10° to 70° took about 7 minutes, so the real time change of

the XRD pattern at 1400 °C could be observed in this way.

4.2.6 Unconfined compressive strength

The unconfined compressive strength of the pyrolyzed anthracite pellets were

determined by a Simpson-Gerosa electronic universal sand strength machine (model

42104). A horizontally moving arm applied pressure on an unconfined pellet until failure.

The final unconfined compressive strength was calculated based on the cross sectional

area of the original pellet sample.

4.3 Results and Discussion

4.3.1 Morphology and crystal structure of the silicon carbide nanowires formed at

1400 °C

Pellets that contained 9% silicon, 1% collagen, and 90% anthracite fines (mesh

#10 × 80) were pyrolyzed at 1400 °C. After the thermal treatment, the surface of

anthracite fines turned from black to light green. This change in color indicated that the

99

surface of the anthracite fines had been covered by the pyrolysis products. The pellets

had a post-pyrolysis unconfined compressive strength of 690 kPa.

The SEM images of these treated anthracite fines in Figure 4-3 exhibited that

intensive SCNWs were formed from the surface of the anthracite fines, and they almost

covered the anthracite fines entirely. The high resolution SEM images in Figure 4-3b&c

show that the SCNWs grew into a massive network. It seemed that many nanowires

connected with each other. These junctions of nanowires could be confirmed from the

TEM image in Figure 4-4. The EDS response from the nanowires indicated that carbon

and silicon were the major elements in the nanowires. Trace amounts of oxygen and

aluminum were also frequently detected by the EDS.

The TEM images of the nanowires (Figure 4-4a) confirmed that the diameters of

the nanowires ranged from 30 to 60 nm. One unique feature of the SCNWs growth found

in this study is that some of the SCNWs joined to one another at junctions. Some

SCNWs jointed almost perpendicular into other SCNWs, and one of the wires often

stopped growing as it teed into another. From the structural view point, these junctions

were perceived as desirable in the binding application as they provided extra strength for

the system. The trace of stacking of the 3C-SiC (111) lattice planes along the [111]

direction could be identified from some portions of the SCNWs with high density stack

faults in Figure 4-4a. Also in the upper-left corner of Figure 4-4a, the [-1, 1, 0] zone

diffraction pattern from a nanowire suggested that stack faults were present on the (111)

planes. This [111] growth is typical for SCNWs (Attolini et al. 2009; Dai et al. 1995;

McMahon et al. 1991; Park et al. 2004).

100

Further investigations of these branching points by HRTEM (Figure 4-4b-d)

revealed that these branching were different from the typical epitaxial growth of branched

nanowires. At the vicinity of a branching point, it seemed that the (111) lattice planes of

the backbone nanowires extruded out and formed the branching nanowires. The stacking

faults in the branching nanowires indicated that growth direction of the branching

nanowire was normal to the nanowires axial. However, this type of growth is not

sustainable. As seen in the TEM images, the branching nanowires either joined into

another backbone nanowire or bent back to the typical growth direction which is parallel

to the axial.

In the lower left corner of Figure 4-4a, twin growth of two SCNWs was also

observed, which normally shows a 70.5° between the growth direction and the (111)

plane (Choi et al. 2004; McMahon et al. 1991; Wang et al. 2008).

XRD patterns appear in Figure 4-5 for raw anthracite; pyrolyzed anthracite; raw

anthracite fines with 9% silicon powder and 1% collagen; and the blend of anthracite, 9%

silicon powder and 1% collagen after the 1400 °C pyrolysis. As shown, some

crystallized silicate species appeared even in the raw anthracite. These were mainly

quartz, muscovite, and kaolinte. After the pyrolysis at 1400 °C, the crystal structure of its

silicates changed dramatically. The distinctive quartz peaks of the raw materials mostly

disappeared in the pyrolyzed anthracite. The kaolinte and muscovite that had been found

in the raw materials transformed into mullite by the thermal treatment (Katsuki et al.

2000). Meanwhile the carbon structure in the anthracite became more organized after the

pyrolysis, as detected in the XRD.

101

When silicon powders were pyrolyzed together with the anthracite fines at 1400

°C for 2 hours, the elemental silicon was depleted. Meanwhile, diffraction peaks from

3C-SiC were found in the XRD pattern. Also, the 1400 °C pyrolysis yielded cristobalite

(SiO2) which had not been present in the raw materials but formed when kaolinite

transformed to mullite with the excess silica.

This formation of silicon carbide at 1400 °C was very fast as shown by the real-

time XRD patterns in Figure 4-6. When the temperature reached 1400 °C, the elemental

silicon had already disappeared from the XRD pattern. Diffraction peaks from the (111)

lattice plane of 3C-SiC appeared in the second scan at 1400 °C. The intensity of this

response did not significantly increase with yet further time than 14 minutes. Indeed,

even after 2 hours at 1400 °C, this SiC intensity had not grown more.

4.3.2 The effect of pyrolysis temperature on the development of the SCNWs

The morphology and crystal structure was a function of temperature, as shown by

the Figure 4-7 and Figure 4-8 data. The tests employed pellets made from anthracite fines

(mesh # 10 x 80) with 9% silicon powder and 1% collagen binder. For these tests, the

pyrolysis chamber ramped temperatures up to the maximum shown (1000-1400 °C), and

held at that temperature for 2 hours; then the temperature was ramped to ambient. As

shown for the ambient temperature condition in Figure 4-8a, the collagen binder adhered

the silicon powders onto the surface of the anthracite fines. The EDS spectrum on a

silicon particle at room temperature exhibited only responses from the elemental silicon

(Figure 4-8b).

After being pyrolyzed at 1000 °C, the anthracite pellet became a pile of fines,

since the collagen adhering bonds had become broken under this high temperature

102

condition. The kaolinite and muscovite in the raw anthracite had also disappeared by this

temperature, but the formation of mullite had not yet taken place. Most of the silicon

powders were still in the elemental cubic crystal structure as detected in the XRD in

Figure 4-7. However, trace amounts of carbon and oxygen were detected by EDS (Figure

4-8d) on the surface of the silicon powders.

When the maximum thermal treatment temperature was increased to 1100 °C, the

anthracite pellet could hold its shape after the pyrolysis. Both XRD and SEM confirmed

that the formation of silicon carbide had already started at a temperature of 1100 °C. The

3C-SiC peaks could be barely recognized from the XRD pattern at this point (Figure 4-7).

SEM images showed some SCNWs around the silicon powders (Figure 4-8f). Most of

the silicon powders were bound on the surface of anthracite fines by the SCNWs. The

binding strength between two anthracite fines was very weak. The whole pellet had an

unconfined compressive strength that was less than 70 kPa.

As the pyrolysis temperature increased to 1200 °C, only a small amount of

element silicon was left in the mixture. The XRD pattern showed a clear presence of

silicon carbide (Figure 4-7). The SEM images of the pyrolysis product revealed that the

SCNWs had already grown very well at 1200 °C (Figure 4-8h). The SCNWs started

providing enough binding strength to hold the anthracite fines together; and the pyrolyzed

pellet had an unconfined compressive strength that was about 400 kPa.

As temperature increased to 1400 °C, the formation of silicon carbide was

accelerated and the elemental silicon was almost depleted during the pyrolysis period.

Other obvious differences between 1200 °C and 1400 °C were the formation of

cristobalite and disappearance of the quartz as discussed above.

103

4.3.3 The effect of silicon content on the growth of SCNWs at 1400 °C

In order to determine the effect of less silicon in these pellet mixtures, pellets with

4.72% and 1.94% silicon were made by using the #10 × 80 anthracite fines. These pellets

were pyrolyzed at 1400 °C, following the same pyrolysis procedure as above. The SEM

images of these pyrolyzed anthracite pellets are shown in Figure 4-9. As compared to

Figure 4-3, it is clear that the proportion of silicon powder decisively impacted the

growth of the nanowires. With 4.72% silicon (Figure 4-9a), the anthracite fines were not

covered by SCNWs as heavily as they had when the pellets included 9% silicon. None-

the-less, the nanowires grew very well (Figure 4-9b); and their morphology was similar

to those that had developed when 9% silicon was included. When the silicon content

decreased to 1.74%, the SCNWs did not grow very well with such low silicon content.

As shown in Figure 4-9d, only a very thin layer of milky material became attached to the

surface of the anthracite fine. It was hard to find a single distinct nanowire from this

pyrolyzed sample.

4.3.4 Discussion on the growth mechanism

It is known that SCNWs could be grown through either VLS or VS mechanism.

One major difference between these two growths is that metal catalysts are usually

important for the formation of the interim liquid phase in VLS growth. Although

metallic catalysts were not added into the system, plenty of natural metals (such as Al and

Fe) were present in the anthracite that was tested in this study. Others have used iron as

the catalyst for SCNW growth (Lee et al. 2008; Li et al. 2010a; Zhang et al. 2001), and

Aluminum also was reported to catalyze the growth of SCNWs (Deng et al. 2002).

However, there was no direct evidence to verify that the SCNWs were grown through the

104

VLS route in this study. No clear image of metal spheres at the nanowire tips was

observed through the SEM and TEM. EDS spectrums at the nanowires tips showed

similar compositions as those in the middles of the nanowires. Also the formation of

massive nanowires on the anthracite surface requires tremendous amount of metal

nanoparticles to lead the growth if VLS was the only growth mechanism.

Furthermore, the branching patterns of the SCNWs in this study can hardly be

explained by VLS growth. Branches of nanowires obtained though VLS growth were

usually epitaxial branching and were achieved through more than one step (Wang et al.

2004). Thus the backbone nanowire was fully grown first, and then the branching

nanowire was grown from the backbone nanowire led by the metal nanopaticle doped on

the backbone nanowire. The branching nanowires in this study, however, showed a

strange growth pattern around the branching point. To grow the nanowire section

normally to its axial through VLS growth, the shape of the liquid nano-droplet will

contradict with basic law of physics. Branches in other types of nanowires formed

through one-step VS growth were also observed by others (Shen et al. 2009).

Another important feature for both VLS and VS growth has been the continuous

supply of raw materials for the nanowire growth from the vapor phase. However, in

contrast, for the study herein, the carbon and silicon instead originated from preloaded

solid-phase precursors (anthracite fines and silicon powders) that were bond together

with collagen. The carbon and silicon sources should transit through a vapor phase as

they participate in the vapor deposition. Conventionally, high temperature pyrolysis can

volatilize some carbon from solid anthracite, yielding hydrocarbons and COx. As for

silicon, although the melting point for silicon is higher than the pyrolysis temperatures

105

used in this study, the silicon powders used herein apparently transformed to vapors at

temperature as low as 1100 °C. It was noticed that after being pyrolyzed at 1100 °C and

1200 °C, some parts on the surface of the pellets were covered by cotton-like fibers.

These amorphous fibers contained mainly silicon and oxygen, as detected by EDS. They

were several microns in diameter and could grow to the length of one centimeter. These

cotton-like fibers were not seen in the 1000 °C and 1400 °C products. The slower SiC

forming rate at 1100 °C and 1200 °C allowed the unused silicon oxide vapor to diffuse

out of the pellets and nucleate on the surfaces to form the SiOx fibers. In the case of

insufficient silicon content or lower pyrolysis temperature, there was not enough supply

of vaporized silicon species. This resulted in the poor growth of SCNWs with either low

silicon content or at temperatures below 1100 °C.

4.3.6 Summary

SCNWs could be grown from anthracite surfaces through a simple protocol that

employed relatively inexpensive solid precursors. The SCNWs were grown through VS

mechanism. The growth of SCNWs started at around 1100 °C which was necessary to

transfer Si-precursors into the vapor phase. The SCNWs produced in this study have the

typical cubic zinc-blende crystal structure. Enough preloaded Si-precursor was also

necessary to form the SCNWs. These nanowires provided strong binding strength to hold

the anthracite fines together even after exposure to extremely high temperature. Results

in this study showed a promising application of SCNWs in reinforcing carbon based

materials.

106

Figure 4-1 EDS spectrum of the ash content from the anthracite used in this study

107

Figure 4-2 The pyrolysis setup

108

Figure 4-3 SEM images of silicon carbide nanowires grown from anthracite fines (#10 × 80, with 9% silicon and 1% collagen) at 1400 °C. Inserts in d are EDS spectrums from

different locations on a nanowire

109

Figure 4-4 TEM image of the SCNWs grown from anthracite fines (#10 × 80, with 9%

silicon and 1% collagen) at 1400 °C. The up corner insert in a is the selected area electron diffraction pattern of the SCNWs

110

Figure 4-5 XRD patterns of raw anthracite fines, and anthracite fines with 9% silicon

powders and 1% collagen before and after thermal treatment at 1400 °C

111

Figure 4-6 Real-time change of the XRD pattern at 1400 °C within mixed anthracite and

silicon powders (9% silicon and 91% anthracite)

112

Figure 4-7 Crystal structure change as temperature increased within anthracite pellets

made from #10 × 80 anthracite fines with 9% silicon content

113

Figure 4-8 Change of morphology as temperature increases within anthracite pellets

made from #10 × 80 anthracite fines with 9% silicon content (a-25 °C; c -1000 °C; e,f-1100 °C; g, h-1200 °C, b and d show the EDS spectrums on silicon particles treated at the

25 °C and 1000 °C respectively)

114

Figure 4-9 SEM images of SCNWs grown from anthracite fines (#10 × 80) with different

silicon contents (a,b-4.72%, c,d-1.74%) at 1400 °C

115

Chapter 5 Conclusions

5.1 Conclusions

The traditional iron casting process consumes vast amount of raw materials and

energy in the production. The linear usage of these raw materials and energy certainly

resulted in tremendous waste of resources. Two engineering solutions were proposal and

tested to increase the sustainability of the iron casting industry. The improved material

and energy flow within the iron casting system achieved by these engineering solutions

are shown in Figure 5-1 and Figure 5-2 respectively.

The first solution is to reclaim the wasted thermal energy from the cupola furnace

exhaust gas to produce porous carbon in-situ; and then apply this porous carbon to

removal the VOC emissions from the iron casting process; finally the saturated porous

carbon could be reused in the green sand molds. The second solution is to modify the

material flow into the cupola furnace, and bind waste anthracite fines with a low-

emission collagen-based binder and ceramic additives as a replacement of the expensive

foundry coke.

The research results from this study confirmed that sustainable engineering

designs can improve the material and energy flow in the iron casting process and benefit

the iron casting industry both environmentally and economically. Results from this

research also made many new original contributions to science and engineering as listed

in the following section.

116

5.1.1 Original contributions to science and engineering

1. The micro-structures of the raw coal materials, especially the

graphitization stage had significant effect on the volume and robustness of the pore

structure developed under variable pyrolysis conditions. The condensed near-graphite

structure of the anthracite micro-crystalline yielded little pore volume after the pyrolysis.

The bituminous coals yield considerable amount of pore volume after releasing the

volatile matters at 600 °C. All these pores were closed when temperature reached 900 °C.

The lignite developed the most pore volume in its random arranged small graphene-layer

groups. And this random arranged structure has good resistance to graphitization effect at

high temperature. Therefore, the pore structure in the lignite is more robust under various

pyrolysis temperatures.

2. The grain size of the raw lignite in fact affected the final pore volume

developed under simple pyrolysis conditions as measured by argon adsorption in the

ASAP 2010. Treated at the same thermal conditions, the lignite with bigger grain size

developed much less pore volume than the lignite with smaller grain size did. Grain size

affected the pore structure development within the lignite because of the restricted

diffusion of several volatile organic species generated during the pyrolysis. TGA-MS

showed that lignite with bigger grain size released much less phenols, naphthalenes,

cresol, and aniline than the lignite with smaller grain size did. These species did not

diffuse out of the bigger lignite grain and blocked some pores. However, these blockages

could be easily broken by steams at high temperature.

3. Pyrolysis developed pores in the coals in two stages. In the first stage, the

pores were developed by releasing the volatile organic compounds at temperature

117

below 600 °C. The second stage was between 600 °C and 800 °C, in which the

oxygen crosslinks between graphene layers were broken to create more pore volume.

Both the bituminous coal and the lignite developed some pore volume at 600 °C, and had

the highest pore volume at 800 °C. Beyond 800 °C some pores were lost in both the

bituminous coal and the lignite. Beyond 800 °C the graphitization took place in the coals

and closed some previous developed pores.

4. At fixed pyrolysis temperature, the pore structures in the coals will finally

be stable after the pore development and graphitization. After that, longer pyrolysis

duration will not change the pore structure any more.

5. Presoaked water with the raw coal materials can have some activation

effect in the simple pyrolysis system. However, the original moisture content in the

lignite did not open more pores during the pyrolysis because of its relative small amount.

By soaking the already pyrolyzed lignite in water and re-pyrolyze in the pyrolysis system,

10% more pore volume could be created.

6. The binding strength of the collagen binder could be increased by

crosslinking the collagen binder with fructose. In this reaction, the amino functional

group on the collagen peptide chain linked with the aldehyde group in fructose.

7. Si-containing materials added into the anthracite pellets provided some

post-pyrolysis strength for the anthracite pellets through different mechanism.

Silicon carbide was not formed from all the Si-containing materials tested; and the redox

potential of the silicon did play an important role in forming silicon carbide. The

crystallized Si (+4) in the kaolinite could not form silicon carbide with anthracite under

the conditions tested in this study. The amorphous Si (+4) in the sodium silicate solution

118

did form trace amount of silicon carbide with anthracite. The zero-valent silicon metal

formed tremendous amount of silicon carbide with the anthracite as detected by both

XRD and SEM.

8. Silicon carbide nanowires could be grown from the anthracite surface

through a simple pyrolysis step when zero-valent silicon metal was added.

Continuous supply of raw materials from the vapor phase is necessary for the silicon

carbide nanowires to grow. The elemental silicon can be transferred easily into the vapor

phase by reacting with oxygen to form silicon monoxide. With the zero-valent silicon, the

anthracite pellets had the highest post-pyrolysis unconfined compressive strength. Thus,

the silicon carbide nanowires generated in-situ had strong mechanical properties (Wong

et al. 1997) on themselves and firmly attached to the anthracite fines. The strength of the

bindered anthracite pellets reached as high as 535 psi (typical foundry coke has ~400 psi)

in this study.

9. The formation of silicon carbide nanowires between silicon powders and

the anthracite fines occurs at temperature as low as 1100 °C. The silicon carbide

nanowires formed between the silicon powders and the anthracite fines at

temperatures from 1100 °C to 1400 °C are highly crystallized 3C-SiC. Real time X-

ray diffraction at 1400 °C confirmed that the formation rate of silicon carbide is very fast

at this temperature. The formation of silicon carbide was accomplished within in ten

minutes at 1400 °C. However at lower temperatures the reaction rate was much lower.

The growth of the silicon carbide nanowires was not assisted by additional catalysts.

10. The grain size of the anthracite fines had significant effect on the growth

of the silicon carbide nanowires as well as the final strength of the nanowire

119

bindered anthracite fines. Smaller void spaces between the pulverized anthracite

powders restricted the length of the silicon carbide nanowires. However, the smaller gaps

between the anthracite powders allowed the silicon carbide nanowires to directly connect

the neighboring anthracite particles.

5.1.2 The impacts on sustainability beyond the iron casting industry

Although the engineering solutions presented in this thesis focused mainly on

addressing the sustainability issues in the iron casting industry, these approaches also

benefit the society and other industries in the following areas:

1. By using the in-situ produced porous carbon adsorbent the usage of commercial

activated carbons can be reduced. As a result the energy consumption and other

environmental pollutions from producing the replaced commercial activated

carbons could be eliminated.

2. By replacing a part of the foundry coke, the energy consumption, the CO2

emission and other pollutions from the coking process can be reduced.

3. Using bindered anthracite fines as coke replacement added value to the otherwise

waste anthracite fines, this provides more revenue to the mining industry

4. Using the collagen-based binders, which came from food by products, provide

more revenue to the food processing industry

5.2 Recommendations for future studies

The silicon carbide nanowires bindered anthracite fines had very strong post-

pyrolysis unconfined compressive strength. However, as mentioned in the section 3.3 that

120

the silicon to anthracite ratio required in the bindered anthracite to maintain an acceptable

post-pyrolysis strength is higher than the elemental silicon to coke ratio originally added

in the cupola furnace for casting quality. Therefore, without adding extra elemental

silicon the amount of foundry coke that can be replaced by the bindered anthracite fines

is limited. According to the experimental results in this study, it is estimated that the

bindered anthracite fines can replace 25% to 40% foundry coke in the cupola furnace

without adding extra elemental silicon. Adding extra elemental silicon will significantly

increase the cost and decrease the environmental impact by replacing coke with the

bindered anthracite fines. The elemental silicon is relative expensive and its production

also consumes energy. Based on this fact, three recommendations were proposed for

future studies on replacing coke by bindered anthracite fines:

1. To seek for low cost and renewable silicon sources that can be utilized in the

binding system.

2. To further improve the silicon carbide nanowire binding strength so that even less

amount of silicon is required for the acceptable binding strength

3. To explore different binding mechanisms

Silicon is abundant in the earth crust and some renewable materials contain

considerable amount of silicon. One promising candidate for renewable silicon sources is

rice husks. On weight base about 20% in the rice husk is amorphous silicon dioxide. Rice

husk is the by product of rice and not edible by human being. Every year huge amount of

rice husks are produced all around the world, for example in the year 2002 more than 115

million tons of rice husks were produce (Bronzeoak_Ltd 2003). Most of these rice husks

were dumped or burned, and currently research is being conducted to utilize rice husks as

121

building materials, fertilizers, insulation materials, or fuel. Based on the high silicon

amount in rice husks, especially in rice husk ashes, it has a great potential to be used as

the silicon source for the silicon carbide nanowires (Adylov et al. 2003; Bhat and Sanghi

1987; Krishnarao 1993; 1995; Krishnarao and Godkhindi 1992; Krishnarao and

Subrahmanyam 1996; Kumar and Godkhindi 1996; Kuskonmaz et al. 1996; Martinez et

al. 2006; Mishra et al. 1995; Nayak et al. 1996; Padhi and Patnaik 1995; Patel 1989;

Rambo et al. 1999; Ray et al. 1991; Rodriguez-Lugo et al. 2002; Singh et al. 2002; Singh

et al. 1995; Singh et al. 1993). Some preliminary experiments were conducted using the

raw rice husks as silicon source in the anthracite pellets and the results could be found in

Appendix B. The raw rice husks did not improve the binding strength significantly, but

only 20% of the raw husks are silicon dioxide. The silicon dioxide amount in the rice

hush ashes is more than 90%. Future studies can focus on using the rice husk ashes

instead of the raw rice husks.

The binding strength of the silicon carbide nanowires may probably be further

increased for the same amount of silicon powders used. The improvement in silicon

carbide nanowire binding strength may be achieved by controlling the growth of the

silicon carbide nanowires to reduce the thickness of the silicon carbide nanowires. In this

way, more silicon carbide nanowires can be produced from the same amount of silicon

precursors. It is know that silicon carbide nanowires are stronger than silicon carbide in

bulk materials (Wong et al. 1997). Therefore it is logical to estimate that several thinner

nanowires will provide more strength than one thicker nanowire. However, because more

sophisticated procedures during the pyrolysis are necessary to produce thinner silicon

carbide nanowires, some modifications may needed in the cupola preheating zone to

122

adapt the more complicated pyrolysis process and utilize the free energy in this zone at

the same time.

Although a significant increased in the nanowire binding strength was obtained by

pulverizing the anthracite fines, further reducing the anthracite grain size may not have

noticeable improvement on the binding strength. The SEM images of the post-pyrolysis

anthracite pellets made from pulverized anthracite fines show that the pulverized

anthracite powders were not totally covered by the silicon carbide nanowires as in the

pellets made by original anthracite.

The finally recommendation is to research on other binding mechanisms in

addition to the silicon carbide nanowire binding. It should be noticed that the high

temperature strength of foundry cokes came from its fused carbon structure which was

generated during the coking process. Anthracite’s carbon structure also has very high

strength at elevated temperatures. Some carbon sources that can be fused at high

temperature may be used to bind the anthracite fines. Green petroleum coke fine is a

potential candidate for this task.

123

Figure 5-1 The change of material and energy flow in the iron casting process by introducing the in-situ porous carbons

124

Figure 5-2 The modified material flow into the cupola furnace by replacing a part of the coke with bindered anthracite fines

125

Appendix A: The crystal structure of 3C-SiC and the stacking sequence of the SiC nanowires

The crystal structure of 3C-SiC (or β-SiC) is a face-centered cubic zinc blender

structure. In this appendix the crystal structure of 3C-SiC was drawn by the crystal

structure molding software “VESTA”. The pictures in this appendix are aimed to explain

some terms used in the main text for describing the 3C-SiC structures.

Table A-1 listed all the parameters used for the crystal structure molding. Figure

A-1 shows a unit cell of 3C-SiC in both “ball and stick” mold and “space filling” mold. It

can be seen from the picture that each unit cell contains 4 carbon atoms (all enclosed in

the boundary). Also each unit cell contains 4 total silicon atoms because each silicon

atom occupying a vertex is shared by 8 unit cells (8x0.125=1) and each silicon atom in

the center of a facet is shared by 2 unite cells (6x0.5=3). In the “ball and stick” mold it

shows that 4 of the vertex silicon atoms are not bonded with any of the carbon atoms

within the cell.

Figure A-2a shows the one of the (111) lattice planes of the 3C-SiC unit cell. This

lattice plane intersects with the three axles at the point of 1 unit lattice constant. The [111]

direction is normal to the (111) lattice planes. Figure A-2b shows the parallel (111) lattice

planes in a 5x5x5 cell group. The 3C-SiC nanowires were grown by stacking the (111)

lattice planes along the [111] direction layer by layer. It can be seen in Figure A-2b that

each layer in the 3C-SiC is actually a bi-layer structure combined by a carbon layer and a

silicon layer.

126

Figure A-3 shows a cross-section of the 5x5x5 cell group from Figure A-2b. It

can be seen that each silicon atom in the first layer occupies a position A. The 3 carbon

atoms in the same bi-layer bonded with one of the silicon atoms will be assigned to a

position B. Then each of these three carbon atoms will connect with a silicon atom in the

second layer, and this silicon atom occupies the same B position. Then the carbon atoms

in the second layer will be assigned to the C positions, which will lead the silicon atoms

in the third layer to the C position. And so on, in the fourth layer, the silicon atoms will

occupy the A positions again. Thus in this system the layers are stacked in the sequence

of ABCABCABC…, and three positions in a whole cycle. This is why the crystal

structure named 3C-SiC. “3” suggests that there are three stacking positions in a cycle,

and “C” means cubic.

127

Table A-1 Parameters used for the 3C-SiC crystal structure molding

Parameters Values (units) Crystal structure Cubic zinc blender Space group number 216 Space group symbol F43m Lattice constant 4.3596 (Å) Bond length 1.89 (Å) Silicon atom radius in space filling 1.18 (Å) Carbon atom radius in space filling 0.77 (Å)

128

Figure A-1. A unit cell of 3C-SiC in “ball and stick” and “space filling” forms.

129

Figure A-2. The (111) lattice plane of the 3C-SiC and the [111] direction

[111]

130

Figure A-3. The stacking sequence of 3C-SiC. [111] is normal to the paper.

131

Appendix B: Appling raw rice husks as an addition renewable silicon sources for binding the anthracite fines

Raw rice husks were grinded into powders and 20 gram of these powders

bindered with 100 gram anthracite fines following the same pellet making procedure used

in this study. Then the pellets were pyrolyzed at 1400 °C by the same pyrolysis process.

After the pyrolysis there was no strength left in the pellets. The SEM images of the

pyrolyzed pellets in Figure B-1 show no sign of the formation of silicon carbide

nanowires. Indeed, many small spheres on the anthracite surface were observed in the

SEM images. These small spheres most likely came from the amorphous silicon dioxide

in the raw rice husks. Figure B-2 is the XRD pattern of the rice husk pellets after the

pyrolysis at 1400 °C. This pattern is very close to that of the pyrolyzed anthracite.

Rice husks powders were also added together with silicon powders in the

anthracite pellets. With 5 gram rice husk powders and 5 gram silicon powders in 100

gram anthracite fines the post-pyrolysis unconfined compressive strength of the

anthracite pellets was about 70 psi which is slightly higher than the pellets made from 5

gram silicon powders and 100 gram anthracite fines.

132

Figure B-1 Anthracite pellets with 16.7% rice husks after the pyrolysis at 1400 °C

133

0

500

1000

1500

2000

2500

0 10 20 30 40 50 60 70 80

Figure B-2 XRD pattern of anthracite pellets with 16.7% rice husks after the pyrolysis at 1400 °C

134

Figure B-3 Anthracite pellets with 4.5% rice husks and 4.5% silicon powder after the

pyrolysis at 1400 °C

135

0

500

1000

1500

2000

2500

0 10 20 30 40 50 60 70 80

Figure B-4 XRD pattern of anthracite pellets with 4.5% rice husks and 4.5% silicon powder after the pyrolysis at 1400 °C

136

References:

Adylov, G. T., Faiziev, S. A., Paizullakhanov, M. S., Mukhsimov, S., and Nodirmatov, E. (2003). "Silicon carbide materials obtained from rice husk." Technical Physics Letters, 29(3), 221-223.

Allen, G. R., Archibald, J. J., and Keenan, T. (1991). "Hazardous air pollutants: a challenge to metal casting industry." Transactions of American Foundry Society, 99, 585-593.

Alonso, M. J. G., Alvarez, D., Borrego, A. G., Menendez, R., and Marban, G. (2001). "Systematic Effects of Coal Rank and Type on the Kinetics of Coal Pyrolysis." Energy & Fuels, 15, 413-428.

Attolini, G., Rossi, F., Bosi, M., Watts, B. E., and Salviati, G. (2008). "Synthesis and characterization of 3C-SiC nanowires." J. Non-Cryst. Solids, 354(47-51), 5227-5229.

Attolini, G., Rossi, F., Fabbri, F., Bosi, M., Watts, B. E., and Salviati, G. (2009). "A new growth method for the synthesis of 3C-SiC nanowires." Mater. Lett., 63(29), 2581-2583.

Avallone, E. A., Baumeister, T., and Sadegh, A. (2006). Standard handbook for mechanical engineers, 11th Ed., McGraw-Hill Professional.

Baek, Y., Ryu, Y., and Yong, K. (2006). "Structural characterization of beta-SiC nanowires synthesized by direct heating method." Materials Science & Engineering C-Biomimetic And Supramolecular Systems, 26(5-7), 805-808.

Bailey, A. J., Sims, T. J., Avery, N. C., and Miles, C. A. (1993). "Chemistry Of Collagen Cross-Links - Glucose-Mediated Covalent Cross-Linking Of Type-Iv Collagen In Lens Capsules." Biochem. J, 296, 489-496.

Bansal, R. C., Donnet, J. B., and Stoeckli, F. (1988). Active Carbon, 1 Ed., Marcel Dekker, New York.

Berkowitz, N. (1985). The chemistry of coal, Elsevier, Amsterdam ; New York.

Bhat, B. V. R., and Sanghi, G. P. (1987). "Increase in the yield of silicon carbide whiskers from rice husk." Bull. Mater. Sci., 9(4), 295-303.

Bronzeoak_Ltd. (2003). "Rice Husk Market Study." Department of Bussiness Innovation and Skills, The United Kingdom.

137

Cai, H. Y., Guell, A. J., Chatzakis, I. N., Lim, J. Y., Dugwell, D. R., and Kandiyoti, R. (1996). "Combustion reactivity and morphological change in coal chars: effect of pyrolysis temperature, heating rate and pressure." Fuel, 75(1), 15-24.

Centeno, T. A., Pis, J. J., Pajares, J. A., and Fuertes, A. B. (1995). "Microporous structure of chars produced by pyrolysis of preoxidized coals." J. Anal. Appl. Pyrolysis, 34(1), 13-28.

Chan, B. P., and So, K. F. (2005). "Photochemical crosslinking improves the physicochemical properties of collagen scaffolds " Journal of Biomedical Materials Research Part A, 75A(3), 689-701.

Charulatha, V., and Rajaram, A. (2003). "Influence of different crosslinking treatments on the physical properties of collagen membranes." Biomaterials, 24, 759-767.

Chattopadhyaya, G., Macdonald, D., Bakhshi, N. N., Mohammadzadeh, J. S. S., and Dalai, A. K. (2006). "Preparation and Characterization of Chars and Activated Carbons from Saskatchewan Lignite." Fuel Process. Technol., 87(11), 997-1006.

Chen, J. J., Wu, R. B., Yang, G. Y., Pan, Y., Lin, J., Wu, L. L., and Zhai, R. (2008). "Synthesis and photoluminescence of needle-shaped 3C-SiC nanowires on the substrate of PAN carbon." J. Alloys Compd., 456(1-2), 320-323.

Chiche, P., Durif, S., and Pregermain, S. (1965). "Development of internal surface area of coals during carbonization." Fuel, 44(1), 5-28.

Chitsora, C. T., Muhlen, H. J., Van Heek, K. H., and Juntgen, H. (1987). "The influence of pyrolysis conditions on the reactivity of char in H2O." Fuel Process. Technol., 15, 17-29.

Chiu, S. C., Huang, C. W., and Li, Y. Y. (2007). "Synthesis of high-purity silicon carbide nanowires by a catalyst-free arc-discharge method." Journal Of Physical Chemistry C, 111(28), 10294-10297.

Choi, H. J., Seong, H. K., Lee, J. C., and Sung, Y. M. (2004). "Growth and modulation of silicon carbide nanowires." J. Cryst. Growth, 269(2-4), 472-478.

Cui, L., Song, W., Zhang, J., Yao, J., and Lin, W. (2007). "Influence of the Gas and Particle Residence Time on Fast Pyrolysis of Lignite." Journal of Energy Resources Technology, 129(2), 152-158.

Cypres, R., Planchon, D., and Braekman-Danheux, C. (1985). "Evolution of pore structure and active surface areas of coal and char during hydrogenation." Fuel, 64(10), 1375-1378.

Dai, H. J., Wong, E. W., Lu, Y. Z., Fan, S. S., and Lieber, C. M. (1995). "Synthesis and Characterization of Carbide Nanorods." Nature, 375(6534), 769-772.

138

Dempsey, T. C., LaFay, V. S., Heltner, S. L., and Taulbee, D. N. (1997). "Understanding properties of carbonaceous additives and their potential to emit benzene." Transactions of American Foundry Society, 105, 109-117.

Deng, S. Z., Li, Z. B., Wang, W. L., Xu, N. S., Zhou, J., Zheng, X. G., Xu, H. T., Chen, J., and She, J. C. (2006). "Field emission study of SiC nanowires/nanorods directly grown on SiC ceramic substrate." Appl. Phys. Lett., 89(2).

Deng, S. Z., Wu, Z. S., Zhou, J., Xu, N. S., Chen, R., and Chen, J. (2002). "Synthesis of silicon carbide nanowires in a catalyst-assisted process." Chem. Phys. Lett., 356(5-6), 511-514.

Duan, X., and Sheardown, H. (2005). "Crosslinking of collagen with dendrimers." Journal of Biomedical Materials Research Part A, 75A(3), 510-518.

Duz, M. Z., Tonbul, Y., Baysal, A., Akba, O., Saydut, A., and Hamamci, C. (2005). "Pyrolysis Kinetics and Chemical Composition of Hazro Coal Accoring to the Particle Size." J. Therm. Anal. Calorim., 81(2), 395-398.

Easler, T. E., Bradt, R. C., and Walker, P. L. (1982). "Mechanical-Properties of Formcoke." Carbon, 20(2), 154-154.

Easler, T. E., Bradt, R. C., and Walker, P. L. (1985). "Mechanical-Properties of Formcoke." Ironmaking & Steelmaking, 12(3), 118-123.

Easler, T. E., Bradt, R. C., and Walker, P. L. (1990). "Gasification Induced Densification of Form Coke." Fuel, 69(1), 124-125.

Easler, T. E., Bradt, R. C., and Walker, P. L. (1991). "Mechanical-Properties of Formcoke after Reaction in CO2 and CO/CO2 Environments." Carbon, 29(8), 1125-1134.

Energetics-Incorporated. (1999). "Energy and environmental profile of the U.S. metalcasting industry." Department of Energy Office of Industrial Technologies.

Ertem, M. E., and Ozdabak, A. (2005). "Energy balance application for Erdemir Coke Plant with thermal camera measurements." Appl. Therm. Eng., 25(2-3), 423-433.

Eyre, D. R., and Wu, J. J. (2005). "Collagen Cross-links." Top. Curr. Chem., 247, 207-229.

Feng, D. H., Jia, T. Q., Li, X., Xu, Z. Z., Chen, J., Deng, S. Z., Wu, Z. S., and Xu, N. S. (2003). "Catalytic synthesis and photoluminescence of needle-shaped 3C-SiC nanowires." Solid State Commun., 128(8), 295-297.

Feng, J., Li, W., and Xie, K. (2006). "Thermal Decomposition Behavior of Lignite by Pyrolysis-FTIR." Energy Sources Part A-Recovery utilization and environmental effects, 28(1-3), 167-175.

139

Figueiredo, J. L., Pereira, M. F. R., Freitas, M. M. A., and Orfao, J. M. M. (1999). "Modification of the surface chemistry of activated carbons." Carbon, 37(9), 1379-1389.

Fox, J. R., Adamovits, M., and Henry, C. (2002). "Strategies for reducing foundry emission." Transactions of American Foundry Society, 110, 1299-1309.

Fox, J. T., Cannon, F. S., Furness, J. C., Smith, P., Lewallen, S., and Goudzwaard, J. (2007). "Simultaneous Bond and Sand Reclamation using Advanced Oxidation, Induced Particle Collision and Discretionary Settling." Transactions of American Foundry Society, 115.

Fox, J. T., Cannon, F. S., Voigt, R. C., Furness, J. C., Goudzwaard, J. E., Wosoba, M., and Smith, P. B. (2008a). "Decreased Bond Consumption by Processing Baghouse Dust through Ultrasonic-Cavitation-Settling coupled to Advanced Oxidation." Transactions of American Foundry Society, 116.

Fox, J. T., Cannon, F. S., Voigt, R. C., Furness, J. C., Headington, F., Coan, D., and Lewallen, S. (2008b). "Waste Green Sand to Core Sand Reclamation, Demonstration via Casting Study, with Simultaneous Clay Recovery via a novel Ultrasonic-Cavitation System." Transactions of American Foundry Society, 116.

Franklin, R. E. (1951). "Crystallite Growth In Graphitizing And Non-Graphitizing Carbons." Proceedings of The Royal Society of London Series A-Mathematical and Physical Sciences, 209(1097), 196-218.

Fu, D. J., Zeng, X. R., Zou, J. Z., Li, L., Li, X. H., and Deng, F. (2009). "In situ synthesis and photoluminescence of SiC nanowires by microwave-assisted pyrolysis of methane." J. Alloys Compd., 486(1-2), 406-409.

Fujimoto, D., and Horiuchi, K. (1986). "Acceleration of glucose-mediated crosslinking of collagen by free lysine." Experientia, 42, 405.

Gao, Y. H., Bando, Y., Kurashima, K., and Sato, T. (2002). "SiC nanorods prepared from SiO and activated carbon." J. Mater. Sci., 37(10), 2023-2029.

Gavalas, G. R. (1982). Coal Pyrolysis, Elsevier Scientific Publishing Company, Amsterdam.

Gill, W. W., and Chaklader, A. C. D. (1984). "Material Characteristics Affecting Formcoke." Fuel, 63(10), 1385-1392.

Glowacki, C. R., Crandell, G. R., Cannon, F. S., Voigt, R. C., Clobes, J. K., Furness, J. C., McComb, B. A., and Knight, S. M. (2003). "Emissions studies at a test foundry using an advanced oxidation-clear water system." Transactions of American Foundry Society, 111, 579-598.

140

Green, R. A., and Olender, A. L. (1980). "An evaluation and comparison of foundry carbonaceous materials." Transactions of American Foundry Society, 88, 817-822.

Green, T., Kovac, J., Brenner, D., and Larsen, J. W. (1982). In: Coal Structure, R. A. Meyers, ed., Acdemic, New York.

Hanson, S., Patrick, J. W., and Walker, A. (2002). "The Effect of Coal Particle Size on Pyrolysis and Steam Gasification." Fuel, 81(5), 531-537.

Hayashi, J., Iwatsuki, M., Morishita, K., Tsutsumi, A., Li, C., and Chiba, T. (2002). "Roles of inherent metallic species in secondary reactions of tar and char during rapid pyrolysis of brown coals in a drop-but reactor." Fuel, 81(15), 1977-1987.

Hayashi, J., Norinaga, K., Yamashita, T., and Chiba, T. (1999). "Effect of Sorbed Water on Conversion of Coal by Rapid Pyrolysis." Energy & Fuels, 13(3), 611-616.

Hayashi, J., Takahashi, H., Doi, S., Kumagai, H., and Chiba, T. (2000). "Reactions in Brown Coal Pyrolysis Responsible for Heating Rate Effect on Tar Yield." Energy & Fuels, 14(2), 400-408.

Hong, S. M., Candelone, J. P., Patterson, C. C., and Boutron, C. F. (1996). "History of ancient copper smelting pollution during Roman and medieval times recorded in Greenland ice." Science, 272(5259), 246-249.

Howard, J. B. (1981). "Fundamentals of Coal Pyrolysis and Hydropyrolysis." In: Chemistry of Coal Utilization, M. A. Elliott, ed., John Wiley & Sons, New York, 665-784.

Huang, H., Wang, Y. J., and Cannon, F. S. (2009). "Pore structure development of in-situ pyrolyzed coals for pollution prevention in iron foundries." Fuel Process. Technol., 90(9), 1183-1191.

Hyun, J. S., Nam, S. H., Kang, B. C., and Boo, J. H. (2009). "Growth of 3C-SiC nanowires on nickel coated Si(100) substrate using dichloromethylvinylsilane and diethylmethylsilane by MOCVD method." Physica Status Solidi C - Current Topics In Solid State Physics, Vol 6, No 4, 6(4), 810-812.

Jenkins, R. G., Nandi, S. P., and Walker, P. L. (1973). "Reactivity of heat-treated coals in air at 500 oC." Fuel, 52(4), 288-293.

Jiang, W. D., Lee, I., and Yang, R. Y. K. (1988). "Microstructural Variations of Lignite, Subbituminous and Bituminous Coals and Their High Temperature Chars." Fuel Process. Technol., 18, 11-23.

Ju, Z. C., Ma, X. C., Fan, N., Li, P., Xu, L. Q., and Qian, Y. T. (2007). "High-yield synthesis of single-crystalline 3C-SiC nanowires by a facile autoclave route." Mater. Lett., 61(18), 3913-3915.

141

Kang, B. C., Lee, S. B., and Boo, J. H. (2004). "Growth of beta-SiC nanowires on Si(100) substrates by MOCVD using nickel as a catalyst." Thin Solid Films, 464, 215-219.

Kang, B. C., Lee, S. B., and Boo, J. H. (2005). "Growth of beta-SiC nanowires and thin films by metalorganic chemical vapor deposition using dichloromethylvinylsilane." Journal Of Vacuum Science & Technology B, 23(4), 1722-1725.

Katsuki, H., Furuta, S., and Komarneni, S. (2000). "Formation of novel ZSM-5/porous mullite composite from sintered kaolin honeycomb by hydrothermal reaction." J. Am. Ceram. Soc., 83(5), 1093-1097.

Khongwong, W., Imai, M., Yoshida, K., and Yano, T. (2009). "Synthesis of beta-SiC/SiO2 core-shell nanowires by simple thermal evaporation." J. Ceram. Soc. Jpn., 117(1362), 194-197.

Kirk, E. (1903). The Cupola Furnace, Henry Carey Baird & Co., Philadelphia.

Kok, M. V. (2003). "Coal Pyrolysis: Thermogravimetric Study and Kinetic Analysis." Energy Sources, 25(10), 1007-1014.

Kok, M. V., Ozbas, E., Karacan, O., and Hicyilmaz, C. (1998). "Effect of Particle Size on Coal Pyrolysis." J. Anal. Appl. Pyrolysis, 45(2), 103-110.

Krishnarao, R. V. (1993). "FORMATION OF SIC WHISKERS FROM RICE HUSK SILICA CARBON-BLACK MIXTURE - EFFECT OF PREHEAT TREATMENT." J. Mater. Sci. Lett., 12(16), 1268-1271.

Krishnarao, R. V. (1995). "EFFECT OF COBALT CATALYST ON THE FORMATION OF SIC FROM RICE HUSK SILICA-CARBON BLACK MIXTURE." Journal Of Materials Science, 30(14), 3645-3651.

Krishnarao, R. V., and Godkhindi, M. M. (1992). "STUDIES ON THE FORMATION OF SIC WHISKERS FROM PULVERIZED RICE HUSK ASHES." Ceram. Int., 18(1), 35-42.

Krishnarao, R. V., and Subrahmanyam, J. (1996). "Formation of SiC from rice husk silica-carbon black mixture: Effect of rapid heating." Ceram. Int., 22(6), 489-492.

Kumar, B., and Godkhindi, M. M. (1996). "Studies on the formation of SiC, Si3N4 and Si2N2O during pyrolysis of rice husk." J. Mater. Sci. Lett., 15(5), 403-405.

Kuskonmaz, N., Sayginer, A., Toy, C., Acma, E., Addemir, O., and Tekin, A. (1996). "Studies on the formation of silicon nitride and silicon carbide from rice husk." High Temp. Mater. Processes (London), 15(1-2), 123-127.

142

Lai, H. L., Wong, N. B., Zhou, X. T., Peng, H. Y., Au, F. C. K., Wang, N., Bello, I., Lee, C. S., Lee, S. T., and Duan, X. F. (2000). "Straight beta-SiC nanorods synthesized by using C-Si-SiO2." Appl. Phys. Lett., 76(3), 294-296.

Lee, J. S., Byeun, Y. K., Lee, S. H., and Choi, S. C. (2008). "In situ growth of SiC nanowires by carbothermal reduction using a mixture of low-purity SiO2 and carbon." J. Alloys Compd., 456(1-2), 257-263.

Lee, T., and Benson, C. H. (2004). "Sorption and degradation of alachor and metolachlor in ground water using green sands." Journal of Environmental Quality, 33(5), 1682-1693.

Leon y Leon D, C. A., and Radovic, L. R. (1994). "Interfacial Chemistry and Electrochemistry of Carbon Surface." In: Chem. Phys. Carbon, P. A. Thrower, ed., Marcel Dekker, Inc., New York, 213-294.

Li, F., and Wen, G. (2007). "A novel method for massive fabrication of beta-SiC nanowires." Journal Of Materials Science, 42(12), 4125-4130.

Li, G. Y., Li, X. D., Wang, H., and Liu, L. (2010a). "Long SiC nanowires synthesized from off-gases of the polycarbosilane-derived SiC preparation." Applied Physics A-Materials Science & Processing, 98(2), 293-298.

Li, G. Y., Li, X. D., Wang, H., Xing, X., and Yang, Y. (2010b). "SiC nanowires grown on activated carbon in a polymer pyrolysis route." Materials Science And Engineering B-Advanced Functional Solid-State Materials, 166(1), 108-112.

Li, P., Xu, L. Q., and Qian, Y. (2008). "Selective synthesis of 3C-SiC hollow nanospheres and nanowires." Crystal Growth & Design, 8(7), 2431-2436.

Li, Y. B., Xie, S. S., Wei, B. Q., Lian, G. D., Zhou, W. Y., Tang, D. S., Zou, X. P., Liu, Z. Q., and Wang, G. (2001a). "Aligned small alpha-SiC nanorods on beta-SiC particles grown in an arc-discharge." Solid State Commun., 119(1), 51-53.

Li, Y. B., Xie, S. S., Zou, X. P., Tang, D. S., Liu, Z. Q., Zhou, W. Y., and Wang, G. (2001b). "Large-scale synthesis of beta-SiC nanorods in the arc-discharge." J. Cryst. Growth, 223(1-2), 125-128.

Li, Z. J., Gao, W. D., Meng, A. A., Geng, Z. D., and Gao, L. (2009). "Large-Scale Synthesis and Raman and Photoluminescence Properties of Single Crystalline beta-SiC Nanowires Periodically Wrapped by Amorphous SiO2 Nanospheres 2." Journal Of Physical Chemistry C, 113(1), 91-96.

Li, Z. J., Li, H. J., Chen, X. L., Meng, A. L., Li, K. Z., Xu, Y. P., and Dai, L. (2003). "Large-scale synthesis of crystalline beta-SiC nanowires." Applied Physics A-Materials Science & Processing, 76(4), 637-640.

143

Liang, C. H., Meng, G. W., Zhang, L. D., Wu, Y. C., and Cui, Z. (2000). "Large-scale synthesis of beta-SiC nanowires by using mesoporous silica embedded with Fe nanoparticles." Chem. Phys. Lett., 329(3-4), 323-328.

Maloney, D. L., and Jenkins, R. G. (1985). "Influence of coal preoxidation and the relation between char structure and gasification potential." Fuel, 64(10), 1415-1522.

Marsh, H., and Menendez, R. (1988). "Carbons from pyrolysis of pitches, coals, and their blends." Fuel Process. Technol., 20(1-3), 269-296.

Martinez, V., Valencia, M. F., Cruz, J., Mejia, J. M., and Chejne, F. (2006). "Production of beta-SiC by pyrolysis of rice husk in gas furnaces." Ceram. Int., 32(8), 891-897.

McMahon, G., Carpenter, G. J. C., and Malis, T. F. (1991). "On The Growth-Mechanism Of Silicon-Carbide Whiskers." Journal Of Materials Science, 26(20), 5655-5663.

Menendez, J. A., Phillips, J., Xia, B., and Radovic, L. R. (1996). "On the Modification and Characterization of Chermical Surface Properties of Activated Carbon: In the Search of Carbons with Stable Basic Properites." Langmuir, 12(18), 4404-4410.

Meng, G. W., Zhang, L. D., Mo, C. M., Zhang, S. Y., Li, H. J., Qin, Y., and Feng, S. P. (1999a). "Synthesis of one-dimensional nanostructures - beta-SiC nanorods with and without amorphous SiO2 wrapping layers." Metallurgical And Materials Transactions A-Physical Metallurgy And Materials Science, 30(1), 213-219.

Meng, G. W., Zhang, L. D., Qin, Y., Feng, S. P., Mo, C. M., Li, H. J., and Zhang, S. Y. (1999b). "Formation of beta-SiC nanorods with amorphous SiO2 wrapping layers." J. Mater. Sci. Lett., 18(15), 1255-1257.

Meng, G. W., Zhang, L. D., Qin, Y., Mo, C. M., and Phillipp, F. (1999c). "Synthesis of beta-SiC nanowires with SiO2 wrappers." Nanostruct. Mater., 12(5-8), 1003-1006.

Meng, G. W., Zhang, L. D., Qin, Y., Phillipp, F., Qiao, S. R., Guo, H. M., and Zhang, S. Y. (1998). "Chemical vapor deposition of beta-SiC nanowires on granular active carbon cylinders loaded with iron nanoparticles inside the pores." Chinese Physics Letters, 15(9), 689-691.

Meng, G. W., Zhang, L. S., Mo, C. M., Phillip, F., Qin, Y., Li, H. J., Feng, S. P., and Zhang, S. Y. (1999d). "beta-SiC nanorods with uniform amorphous SiO2 wrappers on the outside surface." Mater. Res. Bull., 34(5), 783-790.

Miranda, A., Cuevas, J. L., Ramos, A. E., and Cruz-Irisson, M. (2009). "Quantum confinement effects on electronic properties of hydrogenated 3C-SiC nanowires." Microelectron. J., 40(4-5), 796-798.

144

Mishra, P. K., Nayak, B. B., and Mohanty, B. (1995). "SYNTHESIS OF SILICON-CARBIDE FROM RICE HUSK IN A PACKED-BED ARC REACTOR." J. Am. Ceram. Soc., 78(9), 2381-2384.

Moldenke, R. (1917). The Principles of Iron Foundring, McGraw-Hill Book Company Inc., New York.

Moore, B. C., Cannon, F. S., Westrick, J. A., Metz, D. H., Shrive, C. A., Demarco, J., and Hartman, D. J. (2001). "Changes in GAC pore structure during full-scale water treatment at Cincinnati: a comparison between virgin and thermally reactivated GAC." Carbon, 39(6), 789-807.

Munoz, E., Dalton, A. B., Collins, S., Zakhidov, A. A., Baughman, R. H., Zhou, W. L., He, J., O'Connor, C. J., McCarthy, B., and Blau, W. J. (2002). "Synthesis of SiC nanorods from sheets of single-walled carbon nanotubes." Chem. Phys. Lett., 359(5-6), 397-402.

Nandi, S. P., Ramadass, V., and Walker, P. L. (1964). "Changes in the ultrafine structure of anthracites upon heat treatment." Carbon, 2(3), 199-210.

Nayak, B. B., Mohanty, B. C., and Singh, S. K. (1996). "Synthesis of silicon carbide from rice husk in a dc arc plasma reactor." J. Am. Ceram. Soc., 79(5), 1197-1200.

Neill, D. A., Cannon, F. S., Voigt, R. C., Furness, J. C., and Bigge, R. (2001). "Effects of Advanced Oxidants on Green Sand System Performance in a Black Water System." Transactions of American Foundry Society, 109, 937-955.

Nowack, K. O., Cannon, F. S., and Mazyck, D. W. (2004). "Enhancing activated carbon adsorption of 2-methylisoborneol: Methane and steam treatments." Environmental Science & Technology, 38(1), 276-284.

Nsakala, N. Y., Essenhigh, R. H., and Walker, P. L. (1977). "Studies on Coal Reactivity: Kinetics of Lignite Pyrolysis in Nitrogen at 808 oC." Combust. Sci. Technol., 16, 153-163.

Nsakala, N. Y., Essenhigh, R. H., and Walker, P. L. (1978). "Characteristics of chars produced from lignites by pyrolysis at 808 oC following rapid heating." Fuel, 57(10), 605-611.

Oda, H., Takeuchi, M., and Yokokawa, C. (1981). "Effect of air-oxidation on the pore-structure of coals and cokes or chars obtained from oxidized coals." Fuel, 50(5), 390-396.

Ohan, M. P., Weadock, K. S., and Dunn, M. G. (2002). "Synergistic effects of glucose and ultraviolet irradiation on the physical properties of collagen." J. Biomed. Mater. Res., 60(3), 384-391.

145

Padhi, B. K., and Patnaik, C. (1995). "DEVELOPMENT OF SI2N2O, SI3N4 AND SIC CERAMIC MATERIALS USING RICE HUSK." Ceram. Int., 21(3), 213-220.

Park, B. T., Ryu, Y. W., and Yong, K. J. (2004). "Growth and characterization of silicon carbide nanowires." Surf. Rev. Lett., 11(4-5), 373-378.

Patel, M. (1989). "X-RAY PHOTOELECTRON SPECTROSCOPIC STUDIES ON SILICON-CARBIDE FROM RICE HUSK." X-Ray Spectrom., 18(5), 215-218.

Paul, S. A., Hull, A. S., Plancher, H., and Agarwal, P. K. (2002). "Use of asphalts for formcoke briquettes." Fuel Process. Technol., 76(3), 211-230.

Pieper, J. S., Hafmans, T., Veerkamp, J. H., and van Kuppevelt, T. H. (2000). "Development of tailor-made collagen-glycosaminoglycan matrices: EDC/NHS crosslinking, and ultrastructural aspects." Biomaterials, 21(6), 581-593.

Plancher, H., Agarwal, P. K., and Severns, R. (2002). "Improving form coke briquette strength." Fuel Process. Technol., 79(2), 83-92.

Puente, G., Fuente, E., and Pis, J. J. (2000). "Reactivity of Pyrolysis Chars Related to Precursor Coal Chemistry." J. Anal. Appl. Pyrolysis, 53(1), 81-93.

Radovic, L. R., Steczko, K., Walker, P. L., and Jenkins, R. G. (1985). "Combined Effects of Inorganic Constituents and Pyrolysis Conditions on the Gasification Reactivity of Coal Chars." Fuel Processing Technolog, 10(3), 311-326.

Radovic, L. R., Walker, P. L., and Jenkins, R. G. (1983). "Importance of carbon active sites in the gasification of coal chars." Fuel, 62(7), 849-856.

Rambo, C. R., Martinelli, J. R., and Bressiani, A. H. A. (1999). "Synthesis of SiC and cristobalite from rice husk by microwave heating." Adv. Powder Technol., 299-3, 63-69.

Ray, A. K., Mahanty, G., and Ghose, A. (1991). "EFFECT OF CATALYSTS AND TEMPERATURE ON SILICON-CARBIDE WHISKERS FORMATION FROM RICE HUSK." J. Mater. Sci. Lett., 10(4), 227-229.

Rehder, J. E. (1981). "Cupola fuel: a review." Foundry Management Technology, 109(2), 49-53.

Richards, S. R. (1990). "PHYSICAL TESTING OF FUEL BRIQUETTES." Fuel Process. Technol., 25(2), 89-100.

Rodriguez-Lugo, V., Rubio, E., Gomez, I., Torres-Martinez, L., and Castano, V. M. (2002). "Synthesis of silicon carbide from rice husk." International Journal of Environment and Pollution, 18(4), 378-387.

146

Sajithlal, G. B., Chithra, P., and Chandrakasan, G. (1998a). "Advanced glycation end products induce crosslinking of collagen in vitro." Biochimica Et Biophysica Acta-Molecular Basis Of Disease, 1407(3), 215-224.

Sajithlal, G. B., Chithra, P., and Chandrakasan, G. (1998b). "The role of metal-catalyzed oxidation in the formation of Advanced Glycation End products: An in vitro study on collagen." Free Radical Biol. Med., 25(3), 265-269.

Sajithlal, G. B., Chithra, P., and Chandrakasan, G. (1999). "An in vitro study on the role of metal catalyzed oxidation in glycation and crosslinking of collagen." Mol. Cell. Biochem., 194(1-2), 257-263.

Sakintuna, B., Yurum, Y., and Cetinkaya, S. (2004). "Evolution of Carbon Microstructures during the Pyrolysis of Turkish Elbistan Lignite in the Temperature Range 700-1000 oC." Energy & Fuels, 18(3), 883-888.

Sathe, C., Hayashi, J., and Li, C. (2002). "Release of Volatiles from the Pyrolysis of a Victorian Lignite at Elevated Pressures." Fuel, 81(9), 1171-1178.

Schinzel, W., ed. (1981). Wiley, New York.

Sell, D. R., Biemel, K. M., Reihl, O., Lederer, M. O., Strauch, C. M., and Monnier, V. M. (2005). "Glucosepane is a major protein cross-link of the senescent human extracellular matrix - Relationship with diabetes." J. Biol. Chem., 280(13), 12310-12315.

Shen, G. Z., Bando, Y., Ye, C. H., Liu, B. D., and Golberg, D. (2006). "Synthesis, characterization and field-emission properties of bamboo-like beta-SiC nanowires." Nanotechnology, 17(14), 3468-3472.

Shen, G. Z., Chen, D., Chen, P. C., and Zhou, C. W. (2009). "Vapor-Solid Growth of One-Dimensional Layer-Structured Gallium Sulfide Nanostructures." Acs Nano, 3(5), 1115-1120.

Shi, W. S., Zheng, Y. F., Peng, H. Y., Wang, N., Lee, C. S., and Lee, S. T. (2000). "Laser ablation synthesis and optical characterization of silicon carbide nanowires." J. Am. Ceram. Soc., 83(12), 3228-3230.

Simons, G. A. (1979). "The Structure of Coal Char: Part II. Pore Combination." Combust. Sci. Technol., 19(5-6), 227-235.

Simons, G. A. (1983). "The role of pore structure in coal pyrolysis and gasification." Prog. Energy Combust. Sci., 9(4), 269-290.

Simons, G. A., and Finson, M. L. (1979). "The structure of Coal Char: Part I. Pore Branching." Combust. Sci. Technol., 19(5-6), 217-225.

147

Singh, S. K., Mohanty, B. C., and Basu, S. (2002). "Synthesis of SiC from rice husk in a plasma reactor." Bull. Mater. Sci., 25(6), 561-563.

Singh, S. K., Parida, K. M., Mohanty, B. C., and Rao, S. B. (1995). "HIGH-SURFACE-AREA SILICON-CARBIDE FROM RICE HUSK - A SUPPORT MATERIAL FOR CATALYSTS." React. Kinet. Catal. Lett., 54(1), 29-34.

Singh, S. K., Stachowicz, L., Girshick, S. L., and Pfender, E. (1993). "THERMAL PLASMA SYNTHESIS OF SIC FROM RICE HULL (HUSK)." J. Mater. Sci. Lett., 12(9), 659-660.

Singla, P. K., Miura, S., Hudgins, R. R., and Silveston, P. L. (1983). "Pore development during carbonization of coals." Fuel, 62(6), 645-648.

Smoot, L. D., Eatough, S. R., and Miller, A. B. (2007). "Form coke reaction processes in carbon dioxide." Fuel, 86(17-18), 2645-2649.

Solomon, P. R., Fletcher, T. H., and Pugmire, R. J. (1992a). "Progress in Coal Pyrolysis." Fuel, 72(5), 587-597.

Solomon, P. R., Serio, M. A., and Suuberg, E. M. (1992b). "Coal Pyrolysis: Experiments, Kinetic Rates and Mechanisms." Prog. Energy Combust. Sci., 18(2), 133-220.

Strugala, A. (2002). "Changes of porosity during carbonization of bituminous coals: effects due to pores with radii less than 2500 nm." Fuel, 81(9), 1119-1130.

Stubbles, J. (2000). "Energy Use in the U.S. Steel Industry: A Historical Perspective and Future Opportunities." Department of Energy, Office of Industrial Technologies.

Suzuki, M., Misic, D. M., Koyama, O., and Kawazoe, K. (1978). "Study of thermal regeneration of spent activated carbons: thermogravimetric measurement of various single component organics loaded on activated carbons." Chem. Eng. Sci., 33(3), 271-279.

Takahashi, K., Yokouchi, M., Lee, S. K., and Ando, K. (2003). "Crack-healing behavior of Al2O3 toughened by SiC whiskers." J. Am. Ceram. Soc., 86(12), 2143-2147.

Tanzer, M. L. (1973). "Cross-Linking of Collagen." Science, 180(4086), 561-566.

Taylor, J. W., and Coban, A. (1987a). "Factors Affecting The Strength Of Formed-Coke Made From Lignite Char." Fuel, 66(9), 1274-1280.

Taylor, J. W., and Coban, A. (1987b). "Formed-Coke From Lignite, And The Critical Role Of Air." Fuel, 66(1), 141-142.

Toda, Y. (1973a). "A study by density measurement of changes in pore structures of coals with heat treatment. 1. Macropore structure." Fuel, 52(1), 36-40.

148

Toda, Y. (1973b). "A study by density measurement of changes in pore structures of coals with heat treatment. 2. Micropore structure." Fuel, 52(2), 99-104.

Toda, Y., Hatami, M., Toyoda, S., and Honda, H. (1970). "Fine structure of carbonized coals." Carbon, 8(4), 565-571.

Tomeczek, J., and Gil, S. (2003). "Volatiles Release and Porosity Evolution During High Pressure Coal Pyrolysis." Fuel, 82(4), 285-292.

Usha, R., and Ramasami, T. (2000). "Effect of crosslinking agents (basic chromium sulfate and formaldehyd) on the thermal and thermomechanical stability of rat tail tendon collagen fibre." Thermochim. Acta, 356, 59-66.

Verzijl, N., DeGroot, J., Ben Zaken, C., Braun-Benjamin, O., Maroudas, A., Bank, R. A., Mizrahi, J., Schalkwijk, C. G., Thorpe, S. R., Baynes, J. W., Bijlsma, J. W. J., Lafeber, F., and TeKoppele, J. M. (2002). "Crosslinking by advanced glycation end products increases the stiffness of the collagen network in human articular cartilage - A possible mechanism through which age is a risk factor for osteoarthritis." Arthritis Rheum., 46(1), 114-123.

Wagner, R. S., and Ellis, W. C. (1964). "Vapor-Liquid-Solid Mechanism Of Single Crystal Growth ( New Method Growth Catalysis From Impurity Whisker Epitaxial + Large Crystals Si E )." Appl. Phys. Lett., 4(5), 89-&.

Wang, C., Heine, R. W., Schumacher, J. S., and Green, R. A. (1973). "What's the story on seacoal replacements in foundry sands?" Transactions of American Foundry Society, 81, 496-502.

Wang, D., Qian, F., Yang, C., Zhong, Z. H., and Lieber, C. M. (2004). "Rational growth of branched and hyperbranched nanowire structures." Nano Lett., 4(5), 871-874.

Wang, D. H., Wang, D. Q., Hao, Y. J., Jin, G. Q., Guo, X. Y., and Tu, K. N. (2008). "Periodically twinned SiC nanowires." Nanotechnology, 19(21).

Wang, Y., Cannon, F. S., Salama, M., Goudzwaard, J., and Furness, J. C. (2007a). "Characterization of hydrocarbon emissions from green sand foundry core binders by analytical pyrolysis." Environmental Science & Technology, 41, 7922-7927.

Wang, Y. J., Cannon, F. S., Komarneni, S., and Voigt, R. C. (2005). "Mechanisms of advanced oxidation processing on bentonite consumption reduction in foundry." Environmental Science & Technology, 39(19), 7712-7718.

Wang, Y. J., Cannon, F. S., Salama, M., Fonseca, D. A., and Giese, S. (2009). "Characterization of Pyrolysis Products from a Biodiesel Phenolic Urethane Binder." Environmental Science & Technology, 43(5), 1559-1564.

149

Wang, Y. J., Cannon, F. S., Voigt, R. C., and Komarneni, S. (2006). "Effects of advanced oxidation on green sand properties via iron casting into green sand molds." Environmental Science & Technology, 40(9), 3095-3101.

Wang, Y. J., Huang, H., Cannon, F. S., Voigt, R. C., Komarneni, S., and Furness, J. C. (2007b). "Evaluation of volatile hydrocarbon emission characteristics of carbonaceous additives in green sand foundries." Environmental Science & Technology, 41(8), 2957-2963.

Wen, G. W., Li, F., Han, Z. X., and Bai, H. W. (2008). "Growth of beta-SiC nanowires from SiBONC nano powder compacts." Rare Metal Materials And Engineering, 37(3), 561-564.

Wong, E. W., Sheehan, P. E., and Lieber, C. M. (1997). "Nanobeam mechanics: Elasticity, strength, and toughness of nanorods and nanotubes." Science, 277(5334), 1971-1975.

World-Bank. (1998). "Coke Manufacturing." In: Pollution Prevention and Abatement Handbook, World Bank Publications, Washington DC, 286-290.

World-Commission-on-Environment-and-Development. (1987). Our Common Future, Oxford University Press, USA.

Xu, Y., Koga, Y., and Watkinson, A. P. (1994). "Pore size distribution of coals from western Canada." Fuel, 73(11), 1797-1801.

Yang, G. Z., Cui, H., Sun, Y., Gong, L., Chen, J., Jiang, D., and Wang, C. X. (2009). "Simple Catalyst-Free Method to the Synthesis of beta-SiC Nanowires and Their Field Emission Properties." Journal Of Physical Chemistry C, 113(36), 15969-15973.

Yang, L., Zhang, X., Huang, R., Zhang, G. Y., and Xue, C. S. (2006). "Formation of beta-SiC nanowires by annealing SiC films in hydrogen atmosphere." Physica E-Low-Dimensional Systems & Nanostructures, 35(1), 146-150.

Yang, W., Araki, H., Hu, Q. L., Ishikawa, N., Suzuki, H., and Noda, T. (2004). "In situ growth of SiC nanowires on RS-SiC substrate(s)." J. Cryst. Growth, 264(1-3), 278-283.

Yang, W., Araki, H., Thaveethavorn, S., Suzuki, H., and Noda, T. (2005). "In situ synthesis and characterization of pure SiC nanowires on silicon wafer." Appl. Surf. Sci., 241(1-2), 236-240.

Yao, X. M., Tan, S. H., Huang, Z. R., Dong, S. M., and Jiang, D. L. (2007). "Growth mechanism of beta-SiC nanowires in SiC reticulated porous ceramics." Ceram. Int., 33(6), 901-904.

150

Yip, K., Wu, H., and Zhang, D. (2007). "Effect of Inherent Moisture in Collie Coal during Pyrolysis Due to in-situ Steam Gasification." Energy & Fuels, 21(5), 2883-2891.

Yu, J., Lucas, J. A., and Wall, T. F. (2007). "Formation of the structure of chars during devolatilization of pulverized coal and its thermoproperties: A review." Prog. Energy Combust. Sci., 33(2), 135-170.

Zhang, J., Wang, Q. S., Wang, F., Chen, X. H., Lei, W. W., Cui, Q. L., and Zou, G. T. (2009a). "Plasma-assisted self-catalytic vapour-liquid-solid growth of beta-SiC nanowires." Journal Of Physics D-Applied Physics, 42(3).

Zhang, L. G., Yang, W. Y., Jin, H., Zheng, Z. H., Xie, Z. P., Miao, H. Z., and An, L. N. (2006). "Ultraviolet photoluminescence from 3C-SiC nanorods." Appl. Phys. Lett., 89(14).

Zhang, T. Y., Luo, M., and Chan, W. K. (2008). "Size-dependent surface stress, surface stiffness, and Young's modulus of hexagonal prism [111] beta-SiC nanowires." J. Appl. Phys., 103(10).

Zhang, X. H., Xu, L., Han, W. B., Weng, L., Han, J. C., and Du, S. Y. (2009b). "Microstructure and properties of silicon carbide whisker reinforced zirconium diboride ultra-high temperature ceramics." Solid State Sci., 11(1), 156-161.

Zhang, Y. F., Han, X. D., Zheng, K., Zhang, Z., Zhang, X. N., Fu, J. Y., Ji, Y., Hao, Y. J., Guo, X. Y., and Wang, Z. L. (2007). "Direct observation of super-plasticity of beta-SiC nanowires at low temperature." Adv. Funct. Mater., 17(17), 3435-3440.

Zhang, Y. J., Wang, N. L., He, R. R., Chen, X. H., and Zhu, J. (2001). "Synthesis of SiC nanorods using floating catalyst." Solid State Commun., 118(11), 595-598.

Zhou, W. M., Liu, X., and Zhang, Y. F. (2006). "Simple approach to beta-SiC nanowires: Synthesis, optical, and electrical properties." Appl. Phys. Lett., 89(22).

Zhou, X. T., Lai, H. L., Peng, H. Y., Au, F. C. K., Liao, L. S., Wang, N., Bello, I., Lee, C. S., and Lee, S. T. (2000a). "Thin beta-SiC nanorods and their field emission properties." Chem. Phys. Lett., 318(1-3), 58-62.

Zhou, X. T., Wang, N., Au, F. C. K., Lai, H. L., Peng, H. Y., Bello, I., Lee, C. S., and Lee, S. T. (2000b). "Growth and emission properties of beta-SiC nanorods." Materials Science And Engineering A-Structural Materials Properties Microstructure And Processing, 286(1), 119-124.

Zhou, X. T., Wang, N., Lai, H. L., Peng, H. Y., Bello, I., Wong, N. B., Lee, C. S., and Lee, S. T. (1999). "beta-SiC nanorods synthesized by hot filament chemical vapor deposition." Appl. Phys. Lett., 74(26), 3942-3944.

151

Zhu, S. M., Xi, H. A., Li, Q., and Wang, R. D. (2005). "In situ growth of beta-SiC nanowires in porous SiC ceramics." J. Am. Ceram. Soc., 88(9), 2619-2621.

Vitae

He Huang

EDUCATION 2010 Ph.D., Environmental Engineering, The Pennsylvania State University 2005 M.S., Environmental Engineering, University of Cincinnati 2003 M.S., Municipal Engineering, Tsinghua University 2001 B.S., Environmental Engineering, Tsinghua University PROFESSONAL REGISTRATION Engineer-in-Training, E.I.T. SELECTED PUBLICATIONS Huang, H., Y. J. Wang, and F. S. Cannon, 2009, Pore structure development of in-situ pyrolyzed carbon for pollution prevention in iron foundries: Fuel Processing Technology, v. 90, p. 1883-1891 Huang, H., and G. A. Sorial, 2007, Perchlorate remediation in aquatic systems by zero valent iron: Environmental Engineering Science, v. 24, p. 917-926. Wang, Y. J., H. Huang, F. S. Cannon, R. C. Voigt, S. Komarneni, and J. C. Furness, 2007, Evaluation of volatile hydrocarbon emission characteristics of carbonaceous additives in green sand foundries: Environmental Science & Technology, v. 41, p. 2957-2963. Huang, H., and G. A. Sorial, 2006, Statistical evaluation of an analytical IC method for the determination of trace level perchlorate: Chemosphere, v. 64, p. 1150-1156. Jiang, Z. P., H. Y. Wang, H. Huang, and C. C. Cao, 2004, Photocatalysis enhancement by electric field: TiO2 thin film for degradation of dye X-3B: Chemosphere, v. 56, p. 503-508. Huang, H., Z. P. Jiang, H. W. Yang, and R. H. Li, 2004, Synthesis of a kind of photocatalyst carried by magnetic particle by sol-gel method: Techniques and Equipment for Environmental Pollution Control, v. 5, p. 65-68. CONFERENCE PRESENTATIONS He Huang, Fred Cannon, Sridhar Komarneni, In-situ Growth of Silicon Carbide Nanowires from Anthracite Surfaces, International Carbon Conference 2010, Clemson, SC He Huang, Yujue Wang and Fred Cannon, Diminish VOC Emission in Iron Foundries by a Cost Effective In-situ Pyrolyzed Carbon, International Carbon Conference 2008, Nagano, Japan He Huang, Yujue Wang and Fred Cannon, Reducing VOC Emission from Foundries by In-situ Pyrolyzed Coal, International Carbon Conference 2007, Seattle, WA He Huang and George Sorial, Perchlorate Remediation by Zero Valent Iron, AIChE the 2005 Annual Meeting, Cincinnati, OH