High Temp Corrossion of Ceramic in Sulfuric Acid

A Thesis

Entitled

Evaluation of Lanthanum Phosphate Ceramic as Alternative Nozzle Material in Sulfuric

Acid Incineration

By

Jesse C. Wright

Submitted as partial fulfillment of the requirements for

The Master of Science in

Chemical Engineering

Advisor: Dr. Abdul-Majeed Azad

College of Graduate Studies

The University of Toledo

May 2008

The University of Toledo

College of Engineering

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY

SUPERVISION BY Jesse C. Wright

ENTITLED Evaluation of Lanthanum Phosphate Ceramic as Alternative Nozzle

Material in Sulfuric Acid Incineration

BE ACCEPTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE

DEGREE OF MASTER OF SCIENCE IN CHEMICAL ENGINEERING

Thesis Advisor: Dr. Abdul-Majeed Azad

Recommendation Concurred by

Dr. John P. Dismukes

Committee

on

Dr. Michael R. Cameron Final Examination

Dean, College of Engineering

iii

An Abstract of

Evaluation of Lanthanum Phosphate Ceramic as Alternative Nozzle Material in Sulfuric

Acid Incineration

Jesse C Wright

Submitted as partial fulfillment of the requirements for

The Master of Chemical Engineering

The University of Toledo

May 2008

Sulfuric acid is one of the largest produced chemicals in the world and a large

percentage is recycled through thermal decomposition. The nozzles that inject spent

sulfuric acid into a thermal decomposition furnace plague the process with high levels of

corrosion. The interface of high temperature gas and aqueous sulfuric acid creates a

unique mode of corrosion. A search for a nozzle material that is resistant to such rapid

deterioration and expensive replacement is therefore warranted. Ceramics being

somewhat superior to metals and alloys in terms of their corrosion resistance in an acidic

medium, lanthanum phosphate monazite was considered as a new alternative material to

compose the nozzles and thus increase the resistance to corrosive failure.

Lanthanum phosphate monazite was selected for this purpose because of its wide

range of applications in extreme and corrosive atmospheres. The material has been

reported to perform well as a ceramic for radioactive waste storage, ceramic matrix

composites in turbine applications, a machinable ceramic, and a good proton conductor

for sensors and fuel cells. Lanthanum phosphate monazite was thought to be a good

iv

candidate to compose the nozzles because of the similarities with its other applications

under conditions that a nozzle in spent acid treatment typically experiences. Testing was

done to determine if the material could function as a structural ceramic substitute to the

currently used alloys, at the operating temperatures and, if it would withstand the aqueous

mode of corrosion and the high temperature gaseous corrosion. During the material

corrosion analysis details of the corrosion process within the furnace were further

defined. The results were compared to other materials for analysis and recommendation.

The findings establish a clear set of directives for future research considering nozzle

corrosion, including the promising potential of using Al2O3-LaPO4 composite as the new

material.

v

Acknowledgements

I would like to thank my adviser, Dr. Abdul-Majeed Azad, for his support and

assistance. My circumstances of work and school required much patience and

understanding. Also, a special thanks to Sathees Kesavan and Desikan Sundararajan,

each generously donated their expertise and time to help me conduct my research.

I would also like to thank Jim Rapp and Neil Nofziger for sharing their research

and information that assisted me in determining the direction of my research. I would

also like to thank Chris Caspers of Federal Mogul. The equipment in his laboratory was

instrumental in defining the physical properties of the materials being studied. Chris

kindly assisted me in running many tests and procedures.

vi

Table of Contents

ABSTRACT.....................................................................................................................III

ACKNOWLEDGEMENTS ............................................................................................ V

TABLE OF CONTENTS ...............................................................................................VI

LIST OF TABLES .......................................................................................................... IX

LIST OF FIGURES ......................................................................................................... X

INTRODUCTION............................................................................................................. 1

1.1 SULFURIC ACID REGENERATION PROCESS ..................................................................... 1

1.2 INJECTION NOZZLES ....................................................................................................... 2

1.3 MOTIVATION.................................................................................................................. 4

1.4 RESEARCH FOCUS ......................................................................................................... 5

LITERATURE REVIEW ................................................................................................ 6

2.1 LANTHANUM PHOSPHATE MONAZITE ............................................................................ 6

2.1.1 COMBUSTION APPLICATIONS 6

2.1.2 MACHINABILITY CHARACTERISTICS 8

2.1.3 PHYSICAL PROPERTIES 10

2.1.4 STOICHIOMETRY 12

2.2 MICROWAVE SINTERING .............................................................................................. 15

EXPERIMENTAL.......................................................................................................... 17

3.1 EXPERIMENTAL ............................................................................................................ 17

3.1.1 PROCESS CONDITIONS 17

3.1.2 BOILING ACID 20

3.2 SYNTHESIS OF LAPO4 .................................................................................................. 21

3.2.1 POWDER SYNTHESIS 21

vii

3.2.2 SINTERING 25

3.3 CORROSION TEST ......................................................................................................... 26

3.3.1 BOILING ACID EXPERIMENT 26

3.3.1 ACID SAMPLES 28

RESULTS AND DISCUSSION ..................................................................................... 29

4.1 MATERIAL PROPERTIES OF LANTHANUM PHOSPHATE, LAPO4 .................................... 29

4.1.1 COEFFICIENT OF THERMAL EXPANSION 29

4.1.2 THERMAL AND THERMOGRAVIMETRIC ANALYSIS 31

4.1.3 MACHINABILITY 34

4.2 CORROSION TESTS ....................................................................................................... 35

4.2.1 FIELD TEST 35

4.2.2 CONSISTENCY OF BOILING ACID TEST PROCEDURE 37

4.2.3 HASTELLOY C22 BOIL TESTS 39

4.2.4 AL2O3 BOILING ACID TESTS 40

4.2.5 LAPO4 BOILING ACID TESTS 41

4.2.6 LIFETIME ESTIMATION 43

4.3 CORROSION MECHANICS.............................................................................................. 45

4.3.1 LAPO4 46

4.3.2 HASTELLOY C22 51

CONCLUSIONS AND RECOMMENDATIONS........................................................ 56

5.1 CONCLUSIONS .............................................................................................................. 56

5.2 RECOMMENDATIONS .................................................................................................... 58

WORKS CITED.............................................................................................................. 60

APPENDIX...................................................................................................................... 66

viii

A-1 HISTORICAL INDUSTRIAL CORROSION STUDIES .......................................................... 66

A-2 CORROSION DATA ...................................................................................................... 69

AISI 316 HIGH TEMPERATURE ACID 70

A-3 SEM/EDS................................................................................................................... 72

SAMPLES TO BE TESTED 72

A-4 RECOMMENDATIONS FOR FUTURE WORK ................................................................... 75

LANTHANUM PHOSPHATE 75

A-5 ALLOY COMPOSITION ................................................................................................. 77

ix

List of Tables

Table 1. Physical properties of lanthanum phosphate ..................................................... 10

Table 2. The average composition of the process gas in the furnace .............................. 18

Table 3. Typical contaminants in refinery grade sulfuric acid ......................................... 28

Table 4. Average post combustion furnace gas composition .......................................... 37

Table 5. Statistical analysis of variability of corrosion test ............................................. 38

Table 6. Boiling acid corrosion test comparison of Al2O3 and Hastelloy C22............... 43

Table 7. Specific details of boiling acid corrosion tests on Hastelloy C22 ..................... 69

Table 8. Specific details of boiling acid corrosion tests on LaPO4.................................. 69

Table 9. Specific details of boiling acid corrosion tests on Al2O3................................... 69

Table 10. Specific details of high temperature field tests of materials in the furnace..... 70

Table 11. Corrosion information on the AISI 316 immersed in 99 wt. % H2SO4 ........... 71

Table 12. Details of coupons SEM/EDS analyzed .......................................................... 72

Table 13. Industrial utilized alloy compositions for nozzles ........................................... 77

x

List of Figures

Figure 1. Typical spent gun design [14]. ........................................................................... 2

Figure 2. Spent acid injection gun orientation into the furnace. ........................................ 3

Figure 3. Crystal structure of monazite LaPO4 [47]. ....................................................... 11

Figure 4. La2O3 - P2O5 phase diagram. ............................................................................ 13

Figure 5. Vaporization losses of P2O5 from La(O3)3 and LaP5O14. ................................. 14

Figure 6. Location of nozzle corrosion. ........................................................................... 19

Figure 7. Nozzle after an in-service failure. .................................................................... 19

Figure 8. Diagram of combustion furnace used to simulate process. .............................. 20

Figure 9. Firing schedule for the LaPO4 coupons............................................................ 22

Figure 10. XRD of calcined LaPO4 powder. ................................................................... 23

Figure 11. Standard XRD pattern of monazite LaPO4 [71]. ............................................ 23

Figure 12. EDS signature of the commercial LaPO4 calcined at 700°C.......................... 24

Figure 13. EDS signature of LaPO4 synthesized in this work. ........................................ 24

Figure 14. PVA burnout schedule for our LaPO4 coupons.............................................. 25

Figure 15. Illustration of boiling acid corrosion apparatus.............................................. 26

Figure 16. Dependence of boiling point of sulfuric acid on its strength [7].................... 27

Figure 17. CTE of LaPO4................................................................................................. 30

Figure 18. Comparative TGA/DSC plots for the LaPO4 samples. .................................. 31

Figure 19. TGA/DSC analysis with three consecutive heat cycles. ................................ 32

Figure 20. Corrosion results for Al2O3, LaPO4 and Hastelloy C22 in the furnace. ......... 36

Figure 21. Consistency of the corrosion rate for Hastelloy C22 in 99 wt. % H2SO4. .... 38

xi

Figure 22. Hastelloy C22 corrosion rates as a function of acid strength. ........................ 40

Figure 23. Results of Al2O3 corrosion in 90 and 99 wt. % refinery grade sulfuric acid.. 41

Figure 24. Time for complete dissolution of LaPO4 coupons. ........................................ 42

Figure 25. Estimation of nozzle failure time for Hastelloy C22...................................... 44

Figure 26. Comparison of nozzle failure time for alumina and Hastelloy C22.............. 45

Figure 27. LaPO4 EDS scan broad view 1000 µm [61]................................................... 47

Figure 28. LaPO4 EDS scan grain boundary zoom view 100 µm [61]............................ 47

Figure 29. EDS scan of LaPO4 [61]................................................................................. 48

Figure 30. SEM/EDS analysis of as-received Hastelloy C22 [61]. ................................. 53

Figure 31. EDS Scans of Hastelloy C22 corrosion tested in 90 wt. % sulfuric acid. ...... 54

Figure 32. A 36 mild steel tested in room temperature spent sulfuric acid. .................... 67

Figure 33. AISI 304 tested in room temperature spent sulfuric acid. .............................. 67

Figure 34. AISI 316 tested in room temperature spent sulfuric acid. .............................. 68

Figure 35. Corrosion rate of AISI 316 in high temperature 99 wt. % H2SO4.................. 71

Figure 36. Element composition of the alloys considered............................................... 77

1

Chapter 1

Introduction

1.1 Sulfuric Acid Regeneration Process

Sulfuric acid is one of the largest produced chemicals by weight in the world.

Sulfuric acid is used in nearly every industry, including, agro-products, mining, paper and

pulp, steel production, water treatment, surfactants and soaps, plastics, petrochemicals

and refining. Its uses are so broad and diverse that a country’s production of sulfuric acid

was once used as an economic indicator [1, 2]. The sulfuric acid regeneration process

considered in this study is sulfuric acid from alkylation units on refineries.

The sulfuric acid is used as a liquid catalyst to convert olefins to alkylate in the

alkylation process. Alkylation uses primarily the C3 – C5 olefins from the fluid catalytic

cracker unit, to produce a valuable product, alkylate, a blending component used in the

gasoline pool, whose the percentage in the pool varies with season and the region.

Alkylation plants using sulfuric acid often consume 0.3 - 0.6 pounds acid per gallon of

alkylate, and produce more than 500,000 barrels alkylate per day in the United States

alone. Therefore, about 10 million pounds of sulfuric acid are consumed each day in the

refineries [2-13]. This acid must be regenerated for repeated use.

2

The acid is regenerated through a thermal decomposition process. In this process

the acid is injected into the furnace and heated to 980°C, where it is decomposed. The

resulting gas stream is recycled through a modified contact process. The process,

however, is quite corrosive and warrants development and use of corrosion-resistant

materials for injection nozzles as well as their periodic maintenance.

1.2 Injection Nozzles

In a typical decomposition process, the nozzle atomizes the spent acid and sprays

the atomized fluid into a furnace operating in the range of 980-1260°C. A properly

designed furnace and gun configuration recesses the gun in the refractory and cools the

gun with atomizing air. This keeps the gun temperature below about 540°C.

Figure 1. Typical spent gun design [14].

The spent acid guns inject the acid in a radial pattern around the centerline of the

furnace near the front where the burner is located, as shown in Figure 2 below. It is

desired to keep the spray zone of each gun from colliding with the flame which results in

flame quenching causing turbulence and soot formation. The turbulence within the

3

furnace can also create moving hot zones. If a hot zone occurs near a spent gun port, it

can back spray acid onto the gun and overheat it, causing rapid nozzle failure.

Figure 2. Spent acid injection gun orientation into the furnace.

Once the acid is atomized within the furnace, it is rapidly heated to the

decomposition temperature. This is a highly endothermic reaction that requires large

amounts of fuel for heating. The reaction results in sulfur dioxide gas as a desired

product for the regeneration process. The reaction occurring in the furnace is described

below.

H2SO4 (l) � SO2 (g) + H2O (g) + ½ O2 (g) ∆H˚Decomposition +202 kJ/ mol

Currently, the material of choice for the nozzles is Hastelloy C22. This alloy has

shown great resistance to corrosion in the manufacturing of sulfuric acid compared to the

traditional stainless steels such as 304, 316 and alloy 20. Hastelloy C22 is more robust

for use in high corrosion areas. Other acid regeneration processes in the esterfication

industry have reported using Hastelloy B2 (a nickel-molybdenum alloy) with similar

results [15]. Alloy compositions are listed in appendix A-5 for comparison [16-19].

4

1.3 Motivation

The required energy to heat and decompose the spent sulfuric acid is the main

cost associated with the regenerative operations. Great effort is put into optimizing the

process to reduce the amount and cost of fuel required for this purpose. A major fraction

of the fuel is consumed in the atomization of the spent acid in the furnace. Specially

designed guns and injection systems have received considerable attention and are covered

in multiple patents [14, 15, 20-22]. These designs, however, offer improvements only if

the nozzles are in working order; corrosion of the nozzles greatly limits the effectiveness

of these systems. Operators must constantly monitor each gun’s performance and

quickly decommission or change poor performing guns to minimize adverse effects on

the process. Considering an average gun lifetime of 3 months and 10 – 20 guns per

furnace, frequent weekly nozzle failures are observed.

In addition to the monetary losses due to frequent injection nozzle failure due to

wasted fuel, furnace refractory damages also occur; inadequately atomized acid collects

as pool on the refractory. The pooled acid slowly evaporates and burns off causing

advanced corrosion to the refractory, whose replacement is expensive.

Enhancement of corrosion resistance for sulfuric acid injection nozzles would

greatly improve the regeneration process. Long-lasting nozzles would minimize

refractory damage and reduce operating costs through fuel savings, thus adding overall

reliability to the process.

5

1.4 Research Focus

The currently used injection nozzles materials do not adequately withstand the

corrosion of the sulfuric acid regeneration process resulting in frequent and costly nozzle

failures. The development of an improved material for the nozzles could reduce the

impact nozzles have on the process efficiency. With this problem at hand in sulfuric acid

regeneration plants, the present work pertains to the search for and development of

alternative somewhat inert and more robust ceramic materials suitable for the fabrication

of atomizing nozzle. Lanthanum phosphate monazite was considered such a possible

material based on its excellent physical, thermal and chemical integrity under conditions

generally encountered in the acid regeneration furnaces. The research was focused on the

aspects of material degradation as a corrosion and process specific parameter.

6

Chapter 2

Literature Review

2.1 Lanthanum Phosphate Monazite

Due to its remarkable physical and chemical properties, lanthanum phosphate

monazite has been considered a potential material for spent nuclear fuel storage, and

corrosion resistant structural material in turbines and fuel cells, due to its machinability.

Lanthanum and other rare earth oxides are readily available and are quite

inexpensive[23]. Among various rare-earth compounds, lanthanum phosphate (LaPO4)

has been proven to be the most appropriate due to its high melting point, chemical

stability, and inertness towards reaction with combustion gases at high service

temperatures [24, 25]. Lanthanum phosphate also interacts benignly with traditional

ceramics in a manner that improves the system characteristics exhibiting favorable

material properties[24, 26-34].

2.1.1 Combustion Applications

In addition to improving the mechanical properties, lanthanum phosphate

monazite offers refractory properties as well, that are useful in extreme environments.

Applications such as coatings for aircraft engine turbines or heat-resistant tiling for space

vehicles experience temperature extremes. Current materials involved in these

7

applications utilize a composite of lanthanum phosphate with other ceramics with the

former being major constituent [35]. Selection of material for developing nozzles for

sulfuric acid regeneration also requires the ability to withstand high temperatures.

Lanthanum phosphate monazite is an attractive alternative oxidation-resistant

candidate for constituting a ceramic matrix composite (CMCs) in the intermediate and

high temperature regime of turbines. It performs better than the traditional carbon or

boron nitride fiber matrix interfaces, due to its refractory nature and superior thermo-

mechanically stability than that of alumina or mullite. Crack deflection, fiber pushout

and mechanical tests data suggest that lanthanum phosphate bonds weakly with other

oxides [36]. These properties with the combination of cleavage, twinning and dislocation

glide make the material exhibit significant plasticity, thus allowing the material to be

machinable. In addition to excellent high temperature oxidation resistance and increased

toughness, lanthanum phosphate exhibits good corrosion resistance in environments

containing sulfur and vanadium salts [23], which are present in carbon based fuels. The

salts are formed during fuel utilization many of which occur in molten state, and become

entrained in the gas stream. The molten salts impinge and condense on the turbine blades

thereby causing localized corrosion. Such corrosive impurities also exist in spent sulfuric

acid and are the likely source of nozzle corrosion. A material with increased resistance to

the sulfur and vanadium salts would add another layer of protection to the atomizing

nozzle.

Rare-earth phosphates (collectively known as monazites) are stable and

compatible with alumina up to1750°C in air and melt congruently in the vicinity of

2000°C. Lanthanum phosphate is sufficiently weak to allow for crack deflection and

8

debonding at the fiber/matrix interface; its coefficient of thermal expansion (CTE) is

9.6x10-6 deg

-1 which is ~25% higher than that of alumina [36]. A combination of all the

above-stated favorable properties makes lanthanum phosphate a good candidate material

for nozzles in sulfuric acid regeneration furnace.

2.1.2 Machinability Characteristics

The essential requirement for a ceramic to be considered for nozzle fabrication is

its machinability. The guns are dropped, slammed into the burner port and adjusted with

pipe wrenches and hammers during operations. A brittle ceramic would never last the

rigors a nozzles experiences. A machinable ceramic therefore, would offer better

resistance to the harsh handling of spent acid guns. Also, nozzles tend to plug from debris

in the spent acid; other failures of the gun parts require quick field maintenance using

some of the above mentioned heavy-duty tools. Specialty ceramics and advanced

coatings are generally not recommended in the industry; a nozzle made out of machinable

ceramic having properties like a metal nozzles would be preferred. In the light of this

and the above-mentioned factors, a nozzle made out of a machinable ceramic would

increase the likelihood of implementation of new nozzle made out of lanthanum

phosphate.

Machining is an inevitable requirement for the flexible use of advanced ceramics,

especially structural ceramics. However, the extremely high hardness makes conventional

machining very difficult or even impossible in the case of ceramics. Generally, two

methods are used for improving the machinability of ceramic materials. One is the

structural designing, where machinability of ceramics is optimized by adjusting the phase

distribution, porosity and microstructure [37]. The second method is by introducing a

9

weak interface phase or layered- structure material in the matrix, which facilitates crack

deflection during machining, such as mica containing glass-ceramics [34].

Early attempts to make lanthanum phosphate machinable involved its composite

with Al2O3. Fracture energy measurements have shown that Al2O3/LaPO4 interface is

weak enough (compared to Al2O3/Al2O3 interface) to satisfy the He and Hutchison

debonding criteria for bi-material interfaces [30]. There were also multiple dislocations,

dislocation networks and subgrain boundaries inside the Al2O3 grains and between

Al2O3/LaPO4 grains. This type of structural development is responsible for active plastic

deformation that occurs within the Al2O3 grains through grain dislocation. Due to the

weak bonding of Al2O3 and LaPO4 and soft layered phase in LaPO4, this composite can

be easily machined with cemented carbide drills instead of conventional diamond tools

[37].

Davis et al. [24] demonstrated that the reason for LaPO4 to exhibit excellent

machinability is associated with the deformation bands observed within its individual

grain, which is similar to that of dislocations, twinning and martenistic transformation of

other ceramic materials [37]. Since, LaPO4 grains are weak, fracture mode is mainly

transgranular. Multiple dislocations and dislocations networks occur inside the LaPO4

grains. This suggests active plastic deformation via grain dislocation [37]. As stated

above, LaPO4 possesses a layered crystal structure, which can be readily delaminated due

to its low cleavage energy, leading to fracture propagation parallel to the layer plane.

Thus, crack deflections, branching and blunting during machining of layered LaPO4 help

to prevent macroscopic fractures from propagation beyond the local cutting area [34],

10

leading to the belief that it would function well as a machinable ceramic nozzle in the

sulfuric acid regeneration plant.

2.1.3 Physical Properties

The reported physical properties of LaPO4 were thoroughly reviewed to ensure

that the material could comply with the requirement of the intended application, viz., the

sulfuric acid regeneration plant. Some consideration for developing a nozzle from

ceramic materials must be taken into account, such as: (1) high melting point, (2) absence

of phase transformation up to the operating temperature, (3) low thermal conductivity, (4)

chemical inertness, (5) comparable thermal expansion similarity with metallic nozzle, (6)

adequate toughness to allow machining and, (7) high strength to withstand operating

pressures and flows [23, 38-40]. Typical values of some of the physical properties of

lanthanum phosphate are shown in Table 1.

Table 1. Physical properties of lanthanum phosphate [23, 41]

Property Value

Melt point (°C) 2070

Thermal conductivity (W m-1 K

-1) 1.8

Coefficient of Thermal Expansion (K-1) 10.5 x 10

-6

Young’s Modulus(GPa) 133

Poisson’s ratio 0.28

Heat Capacity @ 25°C (J K-1 mol

-1) 101.28

Lanthanum phosphate crystallizes in two modifications: the rhabdophane-type

hexagonal LaPO4*0.5H2O and the monazite-type monoclinic LaPO4, the later being

achieved either through ball milling the LaPO4*0.5H2O powder for extended period or

calcining at temperatures above 600°C [42].

11

LaPO4 (along with ScPO4 and YPO4) is a member of mineral group ABO4, which

includes silicates such as zircon (ZrSiO4), thorite (ThSiO4, tetragonal), huttonite (ThSiO4,

monoclinic), and coffinite (USiO4). Depending on the size of A-site cation, the ABO4

minerals crystallize in either the lanthanum phosphate monazite or zircon type motifs

[43-46].

LaPO4 crystallizes in a P21/n symmetry with four formula units in the unit cell.

Phosphorous is four-fold coordinated with oxygen in a distorted tetrahedral environment.

Lanthanum is nine-fold coordinated with oxygen in an unusual arrangement: five-fold

coordination with oxygen along a plane perpendicular to the c axis and four-fold

coordination with oxygen along the plane parallel to c (two oxygen atoms above and two

below the ab plane). It is geometrically described as a framework of LaO9 polyhedra and

PO4 tetrahedra alternating in 3-D. These fundamental blocks interpenetrate each other

and share oxygen atoms at the corner [27, 43-46].

Figure 3. Crystal structure of monazite LaPO4 [47].

LaPO4 like other ceramics has a direct relationship with density and corrosion

resistance. Increasing the density of a ceramic decrease pores and voids within the

ceramic body, decreasing the overall surface area. Therefore, LaPO4 corrosion resistance

12

depends on the sintering conditions, as this control ceramic microstructure and density.

It has been found that increasing the sintering temperature from 900 to 1300°C resulted

in a decrease in pore volume fraction and increase in pore diameter from ~8 to ~25 nm.

Extended soaking at 1200°C transforms the defective, porous, and poorly crystallized

grains into small (40 nm diameter), well-crystallized ones with low level of porosity;

soaking for 100h at 1200°C caused grains to grow to ~100 nm in diameter with further

reduction in porosity [48]. Density of LaPO4 and its composites decrease with increase

in soak time. This is due to the concomitant increase in the grain size upon long-term

sintering [49].

2.1.4 Stoichiometry

To achieve optimum properties, LaPO4 must be formed with a 1:1 ratio of La: P.

Small deviations from stoichiometry change its solidus temperature from 2070°C to

1580°C on the La-rich side and to P-rich side at 1050°C [37, 50]. LaP3O9 decomposes to

LaPO4 with loss of P2O5 above ~875°C and becomes amorphous in the range of 1050 –

1225°C (the reported melting point of LaP3O9 is 1050°C). Therefore, without

appropriate phosphorous removal, samples sintered near or above the melting point of

LaP3O9, will be phosphate-rich amorphous phases, and may escape detection by XRD.

This phosphate could exist at grain boundaries of LaPO4 [51, 52].

The residual secondary phase hydrogen-phosphate and corresponding lanthanum

trimetaphosphate may impact the LaPO4 sintering in two ways. First, it decreases the

sintering temperature. This parameter could also be taken into account to explain better

sinterability of powders containing trimetaphosphate. Second, it promotes grain growth,

thereby decrease in surface area. Powders containing trimetaphosphate have greater

13

initial surface area (SA ~ 32 m2g-1) than pure LaPO4 (SA ~ 17.5 m

2g-1). The grain

growth takes place with rapid grain boundary migration whereby the grain boundaries get

separated from the pores. To achieve high densification, this phenomenon must be

suppressed for it leads to intragranular pores inhibiting sintering. This detrimental effect

appears particularly important when trimetaphosphate present to the extent of 3.1%, only

78% of the theoretical density is reached [52].

Figure 4. La2O3 - P2O5 phase diagram [53].

14

Figure 5. Vaporization losses of P2O5 from La(PO3)3 and LaP5O14 [53].

Considering La2O3-P2O5 phase diagrams, La(PO3)3 decomposes into LaPO4 with

incongruent melting in the range 1050-1235°C. This produces a phosphorous oxide rich

liquid and P4O10 gas. The presences of the secondary liquid phase may explain the

decrease of sintering temperature but not without grain growth [52, 53]. The off-gassing

of phosphorous pentoxide dimer threatens to damage sintering equipment during

processing. In addition, many different forms of lanthanum (LaPO4*0.5H2O, LaP3O9,

La2O3, La(PO3)3, and La(OH)3) existing greatly alter the ceramics properties. This will

affect the materials performance in high temperatures and will compromise its integrity.

15

2.2 Microwave Sintering

The LaPO4 samples synthesized and employed in this study were sintered using a

microwave, with the aim to produce a dense nonporous ceramic body to enhance the

corrosion resistance. The benefits of microwave energy in thermally activated processes

stem from the specificity of energy absorption. Contrary to other conventional methods,

microwave energy is unique and transforms into heat within the material thus altering the

heating properties and greatly reducing process times [54-56].

Microwave processing also maximizes the mechanical properties of the ceramic

that would not be fully attained by traditional methods. It enables the ceramic to obtain

higher degree of densification and improve the intergranular structure and makes the

ceramic less susceptible to acid penetration and grain boundary dissolution. Both these

attributes contribute to improved resistance to corrosion and nozzle performance.

In conventional furnace heating, energy is transferred to the materials by thermal

electromagnetic radiation, with its maximum intensity being in the infrared range. The

penetration depth of infrared radiation is very small in a majority of solids. Therefore,

energy deposition is localized within a thin layer near the surface of materials. As a

result conventional heating essentially depends on heat transfer from the hotter near-

surface region to the colder bulk of the material [54-56].

Microwave heating on the other hand is based upon the capacity of a material to

absorb the electromagnetic energy. The absorption in this instance is assisted by

employing the hybrid heating scheme. This involves introduction of additional heat

sources into the system. These sources are microwave-absorbing objects (susceptors),

from which the microwave-generated heat is transferred to the low-loss material

16

undergoing processing. The heat is deposited directly inside the sample, and the heating

rate is limited only by the power of the microwave source. The heat is dissipated to the

environment through the surface. Therefore, the temperature inside the sample is always

higher than on the surface. This microwave-specific temperature distribution is known as

an inverse temperature profile [54-56].

At elevated temperatures, microwave absorption in most materials grows sharply,

primarily due to an onset of another absorption mechanism. This is characteristics for

both ionic (Al2O3, ZrO2) and covalent (Si3N4, AlN) solids. A sharp increase in the

microwave loss begins at temperatures of about 0.4-0.5 Tm (where Tm is the melting

temperature of the material). In this range the bonds between ions in ionic crystals start

to break, and the electrons in covalent materials begin to populate the conduction bands.

A correctly designed susceptor system will provide heating at the initial stage, and reflect

most of the power at high temperatures when microwaves are absorbed directly by the

material undergoing processing [54-56].

17

Chapter 3

Experimental

3.1 Experimental

3.1.1 Process Conditions

A number of simulated tests were conducted under conditions close to those

experienced by the sulfuric acid regeneration nozzles during service. The observations

were used to determine how the spent acid gun interacted with the incineration process

within the furnace. The guns were observed through downstream view ports, a video

camera and empty gun ports. These observations established that if the spent acid gun

were installed as designed, each gun would experience similar conditions and a standard

set of conditions were experienced by each spent acid gun.

The employed variables were temperature, gas composition and flow

characteristics. The operating temperature was determined by type-K thermocouple

inserted alongside a spent gun. The thermocouple was placed on the nozzle, but did not

18

come in contact with the acid stream. The nozzle operated reasonably in the temperature

range of 600 –1000°C, preferably at 700°C. The gas composition was determined

through inline process sensors and the wet chemistry method known as the Orsat 1.

Table 2. The average composition of the process gas in the furnace

N2 H2O SO2 CO2 O2 SO3

55% 25% 10% 8% 2% <1%

Over fifty failed nozzles were collected from two different sites and analyzed;

each failure was found to be the result of localized corrosion on the bottom rim of the

nozzle. This confirmed that each gun was exposed to similar conditions and that a unique

mechanism was responsible for the failure. The corrosion led to a small hole at the

bottom rim of the nozzle which grew in size with time. The hole bypassed spent acid

around the atomizing chamber and injected spent acid as a stream into the furnace.

1 Orsat - is for analysis of process gas samples to determine CO2, O2 and sometimes CO concentrations

from fossil fuel emission sources. A liquid filled leveling bottle moves the integrated sample through a

graduated glass burette and absorption pipettes containing absorbing reagents. The resulting changes in

volume measure percent of O2 , CO2 and CO. The burette is water jacketed for temperature stability 57.

Apex Instruments, I. Method 3 Orsat Analyzer. 2002 [cited 2008 1-10-2008]; Available from:

http://www.apexinst.com/products/orsat.htm

19

Figure 6. Location (highlighted in grey) of nozzle corrosion.

Figure 7. Nozzle after an in-service failure.

20

The spent acid pooled at the bottom rim of the nozzle was the source of the

localized corrosion. This was verified through visual observation of the operating guns.

The turbulence in the furnace due to high flow rate caused the spent acid to pool at the

bottom of the nozzle in a cyclic pattern, alternating between acid gain and acid loss.

During acid gain the furnace flow would redirect atomized acid back towards the nozzle,

causing it to collect on the bottom rim. A slight change in the furnace conditions would

not allow the atomized stream to collect on the nozzle. The static pool would therefore

boil off.

3.1.2 Boiling Acid

For the evaluation of how the cyclic pooling of acid affected the corrosion

process, a bench-top combustion furnace (Figure 6) was employed. 99 wt. % acid in a

small crucible was placed in the furnace and heated to 650°C until the acid started to boil.

The temperature was then lowered to 500°C and held steady for 30 min.

Figure 8. Diagram of combustion furnace used to simulate process.

21

This heating interval increased the volume of the acid and decreased the acid

strength. The acid strength decreased to < 90 wt. %, indicating that the hygroscopic acid

absorbed the water produced from combustion and diluted the acid pooled collected at

the bottom of the nozzle. This ameliorated the corrosion rate significantly as the weaker

acid was found to be much more corrosive.

3.2 Synthesis of LaPO4

3.2.1 Powder Synthesis

The lanthanum phosphate (LaPO4) powders were synthesized by reacting

phosphoric acid and lanthanum oxide (La2O3) in a continuously stirred de-ionized water

bath. Phosphoric acid was reagent grade 85% orthophosphoric acid JTBaker

(Phillipsburg, New Jersey) and La2O3 (99.99% purity) was purchased from Molycorp

(Englewood, Colorado). For the synthesis, phosphoric acid was first diluted with

distilled water bath, followed by the slow addition of La2O3 to the aqueous acid. The

reactants were added in the proportion of La to P molar ratio of 1:1 in order to yield

LaPO4 as the final desired product per the following scheme:

2La2O3 + 3H3PO4 � 2LaPO4*0.5H2O + LaPO4 + La(OH)3 + 2H2O

The hydrated form of LaPO4 must be calcined to remove the attached water.

Once the water is removed through dehydration LaPO4 exists in the monazite phase.

This research illustrates that the initial calcination removes the majority of the water, but

minute amounts of water are reabsorbed during cool downs. Also, it was found that the

22

La(OH)3 residual bi-product is removed during the initial sintering and the resulting

weight loss does not reoccur. These findings are discussed further in chapter 4 Figure 17.

The undesired impurity (La(OH)3) must be removed through further processing.

As La2O3 is added to the diluted phosphoric acid, the reaction proceeded

immediately via precipitate formation. The solution is constantly stirred during the

reaction and continued for 24 additional hours to ensure completion of the reaction. The

precipitate was allowed to settle, washed several times with de-ionized water until the pH

of the decanted water was close to 7. The hydrated precipitate was calcined in air using

the following heating schedule

Figure 9. Firing schedule for the LaPO4 coupons.

This firing schedule helped in water removal from hydrated lanthanum phosphate

and its transformation into the monazite form. The X-ray diffraction peaks in the

calcined samples (Figure 8) conformed to those of LaPO4 (JCPDS#320493; monazite

structure) as shown in Figure 9.

23

Position [°2Theta]

20 30 40 50

Counts

0

100

200

300

400

LP700

Figure 10. XRD of calcined (700°°°°C/4h) LaPO4 powder.

Figure 11. Standard XRD pattern of monazite LaPO4 [71]

24

Figures 10 and 11 show the EDS signatures in the commercial monazite from Alfa Aesar

and that synthesized in this work, both caclined at 700°C. As can be seen, the two EDS

scans are very similar.

Figure 12. EDS signature of the commercial LaPO4 calcined at 700°C.

Figure 13. EDS signature of LaPO4 synthesized in this work.

25

3.2.2 Sintering

The calcined lanthanum phosphate powders were ball-milled for 24h as aqueous

slurry using zirconia spheres as the milling media, dried in a vacuum oven and mixed

with 2 wt.% of polyvinyl alcohol (PVA). The powder was uniaxially pressed into pellets

(2cm x 3.5cm) under a pressure of 240 kPa, followed by cold isostatic pressing (CIP)

using a force of 71.2 kN.

The CIPed samples were first fired up to 250°C in a traditional electric furnace to

burn out the organic binder (PVA) prior to microwave (MW) sintering. This is a required

step for the microwave unit because the organic vapors could cause arcing and damage

the unit. A slow ramp rate was used during the binder burnout, as faster rates resulted in

cracking of the pellets. The heating schedule was as follows:

Figure 14. PVA burnout schedule for our LaPO4 coupons.

The samples were sintered using a ThermWave Model III MW Power Unit

microwave furnace Ceralink Inc. (Troy, New York). It operates at 2.45 GHz with two 50

gram silicon carbide susceptors. The samples were heated to 1575°C in about 45

26

minutes, held for 15 min., and allowed to cool to room temperature in the insulated box

[88]. While the unsintered samples were granular and porous, the sintered ones became

semi-glossy and quite dense.

3.3 Corrosion Test

3.3.1 Boiling Acid Experiment

As explained above, the location of nozzle failure was at the bottom where acid

would pool and boil off. Based on this failure mechanism, it was desired to reproduce the

conditions so that alternative materials could be tested prior to fabricating actual nozzles

from these compositions. The boiler apparatus was a 250 mL boiling flask with a water-

cooled reflux condenser attached. The candidate nozzle material sample was placed in

the flask and the acid was added to the flask, submerging the sample. A heating element

was then used to bring the acid to a boil.

Figure 15. Illustration of boiling acid corrosion apparatus.

The factors considered in the boiling unit were: the acid strength (99 and 90 wt.

%), materials (LaPO4, Hastelloy C22, & Al2O3), and the period of time the sample was

27

subjected to boiling acid tests. The acid used in the boiling tests was refinery grade. 60

ml of the H2SO4 acid was used in all the tests; in the case of tests with acid of 90%

strength, the 99% acid was diluted with adequate amount of water. This was done to

standardize the amount of acid present in the solution regardless of strength.

All tests were conducted using a standardized protocol. The flask was cleaned

and dried before each test. The sample was placed in the flask and the acid was added.

The flask was then attached to a condenser tube which was then connect to cooling water

circulation. Finally, the heater was turned on; it took about 30 minutes to boil.

Figure 16. Dependence of boiling point of sulfuric acid on its strength [7].

The materials in the boiling apparatus are subjected to high temperatures based on

the boiling point of concentrated sulfuric acid. As seen from Fig 18, the 90 and 99 wt. %

sulfuric acid boil at about 260°C and 280°C, respectively. Though the boiling apparatus

operated at a constant acid strength, the nozzle however, experienced fluctuating strength

28

during operation. This could expose the nozzle to temperatures in excess of 330°C,

which in turn could influence the corrosion rate significantly.

3.3.1 Acid Samples

The acid used for corrosion tests was taken at random in a 3 month span from a

sulfuric acid regeneration facility. Each aliquot was 350 ml and 99.25 (± 0.20) wt. %

taken from the same source point when sampling large shipments for refiners. In all, 15

samples were used. All the samples were stored individually and used at random. The

nature and concentration of contaminants in a typical fresh refinery grade sulfuric acid is

shown in Table 6.

Table 3. Typical contaminants in refinery grade sulfuric acid

Species SO2 Fe Nitric Cr Pb As Cd Hg Calcined Residue

mg/kg 50 20 10 1.0 1.0 0.50 0.05 0.05 100

Spent sulfuric acid was not used because it contains 4 – 7 wt. % hydrocarbon

carryovers from the alkylation unit, which tends to boil at around 60°C, foaming out of

the reflux unit and forming pumice-like solid substances [58]. This would inhibit the

simulation of the actual process occurring in the furnace (viz., a pool of acid collecting at

the tip of the nozzle and boiling without the formation of solid residue). It was assumed

that the atomization of the spent acid facilitated in the removal of most of the

hydrocarbon.

29

Chapter 4

Results and Discussion

4.1 Material Properties of Lanthanum Phosphate, LaPO4

The material properties of LaPO4 were evaluated to determine if the material was

compatible with the service conditions of a nozzle. Measurement of the coefficient of

thermal expansion (CTE) and simultaneous TGA/DSC analysis were used to determine if

the material underwent unusual expansion, phase changes or chemical instabilities that

would compromise its structure while in service [23, 38-40]. Tests were also conducted

to verify if LaPO4 was machinable as reported.

4.1.1 Coefficient of Thermal Expansion

The coefficient of thermal expansion (CTE) was determined using a dilatometer

DIL 2016 STD (Orton, USA). The test coupon made from the commercial LaPO4 powder

procured from Alfa-Aesar, after calcining and sintering as described in Chapter 3. The

CTE measurements were done in air between 25 and 1125°C and the results are shown in

Figure 15.

30

-35

-30

-25

-20

-15

-10

-5

0

5

10

25 225 425 625 825 1025

Temperature Deg. C

CT

E (

x 1

06 K

-1)

Figure 17. CTE of LaPO4.

As can be seen from Figure 15, between room temperature and 1100°C, the

ceramic expands by 0.75% during heating and contracts by 3% upon cooling of its

original linear dimension.

During thermal cycling, the expansion and contraction however, did not

compromise the structure of LaPO4 (within the operational temperature range of the

nozzles) since exceeding 1000°C in the regeneration process is uncommon. The LaPO4

nozzle must be jointed with a Hastelloy C22 gun body, which has a CTE value of 12.4 x

10-6 K

-1 below 500°C and 15.8 x 10

-6 K

-1 at 1000°C. Obviously, the CTE values for the

ceramic and the alloy differ, but this difference can be accounted for in the gun design.

31

4.1.2 Thermal and thermogravimetric Analysis

Analysis by thermogravimetry and differential scanning calorimetry (TGA/DSC)

was conducted on the commercial and the lab-synthesized LaPO4 powders after

calcination. The analysis assisted in determining if the material experienced a phase

change or chemical instability within the temperature range of nozzle’s operation. The

synthesis process of LaPO4 powder was also analyzed using the TGA/DSC technique

analysis on both commercial and our samples. The results are compared in Figure 16.

0.985

0.987

0.989

0.991

0.993

0.995

0.997

0.999

0 275 550 825 1100

Temperature Deg. C

Weig

ht

Ch

an

ge

-0.6

-0.1

0.4

0.9

1.4

1.9

Heat

Flo

w (

W/g

)

Weight ChangeCommercial

Weight Change ThisWork

Heat Flow (W/g)Commercial

Heat Flow (W/g) ThisWork

Figure 18. Comparative TGA/DSC plots for the LaPO4 samples; our sample

contained 2 wt. % the binder (PVA).

The major weight loss was associated with the release and evaporation of water

absorbed during synthesis and storage [59]. Figure 17 shows the trend of weight loss and

32

heat flow characteristics over several cycles. Again, the first heating cycle shows the

largest weight loss. The second weight loss was not due to dehydration but to residual

reactants and intermediate products decomposing or off-gassing upon heating. This

weight loss only occurred during the first heating cycle and no significant weight change

occurred in subsequent heating cycles.

Figure 19. TGA/DSC analysis with three consecutive heat cycles.

The heat flow analysis suggested that there were structural changes at around 200

and 850°C. The changes in heat flow rates were not large in magnitude and further

diminished and remained small over repeated cycles. Based on TGA/DSC analysis

results, it can be concluded that LaPO4 did not experience any drastic or unusual phase

change or chemical instability that would compromise the functionality of the ceramic in

33

the intended application.

The TGA/DSC profiles in Figure 17 show certain distinction for the two samples

used here. The LaPO4 sample synthesized in this work and used for the analysis,

contained 2 wt% of PVA as binder and residual acid. Hence, it showed a weight loss of

0.013% compared to 0.003% in the case of commercial sample. PVA burn-off in our

sample is obviously the largest contributor to the observed large weight loss. The heat

flow peak at ~850°C is about 1W/g larger for our sample than in the case if its

commercial counterpart from Alfa-Aesar.

Possibly, the difference between the two TGA/DSC signatures of the samples was

due to the residual H3PO4 present and the secondary compounds that correspond with the

excess acid residue in our sample. Side reactions and undesired products were likely

formed when the reactants were mixed. Considering that excess H3PO4 existed, a

polytriphosphate phase could result and its presence could explain the observed peak at

about 850 °C. The polytrioxophosphate forms as the result of the following series of

reactions [59].

H3PO4(absorbed) � HPO3(adsorbed) + H2O(g)

2HPO3(ad) + LaPO4(s) � La(PO3)3(s) + H2O(g)

The net reaction being,

2H3PO4(absorbed) + LaPO4(solid) �La(PO3)3(s) + 3H2O(g)

34

This last net reaction, potentially accounts for the peak seen at about 850°C. Between

1050 and 1350°C the triphosphate decomposes into LaPO4 and gaseous phosphorous

pentoxide dimer [59]:

2La(PO3)3(s) � 2LaPO4(s) + P4O10(liquid, gas)

A weight loss due to the release of P4O10 gas in this high temperature regime

accompanies the incongruent melting of La(PO3)3 [53, 59], accounting for the one-time

weight loss during the first heating step, especially at the higher end of temperature, and

the difference in heat flow characteristics of the two sets of LaPO4 samples. This

incongruent melt and weight loss follows the phase diagram of La2O3 – P2O5 illustrated

in chapter 3. Also, the vaporization of by-products is illustrated in the vaporization losses

of P2O5 from La(PO3)3 and LaP5O14. These figures are useful based on the consideration

of excess La or P.

4.1.3 Machinability

A small coupon of sintered LaPO4 was tested for its machinability. The coupon

was drilled using a carbide-tipped bit. The drill bit cut through the material resulting in

crisp defined edges and minimal chipping. A drill bit without carbide was attempted and

the material was cut through, but chipping and fracturing occurred to a small extent.

Hence, LaPO4 is machinable using carbide tools.

35

4.2 Corrosion Tests

The sintered LaPO4 samples were tested in simulated conditions close to what a

nozzle would experience during service. The two modes of testing, viz., exposure to the

furnace gasses and boiling acid were used.

4.2.1 Field Test

The field tests were conducted to determine if the materials could be used in the

furnace. The leading factor in determining if they were usable was if each material

would retain the required structural properties to compose a nozzle. The coupons were

tested in conditions similar to those experienced within the furnace. The coupons were

placed in a carbon steel container and inserted into an unoccupied spent gun port while

the furnace was operating. The container eliminated direct flame contact and reduced

soot deposition. A hand-held type K thermocouple was periodically used to determine

the temperature within the container. The temperature fluctuated from 540 to 1050°C

depending up on the gas flow rates and flow patterns within the furnace. To normalize

these fluctuations, the samples were left in the furnace for a minimum of two days and

tested during typical plant operating conditions. This ensured each set would experience

very similar conditions.

The first field test used A 36 mild steel, AISI 316 stainless steel, Hastelloy C22,

Al2O3 and LaPO4 as the study materials in furnace gas corrosion experiments. After the

first run, A 36 mild steel and AISI 316 were eliminated from the test because the

corrosion in these two samples was too severe compared to other materials. The mild

steel and stainless steel appeared to carburize and developed brittle flaky coating on the

exterior surface that came off easily. Remaining tests used Hastelloy C22, Al2O3 and

36

LaPO4. The corrosion rate of Hastelloy C22 alloy and the ceramics was comparable, but

the ceramics offered a slight improvement. Al2O3 showed the best resistance to high

temperature environment within the furnace.

Al2O3

LaPO4 (2%PVA)C22

0.00E+00

1.00E-06

2.00E-06

3.00E-06

4.00E-06

5.00E-06

6.00E-06

7.00E-06

1

Sample Material

Weig

ht

Ch

an

ge

(g

/hr)

Al2O3

LaPO4 (2%PVA)

C22

Figure 20. Corrosion results for Al2O3, LaPO4 and Hastelloy C22 in the furnace.

During the field tests conducted in the furnace only the alloys were greatly

affected and the source of corrosion appeared to be carburization. Carburization, a

corrosion process, affects steels and high-temperature alloys mostly at temperatures

above 900 °C. Ingress of carbon leads to the precipitation of stable carbides M23C6 and

M7C3 (M = Cr, Fe, Ni), with Cr as the main metal component. This precipitation causes

embrittlement, a volume increase of the carburized zone, cracking in the outer surface

and also loss of oxidation resistance. This kind of carburization can be widely suppressed

by protective oxide scales, but in cases of oxide failure it is controlled also by surface

37

reaction and carbon diffusion kinetics [60]. This mechanism is viable based on the

temperature and composition of the furnace gas.

Table 4. Average post combustion furnace gas composition

N2 H2O SO2 CO2 O2

50% 27% 12% 10% 2%

Based on the results of the field tests it appears that Al2O3, Hastelloy C22 and

LaPO4 could withstand the furnace gas conditions. Based on the rate of nozzle failure

during operation, the hot gas corrosion is not the leading mechanism of failure. The field

tests in the furnace port showed that the leading cause of failure is boiling acid on the tips

of the nozzles.

The field test also established that these materials could withstand the thermal

shock during insertion into and removal from the furnace. The coupons were placed in

the furnace gun port without preheating or a controlled heating rate, resulting in a rapid

rate of heating. Once the test was concluded the samples were removed and air cooled.

The thermal stress from the rapid heating and cooling did not damage the coupons. This

showed that LaPO4 could withstand the typical procedure of commissioning and

decommissioning of the gun.

4.2.2 Consistency of Boiling Acid Test Procedure

The first set of boiling acid corrosion tests was run to determine the variability of

the test method. This assisted in determining how the testing apparatus and the acid

samplings affected the outcome of the tests. All tests were run for approximately 28

hours (ranging from 24 – 32 hours) with coupons of similar size and weight. Also, each

38

of the Hastelloy C22 coupons were cut from the same plate intended for construction of a

piece of equipment.

-0.02500

-0.01500

-0.00500

0 10 20 30 40 50

Time (hours)

Ma

ss

Lo

ss

(g

/cm

2)

Figure 21. Consistency of the corrosion rate for Hastelloy C22 in 99 wt. % H2SO4.

Based on the experimental data shown in Figure 19 following statistics were

calculated for the variance tests (Table 5).

Table 5. Statistical analysis of variability of corrosion test

Property Value

Average Time (h) 27.7

Average Corrosion Rate (g /cm2 /h) 0.006

Standard Deviation of Corrosion Rate 0.0026

Median Corrosion Rate (g /cm2 /h) 0.0059

Average Surface Area (cm2) 6.56

Standard Deviation Surface Area 0.33

Number of Tests 6

Acid Selection Random

Acid Strength (% wt. H2SO4) 99.25±0.25

Despite slight differences in coupon size and surface area the test results were

39

fairly consistent and reproducible. The boiling acid apparatus provided an accurate

simulation of the field conditions of the nozzle.

4.2.3 Hastelloy C22 Boil Tests

Hastelloy C22 corrosion tests were conducted to develop a mode of comparison

for the new materials under consideration. It was known that advanced corrosion was

occurring, but the rates of corrosion were not known. The boiling test rates provided an

estimation of corrosion rates during operation. Based on protocol, the acid could be

between 87 – 99 % wt. H2SO4, but in common practice the acid would be close to 90 – 99

% wt. H2SO4. For this reason the tests were conducted with 99 % wt. H2SO4 and 90 %

wt. H2SO4, so as to cover a broad ranger

In the case of Hastelloy C22, the acid appeared transparent (clear water) initially

and proceeded to become milky light green in color. For the longer tests, sludge began to

build up at the bottom of the container. The alloy would appear burnt, transforming from

a shiny metallic luster to dull grayish black; bubbles of H2 were also produced. The

longer the coupon was in the acid the darker it became. The dull grayish black coloration

was only on the surface; scoring would reveal a shiny layer directly beneath the surface.

Figure 20 illustrates that the weight loss is linear on a per surface area per time basis.

40

-0.006

-0.005

-0.004

-0.003

-0.002

-0.001

0

0 50 100 150 200

Time (hours)

We

igh

t lo

ss

(g/c

m2/h

)

99 %wt.

90 %wt.

Figure 22. Hastelloy C22 corrosion rates as a function of acid strength.

As can be seen, the corrosion of the Hastelloy C22 coupons was much more rapid

in 90 wt. % acid than in 99 wt. %. This was evident by the discoloration of the acid,

taking on a much darker green hue. The discoloration of the surface was dark for the

shorter duration test, but the longer duration test showed localized spots that were pitted

and shiny. The testing of Hastelloy C22 in sulfuric acid established a standard for future

materials and assessed the variance of different testing method. It also verified that 90

wt. % was much more corrosive than the 99 wt. % and that the corrosion rate was directly

related to the time of exposure.

4.2.4 Al2O3 Boiling Acid Tests

Alumina (Al2O3) was also tested in boiling sulfuric acid. High density

machinable alumina ceramic (99.99% pure) was purchase from McMaster–Carr

(Cleveland, USA) and cut into 5/16″ x 1″ coupons. Without a diamond blade wet saw, the

material fractured and chipped. It was apparent that to compose a nozzle with this

41

material it would have to be molded and net-formed prior to sintering.

-6.00E-04

-5.00E-04

-4.00E-04

-3.00E-04

-2.00E-04

-1.00E-04

0.00E+00

1.00E-04

0 20 40 60 80 100 120 140

Time (h)

We

igh

t lo

ss

(g

/h)

99 wt.% H2SO4

90 wt.% H2SO4

Assumed this is a balance

measuring error and not an

actual gain in mass.

Figure 23. Results of Al2O3 corrosion in 90 and 99 wt. % refinery grade sulfuric

acid.

Despite the foreseen difficulties in fabricating a nozzle, the alumina ceramic

exhibited promising results in the corrosion tests. The influence of acid on alumina

ceramic was not discernable as the coupons appeared unaffected. The acid was clear and

transparent before and after the test. The alumina did not exhibit any of the common

characteristics of degradation as the other samples did, such as breaking off, discoloration

of acid and the sample surface, or acid absorption into the material.

4.2.5 LaPO4 Boiling Acid Tests

Many boiling acid tests were conducted on LaPO4 during the determination of the

optimal synthesis route for LaPO4, which was an iterative process. A batch would be

made, calcined, sintered and then tested in the boiling acid apparatus. Based on the

observations, changes to the synthesis process were made accordingly. The first rounds

42

of tests considered the sintering temperature and soak-time. The samples sintered at

lower temperatures (900-1100°C) did not withstand the boiling acid and rapidly

dissolved. When the soak-time was extended, the longer duration did not have a benign

effect on the material resistance to the boiling acid. The sintering temperature was then

increased to 1300°C and PVA (2 wt. %) was included as a binder to aid densification

during sintering process. These changes made minimal improvements to the resistance

when compared to the samples sintered at 900 or 1100°C.

Finally, a sintering temperature of 1585°C resulted in a significant improvement

in the resistance to the boiling acid. The LaPO4 coupon sintered at 1585°C withstood 99

wt. % boiling acid for nearly 20h without PVA addition and over 100h in the case where

binder was added. This was a vast improvement from previous samples sintered at the

lower temperatures. Further tests employed longer sintering time, but this did not

increase the resistance noticeably.

0.0

20.0

40.0

60.0

80.0

100.0

120.0

Coupon

Ho

urs

LaPO4 No Binder

99%

LaPO4 2% PVA 99%

LaPO4 2% PVA 90%

LaPO4 without PVA

boiled in refinery

grade H2SO4 99 %wt.

LaPO4 w/ 2% wt. PVA

boiled in refinery

grade H2SO4 99 %wt.

LaPO4 w/ 2% wt. PVA

boiled in refinery

grade H2SO4 90 %wt.

Figure 24. Time for complete dissolution of LaPO4 coupons.

43

As can be seen in this case also, the weaker acid degraded LaPO4 ten times faster

than the stronger acid did; LaPO4 with and without PVA dissolved in a consistent

manner. Each coupon would begin as a unified mass, but small pieces would break off.

The first pieces would be about a millimeter square and the subsequent debris would

progressively decrease in size to eventually become a fine grained powder.

4.2.6 Lifetime Estimation

Comparing the performance of Al2O3, Hastelloy C22 and LaPO4, in the

acid boiling unit, only Al2O3 showed appreciable resilience. LaPO4 did not show

promising results; the samples dissolved within a span of time that is too brief to function

as a nozzle. The results of the corrosion resistance between Hastelloy C22 and Al2O3 are

listed in table 6.

Table 6. Boiling acid corrosion test comparison of Al2O3 and Hastelloy C22

Material Acid wt. % g/cm2/hr g/hr

Al2O3 99 0 0

Hastelloy C22 99 1.50E-04 9.70E-04

Al2O3 90 6.38E-05 5.08E-05

Hastelloy C22 90 5.30E-03 2.10E-02

The resistance of alumina ceramic to the boiling acid was an order of magnitude

greater than that of Hastelloy C22; 10-5 as opposed to 10

-4 for

the 90 wt. % H2SO4. In the

99 wt. % acid, alumina did not experience weight loss, whereas the Hastelloy C22 alloy

had appreciable weight loss. The rates may not seem to be great, but considering that the

nozzle is used continuously (24 hours a day, 365 days a year) and considering that the

44

nozzle body is extremely thin (about 0.635 cm) it can be destroyed in a very short period

of time.

Lifetime of a nozzle can be roughly predicted based on the size of a common hole

found in a failed nozzle. When the diameter of the hole reaches ~ 1 mm, the disruption of

spray becomes noticeable. Using this threshold, a time frame for failure can be estimated

using the corrosion rates determined in the laboratory. Figure 23 represents the estimated

failure time for a nozzle using the individual corrosion rates from 99 wt. % H2SO4 boil

tests, the average surface area (based on 1 mm diameter hole) and thickness of the nozzle.

The time taken by the acid to create a 1 mm diameter hole in the nozzle represents the

lifetime.

0.00

10.00

20.00

30.00

40.00

50.00

60.00

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Individual Test

Lif

eti

me

(d

ay

s)

Figure 25. Estimation of nozzle failure time based on the corrosion rates for

Hastelloy C22 in 99 wt. % H2SO4.

45

C22, 0.49

Alumina,

6.60

0.00

2.00

4.00

6.00

8.00

10.00

1

Tests

Lif

eti

me

(d

ay

s)

Figure 26. Comparison of nozzle failure time for alumina and Hastelloy C22

corrosion in 90 wt. % H2SO4.

The prediction of nozzle lifetime for the Hastelloy C22 in 99 wt. % H2SO4 agrees

well with the experiments in the field. The alumina is not represented in figures for 99

wt. % H2SO4 because there was no measurable corrosion. The Hastelloy C22 and

alumina in the 90 wt. % H2SO4 have a much reduced lifetime, but the alumina lasted

nearly 13 times longer than the Hastelloy C22. The predicted failure time represents the

time a nozzle will last when acid begins to pool on the nozzle; if pooling does not occur

the nozzle should last much longer. The results emphasize the importance of proper

operations and material selection.

4.3 Corrosion Mechanics

The results of the boiling acid tests and the furnace tests showed that the leading

cause of failure was corrosion due to acid boiling at the tip of the nozzle. Both

environments were corrosive and aided to the degradation of the material.

46

4.3.1 LaPO4

The SEM/EDS analysis of the post-test samples shows that the corrosion

mechanism responsible for degrading LaPO4 originates at the grain boundaries. The

segregation of alumina to the grain boundaries is illustrated in Figure 26. Al2O3 is an

impurity phase present in LaPO4. It segregates to the grain boundaries but does not react

with the lanthanum phosphate [26, 29-31, 34, 35, 92]. During acid boiling, Al2O3 is

dissolved along with the grain boundaries and material loss occurs which increases with

time. This can be seen in Figure 25.

47

Figure 27. LaPO4 our sample with 2 wt. % PVA sintered at 1575 °C, EDS scan

broad view 1000 µm [61].

Figure 28. LaPO4 our sample with 2 wt. % PVA sintered at 1575 °C, EDS scan

grain boundary zoom view 100 µm [61].

48

The analysis of the sintered LaPO4 after exposure to the boiling 99 wt. % H2SO4

revealed that the acid and LaPO4 were reacting. The EDS pattern shown in Figure 27

identifies the major elements as lanthanum (La), oxygen (O) and sulfur (S). The scan did

not resemble the EDS scans of the coupon before the test.

Figure 29. EDS scan of LaPO4 laboratory synthesized and acid boil tested in 99 wt.

% H2SO4 [61].

Based on the EDS results it appears that the acid boiling removed the alumina

impurity and phosphate. The sample was not cross-sectioned and depth-profiled, so it is

49

not known if this is a surface reaction or if the acid penetrated the pores and reacted with

the bulk of the ceramic.

The first phase of decomposition would be the removal of the phosphate group.

This could be accomplished in a reaction similar to the dissolution of phosphate glasses,

when glass-ceramics are reheated to temperatures above crystallization. In the case of

LaPO4 the residual carbon and acid within the bulk ceramic creates different phase

boundaries and promotes the dissolution of the boundaries, as seen in the SEM of the

corroded sample.

Within the bulk ceramic the following materials exist in some unknown

concentration.

LaPO4 ; La3PO7 ; La2O3 ; La(PO3)3

The impurities at the grain boundaries and the temperatures at which the acid boils could

greatly affect the stability of the phases promoting the formation of a PxOy phase.

Dissolution of PxOy could promote the formation of LaPO4 which will then react with the

sulfur and water.

An estimation of the possible reactions was derived from the use of lanthanum in

flue gas desulphurization and oxygen storage for automotive catalysts in exhaust

conversion. Each system uses a form of lanthanum derived from lanthanum oxide. The

lanthanum oxide is then hydrated and reacted with a form of sulfur.

Derivatives of this material can be formed through the following reaction.

50

2La(OH)3 ↔ La2O3 + 3H2O

This reaction will occur naturally in lanthanum oxide powders exposed to the

atmosphere. The hydration / dehydration ranges are illustrated below.

2La(OH)3 � 2LaOOH � La2O3 377°C (-2H2O) 527°C(-H2O)

La(OH)3 readily transforms into La2O2S upon reaction with sulfur-bearing species, such

as SO2.

2La(OH)3 + SO2 ↔ La2O2S + 3H2O + 3/2O2

A novel method of oxygen storage utilizes lanthanum oxysulfate (La2O2SO4). This

material will produce free oxygen through the following reversible reaction:

La2O2SO4 ↔ La2O2S + 2O2

This is based on the redox of sulfur in a lanthanum oxysulfate / oxysulfide (La2O2SO4 /

La2O2S) system. The redox reaction is based on the S6+ and S

2- cycle [62-64]. The

resulting lanthanum oxysulfate/sulfide is the final product that eventually disintegrates

into the powder found on the bottom of the beaker.

In the present case, the corrosion process started with LaPO4 dissolving in a

51

similar manner as phosphate glass and in the presence of other lanthanum compounds,

the system proceeds to the redox state. The impurities and non-stoichiometric species

greatly increase the negative influences of the acid. Dissolution of the grain boundary

increases the contact of the acid with the ceramic, increasing thereby the rate of

degradation. This proposed mechanism agrees well with the EDS analysis and the high

concentration of La, S, and O.

For further testing and defining the corrosion mechanism the suggested method is

outlined in the appendix A-4.

4.3.2 Hastelloy C22

Hastelloy C22 has better corrosion resistance compared to many other commercial

alloys. However, it does experience some corrosion and does not provide an adequate

lifetime for nozzles used in acid injection. The intention of using it in the current study

was to evaluate LaPO4 ceramic against the Hastelloy C22 for potential use as a nozzle.

In the process of comparison, the test protocol provided sufficient data to suggest a

plausible mode of corrosion and potentially identify corrective measures to improve the

material resistance.

Much consideration has thus far been put into the alloy composition, but little has

gone into how the acid interacts with it. The anodic passivity formed between the acid

and alloy involves the solid-state formation of a chromium-rich oxide–hydroxide passive

film with the participation of undissociated H2SO4 as the source of oxygen [65]. The

surface process of passivation has a strong build up of bisulfate/sulfate (HSO4-/SO4

2-) as a

52

monolayer coverage. A significant part of the sorbed bisulfate/sulfate ions is strongly

bonded to the passive layer thus formed [66]. The SO42- bonds not only to the surface of

an oxide film formed during corrosion of the alloy sample, but are also incorporated into

the layer up to a depth of several monolayers [66, 67]. Bisulfite ions (even in low

concentration) inhibit the formation of the compact and protective passive layer and

increase the rate of alloy surface passivation (formation of a thicker surface layer) [67].

This can be seen in the tested samples of Hastelloy C22. The samples boiled in

the 99 wt % sulfuric acid did not experience advanced corrosion as did the samples

boiled in the 90 wt % sulfuric acid. This becomes evident while comparing the

microscopic features of the surface. The coupons exposed to the 99 wt % acid show a

high concentration of iron, sulfur and oxygen and have a black burnt appearance. This is

assumed to be the passivation layer. In the case of coupons exposed to 90 wt %, the

passivation layer appear gray and has same level of iron, sulfur and oxygen as the 99 wt

%, but also have increased chromium and nickel. The passivation layer was not uniform

on the 90 wt % and had shiny spots dispersed through the gray surface. Electron

microscopy of the shiny areas was found to have high concentration of iron,

molybdenum, nickel, chromium and oxygen. Nickel and molybdenum have much greater

presence than any other elements on the surface. It is assumed that these shiny areas

were located where penetration of the passivation layer had occurred. This is based on

the high concentrations of nickel and molybdenum signifying that the iron and chromium

had experienced advanced dissolution.

The SEM/EDS analysis was also used to further consider how the boiling acid

53

was affecting the Hastelloy C22. For a mode of comparison a coupon of as-received non-

tested Hastelloy C22 was analyzed as shown below.

Figure 30. SEM/EDS analysis of as-received Hastelloy C22 [61]; two large peaks

are due to Fe and Ni.

The EDS microanalysis of the as-received Hastelloy C22 shows the high amounts

of the key alloying elements (iron, chromium, nickel and molybdenum). There is minimal

oxygen present, suggesting that only minimal oxidation occurred on the surface. A few

different locations were scanned which also showed very similar results.

54

Figure 31. EDS Scans of Hastelloy C22 corrosion tested in 90 wt. % sulfuric acid.

Coupons exposed to acid had two levels of exposure, passivation and penetration.

The passivated layers had a uniformed black appearance and minimal weight loss; the

EDS patterns revealed mainly iron, sulfur and oxygen in the layer close to the surface.

Penetrated samples had a non-uniformed grayish appearance with random regions that

were shiny and metallic. EDS analysis of the gray areas revealed results similar to the

passivated layer, but the shiny areas were high in the alloying elements (iron, chromium,

nickel and molybdenum); no sulfur was present.

Passivation layers became more pronounced as acid strength decreased. All

samples exposed to the 99 wt. % H2SO4 had a corrosion rate two orders of magnitude

lower than those exposed to the 90 wt. % acid. All coupons had a black appearance

which is thought to be the passivation layer. EDS revealed that the major elements at the

surface were iron, sulfur and oxygen. It is assumed that the surface process of

55

passivation has a strong build up of bisulfate/sulfate (HSO4-/SO4

2-) as monolayer

coverage. A significant part of the sorbed bisulfate/sulfate ions are strongly bonded to

the passive layers formed and the dominant element in the passive layer was iron. This

agrees with the corrosion mechanism being reaction limited by the mass transfer of iron

in stainless steel. The increased concentration of acid favored the alloy and limited

corrosion, whereas the lower concentration did not and iron was depleted at a much faster

rate.

This is supported by the rates of weight loss determined experimentally. The

presence of iron, chromium, nickel and molybdenum at the surface region show that the

iron concentration has decreased to a level that no longer supports a large buildup of

sulfates monolayer. It is not known as to why the 90 wt. % H2SO4 decreases the

effectiveness of the passivation layer. It is stipulated that the increased water content

increases the solubility of the passive compounds formed by the protective alloys and

increases the electrochemical potential for corrosion.

56

Chapter 5

Conclusions and Recommendations

5.1 Conclusions

The sheer volume of sulfuric acid produced annually in the US as well as

globally, its diverse use and the inherent safety risks to people and the environment

require that robust and reliable processes are developed and implemented to mitigate the

risks associated with the spent/used acid disposal and the efficient production of fresh/

new batch. To improve the reliability of the process, the materials used to fabricate the

nozzles for the regeneration of spent sulfuric acid were analyzed in this thesis, with the

goal of increasing the corrosion resistance in the highly corrosive environment. The

material of specific interest was lanthanum phosphate monazite (LaPO4) due mainly to

many of its attractive physico-chemical attributes. Its performance was compared with

that of currently utilized material, viz., Hastelloy C22 as well as other alternative

substitutes that could potentially be used. While LaPO4 is known to exhibit exceptional

corrosion resistance in many applications under extreme and severe process conditions, it

did not show a corrosion resistance superior to that of the currently available materials for

sulfuric acid injection nozzles; hence, it cannot be considered a suitable replacement

material.

57

LaPO4 was considered because of its reported chemical and thermal stability, in

addition to its easy machinability. To assess its compatibility as a nozzle material for

sulfuric acid injection into a thermal decomposition furnace, sintered coupons were

subjected to systematic tests and analyses, including the determination of its mechanical

properties, such as the coefficient of thermal expansion, phase transitions via

simultaneous thermogravimetry and differential scanning calorimetry. This determined if

the material could function as a bulk ceramic for composing a nozzle in the intended

temperature range.

With that done, it was further tested to assess its resistance to corrosive

environments. A series of boiling sulfuric acid tests with varying strengths was

conducted to quantify its resilience to aqueous acid corrosion. Field tests in a thermal

decomposition furnace examined how well the material performed in the high-

temperature corrosive process gas environment. Concurrently, other materials (Al2O3,

AISI 316, Hastelloy C22) were also subjected to identical tests. The results served as a

mode of comparison of the performance of LaPO4. Hastelloy C22 represented what was

currently in use, AISI 316 confirmed the requirement for specialty alloys and the two

ceramics (LaPO4 and Al2O3) explored future options of material design with LaPO4.

In the testing and analysis period, the LaPO4 synthesis protocol was also refined

to make the material behavior more in line with the required performance criterion. In

that aspect, it was found that using an organic binder (PVA) improved the densification

characteristics, which directly improved its resistance to degradation when subjected to

the boiling test in concentrated sulfuric acid. Improvements in the material synthesis of

58

LaPO4 greatly improved its corrosion resistance, but the improvements were not

significant enough to establish LaPO4 as the material of choice for composing sulfuric

acid injection nozzles. LaPO4 experienced dissolution in sulfuric acid at a rate greater

than Hastelloy C22; Al2O3 on the other hand, exhibited superior resistance to the boiling

test compared to LaPO4 and Hastelloy C22 both.

5.2 Recommendations

The research conducted and reported here improved the understanding of how and

why the nozzles fail in this particular application of immense industrial application. One

of the key findings was the tendency for the pooled acid on the tip of the nozzles to

decrease in acid strength, which greatly increased the corrosion rate, due to the

hygroscopic properties of the acid and the high concentration of water in the combustion

process. The interaction between the nozzle and the corrosion process emphasized the

importance of future nozzles to utilizing geometries that minimize acid pooling on the tip

of the nozzle. Also, design consideration must be given to how the gun is inserted into

the furnace port. The gun must be installed in a manner that shields the nozzle from the

direct heat of the furnace and maximizes the drainage of pooled acid from the tip of the

nozzle. These discoveries are significant and are expected to improve future furnace

design and acid injection systems.

The superior corrosion resistance of Al2O3 system suggests that it should be used

to compose the nozzles, but the material is not machinable and does not fit into the

maintenance structure of the plant. However, based on the results obtained in this work

and the published research on the utilization of Al2O3 mixed with LaPO4 to form a

59

composite could produce a machinable ceramic with improved toughness. It is suggested

that future research consider this composite for fabricating sulfuric acid injection nozzles.

The superior corrosion resistance exhibited by Al2O3 would provide a long, predictable

life for the nozzle, and the incorporation of the LaPO4 would increase the material

toughness and machinability, thus allowing it to be incorporated into the current

maintenance structure.

Not only was important information determined concerning the sulfuric acid

injection nozzles, but the details of material corrosion and concentrated sulfuric acid was

further defined. A better understanding of the mechanisms of corrosion in sulfuric acid

manufacturing were attained and can be used for future selection of materials and

alloying element concentrations for process equipment. The proposal that grain boundary

dissolution was the leading mode of failure for LaPO4 puts increased emphasis on

attaining ceramics of increased purity, reducing contaminants and minimizing side

reactions of lanthanum and phosphorous. Future research should consider the

relationship of sintering and microstructure to the corrosion resistance of LaPO4.

Optimization of the microstructure could reduce grain boundary dissolution and increase

corrosion resistance. The proposal about boundary dissolution also identifies the need to

improve the ceramic synthesis process and alloys with increased contents of

molybdenum, chromium and nickel. The study of injection nozzle corrosion will help

improve corrosion resistance materials in sulfuric acid manufacturing and illustrates the

need for more research in the field of corrosion.

60

Works Cited

1. Limited, R.U. Rhodia North America. 2007 [cited 2007 9-10-2007]; Available

from: http://www.na.rohdia.com.

2. Chemical of the Week Sulfuric Acid. 2001 [cited 2007 January 4]; Available

from: http://scifun.chem.wisc.edu/chemweek/Sulf&top/Sulf&Top.html.

3. Alexandru Platon, W.J.T., Solid acid characteristics and isobutane/butene

alkylation. Applied Catalyst A:General, 2005. 282: p. 93-100.

4. J. Weitkamp, Y.T., Isobutane/butene alkylation on solid catalysts. Where do we

stand? Catalysis Today, 1999. 49: p. 193-199.

5. Hommeltoft, S.I., Isobutane alkylation Recent developments and future

perspectives. Applied Catalyst A:General, 2001. 221: p. 421-428.

6. Albright, D., The Future of Alkylation on Oil Refineries, J. Wright, Editor. 2007:

Hammond.

7. Louie, D., Handbook of Sulphuric Acid Manufacturing. 2005, Toronto: DKL

Engineering Inc.

8. Wang, G., Alkylation & Polymerization Chapter 11. 2003: Rhodia Technology

Group. p. 29.

9. Belco, D. DuPont Belco Clean Air Technologies. 2006 [cited 2007 8-09-2007];

Available from: http://www.belcotech.com/products/wetelec.html.

10. DuPont. DuPont SAR/SGR Technologies. 2007 [cited 2007 August 28];

Available from:

http://www2.dupont.com/Clean_Technologies/en_US/products/sar/sar_sgr.html.

11. Inc., M. Marsulex, Proven Solutions. 2007 [cited 2007 8-14-2007]; Available

from: http://www.marsulex.com/proven_solutions/examples/bp_whiting.html.

12. MECS. MECS Global Home Page. 2007 [cited 2007 9-18-2007]; Available

from: http://www.mecsglobal.com/MECS/layout/default.asp.

13. Rosenberg, H., Topsoe wet gas sulphuric acid (WSA) technology - an attractive

alternative for reduction of sulphur emissions from furnaces and converters.

2006: Lyngby. p. 8.

14. James Slavas; Matthew Betsold; Douglas Dziadzio; Robert Bedaw, S.V.B., Flat

Fan Spray Nozzle, U.P. Office, Editor. 1996, Bete Fog Nozzle, Inc.: USA. p. 10.

61

15. Charles Anthony Dafft, C.R.W., Process for generating recoverable sulfur

containing compounds from a spent acid stream, U.S.P. Office, Editor. 2002,

Rohm and Haas Company: USA.

16. International, H., Hastelloy C-22 Alloy, in Corrosion-Resistant Alloys. 2002,

Haynes International Inc. p. 24.

17. International, H., Hastelloy C-276 Alloy, in Corrosion Resistant Alloys. 2002,

Haynes International Inc. p. 16.

18. International, H. Hastelloy B-2 alloy. Haynes Corrosion-Resistant Alloys [cited

2006 October 10, 2006]; Available from:

http://www.haynesintl.com/mini/B2s/B2.htm.

19. Metals, E.-. Alloy 20 Composition.

20. Charles Anthony Dafft, C.R.W.

21. Fiquet; Olivier (Venelles, F., Croixmarie; Yves (Aixes en Provence, FR), Method

of manufacturing compounds of the monazite type, doped or not doped with

actinides and application fo the packaging of radioactive waste high in actinides

and in lanthanides, U.S. Patent, Editor. 1998, Commissariat A L`Energie

Atomique (Paris, FR): United States of America. p. 15.

22. Soule, L.S., Flat fan spray nozzle, U.P. Office, Editor. 1997, Bete Fog Nozzle,

Inc.: USA. p. 12.

23. X.Q. Cao, R.V., D. Stoever, Ceramic materials for thermal barrier coatings.

Journal of European Ceramic Society, 2004. 24: p. 1-10.

24. J.B. Davis, D.B.M., P.E.D. Morgan, Monazite-containing oxide/oxide composites.

Journal of the European Ceramics Society, 2000. 20: p. 583-587.

25. Suree S. Brown, H.-J.I., Adam J. Rondinone, Sheng Dai, Facile, alternative

synthesis of lanthanum phosphate nanocrystals by ultrasonication. Journal of

Colloid and Interface Science 2005. 292: p. 127-132.

26. Ruigang Wang, W.P., Jian Chen, Minghao Fang, Jun Meng, Effect of LaPO4

content on the microstructure and machinability of Al2O3/LaPO4 composites.

Material Letters, 2002. 57: p. 822-827.

27. Jingyang Wang, Y.Z., Zhijun Lin, First-principle elastic stiffness of LaPO4

monazite. Applied Physics Letters, 2005. 87.

62

28. L. Perri`ere, D.B., B. Naitali, F. Audubert , E. Championb, D.S. Smith, D.

Bernache-Assollant, Microstructural dependence of the thermal and mechanical

properties of monazite LnPO4 (Ln = La to Gd). Journal of the European Ceramic

Society, 2007. 27: p. 3207-3213.

29. Ruigang Wang, W.P., Jian Chen, Minghao Fang, Mengning Jiang, Zhenzhu Cao,

Microstructure and mechanical properties of machinable Al2O3/LaPO4

composites by hot pressing. Ceramics International, 2003. 29: p. 83-89.

30. K.K. Chawlaa, H.L., J. Janczak-Ruschc, S. Sambasivan, Microstructure and

properties of monazite (LaPO4) coated saphikon ®fiber/alumina matrix

composites. Journal of the European Ceramics Society, 2000. 20: p. 551-559.

31. Lo Bo, S.L., Liu Xiaozhen, Zhang Shuihe, Wu Chunfang, Liu Wenjing, Monazite

coatings on short alumina fibers using layer-by-layer assembly technique.

Material Science and Engineering A, 2004. 364: p. 324-332.

32. Henryk Tomaszewski, H.W.e., Anna Wajler, Marek Boniecki, Dariusz Kalinski,

Multilayer ceramic composites with high failure resistance. Journal of the

European Ceramic Society, 2007. 27: p. 1373-1377.

33. Guoliang Gong, B.Z., Hanrui Zhang, Wenlan Li, Pressurless sintering of

machineable Al2O3/LaPO4 composites in N2 atmosphere. Ceramics

International, 2006. 32: p. 349-352.

34. Ruigang Wang, W.P., Jian Chen, Mengning Jiang, Yongming Luo, Minghao

Fang, Properties and microstructures of machinable Al2O3/LaPO4 ceramic

composites. Ceramics International, 2003. 29: p. 19-25.

35. S. Boucher, J.P., R. F. Marzke, B. Takulapalli, G.H. Wolf, P.F. McMillan, W.T.

Petuskey, Melt and glass structure in the Al2O3-CaO-LaPO4 system studied by

27Al and 31P NMR, and by Raman scattering. Journal of the European Ceramic

Society 2005. 25: p. 1333-1340.

36. T.J. Hwang, M.R.H., H. Shao, H.G. Hornis, A. T. Hunt, Combustion chemical

vapor deposition (CCVD) of LaPO4 monazite and beta-alumina on alumina fibers

for ceramic matrix composites. Materials Science and Engineering A 1998. 244:

p. 91-96.

37. Wang Ruigang, P.W., Chen Jian, Fang Minghao, Cao Zhenzhu, Luo Yongming,

Synthesis and sintering of LaPO4 powder and its application. Materials

Chemistry and Physics, 2003. 79: p. 30-36.

63

38. W. Beele, G.M., A. van Lieshout, The evolution of thermal barrier coatings -

status and upcoming solutions for today's key issues. Surface and Coatings

Technology, 1999. 120-121: p. 61-67.

39. N.P. Padture, K.W.S., T. Bhatia, A. Ozturk, B. Cetegen, E.H. Jordan, M. Gell, S.

Jiang, T.D. Xiao, P.R. Strutt, E. Garcia, P. Miranzo, M.I. Osendi, Towards

Durable Thermal Barrier Coatings with Novel Microstructures Deposited by

Solution-Precursor Plasma Spray Acta Materialia, 2001. 49: p. 2251-2257.

40. Lee, K.N., Current status of environmental barrier coatins for Si-Based ceramics.

Surface and Coatings Technology, 2000. 133-134: p. 1-7.

41. C Thiriet, R.J.M.K., P. Javorsky, N. Magnani, F. Wastin, The low temperature

heat capacity of LaPO4 and GdPO4, the thermodynamic functions of the

monazite-type LnPO4 series. Journal of Chemical Thermodynamics, 2005. 37: p.

131-139.

42. Wu Min, D.M., K. Yokoi, T. Yamaguchi, K. Daimon, Yasuo Hikichi, T.

Matsubara, T. Ota, Thermal and mechanical properties of sintered LaPO4-Al2O3

composites. Materials Research Bulletin, 2001. 36: p. 939-945.

43. I.W. Donald, B.L.M., R.N.J. Taylor, Review The immobilization of high level

radioactive wastes using ceramics and glasses. Journal of Material Science, 1997.

32: p. 5851-5887.

44. I. W. Donald, B.L.M., S.K. Fong, L.A. Gerrard, D.M. Strachan, R.D. Scheele, A

glass-encapsulated calcium phosphate wasteform for the immobilization of

actinide-, fluoride-, and chloride-containing radioactive wastes from the

pyrochemical reprocessing of plutonium metal. Journal of Nuclear Materials,

2007. 361: p. 78-93.

45. R.C. Ewing, W.J.W.a.F.W.C., Jr., Radiation Effects in Nuclear Waste Forms for

High-Level Radioactive Waste. Progress in Nuclear Energy, 1995. 29(2): p. 63-

127.

46. Meldrum, B., Ewing, Displacive radiation effects in the monazite - and zircon-

structure orthophosphates. Physical Review B, 1997. 56(21): p. 13805-13814.

47. Yu, R., DeJonghe, L. C., Proton-Transfer Mechanism in LaPO4. Journal of

Physical Chemistry C, 2007. 111(29): p. 11003-11007.

48. R.S. Hay, E.B., M.D. Petry, Effect of coating deposition temperature on monazite

coated fiber. Journal of the European Ceramic Society, 2000. 20: p. 589-597.

64

49. Wu Min, K.D., T. Matsubara, Y. Hikichi, Thermal and mechanical properites of

sintered machinable LaPO4-ZrO2 composites. Materials Research Bulletin, 2002.

37: p. 1107-1115.

50. J.B. Davis, D.B.M., K.S. Oka, R.M. Housley, P.E.D. Morgan, Ceramic

composites for thermal protection systems. Composites: Part A, 1999. 30: p. 483-

488.

51. Gabriel Harley, R.Y., Lutgard C. De Jonghe, Proton transport paths in lanthanum

phosphate electrolytes. Solid State Ionics, 2007. 178: p. 769-773.

52. D. Bregiroux, S.L., E. Champion, F. Audubert, D. Bernache-Assollant, Sintering

and microstructure of rare earth phosphate ceramics REPO4 with RE=La, Ce or

Y. Journal of the European Ceramics Society, 2006. 26: p. 279-287.

53. Hee Dong Park, E.R.K., Phase Equilibria in the System La2O3-P2O5. Journal of

the American Ceramic Society, 1984. 67(1): p. 23-26.

54. Yu V Bykov, K.I.R., V E Semenov, High-temperature microwave processing of

materials. Institute of Physics Publishing, 2001. 34: p. R55-R75.

55. Agrawal, D.K., Microwave processing of ceramics. Current Opinion in Solid

State and Materials Science, 1998. 3: p. 480-485.

56. Romualdo R. Menezes, P.M.S., Ruth H.G.A. Kiminami, Microwave hybrid fast

sintering of porcelain bodies. Journal of Materials Processing Technology, 2007.

190: p. 223-229.

57. Apex Instruments, I. Method 3 Orsat Analyzer. 2002 [cited 2008 1-10-2008];

Available from: http://www.apexinst.com/products/orsat.htm.

58. Stephen Sung, G.S., Lyle F. Albright, Decomposition of Spent Alkylation Sulfuric

Acid to Produce Sulfur Dioxide and Water. Industrial Engineering Chemical

Research, 1993. 32: p. 2490-2494.

59. S. Lucas, E.C., D. Bernache-Assollant, G. Leroy, Rare earth phosphate powders

RePO4*nH2O (Re=La, Ce or Y) II. Thermal behavior. Journal of Solid State

Chemistry, 2004. 177: p. 1312-1320.

60. H.J. Grabke, D.M., E. M. Muller-Lorenz, A. Schneider, Role of sulphur in

carburization, carbide formation and metal dusting of iron. Surface and Interface

Analysis, 2002. 34: p. 369-374

61. Hare, E., SEM/EDS Analysis of Corrosion Test Samples. 2007, SEM Lab, Inc.:

Snohomish. p. 32.

65

62. Machida, M.K., K.; Ito, K.; Ikeue, K., Large-Capacity Oxygen Storage by

Lanthanide Oxysulfate/Oxysulfide Systems. 2005. 17(6): p. 1487-1492.

63. Masato Machida, K.K., Kazuhiro Ito, Novel oxygen storage mechanism based on

redox of sulfur in lanthanum oxysulfate/oxysulfide. Chemical Communications,

2004. 2004(6): p. 662-662.

64. Jianxin Ma, M.F., Ngai Ting lau, Activation of La2O3 for the Catalytic Reduction

of SO2 by CO. Journal of Catalysis, 1996. 163(2): p. 271-278.

65. J.R. Kish, M.B.I., J.R. Rodda, Anodic behaviour of stainless steel S43000 in

concentrated solutions of sulphuric acid. Corrosion Science, 2003. 45: p. 1571-

1594.

66. D. Deb, S.R.I., V. M. Radhakrishnan, A comparative study of oxidation and hot

corrosion of a cast nickel base superalloy in different corrosive environments.

Materials Letters, 1996. 29: p. 19-23.

67. Christian Schmidt, K.R., Donald H. Bilderback, Rong Huang, In situ synchrotron-

radiation XRF study of REE phosphate dissolution in aqueous fluids to 800 °C.

Lithos, 2007. 95: p. 87-102.

66

Appendix

A-1 Historical Industrial Corrosion Studies

The following tests were conducted on different feeds of spent sulfuric acid. The

test procedure was placing the specified material coupon in a beaker sealed from

atmosphere for the allotted time and measuring the weight change at the indicated times.

The same coupons were used for multiple times. The acid was maintained at room

temperature and the spent acid was as received from the refiner or petrochemical facility.

67

Material Tested: A - 36

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0

42 125 166

Time (hours)

Gra

ms

Spent A

Spent B

Spent C

Figure 32. A 36 mild steel tested in room temperature spent sulfuric acid.

Material Tested: SS 304

-0.0025

-0.002

-0.0015

-0.001

-0.0005

0

42 125 166

Time (hours)

Gra

ms

Spent A

Spent B

Spent C

Figure 33. AISI 304 tested in room temperature spent sulfuric acid.

68

Material Tested: SS 316

-0.0018

-0.0016

-0.0014

-0.0012

-0.001

-0.0008

-0.0006

-0.0004

-0.0002

0

42 125 166

Time (hours)

Gra

ms

Spent A

Spent B

Spent C

Figure 34. AISI 316 tested in room temperature spent sulfuric acid.

69

A-2 Corrosion Data

Table 7. Specific details of boiling acid corrosion tests on Hastelloy C22 Initial Final Difference

Coupon Mass (g) a b c t in2

cm2

Mass (g) Mass (g) Hours g / h g / cm2 g / cm2 / hr % Diff. % St.

0 19.0539 0.523 0.528 1.058 0.2555 1.0967 7.0757 19.0206 -0.0333 27.33 -0.001218 -0.00471 -0.000172202 -0.175 99

1 16.5078 0.525 0.532 0.925 0.2555 1.0332 6.6658 16.49505 -0.01275 29.67 -0.00043 -0.00191 -6.44747E-05 -0.077 99

2 16.9512 0.594 0.595 0.978 0.256 1.2385 7.9902 16.9116 -0.0396 45.55 -0.000869 -0.00496 -0.000108805 -0.234 99

3 15.6969 0.493 0.48 0.928 0.2555 0.9461 6.1036 15.6491 -0.0478 32.833 -0.001456 -0.00783 -0.000238519 -0.305 99

4 17.8088 0.529 0.53 0.93 0.2555 1.0427 6.7268 17.7816 -0.0272 45.417 -0.000599 -0.00404 -8.90315E-05 -0.152733 99

5 18.7335 0.496 0.495 1.079 0.255 1.0337 6.6688 18.6917 -0.0418 26.667 -0.001567 -0.00627 -0.00023505 -0.22313 99

6 15.71 0.508 0.488 0.93 0.2555 0.9799 6.3221 15.6495 -0.0605 25.5 -0.002373 -0.00957 -0.000375277 -0.385105 99

7 17.889 0.489 0.473 1.078 0.2565 1.0093 6.5116 17.8534 -0.0356 24.25 -0.001468 -0.00547 -0.000225451 -0.199005 99

8 16.3424 0.492 0.492 0.971 0.255 0.9707 6.2627 16.1996 -0.1428 122.25 -0.001168 -0.02280 -0.000186518 -0.873801 99

20 6.44 0.487 0.617 0.329 0.272 0.7256 4.6815 6.426 -0.014 69.33 -0.000202 -0.00299 -4.31344E-05 -0.217391 99

21 6.2164 0.315 0.331 0.568 0.271 0.4539 2.9285 6.2011 -0.0153 92 -0.000166 -0.00522 -5.67887E-05 -0.246123 99

22 10.411 0.49 0.496 0.6149 0.275 0.8138 5.2506

23 10.6404 0.446 0.469 0.658 0.274 0.7471 4.8197

24 8.6098 0.364 0.3805 0.646 0.272 0.5717 3.6881 7.6978 -0.912 47 -0.019404 -0.24728 -0.005261266 -10.59258 90

25 10.6318 0.4735 0.463 0.636 0.274 0.7799 5.0313

26 9.9289 0.463 0.469 0.5545 0.2725 0.7262 4.6853

27 11.3251 0.4575 0.465 0.7095 0.2747 0.7887 5.0883 11.2969 -0.0282 168 -0.000168 -0.00554 -3.2989E-05 -0.249004 99

28 14.4153 0.508 0.615 0.695 0.272 0.9418 6.0763

29 8.0466 0.3685 0.363 0.639 0.2725 0.5701 3.6779

211 8.8209 0.384 0.386 0.6545 0.272 0.6084 3.9255 8.3486 -0.4723 22.5 -0.020991 -0.12032 -0.005347423 -5.354329 90

INCHES Surface

Table 8. Specific details of boiling acid corrosion tests on LaPO4

(All final weights are zero)

Coupon Int. Wt. Final Wt. Wt. Loss Start Time Finish Time Total Time hours %ST

LaPO4 no Binder 3.5765 0 -3.5765 8/10/2007 18:30 8/11/2007 12:00 17.5 99

LaPO4 2% PVA 8/11/2007 14:15 8/16/2007 12:15 106.00 99

LaPO4 2% PVA 9/8/07 10:20 9/9/07 19:20 9 90 Microwave Sintered 1575'C

Microwave Sintered 1575'C

Microwave Sintered 1575'C

Table 9. Specific details of boiling acid corrosion tests on Al2O3 Coupon Int. Wt. Final Wt. Wt. Loss Time Hours g/hr % St.

1 5.1567 5.1652 0.0085 133.25 6.38E-05 99

2 5.0902 5.0536 -0.0366 73.41666667 -4.99E-04 90

70

Table 10. Specific details of high temperature field tests of materials in the furnace

Material Initial Wt. Final Wt. Delta Wt. Duration (hours) Wt. / hour Temp Range Avg. Temp Comment

Al2O3 5.1091 5.3142 0.2051 504 4.07E-04 700 - 1800 850

Advanced Carburization built up on coupon from carbon steel dish. Coupon was still in original shape

and size, no visible damage.

LaPO4 (2%PVA) 1.3053 2.4806 1.1753 504 2.33E-03 700 - 1800 850

Advanced carburization built up on coupon from carbon steel dish. LaPO4 crumbled into pieces and

entire surface had carbon adhered onto it. No structural properties existed.

C22 14.4153 14.6343 0.219 504 4.35E-04 700 - 1800 850

Carburization occurred on the surface, the carburization from the carbon steel container did not

adhere. Can see carburization penetrated the C22 coupon to some extent. Wt. gain is a

combination of carburization of coupon and external sources of soot and carbon.

A36 Mild Steel 700 - 1800 850

The end cap used to hold the samples was corroded in half. It did not melt but turned into a charcoal

like substance with a large volume increase. This flakie substance (Carburization?) broke off

removing most of the metal. The whole cap was like this, losing the majority of its metal.

Al2O3 5.1798 5.1805 0.0007 64.5 1.09E-05 600 - 850 750

Appears the same as when put into the furnace no noticeable damage. Assume difference is within

error of scale. Assume Wt. change of 0.

LaPO4 (2%PVA) 0.4953 0.4962 0.0009 64.5 1.40E-05 600 - 850 750

Appears the same as when put into the furnace no noticeable damage. Assume difference is within

error of scale. Assume Wt. change of 0.

C22 9.9289 9.9309 0.002 64.5 3.10E-05 600 - 850 750

Exterior surface is extremely blackened and a black scale flaked off. Metal appears burned.

Assume wt. gain is due to carburization, if left in longer assume would be corroded further and start

to loose weight.

A36 Mild Steel 5.7901 5.6719 -0.1182 64.5 -1.83E-03 600 - 850 750

Exterior surface is extremely blackened and a black scale flaked off. Metal appears burned.

Assume wt. gain is due to carburization. This was the worst of all the materials.

304L 16.106 16.1013 -0.0047 64.5 -7.29E-05 600 - 850 750

Exterior surface is extremely blackened and a black scale flaked off. Metal appears burned.

Assume wt. gain is due to carburization.

Al2O3 5.1805 5.1822 0.0017 92.00 1.85E-05 850 - 1100 800

Appears the same as when put into the furnace no noticeable damage. Assume difference is within

error of scale. Assume Wt. change of 0.

LaPO4 (2%PVA) 0.4962 0.4965 0.0003 92.00 3.26E-06 850 - 1100 800

Appears the same as when put into the furnace no noticeable damage. Assume difference is within

error of scale. Assume Wt. change of 0.

C22 9.9309 9.9308 -1E-04 92.00 -1.09E-06 850 - 1100 800

Exterior surface is extremely blackened and a black scale flaked off. Metal appears burned.

Assume wt. gain is due to carburization, if left in longer assume would be corroded further and start

to loose weight.

A36 Mild Steel 5.6719 5.837 0.1651 92.00 1.79E-03 850 - 1100 800

Exterior surface is extremely blackened and a black scale flaked off. Metal appears burned.

Assume wt. gain is due to carburization. This was the worst of all the materials.

304L 16.1013 16.0448 -0.0565 92.00 -6.14E-04 850 - 1100 800

Exterior surface is extremely blackened and a black scale flaked off. Metal appears burned.

Assume wt. gain is due to carburization.

Al2O3 5.1822 5.1822 0 336 0.00E+00 850 - 1100 800 Appears the same as when put into the furnace no noticeable damage.

LaPO4 (2%PVA) 0.4965 0.4983 0.0018 336 5.36E-06 850 - 1100 800 Appears the same as when put inot the furnace no noticeable damage.

C22 9.9308 9.9321 0.0013 336 3.87E-06 850 - 1100 800 Blackened surface

A36 Mild Steel 5.837 6.2244 0.3874 336 1.15E-03 850 - 1100 800 Blackened surface, with a swollen outer layer that dusts off

304L 16.0448 16.0438 -0.001 336 -2.98E-06 850 - 1100 800 Blackened surface

These samples did not see a large change or carburization as did the previous samples. The

conditions seemed to track the lower rates of the furnace.

All Samples were placed in the furnace again after cooled, cleaned and weighed.

Samples below were held in a canister with a 1" hole facing the top of the duct, instead of an open tray.

All Samples were placed in the furnace again after cooled, cleaned and weighed.

AISI 316 High Temperature Acid

A series of tests were conducted on AISI 316 in the reflux unit. It was the original

intention to test the AISI 316 coupons as the other materials were tested in boiling

sulfuric acid. This did not occur. The heating element did not function properly and

heated the acid to just below boiling temperature. Since the acid was not boiling the tests

were not part of the controlled experiment and were not included with the results.

71

Table 11. Corrosion information on the AISI 316 immersed in high temperature,

but not boiling 99 wt. % H2SO4

Initial Final Difference

Coupon Mass (g) a b c t in2

cm2

Mass (g) Mass (g) Hours g / h / cm2 % Diff.

3 8.1991 0.976 0.544 0.497 0.133 1.2756 8.2299 7.905 -0.2941 29.00 -1.23E-03 -3.586979 rolling boil

4 9.9541 1.005 0.6395 0.5671 0.13 1.4739 9.5092 9.9465 -0.0076 5.00 -1.60E-04 -0.07635 Simmer

5 6.9088 0.98 0.453 0.44 0.131 1.1319 7.3026 6.836 -0.0728 32.17 -3.10E-04 -1.053729 Simmer

6 8.4256 0.979 0.547 0.543 0.13 1.3217 8.5268 8.403 -0.0226 12.00 -2.21E-04 -0.26823 Simmer

INCHES Surface

-5.00E-04

-4.50E-04

-4.00E-04

-3.50E-04

-3.00E-04

-2.50E-04

-2.00E-04

-1.50E-04

-1.00E-04

-5.00E-05

0.00E+00

0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00

Time (HOURS)

WT

. L

oss

g/h

/cm

2

Figure 35. Corrosion rate of AISI 316 in high temperature 99 wt. % H2SO4.

The corrosion rates of the stainless steel 316 in high temperature concentrated

sulfuric acid are much greater than the rates of the Hastelloy by over two orders of

magnitude. Based on these results it was determined that further testing was not required.

72

A-3 SEM/EDS

Samples to be Tested

Table 12. Details of coupons SEM/EDS analyzed

Material Condition State Test LaPO4 Purchased,

Calcined Virgin Powder SEM-EDS

LaPO4 Purchased,

Sintered Virgin Pellet SEM-EDS

LaPO4 (2% PVA)

synthesized, calcined Virgin Powder SEM-EDS

LaPO4 (2% PVA)

synthesized, Microwave

Sintered Virgin Pellet SEM-EDS

C22 Hastelloy As Received Coupon SEM-EDS

C22 Hastelloy

99% H2SO4 Boiled 24

hours Coupon SEM-EDS

C22 Hastelloy

99% H2SO4 Boiled 72

hours Coupon SEM-EDS

C22 Hastelloy

90% H2SO4 Boiled 24

hours Coupon SEM-EDS

C22 Hastelloy

90% H2SO4 Boiled 72

hours Coupon SEM-EDS

Methods of Analysis:

EDS:

This technique is used in conjunction with SEM and is not a surface science

technique. An electron beam strikes the surface of a conducting sample. The energy of

the beam is typically in the range 10-20keV. This causes X-rays to be emitted from the

contact point with the material. The energy of the X-rays emitted depends on the material

under examination. The X-rays are generated in a region about 2 microns in depth, and

thus EDS is not a surface science technique. By moving the electron beam across the

material an image of each element in the sample can be acquired in a manner similar to

73

SAM. Due to the low X-ray intensity, images usually take a number of hours to acquire.

Elements of low atomic number are difficult to detect by EDS. The SiLi detector (see

below) is often protected by a Beryllium window. The absorption of the soft X-rays by

the Be precludes the detection of elements below an atomic number of 11 (Na). In

windowless systems, elements with as low atomic number as 4 (Be) have been detected,

but the problems involved get progressively worse as the atomic number is reduced.

SEM:

The scanning electron microscope (SEM) is a type of electron microscope capable of

producing high-resolution images of a sample surface. Due to the manner in which the

image is created, SEM images have a characteristic three-dimensional appearance and are

useful for judging the surface structure of the sample. Using the SEM the following

properties can be discovered.

• Fracture Surfaces

• Intergranular corrosion

• Grain surface

74

Sample Preparation

The C-22 alloy is conductive and the coupons are small enough that they can be

test as is. The alloy will be tested for micro chemical analysis. The powders must be

single layered onto a carbon or copper tape so that during the EDS scan the powder is not

disrupted.

75

A-4 Recommendations for Future Work

Lanthanum Phosphate

It was the intentions of this study to determine if lanthanum phosphate would be

an improved material for composing the sulfuric acid injection nozzles. Based on the

boiling acid test it was determined that the lanthanum phosphate as a bulk ceramic would

not be a good material for the task. The study did not conduct an in-depth analysis of

what was causing the material to degrade in the particular corrosive environment since it

was deemed unusable. Although the material did not work for this application it can be

applied to many useful purposes that do exhibit corrosive environments and an in-depth

study on the corrosion mechanism would be useful. The following outline suggests the

next steps in determining the cause of the corrosion in a highly acidic environment.

Analyze lanthanum phosphate ceramic samples that have been pressed and sintered as

follows.

• XRD - Surface and cross-sectional

• XRD – During a heat up and cool down

• SEM – Surface and cross-sectional

• EDS – Surface and cross-sectional

These tests should be conducted in a manner that allows the details of the bulk and

grain boundaries to be determined. The details of concern would be the structures,

phases, compositions, purities and interactions with each other.

Once this has been determined the samples can be exposed to the boiling acid tests

76

for different exposure times. Reagent grade sulfuric acid and high purity water should be

used in volumes that would allow later analysis of the acid. This would determine what

is being dissolved in the acid. The sample in bulk and dissolved form should be

reclaimed. The reclaimed samples must be washed repeatedly with high purity water and

vacuum dried. The samples if not washed properly will absorb the moisture in the air and

excrete an acidic liquid.

Post analysis would be conducted as described in the pre-analysis. If the same

details are defined for the bulk and grain boundaries of the ceramic at the different

exposure times the mechanism for corrosion should be able to be determined. This

should assist in determining what is depleting the phosphate group and further

transformation to a La, O, and S composition.

77

A-5 Alloy Composition

Table 13. Industrial utilized alloy compositions for nozzles

Alloy Ni Mo Fe Co Cr Mn Si C V W S Cu P Cb B2 69 28 2 1 1 1 0.1 0.01 0.0 0.0 0 0 0 0

C276 57 16 5 2.5 16 1 0.08 0.01 0.35 0.0 0 0 0 0

C22 56 13 3 2.5 22 0.5 0.08 0.01 0.35 3 0 0 0 0

Alloy 20 35 2.5 34.5 0 20 2 1 0.07 0 0 0.35 3.50 0.045 1

AISI304 8 0 71.3 0 17.5 2 1 0.08 0 0 0.03 0 0.045 0

AISI316 12 2.5 65.4 0 17 2 1 0.03 0 0 0.03 0 0.045 0

Alloy Comparison

0 10 20 30 40 50 60 70 80

Ni

Mo

Fe

Co

Cr

Mn

W

Cu

Ele

men

ts

Percent

SS316

SS304

Alloy 20

C-22

C-276

B-2

Figure 36. Element composition of the alloys considered.