Development of Rh Catalysts for Steam Methane Reforming
1. i
UNIVERSITY OF CAPE TOWN
DEPARTMENT OF CHEMICAL ENGINEERING
Development of Rhodium Catalysts for
Steam Methane Reforming
CHE4045Z
Project 33 Report
Prepared by Students:
Sibusiso Msani
Sikhulile Ntaka
Prepared for Supervisors:
Dr Olaf Conrad
Dr Peter Malatji
Walter Bohringer
PLAGIARISM DECLARATION
I know that plagiarism is wrong. Plagiarism is to use another’s work and pretend that
it is my own.
I have used the Harvard system for citation and referencing. In this report, all
contributions to, and quotations from, the work(s) of other people have been cited
and referenced.
This report is my own work. I have not allowed, and will not allow, anyone to copy
my work
Sibusiso Msani: _____S. Msani____________
Sikhulile Ntaka: _______S. Ntaka__________
Dated _______23 September 2013______
2. ii
TABLE OF CONTENTS
Page
TABLE OF CONTENTS ii
LIST OF ILLUSTRATIONS v
SYNOPSIS ............................................................................................................. 1
INTRODUCTION........................................................................................ 21
1.1 Background 2
1.2 Problem Statement 2
1.3 Scope of Study 2
1.4 Aims and Objectives 2
1.5 Key Questions 2
LITERATURE REVIEW ............................................................................. 42
2.1 Hydrogen Production Methods 4
2.1.1 Steam Hydrocarbon or Alcohol Reforming 4
2.1.2 Steam Solid-Feedstock Gasification 5
2.1 Catalysts for SMR 5
2.1.1 Nickel catalysts 5
2.1.2 PGM based catalyst 5
2.1.3 Rh-Ni bimetallic catalysts 5
Poisoning 6
Coke and Carbon Formation/Fouling 6
2.1.4 Measures used to supress coking 6
2.1.5 Effect of Support on Catalyst Activity 7
HYPOTHESIS............................................................................................ 83
METHOD OF INVESTIGATION................................................................. 94
4.1.1 Preparation of catalyst supports 9
4.1.2 Preparation of Catalysts 10
4.1.3 Physico-Chemical catalyst characterisation 11
4.1.4 Catalytic testing 14
RESULTS ................................................................................................ 165
5.1 Characterisation of Catalyst Supports 16
5.1.1 XRD 16
5.1.2 BET 17
5.2 Characterisation of Catalysts 18
5.2.1 XRD 18
3. iii
5.2.2 BET 18
5.2.3 ICP 19
5.2.4 TPR profiles of catalysts 19
5.2.5 H2 Chemisorption 20
5.3 Catalytic Testing 20
20
DISCUSSION........................................................................................... 256
6.1 Characterisation of Catalyst Supports 25
6.1.1 XRD 25
6.1.2 BET 25
6.2 Characterisation of Catalysts 25
6.2.1 XRD 25
6.2.2 BET 25
6.2.3 ICP 25
6.2.4 Catalyst TPR 26
6.2.5 H2 Chemisorption 26
6.3 Catalytic Testing 26
6.3.1 Activity and Hydrogen Selectivity 26
6.3.2 Metal Oxide (MgO) Effects 27
6.3.3 Catalyst Surface Area Effects 27
6.3.4 Particle Size Effects 28
CONCLUSIONS....................................................................................... 297
7.1 Catalyst Supports and Catalyst Preparation 29
7.2 Catalyst Activity 29
7.3 Catalyst Selectivity 29
RECOMMENDATIONS FOR FUTURE WORK ....................................... 308
NOMENCLATURE................................................................................... 319
REFERENCES......................................................................................... 3210
APPENDIX I: PRECIPITATION REACTIONS FOR CATALYST SUPPORT11
PREPARATION.................................................................................................... 34
APPENDIX II: SAMPLE CALCULATIONS.............................................. 3512
APPENDIX III: CATALYST TESTING DATA .......................................... 4013
APPENDIX IV: H2 CHEMISORPTION ANALYSIS METHOD.................. 4414
APPENDIX V: TPR ANALYSIS METHOD .............................................. 4515
APPENDIX VI: STANDARD ANGLO RISK MATRIC.............................. 4616
4. iv
APPENDIX VII: SAFETY, HEALTH AND ENVIRONMENT..................... 4717
APPENDIX VIII: TIMELINES ................................................................... 5018
5. v
LIST OF ILLUSTRATIONS
List of Tables
Table 1: Catalyst types and their Rh compositions..................................................10
Table 2: BET data for the metal oxide supports ......................................................17
Table 3: BET data for Rh catalysts on different metal oxide supports .....................18
Table 4: ICP analysis results for the metal oxide supports......................................19
Table 5: H2 Chemisorption results for Rh catalysts.................................................20
Table 6: The table shows the theoretical MgO content of each of supports based on
the molar mass. .......................................................................................................21
Table 7: Summary of factors affecting catalyst performance...................................24
Table 8: Recipe for catalyst support preparation .....................................................35
Table 9: Recipe for catalyst preparation..................................................................35
Table 10: Rh/Al2O3 catalyst testing data .................................................................40
Table 11: Rh/MgAl2O4 catalyst testing data.............................................................41
Table 12: Rh/MgO-MgAl2O4 catalyst testing results...............................................42
Table 13: Rh/MgO catalyst testing ..........................................................................43
Table 14: Standard Anglo Risk Matric .....................................................................46
Table 15: Hazard inventory pertaining to equipment ...............................................47
Table 16 : Hazard inventory pertaining to chemicals...............................................48
Table 17: September 2013 time lines......................................................................50
Table 18: October 2013 time lines...........................................................................51
Table 19: November 2013 timelines........................................................................51
List of Figures
Figure 1: Fuel processing for fuel cells using methane as a fuel...............................5
Figure 2: The TPR profiles of rhodium on different supports.....................................8
Figure 3: Procedure for support preparation.............................................................9
Figure 4 Procedure for catalyst preparation............................................................10
Figure 5: SMR schematic for catalyst testing..........................................................14
Figure 6: The effect of temperature, pressure and S/C ratio on methane equilibrium
conversion ...............................................................................................................15
Figure 6: XRD results for MgO................................................................................16
Figure 7: XRD results for MgAl2O4..........................................................................16
Figure 8: XRD results for MgO-MgAl2O4.................................................................17
Figure 9: XRD results for rhodium supported catalysts...........................................18
Figure 10: This is a plot of the reduction profiles for the four catalysts prepared. ...19
6. vi
Figure 11: CH4 conversion on different Rh supported catalyst.
Figure 12: H2 selectivity obtained for the various rhodium catalysts. ......................21
Figure 13: This is a figure which shows how the how the conversion varies with
varying amount of MgO added to the support..........................................................22
Figure 14: This is a plot of the activity per metallic surface area for each of the four
catalysts prepared. ..................................................................................................22
Figure 15: The active metallic surface area is plotted against conversion. .............23
Figure 16: A plot of the average particle size and conversion.................................23
Figure 17: The plot of the conversion per active metallic surface area and the
average particle size................................................................................................24
Figure 18: GC calibration curve for determining H2 number of moles .....................37
Figure 19: GC calibration curve for determining H2 number of moles .....................38
Figure 20: GC calibration curve for determining H2 number of moles .....................39
7. 1
SYNOPSIS
Polymer electrolyte fuel cells (PEFCs) are the leading candidates in clean electrical
energy supply for stationery applications or transportation utilisation. The fuel for
these cells is hydrogen. Steam methane reforming (SMR) is by far the most viable
method for hydrogen production in the short to medium term while other hydrogen
production methods are developed. The key challenge is developing a catalyst that
is economical, highly active, coke resistant and selective to hydrogen production. In
industry, nickel based catalysts on aluminium oxide supports are used for SMR but
have a draw-back due to coking. The problem of coking limits the consideration of
nickel based catalysts as catalysts for hydrogen production for fuel cells to power
portable devices. Bimetallic catalysts containing nickel and a noble metal, such as
rhodium, have shown to overcome this disadvantage, however; there is still a need
for further research. For bimetallic catalysts it is crucial that there is interaction
between Rh and Ni in order for the catalysts to be resistant to coking and thus
further enhance stability and performance of the catalyst. The combination of acidic
and basic supports such as magnesium oxide has also shown to reduce coking. The
aim of this investigation is to determine the effect of four different metal oxide
supports on performance of Rh based catalysts. It is hypothesised that adding a
basic component to the support of the Rh catalyst will improve performance during
SMR.
The investigation included preparation of the following supports without MgO and
with increasing amount of MgO: Al2O3, MgAl2O4, MgO-MgAl2O4, and MgO by co-
precipitation method. Rhodium salts were then impregnated onto the calcined
supports by incipient wetness impregnation method. Catalyst carriers and resulting
catalysts were characterized by conventional techniques. Prepared catalysts were
then evaluated for the SMR reaction using a fixed bed reactor at 700 o
C, 1 barg, S/C
of 3, GHSVwet of 200 000/h. The catalysts were evaluated based on CH4 conversion
and H2 selectivity.
8. 2
INTRODUCTION1
1.1 Background
Fuel cells are a viable option for clean electrical energy generation. Hydrogen is
used as a fuel and the only emission from the fuel cell is water. There is a variety of
fuel cell types namely:
proton-exchange membrane (PEM) fuel cells
alkaline fuel cells (AFC)
phosphoric acid fuel cells (PAFC)
solid oxide fuel cells (SOFC)
PEM fuel cells are considered the best energy source for residential and trans-
portable applications because of low operating temperatures (80 -100 °C), sustained
operation at high current density, low weight and compact-ness. Furthermore, they
have a potential for low cost, long stack life, fast start-ups and suitability to
discontinuous operation (Ghenciu, 2002).
Hydrogen as the feedstock for fuel cells can be produced via a multitude of
methods, however, SMR has been found to be superior to other available
techniques, such as electrolysis, gasification of solid feedstocks and steam
reforming of long chain hydrocarbons or alcohols. This is due to high hydrogen yield,
high efficiency and the abundance of natural gas despite relatively high production
cost. Processes involved in hydrogen production or fuel processing for fuel cells are
as follows: SMR, Water Gas Shift (WGS), and Preferential Oxidation (PrOx) of CO
or Selective Methanation (SelMeth).
1.2 Problem Statement
The catalyst used in the SMR step has a significant role on the viability of PEM fuel
cells. The catalyst has primarily been nickel based, however these type of catalysts
are plagued by coking. This can be remedied by using platinum metal groups
(PGMs), unfortunately, precious metals are costly. Therefore their use in a reformer
should be minimal to reduce the final cost of reformer-fuel cell technology.
Consequently, there has been a shift to using Ni-PGM bimetallic catalysts nickel,
which reduces the amount of PGMs needed. Metal oxide supports containing
magnesium oxide have also been found to supress coke formation during SMR
processing thus improving catalyst lifetime.
1.3 Scope of Study
This study will consider only rhodium based monometallic catalysts on four different
supports without MgO and with increasing amount of MgO.
1.4 Aims and Objectives
The aim is to investigate the performance of various Rh based catalysts on four
different supports with acidic or basic characteristics in SMR reaction. The catalysts
will be ranked according to performance i.e. selectivity and activity (CH4
conversion).
1.5 Key Questions
The key questions are centred around the effect that the acidity/basicity of the
catalyst support has on the catalyst performance. Furthermore, if the addition MgO
9. 3
is favourable to catalyst performance, in what form must it be present in the bi-oxidic
carrier?
10. 4
LITERATURE REVIEW2
2.1 Hydrogen Production Methods
Hydrogen is considered a future carrier of energy for transportation and stationary
power systems. Following is an in-depth discussion of the available hydrogen
production methods. Steam is used in one of two ways to produce hydrogen; by
steam coal gasification or steam hydrocarbon or alcohol reforming.
2.1.1 Steam Hydrocarbon or Alcohol Reforming
In this process the natural gas reacts with steam to form H2 and CO at temperatures
in the range 700-900 °C. H2 production via the use of natural gas has a high
efficiency of ca. 72% (Farrauto, 2003). Impurities such as sulphur compounds in
natural gas require pre-treatment upstream SMR. This process yields a H2, and CO
as main products with CO2 as the by product. The cost of producing H2 depends
mainly on the feed stock used for example, natural gas or biogas and scale of
production. SMR has been considered the best because of its high H2 yield, natural
gas abundance and high efficiency.
Fuel processing consist of three main steps, their discussion from (Farrauto et al.,
2003) is presented below with the summary shown in Figure 1.
Steam Methane Reforming (SMR)
Methane is reacted with water at process temperatures of 700-900 °C via the
following endothermic reaction: ΔHo
= 50 kJ/mol
CH4 + H2O CO +3 H2 (1)
In order to achieve high H2 production a low H2O to C ration is used, preferably 1.5
H2O/C. This also helps in minimizing the reverse WGS.
Water Gas Shift
The SMR product is fed into the WGS and this stream is known to contain ca. 12%
CO. The CO is reacted with H2O via the following exothermic reaction: ΔHo
= -44
kJ/mol
CO + H2O H2 + CO2 (2)
This reaction proceeds in two steps; low-temperature shift (LTS) and high-
temperature shift (HTS).Most CO is converted in the HTS and with less than 1% CO
sent to the LTS.
Preferential Oxidation of CO
The exit stream from the WGS containing 0.1-1.0 % CO is fed into the PROX
process to convert CO into CO2 since it poisons the catalyst in PEM cells. The CO
exit concentration is less than 10 ppmvol. The reaction proceeds via reaction (3):
ΔHo
=-285 kJ/mol
2CO+O2 2CO2 (3)
11. 5
SMR
CH4+H2O CO+3H2
ΔHo
=50 kJ/mol, 700-800o
C
Ni/Al2O3
Water Gas Shift
CO+H2O CO2+H2
ΔHo
=-44 kJ/mol
HTS: 400 o
C, Fe,Cr
LTS: 200 o
C,Cu,Zn
Preferential Oxidation of CO
2CO+O2 2CO2
ΔHo
=-285 kJ/mol
140-200 o
C, Pt
Fuel Cell
Figure 1: Fuel processing for fuel cells using methane as a fuel (Adapted from:
(Farrauto et al., 2003))
2.1.2 Steam Solid-Feedstock Gasification
This process involves the steam gasification of coal or biomass at 1000 o
C
producing syngas, which is H2 and CO, this is due to a stoichiometric shortage of O2
and the presence of enough steam during the reaction (van Dyk et al., 2005). Sasol
Company in South Africa has adopted this process with their Sasol-Lurgi fixed bed
gasification and they produced ca. 28 % of total amount of hydrogen produced
worldwide (SFA Pacific for the U.S. Department of Energy, 2000).
2.1 Catalysts for SMR
Steam methane reforming catalysts can be classified as follows:
- Nickel/nickel oxide or cobalt compositions on an aluminium oxide (Al2O3) support
or supports such as magnesium aluminate (MgAl2O4) or magnesia-alumina (MgO-
Al2O3). Alkali & alkali earth compounds are incorporated mainly to accelerate carbon
removal.
- Precious metals (Rh, Ru, Pt, Pd, Re) on alumina or rare earth oxide mostly ceria
(CeO2) (Ghenciu, 2002).
2.1.1 Nickel catalysts
Nickel based catalysts over Al2O3 supports with a nickel loading from 20 wt% are
traditionally the catalysts used for steam methane reforming. These catalysts are
less active than the noble based catalysts and are susceptible to coking and
oxidation (Zeppieri et al., 2010). They are preferred industrially where the process of
syngas production is continuous and also because of their low cost, Fuel cell
application requires stop-start process, therefore nickel based catalysts are not
suitable as catalyst for fuel processing. The acidic support is important in the
decomposition of methane however it also promotes carbon formation (de Miguel et
al., 2012).
2.1.2 PGM based catalyst
Precious metal catalysts are characterised by higher activity than nickel catalysts
per active surface area and are less affected by deactivation due to coking (Frusteri
et al., 2004). The main disadvantage with the use of these catalysts is the high
costof precious metals. According to (Vaidya et al., 2006) the order of performance
of alumina-supported noble metal catalysts (5% metal loading) for bio-ethanol
oxidative steam reforming follows the sequence ,Rh>Ru>>Pd>Pt. In particular Rh
on Al2O3 or MgO as a support shows high activity and selectivity (Ferrauto et al.,
2003).
2.1.3 Rh-Ni bimetallic catalysts
Bimetallic catalysts were introduced in order to reduce the high cost associated with
using only Rh as the active metal. It was observed in (Ferrandon et al., 2010) that
for both steam and autothermal reforming of n-butane, Ni-Rh supported on La-
Al2O3 catalysts with low loading of rhodium performed better than Ni or Rh alone at
12. 6
the same metal loadings. A Rh:Ni atomic ratio as low as 1:100 showed higher
conversion and higher H2 yields for the steam reforming of n-butane than on Rh
catalysts alone.
Factors Affecting Catalyst Activity
Catalyst deactivation occurs in one of three ways; that is poisoning, sintering and
coking/fouling.
Sintering
Catalysts are prone to sintering. This is when the active metal particles on the
catalyst begin to grow into larger particles. Sintering adversely affects the catalyst by
influencing activity. It reduces the surface area available for reaction (Sehested,
2006).
Poisoning
This is the strong chemisorption of reactants, products or impurities on the surface
otherwise available for catalysis (Bartholomew, 2001).
Coke and Carbon Formation/Fouling
Coke is formed through two ways: gas phase reactions which result in formation of
carbonaceous intermediates that condense on the surface of the catalyst. This
occurs at high temperatures and accounts for a small portion of the coking which
occurs. It is more significant when heavier hydrocarbons are used. In the second
mechanism hydrocarbons dissociate on the metal surface and produce highly
reactive atomic carbon species, some of which are gasified while others form coke
(Trimm, 1997). These phenomena result in catalyst deactivation by blockage of
active sites and/or pores (Bartholomew, 2001). There is often a distinction made
between carbon formation and coking and this is related to their origin. Carbon is a
product of CO disproportionation whereas coking is due to condensation or
decomposition of hydrocarbons on the catalyst surface. These hydrocarbons
typically consist of polymerized heavy hydrocarbons.
2.1.4 Measures used to supress coking
Addition of alkaline supports
Little is understood about the chemistry behind the use of alkaline supports for
minimisation of coke formation; however, they are used for catalyst activity
improvement and to facilitate coke gasification. Alkali promoters, such as,(Magnesia
(MgO) and Potassia (KO) based materials) are usually used. The basic character of
MgO inhibits the formation of coke (Ghenciu, 2002). The disadvantage of these
types of supports is deactivation due to pore blocking at high temperatures
(Ghenciu, 2002). . A glassy surface layer KOH, it the case of Potassia, is formed
which prevents access to active sites (Clarke et al., 1997).
A study of a M/MgO (M = Ni, Pd, Rh or Co) steam reforming catalysts showed that
coke formation is drastically inhibited on the Rh/MgO catalyst and is considerably
less when compared to coke formation on an acidic support such as Al2O3. These
catalysts were tested at 650 °C with GHSV = 40000 h−1
.The Rh/MgO catalyst
performed best in terms of activity and stability however it displayed poor hydrogen
production (Frusteri et al., 2004). Rh/MgO was the most resistant to coking. The
Ni/MgO catalyst showed the highest selectivity to hydrogen formation. The higher H2
selectivity observed when using nickel when compared to Rh based catalyst as the
13. 7
active metal can possibly be attributed to the high WGS activity in Ni than in Rh
catalysts.
Addition of rare earth metal oxides
Adding oxides of rare earth metals such as Ce or Zr to the support has been found
to reduce coking significantly at the same time improving catalysts performance for
steam reforming marginally (Trimm, 1997).
2.1.5 Effect of Support on Catalyst Activity
A study was done by (Mizuno et al., 2003), which investigated the effect of six
different supports; CeO2, Al2O3, TiO2, SiO2, ZrO2 and MgO on the activity and
selectivity of a rhodium based catalyst. The catalysts were tested for steam
reforming of 2-propanol.
It was found that the supports affects the reducibility of rhodium oxide (Rh2O3) which
further affects the catalytic stability of the rhodium catalyst for carbon dioxide
reforming and partial oxidation of methane. The reducibility of a catalyst was
examined using Temperature Programmed Reduction (TPR) technique. Rhodium
oxide interaction with the support was shown by the temperature at which the Rh(III)
rhodium oxide was reduced. The peaks between 400 and 420 K represent the
reduction of rhodium from Rh(III) to Rh. Multiple peaks were observed where there
is significant interaction between the support and the active metal, as shown in
Figure 2. The TPR analyses for the catalysts showed reaction between the active
metal and the support for the CeO2 and MgO catalyst. In the case of MgO the peaks
were attributed to the reduction of Rh(III) in Rh2O3/MgO and in MgRh2O4 spinel
(rhodium oxide in solid solution with MgO). The formation of the MgRh2O4 spinel
could be attributed to the temperature at which the catalyst was calcined which was
500 o
C.
14. 8
Figure 2: The TPR profiles of rhodium on different supports (Adapted from Mizuno
et al., 2003)
It was also found that the catalytic activity of the rhodium on acidic supports: Al2O3,
SiO2, ZrO2 and TiO2, was different. Furthermore catalytic performance of the basic
MgO support and the acidic ZrO2 were similar and thus it was concluded that
support type has no effect on the catalytic activity in steam reforming of 2-propanol.
The same was found to be true in a study by (Ligthart et al., 2011) which found that
the various supports used (ZrO2, CeO2, CeZrO2 and SiO2) had no effect on the
intrinsic activity in steam methane reforming. The support only affected the catalytic
activity by indirectly by influencing the dispersion and degree of reduction of the
metal phase. This researcher found there was a linear relationship between the
particle metal dispersion and the intrinsic activity. Furthermore the dispersion is
influenced by particle size. A smaller particle size results in a higher dispersion.
HYPOTHESIS3
It is hypothesised that:
Adding a basic substance such as MgO to the support of the monometallic
catalyst will improve the catalyst performance.
15. 9
METHOD OF INVESTIGATION4
The investigation includes catalyst support preparation, catalyst preparation and
characterization, and finally testing the catalyst in SMR.
4.1.1 Preparation of catalyst supports
The following four catalyst supports were prepared using the precipitation method:
Al2O3, MgAl2O4, MgO-MgAl2O4 and MgO. As an example, the reaction that takes
place during co-precipitation of suitable metal precursors to form MgAl2O4 is
represented by reaction (1) and (2). The other catalyst support precipitation
reactions can be found in Appendix I.
Mg(NO3)2 + 2 Al(NO3)3 + 8 NH4OH 8 NH4NO3 + Mg(OH)2(s) + 2 Al(OH)3(s) (1)
Mg(OH)2(s) + 2 Al(OH)3(s) calcination MgAl2O4 (2)
Detailed support preparation procedure
Figure 3 is the flow diagram showing the steps in the support preparation. The metal
nitrates, magnesium nitrate and aluminium nitrate (supplied by Sigma-Aldrich) were
dissolved in deionised water. The pH of the solution was measured. The metal
nitrates in the solution was precipitated as their respective metal hydroxides using
5M solution of ammonium hydroxide (supplied by Kimix). Ammonium hydroxide was
added until the pH of the slurry was 9. This was to ensure that pH of minimum
solubility of metal hydroxides is reached for complete all precipitation of the metal
salts. The support was dried initially in the 60 o
C for 1h and then at the oven prior to
the 120 o
C for 16 h. The support was calcined at 800 o
C ramping at 1 C/min from
30 C to 800 C and holding at this temperature for 1 h. The calcination temperature is
higher than the conditions under which the catalyst was be tested. This ensured that
the catalyst would be stable in the conditions tested. The support was then crushed
and sieved to particle size of less than 125 microns.
Figure 3: Procedure for support preparation.
16. 10
4.1.2 Preparation of Catalysts
The catalysts which were prepared are shown in Table 1. The catalysts were
prepared using the incipient wetness impregnation method.
The catalysts which were prepared are shown in Table 1. The catalysts were
prepared using the incipient wetness impregnation method. Figure 4 shows the
procedure for catalyst preparation. The pore volume of the support was determined
using BET. An amount of water equivalent to approximately double the pore volume
of support was added. This was found to be sufficient for incipient wetness to occur.
The mixture was evacuated to remove air pockets from the support to allow the salt
solution to get into the pores. Sonication was done depending on the wetness of the
mixture formed. The Rh/Al2O3 formed slurry thus it was sonicated. The sonication
step ensures even distribution of the catalyst as the air in pores is further removed.
The drying process is the same as that of the catalyst support.
Table 1: Catalyst types and their Rh compositions
Catalyst wt % Rh (intended)
1 Rh/MgO 1
2 Rh/MgO-MgAl2O4 1
3 Rh/MgAl2O4
Ni/MgAl2O4
1
4 Rh/Al2O3 1
Figure 4 shows the procedure for catalyst preparation.
Figure 4 Procedure for catalyst preparation.
17. 11
4.1.3 Physico-Chemical catalyst characterisation
The catalysts were characterized by the following methods:
X-Ray Diffraction (XRD)
The data was obtained from Diffraktometer D8, BRUKER. The excitation source was
Co Kα electrode. This method is used to analyse the crystal structure of the catalyst
supports. The X-rays are passed through the solid catalyst which in turn shows a
diffraction pattern. The diffraction pattern is then used to determine each support
type based on their crystal structure. The sample was ground down to particles of
0.002 mm to 0.005 mm cross section (Scintag, 1999).
BET
This technique is used to determine the material surface area and pore volume by
measuring the N2-multilayer adsorption as a function of relative pressure using a
fully automated analyser; obtaining a N2-adsorption isotherm described by equation
(4) (Ceram, 2013).
( ( ) ) ( )
(4)
Equation (4) can be linearized to for equation (5):
( )
(5)
From equation (5) it can be seen that the isotherm form a straight line. The slope
and the intercept are used to estimate the volume of N2 adsorbed to form a
monolayer. This equation is applicable at a pressure range of 0.05-0.3 to determine
Vmonolayer with the value of c typically between 50 and 300 when using nitrogen at 77
K (van Steen, 2013).
Using the volume of the gas adsorbed, the number of moles of N2 can be
determined using the ideal gas law: 1 mol gas occupies 22.4 dm3 at STP. The BET-
surface area can be estimated by using equation (6).
( ) (6)
Mean area per molecule
M Molecular weight
NAV Avogadro’s number
ρ Density of liquid adsorbant
18. 12
Temperature Programmed Reduction (TPR)
Autochem 2010, Micrometrics Inc. was used to obtain TPR profiles of the catalysts.
By this technique the influence of the support material, pre-treatment procedure and
application of promoters on the reducibility of the metal oxides was investigated
(Monti and Baiker, 1983). Effects will show a shift in the reduction temperature (peak
position) for the individual metal oxides.
The following TPR conditions were set for the analysis; these are changed
automatically through analysis.
1. Gas: (5% H2-Ar), Flow Rate: 50ml/min Change Gas Flows. Gas: Preparation
(Argon, Flow Rate: 10 ml/min)
2. Wait for 5.00 minutes
3. Temperature Ramp: Sample Ramp Temp: 120 °C Rate: 10.0 °C/min, hold for
60.00 min.
4. Temperature Ramp: Sample Ramp Temp: 60 °C, Rate: 10.0 °C/min, hold for 5.00
min.
5. Change Gas Flows
6. Gas: Carrier (5% H2-Ar), Flow Rate: 50 ml/min, Trap activated.
7. Wait until baseline is stable
8. Start Recording: One measurement every 1.0 seconds
9. Temperature Ramp: 900 °C, Rate: 10.0 C/min.
10. Stop Recording
11. Change Gas Flows
12. Gas: Preparation (Argon), Flow Rate: 10 ml/min
13. Temperature Ramp: Sample Ramp Temp: 80 °C, Rate: 10.0 ml/min.
14. Done: Return to Ambient, Detector: Off, Flow gas: Default Gas
Inductively Coupled Plasma (ICP) analysis
This method was used for determining the metal loading of the catalyst samples.
The catalyst sample is dissolved in acid before being introduced into the ICP
(nanoComposix, 2012).
The following procedure was followed to prepare the sample for ICP analysis; this
was done in three steps which include sample preparation, acid digestion and
sample analysis.
Sample preparation
50 mg of sample was weighed on Precisa 205 A SCS, SCALETEC analytical
scale and added into the Erlenmeyer conical flask.
19. 13
Acid digestion
6 ml of hydrochloric acid and 2 ml nitric acid was added onto the sample in
the Erlenmeyer conical flask.
The solution was left to stand for 1 hour.
After an hour, the solution was than heated on a hot plate until small amount
of acid was left.
The solution was than cooled.
6 ml of hydrochloric acid, 2ml nitric acid and 4 ml hydrofluoric acid were then
added onto the cool solution.
The solution was left to stand for an hour.
The solution was heated until a small amount of acid remains.
Final preparation and sample analysis
A 100 ml aqueous solution of 100 ml was made.
The solution was filtered with no.1 Whatman filter paper.
Sample was introduced in the ICP and metal content was analysed
H2 Chemisorption
ASAP 2020, Micrometrics Inc. H2 chemisorption data is used to calculate the metal
surface area (Kip et al., 1986). Metal particle size and dispersion was calculated
from this data.
As a measure of the H2 adsorption capacity the Langmuir isotherm with dissociative
adsorption was used to calculate the monolayer volume/chemisorbed (Vm) at normal
conditions (1 atm, 273 K), (Smeds et al., 1996) from equation (3).
( )
(7)
V adsorbed volume
Vm adsorbed monolayer normal volume (at 273 K, 1 atm)
KH overall adsorption equilibrium constant of H2
PH partial pressure for H2
Equation (3) can be linearized to estimate Vm and KH using experimental data, this
equation is written in the following form:
(8)
⁄
and
20. 14
4.1.4 Catalytic testing
Catalyst Testing Experimental Set up
(a) Methane and Argon gas cylinders, (b) Water, (c) HPLC pump, (d) Top half of the
reactor filled with silicon carbide(particle size), (e) Reactor bed with catalyst mixed
with silicon carbide (≈ 1mm average particle size). This is the isothermal zone of
reactor bed (f) Bottom half of the reactor filled with silicon carbide (particle size); (g)
Water catch-pot and (h) Online Gas Chromatography (GC).The reactor bed is
located at the isothermal zone. The stainless steel reactor tube was used and its
dimensions are 15.178 mm ID and length is 899 mm. Water is introduced as liquid
and slowly heated up along the catalyst bed and ensures that there is a constant
flow of water. The 0.1000 g catalyst is mixed will fine silicon carbide to make up 2 ml
for even size distribution across the reactor. The line from the reactor outlet to the
GC is heated at 40 C. All gases pass to the online GC connected to the computer
where all the data is processed. The excess and unreacted water is condensed to
the catch pot.
(b)
(g)
(a)
(c)
(e)
(d)
(f)
Arg CH4
(i)
Figure 5: SMR schematic for catalyst testing.
Catalytic Testing Procedure
Each catalyst was tested in a fixed bed reactor. Testing was carried out by firstly
reducing the catalyst using H2 by the following steps:
1. Catalyst was heated to 750 o
C at 1o
C/min in the flow of H2 and this temperature
was held for 3 hours.
2. The temperature was then reduced to 700 o
C which is the operating temperature
of the reaction. Water was introduced to the reactor at this stage to facilitate cooling
of the reactor bed. This took about 30 minutes.
21. 15
3. The H2 feed was switched off and it is flushed out with Argon for 5 minutes.
4. CH4 was introduced and the flow rate was set such that a S/C ratio of 3 is
obtained.
5. The reactor product was analysed using online gas chromatography.
6. The conversion of CH4 was monitored. The aim was to keep the conversion at
least 20 % below the equilibrium conversion of 97 % under the conditions
investigated by (Joensen & J. R. Rostrup-Nielsen 2002) as seen on Figure 6.
Figure 6: The effect of temperature, pressure and S/C ratio on methane equilibrium
conversion (Joensen & J. R. Rostrup-Nielsen, 2002).
22. 16
RESULTS5
5.1 Characterisation of Catalyst Supports
5.1.1 XRD
The following are the XRD patterns of the four catalyst supports prepared.
Figure 7: XRD results for MgO. The calculated peaks were obtained from the XRD
database.
Figure 8: XRD results for MgAl2O4. The calculated peaks were obtained from the
XRD database.
0 20 40 60 80 100 120
Intensity(a.u.)
2 Theta (degrees)
Experimental results Calculated (data base)
0 20 40 60 80 100 120
Intensity(a.u.)
2 Theta (degrees)
Calculated (data base) Experimental results
23. 17
Figure 9: XRD results for MgO-MgAl2O4.
5.1.2 BET
Table 2: BET data for the metal oxide supports
Units Al2O3 MgAl2O4 MgO-MgAl2O4 MgO
Pore
volume
cm3
/g 0.24 0.32 0.22 0.045
Surface
area (BET)
m2
/g 143 110 73.3 16.4
Density g/cm3
1 1 1 1
0 20 40 60 80 100 120
2 Theta (degrees)
Intensity(a.u)
MgO-MgAl2O4
MgAl2O4
MgO
24. 18
5.2 Characterisation of Catalysts
5.2.1 XRD
Figure 10: XRD results for rhodium supported catalysts. The catalyst peaks are
then matched with catalyst supports peaks.
5.2.2 BET
Table 3: BET data for Rh catalysts on different metal oxide supports
Units Rh/Al2O3 Rh/MgAl2O4 Rh/MgO-
MgAl2O4
Rh/MgO
Surface
area (BET)
m2
/g 130.5 103.5 80 29
Area of
mesopores
m2
/g 128.2 96.8 83.7 24.2
Pore
diameter
A 73.2 128.7 121.0 351.6
Pore
volume
cm3
/g 0.24 0.33 0.24 0.25
Density g/cm3
1 1 1 1
0 20 40 60 80 100 120
Intensity(a.u)
2 Theta o
MgO-MgAl2O4
Rh/Al2O3
Rh/MgAl2O4
MgAl2O4
Rh/MgO-MgAl2O4
25. 19
5.2.3 ICP
Table 4: ICP analysis results for the metal oxide supports
Catalyst Rh content actual (wt %) Intended loading (wt %)
Rh/Al2O3 0.73 1
Rh/MgAl2O4 0.93 1
Rh/MgO-MgAl2O4 0.94 1
Rh/MgO 1.03 1
5.2.4 TPR profiles of catalysts
The following results are TPR results for Rhodium catalyst on different metal oxides
supports.
Figure 11: This is a plot of the reduction profiles for the four catalysts prepared.
0 100 200 300 400 500 600 700
HydrogenConsumption(a.u)
Temperature (oC)
Rh/MgO
Rh/ MgAl2O4
Rh/MgO-MgAl2O4
26. 20
5.2.5 H2 Chemisorption
Table 5: H2 Chemisorption results for Rh catalysts
Units Rh/Al2O3 Rh/MgAl2O4 Rh/MgO-MgAl2O4 Rh/MgO
Metallic
Surface
area
m2
/g
catalyst
0.54 0.73 0.60 0.63
m2
/g
metal
54 73 60 63
Particle size nm 9.01 6.60 8.10 7.7
Metal
Dispersion
% 12.2 16.7 13.6 14.3
5.3 Catalytic Testing
Figure 12: CH4 conversion on different Rh supported catalyst. Pressure = 1 barg,
Temperature = 700 o
C, GSHVwet = 200 000 and a steam to carbon ratio of 3. A
sample calculation for CH4 conversion can be seen in Appendix II.
0
20
40
60
80
100
0 10 20 30 40 50 60 70 80
CH4conversion,%
Time-on-stream, hours
Rh/MgAl2O4 Rh/MgO-MgAl2O4 Rh/MgO Rh/Al2O3
27. 21
Figure 13: H2 selectivity obtained for the various rhodium catalysts. Pressure = 1
barg, Temperature = 700 o
C, GSHVwet = 200 000 and a steam to carbon ratio of 3.
Sample calculation for selectivity can be seen in Appendix IV. The selectivity data
for the MgO Rh/MgO catalyst was not considered for the first 10 hours.
It was investigated if the support type had an effect on the activity. Table 6 shows
the MgO content of each of the catalyst supports. Figure 13 shows the results
obtained.
Table 6: The table shows the theoretical MgO content of each of supports based on
the molar mass.
Catalyst Mg content (wt%)
MgO content (mol%)
Rh/Al2O3 0 0 0
Rh/MgAl2O4 17 0 (MgO·MgAl2O4) 50 (MgO·Al2O3)
Rh/MgO-MgAl2O4 24 50 (MgO·MgAl2O4) 66 (2MgO·Al2O3)
Rh/MgO 60 100 100
0
20
40
60
80
100
0 10 20 30 40 50 60 70 80
H2Selectivity,%
Time-on-stream, hours
Rh/MgO Rh/MgO-MgAl2O4 Rh/MgAl2O4 Rh/Al2O3
28. 22
Figure 14: This is a figure which shows how the how the conversion varies with
varying amount of MgO added to the support. The supports’ structure can be re-
written to show the MgO content in the following manner: Al2O3, MgAl2O4
(MgOAl2O3), MgO-MgAl2O4 ((2MgO)Al2O3) and MgO.
It is useful to observe the trend between the activity per active surface area and the
support type used.
Figure 15: This is a plot of the activity per metallic surface area for each of the four
catalysts prepared.
The next trend which was considered is the correlation between metallic surface
area or the active surface area and conversion. This is plotted in Figure 15.
0
10
20
30
40
50
60
70
80
90
Al2O3 MgAl2O4 MgO-MgAl2O4 MgO
Conversion(%)
Support
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
Al2O3 MgAl2O4 MgO-MgAl2O4 MgO
TOF(m-2s-1)
Support
29. 23
Figure 16: The active metallic surface area is plotted against conversion. The four
data points represent the four catalysts tested with their respective metallic surface
area and conversion.
The trend between average particle size and activity was investigated. This is shown
in figure 16.
Figure 17: A plot of the average particle size and conversion.
0
10
20
30
40
50
60
70
80
90
0.5 0.55 0.6 0.65 0.7 0.75
Conversion(%)
Metallic Surface Area (m2)
Rh/Al2O3
Rh/MgO-MgAl2O4
MgAl2O4MgO
0
10
20
30
40
50
60
70
80
90
6 6.5 7 7.5 8 8.5 9 9.5
Conversion(%)
Average particle diameter (nm)
30. 24
The trend between average particle size and activity per metallic surface area was
investigated. This is shown in figure 17.
Figure 18: The plot of the conversion per active metallic surface area and the
average particle size.
Table 7: Summary of factors affecting catalyst performance.
Catalyst
MgO
content
(mol %)
Metallic
surface
area (m2
/g
catalyst)
Metal
dispersi
on (%)
CH4
Conversion
(%)
H2
Selectivity
(%)
Rh/Al2O3 0 0.54 12.2 80 78
Rh/MgAl2O4 50 0.73 16.7 45 78
Rh/MgO-MgAl2O4 66 0.60 13.6 68 78
Rh/MgO 100 0.63 14.3 45 70
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
6 6.5 7 7.5 8 8.5 9 9.5
TOF(m-2s-1)
Average particle diameter (nm)
31. 25
DISCUSSION6
6.1 Characterisation of Catalyst Supports
6.1.1 XRD
Figure 6, 7 and 8 show the XRD patterns for claimed metal oxide supports (i.e.
MgO, MgAl2O4 and MgO-MgAl2O4). For MgO and MgAl2O4, the experimental peaks
were compared with the calculated peaks from the XRD data base to confirm if the
intended support is the one which was obtained.
On Figure 6 it can be seen that magnesium oxide structure was the only phase
existing in the material as it is a perfect match with the calculated peaks from the
data base. The large and broad peak at about 10o
is a result of amorphous
magnesium oxide. The pattern shows sharp peaks which means that the
magnesium oxide support has large crystals. Figure 7 shows the XRD pattern for
magnesium aluminate (spinel), it can be seen that the spinel is only material
existing; however there is a slight shift of peaks on the right to the peak at 42o
. This
might be as a result of the addition of the aluminium oxide. Figure 8 is the XRD
pattern for the combination of magnesium oxide and spinel. The peaks for this
compound have a perfect match with the magnesium oxide and spinel. The spinel
and magnesium oxide on the spinel have broader peaks compared to MgO and this
shows that in the samples of lower Mg/Al ratios, apparently suggesting that the size
of the MgO crystal decreased.
6.1.2 BET
Table 2 shows the BET results for the metal oxide catalyst supports. It can be seen
from Table 2, an increase in basicity of the catalyst support decreases the surface
area.
6.2 Characterisation of Catalysts
6.2.1 XRD
Figure 9 shows the XRD results for rhodium supported catalysts and the metal oxide
supports. These results show clearly that the addition of rhodium does not affect the
structure of the metal oxide support. The peaks are in the same position as the one
for the metal oxide supports.
6.2.2 BET
Table 3 show the BET results for rhodium on the metal supports. It can be seen
from Table 3, an increase in basicity of the catalyst support decreases the surface
area.
6.2.3 ICP
Table 4 shows the metal loading determined by ICP analysis. It can be seen from
these results that ICP results are slightly off compared to the intended metal loading.
This might be due that not all the salt solution was transferred onto the support
during co-impregnation. If the ICP results are accurate, Rh/Al2O3 has the lowest
loading; however it shows high activity compared to other catalyst as it can be seen
from Figure 11. This implies that the metal loading might not play a crucial role in
activity but the support type is.
32. 26
6.2.4 Catalyst TPR
TPR peaks for the different rhodium supported catalysts are shown in Figure 10.
Rh/MgO-MgAl2O4, Rh/MgO, Rh/MgAl2O4 and Rh/Al2O3 showed reduction peaks at
319 o
C, 312 o
C, 260 o
C and 122 o
C respectively. These peaks are due to the
reduction of rhodium supported catalyst. It can be observed that an increase in the
ratio of Al to Mg results in sharper peaks and the reduction temperature of rhodium
oxide is lowered. This can be seen on Rh/Al2O3 catalyst TPR plot which has zero
magnesium. This suggests the contribution of aluminium on the reduction peak. The
rhodium oxide becomes easier to reduce with the addition of the aluminium. The
reduction peak for Rh/MgO-MgAl2O3 is at a higher temperature than that of Rh/MgO
and this is due to MgO in excess in the MgO-MgAl2O4 support which decreases the
reducibility of the catalyst.
6.2.5 H2 Chemisorption
Table 5 shows the H2 chemisorption results for rhodium supported catalysts. It can
be seen that Rh/Al2O3 has the lowest metallic surface area and dispersion
compared to the other rhodium supported catalysts which implies that it has the
lowest number of active sites. A lower dispersion number result in a lower catalyst
activity but this is contrary to what is seen in Figure 11. This catalyst has the largest
average particle size. It must be noted that the average particle size is reported and
is not representative of the whole sample analysed. The particle size of a sample is
better represented using a particle distribution curve.
6.3 Catalytic Testing
6.3.1 Activity and Hydrogen Selectivity
The plot of conversion of the four catalysts shows an initial high conversion followed
by a precipitous drop in the first few hours of operation. This can be attributed to
highly active metal sites on the surface of the catalyst. These active metal sites have
a high surface energy due to low co-ordination number and tend to sinter shortly
resulting in a loss of active metal sites (Bartholomew, 2001).
The conversion plots do not follow the observed in literature. The general trend
usually noted is constant deactivation of the catalyst. The catalysts tested in this
study are characterised by an initial stage of decreasing conversion followed by an
increase until stability is reached. It is suspected that this phenomenon is brought
upon by oxidation of the rhodium on the catalyst back to its inactive oxidative state
which gets reduced by methane when the feed is introduced. This would explain the
increasing conversion with time as the active sites are being re-reduced. This
oxidation is possibly brought upon by the water which was fed initially after the
catalyst was reduced to aid in cooling. This is corroborated by the experiment done
by (Ligthart, van Santen and Hensen, 2011) where rhodium based catalyst on the
supports ZrO2, CeO2, CeZrO2 and SiO2 were tested in the SMR reaction. The
deactivation which occurred was attributed to oxidation of very small particles
(particles < 2.5 nm). The experiment was run at a steam partial pressure of 0.18 bar.
The steam partial pressure used for this investigation is 1.5 bar. This is ten times
bigger and thus bigger particles might also have been oxidised.
33. 27
Activity
The Rh/Al2O3 shows the highest activity and took the shortest time to stabilise. This
was followed by Rh/MgO-MgAl2O4. The Rh/MgO catalyst did not stabilise however
Rh/MgAl2O4 and Rh/MgO showed similar performance (at the time that the Rh/MgO
catalyst run was stopped).
Selectivity
The maximum molar selectivity which can be attained from the SMR reaction is 75
% and when the WGS reaction is considered the maximum molar theoretical
selectivity increases to 80 %. The selectivity obtained for the catalysts tested is thus
expected to be between 75 and 80 %. The selectivity of the three catalysts;
Rh/MgAl2O4, Rh/MgO-MgAl2O4 and Rh/Al2O3 fall within the expected band however
the Rh/MgO shows selectivity which is below this band. This is unexpected and is
possibly due to machine error of the online gas chromatography equipment. The
SMR reaction was simulated using an Aspen equilibrium reactor. The selectivity of
hydrogen was found to be 77.6 %. This suggests that the WGS reaction in the
experiment reached equilibrium.
6.3.2 Metal Oxide (MgO) Effects
The aim of the study is to determine the effect of adding the MgO to the support of
the catalyst and thus the relationship between MgO addition to the support and the
activity was investigated. The results obtained show addition of MgO to the catalyst
support does not show an accompanied constant increase or decrease in activity.
Figure 13 shows a ziz-zag pattern with an increase in the MgO content of the
support. The activity per active metallic site which is shown in Figure 14 shows the
same trend and thus the difference in conversion in Figure 13 cannot be attribute
difference in the amount of active area of the respective catalysts. The performance
of the Rh/MgO catalyst and Rh/MgAl2O4 is similar in terms of the conversion
reached however the basic character of the supports is significantly different and
thus there is no clear relationship between MgO content in the support and catalyst
performance. The same was found to be true in (Ligthart, van Santen and Hensen,
2011) and (Mizuno et al., 2003) as mentioned in section 2.3.2. It should be noted
however that the Rh/MgO catalyst had not stabilised at the time this run was
stopped.
6.3.3 Catalyst Surface Area Effects
Figure 13 shows the relationship between the metallic surface area and the
conversion. The catalytic activity decreases with increasing metallic surface area.
This is contrary to what is expected. An increase in metallic surface area should be
accompanied by an increase in activity. In the findings of (Ligthart, van Santen and
Hensen, 2011) the intrinsic rate of reaction increased linearly with increasing surface
area, however, this is when the support type used was kept constant. The metallic
surface area was adjusted by changing the metal loading. Figure 13 suggests that
the differences in catalyst performance cannot be attributed to the difference in
metallic areas of the catalysts.
34. 28
6.3.4 Particle Size Effects
Figure 15 shows that the conversion declines slightly between the 6.6 nm and 7.7
nm followed by an increase in conversion thereafter. Figure 16, however, shows an
increase in the activity per active metal area with increasing particle diameter
throughout the particle size ranges observed. An increase in particle size is
expected to lower conversion because this reduces the surface area available for
reaction however the opposite is seen from the plot on Figure 15. This phenomenon
can be explained as mentioned in section 6.3.1. It was suggested that smaller
particles get oxidised and become inactive in the reaction. In a study done by (von
Steen et al., 2005) on the cobalt catalyst for Fischer Tropsch it was found that
particle less than 4.4 nm are likely to be oxidised under realistic Fischer Tropsch
conditions. It is thus possible that in this experiment there was a critical particle size
below which oxidising of the active metal was significant. It is possible that catalysts
with a smaller average particle size have a larger portion of their particles which are
smaller than the critical size. Thus a larger portion of the active surface area
became deactivated. This would then result in the trends observed.
35. 29
CONCLUSIONS7
7.1 Catalyst Supports and Catalyst Preparation
The XRD results show that the catalyst and catalyst supports which were prepared
contains the material they were prepared with and in the intended proportion. The
increase in basicity (that is the addition of magnesium) increases the catalyst and
catalyst support surface area as shown by BET results. An increase in Al/Mg ratio
decreases the reduction temperature. H2 chemisorption reports the average particle
size which makes it difficult to conclude on the trend as some particles. There might
be inaccuracy in the co-impregnation method as the ICP results are way off the
intended metal loading, however the error might be also on the ICP digestion
method.
7.2 Catalyst Activity
The Rh/Al2O3 is the most catalytically active catalyst for SMR in the study carried
out. The difference in performance between the four catalysts is not due the support
type used, thus there is no relationship between the support used and the activity of
the catalyst. The difference in the performance of the catalysts can be partly
attributed to the particle size effect. The catalysts with a smaller average particle
size had lower activity. The metallic surface area has a negligible influence on the
activity of the catalyst when the performance between the four rhodium based
catalysts is compared based on metallic surface area.
7.3 Catalyst Selectivity
The hydrogen selectivity is independent of the support used in the case of the three
rhodium supported catalysts Rh/Al2O3, Rh/MgAl2O4 and Rh/MgO-MgAl2O4. The
WGS reaction reached equilibrium for these catalysts thus they are also active in the
WGS reaction.
36. 30
RECOMMENDATIONS FOR FUTURE WORK8
The following recommendations have been suggested:
The spent catalyst should be analysed using XRD and TEM to determine whether
oxidation of the catalyst did occur.
The effects of metallic surface area and particle size on catalyst activity should be
studied while using the same support as done in Ligthart, van Santen and Hensen,
2011).
The start-up procedure should be adjusted such that water is not used to cool the
reactor after activation which is the proposed mechanism via which the catalyst was
oxidised.
The catalyst preparation and testing must be repeated to test for reproducibility of
results. The catalysts should be prepared in batches so as to minimise difference in
performance due to differences in preparation technique.
The catalyst testing should be done over a longer period to study other effects which
affect catalyst activity that is coking and sintering.
37. 31
NOMENCLATURE9
SMR Steam Methane Reforming
PEFC Polymer electrolyte fuel cell
PGM Platinum metals groups
XRD X-Ray Diffraction
TPR Temperature Programmed Reduction
ICP Inductively Coupled Plasma
GC Gas Chromatography
V Adsorbed volume (m3
)
Vm Adsorbed monolayer normal volume (m3
) (at 273 K, 1 atm)
KH Overall adsorption equilibrium constant of H2
PH Partial pressure of H2
TEM Transmission Electron Microscopy
BET Brunaurer-Emmett-Teller
GSHV Gas Space Hourly Velocity
38. 32
REFERENCES10
1. Bartholomew H.C. 2001. Mechanisms of Catalyst Deactivation. Applied
Catalysis A: General 212: 17 – 60
2. Ceram. 2013. BET surface area analysis and BJH pore size and volume
analysis.
Available:http://www.ceram.com/testing-analysis/techniques/brunauer-
emmett-teller-surface-area-analysis-barrett-joyner-halenda-pore-size-and-
volume-analysis/
Accessed: 03 November 2013
3. Clarke, S.H., Dicksm, A.L., Pointon, K. Smith, T.A. and Swann A. 1997.
Catalytic Aspects of the Steam Reforming of Hydrocarbons in Internal
Reforming Fuel Cells. Catalysis Today 38: 411 – 423.
4. de Miguel, N., Manzanedo, J. and Arias, P. 2012. Active and Stable Ni-MgO
Catalyst Coated on a Metal Monolith for Methane Steam Reforming under
Low Steam-to-Carbon Ratios. Chemical Engineering Technology 35: 2195 -
2203.
5. Farrauto, R., Hwang, S., Shore, L., Ruettinger, W., Lampert, J., Giroux, T.,
Liu, Y. and Ilinich, O. 2003. New Material Needs for Hydrocarbon Fuel
Processing: Generating Hydrogen for the PEM Fuel Cell.33: 1 - 27.
6. Ferrandon, M., Kropf, A.J. and Krause T. 2010. Applied Catalysis A: General
379: 121 – 128.
7. Frusteri, F., Freni, S., Spadaro, L., Chiodo, V., Bonura, G., Donato, S. and
Cavallaro, S. 2004. H2 production for MC fuel cell by steam reforming of
ethanol over MgO supported Pd, Rh, Ni and Co catalysts. Catalysis
Communications: Journal of Catalysis Communication 5: 611 – 615.
8. Ghenciu, A F. 2002. Review of fuel processing catalyst for hydrogen
production in PEM fuel cell systems. Current Opinion in Solid State and
Material Science. 6: 389 - 399.
9. Kip, B. J., Duivenvoorden, B.M., Koningsberger, D.C. and Pris R. 1986.
Determination of Metal Particle Size of Highly Dispersed Rh,Ir, and Pt
Catalyst by Hydrogen Chemisorption and EXAFS. Journal of American
Chemical Society 108: 5633 - 5634.
10.Levin, B.D., Pitt, L. and Love, Murray. 2004. Biohydrogen production:
prospects and limitations to practical application. International Journal of
Hydrogen Energy 29: 173 - 185.
11.Ligthart, D.A.J.M., van Santen R.A., Hensen E.J.M. 2011. Influence of
Particle Size on the Activity and Stability in Steam Methane Reforming of
Supported Rh Nanoparticles. Journal of Catalysis 280: 206 – 220.
12.Lipman, T. 2011. An Overview of Hydrogen Production and Storage Systems
with Renewable Hydrogen Case Studies. California, United States:
Renewable Energy Fuel Cell Technologies Program
13.Lyubovsky, M., Roychoudhury, S. and LaPierre, R. 2004. Catalytic partial
‘‘oxidation of methane to syngas’’ at elevated pressures. Catalysis 99: 113 -
117.
14.Mizuno T., Matsumura Y., Nakajima T. and Mishima S. 2003. Effect of
Support on Catalytic Properties of Rh Catalysts for Steam Reforming of
Propanol. International Journal of Hydrogen Energy 28: 1393 - 1399
39. 33
15.Monti, D.A.M., Baiker, A. 1983. Temperature-Programmed Reduction,
Parametric Sensitivity and Estimation of Kinetic Parameters. Journal of
catalysis 83: 323 - 335.
16.Ni, M., Leung, D.Y.C., Leung, M.K.H. and Sumathy, K. 2005. An overview of
hydrogen production from biomass. Fuel processing technology 87: 461 -
472.
17.Vaidya, P.D. and Rodrigues A.E. 2006. Insight into the Steam Reforming of
Ethanol to Produce Hydrogen for Fuel Cells. Chemical Engineering Journal
117: 39 – 49.
18.Scintag, Inc.1999. Basics of X-ray Diffraction.
Available: http://epswww.unm.edu/xrd/xrdbasics.pdf.
Accessed: 18 September 2013
19.Sehested, J. 2006. Four challenges for nickel steam-reforming catalysts.
Applied Catalysis Today: Geberal 111: 103 – 110.
20.Smeds, S., Salmi, T., Lindfors, L.P. and Krause, O. 1996. Chemisorption and
TPD studies of hydrogen on Ni/AI203. Applied Catalysis A: General 144: 177-
194.
21.Strohm, J.J., Zheng, J. and Song, C. 2006. Low-temperature steam reforming
of jet fuel in the absence and presence of sulfur over Rh and Rh–Ni catalysts
for fuel cells. Journal of Catalysis 238: 309 – 320.
22.Trimm, D.L. 1997. Coke Formation and Minimisation during Steam Reforming
23.van Steen, E., Claeys, M., Dry, M.E., van de Loosdrecht, J., Viljoen, E.L. and
Visagie J.L. 2005. Stability of Nanocrystals: Thermodynamic Analysis of
Oxidation and Re-reduction of Cobalt in Water/Hydrogen Mixtures. Journal of
Physical Chemistry B 109: 3575 – 3577.
24.van Steen., E. 2013. Heterogeneous Catalysis-Basic Principles, Catalyst
Synthesis and Characterization, and Industrial Applications. University of
Cape Town, Department of Chemical Engineering
25.Zeppieri, M., Villa, P.L., Verdone, N., Scarsella, M. and de Filippis, P. 2010.
Kinetic of methane steam reforming reaction over nickel- and rhodium-based
catalysts. Applied Catalysis A: General 387: 147 - 154.
41. 35
APPENDIX II: SAMPLE CALCULATIONS12
Table 7 shows the raw materials used to prepare each support and Table 8 is the
recipe for catalyst preparation.
Table 8: Recipe for catalyst support preparation
support Mass of
Al(NO3)3 (g)
Mass of
(Mg(NO3)2) (g)
Volume of
NH4OH (ml)
1 MgAl2O4 40.05 13.36 35
2 MgO-MgAl2O4 40.05 27.47 84
3 MgO - 54.90 35
4 Al2O3 40.02 - 56
Table 9: Recipe for catalyst preparation
catalyst Mass of
Rh(NO3)3 (g)
Mass of Support (g)
1 Rh/Al2O3 0.0605 1.90
2 Rh/MgAl2O4 0.064 1.91
3 Rh/MgO-MgAl2O4 0.061 1.93
4 Rh/MgO 0.06 1.98
Catalyst preparation: Rhodium requirement calculation
Let the metal composition of the catalyst to be made be Y wt % and the required
mass of the metal be X g. The mass of the catalyst support is known to be Z g.
Therefore:
=> (12)
Since the compound to be added is a metal nitrate, the mass of the compound isbe
calculated as follows:
Let the molar mass of the metal nitrate be A g/mol and the molar mass of metal be
B:
(13)
Example: For Rh/Al2O3
Mass,Al2O3 = 1.9 g
Rh composition = 1 wt %
Molar mass Rh= 102.91 g/mol
Molar mass of Rh(NO3)2
Rearranging:
42. 36
CH4 conversion calculation
The following formula was used to calculate the CH4 conversion:
(14)
Where the area is the area under the peak from the GC data; see Appendix III.
Example: For conversion after 22.36 hours of CH4 using Rh/MgAl2O4:
From Appendix III, Table 14: AreaCH4,in = 7883
AreaCH4,out = 1946.5
AreaAr,in = 193
AreaAr,out = 78.5
Therefore:
Selectivity calculation
The following formula was used to calculate the hydrogen selectivity,
: Number of moles of hydrogen
: Number of carbon monoxide
: Number of carbon dioxide
The number of moles were determined using the equations from Figure 18, 19 and
20:
Areas are obtained from Appendix III.
Example: at time 22.36 hours using Rh/MgAl2O4
= 44774.6
= 685.2
= 618.1
So =135.25 moles
43. 37
14.95 moles
22.52 moles
Figure 19: GC calibration curve for determining H2 number of moles
Area = 133.49 NH2 + 26719
R² = 0.9979
35000
40000
45000
50000
55000
60000
65000
70000
50 100 150 200 250 300 350
Area(UV/min)
Number of moles of H2
44. 38
Figure 20: GC calibration curve for determining H2 number of moles
Area = 17.356 NCO + 425.7
R² = 0.9993
600
650
700
750
800
850
900
950
1000
1050
1100
10 15 20 25 30 35 40
Area(UV/min)
Number of moles of CO
45. 39
Figure 21: GC calibration curve for determining H2 number of moles
Area = 19.028 NCO2 + 189.64
R² = 0.9997
400
500
600
700
800
900
1000
10 15 20 25 30 35 40
Area(UV/min)
Number of moles
51. 45
APPENDIX V: TPR ANALYSIS METHOD15
The following TPR conditions were set for the analysis; these are changed
automatically through analysis.
1. Gas: (5% H2-Ar), Flow Rate: 50ml/min Change Gas Flows. Gas: Preparation
(Argon, Flow Rate: 10 ml/min)
2. Wait for 5.00 minutes
3. Temperature Ramp: Sample Ramp Temp: 120 °C Rate: 10.0 °C/min, hold for
60.00 min.
4. Temperature Ramp: Sample Ramp Temp: 60 °C, Rate: 10.0 °C/min, hold for 5.00
min.
5. Change Gas Flows
6. Gas: Carrier (5% H2-Ar), Flow Rate: 50 ml/min, Trap activated.
7. Wait until baseline is stable
8. Start Recording: One measurement every 1.0 seconds
9. Temperature Ramp: 900 °C, Rate: 10.0 C/min.
10. Stop Recording
11. Change Gas Flows
12. Gas: Preparation (Argon), Flow Rate: 10 ml/min
13. Temperature Ramp: Sample Ramp Temp: 80 °C, Rate: 10.0 ml/min.
14. Done: Return to Ambient, Detector: Off, Flow gas: Default Gas
52. 46
APPENDIX VI: STANDARD ANGLO RISK MATRIC16
Table 14: Standard Anglo Risk Matric
Standard Anglo Risk Matrix
Hazard Effect / Consequence
(Where an event has more than one 'Loss Type', choose the 'Consequence' with the highest rating)
Loss Type
(Additional 'Loss Types' may exist for an event; identify &
rate accordingly)
1
Insignificant
2
Minor
3
Moderate
4
Major
5
Catastrophic
(S/H)
Harm to People (Safety / Health)
First aid case / Exposure to
minor health risk
Medical treatment / Exposure
to major health risk
Loss time injury / Reversible
impact on health
Single fatality or loss of
quality of life / Irreversible
impact on health
Multiple fatalities / Impact
on health ultimately fatal
(EI)
Environmental Impact
Minimal environmental harm - L1
incident
Material environmental harm -
L2 incident remediable short
term
Serious environmental harm -
L2 incident remediable within
LOM
Major environmental harm -
L2 incident remediable post
LOM
Extreme environmental
harm - L3 incident
irreversible
(BI/MD)
Business Disruption / Material Damage & Other
Consequential Losses
No disruption to operation /
R500k to less than R5m
Brief disruption to operation /
R5m to less than R50m
Partial shutdown / R50m to
less than R500m
Partial loss of operation
R500m to less than R5bn
Substantial or total loss of
R5bn and more
(L&R)
Legal & Regulatory
Low level legal issue
Minor legal issue; non
compliance and breaches of
the law
Serious breech of law;
investigation/report to
authority, prosecution and/or
moderate penalty possible
Major breech of the law;
considerable prosecution
and penalties
Very considerable penalties
& prosecutions. Multiple law
suits & jail terms
(R/S/C)
Impact on Reputation/Social/Community
Slight impact - public awareness
may exist but no public concern
Limited impact - local public
concern
Considerable impact - regional
public concern
National impact - national
public concern
International impact -
international public attention
Likelihood
Examples
(Consider near-hits as well as actual events)
Risk Rating
5
Almost
Certain
The unwanted event has occurred frequently;
occurs in order of one or more times per year &
is likely to reoccur within 1 year
11 (M) 16 (H) 20 (H) 23 (Ex) 25 (Ex)
4
Likely
The unwanted event has occurred infrequently;
occurs in order of less than once per year & is
likely to re-occur within 5 years
7 (M) 12 (M) 17 (H) 21 (Ex) 24 (Ex)
3
Possible
The unwanted event could well have occurred in
the business at some point within 10 years 4 (L) 8 (M) 13 (H) 18 (H) 22 (Ex)
2
Unlikely
The unwanted event has happened in the
business at some time; or could happen within
20 years
2 (L) 5 (L) 9 (M) 14 (H) 19 (H)
1
Rare
The unwanted event has never been known to
occur in the business; or is highly unlikely that it
could ever occur beyond 20 years
1 (L) 3 (L) 6 (M) 10 (M) 15 (H)
Interpretation of Risk Level
Risk Rating Risk Level Guidelines for Risk Matrix
21 to 25 (Ex) – Extreme Eliminate, avoid, implement specific action plans / procedures to manage & monitor
13 to 20 (H) – High Proactively manage
6 to 12 (M) – Medium Actively manage
1 to 5 (L) – Low Monitor & manage as appropriate
53. 47
APPENDIX VII: SAFETY, HEALTH AND ENVIRONMENT17
The catalysis centre has had the highest records of safety incidents (von Blotnitz, 2013). Therefore a detailed hazard analysis must be
done before experiments are performed in the laboratory. Table 4 is the analysis of the possible hazards in the laboratory pertaining to
the equipment which will be used. The Standard Anglo Risk Matric, see Appendix VI , was used to determine the loss type, level of
consequence and risk rating.
Table 15: Hazard inventory pertaining to equipment
No Sub
System
Hazard- “What if” Risk Current Controls Loss Type C L R
1
High
pressure
line
Release of high pressure
gas
Injury to body,
head and
limbs
Only authorised
personnel may handle
the pipelines
S/H 4 4 4L
2 X-Ray
Diffraction
Accidental release of x-rays Cause cancer
X-ray automatically
switches off when the
XRD door opens
S/H 4 4 1L
3 Furnace/
Oven/
Reactor
Exposure to hot surface First degree
burn
Has ventilation system for
cooling
S/H 2 2 3L
54. 48
Table 16 : Hazard inventory pertaining to chemicals
Hazard/
Aspect
Chemical name Consequences of release Comments
Chemical Hydrogen A high H2 concentration causes an oxygen deficient
environment. For O2 < 40,000 ppm –
unconsciousness will occur in less than 40 seconds
and death may occur if there is no treatment in the
next 1 minute.
Hydrogen is highly flammable and at high pressures
burns with a colourless flame.
Hydrogen forms an explosive mixture
with air.
Chemical Methane Forms explosive mixtures with air.
May cause suffocation by displacing the oxygen in
the air.
Extremely flammable
Chemical Carbon Monoxide 35ppm = headache & dizziness within 6-8 hrs of
constant exposure.
100ppm = slight headache in 2 to 3 hrs,
400ppm = frontal headache within 1 to 2 hrs,
800ppm = dizziness, nausea, convulsions within
45min & insensible within 2 hrs,
1,600 ppm – headache, tachycardia, dizziness &
nausea within 20 minutes, death in less than 2 hours
Colourless, odorless and tasteless
therefore its leak is not easy to detect.
55. 49
Hazard/
Aspect
Chemical name Consequences of release Comments
Chemical Carbon dioxide 10,000 ppm – slight effect on chemical metabolism
after several hours
30,000 ppm – weak signs of intoxication
40,000 ppm – deeper & more rapid breathing.
50,000 ppm – more laborious breathing, headache &
loss of judgement
100,000 ppm – unconsciousness in less than 1
minute and further exposure leads to death
Chemical Ammonium Hydroxide
(Aqueous solution of
Ammonia)
Extremely destructive to the tissue of the mucous
membranes and upper tract, eyes and skin.
Must always be prepared in the fume
hood and appropriate PPE must be
worn.
56. 50
APPENDIX VIII: TIMELINES18
Table 17: September 2013 time lines
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Meeting With Supervisor
Proposal
Proposal First draft
Proposal second draft
Final proposal due
Presentation
Catalyst supports preparation
Test catalyst start
Sep-13
57. 51
Table 18: October 2013 time lines
Table 19: November 2013 timelines
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Meeting With Supervisor
Prepare and characterize
catalysts
Test Catalysts
Final oral presentation
Experiment stop
Report writing start
Oct-13
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Meeting With Supervisor
Report writing
Final report Due
Poster session
Novemeber 2013
