2
Bachelor Thesis Chemistry
Characterization of Mo-V-W-O catalysts
by
Hannah de Valk
10559132
08 December 2017
18 EC
Research institute
Van 't Hoff Institute for Molecular
Sciences
Research group
Sustainable Materials Characterization
Supervisor/examiner
Moniek Tromp
Second examiner
Stefania Grecea
Daily supervisor
Michelle Hammerton
2
Abstract
Molybdenum and vanadium oxides are known to be effective catalysts for the oxidation of acrolein to acrylic acid. The addition of W in the phase pure structure can make the catalyst more selective and more hydrothermal stable. The aim of this research was to fully characterise four phase pure structures: the orthorhombic, trigonal, tetragonal and
amorphous Mo3VO catalysts. In addition to this, we attempted to verify whether or not W
is redox inactive in the amorphous phase pure structure. The phase composition and redox behaviour of W has been examined by x-ray adsorption and diffraction spectroscopy.
Although orthorhombic Mo3VWyOx has been synthesized previously, the reaction was too
sensitive to recreate this product. In addition to control of pH, temperature and a solution purged of air, the conditions that are necessary to synthesize the required product are the presence of a nucleation surface and the absence of vibrations during the hydrothermal
treatment. Nevertheless, Mo3VWyOx crystalline in the ac-plane but amorphous in the
ab-plane was formed with different amounts of W (5, 7.5 and 10 at.%). Oxidation and reduction reactions of these catalysts were carried out. After analysing the XAS data of the
W L3-edge it can be concluded that W is redox inert in the amorphous phase. Small changes
in XANES spectra after the oxidation and reduction of amorphous Mo3VWyOx can be due to
structural changes in the catalyst due to redox of Mo6+ or V+4/+5, which are also present in
the catalyst. Furthermore, it is likely that a higher amount of W in the MoVW based catalyst made the synthesized catalyst more redox stable compared to the MoV based catalyst. Because of the stronger covalent bonding with oxygen.
3
Populair wetenschappelijke samenvatting
In de industrie wordt er veel gebruik gemaakt van katalysatoren. Een katalysator is een stof die de reactie versneld maar hierin zelf niet wordt verbruikt. Deze katalysatoren kunnen reacties mogelijk maken die in eerste instantie niet lukken of ze zorgen voor het versnellen van het proces zonder dat hierbij de druk, temperatuur of concentratie verhoogd hoeft te worden. In de scheikunde wordt veel onderzoek gedaan naar katalysatoren omdat het ervoor zorgt dat we duurzamer met energie omgaan.
De katalysator die in dit onderzoek is gebruik, is een heterogene katalysator. Dit is een katalysator die zich in een andere fase bevindt dan de reactanten. De katalysator is in dit onderzoek een vaste stof terwijl de reactanten zich in een vloeibaar of gasmengsel bevinden. Tijdens de reactie zal de reactant zich binden aan het oppervlak van de katalysator waarbij sommige verbindingen binnen het substraat verbroken worden en andere bindingen worden opnieuw gemaakt.
In dit onderzoek is er gekeken naar de katalysator die in de industrie wordt gebruik voor het omzetten van acroleïne naar acrylzuur, zie figuur A. Acrylzuur wordt gebruik bij het maken van onder andere plastics, verf, lijm en inkt.
Figuur A: Oxidatie van acroleine naar acrylzuur met behulp van een katalysator.
De katalysator die bij deze reactie wordt gebruik is een metaaloxide. In dit geval levert de katalysator zijn zuurstof uit zijn kristaalrooster aan het reactant zodat deze oxideert. Vervolgens neemt de katalysator weer een nieuw zuurstof atoom op uit zijn omgeving. De katalysator die bij de omzetting van acroleïne naar acrylzuur wordt gebruikt, bestaat uit de metalen molybdenum (Mo), vanadium (V), wolfraam (W) en zuurstof (O). Molybdenum zorgt ervoor dat de katalysator actief is, terwijl vanadium juist voor de selectiviteit zorgt. Daarnaast zorgt wolfraam ervoor dat de katalysator beter bestand is tegen hoge temperaturen.
In de industrie wordt deze katalysator al veel gebruikt. In dit geval zijn de metaalelementen niet gelijkmatig verdeeld, maar bevinden deze zich in verschillende structuren. Wanneer de oxidatie reactie wordt uitgevoerd zullen V en Mo van oxidatie toestand veranderen, terwijl W in dezelfde oxidatietoestand zal blijven. Het is alleen nog niet bekend waar W zich in deze structuur bevindt. Om fundamenteel onderzoek te doen, is er voor gekozen om een katalysator te maken die uit één pure fase bestaat, opgebouwd uit de elementen Mo, V, W en O. Door verschillende analysemethodes zal er gekeken worden of het is gelukt om het pure kristal te maken en om te kijken of W redox actief is. Tijdens het onderzoek is het niet gelukt om een pure fase te maken, maar een kristal zonder een duidelijk gevormde structuur. Uit de analyses bleek dat in deze structuur W niet redox actief is. Maar er is een grote kans dat de andere elementen uit het stof dit wel zijn. Door meer onderzoek te doen zouden we hier beter inzicht in kunnen krijgen.
4
Contents
Abstract 2
Populair wetenschappelijke samenvatting 3
Contents 4
1. Introduction 5
1.1 Oxidation of acrolein 5
1.2 Development of MoVWO based catalyst 5
1.2.1 Phase pure Mo3VOx catalyst 5
1.2.2 Addition of W to the MoV based catalyst 7
1.3 Redox behaviour of Mo-V-W based catalyst 8
1.4 Project Aim 8
2. Results & Discussion 10
2.1 Overview of all experiments 10
2.2 Synthesis of Mo3VOx 12
2.3 Synthesis of Mo3VWyOx 12
2.4 TGA analysis 12
2.5 XRD analysis 13
2.6 XAS analysis 17
3. Conclusion & Outlook 20
4. Experimental Section 21
4.1 Experimental settings and analysing methods 21
4.2 Synthesis of orthorhombic Mo3VOx 22
4.3 Synthesis of amorphous Mo3VWyOx 5.0 at.-%, 7.5 at.-%, 10 at.-% 22
Acknowledgements 23
References 24
5
Chapter 1
Introduction
1.1 Oxidation of acrolein
Nowadays there is an increasing rate of the use of acrylic acid (AA), which is used for paint, lime, ink and plastics.1 AA is the simplest unsaturated monocarboxylic acid that undergoes excellent
polymerization. There are different ways to produce this product, although the most commonly used pathway in the industry is via oxidation of acrolein (ACR) that can be converted into AA by the use of a heterogenous catalyst.2 There are many different metal oxides that can be used as
catalyst for the oxidation from AA to ACR.
1.2 Development of MoVWO based catalyst
The Mo-V based oxides are considered the most active catalysts for the oxidation of acrolein to acrylic acid.1 The reasons for this are the multiple oxidation states of the elements in the catalyst
and the complexity of the catalyst, that makes it possible to stabilize acrylate and oxidize it at the same time.3,4 Moreover, when tungsten (W) is added to the V-Mo structure, the catalyst becomes
more stable at high temperatures.2 The morphology of the catalyst has influence on the activity
of the oxidation reactions. 3 The morphology will change by using different synthesis methods.3 In
industry the MoVW based catalyst is prepared by using a spray-drying or crystallization method. This results in a catalyst with many different phases and different metal stoichiometries.3 To gain
a more fundamental understanding of the MoVW based catalyst, phase pure structures have to be made.
The Mo, V and W monometallic oxides are respectively MoO3, which is inactive, V2O5, which is
active, but unselective, and WO3, which is completely inert.3 But with mixing these elements the
catalyst will become active, selective and stable. Different studies show that the optimal Mo/V ratio for the synthesis of AA is 3:1 in this crystal catalyst.1, 4, 5 The catalyst with this ratio, contains
the largest amount of V+4 ions, instead of V+5, that serves as centre for the stabilization of ACR, and
Mo+6 oxidation states.4-6 Moreover, Mo4VO14 and Mo3VO11 are discussed as active phases.3 The
amorphous parts are important for the partial oxidation of ACR, because they show a higher stability under reduction conditions compared to the crystalline samples.7 To gain a detailed
fundamental understanding of the function of the industrially used catalyst, ordered, phase pure reference structures are needed. The focus must be on the rational approach, rather that the optimization of the catalyst. By using the hydrothermal method, that is not used in industry, it is possible to control the crystal structure of the catalyst and make phase pure. The hydrothermal method was used to make a phase pure crystal. This method is based on high temperatures and high pressure. For these reactions autoclaves are used. Important factors are the initial pH of the solution, the duration of the synthesis, the temperature and the pressure.
1.2.1 Phase pure Mo3VOx catalysts
The group of Ueda synthesized several different pure phase crystalline Mo3VOx using a
hydrothermal method: trigonal, orthorhombic, tetragonal and amorphous phases (amorphous (written as: Tri-Mo3VOx, Orth-Mo3VOx, Tetra-Mo3VOx and Amor-Mo3VOx) are shown in figure 1.1.1
It was showed that not the amorphous phase, but the crystalline trigonal and orthorhombic are phases the most active for the partial oxidation of ACR to AA.
6
Figure 1.1: Structures of (a) orth-Mo3VOx , (b) tri-Mo3VOx, (c) tetra-Mo3VOx, (d) amor-Mo3VOx. 8 Pink is the pentagonal {Mo6O21}6- unit. Orange is the heptagonal channel.
To gain a more fundamental understanding of the active sites of the catalyst, different crystal structrures were investigated.1 These different pure crystaline phases form the same layered
structure in the c-plane, which is confirmed by XRD with the peaks at 22.5 and 45 deg 2-theta.1
The differences between these strucures occur in the ab-plane where they have a different crystallinity. The different arrangements of these structures affect the catalytic performance in the partial oxidation of ACR. . The group of Ueda showed that the conversion of acrolein depends on the external surface area of the catalyst.8 Because, the heptagonal channels micropores in the
catalyst release oxygen to oxidize acrolein. With more heptagonal channels is the ab-plane the catalyst will become more active. Therefore, heptagonal channels are essential for the oxidation of ACR to AA.9 The most active phase consists of the orthorhombic Mo3VO11 crystalline phase. In
the ab-plane of the orth-MoVO there are different layers made of a network based on the pentagonal {Mo6O21}6- unit, MoO6 and VO6 octahedrons. Orth-Mo3VOx forms empty hexagonal and
heptagonal channels in the ab-plane, see figure 1.1 The heptagonal channels micropores consist at the external surface of orth-Mo3VOx and release oxygen to oxidize acrolein to acrylic acid.
A decrease in the amount of heptagonal channels makes the catalyst less active. Comparing the crystal structures of the different phases shows that tri-Mo3VOx and amor-Mo3VOx have a
crystalline system along the ab axis identical to the orthorhombic structure, but they show fewer heptagonal channels along the axis, therefore these morphologies are less active. Tet- Mo3VOx has
no heptagonal channels, and has a low activity toward the oxidation of acrolein.
The morphology of the crystal structure has also an influence on the oxygen mobility through the catalyst and with that on the activity of the oxidation of AA. The reaction of ACR to AA proceeds via Mars-Van Krevelen mechanism.10 After ACR is oxidized by the oxygen from the surface of the
catalyst, the vacancy will be replenished by oxygen form the bulk.3 By having an excess of
oxidation sites, overoxidation of the reactant can happen easily and partial oxidation of AA can only be achieved when the surface is partially reduced.2 Therefore, the degree of oxidation of the
surface influences the selectivity of the catalyst. The catalytic activity also correlates with the amount of oxygen that is available for the oxidation reaction. The different arrangement of the structure affects the catalytic performance, as well as the incorporation of another heteroatom in the phase pure crystal structure.2 W has a strong covalent bonding with oxygen and so reduces
the oxygen mobility through the bulk catalyst, slowing the reoxidation of the surface.2 Moreover,
by the incorporation of W into the structure, the catalyst becomes less redox active and overoxidation by the heptagonal channels will occur less.
c) b)
7
1.2.2 Addition of W to the MoV based catalyst
The addition of W increases the complexity of the catalyst significantly and causes competing effects regarding redox activity, oxygen mobility and crystallinity, which in turn affects the activity and selectivity of the catalyst. The addition of W affects the crystallinity of the orthorhombic structure and causes crystal splitting, that contributed to exposure of more active phases.11
Another effect of crystal splitting is that the oxygen mobility through the catalyst.3 This is
important because the dynamics of bulk oxygen lead to re-oxidation process on the catalyst surface and thus influence the selectivity pattern of the network of acrolein oxidation.2 The
catalyst has highest activity when W is added in small amounts (W0<y<1).3 An excessive amount W
content will deactivate the reaction because the proportion of inert redox stable material is increased. W has a strong bond with oxygen that is not easily broken, in consequence the catalyst will become less active.12 However, a small amount of W is stabilizing but not deactivating.13
Presence of W also prevents clustering of edge sharing metal oxygen octahedrons, that otherwise result in shear structures that reduce oxygen mobility. If the catalyst MoVWO has a lower W content, it will release more oxygen, increasing the activity and reducing the selectivity of the catalyst. In the stoichiometry of Mo8V2WyOx the optimal point between activity and selectivity is
with a W content of y=1.3
The group of Ueda investigated the addition of W into the phase pure crystal structures.11 .
Different amounts of W (adjustment of W content to 2.5, 5.0, 7.5, and 10 at.% in the synthesized Mo-V-W-O) where added to the amorphous, orthorhombic and trigonal structure. W was found to cause cleavage of Mo-O and V-O bonds, resulting in a decrease in the long range order of the ab- plane.11 The crystal size was also thereby decreased, resulting in a larger external surface area
that contributed to exposure of more active sites. The structures of these catalyst were investigated by XRD characterization see figure 1.2. X-ray diffraction patterns of amor-MoVWO and orth-MoVWO were made with different W contents. In figure 2(II), the diffraction patterns of the orthorhombic structure, showed diffraction peaks at 6.6, 7.9, 9.0 and 27.3°, etc. with the planes (020), (120), (210), and (630). With an increasing W content the peaks before 10°, what indicate the orthorhombic structure, became smaller. Therefore, with an increasing W content, the crystals had split into smaller rods. The addition of W decreases the crystal size and has as consequence a larger surface area. This in contrast to figure 2.1(I), the amor-MoVWO, that showed an amorphous structure. Where the XRD barely changed a shift.
Figure 1.2: X-ray diffraction patterns of amor-MoVWO (I) and orth-MoVWO (II) with different W contents. a) W0 b) W2.5 c) W5.0 d) W7.5 e) W10.
8
Moreover, the research of Ueda showed that amor-MoVWO and orth-MoVWO still contained the same pentagonal units {Mo6O21} and {MO6} (M=Mo,V) octahedra.11 Plausibly, W formed {WO6}
octahedra and acted as linkers connecting {Mo6O21} pentagonal units. In contrast to the research
of Ishikawa et al., were was concluded that W in the phase pure tri-MoVWO catalyst was located at the {Mo6O21}6- pentagonal unit by substituting Mo.14 This substitution caused only a slight
change in catalytic activity.
The catalyst with the highest activity towards the oxidation of ACR was the orthorhombic Mo-V-O-W with 7.5 at.% (Mo69.7V23.1W7.2), again due to the higher degree of ordered arrangements of
heptagonal and hexagonal channels and the large external surface area.11 Consequently, the
orthorhombic structure was highest in activity. However, with an equal amount of W, the amorphous MoVW has a similar selectivity toward the oxidation of ACR. Because in the amorphous phase fewer heptagonal channels that can oxidize acrolein are present, but with the same amount of W the selectivity will stay the same.
1.3 Redox behaviour of Mo-V-W based catalyst
The redox activity of W is possibly dependent on the structure of the catalyst. It is known that WO3, WO2 and W are not redox active.7 Although it is known that WO3 is completely reduced to
W0 metal near 800°C, via the formation of WO2.72, at 520°C, and WO2, at 600°C. In previous
research W was also shown to be redox inert in the Mo8V2WyOx mixed phase catalyst at 480°C
(where the reduction was carried out by acrolein) and 400°C (where the oxidations was carried out with O2. ++ Previously, research showed that the mixed phase Mo8V2WyOx catalyst contains
Mo+6, W+6 and V+4/+5.5 The V+5 ions are reduced during the oxidation of ACR to V+4 if the reaction
is carried out below 350 ֯C. When the temperature was increased to 400 ֯C, Mo+6 was also reduced
and became Mo+4. If the W content increases (up until y<1.5) vanadium can be reduced to
oxidation state V+3 while W stays inert. It has not yet been investigated if W in a phase pure crystal
also participates in redox reactions. However, the redox behaviour of W has not yet been investigated when incorporated in a phase pure MoVW based crystal.
1.4 Project aim
To bridge the gap between ideal and real systems and gain a better scientific understanding of the structure and dynamics of the catalyst, different phase pure Mo3VOx and Mo3VWyOx catalysts will
be studied. It is also important to understand the role of W in a phase pure system. This knowledge is needed to optimize and understand the working of the Mo-V-W based catalyst and unravel the industry. It is relevant to know how stabilizing elements influence the active phases of the catalysts.
In this research the synthesis and the stoichiometries of different phase pure Mo3VOx catalysts
will be studied. Moreover, the redox behaviour of W will be studied that will give more insight in the morphology of the catalyst, as it is not known if W takes part in the redox activity in the phase pure Mo3VWyOx catalyst. This report set out the answer of the following research question: In
which way is it possible to synthesize phase pure MoV-based catalyst and how is it possible to introduce W into this phase pure structure? To answer this question orth-MoVOx. tri-MoVOx,
amor-MoVOx, and orth-Mo3VWyO with different amounts of W, will be synthesized and the phase
composition of the catalyst and redox behaviour of W will be examined by x-ray adsorption spectroscopy and x-ray diffraction spectroscopy.
Formerly, only one group managed to synthesise these phase pure catalysts using a hydrothermal method.1, 11 However, it is important to demonstrate the reproducibility of these experiments in
order to test the reliability of their synthesis. Therefore, different amounts of W, 5 at.%, 7.5 at.%, and 10 at.% (also written as W5, W7.5 and W10) will be added to phase pure Mo3VOx forming
9
spray dried Mo3VW0.5Ox : 21.5 < x < 30.5).2 The samples will first be oxidized under air by heating
till 300 ֯C as described previously.2 After the oxidation the sample will be reduced by purging using
30% H2 flow in N2 at 480 ֯C. By combining x-ray absorption spectroscopy and x-ray diffraction
spectroscopy the structure and redox activity of orthorhombic-Mo3VWyOx will be determined.
To investigate the changes in crystal structure due to incorporation of W, XRD and XAS analyses were performed. To compare longhe range order to the local structure of the catalysts. X-ray diffraction is used to provide bulk structural information of systems.15 To investigate if W is
introduced in the MoV framework and investigate the local environment and bulk sample X-rays absorption spectroscopy (XAS) was used. XAS is technique that provides element structural information about the local order of the elements in a crystal and can be used to determine the oxidation state of an element in a compound.15,16These data are obtained by the incoming varying
photon energy across a range of the energies to excite core electrons. The peak intensities will become more intense when there are more unoccupied orbitals for the core electron to be excited into. Absorption energies are characteristic for the elements. In this research W-L3 XANES were
measured up to k=7.3Å-1 to normalize the spectrum. The combination of these methods will give
10
Chapter 2
Results and discussion
2.1 Overview of all experiments
During the research different synthesis were executed, the XRD patterns are included in the appendix 2,3,4 and 5. Table 2.1 summarises which of the experiments led to a desired product. There is made a distinction between the experiments that are synthesized during the weekend and during the week. Also is a distinction made between different batches of the PTFE sheets that are used. In previous research, PTFE sheets with a thickness of 0.1 mm were used.17 Whereas in
this research PTFE sheets of two different batches from the same manufacturer with a thickness of 0.025 mm were used. Also a PTFE sheet with a thickness of 0.05 mm from a different manufacturer was used. Two of the PTFE sheets had the same thickness, there were differences in transparency and texture and only one resulted in crystallisation of the pure orthorhombic phase. Batch to batch variations in terms of colour and surface characteristics result from the manufacturing process used to make this film, where PTFE powder is mixed with naphtha and then calendared, dried and then sintered. This information was gained by communication with the company who produce PTFE sheets. During this research is tried to make the MoV based and MoVW (with different amounts of W) based orthorhombic phase, trigonal phase and the MoV based amorphous phase. All the experiments that are depicted in the table are carried out under argon.
Table 2.1: Overview of the experiments that are done compared to different conditions.
The hydrothermal synthesis was found to be highly sensitive to experimental conditions. Several conditions had influence on the formation of a phase pure crystal structure. For the formation of the orthorhombic phase, it is important to have a nucleation surface, in this case a PTFE sheet. Without the presence of a PTFE sheet, hexagonal molybdenum trioxide (h-MoO3) is the only
crystal that is formed.18 See figure 2.1 for the XRD pattern of synthesised MoO3, while there was
tried to make phase pure orth-Mo3VOx. Figure 2.2 shows a XRD pattern of a reference structure of
h-MoO3, table 2.2 describes the hkl-indices of the peaks.19 No XRD peaks were identified that the
structure contained a V compound.
Conditions PTFE Sheet Amorphous Key: Exp missing (mm) MoV MoV seeds MoVWy (5/7.5/10 at.%)MoV MoVWy (7.5 at%) MoV MoVWy (5/7.5/10 at.%) No material
Weekend 0.025 batch 1 Pure phase
Week 0.025 batch 1 Undesired phase
Weekend 0.05 Very small particles
Week 0.05
Weekend 0.025 batch 2
Week 0.025 batch 2
11
Figure 2.1: XRD pattern of synthesized MoO3
Figure 2.2: XRD pattern of h-MoO3 from literature.19
The 0.05 mm PTFE sheet was found to be unsuitable to control crystal growth as hydrothermal treatment resulted in very small crystallite sites. Since this synthesis is very sensitive to the nucleation surface, both the thickness and film surface appear to influence the crystallization process. Because the nucleation surface is important for the crystallisation process of MoV based catalysts. The experiments that were synthesized during the week did not give phase pure crystals. The reason for this can be that during the week vibrations will disturb the nucleation process. Only the reactions that were hydrothermally treated during the weekend gave the desired product. After discovering this condition, the hydrothermal treatment was carried out in an oven where no vibrations could occur. During the start of the research the synthesis were carried out in air. This gave also MoO3 as result. The conditions were perhaps too oxidising, so
instead of forming the mixed oxide, MoO3 is formed instead as it is possibly more stable under oxidising conditions.
The conditions that did result in a successful synthesis were: firstly, to have a nucleation surface in the Teflon tube. In this experiment a PTFE sheet with a thickness of 0.025 mm was used. Secondly, it was important to execute the addition of the two solutions under argon. Finally, the hydrothermal treatment must be carried out at a place where no vibrations could occur. This synthesis under this conditions was carried out one more time and gave similar results.
0,00 10,00 20,00 30,00 40,00 50,00 60,00
12
2.2 Synthesis of orth-Mo3VOx
In figure 2.4, the XRD pattern of orth-Mo3VOx is shown and compared to the research of Ishikawa
et al., where a reference structure was calculated.20 The firs peak that is shown in figure 2.3 at 5
degrees is not gain by experimental synthesis. By comparing these two patterns it can be concluded that the orthorhombic phase has started to develop but the experimental sample is less crystalline. All the different phases of the Mo3VOx catalyst have the same layered structure in the
ac-plane. These two peaks are ascribed to (001) at 22 ̊2θ and (002) at 45 ̊2θ plane reflections. The three peaks before 10 ̊2θ indicate which Mo3VOx structure is synthesized. In orth-MoVO the
main diffraction peaks that are corresponding to this crystal structure emerged at 6.6, 7.9, 9.0 and 27.3 ̊2θ. These are the planes of (020), (120), (210) and (630). 11
The crystal structures of MoO3 and VO2 are not found in this pattern. The refence structures of
these crystals are included in appendix 5. 2.3 Synthesis of Mo3VWyOx
After the synthesis of orth-Mo3VOx, it was attempted to incorporate W in this structure. To
investigate the changes in crystal structure due to incorporation of W, XRD and XAS analysis was performed.
2.4 TGA and DSC analysis
The results of the MoVW catalyst, with different W contents, show almost the same graphs of the thermal gravimetric analysis (TGA) and differential scanning calorimetry (DSC). During synthesis of the catalyst was dried to 80°C. The TGA showed around 80°C mass loss of 3.1% that is likely due to the balance that has to stabilize. The mass loss around 280°C of 1.7 % in mass is likely due to loss of oxalic acid that was still in the crystal structure.21, 22 With differential scanning
calorimetry (DSC) it became clear that during heat treatment from 35-650°C there was no phase change. So heating to 650 °C will not give phase changes in the catalyst. The graphs of the TGA and DSC are shown in figure 2.4, 2.5 and 2.6.
Figure 2.3: XRD pattern of a) computed orth-Mo3VOx and b) synthesized orth-Mo3VOx
b a (001) (002) (020) (120) (210) (630)
13
2.5 XRD analysis
After hydrothermal treatment of the MoVWO compounds with a different W content (W5, W7.5 and W10) XRD patterns were measured and analysed, see figure 2.7. These XRD patterns were compared to the diffraction patterns of the research of Qui et al., shown in figure 2.8. 11 Comparing
these patterns, it can be concluded that synthesis of the orthorhombic phase was not successful, but the amorphous phase was synthesized instead. The experimental diffractogram shows two sharp peaks at 22°-2θ and 45°-2θ °, implying that the amorphous catalyst is only crystalline along the c axis. Confirming the literature, in this research the synthesis was even more sensitive than the synthesis of orth-Mo3VOx. 11 The addition of W results in crystalline splitting into smaller rods
as lattice contraction results in unstable structures.
In the research of Qui et al. they state that there is a high probability that W forms {WO6} octahedra
and acts as linkers connecting {Mo6O21} pentagonal units, what was indicated by STEM and
RAMAN analysis.11 It is possible that a disordered arrangement of pentagonal {Mo6O21} and {MO6}
octahedra was formed in the ab-plane.
Figure 2.4: TGA and DSC of amor-MoVWO: W5
Figure 2.6: TGA and DSC of amor-MoVWO with a W content of 10 at.%
14
Figure 2.7: XRD patterns of MoVWO with a) W 5 b) W 7.5 c) W 10
Figure 2.8: X-ray diffraction patterns of amor-MoVWO, with different W contents. a) W0 b) W2.5 c) W5.0 d) W7.5 e) W10.
a
b
c
(001)
15
After hydrothermal treatment, all samples were heat treated in air to ensure that they were fully oxidised. Therefore samples were heat treated/oxidised at 300°C in air overnight. After oxidizing the samples they still showed the same diffraction pattern, only with a lower intensity, see figures
2.9, 2.10 and 2.11. The reason for this can be that the layers have been fallen apart or less material
is used. To investigate the incorporation of W in the structure further, the compounds were reduced at 480°C with a 30% H2 flow in N2 with a 3°C/min heating rate over 4 hours. A change in
crystal structure was the result of reduction, as also shown in figures 2.9, 2.10 and 2.11. The data of some diffraction patterns were enlarged in size in order to compare it more easily. And therefore due to heat treatment alone, because the DSC did not show any phase changes up to 650°C under air, see figures 2.5, 2.6 and 2.7. The XRD pattern show a phase change after reductions that was carried out at 480°C, so the reduction process have influenced the phase change of amor-Mo3VWyOx to MoO2.
After reduction, different crystal compositions were formed. By comparing the literature to the standards that were measured, it looks like there is MoO2 character in the experimental diffraction
patterns of the reduced material, see figure 2.9, 2.10 and 2.11: c) for the synthesized product and MoO2 for the reference structure. If the pattern is compared to other structures as VO2, MoO3 and
V2O5 no correlations are identified, see appendix 5. It is possible that the other metals do not form
16
Figure 2.9: X-ray diffraction patterns of Amor-Mo3VWyOx with W5 a) MoVWO b) oxidized MoVWO (x2) c) reduced MoVWO (x2) d) MoO2
Figure 2.10: X-ray diffraction patterns of Amor-Mo3VWyOx with W7.5 a) MoVWO b) oxidized MoVWO (x2) c) reduced MoVWO (x5) d) MoO2
Figure 2.11: X-ray diffraction patterns of Amor-Mo3VWyOx with W10 a) MoVWO b) oxidized MoVWO (x2) c) reduced MoVWO (x2) d) MoO2 0,00 10,00 20,00 30,00 40,00 50,00 60,00 I re l. 2 theta a b c MoO2 0,00 10,00 20,00 30,00 40,00 50,00 60,00 I re l. 2 theta a b c MoO2 0,00 10,00 20,00 30,00 40,00 50,00 60,00 a b c MoO2
17
2.6 XAS analyses
XAS analysis of the catalysts was made directly after hydrothermal synthesis, after oxidation at 300°C and after reduction with H2. To analyse the data it is important to look to the position of
the absorption edge, defined as E0, which can shift to the right for higher oxidation states.23
Furthermore, the peak height, or area that is more accurate, must be analysed to obtain more information about the oxidation state. If the area becomes larger, then W will be in a higher oxidation state as there are more empty states that the exited electron can enter when W is oxidised. The W L3 near-edge spectra of the different catalysts with W5, W 7.5 and W10 are
compared to W standards: W foil, WO2, WO3 and ammoniummetatungstate (starting material)
with oxidation states respectively W0, W4+ and W6+ shown in figure 2.12.
Some of the adsorption shifts show higher energies than that WO3-standard. However, for W it is
impossible to have a higher oxidation state that +6. Therefore, the higher white line could be due to differences in the local environment instead due to a higher oxidation state. The change in peak heights can be an indication that the typical surrounding of W6+, in WO3, is different than in the
hydrothermal samples.
Figure 2.12: W L3 near-edge absorption spectra of reference structures and amor-MoVWO with different W content after hydrothermal treatment.
Figures 2.13, 2.14, 2.15 show the different states of W5, W7.5 and W10 with the standard samples
18
Figure 2.13: W L3 near-edge absorption spectra of reference structures and amor-MoVWO with W5 after hydrothermal treatment, oxidation and reduction.
Figure 2.14: W L3 near-edge absorption spectra of reference structures and amor-MoVWO with W7.5 after hydrothermal treatment, oxidation and reduction.
Figure 2.15: W L3 near-edge absorption spectra of reference structures and amor-MoVWO with W10 after hydrothermal treatment, oxidation and reduction.
19
It is unlikely that W underwent changes in oxidation state, however it is possible that the structural change affects the XANES of W. If W is added in higher amounts it is likely that the catalyst becomes more redox stable, because of more WO3-like bonds form.11 As shown in the
XRD patters, the structure of the samples changed a lot. According to the literature exposure of the samples to the reduction of H2 results probably in a reduction of both Mo6+ and V5+ whereas
W6+ is inert.24 The difference in peak areas can be affected by the coordination environment.5
Moreover, for hexagonal (V, Mo)O3 a direct reduction to MoO2 is expected in literature.7 With that
20
Chapter 3
Conclusion and outlook
To conclude, orth-Mo3VOx has been synthesized. In addition to pH, temperature and synthesis
time, the synthesis was very sensitive to the nucleation surface and vibration during hydrothermal treatment. The synthesis of orth-Mo3VWyOx with W5.0, W7.5 and W10 did not
result in formation of the orthorhombic structure, but in the amorphous structure, according to the diffraction patterns. Amor-MoVWO is only crystalline along the c axis, while a disordered arrangement of pentagonal {Mo6O21} and {MO6} octahedra is known to form in the ab-plane. The
synthesis including W was even more sensitive, probably due to the addition of W in the crystal structure that makes the crystal break in even smaller rods. After the reduction of the three W containing catalysts, the crystal structure changed drastically. h-MoO2 was formed after reduction.
The rest of the structure seemed to become amorphous. After the reduction, the oxidation state of W did not show a significant difference. It is very likely that the oxidation state of Mo and V had changed and with that the geometry of the crystal. This can be confirmed by XRD, with formation of the MoO2 phase, in which Mo is in oxidation state +4 in stead of the oxidation state of +6 where
Mo started in. Futhermore, it is likely that adding W in higher amounts makes the amor-MoVWO more redox stable. The XRD pattern of reduced amor-MoVWO with W10 showed, beside the MoO2
peaks, still a small pattern of amor-MoVWO while the other reduced patterns did not. In further research it would be valuable to use XAS to measure the different oxidation states of V and Mo and how they are influenced by the presence of W in the pure amor-MoVWO and, if possible, in the pure orth-MoVWO structures.
21
Chapter 4
Experimental section
4.1 Experimental settings and analysing methods
The catalysts orth-Mo3VOx and amor-Mo3VWOx with W5, W7.5 and W10 were synthesized as
previously reported in several studies by the Ueda group.11 Because of differences in equipment
a few minor modifications were set. First, in the literature they performed hydrothermal treatment with a solution of 240 mL. During these experiments a solution of 25 mL was made, so the whole synthesis was downscaled. Furthermore, the reaction was carried out under argon, while in the procedure of the Ueda group, they worked under air and used only nitrogen for 10 minutes before the hydrothermal treatment. Finally, in the research of Pyrz et al., that is part of the research group of Ueda, they used for synthesis of MoVO a PTFE sheet with a thickness of 0.1 mm.17 It is likely, unless otherwise stated, that the experiment described by Chen et al. and Qui et
al. also used a PTFE sheet with the same thickness of 0.1 mm. In contrast to this experiment, PTFE sheets with a thickness of 0.025 mm from Goodfellows and 0.05 mm from VWR were used. To explore the scope of the synthesis phase pure orth-MoVWyOx first pure orth-Mo3VOx had to be
synthesized. The precursor solutions consist of clear blue VOSO4 ּ nH2O aqueous solution and a
transparent solution of (NH4)6Mo7O24 ּ nH2O in water. By adding the V compound dropwise to the
Mo compound the solution changes colour immediately to dark brown. This is because polyoxomolybdate is formed.25 During the addition of V the precursor Mo solution changed from
a pH of 5.6 to 3.2. The pH controls the concentration of {Mo72V30} species in the solution. By adding
all the V compound to the Mo compound the pH will be 3.2 and 63% of V in the precursor solution react with Mo and forms {Mo72V30}. Likewise, decreasing the pH will lower the amount of
{Mo72V30} formed. Moreover, lowering the pH results in the trigonal, rather than orthorhombic
structure.
Unless stated otherwise, all the employed reagents and solvents used were purchased from Sigma Aldrich, Alfa Aesar, Pfalz & Bauer inc. and Merck and used as supplied without further purification. All reactions were carried out in glassware under argon atmosphere using standard Schlenk and vacuum line techniques. Liquids were transferred through rubber septa into the reaction flasks via a syringe. The 0.05 mm PTFE was purchased from VWR and the 0.025 mm PTFE from Goodfellows. The influence of redox reaction of amorphous MoVWO was studied by in situ X-ray diffraction (XRD), X-ray absorption spectroscopy (XAS) and thermal analysis (TG/DSC).
In situ XRD studies were performed with a Rigaku Miniflex X-ray diffractometer. XRD measurements were conducted at r.t. Diffraction patterns were recorded in a theta range from 4 to 60°. Phase identification analysis and structure refinement of experimental diffraction patterns was performed using the software Match! Version 3. The Crystallography Open Database (COD) was used to search for reference structures in the literature. 26
The XAS experiments described in this work were performed in the transmission geometry with easyXES100 from 10150 eV up to k 7.3 Å-1. Data analysis of the XANES spectra was performed
using easyXES100 Software and Athena 3.0. 27 The spectra were energy calibrated with respect to
a W metal foil reference spectrum. For background subtraction a first order polynomial was refined to the pre-edge and for normalization a third order polynomial was refined to the EXAFS region up to k 7.3 Å-1. The E0 value for each spectra was defined by the first peak in the first
22
differential in the W L3 edge. Before making the measurements, the product was mixed with boron
nitride.
Thermal gravimetric analysis (TGA) and differential scanning calorimetry (DSC) was performed with a NETZSCH STA 449F3 UvA with a furnace made of SiC (up to 1550 ° C) cooled by means of ventilator. Measurements were carried out at heating rates of 2 K/min, 1:1 air:argon, at a total flow of 20 ml/min the sample was heated from 35-650 ° C. The data treatment was performed via the Proteus Analysis software.
4.2 Synthesis of orthorhombic Mo3VOx
In a three neck 250 mL round bottom flask, VOSO4 ּ nH2O (n = 3.86) (0.336 g, 1.45 mmol, 1 eq.)
was dissolved in deionized water (15 mL), the solution became a clear blue. (NH4)6Mo7O24 (1.11
g, 0.898 mmol, 4 eq.) was also added in a three neck 250 mL round bottom flask and dissolved in deionized water. This solution became transparent. Both solutions where stirred for 10 min. The solution of VOSO4 ּ nH2O was added dropwise to the (NH4)6Mo7O24 solution while stirring. While
adding the vanadium compound the pH dropped from 5.6 to 3.2 and the colour changed to dark brown. The mixture was stirred for 10 minutes and transferred into an autoclave with a Teflon inner tube of 40 mL and a PTFE thin sheet (50cm2, 0.025 mm pore diameter). The reaction
mixtures was purged with argon for 10 min to remove the oxygen. The solution was hydrothermally treated at 175 ̊C for 48h in an oven where no vibration could enter. The brown/grey crystals where filtrated and washed with water and dried overnight at 80 ̊C. The crystals where washed with oxalic acid (0.4 mol L-1) for 30 min at 60 ̊C while stirring. The crystals
where washed with water and dried overnight at 80 ̊C. Yield: 12.3% (0.115 g, 0.27 mmol, purple solid).
4.3 Synthesis of amorphous Mo3VWyOx W5.0, W7.5, W10
Solution A was obtained by adding VOSO4 ּ nH2O (n = 3.86) (0.336 g, 1.45 mmol, 1 eq.) in a three
neck 250 mL round bottom flask. Deionized water (15 mL) was added and the solution became a clear blue. Solution B was obtained by adding (NH4)6Mo7O24 (1.11 g, 0.898 mmol, 4 eq.) and a
certain amount of (NH4)6[H2W12O40] (with adjustment of W content to 5.0 (0.08 g, 26.2 ּ 10-3 mmol)
, 7.5 (0.12 g, 39.3 ּ 10-3 mmol) and 10 (0.16 g, , 52.5 ּ 10-3 mmol) at.% in the synthesized
Mo-V-W-O) in a three neck 250 mL round bottom flask and dissolved in deionized water (15 mL). This solution became transparent. The solutions were stirred for 10 min. Solution B was poured in solution A under stirring conditions. The pH dropped from 5.6 to 3.2 and the colour changed to dark brown. The mixture was stirred for 10 minutes and transferred into an autoclave with a Teflon inner tube of 40 mL and a PTFE thin sheet (50cm2,0.025 mmpore diameter). The reaction
mixtures was purged with argon for 10 min to remove the oxygen. The solution was hydrothermally treated at 175 ̊C for 48h in an oven where no vibration could enter. The dark purple crystals where filtrated and washed with water and dried overnight at 80 ̊C. The crystals where washed with oxalic acid (0.4 mol L-1) for 30 min at 60 ̊C while stirring. The crystals where
washed with water and dried overnight at 80 ̊C. Yield Mo3VWyOx 5.0 at.-%: 28% (0.204 g) Yield
23
Acknowledgements
After more than three months of research, I could write another thesis about the sustainable
materials and characterization research group. First of all, I will like to express my gratitude to
my daily supervisor Michelle. You learned me a lot about research scales and all the useful comments. Furthermore, I would like to thank Moniek for giving me the opportunity to do this project under her supervision. Dave, without all de advise and help during the project, it would not be possible to finish the project without you. Lukas and Ties, thank you so much for assisting me during lab work and your enthusiasm. Also Bas and JP who were always there to ask questions. Dorette, thank you for the safety advise during the project. Andrea thank you for the TGA training. I like to thank Norbert who was always there for finding autoclaves, repairing tubes and using the XRD. It was always a pleasure coming to the lab every day with such lovely and engaging people. Last but not least I would like to thank the research sustainable materials and
characterization group and the Heterogeneous catalysis and sustainable chemistry group
24
References
1. Chen, C., Kosuke, N., Murayama, T. & Ueda, W. Single-Crystalline-Phase Mo3VOx : An Efficient Catalyst for the Partial Oxidation of Acrolein to Acrylic Acid. ChemCatChem 5, 2869–2873 (2013).
2. Heid, M. et al. Dynamics of Bulk Oxygen in the Selective Oxidation of Acrolein.
ChemCatChem 9, 2390–2398 (2017).
3. Endres, S., Kampe, P., Kunert, J., Drochner, A. & Vogel, H. The influence of tungsten on structure and activity of Mo–V–W-mixed oxide catalysts for acrolein oxidation. (2007). doi:10.1016/j.apcata.2007.02.040
4. Andrushkevich, T. V. Heterogeneous Catalytic Oxidation of Acrolein to Acrylic Acid: Mechanism and Catalysts. Catal. Rev. 35, 213–259 (1993).
5. Kampe, P. et al. Heterogeneously catalysed partial oxidation of acrolein to acrylic acid— structure, function and dynamics of the V–Mo–W mixed oxides. Phys. Chem. Chem. Phys. 9, 3577–3589 (2007).
6. Qiu, C., Chen, C., Ishikawa, S., Murayama, T. & Ueda, W. Crystalline Mo-V–W-mixed Oxide with Orthorhombic and Trigonal Structures as Highly Efficient Oxidation Catalysts of Acrolein to Acrylic Acid. Top. Catal. 57, 1163–1170 (2014).
7. Giebeler, L., Wirth, A., Martens, J. A., Vogel, H. & Fuess, H. Phase transitions of V-Mo-W mixed oxides during reduction/re-oxidation cycles. Appl. Catal. A Gen. 379, 155–165 (2010).
8. Ishikawa, S. & Ueda, W. Microporous crystalline Mo–V mixed oxides for selective oxidations. Catal. Sci. Technol. 6, 617–629 (2016).
9. Ishikawa, S. & Ueda, W. Microporous Crystalline Mo‐V Mixed Oxides for Selective Oxidations. ChemInform 47, (2016).
10. Zhao, C. & Wachs, I. E. Selective oxidation of propylene to acrolein over supported V 2 O 5 /Nb 2 O 5 catalysts: An in situ Raman, IR, TPSR and kinetic study. (2006).
doi:10.1016/j.cattod.2006.07.018
11. Qiu, C. et al. Synthesis of crystalline Mo–V–W–O complex oxides with orthorhombic and trigonal structures and their application as catalysts Synthesis of crystalline Mo –V–W –O complex oxides with orthorhombic and trigonal structures and their application as catalysts. Catal. Struct. React 1, 71–77 (2015).
12. Endres, S., Kampe, P., Kunert, J., Drochner, A. & Vogel, H. The influence of tungsten on structure and activity of Mo-V-W-mixed oxide catalysts for acrolein oxidation. Appl. Catal.
A Gen. 325, 237–243 (2007).
13. Phase transitions of V-Mo-W mixed oxides during reduction/re-oxidation cycles. Appl.
Catal. A Gen. 379, 155–165 (2010).
14. Ishikawa, S. et al. Synthesis of Trigonal Mo–V–M3rd–O (M3rd = Fe, W) Catalysts by Using Structure-Directing Agent and Catalytic Performances for Selective Oxidation of Ethane.
Top. Catal. 59, 1477–1488 (2016).
15. Olga Kirilenko aus Weissrussland, D.-C., Hildebrandt Berichter, P., Lerch Berichter, M. & Ressler, P.-D. T. Structural Evolution of Ammonium Paratungstate During Thermal Decomposition.
25
16. Atkins, P. W. (Peter W. Shriver & Atkins’ inorganic chemistry, fifty edition. (Oxford University Press, 2010).
17. Pyrz, W. D. et al. Atomic-level imaging of Mo-V-O complex oxide phase intergrowth, grain boundaries, and defects using HAADF-STEM. Proc. Natl. Acad. Sci. U. S. A. 107, 6152–7 (2010).
18. Konya, T. et al. An orthorhombic Mo 3 VO x catalyst most active for oxidative
dehydrogenation of ethane among related complex metal oxides. 380 Catal. Sci. Technol.
Catal. Sci. Technol 3, 380–387 (2013).
19. Olenkova, I. P. & Plyasova, L. M. Crystal structure hexagonal MoO3. React. Kinet. Catal.
Lett’ 16, 81–85 (1981).
20. Ishikawa, S. et al. Redox Treatment of Orthorhombic Mo 29 V 11 O 112 and Relationships
between Crystal Structure, Microporosity and Catalytic Performance for Selective Oxidation of Ethane. J. Phys. Chem. C 119, 7195–7206 (2015).
21. Shallenberger, R. S. Taste Chemistry. (Springer US, 1993).
22. 6153-56-6 CAS MSDS (Oxalic acid dihydrate) Melting Point Boiling Point Density CAS Chemical Properties. Available at:
http://www.chemicalbook.com/ChemicalProductProperty_US_CB1303757.aspx. (Accessed: 7th December 2017)
23. Niemantsverdriet, J. W. & Wiley InterScience (Online service). Spectroscopy in catalysis :
an introduction. (Wiley-VCH, 2007).
24. Schimanke, G., Martin, M., Kunert, J. & Vogel, H. Characterization of Mo-V-W Mixed Oxide Catalysts byex situ andin situ X-Ray Absorption Spectroscopy. Zeitschrift f�r Anorg. und Allg. Chemie 631, 1289–1296 (2005).
25. Sadakane, M., Watanabe, N., Katou, T., Nodasaka, Y. & Ueda, W. Crystalline Mo3VOx Mixed-Metal-Oxide Catalyst with Trigonal Symmetry. Angew. Chemie 119, 1515–1518 (2007). 26. Crystallography Open Database. Available at: http://www.crystallography.net/cod/.
(Accessed: 7th December 2017)
27. ATHENA: XAS Data Processing — Athena 0.9.26 documentation. Available at:
http://bruceravel.github.io/demeter/documents/Athena/index.html. (Accessed: 7th December 2017)
26
Appendix
Appendix 1: TGA and DSC of orth-Mo3VOx.
Appendix 2: XRD pattern of the failed experiment of the synthesis of orth-Mo3VOx with PTFE
sheet.
Appendix 2: XRD pattern of the failed experiment of the synthesis of orth-Mo3VOx with PTFE
sheet.
0,00 10,00 20,00 30,00 40,00 50,00 60,00
27
Appendix 3: XRD pattern of the failed experiment of the synthesis of orth-Mo3VOx by using
seeds of previous experiments.
Appendix 4: XRD pattern of the failed experiment of the synthesis of amor-Mo3VOx
Appendix 5: XRD patterns of VO2, MoO3 and V2O5, WO2 and WO3
28
0,00 10,00 20,00 30,00 40,00 50,00 60,00 70,00
29
0,00 10,00 20,00 30,00 40,00 50,00 60,00 70,00
30
0,00 10,00 20,00 30,00 40,00 50,00 60,00 70,00 80,00
V2O5
0,00 10,00 20,00 30,00 40,00 50,00 60,00 70,00 80,00
