4.7.3 Growth mechanism
The analysis of the growth temperature range, along with the morphology of the films allows some insights into the growth mechanism of these films to be obtained. While the method used to grow the MoTe2 films here can be considered an extension of previous works on the TAC of metallic films to TMDs, the nature of the reaction interface in this case is quite different. The majority of TAC works to date refer to solid films that are converted through exposure to vapour- phase chalcogens.108,109,226 In this work, the reaction is occurring at a solid-liquid
interface, which results in a much higher chalcogen concentration at the interface. This is expected to be a large part of the reason for the quite different morphology - much larger crystal sizes and more the out-of-plane nature of these crystals.
The proposed growth mechanism is that upon heating the Te begins to melt and form an intermixed layer with Mo. As the concentration of Mo in this layer increases, MoTe2 crystals begin to nucleate and grow, Figure 4.9, in a manner analogous to the solution-liquid-solid growth mechanism.227 By examining the
results from the TEM, discussed in Section 4.7.2, it is clear that rather than the Mo film being converted, crystals of MoTe2 nucleate from the intermixed region and then grow from this. This results in the formation of two distinct layers- the crystalline MoTe2 flakes and an amorphous layer of MoTexbelow this. The growth mechanism for this system will be discussed further in regard to the growth of WTe2 in Chapter 5
4.8 Application as an electrocatalyst
TMDs have been shown to have strong electrocatalytic behaviour and the MoTe2 films grown here possess many promising characteristics in this regard.144As they
are in the 1T’ phase, they have high conductivity, which facilitates improved charge transport to the active sites.228,229 Furthermore, it has been suggested
Figure 4.9: Schematic of the proposed growth mechanism for the MoTe2 thin films
for other TMDs, such as MoS2, that in the 1T phase the basal plane becomes catalytically active.145 Finally, due to these films’ polycrystalline nature and ver-
tically oriented flakes, it is expected that there is a high percentage of under- coordinated and edge sites, both of which have been shown to improve catalytic behaviour.119,146,147 To investigate the applicability of these films, in electrocat-
alytic applications, their performance as a catalyst for the HER was examined. In this case, MoTe2 was grown on PyC which provides an ideal inert, conductive current collector due to its low reactivity and stability at high temperatures. This system of films on PyC has previously been used for the characterisation of other TMDs such as MoS2.230,231
The catalytic performance of the MoTe2 samples for the HER was measured in a three-electrode cell with 0.5 M H2SO4 as the electrolyte. A graphite counter electrode and a Ag/AgCl reference electrode were used. Additional experimental details are provided in Section 2.3.9. The activities of the various MoTe2 films are investigated by measuring LSVs in this cell. The current density measured is proportional to the hydrogen evolved, with higher current at lower applied voltage (OP) indicative of more effective catalytic behaviour.
The results of the LSV for representative films of different thickness can be seen in Figure 4.10(a). From this it is possible to extract the onset potential of the different films, defined here as the OP required to measure a current density of 0.5 mA cm-2. It is evident that the thickness of the films has a strong influence on
4.8 Application as an electrocatalyst
Figure 4.10: (a) LSV of MoTe2 films of different thickness. All thicknesses refer to the starting Mo layer thickness. Also shown is the response of bare PyC and a 30 nm Pt film. (b) Corresponding Tafel plots, with fitted slopes, of the same samples.
their effectiveness as catalysts for the HER, with the thicker films demonstrating markedly-better behaviour. The best performing of the films examined here is the 50 nm MoTe2 film, which has an onset potential of -349 mV. The thinnest 2 nm films have onset potentials of ∼-500 mV.
By plotting the log of the current density versus the OP, the Tafel plot is obtained, as shown in Figure 4(b). Taking the slope of the linear region gives the Tafel slope, which is related to the kinetics of the reaction steps involved in the evolution of hydrogen. For applications, such as HER, it is generally accepted that a lower Tafel slope is desirable as it can be understood to be the voltage increase required to increase the current density, and so hydrogen produced, by one order of magnitude.
Tafel slope values for each film are shown in Figure 4.10(b). The 30 nm and 50 nm films give the best performance by this metric with values below 70 mV dec-1. These comparatively low values indicate that after the onset potential, the
current increases quite quickly, leading to more competitive behaviour at higher voltages. The thinner 2 nm and 5 nm films, have much higher Tafel slopes, on the order of 120 mV dec-1, indicating that it is likely the Volmer reaction step that
Table 4.1: Details of the electrocatalytic properties of MoTe2 films from pub- lished literature
Reference Tafel mV dec-1 OP @ 10 mA cm-2 Note
Seok et al.233 127 -0.356 V Single crystal,
area 10-3 cm2
McGlynn et al.81 78 -0.34 V Grown film,
area 0.071 cm2
Gholamvand ∼100 ∼-0.44 V Liquid exfoliated
et al.80 sheets
Kosmala et al.232 ∼100 -0.53 V No area given
This work 67 -0.48 V area ∼1 cm2
is rate limiting in these. In Figure 4.10(a) and (b), the response of a bare PyC film can also be seen, clearly this would have a negligible effect on any of the measurements due to its very low electrocatalytic activity in the measurement window.
The performance of the 1T’ MoTe2 films is not competitive with state-of- the-art Pt catalysts (a 30 nm film of Pt on PyC measured here has an onset potential of close to 30 mV) but it is similar to the values for other relatively large-area MoTe2 electrodes reported in the literature.81,232 A common metric
used to compare different electrodes in literature is the OP required to measure 10 mA cm-2 (OP @ 10 mA cm-2) of current density. For this reason the following
section will use this metric rather than the previously defined onset potential to compare the films produced here to others reported in literature. A table of results from other published works on MoTe2 are shown in Table 4.1.
The OP @ 10 mA cm-2is higher than some reported works, however the Tafel
slopes displayed by these films compare very favourably with other works. Films of liquid-exfoliated flakes have shown Tafel slopes of∼100 mV dec-1.80Seoket al.
examined single crystals of 1T’ MoTe2, while McGlynn et al. looked at ink-jet deposited MoTe2 films, these reports showed Tafel slopes of 127 and 78 mV dec-1
4.8 Application as an electrocatalyst
unlike this work were not grown directly on the current collector. This system of TMD on PyC offers a cheaper, more scalable alternative to platinum-group catalysts.
It is evident from Figure 4.10(a) that the thickness of the films has a strong influence on their effectiveness as catalysts for HER. Further insight into this behaviour can be gleaned by examining the morphologies of the films. As seen in Section 4.7.1, from both SEM and AFM, there are the noticeable differences between the films of different thicknesses, with thicker films having increased roughness. The extracted roughness was plotted against the respective onset potential (OP needed to measure 0.5 mA cm-2) of each film, Figure 4.11(a).
There is a clear trend that initially the onset potential drops quickly as roughness increases, while above∼35 nm the onset potential changes much more slowly. It is expected that the decrease in onset potential correlates with roughness because a higher roughness indicates a more porous surface. This gives a greater effective surface area, and so, more active sites for the reaction to proceed. This increased porosity also offers improved access to the available active sites along with a reduction in mass transport limitations.
Not all surface information is captured by the roughness however, as the films change morphology it could also be expected that there is change in the type of surface sites exposed e.g. varying density of defects, edge sites or basal plane due to the changing crystal sizes and orientation. It is suggested that this is what is causing the rate at which the onset potential changes with roughness to vary over the measurement window. Further experiments, examining the type and density of active sites would be needed to confirm this. As is, it has been found that electrocatalytic behaviour improves with increasing thickness of the MoTe2 films up to a roughness of ∼35 nm, correlating to films with an initial starting Mo thickness of 20 - 30 nm. Increasing thickness beyond this does not show significant improvements in performance.
Figure 4.11: (a) Plot of the onset potential (OP required to measure 0.5 mA cm-2 of current density) versus the RMS roughness of the films. The onset po-
tential drops noticeably before changing much more slowly with roughness above
∼35 nm. (b) Chronoamperometry of 30 nm MoTe2 sample on PyC showing the measured current density as a function of time for a fixed applied potential of -0.6V with respect to the Ag/AgCl reference electrode.
The stability of these films in the electrochemical set-up was examined using chronoamperometry. A constant voltage of -0.6 V, with respect to the Ag/AgCl reference electrode, was applied to a sample and the current measured over a period of 190 minutes. The results are shown in Figure 4.11(b). The film showed a 40% drop in current density over the applied time period which may be due to mechanical damage caused to the electrode by H2 bubble formation.230