Home
Search
Collections
Journals
About
Contact us
My IOPscience
Field electron emission enhancement of graphenated MWCNTs emitters following their decoration with Au nanoparticles by a pulsed laser ablation process
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 045706 (http://iopscience.iop.org/0957-4484/26/4/045706) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 193.255.248.150 This content was downloaded on 05/02/2015 at 19:30
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 045706 (9pp)
doi:10.1088/0957-4484/26/4/045706
Field electron emission enhancement of graphenated MWCNTs emitters following their decoration with Au nanoparticles by a pulsed laser ablation process L-A Gautier, V Le Borgne, N Delegan, R Pandiyan and M A El Khakani Institut National de la Recherche Scientifique, Centre Énergie, Matériaux et Télécommunications, 1650 Blvd. Lionel–Boulet, Varennes, Quebec J3X-1S2, Canada E-mail: [email protected] Received 6 September 2014, revised 7 November 2014 Accepted for publication 9 December 2014 Published 8 January 2015 Abstract
A plasma-enhanced chemical vapor deposition (PECVD) process was adapted to alter the growth of multiwall carbon nanotubes (MWCNTs) so that graphene sheets grow out of their tips. Gold nanoparticle (Au-NP) decoration of graphenated MWCNTs (g-MWCNTs) was obtained by subsequent decoration by a pulsed laser deposition (PLD) process. By varying the number of laser ablation pulses (NLp) in the PLD process, we were able to control the size of the gold nanoparticles and the surface coverage of the decorated g-MWCNTs. The presence of Au-NPs, preferentially located at the tip of the g-MWCNTs emitters, is shown to significantly improve the field electron emission (FEE) properties of the global g-MWCNT/Au-NP nanohybrid films. Indeed, the electric field needed to extract a current density of 0.1 μA cm−2 from the g-MWCNT/ Au-NP films was decreased from 2.68 V μm−1 to a value as low as 0.96 V μm−1. On the other hand, UV photoelectron spectroscopy (UPS) characterization revealed a decrease in the global work function of the Au-decorated g-MWCNT nanohybrids compared to that of bare gMWCNT emitters. Surprisingly, the work function of g-MWCNT was found to decrease from 4.9 to 4.7 eV with the addition of Au-NPs—a value lower than the work function of both materials worth 5.2 and 4.9 eV for gold and g-MWCNT, respectively. Our results show that the NLp dependence of the FEE characteristics of the g-MWCNT/Au-NP emitters correlates well with their work function changes. Fowler-Nordheim-theory-based calculations suggest that the significant FEE enhancement of the emitters is also caused by the Au-NPs acting as nanoscale electric field enhancers. Keywords: vertically aligned carbon nanotubes, graphenated carbon nanotube, field electron emission, pulsed laser deposition, nanoparticles decoration, UV photoelectron spectroscopy (Some figures may appear in colour only in the online journal) 1. Introduction
would benefit from improved cold cathode emitters. CNTs exhibit unique properties such as high carrier mobility, high stiffness, and temperature and chemical stability that make them excellent candidates [6] for FEE. Though the basic design of CNT-based FEE devices has remained unchanged since their introduction in 1999 [1], recent work has focused on using a variety of methods to improve the FEE properties of these devices. Therefore, attention has been devoted to
For more than two decades, carbon nanotubes (CNTs) have been widely used in various research and applications fields. More recently, important attention has been given to using CNTs for field electron emission (FEE) devices such as field emission displays [1], lighting devices [2, 3], and x-ray sources [4, 5], all of which are potential applications that 0957-4484/15/045706+09$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
L-A Gautier et al
Nanotechnology 26 (2015) 045706
2. Experimental setup
FEE optimization of CNTs using substrate and/or emitter geometry tailoring. These methods typically rely on processes that create smaller-diameter emitter arrays, which enhance the electric field at the emitter tip. Geometry tailoring is shown to significantly improve the FEE properties, but such approaches generally require complex, laborious, and costly microfabrication processing [7–9]. Alternatively, other methods that rely on altering the properties of CNTs have been introduced. Through a variety of decoration approaches that can rely on either chemical [10] or physical [11, 12] processing, it is possible to add low work function (φ) materials to the emitters to improve their FEE properties. Indeed, cursory inspection of the Fowler-Nordheim equation shows that FEE current is inversely correlated to the work function. Therefore, the addition of low work function metal coatings and/or nanoparticles to the CNT emitters has been investigated. Ti (φ = 4.33 eV) [13], Hf (φ = 3.9 eV) [14], Ru (φ = 4.7 eV) [15], Cs (φ = 2.1 eV) [12], and Er (φ = 3 eV) [16] are among the metal coatings that have been reported to enhance FEE properties. For instance, by coating multiwall carbon nanotube (MWCNT) arrays with 4 nm Ti nanoparticles, the turn-on field to obtain emission was decreased by 1 V μm−1[13]. However, Ti and the other previously cited metals have poor stability under air, which leads to the formation of oxides with low thermal and electrical conductivity. As such, though the addition of low work function metal has proven to be effective in improving the FEE properties of MWCNTs-based devices, this concern has not yet been meaningfully addressed. Furthermore, the good FEE improvements reported in the literature for these low work function metals have steered attention away from the inclusion of noble metals. Noble metals have very high thermal and chemical stability as well as electrical and thermal conductivity, but they tend to have high work functions. In this work, we synthesized MWCNTs by means of plasma-enhanced chemical vapor deposition (PECVD). During this process, graphene sheets grow out of the MWCNTs, yielding graphenated MWCNTs (g-MWCNTs). This novel structure has recently been reported and exhibits good electronic and opto-electronic properties, leading to a highly transparent conductive film with an optical transmission as high as 90.4% (at 550 nm) coupled with a sheet resistance 12.5 kΩ/□ [17]. This structure could be implemented in devices such as transparent electrodes [18] and supercapacitors [19, 20]. However, it has not yet drawn attention as an FEE material. Graphene sheets that spread out from the tip of the MWCNTs act as many emission sites, improving the FEE properties of raw MWCNTs. In addition, this unique structure lends itself to efficient gold nanoparticle (Au-NP) decoration using pulsed laser deposition (PLD). In this study, we demonstrate the potential of Au-NP decoration as a method for improving the FEE properties of g-MWCNTs. We show a significant decrease in the turn-on and threshold fields that is directly correlated to the amount of Au-NPs that decorate the g-MWCNTs. The origins of these improvements are investigated in light of the Fowler-Nordheim theory and ultraviolet photoelectron spectroscopy (UPS) measurements.
FEE devices are prepared using the following multistep procedure: First, single-side polished n-doped Si (0.001–0.005 Ω.cm) wafers were coated with 20 nm Al, deposited by RF magnetron sputtering. Next, a 30 min annealing at 500 °C under air is carried out to oxidize the Al into an AlxOy buffer layer. A 25 nm Fe layer is subsequently deposited by PLD with a KrF laser (248 nm, 20 ns, 140 mJ) to serve as catalytic layer for g-MWCNT growth. The prepared Si substrates are then inserted into a home-built 13.56 MHz RF-PECVD system. The typical growth of g-MWCNT parameters are: 700 °C sample temperature, 600 mTorr pressure, RF power density of 0.88 W cm−2, and −110 V biasing. The g-MWCNTs are grown from C2H2 feedstock (10 sccm), with additional H2 (20 sccm) as a reducing agent carried in an Ar background (500 sccm) gas. After a 10 min synthesis time, vertically aligned g-MWCNTs of about 2.2 μm are obtained. The g-MWCNT/Au-NP nanohybrid structures were prepared by decorating the g-MWCNTs with 15, 30, 50, and 100 laser pulses (NLp) of gold nanoparticles. A background pressure of 300 mTorr of He was used to favor nanoparticle formation [21, 22]. A Jeol-6900 scanning electron microscope (SEM) and a Jeol 2100-F transmission electron microscope (TEM) were used to study the morphology of the carbon nanotube emitters and the size, aspect, and crystallinity of the gold nanoparticles. UPS measurements were carried out at room temperature under ultra-high vacuum using a conventional He-discharge lamp (He I: 21.22 eV) with a resolution of 0.06 eV in an Escalab 2201-XL. The position of the Fermi level was calibrated using a gold reference, and a bias of 3 V was applied to the samples to accelerate low-energy electrons. The FEE measurements were conducted with a flat copper anode fixed on a high-precision translation stage (0.14 μm precision) at a distance of 150 μm of the gMWCNT/Au-NP film. I–V curves were acquired with a Keithley 2410 source-measure unit. The FEE properties of the samples, with different NLp values, were characterized at a base pressure of 10−5 Torr. Samples were cycled several times from 0 to 800 V to ensure accurate and comparable measurements.
3. Results and discussion Figure 1 shows a typical SEM image of our MWCNT films grown by the PECVD process. The MWCNTs uniformly cover the entire silicon surface and have an average length of 2.15 μm. Close inspection of the MWCNTs shows that the tips are wrapped with few-layers graphene (FLG) structures that deploy around the MWCNTs (see the inset of figure 1). These FLGs and MWCNTs combine to form what we refer to in this paper as g-MWCNTs. With their deployed ‘wings’, these MWCNTs offer a unique nanoarchitecture with a large surface area available for decoration with nanoparticles, such as the Au-NPs used in the present study. High-resolution TEM (HRTEM) images (figure 2(a)) reveal that the MWCNT’s ‘tree trunks’ have an average diameter of ≈20 nm 2
L-A Gautier et al
Nanotechnology 26 (2015) 045706
and an interwall distance of 0.35 nm. TEM images show that the MWCNTs grow following a ‘root-growth’ mechanism. Indeed, it is possible to assert that in this work, the catalyst seed (circled in red in figure 2(b)) is at the root of the gMWCNTs, as the FLGs are known to be preferentially found near the tip. Figure 2(b) shows a typical g-MWCNT, in which one can see that the bottom part of the tubes—zone 1 in figure 2(b)—is clean and devoid of any FLG. However, graphene sheets appear to grow once the MWCNTs reach a length of ∼0.85 μm—zone 2 in figure 2(b). The graphene sheets that surround the MWCNTs spread out to distances that range from 10 to 100 nm from the tube, giving structures with a total width of about 130 nm (figure 2(c)). It has been previously reported that the formation of FLG structures is facilitated by the presence of hydrogen during the synthesis; the addition of an H2 flow and the use of H-rich carbon feedstock ensure high availability of H species. Also, longer reaction times seem to favor the growth of FLGs [23, 24]. During the MWCNTs’ growth, the hydrogen plasma increases the number of surface defects by plasma bombardment. This bombardment breaks open C-C bonds, leaving dangling bonds that are either filled with C-H or left open. The defects then act as nucleation sites available for graphene sheet growth. As more bombardment takes place at the extremity of the MWCNTs, the density of FLG structures increases along the tubes and finally covers the tip entirely. In contrast, the MWCNTs without graphenation were obtained under PECVD growth conditions, ensuring less hydrogen in the plasma (7 sccm of C2H2 and 15 sccm of H2) and softer bombardment conditions (biasing voltage of −36 V) in comparison to the growth conditions of the g-MWCNTs described above. The structural effect of the graphene layers surrounding the MWCNTs was also investigated using Raman spectroscopy. Figure 2(d) compares the Raman spectra for gMWCNTs (top curve) and for MWCNTs without FLG structures (bottom curve). The clear separation between D and G peaks located at 1350 and 1589 cm−1, respectively, as well as their narrowness, is characteristic of good and well-crystallized MWCNTs. The D/G ratios for both samples are similar and are worth 1.27 and 1.14, respectively, which denotes similar crystalline structures. From the D/G ratio, it can thus be inferred that the defects responsible for the growth of FLG structures are mostly filled during the synthesis—thus leading to FLG growth—and have little effect on the D/G ratio. The main difference between the two Raman spectra lies in the 2D peak located at about 2690 cm−1, whose intensity is seen to be ∼2 times higher in g-MWCNTs than in MWCNTs. Graphene and FLG have a very intense 2D band; pristine graphene has a 2D band several times more intense than the G band, and this intensity decreases somewhat as the number of layers increases [25]. We can therefore attribute the more intense 2D band of the g-MWCNTs to the presence of graphene and FLG near the tip of the MWCNTs. The g-MWCNTs films were decorated by means of the above-described PLD process. Figures 3(a) and (b) present the TEM images of the Au-NP-decorated g-MWCNTs at NLp = 15 and 100, respectively. Note that the Au-NPs are
Figure 1. Typical SEM image of the g-MWCNTs. (Inset) Detail of
the FLG structures that adorn the MWCNTs.
Figure 2. (a) HRTEM image of a single MWCNT. (b) TEM image of a single g-MWCNT structure, with the growth seed circled. (c) HRTEM image of FLG structures. (d) Raman spectrum of MWCNT without (bottom, black) and with (top, red) FLG structures.
3
L-A Gautier et al
Nanotechnology 26 (2015) 045706
of figure 3(a) shows an HRTEM image of an Au-NP with a constant interplanar distance of 2.4 Å. This distance is consistent with the 〈111〉 plane distance in face-centered cubic (FCC) structure of gold [27]. Neither defects nor crystal boundaries could be found, indicating that the Au-NPs have good crystalline quality. Figure 3(b) shows that at higher values of NLp, the particle size and surface density increase significantly. The Au-NPs’ size increase with NLp is illustrated in the histograms of figure 3(c). The median size of the Au-NPs is seen to increase linearly from 2 to 4.7 nm when NLp increases from 15 to 100. Increasing NLp leads to an increased amount of ablated Au species ‘landing’ on the gMWCNTs’ surface. In this cumulative deposition process, not only does the density of nucleation sites naturally increase with NLp, but also pre-existing nucleation sites (NP embryos) tend to grow in size as more material is made available through consecutive ablation pulses. The general tendency is that the NP’s size continuously increases with NLp up to NLp of a few hundreds, and then tends to saturate when a full coverage of the substrate surface is reached, as has been reported in the PLD of NPs of other materials such as Pt, CoNi, and PbS [26, 28–30]. Moreover, note that the size distributions of figure 3(c) tend to widen as NLp is increased. In fact, starting from NLp ⩾ 50, some Au nanoparticles agglomerate, creating larger elongated clusters with lengths approximately twice their width, as one can see in some of the Au-NPs in figure 3(b). Despite the enlargement of the AuNPs, the PLD decoration at NLp = 100 is still far from reaching full surface coverage and/or forming a continuous Au film on the g-MWCNTs. The fact that most of the Au-NPs are located on the FLG is thought to result from the particular structure of our g-MWCNTs, where the FLG deployed around the g-MWCNTs acts as a net, catching the deposited nanoparticles. The Au-NP-decorated g-MWCNTs films were integrated into functional devices, and their FEE properties were systematically investigated. Many sets of devices with similar gMWCNTs films (i.e., identical PECVD growth conditions leading to a similar surface morphology, as systematically assessed by SEM observations) were submitted to the same Au-NP decoration conditions and their characteristic J-E curves were repeatedly measured. Figure 4 shows a typical effect of Au-NP decoration of the g-MWCNTs emitters on their J-E curves as NLp is increased from 15 to 100. The FEE of the g-MWCNT/Au-NP nanohybrid-based devices is seen to be significantly affected by the value of NLp. Interestingly, the J-E curve of the devices decorated with NLp = 30 is clearly seen to be significantly shifted to lower fields with a more abrupt increase in the emitted current density. Consequently, NLp = 30 arises as the optimal value at which the turn-on field (Eon is defined as the field needed to extract a current density of 0.1 μA cm−2) is the lowest (0.96 V um−1 versus 2.68 V um−1 for pristine g-MWCNTs). Increasing further NLp to larger values does not continue to improve the FEE properties. In fact, at NLp = 100, Eon is seen to be slightly larger than that of g-MWCNTs, indicating that no FEE improvement occurs for NLp ⩾ 100. It is worth mentioning that the results were found to be quite reproducible, and all
Figure 3. TEM images of g-MWCNTs with (a) NLp = 15 and (b) NLp = 100 (inset) HRTEM of a single Au-NP. (c) NP diameter distribution for several values of NLp.
mostly located on the deployed graphene layers. The size and shape of the Au-NPs are directly affected by NLp, as has been shown for other PLD-deposited NPs [26]. Figure 3(a) shows rounded Au-NPs of ∼2 nm in diameter well dispersed over the surface of g-MWCNTs, obtained for NLp = 15. The inset 4
L-A Gautier et al
Nanotechnology 26 (2015) 045706
Figure 4. FEE measurements of g-MWCNTs for several values
Figure 6. Comparison of UPS spectra of g-MWCNTs with and
of NLp.
without Au-NP. (Inset) Enlargement of the Fermi level region.
local electric field enhancers given their nanometric size. This would be accounted for as a ‘geometrical’ contribution [13]. The other main possible source of this improvement would be the alteration of the electronic properties of the g-MWCNTs through the addition of gold. Indeed, the two main parameters that have an effect on the slope of the FN plots are the geometrical enhancement factor (β) and the work function. To differentiate between these two effects, UPS measurements were systematically conducted on pristine and Au-decorated g-MWCNTs. Figure 6 shows the obtained UPS spectra of the g-MWCNT/Au-NP nanohybrids for several values of NLp. The work function of each sample was calculated by subtracting the difference between the Fermi level and the cut-off energy value from the He I incident energy of 21.2 eV. The measured Fermi level stays very close to 0 eV for all the AuNP-decorated g-MWCNT samples, as one can clearly see in the inset of figure 6 (zoomed-in Fermi level region in Ylogarithmic scale). However, the cut-off energy shifts as NLp is increased. The measured work function of pristine g-MWCNTs is about 4.9 eV. As a comparison, work function values of graphene, graphite, and MWCNTs were reported to be around 4.5 eV [33], 4.75–4.85 eV [34], and 4.6–4.9 eV [34–36], respectively. At NLp = 30, the global work function of the gMWCNT/Au-NP nanohybrid was determined to be 4.7 eV, lower than the 4.9 eV measured for undecorated g-MWCNTs films. When more gold is added (NLp > 30), the work function was found to increase and reach 5.14 eV for NLp = 100 (see figure 7), a value close to that quoted for bulk Au (5.31 eV) [37]. While no chemical bonds are expected to form between Au and C [38], this work function variation could be explained by the work function difference between gMWCNT and the gold nanoparticles, which induces charge migration at the interface. Charge migration at the interface forces the Fermi levels to equilibrate, resulting in a lower effective work function. According to the work of Giovannetti et al [33], charge migration is expected to occur from graphene to Au-NPs, since the work function difference
Figure 5. Fowler-Nordheim plots of the g-MWCNTs for several
values of NLp.
the sets of samples showed that significant FEE enhancement occurs at NLp = 30. The J-E curves of our FEE devices were analyzed as a function of NLp, in light of the Fowler-Nordheim formalism [31]. Figure 5 shows the Fowler-Nordheim (FN) plot according to the following FN equation (1): J=
3⎞ ⎛ Aβ 2E 2 − Bφ 2 ⎟ exp ⎜⎜ ⎟, φ ⎝ βE ⎠
(1)
where J is the current density (A/m2), β is the field enhancement factor related to the emitter geometry, φ is the gMWCNT work function (eV), E is the electric field (V m−1), and A and B constants are worth 1.56 × 10−6 (A eV V−2) and 6.83 × 109 (V eV−3/2 m−1), respectively [32]. One can see that NLp = 30 has the smallest slope in the high field region [7], indicating that these conditions lead to the best FEE properties. The observed improvement of the FEE properties could be attributed to two main critical factors. Au-NPs could act as 5
L-A Gautier et al
Nanotechnology 26 (2015) 045706
similar trends as a function of NLp. Indeed, Eon decreases from 2.68 V μm−1 for pristine g-MWCNTs to an optimal value of 0.96 V μm−1 at NLp = 30. At higher values of NLp, Eon increases again to reach values of 1.83 V μm−1 (at NLp = 50) and 2.7 V μm−1 (at NLp = 100). A very similar behavior is observed for Eth that goes from 5.7 V μm−1 to the optimal value of 2.3 V μm−1 at NLp = 30. The behaviors of the FEE properties and work function suggest the presence of a correlation. It can be thus inferred that the lower global work function is, to some extent, responsible for the improvement of the FEE properties. As a complementary analysis, the measured work function for every NLp condition was introduced into the FN equation to derive associated β values, which are directly accessible from the slope of the FN plots in figure 5. These FN plots show that there are two regimes, each with its own slope, that lead to β values for high and low field (βHF and βLF) regions, respectively. The presence of a kink in the FN plots, which marks a deviation from the ‘normal’ linear FN behavior, has been debated in the literature, and different explanations have been offered. These include space charge effects in the vacuum spacing [40–42], interfacial resistance [43, 44], and desorption of adsorbates at low fields [45]. In fact, emission current densities obtained in this work (∼mA cm−2) remain very low in comparison with the saturation current expected for metal emission (∼106 A cm−2). This implies that we are not really reaching a suppression regime where the emitted current saturates (see figure 4). The presence of a kink in our FN plots, which separate what we referred to as ‘low field’ and ‘high field’ regions, is believed to be due to the interface barrier between CNTs and their underlying substrate (AlxOy buffer layer), rather than a real suppression regime. A double-barrier model has been proposed to explain the presence of such a knee in the FN plots of CNTs when in contact with oxide layers [46, 47], as is the case in our FEE devices. A more recent work has elegantly confirmed that the difference of the electron tunneling probability through the two barriers is responsible for the appearance of a kink in the FN law in the high fields region [43]. Thus, at low fields, the FN emission is somewhat hindered by the interface barrier between the CNTs and the underlying AlxOy oxide layer, while at high fields, electrons can overcome both barriers, leading to a higher FN tunneling. As a consequence, both βHF and βLF factors can be used to characterize the FEE behavior of MWCNTs, as previously reported in the literature by our group [7] and by others [48–50]. In any case, regardless of the absolute values, both βHF and βLF are found to show the same dependence on NLp, even if the βHF have shown a somewhat more pronounced variation (see figure 7(b)). This similarity in the relative NLp dependence of both βLF and βHF confirms that we are still far from any current saturation regime. Figure 7(b) shows that both βHF and βLF enhancement factors reach their maximum values of 8180 and 3530, respectively, at the same NLp = 30. This corresponds to an increase of β by a factor 2.8 and 2.5 compared to pristine g-MWCNT (NLp = 0). After this optimal point, βHF and βLF decrease for higher NLp values. The NLp dependence of βHF and βLF, as plotted in figure 7(b), shows
Figure 7. (a) Eon (black circles), Eth (black triangles), and work function (red squares) obtained for structures with increasing NLp values. (b) Enhancement factors (β) as a function of NLp.
between gold and CNTs is less than 0.9 eV. The electron transfer from graphene would populate the high energy levels of the Au-NPs, and consequently lower their work function. Note that the efficiency of this transfer of electrons from graphene to Au-NPs—and the resulting decrease of their work function—is found to be dependent on the Au-NP size, as it reaches its maximum for Au-NPs with a diameter of ∼2.5 nm (NLp = 30). Smaller Au-NPs have a higher surfaceto-volume ratio, favoring more charge injection [33, 39], as the C–Au interface occupies the majority of the Au-NP volume. As the Au-NPs grow in size, their surface-to-volume ratio decreases with less charge exchange (with respect to their actual volume), leading to a global work function that tends toward that of bulk gold (when the FLG will be completely covered by a continuous gold layer). In addition to size-induced effects, the surface coverage of the g-MWCNTs by Au-NPs increases as NLp is increased to 100. This means that an increasing part of the UPS signal is due to Au-NPs with respect to the carbon, leading to a higher measured work function. The effect of the addition of Au-NPs to the g-MWCNTs on their FEE characteristics and work functions are summarized in figure 7. Figure 7(a) gives Eon and Eth (Eth is defined as the threshold field needed to extract 1 mA cm−2) and the work function as a function of NLp. Interestingly, both the work function and FEE characteristics are seen to follow 6
L-A Gautier et al
Nanotechnology 26 (2015) 045706
Figure 8. Typical TEM images taken on g-MWCNT samples decorated with Au-NPs at (a) NLp = 30 and (b) NLp = 100 after being subjected to several FEE measurements. (c) Current density variation over time for g-MWCNT emitters decorated with Au-NPs at NLp = 0, 30, and 100.
that the FEE enhancements do not arise solely from the decreased work function, but mainly from geometrical considerations. Indeed, had the improvement been strictly caused by the lower work function, β would have remained unchanged with respect to g-MWCNTs. It appears that at low NLp values, the Au-NPs act as local field enhancement sites that increase βHF and βLF. For higher NLp values, screening effects between Au-NPs that can cancel out any geometrical gain from the Au-NPs would explain the observed decrease of βHF and βLF. In sum, the FEE improvements brought by the Au-NP decoration are thought to come from the combination of both lower work function and a better geometrical factor at the optimal Au-NP size and surface density. Finally, the nanostructure of the Au-NP-decorated gMWCNT emitters were examined by means of TEM observations after being subjected to several FEE sets of measurements in order to check for the presence of any change or damage induced by the FEE. Figures 8(a) and (b) show the TEM images of the g-MWCNTs samples decorated with NLp = 30 and 100, respectively, after FEE measurements. The Au-NPs remain distributed on the deployed graphene foils, as in the as-prepared samples prior to the FEE measurements.
Given the local nature of the TEM observations, it is hard to assert with precision on any possible modification of the density of Au-NPs onto the g-MWCNTs. However, these images resemble those shown in figure 3 (before FEE measurements), and no significant changes can be noticed in the nanostructure of the Au-NP-decorated g-MWCNTs after FEE measurements. This suggests that the Au-NP-decorated gMWCNT emitters developed here are structurally stable. This nanostructural stability is consistent with the FEE measurements (J-E), which were found to be fairly reproducible when repeating numerous measurement cycles on devices with the same NLp. The time stability of the FEE of our devices was also assessed over more than 2 h of continuous emission at a fixed electric field. Figure 8(c) shows that regardless of the NLp value, the general trend is that the emitted current density undergoes an initial drop over the first 5 to 20 min, and then stays more or less stable around an average value. The g-MWCNTs decorated with NLp = 30 are confirmed to give the highest current density value, even if it drops by ∼50% over the first ∼20 min and then fluctuates, with an overall tendency to decrease slightly after 2 h of current emission. The frequent rises of the emitted current, particularly in the 7
L-A Gautier et al
Nanotechnology 26 (2015) 045706
case of NLp = 30, are believed to be due to some protruding gMWCNTs that randomly align in the electric field and produce more current before they get burnt or eventually broken.
[8]
4. Conclusion
[9]
The field emission properties of vertically aligned gMWCNTs synthesised by PECVD have been improved by means of Au-NP decoration by PLD. The PLD process is shown to be very effective in decorating g-MWCNTs with Au-NPs with controllable size and surface coverage. The PLD-deposited Au-NPs are shown to have very small diameters, on the order of 2–4 nm, and to uniformly cover the FLG deployed at the top of the g-MWCNTs. We have demonstrated that this decoration has a positive effect on the turn-on and threshold FEE values, as they were decreased by factors of 2.8 and 2.5, respectively. The decrease of the field needed to extract charges from these nanohybrid structures at the optimal NLp value of 30 is attributed to a combination of a decrease in their global work function (as derived from UPS measurements) and geometry enhancement due to the presence of Au-NPs at the tips of the emitters.
[10]
[11]
[12]
[13] [14]
[15]
Acknowledgments The authors would like to acknowledge financial support from the Natural Science and Engineering Research Council (NSERC) of Canada, Le Fonds de Recherche du QuébecNature et Technologies (FRQNT) through its strategic Network ‘Plasma-Québec’, and Nano-Québec (the Québec Organization for the promotion of nanoscience and nanotechnologies).
[16] [17] [18]
References
[19]
[1] Choi W B et al 1999 Fully sealed, high-brightness carbonnanotube field-emission display Appl. Phys. Lett. 75 3129 [2] Bonard J-M, Stöckli T, Noury O and Châtelain A 2001 Field emission from cylindrical carbon nanotube cathodes: possibilities for luminescent tubes Appl. Phys. Lett. 78 2775 [3] Cho W-S, Lee H-J, Lee Y-D, Park J-H, Kim J-K, Lee Y-H and Ju B-K 2007 Carbon nanotube-based triode field emission lamps using metal meshes with spacers IEEE Electron Device Lett. 28 386–8 [4] Yue G Z, Qiu Q, Gao B, Cheng Y, Zhang J, Shimoda H, Chang S, Lu J P and Zhou O 2002 Generation of continuous and pulsed diagnostic imaging x-ray radiation using a carbon-nanotube-based field-emission cathode Appl. Phys. Lett. 81 355 [5] Jeong J-W, Kang J-T, Choi S, Kim J-W, Ahn S and Song Y-H 2013 A digital miniature x-ray tube with a high-density triode carbon nanotube field emitter Appl. Phys. Lett. 102 023504 [6] Popov V 2004 Carbon nanotubes: properties and application Mater. Sci. Eng. Rep. 43 61–102 [7] Gautier L-A, Le Borgne V, Al Moussalami S and El Khakani M A 2014 Enhanced field electron emission
[20]
[21] [22]
[23]
[24]
8
properties of hierarchically structured MWCNT-based cold cathodes Nanoscale Res. Lett. 9 55 Sato H, Haruki K, Watanabe M, Hata K and Saito Y 2012 Effect of geometry of a vertically aligned carbon nanotube pillar array on its field-emission properties Surf. Interface Anal. 44 776–9 Hazra K S, Gigras T and Misra D S 2011 Tailoring the electrostatic screening effect during field emission from hollow multiwalled carbon nanotube pillars Appl. Phys. Lett. 98 123116 Yu C, Ryu Y, Yin L and Yang H 2011 Modulating eletronic transport properties of carbon nanotubes to improve the thermoelectric power factor via nanoparticle decoration ACS Nano 5 1297–303 Chen Y M, Chen C A, Huang Y S, Lee K Y and Tiong K K 2010 Characterization and enhanced field emission properties of IrO2-coated carbon nanotube bundle arrays Nanotechnology 21 035702 Shahi M, Gautam S, Shah P V, Rawat J S, Chaudhury P K and Tandon R P 2013 Decoration of cesium iodide nano particles on patterned carbon nanotube emitter arrays to improve their field emission J. Nanoparticle Res. 15 1497 Zuo Y, Ren Y, Wang Z, Han X and Xi L 2013 Enhanced field emission and hysteresis characteristics of aligned carbon nanotubes with Ti decoration Org. Electron. 14 2306–14 Zhang J, Yang C, Wang Y, Feng T, Yu W, Jiang J, Wang X and Liu X 2006 Improvement of the field emission of carbon nanotubes by hafnium coating and annealing Nanotechnology 17 257–60 Liu C, Kim K S, Baek J, Cho Y, Han S, Kim S-W, Min N-K, Choi Y, Kim J-U and Lee C J 2009 Improved field emission properties of double-walled carbon nanotubes decorated with Ru nanoparticles Carbon 47 1158–64 Shrestha S, Choi W C, Song W, Kwon Y T, Shrestha S P and Park C-Y 2010 Preparation and field emission properties of Er-decorated multiwalled carbon nanotubes Carbon 48 54–9 Yu K, Lu G, Bo Z, Mao S and Chen J 2011 Carbon nanotube with chemically bonded graphene leaves for electronic and optoelectronic applications J. Phys. Chem. Lett. 2 1556–62 Tung V C, Chen L-M, Allen M J, Wassei J K, Nelson K, Kaner R B and Yang Y 2009 Low-temperature solution processing of graphene-carbon nanotube hybrid materials for high-performance transparent conductors Nano Lett. 9 1949–55 Yu D and Dai L 2010 Self-assembled graphene/carbon nanotube hybrid films for supercapacitors J. Phys. Chem. Lett. 1 467–70 Fan Z, Yan J, Zhi L, Zhang Q, Wei T, Feng J, Zhang M, Qian W and Wei F 2010 A three-dimensional carbon nanotube/graphene sandwich and its application as electrode in supercapacitors Adv. Mater. 22 3723–8 Yi J H and El Khakani M A 2005 Highly oriented planar arrays of SWCNTs grown onto HOPG substrates by means of an ‘all-laser’ process Chem. Phys. Lett. 413 182–7 El Khakani M A and Yi J H 2004 The nanostructure and electrical properties of SWNT bundle networks grown by an all-laser growth process for nanoelectronic device applications Nanotechnology 15 S534–9 Zeng L, Lei D, Wang W, Liang J, Wang Z, Yao N and Zhang B 2008 Preparation of carbon nanosheets deposited on carbon nanotubes by microwave plasma-enhanced chemical vapor deposition method Appl. Surf. Sci. 254 1700–4 Ubnoske S M, Raut A S, Brown B, Parker C B, Stoner B R and Glass J T 2014 Perspectives on the growth of high edge density carbon nanostructures: transitions from vertically oriented graphene nanosheets to graphenated carbon nanotubes J. Phys. Chem. C. Nanomater. Interfaces 118 16126–32
L-A Gautier et al
Nanotechnology 26 (2015) 045706
[25] Ferrari A C et al 2006 Raman spectrum of graphene and graphene layers Phys. Rev. Lett. 97 187401 [26] Ka I, Ma D and El Khakani M A 2011 Tailoring the photoluminescence of PbS-nanoparticles layers deposited by means of the pulsed laser ablation technique J. Nanoparticle Res. 13 2269–74 [27] Stoeva S I, Prasad B L V, Uma S, Stoimenov P K, Zaikovski V, Sorensen C M and Klabunde K J 2003 Facecentered cubic and hexagonal closed-packed nanocrystal superlattices of gold nanoparticles prepared by different methods J. Phys. Chem. B 107 7441–8 [28] Dolbec R, Irissou E, Chaker M, Guay D, Rosei F and El Khakani M A 2004 Growth dynamics of pulsed laser deposited Pt nanoparticles on highly oriented pyrolitic graphite substrates Phys. Rev. B 70 201406 [29] Hamoudi Z, El Khakani M A and Mohamedi M 2012 Influence of pressure on the structural and electrocatalytic properties of pt nanoparticles grown by pulsed laser ablation onto carbon paper substrate Int. J. Electrochem. Sci. 7 1666–76 [30] Aïssa B, Therriault D and El Khakani M A 2011 On-substrate growth of single-walled carbon nanotube networks by an ‘all-laser’ processing route Carbon 49 2795–808 [31] Fowler R H and Nordheim L 1928 Electron emission in intense electric fields Proc. R. Soc. A Math. Phys. Eng. Sci. 119 173–81 [32] Mayer A 2011 Exact solutions for the field electron emission achieved from a flat metal using the standard Fowler– Nordheim equation with a correction factor that accounts for the electric field, the work function, and the Fermi energy of the emitter J. Vac. Sci. Technol. B 29 021803 [33] Giovannetti G, Khomyakov P, Brocks G, Karpan V, van den Brink J and Kelly P 2008 Doping graphene with metal contacts Phys. Rev. Lett. 101 026803 [34] Shiraishi M and Ata M 2001 Work function of carbon nanotubes Carbon 39 1913–7 [35] Ago H, Kugler T, Cacialli F, Salaneck W R, Shaffer M S P, Windle A H and Friend R H 1999 Work functions and surface functional groups of multiwall carbon nanotubes J. Phys. Chem. B 103 8116–21 [36] Su W, Leung T and Chan C 2007 Work function of singlewalled and multiwalled carbon nanotubes: first-principles study Phys. Rev. B 76 235413 [37] Lide D R 2005 CRC Handbook of Chemistry and Physics (Boca Raton, FL: CRC Press) [38] Bittencourt C, Felten A, Douhard B, Ghijsen J, Johnson R L, Drube W and Pireaux J-J 2006 Photoemission studies of
[39]
[40]
[41] [42]
[43]
[44] [45]
[46]
[47]
[48]
[49] [50]
9
gold clusters thermally evaporated on multiwall carbon nanotubes Chem. Phys. 328 385–91 Khoa N T, Kim S W, Yoo D-H, Kim E J and Hahn S H 2014 Size-dependent work function and catalytic performance of gold nanoparticles decorated graphene oxide sheets Appl. Catal. A Gen. 469 159–64 Xu N S, Chen Y, Deng S Z, Chen J, Ma X C and Wang E G 2001 Vacuum gap dependence of field electron emission properties of large area multi-walled carbon nanotube films J. Phys. D: Appl. Phys. 34 1597–601 Barbour J, Dolan W, Trolan J, Martin E and Dyke W 1953 Space-charge effects in field emission Phys. Rev. 92 45–51 Chen Y, Deng S Z, Xu N S, Chen J, Ma X C and Wang E G 2002 Physical origin of non-linearity in Fowler–Nordheim plots of aligned large area multi-walled nitrogen-containing carbon nanotubes Mater. Sci. Eng. A 327 16–9 Chen L F, Song H, Cao L Z, Jiang H, Li D B, Guo W G, Liu X, Zhao H F and Li Z M 2009 Effect of interface barrier between carbon nanotube film and substrate on field emission J. Appl. Phys. 106 033703 Bonard J-M, Klinke C, Dean K and Coll B 2003 Degradation and failure of carbon nanotube field emitters Phys. Rev. B 67 115406 Kim D-H, Yang H-S, Kang H-D and Lee H-R 2003 Novel emission degradation behavior of patterned carbon nanotubes by field emission Chem. Phys. Lett. 368 439–44 Zhang J, Yang C, Yang W, Feng T, Wang X and Liu X 2006 Appearance of a knee on the Fowler–Nordheim plot of carbon nanotubes on a substrate Solid State Commun. 138 13–6 Zhang J, Wang X, Yang W, Yu W, Feng T, Li Q, Liu X and Yang C 2006 Interaction between carbon nanotubes and substrate and its implication on field emission mechanism Carbon 44 418–22 Neupane S, Lastres M, Chiarella M, Li W, Su Q and Du G 2012 Synthesis and field emission properties of vertically aligned carbon nanotube arrays on copper Carbon 50 2641–50 Altman I S, Pikhitsa P V and Choi M 2004 Electron field emission from nanocarbons: a two-process model Appl. Phys. Lett. 84 1126 Rakhi R B, Reddy A L M, Shaijumon M M, Sethupathi K and Ramaprabhu S 2007 Electron field emitters based on multiwalled carbon nanotubes decorated with nanoscale metal clusters J. Nanoparticle Res. 10 179–89
Search
Collections
Journals
About
Contact us
My IOPscience
Field electron emission enhancement of graphenated MWCNTs emitters following their decoration with Au nanoparticles by a pulsed laser ablation process
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Nanotechnology 26 045706 (http://iopscience.iop.org/0957-4484/26/4/045706) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 193.255.248.150 This content was downloaded on 05/02/2015 at 19:30
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 26 (2015) 045706 (9pp)
doi:10.1088/0957-4484/26/4/045706
Field electron emission enhancement of graphenated MWCNTs emitters following their decoration with Au nanoparticles by a pulsed laser ablation process L-A Gautier, V Le Borgne, N Delegan, R Pandiyan and M A El Khakani Institut National de la Recherche Scientifique, Centre Énergie, Matériaux et Télécommunications, 1650 Blvd. Lionel–Boulet, Varennes, Quebec J3X-1S2, Canada E-mail: [email protected] Received 6 September 2014, revised 7 November 2014 Accepted for publication 9 December 2014 Published 8 January 2015 Abstract
A plasma-enhanced chemical vapor deposition (PECVD) process was adapted to alter the growth of multiwall carbon nanotubes (MWCNTs) so that graphene sheets grow out of their tips. Gold nanoparticle (Au-NP) decoration of graphenated MWCNTs (g-MWCNTs) was obtained by subsequent decoration by a pulsed laser deposition (PLD) process. By varying the number of laser ablation pulses (NLp) in the PLD process, we were able to control the size of the gold nanoparticles and the surface coverage of the decorated g-MWCNTs. The presence of Au-NPs, preferentially located at the tip of the g-MWCNTs emitters, is shown to significantly improve the field electron emission (FEE) properties of the global g-MWCNT/Au-NP nanohybrid films. Indeed, the electric field needed to extract a current density of 0.1 μA cm−2 from the g-MWCNT/ Au-NP films was decreased from 2.68 V μm−1 to a value as low as 0.96 V μm−1. On the other hand, UV photoelectron spectroscopy (UPS) characterization revealed a decrease in the global work function of the Au-decorated g-MWCNT nanohybrids compared to that of bare gMWCNT emitters. Surprisingly, the work function of g-MWCNT was found to decrease from 4.9 to 4.7 eV with the addition of Au-NPs—a value lower than the work function of both materials worth 5.2 and 4.9 eV for gold and g-MWCNT, respectively. Our results show that the NLp dependence of the FEE characteristics of the g-MWCNT/Au-NP emitters correlates well with their work function changes. Fowler-Nordheim-theory-based calculations suggest that the significant FEE enhancement of the emitters is also caused by the Au-NPs acting as nanoscale electric field enhancers. Keywords: vertically aligned carbon nanotubes, graphenated carbon nanotube, field electron emission, pulsed laser deposition, nanoparticles decoration, UV photoelectron spectroscopy (Some figures may appear in colour only in the online journal) 1. Introduction
would benefit from improved cold cathode emitters. CNTs exhibit unique properties such as high carrier mobility, high stiffness, and temperature and chemical stability that make them excellent candidates [6] for FEE. Though the basic design of CNT-based FEE devices has remained unchanged since their introduction in 1999 [1], recent work has focused on using a variety of methods to improve the FEE properties of these devices. Therefore, attention has been devoted to
For more than two decades, carbon nanotubes (CNTs) have been widely used in various research and applications fields. More recently, important attention has been given to using CNTs for field electron emission (FEE) devices such as field emission displays [1], lighting devices [2, 3], and x-ray sources [4, 5], all of which are potential applications that 0957-4484/15/045706+09$33.00
1
© 2015 IOP Publishing Ltd Printed in the UK
L-A Gautier et al
Nanotechnology 26 (2015) 045706
2. Experimental setup
FEE optimization of CNTs using substrate and/or emitter geometry tailoring. These methods typically rely on processes that create smaller-diameter emitter arrays, which enhance the electric field at the emitter tip. Geometry tailoring is shown to significantly improve the FEE properties, but such approaches generally require complex, laborious, and costly microfabrication processing [7–9]. Alternatively, other methods that rely on altering the properties of CNTs have been introduced. Through a variety of decoration approaches that can rely on either chemical [10] or physical [11, 12] processing, it is possible to add low work function (φ) materials to the emitters to improve their FEE properties. Indeed, cursory inspection of the Fowler-Nordheim equation shows that FEE current is inversely correlated to the work function. Therefore, the addition of low work function metal coatings and/or nanoparticles to the CNT emitters has been investigated. Ti (φ = 4.33 eV) [13], Hf (φ = 3.9 eV) [14], Ru (φ = 4.7 eV) [15], Cs (φ = 2.1 eV) [12], and Er (φ = 3 eV) [16] are among the metal coatings that have been reported to enhance FEE properties. For instance, by coating multiwall carbon nanotube (MWCNT) arrays with 4 nm Ti nanoparticles, the turn-on field to obtain emission was decreased by 1 V μm−1[13]. However, Ti and the other previously cited metals have poor stability under air, which leads to the formation of oxides with low thermal and electrical conductivity. As such, though the addition of low work function metal has proven to be effective in improving the FEE properties of MWCNTs-based devices, this concern has not yet been meaningfully addressed. Furthermore, the good FEE improvements reported in the literature for these low work function metals have steered attention away from the inclusion of noble metals. Noble metals have very high thermal and chemical stability as well as electrical and thermal conductivity, but they tend to have high work functions. In this work, we synthesized MWCNTs by means of plasma-enhanced chemical vapor deposition (PECVD). During this process, graphene sheets grow out of the MWCNTs, yielding graphenated MWCNTs (g-MWCNTs). This novel structure has recently been reported and exhibits good electronic and opto-electronic properties, leading to a highly transparent conductive film with an optical transmission as high as 90.4% (at 550 nm) coupled with a sheet resistance 12.5 kΩ/□ [17]. This structure could be implemented in devices such as transparent electrodes [18] and supercapacitors [19, 20]. However, it has not yet drawn attention as an FEE material. Graphene sheets that spread out from the tip of the MWCNTs act as many emission sites, improving the FEE properties of raw MWCNTs. In addition, this unique structure lends itself to efficient gold nanoparticle (Au-NP) decoration using pulsed laser deposition (PLD). In this study, we demonstrate the potential of Au-NP decoration as a method for improving the FEE properties of g-MWCNTs. We show a significant decrease in the turn-on and threshold fields that is directly correlated to the amount of Au-NPs that decorate the g-MWCNTs. The origins of these improvements are investigated in light of the Fowler-Nordheim theory and ultraviolet photoelectron spectroscopy (UPS) measurements.
FEE devices are prepared using the following multistep procedure: First, single-side polished n-doped Si (0.001–0.005 Ω.cm) wafers were coated with 20 nm Al, deposited by RF magnetron sputtering. Next, a 30 min annealing at 500 °C under air is carried out to oxidize the Al into an AlxOy buffer layer. A 25 nm Fe layer is subsequently deposited by PLD with a KrF laser (248 nm, 20 ns, 140 mJ) to serve as catalytic layer for g-MWCNT growth. The prepared Si substrates are then inserted into a home-built 13.56 MHz RF-PECVD system. The typical growth of g-MWCNT parameters are: 700 °C sample temperature, 600 mTorr pressure, RF power density of 0.88 W cm−2, and −110 V biasing. The g-MWCNTs are grown from C2H2 feedstock (10 sccm), with additional H2 (20 sccm) as a reducing agent carried in an Ar background (500 sccm) gas. After a 10 min synthesis time, vertically aligned g-MWCNTs of about 2.2 μm are obtained. The g-MWCNT/Au-NP nanohybrid structures were prepared by decorating the g-MWCNTs with 15, 30, 50, and 100 laser pulses (NLp) of gold nanoparticles. A background pressure of 300 mTorr of He was used to favor nanoparticle formation [21, 22]. A Jeol-6900 scanning electron microscope (SEM) and a Jeol 2100-F transmission electron microscope (TEM) were used to study the morphology of the carbon nanotube emitters and the size, aspect, and crystallinity of the gold nanoparticles. UPS measurements were carried out at room temperature under ultra-high vacuum using a conventional He-discharge lamp (He I: 21.22 eV) with a resolution of 0.06 eV in an Escalab 2201-XL. The position of the Fermi level was calibrated using a gold reference, and a bias of 3 V was applied to the samples to accelerate low-energy electrons. The FEE measurements were conducted with a flat copper anode fixed on a high-precision translation stage (0.14 μm precision) at a distance of 150 μm of the gMWCNT/Au-NP film. I–V curves were acquired with a Keithley 2410 source-measure unit. The FEE properties of the samples, with different NLp values, were characterized at a base pressure of 10−5 Torr. Samples were cycled several times from 0 to 800 V to ensure accurate and comparable measurements.
3. Results and discussion Figure 1 shows a typical SEM image of our MWCNT films grown by the PECVD process. The MWCNTs uniformly cover the entire silicon surface and have an average length of 2.15 μm. Close inspection of the MWCNTs shows that the tips are wrapped with few-layers graphene (FLG) structures that deploy around the MWCNTs (see the inset of figure 1). These FLGs and MWCNTs combine to form what we refer to in this paper as g-MWCNTs. With their deployed ‘wings’, these MWCNTs offer a unique nanoarchitecture with a large surface area available for decoration with nanoparticles, such as the Au-NPs used in the present study. High-resolution TEM (HRTEM) images (figure 2(a)) reveal that the MWCNT’s ‘tree trunks’ have an average diameter of ≈20 nm 2
L-A Gautier et al
Nanotechnology 26 (2015) 045706
and an interwall distance of 0.35 nm. TEM images show that the MWCNTs grow following a ‘root-growth’ mechanism. Indeed, it is possible to assert that in this work, the catalyst seed (circled in red in figure 2(b)) is at the root of the gMWCNTs, as the FLGs are known to be preferentially found near the tip. Figure 2(b) shows a typical g-MWCNT, in which one can see that the bottom part of the tubes—zone 1 in figure 2(b)—is clean and devoid of any FLG. However, graphene sheets appear to grow once the MWCNTs reach a length of ∼0.85 μm—zone 2 in figure 2(b). The graphene sheets that surround the MWCNTs spread out to distances that range from 10 to 100 nm from the tube, giving structures with a total width of about 130 nm (figure 2(c)). It has been previously reported that the formation of FLG structures is facilitated by the presence of hydrogen during the synthesis; the addition of an H2 flow and the use of H-rich carbon feedstock ensure high availability of H species. Also, longer reaction times seem to favor the growth of FLGs [23, 24]. During the MWCNTs’ growth, the hydrogen plasma increases the number of surface defects by plasma bombardment. This bombardment breaks open C-C bonds, leaving dangling bonds that are either filled with C-H or left open. The defects then act as nucleation sites available for graphene sheet growth. As more bombardment takes place at the extremity of the MWCNTs, the density of FLG structures increases along the tubes and finally covers the tip entirely. In contrast, the MWCNTs without graphenation were obtained under PECVD growth conditions, ensuring less hydrogen in the plasma (7 sccm of C2H2 and 15 sccm of H2) and softer bombardment conditions (biasing voltage of −36 V) in comparison to the growth conditions of the g-MWCNTs described above. The structural effect of the graphene layers surrounding the MWCNTs was also investigated using Raman spectroscopy. Figure 2(d) compares the Raman spectra for gMWCNTs (top curve) and for MWCNTs without FLG structures (bottom curve). The clear separation between D and G peaks located at 1350 and 1589 cm−1, respectively, as well as their narrowness, is characteristic of good and well-crystallized MWCNTs. The D/G ratios for both samples are similar and are worth 1.27 and 1.14, respectively, which denotes similar crystalline structures. From the D/G ratio, it can thus be inferred that the defects responsible for the growth of FLG structures are mostly filled during the synthesis—thus leading to FLG growth—and have little effect on the D/G ratio. The main difference between the two Raman spectra lies in the 2D peak located at about 2690 cm−1, whose intensity is seen to be ∼2 times higher in g-MWCNTs than in MWCNTs. Graphene and FLG have a very intense 2D band; pristine graphene has a 2D band several times more intense than the G band, and this intensity decreases somewhat as the number of layers increases [25]. We can therefore attribute the more intense 2D band of the g-MWCNTs to the presence of graphene and FLG near the tip of the MWCNTs. The g-MWCNTs films were decorated by means of the above-described PLD process. Figures 3(a) and (b) present the TEM images of the Au-NP-decorated g-MWCNTs at NLp = 15 and 100, respectively. Note that the Au-NPs are
Figure 1. Typical SEM image of the g-MWCNTs. (Inset) Detail of
the FLG structures that adorn the MWCNTs.
Figure 2. (a) HRTEM image of a single MWCNT. (b) TEM image of a single g-MWCNT structure, with the growth seed circled. (c) HRTEM image of FLG structures. (d) Raman spectrum of MWCNT without (bottom, black) and with (top, red) FLG structures.
3
L-A Gautier et al
Nanotechnology 26 (2015) 045706
of figure 3(a) shows an HRTEM image of an Au-NP with a constant interplanar distance of 2.4 Å. This distance is consistent with the 〈111〉 plane distance in face-centered cubic (FCC) structure of gold [27]. Neither defects nor crystal boundaries could be found, indicating that the Au-NPs have good crystalline quality. Figure 3(b) shows that at higher values of NLp, the particle size and surface density increase significantly. The Au-NPs’ size increase with NLp is illustrated in the histograms of figure 3(c). The median size of the Au-NPs is seen to increase linearly from 2 to 4.7 nm when NLp increases from 15 to 100. Increasing NLp leads to an increased amount of ablated Au species ‘landing’ on the gMWCNTs’ surface. In this cumulative deposition process, not only does the density of nucleation sites naturally increase with NLp, but also pre-existing nucleation sites (NP embryos) tend to grow in size as more material is made available through consecutive ablation pulses. The general tendency is that the NP’s size continuously increases with NLp up to NLp of a few hundreds, and then tends to saturate when a full coverage of the substrate surface is reached, as has been reported in the PLD of NPs of other materials such as Pt, CoNi, and PbS [26, 28–30]. Moreover, note that the size distributions of figure 3(c) tend to widen as NLp is increased. In fact, starting from NLp ⩾ 50, some Au nanoparticles agglomerate, creating larger elongated clusters with lengths approximately twice their width, as one can see in some of the Au-NPs in figure 3(b). Despite the enlargement of the AuNPs, the PLD decoration at NLp = 100 is still far from reaching full surface coverage and/or forming a continuous Au film on the g-MWCNTs. The fact that most of the Au-NPs are located on the FLG is thought to result from the particular structure of our g-MWCNTs, where the FLG deployed around the g-MWCNTs acts as a net, catching the deposited nanoparticles. The Au-NP-decorated g-MWCNTs films were integrated into functional devices, and their FEE properties were systematically investigated. Many sets of devices with similar gMWCNTs films (i.e., identical PECVD growth conditions leading to a similar surface morphology, as systematically assessed by SEM observations) were submitted to the same Au-NP decoration conditions and their characteristic J-E curves were repeatedly measured. Figure 4 shows a typical effect of Au-NP decoration of the g-MWCNTs emitters on their J-E curves as NLp is increased from 15 to 100. The FEE of the g-MWCNT/Au-NP nanohybrid-based devices is seen to be significantly affected by the value of NLp. Interestingly, the J-E curve of the devices decorated with NLp = 30 is clearly seen to be significantly shifted to lower fields with a more abrupt increase in the emitted current density. Consequently, NLp = 30 arises as the optimal value at which the turn-on field (Eon is defined as the field needed to extract a current density of 0.1 μA cm−2) is the lowest (0.96 V um−1 versus 2.68 V um−1 for pristine g-MWCNTs). Increasing further NLp to larger values does not continue to improve the FEE properties. In fact, at NLp = 100, Eon is seen to be slightly larger than that of g-MWCNTs, indicating that no FEE improvement occurs for NLp ⩾ 100. It is worth mentioning that the results were found to be quite reproducible, and all
Figure 3. TEM images of g-MWCNTs with (a) NLp = 15 and (b) NLp = 100 (inset) HRTEM of a single Au-NP. (c) NP diameter distribution for several values of NLp.
mostly located on the deployed graphene layers. The size and shape of the Au-NPs are directly affected by NLp, as has been shown for other PLD-deposited NPs [26]. Figure 3(a) shows rounded Au-NPs of ∼2 nm in diameter well dispersed over the surface of g-MWCNTs, obtained for NLp = 15. The inset 4
L-A Gautier et al
Nanotechnology 26 (2015) 045706
Figure 4. FEE measurements of g-MWCNTs for several values
Figure 6. Comparison of UPS spectra of g-MWCNTs with and
of NLp.
without Au-NP. (Inset) Enlargement of the Fermi level region.
local electric field enhancers given their nanometric size. This would be accounted for as a ‘geometrical’ contribution [13]. The other main possible source of this improvement would be the alteration of the electronic properties of the g-MWCNTs through the addition of gold. Indeed, the two main parameters that have an effect on the slope of the FN plots are the geometrical enhancement factor (β) and the work function. To differentiate between these two effects, UPS measurements were systematically conducted on pristine and Au-decorated g-MWCNTs. Figure 6 shows the obtained UPS spectra of the g-MWCNT/Au-NP nanohybrids for several values of NLp. The work function of each sample was calculated by subtracting the difference between the Fermi level and the cut-off energy value from the He I incident energy of 21.2 eV. The measured Fermi level stays very close to 0 eV for all the AuNP-decorated g-MWCNT samples, as one can clearly see in the inset of figure 6 (zoomed-in Fermi level region in Ylogarithmic scale). However, the cut-off energy shifts as NLp is increased. The measured work function of pristine g-MWCNTs is about 4.9 eV. As a comparison, work function values of graphene, graphite, and MWCNTs were reported to be around 4.5 eV [33], 4.75–4.85 eV [34], and 4.6–4.9 eV [34–36], respectively. At NLp = 30, the global work function of the gMWCNT/Au-NP nanohybrid was determined to be 4.7 eV, lower than the 4.9 eV measured for undecorated g-MWCNTs films. When more gold is added (NLp > 30), the work function was found to increase and reach 5.14 eV for NLp = 100 (see figure 7), a value close to that quoted for bulk Au (5.31 eV) [37]. While no chemical bonds are expected to form between Au and C [38], this work function variation could be explained by the work function difference between gMWCNT and the gold nanoparticles, which induces charge migration at the interface. Charge migration at the interface forces the Fermi levels to equilibrate, resulting in a lower effective work function. According to the work of Giovannetti et al [33], charge migration is expected to occur from graphene to Au-NPs, since the work function difference
Figure 5. Fowler-Nordheim plots of the g-MWCNTs for several
values of NLp.
the sets of samples showed that significant FEE enhancement occurs at NLp = 30. The J-E curves of our FEE devices were analyzed as a function of NLp, in light of the Fowler-Nordheim formalism [31]. Figure 5 shows the Fowler-Nordheim (FN) plot according to the following FN equation (1): J=
3⎞ ⎛ Aβ 2E 2 − Bφ 2 ⎟ exp ⎜⎜ ⎟, φ ⎝ βE ⎠
(1)
where J is the current density (A/m2), β is the field enhancement factor related to the emitter geometry, φ is the gMWCNT work function (eV), E is the electric field (V m−1), and A and B constants are worth 1.56 × 10−6 (A eV V−2) and 6.83 × 109 (V eV−3/2 m−1), respectively [32]. One can see that NLp = 30 has the smallest slope in the high field region [7], indicating that these conditions lead to the best FEE properties. The observed improvement of the FEE properties could be attributed to two main critical factors. Au-NPs could act as 5
L-A Gautier et al
Nanotechnology 26 (2015) 045706
similar trends as a function of NLp. Indeed, Eon decreases from 2.68 V μm−1 for pristine g-MWCNTs to an optimal value of 0.96 V μm−1 at NLp = 30. At higher values of NLp, Eon increases again to reach values of 1.83 V μm−1 (at NLp = 50) and 2.7 V μm−1 (at NLp = 100). A very similar behavior is observed for Eth that goes from 5.7 V μm−1 to the optimal value of 2.3 V μm−1 at NLp = 30. The behaviors of the FEE properties and work function suggest the presence of a correlation. It can be thus inferred that the lower global work function is, to some extent, responsible for the improvement of the FEE properties. As a complementary analysis, the measured work function for every NLp condition was introduced into the FN equation to derive associated β values, which are directly accessible from the slope of the FN plots in figure 5. These FN plots show that there are two regimes, each with its own slope, that lead to β values for high and low field (βHF and βLF) regions, respectively. The presence of a kink in the FN plots, which marks a deviation from the ‘normal’ linear FN behavior, has been debated in the literature, and different explanations have been offered. These include space charge effects in the vacuum spacing [40–42], interfacial resistance [43, 44], and desorption of adsorbates at low fields [45]. In fact, emission current densities obtained in this work (∼mA cm−2) remain very low in comparison with the saturation current expected for metal emission (∼106 A cm−2). This implies that we are not really reaching a suppression regime where the emitted current saturates (see figure 4). The presence of a kink in our FN plots, which separate what we referred to as ‘low field’ and ‘high field’ regions, is believed to be due to the interface barrier between CNTs and their underlying substrate (AlxOy buffer layer), rather than a real suppression regime. A double-barrier model has been proposed to explain the presence of such a knee in the FN plots of CNTs when in contact with oxide layers [46, 47], as is the case in our FEE devices. A more recent work has elegantly confirmed that the difference of the electron tunneling probability through the two barriers is responsible for the appearance of a kink in the FN law in the high fields region [43]. Thus, at low fields, the FN emission is somewhat hindered by the interface barrier between the CNTs and the underlying AlxOy oxide layer, while at high fields, electrons can overcome both barriers, leading to a higher FN tunneling. As a consequence, both βHF and βLF factors can be used to characterize the FEE behavior of MWCNTs, as previously reported in the literature by our group [7] and by others [48–50]. In any case, regardless of the absolute values, both βHF and βLF are found to show the same dependence on NLp, even if the βHF have shown a somewhat more pronounced variation (see figure 7(b)). This similarity in the relative NLp dependence of both βLF and βHF confirms that we are still far from any current saturation regime. Figure 7(b) shows that both βHF and βLF enhancement factors reach their maximum values of 8180 and 3530, respectively, at the same NLp = 30. This corresponds to an increase of β by a factor 2.8 and 2.5 compared to pristine g-MWCNT (NLp = 0). After this optimal point, βHF and βLF decrease for higher NLp values. The NLp dependence of βHF and βLF, as plotted in figure 7(b), shows
Figure 7. (a) Eon (black circles), Eth (black triangles), and work function (red squares) obtained for structures with increasing NLp values. (b) Enhancement factors (β) as a function of NLp.
between gold and CNTs is less than 0.9 eV. The electron transfer from graphene would populate the high energy levels of the Au-NPs, and consequently lower their work function. Note that the efficiency of this transfer of electrons from graphene to Au-NPs—and the resulting decrease of their work function—is found to be dependent on the Au-NP size, as it reaches its maximum for Au-NPs with a diameter of ∼2.5 nm (NLp = 30). Smaller Au-NPs have a higher surfaceto-volume ratio, favoring more charge injection [33, 39], as the C–Au interface occupies the majority of the Au-NP volume. As the Au-NPs grow in size, their surface-to-volume ratio decreases with less charge exchange (with respect to their actual volume), leading to a global work function that tends toward that of bulk gold (when the FLG will be completely covered by a continuous gold layer). In addition to size-induced effects, the surface coverage of the g-MWCNTs by Au-NPs increases as NLp is increased to 100. This means that an increasing part of the UPS signal is due to Au-NPs with respect to the carbon, leading to a higher measured work function. The effect of the addition of Au-NPs to the g-MWCNTs on their FEE characteristics and work functions are summarized in figure 7. Figure 7(a) gives Eon and Eth (Eth is defined as the threshold field needed to extract 1 mA cm−2) and the work function as a function of NLp. Interestingly, both the work function and FEE characteristics are seen to follow 6
L-A Gautier et al
Nanotechnology 26 (2015) 045706
Figure 8. Typical TEM images taken on g-MWCNT samples decorated with Au-NPs at (a) NLp = 30 and (b) NLp = 100 after being subjected to several FEE measurements. (c) Current density variation over time for g-MWCNT emitters decorated with Au-NPs at NLp = 0, 30, and 100.
that the FEE enhancements do not arise solely from the decreased work function, but mainly from geometrical considerations. Indeed, had the improvement been strictly caused by the lower work function, β would have remained unchanged with respect to g-MWCNTs. It appears that at low NLp values, the Au-NPs act as local field enhancement sites that increase βHF and βLF. For higher NLp values, screening effects between Au-NPs that can cancel out any geometrical gain from the Au-NPs would explain the observed decrease of βHF and βLF. In sum, the FEE improvements brought by the Au-NP decoration are thought to come from the combination of both lower work function and a better geometrical factor at the optimal Au-NP size and surface density. Finally, the nanostructure of the Au-NP-decorated gMWCNT emitters were examined by means of TEM observations after being subjected to several FEE sets of measurements in order to check for the presence of any change or damage induced by the FEE. Figures 8(a) and (b) show the TEM images of the g-MWCNTs samples decorated with NLp = 30 and 100, respectively, after FEE measurements. The Au-NPs remain distributed on the deployed graphene foils, as in the as-prepared samples prior to the FEE measurements.
Given the local nature of the TEM observations, it is hard to assert with precision on any possible modification of the density of Au-NPs onto the g-MWCNTs. However, these images resemble those shown in figure 3 (before FEE measurements), and no significant changes can be noticed in the nanostructure of the Au-NP-decorated g-MWCNTs after FEE measurements. This suggests that the Au-NP-decorated gMWCNT emitters developed here are structurally stable. This nanostructural stability is consistent with the FEE measurements (J-E), which were found to be fairly reproducible when repeating numerous measurement cycles on devices with the same NLp. The time stability of the FEE of our devices was also assessed over more than 2 h of continuous emission at a fixed electric field. Figure 8(c) shows that regardless of the NLp value, the general trend is that the emitted current density undergoes an initial drop over the first 5 to 20 min, and then stays more or less stable around an average value. The g-MWCNTs decorated with NLp = 30 are confirmed to give the highest current density value, even if it drops by ∼50% over the first ∼20 min and then fluctuates, with an overall tendency to decrease slightly after 2 h of current emission. The frequent rises of the emitted current, particularly in the 7
L-A Gautier et al
Nanotechnology 26 (2015) 045706
case of NLp = 30, are believed to be due to some protruding gMWCNTs that randomly align in the electric field and produce more current before they get burnt or eventually broken.
[8]
4. Conclusion
[9]
The field emission properties of vertically aligned gMWCNTs synthesised by PECVD have been improved by means of Au-NP decoration by PLD. The PLD process is shown to be very effective in decorating g-MWCNTs with Au-NPs with controllable size and surface coverage. The PLD-deposited Au-NPs are shown to have very small diameters, on the order of 2–4 nm, and to uniformly cover the FLG deployed at the top of the g-MWCNTs. We have demonstrated that this decoration has a positive effect on the turn-on and threshold FEE values, as they were decreased by factors of 2.8 and 2.5, respectively. The decrease of the field needed to extract charges from these nanohybrid structures at the optimal NLp value of 30 is attributed to a combination of a decrease in their global work function (as derived from UPS measurements) and geometry enhancement due to the presence of Au-NPs at the tips of the emitters.
[10]
[11]
[12]
[13] [14]
[15]
Acknowledgments The authors would like to acknowledge financial support from the Natural Science and Engineering Research Council (NSERC) of Canada, Le Fonds de Recherche du QuébecNature et Technologies (FRQNT) through its strategic Network ‘Plasma-Québec’, and Nano-Québec (the Québec Organization for the promotion of nanoscience and nanotechnologies).
[16] [17] [18]
References
[19]
[1] Choi W B et al 1999 Fully sealed, high-brightness carbonnanotube field-emission display Appl. Phys. Lett. 75 3129 [2] Bonard J-M, Stöckli T, Noury O and Châtelain A 2001 Field emission from cylindrical carbon nanotube cathodes: possibilities for luminescent tubes Appl. Phys. Lett. 78 2775 [3] Cho W-S, Lee H-J, Lee Y-D, Park J-H, Kim J-K, Lee Y-H and Ju B-K 2007 Carbon nanotube-based triode field emission lamps using metal meshes with spacers IEEE Electron Device Lett. 28 386–8 [4] Yue G Z, Qiu Q, Gao B, Cheng Y, Zhang J, Shimoda H, Chang S, Lu J P and Zhou O 2002 Generation of continuous and pulsed diagnostic imaging x-ray radiation using a carbon-nanotube-based field-emission cathode Appl. Phys. Lett. 81 355 [5] Jeong J-W, Kang J-T, Choi S, Kim J-W, Ahn S and Song Y-H 2013 A digital miniature x-ray tube with a high-density triode carbon nanotube field emitter Appl. Phys. Lett. 102 023504 [6] Popov V 2004 Carbon nanotubes: properties and application Mater. Sci. Eng. Rep. 43 61–102 [7] Gautier L-A, Le Borgne V, Al Moussalami S and El Khakani M A 2014 Enhanced field electron emission
[20]
[21] [22]
[23]
[24]
8
properties of hierarchically structured MWCNT-based cold cathodes Nanoscale Res. Lett. 9 55 Sato H, Haruki K, Watanabe M, Hata K and Saito Y 2012 Effect of geometry of a vertically aligned carbon nanotube pillar array on its field-emission properties Surf. Interface Anal. 44 776–9 Hazra K S, Gigras T and Misra D S 2011 Tailoring the electrostatic screening effect during field emission from hollow multiwalled carbon nanotube pillars Appl. Phys. Lett. 98 123116 Yu C, Ryu Y, Yin L and Yang H 2011 Modulating eletronic transport properties of carbon nanotubes to improve the thermoelectric power factor via nanoparticle decoration ACS Nano 5 1297–303 Chen Y M, Chen C A, Huang Y S, Lee K Y and Tiong K K 2010 Characterization and enhanced field emission properties of IrO2-coated carbon nanotube bundle arrays Nanotechnology 21 035702 Shahi M, Gautam S, Shah P V, Rawat J S, Chaudhury P K and Tandon R P 2013 Decoration of cesium iodide nano particles on patterned carbon nanotube emitter arrays to improve their field emission J. Nanoparticle Res. 15 1497 Zuo Y, Ren Y, Wang Z, Han X and Xi L 2013 Enhanced field emission and hysteresis characteristics of aligned carbon nanotubes with Ti decoration Org. Electron. 14 2306–14 Zhang J, Yang C, Wang Y, Feng T, Yu W, Jiang J, Wang X and Liu X 2006 Improvement of the field emission of carbon nanotubes by hafnium coating and annealing Nanotechnology 17 257–60 Liu C, Kim K S, Baek J, Cho Y, Han S, Kim S-W, Min N-K, Choi Y, Kim J-U and Lee C J 2009 Improved field emission properties of double-walled carbon nanotubes decorated with Ru nanoparticles Carbon 47 1158–64 Shrestha S, Choi W C, Song W, Kwon Y T, Shrestha S P and Park C-Y 2010 Preparation and field emission properties of Er-decorated multiwalled carbon nanotubes Carbon 48 54–9 Yu K, Lu G, Bo Z, Mao S and Chen J 2011 Carbon nanotube with chemically bonded graphene leaves for electronic and optoelectronic applications J. Phys. Chem. Lett. 2 1556–62 Tung V C, Chen L-M, Allen M J, Wassei J K, Nelson K, Kaner R B and Yang Y 2009 Low-temperature solution processing of graphene-carbon nanotube hybrid materials for high-performance transparent conductors Nano Lett. 9 1949–55 Yu D and Dai L 2010 Self-assembled graphene/carbon nanotube hybrid films for supercapacitors J. Phys. Chem. Lett. 1 467–70 Fan Z, Yan J, Zhi L, Zhang Q, Wei T, Feng J, Zhang M, Qian W and Wei F 2010 A three-dimensional carbon nanotube/graphene sandwich and its application as electrode in supercapacitors Adv. Mater. 22 3723–8 Yi J H and El Khakani M A 2005 Highly oriented planar arrays of SWCNTs grown onto HOPG substrates by means of an ‘all-laser’ process Chem. Phys. Lett. 413 182–7 El Khakani M A and Yi J H 2004 The nanostructure and electrical properties of SWNT bundle networks grown by an all-laser growth process for nanoelectronic device applications Nanotechnology 15 S534–9 Zeng L, Lei D, Wang W, Liang J, Wang Z, Yao N and Zhang B 2008 Preparation of carbon nanosheets deposited on carbon nanotubes by microwave plasma-enhanced chemical vapor deposition method Appl. Surf. Sci. 254 1700–4 Ubnoske S M, Raut A S, Brown B, Parker C B, Stoner B R and Glass J T 2014 Perspectives on the growth of high edge density carbon nanostructures: transitions from vertically oriented graphene nanosheets to graphenated carbon nanotubes J. Phys. Chem. C. Nanomater. Interfaces 118 16126–32
L-A Gautier et al
Nanotechnology 26 (2015) 045706
[25] Ferrari A C et al 2006 Raman spectrum of graphene and graphene layers Phys. Rev. Lett. 97 187401 [26] Ka I, Ma D and El Khakani M A 2011 Tailoring the photoluminescence of PbS-nanoparticles layers deposited by means of the pulsed laser ablation technique J. Nanoparticle Res. 13 2269–74 [27] Stoeva S I, Prasad B L V, Uma S, Stoimenov P K, Zaikovski V, Sorensen C M and Klabunde K J 2003 Facecentered cubic and hexagonal closed-packed nanocrystal superlattices of gold nanoparticles prepared by different methods J. Phys. Chem. B 107 7441–8 [28] Dolbec R, Irissou E, Chaker M, Guay D, Rosei F and El Khakani M A 2004 Growth dynamics of pulsed laser deposited Pt nanoparticles on highly oriented pyrolitic graphite substrates Phys. Rev. B 70 201406 [29] Hamoudi Z, El Khakani M A and Mohamedi M 2012 Influence of pressure on the structural and electrocatalytic properties of pt nanoparticles grown by pulsed laser ablation onto carbon paper substrate Int. J. Electrochem. Sci. 7 1666–76 [30] Aïssa B, Therriault D and El Khakani M A 2011 On-substrate growth of single-walled carbon nanotube networks by an ‘all-laser’ processing route Carbon 49 2795–808 [31] Fowler R H and Nordheim L 1928 Electron emission in intense electric fields Proc. R. Soc. A Math. Phys. Eng. Sci. 119 173–81 [32] Mayer A 2011 Exact solutions for the field electron emission achieved from a flat metal using the standard Fowler– Nordheim equation with a correction factor that accounts for the electric field, the work function, and the Fermi energy of the emitter J. Vac. Sci. Technol. B 29 021803 [33] Giovannetti G, Khomyakov P, Brocks G, Karpan V, van den Brink J and Kelly P 2008 Doping graphene with metal contacts Phys. Rev. Lett. 101 026803 [34] Shiraishi M and Ata M 2001 Work function of carbon nanotubes Carbon 39 1913–7 [35] Ago H, Kugler T, Cacialli F, Salaneck W R, Shaffer M S P, Windle A H and Friend R H 1999 Work functions and surface functional groups of multiwall carbon nanotubes J. Phys. Chem. B 103 8116–21 [36] Su W, Leung T and Chan C 2007 Work function of singlewalled and multiwalled carbon nanotubes: first-principles study Phys. Rev. B 76 235413 [37] Lide D R 2005 CRC Handbook of Chemistry and Physics (Boca Raton, FL: CRC Press) [38] Bittencourt C, Felten A, Douhard B, Ghijsen J, Johnson R L, Drube W and Pireaux J-J 2006 Photoemission studies of
[39]
[40]
[41] [42]
[43]
[44] [45]
[46]
[47]
[48]
[49] [50]
9
gold clusters thermally evaporated on multiwall carbon nanotubes Chem. Phys. 328 385–91 Khoa N T, Kim S W, Yoo D-H, Kim E J and Hahn S H 2014 Size-dependent work function and catalytic performance of gold nanoparticles decorated graphene oxide sheets Appl. Catal. A Gen. 469 159–64 Xu N S, Chen Y, Deng S Z, Chen J, Ma X C and Wang E G 2001 Vacuum gap dependence of field electron emission properties of large area multi-walled carbon nanotube films J. Phys. D: Appl. Phys. 34 1597–601 Barbour J, Dolan W, Trolan J, Martin E and Dyke W 1953 Space-charge effects in field emission Phys. Rev. 92 45–51 Chen Y, Deng S Z, Xu N S, Chen J, Ma X C and Wang E G 2002 Physical origin of non-linearity in Fowler–Nordheim plots of aligned large area multi-walled nitrogen-containing carbon nanotubes Mater. Sci. Eng. A 327 16–9 Chen L F, Song H, Cao L Z, Jiang H, Li D B, Guo W G, Liu X, Zhao H F and Li Z M 2009 Effect of interface barrier between carbon nanotube film and substrate on field emission J. Appl. Phys. 106 033703 Bonard J-M, Klinke C, Dean K and Coll B 2003 Degradation and failure of carbon nanotube field emitters Phys. Rev. B 67 115406 Kim D-H, Yang H-S, Kang H-D and Lee H-R 2003 Novel emission degradation behavior of patterned carbon nanotubes by field emission Chem. Phys. Lett. 368 439–44 Zhang J, Yang C, Yang W, Feng T, Wang X and Liu X 2006 Appearance of a knee on the Fowler–Nordheim plot of carbon nanotubes on a substrate Solid State Commun. 138 13–6 Zhang J, Wang X, Yang W, Yu W, Feng T, Li Q, Liu X and Yang C 2006 Interaction between carbon nanotubes and substrate and its implication on field emission mechanism Carbon 44 418–22 Neupane S, Lastres M, Chiarella M, Li W, Su Q and Du G 2012 Synthesis and field emission properties of vertically aligned carbon nanotube arrays on copper Carbon 50 2641–50 Altman I S, Pikhitsa P V and Choi M 2004 Electron field emission from nanocarbons: a two-process model Appl. Phys. Lett. 84 1126 Rakhi R B, Reddy A L M, Shaijumon M M, Sethupathi K and Ramaprabhu S 2007 Electron field emitters based on multiwalled carbon nanotubes decorated with nanoscale metal clusters J. Nanoparticle Res. 10 179–89
