The MWCNT yarn can be made out of super-aligned MWCNTs using a spinning technique (Jiang et al., 2002; Zhang et al., 2006). The yarns were obtained from Tsinghua University, Beijing, China. The pristine MWCNT yarn was pulled through an ethylene glycol solvent (purchased from Sigma-Aldrich) to make it stiffer and packed tighter. Then a thermal treatment at 200–250°C for 30 min in air in a forced convection oven (JEIO TECH OF-22GW) was conducted to completely evaporate the solvent. The morphology of the yarn was characterized by scanning electron microscopy (SEM, Hitachi, S-4700).

Since the MWCNT yarn has a macroscopic dimension, it can be easily manipulated by hand. A schematic diagram of the lateral MWCNT yarn emitter is shown in Figure 1. A nearly 20 mm-long yarn was attached on a stainless steel support using a conducting epoxy (silver paste, purchased from MG Chemicals). A simple manipulation system is enough to attach the yarn on the support because the segment of the yarn can be handled easily using tweezers. Thus, the fabrication of CNT emitters is very simple and easy. The sample was put in the oven to subject a thermal annealing under atmosphere pressure at 80°C for 15 min, which can enhance the mechanical adhesion between the yarn and the support. Finally, the sample was put in a vacuum chamber at a pressure less than 2 × 10⁻⁷ torr to investigate field emission properties. Field emission measurements were carried out using a diode configuration. The anode was a homemade phosphor-coated indium tin oxide (ITO) glass with a width of ~10 mm. The phosphor powder and binder were purchased from Alibaba and then screen-printed on ITO-coated glass (purchased from Sigma-Aldrich). The gap between the anode and the yarn was adjusted to 400 μm by ceramic spacers. The emission current was monitored with a Keithley 6485, and the direct current power was supplied by a constant power voltage and current controller (HCN140-3500).

Figure 2A shows a SEM image of the MWCNT yarn with low magnification. The diameter of the MWCNT yarn is about 50 μm. It is obvious that the surface of the MWCNT yarn is very smooth and even. We observed the yarn throughout the entire length and could not find noticeable protruding CNT tips on the surface at this magnification. This indicates that MWCNTs are tightly bound together within the yarn. The high magnification SEM images are shown in Figures 2B,C. The yarn consists of arrays of aligned MWCNTs, but some of MWCNTs show the curly shape. The two ends (tips) of the curly CNTs are just buried under the neighboring CNTs. Figure 2D shows the highly magnified SEM image of the yarn surface. There are many outmost MWCNTs protruding out from the yarn surface, indicated by the white arrows. An exaggerated example was also presented in the inset of Figure 2D. It is worth noting that most of the protruding CNTs are not so far away from the surface. This typical microstructure can allow field emission possible from the MWCNT yarn.

We investigated field emission properties of the lateral MWCNT yarn emitter. A strong uniform line emission pattern was observed from the lateral emitter, as shown in Figure 3A. The two joints were kept uncovered by the ITO glass to avoid field emission from the silver paste (Ag paste) or other potential undesirable parts. It clearly reveals that the field emission is from the lateral emitter. This uniform line emission pattern can be used for lamps, figure display boards, cylindrical cathodes, etc.

Field emission characteristics of the lateral emitter are shown in Figure 3B. The lateral emitter demonstrated excellent field emission performance with a turn-on electric field of 1.13 V/μm at an emission current of 1 μA and an emission current of 0.2 mA at an applied voltage of 700 V. It was assumed that the field emission occurred from the top half of the yarn surface. Based on the diameter of the yarn (50 μm) and the length (10 mm), the corresponding emission current density was calculated as ~25.5 mA/cm². This current level is enough for field emission display applications. The lateral emitter exhibited a better or comparable field emission performance in comparison with other reported lateral emitters or vertical emitters (Gao et al., 2001; Jo et al., 2003; Seelaboyina et al., 2006; Jang et al., 2010; Chen et al., 2013). For example, the electron emission threshold for a current density of 1 mA/cm² was 4 and 6.5 V/μm for the single-walled CNT film prepared using methanol and DMF, respectively, with the gap distance of 165 μm and emission area of 6 mm² (Gao et al., 2001). It means that the emitters reached 60 μA at the voltage of 660 or 1072 V. In addition, the values of electric field required to obtain a current density of 1 mA/cm² were in the range of 5.40–12.8 V/μm for samples with different CNT lengths and number densities, with a gap distance of 300 μm, an emission size of 5-mm diameter, and the maximum emission current density is about 2.2 mA/cm² (Jo et al., 2003). It implies that the emitters reached about 0.2 mA at the voltage range of 1620–3840 V. Furthermore, the thin MWCNTs reached the maximum emission current of less than 0.2 mA at an applied voltage of 1500 V with a gap distance of 380 μm (Seelaboyina et al., 2006). The low turn-on electric field and relatively high current (and current density) at low applied voltage suggested a better field emission performance of our lateral emitter compared to these reports. The inset in Figure 3B shows the corresponding Fowler–Nordheim (F–N) plot. The straight line indicates the quantum mechanical tunneling characteristic of field electron emission. The field enhancement factor can be calculated using the F–N equation: (1)I=(Aαβ2V2/d2φ)exp(−Bφ32d/βV) where I is the emission current, A = 1.56 × 10⁻⁶ AV⁻² eV, B = 6.83 × 10⁹ eV^−3/2 Vm⁻¹, α is the effective emission area, β is the field enhancement factor, φ is the work function, V is the applied voltage, and d is the gap distance between the anode and the emitter. Using Eq. 1, the slope of the F–N plot [log(I/V²) versus 1/V] is given by: (2)Slop=−Bφ32d/β

Therefore, the field enhancement factor can be calculated from the F–N plot assuming the work function of the MWCNTs to be 5.0 eV (de Jonge et al., 2004). The field enhancement factor of our lateral emitter is 8300. This large field enhancement factor indicates that the lateral emitter can start to emit electrons at a considerably low applied voltage.

Field emission stability is one of the important parameters for practical applications of CNT-based field emission cold cathodes. We evaluated emission current stability of the lateral emitter at an initial emission current of about 0.10 mA under a direct current bias. Then, the applied voltage was kept constant for over 20 h. As a matter of fact, this initial current is strong enough to perform acceleration test for field emission stability (the corresponding current density is about 12.7 mA/cm²) compared with the previous reports (Moon et al., 2006; Jung et al., 2007a). As shown in Figure 4, the lateral emitter showed a very stable emission current for over 20 h without any significant current decay, indicating a better field emission stability (Moon et al., 2006; Jung et al., 2007b). The excellent field emission stability is mainly attributed to the tight bonding of neighboring MWCNTs within the yarn. The yarn was densified and shrank into a tight structure with several tens of micrometers in diameter after ethylene glycol treatment. The tightly bound MWCNT yarn would not easily degrade as applying high electric field on it. As a result, the lateral emitter demonstrated a long-term emission stability.

The surface morphology of the lateral emitter after field emission measurement was observed as well. There was no noticeable damage to the surface of the MWCNT yarn after field emission stability test. In addition, most importantly, no protruding CNT tips were found. When we consider the morphology of the yarn, some outmost MWCNTs protrude out from the surface, as shown in Figure 2D. They look just like CNT loops in other reports (Konishi et al., 2005; Futaba et al., 2012). Electrons can be easily extracted out from these protruding MWCNTs due to the concentrated electric field compared with other MWCNTs inside the yarn. Literature survey shows that field emission can either from the tip of CNTs or even from the body of CNTs (Zhou and Kawazoe, 2001; Zhou et al., 2001; Konishi et al., 2005). Here, it is suggested that the field emission occurs from the outmost MWCNTs which are protruding from the yarn surface. When the electrical bias is applied, the electric field is concentrated on the surface of the yarn, especially on the outmost MWCNTs which are protruding from the surface. This phenomenon is similar to the concentrated electric field at the protruding tips of the vertically aligned CNTs. The lateral yarn emitter showed high emission current, long-term emission stability, and uniform line emission pattern, suggesting its potential field emission applications such as line-shaped emitters, figure display boards, and cylindrical lamps.

We also investigated the field emission properties as a function of the gap between the anode and the lateral emitter. The same lateral MWCNT yarn emitter was used to study the gap dependence. The gap was adjusted as 100, 200, 400, 600, and 800 μm, separately, by ceramic spacers. The emission current–applied voltage (I − V) curves of the lateral emitter as a function of the gap are shown in Figure 5. It can be seen that the I − V curves shift to the right (higher voltage) as increasing the gap. The turn-on electric field was 2.30 V/μm for the gap of 100 μm. As the gap was increased to 200 μm, the turn-on electric field dramatically decreased to 1.46 V/μm. Continuous increasing of the gap to 400, 600, and 800 μm resulted in the turn-on field of 1.13, 1.11, and 0.96 V/μm, respectively. These values indicate lower turn-on electric fields than those of the vertical emitters. The lateral emitter shows a same gap dependence of field emission properties as the conventional vertical emitters (Choi et al., 2000; Moon et al., 2006; Chen et al., 2009).

Some potential applications of our lateral emitter are demonstrated in Figure 6. First, a field emission cylindrical lamp based on the lateral emitter is presented in Figure 6A. A phosphor-coated flexible transparent film can be easily made and used as the anode, and therefore a cylindrical lamp can be assembled without using hardly fabricated uniform flexible emission cathode. Second, a figure display board is also possible using our line-shaped emitter (Figure 6B). A patterned phosphor-coated ITO glass can be used as the anode. The spacers are used not only to separate the anode and the emitter but also to isolate the influence from other MWCNT yarns nearby. It is worth noting that the specifications need to be evaluated once the real device is made, such as power assumption, output luminosity as in reference (Croci et al., 2004).