We discuss the influence of few-walled carbon nanotubes (FWCNTs) treated with nitric acid and/or sulfuric acid on field emission characteristics. FWCNTs/tetraethyl orthosilicate TEOS) thin film field emitters were fabricated by a spray method using FWCNTs/TEOS sol one-component solution onto indium tin oxide (ITO) glass. After thermal curing, they were found tightly adhered to the ITO glass, and after an activation process by a taping method, numerous FWCNTs were aligned preferentially in the vertical direction. Pristine FWCNT/TEOS-based field emitters revealed higher current density, lower turn-on field, and a higher field enhancement factor than the oxidized FWCNTs-based field emitters. However, the un-stable dispersion of pristine FWCNT in TEOS/N,N-dimethylformamide solution was not applicable to the field emitter fabrication using a spray method. Although the field emitter of nitric acid-treated FWCNT showed slightly lower field emission characteristics, this could be improved by the introduction of metal nanoparticles or resistive layer coating. Thus, we can conclude that our spray method using nitric acid-treated FWCNT could be useful for fabricating a field emitter and offers several advantages compared to previously reported techniques such as chemical vapor deposition and screen printing.
Carbon nanotubes (CNTs) have many exceptional properties that make them attractive for a variety of applications. Previous works have shown that CNTs have novel physical and chemical properties making them promising materials in field emission devices. In 1995, four years after the discovery of CNTs by Iijima
, three groups reported field emission from CNTs at low turn-on fields and high current densities
. From 1998, the concept of using nanotubes in field emission devices spurred on efforts worldwide: a first display
as well as a lighting element
was presented. This area has been experiencing a rapid growth the past few years, and it is encouraging to note that two commercial products, namely high-brightness luminescent elements and an X-ray tube, have hit the market. Applications such as flat displays
, X-ray tube sources
, electron sources for high-resolution electron beam instruments such as electron beam lithography machines and electron microscopes
are under active consideration.
There are several properties of CNTs that make them extraordinary materials for field emission. Firstly, the graphene walls in them are parallel to the filament axis resulting in the nanotubes (whether metallic single-walled or multi-walled) exhibiting high electrical conductivity at room temperature. Secondly, CNTs are high in aspect ratio and whisker-like in shape. Utsumi
evaluated commonly used field emission tip shapes and concluded that the best field emission tip should be whisker-like, followed by the sharpened pyramid, hemi-spheroidal,and pyramidal shapes. It has been reported that even curly 'spaghetti-like' nanotubes stand up vertically like whiskers during emission under the application of an electric field
. Thirdly, CNTs can be very stable emitters, even at high tem-peratures. Previous researchers demonstrated that a multi-walled carbon nanotube (MWCNT) emitter could be heated up to 2000 K by its field-emitted current and remain stable. They claimed that this was the first reported observation of field emission in-duced stable heating. This characteristic is distinctively differ-ent from metal emitters. In metals, the resistance,
, increases with temperature, which means that more heat,
, is produced as higher currents,
, are drawn
(Q = I2R)
. The combination of high temperature and electric field causes the well-known mechanism of field-sharpening of tips by surface diffusion, which in turn increases the local field, current. and temperature. This posi-tive feedback mechanism causes an unstable thermal runaway, which inevitably leads to emitter destruction for metal-based emitters. In contrast, the resistance of a CNT decreases with temperature, which limits I
R heat generation. Consequently, its temperature varies sub-linearly with the current
. Addition-ally, surface diffusion is less likely in the strong C-C covalent bonds of the CNTs. These unique characteristics of CNTs make them a remarkable field emitter.
In general, single-walled carbon nanotubes (SWCNTs) have been studied to obtain low threshold electric field and high emission current density
, but abrupt degradation of emis-sion current becomes a serious obstacle for field emission ap-plications. In contrast to SWCNTs, MWCNTs have shown bet-ter emission stability; however, a small field enhancement factor has been attributed to low emission current
. Recently, dou-ble-walled carbon nanotubes have been studied for this purpose due to high current density with long-term emission stability, but the poor production rate and following purification process of catalysts have been a serious bottleneck
. Few-walled car-bon nanotubes (FWCNTs), with small number of walls (2 to 6) and small diameters (below 5 nm), have been synthesized with high catalyst efficiency such that no further purification process was required
. Therefore, these FWCNTs could be the most promising candidates for the field emitter due to their intermedi-ate structural properties between SWCNTs and MWCNTs
and high production rate.
To fabricate CNT field emitters, there are several available methods including direct growth of CNT arrays using chemical vapor deposition (CVD), paste screen-printing of CNT powders, and electrophoretic deposition. However, each of them has sev-eral problems. The catalyst CVD method has to be performed at a higher temperature for CNT growth. Its fabricating process is limited by the scalability of the substrate size and it is diffi-cult to control CNT morphologies such as density and spacing. The paste screen-printing method, by contrast, is known as a more effective technology for large-scale production and has the advantage of a low cost process. However, substantial degrada-tion of emission tips caused by oxygen gases that are out-gassed from the organic vehicles in the paste cannot be avoided and as a result, poor uniformity of emission, poor emission spot density,and short lifetime will limit its application in commercial pro-duction. Furthermore, because the vapor from residual organic solvent will severely deteriorate the vacuum of field emission devices after the sealing process, a post-heat-treatment to re-move the residual organic solvent from the paste is necessary. The electrophoretic method has the problems of weak adhesion between CNT emitters and substrates, and non-uniformity in the thickness of the deposited layer, which will cause irregularity in the distribution of emission sites. The spray method is an easy and convenient method to deposit CNTs on the cathode sub-strates with a large area. However, the poor adhesion of CNTs to the substrate is also a serious drawback for long-term electron emission. Moreover, the spray method has been largely ham-pered by poor dispersion of CNTs, which was caused by their in-herent attractive van der Waals interactions between CNTs. One of the best methods to overcome this problem is to introduce carboxyl groups (-COOH) and hydroxyl groups (-OH) onto the surface of the CNTs through oxidation by nitric acid/sulfuric acid treatment. While this technique is a powerful method for improving the dispersion of CNTs, the acid treatment induces structural defects that change the property and morphology of CNTs.
In this study, we describe the fabrication of the CNT cath-odes by a spray method using a FWCNTs/binder one-compo-nent solution to improve the adhesion between the FWCNTs and the substrate for long-term emission stability. Moreover, various oxidation processes were applied to modify the sur-faces of pristine FW--CNTs. The hydration properties of FWCNTs and their effects on electron field emission property were studied.
- 2.1. Materials
The FWCNTs fabricated by CVD method were purchased from Hanwha Nanotech Co. We used a CNT product (CMP series) consisting of highly purified agglomerates of FW-CNTs with an outer mean diameter of 3 to 5 nm and a typical length ranging from hundreds of nanometers to a micrometer. Tetraethyl orthosilicate (TEOS) was purchased from Aldrich and used as received for improving adhesion between the FWCNTs and the substrate. Sulfuric acid (H
) and nitric acid (HNO
) (Daejung Chemicals Co., Korea) were used for functionalization of CNTs on their side wall by chemical oxidation process. We used
-dimethylformamide (DMF) as a dispersion solvent of CNTs due to the high value of dis-persion component (δ
) of the Hildebrand solubility param-eter (δ
). Ham et al.
noted that solvents with high values of δ
are the best for making homogeneous and agglomerate-free dispersions of CNTs.
- 2.2. Surface functionalization of FWCNTs
For the surface functionalization of FWCNTs, 50 mg of the agglomerated FWCNTs were mixed with 100 mL of one of the acid solutions listed in
by stirring for 5 min at room temperature and sonicating in a conventional ultrasonic bath (100 W) for 10 min; sonication longer than 10 min with high
Acid treatment conditions for each sampleFWCNT: few-walled carbon nanotube.
Acid treatment conditions for each sample FWCNT: few-walled carbon nanotube.
(a) Schematic of the acid treatment process for functionalization on FWCNTs surface and (b) illustration of the vacuum tube diode field emissiontest setup. FWCNT: few-walled carbon nanotube, ITO: indium tin oxide.
power could induce surface defects, not a functionalization, of FWCNTs. The sonicated FWCNTs/acid mixture was refluxed on a hot plate at 150℃ and then the slurry was filtered by vacuum filtration using an alumina filter and thoroughly washed with distilled water until the solution reached a pH value of 7. Fi-nally, the functionalized FWCNTs were dried by a freeze dryer for 48 h to remove moisture.
a describes the acid treatment process schematically.
- 2.3. Preparation of FWCNTs/TEOS sol onecomponent solution
TEOS sol solution was prepared by mixing 5 g of tetrae-thoxy silane, 1.73 g of water, 3.33 g of ethanol, and 8 mg of 30.9 M hydrochloric acid, and performing a continuous sol-gel reaction at 60
C under 250 rpm (
). After that,pristine and functionalized FWCNTs were dispersed in the DMF for 2 h in a conventional ultrasonic bath (100 W) at a definite concentration. Next, we centrifuged the solution 5000 rpm for 30 min to remove graphitic particles and large CNT bundles. Lastly, dispersed FWCNTs solution was mixed with TEOS sol and sonicated in a conventional ultrasonic bath (100 W) for 30 min (
Formulation of the TEOS solTEOS: tetraethyl orthosilicate, HCl: hydrochloric acid.
Formulation of the TEOS sol TEOS: tetraethyl orthosilicate, HCl: hydrochloric acid.
Formulation of the FWCNTs/TEOS one component solu-tion for spray coatingFWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate, DMF:N,N-dimethylformamide.
Formulation of the FWCNTs/TEOS one component solu-tion for spray coating FWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate, DMF: N,N-dimethylformamide.
- 2.4. Preparation of FWCNTs/TEOS thin film field emitter
The field emitters using FWCNTs/TEOS one component solution were prepared by the following method. Firstly, the FWCNTs/TEOS one component solution was sprayed onto in-dium tin oxide (ITO) coated glass by an automatic spray coater (NVD-200, Fujimori, Japan) with a nozzle of 1.2 mm diameter at 135℃. The prepared FWCNTs/TEOS thin films (2 × 2 cm
)were then cured in a vacuum oven at 150℃ for 90 min. After thermal curing, they were found tightly adhered to the ITO glass. We performed the activation process by a taping method that creates many nanotubes aligned in the vertical direction.
- 2.5. Characterizations
To get information about functional groups on the nanotube surface, X-ray photoelectron spectroscopy (XPS) measurements were performed on a MultiLab2000 (Thermo VG Scientific, UK) with monochromatized Al Kα X-ray radiation as the X-ray source for excitation under ultrahigh vacuum system. The XPS core level spectra were analyzed with a fitting routine, which can decom-pose each spectrum into individual mixed Gaussian Lorentzian peaks. Raman spectroscopy (FRS-100S; Bruker, USA) with an Ar laser excitation of 633 nm wavelength was used to assess the quality of the graphitic structure of the FWCNTs. We measured the Raman spectrum from more than ten different points of each sample for correct information. The corresponding images of the samples were obtained by field-emission scanning electron mi-croscopy (FE-SEM, 10 kV, S-4800, Hitachi, Japan). The well-dispersed FWCNTs solution after centrifugation was analyzed by UV visible (UV-VIS) spectroscopy (cary 5000; Varian, USA) to determine the effect of acid treatment on the state of the disper-sion and dispersability of the nanotubes in DMF.
Deconvolution of the X-ray photoelectron spectroscopy C1s peak for the (a) pristine FWCNT (b) N_FWCNT, (c) SN-a FWCNT, and (d) SN-b FWCNT. FWCNT: few-walled carbon nanotube.
The field emission properties were measured in a diode type with a direct current bias inside a vacuum chamber un-der less than 5 × 10
Torr of pressure, as shown in
b. ITO glass was used as an anode electrode separated by a 200μm polyethylene terephthalate (PET) spacer. To obtain stable field emission properties, an electrical aging process was per-formed by the following method: first, 100 repetitions of I-V sweeping up to 0.5 mA/cm
of emission cur-rent density, then constant voltage annealing at 0.35 mA/cm
of emission current density for 2 h.
During this electrical aging process, some protruded FW-CNTs were removed so that the length uniformity of FWCNT emitters was enhanced, which resulted in stable electron emis-sion. White phosphor-coated ITO glass as an anode was used to determine the emission pattern of FWCNT emitters.
3. Results and Discussion
- 3.1. Characterization of pristine and acid-treated FWCNTs
- 3.1.1. XPS analysis of FWCNTs
With reference to functionality grafting, XPS is a surface analytical technique that can provide useful information on the nature of functional groups and also on the presence of structural defects on the nanotube surface. The intensity
of photoelec-trons of kinetic energies
is given by
Relative percentage of functional components obtainedfrom curve fitting the C1s peak for the pristine FWCNT and oxidized FWCNTsFWCNT: few-walled carbon nanotube.
Relative percentage of functional components obtainedfrom curve fitting the C1s peak for the pristine FWCNT and oxidized FWCNTs FWCNT: few-walled carbon nanotube.
Summary of the elemental composition for the pristine FWCNT and oxidized FWCNTsFWCNT: few-walled carbon nanotube
Summary of the elemental composition for the pristine FWCNT and oxidized FWCNTs FWCNT: few-walled carbon nanotube
is the atomic density, σ is the ionization cross-section of the observed photoelectron line according to a defined core level and the photon energy
) is the inelastic mean free path, and
is the transmission function of the spectrometer used. Thus,
I ∝ N
, and we can quantitatively analyze the atomic concentration of carbon atoms functionalized in the oxidizing process. Atomic composition of functionalities was obtained from curve fitting results based on ratios of peak areas. These values could only give us basic information about the functional groups attached to nanotubes.
The C1s peak of pristine and acid-treated FWCNTs at 285 eV can be deconvoluted into four fitting curves with binding energy at ~285 eV, ~285.63 eV, ~286.59 eV, and ~289.16 eV, which are assigned to C-C, C-O, C=O and O-C=O, respectively (
. The relative amount of functional groups could be calculated from the area of each deconvoluted peak. The rela-tive amount of functional groups and the atomic element in the samples is listed in
As seen in
, we found that the relative amounts of C-O and O-C=O on the surface of acid-treated FWCNTs are higher than that of pristine FWCNT. Furthermore, sulfuric/nitric acid-treated FWCNTs (indicated by SN-a FWCNT and SN-b FW-CNT) have much higher relative amounts of C-O and O-C=O than nitric acid-treated FWCNT (indicated by N_FWCNT). Moreover, the atomic percentage of the oxygen element (
) on the surface of acid-treated FWCNTs noticeably increases compared to pristine FWCNT. These results indicate that acid treatment with sulfuric/nitric mixture acid could effectively in-
(a) Raman spectra (excitation at λ = 633 nm) for pristine few-walledcarbon nanotube (FWCNT) and oxidized FWCNTs at various acidtreatment conditions. (b) Ratio of the D-to-G band intensity and the Gband shift.
duce more polar oxygen-containing groups on the surface of nanotubes, which will improve the stability of CNT dispersion in DMF.
- 3.1.2. Carbon ordering
Raman spectroscopy is a powerful measurement method for evaluating the structural properties of CNT because it is a non-destructive and contactless measurement method. The CNT was formed through the crimping of flat graphite and has an unsatu-rated double-bond structure on the graphitic structure; thus a characteristic band can be measured using the Raman spectra. The peak near 1580 cm
represents the graphite E
symmetry of the interlayer mode, which reflects the structural intensity of the
hybridized carbon atoms of the CNTs, and the absorption peak is referred to as the
band (graphitic band). A perfect gra-phitic sheet is composed of hexagons of carbon atoms, and when pentagons and heptagons or defects occur in other positions, de-fect
bands (disorder band) can take place at 1350 cm
band accounts for the crystal defect degree of graphitic sheets. Therefore, the extent of the defects in CNTs can be evaluated by the ratio of the
band intensities (I
In this study, Raman spectroscopy was used to assess the qual-ity of FWCNTs depending on the acid treatments.
a shows the Raman spectrum of pristine and acid-treated FWCNTs. The maximum Raman peak near 1570 cm
band) related to the graphite E
symmetric intra-layer mode was shifted to higher wavenumbers after acid treatment (
b). This phenomenon surcould be explained by the following two reasons. First, when the CNTs are treated by nitric acid or sulfuric/nitric acid π electrons are withdrawn from the nanotube to nitrate ions (NO
), which resulted in the blue shift of Raman
. Second, it has been reported that the tangential Raman E
mode can be shifted to higher wavenumbers upon hydrostatic pressure, which cor-responds to a reduction in the bond length and an associated stiffening of the bonds
. In our experiment, compressive strain of the C-C bond or a change in the bond angles resulting from graphitic layer stacking caused by strong hydrogen bond-ing could lead to an up-shift in the tangential mode frequency. In this case, we believe the strong intermolecular interaction between hydroxyl groups could have arisen from the acid treat-ment, thus enhancing graphitic layer stacking.
In addition, as shown in
b, sulfuric/nitric acid-treated samples (SN-a FWCNT, SN-b FWCNT) have a higher I
ratio than the nitric acid-treated sample (N_FWCNT), and the I
ratio increases with increased refluxing time. These results in-dicate that the crystallinity of nanotubes could be triggered by a strong acid treatment and its experimental condition. In gen-eral, defects in nanotubes are mainly located at the tip due to the hemisphere structure of the carbon buckyball. Some defects can form on the side wall of the nanotube because of bending or kinking. These defects can be oxidized by a strong acid and then chemical functionalization can easily occur. Furthermore, these side wall defects can be an electron trapping site, which can be attributed to poor electron emission, even though the open tip of a nanotube has better emission properties than a closed one. In the Section 3.2, we will discuss the effects of nanotube crystal-linity on field emission properties.
- 3.1.3. Solubility and stability of FWCNT dispersion in DMF
It is well known that the bundled structure of nano- tubes does not actively absorb light in the UV-Vis region; individual nano-tubes show the opposite effect
. Therefore, dispersion of CNTs can be characterized using UV-Vis absorption spectros-copy. In this study, to characterize the dispersion of FWCNTs in DMF using UV-Vis spectroscopy, absorbance values were re-corded at 550nm, as reported in previous studies
, which is shown in
According to the Beer-Lambert's law (
/I = ε
), the absorbance
is expressed in terms of extinc-tion coefficient ε, the concentration of nanotubes
, and the light path length
. In general, the absorbance is measured from ab-sorption spectroscopy at a given wavelength. The path length
is fixed for a given optical cell in the experiment. Therefore, if the concentration of the CNT in solution is known, the extinction coefficient can be easily obtained using the equation.
a depicts the UV-Vis spectra of pristine and oxidized FWCNTs solution with various concentrations of nanotubes in DMF. We note that the relationship between the absorbance and the concentration of the FWCNT solutions was linear, which is in good agreement with Beer-Lambert's law
. In addition, it is noteworthy that the UV absorbance value of pristine FWCNT so-lution was higher than that of the oxidized FWCNTs solution;therefore, the extinction coefficient of the pristine FWCNT solution at a nanotube concentration of 100 mg/L is the largest (
b). We note that the absorbance of the nitric acid-treated sample (N_FWCNT) was higher than that of the sulfuric/nitric acid-treated samples (SN-a FWCNT, SN-b FWCNT) at the
(a) UV-visible absorbance at 550 nm for pristine and oxidizedfew-walled carbon nanotubes (FWCNTs) solutions depending on the acidtreatment at the same FWCNTs concentration and (b) extinction coeffi-cient at 550 nm for pristine and oxidized FWCNTs solutions depending onthe acid treatment at the same FWCNTs concentration (100 mg/L).
same nanotube concentration (
). In general, carbon atoms in the pristine CNTs are in an
state, but an
carbon con-figuration is generated during the oxidization in the defect sites of CNTs. The geometrical tension between the hybridization of oxidized and non-oxidized carbon induced local deformations. One would expect that extended acid treatment, in which a greater number of functional groups are produced, would be ac-companied by more deformation. Therefore, we believe that this phenomenon is due to the bent, shortened, and bundled structure in several areas resulting from the acid treatment.
shows the visual examination of pristine and oxidized FWCNTs in DMF at a nanotube concentration of 100 mg/L and 200 mg/L, respectively. Digital photo-images were taken after 24 h. It was found that the clear dispersion of pristine FWCNT in DMF was formed when the nanotube concentration was 100 mg/L, but in a higher concentration of 200 mg/L, the nanotubes were aggregated and precipitated. On the other hand, the oxi-dized FWCNTs showed a clear dispersion independent of the nanotube concentration in DMF. The stability of oxidized FW-CNTs in DMF may be due to the fact that the oxidation process introduces carboxyl and hydroxyl groups on the nanotube sur-
Digital photo-images of pristine and oxidized few-walled carbon nano tubes (FWCNTs) dispersed in N-dimethylformamide at nanotube concentrations of (a) 100 mg/L and (b) 200 mg/L, respectively. Pictureswere taken after 24 h.
faces. These groups ionize in DMF and the oxygen-containing groups are negatively charged. The electrostatic repulsive forces between negative surface charges of the oxygen-containing groups may lead to the stability of oxidized nanotubes so that the oxidized nanotubes can form a stable dispersion in DMF. This suggests that the oxidized FWCNTs have high dispersion stabil-ity in DMF so that the aggregation could be minimized. Conse-quently, when the CNTs are oxidized by an acid treatment, the dispersion stability in DMF is significantly enhanced due to the combination of polar-polar affinity and electrostatic repulsion.
- 3.2. Characterization of FWCNTs/TEOS thin film field emitter
- 3.2.1. Morphological property
shows the FE-SEM images of the spray coated FW-CNTs/TEOS films before the activation process. As increasing the oxidation level (or I
), the TEOS binder was shown in the top of the film. The intermolecular interaction of TEOS with nanotubes is quite poor, so that the TEOS binder should be pre-cipitated to the bottom of the film. However, CNTs functional-ized with carboxyl or hydroxyl groups could interact with the TEOS layer precipitated to the bottom of the film, which is very low in the case of highly oxidized sample, as shown in the Figs.
Field-emission scanning electron microscopy images of (a) pris-tine FWCNT/TEOS, (b) N_FWCNT/TEOS, (c) SN-a FWCNT/TEOS and (d) SN-bFWCNT/TEOS based field emitters before the surface activation, respec-tively. FWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate
Cross sectional field-emission scanning electron microscopyimages of (a) pristine FWCNT/TEOS, (b) N_FWCNT/TEOS, (c) SN-a FWCNT/TEOS and (d) SN-b FWCNT/TEOS based field emitters before the surface activation, respectively. FWCNT: few-walled carbon nanotube, TEOS: tetra-ethyl orthosilicate.
6c and d. We also realized that the adhesion strength between CNTs and the substrate is varied depending on these oxidation levels. Therefore, following the activation process using the tap-ing should be optimized to minimize this adhesion effect on the field emission of CNTs.
show the cross-sectional FE-SEM image of the spray coated FWCNTs/TEOS-based field emitter before and after the activation process. The thickness of FWCNTs/TEOS film was about 0.3 μm. It was difficult to observe the protruded CNTs at the film surface before the activation process. The bur-ied nanotubes on the surface have been aligned by the removal of a covered layer, by applying the adhesive tape technique. As explained previously, this taping process was individually optimized to all samples. The surface morphology of activated FWCNTs/TEOS-based field emitter is shown in
Field-emission scanning electron microscopy images of (a) pris-tine FWCNT/TEOS, (b) N_FWCNT/TEOS, (c) SN-a FWCNT/TEOS and (d) SN-b FWCNT/TEOS based field emitters cross-section after the surface activa-tion, respectively. FWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate.
activation process, the film thickness slightly decreased from its original value of about 0.3 μm to about 0.2 μm, and many CNTs were preferentially aligned to the vertical direction. The aligned nanotube density of pristine FWCNT (
a), nitric acid-treated FWCNT (
b) and FWCNTs treated with sulfu-ric/nitric acid for a short time (
c) is quite similar, but the FWCNTs treated by sulfuric/nitric acid for a long time (
d) have low aligned nanotube density even though we sprayed the same amount of nanotube solution and optimized the activation process. The vertical alignment of the nanotube is necessary for its field emission application because high field enhancement is strongly related to the aspect ratio of the nanotube in the ver-tical direction. The emission stability of CNT emitters is also enhanced after the activation process because some loosely bounded nanotubes could be efficiently removed.
Therefore, the activation process of the CNT film is an important step in making a high performance nanotube field emitter.Vertically aligned CNT emitters are aged electrically inside a vacuum chamber to produce a uniform length distribution,which results in the stable emission of electrons even at a high electric field. During this electrical aging step, some protruded nanotube emitters, as shown in the
a and b are oxidized through Joule heating.
- 3.2.2. Field emission properties of FWCNTs/TEOS thin film field emitter
Field emission results are usually analyzed by the
theory, which describes the emission of elec-trons from a metal as quantum-mechanical tunneling enhanced by a strong electric field
equation can be writ-ten as:
is the current density,
is the electric field strength, Φ is the work function, and
are constants (when
and Φ are expressed in mA/cm
, V/cm and eV, respectively, the con-
(a) J-E characteristics and (b) their Fowler-Norheim plot of thepristine FWCNT/TEOS and oxidized FWCNTs/TEOS based field emitters. FWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate.
assume the values of 1.54 × 10
and 6.87× 10
, respectively). According to the
model, a plot of ln
has a linear behavior with a slope that can used to evaluate the field enhancement factor (β).
The field emission characteristics of pristine and oxidized FWCNTs were measured after electrical aging in each sample as shown in
The vertically aligned pristine FWCNT/TEOS field emitter shows a low turn-on voltage of 1.90 V/μm, high field enhancement factor of 3519.43, and high emission current density of 0.5 mA/cm
at an applied field of about 3.37 V/μm. On the other hand, the nitric acid-treated FW-CNT/TEOS (N_FWCNT) field emitter shows a turn-on voltage of 2.38 V/μm, field enhancement factor of 1055.24, and emis-sion current density of 0.5 mA/cm
at an applied field of about 3.63 V/μm. A more excessively acid-treated FWCNT (SN-a FWCNT) compared to N_FWCNT shows the turn-on voltage of 2.83 V/μm, field enhancement factor of 580.23, and emission current density of 0.5 mA/cm
at an applied field of about 3.9 V/μm. The field emission of FWCNT treated with sulfuric/nitric acid for a long time (SN-b FWCNT) was not observed due to the low aligned nanotube density and high defect ratio. The oxidized FWCNTs-based field emitter indicates higher turn-on voltage than the pristine FWCNT-based field emitter because
The ratio of the D-to-G band intensity and field emis-sion characteristics of the pristine FWCNT/TEOS and oxidized FW-CNTs/TEOS based field emittersTurn-on field was calculated as a starting field of the linear regionat the Fowler-Norheim (F-N) plot. The field enhancement factor wasdetermined from the slope of the linear region, β = -BΦ1.5 / S where Bis just a constant, Φ is the work function, and S is the slope of the F-Nplot. We assumed the work function of the nanotube to be 5.0 eV. FWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate.
The ratio of the D-to-G band intensity and field emis-sion characteristics of the pristine FWCNT/TEOS and oxidized FW-CNTs/TEOS based field emitters Turn-on field was calculated as a starting field of the linear regionat the Fowler-Norheim (F-N) plot. The field enhancement factor wasdetermined from the slope of the linear region, β = -BΦ1.5 / S where Bis just a constant, Φ is the work function, and S is the slope of the F-Nplot. We assumed the work function of the nanotube to be 5.0 eV. FWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate.
of the defects on the nanotube surface. There are two possible explanations. First, the defects located on the side wall of nano-tubes can act as an electron trapping site. It is well known that in an ideal nanotube, ballistic electron transfer is realized. That is why the electrical conductivity of nanotubes is extremely high. However, when the nanotube is treated with a strong acid, defects are formed in its side wall, and then the electrons are scattered and/or trapped. Finally, the conductivity of a defective nanotube is lower than that of an ideal one. In the field emission,the conductivity of emitters is also a very important physical property. Second, vertical alignment of nanotubes could be al-tered by a defect. The defects in the side wall of the nanotube mean that pentagons and heptagons are generated, and this re-sults in the bending of the nanotube with a certain angle to the vertical nanotube axis. In general, it is well known that vertically aligned nanotubes have much lower turn-on voltage and higher field enhancement factors than randomly entangled CNTs, indi-cating that most electrons are emitted from the tips of FWCNTs and the electronic DOS of tips near the Fermi energy is different from that of side walls
. Therefore, vertically aligned CNTs without defects can be used as highly effective field emitters.
plot is shown in
b. To calculate the field en-hancement factor (β), we use the
equation. When assuming the work function of CNTs to be 5.0 eV, the pristine FWCNT-based field emitter showed the higher field enhancement factor (β) of 3519.43 compared to the oxidized FWCNTs. We can con-sider that the field enhancement factor is strongly related to the aspect ratio of nanotubes in the vertical direction. As mentioned above, a defective nanotube is bent and thus the vertical height is smaller compared to an ideal one. Also, the bundled structure is enhanced upon acid treatment due to the strong interaction of functional groups. Therefore, the aspect ratio of acid-treated FWCNTs is smaller than that of pristine FWCNT, which ex-plains the high field enhancement factor for pristine FWCNT. It is worth noting that the field emission properties of acid-treated FWCNT at a high electric field are different from those at a low electric field. The field emission properties of acid-treated FW-CNT are significantly enhanced at high electric field. There are
(a) Emission current stability of the pristine FWCNT/TEOS and oxidized FWCNTs/TEOS based field emitters for 2 h at constant voltage. (b) Digital photo-image of emission pattern for N_FWCNT/TEOS based fieldemitter with a size of 2 cm × 2 cm (the corresponding emission currentwas 0.5 mA/cm2). FWCNT: few-walled carbon nanotube, TEOS: tetraethyl orthosilicate.
several papers to explain this phenomenon
. In our case, we can claim that this is related to desorption at a high elec-tric field. At a high electric field, the emission current increases, and thus Joule heating occurs especially at the nanotube tip. In general, intrinsic nanotube emission properties are revealed af-ter desorption of water molecules. However, water absorption is favorable to acid-treated nanotubes due to the polar-polar in-teraction between functional groups and water molecules. That is why pristine FWCNTs do not have a hetero-behavior of the field emission.
We have also compared the emission current stability of oxi-dized FWCNT-based field emitters with that of pristine FW-CNT-based field emitters with a constant field at 0.35 mA/cm
of current density for 2 h.
a shows the degradation rate of the emission current density
at the constant field as a function of time. We found that the emission stability is dra-matically enhanced by using the TEOS binder for improving the adhesion of CNTs to the substrate. We also note that the current degradation rate of N_FWCNT-based field emitters is similar to that of pristine FWCNT-based field emitters, even though the initial current density of pristine FWCNTs is higher than that of N_FWCNTs. Moreover, we consider the emission current stability (
= 68.9% after 2 h) of the SN-a FWCNT-based field emitter to be the worst among the three samples. It is well known that the emission current of nanotubes decreases when a nanotube tip with attached residual oxygen gases is degraded by Joule heating under a high electric field
. In fact, the out-gassing from the cathode is one of the main reasons for vacuum deterioration, resulting in degradation of the emission current. Even though the FWCNTs/TEOS thin films were annealed and cured at 150
C for 90 min, the residual oxygen gases are more favorably absorbed at the tip of acid-treated FWCNTs. There-fore, the current degradation rate of CNTs would mainly depend on the defect level of CNT tips, and thus we can conclude that the decreased emission current stability is mainly attributed to existence of defects on the nanotube surface, which can act as resistance to electron emission, and electrical stress loaded on defects, resulting in failure of field emission. The digital photo-image of the emission pattern from N_FWCNT/TEOS-based field emitter with a size of 2 cm × 2 cm is shown in
b. We observed uniform and extremely bright electron emission from N_FWCNT using phosphor-coated ITO glass as an anode. This plane-type field emission, not spot-type, is attributed to the uniform length, diameter, and density of N_FWCNT-based field emitter.
We have discussed the influence of FWCNTs treated with ni-tric acid/sulfuric acid on field emission characteristics. Carboxyl and hydroxyl groups were functionalized on the surface of the FWCNTs using nitric acid and sulfuric acid in different propor-tions. When the FWCNTs are functionalized with carboxyl and hydroxyl groups, the dispersion stability in DMF was signifi-cantly enhanced due to the combination of polar-polar affinity and electrostatic repulsion.
In addition, FWCNTs/TEOS thin film field emitters were fab-ricated by a spray method using FWCNTs/TEOS one compo-nent solution onto ITO glass. After thermal curing, they were found tightly adhered to ITO glass, and after the activation pro-cess by a taping method, CNTs were aligned preferentially in the vertical direction. The FWCNTs/TEOS-based field emitter with uniform surface, small thickness and strong adhesion to the substrate was successfully fabricated using a spray method. The pristine FWCNT/TEOS-based field emitters revealed a higher current density, lower turn-on field, and higher field enhance-ment factor than the oxidized FWCNTs-based field emitters. However, the unstable dispersion of pristine FWCNTs in TEOS/DMF solution made them poorly applicable to field emitter fab-rication using the spray method. Although the field emitter of nitric acid-treated FWCNT showed slightly lower field emis-sion characteristics, this can be improved by introducing metal nanoparticles or a resistive layer coating.
Thus, we can conclude that our spray method could be use-ful for fabricating field emitters, which offer several advantages compared to previously reported techniques such as CVD and screen printing. For example, as well as a uniform surface, a convenient method to deposit CNTs for cathodes, and good re-producibility, it is cost-effective and amenable to mass produc-tion, and it could potentially lead to the use of CNTs as func-tional devices.
This work was supported by the research fund of Kumoh Na-tional Institute of Technology.
Electron field emission from nanofilament carbon films
Chem Phys Lett
DOI : 10.1016/0009-2614(94)01418-U
A nanotube-based field-emission flat panel display
DOI : 10.1063/1.121493
Patternedselective growth of carbon nanotubes and large field emission fromvertically well-aligned carbon nanotube field emitter arrays
DOI : 10.1063/1.1335846
Carbon nanotubes as electron source in an x-ray tube
DOI : 10.1063/1.1367278
Generation of continuous and pulsed diagnos-tic imaging x-ray radiation using a carbon-nanotube-based field-emission cathode
Appl Phys Lett
DOI : 10.1063/1.1492305
Fieldemission from carbon nanotubes: perspectives for applications andclues to the emission mechanism
Appl Phys A: Mater Sci Process
DOI : 10.1007/s003390050998
Vacuum microelectronics: what’s new and exciting.
IEEE Trans Electron Devices
DOI : 10.1109/16.88510
Stability of carbon nanotubesunder electric field studied by scanning electron microscopy
DOI : 10.1063/1.1429300
Hot nanotubes: stable heating of individual multiwall carbon nanotubes to 2000 K in-duced by the field-emission current
Phys Rev Lett
DOI : 10.1103/PhysRevLett.88.105502
Fully sealed high-brightnesscarbon-nanotube field-emission display
Appl Phys Lett
DOI : 10.1063/1.125253
39.4: large size FEDwith carbon nanotube emitter.
SID Symposium Digest of Technical Papers
DOI : 10.1889/1.1830144
Growth conditions of double-walled carbon nanotubes in arc discharge.
J Phys Chem B
DOI : 10.1021/jp021367o
High-yield catalytic synthesis of thin multiwalled carbon nanotubes.
J Phys Chem B
DOI : 10.1021/jp046152o
Fabrication of efficient field emitters with thin multiwalled carbonnanotubes using spray method
DOI : 10.1016/j.carbon.2006.04.009
An explanation of dispersion states of single-walled carbon nanotubes in solvents and aqueous surfac-tant solutions using solubility parameters.
J Colloid Interface Sci
DOI : 10.1016/j.jcis.2005.01.002
Improved accuracy of quantita-tive XPS analysis using predetermined spectrometer transmission functions with UNIFIT 2004.
Surf Interface Anal
DOI : 10.1002/sia.2056
Fabrication and characterization of polycarbonate/carbon nano-tubes composites
Composites Part A: Appl Sci Manuf
DOI : 10.1016/j.compositesa.2005.08.009
Raman andIR spectroscopy of chemically processed single-walled carbonnanotubes
J Am Chem Soc
DOI : 10.1021/ja052951o
Variations in the Raman peak shift as afunction of hydrostatic pressure for various carbon nanostructures:a simple geometric effect
Phys Rev B
DOI : 10.1103/PhysRevB.67.035417
Controlling the disper-sion of multi-wall carbon nanotubes in aqueous surfactant solu-tion.
DOI : 10.1016/j.carbon.2006.10.010
Attaching proteins to carbon nanotubes via diimide-activated amidation.
DOI : 10.1021/nl010095i
Organic solvent dispersions of single-walled carbon nanotubes: toward so-lutions of pristine nanotubes.
J Phys Chem B
DOI : 10.1021/jp002555m
Electrical Properties of Materials
Oxford University Press
Field Emission in Vacuum Microelectronics
Deviations from the Fowler-Nordheim theory and peculiarities of field electron emission from small-scale objects.
J Vac Sci Technol B
DOI : 10.1116/1.589929
De Heer WA
Field emission from single-wall carbon nanotube films
DOI : 10.1063/1.122037
Electrochemical investigation of single-walled carbon nano-tubes for hydrogen storage.
DOI : 10.1016/s0013-4686(00)00510-7
Oxygen and ozone oxidation-en-hanced field emission of carbon nanotubes
Appl Phys Lett
DOI : 10.1063/1.1485315