Characterization of X-ray charge neutralizer using carbon-nanotube field emitter
Shuhei Okawaki, Satoshi Abo, Fujio Wakaya, Hayato Yamashita, Masayuki Abe, Mikio Takai
CCharacterization of X-ray charge neutralizer using carbon-nanotube fieldemitter
Shuhei Okawaki, Satoshi Abo, Fujio Wakaya, a) Hayato Yamashita, Masayuki Abe, and Mikio Takai
Graduate School of Engineering Science, Osaka University, Toyonaka, Osaka 560-8531 Japan (Dated: Dec. 9, 2015 submitted to Jpn. J. Appl. Phys; revised on March 31, 2016)
An X-ray charge neutralizer using a screen-printed carbon-nanotube field emitter is demonstrated to show thepossibility of a large-area flat-panel charge neutralizer, although the device dimensions in the present work are not verylarge. The X-ray yields and spectra are characterized to estimate the ion generation rate as one of the figures of meritof neutralizers. Charge neutralization characteristics are measured and show good performance.Journal reference: Jpn. J. Appl. Phys. , 06GF10 (2016) DOI: 10.7567 / JJAP.55.06GF10
I. INTRODUCTION
It is widely known that static electricity causes troublesnot only in high-tech industries but also in many fields in-cluding the manufacture of plastic, rubber, powder, and pa-per, and industries of textile, spinning, and printing, meaningthat the management of static electricity has been importantin many industries for a long time.
In the ULSI industry,the device dimensions become continuously small , leadingto a more fragile device against static electricity. In the indus-try of flat-panel displays, the screen size and pixel dimensionbecome continuously large and small, respectively, meaningthat the product becomes more sensitive to particle contami-nations caused by static electricity. These outstanding trendsin high-tech industries suggest that the management of staticelectricity becomes more critical presently.Although material modification by antistatic additive dop-ing or surface coating is e ff ective for overcoming static elec-tricity problems , such a technique cannot always be adoptedbecause, especially for electronic devices, an insulating mate-rial is necessary in many cases for realizing device functions.Charge neutralizers using air molecules ionized by corona dis-charge or soft X-ray irradiation are, therefore, often used forsolving static electricity problems. Corona-discharge-type charge neutralizers have the follow-ing disadvantages, although they are widely used: (1) thecharge balance of ionized air is not good, (2) the dischargeprocess generates particles and causes contamination prob-lems, (3) the density of ionized air is low on the target materialsurface, although it is high around the discharge electrodes,(4) the discharge process generates electromagnetic noise thatmay cause troubles in the target devices to be neutralized. TheX-ray charge neutralizer has advantages concerning all theabove problems. Particularly when the X-ray source is shapedto a large-area flat panel, it should be useful for present large-area flat-panel devices because of the uniform ion density overa large area.A carbon nanotube (CNT) is one of the promising materialsfor electron field emitters owing to its high aspect ratio, high a) Corresponding author: [email protected] current tolerance, high mechanical strength, and high chemi-cal stability. . Many applications using CNTs as field emit-ters, such as a field emission display, a backlight unit fora liquid crystal display, and an X-ray source, are re-ported. A screen-printed CNT mat with appropriate post sur-face treatments is one of the best candidates for realizing alarge-area field emitter and can be applied to a large-areaX-ray source . Although the X-ray generated at the large-area source may not be very suitable for high-resolution X-rayimaging, it should be useful for realizing a large-area X-raycharge neutralizer for large-area flat-panel devices.In this study, an X-ray charge neutralizer using a screen-printed CNT cathode is demonstrated. Characterization re-sults show good performance for a charge neutralizer.
II. EXPERIMENTAL METHODS
To apply electron field emitters to an X-ray source, a three-terminal configuration with emitters, gate electrodes, and ananode is necessary; such a configuration enables us to controlseparately the anode current from the anode voltage, leadingto the separate control of the X-ray intensity and X-ray en-ergy. An in-plane side-gate structure is adopted in the presentwork because it is easily applied to a large-area flat-panelemitter. The fabrication process for the emitter, which was al-ready reported, is summarized as below. Indium-tin-oxide(ITO) electrodes of 30 × , some of which were used asside-gate electrodes and others as back electrodes for CNTcathodes, were deposited on a glass substrate. A CNT pastewas screen-printed on a part of the ITO electrodes with an areaof 10 × . The gap between the edges of the CNT cathodeand side-gate electrode is 100 µ m. The schematic top view ofthe cathode is shown in Fig. 1. The e ff ective area of the cath-ode, which is ∼ ×
10 mm , is not very large but it can easilybe enlarged because screen printing can easily be applied toa large-area process. The tape-peeling surface treatment wasperformed to improve the electron emission property. The experimental setup used to generate and detect X-raysis schematically shown in Fig. 2. In a vacuum of ∼ − Pa, a10- µ m-thick Cu thin film, placed 8 cm away from the emitter,was irradiated by field-emitted electrons from the CNT cath-ode typically at 10 keV to generate X-rays. To characterize a r X i v : . [ phy s i c s . a pp - ph ] D ec glass substrateITO electrode screen-printed CNTside-gate voltage
10 mm1 mm
FIG. 1. (Color online) Schematic top view of screen-printed CNTcathode with side-gate electrodes. The bias setup for extracting elec-trons is also shown. the X-ray spectrum and yield in vacuum, the X-ray detectorwas set 4 cm away from the Cu film as shown in Fig. 2(a).To characterize X-rays in air, a 250- µ m-thick Be window wasused and the X-ray detector was set in air 5 cm away from theBe window shown in Fig. 2(b). The X-ray detection systemused is Amptek XR-100CR / PX4 with a detector area of 13mm .Ion current was measured in air by a metal plate of 40 × , placed 5 cm away from the Be window instead of the X-ray detector shown in Fig. 2(b). The bias voltage of the platewas kept constant at − Charge neu-tralization performance as a function of time was character-ized by using the charged plate monitor shown schematicallyin Fig. 3 placed in air 5 cm away from the Be window. Thearea and capacitance of the charged plate monitor are 25 × and 25 pF, respectively. The current I gr defined in Fig. 3was measured as a function of time t , since the high-voltagesource was disconnected from the circuit. The electric poten-tial of the charged plate, V cp , can be estimated as V cp ( t ) = V cp (0) − C (cid:90) t I gr d t , (1)with V cp (0) = C =
25 pF. The direct measurementof V cp is di ffi cult because the ion current and I gr are typically10 − A at the electric potential of 10 V as discussed in the fol-lowing section, which means that a voltmeter input impedanceof more than 10 Ω is necessary. This is the reason why theestimation using Eq. (1) is adopted in this work. III. RESULTS AND DISCUSSION
The anode current of the CNT cathode with side-gate elec-trodes was controlled by the side-gate voltage and reached the anode voltageelectronX-ray10 -5 PaCu
X-ray detector(a) anode voltageelectronX-ray10 -5 PaCu
13 cm8 cm
X-ray detectorBe air
X-ray(b)
FIG. 2. (Color online) Experimental setup for generating and detect-ing X-rays. The X-ray detector is placed in vacuum (a) or in air (b). highest value of 300 µ A. The detailed field-emission proper-ties were similar to those observed in the previous work .Figure 4 shows the X-ray spectra obtained in vacuum, thesetup for which is shown in Fig. 2(a). The anode current dur-ing the measurement was reduced and kept at 1 nA, althoughit can be increased up to 300 µ A as described previously, be-cause the X-ray spectrum cannot be measured if it is verystrong. The characteristic X-ray peaks of Cu were observedwhen the anode voltage was 10 kV. For all anode voltages, themaximum energy of the bremsstrahlung X-ray correspondedto the anode voltage as expected.The X-ray spectra obtained in air are shown in Fig. 5. Whenthe spectrum was measured in air, the X-ray was absorbed byair and its intensity decreased. The anode current was, there-fore, increased to 400 nA, while the other parameters weremaintained the same as in the in-vacuum measurement shownin Fig. 4. The maximum energies of the bremsstrahlung X- A air charged plate ( )ground plateinsulator FIG. 3. (Color online) Schematic of charged plate monitor with mea-surement setup. Energy (keV) Y i e l d ( c oun t / E ) FIG. 4. (Color online) Energy spectra of X-rays measured in vacuumas in Fig. 2(a). The channel width ∆ E of the multichannel analyzer is89.9 eV. The anode current is fixed to 1 nA during the measurement.The measurement time for each spectrum is 60 s. rays with anode voltages of 8 and 10 kV exceed the corre-sponding anode voltages not as expected. This is due to thedouble or multiple counting of photons during the time con-stant of the detection system.For the charge neutralizer, the ion generation rate G is oneof the important figures of merit, which can be estimated as G ( T , P ) = S ∆ t (cid:88) i E i W Y ( E i ) (cid:34) µ en ( E i ) ρ (cid:35) ρ ( T , P ) , (2)where T and P are the air temperature and pressure, respec-tively, S is the area of the X-ray detector, ∆ t is the measure-ment time for obtaining the spectrum, E i is the X-ray energyof the i th channel of the detector, W is the average ioniza-tion energy of air, Y ( E i ) is the X-ray yield, µ en ( E i ) is the X-ray absorption coe ffi cient of air, ρ is the mass density of air.Figure 6 shows the ion generation rate estimated from thespectrum shown in Fig. 5 using Eq. (2) with S =
13 mm , ∆ t =
60 s, W =
34 eV, µ en /ρ from the database, and ρ = . × − g / cm at 760 Torr and 20 ◦ C. The maximum iongeneration rate in Fig. 6 is 1 . × cm − s − , corresponding Energy (keV) Y i e l d ( c oun t / E ) FIG. 5. (Color online) Energy spectra of X-ray measured in air as inFig. 2(b). The channel width ∆ E of the multichannel analyzer is 89.9eV. The anode current is fixed to 400 nA during the measurement.The measurement time for each spectrum is 60 s. Anode Voltage (kV) I on G ene r a t i on R a t e ( c m - s - ) FIG. 6. (Color online) Estimated ion generation rate at 760 Torr and20 ◦ C as a function of anode voltage. to the ion concentration C ion of 5 . × cm − . It is reported that C ion = cm − or G = . × cm − s − is enough for acharge neutralizer. The maximum G estimated in the presentwork is two orders of magnitude lower than this value. The iongeneration rate can, however, be increased by increasing theanode current, which was intendedly reduced from 300 µ A to400 nA in order to obtain the X-ray spectrum shown in Fig. 5.Assuming that G is proportional to the anode current, we canexpect G to be 750 times larger than those in the present workwith an anode current of 300 µ A, the upper limit of the emitterused in the present work. This means that the charge neutral-ization performance of the present device is fairly good.The ion currents observed in air by using the metal platebiased at − A) I on C u rr en t ( n A )
10 kV 8 kV 5 kVAnode Voltage
FIG. 7. (Color online) Ion current measured in the air as a functionof anode current. C u rr en t ( n A ) with X-ray irradiation without X-ray irradiation FIG. 8. (Color online) Current measured by the charged plate mon-itor shown in Fig. 3. The distance between the Be window and thecharged plate is 5 cm. The anode voltage and anode current were 10kV and 300 µ A, respectively, during the measurement. than that of the corona-discharge-type neutralizer , althoughthe measurement with positive bias was not performed owingto the voltage source restriction.The measured current to the ground plate I gr , defined inFig. 3, is shown in Fig. 8. For this measurement, the anodecurrent was increased and fixed to 300 µ A to obtain the high-est neutralization performance. When the current I gr shown inthis figure is used with Eq. (1), V cp (cid:44) ffi cientlylong time. This is probably due to the unreliable value of thecapacitance C . If C is assumed to be ∼
10% larger than 25pF, V cp = ffi ciently long time. The following nor-malization was, therefore, used instead of Eq. (1) to avoid theproblem: V cp ( t ) = V cp (0) − (cid:82) t I gr d t (cid:82) ∞ I gr d t . (3)The resulting estimated electronic potentials at the chargedplate with and without X-ray irradiation are shown in Fig. 9. E s t i m a t ed V o l t age ( k V ) with X-ray irradiation without X-ray irradiation FIG. 9. (Color online) Voltage at the charged plate as a function oftime estimated from the observed current shown in Fig. 8.
The decay observed without X-ray irradiation should be due tothe natural neutralization of the charged plate interacting withthe surrounding air. If the decay time is defined as the time atwhich the voltage becomes 1 /
10 of the initial value, it is esti-mated to be 17.5 and 215 s with and without X-ray irradiation,respectively. The decay time of 17.5 s with X-ray irradiationis not very good compared with the reference , but can be im-proved by increasing the X-ray intensity. The X-ray intensitycan be increased by decreasing the distances between the Cutarget and the Be window and between the Be window andthe object material to be neutralized, which are 13 and 5 cm,respectively, in the present work for the preliminary demon-stration. IV. CONCLUSIONS
An X-ray charge neutralizer was demonstrated by usinga screen-printed CNT field emitter. The e ff ective area of theemitter is ∼ ×
10 mm . This is not very large and is almosta point source because the distances in the present work be-tween the emitter and the Cu target, between the Cu targetand the Be window, and between the Be window and the ob-ject material to be neutralized are 8, 13, and 5 cm, respec-tively. To realize a large-area flat-panel source is not di ffi -cult because screen printing can be easily applied to a large-area process. Charge neutralization characteristics were mea-sured and showed good performance even under such an al-most point-source condition, suggesting that the performanceis much improved when a vacuum-sealed large-area flat-panelcharge neutralizer is realized using the screen-printed CNTfield emitter. ACKNOWLEDGMENTS
This work was partially supported by JSPS KAKENHIGrant Number 23360022. C. G. Noll, in
Handbook of Electrostatic Processes , ed. J.-S. Chang, A. J.Kelly, and J. M. Crowley, (Marcel Dekker, New York, 1995) Chap. 33. Seidenki Joden no Sochi to Gijutsu (Ionizer and Tehnique of Charge Elimi-nating), ed. Y. Murata (CMC, Tokyo, 2004) [in Japanerse]. International Technology Roadmap for Semiconductors 2013 Edition Ex-ecutive Summary (2013) [http: // K.-S. Choi, T. Fujiki, and Y. Murata, Jpn. J. Appl. Phys. , 7693 (2004). K.-S. Choi, S. Nakamura, and Y. Murata, Jpn. J. Appl. Phys. , 3248(2005). H. Inaba, T. Ohmi, T. Yoshida, and T. Okada, J. Electrostat. , 15 (1994). H. Inaba, T. Ohmi, T. Yoshida, and T. Okada, IEICE Trans. Electron.
E79-C , 328 (1996). J. Kawai, H. Ishii, and Y. Hosokawa,