A method for electrochemically etching field emission tips is presented. Etching parameters are characterized and the operation of the tips in field emission mode is investigated.
A new variation of the drop-off method for fabricating field emission points by electrochemically etching tungsten rods in a NaOH solution is described. The results of studies in which the etching current and the molarity of the NaOH solution used in the etching process were varied are presented. The investigation of the geometry of the tips, by imaging them with a scanning electron microscope, and by operating them in field emission mode is also described. The field emission tips produced are intended to be used as an electron beam source for ion production via electron impact ionization of background gas or vapor in Penning trap mass spectrometry applications.
Sharp tips or points have long been used in microscopy applications, such as the field ion microscope (FIM)1 and the scanning tunneling microscope (STM)2, and a range of techniques for producing sharp tips of various materials have been developed3. These sharp tips can also be operated as field emission points (FEPs) by applying a high voltage to them, and serve as a convenient electron beam source. One application of such as source is ion production via electron impact ionization (EII). The FEP is particularly advantageous in applications where temperature fluctuations produced by thermal emitters are undesirable. For example, ion production via EII of background gas or vapor in high precision Penning traps4,5.
A simple method for fabricating FEPs is to electrochemically etch tungsten rods in a sodium hydroxide (NaOH) solution. This technique is relatively straightforward to implement with modest equipment and has been shown to be quite reproducible and reliable. A number of methods are described in the literature and improvements to these techniques continue to appear6. Here we describe a method for the electrochemical etching of tungsten tips in a NaOH solution. Our method is a variation of the lamella drop-off technique7,8 and the floating layer technique9,10. Like these two methods it enables the production of two tips from a single etching procedure. A picture of the experimental apparatus for etching the tips is shown in Figure 1.
Figure 1. Etching apparatus. Photograph of the experimental apparatus used for electrochemical etching of tungsten rods with NaOH solution. Please click here to view a larger version of this figure.
Electrochemical etching of tungsten in the aqueous NaOH base occurs via a two-stage process. First, intermediate tungsten oxides are formed, and second, these oxides are non-electrochemically dissolved to form the soluble tungstate anion. This process is described, in simplified form, by the two reactions
(1) W + 6OH– → WO3(S) + 3H2O + 6e–, and
(2) WO3(S) + 2OH– → WO42- + H2O.
The etching current and the NaOH solution molarity used affect the time and voltage required to etch through the tungsten rod. Studies of these effects are presented and discussed. More importantly, the etching parameters have an effect on the geometry of the tips and, as such, on their operation in field emission mode. The geometry of the tips we produced were characterized by imaging them with a scanning electron microscope (SEM). These images can be used to estimate, for example, the tip radius. In addition, the tips were operated in field emission mode by applying a negative voltage of typically a few hundred volts to a few kilovolts to them and monitoring the resulting electron emission current. The relationship between field emission current, I, and applied bias voltage, V, can be described by the Fowler-Nordheim equation11
(3) I = AV2 e-Creff/V,
where reff is the effective radius of the tip, A is a constant, and C is the second Fowler-Nordheim constant , in which b = 6.83 eV–3/2V/nm, is the work function of tungsten ( ≈ 4.5 eV), k is a factor that depends on geometry (k ≈ 5), and is the Nordheim image correction term ( ≈ 1)12. Hence, the effective radius of the tip can be determined by measuring the electron current as a function of bias voltage. Specifically, it can be obtained from the slope of a so-called Fowler-Nordheim (FN) plot of ln(I/V2) vs 1/V.
1. Electrochemical Etching
Figure 2. Schematic of etching circuit. A schematic drawing of the etching circuit used to provide the constant DC etching current. The current is determined by monitoring the voltage across a low resistance resistor and the voltage is recorded by monitoring the voltage across a high resistance resistor using an ADC. A computer program monitors the current and provides a 5 V output signal to a relay that opens the etching circuit once the current drops below a specified value. Please click here to view a larger version of this figure.
2. Characterization of Field Emission Points
Figure 3. Optical image of FEP tips. Picture of (a) a good tip and (b) a bad tip, as viewed through an optical microscope. Please click here to view a larger version of this figure.
Figure 4. FEP holder for SEM imaging. A picture of (a) the top and (b) the bottom of the holder used to secure FEPs while imaging with the SEM. Please click here to view a larger version of this figure.
Figure 5. Field emission apparatus. Schematic of the apparatus used to apply a HV to the FEPs while under vacuum to produce an electron beam. The electron beam current is monitored on the Faraday cup with a picoammeter. Please click here to view a larger version of this figure.
Study of etching parameters
During the etching process the power supply is operated in constant current mode. The voltage required to maintain this constant current increases slightly as the tungsten rod is etched away (due to the increase in resistance of the rod). The current drops almost to zero when the tip etches all the way through. A small current continues to flow due to the fact that the upper tip is still in contact with the etching solution. A plot of current and voltage as a function of time during the etching process is shown in Figure 6.
Figure 6. Current and voltage during etching process. The current and voltage supplied by the power supply during the etching process. The voltage required to maintain the constant current increases slightly during the etching process due to the increase in resistance as the tungsten rod etches away. The error bars on the voltage data points, determined as the standard uncertainty in the voltage data averaged in 15 sec bins, also increase in size during the etching period as a result of larger voltage fluctuations. The current drops almost to zero when the tungsten rod etches all the way through and the bottom tip drops off. Please click here to view a larger version of this figure.
Reprinted from Int. J. Mass Spectrom., Vol. 379, M. Redshaw, et al., Fabrication and characterization of field emission points for ion production in Penning trap applications, Pages 187 – 193, Copyright (2015), with permission from Elsevier.
The time required to etch through the tungsten rod depends on the etching current used and on the molarity of the solution. Figure 7 (a) shows the time required to etch through a 0.5 mm diameter tungsten rod as a function of etching current for three different molarity NaOH solutions. The etching rate increases linearly with current. Power law fits of etching time as a function of current gave exponents of 1 for all three NaOH solution molarities. Figure 7 (b) shows that the etching voltage is linearly proportional to current and that the voltage required to deliver the constant current decreases with increasing molarity. This relationship is to be expected from Ohm's Law: the number of charge carriers available in the solution, and hence the effective conductance, is determined by the molarity of the solution. The dependence of etching time, or inverse etching rate, on current, as seen in Figure 7 (a) is expected based on Eqn. (1). However, Figure 7 (a) also shows that, for low current settings of 100 mA, the etching rate decreases with increasing molarity. This could result from the lower potential required to maintain this current for the high molarity solution, since etching current also depends on the potential required to drive the reaction15.
Figure 7. Etching time and voltage vs current and molarity. (a) Main: Time required to etch through 0.5 mm diameter tungsten rods as a function of etching current for NaOH solution molarities of 0.75, 1.5, and 3.0. (b) Inset: Average voltage supplied by the constant-current power supply during the etching process. Please click here to view a larger version of this figure.
Reprinted from Int. J. Mass Spectrom., Vol. 379, M. Redshaw, et al., Fabrication and characterization of field emission points for ion production in Penning trap applications, Pages 187 – 193, Copyright (2015), with permission from Elsevier.
SEM imaging:
SEM imaging can be used to reveal the structure of the tip. Figure 8 shows SEM images of (a) top and (b) bottom tips. In (i), the bottom tips can be seen to have a larger aspect ratio than the top tips. This is due to the fact that some etching solution runs down the tungsten rod, etching or polishing the surface. The images in (ii) and (iii) show that the bottom tips generally have an acute cone angle and in many cases have a large bulb at the tip, increasing the effective radius of curvature. The upper tips on the other hand generally taper to a fine point.
Figure 8. SEM images of field emission tips. SEM images of (a) top and (b) bottom tips etched from a 0.5 mm diameter tungsten rod using 0.75 M NaOH solution and a nominally 200 mA etching current, shown with magnifications of (i) 35X, (ii) 1,800X, and (iii) 37,000X. Please click here to view a larger version of this figure.
Reprinted from Int. J. Mass Spectrom., Vol. 379, M. Redshaw, et al., Fabrication and characterization of field emission points for ion production in Penning trap applications, Pages 187 – 193, Copyright (2015), with permission from Elsevier.
The bulb structure seen on the lower tips has been observed by other researchers, e.g., Ibe et al.15, and is attributed to the recoil force on the tip as it fractures and the bottom piece drops off. In this scenario, the energy released during fracturing can cause local melting, deforming the tip. The upper tips do not show a corresponding bulb. We attribute this to the post-drop off etching period after the lower tips drops off, but before the current is completely switched off (the current decreases significantly after the lower tip falls off, but does not go completely to zero since the upper tip is still in contact with the etching solution).
Field emission tests:
The FEPs were operated in field emission mode by applying a negative bias of between a few hundred volts and a few kilovolts between the FEP and ground. The field emission electrons struck a Faraday cup and the current was recorded. The field emission current as a function of bias voltage was investigated. A plot of ln(I/V2) vs 1/V shows a linearly decreasing dependence. This relationship is well described by the Fowler-Nordheim equation. Using this equation and the slope of the data in the Fowler-Nordheim (FN) plot, the effective radius of the tip can be extracted. These measurements were consistent with results obtained from the SEM images14. The tips were conditioned for ~1 hr by operating them in field emission mode with a constant current of ~5 nA. After this time, the measurement of field emission current vs bias voltage was repeated. In general, the location of the data on the FN plot and the slope changed. In Figure 9, it can be seen that after the conditioning process the tip fires at a lower voltage and the slope has decreased. This indicates that the effective radius of the tip has decreased and hence the electric field required to remove electrons from the tip can be attained at a lower bias potential.
Figure 9. Fowler-Nordheim plot. Plot of ln(I/V2) as a function of 1/V obtained by scanning the bias voltage, V, applied to the FEP and recording the average field emission current, I, produced by the tip. The two data sets correspond to measurements taken after the FEP first fired and after conditioning it for 1 hr. The straight lines are linear least-squares fits to the data, the slope of which is proportional to the effective radius of the tip. Please click here to view a larger version of this figure.
Reprinted from Int. J. Mass Spectrom., Vol. 379, M. Redshaw, et al., Fabrication and characterization of field emission points for ion production in Penning trap applications, Pages 187 – 193, Copyright (2015), with permission from Elsevier.
The conditioning of the FEPs, inferred from the data shown in Figure 9, suggests that the field emission process can reduce the effective radius of the tip of the FEP. This behavior has been observed by other researchers and is attributed to heating of the tip by the electron current, and sputtering by atoms and molecules in the background gas of the vacuum that are ionized by the electron beam and accelerated toward the tip of the FEP16,17. In our apparatus, the main background gas (determined with a residual gas analyzer) was H2O, and the most abundant ion species produced was H3O+ (determined via the cyclotron frequency of ions in a Penning trap14). Heating can clean the end of the FEP and also melt the tip. The results of melting range from a rearrangement of the atoms at the apex, which can sharpen the tip, to producing a molten blob of material at the end of the tip, blunting it. Sputtering can remove material from the tip, hence sharpening it, and can also decapitate the tip of the FEP. Significant changes in the field emission current were often observed during the conditioning process and SEM images of FEPs after field emission showed significant changes in the geometry of the tip, including the formation of blobs of melted metal at the tip, bent tips, and tips that had been decapitated-see Redshaw et al. for further details14.
We have described straightforward procedures to electrochemically etch sharp field emission points (FEPs) in a NaOH solution, and to test the FEPs by operating them in field emission mode. The etching procedure described is a variation of existing techniques-the lamella drop-off technique7,8 and the floating layer technique9,10. However, we found it to be more convenient and reliable to implement than the aforementioned methods.
Before starting the etching procedure, to minimize the likelihood of producing tips with gross deformations, e.g., a bent tip, as shown in Figure 2, the tungsten rod must be aligned through the hole in the copper cathode as vertically as possible. During etching, the NaOH drip rate from the separatory funnel should be monitored to ensure that the level of NaOH in the small reservoir in the copper cathode plate remains approximately constant. At the end of the etching procedure, the bottom tip will drop off, and the etching current will be greatly reduced. Shortly after this drop-off, the etching current should be switched off completely to avoid blunting the tip by continued etching. However, some etching/polishing of the tip at this stage is beneficial to the production of FEPs to be used as electron beam sources, since it appears that this polishing stage can smooth the tip and remove irregularities14. In our set-up a cut-off time of ~100 msec after the bottom tip drops off was used to produce tips with radii of ~100 nm. Other researchers have used fast transistor-based cut-off circuits to stop the etching process in as little as 500 nsec after the drop-off of the lower part of the tip, resulting in tips with radii down to ~10 nm to be used for STM applications12,15. Such a circuit was also tested in our set-up, and enabled tips with <100 nm radii to be produced. However, we found that these tips were less uniform at the tip and did not perform so well in field emission mode because, we believe, the smaller tips made them more susceptible to being melted by the electron beam current.
Field emission was initiated by applying a negative HV to the FEP, which was gradually increased until the FEP fired. The voltage required to initiate field emission depends on the geometry of the tip, and is typically lower for sharper tips14. While firing the FEP for the first time, the HV should not be scanned too quickly (~250 V/sec) to avoid a sudden current spike. We generally kept the electron beam current below 1 μA to avoid melting the tip. After the tip had fired, we conditioned it for 1 hr by operating it in field emission mode with an electron beam current of ~5 nA. We found that this procedure made the tip more stable, i.e., the HV required to produce a given electron beam current (typically 1 nA or less in our application) remained fairly constant.
In summary, we have presented a straight-forward technique for electrochemically etching sharp FEPs from tungsten rods. These FEPs have been successfully operated in field emission mode with voltages ranging from a few hundred volts to a few kilovolts to produce an emission current on the order of nA. These FEP tips have also been implemented in a Penning trap mass spectrometry application14.
The authors have nothing to disclose.
We acknowledge the services of Stanley Flegler, Carol Flegler, and Abigail Tirrell at the MSU Center for Advanced Microscopy. We thank Ray Clark and Mark Wilson for technical assistance with the set-up of the electrochemical etching apparatus. Earlier contributions from Anne Benjamin, Georg Bollen, Rafael Ferrer, David Lincoln, Stefan Schwarz and Adrian Valverde, and technical assistance from John Yurkon are also acknowledged. This work was partially supported by the National Science Foundation contract no. PHY-1102511 and PHY-1307233, Michigan State University and the Facility for Rare Isotope Beams, and Central Michigan University.
Tungsten Rod 0.020" x 12" | ESPI Metals | http://www.espimetals.com/index.php/online-catalog/467-Tungsten | 3N8 Purity |
NaOH salt | Cole-Parmer | Item # WU-88404-71 | 100 g |
Separatory funnel | Cole-Parmer | Item# WU-34506-03 | 250 mL |
DC Power supply | BK Precision | 1672 | Triple Output 0 – 32 V, 0 – 3 A DC Power Supply |
Acetone | Cole-Parmer | Item# WU-88000-68 | 500 mL |
Data Acquisition Card | National Instruments | NI PXI-6221 | 16 AI, 24 DIO, 2 AO |
Relay | Magnecraft | 276 XAXH-5D | 7 A, 30 V DC Reed Relay |
6-way 6" conflat flange cross | Kurt J Lesker | C6-0600 | |
6" to 2-3/4" conflat zero length reducer flange (x3) | Kurt J Lesker | RF600X275 | |
2-3/4" conflat flange SHV feedthrough | Kurt J Lesker | IFTSG041033 | |
2-3/4" conflat flange BNC feedthrough | Kurt J Lesker | IFTBG042033 | |
2-3/4" conflat flange linear feedthrough | MDC | 660006, REF# BLM-275-2 | |
6" conflat flange blankoff | Kurt J Lesker | F0600X000N | |
6" conflat flange window | Kurt J Lesker | VPZL-600 | |
HV Power supply | Keithley Instruments | Keithley Model #2290-5 | 0 – 5 kV DC HV Power Supply |
Picoammeter | Keithley Instruments | Keithley Model #6485 | |
Faraday Cup | Beam Imaging Solutions | Model FC-1 Faraday Cup |