In this regard, bet-ter resolutions are realized by increasing the acceleration voltage up to mega-electron-volt level and developing an aberration corrector.2,3 Another method by which
Trang 1Low-carbon steel ultra-high-vacuum Schottky emitter electron gun with double O-rings for axis adjustment
In-Yong Park and Boklae Cho
Citation: J Vac Sci Technol A 35, 020604 (2017); doi: 10.1116/1.4971413
View online: http://dx.doi.org/10.1116/1.4971413
View Table of Contents: http://avs.scitation.org/toc/jva/35/2
Published by the American Vacuum Society
Articles you may be interested in
Metal-vapor integration/transportation based on metal-atom desorption from polymer surfaces with a low glass-transition temperature
J Vac Sci Technol A 35, 020603020603 (2016); 10.1116/1.4971415
Patterning optically clear films: Coplanar transparent and color-contrasted thin films from interdiffused
electrodeposited and solution-processed metal oxides
J Vac Sci Technol A 35, 020602020602 (2016); 10.1116/1.4968549
Evolution of microstructure and macrostress in sputtered hard Ti(Al,V)N films with increasing energy delivered during their growth by bombarding ions
J Vac Sci Technol A 35, 020601020601 (2016); 10.1116/1.4967935
Atomic fluorine densities in electron beam generated plasmas: A high ion to radical ratio source for etching with atomic level precision
J Vac Sci Technol A 35, 01A10401A104 (2016); 10.1116/1.4971416
Trang 2with double O-rings for axis adjustment
In-YongParkand BoklaeChoa)
Division of Industrial Metrology, Korea Research Institute of Standards and Science, 267 Gajeong-ro,
Yuseong-gu, Daejeon 34113, South Korea
(Received 26 September 2016; accepted 14 November 2016; published 6 December 2016)
With the aim to create a simpler structure and reduce the production cost of an existing Schottky
emitter-scanning electron microscope (SE-SEM), the authors have built and tested a double-O-ring
electron gun which is also compatible with ultrahigh vacuum (UHV) Specifically, the gun and
col-umn of the SEM consist of low-carbon steel, of which the magnetic shielding effect is greater than
that of stainless steel, allowing magnification of200 000 in the adapted SEM base without
addi-tional magnetic shielding material, such as permalloy or mu-metal The position of the electron gun
can be adjusted along the horizontal axis while maintaining the UHV condition Excellent beam
current stability with less than 1% variation for more than 1 h was noted Therefore, the authors
anticipate that the double-O-ring electron gun and column of low-carbon steel together represent an
inexpensive and uncomplicated SE-SEM compared to existing types.V C 2016 Author(s) All article
content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY)
license (http://creativecommons.org/licenses/by/4.0/) [http://dx.doi.org/10.1116/1.4971413]
I INTRODUCTION
Since Ruska initially developed the first electron
scope (EM) that exceeded the resolution of an optical
micro-scope in 1933,1 EMs have played a vital role in basic
research and in many industrial applications by providing
shape, structural, and compositional analyses of materials at
a high spatial resolution In cutting-edge research and
indus-tries, EMs capable of atomic-scale imaging resolution have
seen strong levels of demand A current goal has been to
improve the imaging resolutions of EMs In this regard,
bet-ter resolutions are realized by increasing the acceleration
voltage up to mega-electron-volt level and developing an
aberration corrector.2,3 Another method by which to reach
the ultimate resolution involves the electron source, such as
a Schottky emission (SE) or a cold field emission (CFE),
with an even smaller source size and higher angular current
density.4 6 Devising new methods to generate the beam is
very important, as the intrinsic properties of the source can
restrict the resolution fundamentally in spite of a high
accel-eration voltage and the removal of aberrations Tip-based SE
and CFE utilize not only a nanoscale source size but also a
bright electron beam such that a high resolution is possible
despite a low demagnification level as compared to a
tung-sten filament source However, a tip-based electron source
should operate in a UHV; otherwise, the generation of the
electron beam can fluctuate because the surface of the tip
would be contaminated by gas adsorption
Therefore, it is very important to maintain the UHV in the
vicinity of the tip to generate a stable electron beam
Generally, a level in the low 1010Torr range allows for the
stable operation of CFE, and a SE requires the low 109
Torr range for the generation of a stable electron beam
ConFlat (CF) knife-edge flanges are the most common type
for UHVs; they use usually a copper gasket between the two
CF flanges for sealing, allowing the CF flange to be adopted for use with most gun chambers of SE and CFE devices However, opposing knife-edges should impress grooves into the copper gasket, meaning that the bolts between the mating
CF flanges are strongly tightened Therefore, when flanges
or chambers are assembled and disassembled, space is neces-sary for a tool, such as a wrench Furthermore, reusing the copper gasket runs the risk of leaks Thus, the installation of
a new copper gasket is strongly recommended whenever UHV components based on a knife-edge sealing are assem-bled and combined In addition, a mechanical alignment is required in an electron gun when changing its position.7,8In this regard, for CF-type electron guns, flexible bellows and related supports between the flanges are added This can make the structure complicated and can increase the produc-tion cost Furthermore, it is very important to protect a SEM from the external stray magnetic fields, as the electron beam
is easily affected by magnetic fields.9,10 Most UHV SEM columns consist of stainless steel, of which the outgassing rate is low However, stainless steel cannot act as a shield against stray magnetic fields Therefore, permalloy, a high-permeability magnetic material, is inserted inside the stain-less column From the point of construction, a double-wall structure of an electron gun may lead to a more complicated design and a higher production cost Recently, Park et al showed that low-carbon steel (C 0.2 wt %) has a suffi-ciently low outgassing rate such that it is possible to con-struct a UHV chamber of this material.11 Low-carbon steel
is a less expensive and soft magnetic material with high magnetic permeability This allows it in itself to act as a shield against stray magnetic fields
In this study, we present a more compact and less expen-sive structure as compared to that of existing SE-SEM devi-ces We applied differentially pumped double-O-rings to an adjustable electron gun structure along the x and y axes
a)
Electronic mail: blcho@kriss.re.kr
Trang 3while maintaining a proper UHV condition This structure
does not use a CF flange at the electron gun, making
assem-bly and disassemassem-bly relatively easy and quick compared to a
typical electron gun Furthermore, we show that 200 000
imaging is possible without permalloy inside the gun
cham-ber to act as a shield against external stray magnetic fields,
as low-carbon steel is used as the gun chamber material
II EXPERIMENTAL APPARATUS
Currently, SE is widely used as an electron source
because it can emit highly stable electron beam with a
nar-row energy spread and high brightness.12,13 Therefore, we
used SE in our system as an electron source Figure1
illus-trates a cross-sectional view of the designed SE-SEM, which
consists of two condenser lenses, a stigmator, a scanner, and
an objective lens, as in a typical SE-SEM
In order to measure the stability of the generated electron
beam and the SE-SEM image quality, we replaced the
elec-tron gun and condenser lenses of a commercial SM300
(Topcon) device with the aforementioned developed parts
The SE-SEM column is divided into three parts: an electron
gun chamber, an intermediate chamber, and a sample
cham-ber Molybdenum apertures (A 500 lm) are inserted between
the chambers to restrict the flow through the column The
electron gun chamber is pumped out and maintained at
1010Torr using a NEXTorrV R
device (SAES Getters, D 200-5), which combines an ion pump and a nonevaporable
getter (NEG) pump The first condenser lens and variable
apertures (A 30, 50, 100, and 200 lm) are in the intermediate
chamber, which is maintained at108Torr by the ion pump
(pumping speed: 10 l/s) Lastly, the sample chamber, which
has a second condenser lens, a scanner, and an objective lens,
is held to 106Torr with a combination of a diffusion pump and a rotary pump during the imaging process Although vac-uum valves and metal hoses are not shown in Fig.1, the three pumps are connected to each other such that the valves are opened for rough pumping and then closed after the baking process The design described here incorporates double-O-rings in the geometry of the electron gun, making it quite compact and enabling it to slide transversely An O-ring is usually suitable not for UHV applications but the high vac-uum range and rough conditions due to the permeation and the outgassing of materials Therefore, although the chamber sealed with a single O-ring is baked out at the temperature limit of the O-ring, the approximate ultimate pressure reaches only a Torr range of 107–108 In order to decrease the ultimate pressure of the O-ring-based vacuum sealing, two O-rings are used in a conventional UHV feedthrough as sliding seals, and a pumping port between the O-rings is installed to extract any permeated and/or outgassed gases.14,15 Similarly, we applied double-O-rings to the electron gun structure for UHV compatibility and to enable sliding motions Four through holes (A 3 mm) connected to the inter-mediate chamber and positioned so as not to interfere with other ports constitute differential pumping ports and a gauge,
as shown in Fig.1 A SE (DENKA TFE) is attached onto the bottom side of a plate which is insulated to more than 30 kV The DENKA TFE is a ZrO/W Schottky electron emitter of which the work function is reduced by the coating of the tung-sten surface with a layer of zirconium oxide The alumina plate
is in contact with the double-O-rings and slides along the hori-zontal axis by means of four clamping screws We used an alumina content of more than 99% and did not apply a glaze substance onto the surface of the alumina For a good seal with the O-ring, we processed the alumina surface and ensured peak-to-valley roughness measurements of less than 3.2 lm
To control the electron beam current and shape, an extractor and a grounded circular hole plate are positioned under the SE tip The gun valve is closed and maintained at the UHV when the sample chamber is opened to load the sample
III RESULTS AND DISCUSSION
A Vacuum level of the electron gun chamber
As a high-temperature pretreatment of the raw materials, the low-carbon steel was baked at 850C for 1 h in a vacuum furnace to degas the adsorbed gas molecules from the inside
of the material After assembling with machined parts, only the electron gun of the SE-SEM was baked at 150C for approximately 120 h to remove any adsorbed water mole-cules from the inner walls, as water molemole-cules become adsorbed onto the surface after being exposed to the atmo-sphere Both the heating temperature and time are important
in a baking process If possible, a higher temperature and a longer time are better, but when this is not feasible, the two conditions should be controlled to complement each other
We set and maintained the baking temperature to 150C because the recommended maximum operating temperature
of a fluoroelastomer is 150C This temperature is lower
F IG 1 (Color online) Schematic diagram of a designed SEM and a picture
of double O-rings electron gun chamber Pump 1 is NEXTorr V R
, pump 2 is ion pump, and pump 3 consists of diffusion pump and rotary pump Pump 1,
pump 2, and pump 3 are connected with metal hoses and vacuum valve for
rough pumping process at first CL1: first condenser lens, CL2: second
con-denser lens, OL: objective lens, HVC: high voltage cable, EG: electron gun,
GV: gun valve, and VA: variable aperture.
020604-2 I.-Y Park and B Cho: Low-carbon steel ultra-high-vacuum SE electron gun 020604-2
J Vac Sci Technol A, Vol 35, No 2, Mar/Apr 2017
Trang 4than the temperature of a normal baking process for UHV;
therefore, we extended the time as compared to the normal
baking time to compensate for the lower baking temperature
Activation of the NEG pump was accomplished at 450C
for 1 h immediately before the end of the baking step After
the entire baking process, all of the ion pumps were turned
on and the heated SEM was cooled to room temperature
nat-urally Perfluoroelastomer (Kalrez) and fluoroelastomer
(Viton) are widely used as O-ring seals Their characteristics
were compared according to applications.16,17 As shown in
Fig.2, in order to compare both fluorinated elastomers in our
system, the vacuum pressure was carefully examined with
double fluoroelastomers and perfluoroelastomers before,
dur-ing, and after the baking process
The pressure of the double fluoroelastomers gun chamber
was higher than that of the double perfluoroelastomers gun
chamber during the baking process This phenomenon
appears to be related to the outgassing rates of the elastomers
because the fluoroelastomer normally have a higher
outgas-sing rate than the perfluoroelastomer types However, there
were no differences between them during the NEG
activa-tion process due to the large amount of degassed gases from
the NEG The pressure tendencies are similar until the
tem-perature dropped to room temtem-perature; however, the pressure
of the fluoroelastomer type 5 h after the end of the baking
process was 1 order of magnitude lower than that of the
per-fluoroelastomer These remarkable results can be attributed
to the relatively high permeation rate of the
perfluoroelasto-mer type as compared to the fluoroelastoperfluoroelasto-mers type Hence,
the fluoroelastomer is appropriate for our double-O-ring type
of electron gun because the generated electron beam is more
stable at a low pressure level
B Electron beam current stability
Deterioration of the vacuum inside of the SE gun
increases the level of instability of the beam current, which
causes streaks on SEM images That is to say, it is
practi-cally impossible to obtain an ultra-high-resolution image
with an unstable electron beam We determined whether the
beam stability of the double-O-ring electron gun was reliable
or not by measuring the beam current at an extractor and a Faraday cup which was placed under the objective lens The current of the extractor was nearly identical to that of the total emission current from the tip The current of the Faraday cup indicates that the probe current is focused on the target The extractor current was measured using the power supply input current of the extractor (Spellman, EBM30N6/718), and the probe current was measured using
a picoammeter (Keithley, 6485) Figure 3(a) shows a SEM image of the aperture of the Faraday cup
To confirm that the entire probe beam goes inside of the Faraday cup, we increased the magnification of the SEM such that the boundary of the aperture disappeared in the image, indicating that the scanning area of the probe beam is smaller than the size of the aperture of the Faraday cup Moreover, the back plate of the Faraday cup is conical such that incident electrons are not reflected directly back toward the aperture This guaranteed that the total current of the probe beam was measured accurately Figure 3(b)presents the current stability levels of the extractor and the probe beam over a time of 1 h at an acceleration voltage of 20 kV,
an extractor voltage ofþ2.6 kV, and a suppressor voltage of
0.3 kV The condenser lens and variable aperture settings were identical to the conditions of the image at a magnifica-tion of100 000 For stabilization of the SE tip, the electron beam was generated for 48 h before the measurement Although the extractor current was nearly 100 lA, the probe current, actually used for imaging, is only 17 pA, approxi-mately, as most of the beam current is blocked out when passing through the aperture in the beam path The stability was estimated with Eq.(1), in whichS is the beam stability (%), r is the standard deviation, andIaverefers to the average
of the measured current
S¼ 2r
Iave
We compared the long-term stability (1 h) and short-term stability (2 min) of the probe beam, as short-term stability is sometimes very important when attempting to capture scanned images The measured and estimated values are summarized in TableI The long-term stability of the SE is known to be approximately 1% for 1 h under normal operat-ing conditions.4,12,18 Our designed double-O-ring electron gun also provides stability of less than 1% for 1 h, indicating that the proposed method to generate an electron beam can support similar circumstances compared to those of the con-ventional method In addition, the short-term fluctuation of probe beam was observed to be less than 0.2% for 2 min, as shown in Fig.3(c)and in Table I Therefore, it appears that the stability levels of the long- and short-term probe beam are sufficient for application to a common SE-SEM system
C Electron gun alignment and gold particle SE-SEM imaging
The optical components of a SE-SEM device should be aligned along the optical axis in order to minimize distortion
F IG 2 (Color online) Comparison of gun chamber pressure with double
flu-oroelastomers and double perfluflu-oroelastomers before, during, and after
bak-ing process The pressure is measured by an ion gauge.
Trang 5and aberrations of the beam shape With these considerations,
mechanical alignment of SE-SEM components can be
achieved precisely by physically shifting To ascertain the
position of the gun on the optical axis, we acquired a beam
image at the aperture positioned between the first condenser
lens and the second lens by scanning the electron beam at the
aperture The beam position in the image is determined by the
relative position depending on the SE tip and the aperture
Therefore, if the beam direction generated from the SE tip is
not aligned with the aperture along the optical axis, the image
of the beam does not appear in the middle of the scanning
image For the alignment of electron gun by screws, we
designed the mechanical translation range of electron gun to
be approximately 6500 lm because this value can correct the error range that occurs when components are manufactured and assembled We intentionally misaligned the electron gun position by about 500 lm in Fig 4(a) Subsequently, we shifted the electron gun in the horizontal direction using four screw-thread adjustments, thereby moving the beam image to the middle of the image, as shown in Fig 4(b) We also checked whether the afterimages of the beam were concentric
or not when the magnification was changed continuously If misalignment arose, the afterimages could not be concentric
T ABLE I Characteristics of generated electron beam current.
a EC is the current measured at extractor.
b LPC is the probe current measured by Faraday cup for 1 h as long-term.
c SPS is the probe current measured by Faraday cup for 2 min as short-term.
F IG 3 (Color online) (a) SEM image of an aperture of Faraday cup (b)
Current curve of probe beam and extractor for 1 h Solid line indicates the
probe beam current and dot line means extractor current (c) Short term plot
of probe beam current for 2 min.
F IG 4 Beam images according to the scanning aperture which is positioned
at the between first condenser lens and second condenser lens at 10 kV acceleration voltages, (a) when the electron gun is misaligned with aperture
at optical axis, (b) electron gun position is adjusted mechanically approxi-mately 500 lm along the horizontal axis by screw, thereby electron gun and aperture are aligned well at the optical axis.
020604-4 I.-Y Park and B Cho: Low-carbon steel ultra-high-vacuum SE electron gun 020604-4
J Vac Sci Technol A, Vol 35, No 2, Mar/Apr 2017
Trang 6When the SE gun was aligned mechanically by screws,
the alumina plates at which the SE was fixed moved at the
contact surface with the double-O-rings while keeping the
vacuum state Generally, the pressure could change slightly
when sliding motion occurred on the surface of the O-rings
Thus, in order to apply the double-O-ring structure to an
electron gun system, it is very important to ensure
mainte-nance of the vacuum pressure in the UHV range during the
mechanical movement of the alumina plate for proper beam
alignment Therefore, we measured the pressure-time trace
experimentally while moving the alumina plate using a
screw, as shown in Fig.5
Figure 5(a) presents the pressure change when the
alu-mina plate was shifted without the generation of an electron
beam There were small fluctuations, but the range of
1010Torr was still preserved When the electron beam was
generated from a SE tip which was heated to a temperature
of 1800 K, the pressure of the gun chamber was increased to
the 109Torr range because gas molecules are desorbed
from the anode by electron bombardment However, the
pressure still remained in the 109Torr range during the
movement of the alumina plate as shown in Fig.5(b), which
supports the contention of the need for a vacuum condition
to generate an electron beam from the SE tip In our system, although small-scale pressure fluctuations were observed in the UHV range during the mechanical movement step, the double-O-ring design can provide a UHV condition and mechanical alignment simultaneously for an electron gun system
To assess the quality and resolution of the resulting SE-SEM images, we observed evaporated gold particles with an acceleration voltage of 20 kV, as shown in Fig.6 The image quality at100 000 is similar when compared to images of a commercial SE-SEM; however, the resolving power at
200 000 provides less visibility The image resolution in SEM is determined not only by the diameter of the probe beam but also by aberration effects A series of electromag-netic lenses, usually composed of condenser lenses and an objective lens, is used to create a smaller probe size However, the size of the probe beam generally becomes larger than expected due to the aberration effect which arises while the electron beam passes through the series of lenses In order to reduce these aberration effects, the optimum divergence and convergence angles of the cone of the electron beam should be determined by varying the lens power, which affects the beam angle, path and cross-over point in the column We attempted
to find optimized divergence and convergence angles to lessen the aberration influence at 200 000 magnification, but the aber-ration was not satisfactorily eliminated In addition, we reno-vated not a SE-SEM but a tungsten-filament-based SEM with the designed SE electron gun It appeared that the assembly of
a platform for a tungsten-filament-based SEM and SE electron
F IG 5 (Color online) Variations of the gun chamber pressure trace over
time when the SE gun was moved mechanically by a screw (a) Mechanical
moving without the generation of an electron beam, and (b) mechanical
moving when the electron beam was generated from the SE tip.
F IG 6 SEM images of evaporated gold particles at 20 kV acceleration vol-tages: (a) 100 000 magnification and (b) 200 000 magnification.
Trang 7gun is not sufficient for high magnification Although we did
not obtain a high-magnification image such as that produced
by a commercial SE-SEM, we verified that the SE can
gener-ate a stable electron beam and that it can cregener-ate images with
the double-O-ring gun
IV SUMMARY AND CONCLUSIONS
We developed a simple and inexpensive electron gun for
SE-SEM with double-O-rings and a column consisting of a
low-carbon steel material This simple structure allows a
UHV condition for electron beam generation from SE and
provides mechanical movement for an adjustment of the
electron beam while maintaining the pressure, with beam
stability of less than 1% for 1 h By mounting the designed
gun on a commercial tungsten-filament-based SEM, we
showed an image of 200 000 magnification without
per-malloy to shield against stray magnetic fields It is expected
that the developed double-O-ring system can be used not
only with existing SE-SEM systems to decrease the
produc-tion cost but also with UHV devices in cases where the
movement of components is strongly required
ACKNOWLEDGMENTS
The authors would like to acknowledge the financial
support from the R&D Convergence Program of NST
(National Research Council of Science and Technology) of Republic of Korea (Grant No CAP-14-3-KRISS) The authors appreciate B You and J Lim for comment and support in assembly and experimental result analysis
1
E Ruska, Biosci Rep 7, 607 (1987).
2
M A O’Keefe, Ultramicroscopy 108, 196 (2008).
3
T Akash et al., Appl Phys Lett 106, 074101 (2015).
4 L Swanson and N A Martin, J Appl Phys 46, 2029 (1975).
5
G A Schwind, G Magera, and L W Swanson, J Vac Sci Technol., B
24, 2897 (2006).
6 B Cho, K Shigeru, and C Oshima, Rev Sci Instrum 84, 013305 (2013).
7
E N Beebe and V O Kostroun, Rev Sci Instrum 63, 3399 (1992).
8
I.-Y Park, B Cho, C Han, S Shin, D Lee, and S Ahn, Rev Sci Instrum.
86, 016110 (2015).
9
I M€ ullerov a and J Fiser, Rev Sci Instrum 65, 1968 (1994).
10 H Hilbrecht, M Knappertsbusch, and H R Thierstein, Mar Micropaleontol 24, 201 (1995).
11
C Park, T Ha, and B Cho, J Vac Sci Technol., A 34, 021601 (2016).
12 H S Kim, M L Yu, M G R Thomson, E Kratschmer, and T H P Chang, J Vac Sci Technol., B 15, 2284 (1997).
13 D W Tuggle and L W Swanson, J Vac Sci Technol., B 3, 220 (1985).
14
P J Silverman, J Vac Sci Technol., A 2, 76 (1984).
15
Y Aiura and K Kitano, Rev Sci Instrum 83, 035106 (2012).
16
C Ma, E Shero, N Verma, S Gilbert, and F Shadman, J IES 38,
43 (1995), available at http://iestjournal.org/doi/abs/10.17764/jiet.2.38.2 k6491358g0l83691?journalCode=jiet.2
17
S Wang and J M Legare, J Fluorine Chem 122, 113 (2003).
18
Handbook of Charged Particle Optics, 2nd ed., edited by J Orloff (CRC, Boca Raton, 2008).
020604-6 I.-Y Park and B Cho: Low-carbon steel ultra-high-vacuum SE electron gun 020604-6
J Vac Sci Technol A, Vol 35, No 2, Mar/Apr 2017