10 Low Voltage Surface Electron Microscopy
Dr. Ajit K. Mahapatro
Learning Objectives
From this module students may get to know about the following
i. Changes in beam-surface interaction for LVSEM
ii. Advantages and limitations of LVSEM
iii. Contrast formation in LVSEM
1. Introduction
In scanning electron microscopy (SEM), as the beam voltage V0 is reduced in the range 1-30 kV, three physical parameters relevant to specimen damage, surface charging, and topographic contrast also change.
0.5 The secondary electron coefficient (δ) increases to unity, the electron range (R = kV0) and the energy deposited per electron (eV0) both decrease. When ‘δ’ is large, there is less surface charging on the insulating samples and there is more signal per beam electron. A smaller electron range ‘R’ means the beam-specimen interaction is more localized and the image contrast is higher. The possible technical limitations can be grouped in four categories: (1) low source brightness, (2) increased effect of chromatic aberration, (3) increased sensitivity to stray fields, and (4) defocusing of the probe by the secondary electron (SE) collection field.
As the beam voltage is reduced to a value where secondary emission coefficient is equal to or greater than unity, the surface potential gets stabilized. These conditions are independent of the angle of incidence of the primary beam and the collector is controlling the image contrast. The image contrast is resulting from the changes in the secondary electron coefficient, δ from different elements, and this effect would be largest when the SE yield is maximum, i.e. at low beam energies.
2. Choice of Electron Energy Range
The interaction between the electrons and the material is much stronger in the energy range of 100 eV to keV than at higher energies. The consequence of this stronger interaction is that the penetrating depth of the electron fall rapidly as the incident energy is reduced. The penetration depth is measured using Bethe range (RB),
The low voltage image is more sensitive to the chemical nature and topographic form of the specimen surface. In the low voltage SEM, since all the signals are produced near the surface of the materials, even very thin specimens could behave as bulk.
The relationship between the secondary electrons (SE) and back scattered electrons (BSE) images could be reversed under low voltage operation. The contrast information of the SE image for typical SEM energies of 20-30 keV, generates from a depth of about 4λ from the surface, i.e. about 15-20 nm. However, the average BSE electron emerges from a depth of about 0.2 RB, i.e. approximately 0.25 µm or more. BSE carries the information and SE carries the surface information of the materials. At low voltages, the energy distribution of the SE does not change with the incident beam energy, leading to the identical depth of information for the SE i.e. 15-20 nm. However, the depth for backscattered signal (0.2 RB) reduces to 3-5 nm due to the fall in electron range. Hence, SEs are emitted from the whole of the interaction volume at low energies while the BSEs emerge from the near top of the specimen’s surface. Thus at low voltages, the SEs perform the task of bulk imaging.
Figure 1. Schematic diagram of the beam-sample interactions in a large-chamber of SEM. At low voltages the SEI and SEII components provide high-resolution information, and the SEIII and SEIV components are noise that degrades the signal-to-noise ratio.
3. Advantages of LVSEM
In LVSEM (V≈1 kV), the electrons used for impinging the specimen’s surface with low energy, penetrate a shorter distance through the surface, possessing a larger cross-section and produce SEs near the surface where they can escape.
- As the signal/beam electron ratio increases i.e. the secondary electron coefficient (δ) approaches unity. The charging effect also diminishes due to the decrease in amount of energy deposited on the sample. This proves to be of great advantage for non-conducting photo resist and passivation layer.
- Another advantage is the result of reduction in beam penetration, results preventing trapping of charges that might distort the energy band structure of the sample, and degrades its functional properties. At 1 kV, such degradation is restricted to the outer surface of 0.02 µm and goes 130 times deeper at 30 kV, making a significant difference in samples of few micrometer dimension.
4. Contrast formation in LVSEM
As discussed, the changes in the electron-solid interaction with the incident beam energy effects the the contrast formation in the SEM images. The most significant effects are the reduction and eventual loss of contrast from topographic image and beam penetration.
The beam penetration into the sample contributes on the SEM image contrast. At high beam energies this occurs over a scale of many micrometers and features raised above the general surrounding surface are exposed by the incident beam irradiation, leading to enhanced SE emission from all the free surfaces of the feature. The 20-30 keV micrograph shows all the expected features of a typical SE image, contrasting from the topography of the sample surface, edge brightness enhancing the shape of features, and a three- dimensional appearance that makes easy interpretation of the image.
Figure 2. Interaction volume of electrons in carbon at 1, 3 and 10 keV beam energies. Monte Carlo simulation [6] using experimental stopping power data and Mott cross-sections.
As the primary beam voltage is reduced, the beam penetration becomes smaller than the typical dimensions of the sample’s feature. The resulting image flattens and provides a two dimensional appearance, which can be explained as the replacement of the earlier image contrast by a simpler directional shading, where one side of a feature is always brighter than the other side. Thus, analyzing the low voltage 1 keV SEM micrograph in comparision with the higher voltage SEM image would not be appropriate. It is also evident that the surface contamination marks, scratches and dust particles which are barely visible at 10 keV become dominant for contrast formation in the 1 keV image due to a significant change in the form of the image contrast associated with a subsequent rise in the SE yield due to fall in the electron range. At high energies most of secondary electrons generated within the interaction volume are far deep inside the surface to escape, and do not contribute to the observed signal. The efficiency with which the SE detector collects the SEs from the materials also contributes to SEM image contrast. At high beam energies more SEs are emitted so the beam penetration contrast, the topographic contrast, and the detector efficiency contrast all contribute to the SEM image, and face tilted towards the detector is bright because these collected most of the SEs. Also the face tilted away from the detector emits more SE than at normal incidence but a smaller fractions are collected by the detector. Here, the topographic contrast is modulated by the detector contrast effects to produce the characteristic shading that makes surface the topography readily interpretable. The features of the sample comparable to or smaller than the beam range are filled by the interaction volume producing enhanced SE emissions that contribute to the three-dimensional appearance. At low voltages, the beam penetration is small, all faces of the sample emit about the same number of SEs. The features raised on the surface no longer display enhanced SE emission, leaving only the detector contrast contributions to measure. This results the SEM image appears flatter, less dramatic, and the true nature of the specimen’s surface is hard to discern.
Crossover Voltage
Electron yield is defined as the total number of electrons exiting a material divided by the number of incident primary electrons. The electron yield varies as a function of accelerating voltage as shown in Figure 3. Electron yield gives a value of unity twice for this plot and these points are termed as crossover voltages. The lower crossover voltage is designated as E1 and the higher crossover voltage is E2. E1 is too small to be measured with existing LVSEM technology. However E2 ranges from 500 eV-3.5 keV for most insulators.
Figure 3. Total electron yield verses primary beam energy showing the E1 and E2 crossover voltages at which the electron yield equals 1.0.
The part of the curve between E1 and E2, defines the range of voltage where positive charging occurs, in which the number of electrons exiting the sample are more than the electrons entering, resulting in a net charge depletion at the surface. Positive charging results as a darkened region in the sample while imaging, as the net positive field locally holds back the electron yield. The part of the curve for voltage values greater than E2, and electron yield less than 1.0 is known as the negatively charged region, in which the number of electrons exiting the sample is less than the electrons entering. Negative charges build-up charges within the uncoated specimen, which behaves as an insulator and displays as bright regions in the image. Negative charging can produce fields large enough to deflect the incident electron beam away from the sample or even back up the microscope column. Thus to prevent charging on the surface of the sample, an acceleration voltage equal to E2 should be used.
5. Tilting effects on surface topography
Figure 4. Tilting and edge effects as a function of material and voltage. (a) At high voltage, both the metal and polymers have large secondary electron emission when the sample is tilted and at edges. (b) At low voltage, the contrast in metal specimen decreases because the electron range is approximately the same as the secondary electron escape depth. (c) At low voltage, contrast in polymer samples improves because the electron range is small, but still larger than the secondary electron.
Topographic contrast arises when there are,
(i) Local differences in the secondary electron coefficients (local variations in crossover points)
(ii) Local differences in the efficiency at which the secondary electrons are collected
Local differences in the secondary electron coefficient arise when there is an edge or a local tilting of the sample’s surface with respect to the primary beam direction, as it happens in surface roughness. At an edge, the secondary electrons can escape from the top surface as well as the side surface (Figure 4a), as a greater percentage of the interaction volume is within the secondary electron escape depth. Tilting also increases the secondary electron yield by moving the interaction volume closer to the surface. The secondary electron coefficient is increased with tilt, as shown by the equation:
δ(θ) = δ0 secθ
where δ, is the secondary electron coefficient at a normal angle of incidence and θ is the tilt angle. Tilting is therefore an effective way of reducing charging in polymer samples examined by shifting the crossover to higher voltages i.e. at a voltage greater than E2.
In metals at low accelerating voltage, the electron range is approximately the same distance as the secondary electron escape depth. For example, in gold at 1 keV the electron range and the secondary electron escape depth are approximately 50 Å. The result is a loss of edge contrast because the electron range is within the secondary electron escape depth at every place on the sample. Therefore tilting neither changes the secondary electron coefficient nor the image contrast. A schematic of this effect is shown in Figure 4b.
In polymers, however, topographic contrast is present and even enhanced at low keV operation. Even at low accelerating voltages the interaction volume in a polymer is always much larger than the 200 Å secondary electron escape depth. Therefore, topographic contrast has the same origins as at high voltage. At low voltages the topographic contrast is enhanced because the total interaction volume is smaller and closer to the surface. A schematic of this effect is shown in Figure 4c.
6. Limitations in LVSEM
6.1. Brightness of thermionic sources at low voltage
The source brightness is the practical limit on the performance of SEM with a heated tungsten. The effect of spherical or chromatic aberration on spot size can be minimized by reducing the acceptance angle of the final lens (a) until the lens becomes diffraction limited. In instruments with conventional tungsten sources, the image becomes too noisy for convenient use and it is reduced to the diffraction limit. The brightness (B) of a thermionic electron emission source is given by the Langmuir equation for small a,
where, j0 is the current density at the surface of the source in amps/cm2, T is the source temperature in K, and Vo is the beam voltage in volts. From equation (1), the low Vo operation produces reduced brightness, but in practice the brightness actually obtained is even lower than the value estimated from Eqn, (1) because this equation is only valid in absence of space charges near the cathode surface. Such space charge shields the cathode from the accelerating field and further reduces B.
For a specific gun geometry, the virtual free of space charge effects near its highest operating voltage (20-30 kV), the field present at the filament surface is proportionately less at 1 kV and the gun brightness is limited by the space charge unless the geometry is changed. The practical measures to improve thermionic gun brightness at low voltages (kV) therefore includes: (i) use of LaB6 cathodes and (i) changing the gun geometry.
LaB6 cathodes have a comparably more pointed tip that reduces the space charge effect and are about 6-10 times brighter than the tungsten source for comparable lifetime. . They also operate at a lower temperatures than normal tungsten (T = 1500 K to 2500 K). Changing the gun geometry involves, either reducing the gap between the Wehnelt and the anode using an anode spacer, or a mechanism to actually move the anode towards the cathode, or addition of more anodes to the gun. A “double-anode approach” and installed extraction anodes of various shapes and spacings between Wehnelt and normal anode are generally used. At low beam voltage (Vo), this electrode runs a few kV above the ground to produce higher constant voltage V1 and reduces the effects of space charge. At 1 kV and 2 kV, this set up produces ten times higher brightness than the typical 30 kV electron gun with tungsten and eight times higher brightness with LaB6 using the geometry depicted in Figure 5.
Figure 5. Double anode system for improved brightness from thermionic cathodes at low voltage.
6.2. Brightness of field-emission sources at low voltage
Field-emission (FE) sources have been known to produce high brightness, but are commercially limited due to the following reasons: (i) requirement of 10-8 Pa (10-10 Torr) vacuum in the range around the emitter tip, and (ii) the current generated by the source is subjected to temporal instability that produces streaky images. Most FE sources utilize a double anode design with V1 in the range 3-7 kV and is used to adjust the beam current. A second supply between the ground and the cathode adjusts the primary beam voltage (Vo) to the desired level and reduces it for LVSEM, producing an electrostatic lens. In principle, the geometric parameters can be adjusted for an efficient emission from the tip with V1= 1 kV. However, in practice, tips with sufficiently small tip radius (r0) usually unstable and subject to catastrophic failure while a suitable choice of the spacing and shape of the two anodes can reduce the lens effect to a low level. Here, the brightness of the source depends only on V1 until the lens effect degrades the source image, and is given by:
where, Cc is the chromatic aberration coefficient and ΔV is the energy spread of the beam. Here, the issue arises as ΔV/Vo increases rapidly at low beam voltage Vo that can be resolved by lowering Cc, ?or ΔV. Lenses can be designed to reduce Cc, by shortening their focal length. It is practically difficult to design a system for Cc, less than 0.2 mm. However, it is relatively easy, in terms of the total magnetic flux required to construct a lens of short focal length at low energy.
6.4. Issues with signal collection in LVSEM
In most SEMs, the secondary electrons produced by collision between the beam electrons and the sample are collected by a field imposed at about 300 V to attract electrons at the entrance of the scintillator/photomultiplier signal amplifier. The collection of SEs using this procedure works well when the sample is 5-10 mm below the objective lens pole-piece. As the working distance is reduced to diminish chromatic aberration (Cc), the same field at the sample surface becomes less efficient for collecting signal electrons as this field becomes relatively large compared to the beam energy, producing disturbances from the beam. To overcome this drawback following measures are used:
- The first method is using an objective lens with a sharp conical lower pole-piece, which allows the collected field to penetrate the electron optical axis easily. This approach also permits observation of highly-tilted large and flat samples. It has the disadvantage that conical lenses often have reduced electron optical properties.
- The second method is to collect the signals with the use of systems employed in the TEM/SEM, where initially the sample is immersed in the lens field with the low energy electrons spiral up the field lines through the hole in the upper pole piece and are finally collected by a small transverse electric field. The advantages of this system include: (i) it could work with very short focal length lenses, (ii) the area of transverse field can be carefully controlled and protected from inhomogeneities produced by irregularities in specimens’ topography. (iii) It is immune to the variations in collection efficiency caused differences in specimen’s surface potential. Also, there exists few disadvantages of this method: (i) magnetic samples can’t be viewed due to lack of directional collection field at the sample, and (ii) the shadowing effect is absent.
- The third method involves using a pair of scintillator/photomultiplier detectors, one on either side of the sample specimen, which helps shielding the axis from the field. An axial metal tube protects the beam from the effects of the collection field until the time it doesn’t reaches the sample.
Summary
- In LVSEM, the image contrast results due to the changes in the secondary electron coefficient (δ) from different elements, and this effect is largest when the SE yield is maximum, i.e. at low beam energies.
- At low energies, charging effects diminishes and the trapping of charges reduces, and results protecting the sample from any structural disorder.
- To improve the brightness at low energies, different gun geometry and use of LaB6 cathodes are the primary steps taken.
- Using pair of scintillator/photomultiplier detectors and objective lens with a sharp conical lower pole-piece are few methods taken into account to overcome the drawback of signal collection in LVSEM.
you can view video on Low Voltage Surface Electron Microscopy |
References
1. E. Bauer, Proceedings of the 5th International Congress on Electron Microscopy, S.S. Bresse, jr. ed., AcademicPress, New York, 1962, Vol. 1, pp. D11-D12.
2. David C. Joy., Low Voltage Scanning Electron Microscopy, Hitachi Instrument News, July 1989.
3. Drummy, Lawrence, F.; Yang, Junyan; Martin, David C. (2004). “Low-voltage electron microscopy of polymer andorganic molecular thin films” Ultramicroscopy99 (4).
4. David C. Joyand Carolyn S. Joy, Micron,LowVoltage Scanning Electron Microscopy,Vol.27, No. 3-4, pp. 247-263,1996.
5. Ichinokawa, T., Sekine, M., Gur, Z.S. & Ishikawa, A. (1982) Electron optical properties of low energyfield emission gun in the energy range from 100 to 2000 eV. Proc. Europ. Reg. Conf. Electron Microscopy, I, 351-352.
6. D. C.Joy,Monte Carlo Modelingfor Electron Microscopy andMicroanalysis, Oxford University Press, New York,1995a.