Our patented development of a pulsed RF source (U.S. Patent number 10 52883, 2010), with the speed and capability of matching impedance changes in pulsed mode, represented a major step forward for many applications.
Pulsed operation allows minimizing the average power (and so the thermal constraints) while keeping high instantaneous power for sensitivity. Deep craters in glasses (down to 150 microns) have been obtained, PV cells on glass are readily measured: pulsed operation assures that no Na diffusion will happen during the measurement.
Pulsed operation is also beneficial for the precise control of the crater shape, offering enhanced depth resolution.
The following example illustrates the excellent depth resolution offered by the pulsed RF operation with OES detection. The sample is a mirror for X-ray and features 60 stacks each with Si/B4C/Mo multilayers. Each lstack is 7 nm thick. Details on these results and experiments had been shown at the 6th GD day.
Magnetic fields are used in plasma deposition (magnetron sputtering) and are known to enhance the signals, but were seldom tried in GD as people were simply transposing the magnetron configurations and obtaining non flat crater shapes.
HORIBA Scientific has patented an RF coupler that takes advantage of advanced knowledge on magnetic fields to provide enhanced signals and flat crater shapes. This has been shown to be beneficial to measure notably thick glasses and ceramics (over 5 mm thick). Presentations were shown at the 6th and 7th GD days.
The analytical RF GD plasma is dense - about 1014 (charged particles/cm3) but incident ions bombarding the surface have a low energy – about 50 eV, and due to multiple collisions, they are not unidirectional. These are the reasons for fast sputtering as well as for low surface damage compared to higher energy sputtering ion beams.
In average operating conditions, metals are sputtered at a rate of 1-5 μm/minute. A 100 nm layer could be sputtered in 3s – 10/15 s in pulsed mode. A thermal treatment on steel in which elements diffuse down to 50 μm could be checked in 12 min.
Such fast sputtering requires a fast detection system, and a very fast detection system for thin films analysis, to be able to adequately follow the varying signals.
The vacuum pumps are needed for the source only. A GD plasma operates at low pressure. Pumps establish the vacuum level, and a small flow of gas is then introduced – the flow is about 0.1 l/mn during analysis – so a gas cylinder will last for several months.
We use 2 independent vacuum pumps in our source to maintain a uniform pressure control using the GD source as sample preparation for SEM.
The double pumping assures that deep craters down to 100-150 μm (sometimes deeper – current record is over 500 microns in a special application) can be obtained by evacuating away the sputtered particles.
Double pumping using two pumps assures adequate control of the uniformity of the pressure at the surface of the sample during the depth profile. It is crucial for depth resolution of ultra thin layers. The combination of the Pulsed RF source and double pumping with 2 pumps that achieves the nanometer depth resolution shown above.
Differential pumping with 2 pumps is also mandatory for the use of RF GD for FESEM as shown by Prof. Ken Shimizu in “New Horizons of Applied Scanning Electron Microscope.”
In standard operating conditions, the plasma covers the entire surface of the sample facing the anode and it is usually relatively straightforward to find operating conditions leading to a flat crater required for achieving high depth resolution.
However, if the conditions are not optimized, the crater shape can be concave or convex, and the depth resolution degraded.
The optical spectrometer(s) collects and analyzes the light from the plasma. The role of the spectrometer is to continuously monitor compositional changes in the sample through the analysis of the light produced by the plasma. As sputtering occurs, the sputtered species enter the plasma where they are excited by collisions and emit photons. If the sample is homogeneous in depth, the spectrum is constant; if it is a multilayer, variations of emitted light eflect the changes in sample composition through its depth.
The optical spectrometer must therefore be ultra fast and offer simultaneous measurement with high dynamic range.
HORIBA Scientific is the world leader for gratings. Gratings, optical configurations and detector operation are described in length in academic books and HORIBA Scientific guides. We will only indicate here the specificities of optics related to GD operation, mainly the need to collect the maximum of light, the need to simultaneously cover a large spectral range, and the need for fast detection in the case of depth profile analysis.
Gratings are of utmost importance and should provide optimum light efficiency. The selected ones for GD are proprietary HORIBA Scientific ratings with enhanced efficiency in the VUV range.
The spectral range to cover is large, as it goes from H at 121 nm to K at 766 nm, covering VUV, visible and near IR ranges.
High resolution polychromators with Paschen Runge mountings are used. This is understandable as depth profile is central for these instruments. The polychromators are truly simultaneous systems and allow following all elements of interest as a function of depth.
The need to cover a large spectral range from VUV to IR while keeping a good resolution requires a slightly more complex optical configuration combining several gratings, each one dedicated to a special part of the spectrum. For instance, in the specific design of the GD Profiler 2, HORIBA Scientific uses a double order grating for the VUV and UV, and a second grating for the near IR region.
Finally VUV light transmission requires the use of dedicated optics (MgF2) and the need of transparent atmosphere (by purging optics with nitrogen).
The polychromator is central in GD, as high dynamic simultaneous measurement is mandatory for depth profiling.
Line selection in the polychromator is totally flexible. We use a thin metallic mask with more than 200 slits pre-engraved (offering multiple selection for all elements). At the start only the lines of interest are equipped with detectors while others are covered, but this cover can be removed and extra detectors added on site if demand arises.
Being flexible, they could be tuned to any line, and also offer the possibility to measure extra elements with high resolution (5-10 pm) and sensitivity. They could be equipped with various gratings. High Dynamic Detection is of course available on the monochromator. In the Image mode, they record the full emission spectrum (from 180 nm to the max range of the selected grating) of a bulk sample or of a layer.
“HDD” means High Dynamic Detection. HDD can measure high concentrations and low concentrations. This is a major invention for GD operation that has been awarded the U.S. Patent 5,726,438 in 1998.
In a depth profile analysis, an element to be measured could be at the ppm level in one layer and 100% in the next.
It is impossible, of course, to stop the measurement between layers and adjust the gain of the detectors due to the very rapid sputtering process employed, so preset values must be used posing serious limitations to the technique. If the light signal is low, a high value is preset. If the signal is high, a lower value will be preferred. But how does one manage when the concentration changes from layer to layer?
HDD is the answer to this challenge. A proprietary design adjusts in real time the applied voltage to the output current of the detector, offering a true dynamic range of over 109 on all lines.
The analytical benefits are below:
The monochromator offers the flexibility to measure any extra element with high resolution and high dynamic range. Addition of lines on site is also ossible if a new element becomes of regular interest.
The image combines the benefits of a high resolution monochromator, ultra fast scanning, high dynamic detectors and proprietary software to offer the record of the full emission spectrum of a bulk sample or a thick layer (over 1 μm). The outstanding precision of the optical system assures that line positions are perfectly in agreement with the theoretical values in the DB (at picometer level!) assuring unambiguous detection of the presence in the material and giving concentration information by comparison to references.
Since GD is a non-thermal plasma, the GD spectrum is less rich in lines than the ICP or Spark ones. In GD atomic lines (lines I) are usually more intense than ionic lines (II). The line selection is therefore usually straightforward and most of the time the measurement of a single line per element is enough in GD, provided the spectrometer resolution is sufficient. Backgrounds are also insignificant compared to ICP. Background correction is therefore of minimal use in GD, even for bulk analysis.
The useful spectral range in analytical GD is very large, typically from 120 nm to 766 nm and so covers VUV, UV, visible and near-IR ranges.
The gas elements which are crucial for depth profile characterization have their most sensitive lines in the VUV range, H (121 nm), O (130 nm), Cl (134 nm), N (149 nm), C (156 nm) when the alkali emit in the red region, Li (670 nm), (K 766 nm).
The spectrometer should therefore easily cope with these stringent requirements.
Measurement in the UV range requires the optics to be transparent at these low wavelengths. This can be done either by evacuating the polychromator (with a vacuum pump) or filling it with a neutral gas N2. Purging with N2 is superior to vacuum as it insures that the optical surfaces will not degrade with time.
Water cooling is useful notably when the source is operated in the non-pulsed mode. Cooling of the sample minimizes the heat generated by the Ar sputtering. Sample temperature therefore will be kept low, avoiding possible melting (when Sn, In or Zn layers are measured, for instance).