5. Spectral Enhancement
The issue of not having full control of sensor gating took me in a new direction. I began looking for ways to amplify and extend the atomic spectra to get a bigger signal to the PDA to compensate for the slow PCM-401. A down side to this is I imagine any improvement in atomic signature intensity would also be preceded by an equivalent increase in the intensity of Bremsstrahlung radiation. However LIBs in itself is not an exact science and accuracy of measurements varies wildly depending upon target composition [L47,L48,L49] as well as analytical methodology, and a lot of work has been done to investigate ways of mitigating these issues.
I researched LIBS papers in depth looking for all experimental work in this area, and found several mechanisms had been investigated. I decided to expand the project to research and include all practical forms of enhancement. A large driving factor behind this was I could put off experimentation that might further endanger the MCP or PDA. As a backup plan, I also researched in detail the feasibility of mechanically shuttering laser and Bremsstrahlung radiation from the MCP.
Additionally, I set about determining how to upgrade my lab to measure and confirm the enhancements.
From this point onward the project was elevated from the initial goal of building a simple LIBS system, to a huge experiment encompassing a broad learning process covering many topics.
I identified a number of potential refinements:
1. MULTIPLE LASER PULSES: 2 TO 7 PULSES MAX
Many papers indicate LIBS plasma signal strength can be increased up to an order of magnitude by
increasing the number of laser pulses. However the improvement is element-fixed and not across the
range. Detection of some elements is adversely affected by increased pulses but generally there is
an advantage. Papers clarify which elements are advantaged and which are not. With the MK367
higher voltage produces more laser pulses, but laser life is considerably reduced. A multi-pulse
system necessitates inclusion of a 'simmer' power supply in the laser drive electronics to increase
flashlamp life. The project will initially run with single pulses but will include a simmer to
accommodate multiple pulses.
2. RESONANT EXCITATION DUAL PULSE SYSTEM
The following paper mentions a number of these enhancements and describes in detail, a specific
variant of a dual pulse system using two separate lasers arranged orthogonally, one 532nm to ablate
the target, the other 1064nm to reheat the target. The reheating phase leads to enhanced
intensities and is particularly useful for use at the lower laser powers required for organic
targets: [L4], Double-Pulse Laser Induced Breakdown Spectroscopy in Orthogonal Beam Geometry to
Enhance Line Emission Intensity from Agricultural Samples.
3. MAGNETIC FIELD
The following 2011 paper reported a 50% improvement in spectral signature by encompassing the
target in a high 2.25T magnetic field: [L15], Study of the Feasibility of Applying LIBS for In-Situ
Characterization of Deposited Layers in Fusion Devices. Creating a magnetic field of this
intensity is difficult and this enhancement has been put on a back-burner until a feasible
mechanism can be obtained.
4. ELECTRIC FIELD, HV AC
I could find no papers that had subjected the plasma to a high voltage alternating electrical
field. Novelty plasma balls use this method to generate coronal streamers, so there is a
possibility it could affect laser plasma too. Flooding the target chamber with an inert gas means
the target chamber can be made electrically conductive, aiding this. I decided to explore this
as a personal research experiment.
5. UV LASER
Most LIBS systems initially used 1064nm Nd:YAG lasers, but research papers discovered other
wavelengths offered advantages, namely 10.6µm (CO2) and UV. UV is particularly useful for
transparent material that is difficult to ablate with IR. UV also has the advantage of being more
efficient than IR, so less energy is needed for the same effect. CO2 lasers often come on the
market at low prices but these are generally large and high power. The same is true of UV lasers
which are typically TEA (Transversely Excited Atmospheric), air, Nitrogen, or Excimer gas lasers.
Of these the TEA is the easiest to build, see http://en.wikipedia.org/wiki/TEA_laser.
One publication advocated 213nm, but at the same time hinted the shorter pulse width might be the source of advantage: [L14], Laser Ablation Solid Sampling for Plasma Spectrochemistry - The Importance of Matching the Hardware to the Application : 'There is additional evidence that laser irradiance (fluence/pulse width), measured in GW/cm², may play a significant role in this process. This might explain why a 213nm Nd:YAG laser operating at a maximum energy density of 25J/cm² with a pulse width of 3ns compares favorably with a 193nm excimer laser operating at a maximum energy density of 45J/cm² with a pulse width of 20ns. The increased irradiance inherent in a short-pulse width laser 8.33GW/cm² for the 213nm wavelength versus 2.25GW/cm² for the 193nm wavelength may play a role in improving the analytical capabilities of laser solid sampling.'
Another paper advocated 193nm, the preserve of the Argon Fluoride Excimer laser. as the ideal wavelength because it produces the lowest intrinsic elemental fractionation.
In Google Books, [L18], Practical Guide to ICP-MS: A Tutorial for Beginners, 'There is evidence to suggest shorter wavelength lasers exhibit better elemental fractionation characteristics (typically defined as the intensity of certain elements varying with time, relative to the dry aerosol volume) than longer wavelengths because they produce smaller particles that are easier to volatize.' Subsequent papers also argued in favour of 193nm.
Excimer lasers are expensive, and the older ones prone to failure due to build-up of impurities in the reaction chamber over time, necessitating frequent replacement and mixing of the gases in precise quantities at precise pressures. Fluoride is also hazardous to health and expensive, and like TEA, the laser requires very high voltages to work. All of these made excimer an impractical choice.
A 2003 paper [O12], Evaluation and Design of a Solid-State 193nm OPO-Nd:YAG Laser Ablation System, revealed it was possible to produce 193nm laser pulses from a conventional Nd:YAG with power to match an excimer gas laser. By feeding the Nd:YAG 1064nm fundamental wavelength through a series of crystals which magically generated harmonic (HG) wavelength of the 1064nm: 512nm, 266nm, 213nm, and by further trickery using an oscillator crystal and a summing crystal, 193nm could be produced. These crystals are referred to as Non Linear Optics (NLOs). The problem with this approach is whilst the harmonic crystals pop up quite often on eBay, the remaining crystals are very specialised and extremely unlikely to appear.
A disadvantage of these HG NLOs, is the output of each stage has a marked reduction in output power, in
my case with older used crystals, probably a reduction of 70%, but an advantage is at each stage the
output pulse gets shorter, which increases the overall power of each shot. A significant advantage of this system is all wavelengths are derivatives of the same originating pulse, so synchronisation is assured although different wavelengths focus at different distances, but the difference is so small as to be of little impact for my application.
6. PLASMA GAS IMMERSION
Immersing the target in gas can increase the plasma signal strength and duration whilst
simultaneously removing the spectral signatures of air and all of its impurities. Helium makes it
worse but Argon improves it due its high atomic weight making it denser than air, restricting the
spark from irradiating outward and helping to retain the high temperature of the original reaction
for longer. [L28], Effect of Atmospheric Conditions on LIBS Spectra, summarises the
conclusions of various papers, including [L28], Effects of Atmosphere on Laser Vaporization and
Excitation Processes of Solid Samples, and [L3], Laser-Induced Breakdown Spectroscopy,
Page 71, 4.1.3 Ambient: 'The laser-induced plasma size, propagation speed, stability, energy and
emission properties depend strongly on the gas ambient into which the plasma expands. Plasma
properties have been studied in different gasses [81,81-89], as a function of pressure [85,90],
even in liquids [91-93].'; '...a shorter plasma lifetime with lower temperatures was found with air
than argon [94]. This observation was explained by lower conductivity and specific heat of argon
with respect to air.','Wisburn et al. [95] found that the collisional translation energy also was
dependent on the atomic mass of the ambient gas, being less effective when the atomic mass
increased, thereby causing a plasma with a longer lifetime.' References are as follows:
[81] A.Bogaerts, Z.Chen, J. Anal. Atomic Spectrometry 19 (2004) 1169
[82] G.Dumitru, V.Romano, H.Weber. Appl. Phys. A79 (2004) 1225
[83] F.Colao, R.Fantoni, V.Lazic, A.Paolini, Appl. Phys. A79 (2004) 143
[84] Z.Arp, D.Cremers, R.Harris, D.Oschwald, G.Parker, D.Wayne, Spectrochim Acta B59 (2004) 987
[85] Y.Iida, Spectrochim. Acta B45 (1990) 1353 [L28]
[86] S.Harilal, C.Bindhu, V.Nampoori, C.Vallabhan, Appl. Phys. Letts. 72 (1998) 167
[87] V.Detalle, M.Sabsabi, L.St-Onge, A.Hamel, R.Heon. Appl. Optics 42 (2003) 5971
[88] S.Amoruso, B.Toftmann, J.Schou. Applied Physics A79 (2004) 1311
[89] R.Russo, X.Mao, M.Caetano, M.Shannon, Appl. Surf. Sci. 96-98 (1996) 144
[90] G.Cristoforetti, S.Legnaioli, V.Palleschi, A.Salvetti, E.Tognoni, Sp'chim Acta B59 (2004) 1907
[91] R.Nyga, W.Neu. Optics Lett. 18 (1993) 747
[92] A.Pichahchy, D.Cremers, and M.Ferris. Spectrochim. Acta B52 (1997) 25
[93] D.Beddows, O.Samek, M.Lizka, H.Telle, Spectrochim. Acta B57 (2002) 1461
[94] D.Kim, K.Yoo, H.Park, K.Oh, D.Kim. Appl. Spectrosc. 51 (1997) 22
[95] R.Wisbrun, I.Schechter, R.Niessner, H.Schroder, K.Kompa, Anal. Chem. 66 (1994) 2964
The project will experiment with target immersion in argon and CO2 gas.
7. USE 10.64µm AS WELL
LIBS systems have also used CO2 10.64µm lasers. TEPS systems also combine
multiple pulses. [L26], Enhancement of Nd:YAG LIBS Emission of a Remote Target Using a Simultaneous
CO2 Laser Pulse in particular, notes a 25x to 300x improvement when used in parallel with a
Nd:YAG. Some papers hinted an advantage at running UV wavelengths with CO2 but the large focus
disparity would likely necessitate a twin lens system, adding to the complexity. Later reading
revealed a minimum power of 100W was needed. The significant expense and poor reputation of CO2
lasers from eBay China lead me to investigate the feasibility of building a DIY high power pulsed
TEA CO2 laser instead (see Lasers section).
8. RESONANCE ENHANCED LIBS: RELIBS / RELIPS
Ionising the encompassing gas increases the plasma signal strength by up to 10%, element-specific.
Papers have investigated various gasses, emphasising the heavier the gas the greater the effect.
Krypton has a heavier weight than Argon so should be better, but an Argon ion laser is easier to
obtain than a Krypton laser. The US Army Research Lab used an unfocused multi-line 50mW Ar laser to
ionise Ar gas around the target with limited but perceivable success: [L16], Tailored Ultra-Fast
Pulses for Selective Energetic Residue Sampling.
Having learned about CO2 enhancement long after this and given CO2 is heavier than Argon, what
about CO2 with a CO2 laser as well as Nd:YAG? See [References: Optical Materials, Gas Densities].
Page 18 of [L25] shows photographic evidence of the effects of argon and air on the target - are
there similar photographs for CO2?
9. PRESSURISED PLASMA TARGET CHAMBER
Pressurising the plasma chamber can improve plasma signal strength by up to a factor of 5 and
optimally it needs to be kept at 2psi. Lower or higher has the opposite effect. Therefore a 0-6psi
dual pressure regulator was bought together with a cheap Chinese USB-based pressure sensor.
[L3], Laser-Induced Breakdown Spectroscopy, Page 72, 'Fig 15: Dependence of the emission
intensities of Fe I 374.949nm on the pressure of the ambient gases...', shows argon emission
intensities peaking at 100 torr, way above air and helium. This figure was taken from
[L28], Effects of Atmosphere on Laser Vaporization and Excitation Processes of Solid Samples.
More applicable papers are:
[L12], Resonance-Enhanced Laser-Induced Plasma Spectroscopy: Ambient Gas Effects, and
[L13], Effect of Atmospheric Conditions on LIBS Spectra.
10. REHEAT THE PLASMA USING A HV DC SPARK GAP
[L13], Single-Shot Spark Assisted Laser-Induced Breakdown Spectroscopy,
'In spark assisted LIBS (SA-LIBS), the electrical discharge improves the relative signal intensity
ratio by reheating the plasma and increasing its lifetime [46]'. Uses a 12kV dc spark gap.
[46]: Comparison of two LIBS techniques for total carbon measurement in soils, Spectrochim.
Acta B 64, 899–904 (2009).'M.Belkov, V.Burakov, A.De Giacomo, V.Kiris, S.Raikov & N.Tarasenko.
11. USE MULTIPLE WAVELENGTHS
I have seen papers describing the removal of residual laser wavelengths by harmonic mirrors or a
pellin broca prism, but I have not yet seen any paper explaining why residual laser wavelengths
need to be removed. Given resonance LIBS utilising a multi-line argon laser has been used to
ionise argon gas to enhance spectra: [L16], Tailored Ultra-Fast Pulses for Selective Energetic
Residue Sampling, it occurred to me if I attempted to generate UV from the fundamental 1064nm
Nd:YAG using harmonic crystals, given different wavelengths offer advantages to different
materials, there might be an advantage to leaving all sub-harmonic wavelengths present, even if
only one wavelength could be sharply focused.
12. ALWAYS FIRE AT A NEW SPOT
Papers describe hitting the same spot on the target reduces spectral intensity, the solution
usually being to rotate the target after each laser pulse. I will therefore place the target on a
stepper-motor based rotary stage. A stepper motor will simplify the rotation and will only need a
simple sensor to detect when one rotation has been completed.
13. USE A TOP HAT BEAM PROFILE
This involves converting the typical Nd:YAG conventional Gaussian (mountain top) beam profile to
one with a flat top using a special beam shaping lens. [L1], Laser Induced Breakdown Spectroscopy -
Fundamentals & Applications, P.343: 'The ideal laser beam would have a “top-hat” profile and all of
the intensity of each succeeding laser shot would impinge on fresh surface at a new depth.'
[L20], Capabilities of a Homogenized 266nm NdYAG Laser Ablation System for LA-ICP-MS, describes
the ideal ablation beam as one with uniform intensity across the beam instead of gaussian with its
peak intensity at the centre. [L20], Laser Ablation Inductively Coupled Plasma Mass Spectrometry,
goes on to indicate a flat top beam profile reduces LIBS fractionation. I did eventually find some
tophat profiling lenses at a good price from a generous seller on eBay. The lens requires a good
quality Gaussian profile to convert to tophat, and I will use LaseView software to help achieve
this and verify the result. Sadly it's no longer free (although they do offer a free one month
trial of their latest version). I got a free copy of Version 3 before it disappeared [S3].
Of these enhancements, the most rewarding but also the most complex to incorporate was 5, the UV laser variant and in particular, the seminal paper [O12] describing how to generate 193nm from a 1064nm fundamental wavelength.
I knew I stood no chance of obtaining the exotic crystals to generate 193nm, but I should be able obtain those necessary to produce 1064nm harmonics down to 266nm, and possibly even 213nm.
Paper [L19], Laser Ablation Solid Sampling for Plasma Spectrochemistry, describes 213nm as being almost as good as 193nm.
Late 2020 I discovered the following short article that summaries many of these enhancements:
[L44] A brief history of laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS)
In order to traverse several crystals in sequence, the Nd:YAG would have to be much more powerful than the MK367, and this time it would be an actively Q-switched laser.
I felt I had come full circle: I would need to acquire another Nd:YAG.