11 Beam Profiling | discovery

11. Beam Profiling

This section covers research into producing the ideal beam shape for LIBS.

Beam shape and quality affects LIBS. The following papers explore ways of optimising this.

[L18], Practical Guide to ICP-MS, says 213nm is the best Nd:YAG wavelenth, but ArF is better still due to its homogenised beam.

P.184: 'The weaknesses in 266nm technology eventually led to the development of 213nm lasers (13) because of the recognized superiority of shorter wavelengths to exhibit a higher degree of absorbance in transparent materials (14).'

P.185, Fig.17-1: 'Craters produced with the 213nm laser system are relatively round and symmetrical, whereas craters produced using the 266nm are more irregular and show excess ablated material around the sides of the craters. (Courtesy of New Wave Research.)

It can be seen that the craters produced with the 213nm laser are relatively round and symmetrical, whereas the 266nm craters are irregular and show ablated material around the sides of the craters. This significant difference in crater geometry between the two systems is translated into a difference in the rate of depth penetration, size distribution, and volume of particles reaching the plasma.

With the 266nm laser system, a high volume burst of material is initially observed producing a spike in the signal whereas with the 213nm laser, the signal gradually increases and levels off quickly, indicating a more consistent stream of small particles being delivered to the plasma. The benefits of 213nm lasers emphasize that matrix independence, high spatial resolution, and the ability to couple with UV transparent materials without fracturing. ...the less coherent nature of the [193nm ArF] excimer beam enables better optical homogenization resulting in an even flatter beam profile. The overall benefit is that cleaner, flatter craters are produced down to approximately 3-4µm in diameter. This provides far better control of the ablation process...'

'Figure 17.2: '...shows scanning electron microscope (SEM) images of a sample of glass ablated with a 213nm Nd:YAG laser on the left (A) and a crater ablated with a flat beam 193nm ArF excimer laser using an internally homogenized [i.e. uniform] beam delivery system on the right (B). It can be seen that the excimer laser produces a much flatter and smoother crater than the Nd:YAG laser system (15).'

In fact, P.186 Fig.17-3 suggests 213nm is actually superior to 193nm. Given the above table, it occurs to me if a tophat beam shaper was applied to the 213nm Gaussian Nd:YAG beam, it might better even ArF.

By luck I managed to acquire some 266nm and uncoated tophat beam shapers to experiment with, but then came across the following thesis [L14], LIBS & Applications Toward Thin Film Analysis:

P.33 'Chapter 5, Depth profiling with LIBS. Most solid state lasers in use have a Gaussian beam profile which produces concave ablation craters...In an effort to homogenize the beams, a set of telescopic non-Gaussian optics were used in conjunction with a beam expander. In order to test the effectiveness of this system, a poorly maintained Nd:YAG...1064nm laser was used. The raw beam profile ...has a very poor Gaussian shape with localized regions of high and low intensity. This laser produces very irregular ablation craters. The modified beam profile using commercially available beam shaping optics (pShaper 6.6.1064).

This system required extensive initialization efforts and was very sensitive to use. Each of the two components, beam expander, and beam shaper, require the use of 5-degree of freedom optical mounts. This system made using the device very difficult. Additionally, the system did not offer any beam homogenization as seen in Figure 5.1(b). It was determined through extensive testing using a beam profilometer and actual ablation of various materials that this type of beam shaping optics did not provide a robust solution.'

'Instead, a 10x beam expander was used in conjunction with two irises separated by 1m to select the center region of a Gaussian beam...This expander and iris combination was applied to the output from [a] Nd:YAG laser (New Wave Research Minilase-III) with a high quality beam Gaussian beam profile. This combination provides a uniform nearly flat beam profile with a long focal waist, prefect for depth profiling of multi layered structures.'

This alternative approach is ok if you have a powerful enough laser to allow you to effectively throw away a large proportion of its energy by blocking it in the apertures. Long before this thesis I read [L1], Laser Induced Breakdown Spectroscopy - Fundamentals & Applications; Page 99 says you can make a crude top-hat using just a pinhole, but you lose 95% of the energy. The article does not mention the aperture size.

I also have to ask why they didn't try the original beam shaper optics on the good Nd:YAG!

At least it warns me I must ensure my Gaussian beam is high quality before feeding it through the tophap profiling lens.

BEAM SHAPING - TOPHAT LENSES

http://en.wikipedia.org/wiki/Tophat_beam
'A tophat laser beam has a near-uniform fluence (energy density) within a circular (or other shape) disk. It is typically formed by diffractive optical elements from a Gaussian beam - but the original beam must have a good Gaussian profile for the conversion to occur. Tophat beams are often used in industry, for example for laser drilling of holes in printed circuit boards. They are also used in very high power laser systems, which use chains of optical amplifiers to produce an intense beam. Tophat beams are named for their resemblance to the shape of a top hat.'

[L1], Laser Induced Breakdown Spectroscopy (LIBS) Fundamentals & Applications, P.343:

'Typically, depth resolution studies are qualitative in nature owing to the difficulty of accurate calibration of the procedure. The ideal laser beam would have a tophat profile and all of the intensity of each succeeding laser shot would impinge on a fresh surface at a new depth. However, the profile of a typical laser is, at best, Gaussian and often may have an irregular profile or even hot spots. Under these circumstances, each successive laser pulse will sample the sides of the evolving crater in addition to the bottom of the hole. Some fresh surface will also be sampled as the hole diameter typically increases with the number of pulses. So the observed intensities of the components will be a complex mixture of signals from different depths, which is hard to calibrate. At higher laser power values, some partial melting may also complicate the picture. In spite of these problems with the analysis of elemental composition in the interior of samples, if the thickness of a layer is determined by noting the "breakthrough point" where a new element appears, for instance, the results can be quite accurate and reproducible. Several examples of depth profile analysis are shown in the next section.'

[L35], Laser Beam Profile Influence on LIBS Analytical Capabilities: Single vs. Multi-mode Beam,
Page 2 says: 'Several theoretical and experimental studies were carried out to find out the best beam (Gaussian or flat-top)for high resolution depth profile analysis by LIBS ref11, ref12. Comparison of laser ablation with flat-top and super-Gaussian beam profiles were carried out by Laserna’s group ref13. For laser ablation sampling at ICP - MS 14 a multi-mode beam profile was transformed to the flat-top profile in order to improve laser ablation. Better reproducibility of sampling and decrease of fractionation were achieved in this work. It was explained that flat-top beam profile resulted in more stable ablation, less droplets was formed and better atomization of sample was obtained.'
BUT...
'In most recent work concerning beam profile influence a Gaussian and a spoiled beams were used for laser ablation in resonant enhanced LIBS. 15. It was observed that Gaussian profile give better reproducibility, signal and longer emission time compared to spoiled profile. Better analytical results were also achieved for laser ablation with Gaussian profile than with spoiled profile.'

P.4: 'Laser crater profiles were measured with white light interferometer microscope NewView 6200,

Zygo Corp.)'

[L36], Laser-Induced Breakdown Spectroscopy: Fundamentals and Applications,
Chap 7.1. 'A Gaussian beam profile of a laser source can be formed to a top hat profile by., e.g., slightly focusing the beam on a 600µm pinhole which cuts off the outer part of the Gaussian beam and leads to a near flat top beam. A disadvantage of this method is the loss of 95% of the burst energy. The aperture is then imaged onto the surface of the specimen with a demagnification factor of 10.'

[L40], Capabilities of a Homogenized 266nm Nd:YAG Laser Ablation System for LA-ICP-MS,

Page 1: 'A 266nm Nd:YAG...with homogenizer optics to...a "flat-top" laser profile is described';
Page 7: 'The time-resolved ablation signal structure is significantly improved by using a homogeneous energy density across the beam in comparison to gaussian beam profile...The effect of elemental fractionation, which previously led to biased data for fractionating elements...Pb,Ni,Cu, is reduced.'

I acquired 266nm tophat beam shapers to shape the final DUV beam, but one has no coating and presumably could be used on the fundamental 1064nm prior to the HG NLOs which may be a better overall solution.

Comment from the seller: 'The damage threshold is around 2J/cm², depends also on pulse duration. These were made for 20ns to 40ns lasers. However I have used them down towards 10ns and lower.

Three variants of the Gaussian to round flat top converters are included:

266nm [No AR coatings],

266nm [Thin AR Coatings],
266nm [Thick AR Coatings].

The design is a shaper/compressor/collimator design; Incoming beam is designed for 5.6mm 1/e² and output is 1.7mm dia pseudo collimated. At 266nm the AR coating will help 3% to 4%. I had a variety of coatings to test threshold survivability (Thin and Thick are AR stack design of the coatings).'

What I received:

square 2Jcm² 20-40ns 6.5mm² to 5.6mm dia (no AR)
square 2Jcm² 20-40ns 6.5mm² to 5.6mm dia (266nm thin)
square 2Jcm² 20-40ns 6.5mm² to 5.6mm dia (266nm thick)
square 2Jcm² 20-40ns ?.?mm² to 1.7mm dia (no AR)
square 2Jcm² 20-40ns ?.?mm² to 1.7mm dia (266nm thin)
square 2Jcm² 20-40ns ?.?mm² to 1.7mm dia (266nm thick)

Experimental UV tophat beam profiling lenses (SUSS MicroOptics, Switzerland)

ADD PHOTOS BELOW:

The MK367 is a 3mm beam and the MK580 is 5mm. If I get the CW LD I can use a x8 expander to increase it to 5.6mm and feed the tophat to produce a beam with no hotspots (also see [N15]: NLO hotspots) as a precursor to applying it to the MK580, but I would then need another expander to increase the beam to cover the NLO. Additionally Tophat shapers need a good Gaussian beam to begin with. This is where the beam profiling cameras and software come in handy. I assume I need an aperture to tidy the beam to TEMoo?

[L36], Laser Beam Profile Influence on LIBS: Gaussian vs Multi-mode Beams, slide show, 25 slides:
Although interesting it doesn't say much that isn't obvious: a wide multi-mode laser beam with multiple points produces a far more erratic result than a single Gaussian peak. It does mention tophat at the beginning but does not compare this with Gaussian or multi-mode. Their conclusions, based upon multiple tests using Gaussian and multi-mode lasers:
1. Laser beam profile and beam quality is a crucial parameter for laser ablation, plasma properties and LIBS analysis [so why didn't they look at tophat, which is even better than Gaussian?]
2. Plasma temperature and electron density were higher for the plasma formed by Gaussian beam compared to the plasma formed multi-mode beam while pulse energy was 14 times smaller.
3. Multi-mode laser beam sampling results in poor reproducibility of the signals due to self-induced instability of laser ablation
4. Better beam (Gaussian vs multi-mode) for analysis:
drilling sampling: competition between precision (Gaussian) and sensitivity (multi-mode)
Scanning sampling: Gaussian beam is the ultimate choice [again, what about tophat?]

But slide 18 really says it all:

'Gaussian has good repeatability with precision vs multi-mode unpredictable beam profile and equally unpredictable crater profile leading to self-induced instability'.

[L38], Analysis of Solid Materials by LIBS, Thesis, T.Čtvrtníčková, Page 6: 'This thesis deals with the analysis of solid samples with...(LIBS).';'The optically enhanced LIBS single-pulse instrumental set-up was tested on depth profiling of multi-layered metal sample. The optical restriction (pinhole) placed in the instrumental set-up was proved to eliminate partially the tailing effects and improve thus the depth resolution.'

Page 116, 4.2: 'Optical restriction evaluation for depth profiling of multi-layered metal samples: Improvement in depth profiling capabilities of LIBS for multi-layered samples has been attempted. For this purpose, in a typical single-pulse LIBS experiment, an optical restriction consisting of a pinhole placed between the collecting lens and the entrance slit of the spectrograph has been used. This new optical approach allows observing only the light emission coming from the central region of the plume. The micro-plasma was created on the sample by a pulsed Nd:YAG laser operating at 1064nm with a homogeneous distribution of energy across the beam. Light emitted by the micro-plasma was detected with an ICCD multichannel detector. The effect of pinhole diameter and the delay time influence on depth analysis have been assessed. An ablation range of only a few nanometres per pulse have been achieved. Depth profiles of various metals (Cr, Ni, Cu) from multi-layered samples have been generated by LIBS and depth resolution at different delay times using various pinhole diameters have been calculated and compared. In this work, a new arrangement of LIBS with a special type of diaphragm (pinhole) has been evaluated in order to improve the depth profiling capability of LIBS by reducing the contribution of light emission from the borders of the plume. The depth resolution for multilayer sample have been calculated and compared for different delay times and various pinhole diameters.'

Page 118: 'In this configuration, the image of plasma was created at approximately 350mm after the focusing lens and the magnification of the plasma image was approximately 2.5mm. With this arrangement and the pinhole diameter smaller than the image of the plasma, only the central region of plume emission was detected by an intensified charge-coupled device (ICCD, Stanford Computer Optics, model 4Quik 05) with 768 x 512 pixels, each 7.8 x 8.7µm². This configuration provides a spectral coverage of ~30nm with a spectral resolution of 0.04nm/pixel using the entrance slit width of 40µm and the grating of 1200 grooves/mm. Operation of the detector was controlled by a personal computer with 4Spec software [139]. For data acquisition, the ICCD system was triggered by the output TTL pulse from the laser Q-switch system.'

Page 121, Fig 4.11: 'As expected, the size of the plasma images on the CCD decreases when the optical restriction diameter diminishes, with plasma areas from 0.49 to 0.022mm² for pinhole diameters from 0.5 to 5mm, respectively. It should be noted that pinhole diameter larger than 3.5mm does not produce any significant decrease of the plasma image size.'

Page 121, 4.2.4, Depth profiles: 'For the sake of comparison, depth profiles of a multi-layered sample for each pinhole diameter were carried out.'

Page 122: 'These results confirm that the contribution of emission light provided from the borders of the plume is reduced by using this optical device. Furthermore, from this figure it is noticeable that with pinhole diameter larger than 3.5mm no evidence of the expected effect (reduction of tailing effect) was observed, which is in agreement with the plasma images studied (Fig. 4.10). Although the best result was obtained with the smallest pinhole diameter (0.5mm), a compromise between micro-plasma signal coming to the spectrograph and restricting effect should be found.'

Page 126, 4.2.6, Conclusion:
'In this work, the capability of laser induced plasmas with an optical restriction for improvement of depth profiling analysis of multi-layered samples has been demonstrated. The use of a pinhole placed between the dichroic mirror and the collection lenses has allowed us to reduce the contribution of emission light from the border of the plasma resulting from the crater edges, which produces the "tailing effect". The optimization of experimental conditions including suitable pinhole inner diameter selection was discussed, and the delay time was studied. The results demonstrate that the optimal pinhole inner diameter is of 1.5mm. On the other hand, working at short delay times seems to be the best choice in terms of depth resolution due to the confining of the plasma close to the sample surface and non-mixing of specimens.'

Page 128: 'The optically enhanced LIBS single-pulse instrumental set-up was tested on depth profiling of multi-layered metal sample. The optical restriction (pinhole) of 1.5mm of inner diameter placed in front of the collection lens was proved to eliminate partially the tailing effects and improve thus the depth resolution. The ablation rate in order of nanoseconds resulted in depth resolution of approximately 1µm. It would be convenient to couple the effect of pinhole and of the double-pulse configuration, while both approaches demonstrated the improvement of the depth resolution, to obtain even better results. Unfortunately, the connection of double-pulse with the restriction effect of pinhole can not lead to more improvement in presented instrumentation. In orthogonal double pulse method the re-heating of the micro-plasma leads to spatial mixing of particles in the micro-plasma. The application of pinhole would not have any sense. It can be generally concluded that LIBS fulfils the requirements demanded for an analytical technique.'

Paper [L39] below seems to be based on the above thesis which is only available in text with no diagrams. However I think these diagrams are all reproduced in the pdf below as the conclusion is almost word for word.:

[L39], Optical Restriction of Plasma Emission Light for Nanometric Sampling Depth and Depth Profiling of Multilayered Metal Samples, Page 1:

'Improvement in depth profiling capabilities of laser-induced breakdown spectrometry (LIBS) for multi-layered samples has been attempted. For this purpose, in a typical LIBS experiment, an optical restriction consisting of a pinhole placed between the dichroic mirror and the collecting lenses has been used. This new optical approach allows observing only the light emission coming from the central region of the plume. The results demonstrate that the optimal pinhole inner diameter is 1.5mm. Averaged ablation rates in the range of nanometres and depth resolution of approximately 1µm have been obtained.'

[L42], Spectroscopic Investigations of Plasma Emission Induced During Laser Material Processing,
Page 29 shows a tophat converter mounted immediately below the focusing lens from a Nd:YAG for LIBS. It also mentions the spectrometer they used apparently didn't have a gate and its integration time only went down to 1ms making it impossible to capture 1kHz pulses.

[L43], Laser Ablation in Analytical Chemistry - A Review,

Page 7: 'Excimer lasers generally have ‘flat-top’ beam profiles. With appropriate imaging optics, both Nd:YAG and excimer lasers can generate flat-bottom craters (figure 2). The shape of the crater walls will influence depth resolution and fractionation may increase with the development of the ablation crater [31]. However, the degree of fractionation is not strictly related to the beam profile; fractionation is not eliminated by having a flat-top beam profile.'

Page 47: 'Fig 2 [missing] White light interferometric microscope images...flat-top laser beam profiles are capable of producing straight-wall craters in a wide range of materials' (4,32).

4. D. Günther, I.Horn, B.Hattendorf, Fres. J. Anal. Chem., 368 (2000) 4.

32. D. Bleiner, A.Plotnikov, C.Vogt, K.Wetzig, and D.Günther, Fres. J. Anal. Chem., 368 (2000) 221.

FRACTIONATION

In general, fractionation is related to the optical properties of the sample, the wavelength of the laser and significantly, the laser irradiance.

d. BEAM SHAPING (TOPHAT) LENSES

http://en.wikipedia.org/wiki/Tophat_beam
'A tophat laser beam has a near-uniform fluence (energy density) within a circular (or other shape) disk. It is typically formed by diffractive optical elements from a Gaussian beam - but the original beam must have a good gaussian profile for the conversion to occur. Tophat beams are often used in industry, for example for laser drilling of holes in printed circuit boards. They are also used in very high power laser systems, which use chains of optical amplifiers to produce an intense beam. Tophat beams are named for their resemblance to the shape of a top hat.'

THIS ALL NEEDS SORTING

The tophat profile has been recommended in LIBS papers an books; page 99 of this one says you can make a crude top-hat using just a pinhole, but you lose 95% of the energy:
'Laser Induced Breakdown Spectroscopy - Fundamentals & applications', [L1].

A paper, 'Laser beam profile influence on LIBS analytical capabilities: single vs. multimode beam', Vasily Lednev, Sergey M. Pershin, Alexey F. Bunkin, Journal of Analytical Atomic Spectrometry, 29/07/10 [L6], page 2 says:
'Several theoretical and experimental studies were carried out to find out the best beam (Gaussian or flat-top)for high resolution depth profile analysis by LIBS ref11, ref12. Comparison of laser ablation with flat-top and super-Gaussian beam profiles were carried out by Laserna’s group ref13. For laser ablation sampling at ICP - MS 14 a multimode beam profile was transformed to the flat-top profile in order to improve laser ablation. Better reproducibility of sampling and decrease of fractionation were achieved in this work. It was explained that flat-top beam profile resulted in more stable ablation, less droplets was formed and better atomization of sample was obtained.'
BUT...
'In most recent work concerning beam profile influence a Gaussian and a spoiled beams were used for laser ablation in resonant enhanced LIBS. 15. It was observed that Gaussian profile give better reproducibility, signal and longer emission time compared to spoiled profile. Better analytical results were also achieved for laser ablation with Gaussian profile than with spoiled profile.'

11 GOT 'Irradiance-dependent depth profiling of layered materials using laser-induced plasma spectrometry', M. P. Mateo, J. M. Vadillo and J. J. Laserna;J.Anal. At. Spectrom., 2001, 16, 1317 - 1321
IRRELEVANT - does not mention beam profile
12 GOT 'Nanometric range depth-resolved analysis of coated-steels using laser-induced breakdown spectrometry with a 308 nm collimated beam', J.M. Vadillo, C.C. Garcia, S. Palanco and J. J. Laserna;
J. Anal. At. Spectrom., 1998, 13, 793 - 797
USELESS - says mere collimation produces a flat rectangular spot. Unlikely. Lost in Spanish translation?
13 'Atomic emission spectroscopy of laser-induced plasmas generated with an annular-shaped laser beam',L. M. Cabalin J. J. Laserna; J. Anal. At. Spectrom., 2004, 19, 445 - 450
USELESS - just says the target's cold inside the ring - obviously!
http://books.google.co.uk/books?id=M6Fo4MEVaGsC&pg=PA99&dq=libs+beam+profile+top+hat&hl=en&sa=X&ved=0ahUKEwiFoeu8xJ_TAhXWFsAKHSzIADkQ6AEIHDAA#v=onepage&q=libs%20beam%20profile%20top%20hat&f=false
Laser-Induced Breakdown Spectroscopy: Fundamentals and Applications By Reinhard Noll
Chap 7.1. 'A Gaussian beam profile of a laser source can be formed to a top hat profile by., e.g., slightly focusing the beam on a 600µm pinhole which cuts off the outer part of the gaussian beam and leads to a near flat top beam. A disadvantage of this method is the loss of 95% of the burst energy. The aperture is then imaged onto the surface of the specimen with a demagnification factor of 10.' BUT this book only talks about top hat as a means of easily measuring crater depth.
Here is probably what I first read, it's also a Google book so I couldn't save it:
http://books.google.co.uk/books?id=PKJWgvQ6wAYC&pg=PA343&dq=libs+beam+profile+top+hat&hl=en&sa=X&ved=0ahUKEwiFoeu8xJ_TAhXWFsAKHSzIADkQ6AEILzAD#v=onepage&q=libs%20beam%20profile%20top%20hat&f=false
P.343: 'Typically, depth resolution studies are qualitative in nature owing to the difficulty of accurate calibration of the procedure. The ideal laser beam would have a tophat profile and all of the intensity of each succeeding laser shot would impinge on a fresh surface at a new depth. However, the profile of a typical laser is, at best, Gaussian and often may have an irregular profile or even hot spots. Under these circumstances, each successive laser pulse will sample the sides of the evolving crater in addition to the bottom of the hole. Some fresh surface will also be sampled as the hole diameter typically increases with the number of pulses. So the observed intensities of the components will be a complex mixture of signals from different depths, which is hard to calibrate. At higher laser power values, some partial melting may also complicate the picture. In spite of these problems with the analysis of elemental composition in the interior of samples, if the thickness of a layer is determined by noting the "breakthrough point" where a new element appears, for instance, the results can be quite accurate and reproducible. Several examples of depth profile analysis are shown in the next section.'[which is 9.6.4 Quantitative analysis, however there are no more mentions in the book of top hat].

Hot spots are also an issue with NLOs, as described here:

http://www.rp-photonics.com/nonlinear_crystal_materials.html

'Note that even if the nominal intensity is below the nominal damage threshold, there may be problems due to fluctuations of the beam power or local intensity (e.g., if a beam profile has "hot spots"), or due to isolated defects in a crystal, which are more sensitive than the regular crystal material.'

Comment from the seller: 'The damage threshold is around 2J/cm², depends also on pulse duration. These were made for 20ns to 40ns lasers. However, I have used them down towards 10ns and lower.

I have a Gaussian to round flat top: Design 266 nm [No AR coatings], Design 266 nm [Thin AR Coatings], Design 266 nm [Thick AR Coatings]. The design is a shaper/compressor/collimator design; Incoming beam is designed for 5.6 mm 1/e² and output is 1.7 mm dia pseudo collimated. At 266 nm, the AR coating will help, 3% to 4%. I had a variety of coatings to test threshold survivability (Thin and Thick are AR stack design of the coatings).'

The MK367 is a 3mm beam and the MK580 is 5mm. If I get the CW LD I can use a x8 expander to increase it to 5.6mm and feed the tophat to produce a beam with no hotspots as a precursor to applying it to the MK580, but I would then need another expander to increase the beam to cover the NLO. Additionally Tophat shapers need a good gaussian beam to begin with. This is where the beam profiling cameras and software come in handy. I assume I need an aperture to tidy the beam to TEMoo?

Hmmm. Big big experiment!

--------------

Tophats for use in LIBS:

http://arxiv.org/ftp/arxiv/papers/1308/1308.3051.pdf
Laser Beam Profile Influence on LIBS Analytical Capabilities: Single vs. Multimode Beam
Vasily Ledneva, Sergey M. Pershina, Alexey F. Bunkina
P.4: Laser crater profiles were measured with white light interferometer microscope (NewView 6200, Zygo Corp.).
Saved as: 'Laser Beam Profile Influence on LIBS Analytical Capabilities - Single vs. Multimode Beam.pdf'
http://www.myshared.ru/slide/1324038
Laser Beam Profile Influence on LIBS: Gaussian vs Multimode Beams [Russian slide show 25 slides]
Prokhorov General Physics Institute, russian Academy of Sciences, Moscow, Sep 30, 2012, Luxor, Egypt.
Although interesting it doesn't say much that isn't obvious: a wide multimode laser beam with multiple points produces a far more erratic result than a single Gaussian peak. It does mention tophat at the beginning but does not compare this with Gaussian or multimode. Their conclusions, based upon multiple tests using Gaussian and multimode lasers:
1. Laser beam profile and beam quality is a crucial parameter for laser ablation, plasma properties and LIBS analysis [so why didn't they look at tophat, which is even better than Gaussian?]
2. Plasma temperature and electron density were higher for the plasma formed by Gaussian beam compared to the plasma formed multimode beam while pulse energy was 14 times smaller.
3. Multimode laser beam sampling results in poor reproducibility of the signals due to self-induced instability of laser ablation
4. Better beam (Gaussian vs multimode) for analysis:
drilling sampling: competition between precision (Gaussian) and sensitivity (multimode)
Scanning sampling: Gaussian beam is the ultimate choice [again, what about tophat?]
But slide 18 really says it all: Gaussian has good repeatability with precision vs multimode unpredictable beam profile and equally unpredictable crater profile leading to self-induced instability.
http://www.fonon.us/3d-fusion
Unrelated to LIBS but still relevant: this is an advert for a 3D metal printing system:
'The company uses a proprietary multi-kilowatt laser optimized for metal AM. The Flat-Top profile provides a predictable uniform illumination across the powder bed with a scalable beam-spot, no hot-spots or peripheral fall-off. With the use of a multi-kilowatt laser, processing speeds are increased when compared to 400W/500W Gaussian profile lasers used in current generation systems.
Gaussian - over-heated point, under-heated periphery, middle "donut" is ideal heating
flat top - all of beam is ideal heating range and there is no periphery...'
http://is.muni.cz/th/16131/prif_d/Thesis_RZK_221208.pdf
P6 This thesis deals with the analysis of solid samples with laser-induced breakdown spectroscopy (LIBS). 'The optically enhanced LIBS single-pulse instrumental set-up was tested on depth profiling of multilayered metal sample. The optical restriction (pinhole) placed in the instrumental set-up was proved to eliminate partially the tailing effects and improve thus the depth resolution.'
HERE:
P116 4.2. 'Optical restriction evaluation for depth profiling of multilayered metal samples Improvement in depth profiling capabilities of LIBS for multilayered samples has been attempted. For this purpose, in a typical single-pulse LIBS experiment, an optical restriction consisting of a pinhole placed between the collecting lens and the entrance slit of the spectrograph has been used. This new optical approach allows observing only the light emission coming from the central region of the plume. The microplasma was created on the sample by a pulsed Nd:YAG laser operating at 1064nm with a homogeneous distribution of energy across the beam. Light emitted by the microplasma was detected with an ICCD multichannel detector. The effect of pinhole diameter and the delay time influence on depth analysis have been assessed. An ablation range of only a few nanometers per pulse have been achieved. Depth profiles of various metals (Cr, Ni, Cu) from multilayered samples have been generated by LIBS and depth resolution at different delay times using various pinhole diameters have been calculated and compared. In this work, a new arrangement of LIBS with a special type of diaphragm (pinhole) has been evaluated in order to improve the depth profiling capability of LIBS by reducing the contribution of light emission from the borders of the plume. The depth resolution for multilayer sample have been calculated and compared for different delay times and various pinhole diameters.'
P118 'In this configuration, the image of plasma was created at approximately 350mm after the focusing lens and the magnification of the plasma image was approximately 2.5mm. With this arrangement and the pinhole diameter smaller than the image of the plasma, only the central region of plume emission was detected by an intensified charge-coupled device (ICCD, Stanford Computer Optics, model 4Quik 05) with 768 x 512 pixels, each 7.8 x 8.7m². This configuration provides a spectral coverage of ~30nm with a spectral resolution of 0.04nm/pixel using the entrance slit width of 40m and the grating of 1200 grooves/mm. Operation of the detector was controlled by a personal computer with 4Spec software [139]. For data acquisition, the ICCD system was triggered by the output TTL pulse from the laser Q-switch system.'
P121 Fig 4.11 'As expected, the size of the plasma images on the CCD decreases when the optical restriction diameter diminishes, with plasma areas from 0.49 to 0.022mm² for pinhole diameters from 0.5 to 5mm, respectively. It should be noted that pinhole diameter larger than 3.5mm does not produce any significant decrease of the plasma image size.'
4.2.4. Depth profiles
'For the sake of comparison, depth profiles of a multilayered sample for each pinhole diameter were carried out.'
P.122 'These results confirm that the contribution of emission light provided from the borders of the plume is reduced by using this optical device. Furthermore, from this figure it is noticeable that with pinhole diameter larger than 3.5mm no evidence of the expected effect (reduction of tailing effect) was observed, which is in agreement with the plasma images studied (Fig. 4.10). Although the best result was obtained with the smallest pinhole diameter (0.5mm), a compromise between microplasma signal coming to the spectrograph and restricting effect should be found.'
P126 4.2.6. Conclusion
'In this work, the capability of laser induced plasmas with an optical restriction for improvement of depth profiling analysis of multilayered samples has been demonstrated. The use of a pinhole placed between the dichroic mirror and the collection lenses has allowed us to reduce the contribution of emission light from the border of the plasma resulting from the crater edges, which produces the "tailing effect". The optimization of experimental conditions including suitable pinhole inner diameter selection was discussed, and the delay time was studied. The results demonstrate that the optimal pinhole inner diameter is of 1.5mm. On the other hand, working at short delay times seems to be the best choice in terms of depth resolution due to the confining of the plasma close to the sample surface and non-mixing of specimens.'
P128 'The optically enhanced LIBS single-pulse instrumental set-up was tested on depth profiling of multilayered metal sample. The optical restriction (pinhole) of 1.5mm of inner diameter placed in front of the collection lens was proved to eliminate partially the tailing effects and improve thus the depth resolution. The ablation rate in order of nanoseconds resulted in depth resolution of approximately 1m. It would be convenient to couple the effect of pinhole and of the double-pulse configuration, while both approaches demonstrated the improvement of the depth resolution, to obtain even better results. Unfortunately, the connection of double-pulse with the restriction effect of pinhole can not lead to more improvement in presented instrumentation. In orthogonal double pulse method the re-heating of the microplasma leads to spatial mixing of particles in the microplasma. The application of pinhole would not have any sense. It can be generally concluded that LIBS fulfils the requirements demanded for an analytical technique.'
The pdf below seems to be a paper based on the above thesis which is only available in text with no diagrams. However I think these diagrams are all reproduced in the pdf below as the conclusion is almost word for word. No date but last ref is dated 2007:
http://documents.mx/documents/optical-restriction-of-plasma-emission-light-for-nanometric-sampling-depth.html
Optical Restriction of Plasma Emission Light for Nanometric Sampling Depth and Depth Profiling of Multilayered Metal Samples,
T. CTVRTNICKOVA , F. J. FORTES, L. M. CABALIN, and J. J. LASERNA
Department of Analytical Chemistry, Faculty of Sciences, University of Malaga, 29071 Malaga, Spain
'Improvement in depth profiling capabilities of laser-induced breakdown spectrometry (LIBS) for multilayered samples has been attempted. For this purpose, in a typical LIBS experiment, an optical restriction consisting of a pinhole placed between the dichroic mirror and the collecting lenses has been used. This new optical approach allows observing only the light emission coming from the central region of the plume. The results demonstrate that the optimal pinhole inner diameter is 1.5mm. Averaged ablation rates in the range of nanometers and depth resolution of approximately 1um have been obtained.'
http://www.researchgate.net/publication/248394646_Reproducibility_of_CIGS_thin_film_analysis_by_laser-induced_breakdown_spectroscopy
Reproducibility of CIGS thin film analysis by LIBS with a tophat NdYAG 2013.pdf
THIS ONE IS PARTICULARLY INTERESTING:
http://books.google.co.uk/books?id=s6qaAgAAQBAJ&pg=PA26&lpg=PA26&dq=%22laser+induced+breakdown+spectroscopy%22+%22top+hat%22&source=bl&ots=nJdQoH_CYS&sig=U36Uxndsght6XjDRitslGEdQAi0&hl=en&sa=X&ved=0ahUKEwikp53o6p_TAhWMOsAKHRDID3o4ChDoAQgiMAE#v=onepage&q=%22laser%20induced%20breakdown%20spectroscopy%22%20%22top%20hat%22&f=false
Spectroscopic Investigations of Plasma Emission Induced During Laser Material Processing By David Diego Vallejo, 15/01/15, 140pp
Page 29 shows a tophat converter mounted immediately below the focusing lens from a Nd:YAG for LIBS.
This also mentions the spectrometer they used apparently didn't have a gate and its integration time only went down to 1ms making it impossible to capture 1kHz pulses. More to read [these were found by Googling 'laser induced breakdown spectroscopy' 'top hat']
It's £23 on Amazon [but it's more about processing than LIBS]:
http://www.amazon.co.uk/gp/search?index=books&linkCode=qs&keywords=9783844278804
THE ONLY DOC I COULD GET RELATED TO IT WAS:
http://www.researchgate.net/publication/270999360_Monitoring_and_Adaptive_Control_of_Laser_Processes
[I think it was a freely available thesis before]
------------------------------
Laser Ablation in Analytical Chemistry - A Review LBNL-48521.pdf
Fig 2: 'White light interferometric microscope images...flat-top laser beam profiles are capable of producing straight-wall craters in a wide range of materials' [4, 32].
4. D. Günther, I.Horn, B.Hattendorf, Fres. J. Anal. Chem., 368 (2000) 4.
32. D. Bleiner, A.Plotnikov, C.Vogt, K.Wetzig, and D.Günther, Fres. J. Anal. Chem., 368 (2000) 221.
'Excimer lasers generally have ‘flat-top’ beam profiles. With appropriate imaging optics, both Nd:YAG and excimer lasers can generate flat-bottom craters (figure 2). The shape of the crater walls will influence depth resolution and fractionation may increase with the development of the ablation crater [31]. However, the degree of fractionation is not strictly related to the beam profile; fractionation is not eliminated by having a flat-top beam profile.'

FRACTIONATION
In general, fractionation is related to the optical properties of the sample, the wavelength of the laser and significantly, the laser irradiance.

Experimental UV tophat beam profiling lenses (SUSS MicroOptics, Switzerland)