Laser Ablation ICP-MS
The Department of Earth Sciences houses a Resonetics S-155-LR 193nm Excimer laser ablation system coupled to an Agilent 7700x quadrupole ICP-MS. The S-155 cell, designed by Laurin Technic Pty, is a two-volumes small volume ablation cell that provides unmatched signal washout and stability. We recently (March 2016) added a Bruker M4 Tornado micro-XRF (20 micron spot) instrument to support routine major and minor-element mapping over large areas. The instruments are being housed in a purposed-built laboratory with tightly controlled air conditioning and ventilation to guarantee minimal instrument drift. The existing Spectra CIROS CCD ICP-IOES is also housed in this lab. Major funding for this facility was provided by CFI (Leaders Opportunity Fund), NBIF (Research Innovation Fund), Xstrata Zinc, Freewest Resources, Elmtree Resources, NB DNR, and the University of New Brunswick.
Lab bookings are managed through Mr. Brandon Boucher, the lab's research technician. A fee structure is listed at the bottom of this page. New users should also read the lab policies document (currently being revised).
Since commissioning these instruments in 2010, we have established a number of analytical procedures with applications to geochemistry, geochronology, geo-metallurgy, and tracer-isotope studies. The information below describes some of the pertinent components of the system
- Conventional geochronometers with negligible common-Pb: zircon, monazite & xenotime
- Conventional geochronometers with common-Pb: Allanite, titanite, rutile, apatite
- Unconventional geochronometers under development: garnet, uraninite, cassiterite, carbonates
- Micro-XRF mapping
- Trace-element measurements and mapping
- Pb-Pb isotopes in galena
- Optical cathodoluminescence
- Laser milling
- Whole rock geochemistry
- Magnetic separation
- Sample preparation facilities
Recent news and advances
1. October 2017. Fast-mapping LA-ICP-MS setup
Adjustments have been made to our tubing diameter and material (1.5mm ID ETFE), and tubing length (1 m) to increase the speed of washout to the ICP-MS. Under these new conditions the signal washout is roughly 2.5 orders of magnitude/sec. As a result we are able to increase the scan rate for mapping while still capturing chemical variations. The following example Ni and Sr maps of a roughly equant 500 um wide opx grain was generated in 12 minutes using a 17 um crater, 34 um/sec scan rate, and 15 Hz pulse rate. The data was processed using Iolite3.5 and the Cellspace imaging technique. A typical signal washout at the end of a NIST610 tuning line is also shown. We are currently experimenting with 1.0 and 1.2mm ID tubing for even faster washout.
2. June 2017. New advances in in-situ U-Pb geochronology of unconventional minerals
We have been using well-characterized geological settings to test the viability in-situ U-Pb geochronology of U-bearing minerals such as:
See the 'unconventional chronometers' page for mor information
3. May 2017. New Monocle software for Iolite3.5 incorporated into routine mapping applications
The recently published description of Monocle by Kamber et al (2017) provides a new way forward for trace-element and geochronology studies.
4. March 2017 Laser-milling of garnet for spatially-controlled Lu-Hf and Sm-Nd geochronology
In collaboration with colleagues from University of Albama, we are refining a process to liberate domains (e.g., cores) from garnet in doubly-polished thick (100um) sections guided by major, minor, and trace-element mapping. The goal is to refine this process to help transform the robustness of Lu-Hf and Sm-Nd studies involving garnet-wholerock isochrons.
5. January 2017. In-situ Pb-Pb isotope measurements in galena
The lab has established and verified a methodology to measure Pb-Pb isotopes in galena with a grand weighted mean precision of 0.05-0.1%. The method is incredibly flexible, being applicable to 1mm diameter grains exposed in polished thin sections are liberated and mounted to exposed a flat cleavage face.
6. Increasing user fees effective April 1st, 2016
The significant increase in the cost of laser ArF premix (>100% increase!) has forced us to increase our user fees by 20% for the 2016 fiscal year. This will hopefully be a temporary increase if premix prices recover to something remotely closer to previous costs. Please contact Chris or Brandon for new fees for internal, external academic (NSERC), external academic (NSF), and external commerical work.
7. March 21, 2016. Installation and commissioning of Bruker M4 Tornado micro-XRF
This new CFI and NBIF suppored instrument provides fast elemental mapping capabilities and EDS-based quantitation of materials at much higher sensitivity that electron-beam instruments. The instrument has a Rh X-ray tube with polycapillary focusing to 20 um, dual SDD detectors (130-140 eV resolution), a large stage and vacuum mode to enhance light element performance. Sample preparation is minimal, but flat samples are still necessary. A single polished thin section (60 micron thick recommended) can be mapped for element abundances and phase proporations in roughly 1.5 hours at 30 micron resolution using a 40kV 500uA beam. This is sufficient to map the distribution of large accessory minerals (e.g., >50 um diameter) we routinely use for geochronology (see below). The elemental maps can be exported (as .jpg) and used as image overlays in the laser software (Geostar v7). Detection limiits are lowest for transition metals (10s of ppm). The instrument is available for use on a cost-recovery basis with technical support from lab staff.
Images (click on images for a larger version): installed M4 Tornado; mosaic of polished thick(60 um) section; multi-element map of sample; distribution of apatite and zircon revealed by P (Ca not shown) and Zr; overlay of element map in Geostar with reflected mosaic of an area containing several 50-100 um apatite grains (green in element map).
8. Spring - 2015. Resurrection and integration of cold-cathode CL system (Nuclide Corporation Luminoscope: http://serc.carleton.edu/research_education/geochemsheets/opticalcl.html)
This cold-cathode optical CL system allows us to rapidly collect true-colour CL mosaics over larger areas. An array of CL photomicrographs collected within a conventional DSLR (operated high ISO800 and 5sec exposure with noise reduction) are collected, automatically stitched using Microsoft Image Composite Editor (ICE), and enhanced in Photoshop. the stitched CL image is then imported into offline Geostar software and used to guide the locations of ablation targets. The combined ablation sequence and background image are saved as a Geostar .sqnimg file and geo-referenced to the laser stage coordinates for online data acquisition.
In contrast to our SEM-CL system (Gatan ChromaCL) operated by UNB MMF, the cold-cathode system has the advantage of 1) much higher throughput (i.e. 10 secs per image); 2) wider field of view (2mm); 3) and true colour imaging. The main disadvantage is the lower resolution of the images - but this will continue to improve with planned hardware upgrades to the microscope objectives and digital camera head. Three views of a detrtial zircon sample are shown below (transmiited light, reflected light, and cold-cathode true colour CL).
9. Winter 2015 - Synthesis of glass reference standards using high-T flux-free fusion
We have begun experimenting with the creation of matrix-matched glasses using powdered minerals as starting materials. For example, several large grains of titanite from an Otter Lake skarn sample, which have been demonstrated to be chemically and isotopically heterogeneous, were powdered in an agate mortar and fused on our in-house glassy-carbon strip heater. This titanite powder melted at ca. 1500C and formed perfect glass spheres upon quenching. This approach produces homogeneous glasses with major element chemistry (measured by SEM-EDS) that is identical to natural titanite. Working values for Pb/U, Pb/Th, and Pb/Pb will be established by LA-ICP-MS using NIST610 as external standard whereas more precise Pb/U, Pb/Th, and Pb/Pb isotopic compositions will eventually be measured by TIMS. The images below show our glassy carbon fusion 'boat' and a spherical glass bead (left) and an enhanced backscattered electron image (SEM-BSE) of the resulting Otter Lake titanite glass.
Description of instrumentation
Some key benefits of the Resonetics S-155 laser ablation system:
- Aerosol washout is rapid even with in-line signal smoothing (5 orders of magnitude in about 5 seconds). This allows us to employ short background counting times (typically 30 to 40 seconds) between each ablation. For high-spatial resolution techniques such as linescans or 2D elemental mapping, the smoothing device is removed to obtain even faster washout times.
- Large format sample holders capable of accommodating: 1) six 45x25mm thin sections and five 1” round standard blocks; 2) fifteen 1” round pucks; 3) irregular shapes up to 100mm x 120mm and <20mm height in the universal holder. This allows us to load several samples along with all the standards we need to carry out an analyses: tuning standards, external standards, and consistency (or QC) standards.
- The design of the S-155 results in identical ablation anywhere within the cell. As a result, standards do not need to be mounted together with unknowns. We simply drive back and forth between unknowns and standards.
- Sensitivity enhancement using N2 in the carrier gas and 2nd external rotary pump yields ~1300-2000 cps/ppm for U and Pb using a 24µm crater, 4 Hz pulse rate and laser fluence of 4 -6 J/cm2. This allows us to analyze minerals at high spatial resolution using shallow craters to minimize down-hole elemental fractionation.
- Smooth, stable, ablation signals with raw %RSD for inter-element ratios <2%.
- Climate-controlled instrument lab maintains 12 hour stability <2% RSD as monitored by replicate standard analyses.
- Stage reproducibility of <5 µm, with backlash corrections calibrated routinely.
- Laser control software (Resonetics Geostar) for online and offline sequence definition and automated acquisition. Options for unsupervised and supervised ablation.
- Between 350 and 400 craters per 8-hour day for routine analyses.
Samples must be well-polished to 1µm. Carbon coat should be removed. Normal thickness (30 µm) polished thin sections are suitable for most applications. Analysis of quartz, feldspars, fluorite, carbonate and fluid or melt inclusions require 60-80 µm thick samples. The locations of targets must be well-defined and identified. For obvious targets it is acceptable to simply circle the areas of interest. For small targets, detailed reflected light images (e.g., taken using a 10x objective) or SEM/BSE images are required to help guide the identification of target grains.
Visiting the lab
For collaborative projects we encourage PIs to be on-hand if possible to help direct the acquisition of LA-ICP-MS data. Fredericton offers a variety of low-cost accommodation including the nearby Rosary Hall Hostel that provides a quiet, convenient, and accessible stay for visiting students. Visitors to the lab over the summer months can take advantage of the UNB Summer Residence and Hotel services for on-campus housing. Off-campus hotels are also located near the UNB campus (e.g. Crowne Plaza 'Lord Beaverbrook' Hotel).
A typical analytical day (8:30 to 5:00) involves loading samples and standards into the holder, evacuating the cell, and defining an ablation sequence (either online or importing a sequence defined offline). The laser and ICP-MS are then tuned using NIST610 to maximize sensitivity while also minimizing oxide production to <0.3% (as monitored by ThO+/Th+) and, for trace-element analyses, double-charged production to <0.3% (as monitored by 22M+/44Ca++). These tuning conditions help minimize the most common isotope interferences. The ICP-MS method is checked and isotopes and dwell times adjusted based on the likely abundance of each analyte (plus consideration of isobaric interferences or tailing from high abundance peaks such as 40Ar under 39K).
The ablation sequence is then defined. This will typically comprise a large number of unknowns interspersed with external calibration standards and consistency standards. The total time required to carry out the ablation sequence is calculated. This value is entered in the ICP-MS method definition. The intensity data for each ablation is, therefore, collected in a signal ICP-MS file that may be up to 9990s long (about 2.75 hours). This is typically enough time to collect between 140 and 160 ablations per run. For U-Pb geochronology we use a minimum of 15 calibration standards interspersed with unknowns as well as 5 consistency standards.
At the end of the ablation sequence, the laser control software (Geostar) generates a laser log file which is meshed to the ICP-MS output file using Iolite (Paton et al., 2011) data reduction software. U-Pb geochronology is additionally reduced using VizualAge (Petrus and Kamber, 2012). This data reduction can be done quickly enough (e.g., 5 to 10 minutes) so as to check the results for quality and consistency prior to embarking on the next ablation sequence. Because preliminary data reduction can be carried out at the end of each ablation sequence, users typically leave the lab at the end of the day with a large dataset in hand.
Offline sequence definition
The laser ablation system is designed to import ablation sequences defined offline. This requires a reflected light image that can be used to identify targets. We use a Zeiss AxioImager polarizing microscope with a motorized X-Y stage to create seamless transmitted and reflected-light mosaics of polished thin sections up to 75mm × 100mm. These mosaics are used as a basemap to identify targets and define offline ablation sequences. The locations and labels for each ablation target are stored as a hybrid image-sequence file. This hybrid file is then imported into the online Geostar software, georeferenced to the stage coordinates, and the ablation sequences loaded with the correct X-Y laser stage coordinates. Completing the sequences involves only the addition of standards.
This offline ablation sequence definition approach can save a significant amount of time for studies involving large targets that are easily visible in reflected light mosaics (e.g., detrital zircon studies or trace element studies of minerals). In these caes, data collection can be expedited using the following approach:
- User couriers polished pucks ahead of laser session to allow time to create reflected light mosaics of the target grains.
- User downloads offline Geostar software, which is freely distributable to our lab users.
- User receives reflected light mosaics via email (or Dropbox etc.) and generates an ablation sequences for each sample from the comfort of their own desk (keeping in mind that 140 points per sample is maximum for a single run). Instruction videos are available to guide you through this process.
- User arrives for laser ablation session with offline sequences in hand, allowing us to begin data acquisition immediate.
Alternatively, if users are able to arrive 1 or 2 days early, they are welcome to spend time on the Zeiss AxioImager doing the imaging and sequence definitions.
Description of methods
Trace element characterization of minerals and whole-rock powders
Trace element analyses are ideally carried out on mineral sub-domains previously analyzed by another microbeam technique (e.g., SEM-EDS or EPMA-WDS) to obtain an independent measure of the major and minor element concentration of the target. This allows us to use a robust internally-standardized data reduction scheme to provide the most accurate trace element data. Electron microbeam imaging techniques such as backscattered electron (BSE) and cathodoluminescence (CL) imaging are also typically required to reveal mineral-scale heterogeneity to guide ablation targets. For some minerals that display near-stoichiometric concentrations of one or more element, we can also use an assumed stoichiometric value for an internal standard concentration. In general, extensive imaging and independent characterization of the targets significantly increase the reliability of LA-ICP-MS trace element data.
For whole-rock geochemistry, we use a custom-built strip heater capable of passing up to 300A @ 12V through metal strips with high melting points. Strips made out of Ir, Pt, W, or C can be used depending on the application. The method requires no fluxing agents so that Li, B, and Br are also analyzed. Users typically provided the powders. Once fused, the glass beads are mounted in epoxy, cured, and polished to their centers to avoid areas affected by volatile loss or contamination from the strip material. The major elements (Si, Al, Ti, Fe, Mn, Mg, Ca, K, Na) as well as F and Cl are then analyzed using an automated overnight run on the UNB Microscopy and Microanalysis Facility JEOL 733 electron microprobe. The pucks are collected in the morning cleaned of carbon coat and placed into the LA-ICP-MS system for full (50+ element) trace element characterization using a large crater (e.g., 100 µm) raster analysis. Trace element concentrations are calibrated against NIST glasses using a major element analyzed on the microprobe as an internal standard (e.g., Ca). Results are QA/QC checked against well-known USGS standards such as GSE-1G, BCR-2G or MPI-DING glasses.
We use a variety of widely distributed and certified reference materials as external calibration standards for LA-ICP-MS. USGS, NIST, MPI-DING, and NRCAN standards are used extensively as primary and quality control standards.
Independent of the number of target grains, a minimum of 15 zircon standards is included in each ablation sequences to ensure that propagated errors (i.e., internal + external) are minimized. We use either 91500 or FC-1 depending on the required crater diameter and U-content of the targets. ‘Normal’ zircon is typically analyzed using 20-30 µm craters, 4.5 Hz pulse rate, and 5 J/cm2 fluence.
Zircon can be analyzed in either grain mount or in-situ in standard polished thin sections. For defining magmatic crystallization ages, target precision of 0.5 to 1% (2σ) can typically be achieved for populations of at least 10 concordant grains requiring no common-Pb correction. For in-situ dating using thin sections, a minimum of 30 ablations is typically required to obtain coherent magmatic or metamorphic zircon populations. Our ability to load up to 6 thin sections into the ablation cell means that even samples with sparse zircon can be reliably dated in-situ by pooling results from multiple thin sections. Figure 1 shows the results of a near-ideal zircon geochronology run in which the zircons were both U-rich and concordant.
For detrital zircon studies, a minimum of 120 spots is recommended. We also run a minimum of 15 primary standards (e.g., 91500) and 5 consistency standards (e.g., Temora) for a total of 140 craters per sequence. Each point takes 1 minute (30s bkgd, 30s ablations). The fast washout and low-contamination properties of the S-155 cell ensure that background intensities are recovered <5s after the end of each ablation. It is typical to obtain 3 detrital zircon analyses in an 8-hour work day, with an additional sequence possible after hours. Hence a total of 420 ablations per 8-hour day can typically be achieved yielding a cost/crater (at current rates) between ~$2.50 to $3.50 for collaborative and non-collaborative work, and $4.75 for industrial clients. Figure 2 is an example of detrital zircon provenance.
The stability of the lab environment and the instrumentation is well demonstrated by replicate analyses of a homogeneous standard (91500 zircon in the example in Figure 3). The variability of the raw 206Pb/238U (i.e., without any down-hole fractionation correction) is better than 2% (excluding two outliers in red) over the course of almost 12 hours of running time.
In contrast to zircon, monazite can contain up to several thousand ppm U and hundreds of ppm Pb so that very small craters can be used for in-situ dating. We prefer to work with crater sizes in the range of 5 to 15 µm as we have found this minimizes instrumental biases (e.g., reverse discordance) that have hampered in-situ monazite geochronology in the past. We have a collection of monazite standards (Trebilcock, 8153, 44069, Thompson Mine) spanning a range of U, Th, Y, and REE concentrations. This allows us to matrix-match the standards and unknowns as closely as possible to minimize laser-induced Pb/U fractionation and mass-load effects in the ICP-MS. For each new monazite dating project, the first run involves the analyses of a few representative unknowns in addition to at least two monazite standards, each with 12 to 15 analyses. This helps establish which standard will be best matched to the target monazite. Subsequent sequences are then defined to include the most appropriate standard and one or more quality control materials to verify the accuracy of the fractionation and mass-bias corrections. Figure 4 shows ages for Trebilcock and 44069 monazite standardized against 8153.
Overlays of compositional maps obtained by electron microprobe can be used to help target the locations of in-situ U-Pb ages. Figure 5 shows an X-ray map overlay on a reflected light laser video feed used to help guide the positioning of craters for U-Pb dating.
Titanite and rutile geochronology
In contrast to monazite, titanite and rutile contain much lower U concentrations as well as variable initial common-Pb concentrations. As a result, we use longer background counting times (40s to 60s) in order to more accurately measure 204Pb. All of our gas lines are fitted with high-capacity VICI Metronics Hg-traps. Our typical mass-204 gas background resulting primarily from 204Hg contamination is 120-170cps, equivalent to ~100ppb (background-equivalent concentration or BEC) at maximum sensitivity. In cases where 204Pb is measurable above the gas background but <10% of the total Pb budget, we are typically able to make an accurate 204Pb-based common-Pb correction. For analyses containing large amounts of common-Pb, an inverse-isochron approach is used to calculate a lower-intercept age using the 207Pb/206Pb and 238U/206Pb ratios. Down-hole laser-induced elemental fractionation is reliably corrected using BLR-1 titanite and R10 rutile. Quality control standards for titanite include OTL-1, TCB, and Khan Mine. Rutile standard R13 is used to verify the accuracy of rutile analyses. Figure 6 shows 204Pb-corrected and -uncorrected data for magmatic titanite from the Howard Peak Granodiorite, NB.
Apatite, allanite, and epidote geochronology
Although no widely-distributed apatite standards exist, we have successfully used a combination of zircon 91500 and a raster-ablation approach to obtain U-Pb apatite ages. We have recently established ablation protocols to accurately and precisely date allanite using NIST610 as an external standard. We are currently developing in-house allanite and apatite standards to verify the accuracy of the analytical protocols.
Trace element mapping
The very fast and stable aerosol washout performance of the S-155 sample cell allows us to conduct routine trace element mapping using a variety of different ablation methods. The simplest approach is to cover an area of interest with a series of raster scans which are then integrated offline into concentration maps using Iolite software. Crater size and scan rate are varied in proportion to the size of the target area. Large targets (e.g., 1mm × 1mm) are typically scanned using a 25 to 50µm crater, scan speeds between 12 to 25 µm/s, and 10 Hz pulse rate. Total acquisition time is about 1 hour and we typically analyses between 20 and 25 elements. Larger targets (a few mm) require the use of larger craters and faster scan speeds to ensure that the total acquisition time per target is <2.75 hours. From left to right, the images in Figure 7 show the distribution of V, Sc, Lu, and Y in garnet.
Zoning in speleothems and otoliths
Calcareous materials with annual growth domains can be analyzed using our rotating rectangular slit. A combination of horizontal and vertical slits is superimposed and the image demagnified onto the sample surface to generate narrow rectangular ablation pattern that can be scanned across growth bands. For example, to core to rim otolith zoning patterns we use a rectangular pattern that is 30µm × 5µm, scan at 5 µm/s, and use a pulse rate of 10 to 15Hz. Much larger speleothem samples requiring long continuous raster scans are analyzed with a 100µm × 10µm rectangular pattern and scan speed of 10 µm/s. By removing the in-line smoothing device (the ‘squid’) fast washout of >6 orders of magnitude in 1 second ensure that compositional variations are captured at the same scale as the width of the slit. Figure 8 shows variations in Zn and Sr moving outward form the nucleus on an otolith. We also no use the Luminoscope to record true-colour CL images of otoliths prior to laser ablation. The CL reveals fine-scale oscillatory zoning and changes in otolith chemistry related primarily to Mn concentration.
LA-ICP-MS for U-Pb geochronology and trace-elements (spots and maps)Fee structure (effective April 1 2016 to April 1 2017)
Per day fees for LA-ICP-MS include: technician time, instrument running costs, consumables, standards, data reduction, and reporting. A working day is 8:30 to 5:00. Collaborative projects are expected to include co-authorship on derivative publications. 10% discount typically applied if end-users visit to aid in setup and acquisition of data. A summary of fee structure.
|External academic (NSERC) collaborative||$1632 per day|
|External academic (non-NSERC) collaborative||$1920 per day|
|External commercial||$2500 per day|
|All raster mapping projects will incur a 20% surchage on hourly laser ablation instrument time. This amounts to an additional $96, $115, and $150 per full day of mapping for the three service categories above|
Otolith analyses package
Solution-ICP-MS analyses are charged based on the number of samples. The cost per sample includes cost-recovery of consumables (tubing, cones, lenses, solutions, labware) and technical support. We have two categories for solution work:
|Solution ICP-MS: <20 elements|
|External academic (NSERC) collaborative||$28 per sample|
|External academic (non-NSERC) collaborative||$34 per sample|
|External academic non-collaborative||$56 per sample|
|Solution ICP-MS: >20 elements|
|External academic (NSERC) collaborative||$34 per sample|
|External academic (non-NSERC) collaborative||$56 per sample|
|External academic non-collaborative||$112 per sample|
Epoxy mounting and polishing
We use Struers Epofix epoxy for mounting. Grinding and polishing is carried out using a combination of 300, 600, and 1200 grit wet/dry sandpaper and 6µm, 3µm, and 1µm diamond paste on a Buehler Minimet polishing machine. Specimen mounting, flattening, and polishing is $50 per sample.
Zeiss AxioImager imaging
$25 per hour for all projects.
Strip-heater fusion and whole rock geochemistry
Strip-heater fee includes: drying (if necessary) and weighing powders, fusion, epoxy mounting, grinding and polishing, EPMA, LA-ICP-MS, data reduction and reporting.
|External academic (NSERC) collaborative||$120 per sample|
|External academic (non-NSERC) collaborative||$130 per sample|
|External academic non-collaborative||$150 per sample|
|External commercial||$170 per sample|
Paton, C., Hellstrom, J.C., Paul, B., Woodhead, J.D., and Hergt, J.M. (2011) Iolite: Freeware for the visualisation and processing of mass spectrometric data. Journal of Analytical Atomic Spectrometry, 26, 2508-2518.
Petrus, J.A., and Kamber, B.S. (2012) VizualAge: A novel approach to laser ablation ICP-MS U-Pb geochronology data reduction. Geostandards and Geoanalytical Research, 36, 247-270.