Chemical Analysis of a Lunar Meteorite by Laser Ablation Mass Spectrometry


Master's Thesis, 2016

72 Pages, Grade: 5.5


Free online reading

Contents

1 Introduction

2 Basic Principles of Time-of-Flight (TOF) Mass Spectrometry & Laser Mass Spectrometer (LMS) for space research
2.1 Ion Source
2.2 TOF Mass Spectrometer
2.2.1 Linear TOF Analyzer
2.2.2 Reflectron
2.2.3 Time-to-Mass Calibration
2.2.4 Mass Resolution
2.3 Detector
2.4 The LMS Instrument

3 Measurement Procedure and Data Processing
3.1 Sample Preparation
3.2 Signal Optimization
3.3 Data Analysis

4 Measurements on a Lunar Meteorite
4.1 Fundamentals of the Sayh al Uhaymir 169 Meteorite
4.2 Energy-Dispersive X-Ray Spectroscopy
4.3 Determination of the Relative Sensitivity Coefficients and Chemical Mapping of an Unknown Area
4.3.1 Introduction
4.3.2 Measurements and Results
4.3.3 Conclusion
4.3.4 Element Composition Maps-Regolith Sample . .
4.3.5 Element Composition Maps-2nd Sample
4.4 Chemical Composition Measurements on the K-REE-P Part of the SaU 169 Meteorite
4.4.1 Introduction
4.4.2 Measurements and Results
4.4.3 Conclusion
4.4.4 Element Composition Maps of the K-REE-P Sample
4.5 Crystallization Temperature of Zircon
4.5.1 Introduction
4.5.2 Measurements and Results
4.5.3 Conclusion

5 Measurements on a K-REE-P Sample with a HV Pulser: Pre- liminary Results
5.1 Working Principles of the HV Pulser
5.2 Measuring Trace Elements by using the HV Pulser
5.2.1 Instrument Settings
5.2.2 Measurements and Results
5.3 Radiometric Dating Methods
5.3.1 Theoretical Principles
5.3.2 Measurement and Results
5.4 Conclusion and Outlook

Abstract

The lunar meteorite SaU 169 was investigated with a miniaturized laser mass spectrometer (LMS). SaU 169 consists of two different lithologies, a regolith breccia and an impact-melted breccia. The latter is extremely enriched in potas- sium, rare earth elements and phosphorus (K-REE-P) as well as the radioactive elements uranium and thorium, which makes it special compared to other known meteorites. The chemical composition of a meteorite can bring insight into its formation conditions. In a first step a set of relative sensitivity coefficients (RSCs) was determined by measurements on the regolith rock. Afterwards the chemical composition of the K-REE-P rock was successfully measured and mapped in an elemental composition map. The present minerals (e.g. zircon, ilmenite, pyroxene and K-feldspar) could be identified by analyzing elemental ratios. Special emphasis was put on the mineral zircon. Rare earth element patterns could be measured inside a zircon grain. With the amount of titanium found inside the zircon grain, the crystallization temperature of the grain could be estimated. This again can help determine whether or not the analyzed grain crystallized during an impact event or not. In the last part of this thesis the LMS instrument was coupled with a HV pulser. The HV pulser greatly enhances the detection sensitivity for heavier elements by preventing the lighter elements from reaching the detector and thus avoiding saturation effects on the detec- tor. With the help of the HV pulser the mass spectra of rare earth elements could be readily recorded with sufficiently high signal-to-noise ratio and mass resolution to make a quantitative analysis possible. Uranium and thorium were detected as well at limited sample locations. Also lead was measured but only in one spectrum. With the lead and uranium abundance determined from the spectrum an estimation of the age of the impact event, where the zircon grain crystallized, could be made and is found to be in good agreement with the age reported in literature.

Chapter 1 Introduction

The chemical composition of planetary surface material is an important scien- tific question on every space mission to a planet, asteroid or the Moon. Chemical composition measurements of rocky material on the surface help to understand the origin and evolution of the planetary body. Spectroscopic methods such as gamma-ray, X-ray or neutron spectroscopy have yielded valuable scientific data in the past. However, these methods are limited by their low sensitivity as well as their inability to investigate subsurface material. The study of extraterres- trial material in a laboratory environment is limited to the study of meteorites found on Earth or collected by a sample return mission. Missions to bring a sample from a planetary body bear the risk of contamination or any other kind of alteration as well as being expensive and become increasingly difficult for far away planetary objects. For these reasons, powerful in-situ instruments are needed.

Laser ablation and ionization time-of-flight spectrometry is one of the most important analytical technique for applications in space research. To date sev- eral miniaturized laser ablation mass analyzers have been developed for the use of in-situ investigations of planetary material. The miniature laser abla- tion mass spectrometer (LASMA) was part of the Phobos-Grunt mission to study the chemical composition of the Phobos surface1. Unfortunately, the mission failed due to technical malfunctioning of the thrusters, leaving Phobos- Grunt stranded in the earth orbit. Another miniature laser ablation/ionization reflectron-type time-of-flight mass spectrometer (LMS) was designed and built at the University of Bern in 2003 for planetary research2.

The LMS instrument is able to provide the same capabilities as large labo- ratory systems despite its small size and light weight, making it suitable for its application on planetary space missions. A dynamic range of about 8 orders of magnitudes as well as a high lateral (μ m-level) and vertical (sub-nm level) res- olution and a high detection sensitivity for almost all elements (10 ppb, atomic fraction) make the LMS instrument useful for numerous applications. LMS is suitable for in-situ measurements of the elemental and isotopic composition with high precision and accuracy. By measuring the lead isotope abundances the age of planetary material can be determined. Isotopic ratios in bio-relevant el- ements help with the search of past or present life on planetary surfaces3 4 5.

Studies on meteorites provide crucial information about the history of the solar system. For this reason the lunar meteorite Sayh al Uhaymir (SaU) 1696 is investigated in this thesis. Mass spectrometry on SaU 169 yields information about the mineralogical composition of the meteorite. The possibility of in-situ dating with the LMS instrument is also investigated.

Chapter 2 Basic Principles of Time-of-Flight (TOF) Mass Spectrometry & Laser Mass Spectrometer (LMS) for space research

1 Mass spectrometry is an important analytical technique for the sensitive de- termination of the elemental and isotopic composition of matter and the struc- tural analysis of molecular compounds. There are variety of mass analyzers developed to separate samples into individual atoms or molecules and measure their concentrations. In the present studies, laser ablation ion source and time- of-flight (TOF) mass spectrometer (LMS) are applied. The LMS instrument is a miniature TOF mass analyzer developed for the investigation of the chemi- cal composition of planetary solids for future planed planetary missions. The working principles of the TOF analyzer was first described by Stephens in the 1940’ies8. In 1955 Wiley and McLauren presented the first design of a linear TOF mass spectrometer which later became the first commercial instrument9. The development of ion sources such as electron impact, Matrix Assisted Laser Desorption/Ionization (MALDI) and laser ablation as well as progress in the development of fast electronics renewed the interest in time-of-fight instruments and opened up new applications in various research fields.

TOF analyzers have become increasingly attractive also in space research due to the relatively simple and robust design of the mass analyzer. Fast measure- ments with a dynamic range close to six decades of a complete mass spectrum within seconds is also advantageous over other instruments, where scanning over the entire mass range is necessary. In the following section a brief introduction of the theory of TOF mass analyzers and their basic operation principles will be given below. The principles of a laser ablation ion source and ion detection will also be explained.

2.1 Ion Source

The ion source is an essential component of a mass spectrometer, delivering ionized molecules/elements for the mass analysis. Among several ion sources based on different evaporation and ionization processes known in the analysis of solid samples, including spark ion source, glow discharge ion source, laser ion source, secondary ion source, sputtered neutral ion source and inductively coupled plasma ion source, laser ablation ion source gains increasingly on im- portance.

The laser ablation source allows for a fast and precise, in-situ, spatially re- solved measurement of elements and isotope ratios with high analytical quality and minimal sample preparation. The laser ionization ion source produces a laser plasma, where the ionization of ablated sample material takes place under high vacuum conditions. The solid sample surface, interacting with photons at a laser power density of > 108 W/cm2 (nanosecond-laser) and > 1011 W/cm2 (femtosecond-laser) evaporates, atomizes and ionizes the solid sample. The ex- perimental conditions for LIMS (laser ionization mass spectrometry) can be optimized with respect to a stoichiometric ablation/ionization of solid sample material by varying the laser power density. For laser power densities larger than 1011 W/cm2, typically stoichiometric laser evaporation and ionization of analyzed material is found. In this laser power density range, the relative sen- sitivity coefficients (RSC) of the chemical elements (see Section 4.2 for further explanation) are nearly one for all elements. The ion source applied in this study uses a fs-laser with near-IR radiation output (775 nm, 150 fs, 1 kHz). The laser radiation beam profile is expanded by a Galilean telescope from 5 mm to 30 mm and focused at a right angle onto sample surface by a doublet lens with a focal length of +200 mm (see Fig.2.4).

2.2 TOF Mass Spectrometer

To get an overview of the principles used in time-of-flight mass spectrometry, it is useful to first take a look at the most basic TOF mass spectrometer called linear time of flight mass spectrometer.

2.2.1 Linear TOF Analyzer

In Figure 2.1 a scheme of a linear TOF mass spectrometer is shown. Due to their simple design and easy handling, linear TOF mass spectrometer are often used when moderate mass resolution required (m/Δm 100). A linear time-of- flight mass spectrometer consists of an ion source with an acceleration region, ion optics to focus the ions in time and space and a field free drift tube between the acceleration unit and the detector.

illustration not visible in this excerpt

Figure 2.1: Schematic picture of a basic linear time-of-flight mass spectrometer. The ions get produced from the sample with the help of laser ablation. The positive ions of the ion plume are accelerated by an electric field before entering the drift tube. At the end of the field free region their arrival time is recorded at the detector unit.

After the formation of ions through a pulsed laser beam, the ions enter the acceleration region where they are accelerated by an electric field E s (with the potential φ s) over a distance s towards the extraction grid. The extraction field is switched on long enough to accelerate all ions to the same kinetic energy E k,

illustration not visible in this excerpt

Rearranging Equation 2.1 gives a velocity v of

illustration not visible in this excerpt

After this initial acceleration the ions enter the field free drift tube where the different masses causes a distribution in velocities. The time t D, needed to pass a d rift tube of length D, can be calculated according to

illustration not visible in this excerpt

However, in reality not all ions will start with the same initial energy. The ions initially moving towards the detector will arrive there before the ions which were initially moving away from it. The ions moving away from the detector will first be decelerated to a zero velocity before they are being accelerated towards the detector. The time for this to happen is referred to as the turn around time t u. It can be expressed by an initial translational energy U th,

illustration not visible in this excerpt

The time t 0 needed to enter the drift tube after extraction and acceleration is equal to

illustration not visible in this excerpt

and the flight time in the field free region is now

illustration not visible in this excerpt

Finally, the total flight time t TOF is given by

illustration not visible in this excerpt

The initial spread in energy is what severely limits the mass resolution of the linear time-of-flight mass spectrometer.

2.2.2 Reflectron

The reflectron TOF mass spectrometer is an advanced kind of a TOF-MS, with an increased mass resolution. The idea was first proposed by Mamyrin10 in 1973. The reflectron acts as an ion mirror positioned on the opposite site of the source. A diagram of a reflectron type TOF analyzer is shown in Figure 2.2. The produced ions are extracted into the drift space with the total length L parallel to the axis of the device. The ions entering the retarding field of the reflectron get decelerated until their kinetic energy reaches zero and are then again accelerated back into the field free region before reaching the detector, now positioned on the same side as the ion source. The reflectron usually has two stages with individually adjustable voltages: a decelerating (length d t and potential φ t) and repelling region (length d k and potential φ k), where the ions reverse their direction of motion.

The total flight time will now be given by

illustration not visible in this excerpt

with the flight time in the field free region t L, t T the flight time in the retarding field and t K the flight time in the repelling field. The time in the field free region is described by

illustration not visible in this excerpt

v 1 describes the velocity of the ion when entering the decelerating field and v 2 the velocity when leaving the repelling field,

illustration not visible in this excerpt

The total time spend in the decelerating field is given by

illustration not visible in this excerpt

and the time t k will be equal to

illustration not visible in this excerpt

Resulting in a total flight time t TOF

illustration not visible in this excerpt

The ion mirror improves mass resolution by doubling the length of the flight tube without significantly changing the dimensions of the TOF analyzer. Addi- tionally the kinetic energy dispersion of ions leaving the ion source with the same [illustration not visible in this excerpt]ratio will be corrected. An ion with a higher kinetic energy will penetrate the reflectron more deeply than another with a lower kinetic energy, therefore spending more time in the retarding field and finally reaching the detector at the same time as the other ion.

illustration not visible in this excerpt

Figure 2.2: Schematic drawing of a time-of-flight reflectron mass spectrometer with two grids. The ions enter get decelerated to a kinetic energy of zero before being accelerated back to the detector.

2.2.3 Time-to-Mass Calibration

One advantage of TOF instrument is the easy calibration of the mass scale from the recorded time of flight spectrum. From Equation 2.13 the mass is given by

illustration not visible in this excerpt

[illustration not visible in this excerpt] and[illustration not visible in this excerpt] are now spectrometer dependent constants. To get a calibrated mass scale at least two mass peaks at times[illustration not visible in this excerpt] and[illustration not visible in this excerpt] have to be identified with their masses[illustration not visible in this excerpt] and[illustration not visible in this excerpt] respectively. With

illustration not visible in this excerpt

follows

illustration not visible in this excerpt

and

illustration not visible in this excerpt

Using additional peaks to mass calibrate a time spectrum improves the calibra- tion.

2.2.4 Mass Resolution

One of the most important parameters to judge the quality of any mass spec- trometer is the mass resolution R. It is a measurement of how well different masses can be distinguished. Since the mass is proportional to the square of the flight time, the mass resolution follows directly from the resolution of the time t TOF.

illustration not visible in this excerpt

where c is a constant. The mass resolution[illustration not visible in this excerpt] isthereforegivenby

illustration not visible in this excerpt

The mass resolution of any mass peak at a time t can therefore be calculated by measuring the the time-spread of the peak. This is usually done at the 50% peak level. Equation 2.19 makes it clear that to improve the mass resolution of a time-of-flight spectrometer the flight time should be as long as possible and the time-spread Δ t at the detector as small as possible. One way to improve the mass resolution is by the use of reflectrons.

2.3 Detector

The detector needs to produce a signal proportional to the number of incoming ions per unit of time. One of the most commonly used detectors in combina- tion with a TOF instrument are MCPs (multichannel plates, compare Figure 2.3) due to their very fast response time. Multichannel plates or micro-channel plates are continuous electron multipliers, consisting of many cylindrical tubes (channels) arranged parallel to a plate. These channels can range in diameter from 4 to 25 micrometers and have a length of up to a few millimeters. The inside walls of each channel are covered by a semiconducting substance. An ion entering one of the channel and colliding with the channel wall will cause the release of secondary electrons which will then again collide and release more electrons. A voltage gradient along the channel ensures the acceleration and therefore multiplication of the electrons. At the output side, a metal anode collects the stream of secondary electrons and the generated electric signal can then be measured. Since the path inside the channels is very short, the response time of the detector is extremely fast, making them well suited to time-of-flight analyzers as already mentioned. Additionally, the large detection area of the MCP allows for detection of larger ion beams without additional focusing. How- ever, the MCP detectors have some disadvantages. They are fragile, sensitive to air and expensive.

illustration not visible in this excerpt

(b) Electron multiplication in a single channel. An impinging particle (either an electron or an ion) hits the wall of the channel, producing secondary electrons in the process. A voltage gradient causes the electrons to speed up and release even more electrons when colliding with the wall.

(a) Drawing of a multichannel plate. The single channels are assembled in a honeycomb like structure.

Figure 2.3: Multichannel Plates: Schematic drawing of a whole MCP plate (a), electron multiplication in a single channel (b). Pictures adopted from11

2.4 The LMS Instrument

The Laser Mass Spectrometer (LMS) is a miniaturized reflectron type time-of- flight mass spectrometer. It was developed in 2003 at the University of Bern. A schematic drawing of the LMS instrument is shown in Figure 2.4. The LMS is a 60 mm x 160 mm sized instrument with a weight of about 1 . 5 kg. The analyzer is located within an ultra-high vacuum chamber (UHV) with a typical base pressure of around 10 8 mbar. The pressure is achieved by the use of a turbo-molecular as well as an ion getter pump. Once the pressure is as low as 10 7 mbar the turbo-molecular pump is switched off to prevent mechanical vibrations and electronic perturbations during the measurement procedure.

illustration not visible in this excerpt

Figure 2.4: Schematic drawing of the LMS in- strument. The laser beam is focused on the sam- ple where a plasma plume is formed. The ions from the plume then enter the mass analyzer through an ion optical system. After accelera- tion they travel through the field free drift tube and are reflected by the retarding field towards the MCP detector.

A smaller sized vacuum chamber is also attached to the main vacuum cham- ber as a sample introduc- tion port. It is used to speed up the sample exchange process. The sample can be introduced into the main chamber via a mechanical sample introduction system. The vacuum in the sample port is produced by an addi- tional turbo-molecular pump. Once inside the main cham- ber the sample is placed on a xyz translation stage al- lowing for a spatial accuracy of only a few micrometers. Ions are produced from the sample via focused, pulsed laser radiation. Currently a Ti-sapphire laser (wavelength 775 nm, pulse duration 190 fs, repetition rate 1kHz, maxi- mum output energy 1 mJ) is used for the ablation and ion- ization process. The laser beam is guided to the opti- cal port of the main vacuum chamber and enters the mass spectrometer where it is finally focused onto a ap- proximately 20 micrometer sized spot on the sample surface. The laser system is controlled remotely via the measurement computer. A trigger signal derived from the laser pulse starts the data acquisition after an appropriate delay. The ablated material forms a hot plasma plume consisting of atomized and ionized species. The close proximity of the plasma plume to the entrance of the ana- lyzer ensures the that the majority of the produced ions enter the mass analyzer through the conical extraction electrode. After the initial acceleration, focusing and collimation the ions enter the field-free drift tube and are finally reflected by the ion mirror towards a pair of MCPs arranged in chevron configuration, meaning the micro-channel plates tubes are rotated 180 from each other, pro- ducing a v-like shape reducing ion feedback in the device. At the output side of the multichannel detectors the stream of secondary electrons are collected on four concentric anode rings. The generated electric signal is then measured with two high-speed ADC (analogue-to-digital) cards. The acquisition cards typically sample for 20 microseconds after the ablation and ionization process, allowing for the detection of masses of up to 600 amu. The acquired spectra from the ADC cards are then stored on a measurement computer. Figure 2.5 shows a block diagram of the described measurement procedure.

The entire laboratory is kept at clean room conditions by a laminar flow ceiling to protect the laboratory equipment, in particular the delicate laser optics, and the samples from dust and other particles inside the laboratory. In the laboratory a constant over-pressure with respect to the outside environment is maintained. The temperature is kept at (22 . 0 ± 0 . 4) C and the humidity is controlled at (42 . 0 ± 0 . 5) % relative humidity level12.

illustration not visible in this excerpt

Figure 2.5: Block diagram of the measurement procedure. The laser system is controlled remotely via the measurement computer. The laser beam is guided towards the entrance window of the main vacuum chamber and then focused on the sample. The ions enter the mass analyzer and are guided towards the MCP detectors. A trigger signal starts the data acquisition with the ADC cards. The recorded spectra are then stored on the measurement computer and are ready to be analyzed.

Chapter 3 Measurement Procedure and Data Processing

3.1 Sample Preparation

One of the big advantages of the LMS instrument is that no special sample preparation is needed before the measurement. A raw sample is placed in suit- able holes on a steel holder and held in place with copper tape if needed (see Figure 3.1).

illustration not visible in this excerpt

Figure 3.1: Samples are placed on the steel holder. Diameter of the sample holder: 15 mm.

To avoid contamination, it is important that the sample is kept at clean room condi- tions and placed on the sample holder wear- ing gloves. Breathing on the sample has to be avoided (possible contamination are wa- ter, Na, K, Cl or organics for example). Mi- croscope pictures are taken beforehand to se- lect the areas of interest for the measure- ment. A micro-translation stage is transport- ing the sample against the laser spot inside the main chamber. At the time of this exper- iment the in-situ microscope camera system was not available yet. Consequently a coordi- nate system (relative to the center of the sam- ple holder) had to be build with the help of an external microscope measurement. Once the sample is inside the chamber, the center of the steel holder in stage coordinates can be found by scaling from the coordinates found by detecting mass spectrometrically the edges of the sample holder. Since the holder itself is circular the center can then be easily calculated and the derived coordinates from the external microscope can be transformed into stage coordi- nates. Most natural samples are inhomogeneous but a spatial resolution of the instrument of around 20 μ m allows for the investigation a few micrometer-sized objects. After the measurements for this thesis a microscope-camera system was installed inside the chamber, additionally increasing orientation on the sample.

3.2 Signal Optimization

Inside the chamber, the sample is placed about 1 mm below the entrance plate of the LMS, to be roughly in the focal point of the laser. The distance from the sample to the entrance plate is then optimized. In the ideal case the focal point should lie slightly under the sample surface to guarantee the accumulation of as many spectra as possible at one vertical position before the ablation surface is moving out of the focal point. Usually around 60’000 to 80’000 single laser shot spectra can be accumulated per position before the peak intensities disappear. This large number of spectra is needed when investigating trace elements in a sample. By accumulating a large number of individual mass spectra the statistical inaccuracies are reduces and the effective signal-to-noise (S/N) ratio is increased.

Mass resolution, peak intensity and peak shapes depend highly on the choice of voltage settings for the ion optics. An optimal choice of voltages is needed before starting the measurement campaign. This voltages can be adjusted manually or with the help of an optimizer software. The procedure is described in detail in13 and14. This procedure is applied in the current investigations and usually yields mass resolutions of 500 to 700 for the56 Fe peak (m/ Δ m at FWHM). Figure 3.2 shows an example of the optimization process. The voltages were optimized for zirconium in this case.

illustration not visible in this excerpt

(a) This voltage set is unsuitable for the given sample, causing broad mass peaks with consequently bad resolution.

(b) A good mass resolution of around 420 is obtained with this voltage set. However, the peaks show a slight asymmetry.

(c) Ideal voltage set for the given sample. The mass resolution is around 470. The peaks are almost symmetrical with a high a intensity.

Figure 3.2: Example of three different voltage sets on the zirconium isotopes. The pictures (a),(b) and (c) show the intensity and shape of the zirconium isotopes in dependence of the the voltage set. The measured intensity per bin number (flight time) is shown in yellow and the fit of the optimizer in blue.

As the quantitative measurements require sufficiently high laser fluence (ir- radiance) to measure effectively all available elements, the final laser fluence levels are adjusted to ensure the quantitative detection of all elements of inter- est at the desired mass resolution and to avoid the formation of clusters and molecules (e.g. oxides). After the signal optimization on relevant mass peaks, the measurement can be started. The laser can be operated in two different modes, standard mode, applying pulses at a rate of 1 kHz, or in group mode, where a adjustable number of spectra (typically 100) are accumulated in one file and a number of such files (typically around 400) are accumulated for each position. A raster software can be used to measure several defined positions automatically in one measurement campaign.

Two or three detector channels with different gain factors can be used for the data acquisition. This allows for the detection of the main elements all the way to trace elements at the same time. Each channel is connected to a different anode ring of the detector. With this method, the mass spectrum in high dynamic range can be measured by carefully combining the low gain and high gain spectra. The mass peaks of low abundant elements can be observed in the high gain mode at the expense of the saturation of the mass peaks of the most abundant elements. These peaks are discarded while combining appropriately the high gain with the low gain spectra16.

3.3 Data Analysis

After acquisition, the data are analyzed with the help of an analyzing soft- ware. The software provides automatic data read-in, mass scale calibration, background subtraction and integration options among some other features (see15 and16 for details). Background subtraction is of utter importance when analyzing trace elements. The most effective way is to record the background directly before or after the measurement with the exact same instrument set- tings. Additional filtering of the mass spectrum might also be effective in some cases. In the measurements carried out in the frame of this project the use of a Fourier-filter proved to be useful. Figure 3.3 illustrates the differences between the original data, the data after the background subtraction and the filtered data. After successful background subtraction and filtering, the area of each peak was integrated with the integration software (see15 for further information).

illustration not visible in this excerpt

(a) Mass spectrum of the hafnium isotopes without any background subtraction or filtering.

(b) Mass spectrum of the hafnium isotopes with background subtraction.

(c) Mass spectrum of the hafnium isotopes with background subtraction and filtering.

Figure 3.3: Example of a part of the same recorded spectrum, without background subtraction (a), with background subtraction (b) and with background subtraction and filtering (c).

Chapter 4 Measurements on a Lunar Meteorite

4.1 Fundamentals of the Sayh al Uhaymir 169 Meteorite

The lunar meteorite Sayh al Uhaymir 169 is chosen for the investigation by laser ablation/ionization mass spectrometry. A picture of the whole meteorite is shown in Figure 4.1.

illustration not visible in this excerpt

Figure 4.1: Say al Uhaymir 169, found in the desert of Oman (photo credit: Beda Hofmann).

Lunar materials are in- teresting test materials for the performance evaluation of the LMS instrument. There are currently two missions to the moon in prepara- tion by the laser mass spec- trometry group, LUNA RE- SOURCE and LUNA GLOB. A LASMA instrument, simi- lar to the LMS system, will be carried by both of these mis- sions to measure the chemical composition of the lunar re-golith.

The goal of the current LMS studies are to serve as proof that the LMS instrument can be used to investigate lunar mineralogy, detect heavy trace elements such as rare earth elements (REE) and is suitable to perform radiometric dating by the detec- tion of uranium, thorium and lead. Hence, the current studies will provide information on the present capabilities of the instrument and form a basis for conducting similar investigations on the lunar surface. The Sayh al Uhaymir 169 meteorite was found in 2002 in Oman by Edwin Gnos, Beda A. Hofmann and Ali Al-Kathiri. The rock originally weighted 206.45 g. It consists of two different lithologies. A holocrystalline, fine-grained polymict impact-melt brec- cia makes up around 87 volume percent of the rock. This part of the meteorite is extremely enriched in potassium, rare earth elements and phosphorus (also commonly referred to as K-REE-P material), as well as the two radioactive ele- ments thorium and uranium. Sayh al Uhaymir has been reported to be the most K-REE-P rich lunar rock that has ever been found. The remaining 13 percent is a shock-lithified regolith breccia. The minerals present in the impact-melt part of the meteorite are mainly low-Ca pyroxene, plagioclase and potassium feldspar. Also present are ilmenite, whitlockite, olivine, zircon and traces of troilite, kamacite and tridymite. Table 4.1 provides an overview of the minerals present in the meteorite and their empirical formula.

Table 4.1: Empirical formula of the minerals found in the impact-melted breccia of the SaU 169 Meteorite[17].

illustration not visible in this excerpt

The crystallization age of 12 different zircon grains was determined by E. Gnos et al. (2004)[6]. The measurements yielded a weighted average age of (3909 ± 13) · 106 a. This together with the K-REE-P-rich nature of the meteorite lead to propose the origin of the Sayh al Uhaymir rock from the Lalande crater area. Three small samples, containing parts of both the impact-melted breccia as well as the regolith breccia, lithologies were provided for the investigation with the LMS instrument by Prof. Beda A. Hofmann. Microscope images of the three samples investigated by LMS are shown in Figure 4.2. In the following sections the results from the investigation of the mineralogical and elemental composition of the lunar samples will be presented for the main, highly abundant elements as well as for trace elements. An overview of the chemical composition of the surface was performed by the energy-dispersive X-ray spectroscopy and is utilized here as a guiding study into the spatial distribution of zirconium on the sample surface.

illustration not visible in this excerpt

(a) Sample 1: Impact-melted breccia

(b) Sample 2: Boarder area between the two lithologies

(c) Sample 3: Regolith breccia

Figure 4.2: Three SaU 169 samples. Sample 1 (a) is a part of the K-REE-P rich lithology, Sample 2 (b) shows the boarder area between the impact-melt and the regolith breccia, and Sample 3 (c) is a part of the regolith portion of Sayh al Uhaymir.

4.2 Energy-Dispersive X-Ray Spectroscopy

The three SaU 169 samples were investigated with energy-dispersive X-ray spec- troscopy (EDX) prior to the main measurements. This was done mainly to iden- tify areas rich in zirconium. Zirconium is present in higher abundances only in the zircon mineral, which will be the main focus of the following mass spectro- metric measurements. Energy-dispersive X-ray spectroscopy is a technique used for rapid and a fairly easy identification of the elements present on a given area. EDX is usually coupled with a scanning electron microscope (SEM). During the EDX measurements the investigated sample material is bombarded with a high energy electron beam. These electrons interact with the atoms of the material, eventually knocking out electrons of the sample atoms. When an electron from an inner shell is ejected, an electron from a higher binding energy level will quickly fill the core hole, generating an X-ray with the energy corresponding to the energy difference between the two shells. The energy of this emitted characteristic X-ray will generally be different for different elements. When an X-ray energy spectrum is is recorded, it allows the identification of the present elements18. An example of such a spectrum is given is Figure 4.3.

illustration not visible in this excerpt

Figure 4.3: X-ray energy spectra for phosphorus (pt1) and zirconium (pt2). Because of the overlap at ~ 2 keV the K α 1 , 2 emission line at 15.7 keV has to be used to distinguish these two elements.

It can be seen that the K α 1 , 2 (emission line from an electron falling from the L to the K shell) of phosphorus and the L α (from the M to the L shell) emission line of zirconium are overlapping. The K α 1 , 2of zirconium helps to differentiate between the two elements. The sensitivity of EDX measurements is rather low and usually does not exceed 1000 ppm.

Several zircon grains of different sizes could be identified during the EDX measurements. With the help of SEM pictures the exact positions of these grains could be located on the sample and used for the investigations during the measurement campaigns with the LMS instrument. One such grain is shown in Figure 4.4.

illustration not visible in this excerpt

(a) Zirconium rich area found during EDX mapping.

(b) SEM picture of the grain to help localize it on the sample.

Figure 4.4: Zircon grain, identified during the EDX measurements (a) and a SEM image of the same grain in (b)(b is rotated by around 90 relative to a).

4.3 Determination of the Relative Sensitivity Coefficients and Chemical Mapping of an Unknown Area

4.3.1 Introduction

The stoichiometric ion production for a given sample by laser ablation/ionization source is typically difficult. Laser-matter interaction depends on the chemical and physical properties of the material. When ions are formed by the use of a laser ablation/ionization ion source, the ablation and ionization efficiency varies for each element (the process of laser ablation/ionization is further explained in Chapter 2.1). This is caused by different properties such as absorption and re- flectivity of the material, sublimation temperature, material porosity and chem- ical composition. There are significant differences in the efficiency of plasma formation depending on whether the material is an insulator, semiconductor or metal. Additionally, material structure, such as crystal and polymorphic, plays an important role in the ablation process. Nevertheless, once a sufficiently high plasma temperature (approximately 40 000 K) could be achieved, the stoichio- metric production of ions can be conducted. The process of ion formation also differs greatly between the use of a fs-laser or a ns-laser. The mechanisms of the plasma and ion formation are discussed in the literature (e.g.[31]). To achieve relevant conditions for stoichiometric ion production a careful control of the laser source parameters (laser fluence, pulse duration etc.) and laser optical system (laser beam focusing optics) delivering the laser radiation at the sample loca- tion have to be maintained to achieve sufficiently high irradiance (laser power densities at the laser focal point). Consequently, because of this multiparamet- ric dependance, one needs to prepare relevant calibration to the measurement valid at the applied conditions (laser ion source parameters, ion optical settings) otherwise the composition derived directly from the mass spectrum may not be an accurate representation of the composition of the investigated material. The LMS instrument was so far tested with a number of NIST standards and the correction factors, relative sensitivity coefficients (RSCs), are known to some extent for the most abundant elements[19]. To derive the quantitative results, the data have to be calibrated with RSCs. The RSC of an element is defined as:

illustration not visible in this excerpt

In the most ideal case the RSC value of an element would be equal to one for all samples. This is however most often not the case. RSCs have to be determined for all elements of interest for each sample. This can be done by the use of standard reference materials of known composition. However, the calibration with these artificial samples may be of little use when investigating a natural rocky material. It is advisable to use a sample of a similar texture and composition than the one of interest to determine the sensitivity coefficients. If the bulk composition of a sample is given by the reference, the RSCs can then be calculated by measuring several positions on the sample, preferably on a homogeneous area. With those measurements the RSC of an element X is given by[19]:

illustration not visible in this excerpt

where n is the number of positions which contained the element X,[illustration not visible in this excerpt] thepeak area of the element for the[illustration not visible in this excerpt] position,[illustration not visible in this excerpt] the total area of all the elements in the spectra,[illustration not visible in this excerpt] the expected atomic fraction of the element and finally[illustration not visible in this excerpt] which corresponds to the total fraction of all the measured elements according to the reference. It is clear, from a statistical point of view, that a sufficient amount of measurement positions on the sample is needed to establish a reliable set of RSCs. Once the relative sensitivity coefficients have been determined, quantitative measurements can be conducted.

As it was mentioned before, the K-REE-P meteorite that was analyzed in this thesis consists of two different lithologies. The regolith breccia part, adjacent to the more interesting K-REE-P breccia, can be used to obtain a set of RSCs, the composition of the regolith has been determined by Gnos et al. (2004)[6].

4.3.2 Measurements and Results

To calculate the RSCs a mostly homogeneous region on the regolith breccia was chosen and a total of 225 positions were measured on this area. The settings used in this experiment are given in Table 4.2. A raster software was used for measuring the positions, covering a total area of 450 μ m2. The voltages were optimized following the standard procedure described in Chapter 3. In Figure

4.5 the chosen area is shown before and after the measurements.

Table 4.2: Instrument settings for the RSC measurement on the regolith part (see Chapter 3 for explanation of the measurement procedure).

illustration not visible in this excerpt

illustration not visible in this excerpt

(a) Overview of the selected area prior to the measurement. The red square indicates the exact position of the measurement area.

(b) Zoom view of the red area in (a) after the measurements.

Figure 4.5: Chosen area on the regolith sample, before (a) and after (b) the measurements.

At five positions only mass spectra with a poor quality (isotopes could no longer be distinguished) were measured, possibly due to surface charging, and hence were not used for the RSC calculation. The element composition was determined from the remaining 220 positions and the RSC values were calculated according to Equation 4.2. The reference table with the literature values of the element abundances from6 is provided in the Appendix. Table 4.3 gives the calculated sensitivity coefficients for the main rock forming elements and some trace elements. The error given with each value results from the peak fitting process. No uncertainties for the reference values have been given.

Table 4.3: Calculated RSC values for the main rock forming elements and some trace elements. Used reference values from6 are provided in the appendix.

illustration not visible in this excerpt

Additionally, the element composition at each position was plotted as a map at the end of this section (Figure 4.9). This helps to understand the distribution of elements over the sample surface and provides an insight into the homogene- ity of the investigated area. The maps show that the material consists mostly of aluminum, calcium, magnesium, silicon and iron, i.e., the main rock-forming el- ements. No large mineral clasts are observed in the investigated area. However, the area is still inhomogeneous at smaller scales and consists of micrometer-sized grains. This is of course expected since perfect homogeneity is very unlikely in natural samples. Overall the area is homogeneous enough to yield a reliable set of RSCs.

Afterwards, an area of unknown element composition was measured and the mineralogical composition was derived from the element correlation. The RSC values that were determined from the regolith sample were used to calibrate the data. This serves as a direct test of the calculated RSCs. Since the minerals are defined by certain element ratios inappropriate RSC values would cause difficulties in the identification of the minerals. The chosen area for this test is shown in Figure 4.6. The settings for the measurements were the same as in

Table 4.2 with the exception of the voltage set. The instrument was slightly optimized to conduct the measurements with a sufficiently high mass resolution in the chosen area. The used voltage set, Kreep Sample2 2016 4.opt, is given in the Appendix.

illustration not visible in this excerpt

(a) Measurements on sample 2 (boarder between regolith and breccia part). The red square indicates the area where the measurements will be conducted.

(b) Zoom of the red area indicated in picture (a) after the measurements.

Figure 4.6: Chosen area for the mineral analysis, before (a) and after (b) the measurements.

illustration not visible in this excerpt

Figure 4.7: Mineralogy of the regolith measurement on the sample (see Fig. 4.6). The presence of the most common minerals can be easily noted with the help of this ternary diagram. The ratios are chosen to avoid overlapping of minerals on the diagram as much as possible.

The chemical composition of the surface was plotted on maps for all the main elements and some trace elements. This is displayed in Figure 4.10 at the end of this section. The area appears to be of one large homogeneous mineral rich in magnesium and silicon. Magnesium and iron cations in the mineral have the same charge (2+) and a similar atomic radius. They can therefore substitute for one another quite easily. This can be seen by the anti-correlation of these two elements in Figure 4.10 (c) and (k). A quick and easy way to investigate the mineral composition in a sample area is by the use of ternary diagrams.

In a ternary diagram three different elemental ratios are plotted against each other. The sum of the ratios is normalized to 100 percent. Such a diagram is shown in Figure 4.7. The ratios were chosen to differentiate between the most common minerals present on the Moon. The red dots in the figure represent the theoretical ratios of the different minerals, which can be directly derived from the empirical formula of the minerals (given in Table 4.1) with a shaded area around them to represent an uncertainty of ± 10 percent for all three ratios, since in reality the element composition of minerals can vary to a certain degree due to substitution of different elements.

illustration not visible in this excerpt

Figure 4.8: Ratio of (Mg+Fe) to Si is plotted. A ratio of 2 indicates the presence of olivine. Also small inclusions of tridymite are likely (ratio is close to 0).

Figure 4.7 shows that all except one positions are lying in the orthopyroxene/olivine corner. Both these minerals are rich in silicon and magnesium. Further analysis, by directly looking at the ratio of magnesium and iron to silicon (2 for olivine and 1 for orthopyroxenes) plotted in Figure 4.8, reveals that it is indeed an olivine grain with some small inclusions of tridymite (this was not labeled on the ternary plot since its only a minor mineral in lunar meteorites). The reddish color of the spot is most likely due to oxidation of the iron when the meteorite was already on the Earth.

4.3.3 Conclusion

Measurements on the regolith matrix yielded a robust set of RSC values for all the main rock forming elements as well as for some trace elements. A mineralog- ical analysis on an olivine rich area further supports the accuracy of the calcu- lated sensitivity coefficients. These RSCs are necessary to conduct quantitative measurements on the K-REE-P rich part of the Sayh al Uhaymir meteorite.

4.3.4 Element Composition Maps-Regolith Sample

illustration not visible in this excerpt

(a) Aluminium (b) Calcium (c) Magnesium

(d) Silicon (e) Titanium (f) Potassium

(g) Zirconium (h) Barium (i) Sodium

(j) Chromium (k) Iron (l) Oygen

Figure 4.9: Elemental composition maps for the regolith sample (see Fig. 4.5).

illustration not visible in this excerpt

4.3.5 Element Composition Maps-2nd Sample

illustration not visible in this excerpt

(a) Aluminium (b) Calcium (c) Magnesium

(d) Silicon (e) Titanium (f) Potassium

(g) Oxygen (h) Strontium (i) Sodium

(j) Chromium (k) Iron (l) Phosphorus

Figure 4.10: Element composition maps for an area on sample two (boarder between regolith and impact-melted breccia (see Fig. 4.6).

4.4 Chemical Composition Measurements on the K-REE-P Part of the SaU 169 Meteorite

4.4.1 Introduction

The Sayh al Uhaymir meteorite consists of two different lithologies (Section 4.1). The main focus of this campaign is to perform a chemical analysis of the impact- melted breccia. With the first set of measurements, described in Section 4.3, a set of RSCs for all the main rock forming elements and some trace elements was obtained. The second part of this measurements focuses on the identification of minerals on the K-REE-P-rich part of the meteorite. Areas rich in zirconium were identified prior to this measurement with the help of EDX spectroscopy (see Section 4.2). Zirconium is present in high abundances in the mineral zircon (ZrSiO4). Due to its chemical durability and its capability to incorporate many trace elements, especially the radioactive uranium and thorium, zircon is of considerable interest for the investigation of material dating22.

4.4.2 Measurements and Results

4.4.2.1 Measurement procedure

For the compositional and mineralogical analysis of the K-REE-P rock, a raster campaign with a total of 225 positions, covering an area of 450 μ m2 was conducted. The area of interest was chosen to include parts of a zircon grain. A microscope picture of the area before and after the measurements is shown in Figure 4.11. The measurement conditions are shown in Table 4.4.

Table 4.4: Instrument settings for the measurement on and around a zircon grain. Further explanation of the measurement procedure is given in Chapter 3.

illustration not visible in this excerpt

illustration not visible in this excerpt

(a) Measurement area prior to the measure- ment (indicated through the red square). The zircon grain can be identified by its milky white color.

(b) Same area as the red box in picture

(a) after the raster campaign. The raster covered parts of the zircon grain as well as its surrounding environment.

Figure 4.11: Raster campaign around a zircon mineral grain before (a) and after (b) the measurements.

The voltages were optimized for optimal measurements of the zirconium mass peaks. The quality of the spectra were secured by recording a few mass spectra before starting the raster campaign. Figure 4.12 shows a typical mass spectrum of zirconium isotopes. Also the relative isotope abundances derived from the spectra are compared with the standard isotope abundances[21]. Zirconium has five stable isotopes with masses 90, 91, 92, 94 and 96 amu. Their terrestrial abundances are:90 Zr: 51.45%,91 Zr: 11.22 %,92 Zr: 17.15%,94 Zr: 17.38% and96 Zr: 2.8%[21]. All five isotopes of zirconium are readily measured and their isotope ratios are accurate at the percent level. The mass resolution of the mass spectra is determined to be larger than 400 (m/ Δ m). Some fraction of the zirconium is present at its oxide state ZrO. This will have to be accounted for when calculating the total atomic fraction of zirconium in the chemical composition of the sample.

illustration not visible in this excerpt

(a) Mass spectrum of zirconium with all 5 isotopes (90 Zr,91 Zr,92 Zr,94 Zr and96 Zr ) showing. Also present are the oxides of zirconium and yttrium at mass 89.

(b) Measured isotope abundance (yellow) and their terrestrial abundance (green).

Figure 4.12: Mass spectrum for zirconium in (a) and isotope abundances in (b). 38

4.4.2.2 Mineralogical Analysis

The element abundances were calculated for each position. The same relative sensitivity coefficients that were determined on the regolith sample were used to scale the abundances. The chemical composition at each measured location is again shown in an element composition map in Figure 4.19 at the end of this section. In the maps the areas of relevant minerals with sharp mineral boarders are readily identified and help identifying element distribution over the sample area. The minerals present in the investigated area can again be visualized with the help of a ternary diagram. Figure 4.13 displays results of the elements ratio correlations on a ternary diagram. With the same element ratios as in previous section one can readily identify the most common lunar minerals. The ternary diagram shows also that the area includes mainly zircon, pyroxene (ortho- and clinopyroxene), plagioclase (anorithe), potassium-feldspar as well as some ilmenite. Most of the points displayed in the diagram represent the ratio values intermediate between the ratios characteristic for particular mineral. The grain sizes of the minerals in the sample are generally smaller than the diameter of the laser spot, thus a measurement contains more than one mineral. The chemical composition of most positions will therefore contain a mixture of different mineral grains. Conversely, the detection of a single mineral grain can be made if the grain size exceeds the size of the laser spot. This was possible only for a large zircon grain in the present studies.

illustration not visible in this excerpt

Figure 4.13: Ternary diagram of an area on a K-REE-P rich sample. The diagram shows the presence of zircon, ilmenite, pyroxenes, K-feldspar and plagioclase (anorithe).

The presence of whitlockite, Ca9(Mg,Fe)1(PO4)(PO3OH), is suggested by the amount of phosphorus measured in the investigated area. However, none of the positions show the correct ratio. It is most likely that whitlockite is part of the matrix material but the grain size of the mineral is very small, such that it only shows up mixed up with other minerals. To distinguish between or- thopyroxene and olivine, the magnesium and iron to silicon ratio can be plotted directly. This ratio will be roughly equal to one in the case of orthopyroxene and two in the case of olivine. This was done in Figure 4.14. It can be seen that both, orthopyroxene and olivine are present in the area.

illustration not visible in this excerpt

Figure 4.14: Ratio of magnesium and iron to silicon. A ratio of 2 identifies olivine. Orthopyroxene has a ratio of 1.

illustration not visible in this excerpt

Figures 4.15 and 4.16 show two recorded mass spectra. One of them in an ilmenite/pyroxene region and one of them inside the zircon grain.

illustration not visible in this excerpt

Figure 4.15: Spectrum (HG) in a region with mainly ilmenite and pyroxene mineral grains. The position is rich in magnesium, silicon, calcium, titanium and iron.

illustration not visible in this excerpt

Figure 4.16: Spectrum (HG) inside the zircon grain. The position contains mostly zirconium and silicon with minor magnesium, calcium and potassium.

4.4.2.3 Rare Earth Elements

A variety of trace elements could be measured during this campaign, most of them were localized inside the zircon grain. Hafnium occurs exclusively in the zircon mineral with an abundance of around 300 to 400 ppm. The distribution of hafnium is shown in Figure 4.19 (m). Hafnium commonly substitutes for zirconium due to the similarity of the electron structures of both elements, giving them similar chemical and physical properties. The investigated zircon grain proved to be rich in rare earth elements and yttrium (sometimes counted as a heavy rare earth element). The abundance map of yttrium is shown in picture (n) Figure 4.19. The rare earth elements are commonly divided into light rare earth elements (lanthanum (La), cerium (Ce), praseodymium (Pr), neodymium (Nd), promethium (Pm), samarium (Sm), europium (Eu), gadolinium (Gd) and sometime scandium (Sc)) and heavy rare earth elements (yttrium (Y), terbium (Tb), dysprosium (Dy), holmium (Ho), erbium (Er), thulium (Tm), ytterbium (Yb) and lutetium (Lu)). Due to their chemical similarities they usually all appear together as trace elements in different minerals, most notably in zircon. The pattern of the REE varies depending on the minerals in which they are encountered, with some minerals preferring light- and other heavy rare earth elements. A map showing the abundance of REE elements has been plotted in Figure 4.17.

illustration not visible in this excerpt

Figure 4.17: Abundance map of rare earth elements.

The REE are present inside the zircon grain at almost every location, they can however occasionally be present in other minerals, such as pyroxene. It has to be noted that the map shows the abundance of the total amount of REE, without distinguishing between the individual elements since not all isotopes were nicely resolved from one another, making integration over individual peaks difficult. Additionally, no RSC values are known for these elements from our previous studies so that their actual abundance is uncertain. However, due to their similarities in physical and chemical properties, the sensitivity coefficients are most likely going to be very similar for all the rare earth elements. The abundance distribution pattern of these elements will therefore be likely as it is measured. By comparing the rare earth element pattern measured inside the zircon grain with the one measured inside the pyroxene mineral, it becomes clear that heavy rare earth elements are strongly favored in zircon, with yttrium, ytterbium and erbium having the biggest abundance. The lighter isotopes seem to be below the detection limit. The situation is much different inside the pyroxene, with lanthium and cerium being most abundant. Unfortunately many of the elements are also present as oxides, partially overlapping in mass with other elements.

illustration not visible in this excerpt

(a) REE pattern in the zircon grain, ytterbium, erbium and lutetium are present. The lighter REE are barely visible. Hafnium oxide is also visible.

(b) REE pattern inside a pyroxene grain. Lanthanum and cerium are most abundant. Also present are praseodymium, neodymium and samarium. The heavier rare earth elements are not detected.

Figure 4.18: REE pattern comparison. Rare earth elements measured inside zircon (a) and inside pyroxene (b)

4.4.3 Conclusion

The LMS instrument is a powerful tool for analyzing the chemical composition of a given area. Knowing the relative concentration of the main rock build- ing elements allows for the identification of the minerals present in the sample. With the help of element ratios the minerals zircon, ilmenite, pyroxene, olivine, K-feldspar and whitlockite have been identified in the measured area. Zircon and ilmenite show visible grain boundaries whereas the other minerals make up the matrix around these grains. The K-REE-P sample shows a high inhomo- geneity on very small spatial scales, which could be visualized by the conducted measurements. Trace elements like hafnium and rare earth elements have been measured for the first time with the LMS instrument. Analysis of their distri- bution pattern might help to gain insight in the formation conditions of various minerals. Presence of the oxide mass peaks make the analysis of these patterns frequently difficult. Unfortunately, no lead was measured during the measure- ment campaign. It is likely that a higher detection sensitivity will be needed to conduct the measurements of lead, thorium and uranium as their abundances are expected to be at ppm levels according to the literature reports6.

4.4.4 Element Composition Maps of the K-REE-P Sample

illustration not visible in this excerpt

(a) Zirconium (b) Silicon (c) Oxygen

(d) Aluminium (e) Calcium (f) Potassium

(g) Titanium (h) Sodium (i) Magnesium

(j) Iron (k) Sulfur (l) Chromium

(m) Hafnium (n) Yttrium (o) Phosphorus

Figure 4.19: Element composition maps for the K-REE-P rich sample of SaU 169 (see Fig. 4.11).

4.5 Crystallization Temperature of Zircon

4.5.1 Introduction

The mineral zircon (ZrSiO4) is unique because of its chemical and physical durability. Additionally, it also serves as a host for a variety of trace elements, such as uranium and thorium, commonly used for various dating methods, but also non-radioactive elements such as hafnium, yttrium and rare earth elements as well as titanium.

From the amount of titanium in zircon, the apparent crystallization temperature of the mineral can be calculated as first described by Watson in 200622. The substitution of titanium for either zirconium or silicon shows a loglinear dependence of the temperature, with the amount of titanium increasing for higher temperatures. Knowing the crystallization temperature of a zircon grain can shed light on the origin of the grain, as zircons are commonly hosted by rocks other than those in which they originally crystallized.

The first zircon thermometer by Watson was calibrated for the use on zircons formed in the presence of the mineral rutile (TiO2) with limited accuracy for zircons that crystallized alongside other titanium containing minerals, such as ilmenite (FeTiO4). It was later revised by Ferry and Watson in 2007 to be better applicable for rocks without rutile[23]. The temperature dependence according to Ferry and Watson is given by:

illustration not visible in this excerpt

where the titanium content is given in weight percent. The two values[illustration not visible in this excerpt] and[illustration not visible in this excerpt] are corrections for reduced oxide activities (chemical potential) of silicon- and titaniumoxide. They depend solely on the composition of the parent magma during crystallization. In the case of zircon coexisting with rutile both[illustration not visible in this excerpt] and[illustration not visible in this excerpt] are set to 1. For any other kind of rock they have to be indepen- dently determined. If no such measurement is possible the zircon thermometer can still be applied using the standard conditions [illustration not visible in this excerpt] and[illustration not visible in this excerpt] are set to 1) and the maximum additional uncertainties due to the undefined activities can be estimated. The calculated temperature will then be a slight underestimation of the real crystallization temperature by usually less then 80 C[23].

Additionally, the incorporation of other trace elements in zircon, such as Hf, U and Th, has to be included in the calculation of the crystallization tempera- ture. Its dependence is given by the[illustration not visible in this excerpt] term in Equation 4.4. This effect can be quantitatively corrected by calculating the mole fraction[illustration not visible in this excerpt] of the ZrSiO4 component of zircon. By the use of Raoults’s law1 [24] it follows that:

illustration not visible in this excerpt

4.5.2 Measurements and Results

In this measurement the titanium content of a single zircon grain located on the impact-melted breccia of SaU 169 was measured at 56 different positions inside the grain. To avoid contributions from nearby minerals only positions away from the grain boundary were used that showed a molecular fraction of ZrSiO4 of no less than 95 percent. The instrument settings are given in Table 4.5. The RSC values used to calibrate the obtained data are the same that have been determined in earlier measurements (see Table 4.3 in Chapter 4.3).

Table 4.5: Instrument settings for the measurement of the titanium content in a zircon grain.

illustration not visible in this excerpt

Since the exact oxide activities a SiO 2 and a TiO 2 are unknown for the given sample, their value was set to one. With this an average crystallization tem- perature of 903 C ± 135 C (1 σ) was calculated according to Equation 4.4. The amount of titanium varied within the grain without showing any consistent pat- tern. The results are visualized in the histogram plot shown in Figure 4.20. As mentioned above, this temperature represents the minimal average crystalliza- tion temperature as no correction for reduced oxide activity could be made. The presence of ilmenite would suggest a a TiO 2 of around 0 . 6. This would increase the crystallization temperature by 69 C. However, a silicon oxide activity of less than one would again lower the temperature estimate. The effects would therefore partially cancel each other out.

illustration not visible in this excerpt

Figure 4.20: Histogram of the crystallization temperature derived from the mea- sured titanium content in a zircon grain (see Equation 4.4). The determined temperature follows a normal distribution with μ = 903 C and σ = 135 C.

4.5.3 Conclusion

As it was mentioned earlier zircon grains are extremely stable and they are often found in rocks others than the ones they have originally crystallized in. Zircons are also commonly used for dating methods involving the radioactive elements uranium, thorium and their decay products. The knowledge of the origin of a zircon grain, weather it crystallized during the impact event or was in fact inherited from a different rock, is of utter importance when interpreting the measured age of the grain. A zircon grain that was formed during equilibrium conditions of an impact melt is usually identified by its poikilitic structure (a structure where smaller minerals are embedded inside the main mineral grain). This method has however been criticized in recent years, as it might not be a sufficient criterion for the origin of a grain. Measuring the crystallization tem- perature of a grain might be a better way of identifying an impact-melt grown grain. The crystallization temperature for zircons grown during an impact event on the Moon was modeled by Wielicki et al. in 201226 to lie between 850 C to 1050 C. According to this model the measured temperature of 903 C ± 135 C lies in the predicted range for impact events on the Moon, justifying the use of this grain for dating the impact event. However, a better understanding of how shock and heating during an impact event affects zircon crystallization is needed.

Chapter 5 Measurements on a K-REE-P Sample with a HV Pulser: Preliminary Results

5.1 Working Principles of the HV Pulser

It has been observed that the detection sensitivity suffers greatly from saturation effects on the detector. This problem is particularly noticeable when studying low abundant elements with higher masses. The situation can be significantly improved by preventing high-abundant species at lower masses to reach the detector by the use of a high-voltage (HV) pulser. In the implementation for LMS the HV pulser generates a temporary pulse on the lens electrode, stopping light elements from reaching the detector and therefore increasing detection sensitivity for heavier elements29. The measurements in this chapter will focus on the detection of heavy elements and are all conducted with the help of the HV pulser.

5.2 Measuring Trace Elements by using the HV Pulser

5.2.1 Instrument Settings

The length of the generated pulse can be manually adjusted (pulse delay setting). The delay length was chosen to remove all light elements up to iron. Afterwards the signal was again optimized on the zirconium isotopes. The exact settings are shown in Table 5.1. The laser fluence was slightly higher than in the previous measurements in an attempt to avoid cluster and molecule formation.

Table 5.1: Instrument settings for the pulser measurements (see Chapter 3 for further explanations).

illustration not visible in this excerpt

The measurements were conducted on two different zircon grains, one of them being the same one shown in Chapter 4 in Figure 4.11 and the other one a more round looking grain located on the impacted-melted breccia. The second grain is shown in Figure 5.1.

illustration not visible in this excerpt

(a) Zircon grain before the measurements.

(b) Zircon grain after the measurements

Figure 5.1: Second zircon grain in the impact-melted breccia. The grain is shown before (a) and after the measurements (b).

5.2.2 Measurements and Results

As expected the use of the HV pulser coupled with the LMS instrument sig- nificantly improved the detection sensitivity for heavy elements. Rare earth element pattern show a much higher signal to noise ratio than in previous mea- surements (see Chapter 4 for comparison). The individual peaks of the elements could now be resolved (mass resolution around 400). Figures 5.2 and 5.3 show two recorded mass spectra displaying different rare earth element patterns. The first one was again recorded in a pyroxene rich area, clearly richer in the light REE, while the second spectrum was recorded inside the zircon grain, therefore being rich in hafnium and preferring heavy REE such as ytterbium and erbium. Additionally, tungsten with its isotopes was observed in the spectrum from the zircon grain. Some oxides are still present in both spectra, even though the laser fluence was higher this time.

illustration not visible in this excerpt

Figure 5.2: Rare earth element pattern inside a pyroxene grain, recorded during the pulser campaign.

illustration not visible in this excerpt

Figure 5.3: Hafnium and heavy rare earth elements inside a zircon grain.

During the measurement campaign with the HV pulser uranium and thorium were observed for the first time inside the zircon grain (shown in Figure 5.4). These elements are of particular interest because they offer the possibility of radiometric dating. The principles of radiometric dating are shortly described in the next section.

illustration not visible in this excerpt

Figure 5.4: Uranium-238 and thorium-232 inside the zircon grain. The two other uranium isotopes,234 U and235 U are not observed due to their low abundance.

5.3 Radiometric Dating Methods

5.3.1 Theoretical Principles

1 Radiometric dating is based on the measurements of isotope ratios between radioactive elements and their daughter nuclides. For dating methods to work the system has to remained closed, meaning no parent or daughter nuclides have been gained or lost through an exchange with the environment. The principle of concordance or discordance is used to check whether a measured age is reliable or not. If the age of a rock is determined by two different geological clocks and they both yield a result close together, the age will be considered concordant and geologically significant.

5.3.1.1 Uranium-lead systems

All dating methods based on the uranium-lead decay chain use the fact that[illustration not visible in this excerpt] and[illustration not visible in this excerpt] have two very different decay constants. The basic principles of the different methods shall be illustrated for the mineral zircon. Zircon is an uranium rich system with little to no non-radiogenic lead. The mineral is also generally considered a closed system for both uranium and the generated lead. With the two decay chains[illustration not visible in this excerpt] and[illustration not visible in this excerpt] two ages can be calculated:

illustration not visible in this excerpt

The plotted curve with[illustration not visible in this excerpt] is known as the concordia curve. A concordant age will lie on the curve and a discordant age will lie off the curve. The concordia method can be extended to other pairs such as uranium/lead-thorium/lead, rubidium/strontium-potassium/argon.

The lead-lead method is a reliable dating method when the two isotopes[illustration not visible in this excerpt] and[illustration not visible in this excerpt] are of purely radiogenic origin. From Equation 5.1 follows that:

illustration not visible in this excerpt

In the case of a closed system, as for instance for zirconium, the concordia curve will remain a straight line and the initial age can be determined by measuring its slope. Lead-lead dating would be the preferred option for dating measurements with the LMS instrument since the ratio of the two isotopes is independent of any RSCs.

5.3.2 Measurement and Results

Unfortunately no lead was observed during the pulser campaign. There are several possible explanations. Firstly, it was observed later on that resonances from the generated HV pulse were affecting the mass spectra. The lead might have been accidentally removed from the spectra. Secondly, it is possible that the lead is present at only specific locations in the zircon and was missed each time. Lastly, the lead content inside the zircon might be lower than expected, which would mean that it did indeed not remain a closed system and the lead escaped the zircon crystall structure somehow. After changing the HV pulser delay from 1200 ns to 1400 ns a possible lead peak was observed. The spectrum is shown in Figure 5.5. Since all the lead in zircon is supposed to be radiogenic, the isotope abundances will of course be different than their terrestrial abundances. Due to the richness of uranium,206 Pb will be the most abundant lead peak.

illustration not visible in this excerpt

Figure 5.5: Measured spectrum with different pulser delay settings showing lead, thorium and uranium.

The peaks appearing at masses 206, 207 and 208 were not perfectly resolved. Therefore it was only possible to estimate an age using the[illustration not visible in this excerpt] to[illustration not visible in this excerpt] ratio. Using Equation 5.2 the age is then given by:

illustration not visible in this excerpt

In a first step the age was estimated using the raw values without using any RSC corrections. This yielded an age of t =(2254 ± 129) · 106 a. Of course, for a quantitative analysis RSC corrections have to be applied to the measured data. A correlation between RSC values and the first ionization energy of the elements has been observed. With this RSC values of certain elements can be estimated according to Zhang et al.[30]. This was done to get approximate values for the RSC of lead and uranium. The fit is shown in Figure 5.6.

Figure 5.6: Fit for the estimation of the RSCs of lead and uranium by using their first ionization energies (fit was provided by Dr. M.B. Neuland).

illustration not visible in this excerpt

Again the age was estimated using Equation 5.3, this time the values were corrected using the estimated RSC values. This yields an age of t =(3517 ± 129) · 106 a. The results are summarized in Table 5.2. The given uncertainty takes only the error given from the peak integration process into account.

Table 5.2: Summarized results of the radiometric dating.

illustration not visible in this excerpt

The derived age is considerably younger than the t =(3909 ± 13) · 106 a re- ported by Gnos et al.[6]. The main source of this error comes from the strong dependence of the RSC values. Estimating RSCs from their ionization energy will give an approximate value. This value will however be less accurate than those determined from a set of measurements. This dependence is why the lead- lead dating method is preferred over the uranium-lead method. Additionally, it can be seen from the spectrum in Figure 5.5 that the uranium peak might be overlapping with a small cluster. This makes the peak slightly bigger than it should be and makes an accurate peak integration difficult. It should also be considered that the age was estimated using only one mass spectrum. More data points are needed to guarantee a reliable result. Lastly, it is unclear how the spectrum was affected by the mentioned resonances of the HV pulser. Nev- ertheless, it could be shown that radiometric dating is possible with the help of an HV pulser.

5.4 Conclusion and Outlook

Coupling the LMS instrument with an HV pulser greatly improves detection sensitivity for heavy elements. This opens up many new application of the instrument. Rare earth elements as well as other heavy trace elements such as uranium, thorium or tungsten, which have an abundance of a couple ppm, can be successfully measured. This new feature opens also the possibility of sensitive radiometric dating.

The current measurements are the first example of the measurements which could be prepared on heavier trace elements with a HV pulser. In particular, the measurements of the isotope ratios of the radiogenic lead in zircon could be now measured readily. More advanced dating analysis can be conducted in future studies including the206 Pb / 207 Pb dating method and possible other ones (e.g., W/Hf) to achieve a consistent picture of the material formation times.

Appendix

A: Sayh al Uhaymir Reference Values

Table 5.3: Composition of the regolith sample according to6. Measured with inductively coupled plasma mass spectrometry and opical emission spectrome- try.

illustration not visible in this excerpt

Table 5.4: Voltage Set: Kreep 20160706 2.opt

illustration not visible in this excerpt

Table 5.5: Voltage Set: Kreep Sample2 2016 4.opt

illustration not visible in this excerpt

Table 5.6: Voltage Set: Kreep 1006 1.opt

illustration not visible in this excerpt

Table 5.7: Voltage Set: Kreep Zirkonium 1.opt

illustration not visible in this excerpt

Acknowledgements

I would like to express my gratitude to everyone who supported this master thesis.

A special thanks goes to Prof. Dr. Peter Wurz for giving me the opportunity to be a part of the LMS group and making this master thesis possible and to Dr. Marek Tulej for supporting me throughout this thesis.

I would also like to thank all the other members of the LMS group for their support and constructive inputs and feedbacks. Especially Valentine Grimaudo and Pavel Moreno-Garc´ıa for their patience and the long days they spend in the LMS lab with me and Reto Wiesendanger for his HV pulser that made many of the measurements in this thesis possible and the time spend conducting these measurements.

Additionally, I would like to thank Prof. Dr. Beda Hofmann and Prof. Dr. Klaus Mezger for answering all my geology questions and Dr. Miklós Mohos for providing the EDX measurements for me despite his busy schedule.

Finally, I would like to thank my family and friends for always supporting me with special thanks to my wonderful boyfriend Philipp Zappa for always cheering me up during tough times.

Bibliography

[1] G. Managadze, P. Wurz, R. Sagdeev, A. Chumikov, M. Tulej, M. Iakovleva and A. Bondarenko (2010). Study of the Main Geochemical Char acteristics of Phobos Regolith Using Laser Time-of-Flight Mass Spectroscopy, Solar System Research, 44(5):376-384.

[2] U. Rohner, J.A. Whitby, P. Wurz (2003). A miniature laser ablation time of flight mass spectrometer for in situ planetary exploration, Measurement Science and Technology, 14:2159-2164.

[3] P. Wurz, D. Abplanalp, M. Tulej, M. Iakovleva, V.A. Fernandes, A. Chu- mikov, and G. Managadze (2012). Mass Spectrometric Analysis in Planetary Science: Investigation of the Surface and the Atmo sphere, Sol. Sys. Res., 46, 408-422.

[4] A. Riedo, M. Neuland, S. Meyer, M. Tulej and P. Wurz (2013). Coupling of LMS with a fs-laser ablation ion source: elemental and isotope composition measurements, Journal of Analytical Atomic Spectrometry, 28(8):1133-1356.

[5] M. Tulej, A. Riedo, M.B. Neuland, S. Meyer, P. Wurz, N. Thomas, V. Grimaudo, P. Moreno-Garc´ıa, P. Broekmann, A. Neubeck and M. Ivarsson (2014). CAMAM: A miniature laser ablation ionisation mass spec trometer and microscope-camera system for in situ investigation of the composition and morphology of extraterrestrial materials, Geostandards and Geoanalytical Research, 38(4):441-466.

[6] Edwin Gnos, Beda A. Hofmann, Ali Al-Kathiri et al. (2004). Pinpointing the Source of a Lunar Meteorite: Implications for the Evolution of the Moon, Science Magazine, 305, 657-659.

[7] P. Wurz (2015). Specialist Course: Mass Spectrometry and Ion Op- tics, Lecture, Physics Institute, University of Bern.

[8] W.E. Stephens (1946). A pulsed mass spectrometer with time disper- sion, Phys. Rev. 69, 674-674.

[9] W.C. Wiley and I.H. McLaren (1955). Time-of-flight mass spectrometer with improved resolution, Rev. Sci. Instr., 26(12), 1150-1156.

[10] B.A. Mamyrin, V.I. Karataev, D.V. Shmikk, and V.A. Zagulin (1973). The mass-reflectron, a new nonmagnetic time-of-flight mass spectrom- eter with high resolution, Sov. Phys. JEPT 37(1), 45-48.

[11]Hammamatsu (2009). MCP & MCP Assembly, Selection Guide. Technical Booklet, 314-5, Shimokanzo, Iwata City, Shiuzoka Pref., 438-0193 Japan.

[12]A. Riedo (2014). Development and performance evaluation of a laser ablation ionisation mass spectrometer for in situ planetary exploration, Inauguraldissertation, University of Bern, Physics Institute, Space Research and Planetary Sciences.

[13]A. Bieler, K. Altwegg, L. Hofer, A. Riedo, T. Smon and P. Wurz (2011). Optimization of mass spectrometers using the adaptive particle swarm algorithm, Journal of Mass Spectrometry 46, 1143-1151.

[14]A.Riedo, S. Meyer, B. Heredia, M. Neuland, A. Bieler, M. Tulej, I. Leya, M. Iakovleva, K. Mezger and P. Wurz (2013). Highly accurate isotope com- position measurements by a miniature laser ablation mass spec- trometer designed for in situ investigations on planetary surfaces., Planetary Space Science 87, 1-13.

[15]Stefan Meyer (2013). Fully Automated and Highly Precise Data Analysis for a Miniaturized Laser-Ablation Mass Spectrometer, Maser Thesis, University of Bern, Physics Institute, Space Research and Planetary Sciences.

[16]Maike Brigitte Neuland (2016). In situ mass spectrometry for plan- etary exploration: Quantitative chemical composition measure- ments on planetary surfaces, Inauguraldissertation, University of Bern, Physics Institute, Space Research and Planetary Sciences.

[17]H.Wenk and A. Bulakh (2004). Minerals, Their Constitution and Ori- gin, Cambridge University Press.

[18]John C. Russ, M. Ashby, R. Kiessling and J. Charles (1984). Fundamen- tals of Energy Dispersive X-ray Analysis, Butterworths Monographs in Materials, ISBN: 978-0-408-11031-0.

[19]M.B. Neuland, S. Meyera, K. Mezger, A. Riedo, M. Tulej, P. Wurz (2014). Probing the Allende meteorite with a miniature laser-ablation mass analyser for space application, Planetary and Space Science 101, 196-209.

[20]Claude J. Allègre (2008). Isotope Geology, Cambridge University Press.

[21]J. S. Becker (2007). Inorganic Mass Spectrometry: Principles and Applications, Wiley & Sons.

[22] E.B. Watson, D.A. Wark, J.B. Thomas (2006). Crystallization ther- mometers for zircon and rutile, Contributions to Mineralogy and Petrology 151, 413-433.

[23]J.M. Ferry, E.B. Watson (2007). New thermodynamic models and re- vised calibrations for the Ti-in-zircon and Zr-in-rutile thermome- ters, Contributions to Mineralogy and Petrology 154, 429-437.

[24]Wilder D. Bancroft, H. L. Davis (1928). Raoult ’ s Law, J. Phys. Chem., 33 (3), 361-370.

[25]M.M.Wielicki, T.M. Harrison and Schmitt (2012). Geochemical signa- tures and magmatic stability of terrestrial impact produced zircon, Earth and Planetary Science Letters, v. 321-322, 20-31.

[26]Bin Fu, F. Zeb Page, Aaron J. Cavosie, John Fournelle, Noriko T. Kita, Jade Star lackey, Simon A. Wilde, John W. Valley (2008). Ti-in-zircon thermometry: application and limitations, Contributions to Mineral- ogy and Petrology 156, 197-215.

[27]John W.Valley, Michael J. Spicuzza, Takayuki Ushikubo (2014), Corre- lated δ 18 O and [Ti] in lunar zircons: a terrestrial perspective for magma temperature and water content on the Moon, Contributions to Mineralogy and Petrology 167, 1-15.

[28]M.M. Wielicki and T.M. Harrison (2015). Zircon formation in impact melts: Complications for deciphering planetary impact histories, in Osinski, G.R., and Kring, D.A., eds., Large Meteorite Impacts and Plan- etary Evolution V: Geological Society of America Special Paper 518, p. 127- 134.

[29]R. Wiesendanger, M. Tulej, M. B. Neuland, S. Frey, P. Wurz, A. Riedo, V. Grimaudo, P. Moreno-Garc´ıa, A. Cedeño López and P. Broekmann (2017). A miniature laser mass spectrometer (LMS) with an HV pulser for selective and sensitive element and isotope analysis (to be sub- mitted).

[30]B. Zhang, M. He, W. Hang, B. Huang (2013). Minimizing matrix effect by femtosecond laser ablation and ionization in elemental deter- mination, Anal. Chem. 85, 4507-4511.

[31]R. Huang, Q. Yu, L. Li, Y. Lin, W. Hang, J. He and B. Huang (2011). High irradiance laser ionization orthogonal time-of-flight mass spectrometry: a versatile tool for solid analysis, Mass Spectrom. Rev. 30 (6), 1256-1268.

1 This chapter is based on[7].

1 Raoult’s law is a law of thermodynamics, stating that the partial vapour pressure of each compound of an ideal mixture of liquids is equal to the vapour pressure of the pure component multiplied by its mole fraction in the mixture.

72 of 72 pages

Details

Title
Chemical Analysis of a Lunar Meteorite by Laser Ablation Mass Spectrometry
College
University of Bern
Grade
5.5
Author
Year
2016
Pages
72
Catalog Number
V378511
ISBN (Book)
9783668581821
File size
3125 KB
Language
English
Tags
SaU 169
Quote paper
Samira Frey (Author), 2016, Chemical Analysis of a Lunar Meteorite by Laser Ablation Mass Spectrometry, Munich, GRIN Verlag, https://www.grin.com/document/378511

Comments

  • No comments yet.
Read the ebook
Title: Chemical Analysis of a Lunar Meteorite by Laser Ablation Mass Spectrometry



Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free