Atomic absorption spectroscopy (AAS)
Aqueous magnesium concentrations were measured by Flame Atomic Absorption Spectroscopy (F-AAS). The principle of these analyses were first described by Walsh (1955) and further developed by Russell et al. (1957) and Box and Walsh (1960). The solution sample is introduced as an aerosol into an air-acetylene-flame. The radiation from a hollow-cathode lamp, which emits the specific wavelength of Mg (285.21 nm) crosses through the flame. Some of the photons crossing the aerosol are absorbed by the Mg atoms of the sample present in the flame. The absorption is detected, measured and directly relates to the concentration. An intensity to concentration calibration curve is established by measuring standards solutions prepared in the same matrix as the samples.
Aqueous Mg concentration measurements in this study were carried out for concentrations from 0.1 and 0.6 ppm using a Perkin Elemer AAnalyst 400. Calibration solutions were prepared using the sample matrix (2% HNO3, 0.05 M HCl, EDTA, sulphate, citrate) and to avoid interference was used a 1000 ppm Mg(NO₃)₂ Merck-Certipur® standard solution diluted into aqueous 0.05mol/kg HNO3. Prior to measurements all samples were diluted. Therefore concentrations were close to the middle of the measurement concentration range (± 0.3 ppm). To minimize the effect of possible interference, a small quantity lanthanum oxide (30µL in 10 ml sample) was added to each sample and the standards prior to analysis. The limit of detection (LOD) and the limit of quantification (LOQ) were determined for each matrix to be 5 ppb and 1 ppb respectively. Measurements had an uncertainty of ±1%.
Atomic emission spectroscopy (AES)
The sodium content of samples containing Na were determined by flame atomic emission spectrophotometry (F-AES). The aqueous sample is introduced as a solution into a flame. The electrons of the outer valence layer are excited towards a higher electronic state. When -38- the electrons fall back to their fundamental state, photons are emitted at a characteristic wavelength (e.g. Na= 589.00 nm). The emission spectrum was measured by the atomic absorption spectrometer (Perkin Elemer AAnalyst 400) in the flame emission mode. As previously described (see AAS Chapter 220.127.116.11) a calibration curve is established allowing the determination of the Na concentration.
Sodium measurements were carried out at concentrations ranging from 0.5 to 6 ppm. Sodium concentrations in purified samples were expected to be below the limit of quantification. Due to a high sample consumption (4 mL), samples needed to be diluted by a factor of 2.7 prior to analysis. The LOQ and LOD were determined at 0.06 ppm and 0.02 ppm respectively.
Aqueous silicon concentrations in this thesis were determined by the application of the molybdate blue method (Strickland, 1952) following the procedure of Tréguer and Le Corre (1975). The principle of this method is the formation of a yellow β-molybdosilic acid (silicomolybdate), when ammonium molybdate is added to an acidic Si-rich solution. This reaction can be written as:
Sulfuric acid is added to the molybdate solution as an acidifying agent, and oxalic acid is added to the sample solution to reduce possible phosphate interferences. Finally ascorbic acid is added to the sample solution to reduce the ammonium molybdate complex. This leads to a blue coloration of the solution. The absorbance of this color relative to a blank solution can be measured at 660 nm and is proportional to the dissolved Si concentration. Standards of known concentrations were measured to establish a calibration curve.
Aqueous Si concentrations measurements were carried out at concentrations of 0.05 to 1 ppm, and from 0.1 to 10 ppm using a Bran & Luebbe technicon analyser III coupled to a Seal XY-2 autosampler and an auto analyser II mixing unit. The solution samples were introduced into the mixing unit at a flow rate of 0.32 ml/min.. The ammonium molybdate solution is mixed with this sample solution at a flow rate of ~0.8 ml/min. Subsequently oxalic acid and finally ascorbic acid are added at a flow rate of ~0.23 ml/min to the measuring solution. Calibration solutions were prepared by diluting a 1000 ppm Merck-Certipur® Si standard solution with deionized MQ® water. The sample solutions were likewise diluted with deionized MQ® water so that their concentrations fell in the middle of the measurement range. At the beginning and end of each set of measurements a standard, was measured to monitor for potential instrument drift. For measurements of Si concentrations between 0.05 and 1 ppm, a Mississippi 03 standard was used; for measurements of Si concentrations between 0.1 and 10 ppm a Miramichi 02 and later a Sangamon 03 standard were used. For the range of 0.05 to 1 ppm LOD was determined to be 0.004 ppm and the LOQ to be 0.013 ppm. The LOD and LOQ for measurements between 0.1 and 10 ppm were determined to be 0.01 ppm, and 0.04 ppm respectively. The long-term reproducibility of these measurements was within 3%.
Organic carbon measurements
Organic carbon was determined in aqueous solution samples containing organic ligands. The measurements were performed using a Shimadzu® TOC-VCSN analyser, equipped with an ASI-V auto sampler and TOC-Control V software. The principle is based on a combustion catalytic oxidation method developed by Shimadzu®. The sample is acidified to convert the carbonate and bicarbonate ions into dissolved carbon dioxide. The sample is then automatically injected with a syringe into a quartz tube filled with air and a platinum catalyst, which is then heated to 680°C. This transforms the dissolved organic carbon into CO2, which is transported to the infrared detector. The total organic carbon concentration (TOC) is obtained by comparison of infrared measurements to a calibration curve, which is established from a set of reference solutions under the same conditions as the sample measurements.
TOC measurements were carried out at carbon concentrations ranging from 2 to 30 ppm. The aqueous samples were to concentrated, they were diluted with MQ® water in glass tubes prior to their analysis. Subsequently they were slightly acidified with bi-distilled HCl.
Quadrupole – ICP-MS
The element concentrations of the IAPSO and JDol-1 standard solutions after their cleaning by ion exchange chromatography (see Chapter 18.104.22.168) were measured on the Agilent® 7500 quadrupole ICP-MS. The solution to be analysed is converted into a fine aerosol by a micro nebuliser and introduced under a constant argon gas flow into an argon plasma torch at temperatures between 5000 and 10000°K. The sample is then ionized. The ions are extracted via a high voltage anode through a set of two cones, which filter all the contaminants and create an ion beam. On their way to the detector, the ions and molecular species go through two filtering systems:
1) The collision/reaction cell: a closed space where ions of low kinetic energy and large molecules are stopped, deviated or broken by large ions (e.g. O, H or He) or react with these to form other species of higher mass, to be further measured. The latter function was not used during this thesis.
2) The quadrupole: this consists of four parallel rods (Fig 2.8). Opposite rods are charged with an adjustable high voltage, which is inverted for the normal pair of rods. The charged ions entering the system follow a spiral pathway, which depends on their mass to charge ratio and on the applied voltage. Hence, only the chosen mass/charge ratios are not deviated and can be measured afterwards.
The signal of the ions is measured by an electron multiplier detector, which enables quantification of the abundances of the different ionic species, in counts per second. These abundances are directly proportional to the concentration of this species in the injected sample. Abundance to concentration calibration curves are established by measuring the abundances of known standards.
Prior to the measurements, the samples were diluted in 2% bi-distilled HNO3 to obtain a concentration <500 ppm in the target element. A known volume of an indium and rhenium enriched solution of known concentration was added as an internal standard. The measured uncertainties of measured concentrations vary between 2 and 15%, depending of the concentration of the sample solution. If the uncertainty of the measurements exceeded 15%, the concentrations determined for this element was discarded. The detection limit of the measurements was found to vary from 0.1 to 10 ppt, depending on the element analysed.
High Resolution – ICP-MS (HR-ICP-MS)
The isotope abundance in the spike solutions were determined by high resolution inductively coupled plasma mass spectrometry (HR-ICP-MS). The measurements were performed on a Thermo Sicentific ELEMENT-XR® with a reversed Nier-Johnson geometry. The detailed functioning of an ICP-MS is described in detail in section 22.214.171.124.
The 29Si and 25Mg-spike solutions were diluted to concentrations between 0.05 and 2 ppm in 2% HNO3 prior to the measurements. The magnesium and silicon samples were introduced with a free aspirating stable introduction system (SIS) and measured in medium resolution mode, with a mass resolution of ~4300. Silicon was additionally measured in high mass resolution mode (R ~9300). A 1 ppb Li tuning solution was used. The diluted spike solutions were introduced with Ar gas into the plasma. Blanks of 2% HNO3 were measured for each measurement sequence. The intensities of the 3 stable isotopes with the masses 27.9764, 28.9760, 29.9732, and 23.9845, 24.9853, 25.9821 for silicon and magnesium. respectively, were measured. These were detected in either triple mode (Faraday and electron multiplier in analogue and pulse modes) for the magnesium spike solutions, or in analogue mode for the silicon spike solutions, with a secondary electron multiplier. The samples signals were corrected for a blank using the blank signals. The signal detected for the different isotopes directly depends on the concentration of the Si and Mg isotopes in the solution. The isotopic ratios could therefore be determined from the ratio of the blank corrected signal of the individual isotopes.
All samples for isotopic ratio determination needed to be purified to remove any potentially interfering elements prior to their analysis. These purification procedures are described in this section, along with the principles of the Multi Collector ICP-MS (MC-ICP-MS).
Silicon sample preparation and ion-exchange chromatography
Silicon bearing powders and aqueous solutions had to be purified prior to their analysis by MC-ICP-MS (see Chapter 126.96.36.199). For the amorphous SiO2 powders, the sample purification is based on the alkaline fusion preparation protocol described in Zambardi & Poitrasson (2011). This protocol calls for:
1) 1 to 5 mg of the powders were weighted into silver crucibles (XRF scientific, Montreal, Canada), together with ~200 mg Merck® suprapure NaOH pellets. If the powders were in contact with organic solutions, a calcination was performed prior to the addition of the NaOH pellets. Therefore SiO2 sample powder in the crucibles in a furnace at 500°C for 4h, similar to the approach for soils of Cornu et al. (1999).
2) The crucibles were placed into a 720°C pre-heated furnace for 10 min.
3) After cooling to room temperature, the crucibles were placed into 30 cc Savillex® Teflon beakers filled with 20 ml Milli-Q® water to dissolve the fusion cake..
4) After 24 hours, the resulting solutions were transferred to 60 cc polypropylene bottles and diluted with 20 ml Milli-Q® water to obtain a 40 ml solution.
5) These solutions were then acidified to a pH of 1.5 using ~10 N HCl and placed on a heating plate at 50°C for 24 to 48 h to ensure that possible suspended material is completely and efficiently dissolved.
Aqueous samples of experiments run without the presence of organic ligands were prepared for Si isotope analysis by diluting the solutions – if needed – and subsequently acidifying them to pH ~2 using bi-distilled 3N HCl. Organic bearing fluid samples had to be carefully treated, as the presence of organic matter likely causes interferences during Si isotope measurements using the MC-ICP-MS. Therefore solutions containing organic ligands had to be processed to ensure complete removal of these organic compounds. For this standard methods such as the use of H2O2 and UV-light were tested in a first attempt. These methods proved to be ineffective as confirmed by the organic carbon measurements. The following sample preparation protocol was therefore adopted to purify aqueous solutions containing organic ligands:
1) 5 ml of the organic bearing aqueous sample was evaporated to dryness in silver crucibles in a laminar flow box.
2) Calcination was performed by heating the SiO2 and organic residue in the crucibles in a furnace at 500°C for 4h, similar to the approach of Cornu et al. (1999) for soils.
3) Approximately ~200 mg of Merck® NaOH pellets were added into the silver crucibles.
4) The fusion of the remaining SiO2 powder was performed by placing the crucibles into a furnace at 720°C for 10 min.
5) The crucibles were cooled to room temperature, and subsequently placed in 30 cc Savillex® Teflon beakers filled with 15 ml Milli-Q® water to dissolve the amorphous SiO2 powder.
6) The resulting solutions were transferred after 24 hours into 30 cc polypropylene bottles and diluted with Milli-Q® water to 20 ml.
7) Samples were acidified to a pH of 1.5 using ~10 N aqueous HCl.
After sample preparation, the resulting solutions were purified by the cation-exchange chromatography protocol described by Zambardi and Poitrasson (2011). In this method, 10 ml BioRad® polypropylene columns were filled with 2 ml BioRad® AG50W-12X (200–400 mesh) cationic resin. Before the samples were treated, the resin was cleaned and conditioned with 3 ml HCl and HNO3 in following order: 3 N HCl, 6 N HCl, 7 N HNO3, 10 N HCl, 6 N HCl, and 3 N HCl (Georg et al., 2006). The resin was then flushed with Milli-Q®, so that the pH before the sample was loaded was close to neutral. The sample solutions were loaded (0.5-2 ml, depending on the Si concentration). Si is not retained by the resin, as non ionic H4SiO4o largely dominates Si species in solution below pH 9, and was therefore directly collected. The columns were subsequently eluted twice with Milli-Q® water to obtain a 6 ml sample solution. These sample solutions were then diluted to obtain 3 ppm of Si and acidified with 10 N bi-distilled HCl to obtain a 0.05 M HCl matrix. Silicon recovery after processing was determined by colorimetry to be between 90 to 100 %.
Magnesium sample preparation and ion chromatography
The magnesium bearing powders and aqueous solutions collected from the isotopic exchange experiments had also to be prepared and subsequently purified by ion exchange chromatography prior to their isotopic measurement with the Neptune® MC-ICP-MS. The processing of the brucite powder consisted of:
1) ~1 mg of brucite powder and the JDo-1 standard were weighted and placed in Savillex® beakers.
2) These powders were dissolved overnight in 2 ml of 15 N bi-distilled HNO3.
3) The resulting solutions were evaporated to dryness at 120°C, and the residues were re-dissolved in 5 ml 1N HNO3.
4) Mg concentrations of these latter solutions were determined using AAS.
5) Aliquots of the solutions containing 15 µg Mg were evaporated and re-dissolved in 2 ml 1M HNO3 to be ready for column chemistry.
The Mg bearing aqueous solutions were prepared as follows:
1) Aliquots of the fluid samples (10-500 µl) and the IAPSO seawater standard
containing 15 µg Mg were placed into Savillex® beakers. In case of the presence of organic ligands ~0.5 ml H2O2 was added and left to react for 3-6 hours.
2) Sample solutions were evaporated to dryness at 120°C.
3) The solid residues were re-dissolved in 2 ml of 1M HNO3 to be ready for column chemistry.
The sample preparation follows the Teng et al. (2007) method. All the samples were processed in 10 ml Bio-Rad® polypropylene (PP) columns filled with 1 ml Bio-Rad® AG50W-X8 200–400 mesh cationic exchange resin. The resin was pre-cleaned with 1 N HCl before filled into the columns. Once in the column, the resin was further cleaned with 10 ml of Milli-Q® water, 20 ml of 4M HCl and finally once more with 10 ml of Milli-Q® water. The resin was then conditioned with 10 times the resin volume of aqueous 1N HNO3. 2 ml of the prepared sample solutions were loaded onto the resin and the elements of interest (in this case only Mg) were eluted by the subsequent washing of the resin with 1 M HNO3. Sodium is removed, and eluted within the first 10 ml after loading the sample. The Mg fraction was then collected in the following 20 ml of elution, and Ca was removed lastly from the resin, and then eluted within the next 30 ml (Teng et al., 2007).
As ion exchange chromatography can significantly fractionate Mg isotopes (Chang et al., 2003; Teng et al., 2007), it is essential to obtain a Mg yield close to 100 %. A resin calibration was performed to ensure all the Mg was collected during column chromatography. Seven fractions of the two standards JDol-1 and IAPSO were collected starting from the last 2 ml of the sodium elution until the first 2 ml of the calcium elution. These were subsequently analysed with the Agilent 7500 quadrupole ICP-MS (see Chapter 188.8.131.52) to determine the concentrations of all elements in solution. The elution curve for Mg in Fig. 2.9 shows no shift compared to the protocol of Teng et al. (2007), which indicates that Mg was eluted completely through the columns, validating the method used in this thesis.
For samples collected from the sulphate bearing experiments was additionally recovered fluid analyses were made using AES measurements (see Chapter 184.108.40.206) to assure that the recovered solutions were free of sodium, as element this could affect the Mg isotopic measurements.
Principles of the Neptune® Multi Collector ICP-MS (MC-ICP-MS)
The isotopic compositions of the aqueous solutions and solid powders recovered from the isotopic exchange experiments were obtained using a Thermo Fisher Scientific Neptune® Multi-Collector Inductively Coupled Plasma Mass Spectrometer (MC-ICP-MS). The Neptune® MC-ICP-MS is a double focussing mass spectrometer, which is capable of preforming measurements with a highly accurate mass resolution using its multi collector mode (Weyer and Schwieters, 2003). Its principle is based on the physical separation of isotopes in an ion beam by both an electrostatic analyser (ESA), which correspond to an energy filter, and a magnet that works as a mass filter. The ESA is placed before the magnet, as shown in Fig. 2.10 based on the Nier and Johnson geometry.
All samples were prepared for analysis in weakly acidic aqueous solutions and converted into a fine aerosol in the nebulizer. The sample is subsequently introduced under a constant argon gas flow through the spray chamber into the torch. In the torch, an inductively coupled Argon plasma is generated at high temperatures (> 8000 K; Georg et al., 2006) in which the mixture of sample and gas is ionized. When the ions leave the torch, their energy will vary between 0.5 and 5 eV (Georg et al., 2006). The magnet alone could not be able to achieve a good isotope separation due to this energy variation. Therefore the ESA is placed before the magnet to deflect the ions as a function of their kinetic energy. This means that if the ion beam containing ions with different velocities reaches the ESA entrance only ions with the same kinetic energy will pass. Three entrance slits are located in front of the ESA enabling measurements in low, medium and high mass resolution. The ions leaving the ESA unit will be deflected by the magnet according to their mass to charge ratio. The isotopes thus will have an individual path according to their mass. The beam leaving the magnet, therefore contains various sub-beams, where each beam can be associated with a given isotope. The sub-beams are collected in 9 available Faraday cups. These generate an electrical signal, which can be detected by either a Faraday or SEM detector, allowing the determination of the relative isotopic abundances.
Table of contents :
Chapter 1 Introduction
1.1a. General introduction (English version)
1.1b. Introduction génerale (Version française)
1.2. Aqueous chemistry
1.2.1. Chemical equilibrium
1.2.1. Speciation and complexation in aqueous solution
1.3. Principles of stable isotope fractionation
1.3.1. Stable isotopes
1.3.2. Stable isotope fractionation
1.3.3. Stable isotope fractionation processes
1.3.4. Influence of aqueous speciation on isotope fractionation
1.4. Silicon and magnesium isotopic systems
1.4.1. Silicon isotopes
1.4.2. Magnesium isotopes
Chapter 2 Materials and Methods
2.1. Theoretical Background
2.1.1. Three-isotope method
2.1.2. Kinetics of isotopic exchange reactions
2.2.1. Amorphous Silica (SiO2,am)
2.2.2. Brucite synthesis
2.2.3. Aqueous Solutions
220.127.116.11. Inorganic solutions
18.104.22.168. Aqueous organic solutions
2.2.4. Isotopic spike solutions
2.3. Experimental design
2.3.1. Chemical equilibration of reactive fluids
2.3.2. Isotope exchange experiments
2.4. Analytical methods
2.4.1. Characterization of solid phases
22.214.171.124. Scanning electron microscopy (SEM)
126.96.36.199. Transmission electron microscopy (TEM)
188.8.131.52. Surface area
184.108.40.206. Thermogravimetric analysis
220.127.116.11. X-ray Powder Diffraction (P-XRD)
2.4.2. Characterization of aqueous solutions
18.104.22.168. pH measurements
22.214.171.124. Atomic absorption spectroscopy (AAS)
126.96.36.199. Atomic emission spectroscopy (AES)
188.8.131.52. Organic carbon measurements
184.108.40.206. Quadrupole – ICP-MS
220.127.116.11. High Resolution – ICP-MS (HR-ICP-MS)
2.4.3. Isotopic Analysis
18.104.22.168. Silicon sample preparation and ion-exchange chromatography
22.214.171.124. Magnesium sample preparation and ion chromatography
126.96.36.199. Principles of the Neptune® Multi Collector ICP-MS (MC-ICP-MS)
188.8.131.52. Silicon isotope measurements
184.108.40.206. Magnesium isotope measurements
220.127.116.11. Analytical reproducibility
2.5. Geochemical calculations with PHREEQC
2.5.1. Thermodynamic modelling
2.4.2. Si speciation calculations
2.4.3. Mg speciation calculations
Chapter 3 The experimental determination of equilibrium Si isotope fractionation factors among H4SiO4 o, H3SiO4 – and amorphous silica (SiO2·0.32 H2O) AT 25 and 75 °C using the three isotope method
3.2. Theoretical background
3.2.1. Geochemical calculations of amorphous silica dissolution rates
3.2.2. Si isotope systematics
3.2.3. Three-isotope method
3.2.4. Kinetics of isotopic exchange
3.3.1. Experimental approach
18.104.22.168. Starting powder – amorphous silica
22.214.171.124. Initial aqueous solutions
126.96.36.199. Characterization of the aqueous solutions
188.8.131.52. Experiment design: Step 1 Equilibration of reactive fluids
184.108.40.206. Experiment design: Step 2 Isotopic exchange experiments
3.3.2. Si isotope analysis
3.4.1. Attainment of fluid-amorphous SiO2 equilibrium during the fluid equilibration75
3.4.2. Results of Isotope exchange experiments
220.127.116.11. Observations on the solid phases
18.104.22.168 Chemical and isotopic evolution of the isotope exchange experiments
3.4.3. Silicon isotope fractionation factors
3.4.4. Isotope exchange kinetics
3.5.1. Silicon isotope fractionation between amorphous silica and aqueous solution 83
3.5.2. Isotope fractionation among Si aqueous species
3.5.3. Kinetics of Si isotope exchange
3.5.4. Can Si fractionation be used as a paleo pH and temperature proxy?
Chapter 4 Extreme silicon isotope fractionation between silicic acid and aqueous organosilicon complexes: Implication for silica biomineralization
4.2. Experimental methods
4.2.1. Starting materials
4.2.2. Experimental design
22.214.171.124. Step 1: Equilibration of the reactive aqueous solutions with amorphous SiO2
126.96.36.199. Step 2: Isotopic exchange experiments
4.2.3. Analytical methods
188.8.131.52. Characterization of aqueous solutions
184.108.40.206. Data reporting and Si isotope analysis
4.3.3. Speciation calculations
4.4. Ab initio calculations
4.5.1. Attainment of equilibrium between the fluid and amorphous SiO2
4.5.2. Results of isotope exchange experiments
4.5.3. Experimental and theoretical silicon isotope fractionation factors
4.5.4. Isotope exchange kinetics
4.6.1. Silicon isotope fractionation in the presence of catechol
4.6.2. Si isotope exchange kinetics in the presence of catechol
4.6.3. Implications for biomineralization
Chapter 5 Determination of the equilibrium magnesium isotope fractionation factors between brucite and aqueous Mg inorganic and organic species
5.2.1. Starting materials
220.127.116.11. Brucite synthesis
18.104.22.168. Initial reactive aqueous solutions
22.214.171.124. Characterization of sampled experimental fluids
5.2.2. Experimental design
126.96.36.199. Step 1: Equilibration of initial aqueous solutions with Mg(OH)2
188.8.131.52. Step 2: Isotopic exchange experiments
5.2.3. Data reporting and Mg isotope analysis
5.2.4. Geochemical and speciation calculations with PHREEQC
5.3.1 Equilibrium between fluid and brucite
5.3.2 Results of isotope exchange experiments
5.3.3. Determination of magnesium equilibrium isotope fractionation factors
5.3.4. Isotope exchange kinetics
5.4.1. Magnesium isotope fractionation between brucite and Mg2+
5.4.2. Magnesium isotope fractionation between Mg2+ and aqueous Mg organic and inorganic complexes
5.4.3. Mg isotope exchange kinetics
5.4.3. Implications for natural systems
Chapter 6 Conclusions and outlook
6.1 General conclusions and outlook (English version)
6.2. Conclusion générale et perspectives (Version française)