| Quartz is a common mineral in the Earth's crust,formed in many geological environments.The concentration of trace elements in quartz provides considerable information about how the mineral was formed.For example,the temperature and pressure at the time of formation can be calculated by determining the Ti content of quartz.The change in Al content in hydrothermal quartz can reflect changes in the fluid pH.By combining CL images with the distribution of trace elements such as Al,Ti,and K in quartz,quartz has been used to interpret geologic processes,including those magma genesis and evolution,and hydrothermal ore formation.Therefore,it is of crucial geological significance to accurately determine the concentration and distribution of trace elements in quartz.In situ microanalysis of trace elements by techniques such as laser ablation inductively coupled plasma mass spectrometry?LA-ICP-MS?and secondary ion mass spectroscopy?SIMS?have been widely used to obtain reliable results.Both have trong analytical ability,but will damage the sample in analysis process and the spatial resolution is low.Electron probe microanalysis?EPMA?has the advantages of non-destructive testing,high spatial resolution.However,the achievement of high accuracy and precision can be an issue.In this paper,the main factors affecting the analytical precision and accuracy can be summarized.The main factors affecting the analytical precision include acceleration voltage,beam current and counting time.The main factors affecting the analytical accuracy are as follows:?1?quartz lattices are easily damaged at high beam currents,resulting in the migration of elements which changes the characteristic X-ray count intensity over time and leads to inaccurate analytical results;?2?two-point interpolation method can cause error to elements that are not linear with the background curve?such as the background of Al-K??;?3?some interference factors,including interference peaks?such as the influence of Al-SK?on background test?,secondary fluorescence?such as a neighboring feldspar close to quartz producing an apparent signal of Al-K?within the quartz?,“step”?appearing on the background of Ti-K??and“holes”?caused by secondary Bragg diffraction at other lattice planes at some angle to the preferred lattice plane?.The detection limit of Al and Ti were determined at different accelerating voltages?15,16,17,18,19,20,and 25 kV?and beam currents?100,200,300,400,500,and 600nA?.The results indicate that,with an increase in the accelerating voltage,the detection limit of Al first decreases slightly and then increases slightly.As the accelerating voltage increases,the detection limit of Ti decreases.However,when the beam current and accelerating voltage continues to increase to over a certain value?20 kV,500 nA?,the effect of raising the beam current and accelerating voltage to improve the detection limit becomes less noticeable.Therefore,the experiment was carried out under the conditions of 500 nA beam currenand 20 kV accelerating voltages.Based on the research of voltage,current,background counting time and detection limit,the counting stability of Al and Ti in Q-Std was examined with beam diameters of0?focused?,1,2,5,10,15,and 20?m,at an accelerating voltage of 20 kV and beam current of 500 nA.Finally,the test beam spot diameter of 20?m is selected.In addition,we conducted a TDI study on surrounding background positions of the Al-K?and Ti-K?peak.It is found that under the condition of high current and small beam spot,the background position will also have the change of TDI.Therefore,it is necessary to select a larger beam spot to avoid TDI variation when using a high current.In order to search for potential interference factors,Detailed wavelength scanning was carried out on the background of Al-K?and Ti-K?,which found Al-K?1,3 and Al-SK?on the background of Al-K?and“step”on the background of Ti-K?.These references can be handled by keeping a safe distance from them when choosing background positions.About secondary fluorescence,Monte Carlo simulation was used to find that for Al excited in an adjacent sanidine,the influence of Al-K?secondary fluorescence ends?10?g/g?at a distance of approximately 16?m,and for Ti-K?in quartz-rutile,an elevated signal?10?g/g?was seen up to a distance of 360?m.In this paper,different models of EPMA in different laboratories are used to compare experiments.It is found that“steps”are ubiquitous in PET crystal wavelength scanning curves.The PETJ of some instruments showing“hole”,but similar effects are not always observed on PETJ.“Holes”are always observed on PETL.In this study,based on detailed wavelength scanning,several points on both sides of the Al-K?and Ti-K?peaks were chosen to determine the background value.Then,quadratic polynomial,exponential,logarithmic,and power functions were used to fit the background curve.Background value is obtained by manual calculation.This study used the ZAF correction to obtain the contents of Al were 158.4±3.4?g/g,154.7±3.4?g/g,150.7±3.4?g/g and 155.5±3.4?g/g,respectively.The contents of Ti were55.3±2.0?g/g,55.7±2.0?g/g,55.9±2.0?g/g and 58.0±2.0?g/g,all within the reference value interval.The content of Al calculated by two-point interpolation method were127.0±3.4?g/g,which is obviously lower than the reference value interval.The content of Ti calculated by two-point interpolation method was 53.9±2.0?g/g,which is obviously lower than the multi-point fitting result.The multi-point background method is very obvious to improve the test accuracy.In addition,it is verified that the background obtained from other quartz samples can be used as the background of an unknown quartz sample,which reduces the analysis time and minimizes sample damage.In the test results,the detection limits of Al and Ti are less than 9?g/g?3??,and using PETL for Ti analysis can lower detection limit to 4?g/g?3??,which can meet most of the quartz samples testing requirements. |
PDF Full Text Request