ForewordGlow discharge optical emission spectrometry (GDOes) is a thin film material characterization technique that combines surface peeling and elemental analysis As early as the 30s of the 20th century, the development and application of glow discharge devices and chemical analysis correlation spectroscopy began 1 In 1947, Rand McNally designed a glow discharge test device to achieve extremely high detection sensitivity and achieve a response to 001 Detection of g fluorine 2 However, it was not until the 60s that glow discharge became the focus of analytical chemistry In 1967, the GRIMM light source appeared3 , its anode is cylindrical, the sample is cathode, and the anode ground maintains zero potential, and the sample maintains negative high pressure during operation The stability and repeatability of the GRIMM light source compared with other types of light sources is higher, making GDO a more reliable and versatile characterization technique 4 The subsequent introduction of pulsed radio frequency (RF) power supplies solves the limitation that DC power supplies can only measure conductive materials5, making GDoES more widely used to measure both conductive and non-conductive materials 6-8
Development and application of GDoES
gdoes hairexhibition
Light source
Millisecond or microsecond pulsed glow discharges (PGDS) devices use a discontinuous excitation mode to attenuate the thermal effect on the sample surface As a result, PGDS can use a higher excitation power than traditional light sources in continuous excitation mode, enhancing the excitation or ionization process 9 This not only improves the sensitivity of GDO testing, but also solves the problem of sample destruction due to overheating Although pure argon is the most commonly used glow discharger, analysis with a mixture of gases has also attracted the attention of the academic community Researchers mix helium, krypton, hydrogen, or oxygen with pure argon for glow discharge tests to obtain high-resolution depth spectra 10-13 These basic studies have improved the sensitivity of GDOES analysis, reduced the interference factors in the analysis, and greatly promoted the progress of GDOES technology and the expansion of application fields
Detectors
Traditional glow discharge emission spectrometers generally use photomultiplier tubes (PMTs) as detectors Because the working voltage of PMTs is generally 1000 V, the large volume and the inability to achieve full-wavelength spectral detection limit the development of GDOEs to high-efficiency, small-scale and full-spectrum 14 The GDA150HR of the German Spectruma company and the GDS500 glow discharge emission spectrometer of the American LECO company both use charge-coupled detectors (CCDs), which have the characteristics of low dark current, high sensitivity, and high signal-to-noise ratio 15 Because it is an ultra-small large-scale integrated component, it can be made into a line array and area array detector, which can realize the simultaneous recording of multiple spectral lines For any one element, there are many spectral lines to choose from, which can cover the full content range For a specific content range of elements, selecting several spectral lines for analysis at the same time can improve the reliability of the analysis results However, at present, CCD also has its limitations in detection, that is, it cannot realize the detection of ppm-level trace elements and the response speed is slow For example, the analysis of impurities in high-purity metals, especially for the analysis of non-metallic element impurities, there is still a big gap between its accuracy and PMT, because the response is slow, and the depth of the measured impurities may deviate from its actual position Therefore, at present, CCD is applied to GDoES with relatively low detection requirements, and mainstream high-end GDoES still uses detectors based on PMT technology
Application of gdoes
Features of GDOes:
GDoES has a number of advantages 23-27 : the matrix effect is small, for samples of different compositions and structures, the sputtering process occurs on the surface of the sample, while the excitation process is in the plasma, and the sample matrix has almost no effect on the signal of the measured substance;Low power, low energy of sputtered ions for GDO depth spectroscopy, layer by layer peeling;Low energy level excitation, the measured sample atoms are mainly excited by the collision of electrons in the plasma, and because the energy carried by the electrons is small, the excitation of the atoms is at a low energy level, and the resulting spectral lines are often simple atomic or ion spectral lines, so the interference between the spectral lines is smallAll elements (including hydrogen) can be detected, and when the discharge is stabilized, the sample atoms entering the plasma can be resolvedThe self-absorption effect is small, and the confined light source makes the plasma thickness of the sample excitation small, resulting in a small self-absorption effect, and the linear range of the calibration curve is wideThe analysis speed is fast and the depth resolution is high, it only takes a few minutes to make a sample, the analysis speed reaches a few microns per minute, and the depth resolution can reach the sub-nanometer levelDirect detection of solid samples, no need to dilute, dissolve, etcThe detection depth range is wide, from the initial microscale to the current nanoscale
The fast sputtering rate of GDO depth profiling requires the detector to quickly collect all the spectral information emitted from the glow chamber, and at the same time, due to the rapid change of element content in the process of sputtering to different depths of the sample, especially to various interfaces, the detector is required to have a timely and accurate response to this Horiba Jobin Yvon's GD-Profiler series glow discharge emission spectrometers all use its patented technology of high dynamic detector (HDD), which is essentially a photomultiplier tube, but by automatically adjusting the voltage of the photomultiplier tube, the linear dynamic range can reach 109 orders of magnitude 16, so that the instrument has a linear response to 1 10-6% 100% concentration change, no signal saturation, no need for preset voltage, so as to carry out fast and sensitive detection This response is often not possible with solid-state detectors or ordinary high-voltage fixed PMT inspection systems
Application of gdoes
Due to its extremely high sputtering rate, detectability of all elements, including hydrogen, and high depth resolution, GDOEs are widely used for profiling of elemental compositions such as steel and automotive surface coatings, surface antimicrobial, electroplating, semiconductors, and vapor deposition films17-20 Improvements in light sources have made it possible for GDOes to characterize extremely thin films (less than 5 nm thick)21 Quantitative analysis of the depth spectra of thiourea monolayer GDOs has demonstrated that the depth resolution of GDOs can reach the sub-nanometer level 22 In recent years, nano-multilayer films have developed rapidly in the fields of protective multilayer films, transparent conductive films, photovoltaics, and flexible electronic devices 28-31 , which also brings a huge demand for the analysis and characterization of nano-multilayer films The fast, detectable and high-resolution nature of all elements (including hydrogen) is a feature of GDOEs that is suitable for other commonly used depth profiling techniques such as secondary ion mass spectrometry (SIMS), Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS) do not have
Protective multilayer films are widely used in marine, aviation, and military fields because they can resist harsh working environments 32-33 As the structure of protective multilayer films becomes more complex, researchers have paid more attention to how to distribute internal elements Therefore, the characterization of protective multilayer films using GDoes has become a research hotspot Marin et al.34 analyzed Al2O3 with three different layer structures using GDoES TiO2 nano-protective multilayer film, comparing the film thickness measured by GDoes with the film thickness measured by atomic force microscopy, the measurement results are relatively consistent In addition, through the analysis of the measured GDOES results, it was found that the roughness of the substrate and the roughness of the surface of the single-layer nanolayer film were on the same order of magnitude Liu et al. 35 pairs of CRCNs on stainless steel substrates GDOES analysis of CR nano-protective multilayer film showed that the CR layer was not a pure metal, and the CR layer also contained C and N elements, which played an important role in improving the mechanical and tribological properties of the multilayer film
Transparent conductive multilayer films are widely used in optoelectronic devices due to their ideal characteristics such as good conductivity and optical transparency GDoEs play an important role in the process control and quality inspection of transparent conductive multilayer films Wang Yu et al.36 The prepared azo cu azo composite film was annealed under vacuum conditions of 10-3 pa for 1 h, and the composite film before and after annealing was tested by GDOES, and it was found that the signal intensity of the Cu layer was weakened after annealing, and the azo layer** The Cu signal is present, which indicates that the interlayer diffusion of the composite layer film occurs during the annealing process, which destroys the original structure of the composite layer film Ma et al.37 prepared SiO2 Ga2O3 multilayer films with different SiO2 contents by atomic layer deposition, and analyzed SiO2 with 0%, 5% and 20% SiO2 doped by GDoes The depth distribution of the concentrations of each element was obtained from the Ga2O3 multilayer film, which proved that the samples met the expected Si doping requirements, and the SiO2 layer was uniformly distributed among the Ga2O3 layers
GDOEs are widely used in the internal mechanism of photovoltaics** Lee et al.38 performed GDOES analysis on a solar cell structured as glass ITO PeDo:PSS halogen perovskite PCBM AG, observed the reversible migration of iodide ions at room temperature under positive or negative bias pressure (on the minute scale), and deduced the diffusion coefficient and mobility of iodide ions to be 1., respectively3 10-12 cm2 s and 5 10-11 cm2 v·s Neugebohrn et al. 39 GDOES analysis of the distribution of Na elements in the MO MoSe2 system showed that the content of Na in the Mo layer was related to the amount introduced in the pre-preparation process, and the Na elements in the MOSE2 layer were mainly concentrated on the surface of the sample and at the MO MoSe2 interface Khalil et al. 40 pairs of Cu2Znsns4 (CZTS) The MO samples were tested by GDOES, and it was observed that the elements in the CZTS layer were uniformly distributed, and there was no increase in the ZN signal intensity at the CZTS MO interface, which indicated that there was no segregation of Zns at the interface, but such segregation was reported in other literatures
Most of the flexible electronic devices are based on organic matter, so they are relatively fragile and easy to be destroyed in depth analysis experiments, and PULSED-RF-GDOES can solve such problems well Zhou Gang et al.41 analyzed a flexible optical functional film containing silver under different pulsed-rf-gdo operating parameters, and concluded that the sample was a multilayer film composed of AlxO1-X, ZnxO1-X, Ag and Cu Lv et al. 42 The Pulsed-RF-GDOES was used to analyze the infrared radiation blocking film samples, and the intensity-time spectrum of PULSED-RF-GDOES was combined with the TOF-SMIS data and quantitative analysis, and finally the layer structure and depth distribution of the samples were obtained
MRI model for GDO depth spectrum quantification
How gdoes work
When the voltage on both sides of the electrode exceeds the voltage required to excite argon, the glow discharge phenomenon will be generated, and the ionization will produce ar+ and free electrons, and the light source is a low-temperature plasma at this time Ar+ is accelerated by the electrode and bombarded on the surface of the sample at the cathode, and the surface atoms are sputtered into the plasma, which collides with the electrons and becomes an excited state When the sample atoms in the excited state return to the ground state, the characteristic spectrum of the sample elements is generated The holographic grating splits the spectrum and analyzes it by the detector, and finally the luminous intensity of the corresponding element in the sample can be obtained by computer processing The working principle of GDO is shown in Figure 1
Figure 1Schematic diagram of how gdoes workfig.1schematic diagram of gdoes working principle
Depth Spectrum Quantitative Analysis—MRI Model
The Mixing Roughness Information (MRI) model, proposed by Hofmann 43, takes into account the three physical effects involved in depth profiling techniques that lead to the distortion of the measured depth spectrum, namely, the mixing of atoms during sputtering, the surface and interface roughness of the sample, and the depth of information of the probe signal This model has been widely used in the quantitative analysis of AES, XPS, SIMS, and GDO depth spectra 44-47 In the MRI model, the normalized intensity of the measured signal IIo can be expressed as a convolution of the original concentration distribution x(z) and the depth resolution function g(z-z)48
The depth resolution function g(z) consists of the following three subfunctions 49 :
Eq. (2) In Eq. (3), w is the length of the atom mixture in the collision cascade, is the roughness, is the depth of information, gw describes the atomic mixing produced by ion sputtering, g gives the sample surface (boundary) roughness, is the Gaussian distribution, g describes that the measured signal comes from a certain depth range from the sample surface
There are three basic parameters involved in the MRI model, W, and , all of which have definite physical meanings 50: Atom mixing length w (nm), when ions bombard the surface of the sample, the outermost atoms are stripped off the surface of the sample, and at the same time, due to the cascade effect caused by the atomic bombardment, the atoms on the surface of the sample are mixed with the internal atoms, which changes the distribution of the original elements on the surface of the sample, resulting in the broadening of the depth spectrum, and the mixing length is used to characterize the mixing degree between atoms; The sputtering conditions, such as the energy of the incident ion, the type of ion and the angle of incidence, determine the size of the atom mixing length w, and the longer the mixing length, the greater the distortion of the depth spectrum. Roughness (nm), roughness of the sample surface (boundary) is the main cause of distortion of the measurement depth spectrum When the material to be measured is polycrystalline, the channel effect (i.e., grains with different orientations with different sputtering yields) will cause the sputtering-induced roughness to increase significantly with the increase of sputtering depth, which will greatly reduce the depth resolution Information depth (nm), the probed signal may originate from the depth range below the surface, which can also lead to distortion of the measured depth spectrum In GDoES and SIMS depth profiling, Since the probe signal comes from the most superficial layer, it can be considered to be zero while in AES and XPS, part of the signal comes from the electrons under the surface layer of the sample, the lower the kinetic energy of the collected electrons, the smaller the corresponding depth of information, and the better the resolution of the depth profiling
The effect of roughness on the measurement depth spectrum can be seen as the integral of the original component depth distribution x(z) with the Gaussian function 51
Considering the optimal sputtering in the sputtering process, it is generally assumed that the instantaneous sputtering rate is linearly related to the composition 52-53, which is satisfied
Thus, the sputtering time is converted to the sputtering depth
In Eq. (6), Qi and Xi are the sputtering rate and component concentration of pure element I, and Qt is the average sputtering rate
Taking into account both optimal sputtering and atomic mixing, the surface concentration of element I during sputtering
Satisfy the following differential equation 54
Considering the effect of the depth of information on the intensity, the normalized depth profiling signal strength II IO 55
In order to characterize the degree of distortion of the measurement depth profile, depth resolution is introduced, which is an important indicator to judge the quality of the measurement depth spectrum The definition of the traditional depth resolution δz (16% 18%) 56 is: assuming an ideal, atomic monolayer interface a b, when the normalized intensity of the measured signal decreases from 84% to 16% or rises from 16% to 84%, as shown in Figure 2 The higher the resolution of the depth profile, the closer the measured depth spectrum is to the true component depth distribution Hofmann et al. 57 summarized the definition and measurement of depth resolution in intersectional, δ-layer, monolayer, and multilayer depth profiling spectra, and analyzed the applicability of δz (16% to 18%), width at half maximum (FWHM), and δz (FWHM) for interinterface, δ-layer, monolayer, and multilayer applications
Figure 2Depth resolution δz(16% 18%)fig.2definition of depth resolution δz
In the depth analysis experiment of nano multilayer film, the highest (low) intensity of the measured element may not appear as a plateau, and the traditional definition of 84% to 16% cannot be used to determine the depth resolution, and the corresponding depth resolution can only be calculated by quantitative analysis of the depth spectrum, such as the quantitative analysis of the measured depth spectrum by using the MRI model, and the optimal sputtering parameter r ( is obtained
atomic mixing length w, roughness , and depth of information, the corresponding depth resolution can be expressed as 58:
Assuming that the influence of each distortion factor on the depth resolution is independent of each other, the corresponding depth resolution can be expressed by the following formula, the "tunnel effect" is the main factor causing the distortion of the GDOES depth spectrum, which is caused by the unevenness at the bottom of the sputtering pit 3 How to consider this effect in the quantitative analysis of depth spectrum is the focus of quantitative analysis of GDO depth spectrum The tunnel effect in GDOES depth analysis is directly related to roughness, and can generally be considered by the dynamic change of roughness parameters However, if the tunnel effect is very obvious and cannot be described by the Gaussian function in the MRI model, the specific morphology of the sputtering tunnel must be considered Liu Yi et al. 59 On the basis of the MRI model, the CRer-Simulation (CRAS) model for simulating the tunnel topography 60 was introduced, and the MRI-CRAS model was established and used for the quantitative analysis of GDO depth analysis
In the MRI-CRAS model, the sputtering pit is a circular area with a radius rmax, and any radial position in the pit is represented as Rreal, or as a dimensionless radius R, which is expressed as r = rreal Rmax The measurement signal i(t) can be regarded as the sum of all measurement signals on the surface of the tunnel, where ilocal represents the measurement intensity at r in the tunnel, and k is the normalization factor Considering the influence of various factors, it is rigorously deduced (see Ref. 59 for the detailed derivation process). When p 1 the tunnel is convex, when p 1 the tunnel is concave, when p = 1 the tunnel effect disappears, when p is constant, the larger b is the greater the tunnel curvature
The depth component function dwf(z) is derived, where p 0 and b 0
The MRI model was used to calculate the functional relationship of the surface concentration and mMRI with sputtering time t considering roughness and mixing effects, and the signal intensity calculated by the MRI-CRAS model was 59 by substituting it into the CRAS model
Quantitative analysis of GDOES high-resolution depth spectroscopy
SiO2 that is naturally oxidized on the surface of monocrystalline silicon
The naturally growing SiO2 layer on the surface of a monocrystalline silicon wafer is about 1 2 nm thick, which can be used to evaluate the detection resolution of depth profiling techniques Zhou Gang et al.41 First, Pulsed-RF-GDOES was used to perform depth profiling tests on the standard sample SiO2 (300 nm) Si(111), and the optimal working parameters were obtained: argon working air pressure 650 Pa, sputtering power 20 W, pulse frequency 10000 Hz, duty cycle 05. The obtained depth spectrum is shown in Figure 3 The wavy shape of the SiO2 film signal (i.e., the signal of Si and O) is caused by the interference of reflected light caused by the flat Si substrate, and the Si signal in SiO2 is slightly different from the pure Si signal of the substrate in intensity, so it is determined that the intensity of the SiO2 (300 nm) layer The half-peak width of the oxygen signal (intensity) in the time spectrum is 455 s, the sputtering rate of the SiO2 layer was 66 nm s Under the same operating parameters, the depth profile measurement of the naturally oxidized SiO2 thin layer on the Si(111) substrate is shown in Figure 4 The experimental results clearly show that the signal-to-noise ratio of the oxygen signal is low, so silicon elements are used to determine the thickness of the naturally oxidized SiO2 thin layer Intensity of the naturally oxidized SiO2 thin layer in Figure 4 The half-peak width of the silicon signal (intensity) in the time spectrum is only 015 s, from which a thin layer of SiO2 naturally oxidized on a Si(111) substrate was obtained with a thickness of about 1 nm (66 nm/s，0.15 s) It is clear that the depth resolution of this depth spectrum is less than 1 nm
Figure 3sio(300 nm) Si(111) standard PULSED-RF-GDOES intensity time spectrumfig.3sio（300 nm）/si（111）standard sample pulsed-rf-gdoes intensity-sputtering time spectrum
Figure 4Si(111) substrates are naturally oxidized SiOPulsed-RF-GDO intensity time spectrumfig.4pulsed-rf-gdoes intensity-sputtering time spectrum of sionaturally grown on the si （111） substrates
Monolayer thiourea molecule
Shimizu et al. 61 used pulsed-RF-gdoes for the first time to perform depth analysis tests on the thiourea (CH4N2S) monolayer (less than 1 nm) growing on the vertical copper surface, and obtained depth spectral data Liu Yi et al.22 used the MRI model to reconstruct the monolayer thiourea molecule based on the known molecular structure of thiourea (Fig. 5(a)) The specific fitting method is as follows: the molecular structure of the monolayer thiourea is projected in the direction of the vertical copper substrate, and is regarded as a multi-layered structure; Then, the MRI model was used to preliminarily fit the experimental data of S and N. Considering the impurity contamination of element C on the surface of the sample, the original distribution of element C was changed, and the theoretical calculated value of MRI and the measured data of C were best fitted by iterative method. The original layer structure of thiourea molecule and the obtained carbon contaminated layer structure were substituted into the MRI model to calculate, and the sputtering rate was assumed to depend on the sputtering depth, so as to realize the fitting of the measurement data of each element of thiourea molecule, the best fitting results are shown in the solid line of Fig. 5(b), and the relationship between sputtering rate and sputtering depth is shown in Fig. 5(c). When the best fit was reached, the three MRI parameters were w=025 nm，σ=0.15 nm, =0 nm According to equation (11), the depth resolution of the thiourea monolayer GDOES depth spectrum is calculated as z=05 nm, indicating sub-nanometer resolution
Figure 5(a) Molecular structure diagram and projection of thiourea, (b) Measured depth spectra (data points) and MRI fitting results (solid line) of thiourea single molecule, carbon impurity layer and copper substrate, and (c) Sputtering rate with splash depth obtained by MRI fittingfig.5（a） schematic of thiourea molecular structure and its projected，（b） measured depth profiles ofthiourea monomer layer， carbon impurity layer and copper substrate， and mri fitting results （solid lines）and （c）sputtering rate as a function of the sputtered depthobtained by mri fitting
MO B4C Si multilayer optical film
MO B4C Si nanolayer has been widely used as a reflective coating in free electron lasers and extreme ultraviolet (EUV) devices63, where the extremely thin B4C layer acts as a diffusion barrier layer
BER et al.64 measured 60 cycles of Mo(3 nm) B4C(0.) deposited on a Si(111) substrate using Pulsed-RF-Gdoes3 nm）/si（3.7 nm) nano-multilayer film, the measurement results are shown in Figure 6 As can be seen from Figure 6, 0The structure of the 3 nm-thick B4C (B's signal) layer is still resolvable
Figure 6mo/bC SI SAMPLE PULSED-RF-GDOES INTENSITY TIME SPECTRUMfig.6mo/bc/si sample pulsed-rf-gdoes intensity-sputtering time spectrum
Yang Hao et al. 65 used the MRI model to quantify the Mo, B, and Si signals in the measured GDO depth spectrum at 15 35 s sputtering time The specific process is as follows: The SRIM program 66 calculates the atomic mixing length w of Mo, Si and B, which are 0., respectively3，0.8 and 06 nm, and the MO B4c Si sample roughness is 07 nm；Based on the total thickness of the sample and the total sputtering, the average sputtering rate range is roughly estimated. Based on the sputtering yield of 500 ev argon ions bombarding each substance, the relative sputtering rate ratio of each element was estimated. Under the framework of the MRI model, the best fit was obtained by changing the layer structure and MRI parameters and using the iterative method (shown in the solid line in Fig. 7)65, and the thickness of each layer obtained by reconstruction is listed in Table 1, and the thickness of the layer determined by Table 1 65 is very different from the nominal value. The mixing length w represents the degree to which the surface atoms and the inner atoms are mixed during sputtering, w=06 nm (already less than 1 nm) indicates a low degree of mixing, while parameter = 07 nm indicates that the surface of the sample was originally flat and did not produce an increase in roughness during sputtering, which is mainly due to the low-energy bombardment of GDoES (the energy of Ar+ is only 50 eV); Equation (11) is used to calculate the depth resolution of the depth spectrum of MO B4C Si nanolayer multilayer gdoes, where Z=11 nm．
Figure 7Pulsed-RF-GDOES depth profiling experimental data (hollow circles).fig.7pulsed-rf-gdoes depth profiling data （open circles） and the bestmri fitted results （solid lines）
a）mo；（b）si；(C) Best MRI fitting results (solid line) for B4C layer
a） mo；（b） si；（c） b4c layers
Table 1 The thickness of each layer was determined by using the best-fitting pulsed-rf-gdoes depth spectrumtable 1 the individual layer thickness determined by best fit of pulsed-rf-gdoes depth profiling data
Comparison of depth resolution of MOSI multilayer GDOs and SMIS
Both SIMS and GDOES use ions to bombard the surface of the sample to analyze the sputtered sample particles SIMS detects the sputtered ion mass spectrometry 68, while GDOES detects the sputtered atomic (ion) spectrum The detection time of GDOES is much smaller than the SIMS detection time, but the sensitivity and resolution of SIMS are generally higher than that of GDOES, and SIMS will be a good choice when the concentration of the measured element is extremely low (up to 10-9) 69 In the following section, the depth spectra of MOSI multilayer SIMS and GDOES are quantitatively analyzed to compare the resolution of the depth profiles of the two
MOSI multilayer films are widely used in nanolithography, soft X-ray or extreme ultraviolet microscopy and astronomical observations, etc. 60-73 Ber et al.64 to 60 cycles of MO (35 nm）/si（3.5 nm) nano-multilayer film was measured by GDO and SIMS depth profiling, and the results are shown in Figure 8 It can be seen from Figure 8 that the maximum peak intensity of the MO signal decreases with the sputtering time, while the peak-to-valley intensity shows an upward trend, which is due to the gradual increase of the roughness of the sputtered surface during sputtering due to continuous ion bombardment The MRI model was used to quantitatively analyze the MO signal in the first ten cycles of the depth profile of the MOSI multilayer film and fit the MO B4C with the previous section In order to quantitatively characterize the variation of the depth spectrum of Mo element in MO Si multilayer film with sputtering depth, it can be assumed that the roughness parameter also increases with the increase of sputtering depth Through the best fitting of the depth spectra of Mo element SIMS and GDoes in MO Si multilayer film (Fig. 8 solid line), the roughness parameters are 04 ~0.66 nm and 067~1.The 2 nm interval varies, compared to the sample roughness given in Ref. 6753 nm is close; The mixing length w is 065 nm and 053 nm is taken as the average of the roughness parameters and used to calculate the depth resolution z, i.e., sims=053 nm and gdoes=094 nm using equation (11) to get zsims=138 nm，∆zgdoes=2.At 16 nm, it is clear that the depth resolution of the SIMS depth spectrum is higher than that of GDos, but the difference is very limited for MOSI multilayers
FIGURE 8GDOES, SIMS and best MRI fitting results for MO elementsfig.8gdoes and sims of mo element and optimal mri fitting results
a） gdoes ；（b） sims
Conclusion
The glow discharge emission spectrometer is constantly updated in terms of light source, power supply and detector, which continuously improves the analytical ability and detection accuracy of GDO depth profiling technology, and achieves sub-nanometer depth resolution Therefore, GDO can characterize molecular layers, nano-multilayer films and polymers, etc., and its future development direction will largely depend on the characterization requirements of high-tech products The research of MRI model in the quantitative analysis of depth profiling will promote the deep profiling technology to the research of flexible devices, Development of quantitative analysis of extremely thin film layers and monolayer Due to the advantages of sub-nanometer depth resolution, fast analysis rate (m min), detectable of all elements (including hydrogen) and small matrix effect, it is believed that GDêEs deep profiling technology will receive more attention from industry and academia in the future, and will play a more important role in the field of testing and analysis