FIG 3 - uploaded by Annett Thøgersen
Content may be subject to copyright.
XPS depth profile showing the P2p spectrum from the substrate surface of the sample with Cr dopant. 

XPS depth profile showing the P2p spectrum from the substrate surface of the sample with Cr dopant. 

Source publication
Article
Full-text available
Phosphorus diffusion in p-type silicon wafers with Fe or Cr impurities has been investigated using X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy. Silicon wafers doped with phosphorous are heavily used in semiconductor devices. It is, therefore, of crucial importance to determine their compositions profile. The XPS P2p...

Context in source publication

Context 1
... of charge carriers and solar cell efficiency. 5 Reducing P 2 O 5 to elemental P in the surface region of the Si wafer can cause the concentration briefly to exceed that of a saturated solid solution. This leads to an environment in which a variety of non-equilibrium processes can occur and in turn potentially cause the anomalous behaviour. 6 In order to optimize the emitter and P diffusion process further, it is important to study the mechanisms behind phosphorus diffusion in silicon and the compositional elements involved in the process. Although numerous studies have been devoted to the chemistry of the surface of Si substrates, the chemical states present in P-doped Si substrates have not been investigated. 1 When downscaling the ultra large scale integrated circuit, P doping and distribution in the Si substrate and its native oxide will influence the properties of device heavily. 1,7 In addition, the structure and morphology of the Si surface and its native oxide will both influence the break down field strength and leakage current in the metal-oxide- semiconductor (MOS) device. 1 In this study, we have been examining the various oxidation states present in P gettered Si substrates contaminated with Fe or Cr, using X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy (TEM). P-type monocrystalline 310 l m thick wafers were inten- tionally contaminated with iron (56Fe) or chromium (52Cr) in the float zone (FZ) by ion implantation at room temperature. The implantation was carried out at an energy of 700 keV and dose of 3 : 1 Â 10 10 cm 2 . The samples were then annealed at 900 C for 1 h, yielding a homogeneous im- purity distribution. Subsequently, a surface layer of 10 l m in thickness was removed by chemical etching in a mixture of HNO 3 , HF, and CH 3 COOH in order to eliminate any influence at the implantation-induced damage. A spin on source of P 2 O 5 mixed with SiO 2 was sprayed onto the surface opposite to the one subjected to ion implantation. This was followed by annealing at different time and temperature conditions in the range from 748 C to 994 C. After diffusion, the phosphorus silicate glass (PSG) was removed in 5% HF. Cross-sectional TEM samples were prepared by ion- milling using a Gatan precision ion polishing system with 5 kV gun voltage. The samples were analysed by high resolution TEM (HRTEM) and energy filtered TEM (EFTEM) in a 200 keV JEOL 2010F microscope with a Gatan imaging filter and detector. XPS was performed in a KRATOS AXIS ULTRA DLD using monochromatic Al K a radiation ( h 1⁄4 1486 : 6 eV) on plan-view samples at zero angle of emission (vertical emission). The X-ray source was operated at 10 mA and 15 kV. The spectra were peak fitted using Casa XPS 8 after subtraction of a Shirley type background. Figure 1 shows a SEM image of the surface of the wafer after removing the P O mixed with SiO . Small pores in the surface of the Si wafer are visible in the images. Figure 2 shows a TEM image of two of the pores, with filtered images using only certain reflections. Crystalline material was observed in the bottom of the pores and fast Fourier trans- forms (FFT) of the high resolution images shown in the three bottom photos of Figure 2 indicate the presence of P 4 O 10 . These pores may have been created during the chemical etching in the mixture of HNO 3 , HF, and CH 3 COOH when removing 10 l m of the surface layer after the implantation. The cross-sectional compositions of the wafers were acquired by XPS depth profiling and peak fitting of the XPS peaks. The Si-2p spectra were peak fitted similar to the work of Cerofolini et al. 9 and Lu et al. , 10 as we carried out in our previous work. 11 Si was fitted with an asymmetrical peak with a tail length of 6.5, tail scale of 0.6%, and 70% Gaussian has been found to provide the best fit (GL(30)T(6.5)). The SiO x peaks (Si þ ; Si 2 þ ; Si 3 þ , and Si 4 þ ) were then fitted with a pure Gaussian function GL(0). The full width half maximum (FWHM) for Si-2p 1 = 2 and Si-2p 3 = 2 was set at 0.65 eV as has been reported before by, e.g., Peden et al. 12 and Himpsel et al. 13 for monochromatic X-rays. However, fitting the Si 0 peaks shows a variation in FWHM between 0.35 eV and 0.8 eV with depth. This could be due to amorphization of the structure due to Ar etching. The FWHM values of Si 2 O, SiO, Si 2 O 3 , and SiO 2 were set at 0.8 eV, 1.1 eV, 1.1 eV, and 1.2– 1.5 eV, respectively, as in our previous paper. 11 The binding energies of Si ð Si 0 3 = 2 Þ ; Si 2 O ð Si þ 3 = 2 Þ , SiO ð Si 2 3 þ = 2 Þ ; Si 2 O 3 ð Si 3 3 þ = 2 Þ , and SiO 2 ð Si 4 3 þ = 2 Þ , were found to be 99.4 eV, 100.4 eV, 101.4 eV, 102.5 eV, and 103.6 eV, respectively. 11 An additional Si 3 = 2 peak at about 99.5 eV has been peak fitted to sev- eral of the spectra near the substrate surface. This could be attributed to the presence of either Si-P or SiH x . However, this peak has no compositional and depth correlation with the P-2p (P À ) peak in Figure 3 or 4. It is then more probable that it is due to SiH x as we have previously found near the surface in HF etched Si samples. 11 The orthorhombic P 2 O 5 phase is assembled by discrete molecules that consist of PO 4 tetrahedra. The O1s peak for P O has been reported to be fitted with two components relating to the P O and P-O-P bonds. These two corresponding peaks should have an area ratio 1:1.5, as there should be one non-bridging oxygen atom and 1.5 bridging-oxygen atom for every phosphorus atom. 14 The shift in binding energy between the two components is due to a difference in charge density. The non-bridging oxygen atom has a reported charge density of À 0.96, while the bridging atom 0.65. This difference results in a shift towards a lower binding energy of the non-bridging atom peak with respect to the bridging-oxygen-atom peak. The binding energy of the non-bridging oxygen (P 1⁄4 O) has been reported to be 531.7 eV, while that of the bridging oxygen (P-O-P) peak was 533.5 eV. 14 This fits well with the binding energies found for the peak at the surface of the wafer, 532.2 eV and 533.7 eV, corresponding to P-O-P and P 1⁄4 O, respectively. The O1s peak with a binding energy of 533 eV corresponds to the SiO x compositions. An oxide peak at a binding energy of 532.3 eV is also fitted to the spectra (peak P 1 in Figures 5 and 6). This O1s peak ð O ð P 1 ÞÞ is the only peak at a depth of 8 nm below the substrate for the Cr doped sample, and 18 nm for the Fe doped, and may be due to a SiO x or Si-P-O phase. At a depth range of 4–11 nm below the substrate surface, the spectra have been fitted with a peak ð O ð P 3 ÞÞ at 534.3 eV (Fe doped) and 534.9 eV (Cr doped). This depth corresponds well with the bottom of the holes seen in Figure 2, and is probably due to P 4 O 10 . The plasmon peak of the Si2p peak is located in the region of the P2p peak. The peak position of the plasmon peak is about 133.5 eV and has been fitted in the spectra. The P2p peak in Figures 3 and 4 was fitted with the two components P 1 = 2 and P 3 = 2 , corresponding to spin-orbit coupling. The P2p peak of P 2 O 5 (P 5 3 þ = 2 ) has been reported to have binding energy of 134.3 eV, while unoxidized P (P 0 3 = 2 ) has a binding energy of 129.4 eV. 1 The P2p spectra for the two samples have been fitted with five P 3 = 2 , as well as five P 1 = 2 components. The binding energies of the five P 3 = 2 components are listed in Table I. The P 3 = 2 peak with a binding energy of 134 eV is in good agreement with the literature values for P 2 O 5 (134.3 eV (Ref. 1)). This peak is intense in the outermost surface which has been covered by P 2 O 5 , as explained in the Sec. II. The peak with a binding energy of 129.6 eV is attributed to the 2P 3 = 2 component of pure P (P 0 ). The peaks were fitted with a spin-orbit coupling of 0.8 eV. The peak with a slightly higher binding energy of 130.2 eV is probably due to P þ , since it has been stated that the binding energy for the donor ion (P þ ) should be 1 eV higher than for elemental P (P 0 ). 1 The peaks at lower binding energy, 129 eV, may be due to P with a negative oxidation number in, e.g., Si-P or interstitial or substitutional P. This fits well with the binding energy of 129 eV, which was found for SiP by Perego et al. 15 It has been proposed that the dominant chemical state of P in the surface region is not oxidized-P but elemental P and/or P bound with Si (Si-P). 7 At a depth of 4–11 nm below the substrate surface, a P-oxide peak with a higher binding energy, 135.5 eV, is observed. This value fits well with the reported binding energy of 135.3 eV for P 4 O 10 . 16 It should be mentioned here that at this depth (4–11 nm), the O(P 3 ) peak of the O1s was observed which was assigned to P 4 O 10 . The binding energy shift between the P 0 3 = 2 and the Si 0 3 = 2 peak, D E B ð P À Si Þ 1⁄4 E B P 0 3 = 2 À E B Si 0 3 = 2 , in the top 4 nm of the substrate is 29.1 eV, while deeper down in the substrate D E B ð P À Si Þ is 29 : 6 eV 6 0 : 12 eV. The binding energy reference for P 0 3 = 2 is 129.9 eV (Ref. 16) and for Si 0 3 = 2 99.3 eV, 16 result- ref ing in a value D E B ð P À Si Þ of 30.6 eV. This value is higher than what was found for the two samples. We have discussed the various reasons for changes in binding energy and chemical shift in detail in previous work. 17 In the present work, the energy separation between the peak P-2p 3 = 2 and Si-2p 3 = 2 is used to monitor the relative positions of the two peaks with depth. An increase of the P-2p 3 = 2 - Si-2p 3 = 2 energy separation implies either a reduced screening of the P-2p 3 = 2 and/or an increased screening of the Si-2p 3 = 2 . The reduced energy separation with respect to the literature values is in agreement with the reduced screening of the Si core hole after P-doping. The increased energy separation with depth (from 29.1 eV at surface to 29.6 eV at depth 18 nm) indicates a reduced effect of P with depth. Figure 7 shows a plot ...

Citations

... 27 The binding energy peaks are positioned at 132.62 and 133.34 eV in Figure 2D are assigned to P P and P O bonds, respectively. 28 Figure 2E shows the core level spectrum of O 1s with peaks detected at 530.81 and 531.96 eV which could be attributed to the presence of Mn O, Co O and P O compounds. 29 Figure 2F and π À π * (satellite peak/electrons transition) bonding, respectively. ...
Article
Mn and Co are known for their redox storage mechanism depending on the kinetics of their ions, which is required to boost capacity values of the material electrodes for energy storage applications. However, to the best of our knowledge, Co-Mn phosphate material-based electrochemical capacitors have been rarely investigated. In this work, CoMn(PO4)2 synergized with graphene foam (GF) was synthesized following an easy, direct and low-temperature hydrothermal method. The study investigated the synergy between Co, Mn and GF-supported composite, with a strong P-O covalent tie occurring at the surface of the material, leading to a high degree of polarization and very low relaxation time. The CoMn(PO4)2/GF material tested as a half-cell could achieve a specific capacity of 58.15 mAh g⁻¹ at 0.5 A g⁻¹. A CoMn(PO4)2/GF//activated carbon (AC) hybrid device was made up considering CoMn(PO4)2/GF and an AC material as positive and negative electrodes, respectively. The CoMn(PO4)2/GF//AC hybrid device achieved a specific capacity of 45.89 mAh g⁻¹, specific energy of 51.53 Wh kg⁻¹ and a corresponding specific power of 561.52 W kg⁻¹ at 0.5 A g⁻¹. Furthermore, the device displayed a 95% capacity retention after 10 000 GCD cycles at 12 A g⁻¹ and showed improved performance over its initial specific capacity after a voltage holding test for over 110 hours at 2 A g⁻¹.
... The O1s peak originates from SiO x structures and Si2s/Si2p [34] peaks at binding energies of 154.8/103.7 eV from the SiO 2 bond. Furthermore, phosphorusas the P2p signal corresponding to the characteristic of the P-O bond at a binding energy of 135.5 eV [35] was detected in the annealed samples of the sets 2 and 3. The presence of phosphorus in the B 2 O 3 /Sb 2 O 5 -Si samples is surprising and can be explained only by contamination during the ALD or RTA processes. ...
Article
Full-text available
Oxides containing group III or group V elements (B2O3/Sb2O5and P2O5/Sb2O5) were grown by plasma‐assisted atomic layer deposition (ALD) onto single‐crystalline silicon and serve as dopant sources for conformal and shallow doping. Transport phenomena in ALD‐oxide‐Si structures during rapid thermal annealing (RTA) were investigated subsequently by X‐ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM) and secondary ion mass spectrometry (SIMS). The XPS and TEM analyses of the annealed ALD‐oxide‐Si structures demonstrate that the ALD oxide converts to a silicon oxide and partially evaporates during annealing. In addition, dopant‐containing, spherical, partially crystalline particles were found to form in the oxide, and Si‐P precipitates at the oxide/Si interface. After diffusion annealing at 1000 °C the SIMS analyses reveal phosphorus and boron concentration profiles in the silicon substrate with maximum concentrations exceeding their solid solubility limits by roughly one order of magnitude. Experimental doping profiles of phosphorus and boron in silicon are compared with simulation results considering a slight injection of self‐interstitials and dynamical defect clustering. This article is protected by copyright. All rights reserved.
... In the P 2p spectrum, the binding energy peaks at 130.39 and 131.12 eV inFig. X (d)are correspond to P-P and P-O bonds respectively[46].Fig. 2(e) shows the spectrum of O 1s with fitted peaks at 528.43 and 529.68 eV which could be ascribed to O 1s in (Ni-O), (Co-O)and P-O compounds[47].Fig. ...
... 28 A strong peak observed at 134.6 eV and another peak located at 129.8 eV can be assigned to the P−O bonds and Si−P, respectively, based on the P 2p 1/2 . 29 Two additional peaks were also observed in the spectra, which could be due to the presence of impurities during a growth and/or contamination issue during chemical treatment. Due to spin− orbit interaction, the Ga 2p peak split into two peaks corresponding to Ga 2p 3/2 (1118.8 ...
... The increase in width of the peak with EBI dose suggesting a greater number of irradiation induced defects and the formation of several localized chemical states in Si/SiO x interface. The thickness of the SiO x layer at the surface has been calculated from the intensities of XPS Si2p spectra by using the method described by Watts and Wolstenholme [30]. The calculated SiO x thickness (d) was found to be 9.4 nm for the pristine and 9.2 nm at the irradiation dose of 1500 kGy. ...
Article
The effect of electron beam irradiation (EBI) on Al/n-Si Schottky diode has been studied by I–V characterization at room temperature. The behavior of the metal-semiconductor (MS) interface is analyzed by means of variations in the MS contact parameters such as, Schottky barrier height (ΦB), ideality factor (n) and series resistance (Rs). These parameters were found to depend on the EBI dose having a fixed incident beam of energy 7.5 MeV. At different doses (500, 1000, 1500 kGy) of EBI, the Schottky contacts were prepared and extracted their contact parameters by applying thermionic emission and Cheung models. Remarkably, the tuning of ΦB was observed as a function of EBI dose. The improved n with increased ΦB is seen for all the EBI doses. As a consequence of which the thermionic emission is more favored. However, the competing transport mechanisms such as space charge limited emission, tunneling and tunneling through the trap states were ascribed due to n > 1. The analysis of XPS spectra have shown the presence of native oxide and increased radiation induced defect states. The thickness variation in the MS interface contributing to Schottky contact behavior is discussed. This study explains a new technique to tune Schottky contact parameters by metal deposition on the electron beam irradiated n-Si wafers.
... According to the aforementioned considerations, P species appearing at P2p = 131 eV originates from the reduction of the pyrophosphate species containing significantly reduced oxygen amount. It must be noted that the peak at 131 eV cannot be attributed to elemental P o as the latter appears according to the literature at 130 eV [38,39]. Thus, it is proposed that the low binding energy peak originates from the catalytic reduction of pyro/poly-phosphate species by H 2 on the Pt surface through probably the abstraction of hydroxyl groups, thus leaving a phosphorous-oxygen polymeric structure which can account for the lower binding energy of the P2p. ...
Article
Full-text available
The poisoning effect and the role of H3PO4 (PA) at the anodic electrocatalytic layer of a high temperature polymer electrolyte membrane (HT PEM based on ADVENT TPS®) fuel cell are discussed under the light of cyclic voltammetry, CO stripping, and X-ray photoelectron spectroscopy (XPS) experiments. The catalytic layer was based on both the pyridine-modified multi-wall carbon nanotubes, 30 wt% Pt/(ox.MWCNT)–Py, and on commercial 30 wt% Pt/C, with varying PA loadings on the electrode. At low PA loadings (<3 gPA/gPt), the electrochemically active surface area of Pt decreases significantly under H2 anode long-term operation, approaching surface Pt utilization <10 %. This degradation is attributed to the formation of pyrophosphoric or triphosphoric acid as well as catalytically H2 reduced PA species, which block the Pt surface area. As was explicitly detected by means of XPS PA species were displaced from the Pt surface under H2 or CO exposure. The poisoning effect is reversible as these species can be hydrated back to orthophosphoric acid. The reduced species can be reoxidized into PA at 750 mV versus RHE. On the other hand, the electrochemical interface is stable at PA loadings exceeding 3 gPA/gPt, thus approaching Pt surface utilization >80 % in the long term. This is believed to be a consequence of the more uniform distribution of PA, thus eliminating the PA displacement from the Pt interphase. It is hypothesized that the minimization of the PA poisoning effect at PA > 3 gPA/gPt, may also be a result of more efficient hydration of the catalytic layer that is being achieved through the hydration of the PA in the membrane and in the catalyst layer by the cathodically produced water vapors.
Article
Full-text available
Trionic effect is a vital excitonic physical phenomenon, which intensively affects the optical and optoelectronic properties of 2D materials. Violet phosphorus (VP) is another allotrope of elemental phosphorus with robust photoluminescence (PL) emission in the visible range. So far, experimental investigations of the excitonic behavior in VP are quite scarce. Herein, the evolution of the PL mechanism in synthesized VP crystals against the Ar+ plasma exposure is investigated with emphasis on a conversion from trion to exciton emission. The estimated trion binding energy of VP is ≈109 meV, relatively larger than common layered materials. By analyzing the chemical states and the atomic structures, the conversion mechanism is proposed as follows. The Ar+ plasma treatment reduces the stannous SnIP impurities’ population, which are incorporated into the VP lattices and serve as the n-type dopants leading to the trion formation. Besides, various surface defects (POx) can promote the trion-to-exciton conversion by withdrawing electrons from VP in the process and act as hole-trap centers to enhance the photodetection of VP phototransistors. This work reveals that the layered VP crystal can provide an ideal platform to study the excitonic physics and future trionic devices at the 2D limit.
Article
The development of efficient and durable bifunctional electrocatalyst from non‐precious metals is crucial for electrochemical water splitting, hitherto hindered by sluggish kinetics in alkaline conditions. Herein, echinop‐like bimetallic CoNiP/CoNi alloy was synthesized from the metal‐organic framework (MOF) serving as an efficient catalyst for oxygen evolution reaction (OER), with a low overpotential of 300 mV in 1 M KOH at 10 mA cm −2 ( η 10 ). The cooperative effect of Ni and Co metals, along with interfacial properties of the integrated semiconducting phosphide/metallic alloy, and electronic conductivity of MOF‐derived carbon, regulated the catalytic performance of the catalyst. Besides, the bimetallic CoNiP/CoNi alloy catalyst was interspersed with N‐doped graphene forming triad catalyst, demonstrated superior activity towards hydrogen evolution reaction ( η 10 = 150 mV) and excellent durability, derived from their strong H binding to high valence Co ions, interfacial effects of triad catalyst, large electrochemically active surface area, and enhanced conductivity from N‐doped graphene. The stability of carbon‐contai n ing catalyst during OER (oxidation) is altered by the high reactivity of heteroatom. The assembled CoNiP/CoNi/N‐RGO||CoNiP/CoNi water electrolyzer delivers a reasonable cell potential of 1.76 V at 10 mA cm ‐2 . The synthesized bimetallic CoNiP/CoNi alloy‐based triad catalyst thus demonstrated excellent electrocatalytic activity and high durability suitable for efficient alkaline water splitting.
Article
The adjustability of the incorporation and distribution of P dopants in Si materials has led to their application in various electronic and photovoltaic devices. In particular, P-doped Si films have been investigated for dopant-induced phenomena such as oxidation enhancement and dopant segregation during thermal oxidation. Despite their technical as well as scientific importance of dopant behaviors in P-doped Si films, these were hardly observed due to low P incorporation. In this study, we prepared and oxidized P-doped epitaxial Si (Si:P) films grown on the Si(100) substrate using in-situ P-doped deposition technique to incorporate high P concentration up to 9.6% (4.8 × 10²¹ cm⁻³) in the films, exceeding solid solubility limits (∼3 × 10²⁰ cm⁻³). We found that in oxidized samples with P concentration greater than 6.2%, defects were generated in the near-surface region of Si:P layers and segregated P was observed at the defected areas. The chemical bonding states of P at the oxidized Si:P film increased in the P 2p peak intensity at the interface between the oxide and the Si:P layer, indicating P segregation at this interface. In the strain states of Si:P films, the decrease in the out-of-plane lattice parameter was observed, while maintaining the constant in-plane lattice parameter, implying strain release due to decreased P concentration in the Si:P layer after dry oxidation.
Article
In this paper, castor oil based UV-cured coatings are prepared via thiol-ene click reaction by varying molar content of a self-synthesized phosphorus-containing flame retardant (DA), a self-synthesized castor oil based-mercaptan (COSH) and 1,3,5-triethylene-1,3,5-trimethylcyclotriazane (TTC), which is used for the study of synergistic effectivities for flame retardant component, photopolymerization kinetics, and thermal degradation. The results demonstrate the synergistic effectivities between DA and TTC via effective quantification of Lewin’s parameter (ES). Under the molar ratio of DA, TTC and COSH of 3:3:5, the thermal stability of coatings is significantly better than others and the char yield can reach 14.54% at 800 °C under air condition despite achieving better flame retardant properties, and the conversion of thiol and vinyl groups can reach 70% or more. Further, scanning electron microscope (SEM) and X-ray photoelectron spectroscopy (XPS) tests show that polyphosphoric acid formed through the decomposition of DA can promote SiO and SiC bond into compact and stable char layer timely to protect wooden substrates and delay burning, which indicates the synergistic fire retardant property of P and Si in the condensed phase. Moreover, there are nonflammable gases (CO2 and SO2) released from the coatings to dilute oxygen concentration during thermal degradation by the analysis of thermogravimetry/infrared spectroscopy (TG-FTIR) tests, which also leads to the generation of porous char layer on the surface of wooden substrates. Besides, the oxygen can be also consumed by the phosphorus-containing compounds generated by DA simultaneously, which shows the synergistic fire retardant property of S and P in the gas phase.