Evaluation of ionic conductivity for Mg–Al layered double hydroxide intercalated with inorganic anions

This study demonstrates that humidity, temperature, and the interlayer anions influence ionic conductivities of Mg–Al layered double hydroxides (LDHs) intercalated with inorganic anions. Results show that Mg–Al LDH intercalated with Br^? exhibited the highest ionic conductivity among Mg–Al LDHs intercalated with CO₃^2?, Cl^?, Br^?, NO₃^? and SO₄^2?. Its ionic conductivity was 1.1 × 10^? 2 S cm^? 1 at 80 °C under 80% relative humidity. The electromotive force for the hydroxide ion concentration cell using Mg?Al CO₃^2? LDH showed the same behavior with that using an anion exchange membrane, indicating that Mg–Al CO₃^2? LDH can be a hydroxide ion conductor.     

ESR dosimetric properties of talc and zircon

Two silicate minerals talc (Mg₃Si₄O₁₀(OH)₂) and zircon (ZrSiO₄) with different crystal lattice structures were subjected to ESR dosimetric studies. Zircon shows anisotropic ESR signals at g_xx = 2.0168, g_yy = 2.0076 and g_zz = 2.0033, which have been identified as a hole-center associated with Y³⁺ substituted at Zr⁴⁺ sites. Other characteristic signals were observed and identified. The ESR signal at g = 2.0033 showed positive response to γ-irradiation at 110 Gy and is suitable for use in dosimetry and dating of natural zircons.Talc (a magnesium sheeted silicate) exhibits ESR derivative spectrum characterized by the presence of Fe³⁺ at g = 4.28 and the hf-sixtet Mn²⁺ signals. This is due to the possible substitution of Fe³⁺ and Mn²⁺ in the Mg²⁺ octahedral sites, respectively. The enhancement of the Mn²⁺ sixtet by γ-irradiation increases the area occupied by the signals that makes it difficult for use with dosimetric applications.     

Synthesis,characterisation, luminescence and defect centres in solution combustion synthesised CaZrO3:Tb3+ phosphor

Tb³⁺ doped CaZrO₃ has been prepared by an easy solution combustion synthesis method. The combustion derived powder was investigated by X-ray diffraction, Fourier-transform infrared spectrometry and scanning electron microscopy techniques. A room temperature photoluminescence study showed that the phosphors can be efficiently excited by 251 nm light with a weak emission in the blue and orange region and a strong emission in green light region. CaZrO₃:Tb³⁺ exhibits three thermoluminescence (TL) glow peaks at 126 °C, 200 °C and 480 °C. Electron Spin Resonance (ESR) studies were carried out to study the defect centres induced in the phosphor by gamma irradiation and also to identify the centres responsible for the TL peaks. The room temperature ESR spectrum of irradiated phosphor appears to be a superposition of two distinct centres. One of the centres (centre I) with principal g-value 2.0233 is identified as an O^? ion. Centre II with an axial symmetric g-tensor with principal values g_=1.9986 and g_?=2.0023 is assigned to an F⁺ centre (singly ionised oxygen vacancy). An additional defect centre is observed during thermal annealing experiments and this centre (assigned to F⁺ centre) seems to originate from an F centre (oxygen vacancy with two electrons). The F centre and also the F⁺ centre appear to correlate with the observed high temperature TL peak in CaZrO₃:Tb³⁺ phosphor.     

Ground-based solar absorption studies for the Carbon Cycle science by Fourier Transform Spectroscopy (CC-FTS) mission

Carbon cycle science by Fourier transform spectroscopy (CC-FTS) is an advanced study for a future satellite mission. The goal of the mission is to obtain a better understanding of the carbon cycle in the Earth's atmosphere by monitoring total and partial columns of CO₂, CH₄, N₂O, and CO in the near infrared. CO₂, CH₄, and N₂O are important greenhouse gases, and CO is produced by incomplete combustion. The molecular O₂ column is also needed to obtain the effective optical path of the reflected sunlight and is used to normalize the column densities of the other gases. As part of this advanced study, ground-based Fourier transform spectra are used to evaluate the spectral region and resolution needed. Spectra in the 3950–7140 cm^?1 region with a spectral resolution of 0.0042 cm^?1 recorded at Kiruna (67.84°N, 20.41°E, and 419 m above sea level), Sweden, on 1 April 1998, were degraded to the resolutions of 0.01, 0.1, and 0.3 cm^?1. The effect of spectral resolution on the retrievals has been investigated with these four Kiruna spectra. To obtain further information on the spectral resolution, optical components and spectroscopic parameters required by the future mission, high-resolution solar absorption spectra between 2000 and 15000 cm^?1 were recorded using Fourier transform spectrometers at Kitt Peak (31.9°N, 111.6°W, and 2.1 km above sea level), Arizona, on 25 July 2005 and Waterloo (43.5°N, 80.6°W, and 0.3 km above sea level), Ontario, on 22 November 2006 with spectral resolutions of 0.01 and 0.1 cm^?1, respectively. Dry air volume mixing ratios (VMRs) of CO₂ and CH₄ were retrieved from these ground-based observations. The HITRAN 2004 spectroscopic parameters are used with the SFIT2 package for the spectral analysis. The measurement precisions for CO₂ and CH₄ total columns are better than 1.07% and 1.13%, respectively, for our observations. Based on these results, a Fourier transform spectrometer (maximum spectral resolution of 0.1 cm^?1 or 5 cm maximum optical path difference (MOPD)) operating between 2000 and 15000 cm^?1 is suggested as the primary instrument for the mission. Further progress in improving the atmospheric retrievals for CO₂, CH₄, and O₂ requires new laboratory measurements of the spectroscopic line parameters.     

Absolute line intensities and self-broadening coefficients in the ν3 − ν1 band of carbon dioxide

Using a tunable diode-laser spectrometer, we have measured the self-broadening coefficients and strengths of 26 absorption lines in the ν₃ ? ν₁ band of ¹²CO₂ and ¹³CO₂ at room temperature. These lines, ranging from P(34) to R(40), are located around 960.9 and 913.4 cm^?1, respectively for the ¹²CO₂ and ¹³CO₂ molecules. The collisional widths and the intensities were obtained by fitting Voigt and Rautian and Sobel’man profiles to the measured shapes of the lines. From the individual line intensities and using a least-squares method, we have determined the vibrational band strength as well as the Herman–Wallis factors for the ν₃ ? ν₁ band of ¹²CO₂ and ¹³CO₂.     

Effects of Li2CO3 as a secondary lithium source on the LiFePO4/C composites prepared via solid-state method

Li₂CO₃ was used as the secondary lithium source for the synthesis of LiFePO₄/C composites via a solid-state reaction method by adopting Li₃PO₄ as the main lithium source. The main purpose of using Li₂CO₃ is to compensate for the partial lithium loss during the sintering while reducing the usage of excess Li₃PO₄. In this study, the effects of Li₂CO₃ amount on the phase, structural and electrochemical properties of LiFePO₄/C material were systematically investigated. The samples were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), constant-current charge–discharge test and cyclic voltammetry (CV). The results showed that by adding an appropriate amount of Li₂CO₃, the impurities, e.g. Li₃PO₄, normally appearing in the final product, could be excluded. It was found that LiFePO₄/C with Li₂CO₃ in 6% excess (vs. stoichiometric LiFePO₄) exhibited the best electrochemical performance, which delivered initial discharge capacities of 141.7, 125.2, 119.9 and 108.9 mAh g^?1, respectively, at 0.5, 1, 2 and 5C rates. The capacity was reduced to 113.4 mAh g^?1 after 50 cycles at 2C rate, with capacity retention rate of 94.6%.     

CO2 pressure broadening and shift coefficients for the 1–0 band of HCl and DCl

CO₂ broadened spectra of the 1–0 band of H³⁵Cl and H³⁷Cl, observed near 2886 cm^?1, and the 1–0 band of D³⁵Cl and D³⁷Cl, located near 2089 cm^?1, have been recorded at room temperature and five total pressures between 150 and 700 Torr, using a Bruker IFS125HR Fourier transform spectrometer. Spectra of pure HCl were also recorded. CO₂ broadening and shift coefficients of HCl and DCl have been measured using multi-spectrum non-linear least squares fitting of Voigt profiles. The analysis of the 1–0 band of DCl was complicated by the presence of overlapping CO₂ bands, which were included in the treatment as absorption coefficients calculated taking line-mixing effects into account.     

Effect of high CO2 concentration on char formation through mineral reaction during biomass pyrolysis

O₂/CO₂ combustion has attracted considerable attention as a promising technology for CO₂ capture. Using biomass for fuel is considered carbon neutral, and O₂/CO₂ biomass combustion can mitigate the deleterious environmental effect of greenhouse. In this study, the effect of CO₂, the main component gas in O₂/CO₂ combustion, on the pyrolysis characteristics of biomass is investigated. Cellulose, lignin, and metal-depleted lignin pyrolysis experiments were performed using a thermobalance. Information on the surface chemistry of the chars was obtained by Fourier transform infrared (FTIR) spectroscopy to investigate changes in the surface chemistry during pyrolysis under different surrounding gasses. When the temperature increased to 1073 K at heating rate of 1 K s^?1, the char yield of lignin in the presence of CO₂ increased by about 10% compared with that under Ar. However, for cellulose and metal-depleted lignin, no significant difference appeared between pyrolysis under CO₂ and that under Ar. FT-IR showed that a strong peak corresponding to carbonate ions appeared in the char derived from lignin under CO₂. Therefore, salts such as Na₂CO₃ or K₂CO₃ formed during the lignin pyrolysis under CO₂. At around 1650–1770 cm^?1, a significant difference appeared in the FTIR spectra of chars formed under CO₂ and those formed under Ar. C=O groups not associated with an aromatic ring were found only in chars formed under CO₂. It was suggested that these salts affected the char formation reaction, in that the char formed during lignin pyrolysis under CO₂ had unique chemical bands that did not appear in the lignin-derived char prepared under Ar.     

High-pressure measurements of CO2 absorption near 2.7 μm: Line mixing and finite duration collision effects

Room-temperature, high-pressure (1–30 atm) measurements of CO₂ absorption are carried out near 2.7 μm to study line mixing and finite duration collision effects on transitions in the ν₁+ν₃ and 2ν₂+ν₃ vibrational bands. Two distributed feedback diode lasers are used to measure CO₂ transitions near 3631–3635 cm^?1 and 3644–3646 cm^?1, and an FTIR spectrometer covers the entire ν₁+ν₃ and 2ν₂+ν₃ bands from 3500 to 3800 cm^?1. The experiments are carried out in CO₂–air and CO₂–Ar mixtures to observe the non-ideal effects under the influence of different perturbers. Measurements are compared with simulations using the Voigt line shape to analyze the deviation from the Lorentzian behavior with increasing gas density, and show significant deviation from this model at high gas densities. Line shape models using empirical corrections or dynamically based scaling laws are evaluated by comparison to the measured high-density spectra. Although none of the models is able to predict the measured spectra accurately, the line mixing model of Niro et al. [24] does an overall good job but overestimates the band centers by about 4–9%. In light of these observations, challenges of developing a CO₂ sensor for high-pressure combustion applications are discussed.     

Modeling midwave infrared muzzle flash spectra from unsuppressed and flash-suppressed large caliber munitions

Time-resolved infrared spectra of firings from a 152 mm howitzer were acquired over an 1800–6000 cm^?1 spectral range using a Fourier-transform spectrometer. The instrument collected primarily at 32 cm^?1 spectral and 100 Hz temporal resolutions. Munitions included unsuppressed and chemically flash suppressed propellants. Secondary combustion occurred with unsuppressed propellants resulting in flash emissions lasting ～100 ms and dominated by H₂O and CO₂ spectral structure. Non-combusting plume emissions were one-tenth as intense and approached background levels within 20–40 ms. A low-dimensional phenomenological model was used to reduce the data to temperatures, soot absorbances, and column densities of H₂O, CO₂, CH₄, and CO. The combusting plumes exhibit peak temperatures of ～1400 K, areas of greater than 32 m², low soot emissivity of ～0.04, with nearly all the CO converted to CO₂. The non-combusting plumes exhibit lower temperatures of ～1000 K, areas of ～5 m², soot emissivity of greater than 0.38 and CO as the primary product. Maximum fit residual relative to peak intensity are 14% and 8.9% for combusting and non-combusting plumes, respectively. The model was generalized to account for turbulence-induced variations in the muzzle plumes. Distributions of temperature and concentration in 1–2 spatial regions demonstrate a reduction in maximum residuals by 40%. A two-region model of combusting plumes provides a plausible interpretation as a ～1550 K, optically thick plume core and ～2550 K, thin, surface-layer flame-front. Temperature rate of change was used to characterize timescales and energy release for plume emissions. Heat of combustion was estimated to be ～5 MJ/kg.     

Shock tube/laser absorption studies of the decomposition of methyl formate

Reaction rate coefficients for the major high-temperature methyl formate (MF, CH₃OCHO) decomposition pathways, MF → CH₃OH + CO (1), MF → CH₂O + CH₂O (2), and MF → CH₄ + CO₂ (3), were directly measured in a shock tube using laser absorption of CO (4.6 μm), CH₂O (306 nm) and CH₄ (3.4 μm). Experimental conditions ranged from 1202 to 1607 K and 1.36 to 1.72 atm, with mixtures varying in initial fuel concentration from 0.1% to 3% MF diluted in argon. The decomposition rate coefficients were determined by monitoring the formation rate of each target species immediately behind the reflected shock waves and modeling the species time-histories with a detailed kinetic mechanism [12]. The three measured rate coefficients can be well-described using two-parameter Arrhenius expressions over the temperature range in the present study: k₁ = 1.1 × 10¹³ exp(?29556/T, K) s^?1, k₂ = 2.6 × 10¹² exp(?32052/T, K) s^?1, and k₃ = 4.4 × 10¹¹ exp(?29 078/T, K) s^?1, all thought to be near their high-pressure limits. Uncertainties in the k₁, k₂ and k₃ measurements were estimated to be ±25%, ±35%, and ±40%, respectively. We believe that these are the first direct high-temperature rate measurements for MF decomposition and all are in excellent agreement with the Dooley et al. [12] mechanism. In addition, by also monitoring methanol (CH₃OH) and MF concentration histories using a tunable CO₂ gas laser operating at 9.67 and 9.23 μm, respectively, all the major oxygen-carrying molecules were quantitatively detected in the reaction system. An oxygen balance analysis during MF decomposition shows that the multi-wavelength laser absorption strategy used in this study was able to track more than 97% of the initial oxygen atoms in the fuel.     

Structural and vibrational properties of PVT grown Bi2Te3 microcrystals

High-quality Bi₂Te₃ microcrystals have been grown by physical vapor transport (PVT) method without using a foreign transport agent. The microcrystals grown under optimal temperature gradient are well facetted and they have dimensions up to ～100 μm. The phase composition of grown crystals has been identified by X-ray single crystal structure analysis in space group R3?m, a=4.3896(2) Å, b=30.5019(10) Å, Z=3 (R=0.0271). Raman microspectrometry has been used to describe the vibration parameters of Bi₂Te₃ microcrystals. The FWHM parameters obtained for representative Raman lines at 61 cm^?1 and 101 cm^?1 are as low as 3.5 cm^?1 and 4.5 cm^?1, respectively.     

Structural investigation of Te-based chalcogenide glasses using Raman spectroscopy

Structural changes of metals (Zn, Sb, In, Ga) and metal halides (AgI, ZnI₂, CdI₂, PbI₂, BiI₃) modified GeTe₄ glasses were investigated with the aid of Raman spectroscopy. The Raman spectra of these glasses in the frequency region between 100 cm^?1 and 300 cm^?1 display four main bands at about 124, 140, 159 and 275 cm^?1 which are contributed by Ge–Te, Te–Te, Te–Te and Ge–Ge vibration modes. The intensity of 159 cm^?1 and 275 cm^?1 bands vary with the addition of different glass modifiers. While the relative intensity of the 124 cm^?1 and 140 cm^?1 bands are insensitive to composition changes. Glass modifiers like Zn, In and Sb act as glass network unstabilizer which will disorganize the glass network by opening up the chain structures of Ge–Te and Te–Te. In the case of Ga and metal halides, Ga can open up Ge–(Te–Te)_4/2 tetrahedra and form Ga–(Te–Te)_3/2 triangle. Iodine can form covalent bonds with tellurium and decrease the tendency of microcrystal formation. Thus both Ga and iodine ultimately act as glass network stabilizer.     

ArF excimer laser deposition of wide-band gap solid electrolytes for thin film batteries

High quality solid electrolyte thin films was grown by pulsed laser deposition (PLD) using a high photon energy ArF excimer laser. Various amorphous thin films were successfully deposited on glass substrates from oxide targets; such as Li₃PO₄, LiBO₂, Li₂SiO₃, Li₂CO₃, Li₂SO₄, Li₂ZrO₃, LiAlO₂, Li₂WO₄ and Ohara glass ceramics. The morphology, optical property and ionic conductivity of these films were examined by optical microscope, UV–VIS spectroscopy and impedance analysis. Dramatic improvement of the film morphology was observed by using a high photon energy laser, while the broken film with many droplets was obtained by using lower ones. Ionic conductivity of the films was examined by impedance spectroscopy and dc polarization method. For example, an ionic conductivity of a Li₃PO₄ film was 4.6 × 10^? 6 S cm^? 1 at 25 °C with activation energy of 0.57 eV. Electronic conductivity measurements revealed that most of the films were pure lithium ion conductors, while a Li₂WO₄ film was a mixed conductor.     

First high-resolution analysis of the ν15, ν12, ν5, ν10 and ν2 bands of oxirane

Fourier transform spectra of oxirane (ethylene oxide, c-C₂H₄O) have been recorded in the 730–1560 cm^?1 (6.4–13.7 μm) spectral region using a Bruker IFS125HR spectrometer at a resolution of 0.0019 cm^?1. A total of six vibration bands, ν₁₅, ν₁₂, ν₅, ν₃, ν₁₀ and ν₂, have been observed and analyzed. The corresponding upper state ro-vibrational levels were fit using Hamiltonian matrices accounting for various interactions. Satisfactory fits were obtained using the following polyads {15¹, 12¹, 5¹} and {10¹, 2¹} of interacting states. As a result, an accurate and extended set of Hamiltonian constants were obtained. The following band centers were derived: ν₀ (ν₁₅) = 808.13518(60) cm^?1, ν₀ (ν₁₂) = 822.27955(37) cm^?1, ν₀ (ν₅) = 876.72592(15), ν₀ (ν₃) = 1270.37032(10) cm^?1, ν₀ (ν₁₀) = 1471.35580(50) cm^?1 and ν₀ (ν₂) = 1497.83309(15) cm^?1 where the uncertainties are one standard deviation.     

High resolution emission spectroscopy of the E2Π–X2Σ+ transition of SrH and SrD

Emission spectra of SrH and SrD have been studied at high resolution using a Fourier transform spectrometer. The molecules have been produced in a high temperature furnace from the reaction of strontium metal vapor with H₂/D₂ in the presence of a slow flow of Ar gas. The spectra observed in the 18 000–19 500 cm^?1 region consist of the 0–0 and 1–1 bands of the E²Π–X²Σ⁺ transition of the two isotopologues. A rotational analysis of these bands has been obtained by combining the present measurements with previously available pure rotation and vibration–rotation measurements for the ground state, and improved spectroscopic constants have been obtained for the E²Π state. The present analysis provides spectroscopic constants for the E²Π state as ΔG(½) = 1166.1011(15) cm^?1, B_e = 3.805503(32) cm^?1, α_e = 0.098880(47) cm^?1, r_e = 2.1083727(89) Å for SrH, and ΔG(½) = 839.1283(23) cm^?1, B_e = 1.918564(15) cm^?1, α_e = 0.034719(23) cm^?1, r_e = 2.1121943(83) Å for SrD.     

Comment on “Spectroscopic database of CO2 line parameters: 4300–7000 cm–1”

The “Spectroscopic database of CO₂ line parameters: 4300–7000 cm^–1” constructed by Toth et al., has been considered in relation with our previous and current studies of the absorption spectrum of carbon dioxide (CO₂) by high-sensitivity CW-cavity ring down spectroscopy (CW-CRDS) in the 5850–7000 cm^?1 region. Part of the line parameters of the database are based on accurate spectroscopic measurements by Fourier transform spectroscopy (FTS) but Toth et al. have chosen to fix to a very low value (4×10^?30 cm/molecule) the lower intensity cut off. This value which is far below the FTS detection limit has led to long range extrapolations to high J values and to the inclusion of weak unobserved bands which were theoretically predicted. In the 5850–7000 cm^?1 region, most of these calculated transitions were previously observed by CW-CRDS. The comparison with the CW-CRDS ¹³CO₂ spectrum in this region, has evidenced that (i) many weak bands above the intensity cut off are missing; (ii) there are important deviations between the line parameters provided in the database and our previous observations both for line positions (up to 1.7 cm^?1) and line intensities (up to a factor 80). Our discussion was limited to the three ¹³C species (¹³C¹⁶O₂, ¹⁶O¹³C¹⁸O and ¹⁶O¹³C¹⁷O) but the conclusions should apply to the other isotopologues in particular ¹²C¹⁶O₂ and to the full spectral range of the database.Alternatively, the global effective operators models for CO₂ can reproduce satisfactorily all the experimental line positions and line intensities available in the literature. This polyad model, which has been developed for most of the CO₂ isotopologues, constitutes an interesting alternative for the most accurate and complete CO₂ database. In particular, very weak bands, accidental resonances, intensity transfers and extra lines are accurately accounted for and predicted by this polyad model.     

Determination of l-proline based on anodic electrochemiluminescence of CdTe quantum dots

A novel flow injection method for detection of l-proline was proposed in the presence of CdTe quantum dots (QDs). This method is based on the enhanced anodic electrochemiluminescence (ECL) emission of CdTe QDs l-proline in aqueous system. CdTe QDs were modified with thioglycolic acid to obtain stable water-soluble QDs and intensive anodic ECL emission in Na₂CO₃–NaHCO₃ buffer solution at an indium tin oxide (ITO) electrode, which was used for the sensitive detection of ECL enhancement using our homemade flow cell. Under the optimal conditions, the ECL intensity was correlated linearly with the concentration of l-proline over the range of 1.0×10^?8?1.0×10^?4 g mL^?1 (r=0.9996) and the detection limit was 5.0×10^?9 g mL^?1. The relative standard deviation was 1.12% for 6.0×10^?5 g mL^?1 l-proline (n=11). The possible mechanism was discussed. This method put forward a new efficient ECL methodology for enhancement-related determination of l-proline successfully.     

First analysis of the 3ν9−ν9 hot band of difluoroboric acid (BF2OH). Further evidence of large amplitude effects for the OH torsion

The hot band 3ν₉?ν₉ of the isotopologue ¹¹BF₂OH (difluoroboric acid) located at 1034.78 cm^?1 was investigated for the first time by Fourier transform infrared spectroscopy. During previous studies both, the ν₉ mode (OH-torsion relative to the BF₂ moiety, at 522.87 cm^?1) and the ν₄ mode (in-plane OH bend) had been shown to exert large amplitude motion, and splittings of 0.0051 and 0.0038 cm^?1 had been observed in the interacting 2ν₉ and ν₄ bands located at 1042.87 and 961.49 cm^?1, respectively. The present work establishes large amplitude effects also for the 9³ excited state located at 1557.655 cm^?1. Numerous P and R transitions of the 3ν₉–ν₉ hot band were identified in the 2ν₉ manifold, and doublets corresponding to a torsional splitting of 0.031 cm^?1 in the 9³ state were observed. The vibrational assignment of the 9³ state was confirmed by the detection of the 3ν₉?2ν₉ hot band Q branch in the 19 μm region.     

Morphometric partitioning of the respiratory surface area and diffusion capacity of the gills and swim bladder in juvenile Amazonian air-breathing fish, Arapaima gigas

The gills and the respiratory swim bladders of juvenile specimens (mean body mass 100 g) of the basal teleost Arapaima gigas (Cuvier 1829) were evaluated using stereological methods in vertical sections. The surface areas, harmonic mean barrier thicknesses and morphometric diffusing capacities for oxygen and carbon dioxide were estimated. The average respiratory surface area of the swim bladder (2173 cm² kg^?1) exceeded that of the gills (780 cm² kg^?1) by a factor of 2.79. Due to the extremely thin air–blood barrier in the swim bladder (harmonic mean 0.22 μm) and the much thicker water–blood barrier of the gills (9.61 μm), the morphometric diffusing capacity for oxygen and carbon dioxide was 88 times greater in the swim bladder than in the gills. These data clearly indicate the importance of the swim bladder, even in juvenile A. gigas that still engage in aquatic respiration. Because of the much greater diffusion constant of CO₂ than O₂ in water, the gills also remain important for CO₂ release.