Osmium Nanoparticles-Polypropylene Hollow Fiber Membranes Applied in Redox Processes

Composite membranes play a very important role in the separation, concentration, and purification processes, but especially in membrane reactors and membrane bioreactors. The development of composite membranes has gained momentum especially through the involvement of various nanoparticles, polymeric, oxide, or metal, that have contributed to increasing their reactivity and selectivity. This paper presents the preparation and characterization of an active metal nanoparticle-support polymer type composite membrane, based on osmium nanoparticles obtained in situ on a polypropylene hollow fiber membrane. Osmium nanoparticles are generated from a solution of osmium tetroxide in tert-butyl alcohol by reduction with molecular hydrogen in a contactor with a polypropylene membrane. The composite osmium-polypropylene hollow fiber obtained membranes (Os-PPM) were characterized from the morphological and structural points of view: scanning electron microscopy (SEM), high resolution SEM (HR-SEM), energy dispersive spectroscopy analysis (EDAX), X-ray diffraction analysis (XRD), Fourier transform Infrared (FTIR) spectroscopy, thermal gravimetric analysis, and differential scanning calorimetry (TGA, DSC). The process performance was tested in a redox process of p-nitrophenol and 10-undecylenic (10-undecenoic) acid, as a target substance of biological or biomedical interest, in solutions of lower aliphatic alcohols in a membrane contactor with a prepared composite membrane. The characteristics of osmium nanoparticles-polypropylene hollow fiber membranes open the way to biological and biotechnological applications. These membranes do not contaminate the working environment, operate at relatively low temperatures, provide a large contact area between reactants, allow successive oxidation and reduction operations in the same module, and help to recover the reaction mass by ultrafiltration. The results obtained show that the osmium-polypropylene composite membrane allows the reduction of p-nitrophenol or the oxidation of 10-undecylenic acid, the conversion depending on the concentration in the lower aliphatic alcohol, the nature of the lower aliphatic alcohol, and the oxidant or reducing flow through the membrane contactor.


Introduction
The development of membranes and membrane processes involving composite membranes has gained great importance with the progress made in obtaining nanoparticles with a well-controlled composition, shape, and size distribution [1,2]. Obviously, the nature of the nanoparticles used in obtaining membranes is of overwhelming importance for selec-

Introduction
The development of membranes and membrane processes involving composite membranes has gained great importance with the progress made in obtaining nanoparticles with a well-controlled composition, shape, and size distribution [1,2]. Obviously, the nature of the nanoparticles used in obtaining membranes is of overwhelming importance for selectivity, reactivity, or effect on membrane processes ( Figure 1). Thus, polymeric nanoparticles based on carbon, oxide, metals, or their compounds have been used for the preparation of composite membranes both in separation processes and in various reaction processes: esterification, etherification, reduction, and oxidation [2][3][4]. At the same time, intense studies have followed the influence of the size and shape of nanoparticles from a certain material on the performance of composite membranes [5,6]. The composite membranes made using metal nanoparticles (silver, gold, copper, nickel, or palladium) have a special place, which brought significant improvements in selectivity and reactivity [7][8][9][10]. As outlined in Figure 1, the use of nanoparticles in the realization of composite membranes and the processes based on them has brought important benefits to users, such as:
Composite membranes based on metal nanoparticles fit in a very special way in all these directions of development of membrane processes. However, the reactivity or catalytic activity are among the best studied and used characteristics of metal nanoparticlepolymer based composite membranes [27][28][29][30][31]. Although they have special properties, osmium and its compounds have been used relatively little in membrane processes [32,33], the main application being the metallization of polymeric or biological membranes for studies in electron microscopy [34].
Due to its chemical inertia, even against aggressive environments [35], osmium can be an interesting membrane material for the construction of membrane reactors aimed at hydrogenation or oxidation [30,36].
Making metal-polymer composite membranes is of interest from the perspective of increasing the physical-chemical and process performances, especially when it is represented by a chemical reaction. The use of osmium as a membrane material can be an economic disadvantage due to the high price, but when used as nanoparticles generated from an accessible material, such as osmium tetroxide, this aspect is overcome (Figures S1-S4).
At the same time, osmium obtained by reducing osmium tetroxide waste on a porous polymeric support can be a useful means of recovery, especially since in catalytic reactions, it is known for selectivity [37,38], regiospecificity [39,40], and stereospecificity [41].
The characteristics of osmium nanoparticles-polypropylene hollow fiber membranes open the path to biological and biotechnological applications. We can mention that these membranes:

•
Do not contaminate the working environment, • Operate at relatively low temperatures, • Provide a large contact area between reactants, • Help to recover the reaction mass by ultrafiltration, and • Allow successive oxidation and reduction operations in the same module.
This paper presents the preparation and characterization of a composite membrane of active metal nanoparticle-support polymer type, based on osmium nanoparticles obtained in situ on a polypropylene hollow fiber membrane and its use in the transformation of 10-undecylenic acid or p-nitrophenol by oxidation and reduction.

Materials
The reagents used in the presented work were of analytical grade and were used without other purification.
The purified water characterized by 18.2 µS/cm conductivity was obtained with a RO Millipore system (MilliQ®Direct 8 RO Water Purification System, Merck KGaA, Darmstadt, Germany).
Polypropylene hollow fiber membranes at a length of 700 mm and a diameter of 0.3 mm are assembled so as to achieve a filtering surface of 1 m 2 [52,54]. The fibers have a porosity of 40% [54,55] and a pore diameter of less than 0.2 µm [56] (Figure 2). This type of module with polypropylene hollow fiber membranes allows operation over the entire pH range, in oxidizing or reducing solutions at temperatures up to 100 • C without dimensional changes and can be sterilized by known physical-chemical methods [57][58][59]. Polypropylene hollow fiber membranes at a length of 700 mm and a diameter of 0.3 mm are assembled so as to achieve a filtering surface of 1 m 2 [52,54]. The fibers have a porosity of 40% [54,55] and a pore diameter of less than 0.2 µm [56] (Figure 2). This type of module with polypropylene hollow fiber membranes allows operation over the entire pH range, in oxidizing or reducing solutions at temperatures up to 100 °C without dimensional changes and can be sterilized by known physical-chemical methods [57][58][59].

Preparation of Osmium Nanoparticles on Polypropylene Hollow Fiber Membranes
To make osmium nanoparticles-polypropylene hollow fibers composite membranes, on the support membrane is deposited metallic osmium from the reaction of osmium tetroxide dissolved in tert-butyl alcohol (1.5 and 2.0 g OsO4 dissolved in 2500 mL tert-butyl alcohol), at a temperature of 25 °C (Figure 3), when the chemical reaction takes place according to Equation (1).
OsO4 + 4 H2 → Os + 4 H2O (1) The stages of reduction of osmium tetroxide on the polypropylene membrane are: • Formation of the osmium tetroxide solution by dissolving 1 g of solid substance in 1000 mL of tert-butyl alcohol, • Immersion of the fiber bundle in the alcoholic solution of osmium tetroxide, • Reduction of osmium tetroxide with molecular hydrogen, • Ethanol washing of osmium nanoparticles-polypropylene hollow fibers composite membranes, and • Drying in a vacuum oven (60 °C).

Preparation of Osmium Nanoparticles on Polypropylene Hollow Fiber Membranes
To make osmium nanoparticles-polypropylene hollow fibers composite membranes, on the support membrane is deposited metallic osmium from the reaction of osmium tetroxide dissolved in tert-butyl alcohol (1.5 and 2.0 g OsO 4 dissolved in 2500 mL tert-butyl alcohol), at a temperature of 25 • C (Figure 3), when the chemical reaction takes place according to Equation (1).
The stages of reduction of osmium tetroxide on the polypropylene membrane are: • Formation of the osmium tetroxide solution by dissolving 1 g of solid substance in 1000 mL of tert-butyl alcohol, • Immersion of the fiber bundle in the alcoholic solution of osmium tetroxide, • Reduction of osmium tetroxide with molecular hydrogen, • Ethanol washing of osmium nanoparticles-polypropylene hollow fibers composite membranes, and • Drying in a vacuum oven (60 • C).
The nanocomposite membrane samples containing osmium were characterized by scanning electron microscopy (SEM) and high resolution scanning electron microscopy (HR-SEM), energy dispersive spectroscopy analysis (EDAX), X-ray diffraction analysis (XRD), atomic absorption spectroscopy, and thermal analysis (TG, DSC). The nanocomposite membrane samples containing osmium were characterized by scanning electron microscopy (SEM) and high resolution scanning electron microscopy (HR-SEM), energy dispersive spectroscopy analysis (EDAX), X-ray diffraction analysis (XRD), atomic absorption spectroscopy, and thermal analysis (TG, DSC).
For X-ray diffraction analysis (XRD), atomic absorption spectroscopy, and thermal For X-ray diffraction analysis (XRD), atomic absorption spectroscopy, and thermal analysis (TG, DSC) the samples were ground at a colloidal ball mill (Retsch PM 100, VIOLA-Shimadzu, Bucharest, Romania) provided with ceramic grinding bodies.

Carrying out the Process of Oxidation or Reduction
The impregnated membrane bundle was mounted in a reaction module similar to that extensively described in our previous works [60][61][62]. A bundle of composite fibers was inserted into this housing and subsequently immobilized with acrylic polymer, thus obtaining a membrane contactor similar to a tubular heat exchanger. This module was placed in the working installation ( Figure 4a). The circulation of fluids through a membrane is shown in Figure 4b. The nanocomposite membrane samples containing osmium were characterized by scanning electron microscopy (SEM) and high resolution scanning electron microscopy (HR-SEM), energy dispersive spectroscopy analysis (EDAX), X-ray diffraction analysis (XRD), atomic absorption spectroscopy, and thermal analysis (TG, DSC).
For X-ray diffraction analysis (XRD), atomic absorption spectroscopy, and thermal analysis (TG, DSC) the samples were ground at a colloidal ball mill (Retsch PM 100, VI-OLA-Shimadzu, Bucharest, Romania) provided with ceramic grinding bodies.

Carrying out the Process of Oxidation or Reduction
The impregnated membrane bundle was mounted in a reaction module similar to that extensively described in our previous works [60][61][62]. A bundle of composite fibers was inserted into this housing and subsequently immobilized with acrylic polymer, thus obtaining a membrane contactor similar to a tubular heat exchanger. This module was placed in the working installation ( Figure 4a). The circulation of fluids through a membrane is shown in Figure 4b.   The solution of p-nitrophenol or 10-undecylenic acid in tert-butyl alcohol (ethanol or propanol) was conveyed through the membranes, and hydrogen or molecular oxygen was transported at normal pressure (1 atm) through the composite membranes.
The conversion (η %) of analytes calculated using the solution's concentration (2) or absorbance (3) [64,65] were: where c f is the final concentration of the solute (10-undecylenic acid or p-nitrophenol) and c o is the initial concentration of solute (10-undecylenic acid or p-nitrophenol).
A 0 is the initial sample solution absorbance and A s is the current sample absorbance. If the reduction of p-nitrophenol leads to p-aminophenol, the oxidation of 10-undecylenic acid leads to a variety of reaction products. That is why in this study, only the conversion was followed. However, to illustrate the complexity of the reaction mass, 10 mL of solutes was subjected to chromatography on a 60 µm silica column and the separated compounds were identified through Fourier transform infrared (FTIR) spectrometry (see Table S1).
After the processing, nitrogen was introduced into the module for two hours to remove hydrogen or oxygen. In the specific case of oxidation, after the process, nitrogen was introduced into the module for two hours to remove oxygen, then hydrogen for two hours to reduce any traces of osmium oxide, and, finally, nitrogen was introduced again for two hours. The module was opened for membrane sampling for analysis after the reaction mass was aspirated, through membranes (a vacuum nanofiltration was performed at 100 mm H 2 O to maximize the reaction mass (permeate) and the retention of nanoparticles and nanoparticle aggregates of osmium (concentrate)).
Successive nitrogen washing and hydrogen gas reduction operations ensured an osmium-free reaction mass and recovery of any osmium loss. The operations described are also mandatory to ensure the safety at work of the laboratory and the staff.

Equipment
The microscopy studies, SEM and HR-SEM, were performed on a Hitachi S4500 system (Hitachi High-Technologies Europe GmbH, Mannheim, Germany).
X-ray diffraction analyses (XRD) were recorded using PANalytical X'Pert Pro MPD equipment (PANalytical B.V., Almelo, The Netherlands) with a CuKαCradiation source, and 2θ measurement range, from 10 to 90 • C.
Thermal analysis (TG, DSC) was performed with a STA 449C Jupiter apparatus, from Netzsch (NETZSCH-Gerätebau GmbH, Selb, Germany). Each sample weighed approximately 10 mg. The samples were placed in an open alumina crucible and heated up to 900 • C with 10 K·min −1 rate, under flow of 50 mL·min −1 dried air. As reference, we used an empty alumina crucible. The evolved gases were analyzed with a FTIR Tensor 27 from Bruker (Bruker Co., Ettlingen, Germany), equipped with a thermostat gas cell.
The electrochemical processes were followed up with a PARSTAT 2273 Potentiostat (Princeton Applied Research, AMETEK Inc., Berwyn, PA, USA). It has been used a glass cell with three electrodes setup.
Cyclic voltammetry was performed for a potential sweep between −0.5 V and +1.23 V, at a scan rate of 50 mV/s. The experimental procedure took place at room temperature.
The UV-Vis studies on the nanoparticles samples composition were performed on dual-beam UV equipment-Varian Cary 50 (Agilent Technologies Inc., Santa Clara, CA, USA)-at a resolution of 1 nm, spectral bandwidth 1.5 nm, and 300 nm/s scan rate. The samples' UV-Vis spectra were recorded for a wavelength from 200 nm to 800 nm, at room temperature, using 10 mm quartz cells.
The pH of the medium was followed up with a combined selective electrode (HI 4107, Hanna Instruments Ltd, Leighton Buzzard, UK) and a multi-parameter system (HI 5522, Hanna Instruments Ltd., Leighton Buzzard, UK).
To assess and validate the content in metal ions, the atomic absorption spectrometer AAnalyst 400 AA Spectrometer (Perkin Elmer Inc., Shelton, CT, USA) with a single-element hollow-cathode lamp was used, driven by WinLab32-AA software (Perkin Elmer Inc., Shelton, CT, USA).

Results and Discussions
The present study used osmium nanoparticles-polypropylene hollow fiber membranes selected from previous experiments on osmium nanoparticles-polymer membranes [63], which included three polymers of different hydrophobicity: polyethylene, polysulfone, and cellulose acetate.
The choice of polypropylene hollow fiber membrane was determined both by the nature of the environment that the studies in this paper address (alcoholic solutions) and by the need for a large contact area.
Polypropylene hollow fiber membrane has the advantage of a chemical resistance over the entire pH range, but also in oxidizing, reducing, or temperature environments up to 90-100 • C.
This paper presents the recovery of osmium tetroxide, left from various stages of processing in electron microscopy, as follows:

•
The preparation of composite membranes based on osmium-polypropylene hollow fiber (Os-PP), by in situ reduction of the osmium tetroxide from tert-butyl alcohol solution with molecular hydrogen, • Morpho-structural characterization of the obtained osmium-polypropylene hollow fiber composite membranes, • The reduction of p-nitrophenol with molecular hydrogen in a module using osmiumpolypropylene hollow fiber composite membranes, and • Oxidation of 10-undecylenic acid with osmium-polypropylene hollow fiber composite membranes using molecular oxygen.

The Preparation and Characterization of the Composite Membrane Osmium-Polypropylene Hollow Fiber (Os-PP) by In Situ Reduction of Osmium Tetroxide
Obtaining a composite membrane intended for a membrane reactor must combine the special physical-chemical resistance of the support with the selectivity and versatility of metallic nanoparticles.
The membrane support chosen for the osmium nanoparticles-polymer composite membrane is the polypropylene hollow fiber membrane that ensures the possibility of working in aqueous or non-aqueous environments, over the entire pH range, at temperatures up to 90-100 • C and multiple washing or sterilization variants (hypochlorite, hydrogen peroxide, alkaline, or basic solutions).
In this study, the most important characteristics were the porosity (minimum 40%), pore diameter (less than 0.2 µm), and small thickness of approx. 20 µm (Figures 5 and S4). These characteristics allowed the intimate contact of hydrogen, through the membrane contactor which was also the membrane preparation installation ( Figure 4), with the solution of osmium tetroxides in tert-butyl alcohol, in order to reduce the metallic osmium that thus develops in the pores (Figures 6 and 7). Molecular hydrogen was conveyed through the polypropylene hollow fibers membranes at a flow rate of 1 L/min, and the solution of osmium tetroxide in tert-butyl was maintained outside the fibers.
The concentration of the solution of osmium tetroxide in tert-butyl alcohol was chosen at 0.6 g/L for the Os-PP1 membrane and 0.8 g/L for the Os-PP2 membrane. The concentrations of the solutions were chosen in order to be able to compare these composite membranes with the previously obtained and reported membrane, Os-PP0 [64]. The concentration of the solution of osmium tetroxide in tert-butyl alcohol was 0.4 g/L. The increase in concentration of osmium tetroxide had in view obtaining a high concentration of osmium atoms at the surface of the raw membranes ( Figure 7) highlighted by energy dispersive spectroscopy analysis (EDAX) ( Table 1). through the polypropylene hollow fibers membranes at a flow rate of 1 L/min, and the solution of osmium tetroxide in tert-butyl was maintained outside the fibers.  Although the local analyses (on the surface) showed the presence of osmium nanoparticles and nanoparticles aggregates, the X-ray diffraction analysis (XRD) being a global method (a quantity of Os-PP membranes ground at the mill was analyzed) was more difficult to interpret, but the lines specific to metallic osmium were still highlighted (see Figure S5 in the Supplementary Material).
To overcome this aspect, extensive energy dispersive spectroscopy analysis (EDAX) analyses were performed. The EDAX diagrams, having a formal aspect presented in Figures 8a and S6, showed the presence of carbon and osmium on the surface, but also of oxygen, which comes both from the conditioning agent of polypropylene hollow fibers membranes and from the working solvent, or the hydrating water that appears in the reduction of osmium tetroxide (see also Equation (1)). Nanomaterials 2021, 11, x FOR PEER REVIEW 9 of 26    The dimensions of osmium nanoparticles (Figure 8b-d) were tracked by detailing the nanoparticle aggregates on the raw osmium nanoparticles-polypropylene hollow fibers composite membranes.
By scanning electron microscopy analysis (SEM) coupled with energy dispersive spectroscopy analysis (EDAX), it was possible to follow the distributions of osmium nanoparticles on the membranes recovered after the microfiltration of the respective reaction system: tert-butanol ( Figure 9a); mass of reagents after the reduction process ( Figure 9b); and mass of reagents after the oxidation process (Figure 9c). nanoparticle aggregates on the raw osmium nanoparticles-polypropylene hollow fibers composite membranes.
By scanning electron microscopy analysis (SEM) coupled with energy dispersive spectroscopy analysis (EDAX), it was possible to follow the distributions of osmium nanoparticles on the membranes recovered after the microfiltration of the respective reaction system: tert-butanol ( Figure 9a); mass of reagents after the reduction process ( Figure 9b); and mass of reagents after the oxidation process (Figure 9c).
The distribution of osmium on the surface of the target membrane was performed both before and after the redox processes (Figure 9a-c; see also Figure S6). The relatively uniform distribution on the membrane surface and the size of the nanoparticles were both remarkable. This uniform distribution of osmium nanoparticles on the processed membranes was due to the microfiltration process which, in addition to the disintegration effect, directed the nanoparticles on the surface and towards the interior of the pores, so that the map of nanoparticles reflected the pore distribution of the support membrane. The surface of 1 m 2 represents a disadvantage for the study of nanoparticles, regardless of the analysis technique, due to dispersion (approx. 1.0-1.6 g/m 2 ).
(a) (b) (c) Figure 9. The energy dispersive spectroscopy analysis (EDAX) for the osmium nanoparticles-polypropylene hollow fibers composite membrane before (a); and after reduction (b); or oxidation (c) processes. Figure 9. The energy dispersive spectroscopy analysis (EDAX) for the osmium nanoparticles-polypropylene hollow fibers composite membrane before (a); and after reduction (b); or oxidation (c) processes.
The distribution of osmium on the surface of the target membrane was performed both before and after the redox processes (Figure 9a-c; see also Figure S6). The relatively uniform distribution on the membrane surface and the size of the nanoparticles were both remarkable. This uniform distribution of osmium nanoparticles on the processed membranes was due to the microfiltration process which, in addition to the disintegration effect, directed the nanoparticles on the surface and towards the interior of the pores, so that the map of nanoparticles reflected the pore distribution of the support membrane. The surface of 1 m 2 represents a disadvantage for the study of nanoparticles, regardless of the analysis technique, due to dispersion (approx. 1.0-1.6 g/m 2 ).
The thermal analysis of the composite membranes gives information on the component's interactions and membranes behavior at different temperatures. This is necessary in order to follow the performances in case of use at higher temperatures than the ambient one, and to fully characterize the composite material ( Figure 10 and Table 2). The thermal analysis of the composite membranes gives information on the component's interactions and membranes behavior at different temperatures. This is necessary in order to follow the performances in case of use at higher temperatures than the ambient one, and to fully characterize the composite material ( Figure 10 and Table 2).     Both samples presented a similar degradation pathway but with some differences. The sample Os-PP2 had a higher thermal stability, the temperatures T 10 -T 50 (at which 10-50% mass loss is recorded) being with~20 • C higher.
The samples were stable up to 215 • C, with negligible or no mass loss recorded. On the DSC curve, the small endothermic peaks with onset at 159.7 • C and 160.4 • C, respectively, indicate the melting temperature. The higher value for the Os-PP2 sample indicates that the osmium compound was shielding the PP in a quantifiable measure.
After 215 • C, the complex degradative-oxidative process for the polymeric part took place. The samples were partially oxidized but also some decomposition took place, the evolved gases being composed of CO 2 and H 2 O, but also saturated hydrocarbons and traces of CO and unsaturated hydrocarbons could be identified. Up to 300 • C, the decomposition and oxidation reactions were fairly equal, but after 300 • C, the oxidation reactions became predominant, with the burning taking place around 400 • C. The strong exothermic peak on DSC curve had maxima at 388.6 • C and 399.9 • C for Os-PP1 and Os-PP2, respectively.
In order to obtain more information on the thermal processes, the gases evolved from thermal analysis were investigated by FTIR during heating in the 20-900 • C range. Figure 11 present the 3D FTIR data for the evolved gases.  Both samples presented a similar degradation pathway but with some differences. The sample Os-PP2 had a higher thermal stability, the temperatures T10-T50 (at which 10-50% mass loss is recorded) being with ~20 °C higher.
The samples were stable up to 215 °C, with negligible or no mass loss recorded. On the DSC curve, the small endothermic peaks with onset at 159.7 °C and 160.4 °C, respectively, indicate the melting temperature. The higher value for the Os-PP2 sample indicates that the osmium compound was shielding the PP in a quantifiable measure.
After 215 °C, the complex degradative-oxidative process for the polymeric part took place. The samples were partially oxidized but also some decomposition took place, the evolved gases being composed of CO2 and H2O, but also saturated hydrocarbons and traces of CO and unsaturated hydrocarbons could be identified. Up to 300 °C, the decomposition and oxidation reactions were fairly equal, but after 300 °C, the oxidation reactions became predominant, with the burning taking place around 400 °C. The strong exothermic peak on DSC curve had maxima at 388.6 °C and 399.9 °C for Os-PP1 and Os-PP2, respectively.
In order to obtain more information on the thermal processes, the gases evolved from thermal analysis were investigated by FTIR during heating in the 20-900 °C range. Figure  11 present the 3D FTIR data for the evolved gases. The 3D FTIR chromatograms for Os-PP1 ( Figure 11a) and Os-PP2 (Figure 11b) allow the extraction of individual FTIR spectrum of the exhaust gases at temperatures of interest. In this case, the beginning of the thermal decomposition was taken into account, at 270 °C, in order to have related compositional information. The 3D FTIR chromatograms for Os-PP1 (Figure 11a) and Os-PP2 (Figure 11b) allow the extraction of individual FTIR spectrum of the exhaust gases at temperatures of interest. In this case, the beginning of the thermal decomposition was taken into account, at 270 • C, in order to have related compositional information.
In the FTIR spectra for Os-PP1 at 270 • C, and Os-PP2 at 274 • C (Figure 12), the presence of CO 2 is mostly seen at 2355 cm −1 , traces of CO at 2169 cm −1 , water, but also the corresponding vibration of Csp 3 -H at 2964 cm −1 . A small peak is observed at 3072 cm −1 (corresponding to Csp 2 -H fragments). In the FTIR spectra for Os-PP1 at 270 °C, and Os-PP2 at 274 °C (Figure 12), the presence of CO2 is mostly seen at 2355 cm −1 , traces of CO at 2169 cm −1 , water, but also the corresponding vibration of Csp 3 -H at 2964 cm −1 . A small peak is observed at 3072 cm −1 (corresponding to Csp 2 -H fragments). This indicates that the degradation of the composite membranes started as a complex process, in which both decomposition and oxidation of the polymeric matrix took place simultaneously.
The bidimensional temperature/wavenumber projection of the 3D FTIR chromatogram for Os-PP1 (Figure 13a) can be used to visualize, at a glance, the temperature intervals where decomposition or oxidations took place. Figure 13a marks the areas where the vibration frequencies characteristic for water, CO2, CO, and hydrocarbons appeared, to highlight the elimination in time. The green, vertical line at 2964 cm −1 represents the release profile with temperature of the hydrocarbons. It was observed that the elimination profile of hydrocarbons by decomposition was different between Os-PP1 and Os-PP2. It can be stated that the differences between T10-T50 (Table 2) can be the subject of the study. Thus, in the case of Os-PP1, a more accentuated decomposition of polypropylene (PP) took place, while in the case of Os-PP2, this decomposition was lower in the 200-300 °C temperature interval. This indicates that the degradation of the composite membranes started as a complex process, in which both decomposition and oxidation of the polymeric matrix took place simultaneously.
The bidimensional temperature/wavenumber projection of the 3D FTIR chromatogram for Os-PP1 (Figure 13a) can be used to visualize, at a glance, the temperature intervals where decomposition or oxidations took place. Figure 13a marks the areas where the vibration frequencies characteristic for water, CO 2 , CO, and hydrocarbons appeared, to highlight the elimination in time. The green, vertical line at 2964 cm −1 represents the release profile with temperature of the hydrocarbons. It was observed that the elimination profile of hydrocarbons by decomposition was different between Os-PP1 and Os-PP2. It can be stated that the differences between T 10 -T 50 ( Table 2) can be the subject of the study. Thus, in the case of Os-PP1, a more accentuated decomposition of polypropylene (PP) took place, while in the case of Os-PP2, this decomposition was lower in the 200-300 • C temperature interval.

Reduction of P-Nitrophenol with Molecular Hydrogen with Osmium Nanoparticles-Polypropylene Hollow Fibers Composite Membranes
The reduction of nitroderivatives is a reaction of ennobling of some intermediate compounds for multiple products of biological interest, and the nitrophenols are the most intensively studied target substances both for the accessibility of analytical monitoring and for their biomedical importance and impact on environment [24,47,48]. These two arguments are the basis for determination of the process performance of composite membranes based on osmium nanoparticles-polypropylene hollow fibers Os-PP1 and Os-PP2 in the reduction reaction of p-nitrophenol from lower aliphatic alcohol solutions with molecular hydrogen.

Reduction of P-Nitrophenol with Molecular Hydrogen with Osmium Nanoparticles-Polypropylene Hollow Fibers Composite Membranes
The reduction of nitroderivatives is a reaction of ennobling of some intermediate compounds for multiple products of biological interest, and the nitrophenols are the most intensively studied target substances both for the accessibility of analytical monitoring and for their biomedical importance and impact on environment [24,47,48]. These two arguments are the basis for determination of the process performance of composite membranes based on osmium nanoparticles-polypropylene hollow fibers Os-PP1 and Os-PP2 in the reduction reaction of p-nitrophenol from lower aliphatic alcohol solutions with molecular hydrogen. For the comparison of the results, a composite membrane of osmium nanoparticlespolypropylene hollow fibers Os-PP0 was tested, recently reported for reduction of 5nitrobenzimidazole [64].
The reduction process was carried out according to the chemical reactions (4) and (5), but the equilibria (6) and (7) must also be considered, in which either the hydration water of osmium nanoparticles or the one resulting from reaction can participate.
The reduction of nitro derivatives with osmium compounds is very well known [32,33,66], but the study of this process on osmium nanoparticles with hydrogen gas is much less approached, the in-situ generation of molecular hydrogen especially with sodium borohydride being preferred [67,68].
At the same time, the choice of the alcoholic reaction medium is sustained due to the low solubility of hydrogen gas in the aqueous medium [69].
The obtained results highlight the influence of the membrane type (Os-PP0, Os-PP1 and Os-PP2) (Figure 14), concentration of p-nitrophenol solution (Figure 15), and nature of the primary alcohol ( Figure 16) over the conversion (η %), when reducing pnitrophenol from alcoholic solution for the composite membranes of osmium nanoparticlespolypropylene hollow fibers (Os-PP).
η Os-PP0 < η Os-PP1 < η Os-PP2 (8) In the membrane contactor (Figure 4), at a temperature of 25 • C, 10 L of 1g/L solution p-nitrophenol in t-butanol solution was introduced, which was recirculated at a flow rate of 0.100 L/min through the outside of the composite membranes of nanoparticles of osmiumpolypropylene hollow fibers, while molecular hydrogen was continuously introduced into the system at a flow rate of 1 L/min through the membranes. The experiments performed for each type of composite membrane (Os-PP0, Os-PP1, and Os-PP2) for six hours showed that the evolution of hydrogenation was slow at the beginning of the operation interval, and the conversion value (η %) reached 75% (Os-PP0), 83% (Os-PP1), and 87% (Os-PP2) at six hours ( Figure 14). For the comparison of the results, a composite membrane of osmium nanoparticlespolypropylene hollow fibers Os-PP0 was tested, recently reported for reduction of 5-nitrobenzimidazole [64].
The reduction process was carried out according to the chemical reactions (4) and (5), but the equilibria (6) and (7) must also be considered, in which either the hydration water of osmium nanoparticles or the one resulting from reaction can participate.
At the same time, the choice of the alcoholic reaction medium is sustained due to the low solubility of hydrogen gas in the aqueous medium [69].
The obtained results highlight the influence of the membrane type (Os-PP0, Os-PP1 and Os-PP2) (Figure 14), concentration of p-nitrophenol solution (Figure 15), and nature of the primary alcohol ( Figure 16) over the conversion (η %), when reducing p-nitrophenol from alcoholic solution for the composite membranes of osmium nanoparticles-polypropylene hollow fibers (Os-PP).
η Os-PP0 < η Os-PP1 < η Os-PP2 (8)   In the membrane contactor (Figure 4), at a temperature of 25 °C, 10 L of 1g/L solution p-nitrophenol in t-butanol solution was introduced, which was recirculated at a flow rate of 0.100 L/min through the outside of the composite membranes of nanoparticles of osmium-polypropylene hollow fibers, while molecular hydrogen was continuously introduced into the system at a flow rate of 1 L/min through the membranes. The experiments performed for each type of composite membrane (Os-PP0, Os-PP1, and Os-PP2) for six  In the membrane contactor (Figure 4), at a temperature of 25 °C, 10 L of 1g/L solution p-nitrophenol in t-butanol solution was introduced, which was recirculated at a flow rate of 0.100 L/min through the outside of the composite membranes of nanoparticles of osmium-polypropylene hollow fibers, while molecular hydrogen was continuously introduced into the system at a flow rate of 1 L/min through the membranes. The experiments performed for each type of composite membrane (Os-PP0, Os-PP1, and Os-PP2) for six The conversion values are directly related to the osmium concentration at the membrane surface (Table 1) increasing in the following order: Using the most performant composite membrane obtained (Os-PP2), the hydrogenation of p-nitrophenol was then followed, in concentrations of 1.0-3.0 g/L, under the same experimental conditions (Figure 15). The conversion of p-nitrophenol (η %) decreased with increasing concentration of p-nitrophenol in t-butanol solution. The results of the conversion (η %) at six hours of operation varies from 85% for the 1.0 g/L solution, 56% for the 2.0 g/L solution, and 28% for the 3.0 g/L solution. For a high conversion, it will be necessary to either operate with dilute solutions or increase the working time.
The type of alcohol, used as solvent, significantly influenced the conversion of pnitrophenol for the tested membrane (Os-PP2) ( Figure 16). The results obtained for a p-nitrophenol solution of 1.0 g/L, under the previous experimental conditions, at six hours of operation, showed a conversion (η %) of 85% for t-butanol, 57% for i-propanol, 24% for ethanol, and 19% for methanol.
The study performed on the reduction reaction of p-nitrophenol from solutions of lower aliphatic alcohols, with molecular hydrogen, on composite membranes of osmium nanoparticles-polypropylene hollow fibers (Os-PP) allowed the establishment of the following parameters:

Oxidation of 10-Undecylenic Acid on Osmium Nanoparticles-Polypropylene Hollow Fiber Composite Membranes Using Molecular Oxygen
The 10-undecylenic acid is an accessible substance with multiple industrial applications, in sports biomedicine but also as membrane material (spacer). The oxidation of this compound is of particular interest for technical applications, but especially biomedical, because its properties can be altered when the athletes who use it bring it into contact with various powders (talc, zinc oxide, titanium oxide) used for adhesion and solar UV rays [70][71][72][73].
In the present case, the oxidation of 10-undecylenic acid was achieved from alcoholic solution on osmium nanoparticles-polypropylene hollow fibers composite membranes, using molecular oxygen ( Figure 17). The reaction of oxidation performed in the installation depicted in Figure 4 used an alcoholic solution (ethanol, i-propanol, and t-butanol) of 10-undecylenic acid of concentration of 1.0, 3.0, and 5.0 g/L at a volume of 10 L and a recirculation flow of 0.100 L/min, through the contactor membrane system, through the outside of the osmium-polypropylene hollow fiber composite membranes (Os-PP2). Molecular oxygen was introduced through membranes, at a flow rate of 1 L/minute, a temperature of 25 • C, and a pressure of 1 atm. The operating time was set at a maximum of five hours.  Figure 18 shows the results of 10-undecylenic acid oxidation, depending on time and nature of the primary alcohol used. The conversion (η %) of 10-undecylenic grew from 48% to 97% during five hours for t-butanol, and from 60% to 82% for ethanol. The conversion for i-propanol fell between the other two solvents. The conversion began more abruptly for the solvent with lower molecular weight, but towards the end of the operation, the order of conversion in the presence of the three alcohols was reversed. Most likely, this aspect of the conversion was determined by the two stages of the process: doublebond oxidation followed by alcoholysis.  Figure 18 shows the results of 10-undecylenic acid oxidation, depending on time and nature of the primary alcohol used. The conversion (η %) of 10-undecylenic grew from 48% to 97% during five hours for t-butanol, and from 60% to 82% for ethanol. The conversion for i-propanol fell between the other two solvents. The conversion began more abruptly for the solvent with lower molecular weight, but towards the end of the operation, the order of conversion in the presence of the three alcohols was reversed. Most likely, this aspect of the conversion was determined by the two stages of the process: double-bond oxidation followed by alcoholysis. nature of the primary alcohol used. The conversion (η %) of 10-undecylenic grew from 48% to 97% during five hours for t-butanol, and from 60% to 82% for ethanol. The conversion for i-propanol fell between the other two solvents. The conversion began more abruptly for the solvent with lower molecular weight, but towards the end of the operation, the order of conversion in the presence of the three alcohols was reversed. Most likely, this aspect of the conversion was determined by the two stages of the process: doublebond oxidation followed by alcoholysis. To study the influence of 10-undecylenic acid concentration in the alcohol solution as a function of time, the solvent that showed the maximum conversion at five hours of operation, t-butanol, was chosen ( Figure 19). To study the influence of 10-undecylenic acid concentration in the alcohol solution as a function of time, the solvent that showed the maximum conversion at five hours of operation, t-butanol, was chosen ( Figure 19).  The conversion (η %) as a function of time for 10-undecylenic acid of variable concentration (1.0, 3.0 and 5.0 g/L) at oxidation with molecular oxygen on the osmium nanoparticles-polypropylene hollow fibers (Os-PP2) composite membrane showed that at high concentrations, the conversion was lower at the beginning of oxidation, but during the process, the conversions approached each other, reaching values of over 80% for the entire chosen concentration range after five hours.
The oxidation of 10-undecylenic acid follows a probable three-stage mechanism: 1. Through the intervention of osmium nanoparticles, oxygen is physically-chemically linked to the double-bond; 2. The alcohol (mostly in the environment) produces the alcoholysis of the osmic acid ester producing a variety of reaction products (the reaction was modified in manuscript according to these data); and 3. The products are released into the environment. The conversion (η %) as a function of time for 10-undecylenic acid of variable concentration (1.0, 3.0 and 5.0 g/L) at oxidation with molecular oxygen on the osmium nanoparticles-polypropylene hollow fibers (Os-PP2) composite membrane showed that at high concentrations, the conversion was lower at the beginning of oxidation, but during the process, the conversions approached each other, reaching values of over 80% for the entire chosen concentration range after five hours.
The oxidation of 10-undecylenic acid follows a probable three-stage mechanism: 1.
Through the intervention of osmium nanoparticles, oxygen is physically-chemically linked to the double-bond; The alcohol (mostly in the environment) produces the alcoholysis of the osmic acid ester producing a variety of reaction products (the reaction was modified in manuscript according to these data); and 3.
The products are released into the environment.
The complex qualitative composition of the reaction mass, determined by chromatographic separation and FTIR spectroscopy identification of the components, showed that the glycolic compound obtained was majority (R 1 , R 2 , and R 3 being the hydrogen atoms), but also alkylated compounds with ethyl, propyl, and butyl (in various positions) were detected (see Table S1).
A very special aspect of our research is the problem of loss of osmium and/or osmium nanoparticles in the reaction medium. Beyond the high economic value of osmium, the toxicity of its compounds (especially osmium tetroxide) requires both operational safety measures and rigorous control of operations. For the research carried out in this study, osmium nanoparticles were permanently operated in reaction space outside the fibers. To avoid the loss of osmium nanoparticles from the reaction compartment, the alcoholic reaction mass was extracted from the membrane module by vacuum suction inside the polypropylene hollow fiber membranes. At the same time, the reaction compartment was subjected to a nitrogen gas flow after each reduction operation (hydrogenation) before the recovery of the reaction mass (See Figure S7). After the oxidation process, the reaction compartment underwent three stages of tartar: with nitrogen gas to remove residual oxygen, with hydrogen (at least two hours to reduce any traces of osmium tetroxides), and then with nitrogen to remove residual hydrogen. The filtration characteristics of polypropylene hollow fiber membranes [56][57][58][59][60][61][62][63][64] ensured the complete retention of osmium nanoparticles in the reaction compartment which could be followed by established analytical techniques [74][75][76][77][78].

Conclusions
This paper presented the preparation and characterization of a composite membrane of active metal nanoparticle-polymer support type, based on osmium nanoparticles obtained in situ on a polypropylene hollow fiber membrane. The osmium nanoparticles were generated from a solution of osmium tetroxide of different concentrations in tertbutyl alcohol, by reduction with molecular hydrogen, in a contactor with polypropylene membrane.
The osmium nanoparticles-polypropylene hollow fibers (Os-PP) composite membranes were characterized morphologically and structurally by scanning electron microscopy (SEM), thermal analysis, surface analysis, and Fourier transform infrared (FTIR) spectroscopy.
The performance of osmium nanoparticles-polypropylene hollow fibers (Os-PP2) composite membranes was tested during the processes of reduction or oxidation of two compounds of technological and biomedical interests (p-nitrophenol and 10-undecylenic acid) from lower saturated alcohol solutions.
The results obtained showed that the reduction of p-nitrophenol with molecular hydrogen on composite membranes based on osmium nanoparticles-polypropylene hollow fibers (Os-PP) can be performed with conversions of about 90%, under the following conditions:
Oxidation of 10-undecylenic acid with molecular oxygen on composite membranes of osmium nanoparticles-polypropylene hollow fibers (Os-PP) took place with conversions of over 80% of solutions in t-butanol of concentration 1-5 g/L, at an operating time of five hours. The composition of the reaction mass to the oxidation of 10-undecylenic acid is complex and requires further studies.
To avoid the loss of osmium nanoparticles from the reaction compartment outside the fibers of the membrane module, the alcoholic reaction mass was extracted from the membrane module by vacuum suction inside the polypropylene hollow fiber membranes.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/nano11102526/s1, Figure S1: The appearances of crystallized metallic osmium and osmium powder, Figure S2: Images with the solution of osmium tetroxide in tert-butanol, the obtained metallic osmium and the osmium powder retained on a microporous membrane, Figure S3: Images with the bottle in which the osmium tetroxide was kept tightly sealed and the osmium powder obtained, Figure S4: Scanning electron microscopy (SEM) images of polypropylene hollow fiber membrane support, Figure S5: X-ray diffraction diagrams for polypropylene hollow fiber membrane dust, Figure S6: Scanning electron microscopy (SEM) and energy dispersive spectroscopy analysis (EDAX) on polypropylene hollow fiber membrane support, Figure S7: The specific safety operations on reactional processes and the obtained membranes, Table S1: The composition of the reaction mass to the oxidation with molecular oxygen of 10-undecylenic acid on osmium-polypropylene hollow fiber composite membranes.