Role of precursor composition in the polymorph transformations, morphology control and ferromagnetic properties of nanosized TiO_2
Dmitry Zablotsky, Mikhail M. Maiorov, Aija Krumina, Marina Romanova, Elmars Blums
JJournal Name
Role of precursor composition in the polymorph trans-formations, morphology control and ferromagneticproperties of nanosized TiO Dmitry Zablotsky, ∗ a Mikhail M. Maiorov, a Aija Krumina, a , b Marina Romanova, c and El-mars Blums a Pure phase and mixed phase TiO nanoparticles have been produced using a pyrolytic methodfrom a non-aqueous carboxylate precursor. The precursor was prepared by a multiphase cationexchange using pentanoic acid (C H COOH). The thermal stability, polymorph content, morphol-ogy, size distribution and surface region of the produced nanoparticles were studied by TGA/DSC,XRD, FTIR and TEM. High quality monodisperse nanoparticles have been produced in the sizerange from 7 to 27 nm. The nanoparticles showed room temperature ferromagnetism revealedby VSM within bound polaron model. The carboxylate precursor is a good alternative to standardsol-gel to produce nanoparticles free from impurities.
Titanium dioxide (TiO ) is a multifunctional semiconductingmetal oxide (SMO) that is expected to play a significant role inenvironmental and energy applications . Amongst all SMOs(such as WO , ZnO, Fe O , MnO etc.) TiO is the most in-tensively studied because its chemically inert, reusable, abun-dant and inexpensive, non-toxic and biologically compatible andhas a wide band gap ( ∼ . In thepast decades a tremendous effort has been put into improvingthe efficiency of TiO materials. Titania crystallizes in threedistinct polymorphs with different properties and structure: ru-tile (tetragonal, space group P42/mmm), anatase (tetragonal,space group I41/amd) and brookite (orthorhombic, space groupPcab). Bi-phasic anatase/rutile mixtures are most wanted thanpure phases because of superior properties . The establishedstandard - Aeroxide TiO P25 marketed by Evonik Industries (De-gussa) – a mixed phase photocatalyst consisting of about 70-80wt.% anatase and 20-30 wt.% rutile, has been extensively stud-ied, benefited from synergistic interaction of the phases . Fun-damentally, a considerable problem is a very limited range of pre-cursors and process chemistries to produce nanostructured TiO . a University of Latvia, Jelgavas 3, 1004 Riga, Latvia; E-mail: [email protected] b Institute of Inorganic Chemistry, Riga Technical University, Faculty of Material Scienceand Applied Chemistry, P. Valdena 7, 1048 Riga, Latvia. c Institute of Biomedical Engineering and Nanotechnologies, Riga Technical University,Viskalu 36A, 1006 Riga, Latvia.
The solution-based sol–gel chemistry is commonly used to pro-duce TiO nanocrystallites. It starts from either an organometal-lic titanium alkoxide (Ti ethoxide, butoxide or isopropoxide) orinorganic titanium chloride (TiCl or TiCl ) and proceeds via atwo-step process: hydrolysis of the salt by added water resultingin the formation of intermediate species (monomers) and theirsubsequent self-assembly and polymerization into extended 3Dnetwork. The colloidal suspension (sol) is precipitated, dried andcalcined to complete the phase transformation.Previous studies have shown that during sol-gel the phase con-tent, crystalline structure, size and morphology of particles werefound to be highly sensitive to various parameters, such as theTi:H O ratio, anion content (e.g. Cl − ), the pH, and the choice ofreaction modulator . The hydrolysis of the Ti precursor is acritical step, which determines the initial monomer species andstrongly impacts the resultant phase . For instance, a changeof molar ratios in the chemical environment (such as decrease ininstantaneous concentration of Ti precursor) results in differenthydrolysis speeds and time-dependent shifts of polymorph selec-tivity . The complexity of the method and the lack of detailedunderstanding of the chemical equilibrium of the species in solu-tion and kinetics of nucleation and growth of the different phasesmakes it difficult to achieve and reproduce a substantial controlover the relative phase content, usually obtaining a final mixtureof anatase, rutile and brookite. Secondly, there are safety con-cerns over direct use of Ti alkoxides or chlorides in sol-gel as theyare in general very reactive and sensitive to moisture. The hydrol-ysis is vigorous (even at 0 ◦ C), hence, extensive precautions arenecessary in handling these chemicals. Likewise, the sol-gel gen-erates highly corrosive acidic conditions, since the alkoxide route
Journal Name, [year], [vol.] , a r X i v : . [ c ond - m a t . m t r l - s c i ] F e b equires acid catalysts (e.g. HCl, HNO , H SO ) to controlthe steps in the reaction and phase selectivity, whereas chlorideprecursors (TiCl or TiCl ) produce large amounts of corrosiveHCl in situ during hydrolysis, which results in chloride-rich solu-tions with very low pH. Moreover, extensive dialysis/purificationprocedures are required to remove the byproducts of sol-gel syn-thesis adsorbed on the particle surface and strongly bound (e.g.Cl − , NO − , NH + ) species obtained from precursors cannot be to-tally removed . The doping impurities, secondary impurityphases and surface-adsorbed features definitely influence the sur-face properties, which can hinder the application potential of thematerial. For instancce, Aeroxide (Degussa) P25 - the most suc-cessful and used photocatalyst - is instead produced by pyrogenicflame-hydrolysis, in which vapourised TiCl is combusted in anoxy-hydrogen flame . The solid is then separated and treatedwith steam at 450-550 ◦ C to remove chlorine-containing groups.The main advantage of pyrohydrolysis is that it is scalable to in-dustrial levels. Nevertheless, a toxic chlorine-rich flame presentssignificant corrosive hazards and is unfriendly to the environ-ment. Beyond the obvious concerns with environmental pollu-tion, Ti chlorides are not easy to store and handle. Thus, it is apractically and scientifically important challenge to develop alter-native precursors to control the size, polymorph content and mor-phology of TiO nanocrystals and, therefore, optimize the prop-erties of material. A metal-organic precursor is more favourablethan inorganic titanium chlorides, because it can be used in achlorine-free system without related hazards, however, the widelyused Ti alkoxides suffer from instability and poor cost-efficiency.In this study we report the development of substitute precur-sors to produce nanostructured TiO via pyrolysis or thermaldecomposition using metal-organic carboxylate-based extractionsystems. Having the advantage of being inexpensive and environ-mentally friendly, metal carboxylate complexes have been usedextensively in the production of nanoparticles of SMOs , butthe production of TiO nanoparticles from a carboxylate precur-sor has not been reported. Herein, Ti chloride is used as theinitial titanium source preserving the key advantage of low cost,however, the chloride anions Cl − are avoided with the extractionsystem via aqueous/organic phase segregation and metal cationexchange resulting in in situ dialysis. The resulting precursorsare storable and the nanostructured TiO is produced by non-hydrolytic pyrolysis. Thus, we are able to overcome many of thespecific problems described above to achieve better particle sizeand polymorph control than hydrolytic routes by simply varyingthe temperature of pyrolysis. Initially, an aqueous solution of TiCl was produced by dissolvingfine-grained c.p. titanium powder (1.2 g, particle size 63-100 µ m,-140+230 mesh) in boiling hydrochloric acid solution (60 ml, 1:1vol.). The metal concentration was adjusted to 0.1M by dH O ad-dition (pH 0.5). Titanium-rich carboxylate-based organometal-lic precursor was produced by liquid-liquid extraction: 60 ml of
Fig. 1
Schematic of preparation of carboxylate-based Ti organometal-lic precursor: Left - phase contact of TiCl aqueous solution and pen-tanoic acid, middle - extraction of Ti + from the aqueous solution into theorganic phase by cation exchange, right - produced non-aqueous Ti + -rich carboxylate-based organometallic precursor with characteristic violetcolor. TiCl aqueous solution and 20 ml (3:1 aqueous/organic volumeratio) of pentanoic acid (without diluent) were placed in a sep-aration funnel; 1M NaOH soluion was added to the mixture insmall portions, followed by vigorous shaking after each additionuntil the pH of the aqueous phase reached pH 1, after whichthe phases were allowed to separate for 0.5 h. After completestratification and removal of the aqueous phase followed by filtra-tion, a Ti-rich precursor E1 (0.15M Ti) was produced. To studythe effect of precursor concentration on produced TiO a moreconcentrated precursor E2 (0.5M Ti) was prepared as describedabove, but with 5:1 aqueous/organic volume ratio. The precur-sors E1 and E2 were pyrolyzed for 1 h at 350 ◦ C (E1/E2-350),400 ◦ C (E1/E2-400), 450 ◦ C (E1/E2-450), 550 ◦ C (E1/E2-550),650 ◦ C (E1/E2-650), 750 ◦ C (E1/E2-750) in static air to samplethe phase content of derived TiO nanoparticles . by sol-gel A reference sample was prepared by standard sol-gel method, us-ing 250 ml of 0.1M TiCl aqueous solution produced in the previ-ous step. The gel-like precipitate was obtained by dropwise addi-tion of 0.5M NaOH aqueous solution at a rate of 3 ml/min undervigorous stirring until pH 6 of the solution was reached. The pro-duced gel was aged for 24 h at ambient conditions, followed byfiltration and washing with dH O. The absence of chlorine in thedecantate was checked by AgNO solution. The as-dried precipi-tate (P) was calcined at 450 ◦ C (P-450), 550 ◦ C (P-550), 650 ◦ C(P-650), 750 ◦ C (P-750) in static air to produce TiO nanoparti-cles. The thermal stability of the produced precursors was studiedby thermal gravimetric analysis (TGA) and differential scanningcalorimetry (DSC) using the STA PT1600 (LINSEIS) device. Thesample was heated in static air from ambient temperature to700 ◦ C at a rate of 10 ◦ C/min.IR spectra were recorded at r.t. using Bruker Tensor II FT-IRspectrophotometer. For each spectrum 36 scans were performedin the range 4000-400 cm − with 7 mm KBr discs (TiO /KBr massratio 1:100) prepared under a load of 2000 kg.Powder X-ray diffraction (XRD) patterns of dried samples were Journal Name, [year], [vol.][vol.]
Schematic of preparation of carboxylate-based Ti organometal-lic precursor: Left - phase contact of TiCl aqueous solution and pen-tanoic acid, middle - extraction of Ti + from the aqueous solution into theorganic phase by cation exchange, right - produced non-aqueous Ti + -rich carboxylate-based organometallic precursor with characteristic violetcolor. TiCl aqueous solution and 20 ml (3:1 aqueous/organic volumeratio) of pentanoic acid (without diluent) were placed in a sep-aration funnel; 1M NaOH soluion was added to the mixture insmall portions, followed by vigorous shaking after each additionuntil the pH of the aqueous phase reached pH 1, after whichthe phases were allowed to separate for 0.5 h. After completestratification and removal of the aqueous phase followed by filtra-tion, a Ti-rich precursor E1 (0.15M Ti) was produced. To studythe effect of precursor concentration on produced TiO a moreconcentrated precursor E2 (0.5M Ti) was prepared as describedabove, but with 5:1 aqueous/organic volume ratio. The precur-sors E1 and E2 were pyrolyzed for 1 h at 350 ◦ C (E1/E2-350),400 ◦ C (E1/E2-400), 450 ◦ C (E1/E2-450), 550 ◦ C (E1/E2-550),650 ◦ C (E1/E2-650), 750 ◦ C (E1/E2-750) in static air to samplethe phase content of derived TiO nanoparticles . by sol-gel A reference sample was prepared by standard sol-gel method, us-ing 250 ml of 0.1M TiCl aqueous solution produced in the previ-ous step. The gel-like precipitate was obtained by dropwise addi-tion of 0.5M NaOH aqueous solution at a rate of 3 ml/min undervigorous stirring until pH 6 of the solution was reached. The pro-duced gel was aged for 24 h at ambient conditions, followed byfiltration and washing with dH O. The absence of chlorine in thedecantate was checked by AgNO solution. The as-dried precipi-tate (P) was calcined at 450 ◦ C (P-450), 550 ◦ C (P-550), 650 ◦ C(P-650), 750 ◦ C (P-750) in static air to produce TiO nanoparti-cles. The thermal stability of the produced precursors was studiedby thermal gravimetric analysis (TGA) and differential scanningcalorimetry (DSC) using the STA PT1600 (LINSEIS) device. Thesample was heated in static air from ambient temperature to700 ◦ C at a rate of 10 ◦ C/min.IR spectra were recorded at r.t. using Bruker Tensor II FT-IRspectrophotometer. For each spectrum 36 scans were performedin the range 4000-400 cm − with 7 mm KBr discs (TiO /KBr massratio 1:100) prepared under a load of 2000 kg.Powder X-ray diffraction (XRD) patterns of dried samples were Journal Name, [year], [vol.][vol.] , ecorded in θ range between 20 ◦ and 75 ◦ using D8 Advance(Bruker Corporation) diffractometer with CuK α radiation ( λ =1.5418Å), accelerating voltage 40 kV and current 40 mA; stepsize 0.02 ◦ at a scanning rate 2 s/step. Zero-background Si sam-ple holders were used. The quantitative phase composition ofthe produced materials were determined by Rietvield refinement(FullProf Suite ver. 7.30), a whole pattern fitting method that sys-tematically varies constraints in a simulated pattern to reproducethe experimental pattern. The refinement was performed start-ing from the identified crystal phases and known crystal structuredata using isotropic size broadening (as confirmed by HRTEMsnapshots), Pseudo-Voigt reflection profile approximation withCaglioti equation (FWHM = U tan θ + V tan θ + W ) and experi-mental instrumental broadening parameters with standard refer-ence material. The background was modeled using a 6th-ordershifted polynomial. The parameters refined were zero shift, scalefactors, unit cell and peak shape.The morphology of the produced nanoparticles was checkedby transmission electron microscopy (TEM) (FEI Technai G2 F20operating at 200 kV) in bright field mode. Dynamic light scat-tering (DLS, Malvern Instruments, Zetasizer Nano S90) was usedto measure the hydrodynamic size distribution in particle suspen-sions produced by electric ablation. For the measurement theinitial suspension was diluted with distilled water to achieve op-timum count rate.The ferromagnetic properties of the produced samples were in-vestigated by vibrating sample magnetometry (VSM). The mag-netization curve was recorded at r.t. (Lake Shore Cryotronics in-strument 7404VSM) in a field range up to 1 T. Several full magne-tization cycles have been performed to detect potential hystereticbehavior. The preparation of reference sample (P) within a standard sol-gelprocedure involves the hydrolysis of TiCl , which generates highlyacidic (low pH) and chloride (Cl − ) rich solution. The selectivityof formed TiO polymorphs is highly dependent on instantaneouspH, Ti concentration and reaction time , where small changein molar ratios would result in different phase structure . Athigh Ti concentration ([Ti] > 0.25 M) the solution will predom-inantly contain [TiO(OH ) ] + monomers and the condensationwill proceed by olation by sharing equatorial edges with for-mation of hydroxo bridges (Ti-OH-Ti) . In this case, rutile-richcrystallites are precipitated . To stabilize the production ofanatase, Ti concentration was decreased to 0.1 M . More-over, the pH was raised to 6 to avoid the formation of a sub-stantial amount of brookite . Conversion of the amorphoushydrated precipitate (P) to crystalline anatase is realized by calci-nation treatment.The organometallic Ti precursors E1 and more concentrated E2were produced by liquid-liquid extraction: the extraction of Ti + from aqueous phase (aq) into the organic phase (o) by pentanoicacid is realized by cation exchange (Fig. 1), according to equation Ti + ( aq ) + HOOCR ( o ) ↔ Ti ( OOCR ) ( o ) + H + ( aq ) Fig. 2
TGA-DSC traces of carboxylate organometallic precursors E1(0.15M Ti) and E2 (0.5M Ti) produced by liquid-liquid extraction and sol-gel precipitates P from monophasic precipitation. Symbols indicate sam-pling temperatures T i , ◦ C: 350, 400, 450, 550, 650, 750
Pentanoic acid is an efficient cation exchanger due to itsfavourable hydrophylic/hydrophobic balance. The organometal-lic extracts, which were brightly violet immediately after extrac-tion (Fig. 1), decolorated shortly after production due to thechange of the Ti oxidation state from Ti + to Ti + and were long-term storable. The thermal behaviour of the precursors, phase transition andphase stability of produced TiO nanoparticles were studied us-ing TGA-DSC analysis. The measured thermoanalytical curves inthe range between 20 and 700 ◦ C are shown in Fig. 2. The ther-mogravimetric analysis of the as-dried reference precipitate (P)shows active weight loss (ca. 20wt.%) when the sample is heatedfrom r.t. to 200 ◦ C, which is accompanied by a pronounced en-dothermic peak (peak max at 90 ◦ C) in the DSC trace. This step isattributed to the loss of weakly bound physisorbed water from thesurface of the particles. The second step in the range between 200 ◦ C and 400 ◦ C (peak max at 290 ◦ C) is identified by a broad en-dothermic process related to the liberation of crystallization wa-ter and chemisorbed surface hydroxyls from hydrated precipitates(titanium oxyhydrate). This is accompanied by a small mass lossca. 5 wt.%. According to the XRD results (see below) the crys-tallization of anatase from the precipitates starts at about 450 ◦ C,which does not associate to any DSC event. Further temperatureincrease results in prolonged weight loss of the precipitates at areduced rate, associated with continuing dehydration process upto 1000 ◦ C.The organometallic carboxylate Ti precursors E1 and E2 pro-
Journal Name, [year], [vol.] , ig. 3 X-ray diffraction patterns (symbols) and corresponding Rietvield-refined profiles (black lines): A - calcined precipitates produced by standardsol-gel method. B - TiO samples produced by pyrolyzation of organometallic precursor. The residual lines show the difference of the measured andsimulated XRD profiles. duced by liquid-liquid extraction show a strong endothermic peakin the range between 100 ◦ C and 186 ◦ C related to the activeevaporation of free extractant (pentanoic acid) and is accom-panied by a pronounced mass loss ca. 75 wt.%, which termi-nates around the boiling point of pentanoic acid (T boil =186 ◦ C).At this point the mass loss ends and amounts to 80-85 wt.%.Further temperature increase initiates the thermal decompositionof non-reactive titanium pentanoate contained in the precursor,which should proceed according to the standard route for transi-tion metal carboxylates : Ti(OOCR) → TiO +2R CO+2CO .The decarboxylation of the organic component is accompaniedby the formation of ketones containing 3 to 9 carbon atoms (i.e.dibutyl ketone, dipropyl ketone, diethyl ketone, acetone) .The ketones evolved during this process also undergo decomposi-tion with atmospheric oxygen to yield CO and H O . The thirdstep in the temperature range between 200 ◦ C and ca. 450 ◦ C ismarked by the broad endothermal process associated with break-down of organic groups and release of volatile species, mainlyCO and formed ketones . The signature of this process ismore pronounced in the DSC trace of precursor E2 having higherTi concentration. The liberation of the titanium ion from theorganometallic precursor leads to the formation of amorphousnuclei of TiO nanoparticles. This process is accompanied bythe gradual increase of the sample mass (by 5-6 wt.%) owing tothe oxygen binding from the air during titanium oxidation. Theexothermic event at ca. 320 ◦ C in the DSC trace indicates the en-ergy release from combustion/burning of carbonaceous residues.The more gradual removal of organic groups continues up to 450 ◦ C, as shown by infrared spectroscopy (see below), which indi-cates that the carboxylate (COO − ) functional group from pen-tanoic acid is fully desorbed at this temperature. TGA analysisshows that the conversion of titanium pentanoate to TiO is com-plete at 450 ◦ C with no further mass change. As confirmed byXRD analysis (see below) the produced TiO is amorphous giv-ing no XRD trace of crystalline structure. The formation of highcrystalline anatase starts at ca. 450 ◦ C accompanied by a broadexothermic process (peak max ca. 530 ◦ C) associated with re-structuring and grain growth of anatase particles. The transfor-mation is completed at ca. 560 ◦ C the product fully transformsinto anatase phase.Gravimetric analysis confirmed 0.15M titanium loading in pre-cursor E1 and 0.5M in E2, correponding to the efficiency of cationexchange 45% and 100% respectively. Higher ratio of aqueous toorganic phase volumes (5:1 vs 3:1) achieves complete extractionand efficient loading of metal cations in the organic phase.As evident from thermal analysis the formation mechanismof nanoparticles from produced organometallic precursors differsubstantially from the standard hydrolytic sol-gel process. The de-tails of the phase content, stability and surface functional groupsof produced nanoparticles can be derived from the XRD and FTIRanalysis.
The precipitate (P) from standard "one pot" sol-gel system wasused as precursor to produce TiO nanoparticles by calcination.The XRD patterns of calcined nanopowders are shown in Fig. 3A Journal Name, [year], [vol.][vol.]
The precipitate (P) from standard "one pot" sol-gel system wasused as precursor to produce TiO nanoparticles by calcination.The XRD patterns of calcined nanopowders are shown in Fig. 3A Journal Name, [year], [vol.][vol.] , able 1 Summary of sample phase structure refinement. A - anatase, R - rutile, Imp. - impurity phase Na Ti O T, ◦ C Primary phase Secondary phaseUnit cell, Å wt.% Phase Crystallite Unit cell, Å wt.% Phase Crystallitea, b c size, d
XRD a, b c size, d
XRD
Sol-gel reference sample (P)r.t. - amorphous -450 3.795 9.492 100% A 7 nm550 3.793 9.497 100% A 9 nm650 3.790 9.504 95% A 12 nm 4.593 2.958 5% R 18 nm750 4.596 2.960 79% R 64 nm 15.1093.749 9.181 21% Imp. 45 nmPyrolyzed carboxylate precursor E1 (0.15M)350 - amorphous -400 - amorphous -450 3.794 9.481 100% A 8 nm550 3.790 9.504 100% A 15 nm650 3.788 9.513 65% A 24 nm 4.596 2.961 35% R 27 nm750 4.593 2.960 100% R 59 nmPyrolyzed carboxylate precursor E2 (0.5M)350 - amorphous -400 - amorphous -450 3.789 9.489 100% A 8 nm550 3.788 9.517 77% A 20 nm 4.597 2.960 23% R 27 nm650 4.595 2.961 81% R 46 nm 3.785 9.520 19% A 33 nm750 4.595 2.962 100% R 57 nm and used as reference. The XRD pattern of the as dried precipitatedoes not show clear signs of crystallinity, indicating that the TiO precursor before calcination is in amorphous state. The annealingtemperature of 450 ◦ C induces the crystallization of precipitateand distinctive peaks emerge, attributable to the tetragonal struc-ture (space group: I41/amd) of crystalline anatase. All the peaksare fully indexed within ICDD PDF Pattern 00-021-1272 indicat-ing that anatase titania has been formed at this temperature. Theresults of the Rietvield refinement concerning phase content andcrystallite size are reported in Table 1.Fig. 3B shows the XRD patterns of powders produced by pyrol-ysis of the organometallic precursor (E1). The sample pyrolyzedat 400 ◦ C shows only a weak and diffuse peak at 2 θ = 25.3 ◦ at-tributable to the [101] reflection of anatase, which demonstratesthat the sample still remains in mostly amorphous state. Pyrolyza-tion at 450 ◦ C induces the formation of pure nano-crystallineanatase from this precursor. The corresponding diffractogram inFig. 3B closely reflects the XRD pattern of precipitate (P) calcinedat 450 ◦ C (Fig. 3A) indicating identical phase state. The crys-tallite size of the produced anatase nanoparticles is ca. 7-8 nm(Table 1).When the processing temperature of the precursors is increasedto 650 ◦ C weak [110] (27.4 ◦ , 100% intensity) and [101] (36.1 ◦ )reflections appear in the XRD pattern of calcined precipitates cor-responding to the tetragonal structure (space group: P42/mnm)of crystalline titania rutile polymorph, which marks the beginningof a high-temperature anatase-to-rutile transformation accompa-nied by rapid grain growth (Table 1). The diffractograms of pow-ders obtained by pyrolyzation of organometallic precursor at thistemperature show additional reflections of rutile. All additionalpeaks are readily indexed within standard ICCD PDF Pattern 00-021-1276 for rutile.At 750 ◦ both precursors complete the transformation fromanatase to rutile and rutile becomes the predominant crystalline phase. This course of transformation is consistent with previousreports . No traces of brookite phase are found in any of thesamples as confirmed by the absence of indicative [121] (30.8 ◦ ,90% intensity) reflection (ICDD PDF 00-029-1276).At 750 ◦ C simultaneously with the anatase to rutile transfor-mation the calcination of the precipitate obtained by the refer-ence method finally leads to the crystallization of the impurityphase in P-750 (Fig. 3A) identified as sodium titanate Na Ti O (ICDD PDF 00-014-0277, space group: C12/m1). It is evidentthat an appreciable amount (ca. 20%, Table 1) of by-productsof the standard monophasic "one-pot" synthesis are remaining inthe produced materials despite extensive purification measuresand the confirmed absence of chloride in washwater effluent. Inthis case sodium from the precipitating basic agent NaOH bindsthe chloride from the titanium chloride solution forming a largeamount of sodium chloride in the precipitate, which is difficult toremove. The impurity phase becomes apparent in the XRD as theproduct of the high-temperature reaction between the unremovedresidues of the sodium salt and titanium dioxide during calcina-tion. In industrial pyro-hydrolysis process used to produce com-mercial TiO the powder is treated with steam at 450-550 ◦ C toremove chlorine-containing groups. Our results show that withinthe non-hydrolytic extraction-pyrolysis method the purification isachieved already at the precursor preparation stage, employinga biphasic system, in which one of the phases accumulates thetarget product, Ti-containing organometallic compound, and theother accumulates by-products of the synthesis, in this case chlo-ride anions.Moreover, non-hydrolytic pyrolysis of organometallic extract al-lows facile phase control by just varying the temperature of thepyrolytic treatment. The samples E1-550 and E2-550 producedby pyrolysis at 550 ◦ C mimic the mixed phase content of Aerox-ide P25, consisting of ca. 70-80 wt.% anatase and 20-30 wt.%rutile with ca. 20-27 nm particle size in both phases (Table 1).
Journal Name, [year], [vol.] , ig. 4 TEM images of samples produced by pyrolysis of organometallic precursor at 450 ◦ C (E1-450, top row) and 750 ◦ C (E1-750, bottom row).Middle - magnified high-resolution TEM image of few ca. 10 nm anatase nanocrystals (top) and a secondary aggregate of crystallographically alignedand fused ca. 10 nm rutile grains (bottom). Right - lattice fringes of nanoparticles and corresponding projections of anatase (A) and rutile (R) crystalstructures.
The phase evolution of the particles is independent of the con-centration of organometallic precursors, with just the start of theanatase-to-rutile conversion beginning at ca. 100 ◦ C lower tem-perature, which allows to use concentrated precursors to producephase-pure anatase.
Fig. 4 shows characteristic TEM snapshots of TiO particles pro-duced by pyrolyzation of organometallic precursor E1 at 450 ◦ C(E1-450). The particles are roughly spheroidal, which is charac-teristic for nucleation directly from the amorphous state. A tightsize distribution is measured by particle counting with an aver-age size of 8.3 nm in agreement with XRD assessment based onRietvield modelling (Table 1). High resolution TEM shows lat-tice fringes unambiguously identifiable as projections of anatasecrystal structure.As indicated by DSC analysis (Fig. 2), at ca. 560 ◦ C the amor-phous phase is fully expended and the primary crystallite forma-tion terminates. Pyrolyzation at higher temperature is accom-panied by primary crystallite fusion with simultaneous anatase-to-rutile transition. Fig. 4 shows the particles pyrolyzed at 750 ◦ C (E1-750). The highly irregular-shaped particles are con-structed by joining of a number of ca. 10 nm primary crystallinegrains forming secondary aggregates. According to strict ther-modynamic arguments, single-phase anatase is only stable be-low ca. 11-14 nm size due to its slightly lower surface free en-ergy , with rutile being the macroscopically (ca. >35 nm) sta-ble polymorph . Epitaxial attachment of aggregated phase-pure anatase nanoparticles leads to joining on their facets, wherethe rutile nucleation is initiated assisted by intermediate phase,e.g. brookite or high-pressure TiO − II phase , formationat twinned interface between contacting anatase grains .Rutile transformation then propagates rapidly into the bulk of thecrystallites. The epitaxially fused nanoparticles with the rutilestructure are noted in the high resolution TEM snapshots. Closeinspection of TEM microphotographs (Fig. 4) clearly reveals thatthe particles have many adjacent rutile crystallites sintered to-gether and crystallographically aligned with respect to each other.The average size of the secondary aggregates is 70.3 nm, consis-tent with XRD analysis (Table 1). Journal Name, [year], [vol.][vol.]
Fig. 4 shows characteristic TEM snapshots of TiO particles pro-duced by pyrolyzation of organometallic precursor E1 at 450 ◦ C(E1-450). The particles are roughly spheroidal, which is charac-teristic for nucleation directly from the amorphous state. A tightsize distribution is measured by particle counting with an aver-age size of 8.3 nm in agreement with XRD assessment based onRietvield modelling (Table 1). High resolution TEM shows lat-tice fringes unambiguously identifiable as projections of anatasecrystal structure.As indicated by DSC analysis (Fig. 2), at ca. 560 ◦ C the amor-phous phase is fully expended and the primary crystallite forma-tion terminates. Pyrolyzation at higher temperature is accom-panied by primary crystallite fusion with simultaneous anatase-to-rutile transition. Fig. 4 shows the particles pyrolyzed at 750 ◦ C (E1-750). The highly irregular-shaped particles are con-structed by joining of a number of ca. 10 nm primary crystallinegrains forming secondary aggregates. According to strict ther-modynamic arguments, single-phase anatase is only stable be-low ca. 11-14 nm size due to its slightly lower surface free en-ergy , with rutile being the macroscopically (ca. >35 nm) sta-ble polymorph . Epitaxial attachment of aggregated phase-pure anatase nanoparticles leads to joining on their facets, wherethe rutile nucleation is initiated assisted by intermediate phase,e.g. brookite or high-pressure TiO − II phase , formationat twinned interface between contacting anatase grains .Rutile transformation then propagates rapidly into the bulk of thecrystallites. The epitaxially fused nanoparticles with the rutilestructure are noted in the high resolution TEM snapshots. Closeinspection of TEM microphotographs (Fig. 4) clearly reveals thatthe particles have many adjacent rutile crystallites sintered to-gether and crystallographically aligned with respect to each other.The average size of the secondary aggregates is 70.3 nm, consis-tent with XRD analysis (Table 1). Journal Name, [year], [vol.][vol.] , .4 Infrared (IR) spectroscopy While XRD provides the information about the bulk of the parti-cles, FT-IR is a surface-sensitive method. It probes the the surfaceregion and chemistry of the particles, which is not identified byXRD. FT-IR spectra of nanoparticles produced by pyrolyzation of
Fig. 5
IR spectra of TiO samples produced by pyrolyzation ofcarboxylate-based Ti precursor: (a) - E1-350, (b) - E1-400, (c) E1-450,(d) E1-550 (complete elimination of surface residual organics is noted at450 ◦ C); (e) - electro-dispersed sample calcined at 850 ◦ C (ED-850). carboxylate precursor (E1) are shown in Fig. 5. The broad IR ab-sorption band at 3200-3600 cm − in all samples corresponds tothe stretching vibration of the hydroxyl groups ( ν OH) terminatedon TiO surface (Ti-OH) and adsorbed molecular water. The spec-trum band centered at 1625 cm − is from the bending vibrationof chemisorbed water δ H-O-H. Weak absorbance doublet at 2341cm − and 2360 cm − is due to asymmetric C-O stretching vibra-tion of carbon dioxide. Since this band is present only in thepyrolyzed samples, we ascribe it to CO residues from the oxi-dation of pentanoic acid captured on the pores in the nanopow-der * . Peaks at 1375 cm − and 1520 cm − represent symmet-ric ν COO − sym and asymmetric ν COO − asym stretching modes of thecarboxylate (COO − ) functional group coming from the organiccomponent of the precursors used in the synthesis process. Themagnitude of the separation between carboxylate stretches ∆ = ν COO − asym - ν COO − sym = ∼
145 cm − is consistent with the valuefor ionic carboxylate complexes ( ∼
164 cm − for ionic acetate ).This typically indicates that the carboxylate group is bound to thesurface Ti-centers in bidentate bridging configuration , whereone metal cation is bound to one of the oxygens of the COO − group and another metal cation to the other oxygen. Carboxy-late vibrations disappear with pyrolyzation temperature startingat 450 ◦ C, indicating complete elimination of organic residueswhen the anatase crystallization starts. A remaining broad and * cf. Fig. 5e, showing FT-IR of nanoparticles produced by electrodispersion of ini-tial metallic powder without any chemicals and calcined at high temperature (seebelow) strong absorption band between 1000 and 400 cm − envelops aset of peaks corresponding to the intrinsic Ti-O-Ti, O-Ti-O andTi-O lattice vibrations of nano-crystalline titanium oxides . Fig. 6
Defect-induced room-temperature ferromagnetism in Ti-oxidenanoparticles: A - schematic polaron model of mesoscale ferromagneticcoupling between magnetic moments produced at cationic or anionic de-fect sites in stoichiometric TiO . M-H magnetization curves recordedfrom: B - precipitates in a monophasic system calcined at (a) 550 ◦ C(P-550, anatase phase), (b) 650 ◦ C (P-650), (c) 750 ◦ C (P-750, rutile);B - pyrolyzed Ti-carboxylate precursor at (a) 450 ◦ C (E1-450, anatase),(b) 550 ◦ C (E1-550, anatase), (c) 750 ◦ C (E1-750, rutile); C - electri-cally dispersed nanoparticles calcined at (a) 350 ◦ C (ED-350), (b) 450 ◦ C (ED-450), (c) 550 ◦ C (ED-550), (d) 750 ◦ C (ED-750).
To obtain an estimation of the defect content in the producedmaterials we characterize their ferromagnetic behaviour by VSM.The recorded magnetization curves M-H are shown in Fig. 6 withthe magnetic field varying in the range between -10 kOe to +10kOe. TiO precipitates produced by standard sol-gel are predom-inantly paramagnetic M ≈ χ H (Fig. 6B), irrespective of the cal-cination temperature. The mass magnetic susceptibility χ of thesamples calcined at 550 ◦ C (P-550) and 650 ◦ C (P-650) havinganatase as the dominant phase is ca. × − cm g − . Theanatase-to-rutile transition in the calcined samples is accompa-nied just by a slight increase in χ at 750 ◦ C (P-750, Table 2).In contrast, the M-H curves of powders produced by pyrolyza-tion of organometallic precursor (Fig. 6C) show distinct room-temperature ferromagnetism superposed by the paramagneticprocess. For all recorded magnetization curves the hysteresis isnegligible with small values of coercitive field ranged within 50-90 Oe and the remanence factor varying around 5-8%, indicatingsoft superparamagnetic-like behaviour characteristic for d0 ferro-magnetism . Journal Name, [year], [vol.] , able 2 Summary of magnetic properties. The polaron dipole moment µ in units of Bohr magneton µ B = 9.27 ··· × − erg/G. T, ◦ C χ / − M sat / − N / µµ − B / cm g − emu g − g − Sol-gel reference sample (P)550 1.8 0.3 0.3 12650 2.1 0.1 0.1 15750 2.4 0.3 0.4 9.7Pyrolyzed carboxylate precursor (E1)450 0.6 1.6 1.5 12550 0.5 1.8 2.0 9.7750 1.3 4.8 4.0 13Electrodispersed sample (ED)350 12.0 13 16 8.6450 6.8 15.5 19 8.6550 3.2 9.5 12 8.6750 1.4 4.3 4.8 9.7
Stoichiometric TiO , having all Ti in 4+ oxidation state is non-ferromagnetic due to the lack of unpaired electrons. Likewise,no magnetic impurities are introduced during the preparation orhandling of the samples. The staring metallic Ti powder is chemi-cally pure titanium as confirmed by X-Ray fluorescence. The con-tent of magnetic impurities, i.e. Fe, Ni or Co, is below 0.1 at.%.The doping of the produced particles by the trace elements poten-tially present in the precursors is discarded as the origin of room-temperature ferromagnetism, because all samples were producedfrom the same metallic powder. Moreover, the absence of impu-rities in the sample produced by pyrolyzation of organometallicprecursor is assured, on account of high selectivity of pentanoicacid extractant towards titanium ions and unfavorable conditionsfor the concurrent bi-phasic extraction of other metallics in pre-cursor preparation. Hence, room temperature ferromagnetism(RTFM) in the pyrolyzed samples is due to the intrinsic defectsin the nanoparticles.The origin of RTFM, known also as d0 magnetism, in undopedsemiconducting metal oxides (SMO) is ascribed to intrinsic de-fects in the crystal lattice, generated during the production pro-cess . The emergence of ferromagnetism is typically accompa-nied by a high number of oxygen vacancies V O in the anionicsublattice of TiO films , nanoparticles and single crys-tals . Likewise, the defect complex Ti + -V O was noted, whichforms when oxygen is removed: the charge imbalance makes un-paired excess electrons occupy 3d state of nearby Ti ions, gen-erating reduced Ti + ions and providing the local magnetic mo-ments . In oxygen-defficient TiO − δ nanoparticles the reduc-tion mechanism Ti + → Ti + can also be initiated by Ti intersti-tials, without involving oxygen vacancies . Moreover, cationicvacancies V Ti were shown to produce d0-ferromagnetism .The balance of these defects is strongly dependent on the prepa-ration method.In view of this, we have produced a reference sample withhigh defect loading by physical condensation - direct electric abla-tion of the starting metallic titanium precursor without additionalchemicals, using a device shown schematically in Fig. 7A .Briefly, a 20 g portion of metallic titanium powder was added to200 ml of distilled water forming a particle bed. Flat titaniumelectrodes connected to a spectrometric discharge generator op- Fig. 7
Electric ablation experiment to produce highly defected Ti-oxidenanoparticles: A - schematic of the electric dispersion apparatus withthe discharge ablation zone between a pair of electrodes pressed intothe bed of titanium particles in distilled water. The repeated dischargesablate material from the surface of the particle bed; B - particle size dis-tribution in as prepared colloid measured by DLS.
Fig. 8
TEM snapshots of electrically dispersed nanoparticles calcined at850 ◦ C with varying magnification. HR-TEM image of lattice fringes andcorresponding rutile lattice projection. erating at 100 Hz and 2 kV were pressed onto the particle bed.After turning on the generator a luminous network of intermit-tent sparks appears within the particle bed. Reactive dispersionof titanium using distilled water as the working dielectric is ac-companied by rapid ejection and oxidation of vaporized titaniumand the formation of titanium oxide nanoparticles. The rapid insitu quenching leads to a highly defected oxygen-deficient struc-ture.The initially transparent liquid turned completely black after30 min of device operation, evidencing the condensation of nan-odispersed phase. Dynamic light scattering (DLS, Malvern Instru-ments, Zetasizer Nano S90) (Fig. 7B) showed that the hydrody-namic particle size in the decanted supernatant varies in a broadrange between ca. 100 nm to 1 µ m, which indicates that theparticles form associates in liquid medium due to the lack of sta-bilization . The produced suspension was dried, the elec-trodispersed nanopowder (ED) was calcined at 350 ◦ C (ED-350),450 ◦ C (ED-450), 550 ◦ C (ED-550), 650 ◦ C (ED-650), 750 ◦ C(ED-750), 850 ◦ C (ED-850) in static air for 2 h to sample phasestructure transformation and ferromagnetic properties.The phase content of electrodispersed nanoparticles is non-stoichiometric oxygen-deficient oxides (Fig. 9, Table 3), predom-
Journal Name, [year], [vol.][vol.]
Journal Name, [year], [vol.][vol.] , ig. 9 XRD patterns of electrodispersed nanoparticles (ED) and phasecontent evolution under calcination treatment. inantly titanium monoxide TiO (space group: Fm-3m, indexedwithin ICDD PDF 01-086-2352), which is able to accommodateup to 10-20% of O vacancies in the cubic structure , but alsotrigonal Ti O structure (space group: R3c, ICDD PDF 01-085-0868) and cubic Ti O structure (space group: P-31c, ICDD PDF01-073-1117) have been found with average crystallite size ca.25-35 nm. Anatase appears after calcination at 350 ◦ C and thecontent of non-stoichiometric oxides begins to decrease due to amore complete oxidation. A mixed phase sample ED-750 of stoi-chiometric anatase (14 wt.%) and rutile TiO (86 wt.%) is finallyobtained calcining at 750 ◦ C with grain size ca. 50-60 nm. Highercalcination temperature (850 ◦ C) completes the anatase-to-rutiletransition with particles having slightly irregular rounded shape(Fig. 8) and the lattice fringes are fully indexed to projections ofrutile structure.As expected, due to non-stoichiometry and a high defect con-tent generated by the very rapid reactive quenching of electricallyablated vapourized material in surrounding water, the producednanoparticles show significantly enhanced ferromagnetic proper-ties (Fig. 6D).Theoretical studies indicate that concentrated cationic andanionic defects couple ferromagnetically by exchange interac-tion, forming bound magnetic polarons (BMPs), which are bub-bles of ferromagnetially ordered spins (Fig. 6A) . More quan- Table 3
Summary of electrodispersed sample phase structure refine-ment. A - anatase, R - rutile. T, ◦ C Unit cell, Å wt.% Phase Crystallitea,b c size, d
XRD r.t. 4.190 4.190 62% TiO 26 nm5.092 13.811 25% Ti O
12 nm5.148 9.580 13% Ti O 35 nm350 4.187 4.187 53% TiO 23 nm5.066 13.803 19% Ti O
12 nm5.147 9.548 16% Ti O 26 nm3.787 9.428 12% A 24 nm450 4.589 2.961 45% R 13 nm3.785 9.526 23% A 52 nm4.195 4.195 14% TiO 37 nm5.134 9.626 9% Ti O 28 nm5.149 13.642 9% Ti O
12 nm550 4.593 2.958 71% R 27 nm3.784 9.519 22% A 49 nm5.124 9.659 7% Ti O 27 nm650 4.596 2.961 81% R 33 nm3.786 9.526 18% A 52 nm5.137 9.695 below 5% Ti O 10 nm750 4.596 2.961 86% R 52 nm3.786 9.527 14% A 62 nm850 4.593 2.960 ∼ titative understanding of the polaron-induced magnetism can beobtained within the BMP-model by fitting the measured magneti-zation with the equation of the type M = M sat L ( ξ ) + χ H (1)where the first term describes the superparamagnetic contribu-tion and the second term describes to first order of magnetic fieldH the paramagnetic processes in non-ferromagnetic matrix, whichare not related to the superparamagnetic properties of the sam-ple. Here, M sat is the saturation magnetization [emu g − ], whichis the product of the number of bound magnetic polarons involvedin the polarization process per unit mass of the sample N [g − ]and µ - the spontaneous dipole moment [erg G − ] of a polaron; L ( x ) = coth ( x ) − x is the Langevin function, ξ = µ Hk B T - Langevinparameter, k B = . ... × − erg K − is the Boltzmann con-stant, T - temperature (ca. 300 K) and χ is paramagnetic suscep-tibility [cm g − ]. We have analyzed the recorded M-H curves forall samples in terms of the bound polaron model Eq. (1) usingpolaron concentration N , their dipole moment µ and magneticsusceptibility χ as fitting parameters. The theoretical lines closelyfollow the measured M-H curves for all samples (Fig. 6). Thecalculated parameters are summarized in Table 2.The superparamagnetic saturation magnetization of samplesproduced by pyrolyzation of organometallic precursor is aboutan order of magnitude higher than for reference sol-gel particles,indicating a more defected structure. M sat consistently increaseswith the pyrolyzation temperature, reaching 4.8 memu/g at 750 ◦ C (E1-750).The highest superparamagnetic saturation M sat =15.5 memu/gis obtained for electrodispersed sample ED-450, indicating highlydefected structure was produced. Calcination at high temperature>450 ◦ C leads to the decrease of ferromagnetic properties due tothe gradually diminishing defect content, such as the filling of
Journal Name, [year], [vol.] , xygen vacancies in static air. The remaining superparamagneticsaturation at the highest annealing temperature 750 ◦ C is M sat =4.3 memu/g.The phase composition, crystallite size (ca. 50-60 nm) and po-laron content ( ∼ × g − ) in both rutile-rich samplesE1-750 and ED-750 (Table 2) produced respectively by pyrolyza-tion and electrodispersion is similar. It is worth noting that ac-cording to the TGA traces (Fig. 2) showing continuous mass gain,the pyrolyzation of organometallic precursor relies on oxygen ab-sorption from the environment, which might favor formation ofoxygen vacancies. This is in contrast to the precipitated sample,which shows similar phase composition and crystallite size, butan order of magnitude smaller polaron concentration, indicatinga defect-free crystal lattice, since the stoichiometry of the precip-itates in aqueous sol-gel is assured on account of the hydrolysisand condensation process.Previously, 10 nm TiO − δ nanoparticles with mostly anatasephase produced by sol-gel showed saturation magnetization 9.5memu/g and coercitivity of 96 Oe after annealing under reduc-ing H /Ar atmosphere. The origin of the ferromagnetism wasattributed to a large content of V O . This was contested by Parraset al. , where Magneli phase TiO . was stabilized under sim-ilar conditions. Rather than the influence of oxygen vacancies,the saturation magnetization 10 memu/g was instead ascribed tothe Ti + ions induced by reduction, initiated by formation of Tiinterstitials. Polycrystalline TiO treated under 4 MeV Ar + ion ir-radiation showed 0.4 memu/g saturation with 168 Oe coercitivefield . Oxygen vacancies V O were the dominant defects afterirradiation. Wang et al. produced 10-20 nm anatase nanopar-ticles by sol-gel and reported ∼ O content and charge transfer inter-actions between Ti + and Ti + . Recently , sol-gel produced10 nm single-phase anatase TiO calcined at 400 ◦ C showed ca.0.5-13 memu/g saturation and 11-305 Oe coercitivity, on accountof V O . Wang et al. produced highly defected 10-20 nm anataseTiO nanoparticles with 25 memu/g superparamagnetic satura-tion and 18 Oe coercitivity using solvothermal reaction. More-over, defected TiO showed remarkably higher photocatalytic ac-tivity vs. normal TiO because of the more efficient charge trans-fer .Our results show that the effective polaron magnetic moment µ ∼ µ B ( ∼ × − emu, Table 2) does not distinctly de-pend on the preparation method or thermal treatment. Remark-ably, very similar magnitudes of polaron magnetic moments wereobtained in two previous investigations, in which this quantitywas estimated: µ ∼ . − . × − emu was measured in largenanorods with diameter ∼
100 nm and length of few µ m pro-duced by NaOH-assisted solvothermal processing with vacuumannealing and/or low-pressure calcination. The phase contentvaried from pure brookite to anatase-rutile mix with large con-centration of V O and Ti + states. The same µ = − . × µ B polaron moment was measured in sol-gel produced 10 nm single-phase anatase due to V O , despite saturation magnetizationbeing roughly 2 orders of magnitude smaller than in the firststudy . Our results show that the polaron properties are roughlyunaffected neither by the preparation method, phase content of the produced TiO , nor annealing conditions, which lead just tothe variability of polaron concentration N in the samples.A unambiguous one-to-one linear correlation exists betweenthe concentration of bound magnetic polarons N and the oxygenvacancy content . Theoretical calculations indicate that a sin-gle oxygen vacancy V O can produce magnetic moment of about1.4-2.4 µ B , cation vacancy V Ti – about 2.3-4 µ B and di-vacancy – 2 µ B . For a characteristic moment ∼ µ B per defect, ca. 0.4-0.7% of defective cells are found in the elec-trodispersed powders with defect content decreasing with highercalcination temperature. In contrast, in pyrolyzed samples thedefect content is ca. 0.01-0.04% without an obvious dependenceon the pyrolyzation temperature. An organometallic precursor based on titanium carboxylate com-plex has been developed to produce phase-pure anatase, rutileand mixed phase TiO nanoparticles with controllable polymorphcontent using a simple non-hydrolytic thermal process. The newapproach, based on a different reaction mechanism, eliminatesfrom consideration many reaction parameters, which are difficultto control in the hydrolytic sol-gel processes, particularly, such asinstantaneous Ti:H O ratio, hydrolysis rate, pH, reaction modu-lators, and allows better control of the reaction process. A broadrange of samples with varying structures, sizes and phase compo-sitions has been produced by pyrolyzation of the carboxylate pre-cursor. The temperature of the pyrolytic treatment comes out as asingle parameter determining the grain size and crystal phase tun-ing of the final product in a simple, predictable and reproducibleway.The carboxylate precursor produced by biphasic cation ex-change is long-term storable, hydrolytically stable and not sensi-tive to moisture contained in the air, unlike inorganic Ti chloridesor organometallic alkoxides commonly employed in the standardmethods, which undergo vigorous hydrolysis. On this account ti-tanium carboxylates are also ideal precursors for film deposition.The high chemical purity of the produced materials is assuredon account of the biphasic extraction involved in the preparationof the precursor. Common impurities characteristic for the stan-dard "one-pot" production methods have not been found duringthe investigation. The results are correlated against the tradi-tional sol-gel method and confirm that the reported process isa good technique to produce TiO nanoparticles having surfacesfree from organics.An advantage of the the carboxylate precursor in pyrogenic syn-thesis is that it allows to avoid the commonly used titanium chlo-rides and the related corrosive hazards in chlorine rich flames.Flame synthesis is a widely employed to produce high qual-ity metal oxide nanoparticles through a one-step process. Hence,the carboxylate precursor is a favourable alternative in the pro-duction of TiO nanoparticles. Acknowledgements
We acknowledge technical staff from the University of Latvia,which was involved in the preparation of some of the samplesunder grant agreement No. 1.1.1.1/16/A/085 (2017-2018).
10 | 1–12
Journal Name, [year], [vol.][vol.]
Journal Name, [year], [vol.][vol.] , otes and references Chemical Engineer-ing Journal , 2018, , 2408–2439.2 C. Byrne, G. Subramanian and S. C. Pillai,
Journal of Environ-mental Chemical Engineering , 2018, , 3531–3555.3 I. Ali, M. Suhail, Z. A. Alothman and A. Alwarthan, RSC Ad-vances , 2018, , 30125–30147.4 C. Wu, Y. Yue, X. Deng, W. Hua and Z. Gao, Catalysis Today ,2004, , 863–869.5 Y. K. Kho, A. Iwase, W. Y. Teoh, L. Mädler, A. Kudo andR. Amal,
The Journal of Physical Chemistry C , 2010, ,2821–2829.6 G. Li, C. P. Richter, R. L. Milot, L. Cai, C. A. Schmuttenmaer,R. H. Crabtree, G. W. Brudvig and V. S. Batista,
Dalton Trans-actions , 2009, 10078.7 G. Li, L. Chen, M. E. Graham and K. A. Gray,
Journal of Molec-ular Catalysis A: Chemical , 2007, , 30–35.8 R. Bacsa and J. Kiwi,
Applied Catalysis B: Environmental ,1998, , 19–29.9 X. Jiang, M. Manawan, T. Feng, R. Qian, T. Zhao, G. Zhou,F. Kong, Q. Wang, S. Dai and J. H. Pan, Catalysis Today , 2018, , 12–17.10 R. I. Bickley, T. Gonzalez-Carreno, J. S. Lees, L. Palmisanoand R. J. Tilley,
Journal of Solid State Chemistry , 1991, ,178–190.11 T. Kawahara, Y. Konishi, H. Tada, N. Tohge, J. Nishii andS. Ito, Angewandte Chemie International Edition , 2002, ,2811–2813.12 D. C. Hurum, K. A. Gray, T. Rajh and M. C. Thurnauer, TheJournal of Physical Chemistry B , 2005, , 977–980.13 B. Ohtani, O. Prieto-Mahaney, D. Li and R. Abe,
Journal ofPhotochemistry and Photobiology A: Chemistry , 2010, ,179–182.14 G. Goutailler, C. Guillard, S. Daniele and L. G. Hubert-Pfalzgraf,
Journal of Materials Chemistry , 2002, , 342–346.15 J. Yu, Journal of Catalysis , 2003, , 69–78.16 Y. Bessekhouad, D. Robert and J. V. Weber,
Journal of Photo-chemistry and Photobiology A: Chemistry , 2003, , 47–53.17 S. L. Isley and R. L. Penn,
The Journal of Physical Chemistry B ,2006, , 15134–15139.18 M. R. Ranade, A. Navrotsky, H. Z. Zhang, J. F. Banfield, S. H.Elder, A. Zaban, P. H. Borse, S. K. Kulkarni, G. S. Doran andH. J. Whitfield,
Proceedings of the National Academy of Sci-ences , 2002, , 6476–6481.19 Y.-J. Liu, M. Aizawa, Z.-M. Wang, H. Hatori, N. Uekawa andH. Kanoh, Journal of Colloid and Interface Science , 2008, ,497–504.20 A. D. Paola, G. Cufalo, M. Addamo, M. Bellardita, R. Cam-postrini, M. Ischia, R. Ceccato and L. Palmisano,
Colloids andSurfaces A: Physicochemical and Engineering Aspects , 2008, , 366–376.21 J. Cihlar, V. Kasparek, M. Kralova and K. Castkova,
Interna-tional Journal of Hydrogen Energy , 2015, , 2950–2962. 22 G. Símonarson, S. Sommer, A. Lotsari, B. Elgh, B. B. Iversenand A. E. Palmqvist, ACS Omega , 2019, , 5750–5757.23 A. Pottier, S. Cassaignon, C. Chanéac, F. Villain, E. Tronc andJ.-P. Jolivet, Journal of Materials Chemistry , 2003, , 877–882.24 K. Schumacher, A. Schild and M. Morters, Flame-hydrolyticallyproduced titanium dioxide powder , Jun. 2005.25 W. W. Yu and X. Peng,
Angew. Chem., Int. Ed. , 2002, , 2368.26 W. W. Yu, L. Qu, W. Guo and X. Peng, Chem. Mater. , 2003, ,2854.27 W. W. Yu, J. C. Falkner, C. T. Yavuz and V. L. Colvin, Chem.Commun. , 2004, 2306–2307.28 J. Park, K. An, J.-G. P. Yosun Hwang, H.-J. Noh, J.-Y. Kim, J.-H. Park, N.-M. Hwang and T. Hyeon,
Nature Materials , 2004, , 891–895.29 V. Serga, D. Zablotsky, A. Kr¯umin , a, M. Lubane and G. Heide-mane, Key Engineering Materials , 2018, , 288–293.30 Y. Wang, Y. He, Q. Lai and M. Fan,
Journal of EnvironmentalSciences , 2014, , 2139–2177.31 A. D. Paola, M. Bellardita, R. Ceccato, L. Palmisano and F. Par-rino, The Journal of Physical Chemistry C , 2009, , 15166–15174.32 R. A. Roca and E. R. Leite,
Journal of the American CeramicSociety , 2012, , 96–102.33 D. Reyes-Coronado, G. Rodríguez-Gattorno, M. E. Espinosa-Pesqueira, C. Cab, R. de Coss and G. Oskam, Nanotechnology ,2008, , 145605.34 S. Abbasi, S. M. Zebarjad, S. H. N. Baghban and A. Youssefi, Synthesis and Reactivity in Inorganic, Metal-Organic, andNano-Metal Chemistry , 2015, , 1539–1548.35 Y. Zhou, C. Chen, N. Wang, Y. Li and H. Ding, The Journal ofPhysical Chemistry C , 2016, , 6116–6124.36 A. Rai and G. Parashar,
Thermochimica Acta , 1979, , 175–179.37 A. Seddon and J. Wood, Thermochimica Acta , 1986, , 341–354.38 M. D. Judd, B. A. Plunkett and M. I. Pope,
Journal of ThermalAnalysis , 1974, , 555–563.39 J. Leicester and M. J. Redman, Journal of Applied Chemistry ,2007, , 357–366.40 H. Zhang and J. F. Banfield, Chemistry of Materials , 2002, ,4145–4154.41 R. Nicula, M. Stir, C. Schick and E. Burkel, ThermochimicaActa , 2003, , 129–136.42 H. Zhang and J. F. Banfield,
Journal of Materials Chemistry ,1998, , 2073–2076.43 H. Zhang and J. F. Banfield, The Journal of Physical ChemistryB , 2000, , 3481–3487.44 K.-R. Zhu, M.-S. Zhang, J.-M. Hong and Z. Yin,
Materials Sci-ence and Engineering: A , 2005, , 87–93.45 R. L. Penn and J. F. Banfield,
American Mineralogist , 1998, ,1077–1082.46 Y.-H. Zhang, C. K. Chan, J. F. Porter and W. Guo, Journal ofMaterials Research , 1998, , 2602–2609. Journal Name, [year], [vol.] , American Mineralogist , 1999, , 877–883.48 S.-C. Zhu, S.-H. Xie and Z.-P. Liu, Journal of the AmericanChemical Society , 2015, , 11532–11539.49 H. Zhang and J. F. Banfield,
Journal of Materials Research ,2000, , 437–448.50 G. Oskam, A. Nellore, R. L. Penn and P. C. Searson, The Jour-nal of Physical Chemistry B , 2003, , 1734–1738.51 N. T. Nolan, M. K. Seery and S. C. Pillai,
The Journal of Physi-cal Chemistry C , 2009, , 16151–16157.52 K. Nakamoto,
Infrared and Raman Spectra of Inorganic andCoordination Compounds , John Wiley & Sons, Inc., 2008.53 V. Zeleˇnák, Z. Vargová and K. Györyová,
Spectrochimica ActaPart A: Molecular and Biomolecular Spectroscopy , 2007, ,262–272.54 C. D. C. (editor), The Coblentz Society Desk Book of InfraredSpectra. 2nd ed. , The Coblentz Society, 1982.55 P. D. Esquinazi, W. Hergert, M. Stiller, L. Botsch, H. Ohldag,D. Spemann, M. Hoffmann, W. A. Adeagbo, A. Chassé, S. K.Nayak and H. B. Hamed, physica status solidi (b) , 2020, ,1900623.56 J. M. D. Coey, M. Venkatesan and C. B. Fitzgerald,
NatureMaterials , 2005, , 173–179.57 S. D. Yoon, Y. Chen, A. Yang, T. L. Goodrich, X. Zuo, D. A.Arena, K. Ziemer, C. Vittoria and V. G. Harris, Journal ofPhysics: Condensed Matter , 2006, , L355–L361.58 S. D. Yoon, Y. Chen, A. Yang, T. L. Goodrich, X. Zuo, K. Ziemer,C. Vittoria and V. G. Harris, Journal of Magnetism and Mag-netic Materials , 2007, , 171–175.59 A. K. Rumaiz, B. Ali, A. Ceylan, M. Boggs, T. Beebe and S. I.Shah,
Solid State Communications , 2007, , 334–338.60 Y. Liu, S. Feng, Z. Li, L. Zhang, G. Wang, W. Chen, T. Wangand W. Zhong,
RSC Advances , 2016, , 57013–57018.61 Q. Zhao, P. Wu, B. L. Li, Z. M. Lu and E. Y. Jiang, Journal ofApplied Physics , 2008, , 073911.62 H. Liu, G. P. Li, Q. L. Lin, D. J. E, X. D. Gao, X. B. Wei, X. D.Pan, S. X. Zhang, J. J. Ding and W. Lan,
Journal of Supercon-ductivity and Novel Magnetism , 2019, , 3557–3562.63 H. Liu, G. Li, D. E, N. Xu, Q. Lin, X. Gao, C. Lan, J. Chen,C. Wang, X. Zhan and K. Zhang, RSC Advances , 2020, ,18687–18693.64 S. Zhou, E. ˇCižmár, K. Potzger, M. Krause, G. Talut, M. Helm,J. Fassbender, S. A. Zvyagin, J. Wosnitza and H. Schmidt, Physical Review B , 2009, , 113201.65 B. Santara, P. K. Giri, K. Imakita and M. Fujii, Nanoscale , 2013, , 5476.66 M. Parras, Á. Varela, R. Cortés-Gil, K. Boulahya, A. Hernandoand J. M. González-Calbet, The Journal of Physical ChemistryLetters , 2013, , 2171–2176. 67 H. Peng, J. Li, S.-S. Li and J.-B. Xia, Physical Review B , 2009, , 092411.68 S. Wang, L. Pan, J.-J. Song, W. Mi, J.-J. Zou, L. Wang andX. Zhang, Journal of the American Chemical Society , 2015, , 2975–2983.69 M. M. Maiorov, D. Zablotsky, E. Blums and A. Krumina,
IOPConference Series: Materials Science and Engineering , 2019, , 012029.70 I. Segal, A. Zablotskaya, E. Lukevics, M. Maiorov, D. Zablot-sky, E. Blums, I. Shestakova and I. Domracheva,
AppliedOrganometallic Chemistry , 2008, , 82–88.71 A. Zablotskaya, I. Segal, E. Lukevics, M. Maiorov, D. Zablot-sky, E. Blums, I. Shestakova and I. Domracheva, Journal ofMagnetism and Magnetic Materials , 2009, , 1428–1432.72 I. Segal, A. Zablotskaya, E. Lukevics, M. Maiorov, D. Zablot-sky, E. Blums, A. Mishnev, R. Georgieva, I. Shestakova andA. Gulbe,
Applied Organometallic Chemistry , 2010, , 193–197.73 A. Zablotskaya, I. Segal, M. Maiorov, D. Zablotsky andE. Blums, Inorganic Frameworks as Smart Nanomedicines , El-sevier, 2018, pp. 279–316.74 A. Zablotskaya, I. Segal, Y. Popelis, A. Mishnev, M. Maiorov,D. Zablotsky, E. Blums, V. Nikolajeva and D. Eze,
AppliedOrganometallic Chemistry , 2015, , 376–383.75 M. D. Banus, T. B. Reed and A. J. Strauss, Physical Review B ,1972, , 2775–2784.76 J. Lu, K. Yang, H. Jin, Y. Dai and B. Huang, Journal of SolidState Chemistry , 2011, , 1148–1152.77 D. Zablotsky, S. Kralj, G. Kitenbergs and M. M. Maiorov,
Jour-nal of Non-Newtonian Fluid Mechanics , 2020, , 104248.78 D. Zablotsky, S. Kralj and M. M. Maiorov,
Colloids and Sur-faces A: Physicochemical and Engineering Aspects , 2020, ,125079.79 D. Sanyal, M. Chakrabarti, P. Nath, A. Sarkar, D. Bhowmickand A. Chakrabarti,
Journal of Physics D: Applied Physics ,2013, , 025001.80 Q. Wang, X. Wei, J. Dai, J. Jiang and X. Huo, Materials Sciencein Semiconductor Processing , 2014, , 111–115.81 V. R. Akshay, B. Arun, S. Dash, A. K. Patra, G. Mandal, G. R.Mutta, A. Chanda and M. Vasundhara, RSC Advances , 2018, , 41994–42008.82 A. Chanda, K. Rout, M. Vasundhara, S. R. Joshi and J. Singh, RSC Advances , 2018, , 10939–10947.83 H.-K. Ma and H.-A. Yang, Journal of Alloys and Compounds ,2010, , 115–122.84 M. Y. Manuputty, J. A. H. Dreyer, Y. Sheng, E. J. Bringley, M. L.Botero, J. Akroyd and M. Kraft,
Chemical Science , 2019, ,1342–1350.
12 | 1–12
Journal Name, [year], [vol.][vol.]