Decomposition Pathways of Titanium Isopropoxide Ti (OiPr) 4: New

Novosibirsk State University, 2, Pirogova Str., 630090 Novosibirsk, Russia b ... of Chemical Kinetics and Combustion, 3, Institutskaya Str., 630090 No...
0 downloads 0 Views 2MB Size
Subscriber access provided by UNIVERSITY OF LEEDS

Article i

4

Decomposition Pathways of Titanium Isopropoxide Ti(OPr): New Insights from UV-Photodissociation Experiments and Quantum Chemical Calculations Kirill S. Ershov, Sergei A. Kochubei, Vitaly G. Kiselev, and Alexey V. Baklanov J. Phys. Chem. A, Just Accepted Manuscript • Publication Date (Web): 15 Jan 2018 Downloaded from http://pubs.acs.org on January 16, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Decomposition Pathways of Titanium Isopropoxide Ti(OiPr)4: New Insights from UV-Photodissociation Experiments and Quantum Chemical Calculations Kirill S. Ershov,a,b Sergei A. Kochubei,c Vitaly G. Kiselev,a,b,d* and Alexey V. Baklanova,b* a b

Novosibirsk State University, 2, Pirogova Str., 630090 Novosibirsk, Russia Institute of Chemical Kinetics and Combustion, 3, Institutskaya Str., 630090 Novosibirsk,

Russia c

Institute of Semiconductor Physics, 13, Lavrentyeva Ave., 630090 Novosibirsk, Russia

d

Semenov Institute of Chemical Physics RAS, 4, Kosygin Str., 119991 Moscow, Russia

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

ABSTRACT

UV-photodissociation at 266 nm of a widely used TiO2 precursor, titanium tetraisopropoxide (Ti(OiPr)4, TTIP), was studied under molecular beam conditions. Using the MS-TOF technique, atomic titanium and titanium(II) oxide (TiO) were detected among the most abundant photofragments. Experimental results were rationalized with the aid of quantum chemical calculations (DLPNO-CCSD(T) and DFT). Contrary to existing literature data, the new fourcentered reaction of acetone elimination was found to be the primary decomposition process of TTIP. According to computational results, the effective activation barrier of this channel is ~49 kcal/mol, which is ~13 kcal/mol lower than that of the competing propylene elimination. The former process followed by a dissociative loss of H atom is a dominating channel of TTIP unimolecular decay. The sequential loss of isopropoxy moieties via these two-step processes is supposed to produce the experimentally observed titanium atoms. In turn, combination of these reactions with propylene elimination can lead to another detected species, TiO. These results indicate that the existing mechanisms of TTIP thermal and photoinitiated decomposition used for chemical vapor deposition (CVD) of titanium dioxide should be reconsidered.

ACS Paragon Plus Environment

2

Page 3 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

INTRODUCTION

Titanium dioxide (TiO2) is widely used as a semiconductor, photocatalyst, and pigment.1-4 Among the huge variety of the preparation procedures, the gas-phase synthetic methods allow for better control of the properties of particles or thin films of TiO2. More specifically, the flame oxidation5-8 and the chemical vapor deposition (CVD) along with thermal9-12 or photoinduced decomposition of TTIP13,14 are typically applied for a gas-phase synthesis of TiO2. These techniques yield TiO2 nanoparticles with the desired shape, narrow size distribution, and high crystallinity.9,11 Titanium tetraisopropoxide Ti(OiPr)4 (TTIP) is often used as TiO2 precursor in the gas-phase synthetic methods. Even though TTIP has a lower vapor pressure in comparison with another widely used precursor, TiCl4 (0.1 and 12 Torr, respectively, at 298 K),15 the former species is chlorine-free and allows for TiO2 deposition at lower temperatures.11 The mechanism and kinetics of TTIP thermal decomposition are necessary for modeling of the TiO2 formation in the flame or under CVD conditions. In turn, the proper kinetic model is crucial for reliable extrapolation of obtained kinetic parameters beyond the range of temperatures and other conditions of a particular experiment. However, even though gas-phase thermolysis of TTIP have been extensively studied experimentally and theoretically, both the kinetics and mechanism of initial stages of thermolysis still remain unclear. Okuyama et al.11 studied the thermolysis of TTIP vapor in a flow reactor in the temperature range 500–660 K. The authors proposed the overall process of TTIP consumption Ti(OiPr)4→ TiO2 + 4C3H6 + 2H2O.

(1)

The effective first-order rate constant of TTIP thermal decomposition (the propylene formation was monitored) was estimated to be k = 3.96·105×exp(-8479.7/T) s-1. Apart from this, acetone and propylene with molecular hydrogen16 as well as methanol and ethanol along with radical species15 were detected as the gas-phase products of TTIP decomposition. Kinetics of the TTIP thermal decomposition was also studied experimentally by monitoring either the time dependence of the TiO2 film thickness10,16 or the size and concentration of TiO2 aerosol nanoparticles.11 The reported kinetic parameters, however, refer entirely to the surface reactions.

ACS Paragon Plus Environment

3

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 29

Shmakov et al.6 studied the products and intermediates of the TTIP conversion in the H2/O2 flame using mass-spectrometry and microthermocouple techniques. On the basis of measured concentration profiles of TTIP and simple products, the authors concluded that hydrolysis is the predominant channel of TTIP consumption in the flame. Apart from this, the authors performed DFT calculations for both bimolecular (hydrolysis) and unimolecular channels of TTIP depletion. In a good agreement with observed concentration profiles, the hydrolysis was found to dominate with a moderate activation barrier of ~12 kcal/mol.6 On the other hand, among the unimolecular channels, the lowest activation barrier of ~57 kcal/mol was calculated6 for 4-centered elimination of propylene molecule Ti(OiPr)4 → (iPrO)3TiOH + C3H6.

(2)

According to DFT calculations, the two barrierless radical channels corresponding to C-O and C-C bond cleavage have endothermicity higher than 85 kcal/mol: Ti(OiPr)4 → (iPrO)3Ti• + •OiPr,

(3)

Ti(OiPr)4 → (iPrO)3TiO• + •C3H7.

(4)

17,18

In the recent papers, Buerger et al.

have studied theoretically the mechanism of TTIP thermal

decomposition with the aid of DFT (B97-1/6-311+G(d,p) level of theory) and CBS-Q composite procedure. Using the RRKM and transition state theory calculated rate constants, the authors proposed the kinetic mechanism of TTIP thermolysis.18 In agreement with the previous computational results,6 they concluded18 that the predominant decomposition channel of TTIP is a sequential four-centered propylene elimination leading ultimately to Ti(OH)4 (Scheme 1). Taking into account the previous experimental results (Eq. 1),11 it is natural to expect that Ti(OH)4 finally transforms to TiO2 (Scheme 1).

Scheme 1. Monomolecular decomposition pathways of TTIP proposed in the literature. Using the proposed kinetic mechanism (Scheme 1), the authors18 also calculated the ignition delay times at various temperatures for TTIP in the dry air. However, the calculated values turned out to be almost an order of magnitude higher than the experimental counterparts. This difference can indicate some important pathways missing in the considered mechanism.

ACS Paragon Plus Environment

4

Page 5 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

It is therefore clear that mechanism of TTIP thermal decomposition is incomplete and to some extent contradictory. Moreover, the kinetics of elementary TTIP decomposition reactions have never been studied experimentally, only the indirect measurements10,11,16 or experiments under flame conditions6 were performed. Therefore, in the present contribution we probe the unimolecular steps of TTIP dissociation both experimentally and theoretically. To this end, we employed the UV-photodissociation of TTIP molecules under molecular beam conditions with mass-spectral probing of photofragments. The molecular beam conditions used in the present work allow us to exclude any secondary bimolecular processes. The detected photofragments provided a clear evidence that the mechanism of TTIP decomposition should be reconsidered. The experimental findings were rationalized with the aid of quantum chemical calculations of several relevant stationary points on the potential energy surfaces for TTIP and intermediates of its thermal decomposition.

2. EXPERIMENTAL AND COMPUTATIONAL METHODS 2.1 Experimental The supersonic molecular beam of TTIP was prepared in the pulsed molecular beam apparatus with the self-made electrodynamic valve generating gas jet (pulse duration 200 µsec). The premixed gas mixture comprised of 0.1 Torr of TTIP and helium (balance) as a carrier gas was expanded into the chamber through the 0.27 mm nozzle at a backing pressure of 1 atm. The mixture was prepared by bubbling of helium through a liquid sample of TTIP (Sigma–Aldrich, purity >98%). The central part of the gas jet passed through the 2.5 mm skimmer mounted 60 mm downstream and entered the region of a homogeneous electric field created by the electrodes of the time-of-flight mass-spectrometer (TOF-MS). The molecular beam was directed perpendicular to the TOF axis. The TTIP was irradiated at a wavelength 266 nm with a pulse energy up to 5 mJ (the 4th harmonic of Nd-YAG laser). The laser radiation was focused onto molecular beam by a lens with a focus length of 17.5 cm. The ions were detected with the microchannel plate (MCP) detector. The UV/vis absorption spectra of TTIP were recorded with the spectrophotometer Varian Cary 50. A gas sample with an optical path of 10 cm was prepared in the quartz cylindrical cell. A droplet of liquid TTIP was placed on a bottom of the cell filled with argon. At room temperature (24 ºC), a TTIP vapor pressure is 0.1 Torr.15 2.2. Quantum chemical calculations

ACS Paragon Plus Environment

5

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

Geometry optimizations and frequency calculations of all stationary points on the potential energy surfaces (PES) were performed using the Truhlar’s group M06-2X density functional and the 6-311++G(2df,p) basis set.19 All equilibrium and transition state structures were ascertained to be the minima or saddle points, accordingly, on the PES. Zero-point energies and thermal corrections to enthalpy and Gibbs free energy were computed at the same level of theory. Note that the M06-2X functional was benchmarked to perform uniformly well for various kinetic problems.20,21 For several relevant PES points, the single-point electronic energies were refined using the recently proposed DLPNO-CCSD(T) method22 along with aVTZ basis set. All electronic structure calculations were carried out using the Gaussian 0923 and ORCA24 program packages.

3. RESULTS 3.1. Photofragments of TTIP decomposition: experimental study Figure 1 displays the experimental UV spectrum as well as the TD-B3LYP calculated positions and oscillator strengths of the electronic transitions for the TTIP in the gas phase. It is seen that the calculations reproduce experimental spectrum reasonably well. According to calculations (see more details in the Supporting Information), the band with a maximum at 255 nm originates from a series of electronic transitions, all of them are comprised mainly of electron promotions from three highest occupied MOs to two lowest unoccupied MOs. These occupied orbitals are composed mostly of the lone pairs of the O atoms, while the unoccupied – are predominantly the d-orbitals of the titanium atom (see Supporting Information). Thus, these transitions are characterized by a ligand to metal charge-transfer (LMCT).

ACS Paragon Plus Environment

6

Page 7 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 1. UV-Vis spectrum of gaseous titanium tetraisopropoxide (TTIP) at 24 ºC, the optical path length is 10 cm. Vertical blue bars indicate the positions and oscillator strengths (f, right axis) of the electronic transitions calculated at the TD-B3LYP/6-311++G(2df,p) level. Therefore, for further study of TTIP photodissociation we used the excitation at 266 nm, which lies within the most intensive UV band (red arrow, Fig. 1). Figure 2 represents the massspectrum of photoions observed after irradiation of TTIP at 266 nm. Very interestingly, we detected only the ions Ti+ and TiO+, which appear in the mass-spectra as groups of five isotope peaks (Fig. 2, inset). These peaks were attributed to stable titanium isotopes

46

Ti-50Ti with the

integrals corresponding to their natural abundance with a maximum contribution of 48Ti (natural abundance 73.9 %). The ionization potentials of Ti and TiO are very close to each other: IP(Ti) = IP(TiO) = 6.82 eV.25,26 Thus, the two quanta of radiation at 266 nm (hν=4.66 eV) are necessary for ionization of these fragments. The entire absence of larger fragments definitely indicates fragmentation of TTIP under irradiation conditions employed in our experiments.

ACS Paragon Plus Environment

7

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

Figure 2. Mass-spectrum of photoions produced by irradiation of titanium tetraisopropoxide at 266 nm. Insets: the isotopic structure of Ti+ and TiO+ peaks. The spectral region corresponding to the mass of TiO2 is encircled in red. If the sequential unimolecular decay of TTIP proceeds in accordance with the overall reaction equation (1), we would expect the TiO2 to be a product of TTIP photofragmentation. However, there are no bands in the corresponding spectral region (Fig. 2, encircled in red). Nevertheless, on the basis of these facts, the formation of TiO2 in our experiments cannot be entirely excluded. The reason is a quite high ionization potential of titanium dioxide IP(TiO2)=9.5 eV.27 Consequently, the three laser quanta at 266 nm are necessary for ionization of TiO2 in comparison with the two quanta necessary for ionization of Ti and TiO (recall, their IPs are ~ 6.8 eV).25,26 Therefore, under experimental conditions employed, the yield of TiO2+ ions can be profoundly less than that for Ti+ and TiO+ even in the case of comparable concentrations of neutral species. It should be emphasized that the recently proposed detailed mechanism of TTIP decomposition18 (Scheme 1) cannot describe the observed formation of the photofragmentation

ACS Paragon Plus Environment

8

Page 9 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

products Ti and TiO (Fig. 2). The sequential propylene elimination (Scheme 1) does not yield neither atomic titanium nor TiO. These species can be formed only in decomposition pathways occurring via the cleavage of Ti-O bond. At the same time, the elementary reaction of Ti-O bond rupture (3) has an unfavorable endothermicity (by ~30 kcal/mol higher than the elimination of propylene (2)) and cannot compete with (2). To shed more light on this problem, we employed quantum chemical calculations, which are discussed in detail in the next section.

3.2. Quantum chemical calculations of decomposition channels of TTIP

(

)

Figure 3 represents the relative enthalpies at 0 K ( ∆ ∆H 0K ) for the lowest decomposition channels of TTIP calculated at the M06-2X/6-311++G(2df,p) level of theory (all raw computational data are available in the Supporting Information). Note that some of the species under study have two or more conformers. Only properties of the lowest energy conformers will be discussed henceforth. The most striking result of computations is that the new predominant channel never discussed before is found. This is a four-centered elimination of acetone (TS1, Fig. 3) followed by formation of a strong complex of acetone with a product (OiPr)3Ti-H (Fig. 3, left side). Note that the C=O moiety of an acetone molecule acts as a ligand and coordinates the Ti center in the complex with a Ti-O distance of 2.27 A (cf. Ti-O bond length of 1.80 A in TTIP). This renders the complex quite strong (the enthalpy of formation ~ −11 kcal/mol, Fig. 3). Finally, the effective activation barrier of the whole process Ti(OiPr)4→(iPrO)3Ti-H + C3H6O

(5)

is 48.9 kcal/mol. This is profoundly lower (by more than 12 kcal/mol) than the activation barrier of the above discussed process (2) of the four-centered propylene elimination (TS2, Fig. 3, right side). The branching ratio between the channels yielding propylene (2) and acetone (5) is determined by the corresponding rate constants  and  . This ratio can roughly be estimated as

 

≈ exp (−

∆(∆  ) 

), where ∆(∆  ) is the difference between the effective

activation enthalpies for the channels (2) and (5). Given the latter value ∆(∆  ) = 12.7 kcal/mol (Fig. 3), the estimation yields ~10-9 at 298 K and ~10-3 at 1000 K, respectively. Moreover, elimination of acetone from the intermediate complex is barrierless (Fig. 3, left side) and therefore occurs via a “looser” transition state than the four-centered propylene elimination

ACS Paragon Plus Environment

9

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

(TS2, Fig. 3, right side). Therefore, the present estimation, obtained using solely the enthalpies of the channels, most likely renders the upper bound of the true branching ratio. It is also worth mentioning here that the Ti-O and O-C bond energies in the TTIP calculated at the M06-2X level of theory differ from the reported18 B97-1 counterparts by ~3-4 kcal/mol. The similar difference exists between the M06-2X and B3LYP values6 of the TS2 activation barrier (Fig. 3). In general, these discrepancies lie within the scatter range typical of DFT. Moreover, to provide more convincing evidence, we also recalculated the single-point energies of the most important PES points at the DLPNO-CCSD(T)/aVTZ level of theory. It is seen from Figure 3 that the latter values are quite close to M06-2X results (within ~2 kcal/mol). This fact confirms the reliability of the M06-2X functional employed in the present work.

(

)

Figure 3. The relative enthalpies at 0 K ( ∆ ∆H 0K ) of the stationary points on the PES corresponding to thermal decomposition of TTIP. The TTIP was chosen as a reference compound for the calculations of the relative thermodynamic properties. All values are in kcal/mol and calculated at the M06-2X/6-311++G(2df,p) level of theory. For several PES points, the sums of DLPNO-CCSD(T)/aVTZ single-point electronic energies and M06-2X zero-point vibrational energies are given in parentheses. Recall that the bimolecular reactions do not occur under conditions of a molecular beam. Therefore, the primary product (iPrO)3Ti-H can further dissociate via the unimolecular channels shown in Figure 4. It is seen that the Ti-H bond cleavage with the calculated activation barrier of 48.8 kcal/mol (Fig. 4, left side) (iPrO)3Ti-H→(iPrO)3Ti• + H•

(6)

ACS Paragon Plus Environment

10

Page 11 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

is a dominating channel of unimolecular decay of (OiPr)3Ti-H. A competing channel of acetone elimination (Fig. 4, right side) (iPrO)3Ti-H→(iPrO)2TiH2 + C3H6O

(7)

has a notably higher activation barrier of 55.2 kcal/mol. Moreover, a preexponetial factor of the rate constant for bond cleavage is usually several orders of magnitude higher than that for fourcentered molecular elimination.

(

)

Figure 4. The relative enthalpies at 0 K ( ∆ ∆H 0K ) of the stationary points on the PES corresponding to thermal decomposition of the hydride intermediate (iPrO)3Ti-H. The latter species was chosen as a reference compound for the calculations of the relative thermodynamic properties. Inset: relative enthalpies of the possible biradical intermediates. All values are in kcal/mol and calculated at the M06-2X/6-311++G(2df,p) level of theory. Apart from the channels (6) and (7), we also considered the hypothetical three-centered elimination of isopropanol (8) yielding a Ti-centered biradical species: (iPrO)3Ti-H→(iPrO)2Ti: + iPrOH

(8)

The calculated relative enthalpy of the products of this channel is shown in the inset of Figure 4. Due to open-shell nature of the singlet (iPrO)2Ti: species, we optimized the triplet compound, and estimated the energy of the open-shell singlet (OSS in Fig. 4) using the broken-symmetry approach.28 As seen from Figure 4, the lowest electronic state of this biradical is a triplet, lying only slightly higher in energy (~2 kcal/mol, Fig. 4) than the products of Ti-H bond cleavage reaction (6). The OSS is unfavorably thermodynamically: it is higher than radical products on the

ACS Paragon Plus Environment

11

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

enthalpic scale by 19 kcal/mol (Fig. 4). Although we did not localize the corresponding transition state, the contribution of the triplet channel (8) to the rate of TTIP disappearance is expected to be negligible due to several reasons. First, the preexponetial factor of the rate constant of a threecentered elimination is expected to be essentially lower (by 2-3 orders of magnitude) than that for bond cleavage reactions.29 Second, non-conservation of the spin should further drastically reduce the rate of reaction (8).30 Thus, we infer that the Ti-H cleavage reaction (6) is the predominant process of (iPrO)3Ti-H unimolecular decay. Finally, we conclude that the new channel of TTIP dissociation ultimately leads to elimination of one isopropoxide group in a sequence of two reactions with an overall endothermicity of 97.7 kcal/mol (Scheme 2, cf. Scheme 1).

Ti(OiPr)4

-C3H6O

-H

( iPrO)3TiH

( iPrO)3Ti

Scheme 2. Dominating primary monomolecular channels of thermal decomposition of TTIP proposed in the present work.

(

)

Figure 5. The relative enthalpies at 0 K ( ∆ ∆H 0K ) of the stationary points on the PES corresponding to thermal decomposition of the radical intermediate (iPrO)3Ti•. The latter species was chosen as a reference compound for the calculations of the relative thermodynamic

ACS Paragon Plus Environment

12

Page 13 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

properties. All values are in kcal/mol and calculated at the M06-2X/6-311++G(2df,p) level of theory. This reaction sequence can be repeated with a radical (iPrO)3Ti•. The results of calculations given in Figure 5 confirm that the similar channel of four-centered acetone elimination (9) dominates for decay of this radical: (iPrO)3Ti• → (iPrO)2TiH• + C3H6O

(9)

with the endothermicity of 55.3 kcal/mol (Fig. 5, left side). Note that the intermediate complex (iPrO)2TiH•…C3H6O is unusually strong – its formation enthalpy is ~ −24 kcal/mol (Fig. 5). Similar to the case of reaction (5), the C=O moiety of an acetone molecule coordinates the titanium radical center in the complex with a Ti-O distance of 2.10 A (cf. Ti-O bond length of 2.27 A in the non-radical (iPrO)3TiH…C3H6O), which is not much longer than the Ti-O bond length of 1.82 A in the (iPrO)2TiH• radical. The activation barriers of competing channels are notably higher: viz., 61.3 kcal/mol for the C-O bond cleavage (Fig. 5, right side): (iPrO)3Ti• → (iPrO)2Ti=O + •iPr

(10)

and 66.6 kcal/mol for the channel of 4-centered propylene elimination (Fig. 5, left side) (iPrO)3Ti•

i

→ ( PrO)2Ti-OH

+ C3H6.

(11)

Finally, the hydride radical (iPrO)2TiH• formed in the reaction (9), can further undergo transformations, which are similar to reactions (6) and (9).

4. DISCUSSION According to above discussed computational results, the acetone elimination reaction (5) is the dominating primary reaction of TTIP dissociation. To explain the formation of atomic titanium, we can propose the stepwise process (Scheme 3, upper part) for total decomposition of TTIP to yield free Ti atoms. Taking into account the calculated energy values necessary for the first (97.7 kcal/mol) and second (115 kcal/mol) steps of this process (cf. Fig. 3 and 4), we can estimate the total energy necessary for production of Ti atoms to be higher than the energy of four photons of the 266 nm wavelength (hν≈107.5 kcal/mol) employed in our experiments.

ACS Paragon Plus Environment

13

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

Scheme 3. Outline of TTIP thermal decomposition mechanism under collisionless conditions of the present work. On the other hand, formation of titanium oxide TiO also observed in our experiments (Fig. 2) can be explained by a combination of reaction steps with elimination of acetone and propylene (Scheme 3, lower part). Note that sequential elimination of acetone and propylene agree well with existing experimental data on TTIP thermal decomposition.16 Both species were detected as the major products of temperature-programmed desorption of TTIP adsorbed on copper. These products appeared sequentially: the acetone was observed in the primary process at lower temperatures.16 It is therefore natural to expect notable contributions of similar processes in a mechanism of TTIP photodecomposition. To provide the sequential absorption of radiation of the same laser pulse by photofragments, the excited TTIP molecule should dissociate on a timescale shorter than the UV-pulse duration (~6-7 ns). Unfortunately, to the best of our knowledge, there are no literature data on the photodynamics of TTIP. Possible mechanism can be a conversion to the ground electronic state followed by dissociation. To estimate the rate constant of TTIP dissociation, we applied the simple approach proposed by Klots.31 This approach links the microcanonical k(E*) and canonical rate constants in the Arrhenius form k(T)=A⋅exp(-Ea/RT) (with known Arrhenius parameters A and Ea) via the so-called isokinetic condition k(E*)≡k(Tb). Klots showed that the temperature Tb can be estimated as  ∗ ≈ 〈 where 〈

!" 〉 $ is

!" 〉 $

− %&" +

() *

+ + ∙ (

() -*∙∙$

),

(13)

the average vibrational energy of a molecule, R - universal gas constant and a

product C⋅R is a heat capacity at the temperature Tb. In the case of one-photon absorption,  ∗ =hν. The endothermicity (48.9 kcal/mol) of the dominating primary process (5) of TTIP decomposition (Fig. 3) was used as an estimate of Arrhenius activation energy Ea. The average experimental preexponential factor for 4-centered elimination reactions of log (A/s-1) = 13.5 was

ACS Paragon Plus Environment

14

Page 15 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

employed for TTIP.32 The vibrational energy of the TTIP molecule was estimated using the DFT calculated harmonic wavenumbers for all 129 oscillators. Using these values, the effective temperature was estimated to be Tb ~ 850 K. In turn, the rate constant calculated with Arrhenius equation allowed us to estimate the lifetime of a one-photon excited TTIP molecule to be τ ≈ 0.1 s. This lifetime is much longer that the laser pulse duration (~6-7 ns). Therefore, the absorption of a few extra quanta is necessary for decomposition of an excited molecule. Further similar calculations of k(E*) for E*=n⋅hν with n >1 have shown that only absorption of four photons (n=4) makes an excited TTIP molecule sufficiently short-lived: viz., the lifetime w.r.t. dissociation τ≈10-8 s becomes comparable with the laser pulse duration. Using the absorption spectrum of TTIP (Fig. 1) the cross-section for absorption of one photon at 266 nm can be estimated to be σ ≈ 3⋅10-17 cm2. At the same time, the laser radiation flux at the focus of laser beam in our experiments is estimated to be about F≈1021 photon⋅cm-2. The value of the product σ⋅F≈3⋅104 >> 1 indicates that the necessary number of quanta can be absorbed by TTIP molecule during the laser pulse at our experimental conditions. This fact is in line with a proposed photodecomposition mechanism of TTIP, namely, internal conversion of electronically excited TTIP molecules to the ground electronic state followed by dissociation. However, fast dissociation of one-photon excited TTIP can also take place if the excited state has a repulsive nature or if a fast conversion of an excited state to a repulsive state takes place. We would like to emphasize that the contribution of these processes to the observed TTIP decomposition cannot be entirely excluded. E.g., Halary-Wagner et al.13 photoinitiated chemical vapor deposition of TiO2 by irradiation of TTIP by XeCl laser (308 nm) within the same absorption band as it was used in the present contribution. The deposits of TiO2 appeared at a laser radiation flux values down to F=1 mJ cm-2 ≈ 1.5⋅1015 photons⋅cm-2. The absorption crosssection σ308=3.1⋅10-18 cm2

13

together with this low flux value provide the conditions where

probability of one photon absorption is very low (σ⋅F≈5⋅10-3). Under these conditions, the multiphoton excitation of TTIP cannot occur. We cannot exclude that both processes, viz., a prompt dissociation after one-photon absorption and multiphoton excitation followed by dissociation, proceed under our experimental conditions of laser photodecomposition of TTIP.

5. CONCLUSION

ACS Paragon Plus Environment

15

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

Photodecomposition of TTIP by the pulsed UV-laser irradiation at 266 nm under collisionless conditions of molecular beam yielded Ti and TiO as the most abundant products. These results were rationalized with the aid of DFT calculations. Contrary to existing literature data,6,18 the new reaction of 4-centered acetone elimination was found to be the primary decomposition process of TTIP. According to computational results, the effective activation barrier of this channel is ~49 kcal/mol, which is ~13 kcal/mol less than that of earlier proposed propylene elimination6,18 (Fig. 3). The new process followed by a dissociative loss of H atom (Fig. 4) is a dominating channel of TTIP unimolecular decay. The radical (iPrO)3Ti•, in turn, undergoes the same acetone elimination with an effective activation barrier of ~55 kcal/mol (Fig. 5). In contrast to other channels, the acetone elimination occurs via the formation of strong complexes, where the C-O fragment of acetone coordinates the titanium containing moiety (Figures 3 and 5). The sequential loss of isopropoxy moieties via similar two-step processes is supposed to produce the experimentally observed Ti atoms (Scheme 3). In turn, combination of these reactions with propylene elimination can lead to another detected species, TiO (Scheme 3). These results indicate that the existing mechanisms of TTIP thermal and photoinitiated decomposition used for chemical vapor deposition (CVD) of titanium dioxide must be modified. The proposed kinetic scheme of TiO2 generation from TTIP via a sequence of unimolecular reactions of propylene elimination (Scheme 1) refers to a minor channel. In turn, the proper kinetic model should definitely comprise the unimolecular reactions (5), (6), and (9) (Scheme 2) as well as bimolecular processes with participation of radical intermediates (H atom, (iPrO)3Ti•, etc.) produced in the primary unimolecular channels (Scheme 3).

ASSOCIATED CONTENT Supporting Information Electronic Supplementary Information (ESI) available: the raw computational data (DFT optimized geometries, electronic energies, and zero-point vibrational energy corrections of all compounds under study), theoretical UV-Vis spectrum of TTIP. See DOI: 10.1039/x0xx00000x.

ACS Paragon Plus Environment

16

Page 17 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

AUTHOR INFORMATION Corresponding Author *To whom correspondence should be addressed: A.V.B.: E-mail: [email protected] V.G.K.: E-mail: [email protected] ORCID Vitaly G. Kiselev: 0000-0002-2721-539X Notes The authors declare no competing financial interest. ACKNOWLEDGMENT The authors are thankful to Mr. Alexandr S. Bogomolov and Ms. Margarita V. Shakhova for technical assistance. K.S.E., S.A.K., and A.V.B. acknowledge Russian Science Foundation for a financial support of the experimental part of this work (project 16-13-10024). V.G.K. is indebted to Russian Science Foundation for a financial support of the computational part of this work (project 16-13-10155); support by the Supercomputer Center of Novosibirsk State University and Siberian Supercomputer Center (SB RAS) is also acknowledged.

REFERENCES (1) Chen, X.; Mao, S. S. Titanium dioxide nanomaterials: Synthesis, properties, modifications, and applications. Chem. Rev. 2007, 107, 2891-2959. (2) Fujishima, A.; Zhang, X. T.; Tryk, D. A. TiO2 photocatalysis and related surface phenomena. Surf. Sci. Rep. 2008, 63, 515-582. (3) Henderson, M. A. A surface science perspective on TiO2 photocatalysis. Surf. Sci. Rep. 2011, 66, 185-297. (4) Nosaka, Y.; Nosaka, A. Y. Identification and roles of the active species generated on various photocatalysts. In Photocatalysis and Water Purification: From Fundamentals to Recent Applications; Pichat, P., Ed.; Wiley 2013, p 3-24.

ACS Paragon Plus Environment

17

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 29

(5) Wooldridge, M. S. Gas-phase combustion synthesis of particles. Prog. Energy Combust. Sci. 1998, 24, 63-87. (6) Shmakov, A. G.; Korobeinichev, O. P.; Knyazkov, D. A.; Paletsky, A. A.; Maksutov, R. A.; Gerasimov, I. E.; Bolshova, T. A.; Kiselev, V. G.; Gritsan, N. P. Combustion chemistry of Ti(OC3H7)(4) in premixed flat burner-stabilized H-2/O-2/Ar flame at 1 atm. Proc. Combust. Inst. 2013, 34, 1143-1149. (7) Li, S. Q.; Ren, Y. H.; Biswas, P.; Tse, S. D. Flame aerosol synthesis of nanostructured materials and functional devices: Processing, modeling, and diagnostics. Prog. Energy Combust. Sci. 2016, 55, 1-59. (8) Wang, Y.; Kangasluoma, J.; Attoui, M.; Fang, J.; Junninen, H.; Kulmala, M.; Petaja, T.; Biswas, P. The high charge fraction of flame-generated particles in the size range below 3 nm measured by enhanced particle detectors. Combust. Flame 2017, 176, 72-80. (9) Chin, S.; Park, E.; Kim, M.; Jeong, J.; Bae, G. N.; Jurng, J. Preparation of TiO2 ultrafine nanopowder with large surface area and its photocatalytic activity for gaseous nitrogen oxides. Powder Tech. 2011, 206, 306-311. (10) Siefering, K. L.; Griffin, G. L. Growth-kinetics of CVD TiO2 - influence of carrier gas. J. Electrochem. Soc. 1990, 137, 1206-1208. (11) Okuyama, K.; Kousaka, Y.; Tohge, N.; Yamamoto, S.; Wu, J. J.; Flagan, R. C.; Seinfeld, J. H. Production of ultrafine metal-oxide aerosol-particles by thermal-decomposition of metal alkoxide vapors. AIChE J. 1986, 32, 2010-2019. (12) Komiyama, H.; Kanai, T.; Inoue, H. Preparation of porous, amorphous, and ultrafine TiO2 particles by chemical vapor-deposition. Chem. Lett. 1984, 1283-1286. (13) Halary-Wagner, E.; Bret, T.; Hoffmann, P. Light-induced CVD of titanium dioxide thin films I: Kinetics of deposition. Chem. Vap. Depos. 2005, 11, 21-28. (14) Watanabe, A.; Tsuchiya, T.; Imai, Y. Selective deposition of anatase and rutile films by KrF laser chemical vapor deposition from titanium isopropoxide. Thin Solid Films 2002, 406, 132-137. (15) Filatov, E. S.; Nizard, H.; Semyannikov, P. P.; Sysoev, S. V.; Trubin, S. V.; Morozova, N. B.; Zherikova, K. V.; Gelfond, N. V. Thermal properties of some volatile titanium (IV) precursors. In Eurocvd 17 / Cvd 17; Swihart, M. T., Barreca, D., Adomaitis, R. A., Worhoff, K., Eds. 2009; Vol. 25, p 557-560. (16) Siefering, K. L.; Griffin, G. L. Kinetics of low-pressure chemical vapor-deposition of TiO2 from titanium tetraisopropoxide. J. Electrochem. Soc. 1990, 137, 814-818. (17) Buerger, P.; Nurkowski, D.; Akroyd, J.; Mosbach, S.; Kraft, M. First-principles thermochemistry for the thermal decomposition of titanium tetraisopropoxide. J. Phys. Chem. A 2015, 119, 8376-8387. (18) Buerger, P.; Nurkowski, D.; Akroyd, J.; Kraft, M. A kinetic mechanism for the thermal decomposition of titanium tetraisopropoxide. Proc. Combust. Inst. 2017, 36, 1019-1027. (19) Zhao, Y.; Truhlar, D. G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor. Chem. Acc. 2008, 120, 215-241. (20) Parkhomenko, D. A.; Edeleva, M. V.; Kiselev, V. G.; Bagryanskaya, E. G. pH-Sensitive C-ON bond homolysis of alkoxyamines of imidazoline series: A Theoretical study. J. Phys. Chem. B 2014, 118, 5542-5550.

ACS Paragon Plus Environment

18

Page 19 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

(21) Kiselev, V. G. Comment on "Decomposition mechanisms of trinitroalkyl compounds: a theoretical study from aliphatic to aromatic nitro compounds'' by G. Fayet, P. Rotureau, B. Minisini, Phys. Chem. Chem. Phys., 2014, 16, 6614. Phys. Chem. Chem. Phys. 2015, 17, 1028310284. (22) Liakos, D. G.; Sparta, M.; Kesharwani, M. K.; Martin, J. M. L.; Neese, F. Exploring the accuracy limits of local pair natural orbital coupled-cluster theory. J. Chem. Theor. Comp. 2015, 11, 1525-1539. (23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, revision A.01; Gaussian Inc.: Wallingford, CT, 2009. (24) Neese, F. The ORCA program system. Wiley Interdiscip. Rev.-Comput. Mol. Sci. 2012, 2, 73-78. (25) Rosenstock, H. M.; Draxl, K.; Steiner, B. W.; Herron, J. T. Energetics of gaseous ions. J. Phys. Chem. Ref. Data (Suppl. 1) 1977, 6. (26) Loock, H. P.; Simard, B.; Wallin, S.; Linton, C. Ionization potentials and bond energies of TiO, ZrO, NbO and MoO. J. Chem. Phys. 1998, 109, 8980-8992. (27) Hildenbrand, D. L. Mass-spectrometric studies of thermochemistry of gaseous TiO and TiO2. Chem. Phys. Lett. 1976, 44, 281-284. (28) Neese, F. Prediction of molecular properties and molecular spectroscopy with density functional theory: From fundamental theory to exchange-coupling. Coord. Chem. Rev. 2009, 253, 526-563. (29) Krajnovich, D.; Huisken, F.; Zhang, Z.; Shen, Y. R.; Lee, Y. T. Competition between atomic and molecular chlorine elimination in the infrared multi-photon dissociation of CF2Cl2. J. Chem. Phys. 1982, 77, 5977-5989. (30) Mebel, A. M.; Lin, S. H.; Chang, C. H. Theoretical study of vibronic spectra and photodissociation pathways of methane. J. Chem. Phys. 1997, 106, 2612-2620. (31) Klots, C. E. Some properties of microcanonical rate constants. Int. Rev. Phys. Chem. 1996, 15, 205-217. (32) Robinson, P. J.; Holbrook, K. A. Unimolecular Reactions Wiley-Interscience: London, 1996.

ACS Paragon Plus Environment

19

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

TOC GRAPHIC

ACS Paragon Plus Environment

20

Page 21 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 1 150x105mm (299 x 299 DPI)

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2 99x83mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 22 of 29

Page 23 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 3 392x141mm (300 x 300 DPI)

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4 388x156mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 24 of 29

Page 25 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 5 331x153mm (300 x 300 DPI)

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Scheme 1 253x12mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 26 of 29

Page 27 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Scheme 2 129x11mm (300 x 300 DPI)

ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Scheme 3 234x36mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 28 of 29

Page 29 of 29 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

TOC Graphic 79x69mm (150 x 150 DPI)

ACS Paragon Plus Environment