Demystifying Ziegler–Natta Catalysts: The Origin of Stereoselectivity

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Demystifying Ziegler-Natta Catalysts: the Origin of Stereoselectivity Antonio Vittoria, Anika Meppelder, Nicolaas Friederichs, Vincenzo Busico, and Roberta Cipullo ACS Catal., Just Accepted Manuscript • Publication Date (Web): 26 May 2017 Downloaded from http://pubs.acs.org on May 26, 2017

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.

ACS Catalysis 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 26

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

ACS Catalysis

Demystifying Ziegler-Natta Catalysts: The Origin of Stereoselectivity Antonio Vittoriaa, Anika Meppelderb, Nic. Friederichsb, Vincenzo Busicoa,*, Roberta Cipulloa,* a

Laboratory of Stereoselective Polymerizations (LSP), Department of Chemical Sciences, Federico II University, Via Cintia, 80126 Naples, Italy b

SABIC, NL-6160 AH Geleen, Netherlands

ABSTRACT. Industrial Ziegler-Natta catalysts for polypropylene production are complex formulations with a reputation of ‘black boxes’. In the present paper, we report the results of an extensive investigation of the three latest commercial generations, carried out with advanced High Throughput Experimentation tools and methods. The thus obtained database of structure/properties relationships, of extraordinary width and depth, provided a high-definition picture of the screened systems, enabling us to highlight important details of their inner workings. In particular, the delicate relationship between surface coverage and lateral steric pressure on the stereoselectivity of the catalytic species, as well as the role of chemisorbed donor and Al-alkyl species on said parameters for different formulations, were revealed. KEYWORDS. Ziegler-Natta Catalysts; Isotactic Polypropylene; Internal Donor; External Donor; Stereoselectivity. INTRODUCTION Isotactic polypropylene (i-PP) and Ti-based Ziegler-Natta (ZN) catalysts have parallel success stories. With an annual consumption of more than 60 M tons, i-PP has become the second largestvolume polymer on the market after polyethylene.1 Progress in catalysis was also dramatic: activity 1 ACS Paragon Plus Environment

ACS Catalysis

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 26

boosted from few kg to several tons of polymer per g of Ti, and stereoselectivity from moderate to almost perfect.2-4 The first two catalyst generations, consisting of crystalline TiCl3 in a layered modification3,4, were relatively simple. Stereoselectivity was the consequence of a peculiar crystal lattice with chirotopic Ti both in the bulk and on the side edges of the structural layers. The latter, in particular, exposed linear arrays of enantiomorphous Ti atoms amenable to Cl/R metathesis with an Al-alkyl compound.3-5 Elegant experimental6 and computational7,8 studies highlighted the surface constraints on the thus formed Ti-alkyls, conformationally locked at the first C-C bond so as to define chiral pockets in which the two propene enantiofaces could be discriminated at the insertion step.4,9 Supported homologues with greatly improved performance, but also much more complex formulations, were introduced in the 1970s.2-4 MgCl2 was serendipitously identified as the bestworking support for TiCl4 (the Ti precursor of largest use).2 The addition of certain donor compounds as powerful stereoselectivity enhancers (Table 1), at the precatalyst preparation stage (‘Internal Donors’, IDs) or in combination with the AlR3 activator (‘External Donors’, EDs), was also a trial-and-error achievement.2,3 Table 1. Typical formulations and performance ranges of MgCl2-supported ZN catalysts for i-PP production.2,3 Generation

Internal Donor

Third

Ethylbenzoate

Fourth

Dialkylphthalate 2,2’-dialkyl-1,3dimethoxypropane Dialkylsuccinate

Fifth Sixth (a)

3

-1 (b)

10 kg(PP) g(Ti) .

External Donor Aromatic monoester Alkoxysilane None or Alkoxysilane Alkoxysilane

Productivity(a) XS(b) Mw/Mn 0.5-0.8

3-5

6-9

1-2

1-5

6-8

>2

2-5

4-6

1-2

1-5

>8

Xylene-Soluble Fraction in wt.-%.

Unraveling the cross-interactions between the components of the different formulations that followed one another at intervals of approximately 10 years2,3,9 is a long-standing open problem, which generated the catalysts’ reputation of ‘black-boxes’. The similarity between the crystal 2 ACS Paragon Plus Environment

Page 3 of 26

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

ACS Catalysis

lattices of MgCl2 and TiCl3 originated the idea of an epitaxial relationship between at least part of the TiClx adsorbates and the MgCl2 substrate.2,3,9,10 Rather than representing a constructive input, this hypothesis triggered decades of flawed mechanistic speculations, including that of a competition between TiCl4 and the donors for selective chemisorption on the support.2,3,9,10 MgCl2(10l) crystal terminations, exposing pentacoordinated Mg, have long been claimed to host Ti2Cl8 adducts that, once activated by an AlR3, would mimic the active sites of authentic TiCl3 catalysts. MgCl2(110) terminations, in turn, featuring tetracoordinated Mg and as such more acidic, were postulated to be preferred targets for donor binding, and home to non-stereoselective sites only. The successful introduction, in the 1990s, of 2,2-dialkyl-1,3-dimethoxypropanes as a class of IDs especially prone to chelate tetracoordinated Mg, thus supposedly hampering TiCl4 interaction with MgCl2(110) facets2,3,11, was presented as a compelling demonstration of the hypothesis, and even an achievement of molecular design.11,12 It was only several years later that more critical analyses of the experimental data and Quantum Mechanics modeling studies disproved the concept of MgCl2(10l) and MgCl2(110) as ‘good’ and ‘bad’ surfaces, respectively. As a matter of fact, it became impossible to ignore the unambiguous evidence that donor molecules have a direct and specific impact on polymer microstructure13, and therefore are, if not part of the catalytic species, at least at non-bonded contact with them.9 Several independent Density Functional Theory (DFT) calculations, in turn, concluded that Ti2Cl8 adsorbates on MgCl2(10l)-like edges are not stable14,15, and lately that TiCl4 chemisorption is only feasible in mononuclear form on MgCl2(110)-like edges.16,17 The current view is that the role of donors in MgCl2-supported ZN catalyst systems is twofold: i) Stabilize the primary particles by strong chemisorption, lowering their surface energy.18-21 Mg/donor mole ratios in the range of 10 to 20 are not unusual2,3, which points to lateral dimensions of the structural layers of only few unit cells18,22, and values of surface area in excess of 150 m2 g-1 (unattainable for binary MgCl2/TiCl4 particles because TiCl4 adsorption is too weak16). 3 ACS Paragon Plus Environment

ACS Catalysis

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 26

ii) Impart the necessary steric hindrance to the inherently chiral but otherwise too open catalytic species, very much alike ancillary ligands in molecular catalysts.9 A qualitative model for this function was proposed by some of us in Ref23, and still accounts for all known facts, including the complex stereoblock microstructure of (part of) the polymers.9 Although in principle one single donor might exert both functions, in most cases ID and ED roles are differentiated (Table 1).2,3,9 Surface stabilization is important already at the stage of primary particle formation; this usually entails chlorination of a Mg precursor compound with excess TiCl4 in the presence of the ID, which requires that the latter compounds are mutually unreactive. For the modification of the catalytic species, on the other hand, a proper steric demand, the preference for chemisorption on Mg rather than Ti, and the lack of reactivity with AlR3 are equally important conditions. Some well-functioning IDs (e.g. dialkylphthalates) react irreversibly with the AlR3 activator, and are extracted from the solid catalyst during polymerization2,3,24-26; therefore, they need to be replaced by an ED. By far the most widely used EDs are sterically demanding alkoxysilanes2,3; these are poorly reactive with AlR3 compounds, but do react with TiCl4, which prevents their use as IDs. While the above general picture is sound, what is still missing is an adequate understanding of the details that would enable true catalyst design. In particular, how the catalytic species look like and what determines their diverse behaviors in the different catalyst generations remain largely unanswered questions. Ironically, formulations that work well experimentally are not trivial to validate by means of computational modeling (as we shall see in the following sections), whereas some published models claiming a high stereoselectivity involve compositions that are not used in industrial practice.17,27 In this sense, the aforementioned black-box perception is justified. In this paper, we present the results of an extensive and thorough study of fourth-, fifth- and sixthgeneration ZN catalysts (Table 1), used in combination with AlEt3 and an array of alkoxysilane EDs with large structural diversity. The investigation, carried out with advanced High Throughput 4 ACS Paragon Plus Environment

Page 5 of 26

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

ACS Catalysis

Experimentation (HTE) tools and methods28, consisted of two parts. In a first part, catalyst performance (in terms of polymerization behavior and polymer properties) was determined using a fully automated secondary screening platform with 48 mini-reactors (Freeslate PPR48), integrated with a polymer characterization workflow including Gel Permeation Chromatography (GPC), analytical Crystallization Elution Fractionation (A-CEF), and

13

C NMR spectroscopy. In a second

part, another HTE platform (Freeslate Core Module) was used to follow the evolution in composition of the solid catalysts under conditions closely mimicking those of application. The resulting database of structure/properties relationships, of extraordinary width and depth for a single investigation, provided a high-definition picture of the screened systems enabling us to highlight for the first time important details of their inner workings. In particular, we explored the delicate relationship between surface coverage at saturation and lateral steric pressure on the stereoselectivity of the catalytic species, and clarified the roles of chemisorbed donors and Al-alkyls on said parameters for different formulations (i.e. catalyst surface distributions); this is very important for further progress. In the latter respect, it should be recalled that a recent REACH ban on dialkylphthalates for toxicity issues has generated a growing market demand for their replacement1; considering that fourth-generation ZN systems are the working horses of i-PP industry2, the question is of high relevance and calls for urgent attention. EXPERIMENTAL SECTION The composition of the four (pre)catalysts, prepared according to Ref.29, can be found in Table 2. The eight alkoxysilane EDs, in turn, are listed in Table 3; ED1 was purchased from Gelest Inc., ED6 was custom-made, whereas all others were obtained from Evonik AG. Table 2. Compositions of the screened (pre)catalysts. Code C1 C2 C3 C4

Internal Donor (ID) Dibutylphthalate 2,2-Diisobutyl-1,3-dimethoxypropane 2,2-Dimethyl-1,3-dimethoxypropane 2,3-Diisopropyldiethylsuccinate

Ti (wt%) 2.0 2.7 2.1 2.4

Mg (wt%) 18.6 18.4 19.1 19.2

ID (wt%) 11.5 13.2 9.3 9.5 5

ACS Paragon Plus Environment

ACS Catalysis

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 26

Table 3. The set of screened alkoxysilane EDs. ED1

ED2

ED3

ED4

Dimethyldimethoxy-

Propyltrimethoxy-

Propyltriethoxy-

Cyclohexylmethyldimethoxy-

ED5

ED6

ED7

ED8

Diisobutyldimethoxy-

Diethylaminotriethoxy-

Diisopropyldimethoxy-

Dicyclopentyldimethoxy-

All semi-batch propene polymerization experiments in heptane slurry were run in a Freeslate Parallel Pressure Reactor setup with 48 reaction cells (PPR48), fully contained in a triple MBraun glove-box under Nitrogen. The cells, with a working volume of 5.0 mL, featured an 800 rpm magnetically coupled stirring, and individual on-line reading/control of temperature, pressure, monomer uptake and uptake rate. The setup and the operating protocol were described in full detail in Ref28. Polymerization conditions were as follows: T = 70°C; p(C3H6) = 4.5 bar; p(H2) = 0.20 bar; [Al]/[Ti] = 160; [ED]/[Al] = 0, 0.025, 0.050, 0.10, 0.20; t = 30 min. The p(C3H6)/p(H2) ratio was set at a value resulting into average polymer molecular weights in the typical commercial range. All experiments were performed in duplicate, for a total of 132 pairs. The

polymerization products were

characterized by high-temperature

Gel Permeation

Chromatography (GPC) with a Freeslate Rapid-GPC setup; analytical Crystallization Elution Fractionation (A-CEF) with a Polymer Char setup; quantitative 13C NMR with a Bruker DRX 400 spectrometer equipped with a high-temperature cryoprobe accommodating 5 mm OD tubes and a pre-heated robotic sample changer (see Ref28 and Supporting Information (SI) for detailed operating protocols and conditions). A second High Throughput Experimentation platform, namely a Freeslate Extended Core Module (XCM), also contained in a triple MBraun glove-box and enabling the robotic handling of solids, 6 ACS Paragon Plus Environment

Page 7 of 26

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

ACS Catalysis

liquids and slurries according to fully automated weighing, dispensing and reaction protocols, was utilized to determine the composition of the four solid catalysts following the reaction with AlEt3/ED mixtures or AlEt3 alone in heptane slurry under conditions mimicking those of polymerization (albeit in the absence of the monomer). For this study, aliquots of 15-20 mg of solid (pre)catalyst in a matrix of 24×8 mL pressure-tight vials were suspended in 3.8 mL of heptane, added with predefined amounts of AlEt3(/ED) solutions ([Al]/[Ti] = 25, [ED]/[Al] = 0.10), allowed to react at 70°C under magnetic stirring (800 rpm) for 30 min, and quenched by rapid cooling to -15°C. After centrifugation (1400 rpm) in a Savant SPD121P centrifugal drying station, the vials were opened and re-positioned in the reaction deck, where the supernatants were removed robotically by aspiration through a needle. The solid phases were washed twice with 3.6 mL aliquots of heptane (applying the same sequence of decantation and supernatant aspiration), once with 3.6 mL of pentane, and finally dried under vacuum at 50°C for 10 h in the aforementioned drying station. Each dry solid phase was then dissolved in 1.00 mL of methanol-d4; 0.60 mL were analyzed by 1H NMR to determine ID and ED contents, whereas 0.40 mL were dried again, mineralized, and analyzed by Inductively Coupled Plasma Optical Emission Spectrometry (ICPOES) for Mg, Ti and Al. Quantitative 1H NMR analyses were performed with a Bruker Avance DRX 400 spectrometer operating at 400 MHz. Acquisition conditions were: 5 mm probe; acquisition time, 3.0 s; relaxation delay, 5.0 s; pulse angle, 90°; spectral width, 10 ppm; 8 transients. Resonance assignment was based on the literature, and preliminary 1H NMR characterizations of the neat donor molecules. Quantitative determinations were based on peak integration against that of an aliquot of acetonitrile added as internal standard (methyl peak at δ = 2.05 ppm downfield of TMS). ICP-OES analyses were carried out using an Agilent 700-series spectrometer, on water solutions of the solid phases treated in sequence with 2.0 mL of concentrated H2SO4, 2.0 mL of concentrated HNO3, and (when needed) 2.0 mL of 30 vol% H2O2 (total time 16 h). The spectrometer was calibrated using commercial standard solutions (metal concentrations in the 1-100 ppm range). 7 ACS Paragon Plus Environment

ACS Catalysis

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 26

RESULTS AND DISCUSSION The catalyst systems The four ZN catalysts selected for this study (Table 2) belong in the three latest generations of commercial relevance (Table 1).2,3 Catalyst C1 (ID = dibutylphthalate) is a widely used fourthgeneration representative.2,3 Catalysts C2 (ID = 2,2-diisobutyl-1,3-dimethoxypropane) and C3 (ID = 2,2-dimethyl-1,3-dimethoxypropane) are members of the fifth generation; the former found industrial application, whereas the latter is poorly stereoselective11,12 but very interesting from a mechanistic standpoint, as we shall see in following sections. Catalyst C4 (ID = 2,3diisopropyldiethylsuccinate) was chosen as an example of the sixth generation, the most recent and also the least described in the scientific literature.2 Alkoxysilane EDs are employed with all three generations (Table 1), even though fifth-generation systems may also be used without.2,3,11 The ED set screened in the present study (Table 3) included dimethoxy-, trimethoxy- and triethoxysilanes bearing linear, branched and cyclic substituents with different steric demand; three of them (namely, ED1-ED3) were selected as ‘minimal structures’ for comparative purposes, whereas the remaining five (ED4-ED8) are applied commercially. Polymerization screening All 132 duplicate pairs of slurry propene homopolymerization experiments were run under the same conditions (T = 70°C, p(C3H6) = 4.5 bar, p(H2) = 0.20 bar, [Al]/[Ti] = 160, t = 30 min), except for the [ED]/[Al] ratio that was varied stepwise ([ED]/[Al] = 0, 0.025, 0.050, 0.10, 0.20). Catalyst deactivation was always negligible, and polymerization kinetics could be simply quantified in terms of average catalyst productivity (Rp, in kg(PP) g(catalyst)-1 h-1). Polymer molecular weight, crystallinity and stereosequence distributions were determined by Rapid-GPC, analytical Crystallization Elution Fractionation (A-CEF), and quantitative 13C NMR spectroscopy, all applied in HTE mode with protocols specifically implemented for use downstream of the PPR48 platform (i.e., applicable to polymer amounts of 0.1-0.2 g, representing the typical yields of PPR48 mini8 ACS Paragon Plus Environment

Page 9 of 26

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

ACS Catalysis

reactors).28 In particular, the Amorphous Fraction (AF) measured by A-CEF was used as a replacement of the Xylene-Soluble (XS) Fraction, difficult to obtain rapidly and reliably in PPR48 scale; the two methods were shown to correlate nicely.28 The concentration of stereodefects in the isotactic fraction, in turn, was obtained from the 13C NMR spectra of raw samples, by measuring the fractional amount of the mmmrrmmm nonad in the methyl region.9,28 For selected PP samples, the mrrm pentad and rrrrrr heptad were also quantified to reveal the presence of stereoblock chains.9,23 The full results of the polymerization screening are reported in Tables S1-S4 (SI). An excerpt for all catalyst systems at [ED]/[Al] = 0.10 is given in Table 4 (A-D) and Figures 1, 2. The three catalyst generations are known to yield polymers with different and characteristic Mw/Mn ranges (Table 1)2,3; our results are in line with that.30 The propensity to undergo ED modification was also idiosyncratic: very high for catalyst C1, moderate for catalyst C4, almost negligible for catalysts C2 and C3. We will comment extensively on the overall results in general, and on the latter aspect in particular, in a subsequent section. Table 4. Experimental results for catalyst C1-C4-AlE3/EDx at [ED]/[Al] = 0.10. A) Catalyst C1 Rp EDx (Kg g-1 h-1) 7.2 None 7.9 6.3 ED1 5.0 5.0 ED2 5.5 5.2 ED3 6.3 8.3 ED4 6.4 7.2 ED5 8.3 4.8 ED6 5.1 6.7 ED7 7.1 8.7 ED8 7.6

Mn (KDa) 17 18 29 29 42 29 30 26 42 32 38 41 25 25 40 39 45 46

Mw (KDa) 103 103 175 175 222 229 180 145 257 284 251 242 198 177 319 313 393 399

Mw/Mn 6.1 5.7 6.0 6.0 5.3 7.8 6.0 5.5 6.1 8.8 6.6 5.9 7.9 7.2 8.0 8.0 8.8 8.7

AF (%) 15.2 11.9 6.3 7.3 3.3 3.5 4.8 4.4 4.3 3.8 4.7 3.6 5.2 5.5 3.2 3.6 3.0 3.0

Tel,max (°C) 113.2 113.0 112.9 113.1 115.8 115.9 115.5 115.6 116.2 116.2 116.3 116.3 116.4 116.4 117.4 117.5 117.9 117.8

[mmmrrmmm] (%) 0.90 0.92 1.12 1.06 0.48 0.49 0.49 0.48 0.41 0.47 0.46 0.40 0.43 0.43 0.30 0.29 0.26 0.29

[mrrm] (%) 2.1 2.0 2.0 1.9 1.0 0.9 0.9 1.0 1.0 1.0 0.8 0.8 1.0 0.9 0.7 0.6 0.6 0.6

[rrrrrr] (%) 1.3 1.3 0.4 0.5 0.3 0.2 0.4 0.3 0.3 0.3 0.3 0.3 0.4 0.4 0.2 0.2 0.2 0.2

9 ACS Paragon Plus Environment

ACS Catalysis

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 26

B) Catalyst C2 EDx None

ED1 ED2 ED3 ED4 ED5 ED6 ED7 ED8

Rp (Kg g-1 h-1) 11.1 10.7 11.2 8.8 7.8 6.5 10.4 8.9 10.1 10.1 9.2 7.1 9.5 9.6 10.6 5.9 10.8 10.0

Mn (KDa) 27 33 32 33 32 32 34 32 35 35 35 26 34 34 29 33 38 33

Mw (KDa) 124 162 157 173 169 164 185 169 157 161 194 144 179 205 166 168 189 180

4.6 4.9 4.9 5.3 5.3 5.2 5.5 5.3 4.5 4.6 5.6 5.6 5.2 6.0 5.8 5.1 5.0 5.5

AF (%) 3.9 4.4 3.5 4.1 2.8 2.4 3.6 3.6 3.7 2.9 3.1 3.4 3.7 4.4 3.6 4.9 3.6 3.8

Tel,max (°C) 114.0 113.7 113.5 113.6 114.3 114.2 114.4 114.1 114.1 114.2 114.1 113.9 113.5 112.8 114.0 113.8 114.6 114.3

[mmmrrmmm] (%) 0.75 0.67 0.66 0.65 0.56 0.55 0.65 0.57 0.52 0.53 0.58 0.55 0.56 0.47 0.55 0.58 0.60 0.53

[mrrm] (%) 1.3 1.3 1.1 1.1 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 0.9 1.0 1.0 1.0 0.9

[rrrrrr] (%) 0.1 0.1 0.2 0.1 0.1 0.1 3 Å). Our results (Figure 3 and Table S5) do not endorse such a discrimination; as a matter of fact, the ID of catalyst C3 turned out to be as extensively and firmly bound to the catalyst surface as that of catalyst C2 (which is also in line with recent DFT calculations21). We will come back to fact [c] at a later stage. Regarding fact [d], our ICP-OES determinations were not informative on the chemical nature of the Al adsorbates. TiCl4 reduction by AlR3 compounds is known to proceed with the formation of AlR3-xClx species (in particular, AlEt2Cl).2-4 In the case of catalyst C1, AlEt2OBu was also formed, as a reduction product of the dibutylphthalate ID (See SI). Last but not least, fact [e] confirms the rather weak chemisorption of TiCl4 onto MgCl216; we verified that the fraction of TiClx lost to the liquid phase had negligible activity in propene polymerization under the conditions of this study, as was demonstrated by means of propene polymerization tests on the filtrates. Donor structure/properties analysis Let us now examine the impact of individual donor structures on catalyst performance. As noted above, the screened EDs modulated very effectively the stereoselectivity of systems C1-AlEt3/EDx and, to a lower extent, of C4-AlEt3/EDx ones (Tables 4-A, 4-D, S1, S4 and Figures 1, 2). It is plausible to trace this finding to the observed ID/ED exchange (Figure 3 and Table S5), that should result into a significant fraction of catalytic species in C1 and C4 with neighboring ED molecules. 14 ACS Paragon Plus Environment

Page 15 of 26

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

ACS Catalysis

We note at this point that the x numeral in the EDx identification codes of the screened alkoxysilanes was assigned ex-post, in such a way that a higher x corresponded to a higher stereoselectivity within the C1-AlEt3/EDx series (Table 4-A). Notably, a very similar ordering turned out to hold for the C4-AlEt3/EDx series too (Table 4-D and Figure 1). In a first approximation, the steric crowding at the Si atom, that is to say next to the surface once the ED molecules get adsorbed, also grows with growing x (Table 3). This correlation is less obvious than it may appear; in fact, it suggests that for all alkoxysilanes in the set similar chemisorption modes ended up with comparable degrees of coverage at saturation (in mol per mol of Mg) for the available surfaces of each given catalyst20,31; the adsorption data of ED1 and ED8 (Figure 3) are compatible with such an assumption. If the hypothesis holds, then the lateral steric pressure experienced by the catalytic species, and hence their stereoselectivity according to the model of Ref23 (see below), should grow with growing alkoxysilane steric demand, and attain a characteristic plateau value for each ED once surface saturation is reached; looking at Tables S1 and S4, this seems indeed to occur around [ED]/[Al] ≈ 0.05. What does not seem to fit in the picture, on the other hand, is that systems C1-AlEt3 and C4-AlEt3 turned out to be slightly more stereoselective than C1-AlEt3/ED1 and C4-AlEt3/ED1 (Tables 4-A, 4-D and Figures 1, 2). Our explanation is that the chemisorbed Al species (Figure 3) surrogated the ED as ID replacements. AlEt3 and AlEt2Cl are strong Lewis acids, known for their self-dimerization equilibria32,33; in the monomeric state, they can form hetero-dinuclear adducts with Al-Cl-Mg and Al-Cl-Ti bridges.2-4 Al-alkyl binding to TiCl3 with formation of doubly-bridged Al-[(µ-Cl)(µ-Et)]Ti moieties is strong34,35, and likely one of the reasons for the low concentration of active Ti measured in Quenched-Flow studies.36,37 As far as binding to MgCl2 is concerned, a recent DFT study concluded that AlEt3-xClx chemisorption (x = 0, 1) on MgCl2(104) facets is exergonic17, which is in line with the results in Figure 3. Based on our polymerization data (Figures 1, 2), the ability of adsorbed Al-alkyls to enhance catalyst stereoselectivity is similar or even slightly higher than that of small EDs (like e.g. ED1), but much poorer than for best-in-class EDs (e.g., ED7 or 15 ACS Paragon Plus Environment

ACS Catalysis

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 26

ED8). It should be noted that the amount of chemisorbed Al on catalysts C1 and C4 was lower in the presence of an alkoxysilane (Figure 3), which we interpret as evidence for a competition; as a matter of fact, according to computational modeling data, alkoxysilanes prevail over AlEt3-xClx species for adsorption on plain MgCl2 crystal terminations.17,20,27 A completely different picture emerged for systems based on catalysts C2 and C3, whose 1,3dimethoxypropane IDs were the dominant donors in the adsorbate pool, leaving very limited room for ED action (Tables 4-B, 4-C, S2, S3, S5, and Figures 1-3). Based on conventional wisdom, such ZN systems should be the easiest to interpret; as a matter of fact, the strong preference of their IDs for chemisorption on MgCl2(110) terminations11,12,31 is expected to determine the least differentiated surface environment (which indeed is consistent with the comparatively narrow molecular weight distribution (MWD) of the produced polymers). Yet, our computational modeling studies indicated that 1,3-dimethoxypropanes on plain MgCl2(110) facets, irrespective of the steric bulk of the alkyl substituents on C-2, cannot get close enough to adjacent TiCl2R catalytic species to make them highly stereoselective in propene insertion. The data in Figure 3 and Table S5, demonstrating unexpectedly that large amounts of Al species were chemisorbed on both catalysts C2 and C3, despite the presence of the ID and with little (catalyst C2) or practically no (catalyst C3) evidence of competition with the ED when used, can provide a solution to this puzzling problem, as will be illustrated in the next section. Improving the computational models of catalytic species Let us now discuss how to interpret the phenomenological picture of the previous sections in terms of suitable models of ZN catalytic species. Different approaches have been reported to quantify the number of such species. Simulations of PP MWDs as summations of Schulz-Flory functions ended up with a minimum of four components38; however, this method cannot discriminate between chemical and physical effects on the MWD, and the possibility of overdetermined solutions is high because the poor resolution and limited precision of MWD data complicate the evaluation of model 16 ACS Paragon Plus Environment

Page 17 of 26

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

ACS Catalysis

significance. A more robust approach, in our opinion, is based on the statistical analysis of highresolution

13

C NMR stereosequence distributions.9,23 This identified three basic types of

stereosequences, namely highly isotactic, weakly isotactic (‘isotactoid’), and syndiotactic9,23; the plausible assumption of a corresponding number of distinct families of catalytic species was translated into the three-site model of Ref23. Here we propose an updated version, assuming that all catalytic species are mononuclear Ti(III) surface adducts with the structure of Figure 4, in line with the indications of recent Raman39 and high-resolution ESR40 studies, and of the latest DFT calculations.16,17 The first coordination sphere of Ti is octahedral and C2-symmetric (like in crystalline TiCl3)9; steric hindrance in the second coordination sphere, on the other hand, can vary. Assuming a Cossee-type chain migratory insertion mechanism5,9, highly isotactic chain propagation requires that the active sites are sterically constrained at two diagonal octants out of the four where the first C-C bond of the growing polymer chain can be located in the 1,2 propene insertion TS, thus locking chain conformation in the desired chiral orientation and ensuring site (pseudo)homotopicity (Figure 4-A).9 Such a condition can be met when adsorbates with adequate bulk (vide infra) occupy the adjacent surface just at the limit still allowing fast monomer access to the Ti center with the favored enantioface, and are under strong lateral pressure by the neighboring co-adsorbate pool, freezing diffusion phenomena41,42 or even hindered conformational motions. Should said steric pressure fade, the enantioselectivity will decrease, because the conformational constraints on the growing chain will weaken, and the chiral active pockets become too loose. Depending on the extent of said fading, and whether only one octant or both octants is/are involved, chain propagation will deteriorate to weakly isotactic (Figure 4-B), or even chain-end-controlled syndiotactic (Figure 4-C).9,23 In case of a dynamic character of the interested surfaces, stereoblock chains may form.9,23 The distribution of the three basic cases of Figure 4 (i.e. close/close, close/open, open/open octants) is a function of the adsorbate pool. In the previous sections, we gave evidence that said pool includes not only donors, but also Al-alkyls. Looking at the recent literature17,27, it appears that models of TiCl4 adsorbates at defective locations of MgCl2(104)-like edges exposing 17 ACS Paragon Plus Environment

ACS Catalysis

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 26

tetracoordinated Mg are not incompatible with the hypothesis of an effective steric modification by adjacent alkoxysilane EDs (even if explicit calculations are still pending).

Figure 4. Updated three-site model for ZN catalysts (see text and Ref.23). Mg and Cl atoms are colored in violet and green, respectively. The large spheres in red highlight the two active Ti sites according to the Cossee insertion mechanism; those in light blue, the surface Mg sites where the presence of an adsorbate would hinder one of the two octants (in light grey) where the first chain CC bond could be located. Chain propagation is predicted to be highly isotactic in case (A); weakly isotactic in case (B); chain-end-controlled syndiotactic in case (C).

A case where the cooperation of Al-alkyls is required, on the other hand, seems that of fifthgeneration catalysts. Figure 5-A shows a computational model (for full details see SI) of a portion of plain MgCl2(110) edge accommodating a TiCl4 unit and an adjacent 2,2-dimethyl-1,3dimethoxypropane molecule in the minimum energy structure; it is evident on inspection, and was confirmed by calculation, that the two co-adsorbates are too far apart to give rise to a catalytic species falling under the case of Figure 4-A. Figure 5-B shows the same fragment with an additional AlEt2Cl molecule chemisorbed in between the two aforementioned adsorbates; the calculated free energy of adsorption was ∆Gads = -7.9 kcal mol-1. Notably, even small alkoxysilane molecules like ED1 were estimated to be too bulky to effectively compete with the AlEt2Cl moiety for chemisorption at that specific surface vacancy. For propene insertion at the catalytic species 18 ACS Paragon Plus Environment

Page 19 of 26

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

ACS Catalysis

formed by alkylation and reduction of the TiCl4 precursors in Figure 5-A and 5-B, we calculated ∆Gre/si ≈ 0 and 1.5 kcal mol-1, respectively; the latter is in good agreement with experiment (Table 4-C). Conformational interlocking of ID and AlEt2Cl enhanced stereorigidity; with bulkier 1,3dimethoxypropanes this can only be more severe (calculations are running). The presence of ED molecules at distal surface locations can also contribute to enforce the necessary lateral steric pressure (Figures 1-3).

Figure 5. (A) DFT model of adjacent TiCl4 and 2,2-diisobutyl-1,3-dimethoxypropane co-adsorption on a MgCl2(110) edge. (B) The same after the adsorption of an AlEt2Cl molecule (see text). Color key: Mg/Violet; Ti/Light grey; Al/Pink; Cl/Green; O/Red; C/Dark grey.

CONCLUSIONS In this paper, we re-visited the polymerization behavior of the three latest generations of industrial MgCl2-supported ZN catalysts for i-PP production with a comprehensive integrated HTE approach. The extensive structure/properties database implemented under rigorously controlled experimental conditions enabled us to analyze and compare catalyst behaviors at an unprecedented level of detail; as a result, we highlighted the roles of all individual catalyst components, and how surface coverage and lateral steric pressure on the active sites can be modulated by means of co-adsorbed donor and Al-alklyl molecules for different surface distributions. A full molecular description of these systems is still difficult to imagine, if only because we are far from sorting out quantitatively the competing 19 ACS Paragon Plus Environment

ACS Catalysis

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

interactions between all components of successful formulations. As long as this is the case, deterministic catalyst design remains unfeasible. On the other hand, the working principles of such formulations are now understood, and searching for novel/better ones is possible with a combination of HTE and statistical modeling tools (e.g. of QSAR type). Concerning in particular the implementation of new donors, the pros and cons of the two viable options (i.e., one single donors able to play the two key functions of surface stabilization and active site modification, vs function-specific ID/ED pairs2,3) can be better evaluated in the light of our new results. The single-donor solution is undoubtedly more elegant, and in principle more simple from the catalyst application standpoint. However, donor design is extremely demanding, because the aforementioned functions call for different (albeit not incompatible) requirements; combining all of them without compromising on ultimate performance is virtually impossible, as is demonstrated by the 1,3-dimethoxypropane case history. Moreover, the approach is difficult to implement in HTE mode, because it includes the stage of precatalyst synthesis, involving a harsh chemistry with several lengthy kinetically controlled steps.2,3 The ID/ED pair solution, in turn, looks more complicated on paper, because it entails two independent searches instead of one, and the optimization of additional process variables (in particular, [Ti]/[ED] and [Al]/[ED] ratios as a function of polymerization temperature). Yet, compared with the single-donor approach, each of the two searches is simpler, because the number of performance criteria to be met by the donor is lower. Moreover, extra-functions can be introduced more easily; as an example, ID design can be aimed to steer the formation of MgCl2(10l)-like or MgCl2(110)-like facets in case chemisorption is favored on one of the two31, with interesting opportunities for e.g. polymer MWD modulation. The ED search, in turn, can be carried out with HTE methodologies28, or even be unnecessary because it is unlikely that a novel ID can lead to a (pre)catalyst with unprecedented MgCl2 surfaces, which implies that existing alkoxysilanes should always represent an option. A superior versatility might be claimed as a further advantage of the ID/ED philosophy; diversifying the performance of one single (pre)catalyst by means of a broad ED portfolio looks appealing in fundamental terms, 20 ACS Paragon Plus Environment

Page 20 of 26

Page 21 of 26

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

ACS Catalysis

although plant engineers who know better about the drawbacks of transitions in huge continuous production lines may have a different opinion. As a final remark, we note that in the present paper we did not consider two very important and closely related properties of ZN systems for i-PP, namely regioselectivity and H2 response.2,3,9,37,43 We have recently implemented HTE protocols for determining such properties for the first time; the results will be presented and discussed in an independent article. AUTHOR INFORMATION Corresponding Authors *E-mail: [email protected], [email protected] Notes The Authors declare no competing financial interest. ASSOCIATED CONTENT Supporting Information. Chemicals specifications. Full results of the polymerization screening. Full results of the catalyst composition screening. Kinetic data for the reaction of dibutylphthalate and 2,3-diisopropyldiethylsuccinate with AlEt3 at 70°C. Computational modeling details. This material is available free of charge via the Internet at http://pubs.acs.org. ACKNOWLEDGMENTS The authors are grateful to Emanuele Breuza and Andrea Correa for the computational modeling results of Figure 5, and Peter H. M. Budzelaar for useful discussions.

21 ACS Paragon Plus Environment

ACS Catalysis

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 22 of 26

REFERENCES (1)

Ali, S., Catal. Rev. 2014, 27, 7-14.

(2)

Cecchin, G.; Morini, G.; Piemontesi, F.; Seidel, A., Kirk-Othmer Encyclopedia of Chemical

Technology. Wiley Interscience: New York, 2007; Vol. 26. (3)

Pasquini, N., Ed., Polypropylene Handbook (2nd Ed.). Hanser Publishers: Munich, Germany,

2005. (4)

Busico, V., In Giulio Natta and the Development of Stereoselective Propene Polymerization

in Polyolefins: 50 years after Ziegler and Natta I.; Kaminsky, W. Ed.; Springer: Heidelberg, 2013; Vol. 257, 37-58. (5)

Arlman, E. J.; Cossee, P. J. Catal. 1964, 3, 99-104.

(6)

Zambelli, A.; Sacchi, M. C.; Locatelli, P.; Zannoni, G. Macromolecules 1982, 15, 211-212.

(7)

Corradini, P.; Barone, V.; Fusco, R.; Guerra, G. Eur. Polym. J. 1979, 15, 1133-1141

(8)

Corradini, P.; Guerra, G.; Fusco, R.; Barone, V. Eur. Polym. J. 1980, 16, 835-842.

(9)

Busico, V.; Cipullo, R. Prog. Polym. Sci. 2001, 26, 443-533.

(10)

Corradini, P.; Barone, V.; Fusco, R.; Guerra, G. Gazz. Chim. Ital. 1983, 113, 601-607.

(11)

Albizzati, E.; Giannini, U.; Morini, G.; Galimberti, M.; Barino, L.; Scordamaglia, R.

Macromol. Symp. 1995, 89, 73-89. (12)

Scordamaglia, R.; Barino, L., Macromol. Theory Simul. 1998, 7 (4), 399-405.

(13)

See, e.g.: Morini, G.; Albizzati, E.; Balbontin, G.; Mingozzi, I.; Sacchi, M. C.; Forlini, F.;

Tritto, I., Macromolecules 1996, 29 (18), 5770-5776. (14)

Seth, M.; Margl, P. M.; Ziegler, T., Macromolecules 2002, 35, 7815-7829

(15)

Boero, M.; Parrinello, M.; Weiss, H.; Hüffer, S., J. Phys. Chem. A 2001, 105, 5096-5105. 22 ACS Paragon Plus Environment

Page 23 of 26

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

ACS Catalysis

(16)

D’Amore, M.; Credendino, R.; Budzelaar, P. H. M.; Causá, M.; Busico, V. J. Catal. 2012,

286, 103-110 (17)

Credendino, R.; Liguori, D.; Fan, Z.; Morini, G.; Cavallo, L. ACS Catal. 2015, 5, 5431-

5435. (18)

Busico, V.; Causà, M.; Cipullo, R.; Credendino, R.; Cutillo, F.; Friederichs, N.; Lamanna,

R.; Segre, A.; Van Axel Castelli, V. J. Phys. Chem. C 2008, 112, 1081-1089 (19)

Credendino, R.; Pater, J. T. M.; Correa, A.; Morini, G.; Cavallo, L. J. Phys. Chem. C 2011,

115, 13322-13328 (20)

Capone, F.; Rongo, L.; D’Amore, M.; Budzelaar, P. H. M.; Busico, V. J. Phys. Chem. C

2013, 117, 24345-24353 (21)

Kuklin, M. S.; Bazhenov, A. S.; Denifl, P.; Leinonen, T.; Linnolahti, M.; Pakkanen, T. A.

Surf. Sci. 2015, 635, 5-10. (22)

D’Amore, M.; Thushara, K. S.; Piovano, A.; Causà, M.; Bordiga, S.; Groppo, E., ACS Catal.

2016, 6, 5786-5796. (23)

Busico, V.; Cipullo, R.; Monaco, G.; Talarico, G.; Vacatello, M.; Chadwick, J. C.; Segre, A.

L.; Sudmeijer, O. Macromolecules 1999, 32, 4173-4182. (24)

Busico, V.; Corradini, P.; De Martino, L.; Proto, A.; Savino, V.; Albizzati, E., Makromol.

Chem. 1985, 186, 1279-1288. (25)

Busico, V.; Corradini, P.; Demartino, L.; Proto, A.; Albizzati, E., Makromol. Chem.,

Macromol. Chem. Phys. 1986, 187, 1115-1124. (26)

Noristi, L.; Barbè, P. C.; Baruzzi, G., Makromol. Chem. 1991, 192, 1115-1127.

(27)

Correa, A.; Credendino, R.; Pater, J. T. M.; Morini, G.; Cavallo, L. Macromolecules 2012,

45, 3695-3701. 23 ACS Paragon Plus Environment

ACS Catalysis

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

(28)

Page 24 of 26

Busico, V.; Cipullo, R.; Mingione, A.; Rongo, L. Ind. Eng. Chem. Res. 2016, 55, 2686-

2695. (29)

WO 2007 134851 to SABIC.

(30)

For third-generation catalysts, see e.g.: (a) Zhou, Q.; Zheng, T.; Li, H.; Li, Q.; Zhang, Y.;

Zhang, L.; Hu, Y. Ind. Eng. Chem. Res. 2014, 53, 17929-17936. (b) Kissin, Y. V.; Zhou, Q.; Li, H.; Zhang, L. J. Catal. 2015, 332, 156-163. (31)

Correa, A.; Piemontesi, F.; Morini, G.; Cavallo, L. Macromolecules 2007, 40, 9181-9189.

(32)

Shreve, A. P.; Muelhaupt, R.; Fultz, W.; Calabrese, J.; Robbins, W.; Ittel, S. D.

Organometallics 1988, 7, 409-416. (33)

Ehm, C.; Antinucci, G.; Budzelaar, P. H. M.; Busico, V. J. Organomet. Chem. 2014, 772-

773, 161-171. (34)

Bahri-Laleh, N.; Correa, A.; Mehdipour-Ataei, S.; Arabi, H.; Haghighi, M. N.; Zohuri, G.;

Cavallo, L. Macromolecules 2011, 44, 778-783 (35)

Kumawat, J.; Gupta, V. K.; Vanka, K. Chemcatchem 2016, 8, 1809-1818.

(36)

Chammingkwan, P.; Thang, V.; Terano, M.; Taniike, T. Top. Catal. 1989, 57, 911-917.

(37)

Yu, Y.; Busico, V.; Budzelaar, P. H. M.; Vittoria, A.; Cipullo, R. Angew. Chem. Int. Ed.

2016, 55, 8590-8594. (38)

Kissin, Y. V.; Chadwick, J. C.; Mingozzi, I.; Morini, G. Macromol. Chem. Phys. 2006, 207,

1344-1350. (39)

Brambilla, L.; Zerbi, G.; Piemontesi, F.; Nascetti, S.; Morini, G. J. Phys. Chem. C 2010,

114, 11475-11484. (40)

Morra, E.; Giamello, E.; Van Doorslaer, S.; Antinucci, G.; D'Amore, M.; Busico, V.;

Chiesa, M. Angew. Chem., Int. Ed. 2015, 54, 4857-4860. 24 ACS Paragon Plus Environment

Page 25 of 26

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

ACS Catalysis

(41)

Credendino, R.; Pater, J. T. M.; Liguori, D.; Morini, G.; Cavallo, L. J. Phys. Chem. C 2012,

116, 22980-22986. (42)

Credendino, R.; Liguori, D.; Morini, G.; Cavallo, L. J. Phys. Chem. C 2014, 118, 8050-

8058. (43)

Chadwick, J. C.; van der Burgt, F.; Rastogi, S.; Busico, V.; Cipullo, R.; Talarico, G.; Heere,

J. J. R. Macromolecules 2004, 37, 9722-9727.

25 ACS Paragon Plus Environment

ACS Catalysis

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 26 of 26

TOC Graphics

26 ACS Paragon Plus Environment