A Catalytic Domino Approach toward Oxo-Alkyl Carbonates and Polycarbonates from CO 2 , Propargylic Alcohols, and (Mono- and Di-)Alcohols

. We have explored the domino reaction between propargylic alcohols, carbon dioxide and various alcohols with the double objective to prepare oxo-alkylcarbonates with a high yield and selectivity under mild conditions and to extend the process to the synthesis of phosgene-free polycarbonates. We first searched for a common catalytic system that was highly selective for the two reactions involved in the domino process, i.e. the cycloaddition of CO 2 to propargylic alcohol to yield  -alkylidene cyclic carbonate (  CC), and the alcoholysis of  CC to furnish the oxo-alkylcarbonate. Kinetics studies monitored by operando IR spectroscopy and supported by 1 H-NMR analyses and DFT modeling have permitted to identify an efficient binary catalytic system composed of a combination of tetrabutylammonium phenolate [TBA][OPh] and silver iodide (AgI) (or copper iodide (CuI)) and to understand its action mode. The [TBA][OPh]/AgI catalytic system (5 mol%) was then successfully implemented for the selective preparation of a range of oxo-alkylcarbonates by the domino reaction with alcohols and propargylic alcohols of different structures. Most of these oxo-alkylcarbonates were produced at a high yield (  97 %) under mild operating conditions, i.e. at 60 °C and 1 bar of CO 2 . The one-pot synthesis of various poly(  -oxocarbonate)s from bis(propargylic alcohol)s, diols and CO 2 was finally investigated and the best operating conditions ([TBA][OPh]/AgI (10 mol%), 60 °C, 15 bar) afforded polycarbonate oligomers with weight-average molar masses of 4,300 g/mol. Although the system should be optimized to produce longer polymer chains, this process offers a new phosgene-free alternative to the synthesis of functional polycarbonates (poly(oxo-carbonate)s) under mild conditions.

1 alcohol to yield -alkylidene cyclic carbonate (CC), and the alcoholysis of CC to furnish the oxo-alkylcarbonate. Kinetics studies monitored by operando IR spectroscopy and supported by 1 H-NMR analyses and DFT modeling have permitted to identify an efficient binary catalytic system composed of a combination of tetrabutylammonium phenolate [TBA][OPh] and silver iodide (AgI) (or copper iodide (CuI)) and to understand its action mode. The [TBA][OPh]/AgI catalytic system (5 mol%) was then successfully implemented for the selective preparation of a range of oxo-alkylcarbonates by the domino reaction with alcohols and propargylic alcohols of different structures. Most of these oxo-alkylcarbonates were produced at a high yield ( 97 %) under mild operating conditions, i.e. at 60 °C and 1 bar of CO 2 . The one-pot synthesis of various poly(-oxocarbonate)s from bis(propargylic alcohol)s, diols and CO 2 was finally investigated and the best operating conditions ([TBA][OPh]/AgI (10 mol%), 60 °C, 15 bar) afforded polycarbonate oligomers with weight-average molar masses of 4,300 g/mol. Although the system should be optimized to produce longer polymer chains, this process offers a new phosgene-free alternative to the synthesis of functional polycarbonates (poly(oxo-carbonate)s) under mild conditions.

INTRODUCTION.
Today, the upsurge in CO 2 utilization as renewable C1 feedstock has opened avenues in modern chemistry for the synthesis of a large diversity of (novel) organic scaffolds (e.g. organic carbonates, 1-5 carbamates 6,7 , oxazolidones, 5,8-11 carboxylic acids, [12][13][14] etc.) and polymers 15 (e.g. polycarbonates, [16][17][18][19][20] polyurethanes, [21][22][23][24] polyesters, 25 polyureas 26,27 ). Among them, -alkylidene 5-membered cyclic carbonates (CCs) are emerging as a novel class of CO 2 -sourced building blocks highlighted by BASF as promising molecules for organic and polymer chemistries. 28 They are typically synthesized by a 100 % atom efficiency carboxylative coupling of CO 2 to propargylic alcohols, [29][30][31][32][33] with some of them that are easily accessible from acetylene and industrial formaldehyde waste. 34 Unlike the conventional 5-membered cyclic carbonates, alkylidene 5-membered cyclic carbonates display a remarkable reactivity towards amines, alcohols and thiols to afford urethanes 35 (oxazolidones or β-oxo-carbamates 36,37 ), β-oxoalkylcarbonates, 38,39 and thiocarbonates or sulfur-containing tetrasubstituted ethylene carbonates, 40,41 respectively. The presence of the exocyclic olefin increases the ring-strain of the molecule and ensures the selective ring-opening by the nucleophiles with the formation of an enol intermediate which tautomerizes into a ketone, thus acting as the driving force for the reaction. The unique chemical features of CCs were exploited by our group to pioneer a novel route to CO 2 -sourced polycarbonates with unprecedented microstructures 40 that already showed relevance as solid electrolytes for Li-ion batteries. 42,43 In contrast to the conventional synthesis of CO 2 -sourced polycarbonates by the direct ring-opening copolymerization of CO 2 with epoxides, [44][45][46][47] our approach focused on the polyaddition of bis(-alkylidene cyclic carbonate)s (bis-CCs) to diols in the presence of an organobase as catalyst to afford regioregular and defects-free poly(β-oxo-carbonate)s at room temperature. Early this year, Schaub engineered a novel synthetic protocol to provide bisCCs and also demonstrated their utility for the fabrication of low molar mass poly(β-oxo-carbonate)s (Mn ~ 2,000 g/mol). 48 However, the synthetic routes to bisCCs are still tedious with demanding purification procedures. To push away these boundaries, we introduce in this work a novel domino terpolymerization approach to produce poly(β-oxo-carbonate)s from a mixture of CO 2 , bispropargylic alcohol and diols (Scheme 1b), some of them being selected for their ease of production from lignin biomass. If many publications focused on the one-pot domino synthesis of oxazolidinones and oxopropyl carbamates from CO 2 , propargylic alcohols and amines [49][50][51][52][53][54] , only few examples reported the preparation of oxoalkyl carbonates by reacting propargylic alcohols, CO 2 and alcohols (Scheme 1a). The challenge lies in the identification of suitable catalysts capable of fixing CO 2 to propargylic alcohols to in-situ form CCs in a selective manner and to promote their alcoholysis.
Song reported on the synthesis of oxoalkyl carbonates with yields of 22-76 % using DBU/Zn salts (40:20 mol%) as catalytic systems at 80 °C, 10 bar in 16 h. 55 Hu obtained yields up to 99 % with AgCl/butylmethylimidazolium acetate ([BMIm][OAc]) binary catalytic systems. However the ionic liquid was added in large amount (100 mol%) to promote the reaction at room temperature and atmospheric CO 2 pressure. 39 He's group used a combination of Ag 2 CO 3 and PPh 3 to catalyze the formation of oxoalkyl carbonates scaffolds with yields of 62-78 %. Working with a 0.5 eq excess in the alcohol was required to obtain higher yields (98 %). 56 Zhang reported high yields of 81-93 % and a large product scope using a silver sulfadiazine/EtNBr (5 mol%) dual system at 1 bar, 80 °C for 24 h. 57,58 This later system suffers from the multi-step synthesis of sulfadiazine that is not a straightforward procedure. Although different catalytic systems exist to provide simple CCs and oxoalkyl carbonates, none of them has been utilized for both the quantitative synthesis of bisCCs and their in-situ transformation into poly(-oxocarbonate)s in a cascade reaction. In addition, as previously exemplified, none of them supported the features needed to respond to the requirements of a step-growth polymerization process that is extremely sensitive to the stoichiometry of the reaction. The bisCCs intermediate produced in-situ by carboxylative cycloaddition of CO 2 to bis(propargylic alcohol) will be directly involved in the polyaddition to the diol. Any side reactions will therefore result in a deviation from the perfect stoichiometry and cause the termination of the domino terpolymerization. To pursue our goal, it 5 is therefore crucial to identify operating conditions and a catalytic system compatible for both steps, i.e. the selective formation of bisCC and its copolymerization with the diol.
In this work, we have developed a novel binary catalyst and evaluated its activity and selectivity on model compounds. Kinetic insights via operando FT-IR spectroscopy, correlated to mechanistic DFT calculations enabled us to understand and rationalize the mode of action of the catalyst. We then used it to selectively synthesize oxo-alkylcarbonates of different structures by the domino reaction between CO 2 , propargylic alcohols and mono-alcohols. Finally this process was exploited for the preparation of PCs under mild operative conditions. Scheme 1. One pot domino synthesis of (a) oxo-alkylcarbonates and (b) poly(β-oxo-carbonate)s

RESULTS AND DISCUSSION
Catalyst design and optimization of the reaction parameters. Model carboxylative coupling reactions of CO 2 with 2-methyl-3-butyn-2-ol to provide CC were first screened (Scheme 2).
We selected tetrabutylammonium phenolate ([TBA][OPh]) as the organocatalyst as we previously demonstrated its high efficiency for the envisioned reaction under moderate conditions. 59 This activity resulted from a good compromise between ion-pair separation controlled by steric effect and the basicity of the anion as evidenced in a previous benchmarking study. 59 Moreover, it is easily prepared from cheap phenol, a bio-based product derived from the fractionation of lignin. 60 Scheme 2. Coupling of CO 2 to 2-methyl-3-butyn-2-ol.
The model reaction was carried out at 50 bar and 80 °C in acetonitrile with a low catalyst loading (5 mol%). Under these conditions, the expected CCs 2 was produced with 90 % yield in only 1 h (Table 1, entry 1). However, the main side reaction observed was the formation of the linear oxo-carbonate 2a by the addition of the propargylic alcohol to CCs. To extend the use of this catalytic system for the targeted domino reaction and prevent this side reaction, an optimization of both the reaction conditions and the catalyst structure was required.  Figure S2). The difference in activity for the reactions carried out in the presence of the different silver salts (Ag 2 CO 3 , AgOAc and AgI) is at this stage difficult to rationalize. However, the different solubility of the silver salts in the reaction medium might be at the origin of this observation. Indeed, when preparing the reaction medium before pressurization with CO 2 , none of these cocatalysts were fully soluble. However, quantifying their solubility in the reaction medium under pressure was not possible. Verpoort 63 and Song 32 .
We then screened some solvents using CuI as cocatalyst and found the reaction to be fastest in acetonitrile and slowest in DMSO (Table 2, entries 7 and 11), with the same selectivity in product 2 (100 %).  Figure S1 for the procedure).
In prelude of our subsequent study on the fabrication of poly(oxo-carbonate)s by the domino reaction, one selected DMSO as the solvent to continue our study. Unlike acetonitrile, its choice was justified by its capability to totally solubilize the bis-propargylic alcohol and provide homogeneous conditions at the initial stage of the reaction. we studied the kinetics of the reaction using online high-pressure FT-IR spectroscopy at 40 °C using both CuI and AgI. Figure 1a shows     Figure S4 for the procedure).
Cascade reaction on model compounds. As [TBA][OPh]/AgI was able to catalyze both the selective formation of DMACC by coupling CO 2 to 2-methyl-3-butyn-2-ol and its alcoholysis by primary and secondary alcohols, we then evaluated its capacity to catalyze the one-pot cascade synthesis of the corresponding oxo-alkylcarbonate 3 from CO 2 , 2-methyl-3-butynol and 1butanol (Table 4). The reactions were carried out at 15 bar for 6h, using 5 mol% of [TBA] [OPh] with and without AgI, and different temperatures were screened. Results are presented in Table   4.  Figure S5 for the procedure).
While there is little to no reaction in the absence of AgI ( However, the main product was CC 2 (93%), with only 7% of the desired oxo-alkylcarbonate 3. By raising the reaction temperature to 60 or 80°C, the yield in 3 was increased to 20 or 74%, respectively (Table 4,  After 20 h of reaction, a mixture of product 2 (45 %) and 3 (40 %) was collected. By decreasing the CO 2 pressure to 5 bar, the formation of CC 2 was faster and its conversion into 3 started after 45 min. After 20 h, the yield in 2 was only 3 % while the desired product 3 was formed with 87 % yield. Importantly, when the pressure was further reduced to 1 bar, CC 2 was rapidly formed and was simultaneously converted into 3. The domino reaction was complete in less than 10 h at 1 bar and selective towards the formation of 3. These kinetics studies clearly highlight that a low CO 2 pressure was required for an efficient domino reaction and for the selective production of 3. It was assumed that high CO 2 pressure favoured the carbonation of 1-butanol, consequently deactivating it for the alcoholysis of the in-situ formed CC 2.  Figure S7-S16.
In all cases, all propargylic alcohols were fully consumed. The corresponding oxoalkylcarbonates were formed with a high yield ( 97 %) when primary alcohols such as 1butanol or benzyl alcohol were involved with the less sterically hindered propargylic alcohol, 2methyl-3-butyn-2-ol. When a secondary 2-propanol or tertiary alcohol 2-methyl-2-butanol was employed together with 2-methyl-3-butyn-2-ol, the yields in oxo-alkylcarbonate dropped to 60 % (product 3C) and 12 % (product 3D), respectively. Similarly, when 1-butanol was reacted with more sterically hindered propargylic alcohols such as 3-methyl-1-pentyn-3-ol and 3,5-dimethyl-  (Table 5). When carried out in DMSO at 1 bar of CO 2 , oligomers with a number average molar mass (M n ) of 1,000 g/mol were collected after 48 h at 60 °C (Table 5, entry 1). By increasing the CO 2 pressure to 15 bar, M n was increased to 1,500 g/mol, however no significant evolution was observed when the reaction time was extended to 72 h (Table 5, entries 2 and 3). When the reaction was performed under neat conditions, similar results were obtained.
Interestingly the molar mass was almost doubled when using DMF as solvent (Table 5, (Table 1), with the consequence that the stoichiometry between -alkylidene cyclic carbonate and alcohol groups is not respected. Second, some thermally induced side reactions (e.g. transcarbonation) might consume reactive groups and thus will provoke a deviation from the perfect stoichiometry, thus lowering the polymer molar mass.
The SEC elugrams obtained on the crude reaction media (in order to avoid any fractionation of the sample during purification) are presented in Figure S18-21. They show that polymers prepared in DMF were of higher molar mass compared to those produced in DMSO or in the bulk. In the two latter cases, dimers and trimers were clearly observed. Dispersities were rather low for a step-growth polymerization process but are the result of the low molar mass polymers.

B1D1
The structure of the poly(-oxo-carbonate) was evidenced by 1 H-NMR spectroscopy (Figure 4a) by

CONCLUSIONS.
In this work, we have disclosed a domino approach to construct oxo-alkylcarbonate scaffolds and
The crude reaction mixture was characterized by 1 H-NMR spectroscopy in CDCl 3 . For the screening of the reaction parameters (T, solvent, catalyst), a similar protocol was applied

Monitoring of the carboxylative coupling of CO 2 to 3-methyl-2-butyn-3-ol by online FT-IR.
In a clean dry reactor of 80 mL equipped with a manometer, a heating mantle, gas inlet/outlets, a mechanical stirrer and a high pressure FT-IR probe were introduced 3-methyl-2-butyn-3-ol (

General procedure for the domino synthesis of butyl-(2-methyl-3-oxobutan-2-yl) carbonate from 2-methyl-3-butyn-2-ol, CO 2 and alcohol.
In a stainless-steel reactor of 12 mL equipped with a magnetic bar, a manometer and a gas inlet/outlet were introduced 3-methyl-2-butyn-3-ol bar. The reaction ran for 6 h after which the reactor was slowly depressurized and placed in a water bath to cool it down to room temperature. The crude reaction mixture was characterized by 1 H-NMR spectroscopy in DMSO-d 6 . Then the catalyst was removed by silica gel chromatography with CH 2 Cl 2 , and the product dried under vacuum at room temperature for 24h.

Monitoring of the synthesis of butyl-(2-methyl-3-oxobutan-2-yl) carbonate by on-line FT-IR.
In a clean dry reactor of 80 mL equipped with a manometer, a heating mantle, gas inlet/outlets, a mechanical stirrer and a high pressure FT-IR probe were introduced 3-methyl-2-butyn-3-ol (
After 30 min, CO 2 was added at a constant pressure of 15 bar. The reaction ran for 24 h, after which the reactor was depressurized and placed in a water bath to cool it down to room temperature. A sample of the reaction mixture was characterized by 1 H-and 13 C-NMR in DMSOd 6 to determine the conversion and by SEC in order to evaluate the molecular parameters of the polymer. Then the catalyst was removed by silica gel chromatography with CH 2 Cl 2 . The solvent was evaporated, and the sample washed 4 times with a mixture of water and CH 2 Cl 2 (5:1) to eliminate DMF. The organic phase was collected, and CH 2 Cl 2 removed under vacuum at room temperature.