The Effect of Temperature on Selectivity in the Oscillatory Mode of the Phenylacetylene Oxidative Carbonylation Reaction

Abstract Reaction temperature plays a major role in product selectivity in the oscillatory mode of the palladium‐catalyzed phenylacetylene oxidative carbonylation reaction. At 40 °C, dimethyl (2Z)‐2‐phenyl‐2‐butenedioate is the major product whereas at 0 °C the major product is 5,5‐dimethoxy‐3‐phenyl‐2(5H)‐furanone. The occurrence of oscillations in pH coincides with an increase in the rate of phenylacetylene consumption and associated product formation. Experiments were performed isothermally in a reaction calorimeter to correlate reactant consumption and product formation with the occurrence of pH oscillations and the heat released by the reaction. An increase in the size of the pH drop in a single oscillation correlates with an increase in energy, indicating that this section of a single oscillation relates to reactant consumption. Based on these observations, a reaction pathway responsible for product formation is provided.

Reactiont emperature plays am ajor role in product selectivity in the oscillatorym ode of the palladium-catalyzedp henylacetylene oxidativec arbonylation reaction. At 40 8C, dimethyl (2Z)-2-phenyl-2-butenedioate is the major product whereas at 0 8C the major product is 5,5-dimethoxy-3-phenyl-2(5H)-furanone. The occurrence of oscillations in pH coincides with an increase in the rate of phenylacetylene consumption and associated product formation. Experiments werep erformed isothermally in ar eaction calorimeter to correlate reactantc onsumption and product formation with the occurrence of pH oscillations and the heat released by the reaction. An increasei nt he size of the pH drop in as ingle oscillationc orrelates with an increase in energy,i ndicating that this section of as ingle oscillation relates to reactant consumption.B ased on these observations, ar eaction pathway responsible for product formation is provided.
Carbonylation reactions are important in synthetic and industrial chemistry as CÀC-bond-forming reactions that can directly synthesize carbonyl compounds leading to av ariety of products. [1] As the reaction frequently resultsi ns everalp roducts, costly separation post-synthesis is needed. Palladium-catalyzed phenylacetyleneo xidative carbonylation (PCPOC) is an extraordinary reaction as it can proceed in an oscillatory mode. PCPOC was found to oscillate in redoxp otential, pH and the rate of CO/O 2 gas mixture consumption in ac atalytic system (PdI 2 ,KI, O 2 ,NaOAcinm ethanol) at 40 8C. [2] When run in ac alorimeter simultaneous oscillations in pH and the rate of heat evolution (Q r )w ere captured [3] with the total energy release followingastaircasef unction. [3b] The products reported are shown in Figure 1. [3a] Fascinatingly, at 40 8C, the occurrence of pH oscillations was reported to affect product selectivity. [3c] When operating in an oscillatory pH regime, product formation was suppressed until oscillations occurred,f ollowed by selective formation of 4.
When operating in an on-oscillatory pH regime, selectivity was poor with the main products being 1, 2 and 4.F or the same initial conditions, oscillatorya nd non-oscillatory regimes were achieved by changing the amount of PdI 2 catalyst. The influence of reaction temperature on the period and amplitude of pH oscillations during the PCPOCr eaction was also captured. [3d] Isothermal experiments performed over the temperature range 10-50 8Cd emonstrated the existence of oscillations in the range 10-40 8C, with ad ecrease in reaction temperature resultingi na ni ncreasei nt he period and amplitude of the pH oscillations. However,t he associated effect of reducing reaction temperature on product formationd uring the PCPOCr eaction in oscillatory mode wasnot investigated.
The study presented here investigates the effect of the oscillatory mode of the PCPOCr eaction on the dynamics of product formation over the temperature range of 0-40 8C. The experiments were conducted in ar eactionc alorimeter under isothermalc onditions at 0, 10, 20, 30 and 40 8Cw hile monitoring pH, reactionh eat and the dynamics of reactantc onsumption and product formation using the previously published method (see Supporting Information, SI). [3a, 4] Figure 2a-es hows the product and reactant concentration profiles along with the recorded pH ate ach temperature. As ummary of the oscillatory characteristics is given in Ta ble S1 in the SI.
Reproducibility of the oscillations in pH and key characteristics of the pH profiles at each temperature have been reported previously. [3d] Although batch oscillators are difficult to align [a] Dr.J. Parker KGaA. This is an openaccessarticleunder the termsoft he Creative Commons AttributionL icense, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. fully at specific time points, good reproducibility is demonstrated in the overall pH behavior (see SI, FiguresS1-S4). Long lasting pH oscillations withoutf urther addition of substrate are noted in Figure 2: af eature exclusive to this oscillatorys ystem.
The products detected at all temperatures were:D MO, 3; Zisomer, 4 and E-isomer, 5 ( Figure 1). [3a] Small amountso fM eAt, 2,were detected at 40 8Cwhile MeCin, 1,was not detected. Table S1 (SI) show that the period and amplitude of the pH oscillations increase as temperature decreases, in agreement with previous studies at 10-40 8C.
From Figure 3t he conversion of PhAc to products recorded at 0-20 8Cf ollows kinetic laws:c onversionsi ncrease as the temperature increases. This is lessa pparent at 30 8Ca nd is not followed at4 08C, suggesting thermodynamic reaction control. The carbonylation of PhAc has been shown to result in the formation of 4 as the major product at temperatures of 25-80 8C under ap ressurized atmosphere containing CO and air althoughn om ention is made of whether the reactionw as conductedi na no scillatory mode. [1c] We found that, in oscillatory mode, the major product depends on the reaction temperature. As temperature increases, formation of 4 increases, while at lower temperatures the formation of 3 is favored with 3 being the major product at 0a nd 10 8C. The ratio, r,o f3 and 4 remains constant duringthe reaction ( Figure 4a)and decreases with increasing temperature. This suggestst he activation energies of the two pathways leadingt of ormation of these two products differ,r esulting in reactiont emperature dictating which process will dominate, kinetically or thermodynamically controlled. [5] Formation of 3 is the faster reaction, that is, under kinetic control, and dominates at lower temperatures whilst 4 is the more thermodynamically stable product (i.e. under thermodynamic control) and dominates at higher temperatures. Figure 4b shows the ratio of 5 and 4 at 0-20 8Ci ss imilar and slightly higher than that at 30 and 40 8Co nce appreciable amountso fb oth isomersa re formed suggestingt heir formation follows ar elated reaction pathway.  The lengthy experiments incurred evaporative loss of solvent affecting the accuracyo fm easurements and making it challenging to capturet rends in energy release for the full duration of the experiments. Thisp articularly affected the run at 40 8Cw here methanol wasa dded regularly to compensate for this loss. In addition, at 40 8Ct he perioda nd amplitude of oscillationsw as very smallr eflecting small energy changes which are difficult to measurea ccurately.M oreover,r educing the reaction temperature slows down the rate of the reaction leading to ac orresponding reduction in the rate at which heat is produced. Thus,i tw as only possible to observe the energy release profile over al arge section of the reaction at 30 8C ( Figure 5). Figure 5a shows heaterp ower directly correlates with the pH oscillations. As ar esult, the heat the reactionp ro-   ChemPhysChem 2017ChemPhysChem , 18,1981ChemPhysChem -1986 www.chemphyschem.org 2017 The Authors. Publishedb yWiley-VCH Verlag GmbH &Co. KGaA, Weinheim duces (Q r )i sr eleased in pulses, causing the total energyr elease to increase stepwise ( Figure 5b-d).
Closer examination of proximate oscillations at 30 8Cs hows that the amount of energy released increases as the size of the pH drop increases (Figure 5c). Figures5ca nd 5d show the large drop in pH at approximately 2910 min is accompanied by ar elease of 0.7 kJ of energy compared to the 0.3 kJ of energy released during the previous and subsequent oscillations. The stepwise release of energy that occurs at 30 8Ci s also evident at 0, 10, 20, and 40 8Cw hen shorter sections of the reactiona re examined, verifying the relationship between changes in pH and energy release (Figure 6a-d). It can be con-cluded that the section of as ingle oscillation when the pH drops relatestoreactantconsumption inducing energy release. This accelerated pH fall is therefore associated with the acceleration in reactionr ate and consequently product formation.
An umber of mechanismsf or palladium-catalyzed carbonylation of alkynes have been proposed. [1c, 2, 6] Combined with the observations in this study,areactionp athway that can explain the products detected under the reaction conditions described here is given in Scheme 1.
KI and PdI 2 react to form K 2 PdI 4 which then goes on to react with methanola nd CO to form I 3 PdCOOCH 3 and HI. [1c, 6b] This is supported by experiments which confirm ad ropi np Ho ccurs Scheme1.Reactionpathway. ChemPhysChem 2017ChemPhysChem , 18,1981ChemPhysChem -1986 www.chemphyschem.org 2017 The Authors. Publishedb yWiley-VCH Verlag GmbH &Co. KGaA, Weinheim when PdI 2 ,K Ia nd methanol are purged with CO indicating the formationo fH I. [7] The next stage is the insertion of phenylacetylene into the PdÀCb ond in I 3 PdCOOCH 3 . [1c, 6b] Depending on the orientation of the PhAc insertion this could lead to either product 1 or 2. This is supported by the detection of 2 in this study at 40 8Ca nd 1 in previouss tudies,a lthough in trace amounts. [8] In previous studies 1 appeared to behave as an intermediate, being produced and consumed. However, subsequente xperiments in which 1 was used as the starting materiali nsteado fP hAc failed to produce the diester products indicating that 1 is not ar eaction intermediate. [8] As only trace levels of 2 wered etected in this study at 40 8Ci ts uggests that monoester formation under these conditions is unfavorable.A s the major products of this reaction are the diester products, the reactiono fI 3 Pd(H)C=C(Ph)COOCH 3 with furtherC Oa nd MeOH is required. This reactiono ccurs with the releaseo fH I which is supported by the observed drop in pH and associated energy pulse that occurs within each pH oscillation. This second reaction with CO and MeOH results in 2i someric intermediates which produce 4 and 5 with the release of Pd 0 and I À .T he palladium is then recycled reforming [PdI 4 ] 2À and consuming HI in the process. [9] This is borne out by the observations of oscillations in turbidity which have previously been reported for this system. [4] The isomeric intermediate leading to 4 is able to undergo ac yclisation process which ultimately results in the formation of 3 accompanied by the release of Pd 0 and I À . [1c, 6b] In conclusion, this study has shown that reaction temperature can be used to tune selectivity of the PCPOCr eaction when operated in oscillatory mode. Under otherwise the same conditions, at 0 8C 3 is the major product whilea t4 0 8Ct he major product is 4.T he stepwise releaseo fe nergy that occurs during the reaction has been shownt oc orrelatew ith the pH fall within each oscillation. Ap lausible reaction pathway explainingt he observedp roduct distribution has been given.