amounts of water. The other minor products from the reaction (detected by GC) were diethoxymethane (CH2(OEt)2) formed from the ethanolysis of the substrate, ethyl propanoate (EtC0 2Et) and a trace of ethyl acetate (CH3C0 2Et). The nature of the reaction products was confirmed by GC-MS and / iH NMR. It was shown that the diethoxymethane is formed by a route independent of the catalyst and ethyl propanoate is assumed to be formed by the carbonylation of EtI formed in-situ. The significance of the trace quantities of ethyl acetate in the product mixture will be discussed in chapter four. A typical product yield (wrt initial [CH2I2]) from this reaction carried out under optimised conditions was; DEM (20%), CH2(OEt)2 (0.6%), EtC0 2Et (0.6%). Although the conversion of diiodomethane is low, the selectivity to the malonate ester is very high - ie 97%. It is also important to note here that this is the only reported catalytic double carbonylation system reported that does not require the use of a base - see Chapter One. The effect of bases on this system is discussed in section 2.4.
As will be highlighted in section 2.9 the active catalyst for this reaction only has a short lifetime under the reaction conditions. It has been shown that the standard 4 hour reaction time used for catalytic runs is in excess of the lifetime of the catalytic species and this (along with the other factors discussed in section 2.9) demonstrates a direct relationship between the yield of DEM and the catalyst lifetime. To quantify the lifetime of the catalyst the term catalyst Turn Uver Kumber (TON) is defined as the average number of catalytic cycles the active rhodium species performed before irreversible decomposition. The unstable nature of the catalytic system prevented its examination by kinetic methods. Conclusions that can be drawn from the catalytic
reactions which give evidence pertaining to the mechanism of the title reaction will be summarised in chapter five rather than here.
The catalytic reactions reported here were carried out in a glass liner inside a (250 cm^) stainless steel autoclave, heated to between 50 and 170 °C. The standard conditions were a reaction time of 4 hours at 120 “C under 40 atm CO. The majority of the reactions reported here were carried out using CH2I2 (1 cm^, as substrate) in EtOH (4 cm^) with [Rh2(OAc)4]-2MeOH (ca. 0.01 g) with various amounts of phosphine, the P:Rh ratio of 3:1 being typical. In the result tables it should be assumed that these conditions apply unless otherwise stated.
It should be stated at the outset that the results gained from these catalytic runs were never found to be consistently reproducible. When a series of identical runs were carried out there were obvious modal yield values, but the variation observed under seemingly identical reaction conditions could be very large. Another problem with these data presented is the accuracy of the determination of the final concentrations of CH2I2 and EtI. The quantitative analysis was carried out using a GC using a narrow bore capillary column (BP-1) linked to a FID detector. FID detectors are noted for their insensitivity towards halogenated compounds. Thus the data for the final [CH2I2] and [Etlj for these reactions has a very large margin of error and therefore it will not be presented here.
2 .1 Effects of the P:Rh ratio
Table 2.1.1 shows the product yields from a series of reactions carried out with fixed amounts of the rhodium precursor [Rh2(OAc)4] but with varying amounts of added PEtg. It can be seen that the best yields of DEM are achieved with either a P:Rh ratio of 3:1 or with no phosphine present at all. These results are best viewed graphically and are presented thus in figure 2.1.1.
Table 2.1.1: Showing the effect of the P:Rh ratio upon the reaction. Y ie ld /
R un P:R h DEM C H2(O E t)2 EtC0 2Et TON Selectivity^
31 1:1 4,1 0.6 <0.1 12.9 87% 32 2:1 3,1 0.5 <0.1 9,7 86% 33 3:1 5,9 1.3 <0.1 18.5 82% 34 4:1 3,7 1.5 <0,1 11,6 71% 45 None^ 9,5 2.7 0,9 29.8 78% 51 None^ - 2.4 - - -
a) Calculated from GC peak area, wrt initial [CH2I2]; b) Calculated as ratio of yield of DEM to yield of (DEM + CH2(OEt)2>; c) [Rh2(OAc)4] only, no PEtg; d) PEtg only, no [Rh2(OAc)4],
I
-Due to RhLn(PEtDue to RhL^(CO)^ complex3)2 complex o6 -
4 -
0 1 2 3 4
Equiv of PEtg per Rh Figure 2.1.1: Graph showing yield of DEM vs. amount of PEtg added,
L = non-phosphoms based ligand.
These data show that there are two potential catalytic systems which can operate in this P:Rh ratio region, a rhodium carbonyl species (Run 45) and a rhodium phosphine complex. The synthesis of DEM (—) at low P:Rh is catalysed by a rhodium carbonyl complex(-) whose action is inhibited by PEtg. At higher P:Rh a rhodium phosphine complex(“ ) is the active catalyst thus the second maxima of DEM yield in figure 2.1,1. The dashed lines show how the observed effect of the P:Rh ratio on the DEM yield could potentially be due to the superposition of the action of two different but catalytically active rhodium species.
The reaction conditions for run 45 (No PEtg) are very similar to those used in a report of the synthesis of [Rh(5(CO)i6].^^'^ To assist in understanding this result a separate autoclave run was carried out in the absence of the substrate and the major
organometallic product from this reaction was observed, by IR, MS, and microanalysis, to be [Rh6(CO)i6]. The role of [Rh6(CO)i6] in this process will be discussed in section 2.2.
The work presented in chapter three (section 3.1) showed that the species formed from the precursors [Rh2(OAc)4] / 6-8 PEtg under standard reaction conditions was [Rh(OAc)CO(PEt3)2]. So why is the ideal P:Rh ratio observed as 3:1 and not 2:1? This potential anomaly is explained by the slow reaction that takes place between PEtg and CH2I2 forming the quaternary salt [Et3PCH2l][I] (described in chapter four). Thus, in order to 'assemble' the bis-phosphine species [Rh(OAc)CO(PEt3)2] in the presence of CH2I2 a slight excess of phosphine is required to allow for the removal of some by slow quaternisation by CH2I2. [Et3PCH2l][I] is shown in chapter 4 to be detrimental to the catalyst stability / lifetime.
Since phosphines make excellent spectroscopic handles for NMR studies it was decided to concentrate our study on the potential of rhodium phosphine complexes as catalysts for the title reaction.
2. 2 Studv of other potential catalvsts / catalvst precursors
Having shown that [Rh(OAc)CO(PEt3)2] is the species formed from the catalyst precursors under standard reaction conditions (chapter three, section 3.1), a range of Rh(I) bis-phosphine species and rhodium carbonyls were investigated for their catalytic activity.
Table 2.2.1: Showing the product yields for a variety of catalytic species.
Yield /
%
Run No. Catalyst^ DEM C H2(O E t)2 E tC0 2E t TON
62 A 3.2 1.0 - 10.1
87 A+1 PEt3 2.0 0.3 <0.1 8.0
63 B 3.2 1.0 - 10.1
123 C+1 PEt3 1.0 0 .2 0 .2 3.1
84 D 5.0 0.8 <0.1 15.7
138 E trace trace trace «
51 None^ 2.4 - -
This now shows that [Rh(X)CO(PEt3)2] (X = I, Cl, OAc) are all active catalysts for the double carbonylation of diiodomethane. Note that in runs 87 and 123 the addition of one equivalent of free phosphine inhibits the reaction; A 3:1 P:Rh ratio is only required when the catalyst is made in-situ when the excess phosphine allows the catalyst to 'assemble' in the presence of diiodomethane as discussed above. The above data confirms that rhodium carbonyl species are able to catalyse the reaction. It also suggests that [Rh2(OAc)4] and CH2I2 in the absence of PEtg do not form [Rh6(CO)i6], but instead forms the Monsanto catalyst [Rh(CO)2l2]’ (compare runs 84 and 138).
2 .3 Effects of reaction temperature
A series of reactions using a 6:1 PEt3:[Rh2(OAc)4] ratio (P:Rh = 3:1) with 1 cm^ CH2I2 and 4 cm^ EtOH under 40 atm CO were carried out at a range of temperatures. Figure 2.3.1 below shows the change in the yield of DEM with temperature.
(D
5 -
40 60 80 100 120 140 160 180
Temp / °C Figure 2.3.1: Graph showing DEM yield vs. reaction temperature
No reaction is observed below 90 'C. Above 90 °C the yield of DEM increases with temperature, the optimum reaction temperature range being 120-140 "C. Elevating the temperature above 140 “C presumably decomposes the active [Rh(X)CO(PEt3)2] species.
2 .4 Effects of an added base
The effect of a very wide range of both inorganic bases and N-centred organic bases on the reaction was examined. Without exception all bases were observed to inhibit the double carbonylation reaction. A possible reason for this, linked to the proposed mechanism for this reaction, is discussed in chapter four (section 4.2).
2 .5 Effects of the solvent svstem
It should be noted that the alcohol used in this reaction acts not only as a solvent, but also as a reagent / HI sink. A wide range of alcoholic solvents was employed in this reaction; MeOH, EtOH, i-PrOH and n-BuOH. All these alcohols (ROH) were suitable, with the malonate ester and acetal CH2(OR)2 being observed in all cases. Malonate esters were not observed when either t-BuGH or Me0 CH2CH2 0H were used as solvents. In the case of MeOH, methyl acetate was produced and with EtOH, ethyl propanoate. The activity of [Rh(X)CO(PEt3)2] towards the carbonylation of MeOH will be discussed in section 2.11. Ethyl propanoate is presumed to arise from the carbonylation of EtI. Quantitative data was only obtained for the EtOH based system.
It was reasoned that the addition of a non-protic co-solvent 1:1 v/v with EtOH would reduce the ethoxide concentration and thus inhibit the side reaction which gives rise to the acetal - see figure 2.5.1.