Cent. Eur. J. Chem. • 11(11) • 2013 • 1711-1722 DOI: 10.2478/s11532-013-0311-7

Central European Journal of Chemistry

A theoretical study of the limits of the acidity of carbon acids in phase transfer catalysis in water and in liquid ammonia#

Invited Paper Ibon Alkorta*, José Elguero, Roger Gallo†

Institute of Medical Chemistry, CSIC, E-28006 Madrid, Spain

Received 9 May 2013; Accepted 6 July 2013

Abstract: The acidities of a large number of carbon acids have been theoretically calculated for the gas-phase and for DMSO solution. The gas-phase values, both DH and DG, are very well correlated with the available experimental data. From the calculated DG values in

DMSO and the pKas in the same solvent, a homogeneous set of pKa (DMSO) values was devised that was used to generate pKa (water).

These last pKas were used to establish the limits of the acidity of carbon acids for reactions under PTC conditions both alkylations and

H/D exchange. A step further led to the pKas in liquid ammonia and from them to the virtual use of PTC using liquid ammonia instead of water. Keywords: PTC • Liquid ammonia • Carbon acids • pKa • DFT calculations © Versita Sp. z o.o.

by their pK s. We have selected the C-alkylation 1. Introduction a reactions but we will also discuss another class of Few synthetic procedures have had a greater impact in reactions (deuteration, isomerization and oxidation) that preparative chemistry than[ Phase Transfer Catalysis occur at higher pKa values (less acid) than the reactions (PTC) [1-4]. Let us remind that the most used anionic of C-substitution. For this purpose, we have used nucleophiles in PTC conditions are those generated by extensively the discussion reported in Chapter 8 of the deprotonation of organic compounds having an acidic book of Starks, Liotta and Halpern [1]. C-H bond (carbon acids) because the reaction with The paper is organized in five sections: electrophilic carbon derivatives is used to build up the 1. DFT calculations of the gas-phase acidity (ΔG) will carbon skeleton of many organic molecules. be carried out on a set of 18 molecules whose pKas in Today, PTC together with ultra sounds and water and in DMSO are known (thanks to the work of microwaves activation, constitute three of the most useful Bordwell about 1200 compounds were measured in this approaches in the chemistry arsenal. PTC owes much to solvent while values in water are much rare) [20]. Mąkosza contributions that cover from 1965 [5-7] till very 2. From these comparisons, we will estimate the recently [8,9]. Although many modifications have been pKas in water of all the molecules needed in the following introduced to PTC, except for solvent-free PTC (with or section. without MW irradiation) [10,11], water is used in all other 3. From literature results we will establish the limits variants, triphasic [12-14], supercritical CO2 [15-18], of PTC in water in function of the pKas of the substrates fluorocarbons [19], etc. (carbon acids).

The present work deals with a computational study 4. Discuss the pKas in liquid ammonia based on the of the possibility to replace water by liquid ammonia. We works of Lagowski [21] and Page et al. [22]. will examine the limits of generating carbanions in water 5. Examine the possibility to carry PTC in liquid and its related topic of carbon acids acidity as measured ammonia.

* E-mail: [email protected] #This paper is dedicated to Prof. Mieczysław Mąkosza the discoverer of PTC 1711 dress: 736 La Salle, Impasse Jean de la Fontaine, 13320 Bouc Bel Air, France A theoretical study of the limits of the acidity of carbon acids in phase transfer catalysis in water and in liquid ammonia

Figure 1. The 27 molecules discussed in this part of the manuscript.

2. Computational details The agreement between experimental and calculated gas-phase data is good enough to identify the site of

The calculated values of ΔH and ΔG correspond deprotonation of 1: the CH2 not the CH3. The experimental to B3LYP/6-311++G(d,p) [23-26] and PCM [27-29] values are all from the NIST data-base [31]. (DMSO)/B3LYP/6-311++G(d,p) including DIS (Solute- From Table 1 the following equations are obtained solvent dispersion interaction energy), REP (Solute- (the RMS residuals are also given): solvent dispersion repulsion energy) and CAV (Solute cavitation energy) calculations, respectively. All the ΔG calc. = –(3.4±8.9) + (1.003±0.006) calculations were done using the Gaussian 09 package ΔH calc., n = 37, R2 = 0.999, RMS = 2.9 (1) [30]. ΔH calc. = –(133±54) + (1.08±0.04) 2 3. Results and discussion ΔH exp., n = 23, R = 0.977, RMS = 12.8 (2) ΔG calc. = –(123±43) + (1.09±0.34) 2 3.1. Comparison of pKas in water and DMSO ΔG exp., n = 23, R = 0.986, RMS = 10.3 (3) with calculated acidities in DMSO (ΔG) The 27 molecules chosen for discussion of the Note that our values are the differences in energy boundaries of PTC in water are represented in Fig. 1. between the neutral and the protonated forms without Seventeen supplementary molecules used to establish any correction for the H+ difference. The purpose was linear relationships are reported in Fig. 2. to verify that our computed values represent adequately The numerical data are reported in Table 1 (gas the experimental values, i.e., that they are proportional. phase) and Table 2 (water and DMSO solutions). The Many authors have described that the relationships calculated values (see Computational details) of ΔH between experimental and calculated thermodynamic and ΔG correspond to B3LYP/6-311++G(d,p) and quantities related to the present question, for instance, PCM(DMSO)/B3LYP/6-311++G(d,p), respectively. carbanions, are generally excellent [32-35].

1712 I. Alkorta, J. Elguero, R. Gallo

Figure 2. The 17 supplementary molecules used for establishing linear relationships.

Combining the data of Tables 2 and 3 leads to the following equations

pKa DMSO = (–172±8) + (0.154±0.007) ΔG DMSO, n = 36, R2 = 0.950, RMS = 2.4 (5)

Fitted values of Eq. 5 (fourth column) have been used to calculate Eq. 6:

pK water = (–4.2±1.3) + (1.01±0.06) a 2 pKa DMSO, n = 25, R = 0.929, RMS = 3.0 (6)

The last column of Table 3 contains the fitted values of Eq. 6 and the last seven values of column three (no calculations in the case of 4, 5, 7, 9, 10, 14 and 24).

These fitted values are not very different from the pKa values measured in water (see Fig. 3), but they are more robust having been calculated using the very

Figure 3. Plot of the experimental values of pKa in water against the reliable gas-phase values. fitted values of the last column of Table 3. Deviations larger than 4 pKa units correspond to

There is a Table in reference [47] with several pKas compounds 32 (malonodinitrile, +5.7), 27 (toluene, in DMSO that, when common, are very similar to those +5.2), 16 (indene, +4.7) and 31 (2-nitropropane, –6.0). of Table 2. There is a lot of confusion in the secondary Compound 33 (methylmalono-dinitrile) also deviates like literature about the solvent used to determine the pKas, 32 but in this case the experimental value corresponds often the values are given without indicating the solvent. to benzyl-malonodinitrile. Thus, for instance, Starks, Liotta and Halpern [1] do not The two malonodinitriles 32 and 33 can be related explicitly write that their pKa values are in DMSO, but to the following comment by Pagani et al. [48]: “We are they are identical to those of Bordwell. Eq. 4 is obtained also convinced that cyanocarbanions are reluctant to from the data of Table 2: hydrogen bond. Proof for this is provided by the almost identical pK values of malonodinitrile in water (pK = a a pKa water = (–3.2±1.1) + (0.96±0.05) 11.41) [49] and in DMS0 (pKa= 11.0) [20], a behavior 2 pKa DMSO, n = 25, R = 0.941, RMS = 2.7 (4) which is dramatically different from that of comparably

1713 A theoretical study of the limits of the acidity of carbon acids in phase transfer catalysis in water and in liquid ammonia

Table 1. Calculated and experimental values in the gas phase (all in kJ mol–1).

No Compound ΔH calc. ΔG calc ΔH exp. ΔG exp.

1 Methylbenzylketone (CH2) 1444.2 1449.8 1465.0 1441.0

2 Acetylacetone 1407.6 1406.0 1438.0 1409.0

3 Acetophenone 1507.4 1508.4 1512.0 1487.0

8 Diethyl malonate 1440.1 1441.7 1442.0 1432.0

11 Phenylacetonitrile 1442.4 1447.7 1467.0 1443.0

12 Acetonitrile 1546.9 1544.4 1560.0 1536.0

13 Dimethylsulfoxide 1529.6 1528.7 1563.0 1536.0

15 Cyclopentadiene 1473.7 1472.5 1481.0 1455.0

16 Indene 1464.2 1464.6 1472.0 1451.0

17 Fluorene 1463.0 1461.8 1472.0 1439.0

18 Diphenylmethane 1504.9 1511.1 1512.0 1499.0

19 Dihydroanthracene 1496.7 1495.6 • •

20 Xanthene 1493.4 1491.8 • •

21 Allylbenzene 1503.4 1506.6 • •

22 Thiazole 1497.8 1493.6 • •

23 Thiophene 1600.4 1598.8 1595.0 1561.0

25 Triphenylmethane 1477.6 1482.8 1501.0 1476.0

26 Benzofurane 1585.1 1584.9 • •

27 Toluene 1587.7 1592.0 1577.0 1564.0

28 3-Nitroprop-1-ene 1403.5 1407.7 • •

29 Nitromethane 1472.8 1477.2 1491.0 1467.0

30 Nitroethane 1471.4 1474.5 1489.0 1469.0

31 2-Nitropropane 1473.2 1470.4 1490.0 1466.0

32 Malonodinitrile 1373.3 1372.5 1405.0 1376.0

33 2-Methylpropanedinitrile 1386.2 1383.6 • •

34 Bis(ethylsulfonyl)methane 1401.4 1400.7 • •

35 Bis(ethylsulfonyl)ethane 1419.9 1419.7 • •

36 Dimedone 1369.6 1371.4 1418.0 1385.0

37 Bis(methylsulfonyl)methane 1398.5 1397.2 • •

38 Ethyl acetate 1549.0 1550.0 1543.0 1527.0

39 Acetone 1534.2 1536.3 1544.0 1516.0

40 3-Ethanoylpentane-2,4-dione 1384.0 1386.5 • •

41 Meldrum’s acid 1369.1 1371.3 1389.0 1359.0

42 1,3-Cyclohexanedione 1371.0 1373.9 • •

43 Propanediamide 1446.0 1450.1 • •

44 Methane 1737.2 1741.3 1749.0 1715.0

1714 I. Alkorta, J. Elguero, R. Gallo

Table 2. Experimental pKa values in water and DMSO solution.

Number and name pKa water Ref. pKa DMSO Ref.

1 Methylbenzylketone 15.9 [36] 19.8 [37] 2 Acetylacetone 9.0 [20] 13.3 [20] 3 Acetophenone 18.7 [20] 22.2 [20] 8 Diethyl malonate 13.3 [34] 16.4 [20] 11 Phenylacetonitrile 21.9 [20] 12 Acetonitrile 28.9 [22] 31.3 [20] 13 DMSO 25.0 [34] 31.1 [20] 15 Cyclopentadiene 15.0 [34] 18.0 [20] 16 Indene 22.6 [20] 20.1 [20] 17 Fluorene 22.6 [20] 18 Diphenylmethane 32.2 [20] 19 Dihydroanthracene 27.0 [20] 20 Xanthene 30.0 [20] 21 Allylbenzene 34.0 [20] 22 Thiazole 29.4 [20] 23 Thiophene 38.4 [38] 25 Triphenylmethane 30.6 [20] 26 Benzofuran 32.7a [20,39,40] 27 Toluene 41.0 [34] 43.0 [20] 28 Nitropropene 5.2 [38] 11.1 [41] 29 Nitromethane 10.2 [34] 17.2 [20] 30 Nitroethane 8.6 [34] 16.7 [37] 31 2-Nitropropane 7.7 [34] 16.9 [20] 32 Malonodinitrile 11.4 [38] 11.1 [20] 33 Methylmalonodinitrile 12.8 [20] 34 Bis(ethylsulfonyl)methane 12.2 [38] 14.4 [37] 35 Bis(ethylsulfonyl)ethane 14.6 [38] 16.7 [37] 36 Dimedone 5.3 [22] 11.2 [20]

37 CH2(SO2Me)2 12.7 [20] 15.0 [20] 38 Ethyl acetate 25.6 [20] 29.5 [22] 39 Acetone 19.3 [20] 26.5 [20] 40 3-Ethanoylpentane-2,4-dione 5.8 [20] 8.6 [20] 41 Meldrum’s acid 4.8 [42] 7.3 [20] 42 1,3-Cyclohexanedione 5.2 [42] 10.3 [20] 43 Propanediamide 12.5 [42] 18.0 [20] 44 Methane 48.0 [42] 56.0 [20]

b R-CH2-CHO 4 20 [46]

(CH3)2CHCHO 5 15.5, 15.7 [43,44]

RCH2CO2R’ 7 24-25 [1]

c ZCH2N=CHPh =CPh2 9,10 20.0 [45]

RCH2SO2Me 14 16-23 [1] 1-Methylimidazole 24 33.1d [40] aThe value 32.7 for benzofuran is given both for water and for DMSO. On page 439 in [1], a value of 36.8 is given; b Acetaldehyde, pKa = 16.7 [44]. cRange 17.2-24.3 [45]. d34.1 [46]

1715 A theoretical study of the limits of the acidity of carbon acids in phase transfer catalysis in water and in liquid ammonia

–1 Table 3. Calculated values of ΔG in DMSO (kJ mol ) and fitted and predicted pKa values in water and DMSO solution.

Comp. ΔG calc pKa water from pKa DMSO from pKa water from pKa DMSO

DMSO pKa DMSO ΔG calc. DMSO precedent

1 1244.0 15.7 18.9 14.8 2 1200.3 9.4 12.2 8.1 3 1289.7 18.0 25.9 21.9 8 1234.6 12.4 17.4 13.4 11 1256.7 • 20.8 16.8 12 1312.0 26.7 29.3 25.3 13 1324.8 26.5 31.3 27.3 15 1244.2 13.9 18.9 14.8 16 1264.4 15.9 22.0 18.0 17 1281.6 • 24.6 20.6 18 1320.2 • 30.6 26.6 19 1315.1 • 29.1 25.8 20 1320.2 • 30.6 26.6 21 1310.9 • 29.1 25.2 22 1288.6 • 25.7 21.7 23 1367.0 • 37.7 33.8 25 1354.7 • 35.9 31.9 26 1364.2 • 37.3 33.4 27 1379.8 37.9 39.7 35.8 28 1187.7 7.3 10.2 6.1 29 1234.4 13.2 17.4 13.3 30 1231.3 12.7 16.9 12.8 31 1238.7 12.9 18.0 14.0 32 1185.3 7.3 9.8 5.7 33 1196.8 • 11.6 7.5 34 1221.2 10.5 15.4 11.3 35 1248.1 12.7 19.5 15.4 36 1171.1 7.4 7.7 3.5 37 1224.2 11.1 15.8 11.8 38 1317.9 24.9 30.2 26.2 39 1296.0 22.1 26.8 22.9 40 1184.0 4.9 9.6 5.5 41 1179.7 3.7 9.0 4.9 42 1169.1 6.6 7.4 3.2 43 1242.9 13.9 18.7 14.6 44 1457.1 50.3 51.6 47.8 Model 33 1196.8 • 11.6 7.5 4 • 15.7 • 15.7 5 • 15.6 • 15.6 7 • 20.3 • 20.3 9, 10 • 15.7 • 15.7 14 • 18.5 • 18.5 24 • 27.9 • 27.9

1716 I. Alkorta, J. Elguero, R. Gallo

acidic carbon acids possessing hydrogen bonding is about 16. These compounds are monoalkylated in groups (carbonyl and nitro)”. Other authors have also good yields using reaction times of 12-24 h at room discussed the case of the cyano groups [50,51], therefore temperature [1]. we should be prudent when using such compounds. Other authors have reported comparisons between 3.2.1.5. Nitriles pKas determined in two different solvents, particularly The C-alkylation of phenylacetonitrile (PAN 11, pKa = water, DMSO and THF, but these relationships cover 16.8) was published by Mąkosza in 1965 [7,8]. It is the less and more homogenous compounds [46]. A much prototypical C-alkylation reaction in PTC; it has been larger study, but containing a limited number of carbon used to determine the chemical and physical parameters acids, has been carried out by Pliego et al. [52]. Charif of PTC. In these papers, a kinetic study was reported –1 results [42] were analyzed using B3LYP/6-311++G(d,p) providing an Ea of 84 kJ mol , a value we will use in a and Gaussian-4 (G4) methods by Rayne and Forest [53]. further discussion.

The pKa of acetonitrile 12, the compound that cannot be alkylated, is approximately 25 [34]. However, under 3.2. The limits of carbanion pKas In Fig. 1 the molecules we need for our discussion of the the proper PTC/OH conditions it should be possible to problem of carbon C-H bond acidity have been reported. deprotonate an aliphatic CH bond α to a nitrile. A patent

The water pKas used in this section are the fitted values described the cyclization of X(CH2)nCN (X = Cl, Br) to the of the last column of Table 3 save for a few compounds cycloalkanecarbonitrile. A yield of 99.3% was obtained that having a general formula cannot be calculated or under solid-liquid conditions [1]. Thus, there is no doubt that are not useful (4, 5, 7, 9, 10, 14 and 24). that a hydrogen atom α to a nitrile (without any adjacent withdrawing or aromatic substituent) can be abstracted 3.2.1. C-Alkylation in PTC conditions. If the resulting carbanions could It is generally assumed that the compounds having an be alkylated intermolecularly this would be a result acidic C–H bond with a pKa ~23 can be deprotonated in reported, which is not the case. That the intramolecular the classical conditions of PTC/OH [1]. reaction above reported was possible is probably due to an entropic factor that favors the cyclization (there are 3.2.1.1. Ketones many examples of reaction rate acceleration when intra Alkylation of phenylacetone (1, methylbenzylketone, and inter reactions are compared). pKa = 14.8) and acetylacetone (2, pKa = 8.0) take place This example is of great importance because easily at temperatures ≤ 30°C in good yields. The it shows that the upper limit, was not reached, for problem of the C/O regioselectivity will not be discussed intermolecular C-alkylations corresponds to pKa ≈ 25. here. Acetophenone (3, pKa = 21.9) has been alkylated All carbon acids with pKas lower than 25 could be with allyl chloride (an alkylation reagent activated C-alkylated. compared with alkyl chlorides); reaction conditions were not reported [1]. 3.2.1.6. Sulfones Activated sulfones like benzylsulfones and

3.2.1.2. Aldehydes phenylsulfones with pKas in the 18-27 range [1] (like 14,

The CH group α to an aldehyde 4 has a pKa ≈ 16. pKa = 18.5) have been alkylated with different reagents

Isobutyraldehyde (5, a hindered aldehyde, pKa = 15.6) and different PTC conditions. reacts at 70 ºC with benzyl chloride (an activated chloride) to afford 96% of 6 [1]. 3.2.1.7. Hydrocarbons

Cyclopentadiene (15, pKa = 14.8) and indene (16, pKa

3.2.1.3. Esters = 17.9) are easily alkylated as well as fluorene (17, pKa

The CH bonds α to an ester 7 having pKas ≈ 20 can = 20.6). Fluorene was alkylated by Mąkosza at 70-90ºC be alkylated. For a malonic diester like diethylmalonate [56], this result is to be compared to fluorene deuteration

(8, pKa = 13.3) the reaction is much more easier. The at room temperature (80% deuteration in 6 min) [1]. This competition of alkylation vs. hydrolysis and the choice difference of reactivity allows to estimate the difference of the adequate parameters to control it will not be of rate constants from the same carbanion. Assuming –1 discussed. an Ea of about 84 kJ mol , the rate constant is multiplied by 2.87 for each 10ºC increase. On going from 20 to 3.2.1.4. Imines 90ºC, the ratio of rate constants between deuteration

The range of pKas of activated imines (PhCH=NCH2Z 9 or and alkylation is approximately 600; besides, the

Ph2C=NCH2Z 10; Z = electron-withdrawing substituent) alkylation was carried out in 6 h instead of 6 min, thus

1717 A theoretical study of the limits of the acidity of carbon acids in phase transfer catalysis in water and in liquid ammonia

Figure 4. The water pKa limits for the different reactions.

the ratio of rate constants should be ~36,000. This with an alkylating agent (pKa < 25). It is probably a

explains why a compound with a pKa ≥ 25 that cannot kinetic effect related with the activation energy being be alkylated can be deuterated. But where is the limit? A larger for a C-alkylation (~ 85 kJ mol–1) compared partial answer to this question is gathered from the fact with a deuteration or an isomerization (faster than

that diphenylmethane (18, pKa = 26.6) has been easily the deuteration). An educated guess would be that deuterated and oxidized. The next section will complete deuteration, isomerization and also oxidation in PTC –1 this aspect. have an Ea about 40-65 kJ mol . We have summarized the previous discussion in 3.2.2. Deuteration, isomerization and oxidation Fig. 4.

Indene (16, pKa = 17.9) and fluorene (17, pKa = 20.6) According to M. Mąkosza (probably using DMSO

have been easily di- and tri-deuterated. Fluorene, values in most cases) [55], concerning pKa limitation of

dihydroanthracene (19, pKa = 25.8) and xanthene (20, CH acid for PTC C-alkylation, a simple answer is not

pKa = 26.6) have been deprotonated and oxidized easy, because besides the CH acidity of the carbanion into phenones in PTC conditions in the presence of precursor there are a few factors affecting effectiveness

oxygen. Allylbenzene (21, pKa = 25.2) is deprotonated of PTC alkylation. The simplest answer is - fluorene (17,

in PTC. The isomerization seems to be faster than the pKa = 23) still can be alkylated in the liquid-liquid system deuteration (intra vs. inter) [1]. with 50% aqueous NaOH [56]. Perhaps somewhat A detailed study was carried out by Spillane et al. [54] weaker CH acids can be alkylated in solid-liquid system

on deuteration of heterocycles (mostly thiazoles but also (solid KOH + K2CO3). PTC reactions of carbanions with

pyridines, thiophenes and imidazoles); positions and more active electrophiles (CCl4, PhSCN, aldehydes, percentages of deuteration as well as some kinetic data etc.) and particularly isotope exchange are feasible

were reported but without relationship with pKa data. when the precursors are less acidic. For instance

The less acidic parent compound that was deuterated aliphatic nitriles and sulfones (pKa 31 and 29) can be

is thiazole (22, pKa = 21.7). 2-Methylthiophene was deprotonated and react with aldehydes or undergo

not deuterated (thiophene 23 has a pKa = 33.8) nor intramolecular alkylation under PTC conditions [57,58].

1-methylimidazole 24 (pKa = 27.9; 1H-imidazole was not Using Eq. 6, 31 and 29 pKa units in DMSO became

deuterated because in PTC conditions this compound 26 and 24 pKa units in water, with are compatible with exists as the imidazolate anion). On the other hand Fig. 2. 2-nitrothiophene was deuterated at position 5 (a to the sulfur). 3.3. Differences in basicity in water and in Amongst the less acid compounds that have been liquid ammonia

deuterated are diphenylmethane (18, pKa = 26.6) and Already in 1955, Hall. Piccolini and Roberts reported

triphenylmethane (25, pKa = 31.9). According to Starks, H/D exchanges of aromatic compounds in liquid Liotta and Halpern the less acid compound deuterated ammonia [59]. The exchange rate of deuterobenzene

is thiophene (23, pKa = 33.8) [1]. Benzofuran (26, pKa = and o-deuterotoluene was too slow for convenient 33.4) has also been deuterated. The upper limit appears measurements but it took place. From the data of a

to be toluene (27, pKa = 35.8) that cannot be deuterated. series of papers of Lagowski et al. and Page et al. we

In conclusion, the borderline is between 33 and 35 pKa have built up Table 4. units. Table 4 values led to Eqs. 7 and 8.

The main open question is the origin of the pKas difference between that necessary to form a carbanion pK L NH = (–3.7±0.8) + (0.94±0.06) a 3 2 in PTC (pKa ≈ 34) and that necessary to make it react pKa water (exp.), n = 11, R = 0.96 , RMS = 1.1 (7)

1718 I. Alkorta, J. Elguero, R. Gallo

Table 4. Experimental values of pKa in liquid ammonia (L NH3) and in water (from Tables 2 and 3).

Compound pKa (L NH3) pKa water (exp.) pKa water (and fitted)

4-Methoxyphenol 6.62 [60] 10.27 [60] 10.27 [60]

Phenol 6.02 [60] 9.99 [60] 9.99 [60]

1-Naphthol 4.97 [60] 9.37 [60] 9.37 [60]

4-Chlorophenol 4.69 [60] 9.20 [60] 9.20 [60]

3-Chlorophenol 4.50 [60] 9.02 [60] 9.02 [60]

4-Carbomethoxyphenol 4.04 [60] 8.47 [60] 8.47 [60]

3-Nitrophenol 3.61 [60] 8.36 [60] 8.36 [60]

8 Diethyl malonate (8.4)a 12.9 [22] 13.3 (Table 3)

12 Acetonitrile 18.3 [22] 28.9 [22] 25.3 (Table 3)

38 Ethyl acetate 18.2 [22] 25.6 [22] 25.7 (Table 3)

39 Acetone 16.5 [22] 20.0 [22] 22.8 (Table 3) a Predicted to have a pKa of 8.3-8.5 (measured <10.5) [22].

Figure 5. The liquid NH3 pKa limits for the different reactions (adding 4 pKa units to the values of Fig. 2). The pKa values correspond to water.

pKa L NH3 = (–2.7±0.3) + (0.82±0.02) corresponding anions. We are going a step further to 2 pKa water (fitt.), n = 11, R = 0.994, RMS = 0.5 (8) propose that the scheme of Fig. 4 can be transformed

into Fig. 5 simply by adding 4 pKa units. Although the slopes are different from 1, in a first This means that H/D exchanges in liquid ammonia approximation it is interesting to compare the intercepts (for practical purposes, D/H between a deuterated of Eq. 4 (–3.2, DMSO exp.) and Eq. 7 (–3.7, water carbon acid and NH3) [59] could be done on very weak exp.) on one hand; and Eq. 6 (–4.0, DMSO fitted) and carbon acids, for instance 1-methylimidazole (24, pKa =

Eq. 8 (–2.8, water, fitted) on the other. From DMSO 27.9), thiophene (23, pKa = 33.8) and toluene (27, pKa exp. to L NH3 the difference is –3.2 –3.7 = –6.9 and = 35.8). Furthermore, acetonitrile (12, pKa = 25.3) could from DMSO fitted to L NH3 the difference is –4.0 –2.8 be alkylated. Roberts et al. experiments show that D/H = –6.8. In the NIST [31] the basicities, as measured by exchange takes place in benzene although very slowly – – – DG, of CH3SO2CH2 , OH and NH2 are 1500, 1606 and [59]. The pKa value of benzene has been reported to be 1657 kJ mol–1, respectively. The intercepts are linearly 43 [61,62], but other authors reported 37 or 43 [63]. This related to DG, depending on the intercepts selected: is consistent with Fig. 3; it is expected that the exchange reaction in PTC would be much faster. Intercept = –(0.040±0.004) DG, n = 3, R2 = 0.990 (with –3.2 and –3.7), RMS = 0.6 (9) 3.4. Possibility to carry out PTC in liquid ammonia Intercept = –(0.042±0.002) DG, It is worth remembering that carbon acids are slightly n = 3, R2 = 0.996 (with –4.0 and –2.8), RMS = 0.3 (10) soluble in liquid ammonia over the temperature range –33 to –40ºC but that they are readily dissolved in

Thus, the increase of pKa on going from DMSO anhydrous ammonia solutions of KNH2 to give colored to H2O to NH3 is directly related to basicity of the solutions (presumably arising from the formation of the

1719 A theoretical study of the limits of the acidity of carbon acids in phase transfer catalysis in water and in liquid ammonia

corresponding potassium salts) [64] although NH3 does but the fine properties due to specific solute-solvent not significantly solvate anions. A good account of liquid interactions have been lost. This is a more apparent

ammonia as solvent is to be found in Cox book [65]. than real situation because if the pKas in DMSO are For practical purposes it is important to know that homogeneous being mainly due to Bordwell [20] those

NH3 is an extremely hygroscopic liquid that boils at in water are from different authors showing a spread of –33ºC. Working at ambient pressures requires the use values far beyond the fine solvent effects. For example

of low temperature thermostats, whereas performing the pKas of ethane have been reported to be 40 [68], experiments under the usual temperature conditions 50 [42] and 52.2 [69]. For tricyanomethane there are leads to the use of pressure equipment. The moderate values of 0 [68] and –5.1 [70]. For compounds 3 (18.7), dielectric constant of ammonia suggests that ionic 12 (28.9) and 13 (25.0), besides the values of Table 2, species in this solvent would be appreciably associated there are other values like 18.24 [71] and 19.5 for 3 [34], [21]. 25 for 12 [68], and 23 [68] and 28.5 for 13 [34].

The phase-transfer catalysts can be the same as This extended scale of pKa of C-H acids has been those used in water biphasic experiments while the used to discuss and to locate more precisely the upper organic solvent must be liquid in the –33 to –40ºC limits for alkylation and other reactions under PTC range (for ambient pressure experiments), but several conditions. This scheme was applied to a liquid ammonia solvents used in PTC are liquid at these temperatures, medium; although alkylations in liquid ammonia with toluene (m.p. –95ºC), chlorobenzene (m.p. –45ºC), sodium amide or potassium amide are rather common diethoxymethane (m.p. –66.5ºC) and dichloromethane (including the Eisleb alkylation) [72], their use in PTC (m.p. –96.7ºC) (note that this solvent has been used to conditions has not been reported. generate carbenes [1-4] as well as to carry out double nucleophilic substitutions) [66,67]. Acknowledgments

4. Conclusions We thank the Ministerio de Ciencia e Innovación (Project No. CTQ2009-13129-C02-02) and the Comunidad We have observed an excellent correlation between Autónoma de Madrid (Project MADRISOLAR2, ref.

calculated deprotonation ΔG and experimental pKa S2009/PPQ-1533) for continuing support. Gratitude for 44 C-H acids in DMSO. From these data we have is also due to the CTI (C.S.I.C.) for an allocation of

built up a scale of pKa values in water that are robust computer time.

References

[1] C.M. Starks, C.L. Liotta, M. Halpern, Phase [11] A. Loupy, Ed. Microwaves in organic synthesis Transfer Catalysis, Fundamentals, Applications, (Wiley-VCH, Weinheim, 2006) and Industrial Perspectives (Chapman & Hall, New [12] S.L. Regen, J. Am. Chem. Soc. 97, 5956 (1975) York, 1994) [13] S.L. Regen, J. Am. Chem. Soc. 98, 6270 (1976) [2] R.A. Jones, Quaternary Ammonium Salts. Their [14] H.J.-M. Dou, R. Gallo, P. Hassanaly, J. Metzger, J. Use in Phase-transfer Catalysis (Academic Press, Org. Chem. 42, 4275 (1977) London, 2001) [15] A.K. Dillow, S.L.J. Yun, D. Suleiman, D.L. Boatright, [3] A.G. Volkov, (Ed.), Interfacial Catalysis (Marcel C.L. Liotta, C.A. Eckert, Ind. Eng. Chem. Res. 35, Dekker, New York, 2003) 1801 (1996) [4] K. Maruoka, Asymmetric Phase Transfer Catalysis [16] K. Chandler, C.W. Culp, D.R. Lamb, C.L. Liotta, (Wiley-VCH, Weinheim, 2008) C.A. Eckert, Ind. Eng. Chem. Res. 37, 3252 (1998) [5] M. Mąkosza, B. Serafin, Roczniki Chem. 39, 1401 [17] A. Loris, A. Perosa, M. Selva, P. Tundo, J. Org. (1965) Chem. 68, 4046 (2003) [6] M. Mąkosza, B. Serafin, Roczniki Chem. 39, 1595 [18] W.S. Ryoo, Ph.D Thesis. Emulsions and (1965) Microemulsions of Water and Carbon Dioxide: [7] M. Mąkosza, B. Serafin,T. Urbański, Chim. Ind. Novel Surfactants and Stabilization Mechanisms (Paris) 93, 537 (1965) (University of Texas at Austin, Austin, 2005) [8] M. Mąkosza, Pure Appl. Chem. 72, 1399 (2000) [19] S. Thayumanavan, Patent EP1778614 A2, 2007 [9] M. Mąkosza, M. Fedoryński, Arkivoc iv, 7 (2006) [20] F.G. Bordwell, Acc. Chem. Res. 21, 456 (1988) [10] C.M. Starks, J. Am. Chem. Soc. 93, 195 (1971) [21] J.J. Lagowski, Pure Appl. Chem. 25, 429 (1971)

1720 I. Alkorta, J. Elguero, R. Gallo

[22] P. Ji, N. Powles, J.H. Atherton, M.I. Page, Org. Lett. N3_2011/U/DMSO-technical_bulletin.pdf 13, 6118 (2011) [38] F.G. Bordwell, J.E. Bartmess, J.A. Hautala, J. Org. [23] A.D. Becke, Phys. Rev. A 38, 3098 (1988) Chem. 43, 3095 (1978) [24] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37, 785 [39] W.P. Jencks, http://evans.harvard.edu/pdf/evans_ (1988) pKa_table.pdf [25] R. Ditchfield, W.J. Hehre, J.A. Pople, J. Chem. [40] K. Shen, Y. Fu, J.-N. Li, L. Liu, Q.-X. Guo, Phys. 54, 724 (1971) Tetrahedron 63, 1568 (2007) [26] M.J. Frisch, J.A. Pople, J.S. Binkley, J. Chem. [41] F.G. Bordwell, Pure Appl. Chem. 49, 963 (1977) Phys. 80, 3265 (1984) [42] I.E. Charif, S.M. Mekelleche, D. Villemin, [27] S. Miertus, E. Scrocco, J. Tomasi, Chem. Phys. 55, N. Mora-Diez, J. Mol. Struct. THEOCHEM 818, 1 117 (1981) (2007) [28] J. Tomasi, M. Persico. Chem. Rev. 94, 2027 (1994) [43] Y. Chiang, A.J. Kresge, P.A. Walsh, J. Am. Chem. [29] R. Cammi, J. Tomasi, J. Comput. Chem. 16, 1449 Soc. 108, 6314 (1986) (1995) [44] J.P. Guthrie, J. Cossar, Can. J. Chem. 64, 2470 [30] M.J. Frisch, G.W. Trucks, H.B. Schlegel, (1986) G.E.Scuseria, M.A. Robb, J.R. Cheeseman, [45] M.J. O’Donnell, W.D. Bennett, W.A. Bruder, G. Scalmani, V. Barone, B. Mennucci, W.N. Jacobsen, K. Knuth, B. LeClef, R.L. Polt, G.A. Petersson, H. Nakatsuji, M. Caricato, F.G. Bordwell, S.R. Mrozack, T.A. Cripe, J. Am. X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, Chem. Soc. 110, 8250 (1988) G. Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, [46] V.E. Matulis, Y.S. Halauko, O.A. Ivashkevich, P.N. K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, Gaponik, J. Mol. Struct. THEOCHEM 909, 19 T. Nakajima, Y. Honda, O. Kitao, H. Nakai, (2009) T. Vreven, J.A. Montgomery, Jr., J.E. Peralta, [47] Ionization Constants of Organic Acids. Ionization F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers, constants of carbon acids (Michigan State K.N. Kudin, V.N. Staroverov, R. Kobayashi, University, USA) http://www2.chemistry.msu.edu/ J. Normand, K. Raghavachari, A. Rendell, faculty/reusch/virttxtjml/acidity2.htm J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, [48] A. Abbotto, S. Bradamante, G.A. Pagani, J. Org. N. Rega, J.M. Millam, M. Klene, J.E. Knox, Chem. 58, 449 (1993) J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, [49] M. Hojatti, A.J. Kresge, W.-H. Wang, J. Am. Chem. R. Gomperts, R.E. Stratmann, O. Yazyev, Soc. 109, 4023 (1987) A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, [50] J.P. Richard, G. Williams, J. Gao, J. Am. Chem. R.L. Martin, K. Morokuma, V.G. Zakrzewski, Soc. 121, 715 (1999) G.A. Voth, P. Salvador, J.J. Dannenberg, [51] T. Bug, H. Mayr, J. Am. Chem. Soc. 125, 12980 S. Dapprich, A.D. Daniels, Ö. Farkas, J.B. Foresman, (2003) J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09, [52] G.I. Almerindo, D.W. Tondo, J.R. Pliego, J. Phys. Revision A.1 (Gaussian, Inc., Wallingford CT, 2009) Chem. A 108, 166 (2004) [31] P.J. Lindstrom, W.G. Mallard (Eds.), NIST Chemistry [53] S. Rayne, K. Forest, Nature Precedings (2010) WebBook, NIST Standard Reference Database http://dx.doi.org/10.1038/npre.2010.4381.1 Number 69 (National Institute of Standards and [54] W.J. Spillane, P. Kavanagh, F. Young, H.J-M. Dou, Technology, USA, 2013) J. Metzger, J. Chem. Soc. Perkin Trans. 1 1763 [32] W.J. Hehre, L. Radom, P.v.R. Schleyer, J.A. Pople, (1981) Ab Initio Molecular Orbital Theory (John Wiley & [55] M. Mąkosza, M Fedorynski, Pol. J. Chem. 70, 1093 Sons, New York, 1986) 411 (1996) [33] I. Alkorta, J. Elguero, Tetrahedron 53, 9741 (1997) [56] M. Mąkosza, Bull. Acad. Pol. Sci. 15, 165 (1967) [34] M.C. Rezende, J. Braz. Chem. Soc. 12, 73 (2001) [57] M. Mąkosza, J. Przyborowski, R. Klajn, A. Kwast, [35] I.A. Koppel, J. Koppel, V. Pihl, I. Leito, M. Mishima, Synlett 1773 (2000) V.M. Vlasov, L.M. Yagupolskii, R.W. Taft, J. Chem. [58] M. Judka, A. Wojtasiewicz, W. Danikiewicz, Soc. Perkin 2 1125 (2000) M. Mąkosza, Tetrahedron 63, 8902 (2007) [36] A.M. Ross, D.L. Whalen, S. Eldin, R.M. Pollack, [59] G.E. Hall, R. Piccolini, J.D. Roberts, J. Am. Chem. J. Am. Chem. Soc. 110, 1981 (1988) Soc. 77, 4540 (1955) [37] Technical Bulletin Reaction Solvent Dimethyl [60] P. Ji, J. Atherton, M.I. Page, J. Org. Chem. 76, 1425 Sulfoxide (DMSO) (Gaylor Chemical Corporation, (2011) USA, 2011) http://chemistry-chemists.com/ [61] A. Sievers, R. Wolfenden, J. Am. Chem. Soc. 124,

1721 A theoretical study of the limits of the acidity of carbon acids in phase transfer catalysis in water and in liquid ammonia

13986 (2002) [67] L. Avila, J. Elguero, S. Julià, J.M. del Mazo, [62] S. Sharma, J.K. Lee, J. Org. Chem. 69, 7018 (2004) Heterocycles 20, 1787 (1983) [63] E.V. Anslyn, D.A. Dougherty, Modern Physical [68] R.G. Pearson, R.L. Dillon, J. Am. Chem. Soc. 75, Organic Chemistry (University Science Books, 2439 (1953) Sausalito, 2006) [69] J.R. Pliego, J.M. Riveros, J. Phys. Chem. A 106, [64] J.H. Takemoto, J.J. Lagowski, J. Am. Chem. Soc. 7434 (2002) 91, 3785 (1969) [70] M.C. Rezende, Tetrahedron 57, 5923 (2001) [65] B.G. Cox, Acids and Bases, Solvent Effects on [71] Y. Chiang, A.J. Kresge, J. Wirz, J. Am. Chem. Soc. Acid-Base Strength (Oxford University Press, 106, 6392 (1984) Oxford, UK, 2013) 93 [72] Z. Wang (Ed.), Comprehensive Organic Name [66] S. Julià, P. Sala, J. del Mazo, M. Sancho, C. Ochoa, Reactions and Reagents (John Wiley & Sons, USA, J. Elguero, J.-P. Fayet, M.C. Vertut, J. Heterocycl. 2009) Chem. 19, 1141 (1982)

1722