Research 1..9
Total Page:16
File Type:pdf, Size:1020Kb
pubsdm_prod | ACSJCA | JCA10.0.1465/W Unicode | research.3f (R3.6.i5 HF05:4232 | 2.0 alpha 39) 2014/10/10 09:17:00 | PROD-JCA1 | rq_3130727 | 12/22/2014 01:44:23 | 9 | JCA-DEFAULT Article pubs.acs.org/JPCA 1 Superacidity of closo-Dodecaborate-Based Brønsted Acids: a DFT 2 Study ,† † † ‡ ‡ ,† 3 Lauri Lipping,* Ivo Leito, Ivar Koppel, Ingo Krossing, Daniel Himmel, and Ilmar A. Koppel* † 4 Institute of Chemistry, University of Tartu, 14a Ravila St., Tartu 50411, Estonia ‡ 5 Institute for Inorganic and Analytic Chemistry, University of Freiburg, 21 Albertstr., Freiburg D-79104, Germany 6 *S Supporting Information 7 ABSTRACT: The structures and intrinsic gas-phase acidities (GA) of some dodecaborane − 8 acids, the derivatives of YB H H(Y= PF ,NH,NF, NMe ), B H H , and B H H − 12 11 3 3 3 3 12 12 2 12 12 9 (HA, H2A, and HA , respectively) have been computationally explored with DFT B3LYP 10 method at the 6-311+G** level of theory as new possible directions of creating 11 superstrong Brønsted acids. Depending on the nature and number of the substituents 12 different protonation geometries were investigated. In general, the GA values of the neutral 13 systems varied according to the substituents in the following order: CF3 < F < Cl and in fl 14 case of anionic acids: CF3 < Cl < F. The dodecatri uoromethyl derivative of H2A, 15 B12(CF3)12H1H2, emerges as the strongest among the considered acids and is expected to 16 be in the gas phase at least as strong as the undecatrifluoromethyl carborane, 17 CB11(CF3)11H1H. The GA values of the respective monoanionic forms of the considered 18 acids all, but the (CF3)11 derivative, remained higher than the widely used threshold of ’ ’ 19 superacidity. The HA derivatives (Y = PF3,NF3)GAs were approximately in the same ’ fi 20 range as the H2A acids . In the case Y = NH3 or NMe3 the GA values were signi cantly fl 21 higher. Also, the pKa values of B12H12H2,CB11H12H, and their per uorinated derivatives in 1,2-dichloroethane (DCE) were fl 22 estimated with SMD and cluster-continuum model calculations. The obtained estimates of pKa values of the per uorinated 23 derivatives are by around 30 units lower than that of trifluoromethylsulfonylimide, making these acids the strongest ever fi 24 predicted in solution. The derivatives of B12H12H2 are as a rule not signi cantly weaker acids than the respective derivatives of 25 CB H H. This is important for expanding practical applicability of this type of acids and their anions, as they are synthetically 11 12 − 26 much more accessible than the corresponding CB11H12 derivatives. 27 ■ INTRODUCTION dinating), extremely inert anionic bases whose conjugate acids 51 1−9 1−4 are the strongest Brønsted acids presently known. These 52 28 Practical and fundamental reasons have motivated scientists anions have extremely low nucleophilicity, electrophilicity and 53 29 to search for molecules and molecular systems that are more 5,6 oxidizing activity and have been used as counterions for 54 30 acidic than known before. Several strategies have been strongly electrophilic cationic species that are also extremely 55 31 proposed to design highly acidic molecules. An obvious route is 12 8 strong acids. The first computational evidence that the 56 32 introducing electron withdrawing substituents (e.g., fluorination intrinsic gas-phase superacidity of boron-based acids can exceed 57 33 or trifluoromethylation) into already strong or superstrong that of sulfuric acid, the “classical” basis for definition of 58 34 Brønsted acids. Well-known examples are fluorosulfonic and superacidity, by many powers of ten was published in 2000 in 59 35 trifluoromethanesulfonic acids, which can be regarded as the work of some of the present authors with co-workers. It was 60 36 derivatives of sulfuric acid. Another much used approach is 61 37 increasing the hydrogen ion donor ability of a Brønsted acid followed by a density functional theory (DFT) investigation of − the intrinsic gas-phase acidities of some smaller carborane 62 38 HA by mixing it with a Lewis acid so that the anion A formed 13 63 39 in the ionization of HA is converted into the highly stabilized derivatives. Further computational extension and revision of the intrinsic gas-phase superacidity scale was carried out in 64 40 complex with this Lewis acid. This principle is operational in, 1 2009. 65 41 e.g., magic acid or HSbF6. Although, both of these acids are 42 strong they are prone to forming reaction side products by The practical chemical use of these novel reagents has yet to 66 43 means of fluorination. gather impetus. The main obstacle is the high cost and limited 67 44 The electron-withdrawing effects of substituents are availability of the monocarba-closo-borane acids and their salts. 68 45 especially powerful if synergized with the excellent charge The quantities that presently can be prepared via complex and 69 46 delocalization ability of the different spherical boron com- time-consuming synthetic pathways are suitable for obtaining 70 7−11 47 pounds. Decades of work on boron compounds and their 48 substituted derivatives have resulted in a new generation of Received: June 30, 2014 49 anions−derivatives of the closo-dodecaborate and monocarba- Revised: December 14, 2014 50 closo-dodecaborate anions−superweak (i.e., very weakly coor- © XXXX American Chemical Society A dx.doi.org/10.1021/jp506485x | J. Phys. Chem. A XXXX, XXX, XXX−XXX The Journal of Physical Chemistry A Article 71 small quantities of valuable substances, enough for small scale calculations. In order to obtain reliable results, different possible 134 72 experiments, but not for extensive or large-scale use. However, protonation geometries and the effects of substituents on the 135 7 fi fi 73 there are developments taking place in this eld, as well. protonation site are compared. For the rst time the pKa values 136 14−19 74 In the recent reports some experiments of fundamental for the borane and carborane acid derivatives of that strength 137 75 interest have been reported made possible by the free acids are reported for 1,2-dichloroethane. 138 76 CB11XnR12‑nH (X = Cl, F; R = H, CH3; n = 6, 11). However, 77 the question about the availability of the derivatives of ■ METHODS 139 − 78 CB11H12 remains. Therefore, the quest for anions of similar Unless otherwise indicated, density functional theory (DFT) 140 2− 79 inertness and low basicity, but easier to prepare, is constantly calculations were carried out on B X H ‑ (X = F, Cl, and 141 12 n 12 n − 80 on. The interest in the derivatives of H2(B12X12) comes from a CF3; n = 0, 1, 6, 11, and 12) and YB12XnH11‑n (Y = PF3,NH3, 142 2− 81 fact that the salts of the starting compound B12H12 are NF3, and NMe3,X= F, Cl, CF3; n = 0, 1, 6, and 11) cages and 143 82 commercially available at a reasonable price. Furthermore, as their protonated forms at the B3LYP/6-311+G** level with the 144 2− 83 B12X12 are dianions, a useful approach to further increase Gaussian 09 system of programs with full thermal corrections 145 29 84 their acidity could be decreasing the availability of the negative to Gibbs energies at the optimized structures. In the largest 146 85 charge by using a single positively charged group, which turns systems where X = CF3 and n = 12 (n = 11 with the Y-borates), 147 86 the bianionic closo-dodecaborate into a monoanion. Besides a the vibrational analysis at the B3LYP/6-311+G** level failed, 148 30,31 87 positive charge this group should have electron-withdrawing so thermal corrections were calculated using the RI-BP86 / 149 32 33 88 properties, should not contain any well-defined protonation def-TZVP level with default RI-J auxiliary basis on the 150 89 centers and should be reasonably stable. Based on these corresponding optimized structures using the Turbomole 151 − + − + − + 34−36 90 considerations we have chosen the PF3 , NF3 , NH3 , and 6.4 program system (see the Discussion for details). 152 − + 91 NMe3 groups for this purpose. The starting position of the substituent insertion for the 153 20 92 In a recent report the solution-phase superacidities of two clusters without Y-group was considered as position 1 (Figure 154 f1 93 diprotic acids, based on the closo-dodecaborate anions 1). In the case of the Y-substituted borates the position 1 was 155 f1 94 H2(B12X12) (X = Cl, Br) have been estimated indirectly by 95 Reed et al. using the anions’ ν(NH) basicity scale based on NH + 96 stretching frequency shifts of Oct3NH in CCl4 induced by H- 97 bond formation between the latter cationic proton donor and 21,22 98 the superweak anionic base. Based on these results and the 99 ability of the derivatives of H2(B12X12) (X = Cl, Br) to 20 100 protonate benzene by forming [C6H7]2[B12X12] salts their 101 acid strength, even corresponding to the detachment of the 102 second proton, was considered to be comparable with the 23 103 respective carborane acids. These results were explained with 20 104 the hypothesis that “halogeno substituents on both anions 105 form an effective screen for negative charge that is delocalized ’ 106 and buried within the icosahedral cage”. Therefore, for accurate Figure 1. Numbering of closo-dodecaborate s vertexes. 107 experimental and computational estimation of intrinsic acidity 108 the careful analysis of possible protonation sites is necessary. the vertex with the Y-group. Replacement of the hydrogen 156 109 However, no direct measurement or computational estima- atoms with substituents was done subsequently in the following 157 110 tion of these acids in any solvent has been published according groups of vertexes (belts): 1, 2···6, 7···11, and 12.