Sarah J Pike1, Christopher A Hunter2, Lee Brammer3, Robin N Perutz1. 1. Department of Chemistry, University of York, Heslington, York, YO10 5DD, UK. 2. Department of Chemistry, University of Cambridge, Cambridge, CB2 1EW, UK. 3. Department of Chemistry, University of Sheffield, Brook Hill, Sheffield, S3 7HF, UK.
Abstract
The energetics of halogen bond formation in solution have been investigated for a series of nickel fluoride halogen bond acceptors; trans-[NiF(2-C5 NF4 )(PEt3 )2 ] (A1), trans-[NiF{2-C5 NF3 (4-H)}(PEt3 )2 ] (A2), trans-[NiF{2-C5 NF3 (4-NMe2 )}(PEt3 )2 ] (A3) and trans-[NiF{2-C5 NF2 H(4-CF3 )}(PCy3 )2 ] (A4) with neutral organic halogen bond donors, iodopentafluorobenzene (D1), 1-iodononafluorobutane (D2) and bromopentafluorobenzene (D3), in order to establish the significance of changes from perfluoroaryl to perfluoroalkyl donors and from iodine to bromine donors. 19 F NMR titration experiments have been employed to obtain the association constants, enthalpy, and entropy for the halogen bond formed between these donor-acceptor partners in protiotoluene. For A2-A4, association constants of the halogen bonds formed with iodoperfluoroalkane (D2) are consistently larger than those obtained for analogous complexes with the iodoperfluoroarene (D1). For complexes formed with A2-A4, the strength of the halogen bond is significantly lowered upon modification of the halogen donor atom from I (in D1) to Br (in D3) (for D1: 5≤K285 ≤12 m-1 , for D3: 1.0≤K193 ≤1.6 m-1 ). The presence of the electron donating NMe2 substituent on the pyridyl ring of acceptor A3 led to an increase in -ΔH, and the association constants of the halogen bond complexes formed with D1-D3, compared to those formed by A1, A2 and A4 with the same donors.
The energetics of halogen bond formation in solution have been investigated for a series of nickel fluoride halogen bond acceptors; trans-[NiF(2-C5 NF4 )(PEt3 )2 ] (A1), trans-[NiF{2-C5 NF3 (4-H)}(PEt3 )2 ] (A2), trans-[NiF{2-C5 NF3 (4-NMe2 )}(PEt3 )2 ] (A3) and trans-[NiF{2-C5 NF2 H(4-CF3 )}(PCy3 )2 ] (A4) with neutral organic halogen bond donors, iodopentafluorobenzene (D1), 1-iodononafluorobutane (D2) and bromopentafluorobenzene (D3), in order to establish the significance of changes from perfluoroaryl to perfluoroalkyl donors and from iodine to bromine donors. 19 F NMR titration experiments have been employed to obtain the association constants, enthalpy, and entropy for the halogen bond formed between these donor-acceptor partners in protiotoluene. For A2-A4, association constants of the halogen bonds formed with iodoperfluoroalkane (D2) are consistently larger than those obtained for analogous complexes with the iodoperfluoroarene (D1). For complexes formed with A2-A4, the strength of the halogen bond is significantly lowered upon modification of the halogendonor atom from I (in D1) to Br (in D3) (for D1: 5≤K285 ≤12 m-1 , for D3: 1.0≤K193 ≤1.6 m-1 ). The presence of the electron donating NMe2 substituent on the pyridyl ring of acceptor A3 led to an increase in -ΔH, and the association constants of the halogen bond complexes formed with D1-D3, compared to those formed by A1, A2 and A4 with the same donors.
Halogen bonding interactions are rapidly emerging as key constituents of the molecular recognition toolbox.1, 2, 3 The utility and importance of halogen bonding interactions are evident through their widespread use in applications including crystal engineering,4 materials chemistry,5 supramolecular chemistry,6 and anion recognition7 and through its emergent role in organocatalysis and reactivity.8, 9 Halogen bonding interactions are known to hold great significance in medicinal chemistry10, 11 and are also recognized to be important in achieving function in biological systems.12, 13, 14 The formation of halogen bonding interactions to species in the “ligand domain” has been revealed crystallographically15, 16 and in solution. (The ligand domain consists of those ligand atoms directly bonded to the metal or with a strong electronic interaction with it.)16 There remains, however, a distinct shortage of information about the energetics of these halogen bonding interactions in solution.17, 18, 19 In contrast, the energetics of halogen‐bonded systems involving organic acceptor and donor partners are better documented and include association constants for halogen bonds formed between haloalkynes,20 haloarenes21 and haloalkanes22 as donors with neutral organic bases. Taylor and co‐workers investigated the influence of the type of donor, fluorinated aryl (C6F5I) and fluorinated alkyl (C8F17I), on the strength of the halogen bond formed with a wide range of organic bases in cyclohexane at 298 K.23 The association constants of the binding event were determined by 19F NMR spectroscopy, demonstrating that the equilibrium constants were larger for all the halogen bond donor‐acceptor partners with the iodoperfluoroalkane halogen bond than with the analogous iodoperfluoroarene interaction (e.g., C6F5I⋅⋅⋅OPBu3, 12±2.5 m
−1 and C8F17I⋅⋅⋅OPBu3, 18±4 m
−1).23 Resnati and co‐workers employed 19F NMR spectroscopy to identify that changing the halogendonor atom from iodine in 1,2‐diiodotetrafluorobutane to bromine in 1,2‐dibromotetrafluorobutane significantly weakened the halogen bond formed with quinuclidine in hydrocarbon solvents.24 The influence of the halogen in perfluoroaryl donors (C6F5X, where X=Br and I) on the strength of halogen bonds formed with 1,4‐diazabicyclo[2.2.2.]octane (DABCO) has been studied computationally and experimentally.25 Halogen‐bond complexes formed between C6F5I and DABCO were present in [D8]toluene, but those formed with C6F5Br were less prominent and thus weaker, due to competing solvent interactions. Bowling and co‐workers used 19F and 15N NMR spectroscopy to investigate the formation of intramolecular halogen bonds between a C6F4X unit (X=I and Br) and a pyridyl moiety (py) in which the halogen bond donor and acceptor units are linked by an aryldiyne spacer. The evidence indicated that the C6F4Br⋅⋅⋅py interaction is probably significantly weaker than the corresponding C6F4I⋅⋅⋅py interaction in benzene solution.26We have previously demonstrated that the 19F M−F resonance of trans‐[MF(pyF)(PR3)2] complexes, where M=Ni, Pd or Pt, pyF=fluorinated 2‐pyridyl and R=ethyl (Et) or cyclohexyl (Cy), is extremely sensitive to chemical environment and can be employed as an NMR spectroscopic probe of the energetics of formation of 1:1 halogen bond adducts between iodopentafluorobenzene and the metal fluoride complexes.17 For the most closely related complexes, the enthalpy of dissociation of the halogen bond followed the order: Pt>Pd>Ni.17 These studies established that modification of the electronic nature of the substituents on the fluoropyridyl ligand (pyF), by replacement of one fluorine by hydrogen or CF3, had no significant effect on the thermodynamic data, but the influence of the phosphine ligand was marked. Crystallographic characterization of this class of halogen bonds has been achieved for the closely related self‐complementary nickel fluorides, trans‐[NiF(4‐C6F4I)(PEt3)2] and trans‐[NiF(2‐C6F4I)(PEt3)2], in which a chain of molecules is formed, linked by intermolecular I⋅⋅⋅F halogen bonds.27 Other authors have shown that 1:1 halogen bonded complexes with C6F5I are also formed by nickel fluoride complexes with pincer ligands, and by fluoride complexes of zinc and magnesium.18 The formation of halogen‐bonding interactions in the ligand domain of metal complexes is not restricted to metal halides; a series of bis(η‐cyclopentadienyl)metal hydrides have been shown to form halogen bonds with C6F5I in toluene,19 and metal cyanides have been identified as halogen bond acceptors crystallographically.28Reports on the energetics of halogen bonding in solution mainly focus on the use of iodinated donors, whereas brominated donors feature less frequently,28a owing to the weaker halogen bonds formed with this donor atom,29 which renders acquisition of solution‐based data more challenging.30 Until now, we also lacked information about the behavior of iodoperfluoroalkane donors towards transition metal fluorides. In this paper, we describe a systematic study of the influence of structural variations of the donor and acceptor species on the binding constants and energetics of halogen bond formation between a series of structurally related nickel fluorides A2–A4 and a range of organic iodo‐ and bromo‐perfluorocarbon donors D1–D3 in protiotoluene (Scheme 1). The halogen bond donors are iodopentafluorobenzene (the standard), iodononafluorobutane and bromopentafluorobenzene. The halogen bond acceptors maintain the square planar nickel fluoride geometry, but vary the substituents on the pyridyl ring and, in A4 the phosphine ligand. Although A4 represents a change in the two parameters, both the pyridyl ring and phosphine ligand, we have previously shown that substitution of F by CF3 on the pyridyl ring had little effect on the energetics of halogen bond formation. The results provide a benchmark for these halogen bond donors, which are in common use in supramolecular assemblies directed by halogen bonding.
Scheme 1
a) Metal fluorides A1–A4 employed as halogen bond acceptors in this study. b) halogen bond donors D1–D3.
a) Metal fluorides A1–A4 employed as halogen bond acceptors in this study. b) halogen bond donors D1–D3.
Results and Discussion
Nickel fluorides were employed as halogen bond acceptors as they are soluble in toluene and do not display appreciable self‐association.17a We reported the formation of the halogen‐bonded adduct D1⋅A1 and D1⋅A4 earlier.17
A2–A4 were prepared according to known literature procedures.17, 31 The study of A2–A4 permits investigation of the influence of the substitution pattern of the fluoropyridyl ring on the energetics of the halogen bond formed with a range of organic donors (D1–D3) (Scheme 1). Accordingly, a series of 19F NMR titration experiments were performed on metal fluorides (A2–A4) through the addition of increasing quantities of halogen bond donors (D1–D3) in protiotoluene. The 19F resonances of the fluoride ligand directly bound to the metal center in the adducts of A1–A4 appear at high field (e.g., δ=−339.3 ppm for A3 at 285 K, Figure 1(i)) and no overlap with other signals occurs during the titration experiments.
Figure 1
Stack plot of 19F NMR spectra in the nickel fluoride region (toluene‐h8, 285 K), at different molar ratios of [D2]/[A3]. Molar ratios a) 0, b) 0.6, c) 1.0, d) 2.3, e) 3.3, f) 5.9, g) 7.1, h) 10, i) 15.3.
Stack plot of 19F NMR spectra in the nickel fluoride region (toluene‐h8, 285 K), at different molar ratios of [D2]/[A3]. Molar ratios a) 0, b) 0.6, c) 1.0, d) 2.3, e) 3.3, f) 5.9, g) 7.1, h) 10, i) 15.3.The NMR titration experiments show only one 19F resonance as the halogen‐bond adduct is in fast exchange on the NMR timescale. Upon the addition of D1 or D2, the 19F NMR signal of the metal fluoride in A2–A4 exhibited a marked downfield shift with δ rising by ca. 20–30 ppm at 285 K upon treatment with a large excess of either perfluoroiodine donor (for example, A2
δ=−367.9 and for D2⋅A3
δ=−339.3, see Figure 1). This shifting behavior is attributed to formation of the halogen bond adduct.17 As expected for an equilibrium between monomers and an intermolecular complex, the chemical shifts increase at lower temperature as the equilibrium is shifted towards the adduct (Figure 2). For titrations of D3 against A2–A4, the changes in the spectra are negligible in the temperature range used for the D1 and D2. For this reason, experiments were conducted at 193 K in order to shift the equilibria towards the halogen bonded adduct, but the change in chemical shift was significantly smaller than for D1 and D2 at 8–10 ppm (see Supporting Information). Through fitting the variation of the chemical shift of the 19F NMR resonance with the molar ratio of [donor]/[acceptor], association constants for the halogen bonding interaction were obtained by titrations for D2–3⋅A2–4. For all systems, the titration data fit well to a 1:1 binding isotherm (Figure 2 and Supporting Information) as in Equation (1):
Figure 2
Titration curves at 246, 259, 271, 285, and 294 K for D2 and A3 in toluene‐h8, showing δ(19F) of the metal fluoride vs. [D2]/[A3]. [A3]=17 mmol dm−3. Diamonds are experimental points; dashed line shows best fit to a 1:1 binding isotherm.
Titration curves at 246, 259, 271, 285, and 294 K for D2 and A3 in toluene‐h8, showing δ(19F) of the metal fluoride vs. [D2]/[A3]. [A3]=17 mmol dm−3. Diamonds are experimental points; dashed line shows best fit to a 1:1 binding isotherm.in which R=C6F5 or C4F9 and X=I or Br.There are two parameters to be fitted: the equilibrium constant K and the downfield shift from the signal of the free fluoride for the coordinated fluoride of the adduct, Δδ
19F.32 The fitting routine for each titration curve models the chemical shift difference, Δδ
fit, between the free metal complex and the halogen‐bonded adduct. The value of Δδ
fit lies between 31 and 38 ppm at 285 K for A2–A4 with iodinated donors D1 or D2 and between 13 and 19 ppm for A2–A4 with brominated donor D3 at 193 K (Table 1). The change in the chemical shift (Δδ) of the 19F resonance observed experimentally correlates well with the calculated Δδ
fit values (Supporting Information). The values of Δδ
fit varied with temperature by no more than 1.1 ppm. From the experimental titration data, both the standard enthalpy and entropy of the halogen bonding interactions between A2–A4 and D1 and between A2–A4 and D2 were calculated from Van't Hoff plots. Analysis of the titration data gave excellent fits with correlation coefficients R
2>0.975 for all systems studied (Figure 3 and Supporting Information). The thermodynamic parameters and association constants for all experiments are reported in Table 1.
Table 1
Summary of thermodynamic parameters for halogen bonding of donors D1–D3 with nickel fluorides A1–A4 in protiotoluene.[a]
Donor
Acceptor
ΔH°[kJ mol−1]
ΔS°[J mol−1 K−1]
K285[m−1]
Δδfit285 K
R2
D1
A1
−16±1[b]
−42±4[b]
5.5±0.1[b]
33.4[c]
–
D1
A2
−18±2
−46±8
7.1±0.2
32.6
0.995
D1
A3
−17±5
−39±19
11.3±0.2
33.5
0.975
D1
A4
−19±4[d]
−54±1[d]
4.4±0.2[d]
35.6[e]
0.998[c]
D2
A2
−23±4
−53±14
18.6±0.3
31.4
0.993
D2
A3
−22±3
−50±12
31.8±0.3
32.0
0.994
D2
A4
−19±4
−51±13
6.9±0.1
38.0
0.990
D3
A2
–
–
1.0±0.1[f]
17.7[f]
–
D3
A3
–
–
1.6±0.1[f]
12.5[f]
–
D3
A4
–
–
1.3±0.1[f]
18.9[f]
–
[a] Errors at 95 % confidence level. Δδ=Chemical shift difference between free metal fluoride, Ni−F, and R−X⋅⋅⋅F−Ni adduct calculated by the fitting routine. [b] From Ref. 17a. [c] At 289 K from Ref. 17a. [d] From Ref. 17b. [e] At 303 K from Ref. 17b. [f] Determined at 193 K.
Figure 3
Van't Hoff plots for halogen‐bonded pairs D1⋅A2, D2⋅A2, D1⋅A3, D2⋅A3 and D2⋅A4.
Summary of thermodynamic parameters for halogen bonding of donors D1–D3 with nickel fluorides A1–A4 in protiotoluene.[a][a] Errors at 95 % confidence level. Δδ=Chemical shift difference between free metal fluoride, Ni−F, and R−X⋅⋅⋅F−Ni adduct calculated by the fitting routine. [b] From Ref. 17a. [c] At 289 K from Ref. 17a. [d] From Ref. 17b. [e] At 303 K from Ref. 17b. [f] Determined at 193 K.Van't Hoff plots for halogen‐bonded pairs D1⋅A2, D2⋅A2, D1⋅A3, D2⋅A3 and D2⋅A4.The presence of the strongly electron‐donating group, NMe2, at the 4‐position of the pyridyl ring in A3 leads to larger K values for D1⋅A3 and D2⋅A3 than for analogous adducts formed with A2, which bear a hydrogen at the same position on the ring (Table 1 and Figure 4). The PCy3 complex, A4, forms halogen bonds with the iodoperfluorocarbon donors, D1 and D2, that have lower K values than with any of the PEt3 bearing complexes A1–A3 (Table 1). The electronic nature of the donor influences the strength of the interaction with complexes following the order D2>D1≫D3 (Table 1).23, 33 Use of the iodoperfluoroalkanedonorD2 results in considerably larger equilibrium constants than those for iodopentafluorobenzene,1d, 19, 23, 30 whereas modification of the halogendonor atom from iodo‐ to bromo‐ in the perfluoroarenes D1 and D3 greatly reduces the strength of the interaction of the halogen bond adduct formed (Table 1).24, 26, 34 The titration data also show a significant reduction in the magnitude of chemical shift change of the 19F resonance for D3⋅A versus D1⋅A (n=1–4). The association constants obtained for the halogen bonding interaction with D3 were all recorded at a single temperature (193 K), as above this temperature the binding constant was too low to be measured accurately. The differences between K
193 values for D3 with different acceptors are small.
Figure 4
Variation in K with donor and acceptor. Measurements at 285 K for D1 and D2 but 193 K for D3.
Variation in K with donor and acceptor. Measurements at 285 K for D1 and D2 but 193 K for D3.The halogen bonding interactions of D1 with A1–A4 and D2 with A2–A4 have favorable enthalpic terms and unfavorable entropic terms, (D1: −19≤ΔH°≤−16 kJ mol−1 and −54≤ΔS°≤−39 J K−1 mol−1 and D2: −23 ≤ ΔH° ≤ −19 kJ mol−1 and −53≤ΔS°≤−50 J K−1 mol−1) in line with literature reports.17 The enthalpic contribution for the D2⋅A2 complex is larger than for D1⋅A2, with a difference that is just significant at the 95 % confidence level. The ΔH° and ΔS° terms of the halogen bonding interactions of A2 and A3 are each comparable for both the aromatic donorD1,17b and aliphatic donorD2, showing that the changes in the energetics on introducing the NMe2 substituent on the fluoropyridyl ring are too small to identify the source of the effect. The ΔH° and ΔS° terms for the halogen bonding interactions formed between donors D1 and D2 and acceptor A4 are within error of the analogous interactions formed with A1–A3, despite the smaller binding constants for the former. As only one temperature was employed to study the halogen bond formation of D3 with A2–A4, the enthalpic and entropic terms could not be calculated for this interaction.
Conclusion
The abilities of a series of structurally related nickel fluorides, trans‐[NiF{2‐C5NF2H(4‐CF3)}(PEt3)2] (A2), trans‐[NiF{2‐NC5F3(4‐NMe2)}(PEt3)2] (A3) and trans‐[NiF{2‐C5NF2H(4‐CF3)}(PCy3)2] (A4), to accept halogen bonds from a range of organic halogen‐bond donors, iodopentafluorobenzene (D1), 1‐iodononafluorobutane (D2) and bromopentafluorobenzene (D3) in protiotoluene have been established using a series of 19F NMR titration experiments. These measurements supplement previous studies of trans‐[NiF{2‐C5NF4}(PEt3)2] (A1) and trans‐[NiF(2‐C5NF4)(PCy3)2] with D1. Binding constants have been determined for the interactions between D1–D3 and A2–A4. Enthalpies and entropies of halogen bond formation between iodinated halogen bond donors and nickel fluoride acceptors have been determined.For halogen bonds formed with A1–A4, the aliphatic donorD2 has association constants close to three times greater than those observed for the aromatic donorD1, which is in accordance with the relative strengths of the two donors observed in halogen bonds with organic acceptors and transition metal hydrides.19 There is a corresponding and consistent increase in −ΔH°. The association constants for the halogen bond interaction with D1 are significantly higher than those observed with the brominated analogue D3, which reflects the markedly different donor capabilities of the halogen atoms, I and Br. These observations are in line with reports of corresponding complexes formed with DABCO,25 and correlate with studies of intramolecular halogen bonding.26 The introduction of an NMe2 electron‐donor group on the fluoropyridyl ring results in a marked increase in association constant.This investigation provides the first determination of energetics of halogen bond formation for aliphatic donors and for bromine donors with metal‐fluoride acceptors. These findings emphasize the utility of metal‐fluorides in providing a benchmark for strengths of halogen bonds to metal complexes and allow comparisons to the strengths of corresponding hydrogen bonds. As for organic systems, the strongest halogen bonds are formed with an iodoperfluoroalkanedonor. We anticipate that this study could also play an important role in the future design of synthetic supramolecular systems which exploit halogen bonding interactions.
Conflict of interest
The authors declare no conflict of interest.As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.SupplementaryClick here for additional data file.
Authors: Marta E G Mosquera; Irene Egido; Carlos Hortelano; María López-López; Pilar Gómez-Sal Journal: Faraday Discuss Date: 2017-10-13 Impact factor: 4.008
Authors: Mohammed G Sarwar; Bojan Dragisic; Lee J Salsberg; Christina Gouliaras; Mark S Taylor Journal: J Am Chem Soc Date: 2010-02-10 Impact factor: 15.419
Authors: Vargini Thangavadivale; Pedro M Aguiar; Naseralla A Jasim; Sarah J Pike; Dan A Smith; Adrian C Whitwood; Lee Brammer; Robin N Perutz Journal: Chem Sci Date: 2018-03-23 Impact factor: 9.825