Reactions of Indene and Indoles with Platinum Methyl Cations: Indene C-H Activation, Indole π vs. Nitrogen Lone-Pair Coordination

Abstract

Reactions of indene and various substituted indoles with [(diimine)Pt^II(Me)(TFE)]⁺ cations have been studied (diimine = ArN = C(Me) − C(Me) = NAr; TFE = 2,2,2-trifluoroethanol). Indene displaces the TFE ligand from platinum to form a stable π coordination complex that, upon heating, undergoes C-H activation with first order kinetics, ΔH^‡ = 29 kcal/mol, ΔS^‡ = 10 eu, and a kinetic isotope effect of 1.1 at 60 °C. Indoles also initially form coordination complexes through the C2=C3 olefin, but these undergo rearrangement to the corresponding N-bound complexes. The relative rates of initial coordination and rearrangement are affected by excess acid or methyl substitution on indole.

Introduction

Selective C-H bond activation is a potentially valuable approach to synthetic problems in areas ranging from fuels and bulk chemicals to fine chemicals and pharmaceutical synthesis.¹ Studies of C-H activation in our laboratory have focused on models of the Shilov system,² particularly [(diimine)Pt^II(Me)(solv)]⁺ (2, diimine = ArN = C(Me) − C(Me) = NAr; solv = 2,2,2-trifluoroethanol (TFE), H₂O).³ These cations are capable of activating a variety of carbon-hydrogen bonds.⁴ Cations 2 can be generated by protonolysis of (diimine)Pt^IIMe₂ species 1 in TFE with aqueous HBF₄^{3, 4ab} or BX₃ (X = C₆F₅,^4cd F⁵), the latter producing H⁺ by boron coordination to TFE.^4c In deuterated solvent 2 is formed as a mixture of two isotopologs. Reaction of 2 with a C-H group then results in liberation of methane as a mixture of CH₄ and CH₃D along with the formation of a new alkyl or aryl platinum(II) complex [(diimine)Pt^II(R)(solv)]⁺ (Scheme 1).⁶

Scheme 1.

The presence of donor heteroatoms (N, O, S, etc.) in the substrate may inhibit C-H activation because these heteroatoms tend to bind tightly to electrophilic metal centers and block C-H bond coordination. In the cases of methanol and dimethyl ether,^4e stable complexes of the organic substrate with platinum are observed, but kinetics studies indicate that substrate-platinum binding is rapid and reversible, and that re-binding (or rearrangement) of the substrate to a C-H bond-coordinated intermediate is rate-determining. Heteroatom binding can be used to advantage, however, in coordinatively directed cyclometallation reactions,⁷ the basis of many synthetic methods. Several recent examples of these involving nitrogen-containing heterocycles have appeared.⁸ Specifically regarding indole, a functionality found frequently in natural products and pharmaceutical agents, several groups have identified conditions for cyclization of indoles with pendant olefins,⁹ and both Sames¹⁰ and Sanford¹¹ have reported selective arylation of simple indoles via palladium-catalyzed C-H activation.

Prior literature reports of stable group 10 metal-indole complexes are few in number. Platinum-coordinated tryptophan and tryptamine derivatives have been studied,¹² but only a single platinum complex, cis-Pt(indole)₂Cl₂, having a neutral, simple L-type donor indole ligand¹³ has been reported. Unfortunately, it was characterized only by elemental analysis and infrared, and neither offers insight into the binding of the indole ligand.¹⁴ Moreover, although (1-indolyl)metal amides (functioning as X-type ligands) are common,¹⁵ and indole-metal arene complexes have been described for several transition metals,¹⁶ we are aware of no examples of transition metal complexes having indole C2=C3 olefin ligation.

The present work examines the reactions of indoles with cations 2. Towards this end, we describe the reaction of these cations with indene, indole’s hydrocarbon analog, as well as with various methyl-substituted indoles. We find that indene forms a stable olefin coordination complex that undergoes C-H activation upon heating. Analogously, indoles initially form olefin coordination complexes, but rather than undergoing C-H activation, these complexes rearrange to stable N-bound 3H-indole species. We expect that C-H activation of nitrogen-containing unsaturated heterocycles is more likely to proceed from a metal-heterocycle π complex than an N-complex, so strategies to extend the lifetime of platinum(II)-indoleπ complexes have been pursued. We find that the rate of rearrangement of indole π complexes to N-complexes can be diminished by introducing steric bulk on the indole or by increasing [H⁺].

Results and Discussion

Reactions of Indene with Cations 2

Treatment of platinum cations 2 with one equivalent of indene results in the formation of stable olefin complexes, 3 (Scheme 2). In each case, complexation is efficient and rapid. Exchange of coordinated indene with 1 equivalent (15 mM) of added indene-d₃ under the same conditions is slower, with k_obs ~ 5 × 10⁻³ s⁻¹, t_1/2 ~ 1 min (from McKay analysis).¹⁷ The resulting products 3 can be characterized by NMR and electrospray MS methods. The ¹H NMR spectrum for 3a is sharp at ambient temperature, and ¹⁹⁵Pt-H coupling constants of 74 and 79 Hz are observed for the C2 and C3 protons of indene. The C1 methylene protons appear as a pair of doublets (2J_H-H = 23 Hz) with weak ¹⁹⁵Pt side bands (J_Pt-H = 41 Hz). By contrast, ambient temperature ¹H NMR spectra of 3b, 3c, and 3d give broad signals for the C1, C2, and C3 indenyl C—H protons near 3.2, 6.5, and 5.6 ppm, respectively. In all cases, Pt-CH₃ signals were shifted significantly upfield from the parent cation 2: whereas 2 typically displays δ(Pt-CH₃) in the range 0.6 – 1.3 ppm, indene complexes 3 have δ(Pt-CH₃) between −0.8 and −0.3 ppm. This shift is similar to that for a previously-reported η²-benzene adduct, which has δ(Pt-CH₃) = −1.3 ppm at −33 °C.^4a Upon gentle heating, complexes 3 react to form η³-indenyl complex 4 and methane (Scheme 2). In the case of 3a, the reaction is clean (> 95% yield) as determined by ¹H NMR. By contrast, heating complexes 3b and 3c results in formation of a complex mixture of products, including 4.

Scheme 2.

Mechanism of Indene Activation

First-order rate constants were measured for the conversion of indene complex 3a to indenyl complex 4a (Table 1). Importantly, two trials at 60 °C with 1 and 5 equivalents of indene gave the same rate constants (1.24 (9) × 10⁻⁴ s⁻¹ and 1.37 (9) × 10⁻⁴ s⁻¹) within error, indicating a zero order rate dependence on [indene]. Activation parameters for the conversion of 3a to 4a, measured by Eyring analysis for kinetic runs conducted at temperatures ranging from 50 to 90 °C (Figure 1), are ΔH^‡ = 28.7 ± 0.7 kcal·mol⁻¹ and ΔS^‡ = 9.7 ± 2.2 e.u.¹⁸ A kinetic isotope effect was determined for the conversion indene-d₃ to deuterated indenyl complex 4a–d₂ (Scheme 3). 1,1,3-Trideuteroindene was prepared according to a previously reported procedure¹⁹ and was complexed to 2a as above. Conversion of 3a-d₃ to 4a-d₂ proceeded smoothly to afford a selectively deuterated product. Comparisons of measured rate constants for parallel runs gave a kinetic isotope effect of 1.1.

Table 1.

First-order rate constants for the conversion of indene π complex 3a to η³-indenyl complex 4a.

Entry	T (°C)	k (s−1)	t(1/2) (min)
1	50	3.38 (11) × 10−5	342(11)
2	60	1.31 (9) × 10−4	88.5(58)
3	70	4.60 (18) × 10−4	25.1(10)
4	80	1.84 (10) × 10−3	6.28(32)
5	90	5.10 (10) × 10−3	2.28(38)

Scheme 3.

The small, positive value for ΔS^‡ and the small KIE are suggestive of an intramolecular 3a-to-4a process not involving rate-determining C-H oxidative addition. A mechanistic scenario that is in accord with these features involves rate-determining rearrangement of π complex 3a to C-H σ complex 5, followed by more rapid C-H oxidative cleavage and reductive elimination of methane (Scheme 4). Analogous reaction of 2e (solv = CF₃CD₂OD) with p-xylene was found to afford a mixture of products resulting from competitive aryl (k_H/k_D ≈ 4) and benzylic (k_H/k_D ≈ 1) C-H activation processes, the latter affording an η³-benzylic product. ^4c Rearrangement of 3a to 4a thus resembles the benzylic activation process, and suggests that benzylic C-H bond activation by 2e also proceeds via an initially formed arene π complex that undergoes rate-determining rearrangement to a C-H σ complex, followed by C-H oxidative cleavage.

Scheme 4.

Reactions of Cations 2 with Indoles

Indole reacts rapidly with platinum cations 2, ultimately to give N-complexation products 8 (Scheme 5). Analogously to reaction with indene, the initially formed adduct is a C2=C3 π complex 6; however, N-ligation, likely via transient 7, inhibits C-H activation, and tautomer 8 is the stable adduct that is generated.²⁰ Reaction of 2a with indole at 40 °C begins with a buildup of 6a-d₂ (¹H NMR), which converts to 8a in minutes. H/D exchange of N1-H with solvent is rapid relative to complexation, but exchange of C3-H may occur before or after π complexation.²¹ Although the participation of platinum in the H/D exchange cannot be excluded, this exchange is rapid in the presence of 0.2 mole percent BF₃-TFE solution (not containing 1a) and is complete in minutes at room temperature. NMR spectra are consistent with assignment of the 3D tautomer 8: in all three complexes (8a, 8b, 8c) the ¹³C signal for the C3 methylene carbon appears as a multiplet at ~44 ppm, and the ¹H signal for the C2 hydrogen appears as a singlet with ¹⁹⁵Pt satellites (J_Pt-H ~ 40 Hz) at ~ 8 ppm.

Scheme 5.

Reactions of methyl-substituted indoles were studied with cation 2a. Reactions of 2a with 1-methylindole and 2-methylindole proceed through the intermediacy of observable and more stable π complexes (9, 11; Scheme 6).^{22 1}H NMR spectra for 9 and 11 share important features with indene complex 3a: each shows a significantly upfield Pt-CH₃ chemical shift relative to cation 2a. Upon rearrangement to the corresponding N-bound complexes 10 and 12, the Pt-CH₃ signal returns downfield (Table 2). Other heterocycles (pyrrole, carbazole, benzofuran) did not afford clean results under these conditions; most notably, N-protected indoles gave rise to complex mixtures of products. ₂₃

Scheme 6.

Table 2.

NMR data for Pt-Me groups of TFE adduct 2a, indene π complex 3a, and the π and N-indole adducts.

Complex	1H δa (ppm)	13C δ (ppm)	JPt-H b (Hz)
2a (TFE)c	+0.66	21.5	72
3a (indene)	−0.74	4.7	72
6a (indole π)	−1.02	−d	62a
8a (indole N)	0.74	−12.1	78
9	−0.75	−3.6	73
10	0.53	−13.7	78
11	−1.16	−3.0	73
12	0.26	−16.8	78

^aData recorded at 500 or 600 MHz.

^bRecorded at 300 MHz.

^cNotes 4bc.

^dThe lifetime of this species is too short to record ¹³C NMR.

The mechanism of conversion of π-bound indole complexes to final N-bound complexes 8, 10, and 12 likely involves direct (intramolecular) slippage of platinum from the C2 = C3 π bond to N1 because the rate of π to N1 conversion is independent of [indole]. Moreover, in a simple crossover experiment, a solution of the 2-methyl indole adduct 9 (~90% conversion from 2a) was treated with 1 equivalent of 1-methylindole. Upon further reaction a distribution of 81(1)% 9, 10(1)% 10, 9(1)% 11 was observed, and finally a mixture containing 91(1)% 10 and 9(1)% 12 was observed. Thus little crossover is observed, excluding a mechanism involving indole displacement from the metal center.

Rearrangements of complexes 9 and 11 to complexes 10 and 12 are spontaneous at room temperature (also zero-order in indole). For the conversion of 2a to 10 at 30 °C, the rate of rearrangement of 9 to 10 is close to the rate for the binding of 2-methylindole to cation 2a, thus enabling NMR observation of all three species simultaneously in a single experiment (Figure 2). First order rate constants for the rearrangements of π complexes 9 and 11 were measured at 40 °C in the presence of excess indole (Table 3). Strikingly, rearrangement of these π complexes requires hours at 40 °C, whereas N-ligated complexes 8 are fully formed at room temperature in only minutes, indicating that methylation of the indole in the 1 or 2 position substantially retards the rate of rearrangement of indole π complexes.

Table 3.

First order rate constants for rearrangement of π complexes at 40 °C.

Entry π	Complex	product	k (s−1)	T1/2 (min)
1	6a	8a	1.24(16) × 10−2	0.94(12)	π → N
2	9	10	3.05(7) × 10−4	37.9(86)	π → N
3	11	12	7.87(37) × 10−5	147(71)	π → N
4	3a	4a	8.4(6) × 10−6a	1380(98)	C-H activation

^aExtrapolated from Eyring plot.

The rate of rearrangement of indole π complex 6a to N-ligated complex 8a is significantly diminished by added acid. When 2a is mixed with one equivalent of indole in the presence of a small amount (15 mole percent) of excess acid at room temperature, 6a is completely converted to 8a in minutes at room temperature, but conversion takes hours in the presence of 50 mole percent excess acid. In a more systematic set of experiments, 2a is prepared from 1a with a variable excess of acid, then five equivalents of indole is added, and reaction progress monitored by NMR at 40 °C; apparent rate constants for rearrangement of 6a are shown in Table 4. Only modest reductions of rate are observed up to about three-fold excess acid, but when the excess acid becomes comparable to the amount of excess indole the rate slows dramatically (entry 4). Under these conditions indole complexation to form 6a is also slowed, to only approximately three times the rate of rearrangement (compared to at least two orders of magnitude faster in the absence of acid). Furthermore, the initial product of π to N1 rearrangement is not 8a but a different N–adduct (presumably 7a); the rate of tautomerization of the first-formed N–adduct to the ultimate product 8a is slow under these acidic conditions.²⁴ Thus protonation of the nitrogen lone pair slows both coordination of platinum to the C=C double bond and subsequent migration of platinum to nitrogen; the inhibition of [1D] to [3D] indole tautomerization suggests that it is mediated by free (unprotonated) indole.

Table 4.

Acid dependence of rate constant for rearrangement 6a.

Entry	BF3	k (s−1)	t1/2 (min)
1	1.1 equiv.	1.24(16) × 10−2	0.94(12)
2	2.5 equiv.	1.09(21) × 10−2	1.09(19)
3	4.0 equiv.	3.51(12) × 10−3	3.29(11)
4a	5.0 equiv.	1.48(5) × 10−4	78.3(27)

Conditions: BF₃ is added to 1a in TFE. 5 equivalents of indole in TFE is then injected.

^aIndole binding and tautomerization are slow relative to k_RAR.

Conclusions

The new indole complexes generated by addition of indole, 1-methylindole and 2-methylindole to [(diimine)Pt(Me)(solv)]⁺ cations are very rare examples of stable complexes of this class of heterocycle with late transition metals. The initial interaction is not with the nitrogen lone pair, but rather indole C2=C3 olefin ligation, the first examples of such complexation. Indene forms a more stable π complex with cation 2a that undergoes subsequent activation of the indenyl C-H group in hours at 60 °C to afford methane and an η³-indenyl adduct. By contrast, indole π complexes of cation 2a rearrange to N-ligated species. The rate of rearrangement can be controlled by substitution of the indole or by added H⁺. Presumably, productive C-H activation chemistry on the indole nucleus is more likely to proceed from a complex of the indole C2=C3 olefin rather than N1; therefore strategies to prevent the formation of apparently inert N complexes from the corresponding π complexes are of interest to the continued development of synthetic strategies for functionalization of indoles by metal-mediated C-H activation, although we have not yet achieved the latter. Progress toward these objectives is ongoing in our laboratories.

Experimental Section

General Considerations

¹H NMR and ¹³C NMR spectra were recorded at ambient temperature using a Varian Inova 500, 600, or Mercury 300 spectrometer. The data are reported by chemical shift (ppm) from tetramethylsilane, multiplicity (s, singlet; d, doublet; t, triplet; m, multiplet; dd, double doublet; dt, double triplet), coupling constants (Hz), and integration. All ¹³C NMR data were collected proton-decoupled (¹³C{¹H}), except for those for 4a, as specified. Mass spectra were acquired on a Finnigan LCQ ion trap or Agilent 5973 Network mass selective detector and were obtained by peak matching. All reactions were carried out under an atmosphere of nitrogen in glassware that had been oven-dried. Unless otherwise noted, all reagents were commercially obtained and, where appropriate, purified prior to use. Tris(pentafluorophenyl)borane [B(C₆F₅)₃] was purified by sublimation (90 °C, 0.5 Torr). 2,2,2-Trifluoroethanol-d₃ (TFE-d₃) was dried over 3 Å molecular sieves for at least 5 days and then vacuum distilled onto B(C₆F₅)₃. After 6 h, the trifluoroethanol-d₃ was vacuum distilled and stored in a Teflon needle-valved vessel. Boron trifluoride was purified according to Brown’s procedure²⁵ and stored as a solution in TFE-d₃. The platinum dimethyl complexes were synthesized following earlier reported procedures, as noted. 2,2,2-Trifluoroethanol-d₃, B(C₆F₅)₃, and platinum dimethyl complexes were stored in a Vacuum Atmospheres dinitrogen atmosphere drybox.

Indene complex 3a

Platinum dimethyl complex 1a^4a (19.2 μmol, 10.5 mg) was weighed out in an oven-dried 4 mL vial in the drybox. TFE-d₃ (700 μL) and BF₃ (0.477 M in TFE-d₃, 18 μmol, 38 μL) were then added and the suspension was stirred until it became a homogeneous orange solution. Indene (19.2 μmol, 2.24 mg, 2.25 μL) was added, and the solution was transferred to an oven-dried J-Young NMR tube. After incubation at room temperature (24 hours in this case) the solution was analyzed by NMR and found to contain 3a in > 95% yield.

¹H NMR (600 MHz, TFE-d₃) δ: 7.51 (dt, J_H-H = 6.8 Hz (doublet), 1.5 Hz (triplet), 1H), 7.26-7.23 (m, 3H), 7.15 (s, 1H), 7.10 (s, 1H), 6.99 (s, 1H), 6.96 (s, 1H), 6.16 (d, J_H-H = 4 Hz, J_Pt-H = 79 Hz, 1H), 5.24 (apparent t, J_H-H ~ 6.4 Hz, J_Pt-H = 74 Hz, 1H), 3.30 (d, J_H-H = 23 Hz, J_Pt-H = 41 Hz, 1H), 3.07 (d, J_H-H = 23 Hz, J_Pt-H = 38 Hz, 1H), 2.32 (s, 6H), 2.25 (s, 3H), 2.20 (s, 3H), 2.13 (s, 3H), 2.12 (s, 3H), 2.07 (s, 3H), 1.97 (s, 3H), −0.74 (s, J_Pt-H = 53 Hz, 3H). ¹³C NMR (150 MHz, TFE-d₃) δ: 185.6, 178.7, 148.4, 142.7, 141.5, 141.0, 140.9, 139.7, 131.8, 131.6, 131.0, 131.0, 130.8, 130.5, 130.4, 129.5, 129.2, 128.9, 126.2, 125.5, 103.1, 92.4, 40.7, 21.1, 21.1, 20.4, 20.2, 17.9, 17.9, 17.9, 17.8, 4.7. FTIR (neat): 2920 (w), 1476 (w), 1384 (w), 1237 (w), 1180 (m), 1160 (m), 1051 (s), 847 (w), 766 (w), 718 (w) cm⁻¹. FAB⁺ MS, calcd for C₃₂H₄₀N₂Pt ([M + H]⁺): 647.2840, found 647.2822. Calcd for C₃₁H₃₆N₂Pt ([M-Me]⁺): 631.2526, found 631.2302.

Indene complex 3a-d₃ was prepared as above. ¹H NMR (500 MHz, TFE-d₃) matches the data above, except that signals at δ = 6.16, 3.30, and 3.07 ppm are absent.

Indene complexes 3b, 3c, 3d

Complexes 3b, 3c, 3d were prepared by direct analogy to complex 3a above, starting from platinum dimethyl species 1b, 1d^4a and 1c.³ 3b: ¹H NMR (300 MHz, TFE-d₃) δ: 7.51 (d, 7.23 Hz, 1H), 7.29-7.19 (m, 3H), 7.02 (broad s, 2H), 6.57 (broad s, 4H), 6.46 (broad s, 1H), 5.64 (broad s, 1H), 3.20 (broad s, 2H), 2.32 (broad s, 12H), 2.07 (broad s, 6H), −0.59 (s, J_Pt-H = 71 Hz, 3H). 3c: ¹H NMR (300 MHz, TFE-d₃) δ: 7.96 (s, 2H), 7.56-7.52 (m, 1H), 7.53 (s, 4H), 7.29-7.11 (m, 3H), 6.53 (s, 1H), 5.68 (s, 1H), 3.22 (s, 2H), 2.12 (s, 6H), −0.34 (s, = 69 Hz, 3H). 3d: ¹H NMR (300 MHz, TFE-J_Pt-H d₃) δ: 7.52-7.47 (m, 3H), 7.31-7.20 (m, 3H), 6.84 (broad s, 4H), 6.33 (broad s, 1H), 5.54 (broad s, 1H), 3.13 (broad s, 2H), 2.11 (broad s, 6H), 1.33 (broad s, 36H), −0.66 (s, J_Pt-H = 69 Hz, 3H).

Eyring Analysis for the conversion of 3a to 4a

3a was prepared as a solution in TFE-d₃ as described above, with the exception that 2.50 eq of B(C₆F₅)₃ was used as the acid source. Conversion and rate were then recorded as described in the main text.

Indenyl complex 4a

The solution of 3a prepared above was heated in its J-Young tube to the temperature and time described in the main text. ¹H NMR (500 MHz, TFE-d₃) δ: 7.12-7.10 (m, 6H), 6.79 (dd J_H-H = 5.7, 3.1 Hz, ~3H), 5.01 (s,²⁶ 2H), 2.40 (s, 6H), 2.12 (s, 6H), 1.97 (s, 12H). ¹³C NMR (150 MHz, TFE-d₃) δ: 172.6 (2C), 147.6 (2C), 140.8 (2C), 134.6, 131.2 (4C), 129.9 (4C), 129.3 (4C), 120.0 (2C), 71.6 (J_Pt-C = 130 Hz, 2C), 21.2 (2C), 18.0 (2C), 17.6 (4C). ¹³C NMR (125 MHz, not {1H}, TFE-d₃) δ: 172.5 (s, 2C), 147.6 (s, 2C), 140.8 (d, J_C-H = 6 Hz, 2C), 134.6, 132-130 (m, J_1,C-H ~ 150 Hz, 2C), 131.2 (d, J_C-H = 152 Hz, 2C), 129.9 (d, J_C-H = 154 Hz, 4C), 129.3 (s, 4C), 120.0 (d, J_C-H = 163 Hz, 2C), 71.6 (d, J_C-H = 181 Hz, 2C), 21.2 (q, J_CH = 126 Hz, 2C), 18.0 (q, J_C-H = 132 Hz, 2C), 17.6 (q, J_C-H = 126 Hz, 4C). FTIR (neat): 2919 (w), 1476 (w), 1331 (w), 1385, 1244 (w), 1182 (m), 1157 (m), 1054 (s), 851, 754 (w) cm⁻¹. FAB⁺ MS, calcd for C₃₁H₃₆N₂Pt ([M + H]⁺): 631.2526, found 631.2530 l. ²H NMR (76.7 MHz, TFE-h₃) for this compound is shown in the Supporting Information.

Indenyl complex 4a-d₂ was prepared as above. ¹H NMR (500 MHz, TFE-d₃) matches the data above, except that the signal at δ = 5.01 ppm is absent. FAB⁺ MS, calcd for C₃₁H₃₄D₂N₂Pt ([M-d₂ + H]⁺): 633.2652, and for C₃₁H₃₂D₃N₂Pt ([M-d₃]⁺): 633.2637; found 33.2631. ²H NMR (76.7 MHz, TFE-h₃) for this compound is shown in the Supporting Information.

Indole π-complex 6a

Complex 6a is observed as an intermediate in kinetic runs (Table 4, entry 4). ¹H NMR (500 MHz, TFE-d₃) δ: 7.62 (s, 1H), 7.60 (d, J_H-H = 7.8 Hz, 1H), 7.25-7.22 (m, 2H), 7.17 (t, J_H-H = 8.0 Hz, 1H), 7.14 (s, 1H), 7.01 (s, 1H), 6.97 (s, 1H), 6.94 (s, 1H), 5.40 (s, J_H-H = 94 Hz, ~0H),²⁷ 2.39 (s, 3H), 2.31 (s, 3H), 2.25 (s, 3H), 2.18 (s, 3H), 2.13 (s, 3H), 2.03 (s, 3H), 1.83 (s, 3H), 1.81 (s, 3H), −1.02 (s, J_Pt-H = 62 Hz, 3H).

Indole N-complex 8a

Platinum dimethyl complex 1a^4a (46.6 μmol, 25.4 mg) was weighed out in an oven-dried 4 mL vial in the drybox. TFE-d₃ (350 μL) and BF₃ (0.455 M in TFE-d₃, 53.5 μmol, 118 μL) were then added and the suspension was stirred until it became a homogeneous orange solution. Indole (47 μmol, 5.5 mg) was added, and the solution was transferred, rinsing with TFE-d₃ (350 μL) to an oven-dried J-Young NMR tube. The solution was analyzed by NMR and found to contain 8a in > 95% yield. ¹H NMR (500 MHz, TFE-d₃) δ: 7.89 (s, J_Pt-H = 34 Hz, 1H), 7.73 (d, J_H-H = 7.3 Hz, 1H), 7.40-7.35 (m, 3H), 7.12 (s, 1H), 7.09 (s, 1H), 6.97 (s, 1H), 6.50 (s, 1H), 2.35 (s, 3H), 2.34 (s, 3H), 2.27 (s, 3H), 2.20 (s, 3H), 2.16 (s, 3H), 1.88 (s, 3H), 1.87 (s, 3H), 1.80 (s, 3H), 0.61 (s, J_Pt-H = 73 Hz, 3H). ¹³C NMR (125 MHz, TFE-d₃) δ: 182.1, 175.6, 175.1, 153.9, 142.7, 140.9, 140.4, 139.8, 134.5, 130.9, 130.9, 130.8, 130.7, 130.6, 130.5, 130.1, 130.0, 129.3, 128.9, 125.8, 122.3, ~44 (m, 1C), 21.1, 20.8, 20.0, 19.5, 17.9, 17.8 (2C), 17.7, −12.1. FTIR (neat): 2921 (m), 2094 (w), 1474 (m), 1449 (m), 1382 (m), 1313 (m), 1240 (m), 1184 (s), 1128 (s), 1069 (s), 1018 (s), 852 (m), 818 (m), 769 (m), 644 (m), 534 (m) cm⁻¹. FAB⁺ MS, calc’d for C₃₁H₃₅N₃Pt²H₂ ([M-H]⁺): 649.2839, found 649.2847.

Indole complex 8b

Platinum dimethyl complex 1b^4a (32.5 μmol, 16.8 mg) was weighed out in an oven-dried 4 mL vial in the drybox. TFE-d₃ (700 μL) and BF₃ (0.478 M in TFE-d₃, 37.3 μmol, 78.1 μL) were then added and the suspension was stirred until it became a homogeneous orange solution. Indole (33 μmol, 3.8 mg) was added, and the solution was transferred to an oven-dried screw-capped NMR tube. The solution was analyzed by NMR and found to contain 8b in > 95% yield.

¹H NMR (500 MHz, TFE-d₃) δ: 8.01 (s, J_Pt-H = 36 Hz, 1H), 7.87 (d, J_H-H = 7.9 Hz, 1H), 7.46-7.41 (m, 2H), 7.38-7.35 (m, 1H), 7.06 (s, 1H), 6.75 (s, 1H), 6.71 (s, 1H), 6.66 (s, 1H), 6.49 (s, broad, 1H), 6.06 (s, broad, 1H), 2.39 (s, 3H), 2.37 (s, 3H), 2.25 (s, broad, 3H), 1.96 (s, 3H), 1.95 (s, 3H), 1.69 (s, 3H), 0.74 (s, J_Pt-H = 73 Hz, 3H). ¹³C NMR (125 MHz, TFE-d₃) δ: 182.3, 174.6, 175.1, 153.9, 147.8, 145.7, 141.9, 141.8, 141.9-141.5 (m, broad, 2C), 134.5, 131.2, 130.5, 130.2, 129.4, 125.9, 122.7, 121.0, 121.0, 120.7 (broad, 1C), 119.5 (broad, 1C), ~44 (m, 1C), 21.6, 21.5, 21.2, 21.9-21.8 (m, 2C), −11.3. FTIR (neat): 3014 (w), 2921 (w), 2875 (w), 1609 (w), 1593 (w), 1468 (w), 1453 (w), 1384 (w), 1308 (w), 1185 (m), 1130 (m), 1060 (s), 863 (w), 771 (w), 685 (w) cm⁻¹. FAB⁺ MS, calcd for C₂₉H₃₁N₃Pt²H₂ ([M-H]⁺): 620.2448, found 620.2459.

Indole complex 8c

Platinum dimethyl complex 1c³ (36.0 μmol, 26.4 mg) was weighed out in an oven-dried 4 mL vial in the drybox. TFE-d₃ (700 μL) and BF₃ (0.478 M in TFE-d₃, 41.4 μmol, 86.6 μL) were then added and the suspension was stirred until it became a homogeneous orange solution. Indole (36 μmol, 9.2 mg) was added, and the solution was transferred to an oven-dried screw-capped NMR tube. The solution was analyzed by NMR and found to contain 8c in > 95% yield.

¹H NMR (500 MHz, TFE-d₃) δ: 8.27 (s, J_Pt-H ~ 40 Hz, 1H), 8.02 (s, 1H), 7.78 (d, J_H-H = 8.5 Hz, 1H), 7.73 (s, 1H), 7.68 (s, 1H), 7.55 (S, 1H), 7.45-7.27 (m, 5H), 2.04 (s, 3H), 2.03 (s, 3H), 0.76 (s, J_Pt-H ~ 60 Hz, 3H). ¹³C NMR (125 MHz, TFE-d₃) δ: 184.5, 177.7, 175.6 (m, 1C), 152.9 (m, 1C), 148.5, 147.0, 136.1-134.1 (m, apparent 4C), 134.1, 130.7, 129.4, 126.2, 124.6 (apparent 2C), 124-122 (m, apparent 4C), 122.3 (q, J_C-F = 227 Hz), 122.0, ~44 (m, 1C), 21.9, 20.5, −9.4. Three CF₃ quartets cannot be identified because of coincidences with the solvent CF₃; each has δ ~ 126 ppm, J_C-F ~280 Hz. FTIR (neat): 3065 (w), 2932 (w), 2882 (w), 1622 (w), 1460 (m), 1373 (s), 1281 (s), 1182 (s), 1141 (s), 1067 (m), 903 (m), 848 (w), 771 (w), 705 (w), 684 (m) cm⁻¹. FAB⁺ MS, calcd for C₂₉H₁₉F₁₂N₃Pt²H₂ ([M-H]⁺): 836.1317, found 836.1333.

2-Methylindole-π-complex 9

Platinum dimethyl complex 1a^4a (33.4 μmol, 18.2 mg) was weighed out in an oven-dried 4 mL vial in the drybox. TFE-d₃ (700 μL) and BF₃ (0.455 M in TFE-d₃, 38.4 μmol, 84.3 μL) were then added and the suspension was stirred until it became a homogeneous orange solution. 2-Methylindole (33 μmol, 4.4 mg) was added, and the solution was transferred to an oven-dried screw-capped NMR tube. The solution was analyzed by NMR and found to contain 9 in > 95% yield.

¹H NMR (500 MHz, TFE-d₃) δ: 7.35 (d, J_H-H = 7.5 Hz, 1H), 7.29 (d, J_H-H = 7.5 Hz, 1H), 7.25 (s, 1H), 7.20-7.14 (m, 3H), 6.97 (s, 1H), 6.90 (s, 1H), 2.53 (s, 3H), 2.43 (s, 3H), 2.35 (s, 3H), 2.29 (s, 3H), 2.25 (s, 3H), 2.10 (s, 3H), 1.90 (s, 3H), 1.83 (s, 3H), 1.63 (s, 3H), −0.75 (s, J_Pt-H = 67, 3H). ¹³C NMR (125 MHz, TFE-d₃) δ: 178.0, 176.8, 143.3, 142.7, 141.9, 141.5, 140.2, 139.7, 131.5, 131.4, 130.7, 130.6, 130.5, 130.4, 130.1, 129.4, 126.7, 126.1, 123.9, 114.1, 21.1, 21.0, 20.2, 19.9, 18.1, 18.0, 17.5 (2C), −3.6. (3 C cannot be uniquely identified because of dispersity in the aryl region and the transitory nature of 9.) FTIR (neat): 2918 (w), 1612 (w), 1505 (w), 1479 (w), 1458 (w), 1383 (w), 1329 (w), 1238 (m), 1157 (m), 1059 (s), 848 (w), 742 (m), 571 (w), 534 (w) cm⁻¹. 9 is not sufficiently long-lived for HRMS.

2-Methylindole-N-complex 10

The crude solution of 9 prepared above was incubated at room temperature for 50 hours to give 10 in > 95% yield.

¹H NMR (500 MHz, TFE-d₃) δ: 7.74 (d, J_H-H = 7.8 Hz, 1H), 7.32 (d, J_H-H = 7.9 Hz, 1H), 7.29-7.27 (m, 2H), 7.11 (s, 1H), 7.10 (s, 1H), 6.87 (s, 1H), 6.44 (s, 1H), 2.41 (s, 3H), 2.35 (s, 3H), 2.33 (s, 3H), 2.24 (s, 3H), 2.19 (s, 3H), 2.12 (s, 3H), 1.86 (s, 3H), 1.85 (s, 3H), 1.83 (s, 3H), 0.53 (s, J_Pt-H = 74, 3H). ¹³C NMR (125 MHz, TFE-d₃) δ: 186.0, 181.7, 175.8, 154.9, 142.8, 142.0, 140.3, 139.6, 135.0, 130.9, 130.7, 130.4, 129.9, 128.9, 128.8, 128.7, 122.5, 121.6, 121.1, 121.0, 112.4, ~45 (m, 1C), 22.1, 21.1, 20.8, 20.1, 19.5, 18.4, 17.9, 17.8, 13.5, −13.7. FTIR (neat): 3512 (broad, w), 3399 (broad, w), 2921 (broad, w), 1613 (w), 1573 (w), 1476 (m), 1456 (m), 1384 (m), 1316 (m), 1240 (m), 1187 (s), 1130 (s), 1069 (s), 852 (w), 756 (w), 646 (w), 535 (w) cm⁻¹. FAB⁺ MS, calc’d for C₃₂H₃₈N₃Pt²H₂ ([M]⁺): 663.2996, found 663.2980.

1-Methylindole- π-complex 11

Platinum dimethyl complex 1a^4a (46.9 μmol, 25.6 mg) was weighed out in an oven-dried 4 mL vial in the drybox. TFE-d₃ (700 μL) and BF₃ (0.455 M in TFE-d₃, 54 μmol, 119 μL) were then added and the suspension was stirred until it became a homogeneous orange solution. 1-Methylindole (46.9 μmol, 6.16 mg, 6.00 μL) was added, and the solution was transferred to an oven-dried screw-capped NMR tube. The solution was analyzed by NMR and found to contain 11 in > 95% yield.

¹H NMR (500 MHz, TFE-d₃) δ: 7.57 (d, J_H-H = 8.7 Hz, 1H), 7.50 (s, J_Pt-H = 49 Hz, 1H), 7.28 (t, J_H-H = 7.8 Hz, 1H), 7.22-7.17 (m, 2H), 7.16 (s, 1H), 7.07 (s, 1H), 6.97 (s, 1H), 6.91 (s, 1H), 5.95 (s, J_H-H = 88 Hz, ~0H),²⁷ 3.65 (s, 3H), 2.41 (s, 3H), 2.32 (s, 3H), 2.24 (s, 6H), 2.14 (s, 3H), 2.00 (s, 3H), 1.86 (s, 3H), 1.80 (s, 3H), −1.16 (s, J_Pt-H = 70 Hz, 3H). ¹³C NMR (125 MHz, TFE-d₃) δ: 179.4, 177.1, 144.9, 142.6, 141.5, 140.4, 140.1, 137.3, 131.7, 131.5, 130.8, 130.7, 130.6, 130.2, 130.1, 129.8, 125.2, 125.0, 128.9, 128.3, 111.9, 34.2, 21.1, 21.0, 20.0, 19.8, 18.0, 17.7 (2C), 17.6, −3.0. Because of dispersion in the 130 ppm region, 1C cannot be located. FTIR (neat): 2919 (w), 1477 (w), 1454 (w), 1383 (w), 1327 (w), 1238 (w), 1184 (m), 1128 (m), 1054 (s), 847 (w), 753 (w), 679 (w) cm⁻¹. FAB⁺ MS, calc’d for C₃₂H₃₉N₃Pt²H₁ ([M]⁺): 662.2933, found 662.2932.

1-Methylindole-N-complex 12

The crude solution of 11 prepared above was incubated at room temperature for 2 days, then heated to 40 °C for 4 hours to give 12 in ~ 78% yield by NMR.

¹H NMR (500 MHz, TFE-d₃) δ: 7.32 (t, J_H-H = 7.6 Hz, 1H), 7.28 (d, J_H-H = 7.4 Hz, 1H), 7.22 (t, J_H-H = 7.4 Hz, 1H), 7.19 (d, J_H-H = 8.4 Hz, 1H), 7.08 (s, 1H), 7.07 (s, 1H), 6.91 (s, 1H), 6.69 (s, 1H), 3.83 (s, 3H), 2.33 (s, 3H), 2.30 (s, 3H), 2.24 (s, 3H), 2.18 (s, 3H), 2.13 (s, 3H), 2.01 (s, 3H), 1.98 (s, 3H), 1.85 (s, 3H), 0.26 (s, J_Pt-H = 77 Hz, 3H). ¹³C NMR (125 MHz, TFE-d₃) δ: 180.1, 176.3, 148.4, 144.6, 142.0, 139.9, 139.7, 137.0, 136.9, 130.9, 130.7, 130.7, 130.4, 130.2, 129.3, 128.8, 128.8, 128.6, 127.6, 125.6, 125.5, 112.7, 39.9 (J_Pt-C = 81 Hz), 21.1, 20.7, 20.0, 19.8, 17.9, 17.9, 17.9, 17.8, −16.8. FTIR (neat): 2923 (m, broad), 2100 (w), 1609 (w), 1515 (m), 1463 (s), 1384 (m), 1312 (m), 1238 (m), 1186 (s), 1130 (s), 1069 (s, broad), 860 (w), 818 (w), 759 (m), 646 (w), 535 (w) cm⁻¹. FAB⁺ MS, calcd for C₃₂H₃₉N₃Pt²H ([M]⁺): 662.2933, found 663.2939.

Crossover experiment between 9 and 1-methylindole

Platinum dimethyl complex 1a (22.4 μmol, 12.2 mg) was weighed out in an oven-dried screw-capped NMR tube in the drybox. TFE-d₃ (511 μL) and BF₃ (0.477 M in TFE-d₃, 24.6 μmol, 51.6 μL) were then added. The suspension was removed from the drybox and sonicated until it became a homogeneous orange solution. The reaction was then thermally equilibrated to 30 °C in a Varian mercury 300 MHz NMR, and 2-methylindole (0.183 M in TFE-d₃, 22.4 μmol, 122 μL) was then injected. [9] was monitored, and reached a maximum in ca. 2 min. at which conversion of 2a to 9 was ~90% complete. The tube was then removed from the instrument, 1-methylindole was injected, and the sample was re-inserted (~ 30 sec.) An initial ratio of 81(1)% 9, 10(1)% 10, 9(1)% 11 was observed with 0(1)% 2a. Kinetic data were recorded for 9.7 hours. After an additional 2 days at RT, 9 was completely converted to 10 (k_RAR = 9.1(3) × 10⁻⁵ sec⁻¹) and 11 was completely converted to 12. A ratio of 91(1)% 10 to 9(1)% 12 was observed.

Conversion of 6a to 8a at high [H⁺]

Platinum dimethyl complex 1a (22.7 μmol, 12.4 mg) was weighed out in a 1 mL graduated cylinder in the drybox. BF₃ (0.477 M in TFE-d₃, 56.8 μmol, 119 μL) and TFE-d₃ (to make 1.00 mL) were then added. 480 μL of this solution (10.9 μmol Pt, 2.5 equiv. BF₃) was then distributed into each of 2 screw-capped NMR tubes. Separately, indole (14.5 mg, 124 μmol), BF₃ (0.477 M in TFE-d₃, 62.0 μmol, 130 μL). and TFE-d₃ (356 μL) were mixed in a 10 mL culture tube. 220 μL of this solution (5 equiv. indole to Pt, 2.5 equiv. BF₃ to Pt) was then distributed into each of two 1 mL syrenges. The NMR tubes and syrenges were removed from the drybox. One NMR tube was then thermally equilibrated to 40 °C in a Varian inova 500 MHz NMR spectrometer. It was then ejected, and the content of 1 syrenge was added to the NMR tube (to make 10.9 μmol Pt, 5 equiv. BF₃, 5 equiv. indole, and 700 μL TFE-d₃). The tube was shaken and re-inserted. After rapid re-shimming, kinetic data were recorded for 7.5 hours. A second trial was then conducted with the other NMR tube and syrenge. Reaction progress was monitored by NMR. Rate constants for this experiment are k_binding = 5.35_10⁻⁴ s⁻¹ and k_RAR = 1.44_10⁻⁴ s⁻¹, where k_binding is the pseudo first order rate constant for the binding of indole to 2a.²⁴

Comment regarding H/D Exchange and Methane Isotopologs Formed in the Conversion of 3 to 4

In the conversion of 3a to 4a the methane that is formed is 65% CH₄, 25% CH₃D due to D⁺ incorporation in the formation of 2a from 1a. A distribution of 53% CH₄, 47% CH₃D is observed in the conversion of 3a-d₃ to 4a-d₂. ²H NMR spectra of 4a and 4a-d₂ were recorded to track ²H incorporation in these compounds (see Supporting Information). ²H incorporation is observed in C2-H (6.8 ppm) in both 4a and, 4a-d₂. Little ²H incorporation is observed in C1,3-H (5.0 ppm) in 4a. As expected, ²H is observed in this position in 4a-d₂. A small amount of H/D exchange with solvent is observed in the diimine backbone (2.4 ppm), but none is observed in the mesityl methyl groups. Accordingly, the high level of CH₄ relative to CH₃D observed in the formation of 4a-d₂ is attributed to some H/D exchange with C2—H in the indenyl fragment prior to methane liberation and a low level of decomposition of the diimine ligand.

Supplementary Material

si20060926_021. Supporting Information Available.

Spectral data for the crossover experiment between 9 and 1-methylindole, spectral and kinetic data for the conversion of 6a to 8a in the presence of excess acid, ²H NMR spectra for 4a and 4a-d₂, and graphical ¹H NMR spectra for 3a, 3a-d₃, 4a, 4a-d₂, 8a, 8b, 8c, 9, 10, 11, and 12 are available free of charge via the Internet at http://pubs.acs.org.