The synthesis and exchange chemistry of frustrated Lewis pair–nitrous oxide complexes

Abstract

Facile activation of nitrous oxide is achieved using a series of fluoroarylboranes, B(C₆F₅)₃, PhB(C₆F₅)₂, MesB(C₆F₅)₂, (C₆F₅)₂BOC₆F₅, B(C₆F₄-p-H)₃, B(C₆H₄-p-F)₃ and 1,4-(C₆F₅)₂BC₆F₄B(C₆F₅)₂ in the presence of the basic, bulky phosphine^tBu₃P. Room temperature reaction yields mono- and bis-zwitterionic species of the form ^tBu₃P(N₂O)B(C₆F₅)₂R (R = C₆F₅1, Ph 2, Mes 3, OC₆F₅4), ^tBu₃P(N₂O)BR₃ (R = C₆F₄-p-H 5, C₆H₄-p-F 6) and ^tBu₃P(N₂O)B(C₆F₅)₂C₆F₄(C₆F₅)₂B(ON₂)P^tBu₃7. N₂O activation is similarly achieved using Cy₃P and B(C₆F₄-p-H)₃ yielding the zwitterionic species Cy₃P(N₂O)B(C₆F₄-p-H)₃8. Reaction of 6 with [Ph₃C][B(C₆F₅)₄] results in facile transfer of the robust ^tBu₃P(N₂O) fragment to the stronger Lewis acid Ph₃C⁺ generating [^tBu₃P(N₂O)CPh₃][B(C₆F₅)₄] 10. Similarly exchange reactions with titanocene and zirconocene cations generate transition metal and phosphine stabilized nitrous oxide salts, of the form [^tBu₃P(N₂O)MCp₂Me][MeB(C₆F₅)₃], (M = Zr 11, Ti 12). The alkoxy zirconocene cation [Cp*₂Zr(OMe)]⁺ forms an FLP in the presence of ^tBu₃P which reacts with N₂O providing a direct synthetic route to the corresponding salt [^tBu₃P(N₂O)ZrCp*₂(OMe)][B(C₆F₅)₄] 13. Kinetic studies of the self-exchange reaction of ^tBu₃P(N₂O)B(C₆H₄-p-F)₃ with B(C₆H₄-p-F)₃ were carried out acquiring information regarding the mechanism of exchange.

---

Introduction

The issues of ozone depletion and climate change have garnered a considerable amount of attention over the past century due to their apparent negative impact on the environment.¹ Anthropogenic emissions of potent greenhouse gases, such as carbon dioxide (CO₂), methane (CH₄) and nitrous oxide (N₂O), continue to rise and contribute to the phenomenon of global warming and to the destruction of the stratospheric ozone layer.^1,2 Considerable attention has focused on CO₂ and its contribution to global warming, while nitrous oxide (N₂O) has remained less recognized despite being roughly 300 times more potent a greenhouse gas than CO₂.²N₂O is inherently stable and can persist for upwards of 150 years in the stratosphere.³ Recent reports indicate that levels of N₂O have risen steadily since industrialization from roughly 270 to 319 ppb, as of 2005, and continue to rise annually.⁴ In fact, anthropogenic N₂O has recently been identified as the largest global ozone depleting gas emission.¹

In light of the effects of this greenhouse gas, methods to sequester and subsequently break down N₂O into less harmful components, such as N₂ and H₂O, are desirable. In nature, N₂O is transformed by metalloenzymes bearing an active site composed of a tetranuclear copper cluster interconnected by a sulfide ligand (Fig. 1).⁵ The metalloenzyme nitrous oxide reductase is an essential component of the microbial denitrification process, which is a key component of the nitrogen cycle.^3,5a,6 This enzyme in bacteria is responsible for catalyzing the transformation of N₂O to N₂ and H₂O and is key to the destruction of N₂O produced by the agricultural sector.⁴

Fig. 1 Resonance structures of N₂O (A), Taube's Ru–N₂O complex (B),⁸ Hillhouse's N₂O insertion product into Cp*₂Ti(C₂Ph₂) (C),¹⁰ and the Cu₄S active site in nitrous oxide reductase (D).¹¹

In synthetic chemistry, N₂O has been shown to be a strong oxidant yet also a kinetically stable species.⁷ In coordination chemistry, N₂O has been identified as an inherently poor ligand due to its inability to act as either a good σ-donor or π-acceptor (the most important resonance structures are shown in Fig. 1).⁷ The first example of a transition metal bound N₂O complex, [Ru(NH₃)₅N₂O]⁺, was proposed by Armor and Taube in 1969 (Fig. 1).⁸ Although this species was never crystallographically confirmed, a number of both computational and spectroscopic studies have since supported this formulation.⁹ Interestingly, few transition metal–N₂O species have been reported in the literature since Armor and Taube's discovery in the late sixties,⁷ and only recently has an example of a species that contains N₂O bound between a phosphine and a Zn center been crystallographically characterized.¹²N₂O is known to undergo reactions at transition metal centers including oxygen insertion into metal–carbon (Fig. 1) and metal–hydride bonds with concomitant N₂ liberation^10,13 and the oxidation of low-valent metal centers.¹⁴ Alternatively N–N scission can be achieved in the presence of metal centers¹⁵ while hydrogenation of N₂O yields N₂ and H₂O.¹⁶ Very recently, nickel(II) carbene complexes have been shown to react with N₂Ovia1,3-dipolar cycloaddition to result in C=O bond formation with liberation of N₂.¹⁷ While it is suspected that these reactions proceed through an initial N₂O–metal adduct, such claims are yet to be unequivocally substantiated.

Recent investigations of the interaction of bulky phosphines with boranes have given rise to the concept of frustrated Lewis pairs (FLPs).¹⁸ These systems encompass sterically hindered acids and bases that are precluded from classical Lewis acid–base adduct formation. The resulting unquenched acidity and basicity offer the possibility for further reactivity with a variety of small molecules, such as H₂, olefins, dienes, alkynes and carbon dioxide. Such activations have been recently reviewed.^18c Herein, the interaction of FLPs with N₂O is shown to result in binding of the N₂O molecule between the Lewis acidic and basic components in a P–N–N–O–B mode. In addition, the use of Lewis acid exchange chemistry is explored as a route to new N₂O compounds. The initial results of this work were communicated previously.¹⁹

Results and discussion

Complexation of N₂O by FLPs

Treatment of frustrated Lewis pairs consisting of an equimolar mixture of the sterically demanding Lewis base ^tBu₃P and the Lewis acidic boranes R₂BR′ with an atmosphere of N₂O in a bromobenzene solution resulted in the formation of novel zwitterionic species over a 12 h period at ambient temperatures (Scheme 1).

Scheme 1 Synthetic method for the generation of 1–6.

Precipitation with hexanes or pentane afforded the zwitterionic species ^tBu₃P(N₂O)BR₂R′ (R = R′ = C₆F₅ (1); R = C₆F₅, R′ = Ph (2); R = C₆F₅, R′ = Mes (3); R = C₆F₅, R′ = OC₆F₅ (4); R = R′ = C₆F₄-p-H (5); R = R′ = C₆H₄-p-F (6)) as analytically pure white microcrystalline solids in yields ranging from 76–91%. In an analogous manner, reaction of the linked bisborane (C₆F₅)₂B(C₆F₄)B(C₆F₅)₂ with two equivalents of ^tBu₃P under an atmosphere of N₂O gave the bis-zwitterionic compound ^tBu₃P(N₂O)B(C₆F₅)₂C₆F₄(C₆F₅)₂B(ON₂)P^tBu₃ (7) in good isolated yield (Scheme 2).

Scheme 2 Reaction of the C₆F₄-linked bisborane with two equivalents of ^tBu₃P and N₂O.

Conversely, attempts were made to isolate the corresponding N₂O compound from reactions of ^tBu₃P with the weak Lewis acid BPh₃ and N₂O. Although the desired product was formed in low yield, the white powder isolated upon precipitation with hexanes was ∼90% pure by NMR spectroscopy (Fig. S1) and attempts to further purify the material invariably led to increasing amounts of decomposition. This suggests that a threshold of Lewis acidity is required to generate stable compounds of this type. NMR spectroscopic analyses of compounds 1–7 in CD₂Cl₂ showed a single resonance in the ³¹P{¹H} NMR spectra between 64 and 69 ppm. ¹¹B{¹H} NMR spectra demonstrated a sharp singlet between 0.67 and 6.69 ppm. For species containing C₆F₅ groups, a significant narrowing of the meta–para gap in comparison to the free boranes was observed in the ¹⁹F NMR spectra, with Δ(δ(m-F)–δ(p-F)) around 5 ppm.These data are consistent with 4-coordinate boron centers.²⁰ Compound 4 bears two distinct C₆F₅ environments (C₆F₅ and OC₆F₅) both of which display meta–para gap narrowing upon the generation of the anionic boron center. Upfield ¹⁹F NMR shifts are noted for 5 (−134.27/−143.10 ppm) and 6 (−120.87 ppm) relative to the free boranes (−129.60/−138.74 and −106.00 ppm, respectively). The synthesis of the ¹⁵N₂O isotopologues 1-¹⁵15N–7-¹⁵15N by reaction of the FLPs with ¹⁵N¹⁵NO further supported the generation of the P–N–N–O–B species.

The observation of two doublets of doublets in the ¹⁵N NMR spectra at roughly 377 and 577 ppm for the terminal and central N atom, respectively, is indicative of two inequivalent and distinct nitrogen environments. ¹J_N–N coupling constants were found to be approximately 16 Hz, while ¹J_N–P and ²J_N–P were determined to be approximately 60 and 20 Hz, respectively. Comparison of the ¹⁵N spectroscopic data of the bound N₂O fragment with free N₂O (135 and 218 ppm, ¹J_N–N = 8.1 Hz) indicates that there is a significant perturbation of the N₂O fragment upon complexation by the borane and phosphine moieties (Fig. 2). A comparison of the ¹⁵N NMR shifts and N–N/N–P coupling constants for compounds 1–7 (Table 1) shows no significant difference upon the reduction of the Lewis acidity of the borane fragment.

Fig. 2 ³¹P{¹H} and ¹⁵N NMR spectra for 4-¹⁵15N.

Table 1 ¹⁵N NMR chemical shifts and coupling constants for compounds 1-¹⁵15N to 11-¹⁵15N and 13-¹⁵15N

	1	2	3	4	5	6	7	8	10	11	13
δ N(1) (ppm)	381.7	377.0	375.3	389.4	367.6	381.3	382.3	376.1	412.3	398.5	386.3
δ N(2) (ppm)	566.6	577.7	574.2	572.5	588.8	571.0	572.1	573.0	548.4	587.6	591.0
^1 J N–N (Hz)	15.6	16.0	15.1	15.8	16.4	15.2	15.5	15.6	14.7	17.1	17.4
^1 J P–N (Hz)	58.7	59.5	59.1	58.6	59.7	58.8	58.8	52.5	58.2	60.6	61.6
^2 J P–N (Hz)	19.6	19.1	19.6	19.7	19.1	19.9	19.6	22.2	20.5	15.8	16.1

---

With the exception of 6, all compounds could be recrystallized by diffusion of pentane or cyclohexane into a CH₂Cl₂ or C₆H₅Br solution at room temperature. Crystal structure determinations were carried out for these compounds, unambiguously confirming their formulation, although the quality of the data for compounds 3 and 4 was quite poor and only served to establish connectivities. The molecular structure of 7 is shown in Fig. 3. Comparison of the relevant metrical parameters (Table 2) reveals no significant variation of the bound N₂O fragment regardless of the nature of the borane.

Fig. 3 POV-ray depiction of the molecular structure of 7. Hydrogen atoms have been omitted for clarity.

Table 2 Metrical parameters for 1, 2, 5, 7 and 8^a

	1	2	5	7	8
a Bond lengths in angstroms (Å), angles in degrees (°).
P–N(1)	1.7087(12)	1.7107(6)	1.7092(16)	1.6905(16)	1.7146(15)
N(1)–N(2)	1.2573(17)	1.2602(8)	1.255(2)	1.253(2)	1.260(2)
N(2)–O	1.3363(15)	1.3270(8)	1.327(2)	1.3198(19)	1.3351(18)
O–B	1.5428(18)	1.5475(9)	1.533(2)	1.536(2)	1.551(1)
P–N(1)–N(2)	117.04(10)	112.85(5)	115.32(13)	116.48(14)	111.10(12)
N(1)–N(2)–O	109.11(11)	111.68(6)	109.49(14)	109.81(15)	111.13(14)
N(2)–O–B	114.43(10)	111.61(5)	114.86(14)	111.26(13)	110.32(13)

---

From the data presented above, it is clear that N₂O complexes can be made from combinations of ^tBu₃P with various boron containing Lewis acids demonstrating that a wide range of Lewis acidities is tolerated. It was thus of interest to examine a series of Lewis basic FLP partners for their ability to form analogous N₂O complexes. Both phosphorus (Mes₃P, (o-tolyl)₃P, HP^tBu₂) and nitrogen bases (2,2,6,6-tetramethylpiperidine, 2,6-lutidine, N-benzylidine-tert-butylamine) were tested in combination with B(C₆F₅)₃, as such FLPs have been shown to effect heterolytic cleavage of H₂.^18c However, in all cases, treatment with N₂O in bromobenzene failed to generate the analogous N₂O complexes. Evidently, N₂O activation is strongly dependent on the nature of the base suggesting that both strong σ-donation and π-acceptance stabilize the PN₂O fragment. Phosphines that are less sterically demanding than ^tBu₃P either form strong adducts with Lewis acids such as B(C₆F₅)₃, or result in nucleophilic aromatic substitution at the para position of a C₆F₅ ring, as is the case with Cy₃P.²¹ Previously, it has been shown that the borane B(C₆F₄-p-H)₃ is immune to attack at the para-carbon atom, forming an FLP with Cy₃P.²² The reaction of an equimolar mixture of Cy₃P and B(C₆F₄-p-H)₃ in bromobenzene with an atmosphere of N₂O results in formation of Cy₃P(N₂O)B(C₆F₄-p-H)₃8 in 59% isolated yield (Scheme 3). The somewhat lower yield of 8 compared to its ^tBu₃P-analogue 6 is presumably the result of the more facile oxidation of Cy₃P by N₂O. Indeed, monitoring a solution of Cy₃P in C₆D₅Br under an atmosphere of N₂O by ³¹P NMR showed >90% conversion to Cy₃P=O overnight, whereas oxidation of ^tBu₃P with N₂O is more sluggish (∼50% conversion overnight).²³

Scheme 3 FLP formation from Cy₃P and B(C₆F₄-p-H)₃ and subsequent reaction with N₂O to give 8.

Spectroscopic analysis of 8 revealed a single resonance in the ³¹P{¹H} NMR spectrum at 56.4 ppm. Two signals, representative of the ortho and meta fluorines, were observed in the ¹⁹F NMR spectrum at −134.3 and −143.0 ppm as compared to −130.2 and −139.3 in the free borane.²² A sharp singlet in the ¹¹B{¹H} NMR spectrum at 0.8 ppm was observed, typical of a four-coordinate boron center. The isotopically labeled compound 8-¹⁵15N shows the expected coupling patterns. Comparison with the ¹⁵N spectroscopic data for the ^tBu₃P analogue 5-¹⁵15N reveals no significant difference in the ¹J_N–N coupling constant, whereas ¹J_P–N is smaller by 7.2 Hz and ²J_N–P larger by 2.1 Hz for 8-¹⁵15N (Table 1). A crystallographic study unambiguously confirmed the structure of 8 (Fig. 4, pertinent bond lengths and angles in Table 2).

Fig. 4 Molecular structure of 8. Hydrogen atoms have been omitted for clarity.

The analogous zwitterionic species using the considerably weaker Lewis acid tris(para-fluorophenyl)borane was sought. An equimolar mixture of B(C₆H₄-p-F)₃ and Cy₃P in bromobenzene was allowed to stir overnight under an atmosphere of N₂O. Precipitation with pentane yielded a microcrystalline colourless solid. NMR spectroscopy in CD₂Cl₂ showed singlets in the ¹¹B{¹H}, ¹⁹F and ³¹P{¹H} NMR spectra at 26.4, −116.4 and 62.6 ppm respectively. Subsequent recrystallization of the crude product from a layered CH₂Cl₂–pentane solution at −35 °C yielded colorless crystals of Cy₃P=O(B(C₆H₄-p-F)₃), 9. A crystallographic study confirmed the identity of 9 (Fig. 5). Efforts to observe an intermediate N₂O adduct en route to 9 were unsuccessful. The results above are in agreement with the notion that the zwitterionic compounds R₃P(N₂O)BR′₃ are intermediates in Staudinger-type oxidations²⁴ that are kinetically trapped by coordination of the Lewis acidic borane fragment. This kinetic stabilization is only successful when Lewis acid coordination is sufficiently strong (irreversible) and faster than phosphine oxidation. In this sense, these compounds are analogues of the phosphazides R₃P(NNN)R′.²⁵

Fig. 5 POV-ray depiction of the molecular structure of 9. Hydrogens have been omitted for clarity. Selected bond distances (Å) and angles (°): P(1)–O(1) 1.5251(8); B(1)–O(1) 1.5854(15); P(1)–O(1)–B(1) 147.26(8).

Lewis acid exchange reactions

Further investigation of the use of the weakly acidic borane B(C₆H₄-p-F)₃ for the activation of N₂O in conjunction with ^tBu₃P revealed that the borane remains only weakly bound to the ^tBu₃PN₂O fragment allowing for facile exchange of the acidic fragment for a stronger Lewis acid.¹² Following this method, 1 can easily be generated via the reaction of 6 with an equivalent of B(C₆F₅)₃. Similarly, treatment of 6 with an equivalent of trityl tetrakis(pentafluorophenyl)borate in CD₂Cl₂ at room temperature immediately yielded a bright yellow solution (Scheme 4).

Scheme 4 Exchange of the borane fragment B(C₆H₄-p-F)₃ of 6 by trityl cation generating 10.

NMR spectroscopic analysis revealed a new product with a single resonance in the ³¹P{¹H} NMR at 76.3 ppm, considerably downfield from that in 6 (65.4 ppm). NMR spectroscopy indicated the liberation of free B(C₆H₄-p-F)₃, which was easily removed by precipitation of the ionic product and washing with pentane. Synthesis of the ¹⁵N isotopologue 10-¹⁵15N supported that the ^tBu₃P(N₂O) fragment remains intact throughout the exchange process, as evidenced by two doublets of doublets in the ¹⁵N NMR spectrum at 548.4 and 412.3 ppm, respectively, with a N–N coupling constant of 14.7 Hz. Similarly the ³¹P{¹H} NMR spectrum revealed a doublet of doublets with ¹J_P–N and ²J_P–N of 58.2 and 20.5 Hz. These data are similar to those observed for 1-¹⁵15N to 7-¹⁵15N although there is a marked shift for the ¹⁵N NMR signals (Table 1). Crystals of 10 were grown from a layered CH₂Cl₂/cyclohexane solution at room temperature and an X-ray crystallographic study unambiguously confirmed its solid state structure as [^tBu₃P(N₂O)CPh₃][B(C₆F₅)₄] (Fig. 6, pertinent bond distances and angles in Table 3).

Fig. 6 Molecular structure of the cation of 10. Hydrogens and the B(C₆F₅)₄ counter ion have been omitted for clarity.

Table 3 Selected metrical parameters for 10–13^a

	10 (X = C)	11 (X = Zr)	13 (X = Zr)
a Bond lengths in angstroms (Å), angles in degrees (°).
P(1)–N(1)	1.7217(13)	1.719(4)	1.683(3)
N(1)–N(2)	1.2418(17)	1.233(5)	1.266(4)
N(2)–O(1)	1.3752(15)	1.372(5)	1.300(4)
O(1)–X(1)	1.4955(17)	2.105(3)	2.151(3)
P(1)–N(1)–N(2)	115.98(10)	111.3(3)	116.9(3)
N(1)–N(2)O(1)	108.52(12)	111.0(3)	111.9(3)
N(2)–O(1)X(1)	110.68(10)	126.2(2)	123.4(2)

---

Throughout the exchange process the ^tBu₃P(N₂O) fragment remains intact and appears to be inherently stable. It is worth noting that this method of borane exchange also offers an easy and clean route to the generation of 2, 3, 4, 5 and 7via exchange of 6 with one equivalent of PhB(C₆F₅)₂, MesB(C₆F₅)₂, (C₆F₅)₂BOC₆F₅, B(C₆F₄-p-H)₃ or with half an equivalent of 1,4-(C₆F₅)₂B(C₆F₄)B(C₆F₅)_2.

Extending this borane exchange strategy to early transition metal complexes is useful as there are few documented cases where intact N₂O interacts with metal complexes without the decomposition of N₂O to O and N₂.¹⁰ Thus, treatment of a bromobenzene solution of [Cp₂ZrMe][MeB(C₆F₅)₃] with an equivalent of 6 resulted in a beige oil that solidified upon trituration with pentane. Recrystallization from CH₂Cl₂–pentane at −35 °C afforded 11 in 83% isolated yield (Scheme 5). NMR spectroscopic studies revealed a singlet in the ³¹P{¹H} NMR at 67.7 ppm. Three peaks were observed in the ¹⁹F NMR spectrum pertaining to the ortho, para and meta signals of the MeB(C₆F₅)₃ anion at −133.1, −165.3 and −167.8 ppm, with no traces of residual B(C₆H₄-p-F)₃. A broad signal in the ¹H NMR at 0.48 ppm was observed pertaining to the methyl group of the counter anion, MeB(C₆F₅)₃. With these spectroscopic data, [^tBu₃P(N₂O)ZrCp₂Me][MeB(C₆F₅)₃] was proposed as the structure of 11, which was confirmed by a single crystal X-ray diffraction study (Fig. 7, Table 3).

Scheme 5 Synthetic scheme for the generation of 11 and 12 by Lewis acid exchange.

Fig. 7 The molecular structure of the cation of 11. Hydrogens and the counter anion MeB(C₆F₅)₃ have been omitted for clarity.

The crystallographic characterization of 11 constitutes the only example of an intact N₂O fragment where the oxygen of the N₂O fragment is bound directly to a zirconium center.¹⁰ As observed in the ^tBu₃P(N₂O)BR₃ compounds described above, the molecular structure displays the same trans orientation of the ZrO and ^tBu₃P fragments relative to the N–N bond. The metric parameters for the PN₂O fragment in 11 are similar to those observed in the borane compounds described above. The N(2)–O(1)–Zr(1) angle of 126.2(2)° is relatively large in comparison to the corresponding N–O–B or N–O–C angles in 1–10 (110.32(13)–114.86(14)°), which may reflect some degree of interaction between an O lone pair and a vacant metal-based orbital that is absent in the main group species 1–10. Coordination of the ^tBu₃P(N₂O) fragment to the cationic zirconocene center results in little structural changes in the Cp₂ZrMe⁺ moiety. The Cp(centroid)–Zr–Cp(centroid) angle in 11 is 130.5°, which may be compared to 129.6 in [Cp₂ZrMe(THF)][BPh₄]²⁶ or 129.2 in Cp₂ZrCl₂.²⁷ As well, the Zr–Me (2.261(4) Å) and Zr–O (2.105(3) Å) bond lengths are statistically indistinguishable from the corresponding distances in Cp₂ZrMe(THF)⁺ (2.256(10) and 2.122(14) Å, respectively).²⁶ In a similar manner, Lewis acid exchange between [Cp₂TiMe][MeB(C₆F₅)₃] and 6 resulted in the formation of [^tBu₃P(N₂O)TiCp₂Me][MeB(C₆F₅)₃] 12, in 81% yield. These species were found to be stable in solution for a number of days.

It is noteworthy that attempts to generate 11 directly from the reaction of the FLP generated by equimolar mixture of [Cp₂ZrMe][MeB(C₆F₅)₃] and with ¹⁵N₂O gave only inseparable and unidentified reaction products. However, in contrast, an analogous species, generated by the zirconocene methoxide cation [Cp*₂Zr(OMe)][B(C₆F₅)₄] and an equivalent of ^tBu₃P in bromobenzene (Scheme 6) reacted under an atmosphere of N₂O for 12 h, to give a yellow solid, 13, which was isolated in 74% yield (Scheme 6). Spectroscopic analysis revealed a single resonance in the ³¹P{¹H} NMR at 64.4 ppm while the ¹H NMR data showed resonances attributable to the Cp* and methoxide ligands at 1.91 and 3.98 ppm, respectively. The ¹¹B{¹H} and ¹⁹F NMR spectra were in keeping with the corresponding B(C₆F₅)₄ counter anion. In addition, ¹⁵N and ³¹P NMR spectroscopy on the isotopically labeled compound 13-¹⁵15N is indicative of an intact ^tBu₃P(N₂O) fragment (Table 1). The sum of these data firmly supports the formulation of 13 as [^tBu₃P(N₂O)ZrCp*₂(OMe)][B(C₆F₅)₄].

Scheme 6 Activation of N₂O employing a Zr/P FLP yielding 13. Cp* = C₅Me₅.

An X-ray structure determination of 13 (Fig. 8, pertinent bond lengths and angles in Table 3) confirmed that the N₂O fragment is O-bound to the Zr center and N-bound to the P center of the phosphine while the orientation remains trans about the N–N bond. Based on the observation of quite disparate Zr–O bond distances, the OMe and ON₂^tBu₃P moieties interact with the metal center in a rather different manner. The Zr–O(alkoxide) bond length in 13 of 1.916(2) Å is typical for zirconocene alkoxides (e.g., [Cp₂Zr(O^tBu)(THF)][BPh₄]: average 1.890(4) Å)²⁸ but significantly shorter than the Zr–ON₂^tBu₃P distance of 2.151(3) Å. In addition, the large Zr(1)–O(2)–C(21) angle of 169.9(2)° in comparison to Zr(1)–O(1)–N(2) (123.4(2)°) suggests that the methoxide ligand, but not the ON₂^tBu₃P fragment, is engaged in substantial backdonation to the electron deficient metal center.^28,29 A comparison between 11 and 13 shows that the Zr(1)–O(1) bond is shortest in 11, in agreement with it having the most Lewis acidic metal center. The P(1)–N(1) and N(2)–O(1) bond lengths in 13 of 1.683(3) and 1.300(4) Å, respectively, were found to be shorter than those in 11 (1.719(4) and 1.372(5) Å). Conversely, the N(1)–N(2) bond in 13 is somewhat longer (1.266(4) vs. 1.233(5) Å in 11). Unlike the P/B systems where variation of the Lewis acidic fragment has been shown to have very little effect on the nature of the N₂O moiety, variation in the Lewis acidity of the metallocene cation causes a more significant perturbation of the PN₂O fragment.

Fig. 8 The molecular structure of 13. Hydrogens and the counter anion B(C₆F₅)₄ have been omitted for clarity.

Mechanism of borane exchange

In order to obtain mechanistic insight, a kinetic study of Lewis acid exchange reactions was undertaken. Thus, to a solution of ^tBu₃P(N₂O)B(C₆H₄-p-F)₃6 in CD₂Cl₂ at room temperature was added an equimolar amount of a stronger Lewis acid (e.g., B(C₆F₅)₃, B(C₆F₅)₂Ph). The NMR spectra taken immediately after mixing (<5 min) in all cases showed only the presence of exchange products, suggesting that these reactions are fast. It is likely that the weakly Lewis acidic nature of the borane in 6 results in facile exchange. Employing a species containing a more tightly bound Lewis acid, the trityl compound [^tBu₃P(N₂O)CPh₃][B(C₆F₅)₄] 10 was shown to react cleanly with B(C₆F₅)₃ to give [Ph₃C][B(C₆F₅)₄] and ^tBu₃P(N₂O)B(C₆F₅)₃ (1). However, even when CD₂Cl₂ was condensed at −196 °C into an NMR tube containing solid 10 and B(C₆F₅)₃, NMR spectroscopic analysis of the reaction at −80 °C showed full conversion to the expected products before the first spectrum could be measured. It was thus not possible to obtain rate data. As an alternative, the exchange between ^tBu₃P(N₂O)B(C₆F₅)₃ (1) and free B(C₆F₅)₃ was examined by ¹⁹F NMR spectroscopy. Well-separated, sharp resonances for both species are observed in the ¹⁹F NMR spectrum, but 2D ¹⁹F EXSY experiments³⁰ (τ_mix = 1.2 s) showed no evidence of exchange between the two species up to 80 °C. Gratifyingly, exchange was observed between ^tBu₃P(N₂O)B(C₆H₄-p-F)₃ (6) and free B(C₆H₄-p-F)₃ at room temperature and analysis of 2D ¹⁹F EXSY spectra obtained between −25 and +25 °C allowed determination of the activation parameters of the exchange process as ΔH^‡ = 71.2(9) kJ mol⁻¹ and ΔS^‡ = 32(3) J mol⁻¹ K⁻¹ (see Supporting Information for details). The small, positive value for ΔS^‡ suggests that the B–O linkage is somewhat weakened before the incoming borane binds. The lack of observable exchange (i.e., a significantly higher activation barrier) between 1, in which the borane is more strongly bound, and B(C₆F₅)₃ is in agreement with this proposal.³¹

Conclusions

FLPs react with N₂O under mild conditions to give zwitterionic products R₃P(N₂O)BR₂R′ in high yield. Phosphines with diminished steric demands or Lewis basicity, or nitrogen Lewis bases do not give analogous compounds, suggesting that a combination of strong σ-donation and π-acceptance is required to capture the N₂O fragment. A range of borane Lewis acidities have been studied. These show little difference in the geometric or electronic properties of the resulting N₂O moieties. The robustness of the R₃P(N₂O) moiety allows the synthesis of otherwise inaccessible N₂O species by exchange of the borane for stronger Lewis acids such as Ph₃C⁺ and [Cp₂ZrMe]⁺. This route to N₂O transition metal derivatives offers a new strategy to exploring and understanding the nature of N₂O binding, transport and reduction in synthetic as well as biological systems. Further studies of both main group and transition N₂O species are underway and will be reported in due course.

Acknowledgements

DWS gratefully acknowledges the financial support of NSERC of Canada and the award of a Canada Research Chair and a Killam Research Fellowship. EO is grateful for the support of a Rubicon postdoctoral fellowship from the Netherlands Organisation for Scientific Research (NWO). RCN is grateful for the award of an Ontario Graduate Scholarship.

Notes and references

1. A. R. Ravishankara, J. S. Daniel and R. W. Portmann, Science, 2009, 326, 123–125.
2. J. Hansen and M. Sato, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 16109–16114.
3. D. Richardson, H. Felgate, N. Watmough, A. Thomson and E. Baggs, Trends Biotechnol., 2009, 27, 388–397.
4. C. S. Snyder, T. W. Bruulsema, T. L. Jensen and P. E. Fixen, Agric. Ecosyst. Environ., 2009, 133, 247–266.
5. (a) W. G. Zumft and P. M. H. Kroneck, in Advances in Microbial Physiology, ed. K. P. Robert, Academic Press, 2006, pp. 107–227; (b) E. I. Solomon, R. Sarangi, J. S. Woertink, A. J. Augustine, J. Yoon and S. Ghosh, Acc. Chem. Res., 2007, 40, 581–591.
6. W. C. Trogler, J. Chem. Educ., 1995, 72, 973.
7. W. B. Tolman, Angew. Chem., Int. Ed., 2010, 49, 1018–1024.
8. J. N. Armor and H. Taube, J. Am. Chem. Soc., 1969, 91, 6874–6876.
9. (a) F. Bottomley and J. R. Crawford, J. Am. Chem. Soc., 1972, 94, 9092–9095; (b) F. Bottomley and W. V. F. Brooks, Inorg. Chem., 1977, 16, 501–502; (c) D. F. T. Tuan and R. Hoffmann, Inorg. Chem., 1985, 24, 871–876; (d) F. Paulat, T. Kuschel, C. Nather, V. K. K. Praneeth, O. Sander and N. Lehnert, Inorg. Chem., 2004, 43, 6979–6994.
10. G. A. Vaughan, C. D. Sofield, G. L. Hillhouse and A. L. Rheingold, J. Am. Chem. Soc., 1989, 111, 5491–5493.
11. P. Chen, S. I. Gorelsky, S. Ghosh and E. I. Solomon, Angew. Chem., Int. Ed., 2004, 43, 4132–4140.
12. R. C. Neu, E. Otten and D. W. Stephan, Angew. Chem., Int. Ed., 2009, 48, 9709–9712.
13. (a) G. A. Vaughan, P. B. Rupert and G. L. Hillhouse, J. Am. Chem. Soc., 1987, 109, 5538–5539; (b) A. W. Kaplan and R. G. Bergman, Organometallics, 1997, 16, 1106–1108.
14. (a) F. Bottomley, Polyhedron, 1992, 11, 1707–1731; (b) W. A. Howard and G. Parkin, J. Am. Chem. Soc., 1994, 116, 606–615; (c) J. T. Groves and J. S. Roman, J. Am. Chem. Soc., 1995, 117, 5594–5595; (d) H. Z. Yu, G. C. Jia and Z. Lin, Organometallics, 2009, 28, 1158–1164.
15. (a) J.-P. F. Cherry, A. R. Johnson, L. M. Baraldo, Y.-C. Tsai, C. C. Cummins, S. V. Kryatov, E. V. Rybak-Akimova, K. B. Capps, C. D. Hoff, C. M. Haar and S. P. Nolan, J. Am. Chem. Soc., 2001, 123, 7271–7286; (b) A. R. Johnson, W. M. Davis, C. C. Cummins, S. Serron, S. P. Nolan, D. G. Musaev and K. Morokuma, J. Am. Chem. Soc., 1998, 120, 2071–2085; (c) C. E. Laplaza, A. L. Odom, W. M. Davis, C. C. Cummins and J. D. Protasiewicz, J. Am. Chem. Soc., 1995, 117, 4999–5000.
16. J.-H. Lee, M. Pink, J. Tomaszewski, H. Fan and K. G. Caulton, J. Am. Chem. Soc., 2007, 129, 8706–8707.
17. N. D. Harrold, R. Waterman, G. L. Hillhouse and T. R. Cundari, J. Am. Chem. Soc., 2009, 131, 12872–12873.
18. (a) D. W. Stephan, Org. Biomol. Chem., 2008, 6, 1535–1539; (b) D. W. Stephan, Dalton Trans., 2009, 3129–3136; (c) D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2010, 49, 46–76.
19. E. Otten, R. C. Neu and D. W. Stephan, J. Am. Chem. Soc., 2009, 131, 9918–9919.
20. W. E. Piers, in Adv. Organomet. Chem., ed. R. West and A. F. Hill, Academic Press, 2004, pp. 1–76.
21. (a) G. C. Welch, L. Cabrera, P. A. Chase, E. Hollink, J. D. Masuda, P. R. Wei and D. W. Stephan, Dalton Trans., 2007, 3407–3414; (b) G. C. Welch, T. Holtrichter-Roessmann and D. W. Stephan, Inorg. Chem., 2008, 47, 1904–1906.
22. M. Ullrich, A. J. Lough and D. W. Stephan, J. Am. Chem. Soc., 2009, 131, 52–53.
23. S. Poh, R. Hernandez, M. Inagaki and P. G. Jessop, Org. Lett., 1999, 1, 583–586.
24. E. H. H. Staudinger, Helv. Chim. Acta, 1921, 4, 861–886.
25. M. W. P. Bebbington and D. Bourissou, Coord. Chem. Rev., 2009, 253, 1248–1261.
26. R. F. Jordan, C. S. Bajgur, R. Willett and B. Scott, J. Am. Chem. Soc., 1986, 108, 7410–7411.
27. J. Y. Corey, X.-H. Zhu, L. Brammer and N. P. Rath, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 1995, 51, 565–567.
28. S. Collins, B. E. Koene, R. Ramachandran and N. J. Taylor, Organometallics, 1991, 10, 2092–2094.
29. J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc., 1976, 98, 1729–1742.
30. (a) C. L. Perrin and T. J. Dwyer, Chem. Rev., 1990, 90, 935–967; (b) P. Espinet, A. C. Albéniz, J. A. Casares and J. M. Martínez-Ilarduya, Coord. Chem. Rev., 2008, 252, 2180–2208.
31. An alternative explanation for the positive activation entropy could be a change in the solvation (CH₂Cl₂) of the zwitterionic compound 6. We were not able to test this hypothesis due to the poor solubility of 6 in non-polar solvents.

---

Footnote

† Electronic supplementary information (ESI) available: Crystallographic data (CIF format) for 1, 2, 5, 7, 8–11 and 13, synthesis and spectroscopic and analytical data for all new compounds. NMR spectra of the product from reaction of ^tBu₃P and BPh₃ with N₂O, and details for the ¹⁹F EXSY NMR analysis. CCDC reference numbers 768908 and 786852–786857. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c0sc00398k

---